From 80492e7d49bee0a280a84a39075a7857b92836b2 Mon Sep 17 00:00:00 2001 From: Roel Kluin Date: Tue, 21 Apr 2009 16:08:39 +0200 Subject: rpcgss: remove redundant test on unsigned Signed-off-by: Roel Kluin Signed-off-by: J. Bruce Fields --- fs/nfsd/export.c | 13 ++++--------- 1 file changed, 4 insertions(+), 9 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/export.c b/fs/nfsd/export.c index 5839b229cd0e..6eb918153fd4 100644 --- a/fs/nfsd/export.c +++ b/fs/nfsd/export.c @@ -464,16 +464,11 @@ static int secinfo_parse(char **mesg, char *buf, struct svc_export *exp) if (err) return err; /* - * Just a quick sanity check; we could also try to check - * whether this pseudoflavor is supported, but at worst - * an unsupported pseudoflavor on the export would just - * be a pseudoflavor that won't match the flavor of any - * authenticated request. The administrator will - * probably discover the problem when someone fails to - * authenticate. + * XXX: It would be nice to also check whether this + * pseudoflavor is supported, so we can discover the + * problem at export time instead of when a client fails + * to authenticate. */ - if (f->pseudoflavor < 0) - return -EINVAL; err = get_int(mesg, &f->flags); if (err) return err; -- cgit v1.2.3 From a9e61e25f9d2e7e43bf17625f5cb56c9e0a89b17 Mon Sep 17 00:00:00 2001 From: Felix Blyakher Date: Tue, 31 Mar 2009 15:12:56 -0500 Subject: lockd: call locks_release_private to cleanup per-filesystem state For every lock request lockd creates a new file_lock object in nlmsvc_setgrantargs() by copying the passed in file_lock with locks_copy_lock(). A filesystem can attach it's own lock_operations vector to the file_lock. It has to be cleaned up at the end of the file_lock's life. However, lockd doesn't do it today, yet it asserts in nlmclnt_release_lockargs() that the per-filesystem state is clean. This patch fixes it by exporting locks_release_private() and adding it to nlmsvc_freegrantargs(), to be symmetrical to creating a file_lock in nlmsvc_setgrantargs(). Signed-off-by: Felix Blyakher Signed-off-by: J. Bruce Fields --- fs/lockd/svclock.c | 2 ++ fs/locks.c | 3 ++- include/linux/fs.h | 1 + 3 files changed, 5 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/lockd/svclock.c b/fs/lockd/svclock.c index 83ee34203bd7..e577a78d7bac 100644 --- a/fs/lockd/svclock.c +++ b/fs/lockd/svclock.c @@ -326,6 +326,8 @@ static void nlmsvc_freegrantargs(struct nlm_rqst *call) { if (call->a_args.lock.oh.data != call->a_owner) kfree(call->a_args.lock.oh.data); + + locks_release_private(&call->a_args.lock.fl); } /* diff --git a/fs/locks.c b/fs/locks.c index ec3deea29e37..b6440f52178f 100644 --- a/fs/locks.c +++ b/fs/locks.c @@ -151,7 +151,7 @@ static struct file_lock *locks_alloc_lock(void) return kmem_cache_alloc(filelock_cache, GFP_KERNEL); } -static void locks_release_private(struct file_lock *fl) +void locks_release_private(struct file_lock *fl) { if (fl->fl_ops) { if (fl->fl_ops->fl_release_private) @@ -165,6 +165,7 @@ static void locks_release_private(struct file_lock *fl) } } +EXPORT_SYMBOL_GPL(locks_release_private); /* Free a lock which is not in use. */ static void locks_free_lock(struct file_lock *fl) diff --git a/include/linux/fs.h b/include/linux/fs.h index 5bed436f4353..5ba615e8f533 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -1108,6 +1108,7 @@ extern void locks_copy_lock(struct file_lock *, struct file_lock *); extern void __locks_copy_lock(struct file_lock *, const struct file_lock *); extern void locks_remove_posix(struct file *, fl_owner_t); extern void locks_remove_flock(struct file *); +extern void locks_release_private(struct file_lock *); extern void posix_test_lock(struct file *, struct file_lock *); extern int posix_lock_file(struct file *, struct file_lock *, struct file_lock *); extern int posix_lock_file_wait(struct file *, struct file_lock *); -- cgit v1.2.3 From 78155ed75f470710f2aecb3e75e3d97107ba8374 Mon Sep 17 00:00:00 2001 From: Bian Naimeng Date: Wed, 22 Apr 2009 18:25:37 +0800 Subject: nfsd4: distinguish expired from stale stateids If we encode the time of client creation into the stateid instead of the time of server boot, then we can determine whether that stateid is from a previous instance of the a server, or from a client that has expired, and return an appropriate error to the client. Signed-off-by: Bian Naimeng Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 62 ++++++++++++++++++++++++++++++++++++++++++----------- 1 file changed, 49 insertions(+), 13 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index c65a27b76a9d..74e822ec34cb 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -206,7 +206,7 @@ alloc_init_deleg(struct nfs4_client *clp, struct nfs4_stateid *stp, struct svc_f dp->dl_recall.cbr_dp = NULL; dp->dl_recall.cbr_ident = cb->cb_ident; dp->dl_recall.cbr_trunc = 0; - dp->dl_stateid.si_boot = boot_time; + dp->dl_stateid.si_boot = get_seconds(); dp->dl_stateid.si_stateownerid = current_delegid++; dp->dl_stateid.si_fileid = 0; dp->dl_stateid.si_generation = 0; @@ -1883,7 +1883,7 @@ init_stateid(struct nfs4_stateid *stp, struct nfs4_file *fp, struct nfsd4_open * stp->st_stateowner = sop; get_nfs4_file(fp); stp->st_file = fp; - stp->st_stateid.si_boot = boot_time; + stp->st_stateid.si_boot = get_seconds(); stp->st_stateid.si_stateownerid = sop->so_id; stp->st_stateid.si_fileid = fp->fi_id; stp->st_stateid.si_generation = 0; @@ -2739,12 +2739,42 @@ nfs4_check_fh(struct svc_fh *fhp, struct nfs4_stateid *stp) static int STALE_STATEID(stateid_t *stateid) { - if (stateid->si_boot == boot_time) - return 0; - dprintk("NFSD: stale stateid (%08x/%08x/%08x/%08x)!\n", - stateid->si_boot, stateid->si_stateownerid, stateid->si_fileid, - stateid->si_generation); - return 1; + if (time_after((unsigned long)boot_time, + (unsigned long)stateid->si_boot)) { + dprintk("NFSD: stale stateid (%08x/%08x/%08x/%08x)!\n", + stateid->si_boot, stateid->si_stateownerid, + stateid->si_fileid, stateid->si_generation); + return 1; + } + return 0; +} + +static int +EXPIRED_STATEID(stateid_t *stateid) +{ + if (time_before((unsigned long)boot_time, + ((unsigned long)stateid->si_boot)) && + time_before((stateid->si_boot + lease_time), get_seconds())) { + dprintk("NFSD: expired stateid (%08x/%08x/%08x/%08x)!\n", + stateid->si_boot, stateid->si_stateownerid, + stateid->si_fileid, stateid->si_generation); + return 1; + } + return 0; +} + +static __be32 +stateid_error_map(stateid_t *stateid) +{ + if (STALE_STATEID(stateid)) + return nfserr_stale_stateid; + if (EXPIRED_STATEID(stateid)) + return nfserr_expired; + + dprintk("NFSD: bad stateid (%08x/%08x/%08x/%08x)!\n", + stateid->si_boot, stateid->si_stateownerid, + stateid->si_fileid, stateid->si_generation); + return nfserr_bad_stateid; } static inline int @@ -2868,8 +2898,10 @@ nfs4_preprocess_stateid_op(struct nfsd4_compound_state *cstate, status = nfserr_bad_stateid; if (is_delegation_stateid(stateid)) { dp = find_delegation_stateid(ino, stateid); - if (!dp) + if (!dp) { + status = stateid_error_map(stateid); goto out; + } status = check_stateid_generation(stateid, &dp->dl_stateid, flags); if (status) @@ -2882,8 +2914,10 @@ nfs4_preprocess_stateid_op(struct nfsd4_compound_state *cstate, *filpp = dp->dl_vfs_file; } else { /* open or lock stateid */ stp = find_stateid(stateid, flags); - if (!stp) + if (!stp) { + status = stateid_error_map(stateid); goto out; + } if (nfs4_check_fh(current_fh, stp)) goto out; if (!stp->st_stateowner->so_confirmed) @@ -2957,7 +2991,7 @@ nfs4_preprocess_seqid_op(struct nfsd4_compound_state *cstate, u32 seqid, */ sop = search_close_lru(stateid->si_stateownerid, flags); if (sop == NULL) - return nfserr_bad_stateid; + return stateid_error_map(stateid); *sopp = sop; goto check_replay; } @@ -3228,8 +3262,10 @@ nfsd4_delegreturn(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, if (!is_delegation_stateid(stateid)) goto out; dp = find_delegation_stateid(inode, stateid); - if (!dp) + if (!dp) { + status = stateid_error_map(stateid); goto out; + } status = check_stateid_generation(stateid, &dp->dl_stateid, flags); if (status) goto out; @@ -3456,7 +3492,7 @@ alloc_init_lock_stateid(struct nfs4_stateowner *sop, struct nfs4_file *fp, struc stp->st_stateowner = sop; get_nfs4_file(fp); stp->st_file = fp; - stp->st_stateid.si_boot = boot_time; + stp->st_stateid.si_boot = get_seconds(); stp->st_stateid.si_stateownerid = sop->so_id; stp->st_stateid.si_fileid = fp->fi_id; stp->st_stateid.si_generation = 0; -- cgit v1.2.3 From 879c5e6b7cb4c689d08ca9b2e353d8ab3dc425d5 Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Wed, 17 Jun 2009 11:47:48 -0400 Subject: jbd2: convert instrumentation from markers to tracepoints Signed-off-by: "Theodore Ts'o" --- fs/jbd2/checkpoint.c | 5 +- fs/jbd2/commit.c | 13 ++-- fs/jbd2/journal.c | 69 ++++++++++++++++++ include/linux/jbd2.h | 6 ++ include/trace/events/jbd2.h | 168 ++++++++++++++++++++++++++++++++++++++++++++ 5 files changed, 252 insertions(+), 9 deletions(-) create mode 100644 include/trace/events/jbd2.h (limited to 'fs') diff --git a/fs/jbd2/checkpoint.c b/fs/jbd2/checkpoint.c index 17159cacbd9e..5d70b3e6d49b 100644 --- a/fs/jbd2/checkpoint.c +++ b/fs/jbd2/checkpoint.c @@ -20,9 +20,9 @@ #include #include #include -#include #include #include +#include /* * Unlink a buffer from a transaction checkpoint list. @@ -358,8 +358,7 @@ int jbd2_log_do_checkpoint(journal_t *journal) * journal straight away. */ result = jbd2_cleanup_journal_tail(journal); - trace_mark(jbd2_checkpoint, "dev %s need_checkpoint %d", - journal->j_devname, result); + trace_jbd2_checkpoint(journal, result); jbd_debug(1, "cleanup_journal_tail returned %d\n", result); if (result <= 0) return result; diff --git a/fs/jbd2/commit.c b/fs/jbd2/commit.c index 0b7d3b8226fd..7b4088b2364d 100644 --- a/fs/jbd2/commit.c +++ b/fs/jbd2/commit.c @@ -16,7 +16,6 @@ #include #include #include -#include #include #include #include @@ -26,6 +25,7 @@ #include #include #include +#include /* * Default IO end handler for temporary BJ_IO buffer_heads. @@ -253,6 +253,7 @@ static int journal_submit_data_buffers(journal_t *journal, * block allocation with delalloc. We need to write * only allocated blocks here. */ + trace_jbd2_submit_inode_data(jinode->i_vfs_inode); err = journal_submit_inode_data_buffers(mapping); if (!ret) ret = err; @@ -394,8 +395,7 @@ void jbd2_journal_commit_transaction(journal_t *journal) commit_transaction = journal->j_running_transaction; J_ASSERT(commit_transaction->t_state == T_RUNNING); - trace_mark(jbd2_start_commit, "dev %s transaction %d", - journal->j_devname, commit_transaction->t_tid); + trace_jbd2_start_commit(journal, commit_transaction); jbd_debug(1, "JBD: starting commit of transaction %d\n", commit_transaction->t_tid); @@ -409,6 +409,7 @@ void jbd2_journal_commit_transaction(journal_t *journal) */ if (commit_transaction->t_synchronous_commit) write_op = WRITE_SYNC_PLUG; + trace_jbd2_commit_locking(journal, commit_transaction); stats.u.run.rs_wait = commit_transaction->t_max_wait; stats.u.run.rs_locked = jiffies; stats.u.run.rs_running = jbd2_time_diff(commit_transaction->t_start, @@ -484,6 +485,7 @@ void jbd2_journal_commit_transaction(journal_t *journal) */ jbd2_journal_switch_revoke_table(journal); + trace_jbd2_commit_flushing(journal, commit_transaction); stats.u.run.rs_flushing = jiffies; stats.u.run.rs_locked = jbd2_time_diff(stats.u.run.rs_locked, stats.u.run.rs_flushing); @@ -520,6 +522,7 @@ void jbd2_journal_commit_transaction(journal_t *journal) commit_transaction->t_state = T_COMMIT; spin_unlock(&journal->j_state_lock); + trace_jbd2_commit_logging(journal, commit_transaction); stats.u.run.rs_logging = jiffies; stats.u.run.rs_flushing = jbd2_time_diff(stats.u.run.rs_flushing, stats.u.run.rs_logging); @@ -1054,9 +1057,7 @@ restart_loop: if (journal->j_commit_callback) journal->j_commit_callback(journal, commit_transaction); - trace_mark(jbd2_end_commit, "dev %s transaction %d head %d", - journal->j_devname, commit_transaction->t_tid, - journal->j_tail_sequence); + trace_jbd2_end_commit(journal, commit_transaction); jbd_debug(1, "JBD: commit %d complete, head %d\n", journal->j_commit_sequence, journal->j_tail_sequence); if (to_free) diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c index 62be7d294ec2..18bfd5dab642 100644 --- a/fs/jbd2/journal.c +++ b/fs/jbd2/journal.c @@ -38,6 +38,10 @@ #include #include #include +#include + +#define CREATE_TRACE_POINTS +#include #include #include @@ -2377,6 +2381,71 @@ static void __exit journal_exit(void) jbd2_journal_destroy_caches(); } +/* + * jbd2_dev_to_name is a utility function used by the jbd2 and ext4 + * tracing infrastructure to map a dev_t to a device name. + * + * The caller should use rcu_read_lock() in order to make sure the + * device name stays valid until its done with it. We use + * rcu_read_lock() as well to make sure we're safe in case the caller + * gets sloppy, and because rcu_read_lock() is cheap and can be safely + * nested. + */ +struct devname_cache { + struct rcu_head rcu; + dev_t device; + char devname[BDEVNAME_SIZE]; +}; +#define CACHE_SIZE_BITS 6 +static struct devname_cache *devcache[1 << CACHE_SIZE_BITS]; +static DEFINE_SPINLOCK(devname_cache_lock); + +static void free_devcache(struct rcu_head *rcu) +{ + kfree(rcu); +} + +const char *jbd2_dev_to_name(dev_t device) +{ + int i = hash_32(device, CACHE_SIZE_BITS); + char *ret; + struct block_device *bd; + + rcu_read_lock(); + if (devcache[i] && devcache[i]->device == device) { + ret = devcache[i]->devname; + rcu_read_unlock(); + return ret; + } + rcu_read_unlock(); + + spin_lock(&devname_cache_lock); + if (devcache[i]) { + if (devcache[i]->device == device) { + ret = devcache[i]->devname; + spin_unlock(&devname_cache_lock); + return ret; + } + call_rcu(&devcache[i]->rcu, free_devcache); + } + devcache[i] = kmalloc(sizeof(struct devname_cache), GFP_KERNEL); + if (!devcache[i]) { + spin_unlock(&devname_cache_lock); + return "NODEV-ALLOCFAILURE"; /* Something non-NULL */ + } + devcache[i]->device = device; + bd = bdget(device); + if (bd) { + bdevname(bd, devcache[i]->devname); + bdput(bd); + } else + __bdevname(device, devcache[i]->devname); + ret = devcache[i]->devname; + spin_unlock(&devname_cache_lock); + return ret; +} +EXPORT_SYMBOL(jbd2_dev_to_name); + MODULE_LICENSE("GPL"); module_init(journal_init); module_exit(journal_exit); diff --git a/include/linux/jbd2.h b/include/linux/jbd2.h index cc02393bfce8..d97eb652d6ca 100644 --- a/include/linux/jbd2.h +++ b/include/linux/jbd2.h @@ -1315,6 +1315,12 @@ extern int jbd_blocks_per_page(struct inode *inode); #define BUFFER_TRACE2(bh, bh2, info) do {} while (0) #define JBUFFER_TRACE(jh, info) do {} while (0) +/* + * jbd2_dev_to_name is a utility function used by the jbd2 and ext4 + * tracing infrastructure to map a dev_t to a device name. + */ +extern const char *jbd2_dev_to_name(dev_t device); + #endif /* __KERNEL__ */ #endif /* _LINUX_JBD2_H */ diff --git a/include/trace/events/jbd2.h b/include/trace/events/jbd2.h new file mode 100644 index 000000000000..845b0b4b48fd --- /dev/null +++ b/include/trace/events/jbd2.h @@ -0,0 +1,168 @@ +#if !defined(_TRACE_JBD2_H) || defined(TRACE_HEADER_MULTI_READ) +#define _TRACE_JBD2_H + +#include +#include + +#undef TRACE_SYSTEM +#define TRACE_SYSTEM jbd2 + +TRACE_EVENT(jbd2_checkpoint, + + TP_PROTO(journal_t *journal, int result), + + TP_ARGS(journal, result), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( int, result ) + ), + + TP_fast_assign( + __entry->dev = journal->j_fs_dev->bd_dev; + __entry->result = result; + ), + + TP_printk("dev %s result %d", + jbd2_dev_to_name(__entry->dev), __entry->result) +); + +TRACE_EVENT(jbd2_start_commit, + + TP_PROTO(journal_t *journal, transaction_t *commit_transaction), + + TP_ARGS(journal, commit_transaction), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( char, sync_commit ) + __field( int, transaction ) + ), + + TP_fast_assign( + __entry->dev = journal->j_fs_dev->bd_dev; + __entry->sync_commit = commit_transaction->t_synchronous_commit; + __entry->transaction = commit_transaction->t_tid; + ), + + TP_printk("dev %s transaction %d sync %d", + jbd2_dev_to_name(__entry->dev), __entry->transaction, + __entry->sync_commit) +); + +TRACE_EVENT(jbd2_commit_locking, + + TP_PROTO(journal_t *journal, transaction_t *commit_transaction), + + TP_ARGS(journal, commit_transaction), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( char, sync_commit ) + __field( int, transaction ) + ), + + TP_fast_assign( + __entry->dev = journal->j_fs_dev->bd_dev; + __entry->sync_commit = commit_transaction->t_synchronous_commit; + __entry->transaction = commit_transaction->t_tid; + ), + + TP_printk("dev %s transaction %d sync %d", + jbd2_dev_to_name(__entry->dev), __entry->transaction, + __entry->sync_commit) +); + +TRACE_EVENT(jbd2_commit_flushing, + + TP_PROTO(journal_t *journal, transaction_t *commit_transaction), + + TP_ARGS(journal, commit_transaction), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( char, sync_commit ) + __field( int, transaction ) + ), + + TP_fast_assign( + __entry->dev = journal->j_fs_dev->bd_dev; + __entry->sync_commit = commit_transaction->t_synchronous_commit; + __entry->transaction = commit_transaction->t_tid; + ), + + TP_printk("dev %s transaction %d sync %d", + jbd2_dev_to_name(__entry->dev), __entry->transaction, + __entry->sync_commit) +); + +TRACE_EVENT(jbd2_commit_logging, + + TP_PROTO(journal_t *journal, transaction_t *commit_transaction), + + TP_ARGS(journal, commit_transaction), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( char, sync_commit ) + __field( int, transaction ) + ), + + TP_fast_assign( + __entry->dev = journal->j_fs_dev->bd_dev; + __entry->sync_commit = commit_transaction->t_synchronous_commit; + __entry->transaction = commit_transaction->t_tid; + ), + + TP_printk("dev %s transaction %d sync %d", + jbd2_dev_to_name(__entry->dev), __entry->transaction, + __entry->sync_commit) +); + +TRACE_EVENT(jbd2_end_commit, + TP_PROTO(journal_t *journal, transaction_t *commit_transaction), + + TP_ARGS(journal, commit_transaction), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( char, sync_commit ) + __field( int, transaction ) + __field( int, head ) + ), + + TP_fast_assign( + __entry->dev = journal->j_fs_dev->bd_dev; + __entry->sync_commit = commit_transaction->t_synchronous_commit; + __entry->transaction = commit_transaction->t_tid; + __entry->head = journal->j_tail_sequence; + ), + + TP_printk("dev %s transaction %d sync %d head %d", + jbd2_dev_to_name(__entry->dev), __entry->transaction, + __entry->sync_commit, __entry->head) +); + +TRACE_EVENT(jbd2_submit_inode_data, + TP_PROTO(struct inode *inode), + + TP_ARGS(inode), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( ino_t, ino ) + ), + + TP_fast_assign( + __entry->dev = inode->i_sb->s_dev; + __entry->ino = inode->i_ino; + ), + + TP_printk("dev %s ino %lu", + jbd2_dev_to_name(__entry->dev), __entry->ino) +); + +#endif /* _TRACE_JBD2_H */ + +/* This part must be outside protection */ +#include -- cgit v1.2.3 From 9bffad1ed2a003a355ed1b42424a0ae3575275ed Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Wed, 17 Jun 2009 11:48:11 -0400 Subject: ext4: convert instrumentation from markers to tracepoints Signed-off-by: "Theodore Ts'o" --- fs/ext4/fsync.c | 8 +- fs/ext4/ialloc.c | 15 +- fs/ext4/inode.c | 69 +---- fs/ext4/mballoc.c | 77 ++--- fs/ext4/mballoc.h | 1 - fs/ext4/super.c | 6 +- include/trace/events/ext4.h | 719 ++++++++++++++++++++++++++++++++++++++++++++ 7 files changed, 774 insertions(+), 121 deletions(-) create mode 100644 include/trace/events/ext4.h (limited to 'fs') diff --git a/fs/ext4/fsync.c b/fs/ext4/fsync.c index 5afe4370840b..83cf6415f599 100644 --- a/fs/ext4/fsync.c +++ b/fs/ext4/fsync.c @@ -28,10 +28,12 @@ #include #include #include -#include + #include "ext4.h" #include "ext4_jbd2.h" +#include + /* * akpm: A new design for ext4_sync_file(). * @@ -52,9 +54,7 @@ int ext4_sync_file(struct file *file, struct dentry *dentry, int datasync) J_ASSERT(ext4_journal_current_handle() == NULL); - trace_mark(ext4_sync_file, "dev %s datasync %d ino %ld parent %ld", - inode->i_sb->s_id, datasync, inode->i_ino, - dentry->d_parent->d_inode->i_ino); + trace_ext4_sync_file(file, dentry, datasync); /* * data=writeback: diff --git a/fs/ext4/ialloc.c b/fs/ext4/ialloc.c index 3743bd849bce..7d502f3be914 100644 --- a/fs/ext4/ialloc.c +++ b/fs/ext4/ialloc.c @@ -23,11 +23,14 @@ #include #include #include + #include "ext4.h" #include "ext4_jbd2.h" #include "xattr.h" #include "acl.h" +#include + /* * ialloc.c contains the inodes allocation and deallocation routines */ @@ -208,11 +211,7 @@ void ext4_free_inode(handle_t *handle, struct inode *inode) ino = inode->i_ino; ext4_debug("freeing inode %lu\n", ino); - trace_mark(ext4_free_inode, - "dev %s ino %lu mode %d uid %lu gid %lu bocks %llu", - sb->s_id, inode->i_ino, inode->i_mode, - (unsigned long) inode->i_uid, (unsigned long) inode->i_gid, - (unsigned long long) inode->i_blocks); + trace_ext4_free_inode(inode); /* * Note: we must free any quota before locking the superblock, @@ -815,8 +814,7 @@ struct inode *ext4_new_inode(handle_t *handle, struct inode *dir, int mode) sb = dir->i_sb; ngroups = ext4_get_groups_count(sb); - trace_mark(ext4_request_inode, "dev %s dir %lu mode %d", sb->s_id, - dir->i_ino, mode); + trace_ext4_request_inode(dir, mode); inode = new_inode(sb); if (!inode) return ERR_PTR(-ENOMEM); @@ -1047,8 +1045,7 @@ got: } ext4_debug("allocating inode %lu\n", inode->i_ino); - trace_mark(ext4_allocate_inode, "dev %s ino %lu dir %lu mode %d", - sb->s_id, inode->i_ino, dir->i_ino, mode); + trace_ext4_allocate_inode(inode, dir, mode); goto really_out; fail: ext4_std_error(sb, err); diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 875db944b22f..2418ad36eab5 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -37,11 +37,14 @@ #include #include #include + #include "ext4_jbd2.h" #include "xattr.h" #include "acl.h" #include "ext4_extents.h" +#include + #define MPAGE_DA_EXTENT_TAIL 0x01 static inline int ext4_begin_ordered_truncate(struct inode *inode, @@ -1466,10 +1469,7 @@ static int ext4_write_begin(struct file *file, struct address_space *mapping, pgoff_t index; unsigned from, to; - trace_mark(ext4_write_begin, - "dev %s ino %lu pos %llu len %u flags %u", - inode->i_sb->s_id, inode->i_ino, - (unsigned long long) pos, len, flags); + trace_ext4_write_begin(inode, pos, len, flags); /* * Reserve one block more for addition to orphan list in case * we allocate blocks but write fails for some reason @@ -1611,10 +1611,7 @@ static int ext4_ordered_write_end(struct file *file, struct inode *inode = mapping->host; int ret = 0, ret2; - trace_mark(ext4_ordered_write_end, - "dev %s ino %lu pos %llu len %u copied %u", - inode->i_sb->s_id, inode->i_ino, - (unsigned long long) pos, len, copied); + trace_ext4_ordered_write_end(inode, pos, len, copied); ret = ext4_jbd2_file_inode(handle, inode); if (ret == 0) { @@ -1658,10 +1655,7 @@ static int ext4_writeback_write_end(struct file *file, struct inode *inode = mapping->host; int ret = 0, ret2; - trace_mark(ext4_writeback_write_end, - "dev %s ino %lu pos %llu len %u copied %u", - inode->i_sb->s_id, inode->i_ino, - (unsigned long long) pos, len, copied); + trace_ext4_writeback_write_end(inode, pos, len, copied); ret2 = ext4_generic_write_end(file, mapping, pos, len, copied, page, fsdata); copied = ret2; @@ -1705,10 +1699,7 @@ static int ext4_journalled_write_end(struct file *file, unsigned from, to; loff_t new_i_size; - trace_mark(ext4_journalled_write_end, - "dev %s ino %lu pos %llu len %u copied %u", - inode->i_sb->s_id, inode->i_ino, - (unsigned long long) pos, len, copied); + trace_ext4_journalled_write_end(inode, pos, len, copied); from = pos & (PAGE_CACHE_SIZE - 1); to = from + len; @@ -2554,9 +2545,7 @@ static int ext4_da_writepage(struct page *page, struct buffer_head *page_bufs; struct inode *inode = page->mapping->host; - trace_mark(ext4_da_writepage, - "dev %s ino %lu page_index %lu", - inode->i_sb->s_id, inode->i_ino, page->index); + trace_ext4_da_writepage(inode, page); size = i_size_read(inode); if (page->index == size >> PAGE_CACHE_SHIFT) len = size & ~PAGE_CACHE_MASK; @@ -2667,19 +2656,7 @@ static int ext4_da_writepages(struct address_space *mapping, int needed_blocks, ret = 0, nr_to_writebump = 0; struct ext4_sb_info *sbi = EXT4_SB(mapping->host->i_sb); - trace_mark(ext4_da_writepages, - "dev %s ino %lu nr_t_write %ld " - "pages_skipped %ld range_start %llu " - "range_end %llu nonblocking %d " - "for_kupdate %d for_reclaim %d " - "for_writepages %d range_cyclic %d", - inode->i_sb->s_id, inode->i_ino, - wbc->nr_to_write, wbc->pages_skipped, - (unsigned long long) wbc->range_start, - (unsigned long long) wbc->range_end, - wbc->nonblocking, wbc->for_kupdate, - wbc->for_reclaim, wbc->for_writepages, - wbc->range_cyclic); + trace_ext4_da_writepages(inode, wbc); /* * No pages to write? This is mainly a kludge to avoid starting @@ -2845,14 +2822,7 @@ out_writepages: if (!no_nrwrite_index_update) wbc->no_nrwrite_index_update = 0; wbc->nr_to_write -= nr_to_writebump; - trace_mark(ext4_da_writepage_result, - "dev %s ino %lu ret %d pages_written %d " - "pages_skipped %ld congestion %d " - "more_io %d no_nrwrite_index_update %d", - inode->i_sb->s_id, inode->i_ino, ret, - pages_written, wbc->pages_skipped, - wbc->encountered_congestion, wbc->more_io, - wbc->no_nrwrite_index_update); + trace_ext4_da_writepages_result(inode, wbc, ret, pages_written); return ret; } @@ -2904,11 +2874,7 @@ static int ext4_da_write_begin(struct file *file, struct address_space *mapping, len, flags, pagep, fsdata); } *fsdata = (void *)0; - - trace_mark(ext4_da_write_begin, - "dev %s ino %lu pos %llu len %u flags %u", - inode->i_sb->s_id, inode->i_ino, - (unsigned long long) pos, len, flags); + trace_ext4_da_write_begin(inode, pos, len, flags); retry: /* * With delayed allocation, we don't log the i_disksize update @@ -3001,10 +2967,7 @@ static int ext4_da_write_end(struct file *file, } } - trace_mark(ext4_da_write_end, - "dev %s ino %lu pos %llu len %u copied %u", - inode->i_sb->s_id, inode->i_ino, - (unsigned long long) pos, len, copied); + trace_ext4_da_write_end(inode, pos, len, copied); start = pos & (PAGE_CACHE_SIZE - 1); end = start + copied - 1; @@ -3255,9 +3218,7 @@ static int ext4_normal_writepage(struct page *page, loff_t size = i_size_read(inode); loff_t len; - trace_mark(ext4_normal_writepage, - "dev %s ino %lu page_index %lu", - inode->i_sb->s_id, inode->i_ino, page->index); + trace_ext4_normal_writepage(inode, page); J_ASSERT(PageLocked(page)); if (page->index == size >> PAGE_CACHE_SHIFT) len = size & ~PAGE_CACHE_MASK; @@ -3343,9 +3304,7 @@ static int ext4_journalled_writepage(struct page *page, loff_t size = i_size_read(inode); loff_t len; - trace_mark(ext4_journalled_writepage, - "dev %s ino %lu page_index %lu", - inode->i_sb->s_id, inode->i_ino, page->index); + trace_ext4_journalled_writepage(inode, page); J_ASSERT(PageLocked(page)); if (page->index == size >> PAGE_CACHE_SHIFT) len = size & ~PAGE_CACHE_MASK; diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index ed8482e22c0e..8d98070b48fb 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -22,6 +22,8 @@ */ #include "mballoc.h" +#include + /* * MUSTDO: * - test ext4_ext_search_left() and ext4_ext_search_right() @@ -340,8 +342,6 @@ static void ext4_mb_generate_from_freelist(struct super_block *sb, void *bitmap, ext4_group_t group); static void release_blocks_on_commit(journal_t *journal, transaction_t *txn); - - static inline void *mb_correct_addr_and_bit(int *bit, void *addr) { #if BITS_PER_LONG == 64 @@ -2859,9 +2859,8 @@ static void release_blocks_on_commit(journal_t *journal, transaction_t *txn) discard_block = (ext4_fsblk_t) entry->group * EXT4_BLOCKS_PER_GROUP(sb) + entry->start_blk + le32_to_cpu(EXT4_SB(sb)->s_es->s_first_data_block); - trace_mark(ext4_discard_blocks, "dev %s blk %llu count %u", - sb->s_id, (unsigned long long) discard_block, - entry->count); + trace_ext4_discard_blocks(sb, (unsigned long long)discard_block, + entry->count); sb_issue_discard(sb, discard_block, entry->count); kmem_cache_free(ext4_free_ext_cachep, entry); @@ -3629,10 +3628,7 @@ ext4_mb_new_inode_pa(struct ext4_allocation_context *ac) mb_debug("new inode pa %p: %llu/%u for %u\n", pa, pa->pa_pstart, pa->pa_len, pa->pa_lstart); - trace_mark(ext4_mb_new_inode_pa, - "dev %s ino %lu pstart %llu len %u lstart %u", - sb->s_id, ac->ac_inode->i_ino, - pa->pa_pstart, pa->pa_len, pa->pa_lstart); + trace_ext4_mb_new_inode_pa(ac, pa); ext4_mb_use_inode_pa(ac, pa); atomic_add(pa->pa_free, &EXT4_SB(sb)->s_mb_preallocated); @@ -3691,9 +3687,8 @@ ext4_mb_new_group_pa(struct ext4_allocation_context *ac) pa->pa_type = MB_GROUP_PA; mb_debug("new group pa %p: %llu/%u for %u\n", pa, - pa->pa_pstart, pa->pa_len, pa->pa_lstart); - trace_mark(ext4_mb_new_group_pa, "dev %s pstart %llu len %u lstart %u", - sb->s_id, pa->pa_pstart, pa->pa_len, pa->pa_lstart); + pa->pa_pstart, pa->pa_len, pa->pa_lstart); + trace_ext4_mb_new_group_pa(ac, pa); ext4_mb_use_group_pa(ac, pa); atomic_add(pa->pa_free, &EXT4_SB(sb)->s_mb_preallocated); @@ -3783,10 +3778,8 @@ ext4_mb_release_inode_pa(struct ext4_buddy *e4b, struct buffer_head *bitmap_bh, ext4_mb_store_history(ac); } - trace_mark(ext4_mb_release_inode_pa, - "dev %s ino %lu block %llu count %u", - sb->s_id, pa->pa_inode->i_ino, grp_blk_start + bit, - next - bit); + trace_ext4_mb_release_inode_pa(ac, pa, grp_blk_start + bit, + next - bit); mb_free_blocks(pa->pa_inode, e4b, bit, next - bit); bit = next + 1; } @@ -3820,8 +3813,7 @@ ext4_mb_release_group_pa(struct ext4_buddy *e4b, if (ac) ac->ac_op = EXT4_MB_HISTORY_DISCARD; - trace_mark(ext4_mb_release_group_pa, "dev %s pstart %llu len %d", - sb->s_id, pa->pa_pstart, pa->pa_len); + trace_ext4_mb_release_group_pa(ac, pa); BUG_ON(pa->pa_deleted == 0); ext4_get_group_no_and_offset(sb, pa->pa_pstart, &group, &bit); BUG_ON(group != e4b->bd_group && pa->pa_len != 0); @@ -3889,6 +3881,8 @@ ext4_mb_discard_group_preallocations(struct super_block *sb, INIT_LIST_HEAD(&list); ac = kmem_cache_alloc(ext4_ac_cachep, GFP_NOFS); + if (ac) + ac->ac_sb = sb; repeat: ext4_lock_group(sb, group); list_for_each_entry_safe(pa, tmp, @@ -3987,12 +3981,15 @@ void ext4_discard_preallocations(struct inode *inode) } mb_debug("discard preallocation for inode %lu\n", inode->i_ino); - trace_mark(ext4_discard_preallocations, "dev %s ino %lu", sb->s_id, - inode->i_ino); + trace_ext4_discard_preallocations(inode); INIT_LIST_HEAD(&list); ac = kmem_cache_alloc(ext4_ac_cachep, GFP_NOFS); + if (ac) { + ac->ac_sb = sb; + ac->ac_inode = inode; + } repeat: /* first, collect all pa's in the inode */ spin_lock(&ei->i_prealloc_lock); @@ -4276,6 +4273,8 @@ ext4_mb_discard_lg_preallocations(struct super_block *sb, INIT_LIST_HEAD(&discard_list); ac = kmem_cache_alloc(ext4_ac_cachep, GFP_NOFS); + if (ac) + ac->ac_sb = sb; spin_lock(&lg->lg_prealloc_lock); list_for_each_entry_rcu(pa, &lg->lg_prealloc_list[order], @@ -4445,8 +4444,7 @@ static int ext4_mb_discard_preallocations(struct super_block *sb, int needed) int ret; int freed = 0; - trace_mark(ext4_mb_discard_preallocations, "dev %s needed %d", - sb->s_id, needed); + trace_ext4_mb_discard_preallocations(sb, needed); for (i = 0; i < ngroups && needed > 0; i++) { ret = ext4_mb_discard_group_preallocations(sb, i, needed); freed += ret; @@ -4475,17 +4473,7 @@ ext4_fsblk_t ext4_mb_new_blocks(handle_t *handle, sb = ar->inode->i_sb; sbi = EXT4_SB(sb); - trace_mark(ext4_request_blocks, "dev %s flags %u len %u ino %lu " - "lblk %llu goal %llu lleft %llu lright %llu " - "pleft %llu pright %llu ", - sb->s_id, ar->flags, ar->len, - ar->inode ? ar->inode->i_ino : 0, - (unsigned long long) ar->logical, - (unsigned long long) ar->goal, - (unsigned long long) ar->lleft, - (unsigned long long) ar->lright, - (unsigned long long) ar->pleft, - (unsigned long long) ar->pright); + trace_ext4_request_blocks(ar); /* * For delayed allocation, we could skip the ENOSPC and @@ -4521,7 +4509,10 @@ ext4_fsblk_t ext4_mb_new_blocks(handle_t *handle, } ac = kmem_cache_alloc(ext4_ac_cachep, GFP_NOFS); - if (!ac) { + if (ac) { + ac->ac_sb = sb; + ac->ac_inode = ar->inode; + } else { ar->len = 0; *errp = -ENOMEM; goto out1; @@ -4594,18 +4585,7 @@ out3: reserv_blks); } - trace_mark(ext4_allocate_blocks, - "dev %s block %llu flags %u len %u ino %lu " - "logical %llu goal %llu lleft %llu lright %llu " - "pleft %llu pright %llu ", - sb->s_id, (unsigned long long) block, - ar->flags, ar->len, ar->inode ? ar->inode->i_ino : 0, - (unsigned long long) ar->logical, - (unsigned long long) ar->goal, - (unsigned long long) ar->lleft, - (unsigned long long) ar->lright, - (unsigned long long) ar->pleft, - (unsigned long long) ar->pright); + trace_ext4_allocate_blocks(ar, (unsigned long long)block); return block; } @@ -4740,10 +4720,7 @@ void ext4_mb_free_blocks(handle_t *handle, struct inode *inode, } ext4_debug("freeing block %lu\n", block); - trace_mark(ext4_free_blocks, - "dev %s block %llu count %lu metadata %d ino %lu", - sb->s_id, (unsigned long long) block, count, metadata, - inode ? inode->i_ino : 0); + trace_ext4_free_blocks(inode, block, count, metadata); ac = kmem_cache_alloc(ext4_ac_cachep, GFP_NOFS); if (ac) { diff --git a/fs/ext4/mballoc.h b/fs/ext4/mballoc.h index 75e34f69215b..c96bb19f58f9 100644 --- a/fs/ext4/mballoc.h +++ b/fs/ext4/mballoc.h @@ -19,7 +19,6 @@ #include #include #include -#include #include #include "ext4_jbd2.h" #include "ext4.h" diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 012c4251397e..e8f0b2af4607 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -37,7 +37,6 @@ #include #include #include -#include #include #include #include @@ -47,6 +46,9 @@ #include "xattr.h" #include "acl.h" +#define CREATE_TRACE_POINTS +#include + static int default_mb_history_length = 1000; module_param_named(default_mb_history_length, default_mb_history_length, @@ -3346,7 +3348,7 @@ static int ext4_sync_fs(struct super_block *sb, int wait) int ret = 0; tid_t target; - trace_mark(ext4_sync_fs, "dev %s wait %d", sb->s_id, wait); + trace_ext4_sync_fs(sb, wait); if (jbd2_journal_start_commit(EXT4_SB(sb)->s_journal, &target)) { if (wait) jbd2_log_wait_commit(EXT4_SB(sb)->s_journal, target); diff --git a/include/trace/events/ext4.h b/include/trace/events/ext4.h new file mode 100644 index 000000000000..acf4cc9cd36d --- /dev/null +++ b/include/trace/events/ext4.h @@ -0,0 +1,719 @@ +#if !defined(_TRACE_EXT4_H) || defined(TRACE_HEADER_MULTI_READ) +#define _TRACE_EXT4_H + +#undef TRACE_SYSTEM +#define TRACE_SYSTEM ext4 + +#include +#include "../../../fs/ext4/ext4.h" +#include "../../../fs/ext4/mballoc.h" +#include + +TRACE_EVENT(ext4_free_inode, + TP_PROTO(struct inode *inode), + + TP_ARGS(inode), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( ino_t, ino ) + __field( umode_t, mode ) + __field( uid_t, uid ) + __field( gid_t, gid ) + __field( blkcnt_t, blocks ) + ), + + TP_fast_assign( + __entry->dev = inode->i_sb->s_dev; + __entry->ino = inode->i_ino; + __entry->mode = inode->i_mode; + __entry->uid = inode->i_uid; + __entry->gid = inode->i_gid; + __entry->blocks = inode->i_blocks; + ), + + TP_printk("dev %s ino %lu mode %d uid %u gid %u blocks %llu", + jbd2_dev_to_name(__entry->dev), __entry->ino, __entry->mode, + __entry->uid, __entry->gid, __entry->blocks) +); + +TRACE_EVENT(ext4_request_inode, + TP_PROTO(struct inode *dir, int mode), + + TP_ARGS(dir, mode), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( ino_t, dir ) + __field( umode_t, mode ) + ), + + TP_fast_assign( + __entry->dev = dir->i_sb->s_dev; + __entry->dir = dir->i_ino; + __entry->mode = mode; + ), + + TP_printk("dev %s dir %lu mode %d", + jbd2_dev_to_name(__entry->dev), __entry->dir, __entry->mode) +); + +TRACE_EVENT(ext4_allocate_inode, + TP_PROTO(struct inode *inode, struct inode *dir, int mode), + + TP_ARGS(inode, dir, mode), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( ino_t, ino ) + __field( ino_t, dir ) + __field( umode_t, mode ) + ), + + TP_fast_assign( + __entry->dev = inode->i_sb->s_dev; + __entry->ino = inode->i_ino; + __entry->dir = dir->i_ino; + __entry->mode = mode; + ), + + TP_printk("dev %s ino %lu dir %lu mode %d", + jbd2_dev_to_name(__entry->dev), __entry->ino, __entry->dir, __entry->mode) +); + +TRACE_EVENT(ext4_write_begin, + + TP_PROTO(struct inode *inode, loff_t pos, unsigned int len, + unsigned int flags), + + TP_ARGS(inode, pos, len, flags), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( ino_t, ino ) + __field( loff_t, pos ) + __field( unsigned int, len ) + __field( unsigned int, flags ) + ), + + TP_fast_assign( + __entry->dev = inode->i_sb->s_dev; + __entry->ino = inode->i_ino; + __entry->pos = pos; + __entry->len = len; + __entry->flags = flags; + ), + + TP_printk("dev %s ino %lu pos %llu len %u flags %u", + jbd2_dev_to_name(__entry->dev), __entry->ino, __entry->pos, __entry->len, + __entry->flags) +); + +TRACE_EVENT(ext4_ordered_write_end, + TP_PROTO(struct inode *inode, loff_t pos, unsigned int len, + unsigned int copied), + + TP_ARGS(inode, pos, len, copied), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( ino_t, ino ) + __field( loff_t, pos ) + __field( unsigned int, len ) + __field( unsigned int, copied ) + ), + + TP_fast_assign( + __entry->dev = inode->i_sb->s_dev; + __entry->ino = inode->i_ino; + __entry->pos = pos; + __entry->len = len; + __entry->copied = copied; + ), + + TP_printk("dev %s ino %lu pos %llu len %u copied %u", + jbd2_dev_to_name(__entry->dev), __entry->ino, __entry->pos, __entry->len, + __entry->copied) +); + +TRACE_EVENT(ext4_writeback_write_end, + TP_PROTO(struct inode *inode, loff_t pos, unsigned int len, + unsigned int copied), + + TP_ARGS(inode, pos, len, copied), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( ino_t, ino ) + __field( loff_t, pos ) + __field( unsigned int, len ) + __field( unsigned int, copied ) + ), + + TP_fast_assign( + __entry->dev = inode->i_sb->s_dev; + __entry->ino = inode->i_ino; + __entry->pos = pos; + __entry->len = len; + __entry->copied = copied; + ), + + TP_printk("dev %s ino %lu pos %llu len %u copied %u", + jbd2_dev_to_name(__entry->dev), __entry->ino, __entry->pos, __entry->len, + __entry->copied) +); + +TRACE_EVENT(ext4_journalled_write_end, + TP_PROTO(struct inode *inode, loff_t pos, unsigned int len, + unsigned int copied), + TP_ARGS(inode, pos, len, copied), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( ino_t, ino ) + __field( loff_t, pos ) + __field( unsigned int, len ) + __field( unsigned int, copied ) + ), + + TP_fast_assign( + __entry->dev = inode->i_sb->s_dev; + __entry->ino = inode->i_ino; + __entry->pos = pos; + __entry->len = len; + __entry->copied = copied; + ), + + TP_printk("dev %s ino %lu pos %llu len %u copied %u", + jbd2_dev_to_name(__entry->dev), __entry->ino, __entry->pos, __entry->len, + __entry->copied) +); + +TRACE_EVENT(ext4_da_writepage, + TP_PROTO(struct inode *inode, struct page *page), + + TP_ARGS(inode, page), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( ino_t, ino ) + __field( pgoff_t, index ) + + ), + + TP_fast_assign( + __entry->dev = inode->i_sb->s_dev; + __entry->ino = inode->i_ino; + __entry->index = page->index; + ), + + TP_printk("dev %s ino %lu page_index %lu", + jbd2_dev_to_name(__entry->dev), __entry->ino, __entry->index) +); + +TRACE_EVENT(ext4_da_writepages, + TP_PROTO(struct inode *inode, struct writeback_control *wbc), + + TP_ARGS(inode, wbc), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( ino_t, ino ) + __field( long, nr_to_write ) + __field( long, pages_skipped ) + __field( loff_t, range_start ) + __field( loff_t, range_end ) + __field( char, nonblocking ) + __field( char, for_kupdate ) + __field( char, for_reclaim ) + __field( char, for_writepages ) + __field( char, range_cyclic ) + ), + + TP_fast_assign( + __entry->dev = inode->i_sb->s_dev; + __entry->ino = inode->i_ino; + __entry->nr_to_write = wbc->nr_to_write; + __entry->pages_skipped = wbc->pages_skipped; + __entry->range_start = wbc->range_start; + __entry->range_end = wbc->range_end; + __entry->nonblocking = wbc->nonblocking; + __entry->for_kupdate = wbc->for_kupdate; + __entry->for_reclaim = wbc->for_reclaim; + __entry->for_writepages = wbc->for_writepages; + __entry->range_cyclic = wbc->range_cyclic; + ), + + TP_printk("dev %s ino %lu nr_t_write %ld pages_skipped %ld range_start %llu range_end %llu nonblocking %d for_kupdate %d for_reclaim %d for_writepages %d range_cyclic %d", + jbd2_dev_to_name(__entry->dev), __entry->ino, __entry->nr_to_write, + __entry->pages_skipped, __entry->range_start, + __entry->range_end, __entry->nonblocking, + __entry->for_kupdate, __entry->for_reclaim, + __entry->for_writepages, __entry->range_cyclic) +); + +TRACE_EVENT(ext4_da_writepages_result, + TP_PROTO(struct inode *inode, struct writeback_control *wbc, + int ret, int pages_written), + + TP_ARGS(inode, wbc, ret, pages_written), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( ino_t, ino ) + __field( int, ret ) + __field( int, pages_written ) + __field( long, pages_skipped ) + __field( char, encountered_congestion ) + __field( char, more_io ) + __field( char, no_nrwrite_index_update ) + ), + + TP_fast_assign( + __entry->dev = inode->i_sb->s_dev; + __entry->ino = inode->i_ino; + __entry->ret = ret; + __entry->pages_written = pages_written; + __entry->pages_skipped = wbc->pages_skipped; + __entry->encountered_congestion = wbc->encountered_congestion; + __entry->more_io = wbc->more_io; + __entry->no_nrwrite_index_update = wbc->no_nrwrite_index_update; + ), + + TP_printk("dev %s ino %lu ret %d pages_written %d pages_skipped %ld congestion %d more_io %d no_nrwrite_index_update %d", + jbd2_dev_to_name(__entry->dev), __entry->ino, __entry->ret, + __entry->pages_written, __entry->pages_skipped, + __entry->encountered_congestion, __entry->more_io, + __entry->no_nrwrite_index_update) +); + +TRACE_EVENT(ext4_da_write_begin, + TP_PROTO(struct inode *inode, loff_t pos, unsigned int len, + unsigned int flags), + + TP_ARGS(inode, pos, len, flags), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( ino_t, ino ) + __field( loff_t, pos ) + __field( unsigned int, len ) + __field( unsigned int, flags ) + ), + + TP_fast_assign( + __entry->dev = inode->i_sb->s_dev; + __entry->ino = inode->i_ino; + __entry->pos = pos; + __entry->len = len; + __entry->flags = flags; + ), + + TP_printk("dev %s ino %lu pos %llu len %u flags %u", + jbd2_dev_to_name(__entry->dev), __entry->ino, __entry->pos, __entry->len, + __entry->flags) +); + +TRACE_EVENT(ext4_da_write_end, + TP_PROTO(struct inode *inode, loff_t pos, unsigned int len, + unsigned int copied), + + TP_ARGS(inode, pos, len, copied), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( ino_t, ino ) + __field( loff_t, pos ) + __field( unsigned int, len ) + __field( unsigned int, copied ) + ), + + TP_fast_assign( + __entry->dev = inode->i_sb->s_dev; + __entry->ino = inode->i_ino; + __entry->pos = pos; + __entry->len = len; + __entry->copied = copied; + ), + + TP_printk("dev %s ino %lu pos %llu len %u copied %u", + jbd2_dev_to_name(__entry->dev), __entry->ino, __entry->pos, __entry->len, + __entry->copied) +); + +TRACE_EVENT(ext4_normal_writepage, + TP_PROTO(struct inode *inode, struct page *page), + + TP_ARGS(inode, page), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( ino_t, ino ) + __field( pgoff_t, index ) + ), + + TP_fast_assign( + __entry->dev = inode->i_sb->s_dev; + __entry->ino = inode->i_ino; + __entry->index = page->index; + ), + + TP_printk("dev %s ino %lu page_index %lu", + jbd2_dev_to_name(__entry->dev), __entry->ino, __entry->index) +); + +TRACE_EVENT(ext4_journalled_writepage, + TP_PROTO(struct inode *inode, struct page *page), + + TP_ARGS(inode, page), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( ino_t, ino ) + __field( pgoff_t, index ) + + ), + + TP_fast_assign( + __entry->dev = inode->i_sb->s_dev; + __entry->ino = inode->i_ino; + __entry->index = page->index; + ), + + TP_printk("dev %s ino %lu page_index %lu", + jbd2_dev_to_name(__entry->dev), __entry->ino, __entry->index) +); + +TRACE_EVENT(ext4_discard_blocks, + TP_PROTO(struct super_block *sb, unsigned long long blk, + unsigned long long count), + + TP_ARGS(sb, blk, count), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( __u64, blk ) + __field( __u64, count ) + + ), + + TP_fast_assign( + __entry->dev = sb->s_dev; + __entry->blk = blk; + __entry->count = count; + ), + + TP_printk("dev %s blk %llu count %llu", + jbd2_dev_to_name(__entry->dev), __entry->blk, __entry->count) +); + +TRACE_EVENT(ext4_mb_new_inode_pa, + TP_PROTO(struct ext4_allocation_context *ac, + struct ext4_prealloc_space *pa), + + TP_ARGS(ac, pa), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( ino_t, ino ) + __field( __u64, pa_pstart ) + __field( __u32, pa_len ) + __field( __u64, pa_lstart ) + + ), + + TP_fast_assign( + __entry->dev = ac->ac_sb->s_dev; + __entry->ino = ac->ac_inode->i_ino; + __entry->pa_pstart = pa->pa_pstart; + __entry->pa_len = pa->pa_len; + __entry->pa_lstart = pa->pa_lstart; + ), + + TP_printk("dev %s ino %lu pstart %llu len %u lstart %llu", + jbd2_dev_to_name(__entry->dev), __entry->ino, __entry->pa_pstart, + __entry->pa_len, __entry->pa_lstart) +); + +TRACE_EVENT(ext4_mb_new_group_pa, + TP_PROTO(struct ext4_allocation_context *ac, + struct ext4_prealloc_space *pa), + + TP_ARGS(ac, pa), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( ino_t, ino ) + __field( __u64, pa_pstart ) + __field( __u32, pa_len ) + __field( __u64, pa_lstart ) + + ), + + TP_fast_assign( + __entry->dev = ac->ac_sb->s_dev; + __entry->ino = ac->ac_inode->i_ino; + __entry->pa_pstart = pa->pa_pstart; + __entry->pa_len = pa->pa_len; + __entry->pa_lstart = pa->pa_lstart; + ), + + TP_printk("dev %s ino %lu pstart %llu len %u lstart %llu", + jbd2_dev_to_name(__entry->dev), __entry->ino, __entry->pa_pstart, + __entry->pa_len, __entry->pa_lstart) +); + +TRACE_EVENT(ext4_mb_release_inode_pa, + TP_PROTO(struct ext4_allocation_context *ac, + struct ext4_prealloc_space *pa, + unsigned long long block, unsigned int count), + + TP_ARGS(ac, pa, block, count), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( ino_t, ino ) + __field( __u64, block ) + __field( __u32, count ) + + ), + + TP_fast_assign( + __entry->dev = ac->ac_sb->s_dev; + __entry->ino = ac->ac_inode->i_ino; + __entry->block = block; + __entry->count = count; + ), + + TP_printk("dev %s ino %lu block %llu count %u", + jbd2_dev_to_name(__entry->dev), __entry->ino, __entry->block, + __entry->count) +); + +TRACE_EVENT(ext4_mb_release_group_pa, + TP_PROTO(struct ext4_allocation_context *ac, + struct ext4_prealloc_space *pa), + + TP_ARGS(ac, pa), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( ino_t, ino ) + __field( __u64, pa_pstart ) + __field( __u32, pa_len ) + + ), + + TP_fast_assign( + __entry->dev = ac->ac_sb->s_dev; + __entry->ino = ac->ac_inode->i_ino; + __entry->pa_pstart = pa->pa_pstart; + __entry->pa_len = pa->pa_len; + ), + + TP_printk("dev %s pstart %llu len %u", + jbd2_dev_to_name(__entry->dev), __entry->pa_pstart, __entry->pa_len) +); + +TRACE_EVENT(ext4_discard_preallocations, + TP_PROTO(struct inode *inode), + + TP_ARGS(inode), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( ino_t, ino ) + + ), + + TP_fast_assign( + __entry->dev = inode->i_sb->s_dev; + __entry->ino = inode->i_ino; + ), + + TP_printk("dev %s ino %lu", + jbd2_dev_to_name(__entry->dev), __entry->ino) +); + +TRACE_EVENT(ext4_mb_discard_preallocations, + TP_PROTO(struct super_block *sb, int needed), + + TP_ARGS(sb, needed), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( int, needed ) + + ), + + TP_fast_assign( + __entry->dev = sb->s_dev; + __entry->needed = needed; + ), + + TP_printk("dev %s needed %d", + jbd2_dev_to_name(__entry->dev), __entry->needed) +); + +TRACE_EVENT(ext4_request_blocks, + TP_PROTO(struct ext4_allocation_request *ar), + + TP_ARGS(ar), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( ino_t, ino ) + __field( unsigned int, flags ) + __field( unsigned int, len ) + __field( __u64, logical ) + __field( __u64, goal ) + __field( __u64, lleft ) + __field( __u64, lright ) + __field( __u64, pleft ) + __field( __u64, pright ) + ), + + TP_fast_assign( + __entry->dev = ar->inode->i_sb->s_dev; + __entry->ino = ar->inode->i_ino; + __entry->flags = ar->flags; + __entry->len = ar->len; + __entry->logical = ar->logical; + __entry->goal = ar->goal; + __entry->lleft = ar->lleft; + __entry->lright = ar->lright; + __entry->pleft = ar->pleft; + __entry->pright = ar->pright; + ), + + TP_printk("dev %s ino %lu flags %u len %u lblk %llu goal %llu lleft %llu lright %llu pleft %llu pright %llu ", + jbd2_dev_to_name(__entry->dev), __entry->ino, __entry->flags, + __entry->len, + (unsigned long long) __entry->logical, + (unsigned long long) __entry->goal, + (unsigned long long) __entry->lleft, + (unsigned long long) __entry->lright, + (unsigned long long) __entry->pleft, + (unsigned long long) __entry->pright) +); + +TRACE_EVENT(ext4_allocate_blocks, + TP_PROTO(struct ext4_allocation_request *ar, unsigned long long block), + + TP_ARGS(ar, block), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( ino_t, ino ) + __field( __u64, block ) + __field( unsigned int, flags ) + __field( unsigned int, len ) + __field( __u64, logical ) + __field( __u64, goal ) + __field( __u64, lleft ) + __field( __u64, lright ) + __field( __u64, pleft ) + __field( __u64, pright ) + ), + + TP_fast_assign( + __entry->dev = ar->inode->i_sb->s_dev; + __entry->ino = ar->inode->i_ino; + __entry->block = block; + __entry->flags = ar->flags; + __entry->len = ar->len; + __entry->logical = ar->logical; + __entry->goal = ar->goal; + __entry->lleft = ar->lleft; + __entry->lright = ar->lright; + __entry->pleft = ar->pleft; + __entry->pright = ar->pright; + ), + + TP_printk("dev %s ino %lu flags %u len %u block %llu lblk %llu goal %llu lleft %llu lright %llu pleft %llu pright %llu ", + jbd2_dev_to_name(__entry->dev), __entry->ino, __entry->flags, + __entry->len, __entry->block, + (unsigned long long) __entry->logical, + (unsigned long long) __entry->goal, + (unsigned long long) __entry->lleft, + (unsigned long long) __entry->lright, + (unsigned long long) __entry->pleft, + (unsigned long long) __entry->pright) +); + +TRACE_EVENT(ext4_free_blocks, + TP_PROTO(struct inode *inode, __u64 block, unsigned long count, + int metadata), + + TP_ARGS(inode, block, count, metadata), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( ino_t, ino ) + __field( __u64, block ) + __field( unsigned long, count ) + __field( int, metadata ) + + ), + + TP_fast_assign( + __entry->dev = inode->i_sb->s_dev; + __entry->ino = inode->i_ino; + __entry->block = block; + __entry->count = count; + __entry->metadata = metadata; + ), + + TP_printk("dev %s ino %lu block %llu count %lu metadata %d", + jbd2_dev_to_name(__entry->dev), __entry->ino, __entry->block, + __entry->count, __entry->metadata) +); + +TRACE_EVENT(ext4_sync_file, + TP_PROTO(struct file *file, struct dentry *dentry, int datasync), + + TP_ARGS(file, dentry, datasync), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( ino_t, ino ) + __field( ino_t, parent ) + __field( int, datasync ) + ), + + TP_fast_assign( + __entry->dev = dentry->d_inode->i_sb->s_dev; + __entry->ino = dentry->d_inode->i_ino; + __entry->datasync = datasync; + __entry->parent = dentry->d_parent->d_inode->i_ino; + ), + + TP_printk("dev %s ino %ld parent %ld datasync %d ", + jbd2_dev_to_name(__entry->dev), __entry->ino, __entry->parent, + __entry->datasync) +); + +TRACE_EVENT(ext4_sync_fs, + TP_PROTO(struct super_block *sb, int wait), + + TP_ARGS(sb, wait), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( int, wait ) + + ), + + TP_fast_assign( + __entry->dev = sb->s_dev; + __entry->wait = wait; + ), + + TP_printk("dev %s wait %d", jbd2_dev_to_name(__entry->dev), + __entry->wait) +); + +#endif /* _TRACE_EXT4_H */ + +/* This part must be outside protection */ +#include -- cgit v1.2.3 From 96159f25112595386c56e09eca90284e85e7ecbf Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Mon, 27 Apr 2009 17:24:22 -0400 Subject: ext3: avoid unnecessary spinlock in critical POSIX ACL path If a filesystem supports POSIX ACL's, the VFS layer expects the filesystem to do POSIX ACL checks on any files not owned by the caller, and it does this for every single pathname component that it looks up. That obviously can be pretty expensive if the filesystem isn't careful about it, especially with locking. That's doubly sad, since the common case tends to be that there are no ACL's associated with the files in question. ext3 already caches the ACL data so that it doesn't have to look it up over and over again, but it does so by taking the inode->i_lock spinlock on every lookup. Which is a noticeable overhead even if it's a private lock, especially on CPU's where the serialization is expensive (eg Intel Netburst aka 'P4'). For the special case of not actually having any ACL's, all that locking is unnecessary. Even if somebody else were to be changing the ACL's on another CPU, we simply don't care - if we've seen a NULL ACL, we might as well use it. So just load the ACL speculatively without any locking, and if it was NULL, just use it. If it's non-NULL (either because we had a cached entry, or because the cache hasn't been filled in at all), it means that we'll need to get the lock and re-load it properly. This is noticeable even on Nehalem, which does locking quite well (much better than P4). From lmbench: Processor, Processes - times in microseconds - smaller is better -------------------------------------------------------------------- Host OS Mhz null null open slct fork exec sh call I/O stat clos TCP proc proc proc --------- ------------- ---- ---- ---- ---- ---- ---- ---- ---- ---- - before: nehalem.l Linux 2.6.30- 3193 0.04 0.09 0.95 1.45 2.18 69.1 273. 1141 nehalem.l Linux 2.6.30- 3193 0.04 0.09 0.95 1.48 2.28 69.9 253. 1140 nehalem.l Linux 2.6.30- 3193 0.04 0.10 0.95 1.42 2.19 68.6 284. 1141 - after: nehalem.l Linux 2.6.30- 3193 0.04 0.09 0.92 1.44 2.12 68.3 282. 1094 nehalem.l Linux 2.6.30- 3193 0.04 0.09 0.92 1.39 2.20 67.0 308. 1123 nehalem.l Linux 2.6.30- 3193 0.04 0.09 0.92 1.39 2.36 67.4 293. 1148 where you can see what appears to be a roughly 3% improvement in stat and open/close latencies from just the removal of the locking overhead. Of course, this only matters for files you don't own (the owner never needs to do the ACL checks), but that's the common case for libraries, header files, and executables. As well as for the base components of any absolute pathname, even if you are the owner of the final file. [ At some point we probably want to move this ACL caching logic entirely into the VFS layer (and only call down to the filesystem when uncached), but in the meantime this improves ext3 a bit. A similar fix to btrfs makes a much bigger difference (15x improvement in lmbench) due to broken caching. ] Signed-off-by: Linus Torvalds Signed-off-by: "Theodore Ts'o" Acked-by: Jan Kara Cc: Al Viro --- fs/ext3/acl.c | 13 ++++++++----- 1 file changed, 8 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/ext3/acl.c b/fs/ext3/acl.c index d81ef2fdb08e..e0c745451715 100644 --- a/fs/ext3/acl.c +++ b/fs/ext3/acl.c @@ -129,12 +129,15 @@ fail: static inline struct posix_acl * ext3_iget_acl(struct inode *inode, struct posix_acl **i_acl) { - struct posix_acl *acl = EXT3_ACL_NOT_CACHED; + struct posix_acl *acl = ACCESS_ONCE(*i_acl); - spin_lock(&inode->i_lock); - if (*i_acl != EXT3_ACL_NOT_CACHED) - acl = posix_acl_dup(*i_acl); - spin_unlock(&inode->i_lock); + if (acl) { + spin_lock(&inode->i_lock); + acl = *i_acl; + if (acl != EXT3_ACL_NOT_CACHED) + acl = posix_acl_dup(acl); + spin_unlock(&inode->i_lock); + } return acl; } -- cgit v1.2.3 From 8b0f9e8f78bd0a65fa001bf18f2c47eef2893a10 Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Mon, 27 Apr 2009 17:33:23 -0400 Subject: ext4: avoid unnecessary spinlock in critical POSIX ACL path If a filesystem supports POSIX ACL's, the VFS layer expects the filesystem to do POSIX ACL checks on any files not owned by the caller, and it does this for every single pathname component that it looks up. That obviously can be pretty expensive if the filesystem isn't careful about it, especially with locking. That's doubly sad, since the common case tends to be that there are no ACL's associated with the files in question. ext4 already caches the ACL data so that it doesn't have to look it up over and over again, but it does so by taking the inode->i_lock spinlock on every lookup. Which is a noticeable overhead even if it's a private lock, especially on CPU's where the serialization is expensive (eg Intel Netburst aka 'P4'). For the special case of not actually having any ACL's, all that locking is unnecessary. Even if somebody else were to be changing the ACL's on another CPU, we simply don't care - if we've seen a NULL ACL, we might as well use it. So just load the ACL speculatively without any locking, and if it was NULL, just use it. If it's non-NULL (either because we had a cached entry, or because the cache hasn't been filled in at all), it means that we'll need to get the lock and re-load it properly. (This commit was ported from a patch originally authored by Linus for ext3.) Signed-off-by: "Theodore Ts'o" --- fs/ext4/acl.c | 13 ++++++++----- 1 file changed, 8 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/ext4/acl.c b/fs/ext4/acl.c index 647e0d65a284..605aeed96d68 100644 --- a/fs/ext4/acl.c +++ b/fs/ext4/acl.c @@ -129,12 +129,15 @@ fail: static inline struct posix_acl * ext4_iget_acl(struct inode *inode, struct posix_acl **i_acl) { - struct posix_acl *acl = EXT4_ACL_NOT_CACHED; + struct posix_acl *acl = ACCESS_ONCE(*i_acl); - spin_lock(&inode->i_lock); - if (*i_acl != EXT4_ACL_NOT_CACHED) - acl = posix_acl_dup(*i_acl); - spin_unlock(&inode->i_lock); + if (acl) { + spin_lock(&inode->i_lock); + acl = *i_acl; + if (acl != EXT4_ACL_NOT_CACHED) + acl = posix_acl_dup(acl); + spin_unlock(&inode->i_lock); + } return acl; } -- cgit v1.2.3 From 4cd5dc751a5889b5b37aa88752e33a58d8f6b624 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 23 Apr 2009 19:31:32 -0400 Subject: NFSD: Refactor transport removal out of __write_ports() Clean up: Refactor transport removal out of __write_ports() to make it easier to understand and maintain. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/nfsd/nfsctl.c | 53 +++++++++++++++++++++++++++++------------------------ 1 file changed, 29 insertions(+), 24 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c index af16849d243a..2c1dce8e9a06 100644 --- a/fs/nfsd/nfsctl.c +++ b/fs/nfsd/nfsctl.c @@ -910,6 +910,31 @@ static ssize_t write_versions(struct file *file, char *buf, size_t size) return rv; } +/* + * A transport listener is removed by writing a "-", it's transport + * name, and it's port number. + */ +static ssize_t __write_ports_delxprt(char *buf) +{ + struct svc_xprt *xprt; + char transport[16]; + int port; + + if (sscanf(&buf[1], "%15s %4u", transport, &port) != 2) + return -EINVAL; + + if (port < 1 || port > USHORT_MAX || nfsd_serv == NULL) + return -EINVAL; + + xprt = svc_find_xprt(nfsd_serv, transport, AF_UNSPEC, port); + if (xprt == NULL) + return -ENOTCONN; + + svc_close_xprt(xprt); + svc_xprt_put(xprt); + return 0; +} + static ssize_t __write_ports(struct file *file, char *buf, size_t size) { if (size == 0) { @@ -984,30 +1009,10 @@ static ssize_t __write_ports(struct file *file, char *buf, size_t size) return err < 0 ? err : 0; } } - /* - * Remove a transport by writing it's transport name and port number - */ - if (buf[0] == '-' && isalpha(buf[1])) { - struct svc_xprt *xprt; - int err = -EINVAL; - char transport[16]; - int port; - if (sscanf(&buf[1], "%15s %4d", transport, &port) == 2) { - if (port < 1 || port > 65535) - return -EINVAL; - if (nfsd_serv) { - xprt = svc_find_xprt(nfsd_serv, transport, - AF_UNSPEC, port); - if (xprt) { - svc_close_xprt(xprt); - svc_xprt_put(xprt); - err = 0; - } else - err = -ENOTCONN; - } - return err < 0 ? err : 0; - } - } + + if (buf[0] == '-' && isalpha(buf[1])) + return __write_ports_delxprt(buf); + return -EINVAL; } -- cgit v1.2.3 From 4eb68c266cb1754ffa0040e882882680ece8cf34 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 23 Apr 2009 19:31:40 -0400 Subject: NFSD: Refactor transport addition out of __write_ports() Clean up: Refactor transport addition out of __write_ports() to make it easier to understand and maintain. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/nfsd/nfsctl.c | 56 +++++++++++++++++++++++++++++++++----------------------- 1 file changed, 33 insertions(+), 23 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c index 2c1dce8e9a06..748532b93fd9 100644 --- a/fs/nfsd/nfsctl.c +++ b/fs/nfsd/nfsctl.c @@ -910,6 +910,36 @@ static ssize_t write_versions(struct file *file, char *buf, size_t size) return rv; } +/* + * A transport listener is added by writing it's transport name and + * a port number. + */ +static ssize_t __write_ports_addxprt(char *buf) +{ + char transport[16]; + int port, err; + + if (sscanf(buf, "%15s %4u", transport, &port) != 2) + return -EINVAL; + + if (port < 1 || port > USHORT_MAX) + return -EINVAL; + + err = nfsd_create_serv(); + if (err != 0) + return err; + + err = svc_create_xprt(nfsd_serv, transport, + PF_INET, port, SVC_SOCK_ANONYMOUS); + if (err < 0) { + /* Give a reasonable perror msg for bad transport string */ + if (err == -ENOENT) + err = -EPROTONOSUPPORT; + return err; + } + return 0; +} + /* * A transport listener is removed by writing a "-", it's transport * name, and it's port number. @@ -986,29 +1016,9 @@ static ssize_t __write_ports(struct file *file, char *buf, size_t size) kfree(toclose); return len; } - /* - * Add a transport listener by writing it's transport name - */ - if (isalpha(buf[0])) { - int err; - char transport[16]; - int port; - if (sscanf(buf, "%15s %4d", transport, &port) == 2) { - if (port < 1 || port > 65535) - return -EINVAL; - err = nfsd_create_serv(); - if (!err) { - err = svc_create_xprt(nfsd_serv, - transport, PF_INET, port, - SVC_SOCK_ANONYMOUS); - if (err == -ENOENT) - /* Give a reasonable perror msg for - * bad transport string */ - err = -EPROTONOSUPPORT; - } - return err < 0 ? err : 0; - } - } + + if (isalpha(buf[0])) + return __write_ports_addxprt(buf); if (buf[0] == '-' && isalpha(buf[1])) return __write_ports_delxprt(buf); -- cgit v1.2.3 From 82d565919aa4138bc481a7473491b71385af4018 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 23 Apr 2009 19:31:48 -0400 Subject: NFSD: Refactor portlist socket closing into a helper Clean up: Refactor the socket closing logic out of __write_ports() to make it easier to understand and maintain. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/nfsd/nfsctl.c | 36 ++++++++++++++++++++++++------------ 1 file changed, 24 insertions(+), 12 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c index 748532b93fd9..fa268d1b775f 100644 --- a/fs/nfsd/nfsctl.c +++ b/fs/nfsd/nfsctl.c @@ -910,6 +910,27 @@ static ssize_t write_versions(struct file *file, char *buf, size_t size) return rv; } +/* + * A '-' followed by the 'name' of a socket means we close the socket. + */ +static ssize_t __write_ports_delfd(char *buf) +{ + char *toclose; + int len = 0; + + toclose = kstrdup(buf + 1, GFP_KERNEL); + if (toclose == NULL) + return -ENOMEM; + + if (nfsd_serv != NULL) + len = svc_sock_names(buf, nfsd_serv, toclose); + if (len >= 0) + lockd_down(); + + kfree(toclose); + return len; +} + /* * A transport listener is added by writing it's transport name and * a port number. @@ -1004,18 +1025,9 @@ static ssize_t __write_ports(struct file *file, char *buf, size_t size) } return err < 0 ? err : 0; } - if (buf[0] == '-' && isdigit(buf[1])) { - char *toclose = kstrdup(buf+1, GFP_KERNEL); - int len = 0; - if (!toclose) - return -ENOMEM; - if (nfsd_serv) - len = svc_sock_names(buf, nfsd_serv, toclose); - if (len >= 0) - lockd_down(); - kfree(toclose); - return len; - } + + if (buf[0] == '-' && isdigit(buf[1])) + return __write_ports_delfd(buf); if (isalpha(buf[0])) return __write_ports_addxprt(buf); -- cgit v1.2.3 From 0b7c2f6fc7e06867885c7b0f256ff5bd494e0653 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 23 Apr 2009 19:31:55 -0400 Subject: NFSD: Refactor socket creation out of __write_ports() Clean up: Refactor the socket creation logic out of __write_ports() to make it easier to understand and maintain. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/nfsd/nfsctl.c | 64 ++++++++++++++++++++++++++++++-------------------------- 1 file changed, 34 insertions(+), 30 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c index fa268d1b775f..b6a847faa0bb 100644 --- a/fs/nfsd/nfsctl.c +++ b/fs/nfsd/nfsctl.c @@ -910,6 +910,37 @@ static ssize_t write_versions(struct file *file, char *buf, size_t size) return rv; } +/* + * A single 'fd' number was written, in which case it must be for + * a socket of a supported family/protocol, and we use it as an + * nfsd listener. + */ +static ssize_t __write_ports_addfd(char *buf) +{ + char *mesg = buf; + int fd, err; + + err = get_int(&mesg, &fd); + if (err != 0 || fd < 0) + return -EINVAL; + + err = nfsd_create_serv(); + if (err != 0) + return err; + + err = svc_addsock(nfsd_serv, fd, buf); + if (err >= 0) { + err = lockd_up(); + if (err < 0) + svc_sock_names(buf + strlen(buf) + 1, nfsd_serv, buf); + + /* Decrease the count, but don't shut down the service */ + nfsd_serv->sv_nrthreads--; + } + + return err < 0 ? err : 0; +} + /* * A '-' followed by the 'name' of a socket means we close the socket. */ @@ -995,36 +1026,9 @@ static ssize_t __write_ports(struct file *file, char *buf, size_t size) len = svc_xprt_names(nfsd_serv, buf, 0); return len; } - /* Either a single 'fd' number is written, in which - * case it must be for a socket of a supported family/protocol, - * and we use it as an nfsd socket, or - * A '-' followed by the 'name' of a socket in which case - * we close the socket. - */ - if (isdigit(buf[0])) { - char *mesg = buf; - int fd; - int err; - err = get_int(&mesg, &fd); - if (err) - return -EINVAL; - if (fd < 0) - return -EINVAL; - err = nfsd_create_serv(); - if (!err) { - err = svc_addsock(nfsd_serv, fd, buf); - if (err >= 0) { - err = lockd_up(); - if (err < 0) - svc_sock_names(buf+strlen(buf)+1, nfsd_serv, buf); - } - /* Decrease the count, but don't shutdown the - * the service - */ - nfsd_serv->sv_nrthreads--; - } - return err < 0 ? err : 0; - } + + if (isdigit(buf[0])) + return __write_ports_addfd(buf); if (buf[0] == '-' && isdigit(buf[1])) return __write_ports_delfd(buf); -- cgit v1.2.3 From c71206a7b4e829601e44c49d8a0f4df0fdfe9302 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 23 Apr 2009 19:32:03 -0400 Subject: NFSD: Note an additional requirement when passing TCP sockets to portlist User space must call listen(3) on SOCK_STREAM sockets passed into /proc/fs/nfsd/portlist, otherwise that listener is ignored. Document this. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/nfsd/nfsctl.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c index b6a847faa0bb..d491fa9df9bd 100644 --- a/fs/nfsd/nfsctl.c +++ b/fs/nfsd/nfsctl.c @@ -1061,7 +1061,9 @@ static ssize_t __write_ports(struct file *file, char *buf, size_t size) * buf: C string containing an unsigned * integer value representing a bound * but unconnected socket that is to be - * used as an NFSD listener + * used as an NFSD listener; listen(3) + * must be called for a SOCK_STREAM + * socket, otherwise it is ignored * size: non-zero length of C string in @buf * Output: * On success: NFS service is started; -- cgit v1.2.3 From 0a5372d8a13c3f670bf7c9c4a68ab04c3664fa28 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 23 Apr 2009 19:32:10 -0400 Subject: NFSD: Finish refactoring __write_ports() Clean up: Refactor transport name listing out of __write_ports() to make it easier to understand and maintain. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/nfsd/nfsctl.c | 20 +++++++++++++------- 1 file changed, 13 insertions(+), 7 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c index d491fa9df9bd..caf4fdc674c0 100644 --- a/fs/nfsd/nfsctl.c +++ b/fs/nfsd/nfsctl.c @@ -910,6 +910,17 @@ static ssize_t write_versions(struct file *file, char *buf, size_t size) return rv; } +/* + * Zero-length write. Return a list of NFSD's current listener + * transports. + */ +static ssize_t __write_ports_names(char *buf) +{ + if (nfsd_serv == NULL) + return 0; + return svc_xprt_names(nfsd_serv, buf, 0); +} + /* * A single 'fd' number was written, in which case it must be for * a socket of a supported family/protocol, and we use it as an @@ -1019,13 +1030,8 @@ static ssize_t __write_ports_delxprt(char *buf) static ssize_t __write_ports(struct file *file, char *buf, size_t size) { - if (size == 0) { - int len = 0; - - if (nfsd_serv) - len = svc_xprt_names(nfsd_serv, buf, 0); - return len; - } + if (size == 0) + return __write_ports_names(buf); if (isdigit(buf[0])) return __write_ports_addfd(buf); -- cgit v1.2.3 From ea068bad27cefc71ab03230dbf01a8f8d98da5ba Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 23 Apr 2009 19:32:18 -0400 Subject: NFSD: move lockd_up() before svc_addsock() Clean up. A couple of years ago, a series of commits, finishing with commit 5680c446, swapped the order of the lockd_up() and svc_addsock() calls in __write_ports(). At that time lockd_up() needed to know the transport protocol of the passed-in socket to start a listener on the same transport protocol. These days, lockd_up() doesn't take a protocol argument; it always starts both a UDP and TCP listener. It's now more straightforward to try the lockd_up() first, then do a lockd_down() if the svc_addsock() fails. Careful review of this code shows that the svc_sock_names() call is used only to close the just-opened socket in case lockd_up() fails. So it is no longer needed if lockd_up() is done first. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/nfsd/nfsctl.c | 19 ++++++++++--------- 1 file changed, 10 insertions(+), 9 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c index caf4fdc674c0..e051847b93fb 100644 --- a/fs/nfsd/nfsctl.c +++ b/fs/nfsd/nfsctl.c @@ -939,17 +939,18 @@ static ssize_t __write_ports_addfd(char *buf) if (err != 0) return err; - err = svc_addsock(nfsd_serv, fd, buf); - if (err >= 0) { - err = lockd_up(); - if (err < 0) - svc_sock_names(buf + strlen(buf) + 1, nfsd_serv, buf); + err = lockd_up(); + if (err != 0) + goto out; - /* Decrease the count, but don't shut down the service */ - nfsd_serv->sv_nrthreads--; - } + err = svc_addsock(nfsd_serv, fd, buf); + if (err < 0) + lockd_down(); - return err < 0 ? err : 0; +out: + /* Decrease the count, but don't shut down the service */ + nfsd_serv->sv_nrthreads--; + return err; } /* -- cgit v1.2.3 From 335c54bdc4d3bacdbd619ec95cd0b352435bd37f Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 23 Apr 2009 19:32:25 -0400 Subject: NFSD: Prevent a buffer overflow in svc_xprt_names() The svc_xprt_names() function can overflow its buffer if it's so near the end of the passed in buffer that the "name too long" string still doesn't fit. Of course, it could never tell if it was near the end of the passed in buffer, since its only caller passes in zero as the buffer length. Let's make this API a little safer. Change svc_xprt_names() so it *always* checks for a buffer overflow, and change its only caller to pass in the correct buffer length. If svc_xprt_names() does overflow its buffer, it now fails with an ENAMETOOLONG errno, instead of trying to write a message at the end of the buffer. I don't like this much, but I can't figure out a clean way that's always safe to return some of the names, *and* an indication that the buffer was not long enough. The displayed error when doing a 'cat /proc/fs/nfsd/portlist' is "File name too long". Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/nfsd/nfsctl.c | 2 +- include/linux/sunrpc/svc_xprt.h | 2 +- net/sunrpc/svc_xprt.c | 56 ++++++++++++++++++++++++++++------------- 3 files changed, 41 insertions(+), 19 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c index e051847b93fb..6a1cd908e6bc 100644 --- a/fs/nfsd/nfsctl.c +++ b/fs/nfsd/nfsctl.c @@ -918,7 +918,7 @@ static ssize_t __write_ports_names(char *buf) { if (nfsd_serv == NULL) return 0; - return svc_xprt_names(nfsd_serv, buf, 0); + return svc_xprt_names(nfsd_serv, buf, SIMPLE_TRANSACTION_LIMIT); } /* diff --git a/include/linux/sunrpc/svc_xprt.h b/include/linux/sunrpc/svc_xprt.h index d790c52525cc..2223ae0b5ed5 100644 --- a/include/linux/sunrpc/svc_xprt.h +++ b/include/linux/sunrpc/svc_xprt.h @@ -83,7 +83,7 @@ int svc_port_is_privileged(struct sockaddr *sin); int svc_print_xprts(char *buf, int maxlen); struct svc_xprt *svc_find_xprt(struct svc_serv *serv, const char *xcl_name, const sa_family_t af, const unsigned short port); -int svc_xprt_names(struct svc_serv *serv, char *buf, int buflen); +int svc_xprt_names(struct svc_serv *serv, char *buf, const int buflen); static inline void svc_xprt_get(struct svc_xprt *xprt) { diff --git a/net/sunrpc/svc_xprt.c b/net/sunrpc/svc_xprt.c index f200393ac877..6f33d33cc064 100644 --- a/net/sunrpc/svc_xprt.c +++ b/net/sunrpc/svc_xprt.c @@ -1098,36 +1098,58 @@ struct svc_xprt *svc_find_xprt(struct svc_serv *serv, const char *xcl_name, } EXPORT_SYMBOL_GPL(svc_find_xprt); -/* - * Format a buffer with a list of the active transports. A zero for - * the buflen parameter disables target buffer overflow checking. +static int svc_one_xprt_name(const struct svc_xprt *xprt, + char *pos, int remaining) +{ + int len; + + len = snprintf(pos, remaining, "%s %u\n", + xprt->xpt_class->xcl_name, + svc_xprt_local_port(xprt)); + if (len >= remaining) + return -ENAMETOOLONG; + return len; +} + +/** + * svc_xprt_names - format a buffer with a list of transport names + * @serv: pointer to an RPC service + * @buf: pointer to a buffer to be filled in + * @buflen: length of buffer to be filled in + * + * Fills in @buf with a string containing a list of transport names, + * each name terminated with '\n'. + * + * Returns positive length of the filled-in string on success; otherwise + * a negative errno value is returned if an error occurs. */ -int svc_xprt_names(struct svc_serv *serv, char *buf, int buflen) +int svc_xprt_names(struct svc_serv *serv, char *buf, const int buflen) { struct svc_xprt *xprt; - char xprt_str[64]; - int totlen = 0; - int len; + int len, totlen; + char *pos; /* Sanity check args */ if (!serv) return 0; spin_lock_bh(&serv->sv_lock); + + pos = buf; + totlen = 0; list_for_each_entry(xprt, &serv->sv_permsocks, xpt_list) { - len = snprintf(xprt_str, sizeof(xprt_str), - "%s %d\n", xprt->xpt_class->xcl_name, - svc_xprt_local_port(xprt)); - /* If the string was truncated, replace with error string */ - if (len >= sizeof(xprt_str)) - strcpy(xprt_str, "name-too-long\n"); - /* Don't overflow buffer */ - len = strlen(xprt_str); - if (buflen && (len + totlen >= buflen)) + len = svc_one_xprt_name(xprt, pos, buflen - totlen); + if (len < 0) { + *buf = '\0'; + totlen = len; + } + if (len <= 0) break; - strcpy(buf+totlen, xprt_str); + + pos += len; totlen += len; } + spin_unlock_bh(&serv->sv_lock); return totlen; } -- cgit v1.2.3 From bfba9ab4c64f0e5c33930711e6c073c285e01fcf Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 23 Apr 2009 19:32:33 -0400 Subject: SUNRPC: pass buffer size to svc_addsock() Adjust the synopsis of svc_addsock() to pass in the size of the output buffer. Add a documenting comment. This is a cosmetic change for now. A subsequent patch will make sure the buffer length is passed to one_sock_name(), where the length will actually be useful. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/nfsd/nfsctl.c | 2 +- include/linux/sunrpc/svcsock.h | 3 ++- net/sunrpc/svcsock.c | 16 +++++++++++++--- 3 files changed, 16 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c index 6a1cd908e6bc..1f1c2159b802 100644 --- a/fs/nfsd/nfsctl.c +++ b/fs/nfsd/nfsctl.c @@ -943,7 +943,7 @@ static ssize_t __write_ports_addfd(char *buf) if (err != 0) goto out; - err = svc_addsock(nfsd_serv, fd, buf); + err = svc_addsock(nfsd_serv, fd, buf, SIMPLE_TRANSACTION_LIMIT); if (err < 0) lockd_down(); diff --git a/include/linux/sunrpc/svcsock.h b/include/linux/sunrpc/svcsock.h index 483e10380aae..e23241c53f42 100644 --- a/include/linux/sunrpc/svcsock.h +++ b/include/linux/sunrpc/svcsock.h @@ -39,7 +39,8 @@ int svc_send(struct svc_rqst *); void svc_drop(struct svc_rqst *); void svc_sock_update_bufs(struct svc_serv *serv); int svc_sock_names(char *buf, struct svc_serv *serv, char *toclose); -int svc_addsock(struct svc_serv *serv, int fd, char *name_return); +int svc_addsock(struct svc_serv *serv, const int fd, + char *name_return, const size_t len); void svc_init_xprt_sock(void); void svc_cleanup_xprt_sock(void); diff --git a/net/sunrpc/svcsock.c b/net/sunrpc/svcsock.c index 8b0832834135..6bec1e25b542 100644 --- a/net/sunrpc/svcsock.c +++ b/net/sunrpc/svcsock.c @@ -1128,9 +1128,19 @@ static struct svc_sock *svc_setup_socket(struct svc_serv *serv, return svsk; } -int svc_addsock(struct svc_serv *serv, - int fd, - char *name_return) +/** + * svc_addsock - add a listener socket to an RPC service + * @serv: pointer to RPC service to which to add a new listener + * @fd: file descriptor of the new listener + * @name_return: pointer to buffer to fill in with name of listener + * @len: size of the buffer + * + * Fills in socket name and returns positive length of name if successful. + * Name is terminated with '\n'. On error, returns a negative errno + * value. + */ +int svc_addsock(struct svc_serv *serv, const int fd, char *name_return, + const size_t len) { int err = 0; struct socket *so = sockfd_lookup(fd, &err); -- cgit v1.2.3 From 8435d34dbbe75678c3cdad3d53b1e7996a79b3bf Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 23 Apr 2009 19:32:40 -0400 Subject: SUNRPC: pass buffer size to svc_sock_names() Adjust the synopsis of svc_sock_names() to pass in the size of the output buffer. Add a documenting comment. This is a cosmetic change for now. A subsequent patch will make sure the buffer length is passed to one_sock_name(), where the length will actually be useful. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/nfsd/nfsctl.c | 3 ++- include/linux/sunrpc/svcsock.h | 4 +++- net/sunrpc/svcsock.c | 19 +++++++++++++++++-- 3 files changed, 22 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c index 1f1c2159b802..b64a7fbfccf5 100644 --- a/fs/nfsd/nfsctl.c +++ b/fs/nfsd/nfsctl.c @@ -966,7 +966,8 @@ static ssize_t __write_ports_delfd(char *buf) return -ENOMEM; if (nfsd_serv != NULL) - len = svc_sock_names(buf, nfsd_serv, toclose); + len = svc_sock_names(nfsd_serv, buf, + SIMPLE_TRANSACTION_LIMIT, toclose); if (len >= 0) lockd_down(); diff --git a/include/linux/sunrpc/svcsock.h b/include/linux/sunrpc/svcsock.h index e23241c53f42..827163138949 100644 --- a/include/linux/sunrpc/svcsock.h +++ b/include/linux/sunrpc/svcsock.h @@ -38,7 +38,9 @@ int svc_recv(struct svc_rqst *, long); int svc_send(struct svc_rqst *); void svc_drop(struct svc_rqst *); void svc_sock_update_bufs(struct svc_serv *serv); -int svc_sock_names(char *buf, struct svc_serv *serv, char *toclose); +int svc_sock_names(struct svc_serv *serv, char *buf, + const size_t buflen, + const char *toclose); int svc_addsock(struct svc_serv *serv, const int fd, char *name_return, const size_t len); void svc_init_xprt_sock(void); diff --git a/net/sunrpc/svcsock.c b/net/sunrpc/svcsock.c index 6bec1e25b542..032b52ea9541 100644 --- a/net/sunrpc/svcsock.c +++ b/net/sunrpc/svcsock.c @@ -259,8 +259,23 @@ static int one_sock_name(char *buf, struct svc_sock *svsk) return len; } -int -svc_sock_names(char *buf, struct svc_serv *serv, char *toclose) +/** + * svc_sock_names - construct a list of listener names in a string + * @serv: pointer to RPC service + * @buf: pointer to a buffer to fill in with socket names + * @buflen: size of the buffer to be filled + * @toclose: pointer to '\0'-terminated C string containing the name + * of a listener to be closed + * + * Fills in @buf with a '\n'-separated list of names of listener + * sockets. If @toclose is not NULL, the socket named by @toclose + * is closed, and is not included in the output list. + * + * Returns positive length of the socket name string, or a negative + * errno value on error. + */ +int svc_sock_names(struct svc_serv *serv, char *buf, const size_t buflen, + const char *toclose) { struct svc_sock *svsk, *closesk = NULL; int len = 0; -- cgit v1.2.3 From 3d72ab8fdd44c872633b210dd1a4afd2910d0bbb Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 23 Apr 2009 19:33:10 -0400 Subject: NFSD: Stricter buffer size checking in write_recoverydir() While it's not likely a pathname will be longer than SIMPLE_TRANSACTION_SIZE, we should be more careful about just plopping it into the output buffer without bounds checking. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/nfsd/nfsctl.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c index b64a7fbfccf5..c4843467cfd4 100644 --- a/fs/nfsd/nfsctl.c +++ b/fs/nfsd/nfsctl.c @@ -1260,8 +1260,9 @@ static ssize_t __write_recoverydir(struct file *file, char *buf, size_t size) status = nfs4_reset_recoverydir(recdir); } - sprintf(buf, "%s\n", nfs4_recoverydir()); - return strlen(buf); + + return scnprintf(buf, SIMPLE_TRANSACTION_LIMIT, "%s\n", + nfs4_recoverydir()); } /** -- cgit v1.2.3 From 261758b5c3dfeac73ca364c47ed538f5ce4250ee Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 23 Apr 2009 19:33:18 -0400 Subject: NFSD: Stricter buffer size checking in write_versions() While it's not likely today that there are enough NFS versions to overflow the output buffer in write_versions(), we should be more careful about detecting the end of the buffer. The number of NFS versions will only increase as NFSv4 minor versions are added. Note that this API doesn't behave the same as portlist. Here we attempt to display as many versions as will fit in the buffer, and do not provide any indication that an overflow would have occurred. I don't have any good rationale for that. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/nfsd/nfsctl.c | 32 +++++++++++++++++++++++++------- 1 file changed, 25 insertions(+), 7 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c index c4843467cfd4..a152694e016e 100644 --- a/fs/nfsd/nfsctl.c +++ b/fs/nfsd/nfsctl.c @@ -793,7 +793,7 @@ static ssize_t __write_versions(struct file *file, char *buf, size_t size) { char *mesg = buf; char *vers, *minorp, sign; - int len, num; + int len, num, remaining; unsigned minor; ssize_t tlen = 0; char *sep; @@ -840,32 +840,50 @@ static ssize_t __write_versions(struct file *file, char *buf, size_t size) } next: vers += len + 1; - tlen += len; } while ((len = qword_get(&mesg, vers, size)) > 0); /* If all get turned off, turn them back on, as * having no versions is BAD */ nfsd_reset_versions(); } + /* Now write current state into reply buffer */ len = 0; sep = ""; + remaining = SIMPLE_TRANSACTION_LIMIT; for (num=2 ; num <= 4 ; num++) if (nfsd_vers(num, NFSD_AVAIL)) { - len += sprintf(buf+len, "%s%c%d", sep, + len = snprintf(buf, remaining, "%s%c%d", sep, nfsd_vers(num, NFSD_TEST)?'+':'-', num); sep = " "; + + if (len > remaining) + break; + remaining -= len; + buf += len; + tlen += len; } if (nfsd_vers(4, NFSD_AVAIL)) - for (minor = 1; minor <= NFSD_SUPPORTED_MINOR_VERSION; minor++) - len += sprintf(buf+len, " %c4.%u", + for (minor = 1; minor <= NFSD_SUPPORTED_MINOR_VERSION; + minor++) { + len = snprintf(buf, remaining, " %c4.%u", (nfsd_vers(4, NFSD_TEST) && nfsd_minorversion(minor, NFSD_TEST)) ? '+' : '-', minor); - len += sprintf(buf+len, "\n"); - return len; + + if (len > remaining) + break; + remaining -= len; + buf += len; + tlen += len; + } + + len = snprintf(buf, remaining, "\n"); + if (len > remaining) + return -EINVAL; + return tlen + len; } /** -- cgit v1.2.3 From e06b64050ebea1220bca3c291d0431f5206b5c95 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 23 Apr 2009 19:33:25 -0400 Subject: NFSD: Stricter buffer size checking in fs/nfsd/nfsctl.c Clean up: For consistency, handle output buffer size checking in a other nfsctl functions the same way it's done for write_versions(). Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- fs/nfsd/nfsctl.c | 14 +++++++++----- 1 file changed, 9 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c index a152694e016e..877e713a0fd6 100644 --- a/fs/nfsd/nfsctl.c +++ b/fs/nfsd/nfsctl.c @@ -695,8 +695,9 @@ static ssize_t write_threads(struct file *file, char *buf, size_t size) if (rv) return rv; } - sprintf(buf, "%d\n", nfsd_nrthreads()); - return strlen(buf); + + return scnprintf(buf, SIMPLE_TRANSACTION_LIMIT, "%d\n", + nfsd_nrthreads()); } /** @@ -1197,7 +1198,9 @@ static ssize_t write_maxblksize(struct file *file, char *buf, size_t size) nfsd_max_blksize = bsize; mutex_unlock(&nfsd_mutex); } - return sprintf(buf, "%d\n", nfsd_max_blksize); + + return scnprintf(buf, SIMPLE_TRANSACTION_LIMIT, "%d\n", + nfsd_max_blksize); } #ifdef CONFIG_NFSD_V4 @@ -1221,8 +1224,9 @@ static ssize_t __write_leasetime(struct file *file, char *buf, size_t size) return -EINVAL; nfs4_reset_lease(lease); } - sprintf(buf, "%ld\n", nfs4_lease_time()); - return strlen(buf); + + return scnprintf(buf, SIMPLE_TRANSACTION_LIMIT, "%ld\n", + nfs4_lease_time()); } /** -- cgit v1.2.3 From bc749ca4c405d507e6ec6e3f3e5475e9a09faf0a Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Tue, 7 Apr 2009 16:55:27 -0700 Subject: nfsd: eliminate ENCODE_HEAD macro This macro doesn't serve any useful purpose. Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4xdr.c | 46 ++++++++++++++++++++++------------------------ 1 file changed, 22 insertions(+), 24 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c index b820c311931c..fe46edeab2f6 100644 --- a/fs/nfsd/nfs4xdr.c +++ b/fs/nfsd/nfs4xdr.c @@ -1503,8 +1503,6 @@ nfsd4_decode_compound(struct nfsd4_compoundargs *argp) * task to translate them into Linux-specific versions which are more * consistent with the style used in NFSv2/v3... */ -#define ENCODE_HEAD __be32 *p - #define WRITE32(n) *p++ = htonl(n) #define WRITE64(n) do { \ *p++ = htonl((u32)((n) >> 32)); \ @@ -2334,7 +2332,7 @@ fail: static void nfsd4_encode_stateid(struct nfsd4_compoundres *resp, stateid_t *sid) { - ENCODE_HEAD; + __be32 *p; RESERVE_SPACE(sizeof(stateid_t)); WRITE32(sid->si_generation); @@ -2345,7 +2343,7 @@ nfsd4_encode_stateid(struct nfsd4_compoundres *resp, stateid_t *sid) static __be32 nfsd4_encode_access(struct nfsd4_compoundres *resp, __be32 nfserr, struct nfsd4_access *access) { - ENCODE_HEAD; + __be32 *p; if (!nfserr) { RESERVE_SPACE(8); @@ -2372,7 +2370,7 @@ nfsd4_encode_close(struct nfsd4_compoundres *resp, __be32 nfserr, struct nfsd4_c static __be32 nfsd4_encode_commit(struct nfsd4_compoundres *resp, __be32 nfserr, struct nfsd4_commit *commit) { - ENCODE_HEAD; + __be32 *p; if (!nfserr) { RESERVE_SPACE(8); @@ -2385,7 +2383,7 @@ nfsd4_encode_commit(struct nfsd4_compoundres *resp, __be32 nfserr, struct nfsd4_ static __be32 nfsd4_encode_create(struct nfsd4_compoundres *resp, __be32 nfserr, struct nfsd4_create *create) { - ENCODE_HEAD; + __be32 *p; if (!nfserr) { RESERVE_SPACE(32); @@ -2421,7 +2419,7 @@ nfsd4_encode_getfh(struct nfsd4_compoundres *resp, __be32 nfserr, struct svc_fh { struct svc_fh *fhp = *fhpp; unsigned int len; - ENCODE_HEAD; + __be32 *p; if (!nfserr) { len = fhp->fh_handle.fh_size; @@ -2440,7 +2438,7 @@ nfsd4_encode_getfh(struct nfsd4_compoundres *resp, __be32 nfserr, struct svc_fh static void nfsd4_encode_lock_denied(struct nfsd4_compoundres *resp, struct nfsd4_lock_denied *ld) { - ENCODE_HEAD; + __be32 *p; RESERVE_SPACE(32 + XDR_LEN(ld->ld_sop ? ld->ld_sop->so_owner.len : 0)); WRITE64(ld->ld_start); @@ -2496,7 +2494,7 @@ nfsd4_encode_locku(struct nfsd4_compoundres *resp, __be32 nfserr, struct nfsd4_l static __be32 nfsd4_encode_link(struct nfsd4_compoundres *resp, __be32 nfserr, struct nfsd4_link *link) { - ENCODE_HEAD; + __be32 *p; if (!nfserr) { RESERVE_SPACE(20); @@ -2510,7 +2508,7 @@ nfsd4_encode_link(struct nfsd4_compoundres *resp, __be32 nfserr, struct nfsd4_li static __be32 nfsd4_encode_open(struct nfsd4_compoundres *resp, __be32 nfserr, struct nfsd4_open *open) { - ENCODE_HEAD; + __be32 *p; ENCODE_SEQID_OP_HEAD; if (nfserr) @@ -2605,7 +2603,7 @@ nfsd4_encode_read(struct nfsd4_compoundres *resp, __be32 nfserr, int v, pn; unsigned long maxcount; long len; - ENCODE_HEAD; + __be32 *p; if (nfserr) return nfserr; @@ -2667,7 +2665,7 @@ nfsd4_encode_readlink(struct nfsd4_compoundres *resp, __be32 nfserr, struct nfsd { int maxcount; char *page; - ENCODE_HEAD; + __be32 *p; if (nfserr) return nfserr; @@ -2716,7 +2714,7 @@ nfsd4_encode_readdir(struct nfsd4_compoundres *resp, __be32 nfserr, struct nfsd4 int maxcount; loff_t offset; __be32 *page, *savep, *tailbase; - ENCODE_HEAD; + __be32 *p; if (nfserr) return nfserr; @@ -2792,7 +2790,7 @@ err_no_verf: static __be32 nfsd4_encode_remove(struct nfsd4_compoundres *resp, __be32 nfserr, struct nfsd4_remove *remove) { - ENCODE_HEAD; + __be32 *p; if (!nfserr) { RESERVE_SPACE(20); @@ -2805,7 +2803,7 @@ nfsd4_encode_remove(struct nfsd4_compoundres *resp, __be32 nfserr, struct nfsd4_ static __be32 nfsd4_encode_rename(struct nfsd4_compoundres *resp, __be32 nfserr, struct nfsd4_rename *rename) { - ENCODE_HEAD; + __be32 *p; if (!nfserr) { RESERVE_SPACE(40); @@ -2825,7 +2823,7 @@ nfsd4_encode_secinfo(struct nfsd4_compoundres *resp, __be32 nfserr, u32 nflavs; struct exp_flavor_info *flavs; struct exp_flavor_info def_flavs[2]; - ENCODE_HEAD; + __be32 *p; if (nfserr) goto out; @@ -2890,7 +2888,7 @@ out: static __be32 nfsd4_encode_setattr(struct nfsd4_compoundres *resp, __be32 nfserr, struct nfsd4_setattr *setattr) { - ENCODE_HEAD; + __be32 *p; RESERVE_SPACE(12); if (nfserr) { @@ -2910,7 +2908,7 @@ nfsd4_encode_setattr(struct nfsd4_compoundres *resp, __be32 nfserr, struct nfsd4 static __be32 nfsd4_encode_setclientid(struct nfsd4_compoundres *resp, __be32 nfserr, struct nfsd4_setclientid *scd) { - ENCODE_HEAD; + __be32 *p; if (!nfserr) { RESERVE_SPACE(8 + sizeof(nfs4_verifier)); @@ -2930,7 +2928,7 @@ nfsd4_encode_setclientid(struct nfsd4_compoundres *resp, __be32 nfserr, struct n static __be32 nfsd4_encode_write(struct nfsd4_compoundres *resp, __be32 nfserr, struct nfsd4_write *write) { - ENCODE_HEAD; + __be32 *p; if (!nfserr) { RESERVE_SPACE(16); @@ -2946,7 +2944,7 @@ static __be32 nfsd4_encode_exchange_id(struct nfsd4_compoundres *resp, int nfserr, struct nfsd4_exchange_id *exid) { - ENCODE_HEAD; + __be32 *p; char *major_id; char *server_scope; int major_id_sz; @@ -3001,7 +2999,7 @@ static __be32 nfsd4_encode_create_session(struct nfsd4_compoundres *resp, int nfserr, struct nfsd4_create_session *sess) { - ENCODE_HEAD; + __be32 *p; if (nfserr) return nfserr; @@ -3057,7 +3055,7 @@ __be32 nfsd4_encode_sequence(struct nfsd4_compoundres *resp, int nfserr, struct nfsd4_sequence *seq) { - ENCODE_HEAD; + __be32 *p; if (nfserr) return nfserr; @@ -3205,7 +3203,7 @@ void nfsd4_encode_operation(struct nfsd4_compoundres *resp, struct nfsd4_op *op) { __be32 *statp; - ENCODE_HEAD; + __be32 *p; RESERVE_SPACE(8); WRITE32(op->opnum); @@ -3239,7 +3237,7 @@ status: void nfsd4_encode_replay(struct nfsd4_compoundres *resp, struct nfsd4_op *op) { - ENCODE_HEAD; + __be32 *p; struct nfs4_replay *rp = op->replay; BUG_ON(!rp); -- cgit v1.2.3 From 3352d2c2d0540955a7bbb3421a28330af7f9d79c Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Tue, 7 Apr 2009 17:03:19 -0700 Subject: nfsd4: delete obsolete xdr comments We don't need comments to tell us these macros are ugly. And we're long past trying to share any of this code with the BSD's. Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4xdr.c | 27 --------------------------- 1 file changed, 27 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c index fe46edeab2f6..4a71fcd3f036 100644 --- a/fs/nfsd/nfs4xdr.c +++ b/fs/nfsd/nfs4xdr.c @@ -83,16 +83,6 @@ check_filename(char *str, int len, __be32 err) return 0; } -/* - * START OF "GENERIC" DECODE ROUTINES. - * These may look a little ugly since they are imported from a "generic" - * set of XDR encode/decode routines which are intended to be shared by - * all of our NFSv4 implementations (OpenBSD, MacOS X...). - * - * If the pain of reading these is too great, it should be a straightforward - * task to translate them into Linux-specific versions which are more - * consistent with the style used in NFSv2/v3... - */ #define DECODE_HEAD \ __be32 *p; \ __be32 status @@ -1489,20 +1479,7 @@ nfsd4_decode_compound(struct nfsd4_compoundargs *argp) DECODE_TAIL; } -/* - * END OF "GENERIC" DECODE ROUTINES. - */ -/* - * START OF "GENERIC" ENCODE ROUTINES. - * These may look a little ugly since they are imported from a "generic" - * set of XDR encode/decode routines which are intended to be shared by - * all of our NFSv4 implementations (OpenBSD, MacOS X...). - * - * If the pain of reading these is too great, it should be a straightforward - * task to translate them into Linux-specific versions which are more - * consistent with the style used in NFSv2/v3... - */ #define WRITE32(n) *p++ = htonl(n) #define WRITE64(n) do { \ *p++ = htonl((u32)((n) >> 32)); \ @@ -3252,10 +3229,6 @@ nfsd4_encode_replay(struct nfsd4_compoundres *resp, struct nfsd4_op *op) ADJUST_ARGS(); } -/* - * END OF "GENERIC" ENCODE ROUTINES. - */ - int nfs4svc_encode_voidres(struct svc_rqst *rqstp, __be32 *p, void *dummy) { -- cgit v1.2.3 From c654b8a9cba6002aad1c01919e4928a79a4a6dcf Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Thu, 16 Apr 2009 17:33:25 -0400 Subject: nfsd: support ext4 i_version ext4 supports a real NFSv4 change attribute, which is bumped whenever the ctime would be updated, including times when two updates arrive within a jiffy of each other. (Note that although ext4 has space for nanosecond-precision ctime, the real resolution is lower: it actually uses jiffies as the time-source.) This ensures clients will invalidate their caches when they need to. There is some fear that keeping the i_version up-to-date could have performance drawbacks, so for now it's turned on only by a mount option. We hope to do something better eventually. Signed-off-by: J. Bruce Fields Cc: Theodore Tso --- fs/nfsd/nfs3xdr.c | 1 + fs/nfsd/nfs4xdr.c | 63 ++++++++++++++++++++++++++++++---------------- include/linux/nfsd/nfsfh.h | 7 ++++++ include/linux/nfsd/xdr4.h | 17 ++++++++++--- 4 files changed, 63 insertions(+), 25 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs3xdr.c b/fs/nfsd/nfs3xdr.c index 17d0dd997204..01d4ec1c88e0 100644 --- a/fs/nfsd/nfs3xdr.c +++ b/fs/nfsd/nfs3xdr.c @@ -272,6 +272,7 @@ void fill_post_wcc(struct svc_fh *fhp) err = vfs_getattr(fhp->fh_export->ex_path.mnt, fhp->fh_dentry, &fhp->fh_post_attr); + fhp->fh_post_change = fhp->fh_dentry->d_inode->i_version; if (err) fhp->fh_post_saved = 0; else diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c index 4a71fcd3f036..12d36a7361cd 100644 --- a/fs/nfsd/nfs4xdr.c +++ b/fs/nfsd/nfs4xdr.c @@ -1490,13 +1490,41 @@ nfsd4_decode_compound(struct nfsd4_compoundargs *argp) memcpy(p, ptr, nbytes); \ p += XDR_QUADLEN(nbytes); \ }} while (0) -#define WRITECINFO(c) do { \ - *p++ = htonl(c.atomic); \ - *p++ = htonl(c.before_ctime_sec); \ - *p++ = htonl(c.before_ctime_nsec); \ - *p++ = htonl(c.after_ctime_sec); \ - *p++ = htonl(c.after_ctime_nsec); \ -} while (0) + +static void write32(__be32 **p, u32 n) +{ + *(*p)++ = n; +} + +static void write64(__be32 **p, u64 n) +{ + write32(p, (u32)(n >> 32)); + write32(p, (u32)n); +} + +static void write_change(__be32 **p, struct kstat *stat, struct inode *inode) +{ + if (IS_I_VERSION(inode)) { + write64(p, inode->i_version); + } else { + write32(p, stat->ctime.tv_sec); + write32(p, stat->ctime.tv_nsec); + } +} + +static void write_cinfo(__be32 **p, struct nfsd4_change_info *c) +{ + write32(p, c->atomic); + if (c->change_supported) { + write64(p, c->before_change); + write64(p, c->after_change); + } else { + write32(p, c->before_ctime_sec); + write32(p, c->before_ctime_nsec); + write32(p, c->after_ctime_sec); + write32(p, c->after_ctime_nsec); + } +} #define RESERVE_SPACE(nbytes) do { \ p = resp->p; \ @@ -1849,16 +1877,9 @@ nfsd4_encode_fattr(struct svc_fh *fhp, struct svc_export *exp, WRITE32(NFS4_FH_PERSISTENT|NFS4_FH_VOL_RENAME); } if (bmval0 & FATTR4_WORD0_CHANGE) { - /* - * Note: This _must_ be consistent with the scheme for writing - * change_info, so any changes made here must be reflected there - * as well. (See xdr4.h:set_change_info() and the WRITECINFO() - * macro above.) - */ if ((buflen -= 8) < 0) goto out_resource; - WRITE32(stat.ctime.tv_sec); - WRITE32(stat.ctime.tv_nsec); + write_change(&p, &stat, dentry->d_inode); } if (bmval0 & FATTR4_WORD0_SIZE) { if ((buflen -= 8) < 0) @@ -2364,7 +2385,7 @@ nfsd4_encode_create(struct nfsd4_compoundres *resp, __be32 nfserr, struct nfsd4_ if (!nfserr) { RESERVE_SPACE(32); - WRITECINFO(create->cr_cinfo); + write_cinfo(&p, &create->cr_cinfo); WRITE32(2); WRITE32(create->cr_bmval[0]); WRITE32(create->cr_bmval[1]); @@ -2475,7 +2496,7 @@ nfsd4_encode_link(struct nfsd4_compoundres *resp, __be32 nfserr, struct nfsd4_li if (!nfserr) { RESERVE_SPACE(20); - WRITECINFO(link->li_cinfo); + write_cinfo(&p, &link->li_cinfo); ADJUST_ARGS(); } return nfserr; @@ -2493,7 +2514,7 @@ nfsd4_encode_open(struct nfsd4_compoundres *resp, __be32 nfserr, struct nfsd4_op nfsd4_encode_stateid(resp, &open->op_stateid); RESERVE_SPACE(40); - WRITECINFO(open->op_cinfo); + write_cinfo(&p, &open->op_cinfo); WRITE32(open->op_rflags); WRITE32(2); WRITE32(open->op_bmval[0]); @@ -2771,7 +2792,7 @@ nfsd4_encode_remove(struct nfsd4_compoundres *resp, __be32 nfserr, struct nfsd4_ if (!nfserr) { RESERVE_SPACE(20); - WRITECINFO(remove->rm_cinfo); + write_cinfo(&p, &remove->rm_cinfo); ADJUST_ARGS(); } return nfserr; @@ -2784,8 +2805,8 @@ nfsd4_encode_rename(struct nfsd4_compoundres *resp, __be32 nfserr, struct nfsd4_ if (!nfserr) { RESERVE_SPACE(40); - WRITECINFO(rename->rn_sinfo); - WRITECINFO(rename->rn_tinfo); + write_cinfo(&p, &rename->rn_sinfo); + write_cinfo(&p, &rename->rn_tinfo); ADJUST_ARGS(); } return nfserr; diff --git a/include/linux/nfsd/nfsfh.h b/include/linux/nfsd/nfsfh.h index afa19016c4a8..8f641c908450 100644 --- a/include/linux/nfsd/nfsfh.h +++ b/include/linux/nfsd/nfsfh.h @@ -151,9 +151,15 @@ typedef struct svc_fh { __u64 fh_pre_size; /* size before operation */ struct timespec fh_pre_mtime; /* mtime before oper */ struct timespec fh_pre_ctime; /* ctime before oper */ + /* + * pre-op nfsv4 change attr: note must check IS_I_VERSION(inode) + * to find out if it is valid. + */ + u64 fh_pre_change; /* Post-op attributes saved in fh_unlock */ struct kstat fh_post_attr; /* full attrs after operation */ + u64 fh_post_change; /* nfsv4 change; see above */ #endif /* CONFIG_NFSD_V3 */ } svc_fh; @@ -298,6 +304,7 @@ fill_pre_wcc(struct svc_fh *fhp) fhp->fh_pre_mtime = inode->i_mtime; fhp->fh_pre_ctime = inode->i_ctime; fhp->fh_pre_size = inode->i_size; + fhp->fh_pre_change = inode->i_version; fhp->fh_pre_saved = 1; } } diff --git a/include/linux/nfsd/xdr4.h b/include/linux/nfsd/xdr4.h index f80d6013fdc3..d0f050f01eca 100644 --- a/include/linux/nfsd/xdr4.h +++ b/include/linux/nfsd/xdr4.h @@ -64,10 +64,13 @@ static inline bool nfsd4_has_session(struct nfsd4_compound_state *cs) struct nfsd4_change_info { u32 atomic; + bool change_supported; u32 before_ctime_sec; u32 before_ctime_nsec; + u64 before_change; u32 after_ctime_sec; u32 after_ctime_nsec; + u64 after_change; }; struct nfsd4_access { @@ -503,10 +506,16 @@ set_change_info(struct nfsd4_change_info *cinfo, struct svc_fh *fhp) { BUG_ON(!fhp->fh_pre_saved || !fhp->fh_post_saved); cinfo->atomic = 1; - cinfo->before_ctime_sec = fhp->fh_pre_ctime.tv_sec; - cinfo->before_ctime_nsec = fhp->fh_pre_ctime.tv_nsec; - cinfo->after_ctime_sec = fhp->fh_post_attr.ctime.tv_sec; - cinfo->after_ctime_nsec = fhp->fh_post_attr.ctime.tv_nsec; + cinfo->change_supported = IS_I_VERSION(fhp->fh_dentry->d_inode); + if (cinfo->change_supported) { + cinfo->before_change = fhp->fh_pre_change; + cinfo->after_change = fhp->fh_post_change; + } else { + cinfo->before_ctime_sec = fhp->fh_pre_ctime.tv_sec; + cinfo->before_ctime_nsec = fhp->fh_pre_ctime.tv_nsec; + cinfo->after_ctime_sec = fhp->fh_post_attr.ctime.tv_sec; + cinfo->after_ctime_nsec = fhp->fh_post_attr.ctime.tv_nsec; + } } int nfs4svc_encode_voidres(struct svc_rqst *, __be32 *, void *); -- cgit v1.2.3 From b8fd47aefa5f13df1edacbc7e68d9874635109e5 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Wed, 29 Apr 2009 11:36:17 -0400 Subject: nfsd: quiet compile warning Stephen Rothwell said: "Today's linux-next build (powerpc ppc64_defconfig) produced this new warning: fs/nfsd/nfs4state.c: In function 'EXPIRED_STATEID': fs/nfsd/nfs4state.c:2757: warning: comparison of distinct pointer types lacks a cast Caused by commit 78155ed75f470710f2aecb3e75e3d97107ba8374 ("nfsd4: distinguish expired from stale stateids")." Signed-off-by: J. Bruce Fields Reported-by: Stephen Rothwell --- fs/nfsd/nfs4state.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 74e822ec34cb..d24dd12ddb4d 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -2754,7 +2754,7 @@ EXPIRED_STATEID(stateid_t *stateid) { if (time_before((unsigned long)boot_time, ((unsigned long)stateid->si_boot)) && - time_before((stateid->si_boot + lease_time), get_seconds())) { + time_before((unsigned long)(stateid->si_boot + lease_time), get_seconds())) { dprintk("NFSD: expired stateid (%08x/%08x/%08x/%08x)!\n", stateid->si_boot, stateid->si_stateownerid, stateid->si_fileid, stateid->si_generation); -- cgit v1.2.3 From f64f79ea5f5e02ba8585f35a10b4a3bcab0cea52 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Wed, 29 Apr 2009 13:45:36 -0400 Subject: nfsd4: setclientid_confirm callback-change fixes This setclientid_confirm case should allow the client to change callbacks, but it currently has a dummy implementation that just turns off callbacks completely. That dummy implementation isn't completely correct either, though: - There's no need to remove any client recovery directory in this case. - New clientid confirm verifiers should be generated (and returned) in setclientid; there's no need to generate a new one here. Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 2 -- 1 file changed, 2 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index d24dd12ddb4d..7e1fcc3aade4 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -1688,8 +1688,6 @@ nfsd4_setclientid_confirm(struct svc_rqst *rqstp, /* XXX: We just turn off callbacks until we can handle * change request correctly. */ atomic_set(&conf->cl_callback.cb_set, 0); - gen_confirm(conf); - nfsd4_remove_clid_dir(unconf); expire_client(unconf); status = nfs_ok; -- cgit v1.2.3 From 595947acaaef373445131471a78650003f5d8e7d Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Thu, 5 Mar 2009 17:18:10 -0500 Subject: nfsd4: set shorter timeout We tried to do something overly complicated with the callback rpc timeouts here. And they're wrong--the result is that by the time a single callback times out, it's already too late to tell the client (using the cb_path_down return to RENEW) that the callback is down. Use a much shorter, simpler timeout. Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4callback.c | 11 +++++++---- 1 file changed, 7 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c index 290289bd44f7..049f052a6eb3 100644 --- a/fs/nfsd/nfs4callback.c +++ b/fs/nfsd/nfs4callback.c @@ -358,6 +358,11 @@ static struct rpc_program cb_program = { .pipe_dir_name = "/nfsd4_cb", }; +static int max_cb_time(void) +{ + return max(NFSD_LEASE_TIME/10, (time_t)1) * HZ; +} + /* Reference counting, callback cleanup, etc., all look racy as heck. * And why is cb_set an atomic? */ @@ -366,10 +371,8 @@ static struct rpc_clnt *setup_callback_client(struct nfs4_client *clp) struct sockaddr_in addr; struct nfs4_callback *cb = &clp->cl_callback; struct rpc_timeout timeparms = { - .to_initval = (NFSD_LEASE_TIME/4) * HZ, - .to_retries = 5, - .to_maxval = (NFSD_LEASE_TIME/2) * HZ, - .to_exponential = 1, + .to_initval = max_cb_time(), + .to_retries = 0, }; struct rpc_create_args args = { .protocol = IPPROTO_TCP, -- cgit v1.2.3 From e1cab5a5896e142190cd66a8287099b52e5855a7 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Mon, 23 Feb 2009 10:45:27 -0800 Subject: nfsd4: set cb_client inside setup_callback_client This is just a minor code simplification. Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4callback.c | 26 ++++++++++++-------------- 1 file changed, 12 insertions(+), 14 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c index 049f052a6eb3..4788d09d9bec 100644 --- a/fs/nfsd/nfs4callback.c +++ b/fs/nfsd/nfs4callback.c @@ -366,7 +366,7 @@ static int max_cb_time(void) /* Reference counting, callback cleanup, etc., all look racy as heck. * And why is cb_set an atomic? */ -static struct rpc_clnt *setup_callback_client(struct nfs4_client *clp) +int setup_callback_client(struct nfs4_client *clp) { struct sockaddr_in addr; struct nfs4_callback *cb = &clp->cl_callback; @@ -389,7 +389,7 @@ static struct rpc_clnt *setup_callback_client(struct nfs4_client *clp) struct rpc_clnt *client; if (!clp->cl_principal && (clp->cl_flavor >= RPC_AUTH_GSS_KRB5)) - return ERR_PTR(-EINVAL); + return -EINVAL; /* Initialize address */ memset(&addr, 0, sizeof(addr)); @@ -399,10 +399,13 @@ static struct rpc_clnt *setup_callback_client(struct nfs4_client *clp) /* Create RPC client */ client = rpc_create(&args); - if (IS_ERR(client)) + if (IS_ERR(client)) { dprintk("NFSD: couldn't create callback client: %ld\n", PTR_ERR(client)); - return client; + return PTR_ERR(client); + } + cb->cb_client = client; + return 0; } @@ -414,28 +417,23 @@ static int do_probe_callback(void *data) .rpc_proc = &nfs4_cb_procedures[NFSPROC4_CLNT_CB_NULL], .rpc_argp = clp, }; - struct rpc_clnt *client; int status; - client = setup_callback_client(clp); - if (IS_ERR(client)) { - status = PTR_ERR(client); - dprintk("NFSD: couldn't create callback client: %d\n", - status); + status = setup_callback_client(clp); + if (status) goto out_err; - } - status = rpc_call_sync(client, &msg, RPC_TASK_SOFT); + status = rpc_call_sync(cb->cb_client, &msg, RPC_TASK_SOFT); if (status) goto out_release_client; - cb->cb_client = client; atomic_set(&cb->cb_set, 1); put_nfs4_client(clp); return 0; out_release_client: - rpc_shutdown_client(client); + rpc_shutdown_client(cb->cb_client); + cb->cb_client = NULL; out_err: dprintk("NFSD: warning: no callback path to client %.*s: error %d\n", (int)clp->cl_name.len, clp->cl_name.data, status); -- cgit v1.2.3 From ecdd03b7914c91ef849e49c4d466c87f4981b5cd Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Mon, 23 Feb 2009 19:35:22 -0800 Subject: nfsd4: create rpc callback client from server thread The code is a little simpler, and it should be easier to avoid races, if we just do all rpc client creation/destruction from nfsd or laundromat threads and do only the rpc calls themselves asynchronously. The rpc creation doesn't involve any significant waiting (it doesn't call the client, for example), so there's no reason not to do this. Also don't bother destroying the client on failure of the rpc null probe. We may want to retry the probe later anyway. Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4callback.c | 31 ++++++++++++++++--------------- 1 file changed, 16 insertions(+), 15 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c index 4788d09d9bec..711c6282151f 100644 --- a/fs/nfsd/nfs4callback.c +++ b/fs/nfsd/nfs4callback.c @@ -409,6 +409,12 @@ int setup_callback_client(struct nfs4_client *clp) } +static void warn_no_callback_path(struct nfs4_client *clp, int reason) +{ + dprintk("NFSD: warning: no callback path to client %.*s: error %d\n", + (int)clp->cl_name.len, clp->cl_name.data, reason); +} + static int do_probe_callback(void *data) { struct nfs4_client *clp = data; @@ -419,24 +425,12 @@ static int do_probe_callback(void *data) }; int status; - status = setup_callback_client(clp); - if (status) - goto out_err; - status = rpc_call_sync(cb->cb_client, &msg, RPC_TASK_SOFT); - if (status) - goto out_release_client; + warn_no_callback_path(clp, status); + else + atomic_set(&cb->cb_set, 1); - atomic_set(&cb->cb_set, 1); - put_nfs4_client(clp); - return 0; -out_release_client: - rpc_shutdown_client(cb->cb_client); - cb->cb_client = NULL; -out_err: - dprintk("NFSD: warning: no callback path to client %.*s: error %d\n", - (int)clp->cl_name.len, clp->cl_name.data, status); put_nfs4_client(clp); return 0; } @@ -448,9 +442,16 @@ void nfsd4_probe_callback(struct nfs4_client *clp) { struct task_struct *t; + int status; BUG_ON(atomic_read(&clp->cl_callback.cb_set)); + status = setup_callback_client(clp); + if (status) { + warn_no_callback_path(clp, status); + return; + } + /* the task holds a reference to the nfs4_client struct */ atomic_inc(&clp->cl_count); -- cgit v1.2.3 From 3cef9ab266a932899e756f7e1ea7a988a97bf3b2 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Mon, 23 Feb 2009 21:42:10 -0800 Subject: nfsd4: lookup up callback cred only once Lookup the callback cred once and then use it for all subsequent callbacks. Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4callback.c | 26 ++++++++++++++++++++++++++ fs/nfsd/nfs4state.c | 4 ++++ include/linux/nfsd/state.h | 1 + 3 files changed, 31 insertions(+) (limited to 'fs') diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c index 711c6282151f..cc10ed35ac81 100644 --- a/fs/nfsd/nfs4callback.c +++ b/fs/nfsd/nfs4callback.c @@ -415,6 +415,22 @@ static void warn_no_callback_path(struct nfs4_client *clp, int reason) (int)clp->cl_name.len, clp->cl_name.data, reason); } +static struct rpc_cred *lookup_cb_cred(struct nfs4_callback *cb) +{ + struct auth_cred acred = { + .machine_cred = 1 + }; + + /* + * Note in the gss case this doesn't actually have to wait for a + * gss upcall (or any calls to the client); this just creates a + * non-uptodate cred which the rpc state machine will fill in with + * a refresh_upcall later. + */ + return rpcauth_lookup_credcache(cb->cb_client->cl_auth, &acred, + RPCAUTH_LOOKUP_NEW); +} + static int do_probe_callback(void *data) { struct nfs4_client *clp = data; @@ -423,9 +439,18 @@ static int do_probe_callback(void *data) .rpc_proc = &nfs4_cb_procedures[NFSPROC4_CLNT_CB_NULL], .rpc_argp = clp, }; + struct rpc_cred *cred; int status; + cred = lookup_cb_cred(cb); + if (IS_ERR(cred)) { + status = PTR_ERR(cred); + goto out; + } + cb->cb_cred = cred; + msg.rpc_cred = cb->cb_cred; status = rpc_call_sync(cb->cb_client, &msg, RPC_TASK_SOFT); +out: if (status) warn_no_callback_path(clp, status); else @@ -475,6 +500,7 @@ nfsd4_cb_recall(struct nfs4_delegation *dp) struct rpc_message msg = { .rpc_proc = &nfs4_cb_procedures[NFSPROC4_CLNT_CB_RECALL], .rpc_argp = cbr, + .rpc_cred = clp->cl_callback.cb_cred }; int retries = 1; int status = 0; diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 7e1fcc3aade4..b205c7d7bc6a 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -643,6 +643,10 @@ shutdown_callback_client(struct nfs4_client *clp) clp->cl_callback.cb_client = NULL; rpc_shutdown_client(clnt); } + if (clp->cl_callback.cb_cred) { + put_rpccred(clp->cl_callback.cb_cred); + clp->cl_callback.cb_cred = NULL; + } } static inline void diff --git a/include/linux/nfsd/state.h b/include/linux/nfsd/state.h index 4d61c873feed..8d882a3eb4b9 100644 --- a/include/linux/nfsd/state.h +++ b/include/linux/nfsd/state.h @@ -97,6 +97,7 @@ struct nfs4_callback { /* RPC client info */ atomic_t cb_set; /* successful CB_NULL call */ struct rpc_clnt * cb_client; + struct rpc_cred * cb_cred; }; /* Maximum number of slots per session. 128 is useful for long haul TCP */ -- cgit v1.2.3 From e300a63ce4ccec073d254d883a3638d5dca1d771 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Thu, 5 Mar 2009 15:01:11 -0500 Subject: nfsd4: replace callback thread by asynchronous rpc We don't really need a synchronous rpc, and moving to an asynchronous rpc allows us to do without this extra kthread. Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4callback.c | 38 ++++++++++++++++++++++---------------- 1 file changed, 22 insertions(+), 16 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c index cc10ed35ac81..0aaf68beedbd 100644 --- a/fs/nfsd/nfs4callback.c +++ b/fs/nfsd/nfs4callback.c @@ -415,6 +415,21 @@ static void warn_no_callback_path(struct nfs4_client *clp, int reason) (int)clp->cl_name.len, clp->cl_name.data, reason); } +static void nfsd4_cb_probe_done(struct rpc_task *task, void *calldata) +{ + struct nfs4_client *clp = calldata; + + if (task->tk_status) + warn_no_callback_path(clp, task->tk_status); + else + atomic_set(&clp->cl_callback.cb_set, 1); + put_nfs4_client(clp); +} + +static const struct rpc_call_ops nfsd4_cb_probe_ops = { + .rpc_call_done = nfsd4_cb_probe_done, +}; + static struct rpc_cred *lookup_cb_cred(struct nfs4_callback *cb) { struct auth_cred acred = { @@ -431,9 +446,8 @@ static struct rpc_cred *lookup_cb_cred(struct nfs4_callback *cb) RPCAUTH_LOOKUP_NEW); } -static int do_probe_callback(void *data) +void do_probe_callback(struct nfs4_client *clp) { - struct nfs4_client *clp = data; struct nfs4_callback *cb = &clp->cl_callback; struct rpc_message msg = { .rpc_proc = &nfs4_cb_procedures[NFSPROC4_CLNT_CB_NULL], @@ -449,15 +463,13 @@ static int do_probe_callback(void *data) } cb->cb_cred = cred; msg.rpc_cred = cb->cb_cred; - status = rpc_call_sync(cb->cb_client, &msg, RPC_TASK_SOFT); + status = rpc_call_async(cb->cb_client, &msg, RPC_TASK_SOFT, + &nfsd4_cb_probe_ops, (void *)clp); out: - if (status) + if (status) { warn_no_callback_path(clp, status); - else - atomic_set(&cb->cb_set, 1); - - put_nfs4_client(clp); - return 0; + put_nfs4_client(clp); + } } /* @@ -466,7 +478,6 @@ out: void nfsd4_probe_callback(struct nfs4_client *clp) { - struct task_struct *t; int status; BUG_ON(atomic_read(&clp->cl_callback.cb_set)); @@ -480,12 +491,7 @@ nfsd4_probe_callback(struct nfs4_client *clp) /* the task holds a reference to the nfs4_client struct */ atomic_inc(&clp->cl_count); - t = kthread_run(do_probe_callback, clp, "nfs4_cb_probe"); - - if (IS_ERR(t)) - atomic_dec(&clp->cl_count); - - return; + do_probe_callback(clp); } /* -- cgit v1.2.3 From c237dc0303bcf6f4cc2e0efe4fe4e341c6f34dac Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Wed, 29 Apr 2009 19:09:19 -0400 Subject: nfsd4: rename callback struct to cb_conn I want to use the name for a struct that actually does represent a single callback. (Actually, I've never been sure it helps to a separate struct for the callback information. Some day maybe those fields could just be dumped into struct nfs4_client. I don't know.) Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4callback.c | 16 ++++++++-------- fs/nfsd/nfs4state.c | 22 +++++++++++----------- include/linux/nfsd/state.h | 4 ++-- 3 files changed, 21 insertions(+), 21 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c index 0aaf68beedbd..ed860d7ddd19 100644 --- a/fs/nfsd/nfs4callback.c +++ b/fs/nfsd/nfs4callback.c @@ -369,7 +369,7 @@ static int max_cb_time(void) int setup_callback_client(struct nfs4_client *clp) { struct sockaddr_in addr; - struct nfs4_callback *cb = &clp->cl_callback; + struct nfs4_cb_conn *cb = &clp->cl_cb_conn; struct rpc_timeout timeparms = { .to_initval = max_cb_time(), .to_retries = 0, @@ -422,7 +422,7 @@ static void nfsd4_cb_probe_done(struct rpc_task *task, void *calldata) if (task->tk_status) warn_no_callback_path(clp, task->tk_status); else - atomic_set(&clp->cl_callback.cb_set, 1); + atomic_set(&clp->cl_cb_conn.cb_set, 1); put_nfs4_client(clp); } @@ -430,7 +430,7 @@ static const struct rpc_call_ops nfsd4_cb_probe_ops = { .rpc_call_done = nfsd4_cb_probe_done, }; -static struct rpc_cred *lookup_cb_cred(struct nfs4_callback *cb) +static struct rpc_cred *lookup_cb_cred(struct nfs4_cb_conn *cb) { struct auth_cred acred = { .machine_cred = 1 @@ -448,7 +448,7 @@ static struct rpc_cred *lookup_cb_cred(struct nfs4_callback *cb) void do_probe_callback(struct nfs4_client *clp) { - struct nfs4_callback *cb = &clp->cl_callback; + struct nfs4_cb_conn *cb = &clp->cl_cb_conn; struct rpc_message msg = { .rpc_proc = &nfs4_cb_procedures[NFSPROC4_CLNT_CB_NULL], .rpc_argp = clp, @@ -480,7 +480,7 @@ nfsd4_probe_callback(struct nfs4_client *clp) { int status; - BUG_ON(atomic_read(&clp->cl_callback.cb_set)); + BUG_ON(atomic_read(&clp->cl_cb_conn.cb_set)); status = setup_callback_client(clp); if (status) { @@ -501,12 +501,12 @@ void nfsd4_cb_recall(struct nfs4_delegation *dp) { struct nfs4_client *clp = dp->dl_client; - struct rpc_clnt *clnt = clp->cl_callback.cb_client; + struct rpc_clnt *clnt = clp->cl_cb_conn.cb_client; struct nfs4_cb_recall *cbr = &dp->dl_recall; struct rpc_message msg = { .rpc_proc = &nfs4_cb_procedures[NFSPROC4_CLNT_CB_RECALL], .rpc_argp = cbr, - .rpc_cred = clp->cl_callback.cb_cred + .rpc_cred = clp->cl_cb_conn.cb_cred }; int retries = 1; int status = 0; @@ -519,7 +519,7 @@ nfsd4_cb_recall(struct nfs4_delegation *dp) switch (status) { case -EIO: /* Network partition? */ - atomic_set(&clp->cl_callback.cb_set, 0); + atomic_set(&clp->cl_cb_conn.cb_set, 0); case -EBADHANDLE: case -NFS4ERR_BAD_STATEID: /* Race: client probably got cb_recall diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index b205c7d7bc6a..d7b5e6b89207 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -182,7 +182,7 @@ alloc_init_deleg(struct nfs4_client *clp, struct nfs4_stateid *stp, struct svc_f { struct nfs4_delegation *dp; struct nfs4_file *fp = stp->st_file; - struct nfs4_callback *cb = &stp->st_stateowner->so_client->cl_callback; + struct nfs4_cb_conn *cb = &stp->st_stateowner->so_client->cl_cb_conn; dprintk("NFSD alloc_init_deleg\n"); if (fp->fi_had_conflict) @@ -633,19 +633,19 @@ static struct nfs4_client *alloc_client(struct xdr_netobj name) static void shutdown_callback_client(struct nfs4_client *clp) { - struct rpc_clnt *clnt = clp->cl_callback.cb_client; + struct rpc_clnt *clnt = clp->cl_cb_conn.cb_client; if (clnt) { /* * Callback threads take a reference on the client, so there * should be no outstanding callbacks at this point. */ - clp->cl_callback.cb_client = NULL; + clp->cl_cb_conn.cb_client = NULL; rpc_shutdown_client(clnt); } - if (clp->cl_callback.cb_cred) { - put_rpccred(clp->cl_callback.cb_cred); - clp->cl_callback.cb_cred = NULL; + if (clp->cl_cb_conn.cb_cred) { + put_rpccred(clp->cl_cb_conn.cb_cred); + clp->cl_cb_conn.cb_cred = NULL; } } @@ -719,7 +719,7 @@ static struct nfs4_client *create_client(struct xdr_netobj name, char *recdir) return NULL; memcpy(clp->cl_recdir, recdir, HEXDIR_LEN); atomic_set(&clp->cl_count, 1); - atomic_set(&clp->cl_callback.cb_set, 0); + atomic_set(&clp->cl_cb_conn.cb_set, 0); INIT_LIST_HEAD(&clp->cl_idhash); INIT_LIST_HEAD(&clp->cl_strhash); INIT_LIST_HEAD(&clp->cl_openowners); @@ -971,7 +971,7 @@ parse_ipv4(unsigned int addr_len, char *addr_val, unsigned int *cbaddrp, unsigne static void gen_callback(struct nfs4_client *clp, struct nfsd4_setclientid *se) { - struct nfs4_callback *cb = &clp->cl_callback; + struct nfs4_cb_conn *cb = &clp->cl_cb_conn; /* Currently, we only support tcp for the callback channel */ if ((se->se_callback_netid_len != 3) || memcmp((char *)se->se_callback_netid_val, "tcp", 3)) @@ -1691,7 +1691,7 @@ nfsd4_setclientid_confirm(struct svc_rqst *rqstp, else { /* XXX: We just turn off callbacks until we can handle * change request correctly. */ - atomic_set(&conf->cl_callback.cb_set, 0); + atomic_set(&conf->cl_cb_conn.cb_set, 0); expire_client(unconf); status = nfs_ok; @@ -2425,7 +2425,7 @@ nfs4_open_delegation(struct svc_fh *fh, struct nfsd4_open *open, struct nfs4_sta { struct nfs4_delegation *dp; struct nfs4_stateowner *sop = stp->st_stateowner; - struct nfs4_callback *cb = &sop->so_client->cl_callback; + struct nfs4_cb_conn *cb = &sop->so_client->cl_cb_conn; struct file_lock fl, *flp = &fl; int status, flag = 0; @@ -2617,7 +2617,7 @@ nfsd4_renew(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, renew_client(clp); status = nfserr_cb_path_down; if (!list_empty(&clp->cl_delegations) - && !atomic_read(&clp->cl_callback.cb_set)) + && !atomic_read(&clp->cl_cb_conn.cb_set)) goto out; status = nfs_ok; out: diff --git a/include/linux/nfsd/state.h b/include/linux/nfsd/state.h index 8d882a3eb4b9..563c367a3013 100644 --- a/include/linux/nfsd/state.h +++ b/include/linux/nfsd/state.h @@ -88,7 +88,7 @@ struct nfs4_delegation { #define dl_fh dl_recall.cbr_fh /* client delegation callback info */ -struct nfs4_callback { +struct nfs4_cb_conn { /* SETCLIENTID info */ u32 cb_addr; unsigned short cb_port; @@ -186,7 +186,7 @@ struct nfs4_client { struct svc_cred cl_cred; /* setclientid principal */ clientid_t cl_clientid; /* generated by server */ nfs4_verifier cl_confirm; /* generated by server */ - struct nfs4_callback cl_callback; /* callback info */ + struct nfs4_cb_conn cl_cb_conn; /* callback info */ atomic_t cl_count; /* ref count */ u32 cl_firststate; /* recovery dir creation */ -- cgit v1.2.3 From b53d40c5070bffde1b2bcaf848412a50d8894794 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Fri, 1 May 2009 19:50:00 -0400 Subject: nfsd4: eliminate struct nfs4_cb_recall The nfs4_cb_recall struct is used only in nfs4_delegation, so its pointer to the containing delegation is unnecessary--we could just use container_of(). But there's no real reason to have this a separate struct at all--just move these fields to nfs4_delegation. Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4callback.c | 24 +++++++++++------------- fs/nfsd/nfs4state.c | 5 ++--- include/linux/nfsd/state.h | 18 +++++------------- 3 files changed, 18 insertions(+), 29 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c index ed860d7ddd19..2509305f6f53 100644 --- a/fs/nfsd/nfs4callback.c +++ b/fs/nfsd/nfs4callback.c @@ -215,18 +215,18 @@ encode_cb_compound_hdr(struct xdr_stream *xdr, struct nfs4_cb_compound_hdr *hdr) } static int -encode_cb_recall(struct xdr_stream *xdr, struct nfs4_cb_recall *cb_rec) +encode_cb_recall(struct xdr_stream *xdr, struct nfs4_delegation *dp) { __be32 *p; - int len = cb_rec->cbr_fh.fh_size; + int len = dp->dl_fh.fh_size; - RESERVE_SPACE(12+sizeof(cb_rec->cbr_stateid) + len); + RESERVE_SPACE(12+sizeof(dp->dl_stateid) + len); WRITE32(OP_CB_RECALL); - WRITE32(cb_rec->cbr_stateid.si_generation); - WRITEMEM(&cb_rec->cbr_stateid.si_opaque, sizeof(stateid_opaque_t)); - WRITE32(cb_rec->cbr_trunc); + WRITE32(dp->dl_stateid.si_generation); + WRITEMEM(&dp->dl_stateid.si_opaque, sizeof(stateid_opaque_t)); + WRITE32(dp->dl_trunc); WRITE32(len); - WRITEMEM(&cb_rec->cbr_fh.fh_base, len); + WRITEMEM(&dp->dl_fh.fh_base, len); return 0; } @@ -241,11 +241,11 @@ nfs4_xdr_enc_cb_null(struct rpc_rqst *req, __be32 *p) } static int -nfs4_xdr_enc_cb_recall(struct rpc_rqst *req, __be32 *p, struct nfs4_cb_recall *args) +nfs4_xdr_enc_cb_recall(struct rpc_rqst *req, __be32 *p, struct nfs4_delegation *args) { struct xdr_stream xdr; struct nfs4_cb_compound_hdr hdr = { - .ident = args->cbr_ident, + .ident = args->dl_ident, .nops = 1, }; @@ -502,17 +502,15 @@ nfsd4_cb_recall(struct nfs4_delegation *dp) { struct nfs4_client *clp = dp->dl_client; struct rpc_clnt *clnt = clp->cl_cb_conn.cb_client; - struct nfs4_cb_recall *cbr = &dp->dl_recall; struct rpc_message msg = { .rpc_proc = &nfs4_cb_procedures[NFSPROC4_CLNT_CB_RECALL], - .rpc_argp = cbr, + .rpc_argp = dp, .rpc_cred = clp->cl_cb_conn.cb_cred }; int retries = 1; int status = 0; - cbr->cbr_trunc = 0; /* XXX need to implement truncate optimization */ - cbr->cbr_dp = dp; + dp->dl_trunc = 0; /* XXX need to implement truncate optimization */ status = rpc_call_sync(clnt, &msg, RPC_TASK_SOFT); while (retries--) { diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index d7b5e6b89207..3e5345e01b13 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -203,9 +203,8 @@ alloc_init_deleg(struct nfs4_client *clp, struct nfs4_stateid *stp, struct svc_f get_file(stp->st_vfs_file); dp->dl_vfs_file = stp->st_vfs_file; dp->dl_type = type; - dp->dl_recall.cbr_dp = NULL; - dp->dl_recall.cbr_ident = cb->cb_ident; - dp->dl_recall.cbr_trunc = 0; + dp->dl_ident = cb->cb_ident; + dp->dl_trunc = 0; dp->dl_stateid.si_boot = get_seconds(); dp->dl_stateid.si_stateownerid = current_delegid++; dp->dl_stateid.si_fileid = 0; diff --git a/include/linux/nfsd/state.h b/include/linux/nfsd/state.h index 563c367a3013..233b60d39b84 100644 --- a/include/linux/nfsd/state.h +++ b/include/linux/nfsd/state.h @@ -61,15 +61,6 @@ typedef struct { #define si_stateownerid si_opaque.so_stateownerid #define si_fileid si_opaque.so_fileid - -struct nfs4_cb_recall { - u32 cbr_ident; - int cbr_trunc; - stateid_t cbr_stateid; - struct knfsd_fh cbr_fh; - struct nfs4_delegation *cbr_dp; -}; - struct nfs4_delegation { struct list_head dl_perfile; struct list_head dl_perclnt; @@ -81,12 +72,13 @@ struct nfs4_delegation { struct file *dl_vfs_file; u32 dl_type; time_t dl_time; - struct nfs4_cb_recall dl_recall; +/* For recall: */ + u32 dl_ident; + int dl_trunc; + stateid_t dl_stateid; + struct knfsd_fh dl_fh; }; -#define dl_stateid dl_recall.cbr_stateid -#define dl_fh dl_recall.cbr_fh - /* client delegation callback info */ struct nfs4_cb_conn { /* SETCLIENTID info */ -- cgit v1.2.3 From 6707bd3d420f53ae8f090dac871843f6f43c9980 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Fri, 1 May 2009 19:57:46 -0400 Subject: nfsd4: remove unused dl_trunc There's no point in keeping this field around--it's always zero. (Background: the protocol allows you to tell the client that the file is about to be truncated, as an optimization to save the client from writing back dirty pages that will just be discarded. We don't implement this hint. If we do some day, adding this field back in will be the least of the work involved.) Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4callback.c | 4 +--- fs/nfsd/nfs4state.c | 1 - include/linux/nfsd/state.h | 1 - 3 files changed, 1 insertion(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c index 2509305f6f53..0420b5e6e20d 100644 --- a/fs/nfsd/nfs4callback.c +++ b/fs/nfsd/nfs4callback.c @@ -224,7 +224,7 @@ encode_cb_recall(struct xdr_stream *xdr, struct nfs4_delegation *dp) WRITE32(OP_CB_RECALL); WRITE32(dp->dl_stateid.si_generation); WRITEMEM(&dp->dl_stateid.si_opaque, sizeof(stateid_opaque_t)); - WRITE32(dp->dl_trunc); + WRITE32(0); /* truncate optimization not implemented */ WRITE32(len); WRITEMEM(&dp->dl_fh.fh_base, len); return 0; @@ -510,8 +510,6 @@ nfsd4_cb_recall(struct nfs4_delegation *dp) int retries = 1; int status = 0; - dp->dl_trunc = 0; /* XXX need to implement truncate optimization */ - status = rpc_call_sync(clnt, &msg, RPC_TASK_SOFT); while (retries--) { switch (status) { diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 3e5345e01b13..cbb16e191d5b 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -204,7 +204,6 @@ alloc_init_deleg(struct nfs4_client *clp, struct nfs4_stateid *stp, struct svc_f dp->dl_vfs_file = stp->st_vfs_file; dp->dl_type = type; dp->dl_ident = cb->cb_ident; - dp->dl_trunc = 0; dp->dl_stateid.si_boot = get_seconds(); dp->dl_stateid.si_stateownerid = current_delegid++; dp->dl_stateid.si_fileid = 0; diff --git a/include/linux/nfsd/state.h b/include/linux/nfsd/state.h index 233b60d39b84..346b603072ce 100644 --- a/include/linux/nfsd/state.h +++ b/include/linux/nfsd/state.h @@ -74,7 +74,6 @@ struct nfs4_delegation { time_t dl_time; /* For recall: */ u32 dl_ident; - int dl_trunc; stateid_t dl_stateid; struct knfsd_fh dl_fh; }; -- cgit v1.2.3 From 3aea09dc9106407d8bc18e593fbffda9ad632844 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Fri, 1 May 2009 20:11:12 -0400 Subject: nfsd4: track recall retries in nfs4_delegation Move this out of a local variable into the nfs4_delegation object in preparation for making this an async rpc call (at which point we'll need any state like this in a common object that's preserved across function calls). Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4callback.c | 4 ++-- include/linux/nfsd/state.h | 1 + 2 files changed, 3 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c index 0420b5e6e20d..b88b207d75d9 100644 --- a/fs/nfsd/nfs4callback.c +++ b/fs/nfsd/nfs4callback.c @@ -507,11 +507,11 @@ nfsd4_cb_recall(struct nfs4_delegation *dp) .rpc_argp = dp, .rpc_cred = clp->cl_cb_conn.cb_cred }; - int retries = 1; int status = 0; + dp->dl_retries = 1; status = rpc_call_sync(clnt, &msg, RPC_TASK_SOFT); - while (retries--) { + while (dp->dl_retries--) { switch (status) { case -EIO: /* Network partition? */ diff --git a/include/linux/nfsd/state.h b/include/linux/nfsd/state.h index 346b603072ce..c0c49215ddc5 100644 --- a/include/linux/nfsd/state.h +++ b/include/linux/nfsd/state.h @@ -76,6 +76,7 @@ struct nfs4_delegation { u32 dl_ident; stateid_t dl_stateid; struct knfsd_fh dl_fh; + int dl_retries; }; /* client delegation callback info */ -- cgit v1.2.3 From 63e4863fabc6e165a6ca813051305be58966da45 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Fri, 1 May 2009 22:36:55 -0400 Subject: nfsd4: make recall callback an asynchronous rpc As with the probe, this removes the need for another kthread. Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4callback.c | 74 +++++++++++++++++++++++++++++++++----------------- fs/nfsd/nfs4state.c | 28 ++----------------- 2 files changed, 52 insertions(+), 50 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c index b88b207d75d9..f4fab69a8c30 100644 --- a/fs/nfsd/nfs4callback.c +++ b/fs/nfsd/nfs4callback.c @@ -494,6 +494,49 @@ nfsd4_probe_callback(struct nfs4_client *clp) do_probe_callback(clp); } +static void nfsd4_cb_recall_done(struct rpc_task *task, void *calldata) +{ + struct nfs4_delegation *dp = calldata; + struct nfs4_client *clp = dp->dl_client; + + switch (task->tk_status) { + case -EIO: + /* Network partition? */ + atomic_set(&clp->cl_cb_conn.cb_set, 0); + warn_no_callback_path(clp, task->tk_status); + case -EBADHANDLE: + case -NFS4ERR_BAD_STATEID: + /* Race: client probably got cb_recall + * before open reply granting delegation */ + break; + default: + /* success, or error we can't handle */ + return; + } + if (dp->dl_retries--) { + rpc_delay(task, 2*HZ); + task->tk_status = 0; + rpc_restart_call(task); + } else { + atomic_set(&clp->cl_cb_conn.cb_set, 0); + warn_no_callback_path(clp, task->tk_status); + } +} + +static void nfsd4_cb_recall_release(void *calldata) +{ + struct nfs4_delegation *dp = calldata; + struct nfs4_client *clp = dp->dl_client; + + nfs4_put_delegation(dp); + put_nfs4_client(clp); +} + +static const struct rpc_call_ops nfsd4_cb_recall_ops = { + .rpc_call_done = nfsd4_cb_recall_done, + .rpc_release = nfsd4_cb_recall_release, +}; + /* * called with dp->dl_count inc'ed. */ @@ -507,32 +550,13 @@ nfsd4_cb_recall(struct nfs4_delegation *dp) .rpc_argp = dp, .rpc_cred = clp->cl_cb_conn.cb_cred }; - int status = 0; + int status; dp->dl_retries = 1; - status = rpc_call_sync(clnt, &msg, RPC_TASK_SOFT); - while (dp->dl_retries--) { - switch (status) { - case -EIO: - /* Network partition? */ - atomic_set(&clp->cl_cb_conn.cb_set, 0); - case -EBADHANDLE: - case -NFS4ERR_BAD_STATEID: - /* Race: client probably got cb_recall - * before open reply granting delegation */ - break; - default: - goto out_put_cred; - } - ssleep(2); - status = rpc_call_sync(clnt, &msg, RPC_TASK_SOFT); + status = rpc_call_async(clnt, &msg, RPC_TASK_SOFT, + &nfsd4_cb_recall_ops, dp); + if (status) { + put_nfs4_client(clp); + nfs4_put_delegation(dp); } -out_put_cred: - /* - * Success or failure, now we're either waiting for lease expiration - * or deleg_return. - */ - put_nfs4_client(clp); - nfs4_put_delegation(dp); - return; } diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index cbb16e191d5b..a4bdf2589b41 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -2059,19 +2059,6 @@ nfs4_file_downgrade(struct file *filp, unsigned int share_access) } } -/* - * Recall a delegation - */ -static int -do_recall(void *__dp) -{ - struct nfs4_delegation *dp = __dp; - - dp->dl_file->fi_had_conflict = true; - nfsd4_cb_recall(dp); - return 0; -} - /* * Spawn a thread to perform a recall on the delegation represented * by the lease (file_lock) @@ -2083,8 +2070,7 @@ do_recall(void *__dp) static void nfsd_break_deleg_cb(struct file_lock *fl) { - struct nfs4_delegation *dp= (struct nfs4_delegation *)fl->fl_owner; - struct task_struct *t; + struct nfs4_delegation *dp = (struct nfs4_delegation *)fl->fl_owner; dprintk("NFSD nfsd_break_deleg_cb: dp %p fl %p\n",dp,fl); if (!dp) @@ -2112,16 +2098,8 @@ void nfsd_break_deleg_cb(struct file_lock *fl) */ fl->fl_break_time = 0; - t = kthread_run(do_recall, dp, "%s", "nfs4_cb_recall"); - if (IS_ERR(t)) { - struct nfs4_client *clp = dp->dl_client; - - printk(KERN_INFO "NFSD: Callback thread failed for " - "for client (clientid %08x/%08x)\n", - clp->cl_clientid.cl_boot, clp->cl_clientid.cl_id); - put_nfs4_client(dp->dl_client); - nfs4_put_delegation(dp); - } + dp->dl_file->fi_had_conflict = true; + nfsd4_cb_recall(dp); } /* -- cgit v1.2.3 From 9064caae8f47bb7ed5d91712e81f01c1aba2fa3c Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Tue, 28 Apr 2009 16:48:25 -0700 Subject: nfsd: use C99 struct initializers Eliminate 56 sparse warnings like this one: fs/nfsd/nfs4xdr.c:1331:15: warning: obsolete array initializer, use C99 syntax Signed-off-by: Randy Dunlap Cc: Neil Brown Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4xdr.c | 112 +++++++++++++++++++++++++++--------------------------- 1 file changed, 56 insertions(+), 56 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c index 12d36a7361cd..ab005fc637e1 100644 --- a/fs/nfsd/nfs4xdr.c +++ b/fs/nfsd/nfs4xdr.c @@ -1318,64 +1318,64 @@ static nfsd4_dec nfsd4_dec_ops[] = { }; static nfsd4_dec nfsd41_dec_ops[] = { - [OP_ACCESS] (nfsd4_dec)nfsd4_decode_access, - [OP_CLOSE] (nfsd4_dec)nfsd4_decode_close, - [OP_COMMIT] (nfsd4_dec)nfsd4_decode_commit, - [OP_CREATE] (nfsd4_dec)nfsd4_decode_create, - [OP_DELEGPURGE] (nfsd4_dec)nfsd4_decode_notsupp, - [OP_DELEGRETURN] (nfsd4_dec)nfsd4_decode_delegreturn, - [OP_GETATTR] (nfsd4_dec)nfsd4_decode_getattr, - [OP_GETFH] (nfsd4_dec)nfsd4_decode_noop, - [OP_LINK] (nfsd4_dec)nfsd4_decode_link, - [OP_LOCK] (nfsd4_dec)nfsd4_decode_lock, - [OP_LOCKT] (nfsd4_dec)nfsd4_decode_lockt, - [OP_LOCKU] (nfsd4_dec)nfsd4_decode_locku, - [OP_LOOKUP] (nfsd4_dec)nfsd4_decode_lookup, - [OP_LOOKUPP] (nfsd4_dec)nfsd4_decode_noop, - [OP_NVERIFY] (nfsd4_dec)nfsd4_decode_verify, - [OP_OPEN] (nfsd4_dec)nfsd4_decode_open, - [OP_OPENATTR] (nfsd4_dec)nfsd4_decode_notsupp, - [OP_OPEN_CONFIRM] (nfsd4_dec)nfsd4_decode_notsupp, - [OP_OPEN_DOWNGRADE] (nfsd4_dec)nfsd4_decode_open_downgrade, - [OP_PUTFH] (nfsd4_dec)nfsd4_decode_putfh, - [OP_PUTPUBFH] (nfsd4_dec)nfsd4_decode_notsupp, - [OP_PUTROOTFH] (nfsd4_dec)nfsd4_decode_noop, - [OP_READ] (nfsd4_dec)nfsd4_decode_read, - [OP_READDIR] (nfsd4_dec)nfsd4_decode_readdir, - [OP_READLINK] (nfsd4_dec)nfsd4_decode_noop, - [OP_REMOVE] (nfsd4_dec)nfsd4_decode_remove, - [OP_RENAME] (nfsd4_dec)nfsd4_decode_rename, - [OP_RENEW] (nfsd4_dec)nfsd4_decode_notsupp, - [OP_RESTOREFH] (nfsd4_dec)nfsd4_decode_noop, - [OP_SAVEFH] (nfsd4_dec)nfsd4_decode_noop, - [OP_SECINFO] (nfsd4_dec)nfsd4_decode_secinfo, - [OP_SETATTR] (nfsd4_dec)nfsd4_decode_setattr, - [OP_SETCLIENTID] (nfsd4_dec)nfsd4_decode_notsupp, - [OP_SETCLIENTID_CONFIRM](nfsd4_dec)nfsd4_decode_notsupp, - [OP_VERIFY] (nfsd4_dec)nfsd4_decode_verify, - [OP_WRITE] (nfsd4_dec)nfsd4_decode_write, - [OP_RELEASE_LOCKOWNER] (nfsd4_dec)nfsd4_decode_notsupp, + [OP_ACCESS] = (nfsd4_dec)nfsd4_decode_access, + [OP_CLOSE] = (nfsd4_dec)nfsd4_decode_close, + [OP_COMMIT] = (nfsd4_dec)nfsd4_decode_commit, + [OP_CREATE] = (nfsd4_dec)nfsd4_decode_create, + [OP_DELEGPURGE] = (nfsd4_dec)nfsd4_decode_notsupp, + [OP_DELEGRETURN] = (nfsd4_dec)nfsd4_decode_delegreturn, + [OP_GETATTR] = (nfsd4_dec)nfsd4_decode_getattr, + [OP_GETFH] = (nfsd4_dec)nfsd4_decode_noop, + [OP_LINK] = (nfsd4_dec)nfsd4_decode_link, + [OP_LOCK] = (nfsd4_dec)nfsd4_decode_lock, + [OP_LOCKT] = (nfsd4_dec)nfsd4_decode_lockt, + [OP_LOCKU] = (nfsd4_dec)nfsd4_decode_locku, + [OP_LOOKUP] = (nfsd4_dec)nfsd4_decode_lookup, + [OP_LOOKUPP] = (nfsd4_dec)nfsd4_decode_noop, + [OP_NVERIFY] = (nfsd4_dec)nfsd4_decode_verify, + [OP_OPEN] = (nfsd4_dec)nfsd4_decode_open, + [OP_OPENATTR] = (nfsd4_dec)nfsd4_decode_notsupp, + [OP_OPEN_CONFIRM] = (nfsd4_dec)nfsd4_decode_notsupp, + [OP_OPEN_DOWNGRADE] = (nfsd4_dec)nfsd4_decode_open_downgrade, + [OP_PUTFH] = (nfsd4_dec)nfsd4_decode_putfh, + [OP_PUTPUBFH] = (nfsd4_dec)nfsd4_decode_notsupp, + [OP_PUTROOTFH] = (nfsd4_dec)nfsd4_decode_noop, + [OP_READ] = (nfsd4_dec)nfsd4_decode_read, + [OP_READDIR] = (nfsd4_dec)nfsd4_decode_readdir, + [OP_READLINK] = (nfsd4_dec)nfsd4_decode_noop, + [OP_REMOVE] = (nfsd4_dec)nfsd4_decode_remove, + [OP_RENAME] = (nfsd4_dec)nfsd4_decode_rename, + [OP_RENEW] = (nfsd4_dec)nfsd4_decode_notsupp, + [OP_RESTOREFH] = (nfsd4_dec)nfsd4_decode_noop, + [OP_SAVEFH] = (nfsd4_dec)nfsd4_decode_noop, + [OP_SECINFO] = (nfsd4_dec)nfsd4_decode_secinfo, + [OP_SETATTR] = (nfsd4_dec)nfsd4_decode_setattr, + [OP_SETCLIENTID] = (nfsd4_dec)nfsd4_decode_notsupp, + [OP_SETCLIENTID_CONFIRM]= (nfsd4_dec)nfsd4_decode_notsupp, + [OP_VERIFY] = (nfsd4_dec)nfsd4_decode_verify, + [OP_WRITE] = (nfsd4_dec)nfsd4_decode_write, + [OP_RELEASE_LOCKOWNER] = (nfsd4_dec)nfsd4_decode_notsupp, /* new operations for NFSv4.1 */ - [OP_BACKCHANNEL_CTL] (nfsd4_dec)nfsd4_decode_notsupp, - [OP_BIND_CONN_TO_SESSION](nfsd4_dec)nfsd4_decode_notsupp, - [OP_EXCHANGE_ID] (nfsd4_dec)nfsd4_decode_exchange_id, - [OP_CREATE_SESSION] (nfsd4_dec)nfsd4_decode_create_session, - [OP_DESTROY_SESSION] (nfsd4_dec)nfsd4_decode_destroy_session, - [OP_FREE_STATEID] (nfsd4_dec)nfsd4_decode_notsupp, - [OP_GET_DIR_DELEGATION] (nfsd4_dec)nfsd4_decode_notsupp, - [OP_GETDEVICEINFO] (nfsd4_dec)nfsd4_decode_notsupp, - [OP_GETDEVICELIST] (nfsd4_dec)nfsd4_decode_notsupp, - [OP_LAYOUTCOMMIT] (nfsd4_dec)nfsd4_decode_notsupp, - [OP_LAYOUTGET] (nfsd4_dec)nfsd4_decode_notsupp, - [OP_LAYOUTRETURN] (nfsd4_dec)nfsd4_decode_notsupp, - [OP_SECINFO_NO_NAME] (nfsd4_dec)nfsd4_decode_notsupp, - [OP_SEQUENCE] (nfsd4_dec)nfsd4_decode_sequence, - [OP_SET_SSV] (nfsd4_dec)nfsd4_decode_notsupp, - [OP_TEST_STATEID] (nfsd4_dec)nfsd4_decode_notsupp, - [OP_WANT_DELEGATION] (nfsd4_dec)nfsd4_decode_notsupp, - [OP_DESTROY_CLIENTID] (nfsd4_dec)nfsd4_decode_notsupp, - [OP_RECLAIM_COMPLETE] (nfsd4_dec)nfsd4_decode_notsupp, + [OP_BACKCHANNEL_CTL] = (nfsd4_dec)nfsd4_decode_notsupp, + [OP_BIND_CONN_TO_SESSION]= (nfsd4_dec)nfsd4_decode_notsupp, + [OP_EXCHANGE_ID] = (nfsd4_dec)nfsd4_decode_exchange_id, + [OP_CREATE_SESSION] = (nfsd4_dec)nfsd4_decode_create_session, + [OP_DESTROY_SESSION] = (nfsd4_dec)nfsd4_decode_destroy_session, + [OP_FREE_STATEID] = (nfsd4_dec)nfsd4_decode_notsupp, + [OP_GET_DIR_DELEGATION] = (nfsd4_dec)nfsd4_decode_notsupp, + [OP_GETDEVICEINFO] = (nfsd4_dec)nfsd4_decode_notsupp, + [OP_GETDEVICELIST] = (nfsd4_dec)nfsd4_decode_notsupp, + [OP_LAYOUTCOMMIT] = (nfsd4_dec)nfsd4_decode_notsupp, + [OP_LAYOUTGET] = (nfsd4_dec)nfsd4_decode_notsupp, + [OP_LAYOUTRETURN] = (nfsd4_dec)nfsd4_decode_notsupp, + [OP_SECINFO_NO_NAME] = (nfsd4_dec)nfsd4_decode_notsupp, + [OP_SEQUENCE] = (nfsd4_dec)nfsd4_decode_sequence, + [OP_SET_SSV] = (nfsd4_dec)nfsd4_decode_notsupp, + [OP_TEST_STATEID] = (nfsd4_dec)nfsd4_decode_notsupp, + [OP_WANT_DELEGATION] = (nfsd4_dec)nfsd4_decode_notsupp, + [OP_DESTROY_CLIENTID] = (nfsd4_dec)nfsd4_decode_notsupp, + [OP_RECLAIM_COMPLETE] = (nfsd4_dec)nfsd4_decode_notsupp, }; struct nfsd4_minorversion_ops { -- cgit v1.2.3 From 6d6cb0d688d0f262cb4fd5771648b0ac01d4f82c Mon Sep 17 00:00:00 2001 From: Adrian Hunter Date: Wed, 8 Apr 2009 14:07:57 +0200 Subject: UBIFS: reset no_space flag after inode deletion When UBIFS runs out of space it spends a lot of time trying to find more space before returning ENOSPC. As there is no point repeating that unless something has changed, UBIFS has an optimization to record that the file system is 100% full and not try to find space. That flag was not being reset when a pending deletion was finally done. Signed-off-by: Adrian Hunter Reviewed-by: Artem Bityutskiy --- fs/ubifs/budget.c | 3 ++- fs/ubifs/super.c | 5 +++++ 2 files changed, 7 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ubifs/budget.c b/fs/ubifs/budget.c index af1914462f02..d0231ba783df 100644 --- a/fs/ubifs/budget.c +++ b/fs/ubifs/budget.c @@ -628,7 +628,7 @@ void ubifs_convert_page_budget(struct ubifs_info *c) * * This function releases budget corresponding to a dirty inode. It is usually * called when after the inode has been written to the media and marked as - * clean. + * clean. It also causes the "no space" flags to be cleared. */ void ubifs_release_dirty_inode_budget(struct ubifs_info *c, struct ubifs_inode *ui) @@ -636,6 +636,7 @@ void ubifs_release_dirty_inode_budget(struct ubifs_info *c, struct ubifs_budget_req req; memset(&req, 0, sizeof(struct ubifs_budget_req)); + /* The "no space" flags will be cleared because dd_growth is > 0 */ req.dd_growth = c->inode_budget + ALIGN(ui->data_len, 8); ubifs_release_budget(c, &req); } diff --git a/fs/ubifs/super.c b/fs/ubifs/super.c index faa44f90608a..f2c1c0b79f66 100644 --- a/fs/ubifs/super.c +++ b/fs/ubifs/super.c @@ -360,6 +360,11 @@ static void ubifs_delete_inode(struct inode *inode) out: if (ui->dirty) ubifs_release_dirty_inode_budget(c, ui); + else { + /* We've deleted something - clean the "no space" flags */ + c->nospace = c->nospace_rp = 0; + smp_wmb(); + } clear_inode(inode); } -- cgit v1.2.3 From 02cb2858dbfe216bd4a91c4afb4e259ef3b9e151 Mon Sep 17 00:00:00 2001 From: Wang Chen Date: Fri, 24 Apr 2009 15:41:57 +0800 Subject: nfsd: nfs4_stat_init cleanup Save some loop time. Signed-off-by: Wang Chen Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index a4bdf2589b41..b276624be66e 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -4002,6 +4002,7 @@ nfs4_state_init(void) INIT_LIST_HEAD(&conf_str_hashtbl[i]); INIT_LIST_HEAD(&unconf_str_hashtbl[i]); INIT_LIST_HEAD(&unconf_id_hashtbl[i]); + INIT_LIST_HEAD(&reclaim_str_hashtbl[i]); } for (i = 0; i < SESSION_HASH_SIZE; i++) INIT_LIST_HEAD(&sessionid_hashtbl[i]); @@ -4024,8 +4025,6 @@ nfs4_state_init(void) INIT_LIST_HEAD(&close_lru); INIT_LIST_HEAD(&client_lru); INIT_LIST_HEAD(&del_recall_lru); - for (i = 0; i < CLIENT_HASH_SIZE; i++) - INIT_LIST_HEAD(&reclaim_str_hashtbl[i]); reclaim_str_hashtbl_size = 0; return 0; } -- cgit v1.2.3 From dd4dc82d4c129babca9b48b219cff0add5ff4132 Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Tue, 12 May 2009 13:28:09 -0700 Subject: lockd: fix FILE_LOCKING=n build error lockd/svclock.c is missing a header file . is missing a definition of locks_release_private() for the config case of FILE_LOCKING=n, causing a build error: fs/lockd/svclock.c:330: error: implicit declaration of function 'locks_release_private' lockd without FILE_LOCKING doesn't make sense, so make LOCKD and LOCKD_V4 depend on FILE_LOCKING, and make NFS depend on FILE_LOCKING. Signed-off-by: Randy Dunlap Cc: Trond Myklebust Signed-off-by: Andrew Morton Signed-off-by: J. Bruce Fields --- fs/Kconfig | 2 ++ fs/nfs/Kconfig | 2 +- 2 files changed, 3 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/Kconfig b/fs/Kconfig index 9f7270f36b2a..44ab328ceb2a 100644 --- a/fs/Kconfig +++ b/fs/Kconfig @@ -226,10 +226,12 @@ source "fs/nfsd/Kconfig" config LOCKD tristate + depends on FILE_LOCKING config LOCKD_V4 bool depends on NFSD_V3 || NFS_V3 + depends on FILE_LOCKING default y config EXPORTFS diff --git a/fs/nfs/Kconfig b/fs/nfs/Kconfig index e67f3ec07736..7dbb8f27b9d6 100644 --- a/fs/nfs/Kconfig +++ b/fs/nfs/Kconfig @@ -1,6 +1,6 @@ config NFS_FS tristate "NFS client support" - depends on INET + depends on INET && FILE_LOCKING select LOCKD select SUNRPC select NFS_ACL_SUPPORT if NFS_V3_ACL -- cgit v1.2.3 From 8b3884a841f398f6e0a0411d6929d8d9381bb265 Mon Sep 17 00:00:00 2001 From: Hunter Adrian Date: Thu, 14 May 2009 06:32:30 +0200 Subject: UBIFS: return error if link and unlink race Consider a scenario when 'vfs_link(dirA/fileA)' and 'vfs_unlink(dirA/fileA, dirB/fileB)' race. 'vfs_link()' does not lock 'dirA->i_mutex', so this is possible. Both of the functions lock 'fileA->i_mutex' though. Suppose 'vfs_unlink()' wins, and takes 'fileA->i_mutex' mutex first. Suppose 'fileA->i_nlink' is 1. In this case 'ubifs_unlink()' will drop the last reference, and put 'inodeA' to the list of orphans. After this, 'vfs_link()' will link 'dirB/fileB' to 'inodeA'. Thir is a problem because, for example, the subsequent 'vfs_unlink(dirB/fileB)' will add the same inode to the list of orphans. This problem was reported by J. R. Okajima [Artem: add more comments, amended commit message] Signed-off-by: Adrian Hunter Signed-off-by: Artem Bityutskiy --- fs/ubifs/dir.c | 19 +++++++++++++++++++ 1 file changed, 19 insertions(+) (limited to 'fs') diff --git a/fs/ubifs/dir.c b/fs/ubifs/dir.c index f55d523c52bb..552fb0111fff 100644 --- a/fs/ubifs/dir.c +++ b/fs/ubifs/dir.c @@ -528,6 +528,25 @@ static int ubifs_link(struct dentry *old_dentry, struct inode *dir, inode->i_nlink, dir->i_ino); ubifs_assert(mutex_is_locked(&dir->i_mutex)); ubifs_assert(mutex_is_locked(&inode->i_mutex)); + + /* + * Return -ENOENT if we've raced with unlink and i_nlink is 0. Doing + * otherwise has the potential to corrupt the orphan inode list. + * + * Indeed, consider a scenario when 'vfs_link(dirA/fileA)' and + * 'vfs_unlink(dirA/fileA, dirB/fileB)' race. 'vfs_link()' does not + * lock 'dirA->i_mutex', so this is possible. Both of the functions + * lock 'fileA->i_mutex' though. Suppose 'vfs_unlink()' wins, and takes + * 'fileA->i_mutex' mutex first. Suppose 'fileA->i_nlink' is 1. In this + * case 'ubifs_unlink()' will drop the last reference, and put 'inodeA' + * to the list of orphans. After this, 'vfs_link()' will link + * 'dirB/fileB' to 'inodeA'. This is a problem because, for example, + * the subsequent 'vfs_unlink(dirB/fileB)' will add the same inode + * to the list of orphans. + */ + if (inode->i_nlink == 0) + return -ENOENT; + err = dbg_check_synced_i_size(inode); if (err) return err; -- cgit v1.2.3 From 8eec2f36fb869f1e6d81d834bbbd487941222fc8 Mon Sep 17 00:00:00 2001 From: Corentin Chary Date: Mon, 25 May 2009 08:49:10 +0200 Subject: UBIFS: return proper error code if the compr is not present If the compressor is not present, mount_ubifs need to return an error code. This way ubifs_fill_super will stop and handle the error. Signed-off-by: Corentin Chary Signed-off-by: Artem Bityutskiy --- fs/ubifs/super.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/ubifs/super.c b/fs/ubifs/super.c index f2c1c0b79f66..052514ca2792 100644 --- a/fs/ubifs/super.c +++ b/fs/ubifs/super.c @@ -1187,6 +1187,7 @@ static int mount_ubifs(struct ubifs_info *c) if (!ubifs_compr_present(c->default_compr)) { ubifs_err("'compressor \"%s\" is not compiled in", ubifs_compr_name(c->default_compr)); + err = -ENOTSUPP; goto out_free; } -- cgit v1.2.3 From 7c83f5cb551b2e5c4934933fda006636f7424123 Mon Sep 17 00:00:00 2001 From: Artem Bityutskiy Date: Mon, 25 May 2009 19:23:04 +0300 Subject: UBIFS: use anonymous device UBIFS has erroneuosly set 'sb->s_dev' to the UBI volume character device major/minor. This may lead to clashes if there is another FS mounted to a block device with the same major/minor numbers. User-space programs which use 'stat->st_dev' may get confused because of this. This problem was found by Al Viro. He also pointed the way to fix the problem - use 'set_anon_super()' and 'kill_anon_super()' VFS helpers. Signed-off-by: Artem Bityutskiy --- fs/ubifs/super.c | 21 ++++----------------- 1 file changed, 4 insertions(+), 17 deletions(-) (limited to 'fs') diff --git a/fs/ubifs/super.c b/fs/ubifs/super.c index 052514ca2792..42b818daa162 100644 --- a/fs/ubifs/super.c +++ b/fs/ubifs/super.c @@ -1945,7 +1945,6 @@ static int ubifs_fill_super(struct super_block *sb, void *data, int silent) sb->s_magic = UBIFS_SUPER_MAGIC; sb->s_blocksize = UBIFS_BLOCK_SIZE; sb->s_blocksize_bits = UBIFS_BLOCK_SHIFT; - sb->s_dev = c->vi.cdev; sb->s_maxbytes = c->max_inode_sz = key_max_inode_size(c); if (c->max_inode_sz > MAX_LFS_FILESIZE) sb->s_maxbytes = c->max_inode_sz = MAX_LFS_FILESIZE; @@ -1990,16 +1989,9 @@ out_free: static int sb_test(struct super_block *sb, void *data) { dev_t *dev = data; + struct ubifs_info *c = sb->s_fs_info; - return sb->s_dev == *dev; -} - -static int sb_set(struct super_block *sb, void *data) -{ - dev_t *dev = data; - - sb->s_dev = *dev; - return 0; + return c->vi.cdev == *dev; } static int ubifs_get_sb(struct file_system_type *fs_type, int flags, @@ -2027,7 +2019,7 @@ static int ubifs_get_sb(struct file_system_type *fs_type, int flags, dbg_gen("opened ubi%d_%d", vi.ubi_num, vi.vol_id); - sb = sget(fs_type, &sb_test, &sb_set, &vi.cdev); + sb = sget(fs_type, &sb_test, &set_anon_super, &vi.cdev); if (IS_ERR(sb)) { err = PTR_ERR(sb); goto out_close; @@ -2068,16 +2060,11 @@ out_close: return err; } -static void ubifs_kill_sb(struct super_block *sb) -{ - generic_shutdown_super(sb); -} - static struct file_system_type ubifs_fs_type = { .name = "ubifs", .owner = THIS_MODULE, .get_sb = ubifs_get_sb, - .kill_sb = ubifs_kill_sb + .kill_sb = kill_anon_super, }; /* -- cgit v1.2.3 From fca4217c5bab31019b5247e977673c9fcc385f6b Mon Sep 17 00:00:00 2001 From: Greg Banks Date: Wed, 1 Apr 2009 07:28:13 +1100 Subject: knfsd: reply cache cleanups Make REQHASH() an inline function. Rename hash_list to cache_hash. Fix an obsolete comment. Signed-off-by: Greg Banks Signed-off-by: J. Bruce Fields --- fs/nfsd/nfscache.c | 29 +++++++++++++++++++---------- include/linux/nfsd/cache.h | 3 +-- 2 files changed, 20 insertions(+), 12 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfscache.c b/fs/nfsd/nfscache.c index 5bfc2ac60d54..6f0aa4989c61 100644 --- a/fs/nfsd/nfscache.c +++ b/fs/nfsd/nfscache.c @@ -29,15 +29,24 @@ */ #define CACHESIZE 1024 #define HASHSIZE 64 -#define REQHASH(xid) (((((__force __u32)xid) >> 24) ^ ((__force __u32)xid)) & (HASHSIZE-1)) -static struct hlist_head * hash_list; +static struct hlist_head * cache_hash; static struct list_head lru_head; static int cache_disabled = 1; +/* + * Calculate the hash index from an XID. + */ +static inline u32 request_hash(u32 xid) +{ + u32 h = xid; + h ^= (xid >> 24); + return h & (HASHSIZE-1); +} + static int nfsd_cache_append(struct svc_rqst *rqstp, struct kvec *vec); -/* +/* * locking for the reply cache: * A cache entry is "single use" if c_state == RC_INPROG * Otherwise, it when accessing _prev or _next, the lock must be held. @@ -62,8 +71,8 @@ int nfsd_reply_cache_init(void) i--; } - hash_list = kcalloc (HASHSIZE, sizeof(struct hlist_head), GFP_KERNEL); - if (!hash_list) + cache_hash = kcalloc (HASHSIZE, sizeof(struct hlist_head), GFP_KERNEL); + if (!cache_hash) goto out_nomem; cache_disabled = 0; @@ -88,8 +97,8 @@ void nfsd_reply_cache_shutdown(void) cache_disabled = 1; - kfree (hash_list); - hash_list = NULL; + kfree (cache_hash); + cache_hash = NULL; } /* @@ -108,7 +117,7 @@ static void hash_refile(struct svc_cacherep *rp) { hlist_del_init(&rp->c_hash); - hlist_add_head(&rp->c_hash, hash_list + REQHASH(rp->c_xid)); + hlist_add_head(&rp->c_hash, cache_hash + request_hash(rp->c_xid)); } /* @@ -138,7 +147,7 @@ nfsd_cache_lookup(struct svc_rqst *rqstp, int type) spin_lock(&cache_lock); rtn = RC_DOIT; - rh = &hash_list[REQHASH(xid)]; + rh = &cache_hash[request_hash(xid)]; hlist_for_each_entry(rp, hn, rh, c_hash) { if (rp->c_state != RC_UNUSED && xid == rp->c_xid && proc == rp->c_proc && @@ -264,7 +273,7 @@ nfsd_cache_update(struct svc_rqst *rqstp, int cachetype, __be32 *statp) len = resv->iov_len - ((char*)statp - (char*)resv->iov_base); len >>= 2; - + /* Don't cache excessive amounts of data and XDR failures */ if (!statp || len > (256 >> 2)) { rp->c_state = RC_UNUSED; diff --git a/include/linux/nfsd/cache.h b/include/linux/nfsd/cache.h index 5bccaab81056..3a3f58934f5e 100644 --- a/include/linux/nfsd/cache.h +++ b/include/linux/nfsd/cache.h @@ -14,8 +14,7 @@ #include /* - * Representation of a reply cache entry. The first two members *must* - * be hash_next and hash_prev. + * Representation of a reply cache entry. */ struct svc_cacherep { struct hlist_node c_hash; -- cgit v1.2.3 From cf0a586cf41a1779edeee7562afb5d0ab46c7cf4 Mon Sep 17 00:00:00 2001 From: Greg Banks Date: Wed, 1 Apr 2009 07:28:15 +1100 Subject: knfsd: fix reply cache memory corruption Fix a regression in the reply cache introduced when the code was converted to use proper Linux lists. When a new entry needs to be inserted, the case where all the entries are currently being used by threads is not correctly detected. This can result in memory corruption and a crash. In the current code this is an extremely unlikely corner case; it would require the machine to have 1024 nfsd threads and all of them to be busy at the same time. However, upcoming reply cache changes make this more likely; a crash due to this problem was actually observed in field. Signed-off-by: Greg Banks Signed-off-by: J. Bruce Fields --- fs/nfsd/nfscache.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfscache.c b/fs/nfsd/nfscache.c index 6f0aa4989c61..4638635c5d87 100644 --- a/fs/nfsd/nfscache.c +++ b/fs/nfsd/nfscache.c @@ -174,8 +174,8 @@ nfsd_cache_lookup(struct svc_rqst *rqstp, int type) } } - /* This should not happen */ - if (rp == NULL) { + /* All entries on the LRU are in-progress. This should not happen */ + if (&rp->c_lru == &lru_head) { static int complaints; printk(KERN_WARNING "nfsd: all repcache entries locked!\n"); -- cgit v1.2.3 From 1dbd0d53f394cd9a86fc801dd68fdbcbcdb45718 Mon Sep 17 00:00:00 2001 From: Greg Banks Date: Wed, 1 Apr 2009 07:28:21 +1100 Subject: knfsd: remove unreported filehandle stats counters The file nfsfh.c contains two static variables nfsd_nr_verified and nfsd_nr_put. These are counters which are incremented as a side effect of the fh_verify() fh_compose() and fh_put() operations, i.e. at least twice per NFS call for any non-trivial workload. Needless to say this makes the cacheline that contains them (and any other innocent victims) a very hot contention point indeed under high call-rate workloads on multiprocessor NFS server. It also turns out that these counters are not used anywhere. They're not reported to userspace, they're not used in logic, they're not even exported from the object file (let alone the module). All they do is waste CPU time. So this patch removes them. Tests on a 16 CPU Altix A4700 with 2 10gige Myricom cards, configured separately (no bonding). Workload is 640 client threads doing directory traverals with random small reads, from server RAM. Before ====== Kernel profile: % cumulative self self total time samples samples calls 1/call 1/call name 6.05 2716.00 2716.00 30406 0.09 1.02 svc_process 4.44 4706.00 1990.00 1975 1.01 1.01 spin_unlock_irqrestore 3.72 6376.00 1670.00 1666 1.00 1.00 svc_export_put 3.41 7907.00 1531.00 1786 0.86 1.02 nfsd_ofcache_lookup 3.25 9363.00 1456.00 10965 0.13 1.01 nfsd_dispatch 3.10 10752.00 1389.00 1376 1.01 1.01 nfsd_cache_lookup 2.57 11907.00 1155.00 4517 0.26 1.03 svc_tcp_recvfrom ... 2.21 15352.00 1003.00 1081 0.93 1.00 nfsd_choose_ofc <---- ^^^^ Here the function nfsd_choose_ofc() reads a global variable which by accident happened to be located in the same cacheline as nfsd_nr_verified. Call rate: nullarbor:~ # pmdumptext nfs3.server.calls ... Thu Dec 13 00:15:27 184780.663 Thu Dec 13 00:15:28 184885.881 Thu Dec 13 00:15:29 184449.215 Thu Dec 13 00:15:30 184971.058 Thu Dec 13 00:15:31 185036.052 Thu Dec 13 00:15:32 185250.475 Thu Dec 13 00:15:33 184481.319 Thu Dec 13 00:15:34 185225.737 Thu Dec 13 00:15:35 185408.018 Thu Dec 13 00:15:36 185335.764 After ===== kernel profile: % cumulative self self total time samples samples calls 1/call 1/call name 6.33 2813.00 2813.00 29979 0.09 1.01 svc_process 4.66 4883.00 2070.00 2065 1.00 1.00 spin_unlock_irqrestore 4.06 6687.00 1804.00 2182 0.83 1.00 nfsd_ofcache_lookup 3.20 8110.00 1423.00 10932 0.13 1.00 nfsd_dispatch 3.03 9456.00 1346.00 1343 1.00 1.00 nfsd_cache_lookup 2.62 10622.00 1166.00 4645 0.25 1.01 svc_tcp_recvfrom [...] 0.10 42586.00 44.00 74 0.59 1.00 nfsd_choose_ofc <--- HA!! ^^^^ Call rate: nullarbor:~ # pmdumptext nfs3.server.calls ... Thu Dec 13 01:45:28 194677.118 Thu Dec 13 01:45:29 193932.692 Thu Dec 13 01:45:30 194294.364 Thu Dec 13 01:45:31 194971.276 Thu Dec 13 01:45:32 194111.207 Thu Dec 13 01:45:33 194999.635 Thu Dec 13 01:45:34 195312.594 Thu Dec 13 01:45:35 195707.293 Thu Dec 13 01:45:36 194610.353 Thu Dec 13 01:45:37 195913.662 Thu Dec 13 01:45:38 194808.675 i.e. about a 5.3% improvement in call rate. Signed-off-by: Greg Banks Reviewed-by: David Chinner Signed-off-by: J. Bruce Fields --- fs/nfsd/nfsfh.c | 6 ------ 1 file changed, 6 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfsfh.c b/fs/nfsd/nfsfh.c index 9f1ca17293d3..8847f3fbfc1e 100644 --- a/fs/nfsd/nfsfh.c +++ b/fs/nfsd/nfsfh.c @@ -27,9 +27,6 @@ #define NFSDDBG_FACILITY NFSDDBG_FH -static int nfsd_nr_verified; -static int nfsd_nr_put; - /* * our acceptability function. * if NOSUBTREECHECK, accept anything @@ -251,7 +248,6 @@ static __be32 nfsd_set_fh_dentry(struct svc_rqst *rqstp, struct svc_fh *fhp) fhp->fh_dentry = dentry; fhp->fh_export = exp; - nfsd_nr_verified++; return 0; out: exp_put(exp); @@ -552,7 +548,6 @@ fh_compose(struct svc_fh *fhp, struct svc_export *exp, struct dentry *dentry, return nfserr_opnotsupp; } - nfsd_nr_verified++; return 0; } @@ -609,7 +604,6 @@ fh_put(struct svc_fh *fhp) fhp->fh_pre_saved = 0; fhp->fh_post_saved = 0; #endif - nfsd_nr_put++; } if (exp) { cache_put(&exp->h, &svc_export_cache); -- cgit v1.2.3 From 428ff9d2e37d3a82af0f56b476f70c244cf550d1 Mon Sep 17 00:00:00 2001 From: Artem Bityutskiy Date: Mon, 25 May 2009 16:59:28 +0300 Subject: UBIFS: remove dead code UBIFS assumes that @c->min_io_size is 8 in case of NOR flash. This is because UBIFS alignes all nodes to 8-byte boundary, and maintaining @c->min_io_size introduced unnecessary complications. This patch removes senseless constructs like: if (c->min_io_size == 1) NOR-specific code Also, few commentaries amendments. Signed-off-by: Artem Bityutskiy --- fs/ubifs/budget.c | 1 - fs/ubifs/recovery.c | 31 ++++--------------------------- 2 files changed, 4 insertions(+), 28 deletions(-) (limited to 'fs') diff --git a/fs/ubifs/budget.c b/fs/ubifs/budget.c index d0231ba783df..eaf6d891d46f 100644 --- a/fs/ubifs/budget.c +++ b/fs/ubifs/budget.c @@ -91,7 +91,6 @@ static int shrink_liability(struct ubifs_info *c, int nr_to_write) return nr_written; } - /** * run_gc - run garbage collector. * @c: UBIFS file-system description object diff --git a/fs/ubifs/recovery.c b/fs/ubifs/recovery.c index 10662975d2ef..805605250f12 100644 --- a/fs/ubifs/recovery.c +++ b/fs/ubifs/recovery.c @@ -343,33 +343,15 @@ int ubifs_write_rcvrd_mst_node(struct ubifs_info *c) * * This function returns %1 if @offs was in the last write to the LEB whose data * is in @buf, otherwise %0 is returned. The determination is made by checking - * for subsequent empty space starting from the next min_io_size boundary (or a - * bit less than the common header size if min_io_size is one). + * for subsequent empty space starting from the next @c->min_io_size boundary. */ static int is_last_write(const struct ubifs_info *c, void *buf, int offs) { - int empty_offs; - int check_len; + int empty_offs, check_len; uint8_t *p; - if (c->min_io_size == 1) { - check_len = c->leb_size - offs; - p = buf + check_len; - for (; check_len > 0; check_len--) - if (*--p != 0xff) - break; - /* - * 'check_len' is the size of the corruption which cannot be - * more than the size of 1 node if it was caused by an unclean - * unmount. - */ - if (check_len > UBIFS_MAX_NODE_SZ) - return 0; - return 1; - } - /* - * Round up to the next c->min_io_size boundary i.e. 'offs' is in the + * Round up to the next @c->min_io_size boundary i.e. @offs is in the * last wbuf written. After that should be empty space. */ empty_offs = ALIGN(offs + 1, c->min_io_size); @@ -392,7 +374,7 @@ static int is_last_write(const struct ubifs_info *c, void *buf, int offs) * * This function pads up to the next min_io_size boundary (if there is one) and * sets empty space to all 0xff. @buf, @offs and @len are updated to the next - * min_io_size boundary (if there is one). + * @c->min_io_size boundary. */ static void clean_buf(const struct ubifs_info *c, void **buf, int lnum, int *offs, int *len) @@ -402,11 +384,6 @@ static void clean_buf(const struct ubifs_info *c, void **buf, int lnum, lnum = lnum; dbg_rcvry("cleaning corruption at %d:%d", lnum, *offs); - if (c->min_io_size == 1) { - memset(*buf, 0xff, c->leb_size - *offs); - return; - } - ubifs_assert(!(*offs & 7)); empty_offs = ALIGN(*offs, c->min_io_size); pad_len = empty_offs - *offs; -- cgit v1.2.3 From 0dc54e9f33e2fbcea28356bc2c8c931cb307d3b3 Mon Sep 17 00:00:00 2001 From: Kevin Cernekee Date: Wed, 8 Apr 2009 22:52:28 -0700 Subject: mtd: add MEMERASE64 ioctl for >4GiB devices New MEMERASE/MEMREADOOB/MEMWRITEOOB ioctls are needed in order to support 64-bit offsets into large NAND flash devices. Signed-off-by: Kevin Cernekee Signed-off-by: Artem Bityutskiy Signed-off-by: David Woodhouse --- drivers/mtd/mtdchar.c | 29 +++++++++++++++++++++-------- fs/compat_ioctl.c | 1 + include/mtd/mtd-abi.h | 6 ++++++ 3 files changed, 28 insertions(+), 8 deletions(-) (limited to 'fs') diff --git a/drivers/mtd/mtdchar.c b/drivers/mtd/mtdchar.c index 763d3f0a1f42..ad4b8618977d 100644 --- a/drivers/mtd/mtdchar.c +++ b/drivers/mtd/mtdchar.c @@ -417,6 +417,7 @@ static int mtd_ioctl(struct inode *inode, struct file *file, break; case MEMERASE: + case MEMERASE64: { struct erase_info *erase; @@ -427,20 +428,32 @@ static int mtd_ioctl(struct inode *inode, struct file *file, if (!erase) ret = -ENOMEM; else { - struct erase_info_user einfo; - wait_queue_head_t waitq; DECLARE_WAITQUEUE(wait, current); init_waitqueue_head(&waitq); - if (copy_from_user(&einfo, argp, - sizeof(struct erase_info_user))) { - kfree(erase); - return -EFAULT; + if (cmd == MEMERASE64) { + struct erase_info_user64 einfo64; + + if (copy_from_user(&einfo64, argp, + sizeof(struct erase_info_user64))) { + kfree(erase); + return -EFAULT; + } + erase->addr = einfo64.start; + erase->len = einfo64.length; + } else { + struct erase_info_user einfo32; + + if (copy_from_user(&einfo32, argp, + sizeof(struct erase_info_user))) { + kfree(erase); + return -EFAULT; + } + erase->addr = einfo32.start; + erase->len = einfo32.length; } - erase->addr = einfo.start; - erase->len = einfo.length; erase->mtd = mtd; erase->callback = mtdchar_erase_callback; erase->priv = (unsigned long)&waitq; diff --git a/fs/compat_ioctl.c b/fs/compat_ioctl.c index b83f6bcfa51a..c603ca2c223a 100644 --- a/fs/compat_ioctl.c +++ b/fs/compat_ioctl.c @@ -2441,6 +2441,7 @@ COMPATIBLE_IOCTL(MEMGETREGIONCOUNT) COMPATIBLE_IOCTL(MEMGETREGIONINFO) COMPATIBLE_IOCTL(MEMGETBADBLOCK) COMPATIBLE_IOCTL(MEMSETBADBLOCK) +COMPATIBLE_IOCTL(MEMERASE64) /* NBD */ ULONG_IOCTL(NBD_SET_SOCK) ULONG_IOCTL(NBD_SET_BLKSIZE) diff --git a/include/mtd/mtd-abi.h b/include/mtd/mtd-abi.h index b6595b3c68b6..2e32be1e3a1e 100644 --- a/include/mtd/mtd-abi.h +++ b/include/mtd/mtd-abi.h @@ -12,6 +12,11 @@ struct erase_info_user { __u32 length; }; +struct erase_info_user64 { + __u64 start; + __u64 length; +}; + struct mtd_oob_buf { __u32 start; __u32 length; @@ -95,6 +100,7 @@ struct otp_info { #define ECCGETLAYOUT _IOR('M', 17, struct nand_ecclayout) #define ECCGETSTATS _IOR('M', 18, struct mtd_ecc_stats) #define MTDFILEMODE _IO('M', 19) +#define MEMERASE64 _IOW('M', 20, struct erase_info_user64) /* * Obsolete legacy interface. Keep it in order not to break userspace -- cgit v1.2.3 From 977185404046afb31d2e18fac0a076de1a20bf0e Mon Sep 17 00:00:00 2001 From: Kevin Cernekee Date: Wed, 8 Apr 2009 22:53:13 -0700 Subject: mtd: compat_ioctl cleanup 1) Move the MEMREADOOB/MEMWRITEOOB compat_ioctl wrappers from fs/compat_ioctl.c into mtdchar.c . Original request was here: http://lkml.org/lkml/2009/4/1/295 2) Add missing COMPATIBLE_IOCTL lines, so that mtd-utils does not error out when running in 64/32 compatibility mode. LKML-Reference: <200904011650.22928.arnd@arndb.de> Signed-off-by: Kevin Cernekee Signed-off-by: Artem Bityutskiy Signed-off-by: David Woodhouse --- drivers/mtd/mtdchar.c | 255 +++++++++++++++++++++++++++++++++----------------- fs/compat_ioctl.c | 51 ++-------- 2 files changed, 180 insertions(+), 126 deletions(-) (limited to 'fs') diff --git a/drivers/mtd/mtdchar.c b/drivers/mtd/mtdchar.c index ad4b8618977d..51bb0b092003 100644 --- a/drivers/mtd/mtdchar.c +++ b/drivers/mtd/mtdchar.c @@ -14,6 +14,7 @@ #include #include #include +#include #include #include @@ -355,6 +356,100 @@ static int otp_select_filemode(struct mtd_file_info *mfi, int mode) # define otp_select_filemode(f,m) -EOPNOTSUPP #endif +static int mtd_do_writeoob(struct file *file, struct mtd_info *mtd, + uint64_t start, uint32_t length, void __user *ptr, + uint32_t __user *retp) +{ + struct mtd_oob_ops ops; + uint32_t retlen; + int ret = 0; + + if (!(file->f_mode & FMODE_WRITE)) + return -EPERM; + + if (length > 4096) + return -EINVAL; + + if (!mtd->write_oob) + ret = -EOPNOTSUPP; + else + ret = access_ok(VERIFY_READ, ptr, length) ? 0 : EFAULT; + + if (ret) + return ret; + + ops.ooblen = length; + ops.ooboffs = start & (mtd->oobsize - 1); + ops.datbuf = NULL; + ops.mode = MTD_OOB_PLACE; + + if (ops.ooboffs && ops.ooblen > (mtd->oobsize - ops.ooboffs)) + return -EINVAL; + + ops.oobbuf = kmalloc(length, GFP_KERNEL); + if (!ops.oobbuf) + return -ENOMEM; + + if (copy_from_user(ops.oobbuf, ptr, length)) { + kfree(ops.oobbuf); + return -EFAULT; + } + + start &= ~((uint64_t)mtd->oobsize - 1); + ret = mtd->write_oob(mtd, start, &ops); + + if (ops.oobretlen > 0xFFFFFFFFU) + ret = -EOVERFLOW; + retlen = ops.oobretlen; + if (copy_to_user(retp, &retlen, sizeof(length))) + ret = -EFAULT; + + kfree(ops.oobbuf); + return ret; +} + +static int mtd_do_readoob(struct mtd_info *mtd, uint64_t start, + uint32_t length, void __user *ptr, uint32_t __user *retp) +{ + struct mtd_oob_ops ops; + int ret = 0; + + if (length > 4096) + return -EINVAL; + + if (!mtd->read_oob) + ret = -EOPNOTSUPP; + else + ret = access_ok(VERIFY_WRITE, ptr, + length) ? 0 : -EFAULT; + if (ret) + return ret; + + ops.ooblen = length; + ops.ooboffs = start & (mtd->oobsize - 1); + ops.datbuf = NULL; + ops.mode = MTD_OOB_PLACE; + + if (ops.ooboffs && ops.ooblen > (mtd->oobsize - ops.ooboffs)) + return -EINVAL; + + ops.oobbuf = kmalloc(length, GFP_KERNEL); + if (!ops.oobbuf) + return -ENOMEM; + + start &= ~((uint64_t)mtd->oobsize - 1); + ret = mtd->read_oob(mtd, start, &ops); + + if (put_user(ops.oobretlen, retp)) + ret = -EFAULT; + else if (ops.oobretlen && copy_to_user(ptr, ops.oobbuf, + ops.oobretlen)) + ret = -EFAULT; + + kfree(ops.oobbuf); + return ret; +} + static int mtd_ioctl(struct inode *inode, struct file *file, u_int cmd, u_long arg) { @@ -487,100 +582,28 @@ static int mtd_ioctl(struct inode *inode, struct file *file, case MEMWRITEOOB: { struct mtd_oob_buf buf; - struct mtd_oob_ops ops; - struct mtd_oob_buf __user *user_buf = argp; - uint32_t retlen; - - if(!(file->f_mode & FMODE_WRITE)) - return -EPERM; - - if (copy_from_user(&buf, argp, sizeof(struct mtd_oob_buf))) - return -EFAULT; - - if (buf.length > 4096) - return -EINVAL; - - if (!mtd->write_oob) - ret = -EOPNOTSUPP; - else - ret = access_ok(VERIFY_READ, buf.ptr, - buf.length) ? 0 : EFAULT; - - if (ret) - return ret; - - ops.ooblen = buf.length; - ops.ooboffs = buf.start & (mtd->oobsize - 1); - ops.datbuf = NULL; - ops.mode = MTD_OOB_PLACE; - - if (ops.ooboffs && ops.ooblen > (mtd->oobsize - ops.ooboffs)) - return -EINVAL; - - ops.oobbuf = kmalloc(buf.length, GFP_KERNEL); - if (!ops.oobbuf) - return -ENOMEM; - - if (copy_from_user(ops.oobbuf, buf.ptr, buf.length)) { - kfree(ops.oobbuf); - return -EFAULT; - } - - buf.start &= ~(mtd->oobsize - 1); - ret = mtd->write_oob(mtd, buf.start, &ops); + struct mtd_oob_buf __user *buf_user = argp; - if (ops.oobretlen > 0xFFFFFFFFU) - ret = -EOVERFLOW; - retlen = ops.oobretlen; - if (copy_to_user(&user_buf->length, &retlen, sizeof(buf.length))) + /* NOTE: writes return length to buf_user->length */ + if (copy_from_user(&buf, argp, sizeof(buf))) ret = -EFAULT; - - kfree(ops.oobbuf); + else + ret = mtd_do_writeoob(file, mtd, buf.start, buf.length, + buf.ptr, &buf_user->length); break; - } case MEMREADOOB: { struct mtd_oob_buf buf; - struct mtd_oob_ops ops; - - if (copy_from_user(&buf, argp, sizeof(struct mtd_oob_buf))) - return -EFAULT; - - if (buf.length > 4096) - return -EINVAL; - - if (!mtd->read_oob) - ret = -EOPNOTSUPP; - else - ret = access_ok(VERIFY_WRITE, buf.ptr, - buf.length) ? 0 : -EFAULT; - if (ret) - return ret; - - ops.ooblen = buf.length; - ops.ooboffs = buf.start & (mtd->oobsize - 1); - ops.datbuf = NULL; - ops.mode = MTD_OOB_PLACE; + struct mtd_oob_buf __user *buf_user = argp; - if (ops.ooboffs && ops.ooblen > (mtd->oobsize - ops.ooboffs)) - return -EINVAL; - - ops.oobbuf = kmalloc(buf.length, GFP_KERNEL); - if (!ops.oobbuf) - return -ENOMEM; - - buf.start &= ~(mtd->oobsize - 1); - ret = mtd->read_oob(mtd, buf.start, &ops); - - if (put_user(ops.oobretlen, (uint32_t __user *)argp)) - ret = -EFAULT; - else if (ops.oobretlen && copy_to_user(buf.ptr, ops.oobbuf, - ops.oobretlen)) + /* NOTE: writes return length to buf_user->start */ + if (copy_from_user(&buf, argp, sizeof(buf))) ret = -EFAULT; - - kfree(ops.oobbuf); + else + ret = mtd_do_readoob(mtd, buf.start, buf.length, + buf.ptr, &buf_user->start); break; } @@ -771,6 +794,67 @@ static int mtd_ioctl(struct inode *inode, struct file *file, return ret; } /* memory_ioctl */ +#ifdef CONFIG_COMPAT + +struct mtd_oob_buf32 { + u_int32_t start; + u_int32_t length; + compat_caddr_t ptr; /* unsigned char* */ +}; + +#define MEMWRITEOOB32 _IOWR('M', 3, struct mtd_oob_buf32) +#define MEMREADOOB32 _IOWR('M', 4, struct mtd_oob_buf32) + +static long mtd_compat_ioctl(struct file *file, unsigned int cmd, + unsigned long arg) +{ + struct mtd_file_info *mfi = file->private_data; + struct mtd_info *mtd = mfi->mtd; + void __user *argp = (void __user *)arg; + int ret = 0; + + lock_kernel(); + + switch (cmd) { + case MEMWRITEOOB32: + { + struct mtd_oob_buf32 buf; + struct mtd_oob_buf32 __user *buf_user = argp; + + if (copy_from_user(&buf, argp, sizeof(buf))) + ret = -EFAULT; + else + ret = mtd_do_writeoob(file, mtd, buf.start, + buf.length, compat_ptr(buf.ptr), + &buf_user->length); + break; + } + + case MEMREADOOB32: + { + struct mtd_oob_buf32 buf; + struct mtd_oob_buf32 __user *buf_user = argp; + + /* NOTE: writes return length to buf->start */ + if (copy_from_user(&buf, argp, sizeof(buf))) + ret = -EFAULT; + else + ret = mtd_do_readoob(mtd, buf.start, + buf.length, compat_ptr(buf.ptr), + &buf_user->start); + break; + } + default: + ret = -ENOIOCTLCMD; + } + + unlock_kernel(); + + return ret; +} + +#endif /* CONFIG_COMPAT */ + /* * try to determine where a shared mapping can be made * - only supported for NOMMU at the moment (MMU can't doesn't copy private @@ -830,6 +914,9 @@ static const struct file_operations mtd_fops = { .read = mtd_read, .write = mtd_write, .ioctl = mtd_ioctl, +#ifdef CONFIG_COMPAT + .compat_ioctl = mtd_compat_ioctl, +#endif .open = mtd_open, .release = mtd_close, .mmap = mtd_mmap, diff --git a/fs/compat_ioctl.c b/fs/compat_ioctl.c index c603ca2c223a..196397bff086 100644 --- a/fs/compat_ioctl.c +++ b/fs/compat_ioctl.c @@ -1411,46 +1411,6 @@ static int ioc_settimeout(unsigned int fd, unsigned int cmd, unsigned long arg) #define HIDPGETCONNLIST _IOR('H', 210, int) #define HIDPGETCONNINFO _IOR('H', 211, int) -struct mtd_oob_buf32 { - u_int32_t start; - u_int32_t length; - compat_caddr_t ptr; /* unsigned char* */ -}; - -#define MEMWRITEOOB32 _IOWR('M',3,struct mtd_oob_buf32) -#define MEMREADOOB32 _IOWR('M',4,struct mtd_oob_buf32) - -static int mtd_rw_oob(unsigned int fd, unsigned int cmd, unsigned long arg) -{ - struct mtd_oob_buf __user *buf = compat_alloc_user_space(sizeof(*buf)); - struct mtd_oob_buf32 __user *buf32 = compat_ptr(arg); - u32 data; - char __user *datap; - unsigned int real_cmd; - int err; - - real_cmd = (cmd == MEMREADOOB32) ? - MEMREADOOB : MEMWRITEOOB; - - if (copy_in_user(&buf->start, &buf32->start, - 2 * sizeof(u32)) || - get_user(data, &buf32->ptr)) - return -EFAULT; - datap = compat_ptr(data); - if (put_user(datap, &buf->ptr)) - return -EFAULT; - - err = sys_ioctl(fd, real_cmd, (unsigned long) buf); - - if (!err) { - if (copy_in_user(&buf32->start, &buf->start, - 2 * sizeof(u32))) - err = -EFAULT; - } - - return err; -} - #ifdef CONFIG_BLOCK struct raw32_config_request { @@ -2439,8 +2399,17 @@ COMPATIBLE_IOCTL(MEMLOCK) COMPATIBLE_IOCTL(MEMUNLOCK) COMPATIBLE_IOCTL(MEMGETREGIONCOUNT) COMPATIBLE_IOCTL(MEMGETREGIONINFO) +COMPATIBLE_IOCTL(MEMSETOOBSEL) +COMPATIBLE_IOCTL(MEMGETOOBSEL) COMPATIBLE_IOCTL(MEMGETBADBLOCK) COMPATIBLE_IOCTL(MEMSETBADBLOCK) +COMPATIBLE_IOCTL(OTPSELECT) +COMPATIBLE_IOCTL(OTPGETREGIONCOUNT) +COMPATIBLE_IOCTL(OTPGETREGIONINFO) +COMPATIBLE_IOCTL(OTPLOCK) +COMPATIBLE_IOCTL(ECCGETLAYOUT) +COMPATIBLE_IOCTL(ECCGETSTATS) +COMPATIBLE_IOCTL(MTDFILEMODE) COMPATIBLE_IOCTL(MEMERASE64) /* NBD */ ULONG_IOCTL(NBD_SET_SOCK) @@ -2551,8 +2520,6 @@ COMPATIBLE_IOCTL(JSIOCGBUTTONS) COMPATIBLE_IOCTL(JSIOCGNAME(0)) /* now things that need handlers */ -HANDLE_IOCTL(MEMREADOOB32, mtd_rw_oob) -HANDLE_IOCTL(MEMWRITEOOB32, mtd_rw_oob) #ifdef CONFIG_NET HANDLE_IOCTL(SIOCGIFNAME, dev_ifname32) HANDLE_IOCTL(SIOCGIFCONF, dev_ifconf) -- cgit v1.2.3 From aea7cea9fa9e39e71f95ad70b3daf98ba9972587 Mon Sep 17 00:00:00 2001 From: Kevin Cernekee Date: Wed, 8 Apr 2009 22:53:49 -0700 Subject: mtd: add OOB ioctls for >4GiB devices Signed-off-by: Kevin Cernekee Signed-off-by: Artem Bityutskiy Signed-off-by: David Woodhouse --- drivers/mtd/mtdchar.c | 28 ++++++++++++++++++++++++++++ fs/compat_ioctl.c | 2 ++ include/mtd/mtd-abi.h | 9 +++++++++ 3 files changed, 39 insertions(+) (limited to 'fs') diff --git a/drivers/mtd/mtdchar.c b/drivers/mtd/mtdchar.c index 51bb0b092003..99d1fbc95011 100644 --- a/drivers/mtd/mtdchar.c +++ b/drivers/mtd/mtdchar.c @@ -607,6 +607,34 @@ static int mtd_ioctl(struct inode *inode, struct file *file, break; } + case MEMWRITEOOB64: + { + struct mtd_oob_buf64 buf; + struct mtd_oob_buf64 __user *buf_user = argp; + + if (copy_from_user(&buf, argp, sizeof(buf))) + ret = -EFAULT; + else + ret = mtd_do_writeoob(file, mtd, buf.start, buf.length, + (void __user *)(uintptr_t)buf.usr_ptr, + &buf_user->length); + break; + } + + case MEMREADOOB64: + { + struct mtd_oob_buf64 buf; + struct mtd_oob_buf64 __user *buf_user = argp; + + if (copy_from_user(&buf, argp, sizeof(buf))) + ret = -EFAULT; + else + ret = mtd_do_readoob(mtd, buf.start, buf.length, + (void __user *)(uintptr_t)buf.usr_ptr, + &buf_user->length); + break; + } + case MEMLOCK: { struct erase_info_user einfo; diff --git a/fs/compat_ioctl.c b/fs/compat_ioctl.c index 196397bff086..8da222eacbaf 100644 --- a/fs/compat_ioctl.c +++ b/fs/compat_ioctl.c @@ -2411,6 +2411,8 @@ COMPATIBLE_IOCTL(ECCGETLAYOUT) COMPATIBLE_IOCTL(ECCGETSTATS) COMPATIBLE_IOCTL(MTDFILEMODE) COMPATIBLE_IOCTL(MEMERASE64) +COMPATIBLE_IOCTL(MEMREADOOB64) +COMPATIBLE_IOCTL(MEMWRITEOOB64) /* NBD */ ULONG_IOCTL(NBD_SET_SOCK) ULONG_IOCTL(NBD_SET_BLKSIZE) diff --git a/include/mtd/mtd-abi.h b/include/mtd/mtd-abi.h index 2e32be1e3a1e..be51ae2bd0ff 100644 --- a/include/mtd/mtd-abi.h +++ b/include/mtd/mtd-abi.h @@ -23,6 +23,13 @@ struct mtd_oob_buf { unsigned char __user *ptr; }; +struct mtd_oob_buf64 { + __u64 start; + __u32 pad; + __u32 length; + __u64 usr_ptr; +}; + #define MTD_ABSENT 0 #define MTD_RAM 1 #define MTD_ROM 2 @@ -101,6 +108,8 @@ struct otp_info { #define ECCGETSTATS _IOR('M', 18, struct mtd_ecc_stats) #define MTDFILEMODE _IO('M', 19) #define MEMERASE64 _IOW('M', 20, struct erase_info_user64) +#define MEMWRITEOOB64 _IOWR('M', 21, struct mtd_oob_buf64) +#define MEMREADOOB64 _IOWR('M', 22, struct mtd_oob_buf64) /* * Obsolete legacy interface. Keep it in order not to break userspace -- cgit v1.2.3 From 668ff9ab45d595222d3f90d7974ccba3518e3bb3 Mon Sep 17 00:00:00 2001 From: Kevin Cernekee Date: Tue, 14 Apr 2009 21:59:22 -0700 Subject: mtd: Handle compat ioctls directly; remove all trace from compat_ioctl.c Remove all references to MTD ioctls from fs/compat_ioctl.c and let them all be handled by mtd_compat_ioctl(). Signed-off-by: Kevin Cernekee Signed-off-by: Artem Bityutskiy Signed-off-by: David Woodhouse --- drivers/mtd/mtdchar.c | 3 ++- fs/compat_ioctl.c | 22 ---------------------- 2 files changed, 2 insertions(+), 23 deletions(-) (limited to 'fs') diff --git a/drivers/mtd/mtdchar.c b/drivers/mtd/mtdchar.c index 99d1fbc95011..5fff04f3303d 100644 --- a/drivers/mtd/mtdchar.c +++ b/drivers/mtd/mtdchar.c @@ -836,6 +836,7 @@ struct mtd_oob_buf32 { static long mtd_compat_ioctl(struct file *file, unsigned int cmd, unsigned long arg) { + struct inode *inode = file->f_path.dentry->d_inode; struct mtd_file_info *mfi = file->private_data; struct mtd_info *mtd = mfi->mtd; void __user *argp = (void __user *)arg; @@ -873,7 +874,7 @@ static long mtd_compat_ioctl(struct file *file, unsigned int cmd, break; } default: - ret = -ENOIOCTLCMD; + ret = mtd_ioctl(inode, file, cmd, arg); } unlock_kernel(); diff --git a/fs/compat_ioctl.c b/fs/compat_ioctl.c index 8da222eacbaf..aa6ba39ff370 100644 --- a/fs/compat_ioctl.c +++ b/fs/compat_ioctl.c @@ -94,7 +94,6 @@ #include #include #include -#include #include #include @@ -2392,27 +2391,6 @@ COMPATIBLE_IOCTL(USBDEVFS_SUBMITURB32) COMPATIBLE_IOCTL(USBDEVFS_REAPURB32) COMPATIBLE_IOCTL(USBDEVFS_REAPURBNDELAY32) COMPATIBLE_IOCTL(USBDEVFS_CLEAR_HALT) -/* MTD */ -COMPATIBLE_IOCTL(MEMGETINFO) -COMPATIBLE_IOCTL(MEMERASE) -COMPATIBLE_IOCTL(MEMLOCK) -COMPATIBLE_IOCTL(MEMUNLOCK) -COMPATIBLE_IOCTL(MEMGETREGIONCOUNT) -COMPATIBLE_IOCTL(MEMGETREGIONINFO) -COMPATIBLE_IOCTL(MEMSETOOBSEL) -COMPATIBLE_IOCTL(MEMGETOOBSEL) -COMPATIBLE_IOCTL(MEMGETBADBLOCK) -COMPATIBLE_IOCTL(MEMSETBADBLOCK) -COMPATIBLE_IOCTL(OTPSELECT) -COMPATIBLE_IOCTL(OTPGETREGIONCOUNT) -COMPATIBLE_IOCTL(OTPGETREGIONINFO) -COMPATIBLE_IOCTL(OTPLOCK) -COMPATIBLE_IOCTL(ECCGETLAYOUT) -COMPATIBLE_IOCTL(ECCGETSTATS) -COMPATIBLE_IOCTL(MTDFILEMODE) -COMPATIBLE_IOCTL(MEMERASE64) -COMPATIBLE_IOCTL(MEMREADOOB64) -COMPATIBLE_IOCTL(MEMWRITEOOB64) /* NBD */ ULONG_IOCTL(NBD_SET_SOCK) ULONG_IOCTL(NBD_SET_BLKSIZE) -- cgit v1.2.3 From 3c8e03166ae234d16e7871f8009638e0946d303c Mon Sep 17 00:00:00 2001 From: Yu Zhiguo Date: Sat, 16 May 2009 16:22:31 +0800 Subject: NFSv4: do exact check about attribute specified Server should return NFS4ERR_ATTRNOTSUPP if an attribute specified is not supported in current environment. Operations CREATE, NVERIFY, OPEN, SETATTR and VERIFY should do this check. This bug is found when do newpynfs tests. The names of the tests that failed are following: CR12 NVF7a NVF7b NVF7c NVF7d NVF7f NVF7r NVF7s OPEN15 VF7a VF7b VF7c VF7d VF7f VF7r VF7s Add function do_check_fattr() to do exact check: 1, Check attribute specified is supported by the NFSv4 server or not. 2, Check FATTR4_WORD0_ACL & FATTR4_WORD0_FS_LOCATIONS are supported in current environment or not. 3, Check attribute specified is writable or not. step 1 and 3 are done in function nfsd4_decode_fattr() but removed to this function now. Signed-off-by: Yu Zhiguo Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4proc.c | 95 +++++++++++++++++++++++++++++++++++++++++++++++++++--- fs/nfsd/nfs4xdr.c | 46 +++++++------------------- 2 files changed, 103 insertions(+), 38 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c index b2883e9c6381..9272e1f806fc 100644 --- a/fs/nfsd/nfs4proc.c +++ b/fs/nfsd/nfs4proc.c @@ -51,6 +51,78 @@ #define NFSDDBG_FACILITY NFSDDBG_PROC +static u32 nfsd_attrmask[] = { + NFSD_WRITEABLE_ATTRS_WORD0, + NFSD_WRITEABLE_ATTRS_WORD1, + NFSD_WRITEABLE_ATTRS_WORD2 +}; + +static u32 nfsd41_ex_attrmask[] = { + NFSD_SUPPATTR_EXCLCREAT_WORD0, + NFSD_SUPPATTR_EXCLCREAT_WORD1, + NFSD_SUPPATTR_EXCLCREAT_WORD2 +}; + +static __be32 +check_attr_support(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, + u32 *bmval, u32 *writable) +{ + struct dentry *dentry = cstate->current_fh.fh_dentry; + struct svc_export *exp = cstate->current_fh.fh_export; + + /* + * Check about attributes are supported by the NFSv4 server or not. + * According to spec, unsupported attributes return ERR_ATTRNOTSUPP. + */ + if ((bmval[0] & ~nfsd_suppattrs0(cstate->minorversion)) || + (bmval[1] & ~nfsd_suppattrs1(cstate->minorversion)) || + (bmval[2] & ~nfsd_suppattrs2(cstate->minorversion))) + return nfserr_attrnotsupp; + + /* + * Check FATTR4_WORD0_ACL & FATTR4_WORD0_FS_LOCATIONS can be supported + * in current environment or not. + */ + if (bmval[0] & FATTR4_WORD0_ACL) { + if (!IS_POSIXACL(dentry->d_inode)) + return nfserr_attrnotsupp; + } + if (bmval[0] & FATTR4_WORD0_FS_LOCATIONS) { + if (exp->ex_fslocs.locations == NULL) + return nfserr_attrnotsupp; + } + + /* + * According to spec, read-only attributes return ERR_INVAL. + */ + if (writable) { + if ((bmval[0] & ~writable[0]) || (bmval[1] & ~writable[1]) || + (bmval[2] & ~writable[2])) + return nfserr_inval; + } + + return nfs_ok; +} + +static __be32 +nfsd4_check_open_attributes(struct svc_rqst *rqstp, + struct nfsd4_compound_state *cstate, struct nfsd4_open *open) +{ + __be32 status = nfs_ok; + + if (open->op_create == NFS4_OPEN_CREATE) { + if (open->op_createmode == NFS4_CREATE_UNCHECKED + || open->op_createmode == NFS4_CREATE_GUARDED) + status = check_attr_support(rqstp, cstate, + open->op_bmval, nfsd_attrmask); + else if (open->op_createmode == NFS4_CREATE_EXCLUSIVE4_1) + status = check_attr_support(rqstp, cstate, + open->op_bmval, nfsd41_ex_attrmask); + } + + return status; +} + static inline void fh_dup2(struct svc_fh *dst, struct svc_fh *src) { @@ -225,6 +297,10 @@ nfsd4_open(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, if (status) goto out; + status = nfsd4_check_open_attributes(rqstp, cstate, open); + if (status) + goto out; + /* Openowner is now set, so sequence id will get bumped. Now we need * these checks before we do any creates: */ status = nfserr_grace; @@ -395,6 +471,11 @@ nfsd4_create(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, if (status) return status; + status = check_attr_support(rqstp, cstate, create->cr_bmval, + nfsd_attrmask); + if (status) + return status; + switch (create->cr_type) { case NF4LNK: /* ugh! we have to null-terminate the linktext, or @@ -689,6 +770,12 @@ nfsd4_setattr(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, if (status) return status; status = nfs_ok; + + status = check_attr_support(rqstp, cstate, setattr->sa_bmval, + nfsd_attrmask); + if (status) + goto out; + if (setattr->sa_acl != NULL) status = nfsd4_set_nfs4_acl(rqstp, &cstate->current_fh, setattr->sa_acl); @@ -763,10 +850,10 @@ _nfsd4_verify(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, if (status) return status; - if ((verify->ve_bmval[0] & ~nfsd_suppattrs0(cstate->minorversion)) - || (verify->ve_bmval[1] & ~nfsd_suppattrs1(cstate->minorversion)) - || (verify->ve_bmval[2] & ~nfsd_suppattrs2(cstate->minorversion))) - return nfserr_attrnotsupp; + status = check_attr_support(rqstp, cstate, verify->ve_bmval, NULL); + if (status) + return status; + if ((verify->ve_bmval[0] & FATTR4_WORD0_RDATTR_ERROR) || (verify->ve_bmval[1] & NFSD_WRITEONLY_ATTRS_WORD1)) return nfserr_inval; diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c index ab005fc637e1..254e5b21b915 100644 --- a/fs/nfsd/nfs4xdr.c +++ b/fs/nfsd/nfs4xdr.c @@ -244,20 +244,8 @@ nfsd4_decode_bitmap(struct nfsd4_compoundargs *argp, u32 *bmval) DECODE_TAIL; } -static u32 nfsd_attrmask[] = { - NFSD_WRITEABLE_ATTRS_WORD0, - NFSD_WRITEABLE_ATTRS_WORD1, - NFSD_WRITEABLE_ATTRS_WORD2 -}; - -static u32 nfsd41_ex_attrmask[] = { - NFSD_SUPPATTR_EXCLCREAT_WORD0, - NFSD_SUPPATTR_EXCLCREAT_WORD1, - NFSD_SUPPATTR_EXCLCREAT_WORD2 -}; - static __be32 -nfsd4_decode_fattr(struct nfsd4_compoundargs *argp, u32 *bmval, u32 *writable, +nfsd4_decode_fattr(struct nfsd4_compoundargs *argp, u32 *bmval, struct iattr *iattr, struct nfs4_acl **acl) { int expected_len, len = 0; @@ -270,18 +258,6 @@ nfsd4_decode_fattr(struct nfsd4_compoundargs *argp, u32 *bmval, u32 *writable, if ((status = nfsd4_decode_bitmap(argp, bmval))) return status; - /* - * According to spec, unsupported attributes return ERR_ATTRNOTSUPP; - * read-only attributes return ERR_INVAL. - */ - if ((bmval[0] & ~nfsd_suppattrs0(argp->minorversion)) || - (bmval[1] & ~nfsd_suppattrs1(argp->minorversion)) || - (bmval[2] & ~nfsd_suppattrs2(argp->minorversion))) - return nfserr_attrnotsupp; - if ((bmval[0] & ~writable[0]) || (bmval[1] & ~writable[1]) || - (bmval[2] & ~writable[2])) - return nfserr_inval; - READ_BUF(4); READ32(expected_len); @@ -414,8 +390,11 @@ nfsd4_decode_fattr(struct nfsd4_compoundargs *argp, u32 *bmval, u32 *writable, goto xdr_error; } } - BUG_ON(bmval[2]); /* no such writeable attr supported yet */ - if (len != expected_len) + if (bmval[0] & ~NFSD_WRITEABLE_ATTRS_WORD0 + || bmval[1] & ~NFSD_WRITEABLE_ATTRS_WORD1 + || bmval[2] & ~NFSD_WRITEABLE_ATTRS_WORD2) + READ_BUF(expected_len - len); + else if (len != expected_len) goto xdr_error; DECODE_TAIL; @@ -508,8 +487,8 @@ nfsd4_decode_create(struct nfsd4_compoundargs *argp, struct nfsd4_create *create if ((status = check_filename(create->cr_name, create->cr_namelen, nfserr_inval))) return status; - status = nfsd4_decode_fattr(argp, create->cr_bmval, nfsd_attrmask, - &create->cr_iattr, &create->cr_acl); + status = nfsd4_decode_fattr(argp, create->cr_bmval, &create->cr_iattr, + &create->cr_acl); if (status) goto out; @@ -672,7 +651,7 @@ nfsd4_decode_open(struct nfsd4_compoundargs *argp, struct nfsd4_open *open) case NFS4_CREATE_UNCHECKED: case NFS4_CREATE_GUARDED: status = nfsd4_decode_fattr(argp, open->op_bmval, - nfsd_attrmask, &open->op_iattr, &open->op_acl); + &open->op_iattr, &open->op_acl); if (status) goto out; break; @@ -686,8 +665,7 @@ nfsd4_decode_open(struct nfsd4_compoundargs *argp, struct nfsd4_open *open) READ_BUF(8); COPYMEM(open->op_verf.data, 8); status = nfsd4_decode_fattr(argp, open->op_bmval, - nfsd41_ex_attrmask, &open->op_iattr, - &open->op_acl); + &open->op_iattr, &open->op_acl); if (status) goto out; break; @@ -883,8 +861,8 @@ nfsd4_decode_setattr(struct nfsd4_compoundargs *argp, struct nfsd4_setattr *seta status = nfsd4_decode_stateid(argp, &setattr->sa_stateid); if (status) return status; - return nfsd4_decode_fattr(argp, setattr->sa_bmval, nfsd_attrmask, - &setattr->sa_iattr, &setattr->sa_acl); + return nfsd4_decode_fattr(argp, setattr->sa_bmval, &setattr->sa_iattr, + &setattr->sa_acl); } static __be32 -- cgit v1.2.3 From 0a93a47f042c459f0f46942c3a920e3c81878031 Mon Sep 17 00:00:00 2001 From: Yu Zhiguo Date: Tue, 19 May 2009 14:09:54 +0800 Subject: NFSv4: kill off complicated macro 'PROC' J. Bruce Fields wrote: ... > (This is extremely confusing code to track down: note that > proc->pc_decode is set to nfs4svc_decode_compoundargs() by the PROC() > macro at the end of fs/nfsd/nfs4proc.c. Which means, for example, that > grepping for nfs4svc_decode_compoundargs() gets you nowhere. Patches to > kill off that macro would be welcomed....) the macro 'PROC' is complicated and obscure, it had better be killed off in order to make the code more clear. Signed-off-by: Yu Zhiguo Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4proc.c | 34 +++++++++++++++++----------------- 1 file changed, 17 insertions(+), 17 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c index 9272e1f806fc..7c8801769a3c 100644 --- a/fs/nfsd/nfs4proc.c +++ b/fs/nfsd/nfs4proc.c @@ -1313,24 +1313,9 @@ static const char *nfsd4_op_name(unsigned opnum) return "unknown_operation"; } -#define nfs4svc_decode_voidargs NULL -#define nfs4svc_release_void NULL #define nfsd4_voidres nfsd4_voidargs -#define nfs4svc_release_compound NULL struct nfsd4_voidargs { int dummy; }; -#define PROC(name, argt, rest, relt, cache, respsize) \ - { (svc_procfunc) nfsd4_proc_##name, \ - (kxdrproc_t) nfs4svc_decode_##argt##args, \ - (kxdrproc_t) nfs4svc_encode_##rest##res, \ - (kxdrproc_t) nfs4svc_release_##relt, \ - sizeof(struct nfsd4_##argt##args), \ - sizeof(struct nfsd4_##rest##res), \ - 0, \ - cache, \ - respsize, \ - } - /* * TODO: At the present time, the NFSv4 server does not do XID caching * of requests. Implementing XID caching would not be a serious problem, @@ -1342,8 +1327,23 @@ struct nfsd4_voidargs { int dummy; }; * better XID's. */ static struct svc_procedure nfsd_procedures4[2] = { - PROC(null, void, void, void, RC_NOCACHE, 1), - PROC(compound, compound, compound, compound, RC_NOCACHE, NFSD_BUFSIZE/4) + [NFSPROC4_NULL] = { + .pc_func = (svc_procfunc) nfsd4_proc_null, + .pc_encode = (kxdrproc_t) nfs4svc_encode_voidres, + .pc_argsize = sizeof(struct nfsd4_voidargs), + .pc_ressize = sizeof(struct nfsd4_voidres), + .pc_cachetype = RC_NOCACHE, + .pc_xdrressize = 1, + }, + [NFSPROC4_COMPOUND] = { + .pc_func = (svc_procfunc) nfsd4_proc_compound, + .pc_decode = (kxdrproc_t) nfs4svc_decode_compoundargs, + .pc_encode = (kxdrproc_t) nfs4svc_encode_compoundres, + .pc_argsize = sizeof(struct nfsd4_compoundargs), + .pc_ressize = sizeof(struct nfsd4_compoundres), + .pc_cachetype = RC_NOCACHE, + .pc_xdrressize = NFSD_BUFSIZE/4, + }, }; struct svc_version nfsd_version4 = { -- cgit v1.2.3 From 8379ea31e991ed2098660954d25f64386adee65c Mon Sep 17 00:00:00 2001 From: Artem Bityutskiy Date: Fri, 29 May 2009 12:34:52 +0300 Subject: UBIFS: allow sync option in rootflags MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit When passing UBIFS parameters via kernel command line, the sync option will be passed to UBIFS as a string, not as an MS_SYNCHRONOUS flag. Teach UBIFS interpreting this flag. Reported-by: Aurélien GÉRÔME Signed-off-by: Artem Bityutskiy --- fs/ubifs/super.c | 40 +++++++++++++++++++++++++++++++++++----- 1 file changed, 35 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/ubifs/super.c b/fs/ubifs/super.c index 42b818daa162..d10fc88c7bbd 100644 --- a/fs/ubifs/super.c +++ b/fs/ubifs/super.c @@ -939,6 +939,27 @@ static const match_table_t tokens = { {Opt_err, NULL}, }; +/** + * parse_standard_option - parse a standard mount option. + * @option: the option to parse + * + * Normally, standard mount options like "sync" are passed to file-systems as + * flags. However, when a "rootflags=" kernel boot parameter is used, they may + * be present in the options string. This function tries to deal with this + * situation and parse standard options. Returns 0 if the option was not + * recognized, and the corresponding integer flag if it was. + * + * UBIFS is only interested in the "sync" option, so do not check for anything + * else. + */ +static int parse_standard_option(const char *option) +{ + ubifs_msg("parse %s", option); + if (!strcmp(option, "sync")) + return MS_SYNCHRONOUS; + return 0; +} + /** * ubifs_parse_options - parse mount parameters. * @c: UBIFS file-system description object @@ -1015,9 +1036,19 @@ static int ubifs_parse_options(struct ubifs_info *c, char *options, break; } default: - ubifs_err("unrecognized mount option \"%s\" " - "or missing value", p); - return -EINVAL; + { + unsigned long flag; + struct super_block *sb = c->vfs_sb; + + flag = parse_standard_option(p); + if (!flag) { + ubifs_err("unrecognized mount option \"%s\" " + "or missing value", p); + return -EINVAL; + } + sb->s_flags |= flag; + break; + } } } @@ -1908,6 +1939,7 @@ static int ubifs_fill_super(struct super_block *sb, void *data, int silent) INIT_LIST_HEAD(&c->orph_list); INIT_LIST_HEAD(&c->orph_new); + c->vfs_sb = sb; c->highest_inum = UBIFS_FIRST_INO; c->lhead_lnum = c->ltail_lnum = UBIFS_LOG_LNUM; @@ -1939,8 +1971,6 @@ static int ubifs_fill_super(struct super_block *sb, void *data, int silent) if (err) goto out_bdi; - c->vfs_sb = sb; - sb->s_fs_info = c; sb->s_magic = UBIFS_SUPER_MAGIC; sb->s_blocksize = UBIFS_BLOCK_SIZE; -- cgit v1.2.3 From b574480507460b8e31b8d38dd4642219fc3b9a10 Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Sat, 20 Jun 2009 23:34:44 -0400 Subject: jbd2: Remove GFP_ATOMIC kmalloc from inside spinlock critical region Fix jbd2_dev_to_name(), a function used when pretty-printting jbd2 and ext4 tracepoints. Signed-off-by: "Theodore Ts'o" --- fs/jbd2/journal.c | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c index 18bfd5dab642..7b545c3b3942 100644 --- a/fs/jbd2/journal.c +++ b/fs/jbd2/journal.c @@ -2410,6 +2410,7 @@ const char *jbd2_dev_to_name(dev_t device) int i = hash_32(device, CACHE_SIZE_BITS); char *ret; struct block_device *bd; + static struct devname_cache *new_dev; rcu_read_lock(); if (devcache[i] && devcache[i]->device == device) { @@ -2419,20 +2420,20 @@ const char *jbd2_dev_to_name(dev_t device) } rcu_read_unlock(); + new_dev = kmalloc(sizeof(struct devname_cache), GFP_KERNEL); + if (!new_dev) + return "NODEV-ALLOCFAILURE"; /* Something non-NULL */ spin_lock(&devname_cache_lock); if (devcache[i]) { if (devcache[i]->device == device) { + kfree(new_dev); ret = devcache[i]->devname; spin_unlock(&devname_cache_lock); return ret; } call_rcu(&devcache[i]->rcu, free_devcache); } - devcache[i] = kmalloc(sizeof(struct devname_cache), GFP_KERNEL); - if (!devcache[i]) { - spin_unlock(&devname_cache_lock); - return "NODEV-ALLOCFAILURE"; /* Something non-NULL */ - } + devcache[i] = new_dev; devcache[i]->device = device; bd = bdget(device); if (bd) { -- cgit v1.2.3 From f4a01017d678fe4baecf480e79d7c4f4b7ebc772 Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Sun, 5 Jul 2009 22:08:16 -0400 Subject: ext4: Fix potential reclaim deadlock when truncating partial block The ext4_block_truncate_page() function previously called grab_cache_page(), which called find_or_create_page() with the __GFP_FS flag potentially set. This could cause a deadlock if the system is low on memory and it attempts a memory reclaim, which could potentially call back into ext4. So we need to call find_or_create_page() directly, and remove the __GFP_FP flag to avoid this potential deadlock. Thanks to Roland Dreier for reporting a lockdep warning which showed this problem. [20786.363249] ================================= [20786.363257] [ INFO: inconsistent lock state ] [20786.363265] 2.6.31-2-generic #14~rbd4gitd960eea9 [20786.363270] --------------------------------- [20786.363276] inconsistent {IN-RECLAIM_FS-W} -> {RECLAIM_FS-ON-W} usage. [20786.363285] http/8397 [HC0[0]:SC0[0]:HE1:SE1] takes: [20786.363291] (jbd2_handle){+.+.?.}, at: [] jbd2_journal_start+0xdb/0x150 [20786.363314] {IN-RECLAIM_FS-W} state was registered at: [20786.363320] [] mark_irqflags+0xc6/0x1a0 [20786.363334] [] __lock_acquire+0x287/0x430 [20786.363345] [] lock_acquire+0xa5/0x150 [20786.363355] [] jbd2_journal_start+0xfa/0x150 [20786.363365] [] ext4_journal_start_sb+0x58/0x90 [20786.363377] [] ext4_delete_inode+0xc5/0x2c0 [20786.363389] [] generic_delete_inode+0xd3/0x1a0 [20786.363401] [] generic_drop_inode+0x25/0x30 [20786.363411] [] iput+0x62/0x70 [20786.363420] [] dentry_iput+0x98/0x110 [20786.363429] [] d_kill+0x50/0x80 [20786.363438] [] dput+0x95/0x180 [20786.363447] [] ecryptfs_d_release+0x2b/0x70 [20786.363459] [] d_free+0x28/0x60 [20786.363468] [] d_kill+0x68/0x80 [20786.363477] [] prune_one_dentry+0xa3/0xc0 [20786.363487] [] __shrink_dcache_sb+0x271/0x290 [20786.363497] [] prune_dcache+0x109/0x1b0 [20786.363506] [] shrink_dcache_memory+0x3f/0x50 [20786.363516] [] shrink_slab+0x12d/0x190 [20786.363527] [] balance_pgdat+0x4d7/0x640 [20786.363537] [] kswapd+0x117/0x170 [20786.363546] [] kthread+0x9e/0xb0 [20786.363558] [] child_rip+0xa/0x20 [20786.363569] [] 0xffffffffffffffff [20786.363598] irq event stamp: 15997 [20786.363603] hardirqs last enabled at (15997): [] kmem_cache_alloc+0xfd/0x1a0 [20786.363617] hardirqs last disabled at (15996): [] kmem_cache_alloc+0x61/0x1a0 [20786.363628] softirqs last enabled at (15966): [] __do_softirq+0x14a/0x220 [20786.363641] softirqs last disabled at (15861): [] call_softirq+0x1c/0x30 [20786.363651] [20786.363653] other info that might help us debug this: [20786.363660] 3 locks held by http/8397: [20786.363665] #0: (&sb->s_type->i_mutex_key#8){+.+.+.}, at: [] do_truncate+0x64/0x90 [20786.363685] #1: (&sb->s_type->i_alloc_sem_key#5){+++++.}, at: [] notify_change+0x250/0x350 [20786.363707] #2: (jbd2_handle){+.+.?.}, at: [] jbd2_journal_start+0xdb/0x150 [20786.363724] [20786.363726] stack backtrace: [20786.363734] Pid: 8397, comm: http Tainted: G C 2.6.31-2-generic #14~rbd4gitd960eea9 [20786.363741] Call Trace: [20786.363752] [] print_usage_bug+0x18c/0x1a0 [20786.363763] [] ? check_usage_backwards+0x0/0xb0 [20786.363773] [] mark_lock_irq+0xf2/0x280 [20786.363783] [] mark_lock+0x137/0x1d0 [20786.363793] [] mark_held_locks+0x6c/0xa0 [20786.363803] [] lockdep_trace_alloc+0xaf/0xe0 [20786.363813] [] __alloc_pages_nodemask+0x7c/0x180 [20786.363824] [] ? find_get_page+0x91/0xf0 [20786.363835] [] alloc_pages_current+0x87/0xd0 [20786.363845] [] __page_cache_alloc+0x67/0x70 [20786.363856] [] find_or_create_page+0x4f/0xb0 [20786.363867] [] ext4_block_truncate_page+0x3e/0x460 [20786.363876] [] ? jbd2_journal_start+0xfa/0x150 [20786.363885] [] ? jbd2_journal_start+0xdb/0x150 [20786.363895] [] ? ext4_meta_trans_blocks+0x75/0xf0 [20786.363905] [] ext4_ext_truncate+0x1bb/0x1e0 [20786.363916] [] ? unmap_mapping_range+0x75/0x290 [20786.363926] [] ext4_truncate+0x498/0x630 [20786.363938] [] ? _raw_spin_unlock+0x5e/0xb0 [20786.363947] [] ? unmap_mapping_range+0xb6/0x290 [20786.363957] [] ? trace_hardirqs_on+0xd/0x10 [20786.363966] [] ? jbd2_journal_stop+0x1f8/0x2e0 [20786.363976] [] vmtruncate+0xb0/0x110 [20786.363986] [] inode_setattr+0x35/0x170 [20786.363995] [] ext4_setattr+0x186/0x370 [20786.364005] [] notify_change+0x16b/0x350 [20786.364014] [] do_truncate+0x70/0x90 [20786.364021] [] T.657+0xeb/0x110 [20786.364021] [] sys_ftruncate+0xe/0x10 [20786.364021] [] system_call_fastpath+0x16/0x1b Reported-by: Roland Dreier Signed-off-by: "Theodore Ts'o" --- fs/ext4/inode.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 60a26f3a6f8b..9760ba09275e 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -3583,7 +3583,8 @@ int ext4_block_truncate_page(handle_t *handle, struct page *page; int err = 0; - page = grab_cache_page(mapping, from >> PAGE_CACHE_SHIFT); + page = find_or_create_page(mapping, from >> PAGE_CACHE_SHIFT, + mapping_gfp_mask(mapping) & ~__GFP_FS); if (!page) return -EINVAL; -- cgit v1.2.3 From 089ceecc1ea4a69ed8bcc5c7c7b96ce487e26b33 Mon Sep 17 00:00:00 2001 From: Eric Sandeen Date: Sun, 5 Jul 2009 22:17:31 -0400 Subject: ext4: mark several more functions in mballoc.c as noinline Ted noticed a stack-deep callchain through writepages->ext4_mb_regular_allocator->ext4_mb_init_cache->submit_bh ... With all the static functions in mballoc.c, gcc helpfully inlines for us, and we get something like this: ext4_mb_regular_allocator (232 bytes stack) ext4_mb_init_cache (232 bytes stack) submit_bh (starts 464 deeper) the 2 ext4 functions here get several others inlined; by telling gcc not to inline them, we can save stack space for when we head off into submit_bh land and associated block layer callchains. The following noinlined functions are only called once, so this won't impact any other callchains: ext4_mb_regular_allocator (104) (was 232) ext4_mb_find_by_goal (56) (noinlined) ext4_mb_init_group (24) (noinlined) ext4_mb_init_cache (136) (was 232) ext4_mb_generate_buddy (88) (noinlined) ext4_mb_generate_from_pa (40) (noinlined) submit_bh ext4_mb_simple_scan_group (24) (noinlined) ext4_mb_scan_aligned (56) (noinlined) ext4_mb_complex_scan_group (40) (noinlined) ext4_mb_try_best_found (24) (noinlined) now when we head off into submit_bh() we're only 264 bytes deeper in stack than when we entered ext4_mb_regular_allocator() (vs. 464 bytes before). Every 200 bytes helps. :) Signed-off-by: Eric Sandeen Signed-off-by: "Theodore Ts'o" --- fs/ext4/mballoc.c | 24 ++++++++++++++++-------- 1 file changed, 16 insertions(+), 8 deletions(-) (limited to 'fs') diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index 519a0a686d94..4a45efabb203 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -657,7 +657,8 @@ static void ext4_mb_mark_free_simple(struct super_block *sb, } } -static void ext4_mb_generate_buddy(struct super_block *sb, +static noinline_for_stack +void ext4_mb_generate_buddy(struct super_block *sb, void *buddy, void *bitmap, ext4_group_t group) { struct ext4_group_info *grp = ext4_get_group_info(sb, group); @@ -1480,7 +1481,8 @@ static void ext4_mb_measure_extent(struct ext4_allocation_context *ac, ext4_mb_check_limits(ac, e4b, 0); } -static int ext4_mb_try_best_found(struct ext4_allocation_context *ac, +static noinline_for_stack +int ext4_mb_try_best_found(struct ext4_allocation_context *ac, struct ext4_buddy *e4b) { struct ext4_free_extent ex = ac->ac_b_ex; @@ -1507,7 +1509,8 @@ static int ext4_mb_try_best_found(struct ext4_allocation_context *ac, return 0; } -static int ext4_mb_find_by_goal(struct ext4_allocation_context *ac, +static noinline_for_stack +int ext4_mb_find_by_goal(struct ext4_allocation_context *ac, struct ext4_buddy *e4b) { ext4_group_t group = ac->ac_g_ex.fe_group; @@ -1566,7 +1569,8 @@ static int ext4_mb_find_by_goal(struct ext4_allocation_context *ac, * The routine scans buddy structures (not bitmap!) from given order * to max order and tries to find big enough chunk to satisfy the req */ -static void ext4_mb_simple_scan_group(struct ext4_allocation_context *ac, +static noinline_for_stack +void ext4_mb_simple_scan_group(struct ext4_allocation_context *ac, struct ext4_buddy *e4b) { struct super_block *sb = ac->ac_sb; @@ -1609,7 +1613,8 @@ static void ext4_mb_simple_scan_group(struct ext4_allocation_context *ac, * In order to optimize scanning, caller must pass number of * free blocks in the group, so the routine can know upper limit. */ -static void ext4_mb_complex_scan_group(struct ext4_allocation_context *ac, +static noinline_for_stack +void ext4_mb_complex_scan_group(struct ext4_allocation_context *ac, struct ext4_buddy *e4b) { struct super_block *sb = ac->ac_sb; @@ -1668,7 +1673,8 @@ static void ext4_mb_complex_scan_group(struct ext4_allocation_context *ac, * we try to find stripe-aligned chunks for stripe-size requests * XXX should do so at least for multiples of stripe size as well */ -static void ext4_mb_scan_aligned(struct ext4_allocation_context *ac, +static noinline_for_stack +void ext4_mb_scan_aligned(struct ext4_allocation_context *ac, struct ext4_buddy *e4b) { struct super_block *sb = ac->ac_sb; @@ -1831,7 +1837,8 @@ void ext4_mb_put_buddy_cache_lock(struct super_block *sb, } -static int ext4_mb_init_group(struct super_block *sb, ext4_group_t group) +static noinline_for_stack +int ext4_mb_init_group(struct super_block *sb, ext4_group_t group) { int ret; @@ -3457,7 +3464,8 @@ static void ext4_mb_generate_from_freelist(struct super_block *sb, void *bitmap, * used in in-core bitmap. buddy must be generated from this bitmap * Need to be called with ext4 group lock held */ -static void ext4_mb_generate_from_pa(struct super_block *sb, void *bitmap, +static noinline_for_stack +void ext4_mb_generate_from_pa(struct super_block *sb, void *bitmap, ext4_group_t group) { struct ext4_group_info *grp = ext4_get_group_info(sb, group); -- cgit v1.2.3 From 726447d803802cd0be8f62d17c4a34421781b938 Mon Sep 17 00:00:00 2001 From: Eric Sandeen Date: Mon, 13 Jul 2009 10:24:17 -0400 Subject: ext4: naturally align struct ext4_allocation_request As Ted noted, the ext4_allocation_request isn't well aligned. Looking at it with pahole we're wasting space on 64-bit arches: struct ext4_allocation_request { struct inode * inode; /* 0 8 */ ext4_lblk_t logical; /* 8 4 */ /* XXX 4 bytes hole, try to pack */ ext4_fsblk_t goal; /* 16 8 */ ext4_lblk_t lleft; /* 24 4 */ /* XXX 4 bytes hole, try to pack */ ext4_fsblk_t pleft; /* 32 8 */ ext4_lblk_t lright; /* 40 4 */ /* XXX 4 bytes hole, try to pack */ ext4_fsblk_t pright; /* 48 8 */ unsigned int len; /* 56 4 */ unsigned int flags; /* 60 4 */ /* --- cacheline 1 boundary (64 bytes) --- */ /* size: 64, cachelines: 1, members: 9 */ /* sum members: 52, holes: 3, sum holes: 12 */ }; Grouping 32-bit members together closes these holes and shrinks the structure by 12 bytes. which is important since ext4 can get on the hairy edge of stack overruns. Signed-off-by: Eric Sandeen Signed-off-by: "Theodore Ts'o" --- fs/ext4/ext4.h | 14 +++++++------- 1 file changed, 7 insertions(+), 7 deletions(-) (limited to 'fs') diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 0ddf7e55abe1..9714db393efe 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -93,20 +93,20 @@ typedef unsigned int ext4_group_t; struct ext4_allocation_request { /* target inode for block we're allocating */ struct inode *inode; + /* how many blocks we want to allocate */ + unsigned int len; /* logical block in target inode */ ext4_lblk_t logical; - /* phys. target (a hint) */ - ext4_fsblk_t goal; /* the closest logical allocated block to the left */ ext4_lblk_t lleft; - /* phys. block for ^^^ */ - ext4_fsblk_t pleft; /* the closest logical allocated block to the right */ ext4_lblk_t lright; - /* phys. block for ^^^ */ + /* phys. target (a hint) */ + ext4_fsblk_t goal; + /* phys. block for the closest logical allocated block to the left */ + ext4_fsblk_t pleft; + /* phys. block for the closest logical allocated block to the right */ ext4_fsblk_t pright; - /* how many blocks we want to allocate */ - unsigned int len; /* flags. see above EXT4_MB_HINT_* */ unsigned int flags; }; -- cgit v1.2.3 From 3e03f9ca6a2599db1823bb0ea24e0845219a0e69 Mon Sep 17 00:00:00 2001 From: Jesper Dangaard Brouer Date: Sun, 5 Jul 2009 22:29:27 -0400 Subject: ext4: Use rcu_barrier() on module unload. The ext4 module uses rcu_call() thus it should use rcu_barrier()on module unload. The kmem cache ext4_pspace_cachep is sometimes free'ed using call_rcu() callbacks. Thus, we must wait for completion of call_rcu() before doing kmem_cache_destroy(). Signed-off-by: Jesper Dangaard Brouer Signed-off-by: "Theodore Ts'o" --- fs/ext4/mballoc.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index 4a45efabb203..2fcaf286f1de 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -2909,7 +2909,11 @@ int __init init_ext4_mballoc(void) void exit_ext4_mballoc(void) { - /* XXX: synchronize_rcu(); */ + /* + * Wait for completion of call_rcu()'s on ext4_pspace_cachep + * before destroying the slab cache. + */ + rcu_barrier(); kmem_cache_destroy(ext4_pspace_cachep); kmem_cache_destroy(ext4_ac_cachep); kmem_cache_destroy(ext4_free_ext_cachep); -- cgit v1.2.3 From f91d1d04171026e56c7e343ee3cdcc801dd85cfb Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Mon, 13 Jul 2009 16:16:20 -0400 Subject: jbd2: Fix a race between checkpointing code and journal_get_write_access() The following race can happen: CPU1 CPU2 checkpointing code checks the buffer, adds it to an array for writeback do_get_write_access() ... lock_buffer() unlock_buffer() flush_batch() submits the buffer for IO __jbd2_journal_file_buffer() So a buffer under writeout is returned from do_get_write_access(). Since the filesystem code relies on the fact that journaled buffers cannot be written out, it does not take the buffer lock and so it can modify buffer while it is under writeout. That can lead to a filesystem corruption if we crash at the right moment. We fix the problem by clearing the buffer dirty bit under buffer_lock even if the buffer is on BJ_None list. Actually, we clear the dirty bit regardless the list the buffer is in and warn about the fact if the buffer is already journalled. Thanks for spotting the problem goes to dingdinghua . Reported-by: dingdinghua Signed-off-by: Jan Kara Signed-off-by: "Theodore Ts'o" --- fs/jbd2/transaction.c | 68 ++++++++++++++++++++++++++------------------------- 1 file changed, 35 insertions(+), 33 deletions(-) (limited to 'fs') diff --git a/fs/jbd2/transaction.c b/fs/jbd2/transaction.c index 494501edba6b..6213ac728f30 100644 --- a/fs/jbd2/transaction.c +++ b/fs/jbd2/transaction.c @@ -499,34 +499,15 @@ void jbd2_journal_unlock_updates (journal_t *journal) wake_up(&journal->j_wait_transaction_locked); } -/* - * Report any unexpected dirty buffers which turn up. Normally those - * indicate an error, but they can occur if the user is running (say) - * tune2fs to modify the live filesystem, so we need the option of - * continuing as gracefully as possible. # - * - * The caller should already hold the journal lock and - * j_list_lock spinlock: most callers will need those anyway - * in order to probe the buffer's journaling state safely. - */ -static void jbd_unexpected_dirty_buffer(struct journal_head *jh) +static void warn_dirty_buffer(struct buffer_head *bh) { - int jlist; - - /* If this buffer is one which might reasonably be dirty - * --- ie. data, or not part of this journal --- then - * we're OK to leave it alone, but otherwise we need to - * move the dirty bit to the journal's own internal - * JBDDirty bit. */ - jlist = jh->b_jlist; + char b[BDEVNAME_SIZE]; - if (jlist == BJ_Metadata || jlist == BJ_Reserved || - jlist == BJ_Shadow || jlist == BJ_Forget) { - struct buffer_head *bh = jh2bh(jh); - - if (test_clear_buffer_dirty(bh)) - set_buffer_jbddirty(bh); - } + printk(KERN_WARNING + "JBD: Spotted dirty metadata buffer (dev = %s, blocknr = %llu). " + "There's a risk of filesystem corruption in case of system " + "crash.\n", + bdevname(bh->b_bdev, b), (unsigned long long)bh->b_blocknr); } /* @@ -593,14 +574,16 @@ repeat: if (jh->b_next_transaction) J_ASSERT_JH(jh, jh->b_next_transaction == transaction); + warn_dirty_buffer(bh); } /* * In any case we need to clean the dirty flag and we must * do it under the buffer lock to be sure we don't race * with running write-out. */ - JBUFFER_TRACE(jh, "Unexpected dirty buffer"); - jbd_unexpected_dirty_buffer(jh); + JBUFFER_TRACE(jh, "Journalling dirty buffer"); + clear_buffer_dirty(bh); + set_buffer_jbddirty(bh); } unlock_buffer(bh); @@ -843,6 +826,15 @@ int jbd2_journal_get_create_access(handle_t *handle, struct buffer_head *bh) J_ASSERT_JH(jh, buffer_locked(jh2bh(jh))); if (jh->b_transaction == NULL) { + /* + * Previous jbd2_journal_forget() could have left the buffer + * with jbddirty bit set because it was being committed. When + * the commit finished, we've filed the buffer for + * checkpointing and marked it dirty. Now we are reallocating + * the buffer so the transaction freeing it must have + * committed and so it's safe to clear the dirty bit. + */ + clear_buffer_dirty(jh2bh(jh)); jh->b_transaction = transaction; /* first access by this transaction */ @@ -1644,8 +1636,13 @@ static int __dispose_buffer(struct journal_head *jh, transaction_t *transaction) if (jh->b_cp_transaction) { JBUFFER_TRACE(jh, "on running+cp transaction"); + /* + * We don't want to write the buffer anymore, clear the + * bit so that we don't confuse checks in + * __journal_file_buffer + */ + clear_buffer_dirty(bh); __jbd2_journal_file_buffer(jh, transaction, BJ_Forget); - clear_buffer_jbddirty(bh); may_free = 0; } else { JBUFFER_TRACE(jh, "on running transaction"); @@ -1896,12 +1893,17 @@ void __jbd2_journal_file_buffer(struct journal_head *jh, if (jh->b_transaction && jh->b_jlist == jlist) return; - /* The following list of buffer states needs to be consistent - * with __jbd_unexpected_dirty_buffer()'s handling of dirty - * state. */ - if (jlist == BJ_Metadata || jlist == BJ_Reserved || jlist == BJ_Shadow || jlist == BJ_Forget) { + /* + * For metadata buffers, we track dirty bit in buffer_jbddirty + * instead of buffer_dirty. We should not see a dirty bit set + * here because we clear it in do_get_write_access but e.g. + * tune2fs can modify the sb and set the dirty bit at any time + * so we try to gracefully handle that. + */ + if (buffer_dirty(bh)) + warn_dirty_buffer(bh); if (test_clear_buffer_dirty(bh) || test_clear_buffer_jbddirty(bh)) was_dirty = 1; -- cgit v1.2.3 From ffacfa7a79d6c00624196b2d13b0a7f72f2b8227 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Mon, 13 Jul 2009 16:22:22 -0400 Subject: ext4: Fix truncation of symlinks after failed write Contents of long symlinks is written via standard write methods. So when the write fails, we add inode to orphan list. But symlinks don't have .truncate method defined so nobody properly removes them from the on disk orphan list. Fix this by calling ext4_truncate() directly instead of calling vmtruncate() (which is saner anyway since we don't need anything vmtruncate() does except from calling .truncate in these paths). We also add inode to orphan list only if ext4_can_truncate() is true (currently, it can be false for symlinks when there are no blocks allocated) - otherwise orphan list processing will complain and ext4_truncate() will not remove inode from on-disk orphan list. Signed-off-by: Jan Kara Signed-off-by: "Theodore Ts'o" --- fs/ext4/inode.c | 26 +++++++++++++------------- 1 file changed, 13 insertions(+), 13 deletions(-) (limited to 'fs') diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 9760ba09275e..ff2afc1909b3 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -1513,14 +1513,14 @@ retry: * Add inode to orphan list in case we crash before * truncate finishes */ - if (pos + len > inode->i_size) + if (pos + len > inode->i_size && ext4_can_truncate(inode)) ext4_orphan_add(handle, inode); ext4_journal_stop(handle); if (pos + len > inode->i_size) { - vmtruncate(inode, inode->i_size); + ext4_truncate(inode); /* - * If vmtruncate failed early the inode might + * If truncate failed early the inode might * still be on the orphan list; we need to * make sure the inode is removed from the * orphan list in that case. @@ -1614,7 +1614,7 @@ static int ext4_ordered_write_end(struct file *file, ret2 = ext4_generic_write_end(file, mapping, pos, len, copied, page, fsdata); copied = ret2; - if (pos + len > inode->i_size) + if (pos + len > inode->i_size && ext4_can_truncate(inode)) /* if we have allocated more blocks and copied * less. We will have blocks allocated outside * inode->i_size. So truncate them @@ -1628,9 +1628,9 @@ static int ext4_ordered_write_end(struct file *file, ret = ret2; if (pos + len > inode->i_size) { - vmtruncate(inode, inode->i_size); + ext4_truncate(inode); /* - * If vmtruncate failed early the inode might still be + * If truncate failed early the inode might still be * on the orphan list; we need to make sure the inode * is removed from the orphan list in that case. */ @@ -1655,7 +1655,7 @@ static int ext4_writeback_write_end(struct file *file, ret2 = ext4_generic_write_end(file, mapping, pos, len, copied, page, fsdata); copied = ret2; - if (pos + len > inode->i_size) + if (pos + len > inode->i_size && ext4_can_truncate(inode)) /* if we have allocated more blocks and copied * less. We will have blocks allocated outside * inode->i_size. So truncate them @@ -1670,9 +1670,9 @@ static int ext4_writeback_write_end(struct file *file, ret = ret2; if (pos + len > inode->i_size) { - vmtruncate(inode, inode->i_size); + ext4_truncate(inode); /* - * If vmtruncate failed early the inode might still be + * If truncate failed early the inode might still be * on the orphan list; we need to make sure the inode * is removed from the orphan list in that case. */ @@ -1722,7 +1722,7 @@ static int ext4_journalled_write_end(struct file *file, unlock_page(page); page_cache_release(page); - if (pos + len > inode->i_size) + if (pos + len > inode->i_size && ext4_can_truncate(inode)) /* if we have allocated more blocks and copied * less. We will have blocks allocated outside * inode->i_size. So truncate them @@ -1733,9 +1733,9 @@ static int ext4_journalled_write_end(struct file *file, if (!ret) ret = ret2; if (pos + len > inode->i_size) { - vmtruncate(inode, inode->i_size); + ext4_truncate(inode); /* - * If vmtruncate failed early the inode might still be + * If truncate failed early the inode might still be * on the orphan list; we need to make sure the inode * is removed from the orphan list in that case. */ @@ -2907,7 +2907,7 @@ retry: * i_size_read because we hold i_mutex. */ if (pos + len > inode->i_size) - vmtruncate(inode, inode->i_size); + ext4_truncate(inode); } if (ret == -ENOSPC && ext4_should_retry_alloc(inode->i_sb, &retries)) -- cgit v1.2.3 From 5887e98b609e96ce61ee0528cf94a2bfdc809dd7 Mon Sep 17 00:00:00 2001 From: "Aneesh Kumar K.V" Date: Sun, 5 Jul 2009 23:12:04 -0400 Subject: ext4: Calculate required journal credits for inserting an extent properly When we have space in the extent tree leaf node we should be able to insert the extent with much less journal credits. The code was doing proper calculation but missed a return statement. Reported-by: Andreas Dilger Signed-off-by: Aneesh Kumar K.V Signed-off-by: "Theodore Ts'o" --- fs/ext4/extents.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index 50322a09bd01..73ebfb44ad75 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c @@ -1977,6 +1977,7 @@ int ext4_ext_calc_credits_for_single_extent(struct inode *inode, int nrblocks, */ /* 1 bitmap, 1 block group descriptor */ ret = 2 + EXT4_META_TRANS_BLOCKS(inode->i_sb); + return ret; } } -- cgit v1.2.3 From 5adfee9c17314c1411095c23191c3cb0c2d25f9f Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Wed, 8 Jul 2009 17:11:24 -0400 Subject: ext4: fix no journal corruption with locale-gen If there is no journal, ext4_should_writeback_data() should return TRUE. This will fix ext4_set_aops() to set ext4_da_ops in the case of delayed allocation; otherwise ext4_journaled_aops gets used by default, which doesn't handle delayed allocation properly. The advantage of using ext4_should_writeback_data() approach is that it should handle nobh better as well. Thanks to Curt Wohlgemuth for investigating this problem, and Aneesh Kumar for suggesting this approach. Signed-off-by: "Theodore Ts'o" --- fs/ext4/ext4_jbd2.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/ext4/ext4_jbd2.h b/fs/ext4/ext4_jbd2.h index be2f426f6805..d574a85aca56 100644 --- a/fs/ext4/ext4_jbd2.h +++ b/fs/ext4/ext4_jbd2.h @@ -281,10 +281,10 @@ static inline int ext4_should_order_data(struct inode *inode) static inline int ext4_should_writeback_data(struct inode *inode) { - if (EXT4_JOURNAL(inode) == NULL) - return 0; if (!S_ISREG(inode->i_mode)) return 0; + if (EXT4_JOURNAL(inode) == NULL) + return 1; if (EXT4_I(inode)->i_flags & EXT4_JOURNAL_DATA_FL) return 0; if (test_opt(inode->i_sb, DATA_FLAGS) == EXT4_MOUNT_WRITEBACK_DATA) -- cgit v1.2.3 From e6462869e4fd88be5141a356ee0c28d8067340cc Mon Sep 17 00:00:00 2001 From: Johann Lombardi Date: Sun, 5 Jul 2009 23:45:11 -0400 Subject: ext4: Fix goal inum check in the inode allocator The goal inode is specificed by inode number which belongs to [1; s_inodes_count]. Signed-off-by: Johann Lombardi Signed-off-by: "Theodore Ts'o" --- fs/ext4/ialloc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ext4/ialloc.c b/fs/ext4/ialloc.c index 2f645732e3b7..29e6dc7299b8 100644 --- a/fs/ext4/ialloc.c +++ b/fs/ext4/ialloc.c @@ -833,7 +833,7 @@ struct inode *ext4_new_inode(handle_t *handle, struct inode *dir, int mode, if (!goal) goal = sbi->s_inode_goal; - if (goal && goal < le32_to_cpu(sbi->s_es->s_inodes_count)) { + if (goal && goal <= le32_to_cpu(sbi->s_es->s_inodes_count)) { group = (goal - 1) / EXT4_INODES_PER_GROUP(sb); ino = (goal - 1) % EXT4_INODES_PER_GROUP(sb); ret2 = 0; -- cgit v1.2.3 From b767e78a179e5ab30fdbff1686d074ac270471eb Mon Sep 17 00:00:00 2001 From: "Aneesh Kumar K.V" Date: Thu, 4 Jun 2009 08:06:06 -0400 Subject: ext4: Don't look at buffer_heads outside i_size. Buffer heads outside i_size will be unmapped. So when we are doing "walk_page_buffers" limit ourself to i_size. Signed-off-by: Aneesh Kumar K.V Reviewed-by: Josef Bacik Acked-by: Jan Kara Signed-off-by: "Theodore Ts'o" ---- --- fs/ext4/inode.c | 28 ++++++++++++++++------------ 1 file changed, 16 insertions(+), 12 deletions(-) (limited to 'fs') diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index ff2afc1909b3..b87b68cd3241 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -2578,7 +2578,7 @@ static int ext4_da_writepage(struct page *page, * all are mapped and non delay. We don't want to * do block allocation here. */ - ret = block_prepare_write(page, 0, PAGE_CACHE_SIZE, + ret = block_prepare_write(page, 0, len, noalloc_get_block_write); if (!ret) { page_bufs = page_buffers(page); @@ -2600,7 +2600,7 @@ static int ext4_da_writepage(struct page *page, return 0; } /* now mark the buffer_heads as dirty and uptodate */ - block_commit_write(page, 0, PAGE_CACHE_SIZE); + block_commit_write(page, 0, len); } if (test_opt(inode->i_sb, NOBH) && ext4_should_writeback_data(inode)) @@ -3246,6 +3246,8 @@ static int ext4_normal_writepage(struct page *page, static int __ext4_journalled_writepage(struct page *page, struct writeback_control *wbc) { + loff_t size; + unsigned int len; struct address_space *mapping = page->mapping; struct inode *inode = mapping->host; struct buffer_head *page_bufs; @@ -3253,14 +3255,17 @@ static int __ext4_journalled_writepage(struct page *page, int ret = 0; int err; - ret = block_prepare_write(page, 0, PAGE_CACHE_SIZE, - noalloc_get_block_write); + size = i_size_read(inode); + if (page->index == size >> PAGE_CACHE_SHIFT) + len = size & ~PAGE_CACHE_MASK; + else + len = PAGE_CACHE_SIZE; + ret = block_prepare_write(page, 0, len, noalloc_get_block_write); if (ret != 0) goto out_unlock; page_bufs = page_buffers(page); - walk_page_buffers(handle, page_bufs, 0, PAGE_CACHE_SIZE, NULL, - bget_one); + walk_page_buffers(handle, page_bufs, 0, len, NULL, bget_one); /* As soon as we unlock the page, it can go away, but we have * references to buffers so we are safe */ unlock_page(page); @@ -3271,19 +3276,18 @@ static int __ext4_journalled_writepage(struct page *page, goto out; } - ret = walk_page_buffers(handle, page_bufs, 0, - PAGE_CACHE_SIZE, NULL, do_journal_get_write_access); + ret = walk_page_buffers(handle, page_bufs, 0, len, NULL, + do_journal_get_write_access); - err = walk_page_buffers(handle, page_bufs, 0, - PAGE_CACHE_SIZE, NULL, write_end_fn); + err = walk_page_buffers(handle, page_bufs, 0, len, NULL, + write_end_fn); if (ret == 0) ret = err; err = ext4_journal_stop(handle); if (!ret) ret = err; - walk_page_buffers(handle, page_bufs, 0, - PAGE_CACHE_SIZE, NULL, bput_one); + walk_page_buffers(handle, page_bufs, 0, len, NULL, bput_one); EXT4_I(inode)->i_state |= EXT4_STATE_JDATA; goto out; -- cgit v1.2.3 From 3f36406f26437afae9f43cc6dcfc264143e21ed0 Mon Sep 17 00:00:00 2001 From: Artem Bityutskiy Date: Fri, 29 May 2009 20:16:27 +0300 Subject: UBIFS: do not forget to register BDI device Reviewed-by: Jens Axboe Signed-off-by: Artem Bityutskiy --- fs/ubifs/super.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'fs') diff --git a/fs/ubifs/super.c b/fs/ubifs/super.c index d10fc88c7bbd..b9b051a4c01e 100644 --- a/fs/ubifs/super.c +++ b/fs/ubifs/super.c @@ -1966,6 +1966,9 @@ static int ubifs_fill_super(struct super_block *sb, void *data, int silent) err = bdi_init(&c->bdi); if (err) goto out_close; + err = bdi_register(&c->bdi, NULL, "ubifs"); + if (err) + goto out_bdi; err = ubifs_parse_options(c, data, 0); if (err) -- cgit v1.2.3 From f2c5dbd7b7396457efc114f825acfdd4db4608f8 Mon Sep 17 00:00:00 2001 From: Artem Bityutskiy Date: Thu, 28 May 2009 16:24:15 +0300 Subject: UBIFS: start using hrtimers UBIFS uses timers for write-buffer write-back. It is not crucial for us to write-back exactly on time. We are fine to write-back a little earlier or later. And this means we may optimize UBIFS timer so that it could be groped with a close timer event, so that the CPU would not be waken up just to do the write back. This is optimization to lessen power consumption, which is important in embedded devices UBIFS is used for. hrtimers have a nice feature: they are effectively range timers, and we may defind the soft and hard limits for it. Standard timers do not have these feature. They may only be made deferrable, but this means there is effectively no hard limit. So, we will better use hrtimers. Signed-off-by: Artem Bityutskiy --- fs/ubifs/io.c | 34 +++++++++++++++++++++------------- fs/ubifs/super.c | 6 +++--- fs/ubifs/ubifs.h | 13 ++++++++----- 3 files changed, 32 insertions(+), 21 deletions(-) (limited to 'fs') diff --git a/fs/ubifs/io.c b/fs/ubifs/io.c index e8e632a1dcdf..bc5857199ec2 100644 --- a/fs/ubifs/io.c +++ b/fs/ubifs/io.c @@ -293,13 +293,14 @@ void ubifs_prep_grp_node(struct ubifs_info *c, void *node, int len, int last) * * This function is called when the write-buffer timer expires. */ -static void wbuf_timer_callback_nolock(unsigned long data) +static enum hrtimer_restart wbuf_timer_callback_nolock(struct hrtimer *timer) { - struct ubifs_wbuf *wbuf = (struct ubifs_wbuf *)data; + struct ubifs_wbuf *wbuf = container_of(timer, struct ubifs_wbuf, timer); wbuf->need_sync = 1; wbuf->c->need_wbuf_sync = 1; ubifs_wake_up_bgt(wbuf->c); + return HRTIMER_NORESTART; } /** @@ -308,13 +309,12 @@ static void wbuf_timer_callback_nolock(unsigned long data) */ static void new_wbuf_timer_nolock(struct ubifs_wbuf *wbuf) { - ubifs_assert(!timer_pending(&wbuf->timer)); + ubifs_assert(!hrtimer_active(&wbuf->timer)); - if (!wbuf->timeout) + if (!ktime_to_ns(wbuf->softlimit)) return; - - wbuf->timer.expires = jiffies + wbuf->timeout; - add_timer(&wbuf->timer); + hrtimer_start_range_ns(&wbuf->timer, wbuf->softlimit, wbuf->delta, + HRTIMER_MODE_REL); } /** @@ -329,7 +329,7 @@ static void cancel_wbuf_timer_nolock(struct ubifs_wbuf *wbuf) * should be canceled. */ wbuf->need_sync = 0; - del_timer(&wbuf->timer); + hrtimer_cancel(&wbuf->timer); } /** @@ -825,6 +825,7 @@ out: int ubifs_wbuf_init(struct ubifs_info *c, struct ubifs_wbuf *wbuf) { size_t size; + ktime_t hardlimit; wbuf->buf = kmalloc(c->min_io_size, GFP_KERNEL); if (!wbuf->buf) @@ -845,14 +846,21 @@ int ubifs_wbuf_init(struct ubifs_info *c, struct ubifs_wbuf *wbuf) wbuf->sync_callback = NULL; mutex_init(&wbuf->io_mutex); spin_lock_init(&wbuf->lock); - wbuf->c = c; - init_timer(&wbuf->timer); - wbuf->timer.function = wbuf_timer_callback_nolock; - wbuf->timer.data = (unsigned long)wbuf; - wbuf->timeout = DEFAULT_WBUF_TIMEOUT; wbuf->next_ino = 0; + hrtimer_init(&wbuf->timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); + wbuf->timer.function = wbuf_timer_callback_nolock; + /* + * Make write-buffer soft limit to be 20% of the hard limit. The + * write-buffer timer is allowed to expire any time between the soft + * and hard limits. + */ + hardlimit = ktime_set(DEFAULT_WBUF_TIMEOUT_SECS, 0); + wbuf->delta = (DEFAULT_WBUF_TIMEOUT_SECS * NSEC_PER_SEC) * 2 / 10; + wbuf->softlimit = ktime_sub_ns(hardlimit, wbuf->delta); + hrtimer_set_expires_range_ns(&wbuf->timer, wbuf->softlimit, + wbuf->delta); return 0; } diff --git a/fs/ubifs/super.c b/fs/ubifs/super.c index b9b051a4c01e..91c91cb7a599 100644 --- a/fs/ubifs/super.c +++ b/fs/ubifs/super.c @@ -799,7 +799,7 @@ static int alloc_wbufs(struct ubifs_info *c) * does not need to be synchronized by timer. */ c->jheads[GCHD].wbuf.dtype = UBI_LONGTERM; - c->jheads[GCHD].wbuf.timeout = 0; + c->jheads[GCHD].wbuf.softlimit = ktime_set(0, 0); return 0; } @@ -1695,7 +1695,7 @@ static void ubifs_remount_ro(struct ubifs_info *c) for (i = 0; i < c->jhead_cnt; i++) { ubifs_wbuf_sync(&c->jheads[i].wbuf); - del_timer_sync(&c->jheads[i].wbuf.timer); + hrtimer_cancel(&c->jheads[i].wbuf.timer); } c->mst_node->flags &= ~cpu_to_le32(UBIFS_MST_DIRTY); @@ -1755,7 +1755,7 @@ static void ubifs_put_super(struct super_block *sb) if (c->jheads) for (i = 0; i < c->jhead_cnt; i++) { ubifs_wbuf_sync(&c->jheads[i].wbuf); - del_timer_sync(&c->jheads[i].wbuf.timer); + hrtimer_cancel(&c->jheads[i].wbuf.timer); } /* diff --git a/fs/ubifs/ubifs.h b/fs/ubifs/ubifs.h index 0a8341e14088..1bf01d820066 100644 --- a/fs/ubifs/ubifs.h +++ b/fs/ubifs/ubifs.h @@ -95,8 +95,8 @@ */ #define BGT_NAME_PATTERN "ubifs_bgt%d_%d" -/* Default write-buffer synchronization timeout (5 secs) */ -#define DEFAULT_WBUF_TIMEOUT (5 * HZ) +/* Default write-buffer synchronization timeout in seconds */ +#define DEFAULT_WBUF_TIMEOUT_SECS 5 /* Maximum possible inode number (only 32-bit inodes are supported now) */ #define MAX_INUM 0xFFFFFFFF @@ -650,8 +650,10 @@ typedef int (*ubifs_lpt_scan_callback)(struct ubifs_info *c, * @io_mutex: serializes write-buffer I/O * @lock: serializes @buf, @lnum, @offs, @avail, @used, @next_ino and @inodes * fields + * @softlimit: soft write-buffer timeout interval + * @delta: hard and soft timeouts delta (the timer expire inteval is @softlimit + * and @softlimit + @delta) * @timer: write-buffer timer - * @timeout: timer expire interval in jiffies * @need_sync: it is set if its timer expired and needs sync * @next_ino: points to the next position of the following inode number * @inodes: stores the inode numbers of the nodes which are in wbuf @@ -678,8 +680,9 @@ struct ubifs_wbuf { int (*sync_callback)(struct ubifs_info *c, int lnum, int free, int pad); struct mutex io_mutex; spinlock_t lock; - struct timer_list timer; - int timeout; + ktime_t softlimit; + unsigned long long delta; + struct hrtimer timer; int need_sync; int next_ino; ino_t *inodes; -- cgit v1.2.3 From 8616e0fc1e27295316f9821a883f0e9fa6f8200f Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Thu, 11 Jun 2009 10:27:28 -0400 Subject: cifs: remove unneeded NULL checks from cifs_show_options show_options is always called with the namespace_sem held. Therefore we don't need to worry about the vfsmount being NULL, or it vanishing while the function is running. By the same token, there's no need to worry about the superblock, tcon, smb or tcp sessions being NULL on entry. Signed-off-by: Jeff Layton Signed-off-by: Steve French --- fs/cifs/cifsfs.c | 130 +++++++++++++++++++++++++------------------------------ 1 file changed, 59 insertions(+), 71 deletions(-) (limited to 'fs') diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c index 0d92114195ab..3f121fbd6c56 100644 --- a/fs/cifs/cifsfs.c +++ b/fs/cifs/cifsfs.c @@ -346,80 +346,68 @@ cifs_show_options(struct seq_file *s, struct vfsmount *m) struct TCP_Server_Info *server; cifs_sb = CIFS_SB(m->mnt_sb); + tcon = cifs_sb->tcon; - if (cifs_sb) { - tcon = cifs_sb->tcon; - if (tcon) { - seq_printf(s, ",unc=%s", cifs_sb->tcon->treeName); - if (tcon->ses) { - if (tcon->ses->userName) - seq_printf(s, ",username=%s", - tcon->ses->userName); - if (tcon->ses->domainName) - seq_printf(s, ",domain=%s", - tcon->ses->domainName); - server = tcon->ses->server; - if (server) { - seq_printf(s, ",addr="); - switch (server->addr.sockAddr6. - sin6_family) { - case AF_INET6: - seq_printf(s, "%pI6", - &server->addr.sockAddr6.sin6_addr); - break; - case AF_INET: - seq_printf(s, "%pI4", - &server->addr.sockAddr.sin_addr.s_addr); - break; - } - } - } - if ((cifs_sb->mnt_cifs_flags & CIFS_MOUNT_OVERR_UID) || - !(tcon->unix_ext)) - seq_printf(s, ",uid=%d", cifs_sb->mnt_uid); - if ((cifs_sb->mnt_cifs_flags & CIFS_MOUNT_OVERR_GID) || - !(tcon->unix_ext)) - seq_printf(s, ",gid=%d", cifs_sb->mnt_gid); - if (!tcon->unix_ext) { - seq_printf(s, ",file_mode=0%o,dir_mode=0%o", + seq_printf(s, ",unc=%s", cifs_sb->tcon->treeName); + if (tcon->ses->userName) + seq_printf(s, ",username=%s", tcon->ses->userName); + if (tcon->ses->domainName) + seq_printf(s, ",domain=%s", tcon->ses->domainName); + + cifs_show_address(s, tcon->ses->server); + + seq_printf(s, ",uid=%d", cifs_sb->mnt_uid); + seq_printf(s, ",gid=%d", cifs_sb->mnt_gid); + + server = tcon->ses->server; + seq_printf(s, ",addr="); + switch (server->addr.sockAddr6.sin6_family) { + case AF_INET6: + seq_printf(s, "%pI6", &server->addr.sockAddr6.sin6_addr); + break; + case AF_INET: + seq_printf(s, "%pI4", &server->addr.sockAddr.sin_addr.s_addr); + break; + } + + if (!tcon->unix_ext) + seq_printf(s, ",file_mode=0%o,dir_mode=0%o", cifs_sb->mnt_file_mode, cifs_sb->mnt_dir_mode); - } - if (tcon->seal) - seq_printf(s, ",seal"); - if (tcon->nocase) - seq_printf(s, ",nocase"); - if (tcon->retry) - seq_printf(s, ",hard"); - } - if (cifs_sb->prepath) - seq_printf(s, ",prepath=%s", cifs_sb->prepath); - if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_POSIX_PATHS) - seq_printf(s, ",posixpaths"); - if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_SET_UID) - seq_printf(s, ",setuids"); - if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_SERVER_INUM) - seq_printf(s, ",serverino"); - if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_DIRECT_IO) - seq_printf(s, ",directio"); - if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NO_XATTR) - seq_printf(s, ",nouser_xattr"); - if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR) - seq_printf(s, ",mapchars"); - if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_UNX_EMUL) - seq_printf(s, ",sfu"); - if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NO_BRL) - seq_printf(s, ",nobrl"); - if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_CIFS_ACL) - seq_printf(s, ",cifsacl"); - if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_DYNPERM) - seq_printf(s, ",dynperm"); - if (m->mnt_sb->s_flags & MS_POSIXACL) - seq_printf(s, ",acl"); - - seq_printf(s, ",rsize=%d", cifs_sb->rsize); - seq_printf(s, ",wsize=%d", cifs_sb->wsize); - } + if (tcon->seal) + seq_printf(s, ",seal"); + if (tcon->nocase) + seq_printf(s, ",nocase"); + if (tcon->retry) + seq_printf(s, ",hard"); + if (cifs_sb->prepath) + seq_printf(s, ",prepath=%s", cifs_sb->prepath); + if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_POSIX_PATHS) + seq_printf(s, ",posixpaths"); + if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_SET_UID) + seq_printf(s, ",setuids"); + if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_SERVER_INUM) + seq_printf(s, ",serverino"); + if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_DIRECT_IO) + seq_printf(s, ",directio"); + if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NO_XATTR) + seq_printf(s, ",nouser_xattr"); + if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR) + seq_printf(s, ",mapchars"); + if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_UNX_EMUL) + seq_printf(s, ",sfu"); + if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NO_BRL) + seq_printf(s, ",nobrl"); + if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_CIFS_ACL) + seq_printf(s, ",cifsacl"); + if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_DYNPERM) + seq_printf(s, ",dynperm"); + if (m->mnt_sb->s_flags & MS_POSIXACL) + seq_printf(s, ",acl"); + + seq_printf(s, ",rsize=%d", cifs_sb->rsize); + seq_printf(s, ",wsize=%d", cifs_sb->wsize); + return 0; } -- cgit v1.2.3 From 340481a36498bf3fe404bcecb2e2d6188e950bff Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Thu, 11 Jun 2009 10:27:29 -0400 Subject: cifs: have cifs_show_options show forceuid/forcegid options Signed-off-by: Jeff Layton Signed-off-by: Steve French --- fs/cifs/cifsfs.c | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'fs') diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c index 3f121fbd6c56..8b315708cb3f 100644 --- a/fs/cifs/cifsfs.c +++ b/fs/cifs/cifsfs.c @@ -357,7 +357,12 @@ cifs_show_options(struct seq_file *s, struct vfsmount *m) cifs_show_address(s, tcon->ses->server); seq_printf(s, ",uid=%d", cifs_sb->mnt_uid); + if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_OVERR_UID) + seq_printf(s, ",forceuid"); + seq_printf(s, ",gid=%d", cifs_sb->mnt_gid); + if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_OVERR_GID) + seq_printf(s, ",forcegid"); server = tcon->ses->server; seq_printf(s, ",addr="); -- cgit v1.2.3 From 1e68b2b2756fc3488ecbade5ad5f13302b3aaafc Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Thu, 11 Jun 2009 10:27:30 -0400 Subject: cifs: add new routine for converting AF_INET and AF_INET6 addrs ...to consolidate some logic used in more than one place. Signed-off-by: Jeff Layton Signed-off-by: Steve French --- fs/cifs/cifsproto.h | 2 +- fs/cifs/connect.c | 21 ++++----------------- fs/cifs/dns_resolve.c | 21 +++------------------ fs/cifs/netmisc.c | 34 ++++++++++++++++++++++++++++++---- 4 files changed, 38 insertions(+), 40 deletions(-) (limited to 'fs') diff --git a/fs/cifs/cifsproto.h b/fs/cifs/cifsproto.h index f9452329bcce..c419416a42ee 100644 --- a/fs/cifs/cifsproto.h +++ b/fs/cifs/cifsproto.h @@ -74,7 +74,7 @@ extern unsigned int smbCalcSize(struct smb_hdr *ptr); extern unsigned int smbCalcSize_LE(struct smb_hdr *ptr); extern int decode_negTokenInit(unsigned char *security_blob, int length, enum securityEnum *secType); -extern int cifs_inet_pton(const int, const char *source, void *dst); +extern int cifs_convert_address(char *src, void *dst); extern int map_smb_to_linux_error(struct smb_hdr *smb, int logErr); extern void header_assemble(struct smb_hdr *, char /* command */ , const struct cifsTconInfo *, int /* length of diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index 97f4311b9a8e..c368ad658236 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -1433,28 +1433,15 @@ cifs_get_tcp_session(struct smb_vol *volume_info) memset(&addr, 0, sizeof(struct sockaddr_storage)); - if (volume_info->UNCip && volume_info->UNC) { - rc = cifs_inet_pton(AF_INET, volume_info->UNCip, - &sin_server->sin_addr.s_addr); - - if (rc <= 0) { - /* not ipv4 address, try ipv6 */ - rc = cifs_inet_pton(AF_INET6, volume_info->UNCip, - &sin_server6->sin6_addr.in6_u); - if (rc > 0) - addr.ss_family = AF_INET6; - } else { - addr.ss_family = AF_INET; - } + cFYI(1, ("UNC: %s ip: %s", volume_info->UNC, volume_info->UNCip)); - if (rc <= 0) { + if (volume_info->UNCip && volume_info->UNC) { + rc = cifs_convert_address(volume_info->UNCip, &addr); + if (!rc) { /* we failed translating address */ rc = -EINVAL; goto out_err; } - - cFYI(1, ("UNC: %s ip: %s", volume_info->UNC, - volume_info->UNCip)); } else if (volume_info->UNCip) { /* BB using ip addr as tcp_ses name to connect to the DFS root below */ diff --git a/fs/cifs/dns_resolve.c b/fs/cifs/dns_resolve.c index df4a306f697e..91b5500755bf 100644 --- a/fs/cifs/dns_resolve.c +++ b/fs/cifs/dns_resolve.c @@ -37,24 +37,9 @@ static int is_ip(const char *name) { - int rc; - struct sockaddr_in sin_server; - struct sockaddr_in6 sin_server6; - - rc = cifs_inet_pton(AF_INET, name, - &sin_server.sin_addr.s_addr); - - if (rc <= 0) { - /* not ipv4 address, try ipv6 */ - rc = cifs_inet_pton(AF_INET6, name, - &sin_server6.sin6_addr.in6_u); - if (rc > 0) - return 1; - } else { - return 1; - } - /* we failed translating address */ - return 0; + struct sockaddr_storage ss; + + return cifs_convert_address(name, &ss); } static int diff --git a/fs/cifs/netmisc.c b/fs/cifs/netmisc.c index 32d6baa0a54f..00e6e357ae88 100644 --- a/fs/cifs/netmisc.c +++ b/fs/cifs/netmisc.c @@ -133,10 +133,12 @@ static const struct smb_to_posix_error mapping_table_ERRHRD[] = { {0, 0} }; -/* Convert string containing dotted ip address to binary form */ -/* returns 0 if invalid address */ - -int +/* + * Convert a string containing text IPv4 or IPv6 address to binary form. + * + * Returns 0 on failure. + */ +static int cifs_inet_pton(const int address_family, const char *cp, void *dst) { int ret = 0; @@ -153,6 +155,30 @@ cifs_inet_pton(const int address_family, const char *cp, void *dst) return ret; } +/* + * Try to convert a string to an IPv4 address and then attempt to convert + * it to an IPv6 address if that fails. Set the family field if either + * succeeds. + * + * Returns 0 on failure. + */ +int +cifs_convert_address(char *src, void *dst) +{ + struct sockaddr_in *s4 = (struct sockaddr_in *) dst; + struct sockaddr_in6 *s6 = (Struct sockaddr_in6 *) dst; + + if (cifs_inet_pton(AF_INET, src, &s4->sin_addr.s_addr)) { + s4->sin_family = AF_INET; + return 1; + } else if (cifs_inet_pton(AF_INET6, src, &s6->sin6_addr.s6_addr)) { + s6->sin6_family = AF_INET6; + return 1; + } + + return 0; +} + /***************************************************************************** convert a NT status code to a dos class/code *****************************************************************************/ -- cgit v1.2.3 From 748de6736c1e482e111f9d1b5a5d5b1787600cad Mon Sep 17 00:00:00 2001 From: Akira Fujita Date: Wed, 17 Jun 2009 19:24:03 -0400 Subject: ext4: online defrag -- Add EXT4_IOC_MOVE_EXT ioctl The EXT4_IOC_MOVE_EXT exchanges the blocks between orig_fd and donor_fd, and then write the file data of orig_fd to donor_fd. ext4_mext_move_extent() is the main fucntion of ext4 online defrag, and this patch includes all functions related to ext4 online defrag. Signed-off-by: Akira Fujita Signed-off-by: Takashi Sato Signed-off-by: Kazuya Mio Signed-off-by: "Theodore Ts'o" --- fs/ext4/Makefile | 2 +- fs/ext4/ext4.h | 15 + fs/ext4/ext4_extents.h | 4 + fs/ext4/extents.c | 4 +- fs/ext4/ioctl.c | 36 ++ fs/ext4/move_extent.c | 1320 ++++++++++++++++++++++++++++++++++++++++++++++++ 6 files changed, 1378 insertions(+), 3 deletions(-) create mode 100644 fs/ext4/move_extent.c (limited to 'fs') diff --git a/fs/ext4/Makefile b/fs/ext4/Makefile index 8a34710ecf40..8867b2a1e5fe 100644 --- a/fs/ext4/Makefile +++ b/fs/ext4/Makefile @@ -6,7 +6,7 @@ obj-$(CONFIG_EXT4_FS) += ext4.o ext4-y := balloc.o bitmap.o dir.o file.o fsync.o ialloc.o inode.o \ ioctl.o namei.o super.o symlink.o hash.o resize.o extents.o \ - ext4_jbd2.o migrate.o mballoc.o block_validity.o + ext4_jbd2.o migrate.o mballoc.o block_validity.o move_extent.o ext4-$(CONFIG_EXT4_FS_XATTR) += xattr.o xattr_user.o xattr_trusted.o ext4-$(CONFIG_EXT4_FS_POSIX_ACL) += acl.o diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index cc7d5edc38c9..276a26f117e6 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -352,6 +352,7 @@ struct ext4_new_group_data { /* note ioctl 10 reserved for an early version of the FIEMAP ioctl */ /* note ioctl 11 reserved for filesystem-independent FIEMAP ioctl */ #define EXT4_IOC_ALLOC_DA_BLKS _IO('f', 12) +#define EXT4_IOC_MOVE_EXT _IOWR('f', 15, struct move_extent) /* * ioctl commands in 32 bit emulation @@ -447,6 +448,15 @@ struct ext4_inode { __le32 i_version_hi; /* high 32 bits for 64-bit version */ }; +struct move_extent { + __u32 reserved; /* should be zero */ + __u32 donor_fd; /* donor file descriptor */ + __u64 orig_start; /* logical start offset in block for orig */ + __u64 donor_start; /* logical start offset in block for donor */ + __u64 len; /* block length to be moved */ + __u64 moved_len; /* moved block length */ +}; +#define MAX_DEFRAG_SIZE ((1UL<<31) - 1) #define EXT4_EPOCH_BITS 2 #define EXT4_EPOCH_MASK ((1 << EXT4_EPOCH_BITS) - 1) @@ -1647,6 +1657,11 @@ extern int ext4_get_blocks(handle_t *handle, struct inode *inode, struct buffer_head *bh, int flags); extern int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, __u64 start, __u64 len); +/* move_extent.c */ +extern int ext4_move_extents(struct file *o_filp, struct file *d_filp, + __u64 start_orig, __u64 start_donor, + __u64 len, __u64 *moved_len); + /* * Add new method to test wether block and inode bitmaps are properly diff --git a/fs/ext4/ext4_extents.h b/fs/ext4/ext4_extents.h index f0c3ec85bd48..20a84105a10b 100644 --- a/fs/ext4/ext4_extents.h +++ b/fs/ext4/ext4_extents.h @@ -221,12 +221,16 @@ static inline int ext4_ext_get_actual_len(struct ext4_extent *ext) } extern int ext4_ext_calc_metadata_amount(struct inode *inode, int blocks); +extern ext4_fsblk_t ext_pblock(struct ext4_extent *ex); extern ext4_fsblk_t idx_pblock(struct ext4_extent_idx *); extern void ext4_ext_store_pblock(struct ext4_extent *, ext4_fsblk_t); extern int ext4_extent_tree_init(handle_t *, struct inode *); extern int ext4_ext_calc_credits_for_single_extent(struct inode *inode, int num, struct ext4_ext_path *path); +extern int ext4_can_extents_be_merged(struct inode *inode, + struct ext4_extent *ex1, + struct ext4_extent *ex2); extern int ext4_ext_try_to_merge(struct inode *inode, struct ext4_ext_path *path, struct ext4_extent *); diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index 2593f748c3a4..50322a09bd01 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c @@ -49,7 +49,7 @@ * ext_pblock: * combine low and high parts of physical block number into ext4_fsblk_t */ -static ext4_fsblk_t ext_pblock(struct ext4_extent *ex) +ext4_fsblk_t ext_pblock(struct ext4_extent *ex) { ext4_fsblk_t block; @@ -1417,7 +1417,7 @@ static int ext4_ext_correct_indexes(handle_t *handle, struct inode *inode, return err; } -static int +int ext4_can_extents_be_merged(struct inode *inode, struct ext4_extent *ex1, struct ext4_extent *ex2) { diff --git a/fs/ext4/ioctl.c b/fs/ext4/ioctl.c index 91e75f7a9e73..bb415408fdb6 100644 --- a/fs/ext4/ioctl.c +++ b/fs/ext4/ioctl.c @@ -14,6 +14,7 @@ #include #include #include +#include #include #include "ext4_jbd2.h" #include "ext4.h" @@ -213,6 +214,41 @@ setversion_out: return err; } + + case EXT4_IOC_MOVE_EXT: { + struct move_extent me; + struct file *donor_filp; + int err; + + if (copy_from_user(&me, + (struct move_extent __user *)arg, sizeof(me))) + return -EFAULT; + + donor_filp = fget(me.donor_fd); + if (!donor_filp) + return -EBADF; + + if (!capable(CAP_DAC_OVERRIDE)) { + if ((current->real_cred->fsuid != inode->i_uid) || + !(inode->i_mode & S_IRUSR) || + !(donor_filp->f_dentry->d_inode->i_mode & + S_IRUSR)) { + fput(donor_filp); + return -EACCES; + } + } + + err = ext4_move_extents(filp, donor_filp, me.orig_start, + me.donor_start, me.len, &me.moved_len); + fput(donor_filp); + + if (!err) + if (copy_to_user((struct move_extent *)arg, + &me, sizeof(me))) + return -EFAULT; + return err; + } + case EXT4_IOC_GROUP_ADD: { struct ext4_new_group_data input; struct super_block *sb = inode->i_sb; diff --git a/fs/ext4/move_extent.c b/fs/ext4/move_extent.c new file mode 100644 index 000000000000..bbf2dd9404dc --- /dev/null +++ b/fs/ext4/move_extent.c @@ -0,0 +1,1320 @@ +/* + * Copyright (c) 2008,2009 NEC Software Tohoku, Ltd. + * Written by Takashi Sato + * Akira Fujita + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of version 2.1 of the GNU Lesser General Public License + * as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#include +#include +#include "ext4_jbd2.h" +#include "ext4_extents.h" +#include "ext4.h" + +#define get_ext_path(path, inode, block, ret) \ + do { \ + path = ext4_ext_find_extent(inode, block, path); \ + if (IS_ERR(path)) { \ + ret = PTR_ERR(path); \ + path = NULL; \ + } \ + } while (0) + +/** + * copy_extent_status - Copy the extent's initialization status + * + * @src: an extent for getting initialize status + * @dest: an extent to be set the status + */ +static void +copy_extent_status(struct ext4_extent *src, struct ext4_extent *dest) +{ + if (ext4_ext_is_uninitialized(src)) + ext4_ext_mark_uninitialized(dest); + else + dest->ee_len = cpu_to_le16(ext4_ext_get_actual_len(dest)); +} + +/** + * mext_next_extent - Search for the next extent and set it to "extent" + * + * @inode: inode which is searched + * @path: this will obtain data for the next extent + * @extent: pointer to the next extent we have just gotten + * + * Search the next extent in the array of ext4_ext_path structure (@path) + * and set it to ext4_extent structure (@extent). In addition, the member of + * @path (->p_ext) also points the next extent. Return 0 on success, 1 if + * ext4_ext_path structure refers to the last extent, or a negative error + * value on failure. + */ +static int +mext_next_extent(struct inode *inode, struct ext4_ext_path *path, + struct ext4_extent **extent) +{ + int ppos, leaf_ppos = path->p_depth; + + ppos = leaf_ppos; + if (EXT_LAST_EXTENT(path[ppos].p_hdr) > path[ppos].p_ext) { + /* leaf block */ + *extent = ++path[ppos].p_ext; + return 0; + } + + while (--ppos >= 0) { + if (EXT_LAST_INDEX(path[ppos].p_hdr) > + path[ppos].p_idx) { + int cur_ppos = ppos; + + /* index block */ + path[ppos].p_idx++; + path[ppos].p_block = idx_pblock(path[ppos].p_idx); + if (path[ppos+1].p_bh) + brelse(path[ppos+1].p_bh); + path[ppos+1].p_bh = + sb_bread(inode->i_sb, path[ppos].p_block); + if (!path[ppos+1].p_bh) + return -EIO; + path[ppos+1].p_hdr = + ext_block_hdr(path[ppos+1].p_bh); + + /* Halfway index block */ + while (++cur_ppos < leaf_ppos) { + path[cur_ppos].p_idx = + EXT_FIRST_INDEX(path[cur_ppos].p_hdr); + path[cur_ppos].p_block = + idx_pblock(path[cur_ppos].p_idx); + if (path[cur_ppos+1].p_bh) + brelse(path[cur_ppos+1].p_bh); + path[cur_ppos+1].p_bh = sb_bread(inode->i_sb, + path[cur_ppos].p_block); + if (!path[cur_ppos+1].p_bh) + return -EIO; + path[cur_ppos+1].p_hdr = + ext_block_hdr(path[cur_ppos+1].p_bh); + } + + /* leaf block */ + path[leaf_ppos].p_ext = *extent = + EXT_FIRST_EXTENT(path[leaf_ppos].p_hdr); + return 0; + } + } + /* We found the last extent */ + return 1; +} + +/** + * mext_double_down_read - Acquire two inodes' read semaphore + * + * @orig_inode: original inode structure + * @donor_inode: donor inode structure + * Acquire read semaphore of the two inodes (orig and donor) by i_ino order. + */ +static void +mext_double_down_read(struct inode *orig_inode, struct inode *donor_inode) +{ + struct inode *first = orig_inode, *second = donor_inode; + + BUG_ON(orig_inode == NULL || donor_inode == NULL); + + /* + * Use the inode number to provide the stable locking order instead + * of its address, because the C language doesn't guarantee you can + * compare pointers that don't come from the same array. + */ + if (donor_inode->i_ino < orig_inode->i_ino) { + first = donor_inode; + second = orig_inode; + } + + down_read(&EXT4_I(first)->i_data_sem); + down_read(&EXT4_I(second)->i_data_sem); +} + +/** + * mext_double_down_write - Acquire two inodes' write semaphore + * + * @orig_inode: original inode structure + * @donor_inode: donor inode structure + * Acquire write semaphore of the two inodes (orig and donor) by i_ino order. + */ +static void +mext_double_down_write(struct inode *orig_inode, struct inode *donor_inode) +{ + struct inode *first = orig_inode, *second = donor_inode; + + BUG_ON(orig_inode == NULL || donor_inode == NULL); + + /* + * Use the inode number to provide the stable locking order instead + * of its address, because the C language doesn't guarantee you can + * compare pointers that don't come from the same array. + */ + if (donor_inode->i_ino < orig_inode->i_ino) { + first = donor_inode; + second = orig_inode; + } + + down_write(&EXT4_I(first)->i_data_sem); + down_write(&EXT4_I(second)->i_data_sem); +} + +/** + * mext_double_up_read - Release two inodes' read semaphore + * + * @orig_inode: original inode structure to be released its lock first + * @donor_inode: donor inode structure to be released its lock second + * Release read semaphore of two inodes (orig and donor). + */ +static void +mext_double_up_read(struct inode *orig_inode, struct inode *donor_inode) +{ + BUG_ON(orig_inode == NULL || donor_inode == NULL); + + up_read(&EXT4_I(orig_inode)->i_data_sem); + up_read(&EXT4_I(donor_inode)->i_data_sem); +} + +/** + * mext_double_up_write - Release two inodes' write semaphore + * + * @orig_inode: original inode structure to be released its lock first + * @donor_inode: donor inode structure to be released its lock second + * Release write semaphore of two inodes (orig and donor). + */ +static void +mext_double_up_write(struct inode *orig_inode, struct inode *donor_inode) +{ + BUG_ON(orig_inode == NULL || donor_inode == NULL); + + up_write(&EXT4_I(orig_inode)->i_data_sem); + up_write(&EXT4_I(donor_inode)->i_data_sem); +} + +/** + * mext_insert_across_blocks - Insert extents across leaf block + * + * @handle: journal handle + * @orig_inode: original inode + * @o_start: first original extent to be changed + * @o_end: last original extent to be changed + * @start_ext: first new extent to be inserted + * @new_ext: middle of new extent to be inserted + * @end_ext: last new extent to be inserted + * + * Allocate a new leaf block and insert extents into it. Return 0 on success, + * or a negative error value on failure. + */ +static int +mext_insert_across_blocks(handle_t *handle, struct inode *orig_inode, + struct ext4_extent *o_start, struct ext4_extent *o_end, + struct ext4_extent *start_ext, struct ext4_extent *new_ext, + struct ext4_extent *end_ext) +{ + struct ext4_ext_path *orig_path = NULL; + ext4_lblk_t eblock = 0; + int new_flag = 0; + int end_flag = 0; + int err = 0; + + if (start_ext->ee_len && new_ext->ee_len && end_ext->ee_len) { + if (o_start == o_end) { + + /* start_ext new_ext end_ext + * donor |---------|-----------|--------| + * orig |------------------------------| + */ + end_flag = 1; + } else { + + /* start_ext new_ext end_ext + * donor |---------|----------|---------| + * orig |---------------|--------------| + */ + o_end->ee_block = end_ext->ee_block; + o_end->ee_len = end_ext->ee_len; + ext4_ext_store_pblock(o_end, ext_pblock(end_ext)); + } + + o_start->ee_len = start_ext->ee_len; + new_flag = 1; + + } else if (start_ext->ee_len && new_ext->ee_len && + !end_ext->ee_len && o_start == o_end) { + + /* start_ext new_ext + * donor |--------------|---------------| + * orig |------------------------------| + */ + o_start->ee_len = start_ext->ee_len; + new_flag = 1; + + } else if (!start_ext->ee_len && new_ext->ee_len && + end_ext->ee_len && o_start == o_end) { + + /* new_ext end_ext + * donor |--------------|---------------| + * orig |------------------------------| + */ + o_end->ee_block = end_ext->ee_block; + o_end->ee_len = end_ext->ee_len; + ext4_ext_store_pblock(o_end, ext_pblock(end_ext)); + + /* + * Set 0 to the extent block if new_ext was + * the first block. + */ + if (new_ext->ee_block) + eblock = le32_to_cpu(new_ext->ee_block); + + new_flag = 1; + } else { + ext4_debug("ext4 move extent: Unexpected insert case\n"); + return -EIO; + } + + if (new_flag) { + get_ext_path(orig_path, orig_inode, eblock, err); + if (orig_path == NULL) + goto out; + + if (ext4_ext_insert_extent(handle, orig_inode, + orig_path, new_ext)) + goto out; + } + + if (end_flag) { + get_ext_path(orig_path, orig_inode, + le32_to_cpu(end_ext->ee_block) - 1, err); + if (orig_path == NULL) + goto out; + + if (ext4_ext_insert_extent(handle, orig_inode, + orig_path, end_ext)) + goto out; + } +out: + if (orig_path) { + ext4_ext_drop_refs(orig_path); + kfree(orig_path); + } + + return err; + +} + +/** + * mext_insert_inside_block - Insert new extent to the extent block + * + * @o_start: first original extent to be moved + * @o_end: last original extent to be moved + * @start_ext: first new extent to be inserted + * @new_ext: middle of new extent to be inserted + * @end_ext: last new extent to be inserted + * @eh: extent header of target leaf block + * @range_to_move: used to decide how to insert extent + * + * Insert extents into the leaf block. The extent (@o_start) is overwritten + * by inserted extents. + */ +static void +mext_insert_inside_block(struct ext4_extent *o_start, + struct ext4_extent *o_end, + struct ext4_extent *start_ext, + struct ext4_extent *new_ext, + struct ext4_extent *end_ext, + struct ext4_extent_header *eh, + int range_to_move) +{ + int i = 0; + unsigned long len; + + /* Move the existing extents */ + if (range_to_move && o_end < EXT_LAST_EXTENT(eh)) { + len = (unsigned long)(EXT_LAST_EXTENT(eh) + 1) - + (unsigned long)(o_end + 1); + memmove(o_end + 1 + range_to_move, o_end + 1, len); + } + + /* Insert start entry */ + if (start_ext->ee_len) + o_start[i++].ee_len = start_ext->ee_len; + + /* Insert new entry */ + if (new_ext->ee_len) { + o_start[i] = *new_ext; + ext4_ext_store_pblock(&o_start[i++], ext_pblock(new_ext)); + } + + /* Insert end entry */ + if (end_ext->ee_len) + o_start[i] = *end_ext; + + /* Increment the total entries counter on the extent block */ + le16_add_cpu(&eh->eh_entries, range_to_move); +} + +/** + * mext_insert_extents - Insert new extent + * + * @handle: journal handle + * @orig_inode: original inode + * @orig_path: path indicates first extent to be changed + * @o_start: first original extent to be changed + * @o_end: last original extent to be changed + * @start_ext: first new extent to be inserted + * @new_ext: middle of new extent to be inserted + * @end_ext: last new extent to be inserted + * + * Call the function to insert extents. If we cannot add more extents into + * the leaf block, we call mext_insert_across_blocks() to create a + * new leaf block. Otherwise call mext_insert_inside_block(). Return 0 + * on success, or a negative error value on failure. + */ +static int +mext_insert_extents(handle_t *handle, struct inode *orig_inode, + struct ext4_ext_path *orig_path, + struct ext4_extent *o_start, + struct ext4_extent *o_end, + struct ext4_extent *start_ext, + struct ext4_extent *new_ext, + struct ext4_extent *end_ext) +{ + struct ext4_extent_header *eh; + unsigned long need_slots, slots_range; + int range_to_move, depth, ret; + + /* + * The extents need to be inserted + * start_extent + new_extent + end_extent. + */ + need_slots = (start_ext->ee_len ? 1 : 0) + (end_ext->ee_len ? 1 : 0) + + (new_ext->ee_len ? 1 : 0); + + /* The number of slots between start and end */ + slots_range = ((unsigned long)(o_end + 1) - (unsigned long)o_start + 1) + / sizeof(struct ext4_extent); + + /* Range to move the end of extent */ + range_to_move = need_slots - slots_range; + depth = orig_path->p_depth; + orig_path += depth; + eh = orig_path->p_hdr; + + if (depth) { + /* Register to journal */ + ret = ext4_journal_get_write_access(handle, orig_path->p_bh); + if (ret) + return ret; + } + + /* Expansion */ + if (range_to_move > 0 && + (range_to_move > le16_to_cpu(eh->eh_max) + - le16_to_cpu(eh->eh_entries))) { + + ret = mext_insert_across_blocks(handle, orig_inode, o_start, + o_end, start_ext, new_ext, end_ext); + if (ret < 0) + return ret; + } else + mext_insert_inside_block(o_start, o_end, start_ext, new_ext, + end_ext, eh, range_to_move); + + if (depth) { + ret = ext4_handle_dirty_metadata(handle, orig_inode, + orig_path->p_bh); + if (ret) + return ret; + } else { + ret = ext4_mark_inode_dirty(handle, orig_inode); + if (ret < 0) + return ret; + } + + return 0; +} + +/** + * mext_leaf_block - Move one leaf extent block into the inode. + * + * @handle: journal handle + * @orig_inode: original inode + * @orig_path: path indicates first extent to be changed + * @dext: donor extent + * @from: start offset on the target file + * + * In order to insert extents into the leaf block, we must divide the extent + * in the leaf block into three extents. The one is located to be inserted + * extents, and the others are located around it. + * + * Therefore, this function creates structures to save extents of the leaf + * block, and inserts extents by calling mext_insert_extents() with + * created extents. Return 0 on success, or a negative error value on failure. + */ +static int +mext_leaf_block(handle_t *handle, struct inode *orig_inode, + struct ext4_ext_path *orig_path, struct ext4_extent *dext, + ext4_lblk_t *from) +{ + struct ext4_extent *oext, *o_start, *o_end, *prev_ext; + struct ext4_extent new_ext, start_ext, end_ext; + ext4_lblk_t new_ext_end; + ext4_fsblk_t new_phys_end; + int oext_alen, new_ext_alen, end_ext_alen; + int depth = ext_depth(orig_inode); + int ret; + + o_start = o_end = oext = orig_path[depth].p_ext; + oext_alen = ext4_ext_get_actual_len(oext); + start_ext.ee_len = end_ext.ee_len = 0; + + new_ext.ee_block = cpu_to_le32(*from); + ext4_ext_store_pblock(&new_ext, ext_pblock(dext)); + new_ext.ee_len = dext->ee_len; + new_ext_alen = ext4_ext_get_actual_len(&new_ext); + new_ext_end = le32_to_cpu(new_ext.ee_block) + new_ext_alen - 1; + new_phys_end = ext_pblock(&new_ext) + new_ext_alen - 1; + + /* + * Case: original extent is first + * oext |--------| + * new_ext |--| + * start_ext |--| + */ + if (le32_to_cpu(oext->ee_block) < le32_to_cpu(new_ext.ee_block) && + le32_to_cpu(new_ext.ee_block) < + le32_to_cpu(oext->ee_block) + oext_alen) { + start_ext.ee_len = cpu_to_le16(le32_to_cpu(new_ext.ee_block) - + le32_to_cpu(oext->ee_block)); + copy_extent_status(oext, &start_ext); + } else if (oext > EXT_FIRST_EXTENT(orig_path[depth].p_hdr)) { + prev_ext = oext - 1; + /* + * We can merge new_ext into previous extent, + * if these are contiguous and same extent type. + */ + if (ext4_can_extents_be_merged(orig_inode, prev_ext, + &new_ext)) { + o_start = prev_ext; + start_ext.ee_len = cpu_to_le16( + ext4_ext_get_actual_len(prev_ext) + + new_ext_alen); + copy_extent_status(prev_ext, &start_ext); + new_ext.ee_len = 0; + } + } + + /* + * Case: new_ext_end must be less than oext + * oext |-----------| + * new_ext |-------| + */ + BUG_ON(le32_to_cpu(oext->ee_block) + oext_alen - 1 < new_ext_end); + + /* + * Case: new_ext is smaller than original extent + * oext |---------------| + * new_ext |-----------| + * end_ext |---| + */ + if (le32_to_cpu(oext->ee_block) <= new_ext_end && + new_ext_end < le32_to_cpu(oext->ee_block) + oext_alen - 1) { + end_ext.ee_len = + cpu_to_le16(le32_to_cpu(oext->ee_block) + + oext_alen - 1 - new_ext_end); + copy_extent_status(oext, &end_ext); + end_ext_alen = ext4_ext_get_actual_len(&end_ext); + ext4_ext_store_pblock(&end_ext, + (ext_pblock(o_end) + oext_alen - end_ext_alen)); + end_ext.ee_block = + cpu_to_le32(le32_to_cpu(o_end->ee_block) + + oext_alen - end_ext_alen); + } + + ret = mext_insert_extents(handle, orig_inode, orig_path, o_start, + o_end, &start_ext, &new_ext, &end_ext); + return ret; +} + +/** + * mext_calc_swap_extents - Calculate extents for extent swapping. + * + * @tmp_dext: the extent that will belong to the original inode + * @tmp_oext: the extent that will belong to the donor inode + * @orig_off: block offset of original inode + * @donor_off: block offset of donor inode + * @max_count: the maximun length of extents + */ +static void +mext_calc_swap_extents(struct ext4_extent *tmp_dext, + struct ext4_extent *tmp_oext, + ext4_lblk_t orig_off, ext4_lblk_t donor_off, + ext4_lblk_t max_count) +{ + ext4_lblk_t diff, orig_diff; + struct ext4_extent dext_old, oext_old; + + dext_old = *tmp_dext; + oext_old = *tmp_oext; + + /* When tmp_dext is too large, pick up the target range. */ + diff = donor_off - le32_to_cpu(tmp_dext->ee_block); + + ext4_ext_store_pblock(tmp_dext, ext_pblock(tmp_dext) + diff); + tmp_dext->ee_block = + cpu_to_le32(le32_to_cpu(tmp_dext->ee_block) + diff); + tmp_dext->ee_len = cpu_to_le16(le16_to_cpu(tmp_dext->ee_len) - diff); + + if (max_count < ext4_ext_get_actual_len(tmp_dext)) + tmp_dext->ee_len = cpu_to_le16(max_count); + + orig_diff = orig_off - le32_to_cpu(tmp_oext->ee_block); + ext4_ext_store_pblock(tmp_oext, ext_pblock(tmp_oext) + orig_diff); + + /* Adjust extent length if donor extent is larger than orig */ + if (ext4_ext_get_actual_len(tmp_dext) > + ext4_ext_get_actual_len(tmp_oext) - orig_diff) + tmp_dext->ee_len = cpu_to_le16(le16_to_cpu(tmp_oext->ee_len) - + orig_diff); + + tmp_oext->ee_len = cpu_to_le16(ext4_ext_get_actual_len(tmp_dext)); + + copy_extent_status(&oext_old, tmp_dext); + copy_extent_status(&dext_old, tmp_oext); +} + +/** + * mext_replace_branches - Replace original extents with new extents + * + * @handle: journal handle + * @orig_inode: original inode + * @donor_inode: donor inode + * @from: block offset of orig_inode + * @count: block count to be replaced + * + * Replace original inode extents and donor inode extents page by page. + * We implement this replacement in the following three steps: + * 1. Save the block information of original and donor inodes into + * dummy extents. + * 2. Change the block information of original inode to point at the + * donor inode blocks. + * 3. Change the block information of donor inode to point at the saved + * original inode blocks in the dummy extents. + * + * Return 0 on success, or a negative error value on failure. + */ +static int +mext_replace_branches(handle_t *handle, struct inode *orig_inode, + struct inode *donor_inode, ext4_lblk_t from, + ext4_lblk_t count) +{ + struct ext4_ext_path *orig_path = NULL; + struct ext4_ext_path *donor_path = NULL; + struct ext4_extent *oext, *dext; + struct ext4_extent tmp_dext, tmp_oext; + ext4_lblk_t orig_off = from, donor_off = from; + int err = 0; + int depth; + int replaced_count = 0; + int dext_alen; + + mext_double_down_write(orig_inode, donor_inode); + + /* Get the original extent for the block "orig_off" */ + get_ext_path(orig_path, orig_inode, orig_off, err); + if (orig_path == NULL) + goto out; + + /* Get the donor extent for the head */ + get_ext_path(donor_path, donor_inode, donor_off, err); + if (donor_path == NULL) + goto out; + depth = ext_depth(orig_inode); + oext = orig_path[depth].p_ext; + tmp_oext = *oext; + + depth = ext_depth(donor_inode); + dext = donor_path[depth].p_ext; + tmp_dext = *dext; + + mext_calc_swap_extents(&tmp_dext, &tmp_oext, orig_off, + donor_off, count); + + /* Loop for the donor extents */ + while (1) { + /* The extent for donor must be found. */ + BUG_ON(!dext || donor_off != le32_to_cpu(tmp_dext.ee_block)); + + /* Set donor extent to orig extent */ + err = mext_leaf_block(handle, orig_inode, + orig_path, &tmp_dext, &orig_off); + if (err < 0) + goto out; + + /* Set orig extent to donor extent */ + err = mext_leaf_block(handle, donor_inode, + donor_path, &tmp_oext, &donor_off); + if (err < 0) + goto out; + + dext_alen = ext4_ext_get_actual_len(&tmp_dext); + replaced_count += dext_alen; + donor_off += dext_alen; + orig_off += dext_alen; + + /* Already moved the expected blocks */ + if (replaced_count >= count) + break; + + if (orig_path) + ext4_ext_drop_refs(orig_path); + get_ext_path(orig_path, orig_inode, orig_off, err); + if (orig_path == NULL) + goto out; + depth = ext_depth(orig_inode); + oext = orig_path[depth].p_ext; + if (le32_to_cpu(oext->ee_block) + + ext4_ext_get_actual_len(oext) <= orig_off) { + err = 0; + goto out; + } + tmp_oext = *oext; + + if (donor_path) + ext4_ext_drop_refs(donor_path); + get_ext_path(donor_path, donor_inode, + donor_off, err); + if (donor_path == NULL) + goto out; + depth = ext_depth(donor_inode); + dext = donor_path[depth].p_ext; + if (le32_to_cpu(dext->ee_block) + + ext4_ext_get_actual_len(dext) <= donor_off) { + err = 0; + goto out; + } + tmp_dext = *dext; + + mext_calc_swap_extents(&tmp_dext, &tmp_oext, orig_off, + donor_off, + count - replaced_count); + } + +out: + if (orig_path) { + ext4_ext_drop_refs(orig_path); + kfree(orig_path); + } + if (donor_path) { + ext4_ext_drop_refs(donor_path); + kfree(donor_path); + } + + mext_double_up_write(orig_inode, donor_inode); + return err; +} + +/** + * move_extent_per_page - Move extent data per page + * + * @o_filp: file structure of original file + * @donor_inode: donor inode + * @orig_page_offset: page index on original file + * @data_offset_in_page: block index where data swapping starts + * @block_len_in_page: the number of blocks to be swapped + * @uninit: orig extent is uninitialized or not + * + * Save the data in original inode blocks and replace original inode extents + * with donor inode extents by calling mext_replace_branches(). + * Finally, write out the saved data in new original inode blocks. Return 0 + * on success, or a negative error value on failure. + */ +static int +move_extent_par_page(struct file *o_filp, struct inode *donor_inode, + pgoff_t orig_page_offset, int data_offset_in_page, + int block_len_in_page, int uninit) +{ + struct inode *orig_inode = o_filp->f_dentry->d_inode; + struct address_space *mapping = orig_inode->i_mapping; + struct buffer_head *bh; + struct page *page = NULL; + const struct address_space_operations *a_ops = mapping->a_ops; + handle_t *handle; + ext4_lblk_t orig_blk_offset; + long long offs = orig_page_offset << PAGE_CACHE_SHIFT; + unsigned long blocksize = orig_inode->i_sb->s_blocksize; + unsigned int w_flags = 0; + unsigned int tmp_data_len, data_len; + void *fsdata; + int ret, i, jblocks; + int blocks_per_page = PAGE_CACHE_SIZE >> orig_inode->i_blkbits; + + /* + * It needs twice the amount of ordinary journal buffers because + * inode and donor_inode may change each different metadata blocks. + */ + jblocks = ext4_writepage_trans_blocks(orig_inode) * 2; + handle = ext4_journal_start(orig_inode, jblocks); + if (IS_ERR(handle)) { + ret = PTR_ERR(handle); + return ret; + } + + if (segment_eq(get_fs(), KERNEL_DS)) + w_flags |= AOP_FLAG_UNINTERRUPTIBLE; + + orig_blk_offset = orig_page_offset * blocks_per_page + + data_offset_in_page; + + /* + * If orig extent is uninitialized one, + * it's not necessary force the page into memory + * and then force it to be written out again. + * Just swap data blocks between orig and donor. + */ + if (uninit) { + ret = mext_replace_branches(handle, orig_inode, + donor_inode, orig_blk_offset, + block_len_in_page); + + /* Clear the inode cache not to refer to the old data */ + ext4_ext_invalidate_cache(orig_inode); + ext4_ext_invalidate_cache(donor_inode); + goto out2; + } + + offs = (long long)orig_blk_offset << orig_inode->i_blkbits; + + /* Calculate data_len */ + if ((orig_blk_offset + block_len_in_page - 1) == + ((orig_inode->i_size - 1) >> orig_inode->i_blkbits)) { + /* Replace the last block */ + tmp_data_len = orig_inode->i_size & (blocksize - 1); + /* + * If data_len equal zero, it shows data_len is multiples of + * blocksize. So we set appropriate value. + */ + if (tmp_data_len == 0) + tmp_data_len = blocksize; + + data_len = tmp_data_len + + ((block_len_in_page - 1) << orig_inode->i_blkbits); + } else { + data_len = block_len_in_page << orig_inode->i_blkbits; + } + + ret = a_ops->write_begin(o_filp, mapping, offs, data_len, w_flags, + &page, &fsdata); + if (unlikely(ret < 0)) + goto out; + + if (!PageUptodate(page)) { + mapping->a_ops->readpage(o_filp, page); + lock_page(page); + } + + /* + * try_to_release_page() doesn't call releasepage in writeback mode. + * We should care about the order of writing to the same file + * by multiple move extent processes. + * It needs to call wait_on_page_writeback() to wait for the + * writeback of the page. + */ + if (PageWriteback(page)) + wait_on_page_writeback(page); + + /* Release old bh and drop refs */ + try_to_release_page(page, 0); + + ret = mext_replace_branches(handle, orig_inode, donor_inode, + orig_blk_offset, block_len_in_page); + if (ret < 0) + goto out; + + /* Clear the inode cache not to refer to the old data */ + ext4_ext_invalidate_cache(orig_inode); + ext4_ext_invalidate_cache(donor_inode); + + if (!page_has_buffers(page)) + create_empty_buffers(page, 1 << orig_inode->i_blkbits, 0); + + bh = page_buffers(page); + for (i = 0; i < data_offset_in_page; i++) + bh = bh->b_this_page; + + for (i = 0; i < block_len_in_page; i++) { + ret = ext4_get_block(orig_inode, + (sector_t)(orig_blk_offset + i), bh, 0); + if (ret < 0) + goto out; + + if (bh->b_this_page != NULL) + bh = bh->b_this_page; + } + + ret = a_ops->write_end(o_filp, mapping, offs, data_len, data_len, + page, fsdata); + page = NULL; + +out: + if (unlikely(page)) { + if (PageLocked(page)) + unlock_page(page); + page_cache_release(page); + } +out2: + ext4_journal_stop(handle); + + return ret < 0 ? ret : 0; +} + +/** + * mext_check_argumants - Check whether move extent can be done + * + * @orig_inode: original inode + * @donor_inode: donor inode + * @orig_start: logical start offset in block for orig + * @donor_start: logical start offset in block for donor + * @len: the number of blocks to be moved + * @moved_len: moved block length + * + * Check the arguments of ext4_move_extents() whether the files can be + * exchanged with each other. + * Return 0 on success, or a negative error value on failure. + */ +static int +mext_check_arguments(struct inode *orig_inode, + struct inode *donor_inode, __u64 orig_start, + __u64 donor_start, __u64 *len, __u64 moved_len) +{ + /* Regular file check */ + if (!S_ISREG(orig_inode->i_mode) || !S_ISREG(donor_inode->i_mode)) { + ext4_debug("ext4 move extent: The argument files should be " + "regular file [ino:orig %lu, donor %lu]\n", + orig_inode->i_ino, donor_inode->i_ino); + return -EINVAL; + } + + /* Ext4 move extent does not support swapfile */ + if (IS_SWAPFILE(orig_inode) || IS_SWAPFILE(donor_inode)) { + ext4_debug("ext4 move extent: The argument files should " + "not be swapfile [ino:orig %lu, donor %lu]\n", + orig_inode->i_ino, donor_inode->i_ino); + return -EINVAL; + } + + /* Files should be in the same ext4 FS */ + if (orig_inode->i_sb != donor_inode->i_sb) { + ext4_debug("ext4 move extent: The argument files " + "should be in same FS [ino:orig %lu, donor %lu]\n", + orig_inode->i_ino, donor_inode->i_ino); + return -EINVAL; + } + + /* orig and donor should be different file */ + if (orig_inode->i_ino == donor_inode->i_ino) { + ext4_debug("ext4 move extent: The argument files should not " + "be same file [ino:orig %lu, donor %lu]\n", + orig_inode->i_ino, donor_inode->i_ino); + return -EINVAL; + } + + /* Ext4 move extent supports only extent based file */ + if (!(EXT4_I(orig_inode)->i_flags & EXT4_EXTENTS_FL)) { + ext4_debug("ext4 move extent: orig file is not extents " + "based file [ino:orig %lu]\n", orig_inode->i_ino); + return -EOPNOTSUPP; + } else if (!(EXT4_I(donor_inode)->i_flags & EXT4_EXTENTS_FL)) { + ext4_debug("ext4 move extent: donor file is not extents " + "based file [ino:donor %lu]\n", donor_inode->i_ino); + return -EOPNOTSUPP; + } + + if ((!orig_inode->i_size) || (!donor_inode->i_size)) { + ext4_debug("ext4 move extent: File size is 0 byte\n"); + return -EINVAL; + } + + /* Start offset should be same */ + if (orig_start != donor_start) { + ext4_debug("ext4 move extent: orig and donor's start " + "offset are not same [ino:orig %lu, donor %lu]\n", + orig_inode->i_ino, donor_inode->i_ino); + return -EINVAL; + } + + if (moved_len) { + ext4_debug("ext4 move extent: moved_len should be 0 " + "[ino:orig %lu, donor %lu]\n", orig_inode->i_ino, + donor_inode->i_ino); + return -EINVAL; + } + + if ((orig_start > MAX_DEFRAG_SIZE) || + (donor_start > MAX_DEFRAG_SIZE) || + (*len > MAX_DEFRAG_SIZE) || + (orig_start + *len > MAX_DEFRAG_SIZE)) { + ext4_debug("ext4 move extent: Can't handle over [%lu] blocks " + "[ino:orig %lu, donor %lu]\n", MAX_DEFRAG_SIZE, + orig_inode->i_ino, donor_inode->i_ino); + return -EINVAL; + } + + if (orig_inode->i_size > donor_inode->i_size) { + if (orig_start >= donor_inode->i_size) { + ext4_debug("ext4 move extent: orig start offset " + "[%llu] should be less than donor file size " + "[%lld] [ino:orig %lu, donor_inode %lu]\n", + orig_start, donor_inode->i_size, + orig_inode->i_ino, donor_inode->i_ino); + return -EINVAL; + } + + if (orig_start + *len > donor_inode->i_size) { + ext4_debug("ext4 move extent: End offset [%llu] should " + "be less than donor file size [%lld]." + "So adjust length from %llu to %lld " + "[ino:orig %lu, donor %lu]\n", + orig_start + *len, donor_inode->i_size, + *len, donor_inode->i_size - orig_start, + orig_inode->i_ino, donor_inode->i_ino); + *len = donor_inode->i_size - orig_start; + } + } else { + if (orig_start >= orig_inode->i_size) { + ext4_debug("ext4 move extent: start offset [%llu] " + "should be less than original file size " + "[%lld] [inode:orig %lu, donor %lu]\n", + orig_start, orig_inode->i_size, + orig_inode->i_ino, donor_inode->i_ino); + return -EINVAL; + } + + if (orig_start + *len > orig_inode->i_size) { + ext4_debug("ext4 move extent: Adjust length " + "from %llu to %lld. Because it should be " + "less than original file size " + "[ino:orig %lu, donor %lu]\n", + *len, orig_inode->i_size - orig_start, + orig_inode->i_ino, donor_inode->i_ino); + *len = orig_inode->i_size - orig_start; + } + } + + if (!*len) { + ext4_debug("ext4 move extent: len shoudld not be 0 " + "[ino:orig %lu, donor %lu]\n", orig_inode->i_ino, + donor_inode->i_ino); + return -EINVAL; + } + + return 0; +} + +/** + * mext_inode_double_lock - Lock i_mutex on both @inode1 and @inode2 + * + * @inode1: the inode structure + * @inode2: the inode structure + * + * Lock two inodes' i_mutex by i_ino order. This function is moved from + * fs/inode.c. + */ +static void +mext_inode_double_lock(struct inode *inode1, struct inode *inode2) +{ + if (inode1 == NULL || inode2 == NULL || inode1 == inode2) { + if (inode1) + mutex_lock(&inode1->i_mutex); + else if (inode2) + mutex_lock(&inode2->i_mutex); + return; + } + + if (inode1->i_ino < inode2->i_ino) { + mutex_lock_nested(&inode1->i_mutex, I_MUTEX_PARENT); + mutex_lock_nested(&inode2->i_mutex, I_MUTEX_CHILD); + } else { + mutex_lock_nested(&inode2->i_mutex, I_MUTEX_PARENT); + mutex_lock_nested(&inode1->i_mutex, I_MUTEX_CHILD); + } +} + +/** + * mext_inode_double_unlock - Release i_mutex on both @inode1 and @inode2 + * + * @inode1: the inode that is released first + * @inode2: the inode that is released second + * + * This function is moved from fs/inode.c. + */ + +static void +mext_inode_double_unlock(struct inode *inode1, struct inode *inode2) +{ + if (inode1) + mutex_unlock(&inode1->i_mutex); + + if (inode2 && inode2 != inode1) + mutex_unlock(&inode2->i_mutex); +} + +/** + * ext4_move_extents - Exchange the specified range of a file + * + * @o_filp: file structure of the original file + * @d_filp: file structure of the donor file + * @orig_start: start offset in block for orig + * @donor_start: start offset in block for donor + * @len: the number of blocks to be moved + * @moved_len: moved block length + * + * This function returns 0 and moved block length is set in moved_len + * if succeed, otherwise returns error value. + * + * Note: ext4_move_extents() proceeds the following order. + * 1:ext4_move_extents() calculates the last block number of moving extent + * function by the start block number (orig_start) and the number of blocks + * to be moved (len) specified as arguments. + * If the {orig, donor}_start points a hole, the extent's start offset + * pointed by ext_cur (current extent), holecheck_path, orig_path are set + * after hole behind. + * 2:Continue step 3 to step 5, until the holecheck_path points to last_extent + * or the ext_cur exceeds the block_end which is last logical block number. + * 3:To get the length of continues area, call mext_next_extent() + * specified with the ext_cur (initial value is holecheck_path) re-cursive, + * until find un-continuous extent, the start logical block number exceeds + * the block_end or the extent points to the last extent. + * 4:Exchange the original inode data with donor inode data + * from orig_page_offset to seq_end_page. + * The start indexes of data are specified as arguments. + * That of the original inode is orig_page_offset, + * and the donor inode is also orig_page_offset + * (To easily handle blocksize != pagesize case, the offset for the + * donor inode is block unit). + * 5:Update holecheck_path and orig_path to points a next proceeding extent, + * then returns to step 2. + * 6:Release holecheck_path, orig_path and set the len to moved_len + * which shows the number of moved blocks. + * The moved_len is useful for the command to calculate the file offset + * for starting next move extent ioctl. + * 7:Return 0 on success, or a negative error value on failure. + */ +int +ext4_move_extents(struct file *o_filp, struct file *d_filp, + __u64 orig_start, __u64 donor_start, __u64 len, + __u64 *moved_len) +{ + struct inode *orig_inode = o_filp->f_dentry->d_inode; + struct inode *donor_inode = d_filp->f_dentry->d_inode; + struct ext4_ext_path *orig_path = NULL, *holecheck_path = NULL; + struct ext4_extent *ext_prev, *ext_cur, *ext_dummy; + ext4_lblk_t block_start = orig_start; + ext4_lblk_t block_end, seq_start, add_blocks, file_end, seq_blocks = 0; + ext4_lblk_t rest_blocks; + pgoff_t orig_page_offset = 0, seq_end_page; + int ret, depth, last_extent = 0; + int blocks_per_page = PAGE_CACHE_SIZE >> orig_inode->i_blkbits; + int data_offset_in_page; + int block_len_in_page; + int uninit; + + /* protect orig and donor against a truncate */ + mext_inode_double_lock(orig_inode, donor_inode); + + mext_double_down_read(orig_inode, donor_inode); + /* Check the filesystem environment whether move_extent can be done */ + ret = mext_check_arguments(orig_inode, donor_inode, orig_start, + donor_start, &len, *moved_len); + mext_double_up_read(orig_inode, donor_inode); + if (ret) + goto out2; + + file_end = (i_size_read(orig_inode) - 1) >> orig_inode->i_blkbits; + block_end = block_start + len - 1; + if (file_end < block_end) + len -= block_end - file_end; + + get_ext_path(orig_path, orig_inode, block_start, ret); + if (orig_path == NULL) + goto out2; + + /* Get path structure to check the hole */ + get_ext_path(holecheck_path, orig_inode, block_start, ret); + if (holecheck_path == NULL) + goto out; + + depth = ext_depth(orig_inode); + ext_cur = holecheck_path[depth].p_ext; + if (ext_cur == NULL) { + ret = -EINVAL; + goto out; + } + + /* + * Get proper extent whose ee_block is beyond block_start + * if block_start was within the hole. + */ + if (le32_to_cpu(ext_cur->ee_block) + + ext4_ext_get_actual_len(ext_cur) - 1 < block_start) { + last_extent = mext_next_extent(orig_inode, + holecheck_path, &ext_cur); + if (last_extent < 0) { + ret = last_extent; + goto out; + } + last_extent = mext_next_extent(orig_inode, orig_path, + &ext_dummy); + if (last_extent < 0) { + ret = last_extent; + goto out; + } + } + seq_start = block_start; + + /* No blocks within the specified range. */ + if (le32_to_cpu(ext_cur->ee_block) > block_end) { + ext4_debug("ext4 move extent: The specified range of file " + "may be the hole\n"); + ret = -EINVAL; + goto out; + } + + /* Adjust start blocks */ + add_blocks = min(le32_to_cpu(ext_cur->ee_block) + + ext4_ext_get_actual_len(ext_cur), block_end + 1) - + max(le32_to_cpu(ext_cur->ee_block), block_start); + + while (!last_extent && le32_to_cpu(ext_cur->ee_block) <= block_end) { + seq_blocks += add_blocks; + + /* Adjust tail blocks */ + if (seq_start + seq_blocks - 1 > block_end) + seq_blocks = block_end - seq_start + 1; + + ext_prev = ext_cur; + last_extent = mext_next_extent(orig_inode, holecheck_path, + &ext_cur); + if (last_extent < 0) { + ret = last_extent; + break; + } + add_blocks = ext4_ext_get_actual_len(ext_cur); + + /* + * Extend the length of contiguous block (seq_blocks) + * if extents are contiguous. + */ + if (ext4_can_extents_be_merged(orig_inode, + ext_prev, ext_cur) && + block_end >= le32_to_cpu(ext_cur->ee_block) && + !last_extent) + continue; + + /* Is original extent is uninitialized */ + uninit = ext4_ext_is_uninitialized(ext_prev); + + data_offset_in_page = seq_start % blocks_per_page; + + /* + * Calculate data blocks count that should be swapped + * at the first page. + */ + if (data_offset_in_page + seq_blocks > blocks_per_page) { + /* Swapped blocks are across pages */ + block_len_in_page = + blocks_per_page - data_offset_in_page; + } else { + /* Swapped blocks are in a page */ + block_len_in_page = seq_blocks; + } + + orig_page_offset = seq_start >> + (PAGE_CACHE_SHIFT - orig_inode->i_blkbits); + seq_end_page = (seq_start + seq_blocks - 1) >> + (PAGE_CACHE_SHIFT - orig_inode->i_blkbits); + seq_start = le32_to_cpu(ext_cur->ee_block); + rest_blocks = seq_blocks; + + /* Discard preallocations of two inodes */ + down_write(&EXT4_I(orig_inode)->i_data_sem); + ext4_discard_preallocations(orig_inode); + up_write(&EXT4_I(orig_inode)->i_data_sem); + + down_write(&EXT4_I(donor_inode)->i_data_sem); + ext4_discard_preallocations(donor_inode); + up_write(&EXT4_I(donor_inode)->i_data_sem); + + while (orig_page_offset <= seq_end_page) { + + /* Swap original branches with new branches */ + ret = move_extent_par_page(o_filp, donor_inode, + orig_page_offset, + data_offset_in_page, + block_len_in_page, uninit); + if (ret < 0) + goto out; + orig_page_offset++; + /* Count how many blocks we have exchanged */ + *moved_len += block_len_in_page; + BUG_ON(*moved_len > len); + + data_offset_in_page = 0; + rest_blocks -= block_len_in_page; + if (rest_blocks > blocks_per_page) + block_len_in_page = blocks_per_page; + else + block_len_in_page = rest_blocks; + } + + /* Decrease buffer counter */ + if (holecheck_path) + ext4_ext_drop_refs(holecheck_path); + get_ext_path(holecheck_path, orig_inode, + seq_start, ret); + if (holecheck_path == NULL) + break; + depth = holecheck_path->p_depth; + + /* Decrease buffer counter */ + if (orig_path) + ext4_ext_drop_refs(orig_path); + get_ext_path(orig_path, orig_inode, seq_start, ret); + if (orig_path == NULL) + break; + + ext_cur = holecheck_path[depth].p_ext; + add_blocks = ext4_ext_get_actual_len(ext_cur); + seq_blocks = 0; + + } +out: + if (orig_path) { + ext4_ext_drop_refs(orig_path); + kfree(orig_path); + } + if (holecheck_path) { + ext4_ext_drop_refs(holecheck_path); + kfree(holecheck_path); + } +out2: + mext_inode_double_unlock(orig_inode, donor_inode); + + if (ret) + return ret; + + /* All of the specified blocks must be exchanged in succeed */ + BUG_ON(*moved_len != len); + + return 0; +} -- cgit v1.2.3 From 7f4520cc6242780ce720aa440ad4b391f998b558 Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Sat, 13 Jun 2009 10:09:41 -0400 Subject: ext4: change s_mount_opt to be an unsigned int We can only fit 32 options in s_mount_opt because an unsigned long is 32-bits on a x86 machine. So use an unsigned int to save space on 64-bit platforms. Signed-off-by: "Theodore Ts'o" --- fs/ext4/ext4.h | 2 +- fs/ext4/inode.c | 2 +- fs/ext4/super.c | 2 +- 3 files changed, 3 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 276a26f117e6..569f527080bf 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -852,7 +852,7 @@ struct ext4_sb_info { struct buffer_head * s_sbh; /* Buffer containing the super block */ struct ext4_super_block *s_es; /* Pointer to the super block in the buffer */ struct buffer_head **s_group_desc; - unsigned long s_mount_opt; + unsigned int s_mount_opt; ext4_fsblk_t s_sb_block; uid_t s_resuid; gid_t s_resgid; diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 2418ad36eab5..f8325a2bc897 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -93,7 +93,7 @@ int ext4_forget(handle_t *handle, int is_metadata, struct inode *inode, BUFFER_TRACE(bh, "enter"); jbd_debug(4, "forgetting bh %p: is_metadata = %d, mode %o, " - "data mode %lx\n", + "data mode %x\n", bh, is_metadata, inode->i_mode, test_opt(inode->i_sb, DATA_FLAGS)); diff --git a/fs/ext4/super.c b/fs/ext4/super.c index e8f0b2af4607..4c364ae7aeb1 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -1655,7 +1655,7 @@ static int ext4_setup_super(struct super_block *sb, struct ext4_super_block *es, ext4_commit_super(sb, 1); if (test_opt(sb, DEBUG)) printk(KERN_INFO "[EXT4 FS bs=%lu, gc=%u, " - "bpg=%lu, ipg=%lu, mo=%04lx]\n", + "bpg=%lu, ipg=%lu, mo=%04x]\n", sb->s_blocksize, sbi->s_groups_count, EXT4_BLOCKS_PER_GROUP(sb), -- cgit v1.2.3 From bc0b0d6d69ee9022f18ae264e62beb30ddeb322a Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Sat, 13 Jun 2009 10:09:48 -0400 Subject: ext4: update the s_last_mounted field in the superblock This field can be very helpful when a system administrator is trying to sort through large numbers of block devices or filesystem images. What is stored in this field can be ambiguous if multiple filesystem namespaces are in play; what we store in practice is the mountpoint interpreted by the process's namespace which first opens a file in the filesystem. Signed-off-by: "Theodore Ts'o" --- fs/ext4/ext4.h | 7 +++++++ fs/ext4/file.c | 36 +++++++++++++++++++++++++++++++++++- 2 files changed, 42 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 569f527080bf..9e268c97eeca 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -834,6 +834,12 @@ struct ext4_super_block { }; #ifdef __KERNEL__ + +/* + * Mount flags + */ +#define EXT4_MF_MNTDIR_SAMPLED 0x0001 + /* * fourth extended-fs super-block data in memory */ @@ -853,6 +859,7 @@ struct ext4_sb_info { struct ext4_super_block *s_es; /* Pointer to the super block in the buffer */ struct buffer_head **s_group_desc; unsigned int s_mount_opt; + unsigned int s_mount_flags; ext4_fsblk_t s_sb_block; uid_t s_resuid; gid_t s_resgid; diff --git a/fs/ext4/file.c b/fs/ext4/file.c index 588af8c77246..3f1873fef1c6 100644 --- a/fs/ext4/file.c +++ b/fs/ext4/file.c @@ -21,6 +21,8 @@ #include #include #include +#include +#include #include "ext4.h" #include "ext4_jbd2.h" #include "xattr.h" @@ -145,6 +147,38 @@ static int ext4_file_mmap(struct file *file, struct vm_area_struct *vma) return 0; } +static int ext4_file_open(struct inode * inode, struct file * filp) +{ + struct super_block *sb = inode->i_sb; + struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); + struct vfsmount *mnt = filp->f_path.mnt; + struct path path; + char buf[64], *cp; + + if (unlikely(!(sbi->s_mount_flags & EXT4_MF_MNTDIR_SAMPLED) && + !(sb->s_flags & MS_RDONLY))) { + sbi->s_mount_flags |= EXT4_MF_MNTDIR_SAMPLED; + /* + * Sample where the filesystem has been mounted and + * store it in the superblock for sysadmin convenience + * when trying to sort through large numbers of block + * devices or filesystem images. + */ + memset(buf, 0, sizeof(buf)); + path.mnt = mnt->mnt_parent; + path.dentry = mnt->mnt_mountpoint; + path_get(&path); + cp = d_path(&path, buf, sizeof(buf)); + path_put(&path); + if (!IS_ERR(cp)) { + memcpy(sbi->s_es->s_last_mounted, cp, + sizeof(sbi->s_es->s_last_mounted)); + sb->s_dirt = 1; + } + } + return generic_file_open(inode, filp); +} + const struct file_operations ext4_file_operations = { .llseek = generic_file_llseek, .read = do_sync_read, @@ -156,7 +190,7 @@ const struct file_operations ext4_file_operations = { .compat_ioctl = ext4_compat_ioctl, #endif .mmap = ext4_file_mmap, - .open = generic_file_open, + .open = ext4_file_open, .release = ext4_release_file, .fsync = ext4_sync_file, .splice_read = generic_file_splice_read, -- cgit v1.2.3 From 4ab2f15b7f709c3626a7eed075a7225b4c775c7e Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Sat, 13 Jun 2009 10:09:36 -0400 Subject: ext4: move the abort flag from s_mount_opts to s_mount_flags We're running out of space in the mount options word, and EXT4_MOUNT_ABORT isn't really a mount option, but a run-time flag. So move it to become EXT4_MF_FS_ABORTED in s_mount_flags. Also remove bogus ext2_fs.h / ext4.h simultaneous #include protection, which can never happen. Signed-off-by: "Theodore Ts'o" --- fs/ext4/ext4.h | 11 ++--------- fs/ext4/inode.c | 4 ++-- fs/ext4/super.c | 10 +++++----- 3 files changed, 9 insertions(+), 16 deletions(-) (limited to 'fs') diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 9e268c97eeca..06ee5a582917 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -684,7 +684,6 @@ struct ext4_inode_info { #define EXT4_MOUNT_ERRORS_PANIC 0x00040 /* Panic on errors */ #define EXT4_MOUNT_MINIX_DF 0x00080 /* Mimics the Minix statfs */ #define EXT4_MOUNT_NOLOAD 0x00100 /* Don't use existing journal*/ -#define EXT4_MOUNT_ABORT 0x00200 /* Fatal error detected */ #define EXT4_MOUNT_DATA_FLAGS 0x00C00 /* Mode for data writes: */ #define EXT4_MOUNT_JOURNAL_DATA 0x00400 /* Write data to journal */ #define EXT4_MOUNT_ORDERED_DATA 0x00800 /* Flush data before commit */ @@ -706,17 +705,10 @@ struct ext4_inode_info { #define EXT4_MOUNT_DATA_ERR_ABORT 0x10000000 /* Abort on file data write */ #define EXT4_MOUNT_BLOCK_VALIDITY 0x20000000 /* Block validity checking */ -/* Compatibility, for having both ext2_fs.h and ext4_fs.h included at once */ -#ifndef _LINUX_EXT2_FS_H #define clear_opt(o, opt) o &= ~EXT4_MOUNT_##opt #define set_opt(o, opt) o |= EXT4_MOUNT_##opt #define test_opt(sb, opt) (EXT4_SB(sb)->s_mount_opt & \ EXT4_MOUNT_##opt) -#else -#define EXT2_MOUNT_NOLOAD EXT4_MOUNT_NOLOAD -#define EXT2_MOUNT_ABORT EXT4_MOUNT_ABORT -#define EXT2_MOUNT_DATA_FLAGS EXT4_MOUNT_DATA_FLAGS -#endif #define ext4_set_bit ext2_set_bit #define ext4_set_bit_atomic ext2_set_bit_atomic @@ -836,9 +828,10 @@ struct ext4_super_block { #ifdef __KERNEL__ /* - * Mount flags + * run-time mount flags */ #define EXT4_MF_MNTDIR_SAMPLED 0x0001 +#define EXT4_MF_FS_ABORTED 0x0002 /* Fatal error detected */ /* * fourth extended-fs super-block data in memory diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index f8325a2bc897..5f927f6a1289 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -2670,13 +2670,13 @@ static int ext4_da_writepages(struct address_space *mapping, * If the filesystem has aborted, it is read-only, so return * right away instead of dumping stack traces later on that * will obscure the real source of the problem. We test - * EXT4_MOUNT_ABORT instead of sb->s_flag's MS_RDONLY because + * EXT4_MF_FS_ABORTED instead of sb->s_flag's MS_RDONLY because * the latter could be true if the filesystem is mounted * read-only, and in that case, ext4_da_writepages should * *never* be called, so if that ever happens, we would want * the stack trace. */ - if (unlikely(sbi->s_mount_opt & EXT4_MOUNT_ABORT)) + if (unlikely(sbi->s_mount_flags & EXT4_MF_FS_ABORTED)) return -EROFS; /* diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 4c364ae7aeb1..04486a53469f 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -303,7 +303,7 @@ static void ext4_handle_error(struct super_block *sb) if (!test_opt(sb, ERRORS_CONT)) { journal_t *journal = EXT4_SB(sb)->s_journal; - EXT4_SB(sb)->s_mount_opt |= EXT4_MOUNT_ABORT; + EXT4_SB(sb)->s_mount_flags |= EXT4_MF_FS_ABORTED; if (journal) jbd2_journal_abort(journal, -EIO); } @@ -416,7 +416,7 @@ void ext4_abort(struct super_block *sb, const char *function, ext4_msg(sb, KERN_CRIT, "Remounting filesystem read-only"); EXT4_SB(sb)->s_mount_state |= EXT4_ERROR_FS; sb->s_flags |= MS_RDONLY; - EXT4_SB(sb)->s_mount_opt |= EXT4_MOUNT_ABORT; + EXT4_SB(sb)->s_mount_flags |= EXT4_MF_FS_ABORTED; if (EXT4_SB(sb)->s_journal) jbd2_journal_abort(EXT4_SB(sb)->s_journal, -EIO); } @@ -1476,7 +1476,7 @@ set_qf_format: break; #endif case Opt_abort: - set_opt(sbi->s_mount_opt, ABORT); + sbi->s_mount_flags |= EXT4_MF_FS_ABORTED; break; case Opt_nobarrier: clear_opt(sbi->s_mount_opt, BARRIER); @@ -3452,7 +3452,7 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data) goto restore_opts; } - if (sbi->s_mount_opt & EXT4_MOUNT_ABORT) + if (sbi->s_mount_flags & EXT4_MF_FS_ABORTED) ext4_abort(sb, __func__, "Abort forced by user"); sb->s_flags = (sb->s_flags & ~MS_POSIXACL) | @@ -3467,7 +3467,7 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data) if ((*flags & MS_RDONLY) != (sb->s_flags & MS_RDONLY) || n_blocks_count > ext4_blocks_count(es)) { - if (sbi->s_mount_opt & EXT4_MOUNT_ABORT) { + if (sbi->s_mount_flags & EXT4_MF_FS_ABORTED) { err = -EROFS; goto restore_opts; } -- cgit v1.2.3 From f157a4aa98a18bd3817a72bea90d48494e2586e7 Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Sat, 13 Jun 2009 11:09:42 -0400 Subject: ext4: Use a hash of the topdir directory name for the Orlov parent group Instead of using a random number to determine the goal parent grop for the Orlov top directories, use a hash of the directory name. This allows for repeatable results when trying to benchmark filesystem layout algorithms. Signed-off-by: "Theodore Ts'o" --- fs/ext4/ext4.h | 3 ++- fs/ext4/ialloc.c | 19 ++++++++++++++----- fs/ext4/migrate.c | 5 ++--- fs/ext4/namei.c | 8 ++++---- 4 files changed, 22 insertions(+), 13 deletions(-) (limited to 'fs') diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 06ee5a582917..d035cf149e0e 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -1315,7 +1315,8 @@ extern int ext4fs_dirhash(const char *name, int len, struct dx_hash_info *hinfo); /* ialloc.c */ -extern struct inode * ext4_new_inode(handle_t *, struct inode *, int); +extern struct inode *ext4_new_inode(handle_t *, struct inode *, int, + const struct qstr *qstr); extern void ext4_free_inode(handle_t *, struct inode *); extern struct inode * ext4_orphan_get(struct super_block *, unsigned long); extern unsigned long ext4_count_free_inodes(struct super_block *); diff --git a/fs/ext4/ialloc.c b/fs/ext4/ialloc.c index 7d502f3be914..3f98ee712ff4 100644 --- a/fs/ext4/ialloc.c +++ b/fs/ext4/ialloc.c @@ -470,7 +470,8 @@ void get_orlov_stats(struct super_block *sb, ext4_group_t g, */ static int find_group_orlov(struct super_block *sb, struct inode *parent, - ext4_group_t *group, int mode) + ext4_group_t *group, int mode, + const struct qstr *qstr) { ext4_group_t parent_group = EXT4_I(parent)->i_block_group; struct ext4_sb_info *sbi = EXT4_SB(sb); @@ -485,6 +486,7 @@ static int find_group_orlov(struct super_block *sb, struct inode *parent, struct ext4_group_desc *desc; struct orlov_stats stats; int flex_size = ext4_flex_bg_size(sbi); + struct dx_hash_info hinfo; ngroups = real_ngroups; if (flex_size > 1) { @@ -506,7 +508,13 @@ static int find_group_orlov(struct super_block *sb, struct inode *parent, int best_ndir = inodes_per_group; int ret = -1; - get_random_bytes(&grp, sizeof(grp)); + if (qstr) { + hinfo.hash_version = DX_HASH_HALF_MD4; + hinfo.seed = sbi->s_hash_seed; + ext4fs_dirhash(qstr->name, qstr->len, &hinfo); + grp = hinfo.hash; + } else + get_random_bytes(&grp, sizeof(grp)); parent_group = (unsigned)grp % ngroups; for (i = 0; i < ngroups; i++) { g = (parent_group + i) % ngroups; @@ -649,7 +657,7 @@ static int find_group_other(struct super_block *sb, struct inode *parent, *group = parent_group + flex_size; if (*group > ngroups) *group = 0; - return find_group_orlov(sb, parent, group, mode); + return find_group_orlov(sb, parent, group, mode, 0); } /* @@ -790,7 +798,8 @@ err_ret: * For other inodes, search forward from the parent directory's block * group to find a free inode. */ -struct inode *ext4_new_inode(handle_t *handle, struct inode *dir, int mode) +struct inode *ext4_new_inode(handle_t *handle, struct inode *dir, int mode, + const struct qstr *qstr) { struct super_block *sb; struct buffer_head *inode_bitmap_bh = NULL; @@ -839,7 +848,7 @@ struct inode *ext4_new_inode(handle_t *handle, struct inode *dir, int mode) if (test_opt(sb, OLDALLOC)) ret2 = find_group_dir(sb, dir, &group); else - ret2 = find_group_orlov(sb, dir, &group, mode); + ret2 = find_group_orlov(sb, dir, &group, mode, qstr); } else ret2 = find_group_other(sb, dir, &group, mode); diff --git a/fs/ext4/migrate.c b/fs/ext4/migrate.c index fe64d9f79852..80d075b8aeaf 100644 --- a/fs/ext4/migrate.c +++ b/fs/ext4/migrate.c @@ -483,9 +483,8 @@ int ext4_ext_migrate(struct inode *inode) retval = PTR_ERR(handle); return retval; } - tmp_inode = ext4_new_inode(handle, - inode->i_sb->s_root->d_inode, - S_IFREG); + tmp_inode = ext4_new_inode(handle, inode->i_sb->s_root->d_inode, + S_IFREG, 0); if (IS_ERR(tmp_inode)) { retval = -ENOMEM; ext4_journal_stop(handle); diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c index 07eb6649e4fa..5f00d2418a83 100644 --- a/fs/ext4/namei.c +++ b/fs/ext4/namei.c @@ -1782,7 +1782,7 @@ retry: if (IS_DIRSYNC(dir)) ext4_handle_sync(handle); - inode = ext4_new_inode (handle, dir, mode); + inode = ext4_new_inode(handle, dir, mode, &dentry->d_name); err = PTR_ERR(inode); if (!IS_ERR(inode)) { inode->i_op = &ext4_file_inode_operations; @@ -1816,7 +1816,7 @@ retry: if (IS_DIRSYNC(dir)) ext4_handle_sync(handle); - inode = ext4_new_inode(handle, dir, mode); + inode = ext4_new_inode(handle, dir, mode, &dentry->d_name); err = PTR_ERR(inode); if (!IS_ERR(inode)) { init_special_inode(inode, inode->i_mode, rdev); @@ -1853,7 +1853,7 @@ retry: if (IS_DIRSYNC(dir)) ext4_handle_sync(handle); - inode = ext4_new_inode(handle, dir, S_IFDIR | mode); + inode = ext4_new_inode(handle, dir, S_IFDIR | mode, &dentry->d_name); err = PTR_ERR(inode); if (IS_ERR(inode)) goto out_stop; @@ -2264,7 +2264,7 @@ retry: if (IS_DIRSYNC(dir)) ext4_handle_sync(handle); - inode = ext4_new_inode(handle, dir, S_IFLNK|S_IRWXUGO); + inode = ext4_new_inode(handle, dir, S_IFLNK|S_IRWXUGO, &dentry->d_name); err = PTR_ERR(inode); if (IS_ERR(inode)) goto out_stop; -- cgit v1.2.3 From 11013911daea4820147ae6d7094dd7c6894e8651 Mon Sep 17 00:00:00 2001 From: Andreas Dilger Date: Sat, 13 Jun 2009 11:45:35 -0400 Subject: ext4: teach the inode allocator to use a goal inode number Enhance the inode allocator to take a goal inode number as a paremeter; if it is specified, it takes precedence over Orlov or parent directory inode allocation algorithms. The extents migration function uses the goal inode number so that the extent trees allocated the migration function use the correct flex_bg. In the future, the goal inode functionality will also be used to allocate an adjacent inode for the extended attributes. Also, for testing purposes the goal inode number can be specified via /sys/fs/{dev}/inode_goal. This can be useful for testing inode allocation beyond 2^32 blocks on very large filesystems. Signed-off-by: Andreas Dilger Signed-off-by: "Theodore Ts'o" --- Documentation/ABI/testing/sysfs-fs-ext4 | 10 ++++++++++ fs/ext4/ext4.h | 3 ++- fs/ext4/ialloc.c | 16 ++++++++++++---- fs/ext4/migrate.c | 5 ++++- fs/ext4/namei.c | 10 ++++++---- fs/ext4/super.c | 2 ++ 6 files changed, 36 insertions(+), 10 deletions(-) (limited to 'fs') diff --git a/Documentation/ABI/testing/sysfs-fs-ext4 b/Documentation/ABI/testing/sysfs-fs-ext4 index 4e79074de282..5fb709997d96 100644 --- a/Documentation/ABI/testing/sysfs-fs-ext4 +++ b/Documentation/ABI/testing/sysfs-fs-ext4 @@ -79,3 +79,13 @@ Description: This file is read-only and shows the number of kilobytes of data that have been written to this filesystem since it was mounted. + +What: /sys/fs/ext4//inode_goal +Date: June 2008 +Contact: "Theodore Ts'o" +Description: + Tuning parameter which (if non-zero) controls the goal + inode used by the inode allocator in p0reference to + all other allocation hueristics. This is intended for + debugging use only, and should be 0 on production + systems. diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index d035cf149e0e..746cdcba969d 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -863,6 +863,7 @@ struct ext4_sb_info { int s_inode_size; int s_first_ino; unsigned int s_inode_readahead_blks; + unsigned int s_inode_goal; spinlock_t s_next_gen_lock; u32 s_next_generation; u32 s_hash_seed[4]; @@ -1316,7 +1317,7 @@ extern int ext4fs_dirhash(const char *name, int len, struct /* ialloc.c */ extern struct inode *ext4_new_inode(handle_t *, struct inode *, int, - const struct qstr *qstr); + const struct qstr *qstr, __u32 goal); extern void ext4_free_inode(handle_t *, struct inode *); extern struct inode * ext4_orphan_get(struct super_block *, unsigned long); extern unsigned long ext4_count_free_inodes(struct super_block *); diff --git a/fs/ext4/ialloc.c b/fs/ext4/ialloc.c index 3f98ee712ff4..2f645732e3b7 100644 --- a/fs/ext4/ialloc.c +++ b/fs/ext4/ialloc.c @@ -799,7 +799,7 @@ err_ret: * group to find a free inode. */ struct inode *ext4_new_inode(handle_t *handle, struct inode *dir, int mode, - const struct qstr *qstr) + const struct qstr *qstr, __u32 goal) { struct super_block *sb; struct buffer_head *inode_bitmap_bh = NULL; @@ -830,6 +830,16 @@ struct inode *ext4_new_inode(handle_t *handle, struct inode *dir, int mode, ei = EXT4_I(inode); sbi = EXT4_SB(sb); + if (!goal) + goal = sbi->s_inode_goal; + + if (goal && goal < le32_to_cpu(sbi->s_es->s_inodes_count)) { + group = (goal - 1) / EXT4_INODES_PER_GROUP(sb); + ino = (goal - 1) % EXT4_INODES_PER_GROUP(sb); + ret2 = 0; + goto got_group; + } + if (sbi->s_log_groups_per_flex && test_opt(sb, OLDALLOC)) { ret2 = find_group_flex(sb, dir, &group); if (ret2 == -1) { @@ -858,7 +868,7 @@ got_group: if (ret2 == -1) goto out; - for (i = 0; i < ngroups; i++) { + for (i = 0; i < ngroups; i++, ino = 0) { err = -EIO; gdp = ext4_get_group_desc(sb, group, &group_desc_bh); @@ -870,8 +880,6 @@ got_group: if (!inode_bitmap_bh) goto fail; - ino = 0; - repeat_in_this_group: ino = ext4_find_next_zero_bit((unsigned long *) inode_bitmap_bh->b_data, diff --git a/fs/ext4/migrate.c b/fs/ext4/migrate.c index 80d075b8aeaf..313a50b39741 100644 --- a/fs/ext4/migrate.c +++ b/fs/ext4/migrate.c @@ -458,6 +458,7 @@ int ext4_ext_migrate(struct inode *inode) struct inode *tmp_inode = NULL; struct list_blocks_struct lb; unsigned long max_entries; + __u32 goal; /* * If the filesystem does not support extents, or the inode @@ -483,8 +484,10 @@ int ext4_ext_migrate(struct inode *inode) retval = PTR_ERR(handle); return retval; } + goal = (((inode->i_ino - 1) / EXT4_INODES_PER_GROUP(inode->i_sb)) * + EXT4_INODES_PER_GROUP(inode->i_sb)) + 1; tmp_inode = ext4_new_inode(handle, inode->i_sb->s_root->d_inode, - S_IFREG, 0); + S_IFREG, 0, goal); if (IS_ERR(tmp_inode)) { retval = -ENOMEM; ext4_journal_stop(handle); diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c index 5f00d2418a83..de04013d16ff 100644 --- a/fs/ext4/namei.c +++ b/fs/ext4/namei.c @@ -1782,7 +1782,7 @@ retry: if (IS_DIRSYNC(dir)) ext4_handle_sync(handle); - inode = ext4_new_inode(handle, dir, mode, &dentry->d_name); + inode = ext4_new_inode(handle, dir, mode, &dentry->d_name, 0); err = PTR_ERR(inode); if (!IS_ERR(inode)) { inode->i_op = &ext4_file_inode_operations; @@ -1816,7 +1816,7 @@ retry: if (IS_DIRSYNC(dir)) ext4_handle_sync(handle); - inode = ext4_new_inode(handle, dir, mode, &dentry->d_name); + inode = ext4_new_inode(handle, dir, mode, &dentry->d_name, 0); err = PTR_ERR(inode); if (!IS_ERR(inode)) { init_special_inode(inode, inode->i_mode, rdev); @@ -1853,7 +1853,8 @@ retry: if (IS_DIRSYNC(dir)) ext4_handle_sync(handle); - inode = ext4_new_inode(handle, dir, S_IFDIR | mode, &dentry->d_name); + inode = ext4_new_inode(handle, dir, S_IFDIR | mode, + &dentry->d_name, 0); err = PTR_ERR(inode); if (IS_ERR(inode)) goto out_stop; @@ -2264,7 +2265,8 @@ retry: if (IS_DIRSYNC(dir)) ext4_handle_sync(handle); - inode = ext4_new_inode(handle, dir, S_IFLNK|S_IRWXUGO, &dentry->d_name); + inode = ext4_new_inode(handle, dir, S_IFLNK|S_IRWXUGO, + &dentry->d_name, 0); err = PTR_ERR(inode); if (IS_ERR(inode)) goto out_stop; diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 04486a53469f..23013d303f81 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -2206,6 +2206,7 @@ EXT4_RO_ATTR(session_write_kbytes); EXT4_RO_ATTR(lifetime_write_kbytes); EXT4_ATTR_OFFSET(inode_readahead_blks, 0644, sbi_ui_show, inode_readahead_blks_store, s_inode_readahead_blks); +EXT4_RW_ATTR_SBI_UI(inode_goal, s_inode_goal); EXT4_RW_ATTR_SBI_UI(mb_stats, s_mb_stats); EXT4_RW_ATTR_SBI_UI(mb_max_to_scan, s_mb_max_to_scan); EXT4_RW_ATTR_SBI_UI(mb_min_to_scan, s_mb_min_to_scan); @@ -2218,6 +2219,7 @@ static struct attribute *ext4_attrs[] = { ATTR_LIST(session_write_kbytes), ATTR_LIST(lifetime_write_kbytes), ATTR_LIST(inode_readahead_blks), + ATTR_LIST(inode_goal), ATTR_LIST(mb_stats), ATTR_LIST(mb_max_to_scan), ATTR_LIST(mb_min_to_scan), -- cgit v1.2.3 From 0610b6e99939828b77eec020ead0e1f44cba38ca Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Mon, 15 Jun 2009 03:45:05 -0400 Subject: ext4: Fix 64-bit block type problem on 32-bit platforms The function ext4_mb_free_blocks() was using an "unsigned long" to pass a block number; this will cause 64-bit block numbers to get truncated on x86 and other 32-bit platforms. Signed-off-by: "Theodore Ts'o" Reviewed-by: Eric Sandeen --- fs/ext4/ext4.h | 2 +- fs/ext4/mballoc.c | 8 ++++---- 2 files changed, 5 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 746cdcba969d..17b9998680e3 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -1341,7 +1341,7 @@ extern void ext4_discard_preallocations(struct inode *); extern int __init init_ext4_mballoc(void); extern void exit_ext4_mballoc(void); extern void ext4_mb_free_blocks(handle_t *, struct inode *, - unsigned long, unsigned long, int, unsigned long *); + ext4_fsblk_t, unsigned long, int, unsigned long *); extern int ext4_mb_add_groupinfo(struct super_block *sb, ext4_group_t i, struct ext4_group_desc *desc); extern void ext4_mb_update_group_info(struct ext4_group_info *grp, diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index 8d98070b48fb..519a0a686d94 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -4689,7 +4689,7 @@ ext4_mb_free_metadata(handle_t *handle, struct ext4_buddy *e4b, * Main entry point into mballoc to free blocks */ void ext4_mb_free_blocks(handle_t *handle, struct inode *inode, - unsigned long block, unsigned long count, + ext4_fsblk_t block, unsigned long count, int metadata, unsigned long *freed) { struct buffer_head *bitmap_bh = NULL; @@ -4715,11 +4715,11 @@ void ext4_mb_free_blocks(handle_t *handle, struct inode *inode, block + count > ext4_blocks_count(es)) { ext4_error(sb, __func__, "Freeing blocks not in datazone - " - "block = %lu, count = %lu", block, count); + "block = %llu, count = %lu", block, count); goto error_return; } - ext4_debug("freeing block %lu\n", block); + ext4_debug("freeing block %llu\n", block); trace_ext4_free_blocks(inode, block, count, metadata); ac = kmem_cache_alloc(ext4_ac_cachep, GFP_NOFS); @@ -4761,7 +4761,7 @@ do_more: ext4_error(sb, __func__, "Freeing blocks in system zone - " - "Block = %lu, count = %lu", block, count); + "Block = %llu, count = %lu", block, count); /* err = 0. ext4_std_error should be a no op */ goto error_return; } -- cgit v1.2.3 From de9a55b841132f7ae097f6e31ccebad2d5030cf5 Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Sun, 14 Jun 2009 17:45:34 -0400 Subject: ext4: Fix up whitespace issues in fs/ext4/inode.c This is a pure cleanup patch. Signed-off-by: "Theodore Ts'o" --- fs/ext4/inode.c | 200 +++++++++++++++++++++++++++++--------------------------- 1 file changed, 103 insertions(+), 97 deletions(-) (limited to 'fs') diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 5f927f6a1289..8d0908afbd5b 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -81,7 +81,7 @@ static int ext4_inode_is_fast_symlink(struct inode *inode) * If the handle isn't valid we're not journaling so there's nothing to do. */ int ext4_forget(handle_t *handle, int is_metadata, struct inode *inode, - struct buffer_head *bh, ext4_fsblk_t blocknr) + struct buffer_head *bh, ext4_fsblk_t blocknr) { int err; @@ -332,8 +332,8 @@ static inline void add_chain(Indirect *p, struct buffer_head *bh, __le32 *v) */ static int ext4_block_to_path(struct inode *inode, - ext4_lblk_t i_block, - ext4_lblk_t offsets[4], int *boundary) + ext4_lblk_t i_block, + ext4_lblk_t offsets[4], int *boundary) { int ptrs = EXT4_ADDR_PER_BLOCK(inode->i_sb); int ptrs_bits = EXT4_ADDR_PER_BLOCK_BITS(inode->i_sb); @@ -365,9 +365,9 @@ static int ext4_block_to_path(struct inode *inode, final = ptrs; } else { ext4_warning(inode->i_sb, "ext4_block_to_path", - "block %lu > max in inode %lu", - i_block + direct_blocks + - indirect_blocks + double_blocks, inode->i_ino); + "block %lu > max in inode %lu", + i_block + direct_blocks + + indirect_blocks + double_blocks, inode->i_ino); } if (boundary) *boundary = final - 1 - (i_block & (ptrs - 1)); @@ -382,25 +382,25 @@ static int __ext4_check_blockref(const char *function, struct inode *inode, while (bref < p+max) { blk = le32_to_cpu(*bref++); - if (blk && - unlikely(!ext4_data_block_valid(EXT4_SB(inode->i_sb), + if (blk && + unlikely(!ext4_data_block_valid(EXT4_SB(inode->i_sb), blk, 1))) { ext4_error(inode->i_sb, function, "invalid block reference %u " "in inode #%lu", blk, inode->i_ino); - return -EIO; - } - } - return 0; + return -EIO; + } + } + return 0; } #define ext4_check_indirect_blockref(inode, bh) \ - __ext4_check_blockref(__func__, inode, (__le32 *)(bh)->b_data, \ + __ext4_check_blockref(__func__, inode, (__le32 *)(bh)->b_data, \ EXT4_ADDR_PER_BLOCK((inode)->i_sb)) #define ext4_check_inode_blockref(inode) \ - __ext4_check_blockref(__func__, inode, EXT4_I(inode)->i_data, \ + __ext4_check_blockref(__func__, inode, EXT4_I(inode)->i_data, \ EXT4_NDIR_BLOCKS) /** @@ -450,7 +450,7 @@ static Indirect *ext4_get_branch(struct inode *inode, int depth, bh = sb_getblk(sb, le32_to_cpu(p->key)); if (unlikely(!bh)) goto failure; - + if (!bh_uptodate_or_lock(bh)) { if (bh_submit_read(bh) < 0) { put_bh(bh); @@ -462,7 +462,7 @@ static Indirect *ext4_get_branch(struct inode *inode, int depth, goto failure; } } - + add_chain(++p, bh, (__le32 *)bh->b_data + *++offsets); /* Reader: end */ if (!p->key) @@ -555,7 +555,7 @@ static ext4_fsblk_t ext4_find_near(struct inode *inode, Indirect *ind) * returns it. */ static ext4_fsblk_t ext4_find_goal(struct inode *inode, ext4_lblk_t block, - Indirect *partial) + Indirect *partial) { /* * XXX need to get goal block from mballoc's data structures @@ -577,7 +577,7 @@ static ext4_fsblk_t ext4_find_goal(struct inode *inode, ext4_lblk_t block, * direct and indirect blocks. */ static int ext4_blks_to_allocate(Indirect *branch, int k, unsigned int blks, - int blocks_to_boundary) + int blocks_to_boundary) { unsigned int count = 0; @@ -613,9 +613,9 @@ static int ext4_blks_to_allocate(Indirect *branch, int k, unsigned int blks, * direct blocks */ static int ext4_alloc_blocks(handle_t *handle, struct inode *inode, - ext4_lblk_t iblock, ext4_fsblk_t goal, - int indirect_blks, int blks, - ext4_fsblk_t new_blocks[4], int *err) + ext4_lblk_t iblock, ext4_fsblk_t goal, + int indirect_blks, int blks, + ext4_fsblk_t new_blocks[4], int *err) { struct ext4_allocation_request ar; int target, i; @@ -686,10 +686,10 @@ static int ext4_alloc_blocks(handle_t *handle, struct inode *inode, } if (!*err) { if (target == blks) { - /* - * save the new block number - * for the first direct block - */ + /* + * save the new block number + * for the first direct block + */ new_blocks[index] = current_block; } blk_allocated += ar.len; @@ -731,9 +731,9 @@ failed_out: * as described above and return 0. */ static int ext4_alloc_branch(handle_t *handle, struct inode *inode, - ext4_lblk_t iblock, int indirect_blks, - int *blks, ext4_fsblk_t goal, - ext4_lblk_t *offsets, Indirect *branch) + ext4_lblk_t iblock, int indirect_blks, + int *blks, ext4_fsblk_t goal, + ext4_lblk_t *offsets, Indirect *branch) { int blocksize = inode->i_sb->s_blocksize; int i, n = 0; @@ -780,7 +780,7 @@ static int ext4_alloc_branch(handle_t *handle, struct inode *inode, * the chain to point to the new allocated * data blocks numbers */ - for (i=1; i < num; i++) + for (i = 1; i < num; i++) *(branch[n].p + i) = cpu_to_le32(++current_block); } BUFFER_TRACE(bh, "marking uptodate"); @@ -823,7 +823,8 @@ failed: * chain to new block and return 0. */ static int ext4_splice_branch(handle_t *handle, struct inode *inode, - ext4_lblk_t block, Indirect *where, int num, int blks) + ext4_lblk_t block, Indirect *where, int num, + int blks) { int i; int err = 0; @@ -924,9 +925,9 @@ err_out: * blocks. */ static int ext4_ind_get_blocks(handle_t *handle, struct inode *inode, - ext4_lblk_t iblock, unsigned int maxblocks, - struct buffer_head *bh_result, - int flags) + ext4_lblk_t iblock, unsigned int maxblocks, + struct buffer_head *bh_result, + int flags) { int err = -EIO; ext4_lblk_t offsets[4]; @@ -942,7 +943,7 @@ static int ext4_ind_get_blocks(handle_t *handle, struct inode *inode, J_ASSERT(!(EXT4_I(inode)->i_flags & EXT4_EXTENTS_FL)); J_ASSERT(handle != NULL || (flags & EXT4_GET_BLOCKS_CREATE) == 0); depth = ext4_block_to_path(inode, iblock, offsets, - &blocks_to_boundary); + &blocks_to_boundary); if (depth == 0) goto out; @@ -990,8 +991,8 @@ static int ext4_ind_get_blocks(handle_t *handle, struct inode *inode, * Block out ext4_truncate while we alter the tree */ err = ext4_alloc_branch(handle, inode, iblock, indirect_blks, - &count, goal, - offsets + (partial - chain), partial); + &count, goal, + offsets + (partial - chain), partial); /* * The ext4_splice_branch call will free and forget any buffers @@ -1002,8 +1003,8 @@ static int ext4_ind_get_blocks(handle_t *handle, struct inode *inode, */ if (!err) err = ext4_splice_branch(handle, inode, iblock, - partial, indirect_blks, count); - else + partial, indirect_blks, count); + else goto cleanup; set_buffer_new(bh_result); @@ -1175,7 +1176,7 @@ int ext4_get_blocks(handle_t *handle, struct inode *inode, sector_t block, up_read((&EXT4_I(inode)->i_data_sem)); if (retval > 0 && buffer_mapped(bh)) { - int ret = check_block_validity(inode, block, + int ret = check_block_validity(inode, block, bh->b_blocknr, retval); if (ret != 0) return ret; @@ -1257,7 +1258,7 @@ int ext4_get_blocks(handle_t *handle, struct inode *inode, sector_t block, up_write((&EXT4_I(inode)->i_data_sem)); if (retval > 0 && buffer_mapped(bh)) { - int ret = check_block_validity(inode, block, + int ret = check_block_validity(inode, block, bh->b_blocknr, retval); if (ret != 0) return ret; @@ -1408,8 +1409,7 @@ static int walk_page_buffers(handle_t *handle, for (bh = head, block_start = 0; ret == 0 && (bh != head || !block_start); - block_start = block_end, bh = next) - { + block_start = block_end, bh = next) { next = bh->b_this_page; block_end = block_start + blocksize; if (block_end <= from || block_start >= to) { @@ -1450,7 +1450,7 @@ static int walk_page_buffers(handle_t *handle, * write. */ static int do_journal_get_write_access(handle_t *handle, - struct buffer_head *bh) + struct buffer_head *bh) { if (!buffer_mapped(bh) || buffer_freed(bh)) return 0; @@ -1458,15 +1458,15 @@ static int do_journal_get_write_access(handle_t *handle, } static int ext4_write_begin(struct file *file, struct address_space *mapping, - loff_t pos, unsigned len, unsigned flags, - struct page **pagep, void **fsdata) + loff_t pos, unsigned len, unsigned flags, + struct page **pagep, void **fsdata) { struct inode *inode = mapping->host; int ret, needed_blocks; handle_t *handle; int retries = 0; struct page *page; - pgoff_t index; + pgoff_t index; unsigned from, to; trace_ext4_write_begin(inode, pos, len, flags); @@ -1475,7 +1475,7 @@ static int ext4_write_begin(struct file *file, struct address_space *mapping, * we allocate blocks but write fails for some reason */ needed_blocks = ext4_writepage_trans_blocks(inode) + 1; - index = pos >> PAGE_CACHE_SHIFT; + index = pos >> PAGE_CACHE_SHIFT; from = pos & (PAGE_CACHE_SIZE - 1); to = from + len; @@ -1523,7 +1523,7 @@ retry: ext4_journal_stop(handle); if (pos + len > inode->i_size) { vmtruncate(inode, inode->i_size); - /* + /* * If vmtruncate failed early the inode might * still be on the orphan list; we need to * make sure the inode is removed from the @@ -1550,9 +1550,9 @@ static int write_end_fn(handle_t *handle, struct buffer_head *bh) } static int ext4_generic_write_end(struct file *file, - struct address_space *mapping, - loff_t pos, unsigned len, unsigned copied, - struct page *page, void *fsdata) + struct address_space *mapping, + loff_t pos, unsigned len, unsigned copied, + struct page *page, void *fsdata) { int i_size_changed = 0; struct inode *inode = mapping->host; @@ -1603,9 +1603,9 @@ static int ext4_generic_write_end(struct file *file, * buffers are managed internally. */ static int ext4_ordered_write_end(struct file *file, - struct address_space *mapping, - loff_t pos, unsigned len, unsigned copied, - struct page *page, void *fsdata) + struct address_space *mapping, + loff_t pos, unsigned len, unsigned copied, + struct page *page, void *fsdata) { handle_t *handle = ext4_journal_current_handle(); struct inode *inode = mapping->host; @@ -1633,7 +1633,7 @@ static int ext4_ordered_write_end(struct file *file, if (pos + len > inode->i_size) { vmtruncate(inode, inode->i_size); - /* + /* * If vmtruncate failed early the inode might still be * on the orphan list; we need to make sure the inode * is removed from the orphan list in that case. @@ -1647,9 +1647,9 @@ static int ext4_ordered_write_end(struct file *file, } static int ext4_writeback_write_end(struct file *file, - struct address_space *mapping, - loff_t pos, unsigned len, unsigned copied, - struct page *page, void *fsdata) + struct address_space *mapping, + loff_t pos, unsigned len, unsigned copied, + struct page *page, void *fsdata) { handle_t *handle = ext4_journal_current_handle(); struct inode *inode = mapping->host; @@ -1675,7 +1675,7 @@ static int ext4_writeback_write_end(struct file *file, if (pos + len > inode->i_size) { vmtruncate(inode, inode->i_size); - /* + /* * If vmtruncate failed early the inode might still be * on the orphan list; we need to make sure the inode * is removed from the orphan list in that case. @@ -1688,9 +1688,9 @@ static int ext4_writeback_write_end(struct file *file, } static int ext4_journalled_write_end(struct file *file, - struct address_space *mapping, - loff_t pos, unsigned len, unsigned copied, - struct page *page, void *fsdata) + struct address_space *mapping, + loff_t pos, unsigned len, unsigned copied, + struct page *page, void *fsdata) { handle_t *handle = ext4_journal_current_handle(); struct inode *inode = mapping->host; @@ -1738,7 +1738,7 @@ static int ext4_journalled_write_end(struct file *file, ret = ret2; if (pos + len > inode->i_size) { vmtruncate(inode, inode->i_size); - /* + /* * If vmtruncate failed early the inode might still be * on the orphan list; we need to make sure the inode * is removed from the orphan list in that case. @@ -1845,7 +1845,7 @@ static void ext4_da_release_space(struct inode *inode, int to_free) } static void ext4_da_page_release_reservation(struct page *page, - unsigned long offset) + unsigned long offset) { int to_release = 0; struct buffer_head *head, *bh; @@ -2854,8 +2854,8 @@ static int ext4_nonda_switch(struct super_block *sb) } static int ext4_da_write_begin(struct file *file, struct address_space *mapping, - loff_t pos, unsigned len, unsigned flags, - struct page **pagep, void **fsdata) + loff_t pos, unsigned len, unsigned flags, + struct page **pagep, void **fsdata) { int ret, retries = 0; struct page *page; @@ -2925,7 +2925,7 @@ out: * when write to the end of file but not require block allocation */ static int ext4_da_should_update_i_disksize(struct page *page, - unsigned long offset) + unsigned long offset) { struct buffer_head *bh; struct inode *inode = page->mapping->host; @@ -2944,9 +2944,9 @@ static int ext4_da_should_update_i_disksize(struct page *page, } static int ext4_da_write_end(struct file *file, - struct address_space *mapping, - loff_t pos, unsigned len, unsigned copied, - struct page *page, void *fsdata) + struct address_space *mapping, + loff_t pos, unsigned len, unsigned copied, + struct page *page, void *fsdata) { struct inode *inode = mapping->host; int ret = 0, ret2; @@ -3044,7 +3044,7 @@ int ext4_alloc_da_blocks(struct inode *inode) * not strictly speaking necessary (and for users of * laptop_mode, not even desirable). However, to do otherwise * would require replicating code paths in: - * + * * ext4_da_writepages() -> * write_cache_pages() ---> (via passed in callback function) * __mpage_da_writepage() --> @@ -3064,7 +3064,7 @@ int ext4_alloc_da_blocks(struct inode *inode) * write out the pages, but rather only collect contiguous * logical block extents, call the multi-block allocator, and * then update the buffer heads with the block allocations. - * + * * For now, though, we'll cheat by calling filemap_flush(), * which will map the blocks, and start the I/O, but not * actually wait for the I/O to complete. @@ -3200,7 +3200,7 @@ static int bput_one(handle_t *handle, struct buffer_head *bh) * */ static int __ext4_normal_writepage(struct page *page, - struct writeback_control *wbc) + struct writeback_control *wbc) { struct inode *inode = page->mapping->host; @@ -3212,7 +3212,7 @@ static int __ext4_normal_writepage(struct page *page, } static int ext4_normal_writepage(struct page *page, - struct writeback_control *wbc) + struct writeback_control *wbc) { struct inode *inode = page->mapping->host; loff_t size = i_size_read(inode); @@ -3248,7 +3248,7 @@ static int ext4_normal_writepage(struct page *page, } static int __ext4_journalled_writepage(struct page *page, - struct writeback_control *wbc) + struct writeback_control *wbc) { struct address_space *mapping = page->mapping; struct inode *inode = mapping->host; @@ -3298,7 +3298,7 @@ out: } static int ext4_journalled_writepage(struct page *page, - struct writeback_control *wbc) + struct writeback_control *wbc) { struct inode *inode = page->mapping->host; loff_t size = i_size_read(inode); @@ -3401,8 +3401,8 @@ static int ext4_releasepage(struct page *page, gfp_t wait) * VFS code falls back into buffered path in that case so we are safe. */ static ssize_t ext4_direct_IO(int rw, struct kiocb *iocb, - const struct iovec *iov, loff_t offset, - unsigned long nr_segs) + const struct iovec *iov, loff_t offset, + unsigned long nr_segs) { struct file *file = iocb->ki_filp; struct inode *inode = file->f_mapping->host; @@ -3722,7 +3722,8 @@ static inline int all_zeroes(__le32 *p, __le32 *q) * (no partially truncated stuff there). */ static Indirect *ext4_find_shared(struct inode *inode, int depth, - ext4_lblk_t offsets[4], Indirect chain[4], __le32 *top) + ext4_lblk_t offsets[4], Indirect chain[4], + __le32 *top) { Indirect *partial, *p; int k, err; @@ -3778,8 +3779,10 @@ no_top: * than `count' because there can be holes in there. */ static void ext4_clear_blocks(handle_t *handle, struct inode *inode, - struct buffer_head *bh, ext4_fsblk_t block_to_free, - unsigned long count, __le32 *first, __le32 *last) + struct buffer_head *bh, + ext4_fsblk_t block_to_free, + unsigned long count, __le32 *first, + __le32 *last) { __le32 *p; if (try_to_extend_transaction(handle, inode)) { @@ -3796,10 +3799,11 @@ static void ext4_clear_blocks(handle_t *handle, struct inode *inode, } /* - * Any buffers which are on the journal will be in memory. We find - * them on the hash table so jbd2_journal_revoke() will run jbd2_journal_forget() - * on them. We've already detached each block from the file, so - * bforget() in jbd2_journal_forget() should be safe. + * Any buffers which are on the journal will be in memory. We + * find them on the hash table so jbd2_journal_revoke() will + * run jbd2_journal_forget() on them. We've already detached + * each block from the file, so bforget() in + * jbd2_journal_forget() should be safe. * * AKPM: turn on bforget in jbd2_journal_forget()!!! */ @@ -4171,7 +4175,7 @@ void ext4_truncate(struct inode *inode) (__le32*)partial->bh->b_data+addr_per_block, (chain+n-1) - partial); BUFFER_TRACE(partial->bh, "call brelse"); - brelse (partial->bh); + brelse(partial->bh); partial--; } do_indirects: @@ -4412,8 +4416,9 @@ void ext4_get_inode_flags(struct ext4_inode_info *ei) if (flags & S_DIRSYNC) ei->i_flags |= EXT4_DIRSYNC_FL; } + static blkcnt_t ext4_inode_blocks(struct ext4_inode *raw_inode, - struct ext4_inode_info *ei) + struct ext4_inode_info *ei) { blkcnt_t i_blocks ; struct inode *inode = &(ei->vfs_inode); @@ -4528,7 +4533,7 @@ struct inode *ext4_iget(struct super_block *sb, unsigned long ino) EXT4_GOOD_OLD_INODE_SIZE + ei->i_extra_isize; if (*magic == cpu_to_le32(EXT4_XATTR_MAGIC)) - ei->i_state |= EXT4_STATE_XATTR; + ei->i_state |= EXT4_STATE_XATTR; } } else ei->i_extra_isize = 0; @@ -4547,7 +4552,7 @@ struct inode *ext4_iget(struct super_block *sb, unsigned long ino) ret = 0; if (ei->i_file_acl && - ((ei->i_file_acl < + ((ei->i_file_acl < (le32_to_cpu(EXT4_SB(sb)->s_es->s_first_data_block) + EXT4_SB(sb)->s_gdb_count)) || (ei->i_file_acl >= ext4_blocks_count(EXT4_SB(sb)->s_es)))) { @@ -4562,15 +4567,15 @@ struct inode *ext4_iget(struct super_block *sb, unsigned long ino) !ext4_inode_is_fast_symlink(inode))) /* Validate extent which is part of inode */ ret = ext4_ext_check_inode(inode); - } else if (S_ISREG(inode->i_mode) || S_ISDIR(inode->i_mode) || + } else if (S_ISREG(inode->i_mode) || S_ISDIR(inode->i_mode) || (S_ISLNK(inode->i_mode) && !ext4_inode_is_fast_symlink(inode))) { - /* Validate block references which are part of inode */ + /* Validate block references which are part of inode */ ret = ext4_check_inode_blockref(inode); } if (ret) { - brelse(bh); - goto bad_inode; + brelse(bh); + goto bad_inode; } if (S_ISREG(inode->i_mode)) { @@ -4601,7 +4606,7 @@ struct inode *ext4_iget(struct super_block *sb, unsigned long ino) } else { brelse(bh); ret = -EIO; - ext4_error(inode->i_sb, __func__, + ext4_error(inode->i_sb, __func__, "bogus i_mode (%o) for inode=%lu", inode->i_mode, inode->i_ino); goto bad_inode; @@ -4754,8 +4759,9 @@ static int ext4_do_update_inode(handle_t *handle, cpu_to_le32(new_encode_dev(inode->i_rdev)); raw_inode->i_block[2] = 0; } - } else for (block = 0; block < EXT4_N_BLOCKS; block++) - raw_inode->i_block[block] = ei->i_data[block]; + } else + for (block = 0; block < EXT4_N_BLOCKS; block++) + raw_inode->i_block[block] = ei->i_data[block]; raw_inode->i_disk_version = cpu_to_le32(inode->i_version); if (ei->i_extra_isize) { @@ -5109,7 +5115,7 @@ int ext4_chunk_trans_blocks(struct inode *inode, int nrblocks) * Give this, we know that the caller already has write access to iloc->bh. */ int ext4_mark_iloc_dirty(handle_t *handle, - struct inode *inode, struct ext4_iloc *iloc) + struct inode *inode, struct ext4_iloc *iloc) { int err = 0; -- cgit v1.2.3 From c364b22c9580a885e0f8c0d0f9710d67dc448958 Mon Sep 17 00:00:00 2001 From: "Aneesh Kumar K.V" Date: Sun, 14 Jun 2009 17:57:10 -0400 Subject: ext4: Fix mmap/truncate race when blocksize < pagesize && delayed allocation It is possible to see buffer_heads which are not mapped in the writepage callback in the following scneario (where the fs blocksize is 1k and the page size is 4k): 1) truncate(f, 1024) 2) mmap(f, 0, 4096) 3) a[0] = 'a' 4) truncate(f, 4096) 5) writepage(...) Now if we get a writepage callback immediately after (4) and before an attempt to write at any other offset via mmap address (which implies we are yet to get a pagefault and do a get_block) what we would have is the page which is dirty have first block allocated and the other three buffer_heads unmapped. In the above case the writepage should go ahead and try to write the first blocks and clear the page_dirty flag. Further attempts to write to the page will again create a fault and result in allocating blocks and marking page dirty. If we don't write any other offset via mmap address we would still have written the first block to the disk and rest of the space will be considered as a hole. So to address this, we change all of the places where we look for delayed, unmapped, or unwritten buffer heads, and only check for delayed or unwritten buffer heads instead. Signed-off-by: Aneesh Kumar K.V Acked-by: Jan Kara Signed-off-by: "Theodore Ts'o" --- fs/ext4/inode.c | 23 ++++++++--------------- 1 file changed, 8 insertions(+), 15 deletions(-) (limited to 'fs') diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index b87b68cd3241..1275f34589c7 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -2305,15 +2305,9 @@ flush_it: return; } -static int ext4_bh_unmapped_or_delay(handle_t *handle, struct buffer_head *bh) +static int ext4_bh_delay_or_unwritten(handle_t *handle, struct buffer_head *bh) { - /* - * unmapped buffer is possible for holes. - * delay buffer is possible with delayed allocation. - * We also need to consider unwritten buffer as unmapped. - */ - return (!buffer_mapped(bh) || buffer_delay(bh) || - buffer_unwritten(bh)) && buffer_dirty(bh); + return (buffer_delay(bh) || buffer_unwritten(bh)) && buffer_dirty(bh); } /* @@ -2400,7 +2394,7 @@ static int __mpage_da_writepage(struct page *page, * Otherwise we won't make progress * with the page in ext4_da_writepage */ - if (ext4_bh_unmapped_or_delay(NULL, bh)) { + if (ext4_bh_delay_or_unwritten(NULL, bh)) { mpage_add_bh_to_extent(mpd, logical, bh->b_size, bh->b_state); @@ -2517,7 +2511,6 @@ static int noalloc_get_block_write(struct inode *inode, sector_t iblock, * so call get_block_wrap with create = 0 */ ret = ext4_get_blocks(NULL, inode, iblock, max_blocks, bh_result, 0); - BUG_ON(create && ret == 0); if (ret > 0) { bh_result->b_size = (ret << inode->i_blkbits); ret = 0; @@ -2533,7 +2526,7 @@ static int noalloc_get_block_write(struct inode *inode, sector_t iblock, * - grab_page_cache when doing write_begin (have journal handle) */ static int ext4_da_writepage(struct page *page, - struct writeback_control *wbc) + struct writeback_control *wbc) { int ret = 0; loff_t size; @@ -2551,7 +2544,7 @@ static int ext4_da_writepage(struct page *page, if (page_has_buffers(page)) { page_bufs = page_buffers(page); if (walk_page_buffers(NULL, page_bufs, 0, len, NULL, - ext4_bh_unmapped_or_delay)) { + ext4_bh_delay_or_unwritten)) { /* * We don't want to do block allocation * So redirty the page and return @@ -2584,7 +2577,7 @@ static int ext4_da_writepage(struct page *page, page_bufs = page_buffers(page); /* check whether all are mapped and non delay */ if (walk_page_buffers(NULL, page_bufs, 0, len, NULL, - ext4_bh_unmapped_or_delay)) { + ext4_bh_delay_or_unwritten)) { redirty_page_for_writepage(wbc, page); unlock_page(page); return 0; @@ -3232,7 +3225,7 @@ static int ext4_normal_writepage(struct page *page, * happily proceed with mapping them and writing the page. */ BUG_ON(walk_page_buffers(NULL, page_buffers(page), 0, len, NULL, - ext4_bh_unmapped_or_delay)); + ext4_bh_delay_or_unwritten)); } if (!ext4_journal_current_handle()) @@ -3322,7 +3315,7 @@ static int ext4_journalled_writepage(struct page *page, * happily proceed with mapping them and writing the page. */ BUG_ON(walk_page_buffers(NULL, page_buffers(page), 0, len, NULL, - ext4_bh_unmapped_or_delay)); + ext4_bh_delay_or_unwritten)); } if (ext4_journal_current_handle()) -- cgit v1.2.3 From 43ce1d23b43330634507a049b55c36e91d27282e Mon Sep 17 00:00:00 2001 From: "Aneesh Kumar K.V" Date: Sun, 14 Jun 2009 17:58:45 -0400 Subject: ext4: Fix mmap/truncate race when blocksize < pagesize && !nodellaoc This patch fixes the mmap/truncate race that was fixed for delayed allocation by merging ext4_{journalled,normal,da}_writepage() into ext4_writepage(). Signed-off-by: Aneesh Kumar K.V Acked-by: Jan Kara Signed-off-by: "Theodore Ts'o" --- fs/ext4/inode.c | 234 +++++++++++--------------------------------- include/trace/events/ext4.h | 45 +-------- 2 files changed, 58 insertions(+), 221 deletions(-) (limited to 'fs') diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 1275f34589c7..97c48b5b0578 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -47,6 +47,10 @@ #define MPAGE_DA_EXTENT_TAIL 0x01 +static int __ext4_journalled_writepage(struct page *page, + struct writeback_control *wbc, + unsigned int len); + static inline int ext4_begin_ordered_truncate(struct inode *inode, loff_t new_size) { @@ -2392,7 +2396,7 @@ static int __mpage_da_writepage(struct page *page, * We need to try to allocate * unmapped blocks in the same page. * Otherwise we won't make progress - * with the page in ext4_da_writepage + * with the page in ext4_writepage */ if (ext4_bh_delay_or_unwritten(NULL, bh)) { mpage_add_bh_to_extent(mpd, logical, @@ -2519,13 +2523,47 @@ static int noalloc_get_block_write(struct inode *inode, sector_t iblock, } /* + * Note that we don't need to start a transaction unless we're journaling data + * because we should have holes filled from ext4_page_mkwrite(). We even don't + * need to file the inode to the transaction's list in ordered mode because if + * we are writing back data added by write(), the inode is already there and if + * we are writing back data modified via mmap(), noone guarantees in which + * transaction the data will hit the disk. In case we are journaling data, we + * cannot start transaction directly because transaction start ranks above page + * lock so we have to do some magic. + * * This function can get called via... * - ext4_da_writepages after taking page lock (have journal handle) * - journal_submit_inode_data_buffers (no journal handle) * - shrink_page_list via pdflush (no journal handle) * - grab_page_cache when doing write_begin (have journal handle) + * + * We don't do any block allocation in this function. If we have page with + * multiple blocks we need to write those buffer_heads that are mapped. This + * is important for mmaped based write. So if we do with blocksize 1K + * truncate(f, 1024); + * a = mmap(f, 0, 4096); + * a[0] = 'a'; + * truncate(f, 4096); + * we have in the page first buffer_head mapped via page_mkwrite call back + * but other bufer_heads would be unmapped but dirty(dirty done via the + * do_wp_page). So writepage should write the first block. If we modify + * the mmap area beyond 1024 we will again get a page_fault and the + * page_mkwrite callback will do the block allocation and mark the + * buffer_heads mapped. + * + * We redirty the page if we have any buffer_heads that is either delay or + * unwritten in the page. + * + * We can get recursively called as show below. + * + * ext4_writepage() -> kmalloc() -> __alloc_pages() -> page_launder() -> + * ext4_writepage() + * + * But since we don't do any block allocation we should not deadlock. + * Page also have the dirty flag cleared so we don't get recurive page_lock. */ -static int ext4_da_writepage(struct page *page, +static int ext4_writepage(struct page *page, struct writeback_control *wbc) { int ret = 0; @@ -2534,7 +2572,7 @@ static int ext4_da_writepage(struct page *page, struct buffer_head *page_bufs; struct inode *inode = page->mapping->host; - trace_ext4_da_writepage(inode, page); + trace_ext4_writepage(inode, page); size = i_size_read(inode); if (page->index == size >> PAGE_CACHE_SHIFT) len = size & ~PAGE_CACHE_MASK; @@ -2596,6 +2634,15 @@ static int ext4_da_writepage(struct page *page, block_commit_write(page, 0, len); } + if (PageChecked(page) && ext4_should_journal_data(inode)) { + /* + * It's mmapped pagecache. Add buffers and journal it. There + * doesn't seem much point in redirtying the page here. + */ + ClearPageChecked(page); + return __ext4_journalled_writepage(page, wbc, len); + } + if (test_opt(inode->i_sb, NOBH) && ext4_should_writeback_data(inode)) ret = nobh_writepage(page, noalloc_get_block_write, wbc); else @@ -3135,112 +3182,10 @@ static int bput_one(handle_t *handle, struct buffer_head *bh) return 0; } -/* - * Note that we don't need to start a transaction unless we're journaling data - * because we should have holes filled from ext4_page_mkwrite(). We even don't - * need to file the inode to the transaction's list in ordered mode because if - * we are writing back data added by write(), the inode is already there and if - * we are writing back data modified via mmap(), noone guarantees in which - * transaction the data will hit the disk. In case we are journaling data, we - * cannot start transaction directly because transaction start ranks above page - * lock so we have to do some magic. - * - * In all journaling modes block_write_full_page() will start the I/O. - * - * Problem: - * - * ext4_writepage() -> kmalloc() -> __alloc_pages() -> page_launder() -> - * ext4_writepage() - * - * Similar for: - * - * ext4_file_write() -> generic_file_write() -> __alloc_pages() -> ... - * - * Same applies to ext4_get_block(). We will deadlock on various things like - * lock_journal and i_data_sem - * - * Setting PF_MEMALLOC here doesn't work - too many internal memory - * allocations fail. - * - * 16May01: If we're reentered then journal_current_handle() will be - * non-zero. We simply *return*. - * - * 1 July 2001: @@@ FIXME: - * In journalled data mode, a data buffer may be metadata against the - * current transaction. But the same file is part of a shared mapping - * and someone does a writepage() on it. - * - * We will move the buffer onto the async_data list, but *after* it has - * been dirtied. So there's a small window where we have dirty data on - * BJ_Metadata. - * - * Note that this only applies to the last partial page in the file. The - * bit which block_write_full_page() uses prepare/commit for. (That's - * broken code anyway: it's wrong for msync()). - * - * It's a rare case: affects the final partial page, for journalled data - * where the file is subject to bith write() and writepage() in the same - * transction. To fix it we'll need a custom block_write_full_page(). - * We'll probably need that anyway for journalling writepage() output. - * - * We don't honour synchronous mounts for writepage(). That would be - * disastrous. Any write() or metadata operation will sync the fs for - * us. - * - */ -static int __ext4_normal_writepage(struct page *page, - struct writeback_control *wbc) -{ - struct inode *inode = page->mapping->host; - - if (test_opt(inode->i_sb, NOBH)) - return nobh_writepage(page, noalloc_get_block_write, wbc); - else - return block_write_full_page(page, noalloc_get_block_write, - wbc); -} - -static int ext4_normal_writepage(struct page *page, - struct writeback_control *wbc) -{ - struct inode *inode = page->mapping->host; - loff_t size = i_size_read(inode); - loff_t len; - - trace_ext4_normal_writepage(inode, page); - J_ASSERT(PageLocked(page)); - if (page->index == size >> PAGE_CACHE_SHIFT) - len = size & ~PAGE_CACHE_MASK; - else - len = PAGE_CACHE_SIZE; - - if (page_has_buffers(page)) { - /* if page has buffers it should all be mapped - * and allocated. If there are not buffers attached - * to the page we know the page is dirty but it lost - * buffers. That means that at some moment in time - * after write_begin() / write_end() has been called - * all buffers have been clean and thus they must have been - * written at least once. So they are all mapped and we can - * happily proceed with mapping them and writing the page. - */ - BUG_ON(walk_page_buffers(NULL, page_buffers(page), 0, len, NULL, - ext4_bh_delay_or_unwritten)); - } - - if (!ext4_journal_current_handle()) - return __ext4_normal_writepage(page, wbc); - - redirty_page_for_writepage(wbc, page); - unlock_page(page); - return 0; -} - static int __ext4_journalled_writepage(struct page *page, - struct writeback_control *wbc) + struct writeback_control *wbc, + unsigned int len) { - loff_t size; - unsigned int len; struct address_space *mapping = page->mapping; struct inode *inode = mapping->host; struct buffer_head *page_bufs; @@ -3248,16 +3193,8 @@ static int __ext4_journalled_writepage(struct page *page, int ret = 0; int err; - size = i_size_read(inode); - if (page->index == size >> PAGE_CACHE_SHIFT) - len = size & ~PAGE_CACHE_MASK; - else - len = PAGE_CACHE_SIZE; - ret = block_prepare_write(page, 0, len, noalloc_get_block_write); - if (ret != 0) - goto out_unlock; - page_bufs = page_buffers(page); + BUG_ON(!page_bufs); walk_page_buffers(handle, page_bufs, 0, len, NULL, bget_one); /* As soon as we unlock the page, it can go away, but we have * references to buffers so we are safe */ @@ -3282,67 +3219,10 @@ static int __ext4_journalled_writepage(struct page *page, walk_page_buffers(handle, page_bufs, 0, len, NULL, bput_one); EXT4_I(inode)->i_state |= EXT4_STATE_JDATA; - goto out; - -out_unlock: - unlock_page(page); out: return ret; } -static int ext4_journalled_writepage(struct page *page, - struct writeback_control *wbc) -{ - struct inode *inode = page->mapping->host; - loff_t size = i_size_read(inode); - loff_t len; - - trace_ext4_journalled_writepage(inode, page); - J_ASSERT(PageLocked(page)); - if (page->index == size >> PAGE_CACHE_SHIFT) - len = size & ~PAGE_CACHE_MASK; - else - len = PAGE_CACHE_SIZE; - - if (page_has_buffers(page)) { - /* if page has buffers it should all be mapped - * and allocated. If there are not buffers attached - * to the page we know the page is dirty but it lost - * buffers. That means that at some moment in time - * after write_begin() / write_end() has been called - * all buffers have been clean and thus they must have been - * written at least once. So they are all mapped and we can - * happily proceed with mapping them and writing the page. - */ - BUG_ON(walk_page_buffers(NULL, page_buffers(page), 0, len, NULL, - ext4_bh_delay_or_unwritten)); - } - - if (ext4_journal_current_handle()) - goto no_write; - - if (PageChecked(page)) { - /* - * It's mmapped pagecache. Add buffers and journal it. There - * doesn't seem much point in redirtying the page here. - */ - ClearPageChecked(page); - return __ext4_journalled_writepage(page, wbc); - } else { - /* - * It may be a page full of checkpoint-mode buffers. We don't - * really know unless we go poke around in the buffer_heads. - * But block_write_full_page will do the right thing. - */ - return block_write_full_page(page, noalloc_get_block_write, - wbc); - } -no_write: - redirty_page_for_writepage(wbc, page); - unlock_page(page); - return 0; -} - static int ext4_readpage(struct file *file, struct page *page) { return mpage_readpage(page, ext4_get_block); @@ -3489,7 +3369,7 @@ static int ext4_journalled_set_page_dirty(struct page *page) static const struct address_space_operations ext4_ordered_aops = { .readpage = ext4_readpage, .readpages = ext4_readpages, - .writepage = ext4_normal_writepage, + .writepage = ext4_writepage, .sync_page = block_sync_page, .write_begin = ext4_write_begin, .write_end = ext4_ordered_write_end, @@ -3504,7 +3384,7 @@ static const struct address_space_operations ext4_ordered_aops = { static const struct address_space_operations ext4_writeback_aops = { .readpage = ext4_readpage, .readpages = ext4_readpages, - .writepage = ext4_normal_writepage, + .writepage = ext4_writepage, .sync_page = block_sync_page, .write_begin = ext4_write_begin, .write_end = ext4_writeback_write_end, @@ -3519,7 +3399,7 @@ static const struct address_space_operations ext4_writeback_aops = { static const struct address_space_operations ext4_journalled_aops = { .readpage = ext4_readpage, .readpages = ext4_readpages, - .writepage = ext4_journalled_writepage, + .writepage = ext4_writepage, .sync_page = block_sync_page, .write_begin = ext4_write_begin, .write_end = ext4_journalled_write_end, @@ -3533,7 +3413,7 @@ static const struct address_space_operations ext4_journalled_aops = { static const struct address_space_operations ext4_da_aops = { .readpage = ext4_readpage, .readpages = ext4_readpages, - .writepage = ext4_da_writepage, + .writepage = ext4_writepage, .writepages = ext4_da_writepages, .sync_page = block_sync_page, .write_begin = ext4_da_write_begin, diff --git a/include/trace/events/ext4.h b/include/trace/events/ext4.h index b456fb0a3c57..dfbc9b0edc88 100644 --- a/include/trace/events/ext4.h +++ b/include/trace/events/ext4.h @@ -190,7 +190,7 @@ TRACE_EVENT(ext4_journalled_write_end, __entry->copied) ); -TRACE_EVENT(ext4_da_writepage, +TRACE_EVENT(ext4_writepage, TP_PROTO(struct inode *inode, struct page *page), TP_ARGS(inode, page), @@ -342,49 +342,6 @@ TRACE_EVENT(ext4_da_write_end, __entry->copied) ); -TRACE_EVENT(ext4_normal_writepage, - TP_PROTO(struct inode *inode, struct page *page), - - TP_ARGS(inode, page), - - TP_STRUCT__entry( - __field( dev_t, dev ) - __field( ino_t, ino ) - __field( pgoff_t, index ) - ), - - TP_fast_assign( - __entry->dev = inode->i_sb->s_dev; - __entry->ino = inode->i_ino; - __entry->index = page->index; - ), - - TP_printk("dev %s ino %lu page_index %lu", - jbd2_dev_to_name(__entry->dev), __entry->ino, __entry->index) -); - -TRACE_EVENT(ext4_journalled_writepage, - TP_PROTO(struct inode *inode, struct page *page), - - TP_ARGS(inode, page), - - TP_STRUCT__entry( - __field( dev_t, dev ) - __field( ino_t, ino ) - __field( pgoff_t, index ) - - ), - - TP_fast_assign( - __entry->dev = inode->i_sb->s_dev; - __entry->ino = inode->i_ino; - __entry->index = page->index; - ), - - TP_printk("dev %s ino %lu page_index %lu", - jbd2_dev_to_name(__entry->dev), __entry->ino, __entry->index) -); - TRACE_EVENT(ext4_discard_blocks, TP_PROTO(struct super_block *sb, unsigned long long blk, unsigned long long count), -- cgit v1.2.3 From 62e086be5d2abef8cad854bc5707329ad345f2ec Mon Sep 17 00:00:00 2001 From: "Aneesh Kumar K.V" Date: Sun, 14 Jun 2009 17:59:34 -0400 Subject: ext4: Move __ext4_journalled_writepage() to avoid forward declaration In addition, fix two unused variable warnings. Signed-off-by: Aneesh Kumar K.V Signed-off-by: "Theodore Ts'o" --- fs/ext4/inode.c | 112 +++++++++++++++++++++++++++----------------------------- 1 file changed, 54 insertions(+), 58 deletions(-) (limited to 'fs') diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 97c48b5b0578..c98e3afea30a 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -47,10 +47,6 @@ #define MPAGE_DA_EXTENT_TAIL 0x01 -static int __ext4_journalled_writepage(struct page *page, - struct writeback_control *wbc, - unsigned int len); - static inline int ext4_begin_ordered_truncate(struct inode *inode, loff_t new_size) { @@ -2522,6 +2518,59 @@ static int noalloc_get_block_write(struct inode *inode, sector_t iblock, return ret; } +static int bget_one(handle_t *handle, struct buffer_head *bh) +{ + get_bh(bh); + return 0; +} + +static int bput_one(handle_t *handle, struct buffer_head *bh) +{ + put_bh(bh); + return 0; +} + +static int __ext4_journalled_writepage(struct page *page, + struct writeback_control *wbc, + unsigned int len) +{ + struct address_space *mapping = page->mapping; + struct inode *inode = mapping->host; + struct buffer_head *page_bufs; + handle_t *handle = NULL; + int ret = 0; + int err; + + page_bufs = page_buffers(page); + BUG_ON(!page_bufs); + walk_page_buffers(handle, page_bufs, 0, len, NULL, bget_one); + /* As soon as we unlock the page, it can go away, but we have + * references to buffers so we are safe */ + unlock_page(page); + + handle = ext4_journal_start(inode, ext4_writepage_trans_blocks(inode)); + if (IS_ERR(handle)) { + ret = PTR_ERR(handle); + goto out; + } + + ret = walk_page_buffers(handle, page_bufs, 0, len, NULL, + do_journal_get_write_access); + + err = walk_page_buffers(handle, page_bufs, 0, len, NULL, + write_end_fn); + if (ret == 0) + ret = err; + err = ext4_journal_stop(handle); + if (!ret) + ret = err; + + walk_page_buffers(handle, page_bufs, 0, len, NULL, bput_one); + EXT4_I(inode)->i_state |= EXT4_STATE_JDATA; +out: + return ret; +} + /* * Note that we don't need to start a transaction unless we're journaling data * because we should have holes filled from ext4_page_mkwrite(). We even don't @@ -2564,7 +2613,7 @@ static int noalloc_get_block_write(struct inode *inode, sector_t iblock, * Page also have the dirty flag cleared so we don't get recurive page_lock. */ static int ext4_writepage(struct page *page, - struct writeback_control *wbc) + struct writeback_control *wbc) { int ret = 0; loff_t size; @@ -3170,59 +3219,6 @@ static sector_t ext4_bmap(struct address_space *mapping, sector_t block) return generic_block_bmap(mapping, block, ext4_get_block); } -static int bget_one(handle_t *handle, struct buffer_head *bh) -{ - get_bh(bh); - return 0; -} - -static int bput_one(handle_t *handle, struct buffer_head *bh) -{ - put_bh(bh); - return 0; -} - -static int __ext4_journalled_writepage(struct page *page, - struct writeback_control *wbc, - unsigned int len) -{ - struct address_space *mapping = page->mapping; - struct inode *inode = mapping->host; - struct buffer_head *page_bufs; - handle_t *handle = NULL; - int ret = 0; - int err; - - page_bufs = page_buffers(page); - BUG_ON(!page_bufs); - walk_page_buffers(handle, page_bufs, 0, len, NULL, bget_one); - /* As soon as we unlock the page, it can go away, but we have - * references to buffers so we are safe */ - unlock_page(page); - - handle = ext4_journal_start(inode, ext4_writepage_trans_blocks(inode)); - if (IS_ERR(handle)) { - ret = PTR_ERR(handle); - goto out; - } - - ret = walk_page_buffers(handle, page_bufs, 0, len, NULL, - do_journal_get_write_access); - - err = walk_page_buffers(handle, page_bufs, 0, len, NULL, - write_end_fn); - if (ret == 0) - ret = err; - err = ext4_journal_stop(handle); - if (!ret) - ret = err; - - walk_page_buffers(handle, page_bufs, 0, len, NULL, bput_one); - EXT4_I(inode)->i_state |= EXT4_STATE_JDATA; -out: - return ret; -} - static int ext4_readpage(struct file *file, struct page *page) { return mpage_readpage(page, ext4_get_block); -- cgit v1.2.3 From 4159175058987cb68aefd0e9eec2598b795363b4 Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Mon, 15 Jun 2009 03:41:23 -0400 Subject: ext4: Don't update ctime for non-extent-mapped inodes MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The VFS handles updating ctime, so we don't need to update the inode's ctime in ext4_splace_branch() to update the direct or indirect blocks. This was harmless when we did this in ext3, but in ext4, thanks to delayed allocation, updating the ctime in ext4_splice_branch() can cause the ctime to mysteriously jump when the blocks are finally allocated. Thanks to Björn Steinbrink for pointing out this problem on the git mailing list. Signed-off-by: "Theodore Ts'o" --- fs/ext4/inode.c | 6 +----- 1 file changed, 1 insertion(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 8d0908afbd5b..7c17ae275af4 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -856,10 +856,6 @@ static int ext4_splice_branch(handle_t *handle, struct inode *inode, } /* We are done with atomic stuff, now do the rest of housekeeping */ - - inode->i_ctime = ext4_current_time(inode); - ext4_mark_inode_dirty(handle, inode); - /* had we spliced it onto indirect block? */ if (where->bh) { /* @@ -878,8 +874,8 @@ static int ext4_splice_branch(handle_t *handle, struct inode *inode, } else { /* * OK, we spliced it into the inode itself on a direct block. - * Inode was dirtied above. */ + ext4_mark_inode_dirty(handle, inode); jbd_debug(5, "splicing direct\n"); } return err; -- cgit v1.2.3 From a2ab0ce09edf20b5228208405dd14bc8790fbdbd Mon Sep 17 00:00:00 2001 From: Christian Engelmayer Date: Sat, 13 Jun 2009 23:06:29 +0200 Subject: jffs2: leaking jffs2_summary in function jffs2_scan_medium In case of an error returned by file_dirty() 's' is not freed as the cleanup path is skipped. Reported by Coverity. Signed-off-by: Christian Engelmayer Signed-off-by: David Woodhouse --- fs/jffs2/scan.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/jffs2/scan.c b/fs/jffs2/scan.c index 1d437de1e9a8..7515e73e2bfb 100644 --- a/fs/jffs2/scan.c +++ b/fs/jffs2/scan.c @@ -196,7 +196,7 @@ int jffs2_scan_medium(struct jffs2_sb_info *c) if (c->nextblock) { ret = file_dirty(c, c->nextblock); if (ret) - return ret; + goto out; /* deleting summary information of the old nextblock */ jffs2_sum_reset_collected(c->summary); } @@ -207,7 +207,7 @@ int jffs2_scan_medium(struct jffs2_sb_info *c) } else { ret = file_dirty(c, jeb); if (ret) - return ret; + goto out; } break; -- cgit v1.2.3 From 61f98ffd74254a95871168bd5a6646b4f3002e31 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Thu, 11 Jun 2009 10:27:32 -0400 Subject: cifs: display scopeid in /proc/mounts Move address display into a new function and display the scopeid as part of the address in /proc/mounts. Signed-off-by: Jeff Layton Signed-off-by: Steve French --- fs/cifs/cifsfs.c | 33 ++++++++++++++++++++++----------- 1 file changed, 22 insertions(+), 11 deletions(-) (limited to 'fs') diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c index 8b315708cb3f..ddef913ff3e6 100644 --- a/fs/cifs/cifsfs.c +++ b/fs/cifs/cifsfs.c @@ -333,6 +333,27 @@ cifs_destroy_inode(struct inode *inode) kmem_cache_free(cifs_inode_cachep, CIFS_I(inode)); } +static void +cifs_show_address(struct seq_file *s, struct TCP_Server_Info *server) +{ + seq_printf(s, ",addr="); + + switch (server->addr.sockAddr.sin_family) { + case AF_INET: + seq_printf(s, "%pI4", &server->addr.sockAddr.sin_addr.s_addr); + break; + case AF_INET6: + seq_printf(s, "%pI6", + &server->addr.sockAddr6.sin6_addr.s6_addr); + if (server->addr.sockAddr6.sin6_scope_id) + seq_printf(s, "%%%u", + server->addr.sockAddr6.sin6_scope_id); + break; + default: + seq_printf(s, "(unknown)"); + } +} + /* * cifs_show_options() is for displaying mount options in /proc/mounts. * Not all settable options are displayed but most of the important @@ -343,7 +364,6 @@ cifs_show_options(struct seq_file *s, struct vfsmount *m) { struct cifs_sb_info *cifs_sb; struct cifsTconInfo *tcon; - struct TCP_Server_Info *server; cifs_sb = CIFS_SB(m->mnt_sb); tcon = cifs_sb->tcon; @@ -364,16 +384,7 @@ cifs_show_options(struct seq_file *s, struct vfsmount *m) if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_OVERR_GID) seq_printf(s, ",forcegid"); - server = tcon->ses->server; - seq_printf(s, ",addr="); - switch (server->addr.sockAddr6.sin6_family) { - case AF_INET6: - seq_printf(s, "%pI6", &server->addr.sockAddr6.sin6_addr); - break; - case AF_INET: - seq_printf(s, "%pI4", &server->addr.sockAddr.sin_addr.s_addr); - break; - } + cifs_show_address(s, tcon->ses->server); if (!tcon->unix_ext) seq_printf(s, ",file_mode=0%o,dir_mode=0%o", -- cgit v1.2.3 From 361ea1ae5451040cd254eee0b6df64581080b2cc Mon Sep 17 00:00:00 2001 From: Steve French Date: Mon, 15 Jun 2009 13:46:12 +0000 Subject: [CIFS] Fix build break Signed-off-by: Steve French --- fs/cifs/netmisc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/cifs/netmisc.c b/fs/cifs/netmisc.c index 00e6e357ae88..f9a54da97d34 100644 --- a/fs/cifs/netmisc.c +++ b/fs/cifs/netmisc.c @@ -166,7 +166,7 @@ int cifs_convert_address(char *src, void *dst) { struct sockaddr_in *s4 = (struct sockaddr_in *) dst; - struct sockaddr_in6 *s6 = (Struct sockaddr_in6 *) dst; + struct sockaddr_in6 *s6 = (struct sockaddr_in6 *) dst; if (cifs_inet_pton(AF_INET, src, &s4->sin_addr.s_addr)) { s4->sin_family = AF_INET; -- cgit v1.2.3 From df59c0ad05182329688e514e5a9c3836fa208ea3 Mon Sep 17 00:00:00 2001 From: Alexey Zaytsev Date: Wed, 10 Jun 2009 12:56:56 -0700 Subject: [SCSI] compat: don't perform unneeded copy in sg_io code The members from 'status' in struct sg_io_hdr to the last are used to transfer information from kernel to user space. The values that user space sets are just ignored. Signed-off-by: Alexey Zaytsev Acked-by: Jens Axboe Acked-by: FUJITA Tomonori Signed-off-by: Andrew Morton Signed-off-by: James Bottomley --- fs/compat_ioctl.c | 6 ------ 1 file changed, 6 deletions(-) (limited to 'fs') diff --git a/fs/compat_ioctl.c b/fs/compat_ioctl.c index b83f6bcfa51a..905523cc281f 100644 --- a/fs/compat_ioctl.c +++ b/fs/compat_ioctl.c @@ -788,12 +788,6 @@ static int sg_ioctl_trans(unsigned int fd, unsigned int cmd, unsigned long arg) if (put_user(compat_ptr(data), &sgio->usr_ptr)) return -EFAULT; - if (copy_in_user(&sgio->status, &sgio32->status, - (4 * sizeof(unsigned char)) + - (2 * sizeof(unsigned short)) + - (3 * sizeof(int)))) - return -EFAULT; - err = sys_ioctl(fd, cmd, (unsigned long) sgio); if (err >= 0) { -- cgit v1.2.3 From 48e03bc515cff75718de5460458680a230ae997e Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Fri, 5 Jun 2009 12:35:15 -0400 Subject: nfsd: Use write gathering only with NFSv2 NFSv3 and above can use unstable writes whenever they are sending more than one write, rather than relying on the flaky write gathering heuristics. More often than not, write gathering is currently getting it wrong when the NFSv3 clients are sending a single write with FILE_SYNC for efficiency reasons. This patch turns off write gathering for NFSv3/v4, and ensures that it only applies to the one case that can actually benefit: namely NFSv2. Signed-off-by: Trond Myklebust Signed-off-by: J. Bruce Fields --- fs/nfsd/vfs.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c index b660435978d2..f30cc4eadb0a 100644 --- a/fs/nfsd/vfs.c +++ b/fs/nfsd/vfs.c @@ -975,6 +975,7 @@ nfsd_vfs_write(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file *file, __be32 err = 0; int host_err; int stable = *stablep; + int use_wgather; #ifdef MSNFS err = nfserr_perm; @@ -993,9 +994,10 @@ nfsd_vfs_write(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file *file, * - the sync export option has been set, or * - the client requested O_SYNC behavior (NFSv3 feature). * - The file system doesn't support fsync(). - * When gathered writes have been configured for this volume, + * When NFSv2 gathered writes have been configured for this volume, * flushing the data to disk is handled separately below. */ + use_wgather = (rqstp->rq_vers == 2) && EX_WGATHER(exp); if (!file->f_op->fsync) {/* COMMIT3 cannot work */ stable = 2; @@ -1004,7 +1006,7 @@ nfsd_vfs_write(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file *file, if (!EX_ISSYNC(exp)) stable = 0; - if (stable && !EX_WGATHER(exp)) { + if (stable && !use_wgather) { spin_lock(&file->f_lock); file->f_flags |= O_SYNC; spin_unlock(&file->f_lock); @@ -1040,7 +1042,7 @@ nfsd_vfs_write(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file *file, * nice and simple solution (IMHO), and it seems to * work:-) */ - if (EX_WGATHER(exp)) { + if (use_wgather) { if (atomic_read(&inode->i_writecount) > 1 || (last_ino == inode->i_ino && last_dev == inode->i_sb->s_dev)) { dprintk("nfsd: write defer %d\n", task_pid_nr(current)); -- cgit v1.2.3 From 9d2a3f31d6d7832cd441eeda08bc2266cdd5d972 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Mon, 15 Jun 2009 18:47:56 -0700 Subject: nfsd: track last inode only in use_wgather case Updating last_ino and last_dev probably isn't useful in the !use_wgather case. Also remove some pointless ifdef'd-out code. Signed-off-by: J. Bruce Fields --- fs/nfsd/vfs.c | 25 ++++++++++--------------- 1 file changed, 10 insertions(+), 15 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c index f30cc4eadb0a..ebf56c6040ca 100644 --- a/fs/nfsd/vfs.c +++ b/fs/nfsd/vfs.c @@ -1026,7 +1026,7 @@ nfsd_vfs_write(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file *file, if (host_err >= 0 && (inode->i_mode & (S_ISUID | S_ISGID))) kill_suid(dentry); - if (host_err >= 0 && stable) { + if (host_err >= 0 && stable && use_wgather) { static ino_t last_ino; static dev_t last_dev; @@ -1042,21 +1042,16 @@ nfsd_vfs_write(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file *file, * nice and simple solution (IMHO), and it seems to * work:-) */ - if (use_wgather) { - if (atomic_read(&inode->i_writecount) > 1 - || (last_ino == inode->i_ino && last_dev == inode->i_sb->s_dev)) { - dprintk("nfsd: write defer %d\n", task_pid_nr(current)); - msleep(10); - dprintk("nfsd: write resume %d\n", task_pid_nr(current)); - } + if (atomic_read(&inode->i_writecount) > 1 + || (last_ino == inode->i_ino && last_dev == inode->i_sb->s_dev)) { + dprintk("nfsd: write defer %d\n", task_pid_nr(current)); + msleep(10); + dprintk("nfsd: write resume %d\n", task_pid_nr(current)); + } - if (inode->i_state & I_DIRTY) { - dprintk("nfsd: write sync %d\n", task_pid_nr(current)); - host_err=nfsd_sync(file); - } -#if 0 - wake_up(&inode->i_wait); -#endif + if (inode->i_state & I_DIRTY) { + dprintk("nfsd: write sync %d\n", task_pid_nr(current)); + host_err=nfsd_sync(file); } last_ino = inode->i_ino; last_dev = inode->i_sb->s_dev; -- cgit v1.2.3 From d911df7b8d44de41661363a4e29ee710180ba025 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Mon, 15 Jun 2009 16:03:53 -0700 Subject: nfsd: Pull write-gathering code out of nfsd_vfs_write This is a relatively self-contained piece of code that handles a special case--move it to its own function. Signed-off-by: J. Bruce Fields --- fs/nfsd/vfs.c | 69 +++++++++++++++++++++++++++++++++-------------------------- 1 file changed, 39 insertions(+), 30 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c index ebf56c6040ca..6ad76a4cfc01 100644 --- a/fs/nfsd/vfs.c +++ b/fs/nfsd/vfs.c @@ -963,6 +963,43 @@ static void kill_suid(struct dentry *dentry) mutex_unlock(&dentry->d_inode->i_mutex); } +/* + * Gathered writes: If another process is currently writing to the file, + * there's a high chance this is another nfsd (triggered by a bulk write + * from a client's biod). Rather than syncing the file with each write + * request, we sleep for 10 msec. + * + * I don't know if this roughly approximates C. Juszak's idea of + * gathered writes, but it's a nice and simple solution (IMHO), and it + * seems to work:-) + * + * Note: we do this only in the NFSv2 case, since v3 and higher have a + * better tool (separate unstable writes and commits) for solving this + * problem. + */ +static int wait_for_concurrent_writes(struct file *file) +{ + struct inode *inode = file->f_path.dentry->d_inode; + static ino_t last_ino; + static dev_t last_dev; + int err = 0; + + if (atomic_read(&inode->i_writecount) > 1 + || (last_ino == inode->i_ino && last_dev == inode->i_sb->s_dev)) { + dprintk("nfsd: write defer %d\n", task_pid_nr(current)); + msleep(10); + dprintk("nfsd: write resume %d\n", task_pid_nr(current)); + } + + if (inode->i_state & I_DIRTY) { + dprintk("nfsd: write sync %d\n", task_pid_nr(current)); + err = nfsd_sync(file); + } + last_ino = inode->i_ino; + last_dev = inode->i_sb->s_dev; + return err; +} + static __be32 nfsd_vfs_write(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file *file, loff_t offset, struct kvec *vec, int vlen, @@ -1026,36 +1063,8 @@ nfsd_vfs_write(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file *file, if (host_err >= 0 && (inode->i_mode & (S_ISUID | S_ISGID))) kill_suid(dentry); - if (host_err >= 0 && stable && use_wgather) { - static ino_t last_ino; - static dev_t last_dev; - - /* - * Gathered writes: If another process is currently - * writing to the file, there's a high chance - * this is another nfsd (triggered by a bulk write - * from a client's biod). Rather than syncing the - * file with each write request, we sleep for 10 msec. - * - * I don't know if this roughly approximates - * C. Juszak's idea of gathered writes, but it's a - * nice and simple solution (IMHO), and it seems to - * work:-) - */ - if (atomic_read(&inode->i_writecount) > 1 - || (last_ino == inode->i_ino && last_dev == inode->i_sb->s_dev)) { - dprintk("nfsd: write defer %d\n", task_pid_nr(current)); - msleep(10); - dprintk("nfsd: write resume %d\n", task_pid_nr(current)); - } - - if (inode->i_state & I_DIRTY) { - dprintk("nfsd: write sync %d\n", task_pid_nr(current)); - host_err=nfsd_sync(file); - } - last_ino = inode->i_ino; - last_dev = inode->i_sb->s_dev; - } + if (host_err >= 0 && stable && use_wgather) + host_err = wait_for_concurrent_writes(file); dprintk("nfsd: write complete host_err=%d\n", host_err); if (host_err >= 0) -- cgit v1.2.3 From e4636d535e32768c8c500641ddb144f56e3dc5c0 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Mon, 15 Jun 2009 19:07:13 -0700 Subject: nfsd: minor nfsd_vfs_write cleanup There's no need to check host_err >= 0 every time here when we could check host_err < 0 once, following the usual kernel style. Signed-off-by: J. Bruce Fields --- fs/nfsd/vfs.c | 15 ++++++++------- 1 file changed, 8 insertions(+), 7 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c index 6ad76a4cfc01..1cf70616a11e 100644 --- a/fs/nfsd/vfs.c +++ b/fs/nfsd/vfs.c @@ -1053,19 +1053,20 @@ nfsd_vfs_write(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file *file, oldfs = get_fs(); set_fs(KERNEL_DS); host_err = vfs_writev(file, (struct iovec __user *)vec, vlen, &offset); set_fs(oldfs); - if (host_err >= 0) { - *cnt = host_err; - nfsdstats.io_write += host_err; - fsnotify_modify(file->f_path.dentry); - } + if (host_err < 0) + goto out_nfserr; + *cnt = host_err; + nfsdstats.io_write += host_err; + fsnotify_modify(file->f_path.dentry); /* clear setuid/setgid flag after write */ - if (host_err >= 0 && (inode->i_mode & (S_ISUID | S_ISGID))) + if (inode->i_mode & (S_ISUID | S_ISGID)) kill_suid(dentry); - if (host_err >= 0 && stable && use_wgather) + if (stable && use_wgather) host_err = wait_for_concurrent_writes(file); +out_nfserr: dprintk("nfsd: write complete host_err=%d\n", host_err); if (host_err >= 0) err = 0; -- cgit v1.2.3 From b9081d90f5b989cd927052084b16b4f950c8c8d7 Mon Sep 17 00:00:00 2001 From: Yu Zhiguo Date: Tue, 9 Jun 2009 17:33:34 +0800 Subject: NFS: kill off complicated macro 'PROC' kill off obscure macro 'PROC' of NFSv2&3 in order to make the code more clear. Among other things, this makes it simpler to grep for callers of these functions--something which has frequently caused confusion among nfs developers. Signed-off-by: Yu Zhiguo Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs3proc.c | 237 +++++++++++++++++++++++++++++++++++++++++++++++------ fs/nfsd/nfsproc.c | 198 ++++++++++++++++++++++++++++++++++++-------- 2 files changed, 379 insertions(+), 56 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs3proc.c b/fs/nfsd/nfs3proc.c index 7c9fe838f038..a713c418a922 100644 --- a/fs/nfsd/nfs3proc.c +++ b/fs/nfsd/nfs3proc.c @@ -652,8 +652,6 @@ nfsd3_proc_commit(struct svc_rqst * rqstp, struct nfsd3_commitargs *argp, * NFSv3 Server procedures. * Only the results of non-idempotent operations are cached. */ -#define nfs3svc_decode_voidargs NULL -#define nfs3svc_release_void NULL #define nfs3svc_decode_fhandleargs nfs3svc_decode_fhandle #define nfs3svc_encode_attrstatres nfs3svc_encode_attrstat #define nfs3svc_encode_wccstatres nfs3svc_encode_wccstat @@ -686,28 +684,219 @@ struct nfsd3_voidargs { int dummy; }; #define WC (7+pAT) /* WCC attributes */ static struct svc_procedure nfsd_procedures3[22] = { - PROC(null, void, void, void, RC_NOCACHE, ST), - PROC(getattr, fhandle, attrstat, fhandle, RC_NOCACHE, ST+AT), - PROC(setattr, sattr, wccstat, fhandle, RC_REPLBUFF, ST+WC), - PROC(lookup, dirop, dirop, fhandle2, RC_NOCACHE, ST+FH+pAT+pAT), - PROC(access, access, access, fhandle, RC_NOCACHE, ST+pAT+1), - PROC(readlink, readlink, readlink, fhandle, RC_NOCACHE, ST+pAT+1+NFS3_MAXPATHLEN/4), - PROC(read, read, read, fhandle, RC_NOCACHE, ST+pAT+4+NFSSVC_MAXBLKSIZE/4), - PROC(write, write, write, fhandle, RC_REPLBUFF, ST+WC+4), - PROC(create, create, create, fhandle2, RC_REPLBUFF, ST+(1+FH+pAT)+WC), - PROC(mkdir, mkdir, create, fhandle2, RC_REPLBUFF, ST+(1+FH+pAT)+WC), - PROC(symlink, symlink, create, fhandle2, RC_REPLBUFF, ST+(1+FH+pAT)+WC), - PROC(mknod, mknod, create, fhandle2, RC_REPLBUFF, ST+(1+FH+pAT)+WC), - PROC(remove, dirop, wccstat, fhandle, RC_REPLBUFF, ST+WC), - PROC(rmdir, dirop, wccstat, fhandle, RC_REPLBUFF, ST+WC), - PROC(rename, rename, rename, fhandle2, RC_REPLBUFF, ST+WC+WC), - PROC(link, link, link, fhandle2, RC_REPLBUFF, ST+pAT+WC), - PROC(readdir, readdir, readdir, fhandle, RC_NOCACHE, 0), - PROC(readdirplus,readdirplus, readdir, fhandle, RC_NOCACHE, 0), - PROC(fsstat, fhandle, fsstat, void, RC_NOCACHE, ST+pAT+2*6+1), - PROC(fsinfo, fhandle, fsinfo, void, RC_NOCACHE, ST+pAT+12), - PROC(pathconf, fhandle, pathconf, void, RC_NOCACHE, ST+pAT+6), - PROC(commit, commit, commit, fhandle, RC_NOCACHE, ST+WC+2), + [NFS3PROC_NULL] = { + .pc_func = (svc_procfunc) nfsd3_proc_null, + .pc_encode = (kxdrproc_t) nfs3svc_encode_voidres, + .pc_argsize = sizeof(struct nfsd3_voidargs), + .pc_ressize = sizeof(struct nfsd3_voidres), + .pc_cachetype = RC_NOCACHE, + .pc_xdrressize = ST, + }, + [NFS3PROC_GETATTR] = { + .pc_func = (svc_procfunc) nfsd3_proc_getattr, + .pc_decode = (kxdrproc_t) nfs3svc_decode_fhandleargs, + .pc_encode = (kxdrproc_t) nfs3svc_encode_attrstatres, + .pc_release = (kxdrproc_t) nfs3svc_release_fhandle, + .pc_argsize = sizeof(struct nfsd3_fhandleargs), + .pc_ressize = sizeof(struct nfsd3_attrstatres), + .pc_cachetype = RC_NOCACHE, + .pc_xdrressize = ST+AT, + }, + [NFS3PROC_SETATTR] = { + .pc_func = (svc_procfunc) nfsd3_proc_setattr, + .pc_decode = (kxdrproc_t) nfs3svc_decode_sattrargs, + .pc_encode = (kxdrproc_t) nfs3svc_encode_wccstatres, + .pc_release = (kxdrproc_t) nfs3svc_release_fhandle, + .pc_argsize = sizeof(struct nfsd3_sattrargs), + .pc_ressize = sizeof(struct nfsd3_wccstatres), + .pc_cachetype = RC_REPLBUFF, + .pc_xdrressize = ST+WC, + }, + [NFS3PROC_LOOKUP] = { + .pc_func = (svc_procfunc) nfsd3_proc_lookup, + .pc_decode = (kxdrproc_t) nfs3svc_decode_diropargs, + .pc_encode = (kxdrproc_t) nfs3svc_encode_diropres, + .pc_release = (kxdrproc_t) nfs3svc_release_fhandle2, + .pc_argsize = sizeof(struct nfsd3_diropargs), + .pc_ressize = sizeof(struct nfsd3_diropres), + .pc_cachetype = RC_NOCACHE, + .pc_xdrressize = ST+FH+pAT+pAT, + }, + [NFS3PROC_ACCESS] = { + .pc_func = (svc_procfunc) nfsd3_proc_access, + .pc_decode = (kxdrproc_t) nfs3svc_decode_accessargs, + .pc_encode = (kxdrproc_t) nfs3svc_encode_accessres, + .pc_release = (kxdrproc_t) nfs3svc_release_fhandle, + .pc_argsize = sizeof(struct nfsd3_accessargs), + .pc_ressize = sizeof(struct nfsd3_accessres), + .pc_cachetype = RC_NOCACHE, + .pc_xdrressize = ST+pAT+1, + }, + [NFS3PROC_READLINK] = { + .pc_func = (svc_procfunc) nfsd3_proc_readlink, + .pc_decode = (kxdrproc_t) nfs3svc_decode_readlinkargs, + .pc_encode = (kxdrproc_t) nfs3svc_encode_readlinkres, + .pc_release = (kxdrproc_t) nfs3svc_release_fhandle, + .pc_argsize = sizeof(struct nfsd3_readlinkargs), + .pc_ressize = sizeof(struct nfsd3_readlinkres), + .pc_cachetype = RC_NOCACHE, + .pc_xdrressize = ST+pAT+1+NFS3_MAXPATHLEN/4, + }, + [NFS3PROC_READ] = { + .pc_func = (svc_procfunc) nfsd3_proc_read, + .pc_decode = (kxdrproc_t) nfs3svc_decode_readargs, + .pc_encode = (kxdrproc_t) nfs3svc_encode_readres, + .pc_release = (kxdrproc_t) nfs3svc_release_fhandle, + .pc_argsize = sizeof(struct nfsd3_readargs), + .pc_ressize = sizeof(struct nfsd3_readres), + .pc_cachetype = RC_NOCACHE, + .pc_xdrressize = ST+pAT+4+NFSSVC_MAXBLKSIZE/4, + }, + [NFS3PROC_WRITE] = { + .pc_func = (svc_procfunc) nfsd3_proc_write, + .pc_decode = (kxdrproc_t) nfs3svc_decode_writeargs, + .pc_encode = (kxdrproc_t) nfs3svc_encode_writeres, + .pc_release = (kxdrproc_t) nfs3svc_release_fhandle, + .pc_argsize = sizeof(struct nfsd3_writeargs), + .pc_ressize = sizeof(struct nfsd3_writeres), + .pc_cachetype = RC_REPLBUFF, + .pc_xdrressize = ST+WC+4, + }, + [NFS3PROC_CREATE] = { + .pc_func = (svc_procfunc) nfsd3_proc_create, + .pc_decode = (kxdrproc_t) nfs3svc_decode_createargs, + .pc_encode = (kxdrproc_t) nfs3svc_encode_createres, + .pc_release = (kxdrproc_t) nfs3svc_release_fhandle2, + .pc_argsize = sizeof(struct nfsd3_createargs), + .pc_ressize = sizeof(struct nfsd3_createres), + .pc_cachetype = RC_REPLBUFF, + .pc_xdrressize = ST+(1+FH+pAT)+WC, + }, + [NFS3PROC_MKDIR] = { + .pc_func = (svc_procfunc) nfsd3_proc_mkdir, + .pc_decode = (kxdrproc_t) nfs3svc_decode_mkdirargs, + .pc_encode = (kxdrproc_t) nfs3svc_encode_createres, + .pc_release = (kxdrproc_t) nfs3svc_release_fhandle2, + .pc_argsize = sizeof(struct nfsd3_mkdirargs), + .pc_ressize = sizeof(struct nfsd3_createres), + .pc_cachetype = RC_REPLBUFF, + .pc_xdrressize = ST+(1+FH+pAT)+WC, + }, + [NFS3PROC_SYMLINK] = { + .pc_func = (svc_procfunc) nfsd3_proc_symlink, + .pc_decode = (kxdrproc_t) nfs3svc_decode_symlinkargs, + .pc_encode = (kxdrproc_t) nfs3svc_encode_createres, + .pc_release = (kxdrproc_t) nfs3svc_release_fhandle2, + .pc_argsize = sizeof(struct nfsd3_symlinkargs), + .pc_ressize = sizeof(struct nfsd3_createres), + .pc_cachetype = RC_REPLBUFF, + .pc_xdrressize = ST+(1+FH+pAT)+WC, + }, + [NFS3PROC_MKNOD] = { + .pc_func = (svc_procfunc) nfsd3_proc_mknod, + .pc_decode = (kxdrproc_t) nfs3svc_decode_mknodargs, + .pc_encode = (kxdrproc_t) nfs3svc_encode_createres, + .pc_release = (kxdrproc_t) nfs3svc_release_fhandle2, + .pc_argsize = sizeof(struct nfsd3_mknodargs), + .pc_ressize = sizeof(struct nfsd3_createres), + .pc_cachetype = RC_REPLBUFF, + .pc_xdrressize = ST+(1+FH+pAT)+WC, + }, + [NFS3PROC_REMOVE] = { + .pc_func = (svc_procfunc) nfsd3_proc_remove, + .pc_decode = (kxdrproc_t) nfs3svc_decode_diropargs, + .pc_encode = (kxdrproc_t) nfs3svc_encode_wccstatres, + .pc_release = (kxdrproc_t) nfs3svc_release_fhandle, + .pc_argsize = sizeof(struct nfsd3_diropargs), + .pc_ressize = sizeof(struct nfsd3_wccstatres), + .pc_cachetype = RC_REPLBUFF, + .pc_xdrressize = ST+WC, + }, + [NFS3PROC_RMDIR] = { + .pc_func = (svc_procfunc) nfsd3_proc_rmdir, + .pc_decode = (kxdrproc_t) nfs3svc_decode_diropargs, + .pc_encode = (kxdrproc_t) nfs3svc_encode_wccstatres, + .pc_release = (kxdrproc_t) nfs3svc_release_fhandle, + .pc_argsize = sizeof(struct nfsd3_diropargs), + .pc_ressize = sizeof(struct nfsd3_wccstatres), + .pc_cachetype = RC_REPLBUFF, + .pc_xdrressize = ST+WC, + }, + [NFS3PROC_RENAME] = { + .pc_func = (svc_procfunc) nfsd3_proc_rename, + .pc_decode = (kxdrproc_t) nfs3svc_decode_renameargs, + .pc_encode = (kxdrproc_t) nfs3svc_encode_renameres, + .pc_release = (kxdrproc_t) nfs3svc_release_fhandle2, + .pc_argsize = sizeof(struct nfsd3_renameargs), + .pc_ressize = sizeof(struct nfsd3_renameres), + .pc_cachetype = RC_REPLBUFF, + .pc_xdrressize = ST+WC+WC, + }, + [NFS3PROC_LINK] = { + .pc_func = (svc_procfunc) nfsd3_proc_link, + .pc_decode = (kxdrproc_t) nfs3svc_decode_linkargs, + .pc_encode = (kxdrproc_t) nfs3svc_encode_linkres, + .pc_release = (kxdrproc_t) nfs3svc_release_fhandle2, + .pc_argsize = sizeof(struct nfsd3_linkargs), + .pc_ressize = sizeof(struct nfsd3_linkres), + .pc_cachetype = RC_REPLBUFF, + .pc_xdrressize = ST+pAT+WC, + }, + [NFS3PROC_READDIR] = { + .pc_func = (svc_procfunc) nfsd3_proc_readdir, + .pc_decode = (kxdrproc_t) nfs3svc_decode_readdirargs, + .pc_encode = (kxdrproc_t) nfs3svc_encode_readdirres, + .pc_release = (kxdrproc_t) nfs3svc_release_fhandle, + .pc_argsize = sizeof(struct nfsd3_readdirargs), + .pc_ressize = sizeof(struct nfsd3_readdirres), + .pc_cachetype = RC_NOCACHE, + }, + [NFS3PROC_READDIRPLUS] = { + .pc_func = (svc_procfunc) nfsd3_proc_readdirplus, + .pc_decode = (kxdrproc_t) nfs3svc_decode_readdirplusargs, + .pc_encode = (kxdrproc_t) nfs3svc_encode_readdirres, + .pc_release = (kxdrproc_t) nfs3svc_release_fhandle, + .pc_argsize = sizeof(struct nfsd3_readdirplusargs), + .pc_ressize = sizeof(struct nfsd3_readdirres), + .pc_cachetype = RC_NOCACHE, + }, + [NFS3PROC_FSSTAT] = { + .pc_func = (svc_procfunc) nfsd3_proc_fsstat, + .pc_decode = (kxdrproc_t) nfs3svc_decode_fhandleargs, + .pc_encode = (kxdrproc_t) nfs3svc_encode_fsstatres, + .pc_argsize = sizeof(struct nfsd3_fhandleargs), + .pc_ressize = sizeof(struct nfsd3_fsstatres), + .pc_cachetype = RC_NOCACHE, + .pc_xdrressize = ST+pAT+2*6+1, + }, + [NFS3PROC_FSINFO] = { + .pc_func = (svc_procfunc) nfsd3_proc_fsinfo, + .pc_decode = (kxdrproc_t) nfs3svc_decode_fhandleargs, + .pc_encode = (kxdrproc_t) nfs3svc_encode_fsinfores, + .pc_argsize = sizeof(struct nfsd3_fhandleargs), + .pc_ressize = sizeof(struct nfsd3_fsinfores), + .pc_cachetype = RC_NOCACHE, + .pc_xdrressize = ST+pAT+12, + }, + [NFS3PROC_PATHCONF] = { + .pc_func = (svc_procfunc) nfsd3_proc_pathconf, + .pc_decode = (kxdrproc_t) nfs3svc_decode_fhandleargs, + .pc_encode = (kxdrproc_t) nfs3svc_encode_pathconfres, + .pc_argsize = sizeof(struct nfsd3_fhandleargs), + .pc_ressize = sizeof(struct nfsd3_pathconfres), + .pc_cachetype = RC_NOCACHE, + .pc_xdrressize = ST+pAT+6, + }, + [NFS3PROC_COMMIT] = { + .pc_func = (svc_procfunc) nfsd3_proc_commit, + .pc_decode = (kxdrproc_t) nfs3svc_decode_commitargs, + .pc_encode = (kxdrproc_t) nfs3svc_encode_commitres, + .pc_release = (kxdrproc_t) nfs3svc_release_fhandle, + .pc_argsize = sizeof(struct nfsd3_commitargs), + .pc_ressize = sizeof(struct nfsd3_commitres), + .pc_cachetype = RC_NOCACHE, + .pc_xdrressize = ST+WC+2, + }, }; struct svc_version nfsd_version3 = { diff --git a/fs/nfsd/nfsproc.c b/fs/nfsd/nfsproc.c index e298e260b5f1..0eb9c820b7a6 100644 --- a/fs/nfsd/nfsproc.c +++ b/fs/nfsd/nfsproc.c @@ -533,45 +533,179 @@ nfsd_proc_statfs(struct svc_rqst * rqstp, struct nfsd_fhandle *argp, * NFSv2 Server procedures. * Only the results of non-idempotent operations are cached. */ -#define nfsd_proc_none NULL -#define nfssvc_release_none NULL struct nfsd_void { int dummy; }; -#define PROC(name, argt, rest, relt, cache, respsize) \ - { (svc_procfunc) nfsd_proc_##name, \ - (kxdrproc_t) nfssvc_decode_##argt, \ - (kxdrproc_t) nfssvc_encode_##rest, \ - (kxdrproc_t) nfssvc_release_##relt, \ - sizeof(struct nfsd_##argt), \ - sizeof(struct nfsd_##rest), \ - 0, \ - cache, \ - respsize, \ - } - #define ST 1 /* status */ #define FH 8 /* filehandle */ #define AT 18 /* attributes */ static struct svc_procedure nfsd_procedures2[18] = { - PROC(null, void, void, none, RC_NOCACHE, ST), - PROC(getattr, fhandle, attrstat, fhandle, RC_NOCACHE, ST+AT), - PROC(setattr, sattrargs, attrstat, fhandle, RC_REPLBUFF, ST+AT), - PROC(none, void, void, none, RC_NOCACHE, ST), - PROC(lookup, diropargs, diropres, fhandle, RC_NOCACHE, ST+FH+AT), - PROC(readlink, readlinkargs, readlinkres, none, RC_NOCACHE, ST+1+NFS_MAXPATHLEN/4), - PROC(read, readargs, readres, fhandle, RC_NOCACHE, ST+AT+1+NFSSVC_MAXBLKSIZE_V2/4), - PROC(none, void, void, none, RC_NOCACHE, ST), - PROC(write, writeargs, attrstat, fhandle, RC_REPLBUFF, ST+AT), - PROC(create, createargs, diropres, fhandle, RC_REPLBUFF, ST+FH+AT), - PROC(remove, diropargs, void, none, RC_REPLSTAT, ST), - PROC(rename, renameargs, void, none, RC_REPLSTAT, ST), - PROC(link, linkargs, void, none, RC_REPLSTAT, ST), - PROC(symlink, symlinkargs, void, none, RC_REPLSTAT, ST), - PROC(mkdir, createargs, diropres, fhandle, RC_REPLBUFF, ST+FH+AT), - PROC(rmdir, diropargs, void, none, RC_REPLSTAT, ST), - PROC(readdir, readdirargs, readdirres, none, RC_NOCACHE, 0), - PROC(statfs, fhandle, statfsres, none, RC_NOCACHE, ST+5), + [NFSPROC_NULL] = { + .pc_func = (svc_procfunc) nfsd_proc_null, + .pc_decode = (kxdrproc_t) nfssvc_decode_void, + .pc_encode = (kxdrproc_t) nfssvc_encode_void, + .pc_argsize = sizeof(struct nfsd_void), + .pc_ressize = sizeof(struct nfsd_void), + .pc_cachetype = RC_NOCACHE, + .pc_xdrressize = ST, + }, + [NFSPROC_GETATTR] = { + .pc_func = (svc_procfunc) nfsd_proc_getattr, + .pc_decode = (kxdrproc_t) nfssvc_decode_fhandle, + .pc_encode = (kxdrproc_t) nfssvc_encode_attrstat, + .pc_release = (kxdrproc_t) nfssvc_release_fhandle, + .pc_argsize = sizeof(struct nfsd_fhandle), + .pc_ressize = sizeof(struct nfsd_attrstat), + .pc_cachetype = RC_NOCACHE, + .pc_xdrressize = ST+AT, + }, + [NFSPROC_SETATTR] = { + .pc_func = (svc_procfunc) nfsd_proc_setattr, + .pc_decode = (kxdrproc_t) nfssvc_decode_sattrargs, + .pc_encode = (kxdrproc_t) nfssvc_encode_attrstat, + .pc_release = (kxdrproc_t) nfssvc_release_fhandle, + .pc_argsize = sizeof(struct nfsd_sattrargs), + .pc_ressize = sizeof(struct nfsd_attrstat), + .pc_cachetype = RC_REPLBUFF, + .pc_xdrressize = ST+AT, + }, + [NFSPROC_ROOT] = { + .pc_decode = (kxdrproc_t) nfssvc_decode_void, + .pc_encode = (kxdrproc_t) nfssvc_encode_void, + .pc_argsize = sizeof(struct nfsd_void), + .pc_ressize = sizeof(struct nfsd_void), + .pc_cachetype = RC_NOCACHE, + .pc_xdrressize = ST, + }, + [NFSPROC_LOOKUP] = { + .pc_func = (svc_procfunc) nfsd_proc_lookup, + .pc_decode = (kxdrproc_t) nfssvc_decode_diropargs, + .pc_encode = (kxdrproc_t) nfssvc_encode_diropres, + .pc_release = (kxdrproc_t) nfssvc_release_fhandle, + .pc_argsize = sizeof(struct nfsd_diropargs), + .pc_ressize = sizeof(struct nfsd_diropres), + .pc_cachetype = RC_NOCACHE, + .pc_xdrressize = ST+FH+AT, + }, + [NFSPROC_READLINK] = { + .pc_func = (svc_procfunc) nfsd_proc_readlink, + .pc_decode = (kxdrproc_t) nfssvc_decode_readlinkargs, + .pc_encode = (kxdrproc_t) nfssvc_encode_readlinkres, + .pc_argsize = sizeof(struct nfsd_readlinkargs), + .pc_ressize = sizeof(struct nfsd_readlinkres), + .pc_cachetype = RC_NOCACHE, + .pc_xdrressize = ST+1+NFS_MAXPATHLEN/4, + }, + [NFSPROC_READ] = { + .pc_func = (svc_procfunc) nfsd_proc_read, + .pc_decode = (kxdrproc_t) nfssvc_decode_readargs, + .pc_encode = (kxdrproc_t) nfssvc_encode_readres, + .pc_release = (kxdrproc_t) nfssvc_release_fhandle, + .pc_argsize = sizeof(struct nfsd_readargs), + .pc_ressize = sizeof(struct nfsd_readres), + .pc_cachetype = RC_NOCACHE, + .pc_xdrressize = ST+AT+1+NFSSVC_MAXBLKSIZE_V2/4, + }, + [NFSPROC_WRITECACHE] = { + .pc_decode = (kxdrproc_t) nfssvc_decode_void, + .pc_encode = (kxdrproc_t) nfssvc_encode_void, + .pc_argsize = sizeof(struct nfsd_void), + .pc_ressize = sizeof(struct nfsd_void), + .pc_cachetype = RC_NOCACHE, + .pc_xdrressize = ST, + }, + [NFSPROC_WRITE] = { + .pc_func = (svc_procfunc) nfsd_proc_write, + .pc_decode = (kxdrproc_t) nfssvc_decode_writeargs, + .pc_encode = (kxdrproc_t) nfssvc_encode_attrstat, + .pc_release = (kxdrproc_t) nfssvc_release_fhandle, + .pc_argsize = sizeof(struct nfsd_writeargs), + .pc_ressize = sizeof(struct nfsd_attrstat), + .pc_cachetype = RC_REPLBUFF, + .pc_xdrressize = ST+AT, + }, + [NFSPROC_CREATE] = { + .pc_func = (svc_procfunc) nfsd_proc_create, + .pc_decode = (kxdrproc_t) nfssvc_decode_createargs, + .pc_encode = (kxdrproc_t) nfssvc_encode_diropres, + .pc_release = (kxdrproc_t) nfssvc_release_fhandle, + .pc_argsize = sizeof(struct nfsd_createargs), + .pc_ressize = sizeof(struct nfsd_diropres), + .pc_cachetype = RC_REPLBUFF, + .pc_xdrressize = ST+FH+AT, + }, + [NFSPROC_REMOVE] = { + .pc_func = (svc_procfunc) nfsd_proc_remove, + .pc_decode = (kxdrproc_t) nfssvc_decode_diropargs, + .pc_encode = (kxdrproc_t) nfssvc_encode_void, + .pc_argsize = sizeof(struct nfsd_diropargs), + .pc_ressize = sizeof(struct nfsd_void), + .pc_cachetype = RC_REPLSTAT, + .pc_xdrressize = ST, + }, + [NFSPROC_RENAME] = { + .pc_func = (svc_procfunc) nfsd_proc_rename, + .pc_decode = (kxdrproc_t) nfssvc_decode_renameargs, + .pc_encode = (kxdrproc_t) nfssvc_encode_void, + .pc_argsize = sizeof(struct nfsd_renameargs), + .pc_ressize = sizeof(struct nfsd_void), + .pc_cachetype = RC_REPLSTAT, + .pc_xdrressize = ST, + }, + [NFSPROC_LINK] = { + .pc_func = (svc_procfunc) nfsd_proc_link, + .pc_decode = (kxdrproc_t) nfssvc_decode_linkargs, + .pc_encode = (kxdrproc_t) nfssvc_encode_void, + .pc_argsize = sizeof(struct nfsd_linkargs), + .pc_ressize = sizeof(struct nfsd_void), + .pc_cachetype = RC_REPLSTAT, + .pc_xdrressize = ST, + }, + [NFSPROC_SYMLINK] = { + .pc_func = (svc_procfunc) nfsd_proc_symlink, + .pc_decode = (kxdrproc_t) nfssvc_decode_symlinkargs, + .pc_encode = (kxdrproc_t) nfssvc_encode_void, + .pc_argsize = sizeof(struct nfsd_symlinkargs), + .pc_ressize = sizeof(struct nfsd_void), + .pc_cachetype = RC_REPLSTAT, + .pc_xdrressize = ST, + }, + [NFSPROC_MKDIR] = { + .pc_func = (svc_procfunc) nfsd_proc_mkdir, + .pc_decode = (kxdrproc_t) nfssvc_decode_createargs, + .pc_encode = (kxdrproc_t) nfssvc_encode_diropres, + .pc_release = (kxdrproc_t) nfssvc_release_fhandle, + .pc_argsize = sizeof(struct nfsd_createargs), + .pc_ressize = sizeof(struct nfsd_diropres), + .pc_cachetype = RC_REPLBUFF, + .pc_xdrressize = ST+FH+AT, + }, + [NFSPROC_RMDIR] = { + .pc_func = (svc_procfunc) nfsd_proc_rmdir, + .pc_decode = (kxdrproc_t) nfssvc_decode_diropargs, + .pc_encode = (kxdrproc_t) nfssvc_encode_void, + .pc_argsize = sizeof(struct nfsd_diropargs), + .pc_ressize = sizeof(struct nfsd_void), + .pc_cachetype = RC_REPLSTAT, + .pc_xdrressize = ST, + }, + [NFSPROC_READDIR] = { + .pc_func = (svc_procfunc) nfsd_proc_readdir, + .pc_decode = (kxdrproc_t) nfssvc_decode_readdirargs, + .pc_encode = (kxdrproc_t) nfssvc_encode_readdirres, + .pc_argsize = sizeof(struct nfsd_readdirargs), + .pc_ressize = sizeof(struct nfsd_readdirres), + .pc_cachetype = RC_NOCACHE, + }, + [NFSPROC_STATFS] = { + .pc_func = (svc_procfunc) nfsd_proc_statfs, + .pc_decode = (kxdrproc_t) nfssvc_decode_fhandle, + .pc_encode = (kxdrproc_t) nfssvc_encode_statfsres, + .pc_argsize = sizeof(struct nfsd_fhandle), + .pc_ressize = sizeof(struct nfsd_statfsres), + .pc_cachetype = RC_NOCACHE, + .pc_xdrressize = ST+5, + }, }; -- cgit v1.2.3 From 6c18ba9f5e506b8115b89b1aa7bdc25178f40b0a Mon Sep 17 00:00:00 2001 From: Alexandros Batsakis Date: Tue, 16 Jun 2009 04:19:13 +0300 Subject: nfsd41: move channel attributes from nfsd4_session to a nfsd4_channel_attr struct the change is valid for both the forechannel and the backchannel (currently dummy) Signed-off-by: Alexandros Batsakis Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 28 +++++++++++++++------------- fs/nfsd/nfs4xdr.c | 2 +- include/linux/nfsd/state.h | 18 +++++++++++++----- include/linux/nfsd/xdr4.h | 11 ----------- 4 files changed, 29 insertions(+), 30 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 89d9ac55c034..d5caf2a709d2 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -444,8 +444,8 @@ static int set_forechannel_maxreqs(struct nfsd4_channel_attrs *fchan) * fchan holds the client values on input, and the server values on output */ static int init_forechannel_attrs(struct svc_rqst *rqstp, - struct nfsd4_session *session, - struct nfsd4_channel_attrs *fchan) + struct nfsd4_channel_attrs *session_fchan, + struct nfsd4_channel_attrs *fchan) { int status = 0; __u32 maxcount = svc_max_payload(rqstp); @@ -455,21 +455,21 @@ static int init_forechannel_attrs(struct svc_rqst *rqstp, /* Use the client's max request and max response size if possible */ if (fchan->maxreq_sz > maxcount) fchan->maxreq_sz = maxcount; - session->se_fmaxreq_sz = fchan->maxreq_sz; + session_fchan->maxreq_sz = fchan->maxreq_sz; if (fchan->maxresp_sz > maxcount) fchan->maxresp_sz = maxcount; - session->se_fmaxresp_sz = fchan->maxresp_sz; + session_fchan->maxresp_sz = fchan->maxresp_sz; /* Set the max response cached size our default which is * a multiple of PAGE_SIZE and small */ - session->se_fmaxresp_cached = NFSD_PAGES_PER_SLOT * PAGE_SIZE; - fchan->maxresp_cached = session->se_fmaxresp_cached; + session_fchan->maxresp_cached = NFSD_PAGES_PER_SLOT * PAGE_SIZE; + fchan->maxresp_cached = session_fchan->maxresp_cached; /* Use the client's maxops if possible */ if (fchan->maxops > NFSD_MAX_OPS_PER_COMPOUND) fchan->maxops = NFSD_MAX_OPS_PER_COMPOUND; - session->se_fmaxops = fchan->maxops; + session_fchan->maxops = fchan->maxops; /* try to use the client requested number of slots */ if (fchan->maxreqs > NFSD_MAX_SLOTS_PER_SESSION) @@ -481,7 +481,7 @@ static int init_forechannel_attrs(struct svc_rqst *rqstp, */ status = set_forechannel_maxreqs(fchan); - session->se_fnumslots = fchan->maxreqs; + session_fchan->maxreqs = fchan->maxreqs; return status; } @@ -495,12 +495,14 @@ alloc_init_session(struct svc_rqst *rqstp, struct nfs4_client *clp, memset(&tmp, 0, sizeof(tmp)); /* FIXME: For now, we just accept the client back channel attributes. */ - status = init_forechannel_attrs(rqstp, &tmp, &cses->fore_channel); + tmp.se_bchannel = cses->back_channel; + status = init_forechannel_attrs(rqstp, &tmp.se_fchannel, + &cses->fore_channel); if (status) goto out; /* allocate struct nfsd4_session and slot table in one piece */ - slotsize = tmp.se_fnumslots * sizeof(struct nfsd4_slot); + slotsize = tmp.se_fchannel.maxreqs * sizeof(struct nfsd4_slot); new = kzalloc(sizeof(*new) + slotsize, GFP_KERNEL); if (!new) goto out; @@ -574,7 +576,7 @@ free_session(struct kref *kref) int i; ses = container_of(kref, struct nfsd4_session, se_ref); - for (i = 0; i < ses->se_fnumslots; i++) { + for (i = 0; i < ses->se_fchannel.maxreqs; i++) { struct nfsd4_cache_entry *e = &ses->se_slots[i].sl_cache_entry; nfsd4_release_respages(e->ce_respages, e->ce_resused); } @@ -1130,7 +1132,7 @@ nfsd4_replay_cache_entry(struct nfsd4_compoundres *resp, * is sent (lease renewal). */ if (seq && nfsd4_not_cached(resp)) { - seq->maxslots = resp->cstate.session->se_fnumslots; + seq->maxslots = resp->cstate.session->se_fchannel.maxreqs; return nfs_ok; } @@ -1473,7 +1475,7 @@ nfsd4_sequence(struct svc_rqst *rqstp, goto out; status = nfserr_badslot; - if (seq->slotid >= session->se_fnumslots) + if (seq->slotid >= session->se_fchannel.maxreqs) goto out; slot = &session->se_slots[seq->slotid]; diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c index d07f704a2ac9..2dcc7feaa6ff 100644 --- a/fs/nfsd/nfs4xdr.c +++ b/fs/nfsd/nfs4xdr.c @@ -3183,7 +3183,7 @@ static int nfsd4_check_drc_limit(struct nfsd4_compoundres *resp) dprintk("%s length %u, xb->page_len %u tlen %u pad %u\n", __func__, length, xb->page_len, tlen, pad); - if (length <= session->se_fmaxresp_cached) + if (length <= session->se_fchannel.maxresp_cached) return status; else return nfserr_rep_too_big_to_cache; diff --git a/include/linux/nfsd/state.h b/include/linux/nfsd/state.h index c0c49215ddc5..105cc100de05 100644 --- a/include/linux/nfsd/state.h +++ b/include/linux/nfsd/state.h @@ -115,6 +115,17 @@ struct nfsd4_slot { struct nfsd4_cache_entry sl_cache_entry; }; +struct nfsd4_channel_attrs { + u32 headerpadsz; + u32 maxreq_sz; + u32 maxresp_sz; + u32 maxresp_cached; + u32 maxops; + u32 maxreqs; + u32 nr_rdma_attrs; + u32 rdma_attrs; +}; + struct nfsd4_session { struct kref se_ref; struct list_head se_hash; /* hash by sessionid */ @@ -122,11 +133,8 @@ struct nfsd4_session { u32 se_flags; struct nfs4_client *se_client; /* for expire_client */ struct nfs4_sessionid se_sessionid; - u32 se_fmaxreq_sz; - u32 se_fmaxresp_sz; - u32 se_fmaxresp_cached; - u32 se_fmaxops; - u32 se_fnumslots; + struct nfsd4_channel_attrs se_fchannel; + struct nfsd4_channel_attrs se_bchannel; struct nfsd4_slot se_slots[]; /* forward channel slots */ }; diff --git a/include/linux/nfsd/xdr4.h b/include/linux/nfsd/xdr4.h index d0f050f01eca..2bacf7535069 100644 --- a/include/linux/nfsd/xdr4.h +++ b/include/linux/nfsd/xdr4.h @@ -366,17 +366,6 @@ struct nfsd4_exchange_id { int spa_how; }; -struct nfsd4_channel_attrs { - u32 headerpadsz; - u32 maxreq_sz; - u32 maxresp_sz; - u32 maxresp_cached; - u32 maxops; - u32 maxreqs; - u32 nr_rdma_attrs; - u32 rdma_attrs; -}; - struct nfsd4_create_session { clientid_t clientid; struct nfs4_sessionid sessionid; -- cgit v1.2.3 From 5d77ddfbcb062f2617ea79d7a371b4bc78f28417 Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Tue, 16 Jun 2009 04:19:38 +0300 Subject: nfsd41: sanity check client drc maxreqs Ensure the client requested maximum requests are between 1 and NFSD_MAX_SLOTS_PER_SESSION Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index d5caf2a709d2..99570c49add5 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -425,6 +425,11 @@ static int set_forechannel_maxreqs(struct nfsd4_channel_attrs *fchan) { int status = 0, np = fchan->maxreqs * NFSD_PAGES_PER_SLOT; + if (fchan->maxreqs < 1) + return nfserr_inval; + else if (fchan->maxreqs > NFSD_MAX_SLOTS_PER_SESSION) + fchan->maxreqs = NFSD_MAX_SLOTS_PER_SESSION; + spin_lock(&nfsd_serv->sv_lock); if (np + nfsd_serv->sv_drc_pages_used > nfsd_serv->sv_drc_max_pages) np = nfsd_serv->sv_drc_max_pages - nfsd_serv->sv_drc_pages_used; -- cgit v1.2.3 From 9c64daff9d5afb102dfe64a26829e26725538e58 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Mon, 8 Jun 2009 15:22:24 -0400 Subject: ext3: avoid unnecessary spinlock in critical POSIX ACL path If a filesystem supports POSIX ACL's, the VFS layer expects the filesystem to do POSIX ACL checks on any files not owned by the caller, and it does this for every single pathname component that it looks up. That obviously can be pretty expensive if the filesystem isn't careful about it, especially with locking. That's doubly sad, since the common case tends to be that there are no ACL's associated with the files in question. ext3 already caches the ACL data so that it doesn't have to look it up over and over again, but it does so by taking the inode->i_lock spinlock on every lookup. Which is a noticeable overhead even if it's a private lock, especially on CPU's where the serialization is expensive (eg Intel Netburst aka 'P4'). For the special case of not actually having any ACL's, all that locking is unnecessary. Even if somebody else were to be changing the ACL's on another CPU, we simply don't care - if we've seen a NULL ACL, we might as well use it. So just load the ACL speculatively without any locking, and if it was NULL, just use it. If it's non-NULL (either because we had a cached entry, or because the cache hasn't been filled in at all), it means that we'll need to get the lock and re-load it properly. This is noticeable even on Nehalem, which does locking quite well (much better than P4). From lmbench: Processor, Processes - times in microseconds - smaller is better -------------------------------------------------------------------- Host OS Mhz null null open slct fork exec sh call I/O stat clos TCP proc proc proc --------- ------------- ---- ---- ---- ---- ---- ---- ---- ---- ---- - before: nehalem.l Linux 2.6.30- 3193 0.04 0.09 0.95 1.45 2.18 69.1 273. 1141 nehalem.l Linux 2.6.30- 3193 0.04 0.09 0.95 1.48 2.28 69.9 253. 1140 nehalem.l Linux 2.6.30- 3193 0.04 0.10 0.95 1.42 2.19 68.6 284. 1141 - after: nehalem.l Linux 2.6.30- 3193 0.04 0.09 0.92 1.44 2.12 68.3 282. 1094 nehalem.l Linux 2.6.30- 3193 0.04 0.09 0.92 1.39 2.20 67.0 308. 1123 nehalem.l Linux 2.6.30- 3193 0.04 0.09 0.92 1.39 2.36 67.4 293. 1148 where you can see what appears to be a roughly 3% improvement in stat and open/close latencies from just the removal of the locking overhead. Of course, this only matters for files you don't own (the owner never needs to do the ACL checks), but that's the common case for libraries, header files, and executables. As well as for the base components of any absolute pathname, even if you are the owner of the final file. [ At some point we probably want to move this ACL caching logic entirely into the VFS layer (and only call down to the filesystem when uncached), but in the meantime this improves ext3 a bit. A similar fix to btrfs makes a much bigger difference (15x improvement in lmbench) due to broken caching. ] Signed-off-by: Linus Torvalds Signed-off-by: "Theodore Ts'o" Acked-by: Jan Kara Cc: Al Viro Signed-off-by: Al Viro --- fs/ext3/acl.c | 13 ++++++++----- 1 file changed, 8 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/ext3/acl.c b/fs/ext3/acl.c index d81ef2fdb08e..e0c745451715 100644 --- a/fs/ext3/acl.c +++ b/fs/ext3/acl.c @@ -129,12 +129,15 @@ fail: static inline struct posix_acl * ext3_iget_acl(struct inode *inode, struct posix_acl **i_acl) { - struct posix_acl *acl = EXT3_ACL_NOT_CACHED; + struct posix_acl *acl = ACCESS_ONCE(*i_acl); - spin_lock(&inode->i_lock); - if (*i_acl != EXT3_ACL_NOT_CACHED) - acl = posix_acl_dup(*i_acl); - spin_unlock(&inode->i_lock); + if (acl) { + spin_lock(&inode->i_lock); + acl = *i_acl; + if (acl != EXT3_ACL_NOT_CACHED) + acl = posix_acl_dup(acl); + spin_unlock(&inode->i_lock); + } return acl; } -- cgit v1.2.3 From 210ad6aedb332e73167ece5af9bd47f0da8c2aca Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Mon, 8 Jun 2009 15:22:25 -0400 Subject: ext4: avoid unnecessary spinlock in critical POSIX ACL path If a filesystem supports POSIX ACL's, the VFS layer expects the filesystem to do POSIX ACL checks on any files not owned by the caller, and it does this for every single pathname component that it looks up. That obviously can be pretty expensive if the filesystem isn't careful about it, especially with locking. That's doubly sad, since the common case tends to be that there are no ACL's associated with the files in question. ext4 already caches the ACL data so that it doesn't have to look it up over and over again, but it does so by taking the inode->i_lock spinlock on every lookup. Which is a noticeable overhead even if it's a private lock, especially on CPU's where the serialization is expensive (eg Intel Netburst aka 'P4'). For the special case of not actually having any ACL's, all that locking is unnecessary. Even if somebody else were to be changing the ACL's on another CPU, we simply don't care - if we've seen a NULL ACL, we might as well use it. So just load the ACL speculatively without any locking, and if it was NULL, just use it. If it's non-NULL (either because we had a cached entry, or because the cache hasn't been filled in at all), it means that we'll need to get the lock and re-load it properly. (This commit was ported from a patch originally authored by Linus for ext3.) Signed-off-by: "Theodore Ts'o" Signed-off-by: Al Viro --- fs/ext4/acl.c | 13 ++++++++----- 1 file changed, 8 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/ext4/acl.c b/fs/ext4/acl.c index 647e0d65a284..605aeed96d68 100644 --- a/fs/ext4/acl.c +++ b/fs/ext4/acl.c @@ -129,12 +129,15 @@ fail: static inline struct posix_acl * ext4_iget_acl(struct inode *inode, struct posix_acl **i_acl) { - struct posix_acl *acl = EXT4_ACL_NOT_CACHED; + struct posix_acl *acl = ACCESS_ONCE(*i_acl); - spin_lock(&inode->i_lock); - if (*i_acl != EXT4_ACL_NOT_CACHED) - acl = posix_acl_dup(*i_acl); - spin_unlock(&inode->i_lock); + if (acl) { + spin_lock(&inode->i_lock); + acl = *i_acl; + if (acl != EXT4_ACL_NOT_CACHED) + acl = posix_acl_dup(acl); + spin_unlock(&inode->i_lock); + } return acl; } -- cgit v1.2.3 From b0895513f499b8f786d292ce48589ca210ca1d6e Mon Sep 17 00:00:00 2001 From: "J. R. Okajima" Date: Wed, 17 Jun 2009 01:16:50 +0900 Subject: remove unlock_kernel() left accidentally commit 337eb00a2c3a421999c39c94ce7e33545ee8baa7 Push BKL down into ->remount_fs() and commit 4aa98cf768b6f2ea4b204620d949a665959214f6 Push BKL down into do_remount_sb() were uncorrectly merged. The former removes one pair of lock/unlock_kernel(), but the latter adds several unlock_kernel(). Finally a few unlock_kernel() calls left. Signed-off-by: J. R. Okajima Signed-off-by: Al Viro --- fs/super.c | 12 +++--------- 1 file changed, 3 insertions(+), 9 deletions(-) (limited to 'fs') diff --git a/fs/super.c b/fs/super.c index 83b47416d006..d40d53a22fb5 100644 --- a/fs/super.c +++ b/fs/super.c @@ -545,24 +545,18 @@ int do_remount_sb(struct super_block *sb, int flags, void *data, int force) if ((flags & MS_RDONLY) && !(sb->s_flags & MS_RDONLY)) { if (force) mark_files_ro(sb); - else if (!fs_may_remount_ro(sb)) { - unlock_kernel(); + else if (!fs_may_remount_ro(sb)) return -EBUSY; - } retval = vfs_dq_off(sb, 1); - if (retval < 0 && retval != -ENOSYS) { - unlock_kernel(); + if (retval < 0 && retval != -ENOSYS) return -EBUSY; - } } remount_rw = !(flags & MS_RDONLY) && (sb->s_flags & MS_RDONLY); if (sb->s_op->remount_fs) { retval = sb->s_op->remount_fs(sb, &flags, data); - if (retval) { - unlock_kernel(); + if (retval) return retval; - } } sb->s_flags = (sb->s_flags & ~MS_RMT_MASK) | (flags & MS_RMT_MASK); if (remount_rw) -- cgit v1.2.3 From fe36adf47eb1f7f4972559efa30ce3d2d3f977f2 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Tue, 16 Jun 2009 13:35:01 -0400 Subject: No instance of ->bmap() needs BKL Signed-off-by: Al Viro --- Documentation/filesystems/Locking | 2 +- fs/ioctl.c | 2 -- 2 files changed, 1 insertion(+), 3 deletions(-) (limited to 'fs') diff --git a/Documentation/filesystems/Locking b/Documentation/filesystems/Locking index 3120f8dd2c31..229d7b7c50a3 100644 --- a/Documentation/filesystems/Locking +++ b/Documentation/filesystems/Locking @@ -187,7 +187,7 @@ readpages: no write_begin: no locks the page yes write_end: no yes, unlocks yes perform_write: no n/a yes -bmap: yes +bmap: no invalidatepage: no yes releasepage: no yes direct_IO: no diff --git a/fs/ioctl.c b/fs/ioctl.c index 286f38dfc6c0..001f8d3118f2 100644 --- a/fs/ioctl.c +++ b/fs/ioctl.c @@ -70,9 +70,7 @@ static int ioctl_fibmap(struct file *filp, int __user *p) res = get_user(block, p); if (res) return res; - lock_kernel(); res = mapping->a_ops->bmap(mapping, block); - unlock_kernel(); return put_user(res, p); } -- cgit v1.2.3 From 66c6af2e8ba55d4d6691c136b42f2423ab9598ec Mon Sep 17 00:00:00 2001 From: Al Viro Date: Tue, 16 Jun 2009 14:15:00 -0400 Subject: fuse doesn't need BKL in ->umount_begin() Signed-off-by: Al Viro --- fs/fuse/inode.c | 3 --- 1 file changed, 3 deletions(-) (limited to 'fs') diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c index f0df55a52929..d8673ccf90b7 100644 --- a/fs/fuse/inode.c +++ b/fs/fuse/inode.c @@ -19,7 +19,6 @@ #include #include #include -#include MODULE_AUTHOR("Miklos Szeredi "); MODULE_DESCRIPTION("Filesystem in Userspace"); @@ -260,9 +259,7 @@ struct inode *fuse_iget(struct super_block *sb, u64 nodeid, static void fuse_umount_begin(struct super_block *sb) { - lock_kernel(); fuse_abort_conn(get_fuse_conn_super(sb)); - unlock_kernel(); } static void fuse_send_destroy(struct fuse_conn *fc) -- cgit v1.2.3 From ee450f796f6c4f3a563c914cb93ccfa91a1f7580 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Tue, 16 Jun 2009 14:17:21 -0400 Subject: 9P doesn't need BKL in ->umount_begin() Signed-off-by: Al Viro --- fs/9p/vfs_super.c | 3 --- 1 file changed, 3 deletions(-) (limited to 'fs') diff --git a/fs/9p/vfs_super.c b/fs/9p/vfs_super.c index ab5547ff29a1..38d695d66a0b 100644 --- a/fs/9p/vfs_super.c +++ b/fs/9p/vfs_super.c @@ -37,7 +37,6 @@ #include #include #include -#include #include #include @@ -231,10 +230,8 @@ v9fs_umount_begin(struct super_block *sb) { struct v9fs_session_info *v9ses; - lock_kernel(); v9ses = sb->s_fs_info; v9fs_session_cancel(v9ses); - unlock_kernel(); } static const struct super_operations v9fs_super_ops = { -- cgit v1.2.3 From 608ba50bd0225d95469154feba8f00a6457848c1 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Tue, 16 Jun 2009 14:52:13 -0400 Subject: Cleanup of adfs headers Signed-off-by: Al Viro --- fs/adfs/adfs.h | 55 ++++++++++++++++++++++++++++++++++++++++++++++ fs/adfs/dir.c | 8 ------- fs/adfs/dir_f.c | 8 ------- fs/adfs/dir_fplus.c | 8 ------- fs/adfs/file.c | 4 ---- fs/adfs/inode.c | 10 --------- fs/adfs/map.c | 6 ----- fs/adfs/super.c | 17 ++------------ include/linux/adfs_fs.h | 13 ----------- include/linux/adfs_fs_i.h | 24 -------------------- include/linux/adfs_fs_sb.h | 38 -------------------------------- 11 files changed, 57 insertions(+), 134 deletions(-) delete mode 100644 include/linux/adfs_fs_i.h delete mode 100644 include/linux/adfs_fs_sb.h (limited to 'fs') diff --git a/fs/adfs/adfs.h b/fs/adfs/adfs.h index a6665f37f456..9cc18775b832 100644 --- a/fs/adfs/adfs.h +++ b/fs/adfs/adfs.h @@ -1,3 +1,6 @@ +#include +#include + /* Internal data structures for ADFS */ #define ADFS_FREE_FRAG 0 @@ -16,6 +19,58 @@ struct buffer_head; +/* + * adfs file system inode data in memory + */ +struct adfs_inode_info { + loff_t mmu_private; + unsigned long parent_id; /* object id of parent */ + __u32 loadaddr; /* RISC OS load address */ + __u32 execaddr; /* RISC OS exec address */ + unsigned int filetype; /* RISC OS file type */ + unsigned int attr; /* RISC OS permissions */ + unsigned int stamped:1; /* RISC OS file has date/time */ + struct inode vfs_inode; +}; + +/* + * Forward-declare this + */ +struct adfs_discmap; +struct adfs_dir_ops; + +/* + * ADFS file system superblock data in memory + */ +struct adfs_sb_info { + struct adfs_discmap *s_map; /* bh list containing map */ + struct adfs_dir_ops *s_dir; /* directory operations */ + + uid_t s_uid; /* owner uid */ + gid_t s_gid; /* owner gid */ + umode_t s_owner_mask; /* ADFS owner perm -> unix perm */ + umode_t s_other_mask; /* ADFS other perm -> unix perm */ + + __u32 s_ids_per_zone; /* max. no ids in one zone */ + __u32 s_idlen; /* length of ID in map */ + __u32 s_map_size; /* sector size of a map */ + unsigned long s_size; /* total size (in blocks) of this fs */ + signed int s_map2blk; /* shift left by this for map->sector */ + unsigned int s_log2sharesize;/* log2 share size */ + __le32 s_version; /* disc format version */ + unsigned int s_namelen; /* maximum number of characters in name */ +}; + +static inline struct adfs_sb_info *ADFS_SB(struct super_block *sb) +{ + return sb->s_fs_info; +} + +static inline struct adfs_inode_info *ADFS_I(struct inode *inode) +{ + return container_of(inode, struct adfs_inode_info, vfs_inode); +} + /* * Directory handling */ diff --git a/fs/adfs/dir.c b/fs/adfs/dir.c index 4d4073447d1a..23aa52f548a0 100644 --- a/fs/adfs/dir.c +++ b/fs/adfs/dir.c @@ -9,15 +9,7 @@ * * Common directory handling for ADFS */ -#include -#include -#include -#include -#include -#include #include -#include /* for file_fsync() */ - #include "adfs.h" /* diff --git a/fs/adfs/dir_f.c b/fs/adfs/dir_f.c index 31df6adf0de6..bafc71222e25 100644 --- a/fs/adfs/dir_f.c +++ b/fs/adfs/dir_f.c @@ -9,15 +9,7 @@ * * E and F format directory handling */ -#include -#include -#include -#include -#include -#include #include -#include - #include "adfs.h" #include "dir_f.h" diff --git a/fs/adfs/dir_fplus.c b/fs/adfs/dir_fplus.c index 139e0f345f18..1796bb352d05 100644 --- a/fs/adfs/dir_fplus.c +++ b/fs/adfs/dir_fplus.c @@ -7,15 +7,7 @@ * it under the terms of the GNU General Public License version 2 as * published by the Free Software Foundation. */ -#include -#include -#include -#include -#include -#include #include -#include - #include "adfs.h" #include "dir_fplus.h" diff --git a/fs/adfs/file.c b/fs/adfs/file.c index 8224d54a2afb..005ea34d1758 100644 --- a/fs/adfs/file.c +++ b/fs/adfs/file.c @@ -19,10 +19,6 @@ * * adfs regular file handling primitives */ -#include -#include /* for file_fsync() */ -#include - #include "adfs.h" const struct file_operations adfs_file_operations = { diff --git a/fs/adfs/inode.c b/fs/adfs/inode.c index 05b3a677201d..798cb071d132 100644 --- a/fs/adfs/inode.c +++ b/fs/adfs/inode.c @@ -7,17 +7,8 @@ * it under the terms of the GNU General Public License version 2 as * published by the Free Software Foundation. */ -#include -#include -#include -#include -#include -#include -#include #include -#include #include - #include "adfs.h" /* @@ -395,4 +386,3 @@ int adfs_write_inode(struct inode *inode, int wait) unlock_kernel(); return ret; } -MODULE_LICENSE("GPL"); diff --git a/fs/adfs/map.c b/fs/adfs/map.c index 568081b93f73..d1a5932bb0f1 100644 --- a/fs/adfs/map.c +++ b/fs/adfs/map.c @@ -7,14 +7,8 @@ * it under the terms of the GNU General Public License version 2 as * published by the Free Software Foundation. */ -#include -#include -#include -#include #include - #include - #include "adfs.h" /* diff --git a/fs/adfs/super.c b/fs/adfs/super.c index 0ec5aaf47aa7..aad92f0a1048 100644 --- a/fs/adfs/super.c +++ b/fs/adfs/super.c @@ -8,26 +8,12 @@ * published by the Free Software Foundation. */ #include -#include -#include -#include -#include -#include -#include -#include #include #include -#include #include -#include #include #include - -#include -#include - -#include - +#include #include "adfs.h" #include "dir_f.h" #include "dir_fplus.h" @@ -534,3 +520,4 @@ static void __exit exit_adfs_fs(void) module_init(init_adfs_fs) module_exit(exit_adfs_fs) +MODULE_LICENSE("GPL"); diff --git a/include/linux/adfs_fs.h b/include/linux/adfs_fs.h index ef788c2085a1..b19801f73890 100644 --- a/include/linux/adfs_fs.h +++ b/include/linux/adfs_fs.h @@ -41,8 +41,6 @@ struct adfs_discrecord { #define ADFS_DR_SIZE_BITS (ADFS_DR_SIZE << 3) #ifdef __KERNEL__ -#include -#include /* * Calculate the boot block checksum on an ADFS drive. Note that this will * appear to be correct if the sector contains all zeros, so also check that @@ -60,17 +58,6 @@ static inline int adfs_checkbblk(unsigned char *ptr) return (result & 0xff) != ptr[511]; } - -static inline struct adfs_sb_info *ADFS_SB(struct super_block *sb) -{ - return sb->s_fs_info; -} - -static inline struct adfs_inode_info *ADFS_I(struct inode *inode) -{ - return container_of(inode, struct adfs_inode_info, vfs_inode); -} - #endif #endif diff --git a/include/linux/adfs_fs_i.h b/include/linux/adfs_fs_i.h deleted file mode 100644 index cb543034e54f..000000000000 --- a/include/linux/adfs_fs_i.h +++ /dev/null @@ -1,24 +0,0 @@ -/* - * linux/include/linux/adfs_fs_i.h - * - * Copyright (C) 1997 Russell King - */ - -#ifndef _ADFS_FS_I -#define _ADFS_FS_I - -/* - * adfs file system inode data in memory - */ -struct adfs_inode_info { - loff_t mmu_private; - unsigned long parent_id; /* object id of parent */ - __u32 loadaddr; /* RISC OS load address */ - __u32 execaddr; /* RISC OS exec address */ - unsigned int filetype; /* RISC OS file type */ - unsigned int attr; /* RISC OS permissions */ - unsigned int stamped:1; /* RISC OS file has date/time */ - struct inode vfs_inode; -}; - -#endif diff --git a/include/linux/adfs_fs_sb.h b/include/linux/adfs_fs_sb.h deleted file mode 100644 index d9bf05c02ccc..000000000000 --- a/include/linux/adfs_fs_sb.h +++ /dev/null @@ -1,38 +0,0 @@ -/* - * linux/include/linux/adfs_fs_sb.h - * - * Copyright (C) 1997-1999 Russell King - */ - -#ifndef _ADFS_FS_SB -#define _ADFS_FS_SB - -/* - * Forward-declare this - */ -struct adfs_discmap; -struct adfs_dir_ops; - -/* - * ADFS file system superblock data in memory - */ -struct adfs_sb_info { - struct adfs_discmap *s_map; /* bh list containing map */ - struct adfs_dir_ops *s_dir; /* directory operations */ - - uid_t s_uid; /* owner uid */ - gid_t s_gid; /* owner gid */ - umode_t s_owner_mask; /* ADFS owner perm -> unix perm */ - umode_t s_other_mask; /* ADFS other perm -> unix perm */ - - __u32 s_ids_per_zone; /* max. no ids in one zone */ - __u32 s_idlen; /* length of ID in map */ - __u32 s_map_size; /* sector size of a map */ - unsigned long s_size; /* total size (in blocks) of this fs */ - signed int s_map2blk; /* shift left by this for map->sector */ - unsigned int s_log2sharesize;/* log2 share size */ - __le32 s_version; /* disc format version */ - unsigned int s_namelen; /* maximum number of characters in name */ -}; - -#endif -- cgit v1.2.3 From 536c94901eb8f2eb6fccf81ae6be814899a9f6e8 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Tue, 16 Jun 2009 23:24:50 -0400 Subject: befs ->pust_super() doesn't need BKL Signed-off-by: Al Viro --- fs/befs/linuxvfs.c | 4 ---- 1 file changed, 4 deletions(-) (limited to 'fs') diff --git a/fs/befs/linuxvfs.c b/fs/befs/linuxvfs.c index 9367b6297d84..02c06138bc6a 100644 --- a/fs/befs/linuxvfs.c +++ b/fs/befs/linuxvfs.c @@ -737,8 +737,6 @@ parse_options(char *options, befs_mount_options * opts) static void befs_put_super(struct super_block *sb) { - lock_kernel(); - kfree(BEFS_SB(sb)->mount_opts.iocharset); BEFS_SB(sb)->mount_opts.iocharset = NULL; @@ -749,8 +747,6 @@ befs_put_super(struct super_block *sb) kfree(sb->s_fs_info); sb->s_fs_info = NULL; - - unlock_kernel(); } /* Allocate private field of the superblock, fill it. -- cgit v1.2.3 From e7ec952f6aa6ac1649ac49eb5e4de5b92c829d1e Mon Sep 17 00:00:00 2001 From: Al Viro Date: Tue, 16 Jun 2009 23:35:46 -0400 Subject: get rid of BKL in fs/efs Only readdir() really needed it, and that's easily fixable by switch to generic_file_llseek() Signed-off-by: Al Viro --- fs/efs/dir.c | 5 +---- fs/efs/namei.c | 9 +-------- fs/efs/symlink.c | 7 +------ 3 files changed, 3 insertions(+), 18 deletions(-) (limited to 'fs') diff --git a/fs/efs/dir.c b/fs/efs/dir.c index 49308a29798a..7ee6f7e3a608 100644 --- a/fs/efs/dir.c +++ b/fs/efs/dir.c @@ -5,12 +5,12 @@ */ #include -#include #include "efs.h" static int efs_readdir(struct file *, void *, filldir_t); const struct file_operations efs_dir_operations = { + .llseek = generic_file_llseek, .read = generic_read_dir, .readdir = efs_readdir, }; @@ -33,8 +33,6 @@ static int efs_readdir(struct file *filp, void *dirent, filldir_t filldir) { if (inode->i_size & (EFS_DIRBSIZE-1)) printk(KERN_WARNING "EFS: WARNING: readdir(): directory size not a multiple of EFS_DIRBSIZE\n"); - lock_kernel(); - /* work out where this entry can be found */ block = filp->f_pos >> EFS_DIRBSIZE_BITS; @@ -107,7 +105,6 @@ static int efs_readdir(struct file *filp, void *dirent, filldir_t filldir) { filp->f_pos = (block << EFS_DIRBSIZE_BITS) | slot; out: - unlock_kernel(); return 0; } diff --git a/fs/efs/namei.c b/fs/efs/namei.c index c3fb5f9c4a44..1511bf9e5f80 100644 --- a/fs/efs/namei.c +++ b/fs/efs/namei.c @@ -8,7 +8,6 @@ #include #include -#include #include #include "efs.h" @@ -63,16 +62,12 @@ struct dentry *efs_lookup(struct inode *dir, struct dentry *dentry, struct namei efs_ino_t inodenum; struct inode * inode = NULL; - lock_kernel(); inodenum = efs_find_entry(dir, dentry->d_name.name, dentry->d_name.len); if (inodenum) { inode = efs_iget(dir->i_sb, inodenum); - if (IS_ERR(inode)) { - unlock_kernel(); + if (IS_ERR(inode)) return ERR_CAST(inode); - } } - unlock_kernel(); return d_splice_alias(inode, dentry); } @@ -115,11 +110,9 @@ struct dentry *efs_get_parent(struct dentry *child) struct dentry *parent = ERR_PTR(-ENOENT); efs_ino_t ino; - lock_kernel(); ino = efs_find_entry(child->d_inode, "..", 2); if (ino) parent = d_obtain_alias(efs_iget(child->d_inode->i_sb, ino)); - unlock_kernel(); return parent; } diff --git a/fs/efs/symlink.c b/fs/efs/symlink.c index 41911ec83aaf..75117d0dac2b 100644 --- a/fs/efs/symlink.c +++ b/fs/efs/symlink.c @@ -9,7 +9,6 @@ #include #include #include -#include #include "efs.h" static int efs_symlink_readpage(struct file *file, struct page *page) @@ -22,9 +21,8 @@ static int efs_symlink_readpage(struct file *file, struct page *page) err = -ENAMETOOLONG; if (size > 2 * EFS_BLOCKSIZE) - goto fail_notlocked; + goto fail; - lock_kernel(); /* read first 512 bytes of link target */ err = -EIO; bh = sb_bread(inode->i_sb, efs_bmap(inode, 0)); @@ -40,14 +38,11 @@ static int efs_symlink_readpage(struct file *file, struct page *page) brelse(bh); } link[size] = '\0'; - unlock_kernel(); SetPageUptodate(page); kunmap(page); unlock_page(page); return 0; fail: - unlock_kernel(); -fail_notlocked: SetPageError(page); kunmap(page); unlock_page(page); -- cgit v1.2.3 From cc46759a8c0ac4c6f13aa4b0f470305c05f600e1 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Tue, 16 Jun 2009 23:47:45 -0400 Subject: get rid of BKL in fs/minix Signed-off-by: Al Viro --- fs/minix/bitmap.c | 25 +++++++++++++------------ fs/minix/dir.c | 5 +---- fs/minix/inode.c | 4 ---- 3 files changed, 14 insertions(+), 20 deletions(-) (limited to 'fs') diff --git a/fs/minix/bitmap.c b/fs/minix/bitmap.c index 3aebe322271a..6ac693faae49 100644 --- a/fs/minix/bitmap.c +++ b/fs/minix/bitmap.c @@ -12,13 +12,14 @@ /* bitmap.c contains the code that handles the inode and block bitmaps */ #include "minix.h" -#include #include #include #include static const int nibblemap[] = { 4,3,3,2,3,2,2,1,3,2,2,1,2,1,1,0 }; +static DEFINE_SPINLOCK(bitmap_lock); + static unsigned long count_free(struct buffer_head *map[], unsigned numblocks, __u32 numbits) { unsigned i, j, sum = 0; @@ -69,11 +70,11 @@ void minix_free_block(struct inode *inode, unsigned long block) return; } bh = sbi->s_zmap[zone]; - lock_kernel(); + spin_lock(&bitmap_lock); if (!minix_test_and_clear_bit(bit, bh->b_data)) printk("minix_free_block (%s:%lu): bit already cleared\n", sb->s_id, block); - unlock_kernel(); + spin_unlock(&bitmap_lock); mark_buffer_dirty(bh); return; } @@ -88,18 +89,18 @@ int minix_new_block(struct inode * inode) struct buffer_head *bh = sbi->s_zmap[i]; int j; - lock_kernel(); + spin_lock(&bitmap_lock); j = minix_find_first_zero_bit(bh->b_data, bits_per_zone); if (j < bits_per_zone) { minix_set_bit(j, bh->b_data); - unlock_kernel(); + spin_unlock(&bitmap_lock); mark_buffer_dirty(bh); j += i * bits_per_zone + sbi->s_firstdatazone-1; if (j < sbi->s_firstdatazone || j >= sbi->s_nzones) break; return j; } - unlock_kernel(); + spin_unlock(&bitmap_lock); } return 0; } @@ -211,10 +212,10 @@ void minix_free_inode(struct inode * inode) minix_clear_inode(inode); /* clear on-disk copy */ bh = sbi->s_imap[ino]; - lock_kernel(); + spin_lock(&bitmap_lock); if (!minix_test_and_clear_bit(bit, bh->b_data)) printk("minix_free_inode: bit %lu already cleared\n", bit); - unlock_kernel(); + spin_unlock(&bitmap_lock); mark_buffer_dirty(bh); out: clear_inode(inode); /* clear in-memory copy */ @@ -237,7 +238,7 @@ struct inode * minix_new_inode(const struct inode * dir, int * error) j = bits_per_zone; bh = NULL; *error = -ENOSPC; - lock_kernel(); + spin_lock(&bitmap_lock); for (i = 0; i < sbi->s_imap_blocks; i++) { bh = sbi->s_imap[i]; j = minix_find_first_zero_bit(bh->b_data, bits_per_zone); @@ -245,17 +246,17 @@ struct inode * minix_new_inode(const struct inode * dir, int * error) break; } if (!bh || j >= bits_per_zone) { - unlock_kernel(); + spin_unlock(&bitmap_lock); iput(inode); return NULL; } if (minix_test_and_set_bit(j, bh->b_data)) { /* shouldn't happen */ - unlock_kernel(); + spin_unlock(&bitmap_lock); printk("minix_new_inode: bit already set\n"); iput(inode); return NULL; } - unlock_kernel(); + spin_unlock(&bitmap_lock); mark_buffer_dirty(bh); j += i * bits_per_zone; if (!j || j > sbi->s_ninodes) { diff --git a/fs/minix/dir.c b/fs/minix/dir.c index e5f206467e40..d407e7a0b6fe 100644 --- a/fs/minix/dir.c +++ b/fs/minix/dir.c @@ -11,7 +11,6 @@ #include "minix.h" #include #include -#include #include typedef struct minix_dir_entry minix_dirent; @@ -20,6 +19,7 @@ typedef struct minix3_dir_entry minix3_dirent; static int minix_readdir(struct file *, void *, filldir_t); const struct file_operations minix_dir_operations = { + .llseek = generic_file_llseek, .read = generic_read_dir, .readdir = minix_readdir, .fsync = simple_fsync, @@ -102,8 +102,6 @@ static int minix_readdir(struct file * filp, void * dirent, filldir_t filldir) char *name; __u32 inumber; - lock_kernel(); - pos = (pos + chunk_size-1) & ~(chunk_size-1); if (pos >= inode->i_size) goto done; @@ -146,7 +144,6 @@ static int minix_readdir(struct file * filp, void * dirent, filldir_t filldir) done: filp->f_pos = (n << PAGE_CACHE_SHIFT) | offset; - unlock_kernel(); return 0; } diff --git a/fs/minix/inode.c b/fs/minix/inode.c index f91a23693597..74ea82d72164 100644 --- a/fs/minix/inode.c +++ b/fs/minix/inode.c @@ -35,8 +35,6 @@ static void minix_put_super(struct super_block *sb) int i; struct minix_sb_info *sbi = minix_sb(sb); - lock_kernel(); - if (!(sb->s_flags & MS_RDONLY)) { if (sbi->s_version != MINIX_V3) /* s_state is now out from V3 sb */ sbi->s_ms->s_state = sbi->s_mount_state; @@ -50,8 +48,6 @@ static void minix_put_super(struct super_block *sb) kfree(sbi->s_imap); sb->s_fs_info = NULL; kfree(sbi); - - unlock_kernel(); } static struct kmem_cache * minix_inode_cachep; -- cgit v1.2.3 From 5ac3455a843d2ca77333c954eea83aa4514c8199 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Tue, 16 Jun 2009 23:59:37 -0400 Subject: get rid of BKL in fs/sysv Signed-off-by: Al Viro --- fs/sysv/dir.c | 5 +---- fs/sysv/inode.c | 11 ----------- 2 files changed, 1 insertion(+), 15 deletions(-) (limited to 'fs') diff --git a/fs/sysv/dir.c b/fs/sysv/dir.c index c7798079e644..4e50286a4cc3 100644 --- a/fs/sysv/dir.c +++ b/fs/sysv/dir.c @@ -15,13 +15,13 @@ #include #include -#include #include #include "sysv.h" static int sysv_readdir(struct file *, void *, filldir_t); const struct file_operations sysv_dir_operations = { + .llseek = generic_file_llseek, .read = generic_read_dir, .readdir = sysv_readdir, .fsync = simple_fsync, @@ -74,8 +74,6 @@ static int sysv_readdir(struct file * filp, void * dirent, filldir_t filldir) unsigned long n = pos >> PAGE_CACHE_SHIFT; unsigned long npages = dir_pages(inode); - lock_kernel(); - pos = (pos + SYSV_DIRSIZE-1) & ~(SYSV_DIRSIZE-1); if (pos >= inode->i_size) goto done; @@ -113,7 +111,6 @@ static int sysv_readdir(struct file * filp, void * dirent, filldir_t filldir) done: filp->f_pos = ((loff_t)n << PAGE_CACHE_SHIFT) | offset; - unlock_kernel(); return 0; } diff --git a/fs/sysv/inode.c b/fs/sysv/inode.c index 479923456a54..9824743832a7 100644 --- a/fs/sysv/inode.c +++ b/fs/sysv/inode.c @@ -21,7 +21,6 @@ * the superblock. */ -#include #include #include #include @@ -37,7 +36,6 @@ static int sysv_sync_fs(struct super_block *sb, int wait) unsigned long time = get_seconds(), old_time; lock_super(sb); - lock_kernel(); /* * If we are going to write out the super block, @@ -52,7 +50,6 @@ static int sysv_sync_fs(struct super_block *sb, int wait) mark_buffer_dirty(sbi->s_bh2); } - unlock_kernel(); unlock_super(sb); return 0; @@ -82,8 +79,6 @@ static void sysv_put_super(struct super_block *sb) { struct sysv_sb_info *sbi = SYSV_SB(sb); - lock_kernel(); - if (sb->s_dirt) sysv_write_super(sb); @@ -99,8 +94,6 @@ static void sysv_put_super(struct super_block *sb) brelse(sbi->s_bh2); kfree(sbi); - - unlock_kernel(); } static int sysv_statfs(struct dentry *dentry, struct kstatfs *buf) @@ -275,7 +268,6 @@ int sysv_write_inode(struct inode *inode, int wait) return -EIO; } - lock_kernel(); raw_inode->i_mode = cpu_to_fs16(sbi, inode->i_mode); raw_inode->i_uid = cpu_to_fs16(sbi, fs_high2lowuid(inode->i_uid)); raw_inode->i_gid = cpu_to_fs16(sbi, fs_high2lowgid(inode->i_gid)); @@ -291,7 +283,6 @@ int sysv_write_inode(struct inode *inode, int wait) for (block = 0; block < 10+1+1+1; block++) write3byte(sbi, (u8 *)&si->i_data[block], &raw_inode->i_data[3*block]); - unlock_kernel(); mark_buffer_dirty(bh); if (wait) { sync_dirty_buffer(bh); @@ -315,9 +306,7 @@ static void sysv_delete_inode(struct inode *inode) truncate_inode_pages(&inode->i_data, 0); inode->i_size = 0; sysv_truncate(inode); - lock_kernel(); sysv_free_inode(inode); - unlock_kernel(); } static struct kmem_cache *sysv_inode_cachep; -- cgit v1.2.3 From 852969b2d273e77dabbc22e1c1058cbafb7ad7d2 Mon Sep 17 00:00:00 2001 From: David Daney Date: Wed, 27 May 2009 17:47:45 -0700 Subject: Hugetlbfs: Enable hugetlbfs for more systems in Kconfig. As part of adding hugetlbfs support for MIPS, I am adding a new kconfig variable 'SYS_SUPPORTS_HUGETLBFS'. Since some mips cpu varients don't yet support it, we can enable selection of HUGETLBFS on a system by system basis from the arch/mips/Kconfig. Signed-off-by: David Daney CC: William Irwin Signed-off-by: Ralf Baechle --- fs/Kconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/Kconfig b/fs/Kconfig index 4044f163035f..d78e950402c1 100644 --- a/fs/Kconfig +++ b/fs/Kconfig @@ -134,7 +134,7 @@ config TMPFS_POSIX_ACL config HUGETLBFS bool "HugeTLB file system support" depends on X86 || IA64 || PPC64 || SPARC64 || (SUPERH && MMU) || \ - (S390 && 64BIT) || BROKEN + (S390 && 64BIT) || SYS_SUPPORTS_HUGETLBFS || BROKEN help hugetlbfs is a filesystem backing for HugeTLB pages, based on ramfs. For architectures that support it, say Y here and read -- cgit v1.2.3 From a566a6b11c86147fe9fc9db7ab15f9eecca3e862 Mon Sep 17 00:00:00 2001 From: Steven Whitehouse Date: Mon, 15 Jun 2009 08:26:48 +0100 Subject: dlm: Fix uninitialised variable warning in lock.c MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit CC [M] fs/dlm/lock.o fs/dlm/lock.c: In function ‘find_rsb’: fs/dlm/lock.c:438: warning: ‘r’ may be used uninitialized in this function Since r is used on the error path to set r_ret, set it to NULL. Signed-off-by: Steven Whitehouse Signed-off-by: David Teigland --- fs/dlm/lock.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/dlm/lock.c b/fs/dlm/lock.c index 205ec95b347e..eb507c453c5f 100644 --- a/fs/dlm/lock.c +++ b/fs/dlm/lock.c @@ -435,7 +435,7 @@ static int search_rsb(struct dlm_ls *ls, char *name, int len, int b, static int find_rsb(struct dlm_ls *ls, char *name, int namelen, unsigned int flags, struct dlm_rsb **r_ret) { - struct dlm_rsb *r, *tmp; + struct dlm_rsb *r = NULL, *tmp; uint32_t hash, bucket; int error = -EINVAL; -- cgit v1.2.3 From 1efae38140546db403845d628db9f2d608caa87e Mon Sep 17 00:00:00 2001 From: Ricardo Labiaga Date: Wed, 1 Apr 2009 09:21:46 -0400 Subject: nfs41: Add Kconfig symbols for NFSv4.1 Added CONFIG_NFS_V4_1 and made it depend upon CONFIG_NFS_V4 and EXPERIMENTAL. Indicate that CONFIG_NFS_V4_1 is for NFS developers at the moment At the moment we're expecting folks trying out nfs41 to actively participate in the development process by helping us debug issues and ideally send patches to fix problems. Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/Kconfig | 9 +++++++++ 1 file changed, 9 insertions(+) (limited to 'fs') diff --git a/fs/nfs/Kconfig b/fs/nfs/Kconfig index e67f3ec07736..5d6d6f415935 100644 --- a/fs/nfs/Kconfig +++ b/fs/nfs/Kconfig @@ -74,6 +74,15 @@ config NFS_V4 If unsure, say N. +config NFS_V4_1 + bool "NFS client support for NFSv4.1 (DEVELOPER ONLY)" + depends on NFS_V4 && EXPERIMENTAL + help + This option enables support for minor version 1 of the NFSv4 protocol + (draft-ietf-nfsv4-minorversion1) in the kernel's NFS client. + + Unless you're an NFS developer, say N. + config ROOT_NFS bool "Root file system on NFS" depends on NFS_FS=y && IP_PNP -- cgit v1.2.3 From 3fd5be9e19921a89d9ed78d6a708a379a6c3c76a Mon Sep 17 00:00:00 2001 From: Mike Sager Date: Wed, 1 Apr 2009 09:21:48 -0400 Subject: nfs41: add mount command option minorversion mount -t nfs4 -o minorversion=[0|1] specifies whether to use 4.0 or 4.1. By default, the minorversion is set to 0. Signed-off-by: Mike Sager [set default minorversion to 0 as per Trond and SteveD's request] Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/internal.h | 1 + fs/nfs/super.c | 10 ++++++++++ 2 files changed, 11 insertions(+) (limited to 'fs') diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h index e4d6a8348adf..ffa6bd54d439 100644 --- a/fs/nfs/internal.h +++ b/fs/nfs/internal.h @@ -44,6 +44,7 @@ struct nfs_parsed_mount_data { unsigned int auth_flavor_len; rpc_authflavor_t auth_flavors[1]; char *client_address; + unsigned int minorversion; char *fscache_uniq; struct { diff --git a/fs/nfs/super.c b/fs/nfs/super.c index d2d67781c579..5a8fdc791cc1 100644 --- a/fs/nfs/super.c +++ b/fs/nfs/super.c @@ -90,6 +90,7 @@ enum { Opt_mountport, Opt_mountvers, Opt_nfsvers, + Opt_minorversion, /* Mount options that take string arguments */ Opt_sec, Opt_proto, Opt_mountproto, Opt_mounthost, @@ -155,6 +156,7 @@ static const match_table_t nfs_mount_option_tokens = { { Opt_mountvers, "mountvers=%u" }, { Opt_nfsvers, "nfsvers=%u" }, { Opt_nfsvers, "vers=%u" }, + { Opt_minorversion, "minorversion=%u" }, { Opt_sec, "sec=%s" }, { Opt_proto, "proto=%s" }, @@ -1211,6 +1213,13 @@ static int nfs_parse_mount_options(char *raw, nfs_parse_invalid_value("nfsvers"); } break; + case Opt_minorversion: + if (match_int(args, &option)) + return 0; + if (option < 0 || option > NFS4_MAX_MINOR_VERSION) + return 0; + mnt->minorversion = option; + break; /* * options that take text values @@ -2261,6 +2270,7 @@ static int nfs4_validate_mount_data(void *options, args->nfs_server.port = NFS_PORT; /* 2049 unless user set port= */ args->auth_flavors[0] = RPC_AUTH_UNIX; args->auth_flavor_len = 0; + args->minorversion = 0; switch (data->version) { case 1: -- cgit v1.2.3 From 94a417f3d7a02478a2d7842e693a61339fb54ea4 Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Wed, 1 Apr 2009 09:21:49 -0400 Subject: nfs41: nfs_client.cl_minorversion This field is set to the nfsv4 minor version for this mount. Signed-off-by: Benny Halevy Note: This patch sets the referral to the same minorversion as the current mount. Revisit in future patch. Signed-off-by: Andy Adamson [removed cl_minorversion assignment in nfs_set_client] Signed-off-by: Benny Halevy [always define nfs_client.cl_minorversion] Signed-off-by: Trond Myklebust --- fs/nfs/client.c | 9 ++++++--- include/linux/nfs_fs_sb.h | 3 ++- 2 files changed, 8 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/nfs/client.c b/fs/nfs/client.c index 75c9cd2aa119..0efcb55c2caa 100644 --- a/fs/nfs/client.c +++ b/fs/nfs/client.c @@ -1101,7 +1101,8 @@ static int nfs4_set_client(struct nfs_server *server, const size_t addrlen, const char *ip_addr, rpc_authflavor_t authflavour, - int proto, const struct rpc_timeout *timeparms) + int proto, const struct rpc_timeout *timeparms, + u32 minorversion) { struct nfs_client_initdata cl_init = { .hostname = hostname, @@ -1164,7 +1165,8 @@ static int nfs4_init_server(struct nfs_server *server, data->client_address, data->auth_flavors[0], data->nfs_server.protocol, - &timeparms); + &timeparms, + data->minorversion); if (error < 0) goto error; @@ -1282,7 +1284,8 @@ struct nfs_server *nfs4_create_referral_server(struct nfs_clone_mount *data, parent_client->cl_ipaddr, data->authflavor, parent_server->client->cl_xprt->prot, - parent_server->client->cl_timeout); + parent_server->client->cl_timeout, + parent_client->cl_minorversion); if (error < 0) goto error; diff --git a/include/linux/nfs_fs_sb.h b/include/linux/nfs_fs_sb.h index 6ad75948cbf7..e9a51fe46aa3 100644 --- a/include/linux/nfs_fs_sb.h +++ b/include/linux/nfs_fs_sb.h @@ -32,6 +32,7 @@ struct nfs_client { const struct nfs_rpc_ops *rpc_ops; /* NFS protocol vector */ int cl_proto; /* Network transport protocol */ + u32 cl_minorversion;/* NFSv4 minorversion */ struct rpc_cred *cl_machine_cred; #ifdef CONFIG_NFS_V4 @@ -63,7 +64,7 @@ struct nfs_client { */ char cl_ipaddr[48]; unsigned char cl_id_uniquifier; -#endif +#endif /* CONFIG_NFS_V4 */ #ifdef CONFIG_NFS_FSCACHE struct fscache_cookie *fscache; /* client index cache cookie */ -- cgit v1.2.3 From 5aae4a9ae0dd55409a42ca61b82ef1f5a840091e Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Wed, 1 Apr 2009 09:21:50 -0400 Subject: nfs41: Use mount minorversion option Use the mount minorversion option to initialize the nfs_client cl_minorversion and match it in nfs_match_client() when looking up a nfs_client. [nfs41: remove ifdefs around nfs_client_initdata.minorversion] Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/client.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfs/client.c b/fs/nfs/client.c index 0efcb55c2caa..a736160046c3 100644 --- a/fs/nfs/client.c +++ b/fs/nfs/client.c @@ -102,6 +102,7 @@ struct nfs_client_initdata { size_t addrlen; const struct nfs_rpc_ops *rpc_ops; int proto; + u32 minorversion; }; /* @@ -150,6 +151,7 @@ static struct nfs_client *nfs_alloc_client(const struct nfs_client_initdata *cl_ rpc_init_wait_queue(&clp->cl_rpcwaitq, "NFS client"); clp->cl_boot_time = CURRENT_TIME; clp->cl_state = 1 << NFS4CLNT_LEASE_EXPIRED; + clp->cl_minorversion = cl_init->minorversion; #endif cred = rpc_lookup_machine_cred(); if (!IS_ERR(cred)) @@ -420,7 +422,9 @@ static struct nfs_client *nfs_match_client(const struct nfs_client_initdata *dat if (clp->cl_proto != data->proto) continue; - + /* Match nfsv4 minorversion */ + if (clp->cl_minorversion != data->minorversion) + continue; /* Match the full socket address */ if (!nfs_sockaddr_cmp(sap, clap)) continue; @@ -1110,6 +1114,7 @@ static int nfs4_set_client(struct nfs_server *server, .addrlen = addrlen, .rpc_ops = &nfs_v4_clientops, .proto = proto, + .minorversion = minorversion, }; struct nfs_client *clp; int error; -- cgit v1.2.3 From c2e713dd83dcabb2ae951ea572e7de68c96e2d48 Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Wed, 1 Apr 2009 09:21:51 -0400 Subject: nfs41: translate NFS4ERR_MINOR_VERS_MISMATCH to EPROTONOSUPPORT To be returned to the mount command when trying to mount a v4 server using minorversion 1. Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4state.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'fs') diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c index 0298e909559f..bc683ed477e1 100644 --- a/fs/nfs/nfs4state.c +++ b/fs/nfs/nfs4state.c @@ -1107,6 +1107,8 @@ static int nfs4_reclaim_lease(struct nfs_client *clp) nfs4_clear_machine_cred(clp); status = -EAGAIN; } + if (status == -NFS4ERR_MINOR_VERS_MISMATCH) + status = -EPROTONOSUPPORT; } return status; } -- cgit v1.2.3 From 557134a39c8d2ab79d8b8d53438e03e29feb5ec4 Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Wed, 1 Apr 2009 09:21:53 -0400 Subject: nfs41: sessions client infrastructure NFSv4.1 Sessions basic data types, initialization, and destruction. The session is always associated with a struct nfs_client that holds the exchange_id results. Signed-off-by: Rahul Iyer Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [remove extraneous rpc_clnt pointer, use the struct nfs_client cl_rpcclient. remove the rpc_clnt parameter from nfs4 nfs4_init_session] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [Use the presence of a session to determine behaviour instead of the minorversion number.] Signed-off-by: Andy Adamson [constified nfs4_has_session's struct nfs_client parameter] Signed-off-by: Benny Halevy [Rename nfs4_put_session() to nfs4_destroy_session() and call it from nfs4_free_client() not nfs4_free_server(). Also get rid of nfs4_get_session() and the ref_count in nfs4_session struct as keeping track of nfs_client should be sufficient] Signed-off-by: Alexandros Batsakis [nfs41: pass rsize and wsize into nfs4_init_session] Signed-off-by: Andy Adamson [separated out removal of rpc_clnt parameter from nfs4_init_session ot a patch of its own] Signed-off-by: Benny Halevy [Pass the nfs_client pointer into nfs4_alloc_session] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: don't assign to session->clp->cl_session in nfs4_destroy_session] [nfs41: fixup nfs4_clear_client_minor_version] [introduce nfs4_clear_client_minor_version() in this patch] Signed-off-by: Benny Halevy [Refactor nfs4_init_session] Moved session allocation into nfs4_init_client_minor_version, called from nfs4_init_client. Leave rwise and wsize initialization in nfs4_init_session, called from nfs4_init_server. Reverted moving of nfs_fsid definition to nfs_fs_sb.h Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: Move NFS4_MAX_SLOT_TABLE define from under CONFIG_NFS_V4_1] [Fix comile error when CONFIG_NFS_V4_1 is not set.] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [moved nfs4_init_slot_table definition to "create_session operation"] Signed-off-by: Benny Halevy [nfs41: alloc session with GFP_KERNEL] Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/client.c | 60 +++++++++++++++++++++++++++++++++++++++++++++++ fs/nfs/internal.h | 12 ++++++++++ fs/nfs/nfs4_fs.h | 4 ++++ fs/nfs/nfs4proc.c | 34 +++++++++++++++++++++++++++ include/linux/nfs_fs_sb.h | 49 ++++++++++++++++++++++++++++++++++++++ include/linux/nfs_xdr.h | 15 ++++++++++++ 6 files changed, 174 insertions(+) (limited to 'fs') diff --git a/fs/nfs/client.c b/fs/nfs/client.c index a736160046c3..f1506f148521 100644 --- a/fs/nfs/client.c +++ b/fs/nfs/client.c @@ -183,6 +183,20 @@ static void nfs4_shutdown_client(struct nfs_client *clp) #endif } +/* + * Clears/puts all minor version specific parts from an nfs_client struct + * reverting it to minorversion 0. + */ +static void nfs4_clear_client_minor_version(struct nfs_client *clp) +{ +#ifdef CONFIG_NFS_V4_1 + if (nfs4_has_session(clp)) { + nfs4_destroy_session(clp->cl_session); + clp->cl_session = NULL; + } +#endif /* CONFIG_NFS_V4_1 */ +} + /* * Destroy a shared client record */ @@ -190,6 +204,7 @@ static void nfs_free_client(struct nfs_client *clp) { dprintk("--> nfs_free_client(%u)\n", clp->rpc_ops->version); + nfs4_clear_client_minor_version(clp); nfs4_shutdown_client(clp); nfs_fscache_release_client_cookie(clp); @@ -1053,6 +1068,30 @@ error: } #ifdef CONFIG_NFS_V4 +/* + * Initialize the minor version specific parts of an NFS4 client record + */ +static int nfs4_init_client_minor_version(struct nfs_client *clp) +{ +#if defined(CONFIG_NFS_V4_1) + if (clp->cl_minorversion) { + struct nfs4_session *session = NULL; + /* + * Create the session and mark it expired. + * When a SEQUENCE operation encounters the expired session + * it will do session recovery to initialize it. + */ + session = nfs4_alloc_session(clp); + if (!session) + return -ENOMEM; + + clp->cl_session = session; + } +#endif /* CONFIG_NFS_V4_1 */ + + return 0; +} + /* * Initialise an NFS4 client record */ @@ -1087,6 +1126,10 @@ static int nfs4_init_client(struct nfs_client *clp, } __set_bit(NFS_CS_IDMAP, &clp->cl_res_state); + error = nfs4_init_client_minor_version(clp); + if (error < 0) + goto error; + nfs_mark_client_ready(clp, NFS_CS_READY); return 0; @@ -1143,6 +1186,21 @@ error: return error; } +/* + * Initialize a session. + * Note: save the mount rsize and wsize for create_server negotiation. + */ +static void nfs4_init_session(struct nfs_client *clp, + unsigned int wsize, unsigned int rsize) +{ +#if defined(CONFIG_NFS_V4_1) + if (nfs4_has_session(clp)) { + clp->cl_session->fc_attrs.max_rqst_sz = wsize; + clp->cl_session->fc_attrs.max_resp_sz = rsize; + } +#endif /* CONFIG_NFS_V4_1 */ +} + /* * Create a version 4 volume record */ @@ -1221,6 +1279,8 @@ struct nfs_server *nfs4_create_server(const struct nfs_parsed_mount_data *data, BUG_ON(!server->nfs_client->rpc_ops); BUG_ON(!server->nfs_client->rpc_ops->file_inode_ops); + nfs4_init_session(server->nfs_client, server->wsize, server->rsize); + /* Probe the root fh to retrieve its FSID */ error = nfs4_path_walk(server, mntfh, data->nfs_server.export_path); if (error < 0) diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h index ffa6bd54d439..7cef45db9257 100644 --- a/fs/nfs/internal.h +++ b/fs/nfs/internal.h @@ -206,6 +206,18 @@ extern int nfs4_path_walk(struct nfs_server *server, const char *path); #endif +/* + * Determine if sessions are in use. + */ +static inline int nfs4_has_session(const struct nfs_client *clp) +{ +#ifdef CONFIG_NFS_V4_1 + if (clp->cl_session) + return 1; +#endif /* CONFIG_NFS_V4_1 */ + return 0; +} + /* * Determine the device name as a string */ diff --git a/fs/nfs/nfs4_fs.h b/fs/nfs/nfs4_fs.h index 84345deab26f..acac6f8c3d39 100644 --- a/fs/nfs/nfs4_fs.h +++ b/fs/nfs/nfs4_fs.h @@ -202,6 +202,10 @@ extern int nfs4_proc_fs_locations(struct inode *dir, const struct qstr *name, extern struct nfs4_state_recovery_ops nfs4_reboot_recovery_ops; extern struct nfs4_state_recovery_ops nfs4_nograce_recovery_ops; +#if defined(CONFIG_NFS_V4_1) +extern void nfs4_destroy_session(struct nfs4_session *session); +extern struct nfs4_session *nfs4_alloc_session(struct nfs_client *clp); +#endif /* CONFIG_NFS_V4_1 */ extern const u32 nfs4_fattr_bitmap[2]; extern const u32 nfs4_statfs_bitmap[2]; diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 4674f8092da8..cdd8e74c47d0 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -3723,6 +3723,40 @@ int nfs4_proc_fs_locations(struct inode *dir, const struct qstr *name, return status; } +#ifdef CONFIG_NFS_V4_1 +/* Destroy the slot table */ +static void nfs4_destroy_slot_table(struct nfs4_session *session) +{ + if (session->fc_slot_table.slots == NULL) + return; + kfree(session->fc_slot_table.slots); + session->fc_slot_table.slots = NULL; + return; +} + +struct nfs4_session *nfs4_alloc_session(struct nfs_client *clp) +{ + struct nfs4_session *session; + struct nfs4_slot_table *tbl; + + session = kzalloc(sizeof(struct nfs4_session), GFP_KERNEL); + if (!session) + return NULL; + tbl = &session->fc_slot_table; + spin_lock_init(&tbl->slot_tbl_lock); + rpc_init_wait_queue(&tbl->slot_tbl_waitq, "Slot table"); + session->clp = clp; + return session; +} + +void nfs4_destroy_session(struct nfs4_session *session) +{ + nfs4_destroy_slot_table(session); + kfree(session); +} + +#endif /* CONFIG_NFS_V4_1 */ + struct nfs4_state_recovery_ops nfs4_reboot_recovery_ops = { .owner_flag_bit = NFS_OWNER_RECLAIM_REBOOT, .state_flag_bit = NFS_STATE_RECLAIM_REBOOT, diff --git a/include/linux/nfs_fs_sb.h b/include/linux/nfs_fs_sb.h index e9a51fe46aa3..b47c0fc55d42 100644 --- a/include/linux/nfs_fs_sb.h +++ b/include/linux/nfs_fs_sb.h @@ -4,9 +4,12 @@ #include #include #include +#include +#include #include +struct nfs4_session; struct nfs_iostats; struct nlm_host; @@ -66,6 +69,10 @@ struct nfs_client { unsigned char cl_id_uniquifier; #endif /* CONFIG_NFS_V4 */ +#ifdef CONFIG_NFS_V4_1 + struct nfs4_session *cl_session; /* sharred session */ +#endif /* CONFIG_NFS_V4_1 */ + #ifdef CONFIG_NFS_FSCACHE struct fscache_cookie *fscache; /* client index cache cookie */ #endif @@ -146,4 +153,46 @@ struct nfs_server { #define NFS_CAP_ACLS (1U << 3) #define NFS_CAP_ATOMIC_OPEN (1U << 4) + +/* maximum number of slots to use */ +#define NFS4_MAX_SLOT_TABLE RPC_MAX_SLOT_TABLE + +#if defined(CONFIG_NFS_V4_1) + +/* Sessions */ +#define SLOT_TABLE_SZ (NFS4_MAX_SLOT_TABLE/(8*sizeof(long))) +struct nfs4_slot_table { + struct nfs4_slot *slots; /* seqid per slot */ + unsigned long used_slots[SLOT_TABLE_SZ]; /* used/unused bitmap */ + spinlock_t slot_tbl_lock; + struct rpc_wait_queue slot_tbl_waitq; /* allocators may wait here */ + int max_slots; /* # slots in table */ + int highest_used_slotid; /* sent to server on each SEQ. + * op for dynamic resizing */ +}; + +static inline int slot_idx(struct nfs4_slot_table *tbl, struct nfs4_slot *sp) +{ + return sp - tbl->slots; +} + +/* + * Session related parameters + */ +struct nfs4_session { + struct nfs4_sessionid sess_id; + u32 flags; + unsigned long session_state; + u32 hash_alg; + u32 ssv_len; + + /* The fore and back channel */ + struct nfs4_channel_attrs fc_attrs; + struct nfs4_slot_table fc_slot_table; + struct nfs4_channel_attrs bc_attrs; + /* back channel has one slot */ + struct nfs_client *clp; +}; + +#endif /* CONFIG_NFS_V4_1 */ #endif diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h index f2c5700c7b6e..8f8c026c558e 100644 --- a/include/linux/nfs_xdr.h +++ b/include/linux/nfs_xdr.h @@ -146,6 +146,21 @@ struct nfs4_change_info { struct nfs_seqid; +/* nfs41 sessions channel attributes */ +struct nfs4_channel_attrs { + u32 headerpadsz; + u32 max_rqst_sz; + u32 max_resp_sz; + u32 max_resp_sz_cached; + u32 max_ops; + u32 max_reqs; +}; + +/* nfs41 sessions slot seqid */ +struct nfs4_slot { + u32 seq_nr; +}; + struct nfs4_sequence_args { /* stub */ }; -- cgit v1.2.3 From 43652ad55342d9146d8035932101a5814b22315a Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Wed, 1 Apr 2009 09:21:54 -0400 Subject: nfs41: use nfs4_server_caps_arg In preparation for nfs41 sequence processing. Signed-off-by: Andy Admason [define nfs4_server_caps_arg] Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 5 ++++- fs/nfs/nfs4xdr.c | 5 +++-- include/linux/nfs_xdr.h | 5 +++++ 3 files changed, 12 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index cdd8e74c47d0..5ef1022b7e6f 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -1584,10 +1584,13 @@ void nfs4_close_context(struct nfs_open_context *ctx, int is_sync) static int _nfs4_server_capabilities(struct nfs_server *server, struct nfs_fh *fhandle) { + struct nfs4_server_caps_arg args = { + .fhandle = fhandle, + }; struct nfs4_server_caps_res res = {}; struct rpc_message msg = { .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_SERVER_CAPS], - .rpc_argp = fhandle, + .rpc_argp = &args, .rpc_resp = &res, }; int status; diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index 1690f0e44b91..91305861037a 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -1900,7 +1900,8 @@ static int nfs4_xdr_enc_statfs(struct rpc_rqst *req, __be32 *p, const struct nfs /* * GETATTR_BITMAP request */ -static int nfs4_xdr_enc_server_caps(struct rpc_rqst *req, __be32 *p, const struct nfs_fh *fhandle) +static int nfs4_xdr_enc_server_caps(struct rpc_rqst *req, __be32 *p, + struct nfs4_server_caps_arg *args) { struct xdr_stream xdr; struct compound_hdr hdr = { @@ -1909,7 +1910,7 @@ static int nfs4_xdr_enc_server_caps(struct rpc_rqst *req, __be32 *p, const struc xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); - encode_putfh(&xdr, fhandle, &hdr); + encode_putfh(&xdr, args->fhandle, &hdr); encode_getattr_one(&xdr, FATTR4_WORD0_SUPPORTED_ATTRS| FATTR4_WORD0_LINK_SUPPORT| FATTR4_WORD0_SYMLINK_SUPPORT| diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h index 8f8c026c558e..a7b7f2a059cc 100644 --- a/include/linux/nfs_xdr.h +++ b/include/linux/nfs_xdr.h @@ -791,6 +791,11 @@ struct nfs4_statfs_arg { struct nfs4_sequence_args seq_args; }; +struct nfs4_server_caps_arg { + struct nfs_fh *fhandle; + struct nfs4_sequence_args seq_args; +}; + struct nfs4_server_caps_res { u32 attr_bitmask[2]; u32 acl_bitmask; -- cgit v1.2.3 From f50c7000817e7cb4e676ac5d911a82c0f3fd226f Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Wed, 1 Apr 2009 09:21:55 -0400 Subject: nfs41: use nfs4_readlink_res In preparation for nfs41 sequence processing. Signed-off-by: Andy Admason [define nfs4_readlink_res] Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 3 ++- fs/nfs/nfs4xdr.c | 3 ++- include/linux/nfs_xdr.h | 4 ++++ 3 files changed, 8 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 5ef1022b7e6f..b0ec8ff96eb7 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -1960,10 +1960,11 @@ static int _nfs4_proc_readlink(struct inode *inode, struct page *page, .pglen = pglen, .pages = &page, }; + struct nfs4_readlink_res res; struct rpc_message msg = { .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_READLINK], .rpc_argp = &args, - .rpc_resp = NULL, + .rpc_resp = &res, }; return rpc_call_sync(NFS_CLIENT(inode), &msg, 0); diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index 91305861037a..1e41420916ad 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -4235,7 +4235,8 @@ out: /* * Decode READLINK response */ -static int nfs4_xdr_dec_readlink(struct rpc_rqst *rqstp, __be32 *p, void *res) +static int nfs4_xdr_dec_readlink(struct rpc_rqst *rqstp, __be32 *p, + struct nfs4_readlink_res *res) { struct xdr_stream xdr; struct compound_hdr hdr; diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h index a7b7f2a059cc..f71260aeb803 100644 --- a/include/linux/nfs_xdr.h +++ b/include/linux/nfs_xdr.h @@ -754,6 +754,10 @@ struct nfs4_readlink { struct nfs4_sequence_args seq_args; }; +struct nfs4_readlink_res { + struct nfs4_sequence_res seq_res; +}; + struct nfs4_rename_arg { const struct nfs_fh * old_dir; const struct nfs_fh * new_dir; -- cgit v1.2.3 From 24ad148a0ff74b1e703a8bc5b3e0793dc7d4e3a9 Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Wed, 1 Apr 2009 09:21:56 -0400 Subject: nfs41: use nfs4_statfs_res In preparation for nfs41 sequence processing. Signed-off-by: Andy Admason [define nfs4_statfs_res] Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 5 ++++- fs/nfs/nfs4xdr.c | 5 +++-- include/linux/nfs_xdr.h | 5 +++++ 3 files changed, 12 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index b0ec8ff96eb7..b715f6057611 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -2426,10 +2426,13 @@ static int _nfs4_proc_statfs(struct nfs_server *server, struct nfs_fh *fhandle, .fh = fhandle, .bitmask = server->attr_bitmask, }; + struct nfs4_statfs_res res = { + .fsstat = fsstat, + }; struct rpc_message msg = { .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_STATFS], .rpc_argp = &args, - .rpc_resp = fsstat, + .rpc_resp = &res, }; nfs_fattr_init(fsstat->fattr); diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index 1e41420916ad..b7871ad82aac 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -4387,7 +4387,8 @@ static int nfs4_xdr_dec_pathconf(struct rpc_rqst *req, __be32 *p, struct nfs_pat /* * STATFS request */ -static int nfs4_xdr_dec_statfs(struct rpc_rqst *req, __be32 *p, struct nfs_fsstat *fsstat) +static int nfs4_xdr_dec_statfs(struct rpc_rqst *req, __be32 *p, + struct nfs4_statfs_res *res) { struct xdr_stream xdr; struct compound_hdr hdr; @@ -4398,7 +4399,7 @@ static int nfs4_xdr_dec_statfs(struct rpc_rqst *req, __be32 *p, struct nfs_fssta if (!status) status = decode_putfh(&xdr); if (!status) - status = decode_statfs(&xdr, fsstat); + status = decode_statfs(&xdr, res->fsstat); return status; } diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h index f71260aeb803..4dac59ef6f4f 100644 --- a/include/linux/nfs_xdr.h +++ b/include/linux/nfs_xdr.h @@ -795,6 +795,11 @@ struct nfs4_statfs_arg { struct nfs4_sequence_args seq_args; }; +struct nfs4_statfs_res { + struct nfs_fsstat *fsstat; + struct nfs4_sequence_res seq_res; +}; + struct nfs4_server_caps_arg { struct nfs_fh *fhandle; struct nfs4_sequence_args seq_args; -- cgit v1.2.3 From 3dda5e434721f942870ee30bc6103761618d410f Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Wed, 1 Apr 2009 09:21:57 -0400 Subject: nfs41: use nfs4_fsinfo_res In preparation for nfs41 sequence processing. Signed-off-by: Andy Admason [define nfs4_fsinfo_res] Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 5 ++++- fs/nfs/nfs4xdr.c | 5 +++-- include/linux/nfs_xdr.h | 5 +++++ 3 files changed, 12 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index b715f6057611..b8915ef533da 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -2458,10 +2458,13 @@ static int _nfs4_do_fsinfo(struct nfs_server *server, struct nfs_fh *fhandle, .fh = fhandle, .bitmask = server->attr_bitmask, }; + struct nfs4_fsinfo_res res = { + .fsinfo = fsinfo, + }; struct rpc_message msg = { .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_FSINFO], .rpc_argp = &args, - .rpc_resp = fsinfo, + .rpc_resp = &res, }; return rpc_call_sync(server->client, &msg, 0); diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index b7871ad82aac..d9ab8209c286 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -4351,7 +4351,8 @@ out: /* * FSINFO request */ -static int nfs4_xdr_dec_fsinfo(struct rpc_rqst *req, __be32 *p, struct nfs_fsinfo *fsinfo) +static int nfs4_xdr_dec_fsinfo(struct rpc_rqst *req, __be32 *p, + struct nfs4_fsinfo_res *res) { struct xdr_stream xdr; struct compound_hdr hdr; @@ -4362,7 +4363,7 @@ static int nfs4_xdr_dec_fsinfo(struct rpc_rqst *req, __be32 *p, struct nfs_fsinf if (!status) status = decode_putfh(&xdr); if (!status) - status = decode_fsinfo(&xdr, fsinfo); + status = decode_fsinfo(&xdr, res->fsinfo); return status; } diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h index 4dac59ef6f4f..7d64913cbb1b 100644 --- a/include/linux/nfs_xdr.h +++ b/include/linux/nfs_xdr.h @@ -675,6 +675,11 @@ struct nfs4_fsinfo_arg { struct nfs4_sequence_args seq_args; }; +struct nfs4_fsinfo_res { + struct nfs_fsinfo *fsinfo; + struct nfs4_sequence_res seq_res; +}; + struct nfs4_getattr_arg { const struct nfs_fh * fh; const u32 * bitmask; -- cgit v1.2.3 From d45b2989a7956ae9e71d584ceac942278c0371c7 Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Wed, 1 Apr 2009 09:21:58 -0400 Subject: nfs41: use nfs4_pathconf_res In preparation for nfs41 sequence processing. Signed-off-by: Andy Admason [define nfs4_pathconf_res] Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 5 ++++- fs/nfs/nfs4xdr.c | 5 +++-- include/linux/nfs_xdr.h | 5 +++++ 3 files changed, 12 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index b8915ef533da..aea2e83d3939 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -2496,10 +2496,13 @@ static int _nfs4_proc_pathconf(struct nfs_server *server, struct nfs_fh *fhandle .fh = fhandle, .bitmask = server->attr_bitmask, }; + struct nfs4_pathconf_res res = { + .pathconf = pathconf, + }; struct rpc_message msg = { .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_PATHCONF], .rpc_argp = &args, - .rpc_resp = pathconf, + .rpc_resp = &res, }; /* None of the pathconf attributes are mandatory to implement */ diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index d9ab8209c286..a77ee3dd0b3d 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -4370,7 +4370,8 @@ static int nfs4_xdr_dec_fsinfo(struct rpc_rqst *req, __be32 *p, /* * PATHCONF request */ -static int nfs4_xdr_dec_pathconf(struct rpc_rqst *req, __be32 *p, struct nfs_pathconf *pathconf) +static int nfs4_xdr_dec_pathconf(struct rpc_rqst *req, __be32 *p, + struct nfs4_pathconf_res *res) { struct xdr_stream xdr; struct compound_hdr hdr; @@ -4381,7 +4382,7 @@ static int nfs4_xdr_dec_pathconf(struct rpc_rqst *req, __be32 *p, struct nfs_pat if (!status) status = decode_putfh(&xdr); if (!status) - status = decode_pathconf(&xdr, pathconf); + status = decode_pathconf(&xdr, res->pathconf); return status; } diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h index 7d64913cbb1b..56523319e14c 100644 --- a/include/linux/nfs_xdr.h +++ b/include/linux/nfs_xdr.h @@ -734,6 +734,11 @@ struct nfs4_pathconf_arg { struct nfs4_sequence_args seq_args; }; +struct nfs4_pathconf_res { + struct nfs_pathconf *pathconf; + struct nfs4_sequence_res seq_res; +}; + struct nfs4_readdir_arg { const struct nfs_fh * fh; u64 cookie; -- cgit v1.2.3 From 663c79b3cd8f5fe21fe7d7565fec0072e3234ddc Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Wed, 1 Apr 2009 09:21:59 -0400 Subject: nfs41: use nfs4_getaclres In preparation for nfs41 sequence processing. Signed-off-by: Andy Admason Signed-off-by: Benny Halevy [nfs41: embed resp_len in nfs_getaclres] Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 20 +++++++++++--------- fs/nfs/nfs4xdr.c | 5 +++-- include/linux/nfs_xdr.h | 5 +++++ 3 files changed, 19 insertions(+), 11 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index aea2e83d3939..20c9acf689fd 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -2755,12 +2755,14 @@ static ssize_t __nfs4_get_acl_uncached(struct inode *inode, void *buf, size_t bu .acl_pages = pages, .acl_len = buflen, }; - size_t resp_len = buflen; + struct nfs_getaclres res = { + .acl_len = buflen, + }; void *resp_buf; struct rpc_message msg = { .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_GETACL], .rpc_argp = &args, - .rpc_resp = &resp_len, + .rpc_resp = &res, }; struct page *localpage = NULL; int ret; @@ -2774,7 +2776,7 @@ static ssize_t __nfs4_get_acl_uncached(struct inode *inode, void *buf, size_t bu return -ENOMEM; args.acl_pages[0] = localpage; args.acl_pgbase = 0; - resp_len = args.acl_len = PAGE_SIZE; + args.acl_len = PAGE_SIZE; } else { resp_buf = buf; buf_to_pages(buf, buflen, args.acl_pages, &args.acl_pgbase); @@ -2782,18 +2784,18 @@ static ssize_t __nfs4_get_acl_uncached(struct inode *inode, void *buf, size_t bu ret = rpc_call_sync(NFS_CLIENT(inode), &msg, 0); if (ret) goto out_free; - if (resp_len > args.acl_len) - nfs4_write_cached_acl(inode, NULL, resp_len); + if (res.acl_len > args.acl_len) + nfs4_write_cached_acl(inode, NULL, res.acl_len); else - nfs4_write_cached_acl(inode, resp_buf, resp_len); + nfs4_write_cached_acl(inode, resp_buf, res.acl_len); if (buf) { ret = -ERANGE; - if (resp_len > buflen) + if (res.acl_len > buflen) goto out_free; if (localpage) - memcpy(buf, resp_buf, resp_len); + memcpy(buf, resp_buf, res.acl_len); } - ret = resp_len; + ret = res.acl_len; out_free: if (localpage) __free_page(localpage); diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index a77ee3dd0b3d..3e777893e2b0 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -4018,7 +4018,8 @@ out: * Decode GETACL response */ static int -nfs4_xdr_dec_getacl(struct rpc_rqst *rqstp, __be32 *p, size_t *acl_len) +nfs4_xdr_dec_getacl(struct rpc_rqst *rqstp, __be32 *p, + struct nfs_getaclres *res) { struct xdr_stream xdr; struct compound_hdr hdr; @@ -4031,7 +4032,7 @@ nfs4_xdr_dec_getacl(struct rpc_rqst *rqstp, __be32 *p, size_t *acl_len) status = decode_putfh(&xdr); if (status) goto out; - status = decode_getacl(&xdr, rqstp, acl_len); + status = decode_getacl(&xdr, rqstp, &res->acl_len); out: return status; diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h index 56523319e14c..6e9ee2848606 100644 --- a/include/linux/nfs_xdr.h +++ b/include/linux/nfs_xdr.h @@ -444,6 +444,11 @@ struct nfs_getaclargs { struct nfs4_sequence_args seq_args; }; +struct nfs_getaclres { + size_t acl_len; + struct nfs4_sequence_res seq_res; +}; + struct nfs_setattrres { struct nfs_fattr * fattr; const struct nfs_server * server; -- cgit v1.2.3 From 9e9ecc03d6b37a5d85e16b357751e3b341de0fef Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Wed, 1 Apr 2009 09:22:00 -0400 Subject: NFS: get rid of unused xdr decode_setattr(, res) argument Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4xdr.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index 3e777893e2b0..27dd25d9ad42 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -3658,7 +3658,7 @@ decode_savefh(struct xdr_stream *xdr) return decode_op_hdr(xdr, OP_SAVEFH); } -static int decode_setattr(struct xdr_stream *xdr, struct nfs_setattrres *res) +static int decode_setattr(struct xdr_stream *xdr) { __be32 *p; uint32_t bmlen; @@ -4009,7 +4009,7 @@ nfs4_xdr_dec_setacl(struct rpc_rqst *rqstp, __be32 *p, void *res) status = decode_putfh(&xdr); if (status) goto out; - status = decode_setattr(&xdr, res); + status = decode_setattr(&xdr); out: return status; } @@ -4162,7 +4162,7 @@ static int nfs4_xdr_dec_setattr(struct rpc_rqst *rqstp, __be32 *p, struct nfs_se status = decode_putfh(&xdr); if (status) goto out; - status = decode_setattr(&xdr, res); + status = decode_setattr(&xdr); if (status) goto out; decode_getfattr(&xdr, res->fattr, res->server); -- cgit v1.2.3 From 73c403a9a93743b068103c13c05ed136dc687d05 Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Wed, 1 Apr 2009 09:22:01 -0400 Subject: nfs41: use nfs4_setaclres In preparation for nfs41 sequence processing. Signed-off-by: Andy Admason [define nfs_setaclres] Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 3 ++- fs/nfs/nfs4xdr.c | 3 ++- include/linux/nfs_xdr.h | 4 ++++ 3 files changed, 8 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 20c9acf689fd..62bbe25d9423 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -2842,10 +2842,11 @@ static int __nfs4_proc_set_acl(struct inode *inode, const void *buf, size_t bufl .acl_pages = pages, .acl_len = buflen, }; + struct nfs_setaclres res; struct rpc_message msg = { .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_SETACL], .rpc_argp = &arg, - .rpc_resp = NULL, + .rpc_resp = &res, }; int ret; diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index 27dd25d9ad42..aa350d5bf207 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -3996,7 +3996,8 @@ nfs4_xdr_enc_setacl(struct rpc_rqst *req, __be32 *p, struct nfs_setaclargs *args * Decode SETACL response */ static int -nfs4_xdr_dec_setacl(struct rpc_rqst *rqstp, __be32 *p, void *res) +nfs4_xdr_dec_setacl(struct rpc_rqst *rqstp, __be32 *p, + struct nfs_setaclres *res) { struct xdr_stream xdr; struct compound_hdr hdr; diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h index 6e9ee2848606..0f2dc8f4cc36 100644 --- a/include/linux/nfs_xdr.h +++ b/include/linux/nfs_xdr.h @@ -436,6 +436,10 @@ struct nfs_setaclargs { struct nfs4_sequence_args seq_args; }; +struct nfs_setaclres { + struct nfs4_sequence_res seq_res; +}; + struct nfs_getaclargs { struct nfs_fh * fh; size_t acl_len; -- cgit v1.2.3 From 22958463d5dca8548e19430779f379e66fd6e4a4 Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Wed, 1 Apr 2009 09:22:02 -0400 Subject: nfs41: use nfs4_fs_locations_res In preparation for nfs41 sequence processing. Signed-off-by: Andy Admason [find nfs4_fs_locations_res] Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 5 ++++- fs/nfs/nfs4xdr.c | 6 ++++-- include/linux/nfs_xdr.h | 5 +++++ 3 files changed, 13 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 62bbe25d9423..e08edc99faac 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -3722,10 +3722,13 @@ int nfs4_proc_fs_locations(struct inode *dir, const struct qstr *name, .page = page, .bitmask = bitmask, }; + struct nfs4_fs_locations_res res = { + .fs_locations = fs_locations, + }; struct rpc_message msg = { .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_FS_LOCATIONS], .rpc_argp = &args, - .rpc_resp = fs_locations, + .rpc_resp = &res, }; int status; diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index aa350d5bf207..e448e33b4d05 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -4504,7 +4504,8 @@ out: /* * FS_LOCATIONS request */ -static int nfs4_xdr_dec_fs_locations(struct rpc_rqst *req, __be32 *p, struct nfs4_fs_locations *res) +static int nfs4_xdr_dec_fs_locations(struct rpc_rqst *req, __be32 *p, + struct nfs4_fs_locations_res *res) { struct xdr_stream xdr; struct compound_hdr hdr; @@ -4519,7 +4520,8 @@ static int nfs4_xdr_dec_fs_locations(struct rpc_rqst *req, __be32 *p, struct nfs if ((status = decode_lookup(&xdr)) != 0) goto out; xdr_enter_page(&xdr, PAGE_SIZE); - status = decode_getfattr(&xdr, &res->fattr, res->server); + status = decode_getfattr(&xdr, &res->fs_locations->fattr, + res->fs_locations->server); out: return status; } diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h index 0f2dc8f4cc36..d837f10c49ef 100644 --- a/include/linux/nfs_xdr.h +++ b/include/linux/nfs_xdr.h @@ -867,6 +867,11 @@ struct nfs4_fs_locations_arg { struct nfs4_sequence_args seq_args; }; +struct nfs4_fs_locations_res { + struct nfs4_fs_locations *fs_locations; + struct nfs4_sequence_res seq_res; +}; + #endif /* CONFIG_NFS_V4 */ struct nfs_page; -- cgit v1.2.3 From cccef3b96a4759ae0790452280c00ea505412157 Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Wed, 1 Apr 2009 09:22:03 -0400 Subject: nfs41: introduce nfs4_call_sync Use nfs4_call_sync rather than rpc_call_sync to provide for a nfs41 sessions-enabled interface for sessions manipulation. The nfs41 rpc logic uses the rpc_call_prepare method to recover and create the session, as well as selecting a free slot id and the rpc_call_done to free the slot and update slot table related metadata. In the coming patches we'll add rpc prepare and done routines for setting up the sequence op and processing the sequence result. Signed-off-by: Benny Halevy [nfs41: nfs4_call_sync] As per 11-14-08 review. Squash into "nfs41: introduce nfs4_call_sync" and "nfs41: nfs4_setup_sequence" Define two functions one for v4 and one for v41 add a pointer to struct nfs4_client to the correct one. Signed-off-by: Andy Adamson [added BUG() in _nfs4_call_sync_session if !CONFIG_NFS_V4_1] Signed-off-by: Benny Halevy [nfs41: check for session not minorversion] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [group minorversion specific stuff together] Signed-off-by: Alexandros Batsakis Signed-off-by: Benny Halevy Signed-off-by: Andy Adamson [nfs41: fixup nfs4_clear_client_minor_version] [introduce nfs4_init_client_minor_version() in this patch] Signed-off-by: Benny Halevy [cleaned-up patch: got rid of nfs_call_sync_t, dprintks, cosmetics, extra server defs] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/client.c | 5 ++++ fs/nfs/internal.h | 12 +++++++++ fs/nfs/nfs4proc.c | 67 +++++++++++++++++++++++++++++++++-------------- include/linux/nfs_fs_sb.h | 8 ++++++ 4 files changed, 73 insertions(+), 19 deletions(-) (limited to 'fs') diff --git a/fs/nfs/client.c b/fs/nfs/client.c index f1506f148521..a9828baaa445 100644 --- a/fs/nfs/client.c +++ b/fs/nfs/client.c @@ -194,6 +194,8 @@ static void nfs4_clear_client_minor_version(struct nfs_client *clp) nfs4_destroy_session(clp->cl_session); clp->cl_session = NULL; } + + clp->cl_call_sync = _nfs4_call_sync; #endif /* CONFIG_NFS_V4_1 */ } @@ -1073,6 +1075,8 @@ error: */ static int nfs4_init_client_minor_version(struct nfs_client *clp) { + clp->cl_call_sync = _nfs4_call_sync; + #if defined(CONFIG_NFS_V4_1) if (clp->cl_minorversion) { struct nfs4_session *session = NULL; @@ -1086,6 +1090,7 @@ static int nfs4_init_client_minor_version(struct nfs_client *clp) return -ENOMEM; clp->cl_session = session; + clp->cl_call_sync = _nfs4_call_sync_session; } #endif /* CONFIG_NFS_V4_1 */ diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h index 7cef45db9257..8d67c2865dc3 100644 --- a/fs/nfs/internal.h +++ b/fs/nfs/internal.h @@ -206,6 +206,18 @@ extern int nfs4_path_walk(struct nfs_server *server, const char *path); #endif +/* nfs4proc.c */ +extern int _nfs4_call_sync(struct nfs_server *server, + struct rpc_message *msg, + struct nfs4_sequence_args *args, + struct nfs4_sequence_res *res, + int cache_reply); +extern int _nfs4_call_sync_session(struct nfs_server *server, + struct rpc_message *msg, + struct nfs4_sequence_args *args, + struct nfs4_sequence_res *res, + int cache_reply); + /* * Determine if sessions are in use. */ diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index e08edc99faac..4fc5b385f61e 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -271,6 +271,33 @@ static void renew_lease(const struct nfs_server *server, unsigned long timestamp spin_unlock(&clp->cl_lock); } +#if defined(CONFIG_NFS_V4_1) + +int _nfs4_call_sync_session(struct nfs_server *server, + struct rpc_message *msg, + struct nfs4_sequence_args *args, + struct nfs4_sequence_res *res, + int cache_reply) +{ + /* in preparation for setting up the sequence op */ + return rpc_call_sync(server->client, msg, 0); +} + +#endif /* CONFIG_NFS_V4_1 */ + +int _nfs4_call_sync(struct nfs_server *server, + struct rpc_message *msg, + struct nfs4_sequence_args *args, + struct nfs4_sequence_res *res, + int cache_reply) +{ + return rpc_call_sync(server->client, msg, 0); +} + +#define nfs4_call_sync(server, msg, args, res, cache_reply) \ + (server)->nfs_client->cl_call_sync((server), (msg), &(args)->seq_args, \ + &(res)->seq_res, (cache_reply)) + static void update_changeattr(struct inode *dir, struct nfs4_change_info *cinfo) { struct nfs_inode *nfsi = NFS_I(dir); @@ -1269,7 +1296,7 @@ static int _nfs4_do_setattr(struct inode *inode, struct rpc_cred *cred, } else memcpy(&arg.stateid, &zero_stateid, sizeof(arg.stateid)); - status = rpc_call_sync(server->client, &msg, 0); + status = nfs4_call_sync(server, &msg, &arg, &res, 1); if (status == 0 && state != NULL) renew_lease(server, timestamp); return status; @@ -1595,7 +1622,7 @@ static int _nfs4_server_capabilities(struct nfs_server *server, struct nfs_fh *f }; int status; - status = rpc_call_sync(server->client, &msg, 0); + status = nfs4_call_sync(server, &msg, &args, &res, 0); if (status == 0) { memcpy(server->attr_bitmask, res.attr_bitmask, sizeof(server->attr_bitmask)); if (res.attr_bitmask[0] & FATTR4_WORD0_ACL) @@ -1609,6 +1636,7 @@ static int _nfs4_server_capabilities(struct nfs_server *server, struct nfs_fh *f server->cache_consistency_bitmask[1] &= FATTR4_WORD1_TIME_METADATA|FATTR4_WORD1_TIME_MODIFY; server->acl_bitmask = res.acl_bitmask; } + return status; } @@ -1641,7 +1669,7 @@ static int _nfs4_lookup_root(struct nfs_server *server, struct nfs_fh *fhandle, .rpc_resp = &res, }; nfs_fattr_init(info->fattr); - return rpc_call_sync(server->client, &msg, 0); + return nfs4_call_sync(server, &msg, &args, &res, 0); } static int nfs4_lookup_root(struct nfs_server *server, struct nfs_fh *fhandle, @@ -1731,7 +1759,7 @@ static int _nfs4_proc_getattr(struct nfs_server *server, struct nfs_fh *fhandle, }; nfs_fattr_init(fattr); - return rpc_call_sync(server->client, &msg, 0); + return nfs4_call_sync(server, &msg, &args, &res, 0); } static int nfs4_proc_getattr(struct nfs_server *server, struct nfs_fh *fhandle, struct nfs_fattr *fattr) @@ -1815,7 +1843,7 @@ static int _nfs4_proc_lookupfh(struct nfs_server *server, const struct nfs_fh *d nfs_fattr_init(fattr); dprintk("NFS call lookupfh %s\n", name->name); - status = rpc_call_sync(server->client, &msg, 0); + status = nfs4_call_sync(server, &msg, &args, &res, 0); dprintk("NFS reply lookupfh: %d\n", status); return status; } @@ -1901,7 +1929,7 @@ static int _nfs4_proc_access(struct inode *inode, struct nfs_access_entry *entry args.access |= NFS4_ACCESS_EXECUTE; } nfs_fattr_init(&fattr); - status = rpc_call_sync(NFS_CLIENT(inode), &msg, 0); + status = nfs4_call_sync(server, &msg, &args, &res, 0); if (!status) { entry->mask = 0; if (res.access & NFS4_ACCESS_READ) @@ -1967,7 +1995,7 @@ static int _nfs4_proc_readlink(struct inode *inode, struct page *page, .rpc_resp = &res, }; - return rpc_call_sync(NFS_CLIENT(inode), &msg, 0); + return nfs4_call_sync(NFS_SERVER(inode), &msg, &args, &res, 0); } static int nfs4_proc_readlink(struct inode *inode, struct page *page, @@ -2061,7 +2089,7 @@ static int _nfs4_proc_remove(struct inode *dir, struct qstr *name) int status; nfs_fattr_init(&res.dir_attr); - status = rpc_call_sync(server->client, &msg, 0); + status = nfs4_call_sync(server, &msg, &args, &res, 1); if (status == 0) { update_changeattr(dir, &res.cinfo); nfs_post_op_update_inode(dir, &res.dir_attr); @@ -2129,7 +2157,7 @@ static int _nfs4_proc_rename(struct inode *old_dir, struct qstr *old_name, nfs_fattr_init(res.old_fattr); nfs_fattr_init(res.new_fattr); - status = rpc_call_sync(server->client, &msg, 0); + status = nfs4_call_sync(server, &msg, &arg, &res, 1); if (!status) { update_changeattr(old_dir, &res.old_cinfo); @@ -2178,7 +2206,7 @@ static int _nfs4_proc_link(struct inode *inode, struct inode *dir, struct qstr * nfs_fattr_init(res.fattr); nfs_fattr_init(res.dir_attr); - status = rpc_call_sync(server->client, &msg, 0); + status = nfs4_call_sync(server, &msg, &arg, &res, 1); if (!status) { update_changeattr(dir, &res.cinfo); nfs_post_op_update_inode(dir, res.dir_attr); @@ -2239,7 +2267,8 @@ static struct nfs4_createdata *nfs4_alloc_createdata(struct inode *dir, static int nfs4_do_create(struct inode *dir, struct dentry *dentry, struct nfs4_createdata *data) { - int status = rpc_call_sync(NFS_CLIENT(dir), &data->msg, 0); + int status = nfs4_call_sync(NFS_SERVER(dir), &data->msg, + &data->arg, &data->res, 1); if (status == 0) { update_changeattr(dir, &data->res.dir_cinfo); nfs_post_op_update_inode(dir, data->res.dir_fattr); @@ -2348,7 +2377,7 @@ static int _nfs4_proc_readdir(struct dentry *dentry, struct rpc_cred *cred, (unsigned long long)cookie); nfs4_setup_readdir(cookie, NFS_COOKIEVERF(dir), dentry, &args); res.pgbase = args.pgbase; - status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0); + status = nfs4_call_sync(NFS_SERVER(dir), &msg, &args, &res, 0); if (status == 0) memcpy(NFS_COOKIEVERF(dir), res.verifier.data, NFS4_VERIFIER_SIZE); @@ -2436,7 +2465,7 @@ static int _nfs4_proc_statfs(struct nfs_server *server, struct nfs_fh *fhandle, }; nfs_fattr_init(fsstat->fattr); - return rpc_call_sync(server->client, &msg, 0); + return nfs4_call_sync(server, &msg, &args, &res, 0); } static int nfs4_proc_statfs(struct nfs_server *server, struct nfs_fh *fhandle, struct nfs_fsstat *fsstat) @@ -2467,7 +2496,7 @@ static int _nfs4_do_fsinfo(struct nfs_server *server, struct nfs_fh *fhandle, .rpc_resp = &res, }; - return rpc_call_sync(server->client, &msg, 0); + return nfs4_call_sync(server, &msg, &args, &res, 0); } static int nfs4_do_fsinfo(struct nfs_server *server, struct nfs_fh *fhandle, struct nfs_fsinfo *fsinfo) @@ -2512,7 +2541,7 @@ static int _nfs4_proc_pathconf(struct nfs_server *server, struct nfs_fh *fhandle } nfs_fattr_init(pathconf->fattr); - return rpc_call_sync(server->client, &msg, 0); + return nfs4_call_sync(server, &msg, &args, &res, 0); } static int nfs4_proc_pathconf(struct nfs_server *server, struct nfs_fh *fhandle, @@ -2781,7 +2810,7 @@ static ssize_t __nfs4_get_acl_uncached(struct inode *inode, void *buf, size_t bu resp_buf = buf; buf_to_pages(buf, buflen, args.acl_pages, &args.acl_pgbase); } - ret = rpc_call_sync(NFS_CLIENT(inode), &msg, 0); + ret = nfs4_call_sync(NFS_SERVER(inode), &msg, &args, &res, 0); if (ret) goto out_free; if (res.acl_len > args.acl_len) @@ -2854,7 +2883,7 @@ static int __nfs4_proc_set_acl(struct inode *inode, const void *buf, size_t bufl return -EOPNOTSUPP; nfs_inode_return_delegation(inode); buf_to_pages(buf, buflen, arg.acl_pages, &arg.acl_pgbase); - ret = rpc_call_sync(NFS_CLIENT(inode), &msg, 0); + ret = nfs4_call_sync(server, &msg, &arg, &res, 1); nfs_access_zap_cache(inode); nfs_zap_acl_cache(inode); return ret; @@ -3143,7 +3172,7 @@ static int _nfs4_proc_getlk(struct nfs4_state *state, int cmd, struct file_lock goto out; lsp = request->fl_u.nfs4_fl.owner; arg.lock_owner.id = lsp->ls_id.id; - status = rpc_call_sync(server->client, &msg, 0); + status = nfs4_call_sync(server, &msg, &arg, &res, 1); switch (status) { case 0: request->fl_type = F_UNLCK; @@ -3736,7 +3765,7 @@ int nfs4_proc_fs_locations(struct inode *dir, const struct qstr *name, nfs_fattr_init(&fs_locations->fattr); fs_locations->server = server; fs_locations->nlocations = 0; - status = rpc_call_sync(server->client, &msg, 0); + status = nfs4_call_sync(server, &msg, &args, &res, 0); nfs_fixup_referral_attributes(&fs_locations->fattr); dprintk("%s: returned status = %d\n", __func__, status); return status; diff --git a/include/linux/nfs_fs_sb.h b/include/linux/nfs_fs_sb.h index b47c0fc55d42..206485e5082f 100644 --- a/include/linux/nfs_fs_sb.h +++ b/include/linux/nfs_fs_sb.h @@ -12,6 +12,9 @@ struct nfs4_session; struct nfs_iostats; struct nlm_host; +struct nfs4_sequence_args; +struct nfs4_sequence_res; +struct nfs_server; /* * The nfs_client identifies our client state to the server. @@ -67,6 +70,11 @@ struct nfs_client { */ char cl_ipaddr[48]; unsigned char cl_id_uniquifier; + int (* cl_call_sync)(struct nfs_server *server, + struct rpc_message *msg, + struct nfs4_sequence_args *args, + struct nfs4_sequence_res *res, + int cache_reply); #endif /* CONFIG_NFS_V4 */ #ifdef CONFIG_NFS_V4_1 -- cgit v1.2.3 From f3752975caa716709c5ea0b0820b86111d921df4 Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Wed, 1 Apr 2009 09:22:04 -0400 Subject: nfs41: nfs41: pass *session in seq_args and seq_res To be used for getting the rpc's minorversion and for nfs41 xdr {en,de}coding of the sequence operation. Reset the seq session ptrs for minorversion=0 rpc calls. Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 1 + include/linux/nfs_xdr.h | 4 ++-- 2 files changed, 3 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 4fc5b385f61e..49cf969417c0 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -291,6 +291,7 @@ int _nfs4_call_sync(struct nfs_server *server, struct nfs4_sequence_res *res, int cache_reply) { + args->sa_session = res->sr_session = NULL; return rpc_call_sync(server->client, msg, 0); } diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h index d837f10c49ef..f5675063f951 100644 --- a/include/linux/nfs_xdr.h +++ b/include/linux/nfs_xdr.h @@ -162,11 +162,11 @@ struct nfs4_slot { }; struct nfs4_sequence_args { - /* stub */ + struct nfs4_session *sa_session; }; struct nfs4_sequence_res { - /* stub */ + struct nfs4_session *sr_session; }; /* -- cgit v1.2.3 From 5f7dbd5c752d88310d8fe1feedefd5c6496eff48 Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Wed, 1 Apr 2009 09:22:05 -0400 Subject: nfs41: set up seq_res.sr_slotid Initialize nfs4_sequence_res sr_slotid to NFS4_MAX_SLOT_TABLE. [was nfs41: sequence res use slotid] Signed-off-by: Andy Adamson [pulled definition of struct nfs4_sequence_res.sr_slotid to here] Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 5 +++++ fs/nfs/read.c | 1 + fs/nfs/unlink.c | 1 + fs/nfs/write.c | 2 ++ include/linux/nfs_xdr.h | 1 + 5 files changed, 10 insertions(+) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 49cf969417c0..5878930b4c3b 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -371,6 +371,7 @@ static struct nfs4_opendata *nfs4_opendata_alloc(struct path *path, p->o_arg.server = server; p->o_arg.bitmask = server->attr_bitmask; p->o_arg.claim = NFS4_OPEN_CLAIM_NULL; + p->o_res.seq_res.sr_slotid = NFS4_MAX_SLOT_TABLE; if (flags & O_EXCL) { u32 *s = (u32 *) p->o_arg.u.verifier.data; s[0] = jiffies; @@ -1463,6 +1464,7 @@ int nfs4_do_close(struct path *path, struct nfs4_state *state, int wait) calldata->res.fattr = &calldata->fattr; calldata->res.seqid = calldata->arg.seqid; calldata->res.server = server; + calldata->res.seq_res.sr_slotid = NFS4_MAX_SLOT_TABLE; calldata->path.mnt = mntget(path->mnt); calldata->path.dentry = dget(path->dentry); @@ -3088,6 +3090,7 @@ static int _nfs4_proc_delegreturn(struct inode *inode, struct rpc_cred *cred, co memcpy(&data->stateid, stateid, sizeof(data->stateid)); data->res.fattr = &data->fattr; data->res.server = server; + data->res.seq_res.sr_slotid = NFS4_MAX_SLOT_TABLE; nfs_fattr_init(data->res.fattr); data->timestamp = jiffies; data->rpc_status = 0; @@ -3240,6 +3243,7 @@ static struct nfs4_unlockdata *nfs4_alloc_unlockdata(struct file_lock *fl, p->arg.fl = &p->fl; p->arg.seqid = seqid; p->res.seqid = seqid; + p->res.seq_res.sr_slotid = NFS4_MAX_SLOT_TABLE; p->arg.stateid = &lsp->ls_stateid; p->lsp = lsp; atomic_inc(&lsp->ls_count); @@ -3412,6 +3416,7 @@ static struct nfs4_lockdata *nfs4_alloc_lockdata(struct file_lock *fl, p->arg.lock_owner.clientid = server->nfs_client->cl_clientid; p->arg.lock_owner.id = lsp->ls_id.id; p->res.lock_seqid = p->arg.lock_seqid; + p->res.seq_res.sr_slotid = NFS4_MAX_SLOT_TABLE; p->lsp = lsp; atomic_inc(&lsp->ls_count); p->ctx = get_nfs_open_context(ctx); diff --git a/fs/nfs/read.c b/fs/nfs/read.c index 4ace3c50a8eb..70ba2b4cb9a4 100644 --- a/fs/nfs/read.c +++ b/fs/nfs/read.c @@ -46,6 +46,7 @@ struct nfs_read_data *nfs_readdata_alloc(unsigned int pagecount) memset(p, 0, sizeof(*p)); INIT_LIST_HEAD(&p->pages); p->npages = pagecount; + p->res.seq_res.sr_slotid = NFS4_MAX_SLOT_TABLE; if (pagecount <= ARRAY_SIZE(p->page_array)) p->pagevec = p->page_array; else { diff --git a/fs/nfs/unlink.c b/fs/nfs/unlink.c index ecc295347775..83d0ce2600ab 100644 --- a/fs/nfs/unlink.c +++ b/fs/nfs/unlink.c @@ -241,6 +241,7 @@ nfs_async_unlink(struct inode *dir, struct dentry *dentry) status = PTR_ERR(data->cred); goto out_free; } + data->res.seq_res.sr_slotid = NFS4_MAX_SLOT_TABLE; status = -EBUSY; spin_lock(&dentry->d_lock); diff --git a/fs/nfs/write.c b/fs/nfs/write.c index e560a78995a3..035e6fb9f57e 100644 --- a/fs/nfs/write.c +++ b/fs/nfs/write.c @@ -52,6 +52,7 @@ struct nfs_write_data *nfs_commitdata_alloc(void) if (p) { memset(p, 0, sizeof(*p)); INIT_LIST_HEAD(&p->pages); + p->res.seq_res.sr_slotid = NFS4_MAX_SLOT_TABLE; } return p; } @@ -71,6 +72,7 @@ struct nfs_write_data *nfs_writedata_alloc(unsigned int pagecount) memset(p, 0, sizeof(*p)); INIT_LIST_HEAD(&p->pages); p->npages = pagecount; + p->res.seq_res.sr_slotid = NFS4_MAX_SLOT_TABLE; if (pagecount <= ARRAY_SIZE(p->page_array)) p->pagevec = p->page_array; else { diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h index f5675063f951..db0d1236aae7 100644 --- a/include/linux/nfs_xdr.h +++ b/include/linux/nfs_xdr.h @@ -167,6 +167,7 @@ struct nfs4_sequence_args { struct nfs4_sequence_res { struct nfs4_session *sr_session; + u8 sr_slotid; /* slot used to send request */ }; /* -- cgit v1.2.3 From 6ce183919b4a09289cb0fe4fce960a9faa1e7c6b Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Wed, 1 Apr 2009 09:22:06 -0400 Subject: NFS: use decode_change_info_maxsz for xdr maxsz calculations Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4xdr.c | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index e448e33b4d05..487a197231fd 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -192,12 +192,16 @@ static int nfs4_stat_to_errno(int); decode_verifier_maxsz) #define encode_remove_maxsz (op_encode_hdr_maxsz + \ nfs4_name_maxsz) +#define decode_remove_maxsz (op_decode_hdr_maxsz + \ + decode_change_info_maxsz) #define encode_rename_maxsz (op_encode_hdr_maxsz + \ 2 * nfs4_name_maxsz) -#define decode_rename_maxsz (op_decode_hdr_maxsz + 5 + 5) +#define decode_rename_maxsz (op_decode_hdr_maxsz + \ + decode_change_info_maxsz + \ + decode_change_info_maxsz) #define encode_link_maxsz (op_encode_hdr_maxsz + \ nfs4_name_maxsz) -#define decode_link_maxsz (op_decode_hdr_maxsz + 5) +#define decode_link_maxsz (op_decode_hdr_maxsz + decode_change_info_maxsz) #define encode_lock_maxsz (op_encode_hdr_maxsz + \ 7 + \ 1 + encode_stateid_maxsz + 8) @@ -414,7 +418,7 @@ static int nfs4_stat_to_errno(int); encode_getattr_maxsz) #define NFS4_dec_remove_sz (compound_decode_hdr_maxsz + \ decode_putfh_maxsz + \ - op_decode_hdr_maxsz + 5 + \ + decode_remove_maxsz + \ decode_getattr_maxsz) #define NFS4_enc_rename_sz (compound_encode_hdr_maxsz + \ encode_putfh_maxsz + \ -- cgit v1.2.3 From 0c4e8c187758258ec58842384fe6a99cf1ce16c7 Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Wed, 1 Apr 2009 09:22:07 -0400 Subject: NFS: define and initialize compound_hdr.replen replen holds the running count of expected reply bytes. repl will then be used by encoding routines for xdr_inline_pages offset after which data bytes are to be received directly into the xdr buffer pages. NOTE: According to the nfsv4 and v4.1 RFCs, the replied tag SHOULD be the same is the one sent, but this is not required as a MUST for the server to do so. The server may screw us if it replies a tag of a different length in the compound result. [NFS: cb_compoundhdr.replen is in words not bytes] Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4xdr.c | 77 +++++++++++++++++++++++++++++++------------------------- 1 file changed, 43 insertions(+), 34 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index 487a197231fd..c85dbee34b8a 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -545,6 +545,7 @@ struct compound_hdr { __be32 * nops_p; uint32_t taglen; char * tag; + uint32_t replen; /* expected reply words */ }; /* @@ -580,9 +581,17 @@ static void encode_string(struct xdr_stream *xdr, unsigned int len, const char * xdr_encode_opaque(p, str, len); } -static void encode_compound_hdr(struct xdr_stream *xdr, struct compound_hdr *hdr) +static void encode_compound_hdr(struct xdr_stream *xdr, + struct rpc_rqst *req, + struct compound_hdr *hdr) { __be32 *p; + struct rpc_auth *auth = req->rq_task->tk_msg.rpc_cred->cr_auth; + + /* initialize running count of expected bytes in reply. + * NOTE: the replied tag SHOULD be the same is the one sent, + * but this is not required as a MUST for the server to do so. */ + hdr->replen = RPC_REPHDRSIZE + auth->au_rslack + 3 + hdr->taglen; dprintk("encode_compound: tag=%.*s\n", (int)hdr->taglen, hdr->tag); BUG_ON(hdr->taglen > NFS4_MAXTAGLEN); @@ -1359,7 +1368,7 @@ static int nfs4_xdr_enc_access(struct rpc_rqst *req, __be32 *p, const struct nfs }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); - encode_compound_hdr(&xdr, &hdr); + encode_compound_hdr(&xdr, req, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_access(&xdr, args->access, &hdr); encode_getfattr(&xdr, args->bitmask, &hdr); @@ -1378,7 +1387,7 @@ static int nfs4_xdr_enc_lookup(struct rpc_rqst *req, __be32 *p, const struct nfs }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); - encode_compound_hdr(&xdr, &hdr); + encode_compound_hdr(&xdr, req, &hdr); encode_putfh(&xdr, args->dir_fh, &hdr); encode_lookup(&xdr, args->name, &hdr); encode_getfh(&xdr, &hdr); @@ -1398,7 +1407,7 @@ static int nfs4_xdr_enc_lookup_root(struct rpc_rqst *req, __be32 *p, const struc }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); - encode_compound_hdr(&xdr, &hdr); + encode_compound_hdr(&xdr, req, &hdr); encode_putrootfh(&xdr, &hdr); encode_getfh(&xdr, &hdr); encode_getfattr(&xdr, args->bitmask, &hdr); @@ -1417,7 +1426,7 @@ static int nfs4_xdr_enc_remove(struct rpc_rqst *req, __be32 *p, const struct nfs }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); - encode_compound_hdr(&xdr, &hdr); + encode_compound_hdr(&xdr, req, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_remove(&xdr, &args->name, &hdr); encode_getfattr(&xdr, args->bitmask, &hdr); @@ -1436,7 +1445,7 @@ static int nfs4_xdr_enc_rename(struct rpc_rqst *req, __be32 *p, const struct nfs }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); - encode_compound_hdr(&xdr, &hdr); + encode_compound_hdr(&xdr, req, &hdr); encode_putfh(&xdr, args->old_dir, &hdr); encode_savefh(&xdr, &hdr); encode_putfh(&xdr, args->new_dir, &hdr); @@ -1459,7 +1468,7 @@ static int nfs4_xdr_enc_link(struct rpc_rqst *req, __be32 *p, const struct nfs4_ }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); - encode_compound_hdr(&xdr, &hdr); + encode_compound_hdr(&xdr, req, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_savefh(&xdr, &hdr); encode_putfh(&xdr, args->dir_fh, &hdr); @@ -1482,7 +1491,7 @@ static int nfs4_xdr_enc_create(struct rpc_rqst *req, __be32 *p, const struct nfs }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); - encode_compound_hdr(&xdr, &hdr); + encode_compound_hdr(&xdr, req, &hdr); encode_putfh(&xdr, args->dir_fh, &hdr); encode_savefh(&xdr, &hdr); encode_create(&xdr, args, &hdr); @@ -1513,7 +1522,7 @@ static int nfs4_xdr_enc_getattr(struct rpc_rqst *req, __be32 *p, const struct nf }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); - encode_compound_hdr(&xdr, &hdr); + encode_compound_hdr(&xdr, req, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_getfattr(&xdr, args->bitmask, &hdr); encode_nops(&hdr); @@ -1531,7 +1540,7 @@ static int nfs4_xdr_enc_close(struct rpc_rqst *req, __be32 *p, struct nfs_closea }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); - encode_compound_hdr(&xdr, &hdr); + encode_compound_hdr(&xdr, req, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_close(&xdr, args, &hdr); encode_getfattr(&xdr, args->bitmask, &hdr); @@ -1550,7 +1559,7 @@ static int nfs4_xdr_enc_open(struct rpc_rqst *req, __be32 *p, struct nfs_openarg }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); - encode_compound_hdr(&xdr, &hdr); + encode_compound_hdr(&xdr, req, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_savefh(&xdr, &hdr); encode_open(&xdr, args, &hdr); @@ -1573,7 +1582,7 @@ static int nfs4_xdr_enc_open_confirm(struct rpc_rqst *req, __be32 *p, struct nfs }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); - encode_compound_hdr(&xdr, &hdr); + encode_compound_hdr(&xdr, req, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_open_confirm(&xdr, args, &hdr); encode_nops(&hdr); @@ -1591,7 +1600,7 @@ static int nfs4_xdr_enc_open_noattr(struct rpc_rqst *req, __be32 *p, struct nfs_ }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); - encode_compound_hdr(&xdr, &hdr); + encode_compound_hdr(&xdr, req, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_open(&xdr, args, &hdr); encode_getfattr(&xdr, args->bitmask, &hdr); @@ -1610,7 +1619,7 @@ static int nfs4_xdr_enc_open_downgrade(struct rpc_rqst *req, __be32 *p, struct n }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); - encode_compound_hdr(&xdr, &hdr); + encode_compound_hdr(&xdr, req, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_open_downgrade(&xdr, args, &hdr); encode_getfattr(&xdr, args->bitmask, &hdr); @@ -1629,7 +1638,7 @@ static int nfs4_xdr_enc_lock(struct rpc_rqst *req, __be32 *p, struct nfs_lock_ar }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); - encode_compound_hdr(&xdr, &hdr); + encode_compound_hdr(&xdr, req, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_lock(&xdr, args, &hdr); encode_nops(&hdr); @@ -1647,7 +1656,7 @@ static int nfs4_xdr_enc_lockt(struct rpc_rqst *req, __be32 *p, struct nfs_lockt_ }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); - encode_compound_hdr(&xdr, &hdr); + encode_compound_hdr(&xdr, req, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_lockt(&xdr, args, &hdr); encode_nops(&hdr); @@ -1665,7 +1674,7 @@ static int nfs4_xdr_enc_locku(struct rpc_rqst *req, __be32 *p, struct nfs_locku_ }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); - encode_compound_hdr(&xdr, &hdr); + encode_compound_hdr(&xdr, req, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_locku(&xdr, args, &hdr); encode_nops(&hdr); @@ -1685,7 +1694,7 @@ static int nfs4_xdr_enc_readlink(struct rpc_rqst *req, __be32 *p, const struct n unsigned int replen; xdr_init_encode(&xdr, &req->rq_snd_buf, p); - encode_compound_hdr(&xdr, &hdr); + encode_compound_hdr(&xdr, req, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_readlink(&xdr, args, req, &hdr); @@ -1713,7 +1722,7 @@ static int nfs4_xdr_enc_readdir(struct rpc_rqst *req, __be32 *p, const struct nf int replen; xdr_init_encode(&xdr, &req->rq_snd_buf, p); - encode_compound_hdr(&xdr, &hdr); + encode_compound_hdr(&xdr, req, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_readdir(&xdr, args, req, &hdr); @@ -1744,7 +1753,7 @@ static int nfs4_xdr_enc_read(struct rpc_rqst *req, __be32 *p, struct nfs_readarg int replen; xdr_init_encode(&xdr, &req->rq_snd_buf, p); - encode_compound_hdr(&xdr, &hdr); + encode_compound_hdr(&xdr, req, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_read(&xdr, args, &hdr); @@ -1771,7 +1780,7 @@ static int nfs4_xdr_enc_setattr(struct rpc_rqst *req, __be32 *p, struct nfs_seta }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); - encode_compound_hdr(&xdr, &hdr); + encode_compound_hdr(&xdr, req, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_setattr(&xdr, args, args->server, &hdr); encode_getfattr(&xdr, args->bitmask, &hdr); @@ -1794,7 +1803,7 @@ nfs4_xdr_enc_getacl(struct rpc_rqst *req, __be32 *p, int replen; xdr_init_encode(&xdr, &req->rq_snd_buf, p); - encode_compound_hdr(&xdr, &hdr); + encode_compound_hdr(&xdr, req, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_getattr_two(&xdr, FATTR4_WORD0_ACL, 0, &hdr); @@ -1817,7 +1826,7 @@ static int nfs4_xdr_enc_write(struct rpc_rqst *req, __be32 *p, struct nfs_writea }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); - encode_compound_hdr(&xdr, &hdr); + encode_compound_hdr(&xdr, req, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_write(&xdr, args, &hdr); req->rq_snd_buf.flags |= XDRBUF_WRITE; @@ -1837,7 +1846,7 @@ static int nfs4_xdr_enc_commit(struct rpc_rqst *req, __be32 *p, struct nfs_write }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); - encode_compound_hdr(&xdr, &hdr); + encode_compound_hdr(&xdr, req, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_commit(&xdr, args, &hdr); encode_getfattr(&xdr, args->bitmask, &hdr); @@ -1856,7 +1865,7 @@ static int nfs4_xdr_enc_fsinfo(struct rpc_rqst *req, __be32 *p, struct nfs4_fsin }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); - encode_compound_hdr(&xdr, &hdr); + encode_compound_hdr(&xdr, req, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_fsinfo(&xdr, args->bitmask, &hdr); encode_nops(&hdr); @@ -1874,7 +1883,7 @@ static int nfs4_xdr_enc_pathconf(struct rpc_rqst *req, __be32 *p, const struct n }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); - encode_compound_hdr(&xdr, &hdr); + encode_compound_hdr(&xdr, req, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_getattr_one(&xdr, args->bitmask[0] & nfs4_pathconf_bitmap[0], &hdr); @@ -1893,7 +1902,7 @@ static int nfs4_xdr_enc_statfs(struct rpc_rqst *req, __be32 *p, const struct nfs }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); - encode_compound_hdr(&xdr, &hdr); + encode_compound_hdr(&xdr, req, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_getattr_two(&xdr, args->bitmask[0] & nfs4_statfs_bitmap[0], args->bitmask[1] & nfs4_statfs_bitmap[1], &hdr); @@ -1913,7 +1922,7 @@ static int nfs4_xdr_enc_server_caps(struct rpc_rqst *req, __be32 *p, }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); - encode_compound_hdr(&xdr, &hdr); + encode_compound_hdr(&xdr, req, &hdr); encode_putfh(&xdr, args->fhandle, &hdr); encode_getattr_one(&xdr, FATTR4_WORD0_SUPPORTED_ATTRS| FATTR4_WORD0_LINK_SUPPORT| @@ -1934,7 +1943,7 @@ static int nfs4_xdr_enc_renew(struct rpc_rqst *req, __be32 *p, struct nfs_client }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); - encode_compound_hdr(&xdr, &hdr); + encode_compound_hdr(&xdr, req, &hdr); encode_renew(&xdr, clp, &hdr); encode_nops(&hdr); return 0; @@ -1951,7 +1960,7 @@ static int nfs4_xdr_enc_setclientid(struct rpc_rqst *req, __be32 *p, struct nfs4 }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); - encode_compound_hdr(&xdr, &hdr); + encode_compound_hdr(&xdr, req, &hdr); encode_setclientid(&xdr, sc, &hdr); encode_nops(&hdr); return 0; @@ -1969,7 +1978,7 @@ static int nfs4_xdr_enc_setclientid_confirm(struct rpc_rqst *req, __be32 *p, str const u32 lease_bitmap[2] = { FATTR4_WORD0_LEASE_TIME, 0 }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); - encode_compound_hdr(&xdr, &hdr); + encode_compound_hdr(&xdr, req, &hdr); encode_setclientid_confirm(&xdr, clp, &hdr); encode_putrootfh(&xdr, &hdr); encode_fsinfo(&xdr, lease_bitmap, &hdr); @@ -1988,7 +1997,7 @@ static int nfs4_xdr_enc_delegreturn(struct rpc_rqst *req, __be32 *p, const struc }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); - encode_compound_hdr(&xdr, &hdr); + encode_compound_hdr(&xdr, req, &hdr); encode_putfh(&xdr, args->fhandle, &hdr); encode_delegreturn(&xdr, args->stateid, &hdr); encode_getfattr(&xdr, args->bitmask, &hdr); @@ -2009,7 +2018,7 @@ static int nfs4_xdr_enc_fs_locations(struct rpc_rqst *req, __be32 *p, struct nfs int replen; xdr_init_encode(&xdr, &req->rq_snd_buf, p); - encode_compound_hdr(&xdr, &hdr); + encode_compound_hdr(&xdr, req, &hdr); encode_putfh(&xdr, args->dir_fh, &hdr); encode_lookup(&xdr, args->name, &hdr); encode_fs_locations(&xdr, args->bitmask, &hdr); @@ -3989,7 +3998,7 @@ nfs4_xdr_enc_setacl(struct rpc_rqst *req, __be32 *p, struct nfs_setaclargs *args int status; xdr_init_encode(&xdr, &req->rq_snd_buf, p); - encode_compound_hdr(&xdr, &hdr); + encode_compound_hdr(&xdr, req, &hdr); encode_putfh(&xdr, args->fh, &hdr); status = encode_setacl(&xdr, args, &hdr); encode_nops(&hdr); -- cgit v1.2.3 From dadf0c2767ce7772fc4ff82044f3ba5823e5b79f Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Wed, 1 Apr 2009 09:22:08 -0400 Subject: NFS: update hdr->replen for every encode op Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4xdr.c | 31 +++++++++++++++++++++++++++++++ 1 file changed, 31 insertions(+) (limited to 'fs') diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index c85dbee34b8a..746c6a52794c 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -749,6 +749,7 @@ static void encode_access(struct xdr_stream *xdr, u32 access, struct compound_hd WRITE32(OP_ACCESS); WRITE32(access); hdr->nops++; + hdr->replen += decode_access_maxsz; } static void encode_close(struct xdr_stream *xdr, const struct nfs_closeargs *arg, struct compound_hdr *hdr) @@ -760,6 +761,7 @@ static void encode_close(struct xdr_stream *xdr, const struct nfs_closeargs *arg WRITE32(arg->seqid->sequence->counter); WRITEMEM(arg->stateid->data, NFS4_STATEID_SIZE); hdr->nops++; + hdr->replen += decode_close_maxsz; } static void encode_commit(struct xdr_stream *xdr, const struct nfs_writeargs *args, struct compound_hdr *hdr) @@ -771,6 +773,7 @@ static void encode_commit(struct xdr_stream *xdr, const struct nfs_writeargs *ar WRITE64(args->offset); WRITE32(args->count); hdr->nops++; + hdr->replen += decode_commit_maxsz; } static void encode_create(struct xdr_stream *xdr, const struct nfs4_create_arg *create, struct compound_hdr *hdr) @@ -802,6 +805,7 @@ static void encode_create(struct xdr_stream *xdr, const struct nfs4_create_arg * WRITE32(create->name->len); WRITEMEM(create->name->name, create->name->len); hdr->nops++; + hdr->replen += decode_create_maxsz; encode_attrs(xdr, create->attrs, create->server); } @@ -815,6 +819,7 @@ static void encode_getattr_one(struct xdr_stream *xdr, uint32_t bitmap, struct c WRITE32(1); WRITE32(bitmap); hdr->nops++; + hdr->replen += decode_getattr_maxsz; } static void encode_getattr_two(struct xdr_stream *xdr, uint32_t bm0, uint32_t bm1, struct compound_hdr *hdr) @@ -827,6 +832,7 @@ static void encode_getattr_two(struct xdr_stream *xdr, uint32_t bm0, uint32_t bm WRITE32(bm0); WRITE32(bm1); hdr->nops++; + hdr->replen += decode_getattr_maxsz; } static void encode_getfattr(struct xdr_stream *xdr, const u32* bitmask, struct compound_hdr *hdr) @@ -854,6 +860,7 @@ static void encode_getfh(struct xdr_stream *xdr, struct compound_hdr *hdr) RESERVE_SPACE(4); WRITE32(OP_GETFH); hdr->nops++; + hdr->replen += decode_getfh_maxsz; } static void encode_link(struct xdr_stream *xdr, const struct qstr *name, struct compound_hdr *hdr) @@ -865,6 +872,7 @@ static void encode_link(struct xdr_stream *xdr, const struct qstr *name, struct WRITE32(name->len); WRITEMEM(name->name, name->len); hdr->nops++; + hdr->replen += decode_link_maxsz; } static inline int nfs4_lock_type(struct file_lock *fl, int block) @@ -912,6 +920,7 @@ static void encode_lock(struct xdr_stream *xdr, const struct nfs_lock_args *args WRITE32(args->lock_seqid->sequence->counter); } hdr->nops++; + hdr->replen += decode_lock_maxsz; } static void encode_lockt(struct xdr_stream *xdr, const struct nfs_lockt_args *args, struct compound_hdr *hdr) @@ -928,6 +937,7 @@ static void encode_lockt(struct xdr_stream *xdr, const struct nfs_lockt_args *ar WRITEMEM("lock id:", 8); WRITE64(args->lock_owner.id); hdr->nops++; + hdr->replen += decode_lockt_maxsz; } static void encode_locku(struct xdr_stream *xdr, const struct nfs_locku_args *args, struct compound_hdr *hdr) @@ -942,6 +952,7 @@ static void encode_locku(struct xdr_stream *xdr, const struct nfs_locku_args *ar WRITE64(args->fl->fl_start); WRITE64(nfs4_lock_length(args->fl)); hdr->nops++; + hdr->replen += decode_locku_maxsz; } static void encode_lookup(struct xdr_stream *xdr, const struct qstr *name, struct compound_hdr *hdr) @@ -954,6 +965,7 @@ static void encode_lookup(struct xdr_stream *xdr, const struct qstr *name, struc WRITE32(len); WRITEMEM(name->name, len); hdr->nops++; + hdr->replen += decode_lookup_maxsz; } static void encode_share_access(struct xdr_stream *xdr, fmode_t fmode) @@ -1093,6 +1105,7 @@ static void encode_open(struct xdr_stream *xdr, const struct nfs_openargs *arg, BUG(); } hdr->nops++; + hdr->replen += decode_open_maxsz; } static void encode_open_confirm(struct xdr_stream *xdr, const struct nfs_open_confirmargs *arg, struct compound_hdr *hdr) @@ -1104,6 +1117,7 @@ static void encode_open_confirm(struct xdr_stream *xdr, const struct nfs_open_co WRITEMEM(arg->stateid->data, NFS4_STATEID_SIZE); WRITE32(arg->seqid->sequence->counter); hdr->nops++; + hdr->replen += decode_open_confirm_maxsz; } static void encode_open_downgrade(struct xdr_stream *xdr, const struct nfs_closeargs *arg, struct compound_hdr *hdr) @@ -1116,6 +1130,7 @@ static void encode_open_downgrade(struct xdr_stream *xdr, const struct nfs_close WRITE32(arg->seqid->sequence->counter); encode_share_access(xdr, arg->fmode); hdr->nops++; + hdr->replen += decode_open_downgrade_maxsz; } static void @@ -1129,6 +1144,7 @@ encode_putfh(struct xdr_stream *xdr, const struct nfs_fh *fh, struct compound_hd WRITE32(len); WRITEMEM(fh->data, len); hdr->nops++; + hdr->replen += decode_putfh_maxsz; } static void encode_putrootfh(struct xdr_stream *xdr, struct compound_hdr *hdr) @@ -1138,6 +1154,7 @@ static void encode_putrootfh(struct xdr_stream *xdr, struct compound_hdr *hdr) RESERVE_SPACE(4); WRITE32(OP_PUTROOTFH); hdr->nops++; + hdr->replen += decode_putrootfh_maxsz; } static void encode_stateid(struct xdr_stream *xdr, const struct nfs_open_context *ctx) @@ -1166,6 +1183,7 @@ static void encode_read(struct xdr_stream *xdr, const struct nfs_readargs *args, WRITE64(args->offset); WRITE32(args->count); hdr->nops++; + hdr->replen += decode_read_maxsz; } static void encode_readdir(struct xdr_stream *xdr, const struct nfs4_readdir_arg *readdir, struct rpc_rqst *req, struct compound_hdr *hdr) @@ -1191,6 +1209,7 @@ static void encode_readdir(struct xdr_stream *xdr, const struct nfs4_readdir_arg WRITE32(attrs[0] & readdir->bitmask[0]); WRITE32(attrs[1] & readdir->bitmask[1]); hdr->nops++; + hdr->replen += decode_readdir_maxsz; dprintk("%s: cookie = %Lu, verifier = %08x:%08x, bitmap = %08x:%08x\n", __func__, (unsigned long long)readdir->cookie, @@ -1207,6 +1226,7 @@ static void encode_readlink(struct xdr_stream *xdr, const struct nfs4_readlink * RESERVE_SPACE(4); WRITE32(OP_READLINK); hdr->nops++; + hdr->replen += decode_readlink_maxsz; } static void encode_remove(struct xdr_stream *xdr, const struct qstr *name, struct compound_hdr *hdr) @@ -1218,6 +1238,7 @@ static void encode_remove(struct xdr_stream *xdr, const struct qstr *name, struc WRITE32(name->len); WRITEMEM(name->name, name->len); hdr->nops++; + hdr->replen += decode_remove_maxsz; } static void encode_rename(struct xdr_stream *xdr, const struct qstr *oldname, const struct qstr *newname, struct compound_hdr *hdr) @@ -1233,6 +1254,7 @@ static void encode_rename(struct xdr_stream *xdr, const struct qstr *oldname, co WRITE32(newname->len); WRITEMEM(newname->name, newname->len); hdr->nops++; + hdr->replen += decode_rename_maxsz; } static void encode_renew(struct xdr_stream *xdr, const struct nfs_client *client_stateid, struct compound_hdr *hdr) @@ -1243,6 +1265,7 @@ static void encode_renew(struct xdr_stream *xdr, const struct nfs_client *client WRITE32(OP_RENEW); WRITE64(client_stateid->cl_clientid); hdr->nops++; + hdr->replen += decode_renew_maxsz; } static void @@ -1253,6 +1276,7 @@ encode_restorefh(struct xdr_stream *xdr, struct compound_hdr *hdr) RESERVE_SPACE(4); WRITE32(OP_RESTOREFH); hdr->nops++; + hdr->replen += decode_restorefh_maxsz; } static int @@ -1272,6 +1296,7 @@ encode_setacl(struct xdr_stream *xdr, struct nfs_setaclargs *arg, struct compoun WRITE32(arg->acl_len); xdr_write_pages(xdr, arg->acl_pages, arg->acl_pgbase, arg->acl_len); hdr->nops++; + hdr->replen += decode_setacl_maxsz; return 0; } @@ -1283,6 +1308,7 @@ encode_savefh(struct xdr_stream *xdr, struct compound_hdr *hdr) RESERVE_SPACE(4); WRITE32(OP_SAVEFH); hdr->nops++; + hdr->replen += decode_savefh_maxsz; } static void encode_setattr(struct xdr_stream *xdr, const struct nfs_setattrargs *arg, const struct nfs_server *server, struct compound_hdr *hdr) @@ -1293,6 +1319,7 @@ static void encode_setattr(struct xdr_stream *xdr, const struct nfs_setattrargs WRITE32(OP_SETATTR); WRITEMEM(arg->stateid.data, NFS4_STATEID_SIZE); hdr->nops++; + hdr->replen += decode_setattr_maxsz; encode_attrs(xdr, arg->iap, server); } @@ -1312,6 +1339,7 @@ static void encode_setclientid(struct xdr_stream *xdr, const struct nfs4_setclie RESERVE_SPACE(4); WRITE32(setclientid->sc_cb_ident); hdr->nops++; + hdr->replen += decode_setclientid_maxsz; } static void encode_setclientid_confirm(struct xdr_stream *xdr, const struct nfs_client *client_state, struct compound_hdr *hdr) @@ -1323,6 +1351,7 @@ static void encode_setclientid_confirm(struct xdr_stream *xdr, const struct nfs_ WRITE64(client_state->cl_clientid); WRITEMEM(client_state->cl_confirm.data, NFS4_VERIFIER_SIZE); hdr->nops++; + hdr->replen += decode_setclientid_confirm_maxsz; } static void encode_write(struct xdr_stream *xdr, const struct nfs_writeargs *args, struct compound_hdr *hdr) @@ -1341,6 +1370,7 @@ static void encode_write(struct xdr_stream *xdr, const struct nfs_writeargs *arg xdr_write_pages(xdr, args->pages, args->pgbase, args->count); hdr->nops++; + hdr->replen += decode_write_maxsz; } static void encode_delegreturn(struct xdr_stream *xdr, const nfs4_stateid *stateid, struct compound_hdr *hdr) @@ -1352,6 +1382,7 @@ static void encode_delegreturn(struct xdr_stream *xdr, const nfs4_stateid *state WRITE32(OP_DELEGRETURN); WRITEMEM(stateid->data, NFS4_STATEID_SIZE); hdr->nops++; + hdr->replen += decode_delegreturn_maxsz; } /* * END OF "GENERIC" ENCODE ROUTINES. -- cgit v1.2.3 From 28f566942c6b1d929f5e240e69e7081b77b238d3 Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Wed, 1 Apr 2009 09:22:09 -0400 Subject: NFS: use dynamically computed compound_hdr.replen for xdr_inline_pages offset As Trond suggested, rather than passing a constant to xdr_inline_pages, keep a running count of the expected reply bytes. In preparation for nfs41, where additional op sequence are expteced when talking to nfs41 servers. [NFS: cb_compoundhdr.replen is in words not bytes] Signed-off-by: Benny Halevy [nfs41: get fs_locations replen before encoding the GETATTR] Signed-off-by: Benny Halevy [nfs41: get getacl replen before encoding the GETATTR] Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4xdr.c | 48 ++++++++++-------------------------------------- 1 file changed, 10 insertions(+), 38 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index 746c6a52794c..68889469d114 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -1721,20 +1721,13 @@ static int nfs4_xdr_enc_readlink(struct rpc_rqst *req, __be32 *p, const struct n struct compound_hdr hdr = { .nops = 0, }; - struct rpc_auth *auth = req->rq_task->tk_msg.rpc_cred->cr_auth; - unsigned int replen; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, req, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_readlink(&xdr, args, req, &hdr); - /* set up reply kvec - * toplevel_status + taglen + rescount + OP_PUTFH + status - * + OP_READLINK + status + string length = 8 - */ - replen = (RPC_REPHDRSIZE + auth->au_rslack + NFS4_dec_readlink_sz) << 2; - xdr_inline_pages(&req->rq_rcv_buf, replen, args->pages, + xdr_inline_pages(&req->rq_rcv_buf, hdr.replen << 2, args->pages, args->pgbase, args->pglen); encode_nops(&hdr); return 0; @@ -1749,23 +1742,16 @@ static int nfs4_xdr_enc_readdir(struct rpc_rqst *req, __be32 *p, const struct nf struct compound_hdr hdr = { .nops = 0, }; - struct rpc_auth *auth = req->rq_task->tk_msg.rpc_cred->cr_auth; - int replen; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, req, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_readdir(&xdr, args, req, &hdr); - /* set up reply kvec - * toplevel_status + taglen + rescount + OP_PUTFH + status - * + OP_READDIR + status + verifer(2) = 9 - */ - replen = (RPC_REPHDRSIZE + auth->au_rslack + NFS4_dec_readdir_sz) << 2; - xdr_inline_pages(&req->rq_rcv_buf, replen, args->pages, + xdr_inline_pages(&req->rq_rcv_buf, hdr.replen << 2, args->pages, args->pgbase, args->count); dprintk("%s: inlined page args = (%u, %p, %u, %u)\n", - __func__, replen, args->pages, + __func__, hdr.replen << 2, args->pages, args->pgbase, args->count); encode_nops(&hdr); return 0; @@ -1776,24 +1762,17 @@ static int nfs4_xdr_enc_readdir(struct rpc_rqst *req, __be32 *p, const struct nf */ static int nfs4_xdr_enc_read(struct rpc_rqst *req, __be32 *p, struct nfs_readargs *args) { - struct rpc_auth *auth = req->rq_task->tk_msg.rpc_cred->cr_auth; struct xdr_stream xdr; struct compound_hdr hdr = { .nops = 0, }; - int replen; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, req, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_read(&xdr, args, &hdr); - /* set up reply kvec - * toplevel status + taglen=0 + rescount + OP_PUTFH + status - * + OP_READ + status + eof + datalen = 9 - */ - replen = (RPC_REPHDRSIZE + auth->au_rslack + NFS4_dec_read_sz) << 2; - xdr_inline_pages(&req->rq_rcv_buf, replen, + xdr_inline_pages(&req->rq_rcv_buf, hdr.replen << 2, args->pages, args->pgbase, args->count); req->rq_rcv_buf.flags |= XDRBUF_READ; encode_nops(&hdr); @@ -1827,20 +1806,18 @@ nfs4_xdr_enc_getacl(struct rpc_rqst *req, __be32 *p, struct nfs_getaclargs *args) { struct xdr_stream xdr; - struct rpc_auth *auth = req->rq_task->tk_msg.rpc_cred->cr_auth; struct compound_hdr hdr = { .nops = 0, }; - int replen; + uint32_t replen; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, req, &hdr); encode_putfh(&xdr, args->fh, &hdr); + replen = hdr.replen + nfs4_fattr_bitmap_maxsz + 1; encode_getattr_two(&xdr, FATTR4_WORD0_ACL, 0, &hdr); - /* set up reply buffer: */ - replen = (RPC_REPHDRSIZE + auth->au_rslack + NFS4_dec_getacl_sz) << 2; - xdr_inline_pages(&req->rq_rcv_buf, replen, + xdr_inline_pages(&req->rq_rcv_buf, replen << 2, args->acl_pages, args->acl_pgbase, args->acl_len); encode_nops(&hdr); return 0; @@ -2045,21 +2022,16 @@ static int nfs4_xdr_enc_fs_locations(struct rpc_rqst *req, __be32 *p, struct nfs struct compound_hdr hdr = { .nops = 0, }; - struct rpc_auth *auth = req->rq_task->tk_msg.rpc_cred->cr_auth; - int replen; + uint32_t replen; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, req, &hdr); encode_putfh(&xdr, args->dir_fh, &hdr); encode_lookup(&xdr, args->name, &hdr); + replen = hdr.replen; /* get the attribute into args->page */ encode_fs_locations(&xdr, args->bitmask, &hdr); - /* set up reply - * toplevel_status + OP_PUTFH + status - * + OP_LOOKUP + status + OP_GETATTR + status = 7 - */ - replen = (RPC_REPHDRSIZE + auth->au_rslack + 7) << 2; - xdr_inline_pages(&req->rq_rcv_buf, replen, &args->page, + xdr_inline_pages(&req->rq_rcv_buf, replen << 2, &args->page, 0, PAGE_SIZE); encode_nops(&hdr); return 0; -- cgit v1.2.3 From 66cc042970f7077c66be65a780eb3a60a9bcbf0b Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Wed, 1 Apr 2009 09:22:10 -0400 Subject: nfs41: encode minorversion in compound header Signed-off-by: Andy Adamdon Signed-off-by: Benny Halevy [nfs41: pass *session in seq_args and seq_res] Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4xdr.c | 70 ++++++++++++++++++++++++++++++++------------------------ 1 file changed, 40 insertions(+), 30 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index 68889469d114..19ca0b519a1d 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -546,6 +546,7 @@ struct compound_hdr { uint32_t taglen; char * tag; uint32_t replen; /* expected reply words */ + u32 minorversion; }; /* @@ -598,7 +599,7 @@ static void encode_compound_hdr(struct xdr_stream *xdr, RESERVE_SPACE(12+(XDR_QUADLEN(hdr->taglen)<<2)); WRITE32(hdr->taglen); WRITEMEM(hdr->tag, hdr->taglen); - WRITE32(NFS4_MINOR_VERSION); + WRITE32(hdr->minorversion); hdr->nops_p = p; WRITE32(hdr->nops); } @@ -1388,6 +1389,15 @@ static void encode_delegreturn(struct xdr_stream *xdr, const nfs4_stateid *state * END OF "GENERIC" ENCODE ROUTINES. */ +static u32 nfs4_xdr_minorversion(const struct nfs4_sequence_args *args) +{ +#if defined(CONFIG_NFS_V4_1) + if (args->sa_session) + return args->sa_session->clp->cl_minorversion; +#endif /* CONFIG_NFS_V4_1 */ + return 0; +} + /* * Encode an ACCESS request */ @@ -1395,7 +1405,7 @@ static int nfs4_xdr_enc_access(struct rpc_rqst *req, __be32 *p, const struct nfs { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 0, + .minorversion = nfs4_xdr_minorversion(&args->seq_args), }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); @@ -1414,7 +1424,7 @@ static int nfs4_xdr_enc_lookup(struct rpc_rqst *req, __be32 *p, const struct nfs { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 0, + .minorversion = nfs4_xdr_minorversion(&args->seq_args), }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); @@ -1434,7 +1444,7 @@ static int nfs4_xdr_enc_lookup_root(struct rpc_rqst *req, __be32 *p, const struc { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 0, + .minorversion = nfs4_xdr_minorversion(&args->seq_args), }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); @@ -1453,7 +1463,7 @@ static int nfs4_xdr_enc_remove(struct rpc_rqst *req, __be32 *p, const struct nfs { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 0, + .minorversion = nfs4_xdr_minorversion(&args->seq_args), }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); @@ -1472,7 +1482,7 @@ static int nfs4_xdr_enc_rename(struct rpc_rqst *req, __be32 *p, const struct nfs { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 0, + .minorversion = nfs4_xdr_minorversion(&args->seq_args), }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); @@ -1495,7 +1505,7 @@ static int nfs4_xdr_enc_link(struct rpc_rqst *req, __be32 *p, const struct nfs4_ { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 0, + .minorversion = nfs4_xdr_minorversion(&args->seq_args), }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); @@ -1518,7 +1528,7 @@ static int nfs4_xdr_enc_create(struct rpc_rqst *req, __be32 *p, const struct nfs { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 0, + .minorversion = nfs4_xdr_minorversion(&args->seq_args), }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); @@ -1549,7 +1559,7 @@ static int nfs4_xdr_enc_getattr(struct rpc_rqst *req, __be32 *p, const struct nf { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 0, + .minorversion = nfs4_xdr_minorversion(&args->seq_args), }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); @@ -1567,7 +1577,7 @@ static int nfs4_xdr_enc_close(struct rpc_rqst *req, __be32 *p, struct nfs_closea { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 0, + .minorversion = nfs4_xdr_minorversion(&args->seq_args), }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); @@ -1586,7 +1596,7 @@ static int nfs4_xdr_enc_open(struct rpc_rqst *req, __be32 *p, struct nfs_openarg { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 0, + .minorversion = nfs4_xdr_minorversion(&args->seq_args), }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); @@ -1627,7 +1637,7 @@ static int nfs4_xdr_enc_open_noattr(struct rpc_rqst *req, __be32 *p, struct nfs_ { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 0, + .minorversion = nfs4_xdr_minorversion(&args->seq_args), }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); @@ -1646,7 +1656,7 @@ static int nfs4_xdr_enc_open_downgrade(struct rpc_rqst *req, __be32 *p, struct n { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 0, + .minorversion = nfs4_xdr_minorversion(&args->seq_args), }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); @@ -1665,7 +1675,7 @@ static int nfs4_xdr_enc_lock(struct rpc_rqst *req, __be32 *p, struct nfs_lock_ar { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 0, + .minorversion = nfs4_xdr_minorversion(&args->seq_args), }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); @@ -1683,7 +1693,7 @@ static int nfs4_xdr_enc_lockt(struct rpc_rqst *req, __be32 *p, struct nfs_lockt_ { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 0, + .minorversion = nfs4_xdr_minorversion(&args->seq_args), }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); @@ -1701,7 +1711,7 @@ static int nfs4_xdr_enc_locku(struct rpc_rqst *req, __be32 *p, struct nfs_locku_ { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 0, + .minorversion = nfs4_xdr_minorversion(&args->seq_args), }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); @@ -1719,7 +1729,7 @@ static int nfs4_xdr_enc_readlink(struct rpc_rqst *req, __be32 *p, const struct n { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 0, + .minorversion = nfs4_xdr_minorversion(&args->seq_args), }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); @@ -1740,7 +1750,7 @@ static int nfs4_xdr_enc_readdir(struct rpc_rqst *req, __be32 *p, const struct nf { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 0, + .minorversion = nfs4_xdr_minorversion(&args->seq_args), }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); @@ -1764,7 +1774,7 @@ static int nfs4_xdr_enc_read(struct rpc_rqst *req, __be32 *p, struct nfs_readarg { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 0, + .minorversion = nfs4_xdr_minorversion(&args->seq_args), }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); @@ -1786,7 +1796,7 @@ static int nfs4_xdr_enc_setattr(struct rpc_rqst *req, __be32 *p, struct nfs_seta { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 0, + .minorversion = nfs4_xdr_minorversion(&args->seq_args), }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); @@ -1807,7 +1817,7 @@ nfs4_xdr_enc_getacl(struct rpc_rqst *req, __be32 *p, { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 0, + .minorversion = nfs4_xdr_minorversion(&args->seq_args), }; uint32_t replen; @@ -1830,7 +1840,7 @@ static int nfs4_xdr_enc_write(struct rpc_rqst *req, __be32 *p, struct nfs_writea { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 0, + .minorversion = nfs4_xdr_minorversion(&args->seq_args), }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); @@ -1850,7 +1860,7 @@ static int nfs4_xdr_enc_commit(struct rpc_rqst *req, __be32 *p, struct nfs_write { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 0, + .minorversion = nfs4_xdr_minorversion(&args->seq_args), }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); @@ -1869,7 +1879,7 @@ static int nfs4_xdr_enc_fsinfo(struct rpc_rqst *req, __be32 *p, struct nfs4_fsin { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 0, + .minorversion = nfs4_xdr_minorversion(&args->seq_args), }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); @@ -1887,7 +1897,7 @@ static int nfs4_xdr_enc_pathconf(struct rpc_rqst *req, __be32 *p, const struct n { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 0, + .minorversion = nfs4_xdr_minorversion(&args->seq_args), }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); @@ -1906,7 +1916,7 @@ static int nfs4_xdr_enc_statfs(struct rpc_rqst *req, __be32 *p, const struct nfs { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 0, + .minorversion = nfs4_xdr_minorversion(&args->seq_args), }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); @@ -1926,7 +1936,7 @@ static int nfs4_xdr_enc_server_caps(struct rpc_rqst *req, __be32 *p, { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 0, + .minorversion = nfs4_xdr_minorversion(&args->seq_args), }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); @@ -2001,7 +2011,7 @@ static int nfs4_xdr_enc_delegreturn(struct rpc_rqst *req, __be32 *p, const struc { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 0, + .minorversion = nfs4_xdr_minorversion(&args->seq_args), }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); @@ -2020,7 +2030,7 @@ static int nfs4_xdr_enc_fs_locations(struct rpc_rqst *req, __be32 *p, struct nfs { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 0, + .minorversion = nfs4_xdr_minorversion(&args->seq_args), }; uint32_t replen; @@ -3996,7 +4006,7 @@ nfs4_xdr_enc_setacl(struct rpc_rqst *req, __be32 *p, struct nfs_setaclargs *args { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 0, + .minorversion = nfs4_xdr_minorversion(&args->seq_args), }; int status; -- cgit v1.2.3 From 9b7b9fcc9c124b8a2a079f748239ce9b7a8d8304 Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Wed, 1 Apr 2009 09:22:11 -0400 Subject: nfs41: xdr {encode,decode}_sequence Implement stubs for encode and decode sequence, defined as no-ops when CONFIG_NFS_V4_1 is not defined. Add the nfsv41 encode and decode sizes. Add encode_sequence to all nfs4_enc_* routines and decode_sequence to all nfs4_dec_* routines as required by v41. [was nfs41: minorversion support for xdr] [added nfs_client argument to encode_sequence so not to use sequence_args to pass sa_session] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: pass *session in seq_args and seq_res] Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4xdr.c | 242 ++++++++++++++++++++++++++++++++++++++++++++++++++++--- 1 file changed, 232 insertions(+), 10 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index 19ca0b519a1d..5b944cd57218 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -244,43 +244,63 @@ static int nfs4_stat_to_errno(int); (encode_getattr_maxsz) #define decode_fs_locations_maxsz \ (0) + +#if defined(CONFIG_NFS_V4_1) +#define encode_sequence_maxsz 0 /* stub */ +#define decode_sequence_maxsz 0 /* stub */ +#else /* CONFIG_NFS_V4_1 */ +#define encode_sequence_maxsz 0 +#define decode_sequence_maxsz 0 +#endif /* CONFIG_NFS_V4_1 */ + #define NFS4_enc_compound_sz (1024) /* XXX: large enough? */ #define NFS4_dec_compound_sz (1024) /* XXX: large enough? */ #define NFS4_enc_read_sz (compound_encode_hdr_maxsz + \ + encode_sequence_maxsz + \ encode_putfh_maxsz + \ encode_read_maxsz) #define NFS4_dec_read_sz (compound_decode_hdr_maxsz + \ + decode_sequence_maxsz + \ decode_putfh_maxsz + \ decode_read_maxsz) #define NFS4_enc_readlink_sz (compound_encode_hdr_maxsz + \ + encode_sequence_maxsz + \ encode_putfh_maxsz + \ encode_readlink_maxsz) #define NFS4_dec_readlink_sz (compound_decode_hdr_maxsz + \ + decode_sequence_maxsz + \ decode_putfh_maxsz + \ decode_readlink_maxsz) #define NFS4_enc_readdir_sz (compound_encode_hdr_maxsz + \ + encode_sequence_maxsz + \ encode_putfh_maxsz + \ encode_readdir_maxsz) #define NFS4_dec_readdir_sz (compound_decode_hdr_maxsz + \ + decode_sequence_maxsz + \ decode_putfh_maxsz + \ decode_readdir_maxsz) #define NFS4_enc_write_sz (compound_encode_hdr_maxsz + \ + encode_sequence_maxsz + \ encode_putfh_maxsz + \ encode_write_maxsz + \ encode_getattr_maxsz) #define NFS4_dec_write_sz (compound_decode_hdr_maxsz + \ + decode_sequence_maxsz + \ decode_putfh_maxsz + \ decode_write_maxsz + \ decode_getattr_maxsz) #define NFS4_enc_commit_sz (compound_encode_hdr_maxsz + \ + encode_sequence_maxsz + \ encode_putfh_maxsz + \ encode_commit_maxsz + \ encode_getattr_maxsz) #define NFS4_dec_commit_sz (compound_decode_hdr_maxsz + \ + decode_sequence_maxsz + \ decode_putfh_maxsz + \ decode_commit_maxsz + \ decode_getattr_maxsz) #define NFS4_enc_open_sz (compound_encode_hdr_maxsz + \ + encode_sequence_maxsz + \ encode_putfh_maxsz + \ encode_savefh_maxsz + \ encode_open_maxsz + \ @@ -289,6 +309,7 @@ static int nfs4_stat_to_errno(int); encode_restorefh_maxsz + \ encode_getattr_maxsz) #define NFS4_dec_open_sz (compound_decode_hdr_maxsz + \ + decode_sequence_maxsz + \ decode_putfh_maxsz + \ decode_savefh_maxsz + \ decode_open_maxsz + \ @@ -305,43 +326,53 @@ static int nfs4_stat_to_errno(int); decode_putfh_maxsz + \ decode_open_confirm_maxsz) #define NFS4_enc_open_noattr_sz (compound_encode_hdr_maxsz + \ + encode_sequence_maxsz + \ encode_putfh_maxsz + \ encode_open_maxsz + \ encode_getattr_maxsz) #define NFS4_dec_open_noattr_sz (compound_decode_hdr_maxsz + \ + decode_sequence_maxsz + \ decode_putfh_maxsz + \ decode_open_maxsz + \ decode_getattr_maxsz) #define NFS4_enc_open_downgrade_sz \ (compound_encode_hdr_maxsz + \ + encode_sequence_maxsz + \ encode_putfh_maxsz + \ encode_open_downgrade_maxsz + \ encode_getattr_maxsz) #define NFS4_dec_open_downgrade_sz \ (compound_decode_hdr_maxsz + \ + decode_sequence_maxsz + \ decode_putfh_maxsz + \ decode_open_downgrade_maxsz + \ decode_getattr_maxsz) #define NFS4_enc_close_sz (compound_encode_hdr_maxsz + \ + encode_sequence_maxsz + \ encode_putfh_maxsz + \ encode_close_maxsz + \ encode_getattr_maxsz) #define NFS4_dec_close_sz (compound_decode_hdr_maxsz + \ + decode_sequence_maxsz + \ decode_putfh_maxsz + \ decode_close_maxsz + \ decode_getattr_maxsz) #define NFS4_enc_setattr_sz (compound_encode_hdr_maxsz + \ + encode_sequence_maxsz + \ encode_putfh_maxsz + \ encode_setattr_maxsz + \ encode_getattr_maxsz) #define NFS4_dec_setattr_sz (compound_decode_hdr_maxsz + \ + decode_sequence_maxsz + \ decode_putfh_maxsz + \ decode_setattr_maxsz + \ decode_getattr_maxsz) #define NFS4_enc_fsinfo_sz (compound_encode_hdr_maxsz + \ + encode_sequence_maxsz + \ encode_putfh_maxsz + \ encode_fsinfo_maxsz) #define NFS4_dec_fsinfo_sz (compound_decode_hdr_maxsz + \ + decode_sequence_maxsz + \ decode_putfh_maxsz + \ decode_fsinfo_maxsz) #define NFS4_enc_renew_sz (compound_encode_hdr_maxsz + \ @@ -363,64 +394,81 @@ static int nfs4_stat_to_errno(int); decode_putrootfh_maxsz + \ decode_fsinfo_maxsz) #define NFS4_enc_lock_sz (compound_encode_hdr_maxsz + \ + encode_sequence_maxsz + \ encode_putfh_maxsz + \ encode_lock_maxsz) #define NFS4_dec_lock_sz (compound_decode_hdr_maxsz + \ + decode_sequence_maxsz + \ decode_putfh_maxsz + \ decode_lock_maxsz) #define NFS4_enc_lockt_sz (compound_encode_hdr_maxsz + \ + encode_sequence_maxsz + \ encode_putfh_maxsz + \ encode_lockt_maxsz) #define NFS4_dec_lockt_sz (compound_decode_hdr_maxsz + \ + decode_sequence_maxsz + \ decode_putfh_maxsz + \ decode_lockt_maxsz) #define NFS4_enc_locku_sz (compound_encode_hdr_maxsz + \ + encode_sequence_maxsz + \ encode_putfh_maxsz + \ encode_locku_maxsz) #define NFS4_dec_locku_sz (compound_decode_hdr_maxsz + \ + decode_sequence_maxsz + \ decode_putfh_maxsz + \ decode_locku_maxsz) #define NFS4_enc_access_sz (compound_encode_hdr_maxsz + \ + encode_sequence_maxsz + \ encode_putfh_maxsz + \ encode_access_maxsz + \ encode_getattr_maxsz) #define NFS4_dec_access_sz (compound_decode_hdr_maxsz + \ + decode_sequence_maxsz + \ decode_putfh_maxsz + \ decode_access_maxsz + \ decode_getattr_maxsz) #define NFS4_enc_getattr_sz (compound_encode_hdr_maxsz + \ + encode_sequence_maxsz + \ encode_putfh_maxsz + \ encode_getattr_maxsz) #define NFS4_dec_getattr_sz (compound_decode_hdr_maxsz + \ + decode_sequence_maxsz + \ decode_putfh_maxsz + \ decode_getattr_maxsz) #define NFS4_enc_lookup_sz (compound_encode_hdr_maxsz + \ + encode_sequence_maxsz + \ encode_putfh_maxsz + \ encode_lookup_maxsz + \ encode_getattr_maxsz + \ encode_getfh_maxsz) #define NFS4_dec_lookup_sz (compound_decode_hdr_maxsz + \ + decode_sequence_maxsz + \ decode_putfh_maxsz + \ decode_lookup_maxsz + \ decode_getattr_maxsz + \ decode_getfh_maxsz) #define NFS4_enc_lookup_root_sz (compound_encode_hdr_maxsz + \ + encode_sequence_maxsz + \ encode_putrootfh_maxsz + \ encode_getattr_maxsz + \ encode_getfh_maxsz) #define NFS4_dec_lookup_root_sz (compound_decode_hdr_maxsz + \ + decode_sequence_maxsz + \ decode_putrootfh_maxsz + \ decode_getattr_maxsz + \ decode_getfh_maxsz) #define NFS4_enc_remove_sz (compound_encode_hdr_maxsz + \ + encode_sequence_maxsz + \ encode_putfh_maxsz + \ encode_remove_maxsz + \ encode_getattr_maxsz) #define NFS4_dec_remove_sz (compound_decode_hdr_maxsz + \ + decode_sequence_maxsz + \ decode_putfh_maxsz + \ decode_remove_maxsz + \ decode_getattr_maxsz) #define NFS4_enc_rename_sz (compound_encode_hdr_maxsz + \ + encode_sequence_maxsz + \ encode_putfh_maxsz + \ encode_savefh_maxsz + \ encode_putfh_maxsz + \ @@ -429,6 +477,7 @@ static int nfs4_stat_to_errno(int); encode_restorefh_maxsz + \ encode_getattr_maxsz) #define NFS4_dec_rename_sz (compound_decode_hdr_maxsz + \ + decode_sequence_maxsz + \ decode_putfh_maxsz + \ decode_savefh_maxsz + \ decode_putfh_maxsz + \ @@ -437,6 +486,7 @@ static int nfs4_stat_to_errno(int); decode_restorefh_maxsz + \ decode_getattr_maxsz) #define NFS4_enc_link_sz (compound_encode_hdr_maxsz + \ + encode_sequence_maxsz + \ encode_putfh_maxsz + \ encode_savefh_maxsz + \ encode_putfh_maxsz + \ @@ -445,6 +495,7 @@ static int nfs4_stat_to_errno(int); encode_restorefh_maxsz + \ decode_getattr_maxsz) #define NFS4_dec_link_sz (compound_decode_hdr_maxsz + \ + decode_sequence_maxsz + \ decode_putfh_maxsz + \ decode_savefh_maxsz + \ decode_putfh_maxsz + \ @@ -453,16 +504,19 @@ static int nfs4_stat_to_errno(int); decode_restorefh_maxsz + \ decode_getattr_maxsz) #define NFS4_enc_symlink_sz (compound_encode_hdr_maxsz + \ + encode_sequence_maxsz + \ encode_putfh_maxsz + \ encode_symlink_maxsz + \ encode_getattr_maxsz + \ encode_getfh_maxsz) #define NFS4_dec_symlink_sz (compound_decode_hdr_maxsz + \ + decode_sequence_maxsz + \ decode_putfh_maxsz + \ decode_symlink_maxsz + \ decode_getattr_maxsz + \ decode_getfh_maxsz) #define NFS4_enc_create_sz (compound_encode_hdr_maxsz + \ + encode_sequence_maxsz + \ encode_putfh_maxsz + \ encode_savefh_maxsz + \ encode_create_maxsz + \ @@ -471,6 +525,7 @@ static int nfs4_stat_to_errno(int); encode_restorefh_maxsz + \ encode_getattr_maxsz) #define NFS4_dec_create_sz (compound_decode_hdr_maxsz + \ + decode_sequence_maxsz + \ decode_putfh_maxsz + \ decode_savefh_maxsz + \ decode_create_maxsz + \ @@ -479,49 +534,63 @@ static int nfs4_stat_to_errno(int); decode_restorefh_maxsz + \ decode_getattr_maxsz) #define NFS4_enc_pathconf_sz (compound_encode_hdr_maxsz + \ + encode_sequence_maxsz + \ encode_putfh_maxsz + \ encode_getattr_maxsz) #define NFS4_dec_pathconf_sz (compound_decode_hdr_maxsz + \ + decode_sequence_maxsz + \ decode_putfh_maxsz + \ decode_getattr_maxsz) #define NFS4_enc_statfs_sz (compound_encode_hdr_maxsz + \ + encode_sequence_maxsz + \ encode_putfh_maxsz + \ encode_statfs_maxsz) #define NFS4_dec_statfs_sz (compound_decode_hdr_maxsz + \ + decode_sequence_maxsz + \ decode_putfh_maxsz + \ decode_statfs_maxsz) #define NFS4_enc_server_caps_sz (compound_encode_hdr_maxsz + \ + encode_sequence_maxsz + \ encode_putfh_maxsz + \ encode_getattr_maxsz) #define NFS4_dec_server_caps_sz (compound_decode_hdr_maxsz + \ + decode_sequence_maxsz + \ decode_putfh_maxsz + \ decode_getattr_maxsz) #define NFS4_enc_delegreturn_sz (compound_encode_hdr_maxsz + \ + encode_sequence_maxsz + \ encode_putfh_maxsz + \ encode_delegreturn_maxsz + \ encode_getattr_maxsz) #define NFS4_dec_delegreturn_sz (compound_decode_hdr_maxsz + \ + decode_sequence_maxsz + \ decode_delegreturn_maxsz + \ decode_getattr_maxsz) #define NFS4_enc_getacl_sz (compound_encode_hdr_maxsz + \ + encode_sequence_maxsz + \ encode_putfh_maxsz + \ encode_getacl_maxsz) #define NFS4_dec_getacl_sz (compound_decode_hdr_maxsz + \ + decode_sequence_maxsz + \ decode_putfh_maxsz + \ decode_getacl_maxsz) #define NFS4_enc_setacl_sz (compound_encode_hdr_maxsz + \ + encode_sequence_maxsz + \ encode_putfh_maxsz + \ encode_setacl_maxsz) #define NFS4_dec_setacl_sz (compound_decode_hdr_maxsz + \ + decode_sequence_maxsz + \ decode_putfh_maxsz + \ decode_setacl_maxsz) #define NFS4_enc_fs_locations_sz \ (compound_encode_hdr_maxsz + \ + encode_sequence_maxsz + \ encode_putfh_maxsz + \ encode_lookup_maxsz + \ encode_fs_locations_maxsz) #define NFS4_dec_fs_locations_sz \ (compound_decode_hdr_maxsz + \ + decode_sequence_maxsz + \ decode_putfh_maxsz + \ decode_lookup_maxsz + \ decode_fs_locations_maxsz) @@ -1385,6 +1454,24 @@ static void encode_delegreturn(struct xdr_stream *xdr, const nfs4_stateid *state hdr->nops++; hdr->replen += decode_delegreturn_maxsz; } + +/* NFSv4.1 operations */ +static void encode_sequence(struct xdr_stream *xdr, + const struct nfs4_sequence_args *args, + struct compound_hdr *hdr) +{ +#if defined(CONFIG_NFS_V4_1) + struct nfs4_session *session = args->sa_session; + + if (!session) + return; + + /* stub */ + hdr->nops++; + hdr->replen += decode_sequence_maxsz; +#endif /* CONFIG_NFS_V4_1 */ +} + /* * END OF "GENERIC" ENCODE ROUTINES. */ @@ -1410,6 +1497,7 @@ static int nfs4_xdr_enc_access(struct rpc_rqst *req, __be32 *p, const struct nfs xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, req, &hdr); + encode_sequence(&xdr, &args->seq_args, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_access(&xdr, args->access, &hdr); encode_getfattr(&xdr, args->bitmask, &hdr); @@ -1429,6 +1517,7 @@ static int nfs4_xdr_enc_lookup(struct rpc_rqst *req, __be32 *p, const struct nfs xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, req, &hdr); + encode_sequence(&xdr, &args->seq_args, &hdr); encode_putfh(&xdr, args->dir_fh, &hdr); encode_lookup(&xdr, args->name, &hdr); encode_getfh(&xdr, &hdr); @@ -1449,6 +1538,7 @@ static int nfs4_xdr_enc_lookup_root(struct rpc_rqst *req, __be32 *p, const struc xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, req, &hdr); + encode_sequence(&xdr, &args->seq_args, &hdr); encode_putrootfh(&xdr, &hdr); encode_getfh(&xdr, &hdr); encode_getfattr(&xdr, args->bitmask, &hdr); @@ -1468,6 +1558,7 @@ static int nfs4_xdr_enc_remove(struct rpc_rqst *req, __be32 *p, const struct nfs xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, req, &hdr); + encode_sequence(&xdr, &args->seq_args, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_remove(&xdr, &args->name, &hdr); encode_getfattr(&xdr, args->bitmask, &hdr); @@ -1487,6 +1578,7 @@ static int nfs4_xdr_enc_rename(struct rpc_rqst *req, __be32 *p, const struct nfs xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, req, &hdr); + encode_sequence(&xdr, &args->seq_args, &hdr); encode_putfh(&xdr, args->old_dir, &hdr); encode_savefh(&xdr, &hdr); encode_putfh(&xdr, args->new_dir, &hdr); @@ -1510,6 +1602,7 @@ static int nfs4_xdr_enc_link(struct rpc_rqst *req, __be32 *p, const struct nfs4_ xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, req, &hdr); + encode_sequence(&xdr, &args->seq_args, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_savefh(&xdr, &hdr); encode_putfh(&xdr, args->dir_fh, &hdr); @@ -1533,6 +1626,7 @@ static int nfs4_xdr_enc_create(struct rpc_rqst *req, __be32 *p, const struct nfs xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, req, &hdr); + encode_sequence(&xdr, &args->seq_args, &hdr); encode_putfh(&xdr, args->dir_fh, &hdr); encode_savefh(&xdr, &hdr); encode_create(&xdr, args, &hdr); @@ -1564,6 +1658,7 @@ static int nfs4_xdr_enc_getattr(struct rpc_rqst *req, __be32 *p, const struct nf xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, req, &hdr); + encode_sequence(&xdr, &args->seq_args, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_getfattr(&xdr, args->bitmask, &hdr); encode_nops(&hdr); @@ -1582,6 +1677,7 @@ static int nfs4_xdr_enc_close(struct rpc_rqst *req, __be32 *p, struct nfs_closea xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, req, &hdr); + encode_sequence(&xdr, &args->seq_args, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_close(&xdr, args, &hdr); encode_getfattr(&xdr, args->bitmask, &hdr); @@ -1601,6 +1697,7 @@ static int nfs4_xdr_enc_open(struct rpc_rqst *req, __be32 *p, struct nfs_openarg xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, req, &hdr); + encode_sequence(&xdr, &args->seq_args, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_savefh(&xdr, &hdr); encode_open(&xdr, args, &hdr); @@ -1642,6 +1739,7 @@ static int nfs4_xdr_enc_open_noattr(struct rpc_rqst *req, __be32 *p, struct nfs_ xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, req, &hdr); + encode_sequence(&xdr, &args->seq_args, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_open(&xdr, args, &hdr); encode_getfattr(&xdr, args->bitmask, &hdr); @@ -1661,6 +1759,7 @@ static int nfs4_xdr_enc_open_downgrade(struct rpc_rqst *req, __be32 *p, struct n xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, req, &hdr); + encode_sequence(&xdr, &args->seq_args, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_open_downgrade(&xdr, args, &hdr); encode_getfattr(&xdr, args->bitmask, &hdr); @@ -1680,6 +1779,7 @@ static int nfs4_xdr_enc_lock(struct rpc_rqst *req, __be32 *p, struct nfs_lock_ar xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, req, &hdr); + encode_sequence(&xdr, &args->seq_args, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_lock(&xdr, args, &hdr); encode_nops(&hdr); @@ -1698,6 +1798,7 @@ static int nfs4_xdr_enc_lockt(struct rpc_rqst *req, __be32 *p, struct nfs_lockt_ xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, req, &hdr); + encode_sequence(&xdr, &args->seq_args, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_lockt(&xdr, args, &hdr); encode_nops(&hdr); @@ -1716,6 +1817,7 @@ static int nfs4_xdr_enc_locku(struct rpc_rqst *req, __be32 *p, struct nfs_locku_ xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, req, &hdr); + encode_sequence(&xdr, &args->seq_args, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_locku(&xdr, args, &hdr); encode_nops(&hdr); @@ -1734,6 +1836,7 @@ static int nfs4_xdr_enc_readlink(struct rpc_rqst *req, __be32 *p, const struct n xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, req, &hdr); + encode_sequence(&xdr, &args->seq_args, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_readlink(&xdr, args, req, &hdr); @@ -1755,6 +1858,7 @@ static int nfs4_xdr_enc_readdir(struct rpc_rqst *req, __be32 *p, const struct nf xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, req, &hdr); + encode_sequence(&xdr, &args->seq_args, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_readdir(&xdr, args, req, &hdr); @@ -1779,6 +1883,7 @@ static int nfs4_xdr_enc_read(struct rpc_rqst *req, __be32 *p, struct nfs_readarg xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, req, &hdr); + encode_sequence(&xdr, &args->seq_args, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_read(&xdr, args, &hdr); @@ -1801,6 +1906,7 @@ static int nfs4_xdr_enc_setattr(struct rpc_rqst *req, __be32 *p, struct nfs_seta xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, req, &hdr); + encode_sequence(&xdr, &args->seq_args, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_setattr(&xdr, args, args->server, &hdr); encode_getfattr(&xdr, args->bitmask, &hdr); @@ -1823,6 +1929,7 @@ nfs4_xdr_enc_getacl(struct rpc_rqst *req, __be32 *p, xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, req, &hdr); + encode_sequence(&xdr, &args->seq_args, &hdr); encode_putfh(&xdr, args->fh, &hdr); replen = hdr.replen + nfs4_fattr_bitmap_maxsz + 1; encode_getattr_two(&xdr, FATTR4_WORD0_ACL, 0, &hdr); @@ -1845,6 +1952,7 @@ static int nfs4_xdr_enc_write(struct rpc_rqst *req, __be32 *p, struct nfs_writea xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, req, &hdr); + encode_sequence(&xdr, &args->seq_args, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_write(&xdr, args, &hdr); req->rq_snd_buf.flags |= XDRBUF_WRITE; @@ -1865,6 +1973,7 @@ static int nfs4_xdr_enc_commit(struct rpc_rqst *req, __be32 *p, struct nfs_write xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, req, &hdr); + encode_sequence(&xdr, &args->seq_args, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_commit(&xdr, args, &hdr); encode_getfattr(&xdr, args->bitmask, &hdr); @@ -1884,6 +1993,7 @@ static int nfs4_xdr_enc_fsinfo(struct rpc_rqst *req, __be32 *p, struct nfs4_fsin xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, req, &hdr); + encode_sequence(&xdr, &args->seq_args, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_fsinfo(&xdr, args->bitmask, &hdr); encode_nops(&hdr); @@ -1902,6 +2012,7 @@ static int nfs4_xdr_enc_pathconf(struct rpc_rqst *req, __be32 *p, const struct n xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, req, &hdr); + encode_sequence(&xdr, &args->seq_args, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_getattr_one(&xdr, args->bitmask[0] & nfs4_pathconf_bitmap[0], &hdr); @@ -1921,6 +2032,7 @@ static int nfs4_xdr_enc_statfs(struct rpc_rqst *req, __be32 *p, const struct nfs xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, req, &hdr); + encode_sequence(&xdr, &args->seq_args, &hdr); encode_putfh(&xdr, args->fh, &hdr); encode_getattr_two(&xdr, args->bitmask[0] & nfs4_statfs_bitmap[0], args->bitmask[1] & nfs4_statfs_bitmap[1], &hdr); @@ -1941,6 +2053,7 @@ static int nfs4_xdr_enc_server_caps(struct rpc_rqst *req, __be32 *p, xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, req, &hdr); + encode_sequence(&xdr, &args->seq_args, &hdr); encode_putfh(&xdr, args->fhandle, &hdr); encode_getattr_one(&xdr, FATTR4_WORD0_SUPPORTED_ATTRS| FATTR4_WORD0_LINK_SUPPORT| @@ -2016,6 +2129,7 @@ static int nfs4_xdr_enc_delegreturn(struct rpc_rqst *req, __be32 *p, const struc xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, req, &hdr); + encode_sequence(&xdr, &args->seq_args, &hdr); encode_putfh(&xdr, args->fhandle, &hdr); encode_delegreturn(&xdr, args->stateid, &hdr); encode_getfattr(&xdr, args->bitmask, &hdr); @@ -2036,6 +2150,7 @@ static int nfs4_xdr_enc_fs_locations(struct rpc_rqst *req, __be32 *p, struct nfs xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, req, &hdr); + encode_sequence(&xdr, &args->seq_args, &hdr); encode_putfh(&xdr, args->dir_fh, &hdr); encode_lookup(&xdr, args->name, &hdr); replen = hdr.replen; /* get the attribute into args->page */ @@ -3762,6 +3877,20 @@ static int decode_delegreturn(struct xdr_stream *xdr) return decode_op_hdr(xdr, OP_DELEGRETURN); } +static int decode_sequence(struct xdr_stream *xdr, + struct nfs4_sequence_res *res, + struct rpc_rqst *rqstp) +{ +#if defined(CONFIG_NFS_V4_1) + if (!res->sr_session) + return 0; + + /* stub */ +#endif /* CONFIG_NFS_V4_1 */ + + return 0; +} + /* * END OF "GENERIC" DECODE ROUTINES. */ @@ -3777,6 +3906,9 @@ static int nfs4_xdr_dec_open_downgrade(struct rpc_rqst *rqstp, __be32 *p, struct xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); status = decode_compound_hdr(&xdr, &hdr); + if (status) + goto out; + status = decode_sequence(&xdr, &res->seq_res, rqstp); if (status) goto out; status = decode_putfh(&xdr); @@ -3800,7 +3932,11 @@ static int nfs4_xdr_dec_access(struct rpc_rqst *rqstp, __be32 *p, struct nfs4_ac int status; xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); - if ((status = decode_compound_hdr(&xdr, &hdr)) != 0) + status = decode_compound_hdr(&xdr, &hdr); + if (status) + goto out; + status = decode_sequence(&xdr, &res->seq_res, rqstp); + if (status) goto out; status = decode_putfh(&xdr); if (status != 0) @@ -3823,7 +3959,11 @@ static int nfs4_xdr_dec_lookup(struct rpc_rqst *rqstp, __be32 *p, struct nfs4_lo int status; xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); - if ((status = decode_compound_hdr(&xdr, &hdr)) != 0) + status = decode_compound_hdr(&xdr, &hdr); + if (status) + goto out; + status = decode_sequence(&xdr, &res->seq_res, rqstp); + if (status) goto out; if ((status = decode_putfh(&xdr)) != 0) goto out; @@ -3846,7 +3986,11 @@ static int nfs4_xdr_dec_lookup_root(struct rpc_rqst *rqstp, __be32 *p, struct nf int status; xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); - if ((status = decode_compound_hdr(&xdr, &hdr)) != 0) + status = decode_compound_hdr(&xdr, &hdr); + if (status) + goto out; + status = decode_sequence(&xdr, &res->seq_res, rqstp); + if (status) goto out; if ((status = decode_putrootfh(&xdr)) != 0) goto out; @@ -3866,7 +4010,11 @@ static int nfs4_xdr_dec_remove(struct rpc_rqst *rqstp, __be32 *p, struct nfs_rem int status; xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); - if ((status = decode_compound_hdr(&xdr, &hdr)) != 0) + status = decode_compound_hdr(&xdr, &hdr); + if (status) + goto out; + status = decode_sequence(&xdr, &res->seq_res, rqstp); + if (status) goto out; if ((status = decode_putfh(&xdr)) != 0) goto out; @@ -3887,7 +4035,11 @@ static int nfs4_xdr_dec_rename(struct rpc_rqst *rqstp, __be32 *p, struct nfs4_re int status; xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); - if ((status = decode_compound_hdr(&xdr, &hdr)) != 0) + status = decode_compound_hdr(&xdr, &hdr); + if (status) + goto out; + status = decode_sequence(&xdr, &res->seq_res, rqstp); + if (status) goto out; if ((status = decode_putfh(&xdr)) != 0) goto out; @@ -3917,7 +4069,11 @@ static int nfs4_xdr_dec_link(struct rpc_rqst *rqstp, __be32 *p, struct nfs4_link int status; xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); - if ((status = decode_compound_hdr(&xdr, &hdr)) != 0) + status = decode_compound_hdr(&xdr, &hdr); + if (status) + goto out; + status = decode_sequence(&xdr, &res->seq_res, rqstp); + if (status) goto out; if ((status = decode_putfh(&xdr)) != 0) goto out; @@ -3950,7 +4106,11 @@ static int nfs4_xdr_dec_create(struct rpc_rqst *rqstp, __be32 *p, struct nfs4_cr int status; xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); - if ((status = decode_compound_hdr(&xdr, &hdr)) != 0) + status = decode_compound_hdr(&xdr, &hdr); + if (status) + goto out; + status = decode_sequence(&xdr, &res->seq_res, rqstp); + if (status) goto out; if ((status = decode_putfh(&xdr)) != 0) goto out; @@ -3988,6 +4148,9 @@ static int nfs4_xdr_dec_getattr(struct rpc_rqst *rqstp, __be32 *p, struct nfs4_g xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); status = decode_compound_hdr(&xdr, &hdr); + if (status) + goto out; + status = decode_sequence(&xdr, &res->seq_res, rqstp); if (status) goto out; status = decode_putfh(&xdr); @@ -4012,6 +4175,7 @@ nfs4_xdr_enc_setacl(struct rpc_rqst *req, __be32 *p, struct nfs_setaclargs *args xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, req, &hdr); + encode_sequence(&xdr, &args->seq_args, &hdr); encode_putfh(&xdr, args->fh, &hdr); status = encode_setacl(&xdr, args, &hdr); encode_nops(&hdr); @@ -4031,6 +4195,9 @@ nfs4_xdr_dec_setacl(struct rpc_rqst *rqstp, __be32 *p, xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); status = decode_compound_hdr(&xdr, &hdr); + if (status) + goto out; + status = decode_sequence(&xdr, &res->seq_res, rqstp); if (status) goto out; status = decode_putfh(&xdr); @@ -4054,6 +4221,9 @@ nfs4_xdr_dec_getacl(struct rpc_rqst *rqstp, __be32 *p, xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); status = decode_compound_hdr(&xdr, &hdr); + if (status) + goto out; + status = decode_sequence(&xdr, &res->seq_res, rqstp); if (status) goto out; status = decode_putfh(&xdr); @@ -4076,6 +4246,9 @@ static int nfs4_xdr_dec_close(struct rpc_rqst *rqstp, __be32 *p, struct nfs_clos xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); status = decode_compound_hdr(&xdr, &hdr); + if (status) + goto out; + status = decode_sequence(&xdr, &res->seq_res, rqstp); if (status) goto out; status = decode_putfh(&xdr); @@ -4106,6 +4279,9 @@ static int nfs4_xdr_dec_open(struct rpc_rqst *rqstp, __be32 *p, struct nfs_openr xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); status = decode_compound_hdr(&xdr, &hdr); + if (status) + goto out; + status = decode_sequence(&xdr, &res->seq_res, rqstp); if (status) goto out; status = decode_putfh(&xdr); @@ -4160,6 +4336,9 @@ static int nfs4_xdr_dec_open_noattr(struct rpc_rqst *rqstp, __be32 *p, struct nf xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); status = decode_compound_hdr(&xdr, &hdr); + if (status) + goto out; + status = decode_sequence(&xdr, &res->seq_res, rqstp); if (status) goto out; status = decode_putfh(&xdr); @@ -4184,6 +4363,9 @@ static int nfs4_xdr_dec_setattr(struct rpc_rqst *rqstp, __be32 *p, struct nfs_se xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); status = decode_compound_hdr(&xdr, &hdr); + if (status) + goto out; + status = decode_sequence(&xdr, &res->seq_res, rqstp); if (status) goto out; status = decode_putfh(&xdr); @@ -4208,6 +4390,9 @@ static int nfs4_xdr_dec_lock(struct rpc_rqst *rqstp, __be32 *p, struct nfs_lock_ xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); status = decode_compound_hdr(&xdr, &hdr); + if (status) + goto out; + status = decode_sequence(&xdr, &res->seq_res, rqstp); if (status) goto out; status = decode_putfh(&xdr); @@ -4229,6 +4414,9 @@ static int nfs4_xdr_dec_lockt(struct rpc_rqst *rqstp, __be32 *p, struct nfs_lock xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); status = decode_compound_hdr(&xdr, &hdr); + if (status) + goto out; + status = decode_sequence(&xdr, &res->seq_res, rqstp); if (status) goto out; status = decode_putfh(&xdr); @@ -4250,6 +4438,9 @@ static int nfs4_xdr_dec_locku(struct rpc_rqst *rqstp, __be32 *p, struct nfs_lock xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); status = decode_compound_hdr(&xdr, &hdr); + if (status) + goto out; + status = decode_sequence(&xdr, &res->seq_res, rqstp); if (status) goto out; status = decode_putfh(&xdr); @@ -4272,6 +4463,9 @@ static int nfs4_xdr_dec_readlink(struct rpc_rqst *rqstp, __be32 *p, xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); status = decode_compound_hdr(&xdr, &hdr); + if (status) + goto out; + status = decode_sequence(&xdr, &res->seq_res, rqstp); if (status) goto out; status = decode_putfh(&xdr); @@ -4293,6 +4487,9 @@ static int nfs4_xdr_dec_readdir(struct rpc_rqst *rqstp, __be32 *p, struct nfs4_r xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); status = decode_compound_hdr(&xdr, &hdr); + if (status) + goto out; + status = decode_sequence(&xdr, &res->seq_res, rqstp); if (status) goto out; status = decode_putfh(&xdr); @@ -4314,6 +4511,9 @@ static int nfs4_xdr_dec_read(struct rpc_rqst *rqstp, __be32 *p, struct nfs_readr xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); status = decode_compound_hdr(&xdr, &hdr); + if (status) + goto out; + status = decode_sequence(&xdr, &res->seq_res, rqstp); if (status) goto out; status = decode_putfh(&xdr); @@ -4337,6 +4537,9 @@ static int nfs4_xdr_dec_write(struct rpc_rqst *rqstp, __be32 *p, struct nfs_writ xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); status = decode_compound_hdr(&xdr, &hdr); + if (status) + goto out; + status = decode_sequence(&xdr, &res->seq_res, rqstp); if (status) goto out; status = decode_putfh(&xdr); @@ -4363,6 +4566,9 @@ static int nfs4_xdr_dec_commit(struct rpc_rqst *rqstp, __be32 *p, struct nfs_wri xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); status = decode_compound_hdr(&xdr, &hdr); + if (status) + goto out; + status = decode_sequence(&xdr, &res->seq_res, rqstp); if (status) goto out; status = decode_putfh(&xdr); @@ -4388,6 +4594,8 @@ static int nfs4_xdr_dec_fsinfo(struct rpc_rqst *req, __be32 *p, xdr_init_decode(&xdr, &req->rq_rcv_buf, p); status = decode_compound_hdr(&xdr, &hdr); + if (!status) + status = decode_sequence(&xdr, &res->seq_res, req); if (!status) status = decode_putfh(&xdr); if (!status) @@ -4407,6 +4615,8 @@ static int nfs4_xdr_dec_pathconf(struct rpc_rqst *req, __be32 *p, xdr_init_decode(&xdr, &req->rq_rcv_buf, p); status = decode_compound_hdr(&xdr, &hdr); + if (!status) + status = decode_sequence(&xdr, &res->seq_res, req); if (!status) status = decode_putfh(&xdr); if (!status) @@ -4426,6 +4636,8 @@ static int nfs4_xdr_dec_statfs(struct rpc_rqst *req, __be32 *p, xdr_init_decode(&xdr, &req->rq_rcv_buf, p); status = decode_compound_hdr(&xdr, &hdr); + if (!status) + status = decode_sequence(&xdr, &res->seq_res, req); if (!status) status = decode_putfh(&xdr); if (!status) @@ -4443,7 +4655,11 @@ static int nfs4_xdr_dec_server_caps(struct rpc_rqst *req, __be32 *p, struct nfs4 int status; xdr_init_decode(&xdr, &req->rq_rcv_buf, p); - if ((status = decode_compound_hdr(&xdr, &hdr)) != 0) + status = decode_compound_hdr(&xdr, &hdr); + if (status) + goto out; + status = decode_sequence(&xdr, &res->seq_res, req); + if (status) goto out; if ((status = decode_putfh(&xdr)) != 0) goto out; @@ -4516,7 +4732,10 @@ static int nfs4_xdr_dec_delegreturn(struct rpc_rqst *rqstp, __be32 *p, struct nf xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); status = decode_compound_hdr(&xdr, &hdr); - if (status != 0) + if (status) + goto out; + status = decode_sequence(&xdr, &res->seq_res, rqstp); + if (status) goto out; status = decode_putfh(&xdr); if (status != 0) @@ -4539,7 +4758,10 @@ static int nfs4_xdr_dec_fs_locations(struct rpc_rqst *req, __be32 *p, xdr_init_decode(&xdr, &req->rq_rcv_buf, p); status = decode_compound_hdr(&xdr, &hdr); - if (status != 0) + if (status) + goto out; + status = decode_sequence(&xdr, &res->seq_res, req); + if (status) goto out; if ((status = decode_putfh(&xdr)) != 0) goto out; -- cgit v1.2.3 From ce5039c1be1fd04c2d9b1af80d41a6d003f2e20c Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Wed, 1 Apr 2009 09:22:13 -0400 Subject: nfs41: nfs4_setup_sequence Perform the nfs4_setup_sequence in the rpc_call_prepare state. If a session slot is not available, we will rpc_sleep_on the slot wait queue leaving the tk_action as rpc_call_prepare. Once we have a session slot, hang on to it even through rpc_restart_calls. Ensure the nfs41_sequence_res sr_slot pointer is NULL before rpc_run_task is called as nfs41_setup_sequence will only find a new slot if it is NULL. A future patch will call free slot after any rpc_restart_calls, and handle the rpc restart that result from a sequence operation error. Signed-off-by: Rahul Iyer [nfs41: sequence res use slotid] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: simplify nfs4_call_sync] Signed-off-by: Benny Halevy [nfs41: nfs4_call_sync] [nfs41: check for session not minorversion] [nfs41: remove rpc_message from nfs41_call_sync_args] [moved NFS4_MAX_SLOT_TABLE logic into nfs41_setup_sequence] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: nfs41_call_sync_data use nfs_client not nfs_server] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: expose nfs4_call_sync_session for lease renewal] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: remove unnecessary return check] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 100 ++++++++++++++++++++++++++++++++++++++++++++++++++++-- 1 file changed, 98 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 5878930b4c3b..90aed76e593b 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -273,14 +273,110 @@ static void renew_lease(const struct nfs_server *server, unsigned long timestamp #if defined(CONFIG_NFS_V4_1) +static int nfs41_setup_sequence(struct nfs4_session *session, + struct nfs4_sequence_args *args, + struct nfs4_sequence_res *res, + int cache_reply, + struct rpc_task *task) +{ + /* slot already allocated? */ + if (res->sr_slotid != NFS4_MAX_SLOT_TABLE) + return 0; + + memset(res, 0, sizeof(*res)); + res->sr_slotid = NFS4_MAX_SLOT_TABLE; + return 0; +} + +int nfs4_setup_sequence(struct nfs_client *clp, + struct nfs4_sequence_args *args, + struct nfs4_sequence_res *res, + int cache_reply, + struct rpc_task *task) +{ + int ret = 0; + + dprintk("--> %s clp %p session %p sr_slotid %d\n", + __func__, clp, clp->cl_session, res->sr_slotid); + + if (!nfs4_has_session(clp)) + goto out; + ret = nfs41_setup_sequence(clp->cl_session, args, res, cache_reply, + task); + if (ret != -EAGAIN) { + /* terminate rpc task */ + task->tk_status = ret; + task->tk_action = NULL; + } +out: + dprintk("<-- %s status=%d\n", __func__, ret); + return ret; +} + +struct nfs41_call_sync_data { + struct nfs_client *clp; + struct nfs4_sequence_args *seq_args; + struct nfs4_sequence_res *seq_res; + int cache_reply; +}; + +static void nfs41_call_sync_prepare(struct rpc_task *task, void *calldata) +{ + struct nfs41_call_sync_data *data = calldata; + + dprintk("--> %s data->clp->cl_session %p\n", __func__, + data->clp->cl_session); + if (nfs4_setup_sequence(data->clp, data->seq_args, + data->seq_res, data->cache_reply, task)) + return; + rpc_call_start(task); +} + +struct rpc_call_ops nfs41_call_sync_ops = { + .rpc_call_prepare = nfs41_call_sync_prepare, +}; + +static int nfs4_call_sync_sequence(struct nfs_client *clp, + struct rpc_clnt *clnt, + struct rpc_message *msg, + struct nfs4_sequence_args *args, + struct nfs4_sequence_res *res, + int cache_reply) +{ + int ret; + struct rpc_task *task; + struct nfs41_call_sync_data data = { + .clp = clp, + .seq_args = args, + .seq_res = res, + .cache_reply = cache_reply, + }; + struct rpc_task_setup task_setup = { + .rpc_client = clnt, + .rpc_message = msg, + .callback_ops = &nfs41_call_sync_ops, + .callback_data = &data + }; + + res->sr_slotid = NFS4_MAX_SLOT_TABLE; + task = rpc_run_task(&task_setup); + if (IS_ERR(task)) + ret = PTR_ERR(task); + else { + ret = task->tk_status; + rpc_put_task(task); + } + return ret; +} + int _nfs4_call_sync_session(struct nfs_server *server, struct rpc_message *msg, struct nfs4_sequence_args *args, struct nfs4_sequence_res *res, int cache_reply) { - /* in preparation for setting up the sequence op */ - return rpc_call_sync(server->client, msg, 0); + return nfs4_call_sync_sequence(server->nfs_client, server->client, + msg, args, res, cache_reply); } #endif /* CONFIG_NFS_V4_1 */ -- cgit v1.2.3 From 510b81756f18922a4c5b555e8145f4fed5beb569 Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Wed, 1 Apr 2009 09:22:14 -0400 Subject: nfs41: find slot Find a free slot using bitmap-based allocation. Use the optimized ffz function to find a zero bit in the bitmap that indicates a free slot, starting the search from the 'lowest_free_slotid' position. If found, mark the slot as used in the bitmap, get the slot's slotid and seqid, and update max_slotid to be used by the SEQUENCE operation. Also, update lowest_free_slotid for next search. If no free slot was found the caller has to wait for a free slot (outside the scope of this function) Signed-off-by: Benny Halevy [nfs41: find slot return slotid] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [use find_first_zero_bit for nfs4_find_slot as per review comment 21/85.] [use NFS4_MAX_SLOT_TABLE rather than NFS4_NO_SLOT] [nfs41: rpc_sleep_on slot_tbl_waitq must be called under slot_tbl_lock] Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 33 +++++++++++++++++++++++++++++++++ 1 file changed, 33 insertions(+) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 90aed76e593b..a0946a0d116e 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -273,6 +273,39 @@ static void renew_lease(const struct nfs_server *server, unsigned long timestamp #if defined(CONFIG_NFS_V4_1) +/* + * nfs4_find_slot - efficiently look for a free slot + * + * nfs4_find_slot looks for an unset bit in the used_slots bitmap. + * If found, we mark the slot as used, update the highest_used_slotid, + * and respectively set up the sequence operation args. + * The slot number is returned if found, or NFS4_MAX_SLOT_TABLE otherwise. + */ +static u8 +nfs4_find_slot(struct nfs4_slot_table *tbl, struct rpc_task *task) +{ + int slotid; + u8 ret_id = NFS4_MAX_SLOT_TABLE; + BUILD_BUG_ON((u8)NFS4_MAX_SLOT_TABLE != (int)NFS4_MAX_SLOT_TABLE); + + spin_lock(&tbl->slot_tbl_lock); + dprintk("--> %s used_slots=%04lx highest_used=%d max_slots=%d\n", + __func__, tbl->used_slots[0], tbl->highest_used_slotid, + tbl->max_slots); + slotid = find_first_zero_bit(tbl->used_slots, tbl->max_slots); + if (slotid >= tbl->max_slots) + goto out; + __set_bit(slotid, tbl->used_slots); + if (slotid > tbl->highest_used_slotid) + tbl->highest_used_slotid = slotid; + ret_id = slotid; +out: + dprintk("<-- %s used_slots=%04lx highest_used=%d slotid=%d \n", + __func__, tbl->used_slots[0], tbl->highest_used_slotid, ret_id); + spin_unlock(&tbl->slot_tbl_lock); + return ret_id; +} + static int nfs41_setup_sequence(struct nfs4_session *session, struct nfs4_sequence_args *args, struct nfs4_sequence_res *res, -- cgit v1.2.3 From fbcd4abcb3841f85578985c09c6df85aa41b0ae8 Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Wed, 1 Apr 2009 09:22:15 -0400 Subject: nfs41: setup_sequence method Allocate a slot in the session slot table and set the sequence op arguments. Called at the rpc prepare stage. Add a status to nfs41_sequence_res, initialize it to one so that we catch rpc level failures which do not go through decode_sequence which sets the new status field. Note that upon an rpc level failure, we don't know if the server processed the sequence operation or not. Proceed as if the server did process the sequence operation. Signed-off-by: Rahul Iyer [nfs41: sequence args use slotid] [nfs41: find slot return slotid] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: remove SEQ4_STATUS_USE_TK_STATUS] As per 11-14-08 review [move extern declaration from nfs41: sequence setup/done support] [removed sa_session definition, changed sa_cache_this into a u8 to reduce footprint] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: rpc_sleep_on slot_tbl_waitq must be called under slot_tbl_lock] Otherwise there's a race (we've hit) with nfs4_free_slot where nfs41_setup_sequence sees a full slot table, unlocks slot_tbl_lock, nfs4_free_slots happen concurrently and call rpc_wake_up_next where there's nobody to wake up yet, context goes back to nfs41_setup_sequence which goes to sleep when the slot table is actually empty now and there's no-one to wake it up anymore. Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4_fs.h | 10 ++++++++++ fs/nfs/nfs4proc.c | 34 ++++++++++++++++++++++++++++++++-- include/linux/nfs_xdr.h | 4 ++++ 3 files changed, 46 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs4_fs.h b/fs/nfs/nfs4_fs.h index acac6f8c3d39..eccf4e93e7d7 100644 --- a/fs/nfs/nfs4_fs.h +++ b/fs/nfs/nfs4_fs.h @@ -203,8 +203,18 @@ extern int nfs4_proc_fs_locations(struct inode *dir, const struct qstr *name, extern struct nfs4_state_recovery_ops nfs4_reboot_recovery_ops; extern struct nfs4_state_recovery_ops nfs4_nograce_recovery_ops; #if defined(CONFIG_NFS_V4_1) +extern int nfs4_setup_sequence(struct nfs_client *clp, + struct nfs4_sequence_args *args, struct nfs4_sequence_res *res, + int cache_reply, struct rpc_task *task); extern void nfs4_destroy_session(struct nfs4_session *session); extern struct nfs4_session *nfs4_alloc_session(struct nfs_client *clp); +#else /* CONFIG_NFS_v4_1 */ +static inline int nfs4_setup_sequence(struct nfs_client *clp, + struct nfs4_sequence_args *args, struct nfs4_sequence_res *res, + int cache_reply, struct rpc_task *task) +{ + return 0; +} #endif /* CONFIG_NFS_V4_1 */ extern const u32 nfs4_fattr_bitmap[2]; diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index a0946a0d116e..c9618080317e 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -280,6 +280,8 @@ static void renew_lease(const struct nfs_server *server, unsigned long timestamp * If found, we mark the slot as used, update the highest_used_slotid, * and respectively set up the sequence operation args. * The slot number is returned if found, or NFS4_MAX_SLOT_TABLE otherwise. + * + * Note: must be called with under the slot_tbl_lock. */ static u8 nfs4_find_slot(struct nfs4_slot_table *tbl, struct rpc_task *task) @@ -288,7 +290,6 @@ nfs4_find_slot(struct nfs4_slot_table *tbl, struct rpc_task *task) u8 ret_id = NFS4_MAX_SLOT_TABLE; BUILD_BUG_ON((u8)NFS4_MAX_SLOT_TABLE != (int)NFS4_MAX_SLOT_TABLE); - spin_lock(&tbl->slot_tbl_lock); dprintk("--> %s used_slots=%04lx highest_used=%d max_slots=%d\n", __func__, tbl->used_slots[0], tbl->highest_used_slotid, tbl->max_slots); @@ -302,7 +303,6 @@ nfs4_find_slot(struct nfs4_slot_table *tbl, struct rpc_task *task) out: dprintk("<-- %s used_slots=%04lx highest_used=%d slotid=%d \n", __func__, tbl->used_slots[0], tbl->highest_used_slotid, ret_id); - spin_unlock(&tbl->slot_tbl_lock); return ret_id; } @@ -312,12 +312,42 @@ static int nfs41_setup_sequence(struct nfs4_session *session, int cache_reply, struct rpc_task *task) { + struct nfs4_slot *slot; + struct nfs4_slot_table *tbl; + u8 slotid; + + dprintk("--> %s\n", __func__); /* slot already allocated? */ if (res->sr_slotid != NFS4_MAX_SLOT_TABLE) return 0; memset(res, 0, sizeof(*res)); res->sr_slotid = NFS4_MAX_SLOT_TABLE; + tbl = &session->fc_slot_table; + + spin_lock(&tbl->slot_tbl_lock); + slotid = nfs4_find_slot(tbl, task); + if (slotid == NFS4_MAX_SLOT_TABLE) { + rpc_sleep_on(&tbl->slot_tbl_waitq, task, NULL); + spin_unlock(&tbl->slot_tbl_lock); + dprintk("<-- %s: no free slots\n", __func__); + return -EAGAIN; + } + spin_unlock(&tbl->slot_tbl_lock); + + slot = tbl->slots + slotid; + args->sa_slotid = slotid; + args->sa_cache_this = cache_reply; + + dprintk("<-- %s slotid=%d seqid=%d\n", __func__, slotid, slot->seq_nr); + + res->sr_slotid = slotid; + res->sr_renewal_time = jiffies; + /* + * sr_status is only set in decode_sequence, and so will remain + * set to 1 if an rpc level failure occurs. + */ + res->sr_status = 1; return 0; } diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h index db0d1236aae7..4ac14b40efc9 100644 --- a/include/linux/nfs_xdr.h +++ b/include/linux/nfs_xdr.h @@ -163,11 +163,15 @@ struct nfs4_slot { struct nfs4_sequence_args { struct nfs4_session *sa_session; + u8 sa_slotid; + u8 sa_cache_this; }; struct nfs4_sequence_res { struct nfs4_session *sr_session; u8 sr_slotid; /* slot used to send request */ + unsigned long sr_renewal_time; + int sr_status; /* sequence operation status */ }; /* -- cgit v1.2.3 From e2c4ab3ce2ecc527672bd7e29a594c50d3ec0477 Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Wed, 1 Apr 2009 09:22:16 -0400 Subject: nfs41: free slot Free a slot in the slot table. Mark the slot as free in the bitmap-based allocation table by clearing a bit corresponding to the slotid. Update lowest_free_slotid if freed slotid is lower than that. Update highest_used_slotid. In the case the freed slotid equals the highest_used_slotid, scan downwards for the next highest used slotid using the optimized fls* functions. Finally, wake up thread waiting on slot_tbl_waitq for a free slot to become available. Signed-off-by: Benny Halevy [nfs41: free slot use slotid] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: use find_first_zero_bit for nfs4_find_slot] While at it, obliterate lowest_free_slotid and fix-up related comments. As per review comment 21/85. Signed-off-by: Benny Halevy [nfs41: use __clear_bit for nfs4_free_slot] While at it, fix-up function comment. Part of review comment 22/85. Signed-off-by: Benny Halevy [nfs41: use find_last_bit in nfs4_free_slot to determine highest used slot.] Signed-off-by: Benny Halevy [nfs41: rpc_sleep_on slot_tbl_waitq must be called under slot_tbl_lock] Otherwise there's a race (we've hit) with nfs4_free_slot where nfs41_setup_sequence sees a full slot table, unlocks slot_tbl_lock, nfs4_free_slots happen concurrently and call rpc_wake_up_next where there's nobody to wake up yet, context goes back to nfs41_setup_sequence which goes to sleep when the slot table is actually empty now and there's no-one to wake it up anymore. Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 36 ++++++++++++++++++++++++++++++++++++ 1 file changed, 36 insertions(+) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index c9618080317e..b6308f6740ec 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -273,6 +273,42 @@ static void renew_lease(const struct nfs_server *server, unsigned long timestamp #if defined(CONFIG_NFS_V4_1) +/* + * nfs4_free_slot - free a slot and efficiently update slot table. + * + * freeing a slot is trivially done by clearing its respective bit + * in the bitmap. + * If the freed slotid equals highest_used_slotid we want to update it + * so that the server would be able to size down the slot table if needed, + * otherwise we know that the highest_used_slotid is still in use. + * When updating highest_used_slotid there may be "holes" in the bitmap + * so we need to scan down from highest_used_slotid to 0 looking for the now + * highest slotid in use. + * If none found, highest_used_slotid is set to -1. + */ +static void +nfs4_free_slot(struct nfs4_slot_table *tbl, u8 free_slotid) +{ + int slotid = free_slotid; + + spin_lock(&tbl->slot_tbl_lock); + /* clear used bit in bitmap */ + __clear_bit(slotid, tbl->used_slots); + + /* update highest_used_slotid when it is freed */ + if (slotid == tbl->highest_used_slotid) { + slotid = find_last_bit(tbl->used_slots, tbl->max_slots); + if (slotid >= 0 && slotid < tbl->max_slots) + tbl->highest_used_slotid = slotid; + else + tbl->highest_used_slotid = -1; + } + rpc_wake_up_next(&tbl->slot_tbl_waitq); + spin_unlock(&tbl->slot_tbl_lock); + dprintk("%s: free_slotid %u highest_used_slotid %d\n", __func__, + free_slotid, tbl->highest_used_slotid); +} + /* * nfs4_find_slot - efficiently look for a free slot * -- cgit v1.2.3 From 13615871cdf8e8263626a817c350c6978a6483fe Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Wed, 1 Apr 2009 09:22:17 -0400 Subject: nfs41: nfs41_sequence_free_slot [from nfs41: separate free slot from sequence done] Don't free the slot until after all rpc_restart_calls have completed. Session reset will require more work. As noted by Trond, since we're using rpc_wake_up_next rather than rpc_wake_up() we must always wake up the next task in the queue either by going through nfs4_free_slot, or just calling rpc_wake_up_next if no slot is to be freed. [nfs41: sequence res use slotid] [nfs41: remove SEQ4_STATUS_USE_TK_STATUS] [got rid of nfs4_sequence_res.sr_session, use nfs_client.cl_session instead] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: rpc_wake_up_next if sessions slot was not consumed.] Signed-off-by: Benny Halevy [nfs41: nfs4_sequence_free_slot use nfs_client for data server] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/internal.h | 14 ++++++++++++++ fs/nfs/nfs4proc.c | 21 +++++++++++++++++++++ 2 files changed, 35 insertions(+) (limited to 'fs') diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h index 8d67c2865dc3..12f4c5e6fd39 100644 --- a/fs/nfs/internal.h +++ b/fs/nfs/internal.h @@ -230,6 +230,20 @@ static inline int nfs4_has_session(const struct nfs_client *clp) return 0; } +#ifdef CONFIG_NFS_V4_1 +extern void nfs41_sequence_free_slot(const struct nfs_client *, + struct nfs4_sequence_res *res); +#endif /* CONFIG_NFS_V4_1 */ + +static inline void nfs4_sequence_free_slot(const struct nfs_client *clp, + struct nfs4_sequence_res *res) +{ +#ifdef CONFIG_NFS_V4_1 + if (nfs4_has_session(clp)) + nfs41_sequence_free_slot(clp, res); +#endif /* CONFIG_NFS_V4_1 */ +} + /* * Determine the device name as a string */ diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index b6308f6740ec..861b103c9e3c 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -309,6 +309,27 @@ nfs4_free_slot(struct nfs4_slot_table *tbl, u8 free_slotid) free_slotid, tbl->highest_used_slotid); } +void nfs41_sequence_free_slot(const struct nfs_client *clp, + struct nfs4_sequence_res *res) +{ + struct nfs4_slot_table *tbl; + + if (!nfs4_has_session(clp)) { + dprintk("%s: No session\n", __func__); + return; + } + tbl = &clp->cl_session->fc_slot_table; + if (res->sr_slotid == NFS4_MAX_SLOT_TABLE) { + dprintk("%s: No slot\n", __func__); + /* just wake up the next guy waiting since + * we may have not consumed a slot after all */ + rpc_wake_up_next(&tbl->slot_tbl_waitq); + return; + } + nfs4_free_slot(tbl, res->sr_slotid); + res->sr_slotid = NFS4_MAX_SLOT_TABLE; +} + /* * nfs4_find_slot - efficiently look for a free slot * -- cgit v1.2.3 From b0df806c0f9d61e5fcfdb0f4e606e1a59303aaa6 Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Wed, 1 Apr 2009 09:22:18 -0400 Subject: nfs41: nfs41_sequence_done Handle session level errors, update slot sequence id and sessions bookeeping, free slot. [nfs41: sequence res use slotid] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: remove SEQ4_STATUS_USE_TK_STATUS] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: check for session not minorversion] Signed-off-by: Andy Adamson [nfs41: bail out early out of nfs41_sequence_done if !res->sr_session] Signed-off-by: Benny Halevy [move nfs4_sequence_done from nfs41: nfs41_call_sync_done] Signed-off-by: Andy Adamson [move nfs4_sequence_free_slot from nfs41: separate free slot from sequence done] Don't free the slot until after all rpc_restart_calls have completed. Session reset will require more work. Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [moved reset sr_slotid to nfs41_sequence_free_slot] [free slot also on unexpectecd error] [remove seq_res.sr_session member, use nfs_client's instead] [ditch seq_res.sr_flags until used] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [look at sr_slotid for bailing out early from nfs41_sequence_done] [nfs41: rpc_wake_up_next if sessions slot was not consumed.] Signed-off-by: Benny Halevy [nfs41: nfs4_sequence_free_slot use nfs_client for data server] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: remove unused error checking in nfs41_sequence_done] Signed-off-by: Andy Adamson [nfs41: remove nfs4_has_session check in nfs41_sequence_done] Signed-off-by: Andy Adamson [nfs41: remove nfs_client pointer check] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 58 +++++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 58 insertions(+) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 861b103c9e3c..188f2ce593ec 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -330,6 +330,46 @@ void nfs41_sequence_free_slot(const struct nfs_client *clp, res->sr_slotid = NFS4_MAX_SLOT_TABLE; } +static void nfs41_sequence_done(struct nfs_client *clp, + struct nfs4_sequence_res *res, + int rpc_status) +{ + unsigned long timestamp; + struct nfs4_slot_table *tbl; + struct nfs4_slot *slot; + + /* + * sr_status remains 1 if an RPC level error occurred. The server + * may or may not have processed the sequence operation.. + * Proceed as if the server received and processed the sequence + * operation. + */ + if (res->sr_status == 1) + res->sr_status = NFS_OK; + + /* -ERESTARTSYS can result in skipping nfs41_sequence_setup */ + if (res->sr_slotid == NFS4_MAX_SLOT_TABLE) + goto out; + + tbl = &clp->cl_session->fc_slot_table; + slot = tbl->slots + res->sr_slotid; + + if (res->sr_status == 0) { + /* Update the slot's sequence and clientid lease timer */ + ++slot->seq_nr; + timestamp = res->sr_renewal_time; + spin_lock(&clp->cl_lock); + if (time_before(clp->cl_last_renewal, timestamp)) + clp->cl_last_renewal = timestamp; + spin_unlock(&clp->cl_lock); + return; + } +out: + /* The session may be reset by one of the error handlers. */ + dprintk("%s: Error %d free the slot \n", __func__, res->sr_status); + nfs41_sequence_free_slot(clp, res); +} + /* * nfs4_find_slot - efficiently look for a free slot * @@ -515,6 +555,24 @@ int _nfs4_call_sync(struct nfs_server *server, (server)->nfs_client->cl_call_sync((server), (msg), &(args)->seq_args, \ &(res)->seq_res, (cache_reply)) +static void nfs4_sequence_done(const struct nfs_server *server, + struct nfs4_sequence_res *res, int rpc_status) +{ +#ifdef CONFIG_NFS_V4_1 + if (nfs4_has_session(server->nfs_client)) + nfs41_sequence_done(server->nfs_client, res, rpc_status); +#endif /* CONFIG_NFS_V4_1 */ +} + +/* no restart, therefore free slot here */ +static void nfs4_sequence_done_free_slot(const struct nfs_server *server, + struct nfs4_sequence_res *res, + int rpc_status) +{ + nfs4_sequence_done(server, res, rpc_status); + nfs4_sequence_free_slot(server->nfs_client, res); +} + static void update_changeattr(struct inode *dir, struct nfs4_change_info *cinfo) { struct nfs_inode *nfsi = NFS_I(dir); -- cgit v1.2.3 From 69ab40c4c3b9622232e31e0ee145875a2853c113 Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Wed, 1 Apr 2009 09:22:19 -0400 Subject: nfs41: nfs41_call_sync_done Implement nfs4.1 synchronous rpc_call_done method that essentially just calls nfs4_sequence_done, that turns around and calls nfs41_sequence_done for minorversion1 rpcs. Signed-off-by: Benny Halevy [nfs41: check for session not minorversion] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [move adding nfs4_sequence_free_slot from nfs41-separate-free-slot-from-sequence-done] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: nfs41_call_sync_data use nfs_client not nfs_server] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 9 +++++++++ 1 file changed, 9 insertions(+) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 188f2ce593ec..a8e0b3573594 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -492,8 +492,17 @@ static void nfs41_call_sync_prepare(struct rpc_task *task, void *calldata) rpc_call_start(task); } +static void nfs41_call_sync_done(struct rpc_task *task, void *calldata) +{ + struct nfs41_call_sync_data *data = calldata; + + nfs41_sequence_done(data->clp, data->seq_res, task->tk_status); + nfs41_sequence_free_slot(data->clp, data->seq_res); +} + struct rpc_call_ops nfs41_call_sync_ops = { .rpc_call_prepare = nfs41_call_sync_prepare, + .rpc_call_done = nfs41_call_sync_done, }; static int nfs4_call_sync_sequence(struct nfs_client *clp, -- cgit v1.2.3 From 19ddab06ed923d27339b8009d37b0ab3de93cadf Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Wed, 1 Apr 2009 09:22:20 -0400 Subject: nfs41: close sequence setup/done support Separate nfs4_close calls from nfs41: sequence setup/done support Call nfs4_sequence_done from respective rpc_call_done methods. Note that we need to pass a pointer to the nfs_server in calls data for passing on to nfs4_sequence_done. Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [pnfs: client data server write validate and release] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: separate free slot from sequence done] [nfs41: sequence res use slotid] [nfs41: remove SEQ4_STATUS_USE_TK_STATUS] [nfs41: nfs4_sequence_free_slot use nfs_client for data server] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index a8e0b3573594..a34fada51446 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -1630,6 +1630,7 @@ static void nfs4_close_done(struct rpc_task *task, void *data) struct nfs4_state *state = calldata->state; struct nfs_server *server = NFS_SERVER(calldata->inode); + nfs4_sequence_done(server, &calldata->res.seq_res, task->tk_status); if (RPC_ASSASSINATED(task)) return; /* hmm. we are done with the inode, and in the process of freeing @@ -1652,6 +1653,7 @@ static void nfs4_close_done(struct rpc_task *task, void *data) return; } } + nfs4_sequence_free_slot(server->nfs_client, &calldata->res.seq_res); nfs_refresh_inode(calldata->inode, calldata->res.fattr); } @@ -1692,6 +1694,10 @@ static void nfs4_close_prepare(struct rpc_task *task, void *data) calldata->arg.fmode = FMODE_WRITE; } calldata->timestamp = jiffies; + if (nfs4_setup_sequence((NFS_SERVER(calldata->inode))->nfs_client, + &calldata->arg.seq_args, &calldata->res.seq_res, + 1, task)) + return; rpc_call_start(task); } @@ -1731,7 +1737,7 @@ int nfs4_do_close(struct path *path, struct nfs4_state *state, int wait) }; int status = -ENOMEM; - calldata = kmalloc(sizeof(*calldata), GFP_KERNEL); + calldata = kzalloc(sizeof(*calldata), GFP_KERNEL); if (calldata == NULL) goto out; calldata->inode = state->inode; -- cgit v1.2.3 From d898528cdb338b145c71049c3cb7b1dfa542a48a Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Wed, 1 Apr 2009 09:22:21 -0400 Subject: nfs41: open sequence setup/done support Separate nfs4_open calls from nfs41: sequence setup/done support Call nfs4_sequence_done from respective rpc_call_done methods. Note that we need to pass a pointer to the nfs_server in calls data for passing on to nfs4_sequence_done. Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [pnfs: client data server write validate and release] [use nfs4_sequence_done_free_slot] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 8 ++++++++ 1 file changed, 8 insertions(+) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index a34fada51446..46e2509452a5 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -1241,6 +1241,10 @@ static void nfs4_open_prepare(struct rpc_task *task, void *calldata) nfs_copy_fh(&data->o_res.fh, data->o_arg.fh); } data->timestamp = jiffies; + if (nfs4_setup_sequence(data->o_arg.server->nfs_client, + &data->o_arg.seq_args, + &data->o_res.seq_res, 1, task)) + return; rpc_call_start(task); return; out_no_action: @@ -1253,6 +1257,10 @@ static void nfs4_open_done(struct rpc_task *task, void *calldata) struct nfs4_opendata *data = calldata; data->rpc_status = task->tk_status; + + nfs4_sequence_done_free_slot(data->o_arg.server, &data->o_res.seq_res, + task->tk_status); + if (RPC_ASSASSINATED(task)) return; if (task->tk_status == 0) { -- cgit v1.2.3 From 66179efee31f431ab538016b4ac05edaf87525dc Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Wed, 1 Apr 2009 09:22:22 -0400 Subject: nfs41: lock sequence setup/done support Separate nfs4_lock calls from nfs41: sequence setup/done support Call nfs4_sequence_done from respective rpc_call_done methods. Note that we need to pass a pointer to the nfs_server in calls data for passing on to nfs4_sequence_done. Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [pnfs: client data server write validate and release] [use nfs4_sequence_done_free_slot] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 8 ++++++++ 1 file changed, 8 insertions(+) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 46e2509452a5..dc070819096a 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -3688,6 +3688,7 @@ struct nfs4_lockdata { unsigned long timestamp; int rpc_status; int cancelled; + struct nfs_server *server; }; static struct nfs4_lockdata *nfs4_alloc_lockdata(struct file_lock *fl, @@ -3715,6 +3716,7 @@ static struct nfs4_lockdata *nfs4_alloc_lockdata(struct file_lock *fl, p->res.lock_seqid = p->arg.lock_seqid; p->res.seq_res.sr_slotid = NFS4_MAX_SLOT_TABLE; p->lsp = lsp; + p->server = server; atomic_inc(&lsp->ls_count); p->ctx = get_nfs_open_context(ctx); memcpy(&p->fl, fl, sizeof(p->fl)); @@ -3744,6 +3746,9 @@ static void nfs4_lock_prepare(struct rpc_task *task, void *calldata) } else data->arg.new_lock_owner = 0; data->timestamp = jiffies; + if (nfs4_setup_sequence(data->server->nfs_client, &data->arg.seq_args, + &data->res.seq_res, 1, task)) + return; rpc_call_start(task); dprintk("%s: done!, ret = %d\n", __func__, data->rpc_status); } @@ -3754,6 +3759,9 @@ static void nfs4_lock_done(struct rpc_task *task, void *calldata) dprintk("%s: begin!\n", __func__); + nfs4_sequence_done_free_slot(data->server, &data->res.seq_res, + task->tk_status); + data->rpc_status = task->tk_status; if (RPC_ASSASSINATED(task)) goto out; -- cgit v1.2.3 From a893693c151433a1375060ecb13419080426bc08 Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Wed, 1 Apr 2009 09:22:23 -0400 Subject: nfs41: locku sequence setup/done support Separate nfs4_locku calls from nfs41: sequence setup/done support Call nfs4_sequence_done from respective rpc_call_done methods. Note that we need to pass a pointer to the nfs_server in calls data for passing on to nfs4_sequence_done. Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [pnfs: client data server write validate and release] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: nfs4_sequence_free_slot use nfs_client for data server] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index dc070819096a..d2cfa4f86537 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -3533,7 +3533,7 @@ static struct nfs4_unlockdata *nfs4_alloc_unlockdata(struct file_lock *fl, struct nfs4_unlockdata *p; struct inode *inode = lsp->ls_state->inode; - p = kmalloc(sizeof(*p), GFP_KERNEL); + p = kzalloc(sizeof(*p), GFP_KERNEL); if (p == NULL) return NULL; p->arg.fh = NFS_FH(inode); @@ -3564,6 +3564,8 @@ static void nfs4_locku_done(struct rpc_task *task, void *data) { struct nfs4_unlockdata *calldata = data; + nfs4_sequence_done(calldata->server, &calldata->res.seq_res, + task->tk_status); if (RPC_ASSASSINATED(task)) return; switch (task->tk_status) { @@ -3582,6 +3584,8 @@ static void nfs4_locku_done(struct rpc_task *task, void *data) if (nfs4_async_handle_error(task, calldata->server, NULL) == -EAGAIN) rpc_restart_call(task); } + nfs4_sequence_free_slot(calldata->server->nfs_client, + &calldata->res.seq_res); } static void nfs4_locku_prepare(struct rpc_task *task, void *data) @@ -3596,6 +3600,10 @@ static void nfs4_locku_prepare(struct rpc_task *task, void *data) return; } calldata->timestamp = jiffies; + if (nfs4_setup_sequence(calldata->server->nfs_client, + &calldata->arg.seq_args, + &calldata->res.seq_res, 1, task)) + return; rpc_call_start(task); } -- cgit v1.2.3 From 472cfbd9b97be210a9b61ac1c6a774cd456ea9d2 Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Wed, 1 Apr 2009 09:22:24 -0400 Subject: nfs41: unlink sequence setup/done support Implement the rpc_call_prepare methods for asynchronuos nfs rpcs, call nfs41_setup_sequence from respective rpc_call_validate_args methods. Call nfs4_sequence_done from respective rpc_call_done methods. Note that we need to pass a pointer to the nfs_server in calls data for passing on to nfs4_sequence_done. Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [pnfs: client data server write validate and release] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: separate free slot from sequence done] [nfs41: sequence res use slotid] [nfs41: remove SEQ4_STATUS_USE_TK_STATUS] [nfs41: nfs4_sequence_free_slot use nfs_client for data server] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 2 ++ fs/nfs/unlink.c | 17 +++++++++++++++++ 2 files changed, 19 insertions(+) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index d2cfa4f86537..97521e47e437 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -2424,8 +2424,10 @@ static int nfs4_proc_unlink_done(struct rpc_task *task, struct inode *dir) { struct nfs_removeres *res = task->tk_msg.rpc_resp; + nfs4_sequence_done(res->server, &res->seq_res, task->tk_status); if (nfs4_async_handle_error(task, res->server, NULL) == -EAGAIN) return 0; + nfs4_sequence_free_slot(res->server->nfs_client, &res->seq_res); update_changeattr(dir, &res->cinfo); nfs_post_op_update_inode(dir, &res->dir_attr); return 1; diff --git a/fs/nfs/unlink.c b/fs/nfs/unlink.c index 83d0ce2600ab..089a21b4e3f9 100644 --- a/fs/nfs/unlink.c +++ b/fs/nfs/unlink.c @@ -15,6 +15,7 @@ #include #include "internal.h" +#include "nfs4_fs.h" struct nfs_unlinkdata { struct hlist_node list; @@ -102,9 +103,25 @@ static void nfs_async_unlink_release(void *calldata) nfs_sb_deactive(sb); } +#if defined(CONFIG_NFS_V4_1) +void nfs_unlink_prepare(struct rpc_task *task, void *calldata) +{ + struct nfs_unlinkdata *data = calldata; + struct nfs_server *server = NFS_SERVER(data->dir); + + if (nfs4_setup_sequence(server->nfs_client, &data->args.seq_args, + &data->res.seq_res, 1, task)) + return; + rpc_call_start(task); +} +#endif /* CONFIG_NFS_V4_1 */ + static const struct rpc_call_ops nfs_unlink_ops = { .rpc_call_done = nfs_async_unlink_done, .rpc_release = nfs_async_unlink_release, +#if defined(CONFIG_NFS_V4_1) + .rpc_call_prepare = nfs_unlink_prepare, +#endif /* CONFIG_NFS_V4_1 */ }; static int nfs_do_call_unlink(struct dentry *parent, struct inode *dir, struct nfs_unlinkdata *data) -- cgit v1.2.3 From f11c88af26453aee2823a1fd9120d0cd8dae7b9a Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Wed, 1 Apr 2009 09:22:25 -0400 Subject: nfs41: read sequence setup/done support Implement the read rpc_call_prepare method for asynchronuos nfs rpcs, call nfs41_setup_sequence from respective rpc_call_validate_args methods. Call nfs4_sequence_done from respective rpc_call_done methods. Note that we need to pass a pointer to the nfs_server in calls data for passing on to nfs4_sequence_done. Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [pnfs: client data server write validate and release] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [move the nfs4_sequence_free_slot call in nfs_readpage_retry from] [nfs41: separate free slot from sequence done] [remove nfs_readargs.nfs_server, use calldata->inode instead] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: Support sessions with O_DIRECT] Signed-off-by: Dean Hildebrand Signed-off-by: Benny Halevy [nfs41: nfs4_sequence_free_slot use nfs_client for data server] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/direct.c | 3 +++ fs/nfs/internal.h | 3 +++ fs/nfs/nfs4proc.c | 5 +++++ fs/nfs/read.c | 30 ++++++++++++++++++++++++++++-- 4 files changed, 39 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/nfs/direct.c b/fs/nfs/direct.c index 08f6b040d289..c8c53a03a585 100644 --- a/fs/nfs/direct.c +++ b/fs/nfs/direct.c @@ -259,6 +259,9 @@ static void nfs_direct_read_release(void *calldata) } static const struct rpc_call_ops nfs_read_direct_ops = { +#if defined(CONFIG_NFS_V4_1) + .rpc_call_prepare = nfs_read_prepare, +#endif /* CONFIG_NFS_V4_1 */ .rpc_call_done = nfs_direct_read_result, .rpc_release = nfs_direct_read_release, }; diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h index 12f4c5e6fd39..988c8b9aa78b 100644 --- a/fs/nfs/internal.h +++ b/fs/nfs/internal.h @@ -206,6 +206,9 @@ extern int nfs4_path_walk(struct nfs_server *server, const char *path); #endif +/* read.c */ +extern void nfs_read_prepare(struct rpc_task *task, void *calldata); + /* nfs4proc.c */ extern int _nfs4_call_sync(struct nfs_server *server, struct rpc_message *msg, diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 97521e47e437..6c0ac539d9c7 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -2864,6 +2864,11 @@ static int nfs4_read_done(struct rpc_task *task, struct nfs_read_data *data) { struct nfs_server *server = NFS_SERVER(data->inode); + dprintk("--> %s\n", __func__); + + /* nfs4_sequence_free_slot called in the read rpc_call_done */ + nfs4_sequence_done(server, &data->res.seq_res, task->tk_status); + if (nfs4_async_handle_error(task, server, data->args.context->state) == -EAGAIN) { rpc_restart_call(task); return -EAGAIN; diff --git a/fs/nfs/read.c b/fs/nfs/read.c index 70ba2b4cb9a4..d71f0d55ebde 100644 --- a/fs/nfs/read.c +++ b/fs/nfs/read.c @@ -22,6 +22,7 @@ #include +#include "nfs4_fs.h" #include "internal.h" #include "iostat.h" #include "fscache.h" @@ -358,19 +359,25 @@ static void nfs_readpage_retry(struct rpc_task *task, struct nfs_read_data *data struct nfs_readres *resp = &data->res; if (resp->eof || resp->count == argp->count) - return; + goto out; /* This is a short read! */ nfs_inc_stats(data->inode, NFSIOS_SHORTREAD); /* Has the server at least made some progress? */ if (resp->count == 0) - return; + goto out; /* Yes, so retry the read at the end of the data */ argp->offset += resp->count; argp->pgbase += resp->count; argp->count -= resp->count; rpc_restart_call(task); + return; +out: + nfs4_sequence_free_slot(NFS_SERVER(data->inode)->nfs_client, + &data->res.seq_res); + return; + } /* @@ -407,7 +414,23 @@ static void nfs_readpage_release_partial(void *calldata) nfs_readdata_release(calldata); } +#if defined(CONFIG_NFS_V4_1) +void nfs_read_prepare(struct rpc_task *task, void *calldata) +{ + struct nfs_read_data *data = calldata; + + if (nfs4_setup_sequence(NFS_SERVER(data->inode)->nfs_client, + &data->args.seq_args, &data->res.seq_res, + 0, task)) + return; + rpc_call_start(task); +} +#endif /* CONFIG_NFS_V4_1 */ + static const struct rpc_call_ops nfs_read_partial_ops = { +#if defined(CONFIG_NFS_V4_1) + .rpc_call_prepare = nfs_read_prepare, +#endif /* CONFIG_NFS_V4_1 */ .rpc_call_done = nfs_readpage_result_partial, .rpc_release = nfs_readpage_release_partial, }; @@ -471,6 +494,9 @@ static void nfs_readpage_release_full(void *calldata) } static const struct rpc_call_ops nfs_read_full_ops = { +#if defined(CONFIG_NFS_V4_1) + .rpc_call_prepare = nfs_read_prepare, +#endif /* CONFIG_NFS_V4_1 */ .rpc_call_done = nfs_readpage_result_full, .rpc_release = nfs_readpage_release_full, }; -- cgit v1.2.3 From def6ed7ef45ed19c3d6ca765f3bfdff1fe4c6bba Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Wed, 1 Apr 2009 09:22:26 -0400 Subject: nfs41 write sequence setup done support Separate write calls from nfs41: sequence setup/done support Implement the write rpc_call_prepare method for asynchronuos nfs rpcs, call nfs41_setup_sequence from respective rpc_call_validate_args methods. Call nfs4_sequence_done from respective rpc_call_done methods. Note that we need to pass a pointer to the nfs_server in calls data for passing on to nfs4_sequence_done. Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [pnfs: client data server write validate and release] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [move the nfs4_sequence_free_slot call in nfs_readpage_retry from] [nfs41: separate free slot from sequence done Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: Support sessions with O_DIRECT.] Signed-off-by: Dean Hildebrand Signed-off-by: Benny Halevy [nfs41: nfs4_sequence_free_slot use nfs_client for data server] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/direct.c | 3 +++ fs/nfs/internal.h | 3 +++ fs/nfs/nfs4proc.c | 4 ++++ fs/nfs/write.c | 22 ++++++++++++++++++++++ 4 files changed, 32 insertions(+) (limited to 'fs') diff --git a/fs/nfs/direct.c b/fs/nfs/direct.c index c8c53a03a585..1955bfeb3551 100644 --- a/fs/nfs/direct.c +++ b/fs/nfs/direct.c @@ -676,6 +676,9 @@ out_unlock: } static const struct rpc_call_ops nfs_write_direct_ops = { +#if defined(CONFIG_NFS_V4_1) + .rpc_call_prepare = nfs_write_prepare, +#endif /* CONFIG_NFS_V4_1 */ .rpc_call_done = nfs_direct_write_result, .rpc_release = nfs_direct_write_release, }; diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h index 988c8b9aa78b..f62bc5226155 100644 --- a/fs/nfs/internal.h +++ b/fs/nfs/internal.h @@ -209,6 +209,9 @@ extern int nfs4_path_walk(struct nfs_server *server, /* read.c */ extern void nfs_read_prepare(struct rpc_task *task, void *calldata); +/* write.c */ +extern void nfs_write_prepare(struct rpc_task *task, void *calldata); + /* nfs4proc.c */ extern int _nfs4_call_sync(struct nfs_server *server, struct rpc_message *msg, diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 6c0ac539d9c7..5f71832436fd 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -2890,6 +2890,10 @@ static int nfs4_write_done(struct rpc_task *task, struct nfs_write_data *data) { struct inode *inode = data->inode; + /* slot is freed in nfs_writeback_done */ + nfs4_sequence_done(NFS_SERVER(inode), &data->res.seq_res, + task->tk_status); + if (nfs4_async_handle_error(task, NFS_SERVER(inode), data->args.context->state) == -EAGAIN) { rpc_restart_call(task); return -EAGAIN; diff --git a/fs/nfs/write.c b/fs/nfs/write.c index 035e6fb9f57e..195fe7667529 100644 --- a/fs/nfs/write.c +++ b/fs/nfs/write.c @@ -25,6 +25,7 @@ #include "delegation.h" #include "internal.h" #include "iostat.h" +#include "nfs4_fs.h" #define NFSDBG_FACILITY NFSDBG_PAGECACHE @@ -1050,7 +1051,23 @@ out: nfs_writedata_release(calldata); } +#if defined(CONFIG_NFS_V4_1) +void nfs_write_prepare(struct rpc_task *task, void *calldata) +{ + struct nfs_write_data *data = calldata; + struct nfs_client *clp = (NFS_SERVER(data->inode))->nfs_client; + + if (nfs4_setup_sequence(clp, &data->args.seq_args, + &data->res.seq_res, 1, task)) + return; + rpc_call_start(task); +} +#endif /* CONFIG_NFS_V4_1 */ + static const struct rpc_call_ops nfs_write_partial_ops = { +#if defined(CONFIG_NFS_V4_1) + .rpc_call_prepare = nfs_write_prepare, +#endif /* CONFIG_NFS_V4_1 */ .rpc_call_done = nfs_writeback_done_partial, .rpc_release = nfs_writeback_release_partial, }; @@ -1113,6 +1130,9 @@ remove_request: } static const struct rpc_call_ops nfs_write_full_ops = { +#if defined(CONFIG_NFS_V4_1) + .rpc_call_prepare = nfs_write_prepare, +#endif /* CONFIG_NFS_V4_1 */ .rpc_call_done = nfs_writeback_done_full, .rpc_release = nfs_writeback_release_full, }; @@ -1195,6 +1215,8 @@ int nfs_writeback_done(struct rpc_task *task, struct nfs_write_data *data) /* Can't do anything about it except throw an error. */ task->tk_status = -EIO; } + nfs4_sequence_free_slot(NFS_SERVER(data->inode)->nfs_client, + &data->res.seq_res); return 0; } -- cgit v1.2.3 From 21d9a851aaa4161a9fddde720594659f3dae7fdd Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Wed, 1 Apr 2009 09:22:27 -0400 Subject: nfs41 commit sequence setup done support Separate commit calls from nfs41: sequence setup/done support Implement the commit rpc_call_prepare method for asynchronuos nfs rpcs, call nfs41_setup_sequence from respective rpc_call_validate_args methods. Call nfs4_sequence_done from respective rpc_call_done methods. Note that we need to pass a pointer to the nfs_server in calls data for passing on to nfs4_sequence_done. Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [pnfs: client data server write validate and release] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: Support sessions with O_DIRECT.] Signed-off-by: Dean Hildebrand Signed-off-by: Benny Halevy [nfs41: separate free slot from sequence done] [nfs41: nfs4_sequence_free_slot use nfs_client for data server] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/direct.c | 3 +++ fs/nfs/nfs4proc.c | 4 ++++ fs/nfs/write.c | 3 +++ 3 files changed, 10 insertions(+) (limited to 'fs') diff --git a/fs/nfs/direct.c b/fs/nfs/direct.c index 1955bfeb3551..489fc01a3204 100644 --- a/fs/nfs/direct.c +++ b/fs/nfs/direct.c @@ -538,6 +538,9 @@ static void nfs_direct_commit_release(void *calldata) } static const struct rpc_call_ops nfs_commit_direct_ops = { +#if defined(CONFIG_NFS_V4_1) + .rpc_call_prepare = nfs_write_prepare, +#endif /* CONFIG_NFS_V4_1 */ .rpc_call_done = nfs_direct_commit_result, .rpc_release = nfs_direct_commit_release, }; diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 5f71832436fd..fbd6274f14f8 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -2920,10 +2920,14 @@ static int nfs4_commit_done(struct rpc_task *task, struct nfs_write_data *data) { struct inode *inode = data->inode; + nfs4_sequence_done(NFS_SERVER(inode), &data->res.seq_res, + task->tk_status); if (nfs4_async_handle_error(task, NFS_SERVER(inode), NULL) == -EAGAIN) { rpc_restart_call(task); return -EAGAIN; } + nfs4_sequence_free_slot(NFS_SERVER(inode)->nfs_client, + &data->res.seq_res); nfs_refresh_inode(inode, data->res.fattr); return 0; } diff --git a/fs/nfs/write.c b/fs/nfs/write.c index 195fe7667529..85a76409de13 100644 --- a/fs/nfs/write.c +++ b/fs/nfs/write.c @@ -1373,6 +1373,9 @@ static void nfs_commit_release(void *calldata) } static const struct rpc_call_ops nfs_commit_ops = { +#if defined(CONFIG_NFS_V4_1) + .rpc_call_prepare = nfs_write_prepare, +#endif /* CONFIG_NFS_V4_1 */ .rpc_call_done = nfs_commit_done, .rpc_release = nfs_commit_release, }; -- cgit v1.2.3 From 938e10109115a71cc69d475122f21cf75e5046cd Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Wed, 1 Apr 2009 09:22:28 -0400 Subject: nfs41 delegreturn sequence setup done support Separate delegreturn calls from nfs41: sequence setup/done support Implement the delegreturn rpc_call_prepare method for asynchronuos nfs rpcs, call nfs41_setup_sequence from respective rpc_call_validate_args methods. Call nfs4_sequence_done from respective rpc_call_done methods. Note that we need to pass a pointer to the nfs_server in calls data for passing on to nfs4_sequence_done. Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [pnfs: client data server write validate and release] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 24 +++++++++++++++++++++++- 1 file changed, 23 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index fbd6274f14f8..dc0feb5837b1 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -3360,6 +3360,10 @@ struct nfs4_delegreturndata { static void nfs4_delegreturn_done(struct rpc_task *task, void *calldata) { struct nfs4_delegreturndata *data = calldata; + + nfs4_sequence_done_free_slot(data->res.server, &data->res.seq_res, + task->tk_status); + data->rpc_status = task->tk_status; if (data->rpc_status == 0) renew_lease(data->res.server, data->timestamp); @@ -3370,7 +3374,25 @@ static void nfs4_delegreturn_release(void *calldata) kfree(calldata); } +#if defined(CONFIG_NFS_V4_1) +static void nfs4_delegreturn_prepare(struct rpc_task *task, void *data) +{ + struct nfs4_delegreturndata *d_data; + + d_data = (struct nfs4_delegreturndata *)data; + + if (nfs4_setup_sequence(d_data->res.server->nfs_client, + &d_data->args.seq_args, + &d_data->res.seq_res, 1, task)) + return; + rpc_call_start(task); +} +#endif /* CONFIG_NFS_V4_1 */ + static const struct rpc_call_ops nfs4_delegreturn_ops = { +#if defined(CONFIG_NFS_V4_1) + .rpc_call_prepare = nfs4_delegreturn_prepare, +#endif /* CONFIG_NFS_V4_1 */ .rpc_call_done = nfs4_delegreturn_done, .rpc_release = nfs4_delegreturn_release, }; @@ -3392,7 +3414,7 @@ static int _nfs4_proc_delegreturn(struct inode *inode, struct rpc_cred *cred, co }; int status = 0; - data = kmalloc(sizeof(*data), GFP_KERNEL); + data = kzalloc(sizeof(*data), GFP_KERNEL); if (data == NULL) return -ENOMEM; data->args.fhandle = &data->fh; -- cgit v1.2.3 From 99fe60d062cfecf382c036065b3278b82b6c5eff Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Wed, 1 Apr 2009 09:22:29 -0400 Subject: nfs41: exchange_id operation Implement the exchange_id operation conforming to http://tools.ietf.org/html/draft-ietf-nfsv4-minorversion1-26 Unlike NFSv4.0, NFSv4.1 requires machine credentials. RPC_AUTH_GSS machine credentials will be passed into the kernel at mount time to be available for the exchange_id operation. RPC_AUTH_UNIX root mounts can use the UNIX root credential. Store the root credential in the nfs_client struct. Without a credential, NFSv4.1 state renewal fails. [nfs41: establish clientid via exchange id only if cred != NULL] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfsd41: move nfstime4 from under CONFIG_NFS_V4_1] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: do not wait a lease time in exchange id] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: pass *session in seq_args and seq_res] Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust [nfs41: Ignoring impid in decode_exchange_id is missing a READ_BUF] Signed-off-by: Benny Halevy [nfs41: fix Xcode_exchange_id's xdr Xcoding pointer type] [nfs41: get rid of unused struct nfs41_exchange_id_res members] Signed-off-by: Benny Halevy --- fs/nfs/nfs4proc.c | 61 ++++++++++++++++++++ fs/nfs/nfs4xdr.c | 139 +++++++++++++++++++++++++++++++++++++++++++++ include/linux/nfs4.h | 1 + include/linux/nfs_fs_sb.h | 6 ++ include/linux/nfs_xdr.h | 40 +++++++++++++ include/linux/nfsd/state.h | 1 - 6 files changed, 247 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index dc0feb5837b1..6f384e290753 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -48,6 +48,7 @@ #include #include #include +#include #include "nfs4_fs.h" #include "delegation.h" @@ -433,11 +434,13 @@ static int nfs41_setup_sequence(struct nfs4_session *session, spin_unlock(&tbl->slot_tbl_lock); slot = tbl->slots + slotid; + args->sa_session = session; args->sa_slotid = slotid; args->sa_cache_this = cache_reply; dprintk("<-- %s slotid=%d seqid=%d\n", __func__, slotid, slot->seq_nr); + res->sr_session = session; res->sr_slotid = slotid; res->sr_renewal_time = jiffies; /* @@ -4128,6 +4131,64 @@ int nfs4_proc_fs_locations(struct inode *dir, const struct qstr *name, } #ifdef CONFIG_NFS_V4_1 +/* + * nfs4_proc_exchange_id() + * + * Since the clientid has expired, all compounds using sessions + * associated with the stale clientid will be returning + * NFS4ERR_BADSESSION in the sequence operation, and will therefore + * be in some phase of session reset. + */ +static int nfs4_proc_exchange_id(struct nfs_client *clp, struct rpc_cred *cred) +{ + nfs4_verifier verifier; + struct nfs41_exchange_id_args args = { + .client = clp, + .flags = clp->cl_exchange_flags, + }; + struct nfs41_exchange_id_res res = { + .client = clp, + }; + int status; + struct rpc_message msg = { + .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_EXCHANGE_ID], + .rpc_argp = &args, + .rpc_resp = &res, + .rpc_cred = cred, + }; + __be32 *p; + + dprintk("--> %s\n", __func__); + BUG_ON(clp == NULL); + p = (u32 *)verifier.data; + *p++ = htonl((u32)clp->cl_boot_time.tv_sec); + *p = htonl((u32)clp->cl_boot_time.tv_nsec); + args.verifier = &verifier; + + while (1) { + args.id_len = scnprintf(args.id, sizeof(args.id), + "%s/%s %u", + clp->cl_ipaddr, + rpc_peeraddr2str(clp->cl_rpcclient, + RPC_DISPLAY_ADDR), + clp->cl_id_uniquifier); + + status = rpc_call_sync(clp->cl_rpcclient, &msg, 0); + + if (status != NFS4ERR_CLID_INUSE) + break; + + if (signalled()) + break; + + if (++clp->cl_id_uniquifier == 0) + break; + } + + dprintk("<-- %s status= %d\n", __func__, status); + return status; +} + /* Destroy the slot table */ static void nfs4_destroy_slot_table(struct nfs4_session *session) { diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index 5b944cd57218..783c4214dccd 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -246,6 +246,27 @@ static int nfs4_stat_to_errno(int); (0) #if defined(CONFIG_NFS_V4_1) +#define encode_exchange_id_maxsz (op_encode_hdr_maxsz + \ + encode_verifier_maxsz + \ + 1 /* co_ownerid.len */ + \ + XDR_QUADLEN(NFS4_EXCHANGE_ID_LEN) + \ + 1 /* flags */ + \ + 1 /* spa_how */ + \ + 0 /* SP4_NONE (for now) */ + \ + 1 /* zero implemetation id array */) +#define decode_exchange_id_maxsz (op_decode_hdr_maxsz + \ + 2 /* eir_clientid */ + \ + 1 /* eir_sequenceid */ + \ + 1 /* eir_flags */ + \ + 1 /* spr_how */ + \ + 0 /* SP4_NONE (for now) */ + \ + 2 /* eir_server_owner.so_minor_id */ + \ + /* eir_server_owner.so_major_id<> */ \ + XDR_QUADLEN(NFS4_OPAQUE_LIMIT) + 1 + \ + /* eir_server_scope<> */ \ + XDR_QUADLEN(NFS4_OPAQUE_LIMIT) + 1 + \ + 1 /* eir_server_impl_id array length */ + \ + 0 /* ignored eir_server_impl_id contents */) #define encode_sequence_maxsz 0 /* stub */ #define decode_sequence_maxsz 0 /* stub */ #else /* CONFIG_NFS_V4_1 */ @@ -594,6 +615,14 @@ static int nfs4_stat_to_errno(int); decode_putfh_maxsz + \ decode_lookup_maxsz + \ decode_fs_locations_maxsz) +#if defined(CONFIG_NFS_V4_1) +#define NFS4_enc_exchange_id_sz \ + (compound_encode_hdr_maxsz + \ + encode_exchange_id_maxsz) +#define NFS4_dec_exchange_id_sz \ + (compound_decode_hdr_maxsz + \ + decode_exchange_id_maxsz) +#endif /* CONFIG_NFS_V4_1 */ static const umode_t nfs_type2fmt[] = { [NF4BAD] = 0, @@ -1455,7 +1484,29 @@ static void encode_delegreturn(struct xdr_stream *xdr, const nfs4_stateid *state hdr->replen += decode_delegreturn_maxsz; } +#if defined(CONFIG_NFS_V4_1) /* NFSv4.1 operations */ +static void encode_exchange_id(struct xdr_stream *xdr, + struct nfs41_exchange_id_args *args, + struct compound_hdr *hdr) +{ + __be32 *p; + + RESERVE_SPACE(4 + sizeof(args->verifier->data)); + WRITE32(OP_EXCHANGE_ID); + WRITEMEM(args->verifier->data, sizeof(args->verifier->data)); + + encode_string(xdr, args->id_len, args->id); + + RESERVE_SPACE(12); + WRITE32(args->flags); + WRITE32(0); /* zero length state_protect4_a */ + WRITE32(0); /* zero length implementation id array */ + hdr->nops++; + hdr->replen += decode_exchange_id_maxsz; +} +#endif /* CONFIG_NFS_V4_1 */ + static void encode_sequence(struct xdr_stream *xdr, const struct nfs4_sequence_args *args, struct compound_hdr *hdr) @@ -2162,6 +2213,26 @@ static int nfs4_xdr_enc_fs_locations(struct rpc_rqst *req, __be32 *p, struct nfs return 0; } +#if defined(CONFIG_NFS_V4_1) +/* + * EXCHANGE_ID request + */ +static int nfs4_xdr_enc_exchange_id(struct rpc_rqst *req, uint32_t *p, + struct nfs41_exchange_id_args *args) +{ + struct xdr_stream xdr; + struct compound_hdr hdr = { + .minorversion = args->client->cl_minorversion, + }; + + xdr_init_encode(&xdr, &req->rq_snd_buf, p); + encode_compound_hdr(&xdr, req, &hdr); + encode_exchange_id(&xdr, args, &hdr); + encode_nops(&hdr); + return 0; +} +#endif /* CONFIG_NFS_V4_1 */ + /* * START OF "GENERIC" DECODE ROUTINES. * These may look a little ugly since they are imported from a "generic" @@ -3877,6 +3948,52 @@ static int decode_delegreturn(struct xdr_stream *xdr) return decode_op_hdr(xdr, OP_DELEGRETURN); } +#if defined(CONFIG_NFS_V4_1) +static int decode_exchange_id(struct xdr_stream *xdr, + struct nfs41_exchange_id_res *res) +{ + __be32 *p; + uint32_t dummy; + int status; + struct nfs_client *clp = res->client; + + status = decode_op_hdr(xdr, OP_EXCHANGE_ID); + if (status) + return status; + + READ_BUF(8); + READ64(clp->cl_ex_clid); + READ_BUF(12); + READ32(clp->cl_seqid); + READ32(clp->cl_exchange_flags); + + /* We ask for SP4_NONE */ + READ32(dummy); + if (dummy != SP4_NONE) + return -EIO; + + /* Throw away minor_id */ + READ_BUF(8); + + /* Throw away Major id */ + READ_BUF(4); + READ32(dummy); + READ_BUF(dummy); + + /* Throw away server_scope */ + READ_BUF(4); + READ32(dummy); + READ_BUF(dummy); + + /* Throw away Implementation id array */ + READ_BUF(4); + READ32(dummy); + READ_BUF(dummy); + + return 0; +} +#endif /* CONFIG_NFS_V4_1 */ + static int decode_sequence(struct xdr_stream *xdr, struct nfs4_sequence_res *res, struct rpc_rqst *rqstp) @@ -4774,6 +4891,25 @@ out: return status; } +#if defined(CONFIG_NFS_V4_1) +/* + * EXCHANGE_ID request + */ +static int nfs4_xdr_dec_exchange_id(struct rpc_rqst *rqstp, uint32_t *p, + void *res) +{ + struct xdr_stream xdr; + struct compound_hdr hdr; + int status; + + xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); + status = decode_compound_hdr(&xdr, &hdr); + if (!status) + status = decode_exchange_id(&xdr, res); + return status; +} +#endif /* CONFIG_NFS_V4_1 */ + __be32 *nfs4_decode_dirent(__be32 *p, struct nfs_entry *entry, int plus) { uint32_t bitmap[2] = {0}; @@ -4943,6 +5079,9 @@ struct rpc_procinfo nfs4_procedures[] = { PROC(GETACL, enc_getacl, dec_getacl), PROC(SETACL, enc_setacl, dec_setacl), PROC(FS_LOCATIONS, enc_fs_locations, dec_fs_locations), +#if defined(CONFIG_NFS_V4_1) + PROC(EXCHANGE_ID, enc_exchange_id, dec_exchange_id), +#endif /* CONFIG_NFS_V4_1 */ }; struct rpc_version nfs_version4 = { diff --git a/include/linux/nfs4.h b/include/linux/nfs4.h index 7c36fcf2dfb7..ad65709ed8d3 100644 --- a/include/linux/nfs4.h +++ b/include/linux/nfs4.h @@ -21,6 +21,7 @@ #define NFS4_FHSIZE 128 #define NFS4_MAXPATHLEN PATH_MAX #define NFS4_MAXNAMLEN NAME_MAX +#define NFS4_OPAQUE_LIMIT 1024 #define NFS4_MAX_SESSIONID_LEN 16 #define NFS4_ACCESS_READ 0x0001 diff --git a/include/linux/nfs_fs_sb.h b/include/linux/nfs_fs_sb.h index 206485e5082f..435ed556efb5 100644 --- a/include/linux/nfs_fs_sb.h +++ b/include/linux/nfs_fs_sb.h @@ -78,6 +78,12 @@ struct nfs_client { #endif /* CONFIG_NFS_V4 */ #ifdef CONFIG_NFS_V4_1 + /* clientid returned from EXCHANGE_ID, used by session operations */ + u64 cl_ex_clid; + /* The sequence id to use for the next CREATE_SESSION */ + u32 cl_seqid; + /* The flags used for obtaining the clientid during EXCHANGE_ID */ + u32 cl_exchange_flags; struct nfs4_session *cl_session; /* sharred session */ #endif /* CONFIG_NFS_V4_1 */ diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h index 4ac14b40efc9..5d70b924af5e 100644 --- a/include/linux/nfs_xdr.h +++ b/include/linux/nfs_xdr.h @@ -879,6 +879,46 @@ struct nfs4_fs_locations_res { #endif /* CONFIG_NFS_V4 */ +struct nfstime4 { + u64 seconds; + u32 nseconds; +}; + +#ifdef CONFIG_NFS_V4_1 +struct nfs_impl_id4 { + u32 domain_len; + char *domain; + u32 name_len; + char *name; + struct nfstime4 date; +}; + +#define NFS4_EXCHANGE_ID_LEN (48) +struct nfs41_exchange_id_args { + struct nfs_client *client; + nfs4_verifier *verifier; + unsigned int id_len; + char id[NFS4_EXCHANGE_ID_LEN]; + u32 flags; +}; + +struct server_owner { + uint64_t minor_id; + uint32_t major_id_sz; + char major_id[NFS4_OPAQUE_LIMIT]; +}; + +struct server_scope { + uint32_t server_scope_sz; + char server_scope[NFS4_OPAQUE_LIMIT]; +}; + +struct nfs41_exchange_id_res { + struct nfs_client *client; + u32 flags; +}; +#endif /* CONFIG_NFS_V4_1 */ + struct nfs_page; #define NFS_PAGEVEC_SIZE (8U) diff --git a/include/linux/nfsd/state.h b/include/linux/nfsd/state.h index 4d61c873feed..7ef4b7ad1214 100644 --- a/include/linux/nfsd/state.h +++ b/include/linux/nfsd/state.h @@ -41,7 +41,6 @@ #include #include -#define NFS4_OPAQUE_LIMIT 1024 typedef struct { u32 cl_boot; u32 cl_id; -- cgit v1.2.3 From 2050f0cc0703aab7cee798b3cb47037754f368bc Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Wed, 1 Apr 2009 09:22:30 -0400 Subject: nfs41: get_lease_time get_lease_time uses the FSINFO rpc operation to get the lease time attribute. nfs4_get_lease_time() is only called from the state manager on session setup so don't recover from clientid or sequence level errors. We do need to recover from NFS4ERR_DELAY or NFS4ERR_GRACE. Use NFS4_POLL_RETRY_MIN - the Linux server returns NFS4ERR_DELAY when an upcall is needed to resolve an uncached export referenced by a file handle. [nfs41: sequence res use slotid] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: remove extraneous rpc_clnt pointer] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: have get_lease_time work on nfs_client] Signed-off-by: Benny Halevy [nfs41: get_lease_time recover from NFS4ERR_DELAY] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: pass *session in seq_args and seq_res] [define nfs4_get_lease_time_{args,res}] Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 94 +++++++++++++++++++++++++++++++++++++++++++++++++ fs/nfs/nfs4xdr.c | 51 +++++++++++++++++++++++++++ include/linux/nfs_xdr.h | 9 +++++ 3 files changed, 154 insertions(+) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 6f384e290753..eafc99afd356 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -4189,6 +4189,100 @@ static int nfs4_proc_exchange_id(struct nfs_client *clp, struct rpc_cred *cred) return status; } +struct nfs4_get_lease_time_data { + struct nfs4_get_lease_time_args *args; + struct nfs4_get_lease_time_res *res; + struct nfs_client *clp; +}; + +static void nfs4_get_lease_time_prepare(struct rpc_task *task, + void *calldata) +{ + int ret; + struct nfs4_get_lease_time_data *data = + (struct nfs4_get_lease_time_data *)calldata; + + dprintk("--> %s\n", __func__); + /* just setup sequence, do not trigger session recovery + since we're invoked within one */ + ret = nfs41_setup_sequence(data->clp->cl_session, + &data->args->la_seq_args, + &data->res->lr_seq_res, 0, task); + + BUG_ON(ret == -EAGAIN); + rpc_call_start(task); + dprintk("<-- %s\n", __func__); +} + +/* + * Called from nfs4_state_manager thread for session setup, so don't recover + * from sequence operation or clientid errors. + */ +static void nfs4_get_lease_time_done(struct rpc_task *task, void *calldata) +{ + struct nfs4_get_lease_time_data *data = + (struct nfs4_get_lease_time_data *)calldata; + + dprintk("--> %s\n", __func__); + nfs41_sequence_done(data->clp, &data->res->lr_seq_res, task->tk_status); + switch (task->tk_status) { + case -NFS4ERR_DELAY: + case -NFS4ERR_GRACE: + dprintk("%s Retry: tk_status %d\n", __func__, task->tk_status); + rpc_delay(task, NFS4_POLL_RETRY_MIN); + task->tk_status = 0; + rpc_restart_call(task); + return; + } + nfs41_sequence_free_slot(data->clp, &data->res->lr_seq_res); + dprintk("<-- %s\n", __func__); +} + +struct rpc_call_ops nfs4_get_lease_time_ops = { + .rpc_call_prepare = nfs4_get_lease_time_prepare, + .rpc_call_done = nfs4_get_lease_time_done, +}; + +int nfs4_proc_get_lease_time(struct nfs_client *clp, struct nfs_fsinfo *fsinfo) +{ + struct rpc_task *task; + struct nfs4_get_lease_time_args args; + struct nfs4_get_lease_time_res res = { + .lr_fsinfo = fsinfo, + }; + struct nfs4_get_lease_time_data data = { + .args = &args, + .res = &res, + .clp = clp, + }; + struct rpc_message msg = { + .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_GET_LEASE_TIME], + .rpc_argp = &args, + .rpc_resp = &res, + }; + struct rpc_task_setup task_setup = { + .rpc_client = clp->cl_rpcclient, + .rpc_message = &msg, + .callback_ops = &nfs4_get_lease_time_ops, + .callback_data = &data + }; + int status; + + res.lr_seq_res.sr_slotid = NFS4_MAX_SLOT_TABLE; + dprintk("--> %s\n", __func__); + task = rpc_run_task(&task_setup); + + if (IS_ERR(task)) + status = PTR_ERR(task); + else { + status = task->tk_status; + rpc_put_task(task); + } + dprintk("<-- %s return %d\n", __func__, status); + + return status; +} + /* Destroy the slot table */ static void nfs4_destroy_slot_table(struct nfs4_session *session) { diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index 783c4214dccd..85ee1d17a461 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -622,6 +622,14 @@ static int nfs4_stat_to_errno(int); #define NFS4_dec_exchange_id_sz \ (compound_decode_hdr_maxsz + \ decode_exchange_id_maxsz) +#define NFS4_enc_get_lease_time_sz (compound_encode_hdr_maxsz + \ + encode_sequence_maxsz + \ + encode_putrootfh_maxsz + \ + encode_fsinfo_maxsz) +#define NFS4_dec_get_lease_time_sz (compound_decode_hdr_maxsz + \ + decode_sequence_maxsz + \ + decode_putrootfh_maxsz + \ + decode_fsinfo_maxsz) #endif /* CONFIG_NFS_V4_1 */ static const umode_t nfs_type2fmt[] = { @@ -2231,6 +2239,27 @@ static int nfs4_xdr_enc_exchange_id(struct rpc_rqst *req, uint32_t *p, encode_nops(&hdr); return 0; } + +/* + * a GET_LEASE_TIME request + */ +static int nfs4_xdr_enc_get_lease_time(struct rpc_rqst *req, uint32_t *p, + struct nfs4_get_lease_time_args *args) +{ + struct xdr_stream xdr; + struct compound_hdr hdr = { + .minorversion = nfs4_xdr_minorversion(&args->la_seq_args), + }; + const u32 lease_bitmap[2] = { FATTR4_WORD0_LEASE_TIME, 0 }; + + xdr_init_encode(&xdr, &req->rq_snd_buf, p); + encode_compound_hdr(&xdr, req, &hdr); + encode_sequence(&xdr, &args->la_seq_args, &hdr); + encode_putrootfh(&xdr, &hdr); + encode_fsinfo(&xdr, lease_bitmap, &hdr); + encode_nops(&hdr); + return 0; +} #endif /* CONFIG_NFS_V4_1 */ /* @@ -4908,6 +4937,27 @@ static int nfs4_xdr_dec_exchange_id(struct rpc_rqst *rqstp, uint32_t *p, status = decode_exchange_id(&xdr, res); return status; } + +/* + * a GET_LEASE_TIME request + */ +static int nfs4_xdr_dec_get_lease_time(struct rpc_rqst *rqstp, uint32_t *p, + struct nfs4_get_lease_time_res *res) +{ + struct xdr_stream xdr; + struct compound_hdr hdr; + int status; + + xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); + status = decode_compound_hdr(&xdr, &hdr); + if (!status) + status = decode_sequence(&xdr, &res->lr_seq_res, rqstp); + if (!status) + status = decode_putrootfh(&xdr); + if (!status) + status = decode_fsinfo(&xdr, res->lr_fsinfo); + return status; +} #endif /* CONFIG_NFS_V4_1 */ __be32 *nfs4_decode_dirent(__be32 *p, struct nfs_entry *entry, int plus) @@ -5081,6 +5131,7 @@ struct rpc_procinfo nfs4_procedures[] = { PROC(FS_LOCATIONS, enc_fs_locations, dec_fs_locations), #if defined(CONFIG_NFS_V4_1) PROC(EXCHANGE_ID, enc_exchange_id, dec_exchange_id), + PROC(GET_LEASE_TIME, enc_get_lease_time, dec_get_lease_time), #endif /* CONFIG_NFS_V4_1 */ }; diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h index 5d70b924af5e..ca643aa87d46 100644 --- a/include/linux/nfs_xdr.h +++ b/include/linux/nfs_xdr.h @@ -174,6 +174,15 @@ struct nfs4_sequence_res { int sr_status; /* sequence operation status */ }; +struct nfs4_get_lease_time_args { + struct nfs4_sequence_args la_seq_args; +}; + +struct nfs4_get_lease_time_res { + struct nfs_fsinfo *lr_fsinfo; + struct nfs4_sequence_res lr_seq_res; +}; + /* * Arguments to the open call. */ -- cgit v1.2.3 From fc931582c260e53ca5ca23bd70ccc9b2265cca9f Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Wed, 1 Apr 2009 09:22:31 -0400 Subject: nfs41: create_session operation Implement the create_session operation conforming to http://tools.ietf.org/html/draft-ietf-nfsv4-minorversion1-26 Set the real fore channel max operations to preserve server resources. Note: If the server returns < NFS4_MAX_OPS, the client will very soon get an NFS4ERR_TOO_MANY_OPS. A later patch will handle this. Set the max_rqst_sz and max_resp_sz to PAGE_SIZE - we preallocate the buffers. Set the back channel max_resp_sz_cached to zero to force the client to always set csa_cachethis to FALSE because the current implementation of the back channel DRC only supports caching the CB_SEQUENCE operation. The client back channel server supports one slot, and desires 2 operations per compound. Signed-off-by: Ricardo Labiaga Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: remove extraneous rpc_clnt pointer] Use the struct nfs_client cl_rpcclient. Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: nfs4_init_channel_attrs, just use nfs41_create_session_args] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: use rsize and wsize for session channel attributes] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: set channel max operations] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: set back channel attributes] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: obliterate nfs4_adjust_channel_attrs] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: have create_session work on nfs_client] Signed-off-by: Benny Halevy [nfs41: move CONFIG_NFS_V4_1 endif] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: pass *session in seq_args and seq_res] [moved nfs4_init_slot_table definition here] Signed-off-by: Benny Halevy [nfs41: use kcalloc to allocate slot table] Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust [nfs41: fix Xcode_create_session's xdr Xcoding pointer type] [nfs41: refactor decoding of channel attributes] Signed-off-by: Benny Halevy --- fs/nfs/nfs4proc.c | 172 ++++++++++++++++++++++++++++++++++++++++++++ fs/nfs/nfs4xdr.c | 185 ++++++++++++++++++++++++++++++++++++++++++++++++ include/linux/nfs4.h | 10 +++ include/linux/nfs_xdr.h | 12 ++++ 4 files changed, 379 insertions(+) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index eafc99afd356..7d81d6e57533 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -54,6 +54,7 @@ #include "delegation.h" #include "internal.h" #include "iostat.h" +#include "callback.h" #define NFSDBG_FACILITY NFSDBG_PROC @@ -4283,6 +4284,50 @@ int nfs4_proc_get_lease_time(struct nfs_client *clp, struct nfs_fsinfo *fsinfo) return status; } +/* + * Initialize slot table + */ +static int nfs4_init_slot_table(struct nfs4_session *session) +{ + struct nfs4_slot_table *tbl = &session->fc_slot_table; + int i, max_slots = session->fc_attrs.max_reqs; + struct nfs4_slot *slot; + int ret = -ENOMEM; + + BUG_ON(max_slots > NFS4_MAX_SLOT_TABLE); + + dprintk("--> %s: max_reqs=%u\n", __func__, + session->fc_attrs.max_reqs); + + slot = kcalloc(max_slots, sizeof(struct nfs4_slot), GFP_KERNEL); + if (!slot) + goto out; + for (i = 0; i < max_slots; ++i) + slot[i].seq_nr = 1; + ret = 0; + + spin_lock(&tbl->slot_tbl_lock); + if (tbl->slots != NULL) { + spin_unlock(&tbl->slot_tbl_lock); + dprintk("%s: slot table already initialized. tbl=%p slots=%p\n", + __func__, tbl, tbl->slots); + WARN_ON(1); + goto out_free; + } + tbl->max_slots = max_slots; + tbl->slots = slot; + tbl->highest_used_slotid = -1; /* no slot is currently used */ + spin_unlock(&tbl->slot_tbl_lock); + dprintk("%s: tbl=%p slots=%p max_slots=%d\n", __func__, + tbl, tbl->slots, tbl->max_slots); +out: + dprintk("<-- %s: return %d\n", __func__, ret); + return ret; +out_free: + kfree(slot); + goto out; +} + /* Destroy the slot table */ static void nfs4_destroy_slot_table(struct nfs4_session *session) { @@ -4314,6 +4359,133 @@ void nfs4_destroy_session(struct nfs4_session *session) kfree(session); } +/* + * Initialize the values to be used by the client in CREATE_SESSION + * If nfs4_init_session set the fore channel request and response sizes, + * use them. + * + * Set the back channel max_resp_sz_cached to zero to force the client to + * always set csa_cachethis to FALSE because the current implementation + * of the back channel DRC only supports caching the CB_SEQUENCE operation. + */ +static void nfs4_init_channel_attrs(struct nfs41_create_session_args *args) +{ + struct nfs4_session *session = args->client->cl_session; + unsigned int mxrqst_sz = session->fc_attrs.max_rqst_sz, + mxresp_sz = session->fc_attrs.max_resp_sz; + + if (mxrqst_sz == 0) + mxrqst_sz = NFS_MAX_FILE_IO_SIZE; + if (mxresp_sz == 0) + mxresp_sz = NFS_MAX_FILE_IO_SIZE; + /* Fore channel attributes */ + args->fc_attrs.headerpadsz = 0; + args->fc_attrs.max_rqst_sz = mxrqst_sz; + args->fc_attrs.max_resp_sz = mxresp_sz; + args->fc_attrs.max_resp_sz_cached = mxresp_sz; + args->fc_attrs.max_ops = NFS4_MAX_OPS; + args->fc_attrs.max_reqs = session->clp->cl_rpcclient->cl_xprt->max_reqs; + + dprintk("%s: Fore Channel : max_rqst_sz=%u max_resp_sz=%u " + "max_resp_sz_cached=%u max_ops=%u max_reqs=%u\n", + __func__, + args->fc_attrs.max_rqst_sz, args->fc_attrs.max_resp_sz, + args->fc_attrs.max_resp_sz_cached, args->fc_attrs.max_ops, + args->fc_attrs.max_reqs); + + /* Back channel attributes */ + args->bc_attrs.headerpadsz = 0; + args->bc_attrs.max_rqst_sz = PAGE_SIZE; + args->bc_attrs.max_resp_sz = PAGE_SIZE; + args->bc_attrs.max_resp_sz_cached = 0; + args->bc_attrs.max_ops = NFS4_MAX_BACK_CHANNEL_OPS; + args->bc_attrs.max_reqs = 1; + + dprintk("%s: Back Channel : max_rqst_sz=%u max_resp_sz=%u " + "max_resp_sz_cached=%u max_ops=%u max_reqs=%u\n", + __func__, + args->bc_attrs.max_rqst_sz, args->bc_attrs.max_resp_sz, + args->bc_attrs.max_resp_sz_cached, args->bc_attrs.max_ops, + args->bc_attrs.max_reqs); +} + +static int _nfs4_proc_create_session(struct nfs_client *clp) +{ + struct nfs4_session *session = clp->cl_session; + struct nfs41_create_session_args args = { + .client = clp, + .cb_program = NFS4_CALLBACK, + }; + struct nfs41_create_session_res res = { + .client = clp, + }; + struct rpc_message msg = { + .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_CREATE_SESSION], + .rpc_argp = &args, + .rpc_resp = &res, + }; + int status; + + nfs4_init_channel_attrs(&args); + args.flags = (SESSION4_PERSIST); + + status = rpc_call_sync(session->clp->cl_rpcclient, &msg, 0); + + /* Set the negotiated values in the session's channel_attrs struct */ + + if (!status) { + /* Increment the clientid slot sequence id */ + clp->cl_seqid++; + } + + return status; +} + +/* + * Issues a CREATE_SESSION operation to the server. + * It is the responsibility of the caller to verify the session is + * expired before calling this routine. + */ +int nfs4_proc_create_session(struct nfs_client *clp) +{ + int status; + unsigned *ptr; + struct nfs_fsinfo fsinfo; + struct nfs4_session *session = clp->cl_session; + + dprintk("--> %s clp=%p session=%p\n", __func__, clp, session); + + status = _nfs4_proc_create_session(clp); + if (status) + goto out; + + /* Init the fore channel */ + status = nfs4_init_slot_table(session); + dprintk("fore channel slot table initialization returned %d\n", status); + if (status) + goto out; + + ptr = (unsigned *)&session->sess_id.data[0]; + dprintk("%s client>seqid %d sessionid %u:%u:%u:%u\n", __func__, + clp->cl_seqid, ptr[0], ptr[1], ptr[2], ptr[3]); + + /* Get the lease time */ + status = nfs4_proc_get_lease_time(clp, &fsinfo); + if (status == 0) { + /* Update lease time and schedule renewal */ + spin_lock(&clp->cl_lock); + clp->cl_lease_time = fsinfo.lease_time * HZ; + clp->cl_last_renewal = jiffies; + clear_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state); + spin_unlock(&clp->cl_lock); + + nfs4_schedule_state_renewal(clp); + } +out: + dprintk("<-- %s\n", __func__); + return status; +} + #endif /* CONFIG_NFS_V4_1 */ struct nfs4_state_recovery_ops nfs4_reboot_recovery_ops = { diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index 85ee1d17a461..7a243a2cf0be 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -246,6 +246,8 @@ static int nfs4_stat_to_errno(int); (0) #if defined(CONFIG_NFS_V4_1) +#define NFS4_MAX_MACHINE_NAME_LEN (64) + #define encode_exchange_id_maxsz (op_encode_hdr_maxsz + \ encode_verifier_maxsz + \ 1 /* co_ownerid.len */ + \ @@ -267,6 +269,31 @@ static int nfs4_stat_to_errno(int); XDR_QUADLEN(NFS4_OPAQUE_LIMIT) + 1 + \ 1 /* eir_server_impl_id array length */ + \ 0 /* ignored eir_server_impl_id contents */) +#define encode_channel_attrs_maxsz (6 + 1 /* ca_rdma_ird.len (0) */) +#define decode_channel_attrs_maxsz (6 + \ + 1 /* ca_rdma_ird.len */ + \ + 1 /* ca_rdma_ird */) +#define encode_create_session_maxsz (op_encode_hdr_maxsz + \ + 2 /* csa_clientid */ + \ + 1 /* csa_sequence */ + \ + 1 /* csa_flags */ + \ + encode_channel_attrs_maxsz + \ + encode_channel_attrs_maxsz + \ + 1 /* csa_cb_program */ + \ + 1 /* csa_sec_parms.len (1) */ + \ + 1 /* cb_secflavor (AUTH_SYS) */ + \ + 1 /* stamp */ + \ + 1 /* machinename.len */ + \ + XDR_QUADLEN(NFS4_MAX_MACHINE_NAME_LEN) + \ + 1 /* uid */ + \ + 1 /* gid */ + \ + 1 /* gids.len (0) */) +#define decode_create_session_maxsz (op_decode_hdr_maxsz + \ + XDR_QUADLEN(NFS4_MAX_SESSIONID_LEN) + \ + 1 /* csr_sequence */ + \ + 1 /* csr_flags */ + \ + decode_channel_attrs_maxsz + \ + decode_channel_attrs_maxsz) #define encode_sequence_maxsz 0 /* stub */ #define decode_sequence_maxsz 0 /* stub */ #else /* CONFIG_NFS_V4_1 */ @@ -622,6 +649,12 @@ static int nfs4_stat_to_errno(int); #define NFS4_dec_exchange_id_sz \ (compound_decode_hdr_maxsz + \ decode_exchange_id_maxsz) +#define NFS4_enc_create_session_sz \ + (compound_encode_hdr_maxsz + \ + encode_create_session_maxsz) +#define NFS4_dec_create_session_sz \ + (compound_decode_hdr_maxsz + \ + decode_create_session_maxsz) #define NFS4_enc_get_lease_time_sz (compound_encode_hdr_maxsz + \ encode_sequence_maxsz + \ encode_putrootfh_maxsz + \ @@ -712,6 +745,7 @@ static void encode_compound_hdr(struct xdr_stream *xdr, static void encode_nops(struct compound_hdr *hdr) { + BUG_ON(hdr->nops > NFS4_MAX_OPS); *hdr->nops_p = htonl(hdr->nops); } @@ -1513,6 +1547,68 @@ static void encode_exchange_id(struct xdr_stream *xdr, hdr->nops++; hdr->replen += decode_exchange_id_maxsz; } + +static void encode_create_session(struct xdr_stream *xdr, + struct nfs41_create_session_args *args, + struct compound_hdr *hdr) +{ + __be32 *p; + char machine_name[NFS4_MAX_MACHINE_NAME_LEN]; + uint32_t len; + struct nfs_client *clp = args->client; + + RESERVE_SPACE(4); + WRITE32(OP_CREATE_SESSION); + + RESERVE_SPACE(8); + WRITE64(clp->cl_ex_clid); + + RESERVE_SPACE(8); + WRITE32(clp->cl_seqid); /*Sequence id */ + WRITE32(args->flags); /*flags */ + + RESERVE_SPACE(2*28); /* 2 channel_attrs */ + /* Fore Channel */ + WRITE32(args->fc_attrs.headerpadsz); /* header padding size */ + WRITE32(args->fc_attrs.max_rqst_sz); /* max req size */ + WRITE32(args->fc_attrs.max_resp_sz); /* max resp size */ + WRITE32(args->fc_attrs.max_resp_sz_cached); /* Max resp sz cached */ + WRITE32(args->fc_attrs.max_ops); /* max operations */ + WRITE32(args->fc_attrs.max_reqs); /* max requests */ + WRITE32(0); /* rdmachannel_attrs */ + + /* Back Channel */ + WRITE32(args->fc_attrs.headerpadsz); /* header padding size */ + WRITE32(args->bc_attrs.max_rqst_sz); /* max req size */ + WRITE32(args->bc_attrs.max_resp_sz); /* max resp size */ + WRITE32(args->bc_attrs.max_resp_sz_cached); /* Max resp sz cached */ + WRITE32(args->bc_attrs.max_ops); /* max operations */ + WRITE32(args->bc_attrs.max_reqs); /* max requests */ + WRITE32(0); /* rdmachannel_attrs */ + + RESERVE_SPACE(4); + WRITE32(args->cb_program); /* cb_program */ + + RESERVE_SPACE(4); /* # of security flavors */ + WRITE32(1); + + RESERVE_SPACE(4); + WRITE32(RPC_AUTH_UNIX); /* auth_sys */ + + /* authsys_parms rfc1831 */ + RESERVE_SPACE(4); + WRITE32((u32)clp->cl_boot_time.tv_nsec); /* stamp */ + len = scnprintf(machine_name, sizeof(machine_name), "%s", + clp->cl_ipaddr); + RESERVE_SPACE(16 + len); + WRITE32(len); + WRITEMEM(machine_name, len); + WRITE32(0); /* UID */ + WRITE32(0); /* GID */ + WRITE32(0); /* No more gids */ + hdr->nops++; + hdr->replen += decode_create_session_maxsz; +} #endif /* CONFIG_NFS_V4_1 */ static void encode_sequence(struct xdr_stream *xdr, @@ -2240,6 +2336,24 @@ static int nfs4_xdr_enc_exchange_id(struct rpc_rqst *req, uint32_t *p, return 0; } +/* + * a CREATE_SESSION request + */ +static int nfs4_xdr_enc_create_session(struct rpc_rqst *req, uint32_t *p, + struct nfs41_create_session_args *args) +{ + struct xdr_stream xdr; + struct compound_hdr hdr = { + .minorversion = args->client->cl_minorversion, + }; + + xdr_init_encode(&xdr, &req->rq_snd_buf, p); + encode_compound_hdr(&xdr, req, &hdr); + encode_create_session(&xdr, args, &hdr); + encode_nops(&hdr); + return 0; +} + /* * a GET_LEASE_TIME request */ @@ -4021,6 +4135,59 @@ static int decode_exchange_id(struct xdr_stream *xdr, return 0; } + +static int decode_chan_attrs(struct xdr_stream *xdr, + struct nfs4_channel_attrs *attrs) +{ + __be32 *p; + u32 nr_attrs; + + READ_BUF(28); + READ32(attrs->headerpadsz); + READ32(attrs->max_rqst_sz); + READ32(attrs->max_resp_sz); + READ32(attrs->max_resp_sz_cached); + READ32(attrs->max_ops); + READ32(attrs->max_reqs); + READ32(nr_attrs); + if (unlikely(nr_attrs > 1)) { + printk(KERN_WARNING "%s: Invalid rdma channel attrs count %u\n", + __func__, nr_attrs); + return -EINVAL; + } + if (nr_attrs == 1) + READ_BUF(4); /* skip rdma_attrs */ + return 0; +} + +static int decode_create_session(struct xdr_stream *xdr, + struct nfs41_create_session_res *res) +{ + __be32 *p; + int status; + struct nfs_client *clp = res->client; + struct nfs4_session *session = clp->cl_session; + + status = decode_op_hdr(xdr, OP_CREATE_SESSION); + + if (status) + return status; + + /* sessionid */ + READ_BUF(NFS4_MAX_SESSIONID_LEN); + COPYMEM(&session->sess_id, NFS4_MAX_SESSIONID_LEN); + + /* seqid, flags */ + READ_BUF(8); + READ32(clp->cl_seqid); + READ32(session->flags); + + /* Channel attributes */ + status = decode_chan_attrs(xdr, &session->fc_attrs); + if (!status) + status = decode_chan_attrs(xdr, &session->bc_attrs); + return status; +} #endif /* CONFIG_NFS_V4_1 */ static int decode_sequence(struct xdr_stream *xdr, @@ -4938,6 +5105,23 @@ static int nfs4_xdr_dec_exchange_id(struct rpc_rqst *rqstp, uint32_t *p, return status; } +/* + * a CREATE_SESSION request + */ +static int nfs4_xdr_dec_create_session(struct rpc_rqst *rqstp, uint32_t *p, + struct nfs41_create_session_res *res) +{ + struct xdr_stream xdr; + struct compound_hdr hdr; + int status; + + xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); + status = decode_compound_hdr(&xdr, &hdr); + if (!status) + status = decode_create_session(&xdr, res); + return status; +} + /* * a GET_LEASE_TIME request */ @@ -5131,6 +5315,7 @@ struct rpc_procinfo nfs4_procedures[] = { PROC(FS_LOCATIONS, enc_fs_locations, dec_fs_locations), #if defined(CONFIG_NFS_V4_1) PROC(EXCHANGE_ID, enc_exchange_id, dec_exchange_id), + PROC(CREATE_SESSION, enc_create_session, dec_create_session), PROC(GET_LEASE_TIME, enc_get_lease_time, dec_get_lease_time), #endif /* CONFIG_NFS_V4_1 */ }; diff --git a/include/linux/nfs4.h b/include/linux/nfs4.h index ad65709ed8d3..bd2eba530667 100644 --- a/include/linux/nfs4.h +++ b/include/linux/nfs4.h @@ -131,6 +131,16 @@ #define NFS4_MAX_UINT64 (~(u64)0) +/* An NFS4 sessions server must support at least NFS4_MAX_OPS operations. + * If a compound requires more operations, adjust NFS4_MAX_OPS accordingly. + */ +#define NFS4_MAX_OPS 8 + +/* Our NFS4 client back channel server only wants the cb_sequene and the + * actual operation per compound + */ +#define NFS4_MAX_BACK_CHANNEL_OPS 2 + enum nfs4_acl_whotype { NFS4_ACL_WHO_NAMED = 0, NFS4_ACL_WHO_OWNER, diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h index ca643aa87d46..62f63fb0c4c8 100644 --- a/include/linux/nfs_xdr.h +++ b/include/linux/nfs_xdr.h @@ -926,6 +926,18 @@ struct nfs41_exchange_id_res { struct nfs_client *client; u32 flags; }; + +struct nfs41_create_session_args { + struct nfs_client *client; + uint32_t flags; + uint32_t cb_program; + struct nfs4_channel_attrs fc_attrs; /* Fore Channel */ + struct nfs4_channel_attrs bc_attrs; /* Back Channel */ +}; + +struct nfs41_create_session_res { + struct nfs_client *client; +}; #endif /* CONFIG_NFS_V4_1 */ struct nfs_page; -- cgit v1.2.3 From 8d35301d7db4165f2430c428865f7b0add47615d Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Wed, 1 Apr 2009 09:22:32 -0400 Subject: nfs41: verify session channel attribues Invalidate the session if the server returns invalid fore or back channel attributes. Use a KERN_WARNING to report the fatal session estabishment error. Signed-off-by: Andy Adamson [refactor nfs4_verify_channel_attrs] Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 50 ++++++++++++++++++++++++++++++++++++++++++++++++-- 1 file changed, 48 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 7d81d6e57533..4d7a8b9e1291 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -4409,6 +4409,51 @@ static void nfs4_init_channel_attrs(struct nfs41_create_session_args *args) args->bc_attrs.max_reqs); } +static int _verify_channel_attr(char *chan, char *attr_name, u32 sent, u32 rcvd) +{ + if (rcvd <= sent) + return 0; + printk(KERN_WARNING "%s: Session INVALID: %s channel %s increased. " + "sent=%u rcvd=%u\n", __func__, chan, attr_name, sent, rcvd); + return -EINVAL; +} + +#define _verify_fore_channel_attr(_name_) \ + _verify_channel_attr("fore", #_name_, \ + args->fc_attrs._name_, \ + session->fc_attrs._name_) + +#define _verify_back_channel_attr(_name_) \ + _verify_channel_attr("back", #_name_, \ + args->bc_attrs._name_, \ + session->bc_attrs._name_) + +/* + * The server is not allowed to increase the fore channel header pad size, + * maximum response size, or maximum number of operations. + * + * The back channel attributes are only negotiatied down: We send what the + * (back channel) server insists upon. + */ +static int nfs4_verify_channel_attrs(struct nfs41_create_session_args *args, + struct nfs4_session *session) +{ + int ret = 0; + + ret |= _verify_fore_channel_attr(headerpadsz); + ret |= _verify_fore_channel_attr(max_resp_sz); + ret |= _verify_fore_channel_attr(max_ops); + + ret |= _verify_back_channel_attr(headerpadsz); + ret |= _verify_back_channel_attr(max_rqst_sz); + ret |= _verify_back_channel_attr(max_resp_sz); + ret |= _verify_back_channel_attr(max_resp_sz_cached); + ret |= _verify_back_channel_attr(max_ops); + ret |= _verify_back_channel_attr(max_reqs); + + return ret; +} + static int _nfs4_proc_create_session(struct nfs_client *clp) { struct nfs4_session *session = clp->cl_session; @@ -4431,8 +4476,9 @@ static int _nfs4_proc_create_session(struct nfs_client *clp) status = rpc_call_sync(session->clp->cl_rpcclient, &msg, 0); - /* Set the negotiated values in the session's channel_attrs struct */ - + if (!status) + /* Verify the session's negotiated channel_attrs values */ + status = nfs4_verify_channel_attrs(&args, session); if (!status) { /* Increment the clientid slot sequence id */ clp->cl_seqid++; -- cgit v1.2.3 From 96b09e024fedf0a6604c8c688a3994d5ed991434 Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Wed, 1 Apr 2009 09:22:33 -0400 Subject: nfs41: use session attributes for rsize and wsize Set the mount points rsize and wsize to the negotiated session fore channel maximum response and requeset size. These values will be bound checked in nfs_server_set_fsinfo. Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [move nfs4_session_set_rwsize into CONFIG_NFS_V4] Signed-off-by: Trond Myklebust --- fs/nfs/client.c | 17 +++++++++++++++++ 1 file changed, 17 insertions(+) (limited to 'fs') diff --git a/fs/nfs/client.c b/fs/nfs/client.c index a9828baaa445..bb7432d83b5a 100644 --- a/fs/nfs/client.c +++ b/fs/nfs/client.c @@ -1206,6 +1206,21 @@ static void nfs4_init_session(struct nfs_client *clp, #endif /* CONFIG_NFS_V4_1 */ } +/* + * Session has been established, and the client marked ready. + * Set the mount rsize and wsize with negotiated fore channel + * attributes which will be bound checked in nfs_server_set_fsinfo. + */ +static void nfs4_session_set_rwsize(struct nfs_server *server) +{ +#ifdef CONFIG_NFS_V4_1 + if (!nfs4_has_session(server->nfs_client)) + return; + server->rsize = server->nfs_client->cl_session->fc_attrs.max_resp_sz; + server->wsize = server->nfs_client->cl_session->fc_attrs.max_rqst_sz; +#endif /* CONFIG_NFS_V4_1 */ +} + /* * Create a version 4 volume record */ @@ -1296,6 +1311,8 @@ struct nfs_server *nfs4_create_server(const struct nfs_parsed_mount_data *data, (unsigned long long) server->fsid.minor); dprintk("Mount FH: %d\n", mntfh->size); + nfs4_session_set_rwsize(server); + error = nfs_probe_fsinfo(server, mntfh, &fattr); if (error < 0) goto error; -- cgit v1.2.3 From 0f3e66c6a6cae479028d31198288e524fe5ff90d Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Wed, 1 Apr 2009 09:22:34 -0400 Subject: nfs41: destroy_session operation Implement the destroy_session operation conforming to http://tools.ietf.org/html/draft-ietf-nfsv4-minorversion1-26 Signed-off-by: Ricardo Labiaga Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: remove extraneous rpc_clnt pointer] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41; NFS_CS_READY required for DESTROY_SESSION] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: pass *session in seq_args and seq_res] Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust [nfs41: fix encode_destroy_session's xdr Xcoding pointer type] Signed-off-by: Benny Halevy --- fs/nfs/nfs4proc.c | 30 ++++++++++++++++++++++++++++ fs/nfs/nfs4xdr.c | 59 +++++++++++++++++++++++++++++++++++++++++++++++++++++++ fs/nfs/super.c | 2 ++ 3 files changed, 91 insertions(+) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 4d7a8b9e1291..59e8bf5b10bb 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -4532,6 +4532,36 @@ out: return status; } +/* + * Issue the over-the-wire RPC DESTROY_SESSION. + * The caller must serialize access to this routine. + */ +int nfs4_proc_destroy_session(struct nfs4_session *session) +{ + int status = 0; + struct rpc_message msg; + + dprintk("--> nfs4_proc_destroy_session\n"); + + /* session is still being setup */ + if (session->clp->cl_cons_state != NFS_CS_READY) + return status; + + msg.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_DESTROY_SESSION]; + msg.rpc_argp = session; + msg.rpc_resp = NULL; + msg.rpc_cred = NULL; + status = rpc_call_sync(session->clp->cl_rpcclient, &msg, 0); + + if (status) + printk(KERN_WARNING + "Got error %d from the server on DESTROY_SESSION. " + "Session has been destroyed regardless...\n", status); + + dprintk("<-- nfs4_proc_destroy_session\n"); + return status; +} + #endif /* CONFIG_NFS_V4_1 */ struct nfs4_state_recovery_ops nfs4_reboot_recovery_ops = { diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index 7a243a2cf0be..a81e49648ccb 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -294,6 +294,8 @@ static int nfs4_stat_to_errno(int); 1 /* csr_flags */ + \ decode_channel_attrs_maxsz + \ decode_channel_attrs_maxsz) +#define encode_destroy_session_maxsz (op_encode_hdr_maxsz + 4) +#define decode_destroy_session_maxsz (op_decode_hdr_maxsz) #define encode_sequence_maxsz 0 /* stub */ #define decode_sequence_maxsz 0 /* stub */ #else /* CONFIG_NFS_V4_1 */ @@ -655,6 +657,10 @@ static int nfs4_stat_to_errno(int); #define NFS4_dec_create_session_sz \ (compound_decode_hdr_maxsz + \ decode_create_session_maxsz) +#define NFS4_enc_destroy_session_sz (compound_encode_hdr_maxsz + \ + encode_destroy_session_maxsz) +#define NFS4_dec_destroy_session_sz (compound_decode_hdr_maxsz + \ + decode_destroy_session_maxsz) #define NFS4_enc_get_lease_time_sz (compound_encode_hdr_maxsz + \ encode_sequence_maxsz + \ encode_putrootfh_maxsz + \ @@ -1609,6 +1615,18 @@ static void encode_create_session(struct xdr_stream *xdr, hdr->nops++; hdr->replen += decode_create_session_maxsz; } + +static void encode_destroy_session(struct xdr_stream *xdr, + struct nfs4_session *session, + struct compound_hdr *hdr) +{ + __be32 *p; + RESERVE_SPACE(4 + NFS4_MAX_SESSIONID_LEN); + WRITE32(OP_DESTROY_SESSION); + WRITEMEM(session->sess_id.data, NFS4_MAX_SESSIONID_LEN); + hdr->nops++; + hdr->replen += decode_destroy_session_maxsz; +} #endif /* CONFIG_NFS_V4_1 */ static void encode_sequence(struct xdr_stream *xdr, @@ -2354,6 +2372,24 @@ static int nfs4_xdr_enc_create_session(struct rpc_rqst *req, uint32_t *p, return 0; } +/* + * a DESTROY_SESSION request + */ +static int nfs4_xdr_enc_destroy_session(struct rpc_rqst *req, uint32_t *p, + struct nfs4_session *session) +{ + struct xdr_stream xdr; + struct compound_hdr hdr = { + .minorversion = session->clp->cl_minorversion, + }; + + xdr_init_encode(&xdr, &req->rq_snd_buf, p); + encode_compound_hdr(&xdr, req, &hdr); + encode_destroy_session(&xdr, session, &hdr); + encode_nops(&hdr); + return 0; +} + /* * a GET_LEASE_TIME request */ @@ -4188,6 +4224,11 @@ static int decode_create_session(struct xdr_stream *xdr, status = decode_chan_attrs(xdr, &session->bc_attrs); return status; } + +static int decode_destroy_session(struct xdr_stream *xdr, void *dummy) +{ + return decode_op_hdr(xdr, OP_DESTROY_SESSION); +} #endif /* CONFIG_NFS_V4_1 */ static int decode_sequence(struct xdr_stream *xdr, @@ -5122,6 +5163,23 @@ static int nfs4_xdr_dec_create_session(struct rpc_rqst *rqstp, uint32_t *p, return status; } +/* + * a DESTROY_SESSION request + */ +static int nfs4_xdr_dec_destroy_session(struct rpc_rqst *rqstp, uint32_t *p, + void *dummy) +{ + struct xdr_stream xdr; + struct compound_hdr hdr; + int status; + + xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); + status = decode_compound_hdr(&xdr, &hdr); + if (!status) + status = decode_destroy_session(&xdr, dummy); + return status; +} + /* * a GET_LEASE_TIME request */ @@ -5316,6 +5374,7 @@ struct rpc_procinfo nfs4_procedures[] = { #if defined(CONFIG_NFS_V4_1) PROC(EXCHANGE_ID, enc_exchange_id, dec_exchange_id), PROC(CREATE_SESSION, enc_create_session, dec_create_session), + PROC(DESTROY_SESSION, enc_destroy_session, dec_destroy_session), PROC(GET_LEASE_TIME, enc_get_lease_time, dec_get_lease_time), #endif /* CONFIG_NFS_V4_1 */ }; diff --git a/fs/nfs/super.c b/fs/nfs/super.c index 5a8fdc791cc1..73db6f8c145a 100644 --- a/fs/nfs/super.c +++ b/fs/nfs/super.c @@ -2485,12 +2485,14 @@ static void nfs4_kill_super(struct super_block *sb) { struct nfs_server *server = NFS_SB(sb); + dprintk("--> %s\n", __func__); nfs_super_return_all_delegations(sb); kill_anon_super(sb); nfs4_renewd_prepare_shutdown(server); nfs_fscache_release_super_cookie(sb); nfs_free_server(server); + dprintk("<-- %s\n", __func__); } /* -- cgit v1.2.3 From 8328d59f380e26477b9c1ede99e33d021365bcd1 Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Wed, 1 Apr 2009 09:22:35 -0400 Subject: nfs41: enable nfs_client only nfs4_async_handle_error The session is per struct nfs_client, not per nfs_server. Allow the handler to be called with no nfs_server which simplifies the nfs4_proc_async_sequence session renewal call and will let it be used by pnfs file layout data servers. Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 13 +++++++++---- 1 file changed, 9 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 59e8bf5b10bb..5fe208b486d4 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -3221,10 +3221,8 @@ static int nfs4_proc_set_acl(struct inode *inode, const void *buf, size_t buflen } static int -nfs4_async_handle_error(struct rpc_task *task, const struct nfs_server *server, struct nfs4_state *state) +_nfs4_async_handle_error(struct rpc_task *task, const struct nfs_server *server, struct nfs_client *clp, struct nfs4_state *state) { - struct nfs_client *clp = server->nfs_client; - if (!clp || task->tk_status >= 0) return 0; switch(task->tk_status) { @@ -3244,7 +3242,8 @@ nfs4_async_handle_error(struct rpc_task *task, const struct nfs_server *server, task->tk_status = 0; return -EAGAIN; case -NFS4ERR_DELAY: - nfs_inc_server_stats(server, NFSIOS_DELAY); + if (server) + nfs_inc_server_stats(server, NFSIOS_DELAY); case -NFS4ERR_GRACE: rpc_delay(task, NFS4_POLL_RETRY_MAX); task->tk_status = 0; @@ -3257,6 +3256,12 @@ nfs4_async_handle_error(struct rpc_task *task, const struct nfs_server *server, return 0; } +static int +nfs4_async_handle_error(struct rpc_task *task, const struct nfs_server *server, struct nfs4_state *state) +{ + return _nfs4_async_handle_error(task, server, server->nfs_client, state); +} + int nfs4_proc_setclientid(struct nfs_client *clp, u32 program, unsigned short port, struct rpc_cred *cred) { nfs4_verifier sc_verifier; -- cgit v1.2.3 From fc01cea963a246742ff15e118ce5259e3091352c Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Wed, 1 Apr 2009 09:22:36 -0400 Subject: nfs41: sequence operation Implement the sequence operation conforming to http://tools.ietf.org/html/draft-ietf-nfsv4-minorversion1-26 Check returned sessionid, slotid and slot sequenceid in decode_sequence. If the server returns different values for sessionID, slotID or slot sequence number than what was sent, the server is looney tunes. Pass the sequence operation status to nfs41_sequence_done in order to determine when to increment the slot sequence ID. Free slot is separated from sequence done. Signed-off-by: Rahul Iyer Signed-off-by: Ricardo Labiaga Signed-off-by: Andy Adamson [nfs41: sequence res use slotid] Signed-off-by: Andy Adamson [nfs41: deref slot table in decode_sequence only for minorversion!=0] Signed-off-by: Benny Halevy [nfs41: nfs4_call_sync] [nfs41: remove SEQ4_STATUS_USE_TK_STATUS] [nfs41: return ESERVERFAULT in decode_sequence] [no sr_session, no sr_flags] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: use nfs4_call_sync_sequence to renew session lease] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: remove nfs4_call_sync_sequence forward definition] Signed-off-by: Andy Adamson [nfs41: use struct nfs_client for nfs41_proc_async_sequence] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: pass *session in seq_args and seq_res] Signed-off-by: Benny Halevy [nfs41 nfs41_sequence_call_done update error checking] [nfs41 nfs41_sequence_done update error checking] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: remove switch on error from nfs41_sequence_call_done] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 92 +++++++++++++++++++++++++++++++++++++++ fs/nfs/nfs4xdr.c | 127 +++++++++++++++++++++++++++++++++++++++++++++++++++--- 2 files changed, 214 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 5fe208b486d4..17768095f0cc 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -4567,6 +4567,98 @@ int nfs4_proc_destroy_session(struct nfs4_session *session) return status; } +/* + * Renew the cl_session lease. + */ +static int nfs4_proc_sequence(struct nfs_client *clp, struct rpc_cred *cred) +{ + struct nfs4_sequence_args args; + struct nfs4_sequence_res res; + + struct rpc_message msg = { + .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_SEQUENCE], + .rpc_argp = &args, + .rpc_resp = &res, + .rpc_cred = cred, + }; + + args.sa_cache_this = 0; + + return nfs4_call_sync_sequence(clp, clp->cl_rpcclient, &msg, &args, + &res, 0); +} + +void nfs41_sequence_call_done(struct rpc_task *task, void *data) +{ + struct nfs_client *clp = (struct nfs_client *)data; + + nfs41_sequence_done(clp, task->tk_msg.rpc_resp, task->tk_status); + + if (task->tk_status < 0) { + dprintk("%s ERROR %d\n", __func__, task->tk_status); + + if (_nfs4_async_handle_error(task, NULL, clp, NULL) + == -EAGAIN) { + rpc_restart_call(task); + return; + } + } + nfs41_sequence_free_slot(clp, task->tk_msg.rpc_resp); + dprintk("%s rpc_cred %p\n", __func__, task->tk_msg.rpc_cred); + + put_rpccred(task->tk_msg.rpc_cred); + kfree(task->tk_msg.rpc_argp); + kfree(task->tk_msg.rpc_resp); + + dprintk("<-- %s\n", __func__); +} + +static void nfs41_sequence_prepare(struct rpc_task *task, void *data) +{ + struct nfs_client *clp; + struct nfs4_sequence_args *args; + struct nfs4_sequence_res *res; + + clp = (struct nfs_client *)data; + args = task->tk_msg.rpc_argp; + res = task->tk_msg.rpc_resp; + + if (nfs4_setup_sequence(clp, args, res, 0, task)) + return; + rpc_call_start(task); +} + +static const struct rpc_call_ops nfs41_sequence_ops = { + .rpc_call_done = nfs41_sequence_call_done, + .rpc_call_prepare = nfs41_sequence_prepare, +}; + +static int nfs41_proc_async_sequence(struct nfs_client *clp, + struct rpc_cred *cred) +{ + struct nfs4_sequence_args *args; + struct nfs4_sequence_res *res; + struct rpc_message msg = { + .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_SEQUENCE], + .rpc_cred = cred, + }; + + args = kzalloc(sizeof(*args), GFP_KERNEL); + if (!args) + return -ENOMEM; + res = kzalloc(sizeof(*res), GFP_KERNEL); + if (!res) { + kfree(args); + return -ENOMEM; + } + res->sr_slotid = NFS4_MAX_SLOT_TABLE; + msg.rpc_argp = args; + msg.rpc_resp = res; + + return rpc_call_async(clp->cl_rpcclient, &msg, RPC_TASK_SOFT, + &nfs41_sequence_ops, (void *)clp); +} + #endif /* CONFIG_NFS_V4_1 */ struct nfs4_state_recovery_ops nfs4_reboot_recovery_ops = { diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index a81e49648ccb..617273e7d47f 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -296,8 +296,10 @@ static int nfs4_stat_to_errno(int); decode_channel_attrs_maxsz) #define encode_destroy_session_maxsz (op_encode_hdr_maxsz + 4) #define decode_destroy_session_maxsz (op_decode_hdr_maxsz) -#define encode_sequence_maxsz 0 /* stub */ -#define decode_sequence_maxsz 0 /* stub */ +#define encode_sequence_maxsz (op_encode_hdr_maxsz + \ + XDR_QUADLEN(NFS4_MAX_SESSIONID_LEN) + 4) +#define decode_sequence_maxsz (op_decode_hdr_maxsz + \ + XDR_QUADLEN(NFS4_MAX_SESSIONID_LEN) + 5) #else /* CONFIG_NFS_V4_1 */ #define encode_sequence_maxsz 0 #define decode_sequence_maxsz 0 @@ -661,6 +663,12 @@ static int nfs4_stat_to_errno(int); encode_destroy_session_maxsz) #define NFS4_dec_destroy_session_sz (compound_decode_hdr_maxsz + \ decode_destroy_session_maxsz) +#define NFS4_enc_sequence_sz \ + (compound_decode_hdr_maxsz + \ + encode_sequence_maxsz) +#define NFS4_dec_sequence_sz \ + (compound_decode_hdr_maxsz + \ + decode_sequence_maxsz) #define NFS4_enc_get_lease_time_sz (compound_encode_hdr_maxsz + \ encode_sequence_maxsz + \ encode_putrootfh_maxsz + \ @@ -1635,11 +1643,39 @@ static void encode_sequence(struct xdr_stream *xdr, { #if defined(CONFIG_NFS_V4_1) struct nfs4_session *session = args->sa_session; + struct nfs4_slot_table *tp; + struct nfs4_slot *slot; + __be32 *p; if (!session) return; - /* stub */ + tp = &session->fc_slot_table; + + WARN_ON(args->sa_slotid == NFS4_MAX_SLOT_TABLE); + slot = tp->slots + args->sa_slotid; + + RESERVE_SPACE(4); + WRITE32(OP_SEQUENCE); + + /* + * Sessionid + seqid + slotid + max slotid + cache_this + */ + dprintk("%s: sessionid=%u:%u:%u:%u seqid=%d slotid=%d " + "max_slotid=%d cache_this=%d\n", + __func__, + ((u32 *)session->sess_id.data)[0], + ((u32 *)session->sess_id.data)[1], + ((u32 *)session->sess_id.data)[2], + ((u32 *)session->sess_id.data)[3], + slot->seq_nr, args->sa_slotid, + tp->highest_used_slotid, args->sa_cache_this); + RESERVE_SPACE(NFS4_MAX_SESSIONID_LEN + 16); + WRITEMEM(session->sess_id.data, NFS4_MAX_SESSIONID_LEN); + WRITE32(slot->seq_nr); + WRITE32(args->sa_slotid); + WRITE32(tp->highest_used_slotid); + WRITE32(args->sa_cache_this); hdr->nops++; hdr->replen += decode_sequence_maxsz; #endif /* CONFIG_NFS_V4_1 */ @@ -2390,6 +2426,24 @@ static int nfs4_xdr_enc_destroy_session(struct rpc_rqst *req, uint32_t *p, return 0; } +/* + * a SEQUENCE request + */ +static int nfs4_xdr_enc_sequence(struct rpc_rqst *req, uint32_t *p, + struct nfs4_sequence_args *args) +{ + struct xdr_stream xdr; + struct compound_hdr hdr = { + .minorversion = nfs4_xdr_minorversion(args), + }; + + xdr_init_encode(&xdr, &req->rq_snd_buf, p); + encode_compound_hdr(&xdr, req, &hdr); + encode_sequence(&xdr, args, &hdr); + encode_nops(&hdr); + return 0; +} + /* * a GET_LEASE_TIME request */ @@ -4236,13 +4290,58 @@ static int decode_sequence(struct xdr_stream *xdr, struct rpc_rqst *rqstp) { #if defined(CONFIG_NFS_V4_1) + struct nfs4_slot *slot; + struct nfs4_sessionid id; + u32 dummy; + int status; + __be32 *p; + if (!res->sr_session) return 0; - /* stub */ -#endif /* CONFIG_NFS_V4_1 */ + status = decode_op_hdr(xdr, OP_SEQUENCE); + if (status) + goto out_err; + /* + * If the server returns different values for sessionID, slotID or + * sequence number, the server is looney tunes. + */ + status = -ESERVERFAULT; + + slot = &res->sr_session->fc_slot_table.slots[res->sr_slotid]; + READ_BUF(NFS4_MAX_SESSIONID_LEN + 20); + COPYMEM(id.data, NFS4_MAX_SESSIONID_LEN); + if (memcmp(id.data, res->sr_session->sess_id.data, + NFS4_MAX_SESSIONID_LEN)) { + dprintk("%s Invalid session id\n", __func__); + goto out_err; + } + /* seqid */ + READ32(dummy); + if (dummy != slot->seq_nr) { + dprintk("%s Invalid sequence number\n", __func__); + goto out_err; + } + /* slot id */ + READ32(dummy); + if (dummy != res->sr_slotid) { + dprintk("%s Invalid slot id\n", __func__); + goto out_err; + } + /* highest slot id - currently not processed */ + READ32(dummy); + /* target highest slot id - currently not processed */ + READ32(dummy); + /* result flags - currently not processed */ + READ32(dummy); + status = 0; +out_err: + res->sr_status = status; + return status; +#else /* CONFIG_NFS_V4_1 */ return 0; +#endif /* CONFIG_NFS_V4_1 */ } /* @@ -5180,6 +5279,23 @@ static int nfs4_xdr_dec_destroy_session(struct rpc_rqst *rqstp, uint32_t *p, return status; } +/* + * a SEQUENCE request + */ +static int nfs4_xdr_dec_sequence(struct rpc_rqst *rqstp, uint32_t *p, + struct nfs4_sequence_res *res) +{ + struct xdr_stream xdr; + struct compound_hdr hdr; + int status; + + xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); + status = decode_compound_hdr(&xdr, &hdr); + if (!status) + status = decode_sequence(&xdr, res, rqstp); + return status; +} + /* * a GET_LEASE_TIME request */ @@ -5375,6 +5491,7 @@ struct rpc_procinfo nfs4_procedures[] = { PROC(EXCHANGE_ID, enc_exchange_id, dec_exchange_id), PROC(CREATE_SESSION, enc_create_session, dec_create_session), PROC(DESTROY_SESSION, enc_destroy_session, dec_destroy_session), + PROC(SEQUENCE, enc_sequence, dec_sequence), PROC(GET_LEASE_TIME, enc_get_lease_time, dec_get_lease_time), #endif /* CONFIG_NFS_V4_1 */ }; -- cgit v1.2.3 From ac72b7b3b3263ce64d55094eac1d1bde5f34e64a Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Wed, 1 Apr 2009 09:22:37 -0400 Subject: nfs41: reset the session slot table Separated from nfs41: schedule async session reset Do not kfree the session slot table upon session reset, just re-initialize it. Add a boolean to nfs4_proc_create_session to inidicate if this is a session reset or a session initialization. Signed-off-by: Benny Halevy Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 44 +++++++++++++++++++++++++++++++++++++++++--- 1 file changed, 41 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 17768095f0cc..0b1214740248 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -4289,6 +4289,37 @@ int nfs4_proc_get_lease_time(struct nfs_client *clp, struct nfs_fsinfo *fsinfo) return status; } +/* Reset a slot table */ +static int nfs4_reset_slot_table(struct nfs4_session *session) +{ + struct nfs4_slot_table *tbl = &session->fc_slot_table; + int i, max_slots = session->fc_attrs.max_reqs; + int old_max_slots = session->fc_slot_table.max_slots; + int ret = 0; + + dprintk("--> %s: max_reqs=%u, tbl %p\n", __func__, + session->fc_attrs.max_reqs, tbl); + + /* Until we have dynamic slot table adjustment, insist + * upon the same slot table size */ + if (max_slots != old_max_slots) { + dprintk("%s reset slot table does't match old\n", + __func__); + ret = -EINVAL; /*XXX NFS4ERR_REQ_TOO_BIG ? */ + goto out; + } + spin_lock(&tbl->slot_tbl_lock); + for (i = 0; i < max_slots; ++i) + tbl->slots[i].seq_nr = 1; + tbl->highest_used_slotid = -1; + spin_unlock(&tbl->slot_tbl_lock); + dprintk("%s: tbl=%p slots=%p max_slots=%d\n", __func__, + tbl, tbl->slots, tbl->max_slots); +out: + dprintk("<-- %s: return %d\n", __func__, ret); + return ret; +} + /* * Initialize slot table */ @@ -4497,7 +4528,7 @@ static int _nfs4_proc_create_session(struct nfs_client *clp) * It is the responsibility of the caller to verify the session is * expired before calling this routine. */ -int nfs4_proc_create_session(struct nfs_client *clp) +int nfs4_proc_create_session(struct nfs_client *clp, int reset) { int status; unsigned *ptr; @@ -4510,8 +4541,11 @@ int nfs4_proc_create_session(struct nfs_client *clp) if (status) goto out; - /* Init the fore channel */ - status = nfs4_init_slot_table(session); + /* Init or reset the fore channel */ + if (reset) + status = nfs4_reset_slot_table(session); + else + status = nfs4_init_slot_table(session); dprintk("fore channel slot table initialization returned %d\n", status); if (status) goto out; @@ -4520,6 +4554,10 @@ int nfs4_proc_create_session(struct nfs_client *clp) dprintk("%s client>seqid %d sessionid %u:%u:%u:%u\n", __func__, clp->cl_seqid, ptr[0], ptr[1], ptr[2], ptr[3]); + if (reset) + /* Lease time is aleady set */ + goto out; + /* Get the lease time */ status = nfs4_proc_get_lease_time(clp, &fsinfo); if (status == 0) { -- cgit v1.2.3 From 76db6d9500caeaa774a3e32a997eba30bbdc176b Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Wed, 1 Apr 2009 09:22:38 -0400 Subject: nfs41: add session setup to the state manager At mount, nfs_alloc_client sets the cl_state NFS4CLNT_LEASE_EXPIRED bit and nfs4_alloc_session sets the NFS4CLNT_SESSION_SETUP bit, so both bits are set when nfs4_lookup_root calls nfs4_recover_expired_lease which schedules the nfs4_state_manager and waits for it to complete. Place the session setup after the clientid establishment in nfs4_state_manager so that the session is setup right after the clientid has been established without rescheduling the state manager. Unlike nfsv4.0, the nfs_client struct is not ready to use until the session has been established. Postpone marking the nfs_client struct to NFS_CS_READY until after a successful CREATE_SESSION call so that other threads cannot use the client until the session is established. If the EXCHANGE_ID call fails and the session has not been setup (the NFS4CLNT_SESSION_SETUP bit is set), mark the client with the error and return. If the session setup CREATE_SESSION call fails with NFS4ERR_STALE_CLIENTID which could occur due to server reboot or network partition inbetween the EXCHANGE_ID and CREATE_SESSION call, reset the NFS4CLNT_LEASE_EXPIRED and NFS4CLNT_SESSION_SETUP bits and try again. If the CREATE_SESSION call fails with other errors, mark the client with the error and return. Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: NFS_CS_SESSION_SETUP cl_cons_state for back channel setup] On session setup, the CREATE_SESSION reply races with the server back channel probe which needs to succeed to setup the back channel. Set a new cl_cons_state NFS_CS_SESSION_SETUP just prior to the CREATE_SESSION call and add it as a valid state to nfs_find_client so that the client back channel can find the nfs_client struct and won't drop the server backchannel probe. Use a new cl_cons_state so that NFSv4.0 back channel behaviour which only sets NFS_CS_READY is unchanged. Adjust waiting on the nfs_client_active_wq accordingly. Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: rename NFS_CS_SESSION_SETUP to NFS_CS_SESSION_INITING] Signed-off-by: Andy Adamson [nfs41: set NFS_CL_SESSION_INITING in alloc_session] Signed-off-by: Andy Adamson [nfs41: move session setup into a function] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [moved nfs4_proc_create_session declaration here] Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/client.c | 10 ++++++---- fs/nfs/internal.h | 1 + fs/nfs/nfs4_fs.h | 2 ++ fs/nfs/nfs4proc.c | 10 ++++++++++ fs/nfs/nfs4state.c | 35 ++++++++++++++++++++++++++++++++++- include/linux/nfs_fs_sb.h | 1 + 6 files changed, 54 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/nfs/client.c b/fs/nfs/client.c index bb7432d83b5a..d28a987f569e 100644 --- a/fs/nfs/client.c +++ b/fs/nfs/client.c @@ -366,7 +366,8 @@ struct nfs_client *nfs_find_client(const struct sockaddr *addr, u32 nfsversion) struct sockaddr *clap = (struct sockaddr *)&clp->cl_addr; /* Don't match clients that failed to initialise properly */ - if (clp->cl_cons_state != NFS_CS_READY) + if (!(clp->cl_cons_state == NFS_CS_READY || + clp->cl_cons_state == NFS_CS_SESSION_INITING)) continue; /* Different NFS versions cannot share the same nfs_client */ @@ -499,7 +500,7 @@ found_client: nfs_free_client(new); error = wait_event_killable(nfs_client_active_wq, - clp->cl_cons_state != NFS_CS_INITING); + clp->cl_cons_state < NFS_CS_INITING); if (error < 0) { nfs_put_client(clp); return ERR_PTR(-ERESTARTSYS); @@ -520,7 +521,7 @@ found_client: /* * Mark a server as ready or failed */ -static void nfs_mark_client_ready(struct nfs_client *clp, int state) +void nfs_mark_client_ready(struct nfs_client *clp, int state) { clp->cl_cons_state = state; wake_up_all(&nfs_client_active_wq); @@ -1135,7 +1136,8 @@ static int nfs4_init_client(struct nfs_client *clp, if (error < 0) goto error; - nfs_mark_client_ready(clp, NFS_CS_READY); + if (!nfs4_has_session(clp)) + nfs_mark_client_ready(clp, NFS_CS_READY); return 0; error: diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h index f62bc5226155..f3b310e8ea03 100644 --- a/fs/nfs/internal.h +++ b/fs/nfs/internal.h @@ -100,6 +100,7 @@ extern void nfs_free_server(struct nfs_server *server); extern struct nfs_server *nfs_clone_server(struct nfs_server *, struct nfs_fh *, struct nfs_fattr *); +extern void nfs_mark_client_ready(struct nfs_client *clp, int state); #ifdef CONFIG_PROC_FS extern int __init nfs_fs_proc_init(void); extern void nfs_fs_proc_exit(void); diff --git a/fs/nfs/nfs4_fs.h b/fs/nfs/nfs4_fs.h index eccf4e93e7d7..288717abaddc 100644 --- a/fs/nfs/nfs4_fs.h +++ b/fs/nfs/nfs4_fs.h @@ -44,6 +44,7 @@ enum nfs4_client_state { NFS4CLNT_RECLAIM_REBOOT, NFS4CLNT_RECLAIM_NOGRACE, NFS4CLNT_DELEGRETURN, + NFS4CLNT_SESSION_SETUP, }; /* @@ -208,6 +209,7 @@ extern int nfs4_setup_sequence(struct nfs_client *clp, int cache_reply, struct rpc_task *task); extern void nfs4_destroy_session(struct nfs4_session *session); extern struct nfs4_session *nfs4_alloc_session(struct nfs_client *clp); +extern int nfs4_proc_create_session(struct nfs_client *, int reset); #else /* CONFIG_NFS_v4_1 */ static inline int nfs4_setup_sequence(struct nfs_client *clp, struct nfs4_sequence_args *args, struct nfs4_sequence_res *res, diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 0b1214740248..7fc0c9c8f5e3 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -4382,6 +4382,16 @@ struct nfs4_session *nfs4_alloc_session(struct nfs_client *clp) session = kzalloc(sizeof(struct nfs4_session), GFP_KERNEL); if (!session) return NULL; + + set_bit(NFS4CLNT_SESSION_SETUP, &clp->cl_state); + /* + * The create session reply races with the server back + * channel probe. Mark the client NFS_CS_SESSION_INITING + * so that the client back channel can find the + * nfs_client struct + */ + clp->cl_cons_state = NFS_CS_SESSION_INITING; + tbl = &session->fc_slot_table; spin_lock_init(&tbl->slot_tbl_lock); rpc_init_wait_queue(&tbl->slot_tbl_waitq, "Slot table"); diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c index bc683ed477e1..df5b4807daa7 100644 --- a/fs/nfs/nfs4state.c +++ b/fs/nfs/nfs4state.c @@ -1113,6 +1113,27 @@ static int nfs4_reclaim_lease(struct nfs_client *clp) return status; } +#ifdef CONFIG_NFS_V4_1 + +static int nfs4_initialize_session(struct nfs_client *clp) +{ + int status; + + status = nfs4_proc_create_session(clp, 0); + if (!status) { + nfs_mark_client_ready(clp, NFS_CS_READY); + } else if (status == -NFS4ERR_STALE_CLIENTID) { + set_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state); + set_bit(NFS4CLNT_SESSION_SETUP, &clp->cl_state); + } else { + nfs_mark_client_ready(clp, status); + } + return status; +} +#else /* CONFIG_NFS_V4_1 */ +static int nfs4_initialize_session(struct nfs_client *clp) { return 0; } +#endif /* CONFIG_NFS_V4_1 */ + static void nfs4_state_manager(struct nfs_client *clp) { int status = 0; @@ -1126,6 +1147,9 @@ static void nfs4_state_manager(struct nfs_client *clp) set_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state); if (status == -EAGAIN) continue; + if (clp->cl_cons_state == + NFS_CS_SESSION_INITING) + nfs_mark_client_ready(clp, status); goto out_error; } clear_bit(NFS4CLNT_CHECK_LEASE, &clp->cl_state); @@ -1136,7 +1160,16 @@ static void nfs4_state_manager(struct nfs_client *clp) if (status != 0) continue; } - + /* Setup the session */ + if (nfs4_has_session(clp) && + test_and_clear_bit(NFS4CLNT_SESSION_SETUP, &clp->cl_state)) { + status = nfs4_initialize_session(clp); + if (status) { + if (status == -NFS4ERR_STALE_CLIENTID) + continue; + goto out_error; + } + } /* First recover reboot state... */ if (test_and_clear_bit(NFS4CLNT_RECLAIM_REBOOT, &clp->cl_state)) { status = nfs4_do_reclaim(clp, &nfs4_reboot_recovery_ops); diff --git a/include/linux/nfs_fs_sb.h b/include/linux/nfs_fs_sb.h index 435ed556efb5..d0902ccec9ce 100644 --- a/include/linux/nfs_fs_sb.h +++ b/include/linux/nfs_fs_sb.h @@ -24,6 +24,7 @@ struct nfs_client { int cl_cons_state; /* current construction state (-ve: init error) */ #define NFS_CS_READY 0 /* ready to be used */ #define NFS_CS_INITING 1 /* busy initialising */ +#define NFS_CS_SESSION_INITING 2 /* busy initialising session */ unsigned long cl_res_state; /* NFS resources state */ #define NFS_CS_CALLBACK 1 /* - callback started */ #define NFS_CS_IDMAP 2 /* - idmap started */ -- cgit v1.2.3 From c3fad1b1aaf850bf692642642ace7cd0d64af0a3 Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Wed, 1 Apr 2009 09:22:39 -0400 Subject: nfs41: add session reset to state manager Move the code to reset a session from the session_reclaimer to the nfs4_state_manager. Destroy the session, and create a new one. Treat NFS4ERR_BADSESSION and NFS4ERR_DEADSESSION as a successful nfs4_proc_destroy_session. Signal nfs4_proc_create_session that this is a session reset so that the session slot table is re-used. If the clientid is stale, set both NFS4CLNT_LEASE_EXPIRED and NFS4CLNT_SESSION_SETUP bits and retry. Use a switch statement in nfs4_session_recovery_handle_error for future patche which will add handling for other errors. Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: session reset in nfs4_recovery_handle_error] Signed-off-by: Andy Adamson [nfs41: reset session on nfs4_do_reclaim session reset error] If nfs4_do_reclaim gets a session reset error, nfs4_recovery_handle_error will set the NFS4CLNT_SESSION_SETUP bit, and the state manager should continue processing to reset the session. Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [move nfs4_proc_destroy_session declaration here] Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4_fs.h | 1 + fs/nfs/nfs4state.c | 51 +++++++++++++++++++++++++++++++++++++++++++++++++-- 2 files changed, 50 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs4_fs.h b/fs/nfs/nfs4_fs.h index 288717abaddc..7cc1cddfd9ba 100644 --- a/fs/nfs/nfs4_fs.h +++ b/fs/nfs/nfs4_fs.h @@ -210,6 +210,7 @@ extern int nfs4_setup_sequence(struct nfs_client *clp, extern void nfs4_destroy_session(struct nfs4_session *session); extern struct nfs4_session *nfs4_alloc_session(struct nfs_client *clp); extern int nfs4_proc_create_session(struct nfs_client *, int reset); +extern int nfs4_proc_destroy_session(struct nfs4_session *); #else /* CONFIG_NFS_v4_1 */ static inline int nfs4_setup_sequence(struct nfs_client *clp, struct nfs4_sequence_args *args, struct nfs4_sequence_res *res, diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c index df5b4807daa7..8b7f007adc42 100644 --- a/fs/nfs/nfs4state.c +++ b/fs/nfs/nfs4state.c @@ -1042,6 +1042,14 @@ static void nfs4_recovery_handle_error(struct nfs_client *clp, int error) case -NFS4ERR_EXPIRED: set_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state); nfs4_state_start_reclaim_nograce(clp); + case -NFS4ERR_BADSESSION: + case -NFS4ERR_BADSLOT: + case -NFS4ERR_BAD_HIGH_SLOT: + case -NFS4ERR_DEADSESSION: + case -NFS4ERR_CONN_NOT_BOUND_TO_SESSION: + case -NFS4ERR_SEQ_FALSE_RETRY: + case -NFS4ERR_SEQ_MISORDERED: + set_bit(NFS4CLNT_SESSION_SETUP, &clp->cl_state); } } @@ -1114,6 +1122,36 @@ static int nfs4_reclaim_lease(struct nfs_client *clp) } #ifdef CONFIG_NFS_V4_1 +static void nfs4_session_recovery_handle_error(struct nfs_client *clp, int err) +{ + switch (err) { + case -NFS4ERR_STALE_CLIENTID: + set_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state); + set_bit(NFS4CLNT_SESSION_SETUP, &clp->cl_state); + } +} + +static int nfs4_reset_session(struct nfs_client *clp) +{ + int status; + + status = nfs4_proc_destroy_session(clp->cl_session); + if (status && status != -NFS4ERR_BADSESSION && + status != -NFS4ERR_DEADSESSION) { + nfs4_session_recovery_handle_error(clp, status); + goto out; + } + + memset(clp->cl_session->sess_id.data, 0, NFS4_MAX_SESSIONID_LEN); + status = nfs4_proc_create_session(clp, 1); + if (status) + nfs4_session_recovery_handle_error(clp, status); + /* fall through*/ +out: + /* Wake up the next rpc task even on error */ + rpc_wake_up_next(&clp->cl_session->fc_slot_table.slot_tbl_waitq); + return status; +} static int nfs4_initialize_session(struct nfs_client *clp) { @@ -1131,6 +1169,7 @@ static int nfs4_initialize_session(struct nfs_client *clp) return status; } #else /* CONFIG_NFS_V4_1 */ +static int nfs4_reset_session(struct nfs_client *clp) { return 0; } static int nfs4_initialize_session(struct nfs_client *clp) { return 0; } #endif /* CONFIG_NFS_V4_1 */ @@ -1160,10 +1199,13 @@ static void nfs4_state_manager(struct nfs_client *clp) if (status != 0) continue; } - /* Setup the session */ + /* Initialize or reset the session */ if (nfs4_has_session(clp) && test_and_clear_bit(NFS4CLNT_SESSION_SETUP, &clp->cl_state)) { - status = nfs4_initialize_session(clp); + if (clp->cl_cons_state == NFS_CS_SESSION_INITING) + status = nfs4_initialize_session(clp); + else + status = nfs4_reset_session(clp); if (status) { if (status == -NFS4ERR_STALE_CLIENTID) continue; @@ -1175,6 +1217,8 @@ static void nfs4_state_manager(struct nfs_client *clp) status = nfs4_do_reclaim(clp, &nfs4_reboot_recovery_ops); if (status == -NFS4ERR_STALE_CLIENTID) continue; + if (test_bit(NFS4CLNT_SESSION_SETUP, &clp->cl_state)) + continue; nfs4_state_end_reclaim_reboot(clp); continue; } @@ -1188,6 +1232,9 @@ static void nfs4_state_manager(struct nfs_client *clp) continue; if (status == -NFS4ERR_EXPIRED) continue; + if (test_bit(NFS4CLNT_SESSION_SETUP, + &clp->cl_state)) + continue; goto out_error; } else nfs4_state_end_reclaim_nograce(clp); -- cgit v1.2.3 From eedc020e718b8ce45381383ec66030f09eb02a1e Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Wed, 1 Apr 2009 09:22:41 -0400 Subject: nfs41: use rpc prepare call state for session reset [nfs41: change nfs4_restart_rpc argument] [nfs41: check for session not minorversion] [nfs41: trigger the state manager for session reset] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [always define nfs4_restart_rpc] Signed-off-by: Trond Myklebust --- fs/nfs/internal.h | 39 +++++++++++++++++++++++++++------------ fs/nfs/nfs4proc.c | 15 ++++++++------- fs/nfs/read.c | 2 +- fs/nfs/unlink.c | 2 +- fs/nfs/write.c | 8 ++++---- 5 files changed, 41 insertions(+), 25 deletions(-) (limited to 'fs') diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h index f3b310e8ea03..78508f29c03a 100644 --- a/fs/nfs/internal.h +++ b/fs/nfs/internal.h @@ -2,6 +2,7 @@ * NFS internal definitions */ +#include "nfs4_fs.h" #include #include @@ -17,6 +18,18 @@ struct nfs_string; */ #define NFS_MAX_READAHEAD (RPC_DEF_SLOT_TABLE - 1) +/* + * Determine if sessions are in use. + */ +static inline int nfs4_has_session(const struct nfs_client *clp) +{ +#ifdef CONFIG_NFS_V4_1 + if (clp->cl_session) + return 1; +#endif /* CONFIG_NFS_V4_1 */ + return 0; +} + struct nfs_clone_mount { const struct super_block *sb; const struct dentry *dentry; @@ -148,6 +161,20 @@ extern __be32 * nfs_decode_dirent(__be32 *, struct nfs_entry *, int); extern struct rpc_procinfo nfs3_procedures[]; extern __be32 *nfs3_decode_dirent(__be32 *, struct nfs_entry *, int); +/* nfs4proc.c */ +static inline void nfs4_restart_rpc(struct rpc_task *task, + const struct nfs_client *clp) +{ +#ifdef CONFIG_NFS_V4_1 + if (nfs4_has_session(clp) && + test_bit(NFS4CLNT_SESSION_SETUP, &clp->cl_state)) { + rpc_restart_call_prepare(task); + return; + } +#endif /* CONFIG_NFS_V4_1 */ + rpc_restart_call(task); +} + /* nfs4xdr.c */ #ifdef CONFIG_NFS_V4 extern __be32 *nfs4_decode_dirent(__be32 *p, struct nfs_entry *entry, int plus); @@ -225,18 +252,6 @@ extern int _nfs4_call_sync_session(struct nfs_server *server, struct nfs4_sequence_res *res, int cache_reply); -/* - * Determine if sessions are in use. - */ -static inline int nfs4_has_session(const struct nfs_client *clp) -{ -#ifdef CONFIG_NFS_V4_1 - if (clp->cl_session) - return 1; -#endif /* CONFIG_NFS_V4_1 */ - return 0; -} - #ifdef CONFIG_NFS_V4_1 extern void nfs41_sequence_free_slot(const struct nfs_client *, struct nfs4_sequence_res *res); diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 7fc0c9c8f5e3..2d45606a8085 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -1661,7 +1661,7 @@ static void nfs4_close_done(struct rpc_task *task, void *data) break; default: if (nfs4_async_handle_error(task, server, state) == -EAGAIN) { - rpc_restart_call(task); + nfs4_restart_rpc(task, server->nfs_client); return; } } @@ -2874,7 +2874,7 @@ static int nfs4_read_done(struct rpc_task *task, struct nfs_read_data *data) nfs4_sequence_done(server, &data->res.seq_res, task->tk_status); if (nfs4_async_handle_error(task, server, data->args.context->state) == -EAGAIN) { - rpc_restart_call(task); + nfs4_restart_rpc(task, server->nfs_client); return -EAGAIN; } @@ -2899,7 +2899,7 @@ static int nfs4_write_done(struct rpc_task *task, struct nfs_write_data *data) task->tk_status); if (nfs4_async_handle_error(task, NFS_SERVER(inode), data->args.context->state) == -EAGAIN) { - rpc_restart_call(task); + nfs4_restart_rpc(task, NFS_SERVER(inode)->nfs_client); return -EAGAIN; } if (task->tk_status >= 0) { @@ -2927,7 +2927,7 @@ static int nfs4_commit_done(struct rpc_task *task, struct nfs_write_data *data) nfs4_sequence_done(NFS_SERVER(inode), &data->res.seq_res, task->tk_status); if (nfs4_async_handle_error(task, NFS_SERVER(inode), NULL) == -EAGAIN) { - rpc_restart_call(task); + nfs4_restart_rpc(task, NFS_SERVER(inode)->nfs_client); return -EAGAIN; } nfs4_sequence_free_slot(NFS_SERVER(inode)->nfs_client, @@ -3628,7 +3628,8 @@ static void nfs4_locku_done(struct rpc_task *task, void *data) break; default: if (nfs4_async_handle_error(task, calldata->server, NULL) == -EAGAIN) - rpc_restart_call(task); + nfs4_restart_rpc(task, + calldata->server->nfs_client); } nfs4_sequence_free_slot(calldata->server->nfs_client, &calldata->res.seq_res); @@ -4237,7 +4238,7 @@ static void nfs4_get_lease_time_done(struct rpc_task *task, void *calldata) dprintk("%s Retry: tk_status %d\n", __func__, task->tk_status); rpc_delay(task, NFS4_POLL_RETRY_MIN); task->tk_status = 0; - rpc_restart_call(task); + nfs4_restart_rpc(task, data->clp); return; } nfs41_sequence_free_slot(data->clp, &data->res->lr_seq_res); @@ -4647,7 +4648,7 @@ void nfs41_sequence_call_done(struct rpc_task *task, void *data) if (_nfs4_async_handle_error(task, NULL, clp, NULL) == -EAGAIN) { - rpc_restart_call(task); + nfs4_restart_rpc(task, clp); return; } } diff --git a/fs/nfs/read.c b/fs/nfs/read.c index d71f0d55ebde..96c4ebfa46f4 100644 --- a/fs/nfs/read.c +++ b/fs/nfs/read.c @@ -371,7 +371,7 @@ static void nfs_readpage_retry(struct rpc_task *task, struct nfs_read_data *data argp->offset += resp->count; argp->pgbase += resp->count; argp->count -= resp->count; - rpc_restart_call(task); + nfs4_restart_rpc(task, NFS_SERVER(data->inode)->nfs_client); return; out: nfs4_sequence_free_slot(NFS_SERVER(data->inode)->nfs_client, diff --git a/fs/nfs/unlink.c b/fs/nfs/unlink.c index 089a21b4e3f9..1064c91ae810 100644 --- a/fs/nfs/unlink.c +++ b/fs/nfs/unlink.c @@ -83,7 +83,7 @@ static void nfs_async_unlink_done(struct rpc_task *task, void *calldata) struct inode *dir = data->dir; if (!NFS_PROTO(dir)->unlink_done(task, dir)) - rpc_restart_call(task); + nfs4_restart_rpc(task, NFS_SERVER(dir)->nfs_client); } /** diff --git a/fs/nfs/write.c b/fs/nfs/write.c index 85a76409de13..ce728829f79a 100644 --- a/fs/nfs/write.c +++ b/fs/nfs/write.c @@ -1145,6 +1145,7 @@ int nfs_writeback_done(struct rpc_task *task, struct nfs_write_data *data) { struct nfs_writeargs *argp = &data->args; struct nfs_writeres *resp = &data->res; + struct nfs_server *server = NFS_SERVER(data->inode); int status; dprintk("NFS: %5u nfs_writeback_done (status %d)\n", @@ -1177,7 +1178,7 @@ int nfs_writeback_done(struct rpc_task *task, struct nfs_write_data *data) if (time_before(complain, jiffies)) { dprintk("NFS: faulty NFS server %s:" " (committed = %d) != (stable = %d)\n", - NFS_SERVER(data->inode)->nfs_client->cl_hostname, + server->nfs_client->cl_hostname, resp->verf->committed, argp->stable); complain = jiffies + 300 * HZ; } @@ -1203,7 +1204,7 @@ int nfs_writeback_done(struct rpc_task *task, struct nfs_write_data *data) */ argp->stable = NFS_FILE_SYNC; } - rpc_restart_call(task); + nfs4_restart_rpc(task, server->nfs_client); return -EAGAIN; } if (time_before(complain, jiffies)) { @@ -1215,8 +1216,7 @@ int nfs_writeback_done(struct rpc_task *task, struct nfs_write_data *data) /* Can't do anything about it except throw an error. */ task->tk_status = -EIO; } - nfs4_sequence_free_slot(NFS_SERVER(data->inode)->nfs_client, - &data->res.seq_res); + nfs4_sequence_free_slot(server->nfs_client, &data->res.seq_res); return 0; } -- cgit v1.2.3 From 4745e3154bf33048d970d1519703089493988f65 Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Wed, 1 Apr 2009 09:22:42 -0400 Subject: nfs41: kick start nfs41 session recovery when handling errors Remove checking for any errors that the SEQUENCE operation does not return. -NFS4ERR_STALE_CLIENTID, NFS4ERR_EXPIRED, NFS4ERR_CB_PATH_DOWN, NFS4ERR_BACK_CHAN_BUSY, NFS4ERR_OP_NOT_IN_SESSION. SEQUENCE operation error recovery is very primative, we only reset the session. Remove checking for any errors that are returned by the SEQUENCE operation, but that resetting the session won't address. NFS4ERR_RETRY_UNCACHED_REP, NFS4ERR_SEQUENCE_POS,NFS4ERR_TOO_MANY_OPS. Add error checking for missing SEQUENCE errors that a session reset will address. NFS4ERR_BAD_HIGH_SLOT, NFS4ERR_DEADSESSION, NFS4ERR_SEQ_FALSE_RETRY. A reset of the session is currently our only response to a SEQUENCE operation error. Don't reset the session on errors where a new session won't help. Don't reset the session on errors where a new session won't help. [nfs41: nfs4_async_handle_error update error checking] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [nfs41: trigger the state manager for session reset] Replace session state bit with nfs_client state bit. Set the NFS4CLNT_SESSION_SETUP bit upon a session related error in the sync/async error handlers. [nfs41: _nfs4_async_handle_error fix session reset error list] Sequence operation errors that session reset could help. NFS4ERR_BADSESSION NFS4ERR_BADSLOT NFS4ERR_BAD_HIGH_SLOT NFS4ERR_DEADSESSION NFS4ERR_CONN_NOT_BOUND_TO_SESSION NFS4ERR_SEQ_FALSE_RETRY NFS4ERR_SEQ_MISORDERED Sequence operation errors that a session reset would not help NFS4ERR_BADXDR NFS4ERR_DELAY NFS4ERR_REP_TOO_BIG NFS4ERR_REP_TOO_BIG_TO_CACHE NFS4ERR_REQ_TOO_BIG NFS4ERR_RETRY_UNCACHED_REP NFS4ERR_SEQUENCE_POS NFS4ERR_TOO_MANY_OPS Signed-off-by: Andy Adamson [nfs41 nfs4_handle_exception fix session reset error list] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [moved nfs41_sequece_call_done code to nfs41: sequence operation] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 32 ++++++++++++++++++++++++++++++++ 1 file changed, 32 insertions(+) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 2d45606a8085..ed6c8899806d 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -249,7 +249,25 @@ static int nfs4_handle_exception(const struct nfs_server *server, int errorcode, ret = nfs4_wait_clnt_recover(clp); if (ret == 0) exception->retry = 1; +#if !defined(CONFIG_NFS_V4_1) break; +#else /* !defined(CONFIG_NFS_V4_1) */ + if (!nfs4_has_session(server->nfs_client)) + break; + /* FALLTHROUGH */ + case -NFS4ERR_BADSESSION: + case -NFS4ERR_BADSLOT: + case -NFS4ERR_BAD_HIGH_SLOT: + case -NFS4ERR_CONN_NOT_BOUND_TO_SESSION: + case -NFS4ERR_DEADSESSION: + case -NFS4ERR_SEQ_FALSE_RETRY: + case -NFS4ERR_SEQ_MISORDERED: + dprintk("%s ERROR: %d Reset session\n", __func__, + errorcode); + set_bit(NFS4CLNT_SESSION_SETUP, &clp->cl_state); + exception->retry = 1; + /* FALLTHROUGH */ +#endif /* !defined(CONFIG_NFS_V4_1) */ case -NFS4ERR_FILE_OPEN: case -NFS4ERR_GRACE: case -NFS4ERR_DELAY: @@ -3241,6 +3259,20 @@ _nfs4_async_handle_error(struct rpc_task *task, const struct nfs_server *server, rpc_wake_up_queued_task(&clp->cl_rpcwaitq, task); task->tk_status = 0; return -EAGAIN; +#if defined(CONFIG_NFS_V4_1) + case -NFS4ERR_BADSESSION: + case -NFS4ERR_BADSLOT: + case -NFS4ERR_BAD_HIGH_SLOT: + case -NFS4ERR_DEADSESSION: + case -NFS4ERR_CONN_NOT_BOUND_TO_SESSION: + case -NFS4ERR_SEQ_FALSE_RETRY: + case -NFS4ERR_SEQ_MISORDERED: + dprintk("%s ERROR %d, Reset session\n", __func__, + task->tk_status); + set_bit(NFS4CLNT_SESSION_SETUP, &clp->cl_state); + task->tk_status = 0; + return -EAGAIN; +#endif /* CONFIG_NFS_V4_1 */ case -NFS4ERR_DELAY: if (server) nfs_inc_server_stats(server, NFSIOS_DELAY); -- cgit v1.2.3 From b069d94af7785750228287c35f9ce5c463515726 Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Wed, 1 Apr 2009 09:22:43 -0400 Subject: nfs41: schedule async session reset Define a new session reset state which is set upon a sequence operation error in both the sync and async error handlers. Place all new requests and all but the last outstanding rpc on the slot_tbl_waitq. Spawn the recovery thread when the last slot is free. Call nfs4_proc_destroy_session, reinitialize the session, call nfs4_proc_create_session, clear the session reset state, and wake up the next task on the slot_tbl_waitq. Return the nfs4_proc_destroy_session status to the session reclaimer and check for NFS4ERR_BADSESSION and NFS4ERR_DEADSESSION. Other destroy session errors should be handled in nfs4_proc_destroy_session where the call can be retried with adjusted arguments. Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy nfs41: make nfs4_wait_bit_killable public] nfs4_wait_bit_killable to be used by NFSv4.1 session recover logic. Signed-off-by: Rahul Iyer Signed-off-by: Benny Halevy [nfs41: have create_session work on nfs_client] Signed-off-by: Andy Adamson [nfs41: trigger the state manager for session reset] Replace the session reset state with the NFS4CLNT_SESSION_SETUP cl_state. Place all rpc tasks to sleep on the slot table waitqueue until the slot table is drained, then schedule state recovery and wait for it to complete. Signed-off-by: Andy Adamson [nfs41: remove nfs41_session_recovery [ch] Replaced by using the nfs4_state_manager. Signed-off-by: Andy Adamson [nfs41: nfs4_wait_bit_killable only used locally] [nfs41: keep nfs4_wait_bit_killable static] [nfs41: keep const nfs_server in nfs4_handle_exception] [nfs41: remove session parameter from nfs4_find_slot] Signed-off-by: Andy Adamson [nfs41: resset the session from nfs41_setup_sequence] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 34 ++++++++++++++++++++++++++++++++++ 1 file changed, 34 insertions(+) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index ed6c8899806d..a2855391b079 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -423,6 +423,22 @@ out: return ret_id; } +static int nfs4_recover_session(struct nfs4_session *session) +{ + struct nfs_client *clp = session->clp; + int ret; + + for (;;) { + ret = nfs4_wait_clnt_recover(clp); + if (ret != 0) + return ret; + if (!test_bit(NFS4CLNT_SESSION_SETUP, &clp->cl_state)) + break; + nfs4_schedule_state_manager(clp); + } + return 0; +} + static int nfs41_setup_sequence(struct nfs4_session *session, struct nfs4_sequence_args *args, struct nfs4_sequence_res *res, @@ -431,6 +447,7 @@ static int nfs41_setup_sequence(struct nfs4_session *session, { struct nfs4_slot *slot; struct nfs4_slot_table *tbl; + int status = 0; u8 slotid; dprintk("--> %s\n", __func__); @@ -443,6 +460,23 @@ static int nfs41_setup_sequence(struct nfs4_session *session, tbl = &session->fc_slot_table; spin_lock(&tbl->slot_tbl_lock); + if (test_bit(NFS4CLNT_SESSION_SETUP, &session->clp->cl_state)) { + if (tbl->highest_used_slotid != -1) { + rpc_sleep_on(&tbl->slot_tbl_waitq, task, NULL); + spin_unlock(&tbl->slot_tbl_lock); + dprintk("<-- %s: Session reset: draining\n", __func__); + return -EAGAIN; + } + + /* The slot table is empty; start the reset thread */ + dprintk("%s Session Reset\n", __func__); + spin_unlock(&tbl->slot_tbl_lock); + status = nfs4_recover_session(session); + if (status) + return status; + spin_lock(&tbl->slot_tbl_lock); + } + slotid = nfs4_find_slot(tbl, task); if (slotid == NFS4_MAX_SLOT_TABLE) { rpc_sleep_on(&tbl->slot_tbl_waitq, task, NULL); -- cgit v1.2.3 From 29fba38b793225798104a1ac870e6807ebd078dd Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Wed, 1 Apr 2009 09:22:44 -0400 Subject: nfs41: lease renewal Send a NFSv4.1 SEQUENCE op rather than RENEW that was deprecated in minorversion 1. Use the nfs_client minorversion to select reboot_recover/ network_partition_recovery/state_renewal ops. Note: we use reclaimer to create the nfs41 session before there are any cl_superblocks for the nfs_client. Signed-off-by: Benny Halevy [nfs41: check for session not minorversion] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy [revamped patch for new nfs4_state_manager design] Signed-off-by: Benny Halevy [nfs41: obliterate nfs4_state_recovery_ops.renew_lease method] moved to nfs4_state_maintenance_ops [also undid per-minorversion nfs4_state_recovery_ops here] Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4_fs.h | 6 ++++++ fs/nfs/nfs4proc.c | 24 ++++++++++++++++++++++++ fs/nfs/nfs4renewd.c | 4 +++- 3 files changed, 33 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfs/nfs4_fs.h b/fs/nfs/nfs4_fs.h index 7cc1cddfd9ba..ae25cc21fe70 100644 --- a/fs/nfs/nfs4_fs.h +++ b/fs/nfs/nfs4_fs.h @@ -180,6 +180,10 @@ struct nfs4_state_recovery_ops { int (*recover_lock)(struct nfs4_state *, struct file_lock *); }; +struct nfs4_state_maintenance_ops { + int (*sched_state_renewal)(struct nfs_client *, struct rpc_cred *); +}; + extern const struct dentry_operations nfs4_dentry_operations; extern const struct inode_operations nfs4_dir_inode_operations; @@ -220,6 +224,8 @@ static inline int nfs4_setup_sequence(struct nfs_client *clp, } #endif /* CONFIG_NFS_V4_1 */ +extern struct nfs4_state_maintenance_ops *nfs4_state_renewal_ops[]; + extern const u32 nfs4_fattr_bitmap[2]; extern const u32 nfs4_statfs_bitmap[2]; extern const u32 nfs4_pathconf_bitmap[2]; diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index a2855391b079..101ba6aed101 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -3016,6 +3016,9 @@ static void nfs4_renew_done(struct rpc_task *task, void *data) if (time_before(clp->cl_last_renewal,timestamp)) clp->cl_last_renewal = timestamp; spin_unlock(&clp->cl_lock); + dprintk("%s calling put_rpccred on rpc_cred %p\n", __func__, + task->tk_msg.rpc_cred); + put_rpccred(task->tk_msg.rpc_cred); } static const struct rpc_call_ops nfs4_renew_ops = { @@ -4790,6 +4793,27 @@ struct nfs4_state_recovery_ops nfs4_nograce_recovery_ops = { .recover_lock = nfs4_lock_expired, }; +struct nfs4_state_maintenance_ops nfs40_state_renewal_ops = { + .sched_state_renewal = nfs4_proc_async_renew, +}; + +#if defined(CONFIG_NFS_V4_1) +struct nfs4_state_maintenance_ops nfs41_state_renewal_ops = { + .sched_state_renewal = nfs41_proc_async_sequence, +}; +#endif + +/* + * Per minor version reboot and network partition recovery ops + */ + +struct nfs4_state_maintenance_ops *nfs4_state_renewal_ops[] = { + &nfs40_state_renewal_ops, +#if defined(CONFIG_NFS_V4_1) + &nfs41_state_renewal_ops, +#endif +}; + static const struct inode_operations nfs4_file_inode_operations = { .permission = nfs_permission, .getattr = nfs_getattr, diff --git a/fs/nfs/nfs4renewd.c b/fs/nfs/nfs4renewd.c index f524e932ff7b..92b43e87267c 100644 --- a/fs/nfs/nfs4renewd.c +++ b/fs/nfs/nfs4renewd.c @@ -59,12 +59,14 @@ void nfs4_renew_state(struct work_struct *work) { + struct nfs4_state_maintenance_ops *ops; struct nfs_client *clp = container_of(work, struct nfs_client, cl_renewd.work); struct rpc_cred *cred; long lease, timeout; unsigned long last, now; + ops = nfs4_state_renewal_ops[clp->cl_minorversion]; dprintk("%s: start\n", __func__); /* Are there any active superblocks? */ if (list_empty(&clp->cl_superblocks)) @@ -86,7 +88,7 @@ nfs4_renew_state(struct work_struct *work) nfs_expire_all_delegations(clp); } else { /* Queue an asynchronous RENEW. */ - nfs4_proc_async_renew(clp, cred); + ops->sched_state_renewal(clp, cred); put_rpccred(cred); } timeout = (2 * lease) / 3; -- cgit v1.2.3 From 8e69514f2981d85108c2bd220ff8e188c0c27cdb Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Wed, 1 Apr 2009 09:22:45 -0400 Subject: nfs41: support minorversion 1 for nfs4_check_lease [moved nfs4_get_renew_cred related changes to "nfs41: introduce get_state_renewal_cred"] Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4_fs.h | 1 + fs/nfs/nfs4proc.c | 2 ++ fs/nfs/nfs4state.c | 4 +++- 3 files changed, 6 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfs/nfs4_fs.h b/fs/nfs/nfs4_fs.h index ae25cc21fe70..095bf4690e84 100644 --- a/fs/nfs/nfs4_fs.h +++ b/fs/nfs/nfs4_fs.h @@ -182,6 +182,7 @@ struct nfs4_state_recovery_ops { struct nfs4_state_maintenance_ops { int (*sched_state_renewal)(struct nfs_client *, struct rpc_cred *); + int (*renew_lease)(struct nfs_client *, struct rpc_cred *); }; extern const struct dentry_operations nfs4_dentry_operations; diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 101ba6aed101..da481bda44f7 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -4795,11 +4795,13 @@ struct nfs4_state_recovery_ops nfs4_nograce_recovery_ops = { struct nfs4_state_maintenance_ops nfs40_state_renewal_ops = { .sched_state_renewal = nfs4_proc_async_renew, + .renew_lease = nfs4_proc_renew, }; #if defined(CONFIG_NFS_V4_1) struct nfs4_state_maintenance_ops nfs41_state_renewal_ops = { .sched_state_renewal = nfs41_proc_async_sequence, + .renew_lease = nfs4_proc_sequence, }; #endif diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c index 8b7f007adc42..dc5599c7f4d9 100644 --- a/fs/nfs/nfs4state.c +++ b/fs/nfs/nfs4state.c @@ -1083,6 +1083,8 @@ restart: static int nfs4_check_lease(struct nfs_client *clp) { struct rpc_cred *cred; + struct nfs4_state_maintenance_ops *ops = + nfs4_state_renewal_ops[clp->cl_minorversion]; int status = -NFS4ERR_EXPIRED; /* Is the client already known to have an expired lease? */ @@ -1094,7 +1096,7 @@ static int nfs4_check_lease(struct nfs_client *clp) if (cred == NULL) goto out; } - status = nfs4_proc_renew(clp, cred); + status = ops->renew_lease(clp, cred); put_rpccred(cred); out: nfs4_recovery_handle_error(clp, status); -- cgit v1.2.3 From a7b721037f898b29a8083da59b1dccd3da385b07 Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Wed, 1 Apr 2009 09:22:46 -0400 Subject: nfs41: introduce get_state_renewal_cred Use the machine cred for sending SEQUENCE to renew the client's lease. [revamp patch for new state management design starting 2.6.29] [nfs41: support minorversion 1 for nfs4_check_lease] Signed-off-by: Benny Halevy [nfs41: get cred in exchange_id when cred arg is NULL] Signed-off-by: Benny Halevy [nfs41: use cl_machined_cred instead of cl_ex_cred] Since EXCHANGE_ID insists on using the machine credential, cl_ex_cred is not needed. nfs4_proc_exchange_id() is only called if the machine credential is available. Remove the credential logic from nfs4_proc_exchange_id. Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4_fs.h | 5 +++++ fs/nfs/nfs4proc.c | 3 +++ fs/nfs/nfs4renewd.c | 2 +- fs/nfs/nfs4state.c | 18 +++++------------- 4 files changed, 14 insertions(+), 14 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs4_fs.h b/fs/nfs/nfs4_fs.h index 095bf4690e84..2b141c5758ec 100644 --- a/fs/nfs/nfs4_fs.h +++ b/fs/nfs/nfs4_fs.h @@ -182,6 +182,7 @@ struct nfs4_state_recovery_ops { struct nfs4_state_maintenance_ops { int (*sched_state_renewal)(struct nfs_client *, struct rpc_cred *); + struct rpc_cred * (*get_state_renewal_cred_locked)(struct nfs_client *); int (*renew_lease)(struct nfs_client *, struct rpc_cred *); }; @@ -240,7 +241,11 @@ extern void nfs4_kill_renewd(struct nfs_client *); extern void nfs4_renew_state(struct work_struct *); /* nfs4state.c */ +struct rpc_cred *nfs4_get_setclientid_cred(struct nfs_client *clp); struct rpc_cred *nfs4_get_renew_cred_locked(struct nfs_client *clp); +#if defined(CONFIG_NFS_V4_1) +struct rpc_cred *nfs4_get_machine_cred_locked(struct nfs_client *clp); +#endif /* CONFIG_NFS_V4_1 */ extern struct nfs4_state_owner * nfs4_get_state_owner(struct nfs_server *, struct rpc_cred *); extern void nfs4_put_state_owner(struct nfs4_state_owner *); diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index da481bda44f7..f862a66e5fb9 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -4236,6 +4236,7 @@ static int nfs4_proc_exchange_id(struct nfs_client *clp, struct rpc_cred *cred) dprintk("--> %s\n", __func__); BUG_ON(clp == NULL); + p = (u32 *)verifier.data; *p++ = htonl((u32)clp->cl_boot_time.tv_sec); *p = htonl((u32)clp->cl_boot_time.tv_nsec); @@ -4795,12 +4796,14 @@ struct nfs4_state_recovery_ops nfs4_nograce_recovery_ops = { struct nfs4_state_maintenance_ops nfs40_state_renewal_ops = { .sched_state_renewal = nfs4_proc_async_renew, + .get_state_renewal_cred_locked = nfs4_get_renew_cred_locked, .renew_lease = nfs4_proc_renew, }; #if defined(CONFIG_NFS_V4_1) struct nfs4_state_maintenance_ops nfs41_state_renewal_ops = { .sched_state_renewal = nfs41_proc_async_sequence, + .get_state_renewal_cred_locked = nfs4_get_machine_cred_locked, .renew_lease = nfs4_proc_sequence, }; #endif diff --git a/fs/nfs/nfs4renewd.c b/fs/nfs/nfs4renewd.c index 92b43e87267c..e27c6cef18f2 100644 --- a/fs/nfs/nfs4renewd.c +++ b/fs/nfs/nfs4renewd.c @@ -78,7 +78,7 @@ nfs4_renew_state(struct work_struct *work) timeout = (2 * lease) / 3 + (long)last - (long)now; /* Are we close to a lease timeout? */ if (time_after(now, last + lease/3)) { - cred = nfs4_get_renew_cred_locked(clp); + cred = ops->get_state_renewal_cred_locked(clp); spin_unlock(&clp->cl_lock); if (cred == NULL) { if (list_empty(&clp->cl_delegations)) { diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c index dc5599c7f4d9..da940abcfaa5 100644 --- a/fs/nfs/nfs4state.c +++ b/fs/nfs/nfs4state.c @@ -77,7 +77,7 @@ static int nfs4_init_client(struct nfs_client *clp, struct rpc_cred *cred) return status; } -static struct rpc_cred *nfs4_get_machine_cred_locked(struct nfs_client *clp) +struct rpc_cred *nfs4_get_machine_cred_locked(struct nfs_client *clp) { struct rpc_cred *cred = NULL; @@ -114,17 +114,7 @@ struct rpc_cred *nfs4_get_renew_cred_locked(struct nfs_client *clp) return cred; } -static struct rpc_cred *nfs4_get_renew_cred(struct nfs_client *clp) -{ - struct rpc_cred *cred; - - spin_lock(&clp->cl_lock); - cred = nfs4_get_renew_cred_locked(clp); - spin_unlock(&clp->cl_lock); - return cred; -} - -static struct rpc_cred *nfs4_get_setclientid_cred(struct nfs_client *clp) +struct rpc_cred *nfs4_get_setclientid_cred(struct nfs_client *clp) { struct nfs4_state_owner *sp; struct rb_node *pos; @@ -1090,7 +1080,9 @@ static int nfs4_check_lease(struct nfs_client *clp) /* Is the client already known to have an expired lease? */ if (test_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state)) return 0; - cred = nfs4_get_renew_cred(clp); + spin_lock(&clp->cl_lock); + cred = ops->get_state_renewal_cred_locked(clp); + spin_unlock(&clp->cl_lock); if (cred == NULL) { cred = nfs4_get_setclientid_cred(clp); if (cred == NULL) -- cgit v1.2.3 From 591d71cbde186cc498c0d9718dc17f2fadf7c643 Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Wed, 1 Apr 2009 09:22:47 -0400 Subject: nfs41: establish sessions-based clientid nfsv4.1 clientid is established via EXCHANGE_ID rather than SETCLIENTID{,_CONFIRM} This is implemented using a new establish_clid method in nfs4_state_recovery_ops. nfs41: establish clientid via exchange id only if cred != NULL >From 2.6.26 reclaimer() uses machine cred for setting up the client id therefore it is never expected to be NULL. Signed-off-by: Rahul Iyer [removed dprintk] Signed-off-by: Benny Halevy [nfs41: lease renewal] [revamped patch for new nfs4_state_manager design] Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4_fs.h | 6 ++++-- fs/nfs/nfs4proc.c | 40 ++++++++++++++++++++++++++++++++++++++-- fs/nfs/nfs4state.c | 12 ++++++++---- 3 files changed, 50 insertions(+), 8 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs4_fs.h b/fs/nfs/nfs4_fs.h index 2b141c5758ec..df9e36319499 100644 --- a/fs/nfs/nfs4_fs.h +++ b/fs/nfs/nfs4_fs.h @@ -178,6 +178,7 @@ struct nfs4_state_recovery_ops { int state_flag_bit; int (*recover_open)(struct nfs4_state_owner *, struct nfs4_state *); int (*recover_lock)(struct nfs4_state *, struct file_lock *); + int (*establish_clid)(struct nfs_client *, struct rpc_cred *); }; struct nfs4_state_maintenance_ops { @@ -200,6 +201,7 @@ extern int nfs4_proc_setclientid(struct nfs_client *, u32, unsigned short, struc extern int nfs4_proc_setclientid_confirm(struct nfs_client *, struct rpc_cred *); extern int nfs4_proc_async_renew(struct nfs_client *, struct rpc_cred *); extern int nfs4_proc_renew(struct nfs_client *, struct rpc_cred *); +extern int nfs4_init_clientid(struct nfs_client *, struct rpc_cred *); extern int nfs4_do_close(struct path *path, struct nfs4_state *state, int wait); extern struct dentry *nfs4_atomic_open(struct inode *, struct dentry *, struct nameidata *); extern int nfs4_open_revalidate(struct inode *, struct dentry *, int, struct nameidata *); @@ -207,8 +209,8 @@ extern int nfs4_server_capabilities(struct nfs_server *server, struct nfs_fh *fh extern int nfs4_proc_fs_locations(struct inode *dir, const struct qstr *name, struct nfs4_fs_locations *fs_locations, struct page *page); -extern struct nfs4_state_recovery_ops nfs4_reboot_recovery_ops; -extern struct nfs4_state_recovery_ops nfs4_nograce_recovery_ops; +extern struct nfs4_state_recovery_ops *nfs4_reboot_recovery_ops[]; +extern struct nfs4_state_recovery_ops *nfs4_nograce_recovery_ops[]; #if defined(CONFIG_NFS_V4_1) extern int nfs4_setup_sequence(struct nfs_client *clp, struct nfs4_sequence_args *args, struct nfs4_sequence_res *res, diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index f862a66e5fb9..dd46339095ce 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -4780,19 +4780,41 @@ static int nfs41_proc_async_sequence(struct nfs_client *clp, #endif /* CONFIG_NFS_V4_1 */ -struct nfs4_state_recovery_ops nfs4_reboot_recovery_ops = { +struct nfs4_state_recovery_ops nfs40_reboot_recovery_ops = { .owner_flag_bit = NFS_OWNER_RECLAIM_REBOOT, .state_flag_bit = NFS_STATE_RECLAIM_REBOOT, .recover_open = nfs4_open_reclaim, .recover_lock = nfs4_lock_reclaim, + .establish_clid = nfs4_init_clientid, }; -struct nfs4_state_recovery_ops nfs4_nograce_recovery_ops = { +#if defined(CONFIG_NFS_V4_1) +struct nfs4_state_recovery_ops nfs41_reboot_recovery_ops = { + .owner_flag_bit = NFS_OWNER_RECLAIM_REBOOT, + .state_flag_bit = NFS_STATE_RECLAIM_REBOOT, + .recover_open = nfs4_open_reclaim, + .recover_lock = nfs4_lock_reclaim, + .establish_clid = nfs4_proc_exchange_id, +}; +#endif /* CONFIG_NFS_V4_1 */ + +struct nfs4_state_recovery_ops nfs40_nograce_recovery_ops = { + .owner_flag_bit = NFS_OWNER_RECLAIM_NOGRACE, + .state_flag_bit = NFS_STATE_RECLAIM_NOGRACE, + .recover_open = nfs4_open_expired, + .recover_lock = nfs4_lock_expired, + .establish_clid = nfs4_init_clientid, +}; + +#if defined(CONFIG_NFS_V4_1) +struct nfs4_state_recovery_ops nfs41_nograce_recovery_ops = { .owner_flag_bit = NFS_OWNER_RECLAIM_NOGRACE, .state_flag_bit = NFS_STATE_RECLAIM_NOGRACE, .recover_open = nfs4_open_expired, .recover_lock = nfs4_lock_expired, + .establish_clid = nfs4_proc_exchange_id, }; +#endif /* CONFIG_NFS_V4_1 */ struct nfs4_state_maintenance_ops nfs40_state_renewal_ops = { .sched_state_renewal = nfs4_proc_async_renew, @@ -4812,6 +4834,20 @@ struct nfs4_state_maintenance_ops nfs41_state_renewal_ops = { * Per minor version reboot and network partition recovery ops */ +struct nfs4_state_recovery_ops *nfs4_reboot_recovery_ops[] = { + &nfs40_reboot_recovery_ops, +#if defined(CONFIG_NFS_V4_1) + &nfs41_reboot_recovery_ops, +#endif +}; + +struct nfs4_state_recovery_ops *nfs4_nograce_recovery_ops[] = { + &nfs40_nograce_recovery_ops, +#if defined(CONFIG_NFS_V4_1) + &nfs41_nograce_recovery_ops, +#endif +}; + struct nfs4_state_maintenance_ops *nfs4_state_renewal_ops[] = { &nfs40_state_renewal_ops, #if defined(CONFIG_NFS_V4_1) diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c index da940abcfaa5..e17bd4412174 100644 --- a/fs/nfs/nfs4state.c +++ b/fs/nfs/nfs4state.c @@ -60,7 +60,7 @@ const nfs4_stateid zero_stateid; static LIST_HEAD(nfs4_clientid_list); -static int nfs4_init_client(struct nfs_client *clp, struct rpc_cred *cred) +int nfs4_init_clientid(struct nfs_client *clp, struct rpc_cred *cred) { unsigned short port; int status; @@ -1098,11 +1098,13 @@ out: static int nfs4_reclaim_lease(struct nfs_client *clp) { struct rpc_cred *cred; + struct nfs4_state_recovery_ops *ops = + nfs4_reboot_recovery_ops[clp->cl_minorversion]; int status = -ENOENT; cred = nfs4_get_setclientid_cred(clp); if (cred != NULL) { - status = nfs4_init_client(clp, cred); + status = ops->establish_clid(clp, cred); put_rpccred(cred); /* Handle case where the user hasn't set up machine creds */ if (status == -EACCES && cred == clp->cl_machine_cred) { @@ -1208,7 +1210,8 @@ static void nfs4_state_manager(struct nfs_client *clp) } /* First recover reboot state... */ if (test_and_clear_bit(NFS4CLNT_RECLAIM_REBOOT, &clp->cl_state)) { - status = nfs4_do_reclaim(clp, &nfs4_reboot_recovery_ops); + status = nfs4_do_reclaim(clp, + nfs4_reboot_recovery_ops[clp->cl_minorversion]); if (status == -NFS4ERR_STALE_CLIENTID) continue; if (test_bit(NFS4CLNT_SESSION_SETUP, &clp->cl_state)) @@ -1219,7 +1222,8 @@ static void nfs4_state_manager(struct nfs_client *clp) /* Now recover expired state... */ if (test_and_clear_bit(NFS4CLNT_RECLAIM_NOGRACE, &clp->cl_state)) { - status = nfs4_do_reclaim(clp, &nfs4_nograce_recovery_ops); + status = nfs4_do_reclaim(clp, + nfs4_nograce_recovery_ops[clp->cl_minorversion]); if (status < 0) { set_bit(NFS4CLNT_RECLAIM_NOGRACE, &clp->cl_state); if (status == -NFS4ERR_STALE_CLIENTID) -- cgit v1.2.3 From 90a16617ee6a052c3a1aac00eb67136324cf4dd0 Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Wed, 1 Apr 2009 09:22:48 -0400 Subject: nfs41: add a get_clid_cred function to nfs4_state_recovery_ops EXCHANGE_ID has different credential requirements than SETCLIENTID. Prepare for a separate credential function. Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4_fs.h | 1 + fs/nfs/nfs4proc.c | 2 ++ fs/nfs/nfs4state.c | 2 +- 3 files changed, 4 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfs/nfs4_fs.h b/fs/nfs/nfs4_fs.h index df9e36319499..1da3e354de22 100644 --- a/fs/nfs/nfs4_fs.h +++ b/fs/nfs/nfs4_fs.h @@ -179,6 +179,7 @@ struct nfs4_state_recovery_ops { int (*recover_open)(struct nfs4_state_owner *, struct nfs4_state *); int (*recover_lock)(struct nfs4_state *, struct file_lock *); int (*establish_clid)(struct nfs_client *, struct rpc_cred *); + struct rpc_cred * (*get_clid_cred)(struct nfs_client *); }; struct nfs4_state_maintenance_ops { diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index dd46339095ce..b4e5442efa6c 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -4786,6 +4786,7 @@ struct nfs4_state_recovery_ops nfs40_reboot_recovery_ops = { .recover_open = nfs4_open_reclaim, .recover_lock = nfs4_lock_reclaim, .establish_clid = nfs4_init_clientid, + .get_clid_cred = nfs4_get_setclientid_cred, }; #if defined(CONFIG_NFS_V4_1) @@ -4804,6 +4805,7 @@ struct nfs4_state_recovery_ops nfs40_nograce_recovery_ops = { .recover_open = nfs4_open_expired, .recover_lock = nfs4_lock_expired, .establish_clid = nfs4_init_clientid, + .get_clid_cred = nfs4_get_setclientid_cred, }; #if defined(CONFIG_NFS_V4_1) diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c index e17bd4412174..90e56072844b 100644 --- a/fs/nfs/nfs4state.c +++ b/fs/nfs/nfs4state.c @@ -1102,7 +1102,7 @@ static int nfs4_reclaim_lease(struct nfs_client *clp) nfs4_reboot_recovery_ops[clp->cl_minorversion]; int status = -ENOENT; - cred = nfs4_get_setclientid_cred(clp); + cred = ops->get_clid_cred(clp); if (cred != NULL) { status = ops->establish_clid(clp, cred); put_rpccred(cred); -- cgit v1.2.3 From b4b82607ffcf09b57301846d154f2c09c0b807c0 Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Wed, 1 Apr 2009 09:22:49 -0400 Subject: nfs41: get_clid_cred for EXCHANGE_ID Unlike SETCLIENTID, EXCHANGE_ID requires a machine credential. Do not search for credentials other than the machine credential. Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4_fs.h | 1 + fs/nfs/nfs4proc.c | 2 ++ fs/nfs/nfs4state.c | 14 ++++++++++++++ 3 files changed, 17 insertions(+) (limited to 'fs') diff --git a/fs/nfs/nfs4_fs.h b/fs/nfs/nfs4_fs.h index 1da3e354de22..61bc3a32e1e2 100644 --- a/fs/nfs/nfs4_fs.h +++ b/fs/nfs/nfs4_fs.h @@ -248,6 +248,7 @@ struct rpc_cred *nfs4_get_setclientid_cred(struct nfs_client *clp); struct rpc_cred *nfs4_get_renew_cred_locked(struct nfs_client *clp); #if defined(CONFIG_NFS_V4_1) struct rpc_cred *nfs4_get_machine_cred_locked(struct nfs_client *clp); +struct rpc_cred *nfs4_get_exchange_id_cred(struct nfs_client *clp); #endif /* CONFIG_NFS_V4_1 */ extern struct nfs4_state_owner * nfs4_get_state_owner(struct nfs_server *, struct rpc_cred *); diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index b4e5442efa6c..64d611658d1f 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -4796,6 +4796,7 @@ struct nfs4_state_recovery_ops nfs41_reboot_recovery_ops = { .recover_open = nfs4_open_reclaim, .recover_lock = nfs4_lock_reclaim, .establish_clid = nfs4_proc_exchange_id, + .get_clid_cred = nfs4_get_exchange_id_cred, }; #endif /* CONFIG_NFS_V4_1 */ @@ -4815,6 +4816,7 @@ struct nfs4_state_recovery_ops nfs41_nograce_recovery_ops = { .recover_open = nfs4_open_expired, .recover_lock = nfs4_lock_expired, .establish_clid = nfs4_proc_exchange_id, + .get_clid_cred = nfs4_get_exchange_id_cred, }; #endif /* CONFIG_NFS_V4_1 */ diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c index 90e56072844b..7dc971335ec2 100644 --- a/fs/nfs/nfs4state.c +++ b/fs/nfs/nfs4state.c @@ -114,6 +114,20 @@ struct rpc_cred *nfs4_get_renew_cred_locked(struct nfs_client *clp) return cred; } +#if defined(CONFIG_NFS_V4_1) + +struct rpc_cred *nfs4_get_exchange_id_cred(struct nfs_client *clp) +{ + struct rpc_cred *cred; + + spin_lock(&clp->cl_lock); + cred = nfs4_get_machine_cred_locked(clp); + spin_unlock(&clp->cl_lock); + return cred; +} + +#endif /* CONFIG_NFS_V4_1 */ + struct rpc_cred *nfs4_get_setclientid_cred(struct nfs_client *clp) { struct nfs4_state_owner *sp; -- cgit v1.2.3 From 008f55d0e019943323c20a03493a2ba5672a4cc8 Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Wed, 1 Apr 2009 09:22:50 -0400 Subject: nfs41: recover lease in _nfs4_lookup_root This creates the nfsv4.1 session on mount. Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/client.c | 16 ++++++++++++++++ fs/nfs/internal.h | 1 + fs/nfs/nfs4proc.c | 9 ++++++++- 3 files changed, 25 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfs/client.c b/fs/nfs/client.c index d28a987f569e..3e232bf56dfb 100644 --- a/fs/nfs/client.c +++ b/fs/nfs/client.c @@ -527,6 +527,22 @@ void nfs_mark_client_ready(struct nfs_client *clp, int state) wake_up_all(&nfs_client_active_wq); } +/* + * With sessions, the client is not marked ready until after a + * successful EXCHANGE_ID and CREATE_SESSION. + * + * Map errors cl_cons_state errors to EPROTONOSUPPORT to indicate + * other versions of NFS can be tried. + */ +int nfs4_check_client_ready(struct nfs_client *clp) +{ + if (!nfs4_has_session(clp)) + return 0; + if (clp->cl_cons_state < NFS_CS_READY) + return -EPROTONOSUPPORT; + return 0; +} + /* * Initialise the timeout values for a connection */ diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h index 78508f29c03a..acee3274d275 100644 --- a/fs/nfs/internal.h +++ b/fs/nfs/internal.h @@ -114,6 +114,7 @@ extern struct nfs_server *nfs_clone_server(struct nfs_server *, struct nfs_fh *, struct nfs_fattr *); extern void nfs_mark_client_ready(struct nfs_client *clp, int state); +extern int nfs4_check_client_ready(struct nfs_client *clp); #ifdef CONFIG_PROC_FS extern int __init nfs_fs_proc_init(void); extern void nfs_fs_proc_exit(void); diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 64d611658d1f..31ce758d53d5 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -2024,8 +2024,15 @@ static int _nfs4_lookup_root(struct nfs_server *server, struct nfs_fh *fhandle, .rpc_argp = &args, .rpc_resp = &res, }; + int status; + nfs_fattr_init(info->fattr); - return nfs4_call_sync(server, &msg, &args, &res, 0); + status = nfs4_recover_expired_lease(server); + if (!status) + status = nfs4_check_client_ready(server->nfs_client); + if (!status) + status = nfs4_call_sync(server, &msg, &args, &res, 0); + return status; } static int nfs4_lookup_root(struct nfs_server *server, struct nfs_fh *fhandle, -- cgit v1.2.3 From 78722e9c9208a312695178f5331511badd190598 Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Tue, 16 Jun 2009 14:43:15 -0400 Subject: nfs41: only retry EXCHANGE_ID on recoverable errors Stops an infinite loop of EXCHANGE_ID. Signed-off-by: Andy Adamson [fixed checkpatch warnings] Signed-off-by: Benny Halevy --- fs/nfs/nfs4state.c | 23 ++++++++++++++++++++++- 1 file changed, 22 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c index 7dc971335ec2..2cc0aca96eb2 100644 --- a/fs/nfs/nfs4state.c +++ b/fs/nfs/nfs4state.c @@ -1183,6 +1183,27 @@ static int nfs4_reset_session(struct nfs_client *clp) { return 0; } static int nfs4_initialize_session(struct nfs_client *clp) { return 0; } #endif /* CONFIG_NFS_V4_1 */ +/* Set NFS4CLNT_LEASE_EXPIRED for all v4.0 errors and for recoverable errors + * on EXCHANGE_ID for v4.1 + */ +static void nfs4_set_lease_expired(struct nfs_client *clp, int status) +{ + if (nfs4_has_session(clp)) { + switch (status) { + case -NFS4ERR_DELAY: + case -NFS4ERR_CLID_INUSE: + case -EAGAIN: + break; + + case -NFS4ERR_NOT_SAME: /* FixMe: implement recovery + * in nfs4_exchange_id */ + default: + return; + } + } + set_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state); +} + static void nfs4_state_manager(struct nfs_client *clp) { int status = 0; @@ -1193,7 +1214,7 @@ static void nfs4_state_manager(struct nfs_client *clp) /* We're going to have to re-establish a clientid */ status = nfs4_reclaim_lease(clp); if (status) { - set_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state); + nfs4_set_lease_expired(clp, status); if (status == -EAGAIN) continue; if (clp->cl_cons_state == -- cgit v1.2.3 From 34dc1ad752ad3f55b2a6e6cd8cfcf3504682fec7 Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Wed, 1 Apr 2009 09:22:52 -0400 Subject: nfs41: increment_{open,lock}_seqid Unlike minorversion0, in nfsv4.1 the open and lock seqids need not be incremented by the client and should always be set to zero. This is implemented using a new nfs_rpc_ops methods - increment_open_seqid and increment_lock_seqid Signed-off-by: Rahul Iyer Signed-off-by: Benny Halevy [nfs41: check for session not minorversion] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 2 ++ fs/nfs/nfs4state.c | 12 +++++++----- 2 files changed, 9 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 31ce758d53d5..f1f087b483ce 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -1808,6 +1808,8 @@ int nfs4_do_close(struct path *path, struct nfs4_state *state, int wait) calldata->state = state; calldata->arg.fh = NFS_FH(state->inode); calldata->arg.stateid = &state->open_stateid; + if (nfs4_has_session(server->nfs_client)) + memset(calldata->arg.stateid->data, 0, 4); /* clear seqid */ /* Serialization for the sequence id */ calldata->arg.seqid = nfs_alloc_seqid(&state->owner->so_seqid); if (calldata->arg.seqid == NULL) diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c index 2cc0aca96eb2..2cfca9929c9a 100644 --- a/fs/nfs/nfs4state.c +++ b/fs/nfs/nfs4state.c @@ -742,12 +742,14 @@ static void nfs_increment_seqid(int status, struct nfs_seqid *seqid) void nfs_increment_open_seqid(int status, struct nfs_seqid *seqid) { - if (status == -NFS4ERR_BAD_SEQID) { - struct nfs4_state_owner *sp = container_of(seqid->sequence, - struct nfs4_state_owner, so_seqid); + struct nfs4_state_owner *sp = container_of(seqid->sequence, + struct nfs4_state_owner, so_seqid); + struct nfs_server *server = sp->so_server; + + if (status == -NFS4ERR_BAD_SEQID) nfs4_drop_state_owner(sp); - } - nfs_increment_seqid(status, seqid); + if (!nfs4_has_session(server->nfs_client)) + nfs_increment_seqid(status, seqid); } /* -- cgit v1.2.3 From 9bdaa86d2a4cbf9a71de3048c0c0a874e7ebc8ad Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Wed, 1 Apr 2009 09:22:55 -0400 Subject: nfs41: Refactor nfs4_{init,destroy}_callback for nfs4.0 Refactor-out code to bring the callback service up and down. Signed-off-by: Benny Halevy --- fs/nfs/client.c | 46 +++++++++++++++++++++++++++++++++++++--------- 1 file changed, 37 insertions(+), 9 deletions(-) (limited to 'fs') diff --git a/fs/nfs/client.c b/fs/nfs/client.c index 3e232bf56dfb..d9657d455730 100644 --- a/fs/nfs/client.c +++ b/fs/nfs/client.c @@ -47,6 +47,9 @@ #include "internal.h" #include "fscache.h" +static int nfs4_init_callback(struct nfs_client *); +static void nfs4_destroy_callback(struct nfs_client *); + #define NFSDBG_FACILITY NFSDBG_CLIENT static DEFINE_SPINLOCK(nfs_client_lock); @@ -121,11 +124,8 @@ static struct nfs_client *nfs_alloc_client(const struct nfs_client_initdata *cl_ clp->rpc_ops = cl_init->rpc_ops; - if (cl_init->rpc_ops->version == 4) { - if (nfs_callback_up() < 0) - goto error_2; - __set_bit(NFS_CS_CALLBACK, &clp->cl_res_state); - } + if (nfs4_init_callback(clp) < 0) + goto error_2; atomic_set(&clp->cl_count, 1); clp->cl_cons_state = NFS_CS_INITING; @@ -162,8 +162,7 @@ static struct nfs_client *nfs_alloc_client(const struct nfs_client_initdata *cl_ return clp; error_3: - if (__test_and_clear_bit(NFS_CS_CALLBACK, &clp->cl_res_state)) - nfs_callback_down(); + nfs4_destroy_callback(clp); error_2: kfree(clp); error_0: @@ -183,6 +182,17 @@ static void nfs4_shutdown_client(struct nfs_client *clp) #endif } +/* + * Destroy the NFS4 callback service + */ +static void nfs4_destroy_callback(struct nfs_client *clp) +{ +#ifdef CONFIG_NFS_V4 + if (__test_and_clear_bit(NFS_CS_CALLBACK, &clp->cl_res_state)) + nfs_callback_down(); +#endif /* CONFIG_NFS_V4 */ +} + /* * Clears/puts all minor version specific parts from an nfs_client struct * reverting it to minorversion 0. @@ -215,8 +225,7 @@ static void nfs_free_client(struct nfs_client *clp) if (!IS_ERR(clp->cl_rpcclient)) rpc_shutdown_client(clp->cl_rpcclient); - if (__test_and_clear_bit(NFS_CS_CALLBACK, &clp->cl_res_state)) - nfs_callback_down(); + nfs4_destroy_callback(clp); if (clp->cl_machine_cred != NULL) put_rpccred(clp->cl_machine_cred); @@ -1087,6 +1096,25 @@ error: } #ifdef CONFIG_NFS_V4 +/* + * Initialize the NFS4 callback service + */ +static int nfs4_init_callback(struct nfs_client *clp) +{ + int error; + + if (clp->rpc_ops->version == 4) { + error = nfs_callback_up(); + if (error < 0) { + dprintk("%s: failed to start callback. Error = %d\n", + __func__, error); + return error; + } + __set_bit(NFS_CS_CALLBACK, &clp->cl_res_state); + } + return 0; +} + /* * Initialize the minor version specific parts of an NFS4 client record */ -- cgit v1.2.3 From 7146851376861fe55c7a48ac8fc1354a5fff6cd0 Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Wed, 1 Apr 2009 09:22:56 -0400 Subject: nfs41: minorversion support for nfs4_{init,destroy}_callback move nfs4_init_callback into nfs4_init_client_minor_version and nfs4_destroy_callback into nfs4_clear_client_minor_version as these need to happen also when auto-negotiating the minorversion once the callback service for nfs41 becomes different than for nfs4.0 Signed-off-by: Benny Halevy [nfs41: Fix checkpatch warning] Signed-off-by: Ricardo Labiaga Signed-off-by: Benny Halevy [Type check arguments of nfs_callback_up] Signed-off-by: Ricardo Labiaga Signed-off-by: Benny Halevy [nfs41: Backchannel: Remove FIXME comment] Signed-off-by: Ricardo Labiaga Signed-off-by: Benny Halevy --- fs/nfs/callback.c | 65 +++++++++++++++++++++++++++++++++++++++---------------- fs/nfs/callback.h | 2 +- fs/nfs/client.c | 21 ++++++------------ 3 files changed, 54 insertions(+), 34 deletions(-) (limited to 'fs') diff --git a/fs/nfs/callback.c b/fs/nfs/callback.c index a886e692ddd0..4b1313eda6f5 100644 --- a/fs/nfs/callback.c +++ b/fs/nfs/callback.c @@ -59,7 +59,7 @@ module_param_call(callback_tcpport, param_set_port, param_get_int, * This is the callback kernel thread. */ static int -nfs_callback_svc(void *vrqstp) +nfs4_callback_svc(void *vrqstp) { int err, preverr = 0; struct svc_rqst *rqstp = vrqstp; @@ -97,20 +97,12 @@ nfs_callback_svc(void *vrqstp) } /* - * Bring up the callback thread if it is not already up. + * Prepare to bring up the NFSv4 callback service */ -int nfs_callback_up(void) +struct svc_rqst * +nfs4_callback_up(struct svc_serv *serv) { - struct svc_serv *serv = NULL; - int ret = 0; - - mutex_lock(&nfs_callback_mutex); - if (nfs_callback_info.users++ || nfs_callback_info.task != NULL) - goto out; - serv = svc_create(&nfs4_callback_program, NFS4_CALLBACK_BUFSIZE, NULL); - ret = -ENOMEM; - if (!serv) - goto out_err; + int ret; ret = svc_create_xprt(serv, "tcp", PF_INET, nfs_callback_set_tcpport, SVC_SOCK_ANONYMOUS); @@ -131,18 +123,53 @@ int nfs_callback_up(void) goto out_err; #endif /* defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE) */ - nfs_callback_info.rqst = svc_prepare_thread(serv, &serv->sv_pools[0]); - if (IS_ERR(nfs_callback_info.rqst)) { - ret = PTR_ERR(nfs_callback_info.rqst); - nfs_callback_info.rqst = NULL; + return svc_prepare_thread(serv, &serv->sv_pools[0]); + +out_err: + if (ret == 0) + ret = -ENOMEM; + return ERR_PTR(ret); +} + +/* + * Bring up the callback thread if it is not already up. + */ +int nfs_callback_up(u32 minorversion, struct rpc_xprt *xprt) +{ + struct svc_serv *serv = NULL; + struct svc_rqst *rqstp; + int (*callback_svc)(void *vrqstp); + char svc_name[12]; + int ret = 0; + + mutex_lock(&nfs_callback_mutex); + if (nfs_callback_info.users++ || nfs_callback_info.task != NULL) + goto out; + serv = svc_create(&nfs4_callback_program, NFS4_CALLBACK_BUFSIZE, NULL); + if (!serv) { + ret = -ENOMEM; + goto out_err; + } + + if (!minorversion) { + rqstp = nfs4_callback_up(serv); + callback_svc = nfs4_callback_svc; + } else { + BUG(); /* for now */ + } + + if (IS_ERR(rqstp)) { + ret = PTR_ERR(rqstp); goto out_err; } svc_sock_update_bufs(serv); - nfs_callback_info.task = kthread_run(nfs_callback_svc, + sprintf(svc_name, "nfsv4.%u-svc", minorversion); + nfs_callback_info.rqst = rqstp; + nfs_callback_info.task = kthread_run(callback_svc, nfs_callback_info.rqst, - "nfsv4-svc"); + svc_name); if (IS_ERR(nfs_callback_info.task)) { ret = PTR_ERR(nfs_callback_info.task); svc_exit_thread(nfs_callback_info.rqst); diff --git a/fs/nfs/callback.h b/fs/nfs/callback.h index e110e286a262..9b907f408b8d 100644 --- a/fs/nfs/callback.h +++ b/fs/nfs/callback.h @@ -63,7 +63,7 @@ extern __be32 nfs4_callback_getattr(struct cb_getattrargs *args, struct cb_getat extern __be32 nfs4_callback_recall(struct cb_recallargs *args, void *dummy); #ifdef CONFIG_NFS_V4 -extern int nfs_callback_up(void); +extern int nfs_callback_up(u32 minorversion, struct rpc_xprt *xprt); extern void nfs_callback_down(void); #else #define nfs_callback_up() (0) diff --git a/fs/nfs/client.c b/fs/nfs/client.c index d9657d455730..df2b40d1883d 100644 --- a/fs/nfs/client.c +++ b/fs/nfs/client.c @@ -47,9 +47,6 @@ #include "internal.h" #include "fscache.h" -static int nfs4_init_callback(struct nfs_client *); -static void nfs4_destroy_callback(struct nfs_client *); - #define NFSDBG_FACILITY NFSDBG_CLIENT static DEFINE_SPINLOCK(nfs_client_lock); @@ -124,9 +121,6 @@ static struct nfs_client *nfs_alloc_client(const struct nfs_client_initdata *cl_ clp->rpc_ops = cl_init->rpc_ops; - if (nfs4_init_callback(clp) < 0) - goto error_2; - atomic_set(&clp->cl_count, 1); clp->cl_cons_state = NFS_CS_INITING; @@ -136,7 +130,7 @@ static struct nfs_client *nfs_alloc_client(const struct nfs_client_initdata *cl_ if (cl_init->hostname) { clp->cl_hostname = kstrdup(cl_init->hostname, GFP_KERNEL); if (!clp->cl_hostname) - goto error_3; + goto error_cleanup; } INIT_LIST_HEAD(&clp->cl_superblocks); @@ -161,9 +155,7 @@ static struct nfs_client *nfs_alloc_client(const struct nfs_client_initdata *cl_ return clp; -error_3: - nfs4_destroy_callback(clp); -error_2: +error_cleanup: kfree(clp); error_0: return NULL; @@ -207,6 +199,8 @@ static void nfs4_clear_client_minor_version(struct nfs_client *clp) clp->cl_call_sync = _nfs4_call_sync; #endif /* CONFIG_NFS_V4_1 */ + + nfs4_destroy_callback(clp); } /* @@ -225,8 +219,6 @@ static void nfs_free_client(struct nfs_client *clp) if (!IS_ERR(clp->cl_rpcclient)) rpc_shutdown_client(clp->cl_rpcclient); - nfs4_destroy_callback(clp); - if (clp->cl_machine_cred != NULL) put_rpccred(clp->cl_machine_cred); @@ -1104,7 +1096,8 @@ static int nfs4_init_callback(struct nfs_client *clp) int error; if (clp->rpc_ops->version == 4) { - error = nfs_callback_up(); + error = nfs_callback_up(clp->cl_minorversion, + clp->cl_rpcclient->cl_xprt); if (error < 0) { dprintk("%s: failed to start callback. Error = %d\n", __func__, error); @@ -1139,7 +1132,7 @@ static int nfs4_init_client_minor_version(struct nfs_client *clp) } #endif /* CONFIG_NFS_V4_1 */ - return 0; + return nfs4_init_callback(clp); } /* -- cgit v1.2.3 From 01c3f05228ce7fc19baa103e4e4bf6c1b5062a53 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Wed, 17 Jun 2009 13:22:58 -0700 Subject: NFSv4: Fix the 'nolock' option regression NFSv4 should just ignore the 'nolock' option. It is an NFSv2/v3 thing... This fixes the Oops in http://bugzilla.kernel.org/show_bug.cgi?id=13330 Signed-off-by: Trond Myklebust --- fs/nfs/super.c | 7 +++++++ 1 file changed, 7 insertions(+) (limited to 'fs') diff --git a/fs/nfs/super.c b/fs/nfs/super.c index 26127b69a275..98b47d17740a 100644 --- a/fs/nfs/super.c +++ b/fs/nfs/super.c @@ -2240,6 +2240,11 @@ static void nfs4_fill_super(struct super_block *sb) nfs_initialise_sb(sb); } +static void nfs4_validate_mount_flags(struct nfs_parsed_mount_data *args) +{ + args->flags &= ~(NFS_MOUNT_NONLM|NFS_MOUNT_NOACL|NFS_MOUNT_VER3); +} + /* * Validate NFSv4 mount options */ @@ -2336,6 +2341,8 @@ static int nfs4_validate_mount_data(void *options, nfs_validate_transport_protocol(args); + nfs4_validate_mount_flags(args); + if (args->auth_flavor_len > 1) goto out_inval_auth; -- cgit v1.2.3 From d5122201a7f90b2aa73092f158b84d1d74f1134d Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Wed, 17 Jun 2009 13:22:58 -0700 Subject: NFSv4: Move error handling out of the delegation generic code The NFSv4 delegation recovery code is required by the protocol to handle more errors. Rather than add NFSv4.0 specific errors into 'generic' delegation code, we should move the error handling into the NFSv4 layer. Signed-off-by: Trond Myklebust --- fs/nfs/delegation.c | 18 +++--------------- fs/nfs/nfs4proc.c | 15 +++++++++++++-- 2 files changed, 16 insertions(+), 17 deletions(-) (limited to 'fs') diff --git a/fs/nfs/delegation.c b/fs/nfs/delegation.c index 968225a88015..d4f669f0683e 100644 --- a/fs/nfs/delegation.c +++ b/fs/nfs/delegation.c @@ -68,7 +68,7 @@ static int nfs_delegation_claim_locks(struct nfs_open_context *ctx, struct nfs4_ { struct inode *inode = state->inode; struct file_lock *fl; - int status; + int status = 0; for (fl = inode->i_flock; fl != NULL; fl = fl->fl_next) { if (!(fl->fl_flags & (FL_POSIX|FL_FLOCK))) @@ -76,21 +76,9 @@ static int nfs_delegation_claim_locks(struct nfs_open_context *ctx, struct nfs4_ if (nfs_file_open_context(fl->fl_file) != ctx) continue; status = nfs4_lock_delegation_recall(state, fl); - if (status >= 0) - continue; - switch (status) { - default: - printk(KERN_ERR "%s: unhandled error %d.\n", - __func__, status); - case -NFS4ERR_EXPIRED: - /* kill_proc(fl->fl_pid, SIGLOST, 1); */ - case -NFS4ERR_STALE_CLIENTID: - nfs4_schedule_state_recovery(NFS_SERVER(inode)->nfs_client); - goto out_err; - } + if (status < 0) + break; } - return 0; -out_err: return status; } diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 4674f8092da8..d326193c9785 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -3630,8 +3630,19 @@ int nfs4_lock_delegation_recall(struct nfs4_state *state, struct file_lock *fl) goto out; do { err = _nfs4_do_setlk(state, F_SETLK, fl, 0); - if (err != -NFS4ERR_DELAY) - break; + switch (err) { + default: + printk(KERN_ERR "%s: unhandled error %d.\n", + __func__, err); + case 0: + goto out; + case -NFS4ERR_EXPIRED: + case -NFS4ERR_STALE_CLIENTID: + nfs4_schedule_state_recovery(server->nfs_client); + goto out; + case -NFS4ERR_DELAY: + break; + } err = nfs4_handle_exception(server, err, &exception); } while (exception.retry); out: -- cgit v1.2.3 From 965b5d679146c9f69bc0325388bb9ed357863c4f Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Wed, 17 Jun 2009 13:22:59 -0700 Subject: NFSv4: Handle more errors when recovering open file and locking state It is possible for servers to return NFS4ERR_BAD_STATEID when the state management code is recovering locks or is reclaiming state when returning a delegation. Ensure that we handle that case. While we're at it, add in handlers for NFS4ERR_STALE, NFS4ERR_ADMIN_REVOKED, NFS4ERR_OPENMODE, NFS4ERR_DENIED and NFS4ERR_STALE_STATEID, since the protocol appears to allow for them too. Also handle ENOMEM... Finally, rather than add new NFSv4.0-specific errors and error handling into the generic delegation code, move that open file and locking state error handling into the NFSv4 layer. Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 44 ++++++++++++++++++++++++++++++++++++++------ fs/nfs/nfs4state.c | 28 ++++++++++++++++++---------- 2 files changed, 56 insertions(+), 16 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index d326193c9785..55314e721632 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -804,16 +804,30 @@ int nfs4_open_delegation_recall(struct nfs_open_context *ctx, struct nfs4_state err = _nfs4_open_delegation_recall(ctx, state, stateid); switch (err) { case 0: - return err; + case -ENOENT: + case -ESTALE: + goto out; case -NFS4ERR_STALE_CLIENTID: case -NFS4ERR_STALE_STATEID: case -NFS4ERR_EXPIRED: /* Don't recall a delegation if it was lost */ nfs4_schedule_state_recovery(server->nfs_client); - return err; + goto out; + case -ERESTARTSYS: + /* + * The show must go on: exit, but mark the + * stateid as needing recovery. + */ + case -NFS4ERR_ADMIN_REVOKED: + case -NFS4ERR_BAD_STATEID: + nfs4_state_mark_reclaim_nograce(server->nfs_client, state); + case -ENOMEM: + err = 0; + goto out; } err = nfs4_handle_exception(server, err, &exception); } while (exception.retry); +out: return err; } @@ -3487,8 +3501,6 @@ static int _nfs4_do_setlk(struct nfs4_state *state, int cmd, struct file_lock *f ret = nfs4_wait_for_completion_rpc_task(task); if (ret == 0) { ret = data->rpc_status; - if (ret == -NFS4ERR_DENIED) - ret = -EAGAIN; } else data->cancelled = 1; rpc_put_task(task); @@ -3576,9 +3588,11 @@ static int nfs4_proc_setlk(struct nfs4_state *state, int cmd, struct file_lock * int err; do { + err = _nfs4_proc_setlk(state, cmd, request); + if (err == -NFS4ERR_DENIED) + err = -EAGAIN; err = nfs4_handle_exception(NFS_SERVER(state->inode), - _nfs4_proc_setlk(state, cmd, request), - &exception); + err, &exception); } while (exception.retry); return err; } @@ -3635,11 +3649,29 @@ int nfs4_lock_delegation_recall(struct nfs4_state *state, struct file_lock *fl) printk(KERN_ERR "%s: unhandled error %d.\n", __func__, err); case 0: + case -ESTALE: goto out; case -NFS4ERR_EXPIRED: case -NFS4ERR_STALE_CLIENTID: + case -NFS4ERR_STALE_STATEID: nfs4_schedule_state_recovery(server->nfs_client); goto out; + case -ERESTARTSYS: + /* + * The show must go on: exit, but mark the + * stateid as needing recovery. + */ + case -NFS4ERR_ADMIN_REVOKED: + case -NFS4ERR_BAD_STATEID: + case -NFS4ERR_OPENMODE: + nfs4_state_mark_reclaim_nograce(server->nfs_client, state); + err = 0; + goto out; + case -ENOMEM: + case -NFS4ERR_DENIED: + /* kill_proc(fl->fl_pid, SIGLOST, 1); */ + err = 0; + goto out; case -NFS4ERR_DELAY: break; } diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c index 0298e909559f..6e094e95cf0f 100644 --- a/fs/nfs/nfs4state.c +++ b/fs/nfs/nfs4state.c @@ -854,25 +854,29 @@ static int nfs4_reclaim_locks(struct nfs4_state *state, const struct nfs4_state_ if (nfs_file_open_context(fl->fl_file)->state != state) continue; status = ops->recover_lock(state, fl); - if (status >= 0) - continue; switch (status) { + case 0: + break; + case -ESTALE: + case -NFS4ERR_ADMIN_REVOKED: + case -NFS4ERR_STALE_STATEID: + case -NFS4ERR_BAD_STATEID: + case -NFS4ERR_EXPIRED: + case -NFS4ERR_NO_GRACE: + case -NFS4ERR_STALE_CLIENTID: + goto out; default: printk(KERN_ERR "%s: unhandled error %d. Zeroing state\n", __func__, status); - case -NFS4ERR_EXPIRED: - case -NFS4ERR_NO_GRACE: + case -ENOMEM: + case -NFS4ERR_DENIED: case -NFS4ERR_RECLAIM_BAD: case -NFS4ERR_RECLAIM_CONFLICT: /* kill_proc(fl->fl_pid, SIGLOST, 1); */ - break; - case -NFS4ERR_STALE_CLIENTID: - goto out_err; + status = 0; } } - up_write(&nfsi->rwsem); - return 0; -out_err: +out: up_write(&nfsi->rwsem); return status; } @@ -918,6 +922,7 @@ restart: printk(KERN_ERR "%s: unhandled error %d. Zeroing state\n", __func__, status); case -ENOENT: + case -ENOMEM: case -ESTALE: /* * Open state on this file cannot be recovered @@ -928,6 +933,9 @@ restart: /* Mark the file as being 'closed' */ state->state = 0; break; + case -NFS4ERR_ADMIN_REVOKED: + case -NFS4ERR_STALE_STATEID: + case -NFS4ERR_BAD_STATEID: case -NFS4ERR_RECLAIM_BAD: case -NFS4ERR_RECLAIM_CONFLICT: nfs4_state_mark_reclaim_nograce(sp->so_client, state); -- cgit v1.2.3 From 3f09df70e3a33590ae5a97b8a15486d3711c7065 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Wed, 17 Jun 2009 13:23:00 -0700 Subject: NFS: Ensure we always hold the BKL when dereferencing inode->i_flock Signed-off-by: Trond Myklebust --- fs/nfs/delegation.c | 16 ++++++++++++++-- fs/nfs/nfs4state.c | 9 +++++++++ 2 files changed, 23 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/nfs/delegation.c b/fs/nfs/delegation.c index d4f669f0683e..af05b918cb5b 100644 --- a/fs/nfs/delegation.c +++ b/fs/nfs/delegation.c @@ -70,15 +70,24 @@ static int nfs_delegation_claim_locks(struct nfs_open_context *ctx, struct nfs4_ struct file_lock *fl; int status = 0; + if (inode->i_flock == NULL) + goto out; + + /* Protect inode->i_flock using the BKL */ + lock_kernel(); for (fl = inode->i_flock; fl != NULL; fl = fl->fl_next) { if (!(fl->fl_flags & (FL_POSIX|FL_FLOCK))) continue; if (nfs_file_open_context(fl->fl_file) != ctx) continue; + unlock_kernel(); status = nfs4_lock_delegation_recall(state, fl); if (status < 0) - break; + goto out; + lock_kernel(); } + unlock_kernel(); +out: return status; } @@ -256,7 +265,10 @@ static int __nfs_inode_return_delegation(struct inode *inode, struct nfs_delegat struct nfs_inode *nfsi = NFS_I(inode); nfs_msync_inode(inode); - /* Guard against new delegated open calls */ + /* + * Guard against new delegated open/lock/unlock calls and against + * state recovery + */ down_write(&nfsi->rwsem); nfs_delegation_claim_opens(inode, &delegation->stateid); up_write(&nfsi->rwsem); diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c index 6e094e95cf0f..0b2a19bfe5a0 100644 --- a/fs/nfs/nfs4state.c +++ b/fs/nfs/nfs4state.c @@ -847,12 +847,19 @@ static int nfs4_reclaim_locks(struct nfs4_state *state, const struct nfs4_state_ struct file_lock *fl; int status = 0; + if (inode->i_flock == NULL) + return 0; + + /* Guard against delegation returns and new lock/unlock calls */ down_write(&nfsi->rwsem); + /* Protect inode->i_flock using the BKL */ + lock_kernel(); for (fl = inode->i_flock; fl != NULL; fl = fl->fl_next) { if (!(fl->fl_flags & (FL_POSIX|FL_FLOCK))) continue; if (nfs_file_open_context(fl->fl_file)->state != state) continue; + unlock_kernel(); status = ops->recover_lock(state, fl); switch (status) { case 0: @@ -875,7 +882,9 @@ static int nfs4_reclaim_locks(struct nfs4_state *state, const struct nfs4_state_ /* kill_proc(fl->fl_pid, SIGLOST, 1); */ status = 0; } + lock_kernel(); } + unlock_kernel(); out: up_write(&nfsi->rwsem); return status; -- cgit v1.2.3 From 5cd973c44a92f4abf8f7084c804089b3eaa7b4bf Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Wed, 17 Jun 2009 13:23:01 -0700 Subject: NFSv4/NLM: Push file locking BKL dependencies down into the NLM layer Signed-off-by: Trond Myklebust --- fs/lockd/clntproc.c | 2 ++ fs/nfs/file.c | 6 ------ 2 files changed, 2 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/lockd/clntproc.c b/fs/lockd/clntproc.c index dd7957064a8c..273e229353f3 100644 --- a/fs/lockd/clntproc.c +++ b/fs/lockd/clntproc.c @@ -165,6 +165,7 @@ int nlmclnt_proc(struct nlm_host *host, int cmd, struct file_lock *fl) /* Set up the argument struct */ nlmclnt_setlockargs(call, fl); + lock_kernel(); if (IS_SETLK(cmd) || IS_SETLKW(cmd)) { if (fl->fl_type != F_UNLCK) { call->a_args.block = IS_SETLKW(cmd) ? 1 : 0; @@ -178,6 +179,7 @@ int nlmclnt_proc(struct nlm_host *host, int cmd, struct file_lock *fl) fl->fl_ops->fl_release_private(fl); fl->fl_ops = NULL; + unlock_kernel(); dprintk("lockd: clnt proc returns %d\n", status); return status; diff --git a/fs/nfs/file.c b/fs/nfs/file.c index ec7e27d00bc6..52ed56762583 100644 --- a/fs/nfs/file.c +++ b/fs/nfs/file.c @@ -592,7 +592,6 @@ static int do_getlk(struct file *filp, int cmd, struct file_lock *fl) struct inode *inode = filp->f_mapping->host; int status = 0; - lock_kernel(); /* Try local locking first */ posix_test_lock(filp, fl); if (fl->fl_type != F_UNLCK) { @@ -608,7 +607,6 @@ static int do_getlk(struct file *filp, int cmd, struct file_lock *fl) status = NFS_PROTO(inode)->lock(filp, cmd, fl); out: - unlock_kernel(); return status; out_noconflict: fl->fl_type = F_UNLCK; @@ -650,13 +648,11 @@ static int do_unlk(struct file *filp, int cmd, struct file_lock *fl) * If we're signalled while cleaning up locks on process exit, we * still need to complete the unlock. */ - lock_kernel(); /* Use local locking if mounted with "-onolock" */ if (!(NFS_SERVER(inode)->flags & NFS_MOUNT_NONLM)) status = NFS_PROTO(inode)->lock(filp, cmd, fl); else status = do_vfs_lock(filp, fl); - unlock_kernel(); return status; } @@ -673,13 +669,11 @@ static int do_setlk(struct file *filp, int cmd, struct file_lock *fl) if (status != 0) goto out; - lock_kernel(); /* Use local locking if mounted with "-onolock" */ if (!(NFS_SERVER(inode)->flags & NFS_MOUNT_NONLM)) status = NFS_PROTO(inode)->lock(filp, cmd, fl); else status = do_vfs_lock(filp, fl); - unlock_kernel(); if (status < 0) goto out; /* -- cgit v1.2.3 From a43cde94feded0f65fce36330614691c650ae8fe Mon Sep 17 00:00:00 2001 From: Ricardo Labiaga Date: Wed, 1 Apr 2009 09:23:08 -0400 Subject: nfs41: Implement NFSv4.1 callback service process. nfs41_callback_up() initializes the necessary queues and creates the new nfs41_callback_svc thread. This thread executes the callback service which waits for requests to arrive on the svc_serv->sv_cb_list. NFS41_BC_MIN_CALLBACKS is set to 1 because we expect callbacks to not cause substantial latency. The actual processing of the callback will be implemented as a separate patch. There is only one NFSv4.1 callback service. The first caller of nfs4_callback_up() creates the service, subsequent callers increment a reference count on the service. The service is destroyed when the last caller invokes nfs_callback_down(). The transport needs to hold a reference to the callback service in order to invoke it during callback processing. Currently this reference is only obtained when the service is first created. This is incorrect, since subsequent registrations for other transports will leave the xprt->serv pointer uninitialized, leading to an oops when a callback arrives on the "unreferenced" transport. This patch fixes the problem by ensuring that a reference to the service is saved in xprt->serv, either because the service is created by this invocation to nfs4_callback_up() or by a prior invocation. Signed-off-by: Ricardo Labiaga Signed-off-by: Benny Halevy [nfs41: Add a reference to svc_serv during callback service bring up] Signed-off-by: Ricardo Labiaga Signed-off-by: Benny Halevy [Type check arguments of nfs_callback_up] Signed-off-by: Ricardo Labiaga Signed-off-by: Benny Halevy [nfs41: save svc_serv in nfs_callback_info] Signed-off-by: Benny Halevy [Removal of ugly #ifdefs] [nfs41: Update to removal of ugly #ifdefs] Signed-off-by: Ricardo Labiaga Signed-off-by: Benny Halevy --- fs/nfs/callback.c | 112 ++++++++++++++++++++++++++++++++++++++++++++++++++++-- fs/nfs/callback.h | 7 ++++ 2 files changed, 115 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/nfs/callback.c b/fs/nfs/callback.c index 4b1313eda6f5..470928898063 100644 --- a/fs/nfs/callback.c +++ b/fs/nfs/callback.c @@ -17,6 +17,9 @@ #include #include #include +#if defined(CONFIG_NFS_V4_1) +#include +#endif #include @@ -28,6 +31,7 @@ struct nfs_callback_data { unsigned int users; + struct svc_serv *serv; struct svc_rqst *rqst; struct task_struct *task; }; @@ -131,6 +135,99 @@ out_err: return ERR_PTR(ret); } +#if defined(CONFIG_NFS_V4_1) +/* + * The callback service for NFSv4.1 callbacks + */ +static int +nfs41_callback_svc(void *vrqstp) +{ + struct svc_rqst *rqstp = vrqstp; + struct svc_serv *serv = rqstp->rq_server; + struct rpc_rqst *req; + int error; + DEFINE_WAIT(wq); + + set_freezable(); + + /* + * FIXME: do we really need to run this under the BKL? If so, please + * add a comment about what it's intended to protect. + */ + lock_kernel(); + while (!kthread_should_stop()) { + prepare_to_wait(&serv->sv_cb_waitq, &wq, TASK_INTERRUPTIBLE); + spin_lock_bh(&serv->sv_cb_lock); + if (!list_empty(&serv->sv_cb_list)) { + req = list_first_entry(&serv->sv_cb_list, + struct rpc_rqst, rq_bc_list); + list_del(&req->rq_bc_list); + spin_unlock_bh(&serv->sv_cb_lock); + dprintk("Invoking bc_svc_process()\n"); + error = bc_svc_process(serv, req, rqstp); + dprintk("bc_svc_process() returned w/ error code= %d\n", + error); + } else { + spin_unlock_bh(&serv->sv_cb_lock); + schedule(); + } + finish_wait(&serv->sv_cb_waitq, &wq); + } + unlock_kernel(); + nfs_callback_info.task = NULL; + svc_exit_thread(rqstp); + return 0; +} + +/* + * Bring up the NFSv4.1 callback service + */ +struct svc_rqst * +nfs41_callback_up(struct svc_serv *serv, struct rpc_xprt *xprt) +{ + /* + * Save the svc_serv in the transport so that it can + * be referenced when the session backchannel is initialized + */ + xprt->bc_serv = serv; + + INIT_LIST_HEAD(&serv->sv_cb_list); + spin_lock_init(&serv->sv_cb_lock); + init_waitqueue_head(&serv->sv_cb_waitq); + return svc_prepare_thread(serv, &serv->sv_pools[0]); +} + +static inline int nfs_minorversion_callback_svc_setup(u32 minorversion, + struct svc_serv *serv, struct rpc_xprt *xprt, + struct svc_rqst **rqstpp, int (**callback_svc)(void *vrqstp)) +{ + if (minorversion) { + *rqstpp = nfs41_callback_up(serv, xprt); + *callback_svc = nfs41_callback_svc; + } + return minorversion; +} + +static inline void nfs_callback_bc_serv(u32 minorversion, struct rpc_xprt *xprt, + struct nfs_callback_data *cb_info) +{ + if (minorversion) + xprt->bc_serv = cb_info->serv; +} +#else +static inline int nfs_minorversion_callback_svc_setup(u32 minorversion, + struct svc_serv *serv, struct rpc_xprt *xprt, + struct svc_rqst **rqstpp, int (**callback_svc)(void *vrqstp)) +{ + return 0; +} + +static inline void nfs_callback_bc_serv(u32 minorversion, struct rpc_xprt *xprt, + struct nfs_callback_data *cb_info) +{ +} +#endif /* CONFIG_NFS_V4_1 */ + /* * Bring up the callback thread if it is not already up. */ @@ -141,21 +238,25 @@ int nfs_callback_up(u32 minorversion, struct rpc_xprt *xprt) int (*callback_svc)(void *vrqstp); char svc_name[12]; int ret = 0; + int minorversion_setup; mutex_lock(&nfs_callback_mutex); - if (nfs_callback_info.users++ || nfs_callback_info.task != NULL) + if (nfs_callback_info.users++ || nfs_callback_info.task != NULL) { + nfs_callback_bc_serv(minorversion, xprt, &nfs_callback_info); goto out; + } serv = svc_create(&nfs4_callback_program, NFS4_CALLBACK_BUFSIZE, NULL); if (!serv) { ret = -ENOMEM; goto out_err; } - if (!minorversion) { + minorversion_setup = nfs_minorversion_callback_svc_setup(minorversion, + serv, xprt, &rqstp, &callback_svc); + if (!minorversion_setup) { + /* v4.0 callback setup */ rqstp = nfs4_callback_up(serv); callback_svc = nfs4_callback_svc; - } else { - BUG(); /* for now */ } if (IS_ERR(rqstp)) { @@ -166,6 +267,7 @@ int nfs_callback_up(u32 minorversion, struct rpc_xprt *xprt) svc_sock_update_bufs(serv); sprintf(svc_name, "nfsv4.%u-svc", minorversion); + nfs_callback_info.serv = serv; nfs_callback_info.rqst = rqstp; nfs_callback_info.task = kthread_run(callback_svc, nfs_callback_info.rqst, @@ -173,6 +275,7 @@ int nfs_callback_up(u32 minorversion, struct rpc_xprt *xprt) if (IS_ERR(nfs_callback_info.task)) { ret = PTR_ERR(nfs_callback_info.task); svc_exit_thread(nfs_callback_info.rqst); + nfs_callback_info.serv = NULL; nfs_callback_info.rqst = NULL; nfs_callback_info.task = NULL; goto out_err; @@ -205,6 +308,7 @@ void nfs_callback_down(void) if (nfs_callback_info.users == 0 && nfs_callback_info.task != NULL) { kthread_stop(nfs_callback_info.task); svc_exit_thread(nfs_callback_info.rqst); + nfs_callback_info.serv = NULL; nfs_callback_info.rqst = NULL; nfs_callback_info.task = NULL; } diff --git a/fs/nfs/callback.h b/fs/nfs/callback.h index 9b907f408b8d..29123b5604f2 100644 --- a/fs/nfs/callback.h +++ b/fs/nfs/callback.h @@ -70,6 +70,13 @@ extern void nfs_callback_down(void); #define nfs_callback_down() do {} while(0) #endif +/* + * nfs41: Callbacks are expected to not cause substantial latency, + * so we limit their concurrency to 1 by setting up the maximum number + * of slots for the backchannel. + */ +#define NFS41_BC_MIN_CALLBACKS 1 + extern unsigned int nfs_callback_set_tcpport; extern unsigned short nfs_callback_tcpport; extern unsigned short nfs_callback_tcpport6; -- cgit v1.2.3 From 8f975242352e92898dc641ebff0d24808f39848a Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Wed, 1 Apr 2009 09:23:11 -0400 Subject: nfs41: create a svc_xprt for nfs41 callback thread and use for incoming callbacks Signed-off-by: Benny Halevy --- fs/nfs/callback.c | 17 ++++++++++++++++- net/sunrpc/svc.c | 1 + 2 files changed, 17 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfs/callback.c b/fs/nfs/callback.c index 470928898063..37815f3216aa 100644 --- a/fs/nfs/callback.c +++ b/fs/nfs/callback.c @@ -185,16 +185,31 @@ nfs41_callback_svc(void *vrqstp) struct svc_rqst * nfs41_callback_up(struct svc_serv *serv, struct rpc_xprt *xprt) { + struct svc_xprt *bc_xprt; + struct svc_rqst *rqstp = ERR_PTR(-ENOMEM); + + dprintk("--> %s\n", __func__); + /* Create a svc_sock for the service */ + bc_xprt = svc_sock_create(serv, xprt->prot); + if (!bc_xprt) + goto out; + /* * Save the svc_serv in the transport so that it can * be referenced when the session backchannel is initialized */ + serv->bc_xprt = bc_xprt; xprt->bc_serv = serv; INIT_LIST_HEAD(&serv->sv_cb_list); spin_lock_init(&serv->sv_cb_lock); init_waitqueue_head(&serv->sv_cb_waitq); - return svc_prepare_thread(serv, &serv->sv_pools[0]); + rqstp = svc_prepare_thread(serv, &serv->sv_pools[0]); + if (IS_ERR(rqstp)) + svc_sock_destroy(bc_xprt); +out: + dprintk("--> %s return %p\n", __func__, rqstp); + return rqstp; } static inline int nfs_minorversion_callback_svc_setup(u32 minorversion, diff --git a/net/sunrpc/svc.c b/net/sunrpc/svc.c index b35048fabe22..6b90ce439c00 100644 --- a/net/sunrpc/svc.c +++ b/net/sunrpc/svc.c @@ -1258,6 +1258,7 @@ bc_svc_process(struct svc_serv *serv, struct rpc_rqst *req, int error; /* Build the svc_rqst used by the common processing routine */ + rqstp->rq_xprt = serv->bc_xprt; rqstp->rq_xid = req->rq_xid; rqstp->rq_prot = req->rq_xprt->prot; rqstp->rq_server = serv; -- cgit v1.2.3 From e82dc22dac6525a2f365a1d53c0483252d4aa38e Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Wed, 1 Apr 2009 09:23:14 -0400 Subject: nfs41: Allow NFSv4 and NFSv4.1 callback services to coexist Tracks the nfs_callback_info for both versions, enabling the callback service for v4 and v4.1 to run concurrently and be stopped independently of each other. Signed-off-by: Ricardo Labiaga Signed-off-by: Benny Halevy --- fs/nfs/callback.c | 50 ++++++++++++++++++++++++-------------------------- fs/nfs/callback.h | 7 ++----- fs/nfs/client.c | 2 +- 3 files changed, 27 insertions(+), 32 deletions(-) (limited to 'fs') diff --git a/fs/nfs/callback.c b/fs/nfs/callback.c index 37815f3216aa..e69b8f61189e 100644 --- a/fs/nfs/callback.c +++ b/fs/nfs/callback.c @@ -36,7 +36,7 @@ struct nfs_callback_data { struct task_struct *task; }; -static struct nfs_callback_data nfs_callback_info; +static struct nfs_callback_data nfs_callback_info[NFS4_MAX_MINOR_VERSION + 1]; static DEFINE_MUTEX(nfs_callback_mutex); static struct svc_program nfs4_callback_program; @@ -60,7 +60,7 @@ module_param_call(callback_tcpport, param_set_port, param_get_int, &nfs_callback_set_tcpport, 0644); /* - * This is the callback kernel thread. + * This is the NFSv4 callback kernel thread. */ static int nfs4_callback_svc(void *vrqstp) @@ -174,8 +174,6 @@ nfs41_callback_svc(void *vrqstp) finish_wait(&serv->sv_cb_waitq, &wq); } unlock_kernel(); - nfs_callback_info.task = NULL; - svc_exit_thread(rqstp); return 0; } @@ -251,13 +249,14 @@ int nfs_callback_up(u32 minorversion, struct rpc_xprt *xprt) struct svc_serv *serv = NULL; struct svc_rqst *rqstp; int (*callback_svc)(void *vrqstp); + struct nfs_callback_data *cb_info = &nfs_callback_info[minorversion]; char svc_name[12]; int ret = 0; int minorversion_setup; mutex_lock(&nfs_callback_mutex); - if (nfs_callback_info.users++ || nfs_callback_info.task != NULL) { - nfs_callback_bc_serv(minorversion, xprt, &nfs_callback_info); + if (cb_info->users++ || cb_info->task != NULL) { + nfs_callback_bc_serv(minorversion, xprt, cb_info); goto out; } serv = svc_create(&nfs4_callback_program, NFS4_CALLBACK_BUFSIZE, NULL); @@ -282,17 +281,14 @@ int nfs_callback_up(u32 minorversion, struct rpc_xprt *xprt) svc_sock_update_bufs(serv); sprintf(svc_name, "nfsv4.%u-svc", minorversion); - nfs_callback_info.serv = serv; - nfs_callback_info.rqst = rqstp; - nfs_callback_info.task = kthread_run(callback_svc, - nfs_callback_info.rqst, - svc_name); - if (IS_ERR(nfs_callback_info.task)) { - ret = PTR_ERR(nfs_callback_info.task); - svc_exit_thread(nfs_callback_info.rqst); - nfs_callback_info.serv = NULL; - nfs_callback_info.rqst = NULL; - nfs_callback_info.task = NULL; + cb_info->serv = serv; + cb_info->rqst = rqstp; + cb_info->task = kthread_run(callback_svc, cb_info->rqst, svc_name); + if (IS_ERR(cb_info->task)) { + ret = PTR_ERR(cb_info->task); + svc_exit_thread(cb_info->rqst); + cb_info->rqst = NULL; + cb_info->task = NULL; goto out_err; } out: @@ -309,23 +305,25 @@ out: out_err: dprintk("NFS: Couldn't create callback socket or server thread; " "err = %d\n", ret); - nfs_callback_info.users--; + cb_info->users--; goto out; } /* * Kill the callback thread if it's no longer being used. */ -void nfs_callback_down(void) +void nfs_callback_down(int minorversion) { + struct nfs_callback_data *cb_info = &nfs_callback_info[minorversion]; + mutex_lock(&nfs_callback_mutex); - nfs_callback_info.users--; - if (nfs_callback_info.users == 0 && nfs_callback_info.task != NULL) { - kthread_stop(nfs_callback_info.task); - svc_exit_thread(nfs_callback_info.rqst); - nfs_callback_info.serv = NULL; - nfs_callback_info.rqst = NULL; - nfs_callback_info.task = NULL; + cb_info->users--; + if (cb_info->users == 0 && cb_info->task != NULL) { + kthread_stop(cb_info->task); + svc_exit_thread(cb_info->rqst); + cb_info->serv = NULL; + cb_info->rqst = NULL; + cb_info->task = NULL; } mutex_unlock(&nfs_callback_mutex); } diff --git a/fs/nfs/callback.h b/fs/nfs/callback.h index 29123b5604f2..88d2e1135b5a 100644 --- a/fs/nfs/callback.h +++ b/fs/nfs/callback.h @@ -64,11 +64,8 @@ extern __be32 nfs4_callback_recall(struct cb_recallargs *args, void *dummy); #ifdef CONFIG_NFS_V4 extern int nfs_callback_up(u32 minorversion, struct rpc_xprt *xprt); -extern void nfs_callback_down(void); -#else -#define nfs_callback_up() (0) -#define nfs_callback_down() do {} while(0) -#endif +extern void nfs_callback_down(int minorversion); +#endif /* CONFIG_NFS_V4 */ /* * nfs41: Callbacks are expected to not cause substantial latency, diff --git a/fs/nfs/client.c b/fs/nfs/client.c index df2b40d1883d..ac6575c4da31 100644 --- a/fs/nfs/client.c +++ b/fs/nfs/client.c @@ -181,7 +181,7 @@ static void nfs4_destroy_callback(struct nfs_client *clp) { #ifdef CONFIG_NFS_V4 if (__test_and_clear_bit(NFS_CS_CALLBACK, &clp->cl_res_state)) - nfs_callback_down(); + nfs_callback_down(clp->cl_minorversion); #endif /* CONFIG_NFS_V4 */ } -- cgit v1.2.3 From 0b5b7ae0a853c91015bb3b1729166ca65f693322 Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Wed, 1 Apr 2009 09:23:15 -0400 Subject: nfs41: Setup the backchannel The NFS v4.1 callback service has already been setup, and rpc_xprt->serv points to the svc_serv structure describing it. Invoke the xprt_setup_backchannel() initialization to pre- allocate the necessary backchannel structures. Signed-off-by: Ricardo Labiaga Signed-off-by: Benny Halevy [nfs41: change nfs4_put_session(nfs4_session**) to nfs4_destroy_session(nfs_session*)] Signed-off-by: Alexandros Batsakis Signed-off-by: Benny Halevy [moved xprt_setup_backchannel from nfs4_init_session to nfs4_init_backchannel] Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy --- fs/nfs/client.c | 9 +++++++++ 1 file changed, 9 insertions(+) (limited to 'fs') diff --git a/fs/nfs/client.c b/fs/nfs/client.c index ac6575c4da31..4f75ec593be8 100644 --- a/fs/nfs/client.c +++ b/fs/nfs/client.c @@ -37,6 +37,7 @@ #include #include #include +#include #include @@ -1096,6 +1097,14 @@ static int nfs4_init_callback(struct nfs_client *clp) int error; if (clp->rpc_ops->version == 4) { + if (nfs4_has_session(clp)) { + error = xprt_setup_backchannel( + clp->cl_rpcclient->cl_xprt, + NFS41_BC_MIN_CALLBACKS); + if (error < 0) + return error; + } + error = nfs_callback_up(clp->cl_minorversion, clp->cl_rpcclient->cl_xprt); if (error < 0) { -- cgit v1.2.3 From 0f91421e8ea986c990c01bf091ceff1d59c165ff Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Wed, 1 Apr 2009 09:23:16 -0400 Subject: nfs41: Client indicates presence of NFSv4.1 callback channel. Set the SESSION4_BACK_CHAN flag to indicate the client supports a backchannel. Signed-off-by: Ricardo Labiaga Signed-off-by: Benny Halevy --- fs/nfs/nfs4proc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index f1f087b483ce..4213c8814997 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -4598,7 +4598,7 @@ static int _nfs4_proc_create_session(struct nfs_client *clp) int status; nfs4_init_channel_attrs(&args); - args.flags = (SESSION4_PERSIST); + args.flags = (SESSION4_PERSIST | SESSION4_BACK_CHAN); status = rpc_call_sync(session->clp->cl_rpcclient, &msg, 0); -- cgit v1.2.3 From 5a0ffe544c54f62be99751e369f4d0f44bd5ee19 Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Wed, 1 Apr 2009 09:23:18 -0400 Subject: nfs41: Release backchannel resources associated with session Frees the preallocated backchannel resources that are associated with this session when the session is destroyed. A backchannel is currently created once per session. Destroy the backchannel only when the session is destroyed. Signed-off-by: Ricardo Labiaga Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy --- fs/nfs/nfs4proc.c | 6 ++++++ fs/nfs/super.c | 1 - 2 files changed, 6 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 4213c8814997..5da939d577d5 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -49,6 +49,7 @@ #include #include #include +#include #include "nfs4_fs.h" #include "delegation.h" @@ -4481,6 +4482,11 @@ struct nfs4_session *nfs4_alloc_session(struct nfs_client *clp) void nfs4_destroy_session(struct nfs4_session *session) { + nfs4_proc_destroy_session(session); + dprintk("%s Destroy backchannel for xprt %p\n", + __func__, session->clp->cl_rpcclient->cl_xprt); + xprt_destroy_backchannel(session->clp->cl_rpcclient->cl_xprt, + NFS41_BC_MIN_CALLBACKS); nfs4_destroy_slot_table(session); kfree(session); } diff --git a/fs/nfs/super.c b/fs/nfs/super.c index 73db6f8c145a..69804a8245f6 100644 --- a/fs/nfs/super.c +++ b/fs/nfs/super.c @@ -2488,7 +2488,6 @@ static void nfs4_kill_super(struct super_block *sb) dprintk("--> %s\n", __func__); nfs_super_return_all_delegations(sb); kill_anon_super(sb); - nfs4_renewd_prepare_shutdown(server); nfs_fscache_release_super_cookie(sb); nfs_free_server(server); -- cgit v1.2.3 From b8f2ef84b079ceb22b42d6d353609db7eb8efa93 Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Wed, 1 Apr 2009 09:23:19 -0400 Subject: nfs41: store minorversion in cb_compound_hdr_arg Signed-off-by: Benny Halevy --- fs/nfs/callback.h | 1 + fs/nfs/callback_xdr.c | 12 +++++++----- 2 files changed, 8 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/nfs/callback.h b/fs/nfs/callback.h index 88d2e1135b5a..51b15c2d3a56 100644 --- a/fs/nfs/callback.h +++ b/fs/nfs/callback.h @@ -26,6 +26,7 @@ enum nfs4_callback_opnum { struct cb_compound_hdr_arg { unsigned int taglen; const char *tag; + unsigned int minorversion; unsigned int callback_ident; unsigned nops; }; diff --git a/fs/nfs/callback_xdr.c b/fs/nfs/callback_xdr.c index dd0ef34b5845..91f6f74ffea7 100644 --- a/fs/nfs/callback_xdr.c +++ b/fs/nfs/callback_xdr.c @@ -132,7 +132,6 @@ static __be32 decode_stateid(struct xdr_stream *xdr, nfs4_stateid *stateid) static __be32 decode_compound_hdr_arg(struct xdr_stream *xdr, struct cb_compound_hdr_arg *hdr) { __be32 *p; - unsigned int minor_version; __be32 status; status = decode_string(xdr, &hdr->taglen, &hdr->tag); @@ -147,15 +146,18 @@ static __be32 decode_compound_hdr_arg(struct xdr_stream *xdr, struct cb_compound p = read_buf(xdr, 12); if (unlikely(p == NULL)) return htonl(NFS4ERR_RESOURCE); - minor_version = ntohl(*p++); + hdr->minorversion = ntohl(*p++); /* Check minor version is zero. */ - if (minor_version != 0) { - printk(KERN_WARNING "%s: NFSv4 server callback with illegal minor version %u!\n", - __func__, minor_version); + if (hdr->minorversion != 0) { + printk(KERN_WARNING "%s: NFSv4 server callback with " + "illegal minor version %u!\n", + __func__, hdr->minorversion); return htonl(NFS4ERR_MINOR_VERS_MISMATCH); } hdr->callback_ident = ntohl(*p++); hdr->nops = ntohl(*p); + dprintk("%s: minorversion %d nops %d\n", __func__, + hdr->minorversion, hdr->nops); return 0; } -- cgit v1.2.3 From 48a9e2d2289ed4b5053b3986d99709e9b07a0923 Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Wed, 1 Apr 2009 09:23:20 -0400 Subject: nfs41: decode minorversion 1 cb_compound header decode cb_compound header conforming to http://tools.ietf.org/html/draft-ietf-nfsv4-minorversion1-26 Get rid of cb_compound_hdr_arg.callback_ident callback_ident is not used anywhere so we shouldn't waste any memory to store it. Signed-off-by: Benny Halevy [nfs41: no need to break read_buf in decode_compound_hdr_arg] See http://linux-nfs.org/pipermail/pnfs/2009-June/007844.html Signed-off-by: Benny Halevy --- fs/nfs/callback.h | 1 - fs/nfs/callback_xdr.c | 7 ++++--- 2 files changed, 4 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/nfs/callback.h b/fs/nfs/callback.h index 51b15c2d3a56..a8ffa9d6ea2f 100644 --- a/fs/nfs/callback.h +++ b/fs/nfs/callback.h @@ -27,7 +27,6 @@ struct cb_compound_hdr_arg { unsigned int taglen; const char *tag; unsigned int minorversion; - unsigned int callback_ident; unsigned nops; }; diff --git a/fs/nfs/callback_xdr.c b/fs/nfs/callback_xdr.c index 91f6f74ffea7..f6cc79b1d1f3 100644 --- a/fs/nfs/callback_xdr.c +++ b/fs/nfs/callback_xdr.c @@ -147,14 +147,15 @@ static __be32 decode_compound_hdr_arg(struct xdr_stream *xdr, struct cb_compound if (unlikely(p == NULL)) return htonl(NFS4ERR_RESOURCE); hdr->minorversion = ntohl(*p++); - /* Check minor version is zero. */ - if (hdr->minorversion != 0) { + /* Check minor version is zero or one. */ + if (hdr->minorversion <= 1) { + p++; /* skip callback_ident */ + } else { printk(KERN_WARNING "%s: NFSv4 server callback with " "illegal minor version %u!\n", __func__, hdr->minorversion); return htonl(NFS4ERR_MINOR_VERS_MISMATCH); } - hdr->callback_ident = ntohl(*p++); hdr->nops = ntohl(*p); dprintk("%s: minorversion %d nops %d\n", __func__, hdr->minorversion, hdr->nops); -- cgit v1.2.3 From 45377b94edea18f53dd3ba4d46d94de4bb7c00b5 Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Wed, 1 Apr 2009 09:23:21 -0400 Subject: nfs41: callback numbers definitions Signed-off-by: Benny Halevy --- fs/nfs/callback.h | 10 ++++++++++ 1 file changed, 10 insertions(+) (limited to 'fs') diff --git a/fs/nfs/callback.h b/fs/nfs/callback.h index a8ffa9d6ea2f..7ba42b0d16c9 100644 --- a/fs/nfs/callback.h +++ b/fs/nfs/callback.h @@ -20,6 +20,16 @@ enum nfs4_callback_procnum { enum nfs4_callback_opnum { OP_CB_GETATTR = 3, OP_CB_RECALL = 4, +/* Callback operations new to NFSv4.1 */ + OP_CB_LAYOUTRECALL = 5, + OP_CB_NOTIFY = 6, + OP_CB_PUSH_DELEG = 7, + OP_CB_RECALL_ANY = 8, + OP_CB_RECALLABLE_OBJ_AVAIL = 9, + OP_CB_RECALL_SLOT = 10, + OP_CB_SEQUENCE = 11, + OP_CB_WANTS_CANCELLED = 12, + OP_CB_NOTIFY_LOCK = 13, OP_CB_ILLEGAL = 10044, }; -- cgit v1.2.3 From 34bc47c941a074f91c2455b4b08503d02c74b878 Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Wed, 1 Apr 2009 09:23:22 -0400 Subject: nfs41: consider minorversion in callback_xdr:process_op Note that this patch changes the nfsv4.0 behavior also when CONFIG_NFS_V4_1 is not defined where NFS4ERR_MINOR_VERS_MISMATCH will be returned if the client received a CB_COMPOUND with minorversion != 0. Previously, it would have returned NFS4ERR_OP_ILLEGAL for CB_SEQUENCE. (or if the server is broken and sent OP_CB_GETATTR or OP_CB_RECALL with minorversion!=0, they would have been processed normally. Signed-off-by: Benny Halevy [nfs41: refactor op preprocessing out of process_op] See http://linux-nfs.org/pipermail/pnfs/2009-June/007845.html [nfs41: define CB_NOTIFY_DEVICEID as not supported] Signed-off-by: Benny Halevy --- fs/nfs/callback.h | 1 + fs/nfs/callback_xdr.c | 85 ++++++++++++++++++++++++++++++++++++++++++--------- 2 files changed, 72 insertions(+), 14 deletions(-) (limited to 'fs') diff --git a/fs/nfs/callback.h b/fs/nfs/callback.h index 7ba42b0d16c9..80fd8a82964f 100644 --- a/fs/nfs/callback.h +++ b/fs/nfs/callback.h @@ -30,6 +30,7 @@ enum nfs4_callback_opnum { OP_CB_SEQUENCE = 11, OP_CB_WANTS_CANCELLED = 12, OP_CB_NOTIFY_LOCK = 13, + OP_CB_NOTIFY_DEVICEID = 14, OP_CB_ILLEGAL = 10044, }; diff --git a/fs/nfs/callback_xdr.c b/fs/nfs/callback_xdr.c index f6cc79b1d1f3..41c5be1c1be5 100644 --- a/fs/nfs/callback_xdr.c +++ b/fs/nfs/callback_xdr.c @@ -356,31 +356,87 @@ out: return status; } -static __be32 process_op(struct svc_rqst *rqstp, +#if defined(CONFIG_NFS_V4_1) + +static __be32 +preprocess_nfs41_op(int nop, unsigned int op_nr, struct callback_op **op) +{ + switch (op_nr) { + case OP_CB_GETATTR: + case OP_CB_RECALL: + *op = &callback_ops[op_nr]; + break; + + case OP_CB_LAYOUTRECALL: + case OP_CB_NOTIFY_DEVICEID: + case OP_CB_NOTIFY: + case OP_CB_PUSH_DELEG: + case OP_CB_RECALL_ANY: + case OP_CB_RECALLABLE_OBJ_AVAIL: + case OP_CB_RECALL_SLOT: + case OP_CB_SEQUENCE: + case OP_CB_WANTS_CANCELLED: + case OP_CB_NOTIFY_LOCK: + return htonl(NFS4ERR_NOTSUPP); + + default: + return htonl(NFS4ERR_OP_ILLEGAL); + } + + return htonl(NFS_OK); +} + +#else /* CONFIG_NFS_V4_1 */ + +static __be32 +preprocess_nfs41_op(int nop, unsigned int op_nr, struct callback_op **op) +{ + return htonl(NFS4ERR_MINOR_VERS_MISMATCH); +} + +#endif /* CONFIG_NFS_V4_1 */ + +static __be32 +preprocess_nfs4_op(unsigned int op_nr, struct callback_op **op) +{ + switch (op_nr) { + case OP_CB_GETATTR: + case OP_CB_RECALL: + *op = &callback_ops[op_nr]; + break; + default: + return htonl(NFS4ERR_OP_ILLEGAL); + } + + return htonl(NFS_OK); +} + +static __be32 process_op(uint32_t minorversion, int nop, + struct svc_rqst *rqstp, struct xdr_stream *xdr_in, void *argp, struct xdr_stream *xdr_out, void *resp) { struct callback_op *op = &callback_ops[0]; unsigned int op_nr = OP_CB_ILLEGAL; - __be32 status = 0; + __be32 status; long maxlen; __be32 res; dprintk("%s: start\n", __func__); status = decode_op_hdr(xdr_in, &op_nr); - if (likely(status == 0)) { - switch (op_nr) { - case OP_CB_GETATTR: - case OP_CB_RECALL: - op = &callback_ops[op_nr]; - break; - default: - op_nr = OP_CB_ILLEGAL; - op = &callback_ops[0]; - status = htonl(NFS4ERR_OP_ILLEGAL); - } + if (unlikely(status)) { + status = htonl(NFS4ERR_OP_ILLEGAL); + goto out; } + dprintk("%s: minorversion=%d nop=%d op_nr=%u\n", + __func__, minorversion, nop, op_nr); + + status = minorversion ? preprocess_nfs41_op(nop, op_nr, &op) : + preprocess_nfs4_op(op_nr, &op); + if (status == htonl(NFS4ERR_OP_ILLEGAL)) + op_nr = OP_CB_ILLEGAL; +out: maxlen = xdr_out->end - xdr_out->p; if (maxlen > 0 && maxlen < PAGE_SIZE) { if (likely(status == 0 && op->decode_args != NULL)) @@ -428,7 +484,8 @@ static __be32 nfs4_callback_compound(struct svc_rqst *rqstp, void *argp, void *r return rpc_system_err; while (status == 0 && nops != hdr_arg.nops) { - status = process_op(rqstp, &xdr_in, argp, &xdr_out, resp); + status = process_op(hdr_arg.minorversion, nops, + rqstp, &xdr_in, argp, &xdr_out, resp); nops++; } -- cgit v1.2.3 From 2d9b9ec344b19b7b65c732b7000114df57684140 Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Wed, 1 Apr 2009 09:23:24 -0400 Subject: nfs41: cb_sequence protocol level data structures Signed-off-by: Benny Halevy --- fs/nfs/callback.h | 35 +++++++++++++++++++++++++++++++++++ 1 file changed, 35 insertions(+) (limited to 'fs') diff --git a/fs/nfs/callback.h b/fs/nfs/callback.h index 80fd8a82964f..2b933ce064dd 100644 --- a/fs/nfs/callback.h +++ b/fs/nfs/callback.h @@ -70,6 +70,41 @@ struct cb_recallargs { uint32_t truncate; }; +#if defined(CONFIG_NFS_V4_1) + +struct referring_call { + uint32_t rc_sequenceid; + uint32_t rc_slotid; +}; + +struct referring_call_list { + struct nfs4_sessionid rcl_sessionid; + uint32_t rcl_nrefcalls; + struct referring_call *rcl_refcalls; +}; + +struct cb_sequenceargs { + struct sockaddr_in *csa_addr; + struct nfs4_sessionid csa_sessionid; + uint32_t csa_sequenceid; + uint32_t csa_slotid; + uint32_t csa_highestslotid; + uint32_t csa_cachethis; + uint32_t csa_nrclists; + struct referring_call_list *csa_rclists; +}; + +struct cb_sequenceres { + uint32_t csr_status; + struct nfs4_sessionid csr_sessionid; + uint32_t csr_sequenceid; + uint32_t csr_slotid; + uint32_t csr_highestslotid; + uint32_t csr_target_highestslotid; +}; + +#endif /* CONFIG_NFS_V4_1 */ + extern __be32 nfs4_callback_getattr(struct cb_getattrargs *args, struct cb_getattrres *res); extern __be32 nfs4_callback_recall(struct cb_recallargs *args, void *dummy); -- cgit v1.2.3 From d49433e1e3bb144a5752ce2a8ba1139dc519df1a Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Wed, 1 Apr 2009 09:23:25 -0400 Subject: nfs41: cb_sequence proc implementation Currently, just free up any referring calls information. Signed-off-by: Benny Halevy [nfs41: fix csr_{,target}highestslotid] Signed-off-by: Benny Halevy --- fs/nfs/callback.h | 4 ++++ fs/nfs/callback_proc.c | 28 ++++++++++++++++++++++++++++ 2 files changed, 32 insertions(+) (limited to 'fs') diff --git a/fs/nfs/callback.h b/fs/nfs/callback.h index 2b933ce064dd..576f51f4dc52 100644 --- a/fs/nfs/callback.h +++ b/fs/nfs/callback.h @@ -103,6 +103,9 @@ struct cb_sequenceres { uint32_t csr_target_highestslotid; }; +extern unsigned nfs4_callback_sequence(struct cb_sequenceargs *args, + struct cb_sequenceres *res); + #endif /* CONFIG_NFS_V4_1 */ extern __be32 nfs4_callback_getattr(struct cb_getattrargs *args, struct cb_getattrres *res); @@ -119,6 +122,7 @@ extern void nfs_callback_down(int minorversion); * of slots for the backchannel. */ #define NFS41_BC_MIN_CALLBACKS 1 +#define NFS41_BC_MAX_CALLBACKS 1 extern unsigned int nfs_callback_set_tcpport; extern unsigned short nfs_callback_tcpport; diff --git a/fs/nfs/callback_proc.c b/fs/nfs/callback_proc.c index f7e83e23cf9f..f731bbe7ac6a 100644 --- a/fs/nfs/callback_proc.c +++ b/fs/nfs/callback_proc.c @@ -101,3 +101,31 @@ out: dprintk("%s: exit with status = %d\n", __func__, ntohl(res)); return res; } + +#if defined(CONFIG_NFS_V4_1) + +/* FIXME: validate args->cbs_{sequence,slot}id */ +/* FIXME: referring calls should be processed */ +unsigned nfs4_callback_sequence(struct cb_sequenceargs *args, + struct cb_sequenceres *res) +{ + int i; + unsigned status = 0; + + for (i = 0; i < args->csa_nrclists; i++) + kfree(args->csa_rclists[i].rcl_refcalls); + kfree(args->csa_rclists); + + memcpy(&res->csr_sessionid, &args->csa_sessionid, + sizeof(res->csr_sessionid)); + res->csr_sequenceid = args->csa_sequenceid; + res->csr_slotid = args->csa_slotid; + res->csr_highestslotid = NFS41_BC_MAX_CALLBACKS - 1; + res->csr_target_highestslotid = NFS41_BC_MAX_CALLBACKS - 1; + + dprintk("%s: exit with status = %d\n", __func__, ntohl(status)); + res->csr_status = status; + return status; +} + +#endif /* CONFIG_NFS_V4_1 */ -- cgit v1.2.3 From 4aece6a19cf7f474f15eb861ba74db4479884ce3 Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Wed, 1 Apr 2009 09:23:26 -0400 Subject: nfs41: cb_sequence xdr implementation [nfs41: get rid of READMEM and COPYMEM for callback_xdr.c] Signed-off-by: Benny Halevy [nfs41: get rid of READ64 in callback_xdr.c] See http://linux-nfs.org/pipermail/pnfs/2009-June/007846.html Signed-off-by: Benny Halevy --- fs/nfs/callback_xdr.c | 172 +++++++++++++++++++++++++++++++++++++++++++++++++- 1 file changed, 170 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/nfs/callback_xdr.c b/fs/nfs/callback_xdr.c index 41c5be1c1be5..56a3cc510107 100644 --- a/fs/nfs/callback_xdr.c +++ b/fs/nfs/callback_xdr.c @@ -20,6 +20,11 @@ 2 + 2 + 3 + 3) #define CB_OP_RECALL_RES_MAXSZ (CB_OP_HDR_RES_MAXSZ) +#if defined(CONFIG_NFS_V4_1) +#define CB_OP_SEQUENCE_RES_MAXSZ (CB_OP_HDR_RES_MAXSZ + \ + 4 + 1 + 3) +#endif /* CONFIG_NFS_V4_1 */ + #define NFSDBG_FACILITY NFSDBG_CALLBACK typedef __be32 (*callback_process_op_t)(void *, void *); @@ -207,6 +212,122 @@ out: return status; } +#if defined(CONFIG_NFS_V4_1) + +static unsigned decode_sessionid(struct xdr_stream *xdr, + struct nfs4_sessionid *sid) +{ + uint32_t *p; + int len = NFS4_MAX_SESSIONID_LEN; + + p = read_buf(xdr, len); + if (unlikely(p == NULL)) + return htonl(NFS4ERR_RESOURCE);; + + memcpy(sid->data, p, len); + return 0; +} + +static unsigned decode_rc_list(struct xdr_stream *xdr, + struct referring_call_list *rc_list) +{ + uint32_t *p; + int i; + unsigned status; + + status = decode_sessionid(xdr, &rc_list->rcl_sessionid); + if (status) + goto out; + + status = htonl(NFS4ERR_RESOURCE); + p = read_buf(xdr, sizeof(uint32_t)); + if (unlikely(p == NULL)) + goto out; + + rc_list->rcl_nrefcalls = ntohl(*p++); + if (rc_list->rcl_nrefcalls) { + p = read_buf(xdr, + rc_list->rcl_nrefcalls * 2 * sizeof(uint32_t)); + if (unlikely(p == NULL)) + goto out; + rc_list->rcl_refcalls = kmalloc(rc_list->rcl_nrefcalls * + sizeof(*rc_list->rcl_refcalls), + GFP_KERNEL); + if (unlikely(rc_list->rcl_refcalls == NULL)) + goto out; + for (i = 0; i < rc_list->rcl_nrefcalls; i++) { + rc_list->rcl_refcalls[i].rc_sequenceid = ntohl(*p++); + rc_list->rcl_refcalls[i].rc_slotid = ntohl(*p++); + } + } + status = 0; + +out: + return status; +} + +static unsigned decode_cb_sequence_args(struct svc_rqst *rqstp, + struct xdr_stream *xdr, + struct cb_sequenceargs *args) +{ + uint32_t *p; + int i; + unsigned status; + + status = decode_sessionid(xdr, &args->csa_sessionid); + if (status) + goto out; + + status = htonl(NFS4ERR_RESOURCE); + p = read_buf(xdr, 5 * sizeof(uint32_t)); + if (unlikely(p == NULL)) + goto out; + + args->csa_addr = svc_addr_in(rqstp); + args->csa_sequenceid = ntohl(*p++); + args->csa_slotid = ntohl(*p++); + args->csa_highestslotid = ntohl(*p++); + args->csa_cachethis = ntohl(*p++); + args->csa_nrclists = ntohl(*p++); + args->csa_rclists = NULL; + if (args->csa_nrclists) { + args->csa_rclists = kmalloc(args->csa_nrclists * + sizeof(*args->csa_rclists), + GFP_KERNEL); + if (unlikely(args->csa_rclists == NULL)) + goto out; + + for (i = 0; i < args->csa_nrclists; i++) { + status = decode_rc_list(xdr, &args->csa_rclists[i]); + if (status) + goto out_free; + } + } + status = 0; + + dprintk("%s: sessionid %x:%x:%x:%x sequenceid %u slotid %u " + "highestslotid %u cachethis %d nrclists %u\n", + __func__, + ((u32 *)&args->csa_sessionid)[0], + ((u32 *)&args->csa_sessionid)[1], + ((u32 *)&args->csa_sessionid)[2], + ((u32 *)&args->csa_sessionid)[3], + args->csa_sequenceid, args->csa_slotid, + args->csa_highestslotid, args->csa_cachethis, + args->csa_nrclists); +out: + dprintk("%s: exit with status = %d\n", __func__, ntohl(status)); + return status; + +out_free: + for (i = 0; i < args->csa_nrclists; i++) + kfree(args->csa_rclists[i].rcl_refcalls); + kfree(args->csa_rclists); + goto out; +} + +#endif /* CONFIG_NFS_V4_1 */ + static __be32 encode_string(struct xdr_stream *xdr, unsigned int len, const char *str) { __be32 *p; @@ -358,12 +479,52 @@ out: #if defined(CONFIG_NFS_V4_1) +static unsigned encode_sessionid(struct xdr_stream *xdr, + const struct nfs4_sessionid *sid) +{ + uint32_t *p; + int len = NFS4_MAX_SESSIONID_LEN; + + p = xdr_reserve_space(xdr, len); + if (unlikely(p == NULL)) + return htonl(NFS4ERR_RESOURCE); + + memcpy(p, sid, len); + return 0; +} + +static unsigned encode_cb_sequence_res(struct svc_rqst *rqstp, + struct xdr_stream *xdr, + const struct cb_sequenceres *res) +{ + uint32_t *p; + unsigned status = res->csr_status; + + if (unlikely(status != 0)) + goto out; + + encode_sessionid(xdr, &res->csr_sessionid); + + p = xdr_reserve_space(xdr, 4 * sizeof(uint32_t)); + if (unlikely(p == NULL)) + return htonl(NFS4ERR_RESOURCE); + + *p++ = htonl(res->csr_sequenceid); + *p++ = htonl(res->csr_slotid); + *p++ = htonl(res->csr_highestslotid); + *p++ = htonl(res->csr_target_highestslotid); +out: + dprintk("%s: exit with status = %d\n", __func__, ntohl(status)); + return status; +} + static __be32 preprocess_nfs41_op(int nop, unsigned int op_nr, struct callback_op **op) { switch (op_nr) { case OP_CB_GETATTR: case OP_CB_RECALL: + case OP_CB_SEQUENCE: *op = &callback_ops[op_nr]; break; @@ -374,7 +535,6 @@ preprocess_nfs41_op(int nop, unsigned int op_nr, struct callback_op **op) case OP_CB_RECALL_ANY: case OP_CB_RECALLABLE_OBJ_AVAIL: case OP_CB_RECALL_SLOT: - case OP_CB_SEQUENCE: case OP_CB_WANTS_CANCELLED: case OP_CB_NOTIFY_LOCK: return htonl(NFS4ERR_NOTSUPP); @@ -512,7 +672,15 @@ static struct callback_op callback_ops[] = { .process_op = (callback_process_op_t)nfs4_callback_recall, .decode_args = (callback_decode_arg_t)decode_recall_args, .res_maxsize = CB_OP_RECALL_RES_MAXSZ, - } + }, +#if defined(CONFIG_NFS_V4_1) + [OP_CB_SEQUENCE] = { + .process_op = (callback_process_op_t)nfs4_callback_sequence, + .decode_args = (callback_decode_arg_t)decode_cb_sequence_args, + .encode_res = (callback_encode_res_t)encode_cb_sequence_res, + .res_maxsize = CB_OP_SEQUENCE_RES_MAXSZ, + }, +#endif /* CONFIG_NFS_V4_1 */ }; /* -- cgit v1.2.3 From 281fe15dc1d6ad46992f18b7a6644269ec5f7138 Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Wed, 1 Apr 2009 09:23:27 -0400 Subject: nfs41: verify CB_SEQUENCE position in callback compound CB_SEQUENCE must appear first in the callback compound RPC. If it is not the first operation NFS4ERR_SEQUENCE_POS must be returned. If the first operation ni the CB_COMPOUND is not CB_SEQUENCE then NFS4ERR_OP_NOT_IN_SESSION must be returned. Signed-off-by: Ricardo Labiaga Signed-off-by: Benny Halevy [nfs41: refactor op preprocessing out of process_op] Signed-off-by: Benny Halevy --- fs/nfs/callback_xdr.c | 8 ++++++++ 1 file changed, 8 insertions(+) (limited to 'fs') diff --git a/fs/nfs/callback_xdr.c b/fs/nfs/callback_xdr.c index 56a3cc510107..537f21da6e5f 100644 --- a/fs/nfs/callback_xdr.c +++ b/fs/nfs/callback_xdr.c @@ -521,6 +521,14 @@ out: static __be32 preprocess_nfs41_op(int nop, unsigned int op_nr, struct callback_op **op) { + if (op_nr == OP_CB_SEQUENCE) { + if (nop != 0) + return htonl(NFS4ERR_SEQUENCE_POS); + } else { + if (nop == 0) + return htonl(NFS4ERR_OP_NOT_IN_SESSION); + } + switch (op_nr) { case OP_CB_GETATTR: case OP_CB_RECALL: -- cgit v1.2.3 From 65fc64e547c794764a441e16e95bb76c0e256bd7 Mon Sep 17 00:00:00 2001 From: Ricardo Labiaga Date: Wed, 1 Apr 2009 09:23:30 -0400 Subject: nfs41: Backchannel: update cb_sequence args and results Change the type of cs_addr and csr_status to 'struct sockaddr' and '__be32' since the cb_sequence processing function will use existing functionality that expects these types. Signed-off-by: Ricardo Labiaga Signed-off-by: Benny Halevy --- fs/nfs/callback.h | 4 ++-- fs/nfs/callback_xdr.c | 2 +- 2 files changed, 3 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/nfs/callback.h b/fs/nfs/callback.h index 576f51f4dc52..07baa8254ca1 100644 --- a/fs/nfs/callback.h +++ b/fs/nfs/callback.h @@ -84,7 +84,7 @@ struct referring_call_list { }; struct cb_sequenceargs { - struct sockaddr_in *csa_addr; + struct sockaddr *csa_addr; struct nfs4_sessionid csa_sessionid; uint32_t csa_sequenceid; uint32_t csa_slotid; @@ -95,7 +95,7 @@ struct cb_sequenceargs { }; struct cb_sequenceres { - uint32_t csr_status; + __be32 csr_status; struct nfs4_sessionid csr_sessionid; uint32_t csr_sequenceid; uint32_t csr_slotid; diff --git a/fs/nfs/callback_xdr.c b/fs/nfs/callback_xdr.c index 537f21da6e5f..e5a2dac5f715 100644 --- a/fs/nfs/callback_xdr.c +++ b/fs/nfs/callback_xdr.c @@ -283,7 +283,7 @@ static unsigned decode_cb_sequence_args(struct svc_rqst *rqstp, if (unlikely(p == NULL)) goto out; - args->csa_addr = svc_addr_in(rqstp); + args->csa_addr = svc_addr(rqstp); args->csa_sequenceid = ntohl(*p++); args->csa_slotid = ntohl(*p++); args->csa_highestslotid = ntohl(*p++); -- cgit v1.2.3 From b73dafa7ac94ca8387f65c57cb63a7ffac91bf2c Mon Sep 17 00:00:00 2001 From: Ricardo Labiaga Date: Wed, 1 Apr 2009 09:23:31 -0400 Subject: nfs41: Backchannel: Refactor nfs4_reset_slot_table() Generalize nfs4_reset_slot_table() so it can be used to reset the backchannel slot table in addition to the forechannel slot table. Signed-off-by: Ricardo Labiaga Signed-off-by: Benny Halevy --- fs/nfs/nfs4proc.c | 38 +++++++++++++++++++++++++++----------- 1 file changed, 27 insertions(+), 11 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 5da939d577d5..a7b1d6c228cd 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -4370,19 +4370,21 @@ int nfs4_proc_get_lease_time(struct nfs_client *clp, struct nfs_fsinfo *fsinfo) return status; } -/* Reset a slot table */ -static int nfs4_reset_slot_table(struct nfs4_session *session) +/* + * Reset a slot table + */ +static int nfs4_reset_slot_table(struct nfs4_slot_table *tbl, int max_slots, + int old_max_slots, int ivalue) { - struct nfs4_slot_table *tbl = &session->fc_slot_table; - int i, max_slots = session->fc_attrs.max_reqs; - int old_max_slots = session->fc_slot_table.max_slots; + int i; int ret = 0; - dprintk("--> %s: max_reqs=%u, tbl %p\n", __func__, - session->fc_attrs.max_reqs, tbl); + dprintk("--> %s: max_reqs=%u, tbl %p\n", __func__, max_slots, tbl); - /* Until we have dynamic slot table adjustment, insist - * upon the same slot table size */ + /* + * Until we have dynamic slot table adjustment, insist + * upon the same slot table size + */ if (max_slots != old_max_slots) { dprintk("%s reset slot table does't match old\n", __func__); @@ -4391,7 +4393,7 @@ static int nfs4_reset_slot_table(struct nfs4_session *session) } spin_lock(&tbl->slot_tbl_lock); for (i = 0; i < max_slots; ++i) - tbl->slots[i].seq_nr = 1; + tbl->slots[i].seq_nr = ivalue; tbl->highest_used_slotid = -1; spin_unlock(&tbl->slot_tbl_lock); dprintk("%s: tbl=%p slots=%p max_slots=%d\n", __func__, @@ -4401,6 +4403,20 @@ out: return ret; } +/* + * Reset the forechannel and backchannel slot tables + */ +static int nfs4_reset_slot_tables(struct nfs4_session *session) +{ + int status; + + status = nfs4_reset_slot_table(&session->fc_slot_table, + session->fc_attrs.max_reqs, + session->fc_slot_table.max_slots, + 1); + return status; +} + /* * Initialize slot table */ @@ -4639,7 +4655,7 @@ int nfs4_proc_create_session(struct nfs_client *clp, int reset) /* Init or reset the fore channel */ if (reset) - status = nfs4_reset_slot_table(session); + status = nfs4_reset_slot_tables(session); else status = nfs4_init_slot_table(session); dprintk("fore channel slot table initialization returned %d\n", status); -- cgit v1.2.3 From 050047ce71bcf60867d2af7a9dc965a9c6f15cb8 Mon Sep 17 00:00:00 2001 From: Ricardo Labiaga Date: Wed, 1 Apr 2009 09:23:32 -0400 Subject: nfs41: Backchannel: Refactor nfs4_init_slot_table() Generalize nfs4_init_slot_table() so it can be used to initialize the backchannel slot table in addition to the forechannel slot table. Signed-off-by: Ricardo Labiaga Signed-off-by: Benny Halevy --- fs/nfs/nfs4proc.c | 26 +++++++++++++++++++------- 1 file changed, 19 insertions(+), 7 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index a7b1d6c228cd..c3019ad85893 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -4420,23 +4420,22 @@ static int nfs4_reset_slot_tables(struct nfs4_session *session) /* * Initialize slot table */ -static int nfs4_init_slot_table(struct nfs4_session *session) +static int nfs4_init_slot_table(struct nfs4_slot_table *tbl, + int max_slots, int ivalue) { - struct nfs4_slot_table *tbl = &session->fc_slot_table; - int i, max_slots = session->fc_attrs.max_reqs; + int i; struct nfs4_slot *slot; int ret = -ENOMEM; BUG_ON(max_slots > NFS4_MAX_SLOT_TABLE); - dprintk("--> %s: max_reqs=%u\n", __func__, - session->fc_attrs.max_reqs); + dprintk("--> %s: max_reqs=%u\n", __func__, max_slots); slot = kcalloc(max_slots, sizeof(struct nfs4_slot), GFP_KERNEL); if (!slot) goto out; for (i = 0; i < max_slots; ++i) - slot[i].seq_nr = 1; + slot[i].seq_nr = ivalue; ret = 0; spin_lock(&tbl->slot_tbl_lock); @@ -4456,11 +4455,24 @@ static int nfs4_init_slot_table(struct nfs4_session *session) out: dprintk("<-- %s: return %d\n", __func__, ret); return ret; + out_free: kfree(slot); goto out; } +/* + * Initialize the forechannel and backchannel tables + */ +static int nfs4_init_slot_tables(struct nfs4_session *session) +{ + int status; + + status = nfs4_init_slot_table(&session->fc_slot_table, + session->fc_attrs.max_reqs, 1); + return status; +} + /* Destroy the slot table */ static void nfs4_destroy_slot_table(struct nfs4_session *session) { @@ -4657,7 +4669,7 @@ int nfs4_proc_create_session(struct nfs_client *clp, int reset) if (reset) status = nfs4_reset_slot_tables(session); else - status = nfs4_init_slot_table(session); + status = nfs4_init_slot_tables(session); dprintk("fore channel slot table initialization returned %d\n", status); if (status) goto out; -- cgit v1.2.3 From f8625a6a4bb76207302be58453603d8e324df490 Mon Sep 17 00:00:00 2001 From: Ricardo Labiaga Date: Wed, 1 Apr 2009 09:23:33 -0400 Subject: nfs41: Backchannel: Add a backchannel slot table to the session Defines a new 'struct nfs4_slot_table' in the 'struct nfs4_session' for use by the backchannel. Initializes, resets, and destroys the backchannel slot table in the same manner the forechannel slot table is initialized, reset, and destroyed. The sequenceid for each slot in the backchannel slot table is initialized to 0, whereas the forechannel slotid's sequenceid is set to 1. Signed-off-by: Ricardo Labiaga Signed-off-by: Benny Halevy --- fs/nfs/nfs4proc.c | 48 +++++++++++++++++++++++++++++++++++------------ include/linux/nfs_fs_sb.h | 2 +- 2 files changed, 37 insertions(+), 13 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index c3019ad85893..57dabb8a048e 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -4414,9 +4414,30 @@ static int nfs4_reset_slot_tables(struct nfs4_session *session) session->fc_attrs.max_reqs, session->fc_slot_table.max_slots, 1); + if (status) + return status; + + status = nfs4_reset_slot_table(&session->bc_slot_table, + session->bc_attrs.max_reqs, + session->bc_slot_table.max_slots, + 0); return status; } +/* Destroy the slot table */ +static void nfs4_destroy_slot_tables(struct nfs4_session *session) +{ + if (session->fc_slot_table.slots != NULL) { + kfree(session->fc_slot_table.slots); + session->fc_slot_table.slots = NULL; + } + if (session->bc_slot_table.slots != NULL) { + kfree(session->bc_slot_table.slots); + session->bc_slot_table.slots = NULL; + } + return; +} + /* * Initialize slot table */ @@ -4470,17 +4491,15 @@ static int nfs4_init_slot_tables(struct nfs4_session *session) status = nfs4_init_slot_table(&session->fc_slot_table, session->fc_attrs.max_reqs, 1); - return status; -} + if (status) + return status; -/* Destroy the slot table */ -static void nfs4_destroy_slot_table(struct nfs4_session *session) -{ - if (session->fc_slot_table.slots == NULL) - return; - kfree(session->fc_slot_table.slots); - session->fc_slot_table.slots = NULL; - return; + status = nfs4_init_slot_table(&session->bc_slot_table, + session->bc_attrs.max_reqs, 0); + if (status) + nfs4_destroy_slot_tables(session); + + return status; } struct nfs4_session *nfs4_alloc_session(struct nfs_client *clp) @@ -4503,7 +4522,12 @@ struct nfs4_session *nfs4_alloc_session(struct nfs_client *clp) tbl = &session->fc_slot_table; spin_lock_init(&tbl->slot_tbl_lock); - rpc_init_wait_queue(&tbl->slot_tbl_waitq, "Slot table"); + rpc_init_wait_queue(&tbl->slot_tbl_waitq, "ForeChannel Slot table"); + + tbl = &session->bc_slot_table; + spin_lock_init(&tbl->slot_tbl_lock); + rpc_init_wait_queue(&tbl->slot_tbl_waitq, "BackChannel Slot table"); + session->clp = clp; return session; } @@ -4515,7 +4539,7 @@ void nfs4_destroy_session(struct nfs4_session *session) __func__, session->clp->cl_rpcclient->cl_xprt); xprt_destroy_backchannel(session->clp->cl_rpcclient->cl_xprt, NFS41_BC_MIN_CALLBACKS); - nfs4_destroy_slot_table(session); + nfs4_destroy_slot_tables(session); kfree(session); } diff --git a/include/linux/nfs_fs_sb.h b/include/linux/nfs_fs_sb.h index d0902ccec9ce..19fe15d12042 100644 --- a/include/linux/nfs_fs_sb.h +++ b/include/linux/nfs_fs_sb.h @@ -205,7 +205,7 @@ struct nfs4_session { struct nfs4_channel_attrs fc_attrs; struct nfs4_slot_table fc_slot_table; struct nfs4_channel_attrs bc_attrs; - /* back channel has one slot */ + struct nfs4_slot_table bc_slot_table; struct nfs_client *clp; }; -- cgit v1.2.3 From 963891ac43ecf9974d82f4c178752e11e007cf87 Mon Sep 17 00:00:00 2001 From: Ricardo Labiaga Date: Wed, 1 Apr 2009 09:23:34 -0400 Subject: nfs41: Backchannel: New find_client_with_session() Finds the 'struct nfs_client' that matches the server's address, major version number, and session ID. Signed-off-by: Ricardo Labiaga Signed-off-by: Benny Halevy --- fs/nfs/callback_proc.c | 36 ++++++++++++++++++++++++++++++++++++ 1 file changed, 36 insertions(+) (limited to 'fs') diff --git a/fs/nfs/callback_proc.c b/fs/nfs/callback_proc.c index f731bbe7ac6a..6b342e82ebb3 100644 --- a/fs/nfs/callback_proc.c +++ b/fs/nfs/callback_proc.c @@ -104,6 +104,42 @@ out: #if defined(CONFIG_NFS_V4_1) +/* + * Returns a pointer to a held 'struct nfs_client' that matches the server's + * address, major version number, and session ID. It is the caller's + * responsibility to release the returned reference. + * + * Returns NULL if there are no connections with sessions, or if no session + * matches the one of interest. + */ + static struct nfs_client *find_client_with_session( + const struct sockaddr *addr, u32 nfsversion, + struct nfs4_sessionid *sessionid) +{ + struct nfs_client *clp; + + clp = nfs_find_client(addr, 4); + if (clp == NULL) + return NULL; + + do { + struct nfs_client *prev = clp; + + if (clp->cl_session != NULL) { + if (memcmp(clp->cl_session->sess_id.data, + sessionid->data, + NFS4_MAX_SESSIONID_LEN) == 0) { + /* Returns a held reference to clp */ + return clp; + } + } + clp = nfs_find_client_next(prev); + nfs_put_client(prev); + } while (clp != NULL); + + return NULL; +} + /* FIXME: validate args->cbs_{sequence,slot}id */ /* FIXME: referring calls should be processed */ unsigned nfs4_callback_sequence(struct cb_sequenceargs *args, -- cgit v1.2.3 From 68f3f90133d56e0c38f04f991e662c2b21592b31 Mon Sep 17 00:00:00 2001 From: Ricardo Labiaga Date: Wed, 1 Apr 2009 09:23:35 -0400 Subject: nfs41: Backchannel: CB_SEQUENCE validation Validates the callback's sessionID, the slot number, and the sequence ID. Increments the slot's sequence. Detects replays, but simply prints a debug message (if debugging is enabled since we don't yet implement a duplicate request cache for the backchannel. This should not present a problem, since only idempotent callbacks are currently implemented. Signed-off-by: Ricardo Labiaga Signed-off-by: Benny Halevy [nfs41: Backchannel: Be more obvious about the return value] [nfs41: Backchannel: dprink in host order] Signed-off-by: Ricardo Labiaga Signed-off-by: Benny Halevy --- fs/nfs/callback_proc.c | 71 +++++++++++++++++++++++++++++++++++++++++++++++--- 1 file changed, 67 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/nfs/callback_proc.c b/fs/nfs/callback_proc.c index 6b342e82ebb3..b7da1f54da68 100644 --- a/fs/nfs/callback_proc.c +++ b/fs/nfs/callback_proc.c @@ -104,6 +104,57 @@ out: #if defined(CONFIG_NFS_V4_1) +/* + * Validate the sequenceID sent by the server. + * Return success if the sequenceID is one more than what we last saw on + * this slot, accounting for wraparound. Increments the slot's sequence. + * + * We don't yet implement a duplicate request cache, so at this time + * we will log replays, and process them as if we had not seen them before, + * but we don't bump the sequence in the slot. Not too worried about it, + * since we only currently implement idempotent callbacks anyway. + * + * We have a single slot backchannel at this time, so we don't bother + * checking the used_slots bit array on the table. The lower layer guarantees + * a single outstanding callback request at a time. + */ +static int +validate_seqid(struct nfs4_slot_table *tbl, u32 slotid, u32 seqid) +{ + struct nfs4_slot *slot; + + dprintk("%s enter. slotid %d seqid %d\n", + __func__, slotid, seqid); + + if (slotid > NFS41_BC_MAX_CALLBACKS) + return htonl(NFS4ERR_BADSLOT); + + slot = tbl->slots + slotid; + dprintk("%s slot table seqid: %d\n", __func__, slot->seq_nr); + + /* Normal */ + if (likely(seqid == slot->seq_nr + 1)) { + slot->seq_nr++; + return htonl(NFS4_OK); + } + + /* Replay */ + if (seqid == slot->seq_nr) { + dprintk("%s seqid %d is a replay - no DRC available\n", + __func__, seqid); + return htonl(NFS4_OK); + } + + /* Wraparound */ + if (seqid == 1 && (slot->seq_nr + 1) == 0) { + slot->seq_nr = 1; + return htonl(NFS4_OK); + } + + /* Misordered request */ + return htonl(NFS4ERR_SEQ_MISORDERED); +} + /* * Returns a pointer to a held 'struct nfs_client' that matches the server's * address, major version number, and session ID. It is the caller's @@ -140,18 +191,27 @@ out: return NULL; } -/* FIXME: validate args->cbs_{sequence,slot}id */ /* FIXME: referring calls should be processed */ unsigned nfs4_callback_sequence(struct cb_sequenceargs *args, struct cb_sequenceres *res) { - int i; - unsigned status = 0; + struct nfs_client *clp; + int i, status; for (i = 0; i < args->csa_nrclists; i++) kfree(args->csa_rclists[i].rcl_refcalls); kfree(args->csa_rclists); + status = htonl(NFS4ERR_BADSESSION); + clp = find_client_with_session(args->csa_addr, 4, &args->csa_sessionid); + if (clp == NULL) + goto out; + + status = validate_seqid(&clp->cl_session->bc_slot_table, + args->csa_slotid, args->csa_sequenceid); + if (status) + goto out_putclient; + memcpy(&res->csr_sessionid, &args->csa_sessionid, sizeof(res->csr_sessionid)); res->csr_sequenceid = args->csa_sequenceid; @@ -159,9 +219,12 @@ unsigned nfs4_callback_sequence(struct cb_sequenceargs *args, res->csr_highestslotid = NFS41_BC_MAX_CALLBACKS - 1; res->csr_target_highestslotid = NFS41_BC_MAX_CALLBACKS - 1; +out_putclient: + nfs_put_client(clp); +out: dprintk("%s: exit with status = %d\n", __func__, ntohl(status)); res->csr_status = status; - return status; + return res->csr_status; } #endif /* CONFIG_NFS_V4_1 */ -- cgit v1.2.3 From 536fc240e7147858255bdb08e7a999a3351a9fb4 Mon Sep 17 00:00:00 2001 From: Hisashi Hifumi Date: Wed, 17 Jun 2009 20:08:51 -0400 Subject: jbd2: clean up jbd2_journal_try_to_free_buffers() This patch reverts 3f31fddf, which is no longer needed because if a race between freeing buffer and committing transaction functionality occurs and dio gets error, currently dio falls back to buffered IO due to the commit 6ccfa806. Signed-off-by: Hisashi Hifumi Cc: Mingming Cao Acked-by: Jan Kara Signed-off-by: Andrew Morton Signed-off-by: "Theodore Ts'o" --- fs/jbd2/transaction.c | 49 ------------------------------------------------- 1 file changed, 49 deletions(-) (limited to 'fs') diff --git a/fs/jbd2/transaction.c b/fs/jbd2/transaction.c index 996ffda06bf3..494501edba6b 100644 --- a/fs/jbd2/transaction.c +++ b/fs/jbd2/transaction.c @@ -1547,36 +1547,6 @@ out: return; } -/* - * jbd2_journal_try_to_free_buffers() could race with - * jbd2_journal_commit_transaction(). The later might still hold the - * reference count to the buffers when inspecting them on - * t_syncdata_list or t_locked_list. - * - * jbd2_journal_try_to_free_buffers() will call this function to - * wait for the current transaction to finish syncing data buffers, before - * try to free that buffer. - * - * Called with journal->j_state_lock hold. - */ -static void jbd2_journal_wait_for_transaction_sync_data(journal_t *journal) -{ - transaction_t *transaction; - tid_t tid; - - spin_lock(&journal->j_state_lock); - transaction = journal->j_committing_transaction; - - if (!transaction) { - spin_unlock(&journal->j_state_lock); - return; - } - - tid = transaction->t_tid; - spin_unlock(&journal->j_state_lock); - jbd2_log_wait_commit(journal, tid); -} - /** * int jbd2_journal_try_to_free_buffers() - try to free page buffers. * @journal: journal for operation @@ -1649,25 +1619,6 @@ int jbd2_journal_try_to_free_buffers(journal_t *journal, ret = try_to_free_buffers(page); - /* - * There are a number of places where jbd2_journal_try_to_free_buffers() - * could race with jbd2_journal_commit_transaction(), the later still - * holds the reference to the buffers to free while processing them. - * try_to_free_buffers() failed to free those buffers. Some of the - * caller of releasepage() request page buffers to be dropped, otherwise - * treat the fail-to-free as errors (such as generic_file_direct_IO()) - * - * So, if the caller of try_to_release_page() wants the synchronous - * behaviour(i.e make sure buffers are dropped upon return), - * let's wait for the current transaction to finish flush of - * dirty data buffers, then try to free those buffers again, - * with the journal locked. - */ - if (ret == 0 && (gfp_mask & __GFP_WAIT) && (gfp_mask & __GFP_FS)) { - jbd2_journal_wait_for_transaction_sync_data(journal); - ret = try_to_free_buffers(page); - } - busy: return ret; } -- cgit v1.2.3 From bf40d3435caf49369058b1ed6bbc92f8e2bd92f1 Mon Sep 17 00:00:00 2001 From: Suresh Jayaraman Date: Wed, 17 Jun 2009 18:02:09 -0700 Subject: NFS: add support for splice writes Adds support for splice writes. It effectively calls generic_file_splice_write() to do the writes. We need not worry about O_APPEND case as the combination of splice() writes and O_APPEND is disallowed. This patch propagates NFS write errors back to the caller. The number of bytes written via splice are being added to NFSIO_NORMALWRITTENBYTES as these are effectively cached writes. Signed-off-by: Suresh Jayaraman Signed-off-by: Trond Myklebust --- fs/nfs/file.c | 31 +++++++++++++++++++++++++++++++ 1 file changed, 31 insertions(+) (limited to 'fs') diff --git a/fs/nfs/file.c b/fs/nfs/file.c index 52ed56762583..0055b813ec2c 100644 --- a/fs/nfs/file.c +++ b/fs/nfs/file.c @@ -48,6 +48,9 @@ static ssize_t nfs_file_splice_read(struct file *filp, loff_t *ppos, size_t count, unsigned int flags); static ssize_t nfs_file_read(struct kiocb *, const struct iovec *iov, unsigned long nr_segs, loff_t pos); +static ssize_t nfs_file_splice_write(struct pipe_inode_info *pipe, + struct file *filp, loff_t *ppos, + size_t count, unsigned int flags); static ssize_t nfs_file_write(struct kiocb *, const struct iovec *iov, unsigned long nr_segs, loff_t pos); static int nfs_file_flush(struct file *, fl_owner_t id); @@ -73,6 +76,7 @@ const struct file_operations nfs_file_operations = { .lock = nfs_lock, .flock = nfs_flock, .splice_read = nfs_file_splice_read, + .splice_write = nfs_file_splice_write, .check_flags = nfs_check_flags, .setlease = nfs_setlease, }; @@ -587,6 +591,33 @@ out_swapfile: goto out; } +static ssize_t nfs_file_splice_write(struct pipe_inode_info *pipe, + struct file *filp, loff_t *ppos, + size_t count, unsigned int flags) +{ + struct dentry *dentry = filp->f_path.dentry; + struct inode *inode = dentry->d_inode; + ssize_t ret; + + dprintk("NFS splice_write(%s/%s, %lu@%llu)\n", + dentry->d_parent->d_name.name, dentry->d_name.name, + (unsigned long) count, (unsigned long long) *ppos); + + /* + * The combination of splice and an O_APPEND destination is disallowed. + */ + + nfs_add_stats(inode, NFSIOS_NORMALWRITTENBYTES, count); + + ret = generic_file_splice_write(pipe, filp, ppos, count, flags); + if (ret >= 0 && nfs_need_sync_write(filp, inode)) { + int err = nfs_do_fsync(nfs_file_open_context(filp), inode); + if (err < 0) + ret = err; + } + return ret; +} + static int do_getlk(struct file *filp, int cmd, struct file_lock *fl) { struct inode *inode = filp->f_mapping->host; -- cgit v1.2.3 From c381ad2cf2d5dcd3991bcc8a18fddd9d5c66ccaa Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Wed, 17 Jun 2009 18:02:09 -0700 Subject: NFS: Do not display the setting of the "intr" mount option The "intr" mount option has been deprecated for a while, but /proc/mounts continues to display "nointr" whether "intr" or "nointr" has been specified for a mount point. Since these options do not have any effect, simply do not display them. Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- fs/nfs/super.c | 1 - 1 file changed, 1 deletion(-) (limited to 'fs') diff --git a/fs/nfs/super.c b/fs/nfs/super.c index 98b47d17740a..539a61a0887b 100644 --- a/fs/nfs/super.c +++ b/fs/nfs/super.c @@ -514,7 +514,6 @@ static void nfs_show_mount_options(struct seq_file *m, struct nfs_server *nfss, const char *nostr; } nfs_info[] = { { NFS_MOUNT_SOFT, ",soft", ",hard" }, - { NFS_MOUNT_INTR, ",intr", ",nointr" }, { NFS_MOUNT_POSIX, ",posix", "" }, { NFS_MOUNT_NOCTO, ",nocto", "" }, { NFS_MOUNT_NOAC, ",noac", "" }, -- cgit v1.2.3 From a21bdd9b960ccce421b63aa0e3efda4fcdc26f10 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Wed, 17 Jun 2009 18:02:10 -0700 Subject: NFS: Return error code from nfs_callback_up() to user space If the kernel cannot start the NFSv4 callback service during a mount request, it returns -ENOMEM to user space, resulting in this message: mount.nfs4: Cannot allocate memory Adjust nfs_alloc_client() and nfs_get_client() to pass NFSv4 callback start-up errors back to user space so a less mysterious error message can be displayed by the mount command. Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- fs/nfs/client.c | 12 ++++++++---- 1 file changed, 8 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/nfs/client.c b/fs/nfs/client.c index 75c9cd2aa119..0e9a1f9743b0 100644 --- a/fs/nfs/client.c +++ b/fs/nfs/client.c @@ -114,6 +114,7 @@ static struct nfs_client *nfs_alloc_client(const struct nfs_client_initdata *cl_ { struct nfs_client *clp; struct rpc_cred *cred; + int err = -ENOMEM; if ((clp = kzalloc(sizeof(*clp), GFP_KERNEL)) == NULL) goto error_0; @@ -121,7 +122,8 @@ static struct nfs_client *nfs_alloc_client(const struct nfs_client_initdata *cl_ clp->rpc_ops = cl_init->rpc_ops; if (cl_init->rpc_ops->version == 4) { - if (nfs_callback_up() < 0) + err = nfs_callback_up(); + if (err < 0) goto error_2; __set_bit(NFS_CS_CALLBACK, &clp->cl_res_state); } @@ -133,6 +135,7 @@ static struct nfs_client *nfs_alloc_client(const struct nfs_client_initdata *cl_ clp->cl_addrlen = cl_init->addrlen; if (cl_init->hostname) { + err = -ENOMEM; clp->cl_hostname = kstrdup(cl_init->hostname, GFP_KERNEL); if (!clp->cl_hostname) goto error_3; @@ -165,7 +168,7 @@ error_3: error_2: kfree(clp); error_0: - return NULL; + return ERR_PTR(err); } static void nfs4_shutdown_client(struct nfs_client *clp) @@ -456,9 +459,10 @@ static struct nfs_client *nfs_get_client(const struct nfs_client_initdata *cl_in spin_unlock(&nfs_client_lock); new = nfs_alloc_client(cl_init); - } while (new); + } while (!IS_ERR(new)); - return ERR_PTR(-ENOMEM); + dprintk("--> nfs_get_client() = %ld [failed]\n", PTR_ERR(new)); + return new; /* install a new client and return with it unready */ install_client: -- cgit v1.2.3 From 18fc31641925867c871bc75270ce642c039188d3 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Wed, 17 Jun 2009 18:02:10 -0700 Subject: NFS: Fix false error return from nfs_callback_up() if ipv6.ko is not available Clear "ret" if the error return from svc_create_xprt(AF_INET6) was -EAFNOSUPORT. Otherwise, callback start-up will succeed, but nfs_callback_up() will return -EAFNOSUPPORT anyway, and the first NFSv4 mount attempt after a reboot will fail. Bug introduced by commit f738f517 in 2.6.30-rc1. Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- fs/nfs/callback.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfs/callback.c b/fs/nfs/callback.c index a886e692ddd0..6e3bd42c21cd 100644 --- a/fs/nfs/callback.c +++ b/fs/nfs/callback.c @@ -127,7 +127,9 @@ int nfs_callback_up(void) nfs_callback_tcpport6 = ret; dprintk("NFS: Callback listener port = %u (af %u)\n", nfs_callback_tcpport6, PF_INET6); - } else if (ret != -EAFNOSUPPORT) + } else if (ret == -EAFNOSUPPORT) + ret = 0; + else goto out_err; #endif /* defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE) */ -- cgit v1.2.3 From 6c9dc4255108bab4ef5c177d369b99c3c23492a7 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Wed, 17 Jun 2009 18:02:10 -0700 Subject: lockd: Update NSM state from SM_MON replies When rpc.statd starts up in user space at boot time, it attempts to write the latest NSM local state number into /proc/sys/fs/nfs/nsm_local_state. If lockd.ko isn't loaded yet (as is the case in most configurations), that file doesn't exist, thus the kernel's NSM state remains set to its initial value of zero during lockd operation. This is a problem because rpc.statd and lockd use the NSM state number to prevent repeated lock recovery on rebooted hosts. If lockd sends a zero NSM state, but then a delayed SM_NOTIFY with a real NSM state number is received, there is no way for lockd or rpc.statd to distinguish that stale SM_NOTIFY from an actual reboot. Thus lock recovery could be performed after the rebooted host has already started reclaiming locks, and those locks will be lost. We could change /etc/init.d/nfslock so it always modprobes lockd.ko before starting rpc.statd. However, if lockd.ko is ever unloaded and reloaded, we are back at square one, since the NSM state is not preserved across an unload/reload cycle. This may happen frequently on clients that use automounter. A period of NFS inactivity causes lockd.ko to be unloaded, and the kernel loses its NSM state setting. Instead, let's use the fact that rpc.statd plants the local system's NSM state in every SM_MON (and SM_UNMON) reply. lockd performs a synchronous SM_MON upcall to the local rpc.statd _before_ sending its first NLM request to a new remote. This would permit rpc.statd to provide the current NSM state to lockd, even after lockd.ko had been unloaded and reloaded. Note that NLMPROC_LOCK arguments are constructed before the nsm_monitor() call, so we have to rearrange argument construction very slightly to make this all work out. And, the kernel appears to treat NSM state as a u32 (see struct nlm_args and nsm_res). Make nsm_local_state a u32 as well, to ensure we don't get bogus comparison results. Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- fs/lockd/clntproc.c | 2 +- fs/lockd/mon.c | 18 ++++++++++++------ include/linux/lockd/lockd.h | 2 +- 3 files changed, 14 insertions(+), 8 deletions(-) (limited to 'fs') diff --git a/fs/lockd/clntproc.c b/fs/lockd/clntproc.c index 273e229353f3..f2fdcbce143e 100644 --- a/fs/lockd/clntproc.c +++ b/fs/lockd/clntproc.c @@ -126,7 +126,6 @@ static void nlmclnt_setlockargs(struct nlm_rqst *req, struct file_lock *fl) struct nlm_lock *lock = &argp->lock; nlmclnt_next_cookie(&argp->cookie); - argp->state = nsm_local_state; memcpy(&lock->fh, NFS_FH(fl->fl_file->f_path.dentry->d_inode), sizeof(struct nfs_fh)); lock->caller = utsname()->nodename; lock->oh.data = req->a_owner; @@ -521,6 +520,7 @@ nlmclnt_lock(struct nlm_rqst *req, struct file_lock *fl) if (nsm_monitor(host) < 0) goto out; + req->a_args.state = nsm_local_state; fl->fl_flags |= FL_ACCESS; status = do_vfs_lock(fl); diff --git a/fs/lockd/mon.c b/fs/lockd/mon.c index 6d5d4a4169e5..38385336614c 100644 --- a/fs/lockd/mon.c +++ b/fs/lockd/mon.c @@ -53,7 +53,7 @@ static DEFINE_SPINLOCK(nsm_lock); /* * Local NSM state */ -int __read_mostly nsm_local_state; +u32 __read_mostly nsm_local_state; int __read_mostly nsm_use_hostnames; static inline struct sockaddr *nsm_addr(const struct nsm_handle *nsm) @@ -184,13 +184,19 @@ int nsm_monitor(const struct nlm_host *host) nsm->sm_mon_name = nsm_use_hostnames ? nsm->sm_name : nsm->sm_addrbuf; status = nsm_mon_unmon(nsm, NSMPROC_MON, &res); - if (res.status != 0) + if (unlikely(res.status != 0)) status = -EIO; - if (status < 0) + if (unlikely(status < 0)) { printk(KERN_NOTICE "lockd: cannot monitor %s\n", nsm->sm_name); - else - nsm->sm_monitored = 1; - return status; + return status; + } + + nsm->sm_monitored = 1; + if (unlikely(nsm_local_state != res.state)) { + nsm_local_state = res.state; + dprintk("lockd: NSM state changed to %d\n", nsm_local_state); + } + return 0; } /** diff --git a/include/linux/lockd/lockd.h b/include/linux/lockd/lockd.h index 51855dfd8adb..c325b187966b 100644 --- a/include/linux/lockd/lockd.h +++ b/include/linux/lockd/lockd.h @@ -195,7 +195,7 @@ extern struct svc_procedure nlmsvc_procedures4[]; extern int nlmsvc_grace_period; extern unsigned long nlmsvc_timeout; extern int nsm_use_hostnames; -extern int nsm_local_state; +extern u32 nsm_local_state; /* * Lockd client functions -- cgit v1.2.3 From 0e5c2632e1c9182f0dadc31bec68d6f42e7905ea Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Wed, 17 Jun 2009 18:02:11 -0700 Subject: lockd: Don't bother with RPC ping for NSM upcalls Cut NSM upcall RPC traffic in half -- don't do a NULL call first. The cases where a ping would be helpful are rare. Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- fs/lockd/mon.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/lockd/mon.c b/fs/lockd/mon.c index 38385336614c..7fce1b525849 100644 --- a/fs/lockd/mon.c +++ b/fs/lockd/mon.c @@ -112,6 +112,7 @@ static struct rpc_clnt *nsm_create(void) .program = &nsm_program, .version = NSM_VERSION, .authflavor = RPC_AUTH_NULL, + .flags = RPC_CLNT_CREATE_NOPING, }; return rpc_create(&args); -- cgit v1.2.3 From 2ad780978b7c0c3e7877949f098cbd06e7c73839 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Wed, 17 Jun 2009 18:02:11 -0700 Subject: NFS: Clean up MNT program definitions Clean up: Relocate MNT program procedure number definitions to the only file that uses them. Relocate the version number definitions, which are shared, to nfs.h. Remove duplicate program number definitions. Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- fs/nfs/mount_clnt.c | 32 ++++++++++++++++++++++++++++---- fs/nfs/nfsroot.c | 3 +++ include/linux/nfs.h | 5 +++-- include/linux/nfs2.h | 7 ------- include/linux/nfs3.h | 5 ----- 5 files changed, 34 insertions(+), 18 deletions(-) (limited to 'fs') diff --git a/fs/nfs/mount_clnt.c b/fs/nfs/mount_clnt.c index ca905a5bb1ba..af45a374d56f 100644 --- a/fs/nfs/mount_clnt.c +++ b/fs/nfs/mount_clnt.c @@ -20,6 +20,30 @@ # define NFSDBG_FACILITY NFSDBG_MOUNT #endif +/* + * Defined by RFC 1094, section A.5 + */ +enum { + MOUNTPROC_NULL = 0, + MOUNTPROC_MNT = 1, + MOUNTPROC_DUMP = 2, + MOUNTPROC_UMNT = 3, + MOUNTPROC_UMNTALL = 4, + MOUNTPROC_EXPORT = 5, +}; + +/* + * Defined by RFC 1813, section 5.2 + */ +enum { + MOUNTPROC3_NULL = 0, + MOUNTPROC3_MNT = 1, + MOUNTPROC3_DUMP = 2, + MOUNTPROC3_UMNT = 3, + MOUNTPROC3_UMNTALL = 4, + MOUNTPROC3_EXPORT = 5, +}; + static struct rpc_program mnt_program; struct mnt_fhstatus { @@ -68,7 +92,7 @@ int nfs_mount(struct nfs_mount_request *info) if (info->version == NFS_MNT3_VERSION) msg.rpc_proc = &mnt_clnt->cl_procinfo[MOUNTPROC3_MNT]; else - msg.rpc_proc = &mnt_clnt->cl_procinfo[MNTPROC_MNT]; + msg.rpc_proc = &mnt_clnt->cl_procinfo[MOUNTPROC_MNT]; status = rpc_call_sync(mnt_clnt, &msg, 0); rpc_shutdown_client(mnt_clnt); @@ -145,13 +169,13 @@ static int xdr_decode_fhstatus3(struct rpc_rqst *req, __be32 *p, #define MNT_fhstatus3_sz (1 + 16) static struct rpc_procinfo mnt_procedures[] = { - [MNTPROC_MNT] = { - .p_proc = MNTPROC_MNT, + [MOUNTPROC_MNT] = { + .p_proc = MOUNTPROC_MNT, .p_encode = (kxdrproc_t) xdr_encode_dirpath, .p_decode = (kxdrproc_t) xdr_decode_fhstatus, .p_arglen = MNT_dirpath_sz, .p_replen = MNT_fhstatus_sz, - .p_statidx = MNTPROC_MNT, + .p_statidx = MOUNTPROC_MNT, .p_name = "MOUNT", }, }; diff --git a/fs/nfs/nfsroot.c b/fs/nfs/nfsroot.c index e3ed5908820b..24c1b93874c4 100644 --- a/fs/nfs/nfsroot.c +++ b/fs/nfs/nfsroot.c @@ -92,6 +92,9 @@ #undef NFSROOT_DEBUG #define NFSDBG_FACILITY NFSDBG_ROOT +/* Default port to use if server is not running a portmapper */ +#define NFS_MNT_PORT 627 + /* Default path we try to mount. "%s" gets replaced by our IP address */ #define NFS_ROOT "/tftpboot/%s" diff --git a/include/linux/nfs.h b/include/linux/nfs.h index 214d499718f7..f387919bbc59 100644 --- a/include/linux/nfs.h +++ b/include/linux/nfs.h @@ -25,8 +25,9 @@ #define NFSMODE_SOCK 0140000 #define NFSMODE_FIFO 0010000 -#define NFS_MNT_PROGRAM 100005 -#define NFS_MNT_PORT 627 +#define NFS_MNT_PROGRAM 100005 +#define NFS_MNT_VERSION 1 +#define NFS_MNT3_VERSION 3 /* * NFS stats. The good thing with these values is that NFSv3 errors are diff --git a/include/linux/nfs2.h b/include/linux/nfs2.h index 0ed9517138fc..fde24b30cc9e 100644 --- a/include/linux/nfs2.h +++ b/include/linux/nfs2.h @@ -64,11 +64,4 @@ struct nfs2_fh { #define NFSPROC_READDIR 16 #define NFSPROC_STATFS 17 -#define NFS_MNT_PROGRAM 100005 -#define NFS_MNT_VERSION 1 -#define MNTPROC_NULL 0 -#define MNTPROC_MNT 1 -#define MNTPROC_UMNT 3 -#define MNTPROC_UMNTALL 4 - #endif /* _LINUX_NFS2_H */ diff --git a/include/linux/nfs3.h b/include/linux/nfs3.h index 539f3b550eab..ac33806ec7f9 100644 --- a/include/linux/nfs3.h +++ b/include/linux/nfs3.h @@ -88,12 +88,7 @@ struct nfs3_fh { #define NFS3PROC_PATHCONF 20 #define NFS3PROC_COMMIT 21 -#define NFS_MNT3_PROGRAM 100005 #define NFS_MNT3_VERSION 3 -#define MOUNTPROC3_NULL 0 -#define MOUNTPROC3_MNT 1 -#define MOUNTPROC3_UMNT 3 -#define MOUNTPROC3_UMNTALL 4 #if defined(__KERNEL__) -- cgit v1.2.3 From 29a1bd6bf8c7a9ff511aaaf7e40fd6ca5198babe Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Wed, 17 Jun 2009 18:02:11 -0700 Subject: NFS: Use xdr_stream-based XDR encoder for MNT's dirpath argument Check the length of the supplied dirpath, and see that it fits properly in the RPC buffer. Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- fs/nfs/mount_clnt.c | 49 ++++++++++++++++++++++++++++++++++++++++++++----- 1 file changed, 44 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/nfs/mount_clnt.c b/fs/nfs/mount_clnt.c index af45a374d56f..93361afcd5da 100644 --- a/fs/nfs/mount_clnt.c +++ b/fs/nfs/mount_clnt.c @@ -20,6 +20,21 @@ # define NFSDBG_FACILITY NFSDBG_MOUNT #endif +/* + * Defined by RFC 1094, section A.3; and RFC 1813, section 5.1.4 + */ +#define MNTPATHLEN (1024) + +/* + * XDR data type sizes + */ +#define encode_dirpath_sz (1 + XDR_QUADLEN(MNTPATHLEN)) + +/* + * XDR argument and result sizes + */ +#define MNT_enc_dirpath_sz encode_dirpath_sz + /* * Defined by RFC 1094, section A.5 */ @@ -135,6 +150,31 @@ static int xdr_encode_dirpath(struct rpc_rqst *req, __be32 *p, return 0; } +static int encode_mntdirpath(struct xdr_stream *xdr, const char *pathname) +{ + const u32 pathname_len = strlen(pathname); + __be32 *p; + + if (unlikely(pathname_len > MNTPATHLEN)) + return -EIO; + + p = xdr_reserve_space(xdr, sizeof(u32) + pathname_len); + if (unlikely(p == NULL)) + return -EIO; + xdr_encode_opaque(p, pathname, pathname_len); + + return 0; +} + +static int mnt_enc_dirpath(struct rpc_rqst *req, __be32 *p, + const char *dirpath) +{ + struct xdr_stream xdr; + + xdr_init_encode(&xdr, &req->rq_snd_buf, p); + return encode_mntdirpath(&xdr, dirpath); +} + static int xdr_decode_fhstatus(struct rpc_rqst *req, __be32 *p, struct mnt_fhstatus *res) { @@ -164,16 +204,15 @@ static int xdr_decode_fhstatus3(struct rpc_rqst *req, __be32 *p, return 0; } -#define MNT_dirpath_sz (1 + 256) #define MNT_fhstatus_sz (1 + 8) #define MNT_fhstatus3_sz (1 + 16) static struct rpc_procinfo mnt_procedures[] = { [MOUNTPROC_MNT] = { .p_proc = MOUNTPROC_MNT, - .p_encode = (kxdrproc_t) xdr_encode_dirpath, + .p_encode = (kxdrproc_t)mnt_enc_dirpath, .p_decode = (kxdrproc_t) xdr_decode_fhstatus, - .p_arglen = MNT_dirpath_sz, + .p_arglen = MNT_enc_dirpath_sz, .p_replen = MNT_fhstatus_sz, .p_statidx = MOUNTPROC_MNT, .p_name = "MOUNT", @@ -183,9 +222,9 @@ static struct rpc_procinfo mnt_procedures[] = { static struct rpc_procinfo mnt3_procedures[] = { [MOUNTPROC3_MNT] = { .p_proc = MOUNTPROC3_MNT, - .p_encode = (kxdrproc_t) xdr_encode_dirpath, + .p_encode = (kxdrproc_t)mnt_enc_dirpath, .p_decode = (kxdrproc_t) xdr_decode_fhstatus3, - .p_arglen = MNT_dirpath_sz, + .p_arglen = MNT_enc_dirpath_sz, .p_replen = MNT_fhstatus3_sz, .p_statidx = MOUNTPROC3_MNT, .p_name = "MOUNT", -- cgit v1.2.3 From 99835db430904e90c0640ebc6b91cd2a90a118f7 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Wed, 17 Jun 2009 18:02:11 -0700 Subject: NFS: remove unused function in fs/nfs/mount_clnt.c Clean up: remove xdr_encode_dirpath() now that it has been replaced. Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- fs/nfs/mount_clnt.c | 8 -------- 1 file changed, 8 deletions(-) (limited to 'fs') diff --git a/fs/nfs/mount_clnt.c b/fs/nfs/mount_clnt.c index 93361afcd5da..79b5954b8a1b 100644 --- a/fs/nfs/mount_clnt.c +++ b/fs/nfs/mount_clnt.c @@ -141,14 +141,6 @@ out_mnt_err: /* * XDR encode/decode functions for MOUNT */ -static int xdr_encode_dirpath(struct rpc_rqst *req, __be32 *p, - const char *path) -{ - p = xdr_encode_string(p, path); - - req->rq_slen = xdr_adjust_iovec(req->rq_svec, p); - return 0; -} static int encode_mntdirpath(struct xdr_stream *xdr, const char *pathname) { -- cgit v1.2.3 From fb12529577541aa02f9c3d9e325329f9568dfb58 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Wed, 17 Jun 2009 18:02:12 -0700 Subject: NFS: Add separate mountd status code decoders for each mountd version Introduce data structures and xdr_stream-based decoding functions for unmarshalling mountd status codes properly. Mountd version 3 uses specific standard error return codes that are not errno values and not NFS3ERR_ values. These have a well-defined standard mapping to local errno values. Introduce data structures and a decoder function that map these status codes to local errno values properly. This is new functionality (but not used yet). Version 1 mountd status values are defined by RFC 1094 as UNIX error values (errno values). Errno values on heterogeneous systems do not necessarily match each other. To avoid exposing possibly incorrect errno values to upper layers, the current XDR decoder converts all non-zero MNT version 1 status codes to -EACCES. The OpenGroup XNFS standard provides a mapping similar to but smaller than the version 3 error codes. Implement a decoder that uses the XNFS error codes, replacing the current decoder. For both mountd protocol versions, map unrecognized errors to -EACCES. Finally we introduce a replacement data structure for mnt_fhstatus at this time, which is used by the new XDR decoders. In addition to documenting that the status value returned by the XDR decoders is always an errno, this new structure will be expanded in subsequent patches. Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- fs/nfs/mount_clnt.c | 116 ++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 116 insertions(+) (limited to 'fs') diff --git a/fs/nfs/mount_clnt.c b/fs/nfs/mount_clnt.c index 79b5954b8a1b..8429885bc729 100644 --- a/fs/nfs/mount_clnt.c +++ b/fs/nfs/mount_clnt.c @@ -29,6 +29,8 @@ * XDR data type sizes */ #define encode_dirpath_sz (1 + XDR_QUADLEN(MNTPATHLEN)) +#define MNT_status_sz (1) +#define MNT_fhs_status_sz (1) /* * XDR argument and result sizes @@ -61,6 +63,65 @@ enum { static struct rpc_program mnt_program; +/* + * Defined by OpenGroup XNFS Version 3W, chapter 8 + */ +enum mountstat { + MNT_OK = 0, + MNT_EPERM = 1, + MNT_ENOENT = 2, + MNT_EACCES = 13, + MNT_EINVAL = 22, +}; + +static struct { + u32 status; + int errno; +} mnt_errtbl[] = { + { .status = MNT_OK, .errno = 0, }, + { .status = MNT_EPERM, .errno = -EPERM, }, + { .status = MNT_ENOENT, .errno = -ENOENT, }, + { .status = MNT_EACCES, .errno = -EACCES, }, + { .status = MNT_EINVAL, .errno = -EINVAL, }, +}; + +/* + * Defined by RFC 1813, section 5.1.5 + */ +enum mountstat3 { + MNT3_OK = 0, /* no error */ + MNT3ERR_PERM = 1, /* Not owner */ + MNT3ERR_NOENT = 2, /* No such file or directory */ + MNT3ERR_IO = 5, /* I/O error */ + MNT3ERR_ACCES = 13, /* Permission denied */ + MNT3ERR_NOTDIR = 20, /* Not a directory */ + MNT3ERR_INVAL = 22, /* Invalid argument */ + MNT3ERR_NAMETOOLONG = 63, /* Filename too long */ + MNT3ERR_NOTSUPP = 10004, /* Operation not supported */ + MNT3ERR_SERVERFAULT = 10006, /* A failure on the server */ +}; + +static struct { + u32 status; + int errno; +} mnt3_errtbl[] = { + { .status = MNT3_OK, .errno = 0, }, + { .status = MNT3ERR_PERM, .errno = -EPERM, }, + { .status = MNT3ERR_NOENT, .errno = -ENOENT, }, + { .status = MNT3ERR_IO, .errno = -EIO, }, + { .status = MNT3ERR_ACCES, .errno = -EACCES, }, + { .status = MNT3ERR_NOTDIR, .errno = -ENOTDIR, }, + { .status = MNT3ERR_INVAL, .errno = -EINVAL, }, + { .status = MNT3ERR_NAMETOOLONG, .errno = -ENAMETOOLONG, }, + { .status = MNT3ERR_NOTSUPP, .errno = -ENOTSUPP, }, + { .status = MNT3ERR_SERVERFAULT, .errno = -ESERVERFAULT, }, +}; + +struct mountres { + int errno; + struct nfs_fh *fh; +}; + struct mnt_fhstatus { u32 status; struct nfs_fh *fh; @@ -179,6 +240,61 @@ static int xdr_decode_fhstatus(struct rpc_rqst *req, __be32 *p, return 0; } +/* + * RFC 1094: "A non-zero status indicates some sort of error. In this + * case, the status is a UNIX error number." This can be problematic + * if the server and client use different errno values for the same + * error. + * + * However, the OpenGroup XNFS spec provides a simple mapping that is + * independent of local errno values on the server and the client. + */ +static int decode_status(struct xdr_stream *xdr, struct mountres *res) +{ + unsigned int i; + u32 status; + __be32 *p; + + p = xdr_inline_decode(xdr, sizeof(status)); + if (unlikely(p == NULL)) + return -EIO; + status = ntohl(*p); + + for (i = 0; i <= ARRAY_SIZE(mnt_errtbl); i++) { + if (mnt_errtbl[i].status == status) { + res->errno = mnt_errtbl[i].errno; + return 0; + } + } + + dprintk("NFS: unrecognized MNT status code: %u\n", status); + res->errno = -EACCES; + return 0; +} + +static int decode_fhs_status(struct xdr_stream *xdr, struct mountres *res) +{ + unsigned int i; + u32 status; + __be32 *p; + + p = xdr_inline_decode(xdr, sizeof(status)); + if (unlikely(p == NULL)) + return -EIO; + status = ntohl(*p); + + for (i = 0; i <= ARRAY_SIZE(mnt3_errtbl); i++) { + if (mnt3_errtbl[i].status == status) { + res->errno = mnt3_errtbl[i].errno; + return 0; + } + } + + dprintk("NFS: unrecognized MNT3 status code: %u\n", status); + res->errno = -EACCES; + return 0; +} + static int xdr_decode_fhstatus3(struct rpc_rqst *req, __be32 *p, struct mnt_fhstatus *res) { -- cgit v1.2.3 From 4fdcd9966d8469be26a6f12122ac21ffce19fc20 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Wed, 17 Jun 2009 18:02:12 -0700 Subject: NFS: add new file handle decoders to in-kernel mountd client Introduce xdr_stream-based XDR file handle decoders to the in-kernel mountd client. These are more careful than the existing decoder functions about buffer overflows and data type and range checking. Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- fs/nfs/mount_clnt.c | 39 +++++++++++++++++++++++++++++++++++++++ 1 file changed, 39 insertions(+) (limited to 'fs') diff --git a/fs/nfs/mount_clnt.c b/fs/nfs/mount_clnt.c index 8429885bc729..dd8c6f448eaa 100644 --- a/fs/nfs/mount_clnt.c +++ b/fs/nfs/mount_clnt.c @@ -31,6 +31,8 @@ #define encode_dirpath_sz (1 + XDR_QUADLEN(MNTPATHLEN)) #define MNT_status_sz (1) #define MNT_fhs_status_sz (1) +#define MNT_fhandle_sz XDR_QUADLEN(NFS2_FHSIZE) +#define MNT_fhandle3_sz (1 + XDR_QUADLEN(NFS3_FHSIZE)) /* * XDR argument and result sizes @@ -272,6 +274,20 @@ static int decode_status(struct xdr_stream *xdr, struct mountres *res) return 0; } +static int decode_fhandle(struct xdr_stream *xdr, struct mountres *res) +{ + struct nfs_fh *fh = res->fh; + __be32 *p; + + p = xdr_inline_decode(xdr, NFS2_FHSIZE); + if (unlikely(p == NULL)) + return -EIO; + + fh->size = NFS2_FHSIZE; + memcpy(fh->data, p, NFS2_FHSIZE); + return 0; +} + static int decode_fhs_status(struct xdr_stream *xdr, struct mountres *res) { unsigned int i; @@ -295,6 +311,29 @@ static int decode_fhs_status(struct xdr_stream *xdr, struct mountres *res) return 0; } +static int decode_fhandle3(struct xdr_stream *xdr, struct mountres *res) +{ + struct nfs_fh *fh = res->fh; + u32 size; + __be32 *p; + + p = xdr_inline_decode(xdr, sizeof(size)); + if (unlikely(p == NULL)) + return -EIO; + + size = ntohl(*p++); + if (size > NFS3_FHSIZE || size == 0) + return -EIO; + + p = xdr_inline_decode(xdr, size); + if (unlikely(p == NULL)) + return -EIO; + + fh->size = size; + memcpy(fh->data, p, size); + return 0; +} + static int xdr_decode_fhstatus3(struct rpc_rqst *req, __be32 *p, struct mnt_fhstatus *res) { -- cgit v1.2.3 From a14017db2852f9393a401a0f64053c331003babf Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Wed, 17 Jun 2009 18:02:12 -0700 Subject: NFS: add XDR decoder for mountd version 3 auth-flavor lists Introduce an xdr_stream-based XDR decoder that can unpack the auth- flavor list returned in a MNT3 reply. The nfs_mount() function's caller allocates an array, and passes the size and a pointer to it. The decoder decodes all the flavors it can into the array, and returns the number of decoded flavors. If the caller is not interested in the auth flavors, it can pass a value of zero as the size of the pre-allocated array. Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- fs/nfs/internal.h | 6 ++++++ fs/nfs/mount_clnt.c | 37 +++++++++++++++++++++++++++++++++++++ 2 files changed, 43 insertions(+) (limited to 'fs') diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h index e4d6a8348adf..0207758de44a 100644 --- a/fs/nfs/internal.h +++ b/fs/nfs/internal.h @@ -29,6 +29,12 @@ struct nfs_clone_mount { rpc_authflavor_t authflavor; }; +/* + * Note: RFC 1813 doesn't limit the number of auth flavors that + * a server can return, so make something up. + */ +#define NFS_MAX_SECFLAVORS (12) + /* * In-kernel mount arguments */ diff --git a/fs/nfs/mount_clnt.c b/fs/nfs/mount_clnt.c index dd8c6f448eaa..ed0a27ed538d 100644 --- a/fs/nfs/mount_clnt.c +++ b/fs/nfs/mount_clnt.c @@ -33,6 +33,7 @@ #define MNT_fhs_status_sz (1) #define MNT_fhandle_sz XDR_QUADLEN(NFS2_FHSIZE) #define MNT_fhandle3_sz (1 + XDR_QUADLEN(NFS3_FHSIZE)) +#define MNT_authflav3_sz (1 + NFS_MAX_SECFLAVORS) /* * XDR argument and result sizes @@ -122,6 +123,8 @@ static struct { struct mountres { int errno; struct nfs_fh *fh; + unsigned int *auth_count; + rpc_authflavor_t *auth_flavors; }; struct mnt_fhstatus { @@ -334,6 +337,40 @@ static int decode_fhandle3(struct xdr_stream *xdr, struct mountres *res) return 0; } +static int decode_auth_flavors(struct xdr_stream *xdr, struct mountres *res) +{ + rpc_authflavor_t *flavors = res->auth_flavors; + unsigned int *count = res->auth_count; + u32 entries, i; + __be32 *p; + + if (*count == 0) + return 0; + + p = xdr_inline_decode(xdr, sizeof(entries)); + if (unlikely(p == NULL)) + return -EIO; + entries = ntohl(*p); + dprintk("NFS: received %u auth flavors\n", entries); + if (entries > NFS_MAX_SECFLAVORS) + entries = NFS_MAX_SECFLAVORS; + + p = xdr_inline_decode(xdr, sizeof(u32) * entries); + if (unlikely(p == NULL)) + return -EIO; + + if (entries > *count) + entries = *count; + + for (i = 0; i < entries; i++) { + flavors[i] = ntohl(*p++); + dprintk("NFS:\tflavor %u: %d\n", i, flavors[i]); + } + *count = i; + + return 0; +} + static int xdr_decode_fhstatus3(struct rpc_rqst *req, __be32 *p, struct mnt_fhstatus *res) { -- cgit v1.2.3 From 8e02f6b9aae9b265064f929c6df15222b9baf256 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Wed, 17 Jun 2009 18:02:13 -0700 Subject: NFS: Update MNT and MNT3 reply decoding functions Solder xdr_stream-based XDR decoding functions into the in-kernel mountd client that are more careful about checking data types and watching for buffer overflows. The new MNT3 decoder includes support for auth-flavor list decoding. The "_sz" macro for MNT3 replies was missing the size of the file handle. I've added this back, and included the size of the auth flavor array. Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- fs/nfs/internal.h | 2 ++ fs/nfs/mount_clnt.c | 63 +++++++++++++++++++++++++++++++++++++++++------------ fs/nfs/nfsroot.c | 2 ++ fs/nfs/super.c | 2 ++ 4 files changed, 55 insertions(+), 14 deletions(-) (limited to 'fs') diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h index 0207758de44a..10cf1110df48 100644 --- a/fs/nfs/internal.h +++ b/fs/nfs/internal.h @@ -83,6 +83,8 @@ struct nfs_mount_request { unsigned short protocol; struct nfs_fh *fh; int noresvport; + unsigned int *auth_flav_len; + rpc_authflavor_t *auth_flavs; }; extern int nfs_mount(struct nfs_mount_request *info); diff --git a/fs/nfs/mount_clnt.c b/fs/nfs/mount_clnt.c index ed0a27ed538d..618d815c3093 100644 --- a/fs/nfs/mount_clnt.c +++ b/fs/nfs/mount_clnt.c @@ -39,6 +39,9 @@ * XDR argument and result sizes */ #define MNT_enc_dirpath_sz encode_dirpath_sz +#define MNT_dec_mountres_sz (MNT_status_sz + MNT_fhandle_sz) +#define MNT_dec_mountres3_sz (MNT_status_sz + MNT_fhandle_sz + \ + MNT_authflav3_sz) /* * Defined by RFC 1094, section A.5 @@ -140,8 +143,10 @@ struct mnt_fhstatus { */ int nfs_mount(struct nfs_mount_request *info) { - struct mnt_fhstatus result = { - .fh = info->fh + struct mountres result = { + .fh = info->fh, + .auth_count = info->auth_flav_len, + .auth_flavors = info->auth_flavs, }; struct rpc_message msg = { .rpc_argp = info->dirpath, @@ -180,7 +185,7 @@ int nfs_mount(struct nfs_mount_request *info) if (status < 0) goto out_call_err; - if (result.status != 0) + if (result.errno != 0) goto out_mnt_err; dprintk("NFS: MNT request succeeded\n"); @@ -191,16 +196,16 @@ out: out_clnt_err: status = PTR_ERR(mnt_clnt); - dprintk("NFS: failed to create RPC client, status=%d\n", status); + dprintk("NFS: failed to create MNT RPC client, status=%d\n", status); goto out; out_call_err: - dprintk("NFS: failed to start MNT request, status=%d\n", status); + dprintk("NFS: MNT request failed, status=%d\n", status); goto out; out_mnt_err: - dprintk("NFS: MNT server returned result %d\n", result.status); - status = nfs_stat_to_errno(result.status); + dprintk("NFS: MNT server returned result %d\n", result.errno); + status = result.errno; goto out; } @@ -291,6 +296,20 @@ static int decode_fhandle(struct xdr_stream *xdr, struct mountres *res) return 0; } +static int mnt_dec_mountres(struct rpc_rqst *req, __be32 *p, + struct mountres *res) +{ + struct xdr_stream xdr; + int status; + + xdr_init_decode(&xdr, &req->rq_rcv_buf, p); + + status = decode_status(&xdr, res); + if (unlikely(status != 0 || res->errno != 0)) + return status; + return decode_fhandle(&xdr, res); +} + static int decode_fhs_status(struct xdr_stream *xdr, struct mountres *res) { unsigned int i; @@ -371,6 +390,25 @@ static int decode_auth_flavors(struct xdr_stream *xdr, struct mountres *res) return 0; } +static int mnt_dec_mountres3(struct rpc_rqst *req, __be32 *p, + struct mountres *res) +{ + struct xdr_stream xdr; + int status; + + xdr_init_decode(&xdr, &req->rq_rcv_buf, p); + + status = decode_fhs_status(&xdr, res); + if (unlikely(status != 0 || res->errno != 0)) + return status; + status = decode_fhandle3(&xdr, res); + if (unlikely(status != 0)) { + res->errno = -EBADHANDLE; + return 0; + } + return decode_auth_flavors(&xdr, res); +} + static int xdr_decode_fhstatus3(struct rpc_rqst *req, __be32 *p, struct mnt_fhstatus *res) { @@ -388,16 +426,13 @@ static int xdr_decode_fhstatus3(struct rpc_rqst *req, __be32 *p, return 0; } -#define MNT_fhstatus_sz (1 + 8) -#define MNT_fhstatus3_sz (1 + 16) - static struct rpc_procinfo mnt_procedures[] = { [MOUNTPROC_MNT] = { .p_proc = MOUNTPROC_MNT, .p_encode = (kxdrproc_t)mnt_enc_dirpath, - .p_decode = (kxdrproc_t) xdr_decode_fhstatus, + .p_decode = (kxdrproc_t)mnt_dec_mountres, .p_arglen = MNT_enc_dirpath_sz, - .p_replen = MNT_fhstatus_sz, + .p_replen = MNT_dec_mountres_sz, .p_statidx = MOUNTPROC_MNT, .p_name = "MOUNT", }, @@ -407,9 +442,9 @@ static struct rpc_procinfo mnt3_procedures[] = { [MOUNTPROC3_MNT] = { .p_proc = MOUNTPROC3_MNT, .p_encode = (kxdrproc_t)mnt_enc_dirpath, - .p_decode = (kxdrproc_t) xdr_decode_fhstatus3, + .p_decode = (kxdrproc_t)mnt_dec_mountres3, .p_arglen = MNT_enc_dirpath_sz, - .p_replen = MNT_fhstatus3_sz, + .p_replen = MNT_dec_mountres3_sz, .p_statidx = MOUNTPROC3_MNT, .p_name = "MOUNT", }, diff --git a/fs/nfs/nfsroot.c b/fs/nfs/nfsroot.c index 24c1b93874c4..8c55b27c0de4 100644 --- a/fs/nfs/nfsroot.c +++ b/fs/nfs/nfsroot.c @@ -490,6 +490,7 @@ static int __init root_nfs_get_handle(void) { struct nfs_fh fh; struct sockaddr_in sin; + unsigned int auth_flav_len = 0; struct nfs_mount_request request = { .sap = (struct sockaddr *)&sin, .salen = sizeof(sin), @@ -499,6 +500,7 @@ static int __init root_nfs_get_handle(void) .protocol = (nfs_data.flags & NFS_MOUNT_TCP) ? XPRT_TRANSPORT_TCP : XPRT_TRANSPORT_UDP, .fh = &fh, + .auth_flav_len = &auth_flav_len, }; int status; diff --git a/fs/nfs/super.c b/fs/nfs/super.c index 539a61a0887b..0b72357cdc41 100644 --- a/fs/nfs/super.c +++ b/fs/nfs/super.c @@ -1380,6 +1380,7 @@ out_security_failure: static int nfs_try_mount(struct nfs_parsed_mount_data *args, struct nfs_fh *root_fh) { + unsigned int auth_flavor_len = 0; struct nfs_mount_request request = { .sap = (struct sockaddr *) &args->mount_server.address, @@ -1387,6 +1388,7 @@ static int nfs_try_mount(struct nfs_parsed_mount_data *args, .protocol = args->mount_server.protocol, .fh = root_fh, .noresvport = args->flags & NFS_MOUNT_NORESVPORT, + .auth_flav_len = &auth_flavor_len, }; int status; -- cgit v1.2.3 From 065015e5efff60884ad600a3e9a5127dbb684429 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Wed, 17 Jun 2009 18:02:13 -0700 Subject: NFS: Remove unused XDR decoder functions Clean up: Remove xdr_decode_fhstatus() and xdr_decode_fhstatus3(), now that they are unused. Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- fs/nfs/mount_clnt.c | 29 ----------------------------- 1 file changed, 29 deletions(-) (limited to 'fs') diff --git a/fs/nfs/mount_clnt.c b/fs/nfs/mount_clnt.c index 618d815c3093..38ef9eaec407 100644 --- a/fs/nfs/mount_clnt.c +++ b/fs/nfs/mount_clnt.c @@ -238,18 +238,6 @@ static int mnt_enc_dirpath(struct rpc_rqst *req, __be32 *p, return encode_mntdirpath(&xdr, dirpath); } -static int xdr_decode_fhstatus(struct rpc_rqst *req, __be32 *p, - struct mnt_fhstatus *res) -{ - struct nfs_fh *fh = res->fh; - - if ((res->status = ntohl(*p++)) == 0) { - fh->size = NFS2_FHSIZE; - memcpy(fh->data, p, NFS2_FHSIZE); - } - return 0; -} - /* * RFC 1094: "A non-zero status indicates some sort of error. In this * case, the status is a UNIX error number." This can be problematic @@ -409,23 +397,6 @@ static int mnt_dec_mountres3(struct rpc_rqst *req, __be32 *p, return decode_auth_flavors(&xdr, res); } -static int xdr_decode_fhstatus3(struct rpc_rqst *req, __be32 *p, - struct mnt_fhstatus *res) -{ - struct nfs_fh *fh = res->fh; - unsigned size; - - if ((res->status = ntohl(*p++)) == 0) { - size = ntohl(*p++); - if (size <= NFS3_FHSIZE && size != 0) { - fh->size = size; - memcpy(fh->data, p, size); - } else - res->status = -EBADHANDLE; - } - return 0; -} - static struct rpc_procinfo mnt_procedures[] = { [MOUNTPROC_MNT] = { .p_proc = MOUNTPROC_MNT, -- cgit v1.2.3 From d23c45fd84f79a3b84899dac053dcafe9d43ebc9 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Wed, 17 Jun 2009 18:02:13 -0700 Subject: NFS: Invalid mount option values should always fail, even with "sloppy" Ian Kent reports: "I've noticed a couple of other regressions with the options vers and proto option of mount.nfs(8). The commands: mount -t nfs -o vers= :/ / mount -t nfs -o proto= :/ / both immediately fail. But if the "-s" option is also used they both succeed with the mount falling back to defaults (by the look of it). In the past these failed even when the sloppy option was given, as I think they should. I believe the sloppy option is meant to allow the mount command to still function for mount options (for example in shared autofs maps) that exist on other Unix implementations but aren't present in the Linux mount.nfs(8). So, an invalid value specified for a known mount option is different to an unknown mount option and should fail appropriately." See RH bugzilla 486266. Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- fs/nfs/super.c | 157 ++++++++++++++++++++++----------------------------------- 1 file changed, 61 insertions(+), 96 deletions(-) (limited to 'fs') diff --git a/fs/nfs/super.c b/fs/nfs/super.c index 0b72357cdc41..a2b2805caf9d 100644 --- a/fs/nfs/super.c +++ b/fs/nfs/super.c @@ -942,11 +942,6 @@ static int nfs_parse_security_flavors(char *value, return 1; } -static void nfs_parse_invalid_value(const char *option) -{ - dfprintk(MOUNT, "NFS: bad value specified for %s option\n", option); -} - /* * Error-check and convert a string of mount options from user space into * a data structure. The whole mount string is processed; bad options are @@ -957,7 +952,7 @@ static int nfs_parse_mount_options(char *raw, struct nfs_parsed_mount_data *mnt) { char *p, *string, *secdata; - int rc, sloppy = 0, errors = 0; + int rc, sloppy = 0, invalid_option = 0; if (!raw) { dfprintk(MOUNT, "NFS: mount options string was NULL.\n"); @@ -1091,113 +1086,82 @@ static int nfs_parse_mount_options(char *raw, */ case Opt_port: if (match_int(args, &option) || - option < 0 || option > USHORT_MAX) { - errors++; - nfs_parse_invalid_value("port"); - } else - mnt->nfs_server.port = option; + option < 0 || option > USHORT_MAX) + goto out_invalid_value; + mnt->nfs_server.port = option; break; case Opt_rsize: - if (match_int(args, &option) || option < 0) { - errors++; - nfs_parse_invalid_value("rsize"); - } else - mnt->rsize = option; + if (match_int(args, &option) || option < 0) + goto out_invalid_value; + mnt->rsize = option; break; case Opt_wsize: - if (match_int(args, &option) || option < 0) { - errors++; - nfs_parse_invalid_value("wsize"); - } else - mnt->wsize = option; + if (match_int(args, &option) || option < 0) + goto out_invalid_value; + mnt->wsize = option; break; case Opt_bsize: - if (match_int(args, &option) || option < 0) { - errors++; - nfs_parse_invalid_value("bsize"); - } else - mnt->bsize = option; + if (match_int(args, &option) || option < 0) + goto out_invalid_value; + mnt->bsize = option; break; case Opt_timeo: - if (match_int(args, &option) || option <= 0) { - errors++; - nfs_parse_invalid_value("timeo"); - } else - mnt->timeo = option; + if (match_int(args, &option) || option <= 0) + goto out_invalid_value; + mnt->timeo = option; break; case Opt_retrans: - if (match_int(args, &option) || option <= 0) { - errors++; - nfs_parse_invalid_value("retrans"); - } else - mnt->retrans = option; + if (match_int(args, &option) || option <= 0) + goto out_invalid_value; + mnt->retrans = option; break; case Opt_acregmin: - if (match_int(args, &option) || option < 0) { - errors++; - nfs_parse_invalid_value("acregmin"); - } else - mnt->acregmin = option; + if (match_int(args, &option) || option < 0) + goto out_invalid_value; + mnt->acregmin = option; break; case Opt_acregmax: - if (match_int(args, &option) || option < 0) { - errors++; - nfs_parse_invalid_value("acregmax"); - } else - mnt->acregmax = option; + if (match_int(args, &option) || option < 0) + goto out_invalid_value; + mnt->acregmax = option; break; case Opt_acdirmin: - if (match_int(args, &option) || option < 0) { - errors++; - nfs_parse_invalid_value("acdirmin"); - } else - mnt->acdirmin = option; + if (match_int(args, &option) || option < 0) + goto out_invalid_value; + mnt->acdirmin = option; break; case Opt_acdirmax: - if (match_int(args, &option) || option < 0) { - errors++; - nfs_parse_invalid_value("acdirmax"); - } else - mnt->acdirmax = option; + if (match_int(args, &option) || option < 0) + goto out_invalid_value; + mnt->acdirmax = option; break; case Opt_actimeo: - if (match_int(args, &option) || option < 0) { - errors++; - nfs_parse_invalid_value("actimeo"); - } else - mnt->acregmin = mnt->acregmax = - mnt->acdirmin = mnt->acdirmax = option; + if (match_int(args, &option) || option < 0) + goto out_invalid_value; + mnt->acregmin = mnt->acregmax = + mnt->acdirmin = mnt->acdirmax = option; break; case Opt_namelen: - if (match_int(args, &option) || option < 0) { - errors++; - nfs_parse_invalid_value("namlen"); - } else - mnt->namlen = option; + if (match_int(args, &option) || option < 0) + goto out_invalid_value; + mnt->namlen = option; break; case Opt_mountport: if (match_int(args, &option) || - option < 0 || option > USHORT_MAX) { - errors++; - nfs_parse_invalid_value("mountport"); - } else - mnt->mount_server.port = option; + option < 0 || option > USHORT_MAX) + goto out_invalid_value; + mnt->mount_server.port = option; break; case Opt_mountvers: if (match_int(args, &option) || option < NFS_MNT_VERSION || - option > NFS_MNT3_VERSION) { - errors++; - nfs_parse_invalid_value("mountvers"); - } else - mnt->mount_server.version = option; + option > NFS_MNT3_VERSION) + goto out_invalid_value; + mnt->mount_server.version = option; break; case Opt_nfsvers: - if (match_int(args, &option)) { - errors++; - nfs_parse_invalid_value("nfsvers"); - break; - } + if (match_int(args, &option)) + goto out_invalid_value; switch (option) { case NFS2_VERSION: mnt->flags &= ~NFS_MOUNT_VER3; @@ -1206,8 +1170,7 @@ static int nfs_parse_mount_options(char *raw, mnt->flags |= NFS_MOUNT_VER3; break; default: - errors++; - nfs_parse_invalid_value("nfsvers"); + goto out_invalid_value; } break; @@ -1221,9 +1184,9 @@ static int nfs_parse_mount_options(char *raw, rc = nfs_parse_security_flavors(string, mnt); kfree(string); if (!rc) { - errors++; dfprintk(MOUNT, "NFS: unrecognized " "security flavor\n"); + return 0; } break; case Opt_proto: @@ -1237,23 +1200,25 @@ static int nfs_parse_mount_options(char *raw, case Opt_xprt_udp: mnt->flags &= ~NFS_MOUNT_TCP; mnt->nfs_server.protocol = XPRT_TRANSPORT_UDP; + kfree(string); break; case Opt_xprt_tcp: mnt->flags |= NFS_MOUNT_TCP; mnt->nfs_server.protocol = XPRT_TRANSPORT_TCP; + kfree(string); break; case Opt_xprt_rdma: /* vector side protocols to TCP */ mnt->flags |= NFS_MOUNT_TCP; mnt->nfs_server.protocol = XPRT_TRANSPORT_RDMA; xprt_load_transport(string); + kfree(string); break; default: - errors++; dfprintk(MOUNT, "NFS: unrecognized " "transport protocol\n"); + return 0; } - kfree(string); break; case Opt_mountproto: string = match_strdup(args); @@ -1272,9 +1237,9 @@ static int nfs_parse_mount_options(char *raw, break; case Opt_xprt_rdma: /* not used for side protocols */ default: - errors++; dfprintk(MOUNT, "NFS: unrecognized " "transport protocol\n"); + return 0; } break; case Opt_addr: @@ -1330,9 +1295,9 @@ static int nfs_parse_mount_options(char *raw, mnt->flags |= NFS_MOUNT_LOOKUP_CACHE_NONEG|NFS_MOUNT_LOOKUP_CACHE_NONE; break; default: - errors++; dfprintk(MOUNT, "NFS: invalid " "lookupcache argument\n"); + return 0; }; break; @@ -1350,20 +1315,20 @@ static int nfs_parse_mount_options(char *raw, break; default: - errors++; + invalid_option = 1; dfprintk(MOUNT, "NFS: unrecognized mount option " "'%s'\n", p); } } - if (errors > 0) { - dfprintk(MOUNT, "NFS: parsing encountered %d error%s\n", - errors, (errors == 1 ? "" : "s")); - if (!sloppy) - return 0; - } + if (!sloppy && invalid_option) + return 0; + return 1; +out_invalid_value: + printk(KERN_INFO "NFS: bad mount option value specified: %s \n", p); + return 0; out_nomem: printk(KERN_INFO "NFS: not enough memory to parse option\n"); return 0; -- cgit v1.2.3 From a5a16bae707cd5d2bc97d7bd1a30079f18113a77 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Wed, 17 Jun 2009 18:02:14 -0700 Subject: NFS: More "sloppy" parsing problems Specifying "port=-5" with the kernel's current mount option parser generates "unrecognized mount option". If "sloppy" is set, this causes the mount to succeed and use the default values; the desired behavior is that, since this is a valid option with an invalid value, the mount should fail, even with "sloppy." To properly handle "sloppy" parsing, we need to distinguish between correct options with invalid values, and incorrect options. We will need to parse integer values by hand, therefore, and not rely on match_token(). For instance, these must all fail with "invalid value": port=12345678 port=-5 port=samuel and not with "unrecognized option," as they do currently. Thus, for the sake of match_token() we need to treat the values for these options as strings, and do the conversion to integers using strict_strtol(). This is basically the same solution we used for the earlier "retry=" fix (commit ecbb3845), except in this case the kernel actually has to parse the value, rather than ignore it. Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- fs/nfs/super.c | 142 +++++++++++++++++++++++++++++++++++++++++++-------------- 1 file changed, 108 insertions(+), 34 deletions(-) (limited to 'fs') diff --git a/fs/nfs/super.c b/fs/nfs/super.c index a2b2805caf9d..b798ed1bd36d 100644 --- a/fs/nfs/super.c +++ b/fs/nfs/super.c @@ -139,22 +139,22 @@ static const match_table_t nfs_mount_option_tokens = { { Opt_fscache_uniq, "fsc=%s" }, { Opt_nofscache, "nofsc" }, - { Opt_port, "port=%u" }, - { Opt_rsize, "rsize=%u" }, - { Opt_wsize, "wsize=%u" }, - { Opt_bsize, "bsize=%u" }, - { Opt_timeo, "timeo=%u" }, - { Opt_retrans, "retrans=%u" }, - { Opt_acregmin, "acregmin=%u" }, - { Opt_acregmax, "acregmax=%u" }, - { Opt_acdirmin, "acdirmin=%u" }, - { Opt_acdirmax, "acdirmax=%u" }, - { Opt_actimeo, "actimeo=%u" }, - { Opt_namelen, "namlen=%u" }, - { Opt_mountport, "mountport=%u" }, - { Opt_mountvers, "mountvers=%u" }, - { Opt_nfsvers, "nfsvers=%u" }, - { Opt_nfsvers, "vers=%u" }, + { Opt_port, "port=%s" }, + { Opt_rsize, "rsize=%s" }, + { Opt_wsize, "wsize=%s" }, + { Opt_bsize, "bsize=%s" }, + { Opt_timeo, "timeo=%s" }, + { Opt_retrans, "retrans=%s" }, + { Opt_acregmin, "acregmin=%s" }, + { Opt_acregmax, "acregmax=%s" }, + { Opt_acdirmin, "acdirmin=%s" }, + { Opt_acdirmax, "acdirmax=%s" }, + { Opt_actimeo, "actimeo=%s" }, + { Opt_namelen, "namlen=%s" }, + { Opt_mountport, "mountport=%s" }, + { Opt_mountvers, "mountvers=%s" }, + { Opt_nfsvers, "nfsvers=%s" }, + { Opt_nfsvers, "vers=%s" }, { Opt_sec, "sec=%s" }, { Opt_proto, "proto=%s" }, @@ -976,7 +976,8 @@ static int nfs_parse_mount_options(char *raw, while ((p = strsep(&raw, ",")) != NULL) { substring_t args[MAX_OPT_ARGS]; - int option, token; + unsigned long option; + int token; if (!*p) continue; @@ -1085,82 +1086,155 @@ static int nfs_parse_mount_options(char *raw, * options that take numeric values */ case Opt_port: - if (match_int(args, &option) || - option < 0 || option > USHORT_MAX) + string = match_strdup(args); + if (string == NULL) + goto out_nomem; + rc = strict_strtoul(string, 10, &option); + kfree(string); + if (rc != 0 || option > USHORT_MAX) goto out_invalid_value; mnt->nfs_server.port = option; break; case Opt_rsize: - if (match_int(args, &option) || option < 0) + string = match_strdup(args); + if (string == NULL) + goto out_nomem; + rc = strict_strtoul(string, 10, &option); + kfree(string); + if (rc != 0) goto out_invalid_value; mnt->rsize = option; break; case Opt_wsize: - if (match_int(args, &option) || option < 0) + string = match_strdup(args); + if (string == NULL) + goto out_nomem; + rc = strict_strtoul(string, 10, &option); + kfree(string); + if (rc != 0) goto out_invalid_value; mnt->wsize = option; break; case Opt_bsize: - if (match_int(args, &option) || option < 0) + string = match_strdup(args); + if (string == NULL) + goto out_nomem; + rc = strict_strtoul(string, 10, &option); + kfree(string); + if (rc != 0) goto out_invalid_value; mnt->bsize = option; break; case Opt_timeo: - if (match_int(args, &option) || option <= 0) + string = match_strdup(args); + if (string == NULL) + goto out_nomem; + rc = strict_strtoul(string, 10, &option); + kfree(string); + if (rc != 0 || option == 0) goto out_invalid_value; mnt->timeo = option; break; case Opt_retrans: - if (match_int(args, &option) || option <= 0) + string = match_strdup(args); + if (string == NULL) + goto out_nomem; + rc = strict_strtoul(string, 10, &option); + kfree(string); + if (rc != 0 || option == 0) goto out_invalid_value; mnt->retrans = option; break; case Opt_acregmin: - if (match_int(args, &option) || option < 0) + string = match_strdup(args); + if (string == NULL) + goto out_nomem; + rc = strict_strtoul(string, 10, &option); + kfree(string); + if (rc != 0) goto out_invalid_value; mnt->acregmin = option; break; case Opt_acregmax: - if (match_int(args, &option) || option < 0) + string = match_strdup(args); + if (string == NULL) + goto out_nomem; + rc = strict_strtoul(string, 10, &option); + kfree(string); + if (rc != 0) goto out_invalid_value; mnt->acregmax = option; break; case Opt_acdirmin: - if (match_int(args, &option) || option < 0) + string = match_strdup(args); + if (string == NULL) + goto out_nomem; + rc = strict_strtoul(string, 10, &option); + kfree(string); + if (rc != 0) goto out_invalid_value; mnt->acdirmin = option; break; case Opt_acdirmax: - if (match_int(args, &option) || option < 0) + string = match_strdup(args); + if (string == NULL) + goto out_nomem; + rc = strict_strtoul(string, 10, &option); + kfree(string); + if (rc != 0) goto out_invalid_value; mnt->acdirmax = option; break; case Opt_actimeo: - if (match_int(args, &option) || option < 0) + string = match_strdup(args); + if (string == NULL) + goto out_nomem; + rc = strict_strtoul(string, 10, &option); + kfree(string); + if (rc != 0) goto out_invalid_value; mnt->acregmin = mnt->acregmax = mnt->acdirmin = mnt->acdirmax = option; break; case Opt_namelen: - if (match_int(args, &option) || option < 0) + string = match_strdup(args); + if (string == NULL) + goto out_nomem; + rc = strict_strtoul(string, 10, &option); + kfree(string); + if (rc != 0) goto out_invalid_value; mnt->namlen = option; break; case Opt_mountport: - if (match_int(args, &option) || - option < 0 || option > USHORT_MAX) + string = match_strdup(args); + if (string == NULL) + goto out_nomem; + rc = strict_strtoul(string, 10, &option); + kfree(string); + if (rc != 0 || option > USHORT_MAX) goto out_invalid_value; mnt->mount_server.port = option; break; case Opt_mountvers: - if (match_int(args, &option) || + string = match_strdup(args); + if (string == NULL) + goto out_nomem; + rc = strict_strtoul(string, 10, &option); + kfree(string); + if (rc != 0 || option < NFS_MNT_VERSION || option > NFS_MNT3_VERSION) goto out_invalid_value; mnt->mount_server.version = option; break; case Opt_nfsvers: - if (match_int(args, &option)) + string = match_strdup(args); + if (string == NULL) + goto out_nomem; + rc = strict_strtoul(string, 10, &option); + kfree(string); + if (rc != 0) goto out_invalid_value; switch (option) { case NFS2_VERSION: -- cgit v1.2.3 From 4bf259e3ae5015e73282ba66716c4a917e1264ac Mon Sep 17 00:00:00 2001 From: James Morris Date: Wed, 17 Jun 2009 18:02:14 -0700 Subject: nfs: remove unnecessary NFS_INO_INVALID_ACL checks Unless I'm mistaken, NFS_INO_INVALID_ACL is being checked twice during getacl calls (i.e. first via nfs_revalidate_inode() and then by each all site). Signed-off-by: Trond Myklebust --- fs/nfs/nfs3acl.c | 2 -- fs/nfs/nfs4proc.c | 2 -- 2 files changed, 4 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs3acl.c b/fs/nfs/nfs3acl.c index 6bbf0e6daad2..bac60515a4b3 100644 --- a/fs/nfs/nfs3acl.c +++ b/fs/nfs/nfs3acl.c @@ -207,8 +207,6 @@ struct posix_acl *nfs3_proc_getacl(struct inode *inode, int type) status = nfs_revalidate_inode(server, inode); if (status < 0) return ERR_PTR(status); - if (NFS_I(inode)->cache_validity & NFS_INO_INVALID_ACL) - nfs_zap_acl_cache(inode); acl = nfs3_get_cached_acl(inode, type); if (acl != ERR_PTR(-EAGAIN)) return acl; diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 55314e721632..42e6a3b05498 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -2824,8 +2824,6 @@ static ssize_t nfs4_proc_get_acl(struct inode *inode, void *buf, size_t buflen) ret = nfs_revalidate_inode(server, inode); if (ret < 0) return ret; - if (NFS_I(inode)->cache_validity & NFS_INO_INVALID_ACL) - nfs_zap_acl_cache(inode); ret = nfs4_read_cached_acl(inode, buf, buflen); if (ret != -ENOENT) return ret; -- cgit v1.2.3 From 24a5d59f3477bcff4c069ff4d0ca9a3e037d0235 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Thu, 18 Jun 2009 12:33:16 +0200 Subject: udf: Use device size when drive reported bogus number of written blocks Some drives report 0 as the number of written blocks when there are some blocks recorded. Use device size in such case so that we can automagically mount such media. Signed-off-by: Jan Kara --- fs/udf/lowlevel.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/udf/lowlevel.c b/fs/udf/lowlevel.c index 703843f30ffd..1b88fd5df05d 100644 --- a/fs/udf/lowlevel.c +++ b/fs/udf/lowlevel.c @@ -56,7 +56,12 @@ unsigned long udf_get_last_block(struct super_block *sb) struct block_device *bdev = sb->s_bdev; unsigned long lblock = 0; - if (ioctl_by_bdev(bdev, CDROM_LAST_WRITTEN, (unsigned long) &lblock)) + /* + * ioctl failed or returned obviously bogus value? + * Try using the device size... + */ + if (ioctl_by_bdev(bdev, CDROM_LAST_WRITTEN, (unsigned long) &lblock) || + lblock == 0) lblock = bdev->bd_inode->i_size >> sb->s_blocksize_bits; if (lblock) -- cgit v1.2.3 From d3a9262e59f7fb83c6d44df3b2b1460ed57d3ea1 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Thu, 18 Jun 2009 12:54:00 +0200 Subject: fs: Provide empty .set_page_dirty() aop for anon inodes .set_page_dirty() is one of those a_ops that defaults to the buffer implementation when not set. Therefore provide a dummy function to make it do nothing. (Uncovered by perfcounters fd's which can now be writable-mmap-ed.) Signed-off-by: Peter Zijlstra Cc: Al Viro Cc: Davide Libenzi LKML-Reference: Signed-off-by: Ingo Molnar --- fs/anon_inodes.c | 15 +++++++++++++++ 1 file changed, 15 insertions(+) (limited to 'fs') diff --git a/fs/anon_inodes.c b/fs/anon_inodes.c index 1dd96d4406c0..47d4a01c5393 100644 --- a/fs/anon_inodes.c +++ b/fs/anon_inodes.c @@ -52,6 +52,19 @@ static const struct dentry_operations anon_inodefs_dentry_operations = { .d_delete = anon_inodefs_delete_dentry, }; +/* + * nop .set_page_dirty method so that people can use .page_mkwrite on + * anon inodes. + */ +static int anon_set_page_dirty(struct page *page) +{ + return 0; +}; + +static const struct address_space_operations anon_aops = { + .set_page_dirty = anon_set_page_dirty, +}; + /** * anon_inode_getfd - creates a new file instance by hooking it up to an * anonymous inode, and a dentry that describe the "class" @@ -151,6 +164,8 @@ static struct inode *anon_inode_mkinode(void) inode->i_fop = &anon_inode_fops; + inode->i_mapping->a_ops = &anon_aops; + /* * Mark the inode dirty from the very beginning, * that way it will never be moved to the dirty -- cgit v1.2.3 From 82e12fe9244ff653f703722a8937b595e10e71f4 Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Tue, 16 Jun 2009 11:03:07 +1000 Subject: nfsd: don't take nfsd_mutex twice when setting number of threads. Currently when we write a number to 'threads' in nfsdfs, we take the nfsd_mutex, update the number of threads, then take the mutex again to read the number of threads. Mostly this isn't a big deal. However if we are write '0', and portmap happens to be dead, then we can get unpredictable behaviour. If the nfsd threads all got killed quickly and the last thread is waiting for portmap to respond, then the second time we take the mutex we will block waiting for the last thread. However if the nfsd threads didn't die quite that fast, then there will be no contention when we try to take the mutex again. Unpredictability isn't fun, and waiting for the last thread to exit is pointless, so avoid taking the lock twice. To achieve this, get nfsd_svc return a non-negative number of active threads when not returning a negative error. Signed-off-by: NeilBrown --- fs/nfsd/nfsctl.c | 14 +++++++++----- fs/nfsd/nfssvc.c | 6 ++++++ 2 files changed, 15 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c index 877e713a0fd6..1250fb978ac1 100644 --- a/fs/nfsd/nfsctl.c +++ b/fs/nfsd/nfsctl.c @@ -207,10 +207,14 @@ static struct file_operations pool_stats_operations = { static ssize_t write_svc(struct file *file, char *buf, size_t size) { struct nfsctl_svc *data; + int err; if (size < sizeof(*data)) return -EINVAL; data = (struct nfsctl_svc*) buf; - return nfsd_svc(data->svc_port, data->svc_nthreads); + err = nfsd_svc(data->svc_port, data->svc_nthreads); + if (err < 0) + return err; + return 0; } /** @@ -692,12 +696,12 @@ static ssize_t write_threads(struct file *file, char *buf, size_t size) if (newthreads < 0) return -EINVAL; rv = nfsd_svc(NFS_PORT, newthreads); - if (rv) + if (rv < 0) return rv; - } + } else + rv = nfsd_nrthreads(); - return scnprintf(buf, SIMPLE_TRANSACTION_LIMIT, "%d\n", - nfsd_nrthreads()); + return scnprintf(buf, SIMPLE_TRANSACTION_LIMIT, "%d\n", rv); } /** diff --git a/fs/nfsd/nfssvc.c b/fs/nfsd/nfssvc.c index cbba4a935786..209eaa0885d1 100644 --- a/fs/nfsd/nfssvc.c +++ b/fs/nfsd/nfssvc.c @@ -413,6 +413,12 @@ nfsd_svc(unsigned short port, int nrservs) goto failure; error = svc_set_num_threads(nfsd_serv, NULL, nrservs); + if (error == 0) + /* We are holding a reference to nfsd_serv which + * we don't want to count in the return value, + * so subtract 1 + */ + error = nfsd_serv->sv_nrthreads - 1; failure: svc_destroy(nfsd_serv); /* Release server */ out: -- cgit v1.2.3 From 671e1fcf63fd115eabcb693b06cbc2e4a3d1a3a3 Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Tue, 16 Jun 2009 11:03:20 +1000 Subject: nfsd: optimise the starting of zero threads when none are running. Currently, if we ask to set then number of nfsd threads to zero when there are none running, we set up all the sockets and register the service, and then tear it all down again. This is pointless. So detect that case and exit promptly. (also remove an assignment to 'error' which was never used. Signed-off-by: NeilBrown Acked-by: Jeff Layton --- fs/nfsd/nfssvc.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfssvc.c b/fs/nfsd/nfssvc.c index 209eaa0885d1..d4c9884cd54b 100644 --- a/fs/nfsd/nfssvc.c +++ b/fs/nfsd/nfssvc.c @@ -390,12 +390,14 @@ nfsd_svc(unsigned short port, int nrservs) mutex_lock(&nfsd_mutex); dprintk("nfsd: creating service\n"); - error = -EINVAL; if (nrservs <= 0) nrservs = 0; if (nrservs > NFSD_MAXSERVS) nrservs = NFSD_MAXSERVS; - + error = 0; + if (nrservs == 0 && nfsd_serv == NULL) + goto out; + /* Readahead param cache - will no-op if it already exists */ error = nfsd_racache_init(2*nrservs); if (error<0) -- cgit v1.2.3 From c78a87d0a1fc885dfdbe21fd5e07787691dfb068 Mon Sep 17 00:00:00 2001 From: David Teigland Date: Thu, 18 Jun 2009 13:20:24 -0500 Subject: dlm: fix plock use-after-free Fix a regression from the original addition of nfs lock support 586759f03e2e9031ac5589912a51a909ed53c30a. When a synchronous (non-nfs) plock completes, the waiting thread will wake up and free the op struct. This races with the user thread in dev_write() which goes on to read the op's callback field to check if the lock is async and needs a callback. This check can happen on the freed op. The fix is to note the callback value before the op can be freed. Signed-off-by: David Teigland --- fs/dlm/plock.c | 17 ++++++++++------- 1 file changed, 10 insertions(+), 7 deletions(-) (limited to 'fs') diff --git a/fs/dlm/plock.c b/fs/dlm/plock.c index 894a32d438d5..16f682e26c07 100644 --- a/fs/dlm/plock.c +++ b/fs/dlm/plock.c @@ -353,7 +353,7 @@ static ssize_t dev_write(struct file *file, const char __user *u, size_t count, { struct dlm_plock_info info; struct plock_op *op; - int found = 0; + int found = 0, do_callback = 0; if (count != sizeof(info)) return -EINVAL; @@ -366,21 +366,24 @@ static ssize_t dev_write(struct file *file, const char __user *u, size_t count, spin_lock(&ops_lock); list_for_each_entry(op, &recv_list, list) { - if (op->info.fsid == info.fsid && op->info.number == info.number && + if (op->info.fsid == info.fsid && + op->info.number == info.number && op->info.owner == info.owner) { + struct plock_xop *xop = (struct plock_xop *)op; list_del_init(&op->list); - found = 1; - op->done = 1; memcpy(&op->info, &info, sizeof(info)); + if (xop->callback) + do_callback = 1; + else + op->done = 1; + found = 1; break; } } spin_unlock(&ops_lock); if (found) { - struct plock_xop *xop; - xop = (struct plock_xop *)op; - if (xop->callback) + if (do_callback) dlm_plock_callback(op); else wake_up(&recv_wq); -- cgit v1.2.3 From d3d64df21d3d0de675a0d3ffa7c10514f3644b30 Mon Sep 17 00:00:00 2001 From: Keika Kobayashi Date: Wed, 17 Jun 2009 16:25:55 -0700 Subject: proc: export statistics for softirq to /proc Export statistics for softirq in /proc/softirqs and /proc/stat. 1. /proc/softirqs Implement /proc/softirqs which shows the number of softirq for each CPU like /proc/interrupts. 2. /proc/stat Add the "softirq" line to /proc/stat. This line shows the number of softirq for all cpu. The first column is the total of all softirqs and each subsequent column is the total for particular softirq. [kosaki.motohiro@jp.fujitsu.com: remove redundant for_each_possible_cpu() loop] Signed-off-by: Keika Kobayashi Reviewed-by: Hiroshi Shimamoto Cc: KOSAKI Motohiro Cc: Ingo Molnar Cc: Eric Dumazet Cc: Alexey Dobriyan Signed-off-by: KOSAKI Motohiro Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- Documentation/filesystems/proc.txt | 26 ++++++++++++++++++++++ fs/proc/Makefile | 1 + fs/proc/softirqs.c | 44 ++++++++++++++++++++++++++++++++++++++ fs/proc/stat.c | 15 +++++++++++++ 4 files changed, 86 insertions(+) create mode 100644 fs/proc/softirqs.c (limited to 'fs') diff --git a/Documentation/filesystems/proc.txt b/Documentation/filesystems/proc.txt index ebff3c10a07f..fb7d649437af 100644 --- a/Documentation/filesystems/proc.txt +++ b/Documentation/filesystems/proc.txt @@ -283,6 +283,7 @@ Table 1-4: Kernel info in /proc rtc Real time clock scsi SCSI info (see text) slabinfo Slab pool info + softirqs softirq usage stat Overall statistics swaps Swap space utilization sys See chapter 2 @@ -597,6 +598,25 @@ on the kind of area : 0xffffffffa0017000-0xffffffffa0022000 45056 sys_init_module+0xc27/0x1d00 ... pages=10 vmalloc N0=10 +.............................................................................. + +softirqs: + +Provides counts of softirq handlers serviced since boot time, for each cpu. + +> cat /proc/softirqs + CPU0 CPU1 CPU2 CPU3 + HI: 0 0 0 0 + TIMER: 27166 27120 27097 27034 + NET_TX: 0 0 0 17 + NET_RX: 42 0 0 39 + BLOCK: 0 0 107 1121 + TASKLET: 0 0 0 290 + SCHED: 27035 26983 26971 26746 + HRTIMER: 0 0 0 0 + RCU: 1678 1769 2178 2250 + + 1.3 IDE devices in /proc/ide ---------------------------- @@ -883,6 +903,7 @@ since the system first booted. For a quick look, simply cat the file: processes 2915 procs_running 1 procs_blocked 0 + softirq 183433 0 21755 12 39 1137 231 21459 2263 The very first "cpu" line aggregates the numbers in all of the other "cpuN" lines. These numbers identify the amount of time the CPU has spent performing @@ -918,6 +939,11 @@ CPUs. The "procs_blocked" line gives the number of processes currently blocked, waiting for I/O to complete. +The "softirq" line gives counts of softirqs serviced since boot time, for each +of the possible system softirqs. The first column is the total of all +softirqs serviced; each subsequent column is the total for that particular +softirq. + 1.9 Ext4 file system parameters ------------------------------ diff --git a/fs/proc/Makefile b/fs/proc/Makefile index 63d965193b22..11a7b5c68153 100644 --- a/fs/proc/Makefile +++ b/fs/proc/Makefile @@ -18,6 +18,7 @@ proc-y += meminfo.o proc-y += stat.o proc-y += uptime.o proc-y += version.o +proc-y += softirqs.o proc-$(CONFIG_PROC_SYSCTL) += proc_sysctl.o proc-$(CONFIG_NET) += proc_net.o proc-$(CONFIG_PROC_KCORE) += kcore.o diff --git a/fs/proc/softirqs.c b/fs/proc/softirqs.c new file mode 100644 index 000000000000..1807c2419f17 --- /dev/null +++ b/fs/proc/softirqs.c @@ -0,0 +1,44 @@ +#include +#include +#include +#include + +/* + * /proc/softirqs ... display the number of softirqs + */ +static int show_softirqs(struct seq_file *p, void *v) +{ + int i, j; + + seq_printf(p, " "); + for_each_possible_cpu(i) + seq_printf(p, "CPU%-8d", i); + seq_printf(p, "\n"); + + for (i = 0; i < NR_SOFTIRQS; i++) { + seq_printf(p, "%8s:", softirq_to_name[i]); + for_each_possible_cpu(j) + seq_printf(p, " %10u", kstat_softirqs_cpu(i, j)); + seq_printf(p, "\n"); + } + return 0; +} + +static int softirqs_open(struct inode *inode, struct file *file) +{ + return single_open(file, show_softirqs, NULL); +} + +static const struct file_operations proc_softirqs_operations = { + .open = softirqs_open, + .read = seq_read, + .llseek = seq_lseek, + .release = single_release, +}; + +static int __init proc_softirqs_init(void) +{ + proc_create("softirqs", 0, NULL, &proc_softirqs_operations); + return 0; +} +module_init(proc_softirqs_init); diff --git a/fs/proc/stat.c b/fs/proc/stat.c index 81e4eb60972e..7cc726c6d70a 100644 --- a/fs/proc/stat.c +++ b/fs/proc/stat.c @@ -29,6 +29,8 @@ static int show_stat(struct seq_file *p, void *v) cputime64_t user, nice, system, idle, iowait, irq, softirq, steal; cputime64_t guest; u64 sum = 0; + u64 sum_softirq = 0; + unsigned int per_softirq_sums[NR_SOFTIRQS] = {0}; struct timespec boottime; unsigned int per_irq_sum; @@ -53,6 +55,13 @@ static int show_stat(struct seq_file *p, void *v) sum += kstat_irqs_cpu(j, i); } sum += arch_irq_stat_cpu(i); + + for (j = 0; j < NR_SOFTIRQS; j++) { + unsigned int softirq_stat = kstat_softirqs_cpu(j, i); + + per_softirq_sums[j] += softirq_stat; + sum_softirq += softirq_stat; + } } sum += arch_irq_stat(); @@ -115,6 +124,12 @@ static int show_stat(struct seq_file *p, void *v) nr_running(), nr_iowait()); + seq_printf(p, "softirq %llu", (unsigned long long)sum_softirq); + + for (i = 0; i < NR_SOFTIRQS; i++) + seq_printf(p, " %u", per_softirq_sums[i]); + seq_printf(p, "\n"); + return 0; } -- cgit v1.2.3 From 3fe4a975d662f11037cb710f8b4b158a3e38f9c0 Mon Sep 17 00:00:00 2001 From: Davide Libenzi Date: Wed, 17 Jun 2009 16:25:58 -0700 Subject: epoll: fix nested calls support This fixes a regression in 2.6.30. I unfortunately accepted a patch time ago, to drop the "current" usage from possible IRQ context, w/out proper thought over it. The patch switched to using the CPU id by bounding the nested call callback with a get_cpu()/put_cpu(). Unfortunately the ep_call_nested() function can be called with a callback that grabs sleepy locks (from own f_op->poll()), that results in epic fails. The following patch uses the proper "context" depending on the path where it is called, and on the kind of callback. This has been reported by Stefan Richter, that has also verified the patch is his previously failing environment. Signed-off-by: Davide Libenzi Reported-by: Stefan Richter Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/eventpoll.c | 21 ++++++++++++--------- 1 file changed, 12 insertions(+), 9 deletions(-) (limited to 'fs') diff --git a/fs/eventpoll.c b/fs/eventpoll.c index 5458e80fc558..085c5c063420 100644 --- a/fs/eventpoll.c +++ b/fs/eventpoll.c @@ -98,7 +98,7 @@ struct epoll_filefd { struct nested_call_node { struct list_head llink; void *cookie; - int cpu; + void *ctx; }; /* @@ -317,17 +317,17 @@ static void ep_nested_calls_init(struct nested_calls *ncalls) * @nproc: Nested call core function pointer. * @priv: Opaque data to be passed to the @nproc callback. * @cookie: Cookie to be used to identify this nested call. + * @ctx: This instance context. * * Returns: Returns the code returned by the @nproc callback, or -1 if * the maximum recursion limit has been exceeded. */ static int ep_call_nested(struct nested_calls *ncalls, int max_nests, int (*nproc)(void *, void *, int), void *priv, - void *cookie) + void *cookie, void *ctx) { int error, call_nests = 0; unsigned long flags; - int this_cpu = get_cpu(); struct list_head *lsthead = &ncalls->tasks_call_list; struct nested_call_node *tncur; struct nested_call_node tnode; @@ -340,7 +340,7 @@ static int ep_call_nested(struct nested_calls *ncalls, int max_nests, * very much limited. */ list_for_each_entry(tncur, lsthead, llink) { - if (tncur->cpu == this_cpu && + if (tncur->ctx == ctx && (tncur->cookie == cookie || ++call_nests > max_nests)) { /* * Ops ... loop detected or maximum nest level reached. @@ -352,7 +352,7 @@ static int ep_call_nested(struct nested_calls *ncalls, int max_nests, } /* Add the current task and cookie to the list */ - tnode.cpu = this_cpu; + tnode.ctx = ctx; tnode.cookie = cookie; list_add(&tnode.llink, lsthead); @@ -364,10 +364,9 @@ static int ep_call_nested(struct nested_calls *ncalls, int max_nests, /* Remove the current task from the list */ spin_lock_irqsave(&ncalls->lock, flags); list_del(&tnode.llink); - out_unlock: +out_unlock: spin_unlock_irqrestore(&ncalls->lock, flags); - put_cpu(); return error; } @@ -408,8 +407,12 @@ static int ep_poll_wakeup_proc(void *priv, void *cookie, int call_nests) */ static void ep_poll_safewake(wait_queue_head_t *wq) { + int this_cpu = get_cpu(); + ep_call_nested(&poll_safewake_ncalls, EP_MAX_NESTS, - ep_poll_wakeup_proc, NULL, wq); + ep_poll_wakeup_proc, NULL, wq, (void *) (long) this_cpu); + + put_cpu(); } /* @@ -663,7 +666,7 @@ static unsigned int ep_eventpoll_poll(struct file *file, poll_table *wait) * could re-enter here. */ pollflags = ep_call_nested(&poll_readywalk_ncalls, EP_MAX_NESTS, - ep_poll_readyevents_proc, ep, ep); + ep_poll_readyevents_proc, ep, ep, current); return pollflags != -1 ? pollflags : 0; } -- cgit v1.2.3 From bcac2b1b7d67f4e7c001b755409fafb37cb0d888 Mon Sep 17 00:00:00 2001 From: Michal Simek Date: Wed, 17 Jun 2009 16:25:59 -0700 Subject: procfs: remove sparse errors in proc_devtree.c CHECK fs/proc/proc_devtree.c fs/proc/proc_devtree.c:197:14: warning: Using plain integer as NULL pointer fs/proc/proc_devtree.c:203:34: warning: Using plain integer as NULL pointer fs/proc/proc_devtree.c:210:14: warning: Using plain integer as NULL pointer fs/proc/proc_devtree.c:223:26: warning: Using plain integer as NULL pointer fs/proc/proc_devtree.c:226:14: warning: Using plain integer as NULL pointer Signed-off-by: Michal Simek Cc: Alexey Dobriyan Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/proc/proc_devtree.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/proc/proc_devtree.c b/fs/proc/proc_devtree.c index fc6c3025befd..7ba79a54948c 100644 --- a/fs/proc/proc_devtree.c +++ b/fs/proc/proc_devtree.c @@ -195,20 +195,20 @@ void proc_device_tree_add_node(struct device_node *np, p = fixup_name(np, de, p); ent = proc_mkdir(p, de); - if (ent == 0) + if (ent == NULL) break; proc_device_tree_add_node(child, ent); } of_node_put(child); - for (pp = np->properties; pp != 0; pp = pp->next) { + for (pp = np->properties; pp != NULL; pp = pp->next) { p = pp->name; if (duplicate_name(de, p)) p = fixup_name(np, de, p); ent = __proc_device_tree_add_prop(de, pp, p); - if (ent == 0) + if (ent == NULL) break; } } @@ -221,10 +221,10 @@ void __init proc_device_tree_init(void) struct device_node *root; proc_device_tree = proc_mkdir("device-tree", NULL); - if (proc_device_tree == 0) + if (proc_device_tree == NULL) return; root = of_find_node_by_path("/"); - if (root == 0) { + if (root == NULL) { printk(KERN_ERR "/proc/device-tree: can't find root\n"); return; } -- cgit v1.2.3 From 2f6d311080c36e30a5fa87adca550dc6b51dbfdc Mon Sep 17 00:00:00 2001 From: Cyrill Gorcunov Date: Wed, 17 Jun 2009 16:26:00 -0700 Subject: proc: vmcore - use kzalloc in get_new_element() Instead of kmalloc+memset better use straight kzalloc Signed-off-by: Cyrill Gorcunov Reviewed-by: WANG Cong Acked-by: Vivek Goyal Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/proc/vmcore.c | 7 +------ 1 file changed, 1 insertion(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/proc/vmcore.c b/fs/proc/vmcore.c index 5edcc3f92ba7..0872afa58d39 100644 --- a/fs/proc/vmcore.c +++ b/fs/proc/vmcore.c @@ -166,12 +166,7 @@ static const struct file_operations proc_vmcore_operations = { static struct vmcore* __init get_new_element(void) { - struct vmcore *p; - - p = kmalloc(sizeof(*p), GFP_KERNEL); - if (p) - memset(p, 0, sizeof(*p)); - return p; + return kzalloc(sizeof(struct vmcore), GFP_KERNEL); } static u64 __init get_vmcore_size_elf64(char *elfptr) -- cgit v1.2.3 From 39fe7557b4d6ab82bafaa7b92b98b806afe6ad0d Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Wed, 17 Jun 2009 16:26:20 -0700 Subject: ext2: Do not update mtime of a moved directory MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit One of our users is complaining that his backup tool is upset on ext2 (while it's happy on ext3, xfs, ...) because of the mtime change. The problem is: mkdir foo mkdir bar mkdir foo/a Now under ext2: mv foo/a foo/b changes mtime of 'foo/a' (foo/b after the move). That does not really make sense and it does not happen under any other filesystem I've seen. More complicated is: mv foo/a bar/a This changes mtime of foo/a (bar/a after the move) and it makes some sense since we had to update parent directory pointer of foo/a. But again, no other filesystem does this. So after some thoughts I'd vote for consistency and change ext2 to behave the same as other filesystems. Do not update mtime of a moved directory. Specs don't say anything about it (neither that it should, nor that it should not be updated) and other common filesystems (ext3, ext4, xfs, reiserfs, fat, ...) don't do it. So let's become more consistent. Spotted by ronny.pretzsch@dfs.de, initial fix by Jörn Engel. Reported-by: Cc: Cc: Jörn Engel Signed-off-by: Jan Kara Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/ext2/dir.c | 5 +++-- fs/ext2/ext2.h | 2 +- fs/ext2/namei.c | 5 +++-- 3 files changed, 7 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/ext2/dir.c b/fs/ext2/dir.c index 003500498c22..6cde970b0a1a 100644 --- a/fs/ext2/dir.c +++ b/fs/ext2/dir.c @@ -450,7 +450,7 @@ ino_t ext2_inode_by_name(struct inode *dir, struct qstr *child) /* Releases the page */ void ext2_set_link(struct inode *dir, struct ext2_dir_entry_2 *de, - struct page *page, struct inode *inode) + struct page *page, struct inode *inode, int update_times) { loff_t pos = page_offset(page) + (char *) de - (char *) page_address(page); @@ -465,7 +465,8 @@ void ext2_set_link(struct inode *dir, struct ext2_dir_entry_2 *de, ext2_set_de_type(de, inode); err = ext2_commit_chunk(page, pos, len); ext2_put_page(page); - dir->i_mtime = dir->i_ctime = CURRENT_TIME_SEC; + if (update_times) + dir->i_mtime = dir->i_ctime = CURRENT_TIME_SEC; EXT2_I(dir)->i_flags &= ~EXT2_BTREE_FL; mark_inode_dirty(dir); } diff --git a/fs/ext2/ext2.h b/fs/ext2/ext2.h index f2e5811936d0..d988a718aedb 100644 --- a/fs/ext2/ext2.h +++ b/fs/ext2/ext2.h @@ -111,7 +111,7 @@ extern struct ext2_dir_entry_2 * ext2_find_entry (struct inode *,struct qstr *, extern int ext2_delete_entry (struct ext2_dir_entry_2 *, struct page *); extern int ext2_empty_dir (struct inode *); extern struct ext2_dir_entry_2 * ext2_dotdot (struct inode *, struct page **); -extern void ext2_set_link(struct inode *, struct ext2_dir_entry_2 *, struct page *, struct inode *); +extern void ext2_set_link(struct inode *, struct ext2_dir_entry_2 *, struct page *, struct inode *, int); /* ialloc.c */ extern struct inode * ext2_new_inode (struct inode *, int); diff --git a/fs/ext2/namei.c b/fs/ext2/namei.c index 90ea17998a73..6524ecaebb7a 100644 --- a/fs/ext2/namei.c +++ b/fs/ext2/namei.c @@ -320,7 +320,7 @@ static int ext2_rename (struct inode * old_dir, struct dentry * old_dentry, if (!new_de) goto out_dir; inode_inc_link_count(old_inode); - ext2_set_link(new_dir, new_de, new_page, old_inode); + ext2_set_link(new_dir, new_de, new_page, old_inode, 1); new_inode->i_ctime = CURRENT_TIME_SEC; if (dir_de) drop_nlink(new_inode); @@ -352,7 +352,8 @@ static int ext2_rename (struct inode * old_dir, struct dentry * old_dentry, inode_dec_link_count(old_inode); if (dir_de) { - ext2_set_link(old_inode, dir_de, dir_page, new_dir); + if (old_dir != new_dir) + ext2_set_link(old_inode, dir_de, dir_page, new_dir, 0); inode_dec_link_count(old_dir); } return 0; -- cgit v1.2.3 From e8ef7aaea79a899be4d7f50e829900c0ce15e52f Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Wed, 17 Jun 2009 16:26:23 -0700 Subject: ext3: fix chain verification in ext3_get_blocks() Chain verification in ext3_get_blocks() has been hosed since it called verify_chain(chain, NULL) which always returns success. As a result readers could in theory race with truncate. On the other hand the race probably cannot happen with the current locking scheme, since by the time ext3_truncate() is called all the pages are already removed and hence get_block() shouldn't be called on such pages... Signed-off-by: Jan Kara Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/ext3/inode.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ext3/inode.c b/fs/ext3/inode.c index b0248c6d5d4c..253c2cdc8d04 100644 --- a/fs/ext3/inode.c +++ b/fs/ext3/inode.c @@ -820,7 +820,7 @@ int ext3_get_blocks_handle(handle_t *handle, struct inode *inode, while (count < maxblocks && count <= blocks_to_boundary) { ext3_fsblk_t blk; - if (!verify_chain(chain, partial)) { + if (!verify_chain(chain, chain + depth - 1)) { /* * Indirect block might be removed by * truncate while we were reading it. -- cgit v1.2.3 From 6f3f1cb21f08fbf757bbbbb0709ee515a7a7c6ad Mon Sep 17 00:00:00 2001 From: Hisashi Hifumi Date: Wed, 17 Jun 2009 16:26:23 -0700 Subject: jbd: clean up journal_try_to_free_buffers() I delete the following patch "commit 3f31fddfa26b7594b44ff2b34f9a04ba409e0f91 Author: Mingming Cao Date: Fri Jul 25 01:46:22 2008 -0700 jbd: fix race between free buffer and commit transaction This patch is no longer needed because if race between freeing buffer and committing transaction functionality occurs and dio gets error, currently dio falls back to buffered IO by the following patch. commit 6ccfa806a9cfbbf1cd43d5b6aa47ef2c0eb518fd Author: Hisashi Hifumi Date: Tue Sep 2 14:35:40 2008 -0700 VFS: fix dio write returning EIO when try_to_release_page fails Signed-off-by: Hisashi Hifumi Cc: Theodore Tso Cc: Mingming Cao Acked-by: Jan Kara Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/jbd/transaction.c | 48 ------------------------------------------------ 1 file changed, 48 deletions(-) (limited to 'fs') diff --git a/fs/jbd/transaction.c b/fs/jbd/transaction.c index ed886e6db399..73242ba7c7b1 100644 --- a/fs/jbd/transaction.c +++ b/fs/jbd/transaction.c @@ -1686,35 +1686,6 @@ out: return; } -/* - * journal_try_to_free_buffers() could race with journal_commit_transaction() - * The latter might still hold the a count on buffers when inspecting - * them on t_syncdata_list or t_locked_list. - * - * journal_try_to_free_buffers() will call this function to - * wait for the current transaction to finish syncing data buffers, before - * tryinf to free that buffer. - * - * Called with journal->j_state_lock held. - */ -static void journal_wait_for_transaction_sync_data(journal_t *journal) -{ - transaction_t *transaction = NULL; - tid_t tid; - - spin_lock(&journal->j_state_lock); - transaction = journal->j_committing_transaction; - - if (!transaction) { - spin_unlock(&journal->j_state_lock); - return; - } - - tid = transaction->t_tid; - spin_unlock(&journal->j_state_lock); - log_wait_commit(journal, tid); -} - /** * int journal_try_to_free_buffers() - try to free page buffers. * @journal: journal for operation @@ -1786,25 +1757,6 @@ int journal_try_to_free_buffers(journal_t *journal, ret = try_to_free_buffers(page); - /* - * There are a number of places where journal_try_to_free_buffers() - * could race with journal_commit_transaction(), the later still - * holds the reference to the buffers to free while processing them. - * try_to_free_buffers() failed to free those buffers. Some of the - * caller of releasepage() request page buffers to be dropped, otherwise - * treat the fail-to-free as errors (such as generic_file_direct_IO()) - * - * So, if the caller of try_to_release_page() wants the synchronous - * behaviour(i.e make sure buffers are dropped upon return), - * let's wait for the current transaction to finish flush of - * dirty data buffers, then try to free those buffers again, - * with the journal locked. - */ - if (ret == 0 && (gfp_mask & __GFP_WAIT) && (gfp_mask & __GFP_FS)) { - journal_wait_for_transaction_sync_data(journal); - ret = try_to_free_buffers(page); - } - busy: return ret; } -- cgit v1.2.3 From ef43618a47179b41e7203a624f2c7445e7da488c Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Wed, 17 Jun 2009 16:26:24 -0700 Subject: ext3: make sure inode is deleted from orphan list after truncate As Ted pointed out, it can happen that ext3_truncate() returns without removing inode from orphan list. This way we could in some rare cases (like when we get ENOMEM from an allocation in ext3_truncate called because of failed ext3_write_begin) leave the inode on orphan list and that triggers assertion failure on umount. So make ext3_truncate() always remove inode from in-memory orphan list. Cc: Theodore Ts'o Signed-off-by: Jan Kara Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/ext3/inode.c | 20 +++++++++++--------- 1 file changed, 11 insertions(+), 9 deletions(-) (limited to 'fs') diff --git a/fs/ext3/inode.c b/fs/ext3/inode.c index 253c2cdc8d04..05dea8132fc0 100644 --- a/fs/ext3/inode.c +++ b/fs/ext3/inode.c @@ -2374,7 +2374,7 @@ void ext3_truncate(struct inode *inode) struct page *page; if (!ext3_can_truncate(inode)) - return; + goto out_notrans; if (inode->i_size == 0 && ext3_should_writeback_data(inode)) ei->i_state |= EXT3_STATE_FLUSH_ON_CLOSE; @@ -2390,7 +2390,7 @@ void ext3_truncate(struct inode *inode) page = grab_cache_page(mapping, inode->i_size >> PAGE_CACHE_SHIFT); if (!page) - return; + goto out_notrans; } handle = start_transaction(inode); @@ -2401,7 +2401,7 @@ void ext3_truncate(struct inode *inode) unlock_page(page); page_cache_release(page); } - return; /* AKPM: return what? */ + goto out_notrans; } last_block = (inode->i_size + blocksize-1) @@ -2525,6 +2525,14 @@ out_stop: ext3_orphan_del(handle, inode); ext3_journal_stop(handle); + return; +out_notrans: + /* + * Delete the inode from orphan list so that it doesn't stay there + * forever and trigger assertion on umount. + */ + if (inode->i_nlink) + ext3_orphan_del(NULL, inode); } static ext3_fsblk_t ext3_get_inode_block(struct super_block *sb, @@ -3122,12 +3130,6 @@ int ext3_setattr(struct dentry *dentry, struct iattr *attr) rc = inode_setattr(inode, attr); - /* If inode_setattr's call to ext3_truncate failed to get a - * transaction handle at all, we need to clean up the in-core - * orphan list manually. */ - if (inode->i_nlink) - ext3_orphan_del(NULL, inode); - if (!rc && (ia_valid & ATTR_MODE)) rc = ext3_acl_chmod(inode); -- cgit v1.2.3 From 52b680c81238ea14693ab893d5d32a4d1c0a987d Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Wed, 17 Jun 2009 16:26:25 -0700 Subject: isofs: let mode and dmode mount options override rock ridge mode setting So far, permissions set via 'mode' and/or 'dmode' mount options were effective only if the medium had no rock ridge extensions (or was mounted without them). Add 'overriderockmode' mount option to indicate that these options should override permissions set in rock ridge extensions. Maybe this should be default but the current behavior is there since mount options were created so I think we should not change how they behave. Cc: Signed-off-by: Jan Kara Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- Documentation/filesystems/isofs.txt | 9 +++++-- fs/isofs/inode.c | 50 ++++++++++++++++++++++++++++--------- fs/isofs/isofs.h | 3 +++ 3 files changed, 48 insertions(+), 14 deletions(-) (limited to 'fs') diff --git a/Documentation/filesystems/isofs.txt b/Documentation/filesystems/isofs.txt index 6973b980ca2a..3c367c3b3608 100644 --- a/Documentation/filesystems/isofs.txt +++ b/Documentation/filesystems/isofs.txt @@ -23,8 +23,13 @@ Mount options unique to the isofs filesystem. map=off Do not map non-Rock Ridge filenames to lower case map=normal Map non-Rock Ridge filenames to lower case map=acorn As map=normal but also apply Acorn extensions if present - mode=xxx Sets the permissions on files to xxx - dmode=xxx Sets the permissions on directories to xxx + mode=xxx Sets the permissions on files to xxx unless Rock Ridge + extensions set the permissions otherwise + dmode=xxx Sets the permissions on directories to xxx unless Rock Ridge + extensions set the permissions otherwise + overriderockperm Set permissions on files and directories according to + 'mode' and 'dmode' even though Rock Ridge extensions are + present. nojoliet Ignore Joliet extensions if they are present. norock Ignore Rock Ridge extensions if they are present. hide Completely strip hidden files from the file system. diff --git a/fs/isofs/inode.c b/fs/isofs/inode.c index 068b34b5a107..8e6fd41eec12 100644 --- a/fs/isofs/inode.c +++ b/fs/isofs/inode.c @@ -148,6 +148,7 @@ struct iso9660_options{ char hide; char showassoc; char nocompress; + char overriderockperm; unsigned char check; unsigned int blocksize; mode_t fmode; @@ -312,7 +313,7 @@ enum { Opt_block, Opt_check_r, Opt_check_s, Opt_cruft, Opt_gid, Opt_ignore, Opt_iocharset, Opt_map_a, Opt_map_n, Opt_map_o, Opt_mode, Opt_nojoliet, Opt_norock, Opt_sb, Opt_session, Opt_uid, Opt_unhide, Opt_utf8, Opt_err, - Opt_nocompress, Opt_hide, Opt_showassoc, Opt_dmode, + Opt_nocompress, Opt_hide, Opt_showassoc, Opt_dmode, Opt_overriderockperm, }; static const match_table_t tokens = { @@ -340,6 +341,7 @@ static const match_table_t tokens = { {Opt_gid, "gid=%u"}, {Opt_mode, "mode=%u"}, {Opt_dmode, "dmode=%u"}, + {Opt_overriderockperm, "overriderockperm"}, {Opt_block, "block=%u"}, {Opt_ignore, "conv=binary"}, {Opt_ignore, "conv=b"}, @@ -367,16 +369,12 @@ static int parse_options(char *options, struct iso9660_options *popt) popt->check = 'u'; /* unset */ popt->nocompress = 0; popt->blocksize = 1024; - popt->fmode = popt->dmode = S_IRUGO | S_IXUGO; /* - * r-x for all. The disc could - * be shared with DOS machines so - * virtually anything could be - * a valid executable. - */ + popt->fmode = popt->dmode = ISOFS_INVALID_MODE; popt->gid = 0; popt->uid = 0; popt->iocharset = NULL; popt->utf8 = 0; + popt->overriderockperm = 0; popt->session=-1; popt->sbsector=-1; if (!options) @@ -466,6 +464,9 @@ static int parse_options(char *options, struct iso9660_options *popt) return 0; popt->dmode = option; break; + case Opt_overriderockperm: + popt->overriderockperm = 1; + break; case Opt_block: if (match_int(&args[0], &option)) return 0; @@ -811,13 +812,20 @@ root_found: sbi->s_gid = opt.gid; sbi->s_utf8 = opt.utf8; sbi->s_nocompress = opt.nocompress; + sbi->s_overriderockperm = opt.overriderockperm; /* * It would be incredibly stupid to allow people to mark every file * on the disk as suid, so we merely allow them to set the default * permissions. */ - sbi->s_fmode = opt.fmode & 0777; - sbi->s_dmode = opt.dmode & 0777; + if (opt.fmode != ISOFS_INVALID_MODE) + sbi->s_fmode = opt.fmode & 0777; + else + sbi->s_fmode = ISOFS_INVALID_MODE; + if (opt.dmode != ISOFS_INVALID_MODE) + sbi->s_dmode = opt.dmode & 0777; + else + sbi->s_dmode = ISOFS_INVALID_MODE; /* * Read the root inode, which _may_ result in changing @@ -1261,7 +1269,10 @@ static int isofs_read_inode(struct inode *inode) ei->i_file_format = isofs_file_normal; if (de->flags[-high_sierra] & 2) { - inode->i_mode = sbi->s_dmode | S_IFDIR; + if (sbi->s_dmode != ISOFS_INVALID_MODE) + inode->i_mode = S_IFDIR | sbi->s_dmode; + else + inode->i_mode = S_IFDIR | S_IRUGO | S_IXUGO; inode->i_nlink = 1; /* * Set to 1. We know there are 2, but * the find utility tries to optimize @@ -1270,8 +1281,16 @@ static int isofs_read_inode(struct inode *inode) * do it the hard way. */ } else { - /* Everybody gets to read the file. */ - inode->i_mode = sbi->s_fmode | S_IFREG; + if (sbi->s_fmode != ISOFS_INVALID_MODE) { + inode->i_mode = S_IFREG | sbi->s_fmode; + } else { + /* + * Set default permissions: r-x for all. The disc + * could be shared with DOS machines so virtually + * anything could be a valid executable. + */ + inode->i_mode = S_IFREG | S_IRUGO | S_IXUGO; + } inode->i_nlink = 1; } inode->i_uid = sbi->s_uid; @@ -1349,6 +1368,13 @@ static int isofs_read_inode(struct inode *inode) test_and_set_uid(&inode->i_uid, sbi->s_uid); test_and_set_gid(&inode->i_gid, sbi->s_gid); } + /* Now set final access rights if overriding rock ridge setting */ + if (S_ISDIR(inode->i_mode) && sbi->s_overriderockperm && + sbi->s_dmode != ISOFS_INVALID_MODE) + inode->i_mode = S_IFDIR | sbi->s_dmode; + if (S_ISREG(inode->i_mode) && sbi->s_overriderockperm && + sbi->s_fmode != ISOFS_INVALID_MODE) + inode->i_mode = S_IFREG | sbi->s_fmode; /* Install the inode operations vector */ if (S_ISREG(inode->i_mode)) { diff --git a/fs/isofs/isofs.h b/fs/isofs/isofs.h index ccbf72faf27a..9679fbcbfc03 100644 --- a/fs/isofs/isofs.h +++ b/fs/isofs/isofs.h @@ -50,6 +50,7 @@ struct isofs_sb_info { unsigned char s_nocompress; unsigned char s_hide; unsigned char s_showassoc; + unsigned char s_overriderockperm; mode_t s_fmode; mode_t s_dmode; @@ -58,6 +59,8 @@ struct isofs_sb_info { struct nls_table *s_nls_iocharset; /* Native language support table */ }; +#define ISOFS_INVALID_MODE ((mode_t) -1) + static inline struct isofs_sb_info *ISOFS_SB(struct super_block *sb) { return sb->s_fs_info; -- cgit v1.2.3 From 5c4a656b7e51503c2b5e7e7310ec326ee38a8389 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Wed, 17 Jun 2009 16:26:27 -0700 Subject: isofs: fix setting of uid and gid to 0 isofs allows setting of default uid and gid of files but value 0 was used to indicate that user did not specify any uid/gid mount option. Since this option also overrides uid/gid set in Rock Ridge extension, it makes sense to allow forcing uid/gid 0. Fix option processing to allow this. Cc: Signed-off-by: Jan Kara Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/isofs/inode.c | 26 ++++++++++++-------------- fs/isofs/isofs.h | 2 ++ 2 files changed, 14 insertions(+), 14 deletions(-) (limited to 'fs') diff --git a/fs/isofs/inode.c b/fs/isofs/inode.c index 8e6fd41eec12..0f4f9ca77f8e 100644 --- a/fs/isofs/inode.c +++ b/fs/isofs/inode.c @@ -153,6 +153,8 @@ struct iso9660_options{ unsigned int blocksize; mode_t fmode; mode_t dmode; + char uid_set; + char gid_set; gid_t gid; uid_t uid; char *iocharset; @@ -370,6 +372,8 @@ static int parse_options(char *options, struct iso9660_options *popt) popt->nocompress = 0; popt->blocksize = 1024; popt->fmode = popt->dmode = ISOFS_INVALID_MODE; + popt->uid_set = 0; + popt->gid_set = 0; popt->gid = 0; popt->uid = 0; popt->iocharset = NULL; @@ -448,11 +452,13 @@ static int parse_options(char *options, struct iso9660_options *popt) if (match_int(&args[0], &option)) return 0; popt->uid = option; + popt->uid_set = 1; break; case Opt_gid: if (match_int(&args[0], &option)) return 0; popt->gid = option; + popt->gid_set = 1; break; case Opt_mode: if (match_int(&args[0], &option)) @@ -810,6 +816,8 @@ root_found: sbi->s_showassoc = opt.showassoc; sbi->s_uid = opt.uid; sbi->s_gid = opt.gid; + sbi->s_uid_set = opt.uid_set; + sbi->s_gid_set = opt.gid_set; sbi->s_utf8 = opt.utf8; sbi->s_nocompress = opt.nocompress; sbi->s_overriderockperm = opt.overriderockperm; @@ -1103,18 +1111,6 @@ static const struct address_space_operations isofs_aops = { .bmap = _isofs_bmap }; -static inline void test_and_set_uid(uid_t *p, uid_t value) -{ - if (value) - *p = value; -} - -static inline void test_and_set_gid(gid_t *p, gid_t value) -{ - if (value) - *p = value; -} - static int isofs_read_level3_size(struct inode *inode) { unsigned long bufsize = ISOFS_BUFFER_SIZE(inode); @@ -1365,8 +1361,10 @@ static int isofs_read_inode(struct inode *inode) if (!high_sierra) { parse_rock_ridge_inode(de, inode); /* if we want uid/gid set, override the rock ridge setting */ - test_and_set_uid(&inode->i_uid, sbi->s_uid); - test_and_set_gid(&inode->i_gid, sbi->s_gid); + if (sbi->s_uid_set) + inode->i_uid = sbi->s_uid; + if (sbi->s_gid_set) + inode->i_gid = sbi->s_gid; } /* Now set final access rights if overriding rock ridge setting */ if (S_ISDIR(inode->i_mode) && sbi->s_overriderockperm && diff --git a/fs/isofs/isofs.h b/fs/isofs/isofs.h index 9679fbcbfc03..e2fc9704f14f 100644 --- a/fs/isofs/isofs.h +++ b/fs/isofs/isofs.h @@ -51,6 +51,8 @@ struct isofs_sb_info { unsigned char s_hide; unsigned char s_showassoc; unsigned char s_overriderockperm; + unsigned char s_uid_set; + unsigned char s_gid_set; mode_t s_fmode; mode_t s_dmode; -- cgit v1.2.3 From 5404ac8e4418ab3d254950ee4f9bcafc1da20b4a Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Wed, 17 Jun 2009 16:26:27 -0700 Subject: isofs: cleanup mount option processing Remove unused variables from isofs_sb_info (used to be some mount options), unify variables for option to use 0/1 (some options used 'y'/'n'), use bit fields for option flags in superblock. Signed-off-by: Jan Kara Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/isofs/dir.c | 5 ++--- fs/isofs/inode.c | 48 ++++++++++++++++++++++++------------------------ fs/isofs/isofs.h | 28 ++++++++++++---------------- fs/isofs/namei.c | 4 ++-- 4 files changed, 40 insertions(+), 45 deletions(-) (limited to 'fs') diff --git a/fs/isofs/dir.c b/fs/isofs/dir.c index 2f0dc5a14633..8ba5441063be 100644 --- a/fs/isofs/dir.c +++ b/fs/isofs/dir.c @@ -195,9 +195,8 @@ static int do_isofs_readdir(struct inode *inode, struct file *filp, * Do not report hidden files if so instructed, or associated * files unless instructed to do so */ - if ((sbi->s_hide == 'y' && - (de->flags[-sbi->s_high_sierra] & 1)) || - (sbi->s_showassoc =='n' && + if ((sbi->s_hide && (de->flags[-sbi->s_high_sierra] & 1)) || + (!sbi->s_showassoc && (de->flags[-sbi->s_high_sierra] & 4))) { filp->f_pos += de_len; continue; diff --git a/fs/isofs/inode.c b/fs/isofs/inode.c index 0f4f9ca77f8e..58a7963e168a 100644 --- a/fs/isofs/inode.c +++ b/fs/isofs/inode.c @@ -141,24 +141,24 @@ static const struct dentry_operations isofs_dentry_ops[] = { }; struct iso9660_options{ - char map; - char rock; + unsigned int rock:1; + unsigned int cruft:1; + unsigned int hide:1; + unsigned int showassoc:1; + unsigned int nocompress:1; + unsigned int overriderockperm:1; + unsigned int uid_set:1; + unsigned int gid_set:1; + unsigned int utf8:1; + unsigned char map; char joliet; - char cruft; - char hide; - char showassoc; - char nocompress; - char overriderockperm; unsigned char check; unsigned int blocksize; mode_t fmode; mode_t dmode; - char uid_set; - char gid_set; gid_t gid; uid_t uid; char *iocharset; - unsigned char utf8; /* LVE */ s32 session; s32 sbsector; @@ -363,11 +363,11 @@ static int parse_options(char *options, struct iso9660_options *popt) int option; popt->map = 'n'; - popt->rock = 'y'; - popt->joliet = 'y'; - popt->cruft = 'n'; - popt->hide = 'n'; - popt->showassoc = 'n'; + popt->rock = 1; + popt->joliet = 1; + popt->cruft = 0; + popt->hide = 0; + popt->showassoc = 0; popt->check = 'u'; /* unset */ popt->nocompress = 0; popt->blocksize = 1024; @@ -395,20 +395,20 @@ static int parse_options(char *options, struct iso9660_options *popt) token = match_token(p, tokens, args); switch (token) { case Opt_norock: - popt->rock = 'n'; + popt->rock = 0; break; case Opt_nojoliet: - popt->joliet = 'n'; + popt->joliet = 0; break; case Opt_hide: - popt->hide = 'y'; + popt->hide = 1; break; case Opt_unhide: case Opt_showassoc: - popt->showassoc = 'y'; + popt->showassoc = 1; break; case Opt_cruft: - popt->cruft = 'y'; + popt->cruft = 1; break; case Opt_utf8: popt->utf8 = 1; @@ -657,7 +657,7 @@ static int isofs_fill_super(struct super_block *s, void *data, int silent) goto out_freebh; sbi->s_high_sierra = 1; - opt.rock = 'n'; + opt.rock = 0; h_pri = (struct hs_primary_descriptor *)vdp; goto root_found; } @@ -680,7 +680,7 @@ static int isofs_fill_super(struct super_block *s, void *data, int silent) root_found: - if (joliet_level && (pri == NULL || opt.rock == 'n')) { + if (joliet_level && (pri == NULL || !opt.rock)) { /* This is the case of Joliet with the norock mount flag. * A disc with both Joliet and Rock Ridge is handled later */ @@ -809,7 +809,7 @@ root_found: s->s_op = &isofs_sops; s->s_export_op = &isofs_export_ops; sbi->s_mapping = opt.map; - sbi->s_rock = (opt.rock == 'y' ? 2 : 0); + sbi->s_rock = (opt.rock ? 2 : 0); sbi->s_rock_offset = -1; /* initial offset, will guess until SP is found*/ sbi->s_cruft = opt.cruft; sbi->s_hide = opt.hide; @@ -1315,7 +1315,7 @@ static int isofs_read_inode(struct inode *inode) * this CDROM was mounted with the cruft option. */ - if (sbi->s_cruft == 'y') + if (sbi->s_cruft) inode->i_size &= 0x00ffffff; if (de->interleave[0]) { diff --git a/fs/isofs/isofs.h b/fs/isofs/isofs.h index e2fc9704f14f..7d33de84f52a 100644 --- a/fs/isofs/isofs.h +++ b/fs/isofs/isofs.h @@ -35,24 +35,20 @@ struct isofs_sb_info { unsigned long s_log_zone_size; unsigned long s_max_size; - unsigned char s_high_sierra; /* A simple flag */ - unsigned char s_mapping; int s_rock_offset; /* offset of SUSP fields within SU area */ - unsigned char s_rock; unsigned char s_joliet_level; - unsigned char s_utf8; - unsigned char s_cruft; /* Broken disks with high - byte of length containing - junk */ - unsigned char s_unhide; - unsigned char s_nosuid; - unsigned char s_nodev; - unsigned char s_nocompress; - unsigned char s_hide; - unsigned char s_showassoc; - unsigned char s_overriderockperm; - unsigned char s_uid_set; - unsigned char s_gid_set; + unsigned char s_mapping; + unsigned int s_high_sierra:1; + unsigned int s_rock:2; + unsigned int s_utf8:1; + unsigned int s_cruft:1; /* Broken disks with high byte of length + * containing junk */ + unsigned int s_nocompress:1; + unsigned int s_hide:1; + unsigned int s_showassoc:1; + unsigned int s_overriderockperm:1; + unsigned int s_uid_set:1; + unsigned int s_gid_set:1; mode_t s_fmode; mode_t s_dmode; diff --git a/fs/isofs/namei.c b/fs/isofs/namei.c index 8299889a835e..eaa831311c9c 100644 --- a/fs/isofs/namei.c +++ b/fs/isofs/namei.c @@ -142,9 +142,9 @@ isofs_find_entry(struct inode *dir, struct dentry *dentry, */ match = 0; if (dlen > 0 && - (sbi->s_hide =='n' || + (!sbi->s_hide || (!(de->flags[-sbi->s_high_sierra] & 1))) && - (sbi->s_showassoc =='y' || + (sbi->s_showassoc || (!(de->flags[-sbi->s_high_sierra] & 4)))) { match = (isofs_cmp(dentry, dpnt, dlen) == 0); } -- cgit v1.2.3 From 37044c86baf8cb894c69bb811e35a7f6f6dbce1c Mon Sep 17 00:00:00 2001 From: Roel Kluin Date: Wed, 17 Jun 2009 16:26:28 -0700 Subject: ufs: sector_t cannot be negative unsigned i_block,fragment cannot be negative. Signed-off-by: Roel Kluin Signed-off-by: Evgeniy Dushistov Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/ufs/inode.c | 10 +--------- 1 file changed, 1 insertion(+), 9 deletions(-) (limited to 'fs') diff --git a/fs/ufs/inode.c b/fs/ufs/inode.c index 3d2512c21f05..7cf33379fd46 100644 --- a/fs/ufs/inode.c +++ b/fs/ufs/inode.c @@ -56,9 +56,7 @@ static int ufs_block_to_path(struct inode *inode, sector_t i_block, sector_t off UFSD("ptrs=uspi->s_apb = %d,double_blocks=%ld \n",ptrs,double_blocks); - if (i_block < 0) { - ufs_warning(inode->i_sb, "ufs_block_to_path", "block < 0"); - } else if (i_block < direct_blocks) { + if (i_block < direct_blocks) { offsets[n++] = i_block; } else if ((i_block -= direct_blocks) < indirect_blocks) { offsets[n++] = UFS_IND_BLOCK; @@ -440,8 +438,6 @@ int ufs_getfrag_block(struct inode *inode, sector_t fragment, struct buffer_head lock_kernel(); UFSD("ENTER, ino %lu, fragment %llu\n", inode->i_ino, (unsigned long long)fragment); - if (fragment < 0) - goto abort_negative; if (fragment > ((UFS_NDADDR + uspi->s_apb + uspi->s_2apb + uspi->s_3apb) << uspi->s_fpbshift)) @@ -504,10 +500,6 @@ abort: unlock_kernel(); return err; -abort_negative: - ufs_warning(sb, "ufs_get_block", "block < 0"); - goto abort; - abort_too_big: ufs_warning(sb, "ufs_get_block", "block > big"); goto abort; -- cgit v1.2.3 From 1d965fe0eb435b3f9a10538815f6a68de0aef03c Mon Sep 17 00:00:00 2001 From: Jeff Mahoney Date: Wed, 17 Jun 2009 16:26:29 -0700 Subject: reiserfs: fix warnings with gcc 4.4 Several code paths in reiserfs have a construct like: if (is_direntry_le_ih(ih = B_N_PITEM_HEAD(src, item_num))) ... which, in addition to being ugly, end up causing compiler warnings with gcc 4.4.0. Previous compilers didn't issue a warning. fs/reiserfs/do_balan.c:1273: warning: operation on `aux_ih' may be undefined fs/reiserfs/lbalance.c:393: warning: operation on `ih' may be undefined fs/reiserfs/lbalance.c:421: warning: operation on `ih' may be undefined fs/reiserfs/lbalance.c:777: warning: operation on `ih' may be undefined I believe this is due to the ih being passed to macros which evaluate the argument more than once. This is old code and we haven't seen any problems with it, but this patch eliminates the warnings. It converts the multiple evaluation macros to static inlines and does a preassignment for the cases that were causing the warnings because that code is just ugly. Reported-by: Chris Mason Signed-off-by: Jeff Mahoney Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/reiserfs/do_balan.c | 5 ++--- fs/reiserfs/lbalance.c | 10 ++++++---- include/linux/reiserfs_fs.h | 47 +++++++++++++++++++++++++++++++++++++-------- 3 files changed, 47 insertions(+), 15 deletions(-) (limited to 'fs') diff --git a/fs/reiserfs/do_balan.c b/fs/reiserfs/do_balan.c index 4beb964a2a3e..128d3f7c8aa5 100644 --- a/fs/reiserfs/do_balan.c +++ b/fs/reiserfs/do_balan.c @@ -1270,9 +1270,8 @@ static int balance_leaf(struct tree_balance *tb, struct item_head *ih, /* item h RFALSE(ih, "PAP-12210: ih must be 0"); - if (is_direntry_le_ih - (aux_ih = - B_N_PITEM_HEAD(tbS0, item_pos))) { + aux_ih = B_N_PITEM_HEAD(tbS0, item_pos); + if (is_direntry_le_ih(aux_ih)) { /* we append to directory item */ int entry_count; diff --git a/fs/reiserfs/lbalance.c b/fs/reiserfs/lbalance.c index 381750a155f6..03d85cbf90bf 100644 --- a/fs/reiserfs/lbalance.c +++ b/fs/reiserfs/lbalance.c @@ -390,7 +390,8 @@ static void leaf_item_bottle(struct buffer_info *dest_bi, if (last_first == FIRST_TO_LAST) { /* if ( if item in position item_num in buffer SOURCE is directory item ) */ - if (is_direntry_le_ih(ih = B_N_PITEM_HEAD(src, item_num))) + ih = B_N_PITEM_HEAD(src, item_num); + if (is_direntry_le_ih(ih)) leaf_copy_dir_entries(dest_bi, src, FIRST_TO_LAST, item_num, 0, cpy_bytes); else { @@ -418,7 +419,8 @@ static void leaf_item_bottle(struct buffer_info *dest_bi, } } else { /* if ( if item in position item_num in buffer SOURCE is directory item ) */ - if (is_direntry_le_ih(ih = B_N_PITEM_HEAD(src, item_num))) + ih = B_N_PITEM_HEAD(src, item_num); + if (is_direntry_le_ih(ih)) leaf_copy_dir_entries(dest_bi, src, LAST_TO_FIRST, item_num, I_ENTRY_COUNT(ih) - cpy_bytes, @@ -774,8 +776,8 @@ void leaf_delete_items(struct buffer_info *cur_bi, int last_first, leaf_delete_items_entirely(cur_bi, first + 1, del_num - 1); - if (is_direntry_le_ih - (ih = B_N_PITEM_HEAD(bh, B_NR_ITEMS(bh) - 1))) + ih = B_N_PITEM_HEAD(bh, B_NR_ITEMS(bh) - 1); + if (is_direntry_le_ih(ih)) /* the last item is directory */ /* len = numbers of directory entries in this item */ len = ih_entry_count(ih); diff --git a/include/linux/reiserfs_fs.h b/include/linux/reiserfs_fs.h index 2245c78d5876..dd31e7bae35c 100644 --- a/include/linux/reiserfs_fs.h +++ b/include/linux/reiserfs_fs.h @@ -660,23 +660,54 @@ static inline void set_le_key_k_type(int version, struct reiserfs_key *key, cpu_to_le32(type2uniqueness(type))) : (void)(set_offset_v2_k_type(&(key->u.k_offset_v2), type)); } + static inline void set_le_ih_k_type(struct item_head *ih, int type) { set_le_key_k_type(ih_version(ih), &(ih->ih_key), type); } -#define is_direntry_le_key(version,key) (le_key_k_type (version, key) == TYPE_DIRENTRY) -#define is_direct_le_key(version,key) (le_key_k_type (version, key) == TYPE_DIRECT) -#define is_indirect_le_key(version,key) (le_key_k_type (version, key) == TYPE_INDIRECT) -#define is_statdata_le_key(version,key) (le_key_k_type (version, key) == TYPE_STAT_DATA) +static inline int is_direntry_le_key(int version, struct reiserfs_key *key) +{ + return le_key_k_type(version, key) == TYPE_DIRENTRY; +} + +static inline int is_direct_le_key(int version, struct reiserfs_key *key) +{ + return le_key_k_type(version, key) == TYPE_DIRECT; +} + +static inline int is_indirect_le_key(int version, struct reiserfs_key *key) +{ + return le_key_k_type(version, key) == TYPE_INDIRECT; +} + +static inline int is_statdata_le_key(int version, struct reiserfs_key *key) +{ + return le_key_k_type(version, key) == TYPE_STAT_DATA; +} // // item header has version. // -#define is_direntry_le_ih(ih) is_direntry_le_key (ih_version (ih), &((ih)->ih_key)) -#define is_direct_le_ih(ih) is_direct_le_key (ih_version (ih), &((ih)->ih_key)) -#define is_indirect_le_ih(ih) is_indirect_le_key (ih_version(ih), &((ih)->ih_key)) -#define is_statdata_le_ih(ih) is_statdata_le_key (ih_version (ih), &((ih)->ih_key)) +static inline int is_direntry_le_ih(struct item_head *ih) +{ + return is_direntry_le_key(ih_version(ih), &ih->ih_key); +} + +static inline int is_direct_le_ih(struct item_head *ih) +{ + return is_direct_le_key(ih_version(ih), &ih->ih_key); +} + +static inline int is_indirect_le_ih(struct item_head *ih) +{ + return is_indirect_le_key(ih_version(ih), &ih->ih_key); +} + +static inline int is_statdata_le_ih(struct item_head *ih) +{ + return is_statdata_le_key(ih_version(ih), &ih->ih_key); +} // // key is pointer to cpu key, result is cpu -- cgit v1.2.3 From 3b34fc5880a2dcc7e5ed9837ef8d6bae051ab266 Mon Sep 17 00:00:00 2001 From: Oleg Nesterov Date: Wed, 17 Jun 2009 16:27:38 -0700 Subject: elf_core_dump: use rcu_read_lock() to access ->real_parent In theory it is not safe to dereference ->parent/real_parent without tasklist or rcu lock, we can race with re-parenting. Signed-off-by: Oleg Nesterov Acked-by: Roland McGrath Cc: David Howells Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/binfmt_elf.c | 8 ++++++-- fs/binfmt_elf_fdpic.c | 8 ++++++-- 2 files changed, 12 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/binfmt_elf.c b/fs/binfmt_elf.c index 40381df34869..9fa212b014a5 100644 --- a/fs/binfmt_elf.c +++ b/fs/binfmt_elf.c @@ -1340,8 +1340,10 @@ static void fill_prstatus(struct elf_prstatus *prstatus, prstatus->pr_info.si_signo = prstatus->pr_cursig = signr; prstatus->pr_sigpend = p->pending.signal.sig[0]; prstatus->pr_sighold = p->blocked.sig[0]; + rcu_read_lock(); + prstatus->pr_ppid = task_pid_vnr(rcu_dereference(p->real_parent)); + rcu_read_unlock(); prstatus->pr_pid = task_pid_vnr(p); - prstatus->pr_ppid = task_pid_vnr(p->real_parent); prstatus->pr_pgrp = task_pgrp_vnr(p); prstatus->pr_sid = task_session_vnr(p); if (thread_group_leader(p)) { @@ -1382,8 +1384,10 @@ static int fill_psinfo(struct elf_prpsinfo *psinfo, struct task_struct *p, psinfo->pr_psargs[i] = ' '; psinfo->pr_psargs[len] = 0; + rcu_read_lock(); + psinfo->pr_ppid = task_pid_vnr(rcu_dereference(p->real_parent)); + rcu_read_unlock(); psinfo->pr_pid = task_pid_vnr(p); - psinfo->pr_ppid = task_pid_vnr(p->real_parent); psinfo->pr_pgrp = task_pgrp_vnr(p); psinfo->pr_sid = task_session_vnr(p); diff --git a/fs/binfmt_elf_fdpic.c b/fs/binfmt_elf_fdpic.c index fdb66faa24f1..20fbeced472b 100644 --- a/fs/binfmt_elf_fdpic.c +++ b/fs/binfmt_elf_fdpic.c @@ -1387,8 +1387,10 @@ static void fill_prstatus(struct elf_prstatus *prstatus, prstatus->pr_info.si_signo = prstatus->pr_cursig = signr; prstatus->pr_sigpend = p->pending.signal.sig[0]; prstatus->pr_sighold = p->blocked.sig[0]; + rcu_read_lock(); + prstatus->pr_ppid = task_pid_vnr(rcu_dereference(p->real_parent)); + rcu_read_unlock(); prstatus->pr_pid = task_pid_vnr(p); - prstatus->pr_ppid = task_pid_vnr(p->real_parent); prstatus->pr_pgrp = task_pgrp_vnr(p); prstatus->pr_sid = task_session_vnr(p); if (thread_group_leader(p)) { @@ -1432,8 +1434,10 @@ static int fill_psinfo(struct elf_prpsinfo *psinfo, struct task_struct *p, psinfo->pr_psargs[i] = ' '; psinfo->pr_psargs[len] = 0; + rcu_read_lock(); + psinfo->pr_ppid = task_pid_vnr(rcu_dereference(p->real_parent)); + rcu_read_unlock(); psinfo->pr_pid = task_pid_vnr(p); - psinfo->pr_ppid = task_pid_vnr(p->real_parent); psinfo->pr_pgrp = task_pgrp_vnr(p); psinfo->pr_sid = task_session_vnr(p); -- cgit v1.2.3 From 0b923606e75f1ab672e25b14ac039a1cdcfa382f Mon Sep 17 00:00:00 2001 From: Peter Oberparleiter Date: Wed, 17 Jun 2009 16:28:05 -0700 Subject: seq_file: add function to write binary data seq_write() can be used to construct seq_files containing arbitrary data. Required by the gcov-profiling interface to synthesize binary profiling data files. Signed-off-by: Peter Oberparleiter Cc: Andi Kleen Cc: Huang Ying Cc: Li Wei Cc: Michael Ellerman Cc: Ingo Molnar Cc: Heiko Carstens Cc: Martin Schwidefsky Cc: Rusty Russell Cc: WANG Cong Cc: Sam Ravnborg Cc: Jeff Dike Cc: Al Viro Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/seq_file.c | 20 ++++++++++++++++++++ include/linux/seq_file.h | 1 + 2 files changed, 21 insertions(+) (limited to 'fs') diff --git a/fs/seq_file.c b/fs/seq_file.c index 7f40f30c55c5..6c959275f2d0 100644 --- a/fs/seq_file.c +++ b/fs/seq_file.c @@ -640,6 +640,26 @@ int seq_puts(struct seq_file *m, const char *s) } EXPORT_SYMBOL(seq_puts); +/** + * seq_write - write arbitrary data to buffer + * @seq: seq_file identifying the buffer to which data should be written + * @data: data address + * @len: number of bytes + * + * Return 0 on success, non-zero otherwise. + */ +int seq_write(struct seq_file *seq, const void *data, size_t len) +{ + if (seq->count + len < seq->size) { + memcpy(seq->buf + seq->count, data, len); + seq->count += len; + return 0; + } + seq->count = seq->size; + return -1; +} +EXPORT_SYMBOL(seq_write); + struct list_head *seq_list_start(struct list_head *head, loff_t pos) { struct list_head *lh; diff --git a/include/linux/seq_file.h b/include/linux/seq_file.h index 004f3b3342c5..0c6a86b79596 100644 --- a/include/linux/seq_file.h +++ b/include/linux/seq_file.h @@ -43,6 +43,7 @@ int seq_release(struct inode *, struct file *); int seq_escape(struct seq_file *, const char *, const char *); int seq_putc(struct seq_file *m, char c); int seq_puts(struct seq_file *m, const char *s); +int seq_write(struct seq_file *seq, const void *data, size_t len); int seq_printf(struct seq_file *, const char *, ...) __attribute__ ((format (printf,2,3))); -- cgit v1.2.3 From 6ddbbbfe52f35301ef5a1b595f912d8d2b3ec143 Mon Sep 17 00:00:00 2001 From: Mike Sager Date: Tue, 16 Jun 2009 04:20:47 +0300 Subject: nfsd41: Remove ip address collision detection case Verified that cthon and pynfs exchange id tests pass (except for the two expected fails: EID8 and EID50) Signed-off-by: Mike Sager Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4state.c | 18 ++++++------------ 1 file changed, 6 insertions(+), 12 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 99570c49add5..ef6944b19f06 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -1247,12 +1247,6 @@ nfsd4_exchange_id(struct svc_rqst *rqstp, expire_client(conf); goto out_new; } - if (ip_addr != conf->cl_addr && - !(exid->flags & EXCHGID4_FLAG_UPD_CONFIRMED_REC_A)) { - /* Client collision. 18.35.4 case 3 */ - status = nfserr_clid_inuse; - goto out; - } /* * Set bit when the owner id and verifier map to an already * confirmed client id (18.35.3). @@ -1266,12 +1260,12 @@ nfsd4_exchange_id(struct svc_rqst *rqstp, copy_verf(conf, &verf); new = conf; goto out_copy; - } else { - /* 18.35.4 case 7 */ - if (exid->flags & EXCHGID4_FLAG_UPD_CONFIRMED_REC_A) { - status = nfserr_noent; - goto out; - } + } + + /* 18.35.4 case 7 */ + if (exid->flags & EXCHGID4_FLAG_UPD_CONFIRMED_REC_A) { + status = nfserr_noent; + goto out; } unconf = find_unconfirmed_client_by_str(dname, strhashval, true); -- cgit v1.2.3 From ef52bff8409bba78b042f1bcf33a0f49debc9774 Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Tue, 16 Jun 2009 04:20:50 +0300 Subject: nfsd41: Backchannel: cleanup nfs4.0 callback encode routines Mimic the client and prepare to share the back channel xdr with NFSv4.1. Bump the number of operations in each encode routine, then backfill the number of operations. Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy Signed-off-by: Ricardo Labiaga Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4callback.c | 24 ++++++++++++++++-------- 1 file changed, 16 insertions(+), 8 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c index f4fab69a8c30..353eb4a0b847 100644 --- a/fs/nfsd/nfs4callback.c +++ b/fs/nfsd/nfs4callback.c @@ -140,8 +140,9 @@ struct nfs4_cb_compound_hdr { int status; u32 ident; u32 nops; + __be32 *nops_p; u32 taglen; - char * tag; + char *tag; }; static struct { @@ -201,7 +202,7 @@ nfs_cb_stat_to_errno(int stat) * XDR encode */ -static int +static void encode_cb_compound_hdr(struct xdr_stream *xdr, struct nfs4_cb_compound_hdr *hdr) { __be32 * p; @@ -210,12 +211,18 @@ encode_cb_compound_hdr(struct xdr_stream *xdr, struct nfs4_cb_compound_hdr *hdr) WRITE32(0); /* tag length is always 0 */ WRITE32(NFS4_MINOR_VERSION); WRITE32(hdr->ident); + hdr->nops_p = p; WRITE32(hdr->nops); - return 0; } -static int -encode_cb_recall(struct xdr_stream *xdr, struct nfs4_delegation *dp) +static void encode_cb_nops(struct nfs4_cb_compound_hdr *hdr) +{ + *hdr->nops_p = htonl(hdr->nops); +} + +static void +encode_cb_recall(struct xdr_stream *xdr, struct nfs4_delegation *dp, + struct nfs4_cb_compound_hdr *hdr) { __be32 *p; int len = dp->dl_fh.fh_size; @@ -227,7 +234,7 @@ encode_cb_recall(struct xdr_stream *xdr, struct nfs4_delegation *dp) WRITE32(0); /* truncate optimization not implemented */ WRITE32(len); WRITEMEM(&dp->dl_fh.fh_base, len); - return 0; + hdr->nops++; } static int @@ -246,12 +253,13 @@ nfs4_xdr_enc_cb_recall(struct rpc_rqst *req, __be32 *p, struct nfs4_delegation * struct xdr_stream xdr; struct nfs4_cb_compound_hdr hdr = { .ident = args->dl_ident, - .nops = 1, }; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_cb_compound_hdr(&xdr, &hdr); - return (encode_cb_recall(&xdr, args)); + encode_cb_recall(&xdr, args, &hdr); + encode_cb_nops(&hdr); + return 0; } -- cgit v1.2.3 From ab52ae6db035fa425f90146327ab7d2c5d3e5654 Mon Sep 17 00:00:00 2001 From: Andy Adamson Date: Tue, 16 Jun 2009 04:20:53 +0300 Subject: nfsd41: Backchannel: minorversion support for the back channel Prepare to share backchannel code with NFSv4.1. Signed-off-by: Andy Adamson Signed-off-by: Benny Halevy Signed-off-by: Ricardo Labiaga [nfsd41: use nfsd4_cb_sequence for callback minorversion] Signed-off-by: Benny Halevy Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4callback.c | 3 ++- fs/nfsd/nfs4state.c | 1 + include/linux/nfsd/state.h | 3 ++- 3 files changed, 5 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c index 353eb4a0b847..3fd23f7aceca 100644 --- a/fs/nfsd/nfs4callback.c +++ b/fs/nfsd/nfs4callback.c @@ -141,6 +141,7 @@ struct nfs4_cb_compound_hdr { u32 ident; u32 nops; __be32 *nops_p; + u32 minorversion; u32 taglen; char *tag; }; @@ -209,7 +210,7 @@ encode_cb_compound_hdr(struct xdr_stream *xdr, struct nfs4_cb_compound_hdr *hdr) RESERVE_SPACE(16); WRITE32(0); /* tag length is always 0 */ - WRITE32(NFS4_MINOR_VERSION); + WRITE32(hdr->minorversion); WRITE32(hdr->ident); hdr->nops_p = p; WRITE32(hdr->nops); diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index ef6944b19f06..980a216a48c8 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -984,6 +984,7 @@ gen_callback(struct nfs4_client *clp, struct nfsd4_setclientid *se) if ( !(parse_ipv4(se->se_callback_addr_len, se->se_callback_addr_val, &cb->cb_addr, &cb->cb_port))) goto out_err; + cb->cb_minorversion = 0; cb->cb_prog = se->se_callback_prog; cb->cb_ident = se->se_callback_ident; return; diff --git a/include/linux/nfsd/state.h b/include/linux/nfsd/state.h index 105cc100de05..f5a95fd34312 100644 --- a/include/linux/nfsd/state.h +++ b/include/linux/nfsd/state.h @@ -85,7 +85,8 @@ struct nfs4_cb_conn { u32 cb_addr; unsigned short cb_port; u32 cb_prog; - u32 cb_ident; + u32 cb_minorversion; + u32 cb_ident; /* minorversion 0 only */ /* RPC client info */ atomic_t cb_set; /* successful CB_NULL call */ struct rpc_clnt * cb_client; -- cgit v1.2.3 From 90c699a9ee4be165966d40f1837909ccb8890a68 Mon Sep 17 00:00:00 2001 From: Bartlomiej Zolnierkiewicz Date: Fri, 19 Jun 2009 08:08:50 +0200 Subject: block: rename CONFIG_LBD to CONFIG_LBDAF Follow-up to "block: enable by default support for large devices and files on 32-bit archs". Rename CONFIG_LBD to CONFIG_LBDAF to: - allow update of existing [def]configs for "default y" change - reflect that it is used also for large files support nowadays Signed-off-by: Bartlomiej Zolnierkiewicz Signed-off-by: Jens Axboe --- Documentation/SubmitChecklist | 2 +- Documentation/ja_JP/SubmitChecklist | 2 +- block/Kconfig | 4 ++-- drivers/md/dm-exception-store.h | 2 +- fs/ext3/resize.c | 2 +- fs/ext3/super.c | 2 +- fs/ext4/resize.c | 2 +- fs/ext4/super.c | 10 +++++----- fs/gfs2/Kconfig | 2 +- fs/ocfs2/super.c | 2 +- fs/xfs/linux-2.6/xfs_linux.h | 2 +- fs/xfs/linux-2.6/xfs_super.c | 2 +- include/linux/kernel.h | 2 +- include/linux/types.h | 2 +- 14 files changed, 19 insertions(+), 19 deletions(-) (limited to 'fs') diff --git a/Documentation/SubmitChecklist b/Documentation/SubmitChecklist index ac5e0b2f1097..78a9168ff377 100644 --- a/Documentation/SubmitChecklist +++ b/Documentation/SubmitChecklist @@ -54,7 +54,7 @@ kernel patches. CONFIG_PREEMPT. 14: If the patch affects IO/Disk, etc: has been tested with and without - CONFIG_LBD. + CONFIG_LBDAF. 15: All codepaths have been exercised with all lockdep features enabled. diff --git a/Documentation/ja_JP/SubmitChecklist b/Documentation/ja_JP/SubmitChecklist index 6c42e071d723..2df4576f1173 100644 --- a/Documentation/ja_JP/SubmitChecklist +++ b/Documentation/ja_JP/SubmitChecklist @@ -75,7 +75,7 @@ Linux カーネルパッチ投稿者向けチェックリスト ビルドした上、動作確認を行ってください。 14: もしパッチがディスクのI/O性能などに影響を与えるようであれば、 - 'CONFIG_LBD'オプションを有効にした場合と無効にした場合の両方で + 'CONFIG_LBDAF'オプションを有効にした場合と無効にした場合の両方で テストを実施してみてください。 15: lockdepの機能を全て有効にした上で、全てのコードパスを評価してください。 diff --git a/block/Kconfig b/block/Kconfig index 2c39527aa7db..95a86adc33a1 100644 --- a/block/Kconfig +++ b/block/Kconfig @@ -23,8 +23,8 @@ menuconfig BLOCK if BLOCK -config LBD - bool "Support for large block devices and files" +config LBDAF + bool "Support for large (2TB+) block devices and files" depends on !64BIT default y help diff --git a/drivers/md/dm-exception-store.h b/drivers/md/dm-exception-store.h index 0a2e6e7f67b3..c92701dc5001 100644 --- a/drivers/md/dm-exception-store.h +++ b/drivers/md/dm-exception-store.h @@ -111,7 +111,7 @@ struct dm_exception_store { /* * Funtions to manipulate consecutive chunks */ -# if defined(CONFIG_LBD) || (BITS_PER_LONG == 64) +# if defined(CONFIG_LBDAF) || (BITS_PER_LONG == 64) # define DM_CHUNK_CONSECUTIVE_BITS 8 # define DM_CHUNK_NUMBER_BITS 56 diff --git a/fs/ext3/resize.c b/fs/ext3/resize.c index 8a0b26340b54..8359e7b3dc89 100644 --- a/fs/ext3/resize.c +++ b/fs/ext3/resize.c @@ -990,7 +990,7 @@ int ext3_group_extend(struct super_block *sb, struct ext3_super_block *es, sb->s_id, n_blocks_count); if (sizeof(sector_t) < 8) ext3_warning(sb, __func__, - "CONFIG_LBD not enabled\n"); + "CONFIG_LBDAF not enabled\n"); return -EINVAL; } diff --git a/fs/ext3/super.c b/fs/ext3/super.c index 26aa64dee6aa..601e881e6105 100644 --- a/fs/ext3/super.c +++ b/fs/ext3/super.c @@ -1812,7 +1812,7 @@ static int ext3_fill_super (struct super_block *sb, void *data, int silent) printk(KERN_ERR "EXT3-fs: filesystem on %s:" " too large to mount safely\n", sb->s_id); if (sizeof(sector_t) < 8) - printk(KERN_WARNING "EXT3-fs: CONFIG_LBD not " + printk(KERN_WARNING "EXT3-fs: CONFIG_LBDAF not " "enabled\n"); goto failed_mount; } diff --git a/fs/ext4/resize.c b/fs/ext4/resize.c index 27eb289eea37..68b0351fc647 100644 --- a/fs/ext4/resize.c +++ b/fs/ext4/resize.c @@ -1002,7 +1002,7 @@ int ext4_group_extend(struct super_block *sb, struct ext4_super_block *es, " too large to resize to %llu blocks safely\n", sb->s_id, n_blocks_count); if (sizeof(sector_t) < 8) - ext4_warning(sb, __func__, "CONFIG_LBD not enabled"); + ext4_warning(sb, __func__, "CONFIG_LBDAF not enabled"); return -EINVAL; } diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 012c4251397e..6e0c2d77c87e 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -1957,7 +1957,7 @@ static loff_t ext4_max_size(int blkbits, int has_huge_files) /* small i_blocks in vfs inode? */ if (!has_huge_files || sizeof(blkcnt_t) < sizeof(u64)) { /* - * CONFIG_LBD is not enabled implies the inode + * CONFIG_LBDAF is not enabled implies the inode * i_block represent total blocks in 512 bytes * 32 == size of vfs inode i_blocks * 8 */ @@ -2000,7 +2000,7 @@ static loff_t ext4_max_bitmap_size(int bits, int has_huge_files) if (!has_huge_files || sizeof(blkcnt_t) < sizeof(u64)) { /* - * !has_huge_files or CONFIG_LBD not enabled implies that + * !has_huge_files or CONFIG_LBDAF not enabled implies that * the inode i_block field represents total file blocks in * 2^32 512-byte sectors == size of vfs inode i_blocks * 8 */ @@ -2436,13 +2436,13 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent) if (has_huge_files) { /* * Large file size enabled file system can only be - * mount if kernel is build with CONFIG_LBD + * mount if kernel is build with CONFIG_LBDAF */ if (sizeof(root->i_blocks) < sizeof(u64) && !(sb->s_flags & MS_RDONLY)) { ext4_msg(sb, KERN_ERR, "Filesystem with huge " "files cannot be mounted read-write " - "without CONFIG_LBD"); + "without CONFIG_LBDAF"); goto failed_mount; } } @@ -2566,7 +2566,7 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent) ext4_msg(sb, KERN_ERR, "filesystem" " too large to mount safely"); if (sizeof(sector_t) < 8) - ext4_msg(sb, KERN_WARNING, "CONFIG_LBD not enabled"); + ext4_msg(sb, KERN_WARNING, "CONFIG_LBDAF not enabled"); goto failed_mount; } diff --git a/fs/gfs2/Kconfig b/fs/gfs2/Kconfig index cad957cdb1e5..5971359d2090 100644 --- a/fs/gfs2/Kconfig +++ b/fs/gfs2/Kconfig @@ -1,6 +1,6 @@ config GFS2_FS tristate "GFS2 file system support" - depends on EXPERIMENTAL && (64BIT || LBD) + depends on EXPERIMENTAL && (64BIT || LBDAF) select DLM if GFS2_FS_LOCKING_DLM select CONFIGFS_FS if GFS2_FS_LOCKING_DLM select SYSFS if GFS2_FS_LOCKING_DLM diff --git a/fs/ocfs2/super.c b/fs/ocfs2/super.c index d33767f17ba3..0d3ed7407a04 100644 --- a/fs/ocfs2/super.c +++ b/fs/ocfs2/super.c @@ -552,7 +552,7 @@ static unsigned long long ocfs2_max_file_offset(unsigned int bbits, */ #if BITS_PER_LONG == 32 -# if defined(CONFIG_LBD) +# if defined(CONFIG_LBDAF) BUILD_BUG_ON(sizeof(sector_t) != 8); /* * We might be limited by page cache size. diff --git a/fs/xfs/linux-2.6/xfs_linux.h b/fs/xfs/linux-2.6/xfs_linux.h index f65a53f8752f..6127e24062d0 100644 --- a/fs/xfs/linux-2.6/xfs_linux.h +++ b/fs/xfs/linux-2.6/xfs_linux.h @@ -24,7 +24,7 @@ * XFS_BIG_BLKNOS needs block layer disk addresses to be 64 bits. * XFS_BIG_INUMS requires XFS_BIG_BLKNOS to be set. */ -#if defined(CONFIG_LBD) || (BITS_PER_LONG == 64) +#if defined(CONFIG_LBDAF) || (BITS_PER_LONG == 64) # define XFS_BIG_BLKNOS 1 # define XFS_BIG_INUMS 1 #else diff --git a/fs/xfs/linux-2.6/xfs_super.c b/fs/xfs/linux-2.6/xfs_super.c index 2e09efbca8db..a220d36f789b 100644 --- a/fs/xfs/linux-2.6/xfs_super.c +++ b/fs/xfs/linux-2.6/xfs_super.c @@ -616,7 +616,7 @@ xfs_max_file_offset( */ #if BITS_PER_LONG == 32 -# if defined(CONFIG_LBD) +# if defined(CONFIG_LBDAF) ASSERT(sizeof(sector_t) == 8); pagefactor = PAGE_CACHE_SIZE; bitshift = BITS_PER_LONG; diff --git a/include/linux/kernel.h b/include/linux/kernel.h index c5a71c38a95f..fac104e7186a 100644 --- a/include/linux/kernel.h +++ b/include/linux/kernel.h @@ -58,7 +58,7 @@ extern const char linux_proc_banner[]; #define _RET_IP_ (unsigned long)__builtin_return_address(0) #define _THIS_IP_ ({ __label__ __here; __here: (unsigned long)&&__here; }) -#ifdef CONFIG_LBD +#ifdef CONFIG_LBDAF # include # define sector_div(a, b) do_div(a, b) #else diff --git a/include/linux/types.h b/include/linux/types.h index 5abe354020f9..c42724f8c802 100644 --- a/include/linux/types.h +++ b/include/linux/types.h @@ -131,7 +131,7 @@ typedef __s64 int64_t; * * blkcnt_t is the type of the inode's block count. */ -#ifdef CONFIG_LBD +#ifdef CONFIG_LBDAF typedef u64 sector_t; typedef u64 blkcnt_t; #else -- cgit v1.2.3 From 528da3e9e237059a84a2625e942811cf824a6efd Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Fri, 12 Jun 2009 16:04:26 -0400 Subject: inotify: inotify_destroy_mark_entry could get called twice inotify_destroy_mark_entry could get called twice for the same mark since it is called directly in inotify_rm_watch and when the mark is being destroyed for another reason. As an example assume that the file being watched was just deleted so inotify_destroy_mark_entry would get called from the path fsnotify_inoderemove() -> fsnotify_destroy_marks_by_inode() -> fsnotify_destroy_mark_entry() -> inotify_destroy_mark_entry(). If this happened at the same time as userspace tried to remove a watch via inotify_rm_watch we could attempt to remove the mark from the idr twice and could thus double dec the ref cnt and potentially could be in a use after free/double free situation. The fix is to have inotify_rm_watch use the generic recursive safe fsnotify_destroy_mark_by_entry() so we are sure the inotify_destroy_mark_entry() function can only be called one. This patch also renames the function to inotify_ingored_remove_idr() so it is clear what is actually going on in the function. Hopefully this fixes: [ 20.342058] idr_remove called for id=20 which is not allocated. [ 20.348000] Pid: 1860, comm: udevd Not tainted 2.6.30-tip #1077 [ 20.353933] Call Trace: [ 20.356410] [] idr_remove+0x115/0x18f [ 20.361737] [] ? _spin_lock+0x6d/0x75 [ 20.367061] [] ? inotify_destroy_mark_entry+0xa3/0xcf [ 20.373771] [] inotify_destroy_mark_entry+0xb7/0xcf [ 20.380306] [] inotify_freeing_mark+0xe/0x10 [ 20.386238] [] fsnotify_destroy_mark_by_entry+0x143/0x170 [ 20.393293] [] inotify_destroy_mark_entry+0x3c/0xcf [ 20.399829] [] sys_inotify_rm_watch+0x9b/0xc6 [ 20.405850] [] system_call_fastpath+0x16/0x1b Reported-by: Peter Zijlstra Signed-off-by: Eric Paris Tested-by: Peter Ziljlstra --- fs/notify/inotify/inotify.h | 3 ++- fs/notify/inotify/inotify_fsnotify.c | 2 +- fs/notify/inotify/inotify_user.c | 32 +++++--------------------------- 3 files changed, 8 insertions(+), 29 deletions(-) (limited to 'fs') diff --git a/fs/notify/inotify/inotify.h b/fs/notify/inotify/inotify.h index ea2605a58b8a..f234f3a4c8ca 100644 --- a/fs/notify/inotify/inotify.h +++ b/fs/notify/inotify/inotify.h @@ -15,7 +15,8 @@ struct inotify_inode_mark_entry { int wd; }; -extern void inotify_destroy_mark_entry(struct fsnotify_mark_entry *entry, struct fsnotify_group *group); +extern void inotify_ignored_and_remove_idr(struct fsnotify_mark_entry *entry, + struct fsnotify_group *group); extern void inotify_free_event_priv(struct fsnotify_event_private_data *event_priv); extern const struct fsnotify_ops inotify_fsnotify_ops; diff --git a/fs/notify/inotify/inotify_fsnotify.c b/fs/notify/inotify/inotify_fsnotify.c index 7ef75b83247e..47cd258fd24d 100644 --- a/fs/notify/inotify/inotify_fsnotify.c +++ b/fs/notify/inotify/inotify_fsnotify.c @@ -81,7 +81,7 @@ static int inotify_handle_event(struct fsnotify_group *group, struct fsnotify_ev static void inotify_freeing_mark(struct fsnotify_mark_entry *entry, struct fsnotify_group *group) { - inotify_destroy_mark_entry(entry, group); + inotify_ignored_and_remove_idr(entry, group); } static bool inotify_should_send_event(struct fsnotify_group *group, struct inode *inode, __u32 mask) diff --git a/fs/notify/inotify/inotify_user.c b/fs/notify/inotify/inotify_user.c index 982a412ac5bc..ff231ad23895 100644 --- a/fs/notify/inotify/inotify_user.c +++ b/fs/notify/inotify/inotify_user.c @@ -363,39 +363,17 @@ static int inotify_find_inode(const char __user *dirname, struct path *path, uns } /* - * When, for whatever reason, inotify is done with a mark (or what used to be a - * watch) we need to remove that watch from the idr and we need to send IN_IGNORED - * for the given wd. - * - * There is a bit of recursion here. The loop looks like: - * inotify_destroy_mark_entry -> fsnotify_destroy_mark_by_entry -> - * inotify_freeing_mark -> inotify_destory_mark_entry -> restart - * But the loop is broken in 2 places. fsnotify_destroy_mark_by_entry sets - * entry->group = NULL before the call to inotify_freeing_mark, so the if (egroup) - * test below will not call back to fsnotify again. But even if that test wasn't - * there this would still be safe since fsnotify_destroy_mark_by_entry() is - * safe from recursion. + * Send IN_IGNORED for this wd, remove this wd from the idr, and drop the + * internal reference help on the mark because it is in the idr. */ -void inotify_destroy_mark_entry(struct fsnotify_mark_entry *entry, struct fsnotify_group *group) +void inotify_ignored_and_remove_idr(struct fsnotify_mark_entry *entry, + struct fsnotify_group *group) { struct inotify_inode_mark_entry *ientry; struct inotify_event_private_data *event_priv; struct fsnotify_event_private_data *fsn_event_priv; - struct fsnotify_group *egroup; struct idr *idr; - spin_lock(&entry->lock); - egroup = entry->group; - - /* if egroup we aren't really done and something might still send events - * for this inode, on the callback we'll send the IN_IGNORED */ - if (egroup) { - spin_unlock(&entry->lock); - fsnotify_destroy_mark_by_entry(entry); - return; - } - spin_unlock(&entry->lock); - ientry = container_of(entry, struct inotify_inode_mark_entry, fsn_entry); event_priv = kmem_cache_alloc(event_priv_cachep, GFP_KERNEL); @@ -699,7 +677,7 @@ SYSCALL_DEFINE2(inotify_rm_watch, int, fd, __s32, wd) fsnotify_get_mark(entry); spin_unlock(&group->inotify_data.idr_lock); - inotify_destroy_mark_entry(entry, group); + fsnotify_destroy_mark_by_entry(entry); fsnotify_put_mark(entry); out: -- cgit v1.2.3 From 3e107603aecf886e1e5bda9dacbd9796eb2a2171 Mon Sep 17 00:00:00 2001 From: OGAWA Hirofumi Date: Sat, 20 Jun 2009 21:50:07 +0900 Subject: fat: Fix the removal of opts->fs_dmask (ce3b0f8d5c2203301fc87f3aaaed73e5819e2a48: New helper - current_umask()) is removing the opts->fs_dmask, probably it's a cut-and-paste miss or something. Signed-off-by: OGAWA Hirofumi --- fs/fat/inode.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/fat/inode.c b/fs/fat/inode.c index 304b411cb8bc..8970d8c49bb0 100644 --- a/fs/fat/inode.c +++ b/fs/fat/inode.c @@ -966,7 +966,7 @@ static int parse_options(char *options, int is_vfat, int silent, int *debug, opts->fs_uid = current_uid(); opts->fs_gid = current_gid(); - opts->fs_fmask = current_umask(); + opts->fs_fmask = opts->fs_dmask = current_umask(); opts->allow_utime = -1; opts->codepage = fat_default_codepage; opts->iocharset = fat_default_iocharset; -- cgit v1.2.3 From 578e4585685410cacd1a4ac86b7e3c12805be918 Mon Sep 17 00:00:00 2001 From: Benny Halevy Date: Thu, 18 Jun 2009 22:01:23 -0400 Subject: nfs41: Move initialization of nfs4_opendata seq_res to nfs4_init_opendata_res nfs4_open_recover_helper clears opendata->o_res before calling nfs4_init_opendata_res, thus causing NFSv4.0 OPEN operations to be sent rather than nfsv4.1. Signed-off-by: Benny Halevy Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 7a750f061c26..92ce43517814 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -680,6 +680,7 @@ static void nfs4_init_opendata_res(struct nfs4_opendata *p) p->o_res.server = p->o_arg.server; nfs_fattr_init(&p->f_attr); nfs_fattr_init(&p->dir_attr); + p->o_res.seq_res.sr_slotid = NFS4_MAX_SLOT_TABLE; } static struct nfs4_opendata *nfs4_opendata_alloc(struct path *path, @@ -711,7 +712,6 @@ static struct nfs4_opendata *nfs4_opendata_alloc(struct path *path, p->o_arg.server = server; p->o_arg.bitmask = server->attr_bitmask; p->o_arg.claim = NFS4_OPEN_CLAIM_NULL; - p->o_res.seq_res.sr_slotid = NFS4_MAX_SLOT_TABLE; if (flags & O_EXCL) { u32 *s = (u32 *) p->o_arg.u.verifier.data; s[0] = jiffies; -- cgit v1.2.3 From b76a3f93d01fc93a87cb6eba4e854ffe378b4bac Mon Sep 17 00:00:00 2001 From: Boaz Harrosh Date: Mon, 8 Jun 2009 19:28:41 +0300 Subject: exofs: Fix bio leak in error handling path (sync read) When failing a read request in the sync path, called from write_begin, I forgot to free the allocated bio, fix it. Signed-off-by: Boaz Harrosh --- fs/exofs/inode.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'fs') diff --git a/fs/exofs/inode.c b/fs/exofs/inode.c index 77d0a295eb1c..bb5d6ed0f7a8 100644 --- a/fs/exofs/inode.c +++ b/fs/exofs/inode.c @@ -295,6 +295,9 @@ static int read_exec(struct page_collect *pcol, bool is_sync) err: if (!is_sync) _unlock_pcol_pages(pcol, ret, READ); + else /* Pages unlocked by caller in sync mode only free bio */ + pcol_free(pcol); + kfree(pcol_copy); if (or) osd_end_request(or); -- cgit v1.2.3 From 27d2e1491985e95c486d991302e399f5c584b4eb Mon Sep 17 00:00:00 2001 From: Boaz Harrosh Date: Sun, 14 Jun 2009 17:23:09 +0300 Subject: exofs: Remove IBM copyrights Boaz, Congrats on getting all the OSD stuff into 2.6.30! I just pulled the git, and saw that the IBM copyrights are still there. Please remove them from all files: * Copyright (C) 2005, 2006 * International Business Machines IBM has revoked all rights on the code - they gave it to me. Thanks! Avishay Signed-off-by: Avishay Traeger Signed-off-by: Boaz Harrosh --- fs/exofs/common.h | 4 +--- fs/exofs/dir.c | 4 +--- fs/exofs/exofs.h | 4 +--- fs/exofs/file.c | 4 +--- fs/exofs/inode.c | 4 +--- fs/exofs/namei.c | 4 +--- fs/exofs/osd.c | 4 +--- fs/exofs/super.c | 4 +--- fs/exofs/symlink.c | 4 +--- 9 files changed, 9 insertions(+), 27 deletions(-) (limited to 'fs') diff --git a/fs/exofs/common.h b/fs/exofs/common.h index 24667eedc023..c6718e4817fe 100644 --- a/fs/exofs/common.h +++ b/fs/exofs/common.h @@ -2,9 +2,7 @@ * common.h - Common definitions for both Kernel and user-mode utilities * * Copyright (C) 2005, 2006 - * Avishay Traeger (avishay@gmail.com) (avishay@il.ibm.com) - * Copyright (C) 2005, 2006 - * International Business Machines + * Avishay Traeger (avishay@gmail.com) * Copyright (C) 2008, 2009 * Boaz Harrosh * diff --git a/fs/exofs/dir.c b/fs/exofs/dir.c index 65b0c8c776a1..4cfab1cc75c0 100644 --- a/fs/exofs/dir.c +++ b/fs/exofs/dir.c @@ -1,8 +1,6 @@ /* * Copyright (C) 2005, 2006 - * Avishay Traeger (avishay@gmail.com) (avishay@il.ibm.com) - * Copyright (C) 2005, 2006 - * International Business Machines + * Avishay Traeger (avishay@gmail.com) * Copyright (C) 2008, 2009 * Boaz Harrosh * diff --git a/fs/exofs/exofs.h b/fs/exofs/exofs.h index 0fd4c7859679..c413b74ecf31 100644 --- a/fs/exofs/exofs.h +++ b/fs/exofs/exofs.h @@ -1,8 +1,6 @@ /* * Copyright (C) 2005, 2006 - * Avishay Traeger (avishay@gmail.com) (avishay@il.ibm.com) - * Copyright (C) 2005, 2006 - * International Business Machines + * Avishay Traeger (avishay@gmail.com) * Copyright (C) 2008, 2009 * Boaz Harrosh * diff --git a/fs/exofs/file.c b/fs/exofs/file.c index 6ed7fe484752..c6810038d637 100644 --- a/fs/exofs/file.c +++ b/fs/exofs/file.c @@ -1,8 +1,6 @@ /* * Copyright (C) 2005, 2006 - * Avishay Traeger (avishay@gmail.com) (avishay@il.ibm.com) - * Copyright (C) 2005, 2006 - * International Business Machines + * Avishay Traeger (avishay@gmail.com) * Copyright (C) 2008, 2009 * Boaz Harrosh * diff --git a/fs/exofs/inode.c b/fs/exofs/inode.c index bb5d6ed0f7a8..6c10f7476699 100644 --- a/fs/exofs/inode.c +++ b/fs/exofs/inode.c @@ -1,8 +1,6 @@ /* * Copyright (C) 2005, 2006 - * Avishay Traeger (avishay@gmail.com) (avishay@il.ibm.com) - * Copyright (C) 2005, 2006 - * International Business Machines + * Avishay Traeger (avishay@gmail.com) * Copyright (C) 2008, 2009 * Boaz Harrosh * diff --git a/fs/exofs/namei.c b/fs/exofs/namei.c index 77fdd765e76d..b7dd0c236863 100644 --- a/fs/exofs/namei.c +++ b/fs/exofs/namei.c @@ -1,8 +1,6 @@ /* * Copyright (C) 2005, 2006 - * Avishay Traeger (avishay@gmail.com) (avishay@il.ibm.com) - * Copyright (C) 2005, 2006 - * International Business Machines + * Avishay Traeger (avishay@gmail.com) * Copyright (C) 2008, 2009 * Boaz Harrosh * diff --git a/fs/exofs/osd.c b/fs/exofs/osd.c index b3d2ccb87aaa..4372542df284 100644 --- a/fs/exofs/osd.c +++ b/fs/exofs/osd.c @@ -1,8 +1,6 @@ /* * Copyright (C) 2005, 2006 - * Avishay Traeger (avishay@gmail.com) (avishay@il.ibm.com) - * Copyright (C) 2005, 2006 - * International Business Machines + * Avishay Traeger (avishay@gmail.com) * Copyright (C) 2008, 2009 * Boaz Harrosh * diff --git a/fs/exofs/super.c b/fs/exofs/super.c index 8216c5b77b53..e47b38e55a26 100644 --- a/fs/exofs/super.c +++ b/fs/exofs/super.c @@ -1,8 +1,6 @@ /* * Copyright (C) 2005, 2006 - * Avishay Traeger (avishay@gmail.com) (avishay@il.ibm.com) - * Copyright (C) 2005, 2006 - * International Business Machines + * Avishay Traeger (avishay@gmail.com) * Copyright (C) 2008, 2009 * Boaz Harrosh * diff --git a/fs/exofs/symlink.c b/fs/exofs/symlink.c index 36e2d7bc7f7b..4dd687c3e747 100644 --- a/fs/exofs/symlink.c +++ b/fs/exofs/symlink.c @@ -1,8 +1,6 @@ /* * Copyright (C) 2005, 2006 - * Avishay Traeger (avishay@gmail.com) (avishay@il.ibm.com) - * Copyright (C) 2005, 2006 - * International Business Machines + * Avishay Traeger (avishay@gmail.com) * Copyright (C) 2008, 2009 * Boaz Harrosh * -- cgit v1.2.3 From baaf94cdc7fe1c61e3c660a3b055724fd9d0a034 Mon Sep 17 00:00:00 2001 From: Boaz Harrosh Date: Sun, 14 Jun 2009 16:52:10 +0300 Subject: exofs: Avoid using file_fsync() The use of file_fsync() in exofs_file_sync() is not necessary since it does some extra stuff not used by exofs. Open code just the parts that are currently needed. TODO: Farther optimization can be done to sync the sb only on inode update of new files, Usually the sb update is not needed in exofs. Signed-off-by: Boaz Harrosh --- fs/exofs/exofs.h | 3 +++ fs/exofs/file.c | 17 ++++++++++++----- fs/exofs/super.c | 2 +- 3 files changed, 16 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/exofs/exofs.h b/fs/exofs/exofs.h index c413b74ecf31..5ec72e020b22 100644 --- a/fs/exofs/exofs.h +++ b/fs/exofs/exofs.h @@ -154,6 +154,9 @@ ino_t exofs_parent_ino(struct dentry *child); int exofs_set_link(struct inode *, struct exofs_dir_entry *, struct page *, struct inode *); +/* super.c */ +int exofs_sync_fs(struct super_block *sb, int wait); + /********************* * operation vectors * *********************/ diff --git a/fs/exofs/file.c b/fs/exofs/file.c index c6810038d637..839b9dc1e70f 100644 --- a/fs/exofs/file.c +++ b/fs/exofs/file.c @@ -45,16 +45,23 @@ static int exofs_file_fsync(struct file *filp, struct dentry *dentry, { int ret; struct address_space *mapping = filp->f_mapping; + struct inode *inode = dentry->d_inode; + struct super_block *sb; ret = filemap_write_and_wait(mapping); if (ret) return ret; - /*Note: file_fsync below also calles sync_blockdev, which is a no-op - * for exofs, but other then that it does sync_inode and - * sync_superblock which is what we need here. - */ - return file_fsync(filp, dentry, datasync); + /* sync the inode attributes */ + ret = write_inode_now(inode, 1); + + /* This is a good place to write the sb */ + /* TODO: Sechedule an sb-sync on create */ + sb = inode->i_sb; + if (sb->s_dirt) + exofs_sync_fs(sb, 1); + + return ret; } static int exofs_flush(struct file *file, fl_owner_t id) diff --git a/fs/exofs/super.c b/fs/exofs/super.c index e47b38e55a26..a343b4ea62f6 100644 --- a/fs/exofs/super.c +++ b/fs/exofs/super.c @@ -198,7 +198,7 @@ static const struct export_operations exofs_export_ops; /* * Write the superblock to the OSD */ -static int exofs_sync_fs(struct super_block *sb, int wait) +int exofs_sync_fs(struct super_block *sb, int wait) { struct exofs_sb_info *sbi; struct exofs_fscb *fscb; -- cgit v1.2.3 From e38be994b9cad09b0d8d78a1875d7e8a2e115d29 Mon Sep 17 00:00:00 2001 From: Hitoshi Mitake Date: Mon, 22 Jun 2009 11:11:20 +0900 Subject: Making fs/minix/minix.h double including safe I happened to find that fs/minix/minix.h doesn't guard double include. Yes, I know this never cause something destructive because this is self-evidence that no source file includes minix.h twice, but I think fixing this is better than disregarding it. Signed-off-by: Hitoshi Mitake Signed-off-by: Linus Torvalds --- fs/minix/minix.h | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'fs') diff --git a/fs/minix/minix.h b/fs/minix/minix.h index cb7fdd11f9a5..9dcf95b42116 100644 --- a/fs/minix/minix.h +++ b/fs/minix/minix.h @@ -1,3 +1,6 @@ +#ifndef FS_MINIX_H +#define FS_MINIX_H + #include #include #include @@ -86,3 +89,5 @@ static inline struct minix_inode_info *minix_i(struct inode *inode) { return list_entry(inode, struct minix_inode_info, vfs_inode); } + +#endif /* FS_MINIX_H */ -- cgit v1.2.3 From 1c520dfbf391e1617ef61553f815b8006a066c44 Mon Sep 17 00:00:00 2001 From: Joel Becker Date: Fri, 19 Jun 2009 15:14:13 -0700 Subject: ocfs2: Provide the ocfs2_dlm_lvb_valid() stack API. The Lock Value Block (LVB) of a DLM lock can be lost when nodes die and the DLM cannot reconstruct its state. Clients of the DLM need to know this. ocfs2's internal DLM, o2dlm, explicitly zeroes out the LVB when it loses track of the state. This is not a standard behavior, but ocfs2 has always relied on it. Thus, an o2dlm LVB is always "valid". ocfs2 now supports both o2dlm and fs/dlm via the stack glue. When fs/dlm loses track of an LVBs state, it sets a flag (DLM_SBF_VALNOTVALID) on the Lock Status Block (LKSB). The contents of the LVB may be garbage or merely stale. ocfs2 doesn't want to try to guess at the validity of the stale LVB. Instead, it should be checking the VALNOTVALID flag. As this is the 'standard' way of treating LVBs, we will promote this behavior. We add a stack glue API ocfs2_dlm_lvb_valid(). It returns non-zero when the LVB is valid. o2dlm will always return valid, while fs/dlm will check VALNOTVALID. Signed-off-by: Joel Becker Acked-by: Mark Fasheh --- fs/ocfs2/dlmglue.c | 9 ++++++--- fs/ocfs2/stack_o2cb.c | 11 +++++++++++ fs/ocfs2/stack_user.c | 8 ++++++++ fs/ocfs2/stackglue.c | 13 +++++++------ fs/ocfs2/stackglue.h | 6 ++++++ 5 files changed, 38 insertions(+), 9 deletions(-) (limited to 'fs') diff --git a/fs/ocfs2/dlmglue.c b/fs/ocfs2/dlmglue.c index 6cdeaa76f27f..83d2ddb27186 100644 --- a/fs/ocfs2/dlmglue.c +++ b/fs/ocfs2/dlmglue.c @@ -1989,7 +1989,8 @@ static inline int ocfs2_meta_lvb_is_trustable(struct inode *inode, { struct ocfs2_meta_lvb *lvb = ocfs2_dlm_lvb(&lockres->l_lksb); - if (lvb->lvb_version == OCFS2_LVB_VERSION + if (ocfs2_dlm_lvb_valid(&lockres->l_lksb) + && lvb->lvb_version == OCFS2_LVB_VERSION && be32_to_cpu(lvb->lvb_igeneration) == inode->i_generation) return 1; return 0; @@ -2382,7 +2383,8 @@ int ocfs2_orphan_scan_lock(struct ocfs2_super *osb, u32 *seqno, int ex) return status; lvb = ocfs2_dlm_lvb(&lockres->l_lksb); - if (lvb->lvb_version == OCFS2_ORPHAN_LVB_VERSION) + if (ocfs2_dlm_lvb_valid(&lockres->l_lksb) && + lvb->lvb_version == OCFS2_ORPHAN_LVB_VERSION) *seqno = be32_to_cpu(lvb->lvb_os_seqno); return status; } @@ -3627,7 +3629,8 @@ static int ocfs2_refresh_qinfo(struct ocfs2_mem_dqinfo *oinfo) struct ocfs2_global_disk_dqinfo *gdinfo; int status = 0; - if (lvb->lvb_version == OCFS2_QINFO_LVB_VERSION) { + if (ocfs2_dlm_lvb_valid(&lockres->l_lksb) && + lvb->lvb_version == OCFS2_QINFO_LVB_VERSION) { info->dqi_bgrace = be32_to_cpu(lvb->lvb_bgrace); info->dqi_igrace = be32_to_cpu(lvb->lvb_igrace); oinfo->dqi_syncms = be32_to_cpu(lvb->lvb_syncms); diff --git a/fs/ocfs2/stack_o2cb.c b/fs/ocfs2/stack_o2cb.c index fcd120f1493a..3f661376a2de 100644 --- a/fs/ocfs2/stack_o2cb.c +++ b/fs/ocfs2/stack_o2cb.c @@ -236,6 +236,16 @@ static int o2cb_dlm_lock_status(union ocfs2_dlm_lksb *lksb) return dlm_status_to_errno(lksb->lksb_o2dlm.status); } +/* + * o2dlm aways has a "valid" LVB. If the dlm loses track of the LVB + * contents, it will zero out the LVB. Thus the caller can always trust + * the contents. + */ +static int o2cb_dlm_lvb_valid(union ocfs2_dlm_lksb *lksb) +{ + return 1; +} + static void *o2cb_dlm_lvb(union ocfs2_dlm_lksb *lksb) { return (void *)(lksb->lksb_o2dlm.lvb); @@ -354,6 +364,7 @@ static struct ocfs2_stack_operations o2cb_stack_ops = { .dlm_lock = o2cb_dlm_lock, .dlm_unlock = o2cb_dlm_unlock, .lock_status = o2cb_dlm_lock_status, + .lvb_valid = o2cb_dlm_lvb_valid, .lock_lvb = o2cb_dlm_lvb, .dump_lksb = o2cb_dump_lksb, }; diff --git a/fs/ocfs2/stack_user.c b/fs/ocfs2/stack_user.c index 9b76d41a8ac6..ff4c798a5635 100644 --- a/fs/ocfs2/stack_user.c +++ b/fs/ocfs2/stack_user.c @@ -738,6 +738,13 @@ static int user_dlm_lock_status(union ocfs2_dlm_lksb *lksb) return lksb->lksb_fsdlm.sb_status; } +static int user_dlm_lvb_valid(union ocfs2_dlm_lksb *lksb) +{ + int invalid = lksb->lksb_fsdlm.sb_flags & DLM_SBF_VALNOTVALID; + + return !invalid; +} + static void *user_dlm_lvb(union ocfs2_dlm_lksb *lksb) { if (!lksb->lksb_fsdlm.sb_lvbptr) @@ -873,6 +880,7 @@ static struct ocfs2_stack_operations ocfs2_user_plugin_ops = { .dlm_lock = user_dlm_lock, .dlm_unlock = user_dlm_unlock, .lock_status = user_dlm_lock_status, + .lvb_valid = user_dlm_lvb_valid, .lock_lvb = user_dlm_lvb, .plock = user_plock, .dump_lksb = user_dlm_dump_lksb, diff --git a/fs/ocfs2/stackglue.c b/fs/ocfs2/stackglue.c index 68b668b0e60a..3f2f1c45b7b6 100644 --- a/fs/ocfs2/stackglue.c +++ b/fs/ocfs2/stackglue.c @@ -6,7 +6,7 @@ * Code which implements an OCFS2 specific interface to underlying * cluster stacks. * - * Copyright (C) 2007 Oracle. All rights reserved. + * Copyright (C) 2007, 2009 Oracle. All rights reserved. * * This program is free software; you can redistribute it and/or * modify it under the terms of the GNU General Public @@ -271,11 +271,12 @@ int ocfs2_dlm_lock_status(union ocfs2_dlm_lksb *lksb) } EXPORT_SYMBOL_GPL(ocfs2_dlm_lock_status); -/* - * Why don't we cast to ocfs2_meta_lvb? The "clean" answer is that we - * don't cast at the glue level. The real answer is that the header - * ordering is nigh impossible. - */ +int ocfs2_dlm_lvb_valid(union ocfs2_dlm_lksb *lksb) +{ + return active_stack->sp_ops->lvb_valid(lksb); +} +EXPORT_SYMBOL_GPL(ocfs2_dlm_lvb_valid); + void *ocfs2_dlm_lvb(union ocfs2_dlm_lksb *lksb) { return active_stack->sp_ops->lock_lvb(lksb); diff --git a/fs/ocfs2/stackglue.h b/fs/ocfs2/stackglue.h index c571af375ef8..03a44d60eac9 100644 --- a/fs/ocfs2/stackglue.h +++ b/fs/ocfs2/stackglue.h @@ -185,6 +185,11 @@ struct ocfs2_stack_operations { */ int (*lock_status)(union ocfs2_dlm_lksb *lksb); + /* + * Return non-zero if the LVB is valid. + */ + int (*lvb_valid)(union ocfs2_dlm_lksb *lksb); + /* * Pull the lvb pointer off of the stack-specific lksb. */ @@ -252,6 +257,7 @@ int ocfs2_dlm_unlock(struct ocfs2_cluster_connection *conn, struct ocfs2_lock_res *astarg); int ocfs2_dlm_lock_status(union ocfs2_dlm_lksb *lksb); +int ocfs2_dlm_lvb_valid(union ocfs2_dlm_lksb *lksb); void *ocfs2_dlm_lvb(union ocfs2_dlm_lksb *lksb); void ocfs2_dlm_dump_lksb(union ocfs2_dlm_lksb *lksb); -- cgit v1.2.3 From 1962f39abbb2d5643a7d59169422661a2d58793d Mon Sep 17 00:00:00 2001 From: Tao Ma Date: Fri, 19 Jun 2009 15:36:52 +0800 Subject: ocfs2: Update atime in splice read if necessary. We should call ocfs2_inode_lock_atime instead of ocfs2_inode_lock in ocfs2_file_splice_read like we do in ocfs2_file_aio_read so that we can update atime in splice read if necessary. Signed-off-by: Tao Ma Signed-off-by: Joel Becker --- fs/ocfs2/file.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c index 07267e0da909..62442e413a00 100644 --- a/fs/ocfs2/file.c +++ b/fs/ocfs2/file.c @@ -2026,7 +2026,7 @@ static ssize_t ocfs2_file_splice_read(struct file *in, size_t len, unsigned int flags) { - int ret = 0; + int ret = 0, lock_level = 0; struct inode *inode = in->f_path.dentry->d_inode; mlog_entry("(0x%p, 0x%p, %u, '%.*s')\n", in, pipe, @@ -2037,12 +2037,12 @@ static ssize_t ocfs2_file_splice_read(struct file *in, /* * See the comment in ocfs2_file_aio_read() */ - ret = ocfs2_inode_lock(inode, NULL, 0); + ret = ocfs2_inode_lock_atime(inode, in->f_vfsmnt, &lock_level); if (ret < 0) { mlog_errno(ret); goto bail; } - ocfs2_inode_unlock(inode, 0); + ocfs2_inode_unlock(inode, lock_level); ret = generic_file_splice_read(in, ppos, pipe, len, flags); -- cgit v1.2.3 From 94e41ecfe0f202df948fdbb19a53308a58cf2184 Mon Sep 17 00:00:00 2001 From: Sunil Mushran Date: Fri, 19 Jun 2009 14:45:54 -0700 Subject: ocfs2: Pin journal head before accessing jh->b_committed_data This patch adds jbd_lock_bh_state() and jbd_unlock_bh_state() around accessses to jh->b_committed_data. Fixes oss bugzilla#1131 http://oss.oracle.com/bugzilla/show_bug.cgi?id=1131 Signed-off-by: Sunil Mushran Signed-off-by: Joel Becker --- fs/ocfs2/suballoc.c | 28 ++++++++++++++++++++++++---- 1 file changed, 24 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/ocfs2/suballoc.c b/fs/ocfs2/suballoc.c index 8439f6b324b9..73a16d4666dc 100644 --- a/fs/ocfs2/suballoc.c +++ b/fs/ocfs2/suballoc.c @@ -923,14 +923,23 @@ static int ocfs2_test_bg_bit_allocatable(struct buffer_head *bg_bh, int nr) { struct ocfs2_group_desc *bg = (struct ocfs2_group_desc *) bg_bh->b_data; + int ret; if (ocfs2_test_bit(nr, (unsigned long *)bg->bg_bitmap)) return 0; - if (!buffer_jbd(bg_bh) || !bh2jh(bg_bh)->b_committed_data) + + if (!buffer_jbd(bg_bh)) return 1; + jbd_lock_bh_state(bg_bh); bg = (struct ocfs2_group_desc *) bh2jh(bg_bh)->b_committed_data; - return !ocfs2_test_bit(nr, (unsigned long *)bg->bg_bitmap); + if (bg) + ret = !ocfs2_test_bit(nr, (unsigned long *)bg->bg_bitmap); + else + ret = 1; + jbd_unlock_bh_state(bg_bh); + + return ret; } static int ocfs2_block_group_find_clear_bits(struct ocfs2_super *osb, @@ -1885,6 +1894,7 @@ static inline int ocfs2_block_group_clear_bits(handle_t *handle, unsigned int tmp; int journal_type = OCFS2_JOURNAL_ACCESS_WRITE; struct ocfs2_group_desc *undo_bg = NULL; + int cluster_bitmap = 0; mlog_entry_void(); @@ -1905,18 +1915,28 @@ static inline int ocfs2_block_group_clear_bits(handle_t *handle, } if (ocfs2_is_cluster_bitmap(alloc_inode)) - undo_bg = (struct ocfs2_group_desc *) bh2jh(group_bh)->b_committed_data; + cluster_bitmap = 1; + + if (cluster_bitmap) { + jbd_lock_bh_state(group_bh); + undo_bg = (struct ocfs2_group_desc *) + bh2jh(group_bh)->b_committed_data; + BUG_ON(!undo_bg); + } tmp = num_bits; while(tmp--) { ocfs2_clear_bit((bit_off + tmp), (unsigned long *) bg->bg_bitmap); - if (ocfs2_is_cluster_bitmap(alloc_inode)) + if (cluster_bitmap) ocfs2_set_bit(bit_off + tmp, (unsigned long *) undo_bg->bg_bitmap); } le16_add_cpu(&bg->bg_free_bits_count, num_bits); + if (cluster_bitmap) + jbd_unlock_bh_state(group_bh); + status = ocfs2_journal_dirty(handle, group_bh); if (status < 0) mlog_errno(status); -- cgit v1.2.3 From c3d38840abaa45c1c5a5fabbb8ffc9a0d1a764d1 Mon Sep 17 00:00:00 2001 From: Sunil Mushran Date: Fri, 19 Jun 2009 14:45:55 -0700 Subject: ocfs2: Fix ocfs2_osb_dump() Skip printing information that is not valid for local mounts. Signed-off-by: Sunil Mushran Signed-off-by: Joel Becker --- fs/ocfs2/super.c | 32 ++++++++++++++++++-------------- 1 file changed, 18 insertions(+), 14 deletions(-) (limited to 'fs') diff --git a/fs/ocfs2/super.c b/fs/ocfs2/super.c index d33767f17ba3..d64739b593e9 100644 --- a/fs/ocfs2/super.c +++ b/fs/ocfs2/super.c @@ -234,20 +234,24 @@ static int ocfs2_osb_dump(struct ocfs2_super *osb, char *buf, int len) "%10s => Opts: 0x%lX AtimeQuanta: %u\n", "Mount", osb->s_mount_opt, osb->s_atime_quantum); - out += snprintf(buf + out, len - out, - "%10s => Stack: %s Name: %*s Version: %d.%d\n", - "Cluster", - (*osb->osb_cluster_stack == '\0' ? - "o2cb" : osb->osb_cluster_stack), - cconn->cc_namelen, cconn->cc_name, - cconn->cc_version.pv_major, cconn->cc_version.pv_minor); + if (cconn) { + out += snprintf(buf + out, len - out, + "%10s => Stack: %s Name: %*s " + "Version: %d.%d\n", "Cluster", + (*osb->osb_cluster_stack == '\0' ? + "o2cb" : osb->osb_cluster_stack), + cconn->cc_namelen, cconn->cc_name, + cconn->cc_version.pv_major, + cconn->cc_version.pv_minor); + } spin_lock(&osb->dc_task_lock); out += snprintf(buf + out, len - out, "%10s => Pid: %d Count: %lu WakeSeq: %lu " "WorkSeq: %lu\n", "DownCnvt", - task_pid_nr(osb->dc_task), osb->blocked_lock_count, - osb->dc_wake_sequence, osb->dc_work_sequence); + (osb->dc_task ? task_pid_nr(osb->dc_task) : -1), + osb->blocked_lock_count, osb->dc_wake_sequence, + osb->dc_work_sequence); spin_unlock(&osb->dc_task_lock); spin_lock(&osb->osb_lock); @@ -267,14 +271,15 @@ static int ocfs2_osb_dump(struct ocfs2_super *osb, char *buf, int len) out += snprintf(buf + out, len - out, "%10s => Pid: %d Interval: %lu Needs: %d\n", "Commit", - task_pid_nr(osb->commit_task), osb->osb_commit_interval, + (osb->commit_task ? task_pid_nr(osb->commit_task) : -1), + osb->osb_commit_interval, atomic_read(&osb->needs_checkpoint)); out += snprintf(buf + out, len - out, - "%10s => State: %d NumTxns: %d TxnId: %lu\n", + "%10s => State: %d TxnId: %lu NumTxns: %d\n", "Journal", osb->journal->j_state, - atomic_read(&osb->journal->j_num_trans), - osb->journal->j_trans_id); + osb->journal->j_trans_id, + atomic_read(&osb->journal->j_num_trans)); out += snprintf(buf + out, len - out, "%10s => GlobalAllocs: %d LocalAllocs: %d " @@ -302,7 +307,6 @@ static int ocfs2_osb_dump(struct ocfs2_super *osb, char *buf, int len) out += snprintf(buf + out, len - out, "%10s => %3s %10s\n", "Slots", "Num", "RecoGen"); - for (i = 0; i < osb->max_slots; ++i) { out += snprintf(buf + out, len - out, "%10s %c %3d %10d\n", -- cgit v1.2.3 From 692684e19e317a374c18e70a44d6413e51f71c11 Mon Sep 17 00:00:00 2001 From: Sunil Mushran Date: Fri, 19 Jun 2009 16:53:17 -0700 Subject: ocfs2: Stop orphan scan as early as possible during umount Currently if the orphan scan fires a tick before the user issues the umount, the umount will wait for the queued orphan scan tasks to complete. This patch makes the umount stop the orphan scan as early as possible so as to reduce the probability of the queued tasks slowing down the umount. Signed-off-by: Sunil Mushran Signed-off-by: Joel Becker --- fs/ocfs2/journal.c | 14 ++++++++++++-- fs/ocfs2/ocfs2.h | 6 ++++++ fs/ocfs2/super.c | 5 +++-- 3 files changed, 21 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/ocfs2/journal.c b/fs/ocfs2/journal.c index 4a3b9e6b31ad..70215a21fb20 100644 --- a/fs/ocfs2/journal.c +++ b/fs/ocfs2/journal.c @@ -1880,6 +1880,9 @@ void ocfs2_queue_orphan_scan(struct ocfs2_super *osb) os = &osb->osb_orphan_scan; + if (atomic_read(&os->os_state) == ORPHAN_SCAN_INACTIVE) + goto out; + status = ocfs2_orphan_scan_lock(osb, &seqno, DLM_LOCK_EX); if (status < 0) { if (status != -EAGAIN) @@ -1887,6 +1890,10 @@ void ocfs2_queue_orphan_scan(struct ocfs2_super *osb) goto out; } + /* Do no queue the tasks if the volume is being umounted */ + if (atomic_read(&os->os_state) == ORPHAN_SCAN_INACTIVE) + goto unlock; + if (os->os_seqno != seqno) { os->os_seqno = seqno; goto unlock; @@ -1920,8 +1927,9 @@ void ocfs2_orphan_scan_work(struct work_struct *work) mutex_lock(&os->os_lock); ocfs2_queue_orphan_scan(osb); - schedule_delayed_work(&os->os_orphan_scan_work, - ocfs2_orphan_scan_timeout()); + if (atomic_read(&os->os_state) == ORPHAN_SCAN_ACTIVE) + schedule_delayed_work(&os->os_orphan_scan_work, + ocfs2_orphan_scan_timeout()); mutex_unlock(&os->os_lock); } @@ -1930,6 +1938,7 @@ void ocfs2_orphan_scan_stop(struct ocfs2_super *osb) struct ocfs2_orphan_scan *os; os = &osb->osb_orphan_scan; + atomic_set(&os->os_state, ORPHAN_SCAN_INACTIVE); mutex_lock(&os->os_lock); cancel_delayed_work(&os->os_orphan_scan_work); mutex_unlock(&os->os_lock); @@ -1940,6 +1949,7 @@ int ocfs2_orphan_scan_init(struct ocfs2_super *osb) struct ocfs2_orphan_scan *os; os = &osb->osb_orphan_scan; + atomic_set(&os->os_state, ORPHAN_SCAN_ACTIVE); os->os_osb = osb; os->os_count = 0; os->os_scantime = CURRENT_TIME; diff --git a/fs/ocfs2/ocfs2.h b/fs/ocfs2/ocfs2.h index 18c1d9ec1c93..60e89503ce5a 100644 --- a/fs/ocfs2/ocfs2.h +++ b/fs/ocfs2/ocfs2.h @@ -154,6 +154,11 @@ struct ocfs2_lock_res { #endif }; +enum ocfs2_orphan_scan_state { + ORPHAN_SCAN_ACTIVE, + ORPHAN_SCAN_INACTIVE +}; + struct ocfs2_orphan_scan { struct mutex os_lock; struct ocfs2_super *os_osb; @@ -162,6 +167,7 @@ struct ocfs2_orphan_scan { struct timespec os_scantime; /* time this node ran the scan */ u32 os_count; /* tracks node specific scans */ u32 os_seqno; /* tracks cluster wide scans */ + atomic_t os_state; /* ACTIVE or INACTIVE */ }; struct ocfs2_dlm_debug { diff --git a/fs/ocfs2/super.c b/fs/ocfs2/super.c index d64739b593e9..3e8a68b103ab 100644 --- a/fs/ocfs2/super.c +++ b/fs/ocfs2/super.c @@ -1814,14 +1814,15 @@ static void ocfs2_dismount_volume(struct super_block *sb, int mnt_err) debugfs_remove(osb->osb_ctxt); + /* Orphan scan should be stopped as early as possible */ + ocfs2_orphan_scan_stop(osb); + ocfs2_disable_quotas(osb); ocfs2_shutdown_local_alloc(osb); ocfs2_truncate_log_shutdown(osb); - ocfs2_orphan_scan_stop(osb); - /* This will disable recovery and flush any recovery work. */ ocfs2_recovery_exit(osb); -- cgit v1.2.3 From 3211949f8998dde71d9fe2e063de045ece5e0473 Mon Sep 17 00:00:00 2001 From: Sunil Mushran Date: Fri, 19 Jun 2009 16:53:18 -0700 Subject: ocfs2: Do not initialize lvb in ocfs2_orphan_scan_lock_res_init() We don't access the LVB in our ocfs2_*_lock_res_init() functions. Since the LVB can become invalid during some cluster recovery operations, the dlmglue must be able to handle an uninitialized LVB. For the orphan scan lock, we initialized an uninitialzed LVB with our scan sequence number plus one. This starts a normal orphan scan cycle. Signed-off-by: Sunil Mushran Signed-off-by: Joel Becker --- fs/ocfs2/dlmglue.c | 7 +++---- fs/ocfs2/journal.c | 1 + 2 files changed, 4 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/ocfs2/dlmglue.c b/fs/ocfs2/dlmglue.c index 83d2ddb27186..d3d1f9372f7e 100644 --- a/fs/ocfs2/dlmglue.c +++ b/fs/ocfs2/dlmglue.c @@ -644,14 +644,10 @@ static void ocfs2_nfs_sync_lock_res_init(struct ocfs2_lock_res *res, static void ocfs2_orphan_scan_lock_res_init(struct ocfs2_lock_res *res, struct ocfs2_super *osb) { - struct ocfs2_orphan_scan_lvb *lvb; - ocfs2_lock_res_init_once(res); ocfs2_build_lock_name(OCFS2_LOCK_TYPE_ORPHAN_SCAN, 0, 0, res->l_name); ocfs2_lock_res_init_common(osb, res, OCFS2_LOCK_TYPE_ORPHAN_SCAN, &ocfs2_orphan_scan_lops, osb); - lvb = ocfs2_dlm_lvb(&res->l_lksb); - lvb->lvb_version = OCFS2_ORPHAN_LVB_VERSION; } void ocfs2_file_lock_res_init(struct ocfs2_lock_res *lockres, @@ -2386,6 +2382,9 @@ int ocfs2_orphan_scan_lock(struct ocfs2_super *osb, u32 *seqno, int ex) if (ocfs2_dlm_lvb_valid(&lockres->l_lksb) && lvb->lvb_version == OCFS2_ORPHAN_LVB_VERSION) *seqno = be32_to_cpu(lvb->lvb_os_seqno); + else + *seqno = osb->osb_orphan_scan.os_seqno + 1; + return status; } diff --git a/fs/ocfs2/journal.c b/fs/ocfs2/journal.c index 70215a21fb20..0b2c27a9485e 100644 --- a/fs/ocfs2/journal.c +++ b/fs/ocfs2/journal.c @@ -1952,6 +1952,7 @@ int ocfs2_orphan_scan_init(struct ocfs2_super *osb) atomic_set(&os->os_state, ORPHAN_SCAN_ACTIVE); os->os_osb = osb; os->os_count = 0; + os->os_seqno = 0; os->os_scantime = CURRENT_TIME; mutex_init(&os->os_lock); -- cgit v1.2.3 From df152c241df9e9d2b9a65d37bd02961abe7f591a Mon Sep 17 00:00:00 2001 From: Sunil Mushran Date: Mon, 22 Jun 2009 11:40:07 -0700 Subject: ocfs2: Disable orphan scanning for local and hard-ro mounts Local and Hard-RO mounts do not need orphan scanning. Signed-off-by: Sunil Mushran Signed-off-by: Joel Becker --- fs/ocfs2/dlmglue.c | 26 ++++++++++++++++---------- fs/ocfs2/dlmglue.h | 4 ++-- fs/ocfs2/journal.c | 30 +++++++++++++++++------------- fs/ocfs2/journal.h | 2 +- fs/ocfs2/super.c | 32 +++++++++++++++----------------- 5 files changed, 51 insertions(+), 43 deletions(-) (limited to 'fs') diff --git a/fs/ocfs2/dlmglue.c b/fs/ocfs2/dlmglue.c index d3d1f9372f7e..1841bbb49cb6 100644 --- a/fs/ocfs2/dlmglue.c +++ b/fs/ocfs2/dlmglue.c @@ -2366,15 +2366,20 @@ void ocfs2_inode_unlock(struct inode *inode, mlog_exit_void(); } -int ocfs2_orphan_scan_lock(struct ocfs2_super *osb, u32 *seqno, int ex) +int ocfs2_orphan_scan_lock(struct ocfs2_super *osb, u32 *seqno) { struct ocfs2_lock_res *lockres; struct ocfs2_orphan_scan_lvb *lvb; - int level = ex ? DLM_LOCK_EX : DLM_LOCK_PR; int status = 0; + if (ocfs2_is_hard_readonly(osb)) + return -EROFS; + + if (ocfs2_mount_local(osb)) + return 0; + lockres = &osb->osb_orphan_scan.os_lockres; - status = ocfs2_cluster_lock(osb, lockres, level, 0, 0); + status = ocfs2_cluster_lock(osb, lockres, DLM_LOCK_EX, 0, 0); if (status < 0) return status; @@ -2388,17 +2393,18 @@ int ocfs2_orphan_scan_lock(struct ocfs2_super *osb, u32 *seqno, int ex) return status; } -void ocfs2_orphan_scan_unlock(struct ocfs2_super *osb, u32 seqno, int ex) +void ocfs2_orphan_scan_unlock(struct ocfs2_super *osb, u32 seqno) { struct ocfs2_lock_res *lockres; struct ocfs2_orphan_scan_lvb *lvb; - int level = ex ? DLM_LOCK_EX : DLM_LOCK_PR; - lockres = &osb->osb_orphan_scan.os_lockres; - lvb = ocfs2_dlm_lvb(&lockres->l_lksb); - lvb->lvb_version = OCFS2_ORPHAN_LVB_VERSION; - lvb->lvb_os_seqno = cpu_to_be32(seqno); - ocfs2_cluster_unlock(osb, lockres, level); + if (!ocfs2_is_hard_readonly(osb) && !ocfs2_mount_local(osb)) { + lockres = &osb->osb_orphan_scan.os_lockres; + lvb = ocfs2_dlm_lvb(&lockres->l_lksb); + lvb->lvb_version = OCFS2_ORPHAN_LVB_VERSION; + lvb->lvb_os_seqno = cpu_to_be32(seqno); + ocfs2_cluster_unlock(osb, lockres, DLM_LOCK_EX); + } } int ocfs2_super_lock(struct ocfs2_super *osb, diff --git a/fs/ocfs2/dlmglue.h b/fs/ocfs2/dlmglue.h index 31b90d7b8f51..30f683107f1e 100644 --- a/fs/ocfs2/dlmglue.h +++ b/fs/ocfs2/dlmglue.h @@ -121,8 +121,8 @@ int ocfs2_super_lock(struct ocfs2_super *osb, int ex); void ocfs2_super_unlock(struct ocfs2_super *osb, int ex); -int ocfs2_orphan_scan_lock(struct ocfs2_super *osb, u32 *seqno, int ex); -void ocfs2_orphan_scan_unlock(struct ocfs2_super *osb, u32 seqno, int ex); +int ocfs2_orphan_scan_lock(struct ocfs2_super *osb, u32 *seqno); +void ocfs2_orphan_scan_unlock(struct ocfs2_super *osb, u32 seqno); int ocfs2_rename_lock(struct ocfs2_super *osb); void ocfs2_rename_unlock(struct ocfs2_super *osb); diff --git a/fs/ocfs2/journal.c b/fs/ocfs2/journal.c index 0b2c27a9485e..f033760ecbea 100644 --- a/fs/ocfs2/journal.c +++ b/fs/ocfs2/journal.c @@ -1883,7 +1883,7 @@ void ocfs2_queue_orphan_scan(struct ocfs2_super *osb) if (atomic_read(&os->os_state) == ORPHAN_SCAN_INACTIVE) goto out; - status = ocfs2_orphan_scan_lock(osb, &seqno, DLM_LOCK_EX); + status = ocfs2_orphan_scan_lock(osb, &seqno); if (status < 0) { if (status != -EAGAIN) mlog_errno(status); @@ -1910,7 +1910,7 @@ void ocfs2_queue_orphan_scan(struct ocfs2_super *osb) os->os_count++; os->os_scantime = CURRENT_TIME; unlock: - ocfs2_orphan_scan_unlock(osb, seqno, DLM_LOCK_EX); + ocfs2_orphan_scan_unlock(osb, seqno); out: return; } @@ -1938,29 +1938,33 @@ void ocfs2_orphan_scan_stop(struct ocfs2_super *osb) struct ocfs2_orphan_scan *os; os = &osb->osb_orphan_scan; - atomic_set(&os->os_state, ORPHAN_SCAN_INACTIVE); - mutex_lock(&os->os_lock); - cancel_delayed_work(&os->os_orphan_scan_work); - mutex_unlock(&os->os_lock); + if (atomic_read(&os->os_state) == ORPHAN_SCAN_ACTIVE) { + atomic_set(&os->os_state, ORPHAN_SCAN_INACTIVE); + mutex_lock(&os->os_lock); + cancel_delayed_work(&os->os_orphan_scan_work); + mutex_unlock(&os->os_lock); + } } -int ocfs2_orphan_scan_init(struct ocfs2_super *osb) +void ocfs2_orphan_scan_init(struct ocfs2_super *osb) { struct ocfs2_orphan_scan *os; os = &osb->osb_orphan_scan; - atomic_set(&os->os_state, ORPHAN_SCAN_ACTIVE); os->os_osb = osb; os->os_count = 0; os->os_seqno = 0; os->os_scantime = CURRENT_TIME; mutex_init(&os->os_lock); + INIT_DELAYED_WORK(&os->os_orphan_scan_work, ocfs2_orphan_scan_work); - INIT_DELAYED_WORK(&os->os_orphan_scan_work, - ocfs2_orphan_scan_work); - schedule_delayed_work(&os->os_orphan_scan_work, - ocfs2_orphan_scan_timeout()); - return 0; + if (ocfs2_is_hard_readonly(osb) || ocfs2_mount_local(osb)) + atomic_set(&os->os_state, ORPHAN_SCAN_INACTIVE); + else { + atomic_set(&os->os_state, ORPHAN_SCAN_ACTIVE); + schedule_delayed_work(&os->os_orphan_scan_work, + ocfs2_orphan_scan_timeout()); + } } struct ocfs2_orphan_filldir_priv { diff --git a/fs/ocfs2/journal.h b/fs/ocfs2/journal.h index 61045eeb3f6e..5432c7f79cc6 100644 --- a/fs/ocfs2/journal.h +++ b/fs/ocfs2/journal.h @@ -144,7 +144,7 @@ static inline void ocfs2_inode_set_new(struct ocfs2_super *osb, } /* Exported only for the journal struct init code in super.c. Do not call. */ -int ocfs2_orphan_scan_init(struct ocfs2_super *osb); +void ocfs2_orphan_scan_init(struct ocfs2_super *osb); void ocfs2_orphan_scan_stop(struct ocfs2_super *osb); void ocfs2_orphan_scan_exit(struct ocfs2_super *osb); diff --git a/fs/ocfs2/super.c b/fs/ocfs2/super.c index 3e8a68b103ab..0746e1abdfc7 100644 --- a/fs/ocfs2/super.c +++ b/fs/ocfs2/super.c @@ -205,11 +205,10 @@ static const match_table_t tokens = { #ifdef CONFIG_DEBUG_FS static int ocfs2_osb_dump(struct ocfs2_super *osb, char *buf, int len) { - int out = 0; - int i; struct ocfs2_cluster_connection *cconn = osb->cconn; struct ocfs2_recovery_map *rm = osb->recovery_map; - struct ocfs2_orphan_scan *os; + struct ocfs2_orphan_scan *os = &osb->osb_orphan_scan; + int i, out = 0; out += snprintf(buf + out, len - out, "%10s => Id: %-s Uuid: %-s Gen: 0x%X Label: %-s\n", @@ -305,6 +304,16 @@ static int ocfs2_osb_dump(struct ocfs2_super *osb, char *buf, int len) atomic_read(&osb->s_num_inodes_stolen)); spin_unlock(&osb->osb_lock); + out += snprintf(buf + out, len - out, "OrphanScan => "); + out += snprintf(buf + out, len - out, "Local: %u Global: %u ", + os->os_count, os->os_seqno); + out += snprintf(buf + out, len - out, " Last Scan: "); + if (atomic_read(&os->os_state) == ORPHAN_SCAN_INACTIVE) + out += snprintf(buf + out, len - out, "Disabled\n"); + else + out += snprintf(buf + out, len - out, "%lu seconds ago\n", + (get_seconds() - os->os_scantime.tv_sec)); + out += snprintf(buf + out, len - out, "%10s => %3s %10s\n", "Slots", "Num", "RecoGen"); for (i = 0; i < osb->max_slots; ++i) { @@ -315,13 +324,6 @@ static int ocfs2_osb_dump(struct ocfs2_super *osb, char *buf, int len) i, osb->slot_recovery_generations[i]); } - os = &osb->osb_orphan_scan; - out += snprintf(buf + out, len - out, "Orphan Scan=> "); - out += snprintf(buf + out, len - out, "Local: %u Global: %u ", - os->os_count, os->os_seqno); - out += snprintf(buf + out, len - out, " Last Scan: %lu seconds ago\n", - (get_seconds() - os->os_scantime.tv_sec)); - return out; } @@ -1179,6 +1181,9 @@ static int ocfs2_fill_super(struct super_block *sb, void *data, int silent) atomic_set(&osb->vol_state, VOLUME_MOUNTED_QUOTAS); wake_up(&osb->osb_mount_event); + /* Start this when the mount is almost sure of being successful */ + ocfs2_orphan_scan_init(osb); + mlog_exit(status); return status; @@ -1983,13 +1988,6 @@ static int ocfs2_initialize_super(struct super_block *sb, goto bail; } - status = ocfs2_orphan_scan_init(osb); - if (status) { - mlog(ML_ERROR, "Unable to initialize delayed orphan scan\n"); - mlog_errno(status); - goto bail; - } - init_waitqueue_head(&osb->checkpoint_event); atomic_set(&osb->needs_checkpoint, 0); -- cgit v1.2.3 From 9a7aa12f3911853a3574d47d567b81a2a5df7208 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Thu, 4 Jun 2009 15:26:49 +0200 Subject: vfs: Set special lockdep map for dirs only if not set by fs Some filesystems need to set lockdep map for i_mutex differently for different directories. For example OCFS2 has system directories (for orphan inode tracking and for gathering all system files like journal or quota files into a single place) which have different locking locking rules than standard directories. For a filesystem setting lockdep map is naturaly done when the inode is read but we have to modify unlock_new_inode() not to overwrite the lockdep map the filesystem has set. Acked-by: peterz@infradead.org CC: mingo@redhat.com Signed-off-by: Jan Kara Signed-off-by: Joel Becker --- fs/inode.c | 17 +++++++++++------ include/linux/lockdep.h | 15 +++++++++++++++ 2 files changed, 26 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/inode.c b/fs/inode.c index f643be565df8..04c785bb63c3 100644 --- a/fs/inode.c +++ b/fs/inode.c @@ -665,12 +665,17 @@ void unlock_new_inode(struct inode *inode) if (inode->i_mode & S_IFDIR) { struct file_system_type *type = inode->i_sb->s_type; - /* - * ensure nobody is actually holding i_mutex - */ - mutex_destroy(&inode->i_mutex); - mutex_init(&inode->i_mutex); - lockdep_set_class(&inode->i_mutex, &type->i_mutex_dir_key); + /* Set new key only if filesystem hasn't already changed it */ + if (!lockdep_match_class(&inode->i_mutex, + &type->i_mutex_key)) { + /* + * ensure nobody is actually holding i_mutex + */ + mutex_destroy(&inode->i_mutex); + mutex_init(&inode->i_mutex); + lockdep_set_class(&inode->i_mutex, + &type->i_mutex_dir_key); + } } #endif /* diff --git a/include/linux/lockdep.h b/include/linux/lockdep.h index da5a5a1f4cd2..b25d1b53df0d 100644 --- a/include/linux/lockdep.h +++ b/include/linux/lockdep.h @@ -258,6 +258,16 @@ extern void lockdep_init_map(struct lockdep_map *lock, const char *name, #define lockdep_set_subclass(lock, sub) \ lockdep_init_map(&(lock)->dep_map, #lock, \ (lock)->dep_map.key, sub) +/* + * Compare locking classes + */ +#define lockdep_match_class(lock, key) lockdep_match_key(&(lock)->dep_map, key) + +static inline int lockdep_match_key(struct lockdep_map *lock, + struct lock_class_key *key) +{ + return lock->key == key; +} /* * Acquire a lock. @@ -326,6 +336,11 @@ static inline void lockdep_on(void) #define lockdep_set_class_and_subclass(lock, key, sub) \ do { (void)(key); } while (0) #define lockdep_set_subclass(lock, sub) do { } while (0) +/* + * We don't define lockdep_match_class() and lockdep_match_key() for !LOCKDEP + * case since the result is not well defined and the caller should rather + * #ifdef the call himself. + */ # define INIT_LOCKDEP # define lockdep_reset() do { debug_locks = 1; } while (0) -- cgit v1.2.3 From cb25797d451dc774d9dbc402a65f16a0e32199fe Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Thu, 4 Jun 2009 15:26:50 +0200 Subject: ocfs2: Add lockdep annotations Add lockdep support to OCFS2. The support also covers all of the cluster locks except for open locks, journal locks, and local quotafile locks. These are special because they are acquired for a node, not for a particular process and lockdep cannot deal with such type of locking. Signed-off-by: Jan Kara Signed-off-by: Joel Becker --- fs/ocfs2/dlmglue.c | 81 +++++++++++++++++++++++++++++++++++++++++++----------- fs/ocfs2/dlmglue.h | 20 ++++++++++++-- fs/ocfs2/inode.c | 11 ++++++++ fs/ocfs2/namei.c | 15 ++++++---- fs/ocfs2/ocfs2.h | 4 +++ fs/ocfs2/sysfile.c | 17 ++++++++++++ 6 files changed, 123 insertions(+), 25 deletions(-) (limited to 'fs') diff --git a/fs/ocfs2/dlmglue.c b/fs/ocfs2/dlmglue.c index 1841bbb49cb6..110bb57c46ab 100644 --- a/fs/ocfs2/dlmglue.c +++ b/fs/ocfs2/dlmglue.c @@ -92,6 +92,9 @@ struct ocfs2_unblock_ctl { enum ocfs2_unblock_action unblock_action; }; +/* Lockdep class keys */ +struct lock_class_key lockdep_keys[OCFS2_NUM_LOCK_TYPES]; + static int ocfs2_check_meta_downconvert(struct ocfs2_lock_res *lockres, int new_level); static void ocfs2_set_meta_lvb(struct ocfs2_lock_res *lockres); @@ -317,9 +320,16 @@ static int ocfs2_lock_create(struct ocfs2_super *osb, u32 dlm_flags); static inline int ocfs2_may_continue_on_blocked_lock(struct ocfs2_lock_res *lockres, int wanted); -static void ocfs2_cluster_unlock(struct ocfs2_super *osb, - struct ocfs2_lock_res *lockres, - int level); +static void __ocfs2_cluster_unlock(struct ocfs2_super *osb, + struct ocfs2_lock_res *lockres, + int level, unsigned long caller_ip); +static inline void ocfs2_cluster_unlock(struct ocfs2_super *osb, + struct ocfs2_lock_res *lockres, + int level) +{ + __ocfs2_cluster_unlock(osb, lockres, level, _RET_IP_); +} + static inline void ocfs2_generic_handle_downconvert_action(struct ocfs2_lock_res *lockres); static inline void ocfs2_generic_handle_convert_action(struct ocfs2_lock_res *lockres); static inline void ocfs2_generic_handle_attach_action(struct ocfs2_lock_res *lockres); @@ -489,6 +499,13 @@ static void ocfs2_lock_res_init_common(struct ocfs2_super *osb, ocfs2_add_lockres_tracking(res, osb->osb_dlm_debug); ocfs2_init_lock_stats(res); +#ifdef CONFIG_DEBUG_LOCK_ALLOC + if (type != OCFS2_LOCK_TYPE_OPEN) + lockdep_init_map(&res->l_lockdep_map, ocfs2_lock_type_strings[type], + &lockdep_keys[type], 0); + else + res->l_lockdep_map.key = NULL; +#endif } void ocfs2_lock_res_init_once(struct ocfs2_lock_res *res) @@ -1252,11 +1269,13 @@ static int ocfs2_wait_for_mask_interruptible(struct ocfs2_mask_waiter *mw, return ret; } -static int ocfs2_cluster_lock(struct ocfs2_super *osb, - struct ocfs2_lock_res *lockres, - int level, - u32 lkm_flags, - int arg_flags) +static int __ocfs2_cluster_lock(struct ocfs2_super *osb, + struct ocfs2_lock_res *lockres, + int level, + u32 lkm_flags, + int arg_flags, + int l_subclass, + unsigned long caller_ip) { struct ocfs2_mask_waiter mw; int wait, catch_signals = !(osb->s_mount_opt & OCFS2_MOUNT_NOINTR); @@ -1399,13 +1418,37 @@ out: } ocfs2_update_lock_stats(lockres, level, &mw, ret); +#ifdef CONFIG_DEBUG_LOCK_ALLOC + if (!ret && lockres->l_lockdep_map.key != NULL) { + if (level == DLM_LOCK_PR) + rwsem_acquire_read(&lockres->l_lockdep_map, l_subclass, + !!(arg_flags & OCFS2_META_LOCK_NOQUEUE), + caller_ip); + else + rwsem_acquire(&lockres->l_lockdep_map, l_subclass, + !!(arg_flags & OCFS2_META_LOCK_NOQUEUE), + caller_ip); + } +#endif mlog_exit(ret); return ret; } -static void ocfs2_cluster_unlock(struct ocfs2_super *osb, - struct ocfs2_lock_res *lockres, - int level) +static inline int ocfs2_cluster_lock(struct ocfs2_super *osb, + struct ocfs2_lock_res *lockres, + int level, + u32 lkm_flags, + int arg_flags) +{ + return __ocfs2_cluster_lock(osb, lockres, level, lkm_flags, arg_flags, + 0, _RET_IP_); +} + + +static void __ocfs2_cluster_unlock(struct ocfs2_super *osb, + struct ocfs2_lock_res *lockres, + int level, + unsigned long caller_ip) { unsigned long flags; @@ -1414,6 +1457,10 @@ static void ocfs2_cluster_unlock(struct ocfs2_super *osb, ocfs2_dec_holders(lockres, level); ocfs2_downconvert_on_unlock(osb, lockres); spin_unlock_irqrestore(&lockres->l_lock, flags); +#ifdef CONFIG_DEBUG_LOCK_ALLOC + if (lockres->l_lockdep_map.key != NULL) + rwsem_release(&lockres->l_lockdep_map, 1, caller_ip); +#endif mlog_exit_void(); } @@ -2159,10 +2206,11 @@ static int ocfs2_assign_bh(struct inode *inode, * returns < 0 error if the callback will never be called, otherwise * the result of the lock will be communicated via the callback. */ -int ocfs2_inode_lock_full(struct inode *inode, - struct buffer_head **ret_bh, - int ex, - int arg_flags) +int ocfs2_inode_lock_full_nested(struct inode *inode, + struct buffer_head **ret_bh, + int ex, + int arg_flags, + int subclass) { int status, level, acquired; u32 dlm_flags; @@ -2200,7 +2248,8 @@ int ocfs2_inode_lock_full(struct inode *inode, if (arg_flags & OCFS2_META_LOCK_NOQUEUE) dlm_flags |= DLM_LKF_NOQUEUE; - status = ocfs2_cluster_lock(osb, lockres, level, dlm_flags, arg_flags); + status = __ocfs2_cluster_lock(osb, lockres, level, dlm_flags, + arg_flags, subclass, _RET_IP_); if (status < 0) { if (status != -EAGAIN && status != -EIOCBRETRY) mlog_errno(status); diff --git a/fs/ocfs2/dlmglue.h b/fs/ocfs2/dlmglue.h index 30f683107f1e..7553836931de 100644 --- a/fs/ocfs2/dlmglue.h +++ b/fs/ocfs2/dlmglue.h @@ -78,6 +78,14 @@ struct ocfs2_orphan_scan_lvb { /* don't block waiting for the downconvert thread, instead return -EAGAIN */ #define OCFS2_LOCK_NONBLOCK (0x04) +/* Locking subclasses of inode cluster lock */ +enum { + OI_LS_NORMAL = 0, + OI_LS_PARENT, + OI_LS_RENAME1, + OI_LS_RENAME2, +}; + int ocfs2_dlm_init(struct ocfs2_super *osb); void ocfs2_dlm_shutdown(struct ocfs2_super *osb, int hangup_pending); void ocfs2_lock_res_init_once(struct ocfs2_lock_res *res); @@ -104,17 +112,23 @@ void ocfs2_open_unlock(struct inode *inode); int ocfs2_inode_lock_atime(struct inode *inode, struct vfsmount *vfsmnt, int *level); -int ocfs2_inode_lock_full(struct inode *inode, +int ocfs2_inode_lock_full_nested(struct inode *inode, struct buffer_head **ret_bh, int ex, - int arg_flags); + int arg_flags, + int subclass); int ocfs2_inode_lock_with_page(struct inode *inode, struct buffer_head **ret_bh, int ex, struct page *page); +/* Variants without special locking class or flags */ +#define ocfs2_inode_lock_full(i, r, e, f)\ + ocfs2_inode_lock_full_nested(i, r, e, f, OI_LS_NORMAL) +#define ocfs2_inode_lock_nested(i, b, e, s)\ + ocfs2_inode_lock_full_nested(i, b, e, 0, s) /* 99% of the time we don't want to supply any additional flags -- * those are for very specific cases only. */ -#define ocfs2_inode_lock(i, b, e) ocfs2_inode_lock_full(i, b, e, 0) +#define ocfs2_inode_lock(i, b, e) ocfs2_inode_lock_full_nested(i, b, e, 0, OI_LS_NORMAL) void ocfs2_inode_unlock(struct inode *inode, int ex); int ocfs2_super_lock(struct ocfs2_super *osb, diff --git a/fs/ocfs2/inode.c b/fs/ocfs2/inode.c index 10e1fa87396a..4dc8890ba316 100644 --- a/fs/ocfs2/inode.c +++ b/fs/ocfs2/inode.c @@ -215,6 +215,8 @@ bail: static int ocfs2_init_locked_inode(struct inode *inode, void *opaque) { struct ocfs2_find_inode_args *args = opaque; + static struct lock_class_key ocfs2_quota_ip_alloc_sem_key, + ocfs2_file_ip_alloc_sem_key; mlog_entry("inode = %p, opaque = %p\n", inode, opaque); @@ -223,6 +225,15 @@ static int ocfs2_init_locked_inode(struct inode *inode, void *opaque) if (args->fi_sysfile_type != 0) lockdep_set_class(&inode->i_mutex, &ocfs2_sysfile_lock_key[args->fi_sysfile_type]); + if (args->fi_sysfile_type == USER_QUOTA_SYSTEM_INODE || + args->fi_sysfile_type == GROUP_QUOTA_SYSTEM_INODE || + args->fi_sysfile_type == LOCAL_USER_QUOTA_SYSTEM_INODE || + args->fi_sysfile_type == LOCAL_GROUP_QUOTA_SYSTEM_INODE) + lockdep_set_class(&OCFS2_I(inode)->ip_alloc_sem, + &ocfs2_quota_ip_alloc_sem_key); + else + lockdep_set_class(&OCFS2_I(inode)->ip_alloc_sem, + &ocfs2_file_ip_alloc_sem_key); mlog_exit(0); return 0; diff --git a/fs/ocfs2/namei.c b/fs/ocfs2/namei.c index 33464c6b60a2..8601f934010b 100644 --- a/fs/ocfs2/namei.c +++ b/fs/ocfs2/namei.c @@ -118,7 +118,7 @@ static struct dentry *ocfs2_lookup(struct inode *dir, struct dentry *dentry, mlog(0, "find name %.*s in directory %llu\n", dentry->d_name.len, dentry->d_name.name, (unsigned long long)OCFS2_I(dir)->ip_blkno); - status = ocfs2_inode_lock(dir, NULL, 0); + status = ocfs2_inode_lock_nested(dir, NULL, 0, OI_LS_PARENT); if (status < 0) { if (status != -ENOENT) mlog_errno(status); @@ -636,7 +636,7 @@ static int ocfs2_link(struct dentry *old_dentry, if (S_ISDIR(inode->i_mode)) return -EPERM; - err = ocfs2_inode_lock(dir, &parent_fe_bh, 1); + err = ocfs2_inode_lock_nested(dir, &parent_fe_bh, 1, OI_LS_PARENT); if (err < 0) { if (err != -ENOENT) mlog_errno(err); @@ -800,7 +800,8 @@ static int ocfs2_unlink(struct inode *dir, return -EPERM; } - status = ocfs2_inode_lock(dir, &parent_node_bh, 1); + status = ocfs2_inode_lock_nested(dir, &parent_node_bh, 1, + OI_LS_PARENT); if (status < 0) { if (status != -ENOENT) mlog_errno(status); @@ -978,7 +979,8 @@ static int ocfs2_double_lock(struct ocfs2_super *osb, inode1 = tmpinode; } /* lock id2 */ - status = ocfs2_inode_lock(inode2, bh2, 1); + status = ocfs2_inode_lock_nested(inode2, bh2, 1, + OI_LS_RENAME1); if (status < 0) { if (status != -ENOENT) mlog_errno(status); @@ -987,7 +989,7 @@ static int ocfs2_double_lock(struct ocfs2_super *osb, } /* lock id1 */ - status = ocfs2_inode_lock(inode1, bh1, 1); + status = ocfs2_inode_lock_nested(inode1, bh1, 1, OI_LS_RENAME2); if (status < 0) { /* * An error return must mean that no cluster locks @@ -1103,7 +1105,8 @@ static int ocfs2_rename(struct inode *old_dir, * won't have to concurrently downconvert the inode and the * dentry locks. */ - status = ocfs2_inode_lock(old_inode, &old_inode_bh, 1); + status = ocfs2_inode_lock_nested(old_inode, &old_inode_bh, 1, + OI_LS_PARENT); if (status < 0) { if (status != -ENOENT) mlog_errno(status); diff --git a/fs/ocfs2/ocfs2.h b/fs/ocfs2/ocfs2.h index 60e89503ce5a..c9345ebb8493 100644 --- a/fs/ocfs2/ocfs2.h +++ b/fs/ocfs2/ocfs2.h @@ -34,6 +34,7 @@ #include #include #include +#include #ifndef CONFIG_OCFS2_COMPAT_JBD # include #else @@ -152,6 +153,9 @@ struct ocfs2_lock_res { unsigned int l_lock_max_exmode; /* Max wait for EX */ unsigned int l_lock_refresh; /* Disk refreshes */ #endif +#ifdef CONFIG_DEBUG_LOCK_ALLOC + struct lockdep_map l_lockdep_map; +#endif }; enum ocfs2_orphan_scan_state { diff --git a/fs/ocfs2/sysfile.c b/fs/ocfs2/sysfile.c index ab713ebdd546..6f53f5e7256a 100644 --- a/fs/ocfs2/sysfile.c +++ b/fs/ocfs2/sysfile.c @@ -50,6 +50,8 @@ static inline int is_in_system_inode_array(struct ocfs2_super *osb, int type, u32 slot); +static struct lock_class_key ocfs2_sysfile_cluster_lock_key[NUM_SYSTEM_INODES]; + static inline int is_global_system_inode(int type) { return type >= OCFS2_FIRST_ONLINE_SYSTEM_INODE && @@ -118,6 +120,21 @@ static struct inode * _ocfs2_get_system_file_inode(struct ocfs2_super *osb, inode = NULL; goto bail; } +#ifdef CONFIG_DEBUG_LOCK_ALLOC + if (type == LOCAL_USER_QUOTA_SYSTEM_INODE || + type == LOCAL_GROUP_QUOTA_SYSTEM_INODE || + type == JOURNAL_SYSTEM_INODE) { + /* Ignore inode lock on these inodes as the lock does not + * really belong to any process and lockdep cannot handle + * that */ + OCFS2_I(inode)->ip_inode_lockres.l_lockdep_map.key = NULL; + } else { + lockdep_init_map(&OCFS2_I(inode)->ip_inode_lockres. + l_lockdep_map, + ocfs2_system_inodes[type].si_name, + &ocfs2_sysfile_cluster_lock_key[type], 0); + } +#endif bail: return inode; -- cgit v1.2.3 From d246ab307d1d003c80fe279897dea22bf52b6e41 Mon Sep 17 00:00:00 2001 From: Tao Ma Date: Thu, 18 Jun 2009 13:12:06 +0800 Subject: ocfs2/trivial: Wrap ocfs2_sysfile_cluster_lock_key within define. MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Actually ocfs2_sysfile_cluster_lock_key is only used if we enable CONFIG_DEBUG_LOCK_ALLOC. Wrap it so that we can avoid a building warning. fs/ocfs2/sysfile.c:53: warning: ‘ocfs2_sysfile_cluster_lock_key’ defined but not used Signed-off-by: Tao Ma Signed-off-by: Joel Becker --- fs/ocfs2/sysfile.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'fs') diff --git a/fs/ocfs2/sysfile.c b/fs/ocfs2/sysfile.c index 6f53f5e7256a..40e53702948c 100644 --- a/fs/ocfs2/sysfile.c +++ b/fs/ocfs2/sysfile.c @@ -50,7 +50,9 @@ static inline int is_in_system_inode_array(struct ocfs2_super *osb, int type, u32 slot); +#ifdef CONFIG_DEBUG_LOCK_ALLOC static struct lock_class_key ocfs2_sysfile_cluster_lock_key[NUM_SYSTEM_INODES]; +#endif static inline int is_global_system_inode(int type) { -- cgit v1.2.3 From 4839641333d4593bfc4fb29aa3af10d36f607d5b Mon Sep 17 00:00:00 2001 From: David Woodhouse Date: Tue, 23 Jun 2009 01:34:19 +0100 Subject: jffs2: fix another potential leak on error path in scan.c Signed-off-by: David Woodhouse --- fs/jffs2/scan.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/jffs2/scan.c b/fs/jffs2/scan.c index 7515e73e2bfb..696686cc206e 100644 --- a/fs/jffs2/scan.c +++ b/fs/jffs2/scan.c @@ -130,9 +130,9 @@ int jffs2_scan_medium(struct jffs2_sb_info *c) if (jffs2_sum_active()) { s = kzalloc(sizeof(struct jffs2_summary), GFP_KERNEL); if (!s) { - kfree(flashbuf); JFFS2_WARNING("Can't allocate memory for summary\n"); - return -ENOMEM; + ret = -ENOMEM; + goto out; } } -- cgit v1.2.3 From 616511d039af402670de8500d0e24495113a9cab Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Mon, 22 Jun 2009 15:09:13 -0400 Subject: VFS: Uninline the function put_mnt_ns() In order to allow modules to use it without having to export vfsmount_lock. Signed-off-by: Trond Myklebust Signed-off-by: Linus Torvalds --- fs/namespace.c | 8 ++++++-- include/linux/mnt_namespace.h | 9 +-------- 2 files changed, 7 insertions(+), 10 deletions(-) (limited to 'fs') diff --git a/fs/namespace.c b/fs/namespace.c index 2dd333b0fe7f..6645846f2056 100644 --- a/fs/namespace.c +++ b/fs/namespace.c @@ -2246,10 +2246,14 @@ void __init mnt_init(void) init_mount_tree(); } -void __put_mnt_ns(struct mnt_namespace *ns) +void put_mnt_ns(struct mnt_namespace *ns) { - struct vfsmount *root = ns->root; + struct vfsmount *root; LIST_HEAD(umount_list); + + if (!atomic_dec_and_lock(&ns->count, &vfsmount_lock)) + return; + root = ns->root; ns->root = NULL; spin_unlock(&vfsmount_lock); down_write(&namespace_sem); diff --git a/include/linux/mnt_namespace.h b/include/linux/mnt_namespace.h index 3a059298cc19..299d11af5f79 100644 --- a/include/linux/mnt_namespace.h +++ b/include/linux/mnt_namespace.h @@ -26,14 +26,7 @@ struct fs_struct; extern struct mnt_namespace *copy_mnt_ns(unsigned long, struct mnt_namespace *, struct fs_struct *); -extern void __put_mnt_ns(struct mnt_namespace *ns); - -static inline void put_mnt_ns(struct mnt_namespace *ns) -{ - if (atomic_dec_and_lock(&ns->count, &vfsmount_lock)) - /* releases vfsmount_lock */ - __put_mnt_ns(ns); -} +extern void put_mnt_ns(struct mnt_namespace *ns); static inline void exit_mnt_ns(struct task_struct *p) { -- cgit v1.2.3 From cf8d2c11cb77f129675478792122f50827e5b0ae Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Mon, 22 Jun 2009 15:09:13 -0400 Subject: VFS: Add VFS helper functions for setting up private namespaces The purpose of this patch is to improve the remote mount path lookup support for distributed filesystems such as the NFSv4 client. When given a mount command of the form "mount server:/foo/bar /mnt", the NFSv4 client is required to look up the filehandle for "server:/", and then look up each component of the remote mount path "foo/bar" in order to find the directory that is actually going to be mounted on /mnt. Following that remote mount path may involve following symlinks, crossing server-side mount points and even following referrals to filesystem volumes on other servers. Since the standard VFS path lookup code already supports walking paths that contain all these features (using in-kernel automounts for following referrals) we would like to be able to reuse that rather than duplicate the full path traversal functionality in the NFSv4 client code. This patch therefore defines a VFS helper function create_mnt_ns(), that sets up a temporary filesystem namespace and attaches a root filesystem to it. It exports the create_mnt_ns() and put_mnt_ns() function for use by filesystem modules. Signed-off-by: Trond Myklebust Signed-off-by: Linus Torvalds --- fs/namespace.c | 45 +++++++++++++++++++++++++++++++++++-------- include/linux/mnt_namespace.h | 1 + 2 files changed, 38 insertions(+), 8 deletions(-) (limited to 'fs') diff --git a/fs/namespace.c b/fs/namespace.c index 6645846f2056..a7bea8c8bd46 100644 --- a/fs/namespace.c +++ b/fs/namespace.c @@ -1937,6 +1937,21 @@ dput_out: return retval; } +static struct mnt_namespace *alloc_mnt_ns(void) +{ + struct mnt_namespace *new_ns; + + new_ns = kmalloc(sizeof(struct mnt_namespace), GFP_KERNEL); + if (!new_ns) + return ERR_PTR(-ENOMEM); + atomic_set(&new_ns->count, 1); + new_ns->root = NULL; + INIT_LIST_HEAD(&new_ns->list); + init_waitqueue_head(&new_ns->poll); + new_ns->event = 0; + return new_ns; +} + /* * Allocate a new namespace structure and populate it with contents * copied from the namespace of the passed in task structure. @@ -1948,14 +1963,9 @@ static struct mnt_namespace *dup_mnt_ns(struct mnt_namespace *mnt_ns, struct vfsmount *rootmnt = NULL, *pwdmnt = NULL; struct vfsmount *p, *q; - new_ns = kmalloc(sizeof(struct mnt_namespace), GFP_KERNEL); - if (!new_ns) - return ERR_PTR(-ENOMEM); - - atomic_set(&new_ns->count, 1); - INIT_LIST_HEAD(&new_ns->list); - init_waitqueue_head(&new_ns->poll); - new_ns->event = 0; + new_ns = alloc_mnt_ns(); + if (IS_ERR(new_ns)) + return new_ns; down_write(&namespace_sem); /* First pass: copy the tree topology */ @@ -2019,6 +2029,24 @@ struct mnt_namespace *copy_mnt_ns(unsigned long flags, struct mnt_namespace *ns, return new_ns; } +/** + * create_mnt_ns - creates a private namespace and adds a root filesystem + * @mnt: pointer to the new root filesystem mountpoint + */ +struct mnt_namespace *create_mnt_ns(struct vfsmount *mnt) +{ + struct mnt_namespace *new_ns; + + new_ns = alloc_mnt_ns(); + if (!IS_ERR(new_ns)) { + mnt->mnt_ns = new_ns; + new_ns->root = mnt; + list_add(&new_ns->list, &new_ns->root->mnt_list); + } + return new_ns; +} +EXPORT_SYMBOL(create_mnt_ns); + SYSCALL_DEFINE5(mount, char __user *, dev_name, char __user *, dir_name, char __user *, type, unsigned long, flags, void __user *, data) { @@ -2264,3 +2292,4 @@ void put_mnt_ns(struct mnt_namespace *ns) release_mounts(&umount_list); kfree(ns); } +EXPORT_SYMBOL(put_mnt_ns); diff --git a/include/linux/mnt_namespace.h b/include/linux/mnt_namespace.h index 299d11af5f79..3beb2592b03f 100644 --- a/include/linux/mnt_namespace.h +++ b/include/linux/mnt_namespace.h @@ -24,6 +24,7 @@ struct proc_mounts { struct fs_struct; +extern struct mnt_namespace *create_mnt_ns(struct vfsmount *mnt); extern struct mnt_namespace *copy_mnt_ns(unsigned long, struct mnt_namespace *, struct fs_struct *); extern void put_mnt_ns(struct mnt_namespace *ns); -- cgit v1.2.3 From c02d7adf8c5429727a98bad1d039bccad4c61c50 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Mon, 22 Jun 2009 15:09:14 -0400 Subject: NFSv4: Replace nfs4_path_walk() with VFS path lookup in a private namespace As noted in the previous patch, the NFSv4 client mount code currently has several limitations. If the mount path contains symlinks, or referrals, or even if it just contains a '..', then the client code in nfs4_path_walk() will fail with an error. This patch replaces the nfs4_path_walk()-based lookup with a helper function that sets up a private namespace to represent the namespace on the server, then uses the ordinary VFS and NFS path lookup code to walk down the mount path in that namespace. Signed-off-by: Trond Myklebust Signed-off-by: Linus Torvalds --- fs/nfs/super.c | 178 ++++++++++++++++++++++++++++++++++++++++++++++++++------- 1 file changed, 157 insertions(+), 21 deletions(-) (limited to 'fs') diff --git a/fs/nfs/super.c b/fs/nfs/super.c index 3d460527daab..8e0673a0d6aa 100644 --- a/fs/nfs/super.c +++ b/fs/nfs/super.c @@ -42,6 +42,8 @@ #include #include #include +#include +#include #include #include #include @@ -272,10 +274,14 @@ static const struct super_operations nfs_sops = { #ifdef CONFIG_NFS_V4 static int nfs4_get_sb(struct file_system_type *fs_type, int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt); +static int nfs4_remote_get_sb(struct file_system_type *fs_type, + int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt); static int nfs4_xdev_get_sb(struct file_system_type *fs_type, int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt); static int nfs4_referral_get_sb(struct file_system_type *fs_type, int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt); +static int nfs4_remote_referral_get_sb(struct file_system_type *fs_type, + int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt); static void nfs4_kill_super(struct super_block *sb); static struct file_system_type nfs4_fs_type = { @@ -286,6 +292,14 @@ static struct file_system_type nfs4_fs_type = { .fs_flags = FS_RENAME_DOES_D_MOVE|FS_REVAL_DOT|FS_BINARY_MOUNTDATA, }; +static struct file_system_type nfs4_remote_fs_type = { + .owner = THIS_MODULE, + .name = "nfs4", + .get_sb = nfs4_remote_get_sb, + .kill_sb = nfs4_kill_super, + .fs_flags = FS_RENAME_DOES_D_MOVE|FS_REVAL_DOT|FS_BINARY_MOUNTDATA, +}; + struct file_system_type nfs4_xdev_fs_type = { .owner = THIS_MODULE, .name = "nfs4", @@ -294,6 +308,14 @@ struct file_system_type nfs4_xdev_fs_type = { .fs_flags = FS_RENAME_DOES_D_MOVE|FS_REVAL_DOT|FS_BINARY_MOUNTDATA, }; +static struct file_system_type nfs4_remote_referral_fs_type = { + .owner = THIS_MODULE, + .name = "nfs4", + .get_sb = nfs4_remote_referral_get_sb, + .kill_sb = nfs4_kill_super, + .fs_flags = FS_RENAME_DOES_D_MOVE|FS_REVAL_DOT|FS_BINARY_MOUNTDATA, +}; + struct file_system_type nfs4_referral_fs_type = { .owner = THIS_MODULE, .name = "nfs4", @@ -2433,12 +2455,12 @@ out_no_client_address: } /* - * Get the superblock for an NFS4 mountpoint + * Get the superblock for the NFS4 root partition */ -static int nfs4_get_sb(struct file_system_type *fs_type, +static int nfs4_remote_get_sb(struct file_system_type *fs_type, int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt) { - struct nfs_parsed_mount_data *data; + struct nfs_parsed_mount_data *data = raw_data; struct super_block *s; struct nfs_server *server; struct nfs_fh *mntfh; @@ -2449,18 +2471,12 @@ static int nfs4_get_sb(struct file_system_type *fs_type, }; int error = -ENOMEM; - data = kzalloc(sizeof(*data), GFP_KERNEL); mntfh = kzalloc(sizeof(*mntfh), GFP_KERNEL); if (data == NULL || mntfh == NULL) goto out_free_fh; security_init_mnt_opts(&data->lsm_opts); - /* Validate the mount data */ - error = nfs4_validate_mount_data(raw_data, data, dev_name); - if (error < 0) - goto out; - /* Get a volume representation */ server = nfs4_create_server(data, mntfh); if (IS_ERR(server)) { @@ -2473,7 +2489,7 @@ static int nfs4_get_sb(struct file_system_type *fs_type, compare_super = NULL; /* Get a superblock - note that we may end up sharing one that already exists */ - s = sget(fs_type, compare_super, nfs_set_super, &sb_mntdata); + s = sget(&nfs4_fs_type, compare_super, nfs_set_super, &sb_mntdata); if (IS_ERR(s)) { error = PTR_ERR(s); goto out_free; @@ -2510,14 +2526,9 @@ static int nfs4_get_sb(struct file_system_type *fs_type, error = 0; out: - kfree(data->client_address); - kfree(data->nfs_server.export_path); - kfree(data->nfs_server.hostname); - kfree(data->fscache_uniq); security_free_mnt_opts(&data->lsm_opts); out_free_fh: kfree(mntfh); - kfree(data); return error; out_free: @@ -2531,6 +2542,102 @@ error_splat_super: goto out; } +static struct vfsmount *nfs_do_root_mount(struct file_system_type *fs_type, + int flags, void *data, const char *hostname) +{ + struct vfsmount *root_mnt; + char *root_devname; + size_t len; + + len = strlen(hostname) + 3; + root_devname = kmalloc(len, GFP_KERNEL); + if (root_devname == NULL) + return ERR_PTR(-ENOMEM); + snprintf(root_devname, len, "%s:/", hostname); + root_mnt = vfs_kern_mount(fs_type, flags, root_devname, data); + kfree(root_devname); + return root_mnt; +} + +static int nfs_follow_remote_path(struct vfsmount *root_mnt, + const char *export_path, struct vfsmount *mnt_target) +{ + struct mnt_namespace *ns_private; + struct nameidata nd; + struct super_block *s; + int ret; + + ns_private = create_mnt_ns(root_mnt); + ret = PTR_ERR(ns_private); + if (IS_ERR(ns_private)) + goto out_mntput; + + ret = vfs_path_lookup(root_mnt->mnt_root, root_mnt, + export_path, LOOKUP_FOLLOW, &nd); + + put_mnt_ns(ns_private); + + if (ret != 0) + goto out_err; + + s = nd.path.mnt->mnt_sb; + atomic_inc(&s->s_active); + mnt_target->mnt_sb = s; + mnt_target->mnt_root = dget(nd.path.dentry); + + path_put(&nd.path); + down_write(&s->s_umount); + return 0; +out_mntput: + mntput(root_mnt); +out_err: + return ret; +} + +/* + * Get the superblock for an NFS4 mountpoint + */ +static int nfs4_get_sb(struct file_system_type *fs_type, + int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt) +{ + struct nfs_parsed_mount_data *data; + char *export_path; + struct vfsmount *root_mnt; + int error = -ENOMEM; + + data = kzalloc(sizeof(*data), GFP_KERNEL); + if (data == NULL) + goto out_free_data; + + /* Validate the mount data */ + error = nfs4_validate_mount_data(raw_data, data, dev_name); + if (error < 0) + goto out; + + export_path = data->nfs_server.export_path; + data->nfs_server.export_path = "/"; + root_mnt = nfs_do_root_mount(&nfs4_remote_fs_type, flags, data, + data->nfs_server.hostname); + data->nfs_server.export_path = export_path; + + error = PTR_ERR(root_mnt); + if (IS_ERR(root_mnt)) + goto out; + + error = nfs_follow_remote_path(root_mnt, export_path, mnt); + +out: + kfree(data->client_address); + kfree(data->nfs_server.export_path); + kfree(data->nfs_server.hostname); + kfree(data->fscache_uniq); +out_free_data: + kfree(data); + dprintk("<-- nfs4_get_sb() = %d%s\n", error, + error != 0 ? " [error]" : ""); + return error; +} + static void nfs4_kill_super(struct super_block *sb) { struct nfs_server *server = NFS_SB(sb); @@ -2627,12 +2734,9 @@ error_splat_super: return error; } -/* - * Create an NFS4 server record on referral traversal - */ -static int nfs4_referral_get_sb(struct file_system_type *fs_type, int flags, - const char *dev_name, void *raw_data, - struct vfsmount *mnt) +static int nfs4_remote_referral_get_sb(struct file_system_type *fs_type, + int flags, const char *dev_name, void *raw_data, + struct vfsmount *mnt) { struct nfs_clone_mount *data = raw_data; struct super_block *s; @@ -2711,4 +2815,36 @@ error_splat_super: return error; } +/* + * Create an NFS4 server record on referral traversal + */ +static int nfs4_referral_get_sb(struct file_system_type *fs_type, + int flags, const char *dev_name, void *raw_data, + struct vfsmount *mnt) +{ + struct nfs_clone_mount *data = raw_data; + char *export_path; + struct vfsmount *root_mnt; + int error; + + dprintk("--> nfs4_referral_get_sb()\n"); + + export_path = data->mnt_path; + data->mnt_path = "/"; + + root_mnt = nfs_do_root_mount(&nfs4_remote_referral_fs_type, + flags, data, data->hostname); + data->mnt_path = export_path; + + error = PTR_ERR(root_mnt); + if (IS_ERR(root_mnt)) + goto out; + + error = nfs_follow_remote_path(root_mnt, export_path, mnt); +out: + dprintk("<-- nfs4_referral_get_sb() = %d%s\n", error, + error != 0 ? " [error]" : ""); + return error; +} + #endif /* CONFIG_NFS_V4 */ -- cgit v1.2.3 From 0b75b35c7cad33e7613f5adf28fa10fe8b09b1c3 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Mon, 22 Jun 2009 15:09:14 -0400 Subject: NFS: Fix nfs_path() to always return a '/' at the beginning of the path Signed-off-by: Trond Myklebust Signed-off-by: Linus Torvalds --- fs/nfs/namespace.c | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'fs') diff --git a/fs/nfs/namespace.c b/fs/nfs/namespace.c index f01caec84463..40c766782891 100644 --- a/fs/nfs/namespace.c +++ b/fs/nfs/namespace.c @@ -65,6 +65,11 @@ char *nfs_path(const char *base, dentry = dentry->d_parent; } spin_unlock(&dcache_lock); + if (*end != '/') { + if (--buflen < 0) + goto Elong; + *--end = '/'; + } namelen = strlen(base); /* Strip off excess slashes in base string */ while (namelen > 0 && base[namelen - 1] == '/') -- cgit v1.2.3 From b88f8a546f5dba213938fdfc11e66bc5c2421623 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Mon, 22 Jun 2009 15:09:14 -0400 Subject: NFS: Correct the NFS mount path when following a referral Signed-off-by: Trond Myklebust Signed-off-by: Linus Torvalds --- fs/nfs/super.c | 24 ++++++++++++++++++++++++ 1 file changed, 24 insertions(+) (limited to 'fs') diff --git a/fs/nfs/super.c b/fs/nfs/super.c index 8e0673a0d6aa..0b4cbdc60abd 100644 --- a/fs/nfs/super.c +++ b/fs/nfs/super.c @@ -2559,6 +2559,27 @@ static struct vfsmount *nfs_do_root_mount(struct file_system_type *fs_type, return root_mnt; } +static void nfs_fix_devname(const struct path *path, struct vfsmount *mnt) +{ + char *page = (char *) __get_free_page(GFP_KERNEL); + char *devname, *tmp; + + if (page == NULL) + return; + devname = nfs_path(path->mnt->mnt_devname, + path->mnt->mnt_root, path->dentry, + page, PAGE_SIZE); + if (devname == NULL) + goto out_freepage; + tmp = kstrdup(devname, GFP_KERNEL); + if (tmp == NULL) + goto out_freepage; + kfree(mnt->mnt_devname); + mnt->mnt_devname = tmp; +out_freepage: + free_page((unsigned long)page); +} + static int nfs_follow_remote_path(struct vfsmount *root_mnt, const char *export_path, struct vfsmount *mnt_target) { @@ -2585,6 +2606,9 @@ static int nfs_follow_remote_path(struct vfsmount *root_mnt, mnt_target->mnt_sb = s; mnt_target->mnt_root = dget(nd.path.dentry); + /* Correct the device pathname */ + nfs_fix_devname(&nd.path, mnt_target); + path_put(&nd.path); down_write(&s->s_umount); return 0; -- cgit v1.2.3 From 3391faa4f18e4e33666d3d24e90e3086fcf9b922 Mon Sep 17 00:00:00 2001 From: Roel Kluin Date: Mon, 22 Jun 2009 23:12:29 +0200 Subject: udf: remove redundant tests on unsigned first_block and goal are unsigned. When negative they are wrapped and caught by the other test. Signed-off-by: Roel Kluin Signed-off-by: Jan Kara --- fs/udf/balloc.c | 9 ++++----- 1 file changed, 4 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/udf/balloc.c b/fs/udf/balloc.c index e48e9a3af763..1e068535b58b 100644 --- a/fs/udf/balloc.c +++ b/fs/udf/balloc.c @@ -238,7 +238,7 @@ static int udf_bitmap_prealloc_blocks(struct super_block *sb, mutex_lock(&sbi->s_alloc_mutex); part_len = sbi->s_partmaps[partition].s_partition_len; - if (first_block < 0 || first_block >= part_len) + if (first_block >= part_len) goto out; if (first_block + block_count > part_len) @@ -297,7 +297,7 @@ static int udf_bitmap_new_block(struct super_block *sb, mutex_lock(&sbi->s_alloc_mutex); repeat: - if (goal < 0 || goal >= sbi->s_partmaps[partition].s_partition_len) + if (goal >= sbi->s_partmaps[partition].s_partition_len) goal = 0; nr_groups = bitmap->s_nr_groups; @@ -666,8 +666,7 @@ static int udf_table_prealloc_blocks(struct super_block *sb, int8_t etype = -1; struct udf_inode_info *iinfo; - if (first_block < 0 || - first_block >= sbi->s_partmaps[partition].s_partition_len) + if (first_block >= sbi->s_partmaps[partition].s_partition_len) return 0; iinfo = UDF_I(table); @@ -743,7 +742,7 @@ static int udf_table_new_block(struct super_block *sb, return newblock; mutex_lock(&sbi->s_alloc_mutex); - if (goal < 0 || goal >= sbi->s_partmaps[partition].s_partition_len) + if (goal >= sbi->s_partmaps[partition].s_partition_len) goal = 0; /* We search for the closest matching block to goal. If we find -- cgit v1.2.3 From b5450d9c84bdd38b261922057cd167da51dfae93 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Fri, 19 Jun 2009 10:30:07 +0200 Subject: reiserfs: remove stray unlock_super in reiserfs_resize Reiserfs doesn't use lock_super anywhere internally, and ->remount_fs which calls reiserfs_resize does have it currently but also expects it to be held on return, so there's no business for the unlock_super here. Signed-off-by: Christoph Hellwig Acked by Edward Shishkin Signed-off-by: Al Viro --- fs/reiserfs/resize.c | 1 - 1 file changed, 1 deletion(-) (limited to 'fs') diff --git a/fs/reiserfs/resize.c b/fs/reiserfs/resize.c index 238e9d9b31e0..18b315d3d104 100644 --- a/fs/reiserfs/resize.c +++ b/fs/reiserfs/resize.c @@ -82,7 +82,6 @@ int reiserfs_resize(struct super_block *s, unsigned long block_count_new) if (reiserfs_allocate_list_bitmaps(s, jbitmap, bmap_nr_new) < 0) { printk ("reiserfs_resize: unable to allocate memory for journal bitmaps\n"); - unlock_super(s); return -ENOMEM; } /* the new journal bitmaps are zero filled, now we copy in the bitmap -- cgit v1.2.3 From 654f562c526cf9dfb8d453f687341fe0777ee454 Mon Sep 17 00:00:00 2001 From: "J. R. Okajima" Date: Thu, 18 Jun 2009 23:30:15 +0900 Subject: vfs: fix nd->root leak in do_filp_open() commit 2a737871108de9ba8930f7650d549f1383767f8b "Cache root in nameidata" introduced a new member nd->root, but forgot to put it in do_filp_open(). Signed-off-by: J. R. Okajima Signed-off-by: Al Viro --- fs/namei.c | 11 ++++++++++- 1 file changed, 10 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/namei.c b/fs/namei.c index 527119afb6a5..5b961eb71cbf 100644 --- a/fs/namei.c +++ b/fs/namei.c @@ -1698,8 +1698,11 @@ struct file *do_filp_open(int dfd, const char *pathname, if (error) return ERR_PTR(error); error = path_walk(pathname, &nd); - if (error) + if (error) { + if (nd.root.mnt) + path_put(&nd.root); return ERR_PTR(error); + } if (unlikely(!audit_dummy_context())) audit_inode(pathname, nd.path.dentry); @@ -1759,6 +1762,8 @@ do_last: } filp = nameidata_to_filp(&nd, open_flag); mnt_drop_write(nd.path.mnt); + if (nd.root.mnt) + path_put(&nd.root); return filp; } @@ -1819,6 +1824,8 @@ ok: */ if (will_write) mnt_drop_write(nd.path.mnt); + if (nd.root.mnt) + path_put(&nd.root); return filp; exit_mutex_unlock: @@ -1859,6 +1866,8 @@ do_link: * with "intent.open". */ release_open_intent(&nd); + if (nd.root.mnt) + path_put(&nd.root); return ERR_PTR(error); } nd.flags &= ~LOOKUP_PARENT; -- cgit v1.2.3 From 3b22edc5730b87d360ee7dd7143397ba09b73a47 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Tue, 23 Jun 2009 17:29:49 -0400 Subject: VFS: Switch init_mount_tree() to use the new create_mnt_ns() helper Eliminates some duplicated code... Signed-off-by: Trond Myklebust Signed-off-by: Al Viro --- fs/namespace.c | 11 ++--------- 1 file changed, 2 insertions(+), 9 deletions(-) (limited to 'fs') diff --git a/fs/namespace.c b/fs/namespace.c index a7bea8c8bd46..4a86b8595164 100644 --- a/fs/namespace.c +++ b/fs/namespace.c @@ -2222,16 +2222,9 @@ static void __init init_mount_tree(void) mnt = do_kern_mount("rootfs", 0, "rootfs", NULL); if (IS_ERR(mnt)) panic("Can't create rootfs"); - ns = kmalloc(sizeof(*ns), GFP_KERNEL); - if (!ns) + ns = create_mnt_ns(mnt); + if (IS_ERR(ns)) panic("Can't allocate initial namespace"); - atomic_set(&ns->count, 1); - INIT_LIST_HEAD(&ns->list); - init_waitqueue_head(&ns->poll); - ns->event = 0; - list_add(&mnt->mnt_list, &ns->list); - ns->root = mnt; - mnt->mnt_ns = ns; init_task.nsproxy->mnt_ns = ns; get_mnt_ns(ns); -- cgit v1.2.3 From f6cc746bbb3b8a8ceb8514a7906ba582607a8cf7 Mon Sep 17 00:00:00 2001 From: "H. Peter Anvin" Date: Tue, 16 Jun 2009 21:15:04 -0700 Subject: devpts: remove module-related code These days, the devpts filesystem is closely integrated with the pty memory management, and cannot be built as a module, even less removed from the kernel. Accordingly, remove all module-related stuff from this filesystem. [ v2: only remove code that's actually dead ] Signed-off-by: H. Peter Anvin Signed-off-by: Al Viro --- fs/devpts/inode.c | 10 ---------- 1 file changed, 10 deletions(-) (limited to 'fs') diff --git a/fs/devpts/inode.c b/fs/devpts/inode.c index 9b1d285f9fe6..75efb028974b 100644 --- a/fs/devpts/inode.c +++ b/fs/devpts/inode.c @@ -423,7 +423,6 @@ static void devpts_kill_sb(struct super_block *sb) } static struct file_system_type devpts_fs_type = { - .owner = THIS_MODULE, .name = "devpts", .get_sb = devpts_get_sb, .kill_sb = devpts_kill_sb, @@ -564,13 +563,4 @@ static int __init init_devpts_fs(void) } return err; } - -static void __exit exit_devpts_fs(void) -{ - unregister_filesystem(&devpts_fs_type); - mntput(devpts_mnt); -} - module_init(init_devpts_fs) -module_exit(exit_devpts_fs) -MODULE_LICENSE("GPL"); -- cgit v1.2.3 From c63e09ecccb50f930e899d7005edc5411ee86d4f Mon Sep 17 00:00:00 2001 From: Al Viro Date: Wed, 24 Jun 2009 02:05:18 -0400 Subject: Make allocation of anon devices cheaper Standard trick - add a new variable (start) such that for each n < start n is known to be busy. Allocation can skip checking everything in [0..start) and if it returns n, we can set start to n + 1. Freeing below start sets start to what we'd just freed. Of course, it still sucks if we do something like free 0 allocate allocate in a loop - still O(n^2) time. However, on saner loads it improves the things a lot and the entire thing is not worth the trouble of switching to something with better worst-case behaviour. Signed-off-by: Al Viro --- fs/super.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/super.c b/fs/super.c index d40d53a22fb5..808ffd59e01b 100644 --- a/fs/super.c +++ b/fs/super.c @@ -608,6 +608,7 @@ void emergency_remount(void) static DEFINE_IDA(unnamed_dev_ida); static DEFINE_SPINLOCK(unnamed_dev_lock);/* protects the above */ +static int unnamed_dev_start = 0; /* don't bother trying below it */ int set_anon_super(struct super_block *s, void *data) { @@ -618,7 +619,8 @@ int set_anon_super(struct super_block *s, void *data) if (ida_pre_get(&unnamed_dev_ida, GFP_ATOMIC) == 0) return -ENOMEM; spin_lock(&unnamed_dev_lock); - error = ida_get_new(&unnamed_dev_ida, &dev); + error = ida_get_new_above(&unnamed_dev_ida, unnamed_dev_start, &dev); + unnamed_dev_start = dev + 1; spin_unlock(&unnamed_dev_lock); if (error == -EAGAIN) /* We raced and lost with another CPU. */ @@ -629,6 +631,7 @@ int set_anon_super(struct super_block *s, void *data) if ((dev & MAX_ID_MASK) == (1 << MINORBITS)) { spin_lock(&unnamed_dev_lock); ida_remove(&unnamed_dev_ida, dev); + unnamed_dev_start = dev; spin_unlock(&unnamed_dev_lock); return -EMFILE; } @@ -645,6 +648,8 @@ void kill_anon_super(struct super_block *sb) generic_shutdown_super(sb); spin_lock(&unnamed_dev_lock); ida_remove(&unnamed_dev_ida, slot); + if (slot < unnamed_dev_start) + unnamed_dev_start = slot; spin_unlock(&unnamed_dev_lock); } -- cgit v1.2.3 From f21f62208a6f60e2e05440b2e438d9541822dc4d Mon Sep 17 00:00:00 2001 From: Al Viro Date: Wed, 24 Jun 2009 03:12:00 -0400 Subject: ... and the same for vfsmount id/mount group id Signed-off-by: Al Viro --- fs/namespace.c | 26 ++++++++++++++++++++++---- fs/super.c | 6 ++++-- 2 files changed, 26 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/namespace.c b/fs/namespace.c index 4a86b8595164..3dc283fd4716 100644 --- a/fs/namespace.c +++ b/fs/namespace.c @@ -42,6 +42,8 @@ __cacheline_aligned_in_smp DEFINE_SPINLOCK(vfsmount_lock); static int event; static DEFINE_IDA(mnt_id_ida); static DEFINE_IDA(mnt_group_ida); +static int mnt_id_start = 0; +static int mnt_group_start = 1; static struct list_head *mount_hashtable __read_mostly; static struct kmem_cache *mnt_cache __read_mostly; @@ -69,7 +71,9 @@ static int mnt_alloc_id(struct vfsmount *mnt) retry: ida_pre_get(&mnt_id_ida, GFP_KERNEL); spin_lock(&vfsmount_lock); - res = ida_get_new(&mnt_id_ida, &mnt->mnt_id); + res = ida_get_new_above(&mnt_id_ida, mnt_id_start, &mnt->mnt_id); + if (!res) + mnt_id_start = mnt->mnt_id + 1; spin_unlock(&vfsmount_lock); if (res == -EAGAIN) goto retry; @@ -79,8 +83,11 @@ retry: static void mnt_free_id(struct vfsmount *mnt) { + int id = mnt->mnt_id; spin_lock(&vfsmount_lock); - ida_remove(&mnt_id_ida, mnt->mnt_id); + ida_remove(&mnt_id_ida, id); + if (mnt_id_start > id) + mnt_id_start = id; spin_unlock(&vfsmount_lock); } @@ -91,10 +98,18 @@ static void mnt_free_id(struct vfsmount *mnt) */ static int mnt_alloc_group_id(struct vfsmount *mnt) { + int res; + if (!ida_pre_get(&mnt_group_ida, GFP_KERNEL)) return -ENOMEM; - return ida_get_new_above(&mnt_group_ida, 1, &mnt->mnt_group_id); + res = ida_get_new_above(&mnt_group_ida, + mnt_group_start, + &mnt->mnt_group_id); + if (!res) + mnt_group_start = mnt->mnt_group_id + 1; + + return res; } /* @@ -102,7 +117,10 @@ static int mnt_alloc_group_id(struct vfsmount *mnt) */ void mnt_release_group_id(struct vfsmount *mnt) { - ida_remove(&mnt_group_ida, mnt->mnt_group_id); + int id = mnt->mnt_group_id; + ida_remove(&mnt_group_ida, id); + if (mnt_group_start > id) + mnt_group_start = id; mnt->mnt_group_id = 0; } diff --git a/fs/super.c b/fs/super.c index 808ffd59e01b..2761d3e22ed9 100644 --- a/fs/super.c +++ b/fs/super.c @@ -620,7 +620,8 @@ int set_anon_super(struct super_block *s, void *data) return -ENOMEM; spin_lock(&unnamed_dev_lock); error = ida_get_new_above(&unnamed_dev_ida, unnamed_dev_start, &dev); - unnamed_dev_start = dev + 1; + if (!error) + unnamed_dev_start = dev + 1; spin_unlock(&unnamed_dev_lock); if (error == -EAGAIN) /* We raced and lost with another CPU. */ @@ -631,7 +632,8 @@ int set_anon_super(struct super_block *s, void *data) if ((dev & MAX_ID_MASK) == (1 << MINORBITS)) { spin_lock(&unnamed_dev_lock); ida_remove(&unnamed_dev_ida, dev); - unnamed_dev_start = dev; + if (unnamed_dev_start > dev) + unnamed_dev_start = dev; spin_unlock(&unnamed_dev_lock); return -EMFILE; } -- cgit v1.2.3 From 01c031945f2755c7afaaf456088543312f2b72ea Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Mon, 8 Jun 2009 13:35:40 +0200 Subject: cleanup __writeback_single_inode There is no reason to for the split between __writeback_single_inode and __sync_single_inode, the former just does a couple of checks before tail-calling the latter. So merge the two, and while we're at it split out the I_SYNC waiting case for data integrity writers, as it's logically separate function. Finally rename __writeback_single_inode to writeback_single_inode. Signed-off-by: Christoph Hellwig Signed-off-by: Al Viro --- fs/fs-writeback.c | 100 +++++++++++++++++++++++++++--------------------------- 1 file changed, 50 insertions(+), 50 deletions(-) (limited to 'fs') diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c index caf049146ca2..c54226be5294 100644 --- a/fs/fs-writeback.c +++ b/fs/fs-writeback.c @@ -278,7 +278,26 @@ int sb_has_dirty_inodes(struct super_block *sb) EXPORT_SYMBOL(sb_has_dirty_inodes); /* - * Write a single inode's dirty pages and inode data out to disk. + * Wait for writeback on an inode to complete. + */ +static void inode_wait_for_writeback(struct inode *inode) +{ + DEFINE_WAIT_BIT(wq, &inode->i_state, __I_SYNC); + wait_queue_head_t *wqh; + + wqh = bit_waitqueue(&inode->i_state, __I_SYNC); + do { + spin_unlock(&inode_lock); + __wait_on_bit(wqh, &wq, inode_wait, TASK_UNINTERRUPTIBLE); + spin_lock(&inode_lock); + } while (inode->i_state & I_SYNC); +} + +/* + * Write out an inode's dirty pages. Called under inode_lock. Either the + * caller has ref on the inode (either via __iget or via syscall against an fd) + * or the inode has I_WILL_FREE set (via generic_forget_inode) + * * If `wait' is set, wait on the writeout. * * The whole writeout design is quite complex and fragile. We want to avoid @@ -288,13 +307,38 @@ EXPORT_SYMBOL(sb_has_dirty_inodes); * Called under inode_lock. */ static int -__sync_single_inode(struct inode *inode, struct writeback_control *wbc) +writeback_single_inode(struct inode *inode, struct writeback_control *wbc) { - unsigned dirty; struct address_space *mapping = inode->i_mapping; int wait = wbc->sync_mode == WB_SYNC_ALL; + unsigned dirty; int ret; + if (!atomic_read(&inode->i_count)) + WARN_ON(!(inode->i_state & (I_WILL_FREE|I_FREEING))); + else + WARN_ON(inode->i_state & I_WILL_FREE); + + if (inode->i_state & I_SYNC) { + /* + * If this inode is locked for writeback and we are not doing + * writeback-for-data-integrity, move it to s_more_io so that + * writeback can proceed with the other inodes on s_io. + * + * We'll have another go at writing back this inode when we + * completed a full scan of s_io. + */ + if (!wait) { + requeue_io(inode); + return 0; + } + + /* + * It's a data-integrity sync. We must wait. + */ + inode_wait_for_writeback(inode); + } + BUG_ON(inode->i_state & I_SYNC); /* Set I_SYNC, reset I_DIRTY */ @@ -389,50 +433,6 @@ __sync_single_inode(struct inode *inode, struct writeback_control *wbc) return ret; } -/* - * Write out an inode's dirty pages. Called under inode_lock. Either the - * caller has ref on the inode (either via __iget or via syscall against an fd) - * or the inode has I_WILL_FREE set (via generic_forget_inode) - */ -static int -__writeback_single_inode(struct inode *inode, struct writeback_control *wbc) -{ - wait_queue_head_t *wqh; - - if (!atomic_read(&inode->i_count)) - WARN_ON(!(inode->i_state & (I_WILL_FREE|I_FREEING))); - else - WARN_ON(inode->i_state & I_WILL_FREE); - - if ((wbc->sync_mode != WB_SYNC_ALL) && (inode->i_state & I_SYNC)) { - /* - * We're skipping this inode because it's locked, and we're not - * doing writeback-for-data-integrity. Move it to s_more_io so - * that writeback can proceed with the other inodes on s_io. - * We'll have another go at writing back this inode when we - * completed a full scan of s_io. - */ - requeue_io(inode); - return 0; - } - - /* - * It's a data-integrity sync. We must wait. - */ - if (inode->i_state & I_SYNC) { - DEFINE_WAIT_BIT(wq, &inode->i_state, __I_SYNC); - - wqh = bit_waitqueue(&inode->i_state, __I_SYNC); - do { - spin_unlock(&inode_lock); - __wait_on_bit(wqh, &wq, inode_wait, - TASK_UNINTERRUPTIBLE); - spin_lock(&inode_lock); - } while (inode->i_state & I_SYNC); - } - return __sync_single_inode(inode, wbc); -} - /* * Write out a superblock's list of dirty inodes. A wait will be performed * upon no inodes, all inodes or the final one, depending upon sync_mode. @@ -526,7 +526,7 @@ void generic_sync_sb_inodes(struct super_block *sb, BUG_ON(inode->i_state & (I_FREEING | I_CLEAR)); __iget(inode); pages_skipped = wbc->pages_skipped; - __writeback_single_inode(inode, wbc); + writeback_single_inode(inode, wbc); if (current_is_pdflush()) writeback_release(bdi); if (wbc->pages_skipped != pages_skipped) { @@ -708,7 +708,7 @@ int write_inode_now(struct inode *inode, int sync) might_sleep(); spin_lock(&inode_lock); - ret = __writeback_single_inode(inode, &wbc); + ret = writeback_single_inode(inode, &wbc); spin_unlock(&inode_lock); if (sync) inode_sync_wait(inode); @@ -732,7 +732,7 @@ int sync_inode(struct inode *inode, struct writeback_control *wbc) int ret; spin_lock(&inode_lock); - ret = __writeback_single_inode(inode, wbc); + ret = writeback_single_inode(inode, wbc); spin_unlock(&inode_lock); return ret; } -- cgit v1.2.3 From 3e63cbb1efca7dd3137de1bb475e2e068e38ef23 Mon Sep 17 00:00:00 2001 From: Ankit Jain Date: Fri, 19 Jun 2009 14:28:07 -0400 Subject: fs: Add new pre-allocation ioctls to vfs for compatibility with legacy xfs ioctls This patch adds ioctls to vfs for compatibility with legacy XFS pre-allocation ioctls (XFS_IOC_*RESVP*). The implementation effectively invokes sys_fallocate for the new ioctls. Also handles the compat_ioctl case. Note: These legacy ioctls are also implemented by OCFS2. [AV: folded fixes from hch] Signed-off-by: Ankit Jain Signed-off-by: Christoph Hellwig Signed-off-by: Al Viro --- fs/compat_ioctl.c | 48 +++++++++++++++++++++++++++++++++++++++++ fs/ioctl.c | 35 ++++++++++++++++++++++++++++++ fs/open.c | 58 +++++++++++++++++++++++++------------------------- include/linux/falloc.h | 21 ++++++++++++++++++ include/linux/fs.h | 6 ++++++ 5 files changed, 139 insertions(+), 29 deletions(-) (limited to 'fs') diff --git a/fs/compat_ioctl.c b/fs/compat_ioctl.c index c135202c38b3..626c7483b4de 100644 --- a/fs/compat_ioctl.c +++ b/fs/compat_ioctl.c @@ -31,6 +31,7 @@ #include #include #include +#include #include #include #include @@ -1779,6 +1780,41 @@ lp_timeout_trans(unsigned int fd, unsigned int cmd, unsigned long arg) return sys_ioctl(fd, cmd, (unsigned long)tn); } +/* on ia32 l_start is on a 32-bit boundary */ +#if defined(CONFIG_IA64) || defined(CONFIG_X86_64) +struct space_resv_32 { + __s16 l_type; + __s16 l_whence; + __s64 l_start __attribute__((packed)); + /* len == 0 means until end of file */ + __s64 l_len __attribute__((packed)); + __s32 l_sysid; + __u32 l_pid; + __s32 l_pad[4]; /* reserve area */ +}; + +#define FS_IOC_RESVSP_32 _IOW ('X', 40, struct space_resv_32) +#define FS_IOC_RESVSP64_32 _IOW ('X', 42, struct space_resv_32) + +/* just account for different alignment */ +static int compat_ioctl_preallocate(struct file *file, unsigned long arg) +{ + struct space_resv_32 __user *p32 = (void __user *)arg; + struct space_resv __user *p = compat_alloc_user_space(sizeof(*p)); + + if (copy_in_user(&p->l_type, &p32->l_type, sizeof(s16)) || + copy_in_user(&p->l_whence, &p32->l_whence, sizeof(s16)) || + copy_in_user(&p->l_start, &p32->l_start, sizeof(s64)) || + copy_in_user(&p->l_len, &p32->l_len, sizeof(s64)) || + copy_in_user(&p->l_sysid, &p32->l_sysid, sizeof(s32)) || + copy_in_user(&p->l_pid, &p32->l_pid, sizeof(u32)) || + copy_in_user(&p->l_pad, &p32->l_pad, 4*sizeof(u32))) + return -EFAULT; + + return ioctl_preallocate(file, p); +} +#endif + typedef int (*ioctl_trans_handler_t)(unsigned int, unsigned int, unsigned long, struct file *); @@ -2756,6 +2792,18 @@ asmlinkage long compat_sys_ioctl(unsigned int fd, unsigned int cmd, case FIOQSIZE: break; +#if defined(CONFIG_IA64) || defined(CONFIG_X86_64) + case FS_IOC_RESVSP_32: + case FS_IOC_RESVSP64_32: + error = compat_ioctl_preallocate(filp, arg); + goto out_fput; +#else + case FS_IOC_RESVSP: + case FS_IOC_RESVSP64: + error = ioctl_preallocate(filp, (void __user *)arg); + goto out_fput; +#endif + case FIBMAP: case FIGETBSZ: case FIONREAD: diff --git a/fs/ioctl.c b/fs/ioctl.c index 001f8d3118f2..5612880fcbe7 100644 --- a/fs/ioctl.c +++ b/fs/ioctl.c @@ -15,6 +15,7 @@ #include #include #include +#include #include @@ -403,6 +404,37 @@ EXPORT_SYMBOL(generic_block_fiemap); #endif /* CONFIG_BLOCK */ +/* + * This provides compatibility with legacy XFS pre-allocation ioctls + * which predate the fallocate syscall. + * + * Only the l_start, l_len and l_whence fields of the 'struct space_resv' + * are used here, rest are ignored. + */ +int ioctl_preallocate(struct file *filp, void __user *argp) +{ + struct inode *inode = filp->f_path.dentry->d_inode; + struct space_resv sr; + + if (copy_from_user(&sr, argp, sizeof(sr))) + return -EFAULT; + + switch (sr.l_whence) { + case SEEK_SET: + break; + case SEEK_CUR: + sr.l_start += filp->f_pos; + break; + case SEEK_END: + sr.l_start += i_size_read(inode); + break; + default: + return -EINVAL; + } + + return do_fallocate(filp, FALLOC_FL_KEEP_SIZE, sr.l_start, sr.l_len); +} + static int file_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) { @@ -414,6 +446,9 @@ static int file_ioctl(struct file *filp, unsigned int cmd, return ioctl_fibmap(filp, p); case FIONREAD: return put_user(i_size_read(inode) - filp->f_pos, p); + case FS_IOC_RESVSP: + case FS_IOC_RESVSP64: + return ioctl_preallocate(filp, p); } return vfs_ioctl(filp, cmd, arg); diff --git a/fs/open.c b/fs/open.c index 7200e23d9258..dd98e8076024 100644 --- a/fs/open.c +++ b/fs/open.c @@ -378,63 +378,63 @@ SYSCALL_ALIAS(sys_ftruncate64, SyS_ftruncate64); #endif #endif /* BITS_PER_LONG == 32 */ -SYSCALL_DEFINE(fallocate)(int fd, int mode, loff_t offset, loff_t len) + +int do_fallocate(struct file *file, int mode, loff_t offset, loff_t len) { - struct file *file; - struct inode *inode; - long ret = -EINVAL; + struct inode *inode = file->f_path.dentry->d_inode; + long ret; if (offset < 0 || len <= 0) - goto out; + return -EINVAL; /* Return error if mode is not supported */ - ret = -EOPNOTSUPP; if (mode && !(mode & FALLOC_FL_KEEP_SIZE)) - goto out; + return -EOPNOTSUPP; - ret = -EBADF; - file = fget(fd); - if (!file) - goto out; if (!(file->f_mode & FMODE_WRITE)) - goto out_fput; + return -EBADF; /* * Revalidate the write permissions, in case security policy has * changed since the files were opened. */ ret = security_file_permission(file, MAY_WRITE); if (ret) - goto out_fput; + return ret; - inode = file->f_path.dentry->d_inode; - - ret = -ESPIPE; if (S_ISFIFO(inode->i_mode)) - goto out_fput; + return -ESPIPE; - ret = -ENODEV; /* * Let individual file system decide if it supports preallocation * for directories or not. */ if (!S_ISREG(inode->i_mode) && !S_ISDIR(inode->i_mode)) - goto out_fput; + return -ENODEV; - ret = -EFBIG; /* Check for wrap through zero too */ if (((offset + len) > inode->i_sb->s_maxbytes) || ((offset + len) < 0)) - goto out_fput; + return -EFBIG; - if (inode->i_op->fallocate) - ret = inode->i_op->fallocate(inode, mode, offset, len); - else - ret = -EOPNOTSUPP; + if (!inode->i_op->fallocate) + return -EOPNOTSUPP; -out_fput: - fput(file); -out: - return ret; + return inode->i_op->fallocate(inode, mode, offset, len); } + +SYSCALL_DEFINE(fallocate)(int fd, int mode, loff_t offset, loff_t len) +{ + struct file *file; + int error = -EBADF; + + file = fget(fd); + if (file) { + error = do_fallocate(file, mode, offset, len); + fput(file); + } + + return error; +} + #ifdef CONFIG_HAVE_SYSCALL_WRAPPERS asmlinkage long SyS_fallocate(long fd, long mode, loff_t offset, loff_t len) { diff --git a/include/linux/falloc.h b/include/linux/falloc.h index 8e912ab6a072..3c155107d61f 100644 --- a/include/linux/falloc.h +++ b/include/linux/falloc.h @@ -3,4 +3,25 @@ #define FALLOC_FL_KEEP_SIZE 0x01 /* default is extend size */ +#ifdef __KERNEL__ + +/* + * Space reservation ioctls and argument structure + * are designed to be compatible with the legacy XFS ioctls. + */ +struct space_resv { + __s16 l_type; + __s16 l_whence; + __s64 l_start; + __s64 l_len; /* len == 0 means until end of file */ + __s32 l_sysid; + __u32 l_pid; + __s32 l_pad[4]; /* reserved area */ +}; + +#define FS_IOC_RESVSP _IOW('X', 40, struct space_resv) +#define FS_IOC_RESVSP64 _IOW('X', 42, struct space_resv) + +#endif /* __KERNEL__ */ + #endif /* _FALLOC_H_ */ diff --git a/include/linux/fs.h b/include/linux/fs.h index 1ff5e4e01952..79e302ddde0f 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -1906,6 +1906,8 @@ static inline int break_lease(struct inode *inode, unsigned int mode) extern int do_truncate(struct dentry *, loff_t start, unsigned int time_attrs, struct file *filp); +extern int do_fallocate(struct file *file, int mode, loff_t offset, + loff_t len); extern long do_sys_open(int dfd, const char __user *filename, int flags, int mode); extern struct file *filp_open(const char *, int, int); @@ -1914,6 +1916,10 @@ extern struct file * dentry_open(struct dentry *, struct vfsmount *, int, extern int filp_close(struct file *, fl_owner_t id); extern char * getname(const char __user *); +/* fs/ioctl.c */ + +extern int ioctl_preallocate(struct file *filp, void __user *argp); + /* fs/dcache.c */ extern void __init vfs_caches_init_early(void); extern void __init vfs_caches_init(unsigned long); -- cgit v1.2.3 From f19d4a8fa6f9b6ccf54df0971c97ffcaa390b7b0 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Mon, 8 Jun 2009 19:50:45 -0400 Subject: add caching of ACLs in struct inode No helpers, no conversions yet. Signed-off-by: Al Viro --- fs/inode.c | 10 ++++++++++ include/linux/fs.h | 7 +++++++ 2 files changed, 17 insertions(+) (limited to 'fs') diff --git a/fs/inode.c b/fs/inode.c index f643be565df8..e193cd592fa8 100644 --- a/fs/inode.c +++ b/fs/inode.c @@ -25,6 +25,7 @@ #include #include #include +#include /* * This is needed for the following functions: @@ -189,6 +190,9 @@ struct inode *inode_init_always(struct super_block *sb, struct inode *inode) } inode->i_private = NULL; inode->i_mapping = mapping; +#ifdef CONFIG_FS_POSIX_ACL + inode->i_acl = inode->i_default_acl = ACL_NOT_CACHED; +#endif #ifdef CONFIG_FSNOTIFY inode->i_fsnotify_mask = 0; @@ -227,6 +231,12 @@ void destroy_inode(struct inode *inode) ima_inode_free(inode); security_inode_free(inode); fsnotify_inode_delete(inode); +#ifdef CONFIG_FS_POSIX_ACL + if (inode->i_acl && inode->i_acl != ACL_NOT_CACHED) + posix_acl_release(inode->i_acl); + if (inode->i_default_acl && inode->i_default_acl != ACL_NOT_CACHED) + posix_acl_release(inode->i_default_acl); +#endif if (inode->i_sb->s_op->destroy_inode) inode->i_sb->s_op->destroy_inode(inode); else diff --git a/include/linux/fs.h b/include/linux/fs.h index 79e302ddde0f..0872372184fe 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -710,6 +710,9 @@ static inline int mapping_writably_mapped(struct address_space *mapping) #define i_size_ordered_init(inode) do { } while (0) #endif +struct posix_acl; +#define ACL_NOT_CACHED ((void *)(-1)) + struct inode { struct hlist_node i_hash; struct list_head i_list; @@ -772,6 +775,10 @@ struct inode { atomic_t i_writecount; #ifdef CONFIG_SECURITY void *i_security; +#endif +#ifdef CONFIG_FS_POSIX_ACL + struct posix_acl *i_acl; + struct posix_acl *i_default_acl; #endif void *i_private; /* fs or device private pointer */ }; -- cgit v1.2.3 From 5e78b435683daaaacadad1b2aeefb8904cf6acfb Mon Sep 17 00:00:00 2001 From: Al Viro Date: Mon, 8 Jun 2009 19:52:55 -0400 Subject: switch ext2 to inode->i_acl Signed-off-by: Al Viro --- fs/ext2/acl.c | 24 +++++++++++------------- fs/ext2/acl.h | 4 ---- fs/ext2/ext2.h | 4 ---- fs/ext2/inode.c | 4 ---- fs/ext2/super.c | 16 ---------------- 5 files changed, 11 insertions(+), 41 deletions(-) (limited to 'fs') diff --git a/fs/ext2/acl.c b/fs/ext2/acl.c index d46e38cb85c5..d2ffddc12117 100644 --- a/fs/ext2/acl.c +++ b/fs/ext2/acl.c @@ -128,10 +128,10 @@ fail: static inline struct posix_acl * ext2_iget_acl(struct inode *inode, struct posix_acl **i_acl) { - struct posix_acl *acl = EXT2_ACL_NOT_CACHED; + struct posix_acl *acl = ACL_NOT_CACHED; spin_lock(&inode->i_lock); - if (*i_acl != EXT2_ACL_NOT_CACHED) + if (*i_acl != ACL_NOT_CACHED) acl = posix_acl_dup(*i_acl); spin_unlock(&inode->i_lock); @@ -143,7 +143,7 @@ ext2_iset_acl(struct inode *inode, struct posix_acl **i_acl, struct posix_acl *acl) { spin_lock(&inode->i_lock); - if (*i_acl != EXT2_ACL_NOT_CACHED) + if (*i_acl != ACL_NOT_CACHED) posix_acl_release(*i_acl); *i_acl = posix_acl_dup(acl); spin_unlock(&inode->i_lock); @@ -155,7 +155,6 @@ ext2_iset_acl(struct inode *inode, struct posix_acl **i_acl, static struct posix_acl * ext2_get_acl(struct inode *inode, int type) { - struct ext2_inode_info *ei = EXT2_I(inode); int name_index; char *value = NULL; struct posix_acl *acl; @@ -166,15 +165,15 @@ ext2_get_acl(struct inode *inode, int type) switch(type) { case ACL_TYPE_ACCESS: - acl = ext2_iget_acl(inode, &ei->i_acl); - if (acl != EXT2_ACL_NOT_CACHED) + acl = ext2_iget_acl(inode, &inode->i_acl); + if (acl != ACL_NOT_CACHED) return acl; name_index = EXT2_XATTR_INDEX_POSIX_ACL_ACCESS; break; case ACL_TYPE_DEFAULT: - acl = ext2_iget_acl(inode, &ei->i_default_acl); - if (acl != EXT2_ACL_NOT_CACHED) + acl = ext2_iget_acl(inode, &inode->i_default_acl); + if (acl != ACL_NOT_CACHED) return acl; name_index = EXT2_XATTR_INDEX_POSIX_ACL_DEFAULT; break; @@ -200,11 +199,11 @@ ext2_get_acl(struct inode *inode, int type) if (!IS_ERR(acl)) { switch(type) { case ACL_TYPE_ACCESS: - ext2_iset_acl(inode, &ei->i_acl, acl); + ext2_iset_acl(inode, &inode->i_acl, acl); break; case ACL_TYPE_DEFAULT: - ext2_iset_acl(inode, &ei->i_default_acl, acl); + ext2_iset_acl(inode, &inode->i_default_acl, acl); break; } } @@ -217,7 +216,6 @@ ext2_get_acl(struct inode *inode, int type) static int ext2_set_acl(struct inode *inode, int type, struct posix_acl *acl) { - struct ext2_inode_info *ei = EXT2_I(inode); int name_index; void *value = NULL; size_t size = 0; @@ -266,11 +264,11 @@ ext2_set_acl(struct inode *inode, int type, struct posix_acl *acl) if (!error) { switch(type) { case ACL_TYPE_ACCESS: - ext2_iset_acl(inode, &ei->i_acl, acl); + ext2_iset_acl(inode, &inode->i_acl, acl); break; case ACL_TYPE_DEFAULT: - ext2_iset_acl(inode, &ei->i_default_acl, acl); + ext2_iset_acl(inode, &inode->i_default_acl, acl); break; } } diff --git a/fs/ext2/acl.h b/fs/ext2/acl.h index b42cf578554b..ecefe478898f 100644 --- a/fs/ext2/acl.h +++ b/fs/ext2/acl.h @@ -53,10 +53,6 @@ static inline int ext2_acl_count(size_t size) #ifdef CONFIG_EXT2_FS_POSIX_ACL -/* Value for inode->u.ext2_i.i_acl and inode->u.ext2_i.i_default_acl - if the ACL has not been cached */ -#define EXT2_ACL_NOT_CACHED ((void *)-1) - /* acl.c */ extern int ext2_permission (struct inode *, int); extern int ext2_acl_chmod (struct inode *); diff --git a/fs/ext2/ext2.h b/fs/ext2/ext2.h index d988a718aedb..9a8a8e27a063 100644 --- a/fs/ext2/ext2.h +++ b/fs/ext2/ext2.h @@ -46,10 +46,6 @@ struct ext2_inode_info { * EAs. */ struct rw_semaphore xattr_sem; -#endif -#ifdef CONFIG_EXT2_FS_POSIX_ACL - struct posix_acl *i_acl; - struct posix_acl *i_default_acl; #endif rwlock_t i_meta_lock; diff --git a/fs/ext2/inode.c b/fs/ext2/inode.c index 29ed682061f6..e27130341d4f 100644 --- a/fs/ext2/inode.c +++ b/fs/ext2/inode.c @@ -1224,10 +1224,6 @@ struct inode *ext2_iget (struct super_block *sb, unsigned long ino) return inode; ei = EXT2_I(inode); -#ifdef CONFIG_EXT2_FS_POSIX_ACL - ei->i_acl = EXT2_ACL_NOT_CACHED; - ei->i_default_acl = EXT2_ACL_NOT_CACHED; -#endif ei->i_block_alloc_info = NULL; raw_inode = ext2_get_inode(inode->i_sb, ino, &bh); diff --git a/fs/ext2/super.c b/fs/ext2/super.c index 458999638c3d..1a9ffee47d56 100644 --- a/fs/ext2/super.c +++ b/fs/ext2/super.c @@ -152,10 +152,6 @@ static struct inode *ext2_alloc_inode(struct super_block *sb) ei = (struct ext2_inode_info *)kmem_cache_alloc(ext2_inode_cachep, GFP_KERNEL); if (!ei) return NULL; -#ifdef CONFIG_EXT2_FS_POSIX_ACL - ei->i_acl = EXT2_ACL_NOT_CACHED; - ei->i_default_acl = EXT2_ACL_NOT_CACHED; -#endif ei->i_block_alloc_info = NULL; ei->vfs_inode.i_version = 1; return &ei->vfs_inode; @@ -198,18 +194,6 @@ static void destroy_inodecache(void) static void ext2_clear_inode(struct inode *inode) { struct ext2_block_alloc_info *rsv = EXT2_I(inode)->i_block_alloc_info; -#ifdef CONFIG_EXT2_FS_POSIX_ACL - struct ext2_inode_info *ei = EXT2_I(inode); - - if (ei->i_acl && ei->i_acl != EXT2_ACL_NOT_CACHED) { - posix_acl_release(ei->i_acl); - ei->i_acl = EXT2_ACL_NOT_CACHED; - } - if (ei->i_default_acl && ei->i_default_acl != EXT2_ACL_NOT_CACHED) { - posix_acl_release(ei->i_default_acl); - ei->i_default_acl = EXT2_ACL_NOT_CACHED; - } -#endif ext2_discard_reservation(inode); EXT2_I(inode)->i_block_alloc_info = NULL; if (unlikely(rsv)) -- cgit v1.2.3 From 6582a0e6f6bc7bf64817b9e1a424782855292ab0 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Mon, 8 Jun 2009 19:53:58 -0400 Subject: switch ext3 to inode->i_acl Signed-off-by: Al Viro --- fs/ext3/acl.c | 22 ++++++++++------------ fs/ext3/acl.h | 4 ---- fs/ext3/inode.c | 4 ---- fs/ext3/super.c | 16 ---------------- include/linux/ext3_fs_i.h | 4 ---- 5 files changed, 10 insertions(+), 40 deletions(-) (limited to 'fs') diff --git a/fs/ext3/acl.c b/fs/ext3/acl.c index e0c745451715..a9707689d9e1 100644 --- a/fs/ext3/acl.c +++ b/fs/ext3/acl.c @@ -134,7 +134,7 @@ ext3_iget_acl(struct inode *inode, struct posix_acl **i_acl) if (acl) { spin_lock(&inode->i_lock); acl = *i_acl; - if (acl != EXT3_ACL_NOT_CACHED) + if (acl != ACL_NOT_CACHED) acl = posix_acl_dup(acl); spin_unlock(&inode->i_lock); } @@ -147,7 +147,7 @@ ext3_iset_acl(struct inode *inode, struct posix_acl **i_acl, struct posix_acl *acl) { spin_lock(&inode->i_lock); - if (*i_acl != EXT3_ACL_NOT_CACHED) + if (*i_acl != ACL_NOT_CACHED) posix_acl_release(*i_acl); *i_acl = posix_acl_dup(acl); spin_unlock(&inode->i_lock); @@ -161,7 +161,6 @@ ext3_iset_acl(struct inode *inode, struct posix_acl **i_acl, static struct posix_acl * ext3_get_acl(struct inode *inode, int type) { - struct ext3_inode_info *ei = EXT3_I(inode); int name_index; char *value = NULL; struct posix_acl *acl; @@ -172,15 +171,15 @@ ext3_get_acl(struct inode *inode, int type) switch(type) { case ACL_TYPE_ACCESS: - acl = ext3_iget_acl(inode, &ei->i_acl); - if (acl != EXT3_ACL_NOT_CACHED) + acl = ext3_iget_acl(inode, &inode->i_acl); + if (acl != ACL_NOT_CACHED) return acl; name_index = EXT3_XATTR_INDEX_POSIX_ACL_ACCESS; break; case ACL_TYPE_DEFAULT: - acl = ext3_iget_acl(inode, &ei->i_default_acl); - if (acl != EXT3_ACL_NOT_CACHED) + acl = ext3_iget_acl(inode, &inode->i_default_acl); + if (acl != ACL_NOT_CACHED) return acl; name_index = EXT3_XATTR_INDEX_POSIX_ACL_DEFAULT; break; @@ -206,11 +205,11 @@ ext3_get_acl(struct inode *inode, int type) if (!IS_ERR(acl)) { switch(type) { case ACL_TYPE_ACCESS: - ext3_iset_acl(inode, &ei->i_acl, acl); + ext3_iset_acl(inode, &inode->i_acl, acl); break; case ACL_TYPE_DEFAULT: - ext3_iset_acl(inode, &ei->i_default_acl, acl); + ext3_iset_acl(inode, &inode->i_default_acl, acl); break; } } @@ -226,7 +225,6 @@ static int ext3_set_acl(handle_t *handle, struct inode *inode, int type, struct posix_acl *acl) { - struct ext3_inode_info *ei = EXT3_I(inode); int name_index; void *value = NULL; size_t size = 0; @@ -274,11 +272,11 @@ ext3_set_acl(handle_t *handle, struct inode *inode, int type, if (!error) { switch(type) { case ACL_TYPE_ACCESS: - ext3_iset_acl(inode, &ei->i_acl, acl); + ext3_iset_acl(inode, &inode->i_acl, acl); break; case ACL_TYPE_DEFAULT: - ext3_iset_acl(inode, &ei->i_default_acl, acl); + ext3_iset_acl(inode, &inode->i_default_acl, acl); break; } } diff --git a/fs/ext3/acl.h b/fs/ext3/acl.h index 42da16b8cac0..07d15a3a5969 100644 --- a/fs/ext3/acl.h +++ b/fs/ext3/acl.h @@ -53,10 +53,6 @@ static inline int ext3_acl_count(size_t size) #ifdef CONFIG_EXT3_FS_POSIX_ACL -/* Value for inode->u.ext3_i.i_acl and inode->u.ext3_i.i_default_acl - if the ACL has not been cached */ -#define EXT3_ACL_NOT_CACHED ((void *)-1) - /* acl.c */ extern int ext3_permission (struct inode *, int); extern int ext3_acl_chmod (struct inode *); diff --git a/fs/ext3/inode.c b/fs/ext3/inode.c index 05dea8132fc0..5f51fed5c750 100644 --- a/fs/ext3/inode.c +++ b/fs/ext3/inode.c @@ -2752,10 +2752,6 @@ struct inode *ext3_iget(struct super_block *sb, unsigned long ino) return inode; ei = EXT3_I(inode); -#ifdef CONFIG_EXT3_FS_POSIX_ACL - ei->i_acl = EXT3_ACL_NOT_CACHED; - ei->i_default_acl = EXT3_ACL_NOT_CACHED; -#endif ei->i_block_alloc_info = NULL; ret = __ext3_get_inode_loc(inode, &iloc, 0); diff --git a/fs/ext3/super.c b/fs/ext3/super.c index 601e881e6105..524b349c6299 100644 --- a/fs/ext3/super.c +++ b/fs/ext3/super.c @@ -464,10 +464,6 @@ static struct inode *ext3_alloc_inode(struct super_block *sb) ei = kmem_cache_alloc(ext3_inode_cachep, GFP_NOFS); if (!ei) return NULL; -#ifdef CONFIG_EXT3_FS_POSIX_ACL - ei->i_acl = EXT3_ACL_NOT_CACHED; - ei->i_default_acl = EXT3_ACL_NOT_CACHED; -#endif ei->i_block_alloc_info = NULL; ei->vfs_inode.i_version = 1; return &ei->vfs_inode; @@ -518,18 +514,6 @@ static void destroy_inodecache(void) static void ext3_clear_inode(struct inode *inode) { struct ext3_block_alloc_info *rsv = EXT3_I(inode)->i_block_alloc_info; -#ifdef CONFIG_EXT3_FS_POSIX_ACL - if (EXT3_I(inode)->i_acl && - EXT3_I(inode)->i_acl != EXT3_ACL_NOT_CACHED) { - posix_acl_release(EXT3_I(inode)->i_acl); - EXT3_I(inode)->i_acl = EXT3_ACL_NOT_CACHED; - } - if (EXT3_I(inode)->i_default_acl && - EXT3_I(inode)->i_default_acl != EXT3_ACL_NOT_CACHED) { - posix_acl_release(EXT3_I(inode)->i_default_acl); - EXT3_I(inode)->i_default_acl = EXT3_ACL_NOT_CACHED; - } -#endif ext3_discard_reservation(inode); EXT3_I(inode)->i_block_alloc_info = NULL; if (unlikely(rsv)) diff --git a/include/linux/ext3_fs_i.h b/include/linux/ext3_fs_i.h index 7894dd0f3b77..ca1bfe90004f 100644 --- a/include/linux/ext3_fs_i.h +++ b/include/linux/ext3_fs_i.h @@ -103,10 +103,6 @@ struct ext3_inode_info { */ struct rw_semaphore xattr_sem; #endif -#ifdef CONFIG_EXT3_FS_POSIX_ACL - struct posix_acl *i_acl; - struct posix_acl *i_default_acl; -#endif struct list_head i_orphan; /* unlinked but open inodes */ -- cgit v1.2.3 From d4bfe2f76d785cc77611a4bda8cedaff358d8c7d Mon Sep 17 00:00:00 2001 From: Al Viro Date: Mon, 8 Jun 2009 19:54:26 -0400 Subject: switch ext4 to inode->i_acl Signed-off-by: Al Viro --- fs/ext4/acl.c | 22 ++++++++++------------ fs/ext4/acl.h | 4 ---- fs/ext4/ext4.h | 4 ---- fs/ext4/inode.c | 4 ---- fs/ext4/super.c | 16 ---------------- 5 files changed, 10 insertions(+), 40 deletions(-) (limited to 'fs') diff --git a/fs/ext4/acl.c b/fs/ext4/acl.c index 605aeed96d68..0084e3a19d86 100644 --- a/fs/ext4/acl.c +++ b/fs/ext4/acl.c @@ -134,7 +134,7 @@ ext4_iget_acl(struct inode *inode, struct posix_acl **i_acl) if (acl) { spin_lock(&inode->i_lock); acl = *i_acl; - if (acl != EXT4_ACL_NOT_CACHED) + if (acl != ACL_NOT_CACHED) acl = posix_acl_dup(acl); spin_unlock(&inode->i_lock); } @@ -147,7 +147,7 @@ ext4_iset_acl(struct inode *inode, struct posix_acl **i_acl, struct posix_acl *acl) { spin_lock(&inode->i_lock); - if (*i_acl != EXT4_ACL_NOT_CACHED) + if (*i_acl != ACL_NOT_CACHED) posix_acl_release(*i_acl); *i_acl = posix_acl_dup(acl); spin_unlock(&inode->i_lock); @@ -161,7 +161,6 @@ ext4_iset_acl(struct inode *inode, struct posix_acl **i_acl, static struct posix_acl * ext4_get_acl(struct inode *inode, int type) { - struct ext4_inode_info *ei = EXT4_I(inode); int name_index; char *value = NULL; struct posix_acl *acl; @@ -172,15 +171,15 @@ ext4_get_acl(struct inode *inode, int type) switch (type) { case ACL_TYPE_ACCESS: - acl = ext4_iget_acl(inode, &ei->i_acl); - if (acl != EXT4_ACL_NOT_CACHED) + acl = ext4_iget_acl(inode, &inode->i_acl); + if (acl != ACL_NOT_CACHED) return acl; name_index = EXT4_XATTR_INDEX_POSIX_ACL_ACCESS; break; case ACL_TYPE_DEFAULT: - acl = ext4_iget_acl(inode, &ei->i_default_acl); - if (acl != EXT4_ACL_NOT_CACHED) + acl = ext4_iget_acl(inode, &inode->i_default_acl); + if (acl != ACL_NOT_CACHED) return acl; name_index = EXT4_XATTR_INDEX_POSIX_ACL_DEFAULT; break; @@ -206,11 +205,11 @@ ext4_get_acl(struct inode *inode, int type) if (!IS_ERR(acl)) { switch (type) { case ACL_TYPE_ACCESS: - ext4_iset_acl(inode, &ei->i_acl, acl); + ext4_iset_acl(inode, &inode->i_acl, acl); break; case ACL_TYPE_DEFAULT: - ext4_iset_acl(inode, &ei->i_default_acl, acl); + ext4_iset_acl(inode, &inode->i_default_acl, acl); break; } } @@ -226,7 +225,6 @@ static int ext4_set_acl(handle_t *handle, struct inode *inode, int type, struct posix_acl *acl) { - struct ext4_inode_info *ei = EXT4_I(inode); int name_index; void *value = NULL; size_t size = 0; @@ -274,11 +272,11 @@ ext4_set_acl(handle_t *handle, struct inode *inode, int type, if (!error) { switch (type) { case ACL_TYPE_ACCESS: - ext4_iset_acl(inode, &ei->i_acl, acl); + ext4_iset_acl(inode, &inode->i_acl, acl); break; case ACL_TYPE_DEFAULT: - ext4_iset_acl(inode, &ei->i_default_acl, acl); + ext4_iset_acl(inode, &inode->i_default_acl, acl); break; } } diff --git a/fs/ext4/acl.h b/fs/ext4/acl.h index cb45257a246e..949789d2bba6 100644 --- a/fs/ext4/acl.h +++ b/fs/ext4/acl.h @@ -53,10 +53,6 @@ static inline int ext4_acl_count(size_t size) #ifdef CONFIG_EXT4_FS_POSIX_ACL -/* Value for inode->u.ext4_i.i_acl and inode->u.ext4_i.i_default_acl - if the ACL has not been cached */ -#define EXT4_ACL_NOT_CACHED ((void *)-1) - /* acl.c */ extern int ext4_permission(struct inode *, int); extern int ext4_acl_chmod(struct inode *); diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 17b9998680e3..0ddf7e55abe1 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -595,10 +595,6 @@ struct ext4_inode_info { */ struct rw_semaphore xattr_sem; #endif -#ifdef CONFIG_EXT4_FS_POSIX_ACL - struct posix_acl *i_acl; - struct posix_acl *i_default_acl; -#endif struct list_head i_orphan; /* unlinked but open inodes */ diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 7c17ae275af4..60a26f3a6f8b 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -4453,10 +4453,6 @@ struct inode *ext4_iget(struct super_block *sb, unsigned long ino) return inode; ei = EXT4_I(inode); -#ifdef CONFIG_EXT4_FS_POSIX_ACL - ei->i_acl = EXT4_ACL_NOT_CACHED; - ei->i_default_acl = EXT4_ACL_NOT_CACHED; -#endif ret = __ext4_get_inode_loc(inode, &iloc, 0); if (ret < 0) diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 8bb9e2d3e4b8..8f4f079e6b9a 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -666,10 +666,6 @@ static struct inode *ext4_alloc_inode(struct super_block *sb) if (!ei) return NULL; -#ifdef CONFIG_EXT4_FS_POSIX_ACL - ei->i_acl = EXT4_ACL_NOT_CACHED; - ei->i_default_acl = EXT4_ACL_NOT_CACHED; -#endif ei->vfs_inode.i_version = 1; ei->vfs_inode.i_data.writeback_index = 0; memset(&ei->i_cached_extent, 0, sizeof(struct ext4_ext_cache)); @@ -735,18 +731,6 @@ static void destroy_inodecache(void) static void ext4_clear_inode(struct inode *inode) { -#ifdef CONFIG_EXT4_FS_POSIX_ACL - if (EXT4_I(inode)->i_acl && - EXT4_I(inode)->i_acl != EXT4_ACL_NOT_CACHED) { - posix_acl_release(EXT4_I(inode)->i_acl); - EXT4_I(inode)->i_acl = EXT4_ACL_NOT_CACHED; - } - if (EXT4_I(inode)->i_default_acl && - EXT4_I(inode)->i_default_acl != EXT4_ACL_NOT_CACHED) { - posix_acl_release(EXT4_I(inode)->i_default_acl); - EXT4_I(inode)->i_default_acl = EXT4_ACL_NOT_CACHED; - } -#endif ext4_discard_preallocations(inode); if (EXT4_JOURNAL(inode)) jbd2_journal_release_jbd_inode(EXT4_SB(inode->i_sb)->s_journal, -- cgit v1.2.3 From 05fc0790b6c9c611129f2f712d00b6a8a364e8d2 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Mon, 8 Jun 2009 19:54:52 -0400 Subject: switch jfs to inode->i_acl Signed-off-by: Al Viro --- fs/jfs/acl.c | 22 +++++++++------------- fs/jfs/jfs_incore.h | 6 ------ fs/jfs/super.c | 16 ---------------- fs/jfs/xattr.c | 12 ++++++------ 4 files changed, 15 insertions(+), 41 deletions(-) (limited to 'fs') diff --git a/fs/jfs/acl.c b/fs/jfs/acl.c index 06ca1b8d2054..5fcfc9857c11 100644 --- a/fs/jfs/acl.c +++ b/fs/jfs/acl.c @@ -31,7 +31,6 @@ static struct posix_acl *jfs_get_acl(struct inode *inode, int type) { struct posix_acl *acl; char *ea_name; - struct jfs_inode_info *ji = JFS_IP(inode); struct posix_acl **p_acl; int size; char *value = NULL; @@ -39,17 +38,17 @@ static struct posix_acl *jfs_get_acl(struct inode *inode, int type) switch(type) { case ACL_TYPE_ACCESS: ea_name = POSIX_ACL_XATTR_ACCESS; - p_acl = &ji->i_acl; + p_acl = &inode->i_acl; break; case ACL_TYPE_DEFAULT: ea_name = POSIX_ACL_XATTR_DEFAULT; - p_acl = &ji->i_default_acl; + p_acl = &inode->i_default_acl; break; default: return ERR_PTR(-EINVAL); } - if (*p_acl != JFS_ACL_NOT_CACHED) + if (*p_acl != ACL_NOT_CACHED) return posix_acl_dup(*p_acl); size = __jfs_getxattr(inode, ea_name, NULL, 0); @@ -80,7 +79,6 @@ static int jfs_set_acl(tid_t tid, struct inode *inode, int type, struct posix_acl *acl) { char *ea_name; - struct jfs_inode_info *ji = JFS_IP(inode); struct posix_acl **p_acl; int rc; int size = 0; @@ -92,11 +90,11 @@ static int jfs_set_acl(tid_t tid, struct inode *inode, int type, switch(type) { case ACL_TYPE_ACCESS: ea_name = POSIX_ACL_XATTR_ACCESS; - p_acl = &ji->i_acl; + p_acl = &inode->i_acl; break; case ACL_TYPE_DEFAULT: ea_name = POSIX_ACL_XATTR_DEFAULT; - p_acl = &ji->i_default_acl; + p_acl = &inode->i_default_acl; if (!S_ISDIR(inode->i_mode)) return acl ? -EACCES : 0; break; @@ -117,7 +115,7 @@ out: kfree(value); if (!rc) { - if (*p_acl && (*p_acl != JFS_ACL_NOT_CACHED)) + if (*p_acl && (*p_acl != ACL_NOT_CACHED)) posix_acl_release(*p_acl); *p_acl = posix_acl_dup(acl); } @@ -126,17 +124,15 @@ out: static int jfs_check_acl(struct inode *inode, int mask) { - struct jfs_inode_info *ji = JFS_IP(inode); - - if (ji->i_acl == JFS_ACL_NOT_CACHED) { + if (inode->i_acl == ACL_NOT_CACHED) { struct posix_acl *acl = jfs_get_acl(inode, ACL_TYPE_ACCESS); if (IS_ERR(acl)) return PTR_ERR(acl); posix_acl_release(acl); } - if (ji->i_acl) - return posix_acl_permission(inode, ji->i_acl, mask); + if (inode->i_acl) + return posix_acl_permission(inode, inode->i_acl, mask); return -EAGAIN; } diff --git a/fs/jfs/jfs_incore.h b/fs/jfs/jfs_incore.h index 439901d205fe..1439f119ec83 100644 --- a/fs/jfs/jfs_incore.h +++ b/fs/jfs/jfs_incore.h @@ -74,10 +74,6 @@ struct jfs_inode_info { /* xattr_sem allows us to access the xattrs without taking i_mutex */ struct rw_semaphore xattr_sem; lid_t xtlid; /* lid of xtree lock on directory */ -#ifdef CONFIG_JFS_POSIX_ACL - struct posix_acl *i_acl; - struct posix_acl *i_default_acl; -#endif union { struct { xtpage_t _xtroot; /* 288: xtree root */ @@ -107,8 +103,6 @@ struct jfs_inode_info { #define i_inline u.link._inline #define i_inline_ea u.link._inline_ea -#define JFS_ACL_NOT_CACHED ((void *)-1) - #define IREAD_LOCK(ip, subclass) \ down_read_nested(&JFS_IP(ip)->rdwrlock, subclass) #define IREAD_UNLOCK(ip) up_read(&JFS_IP(ip)->rdwrlock) diff --git a/fs/jfs/super.c b/fs/jfs/super.c index 09b1b6ee2186..37e6dcda8fc8 100644 --- a/fs/jfs/super.c +++ b/fs/jfs/super.c @@ -128,18 +128,6 @@ static void jfs_destroy_inode(struct inode *inode) ji->active_ag = -1; } spin_unlock_irq(&ji->ag_lock); - -#ifdef CONFIG_JFS_POSIX_ACL - if (ji->i_acl != JFS_ACL_NOT_CACHED) { - posix_acl_release(ji->i_acl); - ji->i_acl = JFS_ACL_NOT_CACHED; - } - if (ji->i_default_acl != JFS_ACL_NOT_CACHED) { - posix_acl_release(ji->i_default_acl); - ji->i_default_acl = JFS_ACL_NOT_CACHED; - } -#endif - kmem_cache_free(jfs_inode_cachep, ji); } @@ -798,10 +786,6 @@ static void init_once(void *foo) init_rwsem(&jfs_ip->xattr_sem); spin_lock_init(&jfs_ip->ag_lock); jfs_ip->active_ag = -1; -#ifdef CONFIG_JFS_POSIX_ACL - jfs_ip->i_acl = JFS_ACL_NOT_CACHED; - jfs_ip->i_default_acl = JFS_ACL_NOT_CACHED; -#endif inode_init_once(&jfs_ip->vfs_inode); } diff --git a/fs/jfs/xattr.c b/fs/jfs/xattr.c index 61dfa8173ebc..f6e90e343593 100644 --- a/fs/jfs/xattr.c +++ b/fs/jfs/xattr.c @@ -727,10 +727,10 @@ static int can_set_system_xattr(struct inode *inode, const char *name, /* * We're changing the ACL. Get rid of the cached one */ - acl =JFS_IP(inode)->i_acl; - if (acl != JFS_ACL_NOT_CACHED) + acl =inode->i_acl; + if (acl != ACL_NOT_CACHED) posix_acl_release(acl); - JFS_IP(inode)->i_acl = JFS_ACL_NOT_CACHED; + inode->i_acl = ACL_NOT_CACHED; return 0; } else if (strcmp(name, POSIX_ACL_XATTR_DEFAULT) == 0) { @@ -746,10 +746,10 @@ static int can_set_system_xattr(struct inode *inode, const char *name, /* * We're changing the default ACL. Get rid of the cached one */ - acl =JFS_IP(inode)->i_default_acl; - if (acl && (acl != JFS_ACL_NOT_CACHED)) + acl = inode->i_default_acl; + if (acl && (acl != ACL_NOT_CACHED)) posix_acl_release(acl); - JFS_IP(inode)->i_default_acl = JFS_ACL_NOT_CACHED; + inode->i_default_acl = ACL_NOT_CACHED; return 0; } -- cgit v1.2.3 From 290c263bf83cd78e53b1aa3b42165f588163f2be Mon Sep 17 00:00:00 2001 From: Al Viro Date: Mon, 8 Jun 2009 19:55:12 -0400 Subject: switch jffs2 to inode->i_acl Signed-off-by: Al Viro --- fs/jffs2/acl.c | 54 ++++++++++++++++++--------------------------------- fs/jffs2/acl.h | 4 ---- fs/jffs2/jffs2_fs_i.h | 4 ---- fs/jffs2/os-linux.h | 4 ---- fs/jffs2/readinode.c | 1 - 5 files changed, 19 insertions(+), 48 deletions(-) (limited to 'fs') diff --git a/fs/jffs2/acl.c b/fs/jffs2/acl.c index 043740dde20c..ac16589ebbd1 100644 --- a/fs/jffs2/acl.c +++ b/fs/jffs2/acl.c @@ -158,10 +158,10 @@ static void *jffs2_acl_to_medium(const struct posix_acl *acl, size_t *size) static struct posix_acl *jffs2_iget_acl(struct inode *inode, struct posix_acl **i_acl) { - struct posix_acl *acl = JFFS2_ACL_NOT_CACHED; + struct posix_acl *acl = ACL_NOT_CACHED; spin_lock(&inode->i_lock); - if (*i_acl != JFFS2_ACL_NOT_CACHED) + if (*i_acl != ACL_NOT_CACHED) acl = posix_acl_dup(*i_acl); spin_unlock(&inode->i_lock); return acl; @@ -170,7 +170,7 @@ static struct posix_acl *jffs2_iget_acl(struct inode *inode, struct posix_acl ** static void jffs2_iset_acl(struct inode *inode, struct posix_acl **i_acl, struct posix_acl *acl) { spin_lock(&inode->i_lock); - if (*i_acl != JFFS2_ACL_NOT_CACHED) + if (*i_acl != ACL_NOT_CACHED) posix_acl_release(*i_acl); *i_acl = posix_acl_dup(acl); spin_unlock(&inode->i_lock); @@ -178,21 +178,20 @@ static void jffs2_iset_acl(struct inode *inode, struct posix_acl **i_acl, struct static struct posix_acl *jffs2_get_acl(struct inode *inode, int type) { - struct jffs2_inode_info *f = JFFS2_INODE_INFO(inode); struct posix_acl *acl; char *value = NULL; int rc, xprefix; switch (type) { case ACL_TYPE_ACCESS: - acl = jffs2_iget_acl(inode, &f->i_acl_access); - if (acl != JFFS2_ACL_NOT_CACHED) + acl = jffs2_iget_acl(inode, &inode->i_acl); + if (acl != ACL_NOT_CACHED) return acl; xprefix = JFFS2_XPREFIX_ACL_ACCESS; break; case ACL_TYPE_DEFAULT: - acl = jffs2_iget_acl(inode, &f->i_acl_default); - if (acl != JFFS2_ACL_NOT_CACHED) + acl = jffs2_iget_acl(inode, &inode->i_default_acl); + if (acl != ACL_NOT_CACHED) return acl; xprefix = JFFS2_XPREFIX_ACL_DEFAULT; break; @@ -218,10 +217,10 @@ static struct posix_acl *jffs2_get_acl(struct inode *inode, int type) if (!IS_ERR(acl)) { switch (type) { case ACL_TYPE_ACCESS: - jffs2_iset_acl(inode, &f->i_acl_access, acl); + jffs2_iset_acl(inode, &inode->i_acl, acl); break; case ACL_TYPE_DEFAULT: - jffs2_iset_acl(inode, &f->i_acl_default, acl); + jffs2_iset_acl(inode, &inode->i_default_acl, acl); break; } } @@ -249,7 +248,6 @@ static int __jffs2_set_acl(struct inode *inode, int xprefix, struct posix_acl *a static int jffs2_set_acl(struct inode *inode, int type, struct posix_acl *acl) { - struct jffs2_inode_info *f = JFFS2_INODE_INFO(inode); int rc, xprefix; if (S_ISLNK(inode->i_mode)) @@ -288,10 +286,10 @@ static int jffs2_set_acl(struct inode *inode, int type, struct posix_acl *acl) if (!rc) { switch(type) { case ACL_TYPE_ACCESS: - jffs2_iset_acl(inode, &f->i_acl_access, acl); + jffs2_iset_acl(inode, &inode->i_acl, acl); break; case ACL_TYPE_DEFAULT: - jffs2_iset_acl(inode, &f->i_acl_default, acl); + jffs2_iset_acl(inode, &inode->i_default_acl, acl); break; } } @@ -321,12 +319,11 @@ int jffs2_permission(struct inode *inode, int mask) int jffs2_init_acl_pre(struct inode *dir_i, struct inode *inode, int *i_mode) { - struct jffs2_inode_info *f = JFFS2_INODE_INFO(inode); struct posix_acl *acl, *clone; int rc; - f->i_acl_default = NULL; - f->i_acl_access = NULL; + inode->i_default_acl = NULL; + inode->i_acl = NULL; if (S_ISLNK(*i_mode)) return 0; /* Symlink always has no-ACL */ @@ -339,7 +336,7 @@ int jffs2_init_acl_pre(struct inode *dir_i, struct inode *inode, int *i_mode) *i_mode &= ~current_umask(); } else { if (S_ISDIR(*i_mode)) - jffs2_iset_acl(inode, &f->i_acl_default, acl); + jffs2_iset_acl(inode, &inode->i_default_acl, acl); clone = posix_acl_clone(acl, GFP_KERNEL); if (!clone) @@ -350,7 +347,7 @@ int jffs2_init_acl_pre(struct inode *dir_i, struct inode *inode, int *i_mode) return rc; } if (rc > 0) - jffs2_iset_acl(inode, &f->i_acl_access, clone); + jffs2_iset_acl(inode, &inode->i_acl, clone); posix_acl_release(clone); } @@ -359,17 +356,16 @@ int jffs2_init_acl_pre(struct inode *dir_i, struct inode *inode, int *i_mode) int jffs2_init_acl_post(struct inode *inode) { - struct jffs2_inode_info *f = JFFS2_INODE_INFO(inode); int rc; - if (f->i_acl_default) { - rc = __jffs2_set_acl(inode, JFFS2_XPREFIX_ACL_DEFAULT, f->i_acl_default); + if (inode->i_default_acl) { + rc = __jffs2_set_acl(inode, JFFS2_XPREFIX_ACL_DEFAULT, inode->i_default_acl); if (rc) return rc; } - if (f->i_acl_access) { - rc = __jffs2_set_acl(inode, JFFS2_XPREFIX_ACL_ACCESS, f->i_acl_access); + if (inode->i_acl) { + rc = __jffs2_set_acl(inode, JFFS2_XPREFIX_ACL_ACCESS, inode->i_acl); if (rc) return rc; } @@ -377,18 +373,6 @@ int jffs2_init_acl_post(struct inode *inode) return 0; } -void jffs2_clear_acl(struct jffs2_inode_info *f) -{ - if (f->i_acl_access && f->i_acl_access != JFFS2_ACL_NOT_CACHED) { - posix_acl_release(f->i_acl_access); - f->i_acl_access = JFFS2_ACL_NOT_CACHED; - } - if (f->i_acl_default && f->i_acl_default != JFFS2_ACL_NOT_CACHED) { - posix_acl_release(f->i_acl_default); - f->i_acl_default = JFFS2_ACL_NOT_CACHED; - } -} - int jffs2_acl_chmod(struct inode *inode) { struct posix_acl *acl, *clone; diff --git a/fs/jffs2/acl.h b/fs/jffs2/acl.h index 8ca058aed384..fc929f2a14f6 100644 --- a/fs/jffs2/acl.h +++ b/fs/jffs2/acl.h @@ -26,13 +26,10 @@ struct jffs2_acl_header { #ifdef CONFIG_JFFS2_FS_POSIX_ACL -#define JFFS2_ACL_NOT_CACHED ((void *)-1) - extern int jffs2_permission(struct inode *, int); extern int jffs2_acl_chmod(struct inode *); extern int jffs2_init_acl_pre(struct inode *, struct inode *, int *); extern int jffs2_init_acl_post(struct inode *); -extern void jffs2_clear_acl(struct jffs2_inode_info *); extern struct xattr_handler jffs2_acl_access_xattr_handler; extern struct xattr_handler jffs2_acl_default_xattr_handler; @@ -43,6 +40,5 @@ extern struct xattr_handler jffs2_acl_default_xattr_handler; #define jffs2_acl_chmod(inode) (0) #define jffs2_init_acl_pre(dir_i,inode,mode) (0) #define jffs2_init_acl_post(inode) (0) -#define jffs2_clear_acl(f) #endif /* CONFIG_JFFS2_FS_POSIX_ACL */ diff --git a/fs/jffs2/jffs2_fs_i.h b/fs/jffs2/jffs2_fs_i.h index 4c41db91eaa4..c6923da98263 100644 --- a/fs/jffs2/jffs2_fs_i.h +++ b/fs/jffs2/jffs2_fs_i.h @@ -50,10 +50,6 @@ struct jffs2_inode_info { uint16_t flags; uint8_t usercompr; struct inode vfs_inode; -#ifdef CONFIG_JFFS2_FS_POSIX_ACL - struct posix_acl *i_acl_access; - struct posix_acl *i_acl_default; -#endif }; #endif /* _JFFS2_FS_I */ diff --git a/fs/jffs2/os-linux.h b/fs/jffs2/os-linux.h index 2228380c47b9..a7f03b7ebcb3 100644 --- a/fs/jffs2/os-linux.h +++ b/fs/jffs2/os-linux.h @@ -56,10 +56,6 @@ static inline void jffs2_init_inode_info(struct jffs2_inode_info *f) f->target = NULL; f->flags = 0; f->usercompr = 0; -#ifdef CONFIG_JFFS2_FS_POSIX_ACL - f->i_acl_access = JFFS2_ACL_NOT_CACHED; - f->i_acl_default = JFFS2_ACL_NOT_CACHED; -#endif } diff --git a/fs/jffs2/readinode.c b/fs/jffs2/readinode.c index 1fc1e92356ee..1a80301004b8 100644 --- a/fs/jffs2/readinode.c +++ b/fs/jffs2/readinode.c @@ -1424,7 +1424,6 @@ void jffs2_do_clear_inode(struct jffs2_sb_info *c, struct jffs2_inode_info *f) struct jffs2_full_dirent *fd, *fds; int deleted; - jffs2_clear_acl(f); jffs2_xattr_delete_inode(c, f->inocache); mutex_lock(&f->sem); deleted = f->inocache && !f->inocache->pino_nlink; -- cgit v1.2.3 From 5affd88a104af43f0063a12ad1ee4c7a587945dc Mon Sep 17 00:00:00 2001 From: Al Viro Date: Mon, 8 Jun 2009 19:55:32 -0400 Subject: switch btrfs to inode->i_acl Signed-off-by: Al Viro --- fs/btrfs/acl.c | 14 +++++++------- fs/btrfs/btrfs_inode.h | 4 ---- fs/btrfs/ctree.h | 2 -- fs/btrfs/inode.c | 16 ++-------------- 4 files changed, 9 insertions(+), 27 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/acl.c b/fs/btrfs/acl.c index 603972576f0f..6db8a42a3e5e 100644 --- a/fs/btrfs/acl.c +++ b/fs/btrfs/acl.c @@ -34,7 +34,7 @@ static void btrfs_update_cached_acl(struct inode *inode, struct posix_acl *acl) { spin_lock(&inode->i_lock); - if (*p_acl && *p_acl != BTRFS_ACL_NOT_CACHED) + if (*p_acl && *p_acl != ACL_NOT_CACHED) posix_acl_release(*p_acl); *p_acl = posix_acl_dup(acl); spin_unlock(&inode->i_lock); @@ -50,11 +50,11 @@ static struct posix_acl *btrfs_get_acl(struct inode *inode, int type) switch (type) { case ACL_TYPE_ACCESS: name = POSIX_ACL_XATTR_ACCESS; - p_acl = &BTRFS_I(inode)->i_acl; + p_acl = &inode->i_acl; break; case ACL_TYPE_DEFAULT: name = POSIX_ACL_XATTR_DEFAULT; - p_acl = &BTRFS_I(inode)->i_default_acl; + p_acl = &inode->i_default_acl; break; default: return ERR_PTR(-EINVAL); @@ -67,11 +67,11 @@ static struct posix_acl *btrfs_get_acl(struct inode *inode, int type) spin_lock(&inode->i_lock); acl = *p_acl; - if (acl != BTRFS_ACL_NOT_CACHED) + if (acl != ACL_NOT_CACHED) acl = posix_acl_dup(acl); spin_unlock(&inode->i_lock); - if (acl != BTRFS_ACL_NOT_CACHED) + if (acl != ACL_NOT_CACHED) return acl; size = __btrfs_getxattr(inode, name, "", 0); @@ -141,13 +141,13 @@ static int btrfs_set_acl(struct inode *inode, struct posix_acl *acl, int type) ret = 0; inode->i_mode = mode; name = POSIX_ACL_XATTR_ACCESS; - p_acl = &BTRFS_I(inode)->i_acl; + p_acl = &inode->i_acl; break; case ACL_TYPE_DEFAULT: if (!S_ISDIR(inode->i_mode)) return acl ? -EINVAL : 0; name = POSIX_ACL_XATTR_DEFAULT; - p_acl = &BTRFS_I(inode)->i_default_acl; + p_acl = &inode->i_default_acl; break; default: return -EINVAL; diff --git a/fs/btrfs/btrfs_inode.h b/fs/btrfs/btrfs_inode.h index acb4f3517582..ea1ea0af8c0e 100644 --- a/fs/btrfs/btrfs_inode.h +++ b/fs/btrfs/btrfs_inode.h @@ -53,10 +53,6 @@ struct btrfs_inode { /* used to order data wrt metadata */ struct btrfs_ordered_inode_tree ordered_tree; - /* standard acl pointers */ - struct posix_acl *i_acl; - struct posix_acl *i_default_acl; - /* for keeping track of orphaned inodes */ struct list_head i_orphan; diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 03441a99ea38..2779c2f5360a 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -41,8 +41,6 @@ struct btrfs_ordered_sum; #define BTRFS_MAGIC "_BHRfS_M" -#define BTRFS_ACL_NOT_CACHED ((void *)-1) - #define BTRFS_MAX_LEVEL 8 #define BTRFS_COMPAT_EXTENT_TREE_V0 diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 8612b3a09811..78ad38ddd01f 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -2123,8 +2123,8 @@ static void btrfs_read_locked_inode(struct inode *inode) */ maybe_acls = acls_after_inode_item(leaf, path->slots[0], inode->i_ino); if (!maybe_acls) { - BTRFS_I(inode)->i_acl = NULL; - BTRFS_I(inode)->i_default_acl = NULL; + inode->i_acl = NULL; + inode->i_default_acl = NULL; } BTRFS_I(inode)->block_group = btrfs_find_block_group(root, 0, @@ -3141,9 +3141,6 @@ static noinline void init_btrfs_i(struct inode *inode) { struct btrfs_inode *bi = BTRFS_I(inode); - bi->i_acl = BTRFS_ACL_NOT_CACHED; - bi->i_default_acl = BTRFS_ACL_NOT_CACHED; - bi->generation = 0; bi->sequence = 0; bi->last_trans = 0; @@ -4640,8 +4637,6 @@ struct inode *btrfs_alloc_inode(struct super_block *sb) ei->last_trans = 0; ei->logged_trans = 0; btrfs_ordered_inode_tree_init(&ei->ordered_tree); - ei->i_acl = BTRFS_ACL_NOT_CACHED; - ei->i_default_acl = BTRFS_ACL_NOT_CACHED; INIT_LIST_HEAD(&ei->i_orphan); INIT_LIST_HEAD(&ei->ordered_operations); return &ei->vfs_inode; @@ -4655,13 +4650,6 @@ void btrfs_destroy_inode(struct inode *inode) WARN_ON(!list_empty(&inode->i_dentry)); WARN_ON(inode->i_data.nrpages); - if (BTRFS_I(inode)->i_acl && - BTRFS_I(inode)->i_acl != BTRFS_ACL_NOT_CACHED) - posix_acl_release(BTRFS_I(inode)->i_acl); - if (BTRFS_I(inode)->i_default_acl && - BTRFS_I(inode)->i_default_acl != BTRFS_ACL_NOT_CACHED) - posix_acl_release(BTRFS_I(inode)->i_default_acl); - /* * Make sure we're properly removed from the ordered operation * lists. -- cgit v1.2.3 From d441b1c293149212045de00f346c8ea6cd41cce4 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Mon, 8 Jun 2009 19:56:34 -0400 Subject: switch nilfs2 to inode->i_acl Actually, get rid of private analog, since nothing in there is using ACLs at all so far. Signed-off-by: Al Viro --- fs/nilfs2/inode.c | 8 -------- fs/nilfs2/nilfs.h | 4 ---- fs/nilfs2/super.c | 10 ---------- 3 files changed, 22 deletions(-) (limited to 'fs') diff --git a/fs/nilfs2/inode.c b/fs/nilfs2/inode.c index 2696d6b513b7..fe9d8f2a13f8 100644 --- a/fs/nilfs2/inode.c +++ b/fs/nilfs2/inode.c @@ -309,10 +309,6 @@ struct inode *nilfs_new_inode(struct inode *dir, int mode) /* ii->i_file_acl = 0; */ /* ii->i_dir_acl = 0; */ ii->i_dir_start_lookup = 0; -#ifdef CONFIG_NILFS_FS_POSIX_ACL - ii->i_acl = NULL; - ii->i_default_acl = NULL; -#endif ii->i_cno = 0; nilfs_set_inode_flags(inode); spin_lock(&sbi->s_next_gen_lock); @@ -434,10 +430,6 @@ static int __nilfs_read_inode(struct super_block *sb, unsigned long ino, raw_inode = nilfs_ifile_map_inode(sbi->s_ifile, ino, bh); -#ifdef CONFIG_NILFS_FS_POSIX_ACL - ii->i_acl = NILFS_ACL_NOT_CACHED; - ii->i_default_acl = NILFS_ACL_NOT_CACHED; -#endif if (nilfs_read_inode_common(inode, raw_inode)) goto failed_unmap; diff --git a/fs/nilfs2/nilfs.h b/fs/nilfs2/nilfs.h index edf6a59d9f2a..724c63766e82 100644 --- a/fs/nilfs2/nilfs.h +++ b/fs/nilfs2/nilfs.h @@ -57,10 +57,6 @@ struct nilfs_inode_info { * EAs. */ struct rw_semaphore xattr_sem; -#endif -#ifdef CONFIG_NILFS_POSIX_ACL - struct posix_acl *i_acl; - struct posix_acl *i_default_acl; #endif struct buffer_head *i_bh; /* i_bh contains a new or dirty disk inode */ diff --git a/fs/nilfs2/super.c b/fs/nilfs2/super.c index ab785f85aa50..8e2ec43b18f4 100644 --- a/fs/nilfs2/super.c +++ b/fs/nilfs2/super.c @@ -189,16 +189,6 @@ static void nilfs_clear_inode(struct inode *inode) { struct nilfs_inode_info *ii = NILFS_I(inode); -#ifdef CONFIG_NILFS_POSIX_ACL - if (ii->i_acl && ii->i_acl != NILFS_ACL_NOT_CACHED) { - posix_acl_release(ii->i_acl); - ii->i_acl = NILFS_ACL_NOT_CACHED; - } - if (ii->i_default_acl && ii->i_default_acl != NILFS_ACL_NOT_CACHED) { - posix_acl_release(ii->i_default_acl); - ii->i_default_acl = NILFS_ACL_NOT_CACHED; - } -#endif /* * Free resources allocated in nilfs_read_inode(), here. */ -- cgit v1.2.3 From e68888bcb60ccba4dc21df9f2d8cd7325b64dce7 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Mon, 8 Jun 2009 20:46:54 -0400 Subject: reiserfs: minimal fix for ACL caching reiserfs uses NULL as "unknown" and ERR_PTR(-ENODATA) as "no ACL"; several codepaths store the former instead of the latter. All those codepaths go through iset_acl() and all cases when it's called with NULL acl are for the second variety, so the minimal fix is to teach iset_acl() to deal with that. Proper fix is to switch to more usual conventions and avoid back and forth between internally used ERR_PTR(-ENODATA) and NULL expected by the rest of the kernel. Signed-off-by: Al Viro --- fs/reiserfs/xattr_acl.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/reiserfs/xattr_acl.c b/fs/reiserfs/xattr_acl.c index c303c426fe2b..a1a7e3530e17 100644 --- a/fs/reiserfs/xattr_acl.c +++ b/fs/reiserfs/xattr_acl.c @@ -194,7 +194,7 @@ static inline void iset_acl(struct inode *inode, struct posix_acl **i_acl, spin_lock(&inode->i_lock); if (*i_acl != ERR_PTR(-ENODATA)) posix_acl_release(*i_acl); - *i_acl = posix_acl_dup(acl); + *i_acl = acl ? posix_acl_dup(acl) : ERR_PTR(-ENODATA); spin_unlock(&inode->i_lock); } -- cgit v1.2.3 From 7a77b15d9294749809de918e24bebc39e0fbc9ab Mon Sep 17 00:00:00 2001 From: Al Viro Date: Mon, 8 Jun 2009 21:01:13 -0400 Subject: switch reiserfs to usual conventions for caching ACLs Signed-off-by: Al Viro --- fs/reiserfs/super.c | 12 ++++++------ fs/reiserfs/xattr_acl.c | 21 ++++++++------------- include/linux/reiserfs_acl.h | 4 ++-- 3 files changed, 16 insertions(+), 21 deletions(-) (limited to 'fs') diff --git a/fs/reiserfs/super.c b/fs/reiserfs/super.c index 2969773cfc22..b194451fc04b 100644 --- a/fs/reiserfs/super.c +++ b/fs/reiserfs/super.c @@ -530,8 +530,8 @@ static void init_once(void *foo) INIT_LIST_HEAD(&ei->i_prealloc_list); inode_init_once(&ei->vfs_inode); #ifdef CONFIG_REISERFS_FS_POSIX_ACL - ei->i_acl_access = NULL; - ei->i_acl_default = NULL; + ei->i_acl_access = ACL_NOT_CACHED; + ei->i_acl_default = ACL_NOT_CACHED; #endif } @@ -586,14 +586,14 @@ static void reiserfs_clear_inode(struct inode *inode) struct posix_acl *acl; acl = REISERFS_I(inode)->i_acl_access; - if (acl && !IS_ERR(acl)) + if (acl && acl != ACL_NOT_CACHED) posix_acl_release(acl); - REISERFS_I(inode)->i_acl_access = NULL; + REISERFS_I(inode)->i_acl_access = ACL_NOT_CACHED; acl = REISERFS_I(inode)->i_acl_default; - if (acl && !IS_ERR(acl)) + if (acl && acl != ACL_NOT_CACHED) posix_acl_release(acl); - REISERFS_I(inode)->i_acl_default = NULL; + REISERFS_I(inode)->i_acl_default = ACL_NOT_CACHED; } #else #define reiserfs_clear_inode NULL diff --git a/fs/reiserfs/xattr_acl.c b/fs/reiserfs/xattr_acl.c index a1a7e3530e17..7b3aeb9327d3 100644 --- a/fs/reiserfs/xattr_acl.c +++ b/fs/reiserfs/xattr_acl.c @@ -192,19 +192,19 @@ static inline void iset_acl(struct inode *inode, struct posix_acl **i_acl, struct posix_acl *acl) { spin_lock(&inode->i_lock); - if (*i_acl != ERR_PTR(-ENODATA)) + if (*i_acl != ACL_NOT_CACHED) posix_acl_release(*i_acl); - *i_acl = acl ? posix_acl_dup(acl) : ERR_PTR(-ENODATA); + *i_acl = posix_acl_dup(acl); spin_unlock(&inode->i_lock); } static inline struct posix_acl *iget_acl(struct inode *inode, struct posix_acl **i_acl) { - struct posix_acl *acl = ERR_PTR(-ENODATA); + struct posix_acl *acl = ACL_NOT_CACHED; spin_lock(&inode->i_lock); - if (*i_acl != ERR_PTR(-ENODATA)) + if (*i_acl != ACL_NOT_CACHED) acl = posix_acl_dup(*i_acl); spin_unlock(&inode->i_lock); @@ -239,15 +239,13 @@ struct posix_acl *reiserfs_get_acl(struct inode *inode, int type) } acl = iget_acl(inode, p_acl); - if (acl && !IS_ERR(acl)) + if (acl != ACL_NOT_CACHED) return acl; - else if (PTR_ERR(acl) == -ENODATA) - return NULL; size = reiserfs_xattr_get(inode, name, NULL, 0); if (size < 0) { if (size == -ENODATA || size == -ENOSYS) { - *p_acl = ERR_PTR(-ENODATA); + *p_acl = NULL; return NULL; } return ERR_PTR(size); @@ -262,7 +260,7 @@ struct posix_acl *reiserfs_get_acl(struct inode *inode, int type) /* This shouldn't actually happen as it should have been caught above.. but just in case */ acl = NULL; - *p_acl = ERR_PTR(-ENODATA); + *p_acl = acl; } else if (retval < 0) { acl = ERR_PTR(retval); } else { @@ -379,11 +377,8 @@ reiserfs_inherit_default_acl(struct reiserfs_transaction_handle *th, } acl = reiserfs_get_acl(dir, ACL_TYPE_DEFAULT); - if (IS_ERR(acl)) { - if (PTR_ERR(acl) == -ENODATA) - goto apply_umask; + if (IS_ERR(acl)) return PTR_ERR(acl); - } if (acl) { struct posix_acl *acl_copy; diff --git a/include/linux/reiserfs_acl.h b/include/linux/reiserfs_acl.h index 8cc65757e47a..8f4d8d718b10 100644 --- a/include/linux/reiserfs_acl.h +++ b/include/linux/reiserfs_acl.h @@ -58,12 +58,12 @@ extern struct xattr_handler reiserfs_posix_acl_access_handler; static inline void reiserfs_init_acl_access(struct inode *inode) { - REISERFS_I(inode)->i_acl_access = NULL; + REISERFS_I(inode)->i_acl_access = ACL_NOT_CACHED; } static inline void reiserfs_init_acl_default(struct inode *inode) { - REISERFS_I(inode)->i_acl_default = NULL; + REISERFS_I(inode)->i_acl_default = ACL_NOT_CACHED; } #else -- cgit v1.2.3 From 281eede0328c84a8f20e0e85b807d5b51c3de4f2 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Mon, 8 Jun 2009 21:07:04 -0400 Subject: switch reiserfs to inode->i_acl Signed-off-by: Al Viro --- fs/reiserfs/inode.c | 4 ---- fs/reiserfs/super.c | 24 ------------------------ fs/reiserfs/xattr_acl.c | 10 ++++------ include/linux/reiserfs_acl.h | 17 ----------------- include/linux/reiserfs_fs_i.h | 4 ---- 5 files changed, 4 insertions(+), 55 deletions(-) (limited to 'fs') diff --git a/fs/reiserfs/inode.c b/fs/reiserfs/inode.c index 6fd0f47e45db..a14d6cd9eeda 100644 --- a/fs/reiserfs/inode.c +++ b/fs/reiserfs/inode.c @@ -1131,8 +1131,6 @@ static void init_inode(struct inode *inode, struct treepath *path) REISERFS_I(inode)->i_trans_id = 0; REISERFS_I(inode)->i_jl = NULL; mutex_init(&(REISERFS_I(inode)->i_mmap)); - reiserfs_init_acl_access(inode); - reiserfs_init_acl_default(inode); reiserfs_init_xattr_rwsem(inode); if (stat_data_v1(ih)) { @@ -1834,8 +1832,6 @@ int reiserfs_new_inode(struct reiserfs_transaction_handle *th, REISERFS_I(dir)->i_attrs & REISERFS_INHERIT_MASK; sd_attrs_to_i_attrs(REISERFS_I(inode)->i_attrs, inode); mutex_init(&(REISERFS_I(inode)->i_mmap)); - reiserfs_init_acl_access(inode); - reiserfs_init_acl_default(inode); reiserfs_init_xattr_rwsem(inode); /* key to search for correct place for new stat data */ diff --git a/fs/reiserfs/super.c b/fs/reiserfs/super.c index b194451fc04b..d3aeb061612b 100644 --- a/fs/reiserfs/super.c +++ b/fs/reiserfs/super.c @@ -529,10 +529,6 @@ static void init_once(void *foo) INIT_LIST_HEAD(&ei->i_prealloc_list); inode_init_once(&ei->vfs_inode); -#ifdef CONFIG_REISERFS_FS_POSIX_ACL - ei->i_acl_access = ACL_NOT_CACHED; - ei->i_acl_default = ACL_NOT_CACHED; -#endif } static int init_inodecache(void) @@ -580,25 +576,6 @@ static void reiserfs_dirty_inode(struct inode *inode) reiserfs_write_unlock(inode->i_sb); } -#ifdef CONFIG_REISERFS_FS_POSIX_ACL -static void reiserfs_clear_inode(struct inode *inode) -{ - struct posix_acl *acl; - - acl = REISERFS_I(inode)->i_acl_access; - if (acl && acl != ACL_NOT_CACHED) - posix_acl_release(acl); - REISERFS_I(inode)->i_acl_access = ACL_NOT_CACHED; - - acl = REISERFS_I(inode)->i_acl_default; - if (acl && acl != ACL_NOT_CACHED) - posix_acl_release(acl); - REISERFS_I(inode)->i_acl_default = ACL_NOT_CACHED; -} -#else -#define reiserfs_clear_inode NULL -#endif - #ifdef CONFIG_QUOTA static ssize_t reiserfs_quota_write(struct super_block *, int, const char *, size_t, loff_t); @@ -612,7 +589,6 @@ static const struct super_operations reiserfs_sops = { .write_inode = reiserfs_write_inode, .dirty_inode = reiserfs_dirty_inode, .delete_inode = reiserfs_delete_inode, - .clear_inode = reiserfs_clear_inode, .put_super = reiserfs_put_super, .write_super = reiserfs_write_super, .sync_fs = reiserfs_sync_fs, diff --git a/fs/reiserfs/xattr_acl.c b/fs/reiserfs/xattr_acl.c index 7b3aeb9327d3..b6e473faa8b8 100644 --- a/fs/reiserfs/xattr_acl.c +++ b/fs/reiserfs/xattr_acl.c @@ -223,16 +223,15 @@ struct posix_acl *reiserfs_get_acl(struct inode *inode, int type) struct posix_acl *acl, **p_acl; int size; int retval; - struct reiserfs_inode_info *reiserfs_i = REISERFS_I(inode); switch (type) { case ACL_TYPE_ACCESS: name = POSIX_ACL_XATTR_ACCESS; - p_acl = &reiserfs_i->i_acl_access; + p_acl = &inode->i_acl; break; case ACL_TYPE_DEFAULT: name = POSIX_ACL_XATTR_DEFAULT; - p_acl = &reiserfs_i->i_acl_default; + p_acl = &inode->i_default_acl; break; default: return ERR_PTR(-EINVAL); @@ -288,7 +287,6 @@ reiserfs_set_acl(struct reiserfs_transaction_handle *th, struct inode *inode, struct posix_acl **p_acl; size_t size = 0; int error; - struct reiserfs_inode_info *reiserfs_i = REISERFS_I(inode); if (S_ISLNK(inode->i_mode)) return -EOPNOTSUPP; @@ -296,7 +294,7 @@ reiserfs_set_acl(struct reiserfs_transaction_handle *th, struct inode *inode, switch (type) { case ACL_TYPE_ACCESS: name = POSIX_ACL_XATTR_ACCESS; - p_acl = &reiserfs_i->i_acl_access; + p_acl = &inode->i_acl; if (acl) { mode_t mode = inode->i_mode; error = posix_acl_equiv_mode(acl, &mode); @@ -311,7 +309,7 @@ reiserfs_set_acl(struct reiserfs_transaction_handle *th, struct inode *inode, break; case ACL_TYPE_DEFAULT: name = POSIX_ACL_XATTR_DEFAULT; - p_acl = &reiserfs_i->i_acl_default; + p_acl = &inode->i_default_acl; if (!S_ISDIR(inode->i_mode)) return acl ? -EACCES : 0; break; diff --git a/include/linux/reiserfs_acl.h b/include/linux/reiserfs_acl.h index 8f4d8d718b10..b4448853900e 100644 --- a/include/linux/reiserfs_acl.h +++ b/include/linux/reiserfs_acl.h @@ -56,15 +56,6 @@ int reiserfs_cache_default_acl(struct inode *dir); extern struct xattr_handler reiserfs_posix_acl_default_handler; extern struct xattr_handler reiserfs_posix_acl_access_handler; -static inline void reiserfs_init_acl_access(struct inode *inode) -{ - REISERFS_I(inode)->i_acl_access = ACL_NOT_CACHED; -} - -static inline void reiserfs_init_acl_default(struct inode *inode) -{ - REISERFS_I(inode)->i_acl_default = ACL_NOT_CACHED; -} #else #define reiserfs_cache_default_acl(inode) 0 @@ -86,12 +77,4 @@ reiserfs_inherit_default_acl(struct reiserfs_transaction_handle *th, { return 0; } - -static inline void reiserfs_init_acl_access(struct inode *inode) -{ -} - -static inline void reiserfs_init_acl_default(struct inode *inode) -{ -} #endif diff --git a/include/linux/reiserfs_fs_i.h b/include/linux/reiserfs_fs_i.h index 76360b36ac33..89f4d3abbf5a 100644 --- a/include/linux/reiserfs_fs_i.h +++ b/include/linux/reiserfs_fs_i.h @@ -54,10 +54,6 @@ struct reiserfs_inode_info { unsigned int i_trans_id; struct reiserfs_journal_list *i_jl; struct mutex i_mmap; -#ifdef CONFIG_REISERFS_FS_POSIX_ACL - struct posix_acl *i_acl_access; - struct posix_acl *i_acl_default; -#endif #ifdef CONFIG_REISERFS_FS_XATTR struct rw_semaphore i_xattr_sem; #endif -- cgit v1.2.3 From 073aaa1b142461d91f83da66db1184d7c1b1edea Mon Sep 17 00:00:00 2001 From: Al Viro Date: Tue, 9 Jun 2009 12:11:54 -0400 Subject: helpers for acl caching + switch to those helpers: get_cached_acl(inode, type), set_cached_acl(inode, type, acl), forget_cached_acl(inode, type). ubifs/xattr.c needed includes reordered, the rest is a plain switchover. Signed-off-by: Al Viro --- fs/btrfs/acl.c | 44 +++++-------------------- fs/ext2/acl.c | 79 ++++++++++---------------------------------- fs/ext3/acl.c | 83 +++++++++++------------------------------------ fs/ext4/acl.c | 65 +++++-------------------------------- fs/jffs2/acl.c | 60 +++++++--------------------------- fs/jfs/acl.c | 32 ++++++++---------- fs/jfs/xattr.c | 10 ++---- fs/reiserfs/xattr_acl.c | 49 ++++++---------------------- fs/ubifs/xattr.c | 2 +- include/linux/posix_acl.h | 64 ++++++++++++++++++++++++++++++++++++ 10 files changed, 155 insertions(+), 333 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/acl.c b/fs/btrfs/acl.c index 6db8a42a3e5e..f128427b995b 100644 --- a/fs/btrfs/acl.c +++ b/fs/btrfs/acl.c @@ -29,51 +29,28 @@ #ifdef CONFIG_FS_POSIX_ACL -static void btrfs_update_cached_acl(struct inode *inode, - struct posix_acl **p_acl, - struct posix_acl *acl) -{ - spin_lock(&inode->i_lock); - if (*p_acl && *p_acl != ACL_NOT_CACHED) - posix_acl_release(*p_acl); - *p_acl = posix_acl_dup(acl); - spin_unlock(&inode->i_lock); -} - static struct posix_acl *btrfs_get_acl(struct inode *inode, int type) { int size; const char *name; char *value = NULL; - struct posix_acl *acl = NULL, **p_acl; + struct posix_acl *acl; + + acl = get_cached_acl(inode, type); + if (acl != ACL_NOT_CACHED) + return acl; switch (type) { case ACL_TYPE_ACCESS: name = POSIX_ACL_XATTR_ACCESS; - p_acl = &inode->i_acl; break; case ACL_TYPE_DEFAULT: name = POSIX_ACL_XATTR_DEFAULT; - p_acl = &inode->i_default_acl; break; default: - return ERR_PTR(-EINVAL); + BUG(); } - /* Handle the cached NULL acl case without locking */ - acl = ACCESS_ONCE(*p_acl); - if (!acl) - return acl; - - spin_lock(&inode->i_lock); - acl = *p_acl; - if (acl != ACL_NOT_CACHED) - acl = posix_acl_dup(acl); - spin_unlock(&inode->i_lock); - - if (acl != ACL_NOT_CACHED) - return acl; - size = __btrfs_getxattr(inode, name, "", 0); if (size > 0) { value = kzalloc(size, GFP_NOFS); @@ -82,13 +59,13 @@ static struct posix_acl *btrfs_get_acl(struct inode *inode, int type) size = __btrfs_getxattr(inode, name, value, size); if (size > 0) { acl = posix_acl_from_xattr(value, size); - btrfs_update_cached_acl(inode, p_acl, acl); + set_cached_acl(inode, type, acl); } kfree(value); } else if (size == -ENOENT || size == -ENODATA || size == 0) { /* FIXME, who returns -ENOENT? I think nobody */ acl = NULL; - btrfs_update_cached_acl(inode, p_acl, acl); + set_cached_acl(inode, type, acl); } else { acl = ERR_PTR(-EIO); } @@ -121,7 +98,6 @@ static int btrfs_set_acl(struct inode *inode, struct posix_acl *acl, int type) { int ret, size = 0; const char *name; - struct posix_acl **p_acl; char *value = NULL; mode_t mode; @@ -141,13 +117,11 @@ static int btrfs_set_acl(struct inode *inode, struct posix_acl *acl, int type) ret = 0; inode->i_mode = mode; name = POSIX_ACL_XATTR_ACCESS; - p_acl = &inode->i_acl; break; case ACL_TYPE_DEFAULT: if (!S_ISDIR(inode->i_mode)) return acl ? -EINVAL : 0; name = POSIX_ACL_XATTR_DEFAULT; - p_acl = &inode->i_default_acl; break; default: return -EINVAL; @@ -172,7 +146,7 @@ out: kfree(value); if (!ret) - btrfs_update_cached_acl(inode, p_acl, acl); + set_cached_acl(inode, type, acl); return ret; } diff --git a/fs/ext2/acl.c b/fs/ext2/acl.c index d2ffddc12117..d636e1297cad 100644 --- a/fs/ext2/acl.c +++ b/fs/ext2/acl.c @@ -125,30 +125,6 @@ fail: return ERR_PTR(-EINVAL); } -static inline struct posix_acl * -ext2_iget_acl(struct inode *inode, struct posix_acl **i_acl) -{ - struct posix_acl *acl = ACL_NOT_CACHED; - - spin_lock(&inode->i_lock); - if (*i_acl != ACL_NOT_CACHED) - acl = posix_acl_dup(*i_acl); - spin_unlock(&inode->i_lock); - - return acl; -} - -static inline void -ext2_iset_acl(struct inode *inode, struct posix_acl **i_acl, - struct posix_acl *acl) -{ - spin_lock(&inode->i_lock); - if (*i_acl != ACL_NOT_CACHED) - posix_acl_release(*i_acl); - *i_acl = posix_acl_dup(acl); - spin_unlock(&inode->i_lock); -} - /* * inode->i_mutex: don't care */ @@ -163,23 +139,19 @@ ext2_get_acl(struct inode *inode, int type) if (!test_opt(inode->i_sb, POSIX_ACL)) return NULL; - switch(type) { - case ACL_TYPE_ACCESS: - acl = ext2_iget_acl(inode, &inode->i_acl); - if (acl != ACL_NOT_CACHED) - return acl; - name_index = EXT2_XATTR_INDEX_POSIX_ACL_ACCESS; - break; - - case ACL_TYPE_DEFAULT: - acl = ext2_iget_acl(inode, &inode->i_default_acl); - if (acl != ACL_NOT_CACHED) - return acl; - name_index = EXT2_XATTR_INDEX_POSIX_ACL_DEFAULT; - break; - - default: - return ERR_PTR(-EINVAL); + acl = get_cached_acl(inode, type); + if (acl != ACL_NOT_CACHED) + return acl; + + switch (type) { + case ACL_TYPE_ACCESS: + name_index = EXT2_XATTR_INDEX_POSIX_ACL_ACCESS; + break; + case ACL_TYPE_DEFAULT: + name_index = EXT2_XATTR_INDEX_POSIX_ACL_DEFAULT; + break; + default: + BUG(); } retval = ext2_xattr_get(inode, name_index, "", NULL, 0); if (retval > 0) { @@ -196,17 +168,9 @@ ext2_get_acl(struct inode *inode, int type) acl = ERR_PTR(retval); kfree(value); - if (!IS_ERR(acl)) { - switch(type) { - case ACL_TYPE_ACCESS: - ext2_iset_acl(inode, &inode->i_acl, acl); - break; + if (!IS_ERR(acl)) + set_cached_acl(inode, type, acl); - case ACL_TYPE_DEFAULT: - ext2_iset_acl(inode, &inode->i_default_acl, acl); - break; - } - } return acl; } @@ -261,17 +225,8 @@ ext2_set_acl(struct inode *inode, int type, struct posix_acl *acl) error = ext2_xattr_set(inode, name_index, "", value, size, 0); kfree(value); - if (!error) { - switch(type) { - case ACL_TYPE_ACCESS: - ext2_iset_acl(inode, &inode->i_acl, acl); - break; - - case ACL_TYPE_DEFAULT: - ext2_iset_acl(inode, &inode->i_default_acl, acl); - break; - } - } + if (!error) + set_cached_acl(inode, type, acl); return error; } diff --git a/fs/ext3/acl.c b/fs/ext3/acl.c index a9707689d9e1..e167bae37ef0 100644 --- a/fs/ext3/acl.c +++ b/fs/ext3/acl.c @@ -126,33 +126,6 @@ fail: return ERR_PTR(-EINVAL); } -static inline struct posix_acl * -ext3_iget_acl(struct inode *inode, struct posix_acl **i_acl) -{ - struct posix_acl *acl = ACCESS_ONCE(*i_acl); - - if (acl) { - spin_lock(&inode->i_lock); - acl = *i_acl; - if (acl != ACL_NOT_CACHED) - acl = posix_acl_dup(acl); - spin_unlock(&inode->i_lock); - } - - return acl; -} - -static inline void -ext3_iset_acl(struct inode *inode, struct posix_acl **i_acl, - struct posix_acl *acl) -{ - spin_lock(&inode->i_lock); - if (*i_acl != ACL_NOT_CACHED) - posix_acl_release(*i_acl); - *i_acl = posix_acl_dup(acl); - spin_unlock(&inode->i_lock); -} - /* * Inode operation get_posix_acl(). * @@ -169,24 +142,21 @@ ext3_get_acl(struct inode *inode, int type) if (!test_opt(inode->i_sb, POSIX_ACL)) return NULL; - switch(type) { - case ACL_TYPE_ACCESS: - acl = ext3_iget_acl(inode, &inode->i_acl); - if (acl != ACL_NOT_CACHED) - return acl; - name_index = EXT3_XATTR_INDEX_POSIX_ACL_ACCESS; - break; - - case ACL_TYPE_DEFAULT: - acl = ext3_iget_acl(inode, &inode->i_default_acl); - if (acl != ACL_NOT_CACHED) - return acl; - name_index = EXT3_XATTR_INDEX_POSIX_ACL_DEFAULT; - break; - - default: - return ERR_PTR(-EINVAL); + acl = get_cached_acl(inode, type); + if (acl != ACL_NOT_CACHED) + return acl; + + switch (type) { + case ACL_TYPE_ACCESS: + name_index = EXT3_XATTR_INDEX_POSIX_ACL_ACCESS; + break; + case ACL_TYPE_DEFAULT: + name_index = EXT3_XATTR_INDEX_POSIX_ACL_DEFAULT; + break; + default: + BUG(); } + retval = ext3_xattr_get(inode, name_index, "", NULL, 0); if (retval > 0) { value = kmalloc(retval, GFP_NOFS); @@ -202,17 +172,9 @@ ext3_get_acl(struct inode *inode, int type) acl = ERR_PTR(retval); kfree(value); - if (!IS_ERR(acl)) { - switch(type) { - case ACL_TYPE_ACCESS: - ext3_iset_acl(inode, &inode->i_acl, acl); - break; + if (!IS_ERR(acl)) + set_cached_acl(inode, type, acl); - case ACL_TYPE_DEFAULT: - ext3_iset_acl(inode, &inode->i_default_acl, acl); - break; - } - } return acl; } @@ -269,17 +231,10 @@ ext3_set_acl(handle_t *handle, struct inode *inode, int type, value, size, 0); kfree(value); - if (!error) { - switch(type) { - case ACL_TYPE_ACCESS: - ext3_iset_acl(inode, &inode->i_acl, acl); - break; - case ACL_TYPE_DEFAULT: - ext3_iset_acl(inode, &inode->i_default_acl, acl); - break; - } - } + if (!error) + set_cached_acl(inode, type, acl); + return error; } diff --git a/fs/ext4/acl.c b/fs/ext4/acl.c index 0084e3a19d86..f6d8967149ca 100644 --- a/fs/ext4/acl.c +++ b/fs/ext4/acl.c @@ -126,33 +126,6 @@ fail: return ERR_PTR(-EINVAL); } -static inline struct posix_acl * -ext4_iget_acl(struct inode *inode, struct posix_acl **i_acl) -{ - struct posix_acl *acl = ACCESS_ONCE(*i_acl); - - if (acl) { - spin_lock(&inode->i_lock); - acl = *i_acl; - if (acl != ACL_NOT_CACHED) - acl = posix_acl_dup(acl); - spin_unlock(&inode->i_lock); - } - - return acl; -} - -static inline void -ext4_iset_acl(struct inode *inode, struct posix_acl **i_acl, - struct posix_acl *acl) -{ - spin_lock(&inode->i_lock); - if (*i_acl != ACL_NOT_CACHED) - posix_acl_release(*i_acl); - *i_acl = posix_acl_dup(acl); - spin_unlock(&inode->i_lock); -} - /* * Inode operation get_posix_acl(). * @@ -169,23 +142,19 @@ ext4_get_acl(struct inode *inode, int type) if (!test_opt(inode->i_sb, POSIX_ACL)) return NULL; + acl = get_cached_acl(inode, type); + if (acl != ACL_NOT_CACHED) + return acl; + switch (type) { case ACL_TYPE_ACCESS: - acl = ext4_iget_acl(inode, &inode->i_acl); - if (acl != ACL_NOT_CACHED) - return acl; name_index = EXT4_XATTR_INDEX_POSIX_ACL_ACCESS; break; - case ACL_TYPE_DEFAULT: - acl = ext4_iget_acl(inode, &inode->i_default_acl); - if (acl != ACL_NOT_CACHED) - return acl; name_index = EXT4_XATTR_INDEX_POSIX_ACL_DEFAULT; break; - default: - return ERR_PTR(-EINVAL); + BUG(); } retval = ext4_xattr_get(inode, name_index, "", NULL, 0); if (retval > 0) { @@ -202,17 +171,9 @@ ext4_get_acl(struct inode *inode, int type) acl = ERR_PTR(retval); kfree(value); - if (!IS_ERR(acl)) { - switch (type) { - case ACL_TYPE_ACCESS: - ext4_iset_acl(inode, &inode->i_acl, acl); - break; + if (!IS_ERR(acl)) + set_cached_acl(inode, type, acl); - case ACL_TYPE_DEFAULT: - ext4_iset_acl(inode, &inode->i_default_acl, acl); - break; - } - } return acl; } @@ -269,17 +230,9 @@ ext4_set_acl(handle_t *handle, struct inode *inode, int type, value, size, 0); kfree(value); - if (!error) { - switch (type) { - case ACL_TYPE_ACCESS: - ext4_iset_acl(inode, &inode->i_acl, acl); - break; + if (!error) + set_cached_acl(inode, type, acl); - case ACL_TYPE_DEFAULT: - ext4_iset_acl(inode, &inode->i_default_acl, acl); - break; - } - } return error; } diff --git a/fs/jffs2/acl.c b/fs/jffs2/acl.c index ac16589ebbd1..edd2ad6416d8 100644 --- a/fs/jffs2/acl.c +++ b/fs/jffs2/acl.c @@ -156,47 +156,25 @@ static void *jffs2_acl_to_medium(const struct posix_acl *acl, size_t *size) return ERR_PTR(-EINVAL); } -static struct posix_acl *jffs2_iget_acl(struct inode *inode, struct posix_acl **i_acl) -{ - struct posix_acl *acl = ACL_NOT_CACHED; - - spin_lock(&inode->i_lock); - if (*i_acl != ACL_NOT_CACHED) - acl = posix_acl_dup(*i_acl); - spin_unlock(&inode->i_lock); - return acl; -} - -static void jffs2_iset_acl(struct inode *inode, struct posix_acl **i_acl, struct posix_acl *acl) -{ - spin_lock(&inode->i_lock); - if (*i_acl != ACL_NOT_CACHED) - posix_acl_release(*i_acl); - *i_acl = posix_acl_dup(acl); - spin_unlock(&inode->i_lock); -} - static struct posix_acl *jffs2_get_acl(struct inode *inode, int type) { struct posix_acl *acl; char *value = NULL; int rc, xprefix; + acl = get_cached_acl(inode, type); + if (acl != ACL_NOT_CACHED) + return acl; + switch (type) { case ACL_TYPE_ACCESS: - acl = jffs2_iget_acl(inode, &inode->i_acl); - if (acl != ACL_NOT_CACHED) - return acl; xprefix = JFFS2_XPREFIX_ACL_ACCESS; break; case ACL_TYPE_DEFAULT: - acl = jffs2_iget_acl(inode, &inode->i_default_acl); - if (acl != ACL_NOT_CACHED) - return acl; xprefix = JFFS2_XPREFIX_ACL_DEFAULT; break; default: - return ERR_PTR(-EINVAL); + BUG(); } rc = do_jffs2_getxattr(inode, xprefix, "", NULL, 0); if (rc > 0) { @@ -214,16 +192,8 @@ static struct posix_acl *jffs2_get_acl(struct inode *inode, int type) } if (value) kfree(value); - if (!IS_ERR(acl)) { - switch (type) { - case ACL_TYPE_ACCESS: - jffs2_iset_acl(inode, &inode->i_acl, acl); - break; - case ACL_TYPE_DEFAULT: - jffs2_iset_acl(inode, &inode->i_default_acl, acl); - break; - } - } + if (!IS_ERR(acl)) + set_cached_acl(inode, type, acl); return acl; } @@ -283,16 +253,8 @@ static int jffs2_set_acl(struct inode *inode, int type, struct posix_acl *acl) return -EINVAL; } rc = __jffs2_set_acl(inode, xprefix, acl); - if (!rc) { - switch(type) { - case ACL_TYPE_ACCESS: - jffs2_iset_acl(inode, &inode->i_acl, acl); - break; - case ACL_TYPE_DEFAULT: - jffs2_iset_acl(inode, &inode->i_default_acl, acl); - break; - } - } + if (!rc) + set_cached_acl(inode, type, acl); return rc; } @@ -336,7 +298,7 @@ int jffs2_init_acl_pre(struct inode *dir_i, struct inode *inode, int *i_mode) *i_mode &= ~current_umask(); } else { if (S_ISDIR(*i_mode)) - jffs2_iset_acl(inode, &inode->i_default_acl, acl); + set_cached_acl(inode, ACL_TYPE_DEFAULT, acl); clone = posix_acl_clone(acl, GFP_KERNEL); if (!clone) @@ -347,7 +309,7 @@ int jffs2_init_acl_pre(struct inode *dir_i, struct inode *inode, int *i_mode) return rc; } if (rc > 0) - jffs2_iset_acl(inode, &inode->i_acl, clone); + set_cached_acl(inode, ACL_TYPE_ACCESS, clone); posix_acl_release(clone); } diff --git a/fs/jfs/acl.c b/fs/jfs/acl.c index 5fcfc9857c11..f272bf032e1e 100644 --- a/fs/jfs/acl.c +++ b/fs/jfs/acl.c @@ -31,26 +31,24 @@ static struct posix_acl *jfs_get_acl(struct inode *inode, int type) { struct posix_acl *acl; char *ea_name; - struct posix_acl **p_acl; int size; char *value = NULL; + acl = get_cached_acl(inode, type); + if (acl != ACL_NOT_CACHED) + return acl; + switch(type) { case ACL_TYPE_ACCESS: ea_name = POSIX_ACL_XATTR_ACCESS; - p_acl = &inode->i_acl; break; case ACL_TYPE_DEFAULT: ea_name = POSIX_ACL_XATTR_DEFAULT; - p_acl = &inode->i_default_acl; break; default: return ERR_PTR(-EINVAL); } - if (*p_acl != ACL_NOT_CACHED) - return posix_acl_dup(*p_acl); - size = __jfs_getxattr(inode, ea_name, NULL, 0); if (size > 0) { @@ -61,17 +59,18 @@ static struct posix_acl *jfs_get_acl(struct inode *inode, int type) } if (size < 0) { - if (size == -ENODATA) { - *p_acl = NULL; + if (size == -ENODATA) acl = NULL; - } else + else acl = ERR_PTR(size); } else { acl = posix_acl_from_xattr(value, size); - if (!IS_ERR(acl)) - *p_acl = posix_acl_dup(acl); } kfree(value); + if (!IS_ERR(acl)) { + set_cached_acl(inode, type, acl); + posix_acl_release(acl); + } return acl; } @@ -79,7 +78,6 @@ static int jfs_set_acl(tid_t tid, struct inode *inode, int type, struct posix_acl *acl) { char *ea_name; - struct posix_acl **p_acl; int rc; int size = 0; char *value = NULL; @@ -90,11 +88,9 @@ static int jfs_set_acl(tid_t tid, struct inode *inode, int type, switch(type) { case ACL_TYPE_ACCESS: ea_name = POSIX_ACL_XATTR_ACCESS; - p_acl = &inode->i_acl; break; case ACL_TYPE_DEFAULT: ea_name = POSIX_ACL_XATTR_DEFAULT; - p_acl = &inode->i_default_acl; if (!S_ISDIR(inode->i_mode)) return acl ? -EACCES : 0; break; @@ -114,11 +110,9 @@ static int jfs_set_acl(tid_t tid, struct inode *inode, int type, out: kfree(value); - if (!rc) { - if (*p_acl && (*p_acl != ACL_NOT_CACHED)) - posix_acl_release(*p_acl); - *p_acl = posix_acl_dup(acl); - } + if (!rc) + set_cached_acl(inode, type, acl); + return rc; } diff --git a/fs/jfs/xattr.c b/fs/jfs/xattr.c index f6e90e343593..fad364548bc9 100644 --- a/fs/jfs/xattr.c +++ b/fs/jfs/xattr.c @@ -727,10 +727,7 @@ static int can_set_system_xattr(struct inode *inode, const char *name, /* * We're changing the ACL. Get rid of the cached one */ - acl =inode->i_acl; - if (acl != ACL_NOT_CACHED) - posix_acl_release(acl); - inode->i_acl = ACL_NOT_CACHED; + forget_cached_acl(inode, ACL_TYPE_ACCESS); return 0; } else if (strcmp(name, POSIX_ACL_XATTR_DEFAULT) == 0) { @@ -746,10 +743,7 @@ static int can_set_system_xattr(struct inode *inode, const char *name, /* * We're changing the default ACL. Get rid of the cached one */ - acl = inode->i_default_acl; - if (acl && (acl != ACL_NOT_CACHED)) - posix_acl_release(acl); - inode->i_default_acl = ACL_NOT_CACHED; + forget_cached_acl(inode, ACL_TYPE_DEFAULT); return 0; } diff --git a/fs/reiserfs/xattr_acl.c b/fs/reiserfs/xattr_acl.c index b6e473faa8b8..35d6e672a279 100644 --- a/fs/reiserfs/xattr_acl.c +++ b/fs/reiserfs/xattr_acl.c @@ -188,29 +188,6 @@ static void *posix_acl_to_disk(const struct posix_acl *acl, size_t * size) return ERR_PTR(-EINVAL); } -static inline void iset_acl(struct inode *inode, struct posix_acl **i_acl, - struct posix_acl *acl) -{ - spin_lock(&inode->i_lock); - if (*i_acl != ACL_NOT_CACHED) - posix_acl_release(*i_acl); - *i_acl = posix_acl_dup(acl); - spin_unlock(&inode->i_lock); -} - -static inline struct posix_acl *iget_acl(struct inode *inode, - struct posix_acl **i_acl) -{ - struct posix_acl *acl = ACL_NOT_CACHED; - - spin_lock(&inode->i_lock); - if (*i_acl != ACL_NOT_CACHED) - acl = posix_acl_dup(*i_acl); - spin_unlock(&inode->i_lock); - - return acl; -} - /* * Inode operation get_posix_acl(). * @@ -220,31 +197,29 @@ static inline struct posix_acl *iget_acl(struct inode *inode, struct posix_acl *reiserfs_get_acl(struct inode *inode, int type) { char *name, *value; - struct posix_acl *acl, **p_acl; + struct posix_acl *acl; int size; int retval; + acl = get_cached_acl(inode, type); + if (acl != ACL_NOT_CACHED) + return acl; + switch (type) { case ACL_TYPE_ACCESS: name = POSIX_ACL_XATTR_ACCESS; - p_acl = &inode->i_acl; break; case ACL_TYPE_DEFAULT: name = POSIX_ACL_XATTR_DEFAULT; - p_acl = &inode->i_default_acl; break; default: - return ERR_PTR(-EINVAL); + BUG(); } - acl = iget_acl(inode, p_acl); - if (acl != ACL_NOT_CACHED) - return acl; - size = reiserfs_xattr_get(inode, name, NULL, 0); if (size < 0) { if (size == -ENODATA || size == -ENOSYS) { - *p_acl = NULL; + set_cached_acl(inode, type, NULL); return NULL; } return ERR_PTR(size); @@ -259,14 +234,13 @@ struct posix_acl *reiserfs_get_acl(struct inode *inode, int type) /* This shouldn't actually happen as it should have been caught above.. but just in case */ acl = NULL; - *p_acl = acl; } else if (retval < 0) { acl = ERR_PTR(retval); } else { acl = posix_acl_from_disk(value, retval); - if (!IS_ERR(acl)) - iset_acl(inode, p_acl, acl); } + if (!IS_ERR(acl)) + set_cached_acl(inode, type, acl); kfree(value); return acl; @@ -284,7 +258,6 @@ reiserfs_set_acl(struct reiserfs_transaction_handle *th, struct inode *inode, { char *name; void *value = NULL; - struct posix_acl **p_acl; size_t size = 0; int error; @@ -294,7 +267,6 @@ reiserfs_set_acl(struct reiserfs_transaction_handle *th, struct inode *inode, switch (type) { case ACL_TYPE_ACCESS: name = POSIX_ACL_XATTR_ACCESS; - p_acl = &inode->i_acl; if (acl) { mode_t mode = inode->i_mode; error = posix_acl_equiv_mode(acl, &mode); @@ -309,7 +281,6 @@ reiserfs_set_acl(struct reiserfs_transaction_handle *th, struct inode *inode, break; case ACL_TYPE_DEFAULT: name = POSIX_ACL_XATTR_DEFAULT; - p_acl = &inode->i_default_acl; if (!S_ISDIR(inode->i_mode)) return acl ? -EACCES : 0; break; @@ -342,7 +313,7 @@ reiserfs_set_acl(struct reiserfs_transaction_handle *th, struct inode *inode, kfree(value); if (!error) - iset_acl(inode, p_acl, acl); + set_cached_acl(inode, type, acl); return error; } diff --git a/fs/ubifs/xattr.c b/fs/ubifs/xattr.c index cfd31e229c89..adafcf556531 100644 --- a/fs/ubifs/xattr.c +++ b/fs/ubifs/xattr.c @@ -55,9 +55,9 @@ * ACL support is not implemented. */ +#include "ubifs.h" #include #include -#include "ubifs.h" /* * Limit the number of extended attributes per inode so that the total size diff --git a/include/linux/posix_acl.h b/include/linux/posix_acl.h index 4bc241290c24..0cdba01b7756 100644 --- a/include/linux/posix_acl.h +++ b/include/linux/posix_acl.h @@ -83,4 +83,68 @@ extern int posix_acl_chmod_masq(struct posix_acl *, mode_t); extern struct posix_acl *get_posix_acl(struct inode *, int); extern int set_posix_acl(struct inode *, int, struct posix_acl *); +static inline struct posix_acl *get_cached_acl(struct inode *inode, int type) +{ + struct posix_acl **p, *acl; + switch (type) { + case ACL_TYPE_ACCESS: + p = &inode->i_acl; + break; + case ACL_TYPE_DEFAULT: + p = &inode->i_default_acl; + break; + default: + return ERR_PTR(-EINVAL); + } + acl = ACCESS_ONCE(*p); + if (acl) { + spin_lock(&inode->i_lock); + acl = *p; + if (acl != ACL_NOT_CACHED) + acl = posix_acl_dup(acl); + spin_unlock(&inode->i_lock); + } + return acl; +} + +static inline void set_cached_acl(struct inode *inode, + int type, + struct posix_acl *acl) +{ + struct posix_acl *old = NULL; + spin_lock(&inode->i_lock); + switch (type) { + case ACL_TYPE_ACCESS: + old = inode->i_acl; + inode->i_acl = posix_acl_dup(acl); + break; + case ACL_TYPE_DEFAULT: + old = inode->i_default_acl; + inode->i_default_acl = posix_acl_dup(acl); + break; + } + spin_unlock(&inode->i_lock); + if (old != ACL_NOT_CACHED) + posix_acl_release(old); +} + +static inline void forget_cached_acl(struct inode *inode, int type) +{ + struct posix_acl *old = NULL; + spin_lock(&inode->i_lock); + switch (type) { + case ACL_TYPE_ACCESS: + old = inode->i_acl; + inode->i_acl = ACL_NOT_CACHED; + break; + case ACL_TYPE_DEFAULT: + old = inode->i_default_acl; + inode->i_default_acl = ACL_NOT_CACHED; + break; + } + spin_unlock(&inode->i_lock); + if (old != ACL_NOT_CACHED) + posix_acl_release(old); +} + #endif /* __LINUX_POSIX_ACL_H */ -- cgit v1.2.3 From 1cbd20d820c36f52543e3e4cd0067ebf52aa388f Mon Sep 17 00:00:00 2001 From: Al Viro Date: Tue, 9 Jun 2009 13:29:39 -0400 Subject: switch xfs to generic acl caching helpers Signed-off-by: Al Viro --- fs/xfs/linux-2.6/xfs_acl.c | 73 ++++++---------------------------------------- fs/xfs/xfs_acl.h | 4 --- fs/xfs/xfs_iget.c | 2 -- fs/xfs/xfs_inode.h | 5 ---- 4 files changed, 9 insertions(+), 75 deletions(-) (limited to 'fs') diff --git a/fs/xfs/linux-2.6/xfs_acl.c b/fs/xfs/linux-2.6/xfs_acl.c index 1e9d1246eebc..b23a54506446 100644 --- a/fs/xfs/linux-2.6/xfs_acl.c +++ b/fs/xfs/linux-2.6/xfs_acl.c @@ -25,14 +25,10 @@ #include -#define XFS_ACL_NOT_CACHED ((void *)-1) - /* * Locking scheme: * - all ACL updates are protected by inode->i_mutex, which is taken before * calling into this file. - * - access and updates to the ip->i_acl and ip->i_default_acl pointers are - * protected by inode->i_lock. */ STATIC struct posix_acl * @@ -102,59 +98,35 @@ xfs_acl_to_disk(struct xfs_acl *aclp, const struct posix_acl *acl) } } -/* - * Update the cached ACL pointer in the inode. - * - * Because we don't hold any locks while reading/writing the attribute - * from/to disk another thread could have raced and updated the cached - * ACL value before us. In that case we release the previous cached value - * and update it with our new value. - */ -STATIC void -xfs_update_cached_acl(struct inode *inode, struct posix_acl **p_acl, - struct posix_acl *acl) -{ - spin_lock(&inode->i_lock); - if (*p_acl && *p_acl != XFS_ACL_NOT_CACHED) - posix_acl_release(*p_acl); - *p_acl = posix_acl_dup(acl); - spin_unlock(&inode->i_lock); -} - struct posix_acl * xfs_get_acl(struct inode *inode, int type) { struct xfs_inode *ip = XFS_I(inode); - struct posix_acl *acl = NULL, **p_acl; + struct posix_acl *acl; struct xfs_acl *xfs_acl; int len = sizeof(struct xfs_acl); char *ea_name; int error; + acl = get_cached_acl(inode, type); + if (acl != ACL_NOT_CACHED) + return acl; + switch (type) { case ACL_TYPE_ACCESS: ea_name = SGI_ACL_FILE; - p_acl = &ip->i_acl; break; case ACL_TYPE_DEFAULT: ea_name = SGI_ACL_DEFAULT; - p_acl = &ip->i_default_acl; break; default: - return ERR_PTR(-EINVAL); + BUG(); } - spin_lock(&inode->i_lock); - if (*p_acl != XFS_ACL_NOT_CACHED) - acl = posix_acl_dup(*p_acl); - spin_unlock(&inode->i_lock); - /* * If we have a cached ACLs value just return it, not need to * go out to the disk. */ - if (acl) - return acl; xfs_acl = kzalloc(sizeof(struct xfs_acl), GFP_KERNEL); if (!xfs_acl) @@ -165,7 +137,7 @@ xfs_get_acl(struct inode *inode, int type) /* * If the attribute doesn't exist make sure we have a negative * cache entry, for any other error assume it is transient and - * leave the cache entry as XFS_ACL_NOT_CACHED. + * leave the cache entry as ACL_NOT_CACHED. */ if (error == -ENOATTR) { acl = NULL; @@ -179,7 +151,7 @@ xfs_get_acl(struct inode *inode, int type) goto out; out_update_cache: - xfs_update_cached_acl(inode, p_acl, acl); + set_cached_acl(inode, type, acl); out: kfree(xfs_acl); return acl; @@ -189,7 +161,6 @@ STATIC int xfs_set_acl(struct inode *inode, int type, struct posix_acl *acl) { struct xfs_inode *ip = XFS_I(inode); - struct posix_acl **p_acl; char *ea_name; int error; @@ -199,13 +170,11 @@ xfs_set_acl(struct inode *inode, int type, struct posix_acl *acl) switch (type) { case ACL_TYPE_ACCESS: ea_name = SGI_ACL_FILE; - p_acl = &ip->i_acl; break; case ACL_TYPE_DEFAULT: if (!S_ISDIR(inode->i_mode)) return acl ? -EACCES : 0; ea_name = SGI_ACL_DEFAULT; - p_acl = &ip->i_default_acl; break; default: return -EINVAL; @@ -242,7 +211,7 @@ xfs_set_acl(struct inode *inode, int type, struct posix_acl *acl) } if (!error) - xfs_update_cached_acl(inode, p_acl, acl); + set_cached_acl(inode, type, acl); return error; } @@ -384,30 +353,6 @@ xfs_acl_chmod(struct inode *inode) return error; } -void -xfs_inode_init_acls(struct xfs_inode *ip) -{ - /* - * No need for locking, inode is not live yet. - */ - ip->i_acl = XFS_ACL_NOT_CACHED; - ip->i_default_acl = XFS_ACL_NOT_CACHED; -} - -void -xfs_inode_clear_acls(struct xfs_inode *ip) -{ - /* - * No need for locking here, the inode is not live anymore - * and just about to be freed. - */ - if (ip->i_acl != XFS_ACL_NOT_CACHED) - posix_acl_release(ip->i_acl); - if (ip->i_default_acl != XFS_ACL_NOT_CACHED) - posix_acl_release(ip->i_default_acl); -} - - /* * System xattr handlers. * diff --git a/fs/xfs/xfs_acl.h b/fs/xfs/xfs_acl.h index 63dc1f2efad5..947b150df8ed 100644 --- a/fs/xfs/xfs_acl.h +++ b/fs/xfs/xfs_acl.h @@ -46,8 +46,6 @@ extern int xfs_check_acl(struct inode *inode, int mask); extern struct posix_acl *xfs_get_acl(struct inode *inode, int type); extern int xfs_inherit_acl(struct inode *inode, struct posix_acl *default_acl); extern int xfs_acl_chmod(struct inode *inode); -extern void xfs_inode_init_acls(struct xfs_inode *ip); -extern void xfs_inode_clear_acls(struct xfs_inode *ip); extern int posix_acl_access_exists(struct inode *inode); extern int posix_acl_default_exists(struct inode *inode); @@ -57,8 +55,6 @@ extern struct xattr_handler xfs_xattr_system_handler; # define xfs_get_acl(inode, type) NULL # define xfs_inherit_acl(inode, default_acl) 0 # define xfs_acl_chmod(inode) 0 -# define xfs_inode_init_acls(ip) -# define xfs_inode_clear_acls(ip) # define posix_acl_access_exists(inode) 0 # define posix_acl_default_exists(inode) 0 #endif /* CONFIG_XFS_POSIX_ACL */ diff --git a/fs/xfs/xfs_iget.c b/fs/xfs/xfs_iget.c index 76c540f719e4..5fcec6f020a7 100644 --- a/fs/xfs/xfs_iget.c +++ b/fs/xfs/xfs_iget.c @@ -83,7 +83,6 @@ xfs_inode_alloc( memset(&ip->i_d, 0, sizeof(xfs_icdinode_t)); ip->i_size = 0; ip->i_new_size = 0; - xfs_inode_init_acls(ip); /* * Initialize inode's trace buffers. @@ -560,7 +559,6 @@ xfs_ireclaim( ASSERT(atomic_read(&ip->i_pincount) == 0); ASSERT(!spin_is_locked(&ip->i_flags_lock)); ASSERT(completion_done(&ip->i_flush)); - xfs_inode_clear_acls(ip); kmem_zone_free(xfs_inode_zone, ip); } diff --git a/fs/xfs/xfs_inode.h b/fs/xfs/xfs_inode.h index 77016702938b..1804f866a71d 100644 --- a/fs/xfs/xfs_inode.h +++ b/fs/xfs/xfs_inode.h @@ -273,11 +273,6 @@ typedef struct xfs_inode { /* VFS inode */ struct inode i_vnode; /* embedded VFS inode */ -#ifdef CONFIG_XFS_POSIX_ACL - struct posix_acl *i_acl; - struct posix_acl *i_default_acl; -#endif - /* Trace buffers per inode. */ #ifdef XFS_INODE_TRACE struct ktrace *i_trace; /* general inode trace */ -- cgit v1.2.3 From 72c04902d1e27c8a324014cff1d4475c11b1cecd Mon Sep 17 00:00:00 2001 From: Al Viro Date: Wed, 24 Jun 2009 16:58:48 -0400 Subject: Get "no acls for this inode" right, fix shmem breakage Signed-off-by: Al Viro --- fs/btrfs/inode.c | 6 ++---- fs/jffs2/acl.c | 3 +-- include/linux/posix_acl.h | 9 +++++++++ mm/shmem.c | 5 +---- 4 files changed, 13 insertions(+), 10 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 78ad38ddd01f..dbe1aabf96cd 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -2122,10 +2122,8 @@ static void btrfs_read_locked_inode(struct inode *inode) * any xattrs or acls */ maybe_acls = acls_after_inode_item(leaf, path->slots[0], inode->i_ino); - if (!maybe_acls) { - inode->i_acl = NULL; - inode->i_default_acl = NULL; - } + if (!maybe_acls) + cache_no_acl(inode); BTRFS_I(inode)->block_group = btrfs_find_block_group(root, 0, alloc_group_block, 0); diff --git a/fs/jffs2/acl.c b/fs/jffs2/acl.c index edd2ad6416d8..8fcb6239218e 100644 --- a/fs/jffs2/acl.c +++ b/fs/jffs2/acl.c @@ -284,8 +284,7 @@ int jffs2_init_acl_pre(struct inode *dir_i, struct inode *inode, int *i_mode) struct posix_acl *acl, *clone; int rc; - inode->i_default_acl = NULL; - inode->i_acl = NULL; + cache_no_acl(inode); if (S_ISLNK(*i_mode)) return 0; /* Symlink always has no-ACL */ diff --git a/include/linux/posix_acl.h b/include/linux/posix_acl.h index c513466c7dc7..065a3652a3ea 100644 --- a/include/linux/posix_acl.h +++ b/include/linux/posix_acl.h @@ -148,4 +148,13 @@ static inline void forget_cached_acl(struct inode *inode, int type) posix_acl_release(old); } #endif + +static inline void cache_no_acl(struct inode *inode) +{ +#ifdef CONFIG_FS_POSIX_ACL + inode->i_acl = NULL; + inode->i_default_acl = NULL; +#endif +} + #endif /* __LINUX_POSIX_ACL_H */ diff --git a/mm/shmem.c b/mm/shmem.c index 5f2019fc7895..d713239ce2ce 100644 --- a/mm/shmem.c +++ b/mm/shmem.c @@ -1558,6 +1558,7 @@ static struct inode *shmem_get_inode(struct super_block *sb, int mode, spin_lock_init(&info->lock); info->flags = flags & VM_NORESERVE; INIT_LIST_HEAD(&info->swaplist); + cache_no_acl(inode); switch (mode & S_IFMT) { default: @@ -2379,10 +2380,6 @@ static struct inode *shmem_alloc_inode(struct super_block *sb) p = (struct shmem_inode_info *)kmem_cache_alloc(shmem_inode_cachep, GFP_KERNEL); if (!p) return NULL; -#ifdef CONFIG_TMPFS_POSIX_ACL - p->vfs_inode.i_acl = NULL; - p->vfs_inode.i_default_acl = NULL; -#endif return &p->vfs_inode; } -- cgit v1.2.3 From d5bb68adda7cc179e8efadeaa3a283cb470f13a6 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Wed, 24 Jun 2009 17:02:42 -0400 Subject: another race fix in jfs_check_acl() Signed-off-by: Al Viro --- fs/jfs/acl.c | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/jfs/acl.c b/fs/jfs/acl.c index f272bf032e1e..91fa3ad6e8c2 100644 --- a/fs/jfs/acl.c +++ b/fs/jfs/acl.c @@ -118,15 +118,16 @@ out: static int jfs_check_acl(struct inode *inode, int mask) { - if (inode->i_acl == ACL_NOT_CACHED) { - struct posix_acl *acl = jfs_get_acl(inode, ACL_TYPE_ACCESS); - if (IS_ERR(acl)) - return PTR_ERR(acl); + struct posix_acl *acl = jfs_get_acl(inode, ACL_TYPE_ACCESS); + + if (IS_ERR(acl)) + return PTR_ERR(acl); + if (acl) { + int error = posix_acl_permission(inode, acl, mask); posix_acl_release(acl); + return error; } - if (inode->i_acl) - return posix_acl_permission(inode, inode->i_acl, mask); return -EAGAIN; } -- cgit v1.2.3 From 268875b9d1dd1bf0b523c59e736da9bc20c8ce1f Mon Sep 17 00:00:00 2001 From: Steve French Date: Thu, 25 Jun 2009 00:29:21 +0000 Subject: [CIFS] Do not send tree disconnect if session is already disconnected Noticed this when tree connect timed out (due to Samba server crash) - we try to send a tree disconnect for a tid that does not exist since we don't have a valid tree id yet. This checks that the session is valid before sending the tree disconnect to handle this case. Signed-off-by: Steve French --- fs/cifs/CHANGES | 2 +- fs/cifs/cifssmb.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/cifs/CHANGES b/fs/cifs/CHANGES index b48689839428..3a9b7a58a51d 100644 --- a/fs/cifs/CHANGES +++ b/fs/cifs/CHANGES @@ -5,7 +5,7 @@ client generated ones by default (mount option "serverino" turned on by default if server supports it). Add forceuid and forcegid mount options (so that when negotiating unix extensions specifying which uid mounted does not immediately force the server's reported -uids to be overridden). +uids to be overridden). Add support for scope moutn parm. Version 1.58 ------------ diff --git a/fs/cifs/cifssmb.c b/fs/cifs/cifssmb.c index b84c61d5bca4..58d65a9a79c3 100644 --- a/fs/cifs/cifssmb.c +++ b/fs/cifs/cifssmb.c @@ -729,7 +729,7 @@ CIFSSMBTDis(const int xid, struct cifsTconInfo *tcon) * the tcon is no longer on the list, so no need to take lock before * checking this. */ - if (tcon->need_reconnect) + if ((tcon->need_reconnect) || (tcon->ses->need_reconnect)) return 0; rc = small_smb_init(SMB_COM_TREE_DISCONNECT, 0, tcon, -- cgit v1.2.3 From 681bf72e4893a187cf6b6b62c08fc193f81c8c2f Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Thu, 11 Jun 2009 10:27:31 -0400 Subject: cifs: have cifs parse scope_id out of IPv6 addresses and use it This patch has CIFS look for a '%' in an IPv6 address. If one is present then it will try to treat that value as a numeric interface index suitable for stuffing into the sin6_scope_id field. This should allow people to mount servers on IPv6 link-local addresses. Signed-off-by: Jeff Layton Acked-by: David Holder Signed-off-by: Steve French --- fs/cifs/connect.c | 6 ++++-- fs/cifs/dns_resolve.c | 4 ++-- fs/cifs/netmisc.c | 32 +++++++++++++++++++++++++++----- 3 files changed, 33 insertions(+), 9 deletions(-) (limited to 'fs') diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index c368ad658236..3fb799ff55c9 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -1386,8 +1386,10 @@ cifs_find_tcp_session(struct sockaddr_storage *addr) server->addr.sockAddr.sin_addr.s_addr)) continue; else if (addr->ss_family == AF_INET6 && - !ipv6_addr_equal(&server->addr.sockAddr6.sin6_addr, - &addr6->sin6_addr)) + (!ipv6_addr_equal(&server->addr.sockAddr6.sin6_addr, + &addr6->sin6_addr) || + server->addr.sockAddr6.sin6_scope_id != + addr6->sin6_scope_id)) continue; ++server->srv_count; diff --git a/fs/cifs/dns_resolve.c b/fs/cifs/dns_resolve.c index 91b5500755bf..87948147d7ec 100644 --- a/fs/cifs/dns_resolve.c +++ b/fs/cifs/dns_resolve.c @@ -35,7 +35,7 @@ * 0 - name is not IP */ static int -is_ip(const char *name) +is_ip(char *name) { struct sockaddr_storage ss; @@ -57,7 +57,7 @@ dns_resolver_instantiate(struct key *key, const void *data, ip[datalen] = '\0'; /* make sure this looks like an address */ - if (!is_ip((const char *) ip)) { + if (!is_ip(ip)) { kfree(ip); return -EINVAL; } diff --git a/fs/cifs/netmisc.c b/fs/cifs/netmisc.c index f9a54da97d34..bd6d6895730d 100644 --- a/fs/cifs/netmisc.c +++ b/fs/cifs/netmisc.c @@ -158,25 +158,47 @@ cifs_inet_pton(const int address_family, const char *cp, void *dst) /* * Try to convert a string to an IPv4 address and then attempt to convert * it to an IPv6 address if that fails. Set the family field if either - * succeeds. + * succeeds. If it's an IPv6 address and it has a '%' sign in it, try to + * treat the part following it as a numeric sin6_scope_id. * * Returns 0 on failure. */ int cifs_convert_address(char *src, void *dst) { + int rc; + char *pct, *endp; struct sockaddr_in *s4 = (struct sockaddr_in *) dst; struct sockaddr_in6 *s6 = (struct sockaddr_in6 *) dst; + /* IPv4 address */ if (cifs_inet_pton(AF_INET, src, &s4->sin_addr.s_addr)) { s4->sin_family = AF_INET; return 1; - } else if (cifs_inet_pton(AF_INET6, src, &s6->sin6_addr.s6_addr)) { - s6->sin6_family = AF_INET6; - return 1; } - return 0; + /* temporarily terminate string */ + pct = strchr(src, '%'); + if (pct) + *pct = '\0'; + + rc = cifs_inet_pton(AF_INET6, src, &s6->sin6_addr.s6_addr); + + /* repair temp termination (if any) and make pct point to scopeid */ + if (pct) + *pct++ = '%'; + + if (!rc) + return rc; + + s6->sin6_family = AF_INET6; + if (pct) { + s6->sin6_scope_id = (u32) simple_strtoul(pct, &endp, 0); + if (!*pct || *endp) + return 0; + } + + return rc; } /***************************************************************************** -- cgit v1.2.3 From b48a485884b5afb3e33b1871bcbd246b67491923 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Thu, 25 Jun 2009 00:56:54 -0400 Subject: cifs: fix problems with earlier patches cifs: fix problems with earlier patches cifs_show_address hasn't been introduced yet, and fix a typo that was silently fixed by a later patch in the series. Signed-off-by: Jeff Layton Signed-off-by: Steve French --- fs/cifs/cifsfs.c | 2 -- 1 file changed, 2 deletions(-) (limited to 'fs') diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c index ddef913ff3e6..b5e9f398c2e5 100644 --- a/fs/cifs/cifsfs.c +++ b/fs/cifs/cifsfs.c @@ -374,8 +374,6 @@ cifs_show_options(struct seq_file *s, struct vfsmount *m) if (tcon->ses->domainName) seq_printf(s, ",domain=%s", tcon->ses->domainName); - cifs_show_address(s, tcon->ses->server); - seq_printf(s, ",uid=%d", cifs_sb->mnt_uid); if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_OVERR_UID) seq_printf(s, ",forceuid"); -- cgit v1.2.3 From 6459340cfcc6f6d165b27c3dd955aeb55a1b73d3 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Thu, 25 Jun 2009 00:56:55 -0400 Subject: cifs: remove rw/ro options cifs: remove rw/ro options These options are handled at the VFS layer. They only ever set the option in the smb_vol struct. Nothing was ever done with them afterward anyway. Signed-off-by: Jeff Layton Signed-off-by: Steve French --- fs/cifs/connect.c | 6 ------ 1 file changed, 6 deletions(-) (limited to 'fs') diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index 3fb799ff55c9..a581cfa2ba82 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -70,7 +70,6 @@ struct smb_vol { mode_t file_mode; mode_t dir_mode; unsigned secFlg; - bool rw:1; bool retry:1; bool intr:1; bool setuids:1; @@ -832,7 +831,6 @@ cifs_parse_mount_options(char *options, const char *devname, vol->dir_mode = vol->file_mode = S_IRUGO | S_IXUGO | S_IWUSR; /* vol->retry default is 0 (i.e. "soft" limited retry not hard retry) */ - vol->rw = true; /* default is always to request posix paths. */ vol->posix_paths = 1; /* default to using server inode numbers where available */ @@ -1198,8 +1196,6 @@ cifs_parse_mount_options(char *options, const char *devname, /* ignore */ } else if (strnicmp(data, "guest", 5) == 0) { /* ignore */ - } else if (strnicmp(data, "rw", 2) == 0) { - vol->rw = true; } else if (strnicmp(data, "noblocksend", 11) == 0) { vol->noblocksnd = 1; } else if (strnicmp(data, "noautotune", 10) == 0) { @@ -1218,8 +1214,6 @@ cifs_parse_mount_options(char *options, const char *devname, parse these options again and set anything and it is ok to just ignore them */ continue; - } else if (strnicmp(data, "ro", 2) == 0) { - vol->rw = false; } else if (strnicmp(data, "hard", 4) == 0) { vol->retry = 1; } else if (strnicmp(data, "soft", 4) == 0) { -- cgit v1.2.3 From 6debdbc0ba6253ac519cd5a3d22e30f1f9f1dd12 Mon Sep 17 00:00:00 2001 From: Simo Leone Date: Thu, 25 Jun 2009 02:44:43 +0000 Subject: [CIFS] Copy struct *after* setting the port, instead of before. Acked-by: Jeff Layton Signed-off-by: Simo Leone Signed-off-by: Steve French --- fs/cifs/connect.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index a581cfa2ba82..12c2cf693555 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -1496,14 +1496,14 @@ cifs_get_tcp_session(struct smb_vol *volume_info) cFYI(1, ("attempting ipv6 connect")); /* BB should we allow ipv6 on port 139? */ /* other OS never observed in Wild doing 139 with v6 */ + sin_server6->sin6_port = htons(volume_info->port); memcpy(&tcp_ses->addr.sockAddr6, sin_server6, sizeof(struct sockaddr_in6)); - sin_server6->sin6_port = htons(volume_info->port); rc = ipv6_connect(tcp_ses); } else { + sin_server->sin_port = htons(volume_info->port); memcpy(&tcp_ses->addr.sockAddr, sin_server, sizeof(struct sockaddr_in)); - sin_server->sin_port = htons(volume_info->port); rc = ipv4_connect(tcp_ses); } if (rc < 0) { -- cgit v1.2.3 From f46c7234e472ceee39afea4fb5a4365843e1850a Mon Sep 17 00:00:00 2001 From: Steve French Date: Thu, 25 Jun 2009 03:04:20 +0000 Subject: [CIFS] cleanup asn handling for ntlmssp Also removes obsolete distinction between rawntlmssp and ntlmssp (in asn/SPNEGO) since as jra noted we can always send raw ntlmssp in session setup now. remove check for experimental runtime flag (/proc/fs/cifs/Experimental) in ntlmssp path. Reviewed-by: Jeff Layton Signed-off-by: Steve French --- fs/cifs/asn1.c | 55 ++++++++++++++++++++++++++++++++++++++++++++++++------ fs/cifs/cifsglob.h | 2 +- fs/cifs/cifssmb.c | 2 +- fs/cifs/sess.c | 2 +- 4 files changed, 52 insertions(+), 9 deletions(-) (limited to 'fs') diff --git a/fs/cifs/asn1.c b/fs/cifs/asn1.c index 1b09f1670061..20692fbfdb24 100644 --- a/fs/cifs/asn1.c +++ b/fs/cifs/asn1.c @@ -49,6 +49,7 @@ #define ASN1_OJI 6 /* Object Identifier */ #define ASN1_OJD 7 /* Object Description */ #define ASN1_EXT 8 /* External */ +#define ASN1_ENUM 10 /* Enumerated */ #define ASN1_SEQ 16 /* Sequence */ #define ASN1_SET 17 /* Set */ #define ASN1_NUMSTR 18 /* Numerical String */ @@ -78,10 +79,12 @@ #define SPNEGO_OID_LEN 7 #define NTLMSSP_OID_LEN 10 #define KRB5_OID_LEN 7 +#define KRB5U2U_OID_LEN 8 #define MSKRB5_OID_LEN 7 static unsigned long SPNEGO_OID[7] = { 1, 3, 6, 1, 5, 5, 2 }; static unsigned long NTLMSSP_OID[10] = { 1, 3, 6, 1, 4, 1, 311, 2, 2, 10 }; static unsigned long KRB5_OID[7] = { 1, 2, 840, 113554, 1, 2, 2 }; +static unsigned long KRB5U2U_OID[8] = { 1, 2, 840, 113554, 1, 2, 2, 3 }; static unsigned long MSKRB5_OID[7] = { 1, 2, 840, 48018, 1, 2, 2 }; /* @@ -122,6 +125,28 @@ asn1_octet_decode(struct asn1_ctx *ctx, unsigned char *ch) return 1; } +#if 0 /* will be needed later by spnego decoding/encoding of ntlmssp */ +static unsigned char +asn1_enum_decode(struct asn1_ctx *ctx, __le32 *val) +{ + unsigned char ch; + + if (ctx->pointer >= ctx->end) { + ctx->error = ASN1_ERR_DEC_EMPTY; + return 0; + } + + ch = *(ctx->pointer)++; /* ch has 0xa, ptr points to lenght octet */ + if ((ch) == ASN1_ENUM) /* if ch value is ENUM, 0xa */ + *val = *(++(ctx->pointer)); /* value has enum value */ + else + return 0; + + ctx->pointer++; + return 1; +} +#endif + static unsigned char asn1_tag_decode(struct asn1_ctx *ctx, unsigned int *tag) { @@ -476,10 +501,9 @@ decode_negTokenInit(unsigned char *security_blob, int length, unsigned int cls, con, tag, oidlen, rc; bool use_ntlmssp = false; bool use_kerberos = false; + bool use_kerberosu2u = false; bool use_mskerberos = false; - *secType = NTLM; /* BB eventually make Kerberos or NLTMSSP the default*/ - /* cifs_dump_mem(" Received SecBlob ", security_blob, length); */ asn1_open(&ctx, security_blob, length); @@ -515,6 +539,7 @@ decode_negTokenInit(unsigned char *security_blob, int length, return 0; } + /* SPNEGO */ if (asn1_header_decode(&ctx, &end, &cls, &con, &tag) == 0) { cFYI(1, ("Error decoding negTokenInit")); return 0; @@ -526,6 +551,7 @@ decode_negTokenInit(unsigned char *security_blob, int length, return 0; } + /* negTokenInit */ if (asn1_header_decode(&ctx, &end, &cls, &con, &tag) == 0) { cFYI(1, ("Error decoding negTokenInit")); return 0; @@ -537,6 +563,7 @@ decode_negTokenInit(unsigned char *security_blob, int length, return 0; } + /* sequence */ if (asn1_header_decode(&ctx, &end, &cls, &con, &tag) == 0) { cFYI(1, ("Error decoding 2nd part of negTokenInit")); return 0; @@ -548,6 +575,7 @@ decode_negTokenInit(unsigned char *security_blob, int length, return 0; } + /* sequence of */ if (asn1_header_decode (&ctx, &sequence_end, &cls, &con, &tag) == 0) { cFYI(1, ("Error decoding 2nd part of negTokenInit")); @@ -560,6 +588,7 @@ decode_negTokenInit(unsigned char *security_blob, int length, return 0; } + /* list of security mechanisms */ while (!asn1_eoc_decode(&ctx, sequence_end)) { rc = asn1_header_decode(&ctx, &end, &cls, &con, &tag); if (!rc) { @@ -576,11 +605,15 @@ decode_negTokenInit(unsigned char *security_blob, int length, if (compare_oid(oid, oidlen, MSKRB5_OID, MSKRB5_OID_LEN) && - !use_kerberos) + !use_mskerberos) use_mskerberos = true; + else if (compare_oid(oid, oidlen, KRB5U2U_OID, + KRB5U2U_OID_LEN) && + !use_kerberosu2u) + use_kerberosu2u = true; else if (compare_oid(oid, oidlen, KRB5_OID, KRB5_OID_LEN) && - !use_mskerberos) + !use_kerberos) use_kerberos = true; else if (compare_oid(oid, oidlen, NTLMSSP_OID, NTLMSSP_OID_LEN)) @@ -593,7 +626,12 @@ decode_negTokenInit(unsigned char *security_blob, int length, } } + /* mechlistMIC */ if (asn1_header_decode(&ctx, &end, &cls, &con, &tag) == 0) { + /* Check if we have reached the end of the blob, but with + no mechListMic (e.g. NTLMSSP instead of KRB5) */ + if (ctx.error == ASN1_ERR_DEC_EMPTY) + goto decode_negtoken_exit; cFYI(1, ("Error decoding last part negTokenInit exit3")); return 0; } else if ((cls != ASN1_CTX) || (con != ASN1_CON)) { @@ -602,6 +640,8 @@ decode_negTokenInit(unsigned char *security_blob, int length, cls, con, tag, end, *end)); return 0; } + + /* sequence */ if (asn1_header_decode(&ctx, &end, &cls, &con, &tag) == 0) { cFYI(1, ("Error decoding last part negTokenInit exit5")); return 0; @@ -611,6 +651,7 @@ decode_negTokenInit(unsigned char *security_blob, int length, cls, con, tag, end, *end)); } + /* sequence of */ if (asn1_header_decode(&ctx, &end, &cls, &con, &tag) == 0) { cFYI(1, ("Error decoding last part negTokenInit exit 7")); return 0; @@ -619,6 +660,8 @@ decode_negTokenInit(unsigned char *security_blob, int length, cls, con, tag, end, *end)); return 0; } + + /* general string */ if (asn1_header_decode(&ctx, &end, &cls, &con, &tag) == 0) { cFYI(1, ("Error decoding last part negTokenInit exit9")); return 0; @@ -630,13 +673,13 @@ decode_negTokenInit(unsigned char *security_blob, int length, } cFYI(1, ("Need to call asn1_octets_decode() function for %s", ctx.pointer)); /* is this UTF-8 or ASCII? */ - +decode_negtoken_exit: if (use_kerberos) *secType = Kerberos; else if (use_mskerberos) *secType = MSKerberos; else if (use_ntlmssp) - *secType = NTLMSSP; + *secType = RawNTLMSSP; return 1; } diff --git a/fs/cifs/cifsglob.h b/fs/cifs/cifsglob.h index a61ab772c6f6..e1225e6ded2f 100644 --- a/fs/cifs/cifsglob.h +++ b/fs/cifs/cifsglob.h @@ -83,7 +83,7 @@ enum securityEnum { NTLM, /* Legacy NTLM012 auth with NTLM hash */ NTLMv2, /* Legacy NTLM auth with NTLMv2 hash */ RawNTLMSSP, /* NTLMSSP without SPNEGO, NTLMv2 hash */ - NTLMSSP, /* NTLMSSP via SPNEGO, NTLMv2 hash */ +/* NTLMSSP, */ /* can use rawNTLMSSP instead of NTLMSSP via SPNEGO */ Kerberos, /* Kerberos via SPNEGO */ MSKerberos, /* MS Kerberos via SPNEGO */ }; diff --git a/fs/cifs/cifssmb.c b/fs/cifs/cifssmb.c index 58d65a9a79c3..61007c627497 100644 --- a/fs/cifs/cifssmb.c +++ b/fs/cifs/cifssmb.c @@ -594,7 +594,7 @@ CIFSSMBNegotiate(unsigned int xid, struct cifsSesInfo *ses) else if (secFlags & CIFSSEC_MAY_KRB5) server->secType = Kerberos; else if (secFlags & CIFSSEC_MAY_NTLMSSP) - server->secType = NTLMSSP; + server->secType = RawNTLMSSP; else if (secFlags & CIFSSEC_MAY_LANMAN) server->secType = LANMAN; /* #ifdef CONFIG_CIFS_EXPERIMENTAL diff --git a/fs/cifs/sess.c b/fs/cifs/sess.c index 897a052270f9..7085a6275c4c 100644 --- a/fs/cifs/sess.c +++ b/fs/cifs/sess.c @@ -802,7 +802,7 @@ ssetup_ntlmssp_authenticate: #endif /* CONFIG_CIFS_UPCALL */ } else { #ifdef CONFIG_CIFS_EXPERIMENTAL - if ((experimEnabled > 1) && (type == RawNTLMSSP)) { + if (type == RawNTLMSSP) { if ((pSMB->req.hdr.Flags2 & SMBFLG2_UNICODE) == 0) { cERROR(1, ("NTLMSSP requires Unicode support")); rc = -ENOSYS; -- cgit v1.2.3 From 0f3bc09ee1b7fcadd5bfdc5ed2e1643f658fe23d Mon Sep 17 00:00:00 2001 From: Suresh Jayaraman Date: Thu, 25 Jun 2009 18:12:34 +0530 Subject: cifs: Fix incorrect return code being printed in cFYI messages FreeXid() along with freeing Xid does add a cifsFYI debug message that prints rc (return code) as well. In some code paths where we set/return error code after calling FreeXid(), incorrect error code is being printed when cifsFYI is enabled. This could be misleading in few cases. For eg. In cifs_open() if cifs_fill_filedata() returns a valid pointer to cifsFileInfo, FreeXid() prints rc=-13 whereas 0 is actually being returned. Fix this by setting rc before calling FreeXid(). Basically convert FreeXid(xid); rc = -ERR; return -ERR; => FreeXid(xid); return rc; [Note that Christoph would like to replace the GetXid/FreeXid calls, which are primarily used for debugging. This seems like a good longer term goal, but although there is an alternative tracing facility, there are no examples yet available that I know of that we can use (yet) to convert this cifs function entry/exit logging, and for creating an identifier that we can use to correlate all dmesg log entries for a particular vfs operation (ie identify all log entries for a particular vfs request to cifs: e.g. a particular close or read or write or byte range lock call ... and just using the thread id is harder). Eventually when a replacement for this is available (e.g. when NFS switches over and various samples to look at in other file systems) we can remove the GetXid/FreeXid macro but in the meantime multiple people use this run time configurable logging all the time for debugging, and Suresh's patch fixes a problem which made it harder to notice some low memory problems in the log so it is worthwhile to fix this problem until a better logging approach is able to be used] Acked-by: Jeff Layton Signed-off-by: Suresh Jayaraman Signed-off-by: Steve French --- fs/cifs/dir.c | 6 ++++-- fs/cifs/file.c | 24 ++++++++++++++++-------- fs/cifs/inode.c | 15 ++++++++++----- fs/cifs/link.c | 3 ++- fs/cifs/xattr.c | 12 ++++++++---- 5 files changed, 40 insertions(+), 20 deletions(-) (limited to 'fs') diff --git a/fs/cifs/dir.c b/fs/cifs/dir.c index 3758965d73d5..7dc6b74f9def 100644 --- a/fs/cifs/dir.c +++ b/fs/cifs/dir.c @@ -307,8 +307,9 @@ cifs_create(struct inode *inode, struct dentry *direntry, int mode, full_path = build_path_from_dentry(direntry); if (full_path == NULL) { + rc = -ENOMEM; FreeXid(xid); - return -ENOMEM; + return rc; } if (oplockEnabled) @@ -540,8 +541,9 @@ int cifs_mknod(struct inode *inode, struct dentry *direntry, int mode, buf = kmalloc(sizeof(FILE_ALL_INFO), GFP_KERNEL); if (buf == NULL) { kfree(full_path); + rc = -ENOMEM; FreeXid(xid); - return -ENOMEM; + return rc; } rc = CIFSSMBOpen(xid, pTcon, full_path, diff --git a/fs/cifs/file.c b/fs/cifs/file.c index 06866841b97f..ebdbe62a829c 100644 --- a/fs/cifs/file.c +++ b/fs/cifs/file.c @@ -300,14 +300,16 @@ int cifs_open(struct inode *inode, struct file *file) pCifsInode = CIFS_I(file->f_path.dentry->d_inode); pCifsFile = cifs_fill_filedata(file); if (pCifsFile) { + rc = 0; FreeXid(xid); - return 0; + return rc; } full_path = build_path_from_dentry(file->f_path.dentry); if (full_path == NULL) { + rc = -ENOMEM; FreeXid(xid); - return -ENOMEM; + return rc; } cFYI(1, ("inode = 0x%p file flags are 0x%x for %s", @@ -494,8 +496,9 @@ static int cifs_reopen_file(struct file *file, bool can_flush) mutex_unlock(&pCifsFile->fh_mutex); if (!pCifsFile->invalidHandle) { mutex_lock(&pCifsFile->fh_mutex); + rc = 0; FreeXid(xid); - return 0; + return rc; } if (file->f_path.dentry == NULL) { @@ -845,8 +848,9 @@ int cifs_lock(struct file *file, int cmd, struct file_lock *pfLock) tcon = cifs_sb->tcon; if (file->private_data == NULL) { + rc = -EBADF; FreeXid(xid); - return -EBADF; + return rc; } netfid = ((struct cifsFileInfo *)file->private_data)->netfid; @@ -1805,8 +1809,9 @@ ssize_t cifs_user_read(struct file *file, char __user *read_data, pTcon = cifs_sb->tcon; if (file->private_data == NULL) { + rc = -EBADF; FreeXid(xid); - return -EBADF; + return rc; } open_file = (struct cifsFileInfo *)file->private_data; @@ -1885,8 +1890,9 @@ static ssize_t cifs_read(struct file *file, char *read_data, size_t read_size, pTcon = cifs_sb->tcon; if (file->private_data == NULL) { + rc = -EBADF; FreeXid(xid); - return -EBADF; + return rc; } open_file = (struct cifsFileInfo *)file->private_data; @@ -2019,8 +2025,9 @@ static int cifs_readpages(struct file *file, struct address_space *mapping, xid = GetXid(); if (file->private_data == NULL) { + rc = -EBADF; FreeXid(xid); - return -EBADF; + return rc; } open_file = (struct cifsFileInfo *)file->private_data; cifs_sb = CIFS_SB(file->f_path.dentry->d_sb); @@ -2185,8 +2192,9 @@ static int cifs_readpage(struct file *file, struct page *page) xid = GetXid(); if (file->private_data == NULL) { + rc = -EBADF; FreeXid(xid); - return -EBADF; + return rc; } cFYI(1, ("readpage %p at offset %d 0x%x\n", diff --git a/fs/cifs/inode.c b/fs/cifs/inode.c index fad882b075ba..155c9e785d0c 100644 --- a/fs/cifs/inode.c +++ b/fs/cifs/inode.c @@ -988,8 +988,9 @@ int cifs_unlink(struct inode *dir, struct dentry *dentry) * sb->s_vfs_rename_mutex here */ full_path = build_path_from_dentry(dentry); if (full_path == NULL) { + rc = -ENOMEM; FreeXid(xid); - return -ENOMEM; + return rc; } if ((tcon->ses->capabilities & CAP_UNIX) && @@ -1118,8 +1119,9 @@ int cifs_mkdir(struct inode *inode, struct dentry *direntry, int mode) full_path = build_path_from_dentry(direntry); if (full_path == NULL) { + rc = -ENOMEM; FreeXid(xid); - return -ENOMEM; + return rc; } if ((pTcon->ses->capabilities & CAP_UNIX) && @@ -1303,8 +1305,9 @@ int cifs_rmdir(struct inode *inode, struct dentry *direntry) full_path = build_path_from_dentry(direntry); if (full_path == NULL) { + rc = -ENOMEM; FreeXid(xid); - return -ENOMEM; + return rc; } rc = CIFSSMBRmDir(xid, pTcon, full_path, cifs_sb->local_nls, @@ -1508,8 +1511,9 @@ int cifs_revalidate(struct dentry *direntry) since that would deadlock */ full_path = build_path_from_dentry(direntry); if (full_path == NULL) { + rc = -ENOMEM; FreeXid(xid); - return -ENOMEM; + return rc; } cFYI(1, ("Revalidate: %s inode 0x%p count %d dentry: 0x%p d_time %ld " "jiffies %ld", full_path, direntry->d_inode, @@ -1911,8 +1915,9 @@ cifs_setattr_nounix(struct dentry *direntry, struct iattr *attrs) full_path = build_path_from_dentry(direntry); if (full_path == NULL) { + rc = -ENOMEM; FreeXid(xid); - return -ENOMEM; + return rc; } /* diff --git a/fs/cifs/link.c b/fs/cifs/link.c index cd83c53fcbb5..fc1e0487eaee 100644 --- a/fs/cifs/link.c +++ b/fs/cifs/link.c @@ -172,8 +172,9 @@ cifs_symlink(struct inode *inode, struct dentry *direntry, const char *symname) full_path = build_path_from_dentry(direntry); if (full_path == NULL) { + rc = -ENOMEM; FreeXid(xid); - return -ENOMEM; + return rc; } cFYI(1, ("Full path: %s", full_path)); diff --git a/fs/cifs/xattr.c b/fs/cifs/xattr.c index e9527eedc639..a75afa3dd9e1 100644 --- a/fs/cifs/xattr.c +++ b/fs/cifs/xattr.c @@ -64,8 +64,9 @@ int cifs_removexattr(struct dentry *direntry, const char *ea_name) full_path = build_path_from_dentry(direntry); if (full_path == NULL) { + rc = -ENOMEM; FreeXid(xid); - return -ENOMEM; + return rc; } if (ea_name == NULL) { cFYI(1, ("Null xattr names not supported")); @@ -118,8 +119,9 @@ int cifs_setxattr(struct dentry *direntry, const char *ea_name, full_path = build_path_from_dentry(direntry); if (full_path == NULL) { + rc = -ENOMEM; FreeXid(xid); - return -ENOMEM; + return rc; } /* return dos attributes as pseudo xattr */ /* return alt name if available as pseudo attr */ @@ -225,8 +227,9 @@ ssize_t cifs_getxattr(struct dentry *direntry, const char *ea_name, full_path = build_path_from_dentry(direntry); if (full_path == NULL) { + rc = -ENOMEM; FreeXid(xid); - return -ENOMEM; + return rc; } /* return dos attributes as pseudo xattr */ /* return alt name if available as pseudo attr */ @@ -351,8 +354,9 @@ ssize_t cifs_listxattr(struct dentry *direntry, char *data, size_t buf_size) full_path = build_path_from_dentry(direntry); if (full_path == NULL) { + rc = -ENOMEM; FreeXid(xid); - return -ENOMEM; + return rc; } /* return dos attributes as pseudo xattr */ /* return alt name if available as pseudo attr */ -- cgit v1.2.3 From ad8034f19792736db5c259103c2eaaf72887bbb4 Mon Sep 17 00:00:00 2001 From: Steve French Date: Fri, 26 Jun 2009 03:25:49 +0000 Subject: [CIFS] remove bkl usage from umount begin The lock_kernel call moved into the fs for umount_begin is not needed. This adds a check to make sure we don't call umount_begin twice on the same fs. umount_begin for cifs is probably not needed and may eventually be able to be removed, but in the meantime this smaller patch is safe and gets rid of the bkl from this path which provides some benefit. Acked-by: Jeff Layton Signed-off-by: Steve French --- fs/cifs/cifsfs.c | 11 +++++++---- 1 file changed, 7 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c index b5e9f398c2e5..9f669f982c4d 100644 --- a/fs/cifs/cifsfs.c +++ b/fs/cifs/cifsfs.c @@ -537,9 +537,14 @@ static void cifs_umount_begin(struct super_block *sb) if (tcon == NULL) return; - lock_kernel(); read_lock(&cifs_tcp_ses_lock); - if (tcon->tc_count == 1) + if ((tcon->tc_count > 1) || (tcon->tidStatus == CifsExiting)) { + /* we have other mounts to same share or we have + already tried to force umount this and woken up + all waiting network requests, nothing to do */ + read_unlock(&cifs_tcp_ses_lock); + return; + } else if (tcon->tc_count == 1) tcon->tidStatus = CifsExiting; read_unlock(&cifs_tcp_ses_lock); @@ -554,9 +559,7 @@ static void cifs_umount_begin(struct super_block *sb) wake_up_all(&tcon->ses->server->response_q); msleep(1); } -/* BB FIXME - finish add checks for tidStatus BB */ - unlock_kernel(); return; } -- cgit v1.2.3 From 71a394faaad07090af5de5c075ec2f5bca0fbb35 Mon Sep 17 00:00:00 2001 From: Steve French Date: Fri, 26 Jun 2009 04:07:18 +0000 Subject: [CIFS] remove unknown mount option warning message Jeff's previous patch which removed the unneeded rw/ro parsing can cause a minor warning in dmesg (about the unknown rw or ro mount option) at mount time. This patch makes cifs ignore them in kernel to remove the warning (they are already handled in the mount helper and VFS). Signed-off-by: Steve French --- fs/cifs/connect.c | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'fs') diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index 12c2cf693555..e16d7592116a 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -1196,6 +1196,10 @@ cifs_parse_mount_options(char *options, const char *devname, /* ignore */ } else if (strnicmp(data, "guest", 5) == 0) { /* ignore */ + } else if (strnicmp(data, "rw", 2) == 0) { + /* ignore */ + } else if (strnicmp(data, "ro", 2) == 0) { + /* ignore */ } else if (strnicmp(data, "noblocksend", 11) == 0) { vol->noblocksnd = 1; } else if (strnicmp(data, "noautotune", 10) == 0) { -- cgit v1.2.3 From f0a71eb820596bd8f6abf64beb4cb181edaa2341 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Sat, 27 Jun 2009 07:04:55 -0400 Subject: cifs: fix fh_mutex locking in cifs_reopen_file Fixes a regression caused by commit a6ce4932fbdbcd8f8e8c6df76812014351c32892 When this lock was converted to a mutex, the locks were turned into unlocks and vice-versa. Signed-off-by: Jeff Layton Acked-by: Shirish Pargaonkar Cc: Stable Tree Signed-off-by: Steve French --- fs/cifs/file.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/cifs/file.c b/fs/cifs/file.c index ebdbe62a829c..97ce4bf89d15 100644 --- a/fs/cifs/file.c +++ b/fs/cifs/file.c @@ -493,9 +493,9 @@ static int cifs_reopen_file(struct file *file, bool can_flush) return -EBADF; xid = GetXid(); - mutex_unlock(&pCifsFile->fh_mutex); + mutex_lock(&pCifsFile->fh_mutex); if (!pCifsFile->invalidHandle) { - mutex_lock(&pCifsFile->fh_mutex); + mutex_unlock(&pCifsFile->fh_mutex); rc = 0; FreeXid(xid); return rc; @@ -527,7 +527,7 @@ static int cifs_reopen_file(struct file *file, bool can_flush) if (full_path == NULL) { rc = -ENOMEM; reopen_error_exit: - mutex_lock(&pCifsFile->fh_mutex); + mutex_unlock(&pCifsFile->fh_mutex); FreeXid(xid); return rc; } @@ -569,14 +569,14 @@ reopen_error_exit: cifs_sb->local_nls, cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR); if (rc) { - mutex_lock(&pCifsFile->fh_mutex); + mutex_unlock(&pCifsFile->fh_mutex); cFYI(1, ("cifs_open returned 0x%x", rc)); cFYI(1, ("oplock: %d", oplock)); } else { reopen_success: pCifsFile->netfid = netfid; pCifsFile->invalidHandle = false; - mutex_lock(&pCifsFile->fh_mutex); + mutex_unlock(&pCifsFile->fh_mutex); pCifsInode = CIFS_I(inode); if (pCifsInode) { if (can_flush) { -- cgit v1.2.3 From 94e5d714f604d4cb4cb13163f01ede278e69258b Mon Sep 17 00:00:00 2001 From: Mimi Zohar Date: Fri, 26 Jun 2009 14:05:27 -0400 Subject: integrity: add ima_counts_put (updated) This patch fixes an imbalance message as reported by J.R. Okajima. The IMA file counters are incremented in ima_path_check. If the actual open fails, such as ETXTBSY, decrement the counters to prevent unnecessary imbalance messages. Reported-by: J.R. Okajima Signed-off-by: Mimi Zohar Signed-off-by: James Morris --- fs/namei.c | 7 +++++++ include/linux/ima.h | 6 ++++++ security/integrity/ima/ima_main.c | 29 ++++++++++++++++++++++++++++- 3 files changed, 41 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/namei.c b/fs/namei.c index 5b961eb71cbf..f3c5b278895a 100644 --- a/fs/namei.c +++ b/fs/namei.c @@ -1761,6 +1761,10 @@ do_last: goto exit; } filp = nameidata_to_filp(&nd, open_flag); + if (IS_ERR(filp)) + ima_counts_put(&nd.path, + acc_mode & (MAY_READ | MAY_WRITE | + MAY_EXEC)); mnt_drop_write(nd.path.mnt); if (nd.root.mnt) path_put(&nd.root); @@ -1817,6 +1821,9 @@ ok: goto exit; } filp = nameidata_to_filp(&nd, open_flag); + if (IS_ERR(filp)) + ima_counts_put(&nd.path, + acc_mode & (MAY_READ | MAY_WRITE | MAY_EXEC)); /* * It is now safe to drop the mnt write * because the filp has had a write taken diff --git a/include/linux/ima.h b/include/linux/ima.h index b1b827d091a9..0e3f2a4c25f6 100644 --- a/include/linux/ima.h +++ b/include/linux/ima.h @@ -24,6 +24,7 @@ extern int ima_path_check(struct path *path, int mask, int update_counts); extern void ima_file_free(struct file *file); extern int ima_file_mmap(struct file *file, unsigned long prot); extern void ima_counts_get(struct file *file); +extern void ima_counts_put(struct path *path, int mask); #else static inline int ima_bprm_check(struct linux_binprm *bprm) @@ -60,5 +61,10 @@ static inline void ima_counts_get(struct file *file) { return; } + +static inline void ima_counts_put(struct path *path, int mask) +{ + return; +} #endif /* CONFIG_IMA_H */ #endif /* _LINUX_IMA_H */ diff --git a/security/integrity/ima/ima_main.c b/security/integrity/ima/ima_main.c index 6f611874d10e..101c512564ec 100644 --- a/security/integrity/ima/ima_main.c +++ b/security/integrity/ima/ima_main.c @@ -238,7 +238,34 @@ out: } /* - * ima_opens_get - increment file counts + * ima_counts_put - decrement file counts + * + * File counts are incremented in ima_path_check. On file open + * error, such as ETXTBSY, decrement the counts to prevent + * unnecessary imbalance messages. + */ +void ima_counts_put(struct path *path, int mask) +{ + struct inode *inode = path->dentry->d_inode; + struct ima_iint_cache *iint; + + if (!ima_initialized || !S_ISREG(inode->i_mode)) + return; + iint = ima_iint_find_insert_get(inode); + if (!iint) + return; + + mutex_lock(&iint->mutex); + iint->opencount--; + if ((mask & MAY_WRITE) || (mask == 0)) + iint->writecount--; + else if (mask & (MAY_READ | MAY_EXEC)) + iint->readcount--; + mutex_unlock(&iint->mutex); +} + +/* + * ima_counts_get - increment file counts * * - for IPC shm and shmat file. * - for nfsd exported files. -- cgit v1.2.3 From b4c458b3a23d76936e76678f2074b1528f129f7a Mon Sep 17 00:00:00 2001 From: Csaba Henk Date: Mon, 29 Jun 2009 03:26:53 +0200 Subject: fuse: fix return value of fuse_dev_write() On 64 bit systems -- where sizeof(ssize_t) > sizeof(int) -- the following test exposes a bug due to a non-careful return of an int or unsigned value: implement a FUSE filesystem which sends an unsolicited notification to the kernel with invalid opcode. The respective write to /dev/fuse will return (1 << 32) - EINVAL with errno == 0 instead of -1 with errno == EINVAL. Signed-off-by: Miklos Szeredi CC: stable@kernel.org --- fs/fuse/dev.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c index 8fed2ed12f38..8a11a8c67c42 100644 --- a/fs/fuse/dev.c +++ b/fs/fuse/dev.c @@ -910,7 +910,7 @@ static ssize_t fuse_dev_write(struct kiocb *iocb, const struct iovec *iov, unsigned long nr_segs, loff_t pos) { int err; - unsigned nbytes = iov_length(iov, nr_segs); + size_t nbytes = iov_length(iov, nr_segs); struct fuse_req *req; struct fuse_out_header oh; struct fuse_copy_state cs; -- cgit v1.2.3 From 201fa69a2849536ef2912e8e971ec0b01c04eff4 Mon Sep 17 00:00:00 2001 From: Miklos Szeredi Date: Tue, 30 Jun 2009 20:06:24 +0200 Subject: fuse: fix bad return value in fuse_file_poll() Fix fuse_file_poll() which returned a -errno value instead of a poll mask. Signed-off-by: Miklos Szeredi CC: stable@kernel.org --- fs/fuse/file.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/fuse/file.c b/fs/fuse/file.c index fce6ce694fde..cbc464043b6f 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -1922,7 +1922,7 @@ unsigned fuse_file_poll(struct file *file, poll_table *wait) req = fuse_get_req(fc); if (IS_ERR(req)) - return PTR_ERR(req); + return POLLERR; req->in.h.opcode = FUSE_POLL; req->in.h.nodeid = ff->nodeid; -- cgit v1.2.3 From e0a43ddcc08c34dbd666d93600fd23914505f4aa Mon Sep 17 00:00:00 2001 From: Miklos Szeredi Date: Tue, 30 Jun 2009 20:12:23 +0200 Subject: fuse: allow umask processing in userspace MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This patch lets filesystems handle masking the file mode on creation. This is needed if filesystem is using ACLs. - The CREATE, MKDIR and MKNOD requests are extended with a "umask" parameter. - A new FUSE_DONT_MASK flag is added to the INIT request/reply. With this the filesystem may request that the create mode is not masked. CC: Jean-Pierre André Signed-off-by: Miklos Szeredi --- fs/fuse/dir.c | 20 +++++++++++++++++--- fs/fuse/fuse_i.h | 3 +++ fs/fuse/inode.c | 9 ++++++++- include/linux/fuse.h | 20 ++++++++++++++++++-- 4 files changed, 46 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c index b3089a083d30..6b700734e519 100644 --- a/fs/fuse/dir.c +++ b/fs/fuse/dir.c @@ -375,7 +375,7 @@ static int fuse_create_open(struct inode *dir, struct dentry *entry, int mode, struct fuse_conn *fc = get_fuse_conn(dir); struct fuse_req *req; struct fuse_req *forget_req; - struct fuse_open_in inarg; + struct fuse_create_in inarg; struct fuse_open_out outopen; struct fuse_entry_out outentry; struct fuse_file *ff; @@ -399,15 +399,20 @@ static int fuse_create_open(struct inode *dir, struct dentry *entry, int mode, if (!ff) goto out_put_request; + if (!fc->dont_mask) + mode &= ~current_umask(); + flags &= ~O_NOCTTY; memset(&inarg, 0, sizeof(inarg)); memset(&outentry, 0, sizeof(outentry)); inarg.flags = flags; inarg.mode = mode; + inarg.umask = current_umask(); req->in.h.opcode = FUSE_CREATE; req->in.h.nodeid = get_node_id(dir); req->in.numargs = 2; - req->in.args[0].size = sizeof(inarg); + req->in.args[0].size = fc->minor < 12 ? sizeof(struct fuse_open_in) : + sizeof(inarg); req->in.args[0].value = &inarg; req->in.args[1].size = entry->d_name.len + 1; req->in.args[1].value = entry->d_name.name; @@ -546,12 +551,17 @@ static int fuse_mknod(struct inode *dir, struct dentry *entry, int mode, if (IS_ERR(req)) return PTR_ERR(req); + if (!fc->dont_mask) + mode &= ~current_umask(); + memset(&inarg, 0, sizeof(inarg)); inarg.mode = mode; inarg.rdev = new_encode_dev(rdev); + inarg.umask = current_umask(); req->in.h.opcode = FUSE_MKNOD; req->in.numargs = 2; - req->in.args[0].size = sizeof(inarg); + req->in.args[0].size = fc->minor < 12 ? FUSE_COMPAT_MKNOD_IN_SIZE : + sizeof(inarg); req->in.args[0].value = &inarg; req->in.args[1].size = entry->d_name.len + 1; req->in.args[1].value = entry->d_name.name; @@ -578,8 +588,12 @@ static int fuse_mkdir(struct inode *dir, struct dentry *entry, int mode) if (IS_ERR(req)) return PTR_ERR(req); + if (!fc->dont_mask) + mode &= ~current_umask(); + memset(&inarg, 0, sizeof(inarg)); inarg.mode = mode; + inarg.umask = current_umask(); req->in.h.opcode = FUSE_MKDIR; req->in.numargs = 2; req->in.args[0].size = sizeof(inarg); diff --git a/fs/fuse/fuse_i.h b/fs/fuse/fuse_i.h index aaf2f9ff970e..ede4f77b2d6c 100644 --- a/fs/fuse/fuse_i.h +++ b/fs/fuse/fuse_i.h @@ -446,6 +446,9 @@ struct fuse_conn { /** Do multi-page cached writes */ unsigned big_writes:1; + /** Don't apply umask to creation modes */ + unsigned dont_mask:1; + /** The number of requests waiting for completion */ atomic_t num_waiting; diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c index d8673ccf90b7..6cc501bd0187 100644 --- a/fs/fuse/inode.c +++ b/fs/fuse/inode.c @@ -725,6 +725,8 @@ static void process_init_reply(struct fuse_conn *fc, struct fuse_req *req) } if (arg->flags & FUSE_BIG_WRITES) fc->big_writes = 1; + if (arg->flags & FUSE_DONT_MASK) + fc->dont_mask = 1; } else { ra_pages = fc->max_read / PAGE_CACHE_SIZE; fc->no_lock = 1; @@ -748,7 +750,7 @@ static void fuse_send_init(struct fuse_conn *fc, struct fuse_req *req) arg->minor = FUSE_KERNEL_MINOR_VERSION; arg->max_readahead = fc->bdi.ra_pages * PAGE_CACHE_SIZE; arg->flags |= FUSE_ASYNC_READ | FUSE_POSIX_LOCKS | FUSE_ATOMIC_O_TRUNC | - FUSE_EXPORT_SUPPORT | FUSE_BIG_WRITES; + FUSE_EXPORT_SUPPORT | FUSE_BIG_WRITES | FUSE_DONT_MASK; req->in.h.opcode = FUSE_INIT; req->in.numargs = 1; req->in.args[0].size = sizeof(*arg); @@ -864,6 +866,11 @@ static int fuse_fill_super(struct super_block *sb, void *data, int silent) if (err) goto err_put_conn; + /* Handle umasking inside the fuse code */ + if (sb->s_flags & MS_POSIXACL) + fc->dont_mask = 1; + sb->s_flags |= MS_POSIXACL; + fc->release = fuse_free_conn; fc->flags = d.flags; fc->user_id = d.user_id; diff --git a/include/linux/fuse.h b/include/linux/fuse.h index d41ed593f79f..e2b816a62488 100644 --- a/include/linux/fuse.h +++ b/include/linux/fuse.h @@ -25,6 +25,9 @@ * - add IOCTL message * - add unsolicited notification support * - add POLL message and NOTIFY_POLL notification + * + * 7.12 + * - add umask flag to input argument of open, mknod and mkdir */ #ifndef _LINUX_FUSE_H @@ -36,7 +39,7 @@ #define FUSE_KERNEL_VERSION 7 /** Minor version number of this interface */ -#define FUSE_KERNEL_MINOR_VERSION 11 +#define FUSE_KERNEL_MINOR_VERSION 12 /** The node ID of the root inode */ #define FUSE_ROOT_ID 1 @@ -112,6 +115,7 @@ struct fuse_file_lock { * INIT request/reply flags * * FUSE_EXPORT_SUPPORT: filesystem handles lookups of "." and ".." + * FUSE_DONT_MASK: don't apply umask to file mode on create operations */ #define FUSE_ASYNC_READ (1 << 0) #define FUSE_POSIX_LOCKS (1 << 1) @@ -119,6 +123,7 @@ struct fuse_file_lock { #define FUSE_ATOMIC_O_TRUNC (1 << 3) #define FUSE_EXPORT_SUPPORT (1 << 4) #define FUSE_BIG_WRITES (1 << 5) +#define FUSE_DONT_MASK (1 << 6) /** * CUSE INIT request/reply flags @@ -262,14 +267,18 @@ struct fuse_attr_out { struct fuse_attr attr; }; +#define FUSE_COMPAT_MKNOD_IN_SIZE 8 + struct fuse_mknod_in { __u32 mode; __u32 rdev; + __u32 umask; + __u32 padding; }; struct fuse_mkdir_in { __u32 mode; - __u32 padding; + __u32 umask; }; struct fuse_rename_in { @@ -300,8 +309,15 @@ struct fuse_setattr_in { }; struct fuse_open_in { + __u32 flags; + __u32 unused; +}; + +struct fuse_create_in { __u32 flags; __u32 mode; + __u32 umask; + __u32 padding; }; struct fuse_open_out { -- cgit v1.2.3 From 3b463ae0c6264f70e5d4c0a9c46af20fed43c96e Mon Sep 17 00:00:00 2001 From: John Muir Date: Sun, 31 May 2009 11:13:57 -0400 Subject: fuse: invalidation reverse calls Add notification messages that allow the filesystem to invalidate VFS caches. Two notifications are added: 1) inode invalidation - invalidate cached attributes - invalidate a range of pages in the page cache (this is optional) 2) dentry invalidation - try to invalidate a subtree in the dentry cache Care must be taken while accessing the 'struct super_block' for the mount, as it can go away while an invalidation is in progress. To prevent this, introduce a rw-semaphore, that is taken for read during the invalidation and taken for write in the ->kill_sb callback. Cc: Csaba Henk Cc: Anand Avati Signed-off-by: Miklos Szeredi --- fs/fuse/dev.c | 81 ++++++++++++++++++++++++++++++++++++++++++++++++++++ fs/fuse/dir.c | 37 ++++++++++++++++++++++++ fs/fuse/fuse_i.h | 24 ++++++++++++++++ fs/fuse/inode.c | 59 ++++++++++++++++++++++++++++++++++++-- include/linux/fuse.h | 16 +++++++++++ 5 files changed, 214 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c index 8a11a8c67c42..f58ecbc416c8 100644 --- a/fs/fuse/dev.c +++ b/fs/fuse/dev.c @@ -849,6 +849,81 @@ err: return err; } +static int fuse_notify_inval_inode(struct fuse_conn *fc, unsigned int size, + struct fuse_copy_state *cs) +{ + struct fuse_notify_inval_inode_out outarg; + int err = -EINVAL; + + if (size != sizeof(outarg)) + goto err; + + err = fuse_copy_one(cs, &outarg, sizeof(outarg)); + if (err) + goto err; + fuse_copy_finish(cs); + + down_read(&fc->killsb); + err = -ENOENT; + if (!fc->sb) + goto err_unlock; + + err = fuse_reverse_inval_inode(fc->sb, outarg.ino, + outarg.off, outarg.len); + +err_unlock: + up_read(&fc->killsb); + return err; + +err: + fuse_copy_finish(cs); + return err; +} + +static int fuse_notify_inval_entry(struct fuse_conn *fc, unsigned int size, + struct fuse_copy_state *cs) +{ + struct fuse_notify_inval_entry_out outarg; + int err = -EINVAL; + char buf[FUSE_NAME_MAX+1]; + struct qstr name; + + if (size < sizeof(outarg)) + goto err; + + err = fuse_copy_one(cs, &outarg, sizeof(outarg)); + if (err) + goto err; + + err = -ENAMETOOLONG; + if (outarg.namelen > FUSE_NAME_MAX) + goto err; + + name.name = buf; + name.len = outarg.namelen; + err = fuse_copy_one(cs, buf, outarg.namelen + 1); + if (err) + goto err; + fuse_copy_finish(cs); + buf[outarg.namelen] = 0; + name.hash = full_name_hash(name.name, name.len); + + down_read(&fc->killsb); + err = -ENOENT; + if (!fc->sb) + goto err_unlock; + + err = fuse_reverse_inval_entry(fc->sb, outarg.parent, &name); + +err_unlock: + up_read(&fc->killsb); + return err; + +err: + fuse_copy_finish(cs); + return err; +} + static int fuse_notify(struct fuse_conn *fc, enum fuse_notify_code code, unsigned int size, struct fuse_copy_state *cs) { @@ -856,6 +931,12 @@ static int fuse_notify(struct fuse_conn *fc, enum fuse_notify_code code, case FUSE_NOTIFY_POLL: return fuse_notify_poll(fc, size, cs); + case FUSE_NOTIFY_INVAL_INODE: + return fuse_notify_inval_inode(fc, size, cs); + + case FUSE_NOTIFY_INVAL_ENTRY: + return fuse_notify_inval_entry(fc, size, cs); + default: fuse_copy_finish(cs); return -EINVAL; diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c index 6b700734e519..e703654e7f40 100644 --- a/fs/fuse/dir.c +++ b/fs/fuse/dir.c @@ -859,6 +859,43 @@ int fuse_update_attributes(struct inode *inode, struct kstat *stat, return err; } +int fuse_reverse_inval_entry(struct super_block *sb, u64 parent_nodeid, + struct qstr *name) +{ + int err = -ENOTDIR; + struct inode *parent; + struct dentry *dir; + struct dentry *entry; + + parent = ilookup5(sb, parent_nodeid, fuse_inode_eq, &parent_nodeid); + if (!parent) + return -ENOENT; + + mutex_lock(&parent->i_mutex); + if (!S_ISDIR(parent->i_mode)) + goto unlock; + + err = -ENOENT; + dir = d_find_alias(parent); + if (!dir) + goto unlock; + + entry = d_lookup(dir, name); + dput(dir); + if (!entry) + goto unlock; + + fuse_invalidate_attr(parent); + fuse_invalidate_entry(entry); + dput(entry); + err = 0; + + unlock: + mutex_unlock(&parent->i_mutex); + iput(parent); + return err; +} + /* * Calling into a user-controlled filesystem gives the filesystem * daemon ptrace-like capabilities over the requester process. This diff --git a/fs/fuse/fuse_i.h b/fs/fuse/fuse_i.h index ede4f77b2d6c..52b641fc0faf 100644 --- a/fs/fuse/fuse_i.h +++ b/fs/fuse/fuse_i.h @@ -484,6 +484,12 @@ struct fuse_conn { /** Called on final put */ void (*release)(struct fuse_conn *); + + /** Super block for this connection. */ + struct super_block *sb; + + /** Read/write semaphore to hold when accessing sb. */ + struct rw_semaphore killsb; }; static inline struct fuse_conn *get_fuse_conn_super(struct super_block *sb) @@ -511,6 +517,11 @@ extern const struct file_operations fuse_dev_operations; extern const struct dentry_operations fuse_dentry_operations; +/** + * Inode to nodeid comparison. + */ +int fuse_inode_eq(struct inode *inode, void *_nodeidp); + /** * Get a filled in inode */ @@ -711,6 +722,19 @@ void fuse_release_nowrite(struct inode *inode); u64 fuse_get_attr_version(struct fuse_conn *fc); +/** + * File-system tells the kernel to invalidate cache for the given node id. + */ +int fuse_reverse_inval_inode(struct super_block *sb, u64 nodeid, + loff_t offset, loff_t len); + +/** + * File-system tells the kernel to invalidate parent attributes and + * the dentry matching parent/name. + */ +int fuse_reverse_inval_entry(struct super_block *sb, u64 parent_nodeid, + struct qstr *name); + int fuse_do_open(struct fuse_conn *fc, u64 nodeid, struct file *file, bool isdir); ssize_t fuse_direct_io(struct file *file, const char __user *buf, diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c index 6cc501bd0187..f91ccc4a189d 100644 --- a/fs/fuse/inode.c +++ b/fs/fuse/inode.c @@ -206,7 +206,7 @@ static void fuse_init_inode(struct inode *inode, struct fuse_attr *attr) BUG(); } -static int fuse_inode_eq(struct inode *inode, void *_nodeidp) +int fuse_inode_eq(struct inode *inode, void *_nodeidp) { u64 nodeid = *(u64 *) _nodeidp; if (get_node_id(inode) == nodeid) @@ -257,6 +257,31 @@ struct inode *fuse_iget(struct super_block *sb, u64 nodeid, return inode; } +int fuse_reverse_inval_inode(struct super_block *sb, u64 nodeid, + loff_t offset, loff_t len) +{ + struct inode *inode; + pgoff_t pg_start; + pgoff_t pg_end; + + inode = ilookup5(sb, nodeid, fuse_inode_eq, &nodeid); + if (!inode) + return -ENOENT; + + fuse_invalidate_attr(inode); + if (offset >= 0) { + pg_start = offset >> PAGE_CACHE_SHIFT; + if (len <= 0) + pg_end = -1; + else + pg_end = (offset + len - 1) >> PAGE_CACHE_SHIFT; + invalidate_inode_pages2_range(inode->i_mapping, + pg_start, pg_end); + } + iput(inode); + return 0; +} + static void fuse_umount_begin(struct super_block *sb) { fuse_abort_conn(get_fuse_conn_super(sb)); @@ -480,6 +505,7 @@ void fuse_conn_init(struct fuse_conn *fc) memset(fc, 0, sizeof(*fc)); spin_lock_init(&fc->lock); mutex_init(&fc->inst_mutex); + init_rwsem(&fc->killsb); atomic_set(&fc->count, 1); init_waitqueue_head(&fc->waitq); init_waitqueue_head(&fc->blocked_waitq); @@ -862,6 +888,7 @@ static int fuse_fill_super(struct super_block *sb, void *data, int silent) fuse_conn_init(fc); fc->dev = sb->s_dev; + fc->sb = sb; err = fuse_bdi_init(fc, sb); if (err) goto err_put_conn; @@ -948,12 +975,25 @@ static int fuse_get_sb(struct file_system_type *fs_type, return get_sb_nodev(fs_type, flags, raw_data, fuse_fill_super, mnt); } +static void fuse_kill_sb_anon(struct super_block *sb) +{ + struct fuse_conn *fc = get_fuse_conn_super(sb); + + if (fc) { + down_write(&fc->killsb); + fc->sb = NULL; + up_write(&fc->killsb); + } + + kill_anon_super(sb); +} + static struct file_system_type fuse_fs_type = { .owner = THIS_MODULE, .name = "fuse", .fs_flags = FS_HAS_SUBTYPE, .get_sb = fuse_get_sb, - .kill_sb = kill_anon_super, + .kill_sb = fuse_kill_sb_anon, }; #ifdef CONFIG_BLOCK @@ -965,11 +1005,24 @@ static int fuse_get_sb_blk(struct file_system_type *fs_type, mnt); } +static void fuse_kill_sb_blk(struct super_block *sb) +{ + struct fuse_conn *fc = get_fuse_conn_super(sb); + + if (fc) { + down_write(&fc->killsb); + fc->sb = NULL; + up_write(&fc->killsb); + } + + kill_block_super(sb); +} + static struct file_system_type fuseblk_fs_type = { .owner = THIS_MODULE, .name = "fuseblk", .get_sb = fuse_get_sb_blk, - .kill_sb = kill_block_super, + .kill_sb = fuse_kill_sb_blk, .fs_flags = FS_REQUIRES_DEV | FS_HAS_SUBTYPE, }; diff --git a/include/linux/fuse.h b/include/linux/fuse.h index e2b816a62488..cf593bf9fd32 100644 --- a/include/linux/fuse.h +++ b/include/linux/fuse.h @@ -28,6 +28,8 @@ * * 7.12 * - add umask flag to input argument of open, mknod and mkdir + * - add notification messages for invalidation of inodes and + * directory entries */ #ifndef _LINUX_FUSE_H @@ -229,6 +231,8 @@ enum fuse_opcode { enum fuse_notify_code { FUSE_NOTIFY_POLL = 1, + FUSE_NOTIFY_INVAL_INODE = 2, + FUSE_NOTIFY_INVAL_ENTRY = 3, FUSE_NOTIFY_CODE_MAX, }; @@ -524,4 +528,16 @@ struct fuse_dirent { #define FUSE_DIRENT_SIZE(d) \ FUSE_DIRENT_ALIGN(FUSE_NAME_OFFSET + (d)->namelen) +struct fuse_notify_inval_inode_out { + __u64 ino; + __s64 off; + __s64 len; +}; + +struct fuse_notify_inval_entry_out { + __u64 parent; + __u32 namelen; + __u32 padding; +}; + #endif /* _LINUX_FUSE_H */ -- cgit v1.2.3 From f7c2df9b55212d5ec94169a4de11e44c683e0af4 Mon Sep 17 00:00:00 2001 From: Jiri Slaby Date: Tue, 30 Jun 2009 21:10:13 +0100 Subject: AFS: Fix lock imbalance Don't unlock on vfs_rejected_lock path in afs_do_setlk, since the lock is unlocked after abort_attempt label. Signed-off-by: Jiri Slaby Signed-off-by: David Howells Signed-off-by: Linus Torvalds --- fs/afs/flock.c | 1 - 1 file changed, 1 deletion(-) (limited to 'fs') diff --git a/fs/afs/flock.c b/fs/afs/flock.c index 210acafe4a9b..3ff8bdd18fb3 100644 --- a/fs/afs/flock.c +++ b/fs/afs/flock.c @@ -432,7 +432,6 @@ vfs_rejected_lock: list_del_init(&fl->fl_u.afs.link); if (list_empty(&vnode->granted_locks)) afs_defer_unlock(vnode, key); - spin_unlock(&vnode->lock); goto abort_attempt; } -- cgit v1.2.3 From 133890103b9de08904f909995973e4b5c08a780e Mon Sep 17 00:00:00 2001 From: Davide Libenzi Date: Tue, 30 Jun 2009 11:41:11 -0700 Subject: eventfd: revised interface and cleanups Change the eventfd interface to de-couple the eventfd memory context, from the file pointer instance. Without such change, there is no clean way to racely free handle the POLLHUP event sent when the last instance of the file* goes away. Also, now the internal eventfd APIs are using the eventfd context instead of the file*. This patch is required by KVM's IRQfd code, which is still under development. Signed-off-by: Davide Libenzi Cc: Gregory Haskins Cc: Rusty Russell Cc: Benjamin LaHaise Cc: Avi Kivity Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/lguest/lg.h | 2 +- drivers/lguest/lguest_user.c | 4 +- fs/aio.c | 24 +++------ fs/eventfd.c | 122 ++++++++++++++++++++++++++++++++++++++----- include/linux/aio.h | 4 +- include/linux/eventfd.h | 35 ++++++++++--- 6 files changed, 149 insertions(+), 42 deletions(-) (limited to 'fs') diff --git a/drivers/lguest/lg.h b/drivers/lguest/lg.h index d4e8979735cb..9c3138265f8e 100644 --- a/drivers/lguest/lg.h +++ b/drivers/lguest/lg.h @@ -82,7 +82,7 @@ struct lg_cpu { struct lg_eventfd { unsigned long addr; - struct file *event; + struct eventfd_ctx *event; }; struct lg_eventfd_map { diff --git a/drivers/lguest/lguest_user.c b/drivers/lguest/lguest_user.c index 32e297121058..9f9a2953b383 100644 --- a/drivers/lguest/lguest_user.c +++ b/drivers/lguest/lguest_user.c @@ -50,7 +50,7 @@ static int add_eventfd(struct lguest *lg, unsigned long addr, int fd) /* Now append new entry. */ new->map[new->num].addr = addr; - new->map[new->num].event = eventfd_fget(fd); + new->map[new->num].event = eventfd_ctx_fdget(fd); if (IS_ERR(new->map[new->num].event)) { kfree(new); return PTR_ERR(new->map[new->num].event); @@ -357,7 +357,7 @@ static int close(struct inode *inode, struct file *file) /* Release any eventfds they registered. */ for (i = 0; i < lg->eventfds->num; i++) - fput(lg->eventfds->map[i].event); + eventfd_ctx_put(lg->eventfds->map[i].event); kfree(lg->eventfds); /* If lg->dead doesn't contain an error code it will be NULL or a diff --git a/fs/aio.c b/fs/aio.c index 76da12537956..d065b2c3273e 100644 --- a/fs/aio.c +++ b/fs/aio.c @@ -485,6 +485,8 @@ static inline void really_put_req(struct kioctx *ctx, struct kiocb *req) { assert_spin_locked(&ctx->ctx_lock); + if (req->ki_eventfd != NULL) + eventfd_ctx_put(req->ki_eventfd); if (req->ki_dtor) req->ki_dtor(req); if (req->ki_iovec != &req->ki_inline_vec) @@ -509,8 +511,6 @@ static void aio_fput_routine(struct work_struct *data) /* Complete the fput(s) */ if (req->ki_filp != NULL) __fput(req->ki_filp); - if (req->ki_eventfd != NULL) - __fput(req->ki_eventfd); /* Link the iocb into the context's free list */ spin_lock_irq(&ctx->ctx_lock); @@ -528,8 +528,6 @@ static void aio_fput_routine(struct work_struct *data) */ static int __aio_put_req(struct kioctx *ctx, struct kiocb *req) { - int schedule_putreq = 0; - dprintk(KERN_DEBUG "aio_put(%p): f_count=%ld\n", req, atomic_long_read(&req->ki_filp->f_count)); @@ -549,24 +547,16 @@ static int __aio_put_req(struct kioctx *ctx, struct kiocb *req) * we would not be holding the last reference to the file*, so * this function will be executed w/out any aio kthread wakeup. */ - if (unlikely(atomic_long_dec_and_test(&req->ki_filp->f_count))) - schedule_putreq++; - else - req->ki_filp = NULL; - if (req->ki_eventfd != NULL) { - if (unlikely(atomic_long_dec_and_test(&req->ki_eventfd->f_count))) - schedule_putreq++; - else - req->ki_eventfd = NULL; - } - if (unlikely(schedule_putreq)) { + if (unlikely(atomic_long_dec_and_test(&req->ki_filp->f_count))) { get_ioctx(ctx); spin_lock(&fput_lock); list_add(&req->ki_list, &fput_head); spin_unlock(&fput_lock); queue_work(aio_wq, &fput_work); - } else + } else { + req->ki_filp = NULL; really_put_req(ctx, req); + } return 1; } @@ -1622,7 +1612,7 @@ static int io_submit_one(struct kioctx *ctx, struct iocb __user *user_iocb, * an eventfd() fd, and will be signaled for each completed * event using the eventfd_signal() function. */ - req->ki_eventfd = eventfd_fget((int) iocb->aio_resfd); + req->ki_eventfd = eventfd_ctx_fdget((int) iocb->aio_resfd); if (IS_ERR(req->ki_eventfd)) { ret = PTR_ERR(req->ki_eventfd); req->ki_eventfd = NULL; diff --git a/fs/eventfd.c b/fs/eventfd.c index 3f0e1974abdc..31d12de83a2a 100644 --- a/fs/eventfd.c +++ b/fs/eventfd.c @@ -14,35 +14,44 @@ #include #include #include -#include #include #include +#include +#include struct eventfd_ctx { + struct kref kref; wait_queue_head_t wqh; /* * Every time that a write(2) is performed on an eventfd, the * value of the __u64 being written is added to "count" and a * wakeup is performed on "wqh". A read(2) will return the "count" * value to userspace, and will reset "count" to zero. The kernel - * size eventfd_signal() also, adds to the "count" counter and + * side eventfd_signal() also, adds to the "count" counter and * issue a wakeup. */ __u64 count; unsigned int flags; }; -/* - * Adds "n" to the eventfd counter "count". Returns "n" in case of - * success, or a value lower then "n" in case of coutner overflow. - * This function is supposed to be called by the kernel in paths - * that do not allow sleeping. In this function we allow the counter - * to reach the ULLONG_MAX value, and we signal this as overflow - * condition by returining a POLLERR to poll(2). +/** + * eventfd_signal - Adds @n to the eventfd counter. + * @ctx: [in] Pointer to the eventfd context. + * @n: [in] Value of the counter to be added to the eventfd internal counter. + * The value cannot be negative. + * + * This function is supposed to be called by the kernel in paths that do not + * allow sleeping. In this function we allow the counter to reach the ULLONG_MAX + * value, and we signal this as overflow condition by returining a POLLERR + * to poll(2). + * + * Returns @n in case of success, a non-negative number lower than @n in case + * of overflow, or the following error codes: + * + * -EINVAL : The value of @n is negative. */ -int eventfd_signal(struct file *file, int n) +int eventfd_signal(struct eventfd_ctx *ctx, int n) { - struct eventfd_ctx *ctx = file->private_data; unsigned long flags; if (n < 0) @@ -59,9 +68,45 @@ int eventfd_signal(struct file *file, int n) } EXPORT_SYMBOL_GPL(eventfd_signal); +static void eventfd_free(struct kref *kref) +{ + struct eventfd_ctx *ctx = container_of(kref, struct eventfd_ctx, kref); + + kfree(ctx); +} + +/** + * eventfd_ctx_get - Acquires a reference to the internal eventfd context. + * @ctx: [in] Pointer to the eventfd context. + * + * Returns: In case of success, returns a pointer to the eventfd context. + */ +struct eventfd_ctx *eventfd_ctx_get(struct eventfd_ctx *ctx) +{ + kref_get(&ctx->kref); + return ctx; +} +EXPORT_SYMBOL_GPL(eventfd_ctx_get); + +/** + * eventfd_ctx_put - Releases a reference to the internal eventfd context. + * @ctx: [in] Pointer to eventfd context. + * + * The eventfd context reference must have been previously acquired either + * with eventfd_ctx_get() or eventfd_ctx_fdget()). + */ +void eventfd_ctx_put(struct eventfd_ctx *ctx) +{ + kref_put(&ctx->kref, eventfd_free); +} +EXPORT_SYMBOL_GPL(eventfd_ctx_put); + static int eventfd_release(struct inode *inode, struct file *file) { - kfree(file->private_data); + struct eventfd_ctx *ctx = file->private_data; + + wake_up_poll(&ctx->wqh, POLLHUP); + eventfd_ctx_put(ctx); return 0; } @@ -185,6 +230,16 @@ static const struct file_operations eventfd_fops = { .write = eventfd_write, }; +/** + * eventfd_fget - Acquire a reference of an eventfd file descriptor. + * @fd: [in] Eventfd file descriptor. + * + * Returns a pointer to the eventfd file structure in case of success, or the + * following error pointer: + * + * -EBADF : Invalid @fd file descriptor. + * -EINVAL : The @fd file descriptor is not an eventfd file. + */ struct file *eventfd_fget(int fd) { struct file *file; @@ -201,6 +256,48 @@ struct file *eventfd_fget(int fd) } EXPORT_SYMBOL_GPL(eventfd_fget); +/** + * eventfd_ctx_fdget - Acquires a reference to the internal eventfd context. + * @fd: [in] Eventfd file descriptor. + * + * Returns a pointer to the internal eventfd context, otherwise the error + * pointers returned by the following functions: + * + * eventfd_fget + */ +struct eventfd_ctx *eventfd_ctx_fdget(int fd) +{ + struct file *file; + struct eventfd_ctx *ctx; + + file = eventfd_fget(fd); + if (IS_ERR(file)) + return (struct eventfd_ctx *) file; + ctx = eventfd_ctx_get(file->private_data); + fput(file); + + return ctx; +} +EXPORT_SYMBOL_GPL(eventfd_ctx_fdget); + +/** + * eventfd_ctx_fileget - Acquires a reference to the internal eventfd context. + * @file: [in] Eventfd file pointer. + * + * Returns a pointer to the internal eventfd context, otherwise the error + * pointer: + * + * -EINVAL : The @fd file descriptor is not an eventfd file. + */ +struct eventfd_ctx *eventfd_ctx_fileget(struct file *file) +{ + if (file->f_op != &eventfd_fops) + return ERR_PTR(-EINVAL); + + return eventfd_ctx_get(file->private_data); +} +EXPORT_SYMBOL_GPL(eventfd_ctx_fileget); + SYSCALL_DEFINE2(eventfd2, unsigned int, count, int, flags) { int fd; @@ -217,6 +314,7 @@ SYSCALL_DEFINE2(eventfd2, unsigned int, count, int, flags) if (!ctx) return -ENOMEM; + kref_init(&ctx->kref); init_waitqueue_head(&ctx->wqh); ctx->count = count; ctx->flags = flags; diff --git a/include/linux/aio.h b/include/linux/aio.h index b16a957030f8..47f7d932a01d 100644 --- a/include/linux/aio.h +++ b/include/linux/aio.h @@ -121,9 +121,9 @@ struct kiocb { /* * If the aio_resfd field of the userspace iocb is not zero, - * this is the underlying file* to deliver event to. + * this is the underlying eventfd context to deliver events to. */ - struct file *ki_eventfd; + struct eventfd_ctx *ki_eventfd; }; #define is_sync_kiocb(iocb) ((iocb)->ki_key == KIOCB_SYNC_KEY) diff --git a/include/linux/eventfd.h b/include/linux/eventfd.h index f45a8ae5f828..3b85ba6479f4 100644 --- a/include/linux/eventfd.h +++ b/include/linux/eventfd.h @@ -8,10 +8,8 @@ #ifndef _LINUX_EVENTFD_H #define _LINUX_EVENTFD_H -#ifdef CONFIG_EVENTFD - -/* For O_CLOEXEC and O_NONBLOCK */ #include +#include /* * CAREFUL: Check include/asm-generic/fcntl.h when defining @@ -27,16 +25,37 @@ #define EFD_SHARED_FCNTL_FLAGS (O_CLOEXEC | O_NONBLOCK) #define EFD_FLAGS_SET (EFD_SHARED_FCNTL_FLAGS | EFD_SEMAPHORE) +#ifdef CONFIG_EVENTFD + +struct eventfd_ctx *eventfd_ctx_get(struct eventfd_ctx *ctx); +void eventfd_ctx_put(struct eventfd_ctx *ctx); struct file *eventfd_fget(int fd); -int eventfd_signal(struct file *file, int n); +struct eventfd_ctx *eventfd_ctx_fdget(int fd); +struct eventfd_ctx *eventfd_ctx_fileget(struct file *file); +int eventfd_signal(struct eventfd_ctx *ctx, int n); #else /* CONFIG_EVENTFD */ -#define eventfd_fget(fd) ERR_PTR(-ENOSYS) -static inline int eventfd_signal(struct file *file, int n) -{ return 0; } +/* + * Ugly ugly ugly error layer to support modules that uses eventfd but + * pretend to work in !CONFIG_EVENTFD configurations. Namely, AIO. + */ +static inline struct eventfd_ctx *eventfd_ctx_fdget(int fd) +{ + return ERR_PTR(-ENOSYS); +} + +static inline int eventfd_signal(struct eventfd_ctx *ctx, int n) +{ + return -ENOSYS; +} + +static inline void eventfd_ctx_put(struct eventfd_ctx *ctx) +{ + +} -#endif /* CONFIG_EVENTFD */ +#endif #endif /* _LINUX_EVENTFD_H */ -- cgit v1.2.3 From 341c87bf346f57748230628c5ad6ee69219250e8 Mon Sep 17 00:00:00 2001 From: KAMEZAWA Hiroyuki Date: Tue, 30 Jun 2009 11:41:23 -0700 Subject: elf: limit max map count to safe value With ELF, at generating coredump, some more headers other than used vmas are added. When max_map_count == 65536, a core generated by following kinds of code can be unreadable because the number of ELF's program header is written in 16bit in Ehdr (please see elf.h) and the number overflows. == ... = mmap(); (munmap, mprotect, etc...) if (failed) abort(); == This can happen in mmap/munmap/mprotect/etc...which calls split_vma(). I think 65536 is not safe as _default_ and reduce it to 65530 is good for avoiding unexpected corrupted core. Anyway, max_map_count can be enlarged by sysctl if a user is brave.. Signed-off-by: KAMEZAWA Hiroyuki Cc: Hugh Dickins Cc: Jakub Jelinek Acked-by: Roland McGrath Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/binfmt_elf.c | 5 ++++- include/linux/sched.h | 16 ++++++++++++++-- 2 files changed, 18 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/binfmt_elf.c b/fs/binfmt_elf.c index 9fa212b014a5..f1867900e459 100644 --- a/fs/binfmt_elf.c +++ b/fs/binfmt_elf.c @@ -1929,7 +1929,10 @@ static int elf_core_dump(long signr, struct pt_regs *regs, struct file *file, un elf = kmalloc(sizeof(*elf), GFP_KERNEL); if (!elf) goto out; - + /* + * The number of segs are recored into ELF header as 16bit value. + * Please check DEFAULT_MAX_MAP_COUNT definition when you modify here. + */ segs = current->mm->map_count; #ifdef ELF_CORE_EXTRA_PHDRS segs += ELF_CORE_EXTRA_PHDRS; diff --git a/include/linux/sched.h b/include/linux/sched.h index 4d0754269884..0085d758d645 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -349,8 +349,20 @@ extern int mutex_spin_on_owner(struct mutex *lock, struct thread_info *owner); struct nsproxy; struct user_namespace; -/* Maximum number of active map areas.. This is a random (large) number */ -#define DEFAULT_MAX_MAP_COUNT 65536 +/* + * Default maximum number of active map areas, this limits the number of vmas + * per mm struct. Users can overwrite this number by sysctl but there is a + * problem. + * + * When a program's coredump is generated as ELF format, a section is created + * per a vma. In ELF, the number of sections is represented in unsigned short. + * This means the number of sections should be smaller than 65535 at coredump. + * Because the kernel adds some informative sections to a image of program at + * generating coredump, we need some margin. The number of extra sections is + * 1-3 now and depends on arch. We use "5" as safe margin, here. + */ +#define MAPCOUNT_ELF_CORE_MARGIN (5) +#define DEFAULT_MAX_MAP_COUNT (USHORT_MAX - MAPCOUNT_ELF_CORE_MARGIN) extern int sysctl_max_map_count; -- cgit v1.2.3 From 4d6c13f87db12ae1ce35ea6a15688ac72419b133 Mon Sep 17 00:00:00 2001 From: Bryan Donlan Date: Tue, 30 Jun 2009 11:41:24 -0700 Subject: ext2: return -EIO not -ESTALE on directory traversal through deleted inode ext2_iget() returns -ESTALE if invoked on a deleted inode, in order to report errors to NFS properly. However, in ext[234]_lookup(), this -ESTALE can be propagated to userspace if the filesystem is corrupted such that a directory entry references a deleted inode. This leads to a misleading error message - "Stale NFS file handle" - and confusion on the part of the admin. The bug can be easily reproduced by creating a new filesystem, making a link to an unused inode using debugfs, then mounting and attempting to ls -l said link. This patch thus changes ext2_lookup to return -EIO if it receives -ESTALE from ext2_iget(), as ext2 does for other filesystem metadata corruption; and also invokes the appropriate ext*_error functions when this case is detected. Signed-off-by: Bryan Donlan Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/ext2/namei.c | 12 ++++++++++-- 1 file changed, 10 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/ext2/namei.c b/fs/ext2/namei.c index 6524ecaebb7a..e1dedb0f7873 100644 --- a/fs/ext2/namei.c +++ b/fs/ext2/namei.c @@ -66,8 +66,16 @@ static struct dentry *ext2_lookup(struct inode * dir, struct dentry *dentry, str inode = NULL; if (ino) { inode = ext2_iget(dir->i_sb, ino); - if (IS_ERR(inode)) - return ERR_CAST(inode); + if (unlikely(IS_ERR(inode))) { + if (PTR_ERR(inode) == -ESTALE) { + ext2_error(dir->i_sb, __func__, + "deleted inode referenced: %lu", + ino); + return ERR_PTR(-EIO); + } else { + return ERR_CAST(inode); + } + } } return d_splice_alias(inode, dentry); } -- cgit v1.2.3 From 752fa51e4c5182c3c257f1cede90577a7e213c58 Mon Sep 17 00:00:00 2001 From: Wolfgang Illmeyer Date: Tue, 30 Jun 2009 11:41:44 -0700 Subject: hostfs: set maximum filesize in superblock for proper LFS support Maximum file size for hostfs mounts defaults to 2GB, so bigger files cannot be read/written through hostfs. This patch initializes the maximum file size to MAX_LFS_SIZE. Addresses http://bugzilla.kernel.org/show_bug.cgi?id=13531 Signed-off-by: Wolfgang Illmeyer Cc: Jeff Dike Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/hostfs/hostfs_kern.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/hostfs/hostfs_kern.c b/fs/hostfs/hostfs_kern.c index fe02ad4740e7..032604e5ef2c 100644 --- a/fs/hostfs/hostfs_kern.c +++ b/fs/hostfs/hostfs_kern.c @@ -972,6 +972,7 @@ static int hostfs_fill_sb_common(struct super_block *sb, void *d, int silent) sb->s_blocksize_bits = 10; sb->s_magic = HOSTFS_SUPER_MAGIC; sb->s_op = &hostfs_sbops; + sb->s_maxbytes = MAX_LFS_FILESIZE; /* NULL is printed as by sprintf: avoid that. */ if (req_root == NULL) -- cgit v1.2.3 From 7878cba9f0037f5599004b03a1260b32d9050360 Mon Sep 17 00:00:00 2001 From: "Martin K. Petersen" Date: Fri, 26 Jun 2009 15:37:49 +0200 Subject: block: Create bip slabs with embedded integrity vectors This patch restores stacking ability to the block layer integrity infrastructure by creating a set of dedicated bip slabs. Each bip slab has an embedded bio_vec array at the end. This cuts down on memory allocations and also simplifies the code compared to the original bvec version. Only the largest bip slab is backed by a mempool. The pool is contained in the bio_set so stacking drivers can ensure forward progress. Signed-off-by: Martin K. Petersen Signed-off-by: Jens Axboe --- block/blk-core.c | 2 +- drivers/md/dm.c | 4 +- fs/bio-integrity.c | 170 ++++++++++++++++++++++++++++++++++++++-------------- fs/bio.c | 11 +++- include/linux/bio.h | 22 +++++-- 5 files changed, 152 insertions(+), 57 deletions(-) (limited to 'fs') diff --git a/block/blk-core.c b/block/blk-core.c index b06cf5c2a829..345d99da8d41 100644 --- a/block/blk-core.c +++ b/block/blk-core.c @@ -2365,7 +2365,7 @@ int blk_rq_prep_clone(struct request *rq, struct request *rq_src, __bio_clone(bio, bio_src); if (bio_integrity(bio_src) && - bio_integrity_clone(bio, bio_src, gfp_mask)) + bio_integrity_clone(bio, bio_src, gfp_mask, bs)) goto free_and_out; if (bio_ctr && bio_ctr(bio, bio_src, data)) diff --git a/drivers/md/dm.c b/drivers/md/dm.c index 3c6d4ee8921d..9acd54a5cffb 100644 --- a/drivers/md/dm.c +++ b/drivers/md/dm.c @@ -1017,7 +1017,7 @@ static struct bio *split_bvec(struct bio *bio, sector_t sector, clone->bi_flags |= 1 << BIO_CLONED; if (bio_integrity(bio)) { - bio_integrity_clone(clone, bio, GFP_NOIO); + bio_integrity_clone(clone, bio, GFP_NOIO, bs); bio_integrity_trim(clone, bio_sector_offset(bio, idx, offset), len); } @@ -1045,7 +1045,7 @@ static struct bio *clone_bio(struct bio *bio, sector_t sector, clone->bi_flags &= ~(1 << BIO_SEG_VALID); if (bio_integrity(bio)) { - bio_integrity_clone(clone, bio, GFP_NOIO); + bio_integrity_clone(clone, bio, GFP_NOIO, bs); if (idx != bio->bi_idx || clone->bi_size < bio->bi_size) bio_integrity_trim(clone, diff --git a/fs/bio-integrity.c b/fs/bio-integrity.c index 31c46a241bac..49a34e7f7306 100644 --- a/fs/bio-integrity.c +++ b/fs/bio-integrity.c @@ -1,7 +1,7 @@ /* * bio-integrity.c - bio data integrity extensions * - * Copyright (C) 2007, 2008 Oracle Corporation + * Copyright (C) 2007, 2008, 2009 Oracle Corporation * Written by: Martin K. Petersen * * This program is free software; you can redistribute it and/or @@ -25,63 +25,121 @@ #include #include -static struct kmem_cache *bio_integrity_slab __read_mostly; -static mempool_t *bio_integrity_pool; -static struct bio_set *integrity_bio_set; +struct integrity_slab { + struct kmem_cache *slab; + unsigned short nr_vecs; + char name[8]; +}; + +#define IS(x) { .nr_vecs = x, .name = "bip-"__stringify(x) } +struct integrity_slab bip_slab[BIOVEC_NR_POOLS] __read_mostly = { + IS(1), IS(4), IS(16), IS(64), IS(128), IS(BIO_MAX_PAGES), +}; +#undef IS + static struct workqueue_struct *kintegrityd_wq; +static inline unsigned int vecs_to_idx(unsigned int nr) +{ + switch (nr) { + case 1: + return 0; + case 2 ... 4: + return 1; + case 5 ... 16: + return 2; + case 17 ... 64: + return 3; + case 65 ... 128: + return 4; + case 129 ... BIO_MAX_PAGES: + return 5; + default: + BUG(); + } +} + +static inline int use_bip_pool(unsigned int idx) +{ + if (idx == BIOVEC_NR_POOLS) + return 1; + + return 0; +} + /** - * bio_integrity_alloc - Allocate integrity payload and attach it to bio + * bio_integrity_alloc_bioset - Allocate integrity payload and attach it to bio * @bio: bio to attach integrity metadata to * @gfp_mask: Memory allocation mask * @nr_vecs: Number of integrity metadata scatter-gather elements + * @bs: bio_set to allocate from * * Description: This function prepares a bio for attaching integrity * metadata. nr_vecs specifies the maximum number of pages containing * integrity metadata that can be attached. */ -struct bio_integrity_payload *bio_integrity_alloc(struct bio *bio, - gfp_t gfp_mask, - unsigned int nr_vecs) +struct bio_integrity_payload *bio_integrity_alloc_bioset(struct bio *bio, + gfp_t gfp_mask, + unsigned int nr_vecs, + struct bio_set *bs) { struct bio_integrity_payload *bip; - struct bio_vec *iv; - unsigned long idx; + unsigned int idx = vecs_to_idx(nr_vecs); BUG_ON(bio == NULL); + bip = NULL; - bip = mempool_alloc(bio_integrity_pool, gfp_mask); - if (unlikely(bip == NULL)) { - printk(KERN_ERR "%s: could not alloc bip\n", __func__); - return NULL; - } + /* Lower order allocations come straight from slab */ + if (!use_bip_pool(idx)) + bip = kmem_cache_alloc(bip_slab[idx].slab, gfp_mask); - memset(bip, 0, sizeof(*bip)); + /* Use mempool if lower order alloc failed or max vecs were requested */ + if (bip == NULL) { + bip = mempool_alloc(bs->bio_integrity_pool, gfp_mask); - iv = bvec_alloc_bs(gfp_mask, nr_vecs, &idx, integrity_bio_set); - if (unlikely(iv == NULL)) { - printk(KERN_ERR "%s: could not alloc bip_vec\n", __func__); - mempool_free(bip, bio_integrity_pool); - return NULL; + if (unlikely(bip == NULL)) { + printk(KERN_ERR "%s: could not alloc bip\n", __func__); + return NULL; + } } - bip->bip_pool = idx; - bip->bip_vec = iv; + memset(bip, 0, sizeof(*bip)); + + bip->bip_slab = idx; bip->bip_bio = bio; bio->bi_integrity = bip; return bip; } +EXPORT_SYMBOL(bio_integrity_alloc_bioset); + +/** + * bio_integrity_alloc - Allocate integrity payload and attach it to bio + * @bio: bio to attach integrity metadata to + * @gfp_mask: Memory allocation mask + * @nr_vecs: Number of integrity metadata scatter-gather elements + * + * Description: This function prepares a bio for attaching integrity + * metadata. nr_vecs specifies the maximum number of pages containing + * integrity metadata that can be attached. + */ +struct bio_integrity_payload *bio_integrity_alloc(struct bio *bio, + gfp_t gfp_mask, + unsigned int nr_vecs) +{ + return bio_integrity_alloc_bioset(bio, gfp_mask, nr_vecs, fs_bio_set); +} EXPORT_SYMBOL(bio_integrity_alloc); /** * bio_integrity_free - Free bio integrity payload * @bio: bio containing bip to be freed + * @bs: bio_set this bio was allocated from * * Description: Used to free the integrity portion of a bio. Usually * called from bio_free(). */ -void bio_integrity_free(struct bio *bio) +void bio_integrity_free(struct bio *bio, struct bio_set *bs) { struct bio_integrity_payload *bip = bio->bi_integrity; @@ -92,8 +150,10 @@ void bio_integrity_free(struct bio *bio) && bip->bip_buf != NULL) kfree(bip->bip_buf); - bvec_free_bs(integrity_bio_set, bip->bip_vec, bip->bip_pool); - mempool_free(bip, bio_integrity_pool); + if (use_bip_pool(bip->bip_slab)) + mempool_free(bip, bs->bio_integrity_pool); + else + kmem_cache_free(bip_slab[bip->bip_slab].slab, bip); bio->bi_integrity = NULL; } @@ -114,7 +174,7 @@ int bio_integrity_add_page(struct bio *bio, struct page *page, struct bio_integrity_payload *bip = bio->bi_integrity; struct bio_vec *iv; - if (bip->bip_vcnt >= bvec_nr_vecs(bip->bip_pool)) { + if (bip->bip_vcnt >= bvec_nr_vecs(bip->bip_slab)) { printk(KERN_ERR "%s: bip_vec full\n", __func__); return 0; } @@ -647,8 +707,8 @@ void bio_integrity_split(struct bio *bio, struct bio_pair *bp, int sectors) bp->iv1 = bip->bip_vec[0]; bp->iv2 = bip->bip_vec[0]; - bp->bip1.bip_vec = &bp->iv1; - bp->bip2.bip_vec = &bp->iv2; + bp->bip1.bip_vec[0] = bp->iv1; + bp->bip2.bip_vec[0] = bp->iv2; bp->iv1.bv_len = sectors * bi->tuple_size; bp->iv2.bv_offset += sectors * bi->tuple_size; @@ -667,17 +727,19 @@ EXPORT_SYMBOL(bio_integrity_split); * @bio: New bio * @bio_src: Original bio * @gfp_mask: Memory allocation mask + * @bs: bio_set to allocate bip from * * Description: Called to allocate a bip when cloning a bio */ -int bio_integrity_clone(struct bio *bio, struct bio *bio_src, gfp_t gfp_mask) +int bio_integrity_clone(struct bio *bio, struct bio *bio_src, + gfp_t gfp_mask, struct bio_set *bs) { struct bio_integrity_payload *bip_src = bio_src->bi_integrity; struct bio_integrity_payload *bip; BUG_ON(bip_src == NULL); - bip = bio_integrity_alloc(bio, gfp_mask, bip_src->bip_vcnt); + bip = bio_integrity_alloc_bioset(bio, gfp_mask, bip_src->bip_vcnt, bs); if (bip == NULL) return -EIO; @@ -693,25 +755,43 @@ int bio_integrity_clone(struct bio *bio, struct bio *bio_src, gfp_t gfp_mask) } EXPORT_SYMBOL(bio_integrity_clone); -static int __init bio_integrity_init(void) +int bioset_integrity_create(struct bio_set *bs, int pool_size) { - kintegrityd_wq = create_workqueue("kintegrityd"); + unsigned int max_slab = vecs_to_idx(BIO_MAX_PAGES); + + bs->bio_integrity_pool = + mempool_create_slab_pool(pool_size, bip_slab[max_slab].slab); + if (!bs->bio_integrity_pool) + return -1; + + return 0; +} +EXPORT_SYMBOL(bioset_integrity_create); + +void bioset_integrity_free(struct bio_set *bs) +{ + if (bs->bio_integrity_pool) + mempool_destroy(bs->bio_integrity_pool); +} +EXPORT_SYMBOL(bioset_integrity_free); + +void __init bio_integrity_init(void) +{ + unsigned int i; + + kintegrityd_wq = create_workqueue("kintegrityd"); if (!kintegrityd_wq) panic("Failed to create kintegrityd\n"); - bio_integrity_slab = KMEM_CACHE(bio_integrity_payload, - SLAB_HWCACHE_ALIGN|SLAB_PANIC); + for (i = 0 ; i < BIOVEC_NR_POOLS ; i++) { + unsigned int size; - bio_integrity_pool = mempool_create_slab_pool(BIO_POOL_SIZE, - bio_integrity_slab); - if (!bio_integrity_pool) - panic("bio_integrity: can't allocate bip pool\n"); + size = sizeof(struct bio_integrity_payload) + + bip_slab[i].nr_vecs * sizeof(struct bio_vec); - integrity_bio_set = bioset_create(BIO_POOL_SIZE, 0); - if (!integrity_bio_set) - panic("bio_integrity: can't allocate bio_set\n"); - - return 0; + bip_slab[i].slab = + kmem_cache_create(bip_slab[i].name, size, 0, + SLAB_HWCACHE_ALIGN|SLAB_PANIC, NULL); + } } -subsys_initcall(bio_integrity_init); diff --git a/fs/bio.c b/fs/bio.c index 24c914043532..1486b19fc431 100644 --- a/fs/bio.c +++ b/fs/bio.c @@ -238,7 +238,7 @@ void bio_free(struct bio *bio, struct bio_set *bs) bvec_free_bs(bs, bio->bi_io_vec, BIO_POOL_IDX(bio)); if (bio_integrity(bio)) - bio_integrity_free(bio); + bio_integrity_free(bio, bs); /* * If we have front padding, adjust the bio pointer before freeing @@ -341,7 +341,7 @@ struct bio *bio_alloc(gfp_t gfp_mask, int nr_iovecs) static void bio_kmalloc_destructor(struct bio *bio) { if (bio_integrity(bio)) - bio_integrity_free(bio); + bio_integrity_free(bio, fs_bio_set); kfree(bio); } @@ -472,7 +472,7 @@ struct bio *bio_clone(struct bio *bio, gfp_t gfp_mask) if (bio_integrity(bio)) { int ret; - ret = bio_integrity_clone(b, bio, gfp_mask); + ret = bio_integrity_clone(b, bio, gfp_mask, fs_bio_set); if (ret < 0) { bio_put(b); @@ -1539,6 +1539,7 @@ void bioset_free(struct bio_set *bs) if (bs->bio_pool) mempool_destroy(bs->bio_pool); + bioset_integrity_free(bs); biovec_free_pools(bs); bio_put_slab(bs); @@ -1579,6 +1580,9 @@ struct bio_set *bioset_create(unsigned int pool_size, unsigned int front_pad) if (!bs->bio_pool) goto bad; + if (bioset_integrity_create(bs, pool_size)) + goto bad; + if (!biovec_create_pools(bs, pool_size)) return bs; @@ -1616,6 +1620,7 @@ static int __init init_bio(void) if (!bio_slabs) panic("bio: can't allocate bios\n"); + bio_integrity_init(); biovec_init_slabs(); fs_bio_set = bioset_create(BIO_POOL_SIZE, 0); diff --git a/include/linux/bio.h b/include/linux/bio.h index 2a04eb54c0dd..2892b710771c 100644 --- a/include/linux/bio.h +++ b/include/linux/bio.h @@ -319,7 +319,6 @@ static inline int bio_has_allocated_vec(struct bio *bio) */ struct bio_integrity_payload { struct bio *bip_bio; /* parent bio */ - struct bio_vec *bip_vec; /* integrity data vector */ sector_t bip_sector; /* virtual start sector */ @@ -328,11 +327,12 @@ struct bio_integrity_payload { unsigned int bip_size; - unsigned short bip_pool; /* pool the ivec came from */ + unsigned short bip_slab; /* slab the bip came from */ unsigned short bip_vcnt; /* # of integrity bio_vecs */ unsigned short bip_idx; /* current bip_vec index */ struct work_struct bip_work; /* I/O completion */ + struct bio_vec bip_vec[0]; /* embedded bvec array */ }; #endif /* CONFIG_BLK_DEV_INTEGRITY */ @@ -430,6 +430,9 @@ struct bio_set { unsigned int front_pad; mempool_t *bio_pool; +#if defined(CONFIG_BLK_DEV_INTEGRITY) + mempool_t *bio_integrity_pool; +#endif mempool_t *bvec_pool; }; @@ -634,8 +637,9 @@ static inline struct bio *bio_list_get(struct bio_list *bl) #define bio_integrity(bio) (bio->bi_integrity != NULL) +extern struct bio_integrity_payload *bio_integrity_alloc_bioset(struct bio *, gfp_t, unsigned int, struct bio_set *); extern struct bio_integrity_payload *bio_integrity_alloc(struct bio *, gfp_t, unsigned int); -extern void bio_integrity_free(struct bio *); +extern void bio_integrity_free(struct bio *, struct bio_set *); extern int bio_integrity_add_page(struct bio *, struct page *, unsigned int, unsigned int); extern int bio_integrity_enabled(struct bio *bio); extern int bio_integrity_set_tag(struct bio *, void *, unsigned int); @@ -645,21 +649,27 @@ extern void bio_integrity_endio(struct bio *, int); extern void bio_integrity_advance(struct bio *, unsigned int); extern void bio_integrity_trim(struct bio *, unsigned int, unsigned int); extern void bio_integrity_split(struct bio *, struct bio_pair *, int); -extern int bio_integrity_clone(struct bio *, struct bio *, gfp_t); +extern int bio_integrity_clone(struct bio *, struct bio *, gfp_t, struct bio_set *); +extern int bioset_integrity_create(struct bio_set *, int); +extern void bioset_integrity_free(struct bio_set *); +extern void bio_integrity_init(void); #else /* CONFIG_BLK_DEV_INTEGRITY */ #define bio_integrity(a) (0) +#define bioset_integrity_create(a, b) (0) #define bio_integrity_prep(a) (0) #define bio_integrity_enabled(a) (0) -#define bio_integrity_clone(a, b, c) (0) -#define bio_integrity_free(a) do { } while (0) +#define bio_integrity_clone(a, b, c, d) (0) +#define bioset_integrity_free(a) do { } while (0) +#define bio_integrity_free(a, b) do { } while (0) #define bio_integrity_endio(a, b) do { } while (0) #define bio_integrity_advance(a, b) do { } while (0) #define bio_integrity_trim(a, b, c) do { } while (0) #define bio_integrity_split(a, b, c) do { } while (0) #define bio_integrity_set_tag(a, b, c) do { } while (0) #define bio_integrity_get_tag(a, b, c) do { } while (0) +#define bio_integrity_init(a) do { } while (0) #endif /* CONFIG_BLK_DEV_INTEGRITY */ -- cgit v1.2.3 From e2dbe12557d85d81f4527879499f55681c3cca4f Mon Sep 17 00:00:00 2001 From: Amerigo Wang Date: Wed, 1 Jul 2009 01:06:26 -0400 Subject: elf: fix one check-after-use Check before use it. Signed-off-by: WANG Cong Cc: Alexander Viro Cc: David Howells Acked-by: Roland McGrath Acked-by: James Morris Signed-off-by: Linus Torvalds --- fs/binfmt_elf.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/binfmt_elf.c b/fs/binfmt_elf.c index f1867900e459..b7c1603cd4bd 100644 --- a/fs/binfmt_elf.c +++ b/fs/binfmt_elf.c @@ -1522,11 +1522,11 @@ static int fill_note_info(struct elfhdr *elf, int phdrs, info->thread = NULL; psinfo = kmalloc(sizeof(*psinfo), GFP_KERNEL); - fill_note(&info->psinfo, "CORE", NT_PRPSINFO, sizeof(*psinfo), psinfo); - if (psinfo == NULL) return 0; + fill_note(&info->psinfo, "CORE", NT_PRPSINFO, sizeof(*psinfo), psinfo); + /* * Figure out how many notes we're going to need for each thread. */ -- cgit v1.2.3 From cc0bad7552308e8905d6ea56e6b7811fa67e716d Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Thu, 25 Jun 2009 00:56:52 -0400 Subject: cifs: add new cifs_iget function and convert unix codepath to use it cifs: add new cifs_iget function and convert unix codepath to use it In order to unify some codepaths, introduce a common cifs_fattr struct for storing inode attributes. The different codepaths (unix, legacy, normal, etc...) can fill out this struct with inode info. It can then be passed as an arg to a common set of routines to get and update inodes. Add a new cifs_iget function that uses iget5_locked to identify inodes. This will compare inodes based on the uniqueid value in a cifs_fattr struct. Rather than filling out an already-created inode, have cifs_get_inode_info_unix instead fill out cifs_fattr and hand that off to cifs_iget. cifs_iget can then properly look for hardlinked inodes. On the readdir side, add a new cifs_readdir_lookup function that spawns populated dentries. Redefine FILE_UNIX_INFO so that it's basically a FILE_UNIX_BASIC_INFO that has a few fields wrapped around it. This allows us to more easily use the same function for filling out the fattr as the non-readdir codepath. With this, we should then have proper hardlink detection and can eventually get rid of some nasty CIFS-specific hacks for handing them. Signed-off-by: Jeff Layton Reviewed-by: Christoph Hellwig Signed-off-by: Steve French --- fs/cifs/cifsfs.h | 13 ++ fs/cifs/cifsglob.h | 25 ++++ fs/cifs/cifspdu.h | 14 +- fs/cifs/cifsproto.h | 9 +- fs/cifs/dir.c | 22 +-- fs/cifs/inode.c | 380 ++++++++++++++++++++++++++-------------------------- fs/cifs/readdir.c | 253 +++++++++++++--------------------- 7 files changed, 339 insertions(+), 377 deletions(-) (limited to 'fs') diff --git a/fs/cifs/cifsfs.h b/fs/cifs/cifsfs.h index 9570a0e8023f..586df24c9abb 100644 --- a/fs/cifs/cifsfs.h +++ b/fs/cifs/cifsfs.h @@ -24,6 +24,19 @@ #define ROOT_I 2 +/* + * ino_t is 32-bits on 32-bit arch. We have to squash the 64-bit value down + * so that it will fit. + */ +static inline ino_t +cifs_uniqueid_to_ino_t(u64 fileid) +{ + ino_t ino = (ino_t) fileid; + if (sizeof(ino_t) < sizeof(u64)) + ino ^= fileid >> (sizeof(u64)-sizeof(ino_t)) * 8; + return ino; +} + extern struct file_system_type cifs_fs_type; extern const struct address_space_operations cifs_addr_ops; extern const struct address_space_operations cifs_addr_ops_smallbuf; diff --git a/fs/cifs/cifsglob.h b/fs/cifs/cifsglob.h index e1225e6ded2f..e6435cba8113 100644 --- a/fs/cifs/cifsglob.h +++ b/fs/cifs/cifsglob.h @@ -371,6 +371,7 @@ struct cifsInodeInfo { bool oplockPending:1; bool delete_pending:1; /* DELETE_ON_CLOSE is set */ u64 server_eof; /* current file size on server */ + u64 uniqueid; /* server inode number */ struct inode vfs_inode; }; @@ -472,6 +473,30 @@ struct dfs_info3_param { char *node_name; }; +/* + * common struct for holding inode info when searching for or updating an + * inode with new info + */ + +#define CIFS_FATTR_DFS_REFERRAL 0x1 + +struct cifs_fattr { + u32 cf_flags; + u32 cf_cifsattrs; + u64 cf_uniqueid; + u64 cf_eof; + u64 cf_bytes; + uid_t cf_uid; + gid_t cf_gid; + umode_t cf_mode; + dev_t cf_rdev; + unsigned int cf_nlink; + unsigned int cf_dtype; + struct timespec cf_atime; + struct timespec cf_mtime; + struct timespec cf_ctime; +}; + static inline void free_dfs_info_param(struct dfs_info3_param *param) { if (param) { diff --git a/fs/cifs/cifspdu.h b/fs/cifs/cifspdu.h index a785f69dbc9f..2d07f890a842 100644 --- a/fs/cifs/cifspdu.h +++ b/fs/cifs/cifspdu.h @@ -2328,19 +2328,7 @@ struct file_attrib_tag { typedef struct { __le32 NextEntryOffset; __u32 ResumeKey; /* as with FileIndex - no need to convert */ - __le64 EndOfFile; - __le64 NumOfBytes; - __le64 LastStatusChange; /*SNIA specs DCE time for the 3 time fields */ - __le64 LastAccessTime; - __le64 LastModificationTime; - __le64 Uid; - __le64 Gid; - __le32 Type; - __le64 DevMajor; - __le64 DevMinor; - __le64 UniqueId; - __le64 Permissions; - __le64 Nlinks; + FILE_UNIX_BASIC_INFO basic; char FileName[1]; } __attribute__((packed)) FILE_UNIX_INFO; /* level 0x202 */ diff --git a/fs/cifs/cifsproto.h b/fs/cifs/cifsproto.h index c419416a42ee..b2bd83fd2aa4 100644 --- a/fs/cifs/cifsproto.h +++ b/fs/cifs/cifsproto.h @@ -98,9 +98,14 @@ extern struct timespec cnvrtDosUnixTm(__le16 le_date, __le16 le_time, extern int cifs_posix_open(char *full_path, struct inode **pinode, struct super_block *sb, int mode, int oflags, int *poplock, __u16 *pnetfid, int xid); -extern void posix_fill_in_inode(struct inode *tmp_inode, - FILE_UNIX_BASIC_INFO *pData, int isNewInode); +extern void cifs_unix_basic_to_fattr(struct cifs_fattr *fattr, + FILE_UNIX_BASIC_INFO *info, + struct cifs_sb_info *cifs_sb); +extern void cifs_fattr_to_inode(struct inode *inode, struct cifs_fattr *fattr); extern struct inode *cifs_new_inode(struct super_block *sb, __u64 *inum); +extern struct inode *cifs_iget(struct super_block *sb, + struct cifs_fattr *fattr); + extern int cifs_get_inode_info(struct inode **pinode, const unsigned char *search_path, FILE_ALL_INFO *pfile_info, diff --git a/fs/cifs/dir.c b/fs/cifs/dir.c index 7dc6b74f9def..a40054faed7f 100644 --- a/fs/cifs/dir.c +++ b/fs/cifs/dir.c @@ -188,6 +188,7 @@ int cifs_posix_open(char *full_path, struct inode **pinode, FILE_UNIX_BASIC_INFO *presp_data; __u32 posix_flags = 0; struct cifs_sb_info *cifs_sb = CIFS_SB(sb); + struct cifs_fattr fattr; cFYI(1, ("posix open %s", full_path)); @@ -236,22 +237,21 @@ int cifs_posix_open(char *full_path, struct inode **pinode, if (presp_data->Type == cpu_to_le32(-1)) goto posix_open_ret; /* open ok, caller does qpathinfo */ - /* get new inode and set it up */ if (!pinode) goto posix_open_ret; /* caller does not need info */ + cifs_unix_basic_to_fattr(&fattr, presp_data, cifs_sb); + + /* get new inode and set it up */ if (*pinode == NULL) { - __u64 unique_id = le64_to_cpu(presp_data->UniqueId); - *pinode = cifs_new_inode(sb, &unique_id); + *pinode = cifs_iget(sb, &fattr); + if (!*pinode) { + rc = -ENOMEM; + goto posix_open_ret; + } + } else { + cifs_fattr_to_inode(*pinode, &fattr); } - /* else an inode was passed in. Update its info, don't create one */ - - /* We do not need to close the file if new_inode fails since - the caller will retry qpathinfo as long as inode is null */ - if (*pinode == NULL) - goto posix_open_ret; - - posix_fill_in_inode(*pinode, presp_data, 1); cifs_fill_fileinfo(*pinode, *pnetfid, cifs_sb->tcon, write_only); diff --git a/fs/cifs/inode.c b/fs/cifs/inode.c index 155c9e785d0c..b22379610d71 100644 --- a/fs/cifs/inode.c +++ b/fs/cifs/inode.c @@ -77,127 +77,146 @@ static void cifs_set_ops(struct inode *inode, const bool is_dfs_referral) } } -static void cifs_unix_info_to_inode(struct inode *inode, - FILE_UNIX_BASIC_INFO *info, int force_uid_gid) +/* populate an inode with info from a cifs_fattr struct */ +void +cifs_fattr_to_inode(struct inode *inode, struct cifs_fattr *fattr) { - struct cifs_sb_info *cifs_sb = CIFS_SB(inode->i_sb); - struct cifsInodeInfo *cifsInfo = CIFS_I(inode); - __u64 num_of_bytes = le64_to_cpu(info->NumOfBytes); - __u64 end_of_file = le64_to_cpu(info->EndOfFile); + struct cifsInodeInfo *cifs_i = CIFS_I(inode); + unsigned long now = jiffies; + + inode->i_atime = fattr->cf_atime; + inode->i_mtime = fattr->cf_mtime; + inode->i_ctime = fattr->cf_ctime; + inode->i_mode = fattr->cf_mode; + inode->i_rdev = fattr->cf_rdev; + inode->i_nlink = fattr->cf_nlink; + inode->i_uid = fattr->cf_uid; + inode->i_gid = fattr->cf_gid; + + cifs_i->cifsAttrs = fattr->cf_cifsattrs; + cifs_i->uniqueid = fattr->cf_uniqueid; + + cFYI(1, ("inode 0x%p old_time=%ld new_time=%ld", inode, + cifs_i->time, now)); + cifs_i->time = now; + + /* + * Can't safely change the file size here if the client is writing to + * it due to potential races. + */ + spin_lock(&inode->i_lock); + if (is_size_safe_to_change(cifs_i, fattr->cf_eof)) { + i_size_write(inode, fattr->cf_eof); + + /* + * i_blocks is not related to (i_size / i_blksize), + * but instead 512 byte (2**9) size is required for + * calculating num blocks. + */ + inode->i_blocks = (512 - 1 + fattr->cf_bytes) >> 9; + } + spin_unlock(&inode->i_lock); + + cifs_set_ops(inode, fattr->cf_flags & CIFS_FATTR_DFS_REFERRAL); +} + +/* Fill a cifs_fattr struct with info from FILE_UNIX_BASIC_INFO. */ +void +cifs_unix_basic_to_fattr(struct cifs_fattr *fattr, FILE_UNIX_BASIC_INFO *info, + struct cifs_sb_info *cifs_sb) +{ + memset(fattr, 0, sizeof(*fattr)); + fattr->cf_uniqueid = le64_to_cpu(info->UniqueId); + fattr->cf_bytes = le64_to_cpu(info->NumOfBytes); + fattr->cf_eof = le64_to_cpu(info->EndOfFile); - inode->i_atime = cifs_NTtimeToUnix(info->LastAccessTime); - inode->i_mtime = - cifs_NTtimeToUnix(info->LastModificationTime); - inode->i_ctime = cifs_NTtimeToUnix(info->LastStatusChange); - inode->i_mode = le64_to_cpu(info->Permissions); + fattr->cf_atime = cifs_NTtimeToUnix(info->LastAccessTime); + fattr->cf_mtime = cifs_NTtimeToUnix(info->LastModificationTime); + fattr->cf_ctime = cifs_NTtimeToUnix(info->LastStatusChange); + fattr->cf_mode = le64_to_cpu(info->Permissions); /* * Since we set the inode type below we need to mask off * to avoid strange results if bits set above. */ - inode->i_mode &= ~S_IFMT; + fattr->cf_mode &= ~S_IFMT; switch (le32_to_cpu(info->Type)) { case UNIX_FILE: - inode->i_mode |= S_IFREG; + fattr->cf_mode |= S_IFREG; + fattr->cf_dtype = DT_REG; break; case UNIX_SYMLINK: - inode->i_mode |= S_IFLNK; + fattr->cf_mode |= S_IFLNK; + fattr->cf_dtype = DT_LNK; break; case UNIX_DIR: - inode->i_mode |= S_IFDIR; + fattr->cf_mode |= S_IFDIR; + fattr->cf_dtype = DT_DIR; break; case UNIX_CHARDEV: - inode->i_mode |= S_IFCHR; - inode->i_rdev = MKDEV(le64_to_cpu(info->DevMajor), - le64_to_cpu(info->DevMinor) & MINORMASK); + fattr->cf_mode |= S_IFCHR; + fattr->cf_dtype = DT_CHR; + fattr->cf_rdev = MKDEV(le64_to_cpu(info->DevMajor), + le64_to_cpu(info->DevMinor) & MINORMASK); break; case UNIX_BLOCKDEV: - inode->i_mode |= S_IFBLK; - inode->i_rdev = MKDEV(le64_to_cpu(info->DevMajor), - le64_to_cpu(info->DevMinor) & MINORMASK); + fattr->cf_mode |= S_IFBLK; + fattr->cf_dtype = DT_BLK; + fattr->cf_rdev = MKDEV(le64_to_cpu(info->DevMajor), + le64_to_cpu(info->DevMinor) & MINORMASK); break; case UNIX_FIFO: - inode->i_mode |= S_IFIFO; + fattr->cf_mode |= S_IFIFO; + fattr->cf_dtype = DT_FIFO; break; case UNIX_SOCKET: - inode->i_mode |= S_IFSOCK; + fattr->cf_mode |= S_IFSOCK; + fattr->cf_dtype = DT_SOCK; break; default: /* safest to call it a file if we do not know */ - inode->i_mode |= S_IFREG; + fattr->cf_mode |= S_IFREG; + fattr->cf_dtype = DT_REG; cFYI(1, ("unknown type %d", le32_to_cpu(info->Type))); break; } - if ((cifs_sb->mnt_cifs_flags & CIFS_MOUNT_OVERR_UID) && - !force_uid_gid) - inode->i_uid = cifs_sb->mnt_uid; + if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_OVERR_UID) + fattr->cf_uid = cifs_sb->mnt_uid; else - inode->i_uid = le64_to_cpu(info->Uid); + fattr->cf_uid = le64_to_cpu(info->Uid); - if ((cifs_sb->mnt_cifs_flags & CIFS_MOUNT_OVERR_GID) && - !force_uid_gid) - inode->i_gid = cifs_sb->mnt_gid; + if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_OVERR_GID) + fattr->cf_gid = cifs_sb->mnt_gid; else - inode->i_gid = le64_to_cpu(info->Gid); + fattr->cf_gid = le64_to_cpu(info->Gid); - inode->i_nlink = le64_to_cpu(info->Nlinks); - - cifsInfo->server_eof = end_of_file; - spin_lock(&inode->i_lock); - if (is_size_safe_to_change(cifsInfo, end_of_file)) { - /* - * We can not safely change the file size here if the client - * is writing to it due to potential races. - */ - i_size_write(inode, end_of_file); - - /* - * i_blocks is not related to (i_size / i_blksize), - * but instead 512 byte (2**9) size is required for - * calculating num blocks. - */ - inode->i_blocks = (512 - 1 + num_of_bytes) >> 9; - } - spin_unlock(&inode->i_lock); + fattr->cf_nlink = le64_to_cpu(info->Nlinks); } - /* - * Needed to setup inode data for the directory which is the - * junction to the new submount (ie to setup the fake directory - * which represents a DFS referral) + * Fill a cifs_fattr struct with fake inode info. + * + * Needed to setup cifs_fattr data for the directory which is the + * junction to the new submount (ie to setup the fake directory + * which represents a DFS referral). */ -static void fill_fake_finddataunix(FILE_UNIX_BASIC_INFO *pfnd_dat, - struct super_block *sb) +void +cifs_create_dfs_fattr(struct cifs_fattr *fattr, struct super_block *sb) { - struct inode *pinode = NULL; - - memset(pfnd_dat, 0, sizeof(FILE_UNIX_BASIC_INFO)); + struct cifs_sb_info *cifs_sb = CIFS_SB(sb); -/* __le64 pfnd_dat->EndOfFile = cpu_to_le64(0); - __le64 pfnd_dat->NumOfBytes = cpu_to_le64(0); - __u64 UniqueId = 0; */ - pfnd_dat->LastStatusChange = - cpu_to_le64(cifs_UnixTimeToNT(CURRENT_TIME)); - pfnd_dat->LastAccessTime = - cpu_to_le64(cifs_UnixTimeToNT(CURRENT_TIME)); - pfnd_dat->LastModificationTime = - cpu_to_le64(cifs_UnixTimeToNT(CURRENT_TIME)); - pfnd_dat->Type = cpu_to_le32(UNIX_DIR); - pfnd_dat->Permissions = cpu_to_le64(S_IXUGO | S_IRWXU); - pfnd_dat->Nlinks = cpu_to_le64(2); - if (sb->s_root) - pinode = sb->s_root->d_inode; - if (pinode == NULL) - return; - - /* fill in default values for the remaining based on root - inode since we can not query the server for this inode info */ - pfnd_dat->DevMajor = cpu_to_le64(MAJOR(pinode->i_rdev)); - pfnd_dat->DevMinor = cpu_to_le64(MINOR(pinode->i_rdev)); - pfnd_dat->Uid = cpu_to_le64(pinode->i_uid); - pfnd_dat->Gid = cpu_to_le64(pinode->i_gid); + cFYI(1, ("creating fake fattr for DFS referral")); + + memset(fattr, 0, sizeof(*fattr)); + fattr->cf_mode = S_IFDIR | S_IXUGO | S_IRWXU; + fattr->cf_uid = cifs_sb->mnt_uid; + fattr->cf_gid = cifs_sb->mnt_gid; + fattr->cf_atime = CURRENT_TIME; + fattr->cf_ctime = CURRENT_TIME; + fattr->cf_mtime = CURRENT_TIME; + fattr->cf_nlink = 2; + fattr->cf_flags |= CIFS_FATTR_DFS_REFERRAL; } /** @@ -244,66 +263,42 @@ cifs_new_inode(struct super_block *sb, __u64 *inum) } int cifs_get_inode_info_unix(struct inode **pinode, - const unsigned char *full_path, struct super_block *sb, int xid) + const unsigned char *full_path, + struct super_block *sb, int xid) { - int rc = 0; + int rc; FILE_UNIX_BASIC_INFO find_data; - struct cifsTconInfo *pTcon; - struct inode *inode; + struct cifs_fattr fattr; + struct cifsTconInfo *tcon; struct cifs_sb_info *cifs_sb = CIFS_SB(sb); - bool is_dfs_referral = false; - struct cifsInodeInfo *cifsInfo; - __u64 num_of_bytes; - __u64 end_of_file; - pTcon = cifs_sb->tcon; + tcon = cifs_sb->tcon; cFYI(1, ("Getting info on %s", full_path)); /* could have done a find first instead but this returns more info */ - rc = CIFSSMBUnixQPathInfo(xid, pTcon, full_path, &find_data, + rc = CIFSSMBUnixQPathInfo(xid, tcon, full_path, &find_data, cifs_sb->local_nls, cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR); - if (rc == -EREMOTE && !is_dfs_referral) { - is_dfs_referral = true; - cFYI(DBG2, ("DFS ref")); - /* for DFS, server does not give us real inode data */ - fill_fake_finddataunix(&find_data, sb); - rc = 0; - } else if (rc) - goto cgiiu_exit; - num_of_bytes = le64_to_cpu(find_data.NumOfBytes); - end_of_file = le64_to_cpu(find_data.EndOfFile); + if (!rc) { + cifs_unix_basic_to_fattr(&fattr, &find_data, cifs_sb); + } else if (rc == -EREMOTE) { + cifs_create_dfs_fattr(&fattr, sb); + rc = 0; + } else { + return rc; + } - /* get new inode */ if (*pinode == NULL) { - __u64 unique_id = le64_to_cpu(find_data.UniqueId); - *pinode = cifs_new_inode(sb, &unique_id); - if (*pinode == NULL) { + /* get new inode */ + *pinode = cifs_iget(sb, &fattr); + if (!*pinode) rc = -ENOMEM; - goto cgiiu_exit; - } + } else { + /* we already have inode, update it */ + cifs_fattr_to_inode(*pinode, &fattr); } - inode = *pinode; - cifsInfo = CIFS_I(inode); - - cFYI(1, ("Old time %ld", cifsInfo->time)); - cifsInfo->time = jiffies; - cFYI(1, ("New time %ld", cifsInfo->time)); - /* this is ok to set on every inode revalidate */ - atomic_set(&cifsInfo->inUse, 1); - - cifs_unix_info_to_inode(inode, &find_data, 0); - - if (num_of_bytes < end_of_file) - cFYI(1, ("allocation size less than end of file")); - cFYI(1, ("Size %ld and blocks %llu", - (unsigned long) inode->i_size, - (unsigned long long)inode->i_blocks)); - - cifs_set_ops(inode, is_dfs_referral); -cgiiu_exit: return rc; } @@ -695,33 +690,85 @@ char *cifs_build_path_to_root(struct cifs_sb_info *cifs_sb) return full_path; } +static int +cifs_find_inode(struct inode *inode, void *opaque) +{ + struct cifs_fattr *fattr = (struct cifs_fattr *) opaque; + + if (CIFS_I(inode)->uniqueid != fattr->cf_uniqueid) + return 0; + + return 1; +} + +static int +cifs_init_inode(struct inode *inode, void *opaque) +{ + struct cifs_fattr *fattr = (struct cifs_fattr *) opaque; + + CIFS_I(inode)->uniqueid = fattr->cf_uniqueid; + return 0; +} + +/* Given fattrs, get a corresponding inode */ +struct inode * +cifs_iget(struct super_block *sb, struct cifs_fattr *fattr) +{ + unsigned long hash; + struct inode *inode; + + cFYI(1, ("looking for uniqueid=%llu", fattr->cf_uniqueid)); + + /* hash down to 32-bits on 32-bit arch */ + hash = cifs_uniqueid_to_ino_t(fattr->cf_uniqueid); + + inode = iget5_locked(sb, hash, cifs_find_inode, cifs_init_inode, fattr); + + /* we have fattrs in hand, update the inode */ + if (inode) { + cifs_fattr_to_inode(inode, fattr); + if (sb->s_flags & MS_NOATIME) + inode->i_flags |= S_NOATIME | S_NOCMTIME; + if (inode->i_state & I_NEW) { + inode->i_ino = hash; + unlock_new_inode(inode); + } + } + + return inode; +} + /* gets root inode */ struct inode *cifs_root_iget(struct super_block *sb, unsigned long ino) { int xid; struct cifs_sb_info *cifs_sb; - struct inode *inode; + struct inode *inode = NULL; long rc; char *full_path; - inode = iget_locked(sb, ino); - if (!inode) - return ERR_PTR(-ENOMEM); - if (!(inode->i_state & I_NEW)) - return inode; - - cifs_sb = CIFS_SB(inode->i_sb); + cifs_sb = CIFS_SB(sb); full_path = cifs_build_path_to_root(cifs_sb); if (full_path == NULL) return ERR_PTR(-ENOMEM); xid = GetXid(); - if (cifs_sb->tcon->unix_ext) - rc = cifs_get_inode_info_unix(&inode, full_path, inode->i_sb, - xid); - else + if (cifs_sb->tcon->unix_ext) { + rc = cifs_get_inode_info_unix(&inode, full_path, sb, xid); + if (!inode) + return ERR_PTR(-ENOMEM); + } else { + inode = iget_locked(sb, ino); + if (!inode) + return ERR_PTR(-ENOMEM); + if (!(inode->i_state & I_NEW)) + return inode; + rc = cifs_get_inode_info(&inode, full_path, NULL, inode->i_sb, xid, NULL); + unlock_new_inode(inode); + } + if (rc && cifs_sb->tcon->ipc) { cFYI(1, ("ipc connection - fake read inode")); inode->i_mode |= S_IFDIR; @@ -737,7 +784,6 @@ struct inode *cifs_root_iget(struct super_block *sb, unsigned long ino) return ERR_PTR(rc); } - unlock_new_inode(inode); kfree(full_path); /* can not call macro FreeXid here since in a void func @@ -1063,44 +1109,6 @@ out_reval: return rc; } -void posix_fill_in_inode(struct inode *tmp_inode, - FILE_UNIX_BASIC_INFO *pData, int isNewInode) -{ - struct cifsInodeInfo *cifsInfo = CIFS_I(tmp_inode); - loff_t local_size; - struct timespec local_mtime; - - cifsInfo->time = jiffies; - atomic_inc(&cifsInfo->inUse); - - /* save mtime and size */ - local_mtime = tmp_inode->i_mtime; - local_size = tmp_inode->i_size; - - cifs_unix_info_to_inode(tmp_inode, pData, 1); - cifs_set_ops(tmp_inode, false); - - if (!S_ISREG(tmp_inode->i_mode)) - return; - - /* - * No sense invalidating pages for new inode - * since we we have not started caching - * readahead file data yet. - */ - if (isNewInode) - return; - - if (timespec_equal(&tmp_inode->i_mtime, &local_mtime) && - (local_size == tmp_inode->i_size)) { - cFYI(1, ("inode exists but unchanged")); - } else { - /* file may have changed on server */ - cFYI(1, ("invalidate inode, readdir detected change")); - invalidate_remote_inode(tmp_inode); - } -} - int cifs_mkdir(struct inode *inode, struct dentry *direntry, int mode) { int rc = 0, tmprc; @@ -1109,6 +1117,7 @@ int cifs_mkdir(struct inode *inode, struct dentry *direntry, int mode) struct cifsTconInfo *pTcon; char *full_path = NULL; struct inode *newinode = NULL; + struct cifs_fattr fattr; cFYI(1, ("In cifs_mkdir, mode = 0x%x inode = 0x%p", mode, inode)); @@ -1148,7 +1157,6 @@ int cifs_mkdir(struct inode *inode, struct dentry *direntry, int mode) cFYI(1, ("posix mkdir returned 0x%x", rc)); d_drop(direntry); } else { - __u64 unique_id; if (pInfo->Type == cpu_to_le32(-1)) { /* no return info, go query for it */ kfree(pInfo); @@ -1162,20 +1170,15 @@ int cifs_mkdir(struct inode *inode, struct dentry *direntry, int mode) else direntry->d_op = &cifs_dentry_ops; - unique_id = le64_to_cpu(pInfo->UniqueId); - newinode = cifs_new_inode(inode->i_sb, &unique_id); - if (newinode == NULL) { + cifs_unix_basic_to_fattr(&fattr, pInfo, cifs_sb); + newinode = cifs_iget(inode->i_sb, &fattr); + if (!newinode) { kfree(pInfo); goto mkdir_get_info; } - newinode->i_nlink = 2; d_instantiate(direntry, newinode); - /* we already checked in POSIXCreate whether - frame was long enough */ - posix_fill_in_inode(direntry->d_inode, - pInfo, 1 /* NewInode */); #ifdef CONFIG_CIFS_DEBUG2 cFYI(1, ("instantiated dentry %p %s to inode %p", direntry, direntry->d_name.name, newinode)); @@ -1622,6 +1625,7 @@ int cifs_getattr(struct vfsmount *mnt, struct dentry *dentry, if (!err) { generic_fillattr(dentry->d_inode, stat); stat->blksize = CIFS_MAX_MSGSIZE; + stat->ino = CIFS_I(dentry->d_inode)->uniqueid; } return err; } diff --git a/fs/cifs/readdir.c b/fs/cifs/readdir.c index 86d0055dc529..231aa6953f83 100644 --- a/fs/cifs/readdir.c +++ b/fs/cifs/readdir.c @@ -63,6 +63,55 @@ static inline void dump_cifs_file_struct(struct file *file, char *label) } #endif /* DEBUG2 */ +/* + * Find the dentry that matches "name". If there isn't one, create one. If it's + * a negative dentry or the uniqueid changed, then drop it and recreate it. + */ +static struct dentry * +cifs_readdir_lookup(struct dentry *parent, struct qstr *name, + struct cifs_fattr *fattr) +{ + struct dentry *dentry, *alias; + struct inode *inode; + struct super_block *sb = parent->d_inode->i_sb; + + cFYI(1, ("For %s", name->name)); + + dentry = d_lookup(parent, name); + if (dentry) { + /* FIXME: check for inode number changes? */ + if (dentry->d_inode != NULL) + return dentry; + d_drop(dentry); + dput(dentry); + } + + dentry = d_alloc(parent, name); + if (dentry == NULL) + return NULL; + + inode = cifs_iget(sb, fattr); + if (!inode) { + dput(dentry); + return NULL; + } + + if (CIFS_SB(sb)->tcon->nocase) + dentry->d_op = &cifs_ci_dentry_ops; + else + dentry->d_op = &cifs_dentry_ops; + + alias = d_materialise_unique(dentry, inode); + if (alias != NULL) { + dput(dentry); + if (IS_ERR(alias)) + return NULL; + dentry = alias; + } + + return dentry; +} + /* Returns 1 if new inode created, 2 if both dentry and inode were */ /* Might check in the future if inode number changed so we can rehash inode */ static int @@ -76,7 +125,6 @@ construct_dentry(struct qstr *qstring, struct file *file, cFYI(1, ("For %s", qstring->name)); - qstring->hash = full_name_hash(qstring->name, qstring->len); tmp_dentry = d_lookup(file->f_path.dentry, qstring); if (tmp_dentry) { /* BB: overwrite old name? i.e. tmp_dentry->d_name and @@ -299,140 +347,6 @@ static void fill_in_inode(struct inode *tmp_inode, int new_buf_type, } } -static void unix_fill_in_inode(struct inode *tmp_inode, - FILE_UNIX_INFO *pfindData, unsigned int *pobject_type, int isNewInode) -{ - loff_t local_size; - struct timespec local_mtime; - - struct cifsInodeInfo *cifsInfo = CIFS_I(tmp_inode); - struct cifs_sb_info *cifs_sb = CIFS_SB(tmp_inode->i_sb); - - __u32 type = le32_to_cpu(pfindData->Type); - __u64 num_of_bytes = le64_to_cpu(pfindData->NumOfBytes); - __u64 end_of_file = le64_to_cpu(pfindData->EndOfFile); - cifsInfo->time = jiffies; - atomic_inc(&cifsInfo->inUse); - - /* save mtime and size */ - local_mtime = tmp_inode->i_mtime; - local_size = tmp_inode->i_size; - - tmp_inode->i_atime = - cifs_NTtimeToUnix(pfindData->LastAccessTime); - tmp_inode->i_mtime = - cifs_NTtimeToUnix(pfindData->LastModificationTime); - tmp_inode->i_ctime = - cifs_NTtimeToUnix(pfindData->LastStatusChange); - - tmp_inode->i_mode = le64_to_cpu(pfindData->Permissions); - /* since we set the inode type below we need to mask off type - to avoid strange results if bits above were corrupt */ - tmp_inode->i_mode &= ~S_IFMT; - if (type == UNIX_FILE) { - *pobject_type = DT_REG; - tmp_inode->i_mode |= S_IFREG; - } else if (type == UNIX_SYMLINK) { - *pobject_type = DT_LNK; - tmp_inode->i_mode |= S_IFLNK; - } else if (type == UNIX_DIR) { - *pobject_type = DT_DIR; - tmp_inode->i_mode |= S_IFDIR; - } else if (type == UNIX_CHARDEV) { - *pobject_type = DT_CHR; - tmp_inode->i_mode |= S_IFCHR; - tmp_inode->i_rdev = MKDEV(le64_to_cpu(pfindData->DevMajor), - le64_to_cpu(pfindData->DevMinor) & MINORMASK); - } else if (type == UNIX_BLOCKDEV) { - *pobject_type = DT_BLK; - tmp_inode->i_mode |= S_IFBLK; - tmp_inode->i_rdev = MKDEV(le64_to_cpu(pfindData->DevMajor), - le64_to_cpu(pfindData->DevMinor) & MINORMASK); - } else if (type == UNIX_FIFO) { - *pobject_type = DT_FIFO; - tmp_inode->i_mode |= S_IFIFO; - } else if (type == UNIX_SOCKET) { - *pobject_type = DT_SOCK; - tmp_inode->i_mode |= S_IFSOCK; - } else { - /* safest to just call it a file */ - *pobject_type = DT_REG; - tmp_inode->i_mode |= S_IFREG; - cFYI(1, ("unknown inode type %d", type)); - } - - if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_OVERR_UID) - tmp_inode->i_uid = cifs_sb->mnt_uid; - else - tmp_inode->i_uid = le64_to_cpu(pfindData->Uid); - if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_OVERR_GID) - tmp_inode->i_gid = cifs_sb->mnt_gid; - else - tmp_inode->i_gid = le64_to_cpu(pfindData->Gid); - tmp_inode->i_nlink = le64_to_cpu(pfindData->Nlinks); - - cifsInfo->server_eof = end_of_file; - spin_lock(&tmp_inode->i_lock); - if (is_size_safe_to_change(cifsInfo, end_of_file)) { - /* can not safely change the file size here if the - client is writing to it due to potential races */ - i_size_write(tmp_inode, end_of_file); - - /* 512 bytes (2**9) is the fake blocksize that must be used */ - /* for this calculation, not the real blocksize */ - tmp_inode->i_blocks = (512 - 1 + num_of_bytes) >> 9; - } - spin_unlock(&tmp_inode->i_lock); - - if (S_ISREG(tmp_inode->i_mode)) { - cFYI(1, ("File inode")); - tmp_inode->i_op = &cifs_file_inode_ops; - - if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_DIRECT_IO) { - if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NO_BRL) - tmp_inode->i_fop = &cifs_file_direct_nobrl_ops; - else - tmp_inode->i_fop = &cifs_file_direct_ops; - } else if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NO_BRL) - tmp_inode->i_fop = &cifs_file_nobrl_ops; - else - tmp_inode->i_fop = &cifs_file_ops; - - if ((cifs_sb->tcon) && (cifs_sb->tcon->ses) && - (cifs_sb->tcon->ses->server->maxBuf < - PAGE_CACHE_SIZE + MAX_CIFS_HDR_SIZE)) - tmp_inode->i_data.a_ops = &cifs_addr_ops_smallbuf; - else - tmp_inode->i_data.a_ops = &cifs_addr_ops; - - if (isNewInode) - return; /* No sense invalidating pages for new inode - since we have not started caching readahead - file data for it yet */ - - if (timespec_equal(&tmp_inode->i_mtime, &local_mtime) && - (local_size == tmp_inode->i_size)) { - cFYI(1, ("inode exists but unchanged")); - } else { - /* file may have changed on server */ - cFYI(1, ("invalidate inode, readdir detected change")); - invalidate_remote_inode(tmp_inode); - } - } else if (S_ISDIR(tmp_inode->i_mode)) { - cFYI(1, ("Directory inode")); - tmp_inode->i_op = &cifs_dir_inode_ops; - tmp_inode->i_fop = &cifs_dir_ops; - } else if (S_ISLNK(tmp_inode->i_mode)) { - cFYI(1, ("Symbolic Link inode")); - tmp_inode->i_op = &cifs_symlink_inode_ops; -/* tmp_inode->i_fop = *//* do not need to set to anything */ - } else { - cFYI(1, ("Special inode")); - init_special_inode(tmp_inode, tmp_inode->i_mode, - tmp_inode->i_rdev); - } -} - /* BB eventually need to add the following helper function to resolve NT_STATUS_STOPPED_ON_SYMLINK return code when we try to do FindFirst on (NTFS) directory symlinks */ @@ -872,7 +786,7 @@ static int cifs_get_name_from_search_buf(struct qstr *pqst, len = strnlen(filename, PATH_MAX); } - *pinum = le64_to_cpu(pFindData->UniqueId); + *pinum = le64_to_cpu(pFindData->basic.UniqueId); } else if (level == SMB_FIND_FILE_DIRECTORY_INFO) { FILE_DIRECTORY_INFO *pFindData = (FILE_DIRECTORY_INFO *)current_entry; @@ -934,9 +848,11 @@ static int cifs_filldir(char *pfindEntry, struct file *file, filldir_t filldir, struct cifsFileInfo *pCifsF; unsigned int obj_type; __u64 inum; + ino_t ino; struct cifs_sb_info *cifs_sb; struct inode *tmp_inode; struct dentry *tmp_dentry; + struct cifs_fattr fattr; /* get filename and len into qstring */ /* get dentry */ @@ -967,39 +883,49 @@ static int cifs_filldir(char *pfindEntry, struct file *file, filldir_t filldir, return rc; /* only these two infolevels return valid inode numbers */ - if (pCifsF->srch_inf.info_level == SMB_FIND_FILE_UNIX || - pCifsF->srch_inf.info_level == SMB_FIND_FILE_ID_FULL_DIR_INFO) - rc = construct_dentry(&qstring, file, &tmp_inode, &tmp_dentry, - &inum); - else - rc = construct_dentry(&qstring, file, &tmp_inode, &tmp_dentry, - NULL); + if (pCifsF->srch_inf.info_level == SMB_FIND_FILE_UNIX) { + cifs_unix_basic_to_fattr(&fattr, + &((FILE_UNIX_INFO *) pfindEntry)->basic, + cifs_sb); + tmp_dentry = cifs_readdir_lookup(file->f_dentry, &qstring, + &fattr); + obj_type = fattr.cf_dtype; + ino = cifs_uniqueid_to_ino_t(fattr.cf_uniqueid); + } else { + if (pCifsF->srch_inf.info_level == + SMB_FIND_FILE_ID_FULL_DIR_INFO) + rc = construct_dentry(&qstring, file, &tmp_inode, + &tmp_dentry, &inum); + else + rc = construct_dentry(&qstring, file, &tmp_inode, + &tmp_dentry, NULL); - if ((tmp_inode == NULL) || (tmp_dentry == NULL)) - return -ENOMEM; + if ((tmp_inode == NULL) || (tmp_dentry == NULL)) { + rc = -ENOMEM; + goto out; + } - /* we pass in rc below, indicating whether it is a new inode, - so we can figure out whether to invalidate the inode cached - data if the file has changed */ - if (pCifsF->srch_inf.info_level == SMB_FIND_FILE_UNIX) - unix_fill_in_inode(tmp_inode, - (FILE_UNIX_INFO *)pfindEntry, - &obj_type, rc); - else if (pCifsF->srch_inf.info_level == SMB_FIND_FILE_INFO_STANDARD) - fill_in_inode(tmp_inode, 0 /* old level 1 buffer type */, - pfindEntry, &obj_type, rc); - else - fill_in_inode(tmp_inode, 1 /* NT */, pfindEntry, &obj_type, rc); + /* we pass in rc below, indicating whether it is a new inode, + * so we can figure out whether to invalidate the inode cached + * data if the file has changed + */ + if (pCifsF->srch_inf.info_level == SMB_FIND_FILE_INFO_STANDARD) + fill_in_inode(tmp_inode, 0, pfindEntry, &obj_type, rc); + else + fill_in_inode(tmp_inode, 1, pfindEntry, &obj_type, rc); - if (rc) /* new inode - needs to be tied to dentry */ { - d_instantiate(tmp_dentry, tmp_inode); - if (rc == 2) - d_rehash(tmp_dentry); - } + /* new inode - needs to be tied to dentry */ + if (rc) { + d_instantiate(tmp_dentry, tmp_inode); + if (rc == 2) + d_rehash(tmp_dentry); + } + ino = cifs_uniqueid_to_ino_t(tmp_inode->i_ino); + } rc = filldir(direntry, qstring.name, qstring.len, file->f_pos, - tmp_inode->i_ino, obj_type); + ino, obj_type); if (rc) { cFYI(1, ("filldir rc = %d", rc)); /* we can not return filldir errors to the caller @@ -1008,6 +934,7 @@ static int cifs_filldir(char *pfindEntry, struct file *file, filldir_t filldir, rc = -EOVERFLOW; } +out: dput(tmp_dentry); return rc; } -- cgit v1.2.3 From bdae997f44535ac4ebe1e055ffe59eeee946f453 Mon Sep 17 00:00:00 2001 From: Keith Packard Date: Wed, 1 Jul 2009 21:56:38 -0700 Subject: fs/notify/inotify: decrement user inotify count on close The per-user inotify_devs value is incremented each time a new file is allocated, but never decremented. This led to inotify_init failing after a limited number of calls. Signed-off-by: Keith Packard Signed-off-by: Eric Paris --- fs/notify/inotify/inotify_user.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'fs') diff --git a/fs/notify/inotify/inotify_user.c b/fs/notify/inotify/inotify_user.c index ff231ad23895..ff27a2965844 100644 --- a/fs/notify/inotify/inotify_user.c +++ b/fs/notify/inotify/inotify_user.c @@ -296,12 +296,15 @@ static int inotify_fasync(int fd, struct file *file, int on) static int inotify_release(struct inode *ignored, struct file *file) { struct fsnotify_group *group = file->private_data; + struct user_struct *user = group->inotify_data.user; fsnotify_clear_marks_by_group(group); /* free this group, matching get was inotify_init->fsnotify_obtain_group */ fsnotify_put_group(group); + atomic_dec(&user->inotify_devs); + return 0; } -- cgit v1.2.3 From f597bb19ccd034cbcf05e1194238e2c8d9505a8a Mon Sep 17 00:00:00 2001 From: Chris Mason Date: Sat, 27 Jun 2009 21:06:22 -0400 Subject: Btrfs: don't log the inode in file_write while growing the file --- fs/btrfs/file.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index 126477eaecf5..7c3cd248d8d6 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -151,7 +151,10 @@ static noinline int dirty_and_release_pages(struct btrfs_trans_handle *trans, } if (end_pos > isize) { i_size_write(inode, end_pos); - btrfs_update_inode(trans, root, inode); + /* we've only changed i_size in ram, and we haven't updated + * the disk i_size. There is no need to log the inode + * at this time. + */ } err = btrfs_end_transaction(trans, root); out_unlock: -- cgit v1.2.3 From c8a894d77de4a1e0a544577fd4eabc9aacd453a8 Mon Sep 17 00:00:00 2001 From: Chris Mason Date: Sat, 27 Jun 2009 21:07:03 -0400 Subject: Btrfs: fix the file clone ioctl for preallocated extents --- fs/btrfs/ioctl.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index eff18f5b5362..9f4db848db10 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -1028,7 +1028,8 @@ static long btrfs_ioctl_clone(struct file *file, unsigned long srcfd, struct btrfs_file_extent_item); comp = btrfs_file_extent_compression(leaf, extent); type = btrfs_file_extent_type(leaf, extent); - if (type == BTRFS_FILE_EXTENT_REG) { + if (type == BTRFS_FILE_EXTENT_REG || + type == BTRFS_FILE_EXTENT_PREALLOC) { disko = btrfs_file_extent_disk_bytenr(leaf, extent); diskl = btrfs_file_extent_disk_num_bytes(leaf, @@ -1051,7 +1052,8 @@ static long btrfs_ioctl_clone(struct file *file, unsigned long srcfd, new_key.objectid = inode->i_ino; new_key.offset = key.offset + destoff - off; - if (type == BTRFS_FILE_EXTENT_REG) { + if (type == BTRFS_FILE_EXTENT_REG || + type == BTRFS_FILE_EXTENT_PREALLOC) { ret = btrfs_insert_empty_item(trans, root, path, &new_key, size); if (ret) -- cgit v1.2.3 From a970b0a16cc416a509d5ae8b1d70978664e6f4fe Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Sat, 27 Jun 2009 21:07:34 -0400 Subject: Btrfs: account for space we may use in fallocate Using Eric Sandeen's xfstest for fallocate, you can easily trigger a ENOSPC panic on btrfs. This is because we do not account for data we may use when doing the fallocate. This patch fixes the problem by properly reserving space, and then just freeing it when we are done. The reservation stuff was made with delalloc in mind, so its a little crude for this case, but it keeps the box from panicing. Signed-off-by: Josef Bacik Signed-off-by: Chris Mason --- fs/btrfs/inode.c | 12 +++++++++++- 1 file changed, 11 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 5b68330f8585..1eacc78f6614 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -5103,6 +5103,7 @@ static long btrfs_fallocate(struct inode *inode, int mode, u64 mask = BTRFS_I(inode)->root->sectorsize - 1; struct extent_map *em; struct btrfs_trans_handle *trans; + struct btrfs_root *root; int ret; alloc_start = offset & ~mask; @@ -5121,6 +5122,13 @@ static long btrfs_fallocate(struct inode *inode, int mode, goto out; } + root = BTRFS_I(inode)->root; + + ret = btrfs_check_data_free_space(root, inode, + alloc_end - alloc_start); + if (ret) + goto out; + locked_end = alloc_end - 1; while (1) { struct btrfs_ordered_extent *ordered; @@ -5128,7 +5136,7 @@ static long btrfs_fallocate(struct inode *inode, int mode, trans = btrfs_start_transaction(BTRFS_I(inode)->root, 1); if (!trans) { ret = -EIO; - goto out; + goto out_free; } /* the extent lock is ordered inside the running @@ -5189,6 +5197,8 @@ static long btrfs_fallocate(struct inode *inode, int mode, GFP_NOFS); btrfs_end_transaction(trans, BTRFS_I(inode)->root); +out_free: + btrfs_free_reserved_data_space(root, inode, alloc_end - alloc_start); out: mutex_unlock(&inode->i_mutex); return ret; -- cgit v1.2.3 From 2c47e605a91dde6b0514f689645e7ab336c8592a Mon Sep 17 00:00:00 2001 From: Yan Zheng Date: Sat, 27 Jun 2009 21:07:35 -0400 Subject: Btrfs: update backrefs while dropping snapshot The new backref format has restriction on type of backref item. If a tree block isn't referenced by its owner tree, full backrefs must be used for the pointers in it. When a tree block loses its owner tree's reference, backrefs for the pointers in it should be updated to full backrefs. Current btrfs_drop_snapshot misses the code that updates backrefs, so it's unsafe for general use. This patch adds backrefs update code to btrfs_drop_snapshot. It isn't a problem in the restricted form btrfs_drop_snapshot is used today, but for general snapshot deletion this update is required. Signed-off-by: Yan Zheng Signed-off-by: Chris Mason --- fs/btrfs/ctree.h | 3 +- fs/btrfs/extent-tree.c | 564 ++++++++++++++++++++++++++++++++++--------------- fs/btrfs/relocation.c | 5 +- fs/btrfs/transaction.c | 4 +- 4 files changed, 395 insertions(+), 181 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 03441a99ea38..a404ecc53eb1 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -2076,8 +2076,7 @@ static inline int btrfs_insert_empty_item(struct btrfs_trans_handle *trans, int btrfs_next_leaf(struct btrfs_root *root, struct btrfs_path *path); int btrfs_prev_leaf(struct btrfs_root *root, struct btrfs_path *path); int btrfs_leaf_free_space(struct btrfs_root *root, struct extent_buffer *leaf); -int btrfs_drop_snapshot(struct btrfs_trans_handle *trans, struct btrfs_root - *root); +int btrfs_drop_snapshot(struct btrfs_root *root, int update_ref); int btrfs_drop_subtree(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct extent_buffer *node, diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index edc7d208c5ce..cd64cfc14f7f 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -990,15 +990,13 @@ static inline int extent_ref_type(u64 parent, u64 owner) return type; } -static int find_next_key(struct btrfs_path *path, struct btrfs_key *key) +static int find_next_key(struct btrfs_path *path, int level, + struct btrfs_key *key) { - int level; - BUG_ON(!path->keep_locks); - for (level = 0; level < BTRFS_MAX_LEVEL; level++) { + for (; level < BTRFS_MAX_LEVEL; level++) { if (!path->nodes[level]) break; - btrfs_assert_tree_locked(path->nodes[level]); if (path->slots[level] + 1 >= btrfs_header_nritems(path->nodes[level])) continue; @@ -1158,7 +1156,8 @@ int lookup_inline_extent_backref(struct btrfs_trans_handle *trans, * For simplicity, we just do not add new inline back * ref if there is any kind of item for this block */ - if (find_next_key(path, &key) == 0 && key.objectid == bytenr && + if (find_next_key(path, 0, &key) == 0 && + key.objectid == bytenr && key.type < BTRFS_BLOCK_GROUP_ITEM_KEY) { err = -EAGAIN; goto out; @@ -4128,6 +4127,7 @@ struct extent_buffer *btrfs_alloc_free_block(struct btrfs_trans_handle *trans, return buf; } +#if 0 int btrfs_drop_leaf_ref(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct extent_buffer *leaf) { @@ -4171,8 +4171,6 @@ int btrfs_drop_leaf_ref(struct btrfs_trans_handle *trans, return 0; } -#if 0 - static noinline int cache_drop_leaf_ref(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct btrfs_leaf_ref *ref) @@ -4553,262 +4551,471 @@ out: } #endif +struct walk_control { + u64 refs[BTRFS_MAX_LEVEL]; + u64 flags[BTRFS_MAX_LEVEL]; + struct btrfs_key update_progress; + int stage; + int level; + int shared_level; + int update_ref; + int keep_locks; +}; + +#define DROP_REFERENCE 1 +#define UPDATE_BACKREF 2 + /* - * helper function for drop_subtree, this function is similar to - * walk_down_tree. The main difference is that it checks reference - * counts while tree blocks are locked. + * hepler to process tree block while walking down the tree. + * + * when wc->stage == DROP_REFERENCE, this function checks + * reference count of the block. if the block is shared and + * we need update back refs for the subtree rooted at the + * block, this function changes wc->stage to UPDATE_BACKREF + * + * when wc->stage == UPDATE_BACKREF, this function updates + * back refs for pointers in the block. + * + * NOTE: return value 1 means we should stop walking down. */ -static noinline int walk_down_tree(struct btrfs_trans_handle *trans, +static noinline int walk_down_proc(struct btrfs_trans_handle *trans, struct btrfs_root *root, - struct btrfs_path *path, int *level) + struct btrfs_path *path, + struct walk_control *wc) { - struct extent_buffer *next; - struct extent_buffer *cur; - struct extent_buffer *parent; - u64 bytenr; - u64 ptr_gen; - u64 refs; - u64 flags; - u32 blocksize; + int level = wc->level; + struct extent_buffer *eb = path->nodes[level]; + struct btrfs_key key; + u64 flag = BTRFS_BLOCK_FLAG_FULL_BACKREF; int ret; - cur = path->nodes[*level]; - ret = btrfs_lookup_extent_info(trans, root, cur->start, cur->len, - &refs, &flags); - BUG_ON(ret); - if (refs > 1) - goto out; + if (wc->stage == UPDATE_BACKREF && + btrfs_header_owner(eb) != root->root_key.objectid) + return 1; - BUG_ON(!(flags & BTRFS_BLOCK_FLAG_FULL_BACKREF)); + /* + * when reference count of tree block is 1, it won't increase + * again. once full backref flag is set, we never clear it. + */ + if ((wc->stage == DROP_REFERENCE && wc->refs[level] != 1) || + (wc->stage == UPDATE_BACKREF && !(wc->flags[level] & flag))) { + BUG_ON(!path->locks[level]); + ret = btrfs_lookup_extent_info(trans, root, + eb->start, eb->len, + &wc->refs[level], + &wc->flags[level]); + BUG_ON(ret); + BUG_ON(wc->refs[level] == 0); + } - while (*level >= 0) { - cur = path->nodes[*level]; - if (*level == 0) { - ret = btrfs_drop_leaf_ref(trans, root, cur); - BUG_ON(ret); - clean_tree_block(trans, root, cur); - break; - } - if (path->slots[*level] >= btrfs_header_nritems(cur)) { - clean_tree_block(trans, root, cur); - break; + if (wc->stage == DROP_REFERENCE && + wc->update_ref && wc->refs[level] > 1) { + BUG_ON(eb == root->node); + BUG_ON(path->slots[level] > 0); + if (level == 0) + btrfs_item_key_to_cpu(eb, &key, path->slots[level]); + else + btrfs_node_key_to_cpu(eb, &key, path->slots[level]); + if (btrfs_header_owner(eb) == root->root_key.objectid && + btrfs_comp_cpu_keys(&key, &wc->update_progress) >= 0) { + wc->stage = UPDATE_BACKREF; + wc->shared_level = level; } + } - bytenr = btrfs_node_blockptr(cur, path->slots[*level]); - blocksize = btrfs_level_size(root, *level - 1); - ptr_gen = btrfs_node_ptr_generation(cur, path->slots[*level]); + if (wc->stage == DROP_REFERENCE) { + if (wc->refs[level] > 1) + return 1; - next = read_tree_block(root, bytenr, blocksize, ptr_gen); - btrfs_tree_lock(next); - btrfs_set_lock_blocking(next); + if (path->locks[level] && !wc->keep_locks) { + btrfs_tree_unlock(eb); + path->locks[level] = 0; + } + return 0; + } - ret = btrfs_lookup_extent_info(trans, root, bytenr, blocksize, - &refs, &flags); + /* wc->stage == UPDATE_BACKREF */ + if (!(wc->flags[level] & flag)) { + BUG_ON(!path->locks[level]); + ret = btrfs_inc_ref(trans, root, eb, 1); BUG_ON(ret); - if (refs > 1) { - parent = path->nodes[*level]; - ret = btrfs_free_extent(trans, root, bytenr, - blocksize, parent->start, - btrfs_header_owner(parent), - *level - 1, 0); + ret = btrfs_dec_ref(trans, root, eb, 0); + BUG_ON(ret); + ret = btrfs_set_disk_extent_flags(trans, root, eb->start, + eb->len, flag, 0); + BUG_ON(ret); + wc->flags[level] |= flag; + } + + /* + * the block is shared by multiple trees, so it's not good to + * keep the tree lock + */ + if (path->locks[level] && level > 0) { + btrfs_tree_unlock(eb); + path->locks[level] = 0; + } + return 0; +} + +/* + * hepler to process tree block while walking up the tree. + * + * when wc->stage == DROP_REFERENCE, this function drops + * reference count on the block. + * + * when wc->stage == UPDATE_BACKREF, this function changes + * wc->stage back to DROP_REFERENCE if we changed wc->stage + * to UPDATE_BACKREF previously while processing the block. + * + * NOTE: return value 1 means we should stop walking up. + */ +static noinline int walk_up_proc(struct btrfs_trans_handle *trans, + struct btrfs_root *root, + struct btrfs_path *path, + struct walk_control *wc) +{ + int ret = 0; + int level = wc->level; + struct extent_buffer *eb = path->nodes[level]; + u64 parent = 0; + + if (wc->stage == UPDATE_BACKREF) { + BUG_ON(wc->shared_level < level); + if (level < wc->shared_level) + goto out; + + BUG_ON(wc->refs[level] <= 1); + ret = find_next_key(path, level + 1, &wc->update_progress); + if (ret > 0) + wc->update_ref = 0; + + wc->stage = DROP_REFERENCE; + wc->shared_level = -1; + path->slots[level] = 0; + + /* + * check reference count again if the block isn't locked. + * we should start walking down the tree again if reference + * count is one. + */ + if (!path->locks[level]) { + BUG_ON(level == 0); + btrfs_tree_lock(eb); + btrfs_set_lock_blocking(eb); + path->locks[level] = 1; + + ret = btrfs_lookup_extent_info(trans, root, + eb->start, eb->len, + &wc->refs[level], + &wc->flags[level]); BUG_ON(ret); - path->slots[*level]++; - btrfs_tree_unlock(next); - free_extent_buffer(next); - continue; + BUG_ON(wc->refs[level] == 0); + if (wc->refs[level] == 1) { + btrfs_tree_unlock(eb); + path->locks[level] = 0; + return 1; + } + } else { + BUG_ON(level != 0); } + } - BUG_ON(!(flags & BTRFS_BLOCK_FLAG_FULL_BACKREF)); + /* wc->stage == DROP_REFERENCE */ + BUG_ON(wc->refs[level] > 1 && !path->locks[level]); - *level = btrfs_header_level(next); - path->nodes[*level] = next; - path->slots[*level] = 0; - path->locks[*level] = 1; - cond_resched(); + if (wc->refs[level] == 1) { + if (level == 0) { + if (wc->flags[level] & BTRFS_BLOCK_FLAG_FULL_BACKREF) + ret = btrfs_dec_ref(trans, root, eb, 1); + else + ret = btrfs_dec_ref(trans, root, eb, 0); + BUG_ON(ret); + } + /* make block locked assertion in clean_tree_block happy */ + if (!path->locks[level] && + btrfs_header_generation(eb) == trans->transid) { + btrfs_tree_lock(eb); + btrfs_set_lock_blocking(eb); + path->locks[level] = 1; + } + clean_tree_block(trans, root, eb); + } + + if (eb == root->node) { + if (wc->flags[level] & BTRFS_BLOCK_FLAG_FULL_BACKREF) + parent = eb->start; + else + BUG_ON(root->root_key.objectid != + btrfs_header_owner(eb)); + } else { + if (wc->flags[level + 1] & BTRFS_BLOCK_FLAG_FULL_BACKREF) + parent = path->nodes[level + 1]->start; + else + BUG_ON(root->root_key.objectid != + btrfs_header_owner(path->nodes[level + 1])); } -out: - if (path->nodes[*level] == root->node) - parent = path->nodes[*level]; - else - parent = path->nodes[*level + 1]; - bytenr = path->nodes[*level]->start; - blocksize = path->nodes[*level]->len; - ret = btrfs_free_extent(trans, root, bytenr, blocksize, parent->start, - btrfs_header_owner(parent), *level, 0); + ret = btrfs_free_extent(trans, root, eb->start, eb->len, parent, + root->root_key.objectid, level, 0); BUG_ON(ret); +out: + wc->refs[level] = 0; + wc->flags[level] = 0; + return ret; +} + +static noinline int walk_down_tree(struct btrfs_trans_handle *trans, + struct btrfs_root *root, + struct btrfs_path *path, + struct walk_control *wc) +{ + struct extent_buffer *next; + struct extent_buffer *cur; + u64 bytenr; + u64 ptr_gen; + u32 blocksize; + int level = wc->level; + int ret; + + while (level >= 0) { + cur = path->nodes[level]; + BUG_ON(path->slots[level] >= btrfs_header_nritems(cur)); - if (path->locks[*level]) { - btrfs_tree_unlock(path->nodes[*level]); - path->locks[*level] = 0; + ret = walk_down_proc(trans, root, path, wc); + if (ret > 0) + break; + + if (level == 0) + break; + + bytenr = btrfs_node_blockptr(cur, path->slots[level]); + blocksize = btrfs_level_size(root, level - 1); + ptr_gen = btrfs_node_ptr_generation(cur, path->slots[level]); + + next = read_tree_block(root, bytenr, blocksize, ptr_gen); + btrfs_tree_lock(next); + btrfs_set_lock_blocking(next); + + level--; + BUG_ON(level != btrfs_header_level(next)); + path->nodes[level] = next; + path->slots[level] = 0; + path->locks[level] = 1; + wc->level = level; } - free_extent_buffer(path->nodes[*level]); - path->nodes[*level] = NULL; - *level += 1; - cond_resched(); return 0; } -/* - * helper for dropping snapshots. This walks back up the tree in the path - * to find the first node higher up where we haven't yet gone through - * all the slots - */ static noinline int walk_up_tree(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct btrfs_path *path, - int *level, int max_level) + struct walk_control *wc, int max_level) { - struct btrfs_root_item *root_item = &root->root_item; - int i; - int slot; + int level = wc->level; int ret; - for (i = *level; i < max_level && path->nodes[i]; i++) { - slot = path->slots[i]; - if (slot + 1 < btrfs_header_nritems(path->nodes[i])) { - /* - * there is more work to do in this level. - * Update the drop_progress marker to reflect - * the work we've done so far, and then bump - * the slot number - */ - path->slots[i]++; - WARN_ON(*level == 0); - if (max_level == BTRFS_MAX_LEVEL) { - btrfs_node_key(path->nodes[i], - &root_item->drop_progress, - path->slots[i]); - root_item->drop_level = i; - } - *level = i; + path->slots[level] = btrfs_header_nritems(path->nodes[level]); + while (level < max_level && path->nodes[level]) { + wc->level = level; + if (path->slots[level] + 1 < + btrfs_header_nritems(path->nodes[level])) { + path->slots[level]++; return 0; } else { - struct extent_buffer *parent; - - /* - * this whole node is done, free our reference - * on it and go up one level - */ - if (path->nodes[*level] == root->node) - parent = path->nodes[*level]; - else - parent = path->nodes[*level + 1]; + ret = walk_up_proc(trans, root, path, wc); + if (ret > 0) + return 0; - clean_tree_block(trans, root, path->nodes[i]); - ret = btrfs_free_extent(trans, root, - path->nodes[i]->start, - path->nodes[i]->len, - parent->start, - btrfs_header_owner(parent), - *level, 0); - BUG_ON(ret); - if (path->locks[*level]) { - btrfs_tree_unlock(path->nodes[i]); - path->locks[i] = 0; + if (path->locks[level]) { + btrfs_tree_unlock(path->nodes[level]); + path->locks[level] = 0; } - free_extent_buffer(path->nodes[i]); - path->nodes[i] = NULL; - *level = i + 1; + free_extent_buffer(path->nodes[level]); + path->nodes[level] = NULL; + level++; } } return 1; } /* - * drop the reference count on the tree rooted at 'snap'. This traverses - * the tree freeing any blocks that have a ref count of zero after being - * decremented. + * drop a subvolume tree. + * + * this function traverses the tree freeing any blocks that only + * referenced by the tree. + * + * when a shared tree block is found. this function decreases its + * reference count by one. if update_ref is true, this function + * also make sure backrefs for the shared block and all lower level + * blocks are properly updated. */ -int btrfs_drop_snapshot(struct btrfs_trans_handle *trans, struct btrfs_root - *root) +int btrfs_drop_snapshot(struct btrfs_root *root, int update_ref) { - int ret = 0; - int wret; - int level; struct btrfs_path *path; - int update_count; + struct btrfs_trans_handle *trans; + struct btrfs_root *tree_root = root->fs_info->tree_root; struct btrfs_root_item *root_item = &root->root_item; + struct walk_control *wc; + struct btrfs_key key; + int err = 0; + int ret; + int level; path = btrfs_alloc_path(); BUG_ON(!path); - level = btrfs_header_level(root->node); + wc = kzalloc(sizeof(*wc), GFP_NOFS); + BUG_ON(!wc); + + trans = btrfs_start_transaction(tree_root, 1); + if (btrfs_disk_key_objectid(&root_item->drop_progress) == 0) { + level = btrfs_header_level(root->node); path->nodes[level] = btrfs_lock_root_node(root); btrfs_set_lock_blocking(path->nodes[level]); path->slots[level] = 0; path->locks[level] = 1; + memset(&wc->update_progress, 0, + sizeof(wc->update_progress)); } else { - struct btrfs_key key; - struct btrfs_disk_key found_key; - struct extent_buffer *node; - btrfs_disk_key_to_cpu(&key, &root_item->drop_progress); + memcpy(&wc->update_progress, &key, + sizeof(wc->update_progress)); + level = root_item->drop_level; + BUG_ON(level == 0); path->lowest_level = level; - wret = btrfs_search_slot(NULL, root, &key, path, 0, 0); - if (wret < 0) { - ret = wret; + ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); + path->lowest_level = 0; + if (ret < 0) { + err = ret; goto out; } - node = path->nodes[level]; - btrfs_node_key(node, &found_key, path->slots[level]); - WARN_ON(memcmp(&found_key, &root_item->drop_progress, - sizeof(found_key))); + btrfs_node_key_to_cpu(path->nodes[level], &key, + path->slots[level]); + WARN_ON(memcmp(&key, &wc->update_progress, sizeof(key))); + /* * unlock our path, this is safe because only this * function is allowed to delete this snapshot */ btrfs_unlock_up_safe(path, 0); + + level = btrfs_header_level(root->node); + while (1) { + btrfs_tree_lock(path->nodes[level]); + btrfs_set_lock_blocking(path->nodes[level]); + + ret = btrfs_lookup_extent_info(trans, root, + path->nodes[level]->start, + path->nodes[level]->len, + &wc->refs[level], + &wc->flags[level]); + BUG_ON(ret); + BUG_ON(wc->refs[level] == 0); + + if (level == root_item->drop_level) + break; + + btrfs_tree_unlock(path->nodes[level]); + WARN_ON(wc->refs[level] != 1); + level--; + } } + + wc->level = level; + wc->shared_level = -1; + wc->stage = DROP_REFERENCE; + wc->update_ref = update_ref; + wc->keep_locks = 0; + while (1) { - unsigned long update; - wret = walk_down_tree(trans, root, path, &level); - if (wret > 0) + ret = walk_down_tree(trans, root, path, wc); + if (ret < 0) { + err = ret; break; - if (wret < 0) - ret = wret; + } - wret = walk_up_tree(trans, root, path, &level, - BTRFS_MAX_LEVEL); - if (wret > 0) + ret = walk_up_tree(trans, root, path, wc, BTRFS_MAX_LEVEL); + if (ret < 0) { + err = ret; break; - if (wret < 0) - ret = wret; - if (trans->transaction->in_commit || - trans->transaction->delayed_refs.flushing) { - ret = -EAGAIN; + } + + if (ret > 0) { + BUG_ON(wc->stage != DROP_REFERENCE); break; } - for (update_count = 0; update_count < 16; update_count++) { + + if (wc->stage == DROP_REFERENCE) { + level = wc->level; + btrfs_node_key(path->nodes[level], + &root_item->drop_progress, + path->slots[level]); + root_item->drop_level = level; + } + + BUG_ON(wc->level == 0); + if (trans->transaction->in_commit || + trans->transaction->delayed_refs.flushing) { + ret = btrfs_update_root(trans, tree_root, + &root->root_key, + root_item); + BUG_ON(ret); + + btrfs_end_transaction(trans, tree_root); + trans = btrfs_start_transaction(tree_root, 1); + } else { + unsigned long update; update = trans->delayed_ref_updates; trans->delayed_ref_updates = 0; if (update) - btrfs_run_delayed_refs(trans, root, update); - else - break; + btrfs_run_delayed_refs(trans, tree_root, + update); } } + btrfs_release_path(root, path); + BUG_ON(err); + + ret = btrfs_del_root(trans, tree_root, &root->root_key); + BUG_ON(ret); + + free_extent_buffer(root->node); + free_extent_buffer(root->commit_root); + kfree(root); out: + btrfs_end_transaction(trans, tree_root); + kfree(wc); btrfs_free_path(path); - return ret; + return err; } +/* + * drop subtree rooted at tree block 'node'. + * + * NOTE: this function will unlock and release tree block 'node' + */ int btrfs_drop_subtree(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct extent_buffer *node, struct extent_buffer *parent) { struct btrfs_path *path; + struct walk_control *wc; int level; int parent_level; int ret = 0; int wret; + BUG_ON(root->root_key.objectid != BTRFS_TREE_RELOC_OBJECTID); + path = btrfs_alloc_path(); BUG_ON(!path); + wc = kzalloc(sizeof(*wc), GFP_NOFS); + BUG_ON(!wc); + btrfs_assert_tree_locked(parent); parent_level = btrfs_header_level(parent); extent_buffer_get(parent); @@ -4817,24 +5024,33 @@ int btrfs_drop_subtree(struct btrfs_trans_handle *trans, btrfs_assert_tree_locked(node); level = btrfs_header_level(node); - extent_buffer_get(node); path->nodes[level] = node; path->slots[level] = 0; + path->locks[level] = 1; + + wc->refs[parent_level] = 1; + wc->flags[parent_level] = BTRFS_BLOCK_FLAG_FULL_BACKREF; + wc->level = level; + wc->shared_level = -1; + wc->stage = DROP_REFERENCE; + wc->update_ref = 0; + wc->keep_locks = 1; while (1) { - wret = walk_down_tree(trans, root, path, &level); - if (wret < 0) + wret = walk_down_tree(trans, root, path, wc); + if (wret < 0) { ret = wret; - if (wret != 0) break; + } - wret = walk_up_tree(trans, root, path, &level, parent_level); + wret = walk_up_tree(trans, root, path, wc, parent_level); if (wret < 0) ret = wret; if (wret != 0) break; } + kfree(wc); btrfs_free_path(path); return ret; } diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c index b23dc209ae10..008397934778 100644 --- a/fs/btrfs/relocation.c +++ b/fs/btrfs/relocation.c @@ -1788,7 +1788,7 @@ static void merge_func(struct btrfs_work *work) btrfs_end_transaction(trans, root); } - btrfs_drop_dead_root(reloc_root); + btrfs_drop_snapshot(reloc_root, 0); if (atomic_dec_and_test(async->num_pending)) complete(async->done); @@ -2075,9 +2075,6 @@ static int do_relocation(struct btrfs_trans_handle *trans, ret = btrfs_drop_subtree(trans, root, eb, upper->eb); BUG_ON(ret); - - btrfs_tree_unlock(eb); - free_extent_buffer(eb); } if (!lowest) { btrfs_tree_unlock(upper->eb); diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index 4e83457ea253..2dbf1c1f56ee 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -593,6 +593,7 @@ int btrfs_defrag_root(struct btrfs_root *root, int cacheonly) return 0; } +#if 0 /* * when dropping snapshots, we generate a ton of delayed refs, and it makes * sense not to join the transaction while it is trying to flush the current @@ -681,6 +682,7 @@ int btrfs_drop_dead_root(struct btrfs_root *root) btrfs_btree_balance_dirty(tree_root, nr); return ret; } +#endif /* * new snapshots need to be created at a very specific time in the @@ -1081,7 +1083,7 @@ int btrfs_clean_old_snapshots(struct btrfs_root *root) while (!list_empty(&list)) { root = list_entry(list.next, struct btrfs_root, root_list); list_del_init(&root->root_list); - btrfs_drop_dead_root(root); + btrfs_drop_snapshot(root, 0); } return 0; } -- cgit v1.2.3 From 9427216476d4fa75103f39d4b228c47d56ba20da Mon Sep 17 00:00:00 2001 From: Chris Mason Date: Thu, 2 Jul 2009 12:26:06 -0400 Subject: Btrfs: honor nodatacow/sum mount options for new files The btrfs attr patches unconditionally inherited the inode flags field without honoring nodatacow and nodatasum. This fix makes sure we properly record the nodatacow/sum mount options in new inodes. Signed-off-by: Chris Mason --- fs/btrfs/inode.c | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 1eacc78f6614..a48c084f6d3a 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -3587,12 +3587,6 @@ static struct inode *btrfs_new_inode(struct btrfs_trans_handle *trans, owner = 1; BTRFS_I(inode)->block_group = btrfs_find_block_group(root, 0, alloc_hint, owner); - if ((mode & S_IFREG)) { - if (btrfs_test_opt(root, NODATASUM)) - BTRFS_I(inode)->flags |= BTRFS_INODE_NODATASUM; - if (btrfs_test_opt(root, NODATACOW)) - BTRFS_I(inode)->flags |= BTRFS_INODE_NODATACOW; - } key[0].objectid = objectid; btrfs_set_key_type(&key[0], BTRFS_INODE_ITEM_KEY); @@ -3647,6 +3641,13 @@ static struct inode *btrfs_new_inode(struct btrfs_trans_handle *trans, btrfs_inherit_iflags(inode, dir); + if ((mode & S_IFREG)) { + if (btrfs_test_opt(root, NODATASUM)) + BTRFS_I(inode)->flags |= BTRFS_INODE_NODATASUM; + if (btrfs_test_opt(root, NODATACOW)) + BTRFS_I(inode)->flags |= BTRFS_INODE_NODATACOW; + } + insert_inode_hash(inode); inode_tree_add(inode); return inode; -- cgit v1.2.3 From 9b627e9bf49ebfeb060dfae0435bdba06cf27cb8 Mon Sep 17 00:00:00 2001 From: Jiri Slaby Date: Thu, 2 Jul 2009 13:50:58 -0400 Subject: Btrfs: fix use after free in btrfs_start_workers fail path worker memory is already freed on one fail path in btrfs_start_workers, but is still dereferenced. Switch the dereference and kfree. Signed-off-by: Jiri Slaby Signed-off-by: Chris Mason --- fs/btrfs/async-thread.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/async-thread.c b/fs/btrfs/async-thread.c index 7f88628a1a72..6e4f6c50a120 100644 --- a/fs/btrfs/async-thread.c +++ b/fs/btrfs/async-thread.c @@ -299,8 +299,8 @@ int btrfs_start_workers(struct btrfs_workers *workers, int num_workers) "btrfs-%s-%d", workers->name, workers->num_workers + i); if (IS_ERR(worker->task)) { - kfree(worker); ret = PTR_ERR(worker->task); + kfree(worker); goto fail; } -- cgit v1.2.3 From 68f5a38c3ea4ae9cc7a40f86ff6d6d031583d93a Mon Sep 17 00:00:00 2001 From: Hu Tao Date: Thu, 2 Jul 2009 13:55:45 -0400 Subject: Btrfs: fix error message formatting Make an error msg look nicer by inserting a space between number and word. Signed-off-by: Hu Tao Signed-off-by: Jiri Kosina Signed-off-by: Chris Mason --- fs/btrfs/extent-tree.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index cd64cfc14f7f..a5aca3997d42 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -2696,7 +2696,7 @@ again: printk(KERN_ERR "no space left, need %llu, %llu delalloc bytes" ", %llu bytes_used, %llu bytes_reserved, " - "%llu bytes_pinned, %llu bytes_readonly, %llu may use" + "%llu bytes_pinned, %llu bytes_readonly, %llu may use " "%llu total\n", (unsigned long long)bytes, (unsigned long long)data_sinfo->bytes_delalloc, (unsigned long long)data_sinfo->bytes_used, -- cgit v1.2.3 From 033a666ccb842ab4134fcd0c861d5ba9f5d6bf3a Mon Sep 17 00:00:00 2001 From: David Howells Date: Thu, 2 Jul 2009 14:35:32 +0100 Subject: NFSD: Don't hold unrefcounted creds over call to nfsd_setuser() nfsd_open() gets an unrefcounted pointer to the current process's effective credentials at the top of the function, then calls nfsd_setuser() via fh_verify() - which may replace and destroy the current process's effective credentials - and then passes the unrefcounted pointer to dentry_open() - but the credentials may have been destroyed by this point. Instead, the value from current_cred() should be passed directly to dentry_open() as one of its arguments, rather than being cached in a variable. Possibly fh_verify() should return the creds to use. This is a regression introduced by 745ca2475a6ac596e3d8d37c2759c0fbe2586227 "CRED: Pass credentials through dentry_open()". Signed-off-by: David Howells Tested-and-Verified-By: Steve Dickson Cc: stable@kernel.org Signed-off-by: J. Bruce Fields --- fs/nfsd/vfs.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c index 4145083dcf88..23341c1063bc 100644 --- a/fs/nfsd/vfs.c +++ b/fs/nfsd/vfs.c @@ -678,7 +678,6 @@ __be32 nfsd_open(struct svc_rqst *rqstp, struct svc_fh *fhp, int type, int access, struct file **filp) { - const struct cred *cred = current_cred(); struct dentry *dentry; struct inode *inode; int flags = O_RDONLY|O_LARGEFILE; @@ -733,7 +732,7 @@ nfsd_open(struct svc_rqst *rqstp, struct svc_fh *fhp, int type, vfs_dq_init(inode); } *filp = dentry_open(dget(dentry), mntget(fhp->fh_export->ex_path.mnt), - flags, cred); + flags, current_cred()); if (IS_ERR(*filp)) host_err = PTR_ERR(*filp); else -- cgit v1.2.3 From 0cfae3d8795f388f9de78adb0171520d19da77e9 Mon Sep 17 00:00:00 2001 From: Ryusuke Konishi Date: Thu, 18 Jun 2009 11:42:53 +0900 Subject: nilfs2: remove unlikely directive causing mis-conversion of error code The following error code handling in nilfs_segctor_write() function wrongly converted negative error codes to a truth value (i.e. 1): err = unlikely(err) ? : res; which originaly meant to be err = err ? : res; This mis-conversion caused that write or sync functions receive the unexpected error code. This fixes the bug by removing the unlikely directive. Signed-off-by: Ryusuke Konishi Cc: stable@kernel.org --- fs/nilfs2/segment.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/nilfs2/segment.c b/fs/nilfs2/segment.c index aa977549919e..c1824915c1c7 100644 --- a/fs/nilfs2/segment.c +++ b/fs/nilfs2/segment.c @@ -1829,8 +1829,8 @@ static int nilfs_segctor_write(struct nilfs_sc_info *sci, err = nilfs_segbuf_write(segbuf, &wi); res = nilfs_segbuf_wait(segbuf, &wi); - err = unlikely(err) ? : res; - if (unlikely(err)) + err = err ? : res; + if (err) return err; } return 0; -- cgit v1.2.3 From 8227b29722fdbac72357aae155d171a5c777670c Mon Sep 17 00:00:00 2001 From: Ryusuke Konishi Date: Thu, 18 Jun 2009 23:52:23 +0900 Subject: nilfs2: fix hang problem of log writer which occurs after write failures Leandro Lucarella gave me a report that nilfs gets stuck after its write function fails. The problem turned out to be caused by bugs which leave writeback flag on pages. This fixes the problem by ensuring to clear the writeback flag in error path. Reported-by: Leandro Lucarella Signed-off-by: Ryusuke Konishi Cc: stable@kernel.org --- fs/nilfs2/segment.c | 26 ++++++-------------------- 1 file changed, 6 insertions(+), 20 deletions(-) (limited to 'fs') diff --git a/fs/nilfs2/segment.c b/fs/nilfs2/segment.c index c1824915c1c7..8b5e4778cf28 100644 --- a/fs/nilfs2/segment.c +++ b/fs/nilfs2/segment.c @@ -1836,19 +1836,6 @@ static int nilfs_segctor_write(struct nilfs_sc_info *sci, return 0; } -static int nilfs_page_has_uncleared_buffer(struct page *page) -{ - struct buffer_head *head, *bh; - - head = bh = page_buffers(page); - do { - if (buffer_dirty(bh) && !list_empty(&bh->b_assoc_buffers)) - return 1; - bh = bh->b_this_page; - } while (bh != head); - return 0; -} - static void __nilfs_end_page_io(struct page *page, int err) { if (!err) { @@ -1872,12 +1859,11 @@ static void nilfs_end_page_io(struct page *page, int err) if (!page) return; - if (buffer_nilfs_node(page_buffers(page)) && - nilfs_page_has_uncleared_buffer(page)) - /* For b-tree node pages, this function may be called twice - or more because they might be split in a segment. - This check assures that cleanup has been done for all - buffers in a split btnode page. */ + if (buffer_nilfs_node(page_buffers(page)) && !PageWriteback(page)) + /* + * For b-tree node pages, this function may be called twice + * or more because they might be split in a segment. + */ return; __nilfs_end_page_io(page, err); @@ -1940,7 +1926,7 @@ static void nilfs_segctor_abort_write(struct nilfs_sc_info *sci, } if (bh->b_page != fs_page) { nilfs_end_page_io(fs_page, err); - if (unlikely(fs_page == failed_page)) + if (fs_page && fs_page == failed_page) goto done; fs_page = bh->b_page; } -- cgit v1.2.3 From 4a52df779700080de4afb0436d9dd9188514a69b Mon Sep 17 00:00:00 2001 From: Ryusuke Konishi Date: Thu, 18 Jun 2009 23:53:25 +0900 Subject: nilfs2: fix incorrect KERN_CRIT messages in case of write failures In case of write-failure retries, the following KERN_CRIT level messages are mistakenly output by nilfs_dat_commit_start() function: nilfs_dat_commit_start: vbn = 408463, start = 12506, end = 18446744073709551615, pbn = 530210 nilfs_dat_commit_start: vbn = 408515, start = 12506, end = 18446744073709551615, pbn = 530211 nilfs_dat_commit_start: vbn = 408464, start = 12506, end = 18446744073709551615, pbn = 530212 ... This suppresses these messages. Signed-off-by: Ryusuke Konishi Cc: stable@kernel.org --- fs/nilfs2/dat.c | 9 --------- 1 file changed, 9 deletions(-) (limited to 'fs') diff --git a/fs/nilfs2/dat.c b/fs/nilfs2/dat.c index 0b2710e2d565..8927ca27e6f7 100644 --- a/fs/nilfs2/dat.c +++ b/fs/nilfs2/dat.c @@ -134,15 +134,6 @@ void nilfs_dat_commit_start(struct inode *dat, struct nilfs_palloc_req *req, entry = nilfs_palloc_block_get_entry(dat, req->pr_entry_nr, req->pr_entry_bh, kaddr); entry->de_start = cpu_to_le64(nilfs_mdt_cno(dat)); - if (entry->de_blocknr != cpu_to_le64(0) || - entry->de_end != cpu_to_le64(NILFS_CNO_MAX)) { - printk(KERN_CRIT - "%s: vbn = %llu, start = %llu, end = %llu, pbn = %llu\n", - __func__, (unsigned long long)req->pr_entry_nr, - (unsigned long long)le64_to_cpu(entry->de_start), - (unsigned long long)le64_to_cpu(entry->de_end), - (unsigned long long)le64_to_cpu(entry->de_blocknr)); - } entry->de_blocknr = cpu_to_le64(blocknr); kunmap_atomic(kaddr, KM_USER0); -- cgit v1.2.3 From ff54de363afa4583e2a6249f25fe21dfaeb11ea2 Mon Sep 17 00:00:00 2001 From: Ryusuke Konishi Date: Fri, 19 Jun 2009 02:53:56 +0900 Subject: nilfs2: fix lockdep warning between regular file and inode file This will fix the following false positive of recursive locking which lockdep has detected: ============================================= [ INFO: possible recursive locking detected ] 2.6.30-nilfs #42 --------------------------------------------- nilfs_cleanerd/10607 is trying to acquire lock: (&bmap->b_sem){++++-.}, at: [] nilfs_bmap_lookup_at_level+0x1a/0x74 [nilfs2] but task is already holding lock: (&bmap->b_sem){++++-.}, at: [] nilfs_bmap_truncate+0x19/0x6a [nilfs2] other info that might help us debug this: 2 locks held by nilfs_cleanerd/10607: #0: (&nilfs->ns_segctor_sem){++++.+}, at: [] nilfs_transaction_begin+0xb6/0x10c [nilfs2] #1: (&bmap->b_sem){++++-.}, at: [] nilfs_bmap_truncate+0x19/0x6a [nilfs2] Signed-off-by: Ryusuke Konishi --- fs/nilfs2/bmap.c | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'fs') diff --git a/fs/nilfs2/bmap.c b/fs/nilfs2/bmap.c index 36df60b6d8a4..99d58a028b94 100644 --- a/fs/nilfs2/bmap.c +++ b/fs/nilfs2/bmap.c @@ -568,6 +568,7 @@ void nilfs_bmap_abort_update_v(struct nilfs_bmap *bmap, } static struct lock_class_key nilfs_bmap_dat_lock_key; +static struct lock_class_key nilfs_bmap_mdt_lock_key; /** * nilfs_bmap_read - read a bmap from an inode @@ -603,7 +604,11 @@ int nilfs_bmap_read(struct nilfs_bmap *bmap, struct nilfs_inode *raw_inode) bmap->b_ptr_type = NILFS_BMAP_PTR_VS; bmap->b_last_allocated_key = 0; bmap->b_last_allocated_ptr = NILFS_BMAP_INVALID_PTR; + lockdep_set_class(&bmap->b_sem, &nilfs_bmap_mdt_lock_key); break; + case NILFS_IFILE_INO: + lockdep_set_class(&bmap->b_sem, &nilfs_bmap_mdt_lock_key); + /* Fall through */ default: bmap->b_ptr_type = NILFS_BMAP_PTR_VM; bmap->b_last_allocated_key = 0; -- cgit v1.2.3 From d9a0a345ab7a58a30ec38e5bb7401a28714914d2 Mon Sep 17 00:00:00 2001 From: Jiro SEKIBA Date: Sat, 4 Jul 2009 23:00:53 +0900 Subject: nilfs2: fix disorder in cp count on error during deleting checkpoints This fixes a bug that checkpoint count gets wrong on errors when deleting a series of checkpoints. The count error is persistent since the checkpoint count is stored on disk. Some userland programs refer to the count via ioctl, and this bugfix is needed to prevent malfunction of such programs. Signed-off-by: Jiro SEKIBA Signed-off-by: Ryusuke Konishi Cc: stable@kernel.org --- fs/nilfs2/cpfile.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/nilfs2/cpfile.c b/fs/nilfs2/cpfile.c index 7d49813f66d6..aec942cf79e3 100644 --- a/fs/nilfs2/cpfile.c +++ b/fs/nilfs2/cpfile.c @@ -307,7 +307,7 @@ int nilfs_cpfile_delete_checkpoints(struct inode *cpfile, ret = nilfs_cpfile_get_checkpoint_block(cpfile, cno, 0, &cp_bh); if (ret < 0) { if (ret != -ENOENT) - goto out_header; + break; /* skip hole */ ret = 0; continue; @@ -340,7 +340,7 @@ int nilfs_cpfile_delete_checkpoints(struct inode *cpfile, continue; printk(KERN_ERR "%s: cannot delete block\n", __func__); - goto out_header; + break; } } @@ -358,7 +358,6 @@ int nilfs_cpfile_delete_checkpoints(struct inode *cpfile, kunmap_atomic(kaddr, KM_USER0); } - out_header: brelse(header_bh); out_sem: -- cgit v1.2.3 From e3dc5a665d39112e98cfd5bbc7fda2963c00c12c Mon Sep 17 00:00:00 2001 From: Adrian Hunter Date: Mon, 22 Jun 2009 17:31:09 +0300 Subject: UBIFS: fix integer overflow warning Fix the following warning: fs/ubifs/io.c: In function 'ubifs_wbuf_init': fs/ubifs/io.c:860: warning: integer overflow in expression And limit maximum hrtimer delta to ULONG_MAX because the argument is 'unsigned long'. Signed-off-by: Adrian Hunter Signed-off-by: Artem Bityutskiy --- fs/ubifs/io.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ubifs/io.c b/fs/ubifs/io.c index bc5857199ec2..2d41ae1d6607 100644 --- a/fs/ubifs/io.c +++ b/fs/ubifs/io.c @@ -857,7 +857,9 @@ int ubifs_wbuf_init(struct ubifs_info *c, struct ubifs_wbuf *wbuf) * and hard limits. */ hardlimit = ktime_set(DEFAULT_WBUF_TIMEOUT_SECS, 0); - wbuf->delta = (DEFAULT_WBUF_TIMEOUT_SECS * NSEC_PER_SEC) * 2 / 10; + wbuf->delta = DEFAULT_WBUF_TIMEOUT_SECS * 1000000000ULL * 2 / 10; + if (wbuf->delta > ULONG_MAX) + wbuf->delta = ULONG_MAX; wbuf->softlimit = ktime_sub_ns(hardlimit, wbuf->delta); hrtimer_set_expires_range_ns(&wbuf->timer, wbuf->softlimit, wbuf->delta); -- cgit v1.2.3 From 70aee2f153972f70fad5f7025134fec063f9efbe Mon Sep 17 00:00:00 2001 From: Artem Bityutskiy Date: Thu, 18 Jun 2009 13:37:15 +0300 Subject: UBIFS: improve debugging messaged 1. Make the I/O debugging message print the journal head number. 2. Add prints to timer functions. Signed-off-by: Artem Bityutskiy --- fs/ubifs/io.c | 26 ++++++++++++++++---------- 1 file changed, 16 insertions(+), 10 deletions(-) (limited to 'fs') diff --git a/fs/ubifs/io.c b/fs/ubifs/io.c index 2d41ae1d6607..2ef689a9a363 100644 --- a/fs/ubifs/io.c +++ b/fs/ubifs/io.c @@ -297,6 +297,7 @@ static enum hrtimer_restart wbuf_timer_callback_nolock(struct hrtimer *timer) { struct ubifs_wbuf *wbuf = container_of(timer, struct ubifs_wbuf, timer); + dbg_io("jhead %d", wbuf->jhead); wbuf->need_sync = 1; wbuf->c->need_wbuf_sync = 1; ubifs_wake_up_bgt(wbuf->c); @@ -313,6 +314,9 @@ static void new_wbuf_timer_nolock(struct ubifs_wbuf *wbuf) if (!ktime_to_ns(wbuf->softlimit)) return; + dbg_io("set timer for jhead %d, %llu-%llu millisecs", wbuf->jhead, + ktime_to_ns(wbuf->softlimit)/USEC_PER_SEC, + (ktime_to_ns(wbuf->softlimit) + wbuf->delta)/USEC_PER_SEC); hrtimer_start_range_ns(&wbuf->timer, wbuf->softlimit, wbuf->delta, HRTIMER_MODE_REL); } @@ -349,8 +353,8 @@ int ubifs_wbuf_sync_nolock(struct ubifs_wbuf *wbuf) /* Write-buffer is empty or not seeked */ return 0; - dbg_io("LEB %d:%d, %d bytes", - wbuf->lnum, wbuf->offs, wbuf->used); + dbg_io("LEB %d:%d, %d bytes, jhead %d", + wbuf->lnum, wbuf->offs, wbuf->used, wbuf->jhead); ubifs_assert(!(c->vfs_sb->s_flags & MS_RDONLY)); ubifs_assert(!(wbuf->avail & 7)); ubifs_assert(wbuf->offs + c->min_io_size <= c->leb_size); @@ -399,7 +403,7 @@ int ubifs_wbuf_seek_nolock(struct ubifs_wbuf *wbuf, int lnum, int offs, { const struct ubifs_info *c = wbuf->c; - dbg_io("LEB %d:%d", lnum, offs); + dbg_io("LEB %d:%d, jhead %d", lnum, offs, wbuf->jhead); ubifs_assert(lnum >= 0 && lnum < c->leb_cnt); ubifs_assert(offs >= 0 && offs <= c->leb_size); ubifs_assert(offs % c->min_io_size == 0 && !(offs & 7)); @@ -506,9 +510,9 @@ int ubifs_wbuf_write_nolock(struct ubifs_wbuf *wbuf, void *buf, int len) struct ubifs_info *c = wbuf->c; int err, written, n, aligned_len = ALIGN(len, 8), offs; - dbg_io("%d bytes (%s) to wbuf at LEB %d:%d", len, - dbg_ntype(((struct ubifs_ch *)buf)->node_type), wbuf->lnum, - wbuf->offs + wbuf->used); + dbg_io("%d bytes (%s) to jhead %d wbuf at LEB %d:%d", len, + dbg_ntype(((struct ubifs_ch *)buf)->node_type), wbuf->jhead, + wbuf->lnum, wbuf->offs + wbuf->used); ubifs_assert(len > 0 && wbuf->lnum >= 0 && wbuf->lnum < c->leb_cnt); ubifs_assert(wbuf->offs >= 0 && wbuf->offs % c->min_io_size == 0); ubifs_assert(!(wbuf->offs & 7) && wbuf->offs <= c->leb_size); @@ -533,8 +537,8 @@ int ubifs_wbuf_write_nolock(struct ubifs_wbuf *wbuf, void *buf, int len) memcpy(wbuf->buf + wbuf->used, buf, len); if (aligned_len == wbuf->avail) { - dbg_io("flush wbuf to LEB %d:%d", wbuf->lnum, - wbuf->offs); + dbg_io("flush jhead %d wbuf to LEB %d:%d", + wbuf->jhead, wbuf->lnum, wbuf->offs); err = ubi_leb_write(c->ubi, wbuf->lnum, wbuf->buf, wbuf->offs, c->min_io_size, wbuf->dtype); @@ -562,7 +566,8 @@ int ubifs_wbuf_write_nolock(struct ubifs_wbuf *wbuf, void *buf, int len) * minimal I/O unit. We have to fill and flush write-buffer and switch * to the next min. I/O unit. */ - dbg_io("flush wbuf to LEB %d:%d", wbuf->lnum, wbuf->offs); + dbg_io("flush jhead %d wbuf to LEB %d:%d", + wbuf->jhead, wbuf->lnum, wbuf->offs); memcpy(wbuf->buf + wbuf->used, buf, wbuf->avail); err = ubi_leb_write(c->ubi, wbuf->lnum, wbuf->buf, wbuf->offs, c->min_io_size, wbuf->dtype); @@ -695,7 +700,8 @@ int ubifs_read_node_wbuf(struct ubifs_wbuf *wbuf, void *buf, int type, int len, int err, rlen, overlap; struct ubifs_ch *ch = buf; - dbg_io("LEB %d:%d, %s, length %d", lnum, offs, dbg_ntype(type), len); + dbg_io("LEB %d:%d, %s, length %d, jhead %d", lnum, offs, + dbg_ntype(type), len, wbuf->jhead); ubifs_assert(wbuf && lnum >= 0 && lnum < c->leb_cnt && offs >= 0); ubifs_assert(!(offs & 7) && offs < c->leb_size); ubifs_assert(type >= 0 && type < UBIFS_NODE_TYPES_CNT); -- cgit v1.2.3 From 0b335b9d7d5f0b832e90ac469480789c07be80ad Mon Sep 17 00:00:00 2001 From: Artem Bityutskiy Date: Tue, 23 Jun 2009 12:30:43 +0300 Subject: UBIFS: slightly optimize write-buffer timer usage This patch adds the following minor optimization: 1. If write-buffer does not use the timer, indicate it with the wbuf->no_timer variable, instead of using the wbuf->softlimit variable. This is better because wbuf->softlimit is of ktime_t type, and the ktime_to_ns function contains 64-bit multiplication. 2. Do not call the 'hrtimer_cancel()' function for write-buffers which do not use timers. 3. Do not cancel the timer in 'ubifs_put_super()' because the synchronization function does this. This patch also removes a confusing comment. Signed-off-by: Artem Bityutskiy --- fs/ubifs/io.c | 9 +++------ fs/ubifs/super.c | 6 ++---- fs/ubifs/ubifs.h | 6 ++++-- 3 files changed, 9 insertions(+), 12 deletions(-) (limited to 'fs') diff --git a/fs/ubifs/io.c b/fs/ubifs/io.c index 2ef689a9a363..9fcf6c38c1bc 100644 --- a/fs/ubifs/io.c +++ b/fs/ubifs/io.c @@ -312,7 +312,7 @@ static void new_wbuf_timer_nolock(struct ubifs_wbuf *wbuf) { ubifs_assert(!hrtimer_active(&wbuf->timer)); - if (!ktime_to_ns(wbuf->softlimit)) + if (wbuf->no_timer) return; dbg_io("set timer for jhead %d, %llu-%llu millisecs", wbuf->jhead, ktime_to_ns(wbuf->softlimit)/USEC_PER_SEC, @@ -327,11 +327,8 @@ static void new_wbuf_timer_nolock(struct ubifs_wbuf *wbuf) */ static void cancel_wbuf_timer_nolock(struct ubifs_wbuf *wbuf) { - /* - * If the syncer is waiting for the lock (from the background thread's - * context) and another task is changing write-buffer then the syncing - * should be canceled. - */ + if (wbuf->no_timer) + return; wbuf->need_sync = 0; hrtimer_cancel(&wbuf->timer); } diff --git a/fs/ubifs/super.c b/fs/ubifs/super.c index 79fad43f3c57..5bb272c56a9b 100644 --- a/fs/ubifs/super.c +++ b/fs/ubifs/super.c @@ -797,7 +797,7 @@ static int alloc_wbufs(struct ubifs_info *c) * does not need to be synchronized by timer. */ c->jheads[GCHD].wbuf.dtype = UBI_LONGTERM; - c->jheads[GCHD].wbuf.softlimit = ktime_set(0, 0); + c->jheads[GCHD].wbuf.no_timer = 1; return 0; } @@ -1754,10 +1754,8 @@ static void ubifs_put_super(struct super_block *sb) /* Synchronize write-buffers */ if (c->jheads) - for (i = 0; i < c->jhead_cnt; i++) { + for (i = 0; i < c->jhead_cnt; i++) ubifs_wbuf_sync(&c->jheads[i].wbuf); - hrtimer_cancel(&c->jheads[i].wbuf.timer); - } /* * On fatal errors c->ro_media is set to 1, in which case we do diff --git a/fs/ubifs/ubifs.h b/fs/ubifs/ubifs.h index 1bf01d820066..97bc9d09d54b 100644 --- a/fs/ubifs/ubifs.h +++ b/fs/ubifs/ubifs.h @@ -654,7 +654,8 @@ typedef int (*ubifs_lpt_scan_callback)(struct ubifs_info *c, * @delta: hard and soft timeouts delta (the timer expire inteval is @softlimit * and @softlimit + @delta) * @timer: write-buffer timer - * @need_sync: it is set if its timer expired and needs sync + * @no_timer: non-zero if this write-buffer does not timer + * @need_sync: non-zero if its timer expired and needs sync * @next_ino: points to the next position of the following inode number * @inodes: stores the inode numbers of the nodes which are in wbuf * @@ -683,7 +684,8 @@ struct ubifs_wbuf { ktime_t softlimit; unsigned long long delta; struct hrtimer timer; - int need_sync; + unsigned int no_timer:1; + unsigned int need_sync:1; int next_ino; ino_t *inodes; }; -- cgit v1.2.3 From 2a35a3a8ab3e94afd631ed4b45878ceb98f7ab28 Mon Sep 17 00:00:00 2001 From: Artem Bityutskiy Date: Tue, 23 Jun 2009 20:26:33 +0300 Subject: UBIFS: set write-buffer timout to 3-5 seconds This patch cleans up write-buffer timeout initialization and sets it to 3-5 interval. Signed-off-by: Artem Bityutskiy --- fs/ubifs/io.c | 17 ++++------------- fs/ubifs/ubifs.h | 5 +++-- 2 files changed, 7 insertions(+), 15 deletions(-) (limited to 'fs') diff --git a/fs/ubifs/io.c b/fs/ubifs/io.c index 9fcf6c38c1bc..48d0af94b26a 100644 --- a/fs/ubifs/io.c +++ b/fs/ubifs/io.c @@ -828,7 +828,6 @@ out: int ubifs_wbuf_init(struct ubifs_info *c, struct ubifs_wbuf *wbuf) { size_t size; - ktime_t hardlimit; wbuf->buf = kmalloc(c->min_io_size, GFP_KERNEL); if (!wbuf->buf) @@ -854,18 +853,10 @@ int ubifs_wbuf_init(struct ubifs_info *c, struct ubifs_wbuf *wbuf) hrtimer_init(&wbuf->timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); wbuf->timer.function = wbuf_timer_callback_nolock; - /* - * Make write-buffer soft limit to be 20% of the hard limit. The - * write-buffer timer is allowed to expire any time between the soft - * and hard limits. - */ - hardlimit = ktime_set(DEFAULT_WBUF_TIMEOUT_SECS, 0); - wbuf->delta = DEFAULT_WBUF_TIMEOUT_SECS * 1000000000ULL * 2 / 10; - if (wbuf->delta > ULONG_MAX) - wbuf->delta = ULONG_MAX; - wbuf->softlimit = ktime_sub_ns(hardlimit, wbuf->delta); - hrtimer_set_expires_range_ns(&wbuf->timer, wbuf->softlimit, - wbuf->delta); + wbuf->softlimit = ktime_set(WBUF_TIMEOUT_SOFTLIMIT, 0); + wbuf->delta = WBUF_TIMEOUT_HARDLIMIT - WBUF_TIMEOUT_SOFTLIMIT; + wbuf->delta *= 1000000000ULL; + ubifs_assert(wbuf->delta <= ULONG_MAX); return 0; } diff --git a/fs/ubifs/ubifs.h b/fs/ubifs/ubifs.h index 97bc9d09d54b..c3a707d458a1 100644 --- a/fs/ubifs/ubifs.h +++ b/fs/ubifs/ubifs.h @@ -95,8 +95,9 @@ */ #define BGT_NAME_PATTERN "ubifs_bgt%d_%d" -/* Default write-buffer synchronization timeout in seconds */ -#define DEFAULT_WBUF_TIMEOUT_SECS 5 +/* Write-buffer synchronization timeout interval in seconds */ +#define WBUF_TIMEOUT_SOFTLIMIT 3 +#define WBUF_TIMEOUT_HARDLIMIT 5 /* Maximum possible inode number (only 32-bit inodes are supported now) */ #define MAX_INUM 0xFFFFFFFF -- cgit v1.2.3 From cb54ef8b1304fe25f3d57031e0f85558a043239f Mon Sep 17 00:00:00 2001 From: Artem Bityutskiy Date: Tue, 23 Jun 2009 20:30:32 +0300 Subject: UBIFS: few spelling fixes Signed-off-by: Artem Bityutskiy --- fs/ubifs/io.c | 6 +++--- fs/ubifs/super.c | 2 +- 2 files changed, 4 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/ubifs/io.c b/fs/ubifs/io.c index 48d0af94b26a..7e4267ecfa56 100644 --- a/fs/ubifs/io.c +++ b/fs/ubifs/io.c @@ -391,7 +391,7 @@ int ubifs_wbuf_sync_nolock(struct ubifs_wbuf *wbuf) * @offs: logical eraseblock offset to seek to * @dtype: data type * - * This function targets the write buffer to logical eraseblock @lnum:@offs. + * This function targets the write-buffer to logical eraseblock @lnum:@offs. * The write-buffer is synchronized if it is not empty. Returns zero in case of * success and a negative error code in case of failure. */ @@ -822,7 +822,7 @@ out: * @c: UBIFS file-system description object * @wbuf: write-buffer to initialize * - * This function initializes write buffer. Returns zero in case of success + * This function initializes write-buffer. Returns zero in case of success * %-ENOMEM in case of failure. */ int ubifs_wbuf_init(struct ubifs_info *c, struct ubifs_wbuf *wbuf) @@ -862,7 +862,7 @@ int ubifs_wbuf_init(struct ubifs_info *c, struct ubifs_wbuf *wbuf) /** * ubifs_wbuf_add_ino_nolock - add an inode number into the wbuf inode array. - * @wbuf: the write-buffer whereto add + * @wbuf: the write-buffer where to add * @inum: the inode number * * This function adds an inode number to the inode array of the write-buffer. diff --git a/fs/ubifs/super.c b/fs/ubifs/super.c index 5bb272c56a9b..76405c457a38 100644 --- a/fs/ubifs/super.c +++ b/fs/ubifs/super.c @@ -986,7 +986,7 @@ static int ubifs_parse_options(struct ubifs_info *c, char *options, switch (token) { /* * %Opt_fast_unmount and %Opt_norm_unmount options are ignored. - * We accepte them in order to be backware-compatible. But this + * We accept them in order to be backward-compatible. But this * should be removed at some point. */ case Opt_fast_unmount: -- cgit v1.2.3 From 44737589442bf69d811e003d9d0064b8fc1541d6 Mon Sep 17 00:00:00 2001 From: Adrian Hunter Date: Wed, 24 Jun 2009 10:15:12 +0300 Subject: UBIFS: fix 64-bit divisions in debug print Signed-off-by: Adrian Hunter --- fs/ubifs/io.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/ubifs/io.c b/fs/ubifs/io.c index 7e4267ecfa56..762a7d6cec73 100644 --- a/fs/ubifs/io.c +++ b/fs/ubifs/io.c @@ -315,8 +315,9 @@ static void new_wbuf_timer_nolock(struct ubifs_wbuf *wbuf) if (wbuf->no_timer) return; dbg_io("set timer for jhead %d, %llu-%llu millisecs", wbuf->jhead, - ktime_to_ns(wbuf->softlimit)/USEC_PER_SEC, - (ktime_to_ns(wbuf->softlimit) + wbuf->delta)/USEC_PER_SEC); + div_u64(ktime_to_ns(wbuf->softlimit), USEC_PER_SEC), + div_u64(ktime_to_ns(wbuf->softlimit) + wbuf->delta, + USEC_PER_SEC)); hrtimer_start_range_ns(&wbuf->timer, wbuf->softlimit, wbuf->delta, HRTIMER_MODE_REL); } -- cgit v1.2.3 From 681947d2fa1a00629de33c4df3ca72c39f06a14c Mon Sep 17 00:00:00 2001 From: Adrian Hunter Date: Wed, 24 Jun 2009 09:59:38 +0300 Subject: UBIFS: minor spelling and grammar fixes Signed-off-by: Adrian Hunter --- fs/ubifs/replay.c | 2 +- fs/ubifs/ubifs.h | 4 ++-- 2 files changed, 3 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/ubifs/replay.c b/fs/ubifs/replay.c index 11cc80125a49..769be42f39d6 100644 --- a/fs/ubifs/replay.c +++ b/fs/ubifs/replay.c @@ -957,7 +957,7 @@ out: return err; out_dump: - ubifs_err("log error detected while replying the log at LEB %d:%d", + ubifs_err("log error detected while replaying the log at LEB %d:%d", lnum, offs + snod->offs); dbg_dump_node(c, snod->node); ubifs_scan_destroy(sleb); diff --git a/fs/ubifs/ubifs.h b/fs/ubifs/ubifs.h index c3a707d458a1..a29349094422 100644 --- a/fs/ubifs/ubifs.h +++ b/fs/ubifs/ubifs.h @@ -655,8 +655,8 @@ typedef int (*ubifs_lpt_scan_callback)(struct ubifs_info *c, * @delta: hard and soft timeouts delta (the timer expire inteval is @softlimit * and @softlimit + @delta) * @timer: write-buffer timer - * @no_timer: non-zero if this write-buffer does not timer - * @need_sync: non-zero if its timer expired and needs sync + * @no_timer: non-zero if this write-buffer does not have a timer + * @need_sync: non-zero if the timer expired and the wbuf needs sync'ing * @next_ino: points to the next position of the following inode number * @inodes: stores the inode numbers of the nodes which are in wbuf * -- cgit v1.2.3 From 1fb8bd01ed0af0d0577e010e8c6b4234de583fa6 Mon Sep 17 00:00:00 2001 From: Artem Bityutskiy Date: Sun, 28 Jun 2009 18:31:58 +0300 Subject: UBIFS: fix assertion warning When debugging is enabled and an unclean file-system is mounter, the following assertion is triggered: UBIFS assert failed in ubifs_tnc_start_commit at 805 (pid 1081) Call Trace: [cfaffbd0] [c0006cf8] show_stack+0x44/0x16c (unreliable) [cfaffc10] [c011b738] ubifs_tnc_start_commit+0xbb8/0xd18 [cfaffc90] [c0112670] do_commit+0x150/0xa44 [cfaffd10] [c0125234] ubifs_rcvry_gc_commit+0xd8/0x544 [cfaffd60] [c0100e9c] ubifs_fill_super+0xe78/0x15f8 [cfaffdf0] [c0102118] ubifs_get_sb+0x20c/0x320 [cfaffe70] [c007f764] vfs_kern_mount+0x58/0xe0 [cfaffe90] [c007f83c] do_kern_mount+0x40/0xf8 [cfaffeb0] [c0095c24] do_mount+0x550/0x758 [cfafff10] [c0095ebc] sys_mount+0x90/0xe0 [cfafff40] [c000ed4c] ret_from_syscall+0x0/0x3c The reason is that we initialize 'c->min_leb_idx' early, and do not re-calculate it after journal replay. Signed-off-by: Artem Bityutskiy --- fs/ubifs/super.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'fs') diff --git a/fs/ubifs/super.c b/fs/ubifs/super.c index 76405c457a38..3507d0ed542d 100644 --- a/fs/ubifs/super.c +++ b/fs/ubifs/super.c @@ -1287,6 +1287,9 @@ static int mount_ubifs(struct ubifs_info *c) if (err) goto out_journal; + /* Calculate 'min_idx_lebs' after journal replay */ + c->min_idx_lebs = ubifs_calc_min_idx_lebs(c); + err = ubifs_mount_orphans(c, c->need_recovery, mounted_read_only); if (err) goto out_orphans; -- cgit v1.2.3 From 7fcd9c3ecbf09c0a77db7ba01aac75b32fb79a93 Mon Sep 17 00:00:00 2001 From: Daniel Mack Date: Thu, 2 Jul 2009 17:15:47 +0200 Subject: UBIFS: allow more than one volume to be mounted UBIFS uses a bdi device per volume, but does not care to hand out unique names to each of them. This causes an error when trying to mount more than one volumes. Append the UBI volume and device ID to avoid that. [Amended a bit by Artem Bityutskiy] Signed-off-by: Daniel Mack Cc: Artem Bityutskiy Cc: Adrian Hunter Cc: linux-mtd@lists.infradead.org Signed-off-by: Artem Bityutskiy --- fs/ubifs/super.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ubifs/super.c b/fs/ubifs/super.c index 3507d0ed542d..26d2e0d80465 100644 --- a/fs/ubifs/super.c +++ b/fs/ubifs/super.c @@ -1976,7 +1976,8 @@ static int ubifs_fill_super(struct super_block *sb, void *data, int silent) err = bdi_init(&c->bdi); if (err) goto out_close; - err = bdi_register(&c->bdi, NULL, "ubifs"); + err = bdi_register(&c->bdi, NULL, "ubifs_%d_%d", + c->vi.ubi_num, c->vi.vol_id); if (err) goto out_bdi; -- cgit v1.2.3 From 3beab0b42413e83a7907db7176b54c840fc75a81 Mon Sep 17 00:00:00 2001 From: "Zhang, Yanmin" Date: Sun, 5 Jul 2009 12:08:08 -0700 Subject: sys_sync(): fix 16% performance regression in ffsb create_4k test I run many ffsb test cases on JBODs (typically 13/12 disks). Comparing with kernel 2.6.30, 2.6.31-rc1 has about 16% regression with ffsb_create_4k. The sub test case creates files continuously for 10 minitues and every file is 1MB. Bisect located below patch. 5cee5815d1564bbbd505fea86f4550f1efdb5cd0 is first bad commit commit 5cee5815d1564bbbd505fea86f4550f1efdb5cd0 Author: Jan Kara Date: Mon Apr 27 16:43:51 2009 +0200 vfs: Make sys_sync() use fsync_super() (version 4) It is unnecessarily fragile to have two places (fsync_super() and do_sync()) doing data integrity sync of the filesystem. Alter __fsync_super() to accommodate needs of both callers and use it. So after this patch __fsync_super() is the only place where we gather all the calls needed to properly send all data on a filesystem to disk. As a matter of fact, ffsb calls sys_sync in the end to make sure all data is flushed to disks and the flushing is counted into the result. vmstat shows ffsb is blocked when syncing for a long time. With 2.6.30, ffsb is blocked for a short time. I checked the patch and did experiments to recover the original methods. Eventually, the root cause is the patch deletes the calling to wakeup_pdflush when syncing, so only ffsb is blocked on disk I/O. wakeup_pdflush could ask pdflush to write back pages with ffsb at the same time. [akpm@linux-foundation.org: restore comment too] Signed-off-by: Zhang Yanmin Cc: Jan Kara Cc: Al Viro Acked-by: Jens Axboe Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/sync.c | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'fs') diff --git a/fs/sync.c b/fs/sync.c index dd200025af85..3422ba61d86d 100644 --- a/fs/sync.c +++ b/fs/sync.c @@ -112,8 +112,13 @@ restart: mutex_unlock(&mutex); } +/* + * sync everything. Start out by waking pdflush, because that writes back + * all queues in parallel. + */ SYSCALL_DEFINE0(sync) { + wakeup_pdflush(0); sync_filesystems(0); sync_filesystems(1); if (unlikely(laptop_mode)) -- cgit v1.2.3 From 793285fcafce4719a05e0c99fa74b188157fe7fe Mon Sep 17 00:00:00 2001 From: Oleg Nesterov Date: Sun, 5 Jul 2009 12:08:26 -0700 Subject: cred_guard_mutex: do not return -EINTR to user-space do_execve() and ptrace_attach() return -EINTR if mutex_lock_interruptible(->cred_guard_mutex) fails. This is not right, change the code to return ERESTARTNOINTR. Perhaps we should also change proc_pid_attr_write(). Signed-off-by: Oleg Nesterov Cc: David Howells Acked-by: Roland McGrath Cc: James Morris Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/compat.c | 4 ++-- fs/exec.c | 4 ++-- kernel/ptrace.c | 4 ++-- 3 files changed, 6 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/compat.c b/fs/compat.c index cdd51a3a7c53..fbadb947727b 100644 --- a/fs/compat.c +++ b/fs/compat.c @@ -1486,8 +1486,8 @@ int compat_do_execve(char * filename, if (!bprm) goto out_files; - retval = mutex_lock_interruptible(¤t->cred_guard_mutex); - if (retval < 0) + retval = -ERESTARTNOINTR; + if (mutex_lock_interruptible(¤t->cred_guard_mutex)) goto out_free; current->in_execve = 1; diff --git a/fs/exec.c b/fs/exec.c index e639957d7a57..4a8849e45b21 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -1277,8 +1277,8 @@ int do_execve(char * filename, if (!bprm) goto out_files; - retval = mutex_lock_interruptible(¤t->cred_guard_mutex); - if (retval < 0) + retval = -ERESTARTNOINTR; + if (mutex_lock_interruptible(¤t->cred_guard_mutex)) goto out_free; current->in_execve = 1; diff --git a/kernel/ptrace.c b/kernel/ptrace.c index 61c78b2c07ba..082c320e4dbf 100644 --- a/kernel/ptrace.c +++ b/kernel/ptrace.c @@ -181,8 +181,8 @@ int ptrace_attach(struct task_struct *task) * interference; SUID, SGID and LSM creds get determined differently * under ptrace. */ - retval = mutex_lock_interruptible(&task->cred_guard_mutex); - if (retval < 0) + retval = -ERESTARTNOINTR; + if (mutex_lock_interruptible(&task->cred_guard_mutex)) goto out; task_lock(task); -- cgit v1.2.3 From d01730d74d2b0155da50d44555001706294014f7 Mon Sep 17 00:00:00 2001 From: Jiaying Zhang Date: Tue, 7 Jul 2009 18:15:21 +0200 Subject: quota: Fix possible deadlock during parallel quotaon and quotaoff The following test script triggers a deadlock on ext2 filesystem: while true; do quotaon /dev/hda >&/dev/null; usleep $RANDOM; done & while true; do quotaoff /dev/hda >&/dev/null; usleep $RANDOM; done & I found there is a potential deadlock between quotaon and quotaoff (or quotasync). Basically, all of quotactl operations need to be protected by dqonoff_mutex. vfs_quota_off and vfs_quota_sync also call sb->s_op->quota_write that needs to grab the i_mutex of the quota file. But in vfs_quota_on_inode (called from quotaon operation), the current code tries to grab the i_mutex of the quota file first before getting quonoff_mutex. Reverse the order in which we take locks in vfs_quota_on_inode(). Jan Kara: Changed changelog to be more readable, made lockdep happy with I_MUTEX_QUOTA. Signed-off-by: Jiaying Zhang Signed-off-by: Jan Kara --- fs/quota/dquot.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/quota/dquot.c b/fs/quota/dquot.c index 607c579e5eca..70f36c043d62 100644 --- a/fs/quota/dquot.c +++ b/fs/quota/dquot.c @@ -2042,8 +2042,8 @@ static int vfs_load_quota_inode(struct inode *inode, int type, int format_id, * changes */ invalidate_bdev(sb->s_bdev); } - mutex_lock(&inode->i_mutex); mutex_lock(&dqopt->dqonoff_mutex); + mutex_lock_nested(&inode->i_mutex, I_MUTEX_QUOTA); if (sb_has_quota_loaded(sb, type)) { error = -EBUSY; goto out_lock; @@ -2094,7 +2094,6 @@ out_file_init: dqopt->files[type] = NULL; iput(inode); out_lock: - mutex_unlock(&dqopt->dqonoff_mutex); if (oldflags != -1) { down_write(&dqopt->dqptr_sem); /* Set the flags back (in the case of accidental quotaon() @@ -2104,6 +2103,7 @@ out_lock: up_write(&dqopt->dqptr_sem); } mutex_unlock(&inode->i_mutex); + mutex_unlock(&dqopt->dqonoff_mutex); out_fmt: put_quota_format(fmt); -- cgit v1.2.3 From b43f3cbd21ffbd719fd4fa6642bfe6af255ded34 Mon Sep 17 00:00:00 2001 From: Alexey Dobriyan Date: Wed, 8 Jul 2009 01:54:37 +0400 Subject: headers: mnt_namespace.h redux Fix various silly problems wrt mnt_namespace.h: - exit_mnt_ns() isn't used, remove it - done that, sched.h and nsproxy.h inclusions aren't needed - mount.h inclusion was need for vfsmount_lock, but no longer - remove mnt_namespace.h inclusion from files which don't use anything from mnt_namespace.h Signed-off-by: Alexey Dobriyan Signed-off-by: Linus Torvalds --- fs/afs/mntpt.c | 1 - fs/namespace.c | 1 + fs/nfs/getroot.c | 1 - fs/reiserfs/super.c | 1 - include/linux/mnt_namespace.h | 13 ++----------- kernel/exit.c | 1 - kernel/fork.c | 1 - kernel/kmod.c | 1 - 8 files changed, 3 insertions(+), 17 deletions(-) (limited to 'fs') diff --git a/fs/afs/mntpt.c b/fs/afs/mntpt.c index c52be53f6946..5ffb570cd3a8 100644 --- a/fs/afs/mntpt.c +++ b/fs/afs/mntpt.c @@ -17,7 +17,6 @@ #include #include #include -#include #include "internal.h" diff --git a/fs/namespace.c b/fs/namespace.c index 3dc283fd4716..277c28a63ead 100644 --- a/fs/namespace.c +++ b/fs/namespace.c @@ -22,6 +22,7 @@ #include #include #include +#include #include #include #include diff --git a/fs/nfs/getroot.c b/fs/nfs/getroot.c index 46177cb87064..b35d2a616066 100644 --- a/fs/nfs/getroot.c +++ b/fs/nfs/getroot.c @@ -30,7 +30,6 @@ #include #include #include -#include #include #include diff --git a/fs/reiserfs/super.c b/fs/reiserfs/super.c index d3aeb061612b..7adea74d6a8a 100644 --- a/fs/reiserfs/super.c +++ b/fs/reiserfs/super.c @@ -24,7 +24,6 @@ #include #include #include -#include #include #include #include diff --git a/include/linux/mnt_namespace.h b/include/linux/mnt_namespace.h index 3beb2592b03f..d74785c2393a 100644 --- a/include/linux/mnt_namespace.h +++ b/include/linux/mnt_namespace.h @@ -2,10 +2,9 @@ #define _NAMESPACE_H_ #ifdef __KERNEL__ -#include -#include -#include +#include #include +#include struct mnt_namespace { atomic_t count; @@ -28,14 +27,6 @@ extern struct mnt_namespace *create_mnt_ns(struct vfsmount *mnt); extern struct mnt_namespace *copy_mnt_ns(unsigned long, struct mnt_namespace *, struct fs_struct *); extern void put_mnt_ns(struct mnt_namespace *ns); - -static inline void exit_mnt_ns(struct task_struct *p) -{ - struct mnt_namespace *ns = p->nsproxy->mnt_ns; - if (ns) - put_mnt_ns(ns); -} - static inline void get_mnt_ns(struct mnt_namespace *ns) { atomic_inc(&ns->count); diff --git a/kernel/exit.c b/kernel/exit.c index 628d41f0dd54..869dc221733e 100644 --- a/kernel/exit.c +++ b/kernel/exit.c @@ -12,7 +12,6 @@ #include #include #include -#include #include #include #include diff --git a/kernel/fork.c b/kernel/fork.c index 467746b3f0aa..bd2959228871 100644 --- a/kernel/fork.c +++ b/kernel/fork.c @@ -17,7 +17,6 @@ #include #include #include -#include #include #include #include diff --git a/kernel/kmod.c b/kernel/kmod.c index 7e95bedb2bfc..385c31a1bdbf 100644 --- a/kernel/kmod.c +++ b/kernel/kmod.c @@ -24,7 +24,6 @@ #include #include #include -#include #include #include #include -- cgit v1.2.3 From d5ce5b40bc66880d1732461d4b47d7fc3331ed30 Mon Sep 17 00:00:00 2001 From: Catalin Marinas Date: Wed, 8 Jul 2009 11:17:34 +0100 Subject: Free the memory allocated by memdup_user() in fs/sysfs/bin.c Commit 1c8542c7bb replaced kmalloc() with memdup_user() in the write() function but also dropped the kfree(temp). The memdup_user() function allocates memory which is never freed. Signed-off-by: Catalin Marinas Cc: Parag Warudkar Signed-off-by: Linus Torvalds --- fs/sysfs/bin.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/sysfs/bin.c b/fs/sysfs/bin.c index 9345806c8853..2524714bece1 100644 --- a/fs/sysfs/bin.c +++ b/fs/sysfs/bin.c @@ -171,6 +171,7 @@ static ssize_t write(struct file *file, const char __user *userbuf, if (count > 0) *off = offs + count; + kfree(temp); return count; } -- cgit v1.2.3 From ad361c9884e809340f6daca80d56a9e9c871690a Mon Sep 17 00:00:00 2001 From: Joe Perches Date: Mon, 6 Jul 2009 13:05:40 -0700 Subject: Remove multiple KERN_ prefixes from printk formats Commit 5fd29d6ccbc98884569d6f3105aeca70858b3e0f ("printk: clean up handling of log-levels and newlines") changed printk semantics. printk lines with multiple KERN_ prefixes are no longer emitted as before the patch. is now included in the output on each additional use. Remove all uses of multiple KERN_s in formats. Signed-off-by: Joe Perches Signed-off-by: Linus Torvalds --- arch/avr32/kernel/traps.c | 13 ++++---- arch/blackfin/kernel/setup.c | 41 ++++++++++++------------- arch/blackfin/kernel/traps.c | 34 ++++++++++----------- arch/m68knommu/kernel/process.c | 21 +++++++------ arch/m68knommu/kernel/traps.c | 6 ++-- arch/mn10300/kernel/traps.c | 21 +++++-------- arch/parisc/kernel/process.c | 2 +- arch/parisc/kernel/traps.c | 20 ++++++------- arch/um/kernel/sysrq.c | 4 +-- arch/x86/kernel/apic/io_apic.c | 1 - arch/x86/kernel/cpu/cpufreq/powernow-k8.c | 2 +- arch/x86/kernel/cpu/mcheck/mce.c | 12 ++++---- arch/x86/kernel/e820.c | 7 ++--- arch/x86/kernel/pci-gart_64.c | 2 +- arch/x86/mm/fault.c | 9 +++--- arch/xtensa/kernel/traps.c | 6 ++-- drivers/block/amiflop.c | 2 +- drivers/block/xsysace.c | 7 +++-- drivers/char/hw_random/intel-rng.c | 9 +++--- drivers/char/isicom.c | 16 +++++----- drivers/i2c/busses/i2c-ibm_iic.c | 9 +++--- drivers/md/md.c | 18 ++++++----- drivers/misc/sgi-xp/xpnet.c | 4 +-- drivers/net/a2065.c | 12 ++------ drivers/net/arcnet/arcnet.c | 26 +++++++--------- drivers/net/bmac.c | 7 +++-- drivers/net/bnx2x_main.c | 7 +++-- drivers/net/dl2k.c | 17 ++++++----- drivers/net/epic100.c | 5 ++-- drivers/net/fealnx.c | 15 ++++++---- drivers/net/hamachi.c | 23 +++++++------- drivers/net/hamradio/baycom_epp.c | 2 +- drivers/net/hamradio/baycom_par.c | 2 +- drivers/net/hamradio/baycom_ser_fdx.c | 2 +- drivers/net/hamradio/baycom_ser_hdx.c | 2 +- drivers/net/natsemi.c | 4 +-- drivers/net/ne.c | 2 +- drivers/net/pci-skeleton.c | 2 +- drivers/net/pcmcia/ibmtr_cs.c | 13 ++++---- drivers/net/pcmcia/nmclan_cs.c | 15 ++++------ drivers/net/pcnet32.c | 50 +++++++++++++++---------------- drivers/net/starfire.c | 2 +- drivers/net/sundance.c | 6 ++-- drivers/net/tsi108_eth.c | 8 +++-- drivers/net/tulip/de2104x.c | 7 +++-- drivers/net/tulip/tulip_core.c | 14 +++++---- drivers/net/tulip/winbond-840.c | 6 ++-- drivers/net/wan/hd64570.c | 3 +- drivers/net/wan/hd64572.c | 3 +- drivers/net/wan/sbni.c | 8 ++--- drivers/net/wireless/ray_cs.c | 9 +++--- drivers/net/wireless/wavelan_cs.c | 13 ++------ drivers/net/yellowfin.c | 23 +++++++------- drivers/parisc/eisa_enumerator.c | 14 ++++----- drivers/pcmcia/tcic.c | 3 +- drivers/scsi/atari_NCR5380.c | 3 +- drivers/scsi/mac53c94.c | 5 ++-- drivers/scsi/sg.c | 2 +- drivers/scsi/sun3_NCR5380.c | 3 +- drivers/serial/8250_pci.c | 11 +++---- drivers/ssb/pcmcia.c | 4 +-- drivers/usb/core/hcd.c | 2 +- drivers/video/amba-clcd.c | 7 +++-- drivers/video/matrox/matroxfb_DAC1064.c | 4 +-- drivers/video/matrox/matroxfb_Ti3026.c | 4 +-- drivers/video/stifb.c | 7 ++--- fs/jffs2/erase.c | 10 ++++--- kernel/module.c | 6 ++-- sound/pci/emu10k1/p16v.c | 2 +- sound/usb/usx2y/usbusx2yaudio.c | 7 +++-- 70 files changed, 330 insertions(+), 338 deletions(-) (limited to 'fs') diff --git a/arch/avr32/kernel/traps.c b/arch/avr32/kernel/traps.c index 6e3d491184ea..b91b2044af9c 100644 --- a/arch/avr32/kernel/traps.c +++ b/arch/avr32/kernel/traps.c @@ -32,22 +32,25 @@ void NORET_TYPE die(const char *str, struct pt_regs *regs, long err) spin_lock_irq(&die_lock); bust_spinlocks(1); - printk(KERN_ALERT "Oops: %s, sig: %ld [#%d]\n" KERN_EMERG, + printk(KERN_ALERT "Oops: %s, sig: %ld [#%d]\n", str, err, ++die_counter); + + printk(KERN_EMERG); + #ifdef CONFIG_PREEMPT - printk("PREEMPT "); + printk(KERN_CONT "PREEMPT "); #endif #ifdef CONFIG_FRAME_POINTER - printk("FRAME_POINTER "); + printk(KERN_CONT "FRAME_POINTER "); #endif if (current_cpu_data.features & AVR32_FEATURE_OCD) { unsigned long did = ocd_read(DID); - printk("chip: 0x%03lx:0x%04lx rev %lu\n", + printk(KERN_CONT "chip: 0x%03lx:0x%04lx rev %lu\n", (did >> 1) & 0x7ff, (did >> 12) & 0x7fff, (did >> 28) & 0xf); } else { - printk("cpu: arch %u r%u / core %u r%u\n", + printk(KERN_CONT "cpu: arch %u r%u / core %u r%u\n", current_cpu_data.arch_type, current_cpu_data.arch_revision, current_cpu_data.cpu_type, diff --git a/arch/blackfin/kernel/setup.c b/arch/blackfin/kernel/setup.c index 298f023bcc09..6136c33e919f 100644 --- a/arch/blackfin/kernel/setup.c +++ b/arch/blackfin/kernel/setup.c @@ -408,13 +408,14 @@ static void __init print_memory_map(char *who) bfin_memmap.map[i].addr + bfin_memmap.map[i].size); switch (bfin_memmap.map[i].type) { case BFIN_MEMMAP_RAM: - printk("(usable)\n"); - break; + printk(KERN_CONT "(usable)\n"); + break; case BFIN_MEMMAP_RESERVED: - printk("(reserved)\n"); - break; - default: printk("type %lu\n", bfin_memmap.map[i].type); - break; + printk(KERN_CONT "(reserved)\n"); + break; + default: + printk(KERN_CONT "type %lu\n", bfin_memmap.map[i].type); + break; } } } @@ -614,19 +615,19 @@ static __init void memory_setup(void) printk(KERN_INFO "Kernel Managed Memory: %ldMB\n", _ramend >> 20); printk(KERN_INFO "Memory map:\n" - KERN_INFO " fixedcode = 0x%p-0x%p\n" - KERN_INFO " text = 0x%p-0x%p\n" - KERN_INFO " rodata = 0x%p-0x%p\n" - KERN_INFO " bss = 0x%p-0x%p\n" - KERN_INFO " data = 0x%p-0x%p\n" - KERN_INFO " stack = 0x%p-0x%p\n" - KERN_INFO " init = 0x%p-0x%p\n" - KERN_INFO " available = 0x%p-0x%p\n" + " fixedcode = 0x%p-0x%p\n" + " text = 0x%p-0x%p\n" + " rodata = 0x%p-0x%p\n" + " bss = 0x%p-0x%p\n" + " data = 0x%p-0x%p\n" + " stack = 0x%p-0x%p\n" + " init = 0x%p-0x%p\n" + " available = 0x%p-0x%p\n" #ifdef CONFIG_MTD_UCLINUX - KERN_INFO " rootfs = 0x%p-0x%p\n" + " rootfs = 0x%p-0x%p\n" #endif #if DMA_UNCACHED_REGION > 0 - KERN_INFO " DMA Zone = 0x%p-0x%p\n" + " DMA Zone = 0x%p-0x%p\n" #endif , (void *)FIXED_CODE_START, (void *)FIXED_CODE_END, _stext, _etext, @@ -859,13 +860,13 @@ void __init setup_arch(char **cmdline_p) #endif printk(KERN_INFO "Hardware Trace "); if (bfin_read_TBUFCTL() & 0x1) - printk("Active "); + printk(KERN_CONT "Active "); else - printk("Off "); + printk(KERN_CONT "Off "); if (bfin_read_TBUFCTL() & 0x2) - printk("and Enabled\n"); + printk(KERN_CONT "and Enabled\n"); else - printk("and Disabled\n"); + printk(KERN_CONT "and Disabled\n"); #if defined(CONFIG_CHR_DEV_FLASH) || defined(CONFIG_BLK_DEV_FLASH) /* we need to initialize the Flashrom device here since we might diff --git a/arch/blackfin/kernel/traps.c b/arch/blackfin/kernel/traps.c index 8eeb457ce5d5..8a1caf2bb5b9 100644 --- a/arch/blackfin/kernel/traps.c +++ b/arch/blackfin/kernel/traps.c @@ -212,7 +212,7 @@ asmlinkage void double_fault_c(struct pt_regs *fp) console_verbose(); oops_in_progress = 1; #ifdef CONFIG_DEBUG_VERBOSE - printk(KERN_EMERG "\n" KERN_EMERG "Double Fault\n"); + printk(KERN_EMERG "Double Fault\n"); #ifdef CONFIG_DEBUG_DOUBLEFAULT_PRINT if (((long)fp->seqstat & SEQSTAT_EXCAUSE) == VEC_UNCOV) { unsigned int cpu = smp_processor_id(); @@ -583,15 +583,14 @@ asmlinkage void trap_c(struct pt_regs *fp) #ifndef CONFIG_DEBUG_BFIN_NO_KERN_HWTRACE if (trapnr == VEC_CPLB_I_M || trapnr == VEC_CPLB_M) verbose_printk(KERN_NOTICE "No trace since you do not have " - "CONFIG_DEBUG_BFIN_NO_KERN_HWTRACE enabled\n" - KERN_NOTICE "\n"); + "CONFIG_DEBUG_BFIN_NO_KERN_HWTRACE enabled\n\n"); else #endif dump_bfin_trace_buffer(); if (oops_in_progress) { /* Dump the current kernel stack */ - verbose_printk(KERN_NOTICE "\n" KERN_NOTICE "Kernel Stack\n"); + verbose_printk(KERN_NOTICE "Kernel Stack\n"); show_stack(current, NULL); print_modules(); #ifndef CONFIG_ACCESS_CHECK @@ -906,7 +905,7 @@ void show_stack(struct task_struct *task, unsigned long *stack) ret_addr = 0; if (!j && i % 8 == 0) - printk("\n" KERN_NOTICE "%p:",addr); + printk(KERN_NOTICE "%p:",addr); /* if it is an odd address, or zero, just skip it */ if (*addr & 0x1 || !*addr) @@ -996,9 +995,9 @@ void dump_bfin_process(struct pt_regs *fp) printk(KERN_NOTICE "CPU = %d\n", current_thread_info()->cpu); if (!((unsigned long)current->mm & 0x3) && (unsigned long)current->mm >= FIXED_CODE_START) - verbose_printk(KERN_NOTICE "TEXT = 0x%p-0x%p DATA = 0x%p-0x%p\n" - KERN_NOTICE " BSS = 0x%p-0x%p USER-STACK = 0x%p\n" - KERN_NOTICE "\n", + verbose_printk(KERN_NOTICE + "TEXT = 0x%p-0x%p DATA = 0x%p-0x%p\n" + " BSS = 0x%p-0x%p USER-STACK = 0x%p\n\n", (void *)current->mm->start_code, (void *)current->mm->end_code, (void *)current->mm->start_data, @@ -1009,8 +1008,8 @@ void dump_bfin_process(struct pt_regs *fp) else verbose_printk(KERN_NOTICE "invalid mm\n"); } else - verbose_printk(KERN_NOTICE "\n" KERN_NOTICE - "No Valid process in current context\n"); + verbose_printk(KERN_NOTICE + "No Valid process in current context\n"); #endif } @@ -1028,7 +1027,7 @@ void dump_bfin_mem(struct pt_regs *fp) addr < (unsigned short *)((unsigned long)erraddr & ~0xF) + 0x10; addr++) { if (!((unsigned long)addr & 0xF)) - verbose_printk("\n" KERN_NOTICE "0x%p: ", addr); + verbose_printk(KERN_NOTICE "0x%p: ", addr); if (!get_instruction(&val, addr)) { val = 0; @@ -1056,9 +1055,9 @@ void dump_bfin_mem(struct pt_regs *fp) oops_in_progress)){ verbose_printk(KERN_NOTICE "Looks like this was a deferred error - sorry\n"); #ifndef CONFIG_DEBUG_HWERR - verbose_printk(KERN_NOTICE "The remaining message may be meaningless\n" - KERN_NOTICE "You should enable CONFIG_DEBUG_HWERR to get a" - " better idea where it came from\n"); + verbose_printk(KERN_NOTICE +"The remaining message may be meaningless\n" +"You should enable CONFIG_DEBUG_HWERR to get a better idea where it came from\n"); #else /* If we are handling only one peripheral interrupt * and current mm and pid are valid, and the last error @@ -1114,9 +1113,10 @@ void show_regs(struct pt_regs *fp) verbose_printk(KERN_NOTICE "%s", linux_banner); - verbose_printk(KERN_NOTICE "\n" KERN_NOTICE "SEQUENCER STATUS:\t\t%s\n", print_tainted()); + verbose_printk(KERN_NOTICE "\nSEQUENCER STATUS:\t\t%s\n", + print_tainted()); verbose_printk(KERN_NOTICE " SEQSTAT: %08lx IPEND: %04lx SYSCFG: %04lx\n", - (long)fp->seqstat, fp->ipend, fp->syscfg); + (long)fp->seqstat, fp->ipend, fp->syscfg); if ((fp->seqstat & SEQSTAT_EXCAUSE) == VEC_HWERR) { verbose_printk(KERN_NOTICE " HWERRCAUSE: 0x%lx\n", (fp->seqstat & SEQSTAT_HWERRCAUSE) >> 14); @@ -1184,7 +1184,7 @@ unlock: verbose_printk(KERN_NOTICE "ICPLB_FAULT_ADDR: %s\n", buf); } - verbose_printk(KERN_NOTICE "\n" KERN_NOTICE "PROCESSOR STATE:\n"); + verbose_printk(KERN_NOTICE "PROCESSOR STATE:\n"); verbose_printk(KERN_NOTICE " R0 : %08lx R1 : %08lx R2 : %08lx R3 : %08lx\n", fp->r0, fp->r1, fp->r2, fp->r3); verbose_printk(KERN_NOTICE " R4 : %08lx R5 : %08lx R6 : %08lx R7 : %08lx\n", diff --git a/arch/m68knommu/kernel/process.c b/arch/m68knommu/kernel/process.c index 1e96c6eb6312..8f8f4abab2ff 100644 --- a/arch/m68knommu/kernel/process.c +++ b/arch/m68knommu/kernel/process.c @@ -290,7 +290,7 @@ void dump(struct pt_regs *fp) unsigned char *tp; int i; - printk(KERN_EMERG "\n" KERN_EMERG "CURRENT PROCESS:\n" KERN_EMERG "\n"); + printk(KERN_EMERG "\nCURRENT PROCESS:\n\n"); printk(KERN_EMERG "COMM=%s PID=%d\n", current->comm, current->pid); if (current->mm) { @@ -301,8 +301,7 @@ void dump(struct pt_regs *fp) (int) current->mm->end_data, (int) current->mm->end_data, (int) current->mm->brk); - printk(KERN_EMERG "USER-STACK=%08x KERNEL-STACK=%08x\n" - KERN_EMERG "\n", + printk(KERN_EMERG "USER-STACK=%08x KERNEL-STACK=%08x\n\n", (int) current->mm->start_stack, (int)(((unsigned long) current) + THREAD_SIZE)); } @@ -313,35 +312,35 @@ void dump(struct pt_regs *fp) fp->d0, fp->d1, fp->d2, fp->d3); printk(KERN_EMERG "d4: %08lx d5: %08lx a0: %08lx a1: %08lx\n", fp->d4, fp->d5, fp->a0, fp->a1); - printk(KERN_EMERG "\n" KERN_EMERG "USP: %08x TRAPFRAME: %08x\n", + printk(KERN_EMERG "\nUSP: %08x TRAPFRAME: %08x\n", (unsigned int) rdusp(), (unsigned int) fp); - printk(KERN_EMERG "\n" KERN_EMERG "CODE:"); + printk(KERN_EMERG "\nCODE:"); tp = ((unsigned char *) fp->pc) - 0x20; for (sp = (unsigned long *) tp, i = 0; (i < 0x40); i += 4) { if ((i % 0x10) == 0) - printk("\n" KERN_EMERG "%08x: ", (int) (tp + i)); + printk(KERN_EMERG "%08x: ", (int) (tp + i)); printk("%08x ", (int) *sp++); } - printk("\n" KERN_EMERG "\n"); + printk(KERN_EMERG "\n"); printk(KERN_EMERG "KERNEL STACK:"); tp = ((unsigned char *) fp) - 0x40; for (sp = (unsigned long *) tp, i = 0; (i < 0xc0); i += 4) { if ((i % 0x10) == 0) - printk("\n" KERN_EMERG "%08x: ", (int) (tp + i)); + printk(KERN_EMERG "%08x: ", (int) (tp + i)); printk("%08x ", (int) *sp++); } - printk("\n" KERN_EMERG "\n"); + printk(KERN_EMERG "\n"); printk(KERN_EMERG "USER STACK:"); tp = (unsigned char *) (rdusp() - 0x10); for (sp = (unsigned long *) tp, i = 0; (i < 0x80); i += 4) { if ((i % 0x10) == 0) - printk("\n" KERN_EMERG "%08x: ", (int) (tp + i)); + printk(KERN_EMERG "%08x: ", (int) (tp + i)); printk("%08x ", (int) *sp++); } - printk("\n" KERN_EMERG "\n"); + printk(KERN_EMERG "\n"); } /* diff --git a/arch/m68knommu/kernel/traps.c b/arch/m68knommu/kernel/traps.c index 51d325343ab5..3739c8f657d7 100644 --- a/arch/m68knommu/kernel/traps.c +++ b/arch/m68knommu/kernel/traps.c @@ -111,7 +111,7 @@ static void print_this_address(unsigned long addr, int i) if (i % 5) printk(KERN_CONT " [%08lx] ", addr); else - printk(KERN_CONT "\n" KERN_EMERG " [%08lx] ", addr); + printk(KERN_EMERG " [%08lx] ", addr); i++; #endif } @@ -137,8 +137,8 @@ static void __show_stack(struct task_struct *task, unsigned long *stack) if (stack + 1 + i > endstack) break; if (i % 8 == 0) - printk("\n" KERN_EMERG " "); - printk(" %08lx", *(stack + i)); + printk(KERN_EMERG " "); + printk(KERN_CONT " %08lx", *(stack + i)); } printk("\n"); i = 0; diff --git a/arch/mn10300/kernel/traps.c b/arch/mn10300/kernel/traps.c index 681ad8c9e4fb..0dfdc5001124 100644 --- a/arch/mn10300/kernel/traps.c +++ b/arch/mn10300/kernel/traps.c @@ -136,8 +136,7 @@ void show_trace(unsigned long *sp) unsigned long *stack, addr, module_start, module_end; int i; - printk(KERN_EMERG "\n" - KERN_EMERG "Call Trace:"); + printk(KERN_EMERG "\nCall Trace:"); stack = sp; i = 0; @@ -153,7 +152,7 @@ void show_trace(unsigned long *sp) printk("\n"); #else if ((i % 6) == 0) - printk("\n" KERN_EMERG " "); + printk(KERN_EMERG " "); printk("[<%08lx>] ", addr); i++; #endif @@ -180,7 +179,7 @@ void show_stack(struct task_struct *task, unsigned long *sp) if (((long) stack & (THREAD_SIZE - 1)) == 0) break; if ((i % 8) == 0) - printk("\n" KERN_EMERG " "); + printk(KERN_EMERG " "); printk("%08lx ", *stack++); } @@ -264,8 +263,7 @@ void show_registers(struct pt_regs *regs) show_stack(current, (unsigned long *) sp); #if 0 - printk(KERN_EMERG "\n" - KERN_EMERG "Code: "); + printk(KERN_EMERG "\nCode: "); if (regs->pc < PAGE_OFFSET) goto bad; @@ -311,16 +309,14 @@ void die(const char *str, struct pt_regs *regs, enum exception_code code) { console_verbose(); spin_lock_irq(&die_lock); - printk(KERN_EMERG "\n" - KERN_EMERG "%s: %04x\n", + printk(KERN_EMERG "\n%s: %04x\n", str, code & 0xffff); show_registers(regs); if (regs->pc >= 0x02000000 && regs->pc < 0x04000000 && (regs->epsw & (EPSW_IM | EPSW_IE)) != (EPSW_IM | EPSW_IE)) { printk(KERN_EMERG "Exception in usermode interrupt handler\n"); - printk(KERN_EMERG "\n" - KERN_EMERG " Please connect to kernel debugger !!\n"); + printk(KERN_EMERG "\nPlease connect to kernel debugger !!\n"); asm volatile ("0: bra 0b"); } @@ -429,9 +425,8 @@ asmlinkage void io_bus_error(u32 bcberr, u32 bcbear, struct pt_regs *regs) { console_verbose(); - printk(KERN_EMERG "\n" - KERN_EMERG "Asynchronous I/O Bus Error\n" - KERN_EMERG "==========================\n"); + printk(KERN_EMERG "Asynchronous I/O Bus Error\n"); + printk(KERN_EMERG "==========================\n"); if (bcberr & BCBERR_BEME) printk(KERN_EMERG "- Multiple recorded errors\n"); diff --git a/arch/parisc/kernel/process.c b/arch/parisc/kernel/process.c index 61c07078c072..1f3aa8db0203 100644 --- a/arch/parisc/kernel/process.c +++ b/arch/parisc/kernel/process.c @@ -156,7 +156,7 @@ void machine_power_off(void) * software. The user has to press the button himself. */ printk(KERN_EMERG "System shut down completed.\n" - KERN_EMERG "Please power this system off now."); + "Please power this system off now."); } void (*pm_power_off)(void) = machine_power_off; diff --git a/arch/parisc/kernel/traps.c b/arch/parisc/kernel/traps.c index c32f5d6d778e..528f0ff9b273 100644 --- a/arch/parisc/kernel/traps.c +++ b/arch/parisc/kernel/traps.c @@ -250,15 +250,14 @@ void die_if_kernel(char *str, struct pt_regs *regs, long err) oops_enter(); /* Amuse the user in a SPARC fashion */ - if (err) printk( -KERN_CRIT " _______________________________ \n" -KERN_CRIT " < Your System ate a SPARC! Gah! >\n" -KERN_CRIT " ------------------------------- \n" -KERN_CRIT " \\ ^__^\n" -KERN_CRIT " \\ (xx)\\_______\n" -KERN_CRIT " (__)\\ )\\/\\\n" -KERN_CRIT " U ||----w |\n" -KERN_CRIT " || ||\n"); + if (err) printk(KERN_CRIT + " _______________________________ \n" + " < Your System ate a SPARC! Gah! >\n" + " ------------------------------- \n" + " \\ ^__^\n" + " (__)\\ )\\/\\\n" + " U ||----w |\n" + " || ||\n"); /* unlock the pdc lock if necessary */ pdc_emergency_unlock(); @@ -797,7 +796,8 @@ void notrace handle_interruption(int code, struct pt_regs *regs) else printk(KERN_DEBUG "User Fault (long pointer) (fault %d) ", code); - printk("pid=%d command='%s'\n", task_pid_nr(current), current->comm); + printk(KERN_CONT "pid=%d command='%s'\n", + task_pid_nr(current), current->comm); show_regs(regs); #endif si.si_signo = SIGSEGV; diff --git a/arch/um/kernel/sysrq.c b/arch/um/kernel/sysrq.c index 56d43d0a3960..0960de54495a 100644 --- a/arch/um/kernel/sysrq.c +++ b/arch/um/kernel/sysrq.c @@ -70,8 +70,8 @@ void show_stack(struct task_struct *task, unsigned long *esp) if (kstack_end(stack)) break; if (i && ((i % 8) == 0)) - printk("\n" KERN_INFO " "); - printk("%08lx ", *stack++); + printk(KERN_INFO " "); + printk(KERN_CONT "%08lx ", *stack++); } show_trace(task, esp); diff --git a/arch/x86/kernel/apic/io_apic.c b/arch/x86/kernel/apic/io_apic.c index 8fd1efb5a0bd..90b5e6efa938 100644 --- a/arch/x86/kernel/apic/io_apic.c +++ b/arch/x86/kernel/apic/io_apic.c @@ -1739,7 +1739,6 @@ __apicdebuginit(void) print_local_APIC(void *dummy) if (apic_verbosity == APIC_QUIET) return; - printk(KERN_DEBUG "\n"); printk(KERN_DEBUG "printing local APIC contents on CPU#%d/%d:\n", smp_processor_id(), hard_smp_processor_id()); v = apic_read(APIC_ID); diff --git a/arch/x86/kernel/cpu/cpufreq/powernow-k8.c b/arch/x86/kernel/cpu/cpufreq/powernow-k8.c index ae068f59603f..2a50ef891000 100644 --- a/arch/x86/kernel/cpu/cpufreq/powernow-k8.c +++ b/arch/x86/kernel/cpu/cpufreq/powernow-k8.c @@ -1259,7 +1259,7 @@ static int __cpuinit powernowk8_cpu_init(struct cpufreq_policy *pol) { static const char ACPI_PSS_BIOS_BUG_MSG[] = KERN_ERR FW_BUG PFX "No compatible ACPI _PSS objects found.\n" - KERN_ERR FW_BUG PFX "Try again with latest BIOS.\n"; + FW_BUG PFX "Try again with latest BIOS.\n"; struct powernow_k8_data *data; struct init_on_cpu init_on_cpu; int rc; diff --git a/arch/x86/kernel/cpu/mcheck/mce.c b/arch/x86/kernel/cpu/mcheck/mce.c index af425b83202b..484c1e5f658e 100644 --- a/arch/x86/kernel/cpu/mcheck/mce.c +++ b/arch/x86/kernel/cpu/mcheck/mce.c @@ -194,14 +194,14 @@ static void print_mce(struct mce *m) m->cs, m->ip); if (m->cs == __KERNEL_CS) print_symbol("{%s}", m->ip); - printk("\n"); + printk(KERN_CONT "\n"); } printk(KERN_EMERG "TSC %llx ", m->tsc); if (m->addr) - printk("ADDR %llx ", m->addr); + printk(KERN_CONT "ADDR %llx ", m->addr); if (m->misc) - printk("MISC %llx ", m->misc); - printk("\n"); + printk(KERN_CONT "MISC %llx ", m->misc); + printk(KERN_CONT "\n"); printk(KERN_EMERG "PROCESSOR %u:%x TIME %llu SOCKET %u APIC %x\n", m->cpuvendor, m->cpuid, m->time, m->socketid, m->apicid); @@ -209,13 +209,13 @@ static void print_mce(struct mce *m) static void print_mce_head(void) { - printk(KERN_EMERG "\n" KERN_EMERG "HARDWARE ERROR\n"); + printk(KERN_EMERG "\nHARDWARE ERROR\n"); } static void print_mce_tail(void) { printk(KERN_EMERG "This is not a software problem!\n" - KERN_EMERG "Run through mcelog --ascii to decode and contact your hardware vendor\n"); + "Run through mcelog --ascii to decode and contact your hardware vendor\n"); } #define PANIC_TIMEOUT 5 /* 5 seconds */ diff --git a/arch/x86/kernel/e820.c b/arch/x86/kernel/e820.c index c4ca89d9aaf4..5cb5725b2bae 100644 --- a/arch/x86/kernel/e820.c +++ b/arch/x86/kernel/e820.c @@ -627,10 +627,9 @@ __init void e820_setup_gap(void) #ifdef CONFIG_X86_64 if (!found) { gapstart = (max_pfn << PAGE_SHIFT) + 1024*1024; - printk(KERN_ERR "PCI: Warning: Cannot find a gap in the 32bit " - "address range\n" - KERN_ERR "PCI: Unassigned devices with 32bit resource " - "registers may break!\n"); + printk(KERN_ERR + "PCI: Warning: Cannot find a gap in the 32bit address range\n" + "PCI: Unassigned devices with 32bit resource registers may break!\n"); } #endif diff --git a/arch/x86/kernel/pci-gart_64.c b/arch/x86/kernel/pci-gart_64.c index cfd9f9063896..d2e56b8f48e7 100644 --- a/arch/x86/kernel/pci-gart_64.c +++ b/arch/x86/kernel/pci-gart_64.c @@ -675,7 +675,7 @@ static __init int init_k8_gatt(struct agp_kern_info *info) nommu: /* Should not happen anymore */ printk(KERN_WARNING "PCI-DMA: More than 4GB of RAM and no IOMMU\n" - KERN_WARNING "falling back to iommu=soft.\n"); + "falling back to iommu=soft.\n"); return -1; } diff --git a/arch/x86/mm/fault.c b/arch/x86/mm/fault.c index 78a5fff857be..85307cc6e45f 100644 --- a/arch/x86/mm/fault.c +++ b/arch/x86/mm/fault.c @@ -426,10 +426,11 @@ static noinline int vmalloc_fault(unsigned long address) } static const char errata93_warning[] = -KERN_ERR "******* Your BIOS seems to not contain a fix for K8 errata #93\n" -KERN_ERR "******* Working around it, but it may cause SEGVs or burn power.\n" -KERN_ERR "******* Please consider a BIOS update.\n" -KERN_ERR "******* Disabling USB legacy in the BIOS may also help.\n"; +KERN_ERR +"******* Your BIOS seems to not contain a fix for K8 errata #93\n" +"******* Working around it, but it may cause SEGVs or burn power.\n" +"******* Please consider a BIOS update.\n" +"******* Disabling USB legacy in the BIOS may also help.\n"; /* * No vm86 mode in 64-bit mode: diff --git a/arch/xtensa/kernel/traps.c b/arch/xtensa/kernel/traps.c index ba9ab9349782..e64efac3b9db 100644 --- a/arch/xtensa/kernel/traps.c +++ b/arch/xtensa/kernel/traps.c @@ -354,10 +354,10 @@ void show_regs(struct pt_regs * regs) for (i = 0; i < 16; i++) { if ((i % 8) == 0) - printk ("\n" KERN_INFO "a%02d: ", i); - printk("%08lx ", regs->areg[i]); + printk(KERN_INFO "a%02d:", i); + printk(KERN_CONT " %08lx", regs->areg[i]); } - printk("\n"); + printk(KERN_CONT "\n"); printk("pc: %08lx, ps: %08lx, depc: %08lx, excvaddr: %08lx\n", regs->pc, regs->ps, regs->depc, regs->excvaddr); diff --git a/drivers/block/amiflop.c b/drivers/block/amiflop.c index 9c6e5b0fe894..2f07b7c99a95 100644 --- a/drivers/block/amiflop.c +++ b/drivers/block/amiflop.c @@ -1645,7 +1645,7 @@ static int __init fd_probe_drives(void) { int drive,drives,nomem; - printk(KERN_INFO "FD: probing units\n" KERN_INFO "found "); + printk(KERN_INFO "FD: probing units\nfound "); drives=0; nomem=0; for(drive=0;drivedev, " ctrl: %.8x seccnt/cmd: %.4x ver:%.4x\n" - KERN_INFO " status:%.8x mpu_lba:%.8x busmode:%4x\n" - KERN_INFO " error: %.8x cfg_lba:%.8x fatstat:%.4x\n", + dev_info(ace->dev, + " ctrl: %.8x seccnt/cmd: %.4x ver:%.4x\n" + " status:%.8x mpu_lba:%.8x busmode:%4x\n" + " error: %.8x cfg_lba:%.8x fatstat:%.4x\n", ace_in32(ace, ACE_CTRL), ace_in(ace, ACE_SECCNTCMD), ace_in(ace, ACE_VERSION), diff --git a/drivers/char/hw_random/intel-rng.c b/drivers/char/hw_random/intel-rng.c index 5dcbe603eca2..91b53eb1c053 100644 --- a/drivers/char/hw_random/intel-rng.c +++ b/drivers/char/hw_random/intel-rng.c @@ -305,10 +305,11 @@ static int __init intel_init_hw_struct(struct intel_rng_hw *intel_rng_hw, (BIOS_CNTL_LOCK_ENABLE_MASK|BIOS_CNTL_WRITE_ENABLE_MASK)) == BIOS_CNTL_LOCK_ENABLE_MASK) { static __initdata /*const*/ char warning[] = - KERN_WARNING PFX "Firmware space is locked read-only. If you can't or\n" - KERN_WARNING PFX "don't want to disable this in firmware setup, and if\n" - KERN_WARNING PFX "you are certain that your system has a functional\n" - KERN_WARNING PFX "RNG, try using the 'no_fwh_detect' option.\n"; + KERN_WARNING +PFX "Firmware space is locked read-only. If you can't or\n" +PFX "don't want to disable this in firmware setup, and if\n" +PFX "you are certain that your system has a functional\n" +PFX "RNG, try using the 'no_fwh_detect' option.\n"; if (no_fwh_detect) return -ENODEV; diff --git a/drivers/char/isicom.c b/drivers/char/isicom.c index 4159292e35cf..621d1184673c 100644 --- a/drivers/char/isicom.c +++ b/drivers/char/isicom.c @@ -1478,10 +1478,10 @@ static int __devinit load_firmware(struct pci_dev *pdev, status = inw(base + 0x4); if (status != 0) { dev_warn(&pdev->dev, "Card%d rejected load header:\n" - KERN_WARNING "Address:0x%x\n" - KERN_WARNING "Count:0x%x\n" - KERN_WARNING "Status:0x%x\n", - index + 1, frame->addr, frame->count, status); + "Address:0x%x\n" + "Count:0x%x\n" + "Status:0x%x\n", + index + 1, frame->addr, frame->count, status); goto errrelfw; } outsw(base, frame->data, word_count); @@ -1526,10 +1526,10 @@ static int __devinit load_firmware(struct pci_dev *pdev, status = inw(base + 0x4); if (status != 0) { dev_warn(&pdev->dev, "Card%d rejected verify header:\n" - KERN_WARNING "Address:0x%x\n" - KERN_WARNING "Count:0x%x\n" - KERN_WARNING "Status: 0x%x\n", - index + 1, frame->addr, frame->count, status); + "Address:0x%x\n" + "Count:0x%x\n" + "Status: 0x%x\n", + index + 1, frame->addr, frame->count, status); goto errrelfw; } diff --git a/drivers/i2c/busses/i2c-ibm_iic.c b/drivers/i2c/busses/i2c-ibm_iic.c index e4476743f203..b1bc6e277d2a 100644 --- a/drivers/i2c/busses/i2c-ibm_iic.c +++ b/drivers/i2c/busses/i2c-ibm_iic.c @@ -85,10 +85,11 @@ static void dump_iic_regs(const char* header, struct ibm_iic_private* dev) { volatile struct iic_regs __iomem *iic = dev->vaddr; printk(KERN_DEBUG "ibm-iic%d: %s\n", dev->idx, header); - printk(KERN_DEBUG " cntl = 0x%02x, mdcntl = 0x%02x\n" - KERN_DEBUG " sts = 0x%02x, extsts = 0x%02x\n" - KERN_DEBUG " clkdiv = 0x%02x, xfrcnt = 0x%02x\n" - KERN_DEBUG " xtcntlss = 0x%02x, directcntl = 0x%02x\n", + printk(KERN_DEBUG + " cntl = 0x%02x, mdcntl = 0x%02x\n" + " sts = 0x%02x, extsts = 0x%02x\n" + " clkdiv = 0x%02x, xfrcnt = 0x%02x\n" + " xtcntlss = 0x%02x, directcntl = 0x%02x\n", in_8(&iic->cntl), in_8(&iic->mdcntl), in_8(&iic->sts), in_8(&iic->extsts), in_8(&iic->clkdiv), in_8(&iic->xfrcnt), in_8(&iic->xtcntlss), in_8(&iic->directcntl)); diff --git a/drivers/md/md.c b/drivers/md/md.c index 0f4a70c43ffc..d4351ff0849f 100644 --- a/drivers/md/md.c +++ b/drivers/md/md.c @@ -1756,9 +1756,10 @@ static void print_sb_1(struct mdp_superblock_1 *sb) __u8 *uuid; uuid = sb->set_uuid; - printk(KERN_INFO "md: SB: (V:%u) (F:0x%08x) Array-ID:<%02x%02x%02x%02x" - ":%02x%02x:%02x%02x:%02x%02x:%02x%02x%02x%02x%02x%02x>\n" - KERN_INFO "md: Name: \"%s\" CT:%llu\n", + printk(KERN_INFO + "md: SB: (V:%u) (F:0x%08x) Array-ID:<%02x%02x%02x%02x" + ":%02x%02x:%02x%02x:%02x%02x:%02x%02x%02x%02x%02x%02x>\n" + "md: Name: \"%s\" CT:%llu\n", le32_to_cpu(sb->major_version), le32_to_cpu(sb->feature_map), uuid[0], uuid[1], uuid[2], uuid[3], @@ -1770,12 +1771,13 @@ static void print_sb_1(struct mdp_superblock_1 *sb) & MD_SUPERBLOCK_1_TIME_SEC_MASK); uuid = sb->device_uuid; - printk(KERN_INFO "md: L%u SZ%llu RD:%u LO:%u CS:%u DO:%llu DS:%llu SO:%llu" + printk(KERN_INFO + "md: L%u SZ%llu RD:%u LO:%u CS:%u DO:%llu DS:%llu SO:%llu" " RO:%llu\n" - KERN_INFO "md: Dev:%08x UUID: %02x%02x%02x%02x:%02x%02x:%02x%02x:%02x%02x" - ":%02x%02x%02x%02x%02x%02x\n" - KERN_INFO "md: (F:0x%08x) UT:%llu Events:%llu ResyncOffset:%llu CSUM:0x%08x\n" - KERN_INFO "md: (MaxDev:%u) \n", + "md: Dev:%08x UUID: %02x%02x%02x%02x:%02x%02x:%02x%02x:%02x%02x" + ":%02x%02x%02x%02x%02x%02x\n" + "md: (F:0x%08x) UT:%llu Events:%llu ResyncOffset:%llu CSUM:0x%08x\n" + "md: (MaxDev:%u) \n", le32_to_cpu(sb->level), (unsigned long long)le64_to_cpu(sb->size), le32_to_cpu(sb->raid_disks), diff --git a/drivers/misc/sgi-xp/xpnet.c b/drivers/misc/sgi-xp/xpnet.c index 8d1c60a3f0df..5d778ec8cdb2 100644 --- a/drivers/misc/sgi-xp/xpnet.c +++ b/drivers/misc/sgi-xp/xpnet.c @@ -235,7 +235,7 @@ xpnet_receive(short partid, int channel, struct xpnet_message *msg) skb->ip_summed = CHECKSUM_UNNECESSARY; dev_dbg(xpnet, "passing skb to network layer\n" - KERN_DEBUG "\tskb->head=0x%p skb->data=0x%p skb->tail=0x%p " + "\tskb->head=0x%p skb->data=0x%p skb->tail=0x%p " "skb->end=0x%p skb->len=%d\n", (void *)skb->head, (void *)skb->data, skb_tail_pointer(skb), skb_end_pointer(skb), skb->len); @@ -399,7 +399,7 @@ xpnet_send(struct sk_buff *skb, struct xpnet_pending_msg *queued_msg, msg->buf_pa = xp_pa((void *)start_addr); dev_dbg(xpnet, "sending XPC message to %d:%d\n" - KERN_DEBUG "msg->buf_pa=0x%lx, msg->size=%u, " + "msg->buf_pa=0x%lx, msg->size=%u, " "msg->leadin_ignore=%u, msg->tailout_ignore=%u\n", dest_partid, XPC_NET_CHANNEL, msg->buf_pa, msg->size, msg->leadin_ignore, msg->tailout_ignore); diff --git a/drivers/net/a2065.c b/drivers/net/a2065.c index 85a18175730b..08787f5a22a3 100644 --- a/drivers/net/a2065.c +++ b/drivers/net/a2065.c @@ -569,16 +569,8 @@ static int lance_start_xmit (struct sk_buff *skb, struct net_device *dev) #ifdef DEBUG_DRIVER /* dump the packet */ - { - int i; - - for (i = 0; i < 64; i++) { - if ((i % 16) == 0) - printk("\n" KERN_DEBUG); - printk ("%2.2x ", skb->data [i]); - } - printk("\n"); - } + print_hex_dump(KERN_DEBUG, "skb->data: ", DUMP_PREFIX_NONE, + 16, 1, skb->data, 64, true); #endif entry = lp->tx_new & lp->tx_ring_mod_mask; ib->btx_ring [entry].length = (-skblen) | 0xf000; diff --git a/drivers/net/arcnet/arcnet.c b/drivers/net/arcnet/arcnet.c index d6d4ab3b430c..7d227cdab9f8 100644 --- a/drivers/net/arcnet/arcnet.c +++ b/drivers/net/arcnet/arcnet.c @@ -158,15 +158,12 @@ module_exit(arcnet_exit); void arcnet_dump_skb(struct net_device *dev, struct sk_buff *skb, char *desc) { - int i; + char hdr[32]; - printk(KERN_DEBUG "%6s: skb dump (%s) follows:", dev->name, desc); - for (i = 0; i < skb->len; i++) { - if (i % 16 == 0) - printk("\n" KERN_DEBUG "[%04X] ", i); - printk("%02X ", ((u_char *) skb->data)[i]); - } - printk("\n"); + /* dump the packet */ + snprintf(hdr, sizeof(hdr), "%6s:%s skb->data:", dev->name, desc); + print_hex_dump(KERN_DEBUG, hdr, DUMP_PREFIX_OFFSET, + 16, 1, skb->data, skb->len, true); } EXPORT_SYMBOL(arcnet_dump_skb); @@ -184,6 +181,7 @@ static void arcnet_dump_packet(struct net_device *dev, int bufnum, int i, length; unsigned long flags = 0; static uint8_t buf[512]; + char hdr[32]; /* hw.copy_from_card expects IRQ context so take the IRQ lock to keep it single threaded */ @@ -197,14 +195,10 @@ static void arcnet_dump_packet(struct net_device *dev, int bufnum, /* if the offset[0] byte is nonzero, this is a 256-byte packet */ length = (buf[2] ? 256 : 512); - printk(KERN_DEBUG "%6s: packet dump (%s) follows:", dev->name, desc); - for (i = 0; i < length; i++) { - if (i % 16 == 0) - printk("\n" KERN_DEBUG "[%04X] ", i); - printk("%02X ", buf[i]); - } - printk("\n"); - + /* dump the packet */ + snprintf(hdr, sizeof(hdr), "%6s:%s packet dump:", dev->name, desc); + print_hex_dump(KERN_DEBUG, hdr, DUMP_PREFIX_OFFSET, + 16, 1, buf, length, true); } #else diff --git a/drivers/net/bmac.c b/drivers/net/bmac.c index 9578a3dfac01..a76315dc7767 100644 --- a/drivers/net/bmac.c +++ b/drivers/net/bmac.c @@ -428,10 +428,11 @@ bmac_init_phy(struct net_device *dev) printk(KERN_DEBUG "phy registers:"); for (addr = 0; addr < 32; ++addr) { if ((addr & 7) == 0) - printk("\n" KERN_DEBUG); - printk(" %.4x", bmac_mif_read(dev, addr)); + printk(KERN_DEBUG); + printk(KERN_CONT " %.4x", bmac_mif_read(dev, addr)); } - printk("\n"); + print(KERN_CONT "\n"); + if (bp->is_bmac_plus) { unsigned int capable, ctrl; diff --git a/drivers/net/bnx2x_main.c b/drivers/net/bnx2x_main.c index 6c67be679764..c36a5f33739f 100644 --- a/drivers/net/bnx2x_main.c +++ b/drivers/net/bnx2x_main.c @@ -484,8 +484,9 @@ static void bnx2x_fw_dump(struct bnx2x *bp) mark = REG_RD(bp, MCP_REG_MCPR_SCRATCH + 0xf104); mark = ((mark + 0x3) & ~0x3); - printk(KERN_ERR PFX "begin fw dump (mark 0x%x)\n" KERN_ERR, mark); + printk(KERN_ERR PFX "begin fw dump (mark 0x%x)\n", mark); + printk(KERN_ERR PFX); for (offset = mark - 0x08000000; offset <= 0xF900; offset += 0x8*4) { for (word = 0; word < 8; word++) data[word] = htonl(REG_RD(bp, MCP_REG_MCPR_SCRATCH + @@ -500,7 +501,7 @@ static void bnx2x_fw_dump(struct bnx2x *bp) data[8] = 0x0; printk(KERN_CONT "%s", (char *)data); } - printk("\n" KERN_ERR PFX "end of fw dump\n"); + printk(KERN_ERR PFX "end of fw dump\n"); } static void bnx2x_panic_dump(struct bnx2x *bp) @@ -7354,7 +7355,7 @@ static void bnx2x_reset_task(struct work_struct *work) #ifdef BNX2X_STOP_ON_ERROR BNX2X_ERR("reset task called but STOP_ON_ERROR defined" " so reset not done to allow debug dump,\n" - KERN_ERR " you will need to reboot when done\n"); + " you will need to reboot when done\n"); return; #endif diff --git a/drivers/net/dl2k.c b/drivers/net/dl2k.c index 895d72143ee0..4b6a219fecea 100644 --- a/drivers/net/dl2k.c +++ b/drivers/net/dl2k.c @@ -268,8 +268,9 @@ rio_probe1 (struct pci_dev *pdev, const struct pci_device_id *ent) printk(KERN_INFO "tx_coalesce:\t%d packets\n", tx_coalesce); if (np->coalesce) - printk(KERN_INFO "rx_coalesce:\t%d packets\n" - KERN_INFO "rx_timeout: \t%d ns\n", + printk(KERN_INFO + "rx_coalesce:\t%d packets\n" + "rx_timeout: \t%d ns\n", np->rx_coalesce, np->rx_timeout*640); if (np->vlan) printk(KERN_INFO "vlan(id):\t%d\n", np->vlan); @@ -1522,9 +1523,9 @@ mii_get_media (struct net_device *dev) printk (KERN_INFO "Operating at 10 Mbps, "); } if (bmcr & MII_BMCR_DUPLEX_MODE) { - printk ("Full duplex\n"); + printk (KERN_CONT "Full duplex\n"); } else { - printk ("Half duplex\n"); + printk (KERN_CONT "Half duplex\n"); } } if (np->tx_flow) @@ -1614,9 +1615,9 @@ mii_set_media (struct net_device *dev) } if (np->full_duplex) { bmcr |= MII_BMCR_DUPLEX_MODE; - printk ("Full duplex\n"); + printk (KERN_CONT "Full duplex\n"); } else { - printk ("Half duplex\n"); + printk (KERN_CONT "Half duplex\n"); } #if 0 /* Set 1000BaseT Master/Slave setting */ @@ -1669,9 +1670,9 @@ mii_get_media_pcs (struct net_device *dev) __u16 bmcr = mii_read (dev, phy_addr, PCS_BMCR); printk (KERN_INFO "Operating at 1000 Mbps, "); if (bmcr & MII_BMCR_DUPLEX_MODE) { - printk ("Full duplex\n"); + printk (KERN_CONT "Full duplex\n"); } else { - printk ("Half duplex\n"); + printk (KERN_CONT "Half duplex\n"); } } if (np->tx_flow) diff --git a/drivers/net/epic100.c b/drivers/net/epic100.c index b60e27dfcfa7..88d7ebf31220 100644 --- a/drivers/net/epic100.c +++ b/drivers/net/epic100.c @@ -338,8 +338,7 @@ static int __devinit epic_init_one (struct pci_dev *pdev, #ifndef MODULE static int printed_version; if (!printed_version++) - printk (KERN_INFO "%s" KERN_INFO "%s", - version, version2); + printk(KERN_INFO "%s%s", version, version2); #endif card_idx++; @@ -1600,7 +1599,7 @@ static int __init epic_init (void) { /* when a module, this is printed whether or not devices are found in probe */ #ifdef MODULE - printk (KERN_INFO "%s" KERN_INFO "%s", + printk (KERN_INFO "%s%s", version, version2); #endif diff --git a/drivers/net/fealnx.c b/drivers/net/fealnx.c index 891be28a7d4f..053fb49820b9 100644 --- a/drivers/net/fealnx.c +++ b/drivers/net/fealnx.c @@ -1209,17 +1209,20 @@ static void fealnx_tx_timeout(struct net_device *dev) unsigned long flags; int i; - printk(KERN_WARNING "%s: Transmit timed out, status %8.8x," - " resetting...\n", dev->name, ioread32(ioaddr + ISR)); + printk(KERN_WARNING + "%s: Transmit timed out, status %8.8x, resetting...\n", + dev->name, ioread32(ioaddr + ISR)); { printk(KERN_DEBUG " Rx ring %p: ", np->rx_ring); for (i = 0; i < RX_RING_SIZE; i++) - printk(" %8.8x", (unsigned int) np->rx_ring[i].status); - printk("\n" KERN_DEBUG " Tx ring %p: ", np->tx_ring); + printk(PR_CONT " %8.8x", + (unsigned int) np->rx_ring[i].status); + printk(KERN_CONT "\n"); + printk(KERN_DEBUG " Tx ring %p: ", np->tx_ring); for (i = 0; i < TX_RING_SIZE; i++) - printk(" %4.4x", np->tx_ring[i].status); - printk("\n"); + printk(PR_CONT " %4.4x", np->tx_ring[i].status); + printk(PR_CONT "\n"); } spin_lock_irqsave(&np->lock, flags); diff --git a/drivers/net/hamachi.c b/drivers/net/hamachi.c index 9d5b62cb30f7..d62378cbc149 100644 --- a/drivers/net/hamachi.c +++ b/drivers/net/hamachi.c @@ -173,8 +173,8 @@ static int tx_params[MAX_UNITS] = {-1, -1, -1, -1, -1, -1, -1, -1}; static const char version[] __devinitconst = KERN_INFO DRV_NAME ".c:v" DRV_VERSION " " DRV_RELDATE " Written by Donald Becker\n" -KERN_INFO " Some modifications by Eric kasten \n" -KERN_INFO " Further modifications by Keith Underwood \n"; +" Some modifications by Eric kasten \n" +" Further modifications by Keith Underwood \n"; /* IP_MF appears to be only defined in , however, @@ -1080,11 +1080,14 @@ static void hamachi_tx_timeout(struct net_device *dev) { printk(KERN_DEBUG " Rx ring %p: ", hmp->rx_ring); for (i = 0; i < RX_RING_SIZE; i++) - printk(" %8.8x", le32_to_cpu(hmp->rx_ring[i].status_n_length)); - printk("\n"KERN_DEBUG" Tx ring %p: ", hmp->tx_ring); + printk(KERN_CONT " %8.8x", + le32_to_cpu(hmp->rx_ring[i].status_n_length)); + printk(KERN_CONT "\n"); + printk(KERN_DEBUG" Tx ring %p: ", hmp->tx_ring); for (i = 0; i < TX_RING_SIZE; i++) - printk(" %4.4x", le32_to_cpu(hmp->tx_ring[i].status_n_length)); - printk("\n"); + printk(KERN_CONT " %4.4x", + le32_to_cpu(hmp->tx_ring[i].status_n_length)); + printk(KERN_CONT "\n"); } /* Reinit the hardware and make sure the Rx and Tx processes @@ -1753,13 +1756,13 @@ static int hamachi_close(struct net_device *dev) #ifdef __i386__ if (hamachi_debug > 2) { - printk("\n"KERN_DEBUG" Tx ring at %8.8x:\n", + printk(KERN_DEBUG " Tx ring at %8.8x:\n", (int)hmp->tx_ring_dma); for (i = 0; i < TX_RING_SIZE; i++) - printk(" %c #%d desc. %8.8x %8.8x.\n", + printk(KERN_DEBUG " %c #%d desc. %8.8x %8.8x.\n", readl(ioaddr + TxCurPtr) == (long)&hmp->tx_ring[i] ? '>' : ' ', i, hmp->tx_ring[i].status_n_length, hmp->tx_ring[i].addr); - printk("\n"KERN_DEBUG " Rx ring %8.8x:\n", + printk(KERN_DEBUG " Rx ring %8.8x:\n", (int)hmp->rx_ring_dma); for (i = 0; i < RX_RING_SIZE; i++) { printk(KERN_DEBUG " %c #%d desc. %4.4x %8.8x\n", @@ -1770,7 +1773,7 @@ static int hamachi_close(struct net_device *dev) u16 *addr = (u16 *) hmp->rx_skbuff[i]->data; int j; - + printk(KERN_DEBUG "Addr: "); for (j = 0; j < 0x50; j++) printk(" %4.4x", addr[j]); printk("\n"); diff --git a/drivers/net/hamradio/baycom_epp.c b/drivers/net/hamradio/baycom_epp.c index 5e4b7afd0683..352703255bba 100644 --- a/drivers/net/hamradio/baycom_epp.c +++ b/drivers/net/hamradio/baycom_epp.c @@ -68,7 +68,7 @@ static const char paranoia_str[] = KERN_ERR static const char bc_drvname[] = "baycom_epp"; static const char bc_drvinfo[] = KERN_INFO "baycom_epp: (C) 1998-2000 Thomas Sailer, HB9JNX/AE4WA\n" -KERN_INFO "baycom_epp: version 0.7 compiled " __TIME__ " " __DATE__ "\n"; +"baycom_epp: version 0.7 compiled " __TIME__ " " __DATE__ "\n"; /* --------------------------------------------------------------------- */ diff --git a/drivers/net/hamradio/baycom_par.c b/drivers/net/hamradio/baycom_par.c index 2e6fc4dc74b1..5f5af9a606f8 100644 --- a/drivers/net/hamradio/baycom_par.c +++ b/drivers/net/hamradio/baycom_par.c @@ -102,7 +102,7 @@ static const char bc_drvname[] = "baycom_par"; static const char bc_drvinfo[] = KERN_INFO "baycom_par: (C) 1996-2000 Thomas Sailer, HB9JNX/AE4WA\n" -KERN_INFO "baycom_par: version 0.9 compiled " __TIME__ " " __DATE__ "\n"; +"baycom_par: version 0.9 compiled " __TIME__ " " __DATE__ "\n"; /* --------------------------------------------------------------------- */ diff --git a/drivers/net/hamradio/baycom_ser_fdx.c b/drivers/net/hamradio/baycom_ser_fdx.c index b6a816e60c0f..aa4488e871b2 100644 --- a/drivers/net/hamradio/baycom_ser_fdx.c +++ b/drivers/net/hamradio/baycom_ser_fdx.c @@ -91,7 +91,7 @@ static const char bc_drvname[] = "baycom_ser_fdx"; static const char bc_drvinfo[] = KERN_INFO "baycom_ser_fdx: (C) 1996-2000 Thomas Sailer, HB9JNX/AE4WA\n" -KERN_INFO "baycom_ser_fdx: version 0.10 compiled " __TIME__ " " __DATE__ "\n"; +"baycom_ser_fdx: version 0.10 compiled " __TIME__ " " __DATE__ "\n"; /* --------------------------------------------------------------------- */ diff --git a/drivers/net/hamradio/baycom_ser_hdx.c b/drivers/net/hamradio/baycom_ser_hdx.c index 3bcc57acbe6d..88c593596020 100644 --- a/drivers/net/hamradio/baycom_ser_hdx.c +++ b/drivers/net/hamradio/baycom_ser_hdx.c @@ -79,7 +79,7 @@ static const char bc_drvname[] = "baycom_ser_hdx"; static const char bc_drvinfo[] = KERN_INFO "baycom_ser_hdx: (C) 1996-2000 Thomas Sailer, HB9JNX/AE4WA\n" -KERN_INFO "baycom_ser_hdx: version 0.10 compiled " __TIME__ " " __DATE__ "\n"; +"baycom_ser_hdx: version 0.10 compiled " __TIME__ " " __DATE__ "\n"; /* --------------------------------------------------------------------- */ diff --git a/drivers/net/natsemi.c b/drivers/net/natsemi.c index c9bfe4eea189..78c088331f57 100644 --- a/drivers/net/natsemi.c +++ b/drivers/net/natsemi.c @@ -130,8 +130,8 @@ static int full_duplex[MAX_UNITS]; static const char version[] __devinitconst = KERN_INFO DRV_NAME " dp8381x driver, version " DRV_VERSION ", " DRV_RELDATE "\n" - KERN_INFO " originally by Donald Becker \n" - KERN_INFO " 2.4.x kernel port by Jeff Garzik, Tjeerd Mulder\n"; + " originally by Donald Becker \n" + " 2.4.x kernel port by Jeff Garzik, Tjeerd Mulder\n"; MODULE_AUTHOR("Donald Becker "); MODULE_DESCRIPTION("National Semiconductor DP8381x series PCI Ethernet driver"); diff --git a/drivers/net/ne.c b/drivers/net/ne.c index 5c3e242428f1..992dbfffdb05 100644 --- a/drivers/net/ne.c +++ b/drivers/net/ne.c @@ -321,7 +321,7 @@ static int __init ne_probe1(struct net_device *dev, unsigned long ioaddr) } if (ei_debug && version_printed++ == 0) - printk(KERN_INFO "%s" KERN_INFO "%s", version1, version2); + printk(KERN_INFO "%s%s", version1, version2); printk(KERN_INFO "NE*000 ethercard probe at %#3lx:", ioaddr); diff --git a/drivers/net/pci-skeleton.c b/drivers/net/pci-skeleton.c index 8c1f6988f398..89f7b2ad5231 100644 --- a/drivers/net/pci-skeleton.c +++ b/drivers/net/pci-skeleton.c @@ -105,7 +105,7 @@ IVc. Errata static char version[] __devinitdata = KERN_INFO NETDRV_DRIVER_LOAD_MSG "\n" -KERN_INFO " Support available from http://foo.com/bar/baz.html\n"; +" Support available from http://foo.com/bar/baz.html\n"; /* define to 1 to enable PIO instead of MMIO */ #undef USE_IO_OPS diff --git a/drivers/net/pcmcia/ibmtr_cs.c b/drivers/net/pcmcia/ibmtr_cs.c index f51944b28cfa..06618af1a468 100644 --- a/drivers/net/pcmcia/ibmtr_cs.c +++ b/drivers/net/pcmcia/ibmtr_cs.c @@ -298,14 +298,11 @@ static int __devinit ibmtr_config(struct pcmcia_device *link) strcpy(info->node.dev_name, dev->name); - printk(KERN_INFO "%s: port %#3lx, irq %d,", - dev->name, dev->base_addr, dev->irq); - printk (" mmio %#5lx,", (u_long)ti->mmio); - printk (" sram %#5lx,", (u_long)ti->sram_base << 12); - printk ("\n" KERN_INFO " hwaddr="); - for (i = 0; i < TR_ALEN; i++) - printk("%02X", dev->dev_addr[i]); - printk("\n"); + printk(KERN_INFO + "%s: port %#3lx, irq %d, mmio %#5lx, sram %#5lx, hwaddr=%pM\n", + dev->name, dev->base_addr, dev->irq, + (u_long)ti->mmio, (u_long)(ti->sram_base << 12), + dev->dev_addr); return 0; cs_failed: diff --git a/drivers/net/pcmcia/nmclan_cs.c b/drivers/net/pcmcia/nmclan_cs.c index 02ef63ed1f99..36de91baf238 100644 --- a/drivers/net/pcmcia/nmclan_cs.c +++ b/drivers/net/pcmcia/nmclan_cs.c @@ -1425,15 +1425,12 @@ static void BuildLAF(int *ladrf, int *adr) ladrf[byte] |= (1 << (hashcode & 7)); #ifdef PCMCIA_DEBUG - if (pc_debug > 2) { - printk(KERN_DEBUG " adr ="); - for (i = 0; i < 6; i++) - printk(" %02X", adr[i]); - printk("\n" KERN_DEBUG " hashcode = %d(decimal), ladrf[0:63]" - " =", hashcode); - for (i = 0; i < 8; i++) - printk(" %02X", ladrf[i]); - printk("\n"); + if (pc_debug > 2) + printk(KERN_DEBUG " adr =%pM\n", adr); + printk(KERN_DEBUG " hashcode = %d(decimal), ladrf[0:63] =", hashcode); + for (i = 0; i < 8; i++) + printk(KERN_CONT " %02X", ladrf[i]); + printk(KERN_CONT "\n"); } #endif } /* BuildLAF */ diff --git a/drivers/net/pcnet32.c b/drivers/net/pcnet32.c index 1c35e1d637a0..28368157dac4 100644 --- a/drivers/net/pcnet32.c +++ b/drivers/net/pcnet32.c @@ -485,7 +485,7 @@ static void pcnet32_realloc_tx_ring(struct net_device *dev, &new_ring_dma_addr); if (new_tx_ring == NULL) { if (netif_msg_drv(lp)) - printk("\n" KERN_ERR + printk(KERN_ERR "%s: Consistent memory allocation failed.\n", dev->name); return; @@ -496,7 +496,7 @@ static void pcnet32_realloc_tx_ring(struct net_device *dev, GFP_ATOMIC); if (!new_dma_addr_list) { if (netif_msg_drv(lp)) - printk("\n" KERN_ERR + printk(KERN_ERR "%s: Memory allocation failed.\n", dev->name); goto free_new_tx_ring; } @@ -505,7 +505,7 @@ static void pcnet32_realloc_tx_ring(struct net_device *dev, GFP_ATOMIC); if (!new_skb_list) { if (netif_msg_drv(lp)) - printk("\n" KERN_ERR + printk(KERN_ERR "%s: Memory allocation failed.\n", dev->name); goto free_new_lists; } @@ -563,7 +563,7 @@ static void pcnet32_realloc_rx_ring(struct net_device *dev, &new_ring_dma_addr); if (new_rx_ring == NULL) { if (netif_msg_drv(lp)) - printk("\n" KERN_ERR + printk(KERN_ERR "%s: Consistent memory allocation failed.\n", dev->name); return; @@ -574,7 +574,7 @@ static void pcnet32_realloc_rx_ring(struct net_device *dev, GFP_ATOMIC); if (!new_dma_addr_list) { if (netif_msg_drv(lp)) - printk("\n" KERN_ERR + printk(KERN_ERR "%s: Memory allocation failed.\n", dev->name); goto free_new_rx_ring; } @@ -583,7 +583,7 @@ static void pcnet32_realloc_rx_ring(struct net_device *dev, GFP_ATOMIC); if (!new_skb_list) { if (netif_msg_drv(lp)) - printk("\n" KERN_ERR + printk(KERN_ERR "%s: Memory allocation failed.\n", dev->name); goto free_new_lists; } @@ -1766,38 +1766,38 @@ pcnet32_probe1(unsigned long ioaddr, int shared, struct pci_dev *pdev) /* Version 0x2623 and 0x2624 */ if (((chip_version + 1) & 0xfffe) == 0x2624) { i = a->read_csr(ioaddr, 80) & 0x0C00; /* Check tx_start_pt */ - printk("\n" KERN_INFO " tx_start_pt(0x%04x):", i); + printk(KERN_INFO " tx_start_pt(0x%04x):", i); switch (i >> 10) { case 0: - printk(" 20 bytes,"); + printk(KERN_CONT " 20 bytes,"); break; case 1: - printk(" 64 bytes,"); + printk(KERN_CONT " 64 bytes,"); break; case 2: - printk(" 128 bytes,"); + printk(KERN_CONT " 128 bytes,"); break; case 3: - printk("~220 bytes,"); + printk(KERN_CONT "~220 bytes,"); break; } i = a->read_bcr(ioaddr, 18); /* Check Burst/Bus control */ - printk(" BCR18(%x):", i & 0xffff); + printk(KERN_CONT " BCR18(%x):", i & 0xffff); if (i & (1 << 5)) - printk("BurstWrEn "); + printk(KERN_CONT "BurstWrEn "); if (i & (1 << 6)) - printk("BurstRdEn "); + printk(KERN_CONT "BurstRdEn "); if (i & (1 << 7)) - printk("DWordIO "); + printk(KERN_CONT "DWordIO "); if (i & (1 << 11)) - printk("NoUFlow "); + printk(KERN_CONT "NoUFlow "); i = a->read_bcr(ioaddr, 25); - printk("\n" KERN_INFO " SRAMSIZE=0x%04x,", i << 8); + printk(KERN_INFO " SRAMSIZE=0x%04x,", i << 8); i = a->read_bcr(ioaddr, 26); - printk(" SRAM_BND=0x%04x,", i << 8); + printk(KERN_CONT " SRAM_BND=0x%04x,", i << 8); i = a->read_bcr(ioaddr, 27); if (i & (1 << 14)) - printk("LowLatRx"); + printk(KERN_CONT "LowLatRx"); } } @@ -1996,7 +1996,7 @@ static int pcnet32_alloc_ring(struct net_device *dev, const char *name) &lp->tx_ring_dma_addr); if (lp->tx_ring == NULL) { if (netif_msg_drv(lp)) - printk("\n" KERN_ERR PFX + printk(KERN_ERR PFX "%s: Consistent memory allocation failed.\n", name); return -ENOMEM; @@ -2008,7 +2008,7 @@ static int pcnet32_alloc_ring(struct net_device *dev, const char *name) &lp->rx_ring_dma_addr); if (lp->rx_ring == NULL) { if (netif_msg_drv(lp)) - printk("\n" KERN_ERR PFX + printk(KERN_ERR PFX "%s: Consistent memory allocation failed.\n", name); return -ENOMEM; @@ -2018,7 +2018,7 @@ static int pcnet32_alloc_ring(struct net_device *dev, const char *name) GFP_ATOMIC); if (!lp->tx_dma_addr) { if (netif_msg_drv(lp)) - printk("\n" KERN_ERR PFX + printk(KERN_ERR PFX "%s: Memory allocation failed.\n", name); return -ENOMEM; } @@ -2027,7 +2027,7 @@ static int pcnet32_alloc_ring(struct net_device *dev, const char *name) GFP_ATOMIC); if (!lp->rx_dma_addr) { if (netif_msg_drv(lp)) - printk("\n" KERN_ERR PFX + printk(KERN_ERR PFX "%s: Memory allocation failed.\n", name); return -ENOMEM; } @@ -2036,7 +2036,7 @@ static int pcnet32_alloc_ring(struct net_device *dev, const char *name) GFP_ATOMIC); if (!lp->tx_skbuff) { if (netif_msg_drv(lp)) - printk("\n" KERN_ERR PFX + printk(KERN_ERR PFX "%s: Memory allocation failed.\n", name); return -ENOMEM; } @@ -2045,7 +2045,7 @@ static int pcnet32_alloc_ring(struct net_device *dev, const char *name) GFP_ATOMIC); if (!lp->rx_skbuff) { if (netif_msg_drv(lp)) - printk("\n" KERN_ERR PFX + printk(KERN_ERR PFX "%s: Memory allocation failed.\n", name); return -ENOMEM; } diff --git a/drivers/net/starfire.c b/drivers/net/starfire.c index 838cce8b8fff..669253c7bd41 100644 --- a/drivers/net/starfire.c +++ b/drivers/net/starfire.c @@ -180,7 +180,7 @@ static int full_duplex[MAX_UNITS] = {0, }; /* These identify the driver base version and may not be removed. */ static const char version[] __devinitconst = KERN_INFO "starfire.c:v1.03 7/26/2000 Written by Donald Becker \n" -KERN_INFO " (unofficial 2.2/2.4 kernel port, version " DRV_VERSION ", " DRV_RELDATE ")\n"; +" (unofficial 2.2/2.4 kernel port, version " DRV_VERSION ", " DRV_RELDATE ")\n"; MODULE_AUTHOR("Donald Becker "); MODULE_DESCRIPTION("Adaptec Starfire Ethernet driver"); diff --git a/drivers/net/sundance.c b/drivers/net/sundance.c index 545f81b34ad7..d1521c3875b2 100644 --- a/drivers/net/sundance.c +++ b/drivers/net/sundance.c @@ -1698,13 +1698,13 @@ static int netdev_close(struct net_device *dev) #ifdef __i386__ if (netif_msg_hw(np)) { - printk("\n"KERN_DEBUG" Tx ring at %8.8x:\n", + printk(KERN_DEBUG " Tx ring at %8.8x:\n", (int)(np->tx_ring_dma)); for (i = 0; i < TX_RING_SIZE; i++) - printk(" #%d desc. %4.4x %8.8x %8.8x.\n", + printk(KERN_DEBUG " #%d desc. %4.4x %8.8x %8.8x.\n", i, np->tx_ring[i].status, np->tx_ring[i].frag[0].addr, np->tx_ring[i].frag[0].length); - printk("\n"KERN_DEBUG " Rx ring %8.8x:\n", + printk(KERN_DEBUG " Rx ring %8.8x:\n", (int)(np->rx_ring_dma)); for (i = 0; i < /*RX_RING_SIZE*/4 ; i++) { printk(KERN_DEBUG " #%d desc. %4.4x %4.4x %8.8x\n", diff --git a/drivers/net/tsi108_eth.c b/drivers/net/tsi108_eth.c index 0f78f99f9b20..7030bd5e9848 100644 --- a/drivers/net/tsi108_eth.c +++ b/drivers/net/tsi108_eth.c @@ -1132,7 +1132,9 @@ static int tsi108_get_mac(struct net_device *dev) } if (!is_valid_ether_addr(dev->dev_addr)) { - printk("KERN_ERR: word1: %08x, word2: %08x\n", word1, word2); + printk(KERN_ERR + "%s: Invalid MAC address. word1: %08x, word2: %08x\n", + dev->name, word1, word2); return -EINVAL; } @@ -1201,8 +1203,8 @@ static void tsi108_set_rx_mode(struct net_device *dev) __set_bit(hash, &data->mc_hash[0]); } else { printk(KERN_ERR - "%s: got multicast address of length %d " - "instead of 6.\n", dev->name, + "%s: got multicast address of length %d instead of 6.\n", + dev->name, mc->dmi_addrlen); } diff --git a/drivers/net/tulip/de2104x.c b/drivers/net/tulip/de2104x.c index 81f054dbb88d..ef49744a5085 100644 --- a/drivers/net/tulip/de2104x.c +++ b/drivers/net/tulip/de2104x.c @@ -944,9 +944,10 @@ static void de_set_media (struct de_private *de) macmode &= ~FullDuplex; if (netif_msg_link(de)) { - printk(KERN_INFO "%s: set link %s\n" - KERN_INFO "%s: mode 0x%x, sia 0x%x,0x%x,0x%x,0x%x\n" - KERN_INFO "%s: set mode 0x%x, set sia 0x%x,0x%x,0x%x\n", + printk(KERN_INFO + "%s: set link %s\n" + "%s: mode 0x%x, sia 0x%x,0x%x,0x%x,0x%x\n" + "%s: set mode 0x%x, set sia 0x%x,0x%x,0x%x\n", de->dev->name, media_name[media], de->dev->name, dr32(MacMode), dr32(SIAStatus), dr32(CSR13), dr32(CSR14), dr32(CSR15), diff --git a/drivers/net/tulip/tulip_core.c b/drivers/net/tulip/tulip_core.c index 2abb5d3becc6..99a63649f4fc 100644 --- a/drivers/net/tulip/tulip_core.c +++ b/drivers/net/tulip/tulip_core.c @@ -570,16 +570,18 @@ static void tulip_tx_timeout(struct net_device *dev) (unsigned int)tp->rx_ring[i].buffer2, buf[0], buf[1], buf[2]); for (j = 0; buf[j] != 0xee && j < 1600; j++) - if (j < 100) printk(" %2.2x", buf[j]); - printk(" j=%d.\n", j); + if (j < 100) + printk(KERN_CONT " %2.2x", buf[j]); + printk(KERN_CONT " j=%d.\n", j); } printk(KERN_DEBUG " Rx ring %8.8x: ", (int)tp->rx_ring); for (i = 0; i < RX_RING_SIZE; i++) - printk(" %8.8x", (unsigned int)tp->rx_ring[i].status); - printk("\n" KERN_DEBUG " Tx ring %8.8x: ", (int)tp->tx_ring); + printk(KERN_CONT " %8.8x", + (unsigned int)tp->rx_ring[i].status); + printk(KERN_DEBUG " Tx ring %8.8x: ", (int)tp->tx_ring); for (i = 0; i < TX_RING_SIZE; i++) - printk(" %8.8x", (unsigned int)tp->tx_ring[i].status); - printk("\n"); + printk(KERN_CONT " %8.8x", (unsigned int)tp->tx_ring[i].status); + printk(KERN_CONT "\n"); } #endif diff --git a/drivers/net/tulip/winbond-840.c b/drivers/net/tulip/winbond-840.c index 842b1a2c40d4..0f15773dae52 100644 --- a/drivers/net/tulip/winbond-840.c +++ b/drivers/net/tulip/winbond-840.c @@ -142,7 +142,7 @@ static int full_duplex[MAX_UNITS] = {-1, -1, -1, -1, -1, -1, -1, -1}; static const char version[] __initconst = KERN_INFO DRV_NAME ".c:v" DRV_VERSION " (2.4 port) " DRV_RELDATE " Donald Becker \n" - KERN_INFO " http://www.scyld.com/network/drivers.html\n"; + " http://www.scyld.com/network/drivers.html\n"; MODULE_AUTHOR("Donald Becker "); MODULE_DESCRIPTION("Winbond W89c840 Ethernet driver"); @@ -939,7 +939,7 @@ static void tx_timeout(struct net_device *dev) printk(KERN_DEBUG " Rx ring %p: ", np->rx_ring); for (i = 0; i < RX_RING_SIZE; i++) printk(" %8.8x", (unsigned int)np->rx_ring[i].status); - printk("\n"KERN_DEBUG" Tx ring %p: ", np->tx_ring); + printk(KERN_DEBUG" Tx ring %p: ", np->tx_ring); for (i = 0; i < TX_RING_SIZE; i++) printk(" %8.8x", np->tx_ring[i].status); printk("\n"); @@ -1520,7 +1520,7 @@ static int netdev_close(struct net_device *dev) printk(KERN_DEBUG " #%d desc. %4.4x %4.4x %8.8x.\n", i, np->tx_ring[i].length, np->tx_ring[i].status, np->tx_ring[i].buffer1); - printk("\n"KERN_DEBUG " Rx ring %8.8x:\n", + printk(KERN_DEBUG " Rx ring %8.8x:\n", (int)np->rx_ring); for (i = 0; i < RX_RING_SIZE; i++) { printk(KERN_DEBUG " #%d desc. %4.4x %4.4x %8.8x\n", diff --git a/drivers/net/wan/hd64570.c b/drivers/net/wan/hd64570.c index 223238de475c..1ea1ef6c3b96 100644 --- a/drivers/net/wan/hd64570.c +++ b/drivers/net/wan/hd64570.c @@ -584,8 +584,9 @@ static void sca_dump_rings(struct net_device *dev) sca_in(DSR_RX(phy_node(port)), card) & DSR_DE ? "" : "in"); for (cnt = 0; cnt < port_to_card(port)->rx_ring_buffers; cnt++) printk(" %02X", readb(&(desc_address(port, cnt, 0)->stat))); + printk(KERN_CONT "\n"); - printk("\n" KERN_DEBUG "TX ring: CDA=%u EDA=%u DSR=%02X in=%u " + printk(KERN_DEBUG "TX ring: CDA=%u EDA=%u DSR=%02X in=%u " "last=%u %sactive", sca_inw(get_dmac_tx(port) + CDAL, card), sca_inw(get_dmac_tx(port) + EDAL, card), diff --git a/drivers/net/wan/hd64572.c b/drivers/net/wan/hd64572.c index 497b003d7239..f099c34a3ae2 100644 --- a/drivers/net/wan/hd64572.c +++ b/drivers/net/wan/hd64572.c @@ -529,8 +529,9 @@ static void sca_dump_rings(struct net_device *dev) sca_in(DSR_RX(port->chan), card) & DSR_DE ? "" : "in"); for (cnt = 0; cnt < port->card->rx_ring_buffers; cnt++) printk(" %02X", readb(&(desc_address(port, cnt, 0)->stat))); + printk(KERN_CONT "\n"); - printk("\n" KERN_DEBUG "TX ring: CDA=%u EDA=%u DSR=%02X in=%u " + printk(KERN_DEBUG "TX ring: CDA=%u EDA=%u DSR=%02X in=%u " "last=%u %sactive", sca_inl(get_dmac_tx(port) + CDAL, card), sca_inl(get_dmac_tx(port) + EDAL, card), diff --git a/drivers/net/wan/sbni.c b/drivers/net/wan/sbni.c index 3fb9dbc88a1a..d14e95a08d66 100644 --- a/drivers/net/wan/sbni.c +++ b/drivers/net/wan/sbni.c @@ -326,11 +326,9 @@ sbni_pci_probe( struct net_device *dev ) } if (pci_irq_line <= 0 || pci_irq_line >= nr_irqs) - printk( KERN_WARNING " WARNING: The PCI BIOS assigned " - "this PCI card to IRQ %d, which is unlikely " - "to work!.\n" - KERN_WARNING " You should use the PCI BIOS " - "setup to assign a valid IRQ line.\n", + printk( KERN_WARNING + " WARNING: The PCI BIOS assigned this PCI card to IRQ %d, which is unlikely to work!.\n" + " You should use the PCI BIOS setup to assign a valid IRQ line.\n", pci_irq_line ); /* avoiding re-enable dual adapters */ diff --git a/drivers/net/wireless/ray_cs.c b/drivers/net/wireless/ray_cs.c index b10b0383dfa5..698b11b1cadb 100644 --- a/drivers/net/wireless/ray_cs.c +++ b/drivers/net/wireless/ray_cs.c @@ -2427,11 +2427,10 @@ static void untranslate(ray_dev_t *local, struct sk_buff *skb, int len) #ifdef PCMCIA_DEBUG if (pc_debug > 3) { - int i; - printk(KERN_DEBUG "skb->data before untranslate"); - for (i = 0; i < 64; i++) - printk("%02x ", skb->data[i]); - printk("\n" KERN_DEBUG + print_hex_dump(KERN_DEBUG, "skb->data before untranslate: ", + DUMP_PREFIX_NONE, 16, 1, + skb->data, 64, true); + printk(KERN_DEBUG "type = %08x, xsap = %02x%02x%02x, org = %02x02x02x\n", ntohs(type), psnap->dsap, psnap->ssap, psnap->ctrl, psnap->org[0], psnap->org[1], psnap->org[2]); diff --git a/drivers/net/wireless/wavelan_cs.c b/drivers/net/wireless/wavelan_cs.c index 6af706408ac0..c6d300666ad8 100644 --- a/drivers/net/wireless/wavelan_cs.c +++ b/drivers/net/wireless/wavelan_cs.c @@ -3556,17 +3556,8 @@ wv_82593_config(struct net_device * dev) cfblk.rcvstop = TRUE; /* Enable Receive Stop Register */ #ifdef DEBUG_I82593_SHOW - { - u_char *c = (u_char *) &cfblk; - int i; - printk(KERN_DEBUG "wavelan_cs: config block:"); - for(i = 0; i < sizeof(struct i82593_conf_block); i++,c++) - { - if((i % 16) == 0) printk("\n" KERN_DEBUG); - printk("%02x ", *c); - } - printk("\n"); - } + print_hex_dump(KERN_DEBUG, "wavelan_cs: config block: ", DUMP_PREFIX_NONE, + 16, 1, &cfblk, sizeof(struct i82593_conf_block), false); #endif /* Copy the config block to the i82593 */ diff --git a/drivers/net/yellowfin.c b/drivers/net/yellowfin.c index 3c7a5053f1da..a07580138e81 100644 --- a/drivers/net/yellowfin.c +++ b/drivers/net/yellowfin.c @@ -109,7 +109,7 @@ static int gx_fix; /* These identify the driver base version and may not be removed. */ static const char version[] __devinitconst = KERN_INFO DRV_NAME ".c:v1.05 1/09/2001 Written by Donald Becker \n" - KERN_INFO " (unofficial 2.4.x port, " DRV_VERSION ", " DRV_RELDATE ")\n"; + " (unofficial 2.4.x port, " DRV_VERSION ", " DRV_RELDATE ")\n"; MODULE_AUTHOR("Donald Becker "); MODULE_DESCRIPTION("Packet Engines Yellowfin G-NIC Gigabit Ethernet driver"); @@ -700,12 +700,15 @@ static void yellowfin_tx_timeout(struct net_device *dev) int i; printk(KERN_WARNING " Rx ring %p: ", yp->rx_ring); for (i = 0; i < RX_RING_SIZE; i++) - printk(" %8.8x", yp->rx_ring[i].result_status); - printk("\n"KERN_WARNING" Tx ring %p: ", yp->tx_ring); + printk(KERN_CONT " %8.8x", + yp->rx_ring[i].result_status); + printk(KERN_CONT "\n"); + printk(KERN_WARNING" Tx ring %p: ", yp->tx_ring); for (i = 0; i < TX_RING_SIZE; i++) - printk(" %4.4x /%8.8x", yp->tx_status[i].tx_errs, - yp->tx_ring[i].result_status); - printk("\n"); + printk(KERN_CONT " %4.4x /%8.8x", + yp->tx_status[i].tx_errs, + yp->tx_ring[i].result_status); + printk(KERN_CONT "\n"); } /* If the hardware is found to hang regularly, we will update the code @@ -1216,20 +1219,20 @@ static int yellowfin_close(struct net_device *dev) #if defined(__i386__) if (yellowfin_debug > 2) { - printk("\n"KERN_DEBUG" Tx ring at %8.8llx:\n", + printk(KERN_DEBUG" Tx ring at %8.8llx:\n", (unsigned long long)yp->tx_ring_dma); for (i = 0; i < TX_RING_SIZE*2; i++) - printk(" %c #%d desc. %8.8x %8.8x %8.8x %8.8x.\n", + printk(KERN_DEBUG " %c #%d desc. %8.8x %8.8x %8.8x %8.8x.\n", ioread32(ioaddr + TxPtr) == (long)&yp->tx_ring[i] ? '>' : ' ', i, yp->tx_ring[i].dbdma_cmd, yp->tx_ring[i].addr, yp->tx_ring[i].branch_addr, yp->tx_ring[i].result_status); printk(KERN_DEBUG " Tx status %p:\n", yp->tx_status); for (i = 0; i < TX_RING_SIZE; i++) - printk(" #%d status %4.4x %4.4x %4.4x %4.4x.\n", + printk(KERN_DEBUG " #%d status %4.4x %4.4x %4.4x %4.4x.\n", i, yp->tx_status[i].tx_cnt, yp->tx_status[i].tx_errs, yp->tx_status[i].total_tx_cnt, yp->tx_status[i].paused); - printk("\n"KERN_DEBUG " Rx ring %8.8llx:\n", + printk(KERN_DEBUG " Rx ring %8.8llx:\n", (unsigned long long)yp->rx_ring_dma); for (i = 0; i < RX_RING_SIZE; i++) { printk(KERN_DEBUG " %c #%d desc. %8.8x %8.8x %8.8x\n", diff --git a/drivers/parisc/eisa_enumerator.c b/drivers/parisc/eisa_enumerator.c index c709ecc2b7f7..0be1d50645ab 100644 --- a/drivers/parisc/eisa_enumerator.c +++ b/drivers/parisc/eisa_enumerator.c @@ -101,7 +101,7 @@ static int configure_memory(const unsigned char *buf, printk("memory %lx-%lx ", (unsigned long)res->start, (unsigned long)res->end); result = request_resource(mem_parent, res); if (result < 0) { - printk("\n" KERN_ERR "EISA Enumerator: failed to claim EISA Bus address space!\n"); + printk(KERN_ERR "EISA Enumerator: failed to claim EISA Bus address space!\n"); return result; } } @@ -191,7 +191,7 @@ static int configure_port(const unsigned char *buf, struct resource *io_parent, printk("ioports %lx-%lx ", (unsigned long)res->start, (unsigned long)res->end); result = request_resource(io_parent, res); if (result < 0) { - printk("\n" KERN_ERR "EISA Enumerator: failed to claim EISA Bus address space!\n"); + printk(KERN_ERR "EISA Enumerator: failed to claim EISA Bus address space!\n"); return result; } } @@ -224,7 +224,7 @@ static int configure_port_init(const unsigned char *buf) case HPEE_PORT_INIT_WIDTH_BYTE: s=1; if (c & HPEE_PORT_INIT_MASK) { - printk("\n" KERN_WARNING "port_init: unverified mask attribute\n"); + printk(KERN_WARNING "port_init: unverified mask attribute\n"); outb((inb(get_16(buf+len+1) & get_8(buf+len+3)) | get_8(buf+len+4)), get_16(buf+len+1)); @@ -249,7 +249,7 @@ static int configure_port_init(const unsigned char *buf) case HPEE_PORT_INIT_WIDTH_DWORD: s=4; if (c & HPEE_PORT_INIT_MASK) { - printk("\n" KERN_WARNING "port_init: unverified mask attribute\n"); + printk(KERN_WARNING "port_init: unverified mask attribute\n"); outl((inl(get_16(buf+len+1) & get_32(buf+len+3)) | get_32(buf+len+7)), get_16(buf+len+1)); @@ -259,7 +259,7 @@ static int configure_port_init(const unsigned char *buf) break; default: - printk("\n" KERN_ERR "Invalid port init word %02x\n", c); + printk(KERN_ERR "Invalid port init word %02x\n", c); return 0; } @@ -297,7 +297,7 @@ static int configure_type_string(const unsigned char *buf) /* just skip past the type field */ len = get_8(buf); if (len > 80) { - printk("\n" KERN_ERR "eisa_enumerator: type info field too long (%d, max is 80)\n", len); + printk(KERN_ERR "eisa_enumerator: type info field too long (%d, max is 80)\n", len); } return 1+len; @@ -398,7 +398,7 @@ static int parse_slot_config(int slot, } if (p0 + function_len < pos) { - printk("\n" KERN_ERR "eisa_enumerator: function %d length mis-match " + printk(KERN_ERR "eisa_enumerator: function %d length mis-match " "got %d, expected %d\n", num_func, pos-p0, function_len); res=-1; diff --git a/drivers/pcmcia/tcic.c b/drivers/pcmcia/tcic.c index 9ad97ea836e8..8eb04230fec7 100644 --- a/drivers/pcmcia/tcic.c +++ b/drivers/pcmcia/tcic.c @@ -472,7 +472,8 @@ static int __init init_tcic(void) init_timer(&poll_timer); /* Build interrupt mask */ - printk(", %d sockets\n" KERN_INFO " irq list (", sockets); + printk(KERN_CONT ", %d sockets\n", sockets); + printk(KERN_INFO " irq list ("); if (irq_list_count == 0) mask = irq_mask; else diff --git a/drivers/scsi/atari_NCR5380.c b/drivers/scsi/atari_NCR5380.c index 0471f8800483..4240b05aef6d 100644 --- a/drivers/scsi/atari_NCR5380.c +++ b/drivers/scsi/atari_NCR5380.c @@ -2826,8 +2826,7 @@ int NCR5380_abort(Scsi_Cmnd *cmd) */ local_irq_restore(flags); - printk(KERN_INFO "scsi%d: warning : SCSI command probably completed successfully\n" - KERN_INFO " before abortion\n", HOSTNO); + printk(KERN_INFO "scsi%d: warning : SCSI command probably completed successfully before abortion\n", HOSTNO); /* Maybe it is sufficient just to release the ST-DMA lock... (if * possible at all) At least, we should check if the lock could be diff --git a/drivers/scsi/mac53c94.c b/drivers/scsi/mac53c94.c index b12ad7c7c673..18735b39b3d3 100644 --- a/drivers/scsi/mac53c94.c +++ b/drivers/scsi/mac53c94.c @@ -75,8 +75,9 @@ static int mac53c94_queue(struct scsi_cmnd *cmd, void (*done)(struct scsi_cmnd * int i; printk(KERN_DEBUG "mac53c94_queue %p: command is", cmd); for (i = 0; i < cmd->cmd_len; ++i) - printk(" %.2x", cmd->cmnd[i]); - printk("\n" KERN_DEBUG "use_sg=%d request_bufflen=%d request_buffer=%p\n", + printk(KERN_CONT " %.2x", cmd->cmnd[i]); + printk(KERN_CONT "\n"); + printk(KERN_DEBUG "use_sg=%d request_bufflen=%d request_buffer=%p\n", scsi_sg_count(cmd), scsi_bufflen(cmd), scsi_sglist(cmd)); } #endif diff --git a/drivers/scsi/sg.c b/drivers/scsi/sg.c index ef142fd47a83..4d6f2fe1cfe9 100644 --- a/drivers/scsi/sg.c +++ b/drivers/scsi/sg.c @@ -619,7 +619,7 @@ sg_write(struct file *filp, const char __user *buf, size_t count, loff_t * ppos) if (strcmp(current->comm, cmd) && printk_ratelimit()) { printk(KERN_WARNING "sg_write: data in/out %d/%d bytes for SCSI command 0x%x--" - "guessing data in;\n" KERN_WARNING " " + "guessing data in;\n " "program %s not setting count and/or reply_len properly\n", old_hdr.reply_len - (int)SZ_SG_HEADER, input_size, (unsigned int) cmnd[0], diff --git a/drivers/scsi/sun3_NCR5380.c b/drivers/scsi/sun3_NCR5380.c index bcaba86060ab..75da6e58ce55 100644 --- a/drivers/scsi/sun3_NCR5380.c +++ b/drivers/scsi/sun3_NCR5380.c @@ -2860,8 +2860,7 @@ static int NCR5380_abort(struct scsi_cmnd *cmd) */ local_irq_restore(flags); - printk(KERN_INFO "scsi%d: warning : SCSI command probably completed successfully\n" - KERN_INFO " before abortion\n", HOSTNO); + printk(KERN_INFO "scsi%d: warning : SCSI command probably completed successfully before abortion\n", HOSTNO); return SCSI_ABORT_NOT_RUNNING; } diff --git a/drivers/serial/8250_pci.c b/drivers/serial/8250_pci.c index 6160e03f410c..e7108e75653d 100644 --- a/drivers/serial/8250_pci.c +++ b/drivers/serial/8250_pci.c @@ -60,11 +60,12 @@ struct serial_private { static void moan_device(const char *str, struct pci_dev *dev) { - printk(KERN_WARNING "%s: %s\n" - KERN_WARNING "Please send the output of lspci -vv, this\n" - KERN_WARNING "message (0x%04x,0x%04x,0x%04x,0x%04x), the\n" - KERN_WARNING "manufacturer and name of serial board or\n" - KERN_WARNING "modem board to rmk+serial@arm.linux.org.uk.\n", + printk(KERN_WARNING + "%s: %s\n" + "Please send the output of lspci -vv, this\n" + "message (0x%04x,0x%04x,0x%04x,0x%04x), the\n" + "manufacturer and name of serial board or\n" + "modem board to rmk+serial@arm.linux.org.uk.\n", pci_name(dev), str, dev->vendor, dev->device, dev->subsystem_vendor, dev->subsystem_device); } diff --git a/drivers/ssb/pcmcia.c b/drivers/ssb/pcmcia.c index fbfadbac67e8..131030f693c7 100644 --- a/drivers/ssb/pcmcia.c +++ b/drivers/ssb/pcmcia.c @@ -583,7 +583,7 @@ static int ssb_pcmcia_sprom_write_all(struct ssb_bus *bus, const u16 *sprom) ssb_printk("."); err = ssb_pcmcia_sprom_write(bus, i, sprom[i]); if (err) { - ssb_printk("\n" KERN_NOTICE PFX + ssb_printk(KERN_NOTICE PFX "Failed to write to SPROM.\n"); failed = 1; break; @@ -591,7 +591,7 @@ static int ssb_pcmcia_sprom_write_all(struct ssb_bus *bus, const u16 *sprom) } err = ssb_pcmcia_sprom_command(bus, SSB_PCMCIA_SPROMCTL_WRITEDIS); if (err) { - ssb_printk("\n" KERN_NOTICE PFX + ssb_printk(KERN_NOTICE PFX "Could not disable SPROM write access.\n"); failed = 1; } diff --git a/drivers/usb/core/hcd.c b/drivers/usb/core/hcd.c index ce3f453f02ef..95ccfa0b9fc5 100644 --- a/drivers/usb/core/hcd.c +++ b/drivers/usb/core/hcd.c @@ -648,7 +648,7 @@ void usb_hcd_poll_rh_status(struct usb_hcd *hcd) struct urb *urb; int length; unsigned long flags; - char buffer[4]; /* Any root hubs with > 31 ports? */ + char buffer[6]; /* Any root hubs with > 31 ports? */ if (unlikely(!hcd->rh_registered)) return; diff --git a/drivers/video/amba-clcd.c b/drivers/video/amba-clcd.c index fb8163d181ab..a21efcd10b78 100644 --- a/drivers/video/amba-clcd.c +++ b/drivers/video/amba-clcd.c @@ -226,9 +226,10 @@ static int clcdfb_set_par(struct fb_info *info) clcdfb_enable(fb, regs.cntl); #ifdef DEBUG - printk(KERN_INFO "CLCD: Registers set to\n" - KERN_INFO " %08x %08x %08x %08x\n" - KERN_INFO " %08x %08x %08x %08x\n", + printk(KERN_INFO + "CLCD: Registers set to\n" + " %08x %08x %08x %08x\n" + " %08x %08x %08x %08x\n", readl(fb->regs + CLCD_TIM0), readl(fb->regs + CLCD_TIM1), readl(fb->regs + CLCD_TIM2), readl(fb->regs + CLCD_TIM3), readl(fb->regs + CLCD_UBAS), readl(fb->regs + CLCD_LBAS), diff --git a/drivers/video/matrox/matroxfb_DAC1064.c b/drivers/video/matrox/matroxfb_DAC1064.c index 0ce3b0a89798..a74e5da17aa0 100644 --- a/drivers/video/matrox/matroxfb_DAC1064.c +++ b/drivers/video/matrox/matroxfb_DAC1064.c @@ -454,9 +454,9 @@ static void DAC1064_restore_2(WPMINFO2) { dprintk(KERN_DEBUG "DAC1064regs "); for (i = 0; i < sizeof(MGA1064_DAC_regs); i++) { dprintk("R%02X=%02X ", MGA1064_DAC_regs[i], ACCESS_FBINFO(hw).DACreg[i]); - if ((i & 0x7) == 0x7) dprintk("\n" KERN_DEBUG "continuing... "); + if ((i & 0x7) == 0x7) dprintk(KERN_DEBUG "continuing... "); } - dprintk("\n" KERN_DEBUG "DAC1064clk "); + dprintk(KERN_DEBUG "DAC1064clk "); for (i = 0; i < 6; i++) dprintk("C%02X=%02X ", i, ACCESS_FBINFO(hw).DACclk[i]); dprintk("\n"); diff --git a/drivers/video/matrox/matroxfb_Ti3026.c b/drivers/video/matrox/matroxfb_Ti3026.c index 13524821e242..4e825112a601 100644 --- a/drivers/video/matrox/matroxfb_Ti3026.c +++ b/drivers/video/matrox/matroxfb_Ti3026.c @@ -651,9 +651,9 @@ static void Ti3026_restore(WPMINFO2) { dprintk(KERN_DEBUG "3026DACregs "); for (i = 0; i < 21; i++) { dprintk("R%02X=%02X ", DACseq[i], hw->DACreg[i]); - if ((i & 0x7) == 0x7) dprintk("\n" KERN_DEBUG "continuing... "); + if ((i & 0x7) == 0x7) dprintk(KERN_DEBUG "continuing... "); } - dprintk("\n" KERN_DEBUG "DACclk "); + dprintk(KERN_DEBUG "DACclk "); for (i = 0; i < 6; i++) dprintk("C%02X=%02X ", i, hw->DACclk[i]); dprintk("\n"); diff --git a/drivers/video/stifb.c b/drivers/video/stifb.c index eec9dcb7f599..6120f0c526fe 100644 --- a/drivers/video/stifb.c +++ b/drivers/video/stifb.c @@ -1115,10 +1115,9 @@ static int __init stifb_init_fb(struct sti_struct *sti, int bpp_pref) if the device name contains the string "DX" and tell the user how to reconfigure the card. */ if (strstr(sti->outptr.dev_name, "DX")) { - printk(KERN_WARNING "WARNING: stifb framebuffer driver does not " - "support '%s' in double-buffer mode.\n" - KERN_WARNING "WARNING: Please disable the double-buffer mode " - "in IPL menu (the PARISC-BIOS).\n", + printk(KERN_WARNING +"WARNING: stifb framebuffer driver does not support '%s' in double-buffer mode.\n" +"WARNING: Please disable the double-buffer mode in IPL menu (the PARISC-BIOS).\n", sti->outptr.dev_name); goto out_err0; } diff --git a/fs/jffs2/erase.c b/fs/jffs2/erase.c index a0244740b75a..b47679be118a 100644 --- a/fs/jffs2/erase.c +++ b/fs/jffs2/erase.c @@ -270,19 +270,21 @@ static inline void jffs2_remove_node_refs_from_ino_list(struct jffs2_sb_info *c, D2({ int i=0; struct jffs2_raw_node_ref *this; - printk(KERN_DEBUG "After remove_node_refs_from_ino_list: \n" KERN_DEBUG); + printk(KERN_DEBUG "After remove_node_refs_from_ino_list: \n"); this = ic->nodes; + printk(KERN_DEBUG); while(this) { - printk( "0x%08x(%d)->", ref_offset(this), ref_flags(this)); + printk(KERN_CONT "0x%08x(%d)->", + ref_offset(this), ref_flags(this)); if (++i == 5) { - printk("\n" KERN_DEBUG); + printk(KERN_DEBUG); i=0; } this = this->next_in_ino; } - printk("\n"); + printk(KERN_CONT "\n"); }); switch (ic->class) { diff --git a/kernel/module.c b/kernel/module.c index 38928fcaff2b..0a049837008e 100644 --- a/kernel/module.c +++ b/kernel/module.c @@ -2451,9 +2451,9 @@ SYSCALL_DEFINE3(init_module, void __user *, umod, return ret; } if (ret > 0) { - printk(KERN_WARNING "%s: '%s'->init suspiciously returned %d, " - "it should follow 0/-E convention\n" - KERN_WARNING "%s: loading module anyway...\n", + printk(KERN_WARNING +"%s: '%s'->init suspiciously returned %d, it should follow 0/-E convention\n" +"%s: loading module anyway...\n", __func__, mod->name, ret, __func__); dump_stack(); diff --git a/sound/pci/emu10k1/p16v.c b/sound/pci/emu10k1/p16v.c index e617acaf10e3..61b8ab39800f 100644 --- a/sound/pci/emu10k1/p16v.c +++ b/sound/pci/emu10k1/p16v.c @@ -644,7 +644,7 @@ int __devinit snd_p16v_pcm(struct snd_emu10k1 *emu, int device, struct snd_pcm * int err; int capture=1; - /* snd_printk("KERN_DEBUG snd_p16v_pcm called. device=%d\n", device); */ + /* snd_printk(KERN_DEBUG "snd_p16v_pcm called. device=%d\n", device); */ emu->p16v_device_offset = device; if (rpcm) *rpcm = NULL; diff --git a/sound/usb/usx2y/usbusx2yaudio.c b/sound/usb/usx2y/usbusx2yaudio.c index dd1ab6177840..9efd27f6b52f 100644 --- a/sound/usb/usx2y/usbusx2yaudio.c +++ b/sound/usb/usx2y/usbusx2yaudio.c @@ -296,9 +296,10 @@ static void usX2Y_error_urb_status(struct usX2Ydev *usX2Y, static void usX2Y_error_sequence(struct usX2Ydev *usX2Y, struct snd_usX2Y_substream *subs, struct urb *urb) { - snd_printk(KERN_ERR "Sequence Error!(hcd_frame=%i ep=%i%s;wait=%i,frame=%i).\n" - KERN_ERR "Most propably some urb of usb-frame %i is still missing.\n" - KERN_ERR "Cause could be too long delays in usb-hcd interrupt handling.\n", + snd_printk(KERN_ERR +"Sequence Error!(hcd_frame=%i ep=%i%s;wait=%i,frame=%i).\n" +"Most propably some urb of usb-frame %i is still missing.\n" +"Cause could be too long delays in usb-hcd interrupt handling.\n", usb_get_current_frame_number(usX2Y->chip.dev), subs->endpoint, usb_pipein(urb->pipe) ? "in" : "out", usX2Y->wait_iso_frame, urb->start_frame, usX2Y->wait_iso_frame); -- cgit v1.2.3 From 5ddf1e0ff00fd808c048d0b920784828276cc516 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Sun, 5 Jul 2009 11:01:02 -0400 Subject: cifs: fix regression with O_EXCL creates and optimize away lookup cifs: fix regression with O_EXCL creates and optimize away lookup Signed-off-by: Jeff Layton Tested-by: Shirish Pargaonkar CC: Stable Kernel Signed-off-by: Steve French --- fs/cifs/dir.c | 9 +++++++++ 1 file changed, 9 insertions(+) (limited to 'fs') diff --git a/fs/cifs/dir.c b/fs/cifs/dir.c index a40054faed7f..ff55fc6932cb 100644 --- a/fs/cifs/dir.c +++ b/fs/cifs/dir.c @@ -643,6 +643,15 @@ cifs_lookup(struct inode *parent_dir_inode, struct dentry *direntry, } } + /* + * O_EXCL: optimize away the lookup, but don't hash the dentry. Let + * the VFS handle the create. + */ + if (nd->flags & LOOKUP_EXCL) { + d_instantiate(direntry, NULL); + return 0; + } + /* can not grab the rename sem here since it would deadlock in the cases (beginning of sys_rename itself) in which we already have the sb rename sem */ -- cgit v1.2.3 From 8b712cd58adfe6aeeb0be4ecc011dc35620719e7 Mon Sep 17 00:00:00 2001 From: Jeff Mahoney Date: Tue, 7 Jul 2009 17:22:12 -0400 Subject: ocfs2: Fixup orphan scan cleanup after failed mount If the mount fails for any reason, ocfs2_dismount_volume calls ocfs2_orphan_scan_stop. It requires that ocfs2_orphan_scan_init be called to setup the mutex and work queues, but that doesn't happen if the mount has failed and we oops accessing an uninitialized work queue. This patch splits the init and startup of the orphan scan, eliminating the oops. Signed-off-by: Jeff Mahoney Signed-off-by: Joel Becker --- fs/ocfs2/journal.c | 8 +++++++- fs/ocfs2/journal.h | 1 + fs/ocfs2/super.c | 4 +++- 3 files changed, 11 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/ocfs2/journal.c b/fs/ocfs2/journal.c index f033760ecbea..c48b93ac6b65 100644 --- a/fs/ocfs2/journal.c +++ b/fs/ocfs2/journal.c @@ -1954,10 +1954,16 @@ void ocfs2_orphan_scan_init(struct ocfs2_super *osb) os->os_osb = osb; os->os_count = 0; os->os_seqno = 0; - os->os_scantime = CURRENT_TIME; mutex_init(&os->os_lock); INIT_DELAYED_WORK(&os->os_orphan_scan_work, ocfs2_orphan_scan_work); +} +void ocfs2_orphan_scan_start(struct ocfs2_super *osb) +{ + struct ocfs2_orphan_scan *os; + + os = &osb->osb_orphan_scan; + os->os_scantime = CURRENT_TIME; if (ocfs2_is_hard_readonly(osb) || ocfs2_mount_local(osb)) atomic_set(&os->os_state, ORPHAN_SCAN_INACTIVE); else { diff --git a/fs/ocfs2/journal.h b/fs/ocfs2/journal.h index 5432c7f79cc6..81e8abcd246e 100644 --- a/fs/ocfs2/journal.h +++ b/fs/ocfs2/journal.h @@ -145,6 +145,7 @@ static inline void ocfs2_inode_set_new(struct ocfs2_super *osb, /* Exported only for the journal struct init code in super.c. Do not call. */ void ocfs2_orphan_scan_init(struct ocfs2_super *osb); +void ocfs2_orphan_scan_start(struct ocfs2_super *osb); void ocfs2_orphan_scan_stop(struct ocfs2_super *osb); void ocfs2_orphan_scan_exit(struct ocfs2_super *osb); diff --git a/fs/ocfs2/super.c b/fs/ocfs2/super.c index 7efb349fb9bd..63af2e36d834 100644 --- a/fs/ocfs2/super.c +++ b/fs/ocfs2/super.c @@ -1182,7 +1182,7 @@ static int ocfs2_fill_super(struct super_block *sb, void *data, int silent) wake_up(&osb->osb_mount_event); /* Start this when the mount is almost sure of being successful */ - ocfs2_orphan_scan_init(osb); + ocfs2_orphan_scan_start(osb); mlog_exit(status); return status; @@ -1981,6 +1981,8 @@ static int ocfs2_initialize_super(struct super_block *sb, snprintf(osb->dev_str, sizeof(osb->dev_str), "%u,%u", MAJOR(osb->sb->s_dev), MINOR(osb->sb->s_dev)); + ocfs2_orphan_scan_init(osb); + status = ocfs2_recovery_init(osb); if (status) { mlog(ML_ERROR, "Unable to initialize recovery state\n"); -- cgit v1.2.3 From 17ae26b669886efe237b77439e43eb390fceb119 Mon Sep 17 00:00:00 2001 From: Jeff Liu Date: Tue, 7 Jul 2009 15:51:40 +0800 Subject: ocfs2: trivial fix for s/migrate/migration/ in dlmrecovery.c logging in dlmrecovery.c:1121, replace 'migrate' to 'migration' to keep the consistency by comparing to other lines with the similar log info in the same file. Signed-off-by: Jeff Liu Signed-off-by: Joel Becker --- fs/ocfs2/dlm/dlmrecovery.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ocfs2/dlm/dlmrecovery.c b/fs/ocfs2/dlm/dlmrecovery.c index bcb9260c3735..43e6e3280569 100644 --- a/fs/ocfs2/dlm/dlmrecovery.c +++ b/fs/ocfs2/dlm/dlmrecovery.c @@ -1118,7 +1118,7 @@ static int dlm_send_mig_lockres_msg(struct dlm_ctxt *dlm, mlog(0, "%s:%.*s: sending mig lockres (%s) to %u\n", dlm->name, res->lockname.len, res->lockname.name, - orig_flags & DLM_MRES_MIGRATION ? "migrate" : "recovery", + orig_flags & DLM_MRES_MIGRATION ? "migration" : "recovery", send_to); /* send it */ -- cgit v1.2.3 From 086b3640c10ab448a6993c4bae1508f496f530c4 Mon Sep 17 00:00:00 2001 From: Artem Bityutskiy Date: Mon, 29 Jun 2009 16:25:33 +0300 Subject: UBIFS: dump a little more in case of corruptions In case of corruptions, dump 8192 bytes instead of 4096. The largest node is 4096+ bytes, so it is better to see a node boundary, which is not always possible when only 4096 bytes are printed. Signed-off-by: Artem Bityutskiy Reviewed-by: Adrian Hunter --- fs/ubifs/scan.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/ubifs/scan.c b/fs/ubifs/scan.c index 0ed82479b44b..165c14ba1a46 100644 --- a/fs/ubifs/scan.c +++ b/fs/ubifs/scan.c @@ -238,12 +238,12 @@ void ubifs_scanned_corruption(const struct ubifs_info *c, int lnum, int offs, { int len; - ubifs_err("corrupted data at LEB %d:%d", lnum, offs); + ubifs_err("corruption at LEB %d:%d", lnum, offs); if (dbg_failure_mode) return; len = c->leb_size - offs; - if (len > 4096) - len = 4096; + if (len > 8192) + len = 8192; dbg_err("first %d bytes from LEB %d:%d", len, lnum, offs); print_hex_dump(KERN_DEBUG, "", DUMP_PREFIX_OFFSET, 32, 4, buf, len, 1); } -- cgit v1.2.3 From ed43f2f06cc1cec7ec2dc235c908530bc8c796eb Mon Sep 17 00:00:00 2001 From: Artem Bityutskiy Date: Mon, 29 Jun 2009 17:59:23 +0300 Subject: UBIFS: small amendments in the LEB scanning code This patch fixes few minor things I've spotted while going through code: 1. Better document return codes 2. If 'ubifs_scan_a_node()' returns some thing we do not expect, treat this as an error. 3. Try to do recovery only when 'ubifs_scan()' returns %-EUCLEAN, not on any error. 4. If empty space starts at a non-aligned address, print a message. Signed-off-by: Artem Bityutskiy Reviewed-by: Adrian Hunter --- fs/ubifs/recovery.c | 7 ++++--- fs/ubifs/replay.c | 7 ++++--- fs/ubifs/scan.c | 14 +++++++++----- 3 files changed, 17 insertions(+), 11 deletions(-) (limited to 'fs') diff --git a/fs/ubifs/recovery.c b/fs/ubifs/recovery.c index 805605250f12..093a1ecb700f 100644 --- a/fs/ubifs/recovery.c +++ b/fs/ubifs/recovery.c @@ -543,8 +543,8 @@ static int drop_incomplete_group(struct ubifs_scan_leb *sleb, int *offs) * * This function does a scan of a LEB, but caters for errors that might have * been caused by the unclean unmount from which we are attempting to recover. - * - * This function returns %0 on success and a negative error code on failure. + * Returns %0 in case of success, %-EUCLEAN if an unrecoverable corruption is + * found, and a negative error code in case of failure. */ struct ubifs_scan_leb *ubifs_recover_leb(struct ubifs_info *c, int lnum, int offs, void *sbuf, int grouped) @@ -643,7 +643,8 @@ struct ubifs_scan_leb *ubifs_recover_leb(struct ubifs_info *c, int lnum, goto corrupted; default: dbg_err("unknown"); - goto corrupted; + err = -EINVAL; + goto error; } } diff --git a/fs/ubifs/replay.c b/fs/ubifs/replay.c index 769be42f39d6..2970500f32df 100644 --- a/fs/ubifs/replay.c +++ b/fs/ubifs/replay.c @@ -837,9 +837,10 @@ static int replay_log_leb(struct ubifs_info *c, int lnum, int offs, void *sbuf) dbg_mnt("replay log LEB %d:%d", lnum, offs); sleb = ubifs_scan(c, lnum, offs, sbuf); - if (IS_ERR(sleb)) { - if (c->need_recovery) - sleb = ubifs_recover_log_leb(c, lnum, offs, sbuf); + if (IS_ERR(sleb) ) { + if (PTR_ERR(sleb) != -EUCLEAN || !c->need_recovery) + return PTR_ERR(sleb); + sleb = ubifs_recover_log_leb(c, lnum, offs, sbuf); if (IS_ERR(sleb)) return PTR_ERR(sleb); } diff --git a/fs/ubifs/scan.c b/fs/ubifs/scan.c index 165c14ba1a46..892ebfee4fe5 100644 --- a/fs/ubifs/scan.c +++ b/fs/ubifs/scan.c @@ -256,7 +256,9 @@ void ubifs_scanned_corruption(const struct ubifs_info *c, int lnum, int offs, * @sbuf: scan buffer (must be c->leb_size) * * This function scans LEB number @lnum and returns complete information about - * its contents. Returns an error code in case of failure. + * its contents. Returns the scaned information in case of success and, + * %-EUCLEAN if the LEB neads recovery, and other negative error codes in case + * of failure. */ struct ubifs_scan_leb *ubifs_scan(const struct ubifs_info *c, int lnum, int offs, void *sbuf) @@ -279,7 +281,6 @@ struct ubifs_scan_leb *ubifs_scan(const struct ubifs_info *c, int lnum, cond_resched(); ret = ubifs_scan_a_node(c, buf, len, lnum, offs, 0); - if (ret > 0) { /* Padding bytes or a valid padding node */ offs += ret; @@ -304,7 +305,8 @@ struct ubifs_scan_leb *ubifs_scan(const struct ubifs_info *c, int lnum, goto corrupted; default: dbg_err("unknown"); - goto corrupted; + err = -EINVAL; + goto error; } err = ubifs_add_snod(c, sleb, buf, offs); @@ -317,8 +319,10 @@ struct ubifs_scan_leb *ubifs_scan(const struct ubifs_info *c, int lnum, len -= node_len; } - if (offs % c->min_io_size) - goto corrupted; + if (offs % c->min_io_size) { + ubifs_err("empty space starts at non-aligned offset %d", offs); + goto corrupted;; + } ubifs_end_scan(c, sleb, lnum, offs); -- cgit v1.2.3 From 431102fed3effe4e4e19678830ddab7f05c34bf9 Mon Sep 17 00:00:00 2001 From: Artem Bityutskiy Date: Mon, 29 Jun 2009 18:58:34 +0300 Subject: UBIFS: clean up free space checking recovery.c has 'is_empty()' helper and it is better to use this helper instead of re-implementing it in several places. This patch does this and removes some amount of unneeded code. Signed-off-by: Artem Bityutskiy Reviewed-by: Adrian Hunter --- fs/ubifs/recovery.c | 22 +++------------------- 1 file changed, 3 insertions(+), 19 deletions(-) (limited to 'fs') diff --git a/fs/ubifs/recovery.c b/fs/ubifs/recovery.c index 093a1ecb700f..fe7af9f676b0 100644 --- a/fs/ubifs/recovery.c +++ b/fs/ubifs/recovery.c @@ -357,11 +357,7 @@ static int is_last_write(const struct ubifs_info *c, void *buf, int offs) empty_offs = ALIGN(offs + 1, c->min_io_size); check_len = c->leb_size - empty_offs; p = buf + empty_offs - offs; - - for (; check_len > 0; check_len--) - if (*p++ != 0xff) - return 0; - return 1; + return is_empty(p, check_len); } /** @@ -814,7 +810,7 @@ struct ubifs_scan_leb *ubifs_recover_log_leb(struct ubifs_info *c, int lnum, static int recover_head(const struct ubifs_info *c, int lnum, int offs, void *sbuf) { - int len, err, need_clean = 0; + int len, err; if (c->min_io_size > 1) len = c->min_io_size; @@ -828,19 +824,7 @@ static int recover_head(const struct ubifs_info *c, int lnum, int offs, /* Read at the head location and check it is empty flash */ err = ubi_read(c->ubi, lnum, sbuf, offs, len); - if (err) - need_clean = 1; - else { - uint8_t *p = sbuf; - - while (len--) - if (*p++ != 0xff) { - need_clean = 1; - break; - } - } - - if (need_clean) { + if (err || !is_empty(sbuf, len)) { dbg_rcvry("cleaning head at %d:%d", lnum, offs); if (offs == 0) return ubifs_leb_unmap(c, lnum); -- cgit v1.2.3 From 061125476039a9a998878468a6abe235b1cee347 Mon Sep 17 00:00:00 2001 From: Artem Bityutskiy Date: Mon, 29 Jun 2009 19:27:14 +0300 Subject: UBIFS: fix corruption dump In the 'ubifs_recover_leb()' function, when we find corrupted empty space, we dump 8K starting from the offset where the last node ends. This is OK if the corrupted empty space is somewhere near that offset. But if the corruption is far at the end of the LEB, we will dump all 0xFF bytes and complitely ignore the interesting data. This is observed on a PPC ("kilauea") with NOR flash. This patch changes the behavior and teaches UBIFS to print only interesting data. I.e., now we find where corruption starts and start dumping from that offset. Signed-off-by: Artem Bityutskiy Reviewed-by: Adrian Hunter --- fs/ubifs/recovery.c | 28 ++++++++++++++++++++++++++-- 1 file changed, 26 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/ubifs/recovery.c b/fs/ubifs/recovery.c index fe7af9f676b0..e5f6cf8a1155 100644 --- a/fs/ubifs/recovery.c +++ b/fs/ubifs/recovery.c @@ -52,6 +52,25 @@ static int is_empty(void *buf, int len) return 1; } +/** + * first_non_ff - find offset of the first non-0xff byte. + * @buf: buffer to search in + * @len: length of buffer + * + * This function returns offset of the first non-0xff byte in @buf or %-1 if + * the buffer contains only 0xff bytes. + */ +static int first_non_ff(void *buf, int len) +{ + uint8_t *p = buf; + int i; + + for (i = 0; i < len; i++) + if (*p++ != 0xff) + return i; + return -1; +} + /** * get_master_node - get the last valid master node allowing for corruption. * @c: UBIFS file-system description object @@ -649,8 +668,13 @@ struct ubifs_scan_leb *ubifs_recover_leb(struct ubifs_info *c, int lnum, clean_buf(c, &buf, lnum, &offs, &len); need_clean = 1; } else { - ubifs_err("corrupt empty space at LEB %d:%d", - lnum, offs); + int corruption = first_non_ff(buf, len); + + ubifs_err("corrupt empty space LEB %d:%d, corruption " + "starts at %d", lnum, offs, corruption); + /* Make sure we dump interesting non-0xFF data */ + offs = corruption; + buf += corruption; goto corrupted; } } -- cgit v1.2.3 From c4c1bff64dfff4e6dd0936a0340f56b9284512c8 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Thu, 9 Jul 2009 20:02:48 -0400 Subject: cifs: add pid of initiating process to spnego upcall info cifs: add pid of initiating process to spnego upcall info This will allow the upcall to poke in /proc//environ and get the value of the $KRB5CCNAME env var for the process. Signed-off-by: Jeff Layton Signed-off-by: Steve French --- fs/cifs/cifs_spnego.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/cifs/cifs_spnego.c b/fs/cifs/cifs_spnego.c index 4a4581cb2b5e..051caecf7d67 100644 --- a/fs/cifs/cifs_spnego.c +++ b/fs/cifs/cifs_spnego.c @@ -86,6 +86,9 @@ struct key_type cifs_spnego_key_type = { /* strlen of ";user=" */ #define USER_KEY_LEN 6 +/* strlen of ";pid=0x" */ +#define PID_KEY_LEN 7 + /* get a key struct with a SPNEGO security blob, suitable for session setup */ struct key * cifs_get_spnego_key(struct cifsSesInfo *sesInfo) @@ -103,7 +106,8 @@ cifs_get_spnego_key(struct cifsSesInfo *sesInfo) IP_KEY_LEN + INET6_ADDRSTRLEN + MAX_MECH_STR_LEN + UID_KEY_LEN + (sizeof(uid_t) * 2) + - USER_KEY_LEN + strlen(sesInfo->userName) + 1; + USER_KEY_LEN + strlen(sesInfo->userName) + + PID_KEY_LEN + (sizeof(pid_t) * 2) + 1; spnego_key = ERR_PTR(-ENOMEM); description = kzalloc(desc_len, GFP_KERNEL); @@ -141,6 +145,9 @@ cifs_get_spnego_key(struct cifsSesInfo *sesInfo) dp = description + strlen(description); sprintf(dp, ";user=%s", sesInfo->userName); + dp = description + strlen(description); + sprintf(dp, ";pid=0x%x", current->pid); + cFYI(1, ("key description = %s", description)); spnego_key = request_key(&cifs_spnego_key_type, description, ""); -- cgit v1.2.3 From 01ea95e3b6b16573a491ef98ad63f7a1bdcb504f Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Thu, 9 Jul 2009 20:02:49 -0400 Subject: cifs: rename CIFSSMBUnixSetInfo to CIFSSMBUnixSetPathInfo cifs: rename CIFSSMBUnixSetInfo to CIFSSMBUnixSetPathInfo ...in preparation of adding a SET_FILE_INFO variant. Signed-off-by: Jeff Layton Signed-off-by: Steve French --- fs/cifs/cifsproto.h | 2 +- fs/cifs/cifssmb.c | 6 +++--- fs/cifs/dir.c | 15 ++++++++------- fs/cifs/file.c | 6 +++--- fs/cifs/inode.c | 16 ++++++++-------- 5 files changed, 23 insertions(+), 22 deletions(-) (limited to 'fs') diff --git a/fs/cifs/cifsproto.h b/fs/cifs/cifsproto.h index b2bd83fd2aa4..d95fd427de57 100644 --- a/fs/cifs/cifsproto.h +++ b/fs/cifs/cifsproto.h @@ -220,7 +220,7 @@ struct cifs_unix_set_info_args { dev_t device; }; -extern int CIFSSMBUnixSetInfo(const int xid, struct cifsTconInfo *pTcon, +extern int CIFSSMBUnixSetPathInfo(const int xid, struct cifsTconInfo *pTcon, char *fileName, const struct cifs_unix_set_info_args *args, const struct nls_table *nls_codepage, diff --git a/fs/cifs/cifssmb.c b/fs/cifs/cifssmb.c index 61007c627497..1cd01ba03656 100644 --- a/fs/cifs/cifssmb.c +++ b/fs/cifs/cifssmb.c @@ -5075,9 +5075,9 @@ SetAttrLgcyRetry: #endif /* temporarily unneeded SetAttr legacy function */ int -CIFSSMBUnixSetInfo(const int xid, struct cifsTconInfo *tcon, char *fileName, - const struct cifs_unix_set_info_args *args, - const struct nls_table *nls_codepage, int remap) +CIFSSMBUnixSetPathInfo(const int xid, struct cifsTconInfo *tcon, char *fileName, + const struct cifs_unix_set_info_args *args, + const struct nls_table *nls_codepage, int remap) { TRANSACTION2_SPI_REQ *pSMB = NULL; TRANSACTION2_SPI_RSP *pSMBr = NULL; diff --git a/fs/cifs/dir.c b/fs/cifs/dir.c index ff55fc6932cb..4326ffd90fa9 100644 --- a/fs/cifs/dir.c +++ b/fs/cifs/dir.c @@ -425,9 +425,10 @@ cifs_create(struct inode *inode, struct dentry *direntry, int mode, args.uid = NO_CHANGE_64; args.gid = NO_CHANGE_64; } - CIFSSMBUnixSetInfo(xid, tcon, full_path, &args, - cifs_sb->local_nls, - cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR); + CIFSSMBUnixSetPathInfo(xid, tcon, full_path, &args, + cifs_sb->local_nls, + cifs_sb->mnt_cifs_flags & + CIFS_MOUNT_MAP_SPECIAL_CHR); } else { /* BB implement mode setting via Windows security descriptors e.g. */ @@ -515,10 +516,10 @@ int cifs_mknod(struct inode *inode, struct dentry *direntry, int mode, args.uid = NO_CHANGE_64; args.gid = NO_CHANGE_64; } - rc = CIFSSMBUnixSetInfo(xid, pTcon, full_path, - &args, cifs_sb->local_nls, - cifs_sb->mnt_cifs_flags & - CIFS_MOUNT_MAP_SPECIAL_CHR); + rc = CIFSSMBUnixSetPathInfo(xid, pTcon, full_path, &args, + cifs_sb->local_nls, + cifs_sb->mnt_cifs_flags & + CIFS_MOUNT_MAP_SPECIAL_CHR); if (!rc) { rc = cifs_get_inode_info_unix(&newinode, full_path, diff --git a/fs/cifs/file.c b/fs/cifs/file.c index 97ce4bf89d15..c34b7f8a217b 100644 --- a/fs/cifs/file.c +++ b/fs/cifs/file.c @@ -448,9 +448,9 @@ int cifs_open(struct inode *inode, struct file *file) .mtime = NO_CHANGE_64, .device = 0, }; - CIFSSMBUnixSetInfo(xid, tcon, full_path, &args, - cifs_sb->local_nls, - cifs_sb->mnt_cifs_flags & + CIFSSMBUnixSetPathInfo(xid, tcon, full_path, &args, + cifs_sb->local_nls, + cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR); } } diff --git a/fs/cifs/inode.c b/fs/cifs/inode.c index b22379610d71..ad19007ea05f 100644 --- a/fs/cifs/inode.c +++ b/fs/cifs/inode.c @@ -1241,10 +1241,10 @@ mkdir_get_info: args.uid = NO_CHANGE_64; args.gid = NO_CHANGE_64; } - CIFSSMBUnixSetInfo(xid, pTcon, full_path, &args, - cifs_sb->local_nls, - cifs_sb->mnt_cifs_flags & - CIFS_MOUNT_MAP_SPECIAL_CHR); + CIFSSMBUnixSetPathInfo(xid, pTcon, full_path, &args, + cifs_sb->local_nls, + cifs_sb->mnt_cifs_flags & + CIFS_MOUNT_MAP_SPECIAL_CHR); } else { if (!(cifs_sb->mnt_cifs_flags & CIFS_MOUNT_CIFS_ACL) && (mode & S_IWUGO) == 0) { @@ -1876,10 +1876,10 @@ cifs_setattr_unix(struct dentry *direntry, struct iattr *attrs) args->ctime = NO_CHANGE_64; args->device = 0; - rc = CIFSSMBUnixSetInfo(xid, pTcon, full_path, args, - cifs_sb->local_nls, - cifs_sb->mnt_cifs_flags & - CIFS_MOUNT_MAP_SPECIAL_CHR); + rc = CIFSSMBUnixSetPathInfo(xid, pTcon, full_path, args, + cifs_sb->local_nls, + cifs_sb->mnt_cifs_flags & + CIFS_MOUNT_MAP_SPECIAL_CHR); if (!rc) rc = inode_setattr(inode, attrs); -- cgit v1.2.3 From 654cf14ac0a71c56c1f0032140c3403382ca076b Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Thu, 9 Jul 2009 20:02:49 -0400 Subject: cifs: make a separate function for filling out FILE_UNIX_BASIC_INFO cifs: make a separate function for filling out FILE_UNIX_BASIC_INFO The SET_FILE_INFO variant will need to do the same thing here. Break this code out into a separate function that both variants can call. Signed-off-by: Jeff Layton Signed-off-by: Steve French --- fs/cifs/cifssmb.c | 74 +++++++++++++++++++++++++++++++------------------------ 1 file changed, 42 insertions(+), 32 deletions(-) (limited to 'fs') diff --git a/fs/cifs/cifssmb.c b/fs/cifs/cifssmb.c index 1cd01ba03656..1f3c8a463fcd 100644 --- a/fs/cifs/cifssmb.c +++ b/fs/cifs/cifssmb.c @@ -5074,6 +5074,47 @@ SetAttrLgcyRetry: } #endif /* temporarily unneeded SetAttr legacy function */ +static void +cifs_fill_unix_set_info(FILE_UNIX_BASIC_INFO *data_offset, + const struct cifs_unix_set_info_args *args) +{ + u64 mode = args->mode; + + /* + * Samba server ignores set of file size to zero due to bugs in some + * older clients, but we should be precise - we use SetFileSize to + * set file size and do not want to truncate file size to zero + * accidently as happened on one Samba server beta by putting + * zero instead of -1 here + */ + data_offset->EndOfFile = cpu_to_le64(NO_CHANGE_64); + data_offset->NumOfBytes = cpu_to_le64(NO_CHANGE_64); + data_offset->LastStatusChange = cpu_to_le64(args->ctime); + data_offset->LastAccessTime = cpu_to_le64(args->atime); + data_offset->LastModificationTime = cpu_to_le64(args->mtime); + data_offset->Uid = cpu_to_le64(args->uid); + data_offset->Gid = cpu_to_le64(args->gid); + /* better to leave device as zero when it is */ + data_offset->DevMajor = cpu_to_le64(MAJOR(args->device)); + data_offset->DevMinor = cpu_to_le64(MINOR(args->device)); + data_offset->Permissions = cpu_to_le64(mode); + + if (S_ISREG(mode)) + data_offset->Type = cpu_to_le32(UNIX_FILE); + else if (S_ISDIR(mode)) + data_offset->Type = cpu_to_le32(UNIX_DIR); + else if (S_ISLNK(mode)) + data_offset->Type = cpu_to_le32(UNIX_SYMLINK); + else if (S_ISCHR(mode)) + data_offset->Type = cpu_to_le32(UNIX_CHARDEV); + else if (S_ISBLK(mode)) + data_offset->Type = cpu_to_le32(UNIX_BLOCKDEV); + else if (S_ISFIFO(mode)) + data_offset->Type = cpu_to_le32(UNIX_FIFO); + else if (S_ISSOCK(mode)) + data_offset->Type = cpu_to_le32(UNIX_SOCKET); +} + int CIFSSMBUnixSetPathInfo(const int xid, struct cifsTconInfo *tcon, char *fileName, const struct cifs_unix_set_info_args *args, @@ -5086,7 +5127,6 @@ CIFSSMBUnixSetPathInfo(const int xid, struct cifsTconInfo *tcon, char *fileName, int bytes_returned = 0; FILE_UNIX_BASIC_INFO *data_offset; __u16 params, param_offset, offset, count, byte_count; - __u64 mode = args->mode; cFYI(1, ("In SetUID/GID/Mode")); setPermsRetry: @@ -5137,38 +5177,8 @@ setPermsRetry: pSMB->InformationLevel = cpu_to_le16(SMB_SET_FILE_UNIX_BASIC); pSMB->Reserved4 = 0; pSMB->hdr.smb_buf_length += byte_count; - /* Samba server ignores set of file size to zero due to bugs in some - older clients, but we should be precise - we use SetFileSize to - set file size and do not want to truncate file size to zero - accidently as happened on one Samba server beta by putting - zero instead of -1 here */ - data_offset->EndOfFile = cpu_to_le64(NO_CHANGE_64); - data_offset->NumOfBytes = cpu_to_le64(NO_CHANGE_64); - data_offset->LastStatusChange = cpu_to_le64(args->ctime); - data_offset->LastAccessTime = cpu_to_le64(args->atime); - data_offset->LastModificationTime = cpu_to_le64(args->mtime); - data_offset->Uid = cpu_to_le64(args->uid); - data_offset->Gid = cpu_to_le64(args->gid); - /* better to leave device as zero when it is */ - data_offset->DevMajor = cpu_to_le64(MAJOR(args->device)); - data_offset->DevMinor = cpu_to_le64(MINOR(args->device)); - data_offset->Permissions = cpu_to_le64(mode); - - if (S_ISREG(mode)) - data_offset->Type = cpu_to_le32(UNIX_FILE); - else if (S_ISDIR(mode)) - data_offset->Type = cpu_to_le32(UNIX_DIR); - else if (S_ISLNK(mode)) - data_offset->Type = cpu_to_le32(UNIX_SYMLINK); - else if (S_ISCHR(mode)) - data_offset->Type = cpu_to_le32(UNIX_CHARDEV); - else if (S_ISBLK(mode)) - data_offset->Type = cpu_to_le32(UNIX_BLOCKDEV); - else if (S_ISFIFO(mode)) - data_offset->Type = cpu_to_le32(UNIX_FIFO); - else if (S_ISSOCK(mode)) - data_offset->Type = cpu_to_le32(UNIX_SOCKET); + cifs_fill_unix_set_info(data_offset, args); pSMB->ByteCount = cpu_to_le16(byte_count); rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, -- cgit v1.2.3 From 3bbeeb3c93a961bd01b969dd4395ecac0c09db8d Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Thu, 9 Jul 2009 20:02:50 -0400 Subject: cifs: add and use CIFSSMBUnixSetFileInfo for setattr calls cifs: add and use CIFSSMBUnixSetFileInfo for setattr calls When there's an open filehandle, SET_FILE_INFO is apparently preferred over SET_PATH_INFO. Add a new variant that sets a FILE_UNIX_INFO_BASIC infolevel via SET_FILE_INFO and switch cifs_setattr_unix to use the new call when there's an open filehandle available. Signed-off-by: Jeff Layton Signed-off-by: Steve French --- fs/cifs/cifsproto.h | 4 ++++ fs/cifs/cifssmb.c | 63 +++++++++++++++++++++++++++++++++++++++++++++++++++++ fs/cifs/inode.c | 11 +++++++++- 3 files changed, 77 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/cifs/cifsproto.h b/fs/cifs/cifsproto.h index d95fd427de57..37c11c08c529 100644 --- a/fs/cifs/cifsproto.h +++ b/fs/cifs/cifsproto.h @@ -220,6 +220,10 @@ struct cifs_unix_set_info_args { dev_t device; }; +extern int CIFSSMBUnixSetFileInfo(const int xid, struct cifsTconInfo *tcon, + const struct cifs_unix_set_info_args *args, + u16 fid, u32 pid_of_opener); + extern int CIFSSMBUnixSetPathInfo(const int xid, struct cifsTconInfo *pTcon, char *fileName, const struct cifs_unix_set_info_args *args, diff --git a/fs/cifs/cifssmb.c b/fs/cifs/cifssmb.c index 1f3c8a463fcd..922f5fe2084c 100644 --- a/fs/cifs/cifssmb.c +++ b/fs/cifs/cifssmb.c @@ -5115,6 +5115,69 @@ cifs_fill_unix_set_info(FILE_UNIX_BASIC_INFO *data_offset, data_offset->Type = cpu_to_le32(UNIX_SOCKET); } +int +CIFSSMBUnixSetFileInfo(const int xid, struct cifsTconInfo *tcon, + const struct cifs_unix_set_info_args *args, + u16 fid, u32 pid_of_opener) +{ + struct smb_com_transaction2_sfi_req *pSMB = NULL; + FILE_UNIX_BASIC_INFO *data_offset; + int rc = 0; + u16 params, param_offset, offset, byte_count, count; + + cFYI(1, ("Set Unix Info (via SetFileInfo)")); + rc = small_smb_init(SMB_COM_TRANSACTION2, 15, tcon, (void **) &pSMB); + + if (rc) + return rc; + + pSMB->hdr.Pid = cpu_to_le16((__u16)pid_of_opener); + pSMB->hdr.PidHigh = cpu_to_le16((__u16)(pid_of_opener >> 16)); + + params = 6; + pSMB->MaxSetupCount = 0; + pSMB->Reserved = 0; + pSMB->Flags = 0; + pSMB->Timeout = 0; + pSMB->Reserved2 = 0; + param_offset = offsetof(struct smb_com_transaction2_sfi_req, Fid) - 4; + offset = param_offset + params; + + data_offset = (FILE_UNIX_BASIC_INFO *) + ((char *)(&pSMB->hdr.Protocol) + offset); + count = sizeof(FILE_UNIX_BASIC_INFO); + + pSMB->MaxParameterCount = cpu_to_le16(2); + /* BB find max SMB PDU from sess */ + pSMB->MaxDataCount = cpu_to_le16(1000); + pSMB->SetupCount = 1; + pSMB->Reserved3 = 0; + pSMB->SubCommand = cpu_to_le16(TRANS2_SET_FILE_INFORMATION); + byte_count = 3 /* pad */ + params + count; + pSMB->DataCount = cpu_to_le16(count); + pSMB->ParameterCount = cpu_to_le16(params); + pSMB->TotalDataCount = pSMB->DataCount; + pSMB->TotalParameterCount = pSMB->ParameterCount; + pSMB->ParameterOffset = cpu_to_le16(param_offset); + pSMB->DataOffset = cpu_to_le16(offset); + pSMB->Fid = fid; + pSMB->InformationLevel = cpu_to_le16(SMB_SET_FILE_UNIX_BASIC); + pSMB->Reserved4 = 0; + pSMB->hdr.smb_buf_length += byte_count; + pSMB->ByteCount = cpu_to_le16(byte_count); + + cifs_fill_unix_set_info(data_offset, args); + + rc = SendReceiveNoRsp(xid, tcon->ses, (struct smb_hdr *) pSMB, 0); + if (rc) + cFYI(1, ("Send error in Set Time (SetFileInfo) = %d", rc)); + + /* Note: On -EAGAIN error only caller can retry on handle based calls + since file handle passed in no longer valid */ + + return rc; +} + int CIFSSMBUnixSetPathInfo(const int xid, struct cifsTconInfo *tcon, char *fileName, const struct cifs_unix_set_info_args *args, diff --git a/fs/cifs/inode.c b/fs/cifs/inode.c index ad19007ea05f..55b616bb381e 100644 --- a/fs/cifs/inode.c +++ b/fs/cifs/inode.c @@ -1790,6 +1790,7 @@ cifs_setattr_unix(struct dentry *direntry, struct iattr *attrs) struct cifs_sb_info *cifs_sb = CIFS_SB(inode->i_sb); struct cifsTconInfo *pTcon = cifs_sb->tcon; struct cifs_unix_set_info_args *args = NULL; + struct cifsFileInfo *open_file; cFYI(1, ("setattr_unix on file %s attrs->ia_valid=0x%x", direntry->d_name.name, attrs->ia_valid)); @@ -1876,10 +1877,18 @@ cifs_setattr_unix(struct dentry *direntry, struct iattr *attrs) args->ctime = NO_CHANGE_64; args->device = 0; - rc = CIFSSMBUnixSetPathInfo(xid, pTcon, full_path, args, + open_file = find_writable_file(cifsInode); + if (open_file) { + u16 nfid = open_file->netfid; + u32 npid = open_file->pid; + rc = CIFSSMBUnixSetFileInfo(xid, pTcon, args, nfid, npid); + atomic_dec(&open_file->wrtPending); + } else { + rc = CIFSSMBUnixSetPathInfo(xid, pTcon, full_path, args, cifs_sb->local_nls, cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR); + } if (!rc) rc = inode_setattr(inode, attrs); -- cgit v1.2.3 From b77863bfa153e886f9f8faf1a791ba57a36efed0 Mon Sep 17 00:00:00 2001 From: Steve French Date: Thu, 9 Jul 2009 22:51:38 +0000 Subject: [CIFS] update cifs version number Signed-off-by: Steve French --- fs/cifs/CHANGES | 6 +++++- fs/cifs/cifsfs.h | 2 +- 2 files changed, 6 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/cifs/CHANGES b/fs/cifs/CHANGES index 3a9b7a58a51d..92888aa90749 100644 --- a/fs/cifs/CHANGES +++ b/fs/cifs/CHANGES @@ -5,7 +5,11 @@ client generated ones by default (mount option "serverino" turned on by default if server supports it). Add forceuid and forcegid mount options (so that when negotiating unix extensions specifying which uid mounted does not immediately force the server's reported -uids to be overridden). Add support for scope moutn parm. +uids to be overridden). Add support for scope mount parm. Improve +hard link detection to use same inode for both. Do not set +read-only dos attribute on directories (for chmod) since Windows +explorer special cases this attribute bit for directories for +a different purpose. Version 1.58 ------------ diff --git a/fs/cifs/cifsfs.h b/fs/cifs/cifsfs.h index 586df24c9abb..6c170948300d 100644 --- a/fs/cifs/cifsfs.h +++ b/fs/cifs/cifsfs.h @@ -113,5 +113,5 @@ extern long cifs_ioctl(struct file *filep, unsigned int cmd, unsigned long arg); extern const struct export_operations cifs_export_ops; #endif /* EXPERIMENTAL */ -#define CIFS_VERSION "1.59" +#define CIFS_VERSION "1.60" #endif /* _CIFSFS_H */ -- cgit v1.2.3 From 0b8f18e358384a52c1ed7fa7129c08e7eaf86bb6 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Thu, 9 Jul 2009 01:46:37 -0400 Subject: cifs: convert cifs_get_inode_info and non-posix readdir to use cifs_iget cifs: convert cifs_get_inode_info and non-posix readdir to use cifs_iget Rather than allocating an inode and filling it out, have cifs_get_inode_info fill out a cifs_fattr and call cifs_iget. This means a pretty hefty reorganization of cifs_get_inode_info. For the readdir codepath, add a couple of new functions for filling out cifs_fattr's from different FindFile response infolevels. Finally, remove cifs_new_inode since there are no more callers. Signed-off-by: Jeff Layton Reviewed-by: Christoph Hellwig Signed-off-by: Steve French --- fs/cifs/cifsacl.c | 26 ++-- fs/cifs/cifsglob.h | 2 + fs/cifs/cifsproto.h | 6 +- fs/cifs/inode.c | 397 ++++++++++++++++++++-------------------------------- fs/cifs/readdir.c | 350 +++++++++++---------------------------------- 5 files changed, 252 insertions(+), 529 deletions(-) (limited to 'fs') diff --git a/fs/cifs/cifsacl.c b/fs/cifs/cifsacl.c index 1403b5d86a73..6941c22398a6 100644 --- a/fs/cifs/cifsacl.c +++ b/fs/cifs/cifsacl.c @@ -327,7 +327,7 @@ static void dump_ace(struct cifs_ace *pace, char *end_of_acl) static void parse_dacl(struct cifs_acl *pdacl, char *end_of_acl, struct cifs_sid *pownersid, struct cifs_sid *pgrpsid, - struct inode *inode) + struct cifs_fattr *fattr) { int i; int num_aces = 0; @@ -340,7 +340,7 @@ static void parse_dacl(struct cifs_acl *pdacl, char *end_of_acl, if (!pdacl) { /* no DACL in the security descriptor, set all the permissions for user/group/other */ - inode->i_mode |= S_IRWXUGO; + fattr->cf_mode |= S_IRWXUGO; return; } @@ -357,7 +357,7 @@ static void parse_dacl(struct cifs_acl *pdacl, char *end_of_acl, /* reset rwx permissions for user/group/other. Also, if num_aces is 0 i.e. DACL has no ACEs, user/group/other have no permissions */ - inode->i_mode &= ~(S_IRWXUGO); + fattr->cf_mode &= ~(S_IRWXUGO); acl_base = (char *)pdacl; acl_size = sizeof(struct cifs_acl); @@ -379,17 +379,17 @@ static void parse_dacl(struct cifs_acl *pdacl, char *end_of_acl, if (compare_sids(&(ppace[i]->sid), pownersid)) access_flags_to_mode(ppace[i]->access_req, ppace[i]->type, - &(inode->i_mode), + &fattr->cf_mode, &user_mask); if (compare_sids(&(ppace[i]->sid), pgrpsid)) access_flags_to_mode(ppace[i]->access_req, ppace[i]->type, - &(inode->i_mode), + &fattr->cf_mode, &group_mask); if (compare_sids(&(ppace[i]->sid), &sid_everyone)) access_flags_to_mode(ppace[i]->access_req, ppace[i]->type, - &(inode->i_mode), + &fattr->cf_mode, &other_mask); /* memcpy((void *)(&(cifscred->aces[i])), @@ -464,7 +464,7 @@ static int parse_sid(struct cifs_sid *psid, char *end_of_acl) /* Convert CIFS ACL to POSIX form */ static int parse_sec_desc(struct cifs_ntsd *pntsd, int acl_len, - struct inode *inode) + struct cifs_fattr *fattr) { int rc; struct cifs_sid *owner_sid_ptr, *group_sid_ptr; @@ -472,7 +472,7 @@ static int parse_sec_desc(struct cifs_ntsd *pntsd, int acl_len, char *end_of_acl = ((char *)pntsd) + acl_len; __u32 dacloffset; - if ((inode == NULL) || (pntsd == NULL)) + if (pntsd == NULL) return -EIO; owner_sid_ptr = (struct cifs_sid *)((char *)pntsd + @@ -497,7 +497,7 @@ static int parse_sec_desc(struct cifs_ntsd *pntsd, int acl_len, if (dacloffset) parse_dacl(dacl_ptr, end_of_acl, owner_sid_ptr, - group_sid_ptr, inode); + group_sid_ptr, fattr); else cFYI(1, ("no ACL")); /* BB grant all or default perms? */ @@ -508,7 +508,6 @@ static int parse_sec_desc(struct cifs_ntsd *pntsd, int acl_len, memcpy((void *)(&(cifscred->gsid)), (void *)group_sid_ptr, sizeof(struct cifs_sid)); */ - return 0; } @@ -671,8 +670,9 @@ static int set_cifs_acl(struct cifs_ntsd *pnntsd, __u32 acllen, } /* Translate the CIFS ACL (simlar to NTFS ACL) for a file into mode bits */ -void acl_to_uid_mode(struct cifs_sb_info *cifs_sb, struct inode *inode, - const char *path, const __u16 *pfid) +void +cifs_acl_to_fattr(struct cifs_sb_info *cifs_sb, struct cifs_fattr *fattr, + struct inode *inode, const char *path, const __u16 *pfid) { struct cifs_ntsd *pntsd = NULL; u32 acllen = 0; @@ -687,7 +687,7 @@ void acl_to_uid_mode(struct cifs_sb_info *cifs_sb, struct inode *inode, /* if we can retrieve the ACL, now parse Access Control Entries, ACEs */ if (pntsd) - rc = parse_sec_desc(pntsd, acllen, inode); + rc = parse_sec_desc(pntsd, acllen, fattr); if (rc) cFYI(1, ("parse sec desc failed rc = %d", rc)); diff --git a/fs/cifs/cifsglob.h b/fs/cifs/cifsglob.h index e6435cba8113..8bcf5a4bcded 100644 --- a/fs/cifs/cifsglob.h +++ b/fs/cifs/cifsglob.h @@ -479,6 +479,8 @@ struct dfs_info3_param { */ #define CIFS_FATTR_DFS_REFERRAL 0x1 +#define CIFS_FATTR_DELETE_PENDING 0x2 +#define CIFS_FATTR_NEED_REVAL 0x4 struct cifs_fattr { u32 cf_flags; diff --git a/fs/cifs/cifsproto.h b/fs/cifs/cifsproto.h index 37c11c08c529..da8fbf565991 100644 --- a/fs/cifs/cifsproto.h +++ b/fs/cifs/cifsproto.h @@ -102,7 +102,6 @@ extern void cifs_unix_basic_to_fattr(struct cifs_fattr *fattr, FILE_UNIX_BASIC_INFO *info, struct cifs_sb_info *cifs_sb); extern void cifs_fattr_to_inode(struct inode *inode, struct cifs_fattr *fattr); -extern struct inode *cifs_new_inode(struct super_block *sb, __u64 *inum); extern struct inode *cifs_iget(struct super_block *sb, struct cifs_fattr *fattr); @@ -113,8 +112,9 @@ extern int cifs_get_inode_info(struct inode **pinode, extern int cifs_get_inode_info_unix(struct inode **pinode, const unsigned char *search_path, struct super_block *sb, int xid); -extern void acl_to_uid_mode(struct cifs_sb_info *cifs_sb, struct inode *inode, - const char *path, const __u16 *pfid); +extern void cifs_acl_to_fattr(struct cifs_sb_info *cifs_sb, + struct cifs_fattr *fattr, struct inode *inode, + const char *path, const __u16 *pfid); extern int mode_to_acl(struct inode *inode, const char *path, __u64); extern int cifs_mount(struct super_block *, struct cifs_sb_info *, char *, diff --git a/fs/cifs/inode.c b/fs/cifs/inode.c index 55b616bb381e..a807397f444e 100644 --- a/fs/cifs/inode.c +++ b/fs/cifs/inode.c @@ -82,23 +82,34 @@ void cifs_fattr_to_inode(struct inode *inode, struct cifs_fattr *fattr) { struct cifsInodeInfo *cifs_i = CIFS_I(inode); - unsigned long now = jiffies; + struct cifs_sb_info *cifs_sb = CIFS_SB(inode->i_sb); + unsigned long oldtime = cifs_i->time; inode->i_atime = fattr->cf_atime; inode->i_mtime = fattr->cf_mtime; inode->i_ctime = fattr->cf_ctime; - inode->i_mode = fattr->cf_mode; inode->i_rdev = fattr->cf_rdev; inode->i_nlink = fattr->cf_nlink; inode->i_uid = fattr->cf_uid; inode->i_gid = fattr->cf_gid; + /* if dynperm is set, don't clobber existing mode */ + if (inode->i_state & I_NEW || + !(cifs_sb->mnt_cifs_flags & CIFS_MOUNT_DYNPERM)) + inode->i_mode = fattr->cf_mode; + cifs_i->cifsAttrs = fattr->cf_cifsattrs; cifs_i->uniqueid = fattr->cf_uniqueid; + if (fattr->cf_flags & CIFS_FATTR_NEED_REVAL) + cifs_i->time = 0; + else + cifs_i->time = jiffies; + cFYI(1, ("inode 0x%p old_time=%ld new_time=%ld", inode, - cifs_i->time, now)); - cifs_i->time = now; + oldtime, cifs_i->time)); + + cifs_i->delete_pending = fattr->cf_flags & CIFS_FATTR_DELETE_PENDING; /* * Can't safely change the file size here if the client is writing to @@ -219,49 +230,6 @@ cifs_create_dfs_fattr(struct cifs_fattr *fattr, struct super_block *sb) fattr->cf_flags |= CIFS_FATTR_DFS_REFERRAL; } -/** - * cifs_new inode - create new inode, initialize, and hash it - * @sb - pointer to superblock - * @inum - if valid pointer and serverino is enabled, replace i_ino with val - * - * Create a new inode, initialize it for CIFS and hash it. Returns the new - * inode or NULL if one couldn't be allocated. - * - * If the share isn't mounted with "serverino" or inum is a NULL pointer then - * we'll just use the inode number assigned by new_inode(). Note that this can - * mean i_ino collisions since the i_ino assigned by new_inode is not - * guaranteed to be unique. - */ -struct inode * -cifs_new_inode(struct super_block *sb, __u64 *inum) -{ - struct inode *inode; - - inode = new_inode(sb); - if (inode == NULL) - return NULL; - - /* - * BB: Is i_ino == 0 legal? Here, we assume that it is. If it isn't we - * stop passing inum as ptr. Are there sanity checks we can use to - * ensure that the server is really filling in that field? Also, - * if serverino is disabled, perhaps we should be using iunique()? - */ - if (inum && (CIFS_SB(sb)->mnt_cifs_flags & CIFS_MOUNT_SERVER_INUM)) - inode->i_ino = (unsigned long) *inum; - - /* - * must set this here instead of cifs_alloc_inode since VFS will - * clobber i_flags - */ - if (sb->s_flags & MS_NOATIME) - inode->i_flags |= S_NOATIME | S_NOCMTIME; - - insert_inode_hash(inode); - - return inode; -} - int cifs_get_inode_info_unix(struct inode **pinode, const unsigned char *full_path, struct super_block *sb, int xid) @@ -302,9 +270,9 @@ int cifs_get_inode_info_unix(struct inode **pinode, return rc; } -static int decode_sfu_inode(struct inode *inode, __u64 size, - const unsigned char *path, - struct cifs_sb_info *cifs_sb, int xid) +static int +cifs_sfu_type(struct cifs_fattr *fattr, const unsigned char *path, + struct cifs_sb_info *cifs_sb, int xid) { int rc; int oplock = 0; @@ -316,10 +284,15 @@ static int decode_sfu_inode(struct inode *inode, __u64 size, pbuf = buf; - if (size == 0) { - inode->i_mode |= S_IFIFO; + fattr->cf_mode &= ~S_IFMT; + + if (fattr->cf_eof == 0) { + fattr->cf_mode |= S_IFIFO; + fattr->cf_dtype = DT_FIFO; return 0; - } else if (size < 8) { + } else if (fattr->cf_eof < 8) { + fattr->cf_mode |= S_IFREG; + fattr->cf_dtype = DT_REG; return -EINVAL; /* EOPNOTSUPP? */ } @@ -331,42 +304,46 @@ static int decode_sfu_inode(struct inode *inode, __u64 size, if (rc == 0) { int buf_type = CIFS_NO_BUFFER; /* Read header */ - rc = CIFSSMBRead(xid, pTcon, - netfid, + rc = CIFSSMBRead(xid, pTcon, netfid, 24 /* length */, 0 /* offset */, &bytes_read, &pbuf, &buf_type); if ((rc == 0) && (bytes_read >= 8)) { if (memcmp("IntxBLK", pbuf, 8) == 0) { cFYI(1, ("Block device")); - inode->i_mode |= S_IFBLK; + fattr->cf_mode |= S_IFBLK; + fattr->cf_dtype = DT_BLK; if (bytes_read == 24) { /* we have enough to decode dev num */ __u64 mjr; /* major */ __u64 mnr; /* minor */ mjr = le64_to_cpu(*(__le64 *)(pbuf+8)); mnr = le64_to_cpu(*(__le64 *)(pbuf+16)); - inode->i_rdev = MKDEV(mjr, mnr); + fattr->cf_rdev = MKDEV(mjr, mnr); } } else if (memcmp("IntxCHR", pbuf, 8) == 0) { cFYI(1, ("Char device")); - inode->i_mode |= S_IFCHR; + fattr->cf_mode |= S_IFCHR; + fattr->cf_dtype = DT_CHR; if (bytes_read == 24) { /* we have enough to decode dev num */ __u64 mjr; /* major */ __u64 mnr; /* minor */ mjr = le64_to_cpu(*(__le64 *)(pbuf+8)); mnr = le64_to_cpu(*(__le64 *)(pbuf+16)); - inode->i_rdev = MKDEV(mjr, mnr); + fattr->cf_rdev = MKDEV(mjr, mnr); } } else if (memcmp("IntxLNK", pbuf, 7) == 0) { cFYI(1, ("Symlink")); - inode->i_mode |= S_IFLNK; + fattr->cf_mode |= S_IFLNK; + fattr->cf_dtype = DT_LNK; } else { - inode->i_mode |= S_IFREG; /* file? */ + fattr->cf_mode |= S_IFREG; /* file? */ + fattr->cf_dtype = DT_REG; rc = -EOPNOTSUPP; } } else { - inode->i_mode |= S_IFREG; /* then it is a file */ + fattr->cf_mode |= S_IFREG; /* then it is a file */ + fattr->cf_dtype = DT_REG; rc = -EOPNOTSUPP; /* or some unknown SFU type */ } CIFSSMBClose(xid, pTcon, netfid); @@ -376,9 +353,13 @@ static int decode_sfu_inode(struct inode *inode, __u64 size, #define SFBITS_MASK (S_ISVTX | S_ISGID | S_ISUID) /* SETFILEBITS valid bits */ -static int get_sfu_mode(struct inode *inode, - const unsigned char *path, - struct cifs_sb_info *cifs_sb, int xid) +/* + * Fetch mode bits as provided by SFU. + * + * FIXME: Doesn't this clobber the type bit we got from cifs_sfu_type ? + */ +static int cifs_sfu_mode(struct cifs_fattr *fattr, const unsigned char *path, + struct cifs_sb_info *cifs_sb, int xid) { #ifdef CONFIG_CIFS_XATTR ssize_t rc; @@ -386,68 +367,80 @@ static int get_sfu_mode(struct inode *inode, __u32 mode; rc = CIFSSMBQueryEA(xid, cifs_sb->tcon, path, "SETFILEBITS", - ea_value, 4 /* size of buf */, cifs_sb->local_nls, - cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR); + ea_value, 4 /* size of buf */, cifs_sb->local_nls, + cifs_sb->mnt_cifs_flags & + CIFS_MOUNT_MAP_SPECIAL_CHR); if (rc < 0) return (int)rc; else if (rc > 3) { mode = le32_to_cpu(*((__le32 *)ea_value)); - inode->i_mode &= ~SFBITS_MASK; - cFYI(1, ("special bits 0%o org mode 0%o", mode, inode->i_mode)); - inode->i_mode = (mode & SFBITS_MASK) | inode->i_mode; + fattr->cf_mode &= ~SFBITS_MASK; + cFYI(1, ("special bits 0%o org mode 0%o", mode, + fattr->cf_mode)); + fattr->cf_mode = (mode & SFBITS_MASK) | fattr->cf_mode; cFYI(1, ("special mode bits 0%o", mode)); - return 0; - } else { - return 0; } + + return 0; #else return -EOPNOTSUPP; #endif } -/* - * Needed to setup inode data for the directory which is the - * junction to the new submount (ie to setup the fake directory - * which represents a DFS referral) - */ -static void fill_fake_finddata(FILE_ALL_INFO *pfnd_dat, - struct super_block *sb) +/* Fill a cifs_fattr struct with info from FILE_ALL_INFO */ +void +cifs_all_info_to_fattr(struct cifs_fattr *fattr, FILE_ALL_INFO *info, + struct cifs_sb_info *cifs_sb, bool adjust_tz) { - memset(pfnd_dat, 0, sizeof(FILE_ALL_INFO)); - -/* __le64 pfnd_dat->AllocationSize = cpu_to_le64(0); - __le64 pfnd_dat->EndOfFile = cpu_to_le64(0); - __u8 pfnd_dat->DeletePending = 0; - __u8 pfnd_data->Directory = 0; - __le32 pfnd_dat->EASize = 0; - __u64 pfnd_dat->IndexNumber = 0; - __u64 pfnd_dat->IndexNumber1 = 0; */ - pfnd_dat->CreationTime = - cpu_to_le64(cifs_UnixTimeToNT(CURRENT_TIME)); - pfnd_dat->LastAccessTime = - cpu_to_le64(cifs_UnixTimeToNT(CURRENT_TIME)); - pfnd_dat->LastWriteTime = - cpu_to_le64(cifs_UnixTimeToNT(CURRENT_TIME)); - pfnd_dat->ChangeTime = - cpu_to_le64(cifs_UnixTimeToNT(CURRENT_TIME)); - pfnd_dat->Attributes = cpu_to_le32(ATTR_DIRECTORY); - pfnd_dat->NumberOfLinks = cpu_to_le32(2); + memset(fattr, 0, sizeof(*fattr)); + fattr->cf_cifsattrs = le32_to_cpu(info->Attributes); + if (info->DeletePending) + fattr->cf_flags |= CIFS_FATTR_DELETE_PENDING; + + if (info->LastAccessTime) + fattr->cf_atime = cifs_NTtimeToUnix(info->LastAccessTime); + else + fattr->cf_atime = CURRENT_TIME; + + fattr->cf_ctime = cifs_NTtimeToUnix(info->ChangeTime); + fattr->cf_mtime = cifs_NTtimeToUnix(info->LastWriteTime); + + if (adjust_tz) { + fattr->cf_ctime.tv_sec += cifs_sb->tcon->ses->server->timeAdj; + fattr->cf_mtime.tv_sec += cifs_sb->tcon->ses->server->timeAdj; + } + + fattr->cf_eof = le64_to_cpu(info->EndOfFile); + fattr->cf_bytes = le64_to_cpu(info->AllocationSize); + + if (fattr->cf_cifsattrs & ATTR_DIRECTORY) { + fattr->cf_mode = S_IFDIR | cifs_sb->mnt_dir_mode; + fattr->cf_dtype = DT_DIR; + } else { + fattr->cf_mode = S_IFREG | cifs_sb->mnt_file_mode; + fattr->cf_dtype = DT_REG; + } + + /* clear write bits if ATTR_READONLY is set */ + if (fattr->cf_cifsattrs & ATTR_READONLY) + fattr->cf_mode &= ~(S_IWUGO); + + fattr->cf_nlink = le32_to_cpu(info->NumberOfLinks); + + fattr->cf_uid = cifs_sb->mnt_uid; + fattr->cf_gid = cifs_sb->mnt_gid; } int cifs_get_inode_info(struct inode **pinode, const unsigned char *full_path, FILE_ALL_INFO *pfindData, struct super_block *sb, int xid, const __u16 *pfid) { - int rc = 0; - __u32 attr; - struct cifsInodeInfo *cifsInfo; + int rc = 0, tmprc; struct cifsTconInfo *pTcon; - struct inode *inode; struct cifs_sb_info *cifs_sb = CIFS_SB(sb); char *buf = NULL; bool adjustTZ = false; - bool is_dfs_referral = false; - umode_t default_mode; + struct cifs_fattr fattr; pTcon = cifs_sb->tcon; cFYI(1, ("Getting info on %s", full_path)); @@ -482,163 +475,82 @@ int cifs_get_inode_info(struct inode **pinode, adjustTZ = true; } } - /* dump_mem("\nQPathInfo return data",&findData, sizeof(findData)); */ - if (rc == -EREMOTE) { - is_dfs_referral = true; - fill_fake_finddata(pfindData, sb); + + if (!rc) { + cifs_all_info_to_fattr(&fattr, (FILE_ALL_INFO *) pfindData, + cifs_sb, adjustTZ); + } else if (rc == -EREMOTE) { + cifs_create_dfs_fattr(&fattr, sb); rc = 0; - } else if (rc) + } else { goto cgii_exit; + } - attr = le32_to_cpu(pfindData->Attributes); - - /* get new inode */ + /* + * If an inode wasn't passed in, then get the inode number + * + * Is an i_ino of zero legal? Can we use that to check if the server + * supports returning inode numbers? Are there other sanity checks we + * can use to ensure that the server is really filling in that field? + * + * We can not use the IndexNumber field by default from Windows or + * Samba (in ALL_INFO buf) but we can request it explicitly. The SNIA + * CIFS spec claims that this value is unique within the scope of a + * share, and the windows docs hint that it's actually unique + * per-machine. + * + * There may be higher info levels that work but are there Windows + * server or network appliances for which IndexNumber field is not + * guaranteed unique? + */ if (*pinode == NULL) { - __u64 inode_num; - __u64 *pinum = &inode_num; - - /* Is an i_ino of zero legal? Can we use that to check - if the server supports returning inode numbers? Are - there other sanity checks we can use to ensure that - the server is really filling in that field? */ - - /* We can not use the IndexNumber field by default from - Windows or Samba (in ALL_INFO buf) but we can request - it explicitly. It may not be unique presumably if - the server has multiple devices mounted under one share */ - - /* There may be higher info levels that work but are - there Windows server or network appliances for which - IndexNumber field is not guaranteed unique? */ - if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_SERVER_INUM) { int rc1 = 0; rc1 = CIFSGetSrvInodeNumber(xid, pTcon, - full_path, pinum, + full_path, &fattr.cf_uniqueid, cifs_sb->local_nls, cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR); if (rc1) { - cFYI(1, ("GetSrvInodeNum rc %d", rc1)); - pinum = NULL; /* BB EOPNOSUPP disable SERVER_INUM? */ + cFYI(1, ("GetSrvInodeNum rc %d", rc1)); + fattr.cf_uniqueid = iunique(sb, ROOT_I); } } else { - pinum = NULL; + fattr.cf_uniqueid = iunique(sb, ROOT_I); } - - *pinode = cifs_new_inode(sb, pinum); - if (*pinode == NULL) { - rc = -ENOMEM; - goto cgii_exit; - } - } - inode = *pinode; - cifsInfo = CIFS_I(inode); - cifsInfo->cifsAttrs = attr; - cifsInfo->delete_pending = pfindData->DeletePending ? true : false; - cFYI(1, ("Old time %ld", cifsInfo->time)); - cifsInfo->time = jiffies; - cFYI(1, ("New time %ld", cifsInfo->time)); - - /* blksize needs to be multiple of two. So safer to default to - blksize and blkbits set in superblock so 2**blkbits and blksize - will match rather than setting to: - (pTcon->ses->server->maxBuf - MAX_CIFS_HDR_SIZE) & 0xFFFFFE00;*/ - - /* Linux can not store file creation time so ignore it */ - if (pfindData->LastAccessTime) - inode->i_atime = cifs_NTtimeToUnix(pfindData->LastAccessTime); - else /* do not need to use current_fs_time - time not stored */ - inode->i_atime = CURRENT_TIME; - inode->i_mtime = cifs_NTtimeToUnix(pfindData->LastWriteTime); - inode->i_ctime = cifs_NTtimeToUnix(pfindData->ChangeTime); - cFYI(DBG2, ("Attributes came in as 0x%x", attr)); - if (adjustTZ && (pTcon->ses) && (pTcon->ses->server)) { - inode->i_ctime.tv_sec += pTcon->ses->server->timeAdj; - inode->i_mtime.tv_sec += pTcon->ses->server->timeAdj; - } - - /* get default inode mode */ - if (attr & ATTR_DIRECTORY) - default_mode = cifs_sb->mnt_dir_mode; - else - default_mode = cifs_sb->mnt_file_mode; - - /* set permission bits */ - if (atomic_read(&cifsInfo->inUse) == 0 || - (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_DYNPERM) == 0) - inode->i_mode = default_mode; - else { - /* just reenable write bits if !ATTR_READONLY */ - if ((inode->i_mode & S_IWUGO) == 0 && - (attr & ATTR_READONLY) == 0) - inode->i_mode |= (S_IWUGO & default_mode); - - inode->i_mode &= ~S_IFMT; - } - /* clear write bits if ATTR_READONLY is set */ - if (attr & ATTR_READONLY) - inode->i_mode &= ~S_IWUGO; - - /* set inode type */ - if ((attr & ATTR_SYSTEM) && - (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_UNX_EMUL)) { - /* no need to fix endianness on 0 */ - if (pfindData->EndOfFile == 0) - inode->i_mode |= S_IFIFO; - else if (decode_sfu_inode(inode, - le64_to_cpu(pfindData->EndOfFile), - full_path, cifs_sb, xid)) - cFYI(1, ("unknown SFU file type\n")); } else { - if (attr & ATTR_DIRECTORY) - inode->i_mode |= S_IFDIR; - else - inode->i_mode |= S_IFREG; + fattr.cf_uniqueid = CIFS_I(*pinode)->uniqueid; } - cifsInfo->server_eof = le64_to_cpu(pfindData->EndOfFile); - spin_lock(&inode->i_lock); - if (is_size_safe_to_change(cifsInfo, cifsInfo->server_eof)) { - /* can not safely shrink the file size here if the - client is writing to it due to potential races */ - i_size_write(inode, cifsInfo->server_eof); - - /* 512 bytes (2**9) is the fake blocksize that must be - used for this calculation */ - inode->i_blocks = (512 - 1 + le64_to_cpu( - pfindData->AllocationSize)) >> 9; + /* query for SFU type info if supported and needed */ + if (fattr.cf_cifsattrs & ATTR_SYSTEM && + cifs_sb->mnt_cifs_flags & CIFS_MOUNT_UNX_EMUL) { + tmprc = cifs_sfu_type(&fattr, full_path, cifs_sb, xid); + if (tmprc) + cFYI(1, ("cifs_sfu_type failed: %d", tmprc)); } - spin_unlock(&inode->i_lock); - inode->i_nlink = le32_to_cpu(pfindData->NumberOfLinks); - - /* BB fill in uid and gid here? with help from winbind? - or retrieve from NTFS stream extended attribute */ #ifdef CONFIG_CIFS_EXPERIMENTAL /* fill in 0777 bits from ACL */ if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_CIFS_ACL) { cFYI(1, ("Getting mode bits from ACL")); - acl_to_uid_mode(cifs_sb, inode, full_path, pfid); + cifs_acl_to_fattr(cifs_sb, &fattr, *pinode, full_path, pfid); } #endif - if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_UNX_EMUL) { - /* fill in remaining high mode bits e.g. SUID, VTX */ - get_sfu_mode(inode, full_path, cifs_sb, xid); - } else if (atomic_read(&cifsInfo->inUse) == 0) { - inode->i_uid = cifs_sb->mnt_uid; - inode->i_gid = cifs_sb->mnt_gid; - /* set so we do not keep refreshing these fields with - bad data after user has changed them in memory */ - atomic_set(&cifsInfo->inUse, 1); - } - - cifs_set_ops(inode, is_dfs_referral); - + /* fill in remaining high mode bits e.g. SUID, VTX */ + if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_UNX_EMUL) + cifs_sfu_mode(&fattr, full_path, cifs_sb, xid); + if (!*pinode) { + *pinode = cifs_iget(sb, &fattr); + if (!*pinode) + rc = -ENOMEM; + } else { + cifs_fattr_to_inode(*pinode, &fattr); + } cgii_exit: kfree(buf); @@ -753,21 +665,14 @@ struct inode *cifs_root_iget(struct super_block *sb, unsigned long ino) return ERR_PTR(-ENOMEM); xid = GetXid(); - if (cifs_sb->tcon->unix_ext) { + if (cifs_sb->tcon->unix_ext) rc = cifs_get_inode_info_unix(&inode, full_path, sb, xid); - if (!inode) - return ERR_PTR(-ENOMEM); - } else { - inode = iget_locked(sb, ino); - if (!inode) - return ERR_PTR(-ENOMEM); - if (!(inode->i_state & I_NEW)) - return inode; - - rc = cifs_get_inode_info(&inode, full_path, NULL, inode->i_sb, + else + rc = cifs_get_inode_info(&inode, full_path, NULL, sb, xid, NULL); - unlock_new_inode(inode); - } + + if (!inode) + return ERR_PTR(-ENOMEM); if (rc && cifs_sb->tcon->ipc) { cFYI(1, ("ipc connection - fake read inode")); diff --git a/fs/cifs/readdir.c b/fs/cifs/readdir.c index 231aa6953f83..f823a4a208a7 100644 --- a/fs/cifs/readdir.c +++ b/fs/cifs/readdir.c @@ -112,239 +112,74 @@ cifs_readdir_lookup(struct dentry *parent, struct qstr *name, return dentry; } -/* Returns 1 if new inode created, 2 if both dentry and inode were */ -/* Might check in the future if inode number changed so we can rehash inode */ -static int -construct_dentry(struct qstr *qstring, struct file *file, - struct inode **ptmp_inode, struct dentry **pnew_dentry, - __u64 *inum) +static void +cifs_fill_common_info(struct cifs_fattr *fattr, struct cifs_sb_info *cifs_sb) { - struct dentry *tmp_dentry = NULL; - struct super_block *sb = file->f_path.dentry->d_sb; - int rc = 0; + fattr->cf_uid = cifs_sb->mnt_uid; + fattr->cf_gid = cifs_sb->mnt_gid; - cFYI(1, ("For %s", qstring->name)); - - tmp_dentry = d_lookup(file->f_path.dentry, qstring); - if (tmp_dentry) { - /* BB: overwrite old name? i.e. tmp_dentry->d_name and - * tmp_dentry->d_name.len?? - */ - cFYI(0, ("existing dentry with inode 0x%p", - tmp_dentry->d_inode)); - *ptmp_inode = tmp_dentry->d_inode; - if (*ptmp_inode == NULL) { - *ptmp_inode = cifs_new_inode(sb, inum); - if (*ptmp_inode == NULL) - return rc; - rc = 1; - } + if (fattr->cf_cifsattrs & ATTR_DIRECTORY) { + fattr->cf_mode = S_IFDIR | cifs_sb->mnt_dir_mode; + fattr->cf_dtype = DT_DIR; } else { - tmp_dentry = d_alloc(file->f_path.dentry, qstring); - if (tmp_dentry == NULL) { - cERROR(1, ("Failed allocating dentry")); - *ptmp_inode = NULL; - return rc; - } - - if (CIFS_SB(sb)->tcon->nocase) - tmp_dentry->d_op = &cifs_ci_dentry_ops; - else - tmp_dentry->d_op = &cifs_dentry_ops; - - *ptmp_inode = cifs_new_inode(sb, inum); - if (*ptmp_inode == NULL) - return rc; - rc = 2; - } - - tmp_dentry->d_time = jiffies; - *pnew_dentry = tmp_dentry; - return rc; -} - -static void fill_in_inode(struct inode *tmp_inode, int new_buf_type, - char *buf, unsigned int *pobject_type, int isNewInode) -{ - loff_t local_size; - struct timespec local_mtime; - - struct cifsInodeInfo *cifsInfo = CIFS_I(tmp_inode); - struct cifs_sb_info *cifs_sb = CIFS_SB(tmp_inode->i_sb); - __u32 attr; - __u64 allocation_size; - __u64 end_of_file; - umode_t default_mode; - - /* save mtime and size */ - local_mtime = tmp_inode->i_mtime; - local_size = tmp_inode->i_size; - - if (new_buf_type) { - FILE_DIRECTORY_INFO *pfindData = (FILE_DIRECTORY_INFO *)buf; - - attr = le32_to_cpu(pfindData->ExtFileAttributes); - allocation_size = le64_to_cpu(pfindData->AllocationSize); - end_of_file = le64_to_cpu(pfindData->EndOfFile); - tmp_inode->i_atime = - cifs_NTtimeToUnix(pfindData->LastAccessTime); - tmp_inode->i_mtime = - cifs_NTtimeToUnix(pfindData->LastWriteTime); - tmp_inode->i_ctime = - cifs_NTtimeToUnix(pfindData->ChangeTime); - } else { /* legacy, OS2 and DOS style */ - int offset = cifs_sb->tcon->ses->server->timeAdj; - FIND_FILE_STANDARD_INFO *pfindData = - (FIND_FILE_STANDARD_INFO *)buf; - - tmp_inode->i_mtime = cnvrtDosUnixTm(pfindData->LastWriteDate, - pfindData->LastWriteTime, - offset); - tmp_inode->i_atime = cnvrtDosUnixTm(pfindData->LastAccessDate, - pfindData->LastAccessTime, - offset); - tmp_inode->i_ctime = cnvrtDosUnixTm(pfindData->LastWriteDate, - pfindData->LastWriteTime, - offset); - attr = le16_to_cpu(pfindData->Attributes); - allocation_size = le32_to_cpu(pfindData->AllocationSize); - end_of_file = le32_to_cpu(pfindData->DataSize); - } - - /* Linux can not store file creation time unfortunately so ignore it */ - - cifsInfo->cifsAttrs = attr; -#ifdef CONFIG_CIFS_EXPERIMENTAL - if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_CIFS_ACL) { - /* get more accurate mode via ACL - so force inode refresh */ - cifsInfo->time = 0; - } else -#endif /* CONFIG_CIFS_EXPERIMENTAL */ - cifsInfo->time = jiffies; - - /* treat dos attribute of read-only as read-only mode bit e.g. 555? */ - /* 2767 perms - indicate mandatory locking */ - /* BB fill in uid and gid here? with help from winbind? - or retrieve from NTFS stream extended attribute */ - if (atomic_read(&cifsInfo->inUse) == 0) { - tmp_inode->i_uid = cifs_sb->mnt_uid; - tmp_inode->i_gid = cifs_sb->mnt_gid; + fattr->cf_mode = S_IFREG | cifs_sb->mnt_file_mode; + fattr->cf_dtype = DT_REG; } - if (attr & ATTR_DIRECTORY) - default_mode = cifs_sb->mnt_dir_mode; - else - default_mode = cifs_sb->mnt_file_mode; - - /* set initial permissions */ - if ((atomic_read(&cifsInfo->inUse) == 0) || - (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_DYNPERM) == 0) - tmp_inode->i_mode = default_mode; - else { - /* just reenable write bits if !ATTR_READONLY */ - if ((tmp_inode->i_mode & S_IWUGO) == 0 && - (attr & ATTR_READONLY) == 0) - tmp_inode->i_mode |= (S_IWUGO & default_mode); - - tmp_inode->i_mode &= ~S_IFMT; - } - - /* clear write bits if ATTR_READONLY is set */ - if (attr & ATTR_READONLY) - tmp_inode->i_mode &= ~S_IWUGO; + if (fattr->cf_cifsattrs & ATTR_READONLY) + fattr->cf_mode &= ~S_IWUGO; - /* set inode type */ - if ((attr & ATTR_SYSTEM) && - (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_UNX_EMUL)) { - if (end_of_file == 0) { - tmp_inode->i_mode |= S_IFIFO; - *pobject_type = DT_FIFO; + if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_UNX_EMUL && + fattr->cf_cifsattrs & ATTR_SYSTEM) { + if (fattr->cf_eof == 0) { + fattr->cf_mode &= ~S_IFMT; + fattr->cf_mode |= S_IFIFO; + fattr->cf_dtype = DT_FIFO; } else { /* - * trying to get the type can be slow, so just call - * this a regular file for now, and mark for reval + * trying to get the type and mode via SFU can be slow, + * so just call those regular files for now, and mark + * for reval */ - tmp_inode->i_mode |= S_IFREG; - *pobject_type = DT_REG; - cifsInfo->time = 0; - } - } else { - if (attr & ATTR_DIRECTORY) { - tmp_inode->i_mode |= S_IFDIR; - *pobject_type = DT_DIR; - } else { - tmp_inode->i_mode |= S_IFREG; - *pobject_type = DT_REG; + fattr->cf_flags |= CIFS_FATTR_NEED_REVAL; } } +} - /* can not fill in nlink here as in qpathinfo version and Unx search */ - if (atomic_read(&cifsInfo->inUse) == 0) - atomic_set(&cifsInfo->inUse, 1); +void +cifs_dir_info_to_fattr(struct cifs_fattr *fattr, FILE_DIRECTORY_INFO *info, + struct cifs_sb_info *cifs_sb) +{ + memset(fattr, 0, sizeof(*fattr)); + fattr->cf_cifsattrs = le32_to_cpu(info->ExtFileAttributes); + fattr->cf_eof = le64_to_cpu(info->EndOfFile); + fattr->cf_bytes = le64_to_cpu(info->AllocationSize); + fattr->cf_atime = cifs_NTtimeToUnix(info->LastAccessTime); + fattr->cf_ctime = cifs_NTtimeToUnix(info->ChangeTime); + fattr->cf_mtime = cifs_NTtimeToUnix(info->LastWriteTime); + + cifs_fill_common_info(fattr, cifs_sb); +} - cifsInfo->server_eof = end_of_file; - spin_lock(&tmp_inode->i_lock); - if (is_size_safe_to_change(cifsInfo, end_of_file)) { - /* can not safely change the file size here if the - client is writing to it due to potential races */ - i_size_write(tmp_inode, end_of_file); +void +cifs_std_info_to_fattr(struct cifs_fattr *fattr, FIND_FILE_STANDARD_INFO *info, + struct cifs_sb_info *cifs_sb) +{ + int offset = cifs_sb->tcon->ses->server->timeAdj; - /* 512 bytes (2**9) is the fake blocksize that must be used */ - /* for this calculation, even though the reported blocksize is larger */ - tmp_inode->i_blocks = (512 - 1 + allocation_size) >> 9; - } - spin_unlock(&tmp_inode->i_lock); - - if (allocation_size < end_of_file) - cFYI(1, ("May be sparse file, allocation less than file size")); - cFYI(1, ("File Size %ld and blocks %llu", - (unsigned long)tmp_inode->i_size, - (unsigned long long)tmp_inode->i_blocks)); - if (S_ISREG(tmp_inode->i_mode)) { - cFYI(1, ("File inode")); - tmp_inode->i_op = &cifs_file_inode_ops; - if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_DIRECT_IO) { - if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NO_BRL) - tmp_inode->i_fop = &cifs_file_direct_nobrl_ops; - else - tmp_inode->i_fop = &cifs_file_direct_ops; - } else if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NO_BRL) - tmp_inode->i_fop = &cifs_file_nobrl_ops; - else - tmp_inode->i_fop = &cifs_file_ops; - - if ((cifs_sb->tcon) && (cifs_sb->tcon->ses) && - (cifs_sb->tcon->ses->server->maxBuf < - PAGE_CACHE_SIZE + MAX_CIFS_HDR_SIZE)) - tmp_inode->i_data.a_ops = &cifs_addr_ops_smallbuf; - else - tmp_inode->i_data.a_ops = &cifs_addr_ops; - - if (isNewInode) - return; /* No sense invalidating pages for new inode - since have not started caching readahead file - data yet */ - - if (timespec_equal(&tmp_inode->i_mtime, &local_mtime) && - (local_size == tmp_inode->i_size)) { - cFYI(1, ("inode exists but unchanged")); - } else { - /* file may have changed on server */ - cFYI(1, ("invalidate inode, readdir detected change")); - invalidate_remote_inode(tmp_inode); - } - } else if (S_ISDIR(tmp_inode->i_mode)) { - cFYI(1, ("Directory inode")); - tmp_inode->i_op = &cifs_dir_inode_ops; - tmp_inode->i_fop = &cifs_dir_ops; - } else if (S_ISLNK(tmp_inode->i_mode)) { - cFYI(1, ("Symbolic Link inode")); - tmp_inode->i_op = &cifs_symlink_inode_ops; - } else { - cFYI(1, ("Init special inode")); - init_special_inode(tmp_inode, tmp_inode->i_mode, - tmp_inode->i_rdev); - } + memset(fattr, 0, sizeof(*fattr)); + fattr->cf_atime = cnvrtDosUnixTm(info->LastAccessDate, + info->LastAccessTime, offset); + fattr->cf_ctime = cnvrtDosUnixTm(info->LastWriteDate, + info->LastWriteTime, offset); + fattr->cf_mtime = cnvrtDosUnixTm(info->LastWriteDate, + info->LastWriteTime, offset); + + fattr->cf_cifsattrs = le16_to_cpu(info->Attributes); + fattr->cf_bytes = le32_to_cpu(info->AllocationSize); + fattr->cf_eof = le32_to_cpu(info->DataSize); + + cifs_fill_common_info(fattr, cifs_sb); } /* BB eventually need to add the following helper function to @@ -846,11 +681,10 @@ static int cifs_filldir(char *pfindEntry, struct file *file, filldir_t filldir, int rc = 0; struct qstr qstring; struct cifsFileInfo *pCifsF; - unsigned int obj_type; - __u64 inum; + u64 inum; ino_t ino; + struct super_block *sb; struct cifs_sb_info *cifs_sb; - struct inode *tmp_inode; struct dentry *tmp_dentry; struct cifs_fattr fattr; @@ -870,71 +704,53 @@ static int cifs_filldir(char *pfindEntry, struct file *file, filldir_t filldir, if (rc != 0) return 0; - cifs_sb = CIFS_SB(file->f_path.dentry->d_sb); + sb = file->f_path.dentry->d_sb; + cifs_sb = CIFS_SB(sb); qstring.name = scratch_buf; rc = cifs_get_name_from_search_buf(&qstring, pfindEntry, pCifsF->srch_inf.info_level, pCifsF->srch_inf.unicode, cifs_sb, - max_len, - &inum /* returned */); + max_len, &inum /* returned */); if (rc) return rc; - /* only these two infolevels return valid inode numbers */ - if (pCifsF->srch_inf.info_level == SMB_FIND_FILE_UNIX) { + if (pCifsF->srch_inf.info_level == SMB_FIND_FILE_UNIX) cifs_unix_basic_to_fattr(&fattr, &((FILE_UNIX_INFO *) pfindEntry)->basic, cifs_sb); - tmp_dentry = cifs_readdir_lookup(file->f_dentry, &qstring, - &fattr); - obj_type = fattr.cf_dtype; - ino = cifs_uniqueid_to_ino_t(fattr.cf_uniqueid); - } else { - if (pCifsF->srch_inf.info_level == - SMB_FIND_FILE_ID_FULL_DIR_INFO) - rc = construct_dentry(&qstring, file, &tmp_inode, - &tmp_dentry, &inum); - else - rc = construct_dentry(&qstring, file, &tmp_inode, - &tmp_dentry, NULL); - - if ((tmp_inode == NULL) || (tmp_dentry == NULL)) { - rc = -ENOMEM; - goto out; - } - - /* we pass in rc below, indicating whether it is a new inode, - * so we can figure out whether to invalidate the inode cached - * data if the file has changed - */ - if (pCifsF->srch_inf.info_level == SMB_FIND_FILE_INFO_STANDARD) - fill_in_inode(tmp_inode, 0, pfindEntry, &obj_type, rc); - else - fill_in_inode(tmp_inode, 1, pfindEntry, &obj_type, rc); + else if (pCifsF->srch_inf.info_level == SMB_FIND_FILE_INFO_STANDARD) + cifs_std_info_to_fattr(&fattr, (FIND_FILE_STANDARD_INFO *) + pfindEntry, cifs_sb); + else + cifs_dir_info_to_fattr(&fattr, (FILE_DIRECTORY_INFO *) + pfindEntry, cifs_sb); - /* new inode - needs to be tied to dentry */ - if (rc) { - d_instantiate(tmp_dentry, tmp_inode); - if (rc == 2) - d_rehash(tmp_dentry); - } + /* FIXME: make _to_fattr functions fill this out */ + if (pCifsF->srch_inf.info_level == SMB_FIND_FILE_ID_FULL_DIR_INFO) + fattr.cf_uniqueid = inum; + else + fattr.cf_uniqueid = iunique(sb, ROOT_I); - ino = cifs_uniqueid_to_ino_t(tmp_inode->i_ino); - } + ino = cifs_uniqueid_to_ino_t(fattr.cf_uniqueid); + tmp_dentry = cifs_readdir_lookup(file->f_dentry, &qstring, &fattr); rc = filldir(direntry, qstring.name, qstring.len, file->f_pos, - ino, obj_type); + ino, fattr.cf_dtype); + + /* + * we can not return filldir errors to the caller since they are + * "normal" when the stat blocksize is too small - we return remapped + * error instead + * + * FIXME: This looks bogus. filldir returns -EOVERFLOW in the above + * case already. Why should we be clobbering other errors from it? + */ if (rc) { cFYI(1, ("filldir rc = %d", rc)); - /* we can not return filldir errors to the caller - since they are "normal" when the stat blocksize - is too small - we return remapped error instead */ rc = -EOVERFLOW; } - -out: dput(tmp_dentry); return rc; } -- cgit v1.2.3 From aeaaf253c4dee7ff9af2f3f0595f3bb66964e944 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Thu, 9 Jul 2009 01:46:39 -0400 Subject: cifs: remove cifsInodeInfo->inUse counter cifs: remove cifsInodeInfo->inUse counter It was purported to be a refcounter of some sort, but was never used that way. It never served any purpose that wasn't served equally well by the I_NEW flag. Signed-off-by: Jeff Layton Reviewed-by: Christoph Hellwig Signed-off-by: Steve French --- fs/cifs/cifsfs.c | 1 - fs/cifs/cifsglob.h | 1 - 2 files changed, 2 deletions(-) (limited to 'fs') diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c index 9f669f982c4d..44f30504b82d 100644 --- a/fs/cifs/cifsfs.c +++ b/fs/cifs/cifsfs.c @@ -308,7 +308,6 @@ cifs_alloc_inode(struct super_block *sb) if (!cifs_inode) return NULL; cifs_inode->cifsAttrs = 0x20; /* default */ - atomic_set(&cifs_inode->inUse, 0); cifs_inode->time = 0; cifs_inode->write_behind_rc = 0; /* Until the file is open and we have gotten oplock diff --git a/fs/cifs/cifsglob.h b/fs/cifs/cifsglob.h index 8bcf5a4bcded..63f6cdfa5638 100644 --- a/fs/cifs/cifsglob.h +++ b/fs/cifs/cifsglob.h @@ -364,7 +364,6 @@ struct cifsInodeInfo { struct list_head openFileList; int write_behind_rc; __u32 cifsAttrs; /* e.g. DOS archive bit, sparse, compressed, system */ - atomic_t inUse; /* num concurrent users (local openers cifs) of file*/ unsigned long time; /* jiffies of last update/check of inode */ bool clientCanCacheRead:1; /* read oplock */ bool clientCanCacheAll:1; /* read and writebehind oplock */ -- cgit v1.2.3 From d0c280d26de9422c9c943f8f486b9830cd9bea70 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Thu, 9 Jul 2009 01:46:44 -0400 Subject: cifs: when ATTR_READONLY is set, only clear write bits on non-directories cifs: when ATTR_READONLY is set, only clear write bits on non-directories On windows servers, ATTR_READONLY apparently either has no meaning or serves as some sort of queue to certain applications for unrelated behavior. This MS kbase article has details: http://support.microsoft.com/kb/326549/ Don't clear the write bits directory mode when ATTR_READONLY is set. Reported-by: pouchat@peewiki.net Signed-off-by: Jeff Layton Signed-off-by: Steve French --- fs/cifs/inode.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/cifs/inode.c b/fs/cifs/inode.c index a807397f444e..18afe57b2461 100644 --- a/fs/cifs/inode.c +++ b/fs/cifs/inode.c @@ -419,11 +419,11 @@ cifs_all_info_to_fattr(struct cifs_fattr *fattr, FILE_ALL_INFO *info, } else { fattr->cf_mode = S_IFREG | cifs_sb->mnt_file_mode; fattr->cf_dtype = DT_REG; - } - /* clear write bits if ATTR_READONLY is set */ - if (fattr->cf_cifsattrs & ATTR_READONLY) - fattr->cf_mode &= ~(S_IWUGO); + /* clear write bits if ATTR_READONLY is set */ + if (fattr->cf_cifsattrs & ATTR_READONLY) + fattr->cf_mode &= ~(S_IWUGO); + } fattr->cf_nlink = le32_to_cpu(info->NumberOfLinks); -- cgit v1.2.3 From 65bc98b0059360e458aebd208587be44641227c1 Mon Sep 17 00:00:00 2001 From: Steve French Date: Fri, 10 Jul 2009 15:27:25 +0000 Subject: [CIFS] Distinguish posix opens and mkdirs from legacy mkdirs in stats Acked-by: Jeff Layton Signed-off-by: Steve French --- fs/cifs/cifs_debug.c | 8 +++++++- fs/cifs/cifsglob.h | 2 ++ fs/cifs/cifssmb.c | 5 ++++- 3 files changed, 13 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/cifs/cifs_debug.c b/fs/cifs/cifs_debug.c index 7f19fefd3d45..42cec2a7c0cf 100644 --- a/fs/cifs/cifs_debug.c +++ b/fs/cifs/cifs_debug.c @@ -261,6 +261,8 @@ static ssize_t cifs_stats_proc_write(struct file *file, atomic_set(&tcon->num_reads, 0); atomic_set(&tcon->num_oplock_brks, 0); atomic_set(&tcon->num_opens, 0); + atomic_set(&tcon->num_posixopens, 0); + atomic_set(&tcon->num_posixmkdirs, 0); atomic_set(&tcon->num_closes, 0); atomic_set(&tcon->num_deletes, 0); atomic_set(&tcon->num_mkdirs, 0); @@ -347,11 +349,15 @@ static int cifs_stats_proc_show(struct seq_file *m, void *v) atomic_read(&tcon->num_locks), atomic_read(&tcon->num_hardlinks), atomic_read(&tcon->num_symlinks)); - seq_printf(m, "\nOpens: %d Closes: %d" + seq_printf(m, "\nOpens: %d Closes: %d " "Deletes: %d", atomic_read(&tcon->num_opens), atomic_read(&tcon->num_closes), atomic_read(&tcon->num_deletes)); + seq_printf(m, "\nPosix Opens: %d " + "Posix Mkdirs: %d", + atomic_read(&tcon->num_posixopens), + atomic_read(&tcon->num_posixmkdirs)); seq_printf(m, "\nMkdirs: %d Rmdirs: %d", atomic_read(&tcon->num_mkdirs), atomic_read(&tcon->num_rmdirs)); diff --git a/fs/cifs/cifsglob.h b/fs/cifs/cifsglob.h index 63f6cdfa5638..6084d6379c03 100644 --- a/fs/cifs/cifsglob.h +++ b/fs/cifs/cifsglob.h @@ -260,6 +260,8 @@ struct cifsTconInfo { atomic_t num_closes; atomic_t num_deletes; atomic_t num_mkdirs; + atomic_t num_posixopens; + atomic_t num_posixmkdirs; atomic_t num_rmdirs; atomic_t num_renames; atomic_t num_t2renames; diff --git a/fs/cifs/cifssmb.c b/fs/cifs/cifssmb.c index 922f5fe2084c..1866bc2927d4 100644 --- a/fs/cifs/cifssmb.c +++ b/fs/cifs/cifssmb.c @@ -1113,7 +1113,10 @@ PsxCreat: psx_create_err: cifs_buf_release(pSMB); - cifs_stats_inc(&tcon->num_mkdirs); + if (posix_flags & SMB_O_DIRECTORY) + cifs_stats_inc(&tcon->num_posixmkdirs); + else + cifs_stats_inc(&tcon->num_posixopens); if (rc == -EAGAIN) goto PsxCreat; -- cgit v1.2.3 From 8aa7e847d834ed937a9ad37a0f2ad5b8584c1ab0 Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Thu, 9 Jul 2009 14:52:32 +0200 Subject: Fix congestion_wait() sync/async vs read/write confusion Commit 1faa16d22877f4839bd433547d770c676d1d964c accidentally broke the bdi congestion wait queue logic, causing us to wait on congestion for WRITE (== 1) when we really wanted BLK_RW_ASYNC (== 0) instead. Signed-off-by: Jens Axboe --- arch/x86/lib/usercopy_32.c | 2 +- drivers/block/pktcdvd.c | 10 ++++++---- drivers/md/dm-crypt.c | 2 +- fs/fat/file.c | 2 +- fs/fuse/dev.c | 8 ++++---- fs/nfs/write.c | 8 +++++--- fs/reiserfs/journal.c | 2 +- fs/xfs/linux-2.6/kmem.c | 4 ++-- fs/xfs/linux-2.6/xfs_buf.c | 2 +- include/linux/backing-dev.h | 6 +++--- include/linux/blkdev.h | 8 ++++---- mm/backing-dev.c | 7 +++---- mm/memcontrol.c | 2 +- mm/page-writeback.c | 8 ++++---- mm/page_alloc.c | 4 ++-- mm/vmscan.c | 8 ++++---- 16 files changed, 43 insertions(+), 40 deletions(-) (limited to 'fs') diff --git a/arch/x86/lib/usercopy_32.c b/arch/x86/lib/usercopy_32.c index 7c8ca91bb9ec..1f118d462acc 100644 --- a/arch/x86/lib/usercopy_32.c +++ b/arch/x86/lib/usercopy_32.c @@ -751,7 +751,7 @@ survive: if (retval == -ENOMEM && is_global_init(current)) { up_read(¤t->mm->mmap_sem); - congestion_wait(WRITE, HZ/50); + congestion_wait(BLK_RW_ASYNC, HZ/50); goto survive; } diff --git a/drivers/block/pktcdvd.c b/drivers/block/pktcdvd.c index 83650e00632d..99a506f619b7 100644 --- a/drivers/block/pktcdvd.c +++ b/drivers/block/pktcdvd.c @@ -1372,8 +1372,10 @@ try_next_bio: wakeup = (pd->write_congestion_on > 0 && pd->bio_queue_size <= pd->write_congestion_off); spin_unlock(&pd->lock); - if (wakeup) - clear_bdi_congested(&pd->disk->queue->backing_dev_info, WRITE); + if (wakeup) { + clear_bdi_congested(&pd->disk->queue->backing_dev_info, + BLK_RW_ASYNC); + } pkt->sleep_time = max(PACKET_WAIT_TIME, 1); pkt_set_state(pkt, PACKET_WAITING_STATE); @@ -2592,10 +2594,10 @@ static int pkt_make_request(struct request_queue *q, struct bio *bio) spin_lock(&pd->lock); if (pd->write_congestion_on > 0 && pd->bio_queue_size >= pd->write_congestion_on) { - set_bdi_congested(&q->backing_dev_info, WRITE); + set_bdi_congested(&q->backing_dev_info, BLK_RW_ASYNC); do { spin_unlock(&pd->lock); - congestion_wait(WRITE, HZ); + congestion_wait(BLK_RW_ASYNC, HZ); spin_lock(&pd->lock); } while(pd->bio_queue_size > pd->write_congestion_off); } diff --git a/drivers/md/dm-crypt.c b/drivers/md/dm-crypt.c index 9933eb861c71..529e2ba505c3 100644 --- a/drivers/md/dm-crypt.c +++ b/drivers/md/dm-crypt.c @@ -776,7 +776,7 @@ static void kcryptd_crypt_write_convert(struct dm_crypt_io *io) * But don't wait if split was due to the io size restriction */ if (unlikely(out_of_pages)) - congestion_wait(WRITE, HZ/100); + congestion_wait(BLK_RW_ASYNC, HZ/100); /* * With async crypto it is unsafe to share the crypto context diff --git a/fs/fat/file.c b/fs/fat/file.c index b28ea646ff60..f042b965c95c 100644 --- a/fs/fat/file.c +++ b/fs/fat/file.c @@ -134,7 +134,7 @@ static int fat_file_release(struct inode *inode, struct file *filp) if ((filp->f_mode & FMODE_WRITE) && MSDOS_SB(inode->i_sb)->options.flush) { fat_flush_inodes(inode->i_sb, inode, NULL); - congestion_wait(WRITE, HZ/10); + congestion_wait(BLK_RW_ASYNC, HZ/10); } return 0; } diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c index f58ecbc416c8..6484eb75acd6 100644 --- a/fs/fuse/dev.c +++ b/fs/fuse/dev.c @@ -286,8 +286,8 @@ __releases(&fc->lock) } if (fc->num_background == FUSE_CONGESTION_THRESHOLD && fc->connected && fc->bdi_initialized) { - clear_bdi_congested(&fc->bdi, READ); - clear_bdi_congested(&fc->bdi, WRITE); + clear_bdi_congested(&fc->bdi, BLK_RW_SYNC); + clear_bdi_congested(&fc->bdi, BLK_RW_ASYNC); } fc->num_background--; fc->active_background--; @@ -414,8 +414,8 @@ static void fuse_request_send_nowait_locked(struct fuse_conn *fc, fc->blocked = 1; if (fc->num_background == FUSE_CONGESTION_THRESHOLD && fc->bdi_initialized) { - set_bdi_congested(&fc->bdi, READ); - set_bdi_congested(&fc->bdi, WRITE); + set_bdi_congested(&fc->bdi, BLK_RW_SYNC); + set_bdi_congested(&fc->bdi, BLK_RW_ASYNC); } list_add_tail(&req->list, &fc->bg_queue); flush_bg_queue(fc); diff --git a/fs/nfs/write.c b/fs/nfs/write.c index ce728829f79a..0a0a2ff767c3 100644 --- a/fs/nfs/write.c +++ b/fs/nfs/write.c @@ -202,8 +202,10 @@ static int nfs_set_page_writeback(struct page *page) struct nfs_server *nfss = NFS_SERVER(inode); if (atomic_long_inc_return(&nfss->writeback) > - NFS_CONGESTION_ON_THRESH) - set_bdi_congested(&nfss->backing_dev_info, WRITE); + NFS_CONGESTION_ON_THRESH) { + set_bdi_congested(&nfss->backing_dev_info, + BLK_RW_ASYNC); + } } return ret; } @@ -215,7 +217,7 @@ static void nfs_end_page_writeback(struct page *page) end_page_writeback(page); if (atomic_long_dec_return(&nfss->writeback) < NFS_CONGESTION_OFF_THRESH) - clear_bdi_congested(&nfss->backing_dev_info, WRITE); + clear_bdi_congested(&nfss->backing_dev_info, BLK_RW_ASYNC); } /* diff --git a/fs/reiserfs/journal.c b/fs/reiserfs/journal.c index 77f5bb746bf0..90622200b39c 100644 --- a/fs/reiserfs/journal.c +++ b/fs/reiserfs/journal.c @@ -997,7 +997,7 @@ static int reiserfs_async_progress_wait(struct super_block *s) DEFINE_WAIT(wait); struct reiserfs_journal *j = SB_JOURNAL(s); if (atomic_read(&j->j_async_throttle)) - congestion_wait(WRITE, HZ / 10); + congestion_wait(BLK_RW_ASYNC, HZ / 10); return 0; } diff --git a/fs/xfs/linux-2.6/kmem.c b/fs/xfs/linux-2.6/kmem.c index 1cd3b55ee3d2..2d3f90afe5f1 100644 --- a/fs/xfs/linux-2.6/kmem.c +++ b/fs/xfs/linux-2.6/kmem.c @@ -53,7 +53,7 @@ kmem_alloc(size_t size, unsigned int __nocast flags) printk(KERN_ERR "XFS: possible memory allocation " "deadlock in %s (mode:0x%x)\n", __func__, lflags); - congestion_wait(WRITE, HZ/50); + congestion_wait(BLK_RW_ASYNC, HZ/50); } while (1); } @@ -130,7 +130,7 @@ kmem_zone_alloc(kmem_zone_t *zone, unsigned int __nocast flags) printk(KERN_ERR "XFS: possible memory allocation " "deadlock in %s (mode:0x%x)\n", __func__, lflags); - congestion_wait(WRITE, HZ/50); + congestion_wait(BLK_RW_ASYNC, HZ/50); } while (1); } diff --git a/fs/xfs/linux-2.6/xfs_buf.c b/fs/xfs/linux-2.6/xfs_buf.c index 1418b916fc27..0c93c7ef3d18 100644 --- a/fs/xfs/linux-2.6/xfs_buf.c +++ b/fs/xfs/linux-2.6/xfs_buf.c @@ -412,7 +412,7 @@ _xfs_buf_lookup_pages( XFS_STATS_INC(xb_page_retries); xfsbufd_wakeup(0, gfp_mask); - congestion_wait(WRITE, HZ/50); + congestion_wait(BLK_RW_ASYNC, HZ/50); goto retry; } diff --git a/include/linux/backing-dev.h b/include/linux/backing-dev.h index 0ec2c594868e..3a52a63c1351 100644 --- a/include/linux/backing-dev.h +++ b/include/linux/backing-dev.h @@ -229,9 +229,9 @@ static inline int bdi_rw_congested(struct backing_dev_info *bdi) (1 << BDI_async_congested)); } -void clear_bdi_congested(struct backing_dev_info *bdi, int rw); -void set_bdi_congested(struct backing_dev_info *bdi, int rw); -long congestion_wait(int rw, long timeout); +void clear_bdi_congested(struct backing_dev_info *bdi, int sync); +void set_bdi_congested(struct backing_dev_info *bdi, int sync); +long congestion_wait(int sync, long timeout); static inline bool bdi_cap_writeback_dirty(struct backing_dev_info *bdi) diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index 49ae07951d55..bb3d39978701 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -779,18 +779,18 @@ extern int sg_scsi_ioctl(struct request_queue *, struct gendisk *, fmode_t, * congested queues, and wake up anyone who was waiting for requests to be * put back. */ -static inline void blk_clear_queue_congested(struct request_queue *q, int rw) +static inline void blk_clear_queue_congested(struct request_queue *q, int sync) { - clear_bdi_congested(&q->backing_dev_info, rw); + clear_bdi_congested(&q->backing_dev_info, sync); } /* * A queue has just entered congestion. Flag that in the queue's VM-visible * state flags and increment the global gounter of congested queues. */ -static inline void blk_set_queue_congested(struct request_queue *q, int rw) +static inline void blk_set_queue_congested(struct request_queue *q, int sync) { - set_bdi_congested(&q->backing_dev_info, rw); + set_bdi_congested(&q->backing_dev_info, sync); } extern void blk_start_queue(struct request_queue *q); diff --git a/mm/backing-dev.c b/mm/backing-dev.c index 493b468a5035..c86edd244294 100644 --- a/mm/backing-dev.c +++ b/mm/backing-dev.c @@ -283,7 +283,6 @@ static wait_queue_head_t congestion_wqh[2] = { __WAIT_QUEUE_HEAD_INITIALIZER(congestion_wqh[1]) }; - void clear_bdi_congested(struct backing_dev_info *bdi, int sync) { enum bdi_state bit; @@ -308,18 +307,18 @@ EXPORT_SYMBOL(set_bdi_congested); /** * congestion_wait - wait for a backing_dev to become uncongested - * @rw: READ or WRITE + * @sync: SYNC or ASYNC IO * @timeout: timeout in jiffies * * Waits for up to @timeout jiffies for a backing_dev (any backing_dev) to exit * write congestion. If no backing_devs are congested then just wait for the * next write to be completed. */ -long congestion_wait(int rw, long timeout) +long congestion_wait(int sync, long timeout) { long ret; DEFINE_WAIT(wait); - wait_queue_head_t *wqh = &congestion_wqh[rw]; + wait_queue_head_t *wqh = &congestion_wqh[sync]; prepare_to_wait(wqh, &wait, TASK_UNINTERRUPTIBLE); ret = io_schedule_timeout(timeout); diff --git a/mm/memcontrol.c b/mm/memcontrol.c index e2fa20dadf40..e717964cb5a0 100644 --- a/mm/memcontrol.c +++ b/mm/memcontrol.c @@ -1973,7 +1973,7 @@ try_to_free: if (!progress) { nr_retries--; /* maybe some writeback is necessary */ - congestion_wait(WRITE, HZ/10); + congestion_wait(BLK_RW_ASYNC, HZ/10); } } diff --git a/mm/page-writeback.c b/mm/page-writeback.c index 7687879253b9..81627ebcd313 100644 --- a/mm/page-writeback.c +++ b/mm/page-writeback.c @@ -575,7 +575,7 @@ static void balance_dirty_pages(struct address_space *mapping) if (pages_written >= write_chunk) break; /* We've done our duty */ - congestion_wait(WRITE, HZ/10); + congestion_wait(BLK_RW_ASYNC, HZ/10); } if (bdi_nr_reclaimable + bdi_nr_writeback < bdi_thresh && @@ -669,7 +669,7 @@ void throttle_vm_writeout(gfp_t gfp_mask) if (global_page_state(NR_UNSTABLE_NFS) + global_page_state(NR_WRITEBACK) <= dirty_thresh) break; - congestion_wait(WRITE, HZ/10); + congestion_wait(BLK_RW_ASYNC, HZ/10); /* * The caller might hold locks which can prevent IO completion @@ -715,7 +715,7 @@ static void background_writeout(unsigned long _min_pages) if (wbc.nr_to_write > 0 || wbc.pages_skipped > 0) { /* Wrote less than expected */ if (wbc.encountered_congestion || wbc.more_io) - congestion_wait(WRITE, HZ/10); + congestion_wait(BLK_RW_ASYNC, HZ/10); else break; } @@ -787,7 +787,7 @@ static void wb_kupdate(unsigned long arg) writeback_inodes(&wbc); if (wbc.nr_to_write > 0) { if (wbc.encountered_congestion || wbc.more_io) - congestion_wait(WRITE, HZ/10); + congestion_wait(BLK_RW_ASYNC, HZ/10); else break; /* All the old data is written */ } diff --git a/mm/page_alloc.c b/mm/page_alloc.c index ad7cd1c56b07..a35eeab2724c 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -1666,7 +1666,7 @@ __alloc_pages_high_priority(gfp_t gfp_mask, unsigned int order, preferred_zone, migratetype); if (!page && gfp_mask & __GFP_NOFAIL) - congestion_wait(WRITE, HZ/50); + congestion_wait(BLK_RW_ASYNC, HZ/50); } while (!page && (gfp_mask & __GFP_NOFAIL)); return page; @@ -1831,7 +1831,7 @@ rebalance: pages_reclaimed += did_some_progress; if (should_alloc_retry(gfp_mask, order, pages_reclaimed)) { /* Wait for some write requests to complete then retry */ - congestion_wait(WRITE, HZ/50); + congestion_wait(BLK_RW_ASYNC, HZ/50); goto rebalance; } diff --git a/mm/vmscan.c b/mm/vmscan.c index 54155268dfca..dea7abd31098 100644 --- a/mm/vmscan.c +++ b/mm/vmscan.c @@ -1104,7 +1104,7 @@ static unsigned long shrink_inactive_list(unsigned long max_scan, */ if (nr_freed < nr_taken && !current_is_kswapd() && lumpy_reclaim) { - congestion_wait(WRITE, HZ/10); + congestion_wait(BLK_RW_ASYNC, HZ/10); /* * The attempt at page out may have made some @@ -1721,7 +1721,7 @@ static unsigned long do_try_to_free_pages(struct zonelist *zonelist, /* Take a nap, wait for some writeback to complete */ if (sc->nr_scanned && priority < DEF_PRIORITY - 2) - congestion_wait(WRITE, HZ/10); + congestion_wait(BLK_RW_ASYNC, HZ/10); } /* top priority shrink_zones still had more to do? don't OOM, then */ if (!sc->all_unreclaimable && scanning_global_lru(sc)) @@ -1960,7 +1960,7 @@ loop_again: * another pass across the zones. */ if (total_scanned && priority < DEF_PRIORITY - 2) - congestion_wait(WRITE, HZ/10); + congestion_wait(BLK_RW_ASYNC, HZ/10); /* * We do this so kswapd doesn't build up large priorities for @@ -2233,7 +2233,7 @@ unsigned long shrink_all_memory(unsigned long nr_pages) goto out; if (sc.nr_scanned && prio < DEF_PRIORITY - 2) - congestion_wait(WRITE, HZ / 10); + congestion_wait(BLK_RW_ASYNC, HZ / 10); } } -- cgit v1.2.3 From ecb554a846f8e9d2a58f6d6c118168a63ac065aa Mon Sep 17 00:00:00 2001 From: FUJITA Tomonori Date: Thu, 9 Jul 2009 14:46:53 +0200 Subject: block: fix sg SG_DXFER_TO_FROM_DEV regression I overlooked SG_DXFER_TO_FROM_DEV support when I converted sg to use the block layer mapping API (2.6.28). Douglas Gilbert explained SG_DXFER_TO_FROM_DEV: http://www.spinics.net/lists/linux-scsi/msg37135.html = The semantics of SG_DXFER_TO_FROM_DEV were: - copy user space buffer to kernel (LLD) buffer - do SCSI command which is assumed to be of the DATA_IN (data from device) variety. This would overwrite some or all of the kernel buffer - copy kernel (LLD) buffer back to the user space. The idea was to detect short reads by filling the original user space buffer with some marker bytes ("0xec" it would seem in this report). The "resid" value is a better way of detecting short reads but that was only added this century and requires co-operation from the LLD. = This patch changes the block layer mapping API to support this semantics. This simply adds another field to struct rq_map_data and enables __bio_copy_iov() to copy data from user space even with READ requests. It's better to add the flags field and kills null_mapped and the new from_user fields in struct rq_map_data but that approach makes it difficult to send this patch to stable trees because st and osst drivers use struct rq_map_data (they were converted to use the block layer in 2.6.29 and 2.6.30). Well, I should clean up the block layer mapping API. zhou sf reported this regiression and tested this patch: http://www.spinics.net/lists/linux-scsi/msg37128.html http://www.spinics.net/lists/linux-scsi/msg37168.html Reported-by: zhou sf Tested-by: zhou sf Cc: stable@kernel.org Signed-off-by: FUJITA Tomonori Signed-off-by: Jens Axboe --- drivers/scsi/sg.c | 4 ++++ fs/bio.c | 22 ++++++++++++---------- include/linux/blkdev.h | 1 + 3 files changed, 17 insertions(+), 10 deletions(-) (limited to 'fs') diff --git a/drivers/scsi/sg.c b/drivers/scsi/sg.c index 4d6f2fe1cfe9..9230402c45af 100644 --- a/drivers/scsi/sg.c +++ b/drivers/scsi/sg.c @@ -1656,6 +1656,10 @@ static int sg_start_req(Sg_request *srp, unsigned char *cmd) md->nr_entries = req_schp->k_use_sg; md->offset = 0; md->null_mapped = hp->dxferp ? 0 : 1; + if (dxfer_dir == SG_DXFER_TO_FROM_DEV) + md->from_user = 1; + else + md->from_user = 0; } if (iov_count) { diff --git a/fs/bio.c b/fs/bio.c index 1486b19fc431..76738005c8e8 100644 --- a/fs/bio.c +++ b/fs/bio.c @@ -705,14 +705,13 @@ static struct bio_map_data *bio_alloc_map_data(int nr_segs, int iov_count, } static int __bio_copy_iov(struct bio *bio, struct bio_vec *iovecs, - struct sg_iovec *iov, int iov_count, int uncopy, - int do_free_page) + struct sg_iovec *iov, int iov_count, + int to_user, int from_user, int do_free_page) { int ret = 0, i; struct bio_vec *bvec; int iov_idx = 0; unsigned int iov_off = 0; - int read = bio_data_dir(bio) == READ; __bio_for_each_segment(bvec, bio, i, 0) { char *bv_addr = page_address(bvec->bv_page); @@ -727,13 +726,14 @@ static int __bio_copy_iov(struct bio *bio, struct bio_vec *iovecs, iov_addr = iov[iov_idx].iov_base + iov_off; if (!ret) { - if (!read && !uncopy) - ret = copy_from_user(bv_addr, iov_addr, - bytes); - if (read && uncopy) + if (to_user) ret = copy_to_user(iov_addr, bv_addr, bytes); + if (from_user) + ret = copy_from_user(bv_addr, iov_addr, + bytes); + if (ret) ret = -EFAULT; } @@ -770,7 +770,8 @@ int bio_uncopy_user(struct bio *bio) if (!bio_flagged(bio, BIO_NULL_MAPPED)) ret = __bio_copy_iov(bio, bmd->iovecs, bmd->sgvecs, - bmd->nr_sgvecs, 1, bmd->is_our_pages); + bmd->nr_sgvecs, bio_data_dir(bio) == READ, + 0, bmd->is_our_pages); bio_free_map_data(bmd); bio_put(bio); return ret; @@ -875,8 +876,9 @@ struct bio *bio_copy_user_iov(struct request_queue *q, /* * success */ - if (!write_to_vm && (!map_data || !map_data->null_mapped)) { - ret = __bio_copy_iov(bio, bio->bi_io_vec, iov, iov_count, 0, 0); + if ((!write_to_vm && (!map_data || !map_data->null_mapped)) || + (map_data && map_data->from_user)) { + ret = __bio_copy_iov(bio, bio->bi_io_vec, iov, iov_count, 0, 1, 0); if (ret) goto cleanup; } diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index bb3d39978701..0146e0fecf1a 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -723,6 +723,7 @@ struct rq_map_data { int nr_entries; unsigned long offset; int null_mapped; + int from_user; }; struct req_iterator { -- cgit v1.2.3 From 812e7a6a43fc34bc8f70c2b80db4ea5997d66ea8 Mon Sep 17 00:00:00 2001 From: Wengang Wang Date: Fri, 10 Jul 2009 13:26:04 +0800 Subject: ocfs2: log the actual return value of ocfs2_file_aio_write() in ocfs2_file_aio_write(), log_exit() could don't log the value which is really returned. this patch fixes it. Signed-off-by: Wengang Wang Signed-off-by: Joel Becker --- fs/ocfs2/file.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c index 62442e413a00..a49fa44aea1f 100644 --- a/fs/ocfs2/file.c +++ b/fs/ocfs2/file.c @@ -1918,8 +1918,10 @@ out_sems: mutex_unlock(&inode->i_mutex); + if (written) + ret = written; mlog_exit(ret); - return written ? written : ret; + return ret; } static int ocfs2_splice_to_file(struct pipe_inode_info *pipe, -- cgit v1.2.3 From 097041e576ee3a50d92dd643ee8ca65bf6a62e21 Mon Sep 17 00:00:00 2001 From: Larry Finger Date: Fri, 10 Jul 2009 20:06:42 -0500 Subject: fuse: Fix build error MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit When building v2.6.31-rc2-344-g69ca06c, the following build errors are found due to missing includes: CC [M] fs/fuse/dev.o fs/fuse/dev.c: In function ‘request_end’: fs/fuse/dev.c:289: error: ‘BLK_RW_SYNC’ undeclared (first use in this function) ... fs/nfs/write.c: In function ‘nfs_set_page_writeback’: fs/nfs/write.c:207: error: ‘BLK_RW_ASYNC’ undeclared (first use in this function) Signed-off-by: Larry Finger@lwfinger.net> Signed-off-by: Linus Torvalds --- fs/fuse/dev.c | 1 + fs/nfs/write.c | 1 + 2 files changed, 2 insertions(+) (limited to 'fs') diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c index 6484eb75acd6..cbceacbc0bf9 100644 --- a/fs/fuse/dev.c +++ b/fs/fuse/dev.c @@ -16,6 +16,7 @@ #include #include #include +#include MODULE_ALIAS_MISCDEV(FUSE_MINOR); diff --git a/fs/nfs/write.c b/fs/nfs/write.c index 0a0a2ff767c3..35d81316163f 100644 --- a/fs/nfs/write.c +++ b/fs/nfs/write.c @@ -19,6 +19,7 @@ #include #include #include +#include #include -- cgit v1.2.3 From 8711c67bee675b4f7a378c71ad5a59c981ec3df0 Mon Sep 17 00:00:00 2001 From: Bartlomiej Zolnierkiewicz Date: Fri, 10 Jul 2009 12:34:27 +0200 Subject: isofs: fix Joliet regression commit 5404ac8e4418ab3d254950ee4f9bcafc1da20b4a ("isofs: cleanup mount option processing") missed conversion of joliet option flag resulting in non-working Joliet support. CC: walt Signed-off-by: Bartlomiej Zolnierkiewicz Signed-off-by: Jan Kara Signed-off-by: Linus Torvalds --- fs/isofs/inode.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/isofs/inode.c b/fs/isofs/inode.c index 58a7963e168a..85f96bc651c7 100644 --- a/fs/isofs/inode.c +++ b/fs/isofs/inode.c @@ -142,6 +142,7 @@ static const struct dentry_operations isofs_dentry_ops[] = { struct iso9660_options{ unsigned int rock:1; + unsigned int joliet:1; unsigned int cruft:1; unsigned int hide:1; unsigned int showassoc:1; @@ -151,7 +152,6 @@ struct iso9660_options{ unsigned int gid_set:1; unsigned int utf8:1; unsigned char map; - char joliet; unsigned char check; unsigned int blocksize; mode_t fmode; @@ -632,7 +632,7 @@ static int isofs_fill_super(struct super_block *s, void *data, int silent) else if (isonum_711(vdp->type) == ISO_VD_SUPPLEMENTARY) { sec = (struct iso_supplementary_descriptor *)vdp; if (sec->escape[0] == 0x25 && sec->escape[1] == 0x2f) { - if (opt.joliet == 'y') { + if (opt.joliet) { if (sec->escape[2] == 0x40) joliet_level = 1; else if (sec->escape[2] == 0x43) -- cgit v1.2.3 From 81e4e1ba7ed4a1fdcf0e2ee944f1575010471464 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Sat, 11 Jul 2009 11:22:34 -0700 Subject: Revert "fuse: Fix build error" as unnecessary This reverts commit 097041e576ee3a50d92dd643ee8ca65bf6a62e21. Trond had a better fix, which is the parent of this one ("Fix compile error due to congestion_wait() changes") Requested-by: Trond Myklebust Acked-by: Larry Finger Signed-off-by: Linus Torvalds --- fs/fuse/dev.c | 1 - fs/nfs/write.c | 1 - 2 files changed, 2 deletions(-) (limited to 'fs') diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c index cbceacbc0bf9..6484eb75acd6 100644 --- a/fs/fuse/dev.c +++ b/fs/fuse/dev.c @@ -16,7 +16,6 @@ #include #include #include -#include MODULE_ALIAS_MISCDEV(FUSE_MINOR); diff --git a/fs/nfs/write.c b/fs/nfs/write.c index 35d81316163f..0a0a2ff767c3 100644 --- a/fs/nfs/write.c +++ b/fs/nfs/write.c @@ -19,7 +19,6 @@ #include #include #include -#include #include -- cgit v1.2.3 From 405f55712dfe464b3240d7816cc4fe4174831be2 Mon Sep 17 00:00:00 2001 From: Alexey Dobriyan Date: Sat, 11 Jul 2009 22:08:37 +0400 Subject: headers: smp_lock.h redux * Remove smp_lock.h from files which don't need it (including some headers!) * Add smp_lock.h to files which do need it * Make smp_lock.h include conditional in hardirq.h It's needed only for one kernel_locked() usage which is under CONFIG_PREEMPT This will make hardirq.h inclusion cheaper for every PREEMPT=n config (which includes allmodconfig/allyesconfig, BTW) Signed-off-by: Alexey Dobriyan Signed-off-by: Linus Torvalds --- arch/alpha/kernel/ptrace.c | 1 - arch/blackfin/kernel/ptrace.c | 1 - arch/blackfin/kernel/sys_bfin.c | 1 - arch/cris/kernel/sys_cris.c | 1 - arch/ia64/kernel/ptrace.c | 1 - arch/m32r/kernel/ptrace.c | 1 - arch/microblaze/kernel/ptrace.c | 1 - arch/microblaze/kernel/signal.c | 1 - arch/microblaze/kernel/sys_microblaze.c | 1 - arch/mips/kernel/ptrace32.c | 1 - arch/mips/mm/hugetlbpage.c | 1 - arch/mn10300/kernel/ptrace.c | 1 - arch/mn10300/kernel/signal.c | 1 - arch/mn10300/kernel/sys_mn10300.c | 1 - arch/mn10300/kernel/traps.c | 1 - arch/mn10300/mm/fault.c | 1 - arch/mn10300/mm/misalignment.c | 1 - arch/powerpc/kernel/ptrace32.c | 1 - arch/s390/kernel/dis.c | 1 - arch/s390/kernel/ptrace.c | 1 - arch/s390/mm/fault.c | 1 - arch/sh/mm/tlb-sh3.c | 1 - arch/sparc/kernel/ptrace_32.c | 1 - arch/sparc/kernel/ptrace_64.c | 1 - arch/sparc/kernel/time_64.c | 1 - arch/sparc/kernel/traps_32.c | 1 - drivers/block/DAC960.c | 1 + drivers/block/cciss.c | 1 + drivers/block/loop.c | 1 - drivers/bluetooth/hci_vhci.c | 1 - drivers/char/amiserial.c | 1 + drivers/char/cyclades.c | 1 + drivers/char/epca.c | 1 + drivers/char/isicom.c | 1 + drivers/char/istallion.c | 1 + drivers/char/moxa.c | 1 + drivers/char/mxser.c | 1 + drivers/char/n_hdlc.c | 1 + drivers/char/n_r3964.c | 1 + drivers/char/pty.c | 1 + drivers/char/rio/rio_linux.c | 1 + drivers/char/riscom8.c | 1 + drivers/char/rocket.c | 1 + drivers/char/serial167.c | 1 + drivers/char/specialix.c | 1 + drivers/char/sx.c | 1 + drivers/char/synclink.c | 1 + drivers/char/synclink_gt.c | 1 + drivers/char/synclinkmp.c | 1 + drivers/char/tpm/tpm.c | 1 - drivers/char/tty_ioctl.c | 1 - drivers/char/tty_ldisc.c | 1 - drivers/char/vt.c | 1 + drivers/char/vt_ioctl.c | 1 + drivers/gpio/vr41xx_giu.c | 1 - drivers/hid/usbhid/hid-core.c | 1 - drivers/isdn/hisax/hfc_usb.c | 1 - drivers/isdn/i4l/isdn_tty.c | 1 + drivers/isdn/mISDN/stack.c | 1 + drivers/media/dvb/bt8xx/dst_ca.c | 1 + drivers/media/dvb/dvb-core/dvbdev.h | 1 - drivers/media/dvb/ttpci/av7110.c | 1 - drivers/media/radio/radio-mr800.c | 1 + drivers/media/radio/radio-si470x.c | 1 + drivers/media/video/bt8xx/bttv-driver.c | 1 + drivers/media/video/cx23885/cx23885-417.c | 1 + drivers/media/video/cx23885/cx23885-video.c | 1 + drivers/media/video/cx88/cx88-blackbird.c | 1 + drivers/media/video/cx88/cx88-video.c | 1 + drivers/media/video/dabusb.c | 1 + drivers/media/video/pwc/pwc-if.c | 1 + drivers/media/video/pwc/pwc.h | 1 - drivers/media/video/s2255drv.c | 1 + drivers/media/video/saa5246a.c | 1 - drivers/media/video/saa5249.c | 1 - drivers/media/video/saa7134/saa7134-empress.c | 1 + drivers/media/video/se401.c | 1 + drivers/media/video/stk-webcam.c | 1 + drivers/media/video/stradis.c | 1 + drivers/media/video/stv680.c | 1 + drivers/media/video/usbvideo/vicam.c | 1 + drivers/media/video/usbvision/usbvision-video.c | 1 + drivers/media/video/v4l2-dev.c | 1 - drivers/media/video/zoran/zoran_driver.c | 1 + drivers/misc/sgi-gru/grufile.c | 1 - drivers/misc/sgi-gru/grukservices.c | 1 - drivers/net/irda/irtty-sir.c | 1 - drivers/pci/hotplug/cpci_hotplug_core.c | 1 - drivers/pci/hotplug/cpqphp_ctrl.c | 1 - drivers/pci/hotplug/cpqphp_sysfs.c | 1 + drivers/pci/hotplug/pciehp_ctrl.c | 1 - drivers/pci/syscall.c | 1 - drivers/s390/block/dasd_ioctl.c | 1 + drivers/scsi/qla2xxx/qla_mid.c | 1 - drivers/telephony/ixj.c | 1 + drivers/telephony/phonedev.c | 1 - drivers/usb/class/cdc-wdm.c | 1 - drivers/usb/gadget/amd5536udc.c | 1 - drivers/usb/gadget/langwell_udc.c | 1 - drivers/usb/gadget/s3c2410_udc.c | 1 - drivers/usb/host/r8a66597-hcd.c | 1 - drivers/usb/misc/iowarrior.c | 1 + drivers/usb/misc/rio500.c | 1 + drivers/usb/misc/usblcd.c | 1 + drivers/usb/musb/cppi_dma.h | 1 - drivers/usb/musb/musb_core.h | 1 - drivers/usb/serial/ftdi_sio.c | 1 + drivers/usb/serial/mos7840.c | 1 + drivers/usb/serial/usb-serial.c | 1 + drivers/video/fbmem.c | 1 - fs/adfs/super.c | 1 + fs/afs/super.c | 1 + fs/autofs4/dev-ioctl.c | 1 - fs/bfs/dir.c | 1 - fs/bfs/file.c | 1 - fs/btrfs/compression.c | 1 - fs/btrfs/file.c | 1 - fs/btrfs/inode.c | 1 - fs/btrfs/ioctl.c | 1 - fs/btrfs/super.c | 1 - fs/char_dev.c | 1 - fs/compat.c | 1 - fs/compat_ioctl.c | 1 + fs/exofs/super.c | 1 + fs/ext2/ioctl.c | 1 - fs/ext4/ioctl.c | 1 - fs/fat/dir.c | 1 - fs/fat/namei_msdos.c | 1 - fs/fat/namei_vfat.c | 1 - fs/fcntl.c | 1 - fs/freevxfs/vxfs_super.c | 1 + fs/hfs/super.c | 1 + fs/hfsplus/super.c | 1 + fs/hpfs/dir.c | 1 + fs/hpfs/file.c | 1 + fs/hpfs/hpfs_fn.h | 1 - fs/hpfs/inode.c | 1 + fs/hpfs/namei.c | 1 + fs/jffs2/super.c | 1 + fs/lockd/clntproc.c | 1 + fs/lockd/svc4proc.c | 1 + fs/lockd/svcproc.c | 1 + fs/nfs/delegation.c | 1 + fs/nfs/dir.c | 1 - fs/nfs/file.c | 1 - fs/nfs/inode.c | 1 - fs/nfs/nfs4proc.c | 1 - fs/nfs/read.c | 1 - fs/nfsd/nfsctl.c | 1 - fs/nfsd/nfssvc.c | 1 - fs/nilfs2/dir.c | 1 - fs/ocfs2/ioctl.c | 1 - fs/reiserfs/xattr.c | 1 - fs/squashfs/super.c | 1 + fs/ubifs/ioctl.c | 1 - fs/xfs/linux-2.6/xfs_file.c | 1 - include/linux/crash_dump.h | 1 - include/linux/hardirq.h | 2 ++ include/linux/quotaops.h | 1 - include/linux/sunrpc/xdr.h | 1 - kernel/power/user.c | 1 - kernel/trace/blktrace.c | 1 + kernel/trace/trace.c | 1 + net/appletalk/ddp.c | 1 + net/ipx/af_ipx.c | 1 + net/irda/af_irda.c | 1 + net/irda/irnet/irnet.h | 1 - net/irda/irnet/irnet_ppp.c | 1 + net/sunrpc/clnt.c | 1 - net/sunrpc/sched.c | 1 - net/sunrpc/svc_xprt.c | 1 + net/wanrouter/wanmain.c | 1 + net/x25/af_x25.c | 1 + 173 files changed, 81 insertions(+), 93 deletions(-) (limited to 'fs') diff --git a/arch/alpha/kernel/ptrace.c b/arch/alpha/kernel/ptrace.c index 1e9ad52c460e..e072041d19f8 100644 --- a/arch/alpha/kernel/ptrace.c +++ b/arch/alpha/kernel/ptrace.c @@ -8,7 +8,6 @@ #include #include #include -#include #include #include #include diff --git a/arch/blackfin/kernel/ptrace.c b/arch/blackfin/kernel/ptrace.c index d76618db50df..6a387eec6b65 100644 --- a/arch/blackfin/kernel/ptrace.c +++ b/arch/blackfin/kernel/ptrace.c @@ -31,7 +31,6 @@ #include #include #include -#include #include #include #include diff --git a/arch/blackfin/kernel/sys_bfin.c b/arch/blackfin/kernel/sys_bfin.c index a8f1329c15a4..3da60fb13ce4 100644 --- a/arch/blackfin/kernel/sys_bfin.c +++ b/arch/blackfin/kernel/sys_bfin.c @@ -29,7 +29,6 @@ * 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA */ -#include #include #include #include diff --git a/arch/cris/kernel/sys_cris.c b/arch/cris/kernel/sys_cris.c index a79fbd87021b..2ad962c7e88e 100644 --- a/arch/cris/kernel/sys_cris.c +++ b/arch/cris/kernel/sys_cris.c @@ -15,7 +15,6 @@ #include #include #include -#include #include #include #include diff --git a/arch/ia64/kernel/ptrace.c b/arch/ia64/kernel/ptrace.c index 92c9689b7d97..9daa87fdb018 100644 --- a/arch/ia64/kernel/ptrace.c +++ b/arch/ia64/kernel/ptrace.c @@ -15,7 +15,6 @@ #include #include #include -#include #include #include #include diff --git a/arch/m32r/kernel/ptrace.c b/arch/m32r/kernel/ptrace.c index bf0abe9e1f73..98b8feb12ed8 100644 --- a/arch/m32r/kernel/ptrace.c +++ b/arch/m32r/kernel/ptrace.c @@ -19,7 +19,6 @@ #include #include #include -#include #include #include #include diff --git a/arch/microblaze/kernel/ptrace.c b/arch/microblaze/kernel/ptrace.c index b86aa623e36d..53ff39af6a5c 100644 --- a/arch/microblaze/kernel/ptrace.c +++ b/arch/microblaze/kernel/ptrace.c @@ -27,7 +27,6 @@ #include #include #include -#include #include #include diff --git a/arch/microblaze/kernel/signal.c b/arch/microblaze/kernel/signal.c index 493819c25fba..1c80e4fc40ce 100644 --- a/arch/microblaze/kernel/signal.c +++ b/arch/microblaze/kernel/signal.c @@ -21,7 +21,6 @@ #include #include #include -#include #include #include #include diff --git a/arch/microblaze/kernel/sys_microblaze.c b/arch/microblaze/kernel/sys_microblaze.c index 8c9ebac5da10..e000bce09b2b 100644 --- a/arch/microblaze/kernel/sys_microblaze.c +++ b/arch/microblaze/kernel/sys_microblaze.c @@ -15,7 +15,6 @@ #include #include #include -#include #include #include #include diff --git a/arch/mips/kernel/ptrace32.c b/arch/mips/kernel/ptrace32.c index c4f9ac17474a..32644b4a0714 100644 --- a/arch/mips/kernel/ptrace32.c +++ b/arch/mips/kernel/ptrace32.c @@ -22,7 +22,6 @@ #include #include #include -#include #include #include diff --git a/arch/mips/mm/hugetlbpage.c b/arch/mips/mm/hugetlbpage.c index 471c09aa1614..8c2834f5919d 100644 --- a/arch/mips/mm/hugetlbpage.c +++ b/arch/mips/mm/hugetlbpage.c @@ -16,7 +16,6 @@ #include #include #include -#include #include #include #include diff --git a/arch/mn10300/kernel/ptrace.c b/arch/mn10300/kernel/ptrace.c index e143339ad28e..cf847dabc1bd 100644 --- a/arch/mn10300/kernel/ptrace.c +++ b/arch/mn10300/kernel/ptrace.c @@ -13,7 +13,6 @@ #include #include #include -#include #include #include #include diff --git a/arch/mn10300/kernel/signal.c b/arch/mn10300/kernel/signal.c index 9f7572a0f578..feb2f2e810db 100644 --- a/arch/mn10300/kernel/signal.c +++ b/arch/mn10300/kernel/signal.c @@ -12,7 +12,6 @@ #include #include #include -#include #include #include #include diff --git a/arch/mn10300/kernel/sys_mn10300.c b/arch/mn10300/kernel/sys_mn10300.c index bca5a84dc72c..29d196b83d25 100644 --- a/arch/mn10300/kernel/sys_mn10300.c +++ b/arch/mn10300/kernel/sys_mn10300.c @@ -13,7 +13,6 @@ #include #include #include -#include #include #include #include diff --git a/arch/mn10300/kernel/traps.c b/arch/mn10300/kernel/traps.c index 0dfdc5001124..91365adba4f5 100644 --- a/arch/mn10300/kernel/traps.c +++ b/arch/mn10300/kernel/traps.c @@ -17,7 +17,6 @@ #include #include #include -#include #include #include #include diff --git a/arch/mn10300/mm/fault.c b/arch/mn10300/mm/fault.c index a62e1e138bc1..53bb17d0f068 100644 --- a/arch/mn10300/mm/fault.c +++ b/arch/mn10300/mm/fault.c @@ -20,7 +20,6 @@ #include #include #include -#include #include #include #include /* For unblank_screen() */ diff --git a/arch/mn10300/mm/misalignment.c b/arch/mn10300/mm/misalignment.c index 94c4a4358065..30016251f658 100644 --- a/arch/mn10300/mm/misalignment.c +++ b/arch/mn10300/mm/misalignment.c @@ -17,7 +17,6 @@ #include #include #include -#include #include #include #include diff --git a/arch/powerpc/kernel/ptrace32.c b/arch/powerpc/kernel/ptrace32.c index 297632cba047..8a6daf4129f6 100644 --- a/arch/powerpc/kernel/ptrace32.c +++ b/arch/powerpc/kernel/ptrace32.c @@ -21,7 +21,6 @@ #include #include #include -#include #include #include #include diff --git a/arch/s390/kernel/dis.c b/arch/s390/kernel/dis.c index d2f270c995d9..db943a7ec513 100644 --- a/arch/s390/kernel/dis.c +++ b/arch/s390/kernel/dis.c @@ -15,7 +15,6 @@ #include #include #include -#include #include #include #include diff --git a/arch/s390/kernel/ptrace.c b/arch/s390/kernel/ptrace.c index 490b39934d65..43acd73105b7 100644 --- a/arch/s390/kernel/ptrace.c +++ b/arch/s390/kernel/ptrace.c @@ -26,7 +26,6 @@ #include #include #include -#include #include #include #include diff --git a/arch/s390/mm/fault.c b/arch/s390/mm/fault.c index 74eb26bf1970..e5e119fe03b2 100644 --- a/arch/s390/mm/fault.c +++ b/arch/s390/mm/fault.c @@ -22,7 +22,6 @@ #include #include #include -#include #include #include #include diff --git a/arch/sh/mm/tlb-sh3.c b/arch/sh/mm/tlb-sh3.c index 7fbfd5a11ffa..17cb7c3adf22 100644 --- a/arch/sh/mm/tlb-sh3.c +++ b/arch/sh/mm/tlb-sh3.c @@ -18,7 +18,6 @@ #include #include #include -#include #include #include diff --git a/arch/sparc/kernel/ptrace_32.c b/arch/sparc/kernel/ptrace_32.c index 8ce6285a06d5..7e3dfd9bb97e 100644 --- a/arch/sparc/kernel/ptrace_32.c +++ b/arch/sparc/kernel/ptrace_32.c @@ -16,7 +16,6 @@ #include #include #include -#include #include #include #include diff --git a/arch/sparc/kernel/ptrace_64.c b/arch/sparc/kernel/ptrace_64.c index a941c610e7ce..4ae91dc2feb9 100644 --- a/arch/sparc/kernel/ptrace_64.c +++ b/arch/sparc/kernel/ptrace_64.c @@ -17,7 +17,6 @@ #include #include #include -#include #include #include #include diff --git a/arch/sparc/kernel/time_64.c b/arch/sparc/kernel/time_64.c index 5c12e79b4bdf..da1218e8ee87 100644 --- a/arch/sparc/kernel/time_64.c +++ b/arch/sparc/kernel/time_64.c @@ -11,7 +11,6 @@ #include #include #include -#include #include #include #include diff --git a/arch/sparc/kernel/traps_32.c b/arch/sparc/kernel/traps_32.c index 358283341b47..c0490c7bbde0 100644 --- a/arch/sparc/kernel/traps_32.c +++ b/arch/sparc/kernel/traps_32.c @@ -13,7 +13,6 @@ #include #include #include -#include #include #include diff --git a/drivers/block/DAC960.c b/drivers/block/DAC960.c index 668dc234b8e2..1e6b7c14f697 100644 --- a/drivers/block/DAC960.c +++ b/drivers/block/DAC960.c @@ -36,6 +36,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/block/cciss.c b/drivers/block/cciss.c index 65a0655e7fc8..a52cc7fe45ea 100644 --- a/drivers/block/cciss.c +++ b/drivers/block/cciss.c @@ -26,6 +26,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/block/loop.c b/drivers/block/loop.c index 801f4ab83302..5757188cd1fb 100644 --- a/drivers/block/loop.c +++ b/drivers/block/loop.c @@ -61,7 +61,6 @@ #include #include #include -#include #include #include #include diff --git a/drivers/bluetooth/hci_vhci.c b/drivers/bluetooth/hci_vhci.c index 1df9dda2e377..d5cde6d86f89 100644 --- a/drivers/bluetooth/hci_vhci.c +++ b/drivers/bluetooth/hci_vhci.c @@ -28,7 +28,6 @@ #include #include #include -#include #include #include #include diff --git a/drivers/char/amiserial.c b/drivers/char/amiserial.c index 72429b6b2fa8..6c32fbf07164 100644 --- a/drivers/char/amiserial.c +++ b/drivers/char/amiserial.c @@ -81,6 +81,7 @@ static char *serial_version = "4.30"; #include #include #include +#include #include #include diff --git a/drivers/char/cyclades.c b/drivers/char/cyclades.c index f3366d3f06cf..2dafc2da0648 100644 --- a/drivers/char/cyclades.c +++ b/drivers/char/cyclades.c @@ -633,6 +633,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/char/epca.c b/drivers/char/epca.c index abef1f7d84fe..ff647ca1c489 100644 --- a/drivers/char/epca.c +++ b/drivers/char/epca.c @@ -36,6 +36,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/char/isicom.c b/drivers/char/isicom.c index 621d1184673c..4f1f4cd670da 100644 --- a/drivers/char/isicom.c +++ b/drivers/char/isicom.c @@ -122,6 +122,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/char/istallion.c b/drivers/char/istallion.c index 0c999f5bb3db..ab2f3349c5c4 100644 --- a/drivers/char/istallion.c +++ b/drivers/char/istallion.c @@ -20,6 +20,7 @@ #include #include +#include #include #include #include diff --git a/drivers/char/moxa.c b/drivers/char/moxa.c index 65b6ff2442c6..dd0083bbb64a 100644 --- a/drivers/char/moxa.c +++ b/drivers/char/moxa.c @@ -34,6 +34,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/char/mxser.c b/drivers/char/mxser.c index 52d953eb30c3..dbf8d52f31d0 100644 --- a/drivers/char/mxser.c +++ b/drivers/char/mxser.c @@ -23,6 +23,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/char/n_hdlc.c b/drivers/char/n_hdlc.c index 1c43c8cdee25..c68118efad84 100644 --- a/drivers/char/n_hdlc.c +++ b/drivers/char/n_hdlc.c @@ -97,6 +97,7 @@ #include #include #include +#include #include /* used in new tty drivers */ #include /* used in new tty drivers */ #include diff --git a/drivers/char/n_r3964.c b/drivers/char/n_r3964.c index 2e99158ebb8a..6934025a1ac1 100644 --- a/drivers/char/n_r3964.c +++ b/drivers/char/n_r3964.c @@ -58,6 +58,7 @@ #include #include #include +#include #include #include #include /* used in new tty drivers */ diff --git a/drivers/char/pty.c b/drivers/char/pty.c index 9d1b4f548f67..6e6942c45f5b 100644 --- a/drivers/char/pty.c +++ b/drivers/char/pty.c @@ -22,6 +22,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/char/rio/rio_linux.c b/drivers/char/rio/rio_linux.c index ce81da5b2da9..d58c2eb07f07 100644 --- a/drivers/char/rio/rio_linux.c +++ b/drivers/char/rio/rio_linux.c @@ -44,6 +44,7 @@ #include #include #include +#include #include #include diff --git a/drivers/char/riscom8.c b/drivers/char/riscom8.c index 217660451237..171711acf5cd 100644 --- a/drivers/char/riscom8.c +++ b/drivers/char/riscom8.c @@ -47,6 +47,7 @@ #include #include #include +#include #include #include diff --git a/drivers/char/rocket.c b/drivers/char/rocket.c index 63d5b628477a..0e29a23ec4c5 100644 --- a/drivers/char/rocket.c +++ b/drivers/char/rocket.c @@ -73,6 +73,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/char/serial167.c b/drivers/char/serial167.c index f1f24f0ee26f..51e7a46787be 100644 --- a/drivers/char/serial167.c +++ b/drivers/char/serial167.c @@ -52,6 +52,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/char/specialix.c b/drivers/char/specialix.c index e72be4190a44..bfe4cdb2febb 100644 --- a/drivers/char/specialix.c +++ b/drivers/char/specialix.c @@ -87,6 +87,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/char/sx.c b/drivers/char/sx.c index 518f2a25d91e..a81ec4fcf6ff 100644 --- a/drivers/char/sx.c +++ b/drivers/char/sx.c @@ -216,6 +216,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/char/synclink.c b/drivers/char/synclink.c index afded3a2379c..813552f14884 100644 --- a/drivers/char/synclink.c +++ b/drivers/char/synclink.c @@ -81,6 +81,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/char/synclink_gt.c b/drivers/char/synclink_gt.c index a2e67e6df3a1..91f20a92fddf 100644 --- a/drivers/char/synclink_gt.c +++ b/drivers/char/synclink_gt.c @@ -62,6 +62,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/char/synclinkmp.c b/drivers/char/synclinkmp.c index 6f727e3c53ad..8d4a2a8a0a70 100644 --- a/drivers/char/synclinkmp.c +++ b/drivers/char/synclinkmp.c @@ -52,6 +52,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/char/tpm/tpm.c b/drivers/char/tpm/tpm.c index ccdd828adcef..b0603b2e5684 100644 --- a/drivers/char/tpm/tpm.c +++ b/drivers/char/tpm/tpm.c @@ -26,7 +26,6 @@ #include #include #include -#include #include "tpm.h" diff --git a/drivers/char/tty_ioctl.c b/drivers/char/tty_ioctl.c index b24f6c6a1ea3..ad6ba4ed2808 100644 --- a/drivers/char/tty_ioctl.c +++ b/drivers/char/tty_ioctl.c @@ -21,7 +21,6 @@ #include #include #include -#include #include #include diff --git a/drivers/char/tty_ldisc.c b/drivers/char/tty_ldisc.c index 913aa8d3f1c5..0ef0dc97ba20 100644 --- a/drivers/char/tty_ldisc.c +++ b/drivers/char/tty_ldisc.c @@ -21,7 +21,6 @@ #include #include #include -#include #include #include #include diff --git a/drivers/char/vt.c b/drivers/char/vt.c index d9113b4c76e3..7947bd1b4cf7 100644 --- a/drivers/char/vt.c +++ b/drivers/char/vt.c @@ -89,6 +89,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/char/vt_ioctl.c b/drivers/char/vt_ioctl.c index 7539bed0f7e0..95189f288f8c 100644 --- a/drivers/char/vt_ioctl.c +++ b/drivers/char/vt_ioctl.c @@ -25,6 +25,7 @@ #include #include #include +#include #include #include diff --git a/drivers/gpio/vr41xx_giu.c b/drivers/gpio/vr41xx_giu.c index b70e06133e78..b16c9a8c03f5 100644 --- a/drivers/gpio/vr41xx_giu.c +++ b/drivers/gpio/vr41xx_giu.c @@ -29,7 +29,6 @@ #include #include #include -#include #include #include diff --git a/drivers/hid/usbhid/hid-core.c b/drivers/hid/usbhid/hid-core.c index 76c4bbe9dccb..3c1fcb7640ab 100644 --- a/drivers/hid/usbhid/hid-core.c +++ b/drivers/hid/usbhid/hid-core.c @@ -22,7 +22,6 @@ #include #include #include -#include #include #include #include diff --git a/drivers/isdn/hisax/hfc_usb.c b/drivers/isdn/hisax/hfc_usb.c index 8df889b0c1a9..9de54202c90c 100644 --- a/drivers/isdn/hisax/hfc_usb.c +++ b/drivers/isdn/hisax/hfc_usb.c @@ -37,7 +37,6 @@ #include #include #include -#include #include #include #include "hisax.h" diff --git a/drivers/isdn/i4l/isdn_tty.c b/drivers/isdn/i4l/isdn_tty.c index b4d4522e5071..2881a66c1aa9 100644 --- a/drivers/isdn/i4l/isdn_tty.c +++ b/drivers/isdn/i4l/isdn_tty.c @@ -13,6 +13,7 @@ #include #include +#include #include "isdn_common.h" #include "isdn_tty.h" #ifdef CONFIG_ISDN_AUDIO diff --git a/drivers/isdn/mISDN/stack.c b/drivers/isdn/mISDN/stack.c index e2f45019ebf0..3e1532a180ff 100644 --- a/drivers/isdn/mISDN/stack.c +++ b/drivers/isdn/mISDN/stack.c @@ -17,6 +17,7 @@ #include #include +#include #include "core.h" static u_int *debug; diff --git a/drivers/media/dvb/bt8xx/dst_ca.c b/drivers/media/dvb/bt8xx/dst_ca.c index 4601b059b2b2..0e246eaad05a 100644 --- a/drivers/media/dvb/bt8xx/dst_ca.c +++ b/drivers/media/dvb/bt8xx/dst_ca.c @@ -21,6 +21,7 @@ #include #include #include +#include #include #include #include "dvbdev.h" diff --git a/drivers/media/dvb/dvb-core/dvbdev.h b/drivers/media/dvb/dvb-core/dvbdev.h index 79927305e84d..487919bea7ae 100644 --- a/drivers/media/dvb/dvb-core/dvbdev.h +++ b/drivers/media/dvb/dvb-core/dvbdev.h @@ -27,7 +27,6 @@ #include #include #include -#include #define DVB_MAJOR 212 diff --git a/drivers/media/dvb/ttpci/av7110.c b/drivers/media/dvb/ttpci/av7110.c index d1d959ed37b7..8d65c652ba50 100644 --- a/drivers/media/dvb/ttpci/av7110.c +++ b/drivers/media/dvb/ttpci/av7110.c @@ -36,7 +36,6 @@ #include #include #include -#include #include #include diff --git a/drivers/media/radio/radio-mr800.c b/drivers/media/radio/radio-mr800.c index 837467f93805..575bf9d89419 100644 --- a/drivers/media/radio/radio-mr800.c +++ b/drivers/media/radio/radio-mr800.c @@ -58,6 +58,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/media/radio/radio-si470x.c b/drivers/media/radio/radio-si470x.c index 46d216329611..e85f318b4d2b 100644 --- a/drivers/media/radio/radio-si470x.c +++ b/drivers/media/radio/radio-si470x.c @@ -127,6 +127,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/media/video/bt8xx/bttv-driver.c b/drivers/media/video/bt8xx/bttv-driver.c index 5eb1464af670..d147d29bb0d3 100644 --- a/drivers/media/video/bt8xx/bttv-driver.c +++ b/drivers/media/video/bt8xx/bttv-driver.c @@ -41,6 +41,7 @@ #include #include #include +#include #include #include #include "bttvp.h" diff --git a/drivers/media/video/cx23885/cx23885-417.c b/drivers/media/video/cx23885/cx23885-417.c index 2943bfd32a94..428f0c45e6b7 100644 --- a/drivers/media/video/cx23885/cx23885-417.c +++ b/drivers/media/video/cx23885/cx23885-417.c @@ -31,6 +31,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/media/video/cx23885/cx23885-video.c b/drivers/media/video/cx23885/cx23885-video.c index 70836af3ab48..5d6093336300 100644 --- a/drivers/media/video/cx23885/cx23885-video.c +++ b/drivers/media/video/cx23885/cx23885-video.c @@ -26,6 +26,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/media/video/cx88/cx88-blackbird.c b/drivers/media/video/cx88/cx88-blackbird.c index 44eacfb0d0d6..356d6896da3f 100644 --- a/drivers/media/video/cx88/cx88-blackbird.c +++ b/drivers/media/video/cx88/cx88-blackbird.c @@ -32,6 +32,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/media/video/cx88/cx88-video.c b/drivers/media/video/cx88/cx88-video.c index b12770848c00..2bb54c3ef5cd 100644 --- a/drivers/media/video/cx88/cx88-video.c +++ b/drivers/media/video/cx88/cx88-video.c @@ -31,6 +31,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/media/video/dabusb.c b/drivers/media/video/dabusb.c index ec2f45dde164..0664d111085f 100644 --- a/drivers/media/video/dabusb.c +++ b/drivers/media/video/dabusb.c @@ -32,6 +32,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/media/video/pwc/pwc-if.c b/drivers/media/video/pwc/pwc-if.c index db25c3034c11..8d17cf613306 100644 --- a/drivers/media/video/pwc/pwc-if.c +++ b/drivers/media/video/pwc/pwc-if.c @@ -62,6 +62,7 @@ #include #include #include +#include #ifdef CONFIG_USB_PWC_INPUT_EVDEV #include #endif diff --git a/drivers/media/video/pwc/pwc.h b/drivers/media/video/pwc/pwc.h index 0be6f814f539..0b658dee05a4 100644 --- a/drivers/media/video/pwc/pwc.h +++ b/drivers/media/video/pwc/pwc.h @@ -29,7 +29,6 @@ #include #include #include -#include #include #include #include diff --git a/drivers/media/video/s2255drv.c b/drivers/media/video/s2255drv.c index 6be845ccc7d7..9e3262c0ba37 100644 --- a/drivers/media/video/s2255drv.c +++ b/drivers/media/video/s2255drv.c @@ -48,6 +48,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/media/video/saa5246a.c b/drivers/media/video/saa5246a.c index 155804b061e9..b624a4c01fdc 100644 --- a/drivers/media/video/saa5246a.c +++ b/drivers/media/video/saa5246a.c @@ -43,7 +43,6 @@ #include #include #include -#include #include #include #include diff --git a/drivers/media/video/saa5249.c b/drivers/media/video/saa5249.c index 271d6e931b75..12835fb82c95 100644 --- a/drivers/media/video/saa5249.c +++ b/drivers/media/video/saa5249.c @@ -46,7 +46,6 @@ #include #include #include -#include #include #include #include diff --git a/drivers/media/video/saa7134/saa7134-empress.c b/drivers/media/video/saa7134/saa7134-empress.c index add1757f8930..296788c3bf0e 100644 --- a/drivers/media/video/saa7134/saa7134-empress.c +++ b/drivers/media/video/saa7134/saa7134-empress.c @@ -22,6 +22,7 @@ #include #include #include +#include #include #include "saa7134-reg.h" diff --git a/drivers/media/video/se401.c b/drivers/media/video/se401.c index c8f05297d0f0..85ffc2cba039 100644 --- a/drivers/media/video/se401.c +++ b/drivers/media/video/se401.c @@ -31,6 +31,7 @@ static const char version[] = "0.24"; #include #include #include +#include #include #include #include "se401.h" diff --git a/drivers/media/video/stk-webcam.c b/drivers/media/video/stk-webcam.c index 2e5937047278..4d6785e63455 100644 --- a/drivers/media/video/stk-webcam.c +++ b/drivers/media/video/stk-webcam.c @@ -27,6 +27,7 @@ #include #include #include +#include #include #include diff --git a/drivers/media/video/stradis.c b/drivers/media/video/stradis.c index 0eb313082c97..eaada39c76fd 100644 --- a/drivers/media/video/stradis.c +++ b/drivers/media/video/stradis.c @@ -26,6 +26,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/media/video/stv680.c b/drivers/media/video/stv680.c index 75f286f7a2e9..8b4e7dafce7b 100644 --- a/drivers/media/video/stv680.c +++ b/drivers/media/video/stv680.c @@ -62,6 +62,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/media/video/usbvideo/vicam.c b/drivers/media/video/usbvideo/vicam.c index 8d73979596f9..45fce39ec9ad 100644 --- a/drivers/media/video/usbvideo/vicam.c +++ b/drivers/media/video/usbvideo/vicam.c @@ -43,6 +43,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/media/video/usbvision/usbvision-video.c b/drivers/media/video/usbvision/usbvision-video.c index 90b58914f984..90d9b5c0e9a7 100644 --- a/drivers/media/video/usbvision/usbvision-video.c +++ b/drivers/media/video/usbvision/usbvision-video.c @@ -50,6 +50,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/media/video/v4l2-dev.c b/drivers/media/video/v4l2-dev.c index 31eac66411d7..a7f1b69a7dab 100644 --- a/drivers/media/video/v4l2-dev.c +++ b/drivers/media/video/v4l2-dev.c @@ -25,7 +25,6 @@ #include #include #include -#include #include #include diff --git a/drivers/media/video/zoran/zoran_driver.c b/drivers/media/video/zoran/zoran_driver.c index 3d7df32a3d87..bcdefb1bcb3d 100644 --- a/drivers/media/video/zoran/zoran_driver.c +++ b/drivers/media/video/zoran/zoran_driver.c @@ -49,6 +49,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/misc/sgi-gru/grufile.c b/drivers/misc/sgi-gru/grufile.c index fa2d93a9fb8d..aed609832bc2 100644 --- a/drivers/misc/sgi-gru/grufile.c +++ b/drivers/misc/sgi-gru/grufile.c @@ -29,7 +29,6 @@ #include #include #include -#include #include #include #include diff --git a/drivers/misc/sgi-gru/grukservices.c b/drivers/misc/sgi-gru/grukservices.c index eedbf9c32760..79689b10f937 100644 --- a/drivers/misc/sgi-gru/grukservices.c +++ b/drivers/misc/sgi-gru/grukservices.c @@ -24,7 +24,6 @@ #include #include #include -#include #include #include #include diff --git a/drivers/net/irda/irtty-sir.c b/drivers/net/irda/irtty-sir.c index d53aa9582137..20f9bc626688 100644 --- a/drivers/net/irda/irtty-sir.c +++ b/drivers/net/irda/irtty-sir.c @@ -31,7 +31,6 @@ #include #include #include -#include #include #include diff --git a/drivers/pci/hotplug/cpci_hotplug_core.c b/drivers/pci/hotplug/cpci_hotplug_core.c index a5b9f6ae507b..d703e73fffa7 100644 --- a/drivers/pci/hotplug/cpci_hotplug_core.c +++ b/drivers/pci/hotplug/cpci_hotplug_core.c @@ -32,7 +32,6 @@ #include #include #include -#include #include #include #include diff --git a/drivers/pci/hotplug/cpqphp_ctrl.c b/drivers/pci/hotplug/cpqphp_ctrl.c index 2fa47af992a8..0ff689afa757 100644 --- a/drivers/pci/hotplug/cpqphp_ctrl.c +++ b/drivers/pci/hotplug/cpqphp_ctrl.c @@ -34,7 +34,6 @@ #include #include #include -#include #include #include #include diff --git a/drivers/pci/hotplug/cpqphp_sysfs.c b/drivers/pci/hotplug/cpqphp_sysfs.c index 8450f4a6568a..e6089bdb6e5b 100644 --- a/drivers/pci/hotplug/cpqphp_sysfs.c +++ b/drivers/pci/hotplug/cpqphp_sysfs.c @@ -33,6 +33,7 @@ #include #include #include +#include #include #include "cpqphp.h" diff --git a/drivers/pci/hotplug/pciehp_ctrl.c b/drivers/pci/hotplug/pciehp_ctrl.c index ff4034502d24..8aab8edf123e 100644 --- a/drivers/pci/hotplug/pciehp_ctrl.c +++ b/drivers/pci/hotplug/pciehp_ctrl.c @@ -30,7 +30,6 @@ #include #include #include -#include #include #include #include "../pci.h" diff --git a/drivers/pci/syscall.c b/drivers/pci/syscall.c index ec22284eed30..e1c1ec540893 100644 --- a/drivers/pci/syscall.c +++ b/drivers/pci/syscall.c @@ -9,7 +9,6 @@ #include #include -#include #include #include #include "pci.h" diff --git a/drivers/s390/block/dasd_ioctl.c b/drivers/s390/block/dasd_ioctl.c index 4ce3f72ee1c1..df918ef27965 100644 --- a/drivers/s390/block/dasd_ioctl.c +++ b/drivers/s390/block/dasd_ioctl.c @@ -16,6 +16,7 @@ #include #include #include +#include #include #include diff --git a/drivers/scsi/qla2xxx/qla_mid.c b/drivers/scsi/qla2xxx/qla_mid.c index 650bcef08f2a..cd78c501803a 100644 --- a/drivers/scsi/qla2xxx/qla_mid.c +++ b/drivers/scsi/qla2xxx/qla_mid.c @@ -9,7 +9,6 @@ #include #include -#include #include #include diff --git a/drivers/telephony/ixj.c b/drivers/telephony/ixj.c index a913efc69669..40de151f2789 100644 --- a/drivers/telephony/ixj.c +++ b/drivers/telephony/ixj.c @@ -257,6 +257,7 @@ #include /* everything... */ #include /* error codes */ #include +#include #include #include #include diff --git a/drivers/telephony/phonedev.c b/drivers/telephony/phonedev.c index b52cc830c0b4..f3873f650bb4 100644 --- a/drivers/telephony/phonedev.c +++ b/drivers/telephony/phonedev.c @@ -23,7 +23,6 @@ #include #include #include -#include #include #include diff --git a/drivers/usb/class/cdc-wdm.c b/drivers/usb/class/cdc-wdm.c index 0fe434505ac4..ba589d4ca8bc 100644 --- a/drivers/usb/class/cdc-wdm.c +++ b/drivers/usb/class/cdc-wdm.c @@ -15,7 +15,6 @@ #include #include #include -#include #include #include #include diff --git a/drivers/usb/gadget/amd5536udc.c b/drivers/usb/gadget/amd5536udc.c index 826f3adde5d8..77352ccc245e 100644 --- a/drivers/usb/gadget/amd5536udc.c +++ b/drivers/usb/gadget/amd5536udc.c @@ -48,7 +48,6 @@ #include #include #include -#include #include #include #include diff --git a/drivers/usb/gadget/langwell_udc.c b/drivers/usb/gadget/langwell_udc.c index 6829d5961359..a3913519fd58 100644 --- a/drivers/usb/gadget/langwell_udc.c +++ b/drivers/usb/gadget/langwell_udc.c @@ -34,7 +34,6 @@ #include #include #include -#include #include #include #include diff --git a/drivers/usb/gadget/s3c2410_udc.c b/drivers/usb/gadget/s3c2410_udc.c index 9a2b8920532d..a9b452fe6221 100644 --- a/drivers/usb/gadget/s3c2410_udc.c +++ b/drivers/usb/gadget/s3c2410_udc.c @@ -28,7 +28,6 @@ #include #include #include -#include #include #include #include diff --git a/drivers/usb/host/r8a66597-hcd.c b/drivers/usb/host/r8a66597-hcd.c index 56976cc0352a..e18f74946e68 100644 --- a/drivers/usb/host/r8a66597-hcd.c +++ b/drivers/usb/host/r8a66597-hcd.c @@ -26,7 +26,6 @@ #include #include #include -#include #include #include #include diff --git a/drivers/usb/misc/iowarrior.c b/drivers/usb/misc/iowarrior.c index 3c5fe5cee05a..90e1a8dedfa9 100644 --- a/drivers/usb/misc/iowarrior.c +++ b/drivers/usb/misc/iowarrior.c @@ -18,6 +18,7 @@ #include #include #include +#include #include #include diff --git a/drivers/usb/misc/rio500.c b/drivers/usb/misc/rio500.c index deb95bb49fd1..d645f3899fe1 100644 --- a/drivers/usb/misc/rio500.c +++ b/drivers/usb/misc/rio500.c @@ -32,6 +32,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/usb/misc/usblcd.c b/drivers/usb/misc/usblcd.c index e0ff9ccd866b..29092b8e59ce 100644 --- a/drivers/usb/misc/usblcd.c +++ b/drivers/usb/misc/usblcd.c @@ -16,6 +16,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/usb/musb/cppi_dma.h b/drivers/usb/musb/cppi_dma.h index 8a39de3e6e47..59bf949e589b 100644 --- a/drivers/usb/musb/cppi_dma.h +++ b/drivers/usb/musb/cppi_dma.h @@ -5,7 +5,6 @@ #include #include -#include #include #include diff --git a/drivers/usb/musb/musb_core.h b/drivers/usb/musb/musb_core.h index f3772ca3b2cf..381d648a36b8 100644 --- a/drivers/usb/musb/musb_core.h +++ b/drivers/usb/musb/musb_core.h @@ -38,7 +38,6 @@ #include #include #include -#include #include #include #include diff --git a/drivers/usb/serial/ftdi_sio.c b/drivers/usb/serial/ftdi_sio.c index 5f08702f672f..5a8ae274d258 100644 --- a/drivers/usb/serial/ftdi_sio.c +++ b/drivers/usb/serial/ftdi_sio.c @@ -33,6 +33,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/usb/serial/mos7840.c b/drivers/usb/serial/mos7840.c index c40f95c1951c..c31940a307f8 100644 --- a/drivers/usb/serial/mos7840.c +++ b/drivers/usb/serial/mos7840.c @@ -26,6 +26,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/usb/serial/usb-serial.c b/drivers/usb/serial/usb-serial.c index a84216464ca0..0c39b55aeef4 100644 --- a/drivers/usb/serial/usb-serial.c +++ b/drivers/usb/serial/usb-serial.c @@ -21,6 +21,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/video/fbmem.c b/drivers/video/fbmem.c index 53ea05645ff8..a85c818be945 100644 --- a/drivers/video/fbmem.c +++ b/drivers/video/fbmem.c @@ -16,7 +16,6 @@ #include #include #include -#include #include #include #include diff --git a/fs/adfs/super.c b/fs/adfs/super.c index aad92f0a1048..6910a98bd73c 100644 --- a/fs/adfs/super.c +++ b/fs/adfs/super.c @@ -13,6 +13,7 @@ #include #include #include +#include #include #include "adfs.h" #include "dir_f.h" diff --git a/fs/afs/super.c b/fs/afs/super.c index ad0514d0115f..e1ea1c240b6a 100644 --- a/fs/afs/super.c +++ b/fs/afs/super.c @@ -18,6 +18,7 @@ #include #include #include +#include #include #include #include diff --git a/fs/autofs4/dev-ioctl.c b/fs/autofs4/dev-ioctl.c index f3da2eb51f56..00bf8fcb245f 100644 --- a/fs/autofs4/dev-ioctl.c +++ b/fs/autofs4/dev-ioctl.c @@ -19,7 +19,6 @@ #include #include #include -#include #include #include #include diff --git a/fs/bfs/dir.c b/fs/bfs/dir.c index 54bd07d44e68..1e41aadb1068 100644 --- a/fs/bfs/dir.c +++ b/fs/bfs/dir.c @@ -8,7 +8,6 @@ #include #include #include -#include #include #include #include "bfs.h" diff --git a/fs/bfs/file.c b/fs/bfs/file.c index 6a021265f018..88b9a3ff44e4 100644 --- a/fs/bfs/file.c +++ b/fs/bfs/file.c @@ -11,7 +11,6 @@ #include #include -#include #include "bfs.h" #undef DEBUG diff --git a/fs/btrfs/compression.c b/fs/btrfs/compression.c index de1e2fd32080..9d8ba4d54a37 100644 --- a/fs/btrfs/compression.c +++ b/fs/btrfs/compression.c @@ -26,7 +26,6 @@ #include #include #include -#include #include #include #include diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index 7c3cd248d8d6..4b833972273a 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -22,7 +22,6 @@ #include #include #include -#include #include #include #include diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 7ffa3d34ea19..791eab19e330 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -26,7 +26,6 @@ #include #include #include -#include #include #include #include diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index 9f4db848db10..bd88f25889f7 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -27,7 +27,6 @@ #include #include #include -#include #include #include #include diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index 9f179d4832d5..6d6d06cb6dfc 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -26,7 +26,6 @@ #include #include #include -#include #include #include #include diff --git a/fs/char_dev.c b/fs/char_dev.c index b7c9d5187a75..a173551e19d7 100644 --- a/fs/char_dev.c +++ b/fs/char_dev.c @@ -13,7 +13,6 @@ #include #include #include -#include #include #include diff --git a/fs/compat.c b/fs/compat.c index fbadb947727b..94502dab972a 100644 --- a/fs/compat.c +++ b/fs/compat.c @@ -32,7 +32,6 @@ #include #include #include -#include #include #include #include diff --git a/fs/compat_ioctl.c b/fs/compat_ioctl.c index 626c7483b4de..f28f070a60fc 100644 --- a/fs/compat_ioctl.c +++ b/fs/compat_ioctl.c @@ -19,6 +19,7 @@ #include #include #include +#include #include #include #include diff --git a/fs/exofs/super.c b/fs/exofs/super.c index a343b4ea62f6..5ab10c3bbebe 100644 --- a/fs/exofs/super.c +++ b/fs/exofs/super.c @@ -31,6 +31,7 @@ * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA */ +#include #include #include #include diff --git a/fs/ext2/ioctl.c b/fs/ext2/ioctl.c index 7cb4badef927..e7431309bdca 100644 --- a/fs/ext2/ioctl.c +++ b/fs/ext2/ioctl.c @@ -13,7 +13,6 @@ #include #include #include -#include #include #include diff --git a/fs/ext4/ioctl.c b/fs/ext4/ioctl.c index bb415408fdb6..24a6abb2aef5 100644 --- a/fs/ext4/ioctl.c +++ b/fs/ext4/ioctl.c @@ -12,7 +12,6 @@ #include #include #include -#include #include #include #include diff --git a/fs/fat/dir.c b/fs/fat/dir.c index 38ff75a0fe22..530b4ca01510 100644 --- a/fs/fat/dir.c +++ b/fs/fat/dir.c @@ -16,7 +16,6 @@ #include #include #include -#include #include #include #include diff --git a/fs/fat/namei_msdos.c b/fs/fat/namei_msdos.c index 82f88733b681..bbc94ae4fd77 100644 --- a/fs/fat/namei_msdos.c +++ b/fs/fat/namei_msdos.c @@ -9,7 +9,6 @@ #include #include #include -#include #include "fat.h" /* Characters that are undesirable in an MS-DOS file name */ diff --git a/fs/fat/namei_vfat.c b/fs/fat/namei_vfat.c index 73471b7ecc8c..cb6e83557112 100644 --- a/fs/fat/namei_vfat.c +++ b/fs/fat/namei_vfat.c @@ -19,7 +19,6 @@ #include #include #include -#include #include #include #include "fat.h" diff --git a/fs/fcntl.c b/fs/fcntl.c index a040b764f8e3..ae413086db97 100644 --- a/fs/fcntl.c +++ b/fs/fcntl.c @@ -19,7 +19,6 @@ #include #include #include -#include #include #include diff --git a/fs/freevxfs/vxfs_super.c b/fs/freevxfs/vxfs_super.c index cdbd1654e4cd..1e8af939b3e4 100644 --- a/fs/freevxfs/vxfs_super.c +++ b/fs/freevxfs/vxfs_super.c @@ -38,6 +38,7 @@ #include #include #include +#include #include #include #include diff --git a/fs/hfs/super.c b/fs/hfs/super.c index 6f833dc8e910..f7fcbe49da72 100644 --- a/fs/hfs/super.c +++ b/fs/hfs/super.c @@ -19,6 +19,7 @@ #include #include #include +#include #include #include "hfs_fs.h" diff --git a/fs/hfsplus/super.c b/fs/hfsplus/super.c index 9fc3af0c0dab..c0759fe0855b 100644 --- a/fs/hfsplus/super.c +++ b/fs/hfsplus/super.c @@ -12,6 +12,7 @@ #include #include #include +#include #include #include diff --git a/fs/hpfs/dir.c b/fs/hpfs/dir.c index 6916c41d7017..8865c94f55f6 100644 --- a/fs/hpfs/dir.c +++ b/fs/hpfs/dir.c @@ -6,6 +6,7 @@ * directory VFS functions */ +#include #include "hpfs_fn.h" static int hpfs_dir_release(struct inode *inode, struct file *filp) diff --git a/fs/hpfs/file.c b/fs/hpfs/file.c index 64ab52259204..3efabff00367 100644 --- a/fs/hpfs/file.c +++ b/fs/hpfs/file.c @@ -6,6 +6,7 @@ * file VFS functions */ +#include #include "hpfs_fn.h" #define BLOCKS(size) (((size) + 511) >> 9) diff --git a/fs/hpfs/hpfs_fn.h b/fs/hpfs/hpfs_fn.h index c2ea31bae313..701ca54c0867 100644 --- a/fs/hpfs/hpfs_fn.h +++ b/fs/hpfs/hpfs_fn.h @@ -13,7 +13,6 @@ #include #include #include -#include #include "hpfs.h" diff --git a/fs/hpfs/inode.c b/fs/hpfs/inode.c index 39a1bfbea312..fe703ae46bc7 100644 --- a/fs/hpfs/inode.c +++ b/fs/hpfs/inode.c @@ -6,6 +6,7 @@ * inode VFS functions */ +#include #include "hpfs_fn.h" void hpfs_init_inode(struct inode *i) diff --git a/fs/hpfs/namei.c b/fs/hpfs/namei.c index b649232dde97..82b9c4ba9ed0 100644 --- a/fs/hpfs/namei.c +++ b/fs/hpfs/namei.c @@ -6,6 +6,7 @@ * adding & removing files & directories */ #include +#include #include "hpfs_fn.h" static int hpfs_mkdir(struct inode *dir, struct dentry *dentry, int mode) diff --git a/fs/jffs2/super.c b/fs/jffs2/super.c index 07a22caf2687..0035c021395a 100644 --- a/fs/jffs2/super.c +++ b/fs/jffs2/super.c @@ -12,6 +12,7 @@ #include #include #include +#include #include #include #include diff --git a/fs/lockd/clntproc.c b/fs/lockd/clntproc.c index f2fdcbce143e..4336adba952a 100644 --- a/fs/lockd/clntproc.c +++ b/fs/lockd/clntproc.c @@ -7,6 +7,7 @@ */ #include +#include #include #include #include diff --git a/fs/lockd/svc4proc.c b/fs/lockd/svc4proc.c index 1725037374c5..bd173a6ca3b1 100644 --- a/fs/lockd/svc4proc.c +++ b/fs/lockd/svc4proc.c @@ -10,6 +10,7 @@ #include #include #include +#include #include #include #include diff --git a/fs/lockd/svcproc.c b/fs/lockd/svcproc.c index 3688e55901fc..e1d28ddd2169 100644 --- a/fs/lockd/svcproc.c +++ b/fs/lockd/svcproc.c @@ -10,6 +10,7 @@ #include #include #include +#include #include #include #include diff --git a/fs/nfs/delegation.c b/fs/nfs/delegation.c index af05b918cb5b..6dd48a4405b4 100644 --- a/fs/nfs/delegation.c +++ b/fs/nfs/delegation.c @@ -10,6 +10,7 @@ #include #include #include +#include #include #include diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c index 89f98e9a024b..38d42c29fb92 100644 --- a/fs/nfs/dir.c +++ b/fs/nfs/dir.c @@ -29,7 +29,6 @@ #include #include #include -#include #include #include #include diff --git a/fs/nfs/file.c b/fs/nfs/file.c index 0055b813ec2c..05062329b678 100644 --- a/fs/nfs/file.c +++ b/fs/nfs/file.c @@ -26,7 +26,6 @@ #include #include #include -#include #include #include diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c index 64f87194d390..bd7938eda6a8 100644 --- a/fs/nfs/inode.c +++ b/fs/nfs/inode.c @@ -30,7 +30,6 @@ #include #include #include -#include #include #include #include diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 92ce43517814..ff0c080db59b 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -45,7 +45,6 @@ #include #include #include -#include #include #include #include diff --git a/fs/nfs/read.c b/fs/nfs/read.c index 96c4ebfa46f4..73ea5e8d66ce 100644 --- a/fs/nfs/read.c +++ b/fs/nfs/read.c @@ -18,7 +18,6 @@ #include #include #include -#include #include diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c index 1250fb978ac1..6d0847562d87 100644 --- a/fs/nfsd/nfsctl.c +++ b/fs/nfsd/nfsctl.c @@ -25,7 +25,6 @@ #include #include #include -#include #include #include diff --git a/fs/nfsd/nfssvc.c b/fs/nfsd/nfssvc.c index d4c9884cd54b..492c79b7800b 100644 --- a/fs/nfsd/nfssvc.c +++ b/fs/nfsd/nfssvc.c @@ -18,7 +18,6 @@ #include #include #include -#include #include #include #include diff --git a/fs/nilfs2/dir.c b/fs/nilfs2/dir.c index 54100acc1102..1a4fa04cf071 100644 --- a/fs/nilfs2/dir.c +++ b/fs/nilfs2/dir.c @@ -43,7 +43,6 @@ */ #include -#include #include "nilfs.h" #include "page.h" diff --git a/fs/ocfs2/ioctl.c b/fs/ocfs2/ioctl.c index 9fcd36dcc9a0..467b413bec21 100644 --- a/fs/ocfs2/ioctl.c +++ b/fs/ocfs2/ioctl.c @@ -7,7 +7,6 @@ #include #include -#include #define MLOG_MASK_PREFIX ML_INODE #include diff --git a/fs/reiserfs/xattr.c b/fs/reiserfs/xattr.c index f3d47d856848..6925b835a43b 100644 --- a/fs/reiserfs/xattr.c +++ b/fs/reiserfs/xattr.c @@ -46,7 +46,6 @@ #include #include #include -#include #include #include diff --git a/fs/squashfs/super.c b/fs/squashfs/super.c index 3b52770f46ff..cb5fc57e370b 100644 --- a/fs/squashfs/super.c +++ b/fs/squashfs/super.c @@ -30,6 +30,7 @@ #include #include #include +#include #include #include #include diff --git a/fs/ubifs/ioctl.c b/fs/ubifs/ioctl.c index 6db7a6be6c97..8aacd64957a2 100644 --- a/fs/ubifs/ioctl.c +++ b/fs/ubifs/ioctl.c @@ -25,7 +25,6 @@ /* This file implements EXT2-compatible extended attribute ioctl() calls */ #include -#include #include #include "ubifs.h" diff --git a/fs/xfs/linux-2.6/xfs_file.c b/fs/xfs/linux-2.6/xfs_file.c index f4e255441574..0542fd507649 100644 --- a/fs/xfs/linux-2.6/xfs_file.c +++ b/fs/xfs/linux-2.6/xfs_file.c @@ -41,7 +41,6 @@ #include "xfs_ioctl.h" #include -#include static struct vm_operations_struct xfs_file_vm_ops; diff --git a/include/linux/crash_dump.h b/include/linux/crash_dump.h index 2dac064d8359..0026f267da20 100644 --- a/include/linux/crash_dump.h +++ b/include/linux/crash_dump.h @@ -3,7 +3,6 @@ #ifdef CONFIG_CRASH_DUMP #include -#include #include #include diff --git a/include/linux/hardirq.h b/include/linux/hardirq.h index 45257475623c..8246c697863d 100644 --- a/include/linux/hardirq.h +++ b/include/linux/hardirq.h @@ -2,7 +2,9 @@ #define LINUX_HARDIRQ_H #include +#ifdef CONFIG_PREEMPT #include +#endif #include #include #include diff --git a/include/linux/quotaops.h b/include/linux/quotaops.h index 7bc457593684..26361c4c037a 100644 --- a/include/linux/quotaops.h +++ b/include/linux/quotaops.h @@ -7,7 +7,6 @@ #ifndef _LINUX_QUOTAOPS_ #define _LINUX_QUOTAOPS_ -#include #include static inline struct quota_info *sb_dqopt(struct super_block *sb) diff --git a/include/linux/sunrpc/xdr.h b/include/linux/sunrpc/xdr.h index d8910b68e1bd..b99c625fddfe 100644 --- a/include/linux/sunrpc/xdr.h +++ b/include/linux/sunrpc/xdr.h @@ -12,7 +12,6 @@ #include #include #include -#include /* * Buffer adjustment diff --git a/kernel/power/user.c b/kernel/power/user.c index ed97375daae9..bf0014d6a5f0 100644 --- a/kernel/power/user.c +++ b/kernel/power/user.c @@ -23,7 +23,6 @@ #include #include #include -#include #include #include diff --git a/kernel/trace/blktrace.c b/kernel/trace/blktrace.c index 39af8af6fc30..1090b0aed9ba 100644 --- a/kernel/trace/blktrace.c +++ b/kernel/trace/blktrace.c @@ -22,6 +22,7 @@ #include #include #include +#include #include #include diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 3aa0a0dfdfa8..8bc8d8afea6a 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -17,6 +17,7 @@ #include #include #include +#include #include #include #include diff --git a/net/appletalk/ddp.c b/net/appletalk/ddp.c index 590b83963622..bfbe13786bb4 100644 --- a/net/appletalk/ddp.c +++ b/net/appletalk/ddp.c @@ -54,6 +54,7 @@ #include #include #include +#include #include /* For TIOCOUTQ/INQ */ #include #include diff --git a/net/ipx/af_ipx.c b/net/ipx/af_ipx.c index 417b0e309495..f1118d92a191 100644 --- a/net/ipx/af_ipx.c +++ b/net/ipx/af_ipx.c @@ -41,6 +41,7 @@ #include #include #include +#include #include #include #include diff --git a/net/irda/af_irda.c b/net/irda/af_irda.c index cb762c8723ea..80cf29aae096 100644 --- a/net/irda/af_irda.c +++ b/net/irda/af_irda.c @@ -45,6 +45,7 @@ #include #include #include +#include #include #include #include diff --git a/net/irda/irnet/irnet.h b/net/irda/irnet/irnet.h index bccf4d0059f0..b001c361ad30 100644 --- a/net/irda/irnet/irnet.h +++ b/net/irda/irnet/irnet.h @@ -241,7 +241,6 @@ #include #include -#include #include #include #include diff --git a/net/irda/irnet/irnet_ppp.c b/net/irda/irnet/irnet_ppp.c index 6d8ae03c14f5..68cbcb19cbd8 100644 --- a/net/irda/irnet/irnet_ppp.c +++ b/net/irda/irnet/irnet_ppp.c @@ -13,6 +13,7 @@ * 2) as a control channel (write commands, read events) */ +#include #include "irnet_ppp.h" /* Private header */ /* Please put other headers in irnet.h - Thanks */ diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index 5bc2f45bddf0..ebfcf9b89909 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -28,7 +28,6 @@ #include #include #include -#include #include #include #include diff --git a/net/sunrpc/sched.c b/net/sunrpc/sched.c index 1102ce1251f7..8f459abe97cf 100644 --- a/net/sunrpc/sched.c +++ b/net/sunrpc/sched.c @@ -16,7 +16,6 @@ #include #include #include -#include #include #include diff --git a/net/sunrpc/svc_xprt.c b/net/sunrpc/svc_xprt.c index 6f33d33cc064..27d44332f017 100644 --- a/net/sunrpc/svc_xprt.c +++ b/net/sunrpc/svc_xprt.c @@ -5,6 +5,7 @@ */ #include +#include #include #include #include diff --git a/net/wanrouter/wanmain.c b/net/wanrouter/wanmain.c index 466e2d22d256..258daa80ad92 100644 --- a/net/wanrouter/wanmain.c +++ b/net/wanrouter/wanmain.c @@ -48,6 +48,7 @@ #include #include /* support for loadable modules */ #include /* kmalloc(), kfree() */ +#include #include #include /* inline mem*, str* functions */ diff --git a/net/x25/af_x25.c b/net/x25/af_x25.c index 21cdc872004e..5e6c072c64d3 100644 --- a/net/x25/af_x25.c +++ b/net/x25/af_x25.c @@ -40,6 +40,7 @@ #include #include #include +#include #include #include #include -- cgit v1.2.3 From dd0d9a46f573b086a67522f819566427dba9c4c7 Mon Sep 17 00:00:00 2001 From: Artem Bityutskiy Date: Thu, 9 Jul 2009 10:44:30 +0100 Subject: AFS: Fix compilation warning Fix the following warning: fs/afs/dir.c: In function 'afs_d_revalidate': fs/afs/dir.c:567: warning: 'fid.vnode' may be used uninitialized in this function fs/afs/dir.c:567: warning: 'fid.unique' may be used uninitialized in this function by marking the 'fid' variable as an uninitialized_var. The problem is that gcc doesn't always manage to work out that fid is always set on the path through the function that uses it. Cc: linux-afs@lists.infradead.org Cc: linux-kernel@vger.kernel.org Signed-off-by: Artem Bityutskiy Signed-off-by: David Howells Signed-off-by: Linus Torvalds --- fs/afs/dir.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/afs/dir.c b/fs/afs/dir.c index 9bd757774c9e..88067f36e5e7 100644 --- a/fs/afs/dir.c +++ b/fs/afs/dir.c @@ -564,7 +564,7 @@ static struct dentry *afs_lookup(struct inode *dir, struct dentry *dentry, static int afs_d_revalidate(struct dentry *dentry, struct nameidata *nd) { struct afs_vnode *vnode, *dir; - struct afs_fid fid; + struct afs_fid uninitialized_var(fid); struct dentry *parent; struct key *key; void *dir_version; -- cgit v1.2.3 From f8c73c790c588fd70fda1632c8927a87b3d31dcd Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Thu, 11 Jun 2009 15:14:40 +0200 Subject: partitions: fix broken uevent_suppress conversion git commit f67f129e "Driver core: implement uevent suppress in kobject" contains this chunk for fs/partitions/check.c: /* suppress uevent if the disk supresses it */ - if (!ddev->uevent_suppress) + if (!dev_get_uevent_suppress(pdev)) kobject_uevent(&pdev->kobj, KOBJ_ADD); However that should have been - if (!ddev->uevent_suppress) + if (!dev_get_uevent_suppress(ddev)) Signed-off-by: Heiko Carstens Acked-by: Ming Lei Cc: stable Signed-off-by: Greg Kroah-Hartman --- fs/partitions/check.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/partitions/check.c b/fs/partitions/check.c index 1a9c7878f864..ea4e6cb29e13 100644 --- a/fs/partitions/check.c +++ b/fs/partitions/check.c @@ -436,7 +436,7 @@ struct hd_struct *add_partition(struct gendisk *disk, int partno, rcu_assign_pointer(ptbl->part[partno], p); /* suppress uevent if the disk supresses it */ - if (!dev_get_uevent_suppress(pdev)) + if (!dev_get_uevent_suppress(ddev)) kobject_uevent(&pdev->kobj, KOBJ_ADD); return p; -- cgit v1.2.3 From d0b6e04a4cd8360e3c9c419f7c30a3081a0c142a Mon Sep 17 00:00:00 2001 From: Li Zefan Date: Mon, 13 Jul 2009 10:33:21 +0800 Subject: tracing/events: Move TRACE_SYSTEM outside of include guard If TRACE_INCLDUE_FILE is defined, will be included and compiled, otherwise it will be So TRACE_SYSTEM should be defined outside of #if proctection, just like TRACE_INCLUDE_FILE. Imaging this scenario: #include -> TRACE_SYSTEM == foo ... #include -> TRACE_SYSTEM == bar ... #define CREATE_TRACE_POINTS #include -> TRACE_SYSTEM == bar !!! and then bar.h will be included and compiled. Signed-off-by: Li Zefan Cc: Steven Rostedt Cc: Frederic Weisbecker LKML-Reference: <4A5A9CF1.2010007@cn.fujitsu.com> Signed-off-by: Ingo Molnar --- fs/gfs2/trace_gfs2.h | 8 +++---- include/trace/events/block.h | 6 ++--- include/trace/events/ext4.h | 6 ++--- include/trace/events/irq.h | 6 ++--- include/trace/events/jbd2.h | 6 ++--- include/trace/events/kmem.h | 6 ++--- include/trace/events/lockdep.h | 6 ++--- include/trace/events/sched.h | 6 ++--- include/trace/events/skb.h | 6 ++--- include/trace/events/workqueue.h | 6 ++--- samples/trace_events/trace-events-sample.h | 37 ++++++++++++++++-------------- 11 files changed, 51 insertions(+), 48 deletions(-) (limited to 'fs') diff --git a/fs/gfs2/trace_gfs2.h b/fs/gfs2/trace_gfs2.h index 98d6ef1c1dc0..148d55c14171 100644 --- a/fs/gfs2/trace_gfs2.h +++ b/fs/gfs2/trace_gfs2.h @@ -1,12 +1,11 @@ +#undef TRACE_SYSTEM +#define TRACE_SYSTEM gfs2 + #if !defined(_TRACE_GFS2_H) || defined(TRACE_HEADER_MULTI_READ) #define _TRACE_GFS2_H #include -#undef TRACE_SYSTEM -#define TRACE_SYSTEM gfs2 -#define TRACE_INCLUDE_FILE trace_gfs2 - #include #include #include @@ -403,5 +402,6 @@ TRACE_EVENT(gfs2_block_alloc, /* This part must be outside protection */ #undef TRACE_INCLUDE_PATH #define TRACE_INCLUDE_PATH . +#define TRACE_INCLUDE_FILE trace_gfs2 #include diff --git a/include/trace/events/block.h b/include/trace/events/block.h index d6b05f42dd44..9a74b468a229 100644 --- a/include/trace/events/block.h +++ b/include/trace/events/block.h @@ -1,3 +1,6 @@ +#undef TRACE_SYSTEM +#define TRACE_SYSTEM block + #if !defined(_TRACE_BLOCK_H) || defined(TRACE_HEADER_MULTI_READ) #define _TRACE_BLOCK_H @@ -5,9 +8,6 @@ #include #include -#undef TRACE_SYSTEM -#define TRACE_SYSTEM block - TRACE_EVENT(block_rq_abort, TP_PROTO(struct request_queue *q, struct request *rq), diff --git a/include/trace/events/ext4.h b/include/trace/events/ext4.h index acf4cc9cd36d..8eb82cc8d149 100644 --- a/include/trace/events/ext4.h +++ b/include/trace/events/ext4.h @@ -1,9 +1,9 @@ -#if !defined(_TRACE_EXT4_H) || defined(TRACE_HEADER_MULTI_READ) -#define _TRACE_EXT4_H - #undef TRACE_SYSTEM #define TRACE_SYSTEM ext4 +#if !defined(_TRACE_EXT4_H) || defined(TRACE_HEADER_MULTI_READ) +#define _TRACE_EXT4_H + #include #include "../../../fs/ext4/ext4.h" #include "../../../fs/ext4/mballoc.h" diff --git a/include/trace/events/irq.h b/include/trace/events/irq.h index b0c7ede55eb1..1cb0c3aa11e6 100644 --- a/include/trace/events/irq.h +++ b/include/trace/events/irq.h @@ -1,12 +1,12 @@ +#undef TRACE_SYSTEM +#define TRACE_SYSTEM irq + #if !defined(_TRACE_IRQ_H) || defined(TRACE_HEADER_MULTI_READ) #define _TRACE_IRQ_H #include #include -#undef TRACE_SYSTEM -#define TRACE_SYSTEM irq - #define softirq_name(sirq) { sirq##_SOFTIRQ, #sirq } #define show_softirq_name(val) \ __print_symbolic(val, \ diff --git a/include/trace/events/jbd2.h b/include/trace/events/jbd2.h index 845b0b4b48fd..10813fa0c8d0 100644 --- a/include/trace/events/jbd2.h +++ b/include/trace/events/jbd2.h @@ -1,12 +1,12 @@ +#undef TRACE_SYSTEM +#define TRACE_SYSTEM jbd2 + #if !defined(_TRACE_JBD2_H) || defined(TRACE_HEADER_MULTI_READ) #define _TRACE_JBD2_H #include #include -#undef TRACE_SYSTEM -#define TRACE_SYSTEM jbd2 - TRACE_EVENT(jbd2_checkpoint, TP_PROTO(journal_t *journal, int result), diff --git a/include/trace/events/kmem.h b/include/trace/events/kmem.h index 9baba50d6512..1493c541f9c4 100644 --- a/include/trace/events/kmem.h +++ b/include/trace/events/kmem.h @@ -1,12 +1,12 @@ +#undef TRACE_SYSTEM +#define TRACE_SYSTEM kmem + #if !defined(_TRACE_KMEM_H) || defined(TRACE_HEADER_MULTI_READ) #define _TRACE_KMEM_H #include #include -#undef TRACE_SYSTEM -#define TRACE_SYSTEM kmem - /* * The order of these masks is important. Matching masks will be seen * first and the left over flags will end up showing by themselves. diff --git a/include/trace/events/lockdep.h b/include/trace/events/lockdep.h index 0e956c9dfd7e..bcf1d209a00d 100644 --- a/include/trace/events/lockdep.h +++ b/include/trace/events/lockdep.h @@ -1,12 +1,12 @@ +#undef TRACE_SYSTEM +#define TRACE_SYSTEM lockdep + #if !defined(_TRACE_LOCKDEP_H) || defined(TRACE_HEADER_MULTI_READ) #define _TRACE_LOCKDEP_H #include #include -#undef TRACE_SYSTEM -#define TRACE_SYSTEM lockdep - #ifdef CONFIG_LOCKDEP TRACE_EVENT(lock_acquire, diff --git a/include/trace/events/sched.h b/include/trace/events/sched.h index 24ab5bcff7b2..8949bb7eb082 100644 --- a/include/trace/events/sched.h +++ b/include/trace/events/sched.h @@ -1,12 +1,12 @@ +#undef TRACE_SYSTEM +#define TRACE_SYSTEM sched + #if !defined(_TRACE_SCHED_H) || defined(TRACE_HEADER_MULTI_READ) #define _TRACE_SCHED_H #include #include -#undef TRACE_SYSTEM -#define TRACE_SYSTEM sched - /* * Tracepoint for calling kthread_stop, performed to end a kthread: */ diff --git a/include/trace/events/skb.h b/include/trace/events/skb.h index 1e8fabb57c06..e499863b9669 100644 --- a/include/trace/events/skb.h +++ b/include/trace/events/skb.h @@ -1,12 +1,12 @@ +#undef TRACE_SYSTEM +#define TRACE_SYSTEM skb + #if !defined(_TRACE_SKB_H) || defined(TRACE_HEADER_MULTI_READ) #define _TRACE_SKB_H #include #include -#undef TRACE_SYSTEM -#define TRACE_SYSTEM skb - /* * Tracepoint for free an sk_buff: */ diff --git a/include/trace/events/workqueue.h b/include/trace/events/workqueue.h index 035f1bff288e..fcfd9a1e4b96 100644 --- a/include/trace/events/workqueue.h +++ b/include/trace/events/workqueue.h @@ -1,3 +1,6 @@ +#undef TRACE_SYSTEM +#define TRACE_SYSTEM workqueue + #if !defined(_TRACE_WORKQUEUE_H) || defined(TRACE_HEADER_MULTI_READ) #define _TRACE_WORKQUEUE_H @@ -5,9 +8,6 @@ #include #include -#undef TRACE_SYSTEM -#define TRACE_SYSTEM workqueue - TRACE_EVENT(workqueue_insertion, TP_PROTO(struct task_struct *wq_thread, struct work_struct *work), diff --git a/samples/trace_events/trace-events-sample.h b/samples/trace_events/trace-events-sample.h index 9977a756fb32..f24ae370e514 100644 --- a/samples/trace_events/trace-events-sample.h +++ b/samples/trace_events/trace-events-sample.h @@ -1,20 +1,3 @@ -/* - * Notice that this file is not protected like a normal header. - * We also must allow for rereading of this file. The - * - * || defined(TRACE_HEADER_MULTI_READ) - * - * serves this purpose. - */ -#if !defined(_TRACE_EVENT_SAMPLE_H) || defined(TRACE_HEADER_MULTI_READ) -#define _TRACE_EVENT_SAMPLE_H - -/* - * All trace headers should include tracepoint.h, until we finally - * make it into a standard header. - */ -#include - /* * If TRACE_SYSTEM is defined, that will be the directory created * in the ftrace directory under /debugfs/tracing/events/ @@ -34,10 +17,30 @@ * #define TRACE_INCLUDE_FILE trace-events-sample * * As we do an the bottom of this file. + * + * Notice that TRACE_SYSTEM should be defined outside of #if + * protection, just like TRACE_INCLUDE_FILE. */ #undef TRACE_SYSTEM #define TRACE_SYSTEM sample +/* + * Notice that this file is not protected like a normal header. + * We also must allow for rereading of this file. The + * + * || defined(TRACE_HEADER_MULTI_READ) + * + * serves this purpose. + */ +#if !defined(_TRACE_EVENT_SAMPLE_H) || defined(TRACE_HEADER_MULTI_READ) +#define _TRACE_EVENT_SAMPLE_H + +/* + * All trace headers should include tracepoint.h, until we finally + * make it into a standard header. + */ +#include + /* * The TRACE_EVENT macro is broken up into 5 parts. * -- cgit v1.2.3 From e6b5d30104db5f34110678ecab14988f1f1eff63 Mon Sep 17 00:00:00 2001 From: Curt Wohlgemuth Date: Mon, 13 Jul 2009 09:07:20 -0400 Subject: ext4: Fix buffer head reference leak in no-journal mode We found a problem with buffer head reference leaks when using an ext4 partition without a journal. In particular, calls to ext4_forget() would not to a brelse() on the input buffer head, which will cause pages they belong to to not be reclaimable. Further investigation showed that all places where ext4_journal_forget() and ext4_journal_revoke() are called are subject to the same problem. The patch below changes __ext4_journal_forget/__ext4_journal_revoke to do an explicit release of the buffer head when the journal handle isn't valid. Signed-off-by: Curt Wohlgemuth Signed-off-by: "Theodore Ts'o" --- fs/ext4/ext4_jbd2.c | 4 ++++ fs/ext4/ext4_jbd2.h | 2 ++ fs/ext4/inode.c | 6 ++---- 3 files changed, 8 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/ext4/ext4_jbd2.c b/fs/ext4/ext4_jbd2.c index ad13a84644e1..eb27fd0f2ee8 100644 --- a/fs/ext4/ext4_jbd2.c +++ b/fs/ext4/ext4_jbd2.c @@ -43,6 +43,8 @@ int __ext4_journal_forget(const char *where, handle_t *handle, ext4_journal_abort_handle(where, __func__, bh, handle, err); } + else + brelse(bh); return err; } @@ -57,6 +59,8 @@ int __ext4_journal_revoke(const char *where, handle_t *handle, ext4_journal_abort_handle(where, __func__, bh, handle, err); } + else + brelse(bh); return err; } diff --git a/fs/ext4/ext4_jbd2.h b/fs/ext4/ext4_jbd2.h index d574a85aca56..139fb8cb87e4 100644 --- a/fs/ext4/ext4_jbd2.h +++ b/fs/ext4/ext4_jbd2.h @@ -131,9 +131,11 @@ int __ext4_journal_get_undo_access(const char *where, handle_t *handle, int __ext4_journal_get_write_access(const char *where, handle_t *handle, struct buffer_head *bh); +/* When called with an invalid handle, this will still do a put on the BH */ int __ext4_journal_forget(const char *where, handle_t *handle, struct buffer_head *bh); +/* When called with an invalid handle, this will still do a put on the BH */ int __ext4_journal_revoke(const char *where, handle_t *handle, ext4_fsblk_t blocknr, struct buffer_head *bh); diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index c98e3afea30a..f9c642b22efa 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -78,16 +78,14 @@ static int ext4_inode_is_fast_symlink(struct inode *inode) * but there may still be a record of it in the journal, and that record * still needs to be revoked. * - * If the handle isn't valid we're not journaling so there's nothing to do. + * If the handle isn't valid we're not journaling, but we still need to + * call into ext4_journal_revoke() to put the buffer head. */ int ext4_forget(handle_t *handle, int is_metadata, struct inode *inode, struct buffer_head *bh, ext4_fsblk_t blocknr) { int err; - if (!ext4_handle_valid(handle)) - return 0; - might_sleep(); BUFFER_TRACE(bh, "enter"); -- cgit v1.2.3 From ac046f1d6121ccdda6db66bd88acd52418f489b2 Mon Sep 17 00:00:00 2001 From: Peng Tao Date: Mon, 13 Jul 2009 09:30:17 -0400 Subject: ext4: fix null handler of ioctls in no journal mode The EXT4_IOC_GROUP_ADD and EXT4_IOC_GROUP_EXTEND ioctls should not flush the journal in no_journal mode. Otherwise, running resize2fs on a mounted no_journal partition triggers the following error messages: BUG: unable to handle kernel NULL pointer dereference at 00000014 IP: [] _spin_lock+0x8/0x19 *pde = 00000000 Oops: 0002 [#1] SMP Signed-off-by: Peng Tao Signed-off-by: "Theodore Ts'o" --- fs/ext4/ioctl.c | 20 ++++++++++++-------- 1 file changed, 12 insertions(+), 8 deletions(-) (limited to 'fs') diff --git a/fs/ext4/ioctl.c b/fs/ext4/ioctl.c index bb415408fdb6..01f149aea841 100644 --- a/fs/ext4/ioctl.c +++ b/fs/ext4/ioctl.c @@ -192,7 +192,7 @@ setversion_out: case EXT4_IOC_GROUP_EXTEND: { ext4_fsblk_t n_blocks_count; struct super_block *sb = inode->i_sb; - int err, err2; + int err, err2=0; if (!capable(CAP_SYS_RESOURCE)) return -EPERM; @@ -205,9 +205,11 @@ setversion_out: return err; err = ext4_group_extend(sb, EXT4_SB(sb)->s_es, n_blocks_count); - jbd2_journal_lock_updates(EXT4_SB(sb)->s_journal); - err2 = jbd2_journal_flush(EXT4_SB(sb)->s_journal); - jbd2_journal_unlock_updates(EXT4_SB(sb)->s_journal); + if (EXT4_SB(sb)->s_journal) { + jbd2_journal_lock_updates(EXT4_SB(sb)->s_journal); + err2 = jbd2_journal_flush(EXT4_SB(sb)->s_journal); + jbd2_journal_unlock_updates(EXT4_SB(sb)->s_journal); + } if (err == 0) err = err2; mnt_drop_write(filp->f_path.mnt); @@ -252,7 +254,7 @@ setversion_out: case EXT4_IOC_GROUP_ADD: { struct ext4_new_group_data input; struct super_block *sb = inode->i_sb; - int err, err2; + int err, err2=0; if (!capable(CAP_SYS_RESOURCE)) return -EPERM; @@ -266,9 +268,11 @@ setversion_out: return err; err = ext4_group_add(sb, &input); - jbd2_journal_lock_updates(EXT4_SB(sb)->s_journal); - err2 = jbd2_journal_flush(EXT4_SB(sb)->s_journal); - jbd2_journal_unlock_updates(EXT4_SB(sb)->s_journal); + if (EXT4_SB(sb)->s_journal) { + jbd2_journal_lock_updates(EXT4_SB(sb)->s_journal); + err2 = jbd2_journal_flush(EXT4_SB(sb)->s_journal); + jbd2_journal_unlock_updates(EXT4_SB(sb)->s_journal); + } if (err == 0) err = err2; mnt_drop_write(filp->f_path.mnt); -- cgit v1.2.3 From 833576b362e15c38be3bfe43942cda693e56287c Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Mon, 13 Jul 2009 09:45:52 -0400 Subject: ext4: Fix ext4_mb_initialize_context() to initialize all fields Pavel Roskin pointed out that kmemcheck indicated that ext4_mb_store_history() was accessing uninitialized values of ac->ac_tail and ac->ac_buddy leading to garbage in the mballoc history. Fix this by initializing the entire structure to all zeros first. Also, two fields were getting doubly initialized by the caller of ext4_mb_initialize_context, so remove them for efficiency's sake. Signed-off-by: "Theodore Ts'o" --- fs/ext4/mballoc.c | 20 ++------------------ 1 file changed, 2 insertions(+), 18 deletions(-) (limited to 'fs') diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index 2fcaf286f1de..cd258463e2a9 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -4227,14 +4227,9 @@ ext4_mb_initialize_context(struct ext4_allocation_context *ac, ext4_get_group_no_and_offset(sb, goal, &group, &block); /* set up allocation goals */ + memset(ac, 0, sizeof(struct ext4_allocation_context)); ac->ac_b_ex.fe_logical = ar->logical; - ac->ac_b_ex.fe_group = 0; - ac->ac_b_ex.fe_start = 0; - ac->ac_b_ex.fe_len = 0; ac->ac_status = AC_STATUS_CONTINUE; - ac->ac_groups_scanned = 0; - ac->ac_ex_scanned = 0; - ac->ac_found = 0; ac->ac_sb = sb; ac->ac_inode = ar->inode; ac->ac_o_ex.fe_logical = ar->logical; @@ -4245,15 +4240,7 @@ ext4_mb_initialize_context(struct ext4_allocation_context *ac, ac->ac_g_ex.fe_group = group; ac->ac_g_ex.fe_start = block; ac->ac_g_ex.fe_len = len; - ac->ac_f_ex.fe_len = 0; ac->ac_flags = ar->flags; - ac->ac_2order = 0; - ac->ac_criteria = 0; - ac->ac_pa = NULL; - ac->ac_bitmap_page = NULL; - ac->ac_buddy_page = NULL; - ac->alloc_semp = NULL; - ac->ac_lg = NULL; /* we have to define context: we'll we work with a file or * locality group. this is a policy, actually */ @@ -4521,10 +4508,7 @@ ext4_fsblk_t ext4_mb_new_blocks(handle_t *handle, } ac = kmem_cache_alloc(ext4_ac_cachep, GFP_NOFS); - if (ac) { - ac->ac_sb = sb; - ac->ac_inode = ar->inode; - } else { + if (!ac) { ar->len = 0; *errp = -ENOMEM; goto out1; -- cgit v1.2.3 From 96577c43827697ca1af5982fa256a34786d0c720 Mon Sep 17 00:00:00 2001 From: dingdinghua Date: Mon, 13 Jul 2009 17:55:35 -0400 Subject: jbd2: fix race between write_metadata_buffer and get_write_access The function jbd2_journal_write_metadata_buffer() calls jbd_unlock_bh_state(bh_in) too early; this could potentially allow another thread to call get_write_access on the buffer head, modify the data, and dirty it, and allowing the wrong data to be written into the journal. Fortunately, if we lose this race, the only time this will actually cause filesystem corruption is if there is a system crash or other unclean shutdown of the system before the next commit can take place. Signed-off-by: dingdinghua Signed-off-by: "Theodore Ts'o" --- fs/jbd2/journal.c | 20 +++++++++++--------- 1 file changed, 11 insertions(+), 9 deletions(-) (limited to 'fs') diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c index 7b545c3b3942..e378cb383979 100644 --- a/fs/jbd2/journal.c +++ b/fs/jbd2/journal.c @@ -297,6 +297,7 @@ int jbd2_journal_write_metadata_buffer(transaction_t *transaction, unsigned int new_offset; struct buffer_head *bh_in = jh2bh(jh_in); struct jbd2_buffer_trigger_type *triggers; + journal_t *journal = transaction->t_journal; /* * The buffer really shouldn't be locked: only the current committing @@ -310,6 +311,11 @@ int jbd2_journal_write_metadata_buffer(transaction_t *transaction, J_ASSERT_BH(bh_in, buffer_jbddirty(bh_in)); new_bh = alloc_buffer_head(GFP_NOFS|__GFP_NOFAIL); + /* keep subsequent assertions sane */ + new_bh->b_state = 0; + init_buffer(new_bh, NULL, NULL); + atomic_set(&new_bh->b_count, 1); + new_jh = jbd2_journal_add_journal_head(new_bh); /* This sleeps */ /* * If a new transaction has already done a buffer copy-out, then @@ -388,14 +394,6 @@ repeat: kunmap_atomic(mapped_data, KM_USER0); } - /* keep subsequent assertions sane */ - new_bh->b_state = 0; - init_buffer(new_bh, NULL, NULL); - atomic_set(&new_bh->b_count, 1); - jbd_unlock_bh_state(bh_in); - - new_jh = jbd2_journal_add_journal_head(new_bh); /* This sleeps */ - set_bh_page(new_bh, new_page, new_offset); new_jh->b_transaction = NULL; new_bh->b_size = jh2bh(jh_in)->b_size; @@ -412,7 +410,11 @@ repeat: * copying is moved to the transaction's shadow queue. */ JBUFFER_TRACE(jh_in, "file as BJ_Shadow"); - jbd2_journal_file_buffer(jh_in, transaction, BJ_Shadow); + spin_lock(&journal->j_list_lock); + __jbd2_journal_file_buffer(jh_in, transaction, BJ_Shadow); + spin_unlock(&journal->j_list_lock); + jbd_unlock_bh_state(bh_in); + JBUFFER_TRACE(new_jh, "file as BJ_IO"); jbd2_journal_file_buffer(new_jh, transaction, BJ_IO); -- cgit v1.2.3 From 4fed598a49c014cbc563179b25f2a4b8565e2a50 Mon Sep 17 00:00:00 2001 From: Ryusuke Konishi Date: Sun, 12 Jul 2009 11:13:55 +0900 Subject: fs/Kconfig: move nilfs2 out fs/Kconfig file was split into individual fs/*/Kconfig files before nilfs was merged. I've found the current config entry of nilfs is tainting the work. Sorry, I didn't notice. This fixes the violation. Signed-off-by: Ryusuke Konishi Cc: Alexey Dobriyan --- fs/Kconfig | 27 +-------------------------- fs/nilfs2/Kconfig | 25 +++++++++++++++++++++++++ 2 files changed, 26 insertions(+), 26 deletions(-) create mode 100644 fs/nilfs2/Kconfig (limited to 'fs') diff --git a/fs/Kconfig b/fs/Kconfig index a97263be6a91..0e7da7bb5d93 100644 --- a/fs/Kconfig +++ b/fs/Kconfig @@ -186,32 +186,7 @@ source "fs/romfs/Kconfig" source "fs/sysv/Kconfig" source "fs/ufs/Kconfig" source "fs/exofs/Kconfig" - -config NILFS2_FS - tristate "NILFS2 file system support (EXPERIMENTAL)" - depends on BLOCK && EXPERIMENTAL - select CRC32 - help - NILFS2 is a log-structured file system (LFS) supporting continuous - snapshotting. In addition to versioning capability of the entire - file system, users can even restore files mistakenly overwritten or - destroyed just a few seconds ago. Since this file system can keep - consistency like conventional LFS, it achieves quick recovery after - system crashes. - - NILFS2 creates a number of checkpoints every few seconds or per - synchronous write basis (unless there is no change). Users can - select significant versions among continuously created checkpoints, - and can change them into snapshots which will be preserved for long - periods until they are changed back to checkpoints. Each - snapshot is mountable as a read-only file system concurrently with - its writable mount, and this feature is convenient for online backup. - - Some features including atime, extended attributes, and POSIX ACLs, - are not supported yet. - - To compile this file system support as a module, choose M here: the - module will be called nilfs2. If unsure, say N. +source "fs/nilfs2/Kconfig" endif # MISC_FILESYSTEMS diff --git a/fs/nilfs2/Kconfig b/fs/nilfs2/Kconfig new file mode 100644 index 000000000000..72da095d4009 --- /dev/null +++ b/fs/nilfs2/Kconfig @@ -0,0 +1,25 @@ +config NILFS2_FS + tristate "NILFS2 file system support (EXPERIMENTAL)" + depends on BLOCK && EXPERIMENTAL + select CRC32 + help + NILFS2 is a log-structured file system (LFS) supporting continuous + snapshotting. In addition to versioning capability of the entire + file system, users can even restore files mistakenly overwritten or + destroyed just a few seconds ago. Since this file system can keep + consistency like conventional LFS, it achieves quick recovery after + system crashes. + + NILFS2 creates a number of checkpoints every few seconds or per + synchronous write basis (unless there is no change). Users can + select significant versions among continuously created checkpoints, + and can change them into snapshots which will be preserved for long + periods until they are changed back to checkpoints. Each + snapshot is mountable as a read-only file system concurrently with + its writable mount, and this feature is convenient for online backup. + + Some features including atime, extended attributes, and POSIX ACLs, + are not supported yet. + + To compile this file system support as a module, choose M here: the + module will be called nilfs2. If unsure, say N. -- cgit v1.2.3 From a89d63a159b1ba5833be2bef00adf8ad8caac8be Mon Sep 17 00:00:00 2001 From: Casey Dahlin Date: Tue, 14 Jul 2009 12:17:51 -0500 Subject: dlm: free socket in error exit path In the tcp_connect_to_sock() error exit path, the socket allocated at the top of the function was not being freed. Signed-off-by: Casey Dahlin Signed-off-by: David Teigland --- fs/dlm/lowcomms.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/dlm/lowcomms.c b/fs/dlm/lowcomms.c index cdb580a9c7a2..618a60f03886 100644 --- a/fs/dlm/lowcomms.c +++ b/fs/dlm/lowcomms.c @@ -902,7 +902,7 @@ static void tcp_connect_to_sock(struct connection *con) int result = -EHOSTUNREACH; struct sockaddr_storage saddr, src_addr; int addr_len; - struct socket *sock; + struct socket *sock = NULL; if (con->nodeid == 0) { log_print("attempt to connect sock 0 foiled"); @@ -962,6 +962,8 @@ out_err: if (con->sock) { sock_release(con->sock); con->sock = NULL; + } else if (sock) { + sock_release(sock); } /* * Some errors are fatal and this list might need adjusting. For other -- cgit v1.2.3 From 9c9ad6162e2aa1e528ed687ccab87fe681ebbef1 Mon Sep 17 00:00:00 2001 From: Abhishek Kulkarni Date: Tue, 14 Jul 2009 13:26:52 -0500 Subject: 9p: Fix incorrect parameters to v9fs_file_readn. Fix v9fs_vfs_readpage. The offset and size parameters to v9fs_file_readn were interchanged and hence passed incorrectly. Signed-off-by: Abhishek Kulkarni Signed-off-by: Eric Van Hensbergen --- fs/9p/vfs_addr.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/9p/vfs_addr.c b/fs/9p/vfs_addr.c index 6fcb1e7095cf..92828281a30b 100644 --- a/fs/9p/vfs_addr.c +++ b/fs/9p/vfs_addr.c @@ -57,7 +57,7 @@ static int v9fs_vfs_readpage(struct file *filp, struct page *page) buffer = kmap(page); offset = page_offset(page); - retval = v9fs_file_readn(filp, buffer, NULL, offset, PAGE_CACHE_SIZE); + retval = v9fs_file_readn(filp, buffer, NULL, PAGE_CACHE_SIZE, offset); if (retval < 0) goto done; -- cgit v1.2.3 From 7447a668a3860b66b3c9db86fdea91e355ba59ac Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Wed, 15 Jul 2009 20:36:08 +0200 Subject: jbd: Fail to load a journal if it is too short Due to on disk corruption, it can happen that journal is too short. Fail to load it in such case so that we don't oops somewhere later. Reported-by: Nageswara R Sastry Signed-off-by: Jan Kara --- fs/jbd/journal.c | 6 ++++++ 1 file changed, 6 insertions(+) (limited to 'fs') diff --git a/fs/jbd/journal.c b/fs/jbd/journal.c index 737f7246a4b5..94a64a199a63 100644 --- a/fs/jbd/journal.c +++ b/fs/jbd/journal.c @@ -848,6 +848,12 @@ static int journal_reset(journal_t *journal) first = be32_to_cpu(sb->s_first); last = be32_to_cpu(sb->s_maxlen); + if (first + JFS_MIN_JOURNAL_BLOCKS > last + 1) { + printk(KERN_ERR "JBD: Journal too short (blocks %lu-%lu).\n", + first, last); + journal_fail_superblock(journal); + return -EINVAL; + } journal->j_first = first; journal->j_last = last; -- cgit v1.2.3 From 9eaaa2d5759837402ec5eee13b2a97921808c3eb Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Mon, 13 Jul 2009 20:26:52 +0200 Subject: ext3: Fix truncation of symlinks after failed write Contents of long symlinks is written via standard write methods. So when the write fails, we add inode to orphan list. But symlinks don't have .truncate method defined so nobody properly removes them from the orphan list (both on disk and in memory). Fix this by calling ext3_truncate() directly instead of calling vmtruncate() (which is saner anyway since we don't need anything vmtruncate() does except from calling .truncate in these paths). We also add inode to orphan list only if ext3_can_truncate() is true (currently, it can be false for symlinks when there are no blocks allocated) - otherwise orphan list processing will complain and ext3_truncate() will not remove inode from on-disk orphan list. Signed-off-by: Jan Kara --- fs/ext3/inode.c | 19 ++++++++++--------- 1 file changed, 10 insertions(+), 9 deletions(-) (limited to 'fs') diff --git a/fs/ext3/inode.c b/fs/ext3/inode.c index 5f51fed5c750..4d7da6f6184b 100644 --- a/fs/ext3/inode.c +++ b/fs/ext3/inode.c @@ -1193,15 +1193,16 @@ write_begin_failed: * i_size_read because we hold i_mutex. * * Add inode to orphan list in case we crash before truncate - * finishes. + * finishes. Do this only if ext3_can_truncate() agrees so + * that orphan processing code is happy. */ - if (pos + len > inode->i_size) + if (pos + len > inode->i_size && ext3_can_truncate(inode)) ext3_orphan_add(handle, inode); ext3_journal_stop(handle); unlock_page(page); page_cache_release(page); if (pos + len > inode->i_size) - vmtruncate(inode, inode->i_size); + ext3_truncate(inode); } if (ret == -ENOSPC && ext3_should_retry_alloc(inode->i_sb, &retries)) goto retry; @@ -1287,7 +1288,7 @@ static int ext3_ordered_write_end(struct file *file, * There may be allocated blocks outside of i_size because * we failed to copy some data. Prepare for truncate. */ - if (pos + len > inode->i_size) + if (pos + len > inode->i_size && ext3_can_truncate(inode)) ext3_orphan_add(handle, inode); ret2 = ext3_journal_stop(handle); if (!ret) @@ -1296,7 +1297,7 @@ static int ext3_ordered_write_end(struct file *file, page_cache_release(page); if (pos + len > inode->i_size) - vmtruncate(inode, inode->i_size); + ext3_truncate(inode); return ret ? ret : copied; } @@ -1315,14 +1316,14 @@ static int ext3_writeback_write_end(struct file *file, * There may be allocated blocks outside of i_size because * we failed to copy some data. Prepare for truncate. */ - if (pos + len > inode->i_size) + if (pos + len > inode->i_size && ext3_can_truncate(inode)) ext3_orphan_add(handle, inode); ret = ext3_journal_stop(handle); unlock_page(page); page_cache_release(page); if (pos + len > inode->i_size) - vmtruncate(inode, inode->i_size); + ext3_truncate(inode); return ret ? ret : copied; } @@ -1358,7 +1359,7 @@ static int ext3_journalled_write_end(struct file *file, * There may be allocated blocks outside of i_size because * we failed to copy some data. Prepare for truncate. */ - if (pos + len > inode->i_size) + if (pos + len > inode->i_size && ext3_can_truncate(inode)) ext3_orphan_add(handle, inode); EXT3_I(inode)->i_state |= EXT3_STATE_JDATA; if (inode->i_size > EXT3_I(inode)->i_disksize) { @@ -1375,7 +1376,7 @@ static int ext3_journalled_write_end(struct file *file, page_cache_release(page); if (pos + len > inode->i_size) - vmtruncate(inode, inode->i_size); + ext3_truncate(inode); return ret ? ret : copied; } -- cgit v1.2.3 From 1e9fd53b783ea646de3ee09a4574afeb6778d504 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Wed, 24 Jun 2009 17:31:40 +0200 Subject: jbd: Fix a race between checkpointing code and journal_get_write_access() The following race can happen: CPU1 CPU2 checkpointing code checks the buffer, adds it to an array for writeback do_get_write_access() ... lock_buffer() unlock_buffer() flush_batch() submits the buffer for IO __jbd_journal_file_buffer() So a buffer under writeout is returned from do_get_write_access(). Since the filesystem code relies on the fact that journaled buffers cannot be written out, it does not take the buffer lock and so it can modify buffer while it is under writeout. That can lead to a filesystem corruption if we crash at the right moment. The similar problem can happen with the journal_get_create_access() path. We fix the problem by clearing the buffer dirty bit under buffer_lock even if the buffer is on BJ_None list. Actually, we clear the dirty bit regardless the list the buffer is in and warn about the fact if the buffer is already journalled. Thanks for spotting the problem goes to dingdinghua . Reported-by: dingdinghua Signed-off-by: Jan Kara --- fs/jbd/transaction.c | 68 +++++++++++++++++++++++++++------------------------- 1 file changed, 35 insertions(+), 33 deletions(-) (limited to 'fs') diff --git a/fs/jbd/transaction.c b/fs/jbd/transaction.c index 73242ba7c7b1..c03ac11f74be 100644 --- a/fs/jbd/transaction.c +++ b/fs/jbd/transaction.c @@ -489,34 +489,15 @@ void journal_unlock_updates (journal_t *journal) wake_up(&journal->j_wait_transaction_locked); } -/* - * Report any unexpected dirty buffers which turn up. Normally those - * indicate an error, but they can occur if the user is running (say) - * tune2fs to modify the live filesystem, so we need the option of - * continuing as gracefully as possible. # - * - * The caller should already hold the journal lock and - * j_list_lock spinlock: most callers will need those anyway - * in order to probe the buffer's journaling state safely. - */ -static void jbd_unexpected_dirty_buffer(struct journal_head *jh) +static void warn_dirty_buffer(struct buffer_head *bh) { - int jlist; - - /* If this buffer is one which might reasonably be dirty - * --- ie. data, or not part of this journal --- then - * we're OK to leave it alone, but otherwise we need to - * move the dirty bit to the journal's own internal - * JBDDirty bit. */ - jlist = jh->b_jlist; + char b[BDEVNAME_SIZE]; - if (jlist == BJ_Metadata || jlist == BJ_Reserved || - jlist == BJ_Shadow || jlist == BJ_Forget) { - struct buffer_head *bh = jh2bh(jh); - - if (test_clear_buffer_dirty(bh)) - set_buffer_jbddirty(bh); - } + printk(KERN_WARNING + "JBD: Spotted dirty metadata buffer (dev = %s, blocknr = %llu). " + "There's a risk of filesystem corruption in case of system " + "crash.\n", + bdevname(bh->b_bdev, b), (unsigned long long)bh->b_blocknr); } /* @@ -583,14 +564,16 @@ repeat: if (jh->b_next_transaction) J_ASSERT_JH(jh, jh->b_next_transaction == transaction); + warn_dirty_buffer(bh); } /* * In any case we need to clean the dirty flag and we must * do it under the buffer lock to be sure we don't race * with running write-out. */ - JBUFFER_TRACE(jh, "Unexpected dirty buffer"); - jbd_unexpected_dirty_buffer(jh); + JBUFFER_TRACE(jh, "Journalling dirty buffer"); + clear_buffer_dirty(bh); + set_buffer_jbddirty(bh); } unlock_buffer(bh); @@ -826,6 +809,15 @@ int journal_get_create_access(handle_t *handle, struct buffer_head *bh) J_ASSERT_JH(jh, buffer_locked(jh2bh(jh))); if (jh->b_transaction == NULL) { + /* + * Previous journal_forget() could have left the buffer + * with jbddirty bit set because it was being committed. When + * the commit finished, we've filed the buffer for + * checkpointing and marked it dirty. Now we are reallocating + * the buffer so the transaction freeing it must have + * committed and so it's safe to clear the dirty bit. + */ + clear_buffer_dirty(jh2bh(jh)); jh->b_transaction = transaction; /* first access by this transaction */ @@ -1782,8 +1774,13 @@ static int __dispose_buffer(struct journal_head *jh, transaction_t *transaction) if (jh->b_cp_transaction) { JBUFFER_TRACE(jh, "on running+cp transaction"); + /* + * We don't want to write the buffer anymore, clear the + * bit so that we don't confuse checks in + * __journal_file_buffer + */ + clear_buffer_dirty(bh); __journal_file_buffer(jh, transaction, BJ_Forget); - clear_buffer_jbddirty(bh); may_free = 0; } else { JBUFFER_TRACE(jh, "on running transaction"); @@ -2041,12 +2038,17 @@ void __journal_file_buffer(struct journal_head *jh, if (jh->b_transaction && jh->b_jlist == jlist) return; - /* The following list of buffer states needs to be consistent - * with __jbd_unexpected_dirty_buffer()'s handling of dirty - * state. */ - if (jlist == BJ_Metadata || jlist == BJ_Reserved || jlist == BJ_Shadow || jlist == BJ_Forget) { + /* + * For metadata buffers, we track dirty bit in buffer_jbddirty + * instead of buffer_dirty. We should not see a dirty bit set + * here because we clear it in do_get_write_access but e.g. + * tune2fs can modify the sb and set the dirty bit at any time + * so we try to gracefully handle that. + */ + if (buffer_dirty(bh)) + warn_dirty_buffer(bh); if (test_clear_buffer_dirty(bh) || test_clear_buffer_jbddirty(bh)) was_dirty = 1; -- cgit v1.2.3 From 43237b5490e8f2f4679decd660064ff35ce490cc Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Wed, 20 May 2009 18:41:58 +0200 Subject: ext3: Get rid of extenddisksize parameter of ext3_get_blocks_handle() Get rid of extenddisksize parameter of ext3_get_blocks_handle(). This seems to be a relict from some old days and setting disksize in this function does not make much sence. Currently it was set only by ext3_getblk(). Since the parameter has some effect only if create == 1, it is easy to check that the three callers which end up calling ext3_getblk() with create == 1 (ext3_append, ext3_quota_write, ext3_mkdir) do the right thing and set disksize themselves. Signed-off-by: Jan Kara --- fs/ext3/dir.c | 3 +-- fs/ext3/inode.c | 13 +++---------- include/linux/ext3_fs.h | 2 +- 3 files changed, 5 insertions(+), 13 deletions(-) (limited to 'fs') diff --git a/fs/ext3/dir.c b/fs/ext3/dir.c index 3d724a95882f..373fa90c796a 100644 --- a/fs/ext3/dir.c +++ b/fs/ext3/dir.c @@ -130,8 +130,7 @@ static int ext3_readdir(struct file * filp, struct buffer_head *bh = NULL; map_bh.b_state = 0; - err = ext3_get_blocks_handle(NULL, inode, blk, 1, - &map_bh, 0, 0); + err = ext3_get_blocks_handle(NULL, inode, blk, 1, &map_bh, 0); if (err > 0) { pgoff_t index = map_bh.b_blocknr >> (PAGE_CACHE_SHIFT - inode->i_blkbits); diff --git a/fs/ext3/inode.c b/fs/ext3/inode.c index 4d7da6f6184b..b49908a167ae 100644 --- a/fs/ext3/inode.c +++ b/fs/ext3/inode.c @@ -788,7 +788,7 @@ err_out: int ext3_get_blocks_handle(handle_t *handle, struct inode *inode, sector_t iblock, unsigned long maxblocks, struct buffer_head *bh_result, - int create, int extend_disksize) + int create) { int err = -EIO; int offsets[4]; @@ -911,13 +911,6 @@ int ext3_get_blocks_handle(handle_t *handle, struct inode *inode, if (!err) err = ext3_splice_branch(handle, inode, iblock, partial, indirect_blks, count); - /* - * i_disksize growing is protected by truncate_mutex. Don't forget to - * protect it if you're about to implement concurrent - * ext3_get_block() -bzzz - */ - if (!err && extend_disksize && inode->i_size > ei->i_disksize) - ei->i_disksize = inode->i_size; mutex_unlock(&ei->truncate_mutex); if (err) goto cleanup; @@ -972,7 +965,7 @@ static int ext3_get_block(struct inode *inode, sector_t iblock, } ret = ext3_get_blocks_handle(handle, inode, iblock, - max_blocks, bh_result, create, 0); + max_blocks, bh_result, create); if (ret > 0) { bh_result->b_size = (ret << inode->i_blkbits); ret = 0; @@ -1005,7 +998,7 @@ struct buffer_head *ext3_getblk(handle_t *handle, struct inode *inode, dummy.b_blocknr = -1000; buffer_trace_init(&dummy.b_history); err = ext3_get_blocks_handle(handle, inode, block, 1, - &dummy, create, 1); + &dummy, create); /* * ext3_get_blocks_handle() returns number of blocks * mapped. 0 in case of a HOLE. diff --git a/include/linux/ext3_fs.h b/include/linux/ext3_fs.h index 634a5e5aba3e..7499b3667798 100644 --- a/include/linux/ext3_fs.h +++ b/include/linux/ext3_fs.h @@ -874,7 +874,7 @@ struct buffer_head * ext3_getblk (handle_t *, struct inode *, long, int, int *); struct buffer_head * ext3_bread (handle_t *, struct inode *, int, int, int *); int ext3_get_blocks_handle(handle_t *handle, struct inode *inode, sector_t iblock, unsigned long maxblocks, struct buffer_head *bh_result, - int create, int extend_disksize); + int create); extern struct inode *ext3_iget(struct super_block *, unsigned long); extern int ext3_write_inode (struct inode *, int); -- cgit v1.2.3 From 90a98b2f3f3647fb17667768a348b2b219f2a9f7 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Mon, 20 Jul 2009 13:40:52 -0400 Subject: cifs: free nativeFileSystem field before allocating a new one ...otherwise, we'll leak this memory if we have to reconnect (e.g. after network failure). Signed-off-by: Jeff Layton CC: Stable Signed-off-by: Steve French --- fs/cifs/connect.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index e16d7592116a..9bb5c8750736 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -2726,6 +2726,7 @@ CIFSTCon(unsigned int xid, struct cifsSesInfo *ses, strncpy(tcon->treeName, tree, MAX_TREE_SIZE); /* mostly informational -- no need to fail on error here */ + kfree(tcon->nativeFileSystem); tcon->nativeFileSystem = cifs_strndup_from_ucs(bcc_ptr, bytes_left, is_unicode, nls_codepage); -- cgit v1.2.3 From cefcb800fa9536bb6f29485c53e6c82a65b0c022 Mon Sep 17 00:00:00 2001 From: Goldwyn Rodrigues Date: Sat, 11 Jul 2009 10:57:27 -0500 Subject: ocfs2: Initialize count in aio_write before generic_write_checks generic_write_checks() expects count to be initialized to the size of the write. Writes to files open with O_DIRECT|O_LARGEFILE write 0 bytes because count is uninitialized. Signed-off-by: Goldwyn Rodrigues Signed-off-by: Joel Becker --- fs/ocfs2/file.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c index a49fa44aea1f..aa501d3f93f1 100644 --- a/fs/ocfs2/file.c +++ b/fs/ocfs2/file.c @@ -1851,6 +1851,7 @@ relock: if (ret) goto out_dio; + count = ocount; ret = generic_write_checks(file, ppos, &count, S_ISBLK(inode->i_mode)); if (ret) -- cgit v1.2.3 From 44d8e4e1e9b941065eb7578669fe51eb65c73e63 Mon Sep 17 00:00:00 2001 From: Subrata Modak Date: Tue, 14 Jul 2009 01:19:31 +0530 Subject: ocfs2: Fix compilation warning for fs/ocfs2/xattr.c MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit gcc 4.4.1 generates the following build warning on i386: CC [M] fs/ocfs2/xattr.o fs/ocfs2/xattr.c: In function ‘ocfs2_xattr_block_get’: fs/ocfs2/xattr.c:1055: warning: ‘block_off’ may be used uninitialized in this function The following fix is based on a similar approach by David Howells few days back: http://lkml.org/lkml/2009/7/9/109, Signed-off-by: Subrata Modak, Signed-off-by: Joel Becker --- fs/ocfs2/xattr.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ocfs2/xattr.c b/fs/ocfs2/xattr.c index ba320e250747..d1a27cda984f 100644 --- a/fs/ocfs2/xattr.c +++ b/fs/ocfs2/xattr.c @@ -1052,7 +1052,8 @@ static int ocfs2_xattr_block_get(struct inode *inode, struct ocfs2_xattr_block *xb; struct ocfs2_xattr_value_root *xv; size_t size; - int ret = -ENODATA, name_offset, name_len, block_off, i; + int ret = -ENODATA, name_offset, name_len, i; + int uninitialized_var(block_off); xs->bucket = ocfs2_xattr_bucket_new(inode); if (!xs->bucket) { -- cgit v1.2.3 From cbfa9639aea5007313b75ec43b689d5f9e0c037d Mon Sep 17 00:00:00 2001 From: Wengang Wang Date: Mon, 13 Jul 2009 11:38:23 +0800 Subject: ocfs2: Fix error return in ocfs2_write_cluster() A typo caused ocfs2_write_cluster() to return 0 in some error cases. Fix it. Signed-off-by: Wengang Wang Signed-off-by: Joel Becker --- fs/ocfs2/aops.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ocfs2/aops.c b/fs/ocfs2/aops.c index b2c52b3a1484..25aced3b0c7d 100644 --- a/fs/ocfs2/aops.c +++ b/fs/ocfs2/aops.c @@ -1301,7 +1301,7 @@ static int ocfs2_write_cluster(struct address_space *mapping, if (tmpret) { mlog_errno(tmpret); if (ret == 0) - tmpret = ret; + ret = tmpret; } } -- cgit v1.2.3 From 1f4cea3790bf44137192f441ee84af256e3bf809 Mon Sep 17 00:00:00 2001 From: Wengang Wang Date: Mon, 13 Jul 2009 11:38:58 +0800 Subject: ocfs2: Fail ocfs2_get_block() immediately when a block needs allocation ocfs2_get_block() does no allocation. Hole filling for writes should have happened farther up in the call chain. We detect this case and print an error, but we then continue with the function. We should be exiting immediately. Signed-off-by: Wengang Wang Signed-off-by: Joel Becker --- fs/ocfs2/aops.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/ocfs2/aops.c b/fs/ocfs2/aops.c index 25aced3b0c7d..e511df101451 100644 --- a/fs/ocfs2/aops.c +++ b/fs/ocfs2/aops.c @@ -193,6 +193,7 @@ static int ocfs2_get_block(struct inode *inode, sector_t iblock, (unsigned long long)OCFS2_I(inode)->ip_blkno); mlog(ML_ERROR, "Size %llu, clusters %u\n", (unsigned long long)i_size_read(inode), OCFS2_I(inode)->ip_clusters); dump_stack(); + goto bail; } past_eof = ocfs2_blocks_for_bytes(inode->i_sb, i_size_read(inode)); -- cgit v1.2.3 From f1015c447781729060c415f5133164c638561f25 Mon Sep 17 00:00:00 2001 From: dingdinghua Date: Wed, 15 Jul 2009 21:42:05 +0200 Subject: jbd: fix race between write_metadata_buffer and get_write_access The function journal_write_metadata_buffer() calls jbd_unlock_bh_state(bh_in) too early; this could potentially allow another thread to call get_write_access on the buffer head, modify the data, and dirty it, and allowing the wrong data to be written into the journal. Fortunately, if we lose this race, the only time this will actually cause filesystem corruption is if there is a system crash or other unclean shutdown of the system before the next commit can take place. Signed-off-by: dingdinghua Acked-by: "Theodore Ts'o" Signed-off-by: Jan Kara --- fs/jbd/journal.c | 20 +++++++++++--------- 1 file changed, 11 insertions(+), 9 deletions(-) (limited to 'fs') diff --git a/fs/jbd/journal.c b/fs/jbd/journal.c index 94a64a199a63..f96f85092d1c 100644 --- a/fs/jbd/journal.c +++ b/fs/jbd/journal.c @@ -287,6 +287,7 @@ int journal_write_metadata_buffer(transaction_t *transaction, struct page *new_page; unsigned int new_offset; struct buffer_head *bh_in = jh2bh(jh_in); + journal_t *journal = transaction->t_journal; /* * The buffer really shouldn't be locked: only the current committing @@ -300,6 +301,11 @@ int journal_write_metadata_buffer(transaction_t *transaction, J_ASSERT_BH(bh_in, buffer_jbddirty(bh_in)); new_bh = alloc_buffer_head(GFP_NOFS|__GFP_NOFAIL); + /* keep subsequent assertions sane */ + new_bh->b_state = 0; + init_buffer(new_bh, NULL, NULL); + atomic_set(&new_bh->b_count, 1); + new_jh = journal_add_journal_head(new_bh); /* This sleeps */ /* * If a new transaction has already done a buffer copy-out, then @@ -361,14 +367,6 @@ repeat: kunmap_atomic(mapped_data, KM_USER0); } - /* keep subsequent assertions sane */ - new_bh->b_state = 0; - init_buffer(new_bh, NULL, NULL); - atomic_set(&new_bh->b_count, 1); - jbd_unlock_bh_state(bh_in); - - new_jh = journal_add_journal_head(new_bh); /* This sleeps */ - set_bh_page(new_bh, new_page, new_offset); new_jh->b_transaction = NULL; new_bh->b_size = jh2bh(jh_in)->b_size; @@ -385,7 +383,11 @@ repeat: * copying is moved to the transaction's shadow queue. */ JBUFFER_TRACE(jh_in, "file as BJ_Shadow"); - journal_file_buffer(jh_in, transaction, BJ_Shadow); + spin_lock(&journal->j_list_lock); + __journal_file_buffer(jh_in, transaction, BJ_Shadow); + spin_unlock(&journal->j_list_lock); + jbd_unlock_bh_state(bh_in); + JBUFFER_TRACE(new_jh, "file as BJ_IO"); journal_file_buffer(new_jh, transaction, BJ_IO); -- cgit v1.2.3 From 5549f7cdf84c02939fd368d0842aa2f472bb6e98 Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Tue, 7 Jul 2009 10:28:23 -0400 Subject: inotify: drop user watch count when a watch is removed The inotify rewrite forgot to drop the inotify watch use cound when a watch was removed. This means that a single inotify fd can only ever register a maximum of /proc/sys/fs/max_user_watches even if some of those had been freed. Signed-off-by: Eric Paris --- fs/notify/inotify/inotify_user.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'fs') diff --git a/fs/notify/inotify/inotify_user.c b/fs/notify/inotify/inotify_user.c index ff27a2965844..1a870f9157b3 100644 --- a/fs/notify/inotify/inotify_user.c +++ b/fs/notify/inotify/inotify_user.c @@ -404,6 +404,8 @@ skip_send_ignore: /* removed from idr, drop that reference */ fsnotify_put_mark(entry); + + atomic_dec(&group->inotify_data.user->inotify_watches); } /* ding dong the mark is dead */ -- cgit v1.2.3 From 75fe2b26394c59c8e16bd7b76f4be5d048103ad1 Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Tue, 7 Jul 2009 10:28:23 -0400 Subject: inotify: do not leak inode marks in inotify_add_watch inotify_add_watch had a couple of problems. The biggest being that if inotify_add_watch was called on the same inode twice (to update or change the event mask) a refence was taken on the original inode mark by fsnotify_find_mark_entry but was not being dropped at the end of the inotify_add_watch call. Thus if inotify_rm_watch was called although the mark was removed from the inode, the refcnt wouldn't hit zero and we would leak memory. Reported-by: Catalin Marinas Signed-off-by: Eric Paris --- fs/notify/inotify/inotify_user.c | 14 ++++++++++---- 1 file changed, 10 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/notify/inotify/inotify_user.c b/fs/notify/inotify/inotify_user.c index 1a870f9157b3..aff4214f16c3 100644 --- a/fs/notify/inotify/inotify_user.c +++ b/fs/notify/inotify/inotify_user.c @@ -463,9 +463,6 @@ retry: goto out_err; spin_lock(&group->inotify_data.idr_lock); - /* if entry is added to the idr we keep the reference obtained - * through fsnotify_mark_add. remember to drop this reference - * when entry is removed from idr */ ret = idr_get_new_above(&group->inotify_data.idr, entry, ++group->inotify_data.last_wd, &ientry->wd); @@ -476,8 +473,13 @@ retry: goto out_err; } atomic_inc(&group->inotify_data.user->inotify_watches); + + /* we put the mark on the idr, take a reference */ + fsnotify_get_mark(entry); } + ret = ientry->wd; + spin_lock(&entry->lock); old_mask = entry->mask; @@ -508,7 +510,11 @@ retry: fsnotify_recalc_group_mask(group); } - return ientry->wd; + /* this either matches fsnotify_find_mark_entry, or init_mark_entry + * depending on which path we took... */ + fsnotify_put_mark(entry); + + return ret; out_err: /* see this isn't supposed to happen, just kill the watch */ -- cgit v1.2.3 From 7e790dd5fc937bc8d2400c30a05e32a9e9eef276 Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Tue, 7 Jul 2009 10:28:24 -0400 Subject: inotify: fix error paths in inotify_update_watch inotify_update_watch could leave things in a horrid state on a number of error paths. We could try to remove idr entries that didn't exist, we could send an IN_IGNORED to userspace for watches that don't exist, and a bit of other stupidity. Clean these up by doing the idr addition before we put the mark on the inode since we can clean that up on error and getting off the inode's mark list is hard. Signed-off-by: Eric Paris --- fs/notify/inotify/inotify_user.c | 79 +++++++++++++++++++++++++--------------- 1 file changed, 49 insertions(+), 30 deletions(-) (limited to 'fs') diff --git a/fs/notify/inotify/inotify_user.c b/fs/notify/inotify/inotify_user.c index aff4214f16c3..726118a5845b 100644 --- a/fs/notify/inotify/inotify_user.c +++ b/fs/notify/inotify/inotify_user.c @@ -365,6 +365,17 @@ static int inotify_find_inode(const char __user *dirname, struct path *path, uns return error; } +static void inotify_remove_from_idr(struct fsnotify_group *group, + struct inotify_inode_mark_entry *ientry) +{ + struct idr *idr; + + spin_lock(&group->inotify_data.idr_lock); + idr = &group->inotify_data.idr; + idr_remove(idr, ientry->wd); + spin_unlock(&group->inotify_data.idr_lock); + ientry->wd = -1; +} /* * Send IN_IGNORED for this wd, remove this wd from the idr, and drop the * internal reference help on the mark because it is in the idr. @@ -375,7 +386,6 @@ void inotify_ignored_and_remove_idr(struct fsnotify_mark_entry *entry, struct inotify_inode_mark_entry *ientry; struct inotify_event_private_data *event_priv; struct fsnotify_event_private_data *fsn_event_priv; - struct idr *idr; ientry = container_of(entry, struct inotify_inode_mark_entry, fsn_entry); @@ -397,10 +407,7 @@ void inotify_ignored_and_remove_idr(struct fsnotify_mark_entry *entry, skip_send_ignore: /* remove this entry from the idr */ - spin_lock(&group->inotify_data.idr_lock); - idr = &group->inotify_data.idr; - idr_remove(idr, ientry->wd); - spin_unlock(&group->inotify_data.idr_lock); + inotify_remove_from_idr(group, ientry); /* removed from idr, drop that reference */ fsnotify_put_mark(entry); @@ -420,6 +427,7 @@ static int inotify_update_watch(struct fsnotify_group *group, struct inode *inod { struct fsnotify_mark_entry *entry = NULL; struct inotify_inode_mark_entry *ientry; + struct inotify_inode_mark_entry *tmp_ientry; int ret = 0; int add = (arg & IN_MASK_ADD); __u32 mask; @@ -430,50 +438,60 @@ static int inotify_update_watch(struct fsnotify_group *group, struct inode *inod if (unlikely(!mask)) return -EINVAL; - ientry = kmem_cache_alloc(inotify_inode_mark_cachep, GFP_KERNEL); - if (unlikely(!ientry)) + tmp_ientry = kmem_cache_alloc(inotify_inode_mark_cachep, GFP_KERNEL); + if (unlikely(!tmp_ientry)) return -ENOMEM; /* we set the mask at the end after attaching it */ - fsnotify_init_mark(&ientry->fsn_entry, inotify_free_mark); - ientry->wd = 0; + fsnotify_init_mark(&tmp_ientry->fsn_entry, inotify_free_mark); + tmp_ientry->wd = -1; find_entry: spin_lock(&inode->i_lock); entry = fsnotify_find_mark_entry(group, inode); spin_unlock(&inode->i_lock); if (entry) { - kmem_cache_free(inotify_inode_mark_cachep, ientry); ientry = container_of(entry, struct inotify_inode_mark_entry, fsn_entry); } else { - if (atomic_read(&group->inotify_data.user->inotify_watches) >= inotify_max_user_watches) { - ret = -ENOSPC; - goto out_err; - } - - ret = fsnotify_add_mark(&ientry->fsn_entry, group, inode); - if (ret == -EEXIST) - goto find_entry; - else if (ret) + ret = -ENOSPC; + if (atomic_read(&group->inotify_data.user->inotify_watches) >= inotify_max_user_watches) goto out_err; - - entry = &ientry->fsn_entry; retry: ret = -ENOMEM; if (unlikely(!idr_pre_get(&group->inotify_data.idr, GFP_KERNEL))) goto out_err; spin_lock(&group->inotify_data.idr_lock); - ret = idr_get_new_above(&group->inotify_data.idr, entry, - ++group->inotify_data.last_wd, - &ientry->wd); + ret = idr_get_new_above(&group->inotify_data.idr, &tmp_ientry->fsn_entry, + group->inotify_data.last_wd, + &tmp_ientry->wd); spin_unlock(&group->inotify_data.idr_lock); if (ret) { if (ret == -EAGAIN) goto retry; goto out_err; } + + ret = fsnotify_add_mark(&tmp_ientry->fsn_entry, group, inode); + if (ret) { + inotify_remove_from_idr(group, tmp_ientry); + if (ret == -EEXIST) + goto find_entry; + goto out_err; + } + + /* tmp_ientry has been added to the inode, so we are all set up. + * now we just need to make sure tmp_ientry doesn't get freed and + * we need to set up entry and ientry so the generic code can + * do its thing. */ + ientry = tmp_ientry; + entry = &ientry->fsn_entry; + tmp_ientry = NULL; + atomic_inc(&group->inotify_data.user->inotify_watches); + /* update the idr hint */ + group->inotify_data.last_wd = ientry->wd; + /* we put the mark on the idr, take a reference */ fsnotify_get_mark(entry); } @@ -514,14 +532,15 @@ retry: * depending on which path we took... */ fsnotify_put_mark(entry); - return ret; - out_err: - /* see this isn't supposed to happen, just kill the watch */ - if (entry) { - fsnotify_destroy_mark_by_entry(entry); - fsnotify_put_mark(entry); + /* could be an error, could be that we found an existing mark */ + if (tmp_ientry) { + /* on the idr but didn't make it on the inode */ + if (tmp_ientry->wd != -1) + inotify_remove_from_idr(group, tmp_ientry); + kmem_cache_free(inotify_inode_mark_cachep, tmp_ientry); } + return ret; } -- cgit v1.2.3 From 520dc2a526fd681337883b6ff1ddcf7c23b1b063 Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Mon, 13 Jul 2009 15:56:54 -0400 Subject: fsnotify: use def_bool in kconfig instead of letting the user choose fsnotify doens't give the user anything. If someone chooses inotify or dnotify it should build fsnotify, if they don't select one it shouldn't be built. This patch changes fsnotify to be a def_bool=n and makes everything else select it. Also fixes the issue people complained about on lwn where gdm hung because they didn't have inotify and they didn't get the inotify build option..... Signed-off-by: Eric Paris --- fs/notify/Kconfig | 12 +----------- fs/notify/dnotify/Kconfig | 2 +- fs/notify/inotify/Kconfig | 2 +- 3 files changed, 3 insertions(+), 13 deletions(-) (limited to 'fs') diff --git a/fs/notify/Kconfig b/fs/notify/Kconfig index 31dac7e3b0f1..dffbb0911d02 100644 --- a/fs/notify/Kconfig +++ b/fs/notify/Kconfig @@ -1,15 +1,5 @@ config FSNOTIFY - bool "Filesystem notification backend" - default y - ---help--- - fsnotify is a backend for filesystem notification. fsnotify does - not provide any userspace interface but does provide the basis - needed for other notification schemes such as dnotify, inotify, - and fanotify. - - Say Y here to enable fsnotify suport. - - If unsure, say Y. + def_bool n source "fs/notify/dnotify/Kconfig" source "fs/notify/inotify/Kconfig" diff --git a/fs/notify/dnotify/Kconfig b/fs/notify/dnotify/Kconfig index 904ff8d5405a..f9c1ca139d8f 100644 --- a/fs/notify/dnotify/Kconfig +++ b/fs/notify/dnotify/Kconfig @@ -1,6 +1,6 @@ config DNOTIFY bool "Dnotify support" - depends on FSNOTIFY + select FSNOTIFY default y help Dnotify is a directory-based per-fd file change notification system diff --git a/fs/notify/inotify/Kconfig b/fs/notify/inotify/Kconfig index 5356884289a1..3e56dbffe729 100644 --- a/fs/notify/inotify/Kconfig +++ b/fs/notify/inotify/Kconfig @@ -15,7 +15,7 @@ config INOTIFY config INOTIFY_USER bool "Inotify support for userspace" - depends on FSNOTIFY + select FSNOTIFY default y ---help--- Say Y here to enable inotify support for userspace, including the -- cgit v1.2.3 From 4a148ba988988b9c400ad0f2cbccc155289b954b Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Mon, 13 Jul 2009 15:56:55 -0400 Subject: inotify: check filename before dropping repeat events inotify drops events if the last event on the queue is the same as the current event. But it does 2 things wrong. First it is comparing old->inode with new->inode. But after an event if put on the queue the ->inode is no longer allowed to be used. It's possible between the last event and this new event the inode could be reused and we would falsely match the inode's memory address between two differing events. The second problem is that when a file is removed fsnotify is passed the negative dentry for the removed object rather than the postive dentry from immediately before the removal. This mean the (broken) inotify tail drop code was matching the NULL ->inode of differing events. The fix is to check the file name which is stored with events when doing the tail drop instead of wrongly checking the address of the stored ->inode. Reported-by: Scott James Remnant Signed-off-by: Eric Paris --- fs/notify/notification.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/notify/notification.c b/fs/notify/notification.c index 959b73e756fd..69391fe8efb1 100644 --- a/fs/notify/notification.c +++ b/fs/notify/notification.c @@ -136,10 +136,15 @@ static bool event_compare(struct fsnotify_event *old, struct fsnotify_event *new { if ((old->mask == new->mask) && (old->to_tell == new->to_tell) && - (old->data_type == new->data_type)) { + (old->data_type == new->data_type) && + (old->name_len == new->name_len)) { switch (old->data_type) { case (FSNOTIFY_EVENT_INODE): - if (old->inode == new->inode) + /* remember, after old was put on the wait_q we aren't + * allowed to look at the inode any more, only thing + * left to check was if the file_name is the same */ + if (old->name_len && + !strcmp(old->file_name, new->file_name)) return true; break; case (FSNOTIFY_EVENT_PATH): -- cgit v1.2.3 From c05594b62125c528d93af3a78229793aae36df7f Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Mon, 13 Jul 2009 15:56:55 -0400 Subject: fsnotify: fix inotify tail drop check with path entries fsnotify drops new events when they are the same as the tail event on the queue to be sent to userspace. The problem is that if the event comes with a path we forget to break out of the switch statement and fall into the code path which matches on events that do not have any type of file backed information (things like IN_UNMOUNT and IN_Q_OVERFLOW). The problem is that this code thinks all such events should be dropped. Fix is to add a break. Signed-off-by: Eric Paris --- fs/notify/notification.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/notify/notification.c b/fs/notify/notification.c index 69391fe8efb1..2b20feaf263a 100644 --- a/fs/notify/notification.c +++ b/fs/notify/notification.c @@ -151,6 +151,7 @@ static bool event_compare(struct fsnotify_event *old, struct fsnotify_event *new if ((old->path.mnt == new->path.mnt) && (old->path.dentry == new->path.dentry)) return true; + break; case (FSNOTIFY_EVENT_NONE): return true; }; -- cgit v1.2.3 From f44aebcc566d1d6275f7191867b9633dc11de2ee Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Wed, 15 Jul 2009 15:49:52 -0400 Subject: inotify: use GFP_NOFS under potential memory pressure inotify can have a watchs removed under filesystem reclaim. ================================= [ INFO: inconsistent lock state ] 2.6.31-rc2 #16 --------------------------------- inconsistent {IN-RECLAIM_FS-W} -> {RECLAIM_FS-ON-W} usage. khubd/217 [HC0[0]:SC0[0]:HE1:SE1] takes: (iprune_mutex){+.+.?.}, at: [] invalidate_inodes+0x20/0xe3 {IN-RECLAIM_FS-W} state was registered at: [] __lock_acquire+0x2c9/0xac4 [] lock_acquire+0x9f/0xc2 [] __mutex_lock_common+0x2d/0x323 [] mutex_lock_nested+0x2e/0x36 [] shrink_icache_memory+0x38/0x1b2 [] shrink_slab+0xe2/0x13c [] kswapd+0x3d1/0x55d [] kthread+0x66/0x6b [] kernel_thread_helper+0x7/0x10 [] 0xffffffff Two things are needed to fix this. First we need a method to tell fsnotify_create_event() to use GFP_NOFS and second we need to stop using one global IN_IGNORED event and allocate them one at a time. This solves current issues with multiple IN_IGNORED on a queue having tail drop problems and simplifies the allocations since we don't have to worry about two tasks opperating on the IGNORED event concurrently. Signed-off-by: Eric Paris --- fs/notify/fsnotify.c | 4 +++- fs/notify/inotify/inotify_user.c | 18 ++++++++++++------ fs/notify/notification.c | 9 +++++---- include/linux/fsnotify_backend.h | 2 +- 4 files changed, 21 insertions(+), 12 deletions(-) (limited to 'fs') diff --git a/fs/notify/fsnotify.c b/fs/notify/fsnotify.c index ec2f7bd76818..037e878e03fc 100644 --- a/fs/notify/fsnotify.c +++ b/fs/notify/fsnotify.c @@ -159,7 +159,9 @@ void fsnotify(struct inode *to_tell, __u32 mask, void *data, int data_is, const if (!group->ops->should_send_event(group, to_tell, mask)) continue; if (!event) { - event = fsnotify_create_event(to_tell, mask, data, data_is, file_name, cookie); + event = fsnotify_create_event(to_tell, mask, data, + data_is, file_name, cookie, + GFP_KERNEL); /* shit, we OOM'd and now we can't tell, maybe * someday someone else will want to do something * here */ diff --git a/fs/notify/inotify/inotify_user.c b/fs/notify/inotify/inotify_user.c index 726118a5845b..f30d9bbc2e1b 100644 --- a/fs/notify/inotify/inotify_user.c +++ b/fs/notify/inotify/inotify_user.c @@ -57,7 +57,6 @@ int inotify_max_user_watches __read_mostly; static struct kmem_cache *inotify_inode_mark_cachep __read_mostly; struct kmem_cache *event_priv_cachep __read_mostly; -static struct fsnotify_event *inotify_ignored_event; /* * When inotify registers a new group it increments this and uses that @@ -384,12 +383,19 @@ void inotify_ignored_and_remove_idr(struct fsnotify_mark_entry *entry, struct fsnotify_group *group) { struct inotify_inode_mark_entry *ientry; + struct fsnotify_event *ignored_event; struct inotify_event_private_data *event_priv; struct fsnotify_event_private_data *fsn_event_priv; + ignored_event = fsnotify_create_event(NULL, FS_IN_IGNORED, NULL, + FSNOTIFY_EVENT_NONE, NULL, 0, + GFP_NOFS); + if (!ignored_event) + return; + ientry = container_of(entry, struct inotify_inode_mark_entry, fsn_entry); - event_priv = kmem_cache_alloc(event_priv_cachep, GFP_KERNEL); + event_priv = kmem_cache_alloc(event_priv_cachep, GFP_NOFS); if (unlikely(!event_priv)) goto skip_send_ignore; @@ -398,7 +404,7 @@ void inotify_ignored_and_remove_idr(struct fsnotify_mark_entry *entry, fsn_event_priv->group = group; event_priv->wd = ientry->wd; - fsnotify_add_notify_event(group, inotify_ignored_event, fsn_event_priv); + fsnotify_add_notify_event(group, ignored_event, fsn_event_priv); /* did the private data get added? */ if (list_empty(&fsn_event_priv->event_list)) @@ -406,6 +412,9 @@ void inotify_ignored_and_remove_idr(struct fsnotify_mark_entry *entry, skip_send_ignore: + /* matches the reference taken when the event was created */ + fsnotify_put_event(ignored_event); + /* remove this entry from the idr */ inotify_remove_from_idr(group, ientry); @@ -748,9 +757,6 @@ static int __init inotify_user_setup(void) inotify_inode_mark_cachep = KMEM_CACHE(inotify_inode_mark_entry, SLAB_PANIC); event_priv_cachep = KMEM_CACHE(inotify_event_private_data, SLAB_PANIC); - inotify_ignored_event = fsnotify_create_event(NULL, FS_IN_IGNORED, NULL, FSNOTIFY_EVENT_NONE, NULL, 0); - if (!inotify_ignored_event) - panic("unable to allocate the inotify ignored event\n"); inotify_max_queued_events = 16384; inotify_max_user_instances = 128; diff --git a/fs/notify/notification.c b/fs/notify/notification.c index 2b20feaf263a..521368574e97 100644 --- a/fs/notify/notification.c +++ b/fs/notify/notification.c @@ -153,7 +153,7 @@ static bool event_compare(struct fsnotify_event *old, struct fsnotify_event *new return true; break; case (FSNOTIFY_EVENT_NONE): - return true; + return false; }; } return false; @@ -345,18 +345,19 @@ static void initialize_event(struct fsnotify_event *event) * @name the filename, if available */ struct fsnotify_event *fsnotify_create_event(struct inode *to_tell, __u32 mask, void *data, - int data_type, const char *name, u32 cookie) + int data_type, const char *name, u32 cookie, + gfp_t gfp) { struct fsnotify_event *event; - event = kmem_cache_alloc(fsnotify_event_cachep, GFP_KERNEL); + event = kmem_cache_alloc(fsnotify_event_cachep, gfp); if (!event) return NULL; initialize_event(event); if (name) { - event->file_name = kstrdup(name, GFP_KERNEL); + event->file_name = kstrdup(name, gfp); if (!event->file_name) { kmem_cache_free(fsnotify_event_cachep, event); return NULL; diff --git a/include/linux/fsnotify_backend.h b/include/linux/fsnotify_backend.h index 6c3de999fb34..4d6f47b51189 100644 --- a/include/linux/fsnotify_backend.h +++ b/include/linux/fsnotify_backend.h @@ -352,7 +352,7 @@ extern void fsnotify_unmount_inodes(struct list_head *list); /* put here because inotify does some weird stuff when destroying watches */ extern struct fsnotify_event *fsnotify_create_event(struct inode *to_tell, __u32 mask, void *data, int data_is, const char *name, - u32 cookie); + u32 cookie, gfp_t gfp); #else -- cgit v1.2.3 From b64aec8d1e1d8482a7b6cca60c8105c756bf1fe4 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Tue, 21 Jul 2009 16:47:46 -0400 Subject: NFSv4: Fix an Oops in nfs4_free_lock_state The oops http://www.kerneloops.org/raw.php?rawid=537858&msgid= appears to be due to the nfs4_lock_state->ls_state field being uninitialised. This happens if the call to nfs4_free_lock_state() is triggered at the end of nfs4_get_lock_state(). The fix is to move the initialisation of ls_state into the allocator. Signed-off-by: Trond Myklebust --- fs/nfs/nfs4state.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c index b73c5a728655..65ca8c18476f 100644 --- a/fs/nfs/nfs4state.c +++ b/fs/nfs/nfs4state.c @@ -553,6 +553,7 @@ static struct nfs4_lock_state *nfs4_alloc_lock_state(struct nfs4_state *state, f INIT_LIST_HEAD(&lsp->ls_sequence.list); lsp->ls_seqid.sequence = &lsp->ls_sequence; atomic_set(&lsp->ls_count, 1); + lsp->ls_state = state; lsp->ls_owner = fl_owner; spin_lock(&clp->cl_lock); nfs_alloc_unique_id(&clp->cl_lockowner_id, &lsp->ls_id, 1, 64); @@ -587,7 +588,6 @@ static struct nfs4_lock_state *nfs4_get_lock_state(struct nfs4_state *state, fl_ if (lsp != NULL) break; if (new != NULL) { - new->ls_state = state; list_add(&new->ls_locks, &state->lock_states); set_bit(LK_STATE_IN_USE, &state->flags); lsp = new; -- cgit v1.2.3 From fccba8045537f7e840d0e7565e1989d465e488a3 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Tue, 21 Jul 2009 16:48:07 -0400 Subject: NFSv4: Fix an NFSv4 mount regression Commit 008f55d0e019943323c20a03493a2ba5672a4cc8 (nfs41: recover lease in _nfs4_lookup_root) forces the state manager to always run on mount. This is a bug in the case of NFSv4.0, which doesn't require us to send a setclientid until we want to grab file state. In any case, this is completely the wrong place to be doing state management. Moving that code into nfs4_init_session... Signed-off-by: Trond Myklebust --- fs/nfs/client.c | 18 +++--------------- fs/nfs/nfs4_fs.h | 6 ++++++ fs/nfs/nfs4proc.c | 24 +++++++++++++++++------- 3 files changed, 26 insertions(+), 22 deletions(-) (limited to 'fs') diff --git a/fs/nfs/client.c b/fs/nfs/client.c index c2d061675d80..8d25ccb2d51d 100644 --- a/fs/nfs/client.c +++ b/fs/nfs/client.c @@ -1242,20 +1242,6 @@ error: return error; } -/* - * Initialize a session. - * Note: save the mount rsize and wsize for create_server negotiation. - */ -static void nfs4_init_session(struct nfs_client *clp, - unsigned int wsize, unsigned int rsize) -{ -#if defined(CONFIG_NFS_V4_1) - if (nfs4_has_session(clp)) { - clp->cl_session->fc_attrs.max_rqst_sz = wsize; - clp->cl_session->fc_attrs.max_resp_sz = rsize; - } -#endif /* CONFIG_NFS_V4_1 */ -} /* * Session has been established, and the client marked ready. @@ -1350,7 +1336,9 @@ struct nfs_server *nfs4_create_server(const struct nfs_parsed_mount_data *data, BUG_ON(!server->nfs_client->rpc_ops); BUG_ON(!server->nfs_client->rpc_ops->file_inode_ops); - nfs4_init_session(server->nfs_client, server->wsize, server->rsize); + error = nfs4_init_session(server); + if (error < 0) + goto error; /* Probe the root fh to retrieve its FSID */ error = nfs4_path_walk(server, mntfh, data->nfs_server.export_path); diff --git a/fs/nfs/nfs4_fs.h b/fs/nfs/nfs4_fs.h index 61bc3a32e1e2..6ea07a3c75d4 100644 --- a/fs/nfs/nfs4_fs.h +++ b/fs/nfs/nfs4_fs.h @@ -220,6 +220,7 @@ extern void nfs4_destroy_session(struct nfs4_session *session); extern struct nfs4_session *nfs4_alloc_session(struct nfs_client *clp); extern int nfs4_proc_create_session(struct nfs_client *, int reset); extern int nfs4_proc_destroy_session(struct nfs4_session *); +extern int nfs4_init_session(struct nfs_server *server); #else /* CONFIG_NFS_v4_1 */ static inline int nfs4_setup_sequence(struct nfs_client *clp, struct nfs4_sequence_args *args, struct nfs4_sequence_res *res, @@ -227,6 +228,11 @@ static inline int nfs4_setup_sequence(struct nfs_client *clp, { return 0; } + +static inline int nfs4_init_session(struct nfs_server *server) +{ + return 0; +} #endif /* CONFIG_NFS_V4_1 */ extern struct nfs4_state_maintenance_ops *nfs4_state_renewal_ops[]; diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index ff0c080db59b..df24f67bca69 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -2040,15 +2040,9 @@ static int _nfs4_lookup_root(struct nfs_server *server, struct nfs_fh *fhandle, .rpc_argp = &args, .rpc_resp = &res, }; - int status; nfs_fattr_init(info->fattr); - status = nfs4_recover_expired_lease(server); - if (!status) - status = nfs4_check_client_ready(server->nfs_client); - if (!status) - status = nfs4_call_sync(server, &msg, &args, &res, 0); - return status; + return nfs4_call_sync(server, &msg, &args, &res, 0); } static int nfs4_lookup_root(struct nfs_server *server, struct nfs_fh *fhandle, @@ -4793,6 +4787,22 @@ int nfs4_proc_destroy_session(struct nfs4_session *session) return status; } +int nfs4_init_session(struct nfs_server *server) +{ + struct nfs_client *clp = server->nfs_client; + int ret; + + if (!nfs4_has_session(clp)) + return 0; + + clp->cl_session->fc_attrs.max_rqst_sz = server->wsize; + clp->cl_session->fc_attrs.max_resp_sz = server->rsize; + ret = nfs4_recover_expired_lease(server); + if (!ret) + ret = nfs4_check_client_ready(clp); + return ret; +} + /* * Renew the cl_session lease. */ -- cgit v1.2.3 From 3c5e10683e684ef45614c9071847e48f633d9806 Mon Sep 17 00:00:00 2001 From: Tao Ma Date: Tue, 21 Jul 2009 15:42:05 +0800 Subject: ocfs2: Add extra credits and access the modified bh in update_edge_lengths. In normal tree rotation left process, we will never touch the tree branch above subtree_index and ocfs2_extend_rotate_transaction doesn't reserve the credits for them either. But when we want to delete the rightmost extent block, we have to update the rightmost records for all the rightmost branch(See ocfs2_update_edge_lengths), so we have to allocate extra credits for them. What's more, we have to access them also. Signed-off-by: Tao Ma Signed-off-by: Joel Becker --- fs/ocfs2/alloc.c | 44 +++++++++++++++++++++++++++++++++++++++----- 1 file changed, 39 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/ocfs2/alloc.c b/fs/ocfs2/alloc.c index 9edcde4974aa..11085af71247 100644 --- a/fs/ocfs2/alloc.c +++ b/fs/ocfs2/alloc.c @@ -2476,15 +2476,37 @@ out_ret_path: return ret; } -static void ocfs2_update_edge_lengths(struct inode *inode, handle_t *handle, - struct ocfs2_path *path) +static int ocfs2_update_edge_lengths(struct inode *inode, handle_t *handle, + int subtree_index, struct ocfs2_path *path) { - int i, idx; + int i, idx, ret; struct ocfs2_extent_rec *rec; struct ocfs2_extent_list *el; struct ocfs2_extent_block *eb; u32 range; + /* + * In normal tree rotation process, we will never touch the + * tree branch above subtree_index and ocfs2_extend_rotate_transaction + * doesn't reserve the credits for them either. + * + * But we do have a special case here which will update the rightmost + * records for all the bh in the path. + * So we have to allocate extra credits and access them. + */ + ret = ocfs2_extend_trans(handle, + handle->h_buffer_credits + subtree_index); + if (ret) { + mlog_errno(ret); + goto out; + } + + ret = ocfs2_journal_access_path(inode, handle, path); + if (ret) { + mlog_errno(ret); + goto out; + } + /* Path should always be rightmost. */ eb = (struct ocfs2_extent_block *)path_leaf_bh(path)->b_data; BUG_ON(eb->h_next_leaf_blk != 0ULL); @@ -2505,6 +2527,8 @@ static void ocfs2_update_edge_lengths(struct inode *inode, handle_t *handle, ocfs2_journal_dirty(handle, path->p_node[i].bh); } +out: + return ret; } static void ocfs2_unlink_path(struct inode *inode, handle_t *handle, @@ -2717,7 +2741,12 @@ static int ocfs2_rotate_subtree_left(struct inode *inode, handle_t *handle, if (del_right_subtree) { ocfs2_unlink_subtree(inode, handle, left_path, right_path, subtree_index, dealloc); - ocfs2_update_edge_lengths(inode, handle, left_path); + ret = ocfs2_update_edge_lengths(inode, handle, subtree_index, + left_path); + if (ret) { + mlog_errno(ret); + goto out; + } eb = (struct ocfs2_extent_block *)path_leaf_bh(left_path)->b_data; ocfs2_et_set_last_eb_blk(et, le64_to_cpu(eb->h_blkno)); @@ -3034,7 +3063,12 @@ static int ocfs2_remove_rightmost_path(struct inode *inode, handle_t *handle, ocfs2_unlink_subtree(inode, handle, left_path, path, subtree_index, dealloc); - ocfs2_update_edge_lengths(inode, handle, left_path); + ret = ocfs2_update_edge_lengths(inode, handle, subtree_index, + left_path); + if (ret) { + mlog_errno(ret); + goto out; + } eb = (struct ocfs2_extent_block *)path_leaf_bh(left_path)->b_data; ocfs2_et_set_last_eb_blk(et, le64_to_cpu(eb->h_blkno)); -- cgit v1.2.3 From f7b1aa69be138ad9d7d3f31fa56f4c9407f56b6a Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Mon, 20 Jul 2009 12:12:36 +0200 Subject: ocfs2: Fix deadlock on umount In commit ea455f8ab68338ba69f5d3362b342c115bea8e13, we moved the dentry lock put process into ocfs2_wq. This causes problems during umount because ocfs2_wq can drop references to inodes while they are being invalidated by invalidate_inodes() causing all sorts of nasty things (invalidate_inodes() ending in an infinite loop, "Busy inodes after umount" messages etc.). We fix the problem by stopping ocfs2_wq from doing any further releasing of inode references on the superblock being unmounted, wait until it finishes the current round of releasing and finally cleaning up all the references in dentry_lock_list from ocfs2_put_super(). The issue was tracked down by Tao Ma . Signed-off-by: Jan Kara Signed-off-by: Joel Becker --- fs/ocfs2/dcache.c | 35 +++++++++++++++++++++++++++-------- fs/ocfs2/dcache.h | 3 +++ fs/ocfs2/ocfs2.h | 22 ++++++++++++++++++---- fs/ocfs2/super.c | 25 ++++++++++++++++++++++--- 4 files changed, 70 insertions(+), 15 deletions(-) (limited to 'fs') diff --git a/fs/ocfs2/dcache.c b/fs/ocfs2/dcache.c index b574431a031d..2f28b7de2c8d 100644 --- a/fs/ocfs2/dcache.c +++ b/fs/ocfs2/dcache.c @@ -310,22 +310,19 @@ out_attach: return ret; } -static DEFINE_SPINLOCK(dentry_list_lock); +DEFINE_SPINLOCK(dentry_list_lock); /* We limit the number of dentry locks to drop in one go. We have * this limit so that we don't starve other users of ocfs2_wq. */ #define DL_INODE_DROP_COUNT 64 /* Drop inode references from dentry locks */ -void ocfs2_drop_dl_inodes(struct work_struct *work) +static void __ocfs2_drop_dl_inodes(struct ocfs2_super *osb, int drop_count) { - struct ocfs2_super *osb = container_of(work, struct ocfs2_super, - dentry_lock_work); struct ocfs2_dentry_lock *dl; - int drop_count = DL_INODE_DROP_COUNT; spin_lock(&dentry_list_lock); - while (osb->dentry_lock_list && drop_count--) { + while (osb->dentry_lock_list && (drop_count < 0 || drop_count--)) { dl = osb->dentry_lock_list; osb->dentry_lock_list = dl->dl_next; spin_unlock(&dentry_list_lock); @@ -333,11 +330,32 @@ void ocfs2_drop_dl_inodes(struct work_struct *work) kfree(dl); spin_lock(&dentry_list_lock); } - if (osb->dentry_lock_list) + spin_unlock(&dentry_list_lock); +} + +void ocfs2_drop_dl_inodes(struct work_struct *work) +{ + struct ocfs2_super *osb = container_of(work, struct ocfs2_super, + dentry_lock_work); + + __ocfs2_drop_dl_inodes(osb, DL_INODE_DROP_COUNT); + /* + * Don't queue dropping if umount is in progress. We flush the + * list in ocfs2_dismount_volume + */ + spin_lock(&dentry_list_lock); + if (osb->dentry_lock_list && + !ocfs2_test_osb_flag(osb, OCFS2_OSB_DROP_DENTRY_LOCK_IMMED)) queue_work(ocfs2_wq, &osb->dentry_lock_work); spin_unlock(&dentry_list_lock); } +/* Flush the whole work queue */ +void ocfs2_drop_all_dl_inodes(struct ocfs2_super *osb) +{ + __ocfs2_drop_dl_inodes(osb, -1); +} + /* * ocfs2_dentry_iput() and friends. * @@ -368,7 +386,8 @@ static void ocfs2_drop_dentry_lock(struct ocfs2_super *osb, /* We leave dropping of inode reference to ocfs2_wq as that can * possibly lead to inode deletion which gets tricky */ spin_lock(&dentry_list_lock); - if (!osb->dentry_lock_list) + if (!osb->dentry_lock_list && + !ocfs2_test_osb_flag(osb, OCFS2_OSB_DROP_DENTRY_LOCK_IMMED)) queue_work(ocfs2_wq, &osb->dentry_lock_work); dl->dl_next = osb->dentry_lock_list; osb->dentry_lock_list = dl; diff --git a/fs/ocfs2/dcache.h b/fs/ocfs2/dcache.h index faa12e75f98d..f5dd1789acf1 100644 --- a/fs/ocfs2/dcache.h +++ b/fs/ocfs2/dcache.h @@ -49,10 +49,13 @@ struct ocfs2_dentry_lock { int ocfs2_dentry_attach_lock(struct dentry *dentry, struct inode *inode, u64 parent_blkno); +extern spinlock_t dentry_list_lock; + void ocfs2_dentry_lock_put(struct ocfs2_super *osb, struct ocfs2_dentry_lock *dl); void ocfs2_drop_dl_inodes(struct work_struct *work); +void ocfs2_drop_all_dl_inodes(struct ocfs2_super *osb); struct dentry *ocfs2_find_local_alias(struct inode *inode, u64 parent_blkno, int skip_unhashed); diff --git a/fs/ocfs2/ocfs2.h b/fs/ocfs2/ocfs2.h index c9345ebb8493..39e1d5a39505 100644 --- a/fs/ocfs2/ocfs2.h +++ b/fs/ocfs2/ocfs2.h @@ -224,10 +224,12 @@ enum ocfs2_mount_options OCFS2_MOUNT_GRPQUOTA = 1 << 10, /* We support group quotas */ }; -#define OCFS2_OSB_SOFT_RO 0x0001 -#define OCFS2_OSB_HARD_RO 0x0002 -#define OCFS2_OSB_ERROR_FS 0x0004 -#define OCFS2_DEFAULT_ATIME_QUANTUM 60 +#define OCFS2_OSB_SOFT_RO 0x0001 +#define OCFS2_OSB_HARD_RO 0x0002 +#define OCFS2_OSB_ERROR_FS 0x0004 +#define OCFS2_OSB_DROP_DENTRY_LOCK_IMMED 0x0008 + +#define OCFS2_DEFAULT_ATIME_QUANTUM 60 struct ocfs2_journal; struct ocfs2_slot_info; @@ -490,6 +492,18 @@ static inline void ocfs2_set_osb_flag(struct ocfs2_super *osb, spin_unlock(&osb->osb_lock); } + +static inline unsigned long ocfs2_test_osb_flag(struct ocfs2_super *osb, + unsigned long flag) +{ + unsigned long ret; + + spin_lock(&osb->osb_lock); + ret = osb->osb_flags & flag; + spin_unlock(&osb->osb_lock); + return ret; +} + static inline void ocfs2_set_ro_flag(struct ocfs2_super *osb, int hard) { diff --git a/fs/ocfs2/super.c b/fs/ocfs2/super.c index 63af2e36d834..f2893878f8ad 100644 --- a/fs/ocfs2/super.c +++ b/fs/ocfs2/super.c @@ -1213,14 +1213,27 @@ static int ocfs2_get_sb(struct file_system_type *fs_type, mnt); } +static void ocfs2_kill_sb(struct super_block *sb) +{ + struct ocfs2_super *osb = OCFS2_SB(sb); + + /* Prevent further queueing of inode drop events */ + spin_lock(&dentry_list_lock); + ocfs2_set_osb_flag(osb, OCFS2_OSB_DROP_DENTRY_LOCK_IMMED); + spin_unlock(&dentry_list_lock); + /* Wait for work to finish and/or remove it */ + cancel_work_sync(&osb->dentry_lock_work); + + kill_block_super(sb); +} + static struct file_system_type ocfs2_fs_type = { .owner = THIS_MODULE, .name = "ocfs2", .get_sb = ocfs2_get_sb, /* is this called when we mount * the fs? */ - .kill_sb = kill_block_super, /* set to the generic one - * right now, but do we - * need to change that? */ + .kill_sb = ocfs2_kill_sb, + .fs_flags = FS_REQUIRES_DEV|FS_RENAME_DOES_D_MOVE, .next = NULL }; @@ -1819,6 +1832,12 @@ static void ocfs2_dismount_volume(struct super_block *sb, int mnt_err) debugfs_remove(osb->osb_ctxt); + /* + * Flush inode dropping work queue so that deletes are + * performed while the filesystem is still working + */ + ocfs2_drop_all_dl_inodes(osb); + /* Orphan scan should be stopped as early as possible */ ocfs2_orphan_scan_stop(osb); -- cgit v1.2.3 From d953126a28f97ec965d23c69fd5795854c048f30 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Tue, 21 Jul 2009 19:22:38 -0400 Subject: NFSv4: Fix a problem whereby a buggy server can oops the kernel We just had a case in which a buggy server occasionally returns the wrong attributes during an OPEN call. While the client does catch this sort of condition in nfs4_open_done(), and causes the nfs4_atomic_open() to return -EISDIR, the logic in nfs_atomic_lookup() is broken, since it causes a fallback to an ordinary lookup instead of just returning the error. When the buggy server then returns a regular file for the fallback lookup, the VFS allows the open, and bad things start to happen, since the open file doesn't have any associated NFSv4 state. The fix is firstly to return the EISDIR/ENOTDIR errors immediately, and secondly to ensure that we are always careful when dereferencing the nfs_open_context state pointer. Signed-off-by: Trond Myklebust --- fs/nfs/dir.c | 2 +- fs/nfs/nfs4proc.c | 16 ++++++++++++---- 2 files changed, 13 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c index 38d42c29fb92..32062c33c859 100644 --- a/fs/nfs/dir.c +++ b/fs/nfs/dir.c @@ -1025,12 +1025,12 @@ static struct dentry *nfs_atomic_lookup(struct inode *dir, struct dentry *dentry res = NULL; goto out; /* This turned out not to be a regular file */ - case -EISDIR: case -ENOTDIR: goto no_open; case -ELOOP: if (!(nd->intent.open.flags & O_NOFOLLOW)) goto no_open; + /* case -EISDIR: */ /* case -EINVAL: */ default: goto out; diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index df24f67bca69..6917311f201c 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -4093,15 +4093,23 @@ nfs4_proc_lock(struct file *filp, int cmd, struct file_lock *request) if (request->fl_start < 0 || request->fl_end < 0) return -EINVAL; - if (IS_GETLK(cmd)) - return nfs4_proc_getlk(state, F_GETLK, request); + if (IS_GETLK(cmd)) { + if (state != NULL) + return nfs4_proc_getlk(state, F_GETLK, request); + return 0; + } if (!(IS_SETLK(cmd) || IS_SETLKW(cmd))) return -EINVAL; - if (request->fl_type == F_UNLCK) - return nfs4_proc_unlck(state, cmd, request); + if (request->fl_type == F_UNLCK) { + if (state != NULL) + return nfs4_proc_unlck(state, cmd, request); + return 0; + } + if (state == NULL) + return -ENOLCK; do { status = nfs4_proc_setlk(state, cmd, request); if ((status != -EAGAIN) || IS_SETLK(cmd)) -- cgit v1.2.3 From 1bec1aed1e7e632b3cc43b6807c2b4dcd1572e28 Mon Sep 17 00:00:00 2001 From: Yan Zheng Date: Wed, 22 Jul 2009 09:59:00 -0400 Subject: Btrfs: fix definition of struct btrfs_extent_inline_ref use __le64 instead of u64 in on-disk structure definition. Signed-off-by: Yan Zheng Signed-off-by: Chris Mason --- fs/btrfs/ctree.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index a404ecc53eb1..da0763135bf0 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -483,7 +483,7 @@ struct btrfs_shared_data_ref { struct btrfs_extent_inline_ref { u8 type; - u64 offset; + __le64 offset; } __attribute__ ((__packed__)); /* old style backrefs item */ -- cgit v1.2.3 From bf1fb512a58d7aeb41aaa40d6d2d2d29e08e506a Mon Sep 17 00:00:00 2001 From: Yan Zheng Date: Wed, 22 Jul 2009 09:59:00 -0400 Subject: Btrfs: properly update space information after shrinking device. Change 'goto done' to 'break' for the case of all device extents have been freed, so that the code updates space information will be execute. Signed-off-by: Yan Zheng Signed-off-by: Chris Mason --- fs/btrfs/volumes.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 3ab80e9cd767..f057730a72bb 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -2007,7 +2007,7 @@ int btrfs_shrink_device(struct btrfs_device *device, u64 new_size) goto done; if (ret) { ret = 0; - goto done; + break; } l = path->nodes[0]; @@ -2015,7 +2015,7 @@ int btrfs_shrink_device(struct btrfs_device *device, u64 new_size) btrfs_item_key_to_cpu(l, &key, path->slots[0]); if (key.objectid != device->devid) - goto done; + break; dev_extent = btrfs_item_ptr(l, slot, struct btrfs_dev_extent); length = btrfs_dev_extent_length(l, dev_extent); -- cgit v1.2.3 From e457afec60fdbd86b963d36f4a8a9285088c6043 Mon Sep 17 00:00:00 2001 From: Yan Zheng Date: Wed, 22 Jul 2009 09:59:00 -0400 Subject: Btrfs: fix double increment of path->slots[0] in btrfs_next_leaf if 1 is returned by btrfs_search_slot, the path already points to the first item with 'key > searching key'. So increasing path->slots[0] by one is superfluous in that case. Signed-off-by: Yan Zheng Signed-off-by: Chris Mason --- fs/btrfs/ctree.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/ctree.c b/fs/btrfs/ctree.c index 60a45f3a4e91..7bb66c65ddfd 100644 --- a/fs/btrfs/ctree.c +++ b/fs/btrfs/ctree.c @@ -4146,7 +4146,8 @@ again: * advance the path if there are now more items available. */ if (nritems > 0 && path->slots[0] < nritems - 1) { - path->slots[0]++; + if (ret == 0) + path->slots[0]++; ret = 0; goto done; } -- cgit v1.2.3 From 33c66f430bfa3a033e70470e4c93f967156b696d Mon Sep 17 00:00:00 2001 From: Yan Zheng Date: Wed, 22 Jul 2009 09:59:00 -0400 Subject: Btrfs: fix locking issue in btrfs_find_next_key When walking up the tree, btrfs_find_next_key assumes the upper level tree block is properly locked. This isn't always true even path->keep_locks is 1. This is because btrfs_find_next_key may advance path->slots[] several times instead of only once. When 'path->slots[level] >= btrfs_header_nritems(path->nodes[level])' is found, we can't guarantee the original value of 'path->slots[level]' is 'btrfs_header_nritems(path->nodes[level]) - 1'. If it's not, the tree block at 'level + 1' isn't locked. This patch fixes the issue by explicitly checking the locking state, re-searching the tree if it's not locked. Signed-off-by: Yan Zheng Signed-off-by: Chris Mason --- fs/btrfs/ctree.c | 95 +++++++++++++++++++++++++++++++++------------------ fs/btrfs/relocation.c | 3 ++ 2 files changed, 65 insertions(+), 33 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/ctree.c b/fs/btrfs/ctree.c index 7bb66c65ddfd..fdd423a550d6 100644 --- a/fs/btrfs/ctree.c +++ b/fs/btrfs/ctree.c @@ -1701,6 +1701,7 @@ int btrfs_search_slot(struct btrfs_trans_handle *trans, struct btrfs_root struct extent_buffer *b; int slot; int ret; + int err; int level; int lowest_unlock = 1; u8 lowest_level = 0; @@ -1737,8 +1738,6 @@ again: p->locks[level] = 1; if (cow) { - int wret; - /* * if we don't really need to cow this block * then we don't want to set the path blocking, @@ -1749,12 +1748,12 @@ again: btrfs_set_path_blocking(p); - wret = btrfs_cow_block(trans, root, b, - p->nodes[level + 1], - p->slots[level + 1], &b); - if (wret) { + err = btrfs_cow_block(trans, root, b, + p->nodes[level + 1], + p->slots[level + 1], &b); + if (err) { free_extent_buffer(b); - ret = wret; + ret = err; goto done; } } @@ -1793,41 +1792,45 @@ cow_done: ret = bin_search(b, key, level, &slot); if (level != 0) { - if (ret && slot > 0) + int dec = 0; + if (ret && slot > 0) { + dec = 1; slot -= 1; + } p->slots[level] = slot; - ret = setup_nodes_for_search(trans, root, p, b, level, + err = setup_nodes_for_search(trans, root, p, b, level, ins_len); - if (ret == -EAGAIN) + if (err == -EAGAIN) goto again; - else if (ret) + if (err) { + ret = err; goto done; + } b = p->nodes[level]; slot = p->slots[level]; unlock_up(p, level, lowest_unlock); - /* this is only true while dropping a snapshot */ if (level == lowest_level) { - ret = 0; + if (dec) + p->slots[level]++; goto done; } - ret = read_block_for_search(trans, root, p, + err = read_block_for_search(trans, root, p, &b, level, slot, key); - if (ret == -EAGAIN) + if (err == -EAGAIN) goto again; - - if (ret == -EIO) + if (err) { + ret = err; goto done; + } if (!p->skip_locking) { - int lret; - btrfs_clear_path_blocking(p, NULL); - lret = btrfs_try_spin_lock(b); + err = btrfs_try_spin_lock(b); - if (!lret) { + if (!err) { btrfs_set_path_blocking(p); btrfs_tree_lock(b); btrfs_clear_path_blocking(p, b); @@ -1837,16 +1840,14 @@ cow_done: p->slots[level] = slot; if (ins_len > 0 && btrfs_leaf_free_space(root, b) < ins_len) { - int sret; - btrfs_set_path_blocking(p); - sret = split_leaf(trans, root, key, - p, ins_len, ret == 0); + err = split_leaf(trans, root, key, + p, ins_len, ret == 0); btrfs_clear_path_blocking(p, NULL); - BUG_ON(sret > 0); - if (sret) { - ret = sret; + BUG_ON(err > 0); + if (err) { + ret = err; goto done; } } @@ -4042,10 +4043,9 @@ out: * calling this function. */ int btrfs_find_next_key(struct btrfs_root *root, struct btrfs_path *path, - struct btrfs_key *key, int lowest_level, + struct btrfs_key *key, int level, int cache_only, u64 min_trans) { - int level = lowest_level; int slot; struct extent_buffer *c; @@ -4058,11 +4058,40 @@ int btrfs_find_next_key(struct btrfs_root *root, struct btrfs_path *path, c = path->nodes[level]; next: if (slot >= btrfs_header_nritems(c)) { - level++; - if (level == BTRFS_MAX_LEVEL) + int ret; + int orig_lowest; + struct btrfs_key cur_key; + if (level + 1 >= BTRFS_MAX_LEVEL || + !path->nodes[level + 1]) return 1; - continue; + + if (path->locks[level + 1]) { + level++; + continue; + } + + slot = btrfs_header_nritems(c) - 1; + if (level == 0) + btrfs_item_key_to_cpu(c, &cur_key, slot); + else + btrfs_node_key_to_cpu(c, &cur_key, slot); + + orig_lowest = path->lowest_level; + btrfs_release_path(root, path); + path->lowest_level = level; + ret = btrfs_search_slot(NULL, root, &cur_key, path, + 0, 0); + path->lowest_level = orig_lowest; + if (ret < 0) + return ret; + + c = path->nodes[level]; + slot = path->slots[level]; + if (ret == 0) + slot++; + goto next; } + if (level == 0) btrfs_item_key_to_cpu(c, key, slot); else { diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c index 008397934778..e71264d1c2c9 100644 --- a/fs/btrfs/relocation.c +++ b/fs/btrfs/relocation.c @@ -670,6 +670,8 @@ again: err = ret; goto out; } + if (ret > 0 && path2->slots[level] > 0) + path2->slots[level]--; eb = path2->nodes[level]; WARN_ON(btrfs_node_blockptr(eb, path2->slots[level]) != @@ -1609,6 +1611,7 @@ static noinline_for_stack int merge_reloc_root(struct reloc_control *rc, BUG_ON(level == 0); path->lowest_level = level; ret = btrfs_search_slot(NULL, reloc_root, &key, path, 0, 0); + path->lowest_level = 0; if (ret < 0) { btrfs_free_path(path); return ret; -- cgit v1.2.3 From 4a8c9a62d7f7f058eed4b8a6f2c890a887778093 Mon Sep 17 00:00:00 2001 From: Yan Zheng Date: Wed, 22 Jul 2009 10:07:05 -0400 Subject: Btrfs: make sure all dirty blocks are written at commit time Write dirty block groups may allocate new block, and so may add new delayed back ref. btrfs_run_delayed_refs may make some block groups dirty. commit_cowonly_roots does not handle the recursion properly, and some dirty blocks can be left unwritten at commit time. This patch moves btrfs_run_delayed_refs into the loop that writes dirty block groups, and makes the code not break out of the loop until there are no dirty block groups or delayed back refs. Signed-off-by: Yan Zheng Signed-off-by: Chris Mason --- fs/btrfs/extent-tree.c | 70 +++++++++++++++++++++++++++++--------------------- fs/btrfs/transaction.c | 9 +------ 2 files changed, 42 insertions(+), 37 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index a5aca3997d42..62a332d34fdb 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -2387,13 +2387,29 @@ fail: } +static struct btrfs_block_group_cache * +next_block_group(struct btrfs_root *root, + struct btrfs_block_group_cache *cache) +{ + struct rb_node *node; + spin_lock(&root->fs_info->block_group_cache_lock); + node = rb_next(&cache->cache_node); + btrfs_put_block_group(cache); + if (node) { + cache = rb_entry(node, struct btrfs_block_group_cache, + cache_node); + atomic_inc(&cache->count); + } else + cache = NULL; + spin_unlock(&root->fs_info->block_group_cache_lock); + return cache; +} + int btrfs_write_dirty_block_groups(struct btrfs_trans_handle *trans, struct btrfs_root *root) { - struct btrfs_block_group_cache *cache, *entry; - struct rb_node *n; + struct btrfs_block_group_cache *cache; int err = 0; - int werr = 0; struct btrfs_path *path; u64 last = 0; @@ -2402,39 +2418,35 @@ int btrfs_write_dirty_block_groups(struct btrfs_trans_handle *trans, return -ENOMEM; while (1) { - cache = NULL; - spin_lock(&root->fs_info->block_group_cache_lock); - for (n = rb_first(&root->fs_info->block_group_cache_tree); - n; n = rb_next(n)) { - entry = rb_entry(n, struct btrfs_block_group_cache, - cache_node); - if (entry->dirty) { - cache = entry; - break; - } + if (last == 0) { + err = btrfs_run_delayed_refs(trans, root, + (unsigned long)-1); + BUG_ON(err); } - spin_unlock(&root->fs_info->block_group_cache_lock); - if (!cache) - break; + cache = btrfs_lookup_first_block_group(root->fs_info, last); + while (cache) { + if (cache->dirty) + break; + cache = next_block_group(root, cache); + } + if (!cache) { + if (last == 0) + break; + last = 0; + continue; + } cache->dirty = 0; - last += cache->key.offset; + last = cache->key.objectid + cache->key.offset; - err = write_one_cache_group(trans, root, - path, cache); - /* - * if we fail to write the cache group, we want - * to keep it marked dirty in hopes that a later - * write will work - */ - if (err) { - werr = err; - continue; - } + err = write_one_cache_group(trans, root, path, cache); + BUG_ON(err); + btrfs_put_block_group(cache); } + btrfs_free_path(path); - return werr; + return 0; } int btrfs_extent_readonly(struct btrfs_root *root, u64 bytenr) diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index 2dbf1c1f56ee..81f7124c3051 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -444,9 +444,6 @@ static int update_cowonly_root(struct btrfs_trans_handle *trans, btrfs_write_dirty_block_groups(trans, root); - ret = btrfs_run_delayed_refs(trans, root, (unsigned long)-1); - BUG_ON(ret); - while (1) { old_root_bytenr = btrfs_root_bytenr(&root->root_item); if (old_root_bytenr == root->node->start) @@ -457,9 +454,8 @@ static int update_cowonly_root(struct btrfs_trans_handle *trans, &root->root_key, &root->root_item); BUG_ON(ret); - btrfs_write_dirty_block_groups(trans, root); - ret = btrfs_run_delayed_refs(trans, root, (unsigned long)-1); + ret = btrfs_write_dirty_block_groups(trans, root); BUG_ON(ret); } free_extent_buffer(root->commit_root); @@ -495,9 +491,6 @@ static noinline int commit_cowonly_roots(struct btrfs_trans_handle *trans, root = list_entry(next, struct btrfs_root, dirty_list); update_cowonly_root(trans, root); - - ret = btrfs_run_delayed_refs(trans, root, (unsigned long)-1); - BUG_ON(ret); } return 0; } -- cgit v1.2.3 From 023d43c7b5a23a81fe8afa9f37296f8ed4be11fb Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Tue, 21 Jul 2009 10:09:23 +0200 Subject: lockdep: Fix lockdep annotation for pipe_double_lock() The presumed use of the pipe_double_lock() routine is to lock 2 locks in a deadlock free way by ordering the locks by their address. However it fails to keep the specified lock classes in order and explicitly annotates a deadlock. Rectify this. Signed-off-by: Peter Zijlstra Acked-by: Miklos Szeredi LKML-Reference: <1248163763.15751.11098.camel@twins> --- fs/pipe.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/pipe.c b/fs/pipe.c index f7dd21ad85a6..52c415114838 100644 --- a/fs/pipe.c +++ b/fs/pipe.c @@ -68,8 +68,8 @@ void pipe_double_lock(struct pipe_inode_info *pipe1, pipe_lock_nested(pipe1, I_MUTEX_PARENT); pipe_lock_nested(pipe2, I_MUTEX_CHILD); } else { - pipe_lock_nested(pipe2, I_MUTEX_CHILD); - pipe_lock_nested(pipe1, I_MUTEX_PARENT); + pipe_lock_nested(pipe2, I_MUTEX_PARENT); + pipe_lock_nested(pipe1, I_MUTEX_CHILD); } } -- cgit v1.2.3 From 29c5e8ce01f9dad7e24b99c21e4f836d6b0289e0 Mon Sep 17 00:00:00 2001 From: Julia Lawall Date: Wed, 22 Jul 2009 16:49:00 -0400 Subject: Btrfs: convert nested spin_lock_irqsave to spin_lock If spin_lock_irqsave is called twice in a row with the same second argument, the interrupt state at the point of the second call overwrites the value saved by the first call. Indeed, the second call does not need to save the interrupt state, so it is changed to a simple spin_lock. Signed-off-by: Julia Lawall Signed-off-by: Chris Mason --- fs/btrfs/async-thread.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/async-thread.c b/fs/btrfs/async-thread.c index 6e4f6c50a120..019e8af449ab 100644 --- a/fs/btrfs/async-thread.c +++ b/fs/btrfs/async-thread.c @@ -424,11 +424,11 @@ int btrfs_requeue_work(struct btrfs_work *work) * list */ if (worker->idle) { - spin_lock_irqsave(&worker->workers->lock, flags); + spin_lock(&worker->workers->lock); worker->idle = 0; list_move_tail(&worker->worker_list, &worker->workers->worker_list); - spin_unlock_irqrestore(&worker->workers->lock, flags); + spin_unlock(&worker->workers->lock); } if (!worker->working) { wake = 1; -- cgit v1.2.3 From 3acada49c2794c5aac21849e2ea05790c6dd2faa Mon Sep 17 00:00:00 2001 From: David Woodhouse Date: Wed, 22 Jul 2009 16:49:01 -0400 Subject: Btrfs: Remove broken sanity check from btrfs_rmap_block() It was never actually doing anything anyway (see the loop condition), and it would be difficult to make it work for RAID[56]. Even if it was actually working, it's checking for the wrong thing anyway. Instead of checking whether we list a block which _doesn't_ land at the relevant physical location, it should be checking that we _have_ listed all the logical blocks which refer to the required physical location on all devices. This function is only called from remove_sb_from_cache() to ensure that we reserve the logical blocks which would reside at the same physical location as the superblock copies. So listing more blocks than we need is actually OK. With RAID[56] we're going to throw away an entire stripe for each block we have to ignore, so we _are_ going to list blocks other than the ones which actually contain the superblock. Signed-off-by: David Woodhouse Signed-off-by: Chris Mason --- fs/btrfs/volumes.c | 20 -------------------- 1 file changed, 20 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index f057730a72bb..55c37276a29f 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -2795,26 +2795,6 @@ int btrfs_rmap_block(struct btrfs_mapping_tree *map_tree, } } - for (i = 0; i > nr; i++) { - struct btrfs_multi_bio *multi; - struct btrfs_bio_stripe *stripe; - int ret; - - length = 1; - ret = btrfs_map_block(map_tree, WRITE, buf[i], - &length, &multi, 0); - BUG_ON(ret); - - stripe = multi->stripes; - for (j = 0; j < multi->num_stripes; j++) { - if (stripe->physical >= physical && - physical < stripe->physical + length) - break; - } - BUG_ON(j >= multi->num_stripes); - kfree(multi); - } - *logical = buf; *naddrs = nr; *stripe_len = map->stripe_len; -- cgit v1.2.3 From 33c17ad5717c887568c1de61f15e5d58ed66d189 Mon Sep 17 00:00:00 2001 From: Julia Lawall Date: Wed, 22 Jul 2009 16:49:01 -0400 Subject: Btrfs: adjust NULL test Move the call to BUG_ON to before the dereference of the tested value. Signed-off-by: Julia Lawall Signed-off-by: Chris Mason --- fs/btrfs/inode.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index a48c084f6d3a..3ea827ddf0fe 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -2607,8 +2607,8 @@ noinline int btrfs_truncate_inode_items(struct btrfs_trans_handle *trans, if (root->ref_cows) btrfs_drop_extent_cache(inode, new_size & (~mask), (u64)-1, 0); path = btrfs_alloc_path(); - path->reada = -1; BUG_ON(!path); + path->reada = -1; /* FIXME, add redo link to tree so we don't leak on crash */ key.objectid = inode->i_ino; -- cgit v1.2.3 From c271b492419a18908ba19ee02b231fb305a27023 Mon Sep 17 00:00:00 2001 From: Daniel Cadete Date: Wed, 22 Jul 2009 16:52:13 -0400 Subject: Btrfs: remove of redundant btrfs_header_level This removes the continues call's of btrfs_header_level. One call of btrfs_header_level(c) its enough. Signed-off-by Daniel Cadete Signed-off-by: Chris Mason --- fs/btrfs/print-tree.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/print-tree.c b/fs/btrfs/print-tree.c index 6d6523da0a30..0d126be22b63 100644 --- a/fs/btrfs/print-tree.c +++ b/fs/btrfs/print-tree.c @@ -309,7 +309,7 @@ void btrfs_print_tree(struct btrfs_root *root, struct extent_buffer *c) } printk(KERN_INFO "node %llu level %d total ptrs %d free spc %u\n", (unsigned long long)btrfs_header_bytenr(c), - btrfs_header_level(c), nr, + level, nr, (u32)BTRFS_NODEPTRS_PER_BLOCK(root) - nr); for (i = 0; i < nr; i++) { btrfs_node_key_to_cpu(c, &key, i); @@ -326,10 +326,10 @@ void btrfs_print_tree(struct btrfs_root *root, struct extent_buffer *c) btrfs_level_size(root, level - 1), btrfs_node_ptr_generation(c, i)); if (btrfs_is_leaf(next) && - btrfs_header_level(c) != 1) + level != 1) BUG(); if (btrfs_header_level(next) != - btrfs_header_level(c) - 1) + level - 1) BUG(); btrfs_print_tree(root, next); free_extent_buffer(next); -- cgit v1.2.3 From 83121942b28daffc9526b14b7843d8cdbd3db641 Mon Sep 17 00:00:00 2001 From: David Woodhouse Date: Wed, 22 Jul 2009 16:52:13 -0400 Subject: Btrfs: Fix crash on read failures at mount If the tree roots hit read errors during mount, btrfs is not properly erroring out. We need to check the uptodate bits after reading in the tree root node. Signed-off-by: David Woodhouse Signed-off-by: Chris Mason --- fs/btrfs/disk-io.c | 10 ++++++++++ 1 file changed, 10 insertions(+) (limited to 'fs') diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 0d50d49d990a..55d9d188e693 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -1783,6 +1783,11 @@ struct btrfs_root *open_ctree(struct super_block *sb, btrfs_super_chunk_root(disk_super), blocksize, generation); BUG_ON(!chunk_root->node); + if (!test_bit(EXTENT_BUFFER_UPTODATE, &chunk_root->node->bflags)) { + printk(KERN_WARNING "btrfs: failed to read chunk root on %s\n", + sb->s_id); + goto fail_chunk_root; + } btrfs_set_root_node(&chunk_root->root_item, chunk_root->node); chunk_root->commit_root = btrfs_root_node(chunk_root); @@ -1810,6 +1815,11 @@ struct btrfs_root *open_ctree(struct super_block *sb, blocksize, generation); if (!tree_root->node) goto fail_chunk_root; + if (!test_bit(EXTENT_BUFFER_UPTODATE, &tree_root->node->bflags)) { + printk(KERN_WARNING "btrfs: failed to read tree root on %s\n", + sb->s_id); + goto fail_tree_root; + } btrfs_set_root_node(&tree_root->root_item, tree_root->node); tree_root->commit_root = btrfs_root_node(tree_root); -- cgit v1.2.3 From ce6e7fcd43aab1f77e56aa36936dd7d2d05a1ffa Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Wed, 22 Jul 2009 15:08:58 -0400 Subject: cifs: disable serverino if server doesn't support it A recent regression when dealing with older servers. This bug was introduced when we made serverino the default... When the server can't provide inode numbers, disable it for the mount. Signed-off-by: Jeff Layton Signed-off-by: Steve French --- fs/cifs/inode.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/cifs/inode.c b/fs/cifs/inode.c index 18afe57b2461..b6a47b32f21e 100644 --- a/fs/cifs/inode.c +++ b/fs/cifs/inode.c @@ -513,9 +513,12 @@ int cifs_get_inode_info(struct inode **pinode, cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR); if (rc1) { - /* BB EOPNOSUPP disable SERVER_INUM? */ cFYI(1, ("GetSrvInodeNum rc %d", rc1)); fattr.cf_uniqueid = iunique(sb, ROOT_I); + /* disable serverino if call not supported */ + if (rc1 == -EINVAL) + cifs_sb->mnt_cifs_flags &= + ~CIFS_MOUNT_SERVER_INUM; } } else { fattr.cf_uniqueid = iunique(sb, ROOT_I); -- cgit v1.2.3 From 03aa3a49ad3592a9e4e1ab19c6da3e852288caf1 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Tue, 21 Jul 2009 19:42:03 -0400 Subject: cifs: fix sb->s_maxbytes so that it casts properly to a signed value This off-by-one bug causes sendfile() to not work properly. When a task calls sendfile() on a file on a CIFS filesystem, the syscall returns -1 and sets errno to EOVERFLOW. do_sendfile uses s_maxbytes to verify the returned offset of the file. The problem there is that this value is cast to a signed value (loff_t). When this is done on the s_maxbytes value that cifs uses, it becomes negative and the comparisons against it fail. Even though s_maxbytes is an unsigned value, it seems that it's not OK to set it in such a way that it'll end up negative when it's cast to a signed value. These casts happen in other codepaths besides sendfile too, but the VFS is a little hard to follow in this area and I can't be sure if there are other bugs that this will fix. It's not clear to me why s_maxbytes isn't just declared as loff_t in the first place, but either way we still need to fix these values to make sendfile work properly. This is also an opportunity to replace the magic bit-shift values here with the standard #defines for this. This fixes the reproducer program I have that does a sendfile and will probably also fix the situation where apache is serving from a CIFS share. Acked-by: Johannes Weiner Signed-off-by: Jeff Layton Signed-off-by: Steve French --- fs/cifs/connect.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index 9bb5c8750736..fc44d316d0bb 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -2452,10 +2452,10 @@ try_mount_again: tcon->local_lease = volume_info->local_lease; } if (pSesInfo) { - if (pSesInfo->capabilities & CAP_LARGE_FILES) { - sb->s_maxbytes = (u64) 1 << 63; - } else - sb->s_maxbytes = (u64) 1 << 31; /* 2 GB */ + if (pSesInfo->capabilities & CAP_LARGE_FILES) + sb->s_maxbytes = MAX_LFS_FILESIZE; + else + sb->s_maxbytes = MAX_NON_LFS; } /* BB FIXME fix time_gran to be larger for LANMAN sessions */ -- cgit v1.2.3 From f1230c97978f52268d8c66e6f88e54c3d2092a75 Mon Sep 17 00:00:00 2001 From: Steve French Date: Wed, 22 Jul 2009 23:13:01 +0000 Subject: [CIFS] fix sparse warning Signed-off-by: Steve French --- fs/cifs/inode.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/cifs/inode.c b/fs/cifs/inode.c index b6a47b32f21e..82d83839655e 100644 --- a/fs/cifs/inode.c +++ b/fs/cifs/inode.c @@ -212,7 +212,7 @@ cifs_unix_basic_to_fattr(struct cifs_fattr *fattr, FILE_UNIX_BASIC_INFO *info, * junction to the new submount (ie to setup the fake directory * which represents a DFS referral). */ -void +static void cifs_create_dfs_fattr(struct cifs_fattr *fattr, struct super_block *sb) { struct cifs_sb_info *cifs_sb = CIFS_SB(sb); @@ -388,7 +388,7 @@ static int cifs_sfu_mode(struct cifs_fattr *fattr, const unsigned char *path, } /* Fill a cifs_fattr struct with info from FILE_ALL_INFO */ -void +static void cifs_all_info_to_fattr(struct cifs_fattr *fattr, FILE_ALL_INFO *info, struct cifs_sb_info *cifs_sb, bool adjust_tz) { -- cgit v1.2.3 From 4a19fb11a90fdbbcb3bc02effa036230d035ca28 Mon Sep 17 00:00:00 2001 From: Stefan Bader Date: Thu, 23 Jul 2009 11:26:05 +0200 Subject: jfs: Fix early release of acl in jfs_get_acl BugLink: http://bugs.launchpad.net/ubuntu/+bug/396780 Commit 073aaa1b142461d91f83da66db1184d7c1b1edea "helpers for acl caching + switch to those" introduced new helper functions for acl handling but seems to have introduced a regression for jfs as the acl is released before returning it to the caller, instead of leaving this for the caller to do. This causes the acl object to be used after freeing it, leading to kernel panics in completely different places. Thanks to Christophe Dumez for reporting and bisecting into this. Reported-by: Christophe Dumez Tested-by: Christophe Dumez Signed-off-by: Stefan Bader Acked-by: Andy Whitcroft Signed-off-by: Dave Kleikamp --- fs/jfs/acl.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/jfs/acl.c b/fs/jfs/acl.c index 91fa3ad6e8c2..a29c7c3e3fb8 100644 --- a/fs/jfs/acl.c +++ b/fs/jfs/acl.c @@ -67,10 +67,8 @@ static struct posix_acl *jfs_get_acl(struct inode *inode, int type) acl = posix_acl_from_xattr(value, size); } kfree(value); - if (!IS_ERR(acl)) { + if (!IS_ERR(acl)) set_cached_acl(inode, type, acl); - posix_acl_release(acl); - } return acl; } -- cgit v1.2.3 From 82e12644cf5227dab15201fbcaf0ca6330ebd70f Mon Sep 17 00:00:00 2001 From: Tao Ma Date: Thu, 23 Jul 2009 08:12:58 +0800 Subject: ocfs2: Use ocfs2_rec_clusters in ocfs2_adjust_adjacent_records. In ocfs2_adjust_adjacent_records, we will adjust adjacent records according to the extent_list in the lower level. But actually the lower level tree will either be a leaf or a branch. If we only use ocfs2_is_empty_extent we will meet with some problem if the lower tree is a branch (tree_depth > 1). So use !ocfs2_rec_clusters instead. And actually only the leaf record can have holes. So add a BUG_ON for non-leaf branch. Signed-off-by: Tao Ma Signed-off-by: Joel Becker --- fs/ocfs2/alloc.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ocfs2/alloc.c b/fs/ocfs2/alloc.c index 11085af71247..f9a3e8942669 100644 --- a/fs/ocfs2/alloc.c +++ b/fs/ocfs2/alloc.c @@ -1914,7 +1914,8 @@ static void ocfs2_adjust_adjacent_records(struct ocfs2_extent_rec *left_rec, * immediately to their right. */ left_clusters = le32_to_cpu(right_child_el->l_recs[0].e_cpos); - if (ocfs2_is_empty_extent(&right_child_el->l_recs[0])) { + if (!ocfs2_rec_clusters(right_child_el, &right_child_el->l_recs[0])) { + BUG_ON(right_child_el->l_tree_depth); BUG_ON(le16_to_cpu(right_child_el->l_next_free_rec) <= 1); left_clusters = le32_to_cpu(right_child_el->l_recs[1].e_cpos); } -- cgit v1.2.3 From b57ac2c43e66cb4aa76c9d2d0e9e0e04f19cc5a6 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Wed, 22 Jul 2009 13:17:15 +0200 Subject: ocfs2: Make global quota files blocksize aligned Change i_size of global quota files so that it always remains aligned to block size. This is mainly because the end of quota block may contain checksum (if checksumming is enabled) and it's a bit awkward for it to be "outside" of quota file (and it makes life harder for ocfs2-tools). Signed-off-by: Jan Kara Signed-off-by: Joel Becker --- fs/ocfs2/quota_global.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/ocfs2/quota_global.c b/fs/ocfs2/quota_global.c index edfa60cd155c..a66cb82fd6e6 100644 --- a/fs/ocfs2/quota_global.c +++ b/fs/ocfs2/quota_global.c @@ -211,14 +211,17 @@ ssize_t ocfs2_quota_write(struct super_block *sb, int type, mutex_lock_nested(&gqinode->i_mutex, I_MUTEX_QUOTA); if (gqinode->i_size < off + len) { + loff_t rounded_end = + ocfs2_align_bytes_to_blocks(sb, off + len); + down_write(&OCFS2_I(gqinode)->ip_alloc_sem); - err = ocfs2_extend_no_holes(gqinode, off + len, off); + err = ocfs2_extend_no_holes(gqinode, rounded_end, off); up_write(&OCFS2_I(gqinode)->ip_alloc_sem); if (err < 0) goto out; err = ocfs2_simple_size_update(gqinode, oinfo->dqi_gqi_bh, - off + len); + rounded_end); if (err < 0) goto out; new = 1; -- cgit v1.2.3 From 4b3fa1904c0d192461ebba692e4940a334b74353 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Wed, 22 Jul 2009 13:17:16 +0200 Subject: ocfs2: Mark buffer uptodate before calling ocfs2_journal_access_dq() In a code path extending local quota files we marked new header buffer uptodate only after calling ocfs2_journal_access_dq() which triggers a bug. Fix it and also call ocfs2 variant of the function marking buffer uptodate. Signed-off-by: Jan Kara Signed-off-by: Joel Becker --- fs/ocfs2/quota_local.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ocfs2/quota_local.c b/fs/ocfs2/quota_local.c index 5a460fa82553..b624f9bc9281 100644 --- a/fs/ocfs2/quota_local.c +++ b/fs/ocfs2/quota_local.c @@ -20,6 +20,7 @@ #include "sysfile.h" #include "dlmglue.h" #include "quota.h" +#include "uptodate.h" /* Number of local quota structures per block */ static inline unsigned int ol_quota_entries_per_block(struct super_block *sb) @@ -979,6 +980,8 @@ static struct ocfs2_quota_chunk *ocfs2_local_quota_add_chunk( mlog_errno(status); goto out; } + ocfs2_set_new_buffer_uptodate(lqinode, bh); + dchunk = (struct ocfs2_local_disk_chunk *)bh->b_data; handle = ocfs2_start_trans(OCFS2_SB(sb), 2); @@ -999,7 +1002,6 @@ static struct ocfs2_quota_chunk *ocfs2_local_quota_add_chunk( memset(dchunk->dqc_bitmap, 0, sb->s_blocksize - sizeof(struct ocfs2_local_disk_chunk) - OCFS2_QBLK_RESERVED_SPACE); - set_buffer_uptodate(bh); unlock_buffer(bh); status = ocfs2_journal_dirty(handle, bh); if (status < 0) { -- cgit v1.2.3 From 0e7f387bf386c99e8ee322649fe4fc23b9cccc6c Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Wed, 22 Jul 2009 13:17:17 +0200 Subject: ocfs2: Initialize blocks allocated to local quota file When we extend local quota file, we should initialize data in newly allocated block. Firstly because on recovery we could parse bogus data, secondly so that block checksums are properly computed. Signed-off-by: Jan Kara Signed-off-by: Joel Becker --- fs/ocfs2/quota_local.c | 98 ++++++++++++++++++++++++++++++++++++++++++-------- 1 file changed, 83 insertions(+), 15 deletions(-) (limited to 'fs') diff --git a/fs/ocfs2/quota_local.c b/fs/ocfs2/quota_local.c index b624f9bc9281..9d2993de1082 100644 --- a/fs/ocfs2/quota_local.c +++ b/fs/ocfs2/quota_local.c @@ -941,7 +941,7 @@ static struct ocfs2_quota_chunk *ocfs2_local_quota_add_chunk( struct ocfs2_local_disk_chunk *dchunk; int status; handle_t *handle; - struct buffer_head *bh = NULL; + struct buffer_head *bh = NULL, *dbh = NULL; u64 p_blkno; /* We are protected by dqio_sem so no locking needed */ @@ -965,34 +965,32 @@ static struct ocfs2_quota_chunk *ocfs2_local_quota_add_chunk( mlog_errno(status); goto out; } + handle = ocfs2_start_trans(OCFS2_SB(sb), 3); + if (IS_ERR(handle)) { + status = PTR_ERR(handle); + mlog_errno(status); + goto out; + } + /* Initialize chunk header */ down_read(&OCFS2_I(lqinode)->ip_alloc_sem); status = ocfs2_extent_map_get_blocks(lqinode, oinfo->dqi_blocks, &p_blkno, NULL, NULL); up_read(&OCFS2_I(lqinode)->ip_alloc_sem); if (status < 0) { mlog_errno(status); - goto out; + goto out_trans; } bh = sb_getblk(sb, p_blkno); if (!bh) { status = -ENOMEM; mlog_errno(status); - goto out; + goto out_trans; } - ocfs2_set_new_buffer_uptodate(lqinode, bh); - dchunk = (struct ocfs2_local_disk_chunk *)bh->b_data; - - handle = ocfs2_start_trans(OCFS2_SB(sb), 2); - if (IS_ERR(handle)) { - status = PTR_ERR(handle); - mlog_errno(status); - goto out; - } - + ocfs2_set_new_buffer_uptodate(lqinode, bh); status = ocfs2_journal_access_dq(handle, lqinode, bh, - OCFS2_JOURNAL_ACCESS_WRITE); + OCFS2_JOURNAL_ACCESS_CREATE); if (status < 0) { mlog_errno(status); goto out_trans; @@ -1009,6 +1007,38 @@ static struct ocfs2_quota_chunk *ocfs2_local_quota_add_chunk( goto out_trans; } + /* Initialize new block with structures */ + down_read(&OCFS2_I(lqinode)->ip_alloc_sem); + status = ocfs2_extent_map_get_blocks(lqinode, oinfo->dqi_blocks + 1, + &p_blkno, NULL, NULL); + up_read(&OCFS2_I(lqinode)->ip_alloc_sem); + if (status < 0) { + mlog_errno(status); + goto out_trans; + } + dbh = sb_getblk(sb, p_blkno); + if (!dbh) { + status = -ENOMEM; + mlog_errno(status); + goto out_trans; + } + ocfs2_set_new_buffer_uptodate(lqinode, dbh); + status = ocfs2_journal_access_dq(handle, lqinode, dbh, + OCFS2_JOURNAL_ACCESS_CREATE); + if (status < 0) { + mlog_errno(status); + goto out_trans; + } + lock_buffer(dbh); + memset(dbh->b_data, 0, sb->s_blocksize - OCFS2_QBLK_RESERVED_SPACE); + unlock_buffer(dbh); + status = ocfs2_journal_dirty(handle, dbh); + if (status < 0) { + mlog_errno(status); + goto out_trans; + } + + /* Update local quotafile info */ oinfo->dqi_blocks += 2; oinfo->dqi_chunks++; status = ocfs2_local_write_info(sb, type); @@ -1033,6 +1063,7 @@ out_trans: ocfs2_commit_trans(OCFS2_SB(sb), handle); out: brelse(bh); + brelse(dbh); kmem_cache_free(ocfs2_qf_chunk_cachep, chunk); return ERR_PTR(status); } @@ -1050,6 +1081,8 @@ static struct ocfs2_quota_chunk *ocfs2_extend_local_quota_file( struct ocfs2_local_disk_chunk *dchunk; int epb = ol_quota_entries_per_block(sb); unsigned int chunk_blocks; + struct buffer_head *bh; + u64 p_blkno; int status; handle_t *handle; @@ -1077,12 +1110,46 @@ static struct ocfs2_quota_chunk *ocfs2_extend_local_quota_file( mlog_errno(status); goto out; } - handle = ocfs2_start_trans(OCFS2_SB(sb), 2); + + /* Get buffer from the just added block */ + down_read(&OCFS2_I(lqinode)->ip_alloc_sem); + status = ocfs2_extent_map_get_blocks(lqinode, oinfo->dqi_blocks, + &p_blkno, NULL, NULL); + up_read(&OCFS2_I(lqinode)->ip_alloc_sem); + if (status < 0) { + mlog_errno(status); + goto out; + } + bh = sb_getblk(sb, p_blkno); + if (!bh) { + status = -ENOMEM; + mlog_errno(status); + goto out; + } + ocfs2_set_new_buffer_uptodate(lqinode, bh); + + handle = ocfs2_start_trans(OCFS2_SB(sb), 3); if (IS_ERR(handle)) { status = PTR_ERR(handle); mlog_errno(status); goto out; } + /* Zero created block */ + status = ocfs2_journal_access_dq(handle, lqinode, bh, + OCFS2_JOURNAL_ACCESS_CREATE); + if (status < 0) { + mlog_errno(status); + goto out_trans; + } + lock_buffer(bh); + memset(bh->b_data, 0, sb->s_blocksize); + unlock_buffer(bh); + status = ocfs2_journal_dirty(handle, bh); + if (status < 0) { + mlog_errno(status); + goto out_trans; + } + /* Update chunk header */ status = ocfs2_journal_access_dq(handle, lqinode, chunk->qc_headerbh, OCFS2_JOURNAL_ACCESS_WRITE); if (status < 0) { @@ -1099,6 +1166,7 @@ static struct ocfs2_quota_chunk *ocfs2_extend_local_quota_file( mlog_errno(status); goto out_trans; } + /* Update file header */ oinfo->dqi_blocks++; status = ocfs2_local_write_info(sb, type); if (status < 0) { -- cgit v1.2.3 From 7669f54c55df225cbb2db939a6c9f111445ffc1c Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Wed, 22 Jul 2009 13:17:18 +0200 Subject: ocfs2: Zero out padding of on disk dquot structure Padding fields of on-disk dquot structure were not zeroed. Zero them so that it's easier to use them later. Signed-off-by: Jan Kara Signed-off-by: Joel Becker --- fs/ocfs2/quota_global.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/ocfs2/quota_global.c b/fs/ocfs2/quota_global.c index a66cb82fd6e6..7248db681351 100644 --- a/fs/ocfs2/quota_global.c +++ b/fs/ocfs2/quota_global.c @@ -69,6 +69,7 @@ static void ocfs2_global_mem2diskdqb(void *dp, struct dquot *dquot) d->dqb_curspace = cpu_to_le64(m->dqb_curspace); d->dqb_btime = cpu_to_le64(m->dqb_btime); d->dqb_itime = cpu_to_le64(m->dqb_itime); + d->dqb_pad1 = d->dqb_pad2 = 0; } static int ocfs2_global_is_id(void *dp, struct dquot *dquot) -- cgit v1.2.3 From 1c1d9793ff6720531c0125a28d321f283716e32f Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Wed, 22 Jul 2009 13:17:19 +0200 Subject: ocfs2: Fix initialization of blockcheck stats We just set blockcheck stats to zeros but we should also properly initialize the spinlock there. Signed-off-by: Jan Kara Signed-off-by: Joel Becker --- fs/ocfs2/super.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/ocfs2/super.c b/fs/ocfs2/super.c index f2893878f8ad..b0ee0fdf799a 100644 --- a/fs/ocfs2/super.c +++ b/fs/ocfs2/super.c @@ -777,6 +777,7 @@ static int ocfs2_sb_probe(struct super_block *sb, } di = (struct ocfs2_dinode *) (*bh)->b_data; memset(stats, 0, sizeof(struct ocfs2_blockcheck_stats)); + spin_lock_init(&stats->b_lock); status = ocfs2_verify_volume(di, *bh, blksize, stats); if (status >= 0) goto bail; -- cgit v1.2.3 From 4539f1df25bcd0fdf0d8a5e2c92de6bece83c7a0 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Wed, 22 Jul 2009 13:17:20 +0200 Subject: ocfs2: Remove syncjiff field from quota info syncjiff is just a converted value of syncms. Some places which are updating syncms forgot to update syncjiff as well. Since the conversion is just a simple division / multiplication and it does not happen frequently, just remove the syncjiff field to avoid forgotten conversions. Signed-off-by: Jan Kara Signed-off-by: Joel Becker --- fs/ocfs2/quota.h | 1 - fs/ocfs2/quota_global.c | 5 ++--- 2 files changed, 2 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/ocfs2/quota.h b/fs/ocfs2/quota.h index 7365e2e08706..3fb96fcd4c81 100644 --- a/fs/ocfs2/quota.h +++ b/fs/ocfs2/quota.h @@ -50,7 +50,6 @@ struct ocfs2_mem_dqinfo { unsigned int dqi_chunks; /* Number of chunks in local quota file */ unsigned int dqi_blocks; /* Number of blocks allocated for local quota file */ unsigned int dqi_syncms; /* How often should we sync with other nodes */ - unsigned int dqi_syncjiff; /* Precomputed dqi_syncms in jiffies */ struct list_head dqi_chunk; /* List of chunks */ struct inode *dqi_gqinode; /* Global quota file inode */ struct ocfs2_lock_res dqi_gqlock; /* Lock protecting quota information structure */ diff --git a/fs/ocfs2/quota_global.c b/fs/ocfs2/quota_global.c index 7248db681351..dab45467b5fd 100644 --- a/fs/ocfs2/quota_global.c +++ b/fs/ocfs2/quota_global.c @@ -346,7 +346,6 @@ int ocfs2_global_read_info(struct super_block *sb, int type) info->dqi_bgrace = le32_to_cpu(dinfo.dqi_bgrace); info->dqi_igrace = le32_to_cpu(dinfo.dqi_igrace); oinfo->dqi_syncms = le32_to_cpu(dinfo.dqi_syncms); - oinfo->dqi_syncjiff = msecs_to_jiffies(oinfo->dqi_syncms); oinfo->dqi_gi.dqi_blocks = le32_to_cpu(dinfo.dqi_blocks); oinfo->dqi_gi.dqi_free_blk = le32_to_cpu(dinfo.dqi_free_blk); oinfo->dqi_gi.dqi_free_entry = le32_to_cpu(dinfo.dqi_free_entry); @@ -356,7 +355,7 @@ int ocfs2_global_read_info(struct super_block *sb, int type) oinfo->dqi_gi.dqi_qtree_depth = qtree_depth(&oinfo->dqi_gi); INIT_DELAYED_WORK(&oinfo->dqi_sync_work, qsync_work_fn); queue_delayed_work(ocfs2_quota_wq, &oinfo->dqi_sync_work, - oinfo->dqi_syncjiff); + msecs_to_jiffies(oinfo->dqi_syncms)); out_err: mlog_exit(status); @@ -611,7 +610,7 @@ static void qsync_work_fn(struct work_struct *work) dquot_scan_active(sb, ocfs2_sync_dquot_helper, oinfo->dqi_type); queue_delayed_work(ocfs2_quota_wq, &oinfo->dqi_sync_work, - oinfo->dqi_syncjiff); + msecs_to_jiffies(oinfo->dqi_syncms)); } /* -- cgit v1.2.3 From 0584974a77796581eb3a64b6c5005edac4a95128 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Wed, 22 Jul 2009 13:17:21 +0200 Subject: ocfs2: Define credit counts for quota operations Numbers of needed credits for some quota operations were written as raw numbers. Create appropriate defines instead. Signed-off-by: Jan Kara Signed-off-by: Joel Becker --- fs/ocfs2/journal.h | 13 ++++++++++--- fs/ocfs2/quota_global.c | 18 +++++++++++++----- fs/ocfs2/quota_local.c | 16 ++++++++++++---- 3 files changed, 35 insertions(+), 12 deletions(-) (limited to 'fs') diff --git a/fs/ocfs2/journal.h b/fs/ocfs2/journal.h index 81e8abcd246e..4a4d3b55fd22 100644 --- a/fs/ocfs2/journal.h +++ b/fs/ocfs2/journal.h @@ -330,20 +330,27 @@ int ocfs2_journal_dirty(handle_t *handle, /* extended attribute block update */ #define OCFS2_XATTR_BLOCK_UPDATE_CREDITS 1 +/* Update of a single quota block */ +#define OCFS2_QUOTA_BLOCK_UPDATE_CREDITS 1 + /* global quotafile inode update, data block */ -#define OCFS2_QINFO_WRITE_CREDITS (OCFS2_INODE_UPDATE_CREDITS + 1) +#define OCFS2_QINFO_WRITE_CREDITS (OCFS2_INODE_UPDATE_CREDITS + \ + OCFS2_QUOTA_BLOCK_UPDATE_CREDITS) +#define OCFS2_LOCAL_QINFO_WRITE_CREDITS OCFS2_QUOTA_BLOCK_UPDATE_CREDITS /* * The two writes below can accidentally see global info dirty due * to set_info() quotactl so make them prepared for the writes. */ /* quota data block, global info */ /* Write to local quota file */ -#define OCFS2_QWRITE_CREDITS (OCFS2_QINFO_WRITE_CREDITS + 1) +#define OCFS2_QWRITE_CREDITS (OCFS2_QINFO_WRITE_CREDITS + \ + OCFS2_QUOTA_BLOCK_UPDATE_CREDITS) /* global quota data block, local quota data block, global quota inode, * global quota info */ -#define OCFS2_QSYNC_CREDITS (OCFS2_INODE_UPDATE_CREDITS + 3) +#define OCFS2_QSYNC_CREDITS (OCFS2_QINFO_WRITE_CREDITS + \ + 2 * OCFS2_QUOTA_BLOCK_UPDATE_CREDITS) static inline int ocfs2_quota_trans_credits(struct super_block *sb) { diff --git a/fs/ocfs2/quota_global.c b/fs/ocfs2/quota_global.c index dab45467b5fd..cc8785cf8f61 100644 --- a/fs/ocfs2/quota_global.c +++ b/fs/ocfs2/quota_global.c @@ -648,10 +648,15 @@ int ocfs2_calc_qdel_credits(struct super_block *sb, int type) return 0; oinfo = sb_dqinfo(sb, type)->dqi_priv; - /* We modify tree, leaf block, global info, local chunk header, - * global and local inode */ - return oinfo->dqi_gi.dqi_qtree_depth + 2 + 1 + - 2 * OCFS2_INODE_UPDATE_CREDITS; + /* + * We modify tree, leaf block, global info, local chunk header, + * global and local inode; OCFS2_QINFO_WRITE_CREDITS already + * accounts for inode update + */ + return oinfo->dqi_gi.dqi_qtree_depth + + OCFS2_QINFO_WRITE_CREDITS + + 2 * OCFS2_QUOTA_BLOCK_UPDATE_CREDITS + + OCFS2_INODE_UPDATE_CREDITS; } static int ocfs2_release_dquot(struct dquot *dquot) @@ -701,7 +706,10 @@ int ocfs2_calc_qinit_credits(struct super_block *sb, int type) * global file we can modify info, tree and leaf block */ return ocfs2_calc_extend_credits(sb, &lfe->id2.i_list, 0) + ocfs2_calc_extend_credits(sb, &gfe->id2.i_list, 0) + - 3 + oinfo->dqi_gi.dqi_qtree_depth + 2; + OCFS2_LOCAL_QINFO_WRITE_CREDITS + + 2 * OCFS2_QUOTA_BLOCK_UPDATE_CREDITS + + oinfo->dqi_gi.dqi_qtree_depth + + 2 * OCFS2_QUOTA_BLOCK_UPDATE_CREDITS; } static int ocfs2_acquire_dquot(struct dquot *dquot) diff --git a/fs/ocfs2/quota_local.c b/fs/ocfs2/quota_local.c index 9d2993de1082..bdb09cb6e1fe 100644 --- a/fs/ocfs2/quota_local.c +++ b/fs/ocfs2/quota_local.c @@ -101,7 +101,8 @@ static int ocfs2_modify_bh(struct inode *inode, struct buffer_head *bh, handle_t *handle; int status; - handle = ocfs2_start_trans(OCFS2_SB(sb), 1); + handle = ocfs2_start_trans(OCFS2_SB(sb), + OCFS2_QUOTA_BLOCK_UPDATE_CREDITS); if (IS_ERR(handle)) { status = PTR_ERR(handle); mlog_errno(status); @@ -611,7 +612,8 @@ int ocfs2_finish_quota_recovery(struct ocfs2_super *osb, goto out_bh; /* Mark quota file as clean if we are recovering quota file of * some other node. */ - handle = ocfs2_start_trans(osb, 1); + handle = ocfs2_start_trans(osb, + OCFS2_LOCAL_QINFO_WRITE_CREDITS); if (IS_ERR(handle)) { status = PTR_ERR(handle); mlog_errno(status); @@ -965,7 +967,10 @@ static struct ocfs2_quota_chunk *ocfs2_local_quota_add_chunk( mlog_errno(status); goto out; } - handle = ocfs2_start_trans(OCFS2_SB(sb), 3); + /* Local quota info and two new blocks we initialize */ + handle = ocfs2_start_trans(OCFS2_SB(sb), + OCFS2_LOCAL_QINFO_WRITE_CREDITS + + 2 * OCFS2_QUOTA_BLOCK_UPDATE_CREDITS); if (IS_ERR(handle)) { status = PTR_ERR(handle); mlog_errno(status); @@ -1128,7 +1133,10 @@ static struct ocfs2_quota_chunk *ocfs2_extend_local_quota_file( } ocfs2_set_new_buffer_uptodate(lqinode, bh); - handle = ocfs2_start_trans(OCFS2_SB(sb), 3); + /* Local quota info, chunk header and the new block we initialize */ + handle = ocfs2_start_trans(OCFS2_SB(sb), + OCFS2_LOCAL_QINFO_WRITE_CREDITS + + 2 * OCFS2_QUOTA_BLOCK_UPDATE_CREDITS); if (IS_ERR(handle)) { status = PTR_ERR(handle); mlog_errno(status); -- cgit v1.2.3 From 963030817060e4f109be1993b9ae8f81dbf5e11a Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Mon, 13 Jul 2009 21:29:25 -0400 Subject: Btrfs: use hybrid extents+bitmap rb tree for free space Currently btrfs has a problem where it can use a ridiculous amount of RAM simply tracking free space. As free space gets fragmented, we end up with thousands of entries on an rb-tree per block group, which usually spans 1 gig of area. Since we currently don't ever flush free space cache back to disk this gets to be a bit unweildly on large fs's with lots of fragmentation. This patch solves this problem by using PAGE_SIZE bitmaps for parts of the free space cache. Initially we calculate a threshold of extent entries we can handle, which is however many extent entries we can cram into 16k of ram. The maximum amount of RAM that should ever be used to track 1 gigabyte of diskspace will be 32k of RAM, which scales much better than we did before. Once we pass the extent threshold, we start adding bitmaps and using those instead for tracking the free space. This patch also makes it so that any free space thats less than 4 * sectorsize we go ahead and put into a bitmap. This is nice since we try and allocate out of the front of a block group, so if the front of a block group is heavily fragmented and then has a huge chunk of free space at the end, we go ahead and add the fragmented areas to bitmaps and use a normal extent entry to track the big chunk at the back of the block group. I've also taken the opportunity to revamp how we search for free space. Previously we indexed free space via an offset indexed rb tree and a bytes indexed rb tree. I've dropped the bytes indexed rb tree and use only the offset indexed rb tree. This cuts the number of tree operations we were doing previously down by half, and gives us a little bit of a better allocation pattern since we will always start from a specific offset and search forward from there, instead of searching for the size we need and try and get it as close as possible to the offset we want. I've given this a healthy amount of testing pre-new format stuff, as well as post-new format stuff. I've booted up my fedora box which is installed on btrfs with this patch and ran with it for a few days without issues. I've not seen any performance regressions in any of my tests. Since the last patch Yan Zheng fixed a problem where we could have overlapping entries, so updating their offset inline would cause problems. Thanks, Signed-off-by: Josef Bacik Signed-off-by: Chris Mason --- fs/btrfs/ctree.h | 8 +- fs/btrfs/extent-tree.c | 25 +- fs/btrfs/free-space-cache.c | 1001 ++++++++++++++++++++++++++++++++++--------- fs/btrfs/free-space-cache.h | 8 + 4 files changed, 826 insertions(+), 216 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index da0763135bf0..0cbf3491bb7c 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -709,6 +709,9 @@ struct btrfs_free_cluster { /* first extent starting offset */ u64 window_start; + /* if this cluster simply points at a bitmap in the block group */ + bool points_to_bitmap; + struct btrfs_block_group_cache *block_group; /* * when a cluster is allocated from a block group, we put the @@ -726,6 +729,10 @@ struct btrfs_block_group_cache { u64 pinned; u64 reserved; u64 flags; + u64 sectorsize; + int extents_thresh; + int free_extents; + int total_bitmaps; int cached; int ro; int dirty; @@ -734,7 +741,6 @@ struct btrfs_block_group_cache { /* free space cache stuff */ spinlock_t tree_lock; - struct rb_root free_space_bytes; struct rb_root free_space_offset; /* block group cache stuff */ diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 62a332d34fdb..98697be6bdde 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -3649,7 +3649,6 @@ refill_cluster: goto loop; checks: search_start = stripe_align(root, offset); - /* move on to the next group */ if (search_start + num_bytes >= search_end) { btrfs_add_free_space(block_group, offset, num_bytes); @@ -7040,6 +7039,16 @@ int btrfs_read_block_groups(struct btrfs_root *root) mutex_init(&cache->cache_mutex); INIT_LIST_HEAD(&cache->list); INIT_LIST_HEAD(&cache->cluster_list); + cache->sectorsize = root->sectorsize; + + /* + * we only want to have 32k of ram per block group for keeping + * track of free space, and if we pass 1/2 of that we want to + * start converting things over to using bitmaps + */ + cache->extents_thresh = ((1024 * 32) / 2) / + sizeof(struct btrfs_free_space); + read_extent_buffer(leaf, &cache->item, btrfs_item_ptr_offset(leaf, path->slots[0]), sizeof(cache->item)); @@ -7091,6 +7100,15 @@ int btrfs_make_block_group(struct btrfs_trans_handle *trans, cache->key.objectid = chunk_offset; cache->key.offset = size; cache->key.type = BTRFS_BLOCK_GROUP_ITEM_KEY; + cache->sectorsize = root->sectorsize; + + /* + * we only want to have 32k of ram per block group for keeping track + * of free space, and if we pass 1/2 of that we want to start + * converting things over to using bitmaps + */ + cache->extents_thresh = ((1024 * 32) / 2) / + sizeof(struct btrfs_free_space); atomic_set(&cache->count, 1); spin_lock_init(&cache->lock); spin_lock_init(&cache->tree_lock); @@ -7103,6 +7121,11 @@ int btrfs_make_block_group(struct btrfs_trans_handle *trans, cache->flags = type; btrfs_set_block_group_flags(&cache->item, type); + cache->cached = 1; + ret = btrfs_add_free_space(cache, chunk_offset, size); + BUG_ON(ret); + remove_sb_from_cache(root, cache); + ret = update_space_info(root->fs_info, cache->flags, size, bytes_used, &cache->space_info); BUG_ON(ret); diff --git a/fs/btrfs/free-space-cache.c b/fs/btrfs/free-space-cache.c index 4538e48581a5..ab8cad8b46c9 100644 --- a/fs/btrfs/free-space-cache.c +++ b/fs/btrfs/free-space-cache.c @@ -16,45 +16,46 @@ * Boston, MA 021110-1307, USA. */ +#include #include +#include #include "ctree.h" #include "free-space-cache.h" #include "transaction.h" -struct btrfs_free_space { - struct rb_node bytes_index; - struct rb_node offset_index; - u64 offset; - u64 bytes; -}; +#define BITS_PER_BITMAP (PAGE_CACHE_SIZE * 8) +#define MAX_CACHE_BYTES_PER_GIG (32 * 1024) -static int tree_insert_offset(struct rb_root *root, u64 offset, - struct rb_node *node) +static inline unsigned long offset_to_bit(u64 bitmap_start, u64 sectorsize, + u64 offset) { - struct rb_node **p = &root->rb_node; - struct rb_node *parent = NULL; - struct btrfs_free_space *info; + BUG_ON(offset < bitmap_start); + offset -= bitmap_start; + return (unsigned long)(div64_u64(offset, sectorsize)); +} - while (*p) { - parent = *p; - info = rb_entry(parent, struct btrfs_free_space, offset_index); +static inline unsigned long bytes_to_bits(u64 bytes, u64 sectorsize) +{ + return (unsigned long)(div64_u64(bytes, sectorsize)); +} - if (offset < info->offset) - p = &(*p)->rb_left; - else if (offset > info->offset) - p = &(*p)->rb_right; - else - return -EEXIST; - } +static inline u64 offset_to_bitmap(struct btrfs_block_group_cache *block_group, + u64 offset) +{ + u64 bitmap_start; + u64 bytes_per_bitmap; - rb_link_node(node, parent, p); - rb_insert_color(node, root); + bytes_per_bitmap = BITS_PER_BITMAP * block_group->sectorsize; + bitmap_start = offset - block_group->key.objectid; + bitmap_start = div64_u64(bitmap_start, bytes_per_bitmap); + bitmap_start *= bytes_per_bitmap; + bitmap_start += block_group->key.objectid; - return 0; + return bitmap_start; } -static int tree_insert_bytes(struct rb_root *root, u64 bytes, - struct rb_node *node) +static int tree_insert_offset(struct rb_root *root, u64 offset, + struct rb_node *node, int bitmap) { struct rb_node **p = &root->rb_node; struct rb_node *parent = NULL; @@ -62,12 +63,34 @@ static int tree_insert_bytes(struct rb_root *root, u64 bytes, while (*p) { parent = *p; - info = rb_entry(parent, struct btrfs_free_space, bytes_index); + info = rb_entry(parent, struct btrfs_free_space, offset_index); - if (bytes < info->bytes) + if (offset < info->offset) { p = &(*p)->rb_left; - else + } else if (offset > info->offset) { p = &(*p)->rb_right; + } else { + /* + * we could have a bitmap entry and an extent entry + * share the same offset. If this is the case, we want + * the extent entry to always be found first if we do a + * linear search through the tree, since we want to have + * the quickest allocation time, and allocating from an + * extent is faster than allocating from a bitmap. So + * if we're inserting a bitmap and we find an entry at + * this offset, we want to go right, or after this entry + * logically. If we are inserting an extent and we've + * found a bitmap, we want to go left, or before + * logically. + */ + if (bitmap) { + WARN_ON(info->bitmap); + p = &(*p)->rb_right; + } else { + WARN_ON(!info->bitmap); + p = &(*p)->rb_left; + } + } } rb_link_node(node, parent, p); @@ -79,110 +102,142 @@ static int tree_insert_bytes(struct rb_root *root, u64 bytes, /* * searches the tree for the given offset. * - * fuzzy == 1: this is used for allocations where we are given a hint of where - * to look for free space. Because the hint may not be completely on an offset - * mark, or the hint may no longer point to free space we need to fudge our - * results a bit. So we look for free space starting at or after offset with at - * least bytes size. We prefer to find as close to the given offset as we can. - * Also if the offset is within a free space range, then we will return the free - * space that contains the given offset, which means we can return a free space - * chunk with an offset before the provided offset. - * - * fuzzy == 0: this is just a normal tree search. Give us the free space that - * starts at the given offset which is at least bytes size, and if its not there - * return NULL. + * fuzzy - If this is set, then we are trying to make an allocation, and we just + * want a section that has at least bytes size and comes at or after the given + * offset. */ -static struct btrfs_free_space *tree_search_offset(struct rb_root *root, - u64 offset, u64 bytes, - int fuzzy) +static struct btrfs_free_space * +tree_search_offset(struct btrfs_block_group_cache *block_group, + u64 offset, int bitmap_only, int fuzzy) { - struct rb_node *n = root->rb_node; - struct btrfs_free_space *entry, *ret = NULL; + struct rb_node *n = block_group->free_space_offset.rb_node; + struct btrfs_free_space *entry, *prev = NULL; + + /* find entry that is closest to the 'offset' */ + while (1) { + if (!n) { + entry = NULL; + break; + } - while (n) { entry = rb_entry(n, struct btrfs_free_space, offset_index); + prev = entry; - if (offset < entry->offset) { - if (fuzzy && - (!ret || entry->offset < ret->offset) && - (bytes <= entry->bytes)) - ret = entry; + if (offset < entry->offset) n = n->rb_left; - } else if (offset > entry->offset) { - if (fuzzy && - (entry->offset + entry->bytes - 1) >= offset && - bytes <= entry->bytes) { - ret = entry; - break; - } + else if (offset > entry->offset) n = n->rb_right; - } else { - if (bytes > entry->bytes) { - n = n->rb_right; - continue; - } - ret = entry; + else break; - } } - return ret; -} - -/* - * return a chunk at least bytes size, as close to offset that we can get. - */ -static struct btrfs_free_space *tree_search_bytes(struct rb_root *root, - u64 offset, u64 bytes) -{ - struct rb_node *n = root->rb_node; - struct btrfs_free_space *entry, *ret = NULL; + if (bitmap_only) { + if (!entry) + return NULL; + if (entry->bitmap) + return entry; - while (n) { - entry = rb_entry(n, struct btrfs_free_space, bytes_index); + /* + * bitmap entry and extent entry may share same offset, + * in that case, bitmap entry comes after extent entry. + */ + n = rb_next(n); + if (!n) + return NULL; + entry = rb_entry(n, struct btrfs_free_space, offset_index); + if (entry->offset != offset) + return NULL; - if (bytes < entry->bytes) { + WARN_ON(!entry->bitmap); + return entry; + } else if (entry) { + if (entry->bitmap) { /* - * We prefer to get a hole size as close to the size we - * are asking for so we don't take small slivers out of - * huge holes, but we also want to get as close to the - * offset as possible so we don't have a whole lot of - * fragmentation. + * if previous extent entry covers the offset, + * we should return it instead of the bitmap entry */ - if (offset <= entry->offset) { - if (!ret) - ret = entry; - else if (entry->bytes < ret->bytes) - ret = entry; - else if (entry->offset < ret->offset) - ret = entry; + n = &entry->offset_index; + while (1) { + n = rb_prev(n); + if (!n) + break; + prev = rb_entry(n, struct btrfs_free_space, + offset_index); + if (!prev->bitmap) { + if (prev->offset + prev->bytes > offset) + entry = prev; + break; + } } - n = n->rb_left; - } else if (bytes > entry->bytes) { - n = n->rb_right; + } + return entry; + } + + if (!prev) + return NULL; + + /* find last entry before the 'offset' */ + entry = prev; + if (entry->offset > offset) { + n = rb_prev(&entry->offset_index); + if (n) { + entry = rb_entry(n, struct btrfs_free_space, + offset_index); + BUG_ON(entry->offset > offset); } else { - /* - * Ok we may have multiple chunks of the wanted size, - * so we don't want to take the first one we find, we - * want to take the one closest to our given offset, so - * keep searching just in case theres a better match. - */ - n = n->rb_right; - if (offset > entry->offset) - continue; - else if (!ret || entry->offset < ret->offset) - ret = entry; + if (fuzzy) + return entry; + else + return NULL; } } - return ret; + if (entry->bitmap) { + n = &entry->offset_index; + while (1) { + n = rb_prev(n); + if (!n) + break; + prev = rb_entry(n, struct btrfs_free_space, + offset_index); + if (!prev->bitmap) { + if (prev->offset + prev->bytes > offset) + return prev; + break; + } + } + if (entry->offset + BITS_PER_BITMAP * + block_group->sectorsize > offset) + return entry; + } else if (entry->offset + entry->bytes > offset) + return entry; + + if (!fuzzy) + return NULL; + + while (1) { + if (entry->bitmap) { + if (entry->offset + BITS_PER_BITMAP * + block_group->sectorsize > offset) + break; + } else { + if (entry->offset + entry->bytes > offset) + break; + } + + n = rb_next(&entry->offset_index); + if (!n) + return NULL; + entry = rb_entry(n, struct btrfs_free_space, offset_index); + } + return entry; } static void unlink_free_space(struct btrfs_block_group_cache *block_group, struct btrfs_free_space *info) { rb_erase(&info->offset_index, &block_group->free_space_offset); - rb_erase(&info->bytes_index, &block_group->free_space_bytes); + block_group->free_extents--; } static int link_free_space(struct btrfs_block_group_cache *block_group, @@ -190,17 +245,311 @@ static int link_free_space(struct btrfs_block_group_cache *block_group, { int ret = 0; - - BUG_ON(!info->bytes); + BUG_ON(!info->bitmap && !info->bytes); ret = tree_insert_offset(&block_group->free_space_offset, info->offset, - &info->offset_index); + &info->offset_index, (info->bitmap != NULL)); if (ret) return ret; - ret = tree_insert_bytes(&block_group->free_space_bytes, info->bytes, - &info->bytes_index); - if (ret) - return ret; + block_group->free_extents++; + return ret; +} + +static void recalculate_thresholds(struct btrfs_block_group_cache *block_group) +{ + u64 max_bytes, possible_bytes; + + /* + * The goal is to keep the total amount of memory used per 1gb of space + * at or below 32k, so we need to adjust how much memory we allow to be + * used by extent based free space tracking + */ + max_bytes = MAX_CACHE_BYTES_PER_GIG * + (div64_u64(block_group->key.offset, 1024 * 1024 * 1024)); + + possible_bytes = (block_group->total_bitmaps * PAGE_CACHE_SIZE) + + (sizeof(struct btrfs_free_space) * + block_group->extents_thresh); + + if (possible_bytes > max_bytes) { + int extent_bytes = max_bytes - + (block_group->total_bitmaps * PAGE_CACHE_SIZE); + + if (extent_bytes <= 0) { + block_group->extents_thresh = 0; + return; + } + + block_group->extents_thresh = extent_bytes / + (sizeof(struct btrfs_free_space)); + } +} + +static void bitmap_clear_bits(struct btrfs_free_space *info, u64 offset, u64 bytes, + u64 sectorsize) +{ + unsigned long start, end; + unsigned long i; + + start = offset_to_bit(info->offset, sectorsize, offset); + end = start + bytes_to_bits(bytes, sectorsize); + BUG_ON(end > BITS_PER_BITMAP); + + for (i = start; i < end; i++) + clear_bit(i, info->bitmap); + + info->bytes -= bytes; +} + +static void bitmap_set_bits(struct btrfs_free_space *info, u64 offset, u64 bytes, + u64 sectorsize) +{ + unsigned long start, end; + unsigned long i; + + start = offset_to_bit(info->offset, sectorsize, offset); + end = start + bytes_to_bits(bytes, sectorsize); + BUG_ON(end > BITS_PER_BITMAP); + + for (i = start; i < end; i++) + set_bit(i, info->bitmap); + + info->bytes += bytes; +} + +static int search_bitmap(struct btrfs_block_group_cache *block_group, + struct btrfs_free_space *bitmap_info, u64 *offset, + u64 *bytes) +{ + unsigned long found_bits = 0; + unsigned long bits, i; + unsigned long next_zero; + + i = offset_to_bit(bitmap_info->offset, block_group->sectorsize, + max_t(u64, *offset, bitmap_info->offset)); + bits = bytes_to_bits(*bytes, block_group->sectorsize); + + for (i = find_next_bit(bitmap_info->bitmap, BITS_PER_BITMAP, i); + i < BITS_PER_BITMAP; + i = find_next_bit(bitmap_info->bitmap, BITS_PER_BITMAP, i + 1)) { + next_zero = find_next_zero_bit(bitmap_info->bitmap, + BITS_PER_BITMAP, i); + if ((next_zero - i) >= bits) { + found_bits = next_zero - i; + break; + } + i = next_zero; + } + + if (found_bits) { + *offset = (u64)(i * block_group->sectorsize) + + bitmap_info->offset; + *bytes = (u64)(found_bits) * block_group->sectorsize; + return 0; + } + + return -1; +} + +static struct btrfs_free_space *find_free_space(struct btrfs_block_group_cache + *block_group, u64 *offset, + u64 *bytes, int debug) +{ + struct btrfs_free_space *entry; + struct rb_node *node; + int ret; + + if (!block_group->free_space_offset.rb_node) + return NULL; + + entry = tree_search_offset(block_group, + offset_to_bitmap(block_group, *offset), + 0, 1); + if (!entry) + return NULL; + + for (node = &entry->offset_index; node; node = rb_next(node)) { + entry = rb_entry(node, struct btrfs_free_space, offset_index); + if (entry->bytes < *bytes) + continue; + + if (entry->bitmap) { + ret = search_bitmap(block_group, entry, offset, bytes); + if (!ret) + return entry; + continue; + } + + *offset = entry->offset; + *bytes = entry->bytes; + return entry; + } + + return NULL; +} + +static void add_new_bitmap(struct btrfs_block_group_cache *block_group, + struct btrfs_free_space *info, u64 offset) +{ + u64 bytes_per_bg = BITS_PER_BITMAP * block_group->sectorsize; + int max_bitmaps = (int)div64_u64(block_group->key.offset + + bytes_per_bg - 1, bytes_per_bg); + BUG_ON(block_group->total_bitmaps >= max_bitmaps); + + info->offset = offset_to_bitmap(block_group, offset); + link_free_space(block_group, info); + block_group->total_bitmaps++; + + recalculate_thresholds(block_group); +} + +static noinline int remove_from_bitmap(struct btrfs_block_group_cache *block_group, + struct btrfs_free_space *bitmap_info, + u64 *offset, u64 *bytes) +{ + u64 end; + +again: + end = bitmap_info->offset + + (u64)(BITS_PER_BITMAP * block_group->sectorsize) - 1; + + if (*offset > bitmap_info->offset && *offset + *bytes > end) { + bitmap_clear_bits(bitmap_info, *offset, + end - *offset + 1, block_group->sectorsize); + *bytes -= end - *offset + 1; + *offset = end + 1; + } else if (*offset >= bitmap_info->offset && *offset + *bytes <= end) { + bitmap_clear_bits(bitmap_info, *offset, + *bytes, block_group->sectorsize); + *bytes = 0; + } + + if (*bytes) { + if (!bitmap_info->bytes) { + unlink_free_space(block_group, bitmap_info); + kfree(bitmap_info->bitmap); + kfree(bitmap_info); + block_group->total_bitmaps--; + recalculate_thresholds(block_group); + } + + bitmap_info = tree_search_offset(block_group, + offset_to_bitmap(block_group, + *offset), + 1, 0); + if (!bitmap_info) + return -EINVAL; + + if (!bitmap_info->bitmap) + return -EAGAIN; + + goto again; + } else if (!bitmap_info->bytes) { + unlink_free_space(block_group, bitmap_info); + kfree(bitmap_info->bitmap); + kfree(bitmap_info); + block_group->total_bitmaps--; + recalculate_thresholds(block_group); + } + + return 0; +} + +static int insert_into_bitmap(struct btrfs_block_group_cache *block_group, + struct btrfs_free_space *info) +{ + struct btrfs_free_space *bitmap_info; + int added = 0; + u64 bytes, offset, end; + int ret; + + /* + * If we are below the extents threshold then we can add this as an + * extent, and don't have to deal with the bitmap + */ + if (block_group->free_extents < block_group->extents_thresh && + info->bytes > block_group->sectorsize * 4) + return 0; + + /* + * some block groups are so tiny they can't be enveloped by a bitmap, so + * don't even bother to create a bitmap for this + */ + if (BITS_PER_BITMAP * block_group->sectorsize > + block_group->key.offset) + return 0; + + bytes = info->bytes; + offset = info->offset; + +again: + bitmap_info = tree_search_offset(block_group, + offset_to_bitmap(block_group, offset), + 1, 0); + if (!bitmap_info) { + BUG_ON(added); + goto new_bitmap; + } + + end = bitmap_info->offset + + (u64)(BITS_PER_BITMAP * block_group->sectorsize); + + if (offset >= bitmap_info->offset && offset + bytes > end) { + bitmap_set_bits(bitmap_info, offset, end - offset, + block_group->sectorsize); + bytes -= end - offset; + offset = end; + added = 0; + } else if (offset >= bitmap_info->offset && offset + bytes <= end) { + bitmap_set_bits(bitmap_info, offset, bytes, + block_group->sectorsize); + bytes = 0; + } else { + BUG(); + } + + if (!bytes) { + ret = 1; + goto out; + } else + goto again; + +new_bitmap: + if (info && info->bitmap) { + add_new_bitmap(block_group, info, offset); + added = 1; + info = NULL; + goto again; + } else { + spin_unlock(&block_group->tree_lock); + + /* no pre-allocated info, allocate a new one */ + if (!info) { + info = kzalloc(sizeof(struct btrfs_free_space), + GFP_NOFS); + if (!info) { + spin_lock(&block_group->tree_lock); + ret = -ENOMEM; + goto out; + } + } + + /* allocate the bitmap */ + info->bitmap = kzalloc(PAGE_CACHE_SIZE, GFP_NOFS); + spin_lock(&block_group->tree_lock); + if (!info->bitmap) { + ret = -ENOMEM; + goto out; + } + goto again; + } + +out: + if (info) { + if (info->bitmap) + kfree(info->bitmap); + kfree(info); + } return ret; } @@ -208,8 +557,8 @@ static int link_free_space(struct btrfs_block_group_cache *block_group, int btrfs_add_free_space(struct btrfs_block_group_cache *block_group, u64 offset, u64 bytes) { - struct btrfs_free_space *right_info; - struct btrfs_free_space *left_info; + struct btrfs_free_space *right_info = NULL; + struct btrfs_free_space *left_info = NULL; struct btrfs_free_space *info = NULL; int ret = 0; @@ -227,18 +576,38 @@ int btrfs_add_free_space(struct btrfs_block_group_cache *block_group, * are adding, if there is remove that struct and add a new one to * cover the entire range */ - right_info = tree_search_offset(&block_group->free_space_offset, - offset+bytes, 0, 0); - left_info = tree_search_offset(&block_group->free_space_offset, - offset-1, 0, 1); + right_info = tree_search_offset(block_group, offset + bytes, 0, 0); + if (right_info && rb_prev(&right_info->offset_index)) + left_info = rb_entry(rb_prev(&right_info->offset_index), + struct btrfs_free_space, offset_index); + else + left_info = tree_search_offset(block_group, offset - 1, 0, 0); - if (right_info) { + /* + * If there was no extent directly to the left or right of this new + * extent then we know we're going to have to allocate a new extent, so + * before we do that see if we need to drop this into a bitmap + */ + if ((!left_info || left_info->bitmap) && + (!right_info || right_info->bitmap)) { + ret = insert_into_bitmap(block_group, info); + + if (ret < 0) { + goto out; + } else if (ret) { + ret = 0; + goto out; + } + } + + if (right_info && !right_info->bitmap) { unlink_free_space(block_group, right_info); info->bytes += right_info->bytes; kfree(right_info); } - if (left_info && left_info->offset + left_info->bytes == offset) { + if (left_info && !left_info->bitmap && + left_info->offset + left_info->bytes == offset) { unlink_free_space(block_group, left_info); info->offset = left_info->offset; info->bytes += left_info->bytes; @@ -248,11 +617,11 @@ int btrfs_add_free_space(struct btrfs_block_group_cache *block_group, ret = link_free_space(block_group, info); if (ret) kfree(info); - +out: spin_unlock(&block_group->tree_lock); if (ret) { - printk(KERN_ERR "btrfs: unable to add free space :%d\n", ret); + printk(KERN_CRIT "btrfs: unable to add free space :%d\n", ret); BUG_ON(ret == -EEXIST); } @@ -263,40 +632,65 @@ int btrfs_remove_free_space(struct btrfs_block_group_cache *block_group, u64 offset, u64 bytes) { struct btrfs_free_space *info; + struct btrfs_free_space *next_info = NULL; int ret = 0; spin_lock(&block_group->tree_lock); - info = tree_search_offset(&block_group->free_space_offset, offset, 0, - 1); - if (info && info->offset == offset) { - if (info->bytes < bytes) { - printk(KERN_ERR "Found free space at %llu, size %llu," - "trying to use %llu\n", - (unsigned long long)info->offset, - (unsigned long long)info->bytes, - (unsigned long long)bytes); +again: + info = tree_search_offset(block_group, offset, 0, 0); + if (!info) { + WARN_ON(1); + goto out_lock; + } + + if (info->bytes < bytes && rb_next(&info->offset_index)) { + u64 end; + next_info = rb_entry(rb_next(&info->offset_index), + struct btrfs_free_space, + offset_index); + + if (next_info->bitmap) + end = next_info->offset + BITS_PER_BITMAP * + block_group->sectorsize - 1; + else + end = next_info->offset + next_info->bytes; + + if (next_info->bytes < bytes || + next_info->offset > offset || offset > end) { + printk(KERN_CRIT "Found free space at %llu, size %llu," + " trying to use %llu\n", + (unsigned long long)info->offset, + (unsigned long long)info->bytes, + (unsigned long long)bytes); WARN_ON(1); ret = -EINVAL; - spin_unlock(&block_group->tree_lock); - goto out; + goto out_lock; } - unlink_free_space(block_group, info); - if (info->bytes == bytes) { - kfree(info); - spin_unlock(&block_group->tree_lock); - goto out; + info = next_info; + } + + if (info->bytes == bytes) { + unlink_free_space(block_group, info); + if (info->bitmap) { + kfree(info->bitmap); + block_group->total_bitmaps--; } + kfree(info); + goto out_lock; + } + if (!info->bitmap && info->offset == offset) { + unlink_free_space(block_group, info); info->offset += bytes; info->bytes -= bytes; + link_free_space(block_group, info); + goto out_lock; + } - ret = link_free_space(block_group, info); - spin_unlock(&block_group->tree_lock); - BUG_ON(ret); - } else if (info && info->offset < offset && - info->offset + info->bytes >= offset + bytes) { + if (!info->bitmap && info->offset <= offset && + info->offset + info->bytes >= offset + bytes) { u64 old_start = info->offset; /* * we're freeing space in the middle of the info, @@ -312,7 +706,9 @@ int btrfs_remove_free_space(struct btrfs_block_group_cache *block_group, info->offset = offset + bytes; info->bytes = old_end - info->offset; ret = link_free_space(block_group, info); - BUG_ON(ret); + WARN_ON(ret); + if (ret) + goto out_lock; } else { /* the hole we're creating ends at the end * of the info struct, just free the info @@ -320,32 +716,22 @@ int btrfs_remove_free_space(struct btrfs_block_group_cache *block_group, kfree(info); } spin_unlock(&block_group->tree_lock); - /* step two, insert a new info struct to cover anything - * before the hole + + /* step two, insert a new info struct to cover + * anything before the hole */ ret = btrfs_add_free_space(block_group, old_start, offset - old_start); - BUG_ON(ret); - } else { - spin_unlock(&block_group->tree_lock); - if (!info) { - printk(KERN_ERR "couldn't find space %llu to free\n", - (unsigned long long)offset); - printk(KERN_ERR "cached is %d, offset %llu bytes %llu\n", - block_group->cached, - (unsigned long long)block_group->key.objectid, - (unsigned long long)block_group->key.offset); - btrfs_dump_free_space(block_group, bytes); - } else if (info) { - printk(KERN_ERR "hmm, found offset=%llu bytes=%llu, " - "but wanted offset=%llu bytes=%llu\n", - (unsigned long long)info->offset, - (unsigned long long)info->bytes, - (unsigned long long)offset, - (unsigned long long)bytes); - } - WARN_ON(1); + WARN_ON(ret); + goto out; } + + ret = remove_from_bitmap(block_group, info, &offset, &bytes); + if (ret == -EAGAIN) + goto again; + BUG_ON(ret); +out_lock: + spin_unlock(&block_group->tree_lock); out: return ret; } @@ -361,10 +747,13 @@ void btrfs_dump_free_space(struct btrfs_block_group_cache *block_group, info = rb_entry(n, struct btrfs_free_space, offset_index); if (info->bytes >= bytes) count++; - printk(KERN_ERR "entry offset %llu, bytes %llu\n", + printk(KERN_CRIT "entry offset %llu, bytes %llu, bitmap %s\n", (unsigned long long)info->offset, - (unsigned long long)info->bytes); + (unsigned long long)info->bytes, + (info->bitmap) ? "yes" : "no"); } + printk(KERN_INFO "block group has cluster?: %s\n", + list_empty(&block_group->cluster_list) ? "no" : "yes"); printk(KERN_INFO "%d blocks of free space at or bigger than bytes is" "\n", count); } @@ -397,26 +786,35 @@ __btrfs_return_cluster_to_free_space( { struct btrfs_free_space *entry; struct rb_node *node; + bool bitmap; spin_lock(&cluster->lock); if (cluster->block_group != block_group) goto out; + bitmap = cluster->points_to_bitmap; + cluster->block_group = NULL; cluster->window_start = 0; + list_del_init(&cluster->block_group_list); + cluster->points_to_bitmap = false; + + if (bitmap) + goto out; + node = rb_first(&cluster->root); - while(node) { + while (node) { entry = rb_entry(node, struct btrfs_free_space, offset_index); node = rb_next(&entry->offset_index); rb_erase(&entry->offset_index, &cluster->root); - link_free_space(block_group, entry); + BUG_ON(entry->bitmap); + tree_insert_offset(&block_group->free_space_offset, + entry->offset, &entry->offset_index, 0); } - list_del_init(&cluster->block_group_list); - - btrfs_put_block_group(cluster->block_group); - cluster->block_group = NULL; cluster->root.rb_node = NULL; + out: spin_unlock(&cluster->lock); + btrfs_put_block_group(block_group); return 0; } @@ -425,20 +823,28 @@ void btrfs_remove_free_space_cache(struct btrfs_block_group_cache *block_group) struct btrfs_free_space *info; struct rb_node *node; struct btrfs_free_cluster *cluster; - struct btrfs_free_cluster *safe; + struct list_head *head; spin_lock(&block_group->tree_lock); - - list_for_each_entry_safe(cluster, safe, &block_group->cluster_list, - block_group_list) { + while ((head = block_group->cluster_list.next) != + &block_group->cluster_list) { + cluster = list_entry(head, struct btrfs_free_cluster, + block_group_list); WARN_ON(cluster->block_group != block_group); __btrfs_return_cluster_to_free_space(block_group, cluster); + if (need_resched()) { + spin_unlock(&block_group->tree_lock); + cond_resched(); + spin_lock(&block_group->tree_lock); + } } - while ((node = rb_last(&block_group->free_space_bytes)) != NULL) { - info = rb_entry(node, struct btrfs_free_space, bytes_index); + while ((node = rb_last(&block_group->free_space_offset)) != NULL) { + info = rb_entry(node, struct btrfs_free_space, offset_index); unlink_free_space(block_group, info); + if (info->bitmap) + kfree(info->bitmap); kfree(info); if (need_resched()) { spin_unlock(&block_group->tree_lock); @@ -446,6 +852,7 @@ void btrfs_remove_free_space_cache(struct btrfs_block_group_cache *block_group) spin_lock(&block_group->tree_lock); } } + spin_unlock(&block_group->tree_lock); } @@ -453,27 +860,37 @@ u64 btrfs_find_space_for_alloc(struct btrfs_block_group_cache *block_group, u64 offset, u64 bytes, u64 empty_size) { struct btrfs_free_space *entry = NULL; + u64 bytes_search = bytes + empty_size; u64 ret = 0; spin_lock(&block_group->tree_lock); - entry = tree_search_offset(&block_group->free_space_offset, offset, - bytes + empty_size, 1); + entry = find_free_space(block_group, &offset, &bytes_search, 0); if (!entry) - entry = tree_search_bytes(&block_group->free_space_bytes, - offset, bytes + empty_size); - if (entry) { + goto out; + + ret = offset; + if (entry->bitmap) { + bitmap_clear_bits(entry, offset, bytes, + block_group->sectorsize); + if (!entry->bytes) { + unlink_free_space(block_group, entry); + kfree(entry->bitmap); + kfree(entry); + block_group->total_bitmaps--; + recalculate_thresholds(block_group); + } + } else { unlink_free_space(block_group, entry); - ret = entry->offset; entry->offset += bytes; entry->bytes -= bytes; - if (!entry->bytes) kfree(entry); else link_free_space(block_group, entry); } - spin_unlock(&block_group->tree_lock); +out: + spin_unlock(&block_group->tree_lock); return ret; } @@ -517,6 +934,47 @@ int btrfs_return_cluster_to_free_space( return ret; } +static u64 btrfs_alloc_from_bitmap(struct btrfs_block_group_cache *block_group, + struct btrfs_free_cluster *cluster, + u64 bytes, u64 min_start) +{ + struct btrfs_free_space *entry; + int err; + u64 search_start = cluster->window_start; + u64 search_bytes = bytes; + u64 ret = 0; + + spin_lock(&block_group->tree_lock); + spin_lock(&cluster->lock); + + if (!cluster->points_to_bitmap) + goto out; + + if (cluster->block_group != block_group) + goto out; + + entry = tree_search_offset(block_group, search_start, 0, 0); + + if (!entry || !entry->bitmap) + goto out; + + search_start = min_start; + search_bytes = bytes; + + err = search_bitmap(block_group, entry, &search_start, + &search_bytes); + if (err) + goto out; + + ret = search_start; + bitmap_clear_bits(entry, ret, bytes, block_group->sectorsize); +out: + spin_unlock(&cluster->lock); + spin_unlock(&block_group->tree_lock); + + return ret; +} + /* * given a cluster, try to allocate 'bytes' from it, returns 0 * if it couldn't find anything suitably large, or a logical disk offset @@ -530,6 +988,10 @@ u64 btrfs_alloc_from_cluster(struct btrfs_block_group_cache *block_group, struct rb_node *node; u64 ret = 0; + if (cluster->points_to_bitmap) + return btrfs_alloc_from_bitmap(block_group, cluster, bytes, + min_start); + spin_lock(&cluster->lock); if (bytes > cluster->max_size) goto out; @@ -567,9 +1029,73 @@ u64 btrfs_alloc_from_cluster(struct btrfs_block_group_cache *block_group, } out: spin_unlock(&cluster->lock); + return ret; } +static int btrfs_bitmap_cluster(struct btrfs_block_group_cache *block_group, + struct btrfs_free_space *entry, + struct btrfs_free_cluster *cluster, + u64 offset, u64 bytes, u64 min_bytes) +{ + unsigned long next_zero; + unsigned long i; + unsigned long search_bits; + unsigned long total_bits; + unsigned long found_bits; + unsigned long start = 0; + unsigned long total_found = 0; + bool found = false; + + i = offset_to_bit(entry->offset, block_group->sectorsize, + max_t(u64, offset, entry->offset)); + search_bits = bytes_to_bits(min_bytes, block_group->sectorsize); + total_bits = bytes_to_bits(bytes, block_group->sectorsize); + +again: + found_bits = 0; + for (i = find_next_bit(entry->bitmap, BITS_PER_BITMAP, i); + i < BITS_PER_BITMAP; + i = find_next_bit(entry->bitmap, BITS_PER_BITMAP, i + 1)) { + next_zero = find_next_zero_bit(entry->bitmap, + BITS_PER_BITMAP, i); + if (next_zero - i >= search_bits) { + found_bits = next_zero - i; + break; + } + i = next_zero; + } + + if (!found_bits) + return -1; + + if (!found) { + start = i; + found = true; + } + + total_found += found_bits; + + if (cluster->max_size < found_bits * block_group->sectorsize) + cluster->max_size = found_bits * block_group->sectorsize; + + if (total_found < total_bits) { + i = find_next_bit(entry->bitmap, BITS_PER_BITMAP, next_zero); + if (i - start > total_bits * 2) { + total_found = 0; + cluster->max_size = 0; + found = false; + } + goto again; + } + + cluster->window_start = start * block_group->sectorsize + + entry->offset; + cluster->points_to_bitmap = true; + + return 0; +} + /* * here we try to find a cluster of blocks in a block group. The goal * is to find at least bytes free and up to empty_size + bytes free. @@ -587,12 +1113,12 @@ int btrfs_find_space_cluster(struct btrfs_trans_handle *trans, struct btrfs_free_space *entry = NULL; struct rb_node *node; struct btrfs_free_space *next; - struct btrfs_free_space *last; + struct btrfs_free_space *last = NULL; u64 min_bytes; u64 window_start; u64 window_free; u64 max_extent = 0; - int total_retries = 0; + bool found_bitmap = false; int ret; /* for metadata, allow allocates with more holes */ @@ -620,30 +1146,79 @@ int btrfs_find_space_cluster(struct btrfs_trans_handle *trans, goto out; } again: - min_bytes = min(min_bytes, bytes + empty_size); - entry = tree_search_bytes(&block_group->free_space_bytes, - offset, min_bytes); + entry = tree_search_offset(block_group, offset, found_bitmap, 1); if (!entry) { ret = -ENOSPC; goto out; } + + /* + * If found_bitmap is true, we exhausted our search for extent entries, + * and we just want to search all of the bitmaps that we can find, and + * ignore any extent entries we find. + */ + while (entry->bitmap || found_bitmap || + (!entry->bitmap && entry->bytes < min_bytes)) { + struct rb_node *node = rb_next(&entry->offset_index); + + if (entry->bitmap && entry->bytes > bytes + empty_size) { + ret = btrfs_bitmap_cluster(block_group, entry, cluster, + offset, bytes + empty_size, + min_bytes); + if (!ret) + goto got_it; + } + + if (!node) { + ret = -ENOSPC; + goto out; + } + entry = rb_entry(node, struct btrfs_free_space, offset_index); + } + + /* + * We already searched all the extent entries from the passed in offset + * to the end and didn't find enough space for the cluster, and we also + * didn't find any bitmaps that met our criteria, just go ahead and exit + */ + if (found_bitmap) { + ret = -ENOSPC; + goto out; + } + + cluster->points_to_bitmap = false; window_start = entry->offset; window_free = entry->bytes; last = entry; max_extent = entry->bytes; - while(1) { + while (1) { /* out window is just right, lets fill it */ if (window_free >= bytes + empty_size) break; node = rb_next(&last->offset_index); if (!node) { + if (found_bitmap) + goto again; ret = -ENOSPC; goto out; } next = rb_entry(node, struct btrfs_free_space, offset_index); + /* + * we found a bitmap, so if this search doesn't result in a + * cluster, we know to go and search again for the bitmaps and + * start looking for space there + */ + if (next->bitmap) { + if (!found_bitmap) + offset = next->offset; + found_bitmap = true; + last = next; + continue; + } + /* * we haven't filled the empty size and the window is * very large. reset and try again @@ -655,19 +1230,6 @@ again: window_free = entry->bytes; last = entry; max_extent = 0; - total_retries++; - if (total_retries % 64 == 0) { - if (min_bytes >= (bytes + empty_size)) { - ret = -ENOSPC; - goto out; - } - /* - * grow our allocation a bit, we're not having - * much luck - */ - min_bytes *= 2; - goto again; - } } else { last = next; window_free += next->bytes; @@ -685,11 +1247,19 @@ again: * The cluster includes an rbtree, but only uses the offset index * of each free space cache entry. */ - while(1) { + while (1) { node = rb_next(&entry->offset_index); - unlink_free_space(block_group, entry); + if (entry->bitmap && node) { + entry = rb_entry(node, struct btrfs_free_space, + offset_index); + continue; + } else if (entry->bitmap && !node) { + break; + } + + rb_erase(&entry->offset_index, &block_group->free_space_offset); ret = tree_insert_offset(&cluster->root, entry->offset, - &entry->offset_index); + &entry->offset_index, 0); BUG_ON(ret); if (!node || entry == last) @@ -697,8 +1267,10 @@ again: entry = rb_entry(node, struct btrfs_free_space, offset_index); } - ret = 0; + cluster->max_size = max_extent; +got_it: + ret = 0; atomic_inc(&block_group->count); list_add_tail(&cluster->block_group_list, &block_group->cluster_list); cluster->block_group = block_group; @@ -718,6 +1290,7 @@ void btrfs_init_free_cluster(struct btrfs_free_cluster *cluster) spin_lock_init(&cluster->refill_lock); cluster->root.rb_node = NULL; cluster->max_size = 0; + cluster->points_to_bitmap = false; INIT_LIST_HEAD(&cluster->block_group_list); cluster->block_group = NULL; } diff --git a/fs/btrfs/free-space-cache.h b/fs/btrfs/free-space-cache.h index 266fb8764054..890a8e79011b 100644 --- a/fs/btrfs/free-space-cache.h +++ b/fs/btrfs/free-space-cache.h @@ -19,6 +19,14 @@ #ifndef __BTRFS_FREE_SPACE_CACHE #define __BTRFS_FREE_SPACE_CACHE +struct btrfs_free_space { + struct rb_node offset_index; + u64 offset; + u64 bytes; + unsigned long *bitmap; + struct list_head list; +}; + int btrfs_add_free_space(struct btrfs_block_group_cache *block_group, u64 bytenr, u64 size); int btrfs_remove_free_space(struct btrfs_block_group_cache *block_group, -- cgit v1.2.3 From 817d52f8dba26d0295c26035531c30ce5f1e3c3e Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Mon, 13 Jul 2009 21:29:25 -0400 Subject: Btrfs: async block group caching This patch moves the caching of the block group off to a kthread in order to allow people to allocate sooner. Instead of blocking up behind the caching mutex, we instead kick of the caching kthread, and then attempt to make an allocation. If we cannot, we wait on the block groups caching waitqueue, which the caching kthread will wake the waiting threads up everytime it finds 2 meg worth of space, and then again when its finished caching. This is how I tested the speedup from this mkfs the disk mount the disk fill the disk up with fs_mark unmount the disk mount the disk time touch /mnt/foo Without my changes this took 11 seconds on my box, with these changes it now takes 1 second. Another change thats been put in place is we lock the super mirror's in the pinned extent map in order to keep us from adding that stuff as free space when caching the block group. This doesn't really change anything else as far as the pinned extent map is concerned, since for actual pinned extents we use EXTENT_DIRTY, but it does mean that when we unmount we have to go in and unlock those extents to keep from leaking memory. I've also added a check where when we are reading block groups from disk, if the amount of space used == the size of the block group, we go ahead and mark the block group as cached. This drastically reduces the amount of time it takes to cache the block groups. Using the same test as above, except doing a dd to a file and then unmounting, it used to take 33 seconds to umount, now it takes 3 seconds. This version uses the commit_root in the caching kthread, and then keeps track of how many async caching threads are running at any given time so if one of the async threads is still running as we cross transactions we can wait until its finished before handling the pinned extents. Thank you, Signed-off-by: Josef Bacik Signed-off-by: Chris Mason --- fs/btrfs/ctree.h | 22 ++- fs/btrfs/disk-io.c | 3 + fs/btrfs/extent-tree.c | 471 +++++++++++++++++++++++++++++++++++--------- fs/btrfs/free-space-cache.c | 42 ++-- fs/btrfs/transaction.c | 23 ++- fs/btrfs/tree-log.c | 2 +- 6 files changed, 439 insertions(+), 124 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 0cbf3491bb7c..42b03c4ee494 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -691,6 +691,7 @@ struct btrfs_space_info { struct list_head block_groups; spinlock_t lock; struct rw_semaphore groups_sem; + atomic_t caching_threads; }; /* @@ -721,11 +722,17 @@ struct btrfs_free_cluster { struct list_head block_group_list; }; +enum btrfs_caching_type { + BTRFS_CACHE_NO = 0, + BTRFS_CACHE_STARTED = 1, + BTRFS_CACHE_FINISHED = 2, +}; + struct btrfs_block_group_cache { struct btrfs_key key; struct btrfs_block_group_item item; + struct btrfs_fs_info *fs_info; spinlock_t lock; - struct mutex cache_mutex; u64 pinned; u64 reserved; u64 flags; @@ -733,15 +740,19 @@ struct btrfs_block_group_cache { int extents_thresh; int free_extents; int total_bitmaps; - int cached; int ro; int dirty; + /* cache tracking stuff */ + wait_queue_head_t caching_q; + int cached; + struct btrfs_space_info *space_info; /* free space cache stuff */ spinlock_t tree_lock; struct rb_root free_space_offset; + u64 free_space; /* block group cache stuff */ struct rb_node cache_node; @@ -834,6 +845,7 @@ struct btrfs_fs_info { atomic_t async_submit_draining; atomic_t nr_async_bios; atomic_t async_delalloc_pages; + atomic_t async_caching_threads; /* * this is used by the balancing code to wait for all the pending @@ -950,6 +962,9 @@ struct btrfs_root { /* the node lock is held while changing the node pointer */ spinlock_t node_lock; + /* taken when updating the commit root */ + struct rw_semaphore commit_root_sem; + struct extent_buffer *commit_root; struct btrfs_root *log_root; struct btrfs_root *reloc_root; @@ -1911,7 +1926,7 @@ int btrfs_run_delayed_refs(struct btrfs_trans_handle *trans, struct btrfs_root *root, unsigned long count); int btrfs_lookup_extent(struct btrfs_root *root, u64 start, u64 len); int btrfs_update_pinned_extents(struct btrfs_root *root, - u64 bytenr, u64 num, int pin); + u64 bytenr, u64 num, int pin, int mark_free); int btrfs_drop_leaf_ref(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct extent_buffer *leaf); int btrfs_cross_ref_exist(struct btrfs_trans_handle *trans, @@ -1996,6 +2011,7 @@ void btrfs_delalloc_reserve_space(struct btrfs_root *root, struct inode *inode, u64 bytes); void btrfs_delalloc_free_space(struct btrfs_root *root, struct inode *inode, u64 bytes); +void btrfs_free_super_mirror_extents(struct btrfs_fs_info *info); /* ctree.c */ int btrfs_bin_search(struct extent_buffer *eb, struct btrfs_key *key, int level, int *slot); diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 55d9d188e693..ec2c915f7f4a 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -907,6 +907,7 @@ static int __setup_root(u32 nodesize, u32 leafsize, u32 sectorsize, spin_lock_init(&root->inode_lock); mutex_init(&root->objectid_mutex); mutex_init(&root->log_mutex); + init_rwsem(&root->commit_root_sem); init_waitqueue_head(&root->log_writer_wait); init_waitqueue_head(&root->log_commit_wait[0]); init_waitqueue_head(&root->log_commit_wait[1]); @@ -1566,6 +1567,7 @@ struct btrfs_root *open_ctree(struct super_block *sb, atomic_set(&fs_info->async_delalloc_pages, 0); atomic_set(&fs_info->async_submit_draining, 0); atomic_set(&fs_info->nr_async_bios, 0); + atomic_set(&fs_info->async_caching_threads, 0); fs_info->sb = sb; fs_info->max_extent = (u64)-1; fs_info->max_inline = 8192 * 1024; @@ -2337,6 +2339,7 @@ int close_ctree(struct btrfs_root *root) free_extent_buffer(root->fs_info->csum_root->commit_root); btrfs_free_block_groups(root->fs_info); + btrfs_free_super_mirror_extents(root->fs_info); del_fs_roots(fs_info); diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 98697be6bdde..9a489cc89fd3 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -21,6 +21,7 @@ #include #include #include +#include #include "compat.h" #include "hash.h" #include "ctree.h" @@ -61,6 +62,13 @@ static int do_chunk_alloc(struct btrfs_trans_handle *trans, struct btrfs_root *extent_root, u64 alloc_bytes, u64 flags, int force); +static noinline int +block_group_cache_done(struct btrfs_block_group_cache *cache) +{ + smp_mb(); + return cache->cached == BTRFS_CACHE_FINISHED; +} + static int block_group_bits(struct btrfs_block_group_cache *cache, u64 bits) { return (cache->flags & bits) == bits; @@ -145,21 +153,64 @@ block_group_cache_tree_search(struct btrfs_fs_info *info, u64 bytenr, return ret; } +void btrfs_free_super_mirror_extents(struct btrfs_fs_info *info) +{ + u64 start, end, last = 0; + int ret; + + while (1) { + ret = find_first_extent_bit(&info->pinned_extents, last, + &start, &end, EXTENT_LOCKED); + if (ret) + break; + + unlock_extent(&info->pinned_extents, start, end, GFP_NOFS); + last = end+1; + } +} + +static int remove_sb_from_cache(struct btrfs_root *root, + struct btrfs_block_group_cache *cache) +{ + struct btrfs_fs_info *fs_info = root->fs_info; + u64 bytenr; + u64 *logical; + int stripe_len; + int i, nr, ret; + + for (i = 0; i < BTRFS_SUPER_MIRROR_MAX; i++) { + bytenr = btrfs_sb_offset(i); + ret = btrfs_rmap_block(&root->fs_info->mapping_tree, + cache->key.objectid, bytenr, + 0, &logical, &nr, &stripe_len); + BUG_ON(ret); + while (nr--) { + try_lock_extent(&fs_info->pinned_extents, + logical[nr], + logical[nr] + stripe_len - 1, GFP_NOFS); + } + kfree(logical); + } + + return 0; +} + /* * this is only called by cache_block_group, since we could have freed extents * we need to check the pinned_extents for any extents that can't be used yet * since their free space will be released as soon as the transaction commits. */ -static int add_new_free_space(struct btrfs_block_group_cache *block_group, +static u64 add_new_free_space(struct btrfs_block_group_cache *block_group, struct btrfs_fs_info *info, u64 start, u64 end) { - u64 extent_start, extent_end, size; + u64 extent_start, extent_end, size, total_added = 0; int ret; while (start < end) { ret = find_first_extent_bit(&info->pinned_extents, start, &extent_start, &extent_end, - EXTENT_DIRTY); + EXTENT_DIRTY|EXTENT_LOCKED| + EXTENT_DELALLOC); if (ret) break; @@ -167,6 +218,7 @@ static int add_new_free_space(struct btrfs_block_group_cache *block_group, start = extent_end + 1; } else if (extent_start > start && extent_start < end) { size = extent_start - start; + total_added += size; ret = btrfs_add_free_space(block_group, start, size); BUG_ON(ret); @@ -178,84 +230,139 @@ static int add_new_free_space(struct btrfs_block_group_cache *block_group, if (start < end) { size = end - start; + total_added += size; ret = btrfs_add_free_space(block_group, start, size); BUG_ON(ret); } - return 0; + return total_added; } -static int remove_sb_from_cache(struct btrfs_root *root, - struct btrfs_block_group_cache *cache) +DEFINE_MUTEX(discard_mutex); + +/* + * if async kthreads are running when we cross transactions, we mark any pinned + * extents with EXTENT_DELALLOC and then let the caching kthreads clean up those + * extents when they are done. Also we run this from btrfs_finish_extent_commit + * in case there were some pinned extents that were missed because we had + * already cached that block group. + */ +static void btrfs_discard_pinned_extents(struct btrfs_fs_info *fs_info, + struct btrfs_block_group_cache *cache) { - u64 bytenr; - u64 *logical; - int stripe_len; - int i, nr, ret; + u64 start, end, last; + int ret; - for (i = 0; i < BTRFS_SUPER_MIRROR_MAX; i++) { - bytenr = btrfs_sb_offset(i); - ret = btrfs_rmap_block(&root->fs_info->mapping_tree, - cache->key.objectid, bytenr, 0, - &logical, &nr, &stripe_len); - BUG_ON(ret); - while (nr--) { - btrfs_remove_free_space(cache, logical[nr], - stripe_len); + if (!cache) + last = 0; + else + last = cache->key.objectid; + + mutex_lock(&discard_mutex); + while (1) { + ret = find_first_extent_bit(&fs_info->pinned_extents, last, + &start, &end, EXTENT_DELALLOC); + if (ret) + break; + + if (cache && start >= cache->key.objectid + cache->key.offset) + break; + + + if (!cache) { + cache = btrfs_lookup_block_group(fs_info, start); + BUG_ON(!cache); + + start = max(start, cache->key.objectid); + end = min(end, cache->key.objectid + cache->key.offset - 1); + + if (block_group_cache_done(cache)) + btrfs_add_free_space(cache, start, + end - start + 1); + cache = NULL; + } else { + start = max(start, cache->key.objectid); + end = min(end, cache->key.objectid + cache->key.offset - 1); + btrfs_add_free_space(cache, start, end - start + 1); + } + + clear_extent_bits(&fs_info->pinned_extents, start, end, + EXTENT_DELALLOC, GFP_NOFS); + last = end + 1; + + if (need_resched()) { + mutex_unlock(&discard_mutex); + cond_resched(); + mutex_lock(&discard_mutex); } - kfree(logical); } - return 0; + mutex_unlock(&discard_mutex); } -static int cache_block_group(struct btrfs_root *root, - struct btrfs_block_group_cache *block_group) +static int caching_kthread(void *data) { + struct btrfs_block_group_cache *block_group = data; + struct btrfs_fs_info *fs_info = block_group->fs_info; + u64 last = 0; struct btrfs_path *path; int ret = 0; struct btrfs_key key; struct extent_buffer *leaf; int slot; - u64 last; - - if (!block_group) - return 0; + u64 total_found = 0; - root = root->fs_info->extent_root; - - if (block_group->cached) - return 0; + BUG_ON(!fs_info); path = btrfs_alloc_path(); if (!path) return -ENOMEM; - path->reada = 2; + atomic_inc(&fs_info->async_caching_threads); + atomic_inc(&block_group->space_info->caching_threads); + last = max_t(u64, block_group->key.objectid, BTRFS_SUPER_INFO_OFFSET); +again: + /* need to make sure the commit_root doesn't disappear */ + down_read(&fs_info->extent_root->commit_root_sem); + /* - * we get into deadlocks with paths held by callers of this function. - * since the alloc_mutex is protecting things right now, just - * skip the locking here + * We don't want to deadlock with somebody trying to allocate a new + * extent for the extent root while also trying to search the extent + * root to add free space. So we skip locking and search the commit + * root, since its read-only */ path->skip_locking = 1; - last = max_t(u64, block_group->key.objectid, BTRFS_SUPER_INFO_OFFSET); + path->search_commit_root = 1; + path->reada = 2; + key.objectid = last; key.offset = 0; btrfs_set_key_type(&key, BTRFS_EXTENT_ITEM_KEY); - ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); + ret = btrfs_search_slot(NULL, fs_info->extent_root, &key, path, 0, 0); if (ret < 0) goto err; while (1) { + smp_mb(); + if (block_group->fs_info->closing) + break; + leaf = path->nodes[0]; slot = path->slots[0]; if (slot >= btrfs_header_nritems(leaf)) { - ret = btrfs_next_leaf(root, path); + ret = btrfs_next_leaf(fs_info->extent_root, path); if (ret < 0) goto err; - if (ret == 0) - continue; - else + else if (ret) break; + + if (need_resched()) { + btrfs_release_path(fs_info->extent_root, path); + up_read(&fs_info->extent_root->commit_root_sem); + cond_resched(); + goto again; + } + + continue; } btrfs_item_key_to_cpu(leaf, &key, slot); if (key.objectid < block_group->key.objectid) @@ -266,24 +373,63 @@ static int cache_block_group(struct btrfs_root *root, break; if (btrfs_key_type(&key) == BTRFS_EXTENT_ITEM_KEY) { - add_new_free_space(block_group, root->fs_info, last, - key.objectid); - + total_found += add_new_free_space(block_group, + fs_info, last, + key.objectid); last = key.objectid + key.offset; } + + if (total_found > (1024 * 1024 * 2)) { + total_found = 0; + wake_up(&block_group->caching_q); + } next: path->slots[0]++; } + ret = 0; - add_new_free_space(block_group, root->fs_info, last, - block_group->key.objectid + - block_group->key.offset); + total_found += add_new_free_space(block_group, fs_info, last, + block_group->key.objectid + + block_group->key.offset); + + spin_lock(&block_group->lock); + block_group->cached = BTRFS_CACHE_FINISHED; + spin_unlock(&block_group->lock); - block_group->cached = 1; - remove_sb_from_cache(root, block_group); - ret = 0; err: btrfs_free_path(path); + up_read(&fs_info->extent_root->commit_root_sem); + atomic_dec(&fs_info->async_caching_threads); + atomic_dec(&block_group->space_info->caching_threads); + wake_up(&block_group->caching_q); + + if (!ret) + btrfs_discard_pinned_extents(fs_info, block_group); + + return 0; +} + +static int cache_block_group(struct btrfs_block_group_cache *cache) +{ + struct task_struct *tsk; + int ret = 0; + + spin_lock(&cache->lock); + if (cache->cached != BTRFS_CACHE_NO) { + spin_unlock(&cache->lock); + return ret; + } + cache->cached = BTRFS_CACHE_STARTED; + spin_unlock(&cache->lock); + + tsk = kthread_run(caching_kthread, cache, "btrfs-cache-%llu\n", + cache->key.objectid); + if (IS_ERR(tsk)) { + ret = PTR_ERR(tsk); + printk(KERN_ERR "error running thread %d\n", ret); + BUG(); + } + return ret; } @@ -1721,7 +1867,7 @@ static int run_one_delayed_ref(struct btrfs_trans_handle *trans, BUG_ON(ret); } btrfs_update_pinned_extents(root, node->bytenr, - node->num_bytes, 1); + node->num_bytes, 1, 0); update_reserved_extents(root, node->bytenr, node->num_bytes, 0); } @@ -2496,6 +2642,7 @@ static int update_space_info(struct btrfs_fs_info *info, u64 flags, found->force_alloc = 0; *space_info = found; list_add_rcu(&found->list, &info->space_info); + atomic_set(&found->caching_threads, 0); return 0; } @@ -2953,7 +3100,7 @@ static u64 first_logical_byte(struct btrfs_root *root, u64 search_start) } int btrfs_update_pinned_extents(struct btrfs_root *root, - u64 bytenr, u64 num, int pin) + u64 bytenr, u64 num, int pin, int mark_free) { u64 len; struct btrfs_block_group_cache *cache; @@ -2988,7 +3135,7 @@ int btrfs_update_pinned_extents(struct btrfs_root *root, spin_unlock(&cache->lock); spin_unlock(&cache->space_info->lock); fs_info->total_pinned -= len; - if (cache->cached) + if (block_group_cache_done(cache) && mark_free) btrfs_add_free_space(cache, bytenr, len); } btrfs_put_block_group(cache); @@ -3034,14 +3181,27 @@ int btrfs_copy_pinned(struct btrfs_root *root, struct extent_io_tree *copy) u64 last = 0; u64 start; u64 end; + bool caching_kthreads = false; struct extent_io_tree *pinned_extents = &root->fs_info->pinned_extents; int ret; + if (atomic_read(&root->fs_info->async_caching_threads)) + caching_kthreads = true; + while (1) { ret = find_first_extent_bit(pinned_extents, last, &start, &end, EXTENT_DIRTY); if (ret) break; + + /* + * we need to make sure that the pinned extents don't go away + * while we are caching block groups + */ + if (unlikely(caching_kthreads)) + set_extent_delalloc(pinned_extents, start, end, + GFP_NOFS); + set_extent_dirty(copy, start, end, GFP_NOFS); last = end + 1; } @@ -3055,6 +3215,12 @@ int btrfs_finish_extent_commit(struct btrfs_trans_handle *trans, u64 start; u64 end; int ret; + int mark_free = 1; + + ret = find_first_extent_bit(&root->fs_info->pinned_extents, 0, + &start, &end, EXTENT_DELALLOC); + if (!ret) + mark_free = 0; while (1) { ret = find_first_extent_bit(unpin, 0, &start, &end, @@ -3065,11 +3231,16 @@ int btrfs_finish_extent_commit(struct btrfs_trans_handle *trans, ret = btrfs_discard_extent(root, start, end + 1 - start); /* unlocks the pinned mutex */ - btrfs_update_pinned_extents(root, start, end + 1 - start, 0); + btrfs_update_pinned_extents(root, start, end + 1 - start, 0, + mark_free); clear_extent_dirty(unpin, start, end, GFP_NOFS); cond_resched(); } + + if (unlikely(!mark_free)) + btrfs_discard_pinned_extents(root->fs_info, NULL); + return ret; } @@ -3110,7 +3281,7 @@ static int pin_down_bytes(struct btrfs_trans_handle *trans, pinit: btrfs_set_path_blocking(path); /* unlocks the pinned mutex */ - btrfs_update_pinned_extents(root, bytenr, num_bytes, 1); + btrfs_update_pinned_extents(root, bytenr, num_bytes, 1, 0); BUG_ON(err < 0); return 0; @@ -3421,7 +3592,7 @@ int btrfs_free_extent(struct btrfs_trans_handle *trans, if (root_objectid == BTRFS_TREE_LOG_OBJECTID) { WARN_ON(owner >= BTRFS_FIRST_FREE_OBJECTID); /* unlocks the pinned mutex */ - btrfs_update_pinned_extents(root, bytenr, num_bytes, 1); + btrfs_update_pinned_extents(root, bytenr, num_bytes, 1, 0); update_reserved_extents(root, bytenr, num_bytes, 0); ret = 0; } else if (owner < BTRFS_FIRST_FREE_OBJECTID) { @@ -3447,6 +3618,45 @@ static u64 stripe_align(struct btrfs_root *root, u64 val) return ret; } +/* + * when we wait for progress in the block group caching, its because + * our allocation attempt failed at least once. So, we must sleep + * and let some progress happen before we try again. + * + * This function will sleep at least once waiting for new free space to + * show up, and then it will check the block group free space numbers + * for our min num_bytes. Another option is to have it go ahead + * and look in the rbtree for a free extent of a given size, but this + * is a good start. + */ +static noinline int +wait_block_group_cache_progress(struct btrfs_block_group_cache *cache, + u64 num_bytes) +{ + DEFINE_WAIT(wait); + + prepare_to_wait(&cache->caching_q, &wait, TASK_UNINTERRUPTIBLE); + + if (block_group_cache_done(cache)) { + finish_wait(&cache->caching_q, &wait); + return 0; + } + schedule(); + finish_wait(&cache->caching_q, &wait); + + wait_event(cache->caching_q, block_group_cache_done(cache) || + (cache->free_space >= num_bytes)); + return 0; +} + +enum btrfs_loop_type { + LOOP_CACHED_ONLY = 0, + LOOP_CACHING_NOWAIT = 1, + LOOP_CACHING_WAIT = 2, + LOOP_ALLOC_CHUNK = 3, + LOOP_NO_EMPTY_SIZE = 4, +}; + /* * walks the btree of allocated extents and find a hole of a given size. * The key ins is changed to record the hole: @@ -3472,6 +3682,7 @@ static noinline int find_free_extent(struct btrfs_trans_handle *trans, struct btrfs_space_info *space_info; int last_ptr_loop = 0; int loop = 0; + bool found_uncached_bg = false; WARN_ON(num_bytes < root->sectorsize); btrfs_set_key_type(ins, BTRFS_EXTENT_ITEM_KEY); @@ -3503,15 +3714,18 @@ static noinline int find_free_extent(struct btrfs_trans_handle *trans, search_start = max(search_start, first_logical_byte(root, 0)); search_start = max(search_start, hint_byte); - if (!last_ptr) { + if (!last_ptr) empty_cluster = 0; - loop = 1; - } if (search_start == hint_byte) { block_group = btrfs_lookup_block_group(root->fs_info, search_start); - if (block_group && block_group_bits(block_group, data)) { + /* + * we don't want to use the block group if it doesn't match our + * allocation bits, or if its not cached. + */ + if (block_group && block_group_bits(block_group, data) && + block_group_cache_done(block_group)) { down_read(&space_info->groups_sem); if (list_empty(&block_group->list) || block_group->ro) { @@ -3534,21 +3748,35 @@ search: down_read(&space_info->groups_sem); list_for_each_entry(block_group, &space_info->block_groups, list) { u64 offset; + int cached; atomic_inc(&block_group->count); search_start = block_group->key.objectid; have_block_group: - if (unlikely(!block_group->cached)) { - mutex_lock(&block_group->cache_mutex); - ret = cache_block_group(root, block_group); - mutex_unlock(&block_group->cache_mutex); - if (ret) { - btrfs_put_block_group(block_group); - break; + if (unlikely(block_group->cached == BTRFS_CACHE_NO)) { + /* + * we want to start caching kthreads, but not too many + * right off the bat so we don't overwhelm the system, + * so only start them if there are less than 2 and we're + * in the initial allocation phase. + */ + if (loop > LOOP_CACHING_NOWAIT || + atomic_read(&space_info->caching_threads) < 2) { + ret = cache_block_group(block_group); + BUG_ON(ret); } } + cached = block_group_cache_done(block_group); + if (unlikely(!cached)) { + found_uncached_bg = true; + + /* if we only want cached bgs, loop */ + if (loop == LOOP_CACHED_ONLY) + goto loop; + } + if (unlikely(block_group->ro)) goto loop; @@ -3627,14 +3855,21 @@ refill_cluster: spin_unlock(&last_ptr->refill_lock); goto checks; } + } else if (!cached && loop > LOOP_CACHING_NOWAIT) { + spin_unlock(&last_ptr->refill_lock); + + wait_block_group_cache_progress(block_group, + num_bytes + empty_cluster + empty_size); + goto have_block_group; } + /* * at this point we either didn't find a cluster * or we weren't able to allocate a block from our * cluster. Free the cluster we've been trying * to use, and go to the next block group */ - if (loop < 2) { + if (loop < LOOP_NO_EMPTY_SIZE) { btrfs_return_cluster_to_free_space(NULL, last_ptr); spin_unlock(&last_ptr->refill_lock); @@ -3645,8 +3880,15 @@ refill_cluster: offset = btrfs_find_space_for_alloc(block_group, search_start, num_bytes, empty_size); - if (!offset) + if (!offset && (cached || (!cached && + loop == LOOP_CACHING_NOWAIT))) { goto loop; + } else if (!offset && (!cached && + loop > LOOP_CACHING_NOWAIT)) { + wait_block_group_cache_progress(block_group, + num_bytes + empty_size); + goto have_block_group; + } checks: search_start = stripe_align(root, offset); /* move on to the next group */ @@ -3694,13 +3936,26 @@ loop: } up_read(&space_info->groups_sem); - /* loop == 0, try to find a clustered alloc in every block group - * loop == 1, try again after forcing a chunk allocation - * loop == 2, set empty_size and empty_cluster to 0 and try again + /* LOOP_CACHED_ONLY, only search fully cached block groups + * LOOP_CACHING_NOWAIT, search partially cached block groups, but + * dont wait foR them to finish caching + * LOOP_CACHING_WAIT, search everything, and wait if our bg is caching + * LOOP_ALLOC_CHUNK, force a chunk allocation and try again + * LOOP_NO_EMPTY_SIZE, set empty_size and empty_cluster to 0 and try + * again */ - if (!ins->objectid && loop < 3 && - (empty_size || empty_cluster || allowed_chunk_alloc)) { - if (loop >= 2) { + if (!ins->objectid && loop < LOOP_NO_EMPTY_SIZE && + (found_uncached_bg || empty_size || empty_cluster || + allowed_chunk_alloc)) { + if (found_uncached_bg) { + found_uncached_bg = false; + if (loop < LOOP_CACHING_WAIT) { + loop++; + goto search; + } + } + + if (loop == LOOP_ALLOC_CHUNK) { empty_size = 0; empty_cluster = 0; } @@ -3713,7 +3968,7 @@ loop: space_info->force_alloc = 1; } - if (loop < 3) { + if (loop < LOOP_NO_EMPTY_SIZE) { loop++; goto search; } @@ -3809,7 +4064,7 @@ again: num_bytes, data, 1); goto again; } - if (ret) { + if (ret == -ENOSPC) { struct btrfs_space_info *sinfo; sinfo = __find_space_info(root->fs_info, data); @@ -3817,7 +4072,6 @@ again: "wanted %llu\n", (unsigned long long)data, (unsigned long long)num_bytes); dump_space_info(sinfo, num_bytes); - BUG(); } return ret; @@ -3855,7 +4109,9 @@ int btrfs_reserve_extent(struct btrfs_trans_handle *trans, ret = __btrfs_reserve_extent(trans, root, num_bytes, min_alloc_size, empty_size, hint_byte, search_end, ins, data); - update_reserved_extents(root, ins->objectid, ins->offset, 1); + if (!ret) + update_reserved_extents(root, ins->objectid, ins->offset, 1); + return ret; } @@ -4017,9 +4273,9 @@ int btrfs_alloc_logged_file_extent(struct btrfs_trans_handle *trans, struct btrfs_block_group_cache *block_group; block_group = btrfs_lookup_block_group(root->fs_info, ins->objectid); - mutex_lock(&block_group->cache_mutex); - cache_block_group(root, block_group); - mutex_unlock(&block_group->cache_mutex); + cache_block_group(block_group); + wait_event(block_group->caching_q, + block_group_cache_done(block_group)); ret = btrfs_remove_free_space(block_group, ins->objectid, ins->offset); @@ -4050,7 +4306,8 @@ static int alloc_tree_block(struct btrfs_trans_handle *trans, ret = __btrfs_reserve_extent(trans, root, num_bytes, num_bytes, empty_size, hint_byte, search_end, ins, 0); - BUG_ON(ret); + if (ret) + return ret; if (root_objectid == BTRFS_TREE_RELOC_OBJECTID) { if (parent == 0) @@ -6966,11 +7223,16 @@ int btrfs_free_block_groups(struct btrfs_fs_info *info) &info->block_group_cache_tree); spin_unlock(&info->block_group_cache_lock); - btrfs_remove_free_space_cache(block_group); down_write(&block_group->space_info->groups_sem); list_del(&block_group->list); up_write(&block_group->space_info->groups_sem); + if (block_group->cached == BTRFS_CACHE_STARTED) + wait_event(block_group->caching_q, + block_group_cache_done(block_group)); + + btrfs_remove_free_space_cache(block_group); + WARN_ON(atomic_read(&block_group->count) != 1); kfree(block_group); @@ -7036,10 +7298,10 @@ int btrfs_read_block_groups(struct btrfs_root *root) atomic_set(&cache->count, 1); spin_lock_init(&cache->lock); spin_lock_init(&cache->tree_lock); - mutex_init(&cache->cache_mutex); + cache->fs_info = info; + init_waitqueue_head(&cache->caching_q); INIT_LIST_HEAD(&cache->list); INIT_LIST_HEAD(&cache->cluster_list); - cache->sectorsize = root->sectorsize; /* * we only want to have 32k of ram per block group for keeping @@ -7057,6 +7319,26 @@ int btrfs_read_block_groups(struct btrfs_root *root) key.objectid = found_key.objectid + found_key.offset; btrfs_release_path(root, path); cache->flags = btrfs_block_group_flags(&cache->item); + cache->sectorsize = root->sectorsize; + + remove_sb_from_cache(root, cache); + + /* + * check for two cases, either we are full, and therefore + * don't need to bother with the caching work since we won't + * find any space, or we are empty, and we can just add all + * the space in and be done with it. This saves us _alot_ of + * time, particularly in the full case. + */ + if (found_key.offset == btrfs_block_group_used(&cache->item)) { + cache->cached = BTRFS_CACHE_FINISHED; + } else if (btrfs_block_group_used(&cache->item) == 0) { + cache->cached = BTRFS_CACHE_FINISHED; + add_new_free_space(cache, root->fs_info, + found_key.objectid, + found_key.objectid + + found_key.offset); + } ret = update_space_info(info, cache->flags, found_key.offset, btrfs_block_group_used(&cache->item), @@ -7112,7 +7394,7 @@ int btrfs_make_block_group(struct btrfs_trans_handle *trans, atomic_set(&cache->count, 1); spin_lock_init(&cache->lock); spin_lock_init(&cache->tree_lock); - mutex_init(&cache->cache_mutex); + init_waitqueue_head(&cache->caching_q); INIT_LIST_HEAD(&cache->list); INIT_LIST_HEAD(&cache->cluster_list); @@ -7121,11 +7403,12 @@ int btrfs_make_block_group(struct btrfs_trans_handle *trans, cache->flags = type; btrfs_set_block_group_flags(&cache->item, type); - cache->cached = 1; - ret = btrfs_add_free_space(cache, chunk_offset, size); - BUG_ON(ret); + cache->cached = BTRFS_CACHE_FINISHED; remove_sb_from_cache(root, cache); + add_new_free_space(cache, root->fs_info, chunk_offset, + chunk_offset + size); + ret = update_space_info(root->fs_info, cache->flags, size, bytes_used, &cache->space_info); BUG_ON(ret); @@ -7184,7 +7467,7 @@ int btrfs_remove_block_group(struct btrfs_trans_handle *trans, rb_erase(&block_group->cache_node, &root->fs_info->block_group_cache_tree); spin_unlock(&root->fs_info->block_group_cache_lock); - btrfs_remove_free_space_cache(block_group); + down_write(&block_group->space_info->groups_sem); /* * we must use list_del_init so people can check to see if they @@ -7193,6 +7476,12 @@ int btrfs_remove_block_group(struct btrfs_trans_handle *trans, list_del_init(&block_group->list); up_write(&block_group->space_info->groups_sem); + if (block_group->cached == BTRFS_CACHE_STARTED) + wait_event(block_group->caching_q, + block_group_cache_done(block_group)); + + btrfs_remove_free_space_cache(block_group); + spin_lock(&block_group->space_info->lock); block_group->space_info->total_bytes -= block_group->key.offset; block_group->space_info->bytes_readonly -= block_group->key.offset; diff --git a/fs/btrfs/free-space-cache.c b/fs/btrfs/free-space-cache.c index ab8cad8b46c9..af99b78b288e 100644 --- a/fs/btrfs/free-space-cache.c +++ b/fs/btrfs/free-space-cache.c @@ -238,6 +238,7 @@ static void unlink_free_space(struct btrfs_block_group_cache *block_group, { rb_erase(&info->offset_index, &block_group->free_space_offset); block_group->free_extents--; + block_group->free_space -= info->bytes; } static int link_free_space(struct btrfs_block_group_cache *block_group, @@ -251,6 +252,7 @@ static int link_free_space(struct btrfs_block_group_cache *block_group, if (ret) return ret; + block_group->free_space += info->bytes; block_group->free_extents++; return ret; } @@ -285,36 +287,40 @@ static void recalculate_thresholds(struct btrfs_block_group_cache *block_group) } } -static void bitmap_clear_bits(struct btrfs_free_space *info, u64 offset, u64 bytes, - u64 sectorsize) +static void bitmap_clear_bits(struct btrfs_block_group_cache *block_group, + struct btrfs_free_space *info, u64 offset, + u64 bytes) { unsigned long start, end; unsigned long i; - start = offset_to_bit(info->offset, sectorsize, offset); - end = start + bytes_to_bits(bytes, sectorsize); + start = offset_to_bit(info->offset, block_group->sectorsize, offset); + end = start + bytes_to_bits(bytes, block_group->sectorsize); BUG_ON(end > BITS_PER_BITMAP); for (i = start; i < end; i++) clear_bit(i, info->bitmap); info->bytes -= bytes; + block_group->free_space -= bytes; } -static void bitmap_set_bits(struct btrfs_free_space *info, u64 offset, u64 bytes, - u64 sectorsize) +static void bitmap_set_bits(struct btrfs_block_group_cache *block_group, + struct btrfs_free_space *info, u64 offset, + u64 bytes) { unsigned long start, end; unsigned long i; - start = offset_to_bit(info->offset, sectorsize, offset); - end = start + bytes_to_bits(bytes, sectorsize); + start = offset_to_bit(info->offset, block_group->sectorsize, offset); + end = start + bytes_to_bits(bytes, block_group->sectorsize); BUG_ON(end > BITS_PER_BITMAP); for (i = start; i < end; i++) set_bit(i, info->bitmap); info->bytes += bytes; + block_group->free_space += bytes; } static int search_bitmap(struct btrfs_block_group_cache *block_group, @@ -414,13 +420,12 @@ again: (u64)(BITS_PER_BITMAP * block_group->sectorsize) - 1; if (*offset > bitmap_info->offset && *offset + *bytes > end) { - bitmap_clear_bits(bitmap_info, *offset, - end - *offset + 1, block_group->sectorsize); + bitmap_clear_bits(block_group, bitmap_info, *offset, + end - *offset + 1); *bytes -= end - *offset + 1; *offset = end + 1; } else if (*offset >= bitmap_info->offset && *offset + *bytes <= end) { - bitmap_clear_bits(bitmap_info, *offset, - *bytes, block_group->sectorsize); + bitmap_clear_bits(block_group, bitmap_info, *offset, *bytes); *bytes = 0; } @@ -495,14 +500,13 @@ again: (u64)(BITS_PER_BITMAP * block_group->sectorsize); if (offset >= bitmap_info->offset && offset + bytes > end) { - bitmap_set_bits(bitmap_info, offset, end - offset, - block_group->sectorsize); + bitmap_set_bits(block_group, bitmap_info, offset, + end - offset); bytes -= end - offset; offset = end; added = 0; } else if (offset >= bitmap_info->offset && offset + bytes <= end) { - bitmap_set_bits(bitmap_info, offset, bytes, - block_group->sectorsize); + bitmap_set_bits(block_group, bitmap_info, offset, bytes); bytes = 0; } else { BUG(); @@ -870,8 +874,7 @@ u64 btrfs_find_space_for_alloc(struct btrfs_block_group_cache *block_group, ret = offset; if (entry->bitmap) { - bitmap_clear_bits(entry, offset, bytes, - block_group->sectorsize); + bitmap_clear_bits(block_group, entry, offset, bytes); if (!entry->bytes) { unlink_free_space(block_group, entry); kfree(entry->bitmap); @@ -891,6 +894,7 @@ u64 btrfs_find_space_for_alloc(struct btrfs_block_group_cache *block_group, out: spin_unlock(&block_group->tree_lock); + return ret; } @@ -967,7 +971,7 @@ static u64 btrfs_alloc_from_bitmap(struct btrfs_block_group_cache *block_group, goto out; ret = search_start; - bitmap_clear_bits(entry, ret, bytes, block_group->sectorsize); + bitmap_clear_bits(block_group, entry, ret, bytes); out: spin_unlock(&cluster->lock); spin_unlock(&block_group->tree_lock); diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index 81f7124c3051..32454d1c566f 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -40,6 +40,14 @@ static noinline void put_transaction(struct btrfs_transaction *transaction) } } +static noinline void switch_commit_root(struct btrfs_root *root) +{ + down_write(&root->commit_root_sem); + free_extent_buffer(root->commit_root); + root->commit_root = btrfs_root_node(root); + up_write(&root->commit_root_sem); +} + /* * either allocate a new transaction or hop into the existing one */ @@ -458,8 +466,7 @@ static int update_cowonly_root(struct btrfs_trans_handle *trans, ret = btrfs_write_dirty_block_groups(trans, root); BUG_ON(ret); } - free_extent_buffer(root->commit_root); - root->commit_root = btrfs_root_node(root); + switch_commit_root(root); return 0; } @@ -537,8 +544,7 @@ static noinline int commit_fs_roots(struct btrfs_trans_handle *trans, btrfs_update_reloc_root(trans, root); if (root->commit_root != root->node) { - free_extent_buffer(root->commit_root); - root->commit_root = btrfs_root_node(root); + switch_commit_root(root); btrfs_set_root_node(&root->root_item, root->node); } @@ -1002,15 +1008,11 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans, btrfs_set_root_node(&root->fs_info->tree_root->root_item, root->fs_info->tree_root->node); - free_extent_buffer(root->fs_info->tree_root->commit_root); - root->fs_info->tree_root->commit_root = - btrfs_root_node(root->fs_info->tree_root); + switch_commit_root(root->fs_info->tree_root); btrfs_set_root_node(&root->fs_info->chunk_root->root_item, root->fs_info->chunk_root->node); - free_extent_buffer(root->fs_info->chunk_root->commit_root); - root->fs_info->chunk_root->commit_root = - btrfs_root_node(root->fs_info->chunk_root); + switch_commit_root(root->fs_info->chunk_root); update_super_roots(root); @@ -1050,6 +1052,7 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans, cur_trans->commit_done = 1; root->fs_info->last_trans_committed = cur_trans->transid; + wake_up(&cur_trans->commit_wait); put_transaction(cur_trans); diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c index c13922206d1b..195606862618 100644 --- a/fs/btrfs/tree-log.c +++ b/fs/btrfs/tree-log.c @@ -264,7 +264,7 @@ static int process_one_buffer(struct btrfs_root *log, { if (wc->pin) btrfs_update_pinned_extents(log->fs_info->extent_root, - eb->start, eb->len, 1); + eb->start, eb->len, 1, 0); if (btrfs_buffer_uptodate(eb, gen)) { if (wc->write) -- cgit v1.2.3 From 20736abaa361bea488df6a1f66f6b37fb01107b9 Mon Sep 17 00:00:00 2001 From: Diego Calleja Date: Fri, 24 Jul 2009 11:06:52 -0400 Subject: Btrfs: Remove code duplication in comp_keys comp_keys is duplicating what is done in btrfs_comp_cpu_keys, so just call it. Signed-off-by: Diego Calleja Signed-off-by: Chris Mason --- fs/btrfs/ctree.c | 14 +------------- 1 file changed, 1 insertion(+), 13 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/ctree.c b/fs/btrfs/ctree.c index fdd423a550d6..91572091c0a4 100644 --- a/fs/btrfs/ctree.c +++ b/fs/btrfs/ctree.c @@ -557,19 +557,7 @@ static int comp_keys(struct btrfs_disk_key *disk, struct btrfs_key *k2) btrfs_disk_key_to_cpu(&k1, disk); - if (k1.objectid > k2->objectid) - return 1; - if (k1.objectid < k2->objectid) - return -1; - if (k1.type > k2->type) - return 1; - if (k1.type < k2->type) - return -1; - if (k1.offset > k2->offset) - return 1; - if (k1.offset < k2->offset) - return -1; - return 0; + return btrfs_comp_cpu_keys(&k1, k2); } /* -- cgit v1.2.3 From 1fcbac581be375ca0a686f72ee2b7fd1dbf386e7 Mon Sep 17 00:00:00 2001 From: Yan Zheng Date: Fri, 24 Jul 2009 11:06:53 -0400 Subject: Btrfs: find_free_dev_extent doesn't handle holes at the start of the device find_free_dev_extent does not properly handle the case where the device is not complete free, and there is a free extent at the beginning of the device. Signed-off-by: Yan Zheng Signed-off-by: Chris Mason --- fs/btrfs/volumes.c | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 55c37276a29f..074c1c56d8c4 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -758,9 +758,13 @@ static noinline int find_free_dev_extent(struct btrfs_trans_handle *trans, ret = btrfs_search_slot(trans, root, &key, path, 0, 0); if (ret < 0) goto error; - ret = btrfs_previous_item(root, path, 0, key.type); - if (ret < 0) - goto error; + if (ret > 0) { + ret = btrfs_previous_item(root, path, key.objectid, key.type); + if (ret < 0) + goto error; + if (ret > 0) + start_found = 1; + } l = path->nodes[0]; btrfs_item_key_to_cpu(l, &key, path->slots[0]); while (1) { -- cgit v1.2.3 From 0a4eefbb745ec0e8a5b694ae3f40cc34082d8f61 Mon Sep 17 00:00:00 2001 From: Yan Zheng Date: Fri, 24 Jul 2009 11:06:53 -0400 Subject: Btrfs: Fix ordering of key field checks in btrfs_previous_item Check objectid of item before checking the item type, otherwise we may return zero for a key that is actually too low. Signed-off-by: Yan Zheng Signed-off-by: Chris Mason --- fs/btrfs/ctree.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/ctree.c b/fs/btrfs/ctree.c index 91572091c0a4..978449af4ccd 100644 --- a/fs/btrfs/ctree.c +++ b/fs/btrfs/ctree.c @@ -4296,10 +4296,10 @@ int btrfs_previous_item(struct btrfs_root *root, path->slots[0]--; btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]); - if (found_key.type == type) - return 0; if (found_key.objectid < min_objectid) break; + if (found_key.type == type) + return 0; if (found_key.objectid == min_objectid && found_key.type < type) break; -- cgit v1.2.3 From d717aa1d31c36cb56059e97966cb76f0be021969 Mon Sep 17 00:00:00 2001 From: Yan Zheng Date: Fri, 24 Jul 2009 12:42:46 -0400 Subject: Btrfs: Avoid delayed reference update looping btrfs_split_leaf and btrfs_del_items can end up in a loop where one is constantly spliting a given leaf and the other is constantly merging it back with the adjacent nodes. There is a better fix for this, but in the interest of something small, this patch just changes btrfs_del_items back to balancing less often. Signed-off-by: Chris Mason --- fs/btrfs/ctree.c | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/ctree.c b/fs/btrfs/ctree.c index 978449af4ccd..3fdcc0512d3a 100644 --- a/fs/btrfs/ctree.c +++ b/fs/btrfs/ctree.c @@ -1040,9 +1040,6 @@ static noinline int balance_level(struct btrfs_trans_handle *trans, BTRFS_NODEPTRS_PER_BLOCK(root) / 4) return 0; - if (btrfs_header_nritems(mid) > 2) - return 0; - if (btrfs_header_nritems(mid) < 2) err_on_enospc = 1; @@ -3796,7 +3793,7 @@ int btrfs_del_items(struct btrfs_trans_handle *trans, struct btrfs_root *root, } /* delete the leaf if it is mostly empty */ - if (used < BTRFS_LEAF_DATA_SIZE(root) / 2) { + if (used < BTRFS_LEAF_DATA_SIZE(root) / 3) { /* push_leaf_left fixes the path. * make sure the path still points to our leaf * for possible call to del_ptr below -- cgit v1.2.3 From ebecd3d9d2adba144c15f1d35c78e0c26ead1bfd Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Fri, 24 Jul 2009 13:17:44 -0400 Subject: Btrfs: make flushoncommit mount option correctly wait on ordered_extents The commit_transaction call to wait_ordered_extents when snap_pending passes nocow_only=1 to process only NOCOW or PREALLOC extents. This isn't correct for the 'flushoncommit' mode, as it skips extents we just started IO on in start_delalloc_inodes. So, in the flushoncommit case, wait on all ordered extents. Otherwise, only pass the nocow_only flag to wait_ordered_extents if snap_pending. Signed-off-by: Sage Weil Signed-off-by: Chris Mason --- fs/btrfs/transaction.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index 32454d1c566f..e51d2bc532f8 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -942,9 +942,11 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans, mutex_unlock(&root->fs_info->trans_mutex); - if (flush_on_commit || snap_pending) { - if (flush_on_commit) - btrfs_start_delalloc_inodes(root); + if (flush_on_commit) { + btrfs_start_delalloc_inodes(root); + ret = btrfs_wait_ordered_extents(root, 0); + BUG_ON(ret); + } else if (snap_pending) { ret = btrfs_wait_ordered_extents(root, 1); BUG_ON(ret); } -- cgit v1.2.3 From 283bb1979fa8580c4037d8df251449368c292a3b Mon Sep 17 00:00:00 2001 From: Chris Mason Date: Fri, 24 Jul 2009 16:30:55 -0400 Subject: Btrfs: clear all space_info->full after removing a block group Btrfs allocates individual extents from block groups, and each block group has a specific type. It may hold metadata, data mirrored or striped etc. When we balance space (btrfs-vol -b) or remove a drive (btrfs-vol -r) we free block groups. Once a block group is freed, the space it was using on the device may be available for use by new block groups. btrfs_remove_block_group was clearing the flag that said 'our devices are full, don't even try to allocate new block groups', but it was only clearing that flag for a specific type of block group. This commit clears the full flag for all of the types of block groups, making it much more likely that we'll be able to balance space when the drive is close to full. Signed-off-by: Chris Mason --- fs/btrfs/extent-tree.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 9a489cc89fd3..508df5f7d2ea 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -7486,7 +7486,8 @@ int btrfs_remove_block_group(struct btrfs_trans_handle *trans, block_group->space_info->total_bytes -= block_group->key.offset; block_group->space_info->bytes_readonly -= block_group->key.offset; spin_unlock(&block_group->space_info->lock); - block_group->space_info->full = 0; + + btrfs_clear_space_info_full(root->fs_info); btrfs_put_block_group(block_group); btrfs_put_block_group(block_group); -- cgit v1.2.3 From 9779b72f0584fd53e0de53f62f205bf0dc0db553 Mon Sep 17 00:00:00 2001 From: Chris Mason Date: Fri, 24 Jul 2009 16:41:41 -0400 Subject: Btrfs: find smallest available device extent during chunk allocation Allocating new block group is easy when the disk has plenty of space. But things get difficult as the disk fills up, especially if the FS has been run through btrfs-vol -b. The balance operation is likely to make the total bytes available on the device greater than the largest extent we'll actually be able to allocate. But the device extent allocation code incorrectly assumes that a device with 5G free will be able to allocate a 5G extent. It isn't normally a problem because device extents don't get freed unless btrfs-vol -b is run. This fixes the device extent allocator to remember the largest free extent it can find, and then uses that value as a fallback. Signed-off-by: Chris Mason --- fs/btrfs/volumes.c | 12 ++++++++++-- 1 file changed, 10 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 074c1c56d8c4..5dbefd11b4af 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -721,7 +721,8 @@ error: */ static noinline int find_free_dev_extent(struct btrfs_trans_handle *trans, struct btrfs_device *device, - u64 num_bytes, u64 *start) + u64 num_bytes, u64 *start, + u64 *max_avail) { struct btrfs_key key; struct btrfs_root *root = device->dev_root; @@ -807,6 +808,10 @@ no_more_items: if (last_byte < search_start) last_byte = search_start; hole_size = key.offset - last_byte; + + if (hole_size > *max_avail) + *max_avail = hole_size; + if (key.offset > last_byte && hole_size >= num_bytes) { *start = last_byte; @@ -1625,6 +1630,7 @@ static int __btrfs_grow_device(struct btrfs_trans_handle *trans, device->fs_devices->total_rw_bytes += diff; device->total_bytes = new_size; + device->disk_total_bytes = new_size; btrfs_clear_space_info_full(device->dev_root->fs_info); return btrfs_update_device(trans, device); @@ -2175,6 +2181,7 @@ static int __btrfs_alloc_chunk(struct btrfs_trans_handle *trans, max_chunk_size); again: + max_avail = 0; if (!map || map->num_stripes != num_stripes) { kfree(map); map = kmalloc(map_lookup_size(num_stripes), GFP_NOFS); @@ -2223,7 +2230,8 @@ again: if (device->in_fs_metadata && avail >= min_free) { ret = find_free_dev_extent(trans, device, - min_free, &dev_offset); + min_free, &dev_offset, + &max_avail); if (ret == 0) { list_move_tail(&device->dev_alloc_list, &private_devs); -- cgit v1.2.3 From 631c07c8d12bcc6ce4a0fbfbd64ea843d78e2b10 Mon Sep 17 00:00:00 2001 From: Julia Lawall Date: Mon, 27 Jul 2009 13:57:00 -0400 Subject: Btrfs: Correct redundant test in add_inode_ref dir has already been tested. It seems that this test should be on the recently returned value inode. A simplified version of the semantic match that finds this problem is as follows: (http://www.emn.fr/x-info/coccinelle/) Signed-off-by: Julia Lawall Signed-off-by: Chris Mason --- fs/btrfs/tree-log.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c index 195606862618..11d0787c6188 100644 --- a/fs/btrfs/tree-log.c +++ b/fs/btrfs/tree-log.c @@ -797,7 +797,7 @@ static noinline int add_inode_ref(struct btrfs_trans_handle *trans, return -ENOENT; inode = read_one_inode(root, key->objectid); - BUG_ON(!dir); + BUG_ON(!inode); ref_ptr = btrfs_item_ptr_offset(eb, slot); ref_end = ref_ptr + btrfs_item_size_nr(eb, slot); -- cgit v1.2.3 From 68b38550ddbea13d296184bf69edff387618b1d3 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Mon, 27 Jul 2009 13:57:01 -0400 Subject: Btrfs: change how we unpin extents We are racy with async block caching and unpinning extents. This patch makes things much less complicated by only unpinning the extent if the block group is cached. We check the block_group->cached var under the block_group->lock spin lock. If it is set to BTRFS_CACHE_FINISHED then we update the pinned counters, and unpin the extent and add the free space back. If it is not set to this, we start the caching of the block group so the next time we unpin extents we can unpin the extent. This keeps us from racing with the async caching threads, lets us kill the fs wide async thread counter, and keeps us from having to set DELALLOC bits for every extent we hit if there are caching kthreads going. One thing that needed to be changed was btrfs_free_super_mirror_extents. Now instead of just looking for LOCKED extents, we also look for DIRTY extents, since we could have left some extents pinned in the previous transaction that will never get freed now that we are unmounting, which would cause us to leak memory. So btrfs_free_super_mirror_extents has been changed to btrfs_free_pinned_extents, and it will clear the extents locked for the super mirror, and any remaining pinned extents that may be present. Thank you, Signed-off-by: Josef Bacik Signed-off-by: Chris Mason --- fs/btrfs/ctree.h | 5 +- fs/btrfs/disk-io.c | 3 +- fs/btrfs/extent-tree.c | 149 ++++++++++++++----------------------------------- fs/btrfs/tree-log.c | 2 +- 4 files changed, 46 insertions(+), 113 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 42b03c4ee494..17ad92c29cfd 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -845,7 +845,6 @@ struct btrfs_fs_info { atomic_t async_submit_draining; atomic_t nr_async_bios; atomic_t async_delalloc_pages; - atomic_t async_caching_threads; /* * this is used by the balancing code to wait for all the pending @@ -1926,7 +1925,7 @@ int btrfs_run_delayed_refs(struct btrfs_trans_handle *trans, struct btrfs_root *root, unsigned long count); int btrfs_lookup_extent(struct btrfs_root *root, u64 start, u64 len); int btrfs_update_pinned_extents(struct btrfs_root *root, - u64 bytenr, u64 num, int pin, int mark_free); + u64 bytenr, u64 num, int pin); int btrfs_drop_leaf_ref(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct extent_buffer *leaf); int btrfs_cross_ref_exist(struct btrfs_trans_handle *trans, @@ -2011,7 +2010,7 @@ void btrfs_delalloc_reserve_space(struct btrfs_root *root, struct inode *inode, u64 bytes); void btrfs_delalloc_free_space(struct btrfs_root *root, struct inode *inode, u64 bytes); -void btrfs_free_super_mirror_extents(struct btrfs_fs_info *info); +void btrfs_free_pinned_extents(struct btrfs_fs_info *info); /* ctree.c */ int btrfs_bin_search(struct extent_buffer *eb, struct btrfs_key *key, int level, int *slot); diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index ec2c915f7f4a..c658397c7473 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -1567,7 +1567,6 @@ struct btrfs_root *open_ctree(struct super_block *sb, atomic_set(&fs_info->async_delalloc_pages, 0); atomic_set(&fs_info->async_submit_draining, 0); atomic_set(&fs_info->nr_async_bios, 0); - atomic_set(&fs_info->async_caching_threads, 0); fs_info->sb = sb; fs_info->max_extent = (u64)-1; fs_info->max_inline = 8192 * 1024; @@ -2339,7 +2338,7 @@ int close_ctree(struct btrfs_root *root) free_extent_buffer(root->fs_info->csum_root->commit_root); btrfs_free_block_groups(root->fs_info); - btrfs_free_super_mirror_extents(root->fs_info); + btrfs_free_pinned_extents(root->fs_info); del_fs_roots(fs_info); diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 508df5f7d2ea..08188f1615d9 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -153,18 +153,26 @@ block_group_cache_tree_search(struct btrfs_fs_info *info, u64 bytenr, return ret; } -void btrfs_free_super_mirror_extents(struct btrfs_fs_info *info) +/* + * We always set EXTENT_LOCKED for the super mirror extents so we don't + * overwrite them, so those bits need to be unset. Also, if we are unmounting + * with pinned extents still sitting there because we had a block group caching, + * we need to clear those now, since we are done. + */ +void btrfs_free_pinned_extents(struct btrfs_fs_info *info) { u64 start, end, last = 0; int ret; while (1) { ret = find_first_extent_bit(&info->pinned_extents, last, - &start, &end, EXTENT_LOCKED); + &start, &end, + EXTENT_LOCKED|EXTENT_DIRTY); if (ret) break; - unlock_extent(&info->pinned_extents, start, end, GFP_NOFS); + clear_extent_bits(&info->pinned_extents, start, end, + EXTENT_LOCKED|EXTENT_DIRTY, GFP_NOFS); last = end+1; } } @@ -209,8 +217,7 @@ static u64 add_new_free_space(struct btrfs_block_group_cache *block_group, while (start < end) { ret = find_first_extent_bit(&info->pinned_extents, start, &extent_start, &extent_end, - EXTENT_DIRTY|EXTENT_LOCKED| - EXTENT_DELALLOC); + EXTENT_DIRTY|EXTENT_LOCKED); if (ret) break; @@ -238,67 +245,6 @@ static u64 add_new_free_space(struct btrfs_block_group_cache *block_group, return total_added; } -DEFINE_MUTEX(discard_mutex); - -/* - * if async kthreads are running when we cross transactions, we mark any pinned - * extents with EXTENT_DELALLOC and then let the caching kthreads clean up those - * extents when they are done. Also we run this from btrfs_finish_extent_commit - * in case there were some pinned extents that were missed because we had - * already cached that block group. - */ -static void btrfs_discard_pinned_extents(struct btrfs_fs_info *fs_info, - struct btrfs_block_group_cache *cache) -{ - u64 start, end, last; - int ret; - - if (!cache) - last = 0; - else - last = cache->key.objectid; - - mutex_lock(&discard_mutex); - while (1) { - ret = find_first_extent_bit(&fs_info->pinned_extents, last, - &start, &end, EXTENT_DELALLOC); - if (ret) - break; - - if (cache && start >= cache->key.objectid + cache->key.offset) - break; - - - if (!cache) { - cache = btrfs_lookup_block_group(fs_info, start); - BUG_ON(!cache); - - start = max(start, cache->key.objectid); - end = min(end, cache->key.objectid + cache->key.offset - 1); - - if (block_group_cache_done(cache)) - btrfs_add_free_space(cache, start, - end - start + 1); - cache = NULL; - } else { - start = max(start, cache->key.objectid); - end = min(end, cache->key.objectid + cache->key.offset - 1); - btrfs_add_free_space(cache, start, end - start + 1); - } - - clear_extent_bits(&fs_info->pinned_extents, start, end, - EXTENT_DELALLOC, GFP_NOFS); - last = end + 1; - - if (need_resched()) { - mutex_unlock(&discard_mutex); - cond_resched(); - mutex_lock(&discard_mutex); - } - } - mutex_unlock(&discard_mutex); -} - static int caching_kthread(void *data) { struct btrfs_block_group_cache *block_group = data; @@ -317,7 +263,6 @@ static int caching_kthread(void *data) if (!path) return -ENOMEM; - atomic_inc(&fs_info->async_caching_threads); atomic_inc(&block_group->space_info->caching_threads); last = max_t(u64, block_group->key.objectid, BTRFS_SUPER_INFO_OFFSET); again: @@ -399,13 +344,9 @@ next: err: btrfs_free_path(path); up_read(&fs_info->extent_root->commit_root_sem); - atomic_dec(&fs_info->async_caching_threads); atomic_dec(&block_group->space_info->caching_threads); wake_up(&block_group->caching_q); - if (!ret) - btrfs_discard_pinned_extents(fs_info, block_group); - return 0; } @@ -1867,7 +1808,7 @@ static int run_one_delayed_ref(struct btrfs_trans_handle *trans, BUG_ON(ret); } btrfs_update_pinned_extents(root, node->bytenr, - node->num_bytes, 1, 0); + node->num_bytes, 1); update_reserved_extents(root, node->bytenr, node->num_bytes, 0); } @@ -3100,19 +3041,15 @@ static u64 first_logical_byte(struct btrfs_root *root, u64 search_start) } int btrfs_update_pinned_extents(struct btrfs_root *root, - u64 bytenr, u64 num, int pin, int mark_free) + u64 bytenr, u64 num, int pin) { u64 len; struct btrfs_block_group_cache *cache; struct btrfs_fs_info *fs_info = root->fs_info; - if (pin) { + if (pin) set_extent_dirty(&fs_info->pinned_extents, bytenr, bytenr + num - 1, GFP_NOFS); - } else { - clear_extent_dirty(&fs_info->pinned_extents, - bytenr, bytenr + num - 1, GFP_NOFS); - } while (num > 0) { cache = btrfs_lookup_block_group(fs_info, bytenr); @@ -3128,14 +3065,34 @@ int btrfs_update_pinned_extents(struct btrfs_root *root, spin_unlock(&cache->space_info->lock); fs_info->total_pinned += len; } else { + int unpin = 0; + + /* + * in order to not race with the block group caching, we + * only want to unpin the extent if we are cached. If + * we aren't cached, we want to start async caching this + * block group so we can free the extent the next time + * around. + */ spin_lock(&cache->space_info->lock); spin_lock(&cache->lock); - cache->pinned -= len; - cache->space_info->bytes_pinned -= len; + unpin = (cache->cached == BTRFS_CACHE_FINISHED); + if (likely(unpin)) { + cache->pinned -= len; + cache->space_info->bytes_pinned -= len; + fs_info->total_pinned -= len; + } spin_unlock(&cache->lock); spin_unlock(&cache->space_info->lock); - fs_info->total_pinned -= len; - if (block_group_cache_done(cache) && mark_free) + + if (likely(unpin)) + clear_extent_dirty(&fs_info->pinned_extents, + bytenr, bytenr + len -1, + GFP_NOFS); + else + cache_block_group(cache); + + if (unpin) btrfs_add_free_space(cache, bytenr, len); } btrfs_put_block_group(cache); @@ -3181,27 +3138,15 @@ int btrfs_copy_pinned(struct btrfs_root *root, struct extent_io_tree *copy) u64 last = 0; u64 start; u64 end; - bool caching_kthreads = false; struct extent_io_tree *pinned_extents = &root->fs_info->pinned_extents; int ret; - if (atomic_read(&root->fs_info->async_caching_threads)) - caching_kthreads = true; - while (1) { ret = find_first_extent_bit(pinned_extents, last, &start, &end, EXTENT_DIRTY); if (ret) break; - /* - * we need to make sure that the pinned extents don't go away - * while we are caching block groups - */ - if (unlikely(caching_kthreads)) - set_extent_delalloc(pinned_extents, start, end, - GFP_NOFS); - set_extent_dirty(copy, start, end, GFP_NOFS); last = end + 1; } @@ -3215,12 +3160,6 @@ int btrfs_finish_extent_commit(struct btrfs_trans_handle *trans, u64 start; u64 end; int ret; - int mark_free = 1; - - ret = find_first_extent_bit(&root->fs_info->pinned_extents, 0, - &start, &end, EXTENT_DELALLOC); - if (!ret) - mark_free = 0; while (1) { ret = find_first_extent_bit(unpin, 0, &start, &end, @@ -3231,16 +3170,12 @@ int btrfs_finish_extent_commit(struct btrfs_trans_handle *trans, ret = btrfs_discard_extent(root, start, end + 1 - start); /* unlocks the pinned mutex */ - btrfs_update_pinned_extents(root, start, end + 1 - start, 0, - mark_free); + btrfs_update_pinned_extents(root, start, end + 1 - start, 0); clear_extent_dirty(unpin, start, end, GFP_NOFS); cond_resched(); } - if (unlikely(!mark_free)) - btrfs_discard_pinned_extents(root->fs_info, NULL); - return ret; } @@ -3281,7 +3216,7 @@ static int pin_down_bytes(struct btrfs_trans_handle *trans, pinit: btrfs_set_path_blocking(path); /* unlocks the pinned mutex */ - btrfs_update_pinned_extents(root, bytenr, num_bytes, 1, 0); + btrfs_update_pinned_extents(root, bytenr, num_bytes, 1); BUG_ON(err < 0); return 0; @@ -3592,7 +3527,7 @@ int btrfs_free_extent(struct btrfs_trans_handle *trans, if (root_objectid == BTRFS_TREE_LOG_OBJECTID) { WARN_ON(owner >= BTRFS_FIRST_FREE_OBJECTID); /* unlocks the pinned mutex */ - btrfs_update_pinned_extents(root, bytenr, num_bytes, 1, 0); + btrfs_update_pinned_extents(root, bytenr, num_bytes, 1); update_reserved_extents(root, bytenr, num_bytes, 0); ret = 0; } else if (owner < BTRFS_FIRST_FREE_OBJECTID) { diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c index 11d0787c6188..d91b0de7c502 100644 --- a/fs/btrfs/tree-log.c +++ b/fs/btrfs/tree-log.c @@ -264,7 +264,7 @@ static int process_one_buffer(struct btrfs_root *log, { if (wc->pin) btrfs_update_pinned_extents(log->fs_info->extent_root, - eb->start, eb->len, 1, 0); + eb->start, eb->len, 1); if (btrfs_buffer_uptodate(eb, gen)) { if (wc->write) -- cgit v1.2.3 From 7b91e2661addd8e2419cb45f6a322aa5dab9bcee Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Thu, 23 Jul 2009 15:22:30 -0400 Subject: cifs: fix error handling in mount-time DFS referral chasing code If the referral is malformed or the hostname can't be resolved, then the current code generates an oops. Fix it to handle these errors gracefully. Reported-by: Sandro Mathys Acked-by: Igor Mammedov CC: Stable Signed-off-by: Jeff Layton Signed-off-by: Steve French --- fs/cifs/cifs_dfs_ref.c | 12 +++++++++--- fs/cifs/connect.c | 13 ++++++++++--- 2 files changed, 19 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/cifs/cifs_dfs_ref.c b/fs/cifs/cifs_dfs_ref.c index 3bb11be8b6a8..606912d8f2a8 100644 --- a/fs/cifs/cifs_dfs_ref.c +++ b/fs/cifs/cifs_dfs_ref.c @@ -55,7 +55,7 @@ void cifs_dfs_release_automount_timer(void) * i.e. strips from UNC trailing path that is not part of share * name and fixup missing '\' in the begining of DFS node refferal * if neccessary. - * Returns pointer to share name on success or NULL on error. + * Returns pointer to share name on success or ERR_PTR on error. * Caller is responsible for freeing returned string. */ static char *cifs_get_share_name(const char *node_name) @@ -68,7 +68,7 @@ static char *cifs_get_share_name(const char *node_name) UNC = kmalloc(len+2 /*for term null and additional \ if it's missed */, GFP_KERNEL); if (!UNC) - return NULL; + return ERR_PTR(-ENOMEM); /* get share name and server name */ if (node_name[1] != '\\') { @@ -87,7 +87,7 @@ static char *cifs_get_share_name(const char *node_name) cERROR(1, ("%s: no server name end in node name: %s", __func__, node_name)); kfree(UNC); - return NULL; + return ERR_PTR(-EINVAL); } /* find sharename end */ @@ -133,6 +133,12 @@ char *cifs_compose_mount_options(const char *sb_mountdata, return ERR_PTR(-EINVAL); *devname = cifs_get_share_name(ref->node_name); + if (IS_ERR(*devname)) { + rc = PTR_ERR(*devname); + *devname = NULL; + goto compose_mount_options_err; + } + rc = dns_resolve_server_name_to_ip(*devname, &srvIP); if (rc != 0) { cERROR(1, ("%s: Failed to resolve server part of %s to IP: %d", diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index fc44d316d0bb..f2486889b7bb 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -2544,11 +2544,20 @@ remote_path_check: if (mount_data != mount_data_global) kfree(mount_data); + mount_data = cifs_compose_mount_options( cifs_sb->mountdata, full_path + 1, referrals, &fake_devname); - kfree(fake_devname); + free_dfs_info_array(referrals, num_referrals); + kfree(fake_devname); + kfree(full_path); + + if (IS_ERR(mount_data)) { + rc = PTR_ERR(mount_data); + mount_data = NULL; + goto mount_fail_check; + } if (tcon) cifs_put_tcon(tcon); @@ -2556,8 +2565,6 @@ remote_path_check: cifs_put_smb_ses(pSesInfo); cleanup_volume_info(&volume_info); - FreeXid(xid); - kfree(full_path); referral_walks_count++; goto try_mount_again; } -- cgit v1.2.3 From f25784b35f590c81d5fb8245a8cd45e1afb6f1b2 Mon Sep 17 00:00:00 2001 From: Yan Zheng Date: Tue, 28 Jul 2009 08:41:57 -0400 Subject: Btrfs: Fix async caching interaction with unmount - don't stop the caching thread until btrfs_commit_super return. - if caching is interrupted by umount, set last to (u64)-1. otherwise the un-scanned range of block group will be considered as free extent. Signed-off-by: Chris Mason --- fs/btrfs/disk-io.c | 3 +++ fs/btrfs/extent-tree.c | 4 +++- 2 files changed, 6 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index c658397c7473..3a9b88759880 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -2317,6 +2317,9 @@ int close_ctree(struct btrfs_root *root) printk(KERN_ERR "btrfs: commit super ret %d\n", ret); } + fs_info->closing = 2; + smp_mb(); + if (fs_info->delalloc_bytes) { printk(KERN_INFO "btrfs: at unmount delalloc count %llu\n", (unsigned long long)fs_info->delalloc_bytes); diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 08188f1615d9..fadf69a2764b 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -288,8 +288,10 @@ again: while (1) { smp_mb(); - if (block_group->fs_info->closing) + if (block_group->fs_info->closing > 1) { + last = (u64)-1; break; + } leaf = path->nodes[0]; slot = path->slots[0]; -- cgit v1.2.3 From 0f58b44582001c8bcdb75f36cf85ebbe5170e959 Mon Sep 17 00:00:00 2001 From: Peter Oberparleiter Date: Tue, 14 Jul 2009 17:56:15 +0200 Subject: sysfs: fix hardlink count on device_move Update directory hardlink count when moving kobjects to a new parent. Fixes the following problem which occurs when several devices are moved to the same parent and then unregistered: > ls -laF /sys/devices/css0/defunct/ > total 0 > drwxr-xr-x 4294967295 root root 0 2009-07-14 17:02 ./ > drwxr-xr-x 114 root root 0 2009-07-14 17:02 ../ > drwxr-xr-x 2 root root 0 2009-07-14 17:01 power/ > -rw-r--r-- 1 root root 4096 2009-07-14 17:01 uevent Signed-off-by: Peter Oberparleiter Cc: stable Signed-off-by: Greg Kroah-Hartman --- fs/sysfs/dir.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'fs') diff --git a/fs/sysfs/dir.c b/fs/sysfs/dir.c index d88d0fac9fa5..14f2d71ea3ce 100644 --- a/fs/sysfs/dir.c +++ b/fs/sysfs/dir.c @@ -939,8 +939,10 @@ again: /* Remove from old parent's list and insert into new parent's list. */ sysfs_unlink_sibling(sd); sysfs_get(new_parent_sd); + drop_nlink(old_parent->d_inode); sysfs_put(sd->s_parent); sd->s_parent = new_parent_sd; + inc_nlink(new_parent->d_inode); sysfs_link_sibling(sd); out_unlock: -- cgit v1.2.3 From 6352a29305373ae6196491e6d4669f301e26492e Mon Sep 17 00:00:00 2001 From: Tyler Hicks Date: Tue, 28 Jul 2009 13:57:01 -0500 Subject: eCryptfs: Check Tag 11 literal data buffer size Tag 11 packets are stored in the metadata section of an eCryptfs file to store the key signature(s) used to encrypt the file encryption key. After extracting the packet length field to determine the key signature length, a check is not performed to see if the length would exceed the key signature buffer size that was passed into parse_tag_11_packet(). Thanks to Ramon de Carvalho Valle for finding this bug using fsfuzzer. Signed-off-by: Tyler Hicks Cc: stable@kernel.org (2.6.27 and 30) Signed-off-by: Linus Torvalds --- fs/ecryptfs/keystore.c | 6 ++++++ 1 file changed, 6 insertions(+) (limited to 'fs') diff --git a/fs/ecryptfs/keystore.c b/fs/ecryptfs/keystore.c index af737bb56cb7..5414253d4c97 100644 --- a/fs/ecryptfs/keystore.c +++ b/fs/ecryptfs/keystore.c @@ -1449,6 +1449,12 @@ parse_tag_11_packet(unsigned char *data, unsigned char *contents, rc = -EINVAL; goto out; } + if (unlikely((*tag_11_contents_size) > max_contents_bytes)) { + printk(KERN_ERR "Literal data section in tag 11 packet exceeds " + "expected size\n"); + rc = -EINVAL; + goto out; + } if (data[(*packet_size)++] != 0x62) { printk(KERN_WARNING "Unrecognizable packet\n"); rc = -EINVAL; -- cgit v1.2.3 From f151cd2c54ddc7714e2f740681350476cda03a28 Mon Sep 17 00:00:00 2001 From: Ramon de Carvalho Valle Date: Tue, 28 Jul 2009 13:58:22 -0500 Subject: eCryptfs: parse_tag_3_packet check tag 3 packet encrypted key size The parse_tag_3_packet function does not check if the tag 3 packet contains a encrypted key size larger than ECRYPTFS_MAX_ENCRYPTED_KEY_BYTES. Signed-off-by: Ramon de Carvalho Valle [tyhicks@linux.vnet.ibm.com: Added printk newline and changed goto to out_free] Signed-off-by: Tyler Hicks Cc: stable@kernel.org (2.6.27 and 30) Signed-off-by: Linus Torvalds --- fs/ecryptfs/keystore.c | 7 +++++++ 1 file changed, 7 insertions(+) (limited to 'fs') diff --git a/fs/ecryptfs/keystore.c b/fs/ecryptfs/keystore.c index 5414253d4c97..259525c9abb8 100644 --- a/fs/ecryptfs/keystore.c +++ b/fs/ecryptfs/keystore.c @@ -1303,6 +1303,13 @@ parse_tag_3_packet(struct ecryptfs_crypt_stat *crypt_stat, } (*new_auth_tok)->session_key.encrypted_key_size = (body_size - (ECRYPTFS_SALT_SIZE + 5)); + if ((*new_auth_tok)->session_key.encrypted_key_size + > ECRYPTFS_MAX_ENCRYPTED_KEY_BYTES) { + printk(KERN_WARNING "Tag 3 packet contains key larger " + "than ECRYPTFS_MAX_ENCRYPTED_KEY_BYTES\n"); + rc = -EINVAL; + goto out_free; + } if (unlikely(data[(*packet_size)++] != 0x04)) { printk(KERN_WARNING "Unknown version number [%d]\n", data[(*packet_size) - 1]); -- cgit v1.2.3 From dddac6a7b445de95515f64fdf82fe5dc36c02f26 Mon Sep 17 00:00:00 2001 From: Alan Jenkins Date: Wed, 29 Jul 2009 21:07:55 +0200 Subject: PM / Hibernate: Replace bdget call with simple atomic_inc of i_count Create bdgrab(). This function copies an existing reference to a block_device. It is safe to call from any context. Hibernation code wishes to copy a reference to the active swap device. Right now it calls bdget() under a spinlock, but this is wrong because bdget() can sleep. It doesn't need a full bdget() because we already hold a reference to active swap devices (and the spinlock protects against swapoff). Fixes http://bugzilla.kernel.org/show_bug.cgi?id=13827 Signed-off-by: Alan Jenkins Signed-off-by: Rafael J. Wysocki --- fs/block_dev.c | 10 ++++++++++ include/linux/fs.h | 1 + mm/swapfile.c | 4 ++-- 3 files changed, 13 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/block_dev.c b/fs/block_dev.c index 3a6d4fb2a329..94dfda24c06e 100644 --- a/fs/block_dev.c +++ b/fs/block_dev.c @@ -564,6 +564,16 @@ struct block_device *bdget(dev_t dev) EXPORT_SYMBOL(bdget); +/** + * bdgrab -- Grab a reference to an already referenced block device + * @bdev: Block device to grab a reference to. + */ +struct block_device *bdgrab(struct block_device *bdev) +{ + atomic_inc(&bdev->bd_inode->i_count); + return bdev; +} + long nr_blockdev_pages(void) { struct block_device *bdev; diff --git a/include/linux/fs.h b/include/linux/fs.h index 0872372184fe..a36ffa5a77a4 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -1946,6 +1946,7 @@ extern void putname(const char *name); extern int register_blkdev(unsigned int, const char *); extern void unregister_blkdev(unsigned int, const char *); extern struct block_device *bdget(dev_t); +extern struct block_device *bdgrab(struct block_device *bdev); extern void bd_set_size(struct block_device *, loff_t size); extern void bd_forget(struct inode *inode); extern void bdput(struct block_device *); diff --git a/mm/swapfile.c b/mm/swapfile.c index d1ade1a48ee7..8ffdc0d23c53 100644 --- a/mm/swapfile.c +++ b/mm/swapfile.c @@ -753,7 +753,7 @@ int swap_type_of(dev_t device, sector_t offset, struct block_device **bdev_p) if (!bdev) { if (bdev_p) - *bdev_p = bdget(sis->bdev->bd_dev); + *bdev_p = bdgrab(sis->bdev); spin_unlock(&swap_lock); return i; @@ -765,7 +765,7 @@ int swap_type_of(dev_t device, sector_t offset, struct block_device **bdev_p) struct swap_extent, list); if (se->start_block == offset) { if (bdev_p) - *bdev_p = bdget(sis->bdev->bd_dev); + *bdev_p = bdgrab(sis->bdev); spin_unlock(&swap_lock); bdput(bdev); -- cgit v1.2.3 From 5c8053652328693d10551131432ef3573e77ed2d Mon Sep 17 00:00:00 2001 From: Catalin Marinas Date: Wed, 29 Jul 2009 15:04:11 -0700 Subject: fs/ramfs/file-nommu.c needs include/linux/sched.h This file makes use of various macros defined in files like asm/current.h or asm-generic/resource.h. All these files can be included via sched.h. The building of the !MMU ARM kernel (with additional patches) fails without this change. Signed-off-by: Catalin Marinas Acked-by: David Howells Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/ramfs/file-nommu.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/ramfs/file-nommu.c b/fs/ramfs/file-nommu.c index ebb2c417912c..11f0c06316de 100644 --- a/fs/ramfs/file-nommu.c +++ b/fs/ramfs/file-nommu.c @@ -20,6 +20,7 @@ #include #include #include +#include #include #include "internal.h" -- cgit v1.2.3 From 5bd9052d79daa4c8beb45436c408b6de672adb82 Mon Sep 17 00:00:00 2001 From: Steve French Date: Thu, 30 Jul 2009 02:26:14 +0000 Subject: [CIFS] Updates fs/cifs/CHANGES Signed-off-by: Steve French --- fs/cifs/CHANGES | 6 ++++++ 1 file changed, 6 insertions(+) (limited to 'fs') diff --git a/fs/cifs/CHANGES b/fs/cifs/CHANGES index 92888aa90749..651cefde385b 100644 --- a/fs/cifs/CHANGES +++ b/fs/cifs/CHANGES @@ -1,3 +1,9 @@ +Version 1.60 +------------- +Fix memory leak in reconnect. Fix oops in DFS mount error path. +Set s_maxbytes to smaller (the max that vfs can handle) so that +sendfile will now work over cifs mounts again. + Version 1.59 ------------ Client uses server inode numbers (which are persistent) rather than -- cgit v1.2.3 From 2163b1e616c41c286f5ab79912671cd4bf52057c Mon Sep 17 00:00:00 2001 From: Steven Whitehouse Date: Thu, 25 Jun 2009 16:30:26 +0100 Subject: GFS2: Shrink the shrinker This patch removes some of the special cases that the shrinker was trying to deal with. As a result we leave fewer items on the list and none at all which cannot be demoted. This makes the list scanning more efficient and solves some issues seen with large numbers of inodes. Signed-off-by: Steven Whitehouse --- fs/gfs2/glock.c | 23 +++++------------------ 1 file changed, 5 insertions(+), 18 deletions(-) (limited to 'fs') diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c index 297421c0427a..fdb796c4f940 100644 --- a/fs/gfs2/glock.c +++ b/fs/gfs2/glock.c @@ -1300,7 +1300,6 @@ static int gfs2_shrink_glock_memory(int nr, gfp_t gfp_mask) struct gfs2_glock *gl; int may_demote; int nr_skipped = 0; - int got_ref = 0; LIST_HEAD(skipped); if (nr == 0) @@ -1318,7 +1317,6 @@ static int gfs2_shrink_glock_memory(int nr, gfp_t gfp_mask) /* Test for being demotable */ if (!test_and_set_bit(GLF_LOCK, &gl->gl_flags)) { gfs2_glock_hold(gl); - got_ref = 1; spin_unlock(&lru_lock); spin_lock(&gl->gl_spin); may_demote = demote_ok(gl); @@ -1327,25 +1325,14 @@ static int gfs2_shrink_glock_memory(int nr, gfp_t gfp_mask) if (may_demote) { handle_callback(gl, LM_ST_UNLOCKED, 0); nr--; - if (queue_delayed_work(glock_workqueue, &gl->gl_work, 0) == 0) - gfs2_glock_put(gl); - got_ref = 0; } + if (queue_delayed_work(glock_workqueue, &gl->gl_work, 0) == 0) + gfs2_glock_put(gl); spin_lock(&lru_lock); - if (may_demote) - continue; - } - if (list_empty(&gl->gl_lru) && - (atomic_read(&gl->gl_ref) <= (2 + got_ref))) { - nr_skipped++; - list_add(&gl->gl_lru, &skipped); - } - if (got_ref) { - spin_unlock(&lru_lock); - gfs2_glock_put(gl); - spin_lock(&lru_lock); - got_ref = 0; + continue; } + nr_skipped++; + list_add(&gl->gl_lru, &skipped); } list_splice(&skipped, &lru_list); atomic_add(nr_skipped, &lru_count); -- cgit v1.2.3 From 1946f70ab5e4eb8b54a8eaaedba2293a3750ab7e Mon Sep 17 00:00:00 2001 From: Benjamin Marzinski Date: Thu, 25 Jun 2009 15:09:51 -0500 Subject: GFS2: keep statfs info in sync on grows GFS2 wasn't syncing its statfs info on grows. This causes a problem when you grow the filesystem on multiple nodes. GFS2 would calculate the new space based on the resource groups (which are always current), and then assume that the filesystem had grown the from the existing statfs size. If you grew the filesystem on two different nodes in a short time, the second node wouldn't see the statfs size change from the first node, and would assume that it was grown by a larger amount than it was. When all these changes were synced out, the total fileystem size would be incorrect (the first grow would be counted twice). This patch syncs makes GFS2 read in the statfs changes from disk before a grow, and write them out after the grow, while the master statfs inode is locked. Signed-off-by: Benjamin Marzinski Signed-off-by: Steven Whitehouse --- fs/gfs2/aops.c | 39 +++++++++++++++++++++++++++++++++++++++ fs/gfs2/super.c | 39 +++++++++++++++++++++++++-------------- fs/gfs2/super.h | 4 ++++ 3 files changed, 68 insertions(+), 14 deletions(-) (limited to 'fs') diff --git a/fs/gfs2/aops.c b/fs/gfs2/aops.c index 03ebb439ace0..7ebae9a4ecc0 100644 --- a/fs/gfs2/aops.c +++ b/fs/gfs2/aops.c @@ -624,6 +624,7 @@ static int gfs2_write_begin(struct file *file, struct address_space *mapping, { struct gfs2_inode *ip = GFS2_I(mapping->host); struct gfs2_sbd *sdp = GFS2_SB(mapping->host); + struct gfs2_inode *m_ip = GFS2_I(sdp->sd_statfs_inode); unsigned int data_blocks = 0, ind_blocks = 0, rblocks; int alloc_required; int error = 0; @@ -637,6 +638,14 @@ static int gfs2_write_begin(struct file *file, struct address_space *mapping, error = gfs2_glock_nq(&ip->i_gh); if (unlikely(error)) goto out_uninit; + if (&ip->i_inode == sdp->sd_rindex) { + error = gfs2_glock_nq_init(m_ip->i_gl, LM_ST_EXCLUSIVE, + GL_NOCACHE, &m_ip->i_gh); + if (unlikely(error)) { + gfs2_glock_dq(&ip->i_gh); + goto out_uninit; + } + } error = gfs2_write_alloc_required(ip, pos, len, &alloc_required); if (error) @@ -667,6 +676,8 @@ static int gfs2_write_begin(struct file *file, struct address_space *mapping, rblocks += data_blocks ? data_blocks : 1; if (ind_blocks || data_blocks) rblocks += RES_STATFS + RES_QUOTA; + if (&ip->i_inode == sdp->sd_rindex) + rblocks += 2 * RES_STATFS; error = gfs2_trans_begin(sdp, rblocks, PAGE_CACHE_SIZE/sdp->sd_sb.sb_bsize); @@ -712,6 +723,10 @@ out_alloc_put: gfs2_alloc_put(ip); } out_unlock: + if (&ip->i_inode == sdp->sd_rindex) { + gfs2_glock_dq(&m_ip->i_gh); + gfs2_holder_uninit(&m_ip->i_gh); + } gfs2_glock_dq(&ip->i_gh); out_uninit: gfs2_holder_uninit(&ip->i_gh); @@ -725,14 +740,21 @@ out_uninit: static void adjust_fs_space(struct inode *inode) { struct gfs2_sbd *sdp = inode->i_sb->s_fs_info; + struct gfs2_inode *m_ip = GFS2_I(sdp->sd_statfs_inode); + struct gfs2_inode *l_ip = GFS2_I(sdp->sd_sc_inode); struct gfs2_statfs_change_host *m_sc = &sdp->sd_statfs_master; struct gfs2_statfs_change_host *l_sc = &sdp->sd_statfs_local; + struct buffer_head *m_bh, *l_bh; u64 fs_total, new_free; /* Total up the file system space, according to the latest rindex. */ fs_total = gfs2_ri_total(sdp); + if (gfs2_meta_inode_buffer(m_ip, &m_bh) != 0) + return; spin_lock(&sdp->sd_statfs_spin); + gfs2_statfs_change_in(m_sc, m_bh->b_data + + sizeof(struct gfs2_dinode)); if (fs_total > (m_sc->sc_total + l_sc->sc_total)) new_free = fs_total - (m_sc->sc_total + l_sc->sc_total); else @@ -741,6 +763,13 @@ static void adjust_fs_space(struct inode *inode) fs_warn(sdp, "File system extended by %llu blocks.\n", (unsigned long long)new_free); gfs2_statfs_change(sdp, new_free, new_free, 0); + + if (gfs2_meta_inode_buffer(l_ip, &l_bh) != 0) + goto out; + update_statfs(sdp, m_bh, l_bh); + brelse(l_bh); +out: + brelse(m_bh); } /** @@ -763,6 +792,7 @@ static int gfs2_stuffed_write_end(struct inode *inode, struct buffer_head *dibh, { struct gfs2_inode *ip = GFS2_I(inode); struct gfs2_sbd *sdp = GFS2_SB(inode); + struct gfs2_inode *m_ip = GFS2_I(sdp->sd_statfs_inode); u64 to = pos + copied; void *kaddr; unsigned char *buf = dibh->b_data + sizeof(struct gfs2_dinode); @@ -794,6 +824,10 @@ static int gfs2_stuffed_write_end(struct inode *inode, struct buffer_head *dibh, brelse(dibh); gfs2_trans_end(sdp); + if (inode == sdp->sd_rindex) { + gfs2_glock_dq(&m_ip->i_gh); + gfs2_holder_uninit(&m_ip->i_gh); + } gfs2_glock_dq(&ip->i_gh); gfs2_holder_uninit(&ip->i_gh); return copied; @@ -823,6 +857,7 @@ static int gfs2_write_end(struct file *file, struct address_space *mapping, struct inode *inode = page->mapping->host; struct gfs2_inode *ip = GFS2_I(inode); struct gfs2_sbd *sdp = GFS2_SB(inode); + struct gfs2_inode *m_ip = GFS2_I(sdp->sd_statfs_inode); struct buffer_head *dibh; struct gfs2_alloc *al = ip->i_alloc; unsigned int from = pos & (PAGE_CACHE_SIZE - 1); @@ -865,6 +900,10 @@ failed: gfs2_quota_unlock(ip); gfs2_alloc_put(ip); } + if (inode == sdp->sd_rindex) { + gfs2_glock_dq(&m_ip->i_gh); + gfs2_holder_uninit(&m_ip->i_gh); + } gfs2_glock_dq(&ip->i_gh); gfs2_holder_uninit(&ip->i_gh); return ret; diff --git a/fs/gfs2/super.c b/fs/gfs2/super.c index 0a6801336470..552e321cee5e 100644 --- a/fs/gfs2/super.c +++ b/fs/gfs2/super.c @@ -353,7 +353,7 @@ fail: return error; } -static void gfs2_statfs_change_in(struct gfs2_statfs_change_host *sc, const void *buf) +void gfs2_statfs_change_in(struct gfs2_statfs_change_host *sc, const void *buf) { const struct gfs2_statfs_change *str = buf; @@ -441,6 +441,29 @@ void gfs2_statfs_change(struct gfs2_sbd *sdp, s64 total, s64 free, brelse(l_bh); } +void update_statfs(struct gfs2_sbd *sdp, struct buffer_head *m_bh, + struct buffer_head *l_bh) +{ + struct gfs2_inode *m_ip = GFS2_I(sdp->sd_statfs_inode); + struct gfs2_inode *l_ip = GFS2_I(sdp->sd_sc_inode); + struct gfs2_statfs_change_host *m_sc = &sdp->sd_statfs_master; + struct gfs2_statfs_change_host *l_sc = &sdp->sd_statfs_local; + + gfs2_trans_add_bh(l_ip->i_gl, l_bh, 1); + + spin_lock(&sdp->sd_statfs_spin); + m_sc->sc_total += l_sc->sc_total; + m_sc->sc_free += l_sc->sc_free; + m_sc->sc_dinodes += l_sc->sc_dinodes; + memset(l_sc, 0, sizeof(struct gfs2_statfs_change)); + memset(l_bh->b_data + sizeof(struct gfs2_dinode), + 0, sizeof(struct gfs2_statfs_change)); + spin_unlock(&sdp->sd_statfs_spin); + + gfs2_trans_add_bh(m_ip->i_gl, m_bh, 1); + gfs2_statfs_change_out(m_sc, m_bh->b_data + sizeof(struct gfs2_dinode)); +} + int gfs2_statfs_sync(struct gfs2_sbd *sdp) { struct gfs2_inode *m_ip = GFS2_I(sdp->sd_statfs_inode); @@ -477,19 +500,7 @@ int gfs2_statfs_sync(struct gfs2_sbd *sdp) if (error) goto out_bh2; - gfs2_trans_add_bh(l_ip->i_gl, l_bh, 1); - - spin_lock(&sdp->sd_statfs_spin); - m_sc->sc_total += l_sc->sc_total; - m_sc->sc_free += l_sc->sc_free; - m_sc->sc_dinodes += l_sc->sc_dinodes; - memset(l_sc, 0, sizeof(struct gfs2_statfs_change)); - memset(l_bh->b_data + sizeof(struct gfs2_dinode), - 0, sizeof(struct gfs2_statfs_change)); - spin_unlock(&sdp->sd_statfs_spin); - - gfs2_trans_add_bh(m_ip->i_gl, m_bh, 1); - gfs2_statfs_change_out(m_sc, m_bh->b_data + sizeof(struct gfs2_dinode)); + update_statfs(sdp, m_bh, l_bh); gfs2_trans_end(sdp); diff --git a/fs/gfs2/super.h b/fs/gfs2/super.h index b56413e3e40d..22e0417ed996 100644 --- a/fs/gfs2/super.h +++ b/fs/gfs2/super.h @@ -40,6 +40,10 @@ extern int gfs2_make_fs_rw(struct gfs2_sbd *sdp); extern int gfs2_statfs_init(struct gfs2_sbd *sdp); extern void gfs2_statfs_change(struct gfs2_sbd *sdp, s64 total, s64 free, s64 dinodes); +extern void gfs2_statfs_change_in(struct gfs2_statfs_change_host *sc, + const void *buf); +extern void update_statfs(struct gfs2_sbd *sdp, struct buffer_head *m_bh, + struct buffer_head *l_bh); extern int gfs2_statfs_sync(struct gfs2_sbd *sdp); extern int gfs2_freeze_fs(struct gfs2_sbd *sdp); -- cgit v1.2.3 From a51b56fff3f04fc5aa66b21a2a6d693ee9862d66 Mon Sep 17 00:00:00 2001 From: Benjamin Marzinski Date: Tue, 30 Jun 2009 13:51:11 -0500 Subject: GFS2: Fix panic in glock memory shrinker It is possible for gfs2_shrink_glock_memory() to check a glock for demotion that's in the process of being freed by gfs2_glock_put(). In this case, gfs2_shrink_glock_memory() will acquire a new reference to this glock, and then try to free the glock itself when it drops the refernce. To solve this, gfs2_shrink_glock_memory() just needs to check if the glock is in the process of being freed, and if so skip it without ever unlocking the lru_lock. Signed-off-by: Benjamin Marzinski Acked-by: Bob Peterson Signed-off-by: Steven Whitehouse --- fs/gfs2/glock.c | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'fs') diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c index fdb796c4f940..827136ee794c 100644 --- a/fs/gfs2/glock.c +++ b/fs/gfs2/glock.c @@ -1314,6 +1314,10 @@ static int gfs2_shrink_glock_memory(int nr, gfp_t gfp_mask) list_del_init(&gl->gl_lru); atomic_dec(&lru_count); + /* Check if glock is about to be freed */ + if (atomic_read(&gl->gl_ref) == 0) + continue; + /* Test for being demotable */ if (!test_and_set_bit(GLF_LOCK, &gl->gl_flags)) { gfs2_glock_hold(gl); -- cgit v1.2.3 From 1e19a19584b332eb92a573b66b7342fb97e67507 Mon Sep 17 00:00:00 2001 From: Steven Whitehouse Date: Fri, 10 Jul 2009 21:13:38 +0100 Subject: GFS2: Don't try and dealloc own inode When searching for unlinked, but still allocated inodes during block allocation, avoid the block relating to the inode that is doing the allocation. This fixes a hang caused when an unlinked, but still open, inode tries to allocate some more blocks and lands up finding itself during the search for deallocatable inodes. Signed-off-by: Steven Whitehouse --- fs/gfs2/rgrp.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/gfs2/rgrp.c b/fs/gfs2/rgrp.c index daa4ae341a29..5e5074176daa 100644 --- a/fs/gfs2/rgrp.c +++ b/fs/gfs2/rgrp.c @@ -961,7 +961,8 @@ static int try_rgrp_fit(struct gfs2_rgrpd *rgd, struct gfs2_alloc *al) * Returns: The inode, if one has been found */ -static struct inode *try_rgrp_unlink(struct gfs2_rgrpd *rgd, u64 *last_unlinked) +static struct inode *try_rgrp_unlink(struct gfs2_rgrpd *rgd, u64 *last_unlinked, + u64 skip) { struct inode *inode; u32 goal = 0, block; @@ -985,6 +986,8 @@ static struct inode *try_rgrp_unlink(struct gfs2_rgrpd *rgd, u64 *last_unlinked) goal++; if (*last_unlinked != NO_BLOCK && no_addr <= *last_unlinked) continue; + if (no_addr == skip) + continue; *last_unlinked = no_addr; inode = gfs2_inode_lookup(rgd->rd_sbd->sd_vfs, DT_UNKNOWN, no_addr, -1, 1); @@ -1104,7 +1107,7 @@ static struct inode *get_local_rgrp(struct gfs2_inode *ip, u64 *last_unlinked) if (try_rgrp_fit(rgd, al)) goto out; if (rgd->rd_flags & GFS2_RDF_CHECK) - inode = try_rgrp_unlink(rgd, last_unlinked); + inode = try_rgrp_unlink(rgd, last_unlinked, ip->i_no_addr); if (!rg_locked) gfs2_glock_dq_uninit(&al->al_rgd_gh); if (inode) @@ -1138,7 +1141,7 @@ static struct inode *get_local_rgrp(struct gfs2_inode *ip, u64 *last_unlinked) if (try_rgrp_fit(rgd, al)) goto out; if (rgd->rd_flags & GFS2_RDF_CHECK) - inode = try_rgrp_unlink(rgd, last_unlinked); + inode = try_rgrp_unlink(rgd, last_unlinked, ip->i_no_addr); if (!rg_locked) gfs2_glock_dq_uninit(&al->al_rgd_gh); if (inode) -- cgit v1.2.3 From 8ff22a6f9bdaac87c0eeb1d56c736181f11b4221 Mon Sep 17 00:00:00 2001 From: Benjamin Marzinski Date: Fri, 10 Jul 2009 18:04:24 -0500 Subject: GFS2: Don't put unlikely reclaim candidates on the reclaim list. GFS2 was placing far too many glocks on the reclaim list that were not good candidates for freeing up from cache. These locks would sit there and repeatedly get scanned to see if they could be reclaimed, wasting a lot of time when there was memory pressure. This fix does more checks on the locks to see if they are actually likely to be removable from cache. Signed-off-by: Benjamin Marzinski Signed-off-by: Steven Whitehouse --- fs/gfs2/glock.c | 72 ++++++++++++++++++++++++++++++++++++--------------------- 1 file changed, 46 insertions(+), 26 deletions(-) (limited to 'fs') diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c index 827136ee794c..f041a89e1ab8 100644 --- a/fs/gfs2/glock.c +++ b/fs/gfs2/glock.c @@ -173,6 +173,26 @@ static void gfs2_glock_hold(struct gfs2_glock *gl) atomic_inc(&gl->gl_ref); } +/** + * demote_ok - Check to see if it's ok to unlock a glock + * @gl: the glock + * + * Returns: 1 if it's ok + */ + +static int demote_ok(const struct gfs2_glock *gl) +{ + const struct gfs2_glock_operations *glops = gl->gl_ops; + + if (gl->gl_state == LM_ST_UNLOCKED) + return 0; + if (!list_empty(&gl->gl_holders)) + return 0; + if (glops->go_demote_ok) + return glops->go_demote_ok(gl); + return 1; +} + /** * gfs2_glock_schedule_for_reclaim - Add a glock to the reclaim list * @gl: the glock @@ -181,14 +201,34 @@ static void gfs2_glock_hold(struct gfs2_glock *gl) static void gfs2_glock_schedule_for_reclaim(struct gfs2_glock *gl) { + int may_reclaim; + may_reclaim = (demote_ok(gl) && + (atomic_read(&gl->gl_ref) == 1 || + (gl->gl_name.ln_type == LM_TYPE_INODE && + atomic_read(&gl->gl_ref) <= 2))); spin_lock(&lru_lock); - if (list_empty(&gl->gl_lru) && gl->gl_state != LM_ST_UNLOCKED) { + if (list_empty(&gl->gl_lru) && may_reclaim) { list_add_tail(&gl->gl_lru, &lru_list); atomic_inc(&lru_count); } spin_unlock(&lru_lock); } +/** + * gfs2_glock_put_nolock() - Decrement reference count on glock + * @gl: The glock to put + * + * This function should only be used if the caller has its own reference + * to the glock, in addition to the one it is dropping. + */ + +static void gfs2_glock_put_nolock(struct gfs2_glock *gl) +{ + if (atomic_dec_and_test(&gl->gl_ref)) + GLOCK_BUG_ON(gl, 1); + gfs2_glock_schedule_for_reclaim(gl); +} + /** * gfs2_glock_put() - Decrement reference count on glock * @gl: The glock to put @@ -214,9 +254,9 @@ int gfs2_glock_put(struct gfs2_glock *gl) rv = 1; goto out; } - /* 1 for being hashed, 1 for having state != LM_ST_UNLOCKED */ - if (atomic_read(&gl->gl_ref) == 2) - gfs2_glock_schedule_for_reclaim(gl); + spin_lock(&gl->gl_spin); + gfs2_glock_schedule_for_reclaim(gl); + spin_unlock(&gl->gl_spin); write_unlock(gl_lock_addr(gl->gl_hash)); out: return rv; @@ -398,7 +438,7 @@ static void state_change(struct gfs2_glock *gl, unsigned int new_state) if (held2) gfs2_glock_hold(gl); else - gfs2_glock_put(gl); + gfs2_glock_put_nolock(gl); } gl->gl_state = new_state; @@ -633,7 +673,7 @@ out: out_sched: gfs2_glock_hold(gl); if (queue_delayed_work(glock_workqueue, &gl->gl_work, 0) == 0) - gfs2_glock_put(gl); + gfs2_glock_put_nolock(gl); out_unlock: clear_bit(GLF_LOCK, &gl->gl_flags); goto out; @@ -1274,26 +1314,6 @@ void gfs2_glock_complete(struct gfs2_glock *gl, int ret) gfs2_glock_put(gl); } -/** - * demote_ok - Check to see if it's ok to unlock a glock - * @gl: the glock - * - * Returns: 1 if it's ok - */ - -static int demote_ok(const struct gfs2_glock *gl) -{ - const struct gfs2_glock_operations *glops = gl->gl_ops; - - if (gl->gl_state == LM_ST_UNLOCKED) - return 0; - if (!list_empty(&gl->gl_holders)) - return 0; - if (glops->go_demote_ok) - return glops->go_demote_ok(gl); - return 1; -} - static int gfs2_shrink_glock_memory(int nr, gfp_t gfp_mask) { -- cgit v1.2.3 From 6b94617024bd6810cde1d0d491202c30d5a38d91 Mon Sep 17 00:00:00 2001 From: Benjamin Marzinski Date: Fri, 10 Jul 2009 18:13:26 -0500 Subject: GFS2: Fix incorrent statfs consistency check Since both linked and unlinked inodes are counted by rgd->rd_dinodes, It makes no sense to count them with the used data blocks (first check that I changed), it makes sense to count them with the linked inodes (second check), and it makes no sense to care if there are more unlinked inodes than linked ones. This fixes these errors. Signed-off-by: Benjamin Marzinski Signed-off-by: Steven Whitehouse --- fs/gfs2/rgrp.c | 14 +++----------- 1 file changed, 3 insertions(+), 11 deletions(-) (limited to 'fs') diff --git a/fs/gfs2/rgrp.c b/fs/gfs2/rgrp.c index 5e5074176daa..fba795798d3a 100644 --- a/fs/gfs2/rgrp.c +++ b/fs/gfs2/rgrp.c @@ -285,27 +285,19 @@ void gfs2_rgrp_verify(struct gfs2_rgrpd *rgd) } tmp = rgd->rd_data - rgd->rd_free - rgd->rd_dinodes; - if (count[1] + count[2] != tmp) { + if (count[1] != tmp) { if (gfs2_consist_rgrpd(rgd)) fs_err(sdp, "used data mismatch: %u != %u\n", count[1], tmp); return; } - if (count[3] != rgd->rd_dinodes) { + if (count[2] + count[3] != rgd->rd_dinodes) { if (gfs2_consist_rgrpd(rgd)) fs_err(sdp, "used metadata mismatch: %u != %u\n", - count[3], rgd->rd_dinodes); + count[2] + count[3], rgd->rd_dinodes); return; } - - if (count[2] > count[3]) { - if (gfs2_consist_rgrpd(rgd)) - fs_err(sdp, "unlinked inodes > inodes: %u\n", - count[2]); - return; - } - } static inline int rgrp_contains_block(struct gfs2_rgrpd *rgd, u64 block) -- cgit v1.2.3 From b94a170e96dc416828af9d350ae2e34b70ae7347 Mon Sep 17 00:00:00 2001 From: Benjamin Marzinski Date: Thu, 23 Jul 2009 18:52:34 -0500 Subject: GFS2: remove dcache entries for remote deleted inodes When a file is deleted from a gfs2 filesystem on one node, a dcache entry for it may still exist on other nodes in the cluster. If this happens, gfs2 will be unable to free this file on disk. Because of this, it's possible to have a gfs2 filesystem with no files on it and no free space. With this patch, when a node receives a callback notifying it that the file is being deleted on another node, it schedules a new workqueue thread to remove the file's dcache entry. Signed-off-by: Benjamin Marzinski Signed-off-by: Steven Whitehouse --- fs/gfs2/glock.c | 43 ++++++++++++++++++++++++++++++++++++++----- fs/gfs2/glock.h | 3 +++ fs/gfs2/glops.c | 21 +++++++++++++++++++++ fs/gfs2/incore.h | 2 ++ fs/gfs2/super.c | 1 + 5 files changed, 65 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c index f041a89e1ab8..8b674b1f3a55 100644 --- a/fs/gfs2/glock.c +++ b/fs/gfs2/glock.c @@ -63,6 +63,7 @@ static void do_xmote(struct gfs2_glock *gl, struct gfs2_holder *gh, unsigned int static DECLARE_RWSEM(gfs2_umount_flush_sem); static struct dentry *gfs2_root; static struct workqueue_struct *glock_workqueue; +struct workqueue_struct *gfs2_delete_workqueue; static LIST_HEAD(lru_list); static atomic_t lru_count = ATOMIC_INIT(0); static DEFINE_SPINLOCK(lru_lock); @@ -167,7 +168,7 @@ static void glock_free(struct gfs2_glock *gl) * */ -static void gfs2_glock_hold(struct gfs2_glock *gl) +void gfs2_glock_hold(struct gfs2_glock *gl) { GLOCK_BUG_ON(gl, atomic_read(&gl->gl_ref) == 0); atomic_inc(&gl->gl_ref); @@ -222,7 +223,7 @@ static void gfs2_glock_schedule_for_reclaim(struct gfs2_glock *gl) * to the glock, in addition to the one it is dropping. */ -static void gfs2_glock_put_nolock(struct gfs2_glock *gl) +void gfs2_glock_put_nolock(struct gfs2_glock *gl) { if (atomic_dec_and_test(&gl->gl_ref)) GLOCK_BUG_ON(gl, 1); @@ -679,6 +680,29 @@ out_unlock: goto out; } +static void delete_work_func(struct work_struct *work) +{ + struct gfs2_glock *gl = container_of(work, struct gfs2_glock, gl_delete); + struct gfs2_sbd *sdp = gl->gl_sbd; + struct gfs2_inode *ip = NULL; + struct inode *inode; + u64 no_addr = 0; + + spin_lock(&gl->gl_spin); + ip = (struct gfs2_inode *)gl->gl_object; + if (ip) + no_addr = ip->i_no_addr; + spin_unlock(&gl->gl_spin); + if (ip) { + inode = gfs2_ilookup(sdp->sd_vfs, no_addr); + if (inode) { + d_prune_aliases(inode); + iput(inode); + } + } + gfs2_glock_put(gl); +} + static void glock_work_func(struct work_struct *work) { unsigned long delay = 0; @@ -757,6 +781,7 @@ int gfs2_glock_get(struct gfs2_sbd *sdp, u64 number, gl->gl_sbd = sdp; gl->gl_aspace = NULL; INIT_DELAYED_WORK(&gl->gl_work, glock_work_func); + INIT_WORK(&gl->gl_delete, delete_work_func); /* If this glock protects actual on-disk data or metadata blocks, create a VFS inode to manage the pages/buffers holding them. */ @@ -898,6 +923,8 @@ static void handle_callback(struct gfs2_glock *gl, unsigned int state, gl->gl_demote_state != state) { gl->gl_demote_state = LM_ST_UNLOCKED; } + if (gl->gl_ops->go_callback) + gl->gl_ops->go_callback(gl); trace_gfs2_demote_rq(gl); } @@ -1344,14 +1371,14 @@ static int gfs2_shrink_glock_memory(int nr, gfp_t gfp_mask) spin_unlock(&lru_lock); spin_lock(&gl->gl_spin); may_demote = demote_ok(gl); - spin_unlock(&gl->gl_spin); - clear_bit(GLF_LOCK, &gl->gl_flags); if (may_demote) { handle_callback(gl, LM_ST_UNLOCKED, 0); nr--; } if (queue_delayed_work(glock_workqueue, &gl->gl_work, 0) == 0) - gfs2_glock_put(gl); + gfs2_glock_put_nolock(gl); + spin_unlock(&gl->gl_spin); + clear_bit(GLF_LOCK, &gl->gl_flags); spin_lock(&lru_lock); continue; } @@ -1738,6 +1765,11 @@ int __init gfs2_glock_init(void) glock_workqueue = create_workqueue("glock_workqueue"); if (IS_ERR(glock_workqueue)) return PTR_ERR(glock_workqueue); + gfs2_delete_workqueue = create_workqueue("delete_workqueue"); + if (IS_ERR(gfs2_delete_workqueue)) { + destroy_workqueue(glock_workqueue); + return PTR_ERR(gfs2_delete_workqueue); + } register_shrinker(&glock_shrinker); @@ -1748,6 +1780,7 @@ void gfs2_glock_exit(void) { unregister_shrinker(&glock_shrinker); destroy_workqueue(glock_workqueue); + destroy_workqueue(gfs2_delete_workqueue); } static int gfs2_glock_iter_next(struct gfs2_glock_iter *gi) diff --git a/fs/gfs2/glock.h b/fs/gfs2/glock.h index a602a28f6f08..c609894ec0d0 100644 --- a/fs/gfs2/glock.h +++ b/fs/gfs2/glock.h @@ -143,6 +143,7 @@ struct lm_lockops { #define GLR_TRYFAILED 13 +extern struct workqueue_struct *gfs2_delete_workqueue; static inline struct gfs2_holder *gfs2_glock_is_locked_by_me(struct gfs2_glock *gl) { struct gfs2_holder *gh; @@ -191,6 +192,8 @@ static inline int gfs2_glock_is_blocking(struct gfs2_glock *gl) int gfs2_glock_get(struct gfs2_sbd *sdp, u64 number, const struct gfs2_glock_operations *glops, int create, struct gfs2_glock **glp); +void gfs2_glock_hold(struct gfs2_glock *gl); +void gfs2_glock_put_nolock(struct gfs2_glock *gl); int gfs2_glock_put(struct gfs2_glock *gl); void gfs2_holder_init(struct gfs2_glock *gl, unsigned int state, unsigned flags, struct gfs2_holder *gh); diff --git a/fs/gfs2/glops.c b/fs/gfs2/glops.c index d5e4ab155ca0..6985eef06c39 100644 --- a/fs/gfs2/glops.c +++ b/fs/gfs2/glops.c @@ -323,6 +323,7 @@ static void trans_go_sync(struct gfs2_glock *gl) if (gl->gl_state != LM_ST_UNLOCKED && test_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags)) { + flush_workqueue(gfs2_delete_workqueue); gfs2_meta_syncfs(sdp); gfs2_log_shutdown(sdp); } @@ -372,6 +373,25 @@ static int trans_go_demote_ok(const struct gfs2_glock *gl) return 0; } +/** + * iopen_go_callback - schedule the dcache entry for the inode to be deleted + * @gl: the glock + * + * gl_spin lock is held while calling this + */ +static void iopen_go_callback(struct gfs2_glock *gl) +{ + struct gfs2_inode *ip = (struct gfs2_inode *)gl->gl_object; + + if (gl->gl_demote_state == LM_ST_UNLOCKED && + gl->gl_state == LM_ST_SHARED && + ip && test_bit(GIF_USER, &ip->i_flags)) { + gfs2_glock_hold(gl); + if (queue_work(gfs2_delete_workqueue, &gl->gl_delete) == 0) + gfs2_glock_put_nolock(gl); + } +} + const struct gfs2_glock_operations gfs2_meta_glops = { .go_type = LM_TYPE_META, }; @@ -406,6 +426,7 @@ const struct gfs2_glock_operations gfs2_trans_glops = { const struct gfs2_glock_operations gfs2_iopen_glops = { .go_type = LM_TYPE_IOPEN, + .go_callback = iopen_go_callback, }; const struct gfs2_glock_operations gfs2_flock_glops = { diff --git a/fs/gfs2/incore.h b/fs/gfs2/incore.h index 225347fbff3c..61801ada36f0 100644 --- a/fs/gfs2/incore.h +++ b/fs/gfs2/incore.h @@ -159,6 +159,7 @@ struct gfs2_glock_operations { int (*go_lock) (struct gfs2_holder *gh); void (*go_unlock) (struct gfs2_holder *gh); int (*go_dump)(struct seq_file *seq, const struct gfs2_glock *gl); + void (*go_callback) (struct gfs2_glock *gl); const int go_type; const unsigned long go_min_hold_time; }; @@ -228,6 +229,7 @@ struct gfs2_glock { struct list_head gl_ail_list; atomic_t gl_ail_count; struct delayed_work gl_work; + struct work_struct gl_delete; }; #define GFS2_MIN_LVB_SIZE 32 /* Min size of LVB that gfs2 supports */ diff --git a/fs/gfs2/super.c b/fs/gfs2/super.c index 552e321cee5e..f522bb017973 100644 --- a/fs/gfs2/super.c +++ b/fs/gfs2/super.c @@ -691,6 +691,7 @@ static int gfs2_make_fs_ro(struct gfs2_sbd *sdp) struct gfs2_holder t_gh; int error; + flush_workqueue(gfs2_delete_workqueue); gfs2_quota_sync(sdp); gfs2_statfs_sync(sdp); -- cgit v1.2.3 From 276e680d192a67d222fcea51af37b056feffb665 Mon Sep 17 00:00:00 2001 From: Yan Zheng Date: Thu, 30 Jul 2009 09:40:40 -0400 Subject: Btrfs: preserve commit_root for async caching The async block group caching code uses the commit_root pointer to get a stable version of the extent allocation tree for scanning. This copy of the tree root isn't going to change and it significantly reduces the complexity of the scanning code. During a commit, we have a loop where we update the extent allocation tree root. We need to loop because updating the root pointer in the tree of tree roots may allocate blocks which may change the extent allocation tree. Right now the commit_root pointer is changed inside this loop. It is more correct to change the commit_root pointer only after all the looping is done. Signed-off-by: Yan Zheng Signed-off-by: Chris Mason --- fs/btrfs/ctree.h | 4 +--- fs/btrfs/disk-io.c | 2 +- fs/btrfs/extent-tree.c | 6 +++--- fs/btrfs/transaction.c | 12 +++++++++--- 4 files changed, 14 insertions(+), 10 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 17ad92c29cfd..38eeb6c49c8a 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -827,6 +827,7 @@ struct btrfs_fs_info { struct mutex drop_mutex; struct mutex volume_mutex; struct mutex tree_reloc_mutex; + struct rw_semaphore extent_commit_sem; /* * this protects the ordered operations list only while we are @@ -961,9 +962,6 @@ struct btrfs_root { /* the node lock is held while changing the node pointer */ spinlock_t node_lock; - /* taken when updating the commit root */ - struct rw_semaphore commit_root_sem; - struct extent_buffer *commit_root; struct btrfs_root *log_root; struct btrfs_root *reloc_root; diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 3a9b88759880..3cf4cfa575c8 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -907,7 +907,6 @@ static int __setup_root(u32 nodesize, u32 leafsize, u32 sectorsize, spin_lock_init(&root->inode_lock); mutex_init(&root->objectid_mutex); mutex_init(&root->log_mutex); - init_rwsem(&root->commit_root_sem); init_waitqueue_head(&root->log_writer_wait); init_waitqueue_head(&root->log_commit_wait[0]); init_waitqueue_head(&root->log_commit_wait[1]); @@ -1624,6 +1623,7 @@ struct btrfs_root *open_ctree(struct super_block *sb, mutex_init(&fs_info->cleaner_mutex); mutex_init(&fs_info->volume_mutex); mutex_init(&fs_info->tree_reloc_mutex); + init_rwsem(&fs_info->extent_commit_sem); btrfs_init_free_cluster(&fs_info->meta_alloc_cluster); btrfs_init_free_cluster(&fs_info->data_alloc_cluster); diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index fadf69a2764b..2fe21fa74913 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -267,7 +267,7 @@ static int caching_kthread(void *data) last = max_t(u64, block_group->key.objectid, BTRFS_SUPER_INFO_OFFSET); again: /* need to make sure the commit_root doesn't disappear */ - down_read(&fs_info->extent_root->commit_root_sem); + down_read(&fs_info->extent_commit_sem); /* * We don't want to deadlock with somebody trying to allocate a new @@ -304,7 +304,7 @@ again: if (need_resched()) { btrfs_release_path(fs_info->extent_root, path); - up_read(&fs_info->extent_root->commit_root_sem); + up_read(&fs_info->extent_commit_sem); cond_resched(); goto again; } @@ -345,7 +345,7 @@ next: err: btrfs_free_path(path); - up_read(&fs_info->extent_root->commit_root_sem); + up_read(&fs_info->extent_commit_sem); atomic_dec(&block_group->space_info->caching_threads); wake_up(&block_group->caching_q); diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index e51d2bc532f8..de48e4ec808c 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -42,10 +42,8 @@ static noinline void put_transaction(struct btrfs_transaction *transaction) static noinline void switch_commit_root(struct btrfs_root *root) { - down_write(&root->commit_root_sem); free_extent_buffer(root->commit_root); root->commit_root = btrfs_root_node(root); - up_write(&root->commit_root_sem); } /* @@ -466,7 +464,10 @@ static int update_cowonly_root(struct btrfs_trans_handle *trans, ret = btrfs_write_dirty_block_groups(trans, root); BUG_ON(ret); } - switch_commit_root(root); + + if (root != root->fs_info->extent_root) + switch_commit_root(root); + return 0; } @@ -499,6 +500,11 @@ static noinline int commit_cowonly_roots(struct btrfs_trans_handle *trans, update_cowonly_root(trans, root); } + + down_write(&fs_info->extent_commit_sem); + switch_commit_root(fs_info->extent_root); + up_write(&fs_info->extent_commit_sem); + return 0; } -- cgit v1.2.3 From f36f3042eae238bdaefe7c79310afe573bfc3622 Mon Sep 17 00:00:00 2001 From: Chris Mason Date: Thu, 30 Jul 2009 10:04:48 -0400 Subject: Btrfs: be more polite in the async caching threads The semaphore used by the async caching threads can prevent a transaction commit, which can make the FS appear to stall. This releases the semaphore more often when a transaction commit is in progress. Signed-off-by: Chris Mason --- fs/btrfs/extent-tree.c | 5 +++-- fs/btrfs/transaction.c | 10 ++++++++++ fs/btrfs/transaction.h | 1 + 3 files changed, 14 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 2fe21fa74913..dc84daee6bc4 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -302,10 +302,11 @@ again: else if (ret) break; - if (need_resched()) { + if (need_resched() || + btrfs_transaction_in_commit(fs_info)) { btrfs_release_path(fs_info->extent_root, path); up_read(&fs_info->extent_commit_sem); - cond_resched(); + schedule_timeout(1); goto again; } diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index de48e4ec808c..cdbb5022da52 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -857,6 +857,16 @@ static void update_super_roots(struct btrfs_root *root) super->root_level = root_item->level; } +int btrfs_transaction_in_commit(struct btrfs_fs_info *info) +{ + int ret = 0; + spin_lock(&info->new_trans_lock); + if (info->running_transaction) + ret = info->running_transaction->in_commit; + spin_unlock(&info->new_trans_lock); + return ret; +} + int btrfs_commit_transaction(struct btrfs_trans_handle *trans, struct btrfs_root *root) { diff --git a/fs/btrfs/transaction.h b/fs/btrfs/transaction.h index 961c3ee5a2e1..663c67404918 100644 --- a/fs/btrfs/transaction.h +++ b/fs/btrfs/transaction.h @@ -107,4 +107,5 @@ int btrfs_record_root_in_trans(struct btrfs_trans_handle *trans, struct btrfs_root *root); int btrfs_write_and_wait_marked_extents(struct btrfs_root *root, struct extent_io_tree *dirty_pages); +int btrfs_transaction_in_commit(struct btrfs_fs_info *info); #endif -- cgit v1.2.3 From 4bf17af0dbfe4cf20cb750e22e8e926273e7a7a4 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Tue, 14 Jul 2009 19:30:23 +0200 Subject: udf: Fix loading of VAT inode when drive wrongly reports number of recorded blocks VAT inode is located in the last block recorded block of the medium. When the drive errorneously reports number of recorded blocks, we failed to load the VAT inode and thus mount the medium. This patch makes kernel try to read VAT inode from the last block of the device if it is different from the last recorded block. Signed-off-by: Jan Kara --- fs/udf/super.c | 12 ++++++++++++ 1 file changed, 12 insertions(+) (limited to 'fs') diff --git a/fs/udf/super.c b/fs/udf/super.c index 6832135159b6..9d1b8c2e6c45 100644 --- a/fs/udf/super.c +++ b/fs/udf/super.c @@ -1087,11 +1087,23 @@ static int udf_load_vat(struct super_block *sb, int p_index, int type1_index) struct udf_inode_info *vati; uint32_t pos; struct virtualAllocationTable20 *vat20; + sector_t blocks = sb->s_bdev->bd_inode->i_size >> sb->s_blocksize_bits; /* VAT file entry is in the last recorded block */ ino.partitionReferenceNum = type1_index; ino.logicalBlockNum = sbi->s_last_block - map->s_partition_root; sbi->s_vat_inode = udf_iget(sb, &ino); + if (!sbi->s_vat_inode && + sbi->s_last_block != blocks - 1) { + printk(KERN_NOTICE "UDF-fs: Failed to read VAT inode from the" + " last recorded block (%lu), retrying with the last " + "block of the device (%lu).\n", + (unsigned long)sbi->s_last_block, + (unsigned long)blocks - 1); + ino.partitionReferenceNum = type1_index; + ino.logicalBlockNum = blocks - 1 - map->s_partition_root; + sbi->s_vat_inode = udf_iget(sb, &ino); + } if (!sbi->s_vat_inode) return 1; -- cgit v1.2.3 From dee865656f2d8b866f8ac22c60d6363b914e9f12 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Wed, 22 Jul 2009 18:12:17 +0200 Subject: quota: Silence lockdep on quota_on Commit d01730d74d2b0155da50d44555001706294014f7 didn't completely fix the problem since we still take dqio_mutex and i_mutex in the wrong order. Move taking of i_mutex further down (luckily it's needed only for updating inode flags) below where dqio_mutex is taken. Tested-by: Valdis Kletnieks Signed-off-by: Jan Kara --- fs/quota/dquot.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/quota/dquot.c b/fs/quota/dquot.c index 70f36c043d62..38f7bd559f35 100644 --- a/fs/quota/dquot.c +++ b/fs/quota/dquot.c @@ -2043,7 +2043,6 @@ static int vfs_load_quota_inode(struct inode *inode, int type, int format_id, invalidate_bdev(sb->s_bdev); } mutex_lock(&dqopt->dqonoff_mutex); - mutex_lock_nested(&inode->i_mutex, I_MUTEX_QUOTA); if (sb_has_quota_loaded(sb, type)) { error = -EBUSY; goto out_lock; @@ -2054,9 +2053,11 @@ static int vfs_load_quota_inode(struct inode *inode, int type, int format_id, * possible) Also nobody should write to the file - we use * special IO operations which ignore the immutable bit. */ down_write(&dqopt->dqptr_sem); + mutex_lock_nested(&inode->i_mutex, I_MUTEX_QUOTA); oldflags = inode->i_flags & (S_NOATIME | S_IMMUTABLE | S_NOQUOTA); inode->i_flags |= S_NOQUOTA | S_NOATIME | S_IMMUTABLE; + mutex_unlock(&inode->i_mutex); up_write(&dqopt->dqptr_sem); sb->dq_op->drop(inode); } @@ -2080,7 +2081,6 @@ static int vfs_load_quota_inode(struct inode *inode, int type, int format_id, goto out_file_init; } mutex_unlock(&dqopt->dqio_mutex); - mutex_unlock(&inode->i_mutex); spin_lock(&dq_state_lock); dqopt->flags |= dquot_state_flag(flags, type); spin_unlock(&dq_state_lock); @@ -2096,13 +2096,14 @@ out_file_init: out_lock: if (oldflags != -1) { down_write(&dqopt->dqptr_sem); + mutex_lock_nested(&inode->i_mutex, I_MUTEX_QUOTA); /* Set the flags back (in the case of accidental quotaon() * on a wrong file we don't want to mess up the flags) */ inode->i_flags &= ~(S_NOATIME | S_NOQUOTA | S_IMMUTABLE); inode->i_flags |= oldflags; + mutex_unlock(&inode->i_mutex); up_write(&dqopt->dqptr_sem); } - mutex_unlock(&inode->i_mutex); mutex_unlock(&dqopt->dqonoff_mutex); out_fmt: put_quota_format(fmt); -- cgit v1.2.3 From e9956fae7dbce6ac770e7a00436c496ddbbd3215 Mon Sep 17 00:00:00 2001 From: Tao Ma Date: Thu, 30 Jul 2009 16:07:10 +0800 Subject: ocfs2/quota: Release lock for error in ocfs2_quota_write. ocfs2_quota_write needs to release the lock if it fails to read quota block. So use "goto out" instead of "return err". Signed-off-by: Tao Ma Acked-by: Jan Kara Signed-off-by: Joel Becker --- fs/ocfs2/quota_global.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ocfs2/quota_global.c b/fs/ocfs2/quota_global.c index cc8785cf8f61..d604a6aa0a22 100644 --- a/fs/ocfs2/quota_global.c +++ b/fs/ocfs2/quota_global.c @@ -238,7 +238,7 @@ ssize_t ocfs2_quota_write(struct super_block *sb, int type, } if (err) { mlog_errno(err); - return err; + goto out; } lock_buffer(bh); if (new) -- cgit v1.2.3 From 97db39a1f6f69e906e98118392400de5217aa33a Mon Sep 17 00:00:00 2001 From: Eric Sandeen Date: Sun, 26 Jul 2009 21:52:01 -0500 Subject: xfs: reduce bmv_count in xfs_vn_fiemap commit 6321e3ed2acf3ee9643cdd403e1c88605d7944ba caused the full bmv_count's worth of getbmapx structures to get allocated; telling it to do MAXEXTNUM was a bit insane, resulting in ENOMEM every time. Chop it down to something reasonable, the number of slots in the caller's input buffer. If this is too large the caller may get ENOMEM but the reason should not be a mystery, and they can try again with something smaller. We add 1 to the value because in the normal getbmap world, bmv_count includes the header and xfs_getbmap does: nex = bmv->bmv_count - 1; if (nex <= 0) return XFS_ERROR(EINVAL); Signed-off-by: Eric Sandeen Reviewed-by: Olaf Weber Reviewed-by: Christoph Hellwig Signed-off-by: Felix Blyakher --- fs/xfs/linux-2.6/xfs_iops.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/xfs/linux-2.6/xfs_iops.c b/fs/xfs/linux-2.6/xfs_iops.c index 58973bb46038..8070b34cc287 100644 --- a/fs/xfs/linux-2.6/xfs_iops.c +++ b/fs/xfs/linux-2.6/xfs_iops.c @@ -680,8 +680,8 @@ xfs_vn_fiemap( else bm.bmv_length = BTOBB(length); - /* our formatter will tell xfs_getbmap when to stop. */ - bm.bmv_count = MAXEXTNUM; + /* We add one because in getbmap world count includes the header */ + bm.bmv_count = fieinfo->fi_extents_max + 1; bm.bmv_iflags = BMV_IF_PREALLOC; if (fieinfo->fi_flags & FIEMAP_FLAG_XATTR) bm.bmv_iflags |= BMV_IF_ATTRFORK; -- cgit v1.2.3 From c8a4051c3731b6db224482218cfd535ab9393ff8 Mon Sep 17 00:00:00 2001 From: Eric Sandeen Date: Fri, 31 Jul 2009 00:02:17 -0500 Subject: xfs: bump up nr_to_write in xfs_vm_writepage VM calculation for nr_to_write seems off. Bump it way up, this gets simple streaming writes zippy again. To be reviewed again after Jens' writeback changes. Signed-off-by: Christoph Hellwig Signed-off-by: Eric Sandeen Cc: Chris Mason Reviewed-by: Felix Blyakher Signed-off-by: Felix Blyakher --- fs/xfs/linux-2.6/xfs_aops.c | 8 ++++++++ 1 file changed, 8 insertions(+) (limited to 'fs') diff --git a/fs/xfs/linux-2.6/xfs_aops.c b/fs/xfs/linux-2.6/xfs_aops.c index 7ec89fc05b2b..aecf2519db76 100644 --- a/fs/xfs/linux-2.6/xfs_aops.c +++ b/fs/xfs/linux-2.6/xfs_aops.c @@ -1268,6 +1268,14 @@ xfs_vm_writepage( if (!page_has_buffers(page)) create_empty_buffers(page, 1 << inode->i_blkbits, 0); + + /* + * VM calculation for nr_to_write seems off. Bump it way + * up, this gets simple streaming writes zippy again. + * To be reviewed again after Jens' writeback changes. + */ + wbc->nr_to_write *= 4; + /* * Convert delayed allocate, unwritten or unmapped space * to real space and flush out to disk. -- cgit v1.2.3 From 6606bb97e146a387932efee263745b7240a11193 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Fri, 31 Jul 2009 11:03:58 -0400 Subject: Btrfs: fix btrfs_remove_from_free_space corner case Yan Zheng hit a problem where we tried to remove some free space but failed because we couldn't find the free space entry. This is because the free space was held within a bitmap that had a starting offset well before the actual offset of the free space, and there were free space extents that were in the same range as that offset, so tree_search_offset returned with NULL because we couldn't find a free space extent that had that offset. This is fixed by making sure that if we fail to find the entry, we re-search again with bitmap_only set to 1 and do an offset_to_bitmap so we can get the appropriate bitmap. A similar problem happens in btrfs_alloc_from_bitmap for the clustering code, but that is not as bad since we will just go and redo our cluster allocation. Also this adds some debugging checks to make sure that the free space we are trying to remove from the bitmap is in fact there. This can probably go away after a while, but since this code is only used by the tree-logging stuff it would be nice to run with it for a while to make sure there are no problems. Signed-off-by: Josef Bacik Signed-off-by: Chris Mason --- fs/btrfs/free-space-cache.c | 73 +++++++++++++++++++++++++++++++++++++++------ 1 file changed, 64 insertions(+), 9 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/free-space-cache.c b/fs/btrfs/free-space-cache.c index af99b78b288e..5edcee3a617f 100644 --- a/fs/btrfs/free-space-cache.c +++ b/fs/btrfs/free-space-cache.c @@ -414,11 +414,29 @@ static noinline int remove_from_bitmap(struct btrfs_block_group_cache *block_gro u64 *offset, u64 *bytes) { u64 end; + u64 search_start, search_bytes; + int ret; again: end = bitmap_info->offset + (u64)(BITS_PER_BITMAP * block_group->sectorsize) - 1; + /* + * XXX - this can go away after a few releases. + * + * since the only user of btrfs_remove_free_space is the tree logging + * stuff, and the only way to test that is under crash conditions, we + * want to have this debug stuff here just in case somethings not + * working. Search the bitmap for the space we are trying to use to + * make sure its actually there. If its not there then we need to stop + * because something has gone wrong. + */ + search_start = *offset; + search_bytes = *bytes; + ret = search_bitmap(block_group, bitmap_info, &search_start, + &search_bytes); + BUG_ON(ret < 0 || search_start != *offset); + if (*offset > bitmap_info->offset && *offset + *bytes > end) { bitmap_clear_bits(block_group, bitmap_info, *offset, end - *offset + 1); @@ -430,6 +448,7 @@ again: } if (*bytes) { + struct rb_node *next = rb_next(&bitmap_info->offset_index); if (!bitmap_info->bytes) { unlink_free_space(block_group, bitmap_info); kfree(bitmap_info->bitmap); @@ -438,16 +457,36 @@ again: recalculate_thresholds(block_group); } - bitmap_info = tree_search_offset(block_group, - offset_to_bitmap(block_group, - *offset), - 1, 0); - if (!bitmap_info) + /* + * no entry after this bitmap, but we still have bytes to + * remove, so something has gone wrong. + */ + if (!next) return -EINVAL; + bitmap_info = rb_entry(next, struct btrfs_free_space, + offset_index); + + /* + * if the next entry isn't a bitmap we need to return to let the + * extent stuff do its work. + */ if (!bitmap_info->bitmap) return -EAGAIN; + /* + * Ok the next item is a bitmap, but it may not actually hold + * the information for the rest of this free space stuff, so + * look for it, and if we don't find it return so we can try + * everything over again. + */ + search_start = *offset; + search_bytes = *bytes; + ret = search_bitmap(block_group, bitmap_info, &search_start, + &search_bytes); + if (ret < 0 || search_start != *offset) + return -EAGAIN; + goto again; } else if (!bitmap_info->bytes) { unlink_free_space(block_group, bitmap_info); @@ -644,8 +683,17 @@ int btrfs_remove_free_space(struct btrfs_block_group_cache *block_group, again: info = tree_search_offset(block_group, offset, 0, 0); if (!info) { - WARN_ON(1); - goto out_lock; + /* + * oops didn't find an extent that matched the space we wanted + * to remove, look for a bitmap instead + */ + info = tree_search_offset(block_group, + offset_to_bitmap(block_group, offset), + 1, 0); + if (!info) { + WARN_ON(1); + goto out_lock; + } } if (info->bytes < bytes && rb_next(&info->offset_index)) { @@ -957,8 +1005,15 @@ static u64 btrfs_alloc_from_bitmap(struct btrfs_block_group_cache *block_group, if (cluster->block_group != block_group) goto out; - entry = tree_search_offset(block_group, search_start, 0, 0); - + /* + * search_start is the beginning of the bitmap, but at some point it may + * be a good idea to point to the actual start of the free area in the + * bitmap, so do the offset_to_bitmap trick anyway, and set bitmap_only + * to 1 to make sure we get the bitmap entry + */ + entry = tree_search_offset(block_group, + offset_to_bitmap(block_group, search_start), + 1, 0); if (!entry || !entry->bitmap) goto out; -- cgit v1.2.3 From 013f1b12f4fc479f697acae2f31bad220162cd03 Mon Sep 17 00:00:00 2001 From: Chris Mason Date: Fri, 31 Jul 2009 14:57:55 -0400 Subject: Btrfs: make sure the async caching thread advances the key The async caching thread can end up looping forever if a given search puts it at the last key in a leaf. It will end up calling btrfs_next_leaf and then checking if it needs to politely drop the read semaphore. Most of the time this looping isn't noticed because it is able to make progress the next time around. But, during log replay, we wait on the async caching thread to finish, and the async thread is waiting on the commit, and no progress is really made. The fix used here is to copy the key out of the next leaf, that way our search lands there properly. Signed-off-by: Chris Mason --- fs/btrfs/extent-tree.c | 21 +++++++++++++++++---- 1 file changed, 17 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index dc84daee6bc4..72a2b9c28e9f 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -265,10 +265,6 @@ static int caching_kthread(void *data) atomic_inc(&block_group->space_info->caching_threads); last = max_t(u64, block_group->key.objectid, BTRFS_SUPER_INFO_OFFSET); -again: - /* need to make sure the commit_root doesn't disappear */ - down_read(&fs_info->extent_commit_sem); - /* * We don't want to deadlock with somebody trying to allocate a new * extent for the extent root while also trying to search the extent @@ -282,6 +278,10 @@ again: key.objectid = last; key.offset = 0; btrfs_set_key_type(&key, BTRFS_EXTENT_ITEM_KEY); +again: + /* need to make sure the commit_root doesn't disappear */ + down_read(&fs_info->extent_commit_sem); + ret = btrfs_search_slot(NULL, fs_info->extent_root, &key, path, 0, 0); if (ret < 0) goto err; @@ -304,6 +304,19 @@ again: if (need_resched() || btrfs_transaction_in_commit(fs_info)) { + leaf = path->nodes[0]; + + /* this shouldn't happen, but if the + * leaf is empty just move on. + */ + if (btrfs_header_nritems(leaf) == 0) + break; + /* + * we need to copy the key out so that + * we are sure the next search advances + * us forward in the btree. + */ + btrfs_item_key_to_cpu(leaf, &key, 0); btrfs_release_path(fs_info->extent_root, path); up_read(&fs_info->extent_commit_sem); schedule_timeout(1); -- cgit v1.2.3 From ab57a40827d99e2d8e59066a56b93bf6c844c916 Mon Sep 17 00:00:00 2001 From: Goldwyn Rodrigues Date: Fri, 31 Jul 2009 14:28:02 -0500 Subject: ocfs2: Remove redundant BUG_ON in __dlm_queue_ast() We BUG_ON() the same thing twice. Signed-off-by: Goldwyn Rodrigues Signed-off-by: Joel Becker --- fs/ocfs2/dlm/dlmast.c | 1 - 1 file changed, 1 deletion(-) (limited to 'fs') diff --git a/fs/ocfs2/dlm/dlmast.c b/fs/ocfs2/dlm/dlmast.c index d07ddbe4b283..81eff8e58322 100644 --- a/fs/ocfs2/dlm/dlmast.c +++ b/fs/ocfs2/dlm/dlmast.c @@ -103,7 +103,6 @@ static void __dlm_queue_ast(struct dlm_ctxt *dlm, struct dlm_lock *lock) lock->ast_pending, lock->ml.type); BUG(); } - BUG_ON(!list_empty(&lock->ast_list)); if (lock->ast_pending) mlog(0, "lock has an ast getting flushed right now\n"); -- cgit v1.2.3 From a97778457f22181e8c38c4cd7d7e528378738a98 Mon Sep 17 00:00:00 2001 From: Ryusuke Konishi Date: Tue, 28 Jul 2009 17:55:29 +0900 Subject: nilfs2: fix oops due to inconsistent state in page with discrete b-tree nodes Andrea Gelmini gave me a report that a kernel oops hit on a nilfs filesystem with a 1KB block size when doing rsync. This turned out to be caused by an inconsistency of dirty state between a page and its buffers storing b-tree node blocks. If the page had multiple buffers split over multiple logs, and if the logs were written at a time, a dirty flag remained in the page even every dirty flag in the buffers was cleared. This will fix the failure by dropping the dirty flag properly for pages with the discrete multiple b-tree nodes. Reported-by: Andrea Gelmini Signed-off-by: Ryusuke Konishi Tested-by: Andrea Gelmini Cc: stable@kernel.org --- fs/nilfs2/segment.c | 16 +++++++++++++++- 1 file changed, 15 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nilfs2/segment.c b/fs/nilfs2/segment.c index 8b5e4778cf28..51ff3d0a4ee2 100644 --- a/fs/nilfs2/segment.c +++ b/fs/nilfs2/segment.c @@ -1859,12 +1859,26 @@ static void nilfs_end_page_io(struct page *page, int err) if (!page) return; - if (buffer_nilfs_node(page_buffers(page)) && !PageWriteback(page)) + if (buffer_nilfs_node(page_buffers(page)) && !PageWriteback(page)) { /* * For b-tree node pages, this function may be called twice * or more because they might be split in a segment. */ + if (PageDirty(page)) { + /* + * For pages holding split b-tree node buffers, dirty + * flag on the buffers may be cleared discretely. + * In that case, the page is once redirtied for + * remaining buffers, and it must be cancelled if + * all the buffers get cleaned later. + */ + lock_page(page); + if (nilfs_page_buffers_clean(page)) + __nilfs_clear_page_dirty(page); + unlock_page(page); + } return; + } __nilfs_end_page_io(page, err); } -- cgit v1.2.3 From 9b9d6b2434fe942895c341b9a982f158529788ec Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Fri, 31 Jul 2009 06:56:09 -0400 Subject: cifs: reinstate original behavior when uid=/gid= options are specified This patch fixes the regression reported here: http://bugzilla.kernel.org/show_bug.cgi?id=13861 commit 4ae1507f6d266d0cc3dd36e474d83aad70fec9e4 changed the default behavior when the uid= or gid= option was specified for a mount. The existing behavior was to always clobber the ownership information provided by the server when these options were specified. The above commit changed this behavior so that these options simply provided defaults when the server did not provide this information (unless "forceuid" or "forcegid" were specified) This patch reverts this change so that the default behavior is restored. It also adds "noforceuid" and "noforcegid" options to make it so that ownership information from the server is preserved, even when the mount has uid= or gid= options specified. It also adds a couple of printk notices that pop up when forceuid or forcegid options are specified without a uid= or gid= option. Reported-by: Tom Chiverton Reviewed-by: Shirish Pargaonkar Signed-off-by: Jeff Layton Signed-off-by: Steve French --- fs/cifs/connect.c | 42 ++++++++++++++++++++++++++++++------------ 1 file changed, 30 insertions(+), 12 deletions(-) (limited to 'fs') diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index f2486889b7bb..1f3345d7fa79 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -803,6 +803,10 @@ cifs_parse_mount_options(char *options, const char *devname, char *data; unsigned int temp_len, i, j; char separator[2]; + short int override_uid = -1; + short int override_gid = -1; + bool uid_specified = false; + bool gid_specified = false; separator[0] = ','; separator[1] = 0; @@ -1093,18 +1097,20 @@ cifs_parse_mount_options(char *options, const char *devname, "too long.\n"); return 1; } - } else if (strnicmp(data, "uid", 3) == 0) { - if (value && *value) - vol->linux_uid = - simple_strtoul(value, &value, 0); - } else if (strnicmp(data, "forceuid", 8) == 0) { - vol->override_uid = 1; - } else if (strnicmp(data, "gid", 3) == 0) { - if (value && *value) - vol->linux_gid = - simple_strtoul(value, &value, 0); - } else if (strnicmp(data, "forcegid", 8) == 0) { - vol->override_gid = 1; + } else if (!strnicmp(data, "uid", 3) && value && *value) { + vol->linux_uid = simple_strtoul(value, &value, 0); + uid_specified = true; + } else if (!strnicmp(data, "forceuid", 8)) { + override_uid = 1; + } else if (!strnicmp(data, "noforceuid", 10)) { + override_uid = 0; + } else if (!strnicmp(data, "gid", 3) && value && *value) { + vol->linux_gid = simple_strtoul(value, &value, 0); + gid_specified = true; + } else if (!strnicmp(data, "forcegid", 8)) { + override_gid = 1; + } else if (!strnicmp(data, "noforcegid", 10)) { + override_gid = 0; } else if (strnicmp(data, "file_mode", 4) == 0) { if (value && *value) { vol->file_mode = @@ -1355,6 +1361,18 @@ cifs_parse_mount_options(char *options, const char *devname, if (vol->UNCip == NULL) vol->UNCip = &vol->UNC[2]; + if (uid_specified) + vol->override_uid = override_uid; + else if (override_uid == 1) + printk(KERN_NOTICE "CIFS: ignoring forceuid mount option " + "specified with no uid= option.\n"); + + if (gid_specified) + vol->override_gid = override_gid; + else if (override_gid == 1) + printk(KERN_NOTICE "CIFS: ignoring forcegid mount option " + "specified with no gid= option.\n"); + return 0; } -- cgit v1.2.3 From 01a261e09a21e0ba342d3907a79cf5c78ee3f37a Mon Sep 17 00:00:00 2001 From: Ryusuke Konishi Date: Sun, 2 Aug 2009 17:45:55 +0900 Subject: nilfs2: fix missing unlock in error path of nilfs_mdt_write_page This adds a missing unlock of nilfs->ns_writer_mutex in nilfs_mdt_write_page() function. Signed-off-by: Ryusuke Konishi --- fs/nilfs2/mdt.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nilfs2/mdt.c b/fs/nilfs2/mdt.c index 3d3ddb3f5177..2dfd47714ae5 100644 --- a/fs/nilfs2/mdt.c +++ b/fs/nilfs2/mdt.c @@ -412,8 +412,10 @@ nilfs_mdt_write_page(struct page *page, struct writeback_control *wbc) return 0; /* Do not request flush for shadow page cache */ if (!sb) { writer = nilfs_get_writer(NILFS_MDT(inode)->mi_nilfs); - if (!writer) + if (!writer) { + nilfs_put_writer(NILFS_MDT(inode)->mi_nilfs); return -EROFS; + } sb = writer->s_super; } -- cgit v1.2.3 From 4486d6ede16b362f89b29845af6fe1a26ae78a54 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Mon, 3 Aug 2009 12:45:10 -0400 Subject: cifs: show noforceuid/noforcegid mount options (try #2) Since forceuid is the default, we now need to show when it's disabled. Signed-off-by: Jeff Layton Signed-off-by: Steve French --- fs/cifs/cifsfs.c | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'fs') diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c index 44f30504b82d..84b75253b05a 100644 --- a/fs/cifs/cifsfs.c +++ b/fs/cifs/cifsfs.c @@ -376,10 +376,14 @@ cifs_show_options(struct seq_file *s, struct vfsmount *m) seq_printf(s, ",uid=%d", cifs_sb->mnt_uid); if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_OVERR_UID) seq_printf(s, ",forceuid"); + else + seq_printf(s, ",noforceuid"); seq_printf(s, ",gid=%d", cifs_sb->mnt_gid); if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_OVERR_GID) seq_printf(s, ",forcegid"); + else + seq_printf(s, ",noforcegid"); cifs_show_address(s, tcon->ses->server); -- cgit v1.2.3 From 24e2fb615fd6b624c320cec9ea9d91a75dad902e Mon Sep 17 00:00:00 2001 From: Roel Kluin Date: Sun, 2 Aug 2009 13:00:18 +0200 Subject: cifs: Read buffer overflow Check whether index is within bounds before testing the element. Acked-by: Jeff Layton Signed-off-by: Roel Kluin Signed-off-by: Steve French --- fs/cifs/cifs_unicode.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/cifs/cifs_unicode.c b/fs/cifs/cifs_unicode.c index 60e3c4253de0..714a542cbafc 100644 --- a/fs/cifs/cifs_unicode.c +++ b/fs/cifs/cifs_unicode.c @@ -44,7 +44,7 @@ cifs_ucs2_bytes(const __le16 *from, int maxbytes, int maxwords = maxbytes / 2; char tmp[NLS_MAX_CHARSET_SIZE]; - for (i = 0; from[i] && i < maxwords; i++) { + for (i = 0; i < maxwords && from[i]; i++) { charlen = codepage->uni2char(le16_to_cpu(from[i]), tmp, NLS_MAX_CHARSET_SIZE); if (charlen > 0) -- cgit v1.2.3 From d098564f3b2b5d555e51bca765a6a9e0dda8f2cd Mon Sep 17 00:00:00 2001 From: Steve French Date: Tue, 4 Aug 2009 03:53:28 +0000 Subject: [CIFS] Update readme to reflect forceuid mount parms Signed-off-by: Steve French --- fs/cifs/CHANGES | 3 ++- fs/cifs/README | 25 ++++++++++++------------- 2 files changed, 14 insertions(+), 14 deletions(-) (limited to 'fs') diff --git a/fs/cifs/CHANGES b/fs/cifs/CHANGES index 651cefde385b..e85b1e4389e0 100644 --- a/fs/cifs/CHANGES +++ b/fs/cifs/CHANGES @@ -2,7 +2,8 @@ Version 1.60 ------------- Fix memory leak in reconnect. Fix oops in DFS mount error path. Set s_maxbytes to smaller (the max that vfs can handle) so that -sendfile will now work over cifs mounts again. +sendfile will now work over cifs mounts again. Add noforcegid +and noforceuid mount parameters. Version 1.59 ------------ diff --git a/fs/cifs/README b/fs/cifs/README index ad92921dbde4..79c1a93400be 100644 --- a/fs/cifs/README +++ b/fs/cifs/README @@ -262,11 +262,11 @@ A partial list of the supported mount options follows: mount. domain Set the SMB/CIFS workgroup name prepended to the username during CIFS session establishment - forceuid Set the default uid for inodes based on the uid - passed in. For mounts to servers + forceuid Set the default uid for inodes to the uid + passed in on mount. For mounts to servers which do support the CIFS Unix extensions, such as a properly configured Samba server, the server provides - the uid, gid and mode so this parameter should not be + the uid, gid and mode so this parameter should not be specified unless the server and clients uid and gid numbering differ. If the server and client are in the same domain (e.g. running winbind or nss_ldap) and @@ -278,11 +278,7 @@ A partial list of the supported mount options follows: of existing files will be the uid (gid) of the person who executed the mount (root, except when mount.cifs is configured setuid for user mounts) unless the "uid=" - (gid) mount option is specified. For the uid (gid) of newly - created files and directories, ie files created since - the last mount of the server share, the expected uid - (gid) is cached as long as the inode remains in - memory on the client. Also note that permission + (gid) mount option is specified. Also note that permission checks (authorization checks) on accesses to a file occur at the server, but there are cases in which an administrator may want to restrict at the client as well. For those @@ -290,12 +286,15 @@ A partial list of the supported mount options follows: (such as Windows), permissions can also be checked at the client, and a crude form of client side permission checking can be enabled by specifying file_mode and dir_mode on - the client. Note that the mount.cifs helper must be - at version 1.10 or higher to support specifying the uid - (or gid) in non-numeric form. - forcegid (similar to above but for the groupid instead of uid) + the client. (default) + forcegid (similar to above but for the groupid instead of uid) (default) + noforceuid Fill in file owner information (uid) by requesting it from + the server if possible. With this option, the value given in + the uid= option (on mount) will only be used if the server + can not support returning uids on inodes. + noforcegid (similar to above but for the group owner, gid, instead of uid) uid Set the default uid for inodes, and indicate to the - cifs kernel driver which local user mounted . If the server + cifs kernel driver which local user mounted. If the server supports the unix extensions the default uid is not used to fill in the owner fields of inodes (files) unless the "forceuid" parameter is specified. -- cgit v1.2.3 From 57ca7deb062abf56168d15f000c16e25f88a9cf3 Mon Sep 17 00:00:00 2001 From: Anders Grafström Date: Tue, 4 Aug 2009 13:11:47 +0200 Subject: jffs2: Fix return value from jffs2_do_readpage_nolock() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This fixes "kernel BUG at fs/jffs2/file.c:251!". This pseudocode hopefully illustrates the scenario that triggers it: jffs2_write_begin { jffs2_do_readpage_nolock { jffs2_read_inode_range { jffs2_read_dnode { Data CRC 33c102e9 != calculated CRC 0ef77e7b for node at 005d42e4 return -EIO; } } ClearPageUptodate(pg); return 0; } } jffs2_write_end { BUG_ON(!PageUptodate(pg)); } Signed-off-by: Anders Grafström Signed-off-by: David Woodhouse --- fs/jffs2/file.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/jffs2/file.c b/fs/jffs2/file.c index 5edc2bf20581..23c947539864 100644 --- a/fs/jffs2/file.c +++ b/fs/jffs2/file.c @@ -99,7 +99,7 @@ static int jffs2_do_readpage_nolock (struct inode *inode, struct page *pg) kunmap(pg); D2(printk(KERN_DEBUG "readpage finished\n")); - return 0; + return ret; } int jffs2_do_readpage_unlock(struct inode *inode, struct page *pg) -- cgit v1.2.3 From 54e346215e4fe2ca8c94c54e546cc61902060510 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Fri, 7 Aug 2009 14:38:25 -0300 Subject: vfs: fix inode_init_always calling convention Currently inode_init_always calls into ->destroy_inode if the additional initialization fails. That's not only counter-intuitive because inode_init_always did not allocate the inode structure, but in case of XFS it's actively harmful as ->destroy_inode might delete the inode from a radix-tree that has never been added. This in turn might end up deleting the inode for the same inum that has been instanciated by another process and cause lots of cause subtile problems. Also in the case of re-initializing a reclaimable inode in XFS it would free an inode we still want to keep alive. Signed-off-by: Christoph Hellwig Reviewed-by: Eric Sandeen --- fs/inode.c | 30 +++++++++++++++++------------- fs/xfs/xfs_iget.c | 17 +++++------------ include/linux/fs.h | 2 +- 3 files changed, 23 insertions(+), 26 deletions(-) (limited to 'fs') diff --git a/fs/inode.c b/fs/inode.c index 901bad1e5f12..af2c05235cc8 100644 --- a/fs/inode.c +++ b/fs/inode.c @@ -120,12 +120,11 @@ static void wake_up_inode(struct inode *inode) * These are initializations that need to be done on every inode * allocation as the fields are not initialised by slab allocation. */ -struct inode *inode_init_always(struct super_block *sb, struct inode *inode) +int inode_init_always(struct super_block *sb, struct inode *inode) { static const struct address_space_operations empty_aops; static struct inode_operations empty_iops; static const struct file_operations empty_fops; - struct address_space *const mapping = &inode->i_data; inode->i_sb = sb; @@ -152,7 +151,7 @@ struct inode *inode_init_always(struct super_block *sb, struct inode *inode) inode->dirtied_when = 0; if (security_inode_alloc(inode)) - goto out_free_inode; + goto out; /* allocate and initialize an i_integrity */ if (ima_inode_alloc(inode)) @@ -198,16 +197,12 @@ struct inode *inode_init_always(struct super_block *sb, struct inode *inode) inode->i_fsnotify_mask = 0; #endif - return inode; + return 0; out_free_security: security_inode_free(inode); -out_free_inode: - if (inode->i_sb->s_op->destroy_inode) - inode->i_sb->s_op->destroy_inode(inode); - else - kmem_cache_free(inode_cachep, (inode)); - return NULL; +out: + return -ENOMEM; } EXPORT_SYMBOL(inode_init_always); @@ -220,9 +215,18 @@ static struct inode *alloc_inode(struct super_block *sb) else inode = kmem_cache_alloc(inode_cachep, GFP_KERNEL); - if (inode) - return inode_init_always(sb, inode); - return NULL; + if (!inode) + return NULL; + + if (unlikely(inode_init_always(sb, inode))) { + if (inode->i_sb->s_op->destroy_inode) + inode->i_sb->s_op->destroy_inode(inode); + else + kmem_cache_free(inode_cachep, inode); + return NULL; + } + + return inode; } void destroy_inode(struct inode *inode) diff --git a/fs/xfs/xfs_iget.c b/fs/xfs/xfs_iget.c index 5fcec6f020a7..719c85b155f4 100644 --- a/fs/xfs/xfs_iget.c +++ b/fs/xfs/xfs_iget.c @@ -64,6 +64,10 @@ xfs_inode_alloc( ip = kmem_zone_alloc(xfs_inode_zone, KM_SLEEP); if (!ip) return NULL; + if (inode_init_always(mp->m_super, VFS_I(ip))) { + kmem_zone_free(xfs_inode_zone, ip); + return NULL; + } ASSERT(atomic_read(&ip->i_iocount) == 0); ASSERT(atomic_read(&ip->i_pincount) == 0); @@ -105,17 +109,6 @@ xfs_inode_alloc( #ifdef XFS_DIR2_TRACE ip->i_dir_trace = ktrace_alloc(XFS_DIR2_KTRACE_SIZE, KM_NOFS); #endif - /* - * Now initialise the VFS inode. We do this after the xfs_inode - * initialisation as internal failures will result in ->destroy_inode - * being called and that will pass down through the reclaim path and - * free the XFS inode. This path requires the XFS inode to already be - * initialised. Hence if this call fails, the xfs_inode has already - * been freed and we should not reference it at all in the error - * handling. - */ - if (!inode_init_always(mp->m_super, VFS_I(ip))) - return NULL; /* prevent anyone from using this yet */ VFS_I(ip)->i_state = I_NEW|I_LOCK; @@ -167,7 +160,7 @@ xfs_iget_cache_hit( * errors cleanly, then tag it so it can be set up correctly * later. */ - if (!inode_init_always(mp->m_super, VFS_I(ip))) { + if (inode_init_always(mp->m_super, VFS_I(ip))) { error = ENOMEM; goto out_error; } diff --git a/include/linux/fs.h b/include/linux/fs.h index a36ffa5a77a4..0c3b5e58a986 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -2137,7 +2137,7 @@ extern loff_t default_llseek(struct file *file, loff_t offset, int origin); extern loff_t vfs_llseek(struct file *file, loff_t offset, int origin); -extern struct inode * inode_init_always(struct super_block *, struct inode *); +extern int inode_init_always(struct super_block *, struct inode *); extern void inode_init_once(struct inode *); extern void inode_add_to_lists(struct super_block *, struct inode *); extern void iput(struct inode *); -- cgit v1.2.3 From 2e00c97e2c1d2ffc9e26252ca26b237678b0b772 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Fri, 7 Aug 2009 14:38:29 -0300 Subject: vfs: add __destroy_inode When we want to tear down an inode that lost the add to the cache race in XFS we must not call into ->destroy_inode because that would delete the inode that won the race from the inode cache radix tree. This patch provides the __destroy_inode helper needed to fix this, the actual fix will be in th next patch. As XFS was the only reason destroy_inode was exported we shift the export to the new __destroy_inode. Signed-off-by: Christoph Hellwig Reviewed-by: Eric Sandeen --- fs/inode.c | 10 +++++++--- include/linux/fs.h | 1 + 2 files changed, 8 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/inode.c b/fs/inode.c index af2c05235cc8..ae7b67e48661 100644 --- a/fs/inode.c +++ b/fs/inode.c @@ -229,7 +229,7 @@ static struct inode *alloc_inode(struct super_block *sb) return inode; } -void destroy_inode(struct inode *inode) +void __destroy_inode(struct inode *inode) { BUG_ON(inode_has_buffers(inode)); ima_inode_free(inode); @@ -241,13 +241,17 @@ void destroy_inode(struct inode *inode) if (inode->i_default_acl && inode->i_default_acl != ACL_NOT_CACHED) posix_acl_release(inode->i_default_acl); #endif +} +EXPORT_SYMBOL(__destroy_inode); + +void destroy_inode(struct inode *inode) +{ + __destroy_inode(inode); if (inode->i_sb->s_op->destroy_inode) inode->i_sb->s_op->destroy_inode(inode); else kmem_cache_free(inode_cachep, (inode)); } -EXPORT_SYMBOL(destroy_inode); - /* * These are initializations that only need to be done diff --git a/include/linux/fs.h b/include/linux/fs.h index 0c3b5e58a986..67888a9e0655 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -2164,6 +2164,7 @@ extern void __iget(struct inode * inode); extern void iget_failed(struct inode *); extern void clear_inode(struct inode *); extern void destroy_inode(struct inode *); +extern void __destroy_inode(struct inode *); extern struct inode *new_inode(struct super_block *); extern int should_remove_suid(struct dentry *); extern int file_remove_suid(struct file *); -- cgit v1.2.3 From b36ec0428a06fcbdb67d61e9e664154e5dd9a8c7 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Fri, 7 Aug 2009 14:38:34 -0300 Subject: xfs: fix freeing of inodes not yet added to the inode cache When freeing an inode that lost race getting added to the inode cache we must not call into ->destroy_inode, because that would delete the inode that won the race from the inode cache radix tree. This patch uses splits a new xfs_inode_free helper out of xfs_ireclaim and uses that plus __destroy_inode to make sure we really only free the memory allocted for the inode that lost the race, and not mess with the inode cache state. Signed-off-by: Christoph Hellwig Reviewed-by: Eric Sandeen Reported-by: Alex Samad Reported-by: Andrew Randrianasulu Reported-by: Stephane Reported-by: Tommy Reported-by: Miah Gregory Reported-by: Gabriel Barazer Reported-by: Leandro Lucarella Reported-by: Daniel Burr Reported-by: Nickolay Reported-by: Michael Guntsche Reported-by: Dan Carley Reported-by: Michael Ole Olsen Reported-by: Michael Weissenbacher Reported-by: Martin Spott Reported-by: Christian Kujau Tested-by: Michael Guntsche Tested-by: Dan Carley Tested-by: Christian Kujau --- fs/xfs/xfs_iget.c | 125 +++++++++++++++++++++++++++++------------------------ fs/xfs/xfs_inode.h | 17 -------- 2 files changed, 68 insertions(+), 74 deletions(-) (limited to 'fs') diff --git a/fs/xfs/xfs_iget.c b/fs/xfs/xfs_iget.c index 719c85b155f4..34ec86923f7e 100644 --- a/fs/xfs/xfs_iget.c +++ b/fs/xfs/xfs_iget.c @@ -116,6 +116,71 @@ xfs_inode_alloc( return ip; } +STATIC void +xfs_inode_free( + struct xfs_inode *ip) +{ + switch (ip->i_d.di_mode & S_IFMT) { + case S_IFREG: + case S_IFDIR: + case S_IFLNK: + xfs_idestroy_fork(ip, XFS_DATA_FORK); + break; + } + + if (ip->i_afp) + xfs_idestroy_fork(ip, XFS_ATTR_FORK); + +#ifdef XFS_INODE_TRACE + ktrace_free(ip->i_trace); +#endif +#ifdef XFS_BMAP_TRACE + ktrace_free(ip->i_xtrace); +#endif +#ifdef XFS_BTREE_TRACE + ktrace_free(ip->i_btrace); +#endif +#ifdef XFS_RW_TRACE + ktrace_free(ip->i_rwtrace); +#endif +#ifdef XFS_ILOCK_TRACE + ktrace_free(ip->i_lock_trace); +#endif +#ifdef XFS_DIR2_TRACE + ktrace_free(ip->i_dir_trace); +#endif + + if (ip->i_itemp) { + /* + * Only if we are shutting down the fs will we see an + * inode still in the AIL. If it is there, we should remove + * it to prevent a use-after-free from occurring. + */ + xfs_log_item_t *lip = &ip->i_itemp->ili_item; + struct xfs_ail *ailp = lip->li_ailp; + + ASSERT(((lip->li_flags & XFS_LI_IN_AIL) == 0) || + XFS_FORCED_SHUTDOWN(ip->i_mount)); + if (lip->li_flags & XFS_LI_IN_AIL) { + spin_lock(&ailp->xa_lock); + if (lip->li_flags & XFS_LI_IN_AIL) + xfs_trans_ail_delete(ailp, lip); + else + spin_unlock(&ailp->xa_lock); + } + xfs_inode_item_destroy(ip); + ip->i_itemp = NULL; + } + + /* asserts to verify all state is correct here */ + ASSERT(atomic_read(&ip->i_iocount) == 0); + ASSERT(atomic_read(&ip->i_pincount) == 0); + ASSERT(!spin_is_locked(&ip->i_flags_lock)); + ASSERT(completion_done(&ip->i_flush)); + + kmem_zone_free(xfs_inode_zone, ip); +} + /* * Check the validity of the inode we just found it the cache */ @@ -292,7 +357,8 @@ out_preload_end: if (lock_flags) xfs_iunlock(ip, lock_flags); out_destroy: - xfs_destroy_inode(ip); + __destroy_inode(VFS_I(ip)); + xfs_inode_free(ip); return error; } @@ -497,62 +563,7 @@ xfs_ireclaim( xfs_qm_dqdetach(ip); xfs_iunlock(ip, XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL); - switch (ip->i_d.di_mode & S_IFMT) { - case S_IFREG: - case S_IFDIR: - case S_IFLNK: - xfs_idestroy_fork(ip, XFS_DATA_FORK); - break; - } - - if (ip->i_afp) - xfs_idestroy_fork(ip, XFS_ATTR_FORK); - -#ifdef XFS_INODE_TRACE - ktrace_free(ip->i_trace); -#endif -#ifdef XFS_BMAP_TRACE - ktrace_free(ip->i_xtrace); -#endif -#ifdef XFS_BTREE_TRACE - ktrace_free(ip->i_btrace); -#endif -#ifdef XFS_RW_TRACE - ktrace_free(ip->i_rwtrace); -#endif -#ifdef XFS_ILOCK_TRACE - ktrace_free(ip->i_lock_trace); -#endif -#ifdef XFS_DIR2_TRACE - ktrace_free(ip->i_dir_trace); -#endif - if (ip->i_itemp) { - /* - * Only if we are shutting down the fs will we see an - * inode still in the AIL. If it is there, we should remove - * it to prevent a use-after-free from occurring. - */ - xfs_log_item_t *lip = &ip->i_itemp->ili_item; - struct xfs_ail *ailp = lip->li_ailp; - - ASSERT(((lip->li_flags & XFS_LI_IN_AIL) == 0) || - XFS_FORCED_SHUTDOWN(ip->i_mount)); - if (lip->li_flags & XFS_LI_IN_AIL) { - spin_lock(&ailp->xa_lock); - if (lip->li_flags & XFS_LI_IN_AIL) - xfs_trans_ail_delete(ailp, lip); - else - spin_unlock(&ailp->xa_lock); - } - xfs_inode_item_destroy(ip); - ip->i_itemp = NULL; - } - /* asserts to verify all state is correct here */ - ASSERT(atomic_read(&ip->i_iocount) == 0); - ASSERT(atomic_read(&ip->i_pincount) == 0); - ASSERT(!spin_is_locked(&ip->i_flags_lock)); - ASSERT(completion_done(&ip->i_flush)); - kmem_zone_free(xfs_inode_zone, ip); + xfs_inode_free(ip); } /* diff --git a/fs/xfs/xfs_inode.h b/fs/xfs/xfs_inode.h index 1804f866a71d..65f24a3cc992 100644 --- a/fs/xfs/xfs_inode.h +++ b/fs/xfs/xfs_inode.h @@ -309,23 +309,6 @@ static inline struct inode *VFS_I(struct xfs_inode *ip) return &ip->i_vnode; } -/* - * Get rid of a partially initialized inode. - * - * We have to go through destroy_inode to make sure allocations - * from init_inode_always like the security data are undone. - * - * We mark the inode bad so that it takes the short cut in - * the reclaim path instead of going through the flush path - * which doesn't make sense for an inode that has never seen the - * light of day. - */ -static inline void xfs_destroy_inode(struct xfs_inode *ip) -{ - make_bad_inode(VFS_I(ip)); - return destroy_inode(VFS_I(ip)); -} - /* * i_flags helper functions */ -- cgit v1.2.3 From 69130c7cf96ea853dc5be599dd6a4b98907d39cc Mon Sep 17 00:00:00 2001 From: Eric Sandeen Date: Thu, 6 Aug 2009 15:07:37 -0700 Subject: compat_ioctl: hook up compat handler for FIEMAP ioctl The FIEMAP_IOC_FIEMAP mapping ioctl was missing a 32-bit compat handler, which means that 32-bit suerspace on 64-bit kernels cannot use this ioctl command. The structure is nicely aligned, padded, and sized, so it is just this simple. Tested w/ 32-bit ioctl tester (from Josef) on a 64-bit kernel on ext4. Signed-off-by: Eric Sandeen Cc: Cc: Mark Lord Cc: Arnd Bergmann Cc: Josef Bacik Cc: Jan Kara Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/compat_ioctl.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/compat_ioctl.c b/fs/compat_ioctl.c index f28f070a60fc..f91fd51b32e3 100644 --- a/fs/compat_ioctl.c +++ b/fs/compat_ioctl.c @@ -1905,6 +1905,7 @@ COMPATIBLE_IOCTL(FIONCLEX) COMPATIBLE_IOCTL(FIOASYNC) COMPATIBLE_IOCTL(FIONBIO) COMPATIBLE_IOCTL(FIONREAD) /* This is also TIOCINQ */ +COMPATIBLE_IOCTL(FS_IOC_FIEMAP) /* 0x00 */ COMPATIBLE_IOCTL(FIBMAP) COMPATIBLE_IOCTL(FIGETBSZ) -- cgit v1.2.3 From 2d8dd38a5aa0cc2490bbad9b75e77fa154e1d145 Mon Sep 17 00:00:00 2001 From: OGAWA Hirofumi Date: Thu, 6 Aug 2009 15:07:39 -0700 Subject: vfs: mnt_want_write_file(): fix special file handling I suspect that mnt_want_write_file() may have wrong assumption. I think mnt_want_write_file() is assuming it increments ->mnt_writers if (file->f_mode & FMODE_WRITE). But, if it's special_file(), it is false? Signed-off-by: OGAWA Hirofumi Acked-by: Dave Hansen Cc: Al Viro Cc: Nick Piggin Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/namespace.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/namespace.c b/fs/namespace.c index 277c28a63ead..7230787d18b0 100644 --- a/fs/namespace.c +++ b/fs/namespace.c @@ -316,7 +316,8 @@ EXPORT_SYMBOL_GPL(mnt_clone_write); */ int mnt_want_write_file(struct file *file) { - if (!(file->f_mode & FMODE_WRITE)) + struct inode *inode = file->f_dentry->d_inode; + if (!(file->f_mode & FMODE_WRITE) || special_file(inode->i_mode)) return mnt_want_write(file->f_path.mnt); else return mnt_clone_write(file->f_path.mnt); -- cgit v1.2.3 From 3440625d78711bee41a84cf29c3d8c579b522666 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Thu, 6 Aug 2009 15:09:34 -0700 Subject: flat: fix uninitialized ptr with shared libs The new credentials code broke load_flat_shared_library() as it now uses an uninitialized cred pointer. Reported-by: Bernd Schmidt Tested-by: Bernd Schmidt Cc: Mike Frysinger Cc: David Howells Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/binfmt_flat.c | 17 ++++++++++++----- 1 file changed, 12 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/binfmt_flat.c b/fs/binfmt_flat.c index 697f6b5f1313..e92f229e3c6e 100644 --- a/fs/binfmt_flat.c +++ b/fs/binfmt_flat.c @@ -828,15 +828,22 @@ static int load_flat_shared_library(int id, struct lib_info *libs) if (IS_ERR(bprm.file)) return res; + bprm.cred = prepare_exec_creds(); + res = -ENOMEM; + if (!bprm.cred) + goto out; + res = prepare_binprm(&bprm); if (res <= (unsigned long)-4096) res = load_flat_file(&bprm, libs, id, NULL); - if (bprm.file) { - allow_write_access(bprm.file); - fput(bprm.file); - bprm.file = NULL; - } + + abort_creds(bprm.cred); + +out: + allow_write_access(bprm.file); + fput(bprm.file); + return(res); } -- cgit v1.2.3 From 4baf8c9201e88546918cbfa61ea8062c38bc1644 Mon Sep 17 00:00:00 2001 From: Bartlomiej Zolnierkiewicz Date: Fri, 7 Aug 2009 13:47:08 -0400 Subject: Btrfs: remove superfluous NULL pointer check in btrfs_rename() This takes care of the following entry from Dan's list: fs/btrfs/inode.c +4788 btrfs_rename(36) warning: variable derefenced before check 'old_inode' Reported-by: Dan Carpenter Cc: Jonathan Corbet Cc: Eugene Teo Cc: Julia Lawall Signed-off-by: Bartlomiej Zolnierkiewicz Signed-off-by: Chris Mason --- fs/btrfs/inode.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 3ea827ddf0fe..04b53b5ebe59 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -4806,8 +4806,7 @@ static int btrfs_rename(struct inode *old_dir, struct dentry *old_dentry, * and the replacement file is large. Start IO on it now so * we don't add too much work to the end of the transaction */ - if (new_inode && old_inode && S_ISREG(old_inode->i_mode) && - new_inode->i_size && + if (new_inode && S_ISREG(old_inode->i_mode) && new_inode->i_size && old_inode->i_size > BTRFS_ORDERED_OPERATIONS_FLUSH_LIMIT) filemap_flush(old_inode->i_mapping); -- cgit v1.2.3 From 60f2e8f8a07331097a57ec4abcdc680405579377 Mon Sep 17 00:00:00 2001 From: Julia Lawall Date: Fri, 7 Aug 2009 13:51:33 -0400 Subject: Btrfs: correct error-handling zlib error handling find_zlib_workspace returns an ERR_PTR value in an error case instead of NULL. A simplified version of the semantic match that finds this problem is as follows: (http://coccinelle.lip6.fr/) // @match exists@ expression x, E; statement S1, S2; @@ x = find_zlib_workspace(...) ... when != x = E ( * if (x == NULL || ...) S1 else S2 | * if (x == NULL && ...) S1 else S2 ) // Signed-off-by: Julia Lawall Signed-off-by: Chris Mason --- fs/btrfs/zlib.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/zlib.c b/fs/btrfs/zlib.c index ecfbce836d32..3e2b90eaa239 100644 --- a/fs/btrfs/zlib.c +++ b/fs/btrfs/zlib.c @@ -208,7 +208,7 @@ int btrfs_zlib_compress_pages(struct address_space *mapping, *total_in = 0; workspace = find_zlib_workspace(); - if (!workspace) + if (IS_ERR(workspace)) return -1; if (Z_OK != zlib_deflateInit(&workspace->def_strm, 3)) { @@ -366,7 +366,7 @@ int btrfs_zlib_decompress_biovec(struct page **pages_in, char *kaddr; workspace = find_zlib_workspace(); - if (!workspace) + if (IS_ERR(workspace)) return -ENOMEM; data_in = kmap(pages_in[page_in_index]); @@ -547,7 +547,7 @@ int btrfs_zlib_decompress(unsigned char *data_in, return -ENOMEM; workspace = find_zlib_workspace(); - if (!workspace) + if (IS_ERR(workspace)) return -ENOMEM; workspace->inf_strm.next_in = data_in; -- cgit v1.2.3 From ceab36edd3d3ad3ffd01d41d6d1e05ac1ff8357e Mon Sep 17 00:00:00 2001 From: Yan Zheng Date: Fri, 7 Aug 2009 13:51:33 -0400 Subject: Btrfs: fix balancing oops when invalidate_inode_pages2 returns EBUSY invalidate_inode_pages2_range may return -EBUSY occasionally which results Oops. This patch fixes the issue by moving invalidate_inode_pages2_range into a loop and keeping calling it until the return value is not -EBUSY. The EBUSY return is temporary, and can happen when the btrfs release page function is unable to release a page because the EXTENT_LOCK bit is set. Signed-off-by: Yan Zheng Signed-off-by: Chris Mason --- fs/btrfs/relocation.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c index e71264d1c2c9..c04f7f212602 100644 --- a/fs/btrfs/relocation.c +++ b/fs/btrfs/relocation.c @@ -2553,8 +2553,13 @@ int relocate_inode_pages(struct inode *inode, u64 start, u64 len) last_index = (start + len - 1) >> PAGE_CACHE_SHIFT; /* make sure the dirty trick played by the caller work */ - ret = invalidate_inode_pages2_range(inode->i_mapping, - first_index, last_index); + while (1) { + ret = invalidate_inode_pages2_range(inode->i_mapping, + first_index, last_index); + if (ret != -EBUSY) + break; + schedule_timeout(HZ/10); + } if (ret) goto out_unlock; -- cgit v1.2.3 From e7432675f8ca868a4af365759a8d4c3779a3d922 Mon Sep 17 00:00:00 2001 From: Sunil Mushran Date: Thu, 6 Aug 2009 16:12:58 -0700 Subject: ocfs2: Initialize the cluster we're writing to in a non-sparse extend In a non-sparse extend, we correctly allocate (and zero) the clusters between the old_i_size and pos, but we don't zero the portions of the cluster we're writing to outside of pos<->len. It handles clustersize > pagesize and blocksize < pagesize. [Cleaned up by Joel Becker.] Signed-off-by: Sunil Mushran Signed-off-by: Joel Becker --- fs/ocfs2/aops.c | 66 ++++++++++++++++++++++++++++++++++++++++----------------- 1 file changed, 47 insertions(+), 19 deletions(-) (limited to 'fs') diff --git a/fs/ocfs2/aops.c b/fs/ocfs2/aops.c index e511df101451..b401654011a2 100644 --- a/fs/ocfs2/aops.c +++ b/fs/ocfs2/aops.c @@ -895,18 +895,17 @@ struct ocfs2_write_cluster_desc { */ unsigned c_new; unsigned c_unwritten; + unsigned c_needs_zero; }; -static inline int ocfs2_should_zero_cluster(struct ocfs2_write_cluster_desc *d) -{ - return d->c_new || d->c_unwritten; -} - struct ocfs2_write_ctxt { /* Logical cluster position / len of write */ u32 w_cpos; u32 w_clen; + /* First cluster allocated in a nonsparse extend */ + u32 w_first_new_cpos; + struct ocfs2_write_cluster_desc w_desc[OCFS2_MAX_CLUSTERS_PER_PAGE]; /* @@ -984,6 +983,7 @@ static int ocfs2_alloc_write_ctxt(struct ocfs2_write_ctxt **wcp, return -ENOMEM; wc->w_cpos = pos >> osb->s_clustersize_bits; + wc->w_first_new_cpos = UINT_MAX; cend = (pos + len - 1) >> osb->s_clustersize_bits; wc->w_clen = cend - wc->w_cpos + 1; get_bh(di_bh); @@ -1218,20 +1218,18 @@ out: */ static int ocfs2_write_cluster(struct address_space *mapping, u32 phys, unsigned int unwritten, + unsigned int should_zero, struct ocfs2_alloc_context *data_ac, struct ocfs2_alloc_context *meta_ac, struct ocfs2_write_ctxt *wc, u32 cpos, loff_t user_pos, unsigned user_len) { - int ret, i, new, should_zero = 0; + int ret, i, new; u64 v_blkno, p_blkno; struct inode *inode = mapping->host; struct ocfs2_extent_tree et; new = phys == 0 ? 1 : 0; - if (new || unwritten) - should_zero = 1; - if (new) { u32 tmp_pos; @@ -1342,7 +1340,9 @@ static int ocfs2_write_cluster_by_desc(struct address_space *mapping, local_len = osb->s_clustersize - cluster_off; ret = ocfs2_write_cluster(mapping, desc->c_phys, - desc->c_unwritten, data_ac, meta_ac, + desc->c_unwritten, + desc->c_needs_zero, + data_ac, meta_ac, wc, desc->c_cpos, pos, local_len); if (ret) { mlog_errno(ret); @@ -1392,14 +1392,14 @@ static void ocfs2_set_target_boundaries(struct ocfs2_super *osb, * newly allocated cluster. */ desc = &wc->w_desc[0]; - if (ocfs2_should_zero_cluster(desc)) + if (desc->c_needs_zero) ocfs2_figure_cluster_boundaries(osb, desc->c_cpos, &wc->w_target_from, NULL); desc = &wc->w_desc[wc->w_clen - 1]; - if (ocfs2_should_zero_cluster(desc)) + if (desc->c_needs_zero) ocfs2_figure_cluster_boundaries(osb, desc->c_cpos, NULL, @@ -1467,13 +1467,28 @@ static int ocfs2_populate_write_desc(struct inode *inode, phys++; } + /* + * If w_first_new_cpos is < UINT_MAX, we have a non-sparse + * file that got extended. w_first_new_cpos tells us + * where the newly allocated clusters are so we can + * zero them. + */ + if (desc->c_cpos >= wc->w_first_new_cpos) { + BUG_ON(phys == 0); + desc->c_needs_zero = 1; + } + desc->c_phys = phys; if (phys == 0) { desc->c_new = 1; + desc->c_needs_zero = 1; *clusters_to_alloc = *clusters_to_alloc + 1; } - if (ext_flags & OCFS2_EXT_UNWRITTEN) + + if (ext_flags & OCFS2_EXT_UNWRITTEN) { desc->c_unwritten = 1; + desc->c_needs_zero = 1; + } num_clusters--; } @@ -1633,10 +1648,13 @@ static int ocfs2_expand_nonsparse_inode(struct inode *inode, loff_t pos, if (newsize <= i_size_read(inode)) return 0; - ret = ocfs2_extend_no_holes(inode, newsize, newsize - len); + ret = ocfs2_extend_no_holes(inode, newsize, pos); if (ret) mlog_errno(ret); + wc->w_first_new_cpos = + ocfs2_clusters_for_bytes(inode->i_sb, i_size_read(inode)); + return ret; } @@ -1645,7 +1663,7 @@ int ocfs2_write_begin_nolock(struct address_space *mapping, struct page **pagep, void **fsdata, struct buffer_head *di_bh, struct page *mmap_page) { - int ret, credits = OCFS2_INODE_UPDATE_CREDITS; + int ret, cluster_of_pages, credits = OCFS2_INODE_UPDATE_CREDITS; unsigned int clusters_to_alloc, extents_to_split; struct ocfs2_write_ctxt *wc; struct inode *inode = mapping->host; @@ -1723,8 +1741,19 @@ int ocfs2_write_begin_nolock(struct address_space *mapping, } - ocfs2_set_target_boundaries(osb, wc, pos, len, - clusters_to_alloc + extents_to_split); + /* + * We have to zero sparse allocated clusters, unwritten extent clusters, + * and non-sparse clusters we just extended. For non-sparse writes, + * we know zeros will only be needed in the first and/or last cluster. + */ + if (clusters_to_alloc || extents_to_split || + wc->w_desc[0].c_needs_zero || + wc->w_desc[wc->w_clen - 1].c_needs_zero) + cluster_of_pages = 1; + else + cluster_of_pages = 0; + + ocfs2_set_target_boundaries(osb, wc, pos, len, cluster_of_pages); handle = ocfs2_start_trans(osb, credits); if (IS_ERR(handle)) { @@ -1757,8 +1786,7 @@ int ocfs2_write_begin_nolock(struct address_space *mapping, * extent. */ ret = ocfs2_grab_pages_for_write(mapping, wc, wc->w_cpos, pos, - clusters_to_alloc + extents_to_split, - mmap_page); + cluster_of_pages, mmap_page); if (ret) { mlog_errno(ret); goto out_quota; -- cgit v1.2.3 From 8a57a9dda760ea7845390f1cd36f3eb2a49391d8 Mon Sep 17 00:00:00 2001 From: Roel Kluin Date: Thu, 6 Aug 2009 16:07:50 -0700 Subject: ocfs2: keep index within status_map[] Do not exceed array status_map[] Signed-off-by: Roel Kluin Cc: Mark Fasheh Signed-off-by: Andrew Morton Signed-off-by: Joel Becker --- fs/ocfs2/stack_o2cb.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ocfs2/stack_o2cb.c b/fs/ocfs2/stack_o2cb.c index 3f661376a2de..e49c41050264 100644 --- a/fs/ocfs2/stack_o2cb.c +++ b/fs/ocfs2/stack_o2cb.c @@ -17,6 +17,7 @@ * General Public License for more details. */ +#include #include #include @@ -153,7 +154,7 @@ static int status_map[] = { static int dlm_status_to_errno(enum dlm_status status) { - BUG_ON(status > (sizeof(status_map) / sizeof(status_map[0]))); + BUG_ON(status < 0 || status >= ARRAY_SIZE(status_map)); return status_map[status]; } -- cgit v1.2.3 From 13f0feafa6b8aead57a2a328e2fca6a5828bf286 Mon Sep 17 00:00:00 2001 From: Oleg Nesterov Date: Tue, 23 Jun 2009 21:25:32 +0200 Subject: mm_for_maps: simplify, use ptrace_may_access() It would be nice to kill __ptrace_may_access(). It requires task_lock(), but this lock is only needed to read mm->flags in the middle. Convert mm_for_maps() to use ptrace_may_access(), this also simplifies the code a little bit. Also, we do not need to take ->mmap_sem in advance. In fact I think mm_for_maps() should not play with ->mmap_sem at all, the caller should take this lock. With or without this patch, without ->cred_guard_mutex held we can race with exec() and get the new ->mm but check old creds. Signed-off-by: Oleg Nesterov Reviewed-by: Serge Hallyn Signed-off-by: James Morris --- fs/proc/base.c | 23 +++++++++++------------ 1 file changed, 11 insertions(+), 12 deletions(-) (limited to 'fs') diff --git a/fs/proc/base.c b/fs/proc/base.c index 3ce5ae9e3d2d..917f338a6739 100644 --- a/fs/proc/base.c +++ b/fs/proc/base.c @@ -237,20 +237,19 @@ struct mm_struct *mm_for_maps(struct task_struct *task) struct mm_struct *mm = get_task_mm(task); if (!mm) return NULL; + if (mm != current->mm) { + /* + * task->mm can be changed before security check, + * in that case we must notice the change after. + */ + if (!ptrace_may_access(task, PTRACE_MODE_READ) || + mm != task->mm) { + mmput(mm); + return NULL; + } + } down_read(&mm->mmap_sem); - task_lock(task); - if (task->mm != mm) - goto out; - if (task->mm != current->mm && - __ptrace_may_access(task, PTRACE_MODE_READ) < 0) - goto out; - task_unlock(task); return mm; -out: - task_unlock(task); - up_read(&mm->mmap_sem); - mmput(mm); - return NULL; } static int proc_pid_cmdline(struct task_struct *task, char * buffer) -- cgit v1.2.3 From 00f89d218523b9bf6b522349c039d5ac80aa536d Mon Sep 17 00:00:00 2001 From: Oleg Nesterov Date: Fri, 10 Jul 2009 03:27:38 +0200 Subject: mm_for_maps: shift down_read(mmap_sem) to the caller mm_for_maps() takes ->mmap_sem after security checks, this looks strange and obfuscates the locking rules. Move this lock to its single caller, m_start(). Signed-off-by: Oleg Nesterov Acked-by: Serge Hallyn Signed-off-by: James Morris --- fs/proc/base.c | 8 +++----- fs/proc/task_mmu.c | 1 + fs/proc/task_nommu.c | 1 + 3 files changed, 5 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/proc/base.c b/fs/proc/base.c index 917f338a6739..f3c2e4085fed 100644 --- a/fs/proc/base.c +++ b/fs/proc/base.c @@ -235,9 +235,8 @@ static int check_mem_permission(struct task_struct *task) struct mm_struct *mm_for_maps(struct task_struct *task) { struct mm_struct *mm = get_task_mm(task); - if (!mm) - return NULL; - if (mm != current->mm) { + + if (mm && mm != current->mm) { /* * task->mm can be changed before security check, * in that case we must notice the change after. @@ -245,10 +244,9 @@ struct mm_struct *mm_for_maps(struct task_struct *task) if (!ptrace_may_access(task, PTRACE_MODE_READ) || mm != task->mm) { mmput(mm); - return NULL; + mm = NULL; } } - down_read(&mm->mmap_sem); return mm; } diff --git a/fs/proc/task_mmu.c b/fs/proc/task_mmu.c index 6f61b7cc32e0..9bd8be1d235c 100644 --- a/fs/proc/task_mmu.c +++ b/fs/proc/task_mmu.c @@ -119,6 +119,7 @@ static void *m_start(struct seq_file *m, loff_t *pos) mm = mm_for_maps(priv->task); if (!mm) return NULL; + down_read(&mm->mmap_sem); tail_vma = get_gate_vma(priv->task); priv->tail_vma = tail_vma; diff --git a/fs/proc/task_nommu.c b/fs/proc/task_nommu.c index 64a72e2e7650..8f5c05d3dbd3 100644 --- a/fs/proc/task_nommu.c +++ b/fs/proc/task_nommu.c @@ -189,6 +189,7 @@ static void *m_start(struct seq_file *m, loff_t *pos) priv->task = NULL; return NULL; } + down_read(&mm->mmap_sem); /* start from the Nth VMA */ for (p = rb_first(&mm->mm_rb); p; p = rb_next(p)) -- cgit v1.2.3 From 704b836cbf19e885f8366bccb2e4b0474346c02d Mon Sep 17 00:00:00 2001 From: Oleg Nesterov Date: Fri, 10 Jul 2009 03:27:40 +0200 Subject: mm_for_maps: take ->cred_guard_mutex to fix the race with exec The problem is minor, but without ->cred_guard_mutex held we can race with exec() and get the new ->mm but check old creds. Now we do not need to re-check task->mm after ptrace_may_access(), it can't be changed to the new mm under us. Strictly speaking, this also fixes another very minor problem. Unless security check fails or the task exits mm_for_maps() should never return NULL, the caller should get either old or new ->mm. Signed-off-by: Oleg Nesterov Acked-by: Serge Hallyn Signed-off-by: James Morris --- fs/proc/base.c | 22 +++++++++++----------- 1 file changed, 11 insertions(+), 11 deletions(-) (limited to 'fs') diff --git a/fs/proc/base.c b/fs/proc/base.c index f3c2e4085fed..175db258942f 100644 --- a/fs/proc/base.c +++ b/fs/proc/base.c @@ -234,19 +234,19 @@ static int check_mem_permission(struct task_struct *task) struct mm_struct *mm_for_maps(struct task_struct *task) { - struct mm_struct *mm = get_task_mm(task); + struct mm_struct *mm; - if (mm && mm != current->mm) { - /* - * task->mm can be changed before security check, - * in that case we must notice the change after. - */ - if (!ptrace_may_access(task, PTRACE_MODE_READ) || - mm != task->mm) { - mmput(mm); - mm = NULL; - } + if (mutex_lock_killable(&task->cred_guard_mutex)) + return NULL; + + mm = get_task_mm(task); + if (mm && mm != current->mm && + !ptrace_may_access(task, PTRACE_MODE_READ)) { + mmput(mm); + mm = NULL; } + mutex_unlock(&task->cred_guard_mutex); + return mm; } -- cgit v1.2.3 From b409d7a0ab46fe530efe52734984b4ed5d46c3eb Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Thu, 6 Aug 2009 23:29:34 +0200 Subject: ocfs2: Fix possible deadlock when extending quota file In OCFS2, allocator locks rank above transaction start. Thus we cannot extend quota file from inside a transaction less we could deadlock. We solve the problem by starting transaction not already in ocfs2_acquire_dquot() but only in ocfs2_local_read_dquot() and ocfs2_global_read_dquot() and we allocate blocks to quota files before starting the transaction. In case we crash, quota files will just have a few blocks more but that's no problem since we just use them next time we extend the quota file. Signed-off-by: Jan Kara Signed-off-by: Joel Becker --- fs/ocfs2/journal.h | 5 --- fs/ocfs2/quota_global.c | 115 ++++++++++++++++++++++++------------------------ 2 files changed, 57 insertions(+), 63 deletions(-) (limited to 'fs') diff --git a/fs/ocfs2/journal.h b/fs/ocfs2/journal.h index 4a4d3b55fd22..2c3222aec622 100644 --- a/fs/ocfs2/journal.h +++ b/fs/ocfs2/journal.h @@ -363,11 +363,6 @@ static inline int ocfs2_quota_trans_credits(struct super_block *sb) return credits; } -/* Number of credits needed for removing quota structure from file */ -int ocfs2_calc_qdel_credits(struct super_block *sb, int type); -/* Number of credits needed for initialization of new quota structure */ -int ocfs2_calc_qinit_credits(struct super_block *sb, int type); - /* group extend. inode update and last group update. */ #define OCFS2_GROUP_EXTEND_CREDITS (OCFS2_INODE_UPDATE_CREDITS + 1) diff --git a/fs/ocfs2/quota_global.c b/fs/ocfs2/quota_global.c index d604a6aa0a22..bf7742d0ee3b 100644 --- a/fs/ocfs2/quota_global.c +++ b/fs/ocfs2/quota_global.c @@ -215,11 +215,7 @@ ssize_t ocfs2_quota_write(struct super_block *sb, int type, loff_t rounded_end = ocfs2_align_bytes_to_blocks(sb, off + len); - down_write(&OCFS2_I(gqinode)->ip_alloc_sem); - err = ocfs2_extend_no_holes(gqinode, rounded_end, off); - up_write(&OCFS2_I(gqinode)->ip_alloc_sem); - if (err < 0) - goto out; + /* Space is already allocated in ocfs2_global_read_dquot() */ err = ocfs2_simple_size_update(gqinode, oinfo->dqi_gqi_bh, rounded_end); @@ -405,13 +401,36 @@ int ocfs2_global_write_info(struct super_block *sb, int type) return err; } +static int ocfs2_global_qinit_alloc(struct super_block *sb, int type) +{ + struct ocfs2_mem_dqinfo *oinfo = sb_dqinfo(sb, type)->dqi_priv; + + /* + * We may need to allocate tree blocks and a leaf block but not the + * root block + */ + return oinfo->dqi_gi.dqi_qtree_depth; +} + +static int ocfs2_calc_global_qinit_credits(struct super_block *sb, int type) +{ + /* We modify all the allocated blocks, tree root, and info block */ + return (ocfs2_global_qinit_alloc(sb, type) + 2) * + OCFS2_QUOTA_BLOCK_UPDATE_CREDITS; +} + /* Read in information from global quota file and acquire a reference to it. * dquot_acquire() has already started the transaction and locked quota file */ int ocfs2_global_read_dquot(struct dquot *dquot) { int err, err2, ex = 0; - struct ocfs2_mem_dqinfo *info = - sb_dqinfo(dquot->dq_sb, dquot->dq_type)->dqi_priv; + struct super_block *sb = dquot->dq_sb; + int type = dquot->dq_type; + struct ocfs2_mem_dqinfo *info = sb_dqinfo(sb, type)->dqi_priv; + struct ocfs2_super *osb = OCFS2_SB(sb); + struct inode *gqinode = info->dqi_gqinode; + int need_alloc = ocfs2_global_qinit_alloc(sb, type); + handle_t *handle = NULL; err = ocfs2_qinfo_lock(info, 0); if (err < 0) @@ -422,14 +441,33 @@ int ocfs2_global_read_dquot(struct dquot *dquot) OCFS2_DQUOT(dquot)->dq_use_count++; OCFS2_DQUOT(dquot)->dq_origspace = dquot->dq_dqb.dqb_curspace; OCFS2_DQUOT(dquot)->dq_originodes = dquot->dq_dqb.dqb_curinodes; + ocfs2_qinfo_unlock(info, 0); + if (!dquot->dq_off) { /* No real quota entry? */ - /* Upgrade to exclusive lock for allocation */ - ocfs2_qinfo_unlock(info, 0); - err = ocfs2_qinfo_lock(info, 1); - if (err < 0) - goto out_qlock; ex = 1; + /* + * Add blocks to quota file before we start a transaction since + * locking allocators ranks above a transaction start + */ + WARN_ON(journal_current_handle()); + down_write(&OCFS2_I(gqinode)->ip_alloc_sem); + err = ocfs2_extend_no_holes(gqinode, + gqinode->i_size + (need_alloc << sb->s_blocksize_bits), + gqinode->i_size); + up_write(&OCFS2_I(gqinode)->ip_alloc_sem); + if (err < 0) + goto out; + } + + handle = ocfs2_start_trans(osb, + ocfs2_calc_global_qinit_credits(sb, type)); + if (IS_ERR(handle)) { + err = PTR_ERR(handle); + goto out; } + err = ocfs2_qinfo_lock(info, ex); + if (err < 0) + goto out_trans; err = qtree_write_dquot(&info->dqi_gi, dquot); if (ex && info_dirty(sb_dqinfo(dquot->dq_sb, dquot->dq_type))) { err2 = __ocfs2_global_write_info(dquot->dq_sb, dquot->dq_type); @@ -441,6 +479,9 @@ out_qlock: ocfs2_qinfo_unlock(info, 1); else ocfs2_qinfo_unlock(info, 0); +out_trans: + if (handle) + ocfs2_commit_trans(osb, handle); out: if (err < 0) mlog_errno(err); @@ -638,24 +679,17 @@ out: return status; } -int ocfs2_calc_qdel_credits(struct super_block *sb, int type) +static int ocfs2_calc_qdel_credits(struct super_block *sb, int type) { - struct ocfs2_mem_dqinfo *oinfo; - int features[MAXQUOTAS] = { OCFS2_FEATURE_RO_COMPAT_USRQUOTA, - OCFS2_FEATURE_RO_COMPAT_GRPQUOTA }; - - if (!OCFS2_HAS_RO_COMPAT_FEATURE(sb, features[type])) - return 0; - - oinfo = sb_dqinfo(sb, type)->dqi_priv; + struct ocfs2_mem_dqinfo *oinfo = sb_dqinfo(sb, type)->dqi_priv; /* * We modify tree, leaf block, global info, local chunk header, * global and local inode; OCFS2_QINFO_WRITE_CREDITS already * accounts for inode update */ - return oinfo->dqi_gi.dqi_qtree_depth + + return (oinfo->dqi_gi.dqi_qtree_depth + 2) * + OCFS2_QUOTA_BLOCK_UPDATE_CREDITS + OCFS2_QINFO_WRITE_CREDITS + - 2 * OCFS2_QUOTA_BLOCK_UPDATE_CREDITS + OCFS2_INODE_UPDATE_CREDITS; } @@ -688,36 +722,10 @@ out: return status; } -int ocfs2_calc_qinit_credits(struct super_block *sb, int type) -{ - struct ocfs2_mem_dqinfo *oinfo; - int features[MAXQUOTAS] = { OCFS2_FEATURE_RO_COMPAT_USRQUOTA, - OCFS2_FEATURE_RO_COMPAT_GRPQUOTA }; - struct ocfs2_dinode *lfe, *gfe; - - if (!OCFS2_HAS_RO_COMPAT_FEATURE(sb, features[type])) - return 0; - - oinfo = sb_dqinfo(sb, type)->dqi_priv; - gfe = (struct ocfs2_dinode *)oinfo->dqi_gqi_bh->b_data; - lfe = (struct ocfs2_dinode *)oinfo->dqi_lqi_bh->b_data; - /* We can extend local file + global file. In local file we - * can modify info, chunk header block and dquot block. In - * global file we can modify info, tree and leaf block */ - return ocfs2_calc_extend_credits(sb, &lfe->id2.i_list, 0) + - ocfs2_calc_extend_credits(sb, &gfe->id2.i_list, 0) + - OCFS2_LOCAL_QINFO_WRITE_CREDITS + - 2 * OCFS2_QUOTA_BLOCK_UPDATE_CREDITS + - oinfo->dqi_gi.dqi_qtree_depth + - 2 * OCFS2_QUOTA_BLOCK_UPDATE_CREDITS; -} - static int ocfs2_acquire_dquot(struct dquot *dquot) { - handle_t *handle; struct ocfs2_mem_dqinfo *oinfo = sb_dqinfo(dquot->dq_sb, dquot->dq_type)->dqi_priv; - struct ocfs2_super *osb = OCFS2_SB(dquot->dq_sb); int status = 0; mlog_entry("id=%u, type=%d", dquot->dq_id, dquot->dq_type); @@ -726,16 +734,7 @@ static int ocfs2_acquire_dquot(struct dquot *dquot) status = ocfs2_lock_global_qf(oinfo, 1); if (status < 0) goto out; - handle = ocfs2_start_trans(osb, - ocfs2_calc_qinit_credits(dquot->dq_sb, dquot->dq_type)); - if (IS_ERR(handle)) { - status = PTR_ERR(handle); - mlog_errno(status); - goto out_ilock; - } status = dquot_acquire(dquot); - ocfs2_commit_trans(osb, handle); -out_ilock: ocfs2_unlock_global_qf(oinfo, 1); out: mlog_exit(status); -- cgit v1.2.3 From 0cc6eee130b0c062feec8446d9cecdb17d2cfad3 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Sat, 18 Jul 2009 18:14:53 -0400 Subject: xfs: avoid memory allocation under m_peraglock in growfs code Allocate the memory for the larger m_perag array before taking the per-AG lock as the per-AG lock can be taken under the i_lock which can be taken from reclaim context. Reported by the new reclaim context tracing in lockdep. Signed-off-by: Christoph Hellwig Reviewed-by: Felix Blyakher Signed-off-by: Felix Blyakher --- fs/xfs/xfs_fsops.c | 20 ++++++++++++++------ 1 file changed, 14 insertions(+), 6 deletions(-) (limited to 'fs') diff --git a/fs/xfs/xfs_fsops.c b/fs/xfs/xfs_fsops.c index cbd451bb4848..2d0b3e1da9e6 100644 --- a/fs/xfs/xfs_fsops.c +++ b/fs/xfs/xfs_fsops.c @@ -167,17 +167,25 @@ xfs_growfs_data_private( new = nb - mp->m_sb.sb_dblocks; oagcount = mp->m_sb.sb_agcount; if (nagcount > oagcount) { + void *new_perag, *old_perag; + xfs_filestream_flush(mp); + + new_perag = kmem_zalloc(sizeof(xfs_perag_t) * nagcount, + KM_MAYFAIL); + if (!new_perag) + return XFS_ERROR(ENOMEM); + down_write(&mp->m_peraglock); - mp->m_perag = kmem_realloc(mp->m_perag, - sizeof(xfs_perag_t) * nagcount, - sizeof(xfs_perag_t) * oagcount, - KM_SLEEP); - memset(&mp->m_perag[oagcount], 0, - (nagcount - oagcount) * sizeof(xfs_perag_t)); + memcpy(new_perag, mp->m_perag, sizeof(xfs_perag_t) * oagcount); + old_perag = mp->m_perag; + mp->m_perag = new_perag; + mp->m_flags |= XFS_MOUNT_32BITINODES; nagimax = xfs_initialize_perag(mp, nagcount); up_write(&mp->m_peraglock); + + kmem_free(old_perag); } tp = xfs_trans_alloc(mp, XFS_TRANS_GROWFS); tp->t_flags |= XFS_TRANS_RESERVE; -- cgit v1.2.3 From ca35dcd6cae7d4a780c484c53f45548c4719f82c Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Sat, 18 Jul 2009 18:14:54 -0400 Subject: xfs: switch to NOFS allocation under i_lock in xfs_getbmap xfs_getbmap allocates memory with i_lock held, but i_lock is taken in reclaim context so all allocations under it must avoid recursions into the filesystem. Reported by the new reclaim context tracing in lockdep. Signed-off-by: Christoph Hellwig Reviewed-by: Felix Blyakher Signed-off-by: Felix Blyakher --- fs/xfs/xfs_bmap.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/xfs/xfs_bmap.c b/fs/xfs/xfs_bmap.c index 7928b9983c1d..8ee5b5a76a2a 100644 --- a/fs/xfs/xfs_bmap.c +++ b/fs/xfs/xfs_bmap.c @@ -6009,7 +6009,7 @@ xfs_getbmap( */ error = ENOMEM; subnex = 16; - map = kmem_alloc(subnex * sizeof(*map), KM_MAYFAIL); + map = kmem_alloc(subnex * sizeof(*map), KM_MAYFAIL | KM_NOFS); if (!map) goto out_unlock_ilock; -- cgit v1.2.3 From f41d7fb9da05b604f8a69fb6cac2a0563c8ede4e Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Sat, 18 Jul 2009 18:14:55 -0400 Subject: xfs: switch to NOFS allocation under i_lock in xfs_da_state_alloc xfs_da_state_alloc is always called with i_lock held, but i_lock is taken in reclaim context so all allocations under it must avoid recursions into the filesystem. Reported by the new reclaim context tracing in lockdep. Signed-off-by: Christoph Hellwig Reviewed-by: Felix Blyakher Signed-off-by: Felix Blyakher --- fs/xfs/xfs_da_btree.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/xfs/xfs_da_btree.c b/fs/xfs/xfs_da_btree.c index 9ff6e57a5075..bd0bb6dfcdfa 100644 --- a/fs/xfs/xfs_da_btree.c +++ b/fs/xfs/xfs_da_btree.c @@ -2201,7 +2201,7 @@ kmem_zone_t *xfs_dabuf_zone; /* dabuf zone */ xfs_da_state_t * xfs_da_state_alloc(void) { - return kmem_zone_zalloc(xfs_da_state_zone, KM_SLEEP); + return kmem_zone_zalloc(xfs_da_state_zone, KM_NOFS); } /* -- cgit v1.2.3 From 73195ed7864ae4a1fb0bea2ed9df59d19b4fde90 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Sat, 18 Jul 2009 18:14:56 -0400 Subject: xfs: switch to NOFS allocation under i_lock in xfs_da_buf_make i_lock is taken in the reclaim context so all allocations under it must avoid recursions into the filesystem. Reported by the new reclaim context tracing in lockdep. Signed-off-by: Christoph Hellwig Reviewed-by: Felix Blyakher Signed-off-by: Felix Blyakher --- fs/xfs/xfs_da_btree.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/xfs/xfs_da_btree.c b/fs/xfs/xfs_da_btree.c index bd0bb6dfcdfa..2847bbc1c534 100644 --- a/fs/xfs/xfs_da_btree.c +++ b/fs/xfs/xfs_da_btree.c @@ -2261,9 +2261,9 @@ xfs_da_buf_make(int nbuf, xfs_buf_t **bps, inst_t *ra) int off; if (nbuf == 1) - dabuf = kmem_zone_alloc(xfs_dabuf_zone, KM_SLEEP); + dabuf = kmem_zone_alloc(xfs_dabuf_zone, KM_NOFS); else - dabuf = kmem_alloc(XFS_DA_BUF_SIZE(nbuf), KM_SLEEP); + dabuf = kmem_alloc(XFS_DA_BUF_SIZE(nbuf), KM_NOFS); dabuf->dirty = 0; #ifdef XFS_DABUF_DEBUG dabuf->ra = ra; -- cgit v1.2.3 From 3f52c2f0a07c23771909cc53f2e9451a7f1bf253 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Sat, 18 Jul 2009 18:14:57 -0400 Subject: xfs: switch to NOFS allocation under i_lock in xfs_dir_cilookup_result xfs_dir_cilookup_result is always called with i_lock held, but i_lock is taken in reclaim context so all allocations under it must avoid recursions into the filesystem. Reported by the new reclaim context tracing in lockdep. Signed-off-by: Christoph Hellwig Reviewed-by: Felix Blyakher Signed-off-by: Felix Blyakher --- fs/xfs/xfs_dir2.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/xfs/xfs_dir2.c b/fs/xfs/xfs_dir2.c index c657bec6d951..bb1d58eb3982 100644 --- a/fs/xfs/xfs_dir2.c +++ b/fs/xfs/xfs_dir2.c @@ -256,7 +256,7 @@ xfs_dir_cilookup_result( !(args->op_flags & XFS_DA_OP_CILOOKUP)) return EEXIST; - args->value = kmem_alloc(len, KM_MAYFAIL); + args->value = kmem_alloc(len, KM_NOFS | KM_MAYFAIL); if (!args->value) return ENOMEM; -- cgit v1.2.3 From 36fae17a648e0aee5d9560514d08477ef48dc87f Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Sat, 18 Jul 2009 18:14:58 -0400 Subject: xfs: switch to NOFS allocation under i_lock in xfs_buf_associate_memory xfs_buf_associate_memory is used for setting up the spare buffer for the log wrap case in xlog_sync which can happen under i_lock when called from xfs_fsync. The i_lock mutex is taken in reclaim context so all allocations under it must avoid recursions into the filesystem. There are a couple more uses of xfs_buf_associate_memory in the log recovery code that are also affected by this, but I'd rather keep the code simple than passing on a gfp_mask argument. Longer term we should just stop requiring the memoery allocation in xlog_sync by some smaller rework of the buffer layer. Reported by the new reclaim context tracing in lockdep. Signed-off-by: Christoph Hellwig Reviewed-by: Felix Blyakher Signed-off-by: Felix Blyakher --- fs/xfs/linux-2.6/xfs_buf.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/xfs/linux-2.6/xfs_buf.c b/fs/xfs/linux-2.6/xfs_buf.c index 1418b916fc27..178c20c13e83 100644 --- a/fs/xfs/linux-2.6/xfs_buf.c +++ b/fs/xfs/linux-2.6/xfs_buf.c @@ -770,7 +770,7 @@ xfs_buf_associate_memory( bp->b_pages = NULL; bp->b_addr = mem; - rval = _xfs_buf_get_pages(bp, page_count, 0); + rval = _xfs_buf_get_pages(bp, page_count, XBF_DONT_BLOCK); if (rval) return rval; -- cgit v1.2.3 From 10746e47e722b5688fcd6eba9fbf9b2e64a248a7 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Sat, 18 Jul 2009 18:14:59 -0400 Subject: xfs: switch to NOFS allocation under i_lock in xfs_attr_rmtval_set xfs_attr_rmtval_set is always called with i_lock held, and i_lock is taken in reclaim context so all allocations under it must avoid recursions into the filesystem. Reported by the new reclaim context tracing in lockdep. Signed-off-by: Christoph Hellwig Reviewed-by: Felix Blyakher Signed-off-by: Felix Blyakher --- fs/xfs/xfs_attr.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/xfs/xfs_attr.c b/fs/xfs/xfs_attr.c index db15feb906ff..bfb583791fe2 100644 --- a/fs/xfs/xfs_attr.c +++ b/fs/xfs/xfs_attr.c @@ -2141,8 +2141,8 @@ xfs_attr_rmtval_set(xfs_da_args_t *args) dblkno = XFS_FSB_TO_DADDR(mp, map.br_startblock), blkcnt = XFS_FSB_TO_BB(mp, map.br_blockcount); - bp = xfs_buf_get_flags(mp->m_ddev_targp, dblkno, - blkcnt, XFS_BUF_LOCK); + bp = xfs_buf_get_flags(mp->m_ddev_targp, dblkno, blkcnt, + XFS_BUF_LOCK | XBF_DONT_BLOCK); ASSERT(bp); ASSERT(!XFS_BUF_GETERROR(bp)); -- cgit v1.2.3 From 7b02ecb3031b192823bc732ae717febc0a59aa92 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Sat, 18 Jul 2009 18:15:00 -0400 Subject: xfs: switch to NOFS allocation under i_lock in xfs_readlink_bmap xfs_readlink_bmap is called with i_lock held, but i_lock is taken in reclaim context so all allocations under it must avoid recursions into the filesystem. Reported by the new reclaim context tracing in lockdep. Signed-off-by: Christoph Hellwig Reviewed-by: Felix Blyakher Signed-off-by: Felix Blyakher --- fs/xfs/xfs_vnodeops.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/xfs/xfs_vnodeops.c b/fs/xfs/xfs_vnodeops.c index c4eca5ed5dab..492d75bae2bf 100644 --- a/fs/xfs/xfs_vnodeops.c +++ b/fs/xfs/xfs_vnodeops.c @@ -538,7 +538,9 @@ xfs_readlink_bmap( d = XFS_FSB_TO_DADDR(mp, mval[n].br_startblock); byte_cnt = XFS_FSB_TO_B(mp, mval[n].br_blockcount); - bp = xfs_buf_read(mp->m_ddev_targp, d, BTOBB(byte_cnt), 0); + bp = xfs_buf_read_flags(mp->m_ddev_targp, d, BTOBB(byte_cnt), + XBF_LOCK | XBF_MAPPED | + XBF_DONT_BLOCK); error = XFS_BUF_GETERROR(bp); if (error) { xfs_ioerror_alert("xfs_readlink", -- cgit v1.2.3 From ddd3a14e0f030f0f7b900621f67532285b8657ef Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Sat, 18 Jul 2009 18:15:01 -0400 Subject: xfs: switch to NOFS allocation under i_lock in xfs_attr_rmtval_get xfs_attr_rmtval_get is always called with i_lock held, but i_lock is taken in reclaim context so all allocations under it must avoid recursions into the filesystem. Reported by the new reclaim context tracing in lockdep. Signed-off-by: Christoph Hellwig Reviewed-by: Felix Blyakher Signed-off-by: Felix Blyakher --- fs/xfs/xfs_attr.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/xfs/xfs_attr.c b/fs/xfs/xfs_attr.c index bfb583791fe2..4ece1906bd41 100644 --- a/fs/xfs/xfs_attr.c +++ b/fs/xfs/xfs_attr.c @@ -2010,7 +2010,9 @@ xfs_attr_rmtval_get(xfs_da_args_t *args) dblkno = XFS_FSB_TO_DADDR(mp, map[i].br_startblock); blkcnt = XFS_FSB_TO_BB(mp, map[i].br_blockcount); error = xfs_read_buf(mp, mp->m_ddev_targp, dblkno, - blkcnt, XFS_BUF_LOCK, &bp); + blkcnt, + XFS_BUF_LOCK | XBF_DONT_BLOCK, + &bp); if (error) return(error); -- cgit v1.2.3 From e0c222c411e22f086e929cd69fdcc89336164ec1 Mon Sep 17 00:00:00 2001 From: Eric Sandeen Date: Mon, 20 Jul 2009 10:52:15 -0500 Subject: use XFS_CORRUPTION_ERROR in xfs_btree_check_sblock In Red Hat Bug 512552 - Can't write to XFS mount during raid5 resync a user ran into corruption while resyncing a raid, and we failed a consistency test, but didn't get much more info; it'd be nice to call XFS_CORRUPTION_ERROR here so we can see the buffer contents. Signed-off-by: Eric Sandeen Reviewed-by: Christoph Hellwig Signed-off-by: Felix Blyakher --- fs/xfs/xfs_btree.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/xfs/xfs_btree.c b/fs/xfs/xfs_btree.c index e9df99574829..26717388acf5 100644 --- a/fs/xfs/xfs_btree.c +++ b/fs/xfs/xfs_btree.c @@ -120,8 +120,8 @@ xfs_btree_check_sblock( XFS_RANDOM_BTREE_CHECK_SBLOCK))) { if (bp) xfs_buftrace("SBTREE ERROR", bp); - XFS_ERROR_REPORT("xfs_btree_check_sblock", XFS_ERRLEVEL_LOW, - cur->bc_mp); + XFS_CORRUPTION_ERROR("xfs_btree_check_sblock", + XFS_ERRLEVEL_LOW, cur->bc_mp, block); return XFS_ERROR(EFSCORRUPTED); } return 0; -- cgit v1.2.3 From b89d4208de3de442c9025919c4261be0b38e79a4 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Mon, 10 Aug 2009 11:32:18 -0300 Subject: xfs: check for dinode realtime flag corruption Ramon tested XFS with a modified version of fsfuzzer and hit a NULL pointer dereference in __xfs_get_blocks due to the RT device target pointer being NULL. To fix this reject inode with the realtime bit set on a a filesystem without an RT subvolume during inode read. Signed-off-by: Christoph Hellwig Reviewed-by: Eric Sandeen Reviewed-by: Felix Blyakher Reported-by: Ramon de Carvalho Valle Tested-by: Ramon de Carvalho Valle Signed-off-by: Felix Blyakher --- fs/xfs/xfs_inode.c | 10 ++++++++++ 1 file changed, 10 insertions(+) (limited to 'fs') diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c index 1f22d65fed0a..da428b3fe0f5 100644 --- a/fs/xfs/xfs_inode.c +++ b/fs/xfs/xfs_inode.c @@ -343,6 +343,16 @@ xfs_iformat( return XFS_ERROR(EFSCORRUPTED); } + if (unlikely((ip->i_d.di_flags & XFS_DIFLAG_REALTIME) && + !ip->i_mount->m_rtdev_targp)) { + xfs_fs_repair_cmn_err(CE_WARN, ip->i_mount, + "corrupt dinode %Lu, has realtime flag set.", + ip->i_ino); + XFS_CORRUPTION_ERROR("xfs_iformat(realtime)", + XFS_ERRLEVEL_LOW, ip->i_mount, dip); + return XFS_ERROR(EFSCORRUPTED); + } + switch (ip->i_d.di_mode & S_IFMT) { case S_IFIFO: case S_IFCHR: -- cgit v1.2.3 From a8914f3a6d72c97328597a556a99daaf5cc288ae Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Mon, 10 Aug 2009 11:32:44 -0300 Subject: xfs: fix spin_is_locked assert on uni-processor builds Without SMP or preemption spin_is_locked always returns false, so we can't do an assert with it. Instead use assert_spin_locked, which does the right thing on all builds. Signed-off-by: Christoph Hellwig Reviewed-by: Eric Sandeen Reported-by: Johannes Engel Tested-by: Johannes Engel Signed-off-by: Felix Blyakher --- fs/xfs/xfs_log.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/xfs/xfs_log.c b/fs/xfs/xfs_log.c index 3750f04ede0b..9dbdff3ea484 100644 --- a/fs/xfs/xfs_log.c +++ b/fs/xfs/xfs_log.c @@ -3180,7 +3180,7 @@ try_again: STATIC void xlog_state_want_sync(xlog_t *log, xlog_in_core_t *iclog) { - ASSERT(spin_is_locked(&log->l_icloglock)); + assert_spin_locked(&log->l_icloglock); if (iclog->ic_state == XLOG_STATE_ACTIVE) { xlog_state_switch_iclogs(log, iclog, 0); -- cgit v1.2.3 From 1ae88b2e446261c038f2c0c3150ffae142b227a2 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Wed, 12 Aug 2009 09:12:30 -0400 Subject: NFS: Fix an O_DIRECT Oops... We can't call nfs_readdata_release()/nfs_writedata_release() without first initialising and referencing args.context. Doing so inside nfs_direct_read_schedule_segment()/nfs_direct_write_schedule_segment() causes an Oops. We should rather be calling nfs_readdata_free()/nfs_writedata_free() in those cases. Looking at the O_DIRECT code, the "struct nfs_direct_req" is already referencing the nfs_open_context for us. Since the readdata and writedata structures carry a reference to that, we can simplify things by getting rid of the extra nfs_open_context references, so that we can replace all instances of nfs_readdata_release()/nfs_writedata_release(). Reported-by: Catalin Marinas Signed-off-by: Trond Myklebust Tested-by: Catalin Marinas Cc: stable@kernel.org Signed-off-by: Linus Torvalds --- fs/nfs/direct.c | 20 ++++++++++---------- fs/nfs/read.c | 6 ++---- fs/nfs/write.c | 6 ++---- include/linux/nfs_fs.h | 5 ++--- 4 files changed, 16 insertions(+), 21 deletions(-) (limited to 'fs') diff --git a/fs/nfs/direct.c b/fs/nfs/direct.c index 489fc01a3204..e4e089a8f294 100644 --- a/fs/nfs/direct.c +++ b/fs/nfs/direct.c @@ -255,7 +255,7 @@ static void nfs_direct_read_release(void *calldata) if (put_dreq(dreq)) nfs_direct_complete(dreq); - nfs_readdata_release(calldata); + nfs_readdata_free(data); } static const struct rpc_call_ops nfs_read_direct_ops = { @@ -314,14 +314,14 @@ static ssize_t nfs_direct_read_schedule_segment(struct nfs_direct_req *dreq, data->npages, 1, 0, data->pagevec, NULL); up_read(¤t->mm->mmap_sem); if (result < 0) { - nfs_readdata_release(data); + nfs_readdata_free(data); break; } if ((unsigned)result < data->npages) { bytes = result * PAGE_SIZE; if (bytes <= pgbase) { nfs_direct_release_pages(data->pagevec, result); - nfs_readdata_release(data); + nfs_readdata_free(data); break; } bytes -= pgbase; @@ -334,7 +334,7 @@ static ssize_t nfs_direct_read_schedule_segment(struct nfs_direct_req *dreq, data->inode = inode; data->cred = msg.rpc_cred; data->args.fh = NFS_FH(inode); - data->args.context = get_nfs_open_context(ctx); + data->args.context = ctx; data->args.offset = pos; data->args.pgbase = pgbase; data->args.pages = data->pagevec; @@ -441,7 +441,7 @@ static void nfs_direct_free_writedata(struct nfs_direct_req *dreq) struct nfs_write_data *data = list_entry(dreq->rewrite_list.next, struct nfs_write_data, pages); list_del(&data->pages); nfs_direct_release_pages(data->pagevec, data->npages); - nfs_writedata_release(data); + nfs_writedata_free(data); } } @@ -534,7 +534,7 @@ static void nfs_direct_commit_release(void *calldata) dprintk("NFS: %5u commit returned %d\n", data->task.tk_pid, status); nfs_direct_write_complete(dreq, data->inode); - nfs_commitdata_release(calldata); + nfs_commit_free(data); } static const struct rpc_call_ops nfs_commit_direct_ops = { @@ -570,7 +570,7 @@ static void nfs_direct_commit_schedule(struct nfs_direct_req *dreq) data->args.fh = NFS_FH(data->inode); data->args.offset = 0; data->args.count = 0; - data->args.context = get_nfs_open_context(dreq->ctx); + data->args.context = dreq->ctx; data->res.count = 0; data->res.fattr = &data->fattr; data->res.verf = &data->verf; @@ -734,14 +734,14 @@ static ssize_t nfs_direct_write_schedule_segment(struct nfs_direct_req *dreq, data->npages, 0, 0, data->pagevec, NULL); up_read(¤t->mm->mmap_sem); if (result < 0) { - nfs_writedata_release(data); + nfs_writedata_free(data); break; } if ((unsigned)result < data->npages) { bytes = result * PAGE_SIZE; if (bytes <= pgbase) { nfs_direct_release_pages(data->pagevec, result); - nfs_writedata_release(data); + nfs_writedata_free(data); break; } bytes -= pgbase; @@ -756,7 +756,7 @@ static ssize_t nfs_direct_write_schedule_segment(struct nfs_direct_req *dreq, data->inode = inode; data->cred = msg.rpc_cred; data->args.fh = NFS_FH(inode); - data->args.context = get_nfs_open_context(ctx); + data->args.context = ctx; data->args.offset = pos; data->args.pgbase = pgbase; data->args.pages = data->pagevec; diff --git a/fs/nfs/read.c b/fs/nfs/read.c index 73ea5e8d66ce..12c9e66d3f1d 100644 --- a/fs/nfs/read.c +++ b/fs/nfs/read.c @@ -60,17 +60,15 @@ struct nfs_read_data *nfs_readdata_alloc(unsigned int pagecount) return p; } -static void nfs_readdata_free(struct nfs_read_data *p) +void nfs_readdata_free(struct nfs_read_data *p) { if (p && (p->pagevec != &p->page_array[0])) kfree(p->pagevec); mempool_free(p, nfs_rdata_mempool); } -void nfs_readdata_release(void *data) +static void nfs_readdata_release(struct nfs_read_data *rdata) { - struct nfs_read_data *rdata = data; - put_nfs_open_context(rdata->args.context); nfs_readdata_free(rdata); } diff --git a/fs/nfs/write.c b/fs/nfs/write.c index 0a0a2ff767c3..a34fae21fe10 100644 --- a/fs/nfs/write.c +++ b/fs/nfs/write.c @@ -87,17 +87,15 @@ struct nfs_write_data *nfs_writedata_alloc(unsigned int pagecount) return p; } -static void nfs_writedata_free(struct nfs_write_data *p) +void nfs_writedata_free(struct nfs_write_data *p) { if (p && (p->pagevec != &p->page_array[0])) kfree(p->pagevec); mempool_free(p, nfs_wdata_mempool); } -void nfs_writedata_release(void *data) +static void nfs_writedata_release(struct nfs_write_data *wdata) { - struct nfs_write_data *wdata = data; - put_nfs_open_context(wdata->args.context); nfs_writedata_free(wdata); } diff --git a/include/linux/nfs_fs.h b/include/linux/nfs_fs.h index fdffb413b192..f6b90240dd41 100644 --- a/include/linux/nfs_fs.h +++ b/include/linux/nfs_fs.h @@ -473,7 +473,6 @@ extern int nfs_writepages(struct address_space *, struct writeback_control *); extern int nfs_flush_incompatible(struct file *file, struct page *page); extern int nfs_updatepage(struct file *, struct page *, unsigned int, unsigned int); extern int nfs_writeback_done(struct rpc_task *, struct nfs_write_data *); -extern void nfs_writedata_release(void *); /* * Try to write back everything synchronously (but check the @@ -488,7 +487,6 @@ extern int nfs_wb_page_cancel(struct inode *inode, struct page* page); extern int nfs_commit_inode(struct inode *, int); extern struct nfs_write_data *nfs_commitdata_alloc(void); extern void nfs_commit_free(struct nfs_write_data *wdata); -extern void nfs_commitdata_release(void *wdata); #else static inline int nfs_commit_inode(struct inode *inode, int how) @@ -507,6 +505,7 @@ nfs_have_writebacks(struct inode *inode) * Allocate nfs_write_data structures */ extern struct nfs_write_data *nfs_writedata_alloc(unsigned int npages); +extern void nfs_writedata_free(struct nfs_write_data *); /* * linux/fs/nfs/read.c @@ -515,7 +514,6 @@ extern int nfs_readpage(struct file *, struct page *); extern int nfs_readpages(struct file *, struct address_space *, struct list_head *, unsigned); extern int nfs_readpage_result(struct rpc_task *, struct nfs_read_data *); -extern void nfs_readdata_release(void *data); extern int nfs_readpage_async(struct nfs_open_context *, struct inode *, struct page *); @@ -523,6 +521,7 @@ extern int nfs_readpage_async(struct nfs_open_context *, struct inode *, * Allocate nfs_read_data structures */ extern struct nfs_read_data *nfs_readdata_alloc(unsigned int npages); +extern void nfs_readdata_free(struct nfs_read_data *); /* * linux/fs/nfs3proc.c -- cgit v1.2.3 From d7e623da1a757fbd8c117fa29190ca8bef14dab3 Mon Sep 17 00:00:00 2001 From: Steven Whitehouse Date: Tue, 11 Aug 2009 11:20:11 +0100 Subject: GFS2: Fix permissions on "recover" file Although this file is only ever written and not read by userspace, it seems that the utils are opening this file O_RDWR, so we need to allow that. Also fixes the whitespace which seemed to be broken. Signed-off-by: Steven Whitehouse Cc: David Teigland --- fs/gfs2/sys.c | 20 ++++++++++---------- 1 file changed, 10 insertions(+), 10 deletions(-) (limited to 'fs') diff --git a/fs/gfs2/sys.c b/fs/gfs2/sys.c index 23419dc3027b..a7cbfbd340c7 100644 --- a/fs/gfs2/sys.c +++ b/fs/gfs2/sys.c @@ -386,16 +386,16 @@ static ssize_t jid_show(struct gfs2_sbd *sdp, char *buf) #define GDLM_ATTR(_name,_mode,_show,_store) \ static struct gfs2_attr gdlm_attr_##_name = __ATTR(_name,_mode,_show,_store) -GDLM_ATTR(proto_name, 0444, proto_name_show, NULL); -GDLM_ATTR(block, 0644, block_show, block_store); -GDLM_ATTR(withdraw, 0644, withdraw_show, withdraw_store); -GDLM_ATTR(id, 0444, lkid_show, NULL); -GDLM_ATTR(jid, 0444, jid_show, NULL); -GDLM_ATTR(first, 0444, lkfirst_show, NULL); -GDLM_ATTR(first_done, 0444, first_done_show, NULL); -GDLM_ATTR(recover, 0200, NULL, recover_store); -GDLM_ATTR(recover_done, 0444, recover_done_show, NULL); -GDLM_ATTR(recover_status, 0444, recover_status_show, NULL); +GDLM_ATTR(proto_name, 0444, proto_name_show, NULL); +GDLM_ATTR(block, 0644, block_show, block_store); +GDLM_ATTR(withdraw, 0644, withdraw_show, withdraw_store); +GDLM_ATTR(id, 0444, lkid_show, NULL); +GDLM_ATTR(jid, 0444, jid_show, NULL); +GDLM_ATTR(first, 0444, lkfirst_show, NULL); +GDLM_ATTR(first_done, 0444, first_done_show, NULL); +GDLM_ATTR(recover, 0600, NULL, recover_store); +GDLM_ATTR(recover_done, 0444, recover_done_show, NULL); +GDLM_ATTR(recover_status, 0444, recover_status_show, NULL); static struct attribute *lock_module_attrs[] = { &gdlm_attr_proto_name.attr, -- cgit v1.2.3 From b2add73dbf93fd50f00564d7abc3e2b9aa9dd20c Mon Sep 17 00:00:00 2001 From: Guillaume Knispel Date: Sat, 15 Aug 2009 19:30:24 +0200 Subject: poll/select: initialize triggered field of struct poll_wqueues The triggered field of struct poll_wqueues introduced in commit 5f820f648c92a5ecc771a96b3c29aa6e90013bba ("poll: allow f_op->poll to sleep"). It was first set to 1 in pollwake() (now __pollwake() ), tested and later set to 0 in poll_schedule_timeout(), but not initialized before. As a result when the process needs to sleep, triggered was likely to be non-zero even if pollwake() is not called before the first poll_schedule_timeout(), meaning schedule_hrtimeout_range() would not be called and an extra loop calling all ->poll() would be done. This patch initialize triggered to 0 in poll_initwait() so the ->poll() are not called twice before the process goes to sleep when it needs to. Signed-off-by: Guillaume Knispel Acked-by: Thomas Gleixner Acked-by: Tejun Heo Cc: stable@kernel.org Signed-off-by: Linus Torvalds --- fs/select.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/select.c b/fs/select.c index d870237e42c7..8084834e123e 100644 --- a/fs/select.c +++ b/fs/select.c @@ -110,6 +110,7 @@ void poll_initwait(struct poll_wqueues *pwq) { init_poll_funcptr(&pwq->pt, __pollwait); pwq->polling_task = current; + pwq->triggered = 0; pwq->error = 0; pwq->table = NULL; pwq->inline_index = 0; -- cgit v1.2.3 From bc990f5cb424cdca9dda866785d088e2c2110ecc Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Sun, 16 Aug 2009 20:36:34 -0400 Subject: xfs: fix locking in xfs_iget_cache_hit The locking in xfs_iget_cache_hit currently has numerous problems: - we clear the reclaim tag without i_flags_lock which protects modifications to it - we call inode_init_always which can sleep with pag_ici_lock held (this is oss.sgi.com BZ #819) - we acquire and drop i_flags_lock a lot and thus provide no consistency between the various flags we set/clear under it This patch fixes all that with a major revamp of the locking in the function. The new version acquires i_flags_lock early and only drops it once we need to call into inode_init_always or before calling xfs_ilock. This patch fixes a bug seen in the wild where we race modifying the reclaim tag. Signed-off-by: Christoph Hellwig Reviewed-by: Felix Blyakher Reviewed-by: Eric Sandeen Signed-off-by: Felix Blyakher --- fs/xfs/linux-2.6/xfs_sync.c | 13 ++++- fs/xfs/linux-2.6/xfs_sync.h | 1 + fs/xfs/xfs_iget.c | 113 +++++++++++++++++++++++--------------------- 3 files changed, 70 insertions(+), 57 deletions(-) (limited to 'fs') diff --git a/fs/xfs/linux-2.6/xfs_sync.c b/fs/xfs/linux-2.6/xfs_sync.c index b619d6b8ca43..98ef624d9baf 100644 --- a/fs/xfs/linux-2.6/xfs_sync.c +++ b/fs/xfs/linux-2.6/xfs_sync.c @@ -708,6 +708,16 @@ xfs_reclaim_inode( return 0; } +void +__xfs_inode_set_reclaim_tag( + struct xfs_perag *pag, + struct xfs_inode *ip) +{ + radix_tree_tag_set(&pag->pag_ici_root, + XFS_INO_TO_AGINO(ip->i_mount, ip->i_ino), + XFS_ICI_RECLAIM_TAG); +} + /* * We set the inode flag atomically with the radix tree tag. * Once we get tag lookups on the radix tree, this inode flag @@ -722,8 +732,7 @@ xfs_inode_set_reclaim_tag( read_lock(&pag->pag_ici_lock); spin_lock(&ip->i_flags_lock); - radix_tree_tag_set(&pag->pag_ici_root, - XFS_INO_TO_AGINO(mp, ip->i_ino), XFS_ICI_RECLAIM_TAG); + __xfs_inode_set_reclaim_tag(pag, ip); __xfs_iflags_set(ip, XFS_IRECLAIMABLE); spin_unlock(&ip->i_flags_lock); read_unlock(&pag->pag_ici_lock); diff --git a/fs/xfs/linux-2.6/xfs_sync.h b/fs/xfs/linux-2.6/xfs_sync.h index 2a10301c99c7..59120602588a 100644 --- a/fs/xfs/linux-2.6/xfs_sync.h +++ b/fs/xfs/linux-2.6/xfs_sync.h @@ -48,6 +48,7 @@ int xfs_reclaim_inode(struct xfs_inode *ip, int locked, int sync_mode); int xfs_reclaim_inodes(struct xfs_mount *mp, int mode); void xfs_inode_set_reclaim_tag(struct xfs_inode *ip); +void __xfs_inode_set_reclaim_tag(struct xfs_perag *pag, struct xfs_inode *ip); void xfs_inode_clear_reclaim_tag(struct xfs_inode *ip); void __xfs_inode_clear_reclaim_tag(struct xfs_mount *mp, struct xfs_perag *pag, struct xfs_inode *ip); diff --git a/fs/xfs/xfs_iget.c b/fs/xfs/xfs_iget.c index 34ec86923f7e..ecbf8b4d2e2e 100644 --- a/fs/xfs/xfs_iget.c +++ b/fs/xfs/xfs_iget.c @@ -191,80 +191,82 @@ xfs_iget_cache_hit( int flags, int lock_flags) __releases(pag->pag_ici_lock) { + struct inode *inode = VFS_I(ip); struct xfs_mount *mp = ip->i_mount; - int error = EAGAIN; + int error; + + spin_lock(&ip->i_flags_lock); /* - * If INEW is set this inode is being set up - * If IRECLAIM is set this inode is being torn down - * Pause and try again. + * If we are racing with another cache hit that is currently + * instantiating this inode or currently recycling it out of + * reclaimabe state, wait for the initialisation to complete + * before continuing. + * + * XXX(hch): eventually we should do something equivalent to + * wait_on_inode to wait for these flags to be cleared + * instead of polling for it. */ - if (xfs_iflags_test(ip, (XFS_INEW|XFS_IRECLAIM))) { + if (ip->i_flags & (XFS_INEW|XFS_IRECLAIM)) { XFS_STATS_INC(xs_ig_frecycle); + error = EAGAIN; goto out_error; } - /* If IRECLAIMABLE is set, we've torn down the vfs inode part */ - if (xfs_iflags_test(ip, XFS_IRECLAIMABLE)) { - - /* - * If lookup is racing with unlink, then we should return an - * error immediately so we don't remove it from the reclaim - * list and potentially leak the inode. - */ - if ((ip->i_d.di_mode == 0) && !(flags & XFS_IGET_CREATE)) { - error = ENOENT; - goto out_error; - } + /* + * If lookup is racing with unlink return an error immediately. + */ + if (ip->i_d.di_mode == 0 && !(flags & XFS_IGET_CREATE)) { + error = ENOENT; + goto out_error; + } + /* + * If IRECLAIMABLE is set, we've torn down the VFS inode already. + * Need to carefully get it back into useable state. + */ + if (ip->i_flags & XFS_IRECLAIMABLE) { xfs_itrace_exit_tag(ip, "xfs_iget.alloc"); /* - * We need to re-initialise the VFS inode as it has been - * 'freed' by the VFS. Do this here so we can deal with - * errors cleanly, then tag it so it can be set up correctly - * later. + * We need to set XFS_INEW atomically with clearing the + * reclaimable tag so that we do have an indicator of the + * inode still being initialized. */ - if (inode_init_always(mp->m_super, VFS_I(ip))) { - error = ENOMEM; - goto out_error; - } + ip->i_flags |= XFS_INEW; + ip->i_flags &= ~XFS_IRECLAIMABLE; + __xfs_inode_clear_reclaim_tag(mp, pag, ip); - /* - * We must set the XFS_INEW flag before clearing the - * XFS_IRECLAIMABLE flag so that if a racing lookup does - * not find the XFS_IRECLAIMABLE above but has the igrab() - * below succeed we can safely check XFS_INEW to detect - * that this inode is still being initialised. - */ - xfs_iflags_set(ip, XFS_INEW); - xfs_iflags_clear(ip, XFS_IRECLAIMABLE); + spin_unlock(&ip->i_flags_lock); + read_unlock(&pag->pag_ici_lock); - /* clear the radix tree reclaim flag as well. */ - __xfs_inode_clear_reclaim_tag(mp, pag, ip); - } else if (!igrab(VFS_I(ip))) { + error = -inode_init_always(mp->m_super, inode); + if (error) { + /* + * Re-initializing the inode failed, and we are in deep + * trouble. Try to re-add it to the reclaim list. + */ + read_lock(&pag->pag_ici_lock); + spin_lock(&ip->i_flags_lock); + + ip->i_flags &= ~XFS_INEW; + ip->i_flags |= XFS_IRECLAIMABLE; + __xfs_inode_set_reclaim_tag(pag, ip); + goto out_error; + } + inode->i_state = I_LOCK|I_NEW; + } else { /* If the VFS inode is being torn down, pause and try again. */ - XFS_STATS_INC(xs_ig_frecycle); - goto out_error; - } else if (xfs_iflags_test(ip, XFS_INEW)) { - /* - * We are racing with another cache hit that is - * currently recycling this inode out of the XFS_IRECLAIMABLE - * state. Wait for the initialisation to complete before - * continuing. - */ - wait_on_inode(VFS_I(ip)); - } + if (!igrab(inode)) { + error = EAGAIN; + goto out_error; + } - if (ip->i_d.di_mode == 0 && !(flags & XFS_IGET_CREATE)) { - error = ENOENT; - iput(VFS_I(ip)); - goto out_error; + /* We've got a live one. */ + spin_unlock(&ip->i_flags_lock); + read_unlock(&pag->pag_ici_lock); } - /* We've got a live one. */ - read_unlock(&pag->pag_ici_lock); - if (lock_flags != 0) xfs_ilock(ip, lock_flags); @@ -274,6 +276,7 @@ xfs_iget_cache_hit( return 0; out_error: + spin_unlock(&ip->i_flags_lock); read_unlock(&pag->pag_ici_lock); return error; } -- cgit v1.2.3 From ada508274b8698a33cb0e5bd037db0f9dc781795 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Mon, 3 Aug 2009 18:24:21 +0200 Subject: ocfs2: Handle quota file corruption more gracefully ocfs2_read_virt_blocks() does BUG when we try to read a block from a file beyond its end. Since this can happen due to filesystem corruption, it is not really an appropriate answer. Make ocfs2_read_quota_block() check the condition and handle it by calling ocfs2_error() and returning EIO. [ Modified to print ip_blkno in the error - Joel ] Reported-by: Tristan Ye Signed-off-by: Jan Kara Signed-off-by: Joel Becker --- fs/ocfs2/quota_global.c | 10 ++++++++++ 1 file changed, 10 insertions(+) (limited to 'fs') diff --git a/fs/ocfs2/quota_global.c b/fs/ocfs2/quota_global.c index bf7742d0ee3b..44f2a5e1d042 100644 --- a/fs/ocfs2/quota_global.c +++ b/fs/ocfs2/quota_global.c @@ -23,6 +23,7 @@ #include "sysfile.h" #include "dlmglue.h" #include "uptodate.h" +#include "super.h" #include "quota.h" static struct workqueue_struct *ocfs2_quota_wq = NULL; @@ -114,6 +115,15 @@ int ocfs2_read_quota_block(struct inode *inode, u64 v_block, int rc = 0; struct buffer_head *tmp = *bh; + if (i_size_read(inode) >> inode->i_sb->s_blocksize_bits <= v_block) { + ocfs2_error(inode->i_sb, + "Quota file %llu is probably corrupted! Requested " + "to read block %Lu but file has size only %Lu\n", + (unsigned long long)OCFS2_I(inode)->ip_blkno, + (unsigned long long)v_block, + (unsigned long long)i_size_read(inode)); + return -EIO; + } rc = ocfs2_read_virt_blocks(inode, v_block, 1, &tmp, 0, ocfs2_validate_quota_block); if (rc) -- cgit v1.2.3 From 60e2ec48665b8495360ca4a6004c5cd52beb2bc1 Mon Sep 17 00:00:00 2001 From: Tao Ma Date: Wed, 12 Aug 2009 14:42:47 +0800 Subject: ocfs2: release the buffer head in ocfs2_do_truncate. In ocfs2_do_truncate, we forget to release last_eb_bh which will cause memleak. So call brelse in the end. Signed-off-by: Tao Ma Signed-off-by: Joel Becker --- fs/ocfs2/alloc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ocfs2/alloc.c b/fs/ocfs2/alloc.c index f9a3e8942669..ab513ddaeff2 100644 --- a/fs/ocfs2/alloc.c +++ b/fs/ocfs2/alloc.c @@ -6851,7 +6851,7 @@ static int ocfs2_do_truncate(struct ocfs2_super *osb, } status = 0; bail: - + brelse(last_eb_bh); mlog_exit(status); return status; } -- cgit v1.2.3 From eef3a116be11d35396efb2a8cc7345fd3221e294 Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Sun, 16 Aug 2009 21:51:44 -0400 Subject: notify: unused event private race inotify decides if private data it passed to get added to an event was used by checking list_empty(). But it's possible that the event may have been dequeued and the private event removed so it would look empty. The fix is to use the return code from fsnotify_add_notify_event rather than looking at the list. Signed-off-by: Eric Paris Signed-off-by: Linus Torvalds --- fs/notify/inotify/inotify_fsnotify.c | 13 +++++++------ fs/notify/inotify/inotify_user.c | 7 +++---- fs/notify/notification.c | 7 +++---- 3 files changed, 13 insertions(+), 14 deletions(-) (limited to 'fs') diff --git a/fs/notify/inotify/inotify_fsnotify.c b/fs/notify/inotify/inotify_fsnotify.c index 47cd258fd24d..5dcbafe72d71 100644 --- a/fs/notify/inotify/inotify_fsnotify.c +++ b/fs/notify/inotify/inotify_fsnotify.c @@ -62,13 +62,14 @@ static int inotify_handle_event(struct fsnotify_group *group, struct fsnotify_ev event_priv->wd = wd; ret = fsnotify_add_notify_event(group, event, fsn_event_priv); - /* EEXIST is not an error */ - if (ret == -EEXIST) - ret = 0; - - /* did event_priv get attached? */ - if (list_empty(&fsn_event_priv->event_list)) + if (ret) { inotify_free_event_priv(fsn_event_priv); + /* EEXIST says we tail matched, EOVERFLOW isn't something + * to report up the stack. */ + if ((ret == -EEXIST) || + (ret == -EOVERFLOW)) + ret = 0; + } /* * If we hold the entry until after the event is on the queue diff --git a/fs/notify/inotify/inotify_user.c b/fs/notify/inotify/inotify_user.c index f30d9bbc2e1b..c172a7a17b17 100644 --- a/fs/notify/inotify/inotify_user.c +++ b/fs/notify/inotify/inotify_user.c @@ -386,6 +386,7 @@ void inotify_ignored_and_remove_idr(struct fsnotify_mark_entry *entry, struct fsnotify_event *ignored_event; struct inotify_event_private_data *event_priv; struct fsnotify_event_private_data *fsn_event_priv; + int ret; ignored_event = fsnotify_create_event(NULL, FS_IN_IGNORED, NULL, FSNOTIFY_EVENT_NONE, NULL, 0, @@ -404,10 +405,8 @@ void inotify_ignored_and_remove_idr(struct fsnotify_mark_entry *entry, fsn_event_priv->group = group; event_priv->wd = ientry->wd; - fsnotify_add_notify_event(group, ignored_event, fsn_event_priv); - - /* did the private data get added? */ - if (list_empty(&fsn_event_priv->event_list)) + ret = fsnotify_add_notify_event(group, ignored_event, fsn_event_priv); + if (ret) inotify_free_event_priv(fsn_event_priv); skip_send_ignore: diff --git a/fs/notify/notification.c b/fs/notify/notification.c index 521368574e97..74b3cf30bc6b 100644 --- a/fs/notify/notification.c +++ b/fs/notify/notification.c @@ -171,9 +171,7 @@ int fsnotify_add_notify_event(struct fsnotify_group *group, struct fsnotify_even struct list_head *list = &group->notification_list; struct fsnotify_event_holder *last_holder; struct fsnotify_event *last_event; - - /* easy to tell if priv was attached to the event */ - INIT_LIST_HEAD(&priv->event_list); + int ret = 0; /* * There is one fsnotify_event_holder embedded inside each fsnotify_event. @@ -194,6 +192,7 @@ alloc_holder: if (group->q_len >= group->max_events) { event = &q_overflow_event; + ret = -EOVERFLOW; /* sorry, no private data on the overflow event */ priv = NULL; } @@ -235,7 +234,7 @@ alloc_holder: mutex_unlock(&group->notification_mutex); wake_up(&group->notification_waitq); - return 0; + return ret; } /* -- cgit v1.2.3 From cd94c8bbef8d4b796a7ed4c551355a334604fd36 Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Sun, 16 Aug 2009 21:51:49 -0400 Subject: inotify: tail drop inotify q_overflow events In f44aebcc the tail drop logic of events with no file backing (q_overflow and in_ignored) was reversed so IN_IGNORED events would never be tail dropped. This now means that Q_OVERFLOW events are NOT tail dropped. The fix is to not tail drop IN_IGNORED, but to tail drop Q_OVERFLOW. Signed-off-by: Eric Paris Signed-off-by: Linus Torvalds --- fs/notify/notification.c | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'fs') diff --git a/fs/notify/notification.c b/fs/notify/notification.c index 74b3cf30bc6b..3816d5750dd5 100644 --- a/fs/notify/notification.c +++ b/fs/notify/notification.c @@ -153,6 +153,10 @@ static bool event_compare(struct fsnotify_event *old, struct fsnotify_event *new return true; break; case (FSNOTIFY_EVENT_NONE): + if (old->mask & FS_Q_OVERFLOW) + return true; + else if (old->mask & FS_IN_IGNORED) + return false; return false; }; } -- cgit v1.2.3 From 08e53fcb0db34baca3db84a457b6d67faabee4c6 Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Sun, 16 Aug 2009 21:51:55 -0400 Subject: inotify: start watch descriptor count at 1 The inotify_add_watch man page specifies that inotify_add_watch() will return a non-negative integer. However, historically the inotify watches started at 1, not at 0. Turns out that the inotifywait program provided by the inotify-tools package doesn't properly handle a 0 watch descriptor. In 7e790dd5 we changed from starting at 1 to starting at 0. This patch starts at 1, just like in previous kernels, but also just like in previous kernels it's possible for it to wrap back to 0. This preserves the kernel functionality exactly like it was before the patch (neither method broke the spec) Signed-off-by: Eric Paris Signed-off-by: Linus Torvalds --- fs/notify/inotify/inotify_user.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/notify/inotify/inotify_user.c b/fs/notify/inotify/inotify_user.c index c172a7a17b17..dc32ed8323ba 100644 --- a/fs/notify/inotify/inotify_user.c +++ b/fs/notify/inotify/inotify_user.c @@ -567,7 +567,7 @@ static struct fsnotify_group *inotify_new_group(struct user_struct *user, unsign spin_lock_init(&group->inotify_data.idr_lock); idr_init(&group->inotify_data.idr); - group->inotify_data.last_wd = 0; + group->inotify_data.last_wd = 1; group->inotify_data.user = user; group->inotify_data.fa = NULL; -- cgit v1.2.3 From 50fb6d2bd7062708892ae7147f30c3ee905b7a3d Mon Sep 17 00:00:00 2001 From: Abhishek Kulkarni Date: Sun, 19 Jul 2009 13:41:57 -0600 Subject: 9p: Check for error in return value of v9fs_fid_add Check if v9fs_fid_add was successful or not based on its return value. Signed-off-by: Abhishek Kulkarni Signed-off-by: Eric Van Hensbergen --- fs/9p/vfs_inode.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/9p/vfs_inode.c b/fs/9p/vfs_inode.c index 81f8bbf12f9f..1fa5f15eaddc 100644 --- a/fs/9p/vfs_inode.c +++ b/fs/9p/vfs_inode.c @@ -470,7 +470,10 @@ v9fs_create(struct v9fs_session_info *v9ses, struct inode *dir, dentry->d_op = &v9fs_dentry_operations; d_instantiate(dentry, inode); - v9fs_fid_add(dentry, fid); + err = v9fs_fid_add(dentry, fid); + if (err < 0) + goto error; + return ofid; error: -- cgit v1.2.3 From 2bb541157fe2602af7b9952096d0524f6f9c1e73 Mon Sep 17 00:00:00 2001 From: Abhishek Kulkarni Date: Sun, 19 Jul 2009 13:41:56 -0600 Subject: 9p: Fix possible inode leak in v9fs_get_inode. Add a missing iput when cleaning up if v9fs_get_inode fails after returning a valid inode. Signed-off-by: Abhishek Kulkarni Signed-off-by: Eric Van Hensbergen --- fs/9p/vfs_inode.c | 105 +++++++++++++++++++++++++++++------------------------- 1 file changed, 56 insertions(+), 49 deletions(-) (limited to 'fs') diff --git a/fs/9p/vfs_inode.c b/fs/9p/vfs_inode.c index 1fa5f15eaddc..0c8af1abf603 100644 --- a/fs/9p/vfs_inode.c +++ b/fs/9p/vfs_inode.c @@ -207,65 +207,72 @@ v9fs_blank_wstat(struct p9_wstat *wstat) struct inode *v9fs_get_inode(struct super_block *sb, int mode) { + int err; struct inode *inode; struct v9fs_session_info *v9ses = sb->s_fs_info; P9_DPRINTK(P9_DEBUG_VFS, "super block: %p mode: %o\n", sb, mode); inode = new_inode(sb); - if (inode) { - inode->i_mode = mode; - inode->i_uid = current_fsuid(); - inode->i_gid = current_fsgid(); - inode->i_blocks = 0; - inode->i_rdev = 0; - inode->i_atime = inode->i_mtime = inode->i_ctime = CURRENT_TIME; - inode->i_mapping->a_ops = &v9fs_addr_operations; - - switch (mode & S_IFMT) { - case S_IFIFO: - case S_IFBLK: - case S_IFCHR: - case S_IFSOCK: - if (!v9fs_extended(v9ses)) { - P9_DPRINTK(P9_DEBUG_ERROR, - "special files without extended mode\n"); - return ERR_PTR(-EINVAL); - } - init_special_inode(inode, inode->i_mode, - inode->i_rdev); - break; - case S_IFREG: - inode->i_op = &v9fs_file_inode_operations; - inode->i_fop = &v9fs_file_operations; - break; - case S_IFLNK: - if (!v9fs_extended(v9ses)) { - P9_DPRINTK(P9_DEBUG_ERROR, - "extended modes used w/o 9P2000.u\n"); - return ERR_PTR(-EINVAL); - } - inode->i_op = &v9fs_symlink_inode_operations; - break; - case S_IFDIR: - inc_nlink(inode); - if (v9fs_extended(v9ses)) - inode->i_op = &v9fs_dir_inode_operations_ext; - else - inode->i_op = &v9fs_dir_inode_operations; - inode->i_fop = &v9fs_dir_operations; - break; - default: + if (!inode) { + P9_EPRINTK(KERN_WARNING, "Problem allocating inode\n"); + return -ENOMEM; + } + + inode->i_mode = mode; + inode->i_uid = current_fsuid(); + inode->i_gid = current_fsgid(); + inode->i_blocks = 0; + inode->i_rdev = 0; + inode->i_atime = inode->i_mtime = inode->i_ctime = CURRENT_TIME; + inode->i_mapping->a_ops = &v9fs_addr_operations; + + switch (mode & S_IFMT) { + case S_IFIFO: + case S_IFBLK: + case S_IFCHR: + case S_IFSOCK: + if (!v9fs_extended(v9ses)) { P9_DPRINTK(P9_DEBUG_ERROR, - "BAD mode 0x%x S_IFMT 0x%x\n", - mode, mode & S_IFMT); - return ERR_PTR(-EINVAL); + "special files without extended mode\n"); + err = -EINVAL; + goto error; } - } else { - P9_EPRINTK(KERN_WARNING, "Problem allocating inode\n"); - return ERR_PTR(-ENOMEM); + init_special_inode(inode, inode->i_mode, inode->i_rdev); + break; + case S_IFREG: + inode->i_op = &v9fs_file_inode_operations; + inode->i_fop = &v9fs_file_operations; + break; + case S_IFLNK: + if (!v9fs_extended(v9ses)) { + P9_DPRINTK(P9_DEBUG_ERROR, + "extended modes used w/o 9P2000.u\n"); + err = -EINVAL; + goto error; + } + inode->i_op = &v9fs_symlink_inode_operations; + break; + case S_IFDIR: + inc_nlink(inode); + if (v9fs_extended(v9ses)) + inode->i_op = &v9fs_dir_inode_operations_ext; + else + inode->i_op = &v9fs_dir_inode_operations; + inode->i_fop = &v9fs_dir_operations; + break; + default: + P9_DPRINTK(P9_DEBUG_ERROR, "BAD mode 0x%x S_IFMT 0x%x\n", + mode, mode & S_IFMT); + err = -EINVAL; + goto error; } + return inode; + +error: + iput(inode); + return ERR_PTR(err); } /* -- cgit v1.2.3 From 0e15597ebfe00e28857185f46aba00f400480ffe Mon Sep 17 00:00:00 2001 From: Abhishek Kulkarni Date: Sun, 19 Jul 2009 13:41:55 -0600 Subject: 9p: minor comment fixes Fix the comments -- mostly the improper and/or missing descriptions of function parameters. Signed-off-by: Abhishek Kulkarni Signed-off-by: Eric Van Hensbergen --- fs/9p/vfs_inode.c | 3 +-- net/9p/client.c | 14 +++++++------- net/9p/trans_fd.c | 8 ++++---- net/9p/trans_rdma.c | 9 +++++---- net/9p/trans_virtio.c | 11 ++++------- 5 files changed, 21 insertions(+), 24 deletions(-) (limited to 'fs') diff --git a/fs/9p/vfs_inode.c b/fs/9p/vfs_inode.c index 0c8af1abf603..f22668afd0d6 100644 --- a/fs/9p/vfs_inode.c +++ b/fs/9p/vfs_inode.c @@ -171,7 +171,6 @@ int v9fs_uflags2omode(int uflags, int extended) /** * v9fs_blank_wstat - helper function to setup a 9P stat structure - * @v9ses: 9P session info (for determining extended mode) * @wstat: structure to initialize * */ @@ -410,9 +409,9 @@ v9fs_open_created(struct inode *inode, struct file *file) * @v9ses: session information * @dir: directory that dentry is being created in * @dentry: dentry that is being created + * @extension: 9p2000.u extension string to support devices, etc. * @perm: create permissions * @mode: open mode - * @extension: 9p2000.u extension string to support devices, etc. * */ static struct p9_fid * diff --git a/net/9p/client.c b/net/9p/client.c index 787ccddb85ea..7bbd2d5ae8d3 100644 --- a/net/9p/client.c +++ b/net/9p/client.c @@ -60,9 +60,9 @@ static struct p9_req_t * p9_client_rpc(struct p9_client *c, int8_t type, const char *fmt, ...); /** - * v9fs_parse_options - parse mount options into session structure - * @options: options string passed from mount - * @v9ses: existing v9fs session information + * parse_options - parse mount options into client structure + * @opts: options string passed from mount + * @clnt: existing v9fs client information * * Return 0 upon success, -ERRNO upon failure */ @@ -232,7 +232,7 @@ EXPORT_SYMBOL(p9_tag_lookup); /** * p9_tag_init - setup tags structure and contents - * @tags: tags structure from the client struct + * @c: v9fs client struct * * This initializes the tags structure for each client instance. * @@ -258,7 +258,7 @@ error: /** * p9_tag_cleanup - cleans up tags structure and reclaims resources - * @tags: tags structure from the client struct + * @c: v9fs client struct * * This frees resources associated with the tags structure * @@ -430,8 +430,8 @@ static int p9_check_errors(struct p9_client *c, struct p9_req_t *req) /** * p9_client_flush - flush (cancel) a request - * c: client state - * req: request to cancel + * @c: client state + * @oldreq: request to cancel * * This sents a flush for a particular requests and links * the flush request to the original request. The current diff --git a/net/9p/trans_fd.c b/net/9p/trans_fd.c index 8c2588e4edc0..8d934dd7fd54 100644 --- a/net/9p/trans_fd.c +++ b/net/9p/trans_fd.c @@ -119,8 +119,8 @@ struct p9_poll_wait { * @wpos: write position for current frame * @wsize: amount of data to write for current frame * @wbuf: current write buffer + * @poll_pending_link: pending links to be polled per conn * @poll_wait: array of wait_q's for various worker threads - * @poll_waddr: ???? * @pt: poll state * @rq: current read work * @wq: current write work @@ -700,9 +700,9 @@ static int p9_fd_cancel(struct p9_client *client, struct p9_req_t *req) } /** - * parse_options - parse mount options into session structure - * @options: options string passed from mount - * @opts: transport-specific structure to parse options into + * parse_opts - parse mount options into p9_fd_opts structure + * @params: options string passed from mount + * @opts: fd transport-specific structure to parse options into * * Returns 0 upon success, -ERRNO upon failure */ diff --git a/net/9p/trans_rdma.c b/net/9p/trans_rdma.c index ac4990041ebb..65cb29db03f8 100644 --- a/net/9p/trans_rdma.c +++ b/net/9p/trans_rdma.c @@ -67,14 +67,15 @@ * @pd: Protection Domain pointer * @qp: Queue Pair pointer * @cq: Completion Queue pointer + * @dm_mr: DMA Memory Region pointer * @lkey: The local access only memory region key * @timeout: Number of uSecs to wait for connection management events * @sq_depth: The depth of the Send Queue * @sq_sem: Semaphore for the SQ * @rq_depth: The depth of the Receive Queue. + * @rq_count: Count of requests in the Receive Queue. * @addr: The remote peer's address * @req_lock: Protects the active request list - * @send_wait: Wait list when the SQ fills up * @cm_done: Completion event for connection management tracking */ struct p9_trans_rdma { @@ -154,9 +155,9 @@ static match_table_t tokens = { }; /** - * parse_options - parse mount options into session structure - * @options: options string passed from mount - * @opts: transport-specific structure to parse options into + * parse_opts - parse mount options into rdma options structure + * @params: options string passed from mount + * @opts: rdma transport-specific structure to parse options into * * Returns 0 upon success, -ERRNO upon failure */ diff --git a/net/9p/trans_virtio.c b/net/9p/trans_virtio.c index a49484e67e1d..9bf0b737aa51 100644 --- a/net/9p/trans_virtio.c +++ b/net/9p/trans_virtio.c @@ -57,11 +57,9 @@ static int chan_index; * @initialized: whether the channel is initialized * @inuse: whether the channel is in use * @lock: protects multiple elements within this structure + * @client: client instance * @vdev: virtio dev associated with this channel * @vq: virtio queue associated with this channel - * @tagpool: accounting for tag ids (and request slots) - * @reqs: array of request slots - * @max_tag: current number of request_slots allocated * @sg: scatter gather list which is used to pack a request (protected?) * * We keep all per-channel information in a structure. @@ -92,7 +90,7 @@ static unsigned int rest_of_page(void *data) /** * p9_virtio_close - reclaim resources of a channel - * @trans: transport state + * @client: client instance * * This reclaims a channel by freeing its resources and * reseting its inuse flag. @@ -181,9 +179,8 @@ static int p9_virtio_cancel(struct p9_client *client, struct p9_req_t *req) /** * p9_virtio_request - issue a request - * @t: transport state - * @tc: &p9_fcall request to transmit - * @rc: &p9_fcall to put reponse into + * @client: client instance issuing the request + * @req: request to be issued * */ -- cgit v1.2.3 From 02bc35672b2fdf251e264adca5407792f63191e4 Mon Sep 17 00:00:00 2001 From: Abhishek Kulkarni Date: Sun, 19 Jul 2009 13:41:54 -0600 Subject: 9p: Fix possible memleak in v9fs_inode_from fid. Add missing p9stat_free in v9fs_inode_from_fid to avoid any possible leaks. Signed-off-by: Abhishek Kulkarni Signed-off-by: Eric Van Hensbergen --- fs/9p/vfs_inode.c | 13 ++++--------- 1 file changed, 4 insertions(+), 9 deletions(-) (limited to 'fs') diff --git a/fs/9p/vfs_inode.c b/fs/9p/vfs_inode.c index f22668afd0d6..fac30d21851f 100644 --- a/fs/9p/vfs_inode.c +++ b/fs/9p/vfs_inode.c @@ -344,30 +344,25 @@ v9fs_inode_from_fid(struct v9fs_session_info *v9ses, struct p9_fid *fid, ret = NULL; st = p9_client_stat(fid); - if (IS_ERR(st)) { - err = PTR_ERR(st); - st = NULL; - goto error; - } + if (IS_ERR(st)) + return ERR_CAST(st); umode = p9mode2unixmode(v9ses, st->mode); ret = v9fs_get_inode(sb, umode); if (IS_ERR(ret)) { err = PTR_ERR(ret); - ret = NULL; goto error; } v9fs_stat2inode(st, ret, sb); ret->i_ino = v9fs_qid2ino(&st->qid); + p9stat_free(st); kfree(st); return ret; error: + p9stat_free(st); kfree(st); - if (ret) - iput(ret); - return ERR_PTR(err); } -- cgit v1.2.3 From 4f4038328da5eb9cc237b51d3fe68138fd3fea14 Mon Sep 17 00:00:00 2001 From: Abhishek Kulkarni Date: Sun, 19 Jul 2009 13:41:53 -0600 Subject: 9p: Fix v9fs show_options Add the delimiter ',' before the options when they are passed and check if no option parameters are passed to prevent displaying NULL in /proc/mounts. Signed-off-by: Abhishek Kulkarni Signed-off-by: Eric Van Hensbergen --- fs/9p/vfs_super.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/9p/vfs_super.c b/fs/9p/vfs_super.c index 38d695d66a0b..a9d7d08cfbe8 100644 --- a/fs/9p/vfs_super.c +++ b/fs/9p/vfs_super.c @@ -220,8 +220,8 @@ static void v9fs_kill_super(struct super_block *s) static int v9fs_show_options(struct seq_file *m, struct vfsmount *mnt) { struct v9fs_session_info *v9ses = mnt->mnt_sb->s_fs_info; - - seq_printf(m, "%s", v9ses->options); + if (v9ses->options != NULL) + seq_printf(m, ",%s", v9ses->options); return 0; } -- cgit v1.2.3 From 1b5ab3e86712b6be38ebbe0d821387c1d8f91d7c Mon Sep 17 00:00:00 2001 From: Abhishek Kulkarni Date: Sun, 19 Jul 2009 13:41:52 -0600 Subject: 9p: Fix possible regressions when ->get_sb fails. ->get_sb can fail causing some badness. this patch fixes * clear sb->fs_s_info in kill_sb. * deactivate_locked_super() calls kill_sb (v9fs_kill_super) which closes the destroys the client, clunks all its fids and closes the v9fs session. Attempting to do it twice will cause an oops. Signed-off-by: Abhishek Kulkarni Signed-off-by: Eric Van Hensbergen --- fs/9p/vfs_super.c | 11 +++++++---- 1 file changed, 7 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/9p/vfs_super.c b/fs/9p/vfs_super.c index a9d7d08cfbe8..2495af4ad9a6 100644 --- a/fs/9p/vfs_super.c +++ b/fs/9p/vfs_super.c @@ -120,7 +120,6 @@ static int v9fs_get_sb(struct file_system_type *fs_type, int flags, P9_DPRINTK(P9_DEBUG_VFS, " \n"); - st = NULL; v9ses = kzalloc(sizeof(struct v9fs_session_info), GFP_KERNEL); if (!v9ses) return -ENOMEM; @@ -173,10 +172,8 @@ P9_DPRINTK(P9_DEBUG_VFS, " simple set mount, return 0\n"); simple_set_mnt(mnt, sb); return 0; -release_sb: - deactivate_locked_super(sb); - free_stat: + p9stat_free(st); kfree(st); clunk_fid: @@ -185,7 +182,12 @@ clunk_fid: close_session: v9fs_session_close(v9ses); kfree(v9ses); + return retval; +release_sb: + p9stat_free(st); + kfree(st); + deactivate_locked_super(sb); return retval; } @@ -207,6 +209,7 @@ static void v9fs_kill_super(struct super_block *s) v9fs_session_close(v9ses); kfree(v9ses); + s->s_fs_info = NULL; P9_DPRINTK(P9_DEBUG_VFS, "exiting kill_super\n"); } -- cgit v1.2.3 From 4d3297ca5bf37ff5956f76fb352e009880aad62d Mon Sep 17 00:00:00 2001 From: Abhishek Kulkarni Date: Sun, 19 Jul 2009 13:41:51 -0600 Subject: 9p: Remove redundant inode uid/gid assignment Remove a redundant update of inode's i_uid and i_gid after v9fs_get_inode() since the latter already sets up a new inode and sets the proper uid and gid values. Signed-off-by: Abhishek Kulkarni Signed-off-by: Eric Van Hensbergen --- fs/9p/vfs_super.c | 5 ----- 1 file changed, 5 deletions(-) (limited to 'fs') diff --git a/fs/9p/vfs_super.c b/fs/9p/vfs_super.c index 2495af4ad9a6..072dce094477 100644 --- a/fs/9p/vfs_super.c +++ b/fs/9p/vfs_super.c @@ -113,8 +113,6 @@ static int v9fs_get_sb(struct file_system_type *fs_type, int flags, struct v9fs_session_info *v9ses = NULL; struct p9_wstat *st = NULL; int mode = S_IRWXUGO | S_ISVTX; - uid_t uid = current_fsuid(); - gid_t gid = current_fsgid(); struct p9_fid *fid; int retval = 0; @@ -149,9 +147,6 @@ static int v9fs_get_sb(struct file_system_type *fs_type, int flags, goto release_sb; } - inode->i_uid = uid; - inode->i_gid = gid; - root = d_alloc_root(inode); if (!root) { iput(inode); -- cgit v1.2.3 From 5fd131893793567c361ae64cbeb28a2a753bbe35 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Thu, 30 Jul 2009 17:01:53 +0200 Subject: ocfs2: Don't oops in ocfs2_kill_sb on a failed mount If we fail to mount the filesystem, we have to be careful not to dereference uninitialized structures in ocfs2_kill_sb. Signed-off-by: Jan Kara Signed-off-by: Joel Becker --- fs/ocfs2/super.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/ocfs2/super.c b/fs/ocfs2/super.c index b0ee0fdf799a..a3f8871d21fd 100644 --- a/fs/ocfs2/super.c +++ b/fs/ocfs2/super.c @@ -1218,13 +1218,17 @@ static void ocfs2_kill_sb(struct super_block *sb) { struct ocfs2_super *osb = OCFS2_SB(sb); + /* Failed mount? */ + if (!osb || atomic_read(&osb->vol_state) == VOLUME_DISABLED) + goto out; + /* Prevent further queueing of inode drop events */ spin_lock(&dentry_list_lock); ocfs2_set_osb_flag(osb, OCFS2_OSB_DROP_DENTRY_LOCK_IMMED); spin_unlock(&dentry_list_lock); /* Wait for work to finish and/or remove it */ cancel_work_sync(&osb->dentry_lock_work); - +out: kill_block_super(sb); } -- cgit v1.2.3 From 48559b4c30708ebdc849483da9fb83ee08c6c908 Mon Sep 17 00:00:00 2001 From: Abhishek Kulkarni Date: Mon, 17 Aug 2009 16:32:18 -0500 Subject: 9p: Add missing cast for the error return value in v9fs_get_inode Cast the error return value (ENOMEM) in v9fs_get_inode() to its correct type using ERR_PTR. Signed-off-by: Abhishek Kulkarni Signed-off-by: Eric Van Hensbergen --- fs/9p/vfs_inode.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/9p/vfs_inode.c b/fs/9p/vfs_inode.c index fac30d21851f..06a223d50a81 100644 --- a/fs/9p/vfs_inode.c +++ b/fs/9p/vfs_inode.c @@ -215,7 +215,7 @@ struct inode *v9fs_get_inode(struct super_block *sb, int mode) inode = new_inode(sb); if (!inode) { P9_EPRINTK(KERN_WARNING, "Problem allocating inode\n"); - return -ENOMEM; + return ERR_PTR(-ENOMEM); } inode->i_mode = mode; -- cgit v1.2.3 From 4b53e4b500779230aedd5355940aeaaed0b5353b Mon Sep 17 00:00:00 2001 From: Abhishek Kulkarni Date: Mon, 17 Aug 2009 16:42:28 -0500 Subject: 9p: remove unnecessary v9fses->options which duplicates the mount string The mount options string is saved in sb->s_options. This patch removes the redundant duplicating of the mount options. Also, since we are not displaying anything special in show options, we replace v9fs_show_options with generic_show_options for now. Signed-off-by: Abhishek Kulkarni Signed-off-by: Eric Van Hensbergen --- fs/9p/v9fs.c | 21 +++++---------------- fs/9p/v9fs.h | 1 - fs/9p/vfs_super.c | 23 +++++------------------ 3 files changed, 10 insertions(+), 35 deletions(-) (limited to 'fs') diff --git a/fs/9p/v9fs.c b/fs/9p/v9fs.c index 332b5ff02fec..f7003cfac63d 100644 --- a/fs/9p/v9fs.c +++ b/fs/9p/v9fs.c @@ -76,7 +76,7 @@ static const match_table_t tokens = { * Return 0 upon success, -ERRNO upon failure. */ -static int v9fs_parse_options(struct v9fs_session_info *v9ses) +static int v9fs_parse_options(struct v9fs_session_info *v9ses, char *opts) { char *options; substring_t args[MAX_OPT_ARGS]; @@ -90,10 +90,10 @@ static int v9fs_parse_options(struct v9fs_session_info *v9ses) v9ses->debug = 0; v9ses->cache = 0; - if (!v9ses->options) + if (!opts) return 0; - options = kstrdup(v9ses->options, GFP_KERNEL); + options = kstrdup(opts, GFP_KERNEL); if (!options) { P9_DPRINTK(P9_DEBUG_ERROR, "failed to allocate copy of option string\n"); @@ -206,24 +206,14 @@ struct p9_fid *v9fs_session_init(struct v9fs_session_info *v9ses, v9ses->uid = ~0; v9ses->dfltuid = V9FS_DEFUID; v9ses->dfltgid = V9FS_DEFGID; - if (data) { - v9ses->options = kstrdup(data, GFP_KERNEL); - if (!v9ses->options) { - P9_DPRINTK(P9_DEBUG_ERROR, - "failed to allocate copy of option string\n"); - retval = -ENOMEM; - goto error; - } - } - rc = v9fs_parse_options(v9ses); + rc = v9fs_parse_options(v9ses, data); if (rc < 0) { retval = rc; goto error; } - v9ses->clnt = p9_client_create(dev_name, v9ses->options); - + v9ses->clnt = p9_client_create(dev_name, data); if (IS_ERR(v9ses->clnt)) { retval = PTR_ERR(v9ses->clnt); v9ses->clnt = NULL; @@ -280,7 +270,6 @@ void v9fs_session_close(struct v9fs_session_info *v9ses) __putname(v9ses->uname); __putname(v9ses->aname); - kfree(v9ses->options); } /** diff --git a/fs/9p/v9fs.h b/fs/9p/v9fs.h index a7d567192998..38762bf102a9 100644 --- a/fs/9p/v9fs.h +++ b/fs/9p/v9fs.h @@ -85,7 +85,6 @@ struct v9fs_session_info { unsigned int afid; unsigned int cache; - char *options; /* copy of mount options */ char *uname; /* user name to mount as */ char *aname; /* name of remote hierarchy being mounted */ unsigned int maxdata; /* max data for client interface */ diff --git a/fs/9p/vfs_super.c b/fs/9p/vfs_super.c index 072dce094477..8961f1a8f668 100644 --- a/fs/9p/vfs_super.c +++ b/fs/9p/vfs_super.c @@ -81,7 +81,7 @@ static int v9fs_set_super(struct super_block *s, void *data) static void v9fs_fill_super(struct super_block *sb, struct v9fs_session_info *v9ses, - int flags) + int flags, void *data) { sb->s_maxbytes = MAX_LFS_FILESIZE; sb->s_blocksize_bits = fls(v9ses->maxdata - 1); @@ -91,6 +91,8 @@ v9fs_fill_super(struct super_block *sb, struct v9fs_session_info *v9ses, sb->s_flags = flags | MS_ACTIVE | MS_SYNCHRONOUS | MS_DIRSYNC | MS_NOATIME; + + save_mount_options(sb, data); } /** @@ -139,7 +141,7 @@ static int v9fs_get_sb(struct file_system_type *fs_type, int flags, retval = PTR_ERR(sb); goto free_stat; } - v9fs_fill_super(sb, v9ses, flags); + v9fs_fill_super(sb, v9ses, flags, data); inode = v9fs_get_inode(sb, S_IFDIR | mode); if (IS_ERR(inode)) { @@ -208,21 +210,6 @@ static void v9fs_kill_super(struct super_block *s) P9_DPRINTK(P9_DEBUG_VFS, "exiting kill_super\n"); } -/** - * v9fs_show_options - Show mount options in /proc/mounts - * @m: seq_file to write to - * @mnt: mount descriptor - * - */ - -static int v9fs_show_options(struct seq_file *m, struct vfsmount *mnt) -{ - struct v9fs_session_info *v9ses = mnt->mnt_sb->s_fs_info; - if (v9ses->options != NULL) - seq_printf(m, ",%s", v9ses->options); - return 0; -} - static void v9fs_umount_begin(struct super_block *sb) { @@ -235,7 +222,7 @@ v9fs_umount_begin(struct super_block *sb) static const struct super_operations v9fs_super_ops = { .statfs = simple_statfs, .clear_inode = v9fs_clear_inode, - .show_options = v9fs_show_options, + .show_options = generic_show_options, .umount_begin = v9fs_umount_begin, }; -- cgit v1.2.3 From 1154ecbd2f8298ef75609f5f8ed5aca96be599fb Mon Sep 17 00:00:00 2001 From: Zhang Qiang Date: Tue, 18 Aug 2009 14:58:24 +0800 Subject: nilfs2: missing a read lock for segment writer in nilfs_attach_checkpoint() 'ns_cno' of structure 'the_nilfs' must be protected from segment writer, in other words, the caller of nilfs_get_checkpoint should hold read lock for nilfs->ns_segctor_sem. This patch adds the lock/unlock operations in nilfs_attach_checkpoint() when calling nilfs_cpfile_get_checkpoint(). Signed-off-by: Zhang Qiang Signed-off-by: Ryusuke Konishi --- fs/nilfs2/super.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'fs') diff --git a/fs/nilfs2/super.c b/fs/nilfs2/super.c index 8e2ec43b18f4..151964f0de4c 100644 --- a/fs/nilfs2/super.c +++ b/fs/nilfs2/super.c @@ -416,8 +416,10 @@ int nilfs_attach_checkpoint(struct nilfs_sb_info *sbi, __u64 cno) if (unlikely(err)) goto failed; + down_read(&nilfs->ns_segctor_sem); err = nilfs_cpfile_get_checkpoint(nilfs->ns_cpfile, cno, 0, &raw_cp, &bh_cp); + up_read(&nilfs->ns_segctor_sem); if (unlikely(err)) { if (err == -ENOENT || err == -EINVAL) { printk(KERN_ERR -- cgit v1.2.3 From a924586036833086b262a371b09d1266c23bb4d1 Mon Sep 17 00:00:00 2001 From: Ryusuke Konishi Date: Wed, 19 Aug 2009 00:29:43 +0900 Subject: nilfs2: fix oopses with doubly mounted snapshots will fix kernel oopses like the following: # mount -t nilfs2 -r -o cp=20 /dev/sdb1 /test1 # mount -t nilfs2 -r -o cp=20 /dev/sdb1 /test2 # umount /test1 # umount /test2 BUG: sleeping function called from invalid context at arch/x86/mm/fault.c:1069 in_atomic(): 0, irqs_disabled(): 1, pid: 3886, name: umount.nilfs2 1 lock held by umount.nilfs2/3886: #0: (&type->s_umount_key#31){+.+...}, at: [] deactivate_super+0x52/0x6c irq event stamp: 1219 hardirqs last enabled at (1219): [] __mutex_unlock_slowpath+0xf8/0x119 hardirqs last disabled at (1218): [] __mutex_unlock_slowpath+0x59/0x119 softirqs last enabled at (1214): [] __do_softirq+0x1a5/0x1ad softirqs last disabled at (1205): [] do_softirq+0x36/0x5a Pid: 3886, comm: umount.nilfs2 Not tainted 2.6.31-rc6 #55 Call Trace: [] __might_sleep+0x107/0x10e [] do_page_fault+0x246/0x397 [] ? do_page_fault+0x0/0x397 [] error_code+0x6b/0x70 [] ? do_page_fault+0x0/0x397 [] ? __lock_acquire+0x91/0x12fd [] ? __lock_acquire+0x12ee/0x12fd [] ? __lock_acquire+0x12ee/0x12fd [] lock_acquire+0xba/0xdd [] ? nilfs_detach_segment_constructor+0x2f/0x2fa [nilfs2] [] down_write+0x2a/0x46 [] ? nilfs_detach_segment_constructor+0x2f/0x2fa [nilfs2] [] nilfs_detach_segment_constructor+0x2f/0x2fa [nilfs2] [] ? mark_held_locks+0x43/0x5b [] ? trace_hardirqs_on_caller+0x10b/0x133 [] ? trace_hardirqs_on+0xb/0xd [] nilfs_put_super+0x2f/0xca [nilfs2] [] generic_shutdown_super+0x49/0xb8 [] kill_block_super+0x1d/0x31 [] ? vfs_quota_off+0x0/0x12 [] deactivate_super+0x57/0x6c [] mntput_no_expire+0x8c/0xb4 [] sys_umount+0x27f/0x2a4 [] sys_oldumount+0xd/0xf [] sysenter_do_call+0x12/0x38 ... This turns out to be a bug brought by an -rc1 patch ("nilfs2: simplify remaining sget() use"). In the patch, a new "put resource" function, nilfs_put_sbinfo() was introduced to delay freeing nilfs_sb_info struct. But the nilfs_put_sbinfo() mistakenly used atomic_dec_and_test() function to check the reference count, and it caused the nilfs_sb_info was freed when user mounted a snapshot twice. This bug also suggests there was unseen memory leak in usual mount /umount operations for nilfs. Signed-off-by: Ryusuke Konishi --- fs/nilfs2/the_nilfs.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nilfs2/the_nilfs.h b/fs/nilfs2/the_nilfs.h index e8adbffc626f..1b9caafb8662 100644 --- a/fs/nilfs2/the_nilfs.h +++ b/fs/nilfs2/the_nilfs.h @@ -253,7 +253,7 @@ nilfs_detach_writer(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi) static inline void nilfs_put_sbinfo(struct nilfs_sb_info *sbi) { - if (!atomic_dec_and_test(&sbi->s_count)) + if (atomic_dec_and_test(&sbi->s_count)) kfree(sbi); } -- cgit v1.2.3 From 89a4eb4b66e8f4d395e14a14d262dac4d6ca52f0 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Tue, 18 Aug 2009 14:11:08 -0700 Subject: vfs: make get_sb_pseudo set s_maxbytes to value that can be cast to signed get_sb_pseudo sets s_maxbytes to ~0ULL which becomes negative when cast to a signed value. Fix it to use MAX_LFS_FILESIZE which casts properly to a positive signed value. Signed-off-by: Jeff Layton Reviewed-by: Johannes Weiner Acked-by: Steve French Reviewed-by: Christoph Hellwig Cc: Al Viro Cc: Robert Love Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/libfs.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/libfs.c b/fs/libfs.c index ddfa89948c3f..dcec3d3ea64f 100644 --- a/fs/libfs.c +++ b/fs/libfs.c @@ -217,7 +217,7 @@ int get_sb_pseudo(struct file_system_type *fs_type, char *name, return PTR_ERR(s); s->s_flags = MS_NOUSER; - s->s_maxbytes = ~0ULL; + s->s_maxbytes = MAX_LFS_FILESIZE; s->s_blocksize = PAGE_SIZE; s->s_blocksize_bits = PAGE_SHIFT; s->s_magic = magic; -- cgit v1.2.3 From 0753ba01e126020bf0f8150934903b48935b697d Mon Sep 17 00:00:00 2001 From: KOSAKI Motohiro Date: Tue, 18 Aug 2009 14:11:10 -0700 Subject: mm: revert "oom: move oom_adj value" The commit 2ff05b2b (oom: move oom_adj value) moveed the oom_adj value to the mm_struct. It was a very good first step for sanitize OOM. However Paul Menage reported the commit makes regression to his job scheduler. Current OOM logic can kill OOM_DISABLED process. Why? His program has the code of similar to the following. ... set_oom_adj(OOM_DISABLE); /* The job scheduler never killed by oom */ ... if (vfork() == 0) { set_oom_adj(0); /* Invoked child can be killed */ execve("foo-bar-cmd"); } .... vfork() parent and child are shared the same mm_struct. then above set_oom_adj(0) doesn't only change oom_adj for vfork() child, it's also change oom_adj for vfork() parent. Then, vfork() parent (job scheduler) lost OOM immune and it was killed. Actually, fork-setting-exec idiom is very frequently used in userland program. We must not break this assumption. Then, this patch revert commit 2ff05b2b and related commit. Reverted commit list --------------------- - commit 2ff05b2b4e (oom: move oom_adj value from task_struct to mm_struct) - commit 4d8b9135c3 (oom: avoid unnecessary mm locking and scanning for OOM_DISABLE) - commit 8123681022 (oom: only oom kill exiting tasks with attached memory) - commit 933b787b57 (mm: copy over oom_adj value at fork time) Signed-off-by: KOSAKI Motohiro Cc: Paul Menage Cc: David Rientjes Cc: KAMEZAWA Hiroyuki Cc: Rik van Riel Cc: Linus Torvalds Cc: Oleg Nesterov Cc: Nick Piggin Cc: Mel Gorman Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- Documentation/filesystems/proc.txt | 15 +++------ fs/proc/base.c | 19 ++--------- include/linux/mm_types.h | 2 -- include/linux/sched.h | 1 + kernel/fork.c | 1 - mm/oom_kill.c | 64 +++++++++++++++++++++++--------------- 6 files changed, 48 insertions(+), 54 deletions(-) (limited to 'fs') diff --git a/Documentation/filesystems/proc.txt b/Documentation/filesystems/proc.txt index fad18f9456e4..ffead13f9443 100644 --- a/Documentation/filesystems/proc.txt +++ b/Documentation/filesystems/proc.txt @@ -1167,13 +1167,11 @@ CHAPTER 3: PER-PROCESS PARAMETERS 3.1 /proc//oom_adj - Adjust the oom-killer score ------------------------------------------------------ -This file can be used to adjust the score used to select which processes should -be killed in an out-of-memory situation. The oom_adj value is a characteristic -of the task's mm, so all threads that share an mm with pid will have the same -oom_adj value. A high value will increase the likelihood of this process being -killed by the oom-killer. Valid values are in the range -16 to +15 as -explained below and a special value of -17, which disables oom-killing -altogether for threads sharing pid's mm. +This file can be used to adjust the score used to select which processes +should be killed in an out-of-memory situation. Giving it a high score will +increase the likelihood of this process being killed by the oom-killer. Valid +values are in the range -16 to +15, plus the special value -17, which disables +oom-killing altogether for this process. The process to be killed in an out-of-memory situation is selected among all others based on its badness score. This value equals the original memory size of the process @@ -1187,9 +1185,6 @@ the parent's score if they do not share the same memory. Thus forking servers are the prime candidates to be killed. Having only one 'hungry' child will make parent less preferable than the child. -/proc//oom_adj cannot be changed for kthreads since they are immune from -oom-killing already. - /proc//oom_score shows process' current badness score. The following heuristics are then applied: diff --git a/fs/proc/base.c b/fs/proc/base.c index 175db258942f..6f742f6658a9 100644 --- a/fs/proc/base.c +++ b/fs/proc/base.c @@ -1003,12 +1003,7 @@ static ssize_t oom_adjust_read(struct file *file, char __user *buf, if (!task) return -ESRCH; - task_lock(task); - if (task->mm) - oom_adjust = task->mm->oom_adj; - else - oom_adjust = OOM_DISABLE; - task_unlock(task); + oom_adjust = task->oomkilladj; put_task_struct(task); len = snprintf(buffer, sizeof(buffer), "%i\n", oom_adjust); @@ -1037,19 +1032,11 @@ static ssize_t oom_adjust_write(struct file *file, const char __user *buf, task = get_proc_task(file->f_path.dentry->d_inode); if (!task) return -ESRCH; - task_lock(task); - if (!task->mm) { - task_unlock(task); - put_task_struct(task); - return -EINVAL; - } - if (oom_adjust < task->mm->oom_adj && !capable(CAP_SYS_RESOURCE)) { - task_unlock(task); + if (oom_adjust < task->oomkilladj && !capable(CAP_SYS_RESOURCE)) { put_task_struct(task); return -EACCES; } - task->mm->oom_adj = oom_adjust; - task_unlock(task); + task->oomkilladj = oom_adjust; put_task_struct(task); if (end - buffer == 0) return -EIO; diff --git a/include/linux/mm_types.h b/include/linux/mm_types.h index 7acc8439d9b3..0042090a4d70 100644 --- a/include/linux/mm_types.h +++ b/include/linux/mm_types.h @@ -240,8 +240,6 @@ struct mm_struct { unsigned long saved_auxv[AT_VECTOR_SIZE]; /* for /proc/PID/auxv */ - s8 oom_adj; /* OOM kill score adjustment (bit shift) */ - cpumask_t cpu_vm_mask; /* Architecture-specific MM context */ diff --git a/include/linux/sched.h b/include/linux/sched.h index 3ab08e4bb6b8..0f1ea4a66957 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -1198,6 +1198,7 @@ struct task_struct { * a short time */ unsigned char fpu_counter; + s8 oomkilladj; /* OOM kill score adjustment (bit shift). */ #ifdef CONFIG_BLK_DEV_IO_TRACE unsigned int btrace_seq; #endif diff --git a/kernel/fork.c b/kernel/fork.c index 021e1138556e..144326b7af50 100644 --- a/kernel/fork.c +++ b/kernel/fork.c @@ -426,7 +426,6 @@ static struct mm_struct * mm_init(struct mm_struct * mm, struct task_struct *p) init_rwsem(&mm->mmap_sem); INIT_LIST_HEAD(&mm->mmlist); mm->flags = (current->mm) ? current->mm->flags : default_dump_filter; - mm->oom_adj = (current->mm) ? current->mm->oom_adj : 0; mm->core_state = NULL; mm->nr_ptes = 0; set_mm_counter(mm, file_rss, 0); diff --git a/mm/oom_kill.c b/mm/oom_kill.c index 175a67a78a99..a7b2460e922b 100644 --- a/mm/oom_kill.c +++ b/mm/oom_kill.c @@ -58,7 +58,6 @@ unsigned long badness(struct task_struct *p, unsigned long uptime) unsigned long points, cpu_time, run_time; struct mm_struct *mm; struct task_struct *child; - int oom_adj; task_lock(p); mm = p->mm; @@ -66,11 +65,6 @@ unsigned long badness(struct task_struct *p, unsigned long uptime) task_unlock(p); return 0; } - oom_adj = mm->oom_adj; - if (oom_adj == OOM_DISABLE) { - task_unlock(p); - return 0; - } /* * The memory size of the process is the basis for the badness. @@ -154,15 +148,15 @@ unsigned long badness(struct task_struct *p, unsigned long uptime) points /= 8; /* - * Adjust the score by oom_adj. + * Adjust the score by oomkilladj. */ - if (oom_adj) { - if (oom_adj > 0) { + if (p->oomkilladj) { + if (p->oomkilladj > 0) { if (!points) points = 1; - points <<= oom_adj; + points <<= p->oomkilladj; } else - points >>= -(oom_adj); + points >>= -(p->oomkilladj); } #ifdef DEBUG @@ -257,8 +251,11 @@ static struct task_struct *select_bad_process(unsigned long *ppoints, *ppoints = ULONG_MAX; } + if (p->oomkilladj == OOM_DISABLE) + continue; + points = badness(p, uptime.tv_sec); - if (points > *ppoints) { + if (points > *ppoints || !chosen) { chosen = p; *ppoints = points; } @@ -307,7 +304,8 @@ static void dump_tasks(const struct mem_cgroup *mem) } printk(KERN_INFO "[%5d] %5d %5d %8lu %8lu %3d %3d %s\n", p->pid, __task_cred(p)->uid, p->tgid, mm->total_vm, - get_mm_rss(mm), (int)task_cpu(p), mm->oom_adj, p->comm); + get_mm_rss(mm), (int)task_cpu(p), p->oomkilladj, + p->comm); task_unlock(p); } while_each_thread(g, p); } @@ -325,8 +323,11 @@ static void __oom_kill_task(struct task_struct *p, int verbose) return; } - if (!p->mm) + if (!p->mm) { + WARN_ON(1); + printk(KERN_WARNING "tried to kill an mm-less task!\n"); return; + } if (verbose) printk(KERN_ERR "Killed process %d (%s)\n", @@ -348,13 +349,28 @@ static int oom_kill_task(struct task_struct *p) struct mm_struct *mm; struct task_struct *g, *q; - task_lock(p); mm = p->mm; - if (!mm || mm->oom_adj == OOM_DISABLE) { - task_unlock(p); + + /* WARNING: mm may not be dereferenced since we did not obtain its + * value from get_task_mm(p). This is OK since all we need to do is + * compare mm to q->mm below. + * + * Furthermore, even if mm contains a non-NULL value, p->mm may + * change to NULL at any time since we do not hold task_lock(p). + * However, this is of no concern to us. + */ + + if (mm == NULL) return 1; - } - task_unlock(p); + + /* + * Don't kill the process if any threads are set to OOM_DISABLE + */ + do_each_thread(g, q) { + if (q->mm == mm && q->oomkilladj == OOM_DISABLE) + return 1; + } while_each_thread(g, q); + __oom_kill_task(p, 1); /* @@ -377,11 +393,10 @@ static int oom_kill_process(struct task_struct *p, gfp_t gfp_mask, int order, struct task_struct *c; if (printk_ratelimit()) { - task_lock(current); printk(KERN_WARNING "%s invoked oom-killer: " - "gfp_mask=0x%x, order=%d, oom_adj=%d\n", - current->comm, gfp_mask, order, - current->mm ? current->mm->oom_adj : OOM_DISABLE); + "gfp_mask=0x%x, order=%d, oomkilladj=%d\n", + current->comm, gfp_mask, order, current->oomkilladj); + task_lock(current); cpuset_print_task_mems_allowed(current); task_unlock(current); dump_stack(); @@ -394,9 +409,8 @@ static int oom_kill_process(struct task_struct *p, gfp_t gfp_mask, int order, /* * If the task is already exiting, don't alarm the sysadmin or kill * its children or threads, just set TIF_MEMDIE so it can die quickly - * if its mm is still attached. */ - if (p->mm && (p->flags & PF_EXITING)) { + if (p->flags & PF_EXITING) { __oom_kill_task(p, 0); return 0; } -- cgit v1.2.3 From a8b88d3d49623ac701b5dc996cbd61219c793c7c Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Thu, 20 Aug 2009 18:26:52 +0200 Subject: ocfs2: Add missing lock name There is missing name for NFSSync cluster lock. This makes lockdep unhappy because we end up passing NULL to lockdep when initializing lock key. Fix it. Signed-off-by: Jan Kara Signed-off-by: Joel Becker --- fs/ocfs2/ocfs2_lockid.h | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/ocfs2/ocfs2_lockid.h b/fs/ocfs2/ocfs2_lockid.h index fcdba091af3d..c212cf5a2bdf 100644 --- a/fs/ocfs2/ocfs2_lockid.h +++ b/fs/ocfs2/ocfs2_lockid.h @@ -108,6 +108,7 @@ static char *ocfs2_lock_type_strings[] = { [OCFS2_LOCK_TYPE_OPEN] = "Open", [OCFS2_LOCK_TYPE_FLOCK] = "Flock", [OCFS2_LOCK_TYPE_QINFO] = "Quota", + [OCFS2_LOCK_TYPE_NFS_SYNC] = "NFSSync", [OCFS2_LOCK_TYPE_ORPHAN_SCAN] = "OrphanScan", }; -- cgit v1.2.3 From c795b33ba171e41563ab7e25105c0cd4edd81cd7 Mon Sep 17 00:00:00 2001 From: Goldwyn Rodrigues Date: Thu, 20 Aug 2009 13:43:19 -0500 Subject: ocfs2/dlm: Wait on lockres instead of erroring cancel requests In case a downconvert is queued, and a flock receives a signal, BUG_ON(lockres->l_action != OCFS2_AST_INVALID) is triggered because a lock cancel triggers a dlmunlock while an AST is scheduled. To avoid this, allow a LKM_CANCEL to pass through, and let it wait on __dlm_wait_on_lockres(). Signed-off-by: Goldwyn Rodrigues Acked-off-by: Mark Fasheh Signed-off-by: Joel Becker --- fs/ocfs2/dlm/dlmunlock.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/ocfs2/dlm/dlmunlock.c b/fs/ocfs2/dlm/dlmunlock.c index fcf879ed6930..756f5b0998e0 100644 --- a/fs/ocfs2/dlm/dlmunlock.c +++ b/fs/ocfs2/dlm/dlmunlock.c @@ -122,7 +122,7 @@ static enum dlm_status dlmunlock_common(struct dlm_ctxt *dlm, * that still has AST's pending... */ in_use = !list_empty(&lock->ast_list); spin_unlock(&dlm->ast_lock); - if (in_use) { + if (in_use && !(flags & LKM_CANCEL)) { mlog(ML_ERROR, "lockres %.*s: Someone is calling dlmunlock " "while waiting for an ast!", res->lockname.len, res->lockname.name); @@ -131,7 +131,7 @@ static enum dlm_status dlmunlock_common(struct dlm_ctxt *dlm, spin_lock(&res->spinlock); if (res->state & DLM_LOCK_RES_IN_PROGRESS) { - if (master_node) { + if (master_node && !(flags & LKM_CANCEL)) { mlog(ML_ERROR, "lockres in progress!\n"); spin_unlock(&res->spinlock); return DLM_FORWARD; -- cgit v1.2.3 From 03e860bd9f6a3cca747b0795bed26279a8b420a0 Mon Sep 17 00:00:00 2001 From: "From: Nick Piggin" Date: Fri, 21 Aug 2009 10:09:44 +0200 Subject: btrfs: fix inode rbtree corruption Node may not be inserted over existing node. This causes inode tree corruption and I was seeing crashes in inode_tree_del which I can not reproduce after this patch. The other way to fix this would be to tie inode lifetime in the rbtree with inode while not in freeing state. I had a look at this but it is not so trivial at this point. At least this patch gets things working again. Signed-off-by: Nick Piggin Cc: Chris Mason Acked-by: Yan Zheng Signed-off-by: Jens Axboe --- fs/btrfs/inode.c | 21 ++++++++++++++------- 1 file changed, 14 insertions(+), 7 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 272b9b2bea86..59cba180fe83 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -3099,8 +3099,12 @@ static void inode_tree_add(struct inode *inode) { struct btrfs_root *root = BTRFS_I(inode)->root; struct btrfs_inode *entry; - struct rb_node **p = &root->inode_tree.rb_node; - struct rb_node *parent = NULL; + struct rb_node **p; + struct rb_node *parent; + +again: + p = &root->inode_tree.rb_node; + parent = NULL; spin_lock(&root->inode_lock); while (*p) { @@ -3108,13 +3112,16 @@ static void inode_tree_add(struct inode *inode) entry = rb_entry(parent, struct btrfs_inode, rb_node); if (inode->i_ino < entry->vfs_inode.i_ino) - p = &(*p)->rb_left; + p = &parent->rb_left; else if (inode->i_ino > entry->vfs_inode.i_ino) - p = &(*p)->rb_right; + p = &parent->rb_right; else { WARN_ON(!(entry->vfs_inode.i_state & (I_WILL_FREE | I_FREEING | I_CLEAR))); - break; + rb_erase(parent, &root->inode_tree); + RB_CLEAR_NODE(parent); + spin_unlock(&root->inode_lock); + goto again; } } rb_link_node(&BTRFS_I(inode)->rb_node, parent, p); @@ -3126,12 +3133,12 @@ static void inode_tree_del(struct inode *inode) { struct btrfs_root *root = BTRFS_I(inode)->root; + spin_lock(&root->inode_lock); if (!RB_EMPTY_NODE(&BTRFS_I(inode)->rb_node)) { - spin_lock(&root->inode_lock); rb_erase(&BTRFS_I(inode)->rb_node, &root->inode_tree); - spin_unlock(&root->inode_lock); RB_CLEAR_NODE(&BTRFS_I(inode)->rb_node); } + spin_unlock(&root->inode_lock); } static noinline void init_btrfs_i(struct inode *inode) -- cgit v1.2.3 From 8e9d78edea3ce5c0036f85b93091483f2f15443a Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Fri, 21 Aug 2009 17:40:08 -0700 Subject: Re-introduce page mapping check in mark_buffer_dirty() In commit a8e7d49aa7be728c4ae241a75a2a124cdcabc0c5 ("Fix race in create_empty_buffers() vs __set_page_dirty_buffers()"), I removed a test for a NULL page mapping unintentionally when some of the code inside __set_page_dirty() was moved to the callers. That removal generally didn't matter, since a filesystem would serialize truncation (which clears the page mapping) against writing (which marks the buffer dirty), so locking at a higher level (either per-page or an inode at a time) should mean that the buffer page would be stable. And indeed, nothing bad seemed to happen. Except it turns out that apparently reiserfs does something odd when under load and writing out the journal, and we have a number of bugzilla entries that look similar: http://bugzilla.kernel.org/show_bug.cgi?id=13556 http://bugzilla.kernel.org/show_bug.cgi?id=13756 http://bugzilla.kernel.org/show_bug.cgi?id=13876 and it looks like reiserfs depended on that check (the common theme seems to be "data=journal", and a journal writeback during a truncate). I suspect reiserfs should have some additional locking, but in the meantime this should get us back to the pre-2.6.29 behavior. Pattern-pointed-out-by: Roland Kletzing Cc: stable@kernel.org (2.6.29 and 2.6.30) Cc: Jeff Mahoney Cc: Nick Piggin Cc: Al Viro Signed-off-by: Linus Torvalds --- fs/buffer.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/buffer.c b/fs/buffer.c index a3ef091a45bd..28f320fac4d4 100644 --- a/fs/buffer.c +++ b/fs/buffer.c @@ -1165,8 +1165,11 @@ void mark_buffer_dirty(struct buffer_head *bh) if (!test_set_buffer_dirty(bh)) { struct page *page = bh->b_page; - if (!TestSetPageDirty(page)) - __set_page_dirty(page, page_mapping(page), 0); + if (!TestSetPageDirty(page)) { + struct address_space *mapping = page_mapping(page); + if (mapping) + __set_page_dirty(page, mapping, 0); + } } } -- cgit v1.2.3 From 6777d773a463ac045d333b989d4e44660f8d92ad Mon Sep 17 00:00:00 2001 From: Mimi Zohar Date: Fri, 21 Aug 2009 14:32:48 -0400 Subject: kernel_read: redefine offset type vfs_read() offset is defined as loff_t, but kernel_read() offset is only defined as unsigned long. Redefine kernel_read() offset as loff_t. Cc: stable@kernel.org Signed-off-by: Mimi Zohar Signed-off-by: James Morris --- fs/exec.c | 4 ++-- include/linux/fs.h | 2 +- 2 files changed, 3 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/exec.c b/fs/exec.c index 4a8849e45b21..fb4f3cdda78c 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -678,8 +678,8 @@ exit: } EXPORT_SYMBOL(open_exec); -int kernel_read(struct file *file, unsigned long offset, - char *addr, unsigned long count) +int kernel_read(struct file *file, loff_t offset, + char *addr, unsigned long count) { mm_segment_t old_fs; loff_t pos = offset; diff --git a/include/linux/fs.h b/include/linux/fs.h index 67888a9e0655..73e9b643e455 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -2123,7 +2123,7 @@ extern struct file *do_filp_open(int dfd, const char *pathname, int open_flag, int mode, int acc_mode); extern int may_open(struct path *, int, int); -extern int kernel_read(struct file *, unsigned long, char *, unsigned long); +extern int kernel_read(struct file *, loff_t, char *, unsigned long); extern struct file * open_exec(const char *); /* fs/dcache.c -- generic fs support functions */ -- cgit v1.2.3 From 6d41807614151829ae17a3a58bff8572af5e407e Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Mon, 10 Aug 2009 16:03:43 -0400 Subject: ext3: Update Kconfig description of EXT3_DEFAULTS_TO_ORDERED The old description for this configuration option was perhaps not completely balanced in terms of describing the tradeoffs of using a default of data=writeback vs. data=ordered. Despite the fact that old description very strongly recomended disabling this feature, all of the major distributions have elected to preserve the existing 'legacy' default, which is a strong hint that it perhaps wasn't telling the whole story. This revised description has been vetted by a number of ext3 developers as being better at informing the user about the tradeoffs of enabling or disabling this configuration feature. Cc: linux-ext4@vger.kernel.org Signed-off-by: "Theodore Ts'o" Signed-off-by: Jan Kara --- fs/ext3/Kconfig | 32 +++++++++++++++++--------------- 1 file changed, 17 insertions(+), 15 deletions(-) (limited to 'fs') diff --git a/fs/ext3/Kconfig b/fs/ext3/Kconfig index fb3c1a21b135..522b15498f45 100644 --- a/fs/ext3/Kconfig +++ b/fs/ext3/Kconfig @@ -29,23 +29,25 @@ config EXT3_FS module will be called ext3. config EXT3_DEFAULTS_TO_ORDERED - bool "Default to 'data=ordered' in ext3 (legacy option)" + bool "Default to 'data=ordered' in ext3" depends on EXT3_FS help - If a filesystem does not explicitly specify a data ordering - mode, and the journal capability allowed it, ext3 used to - historically default to 'data=ordered'. - - That was a rather unfortunate choice, because it leads to all - kinds of latency problems, and the 'data=writeback' mode is more - appropriate these days. - - You should probably always answer 'n' here, and if you really - want to use 'data=ordered' mode, set it in the filesystem itself - with 'tune2fs -o journal_data_ordered'. - - But if you really want to enable the legacy default, you can do - so by answering 'y' to this question. + The journal mode options for ext3 have different tradeoffs + between when data is guaranteed to be on disk and + performance. The use of "data=writeback" can cause + unwritten data to appear in files after an system crash or + power failure, which can be a security issue. However, + "data=ordered" mode can also result in major performance + problems, including seconds-long delays before an fsync() + call returns. For details, see: + + http://ext4.wiki.kernel.org/index.php/Ext3_data_mode_tradeoffs + + If you have been historically happy with ext3's performance, + data=ordered mode will be a safe choice and you should + answer 'y' here. If you understand the reliability and data + privacy issues of data=writeback and are willing to make + that trade off, answer 'n'. config EXT3_FS_XATTR bool "Ext3 extended attributes" -- cgit v1.2.3 From 3c4cec65274481ec6332b0a91f19b4c8c5394801 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Mon, 24 Aug 2009 16:38:43 +0200 Subject: ext3: Improve error message that changing journaling mode on remount is not possible This patch makes the error message about changing journaling mode on remount more descriptive. Some people are going to hit this error now due to commit bbae8bcc49bc4d002221dab52c79a50a82e7cd1f if they configure a kernel to default to data=writeback mode. The problem happens if they have data=ordered set for the root filesystem in /etc/fstab but not in the kernel command line (and they don't use initrd). Their filesystem then gets mounted as data=writeback by kernel but then their boot fails because init scripts won't be able to remount the filesystem rw. Better error message will hopefully make it easier for them to find the error in their setup and bother us less with error reports :). Signed-off-by: Jan Kara --- fs/ext3/super.c | 40 +++++++++++++++++++++++++++------------- 1 file changed, 27 insertions(+), 13 deletions(-) (limited to 'fs') diff --git a/fs/ext3/super.c b/fs/ext3/super.c index 524b349c6299..a8d80a7f1105 100644 --- a/fs/ext3/super.c +++ b/fs/ext3/super.c @@ -543,6 +543,19 @@ static inline void ext3_show_quota_options(struct seq_file *seq, struct super_bl #endif } +static char *data_mode_string(unsigned long mode) +{ + switch (mode) { + case EXT3_MOUNT_JOURNAL_DATA: + return "journal"; + case EXT3_MOUNT_ORDERED_DATA: + return "ordered"; + case EXT3_MOUNT_WRITEBACK_DATA: + return "writeback"; + } + return "unknown"; +} + /* * Show an option if * - it's set to a non-default value OR @@ -616,13 +629,8 @@ static int ext3_show_options(struct seq_file *seq, struct vfsmount *vfs) if (test_opt(sb, NOBH)) seq_puts(seq, ",nobh"); - if (test_opt(sb, DATA_FLAGS) == EXT3_MOUNT_JOURNAL_DATA) - seq_puts(seq, ",data=journal"); - else if (test_opt(sb, DATA_FLAGS) == EXT3_MOUNT_ORDERED_DATA) - seq_puts(seq, ",data=ordered"); - else if (test_opt(sb, DATA_FLAGS) == EXT3_MOUNT_WRITEBACK_DATA) - seq_puts(seq, ",data=writeback"); - + seq_printf(seq, ",data=%s", data_mode_string(sbi->s_mount_opt & + EXT3_MOUNT_DATA_FLAGS)); if (test_opt(sb, DATA_ERR_ABORT)) seq_puts(seq, ",data_err=abort"); @@ -1024,12 +1032,18 @@ static int parse_options (char *options, struct super_block *sb, datacheck: if (is_remount) { if ((sbi->s_mount_opt & EXT3_MOUNT_DATA_FLAGS) - != data_opt) { - printk(KERN_ERR - "EXT3-fs: cannot change data " - "mode on remount\n"); - return 0; - } + == data_opt) + break; + printk(KERN_ERR + "EXT3-fs (device %s): Cannot change " + "data mode on remount. The filesystem " + "is mounted in data=%s mode and you " + "try to remount it in data=%s mode.\n", + sb->s_id, + data_mode_string(sbi->s_mount_opt & + EXT3_MOUNT_DATA_FLAGS), + data_mode_string(data_opt)); + return 0; } else { sbi->s_mount_opt &= ~EXT3_MOUNT_DATA_FLAGS; sbi->s_mount_opt |= data_opt; -- cgit v1.2.3 From 353d5c30c666580347515da609dd74a2b8e9b828 Mon Sep 17 00:00:00 2001 From: Hugh Dickins Date: Mon, 24 Aug 2009 16:30:28 +0100 Subject: mm: fix hugetlb bug due to user_shm_unlock call 2.6.30's commit 8a0bdec194c21c8fdef840989d0d7b742bb5d4bc removed user_shm_lock() calls in hugetlb_file_setup() but left the user_shm_unlock call in shm_destroy(). In detail: Assume that can_do_hugetlb_shm() returns true and hence user_shm_lock() is not called in hugetlb_file_setup(). However, user_shm_unlock() is called in any case in shm_destroy() and in the following atomic_dec_and_lock(&up->__count) in free_uid() is executed and if up->__count gets zero, also cleanup_user_struct() is scheduled. Note that sched_destroy_user() is empty if CONFIG_USER_SCHED is not set. However, the ref counter up->__count gets unexpectedly non-positive and the corresponding structs are freed even though there are live references to them, resulting in a kernel oops after a lots of shmget(SHM_HUGETLB)/shmctl(IPC_RMID) cycles and CONFIG_USER_SCHED set. Hugh changed Stefan's suggested patch: can_do_hugetlb_shm() at the time of shm_destroy() may give a different answer from at the time of hugetlb_file_setup(). And fixed newseg()'s no_id error path, which has missed user_shm_unlock() ever since it came in 2.6.9. Reported-by: Stefan Huber Signed-off-by: Hugh Dickins Tested-by: Stefan Huber Cc: stable@kernel.org Signed-off-by: Linus Torvalds --- fs/hugetlbfs/inode.c | 20 ++++++++++++-------- include/linux/hugetlb.h | 6 ++++-- ipc/shm.c | 8 +++++--- 3 files changed, 21 insertions(+), 13 deletions(-) (limited to 'fs') diff --git a/fs/hugetlbfs/inode.c b/fs/hugetlbfs/inode.c index 941c8425c10b..cb88dac8ccaa 100644 --- a/fs/hugetlbfs/inode.c +++ b/fs/hugetlbfs/inode.c @@ -935,26 +935,28 @@ static int can_do_hugetlb_shm(void) return capable(CAP_IPC_LOCK) || in_group_p(sysctl_hugetlb_shm_group); } -struct file *hugetlb_file_setup(const char *name, size_t size, int acctflag) +struct file *hugetlb_file_setup(const char *name, size_t size, int acctflag, + struct user_struct **user) { int error = -ENOMEM; - int unlock_shm = 0; struct file *file; struct inode *inode; struct dentry *dentry, *root; struct qstr quick_string; - struct user_struct *user = current_user(); + *user = NULL; if (!hugetlbfs_vfsmount) return ERR_PTR(-ENOENT); if (!can_do_hugetlb_shm()) { - if (user_shm_lock(size, user)) { - unlock_shm = 1; + *user = current_user(); + if (user_shm_lock(size, *user)) { WARN_ONCE(1, "Using mlock ulimits for SHM_HUGETLB deprecated\n"); - } else + } else { + *user = NULL; return ERR_PTR(-EPERM); + } } root = hugetlbfs_vfsmount->mnt_root; @@ -996,8 +998,10 @@ out_inode: out_dentry: dput(dentry); out_shm_unlock: - if (unlock_shm) - user_shm_unlock(size, user); + if (*user) { + user_shm_unlock(size, *user); + *user = NULL; + } return ERR_PTR(error); } diff --git a/include/linux/hugetlb.h b/include/linux/hugetlb.h index 2723513a5651..5cbc620bdfe0 100644 --- a/include/linux/hugetlb.h +++ b/include/linux/hugetlb.h @@ -10,6 +10,7 @@ #include struct ctl_table; +struct user_struct; int PageHuge(struct page *page); @@ -146,7 +147,8 @@ static inline struct hugetlbfs_sb_info *HUGETLBFS_SB(struct super_block *sb) extern const struct file_operations hugetlbfs_file_operations; extern struct vm_operations_struct hugetlb_vm_ops; -struct file *hugetlb_file_setup(const char *name, size_t, int); +struct file *hugetlb_file_setup(const char *name, size_t size, int acct, + struct user_struct **user); int hugetlb_get_quota(struct address_space *mapping, long delta); void hugetlb_put_quota(struct address_space *mapping, long delta); @@ -168,7 +170,7 @@ static inline void set_file_hugepages(struct file *file) #define is_file_hugepages(file) 0 #define set_file_hugepages(file) BUG() -#define hugetlb_file_setup(name,size,acctflag) ERR_PTR(-ENOSYS) +#define hugetlb_file_setup(name,size,acct,user) ERR_PTR(-ENOSYS) #endif /* !CONFIG_HUGETLBFS */ diff --git a/ipc/shm.c b/ipc/shm.c index 15dd238e5338..1bc4701ef4f0 100644 --- a/ipc/shm.c +++ b/ipc/shm.c @@ -174,7 +174,7 @@ static void shm_destroy(struct ipc_namespace *ns, struct shmid_kernel *shp) shm_unlock(shp); if (!is_file_hugepages(shp->shm_file)) shmem_lock(shp->shm_file, 0, shp->mlock_user); - else + else if (shp->mlock_user) user_shm_unlock(shp->shm_file->f_path.dentry->d_inode->i_size, shp->mlock_user); fput (shp->shm_file); @@ -369,8 +369,8 @@ static int newseg(struct ipc_namespace *ns, struct ipc_params *params) /* hugetlb_file_setup applies strict accounting */ if (shmflg & SHM_NORESERVE) acctflag = VM_NORESERVE; - file = hugetlb_file_setup(name, size, acctflag); - shp->mlock_user = current_user(); + file = hugetlb_file_setup(name, size, acctflag, + &shp->mlock_user); } else { /* * Do not allow no accounting for OVERCOMMIT_NEVER, even @@ -410,6 +410,8 @@ static int newseg(struct ipc_namespace *ns, struct ipc_params *params) return error; no_id: + if (shp->mlock_user) /* shmflg & SHM_HUGETLB case */ + user_shm_unlock(size, shp->mlock_user); fput(file); no_file: security_shm_free(shp); -- cgit v1.2.3 From 7111dc73923e9737b38a3ef5b5f236109000ff28 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Mon, 24 Aug 2009 19:21:29 -0400 Subject: NFSv4: Fix an infinite looping problem with the nfs4_state_manager Commit 76db6d9500caeaa774a3e32a997eba30bbdc176b (nfs41: add session setup to the state manager) introduces an infinite loop possibility in the NFSv4 state manager. By first checking nfs4_has_session() before clearing the NFS4CLNT_SESSION_SETUP flag, it allows for a situation where someone sets that flag, but it never gets cleared, and so the state manager loops. In fact commit c3fad1b1aaf850bf692642642ace7cd0d64af0a3 (nfs41: add session reset to state manager) causes this to happen every time we get a network partition error. Signed-off-by: Trond Myklebust Tested-by: Daniel J Blueman Signed-off-by: Linus Torvalds --- fs/nfs/nfs4state.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c index 65ca8c18476f..1434080aefeb 100644 --- a/fs/nfs/nfs4state.c +++ b/fs/nfs/nfs4state.c @@ -1250,8 +1250,8 @@ static void nfs4_state_manager(struct nfs_client *clp) continue; } /* Initialize or reset the session */ - if (nfs4_has_session(clp) && - test_and_clear_bit(NFS4CLNT_SESSION_SETUP, &clp->cl_state)) { + if (test_and_clear_bit(NFS4CLNT_SESSION_SETUP, &clp->cl_state) + && nfs4_has_session(clp)) { if (clp->cl_cons_state == NFS_CS_SESSION_INITING) status = nfs4_initialize_session(clp); else -- cgit v1.2.3 From 52cef7555adf5ca09b3b7283097466759120d901 Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Mon, 24 Aug 2009 16:03:35 -0400 Subject: inotify: seperate new watch creation updating existing watches There is nothing known wrong with the inotify watch addition/modification but this patch seperates the two code paths to make them each easy to verify as correct. Signed-off-by: Eric Paris --- fs/notify/inotify/inotify_user.c | 172 +++++++++++++++++++++++---------------- 1 file changed, 103 insertions(+), 69 deletions(-) (limited to 'fs') diff --git a/fs/notify/inotify/inotify_user.c b/fs/notify/inotify/inotify_user.c index dc32ed8323ba..d8f73c253073 100644 --- a/fs/notify/inotify/inotify_user.c +++ b/fs/notify/inotify/inotify_user.c @@ -431,80 +431,29 @@ static void inotify_free_mark(struct fsnotify_mark_entry *entry) kmem_cache_free(inotify_inode_mark_cachep, ientry); } -static int inotify_update_watch(struct fsnotify_group *group, struct inode *inode, u32 arg) +static int inotify_update_existing_watch(struct fsnotify_group *group, + struct inode *inode, + u32 arg) { - struct fsnotify_mark_entry *entry = NULL; + struct fsnotify_mark_entry *entry; struct inotify_inode_mark_entry *ientry; - struct inotify_inode_mark_entry *tmp_ientry; - int ret = 0; - int add = (arg & IN_MASK_ADD); - __u32 mask; __u32 old_mask, new_mask; + __u32 mask; + int add = (arg & IN_MASK_ADD); + int ret; /* don't allow invalid bits: we don't want flags set */ mask = inotify_arg_to_mask(arg); if (unlikely(!mask)) return -EINVAL; - tmp_ientry = kmem_cache_alloc(inotify_inode_mark_cachep, GFP_KERNEL); - if (unlikely(!tmp_ientry)) - return -ENOMEM; - /* we set the mask at the end after attaching it */ - fsnotify_init_mark(&tmp_ientry->fsn_entry, inotify_free_mark); - tmp_ientry->wd = -1; - -find_entry: spin_lock(&inode->i_lock); entry = fsnotify_find_mark_entry(group, inode); spin_unlock(&inode->i_lock); - if (entry) { - ientry = container_of(entry, struct inotify_inode_mark_entry, fsn_entry); - } else { - ret = -ENOSPC; - if (atomic_read(&group->inotify_data.user->inotify_watches) >= inotify_max_user_watches) - goto out_err; -retry: - ret = -ENOMEM; - if (unlikely(!idr_pre_get(&group->inotify_data.idr, GFP_KERNEL))) - goto out_err; - - spin_lock(&group->inotify_data.idr_lock); - ret = idr_get_new_above(&group->inotify_data.idr, &tmp_ientry->fsn_entry, - group->inotify_data.last_wd, - &tmp_ientry->wd); - spin_unlock(&group->inotify_data.idr_lock); - if (ret) { - if (ret == -EAGAIN) - goto retry; - goto out_err; - } + if (!entry) + return -ENOENT; - ret = fsnotify_add_mark(&tmp_ientry->fsn_entry, group, inode); - if (ret) { - inotify_remove_from_idr(group, tmp_ientry); - if (ret == -EEXIST) - goto find_entry; - goto out_err; - } - - /* tmp_ientry has been added to the inode, so we are all set up. - * now we just need to make sure tmp_ientry doesn't get freed and - * we need to set up entry and ientry so the generic code can - * do its thing. */ - ientry = tmp_ientry; - entry = &ientry->fsn_entry; - tmp_ientry = NULL; - - atomic_inc(&group->inotify_data.user->inotify_watches); - - /* update the idr hint */ - group->inotify_data.last_wd = ientry->wd; - - /* we put the mark on the idr, take a reference */ - fsnotify_get_mark(entry); - } - - ret = ientry->wd; + ientry = container_of(entry, struct inotify_inode_mark_entry, fsn_entry); spin_lock(&entry->lock); @@ -536,18 +485,103 @@ retry: fsnotify_recalc_group_mask(group); } - /* this either matches fsnotify_find_mark_entry, or init_mark_entry - * depending on which path we took... */ + /* return the wd */ + ret = ientry->wd; + + /* match the get from fsnotify_find_mark_entry() */ fsnotify_put_mark(entry); + return ret; +} + +static int inotify_new_watch(struct fsnotify_group *group, + struct inode *inode, + u32 arg) +{ + struct inotify_inode_mark_entry *tmp_ientry; + __u32 mask; + int ret; + + /* don't allow invalid bits: we don't want flags set */ + mask = inotify_arg_to_mask(arg); + if (unlikely(!mask)) + return -EINVAL; + + tmp_ientry = kmem_cache_alloc(inotify_inode_mark_cachep, GFP_KERNEL); + if (unlikely(!tmp_ientry)) + return -ENOMEM; + + fsnotify_init_mark(&tmp_ientry->fsn_entry, inotify_free_mark); + tmp_ientry->fsn_entry.mask = mask; + tmp_ientry->wd = -1; + + ret = -ENOSPC; + if (atomic_read(&group->inotify_data.user->inotify_watches) >= inotify_max_user_watches) + goto out_err; +retry: + ret = -ENOMEM; + if (unlikely(!idr_pre_get(&group->inotify_data.idr, GFP_KERNEL))) + goto out_err; + + spin_lock(&group->inotify_data.idr_lock); + ret = idr_get_new_above(&group->inotify_data.idr, &tmp_ientry->fsn_entry, + group->inotify_data.last_wd, + &tmp_ientry->wd); + spin_unlock(&group->inotify_data.idr_lock); + if (ret) { + /* idr was out of memory allocate and try again */ + if (ret == -EAGAIN) + goto retry; + goto out_err; + } + + /* we are on the idr, now get on the inode */ + ret = fsnotify_add_mark(&tmp_ientry->fsn_entry, group, inode); + if (ret) { + /* we failed to get on the inode, get off the idr */ + inotify_remove_from_idr(group, tmp_ientry); + goto out_err; + } + + /* we put the mark on the idr, take a reference */ + fsnotify_get_mark(&tmp_ientry->fsn_entry); + + /* update the idr hint, who cares about races, it's just a hint */ + group->inotify_data.last_wd = tmp_ientry->wd; + + /* increment the number of watches the user has */ + atomic_inc(&group->inotify_data.user->inotify_watches); + + /* return the watch descriptor for this new entry */ + ret = tmp_ientry->wd; + + /* match the ref from fsnotify_init_markentry() */ + fsnotify_put_mark(&tmp_ientry->fsn_entry); + out_err: - /* could be an error, could be that we found an existing mark */ - if (tmp_ientry) { - /* on the idr but didn't make it on the inode */ - if (tmp_ientry->wd != -1) - inotify_remove_from_idr(group, tmp_ientry); + if (ret < 0) kmem_cache_free(inotify_inode_mark_cachep, tmp_ientry); - } + + return ret; +} + +static int inotify_update_watch(struct fsnotify_group *group, struct inode *inode, u32 arg) +{ + int ret = 0; + +retry: + /* try to update and existing watch with the new arg */ + ret = inotify_update_existing_watch(group, inode, arg); + /* no mark present, try to add a new one */ + if (ret == -ENOENT) + ret = inotify_new_watch(group, inode, arg); + /* + * inotify_new_watch could race with another thread which did an + * inotify_new_watch between the update_existing and the add watch + * here, go back and try to update an existing mark again. + */ + if (ret == -EEXIST) + goto retry; return ret; } -- cgit v1.2.3 From cf4374267fbe966e8e4e7db68f5dc7b267439780 Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Mon, 24 Aug 2009 16:03:35 -0400 Subject: inotify: do not BUG on idr entries at inotify destruction If an inotify watch is left in the idr when an fsnotify group is destroyed this will lead to a BUG. This is not a dangerous situation and really indicates a programming bug and leak of memory. This patch changes it to use a WARN and a printk rather than killing people's boxes. Signed-off-by: Eric Paris --- fs/notify/inotify/inotify_fsnotify.c | 33 +++++++++++++++++++++++++++++++-- 1 file changed, 31 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/notify/inotify/inotify_fsnotify.c b/fs/notify/inotify/inotify_fsnotify.c index 5dcbafe72d71..c9ee67b442e1 100644 --- a/fs/notify/inotify/inotify_fsnotify.c +++ b/fs/notify/inotify/inotify_fsnotify.c @@ -105,16 +105,45 @@ static bool inotify_should_send_event(struct fsnotify_group *group, struct inode return send; } +/* + * This is NEVER supposed to be called. Inotify marks should either have been + * removed from the idr when the watch was removed or in the + * fsnotify_destroy_mark_by_group() call when the inotify instance was being + * torn down. This is only called if the idr is about to be freed but there + * are still marks in it. + */ static int idr_callback(int id, void *p, void *data) { - BUG(); + struct fsnotify_mark_entry *entry; + struct inotify_inode_mark_entry *ientry; + static bool warned = false; + + if (warned) + return 0; + + warned = false; + entry = p; + ientry = container_of(entry, struct inotify_inode_mark_entry, fsn_entry); + + WARN(1, "inotify closing but id=%d for entry=%p in group=%p still in " + "idr. Probably leaking memory\n", id, p, data); + + /* + * I'm taking the liberty of assuming that the mark in question is a + * valid address and I'm dereferencing it. This might help to figure + * out why we got here and the panic is no worse than the original + * BUG() that was here. + */ + if (entry) + printk(KERN_WARNING "entry->group=%p inode=%p wd=%d\n", + entry->group, entry->inode, ientry->wd); return 0; } static void inotify_free_group_priv(struct fsnotify_group *group) { /* ideally the idr is empty and we won't hit the BUG in teh callback */ - idr_for_each(&group->inotify_data.idr, idr_callback, NULL); + idr_for_each(&group->inotify_data.idr, idr_callback, group); idr_remove_all(&group->inotify_data.idr); idr_destroy(&group->inotify_data.idr); } -- cgit v1.2.3 From dead537dd8a1c9495322c1d6f7c780697f474af0 Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Mon, 24 Aug 2009 16:03:35 -0400 Subject: inotify: fix locking around inotify watching in the idr The are races around the idr storage of inotify watches. It's possible that a watch could be found from sys_inotify_rm_watch() in the idr, but it could be removed from the idr before that code does it's removal. Move the locking and the refcnt'ing so that these have to happen atomically. Signed-off-by: Eric Paris --- fs/notify/inotify/inotify_user.c | 50 ++++++++++++++++++++++++++++++++-------- 1 file changed, 40 insertions(+), 10 deletions(-) (limited to 'fs') diff --git a/fs/notify/inotify/inotify_user.c b/fs/notify/inotify/inotify_user.c index d8f73c253073..ce1f5823e2c2 100644 --- a/fs/notify/inotify/inotify_user.c +++ b/fs/notify/inotify/inotify_user.c @@ -364,20 +364,53 @@ static int inotify_find_inode(const char __user *dirname, struct path *path, uns return error; } +/* + * Remove the mark from the idr (if present) and drop the reference + * on the mark because it was in the idr. + */ static void inotify_remove_from_idr(struct fsnotify_group *group, struct inotify_inode_mark_entry *ientry) { struct idr *idr; + struct fsnotify_mark_entry *entry; + struct inotify_inode_mark_entry *found_ientry; + int wd; spin_lock(&group->inotify_data.idr_lock); idr = &group->inotify_data.idr; - idr_remove(idr, ientry->wd); - spin_unlock(&group->inotify_data.idr_lock); + wd = ientry->wd; + + if (wd == -1) + goto out; + + entry = idr_find(&group->inotify_data.idr, wd); + if (unlikely(!entry)) + goto out; + + found_ientry = container_of(entry, struct inotify_inode_mark_entry, fsn_entry); + if (unlikely(found_ientry != ientry)) { + /* We found an entry in the idr with the right wd, but it's + * not the entry we were told to remove. eparis seriously + * fucked up somewhere. */ + WARN_ON(1); + ientry->wd = -1; + goto out; + } + + /* One ref for being in the idr, one ref held by the caller */ + BUG_ON(atomic_read(&entry->refcnt) < 2); + + idr_remove(idr, wd); ientry->wd = -1; + + /* removed from the idr, drop that ref */ + fsnotify_put_mark(entry); +out: + spin_unlock(&group->inotify_data.idr_lock); } + /* - * Send IN_IGNORED for this wd, remove this wd from the idr, and drop the - * internal reference help on the mark because it is in the idr. + * Send IN_IGNORED for this wd, remove this wd from the idr. */ void inotify_ignored_and_remove_idr(struct fsnotify_mark_entry *entry, struct fsnotify_group *group) @@ -417,9 +450,6 @@ skip_send_ignore: /* remove this entry from the idr */ inotify_remove_from_idr(group, ientry); - /* removed from idr, drop that reference */ - fsnotify_put_mark(entry); - atomic_dec(&group->inotify_data.user->inotify_watches); } @@ -535,6 +565,9 @@ retry: goto out_err; } + /* we put the mark on the idr, take a reference */ + fsnotify_get_mark(&tmp_ientry->fsn_entry); + /* we are on the idr, now get on the inode */ ret = fsnotify_add_mark(&tmp_ientry->fsn_entry, group, inode); if (ret) { @@ -543,9 +576,6 @@ retry: goto out_err; } - /* we put the mark on the idr, take a reference */ - fsnotify_get_mark(&tmp_ientry->fsn_entry); - /* update the idr hint, who cares about races, it's just a hint */ group->inotify_data.last_wd = tmp_ientry->wd; -- cgit v1.2.3 From 0db501bd0610ee0c0aca84d927f90bcccd09e2bd Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Thu, 27 Aug 2009 03:20:04 -0700 Subject: inotify: Ensure we alwasy write the terminating NULL. Before the rewrite copy_event_to_user always wrote a terqminating '\0' byte to user space after the filename. Since the rewrite that terminating byte was skipped if your filename is exactly a multiple of event_size. Ouch! So add one byte to name_size before we round up and use clear_user to set userspace to zero like /dev/zero does instead of copying the strange nul_inotify_event. I can't quite convince myself len_to_zero will never exceed 16 and even if it doesn't clear_user should be more efficient and a more accurate reflection of what the code is trying to do. Signed-off-by: Eric W. Biederman Signed-off-by: Eric Paris --- fs/notify/inotify/inotify_user.c | 13 ++++++------- 1 file changed, 6 insertions(+), 7 deletions(-) (limited to 'fs') diff --git a/fs/notify/inotify/inotify_user.c b/fs/notify/inotify/inotify_user.c index ce1f5823e2c2..0e781bc88d1e 100644 --- a/fs/notify/inotify/inotify_user.c +++ b/fs/notify/inotify/inotify_user.c @@ -47,9 +47,6 @@ static struct vfsmount *inotify_mnt __read_mostly; -/* this just sits here and wastes global memory. used to just pad userspace messages with zeros */ -static struct inotify_event nul_inotify_event; - /* these are configurable via /proc/sys/fs/inotify/ */ static int inotify_max_user_instances __read_mostly; static int inotify_max_queued_events __read_mostly; @@ -199,8 +196,10 @@ static ssize_t copy_event_to_user(struct fsnotify_group *group, inotify_free_event_priv(fsn_priv); } - /* round up event->name_len so it is a multiple of event_size */ - name_len = roundup(event->name_len, event_size); + /* round up event->name_len so it is a multiple of event_size + * plus an extra byte for the terminating '\0'. + */ + name_len = roundup(event->name_len + 1, event_size); inotify_event.len = name_len; inotify_event.mask = inotify_mask_to_arg(event->mask); @@ -224,8 +223,8 @@ static ssize_t copy_event_to_user(struct fsnotify_group *group, return -EFAULT; buf += event->name_len; - /* fill userspace with 0's from nul_inotify_event */ - if (copy_to_user(buf, &nul_inotify_event, len_to_zero)) + /* fill userspace with 0's */ + if (clear_user(buf, len_to_zero)) return -EFAULT; buf += len_to_zero; event_size += name_len; -- cgit v1.2.3 From 9886e836a6a5dbd273dc55b17e713f0a188d137f Mon Sep 17 00:00:00 2001 From: David Howells Date: Thu, 27 Aug 2009 13:09:06 +0100 Subject: AFS: Stop readlink() on AFS crashing due to NULL 'file' ptr kAFS crashes when asked to read a symbolic link because page_getlink() passes a NULL file pointer to read_mapping_page(), but afs_readpage() expects a file pointer from which to extract a key. Modify afs_readpage() to request the appropriate key from the calling process's keyrings if a file struct is not supplied with one attached. Signed-off-by: David Howells Acked-by: Anton Blanchard Signed-off-by: Linus Torvalds --- fs/afs/file.c | 18 +++++++++++++++--- 1 file changed, 15 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/afs/file.c b/fs/afs/file.c index 0149dab365e7..681c2a7b013f 100644 --- a/fs/afs/file.c +++ b/fs/afs/file.c @@ -134,9 +134,16 @@ static int afs_readpage(struct file *file, struct page *page) inode = page->mapping->host; - ASSERT(file != NULL); - key = file->private_data; - ASSERT(key != NULL); + if (file) { + key = file->private_data; + ASSERT(key != NULL); + } else { + key = afs_request_key(AFS_FS_S(inode->i_sb)->volume->cell); + if (IS_ERR(key)) { + ret = PTR_ERR(key); + goto error_nokey; + } + } _enter("{%x},{%lu},{%lu}", key_serial(key), inode->i_ino, page->index); @@ -207,12 +214,17 @@ static int afs_readpage(struct file *file, struct page *page) unlock_page(page); } + if (!file) + key_put(key); _leave(" = 0"); return 0; error: SetPageError(page); unlock_page(page); + if (!file) + key_put(key); +error_nokey: _leave(" = %d", ret); return ret; } -- cgit v1.2.3 From a1b08e75dff3dc18a88444803753e667bb1d126e Mon Sep 17 00:00:00 2001 From: Tao Ma Date: Thu, 27 Aug 2009 14:46:56 +0800 Subject: ocfs2: invalidate dentry if its dentry_lock isn't initialized. In commit a5a0a630922a2f6a774b6dac19f70cb5abd86bb0, when ocfs2_attch_dentry_lock fails, we call an extra iput and reset dentry->d_fsdata to NULL. This resolve a bug, but it isn't completed and the dentry is still there. When we want to use it again, ocfs2_dentry_revalidate doesn't catch it and return true. That make future ocfs2_dentry_lock panic out. One bug is http://oss.oracle.com/bugzilla/show_bug.cgi?id=1162. The resolution is to add a check for dentry->d_fsdata in revalidate process and return false if dentry->d_fsdata is NULL, so that a new ocfs2_lookup will be called again. Signed-off-by: Tao Ma Signed-off-by: Joel Becker --- fs/ocfs2/dcache.c | 11 +++++++++++ 1 file changed, 11 insertions(+) (limited to 'fs') diff --git a/fs/ocfs2/dcache.c b/fs/ocfs2/dcache.c index 2f28b7de2c8d..b4957c7d9fe2 100644 --- a/fs/ocfs2/dcache.c +++ b/fs/ocfs2/dcache.c @@ -85,6 +85,17 @@ static int ocfs2_dentry_revalidate(struct dentry *dentry, goto bail; } + /* + * If the last lookup failed to create dentry lock, let us + * redo it. + */ + if (!dentry->d_fsdata) { + mlog(0, "Inode %llu doesn't have dentry lock, " + "returning false\n", + (unsigned long long)OCFS2_I(inode)->ip_blkno); + goto bail; + } + ret = 1; bail: -- cgit v1.2.3 From b962e7312ae87006aed6f68ceee94bdf8db08338 Mon Sep 17 00:00:00 2001 From: Brian Rogers Date: Fri, 28 Aug 2009 10:00:05 -0400 Subject: inotify: do not send a block of zeros when no pathname is available When an event has no pathname, there's no need to pad it with a null byte and therefore generate an inotify_event sized block of zeros. This fixes a regression introduced by commit 0db501bd0610ee0c0aca84d927f90bcccd09e2bd where my system wouldn't finish booting because some process was being confused by this. Signed-off-by: Brian Rogers Signed-off-by: Eric Paris --- fs/notify/inotify/inotify_user.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/notify/inotify/inotify_user.c b/fs/notify/inotify/inotify_user.c index 0e781bc88d1e..b547ae17b461 100644 --- a/fs/notify/inotify/inotify_user.c +++ b/fs/notify/inotify/inotify_user.c @@ -180,7 +180,7 @@ static ssize_t copy_event_to_user(struct fsnotify_group *group, struct fsnotify_event_private_data *fsn_priv; struct inotify_event_private_data *priv; size_t event_size = sizeof(struct inotify_event); - size_t name_len; + size_t name_len = 0; /* we get the inotify watch descriptor from the event private data */ spin_lock(&event->lock); @@ -196,10 +196,12 @@ static ssize_t copy_event_to_user(struct fsnotify_group *group, inotify_free_event_priv(fsn_priv); } - /* round up event->name_len so it is a multiple of event_size + /* + * round up event->name_len so it is a multiple of event_size * plus an extra byte for the terminating '\0'. */ - name_len = roundup(event->name_len + 1, event_size); + if (event->name_len) + name_len = roundup(event->name_len + 1, event_size); inotify_event.len = name_len; inotify_event.mask = inotify_mask_to_arg(event->mask); -- cgit v1.2.3 From 83cb10f0ef3c96162be92339ccf8c0c9c9f2d13e Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Fri, 28 Aug 2009 11:57:55 -0400 Subject: inotify: fix length reporting and size checking 0db501bd0610ee0c0 introduced a regresion in that it now sends a nul terminator but the length accounting when checking for space or reporting to userspace did not take this into account. This corrects all of the rounding logic. Signed-off-by: Eric Paris --- fs/notify/inotify/inotify_user.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/notify/inotify/inotify_user.c b/fs/notify/inotify/inotify_user.c index b547ae17b461..6111670b2573 100644 --- a/fs/notify/inotify/inotify_user.c +++ b/fs/notify/inotify/inotify_user.c @@ -154,7 +154,8 @@ static struct fsnotify_event *get_one_event(struct fsnotify_group *group, event = fsnotify_peek_notify_event(group); - event_size += roundup(event->name_len, event_size); + if (event->name_len) + event_size += roundup(event->name_len + 1, event_size); if (event_size > count) return ERR_PTR(-EINVAL); @@ -327,8 +328,9 @@ static long inotify_ioctl(struct file *file, unsigned int cmd, list_for_each_entry(holder, &group->notification_list, event_list) { event = holder->event; send_len += sizeof(struct inotify_event); - send_len += roundup(event->name_len, - sizeof(struct inotify_event)); + if (event->name_len) + send_len += roundup(event->name_len + 1, + sizeof(struct inotify_event)); } mutex_unlock(&group->notification_mutex); ret = put_user(send_len, (int __user *) p); -- cgit v1.2.3 From 750a8870fe4016ef3091fc97e084d58c613c2cc7 Mon Sep 17 00:00:00 2001 From: Eric Paris Date: Fri, 28 Aug 2009 12:50:47 -0400 Subject: inotify: update the group mask on mark addition Seperating the addition and update of marks in inotify resulted in a regression in that inotify never gets events. The inotify group mask is always 0. This mask should be updated any time a new mark is added. Signed-off-by: Eric Paris --- fs/notify/inotify/inotify_user.c | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'fs') diff --git a/fs/notify/inotify/inotify_user.c b/fs/notify/inotify/inotify_user.c index 6111670b2573..dcd2040d330c 100644 --- a/fs/notify/inotify/inotify_user.c +++ b/fs/notify/inotify/inotify_user.c @@ -591,6 +591,10 @@ retry: /* match the ref from fsnotify_init_markentry() */ fsnotify_put_mark(&tmp_ientry->fsn_entry); + /* if this mark added a new event update the group mask */ + if (mask & ~group->mask) + fsnotify_recalc_group_mask(group); + out_err: if (ret < 0) kmem_cache_free(inotify_inode_mark_cachep, tmp_ientry); -- cgit v1.2.3 From b1f1b8ce0a1d71cbc72f7540134d52b79bd8f5ac Mon Sep 17 00:00:00 2001 From: Ryusuke Konishi Date: Sun, 30 Aug 2009 04:21:41 +0900 Subject: nilfs2: fix preempt count underflow in nilfs_btnode_prepare_change_key This will fix the following preempt count underflow reported from users with the title "[NILFS users] segctord problem" (Message-ID: <949415.6494.qm@web58808.mail.re1.yahoo.com> and Message-ID: ): WARNING: at kernel/sched.c:4890 sub_preempt_count+0x95/0xa0() Hardware name: HP Compaq 6530b (KR980UT#ABC) Modules linked in: bridge stp llc bnep rfcomm l2cap xfs exportfs nilfs2 cowloop loop vboxnetadp vboxnetflt vboxdrv btusb bluetooth uvcvideo videodev v4l1_compat v4l2_compat_ioctl32 arc4 snd_hda_codec_analog ecb iwlagn iwlcore rfkill lib80211 mac80211 snd_hda_intel snd_hda_codec ehci_hcd uhci_hcd usbcore snd_hwdep snd_pcm tg3 cfg80211 psmouse snd_timer joydev libphy ohci1394 snd_page_alloc hp_accel lis3lv02d ieee1394 led_class i915 drm i2c_algo_bit video backlight output i2c_core dm_crypt dm_mod Pid: 4197, comm: segctord Not tainted 2.6.30-gentoo-r4-64 #7 Call Trace: [] ? sub_preempt_count+0x95/0xa0 [] warn_slowpath_common+0x78/0xd0 [] warn_slowpath_null+0xf/0x20 [] sub_preempt_count+0x95/0xa0 [] nilfs_btnode_prepare_change_key+0x11b/0x190 [nilfs2] [] nilfs_btree_assign_p+0x19d/0x1e0 [nilfs2] [] nilfs_btree_assign+0xbd/0x130 [nilfs2] [] nilfs_bmap_assign+0x47/0x70 [nilfs2] [] nilfs_segctor_do_construct+0x956/0x20f0 [nilfs2] [] ? _spin_unlock_irqrestore+0x12/0x40 [] ? __up_write+0xe0/0x150 [] ? up_write+0x9/0x10 [] ? nilfs_bmap_test_and_clear_dirty+0x43/0x60 [nilfs2] [] ? nilfs_mdt_fetch_dirty+0x27/0x60 [nilfs2] [] nilfs_segctor_construct+0x8c/0xd0 [nilfs2] [] nilfs_segctor_thread+0x15c/0x3a0 [nilfs2] [] ? nilfs_construction_timeout+0x0/0x10 [nilfs2] [] ? add_timer+0x13/0x20 [] ? __wake_up_common+0x5a/0x90 [] ? autoremove_wake_function+0x0/0x40 [] ? nilfs_segctor_thread+0x0/0x3a0 [nilfs2] [] ? nilfs_segctor_thread+0x0/0x3a0 [nilfs2] [] kthread+0x56/0x90 [] child_rip+0xa/0x20 [] ? kthread+0x0/0x90 [] ? child_rip+0x0/0x20 This problem was caused due to a missing radix_tree_preload() call in the retry path of nilfs_btnode_prepare_change_key() function. Reported-by: Eric A Reported-by: Jerome Poulin Signed-off-by: Ryusuke Konishi Tested-by: Jerome Poulin Cc: stable@kernel.org --- fs/nilfs2/btnode.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/nilfs2/btnode.c b/fs/nilfs2/btnode.c index 7e0b61be212e..c668bca579c1 100644 --- a/fs/nilfs2/btnode.c +++ b/fs/nilfs2/btnode.c @@ -209,6 +209,7 @@ int nilfs_btnode_prepare_change_key(struct address_space *btnc, * We cannot call radix_tree_preload for the kernels older * than 2.6.23, because it is not exported for modules. */ +retry: err = radix_tree_preload(GFP_NOFS & ~__GFP_HIGHMEM); if (err) goto failed_unlock; @@ -219,7 +220,6 @@ int nilfs_btnode_prepare_change_key(struct address_space *btnc, (unsigned long long)oldkey, (unsigned long long)newkey); -retry: spin_lock_irq(&btnc->tree_lock); err = radix_tree_insert(&btnc->page_tree, newkey, obh->b_page); spin_unlock_irq(&btnc->tree_lock); -- cgit v1.2.3 From 37d0892c5a94e208cf863e3b7bac014edee4346d Mon Sep 17 00:00:00 2001 From: Ian Kent Date: Tue, 1 Sep 2009 11:26:22 +0800 Subject: autofs4 - fix missed case when changing to use struct path In the recent change by Al Viro that changes verious subsystems to use "struct path" one case was missed in the autofs4 module which causes mounts to no longer expire. Signed-off-by: Ian Kent Signed-off-by: Linus Torvalds --- fs/autofs4/expire.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/autofs4/expire.c b/fs/autofs4/expire.c index aa39ae83f019..3da18d453488 100644 --- a/fs/autofs4/expire.c +++ b/fs/autofs4/expire.c @@ -77,7 +77,7 @@ static int autofs4_mount_busy(struct vfsmount *mnt, struct dentry *dentry) } /* Update the expiry counter if fs is busy */ - if (!may_umount_tree(mnt)) { + if (!may_umount_tree(path.mnt)) { struct autofs_info *ino = autofs4_dentry_ino(top); ino->last_used = jiffies; goto done; -- cgit v1.2.3 From 3725867dccfb83e4b0cff64e916a04258f300591 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Tue, 1 Sep 2009 14:03:08 -0400 Subject: xfs: actually enable the swapext compat handler Fix a small typo in the compat ioctl handler that cause the swapext compat handler to never be called. Signed-off-by: Christoph Hellwig Reviewed-by: Torsten Kaiser Tested-by: Torsten Kaiser Reviewed-by: Eric Sandeen Reviewed-by: Felix Blyakher Signed-off-by: Felix Blyakher --- fs/xfs/linux-2.6/xfs_ioctl32.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/xfs/linux-2.6/xfs_ioctl32.c b/fs/xfs/linux-2.6/xfs_ioctl32.c index 0882d166239a..eafcc7c18706 100644 --- a/fs/xfs/linux-2.6/xfs_ioctl32.c +++ b/fs/xfs/linux-2.6/xfs_ioctl32.c @@ -619,7 +619,7 @@ xfs_file_compat_ioctl( case XFS_IOC_GETVERSION_32: cmd = _NATIVE_IOC(cmd, long); return xfs_file_ioctl(filp, cmd, p); - case XFS_IOC_SWAPEXT: { + case XFS_IOC_SWAPEXT_32: { struct xfs_swapext sxp; struct compat_xfs_swapext __user *sxu = arg; -- cgit v1.2.3 From bc8cec0dff072f1a45ce7f6b2c5234bb3411ac51 Mon Sep 17 00:00:00 2001 From: Massimo Cirillo Date: Thu, 27 Aug 2009 10:44:09 +0200 Subject: JFFS2: add missing verify buffer allocation/deallocation The function jffs2_nor_wbuf_flash_setup() doesn't allocate the verify buffer if CONFIG_JFFS2_FS_WBUF_VERIFY is defined, so causing a kernel panic when that macro is enabled and the verify function is called. Similarly the jffs2_nor_wbuf_flash_cleanup() must free the buffer if CONFIG_JFFS2_FS_WBUF_VERIFY is enabled. The following patch fixes the problem. The following patch applies to 2.6.30 kernel. Signed-off-by: Massimo Cirillo Signed-off-by: Artem Bityutskiy Signed-off-by: David Woodhouse Cc: stable@kernel.org --- fs/jffs2/wbuf.c | 10 ++++++++++ 1 file changed, 10 insertions(+) (limited to 'fs') diff --git a/fs/jffs2/wbuf.c b/fs/jffs2/wbuf.c index d9a721e6db70..5ef7bac265e5 100644 --- a/fs/jffs2/wbuf.c +++ b/fs/jffs2/wbuf.c @@ -1268,10 +1268,20 @@ int jffs2_nor_wbuf_flash_setup(struct jffs2_sb_info *c) { if (!c->wbuf) return -ENOMEM; +#ifdef CONFIG_JFFS2_FS_WBUF_VERIFY + c->wbuf_verify = kmalloc(c->wbuf_pagesize, GFP_KERNEL); + if (!c->wbuf_verify) { + kfree(c->wbuf); + return -ENOMEM; + } +#endif return 0; } void jffs2_nor_wbuf_flash_cleanup(struct jffs2_sb_info *c) { +#ifdef CONFIG_JFFS2_FS_WBUF_VERIFY + kfree(c->wbuf_verify); +#endif kfree(c->wbuf); } -- cgit v1.2.3 From 8379e7c46cc48f51197dd663fc6676f47f2a1e71 Mon Sep 17 00:00:00 2001 From: Sunil Mushran Date: Fri, 4 Sep 2009 11:12:01 -0700 Subject: ocfs2: ocfs2_write_begin_nolock() should handle len=0 Bug introduced by mainline commit e7432675f8ca868a4af365759a8d4c3779a3d922 The bug causes ocfs2_write_begin_nolock() to oops when len=0. Signed-off-by: Sunil Mushran Cc: stable@kernel.org Signed-off-by: Joel Becker --- fs/ocfs2/aops.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/ocfs2/aops.c b/fs/ocfs2/aops.c index b401654011a2..8a1e61545f41 100644 --- a/fs/ocfs2/aops.c +++ b/fs/ocfs2/aops.c @@ -1747,8 +1747,8 @@ int ocfs2_write_begin_nolock(struct address_space *mapping, * we know zeros will only be needed in the first and/or last cluster. */ if (clusters_to_alloc || extents_to_split || - wc->w_desc[0].c_needs_zero || - wc->w_desc[wc->w_clen - 1].c_needs_zero) + (wc->w_clen && (wc->w_desc[0].c_needs_zero || + wc->w_desc[wc->w_clen - 1].c_needs_zero))) cluster_of_pages = 1; else cluster_of_pages = 0; -- cgit v1.2.3 From a2a8474c3fff88d8dd52d05cb450563fb26fd26c Mon Sep 17 00:00:00 2001 From: Oleg Nesterov Date: Sat, 5 Sep 2009 11:17:13 -0700 Subject: exec: do not sleep in TASK_TRACED under ->cred_guard_mutex Tom Horsley reports that his debugger hangs when it tries to read /proc/pid_of_tracee/maps, this happens since "mm_for_maps: take ->cred_guard_mutex to fix the race with exec" 04b836cbf19e885f8366bccb2e4b0474346c02d commit in 2.6.31. But the root of the problem lies in the fact that do_execve() path calls tracehook_report_exec() which can stop if the tracer sets PT_TRACE_EXEC. The tracee must not sleep in TASK_TRACED holding this mutex. Even if we remove ->cred_guard_mutex from mm_for_maps() and proc_pid_attr_write(), another task doing PTRACE_ATTACH should not hang until it is killed or the tracee resumes. With this patch do_execve() does not use ->cred_guard_mutex directly and we do not hold it throughout, instead: - introduce prepare_bprm_creds() helper, it locks the mutex and calls prepare_exec_creds() to initialize bprm->cred. - install_exec_creds() drops the mutex after commit_creds(), and thus before tracehook_report_exec()->ptrace_stop(). or, if exec fails, free_bprm() drops this mutex when bprm->cred != NULL which indicates install_exec_creds() was not called. Reported-by: Tom Horsley Signed-off-by: Oleg Nesterov Acked-by: David Howells Cc: Roland McGrath Cc: James Morris Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/compat.c | 17 ++++--------- fs/exec.c | 63 +++++++++++++++++++++++++++++-------------------- include/linux/binfmts.h | 1 + 3 files changed, 43 insertions(+), 38 deletions(-) (limited to 'fs') diff --git a/fs/compat.c b/fs/compat.c index 94502dab972a..6d6f98fe64a0 100644 --- a/fs/compat.c +++ b/fs/compat.c @@ -1485,20 +1485,15 @@ int compat_do_execve(char * filename, if (!bprm) goto out_files; - retval = -ERESTARTNOINTR; - if (mutex_lock_interruptible(¤t->cred_guard_mutex)) + retval = prepare_bprm_creds(bprm); + if (retval) goto out_free; - current->in_execve = 1; - - retval = -ENOMEM; - bprm->cred = prepare_exec_creds(); - if (!bprm->cred) - goto out_unlock; retval = check_unsafe_exec(bprm); if (retval < 0) - goto out_unlock; + goto out_free; clear_in_exec = retval; + current->in_execve = 1; file = open_exec(filename); retval = PTR_ERR(file); @@ -1547,7 +1542,6 @@ int compat_do_execve(char * filename, /* execve succeeded */ current->fs->in_exec = 0; current->in_execve = 0; - mutex_unlock(¤t->cred_guard_mutex); acct_update_integrals(current); free_bprm(bprm); if (displaced) @@ -1567,10 +1561,7 @@ out_file: out_unmark: if (clear_in_exec) current->fs->in_exec = 0; - -out_unlock: current->in_execve = 0; - mutex_unlock(¤t->cred_guard_mutex); out_free: free_bprm(bprm); diff --git a/fs/exec.c b/fs/exec.c index fb4f3cdda78c..172ceb6edde4 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -1015,6 +1015,35 @@ out: EXPORT_SYMBOL(flush_old_exec); +/* + * Prepare credentials and lock ->cred_guard_mutex. + * install_exec_creds() commits the new creds and drops the lock. + * Or, if exec fails before, free_bprm() should release ->cred and + * and unlock. + */ +int prepare_bprm_creds(struct linux_binprm *bprm) +{ + if (mutex_lock_interruptible(¤t->cred_guard_mutex)) + return -ERESTARTNOINTR; + + bprm->cred = prepare_exec_creds(); + if (likely(bprm->cred)) + return 0; + + mutex_unlock(¤t->cred_guard_mutex); + return -ENOMEM; +} + +void free_bprm(struct linux_binprm *bprm) +{ + free_arg_pages(bprm); + if (bprm->cred) { + mutex_unlock(¤t->cred_guard_mutex); + abort_creds(bprm->cred); + } + kfree(bprm); +} + /* * install the new credentials for this executable */ @@ -1024,12 +1053,13 @@ void install_exec_creds(struct linux_binprm *bprm) commit_creds(bprm->cred); bprm->cred = NULL; - - /* cred_guard_mutex must be held at least to this point to prevent + /* + * cred_guard_mutex must be held at least to this point to prevent * ptrace_attach() from altering our determination of the task's - * credentials; any time after this it may be unlocked */ - + * credentials; any time after this it may be unlocked. + */ security_bprm_committed_creds(bprm); + mutex_unlock(¤t->cred_guard_mutex); } EXPORT_SYMBOL(install_exec_creds); @@ -1246,14 +1276,6 @@ int search_binary_handler(struct linux_binprm *bprm,struct pt_regs *regs) EXPORT_SYMBOL(search_binary_handler); -void free_bprm(struct linux_binprm *bprm) -{ - free_arg_pages(bprm); - if (bprm->cred) - abort_creds(bprm->cred); - kfree(bprm); -} - /* * sys_execve() executes a new program. */ @@ -1277,20 +1299,15 @@ int do_execve(char * filename, if (!bprm) goto out_files; - retval = -ERESTARTNOINTR; - if (mutex_lock_interruptible(¤t->cred_guard_mutex)) + retval = prepare_bprm_creds(bprm); + if (retval) goto out_free; - current->in_execve = 1; - - retval = -ENOMEM; - bprm->cred = prepare_exec_creds(); - if (!bprm->cred) - goto out_unlock; retval = check_unsafe_exec(bprm); if (retval < 0) - goto out_unlock; + goto out_free; clear_in_exec = retval; + current->in_execve = 1; file = open_exec(filename); retval = PTR_ERR(file); @@ -1340,7 +1357,6 @@ int do_execve(char * filename, /* execve succeeded */ current->fs->in_exec = 0; current->in_execve = 0; - mutex_unlock(¤t->cred_guard_mutex); acct_update_integrals(current); free_bprm(bprm); if (displaced) @@ -1360,10 +1376,7 @@ out_file: out_unmark: if (clear_in_exec) current->fs->in_exec = 0; - -out_unlock: current->in_execve = 0; - mutex_unlock(¤t->cred_guard_mutex); out_free: free_bprm(bprm); diff --git a/include/linux/binfmts.h b/include/linux/binfmts.h index 61ee18c1bdb4..2046b5b8af48 100644 --- a/include/linux/binfmts.h +++ b/include/linux/binfmts.h @@ -117,6 +117,7 @@ extern int setup_arg_pages(struct linux_binprm * bprm, int executable_stack); extern int bprm_mm_init(struct linux_binprm *bprm); extern int copy_strings_kernel(int argc,char ** argv,struct linux_binprm *bprm); +extern int prepare_bprm_creds(struct linux_binprm *bprm); extern void install_exec_creds(struct linux_binprm *bprm); extern void do_coredump(long signr, int exit_code, struct pt_regs *regs); extern int set_binfmt(struct linux_binfmt *new); -- cgit v1.2.3 From 9de6886ec6e37f45807266a702bb7621498395ad Mon Sep 17 00:00:00 2001 From: Nicolas Pitre Date: Sat, 5 Sep 2009 00:25:37 -0400 Subject: ext2: fix unbalanced kmap()/kunmap() In ext2_rename(), dir_page is acquired through ext2_dotdot(). It is then released through ext2_set_link() but only if old_dir != new_dir. Failing that, the pkmap reference count is never decremented and the page remains pinned forever. Repeat that a couple times with highmem pages and all pkmap slots get exhausted, and every further kmap() calls end up stalling on the pkmap_map_wait queue at which point the whole system comes to a halt. Signed-off-by: Nicolas Pitre Acked-by: Theodore Ts'o Cc: stable@kernel.org Signed-off-by: Linus Torvalds --- fs/ext2/namei.c | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'fs') diff --git a/fs/ext2/namei.c b/fs/ext2/namei.c index e1dedb0f7873..78d9b925fc94 100644 --- a/fs/ext2/namei.c +++ b/fs/ext2/namei.c @@ -362,6 +362,10 @@ static int ext2_rename (struct inode * old_dir, struct dentry * old_dentry, if (dir_de) { if (old_dir != new_dir) ext2_set_link(old_inode, dir_de, dir_page, new_dir, 0); + else { + kunmap(dir_page); + page_cache_release(dir_page); + } inode_dec_link_count(old_dir); } return 0; -- cgit v1.2.3