summaryrefslogtreecommitdiffstats
path: root/fs/nilfs2
diff options
context:
space:
mode:
authorLinus Torvalds <torvalds@linux-foundation.org>2024-09-21 17:20:50 +0200
committerLinus Torvalds <torvalds@linux-foundation.org>2024-09-21 17:20:50 +0200
commit7856a565416e0cf091f825b0e25c7a1b7abb650e (patch)
tree0a04a0594167fc997b3b1299610b5ef95ab89f19 /fs/nilfs2
parentMerge tag 'mm-stable-2024-09-20-02-31' of git://git.kernel.org/pub/scm/linux/... (diff)
parentlist: test: increase coverage of list_test_list_replace*() (diff)
downloadlinux-7856a565416e0cf091f825b0e25c7a1b7abb650e.tar.xz
linux-7856a565416e0cf091f825b0e25c7a1b7abb650e.zip
Merge tag 'mm-nonmm-stable-2024-09-21-07-52' of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm
Pull non-MM updates from Andrew Morton: "Many singleton patches - please see the various changelogs for details. Quite a lot of nilfs2 work this time around. Notable patch series in this pull request are: - "mul_u64_u64_div_u64: new implementation" by Nicolas Pitre, with assistance from Uwe Kleine-König. Reimplement mul_u64_u64_div_u64() to provide (much) more accurate results. The current implementation was causing Uwe some issues in the PWM drivers. - "xz: Updates to license, filters, and compression options" from Lasse Collin. Miscellaneous maintenance and kinor feature work to the xz decompressor. - "Fix some GDB command error and add some GDB commands" from Kuan-Ying Lee. Fixes and enhancements to the gdb scripts. - "treewide: add missing MODULE_DESCRIPTION() macros" from Jeff Johnson. Adds lots of MODULE_DESCRIPTIONs, thus fixing lots of warnings about this. - "nilfs2: add support for some common ioctls" from Ryusuke Konishi. Adds various commonly-available ioctls to nilfs2. - "This series fixes a number of formatting issues in kernel doc comments" from Ryusuke Konishi does that. - "nilfs2: prevent unexpected ENOENT propagation" from Ryusuke Konishi. Fix issues where -ENOENT was being unintentionally and inappropriately returned to userspace. - "nilfs2: assorted cleanups" from Huang Xiaojia. - "nilfs2: fix potential issues with empty b-tree nodes" from Ryusuke Konishi fixes some issues which can occur on corrupted nilfs2 filesystems. - "scripts/decode_stacktrace.sh: improve error reporting and usability" from Luca Ceresoli does those things" * tag 'mm-nonmm-stable-2024-09-21-07-52' of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm: (103 commits) list: test: increase coverage of list_test_list_replace*() list: test: fix tests for list_cut_position() proc: use __auto_type more treewide: correct the typo 'retun' ocfs2: cleanup return value and mlog in ocfs2_global_read_info() nilfs2: remove duplicate 'unlikely()' usage nilfs2: fix potential oob read in nilfs_btree_check_delete() nilfs2: determine empty node blocks as corrupted nilfs2: fix potential null-ptr-deref in nilfs_btree_insert() user_namespace: use kmemdup_array() instead of kmemdup() for multiple allocation tools/mm: rm thp_swap_allocator_test when make clean squashfs: fix percpu address space issues in decompressor_multi_percpu.c lib: glob.c: added null check for character class nilfs2: refactor nilfs_segctor_thread() nilfs2: use kthread_create and kthread_stop for the log writer thread nilfs2: remove sc_timer_task nilfs2: do not repair reserved inode bitmap in nilfs_new_inode() nilfs2: eliminate the shared counter and spinlock for i_generation nilfs2: separate inode type information from i_state field nilfs2: use the BITS_PER_LONG macro ...
Diffstat (limited to 'fs/nilfs2')
-rw-r--r--fs/nilfs2/alloc.h2
-rw-r--r--fs/nilfs2/bmap.c2
-rw-r--r--fs/nilfs2/bmap.h20
-rw-r--r--fs/nilfs2/btnode.c63
-rw-r--r--fs/nilfs2/btree.c12
-rw-r--r--fs/nilfs2/btree.h1
-rw-r--r--fs/nilfs2/cpfile.c54
-rw-r--r--fs/nilfs2/dat.c17
-rw-r--r--fs/nilfs2/dir.c44
-rw-r--r--fs/nilfs2/inode.c79
-rw-r--r--fs/nilfs2/ioctl.c109
-rw-r--r--fs/nilfs2/mdt.c6
-rw-r--r--fs/nilfs2/nilfs.h27
-rw-r--r--fs/nilfs2/page.c21
-rw-r--r--fs/nilfs2/page.h4
-rw-r--r--fs/nilfs2/recovery.c11
-rw-r--r--fs/nilfs2/segment.c234
-rw-r--r--fs/nilfs2/segment.h10
-rw-r--r--fs/nilfs2/sufile.c52
-rw-r--r--fs/nilfs2/super.c9
-rw-r--r--fs/nilfs2/the_nilfs.c5
-rw-r--r--fs/nilfs2/the_nilfs.h6
22 files changed, 463 insertions, 325 deletions
diff --git a/fs/nilfs2/alloc.h b/fs/nilfs2/alloc.h
index d825a9faca6d..e19d7eb10084 100644
--- a/fs/nilfs2/alloc.h
+++ b/fs/nilfs2/alloc.h
@@ -37,7 +37,7 @@ void *nilfs_palloc_block_get_entry(const struct inode *, __u64,
int nilfs_palloc_count_max_entries(struct inode *, u64, u64 *);
/**
- * nilfs_palloc_req - persistent allocator request and reply
+ * struct nilfs_palloc_req - persistent allocator request and reply
* @pr_entry_nr: entry number (vblocknr or inode number)
* @pr_desc_bh: buffer head of the buffer containing block group descriptors
* @pr_bitmap_bh: buffer head of the buffer containing a block group bitmap
diff --git a/fs/nilfs2/bmap.c b/fs/nilfs2/bmap.c
index cd14ea25968c..c9e8d9a7d820 100644
--- a/fs/nilfs2/bmap.c
+++ b/fs/nilfs2/bmap.c
@@ -349,7 +349,7 @@ int nilfs_bmap_propagate(struct nilfs_bmap *bmap, struct buffer_head *bh)
}
/**
- * nilfs_bmap_lookup_dirty_buffers -
+ * nilfs_bmap_lookup_dirty_buffers - collect dirty block buffers
* @bmap: bmap
* @listp: pointer to buffer head list
*/
diff --git a/fs/nilfs2/bmap.h b/fs/nilfs2/bmap.h
index 608168a5cb88..4656df392722 100644
--- a/fs/nilfs2/bmap.h
+++ b/fs/nilfs2/bmap.h
@@ -44,6 +44,19 @@ struct nilfs_bmap_stats {
/**
* struct nilfs_bmap_operations - bmap operation table
+ * @bop_lookup: single block search operation
+ * @bop_lookup_contig: consecutive block search operation
+ * @bop_insert: block insertion operation
+ * @bop_delete: block delete operation
+ * @bop_clear: block mapping resource release operation
+ * @bop_propagate: operation to propagate dirty state towards the
+ * mapping root
+ * @bop_lookup_dirty_buffers: operation to collect dirty block buffers
+ * @bop_assign: disk block address assignment operation
+ * @bop_mark: operation to mark in-use blocks as dirty for
+ * relocation by GC
+ * @bop_seek_key: find valid block key operation
+ * @bop_last_key: find last valid block key operation
*/
struct nilfs_bmap_operations {
int (*bop_lookup)(const struct nilfs_bmap *, __u64, int, __u64 *);
@@ -66,7 +79,7 @@ struct nilfs_bmap_operations {
int (*bop_seek_key)(const struct nilfs_bmap *, __u64, __u64 *);
int (*bop_last_key)(const struct nilfs_bmap *, __u64 *);
- /* The following functions are internal use only. */
+ /* private: internal use only */
int (*bop_check_insert)(const struct nilfs_bmap *, __u64);
int (*bop_check_delete)(struct nilfs_bmap *, __u64);
int (*bop_gather_data)(struct nilfs_bmap *, __u64 *, __u64 *, int);
@@ -74,9 +87,8 @@ struct nilfs_bmap_operations {
#define NILFS_BMAP_SIZE (NILFS_INODE_BMAP_SIZE * sizeof(__le64))
-#define NILFS_BMAP_KEY_BIT (sizeof(unsigned long) * 8 /* CHAR_BIT */)
-#define NILFS_BMAP_NEW_PTR_INIT \
- (1UL << (sizeof(unsigned long) * 8 /* CHAR_BIT */ - 1))
+#define NILFS_BMAP_KEY_BIT BITS_PER_LONG
+#define NILFS_BMAP_NEW_PTR_INIT (1UL << (BITS_PER_LONG - 1))
static inline int nilfs_bmap_is_new_ptr(unsigned long ptr)
{
diff --git a/fs/nilfs2/btnode.c b/fs/nilfs2/btnode.c
index c034080c334b..57b4af5ad646 100644
--- a/fs/nilfs2/btnode.c
+++ b/fs/nilfs2/btnode.c
@@ -179,11 +179,32 @@ void nilfs_btnode_delete(struct buffer_head *bh)
}
/**
- * nilfs_btnode_prepare_change_key
- * prepare to move contents of the block for old key to one of new key.
- * the old buffer will not be removed, but might be reused for new buffer.
- * it might return -ENOMEM because of memory allocation errors,
- * and might return -EIO because of disk read errors.
+ * nilfs_btnode_prepare_change_key - prepare to change the search key of a
+ * b-tree node block
+ * @btnc: page cache in which the b-tree node block is buffered
+ * @ctxt: structure for exchanging context information for key change
+ *
+ * nilfs_btnode_prepare_change_key() prepares to move the contents of the
+ * b-tree node block of the old key given in the "oldkey" member of @ctxt to
+ * the position of the new key given in the "newkey" member of @ctxt in the
+ * page cache @btnc. Here, the key of the block is an index in units of
+ * blocks, and if the page and block sizes match, it matches the page index
+ * in the page cache.
+ *
+ * If the page size and block size match, this function attempts to move the
+ * entire folio, and in preparation for this, inserts the original folio into
+ * the new index of the cache. If this insertion fails or if the page size
+ * and block size are different, it falls back to a copy preparation using
+ * nilfs_btnode_create_block(), inserts a new block at the position
+ * corresponding to "newkey", and stores the buffer head pointer in the
+ * "newbh" member of @ctxt.
+ *
+ * Note that the current implementation does not support folio sizes larger
+ * than the page size.
+ *
+ * Return: 0 on success, or the following negative error code on failure.
+ * * %-EIO - I/O error (metadata corruption).
+ * * %-ENOMEM - Insufficient memory available.
*/
int nilfs_btnode_prepare_change_key(struct address_space *btnc,
struct nilfs_btnode_chkey_ctxt *ctxt)
@@ -245,8 +266,21 @@ retry:
}
/**
- * nilfs_btnode_commit_change_key
- * commit the change_key operation prepared by prepare_change_key().
+ * nilfs_btnode_commit_change_key - commit the change of the search key of
+ * a b-tree node block
+ * @btnc: page cache in which the b-tree node block is buffered
+ * @ctxt: structure for exchanging context information for key change
+ *
+ * nilfs_btnode_commit_change_key() executes the key change based on the
+ * context @ctxt prepared by nilfs_btnode_prepare_change_key(). If no valid
+ * block buffer is prepared in "newbh" of @ctxt (i.e., a full folio move),
+ * this function removes the folio from the old index and completes the move.
+ * Otherwise, it copies the block data and inherited flag states of "oldbh"
+ * to "newbh" and clears the "oldbh" from the cache. In either case, the
+ * relocated buffer is marked as dirty.
+ *
+ * As with nilfs_btnode_prepare_change_key(), the current implementation does
+ * not support folio sizes larger than the page size.
*/
void nilfs_btnode_commit_change_key(struct address_space *btnc,
struct nilfs_btnode_chkey_ctxt *ctxt)
@@ -285,8 +319,19 @@ void nilfs_btnode_commit_change_key(struct address_space *btnc,
}
/**
- * nilfs_btnode_abort_change_key
- * abort the change_key operation prepared by prepare_change_key().
+ * nilfs_btnode_abort_change_key - abort the change of the search key of a
+ * b-tree node block
+ * @btnc: page cache in which the b-tree node block is buffered
+ * @ctxt: structure for exchanging context information for key change
+ *
+ * nilfs_btnode_abort_change_key() cancels the key change associated with the
+ * context @ctxt prepared via nilfs_btnode_prepare_change_key() and performs
+ * any necessary cleanup. If no valid block buffer is prepared in "newbh" of
+ * @ctxt, this function removes the folio from the destination index and aborts
+ * the move. Otherwise, it clears "newbh" from the cache.
+ *
+ * As with nilfs_btnode_prepare_change_key(), the current implementation does
+ * not support folio sizes larger than the page size.
*/
void nilfs_btnode_abort_change_key(struct address_space *btnc,
struct nilfs_btnode_chkey_ctxt *ctxt)
diff --git a/fs/nilfs2/btree.c b/fs/nilfs2/btree.c
index 862bdf23120e..ef5061bb56da 100644
--- a/fs/nilfs2/btree.c
+++ b/fs/nilfs2/btree.c
@@ -350,7 +350,7 @@ static int nilfs_btree_node_broken(const struct nilfs_btree_node *node,
if (unlikely(level < NILFS_BTREE_LEVEL_NODE_MIN ||
level >= NILFS_BTREE_LEVEL_MAX ||
(flags & NILFS_BTREE_NODE_ROOT) ||
- nchildren < 0 ||
+ nchildren <= 0 ||
nchildren > NILFS_BTREE_NODE_NCHILDREN_MAX(size))) {
nilfs_crit(inode->i_sb,
"bad btree node (ino=%lu, blocknr=%llu): level = %d, flags = 0x%x, nchildren = %d",
@@ -381,7 +381,8 @@ static int nilfs_btree_root_broken(const struct nilfs_btree_node *node,
if (unlikely(level < NILFS_BTREE_LEVEL_NODE_MIN ||
level >= NILFS_BTREE_LEVEL_MAX ||
nchildren < 0 ||
- nchildren > NILFS_BTREE_ROOT_NCHILDREN_MAX)) {
+ nchildren > NILFS_BTREE_ROOT_NCHILDREN_MAX ||
+ (nchildren == 0 && level > NILFS_BTREE_LEVEL_NODE_MIN))) {
nilfs_crit(inode->i_sb,
"bad btree root (ino=%lu): level = %d, flags = 0x%x, nchildren = %d",
inode->i_ino, level, flags, nchildren);
@@ -1658,13 +1659,16 @@ static int nilfs_btree_check_delete(struct nilfs_bmap *btree, __u64 key)
int nchildren, ret;
root = nilfs_btree_get_root(btree);
+ nchildren = nilfs_btree_node_get_nchildren(root);
+ if (unlikely(nchildren == 0))
+ return 0;
+
switch (nilfs_btree_height(btree)) {
case 2:
bh = NULL;
node = root;
break;
case 3:
- nchildren = nilfs_btree_node_get_nchildren(root);
if (nchildren > 1)
return 0;
ptr = nilfs_btree_node_get_ptr(root, nchildren - 1,
@@ -1673,12 +1677,12 @@ static int nilfs_btree_check_delete(struct nilfs_bmap *btree, __u64 key)
if (ret < 0)
return ret;
node = (struct nilfs_btree_node *)bh->b_data;
+ nchildren = nilfs_btree_node_get_nchildren(node);
break;
default:
return 0;
}
- nchildren = nilfs_btree_node_get_nchildren(node);
maxkey = nilfs_btree_node_get_key(node, nchildren - 1);
nextmaxkey = (nchildren > 1) ?
nilfs_btree_node_get_key(node, nchildren - 2) : 0;
diff --git a/fs/nilfs2/btree.h b/fs/nilfs2/btree.h
index 92868e1a48ca..2a220f716c91 100644
--- a/fs/nilfs2/btree.h
+++ b/fs/nilfs2/btree.h
@@ -24,6 +24,7 @@
* @bp_index: index of child node
* @bp_oldreq: ptr end request for old ptr
* @bp_newreq: ptr alloc request for new ptr
+ * @bp_ctxt: context information for changing the key of a b-tree node block
* @bp_op: rebalance operation
*/
struct nilfs_btree_path {
diff --git a/fs/nilfs2/cpfile.c b/fs/nilfs2/cpfile.c
index 69a5cced1e84..f0ce37552446 100644
--- a/fs/nilfs2/cpfile.c
+++ b/fs/nilfs2/cpfile.c
@@ -125,10 +125,17 @@ static void nilfs_cpfile_block_init(struct inode *cpfile,
}
}
-static inline int nilfs_cpfile_get_header_block(struct inode *cpfile,
- struct buffer_head **bhp)
+static int nilfs_cpfile_get_header_block(struct inode *cpfile,
+ struct buffer_head **bhp)
{
- return nilfs_mdt_get_block(cpfile, 0, 0, NULL, bhp);
+ int err = nilfs_mdt_get_block(cpfile, 0, 0, NULL, bhp);
+
+ if (unlikely(err == -ENOENT)) {
+ nilfs_error(cpfile->i_sb,
+ "missing header block in checkpoint metadata");
+ err = -EIO;
+ }
+ return err;
}
static inline int nilfs_cpfile_get_checkpoint_block(struct inode *cpfile,
@@ -283,14 +290,9 @@ int nilfs_cpfile_create_checkpoint(struct inode *cpfile, __u64 cno)
down_write(&NILFS_MDT(cpfile)->mi_sem);
ret = nilfs_cpfile_get_header_block(cpfile, &header_bh);
- if (unlikely(ret < 0)) {
- if (ret == -ENOENT) {
- nilfs_error(cpfile->i_sb,
- "checkpoint creation failed due to metadata corruption.");
- ret = -EIO;
- }
+ if (unlikely(ret < 0))
goto out_sem;
- }
+
ret = nilfs_cpfile_get_checkpoint_block(cpfile, cno, 1, &cp_bh);
if (unlikely(ret < 0))
goto out_header;
@@ -704,9 +706,15 @@ ssize_t nilfs_cpfile_get_cpinfo(struct inode *cpfile, __u64 *cnop, int mode,
}
/**
- * nilfs_cpfile_delete_checkpoint -
- * @cpfile:
- * @cno:
+ * nilfs_cpfile_delete_checkpoint - delete a checkpoint
+ * @cpfile: checkpoint file inode
+ * @cno: checkpoint number to delete
+ *
+ * Return: 0 on success, or the following negative error code on failure.
+ * * %-EBUSY - Checkpoint in use (snapshot specified).
+ * * %-EIO - I/O error (including metadata corruption).
+ * * %-ENOENT - No valid checkpoint found.
+ * * %-ENOMEM - Insufficient memory available.
*/
int nilfs_cpfile_delete_checkpoint(struct inode *cpfile, __u64 cno)
{
@@ -968,21 +976,15 @@ static int nilfs_cpfile_clear_snapshot(struct inode *cpfile, __u64 cno)
}
/**
- * nilfs_cpfile_is_snapshot -
+ * nilfs_cpfile_is_snapshot - determine if checkpoint is a snapshot
* @cpfile: inode of checkpoint file
- * @cno: checkpoint number
- *
- * Description:
- *
- * Return Value: On success, 1 is returned if the checkpoint specified by
- * @cno is a snapshot, or 0 if not. On error, one of the following negative
- * error codes is returned.
- *
- * %-EIO - I/O error.
- *
- * %-ENOMEM - Insufficient amount of memory available.
+ * @cno: checkpoint number
*
- * %-ENOENT - No such checkpoint.
+ * Return: 1 if the checkpoint specified by @cno is a snapshot, 0 if not, or
+ * the following negative error code on failure.
+ * * %-EIO - I/O error (including metadata corruption).
+ * * %-ENOENT - No such checkpoint.
+ * * %-ENOMEM - Insufficient memory available.
*/
int nilfs_cpfile_is_snapshot(struct inode *cpfile, __u64 cno)
{
diff --git a/fs/nilfs2/dat.c b/fs/nilfs2/dat.c
index fc1caf63a42a..0bef662176a4 100644
--- a/fs/nilfs2/dat.c
+++ b/fs/nilfs2/dat.c
@@ -271,18 +271,15 @@ void nilfs_dat_abort_update(struct inode *dat,
}
/**
- * nilfs_dat_mark_dirty -
- * @dat: DAT file inode
+ * nilfs_dat_mark_dirty - mark the DAT block buffer containing the specified
+ * virtual block address entry as dirty
+ * @dat: DAT file inode
* @vblocknr: virtual block number
*
- * Description:
- *
- * Return Value: On success, 0 is returned. On error, one of the following
- * negative error codes is returned.
- *
- * %-EIO - I/O error.
- *
- * %-ENOMEM - Insufficient amount of memory available.
+ * Return: 0 on success, or the following negative error code on failure.
+ * * %-EINVAL - Invalid DAT entry (internal code).
+ * * %-EIO - I/O error (including metadata corruption).
+ * * %-ENOMEM - Insufficient memory available.
*/
int nilfs_dat_mark_dirty(struct inode *dat, __u64 vblocknr)
{
diff --git a/fs/nilfs2/dir.c b/fs/nilfs2/dir.c
index 4b3e19d74925..fe5b1a30c509 100644
--- a/fs/nilfs2/dir.c
+++ b/fs/nilfs2/dir.c
@@ -231,37 +231,6 @@ static struct nilfs_dir_entry *nilfs_next_entry(struct nilfs_dir_entry *p)
nilfs_rec_len_from_disk(p->rec_len));
}
-static unsigned char
-nilfs_filetype_table[NILFS_FT_MAX] = {
- [NILFS_FT_UNKNOWN] = DT_UNKNOWN,
- [NILFS_FT_REG_FILE] = DT_REG,
- [NILFS_FT_DIR] = DT_DIR,
- [NILFS_FT_CHRDEV] = DT_CHR,
- [NILFS_FT_BLKDEV] = DT_BLK,
- [NILFS_FT_FIFO] = DT_FIFO,
- [NILFS_FT_SOCK] = DT_SOCK,
- [NILFS_FT_SYMLINK] = DT_LNK,
-};
-
-#define S_SHIFT 12
-static unsigned char
-nilfs_type_by_mode[(S_IFMT >> S_SHIFT) + 1] = {
- [S_IFREG >> S_SHIFT] = NILFS_FT_REG_FILE,
- [S_IFDIR >> S_SHIFT] = NILFS_FT_DIR,
- [S_IFCHR >> S_SHIFT] = NILFS_FT_CHRDEV,
- [S_IFBLK >> S_SHIFT] = NILFS_FT_BLKDEV,
- [S_IFIFO >> S_SHIFT] = NILFS_FT_FIFO,
- [S_IFSOCK >> S_SHIFT] = NILFS_FT_SOCK,
- [S_IFLNK >> S_SHIFT] = NILFS_FT_SYMLINK,
-};
-
-static void nilfs_set_de_type(struct nilfs_dir_entry *de, struct inode *inode)
-{
- umode_t mode = inode->i_mode;
-
- de->file_type = nilfs_type_by_mode[(mode & S_IFMT)>>S_SHIFT];
-}
-
static int nilfs_readdir(struct file *file, struct dir_context *ctx)
{
loff_t pos = ctx->pos;
@@ -297,10 +266,7 @@ static int nilfs_readdir(struct file *file, struct dir_context *ctx)
if (de->inode) {
unsigned char t;
- if (de->file_type < NILFS_FT_MAX)
- t = nilfs_filetype_table[de->file_type];
- else
- t = DT_UNKNOWN;
+ t = fs_ftype_to_dtype(de->file_type);
if (!dir_emit(ctx, de->name, de->name_len,
le64_to_cpu(de->inode), t)) {
@@ -444,7 +410,7 @@ void nilfs_set_link(struct inode *dir, struct nilfs_dir_entry *de,
err = nilfs_prepare_chunk(folio, from, to);
BUG_ON(err);
de->inode = cpu_to_le64(inode->i_ino);
- nilfs_set_de_type(de, inode);
+ de->file_type = fs_umode_to_ftype(inode->i_mode);
nilfs_commit_chunk(folio, mapping, from, to);
inode_set_mtime_to_ts(dir, inode_set_ctime_current(dir));
}
@@ -531,7 +497,7 @@ got_it:
de->name_len = namelen;
memcpy(de->name, name, namelen);
de->inode = cpu_to_le64(inode->i_ino);
- nilfs_set_de_type(de, inode);
+ de->file_type = fs_umode_to_ftype(inode->i_mode);
nilfs_commit_chunk(folio, folio->mapping, from, to);
inode_set_mtime_to_ts(dir, inode_set_ctime_current(dir));
nilfs_mark_inode_dirty(dir);
@@ -612,14 +578,14 @@ int nilfs_make_empty(struct inode *inode, struct inode *parent)
de->rec_len = nilfs_rec_len_to_disk(NILFS_DIR_REC_LEN(1));
memcpy(de->name, ".\0\0", 4);
de->inode = cpu_to_le64(inode->i_ino);
- nilfs_set_de_type(de, inode);
+ de->file_type = fs_umode_to_ftype(inode->i_mode);
de = (struct nilfs_dir_entry *)(kaddr + NILFS_DIR_REC_LEN(1));
de->name_len = 2;
de->rec_len = nilfs_rec_len_to_disk(chunk_size - NILFS_DIR_REC_LEN(1));
de->inode = cpu_to_le64(parent->i_ino);
memcpy(de->name, "..\0", 4);
- nilfs_set_de_type(de, inode);
+ de->file_type = fs_umode_to_ftype(inode->i_mode);
kunmap_local(kaddr);
nilfs_commit_chunk(folio, mapping, 0, chunk_size);
fail:
diff --git a/fs/nilfs2/inode.c b/fs/nilfs2/inode.c
index 8661f452dba6..be6acf6e2bfc 100644
--- a/fs/nilfs2/inode.c
+++ b/fs/nilfs2/inode.c
@@ -15,6 +15,7 @@
#include <linux/writeback.h>
#include <linux/uio.h>
#include <linux/fiemap.h>
+#include <linux/random.h>
#include "nilfs.h"
#include "btnode.h"
#include "segment.h"
@@ -28,17 +29,13 @@
* @ino: inode number
* @cno: checkpoint number
* @root: pointer on NILFS root object (mounted checkpoint)
- * @for_gc: inode for GC flag
- * @for_btnc: inode for B-tree node cache flag
- * @for_shadow: inode for shadowed page cache flag
+ * @type: inode type
*/
struct nilfs_iget_args {
u64 ino;
__u64 cno;
struct nilfs_root *root;
- bool for_gc;
- bool for_btnc;
- bool for_shadow;
+ unsigned int type;
};
static int nilfs_iget_test(struct inode *inode, void *opaque);
@@ -162,7 +159,7 @@ static int nilfs_writepages(struct address_space *mapping,
int err = 0;
if (sb_rdonly(inode->i_sb)) {
- nilfs_clear_dirty_pages(mapping, false);
+ nilfs_clear_dirty_pages(mapping);
return -EROFS;
}
@@ -186,7 +183,7 @@ static int nilfs_writepage(struct page *page, struct writeback_control *wbc)
* have dirty pages that try to be flushed in background.
* So, here we simply discard this dirty page.
*/
- nilfs_clear_folio_dirty(folio, false);
+ nilfs_clear_folio_dirty(folio);
folio_unlock(folio);
return -EROFS;
}
@@ -315,8 +312,7 @@ static int nilfs_insert_inode_locked(struct inode *inode,
unsigned long ino)
{
struct nilfs_iget_args args = {
- .ino = ino, .root = root, .cno = 0, .for_gc = false,
- .for_btnc = false, .for_shadow = false
+ .ino = ino, .root = root, .cno = 0, .type = NILFS_I_TYPE_NORMAL
};
return insert_inode_locked4(inode, ino, nilfs_iget_test, &args);
@@ -325,7 +321,6 @@ static int nilfs_insert_inode_locked(struct inode *inode,
struct inode *nilfs_new_inode(struct inode *dir, umode_t mode)
{
struct super_block *sb = dir->i_sb;
- struct the_nilfs *nilfs = sb->s_fs_info;
struct inode *inode;
struct nilfs_inode_info *ii;
struct nilfs_root *root;
@@ -343,25 +338,13 @@ struct inode *nilfs_new_inode(struct inode *dir, umode_t mode)
root = NILFS_I(dir)->i_root;
ii = NILFS_I(inode);
ii->i_state = BIT(NILFS_I_NEW);
+ ii->i_type = NILFS_I_TYPE_NORMAL;
ii->i_root = root;
err = nilfs_ifile_create_inode(root->ifile, &ino, &bh);
if (unlikely(err))
goto failed_ifile_create_inode;
/* reference count of i_bh inherits from nilfs_mdt_read_block() */
-
- if (unlikely(ino < NILFS_USER_INO)) {
- nilfs_warn(sb,
- "inode bitmap is inconsistent for reserved inodes");
- do {
- brelse(bh);
- err = nilfs_ifile_create_inode(root->ifile, &ino, &bh);
- if (unlikely(err))
- goto failed_ifile_create_inode;
- } while (ino < NILFS_USER_INO);
-
- nilfs_info(sb, "repaired inode bitmap for reserved inodes");
- }
ii->i_bh = bh;
atomic64_inc(&root->inodes_count);
@@ -385,9 +368,7 @@ struct inode *nilfs_new_inode(struct inode *dir, umode_t mode)
/* ii->i_dir_acl = 0; */
ii->i_dir_start_lookup = 0;
nilfs_set_inode_flags(inode);
- spin_lock(&nilfs->ns_next_gen_lock);
- inode->i_generation = nilfs->ns_next_generation++;
- spin_unlock(&nilfs->ns_next_gen_lock);
+ inode->i_generation = get_random_u32();
if (nilfs_insert_inode_locked(inode, root, ino) < 0) {
err = -EIO;
goto failed_after_creation;
@@ -546,23 +527,10 @@ static int nilfs_iget_test(struct inode *inode, void *opaque)
return 0;
ii = NILFS_I(inode);
- if (test_bit(NILFS_I_BTNC, &ii->i_state)) {
- if (!args->for_btnc)
- return 0;
- } else if (args->for_btnc) {
+ if (ii->i_type != args->type)
return 0;
- }
- if (test_bit(NILFS_I_SHADOW, &ii->i_state)) {
- if (!args->for_shadow)
- return 0;
- } else if (args->for_shadow) {
- return 0;
- }
- if (!test_bit(NILFS_I_GCINODE, &ii->i_state))
- return !args->for_gc;
-
- return args->for_gc && args->cno == ii->i_cno;
+ return !(args->type & NILFS_I_TYPE_GC) || args->cno == ii->i_cno;
}
static int nilfs_iget_set(struct inode *inode, void *opaque)
@@ -572,15 +540,9 @@ static int nilfs_iget_set(struct inode *inode, void *opaque)
inode->i_ino = args->ino;
NILFS_I(inode)->i_cno = args->cno;
NILFS_I(inode)->i_root = args->root;
+ NILFS_I(inode)->i_type = args->type;
if (args->root && args->ino == NILFS_ROOT_INO)
nilfs_get_root(args->root);
-
- if (args->for_gc)
- NILFS_I(inode)->i_state = BIT(NILFS_I_GCINODE);
- if (args->for_btnc)
- NILFS_I(inode)->i_state |= BIT(NILFS_I_BTNC);
- if (args->for_shadow)
- NILFS_I(inode)->i_state |= BIT(NILFS_I_SHADOW);
return 0;
}
@@ -588,8 +550,7 @@ struct inode *nilfs_ilookup(struct super_block *sb, struct nilfs_root *root,
unsigned long ino)
{
struct nilfs_iget_args args = {
- .ino = ino, .root = root, .cno = 0, .for_gc = false,
- .for_btnc = false, .for_shadow = false
+ .ino = ino, .root = root, .cno = 0, .type = NILFS_I_TYPE_NORMAL
};
return ilookup5(sb, ino, nilfs_iget_test, &args);
@@ -599,8 +560,7 @@ struct inode *nilfs_iget_locked(struct super_block *sb, struct nilfs_root *root,
unsigned long ino)
{
struct nilfs_iget_args args = {
- .ino = ino, .root = root, .cno = 0, .for_gc = false,
- .for_btnc = false, .for_shadow = false
+ .ino = ino, .root = root, .cno = 0, .type = NILFS_I_TYPE_NORMAL
};
return iget5_locked(sb, ino, nilfs_iget_test, nilfs_iget_set, &args);
@@ -631,8 +591,7 @@ struct inode *nilfs_iget_for_gc(struct super_block *sb, unsigned long ino,
__u64 cno)
{
struct nilfs_iget_args args = {
- .ino = ino, .root = NULL, .cno = cno, .for_gc = true,
- .for_btnc = false, .for_shadow = false
+ .ino = ino, .root = NULL, .cno = cno, .type = NILFS_I_TYPE_GC
};
struct inode *inode;
int err;
@@ -677,9 +636,7 @@ int nilfs_attach_btree_node_cache(struct inode *inode)
args.ino = inode->i_ino;
args.root = ii->i_root;
args.cno = ii->i_cno;
- args.for_gc = test_bit(NILFS_I_GCINODE, &ii->i_state) != 0;
- args.for_btnc = true;
- args.for_shadow = test_bit(NILFS_I_SHADOW, &ii->i_state) != 0;
+ args.type = ii->i_type | NILFS_I_TYPE_BTNC;
btnc_inode = iget5_locked(inode->i_sb, inode->i_ino, nilfs_iget_test,
nilfs_iget_set, &args);
@@ -733,8 +690,8 @@ void nilfs_detach_btree_node_cache(struct inode *inode)
struct inode *nilfs_iget_for_shadow(struct inode *inode)
{
struct nilfs_iget_args args = {
- .ino = inode->i_ino, .root = NULL, .cno = 0, .for_gc = false,
- .for_btnc = false, .for_shadow = true
+ .ino = inode->i_ino, .root = NULL, .cno = 0,
+ .type = NILFS_I_TYPE_SHADOW
};
struct inode *s_inode;
int err;
@@ -900,7 +857,7 @@ static void nilfs_clear_inode(struct inode *inode)
if (test_bit(NILFS_I_BMAP, &ii->i_state))
nilfs_bmap_clear(ii->i_bmap);
- if (!test_bit(NILFS_I_BTNC, &ii->i_state))
+ if (!(ii->i_type & NILFS_I_TYPE_BTNC))
nilfs_detach_btree_node_cache(inode);
if (ii->i_root && inode->i_ino == NILFS_ROOT_INO)
diff --git a/fs/nilfs2/ioctl.c b/fs/nilfs2/ioctl.c
index 8be471ce4f19..fa77f78df681 100644
--- a/fs/nilfs2/ioctl.c
+++ b/fs/nilfs2/ioctl.c
@@ -17,6 +17,7 @@
#include <linux/mount.h> /* mnt_want_write_file(), mnt_drop_write_file() */
#include <linux/buffer_head.h>
#include <linux/fileattr.h>
+#include <linux/string.h>
#include "nilfs.h"
#include "segment.h"
#include "bmap.h"
@@ -114,7 +115,11 @@ static int nilfs_ioctl_wrap_copy(struct the_nilfs *nilfs,
}
/**
- * nilfs_fileattr_get - ioctl to support lsattr
+ * nilfs_fileattr_get - retrieve miscellaneous file attributes
+ * @dentry: the object to retrieve from
+ * @fa: fileattr pointer
+ *
+ * Return: always 0 as success.
*/
int nilfs_fileattr_get(struct dentry *dentry, struct fileattr *fa)
{
@@ -126,7 +131,12 @@ int nilfs_fileattr_get(struct dentry *dentry, struct fileattr *fa)
}
/**
- * nilfs_fileattr_set - ioctl to support chattr
+ * nilfs_fileattr_set - change miscellaneous file attributes
+ * @idmap: idmap of the mount
+ * @dentry: the object to change
+ * @fa: fileattr pointer
+ *
+ * Return: 0 on success, or a negative error code on failure.
*/
int nilfs_fileattr_set(struct mnt_idmap *idmap,
struct dentry *dentry, struct fileattr *fa)
@@ -159,6 +169,10 @@ int nilfs_fileattr_set(struct mnt_idmap *idmap,
/**
* nilfs_ioctl_getversion - get info about a file's version (generation number)
+ * @inode: inode object
+ * @argp: userspace memory where the generation number of @inode is stored
+ *
+ * Return: 0 on success, or %-EFAULT on error.
*/
static int nilfs_ioctl_getversion(struct inode *inode, void __user *argp)
{
@@ -1266,6 +1280,91 @@ out:
return ret;
}
+/**
+ * nilfs_ioctl_get_fslabel - get the volume name of the file system
+ * @sb: super block instance
+ * @argp: pointer to userspace memory where the volume name should be stored
+ *
+ * Return: 0 on success, %-EFAULT if copying to userspace memory fails.
+ */
+static int nilfs_ioctl_get_fslabel(struct super_block *sb, void __user *argp)
+{
+ struct the_nilfs *nilfs = sb->s_fs_info;
+ char label[NILFS_MAX_VOLUME_NAME + 1];
+
+ BUILD_BUG_ON(NILFS_MAX_VOLUME_NAME >= FSLABEL_MAX);
+
+ down_read(&nilfs->ns_sem);
+ memtostr_pad(label, nilfs->ns_sbp[0]->s_volume_name);
+ up_read(&nilfs->ns_sem);
+
+ if (copy_to_user(argp, label, sizeof(label)))
+ return -EFAULT;
+ return 0;
+}
+
+/**
+ * nilfs_ioctl_set_fslabel - set the volume name of the file system
+ * @sb: super block instance
+ * @filp: file object
+ * @argp: pointer to userspace memory that contains the volume name
+ *
+ * Return: 0 on success, or the following negative error code on failure.
+ * * %-EFAULT - Error copying input data.
+ * * %-EINVAL - Label length exceeds record size in superblock.
+ * * %-EIO - I/O error.
+ * * %-EPERM - Operation not permitted (insufficient permissions).
+ * * %-EROFS - Read only file system.
+ */
+static int nilfs_ioctl_set_fslabel(struct super_block *sb, struct file *filp,
+ void __user *argp)
+{
+ char label[NILFS_MAX_VOLUME_NAME + 1];
+ struct the_nilfs *nilfs = sb->s_fs_info;
+ struct nilfs_super_block **sbp;
+ size_t len;
+ int ret;
+
+ if (!capable(CAP_SYS_ADMIN))
+ return -EPERM;
+
+ ret = mnt_want_write_file(filp);
+ if (ret)
+ return ret;
+
+ if (copy_from_user(label, argp, NILFS_MAX_VOLUME_NAME + 1)) {
+ ret = -EFAULT;
+ goto out_drop_write;
+ }
+
+ len = strnlen(label, NILFS_MAX_VOLUME_NAME + 1);
+ if (len > NILFS_MAX_VOLUME_NAME) {
+ nilfs_err(sb, "unable to set label with more than %zu bytes",
+ NILFS_MAX_VOLUME_NAME);
+ ret = -EINVAL;
+ goto out_drop_write;
+ }
+
+ down_write(&nilfs->ns_sem);
+ sbp = nilfs_prepare_super(sb, false);
+ if (unlikely(!sbp)) {
+ ret = -EIO;
+ goto out_unlock;
+ }
+
+ strtomem_pad(sbp[0]->s_volume_name, label, 0);
+ if (sbp[1])
+ strtomem_pad(sbp[1]->s_volume_name, label, 0);
+
+ ret = nilfs_commit_super(sb, NILFS_SB_COMMIT_ALL);
+
+out_unlock:
+ up_write(&nilfs->ns_sem);
+out_drop_write:
+ mnt_drop_write_file(filp);
+ return ret;
+}
+
long nilfs_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
{
struct inode *inode = file_inode(filp);
@@ -1308,6 +1407,10 @@ long nilfs_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
return nilfs_ioctl_set_alloc_range(inode, argp);
case FITRIM:
return nilfs_ioctl_trim_fs(inode, argp);
+ case FS_IOC_GETFSLABEL:
+ return nilfs_ioctl_get_fslabel(inode->i_sb, argp);
+ case FS_IOC_SETFSLABEL:
+ return nilfs_ioctl_set_fslabel(inode->i_sb, filp, argp);
default:
return -ENOTTY;
}
@@ -1334,6 +1437,8 @@ long nilfs_compat_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
case NILFS_IOCTL_RESIZE:
case NILFS_IOCTL_SET_ALLOC_RANGE:
case FITRIM:
+ case FS_IOC_GETFSLABEL:
+ case FS_IOC_SETFSLABEL:
break;
default:
return -ENOIOCTLCMD;
diff --git a/fs/nilfs2/mdt.c b/fs/nilfs2/mdt.c
index 4f792a0ad0f0..ceb7dc0b5bad 100644
--- a/fs/nilfs2/mdt.c
+++ b/fs/nilfs2/mdt.c
@@ -411,7 +411,7 @@ nilfs_mdt_write_page(struct page *page, struct writeback_control *wbc)
* have dirty folios that try to be flushed in background.
* So, here we simply discard this dirty folio.
*/
- nilfs_clear_folio_dirty(folio, false);
+ nilfs_clear_folio_dirty(folio);
folio_unlock(folio);
return -EROFS;
}
@@ -638,10 +638,10 @@ void nilfs_mdt_restore_from_shadow_map(struct inode *inode)
if (mi->mi_palloc_cache)
nilfs_palloc_clear_cache(inode);
- nilfs_clear_dirty_pages(inode->i_mapping, true);
+ nilfs_clear_dirty_pages(inode->i_mapping);
nilfs_copy_back_pages(inode->i_mapping, shadow->inode->i_mapping);
- nilfs_clear_dirty_pages(ii->i_assoc_inode->i_mapping, true);
+ nilfs_clear_dirty_pages(ii->i_assoc_inode->i_mapping);
nilfs_copy_back_pages(ii->i_assoc_inode->i_mapping,
NILFS_I(shadow->inode)->i_assoc_inode->i_mapping);
diff --git a/fs/nilfs2/nilfs.h b/fs/nilfs2/nilfs.h
index 4017f7856440..fb1c4c5bae7c 100644
--- a/fs/nilfs2/nilfs.h
+++ b/fs/nilfs2/nilfs.h
@@ -22,6 +22,7 @@
/**
* struct nilfs_inode_info - nilfs inode data in memory
* @i_flags: inode flags
+ * @i_type: inode type (combination of flags that inidicate usage)
* @i_state: dynamic state flags
* @i_bmap: pointer on i_bmap_data
* @i_bmap_data: raw block mapping
@@ -37,6 +38,7 @@
*/
struct nilfs_inode_info {
__u32 i_flags;
+ unsigned int i_type;
unsigned long i_state; /* Dynamic state flags */
struct nilfs_bmap *i_bmap;
struct nilfs_bmap i_bmap_data;
@@ -90,9 +92,16 @@ enum {
NILFS_I_UPDATED, /* The file has been written back */
NILFS_I_INODE_SYNC, /* dsync is not allowed for inode */
NILFS_I_BMAP, /* has bmap and btnode_cache */
- NILFS_I_GCINODE, /* inode for GC, on memory only */
- NILFS_I_BTNC, /* inode for btree node cache */
- NILFS_I_SHADOW, /* inode for shadowed page cache */
+};
+
+/*
+ * Flags to identify the usage of on-memory inodes (i_type)
+ */
+enum {
+ NILFS_I_TYPE_NORMAL = 0,
+ NILFS_I_TYPE_GC = 0x0001, /* For data caching during GC */
+ NILFS_I_TYPE_BTNC = 0x0002, /* For btree node cache */
+ NILFS_I_TYPE_SHADOW = 0x0004, /* For shadowed page cache */
};
/*
@@ -103,6 +112,18 @@ enum {
NILFS_SB_COMMIT_ALL /* Commit both super blocks */
};
+/**
+ * define NILFS_MAX_VOLUME_NAME - maximum number of characters (bytes) in a
+ * file system volume name
+ *
+ * Defined by the size of the volume name field in the on-disk superblocks.
+ * This volume name does not include the terminating NULL byte if the string
+ * length matches the field size, so use (NILFS_MAX_VOLUME_NAME + 1) for the
+ * size of the buffer that requires a NULL byte termination.
+ */
+#define NILFS_MAX_VOLUME_NAME \
+ sizeof_field(struct nilfs_super_block, s_volume_name)
+
/*
* Macros to check inode numbers
*/
diff --git a/fs/nilfs2/page.c b/fs/nilfs2/page.c
index 14e470fb8870..9c0b7cddeaae 100644
--- a/fs/nilfs2/page.c
+++ b/fs/nilfs2/page.c
@@ -262,7 +262,7 @@ repeat:
NILFS_FOLIO_BUG(folio, "inconsistent dirty state");
dfolio = filemap_grab_folio(dmap, folio->index);
- if (unlikely(IS_ERR(dfolio))) {
+ if (IS_ERR(dfolio)) {
/* No empty page is added to the page cache */
folio_unlock(folio);
err = PTR_ERR(dfolio);
@@ -357,9 +357,8 @@ repeat:
/**
* nilfs_clear_dirty_pages - discard dirty pages in address space
* @mapping: address space with dirty pages for discarding
- * @silent: suppress [true] or print [false] warning messages
*/
-void nilfs_clear_dirty_pages(struct address_space *mapping, bool silent)
+void nilfs_clear_dirty_pages(struct address_space *mapping)
{
struct folio_batch fbatch;
unsigned int i;
@@ -380,7 +379,7 @@ void nilfs_clear_dirty_pages(struct address_space *mapping, bool silent)
* was acquired. Skip processing in that case.
*/
if (likely(folio->mapping == mapping))
- nilfs_clear_folio_dirty(folio, silent);
+ nilfs_clear_folio_dirty(folio);
folio_unlock(folio);
}
@@ -392,20 +391,13 @@ void nilfs_clear_dirty_pages(struct address_space *mapping, bool silent)
/**
* nilfs_clear_folio_dirty - discard dirty folio
* @folio: dirty folio that will be discarded
- * @silent: suppress [true] or print [false] warning messages
*/
-void nilfs_clear_folio_dirty(struct folio *folio, bool silent)
+void nilfs_clear_folio_dirty(struct folio *folio)
{
- struct inode *inode = folio->mapping->host;
- struct super_block *sb = inode->i_sb;
struct buffer_head *bh, *head;
BUG_ON(!folio_test_locked(folio));
- if (!silent)
- nilfs_warn(sb, "discard dirty page: offset=%lld, ino=%lu",
- folio_pos(folio), inode->i_ino);
-
folio_clear_uptodate(folio);
folio_clear_mappedtodisk(folio);
@@ -419,11 +411,6 @@ void nilfs_clear_folio_dirty(struct folio *folio, bool silent)
bh = head;
do {
lock_buffer(bh);
- if (!silent)
- nilfs_warn(sb,
- "discard dirty block: blocknr=%llu, size=%zu",
- (u64)bh->b_blocknr, bh->b_size);
-
set_mask_bits(&bh->b_state, clear_bits, 0);
unlock_buffer(bh);
} while (bh = bh->b_this_page, bh != head);
diff --git a/fs/nilfs2/page.h b/fs/nilfs2/page.h
index 7e1a2c455a10..64521a03a19e 100644
--- a/fs/nilfs2/page.h
+++ b/fs/nilfs2/page.h
@@ -41,8 +41,8 @@ void nilfs_folio_bug(struct folio *);
int nilfs_copy_dirty_pages(struct address_space *, struct address_space *);
void nilfs_copy_back_pages(struct address_space *, struct address_space *);
-void nilfs_clear_folio_dirty(struct folio *, bool);
-void nilfs_clear_dirty_pages(struct address_space *, bool);
+void nilfs_clear_folio_dirty(struct folio *folio);
+void nilfs_clear_dirty_pages(struct address_space *mapping);
unsigned int nilfs_page_count_clean_buffers(struct page *, unsigned int,
unsigned int);
unsigned long nilfs_find_uncommitted_extent(struct inode *inode,
diff --git a/fs/nilfs2/recovery.c b/fs/nilfs2/recovery.c
index ec61ce9f29a2..21d81097a89f 100644
--- a/fs/nilfs2/recovery.c
+++ b/fs/nilfs2/recovery.c
@@ -433,8 +433,17 @@ static int nilfs_prepare_segment_for_recovery(struct the_nilfs *nilfs,
* The next segment is invalidated by this recovery.
*/
err = nilfs_sufile_free(sufile, segnum[1]);
- if (unlikely(err))
+ if (unlikely(err)) {
+ if (err == -ENOENT) {
+ nilfs_err(sb,
+ "checkpoint log inconsistency at block %llu (segment %llu): next segment %llu is unallocated",
+ (unsigned long long)nilfs->ns_last_pseg,
+ (unsigned long long)nilfs->ns_segnum,
+ (unsigned long long)segnum[1]);
+ err = -EINVAL;
+ }
goto failed;
+ }
for (i = 1; i < 4; i++) {
err = nilfs_segment_list_add(head, segnum[i]);
diff --git a/fs/nilfs2/segment.c b/fs/nilfs2/segment.c
index 871ec35ea8e8..587251830897 100644
--- a/fs/nilfs2/segment.c
+++ b/fs/nilfs2/segment.c
@@ -519,7 +519,7 @@ static void nilfs_segctor_end_finfo(struct nilfs_sc_info *sci,
ii = NILFS_I(inode);
- if (test_bit(NILFS_I_GCINODE, &ii->i_state))
+ if (ii->i_type & NILFS_I_TYPE_GC)
cno = ii->i_cno;
else if (NILFS_ROOT_METADATA_FILE(inode->i_ino))
cno = 0;
@@ -1102,12 +1102,64 @@ static int nilfs_segctor_scan_file_dsync(struct nilfs_sc_info *sci,
return err;
}
+/**
+ * nilfs_free_segments - free the segments given by an array of segment numbers
+ * @nilfs: nilfs object
+ * @segnumv: array of segment numbers to be freed
+ * @nsegs: number of segments to be freed in @segnumv
+ *
+ * nilfs_free_segments() wraps nilfs_sufile_freev() and
+ * nilfs_sufile_cancel_freev(), and edits the segment usage metadata file
+ * (sufile) to free all segments given by @segnumv and @nsegs at once. If
+ * it fails midway, it cancels the changes so that none of the segments are
+ * freed. If @nsegs is 0, this function does nothing.
+ *
+ * The freeing of segments is not finalized until the writing of a log with
+ * a super root block containing this sufile change is complete, and it can
+ * be canceled with nilfs_sufile_cancel_freev() until then.
+ *
+ * Return: 0 on success, or the following negative error code on failure.
+ * * %-EINVAL - Invalid segment number.
+ * * %-EIO - I/O error (including metadata corruption).
+ * * %-ENOMEM - Insufficient memory available.
+ */
+static int nilfs_free_segments(struct the_nilfs *nilfs, __u64 *segnumv,
+ size_t nsegs)
+{
+ size_t ndone;
+ int ret;
+
+ if (!nsegs)
+ return 0;
+
+ ret = nilfs_sufile_freev(nilfs->ns_sufile, segnumv, nsegs, &ndone);
+ if (unlikely(ret)) {
+ nilfs_sufile_cancel_freev(nilfs->ns_sufile, segnumv, ndone,
+ NULL);
+ /*
+ * If a segment usage of the segments to be freed is in a
+ * hole block, nilfs_sufile_freev() will return -ENOENT.
+ * In this case, -EINVAL should be returned to the caller
+ * since there is something wrong with the given segment
+ * number array. This error can only occur during GC, so
+ * there is no need to worry about it propagating to other
+ * callers (such as fsync).
+ */
+ if (ret == -ENOENT) {
+ nilfs_err(nilfs->ns_sb,
+ "The segment usage entry %llu to be freed is invalid (in a hole)",
+ (unsigned long long)segnumv[ndone]);
+ ret = -EINVAL;
+ }
+ }
+ return ret;
+}
+
static int nilfs_segctor_collect_blocks(struct nilfs_sc_info *sci, int mode)
{
struct the_nilfs *nilfs = sci->sc_super->s_fs_info;
struct list_head *head;
struct nilfs_inode_info *ii;
- size_t ndone;
int err = 0;
switch (nilfs_sc_cstage_get(sci)) {
@@ -1201,14 +1253,10 @@ static int nilfs_segctor_collect_blocks(struct nilfs_sc_info *sci, int mode)
nilfs_sc_cstage_inc(sci);
fallthrough;
case NILFS_ST_SUFILE:
- err = nilfs_sufile_freev(nilfs->ns_sufile, sci->sc_freesegs,
- sci->sc_nfreesegs, &ndone);
- if (unlikely(err)) {
- nilfs_sufile_cancel_freev(nilfs->ns_sufile,
- sci->sc_freesegs, ndone,
- NULL);
+ err = nilfs_free_segments(nilfs, sci->sc_freesegs,
+ sci->sc_nfreesegs);
+ if (unlikely(err))
break;
- }
sci->sc_stage.flags |= NILFS_CF_SUFREED;
err = nilfs_segctor_scan_file(sci, nilfs->ns_sufile,
@@ -2456,7 +2504,7 @@ static void nilfs_construction_timeout(struct timer_list *t)
{
struct nilfs_sc_info *sci = from_timer(sci, t, sc_timer);
- wake_up_process(sci->sc_timer_task);
+ wake_up_process(sci->sc_task);
}
static void
@@ -2582,123 +2630,85 @@ static int nilfs_segctor_flush_mode(struct nilfs_sc_info *sci)
}
/**
- * nilfs_segctor_thread - main loop of the segment constructor thread.
+ * nilfs_log_write_required - determine whether log writing is required
+ * @sci: nilfs_sc_info struct
+ * @modep: location for storing log writing mode
+ *
+ * Return: true if log writing is required, false otherwise. If log writing
+ * is required, the mode is stored in the location pointed to by @modep.
+ */
+static bool nilfs_log_write_required(struct nilfs_sc_info *sci, int *modep)
+{
+ bool timedout, ret = true;
+
+ spin_lock(&sci->sc_state_lock);
+ timedout = ((sci->sc_state & NILFS_SEGCTOR_COMMIT) &&
+ time_after_eq(jiffies, sci->sc_timer.expires));
+ if (timedout || sci->sc_seq_request != sci->sc_seq_done)
+ *modep = SC_LSEG_SR;
+ else if (sci->sc_flush_request)
+ *modep = nilfs_segctor_flush_mode(sci);
+ else
+ ret = false;
+
+ spin_unlock(&sci->sc_state_lock);
+ return ret;
+}
+
+/**
+ * nilfs_segctor_thread - main loop of the log writer thread
* @arg: pointer to a struct nilfs_sc_info.
*
- * nilfs_segctor_thread() initializes a timer and serves as a daemon
- * to execute segment constructions.
+ * nilfs_segctor_thread() is the main loop function of the log writer kernel
+ * thread, which determines whether log writing is necessary, and if so,
+ * performs the log write in the background, or waits if not. It is also
+ * used to decide the background writeback of the superblock.
+ *
+ * Return: Always 0.
*/
static int nilfs_segctor_thread(void *arg)
{
struct nilfs_sc_info *sci = (struct nilfs_sc_info *)arg;
struct the_nilfs *nilfs = sci->sc_super->s_fs_info;
- int timeout = 0;
-
- sci->sc_timer_task = current;
- timer_setup(&sci->sc_timer, nilfs_construction_timeout, 0);
- /* start sync. */
- sci->sc_task = current;
- wake_up(&sci->sc_wait_task); /* for nilfs_segctor_start_thread() */
nilfs_info(sci->sc_super,
"segctord starting. Construction interval = %lu seconds, CP frequency < %lu seconds",
sci->sc_interval / HZ, sci->sc_mjcp_freq / HZ);
set_freezable();
- spin_lock(&sci->sc_state_lock);
- loop:
- for (;;) {
- int mode;
-
- if (sci->sc_state & NILFS_SEGCTOR_QUIT)
- goto end_thread;
-
- if (timeout || sci->sc_seq_request != sci->sc_seq_done)
- mode = SC_LSEG_SR;
- else if (sci->sc_flush_request)
- mode = nilfs_segctor_flush_mode(sci);
- else
- break;
-
- spin_unlock(&sci->sc_state_lock);
- nilfs_segctor_thread_construct(sci, mode);
- spin_lock(&sci->sc_state_lock);
- timeout = 0;
- }
-
- if (freezing(current)) {
- spin_unlock(&sci->sc_state_lock);
- try_to_freeze();
- spin_lock(&sci->sc_state_lock);
- } else {
+ while (!kthread_should_stop()) {
DEFINE_WAIT(wait);
- int should_sleep = 1;
+ bool should_write;
+ int mode;
+
+ if (freezing(current)) {
+ try_to_freeze();
+ continue;
+ }
prepare_to_wait(&sci->sc_wait_daemon, &wait,
TASK_INTERRUPTIBLE);
-
- if (sci->sc_seq_request != sci->sc_seq_done)
- should_sleep = 0;
- else if (sci->sc_flush_request)
- should_sleep = 0;
- else if (sci->sc_state & NILFS_SEGCTOR_COMMIT)
- should_sleep = time_before(jiffies,
- sci->sc_timer.expires);
-
- if (should_sleep) {
- spin_unlock(&sci->sc_state_lock);
+ should_write = nilfs_log_write_required(sci, &mode);
+ if (!should_write)
schedule();
- spin_lock(&sci->sc_state_lock);
- }
finish_wait(&sci->sc_wait_daemon, &wait);
- timeout = ((sci->sc_state & NILFS_SEGCTOR_COMMIT) &&
- time_after_eq(jiffies, sci->sc_timer.expires));
if (nilfs_sb_dirty(nilfs) && nilfs_sb_need_update(nilfs))
set_nilfs_discontinued(nilfs);
+
+ if (should_write)
+ nilfs_segctor_thread_construct(sci, mode);
}
- goto loop;
- end_thread:
/* end sync. */
+ spin_lock(&sci->sc_state_lock);
sci->sc_task = NULL;
timer_shutdown_sync(&sci->sc_timer);
- wake_up(&sci->sc_wait_task); /* for nilfs_segctor_kill_thread() */
spin_unlock(&sci->sc_state_lock);
return 0;
}
-static int nilfs_segctor_start_thread(struct nilfs_sc_info *sci)
-{
- struct task_struct *t;
-
- t = kthread_run(nilfs_segctor_thread, sci, "segctord");
- if (IS_ERR(t)) {
- int err = PTR_ERR(t);
-
- nilfs_err(sci->sc_super, "error %d creating segctord thread",
- err);
- return err;
- }
- wait_event(sci->sc_wait_task, sci->sc_task != NULL);
- return 0;
-}
-
-static void nilfs_segctor_kill_thread(struct nilfs_sc_info *sci)
- __acquires(&sci->sc_state_lock)
- __releases(&sci->sc_state_lock)
-{
- sci->sc_state |= NILFS_SEGCTOR_QUIT;
-
- while (sci->sc_task) {
- wake_up(&sci->sc_wait_daemon);
- spin_unlock(&sci->sc_state_lock);
- wait_event(sci->sc_wait_task, sci->sc_task == NULL);
- spin_lock(&sci->sc_state_lock);
- }
-}
-
/*
* Setup & clean-up functions
*/
@@ -2719,7 +2729,6 @@ static struct nilfs_sc_info *nilfs_segctor_new(struct super_block *sb,
init_waitqueue_head(&sci->sc_wait_request);
init_waitqueue_head(&sci->sc_wait_daemon);
- init_waitqueue_head(&sci->sc_wait_task);
spin_lock_init(&sci->sc_state_lock);
INIT_LIST_HEAD(&sci->sc_dirty_files);
INIT_LIST_HEAD(&sci->sc_segbufs);
@@ -2774,8 +2783,12 @@ static void nilfs_segctor_destroy(struct nilfs_sc_info *sci)
up_write(&nilfs->ns_segctor_sem);
+ if (sci->sc_task) {
+ wake_up(&sci->sc_wait_daemon);
+ kthread_stop(sci->sc_task);
+ }
+
spin_lock(&sci->sc_state_lock);
- nilfs_segctor_kill_thread(sci);
flag = ((sci->sc_state & NILFS_SEGCTOR_COMMIT) || sci->sc_flush_request
|| sci->sc_seq_request != sci->sc_seq_done);
spin_unlock(&sci->sc_state_lock);
@@ -2823,14 +2836,15 @@ static void nilfs_segctor_destroy(struct nilfs_sc_info *sci)
* This allocates a log writer object, initializes it, and starts the
* log writer.
*
- * Return Value: On success, 0 is returned. On error, one of the following
- * negative error code is returned.
- *
- * %-ENOMEM - Insufficient memory available.
+ * Return: 0 on success, or the following negative error code on failure.
+ * * %-EINTR - Log writer thread creation failed due to interruption.
+ * * %-ENOMEM - Insufficient memory available.
*/
int nilfs_attach_log_writer(struct super_block *sb, struct nilfs_root *root)
{
struct the_nilfs *nilfs = sb->s_fs_info;
+ struct nilfs_sc_info *sci;
+ struct task_struct *t;
int err;
if (nilfs->ns_writer) {
@@ -2843,15 +2857,23 @@ int nilfs_attach_log_writer(struct super_block *sb, struct nilfs_root *root)
return 0;
}
- nilfs->ns_writer = nilfs_segctor_new(sb, root);
- if (!nilfs->ns_writer)
+ sci = nilfs_segctor_new(sb, root);
+ if (unlikely(!sci))
return -ENOMEM;
- err = nilfs_segctor_start_thread(nilfs->ns_writer);
- if (unlikely(err))
+ nilfs->ns_writer = sci;
+ t = kthread_create(nilfs_segctor_thread, sci, "segctord");
+ if (IS_ERR(t)) {
+ err = PTR_ERR(t);
+ nilfs_err(sb, "error %d creating segctord thread", err);
nilfs_detach_log_writer(sb);
+ return err;
+ }
+ sci->sc_task = t;
+ timer_setup(&sci->sc_timer, nilfs_construction_timeout, 0);
- return err;
+ wake_up_process(sci->sc_task);
+ return 0;
}
/**
diff --git a/fs/nilfs2/segment.h b/fs/nilfs2/segment.h
index 1060f72ebf5a..f723f47ddc4e 100644
--- a/fs/nilfs2/segment.h
+++ b/fs/nilfs2/segment.h
@@ -22,10 +22,10 @@ struct nilfs_root;
* struct nilfs_recovery_info - Recovery information
* @ri_need_recovery: Recovery status
* @ri_super_root: Block number of the last super root
- * @ri_ri_cno: Number of the last checkpoint
+ * @ri_cno: Number of the last checkpoint
* @ri_lsegs_start: Region for roll-forwarding (start block number)
* @ri_lsegs_end: Region for roll-forwarding (end block number)
- * @ri_lseg_start_seq: Sequence value of the segment at ri_lsegs_start
+ * @ri_lsegs_start_seq: Sequence value of the segment at ri_lsegs_start
* @ri_used_segments: List of segments to be mark active
* @ri_pseg_start: Block number of the last partial segment
* @ri_seq: Sequence number on the last partial segment
@@ -105,9 +105,8 @@ struct nilfs_segsum_pointer {
* @sc_flush_request: inode bitmap of metadata files to be flushed
* @sc_wait_request: Client request queue
* @sc_wait_daemon: Daemon wait queue
- * @sc_wait_task: Start/end wait queue to control segctord task
* @sc_seq_request: Request counter
- * @sc_seq_accept: Accepted request count
+ * @sc_seq_accepted: Accepted request count
* @sc_seq_done: Completion counter
* @sc_sync: Request of explicit sync operation
* @sc_interval: Timeout value of background construction
@@ -158,7 +157,6 @@ struct nilfs_sc_info {
wait_queue_head_t sc_wait_request;
wait_queue_head_t sc_wait_daemon;
- wait_queue_head_t sc_wait_task;
__u32 sc_seq_request;
__u32 sc_seq_accepted;
@@ -171,7 +169,6 @@ struct nilfs_sc_info {
unsigned long sc_watermark;
struct timer_list sc_timer;
- struct task_struct *sc_timer_task;
struct task_struct *sc_task;
};
@@ -192,7 +189,6 @@ enum {
};
/* sc_state */
-#define NILFS_SEGCTOR_QUIT 0x0001 /* segctord is being destroyed */
#define NILFS_SEGCTOR_COMMIT 0x0004 /* committed transaction exists */
/*
diff --git a/fs/nilfs2/sufile.c b/fs/nilfs2/sufile.c
index 6748218be7c5..eea5a6a12f7b 100644
--- a/fs/nilfs2/sufile.c
+++ b/fs/nilfs2/sufile.c
@@ -79,10 +79,17 @@ nilfs_sufile_block_get_segment_usage(const struct inode *sufile, __u64 segnum,
NILFS_MDT(sufile)->mi_entry_size;
}
-static inline int nilfs_sufile_get_header_block(struct inode *sufile,
- struct buffer_head **bhp)
+static int nilfs_sufile_get_header_block(struct inode *sufile,
+ struct buffer_head **bhp)
{
- return nilfs_mdt_get_block(sufile, 0, 0, NULL, bhp);
+ int err = nilfs_mdt_get_block(sufile, 0, 0, NULL, bhp);
+
+ if (unlikely(err == -ENOENT)) {
+ nilfs_error(sufile->i_sb,
+ "missing header block in segment usage metadata");
+ err = -EIO;
+ }
+ return err;
}
static inline int
@@ -506,8 +513,15 @@ int nilfs_sufile_mark_dirty(struct inode *sufile, __u64 segnum)
down_write(&NILFS_MDT(sufile)->mi_sem);
ret = nilfs_sufile_get_segment_usage_block(sufile, segnum, 0, &bh);
- if (ret)
+ if (unlikely(ret)) {
+ if (ret == -ENOENT) {
+ nilfs_error(sufile->i_sb,
+ "segment usage for segment %llu is unreadable due to a hole block",
+ (unsigned long long)segnum);
+ ret = -EIO;
+ }
goto out_sem;
+ }
kaddr = kmap_local_page(bh->b_page);
su = nilfs_sufile_block_get_segment_usage(sufile, segnum, bh, kaddr);
@@ -840,21 +854,17 @@ out:
}
/**
- * nilfs_sufile_get_suinfo -
+ * nilfs_sufile_get_suinfo - get segment usage information
* @sufile: inode of segment usage file
* @segnum: segment number to start looking
- * @buf: array of suinfo
- * @sisz: byte size of suinfo
- * @nsi: size of suinfo array
- *
- * Description:
+ * @buf: array of suinfo
+ * @sisz: byte size of suinfo
+ * @nsi: size of suinfo array
*
- * Return Value: On success, 0 is returned and .... On error, one of the
- * following negative error codes is returned.
- *
- * %-EIO - I/O error.
- *
- * %-ENOMEM - Insufficient amount of memory available.
+ * Return: Count of segment usage info items stored in the output buffer on
+ * success, or the following negative error code on failure.
+ * * %-EIO - I/O error (including metadata corruption).
+ * * %-ENOMEM - Insufficient memory available.
*/
ssize_t nilfs_sufile_get_suinfo(struct inode *sufile, __u64 segnum, void *buf,
unsigned int sisz, size_t nsi)
@@ -1241,9 +1251,15 @@ int nilfs_sufile_read(struct super_block *sb, size_t susize,
if (err)
goto failed;
- err = nilfs_sufile_get_header_block(sufile, &header_bh);
- if (err)
+ err = nilfs_mdt_get_block(sufile, 0, 0, NULL, &header_bh);
+ if (unlikely(err)) {
+ if (err == -ENOENT) {
+ nilfs_err(sb,
+ "missing header block in segment usage metadata");
+ err = -EINVAL;
+ }
goto failed;
+ }
sui = NILFS_SUI(sufile);
kaddr = kmap_local_page(header_bh->b_page);
diff --git a/fs/nilfs2/super.c b/fs/nilfs2/super.c
index e835e1f5a712..eca79cca3803 100644
--- a/fs/nilfs2/super.c
+++ b/fs/nilfs2/super.c
@@ -105,6 +105,10 @@ static void nilfs_set_error(struct super_block *sb)
/**
* __nilfs_error() - report failure condition on a filesystem
+ * @sb: super block instance
+ * @function: name of calling function
+ * @fmt: format string for message to be output
+ * @...: optional arguments to @fmt
*
* __nilfs_error() sets an ERROR_FS flag on the superblock as well as
* reporting an error message. This function should be called when
@@ -156,6 +160,7 @@ struct inode *nilfs_alloc_inode(struct super_block *sb)
return NULL;
ii->i_bh = NULL;
ii->i_state = 0;
+ ii->i_type = 0;
ii->i_cno = 0;
ii->i_assoc_inode = NULL;
ii->i_bmap = &ii->i_bmap_data;
@@ -1063,6 +1068,10 @@ nilfs_fill_super(struct super_block *sb, struct fs_context *fc)
if (err)
goto failed_nilfs;
+ super_set_uuid(sb, nilfs->ns_sbp[0]->s_uuid,
+ sizeof(nilfs->ns_sbp[0]->s_uuid));
+ super_set_sysfs_name_bdev(sb);
+
cno = nilfs_last_cno(nilfs);
err = nilfs_attach_checkpoint(sb, cno, true, &fsroot);
if (err) {
diff --git a/fs/nilfs2/the_nilfs.c b/fs/nilfs2/the_nilfs.c
index e44dde57ab65..ac03fd3c330c 100644
--- a/fs/nilfs2/the_nilfs.c
+++ b/fs/nilfs2/the_nilfs.c
@@ -12,7 +12,6 @@
#include <linux/slab.h>
#include <linux/blkdev.h>
#include <linux/backing-dev.h>
-#include <linux/random.h>
#include <linux/log2.h>
#include <linux/crc32.h>
#include "nilfs.h"
@@ -69,7 +68,6 @@ struct the_nilfs *alloc_nilfs(struct super_block *sb)
INIT_LIST_HEAD(&nilfs->ns_dirty_files);
INIT_LIST_HEAD(&nilfs->ns_gc_inodes);
spin_lock_init(&nilfs->ns_inode_lock);
- spin_lock_init(&nilfs->ns_next_gen_lock);
spin_lock_init(&nilfs->ns_last_segment_lock);
nilfs->ns_cptree = RB_ROOT;
spin_lock_init(&nilfs->ns_cptree_lock);
@@ -754,9 +752,6 @@ int init_nilfs(struct the_nilfs *nilfs, struct super_block *sb)
nilfs->ns_blocksize_bits = sb->s_blocksize_bits;
nilfs->ns_blocksize = blocksize;
- get_random_bytes(&nilfs->ns_next_generation,
- sizeof(nilfs->ns_next_generation));
-
err = nilfs_store_disk_layout(nilfs, sbp);
if (err)
goto failed_sbh;
diff --git a/fs/nilfs2/the_nilfs.h b/fs/nilfs2/the_nilfs.h
index 1e829ed7b0ef..4776a70f01ae 100644
--- a/fs/nilfs2/the_nilfs.h
+++ b/fs/nilfs2/the_nilfs.h
@@ -71,8 +71,6 @@ enum {
* @ns_dirty_files: list of dirty files
* @ns_inode_lock: lock protecting @ns_dirty_files
* @ns_gc_inodes: dummy inodes to keep live blocks
- * @ns_next_generation: next generation number for inodes
- * @ns_next_gen_lock: lock protecting @ns_next_generation
* @ns_mount_opt: mount options
* @ns_resuid: uid for reserved blocks
* @ns_resgid: gid for reserved blocks
@@ -161,10 +159,6 @@ struct the_nilfs {
/* GC inode list */
struct list_head ns_gc_inodes;
- /* Inode allocator */
- u32 ns_next_generation;
- spinlock_t ns_next_gen_lock;
-
/* Mount options */
unsigned long ns_mount_opt;