From a2de733c78fa7af51ba9670482fa7d392aa67c57 Mon Sep 17 00:00:00 2001 From: Arne Jansen Date: Tue, 8 Mar 2011 14:14:00 +0100 Subject: btrfs: scrub This adds an initial implementation for scrub. It works quite straightforward. The usermode issues an ioctl for each device in the fs. For each device, it enumerates the allocated device chunks. For each chunk, the contained extents are enumerated and the data checksums fetched. The extents are read sequentially and the checksums verified. If an error occurs (checksum or EIO), a good copy is searched for. If one is found, the bad copy will be rewritten. All enumerations happen from the commit roots. During a transaction commit, the scrubs get paused and afterwards continue from the new roots. This commit is based on the series originally posted to linux-btrfs with some improvements that resulted from comments from David Sterba, Ilya Dryomov and Jan Schmidt. Signed-off-by: Arne Jansen --- fs/btrfs/ctree.h | 37 +++++++++++++++++++++++++++++++++++-- 1 file changed, 35 insertions(+), 2 deletions(-) (limited to 'fs/btrfs/ctree.h') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 2e61fe1b6b8c..31141ba6072d 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -23,6 +23,7 @@ #include #include #include +#include #include #include #include @@ -33,6 +34,7 @@ #include "extent_io.h" #include "extent_map.h" #include "async-thread.h" +#include "ioctl.h" struct btrfs_trans_handle; struct btrfs_transaction; @@ -510,6 +512,12 @@ struct btrfs_extent_item_v0 { /* use full backrefs for extent pointers in the block */ #define BTRFS_BLOCK_FLAG_FULL_BACKREF (1ULL << 8) +/* + * this flag is only used internally by scrub and may be changed at any time + * it is only declared here to avoid collisions + */ +#define BTRFS_EXTENT_FLAG_SUPER (1ULL << 48) + struct btrfs_tree_block_info { struct btrfs_disk_key key; u8 level; @@ -1077,6 +1085,17 @@ struct btrfs_fs_info { void *bdev_holder; + /* private scrub information */ + struct mutex scrub_lock; + atomic_t scrubs_running; + atomic_t scrub_pause_req; + atomic_t scrubs_paused; + atomic_t scrub_cancel_req; + wait_queue_head_t scrub_pause_wait; + struct rw_semaphore scrub_super_lock; + int scrub_workers_refcnt; + struct btrfs_workers scrub_workers; + /* filesystem state */ u64 fs_state; }; @@ -2472,8 +2491,8 @@ struct btrfs_csum_item *btrfs_lookup_csum(struct btrfs_trans_handle *trans, int btrfs_csum_truncate(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct btrfs_path *path, u64 isize); -int btrfs_lookup_csums_range(struct btrfs_root *root, u64 start, - u64 end, struct list_head *list); +int btrfs_lookup_csums_range(struct btrfs_root *root, u64 start, u64 end, + struct list_head *list, int search_commit); /* inode.c */ /* RHEL and EL kernels have a patch that renames PG_checked to FsMisc */ @@ -2637,4 +2656,18 @@ void btrfs_reloc_pre_snapshot(struct btrfs_trans_handle *trans, u64 *bytes_to_reserve); void btrfs_reloc_post_snapshot(struct btrfs_trans_handle *trans, struct btrfs_pending_snapshot *pending); + +/* scrub.c */ +int btrfs_scrub_dev(struct btrfs_root *root, u64 devid, u64 start, u64 end, + struct btrfs_scrub_progress *progress); +int btrfs_scrub_pause(struct btrfs_root *root); +int btrfs_scrub_pause_super(struct btrfs_root *root); +int btrfs_scrub_continue(struct btrfs_root *root); +int btrfs_scrub_continue_super(struct btrfs_root *root); +int btrfs_scrub_cancel(struct btrfs_root *root); +int btrfs_scrub_cancel_dev(struct btrfs_root *root, struct btrfs_device *dev); +int btrfs_scrub_cancel_devid(struct btrfs_root *root, u64 devid); +int btrfs_scrub_progress(struct btrfs_root *root, u64 devid, + struct btrfs_scrub_progress *progress); + #endif -- cgit v1.2.3 From 475f63874d739d7842a56da94687f18d583ae654 Mon Sep 17 00:00:00 2001 From: Jan Schmidt Date: Fri, 11 Mar 2011 15:41:01 +0100 Subject: btrfs: new ioctls for scrub adds ioctls necessary to start and cancel scrubs, to get current progress and to get info about devices to be scrubbed. Note that the scrub is done per-device and that the ioctl only returns after the scrub for this devices is finished or has been canceled. Signed-off-by: Arne Jansen --- fs/btrfs/ctree.h | 2 - fs/btrfs/ioctl.c | 131 +++++++++++++++++++++++++++++++++++++++++++++++++++++++ fs/btrfs/ioctl.h | 38 ++++++++++++++++ 3 files changed, 169 insertions(+), 2 deletions(-) (limited to 'fs/btrfs/ctree.h') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 31141ba6072d..b7373b14e4cd 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -189,7 +189,6 @@ struct btrfs_mapping_tree { struct extent_map_tree map_tree; }; -#define BTRFS_UUID_SIZE 16 struct btrfs_dev_item { /* the internal btrfs device id */ __le64 devid; @@ -296,7 +295,6 @@ static inline unsigned long btrfs_chunk_item_size(int num_stripes) sizeof(struct btrfs_stripe) * (num_stripes - 1); } -#define BTRFS_FSID_SIZE 16 #define BTRFS_HEADER_FLAG_WRITTEN (1ULL << 0) #define BTRFS_HEADER_FLAG_RELOC (1ULL << 1) diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index f580a3a5d2fc..205cd011d2f3 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -1803,6 +1803,75 @@ static long btrfs_ioctl_rm_dev(struct btrfs_root *root, void __user *arg) return ret; } +static long btrfs_ioctl_fs_info(struct btrfs_root *root, void __user *arg) +{ + struct btrfs_ioctl_fs_info_args fi_args; + struct btrfs_device *device; + struct btrfs_device *next; + struct btrfs_fs_devices *fs_devices = root->fs_info->fs_devices; + + if (!capable(CAP_SYS_ADMIN)) + return -EPERM; + + fi_args.num_devices = fs_devices->num_devices; + fi_args.max_id = 0; + memcpy(&fi_args.fsid, root->fs_info->fsid, sizeof(fi_args.fsid)); + + mutex_lock(&fs_devices->device_list_mutex); + list_for_each_entry_safe(device, next, &fs_devices->devices, dev_list) { + if (device->devid > fi_args.max_id) + fi_args.max_id = device->devid; + } + mutex_unlock(&fs_devices->device_list_mutex); + + if (copy_to_user(arg, &fi_args, sizeof(fi_args))) + return -EFAULT; + + return 0; +} + +static long btrfs_ioctl_dev_info(struct btrfs_root *root, void __user *arg) +{ + struct btrfs_ioctl_dev_info_args *di_args; + struct btrfs_device *dev; + struct btrfs_fs_devices *fs_devices = root->fs_info->fs_devices; + int ret = 0; + char *s_uuid = NULL; + char empty_uuid[BTRFS_UUID_SIZE] = {0}; + + if (!capable(CAP_SYS_ADMIN)) + return -EPERM; + + di_args = memdup_user(arg, sizeof(*di_args)); + if (IS_ERR(di_args)) + return PTR_ERR(di_args); + + if (memcmp(empty_uuid, di_args->uuid, BTRFS_UUID_SIZE) != 0) + s_uuid = di_args->uuid; + + mutex_lock(&fs_devices->device_list_mutex); + dev = btrfs_find_device(root, di_args->devid, s_uuid, NULL); + mutex_unlock(&fs_devices->device_list_mutex); + + if (!dev) { + ret = -ENODEV; + goto out; + } + + di_args->devid = dev->devid; + di_args->bytes_used = dev->bytes_used; + di_args->total_bytes = dev->total_bytes; + memcpy(di_args->uuid, dev->uuid, sizeof(di_args->uuid)); + strncpy(di_args->path, dev->name, sizeof(di_args->path)); + +out: + if (ret == 0 && copy_to_user(arg, di_args, sizeof(*di_args))) + ret = -EFAULT; + + kfree(di_args); + return ret; +} + static noinline long btrfs_ioctl_clone(struct file *file, unsigned long srcfd, u64 off, u64 olen, u64 destoff) { @@ -2465,6 +2534,58 @@ static noinline long btrfs_ioctl_wait_sync(struct file *file, void __user *argp) return btrfs_wait_for_commit(root, transid); } +static long btrfs_ioctl_scrub(struct btrfs_root *root, void __user *arg) +{ + int ret; + struct btrfs_ioctl_scrub_args *sa; + + if (!capable(CAP_SYS_ADMIN)) + return -EPERM; + + sa = memdup_user(arg, sizeof(*sa)); + if (IS_ERR(sa)) + return PTR_ERR(sa); + + ret = btrfs_scrub_dev(root, sa->devid, sa->start, sa->end, + &sa->progress); + + if (copy_to_user(arg, sa, sizeof(*sa))) + ret = -EFAULT; + + kfree(sa); + return ret; +} + +static long btrfs_ioctl_scrub_cancel(struct btrfs_root *root, void __user *arg) +{ + if (!capable(CAP_SYS_ADMIN)) + return -EPERM; + + return btrfs_scrub_cancel(root); +} + +static long btrfs_ioctl_scrub_progress(struct btrfs_root *root, + void __user *arg) +{ + struct btrfs_ioctl_scrub_args *sa; + int ret; + + if (!capable(CAP_SYS_ADMIN)) + return -EPERM; + + sa = memdup_user(arg, sizeof(*sa)); + if (IS_ERR(sa)) + return PTR_ERR(sa); + + ret = btrfs_scrub_progress(root, sa->devid, &sa->progress); + + if (copy_to_user(arg, sa, sizeof(*sa))) + ret = -EFAULT; + + kfree(sa); + return ret; +} + long btrfs_ioctl(struct file *file, unsigned int cmd, unsigned long arg) { @@ -2504,6 +2625,10 @@ long btrfs_ioctl(struct file *file, unsigned int return btrfs_ioctl_add_dev(root, argp); case BTRFS_IOC_RM_DEV: return btrfs_ioctl_rm_dev(root, argp); + case BTRFS_IOC_FS_INFO: + return btrfs_ioctl_fs_info(root, argp); + case BTRFS_IOC_DEV_INFO: + return btrfs_ioctl_dev_info(root, argp); case BTRFS_IOC_BALANCE: return btrfs_balance(root->fs_info->dev_root); case BTRFS_IOC_CLONE: @@ -2527,6 +2652,12 @@ long btrfs_ioctl(struct file *file, unsigned int return btrfs_ioctl_start_sync(file, argp); case BTRFS_IOC_WAIT_SYNC: return btrfs_ioctl_wait_sync(file, argp); + case BTRFS_IOC_SCRUB: + return btrfs_ioctl_scrub(root, argp); + case BTRFS_IOC_SCRUB_CANCEL: + return btrfs_ioctl_scrub_cancel(root, argp); + case BTRFS_IOC_SCRUB_PROGRESS: + return btrfs_ioctl_scrub_progress(root, argp); } return -ENOTTY; diff --git a/fs/btrfs/ioctl.h b/fs/btrfs/ioctl.h index 37ac030d64b4..1a638ceeead8 100644 --- a/fs/btrfs/ioctl.h +++ b/fs/btrfs/ioctl.h @@ -32,6 +32,8 @@ struct btrfs_ioctl_vol_args { #define BTRFS_SUBVOL_CREATE_ASYNC (1ULL << 0) #define BTRFS_SUBVOL_RDONLY (1ULL << 1) +#define BTRFS_FSID_SIZE 16 +#define BTRFS_UUID_SIZE 16 #define BTRFS_SUBVOL_NAME_MAX 4039 struct btrfs_ioctl_vol_args_v2 { @@ -79,6 +81,33 @@ struct btrfs_scrub_progress { * Intermittent error. */ }; +struct btrfs_ioctl_scrub_args { + __u64 devid; /* in */ + __u64 start; /* in */ + __u64 end; /* in */ + __u64 flags; /* in */ + struct btrfs_scrub_progress progress; /* out */ + /* pad to 1k */ + __u64 unused[(1024-32-sizeof(struct btrfs_scrub_progress))/8]; +}; + +#define BTRFS_DEVICE_PATH_NAME_MAX 1024 +struct btrfs_ioctl_dev_info_args { + __u64 devid; /* in/out */ + __u8 uuid[BTRFS_UUID_SIZE]; /* in/out */ + __u64 bytes_used; /* out */ + __u64 total_bytes; /* out */ + __u64 unused[379]; /* pad to 4k */ + __u8 path[BTRFS_DEVICE_PATH_NAME_MAX]; /* out */ +}; + +struct btrfs_ioctl_fs_info_args { + __u64 max_id; /* out */ + __u64 num_devices; /* out */ + __u8 fsid[BTRFS_FSID_SIZE]; /* out */ + __u64 reserved[124]; /* pad to 1k */ +}; + #define BTRFS_INO_LOOKUP_PATH_MAX 4080 struct btrfs_ioctl_ino_lookup_args { __u64 treeid; @@ -240,4 +269,13 @@ struct btrfs_ioctl_space_args { struct btrfs_ioctl_vol_args_v2) #define BTRFS_IOC_SUBVOL_GETFLAGS _IOW(BTRFS_IOCTL_MAGIC, 25, __u64) #define BTRFS_IOC_SUBVOL_SETFLAGS _IOW(BTRFS_IOCTL_MAGIC, 26, __u64) +#define BTRFS_IOC_SCRUB _IOWR(BTRFS_IOCTL_MAGIC, 27, \ + struct btrfs_ioctl_scrub_args) +#define BTRFS_IOC_SCRUB_CANCEL _IO(BTRFS_IOCTL_MAGIC, 28) +#define BTRFS_IOC_SCRUB_PROGRESS _IOWR(BTRFS_IOCTL_MAGIC, 29, \ + struct btrfs_ioctl_scrub_args) +#define BTRFS_IOC_DEV_INFO _IOWR(BTRFS_IOCTL_MAGIC, 30, \ + struct btrfs_ioctl_dev_info_args) +#define BTRFS_IOC_FS_INFO _IOR(BTRFS_IOCTL_MAGIC, 31, \ + struct btrfs_ioctl_fs_info_args) #endif -- cgit v1.2.3 From 8628764e1a5e1998a42b9713e9edea7753653d01 Mon Sep 17 00:00:00 2001 From: Arne Jansen Date: Wed, 23 Mar 2011 16:34:19 +0100 Subject: btrfs: add readonly flag setting the readonly flag prevents writes in case an error is detected Signed-off-by: Arne Jansen --- fs/btrfs/ctree.h | 2 +- fs/btrfs/ioctl.c | 2 +- fs/btrfs/ioctl.h | 1 + fs/btrfs/scrub.c | 23 +++++++++++++---------- 4 files changed, 16 insertions(+), 12 deletions(-) (limited to 'fs/btrfs/ctree.h') diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index b7373b14e4cd..ee904666b766 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -2657,7 +2657,7 @@ void btrfs_reloc_post_snapshot(struct btrfs_trans_handle *trans, /* scrub.c */ int btrfs_scrub_dev(struct btrfs_root *root, u64 devid, u64 start, u64 end, - struct btrfs_scrub_progress *progress); + struct btrfs_scrub_progress *progress, int readonly); int btrfs_scrub_pause(struct btrfs_root *root); int btrfs_scrub_pause_super(struct btrfs_root *root); int btrfs_scrub_continue(struct btrfs_root *root); diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index 205cd011d2f3..f0a74f014748 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -2547,7 +2547,7 @@ static long btrfs_ioctl_scrub(struct btrfs_root *root, void __user *arg) return PTR_ERR(sa); ret = btrfs_scrub_dev(root, sa->devid, sa->start, sa->end, - &sa->progress); + &sa->progress, sa->flags & BTRFS_SCRUB_READONLY); if (copy_to_user(arg, sa, sizeof(*sa))) ret = -EFAULT; diff --git a/fs/btrfs/ioctl.h b/fs/btrfs/ioctl.h index 1a638ceeead8..e5e0ee2cad4e 100644 --- a/fs/btrfs/ioctl.h +++ b/fs/btrfs/ioctl.h @@ -81,6 +81,7 @@ struct btrfs_scrub_progress { * Intermittent error. */ }; +#define BTRFS_SCRUB_READONLY 1 struct btrfs_ioctl_scrub_args { __u64 devid; /* in */ __u64 start; /* in */ diff --git a/fs/btrfs/scrub.c b/fs/btrfs/scrub.c index 6a50801ecfa0..a31f2a9bd2e2 100644 --- a/fs/btrfs/scrub.c +++ b/fs/btrfs/scrub.c @@ -42,7 +42,6 @@ * - In case of a read error on files with nodatasum, map the file and read * the extent to trigger a writeback of the good copy * - track and record media errors, throw out bad devices - * - add a readonly mode * - add a mode to also read unallocated space * - make the prefetch cancellable */ @@ -99,6 +98,7 @@ struct scrub_dev { u16 csum_size; struct list_head csum_list; atomic_t cancel_req; + int readonly; /* * statistics */ @@ -329,14 +329,16 @@ static void scrub_fixup(struct scrub_bio *sbio, int ix) if (i == multi->num_stripes) goto uncorrectable; - /* - * bi_io_vec[ix].bv_page now contains good data, write it back - */ - if (scrub_fixup_io(WRITE, sdev->dev->bdev, - (sbio->physical + ix * PAGE_SIZE) >> 9, - sbio->bio->bi_io_vec[ix].bv_page)) { - /* I/O-error, writeback failed, give up */ - goto uncorrectable; + if (!sdev->readonly) { + /* + * bi_io_vec[ix].bv_page now contains good data, write it back + */ + if (scrub_fixup_io(WRITE, sdev->dev->bdev, + (sbio->physical + ix * PAGE_SIZE) >> 9, + sbio->bio->bi_io_vec[ix].bv_page)) { + /* I/O-error, writeback failed, give up */ + goto uncorrectable; + } } kfree(multi); @@ -1156,7 +1158,7 @@ static noinline_for_stack void scrub_workers_put(struct btrfs_root *root) int btrfs_scrub_dev(struct btrfs_root *root, u64 devid, u64 start, u64 end, - struct btrfs_scrub_progress *progress) + struct btrfs_scrub_progress *progress, int readonly) { struct scrub_dev *sdev; struct btrfs_fs_info *fs_info = root->fs_info; @@ -1209,6 +1211,7 @@ int btrfs_scrub_dev(struct btrfs_root *root, u64 devid, u64 start, u64 end, scrub_workers_put(root); return PTR_ERR(sdev); } + sdev->readonly = readonly; dev->scrub_device = sdev; atomic_inc(&fs_info->scrubs_running); -- cgit v1.2.3