Star64_linux/fs/btrfs
Filipe Manana d36808e0d4 Btrfs: fix directory inconsistency after fsync log replay
If we have an inode (file) with a link count greater than 1, remove
one of its hard links, fsync the inode, power fail/crash and then
replay the fsync log on the next mount, we end up getting the parent
directory's metadata inconsistent - its i_size still reflects the
deleted hard link and has dangling index entries (with no matching
inode reference entries). This prevents the directory from ever being
deletable, as its i_size can never decrease to BTRFS_EMPTY_DIR_SIZE
even if all of its children inodes are deleted, and the dangling index
entries can never be removed (as they point to an inode that does not
exist anymore).

This is easy to reproduce with the following excerpt from the test case
for xfstests that I just made:

    _scratch_mkfs >> $seqres.full 2>&1

    _init_flakey
    _mount_flakey

    # Create a test file with 2 hard links in the same directory.
    mkdir -p $SCRATCH_MNT/a/b
    echo "hello world" > $SCRATCH_MNT/a/b/foo
    ln $SCRATCH_MNT/a/b/foo $SCRATCH_MNT/a/b/bar

    # Make sure all metadata and data are durably persisted.
    sync

    # Now remove one of the hard links and fsync the inode.
    rm -f $SCRATCH_MNT/a/b/bar
    $XFS_IO_PROG -c "fsync" $SCRATCH_MNT/a/b/foo

    # Simulate a crash/power loss. This makes sure the next mount
    # will see an fsync log and will replay that log.

    _load_flakey_table $FLAKEY_DROP_WRITES
    _unmount_flakey

    _load_flakey_table $FLAKEY_ALLOW_WRITES
    _mount_flakey

    # Remove the last hard link of the file and attempt to remove its parent
    # directory - this failed in btrfs because the fsync log and replay code
    # didn't decrement the parent directory's i_size and left dangling directory
    # index entries - this made the btrfs rmdir implementation always fail with
    # the error -ENOTEMPTY.
    #
    # The dangling directory index entries were visible to user space, but it was
    # impossible to do anything on them (unlink, open, read, write, stat, etc)
    # because the inode they pointed to did not exist anymore.
    #
    # The parent directory's metadata inconsistency (stale index entries) was
    # also detected by btrfs' fsck tool, which is run automatically by the fstests
    # framework when the test finishes. The error message reported by fsck was:
    #
    # root 5 inode 259 errors 2001, no inode item, link count wrong
    #   unresolved ref dir 258 index 3 namelen 3 name bar filetype 1 errors 4, no inode ref
    #
    rm -f $SCRATCH_MNT/a/b/*
    rmdir $SCRATCH_MNT/a/b
    rmdir $SCRATCH_MNT/a

To fix this just make sure that after an unlink, if the inode is fsync'ed,
he parent inode is fully logged in the fsync log.

A test case for xfstests follows soon.

Signed-off-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: Chris Mason <clm@fb.com>
2015-01-21 18:02:04 -08:00
..
tests btrfs: switch extent_state state to unsigned 2015-01-21 18:02:04 -08:00
acl.c btrfs: remove useless ACL check 2014-06-09 17:20:42 -07:00
async-thread.c btrfs: remove unlikely from NULL checks 2014-10-02 16:06:19 +02:00
async-thread.h Btrfs: implement repair function when direct read fails 2014-09-17 13:39:01 -07:00
backref.c btrfs: cleanup, remove inode_ref_info helper 2015-01-14 19:23:47 +01:00
backref.h btrfs: cleanup, remove inode_item_info helper 2015-01-14 19:23:47 +01:00
btrfs_inode.h Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs 2014-10-11 08:03:52 -04:00
check-integrity.c Btrfs: include vmalloc.h in check-integrity.c 2014-11-25 06:01:11 -08:00
check-integrity.h block: submit_bio_wait() conversions 2013-11-24 16:33:41 -07:00
compression.c Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs 2014-12-12 11:15:23 -08:00
compression.h btrfs: zero out left over bytes after processing compression streams 2014-11-30 09:33:51 -08:00
ctree.c Btrfs: fix setup_leaf_for_split() to avoid leaf corruption 2015-01-21 18:02:03 -08:00
ctree.h Btrfs: track dirty block groups on their own list 2015-01-21 17:36:52 -08:00
delayed-inode.c Btrfs: don't delay inode ref updates during log replay 2015-01-02 14:47:56 -05:00
delayed-inode.h Btrfs: introduce the delayed inode ref deletion for the single link inode 2014-01-28 13:20:09 -08:00
delayed-ref.c Btrfs: rework qgroup accounting 2014-06-09 17:20:48 -07:00
delayed-ref.h Btrfs: rework qgroup accounting 2014-06-09 17:20:48 -07:00
dev-replace.c Merge branch 'raid56-scrub-replace' of git://github.com/miaoxie/linux-btrfs into for-linus 2014-12-02 18:42:03 -08:00
dev-replace.h
dir-item.c Btrfs: make xattr replace operations atomic 2014-11-20 17:20:07 -08:00
disk-io.c btrfs: set proper message level for skinny metadata 2015-01-21 18:02:03 -08:00
disk-io.h btrfs: sink blocksize parameter to btrfs_find_create_tree_block 2014-12-12 18:07:21 +01:00
export.c btrfs: kill the key type accessor helpers 2014-09-17 13:37:12 -07:00
export.h
extent-tree.c Btrfs: lookup for block group only if needed when freeing a tree block 2015-01-21 18:02:04 -08:00
extent_io.c btrfs: switch extent_state state to unsigned 2015-01-21 18:02:04 -08:00
extent_io.h btrfs: switch extent_state state to unsigned 2015-01-21 18:02:04 -08:00
extent_map.c Btrfs: do not move em to modified list when unpinning 2014-11-21 11:59:54 -08:00
extent_map.h Btrfs: fix NULL pointer crash when running balance and scrub concurrently 2014-06-19 14:20:55 -07:00
file-item.c Btrfs: fix kfree on list_head in btrfs_lookup_csums_range error cleanup 2014-11-04 06:59:04 -08:00
file.c Btrfs: fix snapshot inconsistency after a file write followed by truncate 2014-11-25 07:41:23 -08:00
free-space-cache.c Btrfs: track dirty block groups on their own list 2015-01-21 17:36:52 -08:00
free-space-cache.h Btrfs: fix race between writing free space cache and trimming 2014-12-02 18:35:09 -08:00
hash.c btrfs: LLVMLinux: Remove VLAIS 2014-10-14 10:51:22 +02:00
hash.h Btrfs: fix btrfs boot when compiled as built-in 2014-01-28 13:20:31 -08:00
inode-item.c btrfs: kill the key type accessor helpers 2014-09-17 13:37:12 -07:00
inode-map.c Btrfs: fix race between writing free space cache and trimming 2014-12-02 18:35:09 -08:00
inode-map.h
inode.c btrfs: switch extent_state state to unsigned 2015-01-21 18:02:04 -08:00
ioctl.c Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs 2014-12-12 11:15:23 -08:00
Kconfig Btrfs: fix btrfs boot when compiled as built-in 2014-01-28 13:20:31 -08:00
locking.c btrfs: fix lockups from btrfs_clear_path_blocking 2014-11-19 10:34:35 -08:00
locking.h btrfs: fix lockups from btrfs_clear_path_blocking 2014-11-19 10:34:35 -08:00
lzo.c btrfs: zero out left over bytes after processing compression streams 2014-11-30 09:33:51 -08:00
Makefile Btrfs: add sanity tests for new qgroup accounting code 2014-06-09 17:20:49 -07:00
math.h
ordered-data.c Btrfs: collect only the necessary ordered extents on ranged fsync 2014-11-21 11:59:56 -08:00
ordered-data.h Btrfs: collect only the necessary ordered extents on ranged fsync 2014-11-21 11:59:56 -08:00
orphan.c btrfs: kill the key type accessor helpers 2014-09-17 13:37:12 -07:00
print-tree.c btrfs: remove parameter blocksize from read_tree_block 2014-10-02 17:14:50 +02:00
print-tree.h btrfs: make static code static & remove dead code 2013-05-06 15:55:23 -04:00
props.c Btrfs: add support for inode properties 2014-01-28 13:20:24 -08:00
props.h Btrfs: add support for inode properties 2014-01-28 13:20:24 -08:00
qgroup.c btrfs: move checks for DUMMY_ROOT into a helper 2014-10-02 17:30:33 +02:00
qgroup.h btrfs: qgroup: account shared subtrees during snapshot delete 2014-08-15 07:43:14 -07:00
raid56.c Btrfs, raid56: fix use-after-free problem in the final device replace procedure on raid56 2014-12-03 10:18:47 +08:00
raid56.h Btrfs, raid56: fix use-after-free problem in the final device replace procedure on raid56 2014-12-03 10:18:47 +08:00
rcu-string.h
reada.c btrfs: sink blocksize parameter to reada_tree_block_flagged 2014-12-12 18:07:20 +01:00
relocation.c btrfs: sink blocksize parameter to tree_block_processed 2014-12-12 18:07:22 +01:00
root-tree.c Btrfs: use bitfield instead of integer data type for the some variants in btrfs_root 2014-06-09 17:20:40 -07:00
scrub.c btrfs: expand btrfs_find_item if found_key is NULL 2015-01-14 19:23:48 +01:00
send.c Btrfs: ensure send always works on roots without orphans 2014-11-25 07:41:23 -08:00
send.h btrfs: make static code static & remove dead code 2013-05-06 15:55:23 -04:00
struct-funcs.c
super.c btrfs: remove a no-op unfreeze superbock callback 2015-01-21 18:02:04 -08:00
sysfs.c btrfs: move commit out of sysfs when changing label 2014-11-12 16:53:15 +01:00
sysfs.h btrfs: code optimize: BTRFS_ATTR_RW could set the mode 2014-09-17 13:37:59 -07:00
transaction.c Btrfs: track dirty block groups on their own list 2015-01-21 17:36:52 -08:00
transaction.h Btrfs: track dirty block groups on their own list 2015-01-21 17:36:52 -08:00
tree-defrag.c Btrfs: use bitfield instead of integer data type for the some variants in btrfs_root 2014-06-09 17:20:40 -07:00
tree-log.c Btrfs: fix directory inconsistency after fsync log replay 2015-01-21 18:02:04 -08:00
tree-log.h Btrfs: fix data corruption after fast fsync and writeback error 2014-09-19 06:57:51 -07:00
ulist.c Btrfs: do not export ulist functions 2014-01-29 07:06:27 -08:00
ulist.h Btrfs: Fix memory corruption by ulist_add_merge() on 32bit arch 2014-08-15 07:43:19 -07:00
uuid-tree.c Btrfs: make btrfs_search_forward return with nodes unlocked 2014-09-17 13:38:02 -07:00
volumes.c Merge branch 'cleanup/blocksize-diet-part2' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux into for-linus 2015-01-21 17:49:35 -08:00
volumes.h Merge branch 'raid56-scrub-replace' of git://github.com/miaoxie/linux-btrfs into for-linus 2014-12-02 18:42:03 -08:00
xattr.c Btrfs: make xattr replace operations atomic 2014-11-20 17:20:07 -08:00
xattr.h btrfs: use generic posix ACL infrastructure 2014-01-25 23:58:18 -05:00
zlib.c btrfs: zero out left over bytes after processing compression streams 2014-11-30 09:33:51 -08:00