linux-toradex.git/fs/btrfs/tree-log.c, branch v4.9.117

btrfs: add barriers to btrfs_sync_log before log_commit_wait wakeups

2018-08-03T05:55:17+00:00

[ Upstream commit 3d3a2e610ea5e7c6d4f9481ecce5d8e2d8317843 ]

Currently the code assumes that there's an implied barrier by the
sequence of code preceding the wakeup, namely the mutex unlock.

As Nikolay pointed out:

I think this is wrong (not your code) but the original assumption that
the RELEASE semantics provided by mutex_unlock is sufficient.
According to memory-barriers.txt:

Section 'LOCK ACQUISITION FUNCTIONS' states:

 (2) RELEASE operation implication:

     Memory operations issued before the RELEASE will be completed before the
     RELEASE operation has completed.

     Memory operations issued after the RELEASE *may* be completed before the
     RELEASE operation has completed.

(I've bolded the may portion)

The example given there:

As an example, consider the following:

    *A = a;
    *B = b;
    ACQUIRE
    *C = c;
    *D = d;
    RELEASE
    *E = e;
    *F = f;

The following sequence of events is acceptable:

    ACQUIRE, {*F,*A}, *E, {*C,*D}, *B, RELEASE

So if we assume that *C is modifying the flag which the waitqueue is checking,
and *E is the actual wakeup, then those accesses can be re-ordered...

IMHO this code should be considered broken...
Signed-off-by: Greg Kroah-Hartman

Btrfs: fix copy_items() return value when logging an inode

2018-05-30T05:50:42+00:00

[ Upstream commit 8434ec46c6e3232cebc25a910363b29f5c617820 ]

When logging an inode, at tree-log.c:copy_items(), if we call
btrfs_next_leaf() at the loop which checks for the need to log holes, we
need to make sure copy_items() returns the value 1 to its caller and
not 0 (on success). This is because the path the caller passed was
released and is now different from what is was before, and the caller
expects a return value of 0 to mean both success and that the path
has not changed, while a return value of 1 means both success and
signals the caller that it can not reuse the path, it has to perform
another tree search.

Even though this is a case that should not be triggered on normal
circumstances or very rare at least, its consequences can be very
unpredictable (especially when replaying a log tree).

Fixes: 16e7549f045d ("Btrfs: incompatible format change to remove hole extents")
Signed-off-by: Filipe Manana 
Signed-off-by: David Sterba 
Signed-off-by: Sasha Levin 
Signed-off-by: Greg Kroah-Hartman

Btrfs: fix NULL pointer dereference in log_dir_items

2018-05-30T05:50:41+00:00

[ Upstream commit 80c0b4210a963e31529e15bf90519708ec947596 ]

0, 1 and <0 can be returned by btrfs_next_leaf(), and when <0 is
returned, path->nodes[0] could be NULL, log_dir_items lacks such a
check for <0 and we may run into a null pointer dereference panic.

Fixes: e02119d5a7b4 ("Btrfs: Add a write ahead tree log to optimize synchronous operations")
Reviewed-by: Nikolay Borisov 
Signed-off-by: Liu Bo 
Signed-off-by: David Sterba 
Signed-off-by: Sasha Levin 
Signed-off-by: Greg Kroah-Hartman

Btrfs: bail out on error during replay_dir_deletes

2018-05-30T05:50:41+00:00

[ Upstream commit b98def7ca6e152ee55e36863dddf6f41f12d1dc6 ]

If errors were returned by btrfs_next_leaf(), replay_dir_deletes needs
to bail out, otherwise @ret would be forced to be 0 after 'break;' and
the caller won't be aware of it.

Fixes: e02119d5a7b4 ("Btrfs: Add a write ahead tree log to optimize synchronous operations")
Reviewed-by: Nikolay Borisov 
Signed-off-by: Liu Bo 
Signed-off-by: David Sterba 
Signed-off-by: Sasha Levin 
Signed-off-by: Greg Kroah-Hartman

Btrfs: fix xattr loss after power failure

2018-05-22T14:57:58+00:00

commit 9a8fca62aacc1599fea8e813d01e1955513e4fad upstream.

If a file has xattrs, we fsync it, to ensure we clear the flags
BTRFS_INODE_NEEDS_FULL_SYNC and BTRFS_INODE_COPY_EVERYTHING from its
inode, the current transaction commits and then we fsync it (without
either of those bits being set in its inode), we end up not logging
all its xattrs. This results in deleting all xattrs when replying the
log after a power failure.

Trivial reproducer

  $ mkfs.btrfs -f /dev/sdb
  $ mount /dev/sdb /mnt

  $ touch /mnt/foobar
  $ setfattr -n user.xa -v qwerty /mnt/foobar
  $ xfs_io -c "fsync" /mnt/foobar

  $ sync

  $ xfs_io -c "pwrite -S 0xab 0 64K" /mnt/foobar
  $ xfs_io -c "fsync" /mnt/foobar
  

  $ mount /dev/sdb /mnt
  $ getfattr --absolute-names --dump /mnt/foobar
  
  $

So fix this by making sure all xattrs are logged if we log a file's inode
item and neither the flags BTRFS_INODE_NEEDS_FULL_SYNC nor
BTRFS_INODE_COPY_EVERYTHING were set in the inode.

Fixes: 36283bf777d9 ("Btrfs: fix fsync xattr loss in the fast fsync path")
Cc:  # 4.2+
Signed-off-by: Filipe Manana 
Signed-off-by: David Sterba 
Signed-off-by: Greg Kroah-Hartman

btrfs: Remove extra parentheses from condition in copy_items()

2018-04-08T10:12:47+00:00

commit 0dde10bed2c44a4024eb446cc72fe4e0cb97ec06 upstream.

There is no need for the extra pair of parentheses, remove it. This
fixes the following warning when building with clang:

fs/btrfs/tree-log.c:3694:10: warning: equality comparison with extraneous
  parentheses [-Wparentheses-equality]
                if ((i == (nr - 1)))
                     ~~^~~~~~~~~~~

Also remove the unnecessary parentheses around the substraction.

Signed-off-by: Matthias Kaehlcke 
Signed-off-by: David Sterba 
Cc: Nathan Chancellor 
Signed-off-by: Greg Kroah-Hartman

Btrfs: fix unexpected -EEXIST when creating new inode

2018-02-22T14:43:50+00:00

commit 900c9981680067573671ecc5cbfa7c5770be3a40 upstream.

The highest objectid, which is assigned to new inode, is decided at
the time of initializing fs roots.  However, in cases where log replay
gets processed, the btree which fs root owns might be changed, so we
have to search it again for the highest objectid, otherwise creating
new inode would end up with -EEXIST.

cc:  v4.4-rc6+
Fixes: f32e48e92596 ("Btrfs: Initialize btrfs_root->highest_objectid when loading tree root and subvolume roots")
Signed-off-by: Liu Bo 
Reviewed-by: Josef Bacik 
Signed-off-by: David Sterba 
Signed-off-by: Greg Kroah-Hartman

Btrfs: fix extent state leak from tree log

2018-02-22T14:43:49+00:00

commit 55237a5f2431a72435e3ed39e4306e973c0446b7 upstream.

It's possible that btrfs_sync_log() bails out after one of the two
btrfs_write_marked_extents() which convert extent state's state bit into
EXTENT_NEED_WAIT from EXTENT_DIRTY/EXTENT_NEW, however only EXTENT_DIRTY
and EXTENT_NEW are searched by free_log_tree() so that those extent states
with EXTENT_NEED_WAIT lead to memory leak.

cc: 
Signed-off-by: Liu Bo 
Reviewed-by: Josef Bacik 
Signed-off-by: David Sterba 
Signed-off-by: Greg Kroah-Hartman

Btrfs: fix crash due to not cleaning up tree log block's dirty bits

2018-02-22T14:43:49+00:00

commit 1846430c24d66e85cc58286b3319c82cd54debb2 upstream.

In cases that the whole fs flips into readonly status due to failures in
critical sections, then log tree's blocks are still dirty, and this leads
to a crash during umount time, the crash is about use-after-free,

umount
 -> close_ctree
    -> stop workers
    -> iput(btree_inode)
       -> iput_final
          -> write_inode_now
	     -> ...
	       -> queue job on stop'd workers

cc:  v3.12+
Fixes: 681ae50917df ("Btrfs: cleanup reserved space when freeing tree log on error")
Signed-off-by: Liu Bo 
Reviewed-by: Josef Bacik 
Signed-off-by: David Sterba 
Signed-off-by: Greg Kroah-Hartman

Btrfs: fix lockdep warning about log_mutex

2017-08-07T01:59:47+00:00

[ Upstream commit 781feef7e6befafd4d9787d1f7ada1f9ccd504e4 ]

While checking INODE_REF/INODE_EXTREF for a corner case, we may acquire a
different inode's log_mutex with holding the current inode's log_mutex, and
lockdep has complained this with a possilble deadlock warning.

Fix this by using mutex_lock_nested() when processing the other inode's
log_mutex.

Reviewed-by: Filipe Manana 
Signed-off-by: Liu Bo 
Signed-off-by: David Sterba 
Signed-off-by: Sasha Levin 
Signed-off-by: Greg Kroah-Hartman