linux-stable.git/fs/btrfs, branch linux-5.7.y

btrfs: fix return value mixup in btrfs_get_extent

2020-08-21T11:07:22+00:00

commit 881a3a11c2b858fe9b69ef79ac5ee9978a266dc9 upstream.

btrfs_get_extent() sets variable ret, but out: error path expect error
to be in variable err so the error code is lost.

Fixes: 6bf9e4bd6a27 ("btrfs: inode: Verify inode mode to avoid NULL pointer dereference")
CC: stable@vger.kernel.org # 5.4+
Reviewed-by: Nikolay Borisov 
Signed-off-by: Pavel Machek (CIP) 
Reviewed-by: David Sterba 
Signed-off-by: David Sterba 
Signed-off-by: Greg Kroah-Hartman

btrfs: make sure SB_I_VERSION doesn't get unset by remount

2020-08-21T11:07:22+00:00

commit faa008899a4db21a2df99833cb4ff6fa67009a20 upstream.

There's some inconsistency around SB_I_VERSION handling with mount and
remount.  Since we don't really want it to be off ever just work around
this by making sure we don't get the flag cleared on remount.

There's a tiny cpu cost of setting the bit, otherwise all changes to
i_version also change some of the times (ctime/mtime) so the inode needs
to be synced. We wouldn't save anything by disabling it.

Reported-by: Eric Sandeen 
CC: stable@vger.kernel.org # 5.4+
Signed-off-by: Josef Bacik 
Reviewed-by: David Sterba 
[ add perf impact analysis ]
Signed-off-by: David Sterba 
Signed-off-by: Greg Kroah-Hartman

btrfs: trim: fix underflow in trim length to prevent access beyond device boundary

2020-08-21T11:07:22+00:00

commit c57dd1f2f6a7cd1bb61802344f59ccdc5278c983 upstream.

[BUG]
The following script can lead to tons of beyond device boundary access:

  mkfs.btrfs -f $dev -b 10G
  mount $dev $mnt
  trimfs $mnt
  btrfs filesystem resize 1:-1G $mnt
  trimfs $mnt

[CAUSE]
Since commit 929be17a9b49 ("btrfs: Switch btrfs_trim_free_extents to
find_first_clear_extent_bit"), we try to avoid trimming ranges that's
already trimmed.

So we check device->alloc_state by finding the first range which doesn't
have CHUNK_TRIMMED and CHUNK_ALLOCATED not set.

But if we shrunk the device, that bits are not cleared, thus we could
easily got a range starts beyond the shrunk device size.

This results the returned @start and @end are all beyond device size,
then we call "end = min(end, device->total_bytes -1);" making @end
smaller than device size.

Then finally we goes "len = end - start + 1", totally underflow the
result, and lead to the beyond-device-boundary access.

[FIX]
This patch will fix the problem in two ways:

- Clear CHUNK_TRIMMED | CHUNK_ALLOCATED bits when shrinking device
  This is the root fix

- Add extra safety check when trimming free device extents
  We check and warn if the returned range is already beyond current
  device.

Link: https://github.com/kdave/btrfs-progs/issues/282
Fixes: 929be17a9b49 ("btrfs: Switch btrfs_trim_free_extents to find_first_clear_extent_bit")
CC: stable@vger.kernel.org # 5.4+
Signed-off-by: Qu Wenruo 
Reviewed-by: Filipe Manana 
Signed-off-by: David Sterba 
Signed-off-by: Greg Kroah-Hartman

btrfs: fix memory leaks after failure to lookup checksums during inode logging

2020-08-21T11:07:22+00:00

commit 4f26433e9b3eb7a55ed70d8f882ae9cd48ba448b upstream.

While logging an inode, at copy_items(), if we fail to lookup the checksums
for an extent we release the destination path, free the ins_data array and
then return immediately. However a previous iteration of the for loop may
have added checksums to the ordered_sums list, in which case we leak the
memory used by them.

So fix this by making sure we iterate the ordered_sums list and free all
its checksums before returning.

Fixes: 3650860b90cc2a ("Btrfs: remove almost all of the BUG()'s from tree-log.c")
CC: stable@vger.kernel.org # 4.4+
Reviewed-by: Johannes Thumshirn 
Signed-off-by: Filipe Manana 
Reviewed-by: David Sterba 
Signed-off-by: David Sterba 
Signed-off-by: Greg Kroah-Hartman

btrfs: inode: fix NULL pointer dereference if inode doesn't need compression

2020-08-21T11:07:21+00:00

commit 1e6e238c3002ea3611465ce5f32777ddd6a40126 upstream.

[BUG]
There is a bug report of NULL pointer dereference caused in
compress_file_extent():

  Oops: Kernel access of bad area, sig: 11 [#1]
  LE PAGE_SIZE=64K MMU=Hash SMP NR_CPUS=2048 NUMA pSeries
  Workqueue: btrfs-delalloc btrfs_delalloc_helper [btrfs]
  NIP [c008000006dd4d34] compress_file_range.constprop.41+0x75c/0x8a0 [btrfs]
  LR [c008000006dd4d1c] compress_file_range.constprop.41+0x744/0x8a0 [btrfs]
  Call Trace:
  [c000000c69093b00] [c008000006dd4d1c] compress_file_range.constprop.41+0x744/0x8a0 [btrfs] (unreliable)
  [c000000c69093bd0] [c008000006dd4ebc] async_cow_start+0x44/0xa0 [btrfs]
  [c000000c69093c10] [c008000006e14824] normal_work_helper+0xdc/0x598 [btrfs]
  [c000000c69093c80] [c0000000001608c0] process_one_work+0x2c0/0x5b0
  [c000000c69093d10] [c000000000160c38] worker_thread+0x88/0x660
  [c000000c69093db0] [c00000000016b55c] kthread+0x1ac/0x1c0
  [c000000c69093e20] [c00000000000b660] ret_from_kernel_thread+0x5c/0x7c
  ---[ end trace f16954aa20d822f6 ]---

[CAUSE]
For the following execution route of compress_file_range(), it's
possible to hit NULL pointer dereference:

 compress_file_extent()
 |- pages = NULL;
 |- start = async_chunk->start = 0;
 |- end = async_chunk = 4095;
 |- nr_pages = 1;
 |- inode_need_compress() == false; <<< Possible, see later explanation
 |  Now, we have nr_pages = 1, pages = NULL
 |- cont:
 |- 		ret = cow_file_range_inline();
 |- 		if (ret <= 0) {
 |-		for (i = 0; i < nr_pages; i++) {
 |-			WARN_ON(pages[i]->mapping);	<<< Crash

To enter above call execution branch, we need the following race:

    Thread 1 (chattr)     |            Thread 2 (writeback)
--------------------------+------------------------------
                          | btrfs_run_delalloc_range
                          | |- inode_need_compress = true
                          | |- cow_file_range_async()
btrfs_ioctl_set_flag()    |
|- binode_flags |=        |
   BTRFS_INODE_NOCOMPRESS |
                          | compress_file_range()
                          | |- inode_need_compress = false
                          | |- nr_page = 1 while pages = NULL
                          | |  Then hit the crash

[FIX]
This patch will fix it by checking @pages before doing accessing it.
This patch is only designed as a hot fix and easy to backport.

More elegant fix may make btrfs only check inode_need_compress() once to
avoid such race, but that would be another story.

Reported-by: Luciano Chavez 
Fixes: 4d3a800ebb12 ("btrfs: merge nr_pages input and output parameter in compress_pages")
CC: stable@vger.kernel.org # 4.14.x: cecc8d9038d16: btrfs: Move free_pages_out label in inline extent handling branch in compress_file_range
CC: stable@vger.kernel.org # 4.14+
Signed-off-by: Qu Wenruo 
Signed-off-by: David Sterba 
Signed-off-by: Greg Kroah-Hartman

btrfs: only search for left_info if there is no right_info in try_merge_free_space

2020-08-21T11:07:21+00:00

commit bf53d4687b8f3f6b752f091eb85f62369a515dfd upstream.

In try_to_merge_free_space we attempt to find entries to the left and
right of the entry we are adding to see if they can be merged.  We
search for an entry past our current info (saved into right_info), and
then if right_info exists and it has a rb_prev() we save the rb_prev()
into left_info.

However there's a slight problem in the case that we have a right_info,
but no entry previous to that entry.  At that point we will search for
an entry just before the info we're attempting to insert.  This will
simply find right_info again, and assign it to left_info, making them
both the same pointer.

Now if right_info _can_ be merged with the range we're inserting, we'll
add it to the info and free right_info.  However further down we'll
access left_info, which was right_info, and thus get a use-after-free.

Fix this by only searching for the left entry if we don't find a right
entry at all.

The CVE referenced had a specially crafted file system that could
trigger this use-after-free. However with the tree checker improvements
we no longer trigger the conditions for the UAF.  But the original
conditions still apply, hence this fix.

Reference: CVE-2019-19448
Fixes: 963030817060 ("Btrfs: use hybrid extents+bitmap rb tree for free space")
CC: stable@vger.kernel.org # 4.4+
Signed-off-by: Josef Bacik 
Signed-off-by: David Sterba 
Signed-off-by: Greg Kroah-Hartman

btrfs: fix messages after changing compression level by remount

2020-08-21T11:07:21+00:00

commit 27942c9971cc405c60432eca9395e514a2ae9f5e upstream.

Reported by Forza on IRC that remounting with compression options does
not reflect the change in level, or at least it does not appear to do so
according to the messages:

  mount -o compress=zstd:1 /dev/sda /mnt
  mount -o remount,compress=zstd:15 /mnt

does not print the change to the level to syslog:

  [   41.366060] BTRFS info (device vda): use zstd compression, level 1
  [   41.368254] BTRFS info (device vda): disk space caching is enabled
  [   41.390429] BTRFS info (device vda): disk space caching is enabled

What really happens is that the message is lost but the level is actualy
changed.

There's another weird output, if compression is reset to 'no':

  [   45.413776] BTRFS info (device vda): use no compression, level 4

To fix that, save the previous compression level and print the message
in that case too and use separate message for 'no' compression.

CC: stable@vger.kernel.org # 4.19+
Signed-off-by: David Sterba 
Signed-off-by: Greg Kroah-Hartman

btrfs: don't show full path of bind mounts in subvol=

2020-08-21T11:07:21+00:00

commit 3ef3959b29c4a5bd65526ab310a1a18ae533172a upstream.

Chris Murphy reported a problem where rpm ostree will bind mount a bunch
of things for whatever voodoo it's doing.  But when it does this
/proc/mounts shows something like

  /dev/sda /mnt/test btrfs rw,relatime,subvolid=256,subvol=/foo 0 0
  /dev/sda /mnt/test/baz btrfs rw,relatime,subvolid=256,subvol=/foo/bar 0 0

Despite subvolid=256 being subvol=/foo.  This is because we're just
spitting out the dentry of the mount point, which in the case of bind
mounts is the source path for the mountpoint.  Instead we should spit
out the path to the actual subvol.  Fix this by looking up the name for
the subvolid we have mounted.  With this fix the same test looks like
this

  /dev/sda /mnt/test btrfs rw,relatime,subvolid=256,subvol=/foo 0 0
  /dev/sda /mnt/test/baz btrfs rw,relatime,subvolid=256,subvol=/foo 0 0

Reported-by: Chris Murphy 
CC: stable@vger.kernel.org # 4.4+
Signed-off-by: Josef Bacik 
Reviewed-by: David Sterba 
Signed-off-by: David Sterba 
Signed-off-by: Greg Kroah-Hartman

btrfs: fix race between page release and a fast fsync

2020-08-21T11:07:21+00:00

commit 3d6448e631591756da36efb3ea6355ff6f383c3a upstream.

When releasing an extent map, done through the page release callback, we
can race with an ongoing fast fsync and cause the fsync to miss a new
extent and not log it. The steps for this to happen are the following:

1) A page is dirtied for some inode I;

2) Writeback for that page is triggered by a path other than fsync, for
   example by the system due to memory pressure;

3) When the ordered extent for the extent (a single 4K page) finishes,
   we unpin the corresponding extent map and set its generation to N,
   the current transaction's generation;

4) The btrfs_releasepage() callback is invoked by the system due to
   memory pressure for that no longer dirty page of inode I;

5) At the same time, some task calls fsync on inode I, joins transaction
   N, and at btrfs_log_inode() it sees that the inode does not have the
   full sync flag set, so we proceed with a fast fsync. But before we get
   into btrfs_log_changed_extents() and lock the inode's extent map tree:

6) Through btrfs_releasepage() we end up at try_release_extent_mapping()
   and we remove the extent map for the new 4Kb extent, because it is
   neither pinned anymore nor locked. By calling remove_extent_mapping(),
   we remove the extent map from the list of modified extents, since the
   extent map does not have the logging flag set. We unlock the inode's
   extent map tree;

7) The task doing the fast fsync now enters btrfs_log_changed_extents(),
   locks the inode's extent map tree and iterates its list of modified
   extents, which no longer has the 4Kb extent in it, so it does not log
   the extent;

8) The fsync finishes;

9) Before transaction N is committed, a power failure happens. After
   replaying the log, the 4K extent of inode I will be missing, since
   it was not logged due to the race with try_release_extent_mapping().

So fix this by teaching try_release_extent_mapping() to not remove an
extent map if it's still in the list of modified extents.

Fixes: ff44c6e36dc9dc ("Btrfs: do not hold the write_lock on the extent tree while logging")
CC: stable@vger.kernel.org # 5.4+
Signed-off-by: Filipe Manana 
Signed-off-by: David Sterba 
Signed-off-by: Greg Kroah-Hartman

btrfs: don't WARN if we abort a transaction with EROFS

2020-08-21T11:07:21+00:00

commit f95ebdbed46a4d8b9fdb7bff109fdbb6fc9a6dc8 upstream.

If we got some sort of corruption via a read and call
btrfs_handle_fs_error() we'll set BTRFS_FS_STATE_ERROR on the fs and
complain.  If a subsequent trans handle trips over this it'll get EROFS
and then abort.  However at that point we're not aborting for the
original reason, we're aborting because we've been flipped read only.
We do not need to WARN_ON() here.

CC: stable@vger.kernel.org # 5.4+
Signed-off-by: Josef Bacik 
Reviewed-by: David Sterba 
Signed-off-by: David Sterba 
Signed-off-by: Greg Kroah-Hartman