1
0
mirror of https://github.com/Zygo/bees.git synced 2025-05-17 21:35:45 +02:00

docs: btrfs-kernel: 4.20 adds 32-bit single convert bug, tree mod log issue #4

There was a 4th tree mod log crash that showed up in testing.  It can
be reproduced or eliminated by applying or reverting d2311e698578
("btrfs: relocation: Delay reloc tree deletion after merge_reloc_roots")
to a 5.4.x kernel before 5.4.54.

Unfortunately, the test can only run if several other patches that
fixed other bugs in d2311e698578 are applied or removed at the same time.
Commit d2311e698578 introduces a bug which destroys filesystems under test
long before tree mod log failures can be reproduced in testing.  One of
those patches also fixes tree mod log issue #4.  I do not know which one,
but since kernels after 5.1 cannot run without all of those patches, I do
not think it matters.

Tree mod issue #4 is the reason why the tree mod workaround is still
required on all kernels before 5.4.  The issue still exists on older
LTS kernels, e.g. 4.9.

Signed-off-by: Zygo Blaxell <bees@furryterror.org>
This commit is contained in:
Zygo Blaxell 2020-10-09 21:02:29 -04:00
parent 217f5c781b
commit 1b9b437c11

View File

@ -39,10 +39,10 @@ These bugs are particularly popular among bees users:
| - | 5.0 | deadlock when dedupe and rename are used simultaneously on the same files | 5.0.4, 5.1 and later | 4ea748e1d2c9 Btrfs: fix deadlock between clone/dedupe and rename | - | 5.0 | deadlock when dedupe and rename are used simultaneously on the same files | 5.0.4, 5.1 and later | 4ea748e1d2c9 Btrfs: fix deadlock between clone/dedupe and rename
| - | 5.1 | send failure or kernel crash while running send and dedupe on same snapshot at same time | 5.0.18, 5.1.4, 5.2 and later | 62d54f3a7fa2 Btrfs: fix race between send and deduplication that lead to failures and crashes | - | 5.1 | send failure or kernel crash while running send and dedupe on same snapshot at same time | 5.0.18, 5.1.4, 5.2 and later | 62d54f3a7fa2 Btrfs: fix race between send and deduplication that lead to failures and crashes
| - | 5.2 | alternating send and dedupe results in incremental send failure | 4.9.188, 4.14.137, 4.19.65, 5.2.7, 5.3 and later | b4f9a1a87a48 Btrfs: fix incremental send failure after deduplication | - | 5.2 | alternating send and dedupe results in incremental send failure | 4.9.188, 4.14.137, 4.19.65, 5.2.7, 5.3 and later | b4f9a1a87a48 Btrfs: fix incremental send failure after deduplication
| - | 5.3 | balance convert to single rejected with error on 32-bit CPUs | 5.3.7, 5.4 and later | 7a54789074a5 btrfs: fix balance convert to single on 32-bit host CPUs | 4.20 | 5.3 | balance convert to single rejected with error on 32-bit CPUs | 5.3.7, 5.4 and later | 7a54789074a5 btrfs: fix balance convert to single on 32-bit host CPUs
| - | 5.3 | kernel crash due to tree mod log issue #1 (often triggered by bees) | 3.16.79, 4.4.195, 4.9.195, 4.14.147, 4.19.77, 5.2.19, 5.3.4, 5.4 and later | efad8a853ad2 Btrfs: fix use-after-free when using the tree modification log | - | 5.3 | kernel crash due to tree mod log issue #1 (often triggered by bees) | 3.16.79, 4.4.195, 4.9.195, 4.14.147, 4.19.77, 5.2.19, 5.3.4, 5.4 and later | efad8a853ad2 Btrfs: fix use-after-free when using the tree modification log
| - | 5.4 | kernel crash due to tree mod log issue #2 (often triggered by bees) | 3.16.83, 4.4.208, 4.9.208, 4.14.161, 4.19.92, 5.4.7, 5.5 and later | 6609fee8897a Btrfs: fix removal logic of the tree mod log that leads to use-after-free issues | - | 5.4 | kernel crash due to tree mod log issue #2 (often triggered by bees) | 3.16.83, 4.4.208, 4.9.208, 4.14.161, 4.19.92, 5.4.7, 5.5 and later | 6609fee8897a Btrfs: fix removal logic of the tree mod log that leads to use-after-free issues
| 5.1 | 5.4 | metadata corruption resulting in loss of filesystem when a write operation occurs while balance starts a new block group. **Do not use kernel 5.1 with btrfs.** Kernel 5.2 and 5.3 have workarounds that may detect corruption in progress and abort before it becomes permanent, but do not prevent corruption from occurring. | 5.4.14, 5.5 and later | 6282675e6708 btrfs: relocation: fix reloc_root lifespan and access | 5.1 | 5.4 | metadata corruption resulting in loss of filesystem when a write operation occurs while balance starts a new block group. **Do not use kernel 5.1 with btrfs.** Kernel 5.2 and 5.3 have workarounds that may detect corruption in progress and abort before it becomes permanent, but do not prevent corruption from occurring. Also kernel crash due to tree mod log issue #4. | 5.4.14, 5.5 and later | 6282675e6708 btrfs: relocation: fix reloc_root lifespan and access
| - | 5.4 | send performance failure when shared extents have too many references | 4.9.207, 4.14.159, 4.19.90, 5.3.17, 5.4.4, 5.5 and later | fd0ddbe25095 Btrfs: send, skip backreference walking for extents with many references | - | 5.4 | send performance failure when shared extents have too many references | 4.9.207, 4.14.159, 4.19.90, 5.3.17, 5.4.4, 5.5 and later | fd0ddbe25095 Btrfs: send, skip backreference walking for extents with many references
| 5.0 | 5.5 | dedupe fails to remove the last extent in a file if the file size is not a multiple of 4K | 5.4.19, 5.5.3, 5.6 and later | 831d2fa25ab8 Btrfs: make deduplication with range including the last block work | 5.0 | 5.5 | dedupe fails to remove the last extent in a file if the file size is not a multiple of 4K | 5.4.19, 5.5.3, 5.6 and later | 831d2fa25ab8 Btrfs: make deduplication with range including the last block work
| 4.5, backported to 3.18.31, 4.1.22, 4.4.4 | 5.5 | `df` incorrectly reports 0 free space while data space is available. Triggered by changes in metadata size, including those typical of large-scale dedupe. Occurs more often starting in 5.3 and especially 5.4 | 4.4.213, 4.9.213, 4.14.170, 4.19.102, 5.4.18, 5.5.2, 5.6 and later | d55966c4279b btrfs: do not zero f_bavail if we have available space | 4.5, backported to 3.18.31, 4.1.22, 4.4.4 | 5.5 | `df` incorrectly reports 0 free space while data space is available. Triggered by changes in metadata size, including those typical of large-scale dedupe. Occurs more often starting in 5.3 and especially 5.4 | 4.4.213, 4.9.213, 4.14.170, 4.19.102, 5.4.18, 5.5.2, 5.6 and later | d55966c4279b btrfs: do not zero f_bavail if we have available space