|
Bugzilla – Full Text Bug Listing |
| Summary: | VUL-0: CVE-2022-48734: kernel: btrfs: fix deadlock between quota disable and qgroup rescan worker | ||
|---|---|---|---|
| Product: | [Novell Products] SUSE Security Incidents | Reporter: | SMASH SMASH <smash_bz> |
| Component: | Incidents | Assignee: | Wenruo Qu <wqu> |
| Status: | NEW --- | QA Contact: | Security Team bot <security-team> |
| Severity: | Normal | ||
| Priority: | P3 - Medium | CC: | carlos.lopez, gabriel.bertazi |
| Version: | unspecified | ||
| Target Milestone: | --- | ||
| Hardware: | Other | ||
| OS: | Other | ||
| URL: | https://smash.suse.de/issue/411576/ | ||
| Whiteboard: | CVSSv3.1:SUSE:CVE-2022-48734:4.1:(AV:L/AC:H/PR:H/UI:N/S:U/C:N/I:N/A:H) | ||
| Found By: | Security Response Team | Services Priority: | |
| Business Priority: | Blocker: | --- | |
| Marketing QA Status: | --- | IT Deployment: | --- |
Ok so this is a nightmare to backport to SLE12-SP5 (and probably SLE12-SP3-TD too, but I haven't checked that one).
A large part of the problem is that we have several, quite intrusive, qgroup patches in SLE12-SP5 that were never
upstreamed. From series.conf the patches are:
1) patches.suse/btrfs-suspend-qgroups-during-relocation-recovery.patch
This one is because for some reason I'm not aware of, the following upstream patch was reverted:
patches.suse/revert-btrfs-qgroup-move-half-of-the-qgroup-accounting-time-out-of-commit-trans.patch
It was authored by Jeff.
2) patches.suse/btrfs-qgroups-fix-rescan-worker-running-races.patch
Fixes a race, was submitted upstream but never picked.
Qu submitted upstream:
https://lore.kernel.org/linux-btrfs/20200206081741.9023-1-wqu@suse.com/
But it was nacked and there was never a followup AFAICS (unless the subject changed and the patch is very different).
There was also an earlier submission from Jeff, and Qu commented on it:
https://lore.kernel.org/linux-btrfs/16f05375-38ef-4694-a124-e7e83d83f6ac@gmx.com/
There was no direct reply from Jeff there, but there's a reply somewhere else in the thread
where he (Jeff) mentions to hold off the patches:
https://lore.kernel.org/linux-btrfs/504675d4-2a5c-2972-afd7-17dfdad71845@suse.com/
But there was never a followup, things got stale and we have a rather large and invasive patch
in SLE12-SP5 that conflicts with backports of upstream fixes (more on this below).
3) patches.suse/btrfs-qgroup-Fix-a-bug-that-prevents-qgroup-to-be-re.patch
Authored by Qu, but never upstreamed.
This seems to be needed so that the suspend/resume qgroup accounting during balance works
(introduced with the first patch I mentioned).
Because btrfs_quota_suspend() sets the BTRFS_FS_QUOTA_DISABLING bit, and then
btrfs_run_qgroups() clears the BTRFS_FS_QUOTA_ENABLED bit if it sees that bit,
which prevents the expensive qgroup accounting from running during the commit
of a transaction used by relocation.
However I don't understand how it makes a difference, because quota_root is not NULL
if we suspend balance.
There's at least one more non-upstream patch for qgroups from Qu:
patches.suse/btrfs-qgroup-Mark-qgroup-inconsistent-if-we-re-inher.patch
But that shouldn't cause any problem backporting this CVE fix (and some fixes for that CVE fix).
So the CVE fix from this bugzilla:
commit e804861bd4e69cc5fe1053eedcb024982dde8e48 v5.17-rc3
Author: Shin'ichiro Kawasaki <shinichiro.kawasaki@wdc.com>
Date: Thu Jan 20 20:09:16 2022 +0900
btrfs: fix deadlock between quota disable and qgroup rescan worker
Is valid, it affects SLE12-SP5, but it doesn't apply cleanly to SLE12-SP5, because there's
a lot of changed code. One of the things that we need is to make the transaction to be
started by btrfs_quota_disable(), and that was done upstream by:
commit 340f1aa27f367e0c3d2662d44f356b05087fea05 4.19-rc1
Author: Nikolay Borisov <nborisov@suse.com>
Date: Thu Jul 5 14:50:48 2018 +0300
btrfs: qgroups: Move transaction management inside btrfs_quota_enable/disable
But we can't apply it. It depends on another upstream patch:
commit 5d23515be66904fa3b1b5d6bd72d2199cd2447ab 4.17-rc1
Author: Nikolay Borisov <nborisov@suse.com>
Date: Wed Jan 31 10:52:04 2018 +0200
btrfs: Move qgroup rescan on quota enable to btrfs_quota_enable
And these 2 patches conflict with patch 3 and I think it conflicts with patch 1 as
well, because if we suspended qgroup accounting then after we resume (call btrfs_quota_resume()),
I suppose we expect btrfs_run_qgroups() to start the qgroup worker. But this later patch removed
the starting of the rescan worker...
And that last patch depends on another one that removes BTRFS_FS_QUOTA_DISABLING:
commit c2faff790ccd11ea5be8e3ca99713f116fcd6030 v4.14-rc3
Author: Misono, Tomohiro <misono.tomohiro@jp.fujitsu.com>
Date: Wed Aug 30 16:33:16 2017 +0900
btrfs: remove BTRFS_FS_QUOTA_DISABLING flag
Which is needed by the suspend/resume of patch 1 as mentioned before.
As for patch 2, that fixes races regarding the rescan worker start, I believe it was fixed upstream
in a different way and by several patches:
commit 13fc1d271a2e3ab8a02071e711add01fab9271f6 v5.4-rc1
Author: Filipe Manana <fdmanana@suse.com>
Date: Tue Sep 24 10:49:54 2019 +0100
Btrfs: fix race setting up and completing qgroup rescan workers
commit d61acbbf54c612ea9bf67eed609494cda0857b3a v5.7-rc1
Author: Qu Wenruo <wqu@suse.com>
Date: Fri Feb 7 13:38:20 2020 +0800
btrfs: qgroup: ensure qgroup_rescan_running is only set when the worker is at least queued
commit daf475c9154bb89a7c219e15386eebad8e1142d3 v5.7-rc1
Author: Qu Wenruo <wqu@suse.com>
Date: Fri Feb 7 13:38:21 2020 +0800
btrfs: qgroup: Remove the unnecesaary spin lock for qgroup_rescan_running
commit a855fbe69229078cd8aecd8974fb996a5ca651e6 5.10-rc6
Author: Filipe Manana <fdmanana@suse.com>
Date: Mon Nov 23 18:31:02 2020 +0000
btrfs: fix lockdep splat when enabling and disabling qgroups
But again applying these patches, depends on the patches I mentioned before from Nikolay
and Misono.
Furthermore, that CVE patch introduced some regressions, and these commits fix them:
commit d4aef1e122d8bbdc15ce3bd0bc813d6b44a7d63a v5.17-rc7
Author: Sidong Yang <realwakka@gmail.com>
Date: Mon Feb 28 01:43:40 2022 +0000
btrfs: qgroup: fix deadlock between rescan worker and remove qgroup
commit 331cd9461412e103d07595a10289de90004ac890 v6.1-rc1
Author: Filipe Manana <fdmanana@suse.com>
Date: Tue Aug 23 12:45:42 2022 +0100
btrfs: fix race between quota enable and quota rescan ioctl
commit b7adbf9ada3513d2092362c8eac5cddc5b651f5c v6.2-rc5
Author: Filipe Manana <fdmanana@suse.com>
Date: Thu Jan 12 16:31:08 2023 +0000
btrfs: fix race between quota rescan and disable leading to NULL pointer deref
But again these need to applied on top of all those patches...
Not sure what to do here, and I don't understand why we have these very intrusive and non-upstreamed patches.
I'm not an expert in the qgroup accounting part, far from it, and patches 1 and 3 are all about it.
Qu is an expert in this area, and the reverted upstream patch was authored by him as well as patch 3,
besides following on the upstream threads I mentioned above.
Qu, what can we do here?
Can you help somehow?
Can we get rid of all these non-upstream patches, re-apply your accounting patch, backport any
performance related qgroup accounting patches so that we don't need that suspend/resume hack?
This would allow backporting all these bug/CVE fixes.
I'm afraid that in the future this will become worse and worse due these non-upstream patches.
So the order of patches to apply, after getting rid of those non-upstreamed patches, would be:
commit c2faff790ccd11ea5be8e3ca99713f116fcd6030 4.14-rc3
Author: Misono, Tomohiro <misono.tomohiro@jp.fujitsu.com>
Date: Wed Aug 30 16:33:16 2017 +0900
btrfs: remove BTRFS_FS_QUOTA_DISABLING flag
commit 5d23515be66904fa3b1b5d6bd72d2199cd2447ab 4.17-rc1
Author: Nikolay Borisov <nborisov@suse.com>
Date: Wed Jan 31 10:52:04 2018 +0200
btrfs: Move qgroup rescan on quota enable to btrfs_quota_enable
commit 340f1aa27f367e0c3d2662d44f356b05087fea05 4.19-rc1
Author: Nikolay Borisov <nborisov@suse.com>
Date: Thu Jul 5 14:50:48 2018 +0300
btrfs: qgroups: Move transaction management inside btrfs_quota_enable/disable
commit 13fc1d271a2e3ab8a02071e711add01fab9271f6 v5.4-rc1
Author: Filipe Manana <fdmanana@suse.com>
Date: Tue Sep 24 10:49:54 2019 +0100
Btrfs: fix race setting up and completing qgroup rescan workers
commit 37d02592f11bb76e4ab1dcaa5b8a2a0715403207 v5.6-rc1
Author: Nikolay Borisov <nborisov@suse.com>
Date: Mon Nov 18 14:16:44 2019 +0200
btrfs: Fix error messages in qgroup_rescan_init
commit d61acbbf54c612ea9bf67eed609494cda0857b3a v5.7-rc1
Author: Qu Wenruo <wqu@suse.com>
Date: Fri Feb 7 13:38:20 2020 +0800
btrfs: qgroup: ensure qgroup_rescan_running is only set when the worker is at least queued
commit daf475c9154bb89a7c219e15386eebad8e1142d3 v5.7-rc1
Author: Qu Wenruo <wqu@suse.com>
Date: Fri Feb 7 13:38:21 2020 +0800
btrfs: qgroup: Remove the unnecesaary spin lock for qgroup_rescan_running
commit a855fbe69229078cd8aecd8974fb996a5ca651e6 5.10-rc6
Author: Filipe Manana <fdmanana@suse.com>
Date: Mon Nov 23 18:31:02 2020 +0000
btrfs: fix lockdep splat when enabling and disabling qgroups
commit e804861bd4e69cc5fe1053eedcb024982dde8e48 v5.17-rc3
Author: Shin'ichiro Kawasaki <shinichiro.kawasaki@wdc.com>
Date: Thu Jan 20 20:09:16 2022 +0900
btrfs: fix deadlock between quota disable and qgroup rescan worker
commit d4aef1e122d8bbdc15ce3bd0bc813d6b44a7d63a v5.17-rc7
Author: Sidong Yang <realwakka@gmail.com>
Date: Mon Feb 28 01:43:40 2022 +0000
btrfs: qgroup: fix deadlock between rescan worker and remove qgroup
commit 331cd9461412e103d07595a10289de90004ac890 v6.1-rc1
Author: Filipe Manana <fdmanana@suse.com>
Date: Tue Aug 23 12:45:42 2022 +0100
btrfs: fix race between quota enable and quota rescan ioctl
commit b7adbf9ada3513d2092362c8eac5cddc5b651f5c v6.2-rc5
Author: Filipe Manana <fdmanana@suse.com>
Date: Thu Jan 12 16:31:08 2023 +0000
btrfs: fix race between quota rescan and disable leading to NULL pointer deref
Also:
commit 2f1a6be12ab6c8470d5776e68644726c94257c54 v6.3-rc5
Author: Filipe Manana <fdmanana@suse.com>
Date: Wed Mar 22 10:33:28 2023 +0000
btrfs: fix race between quota disable and quota assign ioctls
This is already in SLE12-SP5 but in the unsorted section because of those non-upstream patches and missing upstream patches.
So I would love to get the non-upstream patches removed so that this could go into the sorted section.
I might be missing some patches, but those should be the main ones.
SUSE-SU-2024:2372-1: An update that solves 249 vulnerabilities, contains three features and has 45 security fixes can now be installed. Category: security (important) Bug References: 1156395, 1190336, 1191958, 1193883, 1194826, 1195065, 1195254, 1195341, 1195349, 1195775, 1196746, 1197915, 1198014, 1199295, 1202767, 1202780, 1205205, 1207361, 1217912, 1218148, 1218570, 1218820, 1219224, 1219633, 1219847, 1220368, 1220812, 1220958, 1221086, 1221282, 1221958, 1222015, 1222072, 1222080, 1222241, 1222254, 1222364, 1222893, 1223013, 1223018, 1223265, 1223384, 1223641, 1224020, 1224331, 1224488, 1224497, 1224498, 1224504, 1224520, 1224539, 1224540, 1224552, 1224583, 1224588, 1224602, 1224603, 1224605, 1224612, 1224614, 1224619, 1224661, 1224662, 1224670, 1224671, 1224674, 1224677, 1224679, 1224696, 1224703, 1224712, 1224716, 1224719, 1224735, 1224749, 1224764, 1224765, 1224766, 1224935, 1224946, 1224951, 1225050, 1225098, 1225105, 1225300, 1225389, 1225391, 1225419, 1225426, 1225448, 1225452, 1225467, 1225475, 1225484, 1225487, 1225514, 1225518, 1225535, 1225585, 1225602, 1225611, 1225681, 1225692, 1225698, 1225699, 1225704, 1225714, 1225726, 1225732, 1225737, 1225749, 1225758, 1225759, 1225760, 1225767, 1225770, 1225823, 1225834, 1225840, 1225866, 1225872, 1225894, 1225945, 1226022, 1226131, 1226145, 1226149, 1226155, 1226211, 1226212, 1226226, 1226514, 1226520, 1226537, 1226538, 1226539, 1226550, 1226552, 1226553, 1226554, 1226556, 1226557, 1226558, 1226559, 1226561, 1226562, 1226563, 1226564, 1226566, 1226567, 1226569, 1226572, 1226575, 1226576, 1226577, 1226579, 1226580, 1226581, 1226582, 1226583, 1226585, 1226587, 1226588, 1226593, 1226595, 1226597, 1226601, 1226602, 1226603, 1226607, 1226610, 1226614, 1226616, 1226617, 1226618, 1226619, 1226621, 1226622, 1226624, 1226626, 1226628, 1226629, 1226632, 1226633, 1226634, 1226637, 1226643, 1226644, 1226645, 1226647, 1226650, 1226653, 1226657, 1226658, 1226669, 1226670, 1226672, 1226673, 1226674, 1226675, 1226678, 1226679, 1226683, 1226685, 1226686, 1226690, 1226691, 1226692, 1226693, 1226696, 1226697, 1226698, 1226699, 1226701, 1226702, 1226703, 1226704, 1226705, 1226706, 1226708, 1226709, 1226710, 1226711, 1226712, 1226713, 1226715, 1226716, 1226718, 1226719, 1226720, 1226721, 1226730, 1226732, 1226734, 1226735, 1226737, 1226738, 1226739, 1226740, 1226744, 1226746, 1226747, 1226749, 1226754, 1226762, 1226764, 1226767, 1226768, 1226769, 1226771, 1226774, 1226777, 1226780, 1226781, 1226785, 1226786, 1226789, 1226791, 1226839, 1226840, 1226841, 1226842, 1226848, 1226852, 1226857, 1226861, 1226863, 1226864, 1226867, 1226868, 1226876, 1226878, 1226883, 1226886, 1226890, 1226891, 1226895, 1226908, 1226915, 1226928, 1226948, 1226949, 1226950, 1226953, 1226962, 1226976, 1226992, 1226994, 1226996, 1227066, 1227096, 1227101, 1227103, 1227274 CVE References: CVE-2021-4439, CVE-2021-47089, CVE-2021-47432, CVE-2021-47515, CVE-2021-47534, CVE-2021-47538, CVE-2021-47539, CVE-2021-47555, CVE-2021-47566, CVE-2021-47571, CVE-2021-47572, CVE-2021-47576, CVE-2021-47577, CVE-2021-47578, CVE-2021-47580, CVE-2021-47582, CVE-2021-47583, CVE-2021-47584, CVE-2021-47585, CVE-2021-47586, CVE-2021-47587, CVE-2021-47589, CVE-2021-47592, CVE-2021-47595, CVE-2021-47596, CVE-2021-47597, CVE-2021-47600, CVE-2021-47601, CVE-2021-47602, CVE-2021-47603, CVE-2021-47604, CVE-2021-47605, CVE-2021-47607, CVE-2021-47608, CVE-2021-47609, CVE-2021-47610, CVE-2021-47611, CVE-2021-47612, CVE-2021-47614, CVE-2021-47615, CVE-2021-47616, CVE-2021-47617, CVE-2021-47618, CVE-2021-47619, CVE-2021-47620, CVE-2022-48711, CVE-2022-48712, CVE-2022-48713, CVE-2022-48714, CVE-2022-48715, CVE-2022-48716, CVE-2022-48717, CVE-2022-48718, CVE-2022-48720, CVE-2022-48721, CVE-2022-48722, CVE-2022-48723, CVE-2022-48724, CVE-2022-48725, CVE-2022-48726, CVE-2022-48727, CVE-2022-48728, CVE-2022-48729, CVE-2022-48730, CVE-2022-48732, CVE-2022-48733, CVE-2022-48734, CVE-2022-48735, CVE-2022-48736, CVE-2022-48737, CVE-2022-48738, CVE-2022-48739, CVE-2022-48740, CVE-2022-48743, CVE-2022-48744, CVE-2022-48745, CVE-2022-48746, CVE-2022-48747, CVE-2022-48748, CVE-2022-48749, CVE-2022-48751, CVE-2022-48752, CVE-2022-48753, CVE-2022-48754, CVE-2022-48755, CVE-2022-48756, CVE-2022-48758, CVE-2022-48759, CVE-2022-48760, CVE-2022-48761, CVE-2022-48763, CVE-2022-48765, CVE-2022-48766, CVE-2022-48767, CVE-2022-48768, CVE-2022-48769, CVE-2022-48770, CVE-2022-48771, CVE-2022-48772, CVE-2023-24023, CVE-2023-52622, CVE-2023-52658, CVE-2023-52667, CVE-2023-52670, CVE-2023-52672, CVE-2023-52675, CVE-2023-52735, CVE-2023-52737, CVE-2023-52752, CVE-2023-52766, CVE-2023-52784, CVE-2023-52787, CVE-2023-52800, CVE-2023-52835, CVE-2023-52837, CVE-2023-52843, CVE-2023-52845, CVE-2023-52846, CVE-2023-52869, CVE-2023-52881, CVE-2023-52882, CVE-2023-52884, CVE-2024-26625, CVE-2024-26644, CVE-2024-26720, CVE-2024-26842, CVE-2024-26845, CVE-2024-26923, CVE-2024-26973, CVE-2024-27432, CVE-2024-33619, CVE-2024-35247, CVE-2024-35789, CVE-2024-35790, CVE-2024-35807, CVE-2024-35814, CVE-2024-35835, CVE-2024-35848, CVE-2024-35857, CVE-2024-35861, CVE-2024-35862, CVE-2024-35864, CVE-2024-35869, CVE-2024-35878, CVE-2024-35884, CVE-2024-35886, CVE-2024-35896, CVE-2024-35898, CVE-2024-35900, CVE-2024-35905, CVE-2024-35925, CVE-2024-35950, CVE-2024-35956, CVE-2024-35958, CVE-2024-35960, CVE-2024-35962, CVE-2024-35997, CVE-2024-36005, CVE-2024-36008, CVE-2024-36017, CVE-2024-36020, CVE-2024-36021, CVE-2024-36025, CVE-2024-36477, CVE-2024-36478, CVE-2024-36479, CVE-2024-36890, CVE-2024-36894, CVE-2024-36899, CVE-2024-36900, CVE-2024-36904, CVE-2024-36915, CVE-2024-36916, CVE-2024-36917, CVE-2024-36919, CVE-2024-36934, CVE-2024-36937, CVE-2024-36940, CVE-2024-36945, CVE-2024-36949, CVE-2024-36960, CVE-2024-36964, CVE-2024-36965, CVE-2024-36967, CVE-2024-36969, CVE-2024-36971, CVE-2024-36975, CVE-2024-36978, CVE-2024-37021, CVE-2024-37078, CVE-2024-37354, CVE-2024-38381, CVE-2024-38388, CVE-2024-38390, CVE-2024-38540, CVE-2024-38541, CVE-2024-38544, CVE-2024-38545, CVE-2024-38546, CVE-2024-38547, CVE-2024-38548, CVE-2024-38549, CVE-2024-38550, CVE-2024-38552, CVE-2024-38553, CVE-2024-38555, CVE-2024-38556, CVE-2024-38557, CVE-2024-38559, CVE-2024-38560, CVE-2024-38564, CVE-2024-38565, CVE-2024-38567, CVE-2024-38568, CVE-2024-38571, CVE-2024-38573, CVE-2024-38578, CVE-2024-38579, CVE-2024-38580, CVE-2024-38581, CVE-2024-38582, CVE-2024-38583, CVE-2024-38587, CVE-2024-38590, CVE-2024-38591, CVE-2024-38594, CVE-2024-38597, CVE-2024-38599, CVE-2024-38600, CVE-2024-38601, CVE-2024-38603, CVE-2024-38605, CVE-2024-38608, CVE-2024-38616, CVE-2024-38618, CVE-2024-38619, CVE-2024-38621, CVE-2024-38627, CVE-2024-38630, CVE-2024-38633, CVE-2024-38634, CVE-2024-38635, CVE-2024-38659, CVE-2024-38661, CVE-2024-38780, CVE-2024-39301, CVE-2024-39468, CVE-2024-39469, CVE-2024-39471 Jira References: PED-8491, PED-8570, PED-8690 Maintenance Incident: [SUSE:Maintenance:34676](https://smelt.suse.de/incident/34676/) Sources used: Public Cloud Module 15-SP5 (src): kernel-source-azure-5.14.21-150500.33.60.1, kernel-syms-azure-5.14.21-150500.33.60.1 openSUSE Leap 15.5 (src): kernel-source-azure-5.14.21-150500.33.60.1, kernel-syms-azure-5.14.21-150500.33.60.1 NOTE: This line indicates an update has been released for the listed product(s). At times this might be only a partial fix. If you have questions please reach out to maintenance coordination. SUSE-SU-2024:2394-1: An update that solves 249 vulnerabilities, contains three features and has 44 security fixes can now be installed. Category: security (important) Bug References: 1156395, 1190336, 1191958, 1193883, 1194826, 1195065, 1195254, 1195341, 1195349, 1195775, 1196746, 1197915, 1198014, 1199295, 1202767, 1202780, 1205205, 1207361, 1217912, 1218148, 1218570, 1218820, 1219224, 1219633, 1219847, 1220368, 1220812, 1220958, 1221086, 1221282, 1221958, 1222015, 1222072, 1222080, 1222241, 1222254, 1222364, 1222893, 1223013, 1223018, 1223265, 1223384, 1223641, 1224020, 1224331, 1224488, 1224497, 1224498, 1224504, 1224520, 1224539, 1224540, 1224552, 1224583, 1224588, 1224602, 1224603, 1224605, 1224612, 1224614, 1224619, 1224661, 1224662, 1224670, 1224671, 1224674, 1224677, 1224679, 1224696, 1224703, 1224712, 1224716, 1224719, 1224735, 1224749, 1224764, 1224765, 1224766, 1224935, 1224946, 1224951, 1225050, 1225098, 1225105, 1225300, 1225389, 1225391, 1225419, 1225426, 1225448, 1225452, 1225467, 1225475, 1225484, 1225487, 1225514, 1225518, 1225535, 1225585, 1225602, 1225611, 1225681, 1225692, 1225698, 1225699, 1225704, 1225714, 1225726, 1225732, 1225737, 1225749, 1225758, 1225759, 1225760, 1225767, 1225770, 1225823, 1225834, 1225840, 1225866, 1225872, 1225894, 1226022, 1226131, 1226145, 1226149, 1226155, 1226211, 1226212, 1226226, 1226514, 1226520, 1226537, 1226538, 1226539, 1226550, 1226552, 1226553, 1226554, 1226556, 1226557, 1226558, 1226559, 1226561, 1226562, 1226563, 1226564, 1226566, 1226567, 1226569, 1226572, 1226575, 1226576, 1226577, 1226579, 1226580, 1226581, 1226582, 1226583, 1226585, 1226587, 1226588, 1226593, 1226595, 1226597, 1226601, 1226602, 1226603, 1226607, 1226610, 1226614, 1226616, 1226617, 1226618, 1226619, 1226621, 1226622, 1226624, 1226626, 1226628, 1226629, 1226632, 1226633, 1226634, 1226637, 1226643, 1226644, 1226645, 1226647, 1226650, 1226653, 1226657, 1226658, 1226669, 1226670, 1226672, 1226673, 1226674, 1226675, 1226678, 1226679, 1226683, 1226685, 1226686, 1226690, 1226691, 1226692, 1226693, 1226696, 1226697, 1226698, 1226699, 1226701, 1226702, 1226703, 1226704, 1226705, 1226706, 1226708, 1226709, 1226710, 1226711, 1226712, 1226713, 1226715, 1226716, 1226718, 1226719, 1226720, 1226721, 1226730, 1226732, 1226734, 1226735, 1226737, 1226738, 1226739, 1226740, 1226744, 1226746, 1226747, 1226749, 1226754, 1226762, 1226764, 1226767, 1226768, 1226769, 1226771, 1226774, 1226777, 1226780, 1226781, 1226785, 1226786, 1226789, 1226791, 1226839, 1226840, 1226841, 1226842, 1226848, 1226852, 1226857, 1226861, 1226863, 1226864, 1226867, 1226868, 1226876, 1226878, 1226883, 1226886, 1226890, 1226891, 1226895, 1226908, 1226915, 1226928, 1226948, 1226949, 1226950, 1226953, 1226962, 1226976, 1226992, 1226994, 1226996, 1227066, 1227096, 1227101, 1227103, 1227274 CVE References: CVE-2021-4439, CVE-2021-47089, CVE-2021-47432, CVE-2021-47515, CVE-2021-47534, CVE-2021-47538, CVE-2021-47539, CVE-2021-47555, CVE-2021-47566, CVE-2021-47571, CVE-2021-47572, CVE-2021-47576, CVE-2021-47577, CVE-2021-47578, CVE-2021-47580, CVE-2021-47582, CVE-2021-47583, CVE-2021-47584, CVE-2021-47585, CVE-2021-47586, CVE-2021-47587, CVE-2021-47589, CVE-2021-47592, CVE-2021-47595, CVE-2021-47596, CVE-2021-47597, CVE-2021-47600, CVE-2021-47601, CVE-2021-47602, CVE-2021-47603, CVE-2021-47604, CVE-2021-47605, CVE-2021-47607, CVE-2021-47608, CVE-2021-47609, CVE-2021-47610, CVE-2021-47611, CVE-2021-47612, CVE-2021-47614, CVE-2021-47615, CVE-2021-47616, CVE-2021-47617, CVE-2021-47618, CVE-2021-47619, CVE-2021-47620, CVE-2022-48711, CVE-2022-48712, CVE-2022-48713, CVE-2022-48714, CVE-2022-48715, CVE-2022-48716, CVE-2022-48717, CVE-2022-48718, CVE-2022-48720, CVE-2022-48721, CVE-2022-48722, CVE-2022-48723, CVE-2022-48724, CVE-2022-48725, CVE-2022-48726, CVE-2022-48727, CVE-2022-48728, CVE-2022-48729, CVE-2022-48730, CVE-2022-48732, CVE-2022-48733, CVE-2022-48734, CVE-2022-48735, CVE-2022-48736, CVE-2022-48737, CVE-2022-48738, CVE-2022-48739, CVE-2022-48740, CVE-2022-48743, CVE-2022-48744, CVE-2022-48745, CVE-2022-48746, CVE-2022-48747, CVE-2022-48748, CVE-2022-48749, CVE-2022-48751, CVE-2022-48752, CVE-2022-48753, CVE-2022-48754, CVE-2022-48755, CVE-2022-48756, CVE-2022-48758, CVE-2022-48759, CVE-2022-48760, CVE-2022-48761, CVE-2022-48763, CVE-2022-48765, CVE-2022-48766, CVE-2022-48767, CVE-2022-48768, CVE-2022-48769, CVE-2022-48770, CVE-2022-48771, CVE-2022-48772, CVE-2023-24023, CVE-2023-52622, CVE-2023-52658, CVE-2023-52667, CVE-2023-52670, CVE-2023-52672, CVE-2023-52675, CVE-2023-52735, CVE-2023-52737, CVE-2023-52752, CVE-2023-52766, CVE-2023-52784, CVE-2023-52787, CVE-2023-52800, CVE-2023-52835, CVE-2023-52837, CVE-2023-52843, CVE-2023-52845, CVE-2023-52846, CVE-2023-52869, CVE-2023-52881, CVE-2023-52882, CVE-2023-52884, CVE-2024-26625, CVE-2024-26644, CVE-2024-26720, CVE-2024-26842, CVE-2024-26845, CVE-2024-26923, CVE-2024-26973, CVE-2024-27432, CVE-2024-33619, CVE-2024-35247, CVE-2024-35789, CVE-2024-35790, CVE-2024-35807, CVE-2024-35814, CVE-2024-35835, CVE-2024-35848, CVE-2024-35857, CVE-2024-35861, CVE-2024-35862, CVE-2024-35864, CVE-2024-35869, CVE-2024-35878, CVE-2024-35884, CVE-2024-35886, CVE-2024-35896, CVE-2024-35898, CVE-2024-35900, CVE-2024-35905, CVE-2024-35925, CVE-2024-35950, CVE-2024-35956, CVE-2024-35958, CVE-2024-35960, CVE-2024-35962, CVE-2024-35997, CVE-2024-36005, CVE-2024-36008, CVE-2024-36017, CVE-2024-36020, CVE-2024-36021, CVE-2024-36025, CVE-2024-36477, CVE-2024-36478, CVE-2024-36479, CVE-2024-36890, CVE-2024-36894, CVE-2024-36899, CVE-2024-36900, CVE-2024-36904, CVE-2024-36915, CVE-2024-36916, CVE-2024-36917, CVE-2024-36919, CVE-2024-36934, CVE-2024-36937, CVE-2024-36940, CVE-2024-36945, CVE-2024-36949, CVE-2024-36960, CVE-2024-36964, CVE-2024-36965, CVE-2024-36967, CVE-2024-36969, CVE-2024-36971, CVE-2024-36975, CVE-2024-36978, CVE-2024-37021, CVE-2024-37078, CVE-2024-37354, CVE-2024-38381, CVE-2024-38388, CVE-2024-38390, CVE-2024-38540, CVE-2024-38541, CVE-2024-38544, CVE-2024-38545, CVE-2024-38546, CVE-2024-38547, CVE-2024-38548, CVE-2024-38549, CVE-2024-38550, CVE-2024-38552, CVE-2024-38553, CVE-2024-38555, CVE-2024-38556, CVE-2024-38557, CVE-2024-38559, CVE-2024-38560, CVE-2024-38564, CVE-2024-38565, CVE-2024-38567, CVE-2024-38568, CVE-2024-38571, CVE-2024-38573, CVE-2024-38578, CVE-2024-38579, CVE-2024-38580, CVE-2024-38581, CVE-2024-38582, CVE-2024-38583, CVE-2024-38587, CVE-2024-38590, CVE-2024-38591, CVE-2024-38594, CVE-2024-38597, CVE-2024-38599, CVE-2024-38600, CVE-2024-38601, CVE-2024-38603, CVE-2024-38605, CVE-2024-38608, CVE-2024-38616, CVE-2024-38618, CVE-2024-38619, CVE-2024-38621, CVE-2024-38627, CVE-2024-38630, CVE-2024-38633, CVE-2024-38634, CVE-2024-38635, CVE-2024-38659, CVE-2024-38661, CVE-2024-38780, CVE-2024-39301, CVE-2024-39468, CVE-2024-39469, CVE-2024-39471 Jira References: PED-8491, PED-8570, PED-8690 Maintenance Incident: [SUSE:Maintenance:34699](https://smelt.suse.de/incident/34699/) Sources used: openSUSE Leap 15.5 (src): kernel-livepatch-SLE15-SP5-RT_Update_17-1-150500.11.3.1, kernel-syms-rt-5.14.21-150500.13.61.1, kernel-source-rt-5.14.21-150500.13.61.1 SUSE Linux Enterprise Micro 5.5 (src): kernel-source-rt-5.14.21-150500.13.61.1 SUSE Linux Enterprise Live Patching 15-SP5 (src): kernel-livepatch-SLE15-SP5-RT_Update_17-1-150500.11.3.1 SUSE Real Time Module 15-SP5 (src): kernel-syms-rt-5.14.21-150500.13.61.1, kernel-source-rt-5.14.21-150500.13.61.1 NOTE: This line indicates an update has been released for the listed product(s). At times this might be only a partial fix. If you have questions please reach out to maintenance coordination. |
Description =========== In the Linux kernel, the following vulnerability has been resolved: btrfs: fix deadlock between quota disable and qgroup rescan worker Quota disable ioctl starts a transaction before waiting for the qgroup rescan worker completes. However, this wait can be infinite and results in deadlock because of circular dependency among the quota disable ioctl, the qgroup rescan worker and the other task with transaction such as block group relocation task. The deadlock happens with the steps following: 1) Task A calls ioctl to disable quota. It starts a transaction and waits for qgroup rescan worker completes. 2) Task B such as block group relocation task starts a transaction and joins to the transaction that task A started. Then task B commits to the transaction. In this commit, task B waits for a commit by task A. 3) Task C as the qgroup rescan worker starts its job and starts a transaction. In this transaction start, task C waits for completion of the transaction that task A started and task B committed. This deadlock was found with fstests test case btrfs/115 and a zoned null_blk device. The test case enables and disables quota, and the block group reclaim was triggered during the quota disable by chance. The deadlock was also observed by running quota enable and disable in parallel with 'btrfs balance' command on regular null_blk devices. An example report of the deadlock: [372.469894] INFO: task kworker/u16:6:103 blocked for more than 122 seconds. [372.479944] Not tainted 5.16.0-rc8 #7 [372.485067] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. [372.493898] task:kworker/u16:6 state:D stack: 0 pid: 103 ppid: 2 flags:0x00004000 [372.503285] Workqueue: btrfs-qgroup-rescan btrfs_work_helper [btrfs] [372.510782] Call Trace: [372.514092] <TASK> [372.521684] __schedule+0xb56/0x4850 [372.530104] ? io_schedule_timeout+0x190/0x190 [372.538842] ? lockdep_hardirqs_on+0x7e/0x100 [372.547092] ? _raw_spin_unlock_irqrestore+0x3e/0x60 [372.555591] schedule+0xe0/0x270 [372.561894] btrfs_commit_transaction+0x18bb/0x2610 [btrfs] [372.570506] ? btrfs_apply_pending_changes+0x50/0x50 [btrfs] [372.578875] ? free_unref_page+0x3f2/0x650 [372.585484] ? finish_wait+0x270/0x270 [372.591594] ? release_extent_buffer+0x224/0x420 [btrfs] [372.599264] btrfs_qgroup_rescan_worker+0xc13/0x10c0 [btrfs] [372.607157] ? lock_release+0x3a9/0x6d0 [372.613054] ? btrfs_qgroup_account_extent+0xda0/0xda0 [btrfs] [372.620960] ? do_raw_spin_lock+0x11e/0x250 [372.627137] ? rwlock_bug.part.0+0x90/0x90 [372.633215] ? lock_is_held_type+0xe4/0x140 [372.639404] btrfs_work_helper+0x1ae/0xa90 [btrfs] [372.646268] process_one_work+0x7e9/0x1320 [372.652321] ? lock_release+0x6d0/0x6d0 [372.658081] ? pwq_dec_nr_in_flight+0x230/0x230 [372.664513] ? rwlock_bug.part.0+0x90/0x90 [372.670529] worker_thread+0x59e/0xf90 [372.676172] ? process_one_work+0x1320/0x1320 [372.682440] kthread+0x3b9/0x490 [372.687550] ? _raw_spin_unlock_irq+0x24/0x50 [372.693811] ? set_kthread_struct+0x100/0x100 [372.700052] ret_from_fork+0x22/0x30 [372.705517] </TASK> [372.709747] INFO: task btrfs-transacti:2347 blocked for more than 123 seconds. [372.729827] Not tainted 5.16.0-rc8 #7 [372.745907] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. [372.767106] task:btrfs-transacti state:D stack: 0 pid: 2347 ppid: 2 flags:0x00004000 [372.787776] Call Trace: [372.801652] <TASK> [372.812961] __schedule+0xb56/0x4850 [372.830011] ? io_schedule_timeout+0x190/0x190 [372.852547] ? lockdep_hardirqs_on+0x7e/0x100 [372.871761] ? _raw_spin_unlock_irqrestore+0x3e/0x60 [372.886792] schedule+0xe0/0x270 [372.901685] wait_current_trans+0x22c/0x310 [btrfs] [372.919743] ? btrfs_put_transaction+0x3d0/0x3d0 [btrfs] [372.938923] ? finish_wait+0x270/0x270 [372.959085] ? join_transaction+0xc75/0xe30 [btrfs] [372.977706] start_transaction+0x938/0x10a0 [btrfs] [372.997168] transaction_kthread+0x19d/0x3c0 [btrfs] [373.013021] ? btrfs_cleanup_transaction.isra.0+0xfc0/0xfc0 [btrfs] [373.031678] kthread+0x3b9/0x490 [373.047420] ? _raw_spin_unlock_irq+0x24/0x50 [373.064645] ? set_kthread_struct+0x100/0x100 [373.078571] ret_from_fork+0x22/0x30 [373.091197] </TASK> [373.105611] INFO: task btrfs:3145 blocked for more than 123 seconds. [373.114147] Not tainted 5.16.0-rc8 #7 [373.120401] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. [373.130393] task:btrfs state:D stack: 0 pid: 3145 ppid: 3141 flags:0x00004000 [373.140998] Call Trace: [373.145501] <TASK> [373.149654] __schedule+0xb56/0x4850 [373.155306] ? io_schedule_timeout+0x190/0x190 [373.161965] ? lockdep_hardirqs_on+0x7e/0x100 [373.168469] ? _raw_spin_unlock_irqrestore+0x3e/0x60 [373.175468] schedule+0xe0/0x270 [373.180814] wait_for_commit+0x104/0x150 [btrfs] [373.187643] ? test_and_set_bit+0x20/0x20 [btrfs] [373.194772] ? kmem_cache_free+0x124/0x550 [373.201191] ? btrfs_put_transaction+0x69/0x3d0 [btrfs] [373.208738] ? finish_wait+0x270/0x270 [373.214704] ? __btrfs_end_transaction+0x347/0x7b0 [btrfs] [373.222342] btrfs_commit_transaction+0x44d/0x2610 [btrfs] [373.230233] ? join_transaction+0x255/0xe30 [btrfs] [373.237334] ? btrfs_record_root_in_trans+0x4d/0x170 [btrfs] [373.245251] ? btrfs_apply_pending_changes+0x50/0x50 [btrfs] [373.253296] relocate_block_group+0x105/0xc20 [btrfs] [373.260533] ? mutex_lock_io_nested+0x1270/0x1270 [373.267516] ? btrfs_wait_nocow_writers+0x85/0x180 [btrfs] [373.275155] ? merge_reloc_roots+0x710/0x710 [btrfs] [373.283602] ? btrfs_wait_ordered_extents+0xd30/0xd30 [btrfs] [373.291934] ? kmem_cache_free+0x124/0x550 [373.298180] btrfs_relocate_block_group+0x35c/0x930 [btrfs] [373.306047] btrfs_relocate_chunk+0x85/0x210 [btrfs] [373.313229] btrfs_balance+0x12f4/0x2d20 [btrfs] [373.320227] ? lock_release+0x3a9/0x6d0 [373.326206] ? btrfs_relocate_chunk+0x210/0x210 [btrfs] [373.333591] ? lock_is_held_type+0xe4/0x140 [373.340031] ? rcu_read_lock_sched_held+0x3f/0x70 [373.346910] btrfs_ioctl_balance+0x548/0x700 [btrfs] [373.354207] btrfs_ioctl+0x7f2/0x71b0 [btrfs] [373.360774] ? lockdep_hardirqs_on_prepare+0x410/0x410 [373.367957] ? lockdep_hardirqs_on_prepare+0x410/0x410 [373.375327] ? btrfs_ioctl_get_supported_features+0x20/0x20 [btrfs] [373.383841] ? find_held_lock+0x2c/0x110 [373.389993] ? lock_release+0x3a9/0x6d0 [373.395828] ? mntput_no_expire+0xf7/0xad0 [373.402083] ? lock_is_held_type+0xe4/0x140 [373.408249] ? vfs_fileattr_set+0x9f0/0x9f0 [373.414486] ? selinux_file_ioctl+0x349/0x4e0 [373.420938] ? trace_raw_output_lock+0xb4/0xe0 [373.427442] ? selinux_inode_getsecctx+0x80/0x80 [373.434224] ? lockdep_hardirqs_on+0x7e/0x100 [373.440660] ? force_qs_rnp+0x2a0/0x6b0 [373.446534] ? lock_is_held_type+0x9b/0x140 [373.452763] ? __blkcg_punt_bio_submit+0x1b0/0x1b0 [373.459732] ? security_file_ioctl+0x50/0x90 [373.466089] __x64_sys_ioctl+0x127/0x190 [373.472022] do_syscall_64+0x3b/0x90 [373.477513] entry_SYSCALL_64_after_hwframe+0x44/0xae [373.484823] RIP: 0033:0x7f8f4af7e2bb [373.490493] RSP: 002b:00007ffcbf936178 EFLAGS: 00000246 ORIG_RAX: 0000000000000010 [373.500197] RAX: ffffffffffffffda RBX: 0000000000000003 RCX: 00007f8f4af7e2bb [373.509451] RDX: 00007ffcbf936220 RSI: 00000000c4009420 RDI: 0000000000000003 [373.518659] RBP: 00007ffcbf93774a R08: 0000000000000013 R09: 00007f8f4b02d4e0 [373.527872] R10: 00007f8f4ae87740 R11: 0000000000000246 R12: 0000000000000001 [373.537222] R13: 00007ffcbf936220 R14: 0000000000000000 R15: 0000000000000002 [373.546506] </TASK> [373.550878] INFO: task btrfs:3146 blocked for more than 123 seconds. [373.559383] Not tainted 5.16.0-rc8 #7 [373.565748] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. [373.575748] task:btrfs state:D stack: 0 pid: 3146 ppid: 2168 flags:0x00000000 [373.586314] Call Trace: [373.590846] <TASK> [373.595121] __schedule+0xb56/0x4850 [373.600901] ? __lock_acquire+0x23db/0x5030 [373.607176] ? io_schedule_timeout+0x190/0x190 [373.613954] schedule+0xe0/0x270 [373.619157] schedule_timeout+0x168/0x220 [373.625170] ? usleep_range_state+0x150/0x150 [373.631653] ? mark_held_locks+0x9e/0xe0 [373.637767] ? do_raw_spin_lock+0x11e/0x250 [373.643993] ? lockdep_hardirqs_on_prepare+0x17b/0x410 [373.651267] ? _raw_spin_unlock_irq+0x24/0x50 [373.657677] ? lockdep_hardirqs_on+0x7e/0x100 [373.664103] wait_for_completion+0x163/0x250 [373.670437] ? bit_wait_timeout+0x160/0x160 [373.676585] btrfs_quota_disable+0x176/0x9a0 [btrfs] [373.683979] ? btrfs_quota_enable+0x12f0/0x12f0 [btrfs] [373.691340] ? down_write+0xd0/0x130 [373.696880] ? down_write_killable+0x150/0x150 [373.703352] btrfs_ioctl+0x3945/0x71b0 [btrfs] [373.710061] ? find_held_lock+0x2c/0x110 [373.716192] ? lock_release+0x3a9/0x6d0 [373.722047] ? __handle_mm_fault+0x23cd/0x3050 [373.728486] ? btrfs_ioctl_get_supported_features+0x20/0x20 [btrfs] [373.737032] ? set_pte+0x6a/0x90 [373.742271] ? do_raw_spin_unlock+0x55/0x1f0 [373.748506] ? lock_is_held_type+0xe4/0x140 [373.754792] ? vfs_fileattr_set+0x9f0/0x9f0 [373.761083] ? selinux_file_ioctl+0x349/0x4e0 [373.767521] ? selinux_inode_getsecctx+0x80/0x80 [373.774247] ? __up_read+0x182/0x6e0 [373.780026] ? count_memcg_events.constprop.0+0x46/0x60 [373.787281] ? up_write+0x460/0x460 [373.792932] ? security_file_ioctl+0x50/0x90 [373.799232] __x64_sys_ioctl+0x127/0x190 [373.805237] do_syscall_64+0x3b/0x90 [373.810947] entry_SYSCALL_64_after_hwframe+0x44/0xae [373.818102] RIP: 0033:0x7f1383ea02bb [373.823847] RSP: 002b:00007fffeb4d71f8 EFLAGS: 00000202 ORIG_RAX: 0000000000000010 [373.833641] RAX: ffffffffffffffda RBX: 0000000000000000 RCX: 00007f1383ea02bb [373.842961] RDX: 00007fffeb4d7210 RSI: 00000000c0109428 RDI: 0000000000000003 [373.852179] RBP: 0000000000000003 R08: 0000000000000003 R09: 0000000000000078 [373.861408] R10: 00007f1383daec78 R11: 0000000000000202 R12: 00007fffeb4d874a [373.870647] R13: 0000000000493099 R14: 0000000000000001 R15: 0000000000000000 [373.879838] </TASK> [373.884018] Showing all locks held in the system: [373.894250] 3 locks held by kworker/4:1/58: [373.900356] 1 lock held by khungtaskd/63: [373.906333] #0: ffffffff8945ff60 (rcu_read_lock){....}-{1:2}, at: debug_show_all_locks+0x53/0x260 [373.917307] 3 locks held by kworker/u16:6/103: [373.923938] #0: ffff888127b4f138 ((wq_completion)btrfs-qgroup-rescan){+.+.}-{0:0}, at: process_one_work+0x712/0x1320 [373.936555] #1: ffff88810b817dd8 ((work_completion)(&work->normal_work)){+.+.}-{0:0}, at: process_one_work+0x73f/0x1320 [373.951109] #2: ffff888102dd4650 (sb_internal#2){.+.+}-{0:0}, at: btrfs_qgroup_rescan_worker+0x1f6/0x10c0 [btrfs] [373.964027] 2 locks held by less/1803: [373.969982] #0: ffff88813ed56098 (&tty->ldisc_sem){++++}-{0:0}, at: tty_ldisc_ref_wait+0x24/0x80 [373.981295] #1: ffffc90000b3b2e8 (&ldata->atomic_read_lock){+.+.}-{3:3}, at: n_tty_read+0x9e2/0x1060 [373.992969] 1 lock held by btrfs-transacti/2347: [373.999893] #0: ffff88813d4887a8 (&fs_info->transaction_kthread_mutex){+.+.}-{3:3}, at: transaction_kthread+0xe3/0x3c0 [btrfs] [374.015872] 3 locks held by btrfs/3145: [374.022298] #0: ffff888102dd4460 (sb_writers#18){.+.+}-{0:0}, at: btrfs_ioctl_balance+0xc3/0x700 [btrfs] [374.034456] #1: ffff88813d48a0a0 (&fs_info->reclaim_bgs_lock){+.+.}-{3:3}, at: btrfs_balance+0xfe5/0x2d20 [btrfs] [374.047646] #2: ffff88813d488838 (&fs_info->cleaner_mutex){+.+.}-{3:3}, at: btrfs_relocate_block_group+0x354/0x930 [btrfs] [374.063295] 4 locks held by btrfs/3146: [374.069647] #0: ffff888102dd4460 (sb_writers#18){.+.+}-{0:0}, at: btrfs_ioctl+0x38b1/0x71b0 [btrfs] [374.081601] #1: ffff88813d488bb8 (&fs_info->subvol_sem){+.+.}-{3:3}, at: btrfs_ioctl+0x38fd/0x71b0 [btrfs] [374.094283] #2: ffff888102dd4650 (sb_internal#2){.+.+}-{0:0}, at: btrfs_quota_disable+0xc8/0x9a0 [btrfs] [374.106885] #3: ffff88813d489800 (&fs_info->qgroup_ioctl_lock){+.+.}-{3:3}, at: btrfs_quota_disable+0xd5/0x9a0 [btrfs] [374.126780] ============================================= To avoid the deadlock, wait for the qgroup rescan worker to complete before starting the transaction for the quota disable ioctl. Clear BTRFS_FS_QUOTA_ENABLE flag before the wait and the transaction to request the worker to complete. On transaction start failure, set the BTRFS_FS_QUOTA_ENABLE flag again. These BTRFS_FS_QUOTA_ENABLE flag changes can be done safely since the function btrfs_quota_disable is not called concurrently because of fs_info->subvol_sem. Also check the BTRFS_FS_QUOTA_ENABLE flag in qgroup_rescan_init to avoid another qgroup rescan worker to start after the previous qgroup worker completed. The Linux kernel CVE team has assigned CVE-2022-48734 to this issue. Affected and fixed versions =========================== Fixed in 5.4.178 with commit 26b3901d20bf Fixed in 5.10.99 with commit 32747e01436a Fixed in 5.15.22 with commit 89d4cca583fc Fixed in 5.16.8 with commit 31198e58c09e Fixed in 5.17 with commit e804861bd4e6 Please see https://www.kernel.org for a full list of currently supported kernel versions by the kernel community. Unaffected versions might change over time as fixes are backported to older supported kernel versions. The official CVE entry at https://cve.org/CVERecord/?id=CVE-2022-48734 will be updated if fixes are backported, please check that for the most up to date information about this issue. Affected files ============== The file(s) affected by this issue are: fs/btrfs/qgroup.c Mitigation ========== The Linux kernel CVE team recommends that you update to the latest stable kernel version for this, and many other bugfixes. Individual changes are never tested alone, but rather are part of a larger kernel release. Cherry-picking individual commits is not recommended or supported by the Linux kernel community at all. If however, updating to the latest release is impossible, the individual changes to resolve this issue can be found at these commits: https://git.kernel.org/stable/c/26b3901d20bf9da2c6a00cb1fb48932166f80a45 https://git.kernel.org/stable/c/32747e01436aac8ef93fe85b5b523b4f3b52f040 https://git.kernel.org/stable/c/89d4cca583fc9594ee7d1a0bc986886d6fb587e6 https://git.kernel.org/stable/c/31198e58c09e21d4f65c49d2361f76b87aca4c3f https://git.kernel.org/stable/c/e804861bd4e69cc5fe1053eedcb024982dde8e48 References: http://web.nvd.nist.gov/view/vuln/detail?vulnId=CVE-2022-48734 https://git.kernel.org/pub/scm/linux/security/vulns.git/plain/cve/published/2022/CVE-2022-48734.mbox https://git.kernel.org/stable/c/26b3901d20bf9da2c6a00cb1fb48932166f80a45 https://git.kernel.org/stable/c/32747e01436aac8ef93fe85b5b523b4f3b52f040 https://git.kernel.org/stable/c/89d4cca583fc9594ee7d1a0bc986886d6fb587e6 https://git.kernel.org/stable/c/31198e58c09e21d4f65c49d2361f76b87aca4c3f https://git.kernel.org/stable/c/e804861bd4e69cc5fe1053eedcb024982dde8e48