CVE-2026-99431: ocfs2: cancel quota sync work before freeing quota info
In the Linux kernel, the following vulnerability has been resolved:
ocfs2: cancel quota sync work before freeing quota info
OCFS2 quota info owns dqisyncwork and must outlive both its timer and qsyncworkfn(). ocfs2localfreeinfo() frees that owner without cancelling the work, relying on the quota suspend/disable wrappers.
Loading quotas on a read-only to read-write remount can successfully queue the work in ocfs2globalreadinfo() and then fail while attaching a dquot to an existing inode in adddquotref(). The quota core unwinds that failure directly through dquotdisable(), bypassing the OCFS2 wrapper:
Remount task (sumount write) Timer/workqueue ocfs2globalreadinfo() schedules dqisyncwork adddquotref() returns an error dquotdisable() ocfs2localfreeinfo() frees oinfo delayed-work timer expires qsyncworkfn() reads oinfo
The timer core can then write to the freed embedded timer, or qsyncworkfn() can read oinfo->dqigqinode before its sumount trylock. The outer ocfs2disablequotas() cleanup skips the failed type because dquotdisable() has already cleared its loaded state.
Cancel dqisyncwork synchronously at the start of ocfs2localfreeinfo(), before dropping the global inode or freeing any owner state. The work is initialized whenever readfileinfo() succeeds. dquotdisable() holds sumount for write, so the callback skips its quota scan via downreadtrylock() and can finish without waiting for teardown. The synchronous cancel also handles its self-rearm.
KASAN report as below:
BUG: KASAN: slab-use-after-free in runtimers+0x9e9/0xa10 Write of size 8 at addr ffff888111b0aad0 by task swapper/3/0
Call Trace: <IRQ> dumpstacklvl+0x93/0xd0 printreport+0xce/0x630 ? runtimers+0x9e9/0xa10 ? srsoaliasreturnthunk+0x5/0xfbef5 ? virtaddrvalid+0x20e/0x420 ? runtimers+0x9e9/0xa10 kasanreport+0xe0/0x110 ? runtimers+0x9e9/0xa10 runtimers+0x9e9/0xa10 ? pfxruntimers+0x10/0x10 ? lockacquire+0x5a3/0x2140 ? pfxdorawspinlock+0x10/0x10 ? srsoaliasreturnthunk+0x5/0xfbef5 timerexpireremote+0xfa/0x160 ? pfxtimerexpireremote+0x10/0x10 ? srsoaliasreturnthunk+0x5/0xfbef5 ? srsoaliasreturnthunk+0x5/0xfbef5 ? tracehardirqson+0x18/0x160 tmigrhandleremoteup+0x4ba/0x9d0 ? pfxtmigrhandleremoteup+0x10/0x10 ? tmigrhandleremote+0x21a/0x340 ? srsoaliasreturnthunk+0x5/0xfbef5 ? lockrelease+0xc8/0x290 ? pfxtmigrhandleremoteup+0x10/0x10 walkgroupsfrom+0x5a/0x160 tmigrhandleremote+0x24f/0x340 ? pfxtmigrhandleremote+0x10/0x10 ? lockacquire+0x190/0x300 ? srsoaliasreturnthunk+0x5/0xfbef5 ? lockrelease+0xc8/0x290 runtimersoftirq+0x222/0x270 ? pfxruntimersoftirq+0x10/0x10 ? srsoaliasreturnthunk+0x5/0xfbef5 ? srsoaliasreturnthunk+0x5/0xfbef5 ? lockdephardirqsonprepare+0xb8/0x1a0 ? srsoaliasreturnthunk+0x5/0xfbef5 handlesoftirqs+0x1fa/0x8d0 ? pfxhandlesoftirqs+0x10/0x10 ? srsoaliasreturnthunk+0x5/0xfbef5 ? rawspinunlock+0x23/0x40 ? srsoaliasreturnthunk+0x5/0xfbef5 ? hrtimerrearmdeferred+0x1fb/0x5d0 irqexitrcu+0x113/0x1e0 irqexitrcu+0xe/0x20 sysvecapictimerinterrupt+0x6c/0x80 </IRQ> <TASK> asmsysvecapictimerinterrupt+0x1a/0x20 ? ctkernelexit.constprop.0+0xc5/0xf0 ? doidle+0x3ae/0x650 defaultidle+0x9/0x10 defaultidlecall+0x6c/0xb0 doidle+0x3ae/0x650 ? cpuhponlineidle+0x9c/0xb0 ? pfxdoidle+0x10/0x10 ? entrytextend+0x1020f6/0x1020f9 ? lockdephardirqsonprepare+0xeb/0x1a0 ? srsoaliasreturnthunk+0x5/0xfbef5 ? tracehardirqson+0x18/0x160 ? srsoa ---truncated---
Affected Software
Remediation
Recommended actions to resolve this vulnerability, in priority order.
- Upgrade
Upgrade
linux/Kernelto a version that resolves this vulnerability.Fixed in 4.15 - Upgrade
Upgrade
linux/Kernelto a version that resolves this vulnerability.Fixed in 6.12.117 - Upgrade
Upgrade
linux/Kernelto a version that resolves this vulnerability.Fixed in 6.18.61 - Upgrade
Upgrade
linux/Kernelto a version that resolves this vulnerability.Fixed in 7.3.2 - Compensating control
In OCFS2, synchronously cancel dqi_sync_work at the start of ocfs2_local_free_info(), before dropping the global inode or freeing quota info.