GHSA-R8HV-J9G9-64G6
Vulnerability from github – Published: 2026-05-28 12:30 – Updated: 2026-05-28 12:30In the Linux kernel, the following vulnerability has been resolved:
cgroup: Defer css percpu_ref kill on rmdir until cgroup is depopulated
A chain of commits going back to v7.0 reworked rmdir to satisfy the controller invariant that a subsystem's ->css_offline() must not run while tasks are still doing kernel-side work in the cgroup.
[1] d245698d727a ("cgroup: Defer task cgroup unlink until after the task is done switching out") [2] a72f73c4dd9b ("cgroup: Don't expose dead tasks in cgroup") [3] 1b164b876c36 ("cgroup: Wait for dying tasks to leave on rmdir") [4] 4c56a8ac6869 ("cgroup: Fix cgroup_drain_dying() testing the wrong condition") [5] 13e786b64bd3 ("cgroup: Increment nr_dying_subsys_* from rmdir context")
[1] moved task cset unlink from do_exit() to finish_task_switch() so a task's cset link drops only after the task has fully stopped scheduling. That made tasks past exit_signals() linger on cset->tasks until their final context switch, which led to a series of problems as what userspace expected to see after rmdir diverged from what the kernel needs to wait for. [2]-[5] tried to bridge that divergence: [2] filtered the exiting tasks from cgroup.procs; [3] had rmdir(2) sleep in TASK_UNINTERRUPTIBLE for them; [4] fixed the wait's condition; [5] made nr_dying_subsys_* visible synchronously.
The cgroup_drain_dying() wait in [3] turned out to be a dead end. When the rmdir caller is also the reaper of a zombie that pins a pidns teardown (e.g. host PID 1 systemd reaping orphan pids that were re-parented to it during the same teardown), rmdir blocks in TASK_UNINTERRUPTIBLE waiting for those pids to free, the pids can't free because PID 1 is the reaper and it's stuck in rmdir, and the system A-A deadlocks. No internal lock ordering breaks this; the wait itself is the bug.
The css killing side that drove the original reorder, however, can be made cleanly asynchronous: ->css_offline() is already async, run from css_killed_work_fn() driven by percpu_ref_kill_and_confirm(). The fix is to make that chain start only after all tasks have left the cgroup. rmdir's user-visible side then returns as soon as cgroup.procs and friends are empty, while ->css_offline() still runs only after the cgroup is fully drained.
Verified by the original reproducer (pidns teardown + zombie reaper, runs under vng) which hangs vanilla and succeeds here, and by per-commit deterministic repros for [2], [3], [4], [5] with a boot parameter that widens the post-exit_signals() window so each state is reliably reachable. Some stress tests on top of that.
cgroup_apply_control_disable() has the same shape of pre-existing race: when a controller is disabled via subtree_control, kill_css() ran synchronously while tasks past exit_signals() could still be linked to the cgroup's csets, and ->css_offline() could fire before they drained. This patch preserves the existing synchronous behavior at that call site (kill_css_sync() + kill_css_finish() back-to-back) and a follow-up patch will defer kill_css_finish() there using a per-css trigger.
This seems like the right approach and I don't see problems with it. The changes are somewhat invasive but not excessively so, so backporting to -stable should be okay. If something does turn out to be wrong, the fallback is to revert the entire chain ([1]-[5]) and rework in the development branch instead.
v2: Pin cgrp across the deferred destroy work with explicit cgroup_get()/cgroup_put() around queue_work() and the work_fn. v1 wasn't actually broken (ordered cgroup_offline_wq + queue_work order in cgroup_task_dead() saved it) but the explicit ref removes the dependency on those non-obvious invariants. Also note the pre-existing cgroup_apply_control_disable() race in the description; a follow-up will defer kill_css_finish() there.
{
"affected": [],
"aliases": [
"CVE-2026-46223"
],
"database_specific": {
"cwe_ids": [],
"github_reviewed": false,
"github_reviewed_at": null,
"nvd_published_at": "2026-05-28T10:16:37Z",
"severity": null
},
"details": "In the Linux kernel, the following vulnerability has been resolved:\n\ncgroup: Defer css percpu_ref kill on rmdir until cgroup is depopulated\n\nA chain of commits going back to v7.0 reworked rmdir to satisfy the\ncontroller invariant that a subsystem\u0027s -\u003ecss_offline() must not run while\ntasks are still doing kernel-side work in the cgroup.\n\n[1] d245698d727a (\"cgroup: Defer task cgroup unlink until after the task is done switching out\")\n[2] a72f73c4dd9b (\"cgroup: Don\u0027t expose dead tasks in cgroup\")\n[3] 1b164b876c36 (\"cgroup: Wait for dying tasks to leave on rmdir\")\n[4] 4c56a8ac6869 (\"cgroup: Fix cgroup_drain_dying() testing the wrong condition\")\n[5] 13e786b64bd3 (\"cgroup: Increment nr_dying_subsys_* from rmdir context\")\n\n[1] moved task cset unlink from do_exit() to finish_task_switch() so a\ntask\u0027s cset link drops only after the task has fully stopped scheduling.\nThat made tasks past exit_signals() linger on cset-\u003etasks until their final\ncontext switch, which led to a series of problems as what userspace expected\nto see after rmdir diverged from what the kernel needs to wait for. [2]-[5]\ntried to bridge that divergence: [2] filtered the exiting tasks from\ncgroup.procs; [3] had rmdir(2) sleep in TASK_UNINTERRUPTIBLE for them; [4]\nfixed the wait\u0027s condition; [5] made nr_dying_subsys_* visible\nsynchronously.\n\nThe cgroup_drain_dying() wait in [3] turned out to be a dead end. When the\nrmdir caller is also the reaper of a zombie that pins a pidns teardown (e.g.\nhost PID 1 systemd reaping orphan pids that were re-parented to it during\nthe same teardown), rmdir blocks in TASK_UNINTERRUPTIBLE waiting for those\npids to free, the pids can\u0027t free because PID 1 is the reaper and it\u0027s stuck\nin rmdir, and the system A-A deadlocks. No internal lock ordering breaks\nthis; the wait itself is the bug.\n\nThe css killing side that drove the original reorder, however, can be made\ncleanly asynchronous: -\u003ecss_offline() is already async, run from\ncss_killed_work_fn() driven by percpu_ref_kill_and_confirm(). The fix is to\nmake that chain start only after all tasks have left the cgroup. rmdir\u0027s\nuser-visible side then returns as soon as cgroup.procs and friends are\nempty, while -\u003ecss_offline() still runs only after the cgroup is fully\ndrained.\n\nVerified by the original reproducer (pidns teardown + zombie reaper, runs\nunder vng) which hangs vanilla and succeeds here, and by per-commit\ndeterministic repros for [2], [3], [4], [5] with a boot parameter that\nwidens the post-exit_signals() window so each state is reliably reachable.\nSome stress tests on top of that.\n\ncgroup_apply_control_disable() has the same shape of pre-existing race:\nwhen a controller is disabled via subtree_control, kill_css() ran\nsynchronously while tasks past exit_signals() could still be linked to\nthe cgroup\u0027s csets, and -\u003ecss_offline() could fire before they drained.\nThis patch preserves the existing synchronous behavior at that call site\n(kill_css_sync() + kill_css_finish() back-to-back) and a follow-up patch\nwill defer kill_css_finish() there using a per-css trigger.\n\nThis seems like the right approach and I don\u0027t see problems with it. The\nchanges are somewhat invasive but not excessively so, so backporting to\n-stable should be okay. If something does turn out to be wrong, the fallback\nis to revert the entire chain ([1]-[5]) and rework in the development branch\ninstead.\n\nv2: Pin cgrp across the deferred destroy work with explicit\n cgroup_get()/cgroup_put() around queue_work() and the work_fn. v1\n wasn\u0027t actually broken (ordered cgroup_offline_wq + queue_work order\n in cgroup_task_dead() saved it) but the explicit ref removes the\n dependency on those non-obvious invariants. Also note the\n pre-existing cgroup_apply_control_disable() race in the description;\n a follow-up will defer kill_css_finish() there.",
"id": "GHSA-r8hv-j9g9-64g6",
"modified": "2026-05-28T12:30:33Z",
"published": "2026-05-28T12:30:33Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2026-46223"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/33fa2e6b1507a0a377a151a8826438bedad1d0b0"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/93618edf753838a727dbff63c7c291dee22d656b"
}
],
"schema_version": "1.4.0",
"severity": []
}
Sightings
| Author | Source | Type | Date | Other |
|---|
Nomenclature
- Seen: The vulnerability was mentioned, discussed, or observed by the user.
- Confirmed: The vulnerability has been validated from an analyst's perspective.
- Published Proof of Concept: A public proof of concept is available for this vulnerability.
- Exploited: The vulnerability was observed as exploited by the user who reported the sighting.
- Patched: The vulnerability was observed as successfully patched by the user who reported the sighting.
- Not exploited: The vulnerability was not observed as exploited by the user who reported the sighting.
- Not confirmed: The user expressed doubt about the validity of the vulnerability.
- Not patched: The vulnerability was not observed as successfully patched by the user who reported the sighting.