Skip to content

Commit 171aac2

Browse files
committed
posix-cpu-timers: fix race between handle_posix_cpu_timers() and posix_cpu_timer_del()
jira LE-4192 cve CVE-2025-38352 Rebuild_History Non-Buildable kernel-6.12.0-55.32.1.el10_0 commit-author Oleg Nesterov <oleg@redhat.com> commit f90fff1 If an exiting non-autoreaping task has already passed exit_notify() and calls handle_posix_cpu_timers() from IRQ, it can be reaped by its parent or debugger right after unlock_task_sighand(). If a concurrent posix_cpu_timer_del() runs at that moment, it won't be able to detect timer->it.cpu.firing != 0: cpu_timer_task_rcu() and/or lock_task_sighand() will fail. Add the tsk->exit_state check into run_posix_cpu_timers() to fix this. This fix is not needed if CONFIG_POSIX_CPU_TIMERS_TASK_WORK=y, because exit_task_work() is called before exit_notify(). But the check still makes sense, task_work_add(&tsk->posix_cputimers_work.work) will fail anyway in this case. Cc: stable@vger.kernel.org Reported-by: Benoît Sevens <bsevens@google.com> Fixes: 0bdd2ed ("sched: run_posix_cpu_timers: Don't check ->exit_state, use lock_task_sighand()") Signed-off-by: Oleg Nesterov <oleg@redhat.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> (cherry picked from commit f90fff1) Signed-off-by: Jonathan Maple <jmaple@ciq.com>
1 parent 5ec8b64 commit 171aac2

File tree

1 file changed

+9
-0
lines changed

1 file changed

+9
-0
lines changed

kernel/time/posix-cpu-timers.c

Lines changed: 9 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -1400,6 +1400,15 @@ void run_posix_cpu_timers(void)
14001400

14011401
lockdep_assert_irqs_disabled();
14021402

1403+
/*
1404+
* Ensure that release_task(tsk) can't happen while
1405+
* handle_posix_cpu_timers() is running. Otherwise, a concurrent
1406+
* posix_cpu_timer_del() may fail to lock_task_sighand(tsk) and
1407+
* miss timer->it.cpu.firing != 0.
1408+
*/
1409+
if (tsk->exit_state)
1410+
return;
1411+
14031412
/*
14041413
* If the actual expiry is deferred to task work context and the
14051414
* work is already scheduled there is no point to do anything here.

0 commit comments

Comments
 (0)