kthread: fix task state in kthread worker if being frozen
PATCH v2 1/1 kthread: fix task state in kthread worker if being frozen
When analyzing a kernel waring message, Peter pointed out that there is a race condition when the kworker is being frozen and falls into try_to_freeze() with TASK_INTERRUPTIBLE, which could trigger a might_sleep() warning in try_to_freeze(). Although the root cause is not related to freeze()[1], it is still worthy to fix this issue ahead.
One possible race scenario:
CPU 0 CPU 1
----- -----
// kthread_worker_fn
set_current_state(TASK_INTERRUPTIBLE);
suspend_freeze_processes()
freeze_processes
static_branch_inc(&freezer_active);
freeze_kernel_threads
pm_nosig_freezing = true;
if (work) { //false
__set_current_state(TASK_RUNNING);
} else if (!freezing(current)) //false, been frozen
freezing():
if (static_branch_unlikely(&freezer_active))
if (pm_nosig_freezing)
return true;
schedule()
}
// state is still TASK_INTERRUPTIBLE
try_to_freeze()
might_sleep() <--- warningFix this by explicitly set the TASK_RUNNING before entering try_to_freeze().
Fixes b56c0d8937e6 ("kthread: implement kthread_worker")
Suggested-by Peter Zijlstra <[email protected]>
Suggested-by Andrew Morton <[email protected]>
Signed-off-by Chen Yu <[email protected]>
---v1->v2: Describe the race condition in commit log and
refined the code. (Andrew Morton)
--- kernel/kthread.c | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-)
kernel/kthread.c +9 -1
@@ -845,8 +845,16 @@ int kthread_worker_fn(void *worker_ptr)* event only cares about the address. */ trace_sched_kthread_work_execute_end (work, func); - }else if (!freezing (current)) + }else if (!freezing (current)) {schedule (); + }else { +/* +* Handle the case where the current remains +* TASK_INTERRUPTIBLE. try_to_freeze() expects +* the current to be TASK_RUNNING. +*/ +__set_current_state (TASK_RUNNING ); + }try_to_freeze ();cond_resched ();
Signature · 1 line
2.25.1