diff options
| author | Sebastian Andrzej Siewior <bigeasy@linutronix.de> | 2020-03-28 00:29:59 +0100 |
|---|---|---|
| committer | Ingo Molnar <mingo@kernel.org> | 2020-04-08 11:35:20 +0200 |
| commit | 62849a9612924a655c67cf6962920544aa5c20db (patch) | |
| tree | f8acbf5978aa3bc2df5eb7f1122426bfe843204c /kernel | |
| parent | 111688ca1c4a43a7e482f5401f82c46326b8ed49 (diff) | |
| download | linux-62849a9612924a655c67cf6962920544aa5c20db.tar.bz2 | |
workqueue: Remove the warning in wq_worker_sleeping()
The kernel test robot triggered a warning with the following race:
task-ctx A interrupt-ctx B
worker
-> process_one_work()
-> work_item()
-> schedule();
-> sched_submit_work()
-> wq_worker_sleeping()
-> ->sleeping = 1
atomic_dec_and_test(nr_running)
__schedule(); *interrupt*
async_page_fault()
-> local_irq_enable();
-> schedule();
-> sched_submit_work()
-> wq_worker_sleeping()
-> if (WARN_ON(->sleeping)) return
-> __schedule()
-> sched_update_worker()
-> wq_worker_running()
-> atomic_inc(nr_running);
-> ->sleeping = 0;
-> sched_update_worker()
-> wq_worker_running()
if (!->sleeping) return
In this context the warning is pointless everything is fine.
An interrupt before wq_worker_sleeping() will perform the ->sleeping
assignment (0 -> 1 > 0) twice.
An interrupt after wq_worker_sleeping() will trigger the warning and
nr_running will be decremented (by A) and incremented once (only by B, A
will skip it). This is the case until the ->sleeping is zeroed again in
wq_worker_running().
Remove the WARN statement because this condition may happen. Document
that preemption around wq_worker_sleeping() needs to be disabled to
protect ->sleeping and not just as an optimisation.
Fixes: 6d25be5782e48 ("sched/core, workqueues: Distangle worker accounting from rq lock")
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: Tejun Heo <tj@kernel.org>
Link: https://lkml.kernel.org/r/20200327074308.GY11705@shao2-debian
Diffstat (limited to 'kernel')
| -rw-r--r-- | kernel/sched/core.c | 3 | ||||
| -rw-r--r-- | kernel/workqueue.c | 6 |
2 files changed, 6 insertions, 3 deletions
diff --git a/kernel/sched/core.c b/kernel/sched/core.c index f6b329bca0c6..c3d12e3762d4 100644 --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -4120,7 +4120,8 @@ static inline void sched_submit_work(struct task_struct *tsk) * it wants to wake up a task to maintain concurrency. * As this function is called inside the schedule() context, * we disable preemption to avoid it calling schedule() again - * in the possible wakeup of a kworker. + * in the possible wakeup of a kworker and because wq_worker_sleeping() + * requires it. */ if (tsk->flags & (PF_WQ_WORKER | PF_IO_WORKER)) { preempt_disable(); diff --git a/kernel/workqueue.c b/kernel/workqueue.c index 3816a18c251e..891ccad5f271 100644 --- a/kernel/workqueue.c +++ b/kernel/workqueue.c @@ -858,7 +858,8 @@ void wq_worker_running(struct task_struct *task) * @task: task going to sleep * * This function is called from schedule() when a busy worker is - * going to sleep. + * going to sleep. Preemption needs to be disabled to protect ->sleeping + * assignment. */ void wq_worker_sleeping(struct task_struct *task) { @@ -875,7 +876,8 @@ void wq_worker_sleeping(struct task_struct *task) pool = worker->pool; - if (WARN_ON_ONCE(worker->sleeping)) + /* Return if preempted before wq_worker_running() was reached */ + if (worker->sleeping) return; worker->sleeping = 1; |