Neither wq_worker_sleeping() nor io_wq_worker_sleeping() require to be invoked
with preemption disabled:
- The worker flag checks operations only need to be serialized against
the worker thread itself.
- The accounting and worker pool operations are serialized with locks.
which means that disabling preemption has neither a reason nor a
value. Remove it.
Signed-off-by: Thomas Gleixner <[email protected]>
---
kernel/sched/core.c | 2 --
1 file changed, 2 deletions(-)
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -6327,12 +6327,10 @@ static inline void sched_submit_work(str
* requires it.
*/
if (task_flags & (PF_WQ_WORKER | PF_IO_WORKER)) {
- preempt_disable();
if (task_flags & PF_WQ_WORKER)
wq_worker_sleeping(tsk);
else
io_wq_worker_sleeping(tsk);
- preempt_enable_no_resched();
}
if (tsk_is_pi_blocked(tsk))
On 9/28/21 8:30 AM, Thomas Gleixner wrote:
> Neither wq_worker_sleeping() nor io_wq_worker_sleeping() require to be invoked
> with preemption disabled:
>
> - The worker flag checks operations only need to be serialized against
> the worker thread itself.
>
> - The accounting and worker pool operations are serialized with locks.
>
> which means that disabling preemption has neither a reason nor a
> value. Remove it.
Reviewed-by: Jens Axboe <[email protected]>
--
Jens Axboe
On Tue, Sep 28, 2021 at 10:31 PM Thomas Gleixner <[email protected]> wrote:
>
> Neither wq_worker_sleeping() nor io_wq_worker_sleeping() require to be invoked
> with preemption disabled:
>
> - The worker flag checks operations only need to be serialized against
> the worker thread itself.
>
> - The accounting and worker pool operations are serialized with locks.
>
> which means that disabling preemption has neither a reason nor a
> value. Remove it.
>
> Signed-off-by: Thomas Gleixner <[email protected]>
> ---
> kernel/sched/core.c | 2 --
> 1 file changed, 2 deletions(-)
>
> --- a/kernel/sched/core.c
> +++ b/kernel/sched/core.c
> @@ -6327,12 +6327,10 @@ static inline void sched_submit_work(str
> * requires it.
> */
> if (task_flags & (PF_WQ_WORKER | PF_IO_WORKER)) {
Hello
The comment about preemption above this if branch and wq_worker_sleeping()
needs to be removed.
Thanks
Lai
Reviewed-by: Lai Jiangshan <[email protected]>
> - preempt_disable();
> if (task_flags & PF_WQ_WORKER)
> wq_worker_sleeping(tsk);
> else
> io_wq_worker_sleeping(tsk);
> - preempt_enable_no_resched();
> }
>
> if (tsk_is_pi_blocked(tsk))
Neither wq_worker_sleeping() nor io_wq_worker_sleeping() require to be invoked
with preemption disabled:
- The worker flag checks operations only need to be serialized against
the worker thread itself.
- The accounting and worker pool operations are serialized with locks.
which means that disabling preemption has neither a reason nor a
value. Remove it and update the stale comment.
Signed-off-by: Thomas Gleixner <[email protected]>
Reviewed-by: Lai Jiangshan <[email protected]>
Reviewed-by: Jens Axboe <[email protected]>
---
V2: Update comment as pointed out by Lai
---
kernel/sched/core.c | 13 ++++---------
1 file changed, 4 insertions(+), 9 deletions(-)
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -6319,20 +6319,14 @@ static inline void sched_submit_work(str
task_flags = tsk->flags;
/*
- * If a worker went to sleep, notify and ask workqueue whether
- * 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 and because wq_worker_sleeping()
- * requires it.
+ * If a worker goes to sleep, notify and ask workqueue whether it
+ * wants to wake up a task to maintain concurrency.
*/
if (task_flags & (PF_WQ_WORKER | PF_IO_WORKER)) {
- preempt_disable();
if (task_flags & PF_WQ_WORKER)
wq_worker_sleeping(tsk);
else
io_wq_worker_sleeping(tsk);
- preempt_enable_no_resched();
}
if (tsk_is_pi_blocked(tsk))
The following commit has been merged into the sched/core branch of tip:
Commit-ID: 579df3f2325191322b96906a5898462253a45ede
Gitweb: https://git.kernel.org/tip/579df3f2325191322b96906a5898462253a45ede
Author: Thomas Gleixner <[email protected]>
AuthorDate: Wed, 29 Sep 2021 11:37:32 +02:00
Committer: Peter Zijlstra <[email protected]>
CommitterDate: Fri, 01 Oct 2021 13:58:07 +02:00
sched: Remove pointless preemption disable in sched_submit_work()
Neither wq_worker_sleeping() nor io_wq_worker_sleeping() require to be invoked
with preemption disabled:
- The worker flag checks operations only need to be serialized against
the worker thread itself.
- The accounting and worker pool operations are serialized with locks.
which means that disabling preemption has neither a reason nor a
value. Remove it and update the stale comment.
Signed-off-by: Thomas Gleixner <[email protected]>
Signed-off-by: Peter Zijlstra (Intel) <[email protected]>
Reviewed-by: Lai Jiangshan <[email protected]>
Reviewed-by: Jens Axboe <[email protected]>
Link: https://lkml.kernel.org/r/8735pnafj7.ffs@tglx
---
kernel/sched/core.c | 10 ++--------
1 file changed, 2 insertions(+), 8 deletions(-)
diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index 8e49b17..d469c7b 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -6097,20 +6097,14 @@ static inline void sched_submit_work(struct task_struct *tsk)
task_flags = tsk->flags;
/*
- * If a worker went to sleep, notify and ask workqueue whether
- * 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 and because wq_worker_sleeping()
- * requires it.
+ * If a worker goes to sleep, notify and ask workqueue whether it
+ * wants to wake up a task to maintain concurrency.
*/
if (task_flags & (PF_WQ_WORKER | PF_IO_WORKER)) {
- preempt_disable();
if (task_flags & PF_WQ_WORKER)
wq_worker_sleeping(tsk);
else
io_wq_worker_sleeping(tsk);
- preempt_enable_no_resched();
}
if (tsk_is_pi_blocked(tsk))
The following commit has been merged into the sched/core branch of tip:
Commit-ID: b945efcdd07d86cece1cce68503aae91f107eacb
Gitweb: https://git.kernel.org/tip/b945efcdd07d86cece1cce68503aae91f107eacb
Author: Thomas Gleixner <[email protected]>
AuthorDate: Wed, 29 Sep 2021 11:37:32 +02:00
Committer: Peter Zijlstra <[email protected]>
CommitterDate: Tue, 05 Oct 2021 15:52:15 +02:00
sched: Remove pointless preemption disable in sched_submit_work()
Neither wq_worker_sleeping() nor io_wq_worker_sleeping() require to be invoked
with preemption disabled:
- The worker flag checks operations only need to be serialized against
the worker thread itself.
- The accounting and worker pool operations are serialized with locks.
which means that disabling preemption has neither a reason nor a
value. Remove it and update the stale comment.
Signed-off-by: Thomas Gleixner <[email protected]>
Signed-off-by: Peter Zijlstra (Intel) <[email protected]>
Reviewed-by: Lai Jiangshan <[email protected]>
Reviewed-by: Jens Axboe <[email protected]>
Link: https://lkml.kernel.org/r/8735pnafj7.ffs@tglx
---
kernel/sched/core.c | 10 ++--------
1 file changed, 2 insertions(+), 8 deletions(-)
diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index e33b03c..e47d7e5 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -6251,20 +6251,14 @@ static inline void sched_submit_work(struct task_struct *tsk)
task_flags = tsk->flags;
/*
- * If a worker went to sleep, notify and ask workqueue whether
- * 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 and because wq_worker_sleeping()
- * requires it.
+ * If a worker goes to sleep, notify and ask workqueue whether it
+ * wants to wake up a task to maintain concurrency.
*/
if (task_flags & (PF_WQ_WORKER | PF_IO_WORKER)) {
- preempt_disable();
if (task_flags & PF_WQ_WORKER)
wq_worker_sleeping(tsk);
else
io_wq_worker_sleeping(tsk);
- preempt_enable_no_resched();
}
if (tsk_is_pi_blocked(tsk))