* [PATCH v3] workqueue: Fix WARN_ON_ONCE() triggers in worker_enter_idle()
@ 2023-05-24 3:53 Zqiang
2023-05-24 13:53 ` Naresh Kamboju
2023-05-24 22:02 ` Tejun Heo
0 siblings, 2 replies; 4+ messages in thread
From: Zqiang @ 2023-05-24 3:53 UTC (permalink / raw
To: tj, jiangshanlai, naresh.kamboju, qiang.zhang1211; +Cc: linux-kernel
Currently, the nr_running can be modified from timer tick, that means
the timer tick can run in not-irq-protected critical section to modify
nr_runnig, consider the following scenario:
CPU0
kworker/0:2 (events)
worker_clr_flags(worker, WORKER_PREP | WORKER_REBOUND);
->pool->nr_running++; (1)
process_one_work()
->worker->current_func(work);
->schedule()
->wq_worker_sleeping()
->worker->sleeping = 1;
->pool->nr_running--; (0)
....
->wq_worker_running()
....
CPU0 by interrupt:
wq_worker_tick()
->worker_set_flags(worker, WORKER_CPU_INTENSIVE);
->pool->nr_running--; (-1)
->worker->flags |= WORKER_CPU_INTENSIVE;
....
->if (!(worker->flags & WORKER_NOT_RUNNING))
->pool->nr_running++; (will not execute)
->worker->sleeping = 0;
....
->worker_clr_flags(worker, WORKER_CPU_INTENSIVE);
->pool->nr_running++; (0)
....
worker_set_flags(worker, WORKER_PREP);
->pool->nr_running--; (-1)
....
worker_enter_idle()
->WARN_ON_ONCE(pool->nr_workers == pool->nr_idle && pool->nr_running);
if the nr_workers is equal to nr_idle, due to the nr_running is not zero,
will trigger WARN_ON_ONCE().
[ 2.460602] WARNING: CPU: 0 PID: 63 at kernel/workqueue.c:1999 worker_enter_idle+0xb2/0xc0
[ 2.462163] Modules linked in:
[ 2.463401] CPU: 0 PID: 63 Comm: kworker/0:2 Not tainted 6.4.0-rc2-next-20230519 #1
[ 2.463771] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.14.0-2 04/01/2014
[ 2.465127] Workqueue: 0x0 (events)
[ 2.465678] RIP: 0010:worker_enter_idle+0xb2/0xc0
...
[ 2.472614] Call Trace:
[ 2.473152] <TASK>
[ 2.474182] worker_thread+0x71/0x430
[ 2.474992] ? _raw_spin_unlock_irqrestore+0x28/0x50
[ 2.475263] kthread+0x103/0x120
[ 2.475493] ? __pfx_worker_thread+0x10/0x10
[ 2.476355] ? __pfx_kthread+0x10/0x10
[ 2.476635] ret_from_fork+0x2c/0x50
[ 2.477051] </TASK>
This commit therefore add the check of worker->sleeping in wq_worker_tick(),
if the worker->sleeping is not zero, directly return.
Reported-by: Naresh Kamboju <naresh.kamboju@linaro.org>
Closes: https://qa-reports.linaro.org/lkft/linux-next-master/build/next-20230519/testrun/17078554/suite/boot/test/clang-nightly-lkftconfig/log
Signed-off-by: Zqiang <qiang.zhang1211@gmail.com>
---
kernel/workqueue.c | 15 ++++++++++-----
1 file changed, 10 insertions(+), 5 deletions(-)
diff --git a/kernel/workqueue.c b/kernel/workqueue.c
index 9c5c1cfa478f..a028b851333e 100644
--- a/kernel/workqueue.c
+++ b/kernel/workqueue.c
@@ -1051,7 +1051,7 @@ void wq_worker_running(struct task_struct *task)
{
struct worker *worker = kthread_data(task);
- if (!worker->sleeping)
+ if (!READ_ONCE(worker->sleeping))
return;
/*
@@ -1071,7 +1071,7 @@ void wq_worker_running(struct task_struct *task)
*/
worker->current_at = worker->task->se.sum_exec_runtime;
- worker->sleeping = 0;
+ WRITE_ONCE(worker->sleeping, 0);
}
/**
@@ -1097,10 +1097,10 @@ void wq_worker_sleeping(struct task_struct *task)
pool = worker->pool;
/* Return if preempted before wq_worker_running() was reached */
- if (worker->sleeping)
+ if (READ_ONCE(worker->sleeping))
return;
- worker->sleeping = 1;
+ WRITE_ONCE(worker->sleeping, 1);
raw_spin_lock_irq(&pool->lock);
/*
@@ -1143,8 +1143,13 @@ void wq_worker_tick(struct task_struct *task)
* If the current worker is concurrency managed and hogged the CPU for
* longer than wq_cpu_intensive_thresh_us, it's automatically marked
* CPU_INTENSIVE to avoid stalling other concurrency-managed work items.
+ *
+ * The worker->sleeping is true means that the worker doing voluntary
+ * switch and will not hogged the CPU, or the worker is running again
+ * but the worker->sleeping has not been reset, in the process of executing
+ * wq_worker_running().
*/
- if ((worker->flags & WORKER_NOT_RUNNING) ||
+ if ((worker->flags & WORKER_NOT_RUNNING) || READ_ONCE(worker->sleeping) ||
worker->task->se.sum_exec_runtime - worker->current_at <
wq_cpu_intensive_thresh_us * NSEC_PER_USEC)
return;
--
2.17.1
^ permalink raw reply related [flat|nested] 4+ messages in thread
* Re: [PATCH v3] workqueue: Fix WARN_ON_ONCE() triggers in worker_enter_idle()
2023-05-24 3:53 [PATCH v3] workqueue: Fix WARN_ON_ONCE() triggers in worker_enter_idle() Zqiang
@ 2023-05-24 13:53 ` Naresh Kamboju
2023-05-24 22:03 ` Tejun Heo
2023-05-24 22:02 ` Tejun Heo
1 sibling, 1 reply; 4+ messages in thread
From: Naresh Kamboju @ 2023-05-24 13:53 UTC (permalink / raw
To: Zqiang; +Cc: tj, jiangshanlai, linux-kernel, Anders Roxell, lkft-triage
+ Anders, LKFT
On Wed, 24 May 2023 at 09:23, Zqiang <qiang.zhang1211@gmail.com> wrote:
>
> Currently, the nr_running can be modified from timer tick, that means
> the timer tick can run in not-irq-protected critical section to modify
> nr_runnig, consider the following scenario:
>
> CPU0
> kworker/0:2 (events)
> worker_clr_flags(worker, WORKER_PREP | WORKER_REBOUND);
> ->pool->nr_running++; (1)
>
> process_one_work()
> ->worker->current_func(work);
> ->schedule()
> ->wq_worker_sleeping()
> ->worker->sleeping = 1;
> ->pool->nr_running--; (0)
> ....
> ->wq_worker_running()
> ....
> CPU0 by interrupt:
> wq_worker_tick()
> ->worker_set_flags(worker, WORKER_CPU_INTENSIVE);
> ->pool->nr_running--; (-1)
> ->worker->flags |= WORKER_CPU_INTENSIVE;
> ....
> ->if (!(worker->flags & WORKER_NOT_RUNNING))
> ->pool->nr_running++; (will not execute)
> ->worker->sleeping = 0;
> ....
> ->worker_clr_flags(worker, WORKER_CPU_INTENSIVE);
> ->pool->nr_running++; (0)
> ....
> worker_set_flags(worker, WORKER_PREP);
> ->pool->nr_running--; (-1)
> ....
> worker_enter_idle()
> ->WARN_ON_ONCE(pool->nr_workers == pool->nr_idle && pool->nr_running);
>
> if the nr_workers is equal to nr_idle, due to the nr_running is not zero,
> will trigger WARN_ON_ONCE().
>
> [ 2.460602] WARNING: CPU: 0 PID: 63 at kernel/workqueue.c:1999 worker_enter_idle+0xb2/0xc0
> [ 2.462163] Modules linked in:
> [ 2.463401] CPU: 0 PID: 63 Comm: kworker/0:2 Not tainted 6.4.0-rc2-next-20230519 #1
> [ 2.463771] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.14.0-2 04/01/2014
> [ 2.465127] Workqueue: 0x0 (events)
> [ 2.465678] RIP: 0010:worker_enter_idle+0xb2/0xc0
> ...
> [ 2.472614] Call Trace:
> [ 2.473152] <TASK>
> [ 2.474182] worker_thread+0x71/0x430
> [ 2.474992] ? _raw_spin_unlock_irqrestore+0x28/0x50
> [ 2.475263] kthread+0x103/0x120
> [ 2.475493] ? __pfx_worker_thread+0x10/0x10
> [ 2.476355] ? __pfx_kthread+0x10/0x10
> [ 2.476635] ret_from_fork+0x2c/0x50
> [ 2.477051] </TASK>
>
> This commit therefore add the check of worker->sleeping in wq_worker_tick(),
> if the worker->sleeping is not zero, directly return.
>
> Reported-by: Naresh Kamboju <naresh.kamboju@linaro.org>
> Closes: https://qa-reports.linaro.org/lkft/linux-next-master/build/next-20230519/testrun/17078554/suite/boot/test/clang-nightly-lkftconfig/log
> Signed-off-by: Zqiang <qiang.zhang1211@gmail.com>
Reported-by: Linux Kernel Functional Testing <lkft@linaro.org>
Tested-by: Anders Roxell <anders.roxell@linaro.org>
Since the probability of occurrence of this problem is only 3%,
Anders took this up and applied this on top of Linux next and
tested for 500 boot tests and all looked good.
Thanks, Anders.
- Naresh
> ---
> kernel/workqueue.c | 15 ++++++++++-----
> 1 file changed, 10 insertions(+), 5 deletions(-)
>
> diff --git a/kernel/workqueue.c b/kernel/workqueue.c
> index 9c5c1cfa478f..a028b851333e 100644
> --- a/kernel/workqueue.c
> +++ b/kernel/workqueue.c
> @@ -1051,7 +1051,7 @@ void wq_worker_running(struct task_struct *task)
> {
> struct worker *worker = kthread_data(task);
>
> - if (!worker->sleeping)
> + if (!READ_ONCE(worker->sleeping))
> return;
>
> /*
> @@ -1071,7 +1071,7 @@ void wq_worker_running(struct task_struct *task)
> */
> worker->current_at = worker->task->se.sum_exec_runtime;
>
> - worker->sleeping = 0;
> + WRITE_ONCE(worker->sleeping, 0);
> }
>
> /**
> @@ -1097,10 +1097,10 @@ void wq_worker_sleeping(struct task_struct *task)
> pool = worker->pool;
>
> /* Return if preempted before wq_worker_running() was reached */
> - if (worker->sleeping)
> + if (READ_ONCE(worker->sleeping))
> return;
>
> - worker->sleeping = 1;
> + WRITE_ONCE(worker->sleeping, 1);
> raw_spin_lock_irq(&pool->lock);
>
> /*
> @@ -1143,8 +1143,13 @@ void wq_worker_tick(struct task_struct *task)
> * If the current worker is concurrency managed and hogged the CPU for
> * longer than wq_cpu_intensive_thresh_us, it's automatically marked
> * CPU_INTENSIVE to avoid stalling other concurrency-managed work items.
> + *
> + * The worker->sleeping is true means that the worker doing voluntary
> + * switch and will not hogged the CPU, or the worker is running again
> + * but the worker->sleeping has not been reset, in the process of executing
> + * wq_worker_running().
> */
> - if ((worker->flags & WORKER_NOT_RUNNING) ||
> + if ((worker->flags & WORKER_NOT_RUNNING) || READ_ONCE(worker->sleeping) ||
> worker->task->se.sum_exec_runtime - worker->current_at <
> wq_cpu_intensive_thresh_us * NSEC_PER_USEC)
> return;
> --
> 2.17.1
>
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [PATCH v3] workqueue: Fix WARN_ON_ONCE() triggers in worker_enter_idle()
2023-05-24 3:53 [PATCH v3] workqueue: Fix WARN_ON_ONCE() triggers in worker_enter_idle() Zqiang
2023-05-24 13:53 ` Naresh Kamboju
@ 2023-05-24 22:02 ` Tejun Heo
1 sibling, 0 replies; 4+ messages in thread
From: Tejun Heo @ 2023-05-24 22:02 UTC (permalink / raw
To: Zqiang; +Cc: jiangshanlai, naresh.kamboju, linux-kernel
Hello,
I updated the comment and description and applied the patch to wq/for-6.5.
Thanks.
From c8f6219be2e58d7f676935ae90b64abef5d0966a Mon Sep 17 00:00:00 2001
From: Zqiang <qiang.zhang1211@gmail.com>
Date: Wed, 24 May 2023 11:53:39 +0800
Subject: [PATCH] workqueue: Fix WARN_ON_ONCE() triggers in worker_enter_idle()
Currently, pool->nr_running can be modified from timer tick, that means the
timer tick can run nested inside a not-irq-protected section that's in the
process of modifying nr_running. Consider the following scenario:
CPU0
kworker/0:2 (events)
worker_clr_flags(worker, WORKER_PREP | WORKER_REBOUND);
->pool->nr_running++; (1)
process_one_work()
->worker->current_func(work);
->schedule()
->wq_worker_sleeping()
->worker->sleeping = 1;
->pool->nr_running--; (0)
....
->wq_worker_running()
....
CPU0 by interrupt:
wq_worker_tick()
->worker_set_flags(worker, WORKER_CPU_INTENSIVE);
->pool->nr_running--; (-1)
->worker->flags |= WORKER_CPU_INTENSIVE;
....
->if (!(worker->flags & WORKER_NOT_RUNNING))
->pool->nr_running++; (will not execute)
->worker->sleeping = 0;
....
->worker_clr_flags(worker, WORKER_CPU_INTENSIVE);
->pool->nr_running++; (0)
....
worker_set_flags(worker, WORKER_PREP);
->pool->nr_running--; (-1)
....
worker_enter_idle()
->WARN_ON_ONCE(pool->nr_workers == pool->nr_idle && pool->nr_running);
if the nr_workers is equal to nr_idle, due to the nr_running is not zero,
will trigger WARN_ON_ONCE().
[ 2.460602] WARNING: CPU: 0 PID: 63 at kernel/workqueue.c:1999 worker_enter_idle+0xb2/0xc0
[ 2.462163] Modules linked in:
[ 2.463401] CPU: 0 PID: 63 Comm: kworker/0:2 Not tainted 6.4.0-rc2-next-20230519 #1
[ 2.463771] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.14.0-2 04/01/2014
[ 2.465127] Workqueue: 0x0 (events)
[ 2.465678] RIP: 0010:worker_enter_idle+0xb2/0xc0
...
[ 2.472614] Call Trace:
[ 2.473152] <TASK>
[ 2.474182] worker_thread+0x71/0x430
[ 2.474992] ? _raw_spin_unlock_irqrestore+0x28/0x50
[ 2.475263] kthread+0x103/0x120
[ 2.475493] ? __pfx_worker_thread+0x10/0x10
[ 2.476355] ? __pfx_kthread+0x10/0x10
[ 2.476635] ret_from_fork+0x2c/0x50
[ 2.477051] </TASK>
This commit therefore add the check of worker->sleeping in wq_worker_tick(),
if the worker->sleeping is not zero, directly return.
tj: Updated comment and description.
Reported-by: Naresh Kamboju <naresh.kamboju@linaro.org>
Reported-by: Linux Kernel Functional Testing <lkft@linaro.org>
Tested-by: Anders Roxell <anders.roxell@linaro.org>
Closes: https://qa-reports.linaro.org/lkft/linux-next-master/build/next-20230519/testrun/17078554/suite/boot/test/clang-nightly-lkftconfig/log
Signed-off-by: Zqiang <qiang.zhang1211@gmail.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
---
kernel/workqueue.c | 17 ++++++++++++-----
1 file changed, 12 insertions(+), 5 deletions(-)
diff --git a/kernel/workqueue.c b/kernel/workqueue.c
index ee16ddb0647c..3ad6806c7161 100644
--- a/kernel/workqueue.c
+++ b/kernel/workqueue.c
@@ -1051,7 +1051,7 @@ void wq_worker_running(struct task_struct *task)
{
struct worker *worker = kthread_data(task);
- if (!worker->sleeping)
+ if (!READ_ONCE(worker->sleeping))
return;
/*
@@ -1071,7 +1071,7 @@ void wq_worker_running(struct task_struct *task)
*/
worker->current_at = worker->task->se.sum_exec_runtime;
- worker->sleeping = 0;
+ WRITE_ONCE(worker->sleeping, 0);
}
/**
@@ -1097,10 +1097,10 @@ void wq_worker_sleeping(struct task_struct *task)
pool = worker->pool;
/* Return if preempted before wq_worker_running() was reached */
- if (worker->sleeping)
+ if (READ_ONCE(worker->sleeping))
return;
- worker->sleeping = 1;
+ WRITE_ONCE(worker->sleeping, 1);
raw_spin_lock_irq(&pool->lock);
/*
@@ -1143,8 +1143,15 @@ void wq_worker_tick(struct task_struct *task)
* If the current worker is concurrency managed and hogged the CPU for
* longer than wq_cpu_intensive_thresh_us, it's automatically marked
* CPU_INTENSIVE to avoid stalling other concurrency-managed work items.
+ *
+ * Set @worker->sleeping means that @worker is in the process of
+ * switching out voluntarily and won't be contributing to
+ * @pool->nr_running until it wakes up. As wq_worker_sleeping() also
+ * decrements ->nr_running, setting CPU_INTENSIVE here can lead to
+ * double decrements. The task is releasing the CPU anyway. Let's skip.
+ * We probably want to make this prettier in the future.
*/
- if ((worker->flags & WORKER_NOT_RUNNING) ||
+ if ((worker->flags & WORKER_NOT_RUNNING) || READ_ONCE(worker->sleeping) ||
worker->task->se.sum_exec_runtime - worker->current_at <
wq_cpu_intensive_thresh_us * NSEC_PER_USEC)
return;
--
2.40.1
^ permalink raw reply related [flat|nested] 4+ messages in thread
* Re: [PATCH v3] workqueue: Fix WARN_ON_ONCE() triggers in worker_enter_idle()
2023-05-24 13:53 ` Naresh Kamboju
@ 2023-05-24 22:03 ` Tejun Heo
0 siblings, 0 replies; 4+ messages in thread
From: Tejun Heo @ 2023-05-24 22:03 UTC (permalink / raw
To: Naresh Kamboju
Cc: Zqiang, jiangshanlai, linux-kernel, Anders Roxell, lkft-triage
On Wed, May 24, 2023 at 07:23:16PM +0530, Naresh Kamboju wrote:
> Reported-by: Linux Kernel Functional Testing <lkft@linaro.org>
> Tested-by: Anders Roxell <anders.roxell@linaro.org>
>
> Since the probability of occurrence of this problem is only 3%,
> Anders took this up and applied this on top of Linux next and
> tested for 500 boot tests and all looked good.
> Thanks, Anders.
This was a tricky bug and I really appreciate the bug report and testing.
Thank you so much.
--
tejun
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2023-05-24 22:03 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2023-05-24 3:53 [PATCH v3] workqueue: Fix WARN_ON_ONCE() triggers in worker_enter_idle() Zqiang
2023-05-24 13:53 ` Naresh Kamboju
2023-05-24 22:03 ` Tejun Heo
2023-05-24 22:02 ` Tejun Heo
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).