Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1637664 > unrolled thread

[PATCH] sched/fair: WARN and refuse to set buddy when !se->on_rq

Back to article view | Back to linux.kernel

Contents

  [PATCH] sched/fair: WARN and refuse to set buddy when !se->on_rq Daniel Axtens <dja@axtens.net> - 2017-05-08 21:30 +0200
    Re: [PATCH] sched/fair: WARN and refuse to set buddy when !se->on_rq Konstantin Khlebnikov <koct9i@gmail.com> - 2017-05-08 21:40 +0200
      Re: [PATCH] sched/fair: WARN and refuse to set buddy when !se->on_rq Daniel Axtens <dja@axtens.net> - 2017-05-09 03:50 +0200
      Re: [PATCH] sched/fair: WARN and refuse to set buddy when !se->on_rq Peter Zijlstra <peterz@infradead.org> - 2017-05-09 08:50 +0200

#1637664 — [PATCH] sched/fair: WARN and refuse to set buddy when !se->on_rq

If we set a next or last buddy for a se that is not on_rq, we will
end up taking a NULL pointer dereference in wakeup_preempt_entity
via pick_next_task_fair.

Detect when we would be about to do that, throw a warning and
then refuse to actually set it.

This has been suggested at least twice[0][1]: just do it.

[0] https://marc.info/?l=linux-kernel&m=146651668921468&w=2
[1] https://lkml.org/lkml/2016/6/16/663

Cc: Konstantin Khlebnikov <khlebnikov@yandex-team.ru>
Cc: Ben Segall <bsegall@google.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Daniel Axtens <dja@axtens.net>

---

I recently had to debug a problem with these (we hadn't backported
Konstantin's patches in this area) and this would have saved a lot
of time/pain.
---
 kernel/sched/fair.c | 10 ++++++++--
 1 file changed, 8 insertions(+), 2 deletions(-)

diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
index d71109321841..9b988815b4ad 100644
--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -6168,8 +6168,11 @@ static void set_last_buddy(struct sched_entity *se)
 	if (entity_is_task(se) && unlikely(task_of(se)->policy == SCHED_IDLE))
 		return;
 
-	for_each_sched_entity(se)
+	for_each_sched_entity(se) {
+		if (WARN_ON_ONCE(!se->on_rq))
+			return;
 		cfs_rq_of(se)->last = se;
+	}
 }
 
 static void set_next_buddy(struct sched_entity *se)
@@ -6177,8 +6180,11 @@ static void set_next_buddy(struct sched_entity *se)
 	if (entity_is_task(se) && unlikely(task_of(se)->policy == SCHED_IDLE))
 		return;
 
-	for_each_sched_entity(se)
+	for_each_sched_entity(se) {
+		if (WARN_ON_ONCE(!se->on_rq))
+			return;
 		cfs_rq_of(se)->next = se;
+	}
 }
 
 static void set_skip_buddy(struct sched_entity *se)
-- 
2.11.0

[toc] | [next] | [standalone]

#1637671

On Mon, May 8, 2017 at 10:22 PM, Daniel Axtens <dja@axtens.net> wrote:
> If we set a next or last buddy for a se that is not on_rq, we will
> end up taking a NULL pointer dereference in wakeup_preempt_entity
> via pick_next_task_fair.
>
> Detect when we would be about to do that, throw a warning and
> then refuse to actually set it.
>
> This has been suggested at least twice[0][1]: just do it.
>
> [0] https://marc.info/?l=linux-kernel&m=146651668921468&w=2
> [1] https://lkml.org/lkml/2016/6/16/663
>
> Cc: Konstantin Khlebnikov <khlebnikov@yandex-team.ru>
> Cc: Ben Segall <bsegall@google.com>
> Cc: Peter Zijlstra <peterz@infradead.org>
> Cc: Thomas Gleixner <tglx@linutronix.de>
> Signed-off-by: Daniel Axtens <dja@axtens.net>
>
> ---
>
> I recently had to debug a problem with these (we hadn't backported
> Konstantin's patches in this area) and this would have saved a lot
> of time/pain.

Yep, this is worth debug. But I think it should be under compile-time
config option.
Something like CONFIG_DEBUG_VM which enables VM_BUG_ON and VM_WARN_ON.

> ---
>  kernel/sched/fair.c | 10 ++++++++--
>  1 file changed, 8 insertions(+), 2 deletions(-)
>
> diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
> index d71109321841..9b988815b4ad 100644
> --- a/kernel/sched/fair.c
> +++ b/kernel/sched/fair.c
> @@ -6168,8 +6168,11 @@ static void set_last_buddy(struct sched_entity *se)
>         if (entity_is_task(se) && unlikely(task_of(se)->policy == SCHED_IDLE))
>                 return;
>
> -       for_each_sched_entity(se)
> +       for_each_sched_entity(se) {
> +               if (WARN_ON_ONCE(!se->on_rq))
> +                       return;
>                 cfs_rq_of(se)->last = se;
> +       }
>  }
>
>  static void set_next_buddy(struct sched_entity *se)
> @@ -6177,8 +6180,11 @@ static void set_next_buddy(struct sched_entity *se)
>         if (entity_is_task(se) && unlikely(task_of(se)->policy == SCHED_IDLE))
>                 return;
>
> -       for_each_sched_entity(se)
> +       for_each_sched_entity(se) {
> +               if (WARN_ON_ONCE(!se->on_rq))
> +                       return;
>                 cfs_rq_of(se)->next = se;
> +       }
>  }
>
>  static void set_skip_buddy(struct sched_entity *se)
> --
> 2.11.0
>

[toc] | [prev] | [next] | [standalone]

#1637812

Konstantin Khlebnikov <koct9i@gmail.com> writes:

> On Mon, May 8, 2017 at 10:22 PM, Daniel Axtens <dja@axtens.net> wrote:
>> If we set a next or last buddy for a se that is not on_rq, we will
>> end up taking a NULL pointer dereference in wakeup_preempt_entity
>> via pick_next_task_fair.
>>
>> Detect when we would be about to do that, throw a warning and
>> then refuse to actually set it.
>>
>> This has been suggested at least twice[0][1]: just do it.
>>
>> [0] https://marc.info/?l=linux-kernel&m=146651668921468&w=2
>> [1] https://lkml.org/lkml/2016/6/16/663
>>
>> Cc: Konstantin Khlebnikov <khlebnikov@yandex-team.ru>
>> Cc: Ben Segall <bsegall@google.com>
>> Cc: Peter Zijlstra <peterz@infradead.org>
>> Cc: Thomas Gleixner <tglx@linutronix.de>
>> Signed-off-by: Daniel Axtens <dja@axtens.net>
>>
>> ---
>>
>> I recently had to debug a problem with these (we hadn't backported
>> Konstantin's patches in this area) and this would have saved a lot
>> of time/pain.
>
> Yep, this is worth debug. But I think it should be under compile-time
> config option.
> Something like CONFIG_DEBUG_VM which enables VM_BUG_ON and VM_WARN_ON.

OK, will respin.

Regards,
Daniel

>
>> ---
>>  kernel/sched/fair.c | 10 ++++++++--
>>  1 file changed, 8 insertions(+), 2 deletions(-)
>>
>> diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
>> index d71109321841..9b988815b4ad 100644
>> --- a/kernel/sched/fair.c
>> +++ b/kernel/sched/fair.c
>> @@ -6168,8 +6168,11 @@ static void set_last_buddy(struct sched_entity *se)
>>         if (entity_is_task(se) && unlikely(task_of(se)->policy == SCHED_IDLE))
>>                 return;
>>
>> -       for_each_sched_entity(se)
>> +       for_each_sched_entity(se) {
>> +               if (WARN_ON_ONCE(!se->on_rq))
>> +                       return;
>>                 cfs_rq_of(se)->last = se;
>> +       }
>>  }
>>
>>  static void set_next_buddy(struct sched_entity *se)
>> @@ -6177,8 +6180,11 @@ static void set_next_buddy(struct sched_entity *se)
>>         if (entity_is_task(se) && unlikely(task_of(se)->policy == SCHED_IDLE))
>>                 return;
>>
>> -       for_each_sched_entity(se)
>> +       for_each_sched_entity(se) {
>> +               if (WARN_ON_ONCE(!se->on_rq))
>> +                       return;
>>                 cfs_rq_of(se)->next = se;
>> +       }
>>  }
>>
>>  static void set_skip_buddy(struct sched_entity *se)
>> --
>> 2.11.0
>>

[toc] | [prev] | [next] | [standalone]

#1637895

On Mon, May 08, 2017 at 10:32:30PM +0300, Konstantin Khlebnikov wrote:
> On Mon, May 8, 2017 at 10:22 PM, Daniel Axtens <dja@axtens.net> wrote:

> > I recently had to debug a problem with these (we hadn't backported
> > Konstantin's patches in this area) and this would have saved a lot
> > of time/pain.
> 
> Yep, this is worth debug. But I think it should be under compile-time
> config option.
> Something like CONFIG_DEBUG_VM which enables VM_BUG_ON and VM_WARN_ON.

See SCHED_WARN_ON() and feel free to introduce SCHED_WARN_ON_ONCE() if
you feel that's needed.

[toc] | [prev] | [standalone]

Back to top | Article view | linux.kernel

csiph-web