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

Groups > linux.kernel > #1724986 > unrolled thread

[PATCH] sched: reset sysctl_sched_time_avg to default when

Started byEthan Zhao <ethan.zhao@oracle.com>
First post2017-09-01 13:40 +0200
Last post2017-09-04 09:50 +0200
Articles 6 — 3 participants

Back to article view | Back to linux.kernel

Contents

  [PATCH] sched: reset sysctl_sched_time_avg to default when  Ethan Zhao <ethan.zhao@oracle.com> - 2017-09-01 13:40 +0200
    Re: [PATCH] sched: reset sysctl_sched_time_avg to default when Peter Zijlstra <peterz@infradead.org> - 2017-09-01 14:40 +0200
      Re: [PATCH] sched: reset sysctl_sched_time_avg to default when Ethan Zhao <ethan.kernel@gmail.com> - 2017-09-02 02:40 +0200
        Re: [PATCH] sched: reset sysctl_sched_time_avg to default when Ethan Zhao <ethan.kernel@gmail.com> - 2017-09-02 09:00 +0200
        Re: [PATCH] sched: reset sysctl_sched_time_avg to default when Peter Zijlstra <peterz@infradead.org> - 2017-09-04 09:40 +0200
          Re: [PATCH] sched: reset sysctl_sched_time_avg to default when Ethan Zhao <ethan.zhao@oracle.com> - 2017-09-04 09:50 +0200

#1724986 — [PATCH] sched: reset sysctl_sched_time_avg to default when

FromEthan Zhao <ethan.zhao@oracle.com>
Date2017-09-01 13:40 +0200
Subject[PATCH] sched: reset sysctl_sched_time_avg to default when
Message-ID<ukSsO-3FA-9@gated-at.bofh.it>
System will hang if user set sysctl_sched_time_avg to 0 by

[root@XXX ~]# sysctl kernel.sched_time_avg_ms=0

Stack traceback for pid 0
0xffff883f6406c600 0 0 1 3 R 0xffff883f6406cf50 *swapper/3
ffff883f7ccc3ae8 0000000000000018 ffffffff810c4dd0 0000000000000000
0000000000017800 ffff883f7ccc3d78 0000000000000003 ffff883f7ccc3bf8
ffffffff810c4fc9 ffff883f7ccc3c08 00000000810c5043 ffff883f7ccc3c08
Call Trace:
<IRQ> [<ffffffff810c4dd0>] ? update_group_capacity+0x110/0x200
[<ffffffff810c4fc9>] ? update_sd_lb_stats+0x109/0x600
[<ffffffff810c5507>] ? find_busiest_group+0x47/0x530
[<ffffffff810c5b84>] ? load_balance+0x194/0x900
[<ffffffff810ad5ca>] ? update_rq_clock.part.83+0x1a/0xe0
[<ffffffff810c6d42>] ? rebalance_domains+0x152/0x290
[<ffffffff810c6f5c>] ? run_rebalance_domains+0xdc/0x1d0
[<ffffffff8108a75b>] ? __do_softirq+0xfb/0x320
[<ffffffff8108ac85>] ? irq_exit+0x125/0x130
[<ffffffff810b3a17>] ? scheduler_ipi+0x97/0x160
[<ffffffff81052709>] ? smp_reschedule_interrupt+0x29/0x30
[<ffffffff8173a1be>] ? reschedule_interrupt+0x6e/0x80
<EOI> [<ffffffff815bc83c>] ? cpuidle_enter_state+0xcc/0x230
[<ffffffff815bc80c>] ? cpuidle_enter_state+0x9c/0x230
[<ffffffff815bc9d7>] ? cpuidle_enter+0x17/0x20
[<ffffffff810cd6dc>] ? cpu_startup_entry+0x38c/0x420
[<ffffffff81053373>] ? start_secondary+0x173/0x1e0

Because divide-by-zero error happens in function

update_group_capacity()
  update_cpu_capacity()
    scale_rt_capacity()
	{
		...
		total = sched_avg_period() + delta;
		used = div_u64(avg, total);
		...
	}

Seems this issue could be reproduced on all I tried stable 4.1 - last
kernel.

To fix this issue, reset sysctl_sched_time_avg to default value
MSEC_PER_SEC if user input invalid value in function

sched_avg_period()

Reported-by: James Puthukattukaran <james.puthukattukaran@oracle.com>
Signed-off-by: Ethan Zhao <ethan.zhao@oracle.com>
---
 Tested on stable 4.1, compiled on stable 4.13-rc5 
 kernel/sched/sched.h | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h
index eeef1a3..b398560 100644
--- a/kernel/sched/sched.h
+++ b/kernel/sched/sched.h
@@ -1620,6 +1620,9 @@ static inline void rq_last_tick_reset(struct rq *rq)
 
 static inline u64 sched_avg_period(void)
 {
+	if (unlikely(sysctl_sched_time_avg <= 0))
+		sysctl_sched_time_avg = MSEC_PER_SEC;
+
 	return (u64)sysctl_sched_time_avg * NSEC_PER_MSEC / 2;
 }
 
-- 
1.8.3.1

[toc] | [next] | [standalone]

#1725021

FromPeter Zijlstra <peterz@infradead.org>
Date2017-09-01 14:40 +0200
Message-ID<ukToS-4zk-13@gated-at.bofh.it>
In reply to#1724986
On Fri, Sep 01, 2017 at 07:31:54PM +0800, Ethan Zhao wrote:
> System will hang if user set sysctl_sched_time_avg to 0 by
> 
> [root@XXX ~]# sysctl kernel.sched_time_avg_ms=0
> 
> Stack traceback for pid 0
> 0xffff883f6406c600 0 0 1 3 R 0xffff883f6406cf50 *swapper/3
> ffff883f7ccc3ae8 0000000000000018 ffffffff810c4dd0 0000000000000000
> 0000000000017800 ffff883f7ccc3d78 0000000000000003 ffff883f7ccc3bf8
> ffffffff810c4fc9 ffff883f7ccc3c08 00000000810c5043 ffff883f7ccc3c08
> Call Trace:
> <IRQ> [<ffffffff810c4dd0>] ? update_group_capacity+0x110/0x200
> [<ffffffff810c4fc9>] ? update_sd_lb_stats+0x109/0x600
> [<ffffffff810c5507>] ? find_busiest_group+0x47/0x530
> [<ffffffff810c5b84>] ? load_balance+0x194/0x900
> [<ffffffff810ad5ca>] ? update_rq_clock.part.83+0x1a/0xe0
> [<ffffffff810c6d42>] ? rebalance_domains+0x152/0x290
> [<ffffffff810c6f5c>] ? run_rebalance_domains+0xdc/0x1d0
> [<ffffffff8108a75b>] ? __do_softirq+0xfb/0x320
> [<ffffffff8108ac85>] ? irq_exit+0x125/0x130
> [<ffffffff810b3a17>] ? scheduler_ipi+0x97/0x160
> [<ffffffff81052709>] ? smp_reschedule_interrupt+0x29/0x30
> [<ffffffff8173a1be>] ? reschedule_interrupt+0x6e/0x80
> <EOI> [<ffffffff815bc83c>] ? cpuidle_enter_state+0xcc/0x230
> [<ffffffff815bc80c>] ? cpuidle_enter_state+0x9c/0x230
> [<ffffffff815bc9d7>] ? cpuidle_enter+0x17/0x20
> [<ffffffff810cd6dc>] ? cpu_startup_entry+0x38c/0x420
> [<ffffffff81053373>] ? start_secondary+0x173/0x1e0
> 
> Because divide-by-zero error happens in function
> 
> update_group_capacity()
>   update_cpu_capacity()
>     scale_rt_capacity()
> 	{
> 		...
> 		total = sched_avg_period() + delta;
> 		used = div_u64(avg, total);
> 		...
> 	}
> 
> Seems this issue could be reproduced on all I tried stable 4.1 - last
> kernel.
> 
> To fix this issue, reset sysctl_sched_time_avg to default value
> MSEC_PER_SEC if user input invalid value in function
> 
> sched_avg_period()
> 
> Reported-by: James Puthukattukaran <james.puthukattukaran@oracle.com>
> Signed-off-by: Ethan Zhao <ethan.zhao@oracle.com>
> ---
>  Tested on stable 4.1, compiled on stable 4.13-rc5 
>  kernel/sched/sched.h | 3 +++
>  1 file changed, 3 insertions(+)
> 
> diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h
> index eeef1a3..b398560 100644
> --- a/kernel/sched/sched.h
> +++ b/kernel/sched/sched.h
> @@ -1620,6 +1620,9 @@ static inline void rq_last_tick_reset(struct rq *rq)
>  
>  static inline u64 sched_avg_period(void)
>  {
> +	if (unlikely(sysctl_sched_time_avg <= 0))
> +		sysctl_sched_time_avg = MSEC_PER_SEC;
> +

Sight, no.. you can set limits in the sysctl table.

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

#1725408

FromEthan Zhao <ethan.kernel@gmail.com>
Date2017-09-02 02:40 +0200
Message-ID<ul4DD-473-1@gated-at.bofh.it>
In reply to#1725021
Yep, that is the first place I considered to set the limit, but that would
break KABI ?

Thanks,
Ethan

On Fri, Sep 1, 2017 at 8:32 PM, Peter Zijlstra <peterz@infradead.org> wrote:
> On Fri, Sep 01, 2017 at 07:31:54PM +0800, Ethan Zhao wrote:
>> System will hang if user set sysctl_sched_time_avg to 0 by
>>
>> [root@XXX ~]# sysctl kernel.sched_time_avg_ms=0
>>
>> Stack traceback for pid 0
>> 0xffff883f6406c600 0 0 1 3 R 0xffff883f6406cf50 *swapper/3
>> ffff883f7ccc3ae8 0000000000000018 ffffffff810c4dd0 0000000000000000
>> 0000000000017800 ffff883f7ccc3d78 0000000000000003 ffff883f7ccc3bf8
>> ffffffff810c4fc9 ffff883f7ccc3c08 00000000810c5043 ffff883f7ccc3c08
>> Call Trace:
>> <IRQ> [<ffffffff810c4dd0>] ? update_group_capacity+0x110/0x200
>> [<ffffffff810c4fc9>] ? update_sd_lb_stats+0x109/0x600
>> [<ffffffff810c5507>] ? find_busiest_group+0x47/0x530
>> [<ffffffff810c5b84>] ? load_balance+0x194/0x900
>> [<ffffffff810ad5ca>] ? update_rq_clock.part.83+0x1a/0xe0
>> [<ffffffff810c6d42>] ? rebalance_domains+0x152/0x290
>> [<ffffffff810c6f5c>] ? run_rebalance_domains+0xdc/0x1d0
>> [<ffffffff8108a75b>] ? __do_softirq+0xfb/0x320
>> [<ffffffff8108ac85>] ? irq_exit+0x125/0x130
>> [<ffffffff810b3a17>] ? scheduler_ipi+0x97/0x160
>> [<ffffffff81052709>] ? smp_reschedule_interrupt+0x29/0x30
>> [<ffffffff8173a1be>] ? reschedule_interrupt+0x6e/0x80
>> <EOI> [<ffffffff815bc83c>] ? cpuidle_enter_state+0xcc/0x230
>> [<ffffffff815bc80c>] ? cpuidle_enter_state+0x9c/0x230
>> [<ffffffff815bc9d7>] ? cpuidle_enter+0x17/0x20
>> [<ffffffff810cd6dc>] ? cpu_startup_entry+0x38c/0x420
>> [<ffffffff81053373>] ? start_secondary+0x173/0x1e0
>>
>> Because divide-by-zero error happens in function
>>
>> update_group_capacity()
>>   update_cpu_capacity()
>>     scale_rt_capacity()
>>       {
>>               ...
>>               total = sched_avg_period() + delta;
>>               used = div_u64(avg, total);
>>               ...
>>       }
>>
>> Seems this issue could be reproduced on all I tried stable 4.1 - last
>> kernel.
>>
>> To fix this issue, reset sysctl_sched_time_avg to default value
>> MSEC_PER_SEC if user input invalid value in function
>>
>> sched_avg_period()
>>
>> Reported-by: James Puthukattukaran <james.puthukattukaran@oracle.com>
>> Signed-off-by: Ethan Zhao <ethan.zhao@oracle.com>
>> ---
>>  Tested on stable 4.1, compiled on stable 4.13-rc5
>>  kernel/sched/sched.h | 3 +++
>>  1 file changed, 3 insertions(+)
>>
>> diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h
>> index eeef1a3..b398560 100644
>> --- a/kernel/sched/sched.h
>> +++ b/kernel/sched/sched.h
>> @@ -1620,6 +1620,9 @@ static inline void rq_last_tick_reset(struct rq *rq)
>>
>>  static inline u64 sched_avg_period(void)
>>  {
>> +     if (unlikely(sysctl_sched_time_avg <= 0))
>> +             sysctl_sched_time_avg = MSEC_PER_SEC;
>> +
>
> Sight, no.. you can set limits in the sysctl table.

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

#1725458

FromEthan Zhao <ethan.kernel@gmail.com>
Date2017-09-02 09:00 +0200
Message-ID<ulazn-81X-3@gated-at.bofh.it>
In reply to#1725408
Peter,

 V2 sent, tested okay on stable 4.1, re-factor a little, could apply
to v4.13-rc5,
please review.

[root@ ~]# sysctl kernel.sched_time_avg_ms
kernel.sched_time_avg_ms = 1000
[root@ ~]# sysctl kernel.sched_time_avg_ms=0
sysctl: setting key "kernel.sched_time_avg_ms": Invalid argument
kernel.sched_time_avg_ms = 0
[root@ ~]# sysctl kernel.sched_time_avg_ms
kernel.sched_time_avg_ms = 1000
[root@ ~]# sysctl kernel.sched_time_avg_ms=900
kernel.sched_time_avg_ms = 900
[root@ ~]# sysctl kernel.sched_time_avg_ms
kernel.sched_time_avg_ms = 900

Thanks,
Ethan

On Sat, Sep 2, 2017 at 8:33 AM, Ethan Zhao <ethan.kernel@gmail.com> wrote:
> Yep, that is the first place I considered to set the limit, but that would
> break KABI ?
>
> Thanks,
> Ethan
>
> On Fri, Sep 1, 2017 at 8:32 PM, Peter Zijlstra <peterz@infradead.org> wrote:
>> On Fri, Sep 01, 2017 at 07:31:54PM +0800, Ethan Zhao wrote:
>>> System will hang if user set sysctl_sched_time_avg to 0 by
>>>
>>> [root@XXX ~]# sysctl kernel.sched_time_avg_ms=0
>>>
>>> Stack traceback for pid 0
>>> 0xffff883f6406c600 0 0 1 3 R 0xffff883f6406cf50 *swapper/3
>>> ffff883f7ccc3ae8 0000000000000018 ffffffff810c4dd0 0000000000000000
>>> 0000000000017800 ffff883f7ccc3d78 0000000000000003 ffff883f7ccc3bf8
>>> ffffffff810c4fc9 ffff883f7ccc3c08 00000000810c5043 ffff883f7ccc3c08
>>> Call Trace:
>>> <IRQ> [<ffffffff810c4dd0>] ? update_group_capacity+0x110/0x200
>>> [<ffffffff810c4fc9>] ? update_sd_lb_stats+0x109/0x600
>>> [<ffffffff810c5507>] ? find_busiest_group+0x47/0x530
>>> [<ffffffff810c5b84>] ? load_balance+0x194/0x900
>>> [<ffffffff810ad5ca>] ? update_rq_clock.part.83+0x1a/0xe0
>>> [<ffffffff810c6d42>] ? rebalance_domains+0x152/0x290
>>> [<ffffffff810c6f5c>] ? run_rebalance_domains+0xdc/0x1d0
>>> [<ffffffff8108a75b>] ? __do_softirq+0xfb/0x320
>>> [<ffffffff8108ac85>] ? irq_exit+0x125/0x130
>>> [<ffffffff810b3a17>] ? scheduler_ipi+0x97/0x160
>>> [<ffffffff81052709>] ? smp_reschedule_interrupt+0x29/0x30
>>> [<ffffffff8173a1be>] ? reschedule_interrupt+0x6e/0x80
>>> <EOI> [<ffffffff815bc83c>] ? cpuidle_enter_state+0xcc/0x230
>>> [<ffffffff815bc80c>] ? cpuidle_enter_state+0x9c/0x230
>>> [<ffffffff815bc9d7>] ? cpuidle_enter+0x17/0x20
>>> [<ffffffff810cd6dc>] ? cpu_startup_entry+0x38c/0x420
>>> [<ffffffff81053373>] ? start_secondary+0x173/0x1e0
>>>
>>> Because divide-by-zero error happens in function
>>>
>>> update_group_capacity()
>>>   update_cpu_capacity()
>>>     scale_rt_capacity()
>>>       {
>>>               ...
>>>               total = sched_avg_period() + delta;
>>>               used = div_u64(avg, total);
>>>               ...
>>>       }
>>>
>>> Seems this issue could be reproduced on all I tried stable 4.1 - last
>>> kernel.
>>>
>>> To fix this issue, reset sysctl_sched_time_avg to default value
>>> MSEC_PER_SEC if user input invalid value in function
>>>
>>> sched_avg_period()
>>>
>>> Reported-by: James Puthukattukaran <james.puthukattukaran@oracle.com>
>>> Signed-off-by: Ethan Zhao <ethan.zhao@oracle.com>
>>> ---
>>>  Tested on stable 4.1, compiled on stable 4.13-rc5
>>>  kernel/sched/sched.h | 3 +++
>>>  1 file changed, 3 insertions(+)
>>>
>>> diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h
>>> index eeef1a3..b398560 100644
>>> --- a/kernel/sched/sched.h
>>> +++ b/kernel/sched/sched.h
>>> @@ -1620,6 +1620,9 @@ static inline void rq_last_tick_reset(struct rq *rq)
>>>
>>>  static inline u64 sched_avg_period(void)
>>>  {
>>> +     if (unlikely(sysctl_sched_time_avg <= 0))
>>> +             sysctl_sched_time_avg = MSEC_PER_SEC;
>>> +
>>
>> Sight, no.. you can set limits in the sysctl table.

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

#1725894

FromPeter Zijlstra <peterz@infradead.org>
Date2017-09-04 09:40 +0200
Message-ID<ulU9c-2BO-23@gated-at.bofh.it>
In reply to#1725408
On Sat, Sep 02, 2017 at 08:33:03AM +0800, Ethan Zhao wrote:
> Yep, that is the first place I considered to set the limit, but that would
> break KABI ?

nah..

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

#1725899

FromEthan Zhao <ethan.zhao@oracle.com>
Date2017-09-04 09:50 +0200
Message-ID<ulUiR-2F5-5@gated-at.bofh.it>
In reply to#1725894
Peter,


On 2017/9/4 15:32, Peter Zijlstra wrote:
> On Sat, Sep 02, 2017 at 08:33:03AM +0800, Ethan Zhao wrote:
>> Yep, that is the first place I considered to set the limit, but that would
>> break KABI ?
> nah..
   V4 sent, please ignore v2 & v3.

   Thanks,
   Ethan

[toc] | [prev] | [standalone]

Back to top | Article view | linux.kernel

csiph-web