linux-kernel - Re: [PATCH 2/2] sched/deadline: fix start high-res preemption tick for a non-running task

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Date:	Mon, 17 Nov 2014 19:02:10 +0800
From:	Wanpeng Li <kernellwp@...il.com>
To:	Juri Lelli <juri.lelli@....com>,
	Wanpeng Li <wanpeng.li@...ux.intel.com>,
	Ingo Molnar <mingo@...hat.com>,
	Peter Zijlstra <peterz@...radead.org>
CC:	Kirill Tkhai <ktkhai@...allels.com>,
	"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
	"juri.lelli@...il.com" <juri.lelli@...il.com>
Subject: Re: [PATCH 2/2] sched/deadline: fix start high-res preemption tick
 for a non-running task

Hi Juri,
On 11/17/14, 5:50 PM, Juri Lelli wrote:
> Hi,
>
> On 13/11/14 10:30, Wanpeng Li wrote:
>> Hi Juri,
>> On 11/13/14, 5:37 PM, Juri Lelli wrote:
>>> Hi,
>>>
>>> not sure I understand what the problem is here.
>>>
>>> On 11/11/14 01:52, Wanpeng Li wrote:
>>>> Queued ticks are scheduled to match the budget, which means the budget
>>>> is overall consumed and the dl task should be throttled.
>>> ... enforce the budget? It means that when the budget is consumed the
>>> task has to be throttled...
>> Agreed.
>>
>>>> Dl task will
>>>> be replenished immediately if fail to start a dl timer.
>>>>
>>>> However, the curr maybe not the left most dl task in the rb tree any
>>>> more after this immediately replenished and reschedule is needed.
>>>> Start high-res preemption tick for this upcoming rescheduled dl task
>>>> is not correct.
>>>>
>>> So, the task that is going to preempt curr is picked by
>>> pick_next_task_dl(), that correctly starts the hrtick for this new task.
>>>
>>> Maybe you can add more information about what you are seeing? A callpath
>>> maybe?
>> The parameter of task_tick_dl() queued == 1 means that hrtick is fired.
>> hrtick() => task_tick_dl( , ,1), so p->dl.runtime should be <= 0 if
>> queued == 1. What I see is queued == 1 && p->dl.runtime > 0 && p is not
>> the left most task and hrtick is start for this task.
>>
> Oh, right, now it makes sense. Good catch! :)
>
> Could you please post a second version with a more explanatory changelog
> and maybe a comment just above the check?

Ok, I will do it and send out a new version tomorrow. ;-) Btw, could you 
review this patch? https://lkml.org/lkml/2014/11/13/51 The bug is 
introduced by deadline.

Regards,
Wanpeng Li

>   
> Thanks,
>
> - Juri
>
>> Regards,
>> Wanpeng Li
>>
>>> Thanks,
>>>
>>> - Juri
>>>
>>>> This patch fix it by not starting high-res preemption tick for a
>>>> non-running dl task.
>>>>
>>>> Signed-off-by: Wanpeng Li <wanpeng.li@...ux.intel.com>
>>>> ---
>>>>    kernel/sched/deadline.c | 3 ++-
>>>>    1 file changed, 2 insertions(+), 1 deletion(-)
>>>>
>>>> diff --git a/kernel/sched/deadline.c b/kernel/sched/deadline.c
>>>> index 56674f6..2a6a5bb 100644
>>>> --- a/kernel/sched/deadline.c
>>>> +++ b/kernel/sched/deadline.c
>>>> @@ -1090,7 +1090,8 @@ static void task_tick_dl(struct rq *rq, struct task_struct *p, int queued)
>>>>    {
>>>>    	update_curr_dl(rq);
>>>>    
>>>> -	if (hrtick_enabled(rq) && queued && p->dl.runtime > 0)
>>>> +	if (hrtick_enabled(rq) && queued && p->dl.runtime > 0 &&
>>>> +	    is_leftmost(p, &rq->dl))
>>>>    		start_hrtick_dl(rq, p);
>>>>    }
>>>>    
>>>>
>>> --
>>> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
>>> the body of a message to majordomo@...r.kernel.org
>>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>>> Please read the FAQ at  http://www.tux.org/lkml/
>>

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/