Re: bug in sched.c:activate_task()

[Nick Piggin <nickpiggin@yahoo.com.au>]

Con Kolivas wrote:
> Ingo Molnar writes:
> 
>>
>> On Mon, 4 Oct 2004, Chen, Kenneth W wrote:
>>
>>> Update p->timestamp to "now" in activate_task() doesn't look right to me
>>> at all.  p->timestamp records last time it was running on a cpu.  
>>> activate_task shouldn't update that variable when it queues a task on
>>> the runqueue.
>>
>>
>> correct, we are overriding it in schedule():
>>
>>         if (likely(prev != next)) {
>>                 next->timestamp = now;
>>                 rq->nr_switches++;
>>
>> the line your patch removes is a remnant of an earlier logic when we
>> timestamped tasks when they touched the runqueue. (vs. timestamping when
>> they actually run on a CPU.) So the patch looks good to me. Andrew, 
>> please
>> apply.
> 
> 
>     unsigned long long delta = now - next->timestamp;
> 
>     if (next->activated == 1)
>         delta = delta * (ON_RUNQUEUE_WEIGHT * 128 / 100) / 128;
> 
> is in schedule() before we update the timestamp, no?
> 

Yeah right, unfortunately.