lkml.org 
[lkml]   [2019]   [Apr]   [6]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
SubjectRe: [PATCH v13 2/3] x86,/proc/pid/status: Add AVX-512 usage elapsed time
From
Date
On 2019/4/6 4:27, Jann Horn wrote:
> On Fri, Apr 5, 2019 at 10:02 PM Aubrey Li <aubrey.li@linux.intel.com> wrote:
>> AVX-512 components use could cause core turbo frequency drop. So
>> it's useful to expose AVX-512 usage elapsed time as a heuristic hint
>> for the user space job scheduler to cluster the AVX-512 using tasks
>> together.
>>
>> Tensorflow example:
>> $ while [ 1 ]; do cat /proc/pid/status | grep AVX; sleep 1; done
>> AVX512_elapsed_ms: 4
>> AVX512_elapsed_ms: 8
>> AVX512_elapsed_ms: 4
>>
>> This means that 4 milliseconds have elapsed since the AVX512 usage
>> of tensorflow task was detected when the task was scheduled out.
>>
>> Or:
>> $ cat /proc/pid/status | grep AVX512_elapsed_ms
>> AVX512_elapsed_ms: -1
>
> (Very nitpicky, feel free to ignore: If you change the /proc/pid to
> /proc/tid in the commit message, it becomes clearer that this status
> is really per-task/thread, not per-process/threadgroup.)

Thanks, I'll refine.

>
> [...]
>> +
>> +/*
>> + * Report the amount of time elapsed in millisecond since last AVX512
>> + * use in the task.
>> + */
>> +static void avx512_status(struct seq_file *m, struct task_struct *task)
>> +{
>> + unsigned long timestamp = task->thread.fpu.avx512_timestamp;
>
> This is theoretically a data race, right? Should this have a READ_ONCE() on it?

Thanks, I'll refine.

>
> Is there something that zeroes out the avx512_timestamp on
> fork()/clone(), or will every child inherit the avx512 timestamp? As
> far as I can tell, the timestamp is inherited; I think it would be
> nicer to zero it out at that point. Either way, It might be worth
> documenting this decision.
>

This timestamp is not inherited, see below:

_do_fork()
->copy_process()
-->dup_task_struct()
--->arch_dup_task_struct()
---->fpu__copy()

Thanks,
-Aubrey

\
 
 \ /
  Last update: 2019-04-06 17:43    [W:0.072 / U:0.976 seconds]
©2003-2018 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site