Re: [RFC PATCH v2 4/8] tick/nohz: keep tick on for a fast idle

[Li, Aubrey]

On 2017/10/16 14:25, Mike Galbraith wrote:
> On Mon, 2017-10-16 at 13:34 +0800, Li, Aubrey wrote:
>> On 2017/10/16 12:45, Mike Galbraith wrote:
>>> On Mon, 2017-10-16 at 11:26 +0800, Li, Aubrey wrote:
>>>>
>>>> I'll try to move quiet_vmstat() into the normal idle branch if this patch series
>>>> are reasonable. Is fast_idle a good indication for it?
>>>
>>> see x86_tip 62cb1188ed86 sched/idle: Move quiet_vmstate() into the NOHZ code
>>
>> It looks like this commit makes tick stop critical as it can be invoked in interrupt
>> exit path?
> 
> do_idle() ain't critical?  It is in my book.  Hopefully, you're about
> to make that idle_stat.fast_idle thingy liberal enough that we cease
> mucking about with the tick on every microscopic idle (and I can then
> trash my years old local patch to avoid needlessly eating that cost).
> 

Welcome any feedback of other patch set, Mike, :)

Thanks,
-Aubrey