RE: [PATCH] mm: vmscan: avoid a unnecessary reschedule in shrink_slab()

[Tianxianting]

Thanks for the explain.
I got it.

-----Original Message-----
From: Michal Hocko [mailto:mhocko@xxxxxxxx] 
Sent: Friday, October 16, 2020 9:45 PM
To: tianxianting (RD) <tian.xianting@xxxxxxx>
Cc: akpm@xxxxxxxxxxxxxxxxxxxx; linux-mm@xxxxxxxxx; linux-kernel@xxxxxxxxxxxxxxx
Subject: Re: [PATCH] mm: vmscan: avoid a unnecessary reschedule in shrink_slab()

On Fri 16-10-20 13:20:41, Tianxianting wrote:
> Thanks
> I understood what you said :)
> But whether it is proper to check reschedule in every loop when lock is taken? 

I do not see any actual problem TBH. cond_resched is mostly to increase interactivity for non preemptible kernel. It can reduce throughput but this is a memory reclaim path and I do not expect this to contribute to any moderate hot paths. Direct reclaim doesn't really count as a hot path.

--
Michal Hocko
SUSE Labs