Re: [PATCH] mm,oom: Don't call schedule_timeout_killable() with oom_lock held.

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On 2018/06/06 18:02, David Rientjes wrote:
> On Mon, 4 Jun 2018, Tetsuo Handa wrote:
> 
>> Is current version of "mm, oom: cgroup-aware OOM killer" patchset going to be
>> dropped for now? I want to know which state should I use for baseline for my patch.
>>
> 
> My patchset to fix the issues with regard to the cgroup-aware oom killer 
> to fix its calculations (current version in -mm is completey buggy for 
> oom_score_adj, fixed in my patch 4/6), its context based errors 
> (discounting mempolicy oom kills, fixed in my patch 6/6) and make it 
> generally useful beyond highly specialized usecases in a backwards 
> compatible way was posted on March 22 at 
> https://marc.info/?l=linux-kernel&m=152175564104466.
> 
> The base patchset is seemingly abandoned in -mm, unfortunately, so I think 
> all oom killer patches should be based on Linus's tree.
> 
OK. I will use linux.git as a base.

By the way, does "[RFC] Getting rid of INFLIGHT_VICTIM" simplify or break
your cgroup-aware oom killer? If it simplifies your work, I'd like to apply
it as well.




[Index of Archives]     [Linux ARM Kernel]     [Linux ARM]     [Linux Omap]     [Fedora ARM]     [IETF Annouce]     [Bugtraq]     [Linux OMAP]     [Linux MIPS]     [eCos]     [Asterisk Internet PBX]     [Linux API]

  Powered by Linux