linux-kernel - Re: [PATCH v2 3/3] sched/fair: do not scribble cache-hot CPU in select_idle

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Fri, 1 Dec 2023 19:26:40 +0530
From:   Madadi Vineeth Reddy <vineethr@...ux.ibm.com>
To:     Chen Yu <yu.chen.surf@...il.com>
Cc:     Chen Yu <yu.c.chen@...el.com>, Tim Chen <tim.c.chen@...el.com>,
        Aaron Lu <aaron.lu@...el.com>,
        Dietmar Eggemann <dietmar.eggemann@....com>,
        Steven Rostedt <rostedt@...dmis.org>,
        Mel Gorman <mgorman@...e.de>,
        K Prateek Nayak <kprateek.nayak@....com>,
        "Gautham R . Shenoy" <gautham.shenoy@....com>,
        linux-kernel@...r.kernel.org,
        Peter Zijlstra <peterz@...radead.org>,
        Mathieu Desnoyers <mathieu.desnoyers@...icios.com>,
        Ingo Molnar <mingo@...hat.com>,
        Vincent Guittot <vincent.guittot@...aro.org>,
        Juri Lelli <juri.lelli@...hat.com>,
        Madadi Vineeth Reddy <vineethr@...ux.ibm.com>
Subject: Re: [PATCH v2 3/3] sched/fair: do not scribble cache-hot CPU in
 select_idle_cpu()

On 30/11/23 12:13, Chen Yu wrote:
> Hi Madadi,
> 
> On Thu, Nov 30, 2023 at 1:26 AM Madadi Vineeth Reddy
> <vineethr@...ux.ibm.com> wrote:
>>
>> Hi Chen Yu,
>>
> [snip...]
> 
>>
>> As per my understanding, if the task which tagged a particular CPU as cache hot has woken up before
>> the cache_hot_timeout, we still don't allow that task to run on that particular CPU. Is this correct?
>>
> 
> Not exactly. When we reached select_idle_cpu(), we have already
> checked if the wakee's previous CPU
> is idle or not in select_idle_sibling(), if it is idle, we don't check
> the cache-hot tag and return the wakee's
> previous CPU directly in select_idle_sibling().

Yeah..got it. Thanks.

So, the way another task(t') can select the cache hot tagged cpu(tagged for task t) is when t' gets 
it's target cpu as this cache hot tagged cpu from wake_affine() in select_task_rq_fair. 
The other way being /* pick the first cache-hot CPU as the last resort */.

> 
> thanks,
> Chenyu
> 
>> If correct, then why don't we let the task to select the CPU if it's the one that tagged it?
>>
>> Thanks and Regards
>> Madadi Vineeth Reddy

Thanks and Regards
Madadi Vineeth Reddy