linux-kernel - Re: [PATCH v4 1/2] sched/fair: set rq->idle_stamp at the end of the sched_balance

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]

Message-ID: <e51f170b-416a-4a57-8e94-0971242c0473@amperemail.onmicrosoft.com>
Date: Fri, 28 Nov 2025 20:54:15 +0800
From: Shijie Huang <shijie@...eremail.onmicrosoft.com>
To: Madadi Vineeth Reddy <vineethr@...ux.ibm.com>,
 Huang Shijie <shijie@...amperecomputing.com>
Cc: mingo@...hat.com, peterz@...radead.org, juri.lelli@...hat.com,
 vincent.guittot@...aro.org, patches@...erecomputing.com, cl@...ux.com,
 Shubhang@...amperecomputing.com, dietmar.eggemann@....com,
 rostedt@...dmis.org, bsegall@...gle.com, mgorman@...e.de,
 linux-kernel@...r.kernel.org, vschneid@...hat.com, kprateek.nayak@....com
Subject: Re: [PATCH v4 1/2] sched/fair: set rq->idle_stamp at the end of the
 sched_balance_newidle


On 28/11/2025 18:07, Madadi Vineeth Reddy wrote:
>>   	 */
>> -	this_rq->idle_stamp = rq_clock(this_rq);
>> +	idle_stamp = rq_clock(this_rq);
>> +
>> +	this_rq->idle_stamp = 0;
> IIUC, by setting this_rq->idle_stamp = 0 at the beginning, any call to update_rq_avg_idle() during
> load balancing when tasks are pulled will fail the if (rq->idle_stamp) check, preventing the average
> idle time from being updated.

  1.) For the newidle balance, it is okay to prevent to update the 
rq->avg_idle.

  2.) For the idle balance, the this_rq->idle_stamp is not zero, and it 
can update the rq->avg_idle.

             Can the idle balance and newidle balance run at the same time?


  3.) For the busy balance, the  this_rq->idle_stamp should be zero, no 
need to update the rq->avg_idle.


Thanks

Huang Shijie