linux-kernel - Re: [PATCH] sched/fair: Add more core cookie check in wake up fast path

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] [day] [month] [year] [list]

Message-ID: <20251121111754.374593-1-sieberf@amazon.com>
Date: Fri, 21 Nov 2025 13:17:54 +0200
From: Fernand Sieber <sieberf@...zon.com>
To: <kprateek.nayak@....com>
CC: <mingo@...hat.com>, <peterz@...radead.org>,
	<linux-kernel@...r.kernel.org>, <juri.lelli@...hat.com>,
	<vincent.guittot@...aro.org>, <dietmar.eggemann@....com>,
	<rostedt@...dmis.org>, <bsegall@...gle.com>, <mgorman@...hsingularity.net>,
	<vschneid@...hat.com>, <dwmw@...zon.co.uk>, <jschoenh@...zon.de>,
	<liuyuxua@...zon.com>
Subject: Re: [PATCH] sched/fair: Add more core cookie check in wake up fast path

Hi Prateek,

Thank you for taking the time to review my patch.

On 11/21/2025 2:54 AM, K Prateek Nayak wrote:
> nit.
>
> You can replace the whole
>
>   (available_idle_cpu() || sched_idle_cpu()) && sched_core_cookie_match()
>
> with "__select_idle_cpu() != 1" [...]

I'm happy to do some code cleanup, however note that `__select_idle_cpu`
is using `sched_cpu_cookie_match` whereas at this point in
`select_idle_sibling` we don't know whether there's a full idle core or not
available so I am proposing to use `sched_core_cookie_match`.

We might be able to do the following alternatives:
* use `__select_idle_cpu` (e.g `sched_cpu_cookie_match`) and ignore the case
  where the cpu might be idle but one of its siblings run an incompatible
  cookie. Perhaps simpler code but could yield more force idle.
* use `test_idle_cores` before deciding whether `sched_cpu_cookie_match` or
  `sched_core_cookie_match` is appropriate. This is the most correct but
  perhaps overly complex.

The approach that I picked has the disadvantages that it might fail some
fast heuristics unnecessarily if there are no full idle core available
anyways but I thought this was acceptable since the `select_idle_cpu`
fallback at the end of the function seems reasonable.

Let me know what you think.

Best regards,
Fernand

Amazon Development Centre (South Africa) (Proprietary) Limited
29 Gogosoa Street, Observatory, Cape Town, Western Cape, 7925, South Africa
Registration Number: 2004 / 034463 / 07