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:   Thu, 2 Jul 2020 13:15:36 +0100
From:   Quentin Perret <qperret@...gle.com>
To:     Valentin Schneider <valentin.schneider@....com>
Cc:     linux-kernel@...r.kernel.org, mingo@...nel.org,
        peterz@...radead.org, vincent.guittot@...aro.org,
        dietmar.eggemann@....com, morten.rasmussen@....com
Subject: Re: [PATCH v3 2/7] sched/topology: Define and assign sched_domain
 flag metadata

Hey Valentin,

On Wednesday 01 Jul 2020 at 20:06:50 (+0100), Valentin Schneider wrote:
> +/*
> + * Domain members have different CPU capacities
> + *
> + * SHARED_PARENT: Set from the topmost domain down to the first domain where
> + * asymmetry is detected.
> + */
> +SD_FLAG(SD_ASYM_CPUCAPACITY,    5, SDF_SHARED_PARENT)

Probably not a huge deal, but I don't think this is _exactly_ how
SD_ASYM_CPUCAPACITY was defined originally, nor how the topology
detection code deals with it (IIRC).

That is, IIRC Morten defined SD_ASYM_CPUCAPACITY as the _lowest_ domain
at which all CPU capacities are visible. On all real systems I can think
of, this domain also happens to be the topmost domain, so that might not
be a huge deal and we can probably change that definition to the one you
suggest. But we should perhaps make the matching changes to
asym_cpu_capacity_level()/build_sched_domains() if we're going down that
path?

Thanks,
Quentin

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ