linux-kernel - Re: [PATCH v6 0/4] Add support for AArch64 AMUv1-based arch_freq_get_on

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: <Zo_lN8jdgADwqvor@arm.com>
Date: Thu, 11 Jul 2024 15:59:19 +0200
From: Beata Michalska <beata.michalska@....com>
To: Catalin Marinas <catalin.marinas@....com>
Cc: linux-kernel@...r.kernel.org, linux-arm-kernel@...ts.infradead.org,
	ionela.voinescu@....com, sudeep.holla@....com, will@...nel.org,
	vincent.guittot@...aro.org, vanshikonda@...amperecomputing.com,
	sumitg@...dia.com, yang@...amperecomputing.com,
	lihuisong@...wei.com, viresh.kumar@...aro.org, rafael@...nel.org
Subject: Re: [PATCH v6 0/4] Add support for AArch64 AMUv1-based
 arch_freq_get_on_cpu

Hi Catalin,

On Mon, Jul 08, 2024 at 06:10:02PM +0100, Catalin Marinas wrote:
> Hi Beata,
> 
> On Mon, Jun 03, 2024 at 09:21:50AM +0100, Beata Michalska wrote:
> > Introducing arm64 specific version of arch_freq_get_on_cpu, cashing on
> > existing implementation for FIE and AMUv1 support: the frequency scale
> > factor, updated on each sched tick, serves as a base for retrieving
> > the frequency for a given CPU, representing an average frequency
> > reported between the ticks - thus its accuracy is limited.
> > 
> > The changes have been rather lightly (due to some limitations) tested on
> > an FVP model. Note that some small discrepancies have been observed while
> > testing (on the model) and this is currently being investigated, though it
> > should not have any significant impact on the overall results.
> 
> What's the plan with this series? Are you still investigating those
> discrepancies or is it good to go?
>
Overall it should be good to go with small caveat:
as per discussion [1] we might need to provide new sysfs attribute exposing an
average frequency instead of plugging new code under existing cpuinfo_cur_freq.
This is to avoid messing up with other archs and make a clean distinction on
which attribute provides what information. 
As such, the arch_freq_get_on_cpu implementation provided within this series
[PATCH v6 3/4] will most probably be shifted to a new function.

Hopefully will be able to send those changes soon.

---
[1] https://lore.kernel.org/all/ZmrB_DqtmVpvG30l@arm.com/
---
BR
Beata

> -- 
> Catalin