linux-kernel - Re: [PATCH v2 2/5] tick/broadcast: Split __tick_broadcast_oneshot

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, 27 May 2021 12:55:30 +0100
From:   Will Deacon <will@...nel.org>
To:     Xin Hao <xhao@...ux.alibaba.com>
Cc:     fweisbec@...il.com, john.stultz@...aro.org,
        kernel-team@...roid.com, linux-arm-kernel@...ts.infradead.org,
        linux-kernel@...r.kernel.org, lorenzo@...gle.com, maz@...nel.org,
        mika.penttila@...tfour.com, sboyd@...nel.org, tglx@...utronix.de
Subject: Re: [PATCH v2 2/5] tick/broadcast: Split
 __tick_broadcast_oneshot_control() into a helper

On Thu, May 27, 2021 at 07:35:03PM +0800, Xin Hao wrote:
> 
> 在 2021/5/27 下午4:22, Will Deacon 写道:
> > On Thu, May 27, 2021 at 03:23:06PM +0800, Xin Hao wrote:
> > >       I  had backport you  tick/broadcast: Prefer per-cpu relatives patches,
> > > 
> > > but i did not get the true result,  the Wakeup Devices are all null, why?
> > Probably because you don't have any suitable per-cpu timers to act as a
> > wakeup. Do you have a per-cpu timer registered with CLOCK_EVT_FEAT_PERCPU
> 
> Yes, you are right, but i want to know why the timer do not support 
> CLOCK_EVT_FEAT_PERCPU.

I suspect there may be drivers with the feature missing simply because it
wasn't really used for much until now (I think it just prevented use for
broadcast). However, before adding that to a timer, you do need to make
sure that it really is banked per-cpu (or at least handles racing accesses)
as well as having the per-cpu irq.

Will