linux-kernel - Re: [PATCH] PM / Domains: Skip latency measurements if timekeeping is suspended

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: <19573969.Pdd3px1nTO@vostro.rjw.lan>
Date:	Wed, 18 Mar 2015 23:30:08 +0100
From:	"Rafael J. Wysocki" <rjw@...ysocki.net>
To:	Geert Uytterhoeven <geert+renesas@...der.be>
Cc:	Kevin Hilman <khilman@...nel.org>,
	Ulf Hansson <ulf.hansson@...aro.org>, linux-pm@...r.kernel.org,
	linux-sh@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] PM / Domains: Skip latency measurements if timekeeping is suspended

On Wednesday, March 18, 2015 05:25:46 PM Geert Uytterhoeven wrote:
> The PM Domain code uses ktime_get() to perform various latency
> measurements.  However, if ktime_get() is called while timekeeping is
> suspended, the following warning is printed:
> 
>     WARNING: CPU: 0 PID: 1340 at kernel/time/timekeeping.c:576 ktime_get+0x30/0xf4()
> 
> This happens when resuming the PM Domain that contains the clock events
> source. Chain of operations is:
> 
>     timekeeping_resume()
>     {
>         clockevents_resume()
>             sh_cmt_clock_event_resume()
>                 pm_genpd_syscore_poweron()
>                     pm_genpd_sync_poweron()
>                         genpd_power_on()
>                             ktime_get(), but timekeeping_suspended == 1
>         ...
>         timekeeping_suspended = 0;
>     }
> 
> Skip all latency measurements if timekeeping is suspended to fix this.

I don't think that this is where we should fix it.  At least using
timekeeping_suspended outside of the timekeeping core would not be
welcome by its maintainers.

> Signed-off-by: Geert Uytterhoeven <geert+renesas@...der.be>
> ---
> I'm not sure if this is needed for all latency measurements.
> So far I only encountered it while powering-on a clock domain during
> resume from s2ram.

The problem seems to be that the clock domain is powered on in a
syscore resume routine which happens to be called before timekeeping_resume().
It looks like we either need to force the right ordering somehow or have a
special variant of GENPD_DEV_TIMED_CALLBACK() for syscore suspend/resume that
won't do the latency measurement at all (which doesn't make much sense at
this point, because time is effectively "frozen" then).

Rafael

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/