linux-kernel - Re: large intermittent latency spike 2.6.28 (and 2.6.27.10), bisect commit ca7e716c7833aeaeb8fedd6d004c5f5d5e14d325 -> Revert "sched_clock: prevent scd->clock from moving backwards"

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]

Date:	Wed, 31 Dec 2008 23:38:01 +0100
From:	"Fabio Comolli" <fabio.comolli@...il.com>
To:	"Jayson King" <dev@...sonking.com>
Cc:	arjan@...ux.intel.com, linux-kernel@...r.kernel.org, mingo@...e.hu,
	torvalds@...ux-foundation.org, a.p.zijlstra@...llo.nl,
	shaggy@...ux.vnet.ibm.com, rjw@...k.pl
Subject: Re: large intermittent latency spike 2.6.28 (and 2.6.27.10), bisect commit ca7e716c7833aeaeb8fedd6d004c5f5d5e14d325 -> Revert "sched_clock: prevent scd->clock from moving backwards"

The revert you did was rejected in the 2.6.28 timeframe becaused it
caused a regression on my laptop. The patch Ingo suggested (originally
from Thomas Gleixner) addresses both your issue and mine.

Regards,
Fabio

On Wed, Dec 31, 2008 at 11:27 PM, Jayson King <dev@...sonking.com> wrote:
> Arjan van de Ven wrote:
>> Jayson King <dev@...sonking.com> wrote:
>>> Please disregard this. ...
>>
>> can you run latencytop to see if it can pinpoint a cause of latency?
>
> Like I said in my other reply I was running the wrong kernel when I wrote
> that (oops)... So now I am still sure it is ca7e716.
>
> Ingo's patch makes the same change that reverting ca7e716 does here:
>
> -       max_clock = scd->tick_gtod + TICK_NSEC;
> +       max_clock = wrap_max(scd->clock, scd->tick_gtod + TICK_NSEC);
>
>
> And with that change (I'm using Ingo's patch now) I can't trigger the hang
> anymore. And before, I could reliably trigger it in a short time.
>
> Jayson
>
>
--
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/