netdev - Re: Soft-Lockup/Race in networking in 2.6.31-rc1+195 ( possibly?caused by netem)

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:	Tue, 7 Jul 2009 11:40:16 +0100
From:	Joao Correia <joaomiguelcorreia@...il.com>
To:	Jarek Poplawski <jarkao2@...il.com>
Cc:	Arun R Bharadwaj <arun@...ux.vnet.ibm.com>,
	Thomas Gleixner <tglx@...utronix.de>,
	Stephen Hemminger <shemminger@...tta.com>,
	netdev@...r.kernel.org, LKML <linux-kernel@...r.kernel.org>,
	Andres Freund <andres@...razel.de>
Subject: Re: Soft-Lockup/Race in networking in 2.6.31-rc1+195 ( 
	possibly?caused by netem)

I am now running 2.6.31-rc2 for a couple of hours, no freeze.

Let me know what/if i can help with tracking down the original source
of the problem.

Thank you very much for your time,
Joao Correia

On Tue, Jul 7, 2009 at 7:50 AM, Jarek Poplawski<jarkao2@...il.com> wrote:
> On Mon, Jul 06, 2009 at 07:26:43PM +0200, Andres Freund wrote:
>> On Monday 06 July 2009 19:23:18 Joao Correia wrote:
>> > Hello
>> >
>> > Since i already had the kernel compiled and ready to boot when i read
>> > this, i gave it a go anyway :-).
>> >
>> > I can reproduce the freeze with those 4 patches applied, so i can
>> > confirm that its, at least, related to, or exposed by, those patches.
>> > There must be something else too, or its just too much fuzziness, but
>> > the freeze takes a bit more time (approximately five minutes, give or
>> > take) compared to the instant freeze before, but its there with the
>> > patches, and without them, no freeze.
>> >
>> > I assume there isnt a "safe" way to get them out of current .31-rc's,
>> > right?
>> `echo 0 > /proc/sys/kernel/timer_migration` should mitigate the problem.
>
> I guess it should fix it entirely. Btw., here is a patch disabling the
> timers' part, so to make it hrtimers only. Could you try?
>
> Thanks,
> Jarek P.
> ---
>
> diff --git a/kernel/timer.c b/kernel/timer.c
> index 0b36b9e..011429c 100644
> --- a/kernel/timer.c
> +++ b/kernel/timer.c
> @@ -634,7 +634,7 @@ __mod_timer(struct timer_list *timer, unsigned long expires,
>
>        cpu = smp_processor_id();
>
> -#if defined(CONFIG_NO_HZ) && defined(CONFIG_SMP)
> +#if 0
>        if (!pinned && get_sysctl_timer_migration() && idle_cpu(cpu)) {
>                int preferred_cpu = get_nohz_load_balancer();
>
>
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html