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:	Mon, 6 Jul 2009 12:28:58 +0100
From:	Joao Correia <joaomiguelcorreia@...il.com>
To:	Jarek Poplawski <jarkao2@...il.com>
Cc:	Andres Freund <andres@...razel.de>,
	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>
Subject: Re: Soft-Lockup/Race in networking in 2.6.31-rc1+195 ( 
	possibly?caused by netem)

Hello

.config is attached. This is a real box, a fedora 11 with rawhide
updates applied. Its a quad core system (amd phenom 9600) with 4gb
ddr2 ram. The board is an asus m3a-h/hdmi. We have a couple of these
boxes just to test run linux before dropping into production, hence
the availability for tests.

Haven't tried without lockdep, will try later and let you know if any
difference.

Thank you for your time,
Joao Correia
Centro de Informatica
Universidade da Beira Interior
Portugal

On Mon, Jul 6, 2009 at 9:14 AM, Jarek Poplawski<jarkao2@...il.com> wrote:
> On Mon, Jul 06, 2009 at 05:53:51AM +0100, Joao Correia wrote:
>> Hello
> Hi,
>
>> System freezes immediatly after grub, no init processing at all, after
>> applying those patches on top of vanilla 2.6.30 on my box.
>>
>> Also, 2nd patch, this piece:
>>
>> +#ifdef CONFIG_SCHED_DEBUG
>> +static inline unsigned int get_sysctl_timer_migration(void)
>> +{
>> +       return sysctl_timer_migration;
>> +}
>> +#else
>> +static inline unsigned int get_sysctl_timer_migration(void)
>> +{
>> +       return 1;
>> +}
>> +#endif
>>
>> doesnt work on top of 2.6.30. It complains, while compiling, that
>> sysctl_timer_migration is not defined. So i just replaced that call
>> with return 1, like on the not debug case. Hope this doesnt defeat
>> your test case, but it wouldnt compile otherwise. Probably that was
>> just introduced after 2.6.30?
>
> Yes, it would be too easy if it had worked just like this... (I only
> test-compiled it, without such warnings, but without CONFIG_SCHED_DEBUG.)
> I'll look at this again.
>
>>
>> (Have also tried latest .31-rc2, original bug still exists.)
>
> Btw., could you send your .config? Do you have lockdep enabled (if so
> did you try without it)? Is it 'a real' box or something virtual?
>
> Thanks,
> Jarek P.
>

Download attachment "config" of type "application/octet-stream" (100278 bytes)