linux-kernel - Re: [BUG] [BISECTED] System gets unresponsive since 2.6.35-rc1

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:	Wed, 15 Sep 2010 16:58:51 +0200
From:	Martin Kepplinger <martinkepplinger@....cc>
To:	Thomas Gleixner <tglx@...utronix.de>
CC:	johnstul@...ibm.com, damm@...nsource.se, akpm@...ux-foundation.org,
	linux-kernel@...r.kernel.org
Subject: Re: [BUG] [BISECTED] System gets unresponsive since 2.6.35-rc1

Am 2010-09-15 13:14, schrieb Thomas Gleixner:
> On Wed, 15 Sep 2010, Martin Kepplinger wrote:
> 
>> Hi,
>>
>> My Ubuntu 10.04, 32bit on a Lenovo G550 Laptop has the following
>> issue since .35-rc1: After a short uptime doing little things like
>> video playback, the system gets slow and stuck after some time. At
>> first, compiz-animations and everything (video playback,
>> terminal-commands) gets unusably slow. /var/log/messages comes up
>> with the following _during_ video playback. These warnings stop when
>> the video is stopped. The system stays (quite) unresponsive /
>> unusable.
>>
>> Sep 15 12:15:51 mobil pulseaudio[1456]: ratelimit.c: 19 events suppressed
>> Sep 15 12:15:57 mobil pulseaudio[1456]: ratelimit.c: 25 events suppressed
>> Sep 15 12:16:02 mobil pulseaudio[1456]: ratelimit.c: 6 events suppressed
>> Sep 15 12:16:07 mobil pulseaudio[1456]: ratelimit.c: 11 events suppressed
>> Sep 15 12:16:17 mobil pulseaudio[1456]: ratelimit.c: 12 events suppressed
>> Sep 15 12:16:23 mobil pulseaudio[1456]: ratelimit.c: 3 events suppressed
>>
>> Shutdown also hangs very soon. The problem is still present in
>> 2.6.36-rc3! I'm about to test -rc4. The last "good" one is 2.6.34.
>>
>> I bisected the problem. Unfortunately I could not revert the
>> patch. Additionally, I'm only 90% sure about the result because I
>> had to skip one step of the bisect (but since the toplevel-Makefile
>> showed "2.6.34-something", I could guess it to be "good")
>>
>> The rest of this E-Mail is the result (git show). I can only report
>> it. I'd happily test patches to it and hope this makes sense to
>> someone of you.
> 
> I doubt that this is the real culprit. We just decoded a HPET problem,
> which might be the real cause. Patch is below.
> 
> Can you provide the output of /proc/timer_list please ?
> 
> Thanks,
> 
> 	tglx
> ---
> commit 54ff7e595d763d894104d421b103a89f7becf47c
> Author: Thomas Gleixner <tglx@...utronix.de>
> Date:   Tue Sep 14 22:10:21 2010 +0200
> 
>     x86: hpet: Work around hardware stupidity
>     

I tested this against 2.6.35. Since I can write this after watching half an
hour of LinuxCon Videos, this fixes the problem. I hope it to get merged soon.
I'll test it against the current tree as well and will stay happy and quiet as
long as I can't find a problem.

Thanks a lot!

	Martin

View attachment "timer_list.dump" of type "text/plain" (9954 bytes)