linux-kernel - Re: perf/ftrace lockup on 3.12-rc6 with trigger code

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 for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:	Fri, 25 Oct 2013 09:17:12 -0400 (EDT)
From:	Vince Weaver <vincent.weaver@...ne.edu>
To:	Ingo Molnar <mingo@...nel.org>
cc:	Steven Rostedt <rostedt@...dmis.org>, linux-kernel@...r.kernel.org,
	Peter Zijlstra <peterz@...radead.org>,
	Ingo Molnar <mingo@...hat.com>,
	Arnaldo Carvalho de Melo <acme@...stprotocols.net>,
	Dave Jones <davej@...hat.com>,
	Frederic Weisbecker <fweisbec@...il.com>
Subject: Re: perf/ftrace lockup on 3.12-rc6 with trigger code

On Fri, 25 Oct 2013, Ingo Molnar wrote:

> 
> Btw., even if various config options are enabled, we don't want the 
> kernel to lock up. So it might not be _the_ bug you are looking, but 
> it certainly looks like _a_ bug.

yes, I agree, but it means I was chasing a different bug than I thought I 
was.

I can still get the same (or similar) bug with the advanced ftrace options 
turned off, it's just it takes a lot longer and is harder to trigger.
I have a 30,000 syscall trace that triggers it but only maybe 1 time in 5 
so it's harder to narrow down the cause.

> A full .config and a SHA1 of the tested tree would definitely be 
> nice!

I was testing e6036c0b88962df82a8853971b86a55f09faef40
but I can crash machines from 3.10 through current with that chunk of code 
I posted.  I am sure it affects before 3.10, that's just when
I first enabled the extra ftrace options.

Attached is the full config.  This is running on a core2 system.  I've 
seen similar bugs on other systems but not sure if the trigger is the same 
or not.

Vince

View attachment "CONFIG-3.12-rc6.bad" of type "TEXT/PLAIN" (98081 bytes)