linux-kernel - Re: WARNING in set_restore

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, 29 Jan 2016 09:32:35 -0500
From:	Steven Rostedt <rostedt@...dmis.org>
To:	Thomas Gleixner <tglx@...utronix.de>
Cc:	Dmitry Vyukov <dvyukov@...gle.com>,
	Oleg Nesterov <oleg@...hat.com>,
	Andrew Morton <akpm@...ux-foundation.org>,
	Richard Weinberger <richard@....at>,
	Amanieu d'Antras <amanieu@...il.com>,
	Chris Metcalf <cmetcalf@...hip.com>,
	Andy Lutomirski <luto@...capital.net>,
	Davidlohr Bueso <dave@...olabs.net>,
	Vladimir Davydov <vdavydov@...allels.com>,
	Palmer Dabbelt <palmer@...belt.com>,
	LKML <linux-kernel@...r.kernel.org>,
	Ingo Molnar <mingo@...hat.com>,
	"H. Peter Anvin" <hpa@...or.com>
Subject: Re: WARNING in set_restore_sigmask

On Fri, 29 Jan 2016 15:10:18 +0100 (CET)
Thomas Gleixner <tglx@...utronix.de> wrote:


> The time depends on the serial speed and the trace buffer size and the number
> of cpus. Don't worry about the uptime. The buffer is a ringbuffer with limited
> capacity, so the dump time is constant.
> 
> If you want that /proc/sys/kernel/ftrace_dump_on_oops spills out on warnings
> as well, you need to enable
> 
> /proc/sys/kernel/traceoff_on_warning
> /proc/sys/kernel/ftrace_dump_on_oops
> 
> and 
> 
> /proc/sys/kernel/panic_on_warn
> 
> because we do not dump on warnings.

To save time in the wait, you may also want to do:

# echo 50 > /sys/kernel/debug/tracing/buffer_size_kb

which will make each per cpu buffer 50 kb, instead of the default of a
meg each. That will cut down the time to spit out the ftrace buffer to
serial console, but it wont have as much data. 50 kb should be enough
to trace back to the problem point though.

-- Steve