linux-kernel - Re: Fwd: Watchdog detected hard lockup.

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] [day] [month] [year] [list]

Date:   Thu, 25 May 2017 11:56:45 +0200
From:   Petr Mladek <pmladek@...e.com>
To:     Sergey Senozhatsky <sergey.senozhatsky.work@...il.com>
Cc:     Sumit Gemini <sumit.gemini0@...il.com>,
        Sergey Senozhatsky <sergey.senozhatsky@...il.com>,
        Peter Zijlstra <peterz@...radead.org>,
        Steven Rostedt <rostedt@...dmis.org>,
        linux-kernel@...r.kernel.org
Subject: Re: Fwd: Watchdog detected hard lockup.

On Thu 2017-05-25 14:50:14, Sergey Senozhatsky wrote:
> On (05/25/17 11:14), Sumit Gemini wrote:
> >    Thanks Sergey and i forgot to add other guys in discussion.
> >     
> > 
> >    [1]lkml.kernel.org/r/20170509082859.854-1-sergey.senozhatsky@...il.com
> > 
> >    Now I'm going for backporting the kernel with RFC, as you suggested me
> >    yesterday.
> 
> or you can try to reduce the amount of data you printk from IRQ.

In fact, you should reduce the amount of data you printk from all
contexts. There might be more sources that produce too many messages.
The system evidently has troubles to keep the speed and get them
on the console.

You might also try to reduce the console_level and show only the more
important messages on the console.

> need to hex dump sbks all the time?

I wanted to suggest using ratelimit but it seems that the hex dump
is already ratelimited, see net_ratelimit() in
ip_handle_martian_source().

This might suggest that are more sources of printk messages on the
system. Or that you modified the ratelimit setting to print too many
messages.

> switching to a deferred printk won't
> do the trick, probably, because you gonna console_unlock() from IRQ in
> the worst case anyway.

Yup.

Best Regards,
Petr