linux-kernel - Re: [PATCH] kernel/hung_task.c: Introduce sysctl to print all traces when a hung task is detected

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:   Mon, 16 Mar 2020 11:01:34 -0300
From:   "Guilherme G. Piccoli" <gpiccoli@...onical.com>
To:     Tetsuo Handa <penguin-kernel@...ove.sakura.ne.jp>,
        Kees Cook <keescook@...omium.org>
Cc:     linux-kernel@...r.kernel.org, linux-fsdevel@...r.kernel.org,
        linux-doc@...r.kernel.org, mcgrof@...nel.org, yzaikin@...gle.com,
        tglx@...utronix.de, kernel@...ccoli.net
Subject: Re: [PATCH] kernel/hung_task.c: Introduce sysctl to print all traces
 when a hung task is detected

On 14/03/2020 01:27, Tetsuo Handa wrote:
> On 2020/03/14 12:12, Kees Cook wrote:
>> On Fri, Mar 13, 2020 at 02:23:37PM -0300, Guilherme G. Piccoli wrote:
>>> Kees / Testsuo, are you OK with this patch once I resend with the
>>> suggestions you gave me?
>>
>> I think so, yes. Send a v2 (to akpm with us in CC).
>>
>>> Is there anybody else I should loop in the patch that should take a
>>> look? Never sent sysctl stuff before, sorry if I forgot somebody heheh
>>
>> akpm usually takes these kinds of things.
>>
> 
> Well, maybe sysctl_hung_task_all_cpu_backtrace = 1 by default is better for
> compatibility? Please CC or BCC kernel-testing people so that they can add
> hung_task_all_cpu_backtrace=1 kernel command line parameter to their testing
> environments if sysctl_hung_task_all_cpu_backtrace = 0 by default.
> 

Thanks a lot Kees and Tetsuo, I'll implement the suggestions and loop
the kernel-testing  and akpm in V2.

About being default, I personally think it's better / more reasonable to
have this parameter defaults to 0, to keep consistency with all other
*_all_cpu_backtrace parameters. Right now kernel shows the backtraces
only when hung_task_panic is set, so the idea of this patch is to
decouple the toggles and allow the user to decide about this policy.

Nevertheless, if people consider that would be good to have it as
enabled by default, I can change it in a potential V3.
Cheers,

Guilherme