linux-kernel - Re: 2.6.19-rc2-mm2 - process_session-helper breaks /sbin/killall5

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:	Fri, 27 Oct 2006 23:18:39 -0400
From:	Valdis.Kletnieks@...edu
To:	Andrew Morton <akpm@...l.org>, Cedric Le Goater <clg@...ibm.com>
Cc:	linux-kernel@...r.kernel.org
Subject: Re: 2.6.19-rc2-mm2 - process_session-helper breaks /sbin/killall5

On Fri, 20 Oct 2006 01:56:41 PDT, Andrew Morton said:
> 
> ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.19-rc2/2.6.19-rc2-mm2/

> +add-process_session-helper-routine.patch
> +add-process_session-helper-routine-deprecate-old-field.patch
> +add-process_session-helper-routine-deprecate-old-field-tidy.patch
> +add-process_session-helper-routine-deprecate-old-field-fix-warnings.patch
> +add-process_session-helper-routine-deprecate-old-field-fix-warnings-2.patch
> +rename-struct-namespace-to-struct-mnt_namespace.patch
> +add-an-identifier-to-nsproxy.patch
> +rename-struct-pspace-to-struct-pid_namespace.patch
> +add-pid_namespace-to-nsproxy.patch
> +use-current-nsproxy-pid_ns.patch
> +add-child-reaper-to-pid_namespace.patch
> 
>  Start work on virtualising process sessions.

Sorry for the delay in spotting this one sooner (and I'm surprised apparently
nobody else has).  Was busy, didn't get a chance to bisect it at all till
today, and I only got as far as "one of the above patches".

System works fine, except when you go to shutdown.  When it hits the /sbin/killall5
call in /etc/init.d/halt, it kills *all* the processes, and we get a
nice message 'INIT: no more processes in current runlevel', and we're dead in
the water.  Checking with alt-sysrg-T shows that in fact, the only things
left running are the various kernel threads.  As near as I can tell, killall5
wasn't able to tell that its parent process was part of its process group,
so didn't refrain from killing it.

Any ideas/clues?

Content of type "application/pgp-signature" skipped