netdev - Re: oom-kill && frozen()

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:	Wed, 13 Nov 2013 19:15:43 +0100
From:	Oleg Nesterov <oleg@...hat.com>
To:	Peter Zijlstra <peterz@...radead.org>
Cc:	Tejun Heo <tj@...nel.org>, David Rientjes <rientjes@...gle.com>,
	David Laight <David.Laight@...LAB.COM>,
	Geert Uytterhoeven <geert@...ux-m68k.org>,
	Ingo Molnar <mingo@...nel.org>, netdev@...r.kernel.org,
	linux-kernel@...r.kernel.org
Subject: Re: oom-kill && frozen()

On 11/13, Peter Zijlstra wrote:
>
> On Wed, Nov 13, 2013 at 06:07:24PM +0100, Oleg Nesterov wrote:
> > 	4. Finally, change try_to_wake_up() path to do
> >
> > 		-	p->state = TASK_WAKING;
> > 		+	p->state &= ~state;
> > 		+	if (p->state & ~(TASK_DEAD | TASK_WAKEKILL | TASK_PARKED))
> > 		+		return;
> > 		+	else
> > 		+		p->state = TASK_WAKING;
> >
> > 	   IOW, if the task sleeps in, say, TASK_INTERRUPTIBLE | __TASK_FROZEN
> > 	   then it need both try_to_wake_up(TASK_INTERRUPTIBLE) and
> > 	   try_to_wake_up(__TASK_FROZEN) to wake up.
>
> > Tejun, Peter, do you think this makes any sense? I am just curious, but
> > "selective wakeup" looks potentially useful.
>
> I've never looked at any of this freeze stuff, so I cannot comment too
> much. However we should be very careful not to add too much to relative
> hot paths for the relative rare case of freezing stuff.

Yes, yes, sure.

Plus my description was confusing and incomplete (and I am sure I missed
something more). I forgot to mention that freeze_task() should also add
the FROZEN state if it sees FROZEN qualifier, this needs more changes.

So please forget. May be I'll _try_ to make something working (at least
for discussion), but I am not sure.

Oleg.

--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html