linux-kernel - Re: [PATCH] Kick CPUS that might be sleeping in cpus_idle

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]

Message-ID: <Pine.LNX.4.58.0801100936320.6420@gandalf.stny.rr.com>
Date:	Thu, 10 Jan 2008 09:43:46 -0500 (EST)
From:	Steven Rostedt <rostedt@...dmis.org>
To:	Ingo Molnar <mingo@...e.hu>
cc:	LKML <linux-kernel@...r.kernel.org>,
	Linus Torvalds <torvalds@...ux-foundation.org>,
	Andrew Morton <akpm@...ux-foundation.org>,
	Thomas Gleixner <tglx@...utronix.de>,
	"Brown, Len" <len.brown@...el.com>,
	Venkatesh Pallipadi <venkatesh.pallipadi@...el.com>,
	Adam Belay <abelay@...ell.com>,
	Peter Zijlstra <a.p.zijlstra@...llo.nl>,
	Andi Kleen <ak@...e.de>
Subject: Re: [PATCH] Kick CPUS that might be sleeping in cpus_idle_wait

On Thu, 10 Jan 2008, Ingo Molnar wrote:
> > FYI, I just hit this hang on 2.6.24-rc6 without any extra patches. So,
> > unless 2.6.24-rc7 did anything to fix this issue, this is a high
> > priority bug (IMHO).
>
> i'm wondering why this only triggered now. Is this something new in
> 2.6.24?

It only triggeres with the switching of the idle governors. And not just
one, you need to switch twice. The first loading of a governor does not
call cpu_idle_wait, but the second one does. NO_HZ must also be enabled,
plus this needs to happen when no events or threads are scheduled to run
on a CPU, which limits this to boot up.

Also, this only seems to happen on my 2x2 (4way) and only once in a while.

I'm surprised that I'm the only one so far to report it. I can boot up the
2.6.23 kernel on this box to see if it also hangs sometimes. But, as I
said, it may take several hundreds of tries to see it.

-- Steve

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/