From: Ingo Molnar <mingo@elte.hu>
To: Pekka Enberg <penberg@kernel.org>
Cc: Tejun Heo <tj@kernel.org>, Thomas Gleixner <tglx@linutronix.de>,
Linus Torvalds <torvalds@linux-foundation.org>,
Jens Axboe <axboe@kernel.dk>,
Andrew Morton <akpm@linux-foundation.org>,
werner <w.landgraf@ru.ru>, "H. Peter Anvin" <hpa@zytor.com>,
Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
Christoph Lameter <cl@linux.com>
Subject: Re: [block IO crash] Re: 2.6.39-rc5-git2 boot crashs
Date: Wed, 4 May 2011 14:57:13 +0200 [thread overview]
Message-ID: <20110504125713.GA9978@elte.hu> (raw)
In-Reply-To: <BANLkTikKZhqkcrG8aqUD0tJxZE5zVu0pMg@mail.gmail.com>
* Pekka Enberg <penberg@kernel.org> wrote:
> Hi Tejun,
>
> On Wed, May 4, 2011 at 2:27 PM, Tejun Heo <tj@kernel.org> wrote:
> > Hmmm... if it's a race caused by preemtion enabled where it shouldn't
> > be, it's most likely the wrong type of this_cpu_cmpxchg_double() being
> > used in SLUB? ie. __this_cpu_cmpxchg_double() where it should have
> > been this_cpu_cmpxchg_double()? Christoph?
>
> There's no __this_cpu_cmpxchg_double() usage in mm/slub.c so I don't
> think it's that simple.
Well, AFAICS the problem is:
earth4:~/tip> grep cmpxchg mm/slub.c
if (unlikely(!this_cpu_cmpxchg_double(
if (unlikely(!this_cpu_cmpxchg_double(
Where this macro resolves to:
# define this_cpu_cmpxchg_double_8(pcp1, pcp2, oval1, oval2, nval1, nval2) \
_this_cpu_generic_cmpxchg_double(pcp1, pcp2, oval1, oval2, nval1, nval2)
where:
#define _this_cpu_generic_cmpxchg_double(pcp1, pcp2, oval1, oval2, nval1, nval2) \
({ \
int ret__; \
preempt_disable(); \
ret__ = __this_cpu_generic_cmpxchg_double(pcp1, pcp2, \
oval1, oval2, nval1, nval2); \
preempt_enable();
where:
#define __this_cpu_generic_cmpxchg_double(pcp1, pcp2, oval1, oval2, nval1, nval2) \
({ \
int __ret = 0; \
if (__this_cpu_read(pcp1) == (oval1) && \
__this_cpu_read(pcp2) == (oval2)) { \
__this_cpu_write(pcp1, (nval1)); \
__this_cpu_write(pcp2, (nval2)); \
__ret = 1; \
} \
(__ret); \
})
With is both IRQ and SMP unsafe.
Thanks,
Ingo
next prev parent reply other threads:[~2011-05-04 12:57 UTC|newest]
Thread overview: 83+ messages / expand[flat|nested] mbox.gz Atom feed top
2011-05-02 22:28 2.6.39-rc5-git2 boot crashs werner
2011-05-02 23:24 ` Linus Torvalds
[not found] ` <web-516990066@zbackend1.aha.ru>
2011-05-03 15:22 ` Linus Torvalds
2011-05-03 19:08 ` Ingo Molnar
2011-05-03 20:17 ` Linus Torvalds
2011-05-03 20:20 ` H. Peter Anvin
2011-05-03 20:50 ` Ingo Molnar
2011-05-03 21:45 ` Linus Torvalds
2011-05-03 22:01 ` H. Peter Anvin
2011-05-04 7:19 ` Borislav Petkov
2011-05-04 7:38 ` Ingo Molnar
2011-05-04 7:55 ` Borislav Petkov
2011-05-04 8:35 ` [block IO crash] " Ingo Molnar
2011-05-04 9:52 ` Thomas Gleixner
2011-05-04 10:19 ` Ingo Molnar
2011-05-04 10:25 ` Ingo Molnar
2011-05-04 10:33 ` Ingo Molnar
2011-05-04 12:37 ` Ingo Molnar
2011-05-04 12:36 ` Ingo Molnar
2011-05-04 11:11 ` Thomas Gleixner
2011-05-04 11:16 ` Pekka Enberg
2011-05-04 11:27 ` Tejun Heo
2011-05-04 12:51 ` Pekka Enberg
2011-05-04 12:57 ` Ingo Molnar [this message]
2011-05-04 13:02 ` Thomas Gleixner
2011-05-04 13:00 ` Thomas Gleixner
2011-05-04 13:20 ` Tejun Heo
2011-05-04 14:10 ` Thomas Gleixner
2011-05-04 14:14 ` Ingo Molnar
2011-05-04 14:36 ` [PATCH] slub: Fix the lockless code on 32-bit platforms with no 64-bit cmpxchg Ingo Molnar
2011-05-04 14:42 ` Christoph Lameter
2011-05-04 16:30 ` Ingo Molnar
2011-05-04 21:52 ` Ben Greear
2011-05-04 22:00 ` Linus Torvalds
2011-05-04 22:22 ` Ben Greear
2011-05-04 14:19 ` [block IO crash] Re: 2.6.39-rc5-git2 boot crashs Christoph Lameter
2011-05-04 14:25 ` Tejun Heo
2011-05-04 14:35 ` Christoph Lameter
2011-05-04 15:20 ` Ingo Molnar
2011-05-04 14:46 ` Thomas Gleixner
2011-05-04 15:00 ` Christoph Lameter
2011-05-04 15:13 ` Linus Torvalds
2011-05-04 15:28 ` Christoph Lameter
2011-05-04 15:37 ` Pekka Enberg
2011-05-04 15:53 ` Linus Torvalds
2011-05-04 18:20 ` Linus Torvalds
2011-05-04 18:49 ` Christoph Lameter
2011-05-04 19:07 ` Linus Torvalds
2011-05-04 19:30 ` Christoph Lameter
2011-05-04 19:38 ` Linus Torvalds
2011-05-04 20:04 ` Christoph Lameter
2011-05-04 20:21 ` Valdis.Kletnieks
2011-05-04 20:32 ` Christoph Lameter
2011-05-04 20:49 ` Ingo Molnar
2011-05-04 21:06 ` Linus Torvalds
2011-05-04 21:19 ` Linus Torvalds
2011-05-04 21:40 ` Thomas Gleixner
2011-05-05 9:54 ` Tejun Heo
2011-05-05 10:18 ` Ingo Molnar
2011-05-05 10:45 ` Thomas Gleixner
2011-05-05 18:20 ` Christoph Lameter
2011-05-05 19:13 ` Ingo Molnar
2011-05-05 19:53 ` werner
2011-05-05 20:09 ` Christoph Lameter
2011-05-05 21:12 ` werner
2011-05-05 22:27 ` Thomas Gleixner
[not found] ` <web-518008166@zbackend1.aha.ru>
[not found] ` <web-518059420@zbackend1.aha.ru>
[not found] ` <20110505060204.GA28015@elte.hu>
2011-05-05 6:46 ` werner
2011-05-04 15:37 ` [block IO crash] " Linus Torvalds
2011-05-04 16:08 ` Christoph Lameter
2011-05-04 16:50 ` Ingo Molnar
2011-05-04 17:12 ` Thomas Gleixner
2011-05-04 15:41 ` Pekka Enberg
2011-05-04 13:22 ` Ingo Molnar
2011-05-04 14:21 ` Christoph Lameter
2011-05-04 14:04 ` Christoph Lameter
2011-05-04 14:07 ` Tejun Heo
2011-05-04 14:21 ` Thomas Gleixner
2011-05-04 10:13 ` Ingo Molnar
2011-05-04 10:41 ` Ingo Molnar
2011-05-04 10:45 ` Ingo Molnar
2011-05-04 11:06 ` Ingo Molnar
2011-05-04 12:37 ` Ingo Molnar
2011-05-04 12:47 ` Ingo Molnar
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20110504125713.GA9978@elte.hu \
--to=mingo@elte.hu \
--cc=akpm@linux-foundation.org \
--cc=axboe@kernel.dk \
--cc=cl@linux.com \
--cc=hpa@zytor.com \
--cc=linux-kernel@vger.kernel.org \
--cc=penberg@kernel.org \
--cc=tglx@linutronix.de \
--cc=tj@kernel.org \
--cc=torvalds@linux-foundation.org \
--cc=w.landgraf@ru.ru \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox