From: Christoph Lameter <clameter@sgi.com>
To: Andrew Morton <akpm@linux-foundation.org>
Cc: Michal Piotrowski <michal.k.k.piotrowski@gmail.com>,
linux-kernel@vger.kernel.org, "Rafael J. Wysocki" <rjw@sisk.pl>,
Satoru Takeuchi <takeuchi_satoru@jp.fujitsu.com>
Subject: Re: 2.6.22-rc3-mm1
Date: Thu, 31 May 2007 11:41:22 -0700 (PDT) [thread overview]
Message-ID: <Pine.LNX.4.64.0705311136200.11008@schroedinger.engr.sgi.com> (raw)
In-Reply-To: <20070531113155.1c6d7185.akpm@linux-foundation.org>
On Thu, 31 May 2007, Andrew Morton wrote:
> > l *0xc0181288
> > 0xc0181288 is in add_partial (/home/devel/linux-mm/mm/slub.c:1193).
> > 1188 }
> > 1189
> > 1190 static void add_partial(struct kmem_cache_node *n, struct page *page)
> > 1191 {
> > 1192 spin_lock(&n->list_lock);
> > 1193 n->nr_partial++;
> > 1194 list_add(&page->lru, &n->partial);
> > 1195 spin_unlock(&n->list_lock);
> > 1196 }
> > 1197
add_partial runs with interrupts disabled. The interrupts are disabled
when we enter SLUB via allocator or free functions.
> > [ 4972.087650] {in-hardirq-W} state was registered at:
> > [ 4972.092656] [<c0143bfe>] mark_lock+0x82/0x557
> > [ 4972.097323] [<c0144c5f>] __lock_acquire+0x476/0xd36
> > [ 4972.102562] [<c01455bd>] lock_acquire+0x9e/0xb8
> > [ 4972.107342] [<c0348993>] _spin_lock+0x38/0x62
> > [ 4972.111993] [<c0181d29>] deactivate_slab+0xb9/0x179
> > [ 4972.117300] [<c0181e56>] flush_slab+0x6d/0x72
> > [ 4972.122063] [<c0181e8c>] __flush_cpu_slab+0x31/0x36
> > [ 4972.127335] [<c0181ea5>] flush_cpu_slab+0x14/0x17
> > [ 4972.132401] [<c0113d6f>] smp_call_function_interrupt+0x3a/0x56
> > [ 4972.138607] [<c0104c73>] call_function_interrupt+0x33/0x38
> > [ 4972.144503] [<c0102b59>] default_idle+0x50/0x69
> > [ 4972.149421] [<c01023eb>] cpu_idle+0xb3/0xf8
> > [ 4972.153889] [<c03454fa>] rest_init+0x56/0x58
> > [ 4972.158402] [<c04f39c7>] start_kernel+0x351/0x359
> > [ 4972.163450] [<ffffffff>] 0xffffffff
> > [ 4972.167221] irq event stamp: 2451
>
> Yep, that's a bug in slub. We take that lock in the IPI handler. If a CPU
> is currently holding that lock and then takes the IPI and enters
> add_partial(), it'll deadlock.
A cpu cannot enter an IPI handler while interrupts are disabled. That
needs to be always the case when the list_lock is held.
> Perhaps a suitable fix would be local_irq_disable() in flush_slab().
As far as I can tell: Interrupts are always disabled when flush_slab is
run.
Sometimes we use spin_lock_irqsave for the list_lock and at other times
spin_lock if interrupts are already disabled. Is that the problem?
next prev parent reply other threads:[~2007-05-31 18:41 UTC|newest]
Thread overview: 108+ messages / expand[flat|nested] mbox.gz Atom feed top
2007-05-31 6:58 2.6.22-rc3-mm1 Andrew Morton
2007-05-31 12:09 ` 2.6.22-rc3-mm1 Cornelia Huck
2007-05-31 12:15 ` 2.6.22-rc3-mm1 Matthew Wilcox
2007-05-31 12:20 ` 2.6.22-rc3-mm1 Cornelia Huck
2007-05-31 12:35 ` 2.6.22-rc3-mm1 Jeff Garzik
2007-05-31 15:11 ` 2.6.22-rc3-mm1 Cornelia Huck
2007-05-31 15:13 ` 2.6.22-rc3-mm1 Christoph Hellwig
2007-05-31 22:10 ` 2.6.22-rc3-mm1 Andrew Morton
2007-06-01 7:09 ` 2.6.22-rc3-mm1 Cornelia Huck
2007-05-31 15:29 ` 2.6.22-rc3-mm1 Michal Piotrowski
2007-05-31 19:58 ` 2.6.22-rc3-mm1 Rafael J. Wysocki
2007-05-31 21:30 ` [PATCH -mm] Freezer: migration_thread should not be freezable (was: Re: [linux-pm] Re: 2.6.22-rc3-mm1) Rafael J. Wysocki
[not found] ` <6bffcb0e0705311425q2e00cc69j5006d95c3f0df158@mail.gmail.com>
2007-05-31 21:34 ` 2.6.22-rc3-mm1 Rafael J. Wysocki
2007-05-31 17:53 ` 2.6.22-rc3-mm1 Michal Piotrowski
2007-05-31 18:08 ` 2.6.22-rc3-mm1 Michal Piotrowski
2007-05-31 18:31 ` 2.6.22-rc3-mm1 Andrew Morton
2007-05-31 18:41 ` Christoph Lameter [this message]
2007-05-31 18:53 ` 2.6.22-rc3-mm1 Andrew Morton
2007-05-31 18:57 ` 2.6.22-rc3-mm1 Christoph Lameter
2007-05-31 20:43 ` 2.6.22-rc3-mm1 Mariusz Kozlowski
2007-05-31 21:19 ` 2.6.22-rc3-mm1 Andrew Morton
2007-06-01 20:50 ` 2.6.22-rc3-mm1 Mariusz Kozlowski
2007-06-01 21:02 ` 2.6.22-rc3-mm1 Andrew Morton
2007-06-01 21:21 ` 2.6.22-rc3-mm1 Mariusz Kozlowski
2007-06-01 23:30 ` 2.6.22-rc3-mm1 Benjamin Herrenschmidt
2007-06-02 8:40 ` 2.6.22-rc3-mm1 Segher Boessenkool
2007-05-31 22:05 ` 2.6.22-rc3-mm1 - works-for-me and a meta-question Valdis.Kletnieks
2007-05-31 22:16 ` Andrew Morton
[not found] ` <6bffcb0e0705311527ge774f48s8014f5108556f651@mail.gmail.com>
2007-05-31 23:21 ` andrea
2007-05-31 23:13 ` 2.6.22-rc3-mm1 - page_mkwrite() breakage Mark Fasheh
2007-06-01 1:01 ` Nick Piggin
2007-06-01 1:24 ` Mark Fasheh
2007-06-01 1:34 ` Nick Piggin
2007-06-01 1:45 ` Mark Fasheh
2007-06-01 1:53 ` Nick Piggin
2007-06-01 5:20 ` Mark Fasheh
2007-06-01 22:01 ` Mark Fasheh
2007-06-01 22:25 ` Andrew Morton
2007-06-01 22:33 ` Mark Fasheh
2007-06-01 22:47 ` Andrew Morton
2007-06-01 22:53 ` Mark Fasheh
2007-06-01 2:01 ` 2.6.22-rc3-mm1 Arnaldo Carvalho de Melo
2007-06-01 2:12 ` 2.6.22-rc3-mm1 Andrew Morton
2007-06-01 2:24 ` 2.6.22-rc3-mm1 Arnaldo Carvalho de Melo
2007-06-01 3:52 ` 2.6.22-rc3-mm1 Michael Ellerman
2007-06-01 5:55 ` 2.6.22-rc3-mm1 Eric W. Biederman
2007-06-01 16:42 ` 2.6.22-rc3-mm1: IA64 make allnoconfig broken Mel Gorman
2007-06-01 17:00 ` Andrew Morton
2007-06-01 18:50 ` Mel Gorman
2007-06-01 20:55 ` [PATCH] Fix problem with IA64 make allnoconfig Mel Gorman
2007-06-02 13:57 ` 2.6.22-rc3-mm1: Xen compile error with X86_CMPXCHG=n Adrian Bunk
2007-06-28 23:36 ` Adrian Bunk
2007-06-29 3:21 ` Jeremy Fitzhardinge
2007-06-02 17:06 ` [-mm patch] fix XEN_BLKDEV_FRONTEND Makefile entry Adrian Bunk
2007-06-02 17:14 ` 2.6.22-rc3-mm1: __attribute__((weak)) considered harmful Adrian Bunk
2007-06-04 21:22 ` Andrew Morton
2007-06-04 23:52 ` Martin Peschke
2007-06-05 3:59 ` Russell King
2007-06-02 19:09 ` [-mm patch] the ASYNC_* options shouldn't be user visible Adrian Bunk
2007-06-04 16:19 ` Williams, Dan J
2007-06-03 20:54 ` [-mm patch] make drivers/firmware/dmi-id.c:dmi_id_init() static Adrian Bunk
2007-06-08 4:38 ` Greg KH
2007-06-03 20:54 ` [-mm patch] drivers/i2c/chips/ds1682.c: make code static Adrian Bunk
2007-06-04 8:15 ` Jean Delvare
2007-06-03 20:54 ` [-mm patch] make xpad_play_effect() static Adrian Bunk
2007-06-03 20:54 ` [-mm patch] fix the tea5761 tuner support Adrian Bunk
2007-06-04 18:00 ` 2.6.22-rc3-mm1 - pppd hanging in netdev_run_todo while holding mutex Valdis.Kletnieks
2007-06-06 6:14 ` Andrew Morton
2007-06-04 22:12 ` [-mm patch] e1000: #if 0 two functions Adrian Bunk
2007-06-04 22:13 ` [-mm patch] drivers/mmc/core/core.{h,c}: cleanups Adrian Bunk
2007-06-06 18:36 ` Pierre Ossman
2007-06-05 9:11 ` 2.6.22-rc3-mm1 Ingo Molnar
2007-06-05 9:18 ` 2.6.22-rc3-mm1 Ingo Molnar
2007-06-05 9:24 ` 2.6.22-rc3-mm1 Andrew Morton
2007-06-05 9:33 ` 2.6.22-rc3-mm1 Ingo Molnar
2007-06-05 9:39 ` 2.6.22-rc3-mm1 Ingo Molnar
2007-06-05 9:42 ` 2.6.22-rc3-mm1 Ingo Molnar
2007-06-05 9:45 ` 2.6.22-rc3-mm1 Ingo Molnar
2007-06-05 9:50 ` 2.6.22-rc3-mm1 Ingo Molnar
2007-06-05 9:56 ` 2.6.22-rc3 nmi watchdog hang Ingo Molnar
2007-06-10 18:10 ` Björn Steinbrink
2007-06-18 12:11 ` Ingo Molnar
2007-06-18 14:31 ` Björn Steinbrink
2007-06-25 6:18 ` Ingo Molnar
2007-06-25 6:59 ` Ingo Molnar
2007-06-25 8:05 ` [patch, 2.6.22-rc6] fix nmi_watchdog=2 bootup hang Ingo Molnar
2007-06-25 8:26 ` Ingo Molnar
2007-06-25 12:45 ` Björn Steinbrink
2007-06-25 12:49 ` Jeremy Fitzhardinge
2007-06-25 13:06 ` Björn Steinbrink
2007-06-25 18:50 ` [patch, 2.6.22-rc6] fix nmi_watchdog=2 bootup hang, take #2 Ingo Molnar
2007-06-25 12:40 ` [patch, 2.6.22-rc6] fix nmi_watchdog=2 bootup hang Jeremy Fitzhardinge
2007-06-25 13:13 ` Björn Steinbrink
2007-06-05 15:16 ` 2.6.22-rc3-mm1 Rusty Russell
2007-06-05 21:50 ` [-mm patch] kernel/power/disk.c: make code static Adrian Bunk
2007-06-05 22:10 ` Rafael J. Wysocki
2007-06-05 21:50 ` [-mm patch] kernel/sched{,_fair}.c: " Adrian Bunk
2007-06-06 6:54 ` Andrew Morton
2007-06-06 7:30 ` Ingo Molnar
2007-06-06 12:31 ` Adrian Bunk
2007-06-06 7:02 ` Ingo Molnar
2007-06-05 21:51 ` [-mm patch] kernel/lockdep_proc.c: make 2 functions static Adrian Bunk
2007-06-05 22:34 ` Peter Zijlstra
2007-06-05 21:51 ` [-mm patch] fs/proc/base.c: make a struct static Adrian Bunk
2007-06-06 7:32 ` Ingo Molnar
2007-06-05 12:05 2.6.22-rc3-mm1 Zoltan Boszormenyi
2007-06-05 16:17 ` 2.6.22-rc3-mm1 Andrew Morton
2007-06-05 16:20 ` 2.6.22-rc3-mm1 Jeff Garzik
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=Pine.LNX.4.64.0705311136200.11008@schroedinger.engr.sgi.com \
--to=clameter@sgi.com \
--cc=akpm@linux-foundation.org \
--cc=linux-kernel@vger.kernel.org \
--cc=michal.k.k.piotrowski@gmail.com \
--cc=rjw@sisk.pl \
--cc=takeuchi_satoru@jp.fujitsu.com \
/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
all inboxes | Powered by JetHome®