From: Ingo Molnar <mingo@elte.hu>
To: Tejun Heo <tj@kernel.org>
Cc: Pekka Enberg <penberg@kernel.org>,
Peter Zijlstra <a.p.zijlstra@chello.nl>,
Linus Torvalds <torvalds@linux-foundation.org>,
Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
Thomas Gleixner <tglx@linutronix.de>,
"H. Peter Anvin" <hpa@zytor.com>,
Andrew Morton <akpm@linux-foundation.org>,
Pekka Enberg <penberg@cs.helsinki.fi>
Subject: Re: percpu related boot crash on x86 (was: Linux 2.6.38-rc1)
Date: Wed, 19 Jan 2011 21:53:53 +0100 [thread overview]
Message-ID: <20110119205353.GA14810@elte.hu> (raw)
In-Reply-To: <20110119131329.GB14096@mtj.dyndns.org>
* Tejun Heo <tj@kernel.org> wrote:
> Hello,
>
> On Wed, Jan 19, 2011 at 02:56:23PM +0200, Pekka Enberg wrote:
> > Is vmalloc() and vfree() supposed to work with interrupts disabled? I
> > always thought they weren't which would mean something in
> > pcpu_mem_alloc() needs changing...
>
> They aren't but percpu allocator doesn't call them with interrupts
> disabled normally. It only happens during early boot before it's
> enabled for the first time but this is an allowed exception. We'll
> probably need to update the WARN_ON to consider the rare call path.
> It also looks like something in that path is turning off/on IRQ
> instead of saving/restoring it, so that probably needs to be changed
> too but both aren't really critical.
>
> I think the real problem is a lot of unsigned int percpu allocations
> added by 6c9ae009 (irq: use per_cpu kstat_irqs). percpu allocator
> should be able to cope but it assumes that it would always be able to
> allocate contiguous allocation map using vmalloc and that might be
> failing, although I cannot readily see how or why that would happen
> frequently on a 64bit machine.
Btw., based on Peter's suggestion i tried the limit-bumping hack below - but it did
not help, the crash still triggers on 5%-10% of all randconfig bootups.
Thanks,
Ingo
>From e2c013d28c8fe81cd30976dcd3058cc86c6285d3 Mon Sep 17 00:00:00 2001
From: Ingo Molnar <mingo@elte.hu>
Date: Wed, 19 Jan 2011 20:15:20 +0100
Subject: [PATCH] <not for upstream> percpu: Increase allocation limits
Might solve the boot crashes ...
Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
mm/percpu.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/mm/percpu.c b/mm/percpu.c
index 3f93001..5cd9cf4 100644
--- a/mm/percpu.c
+++ b/mm/percpu.c
@@ -74,7 +74,7 @@
#include <asm/io.h>
#define PCPU_SLOT_BASE_SHIFT 5 /* 1-31 shares the same slot */
-#define PCPU_DFL_MAP_ALLOC 16 /* start a map with 16 ents */
+#define PCPU_DFL_MAP_ALLOC 64 /* start a map with 64 ents */
#ifdef CONFIG_SMP
/* default addr <-> pcpu_ptr mapping, override in asm/percpu.h if necessary */
next prev parent reply other threads:[~2011-01-19 20:54 UTC|newest]
Thread overview: 69+ messages / expand[flat|nested] mbox.gz Atom feed top
2011-01-18 23:54 Linux 2.6.38-rc1 Linus Torvalds
2011-01-19 2:07 ` Linus Torvalds
2011-01-19 3:42 ` Justin Mattock
2011-01-19 3:53 ` Linus Torvalds
2011-01-19 4:05 ` Justin Mattock
2011-01-19 5:10 ` nobody
2011-01-19 5:42 ` Linus Torvalds
2011-01-19 5:56 ` Linus Torvalds
2011-01-19 18:07 ` Steven Rostedt
2011-01-19 18:26 ` Linus Torvalds
2011-01-21 3:23 ` tag&track [Re: Linux 2.6.38-rc1] nobody
2011-01-21 8:24 ` Alexey Dobriyan
2011-01-23 8:13 ` Linux 2.6.38-rc1 Török Edwin
2011-01-19 7:39 ` Linux 2.6.38-rc1 doesn't boot Markus Trippelsdorf
2011-01-19 7:46 ` Shaohua Li
2011-01-19 7:55 ` Markus Trippelsdorf
2011-01-19 7:49 ` Markus Trippelsdorf
2011-01-19 8:12 ` Shaohua Li
2011-01-19 8:56 ` H. Peter Anvin
2011-01-19 9:09 ` Ingo Molnar
2011-01-20 2:08 ` Shaohua Li
2011-01-20 3:32 ` Lu, Hongjiu
2011-01-20 11:25 ` Ingo Molnar
2011-01-20 15:08 ` Anvin, H Peter
2011-01-21 7:18 ` Shaohua Li
2011-01-21 7:55 ` Shaohua Li
2011-01-21 15:28 ` H. Peter Anvin
2011-01-21 15:37 ` Lu, Hongjiu
2011-01-21 21:09 ` Ingo Molnar
2011-01-19 9:33 ` [tip:x86/urgent] Revert "x86: Make relocatable kernel work with new binutils" tip-bot for Ingo Molnar
2011-01-20 4:59 ` Linux 2.6.38-rc1 doesn't boot Alexandre Courbot
2011-01-19 8:39 ` PPS parport boot lockup: INFO: HARDIRQ-READ-safe -> HARDIRQ-READ-unsafe lock order detected Ingo Molnar
2011-01-20 13:04 ` Alexander Gordeev
2011-01-21 14:44 ` Alexander Gordeev
2011-01-21 16:37 ` Linus Torvalds
2011-01-21 19:43 ` Ingo Molnar
2011-01-24 23:28 ` Alexander Gordeev
2011-01-24 23:46 ` [PATCH] pps: claim parallel port exclusively Alexander Gordeev
2011-01-25 0:19 ` Ingo Molnar
2011-01-24 15:00 ` PPS parport boot lockup: INFO: HARDIRQ-READ-safe -> HARDIRQ-READ-unsafe lock order detected Alexander Gordeev
2011-01-24 15:12 ` [PATCH] parport: make lockdep happy with waitlist_lock Alexander Gordeev
2011-01-24 15:28 ` Ingo Molnar
2011-01-24 15:33 ` Alexander Gordeev
2011-01-19 12:02 ` percpu related boot crash on x86 (was: Linux 2.6.38-rc1) Ingo Molnar
2011-01-19 12:44 ` Tejun Heo
2011-01-19 12:48 ` Peter Zijlstra
2011-01-19 12:56 ` Pekka Enberg
2011-01-19 13:12 ` Peter Zijlstra
2011-01-19 13:13 ` Tejun Heo
2011-01-19 20:53 ` Ingo Molnar [this message]
2011-01-19 23:11 ` Ingo Molnar
2011-01-20 8:31 ` percpu related boot crash on x86 Pekka Enberg
2011-01-20 10:47 ` Peter Zijlstra
2011-01-20 11:12 ` Eric Dumazet
2011-01-20 11:19 ` Tejun Heo
2011-01-20 11:06 ` [PATCH 1/2] lockdep: move early boot local IRQ enable/disable status to init/main.c Tejun Heo
2011-01-20 11:07 ` [PATCH 2/2] smp: allow on_each_cpu() to be called while early_boot_irqs_disabled " Tejun Heo
2011-01-20 20:22 ` [tip:core/urgent] smp: Allow " tip-bot for Tejun Heo
2011-01-20 11:11 ` [PATCH 1/2] lockdep: move early boot local IRQ enable/disable " Tejun Heo
2011-01-20 11:23 ` Peter Zijlstra
2011-01-20 11:26 ` Tejun Heo
2011-01-20 11:30 ` Pekka Enberg
2011-01-20 11:38 ` Peter Zijlstra
2011-01-20 12:00 ` Ingo Molnar
2011-01-20 12:20 ` [PATCH UPDATED " Tejun Heo
2011-01-20 11:51 ` [PATCH " Ingo Molnar
2011-01-20 20:21 ` [tip:core/urgent] lockdep: Move " tip-bot for Tejun Heo
2011-01-19 21:40 ` Linux 2.6.38-rc1 Alan Cox
2011-01-21 15:30 ` Aaro Koskinen
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=20110119205353.GA14810@elte.hu \
--to=mingo@elte.hu \
--cc=a.p.zijlstra@chello.nl \
--cc=akpm@linux-foundation.org \
--cc=hpa@zytor.com \
--cc=linux-kernel@vger.kernel.org \
--cc=penberg@cs.helsinki.fi \
--cc=penberg@kernel.org \
--cc=tglx@linutronix.de \
--cc=tj@kernel.org \
--cc=torvalds@linux-foundation.org \
/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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.