From: Alexey Dobriyan <adobriyan@sw.ru>
To: Christoph Lameter <clameter@sgi.com>
Cc: Andrew Morton <akpm@linux-foundation.org>,
gregkh@suse.de, linux-kernel@vger.kernel.org
Subject: Re: 2.6.23-rc6-mm1: BUG kmalloc-16: Object padding overwritten (sysfs?)
Date: Thu, 20 Sep 2007 14:32:29 +0400 [thread overview]
Message-ID: <20070920103229.GB6781@localhost.sw.ru> (raw)
In-Reply-To: <20070920075353.GA6781@localhost.sw.ru>
On Thu, Sep 20, 2007 at 11:53:53AM +0400, Alexey Dobriyan wrote:
> On Wed, Sep 19, 2007 at 02:36:04PM -0700, Christoph Lameter wrote:
> > So I would think that this is an actual memory corruption issue on your
> > system and the /sys/slab stuff works fine. slabinfo -v will perform a more
> > details scan to help you located the problem.
>
> OK, I do clean boot, ssh to box, then sudo slabinfo -v.
>
>
> =============================================================================
> BUG kmalloc-16: Object padding overwritten
> -----------------------------------------------------------------------------
The winner is slub-avoid-touching-page-struct-when-freeing-to-per-cpu-slab.patch
Blind bisecting pointed to it and reverting the patch from full -mm makes
the problem go away
From: Christoph Lameter <clameter@sgi.com>
Set c->node to -1 if we allocate from a debug slab instead for SlabDebug
which requires access the page struct cacheline.
Signed-off-by: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---
mm/slub.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff -puN mm/slub.c~slub-avoid-touching-page-struct-when-freeing-to-per-cpu-slab mm/slub.c
--- a/mm/slub.c~slub-avoid-touching-page-struct-when-freeing-to-per-cpu-slab
+++ a/mm/slub.c
@@ -1537,6 +1537,7 @@ debug:
c->page->inuse++;
c->page->freelist = object[c->offset];
+ c->node = -1;
slab_unlock(c->page);
return object;
}
@@ -1560,8 +1561,7 @@ static void __always_inline *slab_alloc(
local_irq_save(flags);
c = get_cpu_slab(s, smp_processor_id());
- if (unlikely(!c->page || !c->freelist ||
- !node_match(c, node)))
+ if (unlikely(!c->freelist || !node_match(c, node)))
object = __slab_alloc(s, gfpflags, node, addr, c);
@@ -1670,7 +1670,7 @@ static void __always_inline slab_free(st
local_irq_save(flags);
debug_check_no_locks_freed(object, s->objsize);
c = get_cpu_slab(s, smp_processor_id());
- if (likely(page == c->page && !SlabDebug(page))) {
+ if (likely(page == c->page && c->node >= 0)) {
object[c->offset] = c->freelist;
c->freelist = object;
} else
_
next prev parent reply other threads:[~2007-09-20 10:33 UTC|newest]
Thread overview: 12+ messages / expand[flat|nested] mbox.gz Atom feed top
2007-09-19 12:39 2.6.23-rc6-mm1: BUG kmalloc-16: Object padding overwritten (sysfs?) Alexey Dobriyan
2007-09-19 12:59 ` Alexey Dobriyan
2007-09-19 19:39 ` Andrew Morton
2007-09-19 20:59 ` Christoph Lameter
2007-09-19 21:36 ` Christoph Lameter
2007-09-20 7:53 ` Alexey Dobriyan
2007-09-20 10:32 ` Alexey Dobriyan [this message]
2007-09-20 17:36 ` Christoph Lameter
2007-09-21 7:16 ` Alexey Dobriyan
2007-09-20 17:25 ` Christoph Lameter
2007-09-20 17:37 ` Christoph Lameter
2007-09-20 18:54 ` Eric W. Biederman
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=20070920103229.GB6781@localhost.sw.ru \
--to=adobriyan@sw.ru \
--cc=akpm@linux-foundation.org \
--cc=clameter@sgi.com \
--cc=gregkh@suse.de \
--cc=linux-kernel@vger.kernel.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.