Re: [BUG 3.12.rc4] Oops: unable to handle kernel paging request duringshutdown

From: Knut Petersen
Date: Mon Oct 28 2013 - 10:50:39 EST


On 26.10.2013 13:43, Ingo Molnar wrote:
* Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx> wrote:

On Fri, Oct 25, 2013 at 11:23 AM, Thomas Gleixner <tglx@xxxxxxxxxxxxx> wrote:
Enable DEBUG_OBJECTS, DEBUG_OBJECTS_FREE, DEBUG_OBJECTS_TIMERS
Yes, but nobody has actually been able to trigger it with those.
It's pretty rare, and the debug options are so expensive that they
aren't reasonable to enable generally...

So we need to try to figure out how to trigger it, or narrow
things down some way..
If it's timing then maybe the 1 seconds delay isn't enough
(especially if it's boot related, where we have a couple of delays
already) - the patch below triples it.

(If the bug goes away due to some other side effect then this patch
won't make a difference to reproducability.)

I tried that patch, together with all the kobj debug options enabled.
With that kernel configuration I was unable to reproduce the problem
during 500 typical workload / reboot cycles. Without the debug
options I was able to reproduce the last reported panic after 91 cycles.
The debug options seem to change timing sufficiently to mask the bug ;-(

Knut

Thanks,

Ingo

diff --git a/lib/kobject.c b/lib/kobject.c
index 084f7b1..124188f 100644
--- a/lib/kobject.c
+++ b/lib/kobject.c
@@ -595,7 +595,7 @@ static void kobject_release(struct kref *kref)
pr_info("kobject: '%s' (%p): %s, parent %p (delayed)\n",
kobject_name(kobj), kobj, __func__, kobj->parent);
INIT_DELAYED_WORK(&kobj->release, kobject_delayed_cleanup);
- schedule_delayed_work(&kobj->release, HZ);
+ schedule_delayed_work(&kobj->release, 3*HZ);
#else
kobject_cleanup(kobj);
#endif


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