Skip to contentSkip to conversation

New: Segmentation fault when shutting down

Bugme-new Andrew Morton 2 participants 2 messages Jan 12, 2010 – Jan 13, 2010 1 message not in the archive High-resolution timers, timer wheel, clockevents

lore
Share & tools
lore
#1 lore raw
In reply to a message not in the archive

(switched to email. Please respond via emailed reply-to-all, not via the bugzilla web interface).

Quoted 22 lines · On Thu, 7 Jan 2010 20:59:20 GMT [email protected] wrote:

http://bugzilla.kernel.org/show_bug.cgi?id=15005

           Summary: Segmentation fault when shutting down
           Product: ACPI
           Version: 2.5
    Kernel Version: 2.6.32.3
          Platform: All
        OS/Version: Linux
              Tree: Mainline
            Status: NEW
          Severity: blocking
          Priority: P1
         Component: Power-Off
        AssignedTo: [email protected]
        ReportedBy: [email protected]
        Regression: Yes

Created an attachment (id=24479)
--> (http://bugzilla.kernel.org/attachment.cgi?id=24479)
Screenshot of segmentation fault

Quoted 8 lines

Currently compiled newest kernel 2.6.32.3 for Atom based netbook (datacask Jupiter 1014a). Compiled the kernel with optimizations for Atom CPUs and also with optimizations for i586.
When shutting down the system the kernel produces a segmentation fault at the end of the shutdown process. Powering off fails. Both kernel fail with the same error.
Last working kernel was 2.6.32.2.

It's a shutdown-time oops in clockevents_notify(). A 2.6.32.2 -> 2.6.32.3 regression.

2.6.32.3 included this prime suspect:

: commit fa3f5a5c1c8e6a2cbc7e21755ea7c215f8cf0577 : Author: Thomas Gleixner <[email protected]>
: Date: Thu Dec 10 15:35:10 2009 +0100
:
: clockevents: Prevent clockevent_devices list corruption on cpu hotplug
:
: commit bb6eddf7676e1c1f3e637aa93c5224488d99036f upstream.
:
: which

So we may well have the same regression in 2.6.33-rcX.

Martin, can you please check whether the below revert fixes things up?

Thanks.

From: Andrew Morton <[email protected]>

Revert

: commit bb6eddf7676e1c1f3e637aa93c5224488d99036f
: Author: Thomas Gleixner <[email protected]>
: AuthorDate: Thu Dec 10 15:35:10 2009 +0100
: Commit: Thomas Gleixner <[email protected]>
: CommitDate: Fri Dec 11 10:28:08 2009 +0100 :
: clockevents: Prevent clockevent_devices list corruption on cpu hotplug

due to the regression reported in
http://bugzilla.kernel.org/show_bug.cgi?id=15005

Cc Xiaotian Feng <[email protected]>
Cc Thomas Gleixner <[email protected]>
Cc "Rafael J. Wysocki" <[email protected]>

---

 kernel/time/clockevents.c |   18 +++---------------
 1 file changed, 3 insertions(+), 15 deletions(-)
kernel/time/clockevents.c renamed from kernel/time/clockevents.c~revert-clockevents-prevent-clockevent_devices-list-corruption-on-cpu-hotplug +3 -15
diff -puN kernel/time/clockevents.c~revert-clockevents-prevent-clockevent_devices-list-corruption-on-cpu-hotplug kernel/time/clockevents.c
@@ -238,9 +238,8 @@ void clockevents_exchange_device(struct 
  */
 void clockevents_notify(unsigned long reason, void *arg)
 {
-	struct clock_event_device *dev, *tmp;
+	struct list_head *node, *tmp;
 	unsigned long flags;
-	int cpu;
 
 	raw_spin_lock_irqsave(&clockevents_lock, flags);
 	clockevents_do_notify(reason, arg);
@@ -251,19 +250,8 @@ void clockevents_notify(unsigned long re
 		 * Unregister the clock event devices which were
 		 * released from the users in the notify chain.
 		 */
-		list_for_each_entry_safe(dev, tmp, &clockevents_released, list)
-			list_del(&dev->list);
-		/*
-		 * Now check whether the CPU has left unused per cpu devices
-		 */
-		cpu = *((int *)arg);
-		list_for_each_entry_safe(dev, tmp, &clockevent_devices, list) {
-			if (cpumask_test_cpu(cpu, dev->cpumask) &&
-			    cpumask_weight(dev->cpumask) == 1) {
-				BUG_ON(dev->mode != CLOCK_EVT_MODE_UNUSED);
-				list_del(&dev->list);
-			}
-		}
+		list_for_each_safe(node, tmp, &clockevents_released)
+			list_del(node);
 		break;
 	default:
 		break;

_

#2 lore raw

I'm not sure, but I can't see the start of the seg fault. Is it a kernel NULL pointer deref or trigger the BUG? Mind to try patch in http://bugzilla.kernel.org/show_bug.cgi?id=15037?

Quoted 125 lines · Andrew Morton

(switched to email. Please respond via emailed reply-to-all, not via the bugzilla web interface).

Quoted 22 lines · On Thu, 7 Jan 2010 20:59:20 GMT [email protected] wrote:

http://bugzilla.kernel.org/show_bug.cgi?id=15005

            Summary: Segmentation fault when shutting down
            Product: ACPI
            Version: 2.5
     Kernel Version: 2.6.32.3
           Platform: All
         OS/Version: Linux
               Tree: Mainline
             Status: NEW
           Severity: blocking
           Priority: P1
          Component: Power-Off
         AssignedTo: [email protected]
         ReportedBy: [email protected]
         Regression: Yes

Created an attachment (id=24479)
--> (http://bugzilla.kernel.org/attachment.cgi?id=24479)
Screenshot of segmentation fault

Quoted 8 lines

Currently compiled newest kernel 2.6.32.3 for Atom based netbook (datacask Jupiter 1014a). Compiled the kernel with optimizations for Atom CPUs and also with optimizations for i586.
When shutting down the system the kernel produces a segmentation fault at the end of the shutdown process. Powering off fails. Both kernel fail with the same error.
Last working kernel was 2.6.32.2.

It's a shutdown-time oops in clockevents_notify(). A 2.6.32.2 -> 2.6.32.3 regression.

2.6.32.3 included this prime suspect:

: commit fa3f5a5c1c8e6a2cbc7e21755ea7c215f8cf0577
: Author: Thomas Gleixner<[email protected]>
: Date: Thu Dec 10 15:35:10 2009 +0100
:
: clockevents: Prevent clockevent_devices list corruption on cpu hotplug
:
: commit bb6eddf7676e1c1f3e637aa93c5224488d99036f upstream.
:
: which

So we may well have the same regression in 2.6.33-rcX.

Martin, can you please check whether the below revert fixes things up?

Thanks.

From: Andrew Morton<[email protected]>

Revert

: commit bb6eddf7676e1c1f3e637aa93c5224488d99036f
: Author: Thomas Gleixner<[email protected]>
: AuthorDate: Thu Dec 10 15:35:10 2009 +0100
: Commit: Thomas Gleixner<[email protected]>
: CommitDate: Fri Dec 11 10:28:08 2009 +0100
:
: clockevents: Prevent clockevent_devices list corruption on cpu hotplug

due to the regression reported in
http://bugzilla.kernel.org/show_bug.cgi?id=15005

Cc Xiaotian Feng<[email protected]>
Cc Thomas Gleixner<[email protected]>
Cc "Rafael J. Wysocki"<[email protected]>

---

  kernel/time/clockevents.c |   18 +++---------------
  1 file changed, 3 insertions(+), 15 deletions(-)

diff -puN kernel/time/clockevents.c~revert-clockevents-prevent-clockevent_devices-list-corruption-on-cpu-hotplug kernel/time/clockevents.c

kernel/time/clockevents.c renamed from kernel/time/clockevents.c~revert-clockevents-prevent-clockevent_devices-list-corruption-on-cpu-hotplug +3 -15
@@ -238,9 +238,8 @@ void clockevents_exchange_device(struct
   */
  void clockevents_notify(unsigned long reason, void *arg)
  {
-	struct clock_event_device *dev, *tmp;
+	struct list_head *node, *tmp;
  	unsigned long flags;
-	int cpu;
 
  	raw_spin_lock_irqsave(&clockevents_lock, flags);
  	clockevents_do_notify(reason, arg);
@@ -251,19 +250,8 @@ void clockevents_notify(unsigned long re
  		 * Unregister the clock event devices which were
  		 * released from the users in the notify chain.
  		 */
-		list_for_each_entry_safe(dev, tmp,&clockevents_released, list)
-			list_del(&dev->list);
-		/*
-		 * Now check whether the CPU has left unused per cpu devices
-		 */
-		cpu = *((int *)arg);
-		list_for_each_entry_safe(dev, tmp,&clockevent_devices, list) {
-			if (cpumask_test_cpu(cpu, dev->cpumask)&&
-			    cpumask_weight(dev->cpumask) == 1) {
-				BUG_ON(dev->mode != CLOCK_EVT_MODE_UNUSED);
-				list_del(&dev->list);
-			}
-		}
+		list_for_each_safe(node, tmp,&clockevents_released)
+			list_del(node);
  		break;
  	default:
  		break;

_