Re: WARNING: at block/genhd.c:1556 disk_clear_events+0xdc/0xf0()

[Tejun Heo]

On Thu, Apr 14, 2011 at 05:00:00AM +0900, Tejun Heo wrote:
> Thanks a lot for finding out the test case.  I'll see whether I can
> reproduce the problem in qemu.

Hmm... can't reproduce the problem with qemu.  Can you please apply
the following patch and see which warning messages trigger?

Thanks.

diff --git a/block/genhd.c b/block/genhd.c
index b364bd0..f061975 100644
--- a/block/genhd.c
+++ b/block/genhd.c
@@ -1377,6 +1377,8 @@ struct disk_events {
 
 	long			poll_msecs;	/* interval, -1 for default */
 	struct delayed_work	dwork;
+	struct mutex		mutex;
+	atomic_t		cnt;
 };
 
 static const char *disk_events_strs[] = {
@@ -1572,6 +1574,9 @@ static void disk_events_workfn(struct work_struct *work)
 	unsigned long intv;
 	int nr_events = 0, i;
 
+	WARN_ON_ONCE(atomic_add_return(1, &ev->cnt) != 1);
+	mutex_lock(&ev->mutex);
+
 	/* check events */
 	events = disk->fops->check_events(disk, clearing);
 
@@ -1595,6 +1600,9 @@ static void disk_events_workfn(struct work_struct *work)
 
 	if (nr_events)
 		kobject_uevent_env(&disk_to_dev(disk)->kobj, KOBJ_CHANGE, envp);
+
+	mutex_unlock(&ev->mutex);
+	WARN_ON_ONCE(atomic_add_return(-1, &ev->cnt) != 0);
 }
 
 /*
@@ -1749,6 +1757,8 @@ static void disk_add_events(struct gendisk *disk)
 	ev->block = 1;
 	ev->poll_msecs = -1;
 	INIT_DELAYED_WORK(&ev->dwork, disk_events_workfn);
+	mutex_init(&ev->mutex);
+	atomic_set(&ev->cnt, 0);
 
 	mutex_lock(&disk_events_mutex);
 	list_add_tail(&ev->node, &disk_events);
--
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/