Re: 2.6.18-rc5-mm1 + all hotfixes -- INFO: possible recursive locking detected

[Stefan Richter <stefanr@s5r6.in-berlin.de>]

Andrew Morton wrote:
> On Tue, 5 Sep 2006 10:37:51 -0700
> "Miles Lane" <miles.lane@gmail.com> wrote:
> 
>> ieee1394: Node changed: 0-01:1023 -> 0-00:1023
>> ieee1394: Node changed: 0-02:1023 -> 0-01:1023
>> ieee1394: Node suspended: ID:BUS[0-00:1023]  GUID[0080880002103eae]
>>
>> =============================================
>> [ INFO: possible recursive locking detected ]
>> 2.6.18-rc5-mm1 #2
>> ---------------------------------------------
>> knodemgrd_0/2321 is trying to acquire lock:
>>  (&s->rwsem){----}, at: [<f8958897>] nodemgr_probe_ne+0x311/0x38d [ieee1394]
>>
>> but task is already holding lock:
>>  (&s->rwsem){----}, at: [<f8959078>] nodemgr_host_thread+0x717/0x883 [ieee1394]
[...]

This information confuses me. These places are not supposed to be the
ones where the locks were actually acquired, are they?

> That's a 1394 glitch, possibly introduced by git-ieee1394.patch.

Or maybe it's older. Nodemgr takes class->subsys.rwsem and
device.bus->subsys.rwsem. It always did. Could there be a change in
driver core which makes this recursive? Or has it always been recursive?
For example,

static void nodemgr_update_pdrv(struct node_entry *ne)
{
	struct unit_directory *ud;
	struct hpsb_protocol_driver *pdrv;
	struct class *class = &nodemgr_ud_class;
	struct class_device *cdev;

	down_read(&class->subsys.rwsem);
	list_for_each_entry(cdev, &class->children, node) {
		ud = container_of(cdev, struct unit_directory, class_dev);
		if (ud->ne != ne || !ud->device.driver)
			continue;

		pdrv = container_of(ud->device.driver, struct hpsb_protocol_driver,
driver);

		if (pdrv->update && pdrv->update(ud)) {
			down_write(&ud->device.bus->subsys.rwsem);
			device_release_driver(&ud->device);
			up_write(&ud->device.bus->subsys.rwsem);
		}
	}
	up_read(&class->subsys.rwsem);
}


Miles,

perhaps you should rather unapply all 1394 patches at once.
git-ieee1394.patch is alas the lowermost patch of a stack of dependent
patches. I somehow expect that the "possible recursive locking" persists
even if all the 1394 patches were removed.

Thanks in advance,
-- 
Stefan Richter
-=====-=-==- =--= --=-=
http://arcgraph.de/sr/