Re: kobject leak in next-20080829 (and mainline)

Previous thread: [PATCH 2/2] ne.c msleep not mdelay by David Fries on Friday, August 29, 2008 - 7:44 pm. (7 messages)

Next thread: [PATCH] proc: move sysrq-trigger out of fs/proc/ by Alexey Dobriyan on Friday, August 29, 2008 - 10:28 pm. (1 message)
From: Alexey Dobriyan
Date: Friday, August 29, 2008 - 9:42 pm

size-32: 3511 kvasprintf+0x57/0x90
	size-64: 3059 kobject_create+0x1c/0x40

These two buddies increase after every "modprobe; rmmod" sequence.

--

From: Greg KH
Date: Friday, August 29, 2008 - 10:43 pm

Do they also do so in mainline?

thanks,

greg k-h
--

From: Alexey Dobriyan
Date: Friday, September 5, 2008 - 10:45 pm

Yep, same bug in mainline.

--

From: Greg KH
Date: Friday, September 12, 2008 - 8:51 am

Not good.  Does 2.6.26 show this as well?

thanks,

greg k-h
--

From: Alexey Dobriyan
Date: Sunday, September 14, 2008 - 1:42 am

Uh-oh, this was actually 10-minute exercise.



[PATCH] Fix "notes" kobject leak

It happens every rmmod if KALLSYMS=y and SYSFS=y.

	# modprobe foo

kobject: 'foo' (ffffffffa00743d0): kobject_add_internal: parent: 'module', set: 'module'
kobject: 'holders' (ffff88017e7c5770): kobject_add_internal: parent: 'foo', set: '<NULL>'
kobject: 'foo' (ffffffffa00743d0): kobject_uevent_env
kobject: 'foo' (ffffffffa00743d0): fill_kobj_path: path = '/module/foo'
kobject: 'notes' (ffff88017fa9b668): kobject_add_internal: parent: 'foo', set: '<NULL>'
	  ^^^^^

	# rmmod foo

kobject: 'holders' (ffff88017e7c5770): kobject_cleanup
kobject: 'holders' (ffff88017e7c5770): auto cleanup kobject_del
kobject: 'holders' (ffff88017e7c5770): calling ktype release
kobject: (ffff88017e7c5770): dynamic_kobj_release
kobject: 'holders': free name
kobject: 'foo' (ffffffffa00743d0): kobject_cleanup
kobject: 'foo' (ffffffffa00743d0): does not have a release() function, it is broken and must be fixed.
kobject: 'foo' (ffffffffa00743d0): auto cleanup 'remove' event
kobject: 'foo' (ffffffffa00743d0): kobject_uevent_env
kobject: 'foo' (ffffffffa00743d0): fill_kobj_path: path = '/module/foo'
kobject: 'foo' (ffffffffa00743d0): auto cleanup kobject_del
kobject: 'foo': free name

	[whooops]

Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
---

 kernel/module.c |    1 +
 1 file changed, 1 insertion(+)

--- a/kernel/module.c
+++ b/kernel/module.c
@@ -1174,6 +1174,7 @@ static void free_notes_attrs(struct module_notes_attrs *notes_attrs,
 			sysfs_remove_bin_file(notes_attrs->dir,
 					      &notes_attrs->attrs[i]);
 		kobject_del(notes_attrs->dir);
+		kobject_put(notes_attrs->dir);
 	}
 	kfree(notes_attrs);
 }
--

From: Greg KH
Date: Monday, September 15, 2008 - 3:53 pm

Hm, no, that should just be a call to kobject_put() instead of
kobject_del(), can you try that instead and see if that solves the issue
(am at a conference and can't test that at the moment, sorry.)

thanks,

greg k-h
--

From: Alexey Dobriyan
Date: Monday, September 15, 2008 - 4:35 pm

Why?! kobject_del() puts parent kobject.
--

From: Greg KH
Date: Monday, September 15, 2008 - 4:40 pm

And that's about it (well, it also tears down the sysfs stuff.)

If you call kobject_put() instead, and it's the last put, it will also
call kobject_del() and do everything else that it needs to do.

Try it and let me know if I'm right or not :)

thanks,

greg k-h
--

From: Alexey Dobriyan
Date: Tuesday, September 23, 2008 - 12:51 pm

OK, this works too.


[PATCH v2] Fix "notes" kobject leak

It happens every rmmod if KALLSYMS=y and SYSFS=y.

	# modprobe foo

kobject: 'foo' (ffffffffa00743d0): kobject_add_internal: parent: 'module', set: 'module'
kobject: 'holders' (ffff88017e7c5770): kobject_add_internal: parent: 'foo', set: '<NULL>'
kobject: 'foo' (ffffffffa00743d0): kobject_uevent_env
kobject: 'foo' (ffffffffa00743d0): fill_kobj_path: path = '/module/foo'
kobject: 'notes' (ffff88017fa9b668): kobject_add_internal: parent: 'foo', set: '<NULL>'
	  ^^^^^

	# rmmod foo

kobject: 'holders' (ffff88017e7c5770): kobject_cleanup
kobject: 'holders' (ffff88017e7c5770): auto cleanup kobject_del
kobject: 'holders' (ffff88017e7c5770): calling ktype release
kobject: (ffff88017e7c5770): dynamic_kobj_release
kobject: 'holders': free name
kobject: 'foo' (ffffffffa00743d0): kobject_cleanup
kobject: 'foo' (ffffffffa00743d0): does not have a release() function, it is broken and must be fixed.
kobject: 'foo' (ffffffffa00743d0): auto cleanup 'remove' event
kobject: 'foo' (ffffffffa00743d0): kobject_uevent_env
kobject: 'foo' (ffffffffa00743d0): fill_kobj_path: path = '/module/foo'
kobject: 'foo' (ffffffffa00743d0): auto cleanup kobject_del
kobject: 'foo': free name

	[whooops]

Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
---

 kernel/module.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

--- a/kernel/module.c
+++ b/kernel/module.c
@@ -1173,7 +1173,7 @@ static void free_notes_attrs(struct module_notes_attrs *notes_attrs,
 		while (i-- > 0)
 			sysfs_remove_bin_file(notes_attrs->dir,
 					      &notes_attrs->attrs[i]);
-		kobject_del(notes_attrs->dir);
+		kobject_put(notes_attrs->dir);
 	}
 	kfree(notes_attrs);
 }
--

From: Greg KH
Date: Tuesday, September 23, 2008 - 1:56 pm

Great, thanks for testing, I'll apply this to my trees.

greg k-h
--

Previous thread: [PATCH 2/2] ne.c msleep not mdelay by David Fries on Friday, August 29, 2008 - 7:44 pm. (7 messages)

Next thread: [PATCH] proc: move sysrq-trigger out of fs/proc/ by Alexey Dobriyan on Friday, August 29, 2008 - 10:28 pm. (1 message)