Re: [PATCH] driver core: move uevent call to driver_register

[Sebastian Ott <sebott@linux.vnet.ibm.com>]

Hi,

On Mon, 16 Jul 2012, Greg Kroah-Hartman wrote:
> On Tue, Jul 17, 2012 at 09:35:02AM +0800, Ming Lei wrote:
> > On Tue, Jul 3, 2012 at 1:08 AM, Sebastian Ott <sebott@linux.vnet.ibm.com> wrote:
> > 
> > > --- a/drivers/base/driver.c
> > > +++ b/drivers/base/driver.c
> > > @@ -187,6 +187,9 @@ int driver_register(struct device_driver
> > >         ret = driver_add_groups(drv, drv->groups);
> > >         if (ret)
> > >                 bus_remove_driver(drv);
> > > +
> > > +       kobject_uevent(&drv->p->kobj, KOBJ_ADD);
> > 
> > You should just send the uevent if 'ret' equals to zero., otherwise
> > OOPS may be triggered by kobject_uevent() after the 'drv' has been
> > removed.
> 
> Ugh, just missed that.
> 
> Sebastian, care to send a follow-on patch for this?

Here is the follow-on. (I've also replied with an updated patch..just in
case you want to take that one).

[PATCH] driver core: don't trigger uevent after failure

Do not send the uevent if driver_add_groups failed.

Found-by: Ming Lei <ming.lei@canonical.com>
Signed-off-by: Sebastian Ott <sebott@linux.vnet.ibm.com>
---
 drivers/base/driver.c |    5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

--- a/drivers/base/driver.c
+++ b/drivers/base/driver.c
@@ -185,9 +185,10 @@ int driver_register(struct device_driver
 	if (ret)
 		return ret;
 	ret = driver_add_groups(drv, drv->groups);
-	if (ret)
+	if (ret) {
 		bus_remove_driver(drv);
-
+		return ret;
+	}
 	kobject_uevent(&drv->p->kobj, KOBJ_ADD);
 
 	return ret;