Re: [PATCH 2.6.19-rc4] usb urb unlink/free clenup

[Andrew Morton <akpm@osdl.org>]

On Mon, 6 Nov 2006 22:28:37 +0100
Mariusz Kozlowski <m.kozlowski@tuxland.pl> wrote:

> Hello,
> 
> 
> 	In many places usb_(unlink,kill,free)_urb() are called this way:
> 
> if (urb)
> 	usb_something_urb(...);
> 
> which is not needed because functions like usb_unlink_urb() and usb_free_urb() are defined this way:
> 
> void usb_free_urb(struct urb *urb)
> {
>         if (urb)
>                 kref_put(&urb->kref, urb_destroy);
> }
> 
> int usb_unlink_urb(struct urb *urb)
> {
>         if (!urb)
>                 return -EINVAL;
> 	...
> }
> 
> We do not need to check for urb != NULL before we call them.

Seems reasonable.

Your patch had all its tabs replaced with spaces by your email client.  I
fixed that all up, but it was rather dull work and I'd prefer not to have
to do it again.

 It is also possible to do similar cleanup
> for usb_kill_urb(). The thing is it does urb check at the begining but right before is 
> might_sleep():
> 
> void usb_kill_urb(struct urb *urb)
> {
>         might_sleep();
>         if (!(urb && urb->dev && urb->dev->bus))
>                 return;
> 	...
> 
> which confuses me. I would like to know what to do about it. Can this be rewritten this way?:
> 
> void usb_kill_urb(struct urb *urb)
> {
> 	if (!urb)
> 		return;
>         might_sleep();
>         if (!urb->dev || !urb->dev->bus))
>                 return;
> 	...
> 
> Or maybe there is no need for this?
> 

I think it's OK as-is.  Presumably it's rare for a caller to pass in a NULL
urb.

It's possible that your proposed change will cause new (and incorrect)
warnings to be emitted, but we can handle those if/when they come out.