On Wed, Nov 02, 2011 at 03:10:14PM +0100, Sebastian Andrzej Siewior wrote:
> This patch avoids the compiler spitting out the following warning:
> |drivers/usb/dwc3/gadget.c:1304: warning: â??trbâ?? is used uninitialized in this function
>
> This is only uninitialized if the list of to-cleanup TRBs is empty which
> should not be the case because we call this functions once a transfer
> completed so it should be on list. We return early in the which should
> never happen so the warning vanishes.
>
> Signed-off-by: Sebastian Andrzej Siewior <bigeasy@xxxxxxxxxxxxx>
> ---
>
> * Felipe Balbi | 2011-11-02 15:10:15 [+0200]:
> >but then we will loose the IOC bit check. What do you think ?
>
> It should be okay. The IOC check is only to figure out if the controller
> is still processing packets or not. That means if we got the IOC event
> then the we can cleanup and the controller is still working. In the other
> case (LST set or SHORT OUT) the controller is stopped and waiting for
> us.
> So in the very unlikely case where the list empty I think we should
> behave like we're done and the controller idle. This is undefined anyway
> because we should never be here :) Maybe adding a WARN_ON() would be
> nice no catch this unlikely event.
you're missing one detail, I guess. That break is actually what break us
out of the loop when req_queued isn't empty. See that we have:
do {
req = next_request(&dep->req_queued);
if (!req)
break;
...
dwc3_gadget_giveback(dep, req);
...
} while (true);
Should we change that to:
while (!list_empty(&dep->req_queued) {
}
to avoid problems ?
--
balbi
Attachment:
signature.asc
Description: Digital signature
