On Wed, Apr 20, 2016 at 10:02:33AM +0300, Roger Quadros wrote:
> On 19/04/16 11:06, Peter Chen wrote:
> > On Tue, Apr 05, 2016 at 05:05:12PM +0300, Roger Quadros wrote:
> >> +/**
> >> + * usb_otg_start_host - start/stop the host controller
> >> + * @otg: usb_otg instance
> >> + * @on: true to start, false to stop
> >> + *
> >> + * Start/stop the USB host controller. This function is meant
> >> + * for use by the OTG controller driver.
> >> + */
> >> +int usb_otg_start_host(struct usb_otg *otg, int on)
> >> +{
> >> + struct otg_hcd_ops *hcd_ops = otg->hcd_ops;
> >> +
> >> + dev_dbg(otg->dev, "otg: %s %d\n", __func__, on);
> >> + if (!otg->host) {
> >> + WARN_ONCE(1, "otg: fsm running without host\n");
> >> + return 0;
> >> + }
> >> +
> >> + if (on) {
> >> + if (otg->flags & OTG_FLAG_HOST_RUNNING)
> >> + return 0;
> >> +
> >> + otg->flags |= OTG_FLAG_HOST_RUNNING;
> >> +
> >> + /* start host */
> >> + hcd_ops->add(otg->primary_hcd.hcd, otg->primary_hcd.irqnum,
> >> + otg->primary_hcd.irqflags);
> >> + if (otg->shared_hcd.hcd) {
> >> + hcd_ops->add(otg->shared_hcd.hcd,
> >> + otg->shared_hcd.irqnum,
> >> + otg->shared_hcd.irqflags);
> >> + }
> >
> > Check the return value please.
>
> And what should we do on failure?
> Even if things fail, they could potentially start working on next
> remove/add iteration so I didn't bother checking return values.
>
If usb_add_hcd has failed, the hcd may be released (usb_put_hcd is
called), in that case, we can't call usb_remove_hcd, maybe we may
need to add hcd valid check for primary hcd too. Even we can't
stop fsm, we need to show an error message for user.
Chipidea idea have a bug before:
commit 41314fea2ffb6dc716b7e698a47c68b329602fc0
Author: Russell King - ARM Linux <li...@arm.linux.org.uk>
Date: Wed Oct 16 13:45:15 2013 +0100
usb/chipidea: fix oops on memory allocation failure
--
Best Regards,
Peter Chen
