Re: dwc3 gadget: controller stop times out on system sleep

[Thinh Nguyen]

On Fri, Apr 28, 2023, Roger Quadros wrote:
> 
> 
> On 28/04/2023 01:13, Thinh Nguyen wrote:
> > On Thu, Apr 27, 2023, Roger Quadros wrote:
> >> Hi,
> >>
> >> On 26/04/2023 23:01, Thinh Nguyen wrote:
> >>> Hi,
> >>>
> >>> On Wed, Apr 26, 2023, Roger Quadros wrote:
> >>>> Hi Thinh,
> >>>>
> >>>> On Linux kernel v6.3
> >>>> Test procedure: 
> >>>>
> >>>> - modprobe g_zero
> >>>> - Connect to PC host
> >>>> - systemctl suspend
> >>>>
> >>>> A large delay of 3 seconds is observed. The delay comes from dwc3_gadget_suspend()->dwc3_gadget_run_stop() waiting for DWC3_DSTS_DEVCTRLHLT to be set.
> >>>> It returns -ETIMEDOUT.
> >>>>
> >>>> Are we missing something to do a clean stop during suspend?
> >>>>
> >>>> FYI. Unloading g_zero does not show this delay on stop.
> >>>>
> >>>> cheers,
> >>>> -roger
> >>>
> >>> When clearing run_stop bit and the controller doesn't halt, that usually
> >>> means there are active transfers/endpoints that aren't ended yet.
> >>>
> >>> The dwc3_gadget_suspend() doesn't properly do all the cleanup before
> >>> clearing the run_stop bit. I think you just need to call
> >>> dwc3_gadget_soft_disconnect() in dwc3_gadget_suspend() to fix this.
> >>
> >> That seems to do the trick.
> >> How does this look?
> >>
> >> -------------------------- drivers/usb/dwc3/gadget.c --------------------------
> >> @@ -4674,11 +4676,18 @@ void dwc3_gadget_exit(struct dwc3 *dwc)
> >>  int dwc3_gadget_suspend(struct dwc3 *dwc)
> >>  {
> >>  	unsigned long flags;
> >> +	int ret;
> >>  
> >> -	if (!dwc->gadget_driver)
> >> +	if (!dwc->gadget_driver || !dwc->softconnect)
> >>  		return 0;
> >>  
> >> -	dwc3_gadget_run_stop(dwc, false, false);
> >> +	ret = dwc3_gadget_soft_disconnect(dwc);
> >> +	if (ret)
> >> +		goto err0;
> >> +
> >> +	ret = dwc3_gadget_run_stop(dwc, false, false);
> >> +	if (ret)
> >> +		goto err1;
> >>  
> > 
> > We already clear run_stop in dwc3_gadget_soft_disconnect().
> > 
> > Can you try the following change (not tested):
> > 
> > diff --git a/drivers/usb/dwc3/gadget.c b/drivers/usb/dwc3/gadget.c
> > index c0ca4d12f95d..2996bcb4d53d 100644
> > --- a/drivers/usb/dwc3/gadget.c
> > +++ b/drivers/usb/dwc3/gadget.c
> > @@ -2699,6 +2699,21 @@ static int dwc3_gadget_soft_disconnect(struct dwc3 *dwc)
> >  	return ret;
> >  }
> >  
> > +static int dwc3_gadget_soft_connect(struct dwc3 *dwc)
> > +{
> > +	/*
> > +	 * In the Synopsys DWC_usb31 1.90a programming guide section
> > +	 * 4.1.9, it specifies that for a reconnect after a
> > +	 * device-initiated disconnect requires a core soft reset
> > +	 * (DCTL.CSftRst) before enabling the run/stop bit.
> > +	 */
> > +	dwc3_core_soft_reset(dwc);
> > +
> > +	dwc3_event_buffers_setup(dwc);
> > +	__dwc3_gadget_start(dwc);
> > +	return dwc3_gadget_run_stop(dwc, true);
> 
> return dwc3_gadget_run_stop(dwc, true, false);

You need to rebase your change to Greg's usb-next.

> 
> > +}> +
> >  static int dwc3_gadget_pullup(struct usb_gadget *g, int is_on)
> >  {
> >  	struct dwc3		*dwc = gadget_to_dwc(g);
> > @@ -2737,21 +2752,10 @@ static int dwc3_gadget_pullup(struct usb_gadget *g, int is_on)
> >  
> >  	synchronize_irq(dwc->irq_gadget);
> >  
> > -	if (!is_on) {
> > +	if (!is_on)
> >  		ret = dwc3_gadget_soft_disconnect(dwc);
> > -	} else {
> > -		/*
> > -		 * In the Synopsys DWC_usb31 1.90a programming guide section
> > -		 * 4.1.9, it specifies that for a reconnect after a
> > -		 * device-initiated disconnect requires a core soft reset
> > -		 * (DCTL.CSftRst) before enabling the run/stop bit.
> > -		 */
> > -		dwc3_core_soft_reset(dwc);
> > -
> > -		dwc3_event_buffers_setup(dwc);
> > -		__dwc3_gadget_start(dwc);
> > -		ret = dwc3_gadget_run_stop(dwc, true);
> > -	}
> > +	else
> > +		ret = dwc3_gadget_soft_connect(dwc);
> >  
> >  	pm_runtime_put(dwc->dev);
> >  
> > @@ -4655,42 +4659,39 @@ void dwc3_gadget_exit(struct dwc3 *dwc)
> >  int dwc3_gadget_suspend(struct dwc3 *dwc)
> >  {
> >  	unsigned long flags;
> > +	int ret;
> >  
> >  	if (!dwc->gadget_driver)
> 
> We need to check for dwc->softconnect here. If it is not set that means
> controller has already stopped so we can simply exit.

That makes sense. Perhaps that can be a separate change?

> 
> >  		return 0;
> >  
> > -	dwc3_gadget_run_stop(dwc, false);
> > +	ret = dwc3_gadget_soft_disconnect(dwc);
> > +	if (ret)
> > +		goto err;
> >  
> >  	spin_lock_irqsave(&dwc->lock, flags);
> >  	dwc3_disconnect_gadget(dwc);
> > -	__dwc3_gadget_stop(dwc);
> >  	spin_unlock_irqrestore(&dwc->lock, flags);
> >  
> >  	return 0;
> > +
> > +err:
> > +	/*
> > +	 * Attempt to reset the controller's state. Likely no
> > +	 * communication can be established until the host
> > +	 * performs a port reset.
> > +	 */
> > +	if (dwc->softconnect)

If we check dwc->softconnect at start, we can skip the check here.

> > +		dwc3_gadget_soft_connect(dwc);
> > +
> > +	return ret;
> >  }
> >  
> >  int dwc3_gadget_resume(struct dwc3 *dwc)
> >  {
> > -	int			ret;
> > -
> >  	if (!dwc->gadget_driver || !dwc->softconnect)
> >  		return 0;
> >  
> > -	ret = __dwc3_gadget_start(dwc);
> > -	if (ret < 0)
> > -		goto err0;
> > -
> > -	ret = dwc3_gadget_run_stop(dwc, true);
> > -	if (ret < 0)
> > -		goto err1;
> > -
> > -	return 0;
> > -
> > -err1:
> > -	__dwc3_gadget_stop(dwc);
> > -
> > -err0:
> > -	return ret;
> > +	return dwc3_gadget_soft_connect(dwc);
> >  }
> >  
> >  void dwc3_gadget_process_pending_events(struct dwc3 *dwc)
> 
> Everything else looks ok. I will send a patch soon.
> 

Thanks,
Thinh