Re: [PATCH] usb: dwc3: EP clear halt leading to incorrect submission of delayed_status

[Wesley Cheng]

Hi Pavan,

On 4/11/2022 8:48 AM, Pavan Kondeti wrote:
> Hi Wesley,
>
> On Wed, Apr 06, 2022 at 06:53:36PM -0700, Wesley Cheng wrote:
> > The usb_ep_clear_halt() API can be called from the function driver, and
> > translates to dwc3_gadget_ep_set_halt().  This routine is shared with when
> > the host issues a clear feature ENDPOINT_HALT, and is differentiated by the
> > protocol argument.  If the following sequence occurs, there can be a
> > situation where the delayed_status flag is improperly cleared for the wrong
> > SETUP transaction:
> >
> > 1. Vendor specific control transfer returns USB_GADGET_DELAYED_STATUS.
> > 2. DWC3 gadget sets dwc->delayed_status to '1'.
> > 3. Another function driver issues a usb_ep_clear_halt() call.
> > 4. DWC3 gadget issues dwc3_stop_active_transfer() and sets
> >     DWC3_EP_PENDING_CLEAR_STALL.
> > 5. EP command complete interrupt triggers for the end transfer, and
> >     dwc3_ep0_send_delayed_status() is allowed to run, as delayed_status
> >     is '1' due to step#1.
> > 6. STATUS phase is sent, and delayed_status is cleared.
> > 7. Vendor specific control transfer is finished being handled, and issues
> >     usb_composite_setup_continue().  This results in queuing of a data
> >     phase.
> >
> > Cache the protocol flag so that DWC3 gadget is aware of when the clear halt
> > is due to a SETUP request from the host versus when it is sourced from a
> > function driver.  This allows for the EP command complete interrupt to know
> > if it needs to issue a delayed status phase.
> >
> > Signed-off-by: Wesley Cheng <quic_wcheng@xxxxxxxxxxx>
> > ---
> >   drivers/usb/dwc3/core.h   | 1 +
> >   drivers/usb/dwc3/ep0.c    | 1 +
> >   drivers/usb/dwc3/gadget.c | 3 ++-
> >   3 files changed, 4 insertions(+), 1 deletion(-)
> >
> > diff --git a/drivers/usb/dwc3/core.h b/drivers/usb/dwc3/core.h
> > index 5c9d467195a6..55f98485c54c 100644
> > --- a/drivers/usb/dwc3/core.h
> > +++ b/drivers/usb/dwc3/core.h
> > @@ -1272,6 +1272,7 @@ struct dwc3 {
> >   	unsigned		connected:1;
> >   	unsigned		softconnect:1;
> >   	unsigned		delayed_status:1;
> > +	unsigned		clear_stall_protocol:1;
> >   	unsigned		ep0_bounced:1;
> >   	unsigned		ep0_expect_in:1;
> >   	unsigned		has_hibernation:1;
> > diff --git a/drivers/usb/dwc3/ep0.c b/drivers/usb/dwc3/ep0.c
> > index 1064be5518f6..aa8476da222d 100644
> > --- a/drivers/usb/dwc3/ep0.c
> > +++ b/drivers/usb/dwc3/ep0.c
> > @@ -1080,6 +1080,7 @@ void dwc3_ep0_send_delayed_status(struct dwc3 *dwc)
> >   	unsigned int direction = !dwc->ep0_expect_in;
> >   
> >   	dwc->delayed_status = false;
> > +	dwc->clear_stall_protocol = 0;
> >   
> >   	if (dwc->ep0state != EP0_STATUS_PHASE)
> >   		return;
> > diff --git a/drivers/usb/dwc3/gadget.c b/drivers/usb/dwc3/gadget.c
> > index ab725d2262d6..c427ddae016f 100644
> > --- a/drivers/usb/dwc3/gadget.c
> > +++ b/drivers/usb/dwc3/gadget.c
> > @@ -2152,6 +2152,7 @@ int __dwc3_gadget_ep_set_halt(struct dwc3_ep *dep, int value, int protocol)
> >   		if (dep->flags & DWC3_EP_END_TRANSFER_PENDING ||
> >   		    (dep->flags & DWC3_EP_DELAY_STOP)) {
> >   			dep->flags |= DWC3_EP_PENDING_CLEAR_STALL;
> > +			dwc->clear_stall_protocol = protocol;
> >   			return 0;
> >   		}
> >   
> > @@ -3483,7 +3484,7 @@ static void dwc3_gadget_endpoint_command_complete(struct dwc3_ep *dep,
> >   		}
> >   
> >   		dep->flags &= ~(DWC3_EP_STALL | DWC3_EP_WEDGE);
> > -		if (dwc->delayed_status)
> > +		if (dwc->clear_stall_protocol)
> >   			dwc3_ep0_send_delayed_status(dwc);
> >   	}
> >   
>
> Is it safe to maintain clear_stall_protocol per dwc3 instance? What if
> CLEAR_FEATURE(halt_endpoint) and usb_ep_clear_halt() are interleaved and
> We come here as part of usb_ep_clear_halt()'s endpoint command complete.
> We may simply send the delayed status corresponding to the protocol clear
> stall.
>
> We can still maintain a global flag if we cache endpoint number in it so
> that we can cross check against the endpoint for which completion received.
>
> Thanks,
> Pavan

Thanks for the comments/feedback.  I agree with what you mentioned, and 
will fix that potential condition.  Will resubmit a new rev, after some 
testing.

Thanks
Wesley Cheng