Message ID | 1442508160-20601-1-git-send-email-balbi@ti.com |
---|---|
State | Superseded |
Headers | show |
On Thu, Sep 17, 2015 at 11:42:38AM -0500, Felipe Balbi wrote: > Instead of clearing DWC3_PENDING_REQUEST when > we start transfer, let's do it when the request > is actually queued, that way we know for sure > that we're clearing in the right time. > > Signed-off-by: Felipe Balbi <balbi@ti.com> looks like I forgot one patch, sorry.
diff --git a/drivers/usb/dwc3/gadget.c b/drivers/usb/dwc3/gadget.c index d28dc6df4e55..2cc6a33438dc 100644 --- a/drivers/usb/dwc3/gadget.c +++ b/drivers/usb/dwc3/gadget.c @@ -948,7 +948,6 @@ static int __dwc3_gadget_kick_transfer(struct dwc3_ep *dep, u16 cmd_param, dwc3_trace(trace_dwc3_gadget, "%s: endpoint busy", dep->name); return -EBUSY; } - dep->flags &= ~DWC3_EP_PENDING_REQUEST; /* * If we are getting here after a short-out-packet we don't enqueue any @@ -1117,6 +1116,10 @@ static int __dwc3_gadget_ep_queue(struct dwc3_ep *dep, struct dwc3_request *req) if (ret && ret != -EBUSY) dev_dbg(dwc->dev, "%s: failed to kick transfers\n", dep->name); + + if (!ret) + dep->flags &= ~DWC3_EP_PENDING_REQUEST; + return ret; }
Instead of clearing DWC3_PENDING_REQUEST when we start transfer, let's do it when the request is actually queued, that way we know for sure that we're clearing in the right time. Signed-off-by: Felipe Balbi <balbi@ti.com> --- drivers/usb/dwc3/gadget.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-)