Explicit status-stage requests for USB gadgets

Explicit status-stage requests for USB gadgets

[Alan Stern]

Felipe and Paul:

I finally took the time to put more work into the explicit-status-stage
project.  Now I've got patches adding support to both dummy-hcd and
net2280.  (The two patches for net2280 I sent earlier today address
unrelated issues that turned up while this work was under way.)  Along
with Paul's work on musb, it seems like a good start and worth merging.

Getting someone to do the same for dwc2 and dwc3 shouldn't be very 
hard, but they are out of my ballpark.

Paul, there is one change I think we should make to your original core
patch.  usb_gadget_control_complete() should not check request->length,
because it has to be possible for a gadget driver to send a zero-length
reply as the data stage for a control-IN request and still get an 
implicit status-stage request.

Aside from that, everything seems to be ready to move forward.

Felipe, any comments?

Alan Stern

Re: Explicit status-stage requests for USB gadgets

[Felipe Balbi]

[-- Attachment #1: Type: text/plain, Size: 1196 bytes --]


Hi Alan,

Alan Stern <stern@rowland.harvard.edu> writes:
> I finally took the time to put more work into the explicit-status-stage
> project.  Now I've got patches adding support to both dummy-hcd and
> net2280.  (The two patches for net2280 I sent earlier today address
> unrelated issues that turned up while this work was under way.)  Along
> with Paul's work on musb, it seems like a good start and worth merging.

Thanks for doing all the work.

> Getting someone to do the same for dwc2 and dwc3 shouldn't be very 
> hard, but they are out of my ballpark.

I can do dwc3, no problem.

> Paul, there is one change I think we should make to your original core
> patch.  usb_gadget_control_complete() should not check request->length,
> because it has to be possible for a gadget driver to send a zero-length
> reply as the data stage for a control-IN request and still get an 
> implicit status-stage request.
>
> Aside from that, everything seems to be ready to move forward.
>
> Felipe, any comments?

I haven't had time to look at the patches yet due to some internal
projects happening at this moment. I'll try to get to them still this
week.

-- 
balbi

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 832 bytes --]