On Wed, 30 Oct 2013, Felipe Balbi wrote: > now that I reread documentation on set_wedge() I see that Alan's patch > is correct. dwc3 should ignore CLEAR_FEATURE(HALT_ENDPOINT) from the > host, but if the gadget driver directly calls clear_halt() is should > succeed. And now that I look back on it, I see that we should have arranged to have a usb_ep_clear_wedge operation to match usb_ep_set_wedge. Un-wedging an endpoint should leave it halted (but then allow the host to clear the halt). As it is, the gadget driver has to clear both of them together by calling usb_ep_clear_halt. Oh well... It doesn't seem to matter much. > I guess I didn't see this before because I was running mass storage with > stall=0... Actually, this should should up in the USBCV even with stall=0. It's the one place where g_mass_storage ignores the "stall=" parameter; the spec doesn't allow any choice about it. > Alan, if you can send as a proper patch after Pratyush gives his > Tested-by, that'd be great. Will do. Alan Stern -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
