On 6/17/2020 4:16 AM, Álvaro Fernández Rojas wrote: >> On 6328, the same register space allows the controlling of the USB PHY >> in either host or device mode, so I believe you would need to add a >> #phy-cells = 1 in order to distinguish the consumer (host versus device) >> if we get to the point where drivers/usb/gadget/udc/bcm63xx_udc.c >> becomes DT aware. > > I’ve just realized that I have implemented this wrong in v3, because I implemented the SwapControl USB_DEVICE_SEL value, and you meant the UTMIControl1 USB_DEVICE_MODE_SEL value. Right that is the register I was referring to. > So I have a couple of questions about this, because I haven’t got any bcm63xx with usb device configuration to test: > - Is USB_DEVICE_SEL also needed in SwapControl or do we only need USB_DEVICE_MODE_SEL in UTMIControl1? It looks like it depends on the device, for 6318 and 63268, there is USB_DEVICE_MODE_SEL defined, but not for 6328, 6362 or 6368 for instance. Note that USB_DEVICE_MODE_SEL is relevant for port 2 only for 6318 and 63268 whereas the UTMI_CONTROL1 appears to be for any port. > - Are the rest of the host values needed when configured in device mode? Should I only set the device values when in device mode? They could probably be configured although I am not sure they sure they will be used at all, it's been a while since I looked at this (over 8 years). I don't know if you have any board with USB device mode capability, if you do not please email privately and I will ship you one. -- Florian
