Re: hidraw, sixaxis and output reports.

[Alan Ott <alan@signal11.us>]

On 10/02/2010 11:59 AM, Antonio Ospite wrote:
> Hi,
>
> I am trying to set leds on sixaxis now, and the code I have works OK
> with libusb, but the equivalent for hidraw does not.
>
> The procedure to set leds is to send an output report 0x01 with the
> correct data. After looking at drivers/usbhid/hid-core.c in
> usbhid_output_raw_report() I see that output reports are sent over the
> interrupt endpoint when hid->urbout is defined (that is when the device
> has an interrupt out ep, right?), but it looks like this particular
> device is accepting output reports only on the _control_ endpoint; this
> is exactly what the working libusb version does btw.
>
> Just for the record, this dumb hack makes setting leds work on sixaxis
> with hidraw:
>
> diff --git a/drivers/hid/usbhid/hid-core.c b/drivers/hid/usbhid/hid-core.c
> index df80532..731c08c 100644
> --- a/drivers/hid/usbhid/hid-core.c
> +++ b/drivers/hid/usbhid/hid-core.c
> @@ -841,7 +841,7 @@ static int usbhid_output_raw_report(struct hid_device *hid, __u8 *buf, size_t co
>          struct usb_host_interface *interface = intf->cur_altsetting;
>          int ret;
>
> -       if (usbhid->urbout&&  report_type != HID_FEATURE_REPORT) {
> +       if (0&&  usbhid->urbout&&  report_type != HID_FEATURE_REPORT) {
>                  int actual_length;
>                  int skipped_report_id = 0;
>                  if (buf[0] == 0x0) {
>
> If you are interested in the userspace programs, just ping me.
>
> What can I do to handle this properly? Add another quirk? Something
> like HID_QUIRK_NO_INTERRUPT_OUT_EP or HID_QUIRK_FORCE_CONTROL_OUT_EP?
>    

Sounds right to me. If there's an OUT endpoint, the device SHOULD handle 
output reports on it. Anything else is non-compliant, and I think a 
quirk makes sense.

That SixAxis is quite a pain....

Alan.