Date: Sat, 16 May 2009 22:44:53 +0200 From: Christoph Langguth <christoph@rosenkeller.org> To: Hans Petter Selasky <hselasky@c2i.net> Cc: freebsd-usb@freebsd.org Subject: Re: How to add support for Macbook Pro (USB) keyboard? Message-ID: <4A0F25C5.9020706@rosenkeller.org> In-Reply-To: <200905162158.22210.hselasky@c2i.net> References: <4A0DC89C.6010708@rosenkeller.org> <200905162001.30735.hselasky@c2i.net> <4A0F1039.7050403@rosenkeller.org> <200905162158.22210.hselasky@c2i.net>
next in thread | previous in thread | raw e-mail | index | archive | help
This is a multi-part message in MIME format.
--------------030500090406070609070007
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Hi,
>> Finally, for curiosity, since I didn't really get how the code works:
>> how do you know reliably whether to toggle the APPLE_SWAP flag? Does the
>> HID descriptor give you information about that "anomaly"? (I'm just
>> wondering whether it wouldn't mistakenly swap the keys on unaffected
>> Apple keyboards)
>
> I was just taking your example. I assume that Apple make the keyboards alike.
> If not we will have to adjust. I did not look too close at the HID
> descriptor. I just assume that when the special HID item is present, which
> indicates the EJECT, then it is an Apple keyboard and that it works like an
> Apple keyboard.
Hmm... well, I guess it's fine for now, but I believe that actually not
all of the keyboards have this swap "bug". Apple seems to change their
hardware (and the behavior thereof...) quite often :-(
>
>> And... one final issue here, which is low-priority for me but would be
>> the topping on the cake: CAPS lock is functioning normally, but the LED
>> does not light up when engaged. Is there any way to fix this?
>
> Can you find the bit or byte that is changing in the report which corresponds
> to the CAPS lock key? Also don't forget to print the first byte which is the
> ID byte.
That was it... the data sent to the keyboard also needs to be prefixed
with the ID. The attached patch does this in a sensitive way -- I've
tested it and it behaves correctly both with the Apple, and a "normal"
USB kbd attached.
Hooray! :-)
--Chris
--------------030500090406070609070007
Content-Type: text/plain;
name="ukbd-led.diff"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
filename="ukbd-led.diff"
--- ukbd.c.org 2009-05-16 20:38:08.000000000 +0000
+++ ukbd.c 2009-05-16 20:35:26.000000000 +0000
@@ -589,7 +589,8 @@
ukbd_set_leds_callback(struct usb2_xfer *xfer)
{
struct usb2_device_request req;
- uint8_t buf[1];
+ uint8_t buf[2];
+ uint8_t reqsize = 0;
struct ukbd_softc *sc = xfer->priv_sc;
switch (USB_GET_STATE(xfer)) {
@@ -603,15 +604,19 @@
USETW2(req.wValue, UHID_OUTPUT_REPORT, 0);
req.wIndex[0] = sc->sc_iface_no;
req.wIndex[1] = 0;
- USETW(req.wLength, 1);
- buf[0] = sc->sc_leds;
+ if (sc->sc_hid_id != 0) {
+ buf[reqsize++] = sc->sc_hid_id;
+ }
+ buf[reqsize++] = sc->sc_leds;
+
+ USETW(req.wLength, reqsize);
usb2_copy_in(xfer->frbuffers, 0, &req, sizeof(req));
- usb2_copy_in(xfer->frbuffers + 1, 0, buf, sizeof(buf));
+ usb2_copy_in(xfer->frbuffers + 1, 0, buf, reqsize);
xfer->frlengths[0] = sizeof(req);
- xfer->frlengths[1] = sizeof(buf);
+ xfer->frlengths[1] = reqsize;
xfer->nframes = 2;
usb2_start_hardware(xfer);
}
--------------030500090406070609070007--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?4A0F25C5.9020706>