Hello again, I've identified a large part of it. So the protocol from the trackpad once the alt interface has been set turns out to be synaptics, only with each finger having the standard structure found in synaptics-usb delimited by the 0xCI (finger count/index) byte and all five being inline. As well, the keyboard simply changes HID interface when the alternate setting is done to the touchpad -- and the media keys (and the rest) also become active. I still do not know what the extra NULL bytes at the very end of the touchpad packets are for though, but it could just as easily be padding, I suppose. Looks like a modification of synaptics-usb may be all I need. Thanks, Zach On Sun, May 26, 2013 at 12:11 AM, Zach C. <fxchip@xxxxxxxxx> wrote: > Hey everyone, > > I wasn't entirely sure if there was a better place to ask such a > question, but I figured this might be as good a place as any for a > question like this -- if not, please kindly let me know who I should > ask or even if I should Google better. > > I am using a laptop (specifically the Razer Blade R2), and the > keyboard and mouse share one USB device, differentiating each input by > interface. By default (i.e. when essentially not touched at all), the > trackpad emulates a standard USB HID pointer (with incidental > implementations of certain gestures in-hardware -- pinch-to-zoom will > send button 4 or 5, for instance) and the keyboard a standard USB HID > keyboard. There are four interfaces in total, however: trackpad on 0, > keyboard on 2, "extras" on 1, and vendor-specific bulk-write endpoints > on 3. > > I say "emulates" because it does a very curious thing when the > trackpad interface (0) is set to the USB alternate setting of 1: any > pretense of being HID inputs for keyboard OR mouse is completely > dropped -- yes, even despite the keyboard being on a separate > interface. Keyboard input begins routing via the "extras" interface on > 1 rather than coming in from 2, and while it still looks like HID to > my rather untrained eye, there are actually several more spaces for > keypresses (for the keyboard -- 14 spaces total plus modifier byte for > ctrl/alt/etc) and it appears report IDs (or at least, some sort of ID > corresponding to the input being done) start being prepended to the > incoming data. "Extras" also begins differentiating which macro keys > -- the display-backed keys above the touchpad -- have been pressed (it > does not do this in HID mode) and begins to send input for the Fn and > Razer logo keys. > > The trackpad's enhanced input seems to consist of five structures per > packet, one for each finger, and the last four are prepended with a > byte whose lower half indicates which finger is being reported and > whose higher half indicates how many fingers total are being used. The > structure itself seems to be laid out (in hex): 1P BB XX XX YY YY UU > -- where P *might be* pressure or width, B is mouse button pressed > (and does carry through to all fingers), X and Y are X and Y > coordinates of the finger in big-endian quadrant-1 (higher Y == closer > to top, higher X == closer to right), and I haven't been able to > figure out U yet. After the five structures (which are terminated with > a byte reading x4 -- where "x" in this case is the number of fingers), > there appear to be 24 more bytes that don't ever seem to be anything > other than 0. > > Otherwise, both keyboard and extras "enhanced input" just seem to > report for two different keyboards. If I mash my palm on the touchpad, > though, I get an apparent report ID of 0x05 with some other data that > I'm not sure how to interpret, I assume it's telling me exactly what > I'm doing but nothing more specific than that. > > Has anyone seen an input device like this before? I've also written a > test program which can be found at > http://fxchip.net/RazerBlade/razer-test.tar.bz2 , but it's sort of > rough in a lot of ways. > > If you're interested in the report descriptors, as well, here's a link > to my lsusb -vv -d 1532:0116 output with usbhid detached: > http://pastie.org/7961204 -- if you'd prefer it inline I might be able > to do that too. > > Thanks in advance, > > Zach -- To unsubscribe from this list: send the line "unsubscribe linux-input" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
