notes.txt

registers:

0x5 - gain
0x6 - dac calibration
0x7 - dac calibration 2
0x0 - write protect?: this is set to 0x5a before most operations
0x2a: Function_En: probably something to do with enabling stuff
      0x5f/0x4f are apparently "sensing mode"

SPI Status:

- bit 1: if unset, needs reset
- bit 3: line is ready?
- bit 7: is calibrated

tyler's device seems to always be in state 0x81, or not reset is calibrated

-----

matching ideas:

find stddev of image, if above 10000 there is probably fprint

using background image: do difference, if lower, retake bg image
			
			with difference, do histogram clamp

profit?

perhaps take the swipe assembling technique from elan.c

outputs a 96x96 image --> rotate

---

initialization: since we're missing gpio from acpi, we use HID:

- search using hid vid/pid settings from registry? hardcoded?

use SPI bus with normal fpinit flow

--- 

reset over hid:

- find the thing
- send it a feature report with id 0xe and length 4 with all zeros: should reset the target

verify: did bit 0 go to 1?

---

finding spi info:

- probably will use udev and tagging: ruleset or sometihng to detect the ACPI id ELAN7001, tag the spi device created with something, find that later from libudev.
- HID device: /shrug