On Sat, 21 Dec 2024 13:31:38 +0100,
Mark Kettenis <[email protected]> wrote:
>
> Interesting. Linux seems to define some quirks for this particular
> touchpad; see the code in drivers/hid/hid-multitouch.c. Especially
> the MT_QUIRK_FORCE_GET_FEATURE might be interesting although I'm not
> sure you get far enough for that to matter.
>
Defently not near yet.
Interesting, that with the last hack (or debug ihidev?), when I do shutdown
-h now, machine doesn't shutdown but asking to press any key to reboot.
And after that I have many lines like this:
invalid packet size (65535 vs 156)
The same line that I have when ddb is triggered.
I also discovers some quirk that suspend-and-resume near the same hardware:
-
https://git.kernel.org/pub/scm/linux/kernel/git/hid/hid.git/commit/?h=for-6.13/upstream-fixes&id=e61080220ae7b52920aed292c8d399ea8ce7cfdb
-
https://git.kernel.org/pub/scm/linux/kernel/git/hid/hid.git/commit/?h=for-6.13/upstream-fixes&id=e02b876cf4c171fecc0f3b4910ed3029db566369
But such trick doesn't work here anyway.
I had tried to power on, power off and on, before read the second part of
descriptor, after attach... and many other things like puting sleep for a
0.1 ... 1 seconds.
For a while I run out of ideas and if anyone has any ideas, I willing to try.
--
wbr, Kirill
