Maximilian Luz

@Menci that's just a warning, shouldn't make any difference.

Error `7` should be `AE_ALREADY_EXISTS `, so there is already a handler attached. Makes sense looking back at https://github.com/jakeday/linux-surface/issues/286#issuecomment-453378745. I'll have to see how to replace that one. Edit: Right,...

Okay, I don't think there is a clean way to do this. It looks like we have to patch in a quirk in `i2c-core-acpi.c`.

I've pushed a patch and uploaded a kernel for testing (currently Arch, Debian will follow in a few minutes). There should again be some output prefixed with `surface_sam:`.

The `surface_sam: not a SAM device` line is a bit weird (as far as I can see from the DSDT the adapter should only have the SAM device as child),...

I've updated the release.

Nice! What about the dGPU? The patch should also allow for the I2C communication via ACPI to work (e.g. in HGON).

Awesome! I think with that the ACPI part is done, the rest should work via a HID driver. According to the descriptor from one of your logs, the HID vendor/product...

If you want to, you could try and implement a basic HID driver, I haven't looked into that yet. Not sure when I'll get to that.

I can't help you with the HID driver specifics (haven't read into that yet), but the overall driver structure is always the same, so you should get started with a...