On Tue, Jun 12, 2018 at 6:22 PM, Hanno Zulla <abos@xxxxxxxx> wrote: > Hi Benjamin, > > thanks again for the review. > >>> Roderick, you were right that the buttons were mapped wrongly, so >>> a mapping is added as requested. The descriptor fixup is still >>> needed, as the original descriptor didn't map the analog triggers >>> at all. I kept the analog axis remap in the fixup. Also, the >>> fixup cleans up the output report to avoid the original >>> descriptor's undefined Usage values. >> >> I am not a big fan of having both a rdesc fixup and input_mapping. >> Can't you also fix the button mapping with the fixup? >> If it's too hard, the answer 'no' is fine too. > > The rationale behind it was this: The device sends the analog > left/right trigger data in its raw hid data stream, but the original > descriptor threw those away and didn't map these incoming data to > any axis. For some games, analog triggers are a desirable feature, > so a fixup seemed warranted. > > Thanks for the hint regarding the buttons. I will try to map them > in the fixup, too. > >> we might want to have this failing harder if the report descriptor >> doesn't match. > > Okay, do you have an example for that? Sorry, I should have double checked before saying this. You can't abort a device enumeration here. You should probably put a warning if the VID/PID matches and the report descriptor is not the one you expect. > >> Unless the device is complete garbage with hid-generic, you can just >> remove this hunk. > > Sorry, I do not understand. Why is it not necessary to add that line to > hid_have_special_driver? > Since v4.16 or v4.17, hid-generic will unbind itself nicely if you load an other driver that claims the device. This allows to have your device handled by hid-generic until hid-bigben is loaded, which is helpful for keyboard devices during the initrd. Anyway, if having hid-generic binding the device first doesn't interfere with your driver, I'd rather keep it that way to reduce the hid_have_sepcial_driver list. Cheers, Benjamin > Thanks, > > Hanno -- 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
