Hi, On 7/14/22 01:56, Ben Greening wrote: > Hi Hans, > > Applying the latest > 0001-ACPI-video-Change-how-we-determine-if-brightness-key.patch you > sent me off-list (my fault, forgot to reply all) and > 0001-ACPI-video-Use-native-backlight-on-Dell-Inspiron-N40.patch makes > it all work again. And a bonus that I don't need any extra kernel > parameters anymore. > >> It would also be interesting if you can start evemu-record on the >> Dell WMI Hotkeys device before pressing any of the brightness keys. >> >> There might still be a single duplicate event reported there on >> the first press. I don't really see a way around that (without causing >> all brightness key presses on some panasonic models to be duplicated), >> but I'm curious if it is a problem at all... > > I rebooted and ran evemu-record before pressing the brightness keys > and "Dell WMI hotkeys" didn't show any events at all. Great thank you for reporting and testing this! I will send the fix on its way to Linus tomorrow and I've just submitted the new acpi_backlight=native quirk upstream too. Regards, Hans > On Wed, Jul 13, 2022 at 6:49 AM Hans de Goede <hdegoede@xxxxxxxxxx> wrote: >> >> Hi Ben, >> >> On 7/13/22 15:29, Hans de Goede wrote: >>> Hi, >>> >>> On 7/13/22 15:08, Ben Greening wrote: >>>> Hi Hans, thanks for getting back to me. >>>> >>>> evemu-record shows events for both "Video Bus" and "Dell WMI hotkeys": >>>> >>>> Video Bus >>>> E: 0.000001 0001 00e0 0001 # EV_KEY / KEY_BRIGHTNESSDOWN 1 >>>> E: 0.000001 0000 0000 0000 # ------------ SYN_REPORT (0) ---------- +0ms >>>> E: 0.000020 0001 00e0 0000 # EV_KEY / KEY_BRIGHTNESSDOWN 0 >>>> E: 0.000020 0000 0000 0000 # ------------ SYN_REPORT (0) ---------- +0ms >>>> >>>> Dell WMI hotkeys >>>> E: 0.000001 0004 0004 57349 # EV_MSC / MSC_SCAN 57349 >>>> E: 0.000001 0001 00e0 0001 # EV_KEY / KEY_BRIGHTNESSDOWN 1 >>>> E: 0.000001 0000 0000 0000 # ------------ SYN_REPORT (0) ---------- +0ms >>>> E: 0.000020 0001 00e0 0000 # EV_KEY / KEY_BRIGHTNESSDOWN 0 >>>> E: 0.000020 0000 0000 0000 # ------------ SYN_REPORT (0) ---------- +0ms >>>> >>>> Adding video.report_key_events=1 with acpi_backlight=video makes >>>> things work like you said it would. >>>> >>>> >>>> With acpi_backlight=video just has intel_backlight. >>>> >>>> Without acpi_backlight=video: >>>> intel_backlight: >>>> max_brightness: 4882 >>>> backlight control works with echo >>>> brightness keys make no change to brightness value >>>> >>>> dell_backlight: >>>> max_brightness: 15 >>>> backlight control doesn't work immediately, but does on reboot >>>> to set brightness at POST. >>>> brightness keys change brightness value, but you don't see the >>>> change until reboot. >>> >>> Ok, so your system lacks ACPI video backlight control, yet still reports >>> brightness keypresses through the ACPI Video Bus. Interesting (weird)... >>> >>> I think I believe I know how to fix the regression, 1 patch coming up. >> >> Can you please give the attached patch a try, with >> video.report_key_events=1 *removed* from the commandline ? >> >> It would also be interesting if you can start evemu-record on the >> Dell WMI Hotkeys device before pressing any of the brightness keys. >> >> There might still be a single duplicate event reported there on >> the first press. I don't really see a way around that (without causing >> all brightness key presses on some panasonic models to be duplicated), >> but I'm curious if it is a problem at all... >> >> Regards, >> >> Hans >> >> >> >> >> >
