https://bugzilla.kernel.org/show_bug.cgi?id=215683
--- Comment #2 from w...@tuxedocomputers.com --- Some more information after some digging and printf debugging: This functions registers some stuff when acpi_backlight=video: https://elixir.bootlin.com/linux/v5.16.14/source/drivers/acpi/acpi_video.c#L1859 When acpi_backlight!=video this function deregisters this stuff again, if registered: https://elixir.bootlin.com/linux/v5.16.14/source/drivers/acpi/acpi_video.c#L2259 When setting acpi_backlight explicitly, it either stays registered or it is never registered so the second function is a noop. The actuall call to reduce screen brigtnes happens in this https://elixir.bootlin.com/linux/v5.16.14/source/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c#L818 respectifly this https://elixir.bootlin.com/linux/v5.16.14/source/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c#L448 function. When on acpi_backlight=video the backlicht seems to be controlled via this atif_handler/SBIOS requests (I don't know what atif or SBIOS stands for). When on acpi_backlight!=video only the ATIF_DGPU_DISPLAY_EVENT is triggered by the SBIOS. However when the registering and unregistering mentioned above happened, the first call to amdgpu_atif_handler catches some dangling SBIOS change brightness event not properly deleted by the unregister it seems. This sets the brightness to a value of 4, which is the observed bug in the first post. -- You may reply to this email to add a comment. You are receiving this mail because: You are watching the assignee of the bug. _______________________________________________ acpi-bugzilla mailing list acpi-bugzilla@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/acpi-bugzilla