Hi Mario, On 8/7/23 13:32, Mario Limonciello wrote: > On 8/7/23 04:44, Hans de Goede wrote: >> Linux defaults to picking the non-working ACPI video backlight interface >> on the Apple iMac12,1 and iMac12,2. >> >> Add a DMI quirk to pick the working native radeon_bl0 interface instead. >> >> Link: https://gitlab.freedesktop.org/drm/amd/-/issues/1838 >> Link: https://gitlab.freedesktop.org/drm/amd/-/issues/2753 >> Signed-off-by: Hans de Goede <hdegoede@xxxxxxxxxx> >> --- >> drivers/acpi/video_detect.c | 18 ++++++++++++++++++ >> 1 file changed, 18 insertions(+) > > I'm wondering if you also in the same series should be reverting these two commits? > > 05eacc198c68 ("drm/radeon: Fix eDP for single-display iMac11,2") > 564d8a2cf3ab ("drm/radeon: Fix eDP for single-display iMac10,1 (v2)") Those 2 patches select which encoder to use to driver the eDP link, this seems to be something independent of / orthogonal to which backlight contrl method to use which is what this patch is about. Regards, Hans > >> >> diff --git a/drivers/acpi/video_detect.c b/drivers/acpi/video_detect.c >> index 18cc08c858cf..7cceaf31d928 100644 >> --- a/drivers/acpi/video_detect.c >> +++ b/drivers/acpi/video_detect.c >> @@ -486,6 +486,24 @@ static const struct dmi_system_id video_detect_dmi_table[] = { >> DMI_MATCH(DMI_PRODUCT_NAME, "iMac11,3"), >> }, >> }, >> + { >> + /* https://gitlab.freedesktop.org/drm/amd/-/issues/1838 */ >> + .callback = video_detect_force_native, >> + /* Apple iMac12,1 */ >> + .matches = { >> + DMI_MATCH(DMI_SYS_VENDOR, "Apple Inc."), >> + DMI_MATCH(DMI_PRODUCT_NAME, "iMac12,1"), >> + }, >> + }, >> + { >> + /* https://gitlab.freedesktop.org/drm/amd/-/issues/2753 */ >> + .callback = video_detect_force_native, >> + /* Apple iMac12,2 */ >> + .matches = { >> + DMI_MATCH(DMI_SYS_VENDOR, "Apple Inc."), >> + DMI_MATCH(DMI_PRODUCT_NAME, "iMac12,2"), >> + }, >> + }, >> { >> /* https://bugzilla.redhat.com/show_bug.cgi?id=1217249 */ >> .callback = video_detect_force_native, >
