On Mon, 02 Mar 2026, Suraj Kandpal <[email protected]> wrote:
> This series aims to fix/update some code with respect to AUX backlight
> via luminance values for both VESA and INTEL HDR codepaths.
> The below are the major changes made in this series.
> - Do not allow 0 brightness by default which makes the screen go blank
> leaving user with no way to see the screen and increase brightness.
> - Take into account luminance_set variable when we decide which funcs
> need to be enabled
> - Update debug logs to make them more concise and help debug which
> code path was taken
Seems to me some of the fixes here are regression fixes, but none of
them have Fixes: tags pointing at the commits, which means none of the
fixes will be picked up by the stable team for backports.
I also note that you've eagerly merged most of this series while the
discussion is still ongoing in the other thread. I appreciate not
merging the specific patches, though. In any case, we've usually settled
on only merging full series or none at all, because we only do CI on the
entire series. Which means the CI results here don't match what was
merged.
There's also a regression on the recent commit 40d2f5820951
("drm/i915/backlight: Remove try_vesa_interface") [1]. Whether that's
related to this series, I don't know, because there are no references to
it. But you will have to figure it out.
What I do know is that the DPCD backlight code is fraught with
perils. Basically any change there could lead to a regression, and
generally I would thread very carefully. I appreciate the eagerness, but
now you also get to look after the regressions and bugs, which might
only start flowing in a year from now. It's all yours.
BR,
Jani.
[1] https://gitlab.freedesktop.org/drm/xe/kernel/-/issues/7514
--
Jani Nikula, Intel
