On Mon, Aug 11, 2025 at 02:39:14PM +0100, Andre Przywara wrote: > On Wed, 6 Aug 2025 08:56:46 -0600 > Tom Rini <tr...@konsulko.com> wrote: > > Hi Tom, > > > The logic for how to handle this video driver is slightly odd. Only in > > the case of when CONFIG_VIDEO_LCD_PANEL_EDP_4_LANE_1620M_VIA_ANX9804 is > > enabled do we need to have this file built, and otherwise we have a > > dummy function in use. Correct the logic by only building this file when > > needed. > > Mmh, this whole code looks very "2010ish", and would deserve a proper > (DM) update, but I wonder what this particular patch fixes?
This patch fixes the problem I hit with "allyesconfig" which resulted in compiling the file and then a function prototype mis-match because, iirc, the header logic and the Makefile logic were different. > If anything, I would drop > CONFIG_VIDEO_LCD_PANEL_EDP_4_LANE_1620M_VIA_ANX9804, and just use > CONFIG_VIDEO_LCD_ANX9804 instead everywhere, since both seem to cover the > same thing, at least with the current code, and the latter looks more sane. > > I am fine with any solution, this seems to affect only this one ancient > board, (so go ahead with whatever looks good to you), but I would just be > curious. There's plenty of time to clean this up differently too, if you're inclined to make the overall logic nicer. I need to get the large number of patches I posted already merged before I take another pass at "allyesconfig" and see what breaks again (I got as far as completing through drivers/, on the hash I started this on). -- Tom
signature.asc
Description: PGP signature
