On Mon, Feb 26, 2024 at 4:43 PM AngeloGioacchino Del Regno <angelogioacchino.delregno@xxxxxxxxxxxxx> wrote: > > Il 26/02/24 09:21, Fei Shao ha scritto: > > On Thu, Feb 22, 2024 at 4:43 PM AngeloGioacchino Del Regno > > <angelogioacchino.delregno@xxxxxxxxxxxxx> wrote: > >> > >> Il 21/02/24 17:56, Justin Green ha scritto: > >>> Add MT8188 overlay driver configuration data. This change consequently > >>> enables 10-bit overlay support on MT8188 devices. > >>> > >>> Tested by running ChromeOS UI on MT8188 and using modetest -P. AR30 and > >>> BA30 overlays are confirmed to work from modetest. > >>> > >>> Signed-off-by: Justin Green <greenjustin@xxxxxxxxxxxx> > >>> Tested-by: Justin Green <greenjustin@xxxxxxxxxxxx> > >> > >> Hello Justin, > >> > >> I'm 99.9% sure that you don't need this, you can just use compatibles > >> > >> compatible = "mediatek,mt8188-disp-ovl", "mediatek,mt8195-disp-ovl"; > >> > >> as they *are* indeed compatible, and MT8188 does support AFBC as well. > > > > Hi, > > > > I confirmed that I can lit up the MT8188 display with that plus a > > follow-up patch [1]. > > Otherwise a compatible sequence of mt8188, mt8195 and mt8192 would be > > needed, but that would be somewhat redundant. > > > > Actually, that wouldn't be MT8192, but MT8183 - following MT8195's devicetree, Ah yes, you're right. > > so it would be > > compatible = "mediatek,mt8188-disp-ovl", "mediatek,mt8195-disp-ovl", > "mediatek,mt8183-disp-ovl"; > > And yes that could be redundant, but I'm not sure that adding a compatible > string to the mtk_drm_drv matches is a good idea. > > I'd be more for using the triple compatible strings in there instead, as it'd > be like that for only *one* node and not more than that - IMO, that is totally > acceptable and it's also the best (and lightest) solution. I see, I'm also fine with this approach. Thanks for the feedback! Regards, Fei > > Cheers, > Angelo > > > [1]: https://lore.kernel.org/all/20240226080721.3331649-1-fshao@xxxxxxxxxxxx/ > > > > Regards, > > Fei > > >
