On Fri, 08 Mar 2019, Lucas De Marchi <lucas.demarchi@xxxxxxxxx> wrote: > On Fri, Mar 08, 2019 at 02:39:36PM -0800, Rodrigo Vivi wrote: >>> Given that every platform so far has had different oa configurations, >>> that looks to be a hasty assumption that future platforms will be fixed. >> >>I know... But my hope is that at some point it gets stabilized. >> >>Well, or at least start with this so any other gen11 could reuse and >>gen12 would start with that and change later for >= gen12 and on... > > If it's different, there's no harm in making this assumption now and > then later change to cover the differences. Making it consistent > everywhere allows to more easily add the next platform. At some point in time we deliberately made the if ladders and switch cases avoid matching future platforms, and liberally sprinkled MISSING_CASE() all over the place so we would catch each location that needed updating. So we wouldn't miss any. This is the opposite of that approach. I'm not against the new approach, because we do have a lot of places where the existing code just works, but if you do know something will break, or will need to be updated to work, why not use the MISSING_CASE() there? BR, Jani. -- Jani Nikula, Intel Open Source Graphics Center _______________________________________________ Intel-gfx mailing list Intel-gfx@xxxxxxxxxxxxxxxxxxxxx https://lists.freedesktop.org/mailman/listinfo/intel-gfx
