Ben Widawsky <b...@bwidawsk.net> writes: >> + if (brw->gen >= 9 && >> + brw_format_for_mesa_format(irb->mt->format) != >> + brw->render_target_format[irb->mt->format]) > > Could you just do > !brw->format_supported_as_render_target[irb->mt->format]?
Hrm, yes, you're right, that would be much neater. > I forget, did you find failures for this in the non-MSRT case? If not, > maybe we could skip this patch and just take the rest of the series? > That way we can avoid the "perf regression" of RGBX clears which we do > hit on certain workloads. I couldn't find any regressions for the non-MSRT case. However, it looks like the RGBX formats don't use fast clears on non-MSRT anyway because none of them are marked as having lossless compression support. In fact, if we land the whole series then it should be a performance *improvement* on non-MSRT because the RGBX clears become RGBA clears internally and those do support fast clears. Regards, - Neil _______________________________________________ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
