On Thu, 2023-05-11 at 18:35 -0700, John.C.Harrison@xxxxxxxxx wrote: > From: John Harrison <John.C.Harrison@xxxxxxxxx> > > The GuC has a completely separate engine class enum when referring to > register capture lists, which combines render and compute. The driver > was using the 'normal' GuC specific engine class enum instead. That > meant that it thought it was defining a capture list for compute > engines, the list was actually being applied to the GSC engine. And if > a platform didn't have a render engine, then it would get no compute > register captures at all. > > Fix that. > > Signed-off-by: John Harrison <John.C.Harrison@xxxxxxxxx> alan:snip. LGTM, simple and straight-forward patch - although i can only imagine the pain of debugging this one. So for the benefit of others on the mailing list, because the COMPUTE and RENDER enum of the i915 (not-GuC-ABI) was different, but the GuC-ABI one was using the its own Render for both, (coincidentially '0' == render for both GUC-ABI and i915), it means that ADS popultion of capture-register list would only cause problems for devices that had no render or has a GSC (all have VD/VE/Blt). So MTL is not yet fully POR and none of the existing upstream supported devices had no render engine so therefore the "Fixes" tag isn't needed IIUC. That said, pending CI results, Reviewed-by: Alan Previn <alan.previn.teres.alexis@xxxxxxxxx>
