fabi/Xenia-Canary
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

[GPU] Reuse base+index*stride in vfetch_mini instead of reloading the index GPR

Browse Source 
The wheel shader in 4D530910 does vfetch_full to r0 with the index from r0.x, and then vfetch_mini.
Thanks @Gliniak for the finding :3
Also small formatting cleanup in commented-out code.
This commit is contained in:
Triang3l
2022-01-09 14:58:38 +03:00
parent 600c14b3f0
commit d6188c5d7e
7 changed files with 103 additions and 66 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
src/xenia/gpu/shader_translator.cc
View File

@@ -876,7 +876,7 @@ bool ParseVertexFetchInstruction(const VertexFetchInstruction& op,
instr.attributes.stride = full_op.stride();
instr.attributes.exp_adjust = op.exp_adjust();
instr.attributes.prefetch_count = op.prefetch_count();
instr.attributes.is_index_rounded = op.is_index_rounded();
instr.attributes.is_index_rounded = full_op.is_index_rounded();
instr.attributes.is_signed = op.is_signed();
instr.attributes.is_integer = !op.is_normalized();
instr.attributes.signed_rf_mode = op.signed_rf_mode();