projects / platform / upstream / llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: b035a57) | patch
R600/SI: Implement enableAggressiveFMAFusion
authorMatt Arsenault <Matthew.Arsenault@amd.com>
Thu, 29 Jan 2015 19:34:32 +0000 (19:34 +0000)
committerMatt Arsenault <Matthew.Arsenault@amd.com>
Thu, 29 Jan 2015 19:34:32 +0000 (19:34 +0000)
commit423bf3f64adc86b314aa16fe3d874ec0bc1e59d0
treeef342ba0aef13039442072d803bdfed1ccdd2521tree | snapshot
parentb035a5740ce32a5bddfecb826845a350f589e377commit | diff
R600/SI: Implement enableAggressiveFMAFusion

Add tests for the various combines. This should
always be at least cycle neutral on all subtargets for f64,
and faster on some. For f32 we should prefer selecting
v_mad_f32 over v_fma_f32.

llvm-svn: 227484
llvm/lib/Target/R600/SIISelLowering.cpp diff | blob | history
llvm/lib/Target/R600/SIISelLowering.h diff | blob | history
llvm/test/CodeGen/R600/fma-combine.ll [new file with mode: 0644] blob
Domain: System / Toolchain;