projects / oota-llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: c416d94) | 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)
commitfa711758df9f1f144930bc16478ebf8e15f2e895
tree5e8b0fbeccf2a45b0ed34ae8e4b30fda8ab138eatree | snapshot
parentc416d947350b6c46ca5bed72655fb00a699322cccommit | 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.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@227484 91177308-0d34-0410-b5e6-96231b3b80d8
lib/Target/R600/SIISelLowering.cpp diff | blob | history
lib/Target/R600/SIISelLowering.h diff | blob | history
test/CodeGen/R600/fma-combine.ll [new file with mode: 0644] blob
C/C++ LLVM-based compilers that forbids OOTA behaviors