projects / oota-llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 196ee11) | patch
CostModel: increase the default cost of supported floating point operations from...
authorNadav Rotem <nrotem@apple.com>
Fri, 12 Apr 2013 21:15:03 +0000 (21:15 +0000)
committerNadav Rotem <nrotem@apple.com>
Fri, 12 Apr 2013 21:15:03 +0000 (21:15 +0000)
commit9eb366acba65b5779d2129db3a6fb6a0414572d4
tree993baaba80d0870be37cf9e98091f3a2066b339ctree | snapshot
parent196ee11f85ce0148d2c2e33fbe1f1171ac5a8828commit | diff
CostModel: increase the default cost of supported floating point operations from 1 to two. Fixed a few tests that changes because now the cost of one insert + a vector operation on two doubles is lower than two scalar operations on doubles.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@179413 91177308-0d34-0410-b5e6-96231b3b80d8
lib/CodeGen/BasicTargetTransformInfo.cpp diff | blob | history
test/Analysis/CostModel/X86/arith.ll diff | blob | history
test/Transforms/BBVectorize/X86/loop1.ll diff | blob | history
test/Transforms/BBVectorize/X86/simple.ll diff | blob | history
C/C++ LLVM-based compilers that forbids OOTA behaviors