diff options
| author | Dan Gohman <gohman@apple.com> | 2010-03-02 01:11:08 +0000 |
|---|---|---|
| committer | Dan Gohman <gohman@apple.com> | 2010-03-02 01:11:08 +0000 |
| commit | 6f34abd0926c1d15e723512e7f69ca664c380ee1 (patch) | |
| tree | d64903405f45255604987467abdfd74aa93417e2 /llvm/lib/Transforms/InstCombine/InstCombineMulDivRem.cpp | |
| parent | 6232f73c7b615c87661bd2aad8d6cb2f880a05a3 (diff) | |
| download | bcm5719-llvm-6f34abd0926c1d15e723512e7f69ca664c380ee1.tar.gz bcm5719-llvm-6f34abd0926c1d15e723512e7f69ca664c380ee1.zip | |
Floating-point add, sub, and mul are now spelled fadd, fsub, and fmul,
respectively.
llvm-svn: 97531
Diffstat (limited to 'llvm/lib/Transforms/InstCombine/InstCombineMulDivRem.cpp')
| -rw-r--r-- | llvm/lib/Transforms/InstCombine/InstCombineMulDivRem.cpp | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/llvm/lib/Transforms/InstCombine/InstCombineMulDivRem.cpp b/llvm/lib/Transforms/InstCombine/InstCombineMulDivRem.cpp index 380e1f87b08..b3974e8eeff 100644 --- a/llvm/lib/Transforms/InstCombine/InstCombineMulDivRem.cpp +++ b/llvm/lib/Transforms/InstCombine/InstCombineMulDivRem.cpp @@ -203,7 +203,7 @@ Instruction *InstCombiner::visitFMul(BinaryOperator &I) { // "In IEEE floating point, x*1 is not equivalent to x for nans. However, // ANSI says we can drop signals, so we can do this anyway." (from GCC) if (Op1F->isExactlyValue(1.0)) - return ReplaceInstUsesWith(I, Op0); // Eliminate 'mul double %X, 1.0' + return ReplaceInstUsesWith(I, Op0); // Eliminate 'fmul double %X, 1.0' } else if (Op1C->getType()->isVectorTy()) { if (ConstantVector *Op1V = dyn_cast<ConstantVector>(Op1C)) { // As above, vector X*splat(1.0) -> X in all defined cases. |
