[PATCH] D28405: SimplifyLibCalls: Replace fabs libcalls with intrinsics
Eli Friedman via Phabricator via llvm-commits
llvm-commits at lists.llvm.org
Mon Jan 9 12:03:10 PST 2017
efriedma added inline comments.
================
Comment at: lib/Transforms/InstCombine/InstCombineCalls.cpp:1646
+ NewFabs->takeName(II);
+ return new FPExtInst(NewFabs, II->getType());
+ }
----------------
Do we need to preserve fast-math flags here?
================
Comment at: lib/Transforms/Utils/SimplifyLibCalls.cpp:1198
+ NewCall->takeName(CI);
+ return NewCall;
}
----------------
Do we need to preserve fast-math flags here?
================
Comment at: test/Transforms/InstCombine/double-float-shrink-2.ll:91
+ ; --> fabsf
+ %E = call fast double @fabs(double %D)
+ %F = fptrunc double %E to float
----------------
We allow putting fast-math flags on arbitrary calls...? Please update LangRef with an explanation of what this means.
https://reviews.llvm.org/D28405
More information about the llvm-commits
mailing list