]> granicus.if.org Git - llvm/commit
Late evaluation of the fast-math vectorization requirement.
authorTyler Nowicki <tyler.nowicki@gmail.com>
Mon, 10 Aug 2015 19:51:46 +0000 (19:51 +0000)
committerTyler Nowicki <tyler.nowicki@gmail.com>
Mon, 10 Aug 2015 19:51:46 +0000 (19:51 +0000)
commit9224227bc2310ddda5e4bf7bc8aebde48fae7f4d
treed375848cdc513ca667b5bf7cb9c5a39a26ab6e24
parentbeb8a9efb779a5692909248aa9a7f3b17289838e
Late evaluation of the fast-math vectorization requirement.

This patch moves the verification of fast-math to just before vectorization is done. This way we can tell clang to append the command line options would that allow floating-point commutativity. Specifically those are enableing fast-math or specifying a loop hint.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@244489 91177308-0d34-0410-b5e6-96231b3b80d8
include/llvm/IR/DiagnosticInfo.h
include/llvm/Transforms/Utils/LoopUtils.h
lib/IR/DiagnosticInfo.cpp
lib/IR/LLVMContext.cpp
lib/Transforms/Utils/LoopUtils.cpp
lib/Transforms/Vectorize/LoopVectorize.cpp
test/Transforms/LoopVectorize/no_fpmath.ll [new file with mode: 0644]