projects / oota-llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 68027c4) | patch
Enable constant propagation for more math functions
authorErik Schnetter <schnetter@gmail.com>
Thu, 27 Aug 2015 16:36:37 +0000 (16:36 +0000)
committerErik Schnetter <schnetter@gmail.com>
Thu, 27 Aug 2015 16:36:37 +0000 (16:36 +0000)
commit6164af29bb0332b0d826b2acce6b0c783fb18ef8
tree01a3c432b7e4a277787ce8eb5b4363da23033da4tree | snapshot
parent68027c4618270e93bc8627e78a82557fca9a7babcommit | diff
Enable constant propagation for more math functions

Constant propagation for single precision math functions (such as
tanf) is already working, but was not enabled. This patch enables
these for many single-precision functions, and adds respective test
cases.

Newly handled functions: acosf asinf atanf atan2f ceilf coshf expf
exp2f fabsf floorf fmodf logf log10f powf sinhf tanf tanhf

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@246158 91177308-0d34-0410-b5e6-96231b3b80d8
lib/Analysis/ConstantFolding.cpp diff | blob | history
test/Transforms/ConstProp/calls.ll diff | blob | history
C/C++ LLVM-based compilers that forbids OOTA behaviors