Enable support for floating-point division reductions
authorYichao Yu <yyc1992@gmail.com>
Sat, 10 Jun 2017 12:45:13 +0000 (08:45 -0400)
committerRaspbian forward porter <root@raspbian.org>
Thu, 11 Jul 2019 03:08:05 +0000 (04:08 +0100)
commit531caa8fd879ae6e9c1935c669ef3a94c71ef044
treeb4ac46d9006bb79b3d215dc8e27ec68958ac0cf4
parent5386759bdcbfeb84e92769e5ffd53948301012f5
Enable support for floating-point division reductions

Similar to fsub, fdiv can also be vectorized using fmul.

Gbp-Pq: Topic julia
Gbp-Pq: Name llvm-D34078-vectorize-fdiv.patch
lib/Transforms/Utils/LoopUtils.cpp
test/Transforms/LoopVectorize/float-reduction.ll