diff options
author | Nikita Popov <nikita.ppv@gmail.com> | 2020-10-29 21:27:59 +0100 |
---|---|---|
committer | Nikita Popov <nikita.ppv@gmail.com> | 2020-10-29 21:27:59 +0100 |
commit | a5f172927df578fe06e4b28da48d979458c88e0a (patch) | |
tree | ce96ddc29d9b8c7ae7aaabc62db2bd6c1dc142b4 /llvm/lib/Transforms/Utils/LoopUtils.cpp | |
parent | fa48ff3fc981783483bcf609798fbb48524300ba (diff) | |
download | llvm-a5f172927df578fe06e4b28da48d979458c88e0a.zip llvm-a5f172927df578fe06e4b28da48d979458c88e0a.tar.gz llvm-a5f172927df578fe06e4b28da48d979458c88e0a.tar.bz2 |
[SDAG] Fix neutral value for vecreduce_fadd
The neutral value for FADD is -0.0, not 0.0, so this is what we
need to pad vectors with.
Diffstat (limited to 'llvm/lib/Transforms/Utils/LoopUtils.cpp')
0 files changed, 0 insertions, 0 deletions