diff options
| author | Matt Arsenault <Matthew.Arsenault@amd.com> | 2017-03-31 19:53:03 +0000 |
|---|---|---|
| committer | Matt Arsenault <Matthew.Arsenault@amd.com> | 2017-03-31 19:53:03 +0000 |
| commit | 8edfaee7be8ba17cd054a90055c91a18c4d239e3 (patch) | |
| tree | f2cc356ac8fcce91d9ab50e1eff2e762995700a0 /llvm/lib/IR/Function.cpp | |
| parent | 3c99441ef4c5325ac3d9e771defae20907245400 (diff) | |
| download | bcm5719-llvm-8edfaee7be8ba17cd054a90055c91a18c4d239e3.tar.gz bcm5719-llvm-8edfaee7be8ba17cd054a90055c91a18c4d239e3.zip | |
AMDGPU: Remove unnecessary ands when f16 is legal
Add a new node to act as a fancy bitcast from f16 operations to
i32 that implicitly zero the high 16-bits of the result.
Alternatively could try making v2f16 legal and canonicalizing
on build_vectors.
llvm-svn: 299246
Diffstat (limited to 'llvm/lib/IR/Function.cpp')
0 files changed, 0 insertions, 0 deletions
