| commit | d21382097498ca08dac68e578bb55ed061dce8b2 | [log] [tgz] |
|---|---|---|
| author | Matt Arsenault <Matthew.Arsenault@amd.com> | Thu Aug 31 21:17:22 2017 +0000 |
| committer | Matt Arsenault <Matthew.Arsenault@amd.com> | Thu Aug 31 21:17:22 2017 +0000 |
| tree | 3fd462a6c0116a28df96291f86d19f6c37d401a3 | |
| parent | 2439f12f3102a137b954807c81502a19344c6da8 [diff] |
AMDGPU: Turn int pack pattern into build_vector build_vector is a more useful canonical form when pattern matching packed operations, so turn shift into high element into a build_vector. Should show no change for now. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@312282 91177308-0d34-0410-b5e6-96231b3b80d8