| commit | 7614ec64311822d77d393cdad5d1e3e8d857fc77 | [log] [tgz] |
|---|---|---|
| author | Matt Arsenault <Matthew.Arsenault@amd.com> | Mon May 02 20:07:26 2016 +0000 |
| committer | Matt Arsenault <Matthew.Arsenault@amd.com> | Mon May 02 20:07:26 2016 +0000 |
| tree | e6145d884c24e8ebcee5489b0c611e0cb2e861d9 | |
| parent | 98d359ac65a13bc99c21f38ef8e1d15cf69f7645 [diff] |
AMDGPU: Make i64 loads/stores promote to v2i32 Now that unaligned access expansion should not attempt to produce i64 accesses, we can remove the hack in PreprocessISelDAG where this is done. This allows splitting i64 private accesses while allowing the new add nodes indexing the vector components can be folded with the base pointer arithmetic. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@268293 91177308-0d34-0410-b5e6-96231b3b80d8