| commit | 7d97468a231ea1fa1e9285a7cd62991856a34405 | [log] [tgz] |
|---|---|---|
| author | Matt Arsenault <Matthew.Arsenault@amd.com> | Fri Oct 18 18:26:37 2019 +0000 |
| committer | Matt Arsenault <Matthew.Arsenault@amd.com> | Fri Oct 18 18:26:37 2019 +0000 |
| tree | 7be605221fc9c45baec309b6287756fa8e5ae170 | |
| parent | 90696971b2a4677279b1ebfcaf15ac89f82fefeb [diff] |
AMDGPU: Relax 32-bit SGPR register class Mostly use SReg_32 instead of SReg_32_XM0 for arbitrary values. This will allow the register coalescer to do a better job eliminating copies to m0. For GlobalISel, as a terrible hack, use SGPR_32 for things that should use SCC until booleans are solved. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@375267 91177308-0d34-0410-b5e6-96231b3b80d8