diff options
| author | Marek Olsak <marek.olsak@amd.com> | 2016-12-09 19:49:54 +0000 |
|---|---|---|
| committer | Marek Olsak <marek.olsak@amd.com> | 2016-12-09 19:49:54 +0000 |
| commit | 0f55fbae6c7f40938b27571a20752a376ae49eb1 (patch) | |
| tree | 528ef08804bc9a7243957b77561bef09bacbb01b /llvm/test/CodeGen/AMDGPU/attr-amdgpu-num-sgpr.ll | |
| parent | 693e9be9181eb8b3cc91e254a2acb49accafbd9c (diff) | |
| download | bcm5719-llvm-0f55fbae6c7f40938b27571a20752a376ae49eb1.tar.gz bcm5719-llvm-0f55fbae6c7f40938b27571a20752a376ae49eb1.zip | |
AMDGPU/SI: Don't reserve XNACK when it's disabled
Summary:
This frees 2 additional scalar registers.
These are results from all of my 3 patches combined:
Polaris:
Spilled SGPRs: 2231 -> 1517 (-32.00 %)
Tonga:
Spilled SGPRs: 3829 -> 2608 (-31.89 %)
Spilled VGPRs: 100 -> 84 (-16.00 %)
Tonga even spills SGPRs via VGPRs to scratch. That's a compute shader
limited to 64 VGPRs.
Reviewers: tstellarAMD
Subscribers: arsenm, kzhuravl, wdng, nhaehnle, yaxunl, tony-tye
Differential Revision: https://reviews.llvm.org/D27151
llvm-svn: 289262
Diffstat (limited to 'llvm/test/CodeGen/AMDGPU/attr-amdgpu-num-sgpr.ll')
| -rw-r--r-- | llvm/test/CodeGen/AMDGPU/attr-amdgpu-num-sgpr.ll | 12 |
1 files changed, 6 insertions, 6 deletions
diff --git a/llvm/test/CodeGen/AMDGPU/attr-amdgpu-num-sgpr.ll b/llvm/test/CodeGen/AMDGPU/attr-amdgpu-num-sgpr.ll index da49517e003..e4f6e72e697 100644 --- a/llvm/test/CodeGen/AMDGPU/attr-amdgpu-num-sgpr.ll +++ b/llvm/test/CodeGen/AMDGPU/attr-amdgpu-num-sgpr.ll @@ -9,10 +9,10 @@ ; FIXME: Should be ablo to skip this copying of the private segment ; buffer because all the SGPR spills are to VGPRs. -; ALL: s_mov_b64 s[6:7], s[2:3] -; ALL: s_mov_b64 s[4:5], s[0:1] +; ALL: s_mov_b64 s[10:11], s[2:3] +; ALL: s_mov_b64 s[8:9], s[0:1] ; ALL: SGPRBlocks: 1 -; ALL: NumSGPRsForWavesPerEU: 12 +; ALL: NumSGPRsForWavesPerEU: 14 define void @max_12_sgprs(i32 addrspace(1)* %out1, i32 addrspace(1)* %out2, @@ -46,9 +46,9 @@ define void @max_12_sgprs(i32 addrspace(1)* %out1, ; TOSGPR: SGPRBlocks: 1 ; TOSGPR: NumSGPRsForWavesPerEU: 16 -; TOSMEM: s_mov_b64 s[6:7], s[2:3] -; TOSMEM: s_mov_b32 s9, s13 -; TOSMEM: s_mov_b64 s[4:5], s[0:1] +; TOSMEM: s_mov_b64 s[10:11], s[2:3] +; TOSMEM: s_mov_b64 s[8:9], s[0:1] +; TOSMEM: s_mov_b32 s7, s13 ; TOSMEM: SGPRBlocks: 1 ; TOSMEM: NumSGPRsForWavesPerEU: 16 |
