diff options
| author | Matt Arsenault <Matthew.Arsenault@amd.com> | 2016-09-30 01:50:20 +0000 |
|---|---|---|
| committer | Matt Arsenault <Matthew.Arsenault@amd.com> | 2016-09-30 01:50:20 +0000 |
| commit | 5d8eb25e78c0eefb54dafd88088bd6fb39f2a125 (patch) | |
| tree | 9261d97dc3de6b150a68f685bf89db4034c00486 /llvm/test/CodeGen/AMDGPU/icmp64.ll | |
| parent | b3949ef88534b0f87d1cd675adf6582dfbf6cae5 (diff) | |
| download | bcm5719-llvm-5d8eb25e78c0eefb54dafd88088bd6fb39f2a125.tar.gz bcm5719-llvm-5d8eb25e78c0eefb54dafd88088bd6fb39f2a125.zip | |
AMDGPU: Use unsigned compare for eq/ne
For some reason there are both of these available, except
for scalar 64-bit compares which only has u64. I'm not sure
why there are both (I'm guessing it's for the one bit inputs we
don't use), but for consistency always using the
unsigned one.
llvm-svn: 282832
Diffstat (limited to 'llvm/test/CodeGen/AMDGPU/icmp64.ll')
| -rw-r--r-- | llvm/test/CodeGen/AMDGPU/icmp64.ll | 4 |
1 files changed, 2 insertions, 2 deletions
diff --git a/llvm/test/CodeGen/AMDGPU/icmp64.ll b/llvm/test/CodeGen/AMDGPU/icmp64.ll index 0eaa33ebafe..3d42eac4429 100644 --- a/llvm/test/CodeGen/AMDGPU/icmp64.ll +++ b/llvm/test/CodeGen/AMDGPU/icmp64.ll @@ -2,7 +2,7 @@ ; RUN: llc -march=amdgcn -mcpu=tonga -verify-machineinstrs < %s | FileCheck -check-prefix=SI %s ; SI-LABEL: {{^}}test_i64_eq: -; SI: v_cmp_eq_i64 +; SI: v_cmp_eq_u64 define void @test_i64_eq(i32 addrspace(1)* %out, i64 %a, i64 %b) nounwind { %cmp = icmp eq i64 %a, %b %result = sext i1 %cmp to i32 @@ -11,7 +11,7 @@ define void @test_i64_eq(i32 addrspace(1)* %out, i64 %a, i64 %b) nounwind { } ; SI-LABEL: {{^}}test_i64_ne: -; SI: v_cmp_ne_i64 +; SI: v_cmp_ne_u64 define void @test_i64_ne(i32 addrspace(1)* %out, i64 %a, i64 %b) nounwind { %cmp = icmp ne i64 %a, %b %result = sext i1 %cmp to i32 |
