diff options
| author | Craig Topper <craig.topper@gmail.com> | 2017-03-31 17:24:29 +0000 |
|---|---|---|
| committer | Craig Topper <craig.topper@gmail.com> | 2017-03-31 17:24:29 +0000 |
| commit | 9601168670c12bf45d7037f91d34a187363f15dd (patch) | |
| tree | 364dc3af884f2a774b25068f03c1aadc635d375b /llvm/test/CodeGen/X86 | |
| parent | f771f79b2fbd1e1dc47b7308f0db2c2f8b4bd56f (diff) | |
| download | bcm5719-llvm-9601168670c12bf45d7037f91d34a187363f15dd.tar.gz bcm5719-llvm-9601168670c12bf45d7037f91d34a187363f15dd.zip | |
[AVX-512] Update lowering for gather/scatter prefetch intrinsics to match the immediate encodings the frontend uses based on the _MM_HINT_T0/T1 constant values in clang's headers.
Our _MM_HINT_T0/T1 constant values are 3/2 which matches gcc, but not icc or Intel documentation. Interestingly gcc had this same bug on their implementation of the gather/scatter builtins at one point too.
Fixes PR32411.
llvm-svn: 299234
Diffstat (limited to 'llvm/test/CodeGen/X86')
| -rw-r--r-- | llvm/test/CodeGen/X86/avx512-gather-scatter-intrin.ll | 4 |
1 files changed, 2 insertions, 2 deletions
diff --git a/llvm/test/CodeGen/X86/avx512-gather-scatter-intrin.ll b/llvm/test/CodeGen/X86/avx512-gather-scatter-intrin.ll index 3eaee2d313f..4890afec216 100644 --- a/llvm/test/CodeGen/X86/avx512-gather-scatter-intrin.ll +++ b/llvm/test/CodeGen/X86/avx512-gather-scatter-intrin.ll @@ -285,9 +285,9 @@ define void @prefetch(<8 x i64> %ind, i8* %base) { ; CHECK-NEXT: vscatterpf1qps (%rdi,%zmm0,2) {%k1} ; CHECK-NEXT: vzeroupper ; CHECK-NEXT: retq - call void @llvm.x86.avx512.gatherpf.qps.512(i8 -1, <8 x i64> %ind, i8* %base, i32 4, i32 1) + call void @llvm.x86.avx512.gatherpf.qps.512(i8 -1, <8 x i64> %ind, i8* %base, i32 4, i32 3) call void @llvm.x86.avx512.gatherpf.qps.512(i8 0, <8 x i64> %ind, i8* %base, i32 4, i32 2) - call void @llvm.x86.avx512.scatterpf.qps.512(i8 1, <8 x i64> %ind, i8* %base, i32 2, i32 1) + call void @llvm.x86.avx512.scatterpf.qps.512(i8 1, <8 x i64> %ind, i8* %base, i32 2, i32 3) call void @llvm.x86.avx512.scatterpf.qps.512(i8 120, <8 x i64> %ind, i8* %base, i32 2, i32 2) ret void } |
