diff options
| author | Craig Topper <craig.topper@intel.com> | 2017-08-30 16:38:33 +0000 |
|---|---|---|
| committer | Craig Topper <craig.topper@intel.com> | 2017-08-30 16:38:33 +0000 |
| commit | afce0baacd6a10e891518d3f584600de366a5e0b (patch) | |
| tree | 6f795088f69bd2f3fcfc8edded12ff031c274515 /llvm/test/CodeGen/X86/vector-bitreverse.ll | |
| parent | 89e8d5e9557d29148c9ff118b0e5eb9e6a3a5277 (diff) | |
| download | bcm5719-llvm-afce0baacd6a10e891518d3f584600de366a5e0b.tar.gz bcm5719-llvm-afce0baacd6a10e891518d3f584600de366a5e0b.zip | |
[AVX512] Don't use 32-bit elements version of AND/OR/XOR/ANDN during isel unless we're matching a masked op or broadcast
Selecting 32-bit element logical ops without a select or broadcast requires matching a bitconvert on the inputs to the and. But that's a weird thing to rely on. It's entirely possible that one of the inputs doesn't have a bitcast and one does.
Since there's no functional difference, just remove the extra patterns and save some isel table size.
Differential Revision: https://reviews.llvm.org/D36854
llvm-svn: 312138
Diffstat (limited to 'llvm/test/CodeGen/X86/vector-bitreverse.ll')
| -rw-r--r-- | llvm/test/CodeGen/X86/vector-bitreverse.ll | 6 |
1 files changed, 3 insertions, 3 deletions
diff --git a/llvm/test/CodeGen/X86/vector-bitreverse.ll b/llvm/test/CodeGen/X86/vector-bitreverse.ll index af91d35a6ee..485911280c6 100644 --- a/llvm/test/CodeGen/X86/vector-bitreverse.ll +++ b/llvm/test/CodeGen/X86/vector-bitreverse.ll @@ -2051,17 +2051,17 @@ define <16 x i32> @test_bitreverse_v16i32(<16 x i32> %a) nounwind { ; AVX512F-NEXT: vpslld $4, %zmm1, %zmm1 ; AVX512F-NEXT: vpandd {{.*}}(%rip){1to16}, %zmm0, %zmm0 ; AVX512F-NEXT: vpsrld $4, %zmm0, %zmm0 -; AVX512F-NEXT: vpord %zmm1, %zmm0, %zmm0 +; AVX512F-NEXT: vporq %zmm1, %zmm0, %zmm0 ; AVX512F-NEXT: vpandd {{.*}}(%rip){1to16}, %zmm0, %zmm1 ; AVX512F-NEXT: vpslld $2, %zmm1, %zmm1 ; AVX512F-NEXT: vpandd {{.*}}(%rip){1to16}, %zmm0, %zmm0 ; AVX512F-NEXT: vpsrld $2, %zmm0, %zmm0 -; AVX512F-NEXT: vpord %zmm1, %zmm0, %zmm0 +; AVX512F-NEXT: vporq %zmm1, %zmm0, %zmm0 ; AVX512F-NEXT: vpandd {{.*}}(%rip){1to16}, %zmm0, %zmm1 ; AVX512F-NEXT: vpslld $1, %zmm1, %zmm1 ; AVX512F-NEXT: vpandd {{.*}}(%rip){1to16}, %zmm0, %zmm0 ; AVX512F-NEXT: vpsrld $1, %zmm0, %zmm0 -; AVX512F-NEXT: vpord %zmm1, %zmm0, %zmm0 +; AVX512F-NEXT: vporq %zmm1, %zmm0, %zmm0 ; AVX512F-NEXT: retq ; ; AVX512BW-LABEL: test_bitreverse_v16i32: |
