diff options
Diffstat (limited to 'test/CodeGen/X86/vector-popcnt-256.ll')
| -rw-r--r-- | test/CodeGen/X86/vector-popcnt-256.ll | 42 |
1 files changed, 42 insertions, 0 deletions
diff --git a/test/CodeGen/X86/vector-popcnt-256.ll b/test/CodeGen/X86/vector-popcnt-256.ll index 7a675619d720e..accbad35e9d72 100644 --- a/test/CodeGen/X86/vector-popcnt-256.ll +++ b/test/CodeGen/X86/vector-popcnt-256.ll @@ -1,6 +1,7 @@ ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py ; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx | FileCheck %s --check-prefix=ALL --check-prefix=AVX --check-prefix=AVX1 ; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx2 | FileCheck %s --check-prefix=ALL --check-prefix=AVX --check-prefix=AVX2 +; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx512vpopcntdq | FileCheck %s --check-prefix=ALL --check-prefix=AVX --check-prefix=AVX512VPOPCNTDQ define <4 x i64> @testv4i64(<4 x i64> %in) nounwind { ; AVX1-LABEL: testv4i64: @@ -39,6 +40,13 @@ define <4 x i64> @testv4i64(<4 x i64> %in) nounwind { ; AVX2-NEXT: vpxor %ymm1, %ymm1, %ymm1 ; AVX2-NEXT: vpsadbw %ymm1, %ymm0, %ymm0 ; AVX2-NEXT: retq +; +; AVX512VPOPCNTDQ-LABEL: testv4i64: +; AVX512VPOPCNTDQ: # BB#0: +; AVX512VPOPCNTDQ-NEXT: # kill: %YMM0<def> %YMM0<kill> %ZMM0<def> +; AVX512VPOPCNTDQ-NEXT: vpopcntq %zmm0, %zmm0 +; AVX512VPOPCNTDQ-NEXT: # kill: %YMM0<def> %YMM0<kill> %ZMM0<kill> +; AVX512VPOPCNTDQ-NEXT: retq %out = call <4 x i64> @llvm.ctpop.v4i64(<4 x i64> %in) ret <4 x i64> %out } @@ -92,6 +100,13 @@ define <8 x i32> @testv8i32(<8 x i32> %in) nounwind { ; AVX2-NEXT: vpsadbw %ymm1, %ymm0, %ymm0 ; AVX2-NEXT: vpackuswb %ymm2, %ymm0, %ymm0 ; AVX2-NEXT: retq +; +; AVX512VPOPCNTDQ-LABEL: testv8i32: +; AVX512VPOPCNTDQ: # BB#0: +; AVX512VPOPCNTDQ-NEXT: # kill: %YMM0<def> %YMM0<kill> %ZMM0<def> +; AVX512VPOPCNTDQ-NEXT: vpopcntd %zmm0, %zmm0 +; AVX512VPOPCNTDQ-NEXT: # kill: %YMM0<def> %YMM0<kill> %ZMM0<kill> +; AVX512VPOPCNTDQ-NEXT: retq %out = call <8 x i32> @llvm.ctpop.v8i32(<8 x i32> %in) ret <8 x i32> %out } @@ -137,6 +152,21 @@ define <16 x i16> @testv16i16(<16 x i16> %in) nounwind { ; AVX2-NEXT: vpaddb %ymm0, %ymm1, %ymm0 ; AVX2-NEXT: vpsrlw $8, %ymm0, %ymm0 ; AVX2-NEXT: retq +; +; AVX512VPOPCNTDQ-LABEL: testv16i16: +; AVX512VPOPCNTDQ: # BB#0: +; AVX512VPOPCNTDQ-NEXT: vmovdqa {{.*#+}} ymm1 = [15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15] +; AVX512VPOPCNTDQ-NEXT: vpand %ymm1, %ymm0, %ymm2 +; AVX512VPOPCNTDQ-NEXT: vmovdqa {{.*#+}} ymm3 = [0,1,1,2,1,2,2,3,1,2,2,3,2,3,3,4,0,1,1,2,1,2,2,3,1,2,2,3,2,3,3,4] +; AVX512VPOPCNTDQ-NEXT: vpshufb %ymm2, %ymm3, %ymm2 +; AVX512VPOPCNTDQ-NEXT: vpsrlw $4, %ymm0, %ymm0 +; AVX512VPOPCNTDQ-NEXT: vpand %ymm1, %ymm0, %ymm0 +; AVX512VPOPCNTDQ-NEXT: vpshufb %ymm0, %ymm3, %ymm0 +; AVX512VPOPCNTDQ-NEXT: vpaddb %ymm2, %ymm0, %ymm0 +; AVX512VPOPCNTDQ-NEXT: vpsllw $8, %ymm0, %ymm1 +; AVX512VPOPCNTDQ-NEXT: vpaddb %ymm0, %ymm1, %ymm0 +; AVX512VPOPCNTDQ-NEXT: vpsrlw $8, %ymm0, %ymm0 +; AVX512VPOPCNTDQ-NEXT: retq %out = call <16 x i16> @llvm.ctpop.v16i16(<16 x i16> %in) ret <16 x i16> %out } @@ -173,6 +203,18 @@ define <32 x i8> @testv32i8(<32 x i8> %in) nounwind { ; AVX2-NEXT: vpshufb %ymm0, %ymm3, %ymm0 ; AVX2-NEXT: vpaddb %ymm2, %ymm0, %ymm0 ; AVX2-NEXT: retq +; +; AVX512VPOPCNTDQ-LABEL: testv32i8: +; AVX512VPOPCNTDQ: # BB#0: +; AVX512VPOPCNTDQ-NEXT: vmovdqa {{.*#+}} ymm1 = [15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15,15] +; AVX512VPOPCNTDQ-NEXT: vpand %ymm1, %ymm0, %ymm2 +; AVX512VPOPCNTDQ-NEXT: vmovdqa {{.*#+}} ymm3 = [0,1,1,2,1,2,2,3,1,2,2,3,2,3,3,4,0,1,1,2,1,2,2,3,1,2,2,3,2,3,3,4] +; AVX512VPOPCNTDQ-NEXT: vpshufb %ymm2, %ymm3, %ymm2 +; AVX512VPOPCNTDQ-NEXT: vpsrlw $4, %ymm0, %ymm0 +; AVX512VPOPCNTDQ-NEXT: vpand %ymm1, %ymm0, %ymm0 +; AVX512VPOPCNTDQ-NEXT: vpshufb %ymm0, %ymm3, %ymm0 +; AVX512VPOPCNTDQ-NEXT: vpaddb %ymm2, %ymm0, %ymm0 +; AVX512VPOPCNTDQ-NEXT: retq %out = call <32 x i8> @llvm.ctpop.v32i8(<32 x i8> %in) ret <32 x i8> %out } |