diff --git a/test/CodeGen/X86/vector-pcmp.ll b/test/CodeGen/X86/vector-pcmp.ll index 57d955d0051..4c27b4b4499 100644 --- a/test/CodeGen/X86/vector-pcmp.ll +++ b/test/CodeGen/X86/vector-pcmp.ll @@ -128,3 +128,139 @@ define <1 x i128> @test_strange_type(<1 x i128> %x) { ret <1 x i128> %not } +define <32 x i8> @test_pcmpgtb_256(<32 x i8> %x) { +; SSE-LABEL: test_pcmpgtb_256: +; SSE: # BB#0: +; SSE-NEXT: pcmpeqd %xmm2, %xmm2 +; SSE-NEXT: pcmpgtb %xmm2, %xmm0 +; SSE-NEXT: pcmpgtb %xmm2, %xmm1 +; SSE-NEXT: retq +; +; AVX1-LABEL: test_pcmpgtb_256: +; AVX1: # BB#0: +; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm1 +; AVX1-NEXT: vpxor %xmm2, %xmm2, %xmm2 +; AVX1-NEXT: vpcmpgtb %xmm1, %xmm2, %xmm1 +; AVX1-NEXT: vpcmpgtb %xmm0, %xmm2, %xmm0 +; AVX1-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0 +; AVX1-NEXT: vpcmpeqd %xmm1, %xmm1, %xmm1 +; AVX1-NEXT: vinsertf128 $1, %xmm1, %ymm1, %ymm1 +; AVX1-NEXT: vxorps %ymm1, %ymm0, %ymm0 +; AVX1-NEXT: retq +; +; AVX2-LABEL: test_pcmpgtb_256: +; AVX2: # BB#0: +; AVX2-NEXT: vpxor %ymm1, %ymm1, %ymm1 +; AVX2-NEXT: vpcmpgtb %ymm0, %ymm1, %ymm0 +; AVX2-NEXT: vpcmpeqd %ymm1, %ymm1, %ymm1 +; AVX2-NEXT: vpxor %ymm1, %ymm0, %ymm0 +; AVX2-NEXT: retq + %sign = ashr <32 x i8> %x, + %not = xor <32 x i8> %sign, + ret <32 x i8> %not +} + +define <16 x i16> @test_pcmpgtw_256(<16 x i16> %x) { +; SSE-LABEL: test_pcmpgtw_256: +; SSE: # BB#0: +; SSE-NEXT: pcmpeqd %xmm2, %xmm2 +; SSE-NEXT: pcmpgtw %xmm2, %xmm0 +; SSE-NEXT: pcmpgtw %xmm2, %xmm1 +; SSE-NEXT: retq +; +; AVX1-LABEL: test_pcmpgtw_256: +; AVX1: # BB#0: +; AVX1-NEXT: vpsraw $15, %xmm0, %xmm1 +; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm0 +; AVX1-NEXT: vpsraw $15, %xmm0, %xmm0 +; AVX1-NEXT: vinsertf128 $1, %xmm0, %ymm1, %ymm0 +; AVX1-NEXT: vpcmpeqd %xmm1, %xmm1, %xmm1 +; AVX1-NEXT: vinsertf128 $1, %xmm1, %ymm1, %ymm1 +; AVX1-NEXT: vxorps %ymm1, %ymm0, %ymm0 +; AVX1-NEXT: retq +; +; AVX2-LABEL: test_pcmpgtw_256: +; AVX2: # BB#0: +; AVX2-NEXT: vpsraw $15, %ymm0, %ymm0 +; AVX2-NEXT: vpcmpeqd %ymm1, %ymm1, %ymm1 +; AVX2-NEXT: vpxor %ymm1, %ymm0, %ymm0 +; AVX2-NEXT: retq + %sign = ashr <16 x i16> %x, + %not = xor <16 x i16> %sign, + ret <16 x i16> %not +} + +define <8 x i32> @test_pcmpgtd_256(<8 x i32> %x) { +; SSE-LABEL: test_pcmpgtd_256: +; SSE: # BB#0: +; SSE-NEXT: pcmpeqd %xmm2, %xmm2 +; SSE-NEXT: pcmpgtd %xmm2, %xmm0 +; SSE-NEXT: pcmpgtd %xmm2, %xmm1 +; SSE-NEXT: retq +; +; AVX1-LABEL: test_pcmpgtd_256: +; AVX1: # BB#0: +; AVX1-NEXT: vpsrad $31, %xmm0, %xmm1 +; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm0 +; AVX1-NEXT: vpsrad $31, %xmm0, %xmm0 +; AVX1-NEXT: vinsertf128 $1, %xmm0, %ymm1, %ymm0 +; AVX1-NEXT: vpcmpeqd %xmm1, %xmm1, %xmm1 +; AVX1-NEXT: vinsertf128 $1, %xmm1, %ymm1, %ymm1 +; AVX1-NEXT: vxorps %ymm1, %ymm0, %ymm0 +; AVX1-NEXT: retq +; +; AVX2-LABEL: test_pcmpgtd_256: +; AVX2: # BB#0: +; AVX2-NEXT: vpsrad $31, %ymm0, %ymm0 +; AVX2-NEXT: vpcmpeqd %ymm1, %ymm1, %ymm1 +; AVX2-NEXT: vpxor %ymm1, %ymm0, %ymm0 +; AVX2-NEXT: retq + %sign = ashr <8 x i32> %x, + %not = xor <8 x i32> %sign, + ret <8 x i32> %not +} + +define <4 x i64> @test_pcmpgtq_256(<4 x i64> %x) { +; SSE2-LABEL: test_pcmpgtq_256: +; SSE2: # BB#0: +; SSE2-NEXT: psrad $31, %xmm1 +; SSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm1[1,1,3,3] +; SSE2-NEXT: psrad $31, %xmm0 +; SSE2-NEXT: pshufd {{.*#+}} xmm0 = xmm0[1,1,3,3] +; SSE2-NEXT: pcmpeqd %xmm2, %xmm2 +; SSE2-NEXT: pxor %xmm2, %xmm0 +; SSE2-NEXT: pxor %xmm2, %xmm1 +; SSE2-NEXT: retq +; +; SSE42-LABEL: test_pcmpgtq_256: +; SSE42: # BB#0: +; SSE42-NEXT: pcmpeqd %xmm2, %xmm2 +; SSE42-NEXT: pcmpgtq %xmm2, %xmm0 +; SSE42-NEXT: pcmpgtq %xmm2, %xmm1 +; SSE42-NEXT: retq +; +; AVX1-LABEL: test_pcmpgtq_256: +; AVX1: # BB#0: +; AVX1-NEXT: vpsrad $31, %xmm0, %xmm1 +; AVX1-NEXT: vpshufd {{.*#+}} xmm1 = xmm1[1,1,3,3] +; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm0 +; AVX1-NEXT: vpsrad $31, %xmm0, %xmm0 +; AVX1-NEXT: vpshufd {{.*#+}} xmm0 = xmm0[1,1,3,3] +; AVX1-NEXT: vinsertf128 $1, %xmm0, %ymm1, %ymm0 +; AVX1-NEXT: vpcmpeqd %xmm1, %xmm1, %xmm1 +; AVX1-NEXT: vinsertf128 $1, %xmm1, %ymm1, %ymm1 +; AVX1-NEXT: vxorps %ymm1, %ymm0, %ymm0 +; AVX1-NEXT: retq +; +; AVX2-LABEL: test_pcmpgtq_256: +; AVX2: # BB#0: +; AVX2-NEXT: vpsrad $31, %ymm0, %ymm0 +; AVX2-NEXT: vpshufd {{.*#+}} ymm0 = ymm0[1,1,3,3,5,5,7,7] +; AVX2-NEXT: vpcmpeqd %ymm1, %ymm1, %ymm1 +; AVX2-NEXT: vpxor %ymm1, %ymm0, %ymm0 +; AVX2-NEXT: retq + %sign = ashr <4 x i64> %x, + %not = xor <4 x i64> %sign, + ret <4 x i64> %not +} +