Index: include/llvm/IR/IntrinsicsX86.td =================================================================== --- include/llvm/IR/IntrinsicsX86.td +++ include/llvm/IR/IntrinsicsX86.td @@ -2741,6 +2741,27 @@ def int_x86_avx512_psrl_dq_512 : GCCBuiltin<"__builtin_ia32_psrldq512">, Intrinsic<[llvm_v8i64_ty], [llvm_v8i64_ty, llvm_i32_ty], [IntrNoMem]>; + def int_x86_avx512_mask_psrlv16_hi : GCCBuiltin<"__builtin_ia32_psrlv16hi_mask">, + Intrinsic<[llvm_v16i16_ty], [llvm_v16i16_ty, + llvm_v16i16_ty, llvm_v16i16_ty, llvm_i16_ty], [IntrNoMem]>; + def int_x86_avx512_mask_psrlv2_di : GCCBuiltin<"__builtin_ia32_psrlv2di_mask">, + Intrinsic<[llvm_v2i64_ty], [llvm_v2i64_ty, + llvm_v2i64_ty, llvm_v2i64_ty, llvm_i8_ty], [IntrNoMem]>; + def int_x86_avx512_mask_psrlv32hi : GCCBuiltin<"__builtin_ia32_psrlv32hi_mask">, + Intrinsic<[llvm_v32i16_ty], [llvm_v32i16_ty, + llvm_v32i16_ty, llvm_v32i16_ty, llvm_i32_ty], [IntrNoMem]>; + def int_x86_avx512_mask_psrlv4_di : GCCBuiltin<"__builtin_ia32_psrlv4di_mask">, + Intrinsic<[llvm_v4i64_ty], [llvm_v4i64_ty, + llvm_v4i64_ty, llvm_v4i64_ty, llvm_i8_ty], [IntrNoMem]>; + def int_x86_avx512_mask_psrlv4_si : GCCBuiltin<"__builtin_ia32_psrlv4si_mask">, + Intrinsic<[llvm_v4i32_ty], [llvm_v4i32_ty, + llvm_v4i32_ty, llvm_v4i32_ty, llvm_i8_ty], [IntrNoMem]>; + def int_x86_avx512_mask_psrlv8_hi : GCCBuiltin<"__builtin_ia32_psrlv8hi_mask">, + Intrinsic<[llvm_v8i16_ty], [llvm_v8i16_ty, + llvm_v8i16_ty, llvm_v8i16_ty, llvm_i8_ty], [IntrNoMem]>; + def int_x86_avx512_mask_psrlv8_si : GCCBuiltin<"__builtin_ia32_psrlv8si_mask">, + Intrinsic<[llvm_v8i32_ty], [llvm_v8i32_ty, + llvm_v8i32_ty, llvm_v8i32_ty, llvm_i8_ty], [IntrNoMem]>; } // Gather ops Index: lib/Target/X86/X86IntrinsicsInfo.h =================================================================== --- lib/Target/X86/X86IntrinsicsInfo.h +++ lib/Target/X86/X86IntrinsicsInfo.h @@ -1239,6 +1239,13 @@ X86_INTRINSIC_DATA(avx512_mask_psrl_wi_512, INTR_TYPE_2OP_MASK, X86ISD::VSRLI, 0), X86_INTRINSIC_DATA(avx512_mask_psrli_d, VSHIFT_MASK, X86ISD::VSRLI, 0), X86_INTRINSIC_DATA(avx512_mask_psrli_q, VSHIFT_MASK, X86ISD::VSRLI, 0), + X86_INTRINSIC_DATA(avx512_mask_psrlv16_hi, INTR_TYPE_2OP_MASK, ISD::SRL, 0), + X86_INTRINSIC_DATA(avx512_mask_psrlv2_di, INTR_TYPE_2OP_MASK, ISD::SRL, 0), + X86_INTRINSIC_DATA(avx512_mask_psrlv32hi, INTR_TYPE_2OP_MASK, ISD::SRL, 0), + X86_INTRINSIC_DATA(avx512_mask_psrlv4_di, INTR_TYPE_2OP_MASK, ISD::SRL, 0), + X86_INTRINSIC_DATA(avx512_mask_psrlv4_si, INTR_TYPE_2OP_MASK, ISD::SRL, 0), + X86_INTRINSIC_DATA(avx512_mask_psrlv8_hi, INTR_TYPE_2OP_MASK, ISD::SRL, 0), + X86_INTRINSIC_DATA(avx512_mask_psrlv8_si, INTR_TYPE_2OP_MASK, ISD::SRL, 0), X86_INTRINSIC_DATA(avx512_mask_psrlv_d, INTR_TYPE_2OP_MASK, ISD::SRL, 0), X86_INTRINSIC_DATA(avx512_mask_psrlv_q, INTR_TYPE_2OP_MASK, ISD::SRL, 0), X86_INTRINSIC_DATA(avx512_mask_psub_b_128, INTR_TYPE_2OP_MASK, ISD::SUB, 0), Index: test/CodeGen/X86/avx512bw-intrinsics.ll =================================================================== --- test/CodeGen/X86/avx512bw-intrinsics.ll +++ test/CodeGen/X86/avx512bw-intrinsics.ll @@ -2906,3 +2906,24 @@ %res4 = add <32 x i16> %res3, %res2 ret <32 x i16> %res4 } + +declare <32 x i16> @llvm.x86.avx512.mask.psrlv32hi(<32 x i16>, <32 x i16>, <32 x i16>, i32) + +define <32 x i16>@test_int_x86_avx512_mask_psrlv32hi(<32 x i16> %x0, <32 x i16> %x1, <32 x i16> %x2, i32 %x3) { +; CHECK-LABEL: test_int_x86_avx512_mask_psrlv32hi: +; CHECK: ## BB#0: +; CHECK-NEXT: kmovd %edi, %k1 +; CHECK-NEXT: vpsrlvw %zmm1, %zmm0, %zmm2 {%k1} +; CHECK-NEXT: vpsrlvw %zmm1, %zmm0, %zmm3 {%k1} {z} +; CHECK-NEXT: vpsrlvw %zmm1, %zmm0, %zmm0 +; CHECK-NEXT: vpaddw %zmm3, %zmm2, %zmm1 +; CHECK-NEXT: vpaddw %zmm0, %zmm1, %zmm0 +; CHECK-NEXT: retq + %res = call <32 x i16> @llvm.x86.avx512.mask.psrlv32hi(<32 x i16> %x0, <32 x i16> %x1, <32 x i16> %x2, i32 %x3) + %res1 = call <32 x i16> @llvm.x86.avx512.mask.psrlv32hi(<32 x i16> %x0, <32 x i16> %x1, <32 x i16> zeroinitializer, i32 %x3) + %res2 = call <32 x i16> @llvm.x86.avx512.mask.psrlv32hi(<32 x i16> %x0, <32 x i16> %x1, <32 x i16> %x2, i32 -1) + %res3 = add <32 x i16> %res, %res1 + %res4 = add <32 x i16> %res3, %res2 + ret <32 x i16> %res4 +} + Index: test/CodeGen/X86/avx512bwvl-intrinsics.ll =================================================================== --- test/CodeGen/X86/avx512bwvl-intrinsics.ll +++ test/CodeGen/X86/avx512bwvl-intrinsics.ll @@ -4591,3 +4591,44 @@ %res4 = add <16 x i16> %res3, %res2 ret <16 x i16> %res4 } + +declare <16 x i16> @llvm.x86.avx512.mask.psrlv16.hi(<16 x i16>, <16 x i16>, <16 x i16>, i16) + +define <16 x i16>@test_int_x86_avx512_mask_psrlv16_hi(<16 x i16> %x0, <16 x i16> %x1, <16 x i16> %x2, i16 %x3) { +; CHECK-LABEL: test_int_x86_avx512_mask_psrlv16_hi: +; CHECK: ## BB#0: +; CHECK-NEXT: kmovw %edi, %k1 +; CHECK-NEXT: vpsrlvw %ymm1, %ymm0, %ymm2 {%k1} +; CHECK-NEXT: vpsrlvw %ymm1, %ymm0, %ymm3 {%k1} {z} +; CHECK-NEXT: vpsrlvw %ymm1, %ymm0, %ymm0 +; CHECK-NEXT: vpaddw %ymm3, %ymm2, %ymm1 +; CHECK-NEXT: vpaddw %ymm0, %ymm1, %ymm0 +; CHECK-NEXT: retq + %res = call <16 x i16> @llvm.x86.avx512.mask.psrlv16.hi(<16 x i16> %x0, <16 x i16> %x1, <16 x i16> %x2, i16 %x3) + %res1 = call <16 x i16> @llvm.x86.avx512.mask.psrlv16.hi(<16 x i16> %x0, <16 x i16> %x1, <16 x i16> zeroinitializer, i16 %x3) + %res2 = call <16 x i16> @llvm.x86.avx512.mask.psrlv16.hi(<16 x i16> %x0, <16 x i16> %x1, <16 x i16> %x2, i16 -1) + %res3 = add <16 x i16> %res, %res1 + %res4 = add <16 x i16> %res3, %res2 + ret <16 x i16> %res4 +} + +declare <8 x i16> @llvm.x86.avx512.mask.psrlv8.hi(<8 x i16>, <8 x i16>, <8 x i16>, i8) + +define <8 x i16>@test_int_x86_avx512_mask_psrlv8_hi(<8 x i16> %x0, <8 x i16> %x1, <8 x i16> %x2, i8 %x3) { +; CHECK-LABEL: test_int_x86_avx512_mask_psrlv8_hi: +; CHECK: ## BB#0: +; CHECK-NEXT: movzbl %dil, %eax +; CHECK-NEXT: kmovw %eax, %k1 +; CHECK-NEXT: vpsrlvw %xmm1, %xmm0, %xmm2 {%k1} +; CHECK-NEXT: vpsrlvw %xmm1, %xmm0, %xmm3 {%k1} {z} +; CHECK-NEXT: vpsrlvw %xmm1, %xmm0, %xmm0 +; CHECK-NEXT: vpaddw %xmm3, %xmm2, %xmm1 +; CHECK-NEXT: vpaddw %xmm0, %xmm1, %xmm0 +; CHECK-NEXT: retq + %res = call <8 x i16> @llvm.x86.avx512.mask.psrlv8.hi(<8 x i16> %x0, <8 x i16> %x1, <8 x i16> %x2, i8 %x3) + %res1 = call <8 x i16> @llvm.x86.avx512.mask.psrlv8.hi(<8 x i16> %x0, <8 x i16> %x1, <8 x i16> zeroinitializer, i8 %x3) + %res2 = call <8 x i16> @llvm.x86.avx512.mask.psrlv8.hi(<8 x i16> %x0, <8 x i16> %x1, <8 x i16> %x2, i8 -1) + %res3 = add <8 x i16> %res, %res1 + %res4 = add <8 x i16> %res3, %res2 + ret <8 x i16> %res4 +} Index: test/CodeGen/X86/avx512vl-intrinsics.ll =================================================================== --- test/CodeGen/X86/avx512vl-intrinsics.ll +++ test/CodeGen/X86/avx512vl-intrinsics.ll @@ -5987,3 +5987,87 @@ %res4 = add <16 x i32> %res2, %res3 ret <16 x i32> %res4 } + +declare <2 x i64> @llvm.x86.avx512.mask.psrlv2.di(<2 x i64>, <2 x i64>, <2 x i64>, i8) + +define <2 x i64>@test_int_x86_avx512_mask_psrlv2_di(<2 x i64> %x0, <2 x i64> %x1, <2 x i64> %x2, i8 %x3) { +; CHECK-LABEL: test_int_x86_avx512_mask_psrlv2_di: +; CHECK: ## BB#0: +; CHECK-NEXT: movzbl %dil, %eax +; CHECK-NEXT: kmovw %eax, %k1 +; CHECK-NEXT: vpsrlvq %xmm1, %xmm0, %xmm2 {%k1} +; CHECK-NEXT: vpsrlvq %xmm1, %xmm0, %xmm3 {%k1} {z} +; CHECK-NEXT: vpsrlvq %xmm1, %xmm0, %xmm0 +; CHECK-NEXT: vpaddq %xmm3, %xmm2, %xmm1 +; CHECK-NEXT: vpaddq %xmm0, %xmm1, %xmm0 +; CHECK-NEXT: retq + %res = call <2 x i64> @llvm.x86.avx512.mask.psrlv2.di(<2 x i64> %x0, <2 x i64> %x1, <2 x i64> %x2, i8 %x3) + %res1 = call <2 x i64> @llvm.x86.avx512.mask.psrlv2.di(<2 x i64> %x0, <2 x i64> %x1, <2 x i64> zeroinitializer, i8 %x3) + %res2 = call <2 x i64> @llvm.x86.avx512.mask.psrlv2.di(<2 x i64> %x0, <2 x i64> %x1, <2 x i64> %x2, i8 -1) + %res3 = add <2 x i64> %res, %res1 + %res4 = add <2 x i64> %res3, %res2 + ret <2 x i64> %res4 +} + +declare <4 x i64> @llvm.x86.avx512.mask.psrlv4.di(<4 x i64>, <4 x i64>, <4 x i64>, i8) + +define <4 x i64>@test_int_x86_avx512_mask_psrlv4_di(<4 x i64> %x0, <4 x i64> %x1, <4 x i64> %x2, i8 %x3) { +; CHECK-LABEL: test_int_x86_avx512_mask_psrlv4_di: +; CHECK: ## BB#0: +; CHECK-NEXT: movzbl %dil, %eax +; CHECK-NEXT: kmovw %eax, %k1 +; CHECK-NEXT: vpsrlvq %ymm1, %ymm0, %ymm2 {%k1} +; CHECK-NEXT: vpsrlvq %ymm1, %ymm0, %ymm3 {%k1} {z} +; CHECK-NEXT: vpsrlvq %ymm1, %ymm0, %ymm0 +; CHECK-NEXT: vpaddq %ymm3, %ymm2, %ymm1 +; CHECK-NEXT: vpaddq %ymm0, %ymm1, %ymm0 +; CHECK-NEXT: retq + %res = call <4 x i64> @llvm.x86.avx512.mask.psrlv4.di(<4 x i64> %x0, <4 x i64> %x1, <4 x i64> %x2, i8 %x3) + %res1 = call <4 x i64> @llvm.x86.avx512.mask.psrlv4.di(<4 x i64> %x0, <4 x i64> %x1, <4 x i64> zeroinitializer, i8 %x3) + %res2 = call <4 x i64> @llvm.x86.avx512.mask.psrlv4.di(<4 x i64> %x0, <4 x i64> %x1, <4 x i64> %x2, i8 -1) + %res3 = add <4 x i64> %res, %res1 + %res4 = add <4 x i64> %res3, %res2 + ret <4 x i64> %res4 +} + +declare <4 x i32> @llvm.x86.avx512.mask.psrlv4.si(<4 x i32>, <4 x i32>, <4 x i32>, i8) + +define <4 x i32>@test_int_x86_avx512_mask_psrlv4_si(<4 x i32> %x0, <4 x i32> %x1, <4 x i32> %x2, i8 %x3) { +; CHECK-LABEL: test_int_x86_avx512_mask_psrlv4_si: +; CHECK: ## BB#0: +; CHECK-NEXT: movzbl %dil, %eax +; CHECK-NEXT: kmovw %eax, %k1 +; CHECK-NEXT: vpsrlvd %xmm1, %xmm0, %xmm2 {%k1} +; CHECK-NEXT: vpsrlvd %xmm1, %xmm0, %xmm3 {%k1} {z} +; CHECK-NEXT: vpsrlvd %xmm1, %xmm0, %xmm0 +; CHECK-NEXT: vpaddd %xmm3, %xmm2, %xmm1 +; CHECK-NEXT: vpaddd %xmm0, %xmm1, %xmm0 +; CHECK-NEXT: retq + %res = call <4 x i32> @llvm.x86.avx512.mask.psrlv4.si(<4 x i32> %x0, <4 x i32> %x1, <4 x i32> %x2, i8 %x3) + %res1 = call <4 x i32> @llvm.x86.avx512.mask.psrlv4.si(<4 x i32> %x0, <4 x i32> %x1, <4 x i32> zeroinitializer, i8 %x3) + %res2 = call <4 x i32> @llvm.x86.avx512.mask.psrlv4.si(<4 x i32> %x0, <4 x i32> %x1, <4 x i32> %x2, i8 -1) + %res3 = add <4 x i32> %res, %res1 + %res4 = add <4 x i32> %res3, %res2 + ret <4 x i32> %res4 +} + +declare <8 x i32> @llvm.x86.avx512.mask.psrlv8.si(<8 x i32>, <8 x i32>, <8 x i32>, i8) + +define <8 x i32>@test_int_x86_avx512_mask_psrlv8_si(<8 x i32> %x0, <8 x i32> %x1, <8 x i32> %x2, i8 %x3) { +; CHECK-LABEL: test_int_x86_avx512_mask_psrlv8_si: +; CHECK: ## BB#0: +; CHECK-NEXT: movzbl %dil, %eax +; CHECK-NEXT: kmovw %eax, %k1 +; CHECK-NEXT: vpsrlvd %ymm1, %ymm0, %ymm2 {%k1} +; CHECK-NEXT: vpsrlvd %ymm1, %ymm0, %ymm3 {%k1} {z} +; CHECK-NEXT: vpsrlvd %ymm1, %ymm0, %ymm0 +; CHECK-NEXT: vpaddd %ymm3, %ymm2, %ymm1 +; CHECK-NEXT: vpaddd %ymm0, %ymm1, %ymm0 +; CHECK-NEXT: retq + %res = call <8 x i32> @llvm.x86.avx512.mask.psrlv8.si(<8 x i32> %x0, <8 x i32> %x1, <8 x i32> %x2, i8 %x3) + %res1 = call <8 x i32> @llvm.x86.avx512.mask.psrlv8.si(<8 x i32> %x0, <8 x i32> %x1, <8 x i32> zeroinitializer, i8 %x3) + %res2 = call <8 x i32> @llvm.x86.avx512.mask.psrlv8.si(<8 x i32> %x0, <8 x i32> %x1, <8 x i32> %x2, i8 -1) + %res3 = add <8 x i32> %res, %res1 + %res4 = add <8 x i32> %res3, %res2 + ret <8 x i32> %res4 +}