+declare <4 x i32> @llvm.x86.avx512.mask.pmovsxb.d.128(<16 x i8>, <4 x i32>, i8)
+
+define <4 x i32>@test_int_x86_avx512_mask_pmovsxb_d_128(<16 x i8> %x0, <4 x i32> %x1, i8 %x2) {
+; CHECK-LABEL: test_int_x86_avx512_mask_pmovsxb_d_128:
+; CHECK: ## BB#0:
+; CHECK-NEXT: movzbl %dil, %eax
+; CHECK-NEXT: kmovw %eax, %k1
+; CHECK-NEXT: vpmovsxbd %xmm0, %xmm1 {%k1}
+; CHECK-NEXT: vpmovsxbd %xmm0, %xmm2 {%k1} {z}
+; CHECK-NEXT: vpmovsxbd %xmm0, %xmm0
+; CHECK-NEXT: vpaddd %xmm2, %xmm1, %xmm1
+; CHECK-NEXT: vpaddd %xmm0, %xmm1, %xmm0
+; CHECK-NEXT: retq
+ %res = call <4 x i32> @llvm.x86.avx512.mask.pmovsxb.d.128(<16 x i8> %x0, <4 x i32> %x1, i8 %x2)
+ %res1 = call <4 x i32> @llvm.x86.avx512.mask.pmovsxb.d.128(<16 x i8> %x0, <4 x i32> zeroinitializer, i8 %x2)
+ %res2 = call <4 x i32> @llvm.x86.avx512.mask.pmovsxb.d.128(<16 x i8> %x0, <4 x i32> %x1, 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.pmovsxb.d.256(<16 x i8>, <8 x i32>, i8)
+
+define <8 x i32>@test_int_x86_avx512_mask_pmovsxb_d_256(<16 x i8> %x0, <8 x i32> %x1, i8 %x2) {
+; CHECK-LABEL: test_int_x86_avx512_mask_pmovsxb_d_256:
+; CHECK: ## BB#0:
+; CHECK-NEXT: movzbl %dil, %eax
+; CHECK-NEXT: kmovw %eax, %k1
+; CHECK-NEXT: vpmovsxbd %xmm0, %ymm1 {%k1}
+; CHECK-NEXT: vpmovsxbd %xmm0, %ymm2 {%k1} {z}
+; CHECK-NEXT: vpmovsxbd %xmm0, %ymm0
+; CHECK-NEXT: vpaddd %ymm2, %ymm1, %ymm1
+; CHECK-NEXT: vpaddd %ymm0, %ymm1, %ymm0
+; CHECK-NEXT: retq
+ %res = call <8 x i32> @llvm.x86.avx512.mask.pmovsxb.d.256(<16 x i8> %x0, <8 x i32> %x1, i8 %x2)
+ %res1 = call <8 x i32> @llvm.x86.avx512.mask.pmovsxb.d.256(<16 x i8> %x0, <8 x i32> zeroinitializer, i8 %x2)
+ %res2 = call <8 x i32> @llvm.x86.avx512.mask.pmovsxb.d.256(<16 x i8> %x0, <8 x i32> %x1, i8 -1)
+ %res3 = add <8 x i32> %res, %res1
+ %res4 = add <8 x i32> %res3, %res2
+ ret <8 x i32> %res4
+}
+
+declare <2 x i64> @llvm.x86.avx512.mask.pmovsxb.q.128(<16 x i8>, <2 x i64>, i8)
+
+define <2 x i64>@test_int_x86_avx512_mask_pmovsxb_q_128(<16 x i8> %x0, <2 x i64> %x1, i8 %x2) {
+; CHECK-LABEL: test_int_x86_avx512_mask_pmovsxb_q_128:
+; CHECK: ## BB#0:
+; CHECK-NEXT: movzbl %dil, %eax
+; CHECK-NEXT: kmovw %eax, %k1
+; CHECK-NEXT: vpmovsxbq %xmm0, %xmm1 {%k1}
+; CHECK-NEXT: vpmovsxbq %xmm0, %xmm2 {%k1} {z}
+; CHECK-NEXT: vpmovsxbq %xmm0, %xmm0
+; CHECK-NEXT: vpaddq %xmm2, %xmm1, %xmm1
+; CHECK-NEXT: vpaddq %xmm0, %xmm1, %xmm0
+; CHECK-NEXT: retq
+ %res = call <2 x i64> @llvm.x86.avx512.mask.pmovsxb.q.128(<16 x i8> %x0, <2 x i64> %x1, i8 %x2)
+ %res1 = call <2 x i64> @llvm.x86.avx512.mask.pmovsxb.q.128(<16 x i8> %x0, <2 x i64> zeroinitializer, i8 %x2)
+ %res2 = call <2 x i64> @llvm.x86.avx512.mask.pmovsxb.q.128(<16 x i8> %x0, <2 x i64> %x1, 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.pmovsxb.q.256(<16 x i8>, <4 x i64>, i8)
+
+define <4 x i64>@test_int_x86_avx512_mask_pmovsxb_q_256(<16 x i8> %x0, <4 x i64> %x1, i8 %x2) {
+; CHECK-LABEL: test_int_x86_avx512_mask_pmovsxb_q_256:
+; CHECK: ## BB#0:
+; CHECK-NEXT: movzbl %dil, %eax
+; CHECK-NEXT: kmovw %eax, %k1
+; CHECK-NEXT: vpmovsxbq %xmm0, %ymm1 {%k1}
+; CHECK-NEXT: vpmovsxbq %xmm0, %ymm2 {%k1} {z}
+; CHECK-NEXT: vpmovsxbq %xmm0, %ymm0
+; CHECK-NEXT: vpaddq %ymm2, %ymm1, %ymm1
+; CHECK-NEXT: vpaddq %ymm0, %ymm1, %ymm0
+; CHECK-NEXT: retq
+ %res = call <4 x i64> @llvm.x86.avx512.mask.pmovsxb.q.256(<16 x i8> %x0, <4 x i64> %x1, i8 %x2)
+ %res1 = call <4 x i64> @llvm.x86.avx512.mask.pmovsxb.q.256(<16 x i8> %x0, <4 x i64> zeroinitializer, i8 %x2)
+ %res2 = call <4 x i64> @llvm.x86.avx512.mask.pmovsxb.q.256(<16 x i8> %x0, <4 x i64> %x1, 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.pmovsxw.d.128(<8 x i16>, <4 x i32>, i8)
+
+define <4 x i32>@test_int_x86_avx512_mask_pmovsxw_d_128(<8 x i16> %x0, <4 x i32> %x1, i8 %x2) {
+; CHECK-LABEL: test_int_x86_avx512_mask_pmovsxw_d_128:
+; CHECK: ## BB#0:
+; CHECK-NEXT: movzbl %dil, %eax
+; CHECK-NEXT: kmovw %eax, %k1
+; CHECK-NEXT: vpmovsxwd %xmm0, %xmm1 {%k1}
+; CHECK-NEXT: vpmovsxwd %xmm0, %xmm2 {%k1} {z}
+; CHECK-NEXT: vpmovsxwd %xmm0, %xmm0
+; CHECK-NEXT: vpaddd %xmm2, %xmm1, %xmm1
+; CHECK-NEXT: vpaddd %xmm0, %xmm1, %xmm0
+; CHECK-NEXT: retq
+ %res = call <4 x i32> @llvm.x86.avx512.mask.pmovsxw.d.128(<8 x i16> %x0, <4 x i32> %x1, i8 %x2)
+ %res1 = call <4 x i32> @llvm.x86.avx512.mask.pmovsxw.d.128(<8 x i16> %x0, <4 x i32> zeroinitializer, i8 %x2)
+ %res2 = call <4 x i32> @llvm.x86.avx512.mask.pmovsxw.d.128(<8 x i16> %x0, <4 x i32> %x1, 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.pmovsxw.d.256(<8 x i16>, <8 x i32>, i8)
+
+define <8 x i32>@test_int_x86_avx512_mask_pmovsxw_d_256(<8 x i16> %x0, <8 x i32> %x1, i8 %x2) {
+; CHECK-LABEL: test_int_x86_avx512_mask_pmovsxw_d_256:
+; CHECK: ## BB#0:
+; CHECK-NEXT: movzbl %dil, %eax
+; CHECK-NEXT: kmovw %eax, %k1
+; CHECK-NEXT: vpmovsxwd %xmm0, %ymm1 {%k1}
+; CHECK-NEXT: vpmovsxwd %xmm0, %ymm2 {%k1} {z}
+; CHECK-NEXT: vpmovsxwd %xmm0, %ymm0
+; CHECK-NEXT: vpaddd %ymm2, %ymm1, %ymm1
+; CHECK-NEXT: vpaddd %ymm0, %ymm1, %ymm0
+; CHECK-NEXT: retq
+ %res = call <8 x i32> @llvm.x86.avx512.mask.pmovsxw.d.256(<8 x i16> %x0, <8 x i32> %x1, i8 %x2)
+ %res1 = call <8 x i32> @llvm.x86.avx512.mask.pmovsxw.d.256(<8 x i16> %x0, <8 x i32> zeroinitializer, i8 %x2)
+ %res2 = call <8 x i32> @llvm.x86.avx512.mask.pmovsxw.d.256(<8 x i16> %x0, <8 x i32> %x1, i8 -1)
+ %res3 = add <8 x i32> %res, %res1
+ %res4 = add <8 x i32> %res3, %res2
+ ret <8 x i32> %res4
+}
+
+declare <2 x i64> @llvm.x86.avx512.mask.pmovsxw.q.128(<8 x i16>, <2 x i64>, i8)
+
+define <2 x i64>@test_int_x86_avx512_mask_pmovsxw_q_128(<8 x i16> %x0, <2 x i64> %x1, i8 %x2) {
+; CHECK-LABEL: test_int_x86_avx512_mask_pmovsxw_q_128:
+; CHECK: ## BB#0:
+; CHECK-NEXT: movzbl %dil, %eax
+; CHECK-NEXT: kmovw %eax, %k1
+; CHECK-NEXT: vpmovsxwq %xmm0, %xmm1 {%k1}
+; CHECK-NEXT: vpmovsxwq %xmm0, %xmm2 {%k1} {z}
+; CHECK-NEXT: vpmovsxwq %xmm0, %xmm0
+; CHECK-NEXT: vpaddq %xmm2, %xmm1, %xmm1
+; CHECK-NEXT: vpaddq %xmm0, %xmm1, %xmm0
+; CHECK-NEXT: retq
+ %res = call <2 x i64> @llvm.x86.avx512.mask.pmovsxw.q.128(<8 x i16> %x0, <2 x i64> %x1, i8 %x2)
+ %res1 = call <2 x i64> @llvm.x86.avx512.mask.pmovsxw.q.128(<8 x i16> %x0, <2 x i64> zeroinitializer, i8 %x2)
+ %res2 = call <2 x i64> @llvm.x86.avx512.mask.pmovsxw.q.128(<8 x i16> %x0, <2 x i64> %x1, 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.pmovsxw.q.256(<8 x i16>, <4 x i64>, i8)
+
+define <4 x i64>@test_int_x86_avx512_mask_pmovsxw_q_256(<8 x i16> %x0, <4 x i64> %x1, i8 %x2) {
+; CHECK-LABEL: test_int_x86_avx512_mask_pmovsxw_q_256:
+; CHECK: ## BB#0:
+; CHECK-NEXT: movzbl %dil, %eax
+; CHECK-NEXT: kmovw %eax, %k1
+; CHECK-NEXT: vpmovsxwq %xmm0, %ymm1 {%k1}
+; CHECK-NEXT: vpmovsxwq %xmm0, %ymm2 {%k1} {z}
+; CHECK-NEXT: vpmovsxwq %xmm0, %ymm0
+; CHECK-NEXT: vpaddq %ymm2, %ymm1, %ymm1
+; CHECK-NEXT: vpaddq %ymm0, %ymm1, %ymm0
+; CHECK-NEXT: retq
+ %res = call <4 x i64> @llvm.x86.avx512.mask.pmovsxw.q.256(<8 x i16> %x0, <4 x i64> %x1, i8 %x2)
+ %res1 = call <4 x i64> @llvm.x86.avx512.mask.pmovsxw.q.256(<8 x i16> %x0, <4 x i64> zeroinitializer, i8 %x2)
+ %res2 = call <4 x i64> @llvm.x86.avx512.mask.pmovsxw.q.256(<8 x i16> %x0, <4 x i64> %x1, i8 -1)
+ %res3 = add <4 x i64> %res, %res1
+ %res4 = add <4 x i64> %res3, %res2
+ ret <4 x i64> %res4
+}