; 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=+xop,+avx | FileCheck %s --check-prefix=ALL --check-prefix=XOP --check-prefix=XOPAVX1
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+xop,+avx2 | FileCheck %s --check-prefix=ALL --check-prefix=XOP --check-prefix=XOPAVX2
+; RUN: llc < %s -mtriple=x86_64-apple-darwin -mcpu=knl -mattr=+avx512bw | FileCheck %s --check-prefix=ALL --check-prefix=AVX512 --check-prefix=AVX512BW
+
;
; Just one 32-bit run to make sure we do reasonable things for i64 shifts.
; RUN: llc < %s -mtriple=i686-unknown-unknown -mattr=+sse2 | FileCheck %s --check-prefix=ALL --check-prefix=X32-SSE --check-prefix=X32-SSE2
; XOP-NEXT: vpshaq %xmm1, %xmm0, %xmm0
; XOP-NEXT: retq
;
+; AVX512-LABEL: var_shift_v2i64:
+; AVX512: ## BB#0:
+; AVX512-NEXT: vmovdqa {{.*#+}} xmm2 = [9223372036854775808,9223372036854775808]
+; AVX512-NEXT: vpsrlvq %xmm1, %xmm2, %xmm3
+; AVX512-NEXT: vpxor %xmm2, %xmm0, %xmm0
+; AVX512-NEXT: vpsrlvq %xmm1, %xmm0, %xmm0
+; AVX512-NEXT: vpsubq %xmm3, %xmm0, %xmm0
+; AVX512-NEXT: retq
+;
; X32-SSE-LABEL: var_shift_v2i64:
; X32-SSE: # BB#0:
; X32-SSE-NEXT: pshufd {{.*#+}} xmm2 = xmm1[2,3,0,1]
; XOPAVX2-NEXT: vpsravd %xmm1, %xmm0, %xmm0
; XOPAVX2-NEXT: retq
;
+; AVX512-LABEL: var_shift_v4i32:
+; AVX512: ## BB#0:
+; AVX512-NEXT: vpsravd %xmm1, %xmm0, %xmm0
+; AVX512-NEXT: retq
+;
; X32-SSE-LABEL: var_shift_v4i32:
; X32-SSE: # BB#0:
; X32-SSE-NEXT: movdqa %xmm1, %xmm2
; XOP-NEXT: vpshaw %xmm1, %xmm0, %xmm0
; XOP-NEXT: retq
;
+; AVX512-LABEL: var_shift_v8i16:
+; AVX512: ## BB#0:
+; AVX512-NEXT: vpsravw %zmm1, %zmm0, %zmm0
+; AVX512-NEXT: retq
+;
; X32-SSE-LABEL: var_shift_v8i16:
; X32-SSE: # BB#0:
; X32-SSE-NEXT: psllw $12, %xmm1
; XOP-NEXT: vpshab %xmm1, %xmm0, %xmm0
; XOP-NEXT: retq
;
+; AVX512-LABEL: var_shift_v16i8:
+; AVX512: ## BB#0:
+; AVX512-NEXT: vpsllw $5, %xmm1, %xmm1
+; AVX512-NEXT: vpunpckhbw {{.*#+}} xmm2 = xmm0[8],xmm1[8],xmm0[9],xmm1[9],xmm0[10],xmm1[10],xmm0[11],xmm1[11],xmm0[12],xmm1[12],xmm0[13],xmm1[13],xmm0[14],xmm1[14],xmm0[15],xmm1[15]
+; AVX512-NEXT: vpunpckhbw {{.*#+}} xmm3 = xmm0[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15]
+; AVX512-NEXT: vpsraw $4, %xmm3, %xmm4
+; AVX512-NEXT: vpblendvb %xmm2, %xmm4, %xmm3, %xmm3
+; AVX512-NEXT: vpsraw $2, %xmm3, %xmm4
+; AVX512-NEXT: vpaddw %xmm2, %xmm2, %xmm2
+; AVX512-NEXT: vpblendvb %xmm2, %xmm4, %xmm3, %xmm3
+; AVX512-NEXT: vpsraw $1, %xmm3, %xmm4
+; AVX512-NEXT: vpaddw %xmm2, %xmm2, %xmm2
+; AVX512-NEXT: vpblendvb %xmm2, %xmm4, %xmm3, %xmm2
+; AVX512-NEXT: vpsrlw $8, %xmm2, %xmm2
+; AVX512-NEXT: vpunpcklbw {{.*#+}} xmm1 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3],xmm0[4],xmm1[4],xmm0[5],xmm1[5],xmm0[6],xmm1[6],xmm0[7],xmm1[7]
+; AVX512-NEXT: vpunpcklbw {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; AVX512-NEXT: vpsraw $4, %xmm0, %xmm3
+; AVX512-NEXT: vpblendvb %xmm1, %xmm3, %xmm0, %xmm0
+; AVX512-NEXT: vpsraw $2, %xmm0, %xmm3
+; AVX512-NEXT: vpaddw %xmm1, %xmm1, %xmm1
+; AVX512-NEXT: vpblendvb %xmm1, %xmm3, %xmm0, %xmm0
+; AVX512-NEXT: vpsraw $1, %xmm0, %xmm3
+; AVX512-NEXT: vpaddw %xmm1, %xmm1, %xmm1
+; AVX512-NEXT: vpblendvb %xmm1, %xmm3, %xmm0, %xmm0
+; AVX512-NEXT: vpsrlw $8, %xmm0, %xmm0
+; AVX512-NEXT: vpackuswb %xmm2, %xmm0, %xmm0
+; AVX512-NEXT: retq
+;
; X32-SSE-LABEL: var_shift_v16i8:
; X32-SSE: # BB#0:
; X32-SSE-NEXT: punpckhbw {{.*#+}} xmm2 = xmm2[8],xmm0[8],xmm2[9],xmm0[9],xmm2[10],xmm0[10],xmm2[11],xmm0[11],xmm2[12],xmm0[12],xmm2[13],xmm0[13],xmm2[14],xmm0[14],xmm2[15],xmm0[15]
; XOPAVX2-NEXT: vpshaq %xmm1, %xmm0, %xmm0
; XOPAVX2-NEXT: retq
;
+; AVX512-LABEL: splatvar_shift_v2i64:
+; AVX512: ## BB#0:
+; AVX512-NEXT: vmovdqa {{.*#+}} xmm2 = [9223372036854775808,9223372036854775808]
+; AVX512-NEXT: vpsrlq %xmm1, %xmm2, %xmm2
+; AVX512-NEXT: vpsrlq %xmm1, %xmm0, %xmm0
+; AVX512-NEXT: vpxor %xmm2, %xmm0, %xmm0
+; AVX512-NEXT: vpsubq %xmm2, %xmm0, %xmm0
+; AVX512-NEXT: retq
+;
; X32-SSE-LABEL: splatvar_shift_v2i64:
; X32-SSE: # BB#0:
; X32-SSE-NEXT: movq {{.*#+}} xmm1 = xmm1[0],zero
; XOP-NEXT: vpsrad %xmm1, %xmm0, %xmm0
; XOP-NEXT: retq
;
+; AVX512-LABEL: splatvar_shift_v4i32:
+; AVX512: ## BB#0:
+; AVX512-NEXT: vxorps %xmm2, %xmm2, %xmm2
+; AVX512-NEXT: vmovss %xmm1, %xmm2, %xmm1
+; AVX512-NEXT: vpsrad %xmm1, %xmm0, %xmm0
+; AVX512-NEXT: retq
+;
; X32-SSE-LABEL: splatvar_shift_v4i32:
; X32-SSE: # BB#0:
; X32-SSE-NEXT: xorps %xmm2, %xmm2
; XOP-NEXT: vpsraw %xmm1, %xmm0, %xmm0
; XOP-NEXT: retq
;
+; AVX512-LABEL: splatvar_shift_v8i16:
+; AVX512: ## BB#0:
+; AVX512-NEXT: vpxor %xmm2, %xmm2, %xmm2
+; AVX512-NEXT: vpblendw {{.*#+}} xmm1 = xmm1[0],xmm2[1,2,3,4,5,6,7]
+; AVX512-NEXT: vpsraw %xmm1, %xmm0, %xmm0
+; AVX512-NEXT: retq
+;
; X32-SSE-LABEL: splatvar_shift_v8i16:
; X32-SSE: # BB#0:
; X32-SSE-NEXT: movd %xmm1, %eax
; XOPAVX2-NEXT: vpshab %xmm1, %xmm0, %xmm0
; XOPAVX2-NEXT: retq
;
+; AVX512-LABEL: splatvar_shift_v16i8:
+; AVX512: ## BB#0:
+; AVX512-NEXT: vpbroadcastb %xmm1, %xmm1
+; AVX512-NEXT: vpsllw $5, %xmm1, %xmm1
+; AVX512-NEXT: vpunpckhbw {{.*#+}} xmm2 = xmm0[8],xmm1[8],xmm0[9],xmm1[9],xmm0[10],xmm1[10],xmm0[11],xmm1[11],xmm0[12],xmm1[12],xmm0[13],xmm1[13],xmm0[14],xmm1[14],xmm0[15],xmm1[15]
+; AVX512-NEXT: vpunpckhbw {{.*#+}} xmm3 = xmm0[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15]
+; AVX512-NEXT: vpsraw $4, %xmm3, %xmm4
+; AVX512-NEXT: vpblendvb %xmm2, %xmm4, %xmm3, %xmm3
+; AVX512-NEXT: vpsraw $2, %xmm3, %xmm4
+; AVX512-NEXT: vpaddw %xmm2, %xmm2, %xmm2
+; AVX512-NEXT: vpblendvb %xmm2, %xmm4, %xmm3, %xmm3
+; AVX512-NEXT: vpsraw $1, %xmm3, %xmm4
+; AVX512-NEXT: vpaddw %xmm2, %xmm2, %xmm2
+; AVX512-NEXT: vpblendvb %xmm2, %xmm4, %xmm3, %xmm2
+; AVX512-NEXT: vpsrlw $8, %xmm2, %xmm2
+; AVX512-NEXT: vpunpcklbw {{.*#+}} xmm1 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3],xmm0[4],xmm1[4],xmm0[5],xmm1[5],xmm0[6],xmm1[6],xmm0[7],xmm1[7]
+; AVX512-NEXT: vpunpcklbw {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; AVX512-NEXT: vpsraw $4, %xmm0, %xmm3
+; AVX512-NEXT: vpblendvb %xmm1, %xmm3, %xmm0, %xmm0
+; AVX512-NEXT: vpsraw $2, %xmm0, %xmm3
+; AVX512-NEXT: vpaddw %xmm1, %xmm1, %xmm1
+; AVX512-NEXT: vpblendvb %xmm1, %xmm3, %xmm0, %xmm0
+; AVX512-NEXT: vpsraw $1, %xmm0, %xmm3
+; AVX512-NEXT: vpaddw %xmm1, %xmm1, %xmm1
+; AVX512-NEXT: vpblendvb %xmm1, %xmm3, %xmm0, %xmm0
+; AVX512-NEXT: vpsrlw $8, %xmm0, %xmm0
+; AVX512-NEXT: vpackuswb %xmm2, %xmm0, %xmm0
+; AVX512-NEXT: retq
+;
; X32-SSE-LABEL: splatvar_shift_v16i8:
; X32-SSE: # BB#0:
; X32-SSE-NEXT: punpcklbw {{.*#+}} xmm1 = xmm1[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
; XOP-NEXT: vpshaq %xmm1, %xmm0, %xmm0
; XOP-NEXT: retq
;
+; AVX512-LABEL: constant_shift_v2i64:
+; AVX512: ## BB#0:
+; AVX512-NEXT: vpsrlvq {{.*}}(%rip), %xmm0, %xmm0
+; AVX512-NEXT: vmovdqa {{.*#+}} xmm1 = [4611686018427387904,72057594037927936]
+; AVX512-NEXT: vpxor %xmm1, %xmm0, %xmm0
+; AVX512-NEXT: vpsubq %xmm1, %xmm0, %xmm0
+; AVX512-NEXT: retq
+;
; X32-SSE-LABEL: constant_shift_v2i64:
; X32-SSE: # BB#0:
; X32-SSE-NEXT: movdqa {{.*#+}} xmm1 = [0,2147483648,0,2147483648]
; XOPAVX2-NEXT: vpsravd {{.*}}(%rip), %xmm0, %xmm0
; XOPAVX2-NEXT: retq
;
+; AVX512-LABEL: constant_shift_v4i32:
+; AVX512: ## BB#0:
+; AVX512-NEXT: vpsravd {{.*}}(%rip), %xmm0, %xmm0
+; AVX512-NEXT: retq
+;
; X32-SSE-LABEL: constant_shift_v4i32:
; X32-SSE: # BB#0:
; X32-SSE-NEXT: movdqa %xmm0, %xmm1
; XOP-NEXT: vpshaw %xmm1, %xmm0, %xmm0
; XOP-NEXT: retq
;
+; AVX512-LABEL: constant_shift_v8i16:
+; AVX512: ## BB#0:
+; AVX512-NEXT: vmovdqa {{.*#+}} xmm1 = [0,1,2,3,4,5,6,7]
+; AVX512-NEXT: vpsravw %zmm1, %zmm0, %zmm0
+; AVX512-NEXT: retq
+;
; X32-SSE-LABEL: constant_shift_v8i16:
; X32-SSE: # BB#0:
; X32-SSE-NEXT: movdqa %xmm0, %xmm1
; XOP-NEXT: vpshab %xmm1, %xmm0, %xmm0
; XOP-NEXT: retq
;
+; AVX512-LABEL: constant_shift_v16i8:
+; AVX512: ## BB#0:
+; AVX512-NEXT: vmovdqa {{.*#+}} xmm1 = [0,1,2,3,4,5,6,7,7,6,5,4,3,2,1,0]
+; AVX512-NEXT: vpsllw $5, %xmm1, %xmm1
+; AVX512-NEXT: vpunpckhbw {{.*#+}} xmm2 = xmm0[8],xmm1[8],xmm0[9],xmm1[9],xmm0[10],xmm1[10],xmm0[11],xmm1[11],xmm0[12],xmm1[12],xmm0[13],xmm1[13],xmm0[14],xmm1[14],xmm0[15],xmm1[15]
+; AVX512-NEXT: vpunpckhbw {{.*#+}} xmm3 = xmm0[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15]
+; AVX512-NEXT: vpsraw $4, %xmm3, %xmm4
+; AVX512-NEXT: vpblendvb %xmm2, %xmm4, %xmm3, %xmm3
+; AVX512-NEXT: vpsraw $2, %xmm3, %xmm4
+; AVX512-NEXT: vpaddw %xmm2, %xmm2, %xmm2
+; AVX512-NEXT: vpblendvb %xmm2, %xmm4, %xmm3, %xmm3
+; AVX512-NEXT: vpsraw $1, %xmm3, %xmm4
+; AVX512-NEXT: vpaddw %xmm2, %xmm2, %xmm2
+; AVX512-NEXT: vpblendvb %xmm2, %xmm4, %xmm3, %xmm2
+; AVX512-NEXT: vpsrlw $8, %xmm2, %xmm2
+; AVX512-NEXT: vpunpcklbw {{.*#+}} xmm1 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3],xmm0[4],xmm1[4],xmm0[5],xmm1[5],xmm0[6],xmm1[6],xmm0[7],xmm1[7]
+; AVX512-NEXT: vpunpcklbw {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; AVX512-NEXT: vpsraw $4, %xmm0, %xmm3
+; AVX512-NEXT: vpblendvb %xmm1, %xmm3, %xmm0, %xmm0
+; AVX512-NEXT: vpsraw $2, %xmm0, %xmm3
+; AVX512-NEXT: vpaddw %xmm1, %xmm1, %xmm1
+; AVX512-NEXT: vpblendvb %xmm1, %xmm3, %xmm0, %xmm0
+; AVX512-NEXT: vpsraw $1, %xmm0, %xmm3
+; AVX512-NEXT: vpaddw %xmm1, %xmm1, %xmm1
+; AVX512-NEXT: vpblendvb %xmm1, %xmm3, %xmm0, %xmm0
+; AVX512-NEXT: vpsrlw $8, %xmm0, %xmm0
+; AVX512-NEXT: vpackuswb %xmm2, %xmm0, %xmm0
+; AVX512-NEXT: retq
+;
; X32-SSE-LABEL: constant_shift_v16i8:
; X32-SSE: # BB#0:
; X32-SSE-NEXT: punpckhbw {{.*#+}} xmm1 = xmm1[8],xmm0[8],xmm1[9],xmm0[9],xmm1[10],xmm0[10],xmm1[11],xmm0[11],xmm1[12],xmm0[12],xmm1[13],xmm0[13],xmm1[14],xmm0[14],xmm1[15],xmm0[15]
; XOP-NEXT: vpshaq %xmm1, %xmm0, %xmm0
; XOP-NEXT: retq
;
+; AVX512-LABEL: splatconstant_shift_v2i64:
+; AVX512: ## BB#0:
+; AVX512-NEXT: vpsrad $7, %xmm0, %xmm1
+; AVX512-NEXT: vpsrlq $7, %xmm0, %xmm0
+; AVX512-NEXT: vpblendd {{.*#+}} xmm0 = xmm0[0],xmm1[1],xmm0[2],xmm1[3]
+; AVX512-NEXT: retq
+;
; X32-SSE-LABEL: splatconstant_shift_v2i64:
; X32-SSE: # BB#0:
; X32-SSE-NEXT: movdqa %xmm0, %xmm1
; XOP-NEXT: vpsrad $5, %xmm0, %xmm0
; XOP-NEXT: retq
;
+; AVX512-LABEL: splatconstant_shift_v4i32:
+; AVX512: ## BB#0:
+; AVX512-NEXT: vpsrad $5, %xmm0, %xmm0
+; AVX512-NEXT: retq
+;
; X32-SSE-LABEL: splatconstant_shift_v4i32:
; X32-SSE: # BB#0:
; X32-SSE-NEXT: psrad $5, %xmm0
; XOP-NEXT: vpsraw $3, %xmm0, %xmm0
; XOP-NEXT: retq
;
+; AVX512-LABEL: splatconstant_shift_v8i16:
+; AVX512: ## BB#0:
+; AVX512-NEXT: vpsraw $3, %xmm0, %xmm0
+; AVX512-NEXT: retq
+;
; X32-SSE-LABEL: splatconstant_shift_v8i16:
; X32-SSE: # BB#0:
; X32-SSE-NEXT: psraw $3, %xmm0
; XOP-NEXT: vpshab %xmm1, %xmm0, %xmm0
; XOP-NEXT: retq
;
+; AVX512-LABEL: splatconstant_shift_v16i8:
+; AVX512: ## BB#0:
+; AVX512-NEXT: vpsrlw $3, %xmm0, %xmm0
+; AVX512-NEXT: vpand {{.*}}(%rip), %xmm0, %xmm0
+; AVX512-NEXT: vmovdqa {{.*#+}} xmm1 = [16,16,16,16,16,16,16,16,16,16,16,16,16,16,16,16]
+; AVX512-NEXT: vpxor %xmm1, %xmm0, %xmm0
+; AVX512-NEXT: vpsubb %xmm1, %xmm0, %xmm0
+; AVX512-NEXT: retq
+;
; X32-SSE-LABEL: splatconstant_shift_v16i8:
; X32-SSE: # BB#0:
; X32-SSE-NEXT: psrlw $3, %xmm0