+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mcpu=x86-64 | FileCheck %s --check-prefix=ALL --check-prefix=SSE --check-prefix=SSE2
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mcpu=x86-64 -mattr=+ssse3 | FileCheck %s --check-prefix=ALL --check-prefix=SSE --check-prefix=SSSE3
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mcpu=x86-64 -mattr=+sse4.1 | FileCheck %s --check-prefix=ALL --check-prefix=SSE --check-prefix=SSE41
define <8 x i16> @shuffle_v8i16_z8zzzzzz(i16 %i) {
; SSE-LABEL: shuffle_v8i16_z8zzzzzz:
; SSE: # BB#0:
-; SSE-NEXT: pxor %xmm0, %xmm0
+; SSE-NEXT: pxor %xmm0, %xmm0
; SSE-NEXT: pinsrw $1, %edi, %xmm0
; SSE-NEXT: retq
;
; AVX-LABEL: shuffle_v8i16_z8zzzzzz:
; AVX: # BB#0:
-; AVX-NEXT: vpxor %xmm0, %xmm0
-; AVX-NEXT: vpinsrw $1, %edi, %xmm0
+; AVX-NEXT: vpxor %xmm0, %xmm0, %xmm0
+; AVX-NEXT: vpinsrw $1, %edi, %xmm0, %xmm0
; AVX-NEXT: retq
%a = insertelement <8 x i16> undef, i16 %i, i32 0
%shuffle = shufflevector <8 x i16> zeroinitializer, <8 x i16> %a, <8 x i32> <i32 2, i32 8, i32 3, i32 7, i32 6, i32 5, i32 4, i32 3>
define <8 x i16> @shuffle_v8i16_zzzzz8zz(i16 %i) {
; SSE-LABEL: shuffle_v8i16_zzzzz8zz:
; SSE: # BB#0:
-; SSE-NEXT: pxor %xmm0, %xmm0
+; SSE-NEXT: pxor %xmm0, %xmm0
; SSE-NEXT: pinsrw $5, %edi, %xmm0
; SSE-NEXT: retq
;
; AVX-LABEL: shuffle_v8i16_zzzzz8zz:
; AVX: # BB#0:
-; AVX-NEXT: vpxor %xmm0, %xmm0
-; AVX-NEXT: vpinsrw $5, %edi, %xmm0
+; AVX-NEXT: vpxor %xmm0, %xmm0, %xmm0
+; AVX-NEXT: vpinsrw $5, %edi, %xmm0, %xmm0
; AVX-NEXT: retq
%a = insertelement <8 x i16> undef, i16 %i, i32 0
%shuffle = shufflevector <8 x i16> zeroinitializer, <8 x i16> %a, <8 x i32> <i32 0, i32 0, i32 0, i32 0, i32 0, i32 8, i32 0, i32 0>
define <8 x i16> @shuffle_v8i16_zuuzuuz8(i16 %i) {
; SSE-LABEL: shuffle_v8i16_zuuzuuz8:
; SSE: # BB#0:
-; SSE-NEXT: pxor %xmm0, %xmm0
+; SSE-NEXT: pxor %xmm0, %xmm0
; SSE-NEXT: pinsrw $7, %edi, %xmm0
; SSE-NEXT: retq
;
; AVX-LABEL: shuffle_v8i16_zuuzuuz8:
; AVX: # BB#0:
-; AVX-NEXT: vpxor %xmm0, %xmm0
-; AVX-NEXT: vpinsrw $7, %edi, %xmm0
+; AVX-NEXT: vpxor %xmm0, %xmm0, %xmm0
+; AVX-NEXT: vpinsrw $7, %edi, %xmm0, %xmm0
; AVX-NEXT: retq
%a = insertelement <8 x i16> undef, i16 %i, i32 0
%shuffle = shufflevector <8 x i16> zeroinitializer, <8 x i16> %a, <8 x i32> <i32 0, i32 undef, i32 undef, i32 3, i32 undef, i32 undef, i32 6, i32 8>
define <8 x i16> @shuffle_v8i16_zzBzzzzz(i16 %i) {
; SSE-LABEL: shuffle_v8i16_zzBzzzzz:
; SSE: # BB#0:
-; SSE-NEXT: pxor %xmm0, %xmm0
+; SSE-NEXT: pxor %xmm0, %xmm0
; SSE-NEXT: pinsrw $2, %edi, %xmm0
; SSE-NEXT: retq
;
; AVX-LABEL: shuffle_v8i16_zzBzzzzz:
; AVX: # BB#0:
-; AVX-NEXT: vpxor %xmm0, %xmm0
-; AVX-NEXT: vpinsrw $2, %edi, %xmm0
+; AVX-NEXT: vpxor %xmm0, %xmm0, %xmm0
+; AVX-NEXT: vpinsrw $2, %edi, %xmm0, %xmm0
; AVX-NEXT: retq
%a = insertelement <8 x i16> undef, i16 %i, i32 3
%shuffle = shufflevector <8 x i16> zeroinitializer, <8 x i16> %a, <8 x i32> <i32 0, i32 1, i32 11, i32 3, i32 4, i32 5, i32 6, i32 7>
ret <8 x i16> %shuffle
}
+
+define <8 x i16> @insert_dup_mem_v8i16_i32(i32* %ptr) {
+; SSE2-LABEL: insert_dup_mem_v8i16_i32:
+; SSE2: # BB#0:
+; SSE2-NEXT: movd {{.*#+}} xmm0 = mem[0],zero,zero,zero
+; SSE2-NEXT: pshufd {{.*#+}} xmm0 = xmm0[0,1,0,3]
+; SSE2-NEXT: pshuflw {{.*#+}} xmm0 = xmm0[0,0,0,0,4,5,6,7]
+; SSE2-NEXT: pshufhw {{.*#+}} xmm0 = xmm0[0,1,2,3,4,4,4,4]
+; SSE2-NEXT: retq
+;
+; SSSE3-LABEL: insert_dup_mem_v8i16_i32:
+; SSSE3: # BB#0:
+; SSSE3-NEXT: movd {{.*#+}} xmm0 = mem[0],zero,zero,zero
+; SSSE3-NEXT: pshufb {{.*#+}} xmm0 = xmm0[0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1]
+; SSSE3-NEXT: retq
+;
+; SSE41-LABEL: insert_dup_mem_v8i16_i32:
+; SSE41: # BB#0:
+; SSE41-NEXT: movd {{.*#+}} xmm0 = mem[0],zero,zero,zero
+; SSE41-NEXT: pshufb {{.*#+}} xmm0 = xmm0[0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1]
+; SSE41-NEXT: retq
+;
+; AVX1-LABEL: insert_dup_mem_v8i16_i32:
+; AVX1: # BB#0:
+; AVX1-NEXT: vmovd {{.*#+}} xmm0 = mem[0],zero,zero,zero
+; AVX1-NEXT: vpshufb {{.*#+}} xmm0 = xmm0[0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1]
+; AVX1-NEXT: retq
+;
+; AVX2-LABEL: insert_dup_mem_v8i16_i32:
+; AVX2: # BB#0:
+; AVX2-NEXT: vpbroadcastw (%rdi), %xmm0
+; AVX2-NEXT: retq
+ %tmp = load i32, i32* %ptr, align 4
+ %tmp1 = insertelement <4 x i32> zeroinitializer, i32 %tmp, i32 0
+ %tmp2 = bitcast <4 x i32> %tmp1 to <8 x i16>
+ %tmp3 = shufflevector <8 x i16> %tmp2, <8 x i16> undef, <8 x i32> zeroinitializer
+ ret <8 x i16> %tmp3
+}
+
+define <8 x i16> @insert_dup_mem_v8i16_sext_i16(i16* %ptr) {
+; SSE2-LABEL: insert_dup_mem_v8i16_sext_i16:
+; SSE2: # BB#0:
+; SSE2-NEXT: movswl (%rdi), %eax
+; SSE2-NEXT: movd %eax, %xmm0
+; SSE2-NEXT: pshufd {{.*#+}} xmm0 = xmm0[0,1,0,3]
+; SSE2-NEXT: pshuflw {{.*#+}} xmm0 = xmm0[0,0,0,0,4,5,6,7]
+; SSE2-NEXT: pshufhw {{.*#+}} xmm0 = xmm0[0,1,2,3,4,4,4,4]
+; SSE2-NEXT: retq
+;
+; SSSE3-LABEL: insert_dup_mem_v8i16_sext_i16:
+; SSSE3: # BB#0:
+; SSSE3-NEXT: movswl (%rdi), %eax
+; SSSE3-NEXT: movd %eax, %xmm0
+; SSSE3-NEXT: pshufb {{.*#+}} xmm0 = xmm0[0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1]
+; SSSE3-NEXT: retq
+;
+; SSE41-LABEL: insert_dup_mem_v8i16_sext_i16:
+; SSE41: # BB#0:
+; SSE41-NEXT: movswl (%rdi), %eax
+; SSE41-NEXT: movd %eax, %xmm0
+; SSE41-NEXT: pshufb {{.*#+}} xmm0 = xmm0[0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1]
+; SSE41-NEXT: retq
+;
+; AVX1-LABEL: insert_dup_mem_v8i16_sext_i16:
+; AVX1: # BB#0:
+; AVX1-NEXT: movswl (%rdi), %eax
+; AVX1-NEXT: vmovd %eax, %xmm0
+; AVX1-NEXT: vpshufb {{.*#+}} xmm0 = xmm0[0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1]
+; AVX1-NEXT: retq
+;
+; AVX2-LABEL: insert_dup_mem_v8i16_sext_i16:
+; AVX2: # BB#0:
+; AVX2-NEXT: movswl (%rdi), %eax
+; AVX2-NEXT: vmovd %eax, %xmm0
+; AVX2-NEXT: vpbroadcastw %xmm0, %xmm0
+; AVX2-NEXT: retq
+ %tmp = load i16, i16* %ptr, align 2
+ %tmp1 = sext i16 %tmp to i32
+ %tmp2 = insertelement <4 x i32> zeroinitializer, i32 %tmp1, i32 0
+ %tmp3 = bitcast <4 x i32> %tmp2 to <8 x i16>
+ %tmp4 = shufflevector <8 x i16> %tmp3, <8 x i16> undef, <8 x i32> zeroinitializer
+ ret <8 x i16> %tmp4
+}
+
+define <8 x i16> @insert_dup_elt1_mem_v8i16_i32(i32* %ptr) {
+; SSE2-LABEL: insert_dup_elt1_mem_v8i16_i32:
+; SSE2: # BB#0:
+; SSE2-NEXT: movd {{.*#+}} xmm0 = mem[0],zero,zero,zero
+; SSE2-NEXT: pshufd {{.*#+}} xmm0 = xmm0[0,1,0,3]
+; SSE2-NEXT: pshuflw {{.*#+}} xmm0 = xmm0[1,1,1,1,4,5,6,7]
+; SSE2-NEXT: pshufhw {{.*#+}} xmm0 = xmm0[0,1,2,3,5,5,5,5]
+; SSE2-NEXT: retq
+;
+; SSSE3-LABEL: insert_dup_elt1_mem_v8i16_i32:
+; SSSE3: # BB#0:
+; SSSE3-NEXT: movd {{.*#+}} xmm0 = mem[0],zero,zero,zero
+; SSSE3-NEXT: pshufb {{.*#+}} xmm0 = xmm0[2,3,2,3,2,3,2,3,2,3,2,3,2,3,2,3]
+; SSSE3-NEXT: retq
+;
+; SSE41-LABEL: insert_dup_elt1_mem_v8i16_i32:
+; SSE41: # BB#0:
+; SSE41-NEXT: movd {{.*#+}} xmm0 = mem[0],zero,zero,zero
+; SSE41-NEXT: pshufb {{.*#+}} xmm0 = xmm0[2,3,2,3,2,3,2,3,2,3,2,3,2,3,2,3]
+; SSE41-NEXT: retq
+;
+; AVX1-LABEL: insert_dup_elt1_mem_v8i16_i32:
+; AVX1: # BB#0:
+; AVX1-NEXT: vmovd {{.*#+}} xmm0 = mem[0],zero,zero,zero
+; AVX1-NEXT: vpshufb {{.*#+}} xmm0 = xmm0[2,3,2,3,2,3,2,3,2,3,2,3,2,3,2,3]
+; AVX1-NEXT: retq
+;
+; AVX2-LABEL: insert_dup_elt1_mem_v8i16_i32:
+; AVX2: # BB#0:
+; AVX2-NEXT: vpbroadcastw 2(%rdi), %xmm0
+; AVX2-NEXT: retq
+ %tmp = load i32, i32* %ptr, align 4
+ %tmp1 = insertelement <4 x i32> zeroinitializer, i32 %tmp, i32 0
+ %tmp2 = bitcast <4 x i32> %tmp1 to <8 x i16>
+ %tmp3 = shufflevector <8 x i16> %tmp2, <8 x i16> undef, <8 x i32> <i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1>
+ ret <8 x i16> %tmp3
+}
+
+define <8 x i16> @insert_dup_elt3_mem_v8i16_i32(i32* %ptr) {
+; SSE2-LABEL: insert_dup_elt3_mem_v8i16_i32:
+; SSE2: # BB#0:
+; SSE2-NEXT: movd {{.*#+}} xmm0 = mem[0],zero,zero,zero
+; SSE2-NEXT: pshufd {{.*#+}} xmm0 = xmm0[1,0,1,0]
+; SSE2-NEXT: pshuflw {{.*#+}} xmm0 = xmm0[3,3,3,3,4,5,6,7]
+; SSE2-NEXT: pshufhw {{.*#+}} xmm0 = xmm0[0,1,2,3,7,7,7,7]
+; SSE2-NEXT: retq
+;
+; SSSE3-LABEL: insert_dup_elt3_mem_v8i16_i32:
+; SSSE3: # BB#0:
+; SSSE3-NEXT: movd {{.*#+}} xmm0 = mem[0],zero,zero,zero
+; SSSE3-NEXT: pshufb {{.*#+}} xmm0 = xmm0[2,3,2,3,2,3,2,3,2,3,2,3,2,3,2,3]
+; SSSE3-NEXT: retq
+;
+; SSE41-LABEL: insert_dup_elt3_mem_v8i16_i32:
+; SSE41: # BB#0:
+; SSE41-NEXT: movd {{.*#+}} xmm0 = mem[0],zero,zero,zero
+; SSE41-NEXT: pshufb {{.*#+}} xmm0 = xmm0[2,3,2,3,2,3,2,3,2,3,2,3,2,3,2,3]
+; SSE41-NEXT: retq
+;
+; AVX1-LABEL: insert_dup_elt3_mem_v8i16_i32:
+; AVX1: # BB#0:
+; AVX1-NEXT: vmovd {{.*#+}} xmm0 = mem[0],zero,zero,zero
+; AVX1-NEXT: vpshufb {{.*#+}} xmm0 = xmm0[2,3,2,3,2,3,2,3,2,3,2,3,2,3,2,3]
+; AVX1-NEXT: retq
+;
+; AVX2-LABEL: insert_dup_elt3_mem_v8i16_i32:
+; AVX2: # BB#0:
+; AVX2-NEXT: vpbroadcastw 2(%rdi), %xmm0
+; AVX2-NEXT: retq
+ %tmp = load i32, i32* %ptr, align 4
+ %tmp1 = insertelement <4 x i32> zeroinitializer, i32 %tmp, i32 1
+ %tmp2 = bitcast <4 x i32> %tmp1 to <8 x i16>
+ %tmp3 = shufflevector <8 x i16> %tmp2, <8 x i16> undef, <8 x i32> <i32 3, i32 3, i32 3, i32 3, i32 3, i32 3, i32 3, i32 3>
+ ret <8 x i16> %tmp3
+}
+
+define <8 x i16> @insert_dup_elt1_mem_v8i16_sext_i16(i16* %ptr) {
+; SSE2-LABEL: insert_dup_elt1_mem_v8i16_sext_i16:
+; SSE2: # BB#0:
+; SSE2-NEXT: movswl (%rdi), %eax
+; SSE2-NEXT: movd %eax, %xmm0
+; SSE2-NEXT: pshufd {{.*#+}} xmm0 = xmm0[0,1,0,3]
+; SSE2-NEXT: pshuflw {{.*#+}} xmm0 = xmm0[1,1,1,1,4,5,6,7]
+; SSE2-NEXT: pshufhw {{.*#+}} xmm0 = xmm0[0,1,2,3,5,5,5,5]
+; SSE2-NEXT: retq
+;
+; SSSE3-LABEL: insert_dup_elt1_mem_v8i16_sext_i16:
+; SSSE3: # BB#0:
+; SSSE3-NEXT: movswl (%rdi), %eax
+; SSSE3-NEXT: movd %eax, %xmm0
+; SSSE3-NEXT: pshufb {{.*#+}} xmm0 = xmm0[2,3,2,3,2,3,2,3,2,3,2,3,2,3,2,3]
+; SSSE3-NEXT: retq
+;
+; SSE41-LABEL: insert_dup_elt1_mem_v8i16_sext_i16:
+; SSE41: # BB#0:
+; SSE41-NEXT: movswl (%rdi), %eax
+; SSE41-NEXT: movd %eax, %xmm0
+; SSE41-NEXT: pshufb {{.*#+}} xmm0 = xmm0[2,3,2,3,2,3,2,3,2,3,2,3,2,3,2,3]
+; SSE41-NEXT: retq
+;
+; AVX1-LABEL: insert_dup_elt1_mem_v8i16_sext_i16:
+; AVX1: # BB#0:
+; AVX1-NEXT: movswl (%rdi), %eax
+; AVX1-NEXT: vmovd %eax, %xmm0
+; AVX1-NEXT: vpshufb {{.*#+}} xmm0 = xmm0[2,3,2,3,2,3,2,3,2,3,2,3,2,3,2,3]
+; AVX1-NEXT: retq
+;
+; AVX2-LABEL: insert_dup_elt1_mem_v8i16_sext_i16:
+; AVX2: # BB#0:
+; AVX2-NEXT: movswl (%rdi), %eax
+; AVX2-NEXT: shrl $16, %eax
+; AVX2-NEXT: vmovd %eax, %xmm0
+; AVX2-NEXT: vpbroadcastw %xmm0, %xmm0
+; AVX2-NEXT: retq
+ %tmp = load i16, i16* %ptr, align 2
+ %tmp1 = sext i16 %tmp to i32
+ %tmp2 = insertelement <4 x i32> zeroinitializer, i32 %tmp1, i32 0
+ %tmp3 = bitcast <4 x i32> %tmp2 to <8 x i16>
+ %tmp4 = shufflevector <8 x i16> %tmp3, <8 x i16> undef, <8 x i32> <i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1, i32 1>
+ ret <8 x i16> %tmp4
+}
+
+define <8 x i16> @insert_dup_elt3_mem_v8i16_sext_i16(i16* %ptr) {
+; SSE2-LABEL: insert_dup_elt3_mem_v8i16_sext_i16:
+; SSE2: # BB#0:
+; SSE2-NEXT: movswl (%rdi), %eax
+; SSE2-NEXT: movd %eax, %xmm0
+; SSE2-NEXT: pshufd {{.*#+}} xmm0 = xmm0[1,0,1,0]
+; SSE2-NEXT: pshuflw {{.*#+}} xmm0 = xmm0[3,3,3,3,4,5,6,7]
+; SSE2-NEXT: pshufhw {{.*#+}} xmm0 = xmm0[0,1,2,3,7,7,7,7]
+; SSE2-NEXT: retq
+;
+; SSSE3-LABEL: insert_dup_elt3_mem_v8i16_sext_i16:
+; SSSE3: # BB#0:
+; SSSE3-NEXT: movswl (%rdi), %eax
+; SSSE3-NEXT: movd %eax, %xmm0
+; SSSE3-NEXT: pshufb {{.*#+}} xmm0 = xmm0[2,3,2,3,2,3,2,3,2,3,2,3,2,3,2,3]
+; SSSE3-NEXT: retq
+;
+; SSE41-LABEL: insert_dup_elt3_mem_v8i16_sext_i16:
+; SSE41: # BB#0:
+; SSE41-NEXT: movswl (%rdi), %eax
+; SSE41-NEXT: movd %eax, %xmm0
+; SSE41-NEXT: pshufb {{.*#+}} xmm0 = xmm0[2,3,2,3,2,3,2,3,2,3,2,3,2,3,2,3]
+; SSE41-NEXT: retq
+;
+; AVX1-LABEL: insert_dup_elt3_mem_v8i16_sext_i16:
+; AVX1: # BB#0:
+; AVX1-NEXT: movswl (%rdi), %eax
+; AVX1-NEXT: vmovd %eax, %xmm0
+; AVX1-NEXT: vpshufb {{.*#+}} xmm0 = xmm0[2,3,2,3,2,3,2,3,2,3,2,3,2,3,2,3]
+; AVX1-NEXT: retq
+;
+; AVX2-LABEL: insert_dup_elt3_mem_v8i16_sext_i16:
+; AVX2: # BB#0:
+; AVX2-NEXT: movswl (%rdi), %eax
+; AVX2-NEXT: shrl $16, %eax
+; AVX2-NEXT: vmovd %eax, %xmm0
+; AVX2-NEXT: vpbroadcastw %xmm0, %xmm0
+; AVX2-NEXT: retq
+ %tmp = load i16, i16* %ptr, align 2
+ %tmp1 = sext i16 %tmp to i32
+ %tmp2 = insertelement <4 x i32> zeroinitializer, i32 %tmp1, i32 1
+ %tmp3 = bitcast <4 x i32> %tmp2 to <8 x i16>
+ %tmp4 = shufflevector <8 x i16> %tmp3, <8 x i16> undef, <8 x i32> <i32 3, i32 3, i32 3, i32 3, i32 3, i32 3, i32 3, i32 3>
+ ret <8 x i16> %tmp4
+}