derek.holloway
7502bf64a4
CodeQL / Analyze (actions) (push) Canceled after 0s
CodeQL / Analyze (python) (push) Canceled after 0s
pre-commit / pre-commit (push) Canceled after 0s
Multi-Arch CI / setup (push) Canceled after 0s
Publish build_manylinux_x86_64 images / publish_build_manylinux_x86_64 (push) Canceled after 0s
Setup (Multi-Arch) / setup (push) Canceled after 0s
Multi-Arch CI / Linux::${{ fromJSON(needs.setup.outputs.linux_build_config || '{}').build_variant_label || 'skip' }} (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish no_rocm_image_ubi10 image / publish_no_rocm_image_ubi10 (push) Canceled after 0s
Multi-Arch CI / Windows::${{ fromJSON(needs.setup.outputs.windows_build_config || '{}').build_variant_label || 'skip' }} (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Multi-Arch CI / Manifest Diff (push) Canceled after 0s
TheRock Manifest Diff Report / Generate Manifest Diff Report (push) Canceled after 0s
Multi-Arch CI / CI Summary (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04 (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04_ocl_rt images / publish_no_rocm_image_ubuntu24_04_ocl_rt (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_rocgdb (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04_rocgdb images / publish_no_rocm_image_ubuntu24_04_rocgdb (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_ocl_rt (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Unit Tests / Unit Tests :: ubuntu-24.04 (push) Canceled after 0s
Unit Tests / Unit Tests :: windows-2022 (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_media (push) Canceled after 0s
Unit Tests / Unit Tests Summary (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_openmpi (push) Canceled after 0s
993 lines
55 KiB
LLVM
993 lines
55 KiB
LLVM
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
|
|
; RUN: llc < %s -mcpu=x86-64 -mattr=+avx512f,+avx512dq | FileCheck %s --check-prefixes=ALL,SLOW,AVX512F
|
|
; RUN: llc < %s -mcpu=x86-64 -mattr=+avx512bw,+avx512dq | FileCheck %s --check-prefixes=ALL,SLOW,AVX512BW
|
|
; RUN: llc < %s -mcpu=x86-64 -mattr=+avx512f,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=ALL,FAST,AVX512F
|
|
; RUN: llc < %s -mcpu=x86-64 -mattr=+avx512bw,+avx512dq,+fast-variable-crosslane-shuffle,+fast-variable-perlane-shuffle | FileCheck %s --check-prefixes=ALL,FAST,AVX512BW
|
|
|
|
target triple = "x86_64-unknown-unknown"
|
|
|
|
define <16 x float> @shuffle_v16f32_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00(<16 x float> %a, <16 x float> %b) {
|
|
; ALL-LABEL: shuffle_v16f32_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vbroadcastss %xmm0, %zmm0
|
|
; ALL-NEXT: retq
|
|
%shuffle = shufflevector <16 x float> %a, <16 x float> %b, <16 x i32><i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0>
|
|
ret <16 x float> %shuffle
|
|
}
|
|
|
|
define <16 x float> @shuffle_v16f32_08_08_08_08_08_08_08_08_08_08_08_08_08_08_08_08(<16 x float> %a, <16 x float> %b) {
|
|
; SLOW-LABEL: shuffle_v16f32_08_08_08_08_08_08_08_08_08_08_08_08_08_08_08_08:
|
|
; SLOW: # %bb.0:
|
|
; SLOW-NEXT: vextractf32x4 $2, %zmm0, %xmm0
|
|
; SLOW-NEXT: vbroadcastss %xmm0, %zmm0
|
|
; SLOW-NEXT: retq
|
|
;
|
|
; FAST-LABEL: shuffle_v16f32_08_08_08_08_08_08_08_08_08_08_08_08_08_08_08_08:
|
|
; FAST: # %bb.0:
|
|
; FAST-NEXT: vbroadcastss {{.*#+}} zmm1 = [8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8]
|
|
; FAST-NEXT: vpermps %zmm0, %zmm1, %zmm0
|
|
; FAST-NEXT: retq
|
|
%shuffle = shufflevector <16 x float> %a, <16 x float> %b, <16 x i32><i32 8, i32 8, i32 8, i32 8, i32 8, i32 8, i32 8, i32 8, i32 8, i32 8, i32 8, i32 8, i32 8, i32 8, i32 8, i32 8>
|
|
ret <16 x float> %shuffle
|
|
}
|
|
|
|
define <16 x float> @shuffle_v16f32_08_08_08_08_08_08_08_08_08_08_08_08_08_08_08_08_bc(<16 x i32> %a, <16 x i32> %b) {
|
|
; SLOW-LABEL: shuffle_v16f32_08_08_08_08_08_08_08_08_08_08_08_08_08_08_08_08_bc:
|
|
; SLOW: # %bb.0:
|
|
; SLOW-NEXT: vextractf32x4 $2, %zmm0, %xmm0
|
|
; SLOW-NEXT: vbroadcastss %xmm0, %zmm0
|
|
; SLOW-NEXT: retq
|
|
;
|
|
; FAST-LABEL: shuffle_v16f32_08_08_08_08_08_08_08_08_08_08_08_08_08_08_08_08_bc:
|
|
; FAST: # %bb.0:
|
|
; FAST-NEXT: vbroadcastss {{.*#+}} zmm1 = [8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8]
|
|
; FAST-NEXT: vpermps %zmm0, %zmm1, %zmm0
|
|
; FAST-NEXT: retq
|
|
%tmp0 = bitcast <16 x i32> %a to <16 x float>
|
|
%tmp1 = bitcast <16 x i32> %b to <16 x float>
|
|
%shuffle = shufflevector <16 x float> %tmp0, <16 x float> %tmp1, <16 x i32><i32 8, i32 8, i32 8, i32 8, i32 8, i32 8, i32 8, i32 8, i32 8, i32 8, i32 8, i32 8, i32 8, i32 8, i32 8, i32 8>
|
|
ret <16 x float> %shuffle
|
|
}
|
|
|
|
define <16 x float> @shuffle_v16f32_00_10_01_11_04_14_05_15_08_18_09_19_0c_1c_0d_1d(<16 x float> %a, <16 x float> %b) {
|
|
; ALL-LABEL: shuffle_v16f32_00_10_01_11_04_14_05_15_08_18_09_19_0c_1c_0d_1d:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vunpcklps {{.*#+}} zmm0 = zmm0[0],zmm1[0],zmm0[1],zmm1[1],zmm0[4],zmm1[4],zmm0[5],zmm1[5],zmm0[8],zmm1[8],zmm0[9],zmm1[9],zmm0[12],zmm1[12],zmm0[13],zmm1[13]
|
|
; ALL-NEXT: retq
|
|
%shuffle = shufflevector <16 x float> %a, <16 x float> %b, <16 x i32><i32 0, i32 16, i32 1, i32 17, i32 4, i32 20, i32 5, i32 21, i32 8, i32 24, i32 9, i32 25, i32 12, i32 28, i32 13, i32 29>
|
|
ret <16 x float> %shuffle
|
|
}
|
|
|
|
define <16 x float> @shuffle_v16f32_00_zz_01_zz_04_zz_05_zz_08_zz_09_zz_0c_zz_0d_zz(<16 x float> %a, <16 x float> %b) {
|
|
; ALL-LABEL: shuffle_v16f32_00_zz_01_zz_04_zz_05_zz_08_zz_09_zz_0c_zz_0d_zz:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vxorps %xmm1, %xmm1, %xmm1
|
|
; ALL-NEXT: vunpcklps {{.*#+}} zmm0 = zmm0[0],zmm1[0],zmm0[1],zmm1[1],zmm0[4],zmm1[4],zmm0[5],zmm1[5],zmm0[8],zmm1[8],zmm0[9],zmm1[9],zmm0[12],zmm1[12],zmm0[13],zmm1[13]
|
|
; ALL-NEXT: retq
|
|
%shuffle = shufflevector <16 x float> %a, <16 x float> zeroinitializer, <16 x i32><i32 0, i32 16, i32 1, i32 16, i32 4, i32 16, i32 5, i32 16, i32 8, i32 16, i32 9, i32 16, i32 12, i32 16, i32 13, i32 16>
|
|
ret <16 x float> %shuffle
|
|
}
|
|
|
|
define <16 x float> @shuffle_v16f32_vunpcklps_swap(<16 x float> %a, <16 x float> %b) {
|
|
; ALL-LABEL: shuffle_v16f32_vunpcklps_swap:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vunpcklps {{.*#+}} zmm0 = zmm1[0],zmm0[0],zmm1[1],zmm0[1],zmm1[4],zmm0[4],zmm1[5],zmm0[5],zmm1[8],zmm0[8],zmm1[9],zmm0[9],zmm1[12],zmm0[12],zmm1[13],zmm0[13]
|
|
; ALL-NEXT: retq
|
|
%shuffle = shufflevector <16 x float> %a, <16 x float> %b, <16 x i32> <i32 16, i32 0, i32 17, i32 1, i32 20, i32 4, i32 21, i32 5, i32 24, i32 8, i32 25, i32 9, i32 28, i32 12, i32 29, i32 13>
|
|
ret <16 x float> %shuffle
|
|
}
|
|
|
|
; PR34382
|
|
define <16 x float> @shuffle_v16f32_01_01_03_00_06_04_05_07_08_08_09_09_15_14_14_12(<16 x float> %a0) {
|
|
; ALL-LABEL: shuffle_v16f32_01_01_03_00_06_04_05_07_08_08_09_09_15_14_14_12:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vpermilps {{.*#+}} zmm0 = zmm0[1,1,3,0,6,4,5,7,8,8,9,9,15,14,14,12]
|
|
; ALL-NEXT: retq
|
|
%shuffle = shufflevector <16 x float> %a0, <16 x float> poison, <16 x i32> <i32 1, i32 1, i32 3, i32 0, i32 6, i32 4, i32 5, i32 7, i32 8, i32 8, i32 9, i32 9, i32 15, i32 14, i32 14, i32 12>
|
|
ret <16 x float> %shuffle
|
|
}
|
|
|
|
define <16 x i32> @shuffle_v16i32_00_10_01_11_04_14_05_15_08_18_09_19_0c_1c_0d_1d(<16 x i32> %a, <16 x i32> %b) {
|
|
; ALL-LABEL: shuffle_v16i32_00_10_01_11_04_14_05_15_08_18_09_19_0c_1c_0d_1d:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vunpcklps {{.*#+}} zmm0 = zmm0[0],zmm1[0],zmm0[1],zmm1[1],zmm0[4],zmm1[4],zmm0[5],zmm1[5],zmm0[8],zmm1[8],zmm0[9],zmm1[9],zmm0[12],zmm1[12],zmm0[13],zmm1[13]
|
|
; ALL-NEXT: retq
|
|
%shuffle = shufflevector <16 x i32> %a, <16 x i32> %b, <16 x i32><i32 0, i32 16, i32 1, i32 17, i32 4, i32 20, i32 5, i32 21, i32 8, i32 24, i32 9, i32 25, i32 12, i32 28, i32 13, i32 29>
|
|
ret <16 x i32> %shuffle
|
|
}
|
|
|
|
define <16 x i32> @shuffle_v16i32_zz_10_zz_11_zz_14_zz_15_zz_18_zz_19_zz_1c_zz_1d(<16 x i32> %a, <16 x i32> %b) {
|
|
; ALL-LABEL: shuffle_v16i32_zz_10_zz_11_zz_14_zz_15_zz_18_zz_19_zz_1c_zz_1d:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vxorps %xmm0, %xmm0, %xmm0
|
|
; ALL-NEXT: vunpcklps {{.*#+}} zmm0 = zmm0[0],zmm1[0],zmm0[1],zmm1[1],zmm0[4],zmm1[4],zmm0[5],zmm1[5],zmm0[8],zmm1[8],zmm0[9],zmm1[9],zmm0[12],zmm1[12],zmm0[13],zmm1[13]
|
|
; ALL-NEXT: retq
|
|
%shuffle = shufflevector <16 x i32> zeroinitializer, <16 x i32> %b, <16 x i32><i32 15, i32 16, i32 13, i32 17, i32 11, i32 20, i32 9, i32 21, i32 7, i32 24, i32 5, i32 25, i32 3, i32 28, i32 1, i32 29>
|
|
ret <16 x i32> %shuffle
|
|
}
|
|
|
|
define <16 x float> @shuffle_v16f32_02_12_03_13_06_16_07_17_0a_1a_0b_1b_0e_1e_0f_1f(<16 x float> %a, <16 x float> %b) {
|
|
; ALL-LABEL: shuffle_v16f32_02_12_03_13_06_16_07_17_0a_1a_0b_1b_0e_1e_0f_1f:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vunpckhps {{.*#+}} zmm0 = zmm0[2],zmm1[2],zmm0[3],zmm1[3],zmm0[6],zmm1[6],zmm0[7],zmm1[7],zmm0[10],zmm1[10],zmm0[11],zmm1[11],zmm0[14],zmm1[14],zmm0[15],zmm1[15]
|
|
; ALL-NEXT: retq
|
|
%shuffle = shufflevector <16 x float> %a, <16 x float> %b, <16 x i32><i32 2, i32 18, i32 3, i32 19, i32 6, i32 22, i32 7, i32 23, i32 10, i32 26, i32 11, i32 27, i32 14, i32 30, i32 15, i32 31>
|
|
ret <16 x float> %shuffle
|
|
}
|
|
|
|
define <16 x float> @shuffle_v16f32_zz_12_zz_13_zz_16_zz_17_zz_1a_zz_1b_zz_1e_zz_1f(<16 x float> %a, <16 x float> %b) {
|
|
; ALL-LABEL: shuffle_v16f32_zz_12_zz_13_zz_16_zz_17_zz_1a_zz_1b_zz_1e_zz_1f:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vxorps %xmm0, %xmm0, %xmm0
|
|
; ALL-NEXT: vunpckhps {{.*#+}} zmm0 = zmm0[2],zmm1[2],zmm0[3],zmm1[3],zmm0[6],zmm1[6],zmm0[7],zmm1[7],zmm0[10],zmm1[10],zmm0[11],zmm1[11],zmm0[14],zmm1[14],zmm0[15],zmm1[15]
|
|
; ALL-NEXT: retq
|
|
%shuffle = shufflevector <16 x float> zeroinitializer, <16 x float> %b, <16 x i32><i32 0, i32 18, i32 0, i32 19, i32 4, i32 22, i32 4, i32 23, i32 6, i32 26, i32 6, i32 27, i32 8, i32 30, i32 8, i32 31>
|
|
ret <16 x float> %shuffle
|
|
}
|
|
|
|
define <16 x float> @shuffle_v16f32_00_00_02_02_04_04_06_06_08_08_10_10_12_12_14_14(<16 x float> %a, <16 x float> %b) {
|
|
; ALL-LABEL: shuffle_v16f32_00_00_02_02_04_04_06_06_08_08_10_10_12_12_14_14:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vmovsldup {{.*#+}} zmm0 = zmm0[0,0,2,2,4,4,6,6,8,8,10,10,12,12,14,14]
|
|
; ALL-NEXT: retq
|
|
%shuffle = shufflevector <16 x float> %a, <16 x float> %b, <16 x i32><i32 0, i32 0, i32 2, i32 2, i32 4, i32 4, i32 6, i32 6, i32 8, i32 8, i32 10, i32 10, i32 12, i32 12, i32 14, i32 14>
|
|
ret <16 x float> %shuffle
|
|
}
|
|
|
|
define <16 x float> @shuffle_v16f32_01_01_03_03_05_05_07_07_09_09_11_11_13_13_15_15(<16 x float> %a, <16 x float> %b) {
|
|
; ALL-LABEL: shuffle_v16f32_01_01_03_03_05_05_07_07_09_09_11_11_13_13_15_15:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vmovshdup {{.*#+}} zmm0 = zmm0[1,1,3,3,5,5,7,7,9,9,11,11,13,13,15,15]
|
|
; ALL-NEXT: retq
|
|
%shuffle = shufflevector <16 x float> %a, <16 x float> %b, <16 x i32><i32 1, i32 1, i32 3, i32 3, i32 5, i32 5, i32 7, i32 7, i32 9, i32 9, i32 11, i32 11, i32 13, i32 13, i32 15, i32 15>
|
|
ret <16 x float> %shuffle
|
|
}
|
|
|
|
define <16 x float> @shuffle_v16f32_00_01_00_01_06_07_06_07_08_09_10_11_12_13_12_13(<16 x float> %a, <16 x float> %b) {
|
|
; ALL-LABEL: shuffle_v16f32_00_01_00_01_06_07_06_07_08_09_10_11_12_13_12_13:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vshufpd {{.*#+}} zmm0 = zmm0[0,0,3,3,4,5,6,6]
|
|
; ALL-NEXT: retq
|
|
%shuffle = shufflevector <16 x float> %a, <16 x float> %b, <16 x i32> <i32 0, i32 1, i32 0, i32 1, i32 6, i32 7, i32 6, i32 7, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 12, i32 13>
|
|
ret <16 x float> %shuffle
|
|
}
|
|
|
|
define <16 x float> @shuffle_v16f32_00_00_02_00_04_04_06_04_08_08_10_08_12_12_14_12(<16 x float> %a, <16 x float> %b) {
|
|
; ALL-LABEL: shuffle_v16f32_00_00_02_00_04_04_06_04_08_08_10_08_12_12_14_12:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vshufps {{.*#+}} zmm0 = zmm0[0,0,2,0,4,4,6,4,8,8,10,8,12,12,14,12]
|
|
; ALL-NEXT: retq
|
|
%shuffle = shufflevector <16 x float> %a, <16 x float> %b, <16 x i32> <i32 0, i32 0, i32 2, i32 0, i32 4, i32 4, i32 6, i32 4, i32 8, i32 8, i32 10, i32 8, i32 12, i32 12, i32 14, i32 12>
|
|
ret <16 x float> %shuffle
|
|
}
|
|
|
|
define <16 x float> @shuffle_v16f32_03_uu_uu_uu_uu_04_uu_uu_uu_uu_11_uu_uu_uu_uu_12(<16 x float> %a, <16 x float> %b) {
|
|
; ALL-LABEL: shuffle_v16f32_03_uu_uu_uu_uu_04_uu_uu_uu_uu_11_uu_uu_uu_uu_12:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vshufps {{.*#+}} zmm0 = zmm0[3,0,3,0,7,4,7,4,11,8,11,8,15,12,15,12]
|
|
; ALL-NEXT: retq
|
|
%shuffle = shufflevector <16 x float> %a, <16 x float> %b, <16 x i32> <i32 3, i32 poison, i32 poison, i32 poison, i32 poison, i32 4, i32 poison, i32 poison, i32 poison, i32 poison, i32 11, i32 poison, i32 poison, i32 poison, i32 poison, i32 12>
|
|
ret <16 x float> %shuffle
|
|
}
|
|
|
|
; PR41203
|
|
define <16 x float> @shuffle_v16f32_00_17_02_19_04_21_06_23_08_25_10_27_12_29_14_31(<16 x float> %a) {
|
|
; ALL-LABEL: shuffle_v16f32_00_17_02_19_04_21_06_23_08_25_10_27_12_29_14_31:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vandpd {{\.?LCPI[0-9]+_[0-9]+}}(%rip){1to8}, %zmm0, %zmm0
|
|
; ALL-NEXT: retq
|
|
%tmp1 = shufflevector <16 x float> %a, <16 x float> poison, <16 x i32> <i32 poison, i32 17, i32 poison, i32 19, i32 poison, i32 5, i32 poison, i32 7, i32 poison, i32 9, i32 poison, i32 11, i32 poison, i32 13, i32 poison, i32 15>
|
|
%tmp2 = shufflevector <16 x float> %tmp1, <16 x float> <float 0.000000e+00, float poison, float 0.000000e+00, float poison, float 0.000000e+00, float poison, float 0.000000e+00, float poison, float 0.000000e+00, float poison, float 0.000000e+00, float poison, float 0.000000e+00, float poison, float 0.000000e+00, float poison>, <16 x i32> <i32 16, i32 1, i32 18, i32 3, i32 20, i32 5, i32 22, i32 7, i32 24, i32 9, i32 26, i32 11, i32 28, i32 13, i32 30, i32 15>
|
|
ret <16 x float> %tmp2
|
|
}
|
|
|
|
; PR48322
|
|
define <16 x float> @shuffle_v16f32_02_03_16_17_06_07_20_21_10_11_24_25_14_15_28_29(<16 x float> %a, <16 x float> %b) {
|
|
; ALL-LABEL: shuffle_v16f32_02_03_16_17_06_07_20_21_10_11_24_25_14_15_28_29:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vshufpd {{.*#+}} zmm0 = zmm0[1],zmm1[0],zmm0[3],zmm1[2],zmm0[5],zmm1[4],zmm0[7],zmm1[6]
|
|
; ALL-NEXT: retq
|
|
%shuffle = shufflevector <16 x float> %a, <16 x float> %b, <16 x i32> <i32 2, i32 3, i32 16, i32 17, i32 6, i32 7, i32 20, i32 21, i32 10, i32 11, i32 24, i32 25, i32 14, i32 15, i32 28, i32 29>
|
|
ret <16 x float> %shuffle
|
|
}
|
|
|
|
; PR86076
|
|
define <16 x float> @shuffle_f32_v16f32_00_08_01_09_02_10_03_11_04_12_05_13_06_14_07_15(float %a0, float %a1) {
|
|
; ALL-LABEL: shuffle_f32_v16f32_00_08_01_09_02_10_03_11_04_12_05_13_06_14_07_15:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vbroadcastss %xmm0, %ymm0
|
|
; ALL-NEXT: vbroadcastss %xmm1, %ymm1
|
|
; ALL-NEXT: vunpcklps {{.*#+}} ymm0 = ymm0[0],ymm1[0],ymm0[1],ymm1[1],ymm0[4],ymm1[4],ymm0[5],ymm1[5]
|
|
; ALL-NEXT: vinsertf64x4 $1, %ymm0, %zmm0, %zmm0
|
|
; ALL-NEXT: retq
|
|
%v0 = insertelement <8 x float> poison, float %a0, i64 0
|
|
%v1 = insertelement <8 x float> poison, float %a1, i64 0
|
|
%b0 = shufflevector <8 x float> %v0, <8 x float> poison, <8 x i32> zeroinitializer
|
|
%b1 = shufflevector <8 x float> %v1, <8 x float> poison, <8 x i32> zeroinitializer
|
|
%r = shufflevector <8 x float> %b0, <8 x float> %b1, <16 x i32> <i32 0, i32 8, i32 1, i32 9, i32 2, i32 10, i32 3, i32 11, i32 4, i32 12, i32 5, i32 13, i32 6, i32 14, i32 7, i32 15>
|
|
ret <16 x float> %r
|
|
}
|
|
|
|
; PR86076
|
|
define <16 x float> @shuffle_f32_v16f32_00_08_00_08_00_08_00_08_00_08_00_08_00_08_00_08(float %a0, float %a1) {
|
|
; SLOW-LABEL: shuffle_f32_v16f32_00_08_00_08_00_08_00_08_00_08_00_08_00_08_00_08:
|
|
; SLOW: # %bb.0:
|
|
; SLOW-NEXT: vinsertps {{.*#+}} xmm0 = xmm0[0],xmm1[0],zero,zero
|
|
; SLOW-NEXT: vbroadcastsd %xmm0, %zmm0
|
|
; SLOW-NEXT: retq
|
|
;
|
|
; FAST-LABEL: shuffle_f32_v16f32_00_08_00_08_00_08_00_08_00_08_00_08_00_08_00_08:
|
|
; FAST: # %bb.0:
|
|
; FAST-NEXT: # kill: def $xmm1 killed $xmm1 def $zmm1
|
|
; FAST-NEXT: # kill: def $xmm0 killed $xmm0 def $zmm0
|
|
; FAST-NEXT: vmovaps {{.*#+}} ymm2 = [0,16,0,16,0,16,0,16]
|
|
; FAST-NEXT: vpermt2ps %zmm1, %zmm2, %zmm0
|
|
; FAST-NEXT: vinsertf64x4 $1, %ymm0, %zmm0, %zmm0
|
|
; FAST-NEXT: retq
|
|
%v0 = insertelement <8 x float> poison, float %a0, i64 0
|
|
%v1 = insertelement <8 x float> poison, float %a1, i64 0
|
|
%sv = shufflevector <8 x float> %v0, <8 x float> %v1, <16 x i32> <i32 0, i32 8, i32 0, i32 8, i32 0, i32 8, i32 0, i32 8, i32 0, i32 8, i32 0, i32 8, i32 0, i32 8, i32 0, i32 8>
|
|
ret <16 x float> %sv
|
|
}
|
|
|
|
define <16 x i32> @shuffle_v16i32_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00(<16 x i32> %a, <16 x i32> %b) {
|
|
; ALL-LABEL: shuffle_v16i32_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00_00:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vbroadcastss %xmm0, %zmm0
|
|
; ALL-NEXT: retq
|
|
%shuffle = shufflevector <16 x i32> %a, <16 x i32> %b, <16 x i32><i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0>
|
|
ret <16 x i32> %shuffle
|
|
}
|
|
|
|
define <16 x i32> @shuffle_v16i32_04_04_04_04_04_04_04_04_04_04_04_04_04_04_04_04(<16 x i32> %a, <16 x i32> %b) {
|
|
; SLOW-LABEL: shuffle_v16i32_04_04_04_04_04_04_04_04_04_04_04_04_04_04_04_04:
|
|
; SLOW: # %bb.0:
|
|
; SLOW-NEXT: vextractf128 $1, %ymm0, %xmm0
|
|
; SLOW-NEXT: vbroadcastss %xmm0, %zmm0
|
|
; SLOW-NEXT: retq
|
|
;
|
|
; FAST-LABEL: shuffle_v16i32_04_04_04_04_04_04_04_04_04_04_04_04_04_04_04_04:
|
|
; FAST: # %bb.0:
|
|
; FAST-NEXT: vbroadcastss {{.*#+}} zmm1 = [4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4]
|
|
; FAST-NEXT: vpermps %zmm0, %zmm1, %zmm0
|
|
; FAST-NEXT: retq
|
|
%shuffle = shufflevector <16 x i32> %a, <16 x i32> %b, <16 x i32><i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 4>
|
|
ret <16 x i32> %shuffle
|
|
}
|
|
|
|
define <16 x i32> @shuffle_v16i32_02_12_03_13_06_16_07_17_0a_1a_0b_1b_0e_1e_0f_1f(<16 x i32> %a, <16 x i32> %b) {
|
|
; ALL-LABEL: shuffle_v16i32_02_12_03_13_06_16_07_17_0a_1a_0b_1b_0e_1e_0f_1f:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vunpckhps {{.*#+}} zmm0 = zmm0[2],zmm1[2],zmm0[3],zmm1[3],zmm0[6],zmm1[6],zmm0[7],zmm1[7],zmm0[10],zmm1[10],zmm0[11],zmm1[11],zmm0[14],zmm1[14],zmm0[15],zmm1[15]
|
|
; ALL-NEXT: retq
|
|
%shuffle = shufflevector <16 x i32> %a, <16 x i32> %b, <16 x i32><i32 2, i32 18, i32 3, i32 19, i32 6, i32 22, i32 7, i32 23, i32 10, i32 26, i32 11, i32 27, i32 14, i32 30, i32 15, i32 31>
|
|
ret <16 x i32> %shuffle
|
|
}
|
|
|
|
define <16 x i32> @shuffle_v16i32_02_zz_03_zz_06_zz_07_zz_0a_zz_0b_zz_0e_zz_0f_zz(<16 x i32> %a, <16 x i32> %b) {
|
|
; ALL-LABEL: shuffle_v16i32_02_zz_03_zz_06_zz_07_zz_0a_zz_0b_zz_0e_zz_0f_zz:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vxorps %xmm1, %xmm1, %xmm1
|
|
; ALL-NEXT: vunpckhps {{.*#+}} zmm0 = zmm0[2],zmm1[2],zmm0[3],zmm1[3],zmm0[6],zmm1[6],zmm0[7],zmm1[7],zmm0[10],zmm1[10],zmm0[11],zmm1[11],zmm0[14],zmm1[14],zmm0[15],zmm1[15]
|
|
; ALL-NEXT: retq
|
|
%shuffle = shufflevector <16 x i32> %a, <16 x i32> zeroinitializer, <16 x i32><i32 2, i32 30, i32 3, i32 28, i32 6, i32 26, i32 7, i32 24, i32 10, i32 22, i32 11, i32 20, i32 14, i32 18, i32 15, i32 16>
|
|
ret <16 x i32> %shuffle
|
|
}
|
|
|
|
define <16 x i32> @shuffle_v16i32_01_02_03_16_05_06_07_20_09_10_11_24_13_14_15_28(<16 x i32> %a, <16 x i32> %b) {
|
|
; AVX512F-LABEL: shuffle_v16i32_01_02_03_16_05_06_07_20_09_10_11_24_13_14_15_28:
|
|
; AVX512F: # %bb.0:
|
|
; AVX512F-NEXT: vmovdqa64 {{.*#+}} zmm2 = [1,2,3,16,5,6,7,20,9,10,11,24,13,14,15,28]
|
|
; AVX512F-NEXT: vpermt2d %zmm1, %zmm2, %zmm0
|
|
; AVX512F-NEXT: retq
|
|
;
|
|
; AVX512BW-LABEL: shuffle_v16i32_01_02_03_16_05_06_07_20_09_10_11_24_13_14_15_28:
|
|
; AVX512BW: # %bb.0:
|
|
; AVX512BW-NEXT: vpalignr {{.*#+}} zmm0 = zmm0[4,5,6,7,8,9,10,11,12,13,14,15],zmm1[0,1,2,3],zmm0[20,21,22,23,24,25,26,27,28,29,30,31],zmm1[16,17,18,19],zmm0[36,37,38,39,40,41,42,43,44,45,46,47],zmm1[32,33,34,35],zmm0[52,53,54,55,56,57,58,59,60,61,62,63],zmm1[48,49,50,51]
|
|
; AVX512BW-NEXT: retq
|
|
%shuffle = shufflevector <16 x i32> %a, <16 x i32> %b, <16 x i32><i32 1, i32 2, i32 3, i32 16, i32 5, i32 6, i32 7, i32 20, i32 9, i32 10, i32 11, i32 24, i32 13, i32 14, i32 15, i32 28>
|
|
ret <16 x i32> %shuffle
|
|
}
|
|
|
|
define <16 x float> @shuffle_v16f32_02_05_u_u_07_u_0a_01_00_05_u_04_07_u_0a_01(<16 x float> %a) {
|
|
; ALL-LABEL: shuffle_v16f32_02_05_u_u_07_u_0a_01_00_05_u_04_07_u_0a_01:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vmovaps {{.*#+}} zmm1 = [2,5,u,u,7,u,10,1,0,5,u,4,7,u,10,1]
|
|
; ALL-NEXT: vpermps %zmm0, %zmm1, %zmm0
|
|
; ALL-NEXT: retq
|
|
%c = shufflevector <16 x float> %a, <16 x float> poison, <16 x i32> <i32 2, i32 5, i32 poison, i32 poison, i32 7, i32 poison, i32 10, i32 1, i32 0, i32 5, i32 poison, i32 4, i32 7, i32 poison, i32 10, i32 1>
|
|
ret <16 x float> %c
|
|
}
|
|
|
|
define <16 x i32> @shuffle_v16i32_02_05_u_u_07_u_0a_01_00_05_u_04_07_u_0a_01(<16 x i32> %a) {
|
|
; ALL-LABEL: shuffle_v16i32_02_05_u_u_07_u_0a_01_00_05_u_04_07_u_0a_01:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vmovaps {{.*#+}} zmm1 = [2,5,u,u,7,u,10,1,0,5,u,4,7,u,10,1]
|
|
; ALL-NEXT: vpermps %zmm0, %zmm1, %zmm0
|
|
; ALL-NEXT: retq
|
|
%c = shufflevector <16 x i32> %a, <16 x i32> poison, <16 x i32> <i32 2, i32 5, i32 poison, i32 poison, i32 7, i32 poison, i32 10, i32 1, i32 0, i32 5, i32 poison, i32 4, i32 7, i32 poison, i32 10, i32 1>
|
|
ret <16 x i32> %c
|
|
}
|
|
|
|
define <16 x i32> @shuffle_v16i32_0f_1f_0e_16_0d_1d_04_1e_0b_1b_0a_1a_09_19_08_18(<16 x i32> %a, <16 x i32> %b) {
|
|
; ALL-LABEL: shuffle_v16i32_0f_1f_0e_16_0d_1d_04_1e_0b_1b_0a_1a_09_19_08_18:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vmovdqa64 {{.*#+}} zmm2 = [15,31,14,22,13,29,4,28,11,27,10,26,9,25,8,24]
|
|
; ALL-NEXT: vpermt2d %zmm1, %zmm2, %zmm0
|
|
; ALL-NEXT: retq
|
|
%c = shufflevector <16 x i32> %a, <16 x i32> %b, <16 x i32> <i32 15, i32 31, i32 14, i32 22, i32 13, i32 29, i32 4, i32 28, i32 11, i32 27, i32 10, i32 26, i32 9, i32 25, i32 8, i32 24>
|
|
ret <16 x i32> %c
|
|
}
|
|
|
|
define <16 x float> @shuffle_v16f32_0f_1f_0e_16_0d_1d_04_1e_0b_1b_0a_1a_09_19_08_18(<16 x float> %a, <16 x float> %b) {
|
|
; ALL-LABEL: shuffle_v16f32_0f_1f_0e_16_0d_1d_04_1e_0b_1b_0a_1a_09_19_08_18:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vmovaps {{.*#+}} zmm2 = [15,31,14,22,13,29,4,28,11,27,10,26,9,25,8,24]
|
|
; ALL-NEXT: vpermt2ps %zmm1, %zmm2, %zmm0
|
|
; ALL-NEXT: retq
|
|
%c = shufflevector <16 x float> %a, <16 x float> %b, <16 x i32> <i32 15, i32 31, i32 14, i32 22, i32 13, i32 29, i32 4, i32 28, i32 11, i32 27, i32 10, i32 26, i32 9, i32 25, i32 8, i32 24>
|
|
ret <16 x float> %c
|
|
}
|
|
|
|
; PR46249
|
|
define <16 x i32> @shuffle_v16i32_0b_0a_09_08_0f_0e_0d_0c_03_02_01_00_07_06_05_04(<16 x i32> %a) {
|
|
; SLOW-LABEL: shuffle_v16i32_0b_0a_09_08_0f_0e_0d_0c_03_02_01_00_07_06_05_04:
|
|
; SLOW: # %bb.0:
|
|
; SLOW-NEXT: vpshufd {{.*#+}} zmm0 = zmm0[3,2,1,0,7,6,5,4,11,10,9,8,15,14,13,12]
|
|
; SLOW-NEXT: vshufi64x2 {{.*#+}} zmm0 = zmm0[4,5,6,7,0,1,2,3]
|
|
; SLOW-NEXT: retq
|
|
;
|
|
; FAST-LABEL: shuffle_v16i32_0b_0a_09_08_0f_0e_0d_0c_03_02_01_00_07_06_05_04:
|
|
; FAST: # %bb.0:
|
|
; FAST-NEXT: vmovaps {{.*#+}} zmm1 = [11,10,9,8,15,14,13,12,3,2,1,0,7,6,5,4]
|
|
; FAST-NEXT: vpermps %zmm0, %zmm1, %zmm0
|
|
; FAST-NEXT: retq
|
|
%1 = shufflevector <16 x i32> %a, <16 x i32> poison, <16 x i32> <i32 11, i32 10, i32 9, i32 8, i32 15, i32 14, i32 13, i32 12, i32 3, i32 2, i32 1, i32 0, i32 7, i32 6, i32 5, i32 4>
|
|
ret <16 x i32> %1
|
|
}
|
|
|
|
define <16 x float> @shuffle_v16f32_0b_0a_09_08_0f_0e_0d_0c_03_02_01_00_07_06_05_04(<16 x float> %a) {
|
|
; SLOW-LABEL: shuffle_v16f32_0b_0a_09_08_0f_0e_0d_0c_03_02_01_00_07_06_05_04:
|
|
; SLOW: # %bb.0:
|
|
; SLOW-NEXT: vshufps {{.*#+}} zmm0 = zmm0[3,2,1,0,7,6,5,4,11,10,9,8,15,14,13,12]
|
|
; SLOW-NEXT: vshuff64x2 {{.*#+}} zmm0 = zmm0[4,5,6,7,0,1,2,3]
|
|
; SLOW-NEXT: retq
|
|
;
|
|
; FAST-LABEL: shuffle_v16f32_0b_0a_09_08_0f_0e_0d_0c_03_02_01_00_07_06_05_04:
|
|
; FAST: # %bb.0:
|
|
; FAST-NEXT: vmovaps {{.*#+}} zmm1 = [11,10,9,8,15,14,13,12,3,2,1,0,7,6,5,4]
|
|
; FAST-NEXT: vpermps %zmm0, %zmm1, %zmm0
|
|
; FAST-NEXT: retq
|
|
%1 = shufflevector <16 x float> %a, <16 x float> poison, <16 x i32> <i32 11, i32 10, i32 9, i32 8, i32 15, i32 14, i32 13, i32 12, i32 3, i32 2, i32 1, i32 0, i32 7, i32 6, i32 5, i32 4>
|
|
ret <16 x float> %1
|
|
}
|
|
|
|
define <16 x float> @shuffle_v16f32_load_0f_1f_0e_16_0d_1d_04_1e_0b_1b_0a_1a_09_19_08_18(<16 x float> %a, ptr %b) {
|
|
; ALL-LABEL: shuffle_v16f32_load_0f_1f_0e_16_0d_1d_04_1e_0b_1b_0a_1a_09_19_08_18:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vmovaps {{.*#+}} zmm1 = [15,31,14,22,13,29,4,28,11,27,10,26,9,25,8,24]
|
|
; ALL-NEXT: vpermt2ps (%rdi), %zmm1, %zmm0
|
|
; ALL-NEXT: retq
|
|
%c = load <16 x float>, ptr %b
|
|
%d = shufflevector <16 x float> %a, <16 x float> %c, <16 x i32> <i32 15, i32 31, i32 14, i32 22, i32 13, i32 29, i32 4, i32 28, i32 11, i32 27, i32 10, i32 26, i32 9, i32 25, i32 8, i32 24>
|
|
ret <16 x float> %d
|
|
}
|
|
|
|
define <16 x float> @shuffle_v16f32_load_08_11_10_00_12_15_14_04(<16 x float> %a0, ptr %a1) {
|
|
; SLOW-LABEL: shuffle_v16f32_load_08_11_10_00_12_15_14_04:
|
|
; SLOW: # %bb.0:
|
|
; SLOW-NEXT: vshufps {{.*#+}} zmm1 = zmm0[2,0],mem[0,0],zmm0[6,4],mem[4,4],zmm0[10,8],mem[8,8],zmm0[14,12],mem[12,12]
|
|
; SLOW-NEXT: vshufps {{.*#+}} zmm0 = zmm0[0,3],zmm1[0,2],zmm0[4,7],zmm1[4,6],zmm0[8,11],zmm1[8,10],zmm0[12,15],zmm1[12,14]
|
|
; SLOW-NEXT: retq
|
|
;
|
|
; FAST-LABEL: shuffle_v16f32_load_08_11_10_00_12_15_14_04:
|
|
; FAST: # %bb.0:
|
|
; FAST-NEXT: vmovaps {{.*#+}} zmm1 = [0,3,2,16,4,7,6,20,8,11,10,24,12,15,14,28]
|
|
; FAST-NEXT: vpermt2ps (%rdi), %zmm1, %zmm0
|
|
; FAST-NEXT: retq
|
|
%1 = load <16 x float>, ptr %a1
|
|
%2 = shufflevector <16 x float> %1, <16 x float> %a0, <16 x i32> <i32 16, i32 19, i32 18, i32 0, i32 20, i32 23, i32 22, i32 4, i32 24, i32 27, i32 26, i32 8, i32 28, i32 31, i32 30, i32 12>
|
|
ret <16 x float> %2
|
|
}
|
|
|
|
define <16 x i32> @shuffle_v16i32_load_0f_1f_0e_16_0d_1d_04_1e_0b_1b_0a_1a_09_19_08_18(<16 x i32> %a, ptr %b) {
|
|
; ALL-LABEL: shuffle_v16i32_load_0f_1f_0e_16_0d_1d_04_1e_0b_1b_0a_1a_09_19_08_18:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vmovdqa64 {{.*#+}} zmm1 = [15,31,14,22,13,29,4,28,11,27,10,26,9,25,8,24]
|
|
; ALL-NEXT: vpermt2d (%rdi), %zmm1, %zmm0
|
|
; ALL-NEXT: retq
|
|
%c = load <16 x i32>, ptr %b
|
|
%d = shufflevector <16 x i32> %a, <16 x i32> %c, <16 x i32> <i32 15, i32 31, i32 14, i32 22, i32 13, i32 29, i32 4, i32 28, i32 11, i32 27, i32 10, i32 26, i32 9, i32 25, i32 8, i32 24>
|
|
ret <16 x i32> %d
|
|
}
|
|
|
|
define <16 x i32> @shuffle_v16i32_0_1_2_19_u_u_u_u_u_u_u_u_u_u_u_u(<16 x i32> %a, <16 x i32> %b) {
|
|
; ALL-LABEL: shuffle_v16i32_0_1_2_19_u_u_u_u_u_u_u_u_u_u_u_u:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vblendps {{.*#+}} xmm0 = xmm0[0,1,2],xmm1[3]
|
|
; ALL-NEXT: retq
|
|
%c = shufflevector <16 x i32> %a, <16 x i32> %b, <16 x i32> <i32 0, i32 1, i32 2, i32 19, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
|
|
ret <16 x i32> %c
|
|
}
|
|
|
|
;FIXME: can do better with vpcompress
|
|
define <8 x i32> @test_v16i32_1_3_5_7_9_11_13_15(<16 x i32> %v) {
|
|
; SLOW-LABEL: test_v16i32_1_3_5_7_9_11_13_15:
|
|
; SLOW: # %bb.0:
|
|
; SLOW-NEXT: vextractf64x4 $1, %zmm0, %ymm1
|
|
; SLOW-NEXT: vshufps {{.*#+}} ymm0 = ymm0[1,3],ymm1[1,3],ymm0[5,7],ymm1[5,7]
|
|
; SLOW-NEXT: vpermpd {{.*#+}} ymm0 = ymm0[0,2,1,3]
|
|
; SLOW-NEXT: retq
|
|
;
|
|
; FAST-LABEL: test_v16i32_1_3_5_7_9_11_13_15:
|
|
; FAST: # %bb.0:
|
|
; FAST-NEXT: vmovaps {{.*#+}} ymm1 = [1,3,5,7,9,11,13,15]
|
|
; FAST-NEXT: vpermps %zmm0, %zmm1, %zmm0
|
|
; FAST-NEXT: # kill: def $ymm0 killed $ymm0 killed $zmm0
|
|
; FAST-NEXT: retq
|
|
%res = shufflevector <16 x i32> %v, <16 x i32> poison, <8 x i32> <i32 1, i32 3, i32 5, i32 7, i32 9, i32 11, i32 13, i32 15>
|
|
ret <8 x i32> %res
|
|
}
|
|
|
|
;FIXME: can do better with vpcompress
|
|
define <4 x i32> @test_v16i32_0_1_2_12 (<16 x i32> %v) {
|
|
; SLOW-LABEL: test_v16i32_0_1_2_12:
|
|
; SLOW: # %bb.0:
|
|
; SLOW-NEXT: vextractf32x4 $3, %zmm0, %xmm1
|
|
; SLOW-NEXT: vbroadcastss %xmm1, %xmm1
|
|
; SLOW-NEXT: vblendps {{.*#+}} xmm0 = xmm0[0,1,2],xmm1[3]
|
|
; SLOW-NEXT: vzeroupper
|
|
; SLOW-NEXT: retq
|
|
;
|
|
; FAST-LABEL: test_v16i32_0_1_2_12:
|
|
; FAST: # %bb.0:
|
|
; FAST-NEXT: vmovaps {{.*#+}} xmm1 = [0,1,2,12]
|
|
; FAST-NEXT: vpermps %zmm0, %zmm1, %zmm0
|
|
; FAST-NEXT: # kill: def $xmm0 killed $xmm0 killed $zmm0
|
|
; FAST-NEXT: vzeroupper
|
|
; FAST-NEXT: retq
|
|
%res = shufflevector <16 x i32> %v, <16 x i32> poison, <4 x i32> <i32 0, i32 1, i32 2, i32 12>
|
|
ret <4 x i32> %res
|
|
}
|
|
|
|
;PR31451
|
|
;FIXME: can do better with vpcompress
|
|
define <4 x i32> @test_v16i32_0_4_8_12(<16 x i32> %v) {
|
|
; ALL-LABEL: test_v16i32_0_4_8_12:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vmovaps {{.*#+}} xmm1 = [0,4,8,12]
|
|
; ALL-NEXT: vpermps %zmm0, %zmm1, %zmm0
|
|
; ALL-NEXT: # kill: def $xmm0 killed $xmm0 killed $zmm0
|
|
; ALL-NEXT: vzeroupper
|
|
; ALL-NEXT: retq
|
|
%res = shufflevector <16 x i32> %v, <16 x i32> poison, <4 x i32> <i32 0, i32 4, i32 8, i32 12>
|
|
ret <4 x i32> %res
|
|
}
|
|
|
|
define <8 x float> @shuffle_v16f32_extract_256(ptr %RET, ptr %a) {
|
|
; ALL-LABEL: shuffle_v16f32_extract_256:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vmovups 32(%rsi), %ymm0
|
|
; ALL-NEXT: retq
|
|
%v_a = load <16 x float>, ptr %a, align 4
|
|
%v2 = shufflevector <16 x float> %v_a, <16 x float> poison, <8 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
|
|
ret <8 x float> %v2
|
|
}
|
|
|
|
;FIXME: can do better with vcompressp
|
|
define <8 x float> @test_v16f32_0_1_2_3_4_6_7_10 (<16 x float> %v) {
|
|
; ALL-LABEL: test_v16f32_0_1_2_3_4_6_7_10:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vmovaps {{.*#+}} ymm1 = [0,1,2,3,4,6,7,10]
|
|
; ALL-NEXT: vpermps %zmm0, %zmm1, %zmm0
|
|
; ALL-NEXT: # kill: def $ymm0 killed $ymm0 killed $zmm0
|
|
; ALL-NEXT: retq
|
|
%res = shufflevector <16 x float> %v, <16 x float> poison, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 6, i32 7, i32 10>
|
|
ret <8 x float> %res
|
|
}
|
|
|
|
;FIXME: can do better with vcompressp
|
|
define <4 x float> @test_v16f32_0_1_3_6 (<16 x float> %v) {
|
|
; ALL-LABEL: test_v16f32_0_1_3_6:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vmovaps {{.*#+}} xmm1 = [0,1,3,6]
|
|
; ALL-NEXT: vpermps %ymm0, %ymm1, %ymm0
|
|
; ALL-NEXT: # kill: def $xmm0 killed $xmm0 killed $ymm0
|
|
; ALL-NEXT: vzeroupper
|
|
; ALL-NEXT: retq
|
|
%res = shufflevector <16 x float> %v, <16 x float> poison, <4 x i32> <i32 0, i32 1, i32 3, i32 6>
|
|
ret <4 x float> %res
|
|
}
|
|
|
|
define <16 x i32> @shuffle_v16i16_1_0_0_0_5_4_4_4_9_8_8_8_13_12_12_12(<16 x i32> %a, <16 x i32> %b) {
|
|
; ALL-LABEL: shuffle_v16i16_1_0_0_0_5_4_4_4_9_8_8_8_13_12_12_12:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vshufps {{.*#+}} zmm0 = zmm0[1,0,0,0,5,4,4,4,9,8,8,8,13,12,12,12]
|
|
; ALL-NEXT: retq
|
|
%c = shufflevector <16 x i32> %a, <16 x i32> %b, <16 x i32> <i32 1, i32 0, i32 0, i32 0, i32 5, i32 4, i32 4, i32 4, i32 9, i32 8, i32 8, i32 8, i32 13, i32 12, i32 12, i32 12>
|
|
ret <16 x i32> %c
|
|
}
|
|
|
|
define <16 x i32> @shuffle_v16i16_3_3_0_0_7_7_4_4_11_11_8_8_15_15_12_12(<16 x i32> %a, <16 x i32> %b) {
|
|
; ALL-LABEL: shuffle_v16i16_3_3_0_0_7_7_4_4_11_11_8_8_15_15_12_12:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vshufps {{.*#+}} zmm0 = zmm0[2,3,0,1,6,7,4,5,10,11,8,9,14,15,12,13]
|
|
; ALL-NEXT: retq
|
|
%c = shufflevector <16 x i32> %a, <16 x i32> %b, <16 x i32> <i32 2, i32 3, i32 0, i32 1, i32 6, i32 7, i32 4, i32 5, i32 10, i32 11, i32 8, i32 9, i32 14, i32 15, i32 12, i32 13>
|
|
ret <16 x i32> %c
|
|
}
|
|
|
|
define <16 x float> @shuffle_v16f32_00_01_10_10_04_05_14_14_08_09_18_18_0c_0d_1c_1c(<16 x float> %a, <16 x float> %b) {
|
|
; ALL-LABEL: shuffle_v16f32_00_01_10_10_04_05_14_14_08_09_18_18_0c_0d_1c_1c:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vshufps {{.*#+}} zmm0 = zmm0[0,1],zmm1[0,0],zmm0[4,5],zmm1[4,4],zmm0[8,9],zmm1[8,8],zmm0[12,13],zmm1[12,12]
|
|
; ALL-NEXT: retq
|
|
%shuffle = shufflevector <16 x float> %a, <16 x float> %b, <16 x i32> <i32 0, i32 1, i32 16, i32 16, i32 4, i32 5, i32 20, i32 20, i32 8, i32 9, i32 24, i32 24, i32 12, i32 13, i32 28, i32 28>
|
|
ret <16 x float> %shuffle
|
|
}
|
|
|
|
define <16 x i32> @insert_mem_and_zero_v16i32(ptr %ptr) {
|
|
; ALL-LABEL: insert_mem_and_zero_v16i32:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vmovss {{.*#+}} xmm0 = mem[0],zero,zero,zero
|
|
; ALL-NEXT: retq
|
|
%a = load i32, ptr %ptr
|
|
%v = insertelement <16 x i32> poison, i32 %a, i32 0
|
|
%shuffle = shufflevector <16 x i32> %v, <16 x i32> zeroinitializer, <16 x i32> <i32 0, i32 17, i32 18, i32 19, i32 20, i32 21, i32 22, i32 23, i32 24, i32 25, i32 26, i32 27, i32 28, i32 29, i32 30, i32 31>
|
|
ret <16 x i32> %shuffle
|
|
}
|
|
|
|
|
|
define <16 x i32> @shuffle_v16i32_0zzzzzzzzzzzzzzz(<16 x i32> %a) {
|
|
; ALL-LABEL: shuffle_v16i32_0zzzzzzzzzzzzzzz:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vxorps %xmm1, %xmm1, %xmm1
|
|
; ALL-NEXT: vblendps {{.*#+}} xmm0 = xmm0[0],xmm1[1,2,3]
|
|
; ALL-NEXT: retq
|
|
%shuffle = shufflevector <16 x i32> %a, <16 x i32> zeroinitializer, <16 x i32> <i32 0, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16>
|
|
ret <16 x i32> %shuffle
|
|
}
|
|
|
|
define <16 x float> @shuffle_v16f32_0zzzzzzzzzzzzzzz(<16 x float> %a) {
|
|
; ALL-LABEL: shuffle_v16f32_0zzzzzzzzzzzzzzz:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vxorps %xmm1, %xmm1, %xmm1
|
|
; ALL-NEXT: vblendps {{.*#+}} xmm0 = xmm0[0],xmm1[1,2,3]
|
|
; ALL-NEXT: retq
|
|
%shuffle = shufflevector <16 x float> %a, <16 x float> zeroinitializer, <16 x i32> <i32 0, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16, i32 16>
|
|
ret <16 x float> %shuffle
|
|
}
|
|
|
|
define <16 x i32> @shuffle_v16i32_16_zz_17_zz_18_zz_19_zz_20_zz_21_zz_22_zz_23_zz(<16 x i32> %a) {
|
|
; ALL-LABEL: shuffle_v16i32_16_zz_17_zz_18_zz_19_zz_20_zz_21_zz_22_zz_23_zz:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vpmovzxdq {{.*#+}} zmm0 = ymm0[0],zero,ymm0[1],zero,ymm0[2],zero,ymm0[3],zero,ymm0[4],zero,ymm0[5],zero,ymm0[6],zero,ymm0[7],zero
|
|
; ALL-NEXT: retq
|
|
%shuffle = shufflevector <16 x i32> zeroinitializer, <16 x i32> %a, <16 x i32> <i32 16, i32 0, i32 17, i32 0, i32 18, i32 0, i32 19, i32 0, i32 20, i32 0, i32 21, i32 0, i32 22, i32 0, i32 23, i32 0>
|
|
ret <16 x i32> %shuffle
|
|
}
|
|
|
|
define <16 x i32> @shuffle_v16i32_01_02_03_04_05_06_07_08_09_10_11_12_13_14_15_16(<16 x i32> %a, <16 x i32> %b) {
|
|
; ALL-LABEL: shuffle_v16i32_01_02_03_04_05_06_07_08_09_10_11_12_13_14_15_16:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: valignd {{.*#+}} zmm0 = zmm0[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15],zmm1[0]
|
|
; ALL-NEXT: retq
|
|
%shuffle = shufflevector <16 x i32> %a, <16 x i32> %b, <16 x i32><i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 16>
|
|
ret <16 x i32> %shuffle
|
|
}
|
|
|
|
define <16 x i32> @shuffle_v16i32_01_02_03_04_05_06_07_08_09_10_11_12_13_14_15_00(<16 x i32> %a) {
|
|
; ALL-LABEL: shuffle_v16i32_01_02_03_04_05_06_07_08_09_10_11_12_13_14_15_00:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: valignd {{.*#+}} zmm0 = zmm0[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,0]
|
|
; ALL-NEXT: retq
|
|
%shuffle = shufflevector <16 x i32> %a, <16 x i32> poison, <16 x i32><i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 0>
|
|
ret <16 x i32> %shuffle
|
|
}
|
|
|
|
define <16 x i32> @shuffle_v16i32_00_03_16_19_04_07_20_23_08_11_24_27_12_15_28_31(<16 x i32> %a, <16 x i32> %b) {
|
|
; ALL-LABEL: shuffle_v16i32_00_03_16_19_04_07_20_23_08_11_24_27_12_15_28_31:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vshufps {{.*#+}} zmm0 = zmm0[0,3],zmm1[0,3],zmm0[4,7],zmm1[4,7],zmm0[8,11],zmm1[8,11],zmm0[12,15],zmm1[12,15]
|
|
; ALL-NEXT: retq
|
|
%shuffle = shufflevector <16 x i32> %a, <16 x i32> %b, <16 x i32> <i32 0, i32 3, i32 16, i32 19, i32 4, i32 7, i32 20, i32 23, i32 8, i32 11, i32 24, i32 27, i32 12, i32 15, i32 28, i32 31>
|
|
ret <16 x i32> %shuffle
|
|
}
|
|
|
|
define <16 x i32> @shuffle_v16i32_16_16_02_03_20_20_06_07_24_24_10_11_28_28_uu_uu(<16 x i32> %a, <16 x i32> %b) {
|
|
; ALL-LABEL: shuffle_v16i32_16_16_02_03_20_20_06_07_24_24_10_11_28_28_uu_uu:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vshufps {{.*#+}} zmm0 = zmm1[0,0],zmm0[2,3],zmm1[4,4],zmm0[6,7],zmm1[8,8],zmm0[10,11],zmm1[12,12],zmm0[14,15]
|
|
; ALL-NEXT: retq
|
|
%shuffle = shufflevector <16 x i32> %a, <16 x i32> %b, <16 x i32> <i32 16, i32 16, i32 02, i32 03, i32 20, i32 20, i32 06, i32 07, i32 24, i32 24, i32 10, i32 11, i32 28, i32 28, i32 poison, i32 poison>
|
|
ret <16 x i32> %shuffle
|
|
}
|
|
|
|
; PR48322
|
|
define <16 x i32> @shuffle_v16i32_02_03_16_17_06_07_20_21_10_11_24_25_14_15_28_29(<16 x i32> %a, <16 x i32> %b) {
|
|
; AVX512F-LABEL: shuffle_v16i32_02_03_16_17_06_07_20_21_10_11_24_25_14_15_28_29:
|
|
; AVX512F: # %bb.0:
|
|
; AVX512F-NEXT: vmovdqa64 {{.*#+}} zmm2 = [1,8,3,10,5,12,7,14]
|
|
; AVX512F-NEXT: vpermt2q %zmm1, %zmm2, %zmm0
|
|
; AVX512F-NEXT: retq
|
|
;
|
|
; AVX512BW-LABEL: shuffle_v16i32_02_03_16_17_06_07_20_21_10_11_24_25_14_15_28_29:
|
|
; AVX512BW: # %bb.0:
|
|
; AVX512BW-NEXT: vpalignr {{.*#+}} zmm0 = zmm0[8,9,10,11,12,13,14,15],zmm1[0,1,2,3,4,5,6,7],zmm0[24,25,26,27,28,29,30,31],zmm1[16,17,18,19,20,21,22,23],zmm0[40,41,42,43,44,45,46,47],zmm1[32,33,34,35,36,37,38,39],zmm0[56,57,58,59,60,61,62,63],zmm1[48,49,50,51,52,53,54,55]
|
|
; AVX512BW-NEXT: retq
|
|
%shuffle = shufflevector <16 x i32> %a, <16 x i32> %b, <16 x i32> <i32 2, i32 3, i32 16, i32 17, i32 6, i32 7, i32 20, i32 21, i32 10, i32 11, i32 24, i32 25, i32 14, i32 15, i32 28, i32 29>
|
|
ret <16 x i32> %shuffle
|
|
}
|
|
|
|
define <16 x i32> @shuffle_v8i32_17_16_01_00_21_20_05_04_25_24_09_08_29_28_13_12(<16 x i32> %a, <16 x i32> %b) {
|
|
; ALL-LABEL: shuffle_v8i32_17_16_01_00_21_20_05_04_25_24_09_08_29_28_13_12:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vshufps {{.*#+}} zmm0 = zmm1[1,0],zmm0[1,0],zmm1[5,4],zmm0[5,4],zmm1[9,8],zmm0[9,8],zmm1[13,12],zmm0[13,12]
|
|
; ALL-NEXT: retq
|
|
%shuffle = shufflevector <16 x i32> %a, <16 x i32> %b, <16 x i32> <i32 17, i32 16, i32 01, i32 00, i32 21, i32 20, i32 05, i32 04, i32 25, i32 24, i32 09, i32 08, i32 29, i32 28, i32 13, i32 12>
|
|
ret <16 x i32> %shuffle
|
|
}
|
|
|
|
define <16 x float> @shuffle_v8f32_v16f32_04_04_04_04_04_04_04_04_04_04_04_04_04_04_04_04(<8 x float> %a) {
|
|
; SLOW-LABEL: shuffle_v8f32_v16f32_04_04_04_04_04_04_04_04_04_04_04_04_04_04_04_04:
|
|
; SLOW: # %bb.0:
|
|
; SLOW-NEXT: vextractf128 $1, %ymm0, %xmm0
|
|
; SLOW-NEXT: vbroadcastss %xmm0, %zmm0
|
|
; SLOW-NEXT: retq
|
|
;
|
|
; FAST-LABEL: shuffle_v8f32_v16f32_04_04_04_04_04_04_04_04_04_04_04_04_04_04_04_04:
|
|
; FAST: # %bb.0:
|
|
; FAST-NEXT: # kill: def $ymm0 killed $ymm0 def $zmm0
|
|
; FAST-NEXT: vbroadcastss {{.*#+}} zmm1 = [4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4]
|
|
; FAST-NEXT: vpermps %zmm0, %zmm1, %zmm0
|
|
; FAST-NEXT: retq
|
|
%shuffle = shufflevector <8 x float> %a, <8 x float> poison, <16 x i32> <i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 4, i32 4>
|
|
ret <16 x float> %shuffle
|
|
}
|
|
|
|
define <16 x float> @insert_sub0_0(<16 x float> %base, <4 x float> %sub1, <4 x float> %sub2, <4 x float> %sub3, <4 x float> %sub4) {
|
|
; ALL-LABEL: insert_sub0_0:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vinsertf32x4 $0, %xmm1, %zmm0, %zmm0
|
|
; ALL-NEXT: retq
|
|
%sub12 = shufflevector <4 x float> %sub1, <4 x float> %sub2, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
|
|
%sub34 = shufflevector <4 x float> %sub3, <4 x float> %sub4, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
|
|
%sub1234 = shufflevector <8 x float> %sub12, <8 x float> %sub34, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
|
|
%res = shufflevector <16 x float> %base, <16 x float> %sub1234, <16 x i32> <i32 16, i32 17, i32 18, i32 19, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
|
|
ret <16 x float> %res
|
|
}
|
|
|
|
define <16 x float> @insert_sub1_12(<16 x float> %base, <4 x float> %sub1, <4 x float> %sub2, <4 x float> %sub3, <4 x float> %sub4) {
|
|
; ALL-LABEL: insert_sub1_12:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vinsertf32x4 $3, %xmm2, %zmm0, %zmm0
|
|
; ALL-NEXT: retq
|
|
%sub12 = shufflevector <4 x float> %sub1, <4 x float> %sub2, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
|
|
%sub34 = shufflevector <4 x float> %sub3, <4 x float> %sub4, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
|
|
%sub1234 = shufflevector <8 x float> %sub12, <8 x float> %sub34, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
|
|
%res = shufflevector <16 x float> %base, <16 x float> %sub1234, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i32 10, i32 11, i32 20, i32 21, i32 22, i32 23>
|
|
ret <16 x float> %res
|
|
}
|
|
|
|
define <16 x float> @insert_sub2_4(<16 x float> %base, <4 x float> %sub1, <4 x float> %sub2, <4 x float> %sub3, <4 x float> %sub4) {
|
|
; ALL-LABEL: insert_sub2_4:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vinsertf32x4 $1, %xmm3, %zmm0, %zmm0
|
|
; ALL-NEXT: retq
|
|
%sub12 = shufflevector <4 x float> %sub1, <4 x float> %sub2, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
|
|
%sub34 = shufflevector <4 x float> %sub3, <4 x float> %sub4, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
|
|
%sub1234 = shufflevector <8 x float> %sub12, <8 x float> %sub34, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
|
|
%res = shufflevector <16 x float> %base, <16 x float> %sub1234, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 24, i32 25, i32 26, i32 27, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
|
|
ret <16 x float> %res
|
|
}
|
|
|
|
define <16 x float> @insert_sub01_8(<16 x float> %base, <4 x float> %sub1, <4 x float> %sub2, <4 x float> %sub3, <4 x float> %sub4) {
|
|
; ALL-LABEL: insert_sub01_8:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: # kill: def $xmm1 killed $xmm1 def $ymm1
|
|
; ALL-NEXT: vinsertf128 $1, %xmm2, %ymm1, %ymm1
|
|
; ALL-NEXT: vinsertf64x4 $1, %ymm1, %zmm0, %zmm0
|
|
; ALL-NEXT: retq
|
|
%sub12 = shufflevector <4 x float> %sub1, <4 x float> %sub2, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
|
|
%sub34 = shufflevector <4 x float> %sub3, <4 x float> %sub4, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
|
|
%sub1234 = shufflevector <8 x float> %sub12, <8 x float> %sub34, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
|
|
%res = shufflevector <16 x float> %base, <16 x float> %sub1234, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 16, i32 17, i32 18, i32 19, i32 20, i32 21, i32 22, i32 23>
|
|
ret <16 x float> %res
|
|
}
|
|
|
|
define <16 x float> @insert_sub23_0(<16 x float> %base, <4 x float> %sub1, <4 x float> %sub2, <4 x float> %sub3, <4 x float> %sub4) {
|
|
; ALL-LABEL: insert_sub23_0:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: # kill: def $xmm3 killed $xmm3 def $ymm3
|
|
; ALL-NEXT: vinsertf128 $1, %xmm4, %ymm3, %ymm1
|
|
; ALL-NEXT: vinsertf64x4 $0, %ymm1, %zmm0, %zmm0
|
|
; ALL-NEXT: retq
|
|
%sub12 = shufflevector <4 x float> %sub1, <4 x float> %sub2, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
|
|
%sub34 = shufflevector <4 x float> %sub3, <4 x float> %sub4, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
|
|
%sub1234 = shufflevector <8 x float> %sub12, <8 x float> %sub34, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
|
|
%res = shufflevector <16 x float> %base, <16 x float> %sub1234, <16 x i32> <i32 24, i32 25, i32 26, i32 27, i32 28, i32 29, i32 30, i32 31, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
|
|
ret <16 x float> %res
|
|
}
|
|
|
|
define <16 x i32> @mask_shuffle_v16i32_02_03_04_05_06_07_08_09_10_11_12_13_14_15_00_01(<16 x i32> %a, <16 x i32> %passthru, i16 %mask) {
|
|
; AVX512F-LABEL: mask_shuffle_v16i32_02_03_04_05_06_07_08_09_10_11_12_13_14_15_00_01:
|
|
; AVX512F: # %bb.0:
|
|
; AVX512F-NEXT: kmovw %edi, %k1
|
|
; AVX512F-NEXT: valignd {{.*#+}} zmm1 {%k1} = zmm0[2,3,4,5,6,7,8,9,10,11,12,13,14,15,0,1]
|
|
; AVX512F-NEXT: vmovdqa64 %zmm1, %zmm0
|
|
; AVX512F-NEXT: retq
|
|
;
|
|
; AVX512BW-LABEL: mask_shuffle_v16i32_02_03_04_05_06_07_08_09_10_11_12_13_14_15_00_01:
|
|
; AVX512BW: # %bb.0:
|
|
; AVX512BW-NEXT: kmovd %edi, %k1
|
|
; AVX512BW-NEXT: valignd {{.*#+}} zmm1 {%k1} = zmm0[2,3,4,5,6,7,8,9,10,11,12,13,14,15,0,1]
|
|
; AVX512BW-NEXT: vmovdqa64 %zmm1, %zmm0
|
|
; AVX512BW-NEXT: retq
|
|
%shuffle = shufflevector <16 x i32> %a, <16 x i32> poison, <16 x i32><i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 0, i32 1>
|
|
%mask.cast = bitcast i16 %mask to <16 x i1>
|
|
%res = select <16 x i1> %mask.cast, <16 x i32> %shuffle, <16 x i32> %passthru
|
|
ret <16 x i32> %res
|
|
}
|
|
|
|
define <16 x i32> @mask_shuffle_v16i32_02_03_04_05_06_07_08_09_10_11_12_13_14_15_16_17(<16 x i32> %a, <16 x i32> %b, <16 x i32> %passthru, i16 %mask) {
|
|
; AVX512F-LABEL: mask_shuffle_v16i32_02_03_04_05_06_07_08_09_10_11_12_13_14_15_16_17:
|
|
; AVX512F: # %bb.0:
|
|
; AVX512F-NEXT: kmovw %edi, %k1
|
|
; AVX512F-NEXT: valignd {{.*#+}} zmm2 {%k1} = zmm0[2,3,4,5,6,7,8,9,10,11,12,13,14,15],zmm1[0,1]
|
|
; AVX512F-NEXT: vmovdqa64 %zmm2, %zmm0
|
|
; AVX512F-NEXT: retq
|
|
;
|
|
; AVX512BW-LABEL: mask_shuffle_v16i32_02_03_04_05_06_07_08_09_10_11_12_13_14_15_16_17:
|
|
; AVX512BW: # %bb.0:
|
|
; AVX512BW-NEXT: kmovd %edi, %k1
|
|
; AVX512BW-NEXT: valignd {{.*#+}} zmm2 {%k1} = zmm0[2,3,4,5,6,7,8,9,10,11,12,13,14,15],zmm1[0,1]
|
|
; AVX512BW-NEXT: vmovdqa64 %zmm2, %zmm0
|
|
; AVX512BW-NEXT: retq
|
|
%shuffle = shufflevector <16 x i32> %a, <16 x i32> %b, <16 x i32><i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 16, i32 17>
|
|
%mask.cast = bitcast i16 %mask to <16 x i1>
|
|
%res = select <16 x i1> %mask.cast, <16 x i32> %shuffle, <16 x i32> %passthru
|
|
ret <16 x i32> %res
|
|
}
|
|
|
|
define <16 x i32> @maskz_shuffle_v16i32_02_03_04_05_06_07_08_09_10_11_12_13_14_15_00_01(<16 x i32> %a, i16 %mask) {
|
|
; AVX512F-LABEL: maskz_shuffle_v16i32_02_03_04_05_06_07_08_09_10_11_12_13_14_15_00_01:
|
|
; AVX512F: # %bb.0:
|
|
; AVX512F-NEXT: kmovw %edi, %k1
|
|
; AVX512F-NEXT: valignd {{.*#+}} zmm0 {%k1} {z} = zmm0[2,3,4,5,6,7,8,9,10,11,12,13,14,15,0,1]
|
|
; AVX512F-NEXT: retq
|
|
;
|
|
; AVX512BW-LABEL: maskz_shuffle_v16i32_02_03_04_05_06_07_08_09_10_11_12_13_14_15_00_01:
|
|
; AVX512BW: # %bb.0:
|
|
; AVX512BW-NEXT: kmovd %edi, %k1
|
|
; AVX512BW-NEXT: valignd {{.*#+}} zmm0 {%k1} {z} = zmm0[2,3,4,5,6,7,8,9,10,11,12,13,14,15,0,1]
|
|
; AVX512BW-NEXT: retq
|
|
%shuffle = shufflevector <16 x i32> %a, <16 x i32> poison, <16 x i32><i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 0, i32 1>
|
|
%mask.cast = bitcast i16 %mask to <16 x i1>
|
|
%res = select <16 x i1> %mask.cast, <16 x i32> %shuffle, <16 x i32> zeroinitializer
|
|
ret <16 x i32> %res
|
|
}
|
|
|
|
define <16 x i32> @maskz_shuffle_v16i32_02_03_04_05_06_07_08_09_10_11_12_13_14_15_16_17(<16 x i32> %a, <16 x i32> %b, i16 %mask) {
|
|
; AVX512F-LABEL: maskz_shuffle_v16i32_02_03_04_05_06_07_08_09_10_11_12_13_14_15_16_17:
|
|
; AVX512F: # %bb.0:
|
|
; AVX512F-NEXT: kmovw %edi, %k1
|
|
; AVX512F-NEXT: valignd {{.*#+}} zmm0 {%k1} {z} = zmm0[2,3,4,5,6,7,8,9,10,11,12,13,14,15],zmm1[0,1]
|
|
; AVX512F-NEXT: retq
|
|
;
|
|
; AVX512BW-LABEL: maskz_shuffle_v16i32_02_03_04_05_06_07_08_09_10_11_12_13_14_15_16_17:
|
|
; AVX512BW: # %bb.0:
|
|
; AVX512BW-NEXT: kmovd %edi, %k1
|
|
; AVX512BW-NEXT: valignd {{.*#+}} zmm0 {%k1} {z} = zmm0[2,3,4,5,6,7,8,9,10,11,12,13,14,15],zmm1[0,1]
|
|
; AVX512BW-NEXT: retq
|
|
%shuffle = shufflevector <16 x i32> %a, <16 x i32> %b, <16 x i32><i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 16, i32 17>
|
|
%mask.cast = bitcast i16 %mask to <16 x i1>
|
|
%res = select <16 x i1> %mask.cast, <16 x i32> %shuffle, <16 x i32> zeroinitializer
|
|
ret <16 x i32> %res
|
|
}
|
|
|
|
define <16 x float> @test_vshuff32x4_512(<16 x float> %x, <16 x float> %x1) nounwind {
|
|
; ALL-LABEL: test_vshuff32x4_512:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vshuff64x2 {{.*#+}} zmm0 = zmm0[0,1,2,3],zmm1[2,3,0,1]
|
|
; ALL-NEXT: retq
|
|
%res = shufflevector <16 x float> %x, <16 x float> %x1, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 20, i32 21, i32 22, i32 23, i32 16, i32 17, i32 18, i32 19>
|
|
ret <16 x float> %res
|
|
}
|
|
|
|
define <16 x i32> @test_vshufi32x4_512(<16 x i32> %x, <16 x i32> %x1) nounwind {
|
|
; ALL-LABEL: test_vshufi32x4_512:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: vshufi64x2 {{.*#+}} zmm0 = zmm0[0,1,2,3],zmm1[2,3,0,1]
|
|
; ALL-NEXT: retq
|
|
%res = shufflevector <16 x i32> %x, <16 x i32> %x1, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 20, i32 21, i32 22, i32 23, i32 16, i32 17, i32 18, i32 19>
|
|
ret <16 x i32> %res
|
|
}
|
|
|
|
define <16 x float> @test_vshuff32x4_512_mask(<16 x float> %x, <16 x float> %x1, <16 x float> %y, <16 x i1> %mask) nounwind {
|
|
; AVX512F-LABEL: test_vshuff32x4_512_mask:
|
|
; AVX512F: # %bb.0:
|
|
; AVX512F-NEXT: vpmovsxbd %xmm3, %zmm3
|
|
; AVX512F-NEXT: vpslld $31, %zmm3, %zmm3
|
|
; AVX512F-NEXT: vpmovd2m %zmm3, %k1
|
|
; AVX512F-NEXT: vshuff32x4 {{.*#+}} zmm2 {%k1} = zmm0[0,1,2,3,4,5,6,7],zmm1[4,5,6,7,0,1,2,3]
|
|
; AVX512F-NEXT: vmovaps %zmm2, %zmm0
|
|
; AVX512F-NEXT: retq
|
|
;
|
|
; AVX512BW-LABEL: test_vshuff32x4_512_mask:
|
|
; AVX512BW: # %bb.0:
|
|
; AVX512BW-NEXT: vpsllw $7, %xmm3, %xmm3
|
|
; AVX512BW-NEXT: vpmovb2m %zmm3, %k1
|
|
; AVX512BW-NEXT: vshuff32x4 {{.*#+}} zmm2 {%k1} = zmm0[0,1,2,3,4,5,6,7],zmm1[4,5,6,7,0,1,2,3]
|
|
; AVX512BW-NEXT: vmovaps %zmm2, %zmm0
|
|
; AVX512BW-NEXT: retq
|
|
%x2 = shufflevector <16 x float> %x, <16 x float> %x1, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 20, i32 21, i32 22, i32 23, i32 16, i32 17, i32 18, i32 19>
|
|
%res = select <16 x i1> %mask, <16 x float> %x2, <16 x float> %y
|
|
ret <16 x float> %res
|
|
}
|
|
|
|
define <16 x i32> @test_vshufi32x4_512_mask(<16 x i32> %x, <16 x i32> %x1, <16 x i32> %y, <16 x i1> %mask) nounwind {
|
|
; AVX512F-LABEL: test_vshufi32x4_512_mask:
|
|
; AVX512F: # %bb.0:
|
|
; AVX512F-NEXT: vpmovsxbd %xmm3, %zmm3
|
|
; AVX512F-NEXT: vpslld $31, %zmm3, %zmm3
|
|
; AVX512F-NEXT: vpmovd2m %zmm3, %k1
|
|
; AVX512F-NEXT: vshufi32x4 {{.*#+}} zmm2 {%k1} = zmm0[0,1,2,3,4,5,6,7],zmm1[4,5,6,7,0,1,2,3]
|
|
; AVX512F-NEXT: vmovdqa64 %zmm2, %zmm0
|
|
; AVX512F-NEXT: retq
|
|
;
|
|
; AVX512BW-LABEL: test_vshufi32x4_512_mask:
|
|
; AVX512BW: # %bb.0:
|
|
; AVX512BW-NEXT: vpsllw $7, %xmm3, %xmm3
|
|
; AVX512BW-NEXT: vpmovb2m %zmm3, %k1
|
|
; AVX512BW-NEXT: vshufi32x4 {{.*#+}} zmm2 {%k1} = zmm0[0,1,2,3,4,5,6,7],zmm1[4,5,6,7,0,1,2,3]
|
|
; AVX512BW-NEXT: vmovdqa64 %zmm2, %zmm0
|
|
; AVX512BW-NEXT: retq
|
|
%x2 = shufflevector <16 x i32> %x, <16 x i32> %x1, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 20, i32 21, i32 22, i32 23, i32 16, i32 17, i32 18, i32 19>
|
|
%res = select <16 x i1> %mask, <16 x i32> %x2, <16 x i32> %y
|
|
ret <16 x i32> %res
|
|
}
|
|
|
|
define <16 x float> @mask_shuffle_v16f32_00_01_02_03_04_05_06_07_16_17_18_19_20_21_22_23(<16 x float> %a, <16 x float> %b, <16 x float> %passthru, i16 %mask) {
|
|
; AVX512F-LABEL: mask_shuffle_v16f32_00_01_02_03_04_05_06_07_16_17_18_19_20_21_22_23:
|
|
; AVX512F: # %bb.0:
|
|
; AVX512F-NEXT: kmovw %edi, %k1
|
|
; AVX512F-NEXT: vinsertf32x8 $1, %ymm1, %zmm0, %zmm2 {%k1}
|
|
; AVX512F-NEXT: vmovaps %zmm2, %zmm0
|
|
; AVX512F-NEXT: retq
|
|
;
|
|
; AVX512BW-LABEL: mask_shuffle_v16f32_00_01_02_03_04_05_06_07_16_17_18_19_20_21_22_23:
|
|
; AVX512BW: # %bb.0:
|
|
; AVX512BW-NEXT: kmovd %edi, %k1
|
|
; AVX512BW-NEXT: vinsertf32x8 $1, %ymm1, %zmm0, %zmm2 {%k1}
|
|
; AVX512BW-NEXT: vmovaps %zmm2, %zmm0
|
|
; AVX512BW-NEXT: retq
|
|
%shuffle = shufflevector <16 x float> %a, <16 x float> %b, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 16, i32 17, i32 18, i32 19, i32 20, i32 21, i32 22, i32 23>
|
|
%mask.cast = bitcast i16 %mask to <16 x i1>
|
|
%res = select <16 x i1> %mask.cast, <16 x float> %shuffle, <16 x float> %passthru
|
|
ret <16 x float> %res
|
|
}
|
|
|
|
define <16 x float> @mask_shuffle_v16f32_00_01_02_03_16_17_18_19_08_09_10_11_12_13_14_15(<16 x float> %a, <16 x float> %b, <16 x float> %passthru, i16 %mask) {
|
|
; AVX512F-LABEL: mask_shuffle_v16f32_00_01_02_03_16_17_18_19_08_09_10_11_12_13_14_15:
|
|
; AVX512F: # %bb.0:
|
|
; AVX512F-NEXT: kmovw %edi, %k1
|
|
; AVX512F-NEXT: vinsertf32x4 $1, %xmm1, %zmm0, %zmm2 {%k1}
|
|
; AVX512F-NEXT: vmovaps %zmm2, %zmm0
|
|
; AVX512F-NEXT: retq
|
|
;
|
|
; AVX512BW-LABEL: mask_shuffle_v16f32_00_01_02_03_16_17_18_19_08_09_10_11_12_13_14_15:
|
|
; AVX512BW: # %bb.0:
|
|
; AVX512BW-NEXT: kmovd %edi, %k1
|
|
; AVX512BW-NEXT: vinsertf32x4 $1, %xmm1, %zmm0, %zmm2 {%k1}
|
|
; AVX512BW-NEXT: vmovaps %zmm2, %zmm0
|
|
; AVX512BW-NEXT: retq
|
|
%shuffle = shufflevector <16 x float> %a, <16 x float> %b, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 16, i32 17, i32 18, i32 19, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
|
|
%mask.cast = bitcast i16 %mask to <16 x i1>
|
|
%res = select <16 x i1> %mask.cast, <16 x float> %shuffle, <16 x float> %passthru
|
|
ret <16 x float> %res
|
|
}
|
|
|
|
define <16 x i32> @mask_shuffle_v16i32_00_01_02_03_04_05_06_07_16_17_18_19_20_21_22_23(<16 x i32> %a, <16 x i32> %b, <16 x i32> %passthru, i16 %mask) {
|
|
; AVX512F-LABEL: mask_shuffle_v16i32_00_01_02_03_04_05_06_07_16_17_18_19_20_21_22_23:
|
|
; AVX512F: # %bb.0:
|
|
; AVX512F-NEXT: kmovw %edi, %k1
|
|
; AVX512F-NEXT: vinserti32x8 $1, %ymm1, %zmm0, %zmm2 {%k1}
|
|
; AVX512F-NEXT: vmovdqa64 %zmm2, %zmm0
|
|
; AVX512F-NEXT: retq
|
|
;
|
|
; AVX512BW-LABEL: mask_shuffle_v16i32_00_01_02_03_04_05_06_07_16_17_18_19_20_21_22_23:
|
|
; AVX512BW: # %bb.0:
|
|
; AVX512BW-NEXT: kmovd %edi, %k1
|
|
; AVX512BW-NEXT: vinserti32x8 $1, %ymm1, %zmm0, %zmm2 {%k1}
|
|
; AVX512BW-NEXT: vmovdqa64 %zmm2, %zmm0
|
|
; AVX512BW-NEXT: retq
|
|
%shuffle = shufflevector <16 x i32> %a, <16 x i32> %b, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 16, i32 17, i32 18, i32 19, i32 20, i32 21, i32 22, i32 23>
|
|
%mask.cast = bitcast i16 %mask to <16 x i1>
|
|
%res = select <16 x i1> %mask.cast, <16 x i32> %shuffle, <16 x i32> %passthru
|
|
ret <16 x i32> %res
|
|
}
|
|
|
|
define <16 x i32> @mask_shuffle_v16i32_00_01_02_03_16_17_18_19_08_09_10_11_12_13_14_15(<16 x i32> %a, <16 x i32> %b, <16 x i32> %passthru, i16 %mask) {
|
|
; AVX512F-LABEL: mask_shuffle_v16i32_00_01_02_03_16_17_18_19_08_09_10_11_12_13_14_15:
|
|
; AVX512F: # %bb.0:
|
|
; AVX512F-NEXT: kmovw %edi, %k1
|
|
; AVX512F-NEXT: vinserti32x4 $1, %xmm1, %zmm0, %zmm2 {%k1}
|
|
; AVX512F-NEXT: vmovdqa64 %zmm2, %zmm0
|
|
; AVX512F-NEXT: retq
|
|
;
|
|
; AVX512BW-LABEL: mask_shuffle_v16i32_00_01_02_03_16_17_18_19_08_09_10_11_12_13_14_15:
|
|
; AVX512BW: # %bb.0:
|
|
; AVX512BW-NEXT: kmovd %edi, %k1
|
|
; AVX512BW-NEXT: vinserti32x4 $1, %xmm1, %zmm0, %zmm2 {%k1}
|
|
; AVX512BW-NEXT: vmovdqa64 %zmm2, %zmm0
|
|
; AVX512BW-NEXT: retq
|
|
%shuffle = shufflevector <16 x i32> %a, <16 x i32> %b, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 16, i32 17, i32 18, i32 19, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
|
|
%mask.cast = bitcast i16 %mask to <16 x i1>
|
|
%res = select <16 x i1> %mask.cast, <16 x i32> %shuffle, <16 x i32> %passthru
|
|
ret <16 x i32> %res
|
|
}
|
|
|
|
define <16 x i32> @mask_shuffle_v4i32_v16i32_00_01_02_03_00_01_02_03_00_01_02_03_00_01_02_03(<4 x i32> %a) {
|
|
; ALL-LABEL: mask_shuffle_v4i32_v16i32_00_01_02_03_00_01_02_03_00_01_02_03_00_01_02_03:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: # kill: def $xmm0 killed $xmm0 def $zmm0
|
|
; ALL-NEXT: vshufi64x2 {{.*#+}} zmm0 = zmm0[0,1,0,1,0,1,0,1]
|
|
; ALL-NEXT: retq
|
|
%res = shufflevector <4 x i32> %a, <4 x i32> poison, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 0, i32 1, i32 2, i32 3, i32 0, i32 1, i32 2, i32 3, i32 0, i32 1, i32 2, i32 3>
|
|
ret <16 x i32> %res
|
|
}
|
|
|
|
define <16 x float> @mask_shuffle_v4f32_v16f32_00_01_02_03_00_01_02_03_00_01_02_03_00_01_02_03(<4 x float> %a) {
|
|
; ALL-LABEL: mask_shuffle_v4f32_v16f32_00_01_02_03_00_01_02_03_00_01_02_03_00_01_02_03:
|
|
; ALL: # %bb.0:
|
|
; ALL-NEXT: # kill: def $xmm0 killed $xmm0 def $zmm0
|
|
; ALL-NEXT: vshuff64x2 {{.*#+}} zmm0 = zmm0[0,1,0,1,0,1,0,1]
|
|
; ALL-NEXT: retq
|
|
%res = shufflevector <4 x float> %a, <4 x float> poison, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 0, i32 1, i32 2, i32 3, i32 0, i32 1, i32 2, i32 3, i32 0, i32 1, i32 2, i32 3>
|
|
ret <16 x float> %res
|
|
}
|
|
|
|
%struct.foo = type { [4 x double], [3 x [4 x double]], [4 x double] }
|
|
|
|
; This test previously hung in shuffle combining. https://github.com/ispc/ispc/issues/1864
|
|
define void @ispc_1864(ptr %arg) {
|
|
; ALL-LABEL: ispc_1864:
|
|
; ALL: # %bb.0: # %bb
|
|
; ALL-NEXT: pushq %rbp
|
|
; ALL-NEXT: .cfi_def_cfa_offset 16
|
|
; ALL-NEXT: .cfi_offset %rbp, -16
|
|
; ALL-NEXT: movq %rsp, %rbp
|
|
; ALL-NEXT: .cfi_def_cfa_register %rbp
|
|
; ALL-NEXT: andq $-64, %rsp
|
|
; ALL-NEXT: subq $4864, %rsp # imm = 0x1300
|
|
; ALL-NEXT: vbroadcastss {{.*#+}} ymm0 = [-5.0E+0,-5.0E+0,-5.0E+0,-5.0E+0,-5.0E+0,-5.0E+0,-5.0E+0,-5.0E+0]
|
|
; ALL-NEXT: vmulps 32(%rdi), %ymm0, %ymm0
|
|
; ALL-NEXT: vcvtps2pd %ymm0, %zmm0
|
|
; ALL-NEXT: vshuff64x2 {{.*#+}} zmm0 = zmm0[2,3,4,5,4,5,6,7]
|
|
; ALL-NEXT: vmovapd %ymm0, {{[0-9]+}}(%rsp)
|
|
; ALL-NEXT: movq %rbp, %rsp
|
|
; ALL-NEXT: popq %rbp
|
|
; ALL-NEXT: .cfi_def_cfa %rsp, 8
|
|
; ALL-NEXT: vzeroupper
|
|
; ALL-NEXT: retq
|
|
bb:
|
|
%tmp = alloca [30 x %struct.foo], align 64
|
|
%tmp1 = load <16 x float>, ptr %arg, align 4
|
|
%tmp2 = fmul <16 x float> %tmp1, <float -5.000000e+00, float -5.000000e+00, float -5.000000e+00, float -5.000000e+00, float -5.000000e+00, float -5.000000e+00, float -5.000000e+00, float -5.000000e+00, float -5.000000e+00, float -5.000000e+00, float -5.000000e+00, float -5.000000e+00, float -5.000000e+00, float -5.000000e+00, float -5.000000e+00, float -5.000000e+00>
|
|
%tmp3 = fpext <16 x float> %tmp2 to <16 x double>
|
|
%tmp4 = getelementptr inbounds [30 x %struct.foo], ptr %tmp, i64 0, i64 3, i32 2, i64 0
|
|
%tmp5 = extractelement <16 x double> %tmp3, i32 10
|
|
store double %tmp5, ptr %tmp4, align 32
|
|
%tmp6 = getelementptr inbounds [30 x %struct.foo], ptr %tmp, i64 0, i64 3, i32 2, i64 1
|
|
%tmp7 = extractelement <16 x double> %tmp3, i32 11
|
|
store double %tmp7, ptr %tmp6, align 8
|
|
%tmp8 = getelementptr inbounds [30 x %struct.foo], ptr %tmp, i64 0, i64 3, i32 2, i64 2
|
|
%tmp9 = extractelement <16 x double> %tmp3, i32 12
|
|
store double %tmp9, ptr %tmp8, align 16
|
|
%tmp10 = getelementptr inbounds [30 x %struct.foo], ptr %tmp, i64 0, i64 3, i32 2, i64 3
|
|
%tmp11 = extractelement <16 x double> %tmp3, i32 13
|
|
store double %tmp11, ptr %tmp10, align 8
|
|
ret void
|
|
}
|
|
|