/external/llvm-project/llvm/test/CodeGen/PowerPC/ |
D | vec_conv_fp_to_i_8byte_elts.ll | 145 ; CHECK-P8-NEXT: lxvd2x vs7, 0, r4 153 ; CHECK-P8-NEXT: xvcvdpuxds vs7, vs7 161 ; CHECK-P8-NEXT: stxvd2x vs7, 0, r3 174 ; CHECK-P9-NEXT: lxv vs7, 0(r4) 175 ; CHECK-P9-NEXT: xvcvdpuxds vs7, vs7 189 ; CHECK-P9-NEXT: stxv vs7, 0(r3) 202 ; CHECK-BE-NEXT: lxv vs7, 0(r4) 203 ; CHECK-BE-NEXT: xvcvdpuxds vs7, vs7 217 ; CHECK-BE-NEXT: stxv vs7, 0(r3) 359 ; CHECK-P8-NEXT: lxvd2x vs7, 0, r4 [all …]
|
D | vec_conv_fp32_to_i64_elts.ll | 189 ; CHECK-P8-NEXT: xxmrghw vs7, v2, v2 196 ; CHECK-P8-NEXT: xvcvspdp vs7, vs7 205 ; CHECK-P8-NEXT: xvcvdpuxds v7, vs7 237 ; CHECK-P9-NEXT: xxmrglw vs7, vs0, vs0 245 ; CHECK-P9-NEXT: xvcvspdp vs7, vs7 253 ; CHECK-P9-NEXT: xvcvdpuxds vs7, vs7 256 ; CHECK-P9-NEXT: stxv vs7, 96(r3) 277 ; CHECK-BE-NEXT: xxmrghw vs7, vs0, vs0 285 ; CHECK-BE-NEXT: xvcvspdp vs7, vs7 293 ; CHECK-BE-NEXT: xvcvdpuxds vs7, vs7 [all …]
|
D | vec_conv_fp64_to_i32_elts.ll | 182 ; CHECK-P8-NEXT: lxvd2x vs7, 0, r4 201 ; CHECK-P8-NEXT: xxswapd vs1, vs7 202 ; CHECK-P8-NEXT: xxmrgld vs7, vs4, vs3 210 ; CHECK-P8-NEXT: xvcvdpuxws v4, vs7 229 ; CHECK-P9-NEXT: lxv vs7, 16(r4) 232 ; CHECK-P9-NEXT: xxmrgld vs8, vs7, vs6 233 ; CHECK-P9-NEXT: xxmrghd vs6, vs7, vs6 234 ; CHECK-P9-NEXT: xxmrgld vs7, vs5, vs4 242 ; CHECK-P9-NEXT: xvcvdpuxws v4, vs7 265 ; CHECK-BE-NEXT: lxv vs7, 0(r4) [all …]
|
D | vec_conv_i_to_fp_8byte_elts.ll | 145 ; CHECK-P8-NEXT: lxvd2x vs7, 0, r4 153 ; CHECK-P8-NEXT: xvcvuxddp vs7, vs7 161 ; CHECK-P8-NEXT: stxvd2x vs7, 0, r3 181 ; CHECK-P9-NEXT: xvcvuxddp vs7, v2 182 ; CHECK-P9-NEXT: stxv vs7, 112(r3) 209 ; CHECK-BE-NEXT: xvcvuxddp vs7, v2 210 ; CHECK-BE-NEXT: stxv vs7, 112(r3) 359 ; CHECK-P8-NEXT: lxvd2x vs7, 0, r4 367 ; CHECK-P8-NEXT: xvcvsxddp vs7, vs7 375 ; CHECK-P8-NEXT: stxvd2x vs7, 0, r3 [all …]
|
D | vec_conv_i32_to_fp64_elts.ll | 176 ; CHECK-P8-NEXT: xvcvuxwdp vs7, v3 187 ; CHECK-P8-NEXT: xxswapd vs3, vs7 219 ; CHECK-P9-NEXT: xvcvuxwdp vs7, v2 224 ; CHECK-P9-NEXT: stxv vs7, 96(r3) 251 ; CHECK-BE-NEXT: xvcvuxwdp vs7, v2 256 ; CHECK-BE-NEXT: stxv vs7, 96(r3) 430 ; CHECK-P8-NEXT: xvcvsxwdp vs7, v3 441 ; CHECK-P8-NEXT: xxswapd vs3, vs7 473 ; CHECK-P9-NEXT: xvcvsxwdp vs7, v2 478 ; CHECK-P9-NEXT: stxv vs7, 96(r3) [all …]
|
/external/XNNPACK/src/f32-velu/gen/ |
D | velu-avx512f-rr1-lut16-p3-perm-x128.c | 102 __m512 vs7 = _mm512_castsi512_ps(_mm512_add_epi32(vl7, ven7)); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x128() local 138 vt7 = _mm512_mul_ps(vt7, vs7); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x128() 147 vs7 = _mm512_fmsub_ps(vs7, valpha, valpha); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x128() 173 __m512 vy7 = _mm512_fmadd_ps(vp7, valpha, vs7); in xnn_f32_velu_ukernel__avx512f_rr1_lut16_p3_perm_x128()
|
D | velu-avx2-rr1-p6-x64.c | 85 __m256 vs7 = _mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(vn7), 23)); in xnn_f32_velu_ukernel__avx2_rr1_p6_x64() local 148 vt7 = _mm256_mul_ps(vt7, vs7); in xnn_f32_velu_ukernel__avx2_rr1_p6_x64() 164 vs7 = _mm256_fmsub_ps(vs7, valpha, valpha); in xnn_f32_velu_ukernel__avx2_rr1_p6_x64() 181 const __m256 ve7 = _mm256_fmadd_ps(vp7, valpha, vs7); in xnn_f32_velu_ukernel__avx2_rr1_p6_x64()
|
D | velu-avx512f-rr1-p6-x128.c | 85 __m512 vs7 = _mm512_castsi512_ps(_mm512_slli_epi32(_mm512_castps_si512(vn7), 23)); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x128() local 148 vt7 = _mm512_mul_ps(vt7, vs7); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x128() 157 vs7 = _mm512_fmsub_ps(vs7, valpha, valpha); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x128() 183 __m512 vy7 = _mm512_fmadd_ps(vp7, valpha, vs7); in xnn_f32_velu_ukernel__avx512f_rr1_p6_x128()
|
D | velu-avx2-rr1-lut4-p4-perm-x64.c | 111 __m256 vs7 = _mm256_castsi256_ps(_mm256_add_epi32(vl7, ven7)); in xnn_f32_velu_ukernel__avx2_rr1_lut4_p4_perm_x64() local 147 vt7 = _mm256_mul_ps(vt7, vs7); in xnn_f32_velu_ukernel__avx2_rr1_lut4_p4_perm_x64() 163 vs7 = _mm256_fmsub_ps(vs7, valpha, valpha); in xnn_f32_velu_ukernel__avx2_rr1_lut4_p4_perm_x64() 180 const __m256 ve7 = _mm256_fmadd_ps(vp7, valpha, vs7); in xnn_f32_velu_ukernel__avx2_rr1_lut4_p4_perm_x64()
|
D | velu-avx2-rr1-lut8-p4-perm-x64.c | 110 __m256 vs7 = _mm256_castsi256_ps(_mm256_add_epi32(vl7, ven7)); in xnn_f32_velu_ukernel__avx2_rr1_lut8_p4_perm_x64() local 146 vt7 = _mm256_mul_ps(vt7, vs7); in xnn_f32_velu_ukernel__avx2_rr1_lut8_p4_perm_x64() 162 vs7 = _mm256_fmsub_ps(vs7, valpha, valpha); in xnn_f32_velu_ukernel__avx2_rr1_lut8_p4_perm_x64() 179 const __m256 ve7 = _mm256_fmadd_ps(vp7, valpha, vs7); in xnn_f32_velu_ukernel__avx2_rr1_lut8_p4_perm_x64()
|
D | velu-avx2-rr1-lut16-p3-gather-x64.c | 119 __m256 vs7 = _mm256_castsi256_ps(_mm256_add_epi32(vl7, ven7)); in xnn_f32_velu_ukernel__avx2_rr1_lut16_p3_gather_x64() local 146 vt7 = _mm256_mul_ps(vt7, vs7); in xnn_f32_velu_ukernel__avx2_rr1_lut16_p3_gather_x64() 162 vs7 = _mm256_fmsub_ps(vs7, valpha, valpha); in xnn_f32_velu_ukernel__avx2_rr1_lut16_p3_gather_x64() 179 const __m256 ve7 = _mm256_fmadd_ps(vp7, valpha, vs7); in xnn_f32_velu_ukernel__avx2_rr1_lut16_p3_gather_x64()
|
D | velu-avx2-rr1-p6-x72.c | 88 __m256 vs7 = _mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(vn7), 23)); in xnn_f32_velu_ukernel__avx2_rr1_p6_x72() local 158 vt7 = _mm256_mul_ps(vt7, vs7); in xnn_f32_velu_ukernel__avx2_rr1_p6_x72() 176 vs7 = _mm256_fmsub_ps(vs7, valpha, valpha); in xnn_f32_velu_ukernel__avx2_rr1_p6_x72() 195 const __m256 ve7 = _mm256_fmadd_ps(vp7, valpha, vs7); in xnn_f32_velu_ukernel__avx2_rr1_p6_x72()
|
D | velu-avx2-rr1-lut4-p4-perm-x72.c | 117 __m256 vs7 = _mm256_castsi256_ps(_mm256_add_epi32(vl7, ven7)); in xnn_f32_velu_ukernel__avx2_rr1_lut4_p4_perm_x72() local 157 vt7 = _mm256_mul_ps(vt7, vs7); in xnn_f32_velu_ukernel__avx2_rr1_lut4_p4_perm_x72() 175 vs7 = _mm256_fmsub_ps(vs7, valpha, valpha); in xnn_f32_velu_ukernel__avx2_rr1_lut4_p4_perm_x72() 194 const __m256 ve7 = _mm256_fmadd_ps(vp7, valpha, vs7); in xnn_f32_velu_ukernel__avx2_rr1_lut4_p4_perm_x72()
|
D | velu-avx2-rr1-lut16-p3-gather-x72.c | 126 __m256 vs7 = _mm256_castsi256_ps(_mm256_add_epi32(vl7, ven7)); in xnn_f32_velu_ukernel__avx2_rr1_lut16_p3_gather_x72() local 156 vt7 = _mm256_mul_ps(vt7, vs7); in xnn_f32_velu_ukernel__avx2_rr1_lut16_p3_gather_x72() 174 vs7 = _mm256_fmsub_ps(vs7, valpha, valpha); in xnn_f32_velu_ukernel__avx2_rr1_lut16_p3_gather_x72() 193 const __m256 ve7 = _mm256_fmadd_ps(vp7, valpha, vs7); in xnn_f32_velu_ukernel__avx2_rr1_lut16_p3_gather_x72()
|
D | velu-avx2-rr1-lut8-p4-perm-x72.c | 116 __m256 vs7 = _mm256_castsi256_ps(_mm256_add_epi32(vl7, ven7)); in xnn_f32_velu_ukernel__avx2_rr1_lut8_p4_perm_x72() local 156 vt7 = _mm256_mul_ps(vt7, vs7); in xnn_f32_velu_ukernel__avx2_rr1_lut8_p4_perm_x72() 174 vs7 = _mm256_fmsub_ps(vs7, valpha, valpha); in xnn_f32_velu_ukernel__avx2_rr1_lut8_p4_perm_x72() 193 const __m256 ve7 = _mm256_fmadd_ps(vp7, valpha, vs7); in xnn_f32_velu_ukernel__avx2_rr1_lut8_p4_perm_x72()
|
D | velu-avx2-rr1-lut8-p4-perm-x80.c | 122 __m256 vs7 = _mm256_castsi256_ps(_mm256_add_epi32(vl7, ven7)); in xnn_f32_velu_ukernel__avx2_rr1_lut8_p4_perm_x80() local 166 vt7 = _mm256_mul_ps(vt7, vs7); in xnn_f32_velu_ukernel__avx2_rr1_lut8_p4_perm_x80() 186 vs7 = _mm256_fmsub_ps(vs7, valpha, valpha); in xnn_f32_velu_ukernel__avx2_rr1_lut8_p4_perm_x80() 207 const __m256 ve7 = _mm256_fmadd_ps(vp7, valpha, vs7); in xnn_f32_velu_ukernel__avx2_rr1_lut8_p4_perm_x80()
|
D | velu-avx2-rr1-p6-x80.c | 91 __m256 vs7 = _mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(vn7), 23)); in xnn_f32_velu_ukernel__avx2_rr1_p6_x80() local 168 vt7 = _mm256_mul_ps(vt7, vs7); in xnn_f32_velu_ukernel__avx2_rr1_p6_x80() 188 vs7 = _mm256_fmsub_ps(vs7, valpha, valpha); in xnn_f32_velu_ukernel__avx2_rr1_p6_x80() 209 const __m256 ve7 = _mm256_fmadd_ps(vp7, valpha, vs7); in xnn_f32_velu_ukernel__avx2_rr1_p6_x80()
|
D | velu-avx2-rr1-lut4-p4-perm-x80.c | 123 __m256 vs7 = _mm256_castsi256_ps(_mm256_add_epi32(vl7, ven7)); in xnn_f32_velu_ukernel__avx2_rr1_lut4_p4_perm_x80() local 167 vt7 = _mm256_mul_ps(vt7, vs7); in xnn_f32_velu_ukernel__avx2_rr1_lut4_p4_perm_x80() 187 vs7 = _mm256_fmsub_ps(vs7, valpha, valpha); in xnn_f32_velu_ukernel__avx2_rr1_lut4_p4_perm_x80() 208 const __m256 ve7 = _mm256_fmadd_ps(vp7, valpha, vs7); in xnn_f32_velu_ukernel__avx2_rr1_lut4_p4_perm_x80()
|
D | velu-avx2-rr1-lut16-p3-gather-x80.c | 133 __m256 vs7 = _mm256_castsi256_ps(_mm256_add_epi32(vl7, ven7)); in xnn_f32_velu_ukernel__avx2_rr1_lut16_p3_gather_x80() local 166 vt7 = _mm256_mul_ps(vt7, vs7); in xnn_f32_velu_ukernel__avx2_rr1_lut16_p3_gather_x80() 186 vs7 = _mm256_fmsub_ps(vs7, valpha, valpha); in xnn_f32_velu_ukernel__avx2_rr1_lut16_p3_gather_x80() 207 const __m256 ve7 = _mm256_fmadd_ps(vp7, valpha, vs7); in xnn_f32_velu_ukernel__avx2_rr1_lut16_p3_gather_x80()
|
/external/XNNPACK/src/f32-vscaleexpminusmax/gen/ |
D | avx2-p5-x64.c | 86 const __m256 vs7 = _mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(vn7), 23)); in xnn_f32_vscaleexpminusmax_ukernel__avx2_p5_x64() local 166 vt7 = _mm256_mul_ps(vt7, vs7); in xnn_f32_vscaleexpminusmax_ukernel__avx2_p5_x64() 175 __m256 vf7 = _mm256_fmadd_ps(vt7, vp7, vs7); in xnn_f32_vscaleexpminusmax_ukernel__avx2_p5_x64()
|
/external/XNNPACK/src/f32-raddexpminusmax/gen/ |
D | avx2-p5-x64-acc2.c | 85 const __m256 vs7 = _mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(vn7), 23)); in xnn_f32_raddexpminusmax_ukernel__avx2_p5_x64_acc2() local 165 vt7 = _mm256_mul_ps(vt7, vs7); in xnn_f32_raddexpminusmax_ukernel__avx2_p5_x64_acc2() 174 __m256 vf7 = _mm256_fmadd_ps(vt7, vp7, vs7); in xnn_f32_raddexpminusmax_ukernel__avx2_p5_x64_acc2()
|
D | avx2-p5-x64-acc4.c | 87 const __m256 vs7 = _mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(vn7), 23)); in xnn_f32_raddexpminusmax_ukernel__avx2_p5_x64_acc4() local 167 vt7 = _mm256_mul_ps(vt7, vs7); in xnn_f32_raddexpminusmax_ukernel__avx2_p5_x64_acc4() 176 __m256 vf7 = _mm256_fmadd_ps(vt7, vp7, vs7); in xnn_f32_raddexpminusmax_ukernel__avx2_p5_x64_acc4()
|
D | avx2-p5-x64.c | 84 const __m256 vs7 = _mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(vn7), 23)); in xnn_f32_raddexpminusmax_ukernel__avx2_p5_x64() local 164 vt7 = _mm256_mul_ps(vt7, vs7); in xnn_f32_raddexpminusmax_ukernel__avx2_p5_x64() 173 __m256 vf7 = _mm256_fmadd_ps(vt7, vp7, vs7); in xnn_f32_raddexpminusmax_ukernel__avx2_p5_x64()
|
/external/XNNPACK/src/f32-raddstoreexpminusmax/gen/ |
D | avx2-p5-x64.c | 85 const __m256 vs7 = _mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(vn7), 23)); in xnn_f32_raddstoreexpminusmax_ukernel__avx2_p5_x64() local 165 vt7 = _mm256_mul_ps(vt7, vs7); in xnn_f32_raddstoreexpminusmax_ukernel__avx2_p5_x64() 174 __m256 vf7 = _mm256_fmadd_ps(vt7, vp7, vs7); in xnn_f32_raddstoreexpminusmax_ukernel__avx2_p5_x64()
|
D | avx2-p5-x64-acc2.c | 86 const __m256 vs7 = _mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(vn7), 23)); in xnn_f32_raddstoreexpminusmax_ukernel__avx2_p5_x64_acc2() local 166 vt7 = _mm256_mul_ps(vt7, vs7); in xnn_f32_raddstoreexpminusmax_ukernel__avx2_p5_x64_acc2() 175 __m256 vf7 = _mm256_fmadd_ps(vt7, vp7, vs7); in xnn_f32_raddstoreexpminusmax_ukernel__avx2_p5_x64_acc2()
|