/external/XNNPACK/src/f32-raddextexp/gen/ |
D | avx512f-p5-scalef-x128-acc2.c | 158 const __m512 vdelta_e5 = _mm512_sub_ps(vn5, vmax_e1); in xnn_f32_raddextexp_ukernel__avx512f_p5_scalef_x128_acc2() local 170 vaccv1 = _mm512_add_ps(vaccv1, _mm512_scalef_ps(vp5, vdelta_e5)); in xnn_f32_raddextexp_ukernel__avx512f_p5_scalef_x128_acc2()
|
D | avx512f-p5-scalef-x128.c | 155 const __m512 vdelta_e5 = _mm512_sub_ps(vn5, vmax_e0); in xnn_f32_raddextexp_ukernel__avx512f_p5_scalef_x128() local 166 vaccv0 = _mm512_add_ps(vaccv0, _mm512_scalef_ps(vp5, vdelta_e5)); in xnn_f32_raddextexp_ukernel__avx512f_p5_scalef_x128()
|
D | avx512f-p5-scalef-x144.c | 165 const __m512 vdelta_e5 = _mm512_sub_ps(vn5, vmax_e0); in xnn_f32_raddextexp_ukernel__avx512f_p5_scalef_x144() local 177 vaccv0 = _mm512_add_ps(vaccv0, _mm512_scalef_ps(vp5, vdelta_e5)); in xnn_f32_raddextexp_ukernel__avx512f_p5_scalef_x144()
|
D | avx512f-p5-scalef-x144-acc3.c | 171 const __m512 vdelta_e5 = _mm512_sub_ps(vn5, vmax_e2); in xnn_f32_raddextexp_ukernel__avx512f_p5_scalef_x144_acc3() local 185 vaccv2 = _mm512_add_ps(vaccv2, _mm512_scalef_ps(vp5, vdelta_e5)); in xnn_f32_raddextexp_ukernel__avx512f_p5_scalef_x144_acc3()
|
D | avx512f-p5-scalef-x160.c | 175 const __m512 vdelta_e5 = _mm512_sub_ps(vn5, vmax_e0); in xnn_f32_raddextexp_ukernel__avx512f_p5_scalef_x160() local 188 vaccv0 = _mm512_add_ps(vaccv0, _mm512_scalef_ps(vp5, vdelta_e5)); in xnn_f32_raddextexp_ukernel__avx512f_p5_scalef_x160()
|
D | avx512f-p5-scalef-x160-acc2.c | 178 const __m512 vdelta_e5 = _mm512_sub_ps(vn5, vmax_e1); in xnn_f32_raddextexp_ukernel__avx512f_p5_scalef_x160_acc2() local 192 vaccv1 = _mm512_add_ps(vaccv1, _mm512_scalef_ps(vp5, vdelta_e5)); in xnn_f32_raddextexp_ukernel__avx512f_p5_scalef_x160_acc2()
|
D | avx512f-p5-scalef-x128-acc4.c | 164 const __m512 vdelta_e5 = _mm512_sub_ps(vn5, vmax_e1); in xnn_f32_raddextexp_ukernel__avx512f_p5_scalef_x128_acc4() local 178 vaccv1 = _mm512_add_ps(vaccv1, _mm512_scalef_ps(vp5, vdelta_e5)); in xnn_f32_raddextexp_ukernel__avx512f_p5_scalef_x128_acc4()
|
D | avx2-p5-x64.c | 163 const __m256 vdelta_e5 = _mm256_max_ps(_mm256_sub_ps(vn5, vmax_e0), vmin_exponent); in xnn_f32_raddextexp_ukernel__avx2_p5_x64() local 178 …_mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(_mm256_add_ps(vdelta_e5, vmagic_bias)), … in xnn_f32_raddextexp_ukernel__avx2_p5_x64()
|
D | avx2-p5-x72.c | 173 const __m256 vdelta_e5 = _mm256_max_ps(_mm256_sub_ps(vn5, vmax_e0), vmin_exponent); in xnn_f32_raddextexp_ukernel__avx2_p5_x72() local 189 …_mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(_mm256_add_ps(vdelta_e5, vmagic_bias)), … in xnn_f32_raddextexp_ukernel__avx2_p5_x72()
|
D | avx512f-p5-scalef-x192.c | 195 const __m512 vdelta_e5 = _mm512_sub_ps(vn5, vmax_e0); in xnn_f32_raddextexp_ukernel__avx512f_p5_scalef_x192() local 210 vaccv0 = _mm512_add_ps(vaccv0, _mm512_scalef_ps(vp5, vdelta_e5)); in xnn_f32_raddextexp_ukernel__avx512f_p5_scalef_x192()
|
D | avx512f-p5-scalef-x192-acc2.c | 198 const __m512 vdelta_e5 = _mm512_sub_ps(vn5, vmax_e1); in xnn_f32_raddextexp_ukernel__avx512f_p5_scalef_x192_acc2() local 214 vaccv1 = _mm512_add_ps(vaccv1, _mm512_scalef_ps(vp5, vdelta_e5)); in xnn_f32_raddextexp_ukernel__avx512f_p5_scalef_x192_acc2()
|
D | avx512f-p5-scalef-x160-acc5.c | 187 const __m512 vdelta_e5 = _mm512_sub_ps(vn5, vmax_e0); in xnn_f32_raddextexp_ukernel__avx512f_p5_scalef_x160_acc5() local 204 vaccv0 = _mm512_add_ps(vaccv0, _mm512_scalef_ps(vp5, vdelta_e5)); in xnn_f32_raddextexp_ukernel__avx512f_p5_scalef_x160_acc5()
|
D | avx2-p5-x64-acc2.c | 166 const __m256 vdelta_e5 = _mm256_max_ps(_mm256_sub_ps(vn5, vmax_e1), vmin_exponent); in xnn_f32_raddextexp_ukernel__avx2_p5_x64_acc2() local 182 …_mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(_mm256_add_ps(vdelta_e5, vmagic_bias)), … in xnn_f32_raddextexp_ukernel__avx2_p5_x64_acc2()
|
D | avx2-p5-x64-acc4.c | 172 const __m256 vdelta_e5 = _mm256_max_ps(_mm256_sub_ps(vn5, vmax_e1), vmin_exponent); in xnn_f32_raddextexp_ukernel__avx2_p5_x64_acc4() local 190 …_mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(_mm256_add_ps(vdelta_e5, vmagic_bias)), … in xnn_f32_raddextexp_ukernel__avx2_p5_x64_acc4()
|
D | avx2-p5-x80-acc2.c | 186 const __m256 vdelta_e5 = _mm256_max_ps(_mm256_sub_ps(vn5, vmax_e1), vmin_exponent); in xnn_f32_raddextexp_ukernel__avx2_p5_x80_acc2() local 204 …_mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(_mm256_add_ps(vdelta_e5, vmagic_bias)), … in xnn_f32_raddextexp_ukernel__avx2_p5_x80_acc2()
|
D | avx512f-p5-scalef-x192-acc3.c | 201 const __m512 vdelta_e5 = _mm512_sub_ps(vn5, vmax_e2); in xnn_f32_raddextexp_ukernel__avx512f_p5_scalef_x192_acc3() local 218 vaccv2 = _mm512_add_ps(vaccv2, _mm512_scalef_ps(vp5, vdelta_e5)); in xnn_f32_raddextexp_ukernel__avx512f_p5_scalef_x192_acc3()
|
D | avx2-p5-x72-acc3.c | 179 const __m256 vdelta_e5 = _mm256_max_ps(_mm256_sub_ps(vn5, vmax_e2), vmin_exponent); in xnn_f32_raddextexp_ukernel__avx2_p5_x72_acc3() local 197 …_mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(_mm256_add_ps(vdelta_e5, vmagic_bias)), … in xnn_f32_raddextexp_ukernel__avx2_p5_x72_acc3()
|
D | avx2-p5-x80.c | 183 const __m256 vdelta_e5 = _mm256_max_ps(_mm256_sub_ps(vn5, vmax_e0), vmin_exponent); in xnn_f32_raddextexp_ukernel__avx2_p5_x80() local 200 …_mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(_mm256_add_ps(vdelta_e5, vmagic_bias)), … in xnn_f32_raddextexp_ukernel__avx2_p5_x80()
|
D | avx512f-p5-scalef-x192-acc6.c | 210 const __m512 vdelta_e5 = _mm512_sub_ps(vn5, vmax_e5); in xnn_f32_raddextexp_ukernel__avx512f_p5_scalef_x192_acc6() local 230 vaccv5 = _mm512_add_ps(vaccv5, _mm512_scalef_ps(vp5, vdelta_e5)); in xnn_f32_raddextexp_ukernel__avx512f_p5_scalef_x192_acc6()
|
D | avx2-p5-x96.c | 203 const __m256 vdelta_e5 = _mm256_max_ps(_mm256_sub_ps(vn5, vmax_e0), vmin_exponent); in xnn_f32_raddextexp_ukernel__avx2_p5_x96() local 222 …_mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(_mm256_add_ps(vdelta_e5, vmagic_bias)), … in xnn_f32_raddextexp_ukernel__avx2_p5_x96()
|
D | avx2-p5-x80-acc5.c | 195 const __m256 vdelta_e5 = _mm256_max_ps(_mm256_sub_ps(vn5, vmax_e0), vmin_exponent); in xnn_f32_raddextexp_ukernel__avx2_p5_x80_acc5() local 216 …_mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(_mm256_add_ps(vdelta_e5, vmagic_bias)), … in xnn_f32_raddextexp_ukernel__avx2_p5_x80_acc5()
|
D | avx2-p5-x96-acc3.c | 209 const __m256 vdelta_e5 = _mm256_max_ps(_mm256_sub_ps(vn5, vmax_e2), vmin_exponent); in xnn_f32_raddextexp_ukernel__avx2_p5_x96_acc3() local 230 …_mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(_mm256_add_ps(vdelta_e5, vmagic_bias)), … in xnn_f32_raddextexp_ukernel__avx2_p5_x96_acc3()
|
D | avx2-p5-x96-acc2.c | 206 const __m256 vdelta_e5 = _mm256_max_ps(_mm256_sub_ps(vn5, vmax_e1), vmin_exponent); in xnn_f32_raddextexp_ukernel__avx2_p5_x96_acc2() local 226 …_mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(_mm256_add_ps(vdelta_e5, vmagic_bias)), … in xnn_f32_raddextexp_ukernel__avx2_p5_x96_acc2()
|
D | avx2-p5-x96-acc6.c | 218 const __m256 vdelta_e5 = _mm256_max_ps(_mm256_sub_ps(vn5, vmax_e5), vmin_exponent); in xnn_f32_raddextexp_ukernel__avx2_p5_x96_acc6() local 242 …_mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(_mm256_add_ps(vdelta_e5, vmagic_bias)), … in xnn_f32_raddextexp_ukernel__avx2_p5_x96_acc6()
|