/external/XNNPACK/src/f32-vscaleexpminusmax/gen/ |
D | avx2-p5-x48.c | 78 const __m256 vs5 = _mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(vn5), 23)); in xnn_f32_vscaleexpminusmax_ukernel__avx2_p5_x48() local 142 vt5 = _mm256_mul_ps(vt5, vs5); in xnn_f32_vscaleexpminusmax_ukernel__avx2_p5_x48() 149 __m256 vf5 = _mm256_fmadd_ps(vt5, vp5, vs5); in xnn_f32_vscaleexpminusmax_ukernel__avx2_p5_x48()
|
D | avx2-p5-x56.c | 81 const __m256 vs5 = _mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(vn5), 23)); in xnn_f32_vscaleexpminusmax_ukernel__avx2_p5_x56() local 153 vt5 = _mm256_mul_ps(vt5, vs5); in xnn_f32_vscaleexpminusmax_ukernel__avx2_p5_x56() 161 __m256 vf5 = _mm256_fmadd_ps(vt5, vp5, vs5); in xnn_f32_vscaleexpminusmax_ukernel__avx2_p5_x56()
|
D | avx2-p5-x64.c | 84 const __m256 vs5 = _mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(vn5), 23)); in xnn_f32_vscaleexpminusmax_ukernel__avx2_p5_x64() local 164 vt5 = _mm256_mul_ps(vt5, vs5); in xnn_f32_vscaleexpminusmax_ukernel__avx2_p5_x64() 173 __m256 vf5 = _mm256_fmadd_ps(vt5, vp5, vs5); in xnn_f32_vscaleexpminusmax_ukernel__avx2_p5_x64()
|
D | avx2-p5-x72.c | 87 const __m256 vs5 = _mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(vn5), 23)); in xnn_f32_vscaleexpminusmax_ukernel__avx2_p5_x72() local 175 vt5 = _mm256_mul_ps(vt5, vs5); in xnn_f32_vscaleexpminusmax_ukernel__avx2_p5_x72() 185 __m256 vf5 = _mm256_fmadd_ps(vt5, vp5, vs5); in xnn_f32_vscaleexpminusmax_ukernel__avx2_p5_x72()
|
D | avx2-p5-x80.c | 90 const __m256 vs5 = _mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(vn5), 23)); in xnn_f32_vscaleexpminusmax_ukernel__avx2_p5_x80() local 186 vt5 = _mm256_mul_ps(vt5, vs5); in xnn_f32_vscaleexpminusmax_ukernel__avx2_p5_x80() 197 __m256 vf5 = _mm256_fmadd_ps(vt5, vp5, vs5); in xnn_f32_vscaleexpminusmax_ukernel__avx2_p5_x80()
|
/external/XNNPACK/src/f32-sigmoid/gen/ |
D | avx2-rr1-p5-div-x48.c | 86 const __m256 vs5 = _mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(vn5), 23)); in xnn_f32_sigmoid_ukernel__avx2_rr1_p5_div_x48() local 142 vt5 = _mm256_mul_ps(vt5, vs5); in xnn_f32_sigmoid_ukernel__avx2_rr1_p5_div_x48() 149 const __m256 ve5 = _mm256_fmadd_ps(vt5, vp5, vs5); in xnn_f32_sigmoid_ukernel__avx2_rr1_p5_div_x48()
|
D | avx2-rr1-p5-nr1fma-x48.c | 86 const __m256 vs5 = _mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(vn5), 23)); in xnn_f32_sigmoid_ukernel__avx2_rr1_p5_nr1fma_x48() local 142 vt5 = _mm256_mul_ps(vt5, vs5); in xnn_f32_sigmoid_ukernel__avx2_rr1_p5_nr1fma_x48() 149 const __m256 ve5 = _mm256_fmadd_ps(vt5, vp5, vs5); in xnn_f32_sigmoid_ukernel__avx2_rr1_p5_nr1fma_x48()
|
D | avx2-rr1-p5-div-x56.c | 89 const __m256 vs5 = _mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(vn5), 23)); in xnn_f32_sigmoid_ukernel__avx2_rr1_p5_div_x56() local 152 vt5 = _mm256_mul_ps(vt5, vs5); in xnn_f32_sigmoid_ukernel__avx2_rr1_p5_div_x56() 160 const __m256 ve5 = _mm256_fmadd_ps(vt5, vp5, vs5); in xnn_f32_sigmoid_ukernel__avx2_rr1_p5_div_x56()
|
D | avx2-rr1-p5-nr2fma-x48.c | 86 const __m256 vs5 = _mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(vn5), 23)); in xnn_f32_sigmoid_ukernel__avx2_rr1_p5_nr2fma_x48() local 142 vt5 = _mm256_mul_ps(vt5, vs5); in xnn_f32_sigmoid_ukernel__avx2_rr1_p5_nr2fma_x48() 149 const __m256 ve5 = _mm256_fmadd_ps(vt5, vp5, vs5); in xnn_f32_sigmoid_ukernel__avx2_rr1_p5_nr2fma_x48()
|
D | avx2-rr1-p5-div-x64.c | 92 const __m256 vs5 = _mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(vn5), 23)); in xnn_f32_sigmoid_ukernel__avx2_rr1_p5_div_x64() local 162 vt5 = _mm256_mul_ps(vt5, vs5); in xnn_f32_sigmoid_ukernel__avx2_rr1_p5_div_x64() 171 const __m256 ve5 = _mm256_fmadd_ps(vt5, vp5, vs5); in xnn_f32_sigmoid_ukernel__avx2_rr1_p5_div_x64()
|
D | avx2-rr1-p5-nr1fma-x56.c | 89 const __m256 vs5 = _mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(vn5), 23)); in xnn_f32_sigmoid_ukernel__avx2_rr1_p5_nr1fma_x56() local 152 vt5 = _mm256_mul_ps(vt5, vs5); in xnn_f32_sigmoid_ukernel__avx2_rr1_p5_nr1fma_x56() 160 const __m256 ve5 = _mm256_fmadd_ps(vt5, vp5, vs5); in xnn_f32_sigmoid_ukernel__avx2_rr1_p5_nr1fma_x56()
|
D | avx2-rr1-p5-div-x80.c | 98 const __m256 vs5 = _mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(vn5), 23)); in xnn_f32_sigmoid_ukernel__avx2_rr1_p5_div_x80() local 182 vt5 = _mm256_mul_ps(vt5, vs5); in xnn_f32_sigmoid_ukernel__avx2_rr1_p5_div_x80() 193 const __m256 ve5 = _mm256_fmadd_ps(vt5, vp5, vs5); in xnn_f32_sigmoid_ukernel__avx2_rr1_p5_div_x80()
|
D | avx2-rr1-p5-div-x72.c | 95 const __m256 vs5 = _mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(vn5), 23)); in xnn_f32_sigmoid_ukernel__avx2_rr1_p5_div_x72() local 172 vt5 = _mm256_mul_ps(vt5, vs5); in xnn_f32_sigmoid_ukernel__avx2_rr1_p5_div_x72() 182 const __m256 ve5 = _mm256_fmadd_ps(vt5, vp5, vs5); in xnn_f32_sigmoid_ukernel__avx2_rr1_p5_div_x72()
|
/external/XNNPACK/src/f32-raddexpminusmax/gen/ |
D | avx2-p5-x64.c | 82 const __m256 vs5 = _mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(vn5), 23)); in xnn_f32_raddexpminusmax_ukernel__avx2_p5_x64() local 162 vt5 = _mm256_mul_ps(vt5, vs5); in xnn_f32_raddexpminusmax_ukernel__avx2_p5_x64() 171 __m256 vf5 = _mm256_fmadd_ps(vt5, vp5, vs5); in xnn_f32_raddexpminusmax_ukernel__avx2_p5_x64()
|
D | avx2-p5-x64-acc2.c | 83 const __m256 vs5 = _mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(vn5), 23)); in xnn_f32_raddexpminusmax_ukernel__avx2_p5_x64_acc2() local 163 vt5 = _mm256_mul_ps(vt5, vs5); in xnn_f32_raddexpminusmax_ukernel__avx2_p5_x64_acc2() 172 __m256 vf5 = _mm256_fmadd_ps(vt5, vp5, vs5); in xnn_f32_raddexpminusmax_ukernel__avx2_p5_x64_acc2()
|
D | avx2-p5-x64-acc4.c | 85 const __m256 vs5 = _mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(vn5), 23)); in xnn_f32_raddexpminusmax_ukernel__avx2_p5_x64_acc4() local 165 vt5 = _mm256_mul_ps(vt5, vs5); in xnn_f32_raddexpminusmax_ukernel__avx2_p5_x64_acc4() 174 __m256 vf5 = _mm256_fmadd_ps(vt5, vp5, vs5); in xnn_f32_raddexpminusmax_ukernel__avx2_p5_x64_acc4()
|
D | avx2-p5-x72.c | 85 const __m256 vs5 = _mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(vn5), 23)); in xnn_f32_raddexpminusmax_ukernel__avx2_p5_x72() local 173 vt5 = _mm256_mul_ps(vt5, vs5); in xnn_f32_raddexpminusmax_ukernel__avx2_p5_x72() 183 __m256 vf5 = _mm256_fmadd_ps(vt5, vp5, vs5); in xnn_f32_raddexpminusmax_ukernel__avx2_p5_x72()
|
D | avx2-p5-x72-acc3.c | 87 const __m256 vs5 = _mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(vn5), 23)); in xnn_f32_raddexpminusmax_ukernel__avx2_p5_x72_acc3() local 175 vt5 = _mm256_mul_ps(vt5, vs5); in xnn_f32_raddexpminusmax_ukernel__avx2_p5_x72_acc3() 185 __m256 vf5 = _mm256_fmadd_ps(vt5, vp5, vs5); in xnn_f32_raddexpminusmax_ukernel__avx2_p5_x72_acc3()
|
D | avx2-p5-x80.c | 88 const __m256 vs5 = _mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(vn5), 23)); in xnn_f32_raddexpminusmax_ukernel__avx2_p5_x80() local 184 vt5 = _mm256_mul_ps(vt5, vs5); in xnn_f32_raddexpminusmax_ukernel__avx2_p5_x80() 195 __m256 vf5 = _mm256_fmadd_ps(vt5, vp5, vs5); in xnn_f32_raddexpminusmax_ukernel__avx2_p5_x80()
|
D | avx2-p5-x80-acc2.c | 89 const __m256 vs5 = _mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(vn5), 23)); in xnn_f32_raddexpminusmax_ukernel__avx2_p5_x80_acc2() local 185 vt5 = _mm256_mul_ps(vt5, vs5); in xnn_f32_raddexpminusmax_ukernel__avx2_p5_x80_acc2() 196 __m256 vf5 = _mm256_fmadd_ps(vt5, vp5, vs5); in xnn_f32_raddexpminusmax_ukernel__avx2_p5_x80_acc2()
|
/external/XNNPACK/src/f32-raddstoreexpminusmax/gen/ |
D | avx2-p5-x64-acc2.c | 84 const __m256 vs5 = _mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(vn5), 23)); in xnn_f32_raddstoreexpminusmax_ukernel__avx2_p5_x64_acc2() local 164 vt5 = _mm256_mul_ps(vt5, vs5); in xnn_f32_raddstoreexpminusmax_ukernel__avx2_p5_x64_acc2() 173 __m256 vf5 = _mm256_fmadd_ps(vt5, vp5, vs5); in xnn_f32_raddstoreexpminusmax_ukernel__avx2_p5_x64_acc2()
|
D | avx2-p5-x64.c | 83 const __m256 vs5 = _mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(vn5), 23)); in xnn_f32_raddstoreexpminusmax_ukernel__avx2_p5_x64() local 163 vt5 = _mm256_mul_ps(vt5, vs5); in xnn_f32_raddstoreexpminusmax_ukernel__avx2_p5_x64() 172 __m256 vf5 = _mm256_fmadd_ps(vt5, vp5, vs5); in xnn_f32_raddstoreexpminusmax_ukernel__avx2_p5_x64()
|
D | avx2-p5-x64-acc4.c | 86 const __m256 vs5 = _mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(vn5), 23)); in xnn_f32_raddstoreexpminusmax_ukernel__avx2_p5_x64_acc4() local 166 vt5 = _mm256_mul_ps(vt5, vs5); in xnn_f32_raddstoreexpminusmax_ukernel__avx2_p5_x64_acc4() 175 __m256 vf5 = _mm256_fmadd_ps(vt5, vp5, vs5); in xnn_f32_raddstoreexpminusmax_ukernel__avx2_p5_x64_acc4()
|
D | avx2-p5-x72.c | 86 const __m256 vs5 = _mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(vn5), 23)); in xnn_f32_raddstoreexpminusmax_ukernel__avx2_p5_x72() local 174 vt5 = _mm256_mul_ps(vt5, vs5); in xnn_f32_raddstoreexpminusmax_ukernel__avx2_p5_x72() 184 __m256 vf5 = _mm256_fmadd_ps(vt5, vp5, vs5); in xnn_f32_raddstoreexpminusmax_ukernel__avx2_p5_x72()
|
D | avx2-p5-x72-acc3.c | 88 const __m256 vs5 = _mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(vn5), 23)); in xnn_f32_raddstoreexpminusmax_ukernel__avx2_p5_x72_acc3() local 176 vt5 = _mm256_mul_ps(vt5, vs5); in xnn_f32_raddstoreexpminusmax_ukernel__avx2_p5_x72_acc3() 186 __m256 vf5 = _mm256_fmadd_ps(vt5, vp5, vs5); in xnn_f32_raddstoreexpminusmax_ukernel__avx2_p5_x72_acc3()
|