| /external/XNNPACK/src/f32-vsigmoid/gen/ |
| D | vsigmoid-avx2-rr1-p5-div-x72.c | 58 const __m256 vz8 = _mm256_or_ps(vx8, vsign_mask); in xnn_f32_vsigmoid_ukernel__avx2_rr1_p5_div_x72() local
68 __m256 vn8 = _mm256_fmadd_ps(vz8, vlog2e, vmagic_bias); in xnn_f32_vsigmoid_ukernel__avx2_rr1_p5_div_x72()
98 __m256 vt8 = _mm256_fmadd_ps(vn8, vminus_ln2, vz8); in xnn_f32_vsigmoid_ukernel__avx2_rr1_p5_div_x72()
188 vf8 = _mm256_andnot_ps(_mm256_cmp_ps(vz8, vdenorm_cutoff, _CMP_LT_OS), vf8); in xnn_f32_vsigmoid_ukernel__avx2_rr1_p5_div_x72()
|
| D | vsigmoid-avx2-rr1-p5-nr1fma-x72.c | 58 const __m256 vz8 = _mm256_or_ps(vx8, vsign_mask); in xnn_f32_vsigmoid_ukernel__avx2_rr1_p5_nr1fma_x72() local
68 __m256 vn8 = _mm256_fmadd_ps(vz8, vlog2e, vmagic_bias); in xnn_f32_vsigmoid_ukernel__avx2_rr1_p5_nr1fma_x72()
98 __m256 vt8 = _mm256_fmadd_ps(vn8, vminus_ln2, vz8); in xnn_f32_vsigmoid_ukernel__avx2_rr1_p5_nr1fma_x72()
209 vf8 = _mm256_andnot_ps(_mm256_cmp_ps(vz8, vdenorm_cutoff, _CMP_LT_OS), vf8); in xnn_f32_vsigmoid_ukernel__avx2_rr1_p5_nr1fma_x72()
|
| D | vsigmoid-avx2-rr1-p5-div-x80.c | 59 const __m256 vz8 = _mm256_or_ps(vx8, vsign_mask); in xnn_f32_vsigmoid_ukernel__avx2_rr1_p5_div_x80() local
70 __m256 vn8 = _mm256_fmadd_ps(vz8, vlog2e, vmagic_bias); in xnn_f32_vsigmoid_ukernel__avx2_rr1_p5_div_x80()
103 __m256 vt8 = _mm256_fmadd_ps(vn8, vminus_ln2, vz8); in xnn_f32_vsigmoid_ukernel__avx2_rr1_p5_div_x80()
202 vf8 = _mm256_andnot_ps(_mm256_cmp_ps(vz8, vdenorm_cutoff, _CMP_LT_OS), vf8); in xnn_f32_vsigmoid_ukernel__avx2_rr1_p5_div_x80()
|
| D | vsigmoid-avx2-rr1-p5-nr2fma-x80.c | 59 const __m256 vz8 = _mm256_or_ps(vx8, vsign_mask); in xnn_f32_vsigmoid_ukernel__avx2_rr1_p5_nr2fma_x80() local
70 __m256 vn8 = _mm256_fmadd_ps(vz8, vlog2e, vmagic_bias); in xnn_f32_vsigmoid_ukernel__avx2_rr1_p5_nr2fma_x80()
103 __m256 vt8 = _mm256_fmadd_ps(vn8, vminus_ln2, vz8); in xnn_f32_vsigmoid_ukernel__avx2_rr1_p5_nr2fma_x80()
235 vf8 = _mm256_andnot_ps(_mm256_cmp_ps(vz8, vdenorm_cutoff, _CMP_LT_OS), vf8); in xnn_f32_vsigmoid_ukernel__avx2_rr1_p5_nr2fma_x80()
|
| D | vsigmoid-avx2-rr1-p5-nr1fma-x80.c | 59 const __m256 vz8 = _mm256_or_ps(vx8, vsign_mask); in xnn_f32_vsigmoid_ukernel__avx2_rr1_p5_nr1fma_x80() local
70 __m256 vn8 = _mm256_fmadd_ps(vz8, vlog2e, vmagic_bias); in xnn_f32_vsigmoid_ukernel__avx2_rr1_p5_nr1fma_x80()
103 __m256 vt8 = _mm256_fmadd_ps(vn8, vminus_ln2, vz8); in xnn_f32_vsigmoid_ukernel__avx2_rr1_p5_nr1fma_x80()
225 vf8 = _mm256_andnot_ps(_mm256_cmp_ps(vz8, vdenorm_cutoff, _CMP_LT_OS), vf8); in xnn_f32_vsigmoid_ukernel__avx2_rr1_p5_nr1fma_x80()
|
| D | vsigmoid-avx-rr2-p5-div-x72.c | 59 const __m256 vz8 = _mm256_or_ps(vx8, vsign_mask); in xnn_f32_vsigmoid_ukernel__avx_rr2_p5_div_x72() local
69 __m256 vn8 = _mm256_add_ps(_mm256_mul_ps(vz8, vlog2e), vmagic_bias); in xnn_f32_vsigmoid_ukernel__avx_rr2_p5_div_x72()
117 __m256 vt8 = _mm256_add_ps(_mm256_mul_ps(vn8, vminus_ln2_hi), vz8); in xnn_f32_vsigmoid_ukernel__avx_rr2_p5_div_x72()
217 vf8 = _mm256_andnot_ps(_mm256_cmp_ps(vz8, vdenorm_cutoff, _CMP_LT_OS), vf8); in xnn_f32_vsigmoid_ukernel__avx_rr2_p5_div_x72()
|
| D | vsigmoid-avx2-rr1-p5-nr2fma-x72.c | 58 const __m256 vz8 = _mm256_or_ps(vx8, vsign_mask); in xnn_f32_vsigmoid_ukernel__avx2_rr1_p5_nr2fma_x72() local
68 __m256 vn8 = _mm256_fmadd_ps(vz8, vlog2e, vmagic_bias); in xnn_f32_vsigmoid_ukernel__avx2_rr1_p5_nr2fma_x72()
98 __m256 vt8 = _mm256_fmadd_ps(vn8, vminus_ln2, vz8); in xnn_f32_vsigmoid_ukernel__avx2_rr1_p5_nr2fma_x72()
218 vf8 = _mm256_andnot_ps(_mm256_cmp_ps(vz8, vdenorm_cutoff, _CMP_LT_OS), vf8); in xnn_f32_vsigmoid_ukernel__avx2_rr1_p5_nr2fma_x72()
|
| D | vsigmoid-avx-rr2-p5-div-x80.c | 60 const __m256 vz8 = _mm256_or_ps(vx8, vsign_mask); in xnn_f32_vsigmoid_ukernel__avx_rr2_p5_div_x80() local
71 __m256 vn8 = _mm256_add_ps(_mm256_mul_ps(vz8, vlog2e), vmagic_bias); in xnn_f32_vsigmoid_ukernel__avx_rr2_p5_div_x80()
124 __m256 vt8 = _mm256_add_ps(_mm256_mul_ps(vn8, vminus_ln2_hi), vz8); in xnn_f32_vsigmoid_ukernel__avx_rr2_p5_div_x80()
234 vf8 = _mm256_andnot_ps(_mm256_cmp_ps(vz8, vdenorm_cutoff, _CMP_LT_OS), vf8); in xnn_f32_vsigmoid_ukernel__avx_rr2_p5_div_x80()
|
| D | vsigmoid-avx-rr2-p5-nr2-x72.c | 60 const __m256 vz8 = _mm256_or_ps(vx8, vsign_mask); in xnn_f32_vsigmoid_ukernel__avx_rr2_p5_nr2_x72() local
70 __m256 vn8 = _mm256_add_ps(_mm256_mul_ps(vz8, vlog2e), vmagic_bias); in xnn_f32_vsigmoid_ukernel__avx_rr2_p5_nr2_x72()
118 __m256 vt8 = _mm256_add_ps(_mm256_mul_ps(vn8, vminus_ln2_hi), vz8); in xnn_f32_vsigmoid_ukernel__avx_rr2_p5_nr2_x72()
247 vf8 = _mm256_andnot_ps(_mm256_cmp_ps(vz8, vdenorm_cutoff, _CMP_LT_OS), vf8); in xnn_f32_vsigmoid_ukernel__avx_rr2_p5_nr2_x72()
|
| D | vsigmoid-avx-rr2-p5-nr2-x80.c | 61 const __m256 vz8 = _mm256_or_ps(vx8, vsign_mask); in xnn_f32_vsigmoid_ukernel__avx_rr2_p5_nr2_x80() local
72 __m256 vn8 = _mm256_add_ps(_mm256_mul_ps(vz8, vlog2e), vmagic_bias); in xnn_f32_vsigmoid_ukernel__avx_rr2_p5_nr2_x80()
125 __m256 vt8 = _mm256_add_ps(_mm256_mul_ps(vn8, vminus_ln2_hi), vz8); in xnn_f32_vsigmoid_ukernel__avx_rr2_p5_nr2_x80()
267 vf8 = _mm256_andnot_ps(_mm256_cmp_ps(vz8, vdenorm_cutoff, _CMP_LT_OS), vf8); in xnn_f32_vsigmoid_ukernel__avx_rr2_p5_nr2_x80()
|
| /external/XNNPACK/src/f32-velu/gen/ |
| D | velu-avx2-rr1-p6-x72.c | 59 const __m256 vz8 = _mm256_max_ps(vsat_cutoff, _mm256_mul_ps(vx8, vprescale)); in xnn_f32_velu_ukernel__avx2_rr1_p6_x72() local
69 __m256 vn8 = _mm256_fmadd_ps(vz8, vlog2e, vmagic_bias); in xnn_f32_velu_ukernel__avx2_rr1_p6_x72()
98 __m256 vt8 = _mm256_fmadd_ps(vn8, vminus_ln2, vz8); in xnn_f32_velu_ukernel__avx2_rr1_p6_x72()
|
| D | velu-avx2-rr1-lut16-p3-gather-x72.c | 59 const __m256 vz8 = _mm256_max_ps(vsat_cutoff, _mm256_mul_ps(vx8, vprescale)); in xnn_f32_velu_ukernel__avx2_rr1_lut16_p3_gather_x72() local
69 __m256 vn8 = _mm256_fmadd_ps(vz8, vlog2e, vmagic_bias); in xnn_f32_velu_ukernel__avx2_rr1_lut16_p3_gather_x72()
126 __m256 vt8 = _mm256_fmadd_ps(vn8, vminus_ln2, vz8); in xnn_f32_velu_ukernel__avx2_rr1_lut16_p3_gather_x72()
|
| D | velu-avx2-rr1-lut4-p4-perm-x72.c | 58 const __m256 vz8 = _mm256_max_ps(vsat_cutoff, _mm256_mul_ps(vx8, vprescale)); in xnn_f32_velu_ukernel__avx2_rr1_lut4_p4_perm_x72() local
68 __m256 vn8 = _mm256_fmadd_ps(vz8, vlog2e, vmagic_bias); in xnn_f32_velu_ukernel__avx2_rr1_lut4_p4_perm_x72()
115 __m256 vt8 = _mm256_fmadd_ps(vn8, vminus_ln2, vz8); in xnn_f32_velu_ukernel__avx2_rr1_lut4_p4_perm_x72()
|
| D | velu-avx2-rr1-lut8-p4-perm-x72.c | 58 const __m256 vz8 = _mm256_max_ps(vsat_cutoff, _mm256_mul_ps(vx8, vprescale)); in xnn_f32_velu_ukernel__avx2_rr1_lut8_p4_perm_x72() local
68 __m256 vn8 = _mm256_fmadd_ps(vz8, vlog2e, vmagic_bias); in xnn_f32_velu_ukernel__avx2_rr1_lut8_p4_perm_x72()
115 __m256 vt8 = _mm256_fmadd_ps(vn8, vminus_ln2, vz8); in xnn_f32_velu_ukernel__avx2_rr1_lut8_p4_perm_x72()
|
| D | velu-avx2-rr1-p6-x80.c | 60 const __m256 vz8 = _mm256_max_ps(vsat_cutoff, _mm256_mul_ps(vx8, vprescale)); in xnn_f32_velu_ukernel__avx2_rr1_p6_x80() local
71 __m256 vn8 = _mm256_fmadd_ps(vz8, vlog2e, vmagic_bias); in xnn_f32_velu_ukernel__avx2_rr1_p6_x80()
103 __m256 vt8 = _mm256_fmadd_ps(vn8, vminus_ln2, vz8); in xnn_f32_velu_ukernel__avx2_rr1_p6_x80()
|
| D | velu-avx2-rr1-lut16-p3-gather-x80.c | 60 const __m256 vz8 = _mm256_max_ps(vsat_cutoff, _mm256_mul_ps(vx8, vprescale)); in xnn_f32_velu_ukernel__avx2_rr1_lut16_p3_gather_x80() local
71 __m256 vn8 = _mm256_fmadd_ps(vz8, vlog2e, vmagic_bias); in xnn_f32_velu_ukernel__avx2_rr1_lut16_p3_gather_x80()
133 __m256 vt8 = _mm256_fmadd_ps(vn8, vminus_ln2, vz8); in xnn_f32_velu_ukernel__avx2_rr1_lut16_p3_gather_x80()
|
| D | velu-avx2-rr1-lut4-p4-perm-x80.c | 59 const __m256 vz8 = _mm256_max_ps(vsat_cutoff, _mm256_mul_ps(vx8, vprescale)); in xnn_f32_velu_ukernel__avx2_rr1_lut4_p4_perm_x80() local
70 __m256 vn8 = _mm256_fmadd_ps(vz8, vlog2e, vmagic_bias); in xnn_f32_velu_ukernel__avx2_rr1_lut4_p4_perm_x80()
121 __m256 vt8 = _mm256_fmadd_ps(vn8, vminus_ln2, vz8); in xnn_f32_velu_ukernel__avx2_rr1_lut4_p4_perm_x80()
|
| D | velu-avx2-rr1-lut8-p4-perm-x80.c | 59 const __m256 vz8 = _mm256_max_ps(vsat_cutoff, _mm256_mul_ps(vx8, vprescale)); in xnn_f32_velu_ukernel__avx2_rr1_lut8_p4_perm_x80() local
70 __m256 vn8 = _mm256_fmadd_ps(vz8, vlog2e, vmagic_bias); in xnn_f32_velu_ukernel__avx2_rr1_lut8_p4_perm_x80()
121 __m256 vt8 = _mm256_fmadd_ps(vn8, vminus_ln2, vz8); in xnn_f32_velu_ukernel__avx2_rr1_lut8_p4_perm_x80()
|
| /external/hyphenation-patterns/pl/ |
| D | hyph-pl.pat.txt | 2311 .vz8
|