| /aosp_15_r20/external/XNNPACK/src/f32-vsigmoid/gen/ |
| H A D | vsigmoid-avx2-rr1-p5-nr1fma-x72.c | 78 const __m256 vs8 = _mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(vn8), 23)); in xnn_f32_vsigmoid_ukernel__avx2_rr1_p5_nr1fma_x72() local
|
| H A D | vsigmoid-avx2-rr1-p5-nr2fma-x72.c | 78 const __m256 vs8 = _mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(vn8), 23)); in xnn_f32_vsigmoid_ukernel__avx2_rr1_p5_nr2fma_x72() local
|
| H A D | vsigmoid-avx2-rr1-p5-nr2fma-x80.c | 81 const __m256 vs8 = _mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(vn8), 23)); in xnn_f32_vsigmoid_ukernel__avx2_rr1_p5_nr2fma_x80() local
|
| H A D | vsigmoid-avx2-rr1-p5-nr1fma-x80.c | 81 const __m256 vs8 = _mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(vn8), 23)); in xnn_f32_vsigmoid_ukernel__avx2_rr1_p5_nr1fma_x80() local
|
| H A D | vsigmoid-avx-rr2-p5-div-x72.c | 97 const __m256 vs8 = _mm256_insertf128_ps(_mm256_castps128_ps256(vs_lo8), vs_hi8, 1); in xnn_f32_vsigmoid_ukernel__avx_rr2_p5_div_x72() local
|
| H A D | vsigmoid-avx-rr2-p5-div-x80.c | 100 const __m256 vs8 = _mm256_insertf128_ps(_mm256_castps128_ps256(vs_lo8), vs_hi8, 1); in xnn_f32_vsigmoid_ukernel__avx_rr2_p5_div_x80() local
|
| H A D | vsigmoid-avx-rr2-p5-nr2-x72.c | 98 const __m256 vs8 = _mm256_insertf128_ps(_mm256_castps128_ps256(vs_lo8), vs_hi8, 1); in xnn_f32_vsigmoid_ukernel__avx_rr2_p5_nr2_x72() local
|
| H A D | vsigmoid-avx-rr2-p5-nr2-x80.c | 101 const __m256 vs8 = _mm256_insertf128_ps(_mm256_castps128_ps256(vs_lo8), vs_hi8, 1); in xnn_f32_vsigmoid_ukernel__avx_rr2_p5_nr2_x80() local
|
| /aosp_15_r20/external/XNNPACK/src/f16-raddstoreexpminusmax/gen/ |
| H A D | neonfp16arith-rr2-p2-x96-acc3.c | 90 const float16x8_t vs8 = vreinterpretq_f16_s16(vshlq_n_s16(vreinterpretq_s16_f16(vn8), 10)); in xnn_f16_raddstoreexpminusmax_ukernel__neonfp16arith_rr2_p2_x96_acc3() local
|
| H A D | neonfp16arith-rr2-p2-x96.c | 88 const float16x8_t vs8 = vreinterpretq_f16_s16(vshlq_n_s16(vreinterpretq_s16_f16(vn8), 10)); in xnn_f16_raddstoreexpminusmax_ukernel__neonfp16arith_rr2_p2_x96() local
|
| H A D | neonfp16arith-rr2-p2-x96-acc6.c | 93 const float16x8_t vs8 = vreinterpretq_f16_s16(vshlq_n_s16(vreinterpretq_s16_f16(vn8), 10)); in xnn_f16_raddstoreexpminusmax_ukernel__neonfp16arith_rr2_p2_x96_acc6() local
|
| /aosp_15_r20/external/XNNPACK/src/f32-velu/gen/ |
| H A D | velu-avx2-rr1-lut4-p4-perm-x72.c | 114 __m256 vs8 = _mm256_castsi256_ps(_mm256_add_epi32(vl8, ven8)); in xnn_f32_velu_ukernel__avx2_rr1_lut4_p4_perm_x72() local
|
| H A D | velu-avx2-rr1-lut8-p4-perm-x72.c | 114 __m256 vs8 = _mm256_castsi256_ps(_mm256_add_epi32(vl8, ven8)); in xnn_f32_velu_ukernel__avx2_rr1_lut8_p4_perm_x72() local
|
| H A D | velu-avx2-rr1-lut16-p3-gather-x72.c | 125 __m256 vs8 = _mm256_castsi256_ps(_mm256_add_epi32(vl8, ven8)); in xnn_f32_velu_ukernel__avx2_rr1_lut16_p3_gather_x72() local
|
| H A D | velu-avx2-rr1-lut8-p4-perm-x80.c | 120 __m256 vs8 = _mm256_castsi256_ps(_mm256_add_epi32(vl8, ven8)); in xnn_f32_velu_ukernel__avx2_rr1_lut8_p4_perm_x80() local
|
| H A D | velu-avx2-rr1-lut4-p4-perm-x80.c | 120 __m256 vs8 = _mm256_castsi256_ps(_mm256_add_epi32(vl8, ven8)); in xnn_f32_velu_ukernel__avx2_rr1_lut4_p4_perm_x80() local
|
| H A D | velu-avx2-rr1-lut16-p3-gather-x80.c | 132 __m256 vs8 = _mm256_castsi256_ps(_mm256_add_epi32(vl8, ven8)); in xnn_f32_velu_ukernel__avx2_rr1_lut16_p3_gather_x80() local
|
| /aosp_15_r20/external/XNNPACK/src/f32-raddextexp/gen/ |
| H A D | avx2-p5-x96-acc2.c | 229 …const __m256 vs8 = _mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(_mm256_add_ps(vdelta_… in xnn_f32_raddextexp_ukernel__avx2_p5_x96_acc2() local
|
| H A D | avx2-p5-x96-acc3.c | 233 …const __m256 vs8 = _mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(_mm256_add_ps(vdelta_… in xnn_f32_raddextexp_ukernel__avx2_p5_x96_acc3() local
|
| H A D | avx2-p5-x80-acc5.c | 219 …const __m256 vs8 = _mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(_mm256_add_ps(vdelta_… in xnn_f32_raddextexp_ukernel__avx2_p5_x80_acc5() local
|
| H A D | avx2-p5-x96-acc6.c | 245 …const __m256 vs8 = _mm256_castsi256_ps(_mm256_slli_epi32(_mm256_castps_si256(_mm256_add_ps(vdelta_… in xnn_f32_raddextexp_ukernel__avx2_p5_x96_acc6() local
|