| /external/XNNPACK/src/f32-dwconv/gen/ |
| D | up16x3-minmax-avx.c | 56 __m256 vacc89ABCDEFp0 = _mm256_load_ps(w + 8); in xnn_f32_dwconv_minmax_ukernel_up16x3__avx() local
|
| D | up16x3-minmax-fma3.c | 56 __m256 vacc89ABCDEFp0 = _mm256_load_ps(w + 8); in xnn_f32_dwconv_minmax_ukernel_up16x3__fma3() local
|
| D | up16x3-minmax-avx-acc2.c | 56 __m256 vacc89ABCDEFp0 = _mm256_load_ps(w + 8); in xnn_f32_dwconv_minmax_ukernel_up16x3__avx_acc2() local
|
| D | up16x3-minmax-fma3-acc2.c | 56 __m256 vacc89ABCDEFp0 = _mm256_load_ps(w + 8); in xnn_f32_dwconv_minmax_ukernel_up16x3__fma3_acc2() local
|
| D | up16x4-minmax-avx.c | 61 __m256 vacc89ABCDEFp0 = _mm256_load_ps(w + 8); in xnn_f32_dwconv_minmax_ukernel_up16x4__avx() local
|
| D | up16x4-minmax-fma3.c | 61 __m256 vacc89ABCDEFp0 = _mm256_load_ps(w + 8); in xnn_f32_dwconv_minmax_ukernel_up16x4__fma3() local
|
| D | up16x4-minmax-fma3-acc2.c | 61 __m256 vacc89ABCDEFp0 = _mm256_load_ps(w + 8); in xnn_f32_dwconv_minmax_ukernel_up16x4__fma3_acc2() local
|
| D | up16x4-minmax-avx-acc2.c | 61 __m256 vacc89ABCDEFp0 = _mm256_load_ps(w + 8); in xnn_f32_dwconv_minmax_ukernel_up16x4__avx_acc2() local
|
| /external/XNNPACK/src/f16-dwconv/gen/ |
| D | up16x3-minmax-fma3.c | 59 __m256 vacc89ABCDEFp0 = _mm256_cvtph_ps(_mm_load_si128((const __m128i*) (w + 8))); in xnn_f16_dwconv_minmax_ukernel_up16x3__fma3() local
|
| D | up16x3-minmax-neonfp16arith.c | 58 float16x8_t vacc89ABCDEFp0 = vld1q_f16(w); w += 8; in xnn_f16_dwconv_minmax_ukernel_up16x3__neonfp16arith() local
|
| D | up16x3-minmax-fma3-acc2.c | 59 __m256 vacc89ABCDEFp0 = _mm256_cvtph_ps(_mm_load_si128((const __m128i*) (w + 8))); in xnn_f16_dwconv_minmax_ukernel_up16x3__fma3_acc2() local
|
| D | up16x3-minmax-neonfp16arith-acc2.c | 58 float16x8_t vacc89ABCDEFp0 = vld1q_f16(w); w += 8; in xnn_f16_dwconv_minmax_ukernel_up16x3__neonfp16arith_acc2() local
|
| D | up16x4-minmax-fma3.c | 64 __m256 vacc89ABCDEFp0 = _mm256_cvtph_ps(_mm_load_si128((const __m128i*) (w + 8))); in xnn_f16_dwconv_minmax_ukernel_up16x4__fma3() local
|
| D | up16x4-minmax-neonfp16arith.c | 63 float16x8_t vacc89ABCDEFp0 = vld1q_f16(w); w += 8; in xnn_f16_dwconv_minmax_ukernel_up16x4__neonfp16arith() local
|
| D | up16x4-minmax-neonfp16arith-acc2.c | 63 float16x8_t vacc89ABCDEFp0 = vld1q_f16(w); w += 8; in xnn_f16_dwconv_minmax_ukernel_up16x4__neonfp16arith_acc2() local
|
| D | up16x4-minmax-fma3-acc2.c | 64 __m256 vacc89ABCDEFp0 = _mm256_cvtph_ps(_mm_load_si128((const __m128i*) (w + 8))); in xnn_f16_dwconv_minmax_ukernel_up16x4__fma3_acc2() local
|
| D | up32x3-minmax-neonfp16arith.c | 58 float16x8_t vacc89ABCDEFp0 = vld1q_f16(w); w += 8; in xnn_f16_dwconv_minmax_ukernel_up32x3__neonfp16arith() local
|
| D | up32x3-minmax-fma3.c | 59 __m256 vacc89ABCDEFp0 = _mm256_cvtph_ps(_mm_load_si128((const __m128i*) (w + 8))); in xnn_f16_dwconv_minmax_ukernel_up32x3__fma3() local
|
| D | up32x3-minmax-fma3-acc2.c | 59 __m256 vacc89ABCDEFp0 = _mm256_cvtph_ps(_mm_load_si128((const __m128i*) (w + 8))); in xnn_f16_dwconv_minmax_ukernel_up32x3__fma3_acc2() local
|
| D | up32x3-minmax-neonfp16arith-acc2.c | 58 float16x8_t vacc89ABCDEFp0 = vld1q_f16(w); w += 8; in xnn_f16_dwconv_minmax_ukernel_up32x3__neonfp16arith_acc2() local
|
| D | up32x4-minmax-neonfp16arith.c | 63 float16x8_t vacc89ABCDEFp0 = vld1q_f16(w); w += 8; in xnn_f16_dwconv_minmax_ukernel_up32x4__neonfp16arith() local
|
| D | up32x4-minmax-fma3.c | 64 __m256 vacc89ABCDEFp0 = _mm256_cvtph_ps(_mm_load_si128((const __m128i*) (w + 8))); in xnn_f16_dwconv_minmax_ukernel_up32x4__fma3() local
|
| D | up32x4-minmax-fma3-acc2.c | 64 __m256 vacc89ABCDEFp0 = _mm256_cvtph_ps(_mm_load_si128((const __m128i*) (w + 8))); in xnn_f16_dwconv_minmax_ukernel_up32x4__fma3_acc2() local
|
| D | up32x4-minmax-neonfp16arith-acc2.c | 63 float16x8_t vacc89ABCDEFp0 = vld1q_f16(w); w += 8; in xnn_f16_dwconv_minmax_ukernel_up32x4__neonfp16arith_acc2() local
|
| D | up16x9-minmax-neonfp16arith.c | 88 float16x8_t vacc89ABCDEFp0 = vld1q_f16(w); w += 8; in xnn_f16_dwconv_minmax_ukernel_up16x9__neonfp16arith() local
|