| /external/XNNPACK/src/f32-conv-hwc2chw/ |
| D | 3x3s2p1c3x4-neon-2x2.c | 144 const float32x4_t vi4x2 = vld1q_f32(i4); i4 += 4; in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__neon_2x2() local
396 float32x4_t vi4x2 = vmovq_n_f32(0.0f); in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__neon_2x2() local
|
| D | 3x3s2p1c3x4-sse-2x2.c | 144 const __m128 vi4x2 = _mm_loadu_ps(i4); i4 += 4; in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__sse_2x2() local
398 __m128 vi4x2 = _mm_setzero_ps(); in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__sse_2x2() local
|
| D | 3x3s2p1c3x4-wasmsimd-2x2.c | 145 const v128_t vi4x2 = wasm_v128_load(i4); i4 += 4; in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__wasmsimd_2x2() local
398 v128_t vi4x2 = vzero; in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__wasmsimd_2x2() local
|
| D | 3x3s2p1c3x4-neonfma-2x2.c | 144 const float32x4_t vi4x2 = vld1q_f32(i4); i4 += 4; in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__neonfma_2x2() local
398 float32x4_t vi4x2 = vmovq_n_f32(0.0f); in xnn_f32_conv_hwc2chw_ukernel_3x3s2p1c3x4__neonfma_2x2() local
|
| /external/XNNPACK/src/f32-dwconv2d-chw/gen/ |
| D | 3x3p1-minmax-scalar-3x1.c | 87 const float vi4x2 = *i4++; in xnn_f32_dwconv2d_chw_ukernel_3x3p1__scalar_3x1() local
|
| D | 3x3s2p1-minmax-scalar-2x1.c | 103 const float vi4x2 = i4[1]; in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__scalar_2x1() local
|
| D | 3x3s2p1-minmax-scalar-2x1-acc2.c | 103 const float vi4x2 = i4[1]; in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__scalar_2x1_acc2() local
|
| D | 3x3p1-minmax-scalar-4x1.c | 95 const float vi4x2 = *i4++; in xnn_f32_dwconv2d_chw_ukernel_3x3p1__scalar_4x1() local
|
| D | 3x3s2p1-minmax-scalar-3x1.c | 120 const float vi4x2 = i4[1]; in xnn_f32_dwconv2d_chw_ukernel_3x3s2p1__scalar_3x1() local
|
| D | 3x3p1-minmax-scalar-5x1.c | 103 const float vi4x2 = *i4++; in xnn_f32_dwconv2d_chw_ukernel_3x3p1__scalar_5x1() local
|
| D | 5x5p2-minmax-scalar-1x1.c | 94 float vi4x2 = *i4++; in xnn_f32_dwconv2d_chw_ukernel_5x5p2__scalar_1x1() local
|
| D | 5x5p2-minmax-scalar-1x1-acc2.c | 94 float vi4x2 = *i4++; in xnn_f32_dwconv2d_chw_ukernel_5x5p2__scalar_1x1_acc2() local
|
| D | 5x5s2p2-minmax-scalar-1x1.c | 102 float vi4x2 = *i4++; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__scalar_1x1() local
|
| D | 5x5p2-minmax-scalar-1x1-acc3.c | 94 float vi4x2 = *i4++; in xnn_f32_dwconv2d_chw_ukernel_5x5p2__scalar_1x1_acc3() local
|
| D | 3x3p1-minmax-scalar-6x1.c | 111 const float vi4x2 = *i4++; in xnn_f32_dwconv2d_chw_ukernel_3x3p1__scalar_6x1() local
|
| D | 5x5p2-minmax-scalar-1x1-acc4.c | 94 float vi4x2 = *i4++; in xnn_f32_dwconv2d_chw_ukernel_5x5p2__scalar_1x1_acc4() local
|
| D | 5x5s2p2-minmax-scalar-1x1-acc2.c | 102 float vi4x2 = *i4++; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__scalar_1x1_acc2() local
|
| D | 5x5s2p2-minmax-scalar-1x1-acc4.c | 102 float vi4x2 = *i4++; in xnn_f32_dwconv2d_chw_ukernel_5x5s2p2__scalar_1x1_acc4() local
|
| D | 5x5p2-minmax-scalar-2x1-acc2.c | 102 float vi4x2 = *i4++; in xnn_f32_dwconv2d_chw_ukernel_5x5p2__scalar_2x1_acc2() local
|
| D | 5x5p2-minmax-scalar-2x1.c | 102 float vi4x2 = *i4++; in xnn_f32_dwconv2d_chw_ukernel_5x5p2__scalar_2x1() local
|
| D | 5x5p2-minmax-scalar-1x1-acc5.c | 94 float vi4x2 = *i4++; in xnn_f32_dwconv2d_chw_ukernel_5x5p2__scalar_1x1_acc5() local
|
| /external/XNNPACK/src/f32-conv-hwc/gen/ |
| D | 3x3s2p1c3x4-neonfma-2x2.c | 136 const float32x4_t vi4x2 = vld1q_f32(i4); i4 += 4; in xnn_f32_conv_hwc_ukernel_3x3s2p1c3x4__neonfma_2x2() local
516 const float32x2_t vi4x2 = vld1_f32(i4); i4 += 2; in xnn_f32_conv_hwc_ukernel_3x3s2p1c3x4__neonfma_2x2() local
|
| D | 3x3s2p1c3x4-neon-2x2.c | 134 const float32x4_t vi4x2 = vld1q_f32(i4); i4 += 4; in xnn_f32_conv_hwc_ukernel_3x3s2p1c3x4__neon_2x2() local
516 const float32x2_t vi4x2 = vld1_f32(i4); i4 += 2; in xnn_f32_conv_hwc_ukernel_3x3s2p1c3x4__neon_2x2() local
|
| D | 3x3s2p0p1c3x4-neon-2x2.c | 158 const float32x4_t vi4x2 = vld1q_f32(i4); i4 += 4; in xnn_f32_conv_hwc_ukernel_3x3s2p0p1c3x4__neon_2x2() local
533 const float32x2_t vi4x2 = vld1_f32(i4); i4 += 2; in xnn_f32_conv_hwc_ukernel_3x3s2p0p1c3x4__neon_2x2() local
|
| D | 3x3s2p0p1c3x4-neonfma-2x2.c | 160 const float32x4_t vi4x2 = vld1q_f32(i4); i4 += 4; in xnn_f32_conv_hwc_ukernel_3x3s2p0p1c3x4__neonfma_2x2() local
533 const float32x2_t vi4x2 = vld1_f32(i4); i4 += 2; in xnn_f32_conv_hwc_ukernel_3x3s2p0p1c3x4__neonfma_2x2() local
|