| /aosp_15_r20/external/XNNPACK/src/bf16-gemm/gen/ |
| H A D | 1x4c8-minmax-neonbf16-bfmlal.c | 56 const bfloat16x8_t vb3 = vld1q_bf16(w); w += 8; in xnn_bf16_gemm_minmax_ukernel_1x4c8__neonbf16_bfmlal() local
74 const bfloat16x8_t vb3 = vld1q_bf16(w); w += 8; in xnn_bf16_gemm_minmax_ukernel_1x4c8__neonbf16_bfmlal() local
|
| H A D | 1x4c8-minmax-neonbf16-bfdot.c | 56 const bfloat16x8_t vb3 = vld1q_bf16(w); w += 8; in xnn_bf16_gemm_minmax_ukernel_1x4c8__neonbf16_bfdot() local
69 const bfloat16x8_t vb3 = vld1q_bf16(w); w += 8; in xnn_bf16_gemm_minmax_ukernel_1x4c8__neonbf16_bfdot() local
|
| H A D | 2x4c8-minmax-neonbf16-bfmlal.c | 67 const bfloat16x8_t vb3 = vld1q_bf16(w); w += 8; in xnn_bf16_gemm_minmax_ukernel_2x4c8__neonbf16_bfmlal() local
94 const bfloat16x8_t vb3 = vld1q_bf16(w); w += 8; in xnn_bf16_gemm_minmax_ukernel_2x4c8__neonbf16_bfmlal() local
|
| H A D | 2x4c8-minmax-neonbf16-bfdot.c | 67 const bfloat16x8_t vb3 = vld1q_bf16(w); w += 8; in xnn_bf16_gemm_minmax_ukernel_2x4c8__neonbf16_bfdot() local
85 const bfloat16x8_t vb3 = vld1q_bf16(w); w += 8; in xnn_bf16_gemm_minmax_ukernel_2x4c8__neonbf16_bfdot() local
|
| H A D | 1x4c8-minmax-neonfma-shland.c | 57 const uint16x8_t vb3 = vld1q_u16(w); w += 8; in xnn_bf16_gemm_minmax_ukernel_1x4c8__neonfma_shland() local
89 const uint16x8_t vb3 = vld1q_u16(w); w += 8; in xnn_bf16_gemm_minmax_ukernel_1x4c8__neonfma_shland() local
|
| H A D | 1x4c8-minmax-neonfma-zip.c | 57 const uint16x8_t vb3 = vld1q_u16(w); w += 8; in xnn_bf16_gemm_minmax_ukernel_1x4c8__neonfma_zip() local
89 const uint16x8_t vb3 = vld1q_u16(w); w += 8; in xnn_bf16_gemm_minmax_ukernel_1x4c8__neonfma_zip() local
|
| H A D | 3x4c8-minmax-neonbf16-bfdot.c | 78 const bfloat16x8_t vb3 = vld1q_bf16(w); w += 8; in xnn_bf16_gemm_minmax_ukernel_3x4c8__neonbf16_bfdot() local
101 const bfloat16x8_t vb3 = vld1q_bf16(w); w += 8; in xnn_bf16_gemm_minmax_ukernel_3x4c8__neonbf16_bfdot() local
|
| H A D | 3x4c8-minmax-neonbf16-bfmlal.c | 78 const bfloat16x8_t vb3 = vld1q_bf16(w); w += 8; in xnn_bf16_gemm_minmax_ukernel_3x4c8__neonbf16_bfmlal() local
114 const bfloat16x8_t vb3 = vld1q_bf16(w); w += 8; in xnn_bf16_gemm_minmax_ukernel_3x4c8__neonbf16_bfmlal() local
|
| H A D | 4x4c8-minmax-neonbf16-bfdot.c | 89 const bfloat16x8_t vb3 = vld1q_bf16(w); w += 8; in xnn_bf16_gemm_minmax_ukernel_4x4c8__neonbf16_bfdot() local
117 const bfloat16x8_t vb3 = vld1q_bf16(w); w += 8; in xnn_bf16_gemm_minmax_ukernel_4x4c8__neonbf16_bfdot() local
|
| H A D | 2x4c8-minmax-neonfma-zip.c | 68 const uint16x8_t vb3 = vld1q_u16(w); w += 8; in xnn_bf16_gemm_minmax_ukernel_2x4c8__neonfma_zip() local
111 const uint16x8_t vb3 = vld1q_u16(w); w += 8; in xnn_bf16_gemm_minmax_ukernel_2x4c8__neonfma_zip() local
|
| H A D | 2x4c8-minmax-neonfma-shland.c | 68 const uint16x8_t vb3 = vld1q_u16(w); w += 8; in xnn_bf16_gemm_minmax_ukernel_2x4c8__neonfma_shland() local
111 const uint16x8_t vb3 = vld1q_u16(w); w += 8; in xnn_bf16_gemm_minmax_ukernel_2x4c8__neonfma_shland() local
|
| H A D | 4x4c8-minmax-neonbf16-bfmlal.c | 89 const bfloat16x8_t vb3 = vld1q_bf16(w); w += 8; in xnn_bf16_gemm_minmax_ukernel_4x4c8__neonbf16_bfmlal() local
134 const bfloat16x8_t vb3 = vld1q_bf16(w); w += 8; in xnn_bf16_gemm_minmax_ukernel_4x4c8__neonbf16_bfmlal() local
|
| H A D | 5x4c8-minmax-neonbf16-bfdot.c | 100 const bfloat16x8_t vb3 = vld1q_bf16(w); w += 8; in xnn_bf16_gemm_minmax_ukernel_5x4c8__neonbf16_bfdot() local
133 const bfloat16x8_t vb3 = vld1q_bf16(w); w += 8; in xnn_bf16_gemm_minmax_ukernel_5x4c8__neonbf16_bfdot() local
|
| H A D | 5x4c8-minmax-neonbf16-bfmlal.c | 100 const bfloat16x8_t vb3 = vld1q_bf16(w); w += 8; in xnn_bf16_gemm_minmax_ukernel_5x4c8__neonbf16_bfmlal() local
154 const bfloat16x8_t vb3 = vld1q_bf16(w); w += 8; in xnn_bf16_gemm_minmax_ukernel_5x4c8__neonbf16_bfmlal() local
|
| H A D | 3x4c8-minmax-neonfma-zip.c | 79 const uint16x8_t vb3 = vld1q_u16(w); w += 8; in xnn_bf16_gemm_minmax_ukernel_3x4c8__neonfma_zip() local
133 const uint16x8_t vb3 = vld1q_u16(w); w += 8; in xnn_bf16_gemm_minmax_ukernel_3x4c8__neonfma_zip() local
|
| H A D | 3x4c8-minmax-neonfma-shland.c | 79 const uint16x8_t vb3 = vld1q_u16(w); w += 8; in xnn_bf16_gemm_minmax_ukernel_3x4c8__neonfma_shland() local
133 const uint16x8_t vb3 = vld1q_u16(w); w += 8; in xnn_bf16_gemm_minmax_ukernel_3x4c8__neonfma_shland() local
|
| H A D | 4x4c8-minmax-neonfma-shland.c | 90 const uint16x8_t vb3 = vld1q_u16(w); w += 8; in xnn_bf16_gemm_minmax_ukernel_4x4c8__neonfma_shland() local
155 const uint16x8_t vb3 = vld1q_u16(w); w += 8; in xnn_bf16_gemm_minmax_ukernel_4x4c8__neonfma_shland() local
|
| H A D | 4x4c8-minmax-neonfma-zip.c | 90 const uint16x8_t vb3 = vld1q_u16(w); w += 8; in xnn_bf16_gemm_minmax_ukernel_4x4c8__neonfma_zip() local
155 const uint16x8_t vb3 = vld1q_u16(w); w += 8; in xnn_bf16_gemm_minmax_ukernel_4x4c8__neonfma_zip() local
|
| H A D | 5x4c8-minmax-neonfma-zip.c | 101 const uint16x8_t vb3 = vld1q_u16(w); w += 8; in xnn_bf16_gemm_minmax_ukernel_5x4c8__neonfma_zip() local
177 const uint16x8_t vb3 = vld1q_u16(w); w += 8; in xnn_bf16_gemm_minmax_ukernel_5x4c8__neonfma_zip() local
|
| H A D | 5x4c8-minmax-neonfma-shland.c | 101 const uint16x8_t vb3 = vld1q_u16(w); w += 8; in xnn_bf16_gemm_minmax_ukernel_5x4c8__neonfma_shland() local
177 const uint16x8_t vb3 = vld1q_u16(w); w += 8; in xnn_bf16_gemm_minmax_ukernel_5x4c8__neonfma_shland() local
|
| /aosp_15_r20/external/pytorch/aten/src/ATen/native/quantized/cpu/qnnpack/src/q8conv/ |
| H A D | 4x4c2-sse2.c | 138 const __m128i vb3 = _mm_loadl_epi64((const __m128i*)((uintptr_t)w + 24)); in pytorch_q8conv_ukernel_4x4c2__sse2() local
250 const __m128i vb3 = _mm_loadl_epi64((const __m128i*)w); in pytorch_q8conv_ukernel_4x4c2__sse2() local
|
| /aosp_15_r20/external/pytorch/aten/src/ATen/native/quantized/cpu/qnnpack/src/q8gemm/ |
| H A D | 4x4c2-dq-sse2.c | 146 const __m128i vb3 = _mm_loadl_epi64((const __m128i*)((uintptr_t)w + 24)); in pytorch_q8gemm_dq_ukernel_4x4c2__sse2() local
248 const __m128i vb3 = in pytorch_q8gemm_dq_ukernel_4x4c2__sse2() local
|
| H A D | 4x4c2-sse2.c | 140 const __m128i vb3 = _mm_loadl_epi64((const __m128i*)((uintptr_t)w + 24)); in pytorch_q8gemm_ukernel_4x4c2__sse2() local
242 const __m128i vb3 = in pytorch_q8gemm_ukernel_4x4c2__sse2() local
|
| H A D | 2x4c8-sse2.c | 122 const __m128i vb3 = _mm_loadl_epi64((const __m128i*)b3); in pytorch_q8gemm_ukernel_2x4c8__sse2() local
162 const __m128i vb3 = _mm_loadl_epi64((const __m128i*)b3); in pytorch_q8gemm_ukernel_2x4c8__sse2() local
|
| /aosp_15_r20/external/XNNPACK/src/f32-vbinary/gen/ |
| H A D | vadd-scalar-x4.c | 41 const float vb3 = b[3]; in xnn_f32_vadd_ukernel__scalar_x4() local
|