| /external/XNNPACK/src/x8-transposec/gen/ |
| D | 8x8-reuse-dec-zip-neon.c | 51 const uint8x8_t v3_5 = vld1_u8(i0); i0 = (uint8_t*) ((uintptr_t) i0 + input_stride); in xnn_x8_transposec_ukernel__8x8_reuse_dec_zip_neon() local 128 const uint8x8_t v3_5 = vld1_u8(i5); in xnn_x8_transposec_ukernel__8x8_reuse_dec_zip_neon() local
|
| D | 8x8-multi-switch-zip-neon.c | 58 const uint8x8_t v3_5 = vld1_u8(i5); i5 = (uint8_t*) ((uintptr_t) i5 + input_offset); in xnn_x8_transposec_ukernel__8x8_multi_switch_zip_neon() local 121 const uint8x8_t v3_5 = vld1_u8(i5); in xnn_x8_transposec_ukernel__8x8_multi_switch_zip_neon() local
|
| D | 8x8-reuse-multi-zip-neon.c | 75 const uint8x8_t v3_5 = vld1_u8(i0); i0 = (uint8_t*) ((uintptr_t) i0 + input_stride); in xnn_x8_transposec_ukernel__8x8_reuse_multi_zip_neon() local 129 const uint8x8_t v3_5 = vld1_u8(i5); in xnn_x8_transposec_ukernel__8x8_reuse_multi_zip_neon() local
|
| D | 8x8-multi-dec-zip-neon.c | 59 const uint8x8_t v3_5 = vld1_u8(i5); i5 = (uint8_t*) ((uintptr_t) i5 + input_offset); in xnn_x8_transposec_ukernel__8x8_multi_dec_zip_neon() local 131 const uint8x8_t v3_5 = vld1_u8(i5); in xnn_x8_transposec_ukernel__8x8_multi_dec_zip_neon() local
|
| D | 8x8-reuse-mov-zip-neon.c | 51 const uint8x8_t v3_5 = vld1_u8(i0); i0 = (uint8_t*) ((uintptr_t) i0 + input_stride); in xnn_x8_transposec_ukernel__8x8_reuse_mov_zip_neon() local 135 const uint8x8_t v3_5 = vld1_u8(i5); in xnn_x8_transposec_ukernel__8x8_reuse_mov_zip_neon() local
|
| D | 8x8-reuse-switch-zip-neon.c | 50 const uint8x8_t v3_5 = vld1_u8(i0); i0 = (uint8_t*) ((uintptr_t) i0 + input_stride); in xnn_x8_transposec_ukernel__8x8_reuse_switch_zip_neon() local 118 const uint8x8_t v3_5 = vld1_u8(i5); in xnn_x8_transposec_ukernel__8x8_reuse_switch_zip_neon() local
|
| D | 8x8-multi-mov-zip-neon.c | 59 const uint8x8_t v3_5 = vld1_u8(i5); i5 = (uint8_t*) ((uintptr_t) i5 + input_offset); in xnn_x8_transposec_ukernel__8x8_multi_mov_zip_neon() local 138 const uint8x8_t v3_5 = vld1_u8(i5); in xnn_x8_transposec_ukernel__8x8_multi_mov_zip_neon() local
|
| D | 16x16-reuse-switch-zip-neon.c | 67 const uint8x16x2_t v3_5 = vzipq_u8(v4_5, v4_13); in xnn_x8_transposec_ukernel__16x16_reuse_switch_zip_neon() local 215 const uint8x16x2_t v3_5 = vzipq_u8(v4_5, v4_13); in xnn_x8_transposec_ukernel__16x16_reuse_switch_zip_neon() local
|
| /external/XNNPACK/src/x16-transposec/gen/ |
| D | 8x8-reuse-dec-zip-neon.c | 51 const uint16x8_t v3_5 = vld1q_u16(i0); i0 = (uint16_t*) ((uintptr_t) i0 + input_stride); in xnn_x16_transposec_ukernel__8x8_reuse_dec_zip_neon() local 128 const uint16x8_t v3_5 = vld1q_u16(i5); in xnn_x16_transposec_ukernel__8x8_reuse_dec_zip_neon() local
|
| D | 8x8-multi-switch-zip-neon.c | 58 const uint16x8_t v3_5 = vld1q_u16(i5); i5 = (uint16_t*) ((uintptr_t) i5 + input_offset); in xnn_x16_transposec_ukernel__8x8_multi_switch_zip_neon() local 121 const uint16x8_t v3_5 = vld1q_u16(i5); in xnn_x16_transposec_ukernel__8x8_multi_switch_zip_neon() local
|
| D | 8x8-reuse-multi-zip-neon.c | 75 const uint16x8_t v3_5 = vld1q_u16(i0); i0 = (uint16_t*) ((uintptr_t) i0 + input_stride); in xnn_x16_transposec_ukernel__8x8_reuse_multi_zip_neon() local 129 const uint16x8_t v3_5 = vld1q_u16(i5); in xnn_x16_transposec_ukernel__8x8_reuse_multi_zip_neon() local
|
| D | 8x8-multi-dec-zip-neon.c | 59 const uint16x8_t v3_5 = vld1q_u16(i5); i5 = (uint16_t*) ((uintptr_t) i5 + input_offset); in xnn_x16_transposec_ukernel__8x8_multi_dec_zip_neon() local 131 const uint16x8_t v3_5 = vld1q_u16(i5); in xnn_x16_transposec_ukernel__8x8_multi_dec_zip_neon() local
|
| D | 8x8-reuse-switch-zip-neon.c | 50 const uint16x8_t v3_5 = vld1q_u16(i0); i0 = (uint16_t*) ((uintptr_t) i0 + input_stride); in xnn_x16_transposec_ukernel__8x8_reuse_switch_zip_neon() local 118 const uint16x8_t v3_5 = vld1q_u16(i5); in xnn_x16_transposec_ukernel__8x8_reuse_switch_zip_neon() local
|
| D | 8x8-reuse-mov-zip-neon.c | 51 const uint16x8_t v3_5 = vld1q_u16(i0); i0 = (uint16_t*) ((uintptr_t) i0 + input_stride); in xnn_x16_transposec_ukernel__8x8_reuse_mov_zip_neon() local 135 const uint16x8_t v3_5 = vld1q_u16(i5); in xnn_x16_transposec_ukernel__8x8_reuse_mov_zip_neon() local
|
| D | 8x8-multi-mov-zip-neon.c | 59 const uint16x8_t v3_5 = vld1q_u16(i5); i5 = (uint16_t*) ((uintptr_t) i5 + input_offset); in xnn_x16_transposec_ukernel__8x8_multi_mov_zip_neon() local 138 const uint16x8_t v3_5 = vld1q_u16(i5); in xnn_x16_transposec_ukernel__8x8_multi_mov_zip_neon() local
|
| D | 8x8-reuse-multi-wasmsimd.c | 80 const v128_t v3_5 = wasm_v128_load(i0); in xnn_x16_transposec_ukernel__8x8_reuse_multi_wasmsimd() local 156 const v128_t v3_5 = wasm_v128_load(i5); in xnn_x16_transposec_ukernel__8x8_reuse_multi_wasmsimd() local
|
| D | 8x8-reuse-switch-sse2.c | 57 const __m128i v3_5 = _mm_loadu_si128((const __m128i*) i0); in xnn_x16_transposec_ukernel__8x8_reuse_switch_sse2() local 148 const __m128i v3_5 = _mm_loadu_si128((const __m128i*) i5); in xnn_x16_transposec_ukernel__8x8_reuse_switch_sse2() local
|
| D | 8x8-multi-switch-sse2.c | 65 const __m128i v3_5 = _mm_loadu_si128((const __m128i*) i5); in xnn_x16_transposec_ukernel__8x8_multi_switch_sse2() local 151 const __m128i v3_5 = _mm_loadu_si128((const __m128i*) i5); in xnn_x16_transposec_ukernel__8x8_multi_switch_sse2() local
|
| D | 8x8-reuse-mov-sse2.c | 58 const __m128i v3_5 = _mm_loadu_si128((const __m128i*) i0); in xnn_x16_transposec_ukernel__8x8_reuse_mov_sse2() local 158 const __m128i v3_5 = _mm_loadu_si128((const __m128i*) i5); in xnn_x16_transposec_ukernel__8x8_reuse_mov_sse2() local
|
| D | 8x8-reuse-switch-wasmsimd.c | 55 const v128_t v3_5 = wasm_v128_load(i0); in xnn_x16_transposec_ukernel__8x8_reuse_switch_wasmsimd() local 144 const v128_t v3_5 = wasm_v128_load(i5); in xnn_x16_transposec_ukernel__8x8_reuse_switch_wasmsimd() local
|
| D | 8x8-reuse-multi-sse2.c | 82 const __m128i v3_5 = _mm_loadu_si128((const __m128i*) i0); in xnn_x16_transposec_ukernel__8x8_reuse_multi_sse2() local 160 const __m128i v3_5 = _mm_loadu_si128((const __m128i*) i5); in xnn_x16_transposec_ukernel__8x8_reuse_multi_sse2() local
|
| D | 8x8-multi-switch-wasmsimd.c | 63 const v128_t v3_5 = wasm_v128_load(i5); in xnn_x16_transposec_ukernel__8x8_multi_switch_wasmsimd() local 147 const v128_t v3_5 = wasm_v128_load(i5); in xnn_x16_transposec_ukernel__8x8_multi_switch_wasmsimd() local
|
| D | 8x8-reuse-mov-wasmsimd.c | 56 const v128_t v3_5 = wasm_v128_load(i0); in xnn_x16_transposec_ukernel__8x8_reuse_mov_wasmsimd() local 154 const v128_t v3_5 = wasm_v128_load(i5); in xnn_x16_transposec_ukernel__8x8_reuse_mov_wasmsimd() local
|
| D | 8x8-multi-mov-sse2.c | 66 const __m128i v3_5 = _mm_loadu_si128((const __m128i*) i5); in xnn_x16_transposec_ukernel__8x8_multi_mov_sse2() local 161 const __m128i v3_5 = _mm_loadu_si128((const __m128i*) i5); in xnn_x16_transposec_ukernel__8x8_multi_mov_sse2() local
|
| D | 8x8-multi-mov-wasmsimd.c | 64 const v128_t v3_5 = wasm_v128_load(i5); in xnn_x16_transposec_ukernel__8x8_multi_mov_wasmsimd() local 157 const v128_t v3_5 = wasm_v128_load(i5); in xnn_x16_transposec_ukernel__8x8_multi_mov_wasmsimd() local
|