| /external/XNNPACK/src/x32-transposec/gen/ |
| D | 4x4-reuse-multi-sse2.c | 68 const __m128i v1_3 = _mm_unpackhi_epi32(v2_2, v2_3); in xnn_x32_transposec_ukernel__4x4_reuse_multi_sse2() local
103 const __m128i v1_3 = _mm_unpackhi_epi32(v2_2, v2_3); in xnn_x32_transposec_ukernel__4x4_reuse_multi_sse2() local
|
| D | 4x4-reuse-multi-wasmsimd.c | 66 const v128_t v1_3 = wasm_v32x4_shuffle(v2_1, v2_3, 2, 6, 3, 7); in xnn_x32_transposec_ukernel__4x4_reuse_multi_wasmsimd() local
99 const v128_t v1_3 = wasm_v32x4_shuffle(v2_1, v2_3, 2, 6, 3, 7); in xnn_x32_transposec_ukernel__4x4_reuse_multi_wasmsimd() local
|
| D | 4x4-multi-multi-sse2.c | 72 const __m128i v1_3 = _mm_unpackhi_epi32(v2_2, v2_3); in xnn_x32_transposec_ukernel__4x4_multi_multi_sse2() local
105 const __m128i v1_3 = _mm_unpackhi_epi32(v2_2, v2_3); in xnn_x32_transposec_ukernel__4x4_multi_multi_sse2() local
|
| D | 4x4-multi-multi-wasmsimd.c | 70 const v128_t v1_3 = wasm_v32x4_shuffle(v2_1, v2_3, 2, 6, 3, 7); in xnn_x32_transposec_ukernel__4x4_multi_multi_wasmsimd() local
101 const v128_t v1_3 = wasm_v32x4_shuffle(v2_1, v2_3, 2, 6, 3, 7); in xnn_x32_transposec_ukernel__4x4_multi_multi_wasmsimd() local
|
| D | 4x4-reuse-mov-wasmsimd.c | 58 const v128_t v1_3 = wasm_v32x4_shuffle(v2_1, v2_3, 2, 6, 3, 7); in xnn_x32_transposec_ukernel__4x4_reuse_mov_wasmsimd() local
101 const v128_t v1_3 = wasm_v32x4_shuffle(v2_1, v2_3, 2, 6, 3, 7); in xnn_x32_transposec_ukernel__4x4_reuse_mov_wasmsimd() local
|
| D | 4x4-reuse-switch-wasmsimd.c | 57 const v128_t v1_3 = wasm_v32x4_shuffle(v2_1, v2_3, 2, 6, 3, 7); in xnn_x32_transposec_ukernel__4x4_reuse_switch_wasmsimd() local
99 const v128_t v1_3 = wasm_v32x4_shuffle(v2_1, v2_3, 2, 6, 3, 7); in xnn_x32_transposec_ukernel__4x4_reuse_switch_wasmsimd() local
|
| D | 4x4-reuse-mov-sse2.c | 60 const __m128i v1_3 = _mm_unpackhi_epi32(v2_2, v2_3); in xnn_x32_transposec_ukernel__4x4_reuse_mov_sse2() local
105 const __m128i v1_3 = _mm_unpackhi_epi32(v2_2, v2_3); in xnn_x32_transposec_ukernel__4x4_reuse_mov_sse2() local
|
| D | 4x4-reuse-switch-sse2.c | 59 const __m128i v1_3 = _mm_unpackhi_epi32(v2_2, v2_3); in xnn_x32_transposec_ukernel__4x4_reuse_switch_sse2() local
103 const __m128i v1_3 = _mm_unpackhi_epi32(v2_2, v2_3); in xnn_x32_transposec_ukernel__4x4_reuse_switch_sse2() local
|
| D | 4x4-multi-switch-wasmsimd.c | 61 const v128_t v1_3 = wasm_v32x4_shuffle(v2_1, v2_3, 2, 6, 3, 7); in xnn_x32_transposec_ukernel__4x4_multi_switch_wasmsimd() local
101 const v128_t v1_3 = wasm_v32x4_shuffle(v2_1, v2_3, 2, 6, 3, 7); in xnn_x32_transposec_ukernel__4x4_multi_switch_wasmsimd() local
|
| D | 4x4-multi-mov-sse2.c | 64 const __m128i v1_3 = _mm_unpackhi_epi32(v2_2, v2_3); in xnn_x32_transposec_ukernel__4x4_multi_mov_sse2() local
107 const __m128i v1_3 = _mm_unpackhi_epi32(v2_2, v2_3); in xnn_x32_transposec_ukernel__4x4_multi_mov_sse2() local
|
| D | 4x4-multi-mov-wasmsimd.c | 62 const v128_t v1_3 = wasm_v32x4_shuffle(v2_1, v2_3, 2, 6, 3, 7); in xnn_x32_transposec_ukernel__4x4_multi_mov_wasmsimd() local
103 const v128_t v1_3 = wasm_v32x4_shuffle(v2_1, v2_3, 2, 6, 3, 7); in xnn_x32_transposec_ukernel__4x4_multi_mov_wasmsimd() local
|
| D | 4x4-multi-switch-sse2.c | 63 const __m128i v1_3 = _mm_unpackhi_epi32(v2_2, v2_3); in xnn_x32_transposec_ukernel__4x4_multi_switch_sse2() local
105 const __m128i v1_3 = _mm_unpackhi_epi32(v2_2, v2_3); in xnn_x32_transposec_ukernel__4x4_multi_switch_sse2() local
|
| /external/XNNPACK/src/x24-transposec/ |
| D | 4x4-ssse3.c | 83 const __m128i v1_3 = _mm_unpackhi_epi8(v2, v3); in xnn_x24_transposec_ukernel__4x4_ssse3() local
131 const __m128i v1_3 = _mm_unpackhi_epi8(v2, v2); in xnn_x24_transposec_ukernel__4x4_ssse3() local
|
| /external/XNNPACK/src/x8-transposec/gen/ |
| D | 8x8-reuse-dec-zip-neon.c | 63 const uint8x8x2_t v1_3 = vzip_u8(v2_1.val[1], v2_3.val[1]); in xnn_x8_transposec_ukernel__8x8_reuse_dec_zip_neon() local
144 const uint8x8x2_t v1_3 = vzip_u8(v2_1.val[1], v2_3.val[1]); in xnn_x8_transposec_ukernel__8x8_reuse_dec_zip_neon() local
|
| D | 8x8-multi-switch-zip-neon.c | 70 const uint8x8x2_t v1_3 = vzip_u8(v2_1.val[1], v2_3.val[1]); in xnn_x8_transposec_ukernel__8x8_multi_switch_zip_neon() local
136 const uint8x8x2_t v1_3 = vzip_u8(v2_1.val[1], v2_3.val[1]); in xnn_x8_transposec_ukernel__8x8_multi_switch_zip_neon() local
|
| D | 8x8-reuse-multi-zip-neon.c | 87 const uint8x8x2_t v1_3 = vzip_u8(v2_1.val[1], v2_3.val[1]); in xnn_x8_transposec_ukernel__8x8_reuse_multi_zip_neon() local
145 const uint8x8x2_t v1_3 = vzip_u8(v2_1.val[1], v2_3.val[1]); in xnn_x8_transposec_ukernel__8x8_reuse_multi_zip_neon() local
|
| D | 8x8-multi-dec-zip-neon.c | 71 const uint8x8x2_t v1_3 = vzip_u8(v2_1.val[1], v2_3.val[1]); in xnn_x8_transposec_ukernel__8x8_multi_dec_zip_neon() local
146 const uint8x8x2_t v1_3 = vzip_u8(v2_1.val[1], v2_3.val[1]); in xnn_x8_transposec_ukernel__8x8_multi_dec_zip_neon() local
|
| D | 8x8-reuse-mov-zip-neon.c | 63 const uint8x8x2_t v1_3 = vzip_u8(v2_1.val[1], v2_3.val[1]); in xnn_x8_transposec_ukernel__8x8_reuse_mov_zip_neon() local
151 const uint8x8x2_t v1_3 = vzip_u8(v2_1.val[1], v2_3.val[1]); in xnn_x8_transposec_ukernel__8x8_reuse_mov_zip_neon() local
|
| D | 8x8-reuse-switch-zip-neon.c | 62 const uint8x8x2_t v1_3 = vzip_u8(v2_1.val[1], v2_3.val[1]); in xnn_x8_transposec_ukernel__8x8_reuse_switch_zip_neon() local
134 const uint8x8x2_t v1_3 = vzip_u8(v2_1.val[1], v2_3.val[1]); in xnn_x8_transposec_ukernel__8x8_reuse_switch_zip_neon() local
|
| /external/XNNPACK/src/x16-transposec/gen/ |
| D | 8x8-reuse-dec-zip-neon.c | 63 const uint16x8x2_t v1_3 = vzipq_u16(v2_1.val[1], v2_3.val[1]); in xnn_x16_transposec_ukernel__8x8_reuse_dec_zip_neon() local
144 const uint16x8x2_t v1_3 = vzipq_u16(v2_1.val[1], v2_3.val[1]); in xnn_x16_transposec_ukernel__8x8_reuse_dec_zip_neon() local
|
| D | 8x8-multi-switch-zip-neon.c | 70 const uint16x8x2_t v1_3 = vzipq_u16(v2_1.val[1], v2_3.val[1]); in xnn_x16_transposec_ukernel__8x8_multi_switch_zip_neon() local
136 const uint16x8x2_t v1_3 = vzipq_u16(v2_1.val[1], v2_3.val[1]); in xnn_x16_transposec_ukernel__8x8_multi_switch_zip_neon() local
|
| D | 8x8-reuse-multi-zip-neon.c | 87 const uint16x8x2_t v1_3 = vzipq_u16(v2_1.val[1], v2_3.val[1]); in xnn_x16_transposec_ukernel__8x8_reuse_multi_zip_neon() local
145 const uint16x8x2_t v1_3 = vzipq_u16(v2_1.val[1], v2_3.val[1]); in xnn_x16_transposec_ukernel__8x8_reuse_multi_zip_neon() local
|
| D | 8x8-multi-dec-zip-neon.c | 71 const uint16x8x2_t v1_3 = vzipq_u16(v2_1.val[1], v2_3.val[1]); in xnn_x16_transposec_ukernel__8x8_multi_dec_zip_neon() local
146 const uint16x8x2_t v1_3 = vzipq_u16(v2_1.val[1], v2_3.val[1]); in xnn_x16_transposec_ukernel__8x8_multi_dec_zip_neon() local
|
| D | 8x8-reuse-switch-zip-neon.c | 62 const uint16x8x2_t v1_3 = vzipq_u16(v2_1.val[1], v2_3.val[1]); in xnn_x16_transposec_ukernel__8x8_reuse_switch_zip_neon() local
134 const uint16x8x2_t v1_3 = vzipq_u16(v2_1.val[1], v2_3.val[1]); in xnn_x16_transposec_ukernel__8x8_reuse_switch_zip_neon() local
|
| D | 8x8-reuse-mov-zip-neon.c | 63 const uint16x8x2_t v1_3 = vzipq_u16(v2_1.val[1], v2_3.val[1]); in xnn_x16_transposec_ukernel__8x8_reuse_mov_zip_neon() local
151 const uint16x8x2_t v1_3 = vzipq_u16(v2_1.val[1], v2_3.val[1]); in xnn_x16_transposec_ukernel__8x8_reuse_mov_zip_neon() local
|