| /external/XNNPACK/src/x32-transposec/gen/ |
| D | 4x4-reuse-multi-sse2.c | 67 const __m128i v1_2 = _mm_unpacklo_epi32(v2_2, v2_3); in xnn_x32_transposec_ukernel__4x4_reuse_multi_sse2() local
102 const __m128i v1_2 = _mm_unpacklo_epi32(v2_2, v2_3); in xnn_x32_transposec_ukernel__4x4_reuse_multi_sse2() local
|
| D | 4x4-reuse-multi-wasmsimd.c | 65 const v128_t v1_2 = wasm_v32x4_shuffle(v2_1, v2_3, 0, 4, 1, 5); in xnn_x32_transposec_ukernel__4x4_reuse_multi_wasmsimd() local
98 const v128_t v1_2 = wasm_v32x4_shuffle(v2_1, v2_3, 0, 4, 1, 5); in xnn_x32_transposec_ukernel__4x4_reuse_multi_wasmsimd() local
|
| D | 4x4-multi-multi-sse2.c | 71 const __m128i v1_2 = _mm_unpacklo_epi32(v2_2, v2_3); in xnn_x32_transposec_ukernel__4x4_multi_multi_sse2() local
104 const __m128i v1_2 = _mm_unpacklo_epi32(v2_2, v2_3); in xnn_x32_transposec_ukernel__4x4_multi_multi_sse2() local
|
| D | 4x4-multi-multi-wasmsimd.c | 69 const v128_t v1_2 = wasm_v32x4_shuffle(v2_1, v2_3, 0, 4, 1, 5); in xnn_x32_transposec_ukernel__4x4_multi_multi_wasmsimd() local
100 const v128_t v1_2 = wasm_v32x4_shuffle(v2_1, v2_3, 0, 4, 1, 5); in xnn_x32_transposec_ukernel__4x4_multi_multi_wasmsimd() local
|
| D | 4x4-reuse-mov-wasmsimd.c | 57 const v128_t v1_2 = wasm_v32x4_shuffle(v2_1, v2_3, 0, 4, 1, 5); in xnn_x32_transposec_ukernel__4x4_reuse_mov_wasmsimd() local
100 const v128_t v1_2 = wasm_v32x4_shuffle(v2_1, v2_3, 0, 4, 1, 5); in xnn_x32_transposec_ukernel__4x4_reuse_mov_wasmsimd() local
|
| D | 4x4-reuse-switch-wasmsimd.c | 56 const v128_t v1_2 = wasm_v32x4_shuffle(v2_1, v2_3, 0, 4, 1, 5); in xnn_x32_transposec_ukernel__4x4_reuse_switch_wasmsimd() local
98 const v128_t v1_2 = wasm_v32x4_shuffle(v2_1, v2_3, 0, 4, 1, 5); in xnn_x32_transposec_ukernel__4x4_reuse_switch_wasmsimd() local
|
| D | 4x4-reuse-mov-sse2.c | 59 const __m128i v1_2 = _mm_unpacklo_epi32(v2_2, v2_3); in xnn_x32_transposec_ukernel__4x4_reuse_mov_sse2() local
104 const __m128i v1_2 = _mm_unpacklo_epi32(v2_2, v2_3); in xnn_x32_transposec_ukernel__4x4_reuse_mov_sse2() local
|
| D | 4x4-reuse-switch-sse2.c | 58 const __m128i v1_2 = _mm_unpacklo_epi32(v2_2, v2_3); in xnn_x32_transposec_ukernel__4x4_reuse_switch_sse2() local
102 const __m128i v1_2 = _mm_unpacklo_epi32(v2_2, v2_3); in xnn_x32_transposec_ukernel__4x4_reuse_switch_sse2() local
|
| D | 4x4-multi-switch-wasmsimd.c | 60 const v128_t v1_2 = wasm_v32x4_shuffle(v2_1, v2_3, 0, 4, 1, 5); in xnn_x32_transposec_ukernel__4x4_multi_switch_wasmsimd() local
100 const v128_t v1_2 = wasm_v32x4_shuffle(v2_1, v2_3, 0, 4, 1, 5); in xnn_x32_transposec_ukernel__4x4_multi_switch_wasmsimd() local
|
| D | 4x4-multi-mov-sse2.c | 63 const __m128i v1_2 = _mm_unpacklo_epi32(v2_2, v2_3); in xnn_x32_transposec_ukernel__4x4_multi_mov_sse2() local
106 const __m128i v1_2 = _mm_unpacklo_epi32(v2_2, v2_3); in xnn_x32_transposec_ukernel__4x4_multi_mov_sse2() local
|
| D | 4x4-multi-mov-wasmsimd.c | 61 const v128_t v1_2 = wasm_v32x4_shuffle(v2_1, v2_3, 0, 4, 1, 5); in xnn_x32_transposec_ukernel__4x4_multi_mov_wasmsimd() local
102 const v128_t v1_2 = wasm_v32x4_shuffle(v2_1, v2_3, 0, 4, 1, 5); in xnn_x32_transposec_ukernel__4x4_multi_mov_wasmsimd() local
|
| D | 4x4-multi-switch-sse2.c | 62 const __m128i v1_2 = _mm_unpacklo_epi32(v2_2, v2_3); in xnn_x32_transposec_ukernel__4x4_multi_switch_sse2() local
104 const __m128i v1_2 = _mm_unpacklo_epi32(v2_2, v2_3); in xnn_x32_transposec_ukernel__4x4_multi_switch_sse2() local
|
| /external/XNNPACK/src/x24-transposec/ |
| D | 4x4-ssse3.c | 82 const __m128i v1_2 = _mm_unpacklo_epi8(v2, v3); in xnn_x24_transposec_ukernel__4x4_ssse3() local
130 const __m128i v1_2 = _mm_unpacklo_epi8(v2, v2); in xnn_x24_transposec_ukernel__4x4_ssse3() local
|
| /external/XNNPACK/src/x8-transposec/gen/ |
| D | 8x8-reuse-dec-zip-neon.c | 62 const uint8x8x2_t v1_2 = vzip_u8(v2_1.val[0], v2_3.val[0]); in xnn_x8_transposec_ukernel__8x8_reuse_dec_zip_neon() local
143 const uint8x8x2_t v1_2 = vzip_u8(v2_1.val[0], v2_3.val[0]); in xnn_x8_transposec_ukernel__8x8_reuse_dec_zip_neon() local
|
| D | 8x8-multi-switch-zip-neon.c | 69 const uint8x8x2_t v1_2 = vzip_u8(v2_1.val[0], v2_3.val[0]); in xnn_x8_transposec_ukernel__8x8_multi_switch_zip_neon() local
135 const uint8x8x2_t v1_2 = vzip_u8(v2_1.val[0], v2_3.val[0]); in xnn_x8_transposec_ukernel__8x8_multi_switch_zip_neon() local
|
| D | 8x8-reuse-multi-zip-neon.c | 86 const uint8x8x2_t v1_2 = vzip_u8(v2_1.val[0], v2_3.val[0]); in xnn_x8_transposec_ukernel__8x8_reuse_multi_zip_neon() local
144 const uint8x8x2_t v1_2 = vzip_u8(v2_1.val[0], v2_3.val[0]); in xnn_x8_transposec_ukernel__8x8_reuse_multi_zip_neon() local
|
| D | 8x8-multi-dec-zip-neon.c | 70 const uint8x8x2_t v1_2 = vzip_u8(v2_1.val[0], v2_3.val[0]); in xnn_x8_transposec_ukernel__8x8_multi_dec_zip_neon() local
145 const uint8x8x2_t v1_2 = vzip_u8(v2_1.val[0], v2_3.val[0]); in xnn_x8_transposec_ukernel__8x8_multi_dec_zip_neon() local
|
| D | 8x8-reuse-mov-zip-neon.c | 62 const uint8x8x2_t v1_2 = vzip_u8(v2_1.val[0], v2_3.val[0]); in xnn_x8_transposec_ukernel__8x8_reuse_mov_zip_neon() local
150 const uint8x8x2_t v1_2 = vzip_u8(v2_1.val[0], v2_3.val[0]); in xnn_x8_transposec_ukernel__8x8_reuse_mov_zip_neon() local
|
| D | 8x8-reuse-switch-zip-neon.c | 61 const uint8x8x2_t v1_2 = vzip_u8(v2_1.val[0], v2_3.val[0]); in xnn_x8_transposec_ukernel__8x8_reuse_switch_zip_neon() local
133 const uint8x8x2_t v1_2 = vzip_u8(v2_1.val[0], v2_3.val[0]); in xnn_x8_transposec_ukernel__8x8_reuse_switch_zip_neon() local
|
| /external/XNNPACK/src/x16-transposec/gen/ |
| D | 8x8-reuse-dec-zip-neon.c | 62 const uint16x8x2_t v1_2 = vzipq_u16(v2_1.val[0], v2_3.val[0]); in xnn_x16_transposec_ukernel__8x8_reuse_dec_zip_neon() local
143 const uint16x8x2_t v1_2 = vzipq_u16(v2_1.val[0], v2_3.val[0]); in xnn_x16_transposec_ukernel__8x8_reuse_dec_zip_neon() local
|
| D | 8x8-multi-switch-zip-neon.c | 69 const uint16x8x2_t v1_2 = vzipq_u16(v2_1.val[0], v2_3.val[0]); in xnn_x16_transposec_ukernel__8x8_multi_switch_zip_neon() local
135 const uint16x8x2_t v1_2 = vzipq_u16(v2_1.val[0], v2_3.val[0]); in xnn_x16_transposec_ukernel__8x8_multi_switch_zip_neon() local
|
| D | 8x8-reuse-multi-zip-neon.c | 86 const uint16x8x2_t v1_2 = vzipq_u16(v2_1.val[0], v2_3.val[0]); in xnn_x16_transposec_ukernel__8x8_reuse_multi_zip_neon() local
144 const uint16x8x2_t v1_2 = vzipq_u16(v2_1.val[0], v2_3.val[0]); in xnn_x16_transposec_ukernel__8x8_reuse_multi_zip_neon() local
|
| D | 8x8-multi-dec-zip-neon.c | 70 const uint16x8x2_t v1_2 = vzipq_u16(v2_1.val[0], v2_3.val[0]); in xnn_x16_transposec_ukernel__8x8_multi_dec_zip_neon() local
145 const uint16x8x2_t v1_2 = vzipq_u16(v2_1.val[0], v2_3.val[0]); in xnn_x16_transposec_ukernel__8x8_multi_dec_zip_neon() local
|
| D | 8x8-reuse-switch-zip-neon.c | 61 const uint16x8x2_t v1_2 = vzipq_u16(v2_1.val[0], v2_3.val[0]); in xnn_x16_transposec_ukernel__8x8_reuse_switch_zip_neon() local
133 const uint16x8x2_t v1_2 = vzipq_u16(v2_1.val[0], v2_3.val[0]); in xnn_x16_transposec_ukernel__8x8_reuse_switch_zip_neon() local
|
| D | 8x8-reuse-mov-zip-neon.c | 62 const uint16x8x2_t v1_2 = vzipq_u16(v2_1.val[0], v2_3.val[0]); in xnn_x16_transposec_ukernel__8x8_reuse_mov_zip_neon() local
150 const uint16x8x2_t v1_2 = vzipq_u16(v2_1.val[0], v2_3.val[0]); in xnn_x16_transposec_ukernel__8x8_reuse_mov_zip_neon() local
|