| /external/XNNPACK/src/x64-transposec/gen/ |
| D | 2x2-reuse-multi-sse2.c | 54 const __m128i v0_1 = _mm_unpackhi_epi64(v1_0, v1_1); in xnn_x64_transposec_ukernel__2x2_reuse_multi_sse2() local 69 __m128i v0_1 = _mm_unpackhi_epi64(v1_0, v1_1); in xnn_x64_transposec_ukernel__2x2_reuse_multi_sse2() local
|
| D | 2x2-multi-multi-sse2.c | 56 const __m128i v0_1 = _mm_unpackhi_epi64(v1_0, v1_1); in xnn_x64_transposec_ukernel__2x2_multi_multi_sse2() local 71 __m128i v0_1 = _mm_unpackhi_epi64(v1_0, v1_1); in xnn_x64_transposec_ukernel__2x2_multi_multi_sse2() local
|
| D | 2x2-reuse-switch-sse2.c | 52 const __m128i v0_1 = _mm_unpackhi_epi64(v1_0, v1_1); in xnn_x64_transposec_ukernel__2x2_reuse_switch_sse2() local 74 __m128i v0_1 = _mm_unpackhi_epi64(v1_0, v1_1); in xnn_x64_transposec_ukernel__2x2_reuse_switch_sse2() local
|
| D | 2x2-reuse-mov-sse2.c | 54 const __m128i v0_1 = _mm_unpackhi_epi64(v1_0, v1_1); in xnn_x64_transposec_ukernel__2x2_reuse_mov_sse2() local 73 __m128i v0_1 = _mm_unpackhi_epi64(v1_0, v1_1); in xnn_x64_transposec_ukernel__2x2_reuse_mov_sse2() local
|
| D | 2x2-multi-switch-sse2.c | 54 const __m128i v0_1 = _mm_unpackhi_epi64(v1_0, v1_1); in xnn_x64_transposec_ukernel__2x2_multi_switch_sse2() local 76 __m128i v0_1 = _mm_unpackhi_epi64(v1_0, v1_1); in xnn_x64_transposec_ukernel__2x2_multi_switch_sse2() local
|
| D | 2x2-multi-mov-sse2.c | 56 const __m128i v0_1 = _mm_unpackhi_epi64(v1_0, v1_1); in xnn_x64_transposec_ukernel__2x2_multi_mov_sse2() local 75 __m128i v0_1 = _mm_unpackhi_epi64(v1_0, v1_1); in xnn_x64_transposec_ukernel__2x2_multi_mov_sse2() local
|
| /external/XNNPACK/src/x16-transposec/gen/ |
| D | 4x4-reuse-multi-zip-neon.c | 63 const uint16x4x2_t v0_1 = vzip_u16(v1_0.val[1], v1_1.val[1]); in xnn_x16_transposec_ukernel__4x4_reuse_multi_zip_neon() local 89 const uint16x4x2_t v0_1 = vzip_u16(v1_0.val[1], v1_1.val[1]); in xnn_x16_transposec_ukernel__4x4_reuse_multi_zip_neon() local
|
| D | 4x4-reuse-dec-zip-neon.c | 55 const uint16x4x2_t v0_1 = vzip_u16(v1_0.val[1], v1_1.val[1]); in xnn_x16_transposec_ukernel__4x4_reuse_dec_zip_neon() local 92 const uint16x4x2_t v0_1 = vzip_u16(v1_0.val[1], v1_1.val[1]); in xnn_x16_transposec_ukernel__4x4_reuse_dec_zip_neon() local
|
| D | 4x4-multi-multi-zip-neon.c | 67 const uint16x4x2_t v0_1 = vzip_u16(v1_0.val[1], v1_1.val[1]); in xnn_x16_transposec_ukernel__4x4_multi_multi_zip_neon() local 91 const uint16x4x2_t v0_1 = vzip_u16(v1_0.val[1], v1_1.val[1]); in xnn_x16_transposec_ukernel__4x4_multi_multi_zip_neon() local
|
| D | 4x4-multi-dec-zip-neon.c | 59 const uint16x4x2_t v0_1 = vzip_u16(v1_0.val[1], v1_1.val[1]); in xnn_x16_transposec_ukernel__4x4_multi_dec_zip_neon() local 94 const uint16x4x2_t v0_1 = vzip_u16(v1_0.val[1], v1_1.val[1]); in xnn_x16_transposec_ukernel__4x4_multi_dec_zip_neon() local
|
| D | 4x4-reuse-mov-zip-neon.c | 55 const uint16x4x2_t v0_1 = vzip_u16(v1_0.val[1], v1_1.val[1]); in xnn_x16_transposec_ukernel__4x4_reuse_mov_zip_neon() local 95 const uint16x4x2_t v0_1 = vzip_u16(v1_0.val[1], v1_1.val[1]); in xnn_x16_transposec_ukernel__4x4_reuse_mov_zip_neon() local
|
| D | 4x4-reuse-switch-zip-neon.c | 54 const uint16x4x2_t v0_1 = vzip_u16(v1_0.val[1], v1_1.val[1]); in xnn_x16_transposec_ukernel__4x4_reuse_switch_zip_neon() local 90 const uint16x4x2_t v0_1 = vzip_u16(v1_0.val[1], v1_1.val[1]); in xnn_x16_transposec_ukernel__4x4_reuse_switch_zip_neon() local
|
| D | 4x4-multi-mov-zip-neon.c | 59 const uint16x4x2_t v0_1 = vzip_u16(v1_0.val[1], v1_1.val[1]); in xnn_x16_transposec_ukernel__4x4_multi_mov_zip_neon() local 97 const uint16x4x2_t v0_1 = vzip_u16(v1_0.val[1], v1_1.val[1]); in xnn_x16_transposec_ukernel__4x4_multi_mov_zip_neon() local
|
| D | 4x4-multi-switch-zip-neon.c | 58 const uint16x4x2_t v0_1 = vzip_u16(v1_0.val[1], v1_1.val[1]); in xnn_x16_transposec_ukernel__4x4_multi_switch_zip_neon() local 92 const uint16x4x2_t v0_1 = vzip_u16(v1_0.val[1], v1_1.val[1]); in xnn_x16_transposec_ukernel__4x4_multi_switch_zip_neon() local
|
| /external/XNNPACK/src/x32-transposec/gen/ |
| D | 4x4-reuse-multi-zip-neon.c | 63 const uint32x4x2_t v0_1 = vzipq_u32(v1_0.val[1], v1_1.val[1]); in xnn_x32_transposec_ukernel__4x4_reuse_multi_zip_neon() local 89 const uint32x4x2_t v0_1 = vzipq_u32(v1_0.val[1], v1_1.val[1]); in xnn_x32_transposec_ukernel__4x4_reuse_multi_zip_neon() local
|
| D | 4x4-multi-multi-zip-neon.c | 67 const uint32x4x2_t v0_1 = vzipq_u32(v1_0.val[1], v1_1.val[1]); in xnn_x32_transposec_ukernel__4x4_multi_multi_zip_neon() local 91 const uint32x4x2_t v0_1 = vzipq_u32(v1_0.val[1], v1_1.val[1]); in xnn_x32_transposec_ukernel__4x4_multi_multi_zip_neon() local
|
| D | 4x4-reuse-dec-zip-neon.c | 55 const uint32x4x2_t v0_1 = vzipq_u32(v1_0.val[1], v1_1.val[1]); in xnn_x32_transposec_ukernel__4x4_reuse_dec_zip_neon() local 92 const uint32x4x2_t v0_1 = vzipq_u32(v1_0.val[1], v1_1.val[1]); in xnn_x32_transposec_ukernel__4x4_reuse_dec_zip_neon() local
|
| D | 4x4-reuse-switch-zip-neon.c | 54 const uint32x4x2_t v0_1 = vzipq_u32(v1_0.val[1], v1_1.val[1]); in xnn_x32_transposec_ukernel__4x4_reuse_switch_zip_neon() local 90 const uint32x4x2_t v0_1 = vzipq_u32(v1_0.val[1], v1_1.val[1]); in xnn_x32_transposec_ukernel__4x4_reuse_switch_zip_neon() local
|
| D | 4x4-reuse-mov-zip-neon.c | 55 const uint32x4x2_t v0_1 = vzipq_u32(v1_0.val[1], v1_1.val[1]); in xnn_x32_transposec_ukernel__4x4_reuse_mov_zip_neon() local 95 const uint32x4x2_t v0_1 = vzipq_u32(v1_0.val[1], v1_1.val[1]); in xnn_x32_transposec_ukernel__4x4_reuse_mov_zip_neon() local
|
| D | 4x4-multi-dec-zip-neon.c | 59 const uint32x4x2_t v0_1 = vzipq_u32(v1_0.val[1], v1_1.val[1]); in xnn_x32_transposec_ukernel__4x4_multi_dec_zip_neon() local 94 const uint32x4x2_t v0_1 = vzipq_u32(v1_0.val[1], v1_1.val[1]); in xnn_x32_transposec_ukernel__4x4_multi_dec_zip_neon() local
|
| D | 4x4-reuse-multi-sse2.c | 71 const __m128i v0_1 = _mm_unpackhi_epi64(v1_0, v1_2); in xnn_x32_transposec_ukernel__4x4_reuse_multi_sse2() local 106 __m128i v0_1 = _mm_unpackhi_epi64(v1_0, v1_2); in xnn_x32_transposec_ukernel__4x4_reuse_multi_sse2() local
|
| D | 4x4-reuse-multi-wasmsimd.c | 68 const v128_t v0_1 = wasm_v32x4_shuffle(v1_0, v1_2, 2, 6, 3, 7); in xnn_x32_transposec_ukernel__4x4_reuse_multi_wasmsimd() local 102 v128_t v0_1 = wasm_v32x4_shuffle(v1_0, v1_2, 2, 6, 3, 7); in xnn_x32_transposec_ukernel__4x4_reuse_multi_wasmsimd() local
|
| D | 4x4-multi-mov-zip-neon.c | 59 const uint32x4x2_t v0_1 = vzipq_u32(v1_0.val[1], v1_1.val[1]); in xnn_x32_transposec_ukernel__4x4_multi_mov_zip_neon() local 97 const uint32x4x2_t v0_1 = vzipq_u32(v1_0.val[1], v1_1.val[1]); in xnn_x32_transposec_ukernel__4x4_multi_mov_zip_neon() local
|
| D | 4x4-multi-switch-zip-neon.c | 58 const uint32x4x2_t v0_1 = vzipq_u32(v1_0.val[1], v1_1.val[1]); in xnn_x32_transposec_ukernel__4x4_multi_switch_zip_neon() local 92 const uint32x4x2_t v0_1 = vzipq_u32(v1_0.val[1], v1_1.val[1]); in xnn_x32_transposec_ukernel__4x4_multi_switch_zip_neon() local
|
| D | 4x4-multi-multi-sse2.c | 75 const __m128i v0_1 = _mm_unpackhi_epi64(v1_0, v1_2); in xnn_x32_transposec_ukernel__4x4_multi_multi_sse2() local 108 __m128i v0_1 = _mm_unpackhi_epi64(v1_0, v1_2); in xnn_x32_transposec_ukernel__4x4_multi_multi_sse2() local
|