| /external/XNNPACK/src/x64-transpose/gen/ |
| D | 2x2-reuse-multi-sse2.c | 51 const __m128i v0_0 = _mm_unpacklo_epi64(v1_0, v1_1); in xnn_x64_transpose_ukernel__2x2_reuse_multi_sse2() local 66 __m128i v0_0 = _mm_unpacklo_epi64(v1_0, v1_1); in xnn_x64_transpose_ukernel__2x2_reuse_multi_sse2() local
|
| D | 2x2-multi-multi-sse2.c | 53 const __m128i v0_0 = _mm_unpacklo_epi64(v1_0, v1_1); in xnn_x64_transpose_ukernel__2x2_multi_multi_sse2() local 68 __m128i v0_0 = _mm_unpacklo_epi64(v1_0, v1_1); in xnn_x64_transpose_ukernel__2x2_multi_multi_sse2() local
|
| D | 2x2-reuse-switch-sse2.c | 49 const __m128i v0_0 = _mm_unpacklo_epi64(v1_0, v1_1); in xnn_x64_transpose_ukernel__2x2_reuse_switch_sse2() local 71 __m128i v0_0 = _mm_unpacklo_epi64(v1_0, v1_1); in xnn_x64_transpose_ukernel__2x2_reuse_switch_sse2() local
|
| D | 2x2-reuse-mov-sse2.c | 51 const __m128i v0_0 = _mm_unpacklo_epi64(v1_0, v1_1); in xnn_x64_transpose_ukernel__2x2_reuse_mov_sse2() local 70 __m128i v0_0 = _mm_unpacklo_epi64(v1_0, v1_1); in xnn_x64_transpose_ukernel__2x2_reuse_mov_sse2() local
|
| D | 2x2-multi-switch-sse2.c | 51 const __m128i v0_0 = _mm_unpacklo_epi64(v1_0, v1_1); in xnn_x64_transpose_ukernel__2x2_multi_switch_sse2() local 73 __m128i v0_0 = _mm_unpacklo_epi64(v1_0, v1_1); in xnn_x64_transpose_ukernel__2x2_multi_switch_sse2() local
|
| D | 2x2-multi-mov-sse2.c | 53 const __m128i v0_0 = _mm_unpacklo_epi64(v1_0, v1_1); in xnn_x64_transpose_ukernel__2x2_multi_mov_sse2() local 72 __m128i v0_0 = _mm_unpacklo_epi64(v1_0, v1_1); in xnn_x64_transpose_ukernel__2x2_multi_mov_sse2() local
|
| /external/XNNPACK/src/x32-transpose/gen/ |
| D | 4x4-reuse-multi-zip-neon.c | 62 const uint32x4x2_t v0_0 = vzipq_u32(v1_0.val[0], v1_1.val[0]); in xnn_x32_transpose_ukernel__4x4_reuse_multi_zip_neon() local 88 const uint32x4x2_t v0_0 = vzipq_u32(v1_0.val[0], v1_1.val[0]); in xnn_x32_transpose_ukernel__4x4_reuse_multi_zip_neon() local
|
| D | 4x4-multi-multi-zip-neon.c | 66 const uint32x4x2_t v0_0 = vzipq_u32(v1_0.val[0], v1_1.val[0]); in xnn_x32_transpose_ukernel__4x4_multi_multi_zip_neon() local 90 const uint32x4x2_t v0_0 = vzipq_u32(v1_0.val[0], v1_1.val[0]); in xnn_x32_transpose_ukernel__4x4_multi_multi_zip_neon() local
|
| D | 4x4-reuse-dec-zip-neon.c | 54 const uint32x4x2_t v0_0 = vzipq_u32(v1_0.val[0], v1_1.val[0]); in xnn_x32_transpose_ukernel__4x4_reuse_dec_zip_neon() local 91 const uint32x4x2_t v0_0 = vzipq_u32(v1_0.val[0], v1_1.val[0]); in xnn_x32_transpose_ukernel__4x4_reuse_dec_zip_neon() local
|
| D | 4x4-reuse-multi-sse2.c | 68 const __m128i v0_0 = _mm_unpacklo_epi64(v1_0, v1_2); in xnn_x32_transpose_ukernel__4x4_reuse_multi_sse2() local 103 __m128i v0_0 = _mm_unpacklo_epi64(v1_0, v1_2); in xnn_x32_transpose_ukernel__4x4_reuse_multi_sse2() local
|
| D | 4x4-multi-dec-zip-neon.c | 58 const uint32x4x2_t v0_0 = vzipq_u32(v1_0.val[0], v1_1.val[0]); in xnn_x32_transpose_ukernel__4x4_multi_dec_zip_neon() local 93 const uint32x4x2_t v0_0 = vzipq_u32(v1_0.val[0], v1_1.val[0]); in xnn_x32_transpose_ukernel__4x4_multi_dec_zip_neon() local
|
| D | 4x4-reuse-switch-zip-neon.c | 53 const uint32x4x2_t v0_0 = vzipq_u32(v1_0.val[0], v1_1.val[0]); in xnn_x32_transpose_ukernel__4x4_reuse_switch_zip_neon() local 89 const uint32x4x2_t v0_0 = vzipq_u32(v1_0.val[0], v1_1.val[0]); in xnn_x32_transpose_ukernel__4x4_reuse_switch_zip_neon() local
|
| D | 4x4-reuse-mov-zip-neon.c | 54 const uint32x4x2_t v0_0 = vzipq_u32(v1_0.val[0], v1_1.val[0]); in xnn_x32_transpose_ukernel__4x4_reuse_mov_zip_neon() local 94 const uint32x4x2_t v0_0 = vzipq_u32(v1_0.val[0], v1_1.val[0]); in xnn_x32_transpose_ukernel__4x4_reuse_mov_zip_neon() local
|
| D | 4x4-multi-switch-zip-neon.c | 57 const uint32x4x2_t v0_0 = vzipq_u32(v1_0.val[0], v1_1.val[0]); in xnn_x32_transpose_ukernel__4x4_multi_switch_zip_neon() local 91 const uint32x4x2_t v0_0 = vzipq_u32(v1_0.val[0], v1_1.val[0]); in xnn_x32_transpose_ukernel__4x4_multi_switch_zip_neon() local
|
| D | 4x4-multi-mov-zip-neon.c | 58 const uint32x4x2_t v0_0 = vzipq_u32(v1_0.val[0], v1_1.val[0]); in xnn_x32_transpose_ukernel__4x4_multi_mov_zip_neon() local 96 const uint32x4x2_t v0_0 = vzipq_u32(v1_0.val[0], v1_1.val[0]); in xnn_x32_transpose_ukernel__4x4_multi_mov_zip_neon() local
|
| D | 4x4-multi-multi-sse2.c | 72 const __m128i v0_0 = _mm_unpacklo_epi64(v1_0, v1_2); in xnn_x32_transpose_ukernel__4x4_multi_multi_sse2() local 105 __m128i v0_0 = _mm_unpacklo_epi64(v1_0, v1_2); in xnn_x32_transpose_ukernel__4x4_multi_multi_sse2() local
|
| D | 4x4-reuse-mov-sse2.c | 60 const __m128i v0_0 = _mm_unpacklo_epi64(v1_0, v1_2); in xnn_x32_transpose_ukernel__4x4_reuse_mov_sse2() local 105 __m128i v0_0 = _mm_unpacklo_epi64(v1_0, v1_2); in xnn_x32_transpose_ukernel__4x4_reuse_mov_sse2() local
|
| D | 4x4-reuse-switch-sse2.c | 59 const __m128i v0_0 = _mm_unpacklo_epi64(v1_0, v1_2); in xnn_x32_transpose_ukernel__4x4_reuse_switch_sse2() local 103 __m128i v0_0 = _mm_unpacklo_epi64(v1_0, v1_2); in xnn_x32_transpose_ukernel__4x4_reuse_switch_sse2() local
|
| D | 4x4-multi-switch-sse2.c | 63 const __m128i v0_0 = _mm_unpacklo_epi64(v1_0, v1_2); in xnn_x32_transpose_ukernel__4x4_multi_switch_sse2() local 105 __m128i v0_0 = _mm_unpacklo_epi64(v1_0, v1_2); in xnn_x32_transpose_ukernel__4x4_multi_switch_sse2() local
|
| D | 4x4-multi-mov-sse2.c | 64 const __m128i v0_0 = _mm_unpacklo_epi64(v1_0, v1_2); in xnn_x32_transpose_ukernel__4x4_multi_mov_sse2() local 107 __m128i v0_0 = _mm_unpacklo_epi64(v1_0, v1_2); in xnn_x32_transpose_ukernel__4x4_multi_mov_sse2() local
|
| /external/XNNPACK/src/x16-transpose/gen/ |
| D | 8x8-reuse-dec-zip-neon.c | 64 const uint16x8x2_t v0_0 = vzipq_u16(v1_0.val[0], v1_2.val[0]); in xnn_x16_transpose_ukernel__8x8_reuse_dec_zip_neon() local 145 const uint16x8x2_t v0_0 = vzipq_u16(v1_0.val[0], v1_2.val[0]); in xnn_x16_transpose_ukernel__8x8_reuse_dec_zip_neon() local
|
| D | 8x8-reuse-switch-zip-neon.c | 63 const uint16x8x2_t v0_0 = vzipq_u16(v1_0.val[0], v1_2.val[0]); in xnn_x16_transpose_ukernel__8x8_reuse_switch_zip_neon() local 135 const uint16x8x2_t v0_0 = vzipq_u16(v1_0.val[0], v1_2.val[0]); in xnn_x16_transpose_ukernel__8x8_reuse_switch_zip_neon() local
|
| D | 8x8-multi-dec-zip-neon.c | 72 const uint16x8x2_t v0_0 = vzipq_u16(v1_0.val[0], v1_2.val[0]); in xnn_x16_transpose_ukernel__8x8_multi_dec_zip_neon() local 147 const uint16x8x2_t v0_0 = vzipq_u16(v1_0.val[0], v1_2.val[0]); in xnn_x16_transpose_ukernel__8x8_multi_dec_zip_neon() local
|
| D | 8x8-reuse-multi-zip-neon.c | 88 const uint16x8x2_t v0_0 = vzipq_u16(v1_0.val[0], v1_2.val[0]); in xnn_x16_transpose_ukernel__8x8_reuse_multi_zip_neon() local 146 const uint16x8x2_t v0_0 = vzipq_u16(v1_0.val[0], v1_2.val[0]); in xnn_x16_transpose_ukernel__8x8_reuse_multi_zip_neon() local
|
| D | 8x8-reuse-mov-zip-neon.c | 64 const uint16x8x2_t v0_0 = vzipq_u16(v1_0.val[0], v1_2.val[0]); in xnn_x16_transpose_ukernel__8x8_reuse_mov_zip_neon() local 152 const uint16x8x2_t v0_0 = vzipq_u16(v1_0.val[0], v1_2.val[0]); in xnn_x16_transpose_ukernel__8x8_reuse_mov_zip_neon() local
|