/aosp_15_r20/external/XNNPACK/src/x32-transposec/gen/ |
H A D | 4x4-reuse-multi-zip-neon.c | 57 const uint32x4_t v2_3 = vld1q_u32(i0); i0 = (uint32_t*) ((uintptr_t) i0 + input_stride); in xnn_x32_transposec_ukernel__4x4_reuse_multi_zip_neon() local 83 const uint32x4_t v2_3 = vmovq_n_u32(0); in xnn_x32_transposec_ukernel__4x4_reuse_multi_zip_neon() local
|
H A D | 4x4-multi-multi-zip-neon.c | 61 const uint32x4_t v2_3 = vld1q_u32(i3); i3 = (uint32_t*) ((uintptr_t) i3 + input_offset); in xnn_x32_transposec_ukernel__4x4_multi_multi_zip_neon() local 85 const uint32x4_t v2_3 = vmovq_n_u32(0); in xnn_x32_transposec_ukernel__4x4_multi_multi_zip_neon() local
|
H A D | 4x4-reuse-dec-zip-neon.c | 49 const uint32x4_t v2_3 = vld1q_u32(i0); i0 = (uint32_t*) ((uintptr_t) i0 + input_stride); in xnn_x32_transposec_ukernel__4x4_reuse_dec_zip_neon() local 86 const uint32x4_t v2_3 = vmovq_n_u32(0); in xnn_x32_transposec_ukernel__4x4_reuse_dec_zip_neon() local
|
H A D | 4x4-reuse-multi-wasmsimd.c | 60 const v128_t v2_3 = wasm_v128_load(i0); in xnn_x32_transposec_ukernel__4x4_reuse_multi_wasmsimd() local 94 const v128_t v2_3 = wasm_v128_xor(v2_0, v2_0); in xnn_x32_transposec_ukernel__4x4_reuse_multi_wasmsimd() local
|
H A D | 4x4-reuse-mov-zip-neon.c | 49 const uint32x4_t v2_3 = vld1q_u32(i0); i0 = (uint32_t*) ((uintptr_t) i0 + input_stride); in xnn_x32_transposec_ukernel__4x4_reuse_mov_zip_neon() local 89 const uint32x4_t v2_3 = vmovq_n_u32(0); in xnn_x32_transposec_ukernel__4x4_reuse_mov_zip_neon() local
|
H A D | 4x4-multi-dec-zip-neon.c | 53 const uint32x4_t v2_3 = vld1q_u32(i3); i3 = (uint32_t*) ((uintptr_t) i3 + input_offset); in xnn_x32_transposec_ukernel__4x4_multi_dec_zip_neon() local 88 const uint32x4_t v2_3 = vmovq_n_u32(0); in xnn_x32_transposec_ukernel__4x4_multi_dec_zip_neon() local
|
H A D | 4x4-reuse-multi-sse2.c | 62 const __m128i v2_3 = _mm_loadu_si128((const __m128i*) i0); in xnn_x32_transposec_ukernel__4x4_reuse_multi_sse2() local 98 const __m128i v2_3 = _mm_undefined_si128(); in xnn_x32_transposec_ukernel__4x4_reuse_multi_sse2() local
|
H A D | 4x4-reuse-switch-zip-neon.c | 48 const uint32x4_t v2_3 = vld1q_u32(i0); i0 = (uint32_t*) ((uintptr_t) i0 + input_stride); in xnn_x32_transposec_ukernel__4x4_reuse_switch_zip_neon() local 84 const uint32x4_t v2_3 = vmovq_n_u32(0); in xnn_x32_transposec_ukernel__4x4_reuse_switch_zip_neon() local
|
H A D | 4x4-multi-multi-wasmsimd.c | 64 const v128_t v2_3 = wasm_v128_load(i3); in xnn_x32_transposec_ukernel__4x4_multi_multi_wasmsimd() local 96 const v128_t v2_3 = wasm_v128_xor(v2_0, v2_0); in xnn_x32_transposec_ukernel__4x4_multi_multi_wasmsimd() local
|
H A D | 4x4-multi-mov-zip-neon.c | 53 const uint32x4_t v2_3 = vld1q_u32(i3); i3 = (uint32_t*) ((uintptr_t) i3 + input_offset); in xnn_x32_transposec_ukernel__4x4_multi_mov_zip_neon() local 91 const uint32x4_t v2_3 = vmovq_n_u32(0); in xnn_x32_transposec_ukernel__4x4_multi_mov_zip_neon() local
|
H A D | 4x4-multi-multi-sse2.c | 66 const __m128i v2_3 = _mm_loadu_si128((const __m128i*) i3); in xnn_x32_transposec_ukernel__4x4_multi_multi_sse2() local 100 const __m128i v2_3 = _mm_undefined_si128(); in xnn_x32_transposec_ukernel__4x4_multi_multi_sse2() local
|
H A D | 4x4-multi-switch-zip-neon.c | 52 const uint32x4_t v2_3 = vld1q_u32(i3); i3 = (uint32_t*) ((uintptr_t) i3 + input_offset); in xnn_x32_transposec_ukernel__4x4_multi_switch_zip_neon() local 86 const uint32x4_t v2_3 = vmovq_n_u32(0); in xnn_x32_transposec_ukernel__4x4_multi_switch_zip_neon() local
|
H A D | 4x4-reuse-switch-sse2.c | 53 const __m128i v2_3 = _mm_loadu_si128((const __m128i*) i0); in xnn_x32_transposec_ukernel__4x4_reuse_switch_sse2() local 98 const __m128i v2_3 = _mm_undefined_si128(); in xnn_x32_transposec_ukernel__4x4_reuse_switch_sse2() local
|
H A D | 4x4-reuse-mov-sse2.c | 54 const __m128i v2_3 = _mm_loadu_si128((const __m128i*) i0); in xnn_x32_transposec_ukernel__4x4_reuse_mov_sse2() local 100 const __m128i v2_3 = _mm_undefined_si128(); in xnn_x32_transposec_ukernel__4x4_reuse_mov_sse2() local
|
H A D | 4x4-reuse-mov-wasmsimd.c | 52 const v128_t v2_3 = wasm_v128_load(i0); in xnn_x32_transposec_ukernel__4x4_reuse_mov_wasmsimd() local 96 const v128_t v2_3 = wasm_v128_xor(v2_0, v2_0); in xnn_x32_transposec_ukernel__4x4_reuse_mov_wasmsimd() local
|
H A D | 4x4-reuse-switch-wasmsimd.c | 51 const v128_t v2_3 = wasm_v128_load(i0); in xnn_x32_transposec_ukernel__4x4_reuse_switch_wasmsimd() local 94 const v128_t v2_3 = wasm_v128_xor(v2_0, v2_0); in xnn_x32_transposec_ukernel__4x4_reuse_switch_wasmsimd() local
|
H A D | 4x4-multi-switch-sse2.c | 57 const __m128i v2_3 = _mm_loadu_si128((const __m128i*) i3); in xnn_x32_transposec_ukernel__4x4_multi_switch_sse2() local 100 const __m128i v2_3 = _mm_undefined_si128(); in xnn_x32_transposec_ukernel__4x4_multi_switch_sse2() local
|
/aosp_15_r20/external/XNNPACK/src/x16-transposec/gen/ |
H A D | 4x4-reuse-multi-zip-neon.c | 57 const uint16x4_t v2_3 = vld1_u16(i0); i0 = (uint16_t*) ((uintptr_t) i0 + input_stride); in xnn_x16_transposec_ukernel__4x4_reuse_multi_zip_neon() local 83 const uint16x4_t v2_3 = vmov_n_u16(0); in xnn_x16_transposec_ukernel__4x4_reuse_multi_zip_neon() local
|
H A D | 4x4-multi-multi-zip-neon.c | 61 const uint16x4_t v2_3 = vld1_u16(i3); i3 = (uint16_t*) ((uintptr_t) i3 + input_offset); in xnn_x16_transposec_ukernel__4x4_multi_multi_zip_neon() local 85 const uint16x4_t v2_3 = vmov_n_u16(0); in xnn_x16_transposec_ukernel__4x4_multi_multi_zip_neon() local
|
H A D | 4x4-reuse-dec-zip-neon.c | 49 const uint16x4_t v2_3 = vld1_u16(i0); i0 = (uint16_t*) ((uintptr_t) i0 + input_stride); in xnn_x16_transposec_ukernel__4x4_reuse_dec_zip_neon() local 86 const uint16x4_t v2_3 = vmov_n_u16(0); in xnn_x16_transposec_ukernel__4x4_reuse_dec_zip_neon() local
|
H A D | 4x4-reuse-switch-zip-neon.c | 48 const uint16x4_t v2_3 = vld1_u16(i0); i0 = (uint16_t*) ((uintptr_t) i0 + input_stride); in xnn_x16_transposec_ukernel__4x4_reuse_switch_zip_neon() local 84 const uint16x4_t v2_3 = vmov_n_u16(0); in xnn_x16_transposec_ukernel__4x4_reuse_switch_zip_neon() local
|
H A D | 4x4-multi-dec-zip-neon.c | 53 const uint16x4_t v2_3 = vld1_u16(i3); i3 = (uint16_t*) ((uintptr_t) i3 + input_offset); in xnn_x16_transposec_ukernel__4x4_multi_dec_zip_neon() local 88 const uint16x4_t v2_3 = vmov_n_u16(0); in xnn_x16_transposec_ukernel__4x4_multi_dec_zip_neon() local
|
H A D | 4x4-reuse-mov-zip-neon.c | 49 const uint16x4_t v2_3 = vld1_u16(i0); i0 = (uint16_t*) ((uintptr_t) i0 + input_stride); in xnn_x16_transposec_ukernel__4x4_reuse_mov_zip_neon() local 89 const uint16x4_t v2_3 = vmov_n_u16(0); in xnn_x16_transposec_ukernel__4x4_reuse_mov_zip_neon() local
|
H A D | 4x4-multi-mov-zip-neon.c | 53 const uint16x4_t v2_3 = vld1_u16(i3); i3 = (uint16_t*) ((uintptr_t) i3 + input_offset); in xnn_x16_transposec_ukernel__4x4_multi_mov_zip_neon() local 91 const uint16x4_t v2_3 = vmov_n_u16(0); in xnn_x16_transposec_ukernel__4x4_multi_mov_zip_neon() local
|
H A D | 4x4-multi-switch-zip-neon.c | 52 const uint16x4_t v2_3 = vld1_u16(i3); i3 = (uint16_t*) ((uintptr_t) i3 + input_offset); in xnn_x16_transposec_ukernel__4x4_multi_switch_zip_neon() local 86 const uint16x4_t v2_3 = vmov_n_u16(0); in xnn_x16_transposec_ukernel__4x4_multi_switch_zip_neon() local
|