/external/XNNPACK/src/f32-dwconv/gen/ |
D | up8x25-minmax-avx.c | 222 const __m256 vi9x01234567 = _mm256_loadu_ps(i9); in xnn_f32_dwconv_minmax_ukernel_up8x25__avx() local 370 const __m256 vi9x01234567 = _mm256_maskload_ps(i9, vmask); in xnn_f32_dwconv_minmax_ukernel_up8x25__avx() local
|
D | up8x25-minmax-fma3-acc2.c | 222 const __m256 vi9x01234567 = _mm256_loadu_ps(i9); in xnn_f32_dwconv_minmax_ukernel_up8x25__fma3_acc2() local 372 const __m256 vi9x01234567 = _mm256_maskload_ps(i9, vmask); in xnn_f32_dwconv_minmax_ukernel_up8x25__fma3_acc2() local
|
D | up8x25-minmax-fma3.c | 222 const __m256 vi9x01234567 = _mm256_loadu_ps(i9); in xnn_f32_dwconv_minmax_ukernel_up8x25__fma3() local 370 const __m256 vi9x01234567 = _mm256_maskload_ps(i9, vmask); in xnn_f32_dwconv_minmax_ukernel_up8x25__fma3() local
|
D | up8x25-minmax-avx-acc2.c | 222 const __m256 vi9x01234567 = _mm256_loadu_ps(i9); in xnn_f32_dwconv_minmax_ukernel_up8x25__avx_acc2() local 372 const __m256 vi9x01234567 = _mm256_maskload_ps(i9, vmask); in xnn_f32_dwconv_minmax_ukernel_up8x25__avx_acc2() local
|
/external/XNNPACK/src/f16-dwconv/gen/ |
D | up8x25-minmax-neonfp16arith.c | 206 const float16x8_t vi9x01234567 = vld1q_f16(i9); i9 += 8; in xnn_f16_dwconv_minmax_ukernel_up8x25__neonfp16arith() local 316 const float16x8_t vi9x01234567 = vld1q_f16(i9); in xnn_f16_dwconv_minmax_ukernel_up8x25__neonfp16arith() local
|
D | up8x25-minmax-neonfp16arith-acc2.c | 206 const float16x8_t vi9x01234567 = vld1q_f16(i9); i9 += 8; in xnn_f16_dwconv_minmax_ukernel_up8x25__neonfp16arith_acc2() local 318 const float16x8_t vi9x01234567 = vld1q_f16(i9); in xnn_f16_dwconv_minmax_ukernel_up8x25__neonfp16arith_acc2() local
|
D | up8x25-minmax-fma3-acc2.c | 225 const __m256 vi9x01234567 = _mm256_cvtph_ps(_mm_loadu_si128((const __m128i*) i9)); in xnn_f16_dwconv_minmax_ukernel_up8x25__fma3_acc2() local 383 const __m256 vi9x01234567 = _mm256_cvtph_ps(_mm_loadu_si128((const __m128i*) i9)); in xnn_f16_dwconv_minmax_ukernel_up8x25__fma3_acc2() local
|
D | up8x25-minmax-fma3.c | 225 const __m256 vi9x01234567 = _mm256_cvtph_ps(_mm_loadu_si128((const __m128i*) i9)); in xnn_f16_dwconv_minmax_ukernel_up8x25__fma3() local 381 const __m256 vi9x01234567 = _mm256_cvtph_ps(_mm_loadu_si128((const __m128i*) i9)); in xnn_f16_dwconv_minmax_ukernel_up8x25__fma3() local
|
/external/XNNPACK/src/qc8-dwconv/gen/ |
D | up8x25-minmax-fp32-neon-mul16.c | 225 const int16x8_t vi9x01234567 = vmovl_s8(vld1_s8(i9)); i9 += 8; in xnn_qc8_dwconv_minmax_fp32_ukernel_up8x25__neon_mul16() local 404 const int16x8_t vi9x01234567 = vmovl_s8(vld1_s8(i9)); in xnn_qc8_dwconv_minmax_fp32_ukernel_up8x25__neon_mul16() local
|
D | up8x25-minmax-fp32-avx2-mul32.c | 221 const __m256i vi9x01234567 = _mm256_cvtepi8_epi32(_mm_loadl_epi64((const __m128i*) i9)); in xnn_qc8_dwconv_minmax_fp32_ukernel_up8x25__avx2_mul32() local 391 const __m256i vi9x01234567 = _mm256_cvtepi8_epi32(_mm_loadl_epi64((const __m128i*) i9)); in xnn_qc8_dwconv_minmax_fp32_ukernel_up8x25__avx2_mul32() local
|
D | up8x25-minmax-fp32-neonv8-mul16.c | 225 const int16x8_t vi9x01234567 = vmovl_s8(vld1_s8(i9)); i9 += 8; in xnn_qc8_dwconv_minmax_fp32_ukernel_up8x25__neonv8_mul16() local 403 const int16x8_t vi9x01234567 = vmovl_s8(vld1_s8(i9)); in xnn_qc8_dwconv_minmax_fp32_ukernel_up8x25__neonv8_mul16() local
|
D | up8x25-minmax-fp32-wasmsimd-mul16-add16.c | 238 const v128_t vi9x01234567 = wasm_i16x8_load8x8(i9); in xnn_qc8_dwconv_minmax_fp32_ukernel_up8x25__wasmsimd_mul16_add16() local 471 const v128_t vi9x01234567 = wasm_i16x8_load8x8(i9); in xnn_qc8_dwconv_minmax_fp32_ukernel_up8x25__wasmsimd_mul16_add16() local
|
D | up8x25-minmax-fp32-neonv8-mla8-ld64.c | 223 const int8x8_t vi9x01234567 = vld1_s8(i9); i9 += 8; in xnn_qc8_dwconv_minmax_fp32_ukernel_up8x25__neonv8_mla8_ld64() local 412 const int8x8_t vi9x01234567 = vld1_s8(i9); in xnn_qc8_dwconv_minmax_fp32_ukernel_up8x25__neonv8_mla8_ld64() local
|
D | up8x25-minmax-fp32-neon-mla8-ld64.c | 223 const int8x8_t vi9x01234567 = vld1_s8(i9); i9 += 8; in xnn_qc8_dwconv_minmax_fp32_ukernel_up8x25__neon_mla8_ld64() local 413 const int8x8_t vi9x01234567 = vld1_s8(i9); in xnn_qc8_dwconv_minmax_fp32_ukernel_up8x25__neon_mla8_ld64() local
|
/external/XNNPACK/src/qs8-dwconv/gen/ |
D | up8x25-minmax-fp32-neon-mul16.c | 226 const int16x8_t vi9x01234567 = vmovl_s8(vld1_s8(i9)); i9 += 8; in xnn_qs8_dwconv_minmax_fp32_ukernel_up8x25__neon_mul16() local 402 const int16x8_t vi9x01234567 = vmovl_s8(vld1_s8(i9)); in xnn_qs8_dwconv_minmax_fp32_ukernel_up8x25__neon_mul16() local
|
D | up8x25-minmax-rndnu-neon-mul16.c | 227 const int16x8_t vi9x01234567 = vmovl_s8(vld1_s8(i9)); i9 += 8; in xnn_qs8_dwconv_minmax_rndnu_ukernel_up8x25__neon_mul16() local 402 const int16x8_t vi9x01234567 = vmovl_s8(vld1_s8(i9)); in xnn_qs8_dwconv_minmax_rndnu_ukernel_up8x25__neon_mul16() local
|
D | up8x25-minmax-fp32-wasmsimd-mul16-add16.c | 238 const v128_t vi9x01234567 = wasm_i16x8_load8x8(i9); in xnn_qs8_dwconv_minmax_fp32_ukernel_up8x25__wasmsimd_mul16_add16() local 468 const v128_t vi9x01234567 = wasm_i16x8_load8x8(i9); in xnn_qs8_dwconv_minmax_fp32_ukernel_up8x25__wasmsimd_mul16_add16() local
|
D | up8x25-minmax-fp32-avx2-mul32.c | 221 const __m256i vi9x01234567 = _mm256_cvtepi8_epi32(_mm_loadl_epi64((const __m128i*) i9)); in xnn_qs8_dwconv_minmax_fp32_ukernel_up8x25__avx2_mul32() local 390 const __m256i vi9x01234567 = _mm256_cvtepi8_epi32(_mm_loadl_epi64((const __m128i*) i9)); in xnn_qs8_dwconv_minmax_fp32_ukernel_up8x25__avx2_mul32() local
|
D | up8x25-minmax-fp32-neonv8-mul16.c | 226 const int16x8_t vi9x01234567 = vmovl_s8(vld1_s8(i9)); i9 += 8; in xnn_qs8_dwconv_minmax_fp32_ukernel_up8x25__neonv8_mul16() local 401 const int16x8_t vi9x01234567 = vmovl_s8(vld1_s8(i9)); in xnn_qs8_dwconv_minmax_fp32_ukernel_up8x25__neonv8_mul16() local
|
D | up8x25-minmax-fp32-wasmsimd-mul16.c | 248 const v128_t vi9x01234567 = wasm_i16x8_load8x8(i9); in xnn_qs8_dwconv_minmax_fp32_ukernel_up8x25__wasmsimd_mul16() local 502 const v128_t vi9x01234567 = wasm_i16x8_load8x8(i9); in xnn_qs8_dwconv_minmax_fp32_ukernel_up8x25__wasmsimd_mul16() local
|
D | up8x25-minmax-rndnu-neon-mla8-ld64.c | 225 const int8x8_t vi9x01234567 = vld1_s8(i9); i9 += 8; in xnn_qs8_dwconv_minmax_rndnu_ukernel_up8x25__neon_mla8_ld64() local 411 const int8x8_t vi9x01234567 = vld1_s8(i9); in xnn_qs8_dwconv_minmax_rndnu_ukernel_up8x25__neon_mla8_ld64() local
|
/external/XNNPACK/src/qu8-dwconv/gen/ |
D | up8x25-minmax-fp32-neon-mul16.c | 227 const int16x8_t vi9x01234567 = vreinterpretq_s16_u16(vmovl_u8(vld1_u8(i9))); i9 += 8; in xnn_qu8_dwconv_minmax_fp32_ukernel_up8x25__neon_mul16() local 403 const int16x8_t vi9x01234567 = vreinterpretq_s16_u16(vmovl_u8(vld1_u8(i9))); in xnn_qu8_dwconv_minmax_fp32_ukernel_up8x25__neon_mul16() local
|
D | up8x25-minmax-rndnu-neon-mul8.c | 247 const uint8x8_t vi9x01234567 = vld1_u8(i9); i9 += 8; in xnn_qu8_dwconv_minmax_rndnu_ukernel_up8x25__neon_mul8() local 454 const int16x8_t vi9x01234567 = vreinterpretq_s16_u16(vmovl_u8(vld1_u8(i9))); in xnn_qu8_dwconv_minmax_rndnu_ukernel_up8x25__neon_mul8() local
|
D | up8x25-minmax-fp32-avx2-mul32.c | 222 const __m256i vi9x01234567 = _mm256_cvtepu8_epi32(_mm_loadl_epi64((const __m128i*) i9)); in xnn_qu8_dwconv_minmax_fp32_ukernel_up8x25__avx2_mul32() local 391 const __m256i vi9x01234567 = _mm256_cvtepu8_epi32(_mm_loadl_epi64((const __m128i*) i9)); in xnn_qu8_dwconv_minmax_fp32_ukernel_up8x25__avx2_mul32() local
|
D | up8x25-minmax-fp32-wasmsimd-mul16.c | 257 const v128_t vi9x01234567 = wasm_u16x8_load8x8(i9); in xnn_qu8_dwconv_minmax_fp32_ukernel_up8x25__wasmsimd_mul16() local 537 const v128_t vi9x01234567 = wasm_u16x8_load8x8(i9); in xnn_qu8_dwconv_minmax_fp32_ukernel_up8x25__wasmsimd_mul16() local
|