/external/XNNPACK/src/f32-gemm/gen-inc/ |
D | 8x8s4-neon.c | 112 float32x4_t va6 = vld1q_f32(a6); a6 += 4; in xnn_f32_gemminc_ukernel_8x8s4__neon() local 235 const float32x4_t va6 = vld1q_dup_f32(a6); a6 += 1; in xnn_f32_gemminc_ukernel_8x8s4__neon() local
|
D | 8x8s4-neonfma.c | 112 float32x4_t va6 = vld1q_f32(a6); a6 += 4; in xnn_f32_gemminc_ukernel_8x8s4__neonfma() local 235 const float32x4_t va6 = vld1q_dup_f32(a6); a6 += 1; in xnn_f32_gemminc_ukernel_8x8s4__neonfma() local
|
D | 7x8-avx-broadcast.c | 103 const __m256 va6 = _mm256_broadcast_ss(a6); in xnn_f32_gemminc_ukernel_7x8__avx_broadcast() local
|
D | 7x8-fma3-broadcast.c | 103 const __m256 va6 = _mm256_broadcast_ss(a6); in xnn_f32_gemminc_ukernel_7x8__fma3_broadcast() local
|
D | 8x8-fma3-broadcast.c | 110 const __m256 va6 = _mm256_broadcast_ss(a6); in xnn_f32_gemminc_ukernel_8x8__fma3_broadcast() local
|
/external/XNNPACK/src/f32-igemm/gen/ |
D | 8x8s4-neon.c | 143 float32x4_t va6 = vld1q_f32(a6); a6 += 4; in xnn_f32_igemm_ukernel_8x8s4__neon() local 266 const float32x4_t va6 = vld1q_dup_f32(a6); a6 += 1; in xnn_f32_igemm_ukernel_8x8s4__neon() local
|
D | 8x8s4-neonfma.c | 143 float32x4_t va6 = vld1q_f32(a6); a6 += 4; in xnn_f32_igemm_ukernel_8x8s4__neonfma() local 266 const float32x4_t va6 = vld1q_dup_f32(a6); a6 += 1; in xnn_f32_igemm_ukernel_8x8s4__neonfma() local
|
D | 7x8-fma3-broadcast.c | 135 const __m256 va6 = _mm256_broadcast_ss(a6); in xnn_f32_igemm_ukernel_7x8__fma3_broadcast() local
|
D | 7x8-avx-broadcast.c | 135 const __m256 va6 = _mm256_broadcast_ss(a6); in xnn_f32_igemm_ukernel_7x8__avx_broadcast() local
|
D | 8x8-fma3-broadcast.c | 145 const __m256 va6 = _mm256_broadcast_ss(a6); in xnn_f32_igemm_ukernel_8x8__fma3_broadcast() local
|
/external/XNNPACK/src/f32-gemm/gen/ |
D | 8x8s4-neonfma.c | 110 float32x4_t va6 = vld1q_f32(a6); a6 += 4; in xnn_f32_gemm_ukernel_8x8s4__neonfma() local 233 const float32x4_t va6 = vld1q_dup_f32(a6); a6 += 1; in xnn_f32_gemm_ukernel_8x8s4__neonfma() local
|
D | 8x8s4-neon.c | 110 float32x4_t va6 = vld1q_f32(a6); a6 += 4; in xnn_f32_gemm_ukernel_8x8s4__neon() local 233 const float32x4_t va6 = vld1q_dup_f32(a6); a6 += 1; in xnn_f32_gemm_ukernel_8x8s4__neon() local
|
D | 7x8-fma3-broadcast.c | 101 const __m256 va6 = _mm256_broadcast_ss(a6); in xnn_f32_gemm_ukernel_7x8__fma3_broadcast() local
|
D | 7x8-avx-broadcast.c | 101 const __m256 va6 = _mm256_broadcast_ss(a6); in xnn_f32_gemm_ukernel_7x8__avx_broadcast() local
|
D | 8x8-fma3-broadcast.c | 108 const __m256 va6 = _mm256_broadcast_ss(a6); in xnn_f32_gemm_ukernel_8x8__fma3_broadcast() local
|
/external/XNNPACK/src/f16-gemm/gen/ |
D | 8x8-neonfp16arith-ld64.c | 102 const float16x4_t va6 = vld1_f16(a6); a6 += 4; in xnn_f16_gemm_ukernel_8x8__neonfp16arith_ld64() local 236 const float16x8_t va6 = vld1q_dup_f16(a6); a6 += 1; in xnn_f16_gemm_ukernel_8x8__neonfp16arith_ld64() local
|
/external/XNNPACK/src/f32-spmm/gen/ |
D | 8x1-scalar.c | 55 const float va6 = a[6]; in xnn_f32_spmm_ukernel_8x1__scalar() local 116 const float va6 = a[6]; in xnn_f32_spmm_ukernel_8x1__scalar() local
|
D | 8x2-scalar.c | 63 const float va6 = a[6]; in xnn_f32_spmm_ukernel_8x2__scalar() local 157 const float va6 = a[6]; in xnn_f32_spmm_ukernel_8x2__scalar() local
|
D | 8x1-scalar-pipelined.c | 43 float va6 = a[6]; in xnn_f32_spmm_ukernel_8x1__scalar_pipelined() local
|
D | 8x4-scalar.c | 79 const float va6 = a[6]; in xnn_f32_spmm_ukernel_8x4__scalar() local 239 const float va6 = a[6]; in xnn_f32_spmm_ukernel_8x4__scalar() local
|
/external/XNNPACK/src/q8-igemm/ |
D | 8x8-neon.c | 132 const uint8x8_t va6 = vld1_u8(a6); a6 += 8; in xnn_q8_igemm_ukernel_8x8__neon() local 328 const uint8x8_t va6 = vld1_u8(a6); in xnn_q8_igemm_ukernel_8x8__neon() local
|
/external/XNNPACK/src/q8-gemm/ |
D | 8x8-neon.c | 113 const uint8x8_t va6 = vld1_u8(a6); in xnn_q8_gemm_ukernel_8x8__neon() local 293 const uint8x8_t va6 = vld1_u8(a6); a6 = (const uint8_t*) ((uintptr_t) a6 + k); in xnn_q8_gemm_ukernel_8x8__neon() local
|