Searched refs:MS_MLAQ_F32 (Results 1 – 5 of 5) sorted by relevance
| /third_party/mindspore/mindspore/ccsrc/backend/kernel_compiler/cpu/nnacl/intrinsics/sse/ |
| D | ConvDwFp32Row_sse.c | 43 dst1 = MS_MLAQ_F32(dst1, w1, in1); in ConvDwFp32Row()
44 dst2 = MS_MLAQ_F32(dst2, w2, in2); in ConvDwFp32Row()
45 dst3 = MS_MLAQ_F32(dst3, w3, in3); in ConvDwFp32Row()
46 dst4 = MS_MLAQ_F32(dst4, w4, in4); in ConvDwFp32Row()
62 dst1 = MS_MLAQ_F32(dst1, w1, in1); in ConvDwFp32Row()
63 dst2 = MS_MLAQ_F32(dst2, w2, in2); in ConvDwFp32Row()
74 dst1 = MS_MLAQ_F32(dst1, w1, in1); in ConvDwFp32Row()
|
| /third_party/mindspore/mindspore/ccsrc/backend/kernel_compiler/cpu/nnacl/fp32/ |
| D | scale_fp32.c | 44 MS_FLOAT32X4 result = MS_MLAQ_F32(offset_4, data, scale_4); in ScaleInner()
77 MS_FLOAT32X4 result = MS_MLAQ_F32(offset_4, data, scale_4); in ScaleAxis()
135 MS_FLOAT32X4 tmp = MS_MLAQ_F32(offset_4, data, scale_4); in ScaleInnerRelu()
177 MS_FLOAT32X4 tmp = MS_MLAQ_F32(offset_4, data, scale_4); in ScaleAxisRelu()
239 MS_FLOAT32X4 tmp = MS_MLAQ_F32(offset_4, data, scale_4); in ScaleInnerRelu6()
283 MS_FLOAT32X4 tmp = MS_MLAQ_F32(offset_4, data, scale_4); in ScaleAxisRelu6()
|
| D | conv_depthwise_fp32.c | 606 acc0 = MS_MLAQ_F32(acc0, MS_LDQ_F32(line1), g10); in ConvDw3x3Line()
607 acc1 = MS_MLAQ_F32(acc1, MS_LDQ_F32(line1 + 4), g11); in ConvDw3x3Line()
608 acc2 = MS_MLAQ_F32(acc2, MS_LDQ_F32(line1 + 8), g12); in ConvDw3x3Line()
609 acc3 = MS_MLAQ_F32(acc3, MS_LDQ_F32(line1 + 12), g13); in ConvDw3x3Line()
611 acc0 = MS_MLAQ_F32(acc0, MS_LDQ_F32(line2), g20); in ConvDw3x3Line()
612 acc1 = MS_MLAQ_F32(acc1, MS_LDQ_F32(line2 + 4), g21); in ConvDw3x3Line()
613 acc2 = MS_MLAQ_F32(acc2, MS_LDQ_F32(line2 + 8), g22); in ConvDw3x3Line()
614 acc3 = MS_MLAQ_F32(acc3, MS_LDQ_F32(line2 + 12), g23); in ConvDw3x3Line()
644 acc0 = MS_MLAQ_F32(acc0, MS_LDQ_F32(line1), g10); in ConvDw3x3Line()
645 acc1 = MS_MLAQ_F32(acc1, MS_LDQ_F32(line1 + 4), g11); in ConvDw3x3Line()
[all …]
|
| /third_party/mindspore/mindspore/ccsrc/backend/kernel_compiler/cpu/nnacl/intrinsics/ |
| D | ms_simd_instructions.h | 56 #define MS_MLAQ_F32(src1, src2, src3) vmlaq_f32(src1, src2, src3) macro
132 #define MS_MLAQ_F32(src1, src2, src3) _mm_add_ps(src1, _mm_mul_ps(src2, src3)) macro
|
| /third_party/mindspore/mindspore/ccsrc/backend/kernel_compiler/cpu/nnacl/base/ |
| D | minimal_filtering_generator.c | 246 res = MS_MLAQ_F32(res, matrix_a[h_offset + i], matrix_b[w + i * n]); in MatrixMultiplyVec()
|