vadds/vmuls

功能说明

以block(32Byte)为单位完成以下计算，一次完成8个block的计算。

函数原型

void vadds(__ubuf__ half *dst, __ubuf__ half *src, half a, uint8_t repeat, uint16_t dstBlockStride, uint16_t srcBlockStride, uint16_t dstRepeatStride, uint16_t srcRepeatStride); 
void vadds(__ubuf__ float *dst, __ubuf__ float *src, float a, uint8_t repeat, uint16_t dstBlockStride, uint16_t srcBlockStride, uint16_t dstRepeatStride, uint16_t srcRepeatStride);

void vadds(__ubuf__ int16_t *dst, __ubuf__ int16_t *src, int16_t a, uint8_t repeat, uint16_t dstBlockStride, uint16_t srcBlockStride, uint16_t dstRepeatStride, uint16_t srcRepeatStride); 

void vadds(__ubuf__ int32_t *dst, __ubuf__ int32_t *src, int32_t a, uint8_t repeat, uint16_t dstBlockStride, uint16_t srcBlockStride, uint16_t dstRepeatStride, uint16_t srcRepeatStride);

流水类型

PIPE_V

同类型接口

// vmuls 
void vmuls(__ubuf__ half *dst, __ubuf__ half *src0, half src1, uint8_t repeat, uint16_t dstBlockStride, uint16_t srcBlockStride, uint16_t dstRepeatStride, uint16_t srcRepeatStride); 

void vmuls(__ubuf__ float *dst, __ubuf__ float *src0, float src1, uint8_t repeat, uint16_t dstBlockStride, uint16_t srcBlockStride, uint16_t dstRepeatStride, uint16_t srcRepeatStride); 

void vmuls(__ubuf__ int16_t *dst, __ubuf__ int16_t *src0, int16_t src1, uint8_t repeat, uint16_t dstBlockStride, uint16_t srcBlockStride, uint16_t dstRepeatStride, uint16_t srcRepeatStride); 

void vmuls(__ubuf__ int32_t *dst, __ubuf__ int32_t *src0, int32_t src1, uint8_t repeat, uint16_t dstBlockStride, uint16_t srcBlockStride, uint16_t dstRepeatStride, uint16_t srcRepeatStride);