STM32 USB库中为何要交换wValue和wIndex的字节序

只看该作者 · 2011-9-16 11:00

移植一个AVR的程序到STM32，发现USB库中无聊的交换了wValue和wIndex的字节序。如下代码：

   pInformation->USBbmRequestType = *pBuf.b++; /* bmRequestType */

    pInformation->USBbRequest = *pBuf.b++; /* bRequest */

    pBuf.w += offset;  /* word not accessed because of 32 bits addressing */

    pInformation->USBwValue = ByteSwap(*pBuf.w++); /* wValue */

    pBuf.w += offset;  /* word not accessed because of 32 bits addressing */

    pInformation->USBwIndex  = ByteSwap(*pBuf.w++); /* wIndex */

    pBuf.w += offset;  /* word not accessed because of 32 bits addressing */

    pInformation->USBwLength = *pBuf.w; /* wLength */

搜索所有的历程，只有DFU用到word值，还蛋疼的再交换回来，如下：

复制

  B0 = pInfo->USBwValues.bw.bb0;

  B1 = pInfo->USBwValues.bw.bb1;

  wBlockNum = (uint16_t)B1;

  wBlockNum = wBlockNum * 0x100;

  wBlockNum += (uint16_t)B0; /* wBlockNum value updated*/



  B0 = pInfo->USBwLengths.bw.bb0;

  B1 =  pInfo->USBwLengths.bw.bb1;

  wlength = (uint16_t)B0;

  wlength = wlength * 0x100;

  wlength += (uint16_t)B1; /* wlength value updated*/

USB传输和STM32都是little-endian字节序，为何要多此一举呢？有达人可以给解释一下吗？