#define DUFFS_DEVICE(x,size) \
  { \
  if(size  0) \
  { \
    int __DUFFS_DEVICE_count, __DUFFS_DEVICE_n; \
    __DUFFS_DEVICE_count = size; \
    __DUFFS_DEVICE_n = (__DUFFS_DEVICE_count+7)  3; \
    switch (__DUFFS_DEVICE_count & 7) \
    { \
    case 0: do { x; \
    case 7: x; \
    case 6: x; \
    case 5: x; \
    case 4: x; \
    case 3: x; \
    case 2: x; \
    case 1: x; \
      } while (--__DUFFS_DEVICE_n  0); \
    } \
  } \
  }

Sample usage:
/*
Previous code:
for(int i=0; i<size; i++) { *dest++ = *src++; }
*/
//New code:
DUFFS_DEVICE(*dest++ = *src++, size);

Another example:
/*
Previous code:
for(int i=0; i<size; i++) { do_something_given(i); }
*/
//New code:
//Note: for optimal performance, do_something_given() should be inlined
DUFFS_DEVICE(do_something_given(i); i++;, size);