Search code examples
algorithmbit-manipulation

How to generate lookup table for counting leading zeroes (clzlut)?


I found this function, but there is no explanation of where the clzlut lookup table came from (and I searched for many hours for it on the web, couldn't find anything):

static uint8_t clzlut[256] = {
  8,7,6,6,5,5,5,5,
  4,4,4,4,4,4,4,4,
  3,3,3,3,3,3,3,3,
  3,3,3,3,3,3,3,3,
  2,2,2,2,2,2,2,2,
  2,2,2,2,2,2,2,2,
  2,2,2,2,2,2,2,2,
  2,2,2,2,2,2,2,2,
  1,1,1,1,1,1,1,1,
  1,1,1,1,1,1,1,1,
  1,1,1,1,1,1,1,1,
  1,1,1,1,1,1,1,1,
  1,1,1,1,1,1,1,1,
  1,1,1,1,1,1,1,1,
  1,1,1,1,1,1,1,1,
  1,1,1,1,1,1,1,1,
  0,0,0,0,0,0,0,0,
  0,0,0,0,0,0,0,0,
  0,0,0,0,0,0,0,0,
  0,0,0,0,0,0,0,0,
  0,0,0,0,0,0,0,0,
  0,0,0,0,0,0,0,0,
  0,0,0,0,0,0,0,0,
  0,0,0,0,0,0,0,0,
  0,0,0,0,0,0,0,0,
  0,0,0,0,0,0,0,0,
  0,0,0,0,0,0,0,0,
  0,0,0,0,0,0,0,0,
  0,0,0,0,0,0,0,0,
  0,0,0,0,0,0,0,0,
  0,0,0,0,0,0,0,0,
  0,0,0,0,0,0,0,0
};

uint32_t clz(uint32_t val)
{
  uint32_t accum = 0;

  accum += clzlut[val >> 24];
  accum += (accum == 8 ) ? clzlut[(val >> 16) & 0xFF] : 0;
  accum += (accum == 16) ? clzlut[(val >>  8) & 0xFF] : 0;
  accum += (accum == 24) ? clzlut[ val        & 0xFF] : 0;

  return accum;
}

How do you generate this lookup table? What is the algorithm (in C or JavaScript or the like), for generating this lookup table? I am going to use it for the "count leading zeroes" (clz) implementation. I know there is builtins for clz, I just want to know how to generate a fallback, and for curiosity/learning's sake.


Solution

  • How do you generate this lookup table?
    ...to know how to generate a fallback, and for curiosity/learning's sake.

    For each index 0 to 255: 
      Start at count = 8. (Bit width of uint8_t)
      Copy index to i.
      while i > 0: 
        Divide i by 2.
        Decrement count.
      clzlut[index] = count