4 ms·

Because it includes testcases.

• therein 10y ago
  You're absolutely 100% correct but to play the devil's advocate, what if he is talking about it being glibc: > return ( j >= 0 ) ? j : -j; Musl: > return a>0 ? a : -a;
  • reidrac 10y ago
    It really doesn't matter, the compiler will generate the same code (gcc -S at least does generate the same assembler output for both pieces of code; see: https://gist.github.com/reidrac/afb89b29228a243954bd51281c39ee88 https://gist.github.com/reidrac/afb89b29228a243954bd51281c39...). I guess looking at the definition: https://en.wikipedia.org/wiki/Absolute_value#Real_numbers https://en.wikipedia.org/wiki/Absolute_value#Real_numbers I would say `( j >= 0 ) ? j : -j` is more accurate, although in practical terms is the same result.
    • PDoyle 10y ago
      They'd differ for floating-point NaNs.
      • vidarh 10y ago
        But the input types here are int's.