compiler_builtins::math::libm_math::generic

Module fmax

Source
Expand description

IEEE 754-2011 maxNum. This has been superseded by IEEE 754-2019 maximumNumber.

Per the spec, returns the canonicalized result of:

  • x if x > y
  • y if y > x
  • The other number if one is NaN
  • Otherwise, either x or y, canonicalized
  • -0.0 and +0.0 may be disregarded (unlike newer operations)

Excluded from our implementation is sNaN handling.

More on the differences: link.

Functionsยง

fmax