Module fmax 

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 This has been superseded by IEEE 754-2019 maximumNumber.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

Functions

fmax