Doubles Drop the 's'

Floating Point

floating-pointdouble-precisiontypes

`f64` is double precision

A f32 is single precision; an f64 (double) is double precision, and the hardware has a parallel set of instructions without the trailing s. The same a + b, typed as f64, compiles to fadd instead of fadds:

fadd f1, f1, f2    # f1 = a + b, double precision
blr

So fadd/fmul/fsub/fdiv are the double-precision forms, and fadds/fmuls/fsubs/fdivs are single. The presence or absence of that one letter tells you the operand type — which is exactly the kind of clue you use to recover the original C declarations.

Your task

Write add_d to compile to the fadd above.

Hints

match add_dmwcceppc.exe -O4,p

Loading editor…

Hit “Compile & Check” to diff your code against the target.