AP_Math: correct warning on fabsF

[peterbarker]

This is a bit weird, actually.

Surely this is true of the trig functions in here too. When we explicitly instantiate Vector3<double> and we're using 32-bit maths we'll use the float versions of the trig functions when determining angles and whatnot for that instantiation?

2024-08-08T01:51:53.6780446Z ../../libraries/AP_Math/vector3.cpp:432:9: warning: absolute value function 'fabsf' given an argument of type 'const double' but has parameter of type 'float' which may cause truncation of value [-Wabsolute-value]
2024-08-08T01:51:53.6781336Z     if (fabsF(cosv) >= 1) {
2024-08-08T01:51:53.6781583Z         ^
2024-08-08T01:51:53.6781930Z ../../libraries/AP_Math/ftype.h:50:18: note: expanded from macro 'fabsF'
2024-08-08T01:51:53.6782342Z #define fabsF(x) fabsf(x)
2024-08-08T01:51:53.6782572Z                  ^
2024-08-08T01:51:53.6789178Z ../../libraries/AP_Math/vector3.cpp:633:16: note: in instantiation of member function 'Vector3<double>::angle' requested here
2024-08-08T01:51:53.6789800Z template class Vector3<double>;