unit testing - ldexp or hex floats in C++/CLI -
i'm writing unit tests, using visual studio unit testing framework, expects me utilize managed c++.
in test, want test accuracy , error recovery of floating point helper function, i'd load floating constants specifying mantissa , exponent. in normal c++, i'd using ldexp
, cannot include <math.h>
, or <cmath>
managed code.
is there way load (possibly denormalized) floating point constant?
fwiw, solved relaxing build flags /clr:safe
/clr
.
unit-testing floating-point c++-cli
No comments:
Post a Comment