I can try to blogging my migrating to CCES from VDSP++.
I have VDSP++ project which use C++.
Now I did the superfast interrupt handlers in CCES my peripheries work well (about this I blog later).
When I started my project I noticed (https://ez.analog.com/thread/9912) that fract C win class fract C++ so I decided use built-in functions because I want to take fract support in C++ with maximum performance.
At that time I noticed next:
int x = 0.5r; // x = 0x40000000
and I started it to use I'm my project.
I declared the variable as integer but I used built-in functions so I got fract support in C++.
But when I am migrating to CCES from VDSP++ I want to start my project with minimum modifications
must know the paragraph "Data Type Conversions and Fixed-Point Types".
So now I must use Bit-Pattern Conversion Functions: bitsfx and fxbits:
int x = bitsr(0.5r); // x = 0x40000000