calling convention for C99 bool?

[binarybana]

Whereas it is mentioned in the documentation. I'm not sure whether this is a documentation bug or if Cbool is supposed to be defined. If the former, then what is the recommended way to get the implementation specific size of Bool?

[JeffBezanson]

I think we should remove Cbool since bool is not a standard C type.
Julia Bool is 1 byte; sizeof will tell you.

[stevengj]

bool is a standard C type (defined in <stdbool.h>) as of C99.

[binarybana]

Not sure if this is relevant here, but in my case I'm trying to call into a C++ function with a bool argument. From what I can tell, the size of C99 bool does not necessarily equal a C++ bool, and the behavior of extern "C" is implementation specific. But trying sizeof(Cbool) was my first guess.

[StefanKarpinski]

The difference between C99 and C++ bool makes me think we should maybe not have this.

[pao]

C99bool vs. Cppbool?

[tknopp]

Are there compilers/platforms (C++/C11) where sizeof(bool)!=1?

[binarybana]

According to this (p. 6) it is 2 for 16bit Microsoft, and according to msdn it was 4 in Visual C++ 4.2.

[nolta]

VC++ 4.2 is, what, 17 years old?

[tknopp]

:-) Ok I provoked this answer. My impression was that the Julia FFI makes certain assumptions anyway (e.g. sizeof(int) == 4) which are not required by the C or C++ standard. Thus if all currently by Julia supported platforms have sizeof(bool)==1, it would be safe to just make this assumption.

[nstiurca]

Sorry to necromance this, but I am working on wrapping a C library which uses bool as defined by <stdbool.h> in C99.

Although I understand that a lot of C codes hack their way around bool in ad hoc ways, it seems very strange that Julia buries its head in the sand on the issue. What is the recommended way to handle C99 bool in Julia FFI? Can we add such a recommendation to the Julia docs perhaps?