>> Should be: for integer and pointer types there are _no_ "trap >> representations". > > I don't think that is guaranteed by C99.
That's correct. As I recall, only unsigned char objects are guaranteed to be free of trap representations. > Certainly the practical value of this for gnulib is questionable. I've heard that some C implementations do support trap representations of pointers in hardware, and that some debugging C implementations also have trap representations for integers. See, for example <http://www-128.ibm.com/developerworks/power/library/pa-ctypes3/>. I've never knowingly used such implementations, thought I wouldn't be surprised if I used one without knowing it.
