On Monday 7 July 2003 7:49 pm, Alexandre Oliva wrote: > > Assuming it doesn't exist already, I think sizeof(function) or something > > like it would be a very useful GNU Extension for future GCC versions :) > > Can't really be done, at least not without surprising semantics. > IIRC, sizeof already has a well-defined meaning when the argument is a > function: the function decays to pointer, and you get the size of the > pointer. That's why I said "or something like it". > Besides, sizeof is supposed to return a compile-time > constant, but the size of a function is most often only known at > assembly time, or, worse, at link time, in case linker relaxations are > available. And don't even get me started on ELF dynamic library > symbol binding and replacement rules :-) Well, the same mechanism that resolves references might be able to resolve function sizes too, if the linker is built to handle this. Just a thought :) Ben
