Re: Type casts

JJ> pointer can be returned. So the cast to (void *) in XPNTR is JJ> superfluous. Perhaps, but casts to void* might have the effect of inhibiting compiler aliasing optimizations, and shouldn't hurt if re-cast. JJ> But there, the format strings should be using %p instead of JJ> %lx, in which case the casts to (long) and (unsigned long) would be On some platforms printf implementation of "%p" includes "0x", on others leaves it out. "%lx" is a more predictable way to get consistent pointer printing. It's true that "%lx" has a builtin non-portable assumption, namely that a pointer fits into a long - but that's pervasive in the source code, and true of all current machines. And it looks like we're moving towards a LP64 world, so this assumption in the source code will continue to be true for at least another decade. JJ> unnecessary anyway. The (unsigned long) cast in internal_object_printer JJ> should be (long), anyway, to match the format string. Perhaps the format should be "%ulx", if that is legal. In practice I don't think it matters. Why have I been able to use gcc 2.95.3 to compile xemacs for years on Suns without any problem? (of course, avoiding -mcpu=ultrasparc)