On Wed, Aug 20, 2008 at 8:06 PM, Michael Ash <[EMAIL PROTECTED]> wrote: > On Wed, Aug 20, 2008 at 8:34 PM, Douglas Davidson <[EMAIL PROTECTED]> wrote: >> Well, after all, zero is zero, how much difference can it make? Quite a >> bit, as it turns out, since in 64-bit one of them is four bytes of zero, and >> the other is eight bytes of zero. If you're just comparing against NULL, it >> doesn't matter, but if you're using it in something where size counts--say, >> a list of vararg arguments--then it matters a lot. It's not easy to debug, >> though, because who would think that you need to distinguish one NULL from >> another? > > It is a little known fact that when passing NULL (and by extension nil > or Nil) as a parameter to a vararg function, you *must* cast it to the > appropriate pointer type to guarantee correct behavior. > > Interestingly, Apple's vararg methods which use nil as a terminator > (such as dictionaryWithObjectsAndKeys:) make no mention of this in > their documentation, and have a great deal of officially sanctioned > sample code which doesn't use such a cast. (And none of my code uses > it either.) I suppose Apple must be implicitly making a stronger > guarantee about the pointer-ness of nil than the C language makes > about NULL.
GCC does that for you (i.e. the NULL defined by GCC is already typed to a pointer): [EMAIL PROTECTED]:~]% cat test.c #include <stdio.h> int main() { printf("sizeof(NULL) == %zu\n", sizeof(NULL)); return 0; } [EMAIL PROTECTED]:~]% cc test.c -arch i386 && ./a.out sizeof(NULL) == 4 [EMAIL PROTECTED]:~]% cc test.c -arch x86_64 && ./a.out sizeof(NULL) == 8 -- Clark S. Cox III [EMAIL PROTECTED] _______________________________________________ Cocoa-dev mailing list (Cocoa-dev@lists.apple.com) Please do not post admin requests or moderator comments to the list. Contact the moderators at cocoa-dev-admins(at)lists.apple.com Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com This email sent to [EMAIL PROTECTED]