On Tue, Aug 17, 2004 at 06:29:34PM -0400, Allan Edwards wrote: > I'm wondering why gcc doesn't do the same since > int=4 ptr=8 bytes for both platforms, is gcc just not > as strict? And how to proceed, do we start fixing > these warnings for Win64 even though they may not > represent real life concerns in order to achieve a > warning free /W3 compile for Win 64?
w.r.t. common code, depends what the changes are; if there are places where an int is used to store the difference between pointers, it's probably a good and safe change to make that an apr_ssize_t instead. w.r.t. warnings: yes, gcc is just less picky. But Win64 has a 32-bit long, right? So there is a significant difference between Win64 and 64-bit Unix platforms which are all (AFAIK) LP64. > - rv = collapse_iovec((char**) &ptfb->Tail, > &ptfb->TailLength, > + rv = collapse_iovec((char**) &ptfb->Tail, (apr_size_t > *)&ptfb->TailLength, if that's casting (int *) to (size_t *) that looks dangerous. joe
