Hi,
Neil Jerram <[EMAIL PROTECTED]> writes:
> I wonder if that's because it's generally concerned about
> uninitialized variables, or because it's noticed the apparent bug in
> scm_c_make_socket_address, which does nothing with its address_size
> parameter and instead has another size_t variable on the stack.
It's a bug in `scm_c_make_socket_address ()'. I apologize for this.
> Can you try changing scm_c_make_socket_address to this:
>
> {
> struct sockaddr *soka;
>
> soka = scm_fill_sockaddr (scm_to_ushort (family), address, &args, 1,
> "scm_c_make_socket_address", &address_size);
^^
Here it should read `address_size', not `&address_size'.
Thanks,
Ludovic.
_______________________________________________
Bug-guile mailing list
[email protected]
http://lists.gnu.org/mailman/listinfo/bug-guile
