On Sun, Feb 29, 2004 at 02:41:05PM +1100, Martin Pool wrote:
> On 19 Feb 2004, Albert Chin <[EMAIL PROTECTED]> wrote:
> > Tru64 UNIX was failing the ParseMask_Case test until I applied the
> > patch below.
> >
> > --
> > albert chin ([EMAIL PROTECTED])
> >
> > -- snip snip
> > --- src/access.c.orig 2004-02-19 02:14:33.000000000 -0600
> > +++ src/access.c 2004-02-19 02:14:43.000000000 -0600
> > @@ -95,11 +95,11 @@
> > * is undefined. (And indeed it fails on i386 because the SHR does
> > * not saturate.) */
> > if (mask_bits == 32)
> > - *mask = (in_addr_t) ~0UL;
> > + *mask = (in_addr_t) ~0U;
> > else
> > - *mask = htonl(~(~0UL >> mask_bits));
> > + *mask = htonl(~(~0U >> mask_bits));
> > } else {
> > - *mask = (in_addr_t) ~0UL;
> > + *mask = (in_addr_t) ~0U;
> > }
> > return 0;
> > }
>
> Sorry, I don't understand properly why this is needed.
>
> I suppose long on tru64 is 64 bits. Won't truncation just work when
> it's cast into in_addr_t?
sizeof(long) == 8
in_addr_t is 'unsigned int' and sizeof(int) == 4.
--
albert chin ([EMAIL PROTECTED])
__
distcc mailing list http://distcc.samba.org/
To unsubscribe or change options:
http://lists.samba.org/mailman/listinfo/distcc
