Re: [PATCH] Building the Hurd with gcc-4.0

[Neal H. Walfield]

At Sat, 3 Sep 2005 02:25:18 +0200,
Thomas Schwinge wrote:
> 
> On Fri, Sep 02, 2005 at 12:49:50AM +0200, Thomas Schwinge wrote:
> > Hi!
> > 
> > The attached `hurd_v.s._gcc-4.0.patch' makes the Hurd buildable with
> > gcc-4.0.
> 
> A fix (to be applied after the above patch) is attached:
> `hurd_v.s._gcc-4.0.fixes.patch'
> The patch only fixes my previous patch, so no new ChangeLog entry is
> needed.
> 
> Thanks to Neal Walfield and Peter De Schrijver.
> 
> 
> Imo, a similar fix is also needed (additionally to those that Neal
> already pointed out and Alfred fixed) for Alfred's patch to make glibc
> build with gcc-4.0:
> <URL:http://sources.redhat.com/ml/libc-alpha/2005-08/msg00018.html>
> #v+
> -       *((integer_t *) t)++ = (integer_t) arg;
> +       *(integer_t *) t = (integer_t) arg;
> +       t += sizeof (integer_t);
>  #else
> -       *((integer_t *) p)++ = (integer_t) arg;
> +       *(integer_t *) p = (integer_t) arg;
> +       p += sizeof (integer_t);
> #v-
> ... should rather be...
> #v+
> -       *((integer_t *) t)++ = (integer_t) arg;
> +       *(integer_t *) t = (integer_t) arg;
> +       t = (mach_msg_type_t *) ((integer_t *) t + 1);
>  #else
> -       *((integer_t *) p)++ = (integer_t) arg;
> +       *(integer_t *) p = (integer_t) arg;
> +       p = (void *) ((integer_t *) p + 1);
> #v-
> 
> Correct?

This looks correct.