On Sat, Jun 02, 2001 at 09:17:09PM +0200, stef wrote:
> On Sat, Jun 02, 2001 at 07:57:36PM +0100, Tim Waugh wrote:
> > On architectures that don't have insb, outsb, and the rest (basically
> > everything but i386), the umax_pp backend doesn't build at all. I
> > think they should be autoconf'd for.
> >
>
> In fact, it is:
>
> .........
> static void
> Outsb (int port, unsigned char *source, int size)
> {
> #ifndef IO_SUPPORT_MISSING
> outsb (port, source, size);
> #endif
> }
> ..........
>
> > ../backend/umax_pp_low.o: In function `Insb':
> > ../backend/umax_pp_low.o(.text+0xe92): undefined reference to `insb'
> > ../backend/umax_pp_low.o: In function `Outsb':
> > ../backend/umax_pp_low.o(.text+0xee2): undefined reference to `outsb'
> > ../backend/umax_pp_low.o: In function `Insw':
> > ../backend/umax_pp_low.o(.text+0xf32): undefined reference to `insl'
> > ../backend/umax_pp_low.o: In function `Outsw':
> > ../backend/umax_pp_low.o(.text+0xf82): undefined reference to `outsw'
> > collect2: ld returned 1 exit status
> > make[1]: *** [umax_pp] Error 1
> >
> > Tim.
> > */
>
> Were inb/outb also undefined ? Or was it simply the string version of these ?
>
> So either IO_SUPPORT_MISSING was mis calculated, or your gcc doesn't have
> string I/O.
>
> Regards,
> Stef
>
> --
Yiiks, me silly, haven't took the time to read carefully.... You mean there
should be something like:
static void
Outsb (int port, unsigned char *source, int size)
{
#ifndef IO_SUPPORT_MISSING
#ifdef HAVE_STRING_IO
outsb (port, source, size);
#endif
for(i=0,.....)
outb(..)
#else
#endif
}
Sorry,
Stef
-- Source code, list archive, and docs: http://www.mostang.com/sane/ To unsubscribe: echo unsubscribe sane-devel | mail majordomo@mostang.com
This archive was generated by hypermail 2b29 : Sat Jun 02 2001 - 12:14:24 PDT