Re: Error cross-compiling lib/libc/rpc/svc_vc.c

Tom Ivar Helbekkmo wrote:
> Johnny Billquist <johnny.billquist@xxxxxxxxxx> writes:
>
> > Ok. Here is a patch for the error you reported, Tom...
>
> That way of doing it is cheating, you know!  :-)

Yes, I know. :-)

> > I don't know if this should be in the repository or not.
>
> Not, I think.  It just looks too wrong.  ;-)  What I am thinking, by
> now, is that it is impossible to keep the cast from dropping the
> "__attribute__((__noreturn__))" with the version of GCC that we're
> using on the VAX.  If I am right, it seems that the two only ways to
> handle this is either to have a "#define __attribute__(x)" in some
> suitable header file, #ifdef'ed for the VAX, or to tweak the warning
> levels so this particular warning is turned off on the VAX, and on any
> other port that uses an equally old GCC.
>
> I'd say the latter -- the former disables other uses of __attribute__().

Well, I tried changing it at the source ( include/stdlib.h ) and even 
removing the __attribute__ there didn't improve things. And I can't find 
where else abort() could be defined, so I'm at loss on figuring this out.

But for my purposes, my fix will do. I want it to compile, and obviously 
the VAX keeps getting less and less interest, so I'm not holding my 
breath for it to get fixed any other way.
I suspect before long, the VAX won't even be supported any more. It's 
been several years since you last could run build.sh from start to 
finish on a native build on a VAX. And all the new stuff they keep 
adding just means it's getting slower and slower, for almost no return 
(in my view). Building on a VAX 8650 now takes over a week for just 
/usr/src.

        Johnny

--
Johnny Billquist                  || "I'm on a bus
                                  ||  on a psychedelic trip
email: bqt@xxxxxxxxxxxx           ||  Reading murder books
pdp is alive!                     ||  tryin' to stay hip" - B. Idol