From xemacs-m  Wed Jul  2 19:59:39 1997
Received: from mercury.Sun.COM (mercury.Sun.COM [192.9.25.1])
	by xemacs.org (8.8.5/8.8.5) with SMTP id TAA12213
	for <xemacs-beta@xemacs.org>; Wed, 2 Jul 1997 19:59:39 -0500 (CDT)
Received: from Eng.Sun.COM ([129.146.1.25]) by mercury.Sun.COM (SMI-8.6/mail.byaddr) with SMTP id SAA11455; Wed, 2 Jul 1997 18:24:15 -0700
Received: from kindra.eng.sun.com by Eng.Sun.COM (SMI-8.6/SMI-5.3)
	id RAA11337; Wed, 2 Jul 1997 17:59:09 -0700
Received: from xemacs.eng.sun.com by kindra.eng.sun.com (SMI-8.6/SMI-SVR4)
	id RAA18014; Wed, 2 Jul 1997 17:59:07 -0700
Received: by xemacs.eng.sun.com (SMI-8.6/SMI-SVR4)
	id RAA01329; Wed, 2 Jul 1997 17:59:03 -0700
Date: Wed, 2 Jul 1997 17:59:03 -0700
Message-Id: <199707030059.RAA01329@xemacs.eng.sun.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
From: Martin Buchholz <mrb@Eng.Sun.COM>
To: Anthony Rossini <rossini@stat.sc.edu>
Cc: XEmacs Beta List <xemacs-beta@xemacs.org>, kwzh@gnu.ai.mit.edu,
        bug-gcc@prep.ai.mit.edu
Subject: B10: configuration weirdness
In-Reply-To: <wncg1u05em3.fsf@fhat.stat.sc.edu>
References: <wncg1u05em3.fsf@fhat.stat.sc.edu>
X-Mailer: VM 6.32 under 20.3 "Sofia" XEmacs  Lucid (beta9)
Reply-To: Martin Buchholz <mrb@Eng.Sun.COM>

>>>>> "AR" == Anthony Rossini <rossini@stat.sc.edu> writes:

AR> As I mentioned in my last email, the combination:
AR> 	Sparc Ultra 1, Solaris 2.5, gcc, R6.3, gave me a failures, due 
AR> 	to a problem with tv_sec.   This still happens with beta-10.

AR> Same problems with Beta 10.

AR> However, I tried Sun's CC (V 4.0), and it seems to be running, now...
AR> (i.e. I configured with:

I believe your gcc is misconfigured.  Unfortunately, because gcc tries 
to `fix' the system header files, the resulting binary is bound to the 
specific version of Solaris it was built on.  You have to have
separate gcc binaries for 2.4, 2.5, 2.5.1, and 2.6 (I actually built
all of these recently, to solve this very problem).

Try rebuilding gcc, and using that gcc to build xemacs.

This comment applies to all configurations of gcc on various OSes.
You are much more likely to get good results out of gcc if you build
it on the same machine where you will be using it.  This sensitivity
of gcc is really a bug that should be addressed by the gcc maintainers.

AR> configure --prefix=/p1/apps/X11R6.3 --with-gcc=no --compiler=/opt/SUNWspro/bin/cc

AR> ). The weird thing is that it wouldn't default to "cc" normally.
AR> I.e. if I don't want to use gcc, I need to say WHICH cc to use... In
AR> particular,

AR> 	 configure --prefix=/p1/apps/X11R6.3 --with-gcc=no

AR> was unhappy, as follows:

Some (as usual, undertested) changes to fix this will be in beta11.

Martin

