From xemacs-m  Sat Jan 11 02:00:22 1997
Received: from altair.xemacs.org (steve@xemacs.miranova.com [206.190.83.19])
          by xemacs.org (8.8.4/8.8.4) with ESMTP
	  id CAA26895 for <xemacs-beta@xemacs.org>; Sat, 11 Jan 1997 02:00:21 -0600 (CST)
Received: (from steve@localhost)
          by altair.xemacs.org (8.8.4/8.8.4)
	  id AAA19580; Sat, 11 Jan 1997 00:10:20 -0800
Sender: steve@xemacs.org
To: xemacs-beta@xemacs.org
Subject: Re: Errors with build of beta6 on Linux/AXP
References: <x2vi9cbgl2.fsf@lemcbed.lem.uni-karlsruhe.de>
X-Url: http://www.miranova.com/%7Esteve/
Mail-Copies-To: never
X-Face: #!T9!#9s-3o8)*uHlX{Ug[xW7E7Wr!*L46-OxqMu\xz23v|R9q}lH?cRS{rCNe^'[`^sr5"
 f8*@r4ipO6Jl!:Ccq<xoV[Qz2u8<8-+Vwf2gzJ44lf_/y9OaQ`@#Q65{U4/TC)i2`~/M&QI$X>p:9I
 OSS'2{-)-4wBnVeg0S\O4Al@)uC[pD|+
X-Attribution: sb
From: Steven L Baur <steve@miranova.com>
In-Reply-To: Jens Lautenbacher's message of 05 Jan 1997 10:45:45 +0100
Mime-Version: 1.0 (generated by tm-edit 7.100)
Content-Type: text/plain; charset=US-ASCII
Date: 11 Jan 1997 00:10:19 -0800
Message-ID: <m2g208tyxg.fsf@altair.xemacs.org>
Lines: 65
X-Mailer: Red Gnus v0.80/XEmacs 20.0

Jens Lautenbacher writes:

> % less lib-src/Makefile

> [...]
> ^L                                                          <------ LOOK
> typedef long double __long_double_t;                        <------ HERE

I don't have a clue where the formfeed is coming from.  GNU Make
doesn't have a problem with it on Intel Linux.  Do you think you could
track down which file the typedef is coming from?  I can't tell by
looking at the scripts where it might be from.  It might be part of
the failed test that is improperly detecting HAVE_MATHERR.

> <I Removed the offending lines, also in lwlib/Makefile, src/Makefile, dynodump/Makefile>
> % make

> [...]
> gcc -c -Demacs -DHAVE_CONFIG_H  -I.        -I/server3/users/jens/daten/gnu/xemacs-19.15-b6/src     -DFUNCPROTO=11 -DNARROWPROTO -D_BSD_SOURCE         -I/usr/X11R6/include             -I/server3/users/jens/daten/gnu/xemacs-19.15-b6/src/../lwlib    -g -O  floatfns.c
> In file included from config.h:411,
>                  from floatfns.c:48:
> m/alpha.h:193: warning: `UNEXEC' redefined
> s/linux.h:193: warning: this is the location of the previous definition

This is bad.  I can't tell just by looking which one it's supposed to
be.  linux.h sets it to unexelf.o, and alpha.h sets it to
unexalpha.o.  Try commenting each UNEXEC definition out in turn and
see what happens.  I'm more hopeful of the unexelf working, but I
don't know.

> floatfns.c:978: warning: `struct exception' declared inside parameter list
> floatfns.c:978: warning: its scope is only this definition or declaration,
> floatfns.c:978: warning: which is probably not what you want.

Try commenting out HAVE_MATHERR in config.h.  It does not get defined
on Intel Linux, so maybe the detection for it on alpha is bad.

> floatfns.c: In function `matherr':
> floatfns.c:987: dereferencing pointer to incomplete type
> floatfns.c:988: dereferencing pointer to incomplete type
> floatfns.c:990: dereferencing pointer to incomplete type
> floatfns.c:992: dereferencing pointer to incomplete type
> floatfns.c:994: `DOMAIN' undeclared (first use this function)
> floatfns.c:994: (Each undeclared identifier is reported only once
> floatfns.c:994: for each function it appears in.)
> floatfns.c:995: `SING' undeclared (first use this function)
> floatfns.c:996: `OVERFLOW' undeclared (first use this function)
> floatfns.c:997: `UNDERFLOW' undeclared (first use this function)

Commenting out HAVE_MATHERR will make these go away.

> make[1]: *** [floatfns.o] Error 1
> make[1]: Leaving directory `/server3/users/jens/daten/gnu/xemacs-19.15-b6/src'
> make: *** [src] Error 2



> <This was the point I gave up>

Remove the HAVE_MATHERR, and one of the UNEXECs and see how much
farther you can get. 
-- 
steve@miranova.com baur
Unsolicited commercial e-mail will be billed at $250/message.
Today's environment is 20.0 XEmacs Lucid (beta90) and Red Gnus v0.80

