From xemacs-m  Sat Dec 14 10:17:28 1996
Received: from nukkekoti.cs.hut.fi (nukkekoti.cs.hut.fi [130.233.40.128]) by xemacs.cs.uiuc.edu (8.8.3/8.8.3) with ESMTP id KAA12517 for <xemacs-beta@xemacs.org>; Sat, 14 Dec 1996 10:17:27 -0600 (CST)
Received: from torvi.cs.hut.fi (pema@torvi.cs.hut.fi [130.233.40.149])
          by nukkekoti.cs.hut.fi (8.8.4/8.8.4/1.12) with ESMTP
	  id SAA09452 for <xemacs-beta@xemacs.org>; Sat, 14 Dec 1996 18:17:28 +0200 (EET)
Received: (from pema@localhost)
          by torvi.cs.hut.fi (8.8.4/8.8.4/1.7)
	  id SAA23684; Sat, 14 Dec 1996 18:17:26 +0200 (EET)
Sender: pema@niksula.hut.fi
To: xemacs-beta@xemacs.org
Subject: Re: 20b31 hpux9.05 gcc built OK
References: <u9bubzg0z0.fsf@neal.ctd.comsat.com> 	<of0afrio6g9.fsf@pippin.ntc.nokia.com> <199612132232.OAA00645@xemacs.eng.sun.com>
From: Pekka Marjola <pema@iki.fi>
X-Face: 8>EVnZ'MSz2dI{N`dzbBND,clpk$+(a>y[d"rnI,}HfS4x!)hp+1H&ha6Rj93tVlirvJ;Z@@MSpFN[;;X)aF\mDDlk6N&0t*B^C\]|<z,S8:j26ZbN:WmvgP-DV?l|'C<(|DacsbwYal'dmuMifMOXszyo]J:k{BSASbF-R
X-Attribution: Pema
X-No-Archive: yes
Mime-Version: 1.0 (generated by tm-edit 7.95)
Content-Type: text/plain; charset=US-ASCII
Date: 14 Dec 1996 18:17:25 +0200
In-Reply-To: Martin Buchholz's message of Fri, 13 Dec 1996 14:32:11 -0800
Message-ID: <k9whglpp0bu.fsf@torvi.cs.hut.fi>
Lines: 37
X-Mailer: Red Gnus v0.74/XEmacs 19.14

"Martin" == Martin Buchholz <mrb@Eng.Sun.COM> writes:

>>>>>> "Pema" == Pekka Marjola <pema@iki.fi> writes:
Pema> Gcc (2.7.2, with cpplib IIRC) required something (-P worked :) to get
Pema> it to compile. Otherwise it failed on those DEFUN macros with comments
Pema> inside parameter lists (like buffer.c, line 296).

Martin> Are you really saying that the presence of COMMENTS inside a macro
Martin> call causes problems?  It's been this way for years.

I didn't have much time to investigate, but here's the error:
/tmp_mnt/net/frodo/share/contrib/src/xemacs-20.0-b31/src/buffer.c:448: parse error before `#'

... etc. When I changed things like:

DEFUN ("bufferp", Fbufferp, 1, 1, 0, /*
T if OBJECT is an editor buffer.
*/
       (object))

to

DEFUN ("bufferp", Fbufferp, 1, 1, 0, (object))
/*
T if OBJECT is an editor buffer.
*/

it compiled without problems. Of course, I didn't bother to do that for
all such DEFUNs, so I tried playing with cpp. -P option helped (gcc's
manual page claims it prevents generating #line statements; to be used
with -E). I cannot remember if I needed anything like that for 19.15
betas (propably not). Very strange...

-- 
Pekka Marjola * pema@iki.fi * http://www.iki.fi/pema/ * +358-40-551 4842
PGP: 0x2865206D/C46E6452206FDB42A87D44F4E0E54E65 * GCS U+++$ L+++ E++...
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Quidquid latine dictum sit, altum viditur.

