From owner-freebsd-ports@FreeBSD.ORG Fri Apr 4 11:33:02 2003 Return-Path: Delivered-To: freebsd-ports@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 6E9E937B401 for ; Fri, 4 Apr 2003 11:33:02 -0800 (PST) Received: from rwcrmhc53.attbi.com (rwcrmhc53.attbi.com [204.127.198.39]) by mx1.FreeBSD.org (Postfix) with ESMTP id C92FE43FA3 for ; Fri, 4 Apr 2003 11:33:01 -0800 (PST) (envelope-from davep.freebsd@meduseld.net) Received: from baloo.meduseld.net (h00095b0c686b.ne.client2.attbi.com[66.30.120.27]) by rwcrmhc53.attbi.com (rwcrmhc53) with ESMTP id <20030404193301053002li8de>; Fri, 4 Apr 2003 19:33:01 +0000 Received: from localhost (localhost [127.0.0.1]) by baloo.meduseld.net (8.12.6/8.12.6) with ESMTP id h34JWV7U071509 for ; Fri, 4 Apr 2003 14:32:34 -0500 (EST) (envelope-from davep.freebsd@meduseld.net) Date: Fri, 04 Apr 2003 14:32:31 -0500 (EST) Message-Id: <20030404.143231.466660463.davep@meduseld.net> To: freebsd-ports@freebsd.org From: "David A. Panariti" X-Attribution: davep X-Mailer: Mew version 4.0.51 on XEmacs 21.4.12 (Portable Code) Mime-Version: 1.0 Content-Type: Multipart/Mixed; boundary="--Next_Part(Fri_Apr__4_14:32:31_2003_332)--" Content-Transfer-Encoding: 7bit Subject: apr-0.9.2 and apr-util-0.9.2 break w/symlinked /usr/local X-BeenThere: freebsd-ports@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Porting software to FreeBSD List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 04 Apr 2003 19:33:02 -0000 ----Next_Part(Fri_Apr__4_14:32:31_2003_332)-- Content-Type: Text/Plain; charset=us-ascii Content-Transfer-Encoding: 7bit I searched the mailing lists and couldn't find any references to this problem. I have /usr/local on another slice and a symlink to it, and this is what causes my problem. Inside ap[ur]-config, pathnames are compared to determine in what state the command is running: installed, source or build. The problem is that `realpath' is used on only one path variable used to make the determination, `thisdir' but not `tmpbindir'. Because of this, ap[ur]-config would mistakenly think I was in the `build' state and emit incorrect include and lib locations. The attached patches correct this. I've provided patches to the .in files whereas for testing I originally patched the scripts in /usr/local/bin, so apologies if something went awry. I'm not subscribed, so any responses should email me directly. regards, davep -- The Heineken Uncertainty Principle: You can never be sure how many beers you had last night. ----Next_Part(Fri_Apr__4_14:32:31_2003_332)-- Content-Type: Text/Plain; charset=us-ascii Content-Transfer-Encoding: 7bit Content-Disposition: inline; filename="apr-config.patch" --- apr-config.in.ORIG Fri Apr 4 14:11:48 2003 +++ apr-config.in Fri Apr 4 14:12:39 2003 @@ -131,6 +131,9 @@ fi if test -d $bindir; then tmpbindir="`cd $bindir && pwd`" + if test -x "`which realpath 2>/dev/null`"; then + tmpbindir="`realpath $tmpbindir`" + fi else tmpbindir="" fi ----Next_Part(Fri_Apr__4_14:32:31_2003_332)-- Content-Type: Text/Plain; charset=us-ascii Content-Transfer-Encoding: 7bit Content-Disposition: inline; filename="apu-config.patch" --- apu-config.in.ORIG Fri Apr 4 14:13:51 2003 +++ apu-config.in Fri Apr 4 14:14:27 2003 @@ -119,6 +119,9 @@ fi if test -d $bindir; then tmpbindir="`cd $bindir && pwd`" + if test -x "`which realpath 2>/dev/null`"; then + tmpbindir="`realpath $tmpbindir`" + fi else tmpbindir="" fi ----Next_Part(Fri_Apr__4_14:32:31_2003_332)----