Date: Mon, 2 Oct 2000 09:34:27 +0200 (MET DST) Message-Id: <200010020734.JAA25254@bashir.belgium.eu.net> User-Agent: Pan/0.8.1beta5 (Unix) From: "Tim Van Holder" To: djgpp AT delorie DOT com Subject: Re: Building gdb 5.0 Reply-To: djgpp AT delorie DOT com In a burst of inspiration, "386sx" <386sx AT my-deja DOT com> wrote this on <8FBE61DCC386sxmydejacom AT 130 DOT 133 DOT 1 DOT 4>: > I'm trying to compile gdb 5.0 from the sources downloaded from gnu and I > unpacked everything with the djunpack.bat file in the distribution. I'm > using Windows 95. > > gdb-5.0$ sh ./gdb/config/djgpp/djconfig.sh Checking the unpacked > distribution... ok. Editing configure scripts for DJGPP... > > It edits all of the scripts and everything goes o.k until... > > Running the configure script... Configuring for a i386-pc-msdosdjgpp > host. Created "Makefile" in /usr/tmp/gdb-5.0 using "mh-frag" Configuring > intl... > /usr/tmp/gdb-5.0/configure: cd: intl: No such file or directory (ENOENT) > > And it goes into recursion at this point. The directory is there, so > what could the problem be? Thanks. From the /usr/tmp/gdb-5.0/configure I am assuming you were running it from a linux prompt. If that is the case, the cause will most likely be stray CR's in the configure script (linux' bash can't handle them, as I found out when testing my patched autoconf 2.49a[1] on Linux). If this is the case, Eli can rest easy and concentrate on Emacs :-) Tim Van Holder [1] BTW, this will likely be released next weekend (and the patches submitted to the autoconf team), pending tests on a system other than my own. Automake 1.4a still fails 6 or so tests in the testsuite (some of which my just be a symptom of it being an alpha release); but I'll try to address those and release it together with autoconf.