From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 6374 invoked by alias); 13 Jan 2004 22:59:25 -0000 Mailing-List: contact gdb-help@sources.redhat.com; run by ezmlm Precedence: bulk List-Subscribe: List-Archive: List-Post: List-Help: , Sender: gdb-owner@sources.redhat.com Received: (qmail 6367 invoked from network); 13 Jan 2004 22:59:24 -0000 Received: from unknown (HELO mclean.mail.mindspring.net) (207.69.200.57) by sources.redhat.com with SMTP; 13 Jan 2004 22:59:24 -0000 Received: from user-119a90a.biz.mindspring.com ([66.149.36.10] helo=berman.michael-chastain.com) by mclean.mail.mindspring.net with esmtp (Exim 3.33 #1) id 1AgXVL-0000PY-00; Tue, 13 Jan 2004 17:59:23 -0500 Received: by berman.michael-chastain.com (Postfix, from userid 502) id 17C134B104; Tue, 13 Jan 2004 17:59:40 -0500 (EST) To: cagney@gnu.org Subject: Re: building snapshot on hp-ux with hp tools Cc: gdb@sources.redhat.com Message-Id: <20040113225940.17C134B104@berman.michael-chastain.com> Date: Tue, 13 Jan 2004 22:59:00 -0000 From: mec.gnu@mindspring.com (Michael Elizabeth Chastain) X-SW-Source: 2004-01/txt/msg00172.txt.bz2 > Historically expect and dejagnu have been bundled in with GDB to ensure > that all parties are using a consistent test framework. It might be > time to cut our losses? Gcc has an "unbundled" strategy, so every gcc tester has to install tcl+expect+dejagnu outside of gcc. It works for them; they have a lot of testing activity. Look at http://gcc.gnu.org/ml/gcc-testresults . I wish that gdb-testers looked like that. > Does a more up-to-date expect build? Errr, no. expect 5.39 does not build on hppa2.0w-hp-hpux11.11 with the hp ansi c compiler. Configure craps out with: checking whether the compiler (gcc -pipe) actually works... configure: error: the native compiler is broken and won't compile. I don't know what the gcc-testresults@ people are doing to work around that. Perhaps they always use gcc to build tcl+expect+dejagnu. If we unbundle tcl+expect+dejagnu, then we can separate those problems out. As it is, I can't tell people to try gdb snapshots on hp-ux (see pr gdb/1458, I can't reply to it with "try a snapshot"). The advantage of bundling is that, in principle, we can make things easier for people, if add value to our bundled versions. The disadvantage of bundling is that we're not spending the effort to add value so that the "+dejagnu" part of "gdb+dejagnu" just stops it from building on some systems. I'm favor of separating gdb from dejagnu. Have a separate "tcl+expect+dejagnu" tarball if you want, but no "gdb+dejagnu". Michael C