From owner-freebsd-testing@FreeBSD.ORG Sun May 11 15:32:07 2014 Return-Path: Delivered-To: freebsd-testing@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [8.8.178.115]) (using TLSv1 with cipher ADH-AES256-SHA (256/256 bits)) (No client certificate requested) by hub.freebsd.org (Postfix) with ESMTPS id 213F3E17; Sun, 11 May 2014 15:32:07 +0000 (UTC) Received: from mx1.stack.nl (relay04.stack.nl [IPv6:2001:610:1108:5010::107]) (using TLSv1 with cipher DHE-RSA-CAMELLIA256-SHA (256/256 bits)) (Client CN "mailhost.stack.nl", Issuer "CA Cert Signing Authority" (not verified)) by mx1.freebsd.org (Postfix) with ESMTPS id D27432C81; Sun, 11 May 2014 15:32:06 +0000 (UTC) Received: from snail.stack.nl (snail.stack.nl [IPv6:2001:610:1108:5010::131]) by mx1.stack.nl (Postfix) with ESMTP id DFE4CB808F; Sun, 11 May 2014 17:32:03 +0200 (CEST) Received: by snail.stack.nl (Postfix, from userid 1677) id C107728497; Sun, 11 May 2014 17:32:03 +0200 (CEST) Date: Sun, 11 May 2014 17:32:03 +0200 From: Jilles Tjoelker To: Garrett Cooper Subject: Re: [PATCH v2] convert /bin/sh tests over to ATF Message-ID: <20140511153203.GA81002@stack.nl> References: <20140124162759.GB90996@stack.nl> <0A68C9FE-6A3A-4DCF-92A9-A506D7581D69@gmail.com> <96373EF8-77EA-4AA6-8422-63881C32E455@gmail.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.5.21 (2010-09-15) Cc: "freebsd-testing@freebsd.org" X-BeenThere: freebsd-testing@freebsd.org X-Mailman-Version: 2.1.18 Precedence: list List-Id: Testing on FreeBSD List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sun, 11 May 2014 15:32:07 -0000 On Mon, Apr 21, 2014 at 05:21:22PM -0700, Garrett Cooper wrote: > On Sun, Mar 9, 2014 at 1:42 PM, Julio Merino wrote: > ... > > The tests need to populate /usr/tests/bin/sh/ with the > > functional_test.sh driver and all its supporting data files. As I > > understand it, your driver code assumes that the data files are > > external (i.e. not bundled into the script itself), so those files > > must be available from the same location as where the driver lives. > > Hence you still need to use FILES to ensure the data files get > > installed into /usr/tests/. > > ... but maybe I'm missing something from your patch; haven't reviewed > > it again to double-check this. > I've reduced the diff quite a bit. The end result of it can be > found here: https://github.com/yaneurabeya/freebsd/pull/2 (and it will > evolve over time if needed because github is additive in pull request > reviews). FWIW, it might be better to actually review it on github > instead of using this list as patches attached to any mailing list can > become stale over time. > You can get the diff like so: > https://github.com/yaneurabeya/freebsd/pull/2.patch This is OK with me as long as important use cases for me are not broken: 1. Allow running new tests on a potentially broken new sh without installing anything. This used to be possible using 'make regress' in src/bin/sh/ but this was removed. I now use SH=$PWD/obj/sh sh tests/legacy_test.sh (I have an obj/ subdirectory to make the built objects more easily accessible). This should also work on stable/9 although it is acceptable to need a few local hacks for that (as is already the case for MK_TESTS and src.opts.mk). On head, it should work without local hacks. 2. Allow installing using a read-only /usr/obj (mounted over NFS). This means that everything needs to be built during buildworld. Installworld should not build anything. (I currently have local hacks: Kyuafile.auto is not built so I added all: Kyuafile.auto to some tests/Makefiles; I guess this should be in .) -- Jilles Tjoelker