Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 29 Oct 2013 13:36:58 -0700
From:      "Simon J. Gerraty" <sjg@juniper.net>
To:        Julio Merino <julio@meroh.net>
Cc:        freebsd-testing@freebsd.org, Rui Paulo <rpaulo@fnop.net>, sjg@juniper.net
Subject:   Re: Plugging ATF tests into the build and other cleanups
Message-ID:  <20131029203658.75BF35807E@chaos.jnpr.net>
In-Reply-To: <CADyfeQU7Y8APwTMDo9aTR2NUi2EBq0ytQ3QcF7Ct3xC7_BatBQ@mail.gmail.com>
References:  <CADyfeQU7Y8APwTMDo9aTR2NUi2EBq0ytQ3QcF7Ct3xC7_BatBQ@mail.gmail.com>

next in thread | previous in thread | raw e-mail | index | archive | help

On Sun, 27 Oct 2013 18:12:12 -0400, Julio Merino writes:
>The one concern I have here is having to keep track of all tests in
>tools/build/mk/OptionalObsoleteFiles.inc so that setting
>WITHOUT_TESTS=no cleans up /usr/tests. This will be a pain to maintain

Yes.

>and a sure source of inconsistencies. If we could special-case this to
>make it more automatic, do you have any suggestions?

Iff ATF/Kyua were the only thing populating /usr/tests/ you could just
rm -rf that dir, but that's not how delete-old-* works else

KYUAFILES!= cd $DESTDIR/ && find usr/tests -name Kyuafile
OLD_DIRS+= ${KYUAFILES:H} 

would be all you'd need.
You can't even do OLD_FILES+= ${KYUAFILES:H}/* 

It could be that a new delete-old- target might be useful,

>As usual: http://portal.meroh.net/~jmmv/freebsd-testing/

remove-without-atf.diff 

looks ok - assuming folk are ok with s,ATF,TESTS,

freebsd-testing/recurse-subdir.diff 

I'm not so crazy about.  
What's wrong with requiring folk to set TESTS_SUBDIR[S] ?

freebsd-testing/move-kyuafiles.diff

I'm also not crazy about this one (the change to lib/Makefile)
What is the expected use? and why can't everything that is needed be
done within the context of the individual lib/*/tests/ dir?
That's how we are doing it.

build-atf-tests.diff

FWIW I prefer ${.CURDIR:H:H} etc rather than ${.CURDIR}/../..
makes for much neater paths (also bit quicker on nfs)

It would be nice to be able to rely on SRCTOP being correctly set
(easy with bmake, have to think about fmake...)

ATF_SRC = ${SRCTOP}/contrib/atf

could then be set on one place
but otherwise ok.




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20131029203658.75BF35807E>