From: Junio C Hamano <gitster@pobox.com>
Date: Wed, 02 Jul 2014 20:05:14 GMT
Subject: Re: [PATCH] t7300: repair filesystem permissions with test_when_finished
Message-ID: <xmqq7g3vh6h1.fsf@gitster.dls.corp.google.com>
In-Reply-To: <20140702190331.GA9869@sigill.intra.peff.net>

Jeff King <peff@peff.net> writes:

> On Wed, Jul 02, 2014 at 12:01:59PM -0700, Jonathan Nieder wrote:
>
>> Jeff King wrote:
>> 
>> > Signed-off-by: Jeff King <peff@peff.net>
>> > ---
>> >  t/t7300-clean.sh | 4 ++--
>> >  1 file changed, 2 insertions(+), 2 deletions(-)
>> 
>> Does the later "git clean -d with an unreadable empty directory" test
>> need the same treatment?
>
> I don't think so, because it is also failing during the experiments I'm
> doing and it is not causing the same problem.

It tests that "clean -d" is happy if a blind rmdir(2) removes the
directory.  If it fails for whatever reason (e.g. add "exit(0)" at
the beginning of cmd_clean(), for example) to remove the directory,
we do leave an empty unreadable directory behind.

But as long as that directory is empty, that will not cause us to
remove the trash directory at the end, so we should be OK.