From: Junio C Hamano <gitster@pobox.com>
To: "Sverre Rabbelier" <srabbelier@gmail.com>
Cc: "Jakub Narebski" <jnareb@gmail.com>, "Git list" <git@vger.kernel.org>
Subject: Re: [PATCH 1/3] Modified test-lib.sh to output stats to /tmp/git-test-results
Date: Sun, 08 Jun 2008 11:53:37 -0700 [thread overview]
Message-ID: <7vabhvkbbi.fsf@gitster.siamese.dyndns.org> (raw)
In-Reply-To: bd6139dc0806080745i2ff2489bv70a1596bcc83f700@mail.gmail.com
"Sverre Rabbelier" <srabbelier@gmail.com> writes:
> On Sun, Jun 8, 2008 at 4:42 PM, Jakub Narebski <jnareb@gmail.com> wrote:
>> "Sverre Rabbelier" <sverre@rabbelier.nl> writes:
>>
>>> Because writing to the current directory is not possible, we write to /tmp/.
>>> Suggestions for a better location are welcome.
>>
>>> + test_results_path="../test-results"
>>
>> Errr... it looks like you forgot to update commit message.
>
> Yeah, I intended to fix that when amending the original commit, but I
> forgot. I thought I couldn't write to the current dir, but instead I
> was writing to /t/trash, which of course gets deleted after every
> test. Obviously outputting to ../test-results fixed that.
Some tests may chdir around and when you hit test_done they may not be in
the directory directly under t/ anymore. I'd say that such a test is
ill-mannered, but we would want to protect ourselves against such mischief
somehow.
Also, this won't be an immediate issue, but we may want to lift the
".NOTPARALLEL" limitation in t/Makefile sometime in the future. A handful
tests need to be adjusted for this change, as they currently refer their
test vectors in t/tXXXX/ as ../tXXXX, and with such a change, we would run
each test in "t/trash directory/t$$" where $$ is the pid, or something
like that. A single "test-results" file everybody races to append would
become unwieldy at that point. Something to keep in mind.
next prev parent reply other threads:[~2008-06-08 18:54 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2008-06-08 14:04 [PATCH 1/3] Modified test-lib.sh to output stats to /tmp/git-test-results Sverre Rabbelier
2008-06-08 14:04 ` [PATCH 2/3] A simple script to parse the results from the testcases Sverre Rabbelier
2008-06-08 14:04 ` [PATCH 3/3] Hook up the result aggregation in the test makefile Sverre Rabbelier
2008-06-08 14:42 ` [PATCH 1/3] Modified test-lib.sh to output stats to /tmp/git-test-results Jakub Narebski
2008-06-08 14:45 ` Sverre Rabbelier
2008-06-08 18:53 ` Junio C Hamano [this message]
2008-06-08 19:02 ` Sverre Rabbelier
2008-06-08 20:59 ` Junio C Hamano
2008-06-09 0:44 ` [PATCH 4/4] Make test-lib warn if the directory changes Sverre Rabbelier
-- strict thread matches above, loose matches on Subject: below --
2008-05-12 9:33 [PATCH 0/3] Aggregate testcase results Sverre Rabbelier
2008-05-12 9:33 ` [PATCH 1/3] Modified test-lib.sh to output stats to /tmp/git-test-results Sverre Rabbelier
2008-05-12 14:55 ` Vegard Nossum
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=7vabhvkbbi.fsf@gitster.siamese.dyndns.org \
--to=gitster@pobox.com \
--cc=git@vger.kernel.org \
--cc=jnareb@gmail.com \
--cc=srabbelier@gmail.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).