From: Jeff King <peff@peff.net>
To: Jonathan Nieder <jrnieder@gmail.com>
Cc: Jon Seymour <jon.seymour@gmail.com>,
git@vger.kernel.org, gitster@pobox.com
Subject: Re: [PATCH 1/2] tests: make test_must_fail more verbose
Date: Tue, 31 Aug 2010 14:06:18 -0400 [thread overview]
Message-ID: <20100831180618.GB15020@sigill.intra.peff.net> (raw)
In-Reply-To: <20100831171055.GO2315@burratino>
On Tue, Aug 31, 2010 at 12:10:55PM -0500, Jonathan Nieder wrote:
> > - test $? -gt 0 -a $? -le 129 -o $? -gt 192
> [...]
> Can the exit status (e.g. from a shell function) be negative?
I don't believe so. Both bash and dash complain about "return -1" (and
POSIX says the number must be "an unsigned decimal integer"). More
importantly, though, is how shells interpret an exit status from a
command that has high bits set. Both bash and dash truncate it to 8-bits
and treat it as unsigned. I would think any reasonable shell would do
that (no, I don't have a Solaris 8 box handy to test on. :) ).
> Though your patch does not affect this, a command interrupted by a
> signal will receive exit status > 192 in shells like ksh93.
Yeah, that is from 5a26973 (t/test-lib.sh: exit with small negagive int
is ok with test_must_fail, 2008-07-12). I'm not sure I agree with the
intent of that commit. We should perhaps just eliminate all the places
where we use negative exit codes.
> Posix says:
> As explained in other sections, certain exit status values
> have been reserved for special uses and should be used by
> applications only for those purposes:
>
> 126
> A file to be executed was found, but it was not an
> executable utility.
> 127
> A utility to be executed was not found.
> >128
> A command was interrupted by a signal.
Yup, that's what I have always assumed to be the case.
> Unfortunately that does not agree with git usage.
>
> 129
> Incorrect command-line usage, or help requested
> (rather than SIGHUP).
Yeah, we might consider changing that. We indeed confuse it with SIGHUP:
$ git log -h 2>/dev/null; echo $?
129
$ git log -p >/dev/null; echo $? [now killall -HUP git elsewhere]
129
We also use 128 for our generic die. I don't know if this is ever
problematic. It has the high bit set like a signal, but you don't
actually check the bits in the shell. And the actual representation at
the C level is not the same.
So part of me is in favor of changing these to something more "normal".
I am slightly worried, though, about breaking some caller. It is a
user-visible interface, though I suspect 99% of callers only ever cared
whether it was 0 or not.
> Here's a test_might_fail patch for consistency.
I didn't even know about that one. :)
They should definitely be kept in sync. I wonder if it is worth
implemening must_fail in terms of might_fail.
> test_might_fail () {
> "$@"
> - test $? -ge 0 -a $? -le 129 -o $? -gt 192
> + exit_code=$?
> + if test $exit_code -gt 129 -a $exit_code -le 192; then
> + echo >&2 "test_might_fail: died by signal: $*"
> + return 1
> + fi
> + return 0
I wonder if 2/2 would need the "missing command" bit, too. Hrm.
-Peff
next prev parent reply other threads:[~2010-08-31 18:06 UTC|newest]
Thread overview: 17+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-08-31 14:49 [PATCH 0/2] fix quotes in maint-reflog-beyond-horizon and detached-stash Jon Seymour
2010-08-31 14:49 ` [PATCH 1/2] maint-reflog-beyond-horizon: fix broken test_must_fail calls Jon Seymour
2010-08-31 14:49 ` [PATCH 2/2] detached-stash: " Jon Seymour
2010-08-31 15:54 ` [PATCH 0/2] fix quotes in maint-reflog-beyond-horizon and detached-stash Jeff King
2010-08-31 15:56 ` [PATCH 1/2] tests: make test_must_fail more verbose Jeff King
2010-08-31 17:10 ` Jonathan Nieder
2010-08-31 18:06 ` Jeff King [this message]
2010-08-31 19:35 ` Jonathan Nieder
2010-09-01 3:37 ` Jon Seymour
2010-09-01 10:55 ` Ævar Arnfjörð Bjarmason
2010-08-31 15:56 ` [PATCH 2/2] tests: make test_must_fail fail on missing commands Jeff King
2010-08-31 16:58 ` Junio C Hamano
2010-08-31 17:25 ` Jeff King
2010-08-31 17:26 ` Jonathan Nieder
2010-08-31 18:08 ` Jeff King
2010-08-31 18:14 ` Ævar Arnfjörð Bjarmason
2010-08-31 17:03 ` [PATCH 0/2] fix quotes in maint-reflog-beyond-horizon and detached-stash Junio C Hamano
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
List information: http://vger.kernel.org/majordomo-info.html
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20100831180618.GB15020@sigill.intra.peff.net \
--to=peff@peff.net \
--cc=git@vger.kernel.org \
--cc=gitster@pobox.com \
--cc=jon.seymour@gmail.com \
--cc=jrnieder@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.
Code repositories for project(s) associated with this public inbox
https://80x24.org/mirrors/git.git
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 read-only IMAP folder(s) and NNTP newsgroup(s).