From: "Ævar Arnfjörð Bjarmason" <avarab@gmail.com>
To: Junio C Hamano <gitster@pobox.com>
Cc: git@vger.kernel.org, Vincent Lefevre <vincent@vinc17.net>,
Chris Torek <chris.torek@gmail.com>,
Denton Liu <liu.denton@gmail.com>,
"Jeff Hostetler" <jeffhost@microsoft.com>
Subject: Re: [PATCH 3/3] pager: properly log pager exit code when signalled
Date: Mon, 01 Feb 2021 20:21:20 +0100 [thread overview]
Message-ID: <8735yffvbj.fsf@evledraar.gmail.com> (raw)
In-Reply-To: <xmqqft2fmzk6.fsf@gitster.c.googlers.com>
On Mon, Feb 01 2021, Junio C Hamano wrote:
> Ævar Arnfjörð Bjarmason <avarab@gmail.com> writes:
>
>> When git invokes a pager that exits with non-zero the common case is
>> that we'll already return the correct SIGPIPE failure from git itself,
>> but the exit code logged in trace2 has always been incorrectly
>> reported[1]. Fix that and log the correct exit code in the logs.
>>
>> Since this gives us something to test outside of our recently-added
>> tests needing a !MINGW prerequisite, let's refactor the test to run on
>> MINGW and actually check for SIGPIPE outside of MINGW.
>>
>> The wait_or_whine() is only called with a true "in_signal" from from
>> finish_command_in_signal(), which in turn is only used in pager.c.
>>
>> I'm not quite sure about that BUG() case. Can we have a true in_signal
>> and not have a true WIFEXITED(status)? I haven't been able to think of
>> a test case for it.
>>
>> 1. The incorrect logging of the exit code in was seemingly copy/pasted
>> into finish_command_in_signal() in ee4512ed481 (trace2: create new
>> combined trace facility, 2019-02-22)
>>
>> Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
>> ---
>> run-command.c | 8 +++++--
>> t/t7006-pager.sh | 61 +++++++++++++++++++++++++++++++++++++++++-------
>> 2 files changed, 58 insertions(+), 11 deletions(-)
>>
>> diff --git a/run-command.c b/run-command.c
>> index ea4d0fb4b15..10e1c96c2bd 100644
>> --- a/run-command.c
>> +++ b/run-command.c
>> @@ -551,8 +551,12 @@ static int wait_or_whine(pid_t pid, const char *argv0, int in_signal)
>>
>> while ((waiting = waitpid(pid, &status, 0)) < 0 && errno == EINTR)
>> ; /* nothing */
>> - if (in_signal)
>> - return 0;
>> + if (in_signal && WIFEXITED(status))
>> + return WEXITSTATUS(status);
>> + if (in_signal) {
>> + BUG("was not expecting waitpid() status %d", status);
>> + return -1;
>> + }
>
> Doesn't BUG die, never to return control back to us? How about
> "warning()" or "error()"?
Maybe I shouldn't do that, but I'm doing it reflexively because SunCC
will yell at me otherwise. See 56f56ac50b9 (style: do not "break" in
switch() after "return", 2020-12-16).
Maybe I should just deal with its complaints, or add an "/* unreachable
*/" comment there...
next prev parent reply other threads:[~2021-02-01 19:22 UTC|newest]
Thread overview: 60+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-01-15 16:15 git fails with a broken pipe when one quits the pager Vincent Lefevre
2021-01-29 23:48 ` [PATCH] pager: exit without error on SIGPIPE Denton Liu
2021-01-30 8:29 ` Johannes Sixt
2021-01-30 12:52 ` Johannes Sixt
2021-02-01 15:03 ` Ævar Arnfjörð Bjarmason
2021-02-01 17:47 ` Junio C Hamano
2021-02-01 19:52 ` Ævar Arnfjörð Bjarmason
2021-02-01 20:55 ` Junio C Hamano
2021-02-02 2:05 ` Ævar Arnfjörð Bjarmason
2021-02-02 4:45 ` Junio C Hamano
2021-02-02 5:25 ` Junio C Hamano
2021-02-02 7:45 ` Johannes Sixt
2021-02-02 20:13 ` Junio C Hamano
2021-02-02 22:15 ` Johannes Sixt
2021-02-02 22:21 ` Junio C Hamano
2021-02-03 17:07 ` Johannes Sixt
2021-02-03 18:12 ` Junio C Hamano
2021-02-04 15:10 ` Vincent Lefevre
2021-02-03 2:45 ` Ævar Arnfjörð Bjarmason
2021-02-03 2:54 ` Junio C Hamano
2021-02-03 3:36 ` Ævar Arnfjörð Bjarmason
2021-02-03 17:19 ` Johannes Sixt
2021-01-31 1:47 ` git fails with a broken pipe when one quits the pager Ævar Arnfjörð Bjarmason
2021-01-31 3:36 ` Vincent Lefevre
2021-01-31 3:47 ` Vincent Lefevre
2021-01-31 20:49 ` Ævar Arnfjörð Bjarmason
2021-02-01 10:34 ` Vincent Lefevre
2021-02-01 11:33 ` Chris Torek
2021-02-01 12:36 ` Vincent Lefevre
2021-02-01 12:53 ` Chris Torek
2021-02-01 15:17 ` Vincent Lefevre
2021-02-01 15:00 ` Ævar Arnfjörð Bjarmason
2021-02-01 12:10 ` Ævar Arnfjörð Bjarmason
2021-02-01 14:48 ` Vincent Lefevre
2021-02-01 15:44 ` Ævar Arnfjörð Bjarmason
2021-02-01 22:16 ` Johannes Sixt
2021-02-03 2:48 ` Ævar Arnfjörð Bjarmason
2021-02-03 17:11 ` Johannes Sixt
2021-02-03 15:26 ` Vincent Lefevre
2021-02-04 0:14 ` Ævar Arnfjörð Bjarmason
2021-02-04 15:38 ` Vincent Lefevre
2021-02-01 14:49 ` [PATCH 0/3] pager: test for exit behavior & trace2 bug fix Ævar Arnfjörð Bjarmason
2021-02-02 1:59 ` [PATCH v2 0/5] " Ævar Arnfjörð Bjarmason
2021-02-02 1:59 ` [PATCH v2 1/5] pager: refactor wait_for_pager() function Ævar Arnfjörð Bjarmason
2021-02-02 1:59 ` [PATCH v2 2/5] pager: test for exit code with and without SIGPIPE Ævar Arnfjörð Bjarmason
2021-02-02 8:50 ` Denton Liu
2021-02-05 7:47 ` Johannes Sixt
2021-02-02 1:59 ` [PATCH v2 3/5] run-command: add braces for "if" block in wait_or_whine() Ævar Arnfjörð Bjarmason
2021-02-02 2:00 ` [PATCH v2 4/5] pager: properly log pager exit code when signalled Ævar Arnfjörð Bjarmason
2021-02-05 7:58 ` Johannes Sixt
2021-02-05 11:37 ` Junio C Hamano
2021-02-02 2:00 ` [WIP/PATCH v2 5/5] WIP pager: respect exit code of pager over SIGPIPE Ævar Arnfjörð Bjarmason
2021-02-01 14:49 ` [PATCH 1/3] pager: test for exit code Ævar Arnfjörð Bjarmason
2021-02-01 14:49 ` [PATCH 2/3] pager: refactor wait_for_pager() function Ævar Arnfjörð Bjarmason
2021-02-01 14:49 ` [PATCH 3/3] pager: properly log pager exit code when signalled Ævar Arnfjörð Bjarmason
2021-02-01 18:07 ` Junio C Hamano
2021-02-01 19:21 ` Ævar Arnfjörð Bjarmason [this message]
2021-02-01 18:15 ` Junio C Hamano
2021-02-01 19:23 ` Ævar Arnfjörð Bjarmason
2021-02-01 22:04 ` git fails with a broken pipe when one quits the pager Johannes Sixt
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=8735yffvbj.fsf@evledraar.gmail.com \
--to=avarab@gmail.com \
--cc=chris.torek@gmail.com \
--cc=git@vger.kernel.org \
--cc=gitster@pobox.com \
--cc=jeffhost@microsoft.com \
--cc=liu.denton@gmail.com \
--cc=vincent@vinc17.net \
/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).