From: Jeff King <peff@peff.net>
To: Junio C Hamano <gitster@pobox.com>
Cc: Noam Postavsky <npostavs@users.sourceforge.net>, git@vger.kernel.org
Subject: Re: git-credential-cache--daemon quits on SIGHUP, can we change it to ignore instead?
Date: Fri, 4 Dec 2015 14:06:59 -0500 [thread overview]
Message-ID: <20151204190658.GA16692@sigill.intra.peff.net> (raw)
In-Reply-To: <xmqqpoymrql7.fsf@gitster.mtv.corp.google.com>
On Fri, Dec 04, 2015 at 10:55:32AM -0800, Junio C Hamano wrote:
> >> + int ignore_sighup = 0;
> >> + git_config_get_bool("credentialcache.ignoresighup", &ignore_sighup);
> >> +
> >
> > Style-wise, I think the declaration should go above the options-list.
>
> I was about to merge this to 'master', following your last issue of
> "What's cooking" report.
>
> I was puzzled that git_config_get_bool() is used here without even
> checking if we are inside any Git repository and wondered if it was
> correct. I'd imagine this is not a problem, as this process is
> spawned by "credential-cache" that was spawned by somebody (either
> push or fetch) who has read $GIT_DIR/config for credential.helper to
> determine that credential-cache needs to be used.
That does not have to be the case; I imagine most people would put
credential.helper in their ~/.gitconfig.
But I'm not sure I understand how that is relevant. The config subsystem
should work just fine whether we are in a repository or not (and if not,
return results only from system and user-wide config).
This probably _does_ trigger setup_git_env() when it was not otherwise
called, and it will back to looking at ".git/config" for the repo-level
config. That may fail to find the file if we are in a bare repository,
or a subdirectory of the working tree. IOW, I suspect this:
git init --bare foo.git
cd foo.git
git config credential.helper cache
git config credentialcache.ignoreSIGHUP true ;# goes into local config
git fetch https://example.com/foo.git
may fail to respect the ignoreSIGHUP option.
I guess the solution would be to setup_git_director_gently() in the
daemon process.
-Peff
next prev parent reply other threads:[~2015-12-04 19:07 UTC|newest]
Thread overview: 29+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-10-10 16:45 git-credential-cache--daemon quits on SIGHUP, can we change it to ignore instead? Noam Postavsky
2015-10-18 15:15 ` Noam Postavsky
2015-10-18 17:58 ` Junio C Hamano
2015-10-19 0:51 ` Noam Postavsky
2015-10-21 2:35 ` Noam Postavsky
2015-10-24 21:47 ` Noam Postavsky
2015-10-25 16:58 ` Junio C Hamano
2015-10-26 21:50 ` Jeff King
2015-10-27 0:50 ` Noam Postavsky
2015-10-27 18:41 ` Jeff King
2015-10-27 19:04 ` Junio C Hamano
2015-10-27 17:52 ` Junio C Hamano
2015-10-27 18:47 ` Jeff King
2015-10-28 3:46 ` Noam Postavsky
2015-10-30 0:10 ` Jeff King
2015-10-30 0:43 ` Noam Postavsky
2015-10-30 0:50 ` Jeff King
2015-10-30 1:20 ` Noam Postavsky
2015-10-30 21:08 ` Jeff King
2015-11-09 2:58 ` Noam Postavsky
2015-11-09 15:53 ` Jeff King
2015-11-10 1:05 ` Noam Postavsky
2015-11-10 12:25 ` Jeff King
2015-11-10 12:26 ` Jeff King
2015-11-11 0:22 ` Noam Postavsky
2015-12-04 18:55 ` Junio C Hamano
2015-12-04 19:06 ` Jeff King [this message]
2015-12-04 20:05 ` Junio C Hamano
2015-12-04 23:25 ` Jeff King
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=20151204190658.GA16692@sigill.intra.peff.net \
--to=peff@peff.net \
--cc=git@vger.kernel.org \
--cc=gitster@pobox.com \
--cc=npostavs@users.sourceforge.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).