git@vger.kernel.org mailing list mirror (one of many)
 help / color / mirror / code / Atom feed
From: Evan Priestley <epriestley@epriestley.com>
To: git@vger.kernel.org
Subject: Re: regression: "96b9e0e3 config: treat user and xdg config permission problems as errors" busted git-daemon
Date: Fri, 12 Apr 2013 07:45:56 -0700	[thread overview]
Message-ID: <D960D402-22EA-4C2B-BFDD-9BE512C88D97@epriestley.com> (raw)

Jonathan Nieder <jrnie...@gmail.com> wrote:

> I'm not sure whether to keep 96b9e0e (config: treat user and xdg
> config permission problem as errors) in the long run, BTW.
> 
> Insights welcome.

For what it's worth, here's an anecdote about this:

I work on some open source software which includes a web-based repository browser for Git, somewhat similar to gitweb. We implement this partially by executing `git` commands from the webserver (usually Apache). For example, we run `git cat-file …` to retrieve file content to show to the user.

After this change, a number of users who manage installs of the software are hitting "fatal: unable to access '/root/.config/git/config': Permission denied" while browsing repositories, because their Apache runs as some unprivileged user (like "apache" or "www-data") but with HOME=/root. We've seen about half a dozen reports of this now, so I believe this sort of setup is at least somewhat common and not just a bizarre one-off user with a broken environment. Users generally have difficulty resolving this error on their own, as it's not obvious that this boils down to an Apache environmental issue.

We'll likely resolve this by running `HOME=/ git ...` instead of `git ...` when we execute commands (or some more finessed version of that, but basically pointing HOME at some dummy readable directory). From cursory investigation, it appears we can't avoid this fatal with more surgical settings like GIT_CONFIG or XDG_CONFIG_HOME, since git still ends up looking in HOME and fataling anyway. This fix is a bit clunky, but not really a big deal.

I imagine our use case is fairly unusual, but I wanted to relate it in case it's helpful in balancing concerns here. If I've missed a more reasonable approach to solving this than redirecting HOME, please let me know, but it looks like that's more or less what the git-daemon patch is doing too.

Thanks,
Evan Priestley

             reply	other threads:[~2013-04-12 14:46 UTC|newest]

Thread overview: 34+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-04-12 14:45 Evan Priestley [this message]
  -- strict thread matches above, loose matches on Subject: below --
2013-04-10  5:33 regression: "96b9e0e3 config: treat user and xdg config permission problems as errors" busted git-daemon Mike Galbraith
2013-04-10 13:56 ` W. Trevor King
2013-04-11  3:39   ` Mike Galbraith
2013-04-11  5:42     ` Jeff King
2013-04-11  7:59       ` Mike Galbraith
2013-04-11 15:35       ` Junio C Hamano
2013-04-11 17:24         ` Jeff King
2013-04-11 18:11           ` Jonathan Nieder
2013-04-11 18:14             ` Jeff King
2013-04-11 18:25               ` Jonathan Nieder
2013-04-11 19:54               ` Junio C Hamano
2013-04-11 20:03                 ` W. Trevor King
2013-04-11 22:20                   ` Junio C Hamano
2013-04-11 22:23                     ` Jeff King
2013-04-12  0:57                       ` W. Trevor King
2013-04-12  4:11                         ` Junio C Hamano
2013-04-12  4:35                           ` Jeff King
2013-04-12  4:46                             ` Junio C Hamano
2013-04-12  5:05                               ` Jeff King
2013-04-12  5:46                                 ` Mike Galbraith
2013-04-12 11:26                                 ` W. Trevor King
2013-04-12 14:48                                   ` Jeff King
2013-04-12 16:08                                     ` Junio C Hamano
2013-04-12 16:16                                       ` Jeff King
2013-04-12 17:05                                         ` Jeff King
2013-04-12 18:23                                           ` Junio C Hamano
2013-04-12 19:01                                             ` Jeff King
2013-04-12 19:51                                               ` Junio C Hamano
2013-04-12 19:58                                                 ` Jeff King
2013-04-12 20:45                                                   ` Junio C Hamano
2013-04-12 17:31                                         ` Junio C Hamano
2013-04-12 16:21                                       ` Mike Galbraith
2013-04-11 20:08                 ` 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

  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=D960D402-22EA-4C2B-BFDD-9BE512C88D97@epriestley.com \
    --to=epriestley@epriestley.com \
    --cc=git@vger.kernel.org \
    /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).