From: Phillip Wood <phillip.wood123@gmail.com>
To: Junio C Hamano <gitster@pobox.com>, git@vger.kernel.org
Subject: Re: [PATCH v2 0/3] safe.directory clean-up
Date: Thu, 25 Jul 2024 10:45:46 +0100 [thread overview]
Message-ID: <d9d1115a-d530-479a-acf8-189713632adf@gmail.com> (raw)
In-Reply-To: <20240723021900.388020-1-gitster@pobox.com>
Hi Junio
Thanks for picking this up, I think this looks like a good approach
apart from resolving relative entries in safe.directory which is
intrinsically unsafe as we don't know which directory the user wants to
consider safe. With these changes it should not be necessary to add "."
to safe.directory to get "git daemon" to work and all the other code
paths use an absolute path from getcwd() so I don't think there is any
need to support relative directories.
I'll be off the list for the next couple of weeks
Best Wishes
Phillip
On 23/07/2024 03:18, Junio C Hamano wrote:
> Changes since v1:
>
> - two "extra" patches for optimization that are unnecessary have
> been dropped.
> - a patch to add test for safe.directory="." use case has been
> added.
>
> Recently we discussed what we should do when either the path
> configured in the safe.directory configuration or coming from
> the caller of ensure_valid_ownership() function as a result of
> repository discovery is not normalized and textual equality check is
> not sufficient. See the thread the contains
>
> https://lore.kernel.org/git/6d5b75a6-639d-429b-bd37-232fc6f475af@gmail.com/
>
> Here are three patches that implement the comparison between
> normalized path and configuration value.
>
> Imagine that you have a repository at /mnt/disk4/repos/frotz
> directory but in order to make it simpler to manage and use, you
> have your users use /projects/frotz to access the repository. A
> symlink /projects/frotz pointing at /mnt/disk4/repos/frotz directory
> allows you to do so.
>
> - The first patch normalizes the path to the directory that we
> suspect is a usable repository, before comparing it with the
> safe.directory configuration variable. The safe.directory may
> say /mnt/disk4/repos/frotz or /mnt/disk4/repos/*, but the path to
> the repository for the users may be /mnt/disk4/repos/frotz or
> /projects/frotz, depending on where they come from and what their
> $PWD makes getcwd() to say.
>
> - The second patch normalizes the value of the safe.directory
> variable. This allows safe.directory to say /projects/frotz
> or /projects/* and have them match /mnt/disk4/repos/frotz (which
> is how the first patch normalizes the repository path to).
>
> - The third patch only adds a test to illustrate what happens when
> the safe.directory configuration is set to ".", which was a
> plausible setting for those who run "git daemon". The
> normalization done by the first two patches is sufficient to make
> this just work.
>
> Junio C Hamano (3):
> safe.directory: normalize the checked path
> safe.directory: normalize the configured path
> safe.directory: setting safe.directory="." allows the "current"
> directory
>
> setup.c | 28 ++++++--
> t/t0033-safe-directory.sh | 146 ++++++++++++++++++++++++++++++++++++++
> 2 files changed, 170 insertions(+), 4 deletions(-)
>
> Range-diff against v1:
> 1: 20e1160946 ! 1: 86d5c83ee7 safe.directory: normalize the checked path
> @@ t/t0033-safe-directory.sh: test_expect_success 'local clone of unowned repo acce
> + sane_unset GIT_TEST_ASSUME_DIFFERENT_OWNER &&
> + git config --global safe.directory "$(pwd)/repository"
> + ) &&
> -+ git -C repository/ for-each-ref >/dev/null &&
> -+ git -C repo/ for-each-ref
> ++ git -C repository for-each-ref &&
> ++ git -C repository/ for-each-ref &&
> ++ git -C repo for-each-ref &&
> ++ git -C repo/ for-each-ref &&
> ++ test_must_fail git -C repository/.git for-each-ref &&
> ++ test_must_fail git -C repository/.git/ for-each-ref &&
> ++ test_must_fail git -C repo/.git for-each-ref &&
> ++ test_must_fail git -C repo/.git/ for-each-ref
> +'
> +
> +test_expect_success SYMLINKS 'checked leading paths are normalized' '
> @@ t/t0033-safe-directory.sh: test_expect_success 'local clone of unowned repo acce
> + sane_unset GIT_TEST_ASSUME_DIFFERENT_OWNER &&
> + git config --global safe.directory "$(pwd)/repository/*"
> + ) &&
> -+ git -C repository/s for-each-ref >/dev/null &&
> -+ git -C repo/s for-each-ref
> ++ git -C repository/s for-each-ref &&
> ++ git -C repository/s/ for-each-ref &&
> ++ git -C repo/s for-each-ref &&
> ++ git -C repo/s/ for-each-ref &&
> ++ git -C repository/s/.git for-each-ref &&
> ++ git -C repository/s/.git/ for-each-ref &&
> ++ git -C repo/s/.git for-each-ref &&
> ++ git -C repo/s/.git/ for-each-ref
> +'
> +
> test_done
> 2: 06de9038b7 ! 2: c4998076da safe.directory: normalize the configured path
> @@ setup.c: static int safe_directory_cb(const char *key, const char *value,
>
> ## t/t0033-safe-directory.sh ##
> @@ t/t0033-safe-directory.sh: test_expect_success SYMLINKS 'checked leading paths are normalized' '
> - git -C repo/s for-each-ref
> + git -C repo/s/.git/ for-each-ref
> '
>
> +test_expect_success SYMLINKS 'configured paths are normalized' '
> @@ t/t0033-safe-directory.sh: test_expect_success SYMLINKS 'checked leading paths a
> + sane_unset GIT_TEST_ASSUME_DIFFERENT_OWNER &&
> + git config --global safe.directory "$(pwd)/repo"
> + ) &&
> -+ git -C repository/ for-each-ref >/dev/null &&
> -+ git -C repo/ for-each-ref
> ++ git -C repository for-each-ref &&
> ++ git -C repository/ for-each-ref &&
> ++ git -C repo for-each-ref &&
> ++ git -C repo/ for-each-ref &&
> ++ test_must_fail git -C repository/.git for-each-ref &&
> ++ test_must_fail git -C repository/.git/ for-each-ref &&
> ++ test_must_fail git -C repo/.git for-each-ref &&
> ++ test_must_fail git -C repo/.git/ for-each-ref
> +'
> +
> +test_expect_success SYMLINKS 'configured leading paths are normalized' '
> @@ t/t0033-safe-directory.sh: test_expect_success SYMLINKS 'checked leading paths a
> + git init repository/s &&
> + ln -s repository repo &&
> + (
> -+ cd repository &&
> ++ cd repository/s &&
> + sane_unset GIT_TEST_ASSUME_DIFFERENT_OWNER &&
> + test_commit sample
> + ) &&
> @@ t/t0033-safe-directory.sh: test_expect_success SYMLINKS 'checked leading paths a
> + sane_unset GIT_TEST_ASSUME_DIFFERENT_OWNER &&
> + git config --global safe.directory "$(pwd)/repo/*"
> + ) &&
> -+ git -C repository/s for-each-ref >/dev/null &&
> -+ git -C repo/s for-each-ref
> ++ git -C repository/s for-each-ref &&
> ++ git -C repository/s/ for-each-ref &&
> ++ git -C repository/s/.git for-each-ref &&
> ++ git -C repository/s/.git/ for-each-ref &&
> ++ git -C repo/s for-each-ref &&
> ++ git -C repo/s/ for-each-ref &&
> ++ git -C repo/s/.git for-each-ref &&
> ++ git -C repo/s/.git/ for-each-ref
> +'
> +
> test_done
> -: ---------- > 3: ffc3f7364e safe.directory: setting safe.directory="." allows the "current" directory
next prev parent reply other threads:[~2024-07-25 9:45 UTC|newest]
Thread overview: 39+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-07-20 22:09 [PATCH 0/2] safe.directory clean-up Junio C Hamano
2024-07-20 22:09 ` [PATCH 1/2] safe.directory: normalize the checked path Junio C Hamano
2024-07-20 22:09 ` [PATCH 2/2] safe.directory: normalize the configured path Junio C Hamano
2024-07-20 22:09 ` [PATCH 3/2] setup: use a single return path in setup_git_directory*() Junio C Hamano
2024-07-20 22:09 ` [PATCH 4/2] setup: cache normalized safe.directory configuration Junio C Hamano
2024-07-23 2:18 ` [PATCH v2 0/3] safe.directory clean-up Junio C Hamano
2024-07-23 2:18 ` [PATCH v2 1/3] safe.directory: normalize the checked path Junio C Hamano
2024-07-23 2:18 ` [PATCH v2 2/3] safe.directory: normalize the configured path Junio C Hamano
2024-07-25 9:45 ` Phillip Wood
2024-07-25 16:11 ` Junio C Hamano
2024-08-14 13:20 ` Phillip Wood
2024-08-14 17:15 ` Junio C Hamano
2024-08-15 9:51 ` Phillip Wood
2024-08-15 14:43 ` Junio C Hamano
2024-07-26 5:02 ` Jeff King
2024-07-26 15:02 ` Junio C Hamano
2024-07-27 22:05 ` Jeff King
2024-07-23 2:19 ` [PATCH v2 3/3] safe.directory: setting safe.directory="." allows the "current" directory Junio C Hamano
2024-07-25 9:45 ` Phillip Wood
2024-07-25 16:12 ` Junio C Hamano
2024-07-25 9:45 ` Phillip Wood [this message]
2024-07-25 16:14 ` [PATCH v2 0/3] safe.directory clean-up Junio C Hamano
2024-07-30 1:10 ` [PATCH v3 " Junio C Hamano
2024-07-30 1:10 ` [PATCH v3 1/3] safe.directory: normalize the checked path Junio C Hamano
2024-07-30 1:10 ` [PATCH v3 2/3] safe.directory: normalize the configured path Junio C Hamano
2024-07-30 7:31 ` Jeff King
2024-07-30 16:03 ` Junio C Hamano
2024-07-30 20:08 ` Jeff King
2024-07-30 7:43 ` Jeff King
2024-07-30 16:22 ` Junio C Hamano
2024-07-30 17:56 ` safe.directory: preliminary clean-up Junio C Hamano
2024-07-30 20:13 ` Jeff King
2024-07-30 20:10 ` [PATCH v3 2/3] safe.directory: normalize the configured path Jeff King
2024-07-30 1:10 ` [PATCH v3 3/3] safe.directory: setting safe.directory="." allows the "current" directory Junio C Hamano
2024-07-30 18:43 ` [PATCH v4 0/4] safe.directory clean-up Junio C Hamano
2024-07-30 18:43 ` [PATCH v4 1/4] safe.directory: preliminary clean-up Junio C Hamano
2024-07-30 18:43 ` [PATCH v4 2/4] safe.directory: normalize the checked path Junio C Hamano
2024-07-30 18:43 ` [PATCH v4 3/4] safe.directory: normalize the configured path Junio C Hamano
2024-07-30 18:43 ` [PATCH v4 4/4] safe.directory: setting safe.directory="." allows the "current" directory 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
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=d9d1115a-d530-479a-acf8-189713632adf@gmail.com \
--to=phillip.wood123@gmail.com \
--cc=git@vger.kernel.org \
--cc=gitster@pobox.com \
--cc=phillip.wood@dunelm.org.uk \
/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).