git.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: "Ævar Arnfjörð Bjarmason" <avarab@gmail.com>
To: git@vger.kernel.org
Cc: "Junio C Hamano" <gitster@pobox.com>, "Jeff King" <peff@peff.net>,
	"Jakub Narębski" <jnareb@gmail.com>,
	"Jacob Keller" <jacob.keller@gmail.com>,
	"Matt McCutchen" <matt@mattmccutchen.net>,
	"Ævar Arnfjörð Bjarmason" <avarab@gmail.com>
Subject: [PATCH] push: document & test --force-with-lease with multiple remotes
Date: Sat,  8 Apr 2017 11:41:00 +0000	[thread overview]
Message-ID: <20170408114100.13743-1-avarab@gmail.com> (raw)
In-Reply-To: <487622bf-00d0-e4fc-4a74-08e18d59336a@gmail.com>

Document & test for cases where there are two remotes pointing to the
same URL, and a background fetch & subsequent `git push
--force-with-lease` shouldn't clobber un-updated references we haven't
fetched.

Some editors like Microsoft's VSC have a feature to auto-fetch in the
background, this bypasses the protections offered by
--force-with-lease as noted in the documentation being added here.

See the 'Tools that do an automatic fetch defeat "git push
--force-with-lease"' (<1491617750.2149.10.camel@mattmccutchen.net>)
git mailing list thread for more details. Jakub Narębski suggested
this method of adding another remote to bypass this edge case,
document that & add a test for it.

Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
---

On Sat, Apr 8, 2017 at 11:29 AM, Jeff King <peff@peff.net> wrote:
> On Sat, Apr 08, 2017 at 09:35:04AM +0200, Ævar Arnfjörð Bjarmason wrote:
>
>> Is it correct that you'd essentially want something that works like:
>>
>>     git push --force-with-lease=master:master origin master:master
>
> I don't think that would do anything useful. It would reject any push
> where the remote "master" is not the same as your own master. And of
> course if they _are_ the same, then the push is a noop.
>

Yeah my whole suggestion is obviously dumb & useless. But I liked
Jakub's suggestion to work around this, so here's docs & a test for
that.

According to my eyeballing of the MS VSC code this should work,
i.e. it seems to do a 'fetch' here, not a 'fetch --all':
https://github.com/Microsoft/vscode/blob/master/src/vs/workbench/parts/git/node/git.lib.ts#L505

Of course another way is to just disable autofetching:
https://github.com/Microsoft/vscode/blob/535a3de60023c81d75d0eac22044284f07dbcddf/extensions/git/src/autofetch.ts#L27

But having two remotes allows you to have your cake & eat it too
without all the hassle of tag creation, which I've added to the docs
though for completeness.

 Documentation/git-push.txt | 37 +++++++++++++++++++++++++++++++++++++
 t/t5533-push-cas.sh        | 29 +++++++++++++++++++++++++++++
 2 files changed, 66 insertions(+)

diff --git a/Documentation/git-push.txt b/Documentation/git-push.txt
index 1624a35888..2f2e9c078b 100644
--- a/Documentation/git-push.txt
+++ b/Documentation/git-push.txt
@@ -210,6 +210,43 @@ or we do not even have to have such a remote-tracking branch when
 this form is used).  If `<expect>` is the empty string, then the named ref
 must not already exist.
 +
+This option interacts very badly with anything that implicitly runs
+`git fetch` on the remote to be pushed to in the background. The
+protection it offers over `--force` is ensuring that subsequent
+changes your work wasn't based on aren't clobbered, but this is
+trivially defeated if some background process is updating refs in the
+background. We don't have anything except the remote tracking info to
+go by as a heuristic for refs you're expected to have seen & are
+willing to clobber.
++
+If your editor or some other system is running `git fetch` in the
+background for you a way to mitigate this is to simply set up another
+remote:
++
+	git remote add origin-push $(git config remote.origin.url)
+	git fetch origin-push
++
+Now when the background process runs `git fetch origin` the references
+on `origin-push` won't be updated, and thus commands like:
++
+	git push --force-with-lease origin
++
+Will fail unless you manually run `git fetch origin-push`. This method
+is of course entirely defeated by something that runs `git fetch
+--all`, in that case you'd need to either disable it or do something
+more tedious like:
++
+	git fetch              ;# update 'master' from remote
+	git tag base master    ;# mark our base point
+	git rebase -i master   ;# rewrite some commits
+	git push --force-with-lease=master:base master:master
++
+I.e. create a `base` tag for versions of the upstream code that you've
+seen and are willing to overwrite, then rewrite history, and finally
+force push changes to `master` if the remote version is still at
+`base`, regardless of what your local `remotes/origin/master` has been
+updated to in the background.
++
 Note that all forms other than `--force-with-lease=<refname>:<expect>`
 that specifies the expected current value of the ref explicitly are
 still experimental and their semantics may change as we gain experience
diff --git a/t/t5533-push-cas.sh b/t/t5533-push-cas.sh
index a2c9e7439f..d38ecee217 100755
--- a/t/t5533-push-cas.sh
+++ b/t/t5533-push-cas.sh
@@ -229,4 +229,33 @@ test_expect_success 'new branch already exists' '
 	)
 '
 
+test_expect_success 'background updates of REMOTE can be mitigated with a non-updated REMOTE-push' '
+	rm -rf src dst &&
+	git init --bare src.bare &&
+	test_when_finished "rm -rf src.bare" &&
+	git clone --no-local src.bare dst &&
+	test_when_finished "rm -rf dst" &&
+	(
+		cd dst &&
+		test_commit G &&
+		git remote add origin-push ../src.bare &&
+		git push origin-push master:master
+	) &&
+	git clone --no-local src.bare dst2 &&
+	test_when_finished "rm -rf dst2" &&
+	(
+		cd dst2 &&
+		test_commit H &&
+		git push
+	) &&
+	(
+		cd dst &&
+		test_commit I &&
+		git fetch origin &&
+		test_must_fail git push --force-with-lease origin-push &&
+		git fetch origin-push &&
+		git push --force-with-lease origin-push
+	)
+'
+
 test_done
-- 
2.11.0


  reply	other threads:[~2017-04-08 11:41 UTC|newest]

Thread overview: 49+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-04-08  2:15 Tools that do an automatic fetch defeat "git push --force-with-lease" Matt McCutchen
2017-04-08  7:24 ` Stefan Haller
2017-04-08  7:35 ` Ævar Arnfjörð Bjarmason
2017-04-08  9:29   ` Jeff King
2017-04-08 10:10     ` Jakub Narębski
2017-04-08 11:41       ` Ævar Arnfjörð Bjarmason [this message]
2017-04-09  9:55         ` [PATCH] push: document & test --force-with-lease with multiple remotes Simon Ruderich
2017-04-09 11:40           ` Ævar Arnfjörð Bjarmason
2017-04-17  3:56         ` Junio C Hamano
2017-04-19  9:22           ` [PATCH v2] " Ævar Arnfjörð Bjarmason
2017-04-08 21:54     ` Tools that do an automatic fetch defeat "git push --force-with-lease" Jacob Keller
2017-04-08 22:13       ` Jeff King
2017-04-08 22:21         ` Jacob Keller
2017-04-09  8:38         ` Stefan Haller
2017-04-09  8:49           ` Jacob Keller
2017-04-09 11:00             ` Stefan Haller
2017-04-10  8:08               ` Jacob Keller
2017-04-10  9:58                 ` Ævar Arnfjörð Bjarmason
2017-04-10 23:33                   ` Jacob Keller
2017-04-11  8:51                     ` Junio C Hamano
2017-04-12  9:11                       ` Stefan Haller
2017-07-06 18:56                       ` [PATCH] push: disable lazy --force-with-lease by default Junio C Hamano
2017-07-06 19:38                         ` Stefan Beller
2017-07-06 22:39                           ` Junio C Hamano
2017-07-06 22:42                             ` Stefan Beller
2017-07-10 22:32                             ` Stefan Beller
2017-07-07  9:24                         ` Stefan Haller
2017-07-07  9:42                           ` Jeff King
2017-07-07  9:54                           ` Ævar Arnfjörð Bjarmason
2017-07-07 15:15                             ` Junio C Hamano
2017-07-15 10:45                               ` Ævar Arnfjörð Bjarmason
2017-07-17 17:28                                 ` Junio C Hamano
2017-07-07  9:39                         ` Ævar Arnfjörð Bjarmason
2017-04-11 12:37                   ` Tools that do an automatic fetch defeat "git push --force-with-lease" Stefan Haller
2017-04-11 12:37                 ` Stefan Haller
2017-04-10 18:31           ` Jeff King
2017-04-11 12:37             ` Stefan Haller
2017-04-11 12:50               ` Jeff King
2017-04-12  9:11                 ` Stefan Haller
2017-04-09  8:38       ` Stefan Haller
2017-04-09  8:46         ` Jacob Keller
2017-04-08  8:25 ` Jacob Keller
2017-04-08  9:31   ` Jeff King
2017-04-08 15:03     ` Stefan Haller
2017-04-08 22:03       ` Jeff King
2017-04-08 15:03 ` Stefan Haller
2017-04-08 16:04   ` Ævar Arnfjörð Bjarmason
2017-04-08 17:28     ` Stefan Haller
2017-04-12  9:11   ` Stefan Haller

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=20170408114100.13743-1-avarab@gmail.com \
    --to=avarab@gmail.com \
    --cc=git@vger.kernel.org \
    --cc=gitster@pobox.com \
    --cc=jacob.keller@gmail.com \
    --cc=jnareb@gmail.com \
    --cc=matt@mattmccutchen.net \
    --cc=peff@peff.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).