git/list[1] front-page[2] threads[3] people[4] search[5] about
 

Re: [PATCH v6 5/8] fetch: protect branches checked out in all worktrees

From
Johannes Schindelin <johannes.schindelin@gmx.de>
Date
Nov 22, 2021, 13:13 UTC
Message-ID
<nycvar.QRO.7.76.6.2111221358160.63@tvgsbejvaqbjf.bet>
In-Reply-To
<20211113033358.2179376-6-andersk@mit.edu>
Hi Anders,

tl;dr the patch looks nice! A few remarks below, to prove that I did not merely glance over the patch.

On Fri, 12 Nov 2021, Anders Kaseorg wrote:
Show 6 quoted lines
> Refuse to fetch into the currently checked out branch of any working
> tree, not just the current one.
>
> Fixes this previously reported bug:
>
> https://public-inbox.org/git/cb957174-5e9a-5603-ea9e-ac9b58a2eaad@mathema.de
These days, I think the lore.kernel.org/git/ links are slightly preferred.
Show 33 quoted lines
>
> As a side effect of using find_shared_symref, we’ll also refuse the
> fetch when we’re on a detached HEAD because we’re rebasing or bisecting
> on the branch in question. This seems like a sensible change.
>
> Signed-off-by: Anders Kaseorg <andersk@mit.edu>
> ---
>  builtin/fetch.c       | 75 +++++++++++++++++++++++--------------------
>  t/t5516-fetch-push.sh | 18 +++++++++++
>  2 files changed, 58 insertions(+), 35 deletions(-)
>
> diff --git a/builtin/fetch.c b/builtin/fetch.c
> index e5971fa6e5..f373252490 100644
> --- a/builtin/fetch.c
> +++ b/builtin/fetch.c
> @@ -28,6 +28,7 @@
>  #include "promisor-remote.h"
>  #include "commit-graph.h"
>  #include "shallow.h"
> +#include "worktree.h"
>
>  #define FORCED_UPDATES_DELAY_WARNING_IN_MS (10 * 1000)
>
> @@ -840,14 +841,13 @@ static void format_display(struct strbuf *display, char code,
>
>  static int update_local_ref(struct ref *ref,
>  			    struct ref_transaction *transaction,
> -			    const char *remote,
> -			    const struct ref *remote_ref,
> -			    struct strbuf *display,
> -			    int summary_width)
> +			    const char *remote, const struct ref *remote_ref,
> +			    struct strbuf *display, int summary_width,

As a reviewer, I prefer to spend my time reviewing the actual change. In this case, the re-wrapping is a bit distracting. Not really important here, but maybe for future contributions. The easier a contributor makes it on reviewers, the better (it definitely helps avoid the grumpy judge bias[*1*]).

Show 70 quoted lines
> +			    struct worktree **worktrees)
>  {
>  	struct commit *current = NULL, *updated;
>  	enum object_type type;
> -	struct branch *current_branch = branch_get(NULL);
> +	const struct worktree *wt;
>  	const char *pretty_ref = prettify_refname(ref->name);
>  	int fast_forward = 0;
>
> @@ -862,16 +862,17 @@ static int update_local_ref(struct ref *ref,
>  		return 0;
>  	}
>
> -	if (current_branch &&
> -	    !strcmp(ref->name, current_branch->name) &&
> -	    !(update_head_ok || is_bare_repository()) &&
> -	    !is_null_oid(&ref->old_oid)) {
> +	if (!update_head_ok &&
> +	    (wt = find_shared_symref(worktrees, "HEAD", ref->name)) &&
> +	    !wt->is_bare && !is_null_oid(&ref->old_oid)) {
>  		/*
>  		 * If this is the head, and it's not okay to update
>  		 * the head, and the old value of the head isn't empty...
>  		 */
>  		format_display(display, '!', _("[rejected]"),
> -			       _("can't fetch in current branch"),
> +			       wt->is_current ?
> +				       _("can't fetch in current branch") :
> +				       _("checked out in another worktree"),
>  			       remote, pretty_ref, summary_width);
>  		return 1;
>  	}
> @@ -1071,7 +1072,8 @@ N_("it took %.2f seconds to check forced updates; you can use\n"
>     " to avoid this check\n");
>
>  static int store_updated_refs(const char *raw_url, const char *remote_name,
> -			      int connectivity_checked, struct ref *ref_map)
> +			      int connectivity_checked, struct ref *ref_map,
> +			      struct worktree **worktrees)
>  {
>  	struct fetch_head fetch_head;
>  	struct commit *commit;
> @@ -1188,7 +1190,8 @@ static int store_updated_refs(const char *raw_url, const char *remote_name,
>  			strbuf_reset(&note);
>  			if (ref) {
>  				rc |= update_local_ref(ref, transaction, what,
> -						       rm, &note, summary_width);
> +						       rm, &note, summary_width,
> +						       worktrees);
>  				free(ref);
>  			} else if (write_fetch_head || dry_run) {
>  				/*
> @@ -1301,16 +1304,15 @@ static int fetch_refs(struct transport *transport, struct ref *ref_map)
>  }
>
>  /* Update local refs based on the ref values fetched from a remote */
> -static int consume_refs(struct transport *transport, struct ref *ref_map)
> +static int consume_refs(struct transport *transport, struct ref *ref_map,
> +			struct worktree **worktrees)
>  {
>  	int connectivity_checked = transport->smart_options
>  		? transport->smart_options->connectivity_checked : 0;
>  	int ret;
>  	trace2_region_enter("fetch", "consume_refs", the_repository);
> -	ret = store_updated_refs(transport->url,
> -				 transport->remote->name,
> -				 connectivity_checked,
> -				 ref_map);
> +	ret = store_updated_refs(transport->url, transport->remote->name,
> +				 connectivity_checked, ref_map, worktrees);
Again, this rewrapping is slightly distracting to my brain.
Show 9 quoted lines
>  	transport_unlock_pack(transport);
>  	trace2_region_leave("fetch", "consume_refs", the_repository);
>  	return ret;
> @@ -1643,7 +1647,7 @@ static int do_fetch(struct transport *transport,
>  				  "you need to specify exactly one branch with the --set-upstream option"));
>  		}
>  	}
> - skip:
> +skip:

Okay, this one is distracting _but also_ pleasing. I am only pointing this out to prove that I did not go over your patch sloppily. :-)

Show 11 quoted lines
> @@ -1653,11 +1657,12 @@ static int do_fetch(struct transport *transport,
>  		ref_map = NULL;
>  		find_non_local_tags(remote_refs, &ref_map, &tail);
>  		if (ref_map)
> -			backfill_tags(transport, ref_map);
> +			backfill_tags(transport, ref_map, worktrees);
>  		free_refs(ref_map);
>  	}
>
> - cleanup:
> +cleanup:
Same here.
Show 31 quoted lines
> +	free_worktrees(worktrees);
>  	return retcode;
>  }
>
> diff --git a/t/t5516-fetch-push.sh b/t/t5516-fetch-push.sh
> index 4db8edd9c8..36fb90f4b0 100755
> --- a/t/t5516-fetch-push.sh
> +++ b/t/t5516-fetch-push.sh
> @@ -1770,4 +1770,22 @@ test_expect_success 'denyCurrentBranch and worktrees' '
>  	git -C cloned push origin HEAD:new-wt &&
>  	test_must_fail git -C cloned push --delete origin new-wt
>  '
> +
> +test_expect_success 'refuse fetch to current branch of worktree' '
> +	test_when_finished "git worktree remove --force wt && git branch -D wt" &&
> +	git worktree add wt &&
> +	test_commit apple &&
> +	test_must_fail git fetch . HEAD:wt &&
> +	git fetch -u . HEAD:wt
> +'
> +
> +test_expect_success 'refuse fetch to current branch of bare repository worktree' '
> +	test_when_finished "rm -fr bare.git" &&
> +	git clone --bare . bare.git &&
> +	git -C bare.git worktree add wt &&
> +	test_commit banana &&
> +	test_must_fail git -C bare.git fetch .. HEAD:wt &&
> +	git -C bare.git fetch -u .. HEAD:wt
> +'
> +
>  test_done
Reads nicely.

Thanks, Dscho

Footnote: https://en.wikipedia.org/wiki/Hungry_judge_effect
Previous: Anders KaseorgNext: Anders Kaseorg
Message 12 of 21 in “protect branches checked out in all worktrees”
  1. 0/8 protect branches checked out in all worktreesAnders Kaseorg, Nov 13, 2021
  2. 1/8 fetch: lowercase error messagesAnders Kaseorg, Nov 13, 2021
  3. Junio C HamanoNov 16, 2021
  4. Anders KaseorgNov 16, 2021
  5. Junio C HamanoNov 17, 2021
  6. Jiang XinNov 22, 2021
  7. 2/8 receive-pack: lowercase error messagesAnders Kaseorg, Nov 13, 2021
  8. Junio C HamanoNov 18, 2021
  9. 5/8 fetch: protect branches checked out in all worktreesAnders Kaseorg, Nov 13, 2021
  10. Junio C HamanoNov 16, 2021
  11. Anders KaseorgNov 16, 2021
  12. Johannes SchindelinNov 22, 2021
  13. 3/8 branch: lowercase error messagesAnders Kaseorg, Nov 13, 2021
  14. 6/8 receive-pack: clean dead code from update_worktree()Anders Kaseorg, Nov 13, 2021
  15. Junio C HamanoNov 16, 2021
  16. 7/8 receive-pack: protect current branch for bare repository worktreeAnders Kaseorg, Nov 13, 2021
  17. 4/8 worktree: simplify find_shared_symref() memory ownership modelAnders Kaseorg, Nov 13, 2021
  18. Junio C HamanoNov 16, 2021
  19. Johannes SchindelinNov 22, 2021
  20. 8/8 branch: protect branches checked out in all worktreesAnders Kaseorg, Nov 13, 2021
  21. Johannes SchindelinNov 22, 2021

Read the whole thread, see it on lore, or plain text.

$ cat FOOTERMessages come from the public archive at lore.kernel.org/git, fetched every hour. The front page is chosen and written each morning by an AI editor and can be wrong; the threads themselves are the record. About and API. For agents: an MCP server at https://gitlist.dev/mcp, and any thread, story or person page as Markdown by adding .md to its URL (or sending Accept: text/markdown). Details in /llms.txt.