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

Re: [PATCH 1/3] refs/files: skip updates with errors in batched updates

From
Patrick Steinhardt <ps@pks.im>
Date
Jun 2, 2025, 12:00 UTC
Message-ID
<aD2SRmlSKZm8g8kn@pks.im>
In-Reply-To
<20250602-6769-address-test-failures-in-the-next-branch-caused-by-batched-reference-updates-v1-1-903d1db3f10e@gmail.com>
On Mon, Jun 02, 2025 at 11:57:24AM +0200, Karthik Nayak wrote:
Show 12 quoted lines
> The commit 23fc8e4f61 (refs: implement batch reference update support,
> 2025-04-08) introduced support for batched reference updates. This
> allows users to batch updates together, while allowing some of the
> updates to fail.
> 
> Under the hood, batched updates use the reference transaction mechanism.
> Each update which fails is marked as such. Any failed updates must be
> skipped over in the rest of the code, as they wouldn't apply any more.
> In two of the loops within 'files_transaction_finish()' of the files
> backend, the failed updates aren't skipped over. This can cause a
> SEGFAULT otherwise. Add the missing skips and a test to validate the
> same.
Curious -- we do have tests, so why don't any of them hit this issue?
Show 21 quoted lines
> Signed-off-by: Karthik Nayak <karthik.188@gmail.com>
> ---
>  refs/files-backend.c  |  7 +++++++
>  t/t1400-update-ref.sh | 14 ++++++++++++++
>  2 files changed, 21 insertions(+)
> 
> diff --git a/refs/files-backend.c b/refs/files-backend.c
> index 4d1f65a57a..c4a0f29072 100644
> --- a/refs/files-backend.c
> +++ b/refs/files-backend.c
> @@ -3208,6 +3208,10 @@ static int files_transaction_finish(struct ref_store *ref_store,
>  	 */
>  	for (i = 0; i < transaction->nr; i++) {
>  		struct ref_update *update = transaction->updates[i];
> +
> +		if (update->rejection_err)
> +			continue;
> +
>  		if (update->flags & REF_DELETING &&
>  		    !(update->flags & REF_LOG_ONLY) &&
>  		    !(update->flags & REF_IS_PRUNING)) {

Ok. And the reftable backend doesn't need the same treatment? Probably doesn't because we queue all updates via `queue_transaction_update()`, which knows to not add them to the list of updates in case any error has happened. And in `_finish()` we iterate through that list of queued updates instead of the global list of updates.

Show 20 quoted lines
> diff --git a/t/t1400-update-ref.sh b/t/t1400-update-ref.sh
> index d29d23cb89..e9a605d0ba 100755
> --- a/t/t1400-update-ref.sh
> +++ b/t/t1400-update-ref.sh
> @@ -2299,6 +2299,20 @@ do
>  			test_grep -q "refname conflict" stdout
>  		)
>  	'
> +
> +	test_expect_success "stdin $type batch-updates delete non-existent ref" '
> +		git init repo &&
> +		test_when_finished "rm -fr repo" &&
> +		(
> +			cd repo &&
> +			test_commit commit &&
> +			head=$(git rev-parse HEAD) &&
> +
> +			format_command $type "delete refs/heads/non-existent" "$head" >stdin &&
> +			git update-ref $type --stdin --batch-updates <stdin >stdout &&
> +			test_grep -q "reference does not exist" stdout
We typically don't silence the output of `test_grep`.
Patrick
Previous: Karthik NayakNext: Karthik Nayak
Message 3 of 40 in “refs: fix some bugs with batched-updates”
  1. 0/3 refs: fix some bugs with batched-updatesKarthik Nayak, Jun 2, 2025
  2. 1/3 refs/files: skip updates with errors in batched updatesKarthik Nayak, Jun 2, 2025
  3. Patrick SteinhardtJun 2, 2025
  4. Karthik NayakJun 2, 2025
  5. Junio C HamanoJun 2, 2025
  6. Karthik NayakJun 2, 2025
  7. 2/3 t5516: use double quotes for tests with variablesKarthik Nayak, Jun 2, 2025
  8. Eric SunshineJun 2, 2025
  9. Karthik NayakJun 2, 2025
  10. 3/3 receive-pack: handle reference deletions separatelyKarthik Nayak, Jun 2, 2025
  11. Patrick SteinhardtJun 2, 2025
  12. Karthik NayakJun 2, 2025
  13. Junio C HamanoJun 2, 2025
  14. Patrick SteinhardtJun 2, 2025
  15. Karthik NayakJun 4, 2025
  16. 0/2 refs: fix some bugs with batched-updatesKarthik Nayak, Jun 5, 2025
  17. 1/2 refs/files: skip updates with errors in batched updatesKarthik Nayak, Jun 5, 2025
  18. 2/2 receive-pack: handle reference deletions separatelyKarthik Nayak, Jun 5, 2025
  19. Patrick SteinhardtJun 5, 2025
  20. Karthik NayakJun 5, 2025
  21. 0/2 refs: fix some bugs with batched-updatesKarthik Nayak, Jun 6, 2025
  22. 1/2 refs/files: skip updates with errors in batched updatesKarthik Nayak, Jun 6, 2025
  23. 2/2 receive-pack: handle reference deletions separatelyKarthik Nayak, Jun 6, 2025
  24. Christian CouderJun 12, 2025
  25. Junio C HamanoJun 12, 2025
  26. Karthik NayakJun 13, 2025
  27. 0/2 refs: fix some bugs with batched-updatesKarthik Nayak, Jun 13, 2025
  28. 1/2 refs/files: skip updates with errors in batched updatesKarthik Nayak, Jun 13, 2025
  29. 2/2 receive-pack: handle reference deletions separatelyKarthik Nayak, Jun 13, 2025
  30. Junio C HamanoJun 13, 2025
  31. Karthik NayakJun 19, 2025
  32. Christian CouderJun 13, 2025
  33. Junio C HamanoJun 13, 2025
  34. 0/2 refs: fix some bugs with batched-updatesKarthik Nayak, Jun 20, 2025
  35. 1/2 refs/files: skip updates with errors in batched updatesKarthik Nayak, Jun 20, 2025
  36. 2/2 receive-pack: handle reference deletions separatelyKarthik Nayak, Jun 20, 2025
  37. Junio C HamanoJun 20, 2025
  38. Karthik NayakJun 21, 2025
  39. Junio C HamanoJun 22, 2025
  40. Karthik NayakJun 22, 2025

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.