[PATCH 0/2] odb/source-loose: fix failing existence checks caused by stale cache

[PATCH 0/2] odb/source-loose: fix failing existence checks caused by stale cache

[Patrick Steinhardt]

Hi,

this small patch series fixes the bug reported in [1].

In summary, when git-merge(1) writes merged blobs it may be unable to
read them back because the loose object cache does not get updated.
Consequently, we may think that the blobs do not exist locally, and thus
in a promisor remote we may perform a backfill fetch. But as the remote
doesn't know those objects we'll fail.

Thanks!

Patrick

[1]: yrjh1p92qc7w.fsf@pohly-mobl1.fritz.box

---
Patrick Steinhardt (2):
      odb/source-loose: extract logic to retrieve subdirectory bitmap
      odb/source-loose: fix existence checks for newly written objects

 odb/source-loose.c       | 54 +++++++++++++++++++++++++++++++++++++++---------
 t/t5616-partial-clone.sh | 43 ++++++++++++++++++++++++++++++++++++++
 2 files changed, 87 insertions(+), 10 deletions(-)


---
base-commit: 750f5de7dfe9a4062f95088212b063f189d50e91
change-id: 20261104-pks-odb-loose-stale-cache-c16107987290

[PATCH 1/2] odb/source-loose: extract logic to retrieve subdirectory bitmap

[Patrick Steinhardt]

The loose object cache tracks which sharding directories have already
been scanned via a bitmap. The logic to retrieve the bitmap and mask for
a given object ID is inlined into the function that populates the cache,
but we're about to add a second function that will need to perform the
same logic.

Extract the logic into a separate function to prepare for this.

Signed-off-by: Patrick Steinhardt <ps@pks.im>
---
 odb/source-loose.c | 24 +++++++++++++++++-------
 1 file changed, 17 insertions(+), 7 deletions(-)

diff --git a/odb/source-loose.c b/odb/source-loose.c
index 50cb388c352..da8f9767257 100644
--- a/odb/source-loose.c
+++ b/odb/source-loose.c
@@ -26,29 +26,39 @@ static int append_loose_object(const struct object_id *oid,
 	return 0;
 }
 
-static struct oidtree *odb_source_loose_cache(struct odb_source_loose *loose,
-					      const struct object_id *oid)
+static uint32_t *subdir_seen_bitmap(struct odb_source_loose *loose,
+				    const struct object_id *oid,
+				    uint32_t *mask)
 {
 	int subdir_nr = oid->hash[0];
-	struct strbuf buf = STRBUF_INIT;
 	size_t word_bits = bitsizeof(loose->subdir_seen[0]);
 	size_t word_index = subdir_nr / word_bits;
-	size_t mask = (size_t)1u << (subdir_nr % word_bits);
-	uint32_t *bitmap;
 
 	if (subdir_nr < 0 ||
 	    (size_t) subdir_nr >= bitsizeof(loose->subdir_seen))
 		BUG("subdir_nr out of range");
 
-	bitmap = &loose->subdir_seen[word_index];
+	*mask = (size_t)1u << (subdir_nr % word_bits);
+
+	return &loose->subdir_seen[word_index];
+}
+
+static struct oidtree *odb_source_loose_cache(struct odb_source_loose *loose,
+					      const struct object_id *oid)
+{
+	struct strbuf buf = STRBUF_INIT;
+	uint32_t *bitmap, mask;
+
+	bitmap = subdir_seen_bitmap(loose, oid, &mask);
 	if (*bitmap & mask)
 		return loose->cache;
+
 	if (!loose->cache) {
 		ALLOC_ARRAY(loose->cache, 1);
 		oidtree_init(loose->cache);
 	}
 	strbuf_addstr(&buf, loose->base.path);
-	for_each_file_in_obj_subdir(subdir_nr, &buf,
+	for_each_file_in_obj_subdir(oid->hash[0], &buf,
 				    loose->base.odb->repo->hash_algo,
 				    append_loose_object,
 				    NULL, NULL,

-- 
2.56.0.500.g5d06dc02cd.dirty

[PATCH 2/2] odb/source-loose: fix existence checks for newly written objects

[Patrick Steinhardt]

The loose object source maintains a cache of object IDs that it uses to
answer existence checks fast. This cache is lazily populated: whenever
we are asked about a specific object ID, we scan the whole sharding
directory it would live in and populate the cache with the objects we
find. Subsequent existence checks are then answered from the cache
without hitting the filesystem.

But when writing a new loose object we don't update that cache. That's
fine in the case where the cache for its sharding directory has not yet
been populated, because we would then populate it on a subsequent read
and thus surface the newly written object. But when it _is_ populated
already, then any subsequent existence check will claim that the object
does not exist at all until the cache is flushed.

This issue can cause git-merge(1) to die in a partial clone because we
try to fetch an object that we have just written, as reported in [1]:

  1. `prefetch_for_content_merges()` performs quick existence checks for
     all blobs that need a content merge to figure out which ones need
     to be fetched from the promisor remote. These lookups populate the
     loose object cache for the respective sharding directories.

  2. We lazily fetch the missing blobs, which writes a new packfile.

  3. We perform the merge and write the resulting blobs as loose
     objects. The cache is not updated, so for all blobs whose sharding
     directory was populated in (1), the object existence checks will
     fail.

  4. When checking out the merge result, `prefetch_cache_entries()`
     performs existence checks for all index entries that need to be
     written to the worktree. But due to the stale cache, the merged
     blobs are reported as missing, and consequently we try to fetch
     them from the promisor remote. This fetch fails because the remote
     will most likely not have them.

This issue bisects to 5946a564cdd (odb/source: make `read_object_info()`
function pluggable, 2026-03-05), because in that commit we stopped
flushing the loose object cache when we try to read a missing object.
That was done intentionally though: a failing read is only retried when
it is a non-QUICK read, whereas the cache only affects QUICK reads. So
flushing the cache for a non-QUICK read wouldn't ever cause us to find
that object.

So the logic described here only ever worked by chance, as the cache was
incoherent after writes even before that commit. It only used to work
because of the interleaving sequence of QUICK and non-QUICK reads: when
we read the lazily-fetched blobs in (3) via a non-QUICK read, the first
read fails because the new packfile has not yet been added to the
"packed" source. On the retry we then used to flush the loose object
cache, and that in turn means that the subsequent existence check in (4)
rescanned the sharding directories and found the merged blobs.

We could in theory fix this by flushing all the loose object caches
again, like we used to do before 5946a564cdd. But that wouldn't address
the root cause: the cache would still be stale unless somebody does a
non-QUICK read for a missing object. Instead, fix the bug by updating
the cache whenever we write a new loose object.

[1]: <yrjh1p92qc7w.fsf@pohly-mobl1.fritz.box>

Reported-by: Patrick Ohly <patrick.ohly@intel.com>
Signed-off-by: Patrick Steinhardt <ps@pks.im>
---
 odb/source-loose.c       | 30 +++++++++++++++++++++++++++---
 t/t5616-partial-clone.sh | 43 +++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 70 insertions(+), 3 deletions(-)

diff --git a/odb/source-loose.c b/odb/source-loose.c
index da8f9767257..bcbd29ce3a0 100644
--- a/odb/source-loose.c
+++ b/odb/source-loose.c
@@ -68,6 +68,19 @@ static struct oidtree *odb_source_loose_cache(struct odb_source_loose *loose,
 	return loose->cache;
 }
 
+/*
+ * Mark an object as seen in the loose cache. Note that we only do this in case
+ * the particular subdirectory was scanned already. If it wasn't, then the next
+ * read would populate it anyway.
+ */
+static void loose_cache_mark_seen(struct odb_source_loose *loose,
+				  const struct object_id *oid)
+{
+	uint32_t mask, *bitmap = subdir_seen_bitmap(loose, oid, &mask);
+	if (*bitmap & mask)
+		oidtree_insert(loose->cache, oid, NULL);
+}
+
 static int quick_has_loose(struct odb_source_loose *loose,
 			   const struct object_id *oid)
 {
@@ -824,8 +837,13 @@ static int write_loose_object(struct odb_source_loose *loose,
 			warning_errno(_("failed utime() on %s"), tmp_file.buf);
 	}
 
-	return finalize_object_file_flags(loose->base.odb->repo, tmp_file.buf, filename.buf,
-					  FOF_SKIP_COLLISION_CHECK);
+	ret = finalize_object_file_flags(loose->base.odb->repo, tmp_file.buf, filename.buf,
+					 FOF_SKIP_COLLISION_CHECK);
+	if (ret < 0)
+		return ret;
+
+	loose_cache_mark_seen(loose, oid);
+	return 0;
 }
 
 static int odb_source_loose_write_object(struct odb_source *source,
@@ -967,8 +985,14 @@ static int odb_source_loose_write_object_stream(struct odb_source *source,
 
 	err = finalize_object_file_flags(loose->base.odb->repo, tmp_file.buf, filename.buf,
 					 FOF_SKIP_COLLISION_CHECK);
-	if (!err && compat)
+	if (err < 0)
+		goto cleanup;
+
+	loose_cache_mark_seen(loose, oid);
+
+	if (compat)
 		err = repo_add_loose_object_map(loose, oid, &compat_oid);
+
 cleanup:
 	strbuf_release(&tmp_file);
 	strbuf_release(&filename);
diff --git a/t/t5616-partial-clone.sh b/t/t5616-partial-clone.sh
index 7cc6f58a31f..9636652362f 100755
--- a/t/t5616-partial-clone.sh
+++ b/t/t5616-partial-clone.sh
@@ -747,6 +747,49 @@ test_expect_success 'after fetching descendants of non-promisor commits, gc work
 	git -C partial gc --prune=now
 '
 
+write_many_mergeable_files () {
+	for i in $(test_seq 50)
+	do
+		{
+			echo "$1 $i" &&
+			test_seq 8 | sed "s/^/line /" &&
+			echo "$2 $i"
+		} >"$3/file-$i" || return 1
+	done
+}
+
+test_expect_success 'merge does not lazy-fetch locally written objects' '
+	test_when_finished "rm -rf server client merged" &&
+	git init server &&
+	(
+		cd server &&
+		git config set uploadpack.allowfilter 1 &&
+
+		write_many_mergeable_files top-A bottom-A . &&
+		git add . &&
+		git commit -m A &&
+
+		write_many_mergeable_files top-B bottom-A . &&
+		git commit -a -m B &&
+
+		git checkout -b C main~1 &&
+		write_many_mergeable_files top-A bottom-C . &&
+		echo C >only-in-C &&
+		git add . &&
+		git commit -m C
+	) &&
+
+	git clone --filter=blob:none --branch=main "file://$(pwd)/server" client &&
+	git -C client merge origin/C &&
+
+	mkdir merged &&
+	write_many_mergeable_files top-B bottom-C merged &&
+	for m in merged/*
+	do
+		test_cmp "$m" client/"$(basename "$m")" || return 1
+	done &&
+	test_cmp server/only-in-C client/only-in-C
+'
 
 . "$TEST_DIRECTORY"/lib-httpd.sh
 start_httpd

-- 
2.56.0.500.g5d06dc02cd.dirty

Re: [PATCH 0/2] odb/source-loose: fix failing existence checks caused by stale cache

[Kristofer Karlsson]

On Wed, 4 Nov 2026 at 13:17, Patrick Steinhardt <ps@pks.im> wrote:
>
> Hi,
>
> this small patch series fixes the bug reported in [1].
>
> In summary, when git-merge(1) writes merged blobs it may be unable to
> read them back because the loose object cache does not get updated.
> Consequently, we may think that the blobs do not exist locally, and thus
> in a promisor remote we may perform a backfill fetch. But as the remote
> doesn't know those objects we'll fail.
>
> Thanks!
>
> Patrick

The fix looks good to me, and I verified that the test fails
without it. My own local fix is logically equivalent,
which is a good sign, though perhaps that biased my review.

Thanks,
Kristofer

Re: [PATCH 0/2] odb/source-loose: fix failing existence checks caused by stale cache

[Derrick Stolee]

On 11/4/2026 7:57 AM, Kristofer Karlsson wrote:
> On Wed, 4 Nov 2026 at 13:17, Patrick Steinhardt <ps@pks.im> wrote:
>>
>> Hi,
>>
>> this small patch series fixes the bug reported in [1].
>>
>> In summary, when git-merge(1) writes merged blobs it may be unable to
>> read them back because the loose object cache does not get updated.
>> Consequently, we may think that the blobs do not exist locally, and thus
>> in a promisor remote we may perform a backfill fetch. But as the remote
>> doesn't know those objects we'll fail.
>>
>> Thanks!
>>
>> Patrick
> 
> The fix looks good to me, and I verified that the test fails
> without it. My own local fix is logically equivalent,
> which is a good sign, though perhaps that biased my review.
I also think the fix looks good. (I haven't verified it locally
like Kristofer did, but the test is good.)

Thanks,
-Stolee

Re: [PATCH 2/2] odb/source-loose: fix existence checks for newly written objects

[Junio C Hamano]

Patrick Steinhardt <ps@pks.im> writes:

> @@ -824,8 +837,13 @@ static int write_loose_object(struct odb_source_loose *loose,
>  			warning_errno(_("failed utime() on %s"), tmp_file.buf);
>  	}
>  
> -	return finalize_object_file_flags(loose->base.odb->repo, tmp_file.buf, filename.buf,
> -					  FOF_SKIP_COLLISION_CHECK);
> +	ret = finalize_object_file_flags(loose->base.odb->repo, tmp_file.buf, filename.buf,
> +					 FOF_SKIP_COLLISION_CHECK);
> +	if (ret < 0)
> +		return ret;
> +
> +	loose_cache_mark_seen(loose, oid);
> +	return 0;
>  }
>  
>  static int odb_source_loose_write_object(struct odb_source *source,
> @@ -967,8 +985,14 @@ static int odb_source_loose_write_object_stream(struct odb_source *source,
>  
>  	err = finalize_object_file_flags(loose->base.odb->repo, tmp_file.buf, filename.buf,
>  					 FOF_SKIP_COLLISION_CHECK);
> -	if (!err && compat)
> +	if (err < 0)
> +		goto cleanup;
> +
> +	loose_cache_mark_seen(loose, oid);
> +
> +	if (compat)
>  		err = repo_add_loose_object_map(loose, oid, &compat_oid);
> +
>  cleanup:
>  	strbuf_release(&tmp_file);
>  	strbuf_release(&filename);

Not about this change but the lack of "if (compat)" stuff in the
previous hunk puzzled me.  It turns out that the logic to write a
single loose object is so big that write_loose_object(), which is
the subject of the previous hunk, helps another function,
odb_source_loose_write_object(), by being its helper function, and
this caller has the corresponding "if (compat)" logic to produce
the hash interoperability data.

Looking very good.