[PATCH v2 1/7] midx: start tracking per object database source

From

Patrick Steinhardt <ps@pks.im>

Date

Jul 15, 2025, 11:29 UTC

Message-ID

<20250715-b4-pks-midx-via-odb-alternate-v2-1-b0ca0b4b516e@pks.im>

In-Reply-To

<20250715-b4-pks-midx-via-odb-alternate-v2-0-b0ca0b4b516e@pks.im>

This layout causes two problems:

  - Object databases consist of multiple object sources (e.g. one source
    per alternate object directory), where each multi-pack index is
    specific to one of those sources. Regardless of that though, the
    MIDX is not tracked per source, but tracked globally for the whole
    object database. This creates a mismatch between the on-disk layout
    and how things are organized in the object database subsystems and
    makes some parts, like figuring out whether a source has an MIDX,
    quite awkward.

  - Multi-pack indices are an implementation detail of how efficient
    access for packfiles work. As such, they are neither relevant in the
    context of loose objects, nor in a potential future where we have
    pluggable backends.

Signed-off-by: Patrick Steinhardt <ps@pks.im>
---
 midx.c     | 19 +++++++++++--------
 midx.h     |  3 ++-
 odb.h      |  9 ++++++++-
 packfile.c |  4 +++-
 4 files changed, 24 insertions(+), 11 deletions(-)

diff --git a/midx.c b/midx.c
index 3c5bc821730..2f64c26058f 100644
--- a/midx.c
+++ b/midx.c
@@ -724,28 +724,29 @@ int midx_preferred_pack(struct multi_pack_index *m, uint32_t *pack_int_id)
 	return 0;
 }
 
-int prepare_multi_pack_index_one(struct repository *r, const char *object_dir, int local)
+int prepare_multi_pack_index_one(struct odb_source *source, int local)
 {
+	struct repository *r = source->odb->repo;
 	struct multi_pack_index *m;
-	struct multi_pack_index *m_search;
 
 	prepare_repo_settings(r);
 	if (!r->settings.core_multi_pack_index)
 		return 0;
 
-	for (m_search = r->objects->multi_pack_index; m_search; m_search = m_search->next)
-		if (!strcmp(object_dir, m_search->object_dir))
-			return 1;
-
-	m = load_multi_pack_index(r, object_dir, local);
+	if (source->midx)
+		return 1;
 
+	m = load_multi_pack_index(r, source->path, local);
 	if (m) {
 		struct multi_pack_index *mp = r->objects->multi_pack_index;
 		if (mp) {
 			m->next = mp->next;
 			mp->next = m;
-		} else
+		} else {
 			r->objects->multi_pack_index = m;
+		}
+		source->midx = m;
+
 		return 1;
 	}
 
@@ -837,6 +838,8 @@ void clear_midx_file(struct repository *r)
 	if (r->objects && r->objects->multi_pack_index) {
 		close_midx(r->objects->multi_pack_index);
 		r->objects->multi_pack_index = NULL;
+		for (struct odb_source *source = r->objects->sources; source; source = source->next)
+			source->midx = NULL;
 	}
 
 	if (remove_path(midx.buf))
diff --git a/midx.h b/midx.h
index 9d1374cbd58..639a6f50e45 100644
--- a/midx.h
+++ b/midx.h
@@ -8,6 +8,7 @@ struct pack_entry;
 struct repository;
 struct bitmapped_pack;
 struct git_hash_algo;
+struct odb_source;
 
 #define MIDX_SIGNATURE 0x4d494458 /* "MIDX" */
 #define MIDX_VERSION 1
@@ -123,7 +124,7 @@ int fill_midx_entry(struct repository *r, const struct object_id *oid, struct pa
 int midx_contains_pack(struct multi_pack_index *m,
 		       const char *idx_or_pack_name);
 int midx_preferred_pack(struct multi_pack_index *m, uint32_t *pack_int_id);
-int prepare_multi_pack_index_one(struct repository *r, const char *object_dir, int local);
+int prepare_multi_pack_index_one(struct odb_source *source, int local);
 
 /*
  * Variant of write_midx_file which writes a MIDX containing only the packs
diff --git a/odb.h b/odb.h
index e922f256802..f09dba1fe1d 100644
--- a/odb.h
+++ b/odb.h
@@ -13,6 +13,7 @@ struct oidmap;
 struct oidtree;
 struct strbuf;
 struct repository;
+struct multi_pack_index;
 
 /*
  * Compute the exact path an alternate is at and returns it. In case of
@@ -55,6 +56,13 @@ struct odb_source {
 	/* Map between object IDs for loose objects. */
 	struct loose_object_map *loose_map;
 
+	/*
+	 * private data
+	 *
+	 * should only be accessed directly by packfile.c and midx.c
+	 */
+	struct multi_pack_index *midx;
+
 	/*
 	 * This is a temporary object store created by the tmp_objdir
 	 * facility. Disable ref updates since the objects in the store
@@ -75,7 +83,6 @@ struct odb_source {
 };
 
 struct packed_git;
-struct multi_pack_index;
 struct cached_object_entry;
 
 /*
diff --git a/packfile.c b/packfile.c
index af9ccfdba62..8bdd85fc7e7 100644
--- a/packfile.c
+++ b/packfile.c
@@ -372,6 +372,8 @@ void close_object_store(struct object_database *o)
 	if (o->multi_pack_index) {
 		close_midx(o->multi_pack_index);
 		o->multi_pack_index = NULL;
+		for (struct odb_source *source = o->sources; source; source = source->next)
+			source->midx = NULL;
 	}
 
 	close_commit_graph(o);
@@ -1037,7 +1039,7 @@ static void prepare_packed_git(struct repository *r)
 	odb_prepare_alternates(r->objects);
 	for (source = r->objects->sources; source; source = source->next) {
 		int local = (source == r->objects->sources);
-		prepare_multi_pack_index_one(r, source->path, local);
+		prepare_multi_pack_index_one(source, local);
 		prepare_packed_git_one(r, source->path, local);
 	}
 	rearrange_packed_git(r);

-- 
2.50.1.404.ge9779f6434.dirty