[PATCH 06/13] object-file: move loose object cache into loose source
- From
Patrick Steinhardt <ps@pks.im>
- Date
- Oct 24, 2025, 09:56 UTC
- Message-ID
- <20251024-b4-pks-odb-loose-backend-v1-6-1a4202273c38@pks.im>
- In-Reply-To
- <20251024-b4-pks-odb-loose-backend-v1-0-1a4202273c38@pks.im>
Our loose objects use a cache that (optionally) stores all objects for each of the opened sharding directories. This cache is located in the `struct odb_source`, but now that we have `struct odb_loose_source` it makes sense to move it into the latter structure so that all state that relates to loose objects is entirely self-contained.
Do so. While at it, rename corresponding functions to have a prefix that relates to `struct odb_loose_source`.
Note that despite this prefix, the functions still accept a `struct odb_source` as input. This is done intentionally: once we introduce pluggable object databases, we will continue to accept this struct but then do a cast inside these functions to `struct odb_loose_source`. This design is similar to how we do it for our ref backends.
Signed-off-by: Patrick Steinhardt <ps@pks.im> --- loose.c | 9 +++++---- object-file.c | 35 +++++++++++++++++++---------------- object-file.h | 16 ++++++++++++++-- object-name.c | 2 +- odb.c | 1 - odb.h | 12 ------------ 6 files changed, 39 insertions(+), 36 deletions(-)
diff --git a/loose.c b/loose.c index e8ea6e7e24b..8cc7573ff2b 100644 --- a/loose.c +++ b/loose.c @@ -1,6 +1,7 @@ #include "git-compat-util.h" #include "hash.h" #include "path.h" +#include "object-file.h" #include "odb.h" #include "hex.h" #include "repository.h" @@ -54,7 +55,7 @@ static int insert_loose_map(struct odb_source *source, inserted |= insert_oid_pair(map->to_compat, oid, compat_oid); inserted |= insert_oid_pair(map->to_storage, compat_oid, oid); if (inserted) - oidtree_insert(source->loose_objects_cache, compat_oid); + oidtree_insert(source->loose->cache, compat_oid); return inserted; } @@ -66,9 +67,9 @@ static int load_one_loose_object_map(struct repository *repo, struct odb_source if (!source->loose_map) loose_object_map_init(&source->loose_map); - if (!source->loose_objects_cache) { - ALLOC_ARRAY(source->loose_objects_cache, 1); - oidtree_init(source->loose_objects_cache); + if (!source->loose->cache) { + ALLOC_ARRAY(source->loose->cache, 1); + oidtree_init(source->loose->cache); } insert_loose_map(source, repo->hash_algo->empty_tree, repo->compat_hash_algo->empty_tree); diff --git a/object-file.c b/object-file.c index 38e09262e42..d7c6b1316cc 100644 --- a/object-file.c +++ b/object-file.c @@ -223,7 +223,7 @@ static int quick_has_loose(struct repository *r, odb_prepare_alternates(r->objects); for (source = r->objects->sources; source; source = source->next) { - if (oidtree_contains(odb_loose_cache(source, oid), oid)) + if (oidtree_contains(odb_loose_source_cache(source, oid), oid)) return 1; } return 0; @@ -1802,44 +1802,44 @@ static int append_loose_object(const struct object_id *oid, return 0; } -struct oidtree *odb_loose_cache(struct odb_source *source, - const struct object_id *oid) +struct oidtree *odb_loose_source_cache(struct odb_source *source, + const struct object_id *oid) { int subdir_nr = oid->hash[0]; struct strbuf buf = STRBUF_INIT; - size_t word_bits = bitsizeof(source->loose_objects_subdir_seen[0]); + size_t word_bits = bitsizeof(source->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(source->loose_objects_subdir_seen)) + (size_t) subdir_nr >= bitsizeof(source->loose->subdir_seen)) BUG("subdir_nr out of range"); - bitmap = &source->loose_objects_subdir_seen[word_index]; + bitmap = &source->loose->subdir_seen[word_index]; if (*bitmap & mask) - return source->loose_objects_cache; - if (!source->loose_objects_cache) { - ALLOC_ARRAY(source->loose_objects_cache, 1); - oidtree_init(source->loose_objects_cache); + return source->loose->cache; + if (!source->loose->cache) { + ALLOC_ARRAY(source->loose->cache, 1); + oidtree_init(source->loose->cache); } strbuf_addstr(&buf, source->path); for_each_file_in_obj_subdir(subdir_nr, &buf, source->odb->repo->hash_algo, append_loose_object, NULL, NULL, - source->loose_objects_cache); + source->loose->cache); *bitmap |= mask; strbuf_release(&buf); - return source->loose_objects_cache; + return source->loose->cache; } void odb_clear_loose_cache(struct odb_source *source) { - oidtree_clear(source->loose_objects_cache); - FREE_AND_NULL(source->loose_objects_cache); - memset(&source->loose_objects_subdir_seen, 0, - sizeof(source->loose_objects_subdir_seen)); + oidtree_clear(source->loose->cache); + FREE_AND_NULL(source->loose->cache); + memset(&source->loose->subdir_seen, 0, + sizeof(source->loose->subdir_seen)); } static int check_stream_oid(git_zstream *stream, @@ -2006,5 +2006,8 @@ struct odb_loose_source *odb_loose_source_new(struct odb_source *source) void odb_loose_source_free(struct odb_loose_source *source) { + if (!source) + return; + odb_clear_loose_cache(source->source); free(source); } diff --git a/object-file.h b/object-file.h index aa84fc2a752..9ecb26a4b35 100644 --- a/object-file.h +++ b/object-file.h @@ -20,6 +20,18 @@ struct odb_source; struct odb_loose_source { struct odb_source *source; + + /* + * Used to store the results of readdir(3) calls when we are OK + * sacrificing accuracy due to races for speed. That includes + * object existence with OBJECT_INFO_QUICK, as well as + * our search for unique abbreviated hashes. Don't use it for tasks + * requiring greater accuracy! + * + * Be sure to call odb_load_loose_cache() before using. + */ + uint32_t subdir_seen[8]; /* 256 bits */ + struct oidtree *cache; }; struct odb_loose_source *odb_loose_source_new(struct odb_source *source); @@ -29,8 +41,8 @@ void odb_loose_source_free(struct odb_loose_source *source); * Populate and return the loose object cache array corresponding to the * given object ID. */ -struct oidtree *odb_loose_cache(struct odb_source *source, - const struct object_id *oid); +struct oidtree *odb_loose_source_cache(struct odb_source *source, + const struct object_id *oid); /* Empty the loose object cache for the specified object directory. */ void odb_clear_loose_cache(struct odb_source *source); diff --git a/object-name.c b/object-name.c index f6902e140dd..77e33e693aa 100644 --- a/object-name.c +++ b/object-name.c @@ -116,7 +116,7 @@ static void find_short_object_filename(struct disambiguate_state *ds) struct odb_source *source; for (source = ds->repo->objects->sources; source && !ds->ambiguous; source = source->next) - oidtree_each(odb_loose_cache(source, &ds->bin_pfx), + oidtree_each(odb_loose_source_cache(source, &ds->bin_pfx), &ds->bin_pfx, ds->len, match_prefix, ds); } diff --git a/odb.c b/odb.c index f1b250ceffe..5a5e770dcd0 100644 --- a/odb.c +++ b/odb.c @@ -370,7 +370,6 @@ static void odb_source_free(struct odb_source *source) { free(source->path); odb_loose_source_free(source->loose); - odb_clear_loose_cache(source); loose_object_map_clear(&source->loose_map); free(source); } diff --git a/odb.h b/odb.h index f7e1bf87803..ad57193c66a 100644 --- a/odb.h +++ b/odb.h @@ -51,18 +51,6 @@ struct odb_source { /* Private state for loose objects. */ struct odb_loose_source *loose; - /* - * Used to store the results of readdir(3) calls when we are OK - * sacrificing accuracy due to races for speed. That includes - * object existence with OBJECT_INFO_QUICK, as well as - * our search for unique abbreviated hashes. Don't use it for tasks - * requiring greater accuracy! - * - * Be sure to call odb_load_loose_cache() before using. - */ - uint32_t loose_objects_subdir_seen[8]; /* 256 bits */ - struct oidtree *loose_objects_cache; - /* Map between object IDs for loose objects. */ struct loose_object_map *loose_map;
-- 2.51.1.930.gacf6e81ea2.dirty