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

[PATCH v1 04/10] packfile: inline cache_or_unpack_entry

From
EWEric Wong <e@80x24.org>
Date
Jul 15, 2024, 00:35 UTC
Message-ID
<20240715003519.2671385-5-e@80x24.org>
In-Reply-To
<20240715003519.2671385-1-e@80x24.org>

We need to check delta_base_cache anyways to fill in the `whence' field in `struct object_info'. Inlining cache_or_unpack_entry() makes it easier to only do the hashmap lookup once and avoid a redundant lookup later on.

This code reorganization will also make an optimization to use the cache entry directly easier to implement.

Signed-off-by: Eric Wong <e@80x24.org>
---
 packfile.c | 48 +++++++++++++++++++++---------------------------
 1 file changed, 21 insertions(+), 27 deletions(-)
diff --git a/packfile.c b/packfile.c
index 371da96cdb..1a409ec142 100644
--- a/packfile.c
+++ b/packfile.c
@@ -1444,23 +1444,6 @@ static void detach_delta_base_cache_entry(struct delta_base_cache_entry *ent)
 	free(ent);
 }
 
-static void *cache_or_unpack_entry(struct repository *r, struct packed_git *p,
-				   off_t base_offset, unsigned long *base_size,
-				   enum object_type *type)
-{
-	struct delta_base_cache_entry *ent;
-
-	ent = get_delta_base_cache_entry(p, base_offset);
-	if (!ent)
-		return unpack_entry(r, p, base_offset, type, base_size);
-
-	if (type)
-		*type = ent->type;
-	if (base_size)
-		*base_size = ent->size;
-	return xmemdupz(ent->data, ent->size);
-}
-
 static inline void release_delta_base_cache(struct delta_base_cache_entry *ent)
 {
 	free(ent->data);
@@ -1521,21 +1504,36 @@ int packed_object_info(struct repository *r, struct packed_git *p,
 		       off_t obj_offset, struct object_info *oi)
 {
 	struct pack_window *w_curs = NULL;
-	unsigned long size;
 	off_t curpos = obj_offset;
 	enum object_type type;
+	struct delta_base_cache_entry *ent;
 
 	/*
 	 * We always get the representation type, but only convert it to
 	 * a "real" type later if the caller is interested. Likewise...
 	 * tbd.
 	 */
-	if (oi->contentp && !oi->content_limit) {
-		*oi->contentp = cache_or_unpack_entry(r, p, obj_offset, oi->sizep,
-						      &type);
+	oi->whence = OI_PACKED;
+	ent = get_delta_base_cache_entry(p, obj_offset);
+	if (ent) {
+		oi->whence = OI_DBCACHED;
+		type = ent->type;
+		if (oi->sizep)
+			*oi->sizep = ent->size;
+		if (oi->contentp) {
+			if (!oi->content_limit ||
+					ent->size <= oi->content_limit)
+				*oi->contentp = xmemdupz(ent->data, ent->size);
+			else
+				*oi->contentp = NULL; /* caller must stream */
+		}
+	} else if (oi->contentp && !oi->content_limit) {
+		*oi->contentp = unpack_entry(r, p, obj_offset, &type,
+						oi->sizep);
 		if (!*oi->contentp)
 			type = OBJ_BAD;
 	} else {
+		unsigned long size;
 		type = unpack_object_header(p, &w_curs, &curpos, &size);
 
 		if (oi->sizep) {
@@ -1559,8 +1557,8 @@ int packed_object_info(struct repository *r, struct packed_git *p,
 
 		if (oi->contentp) {
 			if (oi->sizep && *oi->sizep <= oi->content_limit) {
-				*oi->contentp = cache_or_unpack_entry(r, p, obj_offset,
-								      oi->sizep, &type);
+				*oi->contentp = unpack_entry(r, p, obj_offset,
+							&type, oi->sizep);
 				if (!*oi->contentp)
 					type = OBJ_BAD;
 			} else {
@@ -1609,10 +1607,6 @@ int packed_object_info(struct repository *r, struct packed_git *p,
 		} else
 			oidclr(oi->delta_base_oid, the_repository->hash_algo);
 	}
-
-	oi->whence = in_delta_base_cache(p, obj_offset) ? OI_DBCACHED :
-							  OI_PACKED;
-
 out:
 	unuse_pack(&w_curs);
 	return type;
Previous: Eric WongNext: Eric Wong
Message 10 of 51 in “cat-file speedups”
  1. 00/10 cat-file speedupsEric Wong, Jul 15, 2024
  2. 01/10 packfile: move sizep computationEric Wong, Jul 15, 2024
  3. Patrick SteinhardtJul 24, 2024
  4. 02/10 packfile: allow content-limit for cat-fileEric Wong, Jul 15, 2024
  5. Patrick SteinhardtJul 24, 2024
  6. Eric WongJul 26, 2024
  7. 03/10 packfile: fix off-by-one in content_limit comparisonEric Wong, Jul 15, 2024
  8. Patrick SteinhardtJul 24, 2024
  9. Eric WongJul 26, 2024
  10. 04/10 packfile: inline cache_or_unpack_entryEric Wong, Jul 15, 2024
  11. 05/10 cat-file: use delta_base_cache entries directlyEric Wong, Jul 15, 2024
  12. Patrick SteinhardtJul 24, 2024
  13. Eric WongJul 26, 2024
  14. assert vs BUG [was: [PATCH v1 05/10] cat-file: use delta_base_cache entries directly]Eric Wong, Aug 18, 2024
  15. Junio C HamanoAug 19, 2024
  16. 06/10 packfile: packed_object_info avoids packed_to_object_typeEric Wong, Jul 15, 2024
  17. Patrick SteinhardtJul 24, 2024
  18. Eric WongJul 26, 2024
  19. 07/10 object_info: content_limit only applies to blobsEric Wong, Jul 15, 2024
  20. 08/10 cat-file: batch-command uses content_limitEric Wong, Jul 15, 2024
  21. 09/10 cat-file: batch_write: use size_t for lengthEric Wong, Jul 15, 2024
  22. 10/10 cat-file: use writev(2) if availableEric Wong, Jul 15, 2024
  23. Patrick SteinhardtJul 24, 2024
  24. 00/10 cat-file speedupsEric Wong, Aug 23, 2024
  25. 01/10 packfile: move sizep computationEric Wong, Aug 23, 2024
  26. Taylor BlauSep 17, 2024
  27. 02/10 packfile: allow content-limit for cat-fileEric Wong, Aug 23, 2024
  28. Junio C HamanoAug 26, 2024
  29. Eric WongAug 27, 2024
  30. Taylor BlauSep 17, 2024
  31. Junio C HamanoSep 17, 2024
  32. 03/10 packfile: fix off-by-one in content_limit comparisonEric Wong, Aug 23, 2024
  33. Junio C HamanoAug 26, 2024
  34. Taylor BlauSep 17, 2024
  35. 04/10 packfile: inline cache_or_unpack_entryEric Wong, Aug 23, 2024
  36. Junio C HamanoAug 26, 2024
  37. Eric WongOct 6, 2024
  38. 05/10 cat-file: use delta_base_cache entries directlyEric Wong, Aug 23, 2024
  39. Junio C HamanoAug 26, 2024
  40. Junio C HamanoAug 26, 2024
  41. 06/10 packfile: packed_object_info avoids packed_to_object_typeEric Wong, Aug 23, 2024
  42. Junio C HamanoAug 26, 2024
  43. 07/10 object_info: content_limit only applies to blobsEric Wong, Aug 23, 2024
  44. Junio C HamanoAug 26, 2024
  45. 08/10 cat-file: batch-command uses content_limitEric Wong, Aug 23, 2024
  46. Junio C HamanoAug 26, 2024
  47. 09/10 cat-file: batch_write: use size_t for lengthEric Wong, Aug 23, 2024
  48. Junio C HamanoAug 27, 2024
  49. 10/10 cat-file: use writev(2) if availableEric Wong, Aug 23, 2024
  50. Junio C HamanoAug 27, 2024
  51. Junio C HamanoAug 27, 2024

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.