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

[PATCH 10/17] Remove ineffective optimization

From
mhagger@alum.mit.edu <mhagger@alum.mit.edu>
Date
Aug 23, 2012, 08:10 UTC
Message-ID
<1345709442-16046-11-git-send-email-mhagger@alum.mit.edu>
In-Reply-To
<1345709442-16046-1-git-send-email-mhagger@alum.mit.edu>
From: Michael Haggerty <mhagger@alum.mit.edu>

I cannot find a scenario in which this function is called any significant number of times, so simplify the code by always allocating an array for return_refs rather than trying to use a stack-allocated array for small lists.

Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
---
 builtin/fetch-pack.c | 14 +++-----------
 1 file changed, 3 insertions(+), 11 deletions(-)
diff --git a/builtin/fetch-pack.c b/builtin/fetch-pack.c
index 0426cf4..9398059 100644
--- a/builtin/fetch-pack.c
+++ b/builtin/fetch-pack.c
@@ -527,17 +527,10 @@ static void filter_refs(struct ref **refs, int *nr_heads, char **heads)
 	struct ref *newlist = NULL;
 	struct ref **newtail = &newlist;
 	struct ref *ref, *next;
-	struct ref *fastarray[32];
 	int match_pos;
 
-	if (*nr_heads && !args.fetch_all) {
-		if (ARRAY_SIZE(fastarray) < *nr_heads)
-			return_refs = xcalloc(*nr_heads, sizeof(struct ref *));
-		else {
-			return_refs = fastarray;
-			memset(return_refs, 0, sizeof(struct ref *) * *nr_heads);
-		}
-	}
+	if (*nr_heads && !args.fetch_all)
+		return_refs = xcalloc(*nr_heads, sizeof(struct ref *));
 	else
 		return_refs = NULL;
 
@@ -584,8 +577,7 @@ static void filter_refs(struct ref **refs, int *nr_heads, char **heads)
 				newtail = &ref->next;
 			}
 		}
-		if (return_refs != fastarray)
-			free(return_refs);
+		free(return_refs);
 	}
 	*refs = newlist;
 }

-- 
1.7.11.3
Previous: mhagger@alum.mit.eduNext: mhagger@alum.mit.edu
Message 21 of 40 in “Clean up how fetch_pack() handles the heads list”
  1. 00/17 Clean up how fetch_pack() handles the heads listmhagger@alum.mit.edu, Aug 23, 2012
  2. 01/17 t5500: add tests of error output for missing refsmhagger@alum.mit.edu, Aug 23, 2012
  3. 02/17 Rename static function fetch_pack() to http_fetch_pack()mhagger@alum.mit.edu, Aug 23, 2012
  4. 03/17 Fix formatting.mhagger@alum.mit.edu, Aug 23, 2012
  5. 04/17 Name local variables more consistentlymhagger@alum.mit.edu, Aug 23, 2012
  6. Jeff KingAug 23, 2012
  7. Michael HaggertyAug 24, 2012
  8. Junio C HamanoAug 26, 2012
  9. Michael HaggertyAug 27, 2012
  10. Jeff KingAug 27, 2012
  11. Junio C HamanoAug 27, 2012
  12. Junio C HamanoAug 27, 2012
  13. 05/17 Do not check the same match_pos twicemhagger@alum.mit.edu, Aug 23, 2012
  14. Jeff KingAug 23, 2012
  15. 06/17 Let fetch_pack() inform caller about number of unique headsmhagger@alum.mit.edu, Aug 23, 2012
  16. Jeff KingAug 23, 2012
  17. Michael HaggertyAug 25, 2012
  18. 07/17 Pass nr_heads to do_pack_ref() by referencemhagger@alum.mit.edu, Aug 23, 2012
  19. 08/17 Pass nr_heads to everything_local() by referencemhagger@alum.mit.edu, Aug 23, 2012
  20. 09/17 Pass nr_heads to filter_refs() by referencemhagger@alum.mit.edu, Aug 23, 2012
  21. 10/17 Remove ineffective optimizationmhagger@alum.mit.edu, Aug 23, 2012
  22. 11/17 filter_refs(): do not leave gaps in return_refsmhagger@alum.mit.edu, Aug 23, 2012
  23. 12/17 filter_refs(): compress unmatched refs in heads arraymhagger@alum.mit.edu, Aug 23, 2012
  24. 13/17 cmd_fetch_pack: return early if finish_connect() returns an errormhagger@alum.mit.edu, Aug 23, 2012
  25. 14/17 Report missing refs even if no existing refs were receivedmhagger@alum.mit.edu, Aug 23, 2012
  26. 15/17 cmd_fetch_pack(): simplify computation of return valuemhagger@alum.mit.edu, Aug 23, 2012
  27. 16/17 fetch_pack(): free matching headsmhagger@alum.mit.edu, Aug 23, 2012
  28. Jeff KingAug 23, 2012
  29. 17/17 fetch_refs(): simplify logicmhagger@alum.mit.edu, Aug 23, 2012
  30. Jeff KingAug 23, 2012
  31. Michael HaggertyAug 25, 2012
  32. Jeff KingAug 23, 2012
  33. Philip OakleyAug 23, 2012
  34. Jeff KingAug 23, 2012
  35. Jeff KingAug 23, 2012
  36. Michael HaggertyAug 25, 2012
  37. Michael HaggertySep 2, 2012
  38. Philip OakleyAug 23, 2012
  39. Junio C HamanoAug 24, 2012
  40. Philip OakleyAug 24, 2012

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.