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

[PATCH v2 08/11] receive-pack: use oidset to de-duplicate .have lines

From
Jeff King <peff@peff.net>
Date
Feb 8, 2017, 20:53 UTC
Message-ID
<20170208205310.mnjocw4xb22ppl24@sigill.intra.peff.net>
In-Reply-To
<20170208205219.twgm5rggovqbepte@sigill.intra.peff.net>

If you have an alternate object store with a very large number of refs, the peak memory usage of the sha1_array can grow high, even if most of them are duplicates that end up not being printed at all.

The similar for_each_alternate_ref() code-paths in fetch-pack solve this by using flags in "struct object" to de-duplicate (and so are relying on obj_hash at the core).

But we don't have a "struct object" at all in this case. We could call lookup_unknown_object() to get one, but if our goal is reducing memory footprint, it's not great:

 - an unknown object is as large as the largest object type
   (a commit), which is bigger than an oidset entry
 - we can free the memory after our ref advertisement, but
   "struct object" entries persist forever (and the
   receive-pack may hang around for a long time, as the
   bottleneck is often client upload bandwidth).

So let's use an oidset. Note that unlike a sha1-array it doesn't sort the output as a side effect. However, our output is at least stable, because for_each_alternate_ref() will give us the sha1s in ref-sorted order.

In one particularly pathological case with an alternate that has 60,000 unique refs out of 80 million total, this reduced the peak heap usage of "git receive-pack . </dev/null" from 13GB to 14MB.

Signed-off-by: Jeff King <peff@peff.net>
---
 builtin/receive-pack.c | 26 ++++++++++++--------------
 1 file changed, 12 insertions(+), 14 deletions(-)
diff --git a/builtin/receive-pack.c b/builtin/receive-pack.c
index d21332d9e..a4926fcfb 100644
--- a/builtin/receive-pack.c
+++ b/builtin/receive-pack.c
@@ -21,6 +21,7 @@
 #include "sigchain.h"
 #include "fsck.h"
 #include "tmp-objdir.h"
+#include "oidset.h"
 
 static const char * const receive_pack_usage[] = {
 	N_("git receive-pack <git-dir>"),
@@ -271,27 +272,24 @@ static int show_ref_cb(const char *path_full, const struct object_id *oid,
 	return 0;
 }
 
-static int show_one_alternate_sha1(const unsigned char sha1[20], void *unused)
+static void show_one_alternate_ref(const char *refname,
+				   const struct object_id *oid,
+				   void *data)
 {
-	show_ref(".have", sha1);
-	return 0;
-}
+	struct oidset *seen = data;
 
-static void collect_one_alternate_ref(const char *refname,
-				      const struct object_id *oid,
-				      void *data)
-{
-	struct sha1_array *sa = data;
-	sha1_array_append(sa, oid->hash);
+	if (oidset_insert(seen, oid))
+		return;
+
+	show_ref(".have", oid->hash);
 }
 
 static void write_head_info(void)
 {
-	struct sha1_array sa = SHA1_ARRAY_INIT;
+	static struct oidset seen = OIDSET_INIT;
 
-	for_each_alternate_ref(collect_one_alternate_ref, &sa);
-	sha1_array_for_each_unique(&sa, show_one_alternate_sha1, NULL);
-	sha1_array_clear(&sa);
+	for_each_alternate_ref(show_one_alternate_ref, &seen);
+	oidset_clear(&seen);
 	for_each_ref(show_ref_cb, NULL);
 	if (!sent_capabilities)
 		show_ref("capabilities^{}", null_sha1);
-- 
2.12.0.rc0.371.ga6cf8653b
Previous: Jeff KingNext: Jeff King
Message 43 of 48 in “reducing resource usage of for_each_alternate_ref”
  1. 0/12 reducing resource usage of for_each_alternate_refJeff King, Jan 24, 2017
  2. 01/12 for_each_alternate_ref: handle failure from real_pathdup()Jeff King, Jan 24, 2017
  3. Junio C HamanoJan 25, 2017
  4. 02/12 for_each_alternate_ref: stop trimming trailing slashesJeff King, Jan 24, 2017
  5. 03/12 for_each_alternate_ref: use strbuf for path allocationJeff King, Jan 24, 2017
  6. Junio C HamanoJan 25, 2017
  7. Jeff KingJan 25, 2017
  8. 04/12 for_each_alternate_ref: pass name/oid instead of ref structJeff King, Jan 24, 2017
  9. 05/12 for_each_alternate_ref: replace transport code with for-each-refJeff King, Jan 24, 2017
  10. Junio C HamanoJan 25, 2017
  11. 06/12 clone: disable save_commit_bufferJeff King, Jan 24, 2017
  12. Junio C HamanoJan 25, 2017
  13. Jeff KingJan 25, 2017
  14. Jeff KingJan 25, 2017
  15. Jeff KingJan 25, 2017
  16. 07/12 fetch-pack: cache results of for_each_alternate_refJeff King, Jan 24, 2017
  17. Junio C HamanoJan 25, 2017
  18. Jeff KingJan 25, 2017
  19. 08/12 add oidset APIJeff King, Jan 24, 2017
  20. Ramsay JonesJan 24, 2017
  21. Jeff KingJan 24, 2017
  22. 10/12 receive-pack: fix misleading namespace/.have commentJeff King, Jan 24, 2017
  23. 09/12 receive-pack: use oidset to de-duplicate .have linesJeff King, Jan 24, 2017
  24. Junio C HamanoJan 25, 2017
  25. Jeff KingJan 25, 2017
  26. 12/12 receive-pack: avoid duplicates between our refs and alternatesJeff King, Jan 24, 2017
  27. Junio C HamanoJan 25, 2017
  28. Jeff KingJan 25, 2017
  29. 11/12 receive-pack: treat namespace .have lines like alternatesJeff King, Jan 24, 2017
  30. Junio C HamanoJan 25, 2017
  31. Jeff KingJan 25, 2017
  32. Lukas FleischerJan 27, 2017
  33. Jeff KingJan 27, 2017
  34. Junio C HamanoJan 27, 2017
  35. Brandon WilliamsJan 24, 2017
  36. Jeff KingJan 24, 2017
  37. 0/11 reducing resource usage of for_each_alternate_refJeff King, Feb 8, 2017
  38. 01/11 for_each_alternate_ref: handle failure from real_pathdup()Jeff King, Feb 8, 2017
  39. 02/11 for_each_alternate_ref: stop trimming trailing slashesJeff King, Feb 8, 2017
  40. 04/11 for_each_alternate_ref: pass name/oid instead of ref structJeff King, Feb 8, 2017
  41. 03/11 for_each_alternate_ref: use strbuf for path allocationJeff King, Feb 8, 2017
  42. 05/11 for_each_alternate_ref: replace transport code with for-each-refJeff King, Feb 8, 2017
  43. 08/11 receive-pack: use oidset to de-duplicate .have linesJeff King, Feb 8, 2017
  44. 07/11 add oidset APIJeff King, Feb 8, 2017
  45. 06/11 fetch-pack: cache results of for_each_alternate_refJeff King, Feb 8, 2017
  46. 09/11 receive-pack: fix misleading namespace/.have commentJeff King, Feb 8, 2017
  47. 10/11 receive-pack: treat namespace .have lines like alternatesJeff King, Feb 8, 2017
  48. 11/11 receive-pack: avoid duplicates between our refs and alternatesJeff King, Feb 8, 2017

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.