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

[PATCH v3 21/21] pack-bitmap: implement optional name_hash cache

From
Jeff King <peff@peff.net>
Date
Nov 14, 2013, 12:48 UTC
Message-ID
<20131114124855.GB11612@sigill.intra.peff.net>
In-Reply-To
<20131114124157.GA23784@sigill.intra.peff.net>
From: Vicent Marti <tanoku@gmail.com>

When we use pack bitmaps rather than walking the object graph, we end up with the list of objects to include in the packfile, but we do not know the path at which any tree or blob objects would be found.

In a recently packed repository, this is fine. A fetch would use the paths only as a heuristic in the delta compression phase, and a fully packed repository should not need to do much delta compression.

As time passes, though, we may acquire more objects on top of our large bitmapped pack. If clients fetch frequently, then they never even look at the bitmapped history, and all works as usual. However, a client who has not fetched since the last bitmap repack will have "have" tips in the bitmapped history, but "want" newer objects.

The bitmaps themselves degrade gracefully in this circumstance. We manually walk the more recent bits of history, and then use bitmaps when we hit them.

But we would also like to perform delta compression between the newer objects and the bitmapped objects (both to delta against what we know the user already has, but also between "new" and "old" objects that the user is fetching). The lack of pathnames makes our delta heuristics much less effective.

This patch adds an optional cache of the 32-bit name_hash values to the end of the bitmap file. If present, a reader can use it to match bitmapped and non-bitmapped names during delta compression.

Here are perf results for p5310:

Test origin/master HEAD^ HEAD ------------------------------------------------------------------------------------------------- 5310.2: repack to disk 36.81(37.82+1.43) 47.70(48.74+1.41) +29.6% 47.75(48.70+1.51) +29.7% 5310.3: simulated clone 30.78(29.70+2.14) 1.08(0.97+0.10) -96.5% 1.07(0.94+0.12) -96.5% 5310.4: simulated fetch 3.16(6.10+0.08) 3.54(10.65+0.06) +12.0% 1.70(3.07+0.06) -46.2% 5310.6: partial bitmap 36.76(43.19+1.81) 6.71(11.25+0.76) -81.7% 4.08(6.26+0.46) -88.9%

You can see that the time spent on an incremental fetch goes down, as our delta heuristics are able to do their work. And we save time on the partial bitmap clone for the same reason.

Signed-off-by: Vicent Marti <tanoku@gmail.com>
Signed-off-by: Jeff King <peff@peff.net>
---
 Documentation/config.txt                  | 11 +++++++++++
 Documentation/technical/bitmap-format.txt | 33 +++++++++++++++++++++++++++++++
 builtin/pack-objects.c                    | 10 +++++++++-
 pack-bitmap-write.c                       | 21 ++++++++++++++++++--
 pack-bitmap.c                             | 11 +++++++++++
 pack-bitmap.h                             |  4 +++-
 t/perf/p5310-pack-bitmaps.sh              |  3 ++-
 t/t5310-pack-bitmaps.sh                   |  3 ++-
 8 files changed, 90 insertions(+), 6 deletions(-)
diff --git a/Documentation/config.txt b/Documentation/config.txt
index a439a4c..e6d3922 100644
--- a/Documentation/config.txt
+++ b/Documentation/config.txt
@@ -1872,6 +1872,17 @@ pack.writebitmaps::
 	space and extra time spent on the initial repack.  Defaults to
 	false.
 
+pack.writeBitmapHashCache::
+	When true, git will include a "hash cache" section in the bitmap
+	index (if one is written). This cache can be used to feed git's
+	delta heuristics, potentially leading to better deltas between
+	bitmapped and non-bitmapped objects (e.g., when serving a fetch
+	between an older, bitmapped pack and objects that have been
+	pushed since the last gc). The downside is that it consumes 4
+	bytes per object of disk space, and that JGit's bitmap
+	implementation does not understand it, causing it to complain if
+	Git and JGit are used on the same repository. Defaults to false.
+
 pager.<cmd>::
 	If the value is boolean, turns on or off pagination of the
 	output of a particular Git subcommand when writing to a tty.
diff --git a/Documentation/technical/bitmap-format.txt b/Documentation/technical/bitmap-format.txt
index 7a86bd7..f8c18a0 100644
--- a/Documentation/technical/bitmap-format.txt
+++ b/Documentation/technical/bitmap-format.txt
@@ -21,6 +21,12 @@ GIT bitmap v1 format
 			requirement for the bitmap index format, also present in JGit,
 			that greatly reduces the complexity of the implementation.
 
+			- BITMAP_OPT_HASH_CACHE (0x4)
+			If present, the end of the bitmap file contains
+			`N` 32-bit name-hash values, one per object in the
+			pack. The format and meaning of the name-hash is
+			described below.
+
 		4-byte entry count (network byte order)
 
 			The total count of entries (bitmapped commits) in this bitmap index.
@@ -129,3 +135,30 @@ The bitstream represented by the above chunk is then:
 The next word after `L_M` (if any) must again be a RLW, for the next
 chunk.  For efficient appending to the bitstream, the EWAH stores a
 pointer to the last RLW in the stream.
+
+
+== Appendix B: Optional Bitmap Sections
+
+These sections may or may not be present in the `.bitmap` file; their
+presence is indicated by the header flags section described above.
+
+Name-hash cache
+---------------
+
+If the BITMAP_OPT_HASH_CACHE flag is set, the end of the bitmap contains
+a cache of 32-bit values, one per object in the pack. The value at
+position `i` is the hash of the pathname at which the `i`th object
+(counting in index order) in the pack can be found.  This can be fed
+into the delta heuristics to compare objects with similar pathnames.
+
+The hash algorithm used is:
+
+    hash = 0;
+    while ((c = *name++))
+	    if (!isspace(c))
+		    hash = (hash >> 2) + (c << 24);
+
+Note that this hashing scheme is tied to the BITMAP_OPT_HASH_CACHE flag.
+If implementations want to choose a different hashing scheme, they are
+free to do so, but MUST allocate a new header flag (because comparing
+hashes made under two different schemes would be pointless).
diff --git a/builtin/pack-objects.c b/builtin/pack-objects.c
index 26646e7..4504789 100644
--- a/builtin/pack-objects.c
+++ b/builtin/pack-objects.c
@@ -64,6 +64,7 @@ static off_t reuse_packfile_offset;
 
 static int use_bitmap_index = 1;
 static int write_bitmap_index;
+static uint16_t write_bitmap_options;
 
 static unsigned long delta_cache_size = 0;
 static unsigned long max_delta_cache_size = 256 * 1024 * 1024;
@@ -862,7 +863,8 @@ static void write_pack_file(void)
 				bitmap_writer_reuse_bitmaps(&to_pack);
 				bitmap_writer_select_commits(indexed_commits, indexed_commits_nr, -1);
 				bitmap_writer_build(&to_pack);
-				bitmap_writer_finish(written_list, nr_written, tmpname);
+				bitmap_writer_finish(written_list, nr_written,
+						     tmpname, write_bitmap_options);
 				write_bitmap_index = 0;
 			}
 
@@ -2155,6 +2157,12 @@ static int git_pack_config(const char *k, const char *v, void *cb)
 		write_bitmap_index = git_config_bool(k, v);
 		return 0;
 	}
+	if (!strcmp(k, "pack.writebitmaphashcache")) {
+		if (git_config_bool(k, v))
+			write_bitmap_options |= BITMAP_OPT_HASH_CACHE;
+		else
+			write_bitmap_options &= ~BITMAP_OPT_HASH_CACHE;
+	}
 	if (!strcmp(k, "pack.usebitmaps")) {
 		use_bitmap_index = git_config_bool(k, v);
 		return 0;
diff --git a/pack-bitmap-write.c b/pack-bitmap-write.c
index 954a74d..1218bef 100644
--- a/pack-bitmap-write.c
+++ b/pack-bitmap-write.c
@@ -490,6 +490,19 @@ static void write_selected_commits_v1(struct sha1file *f,
 	}
 }
 
+static void write_hash_cache(struct sha1file *f,
+			     struct pack_idx_entry **index,
+			     uint32_t index_nr)
+{
+	uint32_t i;
+
+	for (i = 0; i < index_nr; ++i) {
+		struct object_entry *entry = (struct object_entry *)index[i];
+		uint32_t hash_value = htonl(entry->hash);
+		sha1write(f, &hash_value, sizeof(hash_value));
+	}
+}
+
 void bitmap_writer_set_checksum(unsigned char *sha1)
 {
 	hashcpy(writer.pack_checksum, sha1);
@@ -497,7 +510,8 @@ void bitmap_writer_set_checksum(unsigned char *sha1)
 
 void bitmap_writer_finish(struct pack_idx_entry **index,
 			  uint32_t index_nr,
-			  const char *filename)
+			  const char *filename,
+			  uint16_t options)
 {
 	static char tmp_file[PATH_MAX];
 	static uint16_t default_version = 1;
@@ -514,7 +528,7 @@ void bitmap_writer_finish(struct pack_idx_entry **index,
 
 	memcpy(header.magic, BITMAP_IDX_SIGNATURE, sizeof(BITMAP_IDX_SIGNATURE));
 	header.version = htons(default_version);
-	header.options = htons(flags);
+	header.options = htons(flags | options);
 	header.entry_count = htonl(writer.selected_nr);
 	memcpy(header.checksum, writer.pack_checksum, 20);
 
@@ -525,6 +539,9 @@ void bitmap_writer_finish(struct pack_idx_entry **index,
 	dump_bitmap(f, writer.tags);
 	write_selected_commits_v1(f, index, index_nr);
 
+	if (options & BITMAP_OPT_HASH_CACHE)
+		write_hash_cache(f, index, index_nr);
+
 	sha1close(f, NULL, CSUM_FSYNC);
 
 	if (adjust_shared_perm(tmp_file))
diff --git a/pack-bitmap.c b/pack-bitmap.c
index 7792dfd..078f7c6 100644
--- a/pack-bitmap.c
+++ b/pack-bitmap.c
@@ -66,6 +66,9 @@ static struct bitmap_index {
 	/* Number of bitmapped commits */
 	uint32_t entry_count;
 
+	/* Name-hash cache (or NULL if not present). */
+	uint32_t *hashes;
+
 	/*
 	 * Extended index.
 	 *
@@ -152,6 +155,11 @@ static int load_bitmap_header(struct bitmap_index *index)
 		if ((flags & BITMAP_OPT_FULL_DAG) == 0)
 			return error("Unsupported options for bitmap index file "
 				"(Git requires BITMAP_OPT_FULL_DAG)");
+
+		if (flags & BITMAP_OPT_HASH_CACHE) {
+			unsigned char *end = index->map + index->map_size - 20;
+			index->hashes = ((uint32_t *)end) - index->pack->num_objects;
+		}
 	}
 
 	index->entry_count = ntohl(header->entry_count);
@@ -626,6 +634,9 @@ static void show_objects_for_type(
 			entry = &bitmap_git.reverse_index->revindex[pos + offset];
 			sha1 = nth_packed_object_sha1(bitmap_git.pack, entry->nr);
 
+			if (bitmap_git.hashes)
+				hash = ntohl(bitmap_git.hashes[entry->nr]);
+
 			show_reach(sha1, object_type, 0, hash, bitmap_git.pack, entry->offset);
 		}
 
diff --git a/pack-bitmap.h b/pack-bitmap.h
index 5e3fb9a..e4e1a57 100644
--- a/pack-bitmap.h
+++ b/pack-bitmap.h
@@ -25,6 +25,7 @@ static const char BITMAP_IDX_SIGNATURE[] = {'B', 'I', 'T', 'M'};;
 
 enum pack_bitmap_opts {
 	BITMAP_OPT_FULL_DAG = 1,
+	BITMAP_OPT_HASH_CACHE = 4,
 };
 
 enum pack_bitmap_flags {
@@ -57,6 +58,7 @@ void bitmap_writer_select_commits(struct commit **indexed_commits,
 void bitmap_writer_build(struct packing_data *to_pack);
 void bitmap_writer_finish(struct pack_idx_entry **index,
 			  uint32_t index_nr,
-			  const char *filename);
+			  const char *filename,
+			  uint16_t options);
 
 #endif
diff --git a/t/perf/p5310-pack-bitmaps.sh b/t/perf/p5310-pack-bitmaps.sh
index ac036d0..8811fc4 100755
--- a/t/perf/p5310-pack-bitmaps.sh
+++ b/t/perf/p5310-pack-bitmaps.sh
@@ -9,7 +9,8 @@ test_perf_large_repo
 # since we want to be able to compare bitmap-aware
 # git versus non-bitmap git
 test_expect_success 'setup bitmap config' '
-	git config pack.writebitmaps true
+	git config pack.writebitmaps true &&
+	git config pack.writebitmaphashcache true
 '
 
 test_perf 'repack to disk' '
diff --git a/t/t5310-pack-bitmaps.sh b/t/t5310-pack-bitmaps.sh
index bd743f4..2c2632f 100755
--- a/t/t5310-pack-bitmaps.sh
+++ b/t/t5310-pack-bitmaps.sh
@@ -14,7 +14,8 @@ test_expect_success 'setup repo with moderate-sized history' '
 	git checkout master &&
 	blob=$(echo tagged-blob | git hash-object -w --stdin) &&
 	git tag tagged-blob $blob &&
-	git config pack.writebitmaps true
+	git config pack.writebitmaps true &&
+	git config pack.writebitmaphashcache true
 '
 
 test_expect_success 'full repack creates bitmaps' '

-- 
1.8.5.rc0.443.g2df7f3f
Previous: Jeff KingNext: Thomas Rast
Message 49 of 55 in “pack bitmaps”
  1. 0/21 pack bitmapsJeff King, Nov 14, 2013
  2. 01/21 sha1write: make buffer const-correctJeff King, Nov 14, 2013
  3. 02/21 revindex: Export new APIsJeff King, Nov 14, 2013
  4. 03/21 pack-objects: Refactor the packing listJeff King, Nov 14, 2013
  5. 04/21 pack-objects: factor out name_hashJeff King, Nov 14, 2013
  6. 05/21 revision: allow setting custom limiter functionJeff King, Nov 14, 2013
  7. 06/21 sha1_file: export `git_open_noatime`Jeff King, Nov 14, 2013
  8. 07/21 compat: add endianness helpersJeff King, Nov 14, 2013
  9. 08/21 ewah: compressed bitmap implementationJeff King, Nov 14, 2013
  10. 09/21 documentation: add documentation for the bitmap formatJeff King, Nov 14, 2013
  11. 10/21 pack-bitmap: add support for bitmap indexesJeff King, Nov 14, 2013
  12. Thomas RastNov 24, 2013
  13. Document khashThomas Rast, Nov 25, 2013
  14. Jeff KingNov 28, 2013
  15. Karsten BleesNov 27, 2013
  16. Jeff KingNov 28, 2013
  17. Karsten BleesDec 3, 2013
  18. Jeff KingDec 3, 2013
  19. Karsten BleesDec 7, 2013
  20. Thomas RastNov 29, 2013
  21. Jeff KingDec 2, 2013
  22. Junio C HamanoDec 2, 2013
  23. Jeff KingDec 2, 2013
  24. Junio C HamanoDec 2, 2013
  25. 11/21 pack-objects: use bitmaps when packing objectsJeff King, Nov 14, 2013
  26. Thomas RastDec 7, 2013
  27. Jeff KingDec 21, 2013
  28. 12/21 rev-list: add bitmap mode to speed up object listsJeff King, Nov 14, 2013
  29. Thomas RastDec 7, 2013
  30. 13/21 pack-objects: implement bitmap writingJeff King, Nov 14, 2013
  31. Thomas RastDec 7, 2013
  32. Jeff KingDec 21, 2013
  33. 14/21 repack: stop using magic number for ARRAY_SIZE(exts)Jeff King, Nov 14, 2013
  34. Thomas RastDec 7, 2013
  35. 15/21 repack: turn exts array into array-of-structJeff King, Nov 14, 2013
  36. Thomas RastDec 7, 2013
  37. 16/21 repack: handle optional files created by pack-objectsJeff King, Nov 14, 2013
  38. Thomas RastDec 7, 2013
  39. 17/21 repack: consider bitmaps when performing repacksJeff King, Nov 14, 2013
  40. Thomas RastDec 7, 2013
  41. 18/21 count-objects: recognize .bitmap in garbage-checkingJeff King, Nov 14, 2013
  42. Thomas RastDec 7, 2013
  43. 19/21 t: add basic bitmap functionality testsJeff King, Nov 14, 2013
  44. Thomas RastDec 7, 2013
  45. Jeff KingDec 21, 2013
  46. 20/21 t/perf: add tests for pack bitmapsJeff King, Nov 14, 2013
  47. Thomas RastDec 7, 2013
  48. Jeff KingDec 21, 2013
  49. 21/21 pack-bitmap: implement optional name_hash cacheJeff King, Nov 14, 2013
  50. Thomas RastDec 7, 2013
  51. Ramsay JonesNov 14, 2013
  52. Jeff KingNov 14, 2013
  53. Ramsay JonesNov 14, 2013
  54. Ramsay JonesNov 18, 2013
  55. Thomas RastNov 16, 2013

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.