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

[PATCH v4 02/17] revision.c: consult Bloom filters for root commits

From
Taylor Blau <me@ttaylorr.com>
Date
Oct 18, 2023, 18:32 UTC
Message-ID
<87b09e6266a01e7fa4480d37f22e1ac3f4be6bc3.1697653929.git.me@ttaylorr.com>
In-Reply-To
<cover.1697653929.git.me@ttaylorr.com>

The commit-graph stores changed-path Bloom filters which represent the set of paths included in a tree-level diff between a commit's root tree and that of its parent.

When a commit has no parents, the tree-diff is computed against that commit's root tree and the empty tree. In other words, every path in that commit's tree is stored in the Bloom filter (since they all appear in the diff).

Consult these filters during pathspec-limited traversals in the function `rev_same_tree_as_empty()`. Doing so yields a performance improvement where we can avoid enumerating the full set of paths in a parentless commit's root tree when we know that the path(s) of interest were not listed in that commit's changed-path Bloom filter.

Suggested-by: SZEDER Gábor <szeder.dev@gmail.com>
Original-patch-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Taylor Blau <me@ttaylorr.com>
---
 revision.c           | 26 ++++++++++++++++++++++----
 t/t4216-log-bloom.sh |  8 ++++++--
 2 files changed, 28 insertions(+), 6 deletions(-)
diff --git a/revision.c b/revision.c
index 219dc76716..6e9da518d9 100644
--- a/revision.c
+++ b/revision.c
@@ -834,17 +834,28 @@ static int rev_compare_tree(struct rev_info *revs,
 	return tree_difference;
 }
 
-static int rev_same_tree_as_empty(struct rev_info *revs, struct commit *commit)
+static int rev_same_tree_as_empty(struct rev_info *revs, struct commit *commit,
+				  int nth_parent)
 {
 	struct tree *t1 = repo_get_commit_tree(the_repository, commit);
+	int bloom_ret = 1;
 
 	if (!t1)
 		return 0;
 
+	if (nth_parent == 1 && revs->bloom_keys_nr) {
+		bloom_ret = check_maybe_different_in_bloom_filter(revs, commit);
+		if (!bloom_ret)
+			return 1;
+	}
+
 	tree_difference = REV_TREE_SAME;
 	revs->pruning.flags.has_changes = 0;
 	diff_tree_oid(NULL, &t1->object.oid, "", &revs->pruning);
 
+	if (bloom_ret == 1 && tree_difference == REV_TREE_SAME)
+		count_bloom_filter_false_positive++;
+
 	return tree_difference == REV_TREE_SAME;
 }
 
@@ -882,7 +893,7 @@ static int compact_treesame(struct rev_info *revs, struct commit *commit, unsign
 		if (nth_parent != 0)
 			die("compact_treesame %u", nth_parent);
 		old_same = !!(commit->object.flags & TREESAME);
-		if (rev_same_tree_as_empty(revs, commit))
+		if (rev_same_tree_as_empty(revs, commit, nth_parent))
 			commit->object.flags |= TREESAME;
 		else
 			commit->object.flags &= ~TREESAME;
@@ -978,7 +989,14 @@ static void try_to_simplify_commit(struct rev_info *revs, struct commit *commit)
 		return;
 
 	if (!commit->parents) {
-		if (rev_same_tree_as_empty(revs, commit))
+		/*
+		 * Pretend as if we are comparing ourselves to the
+		 * (non-existent) first parent of this commit object. Even
+		 * though no such parent exists, its changed-path Bloom filter
+		 * (if one exists) is relative to the empty tree, using Bloom
+		 * filters is allowed here.
+		 */
+		if (rev_same_tree_as_empty(revs, commit, 1))
 			commit->object.flags |= TREESAME;
 		return;
 	}
@@ -1059,7 +1077,7 @@ static void try_to_simplify_commit(struct rev_info *revs, struct commit *commit)
 
 		case REV_TREE_NEW:
 			if (revs->remove_empty_trees &&
-			    rev_same_tree_as_empty(revs, p)) {
+			    rev_same_tree_as_empty(revs, p, nth_parent)) {
 				/* We are adding all the specified
 				 * paths from this parent, so the
 				 * history beyond this parent is not
diff --git a/t/t4216-log-bloom.sh b/t/t4216-log-bloom.sh
index 487fc3d6b9..322640feeb 100755
--- a/t/t4216-log-bloom.sh
+++ b/t/t4216-log-bloom.sh
@@ -87,7 +87,11 @@ test_bloom_filters_not_used () {
 		# if the Bloom filter system is initialized, ensure that no
 		# filters were used
 		data="statistics:{"
-		data="$data\"filter_not_present\":0,"
+		# unusable filters (e.g., those computed with a
+		# different value of commitGraph.changedPathsVersion)
+		# are counted in the filter_not_present bucket, so any
+		# value is OK there.
+		data="$data\"filter_not_present\":[0-9][0-9]*,"
 		data="$data\"maybe\":0,"
 		data="$data\"definitely_not\":0,"
 		data="$data\"false_positive\":0}"
@@ -174,7 +178,7 @@ test_expect_success 'setup - add commit-graph to the chain with Bloom filters' '
 
 test_bloom_filters_used_when_some_filters_are_missing () {
 	log_args=$1
-	bloom_trace_prefix="statistics:{\"filter_not_present\":3,\"maybe\":6,\"definitely_not\":9"
+	bloom_trace_prefix="statistics:{\"filter_not_present\":3,\"maybe\":6,\"definitely_not\":10"
 	setup "$log_args" &&
 	grep -q "$bloom_trace_prefix" "$TRASH_DIRECTORY/trace.perf" &&
 	test_cmp log_wo_bloom log_w_bloom
-- 
2.42.0.415.g8942f205c8
Previous: Taylor BlauNext: Taylor Blau
Message 50 of 89 in “merge-ort: implement support for packing objects together”
  1. 0/7 merge-ort: implement support for packing objects togetherTaylor Blau, Oct 6, 2023
  2. 1/7 bulk-checkin: factor out `format_object_header_hash()`Taylor Blau, Oct 6, 2023
  3. 2/7 bulk-checkin: factor out `prepare_checkpoint()`Taylor Blau, Oct 6, 2023
  4. 3/7 bulk-checkin: factor out `truncate_checkpoint()`Taylor Blau, Oct 6, 2023
  5. 4/7 bulk-checkin: factor our `finalize_checkpoint()`Taylor Blau, Oct 6, 2023
  6. 5/7 bulk-checkin: introduce `index_blob_bulk_checkin_incore()`Taylor Blau, Oct 6, 2023
  7. 6/7 bulk-checkin: introduce `index_tree_bulk_checkin_incore()`Taylor Blau, Oct 6, 2023
  8. Eric BiedermanOct 7, 2023
  9. Taylor BlauOct 9, 2023
  10. 7/7 builtin/merge-tree.c: implement support for `--write-pack`Taylor Blau, Oct 6, 2023
  11. Junio C HamanoOct 6, 2023
  12. Taylor BlauOct 6, 2023
  13. Elijah NewrenOct 8, 2023
  14. Taylor BlauOct 8, 2023
  15. Jeff KingOct 8, 2023
  16. Taylor BlauOct 9, 2023
  17. Jeff KingOct 9, 2023
  18. Junio C HamanoOct 9, 2023
  19. Patrick SteinhardtOct 9, 2023
  20. Taylor BlauOct 9, 2023
  21. Patrick SteinhardtOct 10, 2023
  22. 0/7 merge-ort: implement support for packing objects togetherTaylor Blau, Oct 17, 2023
  23. 1/7 bulk-checkin: factor out `format_object_header_hash()`Taylor Blau, Oct 17, 2023
  24. 2/7 bulk-checkin: factor out `prepare_checkpoint()`Taylor Blau, Oct 17, 2023
  25. 3/7 bulk-checkin: factor out `truncate_checkpoint()`Taylor Blau, Oct 17, 2023
  26. 4/7 bulk-checkin: factor our `finalize_checkpoint()`Taylor Blau, Oct 17, 2023
  27. 5/7 bulk-checkin: introduce `index_blob_bulk_checkin_incore()`Taylor Blau, Oct 17, 2023
  28. Junio C HamanoOct 18, 2023
  29. Taylor BlauOct 18, 2023
  30. 6/7 bulk-checkin: introduce `index_tree_bulk_checkin_incore()`Taylor Blau, Oct 17, 2023
  31. 7/7 builtin/merge-tree.c: implement support for `--write-pack`Taylor Blau, Oct 17, 2023
  32. 00/10 merge-ort: implement support for packing objects togetherTaylor Blau, Oct 18, 2023
  33. 03/10 bulk-checkin: factor out `truncate_checkpoint()`Taylor Blau, Oct 18, 2023
  34. 10/10 builtin/merge-tree.c: implement support for `--write-pack`Taylor Blau, Oct 18, 2023
  35. 07/10 bulk-checkin: generify `stream_blob_to_pack()` for arbitrary typesTaylor Blau, Oct 18, 2023
  36. 08/10 bulk-checkin: introduce `index_blob_bulk_checkin_incore()`Taylor Blau, Oct 18, 2023
  37. Junio C HamanoOct 18, 2023
  38. Taylor BlauOct 19, 2023
  39. 09/10 bulk-checkin: introduce `index_tree_bulk_checkin_incore()`Taylor Blau, Oct 18, 2023
  40. 01/10 bulk-checkin: factor out `format_object_header_hash()`Taylor Blau, Oct 18, 2023
  41. 02/10 bulk-checkin: factor out `prepare_checkpoint()`Taylor Blau, Oct 18, 2023
  42. 04/10 bulk-checkin: factor out `finalize_checkpoint()`Taylor Blau, Oct 18, 2023
  43. 05/10 bulk-checkin: extract abstract `bulk_checkin_source`Taylor Blau, Oct 18, 2023
  44. Junio C HamanoOct 18, 2023
  45. Taylor BlauOct 19, 2023
  46. Junio C HamanoOct 19, 2023
  47. 06/10 bulk-checkin: implement `SOURCE_INCORE` mode for `bulk_checkin_source`Taylor Blau, Oct 18, 2023
  48. 00/17 bloom: changed-path Bloom filters v2 (& sundries)Taylor Blau, Oct 18, 2023
  49. 01/17 t/t4216-log-bloom.sh: harden `test_bloom_filters_not_used()`Taylor Blau, Oct 18, 2023
  50. 02/17 revision.c: consult Bloom filters for root commitsTaylor Blau, Oct 18, 2023
  51. 03/17 commit-graph: ensure Bloom filters are read with consistent settingsTaylor Blau, Oct 18, 2023
  52. 04/17 gitformat-commit-graph: describe version 2 of BDATTaylor Blau, Oct 18, 2023
  53. 05/17 t/helper/test-read-graph.c: extract `dump_graph_info()`Taylor Blau, Oct 18, 2023
  54. 06/17 bloom.h: make `load_bloom_filter_from_graph()` publicTaylor Blau, Oct 18, 2023
  55. 07/17 t/helper/test-read-graph: implement `bloom-filters` modeTaylor Blau, Oct 18, 2023
  56. 08/17 t4216: test changed path filters with high bit pathsTaylor Blau, Oct 18, 2023
  57. 09/17 repo-settings: introduce commitgraph.changedPathsVersionTaylor Blau, Oct 18, 2023
  58. 10/17 commit-graph: new filter ver. that fixes murmur3Taylor Blau, Oct 18, 2023
  59. 11/17 bloom: annotate filters with hash versionTaylor Blau, Oct 18, 2023
  60. 12/17 bloom: prepare to discard incompatible Bloom filtersTaylor Blau, Oct 18, 2023
  61. 13/17 commit-graph.c: unconditionally load Bloom filtersTaylor Blau, Oct 18, 2023
  62. 14/17 commit-graph: drop unnecessary `graph_read_bloom_data_context`Taylor Blau, Oct 18, 2023
  63. 15/17 object.h: fix mis-aligned flag bits tableTaylor Blau, Oct 18, 2023
  64. 16/17 commit-graph: reuse existing Bloom filters where possibleTaylor Blau, Oct 18, 2023
  65. 17/17 bloom: introduce `deinit_bloom_filters()`Taylor Blau, Oct 18, 2023
  66. Junio C HamanoOct 18, 2023
  67. Taylor BlauOct 20, 2023
  68. SZEDER GáborOct 23, 2023
  69. Taylor BlauOct 30, 2023
  70. 00/17 bloom: changed-path Bloom filters v2 (& sundries)Taylor Blau, Jan 16, 2024
  71. 01/17 t/t4216-log-bloom.sh: harden `test_bloom_filters_not_used()`Taylor Blau, Jan 16, 2024
  72. 02/17 revision.c: consult Bloom filters for root commitsTaylor Blau, Jan 16, 2024
  73. 03/17 commit-graph: ensure Bloom filters are read with consistent settingsTaylor Blau, Jan 16, 2024
  74. 04/17 gitformat-commit-graph: describe version 2 of BDATTaylor Blau, Jan 16, 2024
  75. 05/17 t/helper/test-read-graph.c: extract `dump_graph_info()`Taylor Blau, Jan 16, 2024
  76. 06/17 bloom.h: make `load_bloom_filter_from_graph()` publicTaylor Blau, Jan 16, 2024
  77. 07/17 t/helper/test-read-graph: implement `bloom-filters` modeTaylor Blau, Jan 16, 2024
  78. 08/17 t4216: test changed path filters with high bit pathsTaylor Blau, Jan 16, 2024
  79. 09/17 repo-settings: introduce commitgraph.changedPathsVersionTaylor Blau, Jan 16, 2024
  80. SZEDER GáborJan 29, 2024
  81. Taylor BlauJan 29, 2024
  82. 10/17 commit-graph: new Bloom filter version that fixes murmur3Taylor Blau, Jan 16, 2024
  83. 11/17 bloom: annotate filters with hash versionTaylor Blau, Jan 16, 2024
  84. 12/17 bloom: prepare to discard incompatible Bloom filtersTaylor Blau, Jan 16, 2024
  85. 13/17 commit-graph.c: unconditionally load Bloom filtersTaylor Blau, Jan 16, 2024
  86. 14/17 commit-graph: drop unnecessary `graph_read_bloom_data_context`Taylor Blau, Jan 16, 2024
  87. 15/17 object.h: fix mis-aligned flag bits tableTaylor Blau, Jan 16, 2024
  88. 16/17 commit-graph: reuse existing Bloom filters where possibleTaylor Blau, Jan 16, 2024
  89. 17/17 bloom: introduce `deinit_bloom_filters()`Taylor Blau, Jan 16, 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.