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

Re: [PATCH v1 0/3] [RFC] Speeding up checkout (and merge, rebase, etc)

From
Duy Nguyen <pclouds@gmail.com>
Date
Jul 27, 2018, 15:42 UTC
Message-ID
<20180727154241.GA21288@duynguyen.home>
In-Reply-To
<xmqqd0v9pyzu.fsf@gitster-ct.c.googlers.com>
On Thu, Jul 26, 2018 at 12:40:05PM -0700, Junio C Hamano wrote:
Show 10 quoted lines
> Duy Nguyen <pclouds@gmail.com> writes:
> 
> > I'm excited so I decided to try out anyway. This is what I've come up
> > with. Switching trees on git.git shows it could skip plenty entries,
> > so promising. It's ugly and it fails at t6020 though, there's still
> > work ahead. But I think it'll stop here.
> 
> We are extremely shallow compared to projects like the kernel and
> stuff from java land, so that is quite an interesting find.
> 

Yeah. I've got a more or less complete patch now with full test suite passed and even with linux.git, the numbers look pretty good.

Ben, is it possible for you to try this one out? I don't suppose it will be that good on a real big repo. But I'm curious how much faster could this patch does.

I'm quite happy that I don't have to make specific code for twoway merge, which means this patch would also help real merges (3way) too. Interestingly this also helps reduce traverse_trees() when diff_index_cached optimization is on. I have no idea how but well.. can't complain.

-- 8< --
Subject: [PATCH] unpack-trees: optimize walking same trees with cache-tree

In order to merge one or many trees with the index, unpack-trees code walk multiple trees in parallel with the index and perform n-way merge. If we find out at start of a directory that all trees are the same (by comparing OID) and cache-tree happens to be available for that directory as well, we could avoid walking the trees.

One nice attribute of cache-tree (and the index) is that the tree is only flattened (and it's called "the index") and we know how many files that directory has. With this information, we could avoid accessing object database to walk tree objects and just take the entries from the index instead.

The upside is of course a lot less I/O since we can potentially skip lots of trees (think subtrees). We also save CPU because we don't have to inflate and the apply deltas. The downside is of course more fragile code since the logic in some functions are now duplicated elsewhere.

WIth this patch, switching between two trees on linux.git where there's only one file changed (toplevel Makefile) seems sped up pretty good. Total checkout time goes down from 0.543 to 0.352 (35%). traverse_trees() one twoway merge (the big one in unpack_trees()) goes from 0.157s to 0.036 (70%).

Note that compared to diff_index_cached optimization (which is very similar to this) we do more work here. This is because diff_index_cached only cares about side effect, it does not modify the index, so we can quickly jump through a big chunk of cache entries. For n-way merge, we need to add entries and verify stuff, so more CPU cycles.

Signed-off-by: Nguyễn Thái Ngọc Duy <pclouds@gmail.com>
---
 unpack-trees.c | 125 +++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 125 insertions(+)
diff --git a/unpack-trees.c b/unpack-trees.c
index 66741130ae..9c791b55b2 100644
--- a/unpack-trees.c
+++ b/unpack-trees.c
@@ -642,6 +642,110 @@ static inline int are_same_oid(struct name_entry *name_j, struct name_entry *nam
 	return name_j->oid && name_k->oid && !oidcmp(name_j->oid, name_k->oid);
 }
 
+static int all_trees_same_as_cache_tree(int n, unsigned long dirmask,
+					struct name_entry *names,
+					struct traverse_info *info)
+{
+	struct unpack_trees_options *o = info->data;
+	int i;
+
+	if (dirmask != ((1 << n) - 1) || !S_ISDIR(names->mode) || !o->merge)
+		return 0;
+
+	for (i = 1; i < n; i++)
+		if (!are_same_oid(names, names + i))
+			return 0;
+
+	return cache_tree_matches_traversal(o->src_index->cache_tree, names, info);
+}
+
+/*
+ * Fast path if we detect that all trees are the same as cache-tree at this
+ * path. We'll walk these trees recursively using cache-tree/index instead of
+ * ODB since already know what these trees contain.
+ */
+static int traverse_by_cache_tree(int pos, int nr_entries, int nr_names,
+				  struct name_entry *names,
+				  struct traverse_info *info)
+{
+	struct cache_entry *src[MAX_UNPACK_TREES + 1] = { NULL, };
+	struct unpack_trees_options *o = info->data;
+	int i, d;
+
+	/*
+	 * Do what unpack_callback() and unpack_nondirectories() normally
+	 * do. But we do it in one function call (for even nested trees)
+	 * instead.
+	 *
+	 * D/F conflicts and staged entries are not a concern because cache-tree
+	 * would be invalidated and we would never get here in the first place.
+	 */
+	for (i = 0; i < nr_entries; i++) {
+		struct cache_entry *tree_ce;
+		int len, rc;
+
+		src[0] = o->src_index->cache[pos + i];
+
+		/* Do what unpack_nondirectories() normally does */
+		len = ce_namelen(src[0]);
+		tree_ce = xcalloc(1, cache_entry_size(len));
+
+		tree_ce->ce_mode = src[0]->ce_mode;
+		tree_ce->ce_flags = create_ce_flags(0);
+		tree_ce->ce_namelen = len;
+		oidcpy(&tree_ce->oid, &src[0]->oid);
+		memcpy(tree_ce->name, src[0]->name, len + 1);
+
+		for (d = 1; d <= nr_names; d++)
+			src[d] = tree_ce;
+
+		rc = call_unpack_fn((const struct cache_entry * const *)src, o);
+		free(tree_ce);
+		if (rc < 0)
+			return rc;
+
+		mark_ce_used(src[0], o);
+	}
+	trace_printf("Quick traverse over %d entries from %s to %s\n",
+		     nr_entries,
+		     o->src_index->cache[pos]->name,
+		     o->src_index->cache[pos + nr_entries - 1]->name);
+	return 0;
+}
+
+static int index_pos_by_traverse_info(struct name_entry *names,
+				      struct traverse_info *info)
+{
+	struct unpack_trees_options *o = info->data;
+	int len = traverse_path_len(info, names);
+	char *name = xmalloc(len + 1);
+	int pos;
+
+	make_traverse_path(name, info, names);
+	pos = index_name_pos(o->src_index, name, len);
+	if (pos >= 0)
+		BUG("This is so wrong. This is a directory and should not exist in index");
+	pos = -pos - 1;
+	/*
+	 * There's no guarantee that pos points to the first entry of the
+	 * directory. If the directory name is "letters" and there's another
+	 * file named "letters.txt" in the index, pos will point to that file
+	 * instead.
+	 */
+	while (pos < o->src_index->cache_nr) {
+		const struct cache_entry *ce = o->src_index->cache[pos];
+		if (ce_namelen(ce) > len &&
+		    ce->name[len] == '/' &&
+		    !memcmp(ce->name, name, len))
+			break;
+		pos++;
+	}
+	if (pos == o->src_index->cache_nr)
+		BUG("This is still wrong");
+	free(name);
+	return pos;
+}
+
 static int traverse_trees_recursive(int n, unsigned long dirmask,
 				    unsigned long df_conflicts,
 				    struct name_entry *names,
@@ -653,6 +757,17 @@ static int traverse_trees_recursive(int n, unsigned long dirmask,
 	void *buf[MAX_UNPACK_TREES];
 	struct traverse_info newinfo;
 	struct name_entry *p;
+	int nr_entries;
+
+	nr_entries = all_trees_same_as_cache_tree(n, dirmask, names, info);
+	if (nr_entries > 0) {
+		struct unpack_trees_options *o = info->data;
+		int pos = index_pos_by_traverse_info(names, info);
+
+		if (!o->merge || df_conflicts)
+			BUG("Wrong condition to get here buddy");
+		return traverse_by_cache_tree(pos, nr_entries, n, names, info);
+	}
 
 	p = names;
 	while (!p->mode)
@@ -812,6 +927,11 @@ static struct cache_entry *create_ce_entry(const struct traverse_info *info, con
 	return ce;
 }
 
+/*
+ * Note that traverse_by_cache_tree() duplicates some logic in this funciton
+ * without actually calling it. If you change the logic here you may need to
+ * check and change there as well.
+ */
 static int unpack_nondirectories(int n, unsigned long mask,
 				 unsigned long dirmask,
 				 struct cache_entry **src,
@@ -996,6 +1116,11 @@ static void debug_unpack_callback(int n,
 		debug_name_entry(i, names + i);
 }
 
+/*
+ * Note that traverse_by_cache_tree() duplicates some logic in this funciton
+ * without actually calling it. If you change the logic here you may need to
+ * check and change there as well.
+ */
 static int unpack_callback(int n, unsigned long mask, unsigned long dirmask, struct name_entry *names, struct traverse_info *info)
 {
 	struct cache_entry *src[MAX_UNPACK_TREES + 1] = { NULL, };
-- 
2.18.0.656.gda699b98b3

-- 8< --
--
Duy
Previous: Junio C HamanoNext: Ben Peart
Message 19 of 121 in “[RFC] Speeding up checkout (and merge, rebase, etc)”
  1. 0/3 [RFC] Speeding up checkout (and merge, rebase, etc)Ben Peart, Jul 18, 2018
  2. 1/3 add unbounded Multi-Producer-Multi-Consumer queueBen Peart, Jul 18, 2018
  3. Stefan BellerJul 18, 2018
  4. Junio C HamanoJul 19, 2018
  5. 2/3 add performance tracing around traverse_trees() in unpack_trees()Ben Peart, Jul 18, 2018
  6. 3/3 Add initial parallel version of unpack_trees()Ben Peart, Jul 18, 2018
  7. Junio C HamanoJul 18, 2018
  8. Stefan BellerJul 18, 2018
  9. Jeff KingJul 18, 2018
  10. Ben PeartJul 23, 2018
  11. Duy NguyenJul 23, 2018
  12. Ben PeartJul 23, 2018
  13. Jeff KingJul 24, 2018
  14. Duy NguyenJul 24, 2018
  15. Ben PeartJul 25, 2018
  16. Duy NguyenJul 26, 2018
  17. Duy NguyenJul 26, 2018
  18. Junio C HamanoJul 26, 2018
  19. Duy NguyenJul 27, 2018
  20. Ben PeartJul 27, 2018
  21. Duy NguyenJul 27, 2018
  22. Junio C HamanoJul 27, 2018
  23. Duy NguyenJul 27, 2018
  24. Duy NguyenJul 29, 2018
  25. 0/4 Speed up unpack_trees()Nguyễn Thái Ngọc Duy, Jul 29, 2018
  26. 1/4 unpack-trees.c: add performance tracingNguyễn Thái Ngọc Duy, Jul 29, 2018
  27. Ben PeartJul 30, 2018
  28. 2/4 unpack-trees: optimize walking same trees with cache-treeNguyễn Thái Ngọc Duy, Jul 29, 2018
  29. Ben PeartJul 30, 2018
  30. 3/4 unpack-trees: reduce malloc in cache-tree walkNguyễn Thái Ngọc Duy, Jul 29, 2018
  31. Ben PeartJul 30, 2018
  32. 4/4 unpack-trees: cheaper index update when walking by cache-treeNguyễn Thái Ngọc Duy, Jul 29, 2018
  33. Elijah NewrenAug 8, 2018
  34. Duy NguyenAug 10, 2018
  35. Elijah NewrenAug 10, 2018
  36. Duy NguyenAug 10, 2018
  37. Elijah NewrenAug 10, 2018
  38. Duy NguyenAug 10, 2018
  39. Ben PeartJul 30, 2018
  40. Duy NguyenJul 31, 2018
  41. Ben PeartJul 31, 2018
  42. Ben PeartJul 31, 2018
  43. Duy NguyenAug 1, 2018
  44. Ben PeartAug 8, 2018
  45. Ben PeartAug 9, 2018
  46. Duy NguyenAug 10, 2018
  47. Duy NguyenAug 10, 2018
  48. Ben PeartJul 30, 2018
  49. 0/4 Speed up unpack_trees()Nguyễn Thái Ngọc Duy, Aug 4, 2018
  50. 1/4 unpack-trees: add performance tracingNguyễn Thái Ngọc Duy, Aug 4, 2018
  51. 2/4 unpack-trees: optimize walking same trees with cache-treeNguyễn Thái Ngọc Duy, Aug 4, 2018
  52. Elijah NewrenAug 8, 2018
  53. Duy NguyenAug 10, 2018
  54. Elijah NewrenAug 10, 2018
  55. 3/4 unpack-trees: reduce malloc in cache-tree walkNguyễn Thái Ngọc Duy, Aug 4, 2018
  56. Elijah NewrenAug 8, 2018
  57. 4/4 unpack-trees: cheaper index update when walking by cache-treeNguyễn Thái Ngọc Duy, Aug 4, 2018
  58. Junio C HamanoAug 6, 2018
  59. Duy NguyenAug 6, 2018
  60. Junio C HamanoAug 6, 2018
  61. Ben PeartAug 8, 2018
  62. Junio C HamanoAug 8, 2018
  63. Junio C HamanoAug 8, 2018
  64. Junio C HamanoAug 8, 2018
  65. Duy NguyenAug 10, 2018
  66. 0/5 Speed up unpack_trees()Nguyễn Thái Ngọc Duy, Aug 12, 2018
  67. 3/5 unpack-trees: optimize walking same trees with cache-treeNguyễn Thái Ngọc Duy, Aug 12, 2018
  68. Ben PeartAug 13, 2018
  69. Duy NguyenAug 15, 2018
  70. 1/5 trace.h: support nested performance tracingNguyễn Thái Ngọc Duy, Aug 12, 2018
  71. Ben PeartAug 13, 2018
  72. 2/5 unpack-trees: add performance tracingNguyễn Thái Ngọc Duy, Aug 12, 2018
  73. Thomas AdamAug 12, 2018
  74. Junio C HamanoAug 13, 2018
  75. Ben PeartAug 13, 2018
  76. Jeff KingAug 13, 2018
  77. Stefan BellerAug 13, 2018
  78. Ben PeartAug 13, 2018
  79. Duy NguyenAug 13, 2018
  80. Jeff KingAug 13, 2018
  81. Junio C HamanoAug 13, 2018
  82. Jeff HostetlerAug 14, 2018
  83. Duy NguyenAug 14, 2018
  84. Stefan BellerAug 14, 2018
  85. Duy NguyenAug 14, 2018
  86. Jeff KingAug 14, 2018
  87. Junio C HamanoAug 14, 2018
  88. Duy NguyenAug 15, 2018
  89. Junio C HamanoAug 15, 2018
  90. Jeff HostetlerAug 14, 2018
  91. 4/5 unpack-trees: reduce malloc in cache-tree walkNguyễn Thái Ngọc Duy, Aug 12, 2018
  92. 5/5 unpack-trees: reuse (still valid) cache-tree from src_indexNguyễn Thái Ngọc Duy, Aug 12, 2018
  93. Elijah NewrenAug 13, 2018
  94. Duy NguyenAug 13, 2018
  95. Ben PeartAug 13, 2018
  96. Duy NguyenAug 13, 2018
  97. Ben PeartAug 13, 2018
  98. Junio C HamanoAug 13, 2018
  99. Ben PeartAug 14, 2018
  100. 0/7 Speed up unpack_trees()Nguyễn Thái Ngọc Duy, Aug 18, 2018
  101. 1/7 trace.h: support nested performance tracingNguyễn Thái Ngọc Duy, Aug 18, 2018
  102. 2/7 unpack-trees: add performance tracingNguyễn Thái Ngọc Duy, Aug 18, 2018
  103. 3/7 unpack-trees: optimize walking same trees with cache-treeNguyễn Thái Ngọc Duy, Aug 18, 2018
  104. Ben PeartAug 20, 2018
  105. 5/7 unpack-trees: reuse (still valid) cache-tree from src_indexNguyễn Thái Ngọc Duy, Aug 18, 2018
  106. 6/7 unpack-trees: add missing cache invalidationNguyễn Thái Ngọc Duy, Aug 18, 2018
  107. 4/7 unpack-trees: reduce malloc in cache-tree walkNguyễn Thái Ngọc Duy, Aug 18, 2018
  108. 7/7 cache-tree: verify valid cache-tree in the test suiteNguyễn Thái Ngọc Duy, Aug 18, 2018
  109. Elijah NewrenAug 18, 2018
  110. Elijah NewrenAug 18, 2018
  111. Duy NguyenAug 19, 2018
  112. Document update for nd/unpack-trees-with-cache-treeNguyễn Thái Ngọc Duy, Aug 25, 2018
  113. Martin ÅgrenAug 25, 2018
  114. Document update for nd/unpack-trees-with-cache-treeNguyễn Thái Ngọc Duy, Aug 25, 2018
  115. Ben PeartJul 27, 2018
  116. Duy NguyenJul 26, 2018
  117. Junio C HamanoJul 24, 2018
  118. Duy NguyenJul 24, 2018
  119. Jeff KingJul 24, 2018
  120. Ben PeartJul 25, 2018
  121. Jeff KingJul 24, 2018

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.