commit-graph: Use the commit-graph in revwalks

lhchavez · lhchavez · commit b3bf58f2d0f8 · 2021-01-06T20:17:05.000-08:00
This change makes revwalks a bit faster by using the `commit-graph` file (if present). This is thanks to the `commit-graph` allow much faster parsing of the commit information by requiring near-zero I/O (aside from reading a few dozen bytes off of a `mmap(2)`-ed file) for each commit, instead of having to read the ODB, inflate the commit, and parse it. Part of: #5757
diff --git a/src/commit_list.c b/src/commit_list.c
@@ -124,6 +124,7 @@ static int commit_quick_parse(
 		return -1;
 	}
 
+	node->generation = 0;
 	node->time = commit->committer->when.time;
 	node->out_degree = (uint16_t) git_array_size(commit->parent_ids);
 	node->parents = alloc_parents(walk, node, node->out_degree);
@@ -148,6 +149,31 @@ int git_commit_list_parse(git_revwalk *walk, git_commit_list_node *commit)
 	if (commit->parsed)
 		return 0;
 
+	/* Let's try to use the commit graph first. */
+	if (walk->cgraph) {
+		git_commit_graph_entry e;
+
+		error = git_commit_graph_entry_find(&e, walk->cgraph, &commit->oid, GIT_OID_RAWSZ);
+		if (error == 0 && git__is_uint16(e.parent_count)) {
+			size_t i;
+			commit->generation = (uint32_t)e.generation;
+			commit->time = e.commit_time;
+			commit->out_degree = (uint16_t)e.parent_count;
+			commit->parents = alloc_parents(walk, commit, commit->out_degree);
+			GIT_ERROR_CHECK_ALLOC(commit->parents);
+
+			for (i = 0; i < commit->out_degree; ++i) {
+				git_commit_graph_entry parent;
+				error = git_commit_graph_entry_parent(&parent, walk->cgraph, &e, i);
+				if (error < 0)
+					return error;
+				commit->parents[i] = git_revwalk__commit_lookup(walk, &parent.sha1);
+			}
+			commit->parsed = 1;
+			return 0;
+		}
+	}
+
 	if ((error = git_odb_read(&obj, walk->odb, &commit->oid)) < 0)
 		return error;
 
diff --git a/src/commit_list.h b/src/commit_list.h
@@ -26,6 +26,7 @@
 typedef struct git_commit_list_node {
 	git_oid oid;
 	int64_t time;
+	uint32_t generation;
 	unsigned int seen:1,
 			 uninteresting:1,
 			 topo_delay:1,
diff --git a/src/revwalk.c b/src/revwalk.c
@@ -672,6 +672,9 @@ static int prepare_walk(git_revwalk *walk)
 int git_revwalk_new(git_revwalk **revwalk_out, git_repository *repo)
 {
 	git_revwalk *walk = git__calloc(1, sizeof(git_revwalk));
+	git_buf commit_graph_path = GIT_BUF_INIT;
+	int error = 0;
+
 	GIT_ERROR_CHECK_ALLOC(walk);
 
 	if (git_oidmap_new(&walk->commits) < 0 ||
@@ -689,6 +692,23 @@ int git_revwalk_new(git_revwalk **revwalk_out, git_repository *repo)
 		return -1;
 	}
 
+	if ((error = git_repository_item_path(&commit_graph_path, repo, GIT_REPOSITORY_ITEM_OBJECTS)) < 0) {
+		git_revwalk_free(walk);
+		return error;
+	}
+	if ((error = git_buf_joinpath(
+			     &commit_graph_path,
+			     git_buf_cstr(&commit_graph_path),
+			     "info/commit-graph"))
+	    < 0) {
+		git_buf_dispose(&commit_graph_path);
+		git_revwalk_free(walk);
+		return error;
+	}
+	/* Best effort */
+	(void)git_commit_graph_open(&walk->cgraph, git_buf_cstr(&commit_graph_path));
+	git_buf_dispose(&commit_graph_path);
+
 	*revwalk_out = walk;
 	return 0;
 }
@@ -698,6 +718,7 @@ void git_revwalk_free(git_revwalk *walk)
 	if (walk == NULL)
 		return;
 
+	git_commit_graph_free(walk->cgraph);
 	git_revwalk_reset(walk);
 	git_odb_free(walk->odb);
 
@@ -817,4 +838,3 @@ int git_revwalk_add_hide_cb(
 
 	return 0;
 }
-
diff --git a/src/revwalk.h b/src/revwalk.h
@@ -11,6 +11,7 @@
 
 #include "git2/revwalk.h"
 #include "oidmap.h"
+#include "commit_graph.h"
 #include "commit_list.h"
 #include "pqueue.h"
 #include "pool.h"
@@ -21,6 +22,7 @@
 struct git_revwalk {
 	git_repository *repo;
 	git_odb *odb;
+	git_commit_graph_file *cgraph;
 
 	git_oidmap *commits;
 	git_pool commit_pool;
