[PATCH v2 4/5] read-tree: use 'skip_cache_tree_update' option

From: Victoria Dye via GitGitGadget <gitgitgadget@gmail.com>
Date: Nov 10, 2022, 01:57 UTC
Message-ID: <5a646bc47c911bb6b58e00574ac30afab1eec00b.1668045438.git.gitgitgadget@gmail.com>
In-Reply-To: <pull.1411.v2.git.1668045438.gitgitgadget@gmail.com>

From: Victoria Dye <vdye@github.com>

When running 'read-tree' with a single tree and no prefix, 'prime_cache_tree()' is called after the tree is unpacked. In that situation, skip a redundant call to 'cache_tree_update()' in 'unpack_trees()' by enabling the 'skip_cache_tree_update' unpack option.

Removing the redundant cache tree update provides a substantial performance improvement to 'git read-tree <tree-ish>', as shown by a test added to 'p0006-read-tree-checkout.sh':

Test before after ---------------------------------------------------------------------- read-tree br_ballast_plus_1 3.94(1.80+1.57) 3.00(1.14+1.28) -23.9%

Note that the 'read-tree' in 't1022-read-tree-partial-clone.sh' is updated to read two trees, rather than one. The test was first introduced in d3da223f221 (cache-tree: prefetch in partial clone read-tree, 2021-07-23) to exercise the 'cache_tree_update()' code path, as used in 'git merge'. Since this patch drops the call to 'cache_tree_update()' in single-tree 'git read-tree', change the test to use the two-tree variant so that 'cache_tree_update()' is called as intended.

Signed-off-by: Victoria Dye <vdye@github.com>
---
 builtin/read-tree.c                | 4 ++++
 t/perf/p0006-read-tree-checkout.sh | 8 ++++++++
 t/t1022-read-tree-partial-clone.sh | 2 +-
 3 files changed, 13 insertions(+), 1 deletion(-)

diff --git a/builtin/read-tree.c b/builtin/read-tree.c
index f4cbe460b97..45c6652444b 100644
--- a/builtin/read-tree.c
+++ b/builtin/read-tree.c
@@ -249,6 +249,10 @@ int cmd_read_tree(int argc, const char **argv, const char *cmd_prefix)
 	if (opts.debug_unpack)
 		opts.fn = debug_merge;
 
+	/* If we're going to prime_cache_tree later, skip cache tree update */
+	if (nr_trees == 1 && !opts.prefix)
+		opts.skip_cache_tree_update = 1;
+
 	cache_tree_free(&active_cache_tree);
 	for (i = 0; i < nr_trees; i++) {
 		struct tree *tree = trees[i];
diff --git a/t/perf/p0006-read-tree-checkout.sh b/t/perf/p0006-read-tree-checkout.sh
index c481c012d2f..325566e18eb 100755
--- a/t/perf/p0006-read-tree-checkout.sh
+++ b/t/perf/p0006-read-tree-checkout.sh
@@ -49,6 +49,14 @@ test_perf "read-tree br_base br_ballast ($nr_files)" '
 	git read-tree -n -m br_base br_ballast
 '
 
+test_perf "read-tree br_ballast_plus_1 ($nr_files)" '
+	# Run read-tree 100 times for clearer performance results & comparisons
+	for i in  $(test_seq 100)
+	do
+		git read-tree -n -m br_ballast_plus_1 || return 1
+	done
+'
+
 test_perf "switch between br_base br_ballast ($nr_files)" '
 	git checkout -q br_base &&
 	git checkout -q br_ballast
diff --git a/t/t1022-read-tree-partial-clone.sh b/t/t1022-read-tree-partial-clone.sh
index a9953b6a71c..da539716359 100755
--- a/t/t1022-read-tree-partial-clone.sh
+++ b/t/t1022-read-tree-partial-clone.sh
@@ -19,7 +19,7 @@ test_expect_success 'read-tree in partial clone prefetches in one batch' '
 	git -C server config uploadpack.allowfilter 1 &&
 	git -C server config uploadpack.allowanysha1inwant 1 &&
 	git clone --bare --filter=blob:none "file://$(pwd)/server" client &&
-	GIT_TRACE_PACKET="$(pwd)/trace" git -C client read-tree $TREE &&
+	GIT_TRACE_PACKET="$(pwd)/trace" git -C client read-tree $TREE $TREE &&
 
 	# "done" marks the end of negotiation (once per fetch). Expect that
 	# only one fetch occurs.

-- 
gitgitgadget