[PATCH 4/4] submodule: submodule_move_head omits old argument in forced case

[Stefan Beller <sbeller@google.com>]

When using hard reset or forced checkout with the option to recurse into
submodules, the submodules need to be reset, too.

It turns out that we need to omit the duplicate old argument to read-tree
in all forced cases to omit the 2 way merge and use the more assertive
behavior of reading the specific new tree into the index and updating
the working tree.

Signed-off-by: Stefan Beller <sbeller@google.com>
---
 submodule.c               |  4 +++-
 t/lib-submodule-update.sh | 11 +++++++++++
 2 files changed, 14 insertions(+), 1 deletion(-)

diff --git a/submodule.c b/submodule.c
index fa25888783..db0f7ac51e 100644
--- a/submodule.c
+++ b/submodule.c
@@ -1653,7 +1653,9 @@ int submodule_move_head(const char *path,
 	else
 		argv_array_push(&cp.args, "-m");
 
-	argv_array_push(&cp.args, old ? old : EMPTY_TREE_SHA1_HEX);
+	if (!(flags & SUBMODULE_MOVE_HEAD_FORCE))
+		argv_array_push(&cp.args, old ? old : EMPTY_TREE_SHA1_HEX);
+
 	argv_array_push(&cp.args, new ? new : EMPTY_TREE_SHA1_HEX);
 
 	if (run_command(&cp)) {
diff --git a/t/lib-submodule-update.sh b/t/lib-submodule-update.sh
index fb0173ea87..380ef4b4ae 100755
--- a/t/lib-submodule-update.sh
+++ b/t/lib-submodule-update.sh
@@ -1015,4 +1015,15 @@ test_submodule_forced_switch_recursing_with_args () {
 			test_submodule_content sub1 origin/modify_sub1
 		)
 	'
+
+	test_expect_success "$command: changed submodule worktree is reset" '
+		prolog &&
+		reset_work_tree_to_interested add_sub1 &&
+		(
+			cd submodule_update &&
+			rm sub1/file1 &&
+			$command HEAD &&
+			test_path_is_file sub1/file1
+		)
+	'
 }
-- 
2.15.1.620.gb9897f4670-goog