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

[PATCH v3 7/7] merge: do not exit restore_state() prematurely

From
Elijah Newren via GitGitGadget <gitgitgadget@gmail.com>
Date
Jul 21, 2022, 08:16 UTC
Message-ID
<81c40492a62e81100c66a8ccc1ec200fb2e6fca5.1658391391.git.gitgitgadget@gmail.com>
In-Reply-To
<pull.1231.v3.git.1658391391.gitgitgadget@gmail.com>
From: Elijah Newren <newren@gmail.com>
Previously, if the user:
* Had no local changes before starting the merge
* A merge strategy makes changes to the working tree/index but returns
  with exit status 2

Then we'd call restore_state() to clean up the changes and either let the next merge strategy run (if there is one), or exit telling the user that no merge strategy could handle the merge. Unfortunately, restore_state() did not clean up the changes as expected; that function was a no-op if the stash was a null, and the stash would be null if there were no local changes before starting the merge. So, instead of "Rewinding the tree to pristine..." as the code claimed, restore_state() would leave garbage around in the index and working tree (possibly including conflicts) for either the next merge strategy or for the user after aborting the merge. And in the case of aborting the merge, the user would be unable to run "git merge --abort" to get rid of the unintended leftover conflicts, because the merge control files were not written as it was presumed that we had restored to a clean state already.

Fix the main problem by making sure that restore_state() only skips the stash application if the stash is null rather than skipping the whole function.

However, there is a secondary problem -- since merge.c forks subprocesses to do the cleanup, the in-memory index is left out-of-sync. While there was a refresh_cache(REFRESH_QUIET) call that attempted to correct that, that function would not handle cases where the previous merge strategy added conflicted entries. We need to drop the index and re-read it to handle such cases.

(Alternatively, we could stop forking subprocesses and instead call some appropriate function to do the work which would update the in-memory index automatically. For now, just do the simple fix.)

Also, add a testcase checking this, one for which the octopus strategy fails on the first commit it attempts to merge, and thus which it cannot handle at all and must completely bail on (as per the "exit 2" code path of commit 98efc8f3d8 ("octopus: allow manual resolve on the last round.", 2006-01-13)).

Reported-by: ZheNing Hu <adlternative@gmail.com>
Signed-off-by: Elijah Newren <newren@gmail.com>
---
 builtin/merge.c        | 10 ++++++----
 t/t7607-merge-state.sh | 32 ++++++++++++++++++++++++++++++++
 2 files changed, 38 insertions(+), 4 deletions(-)
 create mode 100755 t/t7607-merge-state.sh
diff --git a/builtin/merge.c b/builtin/merge.c
index 11bb4bab0a1..7fb4414ebb7 100644
--- a/builtin/merge.c
+++ b/builtin/merge.c
@@ -386,11 +386,11 @@ static void restore_state(const struct object_id *head,
 	const char *args[] = { "stash", "apply", "--index", "--quiet",
 			       NULL, NULL };
 
-	if (is_null_oid(stash))
-		return;
-
 	reset_hard(head, 1);
 
+	if (is_null_oid(stash))
+		goto refresh_cache;
+
 	args[4] = oid_to_hex(stash);
 
 	/*
@@ -399,7 +399,9 @@ static void restore_state(const struct object_id *head,
 	 */
 	run_command_v_opt(args, RUN_GIT_CMD);
 
-	refresh_cache(REFRESH_QUIET);
+refresh_cache:
+	if (discard_cache() < 0 || read_cache() < 0)
+		die(_("could not read index"));
 }
 
 /* This is called when no merge was necessary. */
diff --git a/t/t7607-merge-state.sh b/t/t7607-merge-state.sh
new file mode 100755
index 00000000000..fc33d57357b
--- /dev/null
+++ b/t/t7607-merge-state.sh
@@ -0,0 +1,32 @@
+#!/bin/sh
+
+test_description="Test that merge state is as expected after failed merge"
+
+GIT_TEST_DEFAULT_INITIAL_BRANCH_NAME=main
+export GIT_TEST_DEFAULT_INITIAL_BRANCH_NAME
+. ./test-lib.sh
+
+test_expect_success 'set up custom strategy' '
+	test_commit --no-tag "Initial" base base &&
+
+	for b in branch1 branch2 branch3
+	do
+		git checkout -b $b main &&
+		test_commit --no-tag "Change on $b" base $b || return 1
+	done &&
+
+	git checkout branch1 &&
+	# This is a merge that octopus cannot handle.  Note, that it does not
+	# just hit conflicts, it completely fails and says that it cannot
+	# handle this type of merge.
+	test_expect_code 2 git merge branch2 branch3 >output 2>&1 &&
+	grep "fatal: merge program failed" output &&
+	grep "Should not be doing an octopus" output &&
+
+	# Make sure we did not leave stray changes around when no appropriate
+	# merge strategy was found
+	git diff --exit-code --name-status &&
+	test_path_is_missing .git/MERGE_HEAD
+'
+
+test_done
-- 
gitgitgadget
Previous: Elijah NewrenNext: Junio C Hamano
Message 57 of 87 in “Fix merge restore state”
  1. 0/2 Fix merge restore stateElijah Newren via GitGitGadget, May 19, 2022
  2. 1/2 merge: remove unused variableElijah Newren via GitGitGadget, May 19, 2022
  3. Junio C HamanoMay 19, 2022
  4. 2/2 merge: make restore_state() do as its name saysElijah Newren via GitGitGadget, May 19, 2022
  5. Junio C HamanoMay 19, 2022
  6. Junio C HamanoMay 19, 2022
  7. Elijah NewrenJun 12, 2022
  8. ZheNing HuJun 12, 2022
  9. 0/6 Fix merge restore stateElijah Newren via GitGitGadget, Jun 19, 2022
  10. 1/6 t6424: make sure a failed merge preserves local changesJunio C Hamano via GitGitGadget, Jun 19, 2022
  11. 2/6 merge: remove unused variableElijah Newren via GitGitGadget, Jun 19, 2022
  12. Junio C HamanoJul 19, 2022
  13. 4/6 merge: make restore_state() restore staged state tooElijah Newren via GitGitGadget, Jun 19, 2022
  14. ZheNing HuJul 17, 2022
  15. Junio C HamanoJul 19, 2022
  16. Junio C HamanoJul 19, 2022
  17. Elijah NewrenJul 21, 2022
  18. 3/6 merge: fix save_state() to work when there are racy-dirty filesElijah Newren via GitGitGadget, Jun 19, 2022
  19. ZheNing HuJul 17, 2022
  20. Junio C HamanoJul 19, 2022
  21. Elijah NewrenJul 21, 2022
  22. Junio C HamanoJul 19, 2022
  23. 6/6 merge: do not exit restore_state() prematurelyElijah Newren via GitGitGadget, Jun 19, 2022
  24. ZheNing HuJul 17, 2022
  25. Junio C HamanoJul 19, 2022
  26. Eric SunshineJul 20, 2022
  27. Elijah NewrenJul 21, 2022
  28. Elijah NewrenJul 21, 2022
  29. 5/6 merge: ensure we can actually restore pre-merge stateElijah Newren via GitGitGadget, Jun 19, 2022
  30. ZheNing HuJul 17, 2022
  31. Junio C HamanoJul 19, 2022
  32. Elijah NewrenJul 21, 2022
  33. 0/7 Fix merge restore stateElijah Newren via GitGitGadget, Jul 21, 2022
  34. 1/7 merge-ort-wrappers: make printed message match the one from recursiveElijah Newren via GitGitGadget, Jul 21, 2022
  35. Junio C HamanoJul 21, 2022
  36. Elijah NewrenJul 21, 2022
  37. Junio C HamanoJul 21, 2022
  38. Elijah NewrenJul 21, 2022
  39. 2/7 merge-resolve: abort if index does not match HEADElijah Newren via GitGitGadget, Jul 21, 2022
  40. 3/7 merge: do not abort early if one strategy fails to handle the mergeElijah Newren via GitGitGadget, Jul 21, 2022
  41. Junio C HamanoJul 21, 2022
  42. Ævar Arnfjörð BjarmasonJul 25, 2022
  43. Elijah NewrenJul 26, 2022
  44. Ævar Arnfjörð BjarmasonJul 26, 2022
  45. 4/7 merge: fix save_state() to work when there are stat-dirty filesElijah Newren via GitGitGadget, Jul 21, 2022
  46. 6/7 merge: ensure we can actually restore pre-merge stateElijah Newren via GitGitGadget, Jul 21, 2022
  47. Junio C HamanoJul 21, 2022
  48. Ben HumphreysMar 2, 2023
  49. Elijah NewrenMar 2, 2023
  50. Junio C HamanoMar 2, 2023
  51. Rudy RigotMar 4, 2023
  52. Ben HumphreysMar 6, 2023
  53. 5/7 merge: make restore_state() restore staged state tooElijah Newren via GitGitGadget, Jul 21, 2022
  54. Junio C HamanoJul 21, 2022
  55. Junio C HamanoJul 21, 2022
  56. Elijah NewrenJul 21, 2022
  57. 7/7 merge: do not exit restore_state() prematurelyElijah Newren via GitGitGadget, Jul 21, 2022
  58. Junio C HamanoJul 21, 2022
  59. 0/7 Fix merge restore stateElijah Newren via GitGitGadget, Jul 22, 2022
  60. 1/7 merge-ort-wrappers: make printed message match the one from recursiveElijah Newren via GitGitGadget, Jul 22, 2022
  61. 2/7 merge-resolve: abort if index does not match HEADElijah Newren via GitGitGadget, Jul 22, 2022
  62. Ævar Arnfjörð BjarmasonJul 22, 2022
  63. Elijah NewrenJul 23, 2022
  64. Ævar Arnfjörð BjarmasonJul 23, 2022
  65. Elijah NewrenJul 26, 2022
  66. Ævar Arnfjörð BjarmasonJul 26, 2022
  67. 3/7 merge: do not abort early if one strategy fails to handle the mergeElijah Newren via GitGitGadget, Jul 22, 2022
  68. Ævar Arnfjörð BjarmasonJul 22, 2022
  69. Elijah NewrenJul 23, 2022
  70. 4/7 merge: fix save_state() to work when there are stat-dirty filesElijah Newren via GitGitGadget, Jul 22, 2022
  71. 5/7 merge: make restore_state() restore staged state tooElijah Newren via GitGitGadget, Jul 22, 2022
  72. Ævar Arnfjörð BjarmasonJul 22, 2022
  73. Elijah NewrenJul 23, 2022
  74. 6/7 merge: ensure we can actually restore pre-merge stateElijah Newren via GitGitGadget, Jul 22, 2022
  75. 7/7 merge: do not exit restore_state() prematurelyElijah Newren via GitGitGadget, Jul 22, 2022
  76. 0/8 Fix merge restore stateElijah Newren via GitGitGadget, Jul 23, 2022
  77. 2/8 merge-resolve: abort if index does not match HEADElijah Newren via GitGitGadget, Jul 23, 2022
  78. 1/8 merge-ort-wrappers: make printed message match the one from recursiveElijah Newren via GitGitGadget, Jul 23, 2022
  79. 4/8 merge: do not abort early if one strategy fails to handle the mergeElijah Newren via GitGitGadget, Jul 23, 2022
  80. 3/8 merge: abort if index does not match HEAD for trivial mergesElijah Newren via GitGitGadget, Jul 23, 2022
  81. 5/8 merge: fix save_state() to work when there are stat-dirty filesElijah Newren via GitGitGadget, Jul 23, 2022
  82. 6/8 merge: make restore_state() restore staged state tooElijah Newren via GitGitGadget, Jul 23, 2022
  83. 7/8 merge: ensure we can actually restore pre-merge stateElijah Newren via GitGitGadget, Jul 23, 2022
  84. 8/8 merge: do not exit restore_state() prematurelyElijah Newren via GitGitGadget, Jul 23, 2022
  85. Junio C HamanoJul 25, 2022
  86. Elijah NewrenJul 26, 2022
  87. ZheNing HuJul 26, 2022

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.