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

Re: [PATCH] rebase -i: do not update "done" when rescheduling command

From
Junio C Hamano <gitster@pobox.com>
Date
Mar 20, 2023, 17:46 UTC
Message-ID
<xmqq4jqfmi2k.fsf@gitster.g>
In-Reply-To
<pull.1492.git.1679237337683.gitgitgadget@gmail.com>
"Phillip Wood via GitGitGadget" <gitgitgadget@gmail.com> writes:
Show 17 quoted lines
> From: Phillip Wood <phillip.wood@dunelm.org.uk>
>
> As the sequencer executes todo commands it appends them to
> .git/rebase-merge/done. This file is used by "git status" to show the
> recently executed commands. Unfortunately when a command is rescheduled
> the command preceding it is erroneously appended to the "done" file.
> This means that when rebase stops after rescheduling "pick B" the "done"
> file contains
>
> 	pick A
> 	pick B
> 	pick A
>
> instead of
>
> 	pick A
> 	pick B

Here it may not be clear what you meant with the verb "reschedule" to those who weren't closely following the previous discussion that led to this fix.

Is it the same as "the command attempted to execute a step, couldn't complete it (e.g. due to conflicts), and gave control to the end user until they say 'git rebase --continue'"? What cases, other than interrupted step due to conflicts, involve "rescheduling"?

> Note that the rescheduled command will still be appended to the "done"
> file again when it is successfully executed. Arguably it would be better
> not to do that but fixing it would be more involved.

And without quite understanding what "reschedule" refers to, it is unclear why it is even arguable---it is perfectly sensible that a command that is rescheduled (hence not yet done) would not be sent to 'done'. If a command that was once rescheduled (hence it wasn't finished initially) gets finished now, shouldn't it be sent to 'done'? It is unclear why is it better not to.

> -static int save_todo(struct todo_list *todo_list, struct replay_opts *opts)
> +static int save_todo(struct todo_list *todo_list, struct replay_opts *opts,
> +		     int reschedule)
>  {

OK, all callers to save_todo() are in pick_commits() that knows what the value of "reschedule" is, and it is passed down to this helper ...

Show 18 quoted lines
> @@ -3389,7 +3390,7 @@ static int save_todo(struct todo_list *todo_list, struct replay_opts *opts)
>  	 * rebase -i writes "git-rebase-todo" without the currently executing
>  	 * command, appending it to "done" instead.
>  	 */
> -	if (is_rebase_i(opts))
> +	if (is_rebase_i(opts) && !reschedule)
>  		next++;
>  
>  	fd = hold_lock_file_for_update(&todo_lock, todo_path, 0);
> @@ -3402,7 +3403,7 @@ static int save_todo(struct todo_list *todo_list, struct replay_opts *opts)
>  	if (commit_lock_file(&todo_lock) < 0)
>  		return error(_("failed to finalize '%s'"), todo_path);
>  
> -	if (is_rebase_i(opts) && next > 0) {
> +	if (is_rebase_i(opts) && !reschedule && next > 0) {
>  		const char *done = rebase_path_done();
>  		int fd = open(done, O_CREAT | O_WRONLY | O_APPEND, 0666);
>  		int ret = 0;

... and the change here is quite straight-forward. With reschedule, we do not advance because by definition we haven't finished the step yet. OK.

Show 7 quoted lines
> @@ -4648,7 +4649,7 @@ static int pick_commits(struct repository *r,
>  		const char *arg = todo_item_get_arg(todo_list, item);
>  		int check_todo = 0;
>  
> -		if (save_todo(todo_list, opts))
> +		if (save_todo(todo_list, opts, 0))
>  			return -1;

I wonder why we pass a hardcoded 0 here---shouldn't the value match the local variable 'reschedule'? here?

The same question for the other two callers, but I admit that when the second one is called, the local variable "reschedule" is not set...

Show 10 quoted lines
>  		if (is_rebase_i(opts)) {
>  			if (item->command != TODO_COMMENT) {
> @@ -4695,8 +4696,7 @@ static int pick_commits(struct repository *r,
>  							    todo_list->current),
>  				       get_item_line(todo_list,
>  						     todo_list->current));
> -				todo_list->current--;
> -				if (save_todo(todo_list, opts))
> +				if (save_todo(todo_list, opts, 1))
>  					return -1;
... yet we call the helper with reschedule set to 1.  Puzzled.
Show 8 quoted lines
> @@ -4788,8 +4788,7 @@ static int pick_commits(struct repository *r,
>  			       get_item_line_length(todo_list,
>  						    todo_list->current),
>  			       get_item_line(todo_list, todo_list->current));
> -			todo_list->current--;
> -			if (save_todo(todo_list, opts))
> +			if (save_todo(todo_list, opts, 1))
>  				return -1;

At this point, reschedule is set and passing it instead of 1 would be OK.

Thanks.
Previous: Stefan HallerNext: Phillip Wood
Message 3 of 80 in “rebase -i: do not update "done" when rescheduling command”
  1. rebase -i: do not update "done" when rescheduling commandPhillip Wood via GitGitGadget, Mar 19, 2023
  2. Stefan HallerMar 20, 2023
  3. Junio C HamanoMar 20, 2023
  4. Phillip WoodMar 24, 2023
  5. Junio C HamanoMar 24, 2023
  6. Phillip WoodMar 24, 2023
  7. Johannes SchindelinMar 27, 2023
  8. Phillip WoodAug 3, 2023
  9. Johannes SchindelinAug 23, 2023
  10. 0/6 rebase -i: impove handling of failed commandsPhillip Wood via GitGitGadget, Apr 21, 2023
  11. 1/6 rebase -i: move unlink() callsPhillip Wood via GitGitGadget, Apr 21, 2023
  12. Junio C HamanoApr 21, 2023
  13. Phillip WoodApr 27, 2023
  14. 2/6 rebase -i: remove patch file after conflict resolutionPhillip Wood via GitGitGadget, Apr 21, 2023
  15. Junio C HamanoApr 21, 2023
  16. Phillip WoodApr 27, 2023
  17. Glen ChooJun 21, 2023
  18. Phillip WoodJul 14, 2023
  19. Junio C HamanoJul 14, 2023
  20. Phillip WoodJul 17, 2023
  21. 3/6 sequencer: factor out part of pick_commits()Phillip Wood via GitGitGadget, Apr 21, 2023
  22. Eric SunshineApr 21, 2023
  23. Junio C HamanoApr 21, 2023
  24. Phillip WoodApr 21, 2023
  25. Junio C HamanoApr 21, 2023
  26. 4/6 rebase --continue: refuse to commit after failed commandPhillip Wood via GitGitGadget, Apr 21, 2023
  27. Eric SunshineApr 21, 2023
  28. Junio C HamanoApr 21, 2023
  29. Glen ChooJun 21, 2023
  30. 5/6 rebase: fix rewritten list for failed pickPhillip Wood via GitGitGadget, Apr 21, 2023
  31. Glen ChooJun 21, 2023
  32. Phillip WoodJul 25, 2023
  33. Glen ChooJul 25, 2023
  34. Phillip WoodJul 26, 2023
  35. Glen ChooJul 26, 2023
  36. Phillip WoodJul 28, 2023
  37. 6/6 rebase -i: fix adding failed command to the todo listPhillip Wood via GitGitGadget, Apr 21, 2023
  38. Glen ChooJun 21, 2023
  39. Junio C HamanoApr 21, 2023
  40. Glen ChooJun 21, 2023
  41. 0/7 rebase -i: impove handling of failed commandsPhillip Wood via GitGitGadget, Aug 1, 2023
  42. 2/7 rebase -i: remove patch file after conflict resolutionPhillip Wood via GitGitGadget, Aug 1, 2023
  43. Junio C HamanoAug 1, 2023
  44. Phillip WoodAug 1, 2023
  45. 1/7 rebase -i: move unlink() callsPhillip Wood via GitGitGadget, Aug 1, 2023
  46. Junio C HamanoAug 1, 2023
  47. Phillip WoodAug 1, 2023
  48. Junio C HamanoAug 1, 2023
  49. 3/7 sequencer: use rebase_path_message()Phillip Wood via GitGitGadget, Aug 1, 2023
  50. Junio C HamanoAug 1, 2023
  51. Phillip WoodAug 1, 2023
  52. Junio C HamanoAug 2, 2023
  53. 4/7 sequencer: factor out part of pick_commits()Phillip Wood via GitGitGadget, Aug 1, 2023
  54. Johannes SchindelinAug 23, 2023
  55. 5/7 rebase: fix rewritten list for failed pickPhillip Wood via GitGitGadget, Aug 1, 2023
  56. Johannes SchindelinAug 23, 2023
  57. Phillip WoodSep 4, 2023
  58. 6/7 rebase --continue: refuse to commit after failed commandPhillip Wood via GitGitGadget, Aug 1, 2023
  59. Johannes SchindelinAug 23, 2023
  60. Phillip WoodSep 4, 2023
  61. Johannes SchindelinSep 5, 2023
  62. Junio C HamanoSep 5, 2023
  63. Phillip WoodSep 5, 2023
  64. 7/7 rebase -i: fix adding failed command to the todo listPhillip Wood via GitGitGadget, Aug 1, 2023
  65. Junio C HamanoAug 2, 2023
  66. Phillip WoodAug 3, 2023
  67. Phillip WoodAug 9, 2023
  68. Glen ChooAug 7, 2023
  69. Phillip WoodAug 9, 2023
  70. 0/7 rebase -i: impove handling of failed commandsPhillip Wood via GitGitGadget, Sep 6, 2023
  71. 3/7 sequencer: use rebase_path_message()Phillip Wood via GitGitGadget, Sep 6, 2023
  72. 2/7 rebase -i: remove patch file after conflict resolutionPhillip Wood via GitGitGadget, Sep 6, 2023
  73. 1/7 rebase -i: move unlink() callsPhillip Wood via GitGitGadget, Sep 6, 2023
  74. 4/7 sequencer: factor out part of pick_commits()Phillip Wood via GitGitGadget, Sep 6, 2023
  75. 5/7 rebase: fix rewritten list for failed pickPhillip Wood via GitGitGadget, Sep 6, 2023
  76. 6/7 rebase --continue: refuse to commit after failed commandPhillip Wood via GitGitGadget, Sep 6, 2023
  77. 7/7 rebase -i: fix adding failed command to the todo listPhillip Wood via GitGitGadget, Sep 6, 2023
  78. Junio C HamanoSep 6, 2023
  79. Johannes SchindelinSep 7, 2023
  80. Junio C HamanoSep 7, 2023

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.