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

Re: [PATCH v4 5/9] rebase: add coverage of other incompatible options

From
PWPhillip Wood <phillip.wood123@gmail.com>
Date
Jan 24, 2023, 16:48 UTC
Message-ID
<0418e5b6-8cbd-9dc9-085e-31380beda89b@dunelm.org.uk>
In-Reply-To
<CABPp-BHVUc7EdY9z_TPcHspCak6Yc3mfDXUkivj4zq_fJem3SQ@mail.gmail.com>
Hi Elijah
On 24/01/2023 15:41, Elijah Newren wrote:
Show 65 quoted lines
> Ah, despite looking over the code multiple times to check my
> statements, I somehow kept missing this:
> 
>      if (keep_base && options.reapply_cherry_picks)
>          options.upstream = options.onto;
> 
> which is how --[no-]reapply-cherry-picks is supported in conjunction
> with --keep-base.  Thanks.
> 
>>>> For all four cases (A)-(D), the apply backend will ignore whatever
>>>> --[no-]reapply-cherry-picks flag is provided.
>>>
>>> For (D) the flag is respected, (C) errors out, the other cases
>>> correspond to the default so it's like saying
>>>
>>>       git rebase --merge --no-reapply-cherry-picks
>>>
>>> ignores the flag.
>>
>> On reflection that is only true for (B). I agree that we should error
>> out on (A) which we don't at the moment.
>>
>> I'd support a change that errors out on (A) and (C) but continues to
>> allow (B) and (D). I think we can do that with the diff below
>>
>>
>> diff --git a/builtin/rebase.c b/builtin/rebase.c
>> index 1481c5b6a5..66aef356b8 100644
>> --- a/builtin/rebase.c
>> +++ b/builtin/rebase.c
>> @@ -1230,12 +1230,6 @@ int cmd_rebase(int argc, const char **argv, const
>> char *prefix)
>>                    if (options.fork_point < 0)
>>                            options.fork_point = 0;
>>            }
>> -        /*
>> -         * --keep-base defaults to --reapply-cherry-picks to avoid losing
>> -         * commits when using this option.
>> -         */
>> -        if (options.reapply_cherry_picks < 0)
>> -                options.reapply_cherry_picks = keep_base;
>>
>>            if (options.root && options.fork_point > 0)
>>                    die(_("options '%s' and '%s' cannot be used
>> together"), "--root", "--fork-point");
>> @@ -1412,11 +1406,17 @@ int cmd_rebase(int argc, const char **argv,
>> const char *prefix)
>>            if (options.empty != EMPTY_UNSPECIFIED)
>>                    imply_merge(&options, "--empty");
>>
>> -        /*
>> -         * --keep-base implements --reapply-cherry-picks by altering
>> upstream so
>> -         * it works with both backends.
>> -         */
>> -        if (options.reapply_cherry_picks && !keep_base)
>> +        if (options.reapply_cherry_picks < 0)
>> +                /*
>> +                 * --keep-base defaults to --reapply-cherry-picks to
>> +                 * avoid losing commits when using this option.
>> +                 */
> 
> I know you were copying the previous comment, but this comment is
> really confusing to me.  Shouldn't it read "--reapply-cherry-picks
> defaults to --keep-base" instead of vice-versa?

Clearly this comment has not been as helpful as I indented it to be. I think maybe we should spell out the defaults with and without --keep-base. Perhaps something like

default to --no-reapply-cherry-picks unless --keep-base is given in which case default to --reapply-cherry-picks

Show 7 quoted lines
>> +                options.reapply_cherry_picks = keep_base;
>> +        else if (!keep_base)
>> +                /*
>> +                 * --keep-base implements --reapply-cherry-picks by
> 
> Should this be --[no-]reapply-cherry-picks, to clarify that it handles
> both cases?  Especially given how many times I missed it?

This has obviously proved to be confusing. The aim was to explain that in order to work with the apply backend "[--reapply-cherry-picks] --keep-base" was doing something unusual with `upstream` to reapply cherry picks. "--no-reapply-cherry-picks --keep-base" does not do anything unusual with `upstream`. I don't think changing it to --[no-]reapply-cherry-picks quite captures that. I came up with

To support --reapply-cherry-picks (which is not supported by the apply backend) --keep-base alters upstream to prevent cherry picked commits from being dropped.

but it really needs to mention that --keep-base also supports --no-reapply-cherry-picks in the usual way

Show 9 quoted lines
>> +                 * altering upstream so it works with both backends.
>> +                 */
>>                    imply_merge(&options, "--reapply-cherry-picks");
> 
> And perhaps this should be
> 
>      imply_merge(&options, options.reapply_cherry_picks ?
> "--reapply-cherry-picks" :
>           "--no-reapply-cherry-picks");
That's a good idea
> Also, the comment in git-rebase.txt about incompatibilities would become
> 
>       * --[no-]reapply-cherry-picks, when --keep-base is not provided
Yes, that's good
Thanks again for working on this
Phillip
Previous: Elijah NewrenNext: Elijah Newren
Message 45 of 70 in “rebase: mark --update-refs as requiring the merge backend”
  1. rebase: mark --update-refs as requiring the merge backendElijah Newren via GitGitGadget, Jan 19, 2023
  2. Derrick StoleeJan 19, 2023
  3. Elijah NewrenJan 20, 2023
  4. Junio C HamanoJan 20, 2023
  5. Elijah NewrenJan 20, 2023
  6. 0/2 rebase: mark --update-refs as requiring the merge backendElijah Newren via GitGitGadget, Jan 20, 2023
  7. 1/2 rebase: remove completely useless -C optionElijah Newren via GitGitGadget, Jan 20, 2023
  8. Eric SunshineJan 20, 2023
  9. Elijah NewrenJan 20, 2023
  10. Martin ÅgrenJan 20, 2023
  11. Elijah NewrenJan 20, 2023
  12. Junio C HamanoJan 20, 2023
  13. Elijah NewrenJan 20, 2023
  14. Junio C HamanoJan 20, 2023
  15. Elijah NewrenJan 21, 2023
  16. Junio C HamanoJan 22, 2023
  17. 2/2 rebase: mark --update-refs as requiring the merge backendElijah Newren via GitGitGadget, Jan 20, 2023
  18. Phillip WoodJan 20, 2023
  19. Elijah NewrenJan 21, 2023
  20. 0/7 rebase: fix several code/testing/documentation issues around flag incompatibilitiesElijah Newren via GitGitGadget, Jan 21, 2023
  21. 1/7 rebase: mark --update-refs as requiring the merge backendElijah Newren via GitGitGadget, Jan 21, 2023
  22. 2/7 rebase: flag --apply and --merge as incompatibleElijah Newren via GitGitGadget, Jan 21, 2023
  23. 3/7 rebase: remove --allow-empty-message from incompatible optsElijah Newren via GitGitGadget, Jan 21, 2023
  24. Phillip WoodJan 21, 2023
  25. 4/7 rebase: fix docs about incompatibilities with --rootElijah Newren via GitGitGadget, Jan 21, 2023
  26. 5/7 rebase: add coverage of other incompatible optionsElijah Newren via GitGitGadget, Jan 21, 2023
  27. Phillip WoodJan 21, 2023
  28. Phillip WoodJan 21, 2023
  29. Elijah NewrenJan 22, 2023
  30. 6/7 rebase: clarify the OPT_CMDMODE incompatibilitiesElijah Newren via GitGitGadget, Jan 21, 2023
  31. 7/7 rebase: fix formatting of rebase --reapply-cherry-picks option in docsElijah Newren via GitGitGadget, Jan 21, 2023
  32. Phillip WoodJan 21, 2023
  33. 0/9 rebase: fix several code/testing/documentation issues around flag incompatibilitiesElijah Newren via GitGitGadget, Jan 22, 2023
  34. 2/9 rebase: flag --apply and --merge as incompatibleElijah Newren via GitGitGadget, Jan 22, 2023
  35. 1/9 rebase: mark --update-refs as requiring the merge backendElijah Newren via GitGitGadget, Jan 22, 2023
  36. 3/9 rebase: remove --allow-empty-message from incompatible optsElijah Newren via GitGitGadget, Jan 22, 2023
  37. 8/9 rebase: put rebase_options initialization in single placeElijah Newren via GitGitGadget, Jan 22, 2023
  38. 5/9 rebase: add coverage of other incompatible optionsElijah Newren via GitGitGadget, Jan 22, 2023
  39. Phillip WoodJan 23, 2023
  40. Elijah NewrenJan 24, 2023
  41. Phillip WoodJan 24, 2023
  42. Phillip WoodJan 24, 2023
  43. Junio C HamanoJan 24, 2023
  44. Elijah NewrenJan 24, 2023
  45. Phillip WoodJan 24, 2023
  46. Elijah NewrenJan 24, 2023
  47. Phillip WoodJan 24, 2023
  48. 9/9 rebase: provide better error message for apply options vs. merge configElijah Newren via GitGitGadget, Jan 22, 2023
  49. 4/9 rebase: fix docs about incompatibilities with --rootElijah Newren via GitGitGadget, Jan 22, 2023
  50. 7/9 rebase: fix formatting of rebase --reapply-cherry-picks option in docsElijah Newren via GitGitGadget, Jan 22, 2023
  51. 6/9 rebase: clarify the OPT_CMDMODE incompatibilitiesElijah Newren via GitGitGadget, Jan 22, 2023
  52. Derrick StoleeJan 23, 2023
  53. Elijah NewrenJan 24, 2023
  54. 00/10 rebase: fix several code/testing/documentation issues around flag incompatibilitiesElijah Newren via GitGitGadget, Jan 25, 2023
  55. 01/10 rebase: mark --update-refs as requiring the merge backendElijah Newren via GitGitGadget, Jan 25, 2023
  56. 02/10 rebase: flag --apply and --merge as incompatibleElijah Newren via GitGitGadget, Jan 25, 2023
  57. 03/10 rebase: remove --allow-empty-message from incompatible optsElijah Newren via GitGitGadget, Jan 25, 2023
  58. 04/10 rebase: fix docs about incompatibilities with --rootElijah Newren via GitGitGadget, Jan 25, 2023
  59. 05/10 rebase: fix incompatiblity checks for --[no-]reapply-cherry-picksElijah Newren via GitGitGadget, Jan 25, 2023
  60. Phillip WoodJan 25, 2023
  61. 08/10 rebase: fix formatting of rebase --reapply-cherry-picks option in docsElijah Newren via GitGitGadget, Jan 25, 2023
  62. 06/10 rebase: add coverage of other incompatible optionsElijah Newren via GitGitGadget, Jan 25, 2023
  63. 07/10 rebase: clarify the OPT_CMDMODE incompatibilitiesElijah Newren via GitGitGadget, Jan 25, 2023
  64. 10/10 rebase: provide better error message for apply options vs. merge configElijah Newren via GitGitGadget, Jan 25, 2023
  65. 09/10 rebase: put rebase_options initialization in single placeElijah Newren via GitGitGadget, Jan 25, 2023
  66. Phillip WoodJan 25, 2023
  67. Junio C HamanoJan 25, 2023
  68. Elijah NewrenJan 25, 2023
  69. rebase --merge vs --whitespace=fix, was Re: [PATCH v5 00/10] rebase: fix several code/testing/documentation issues around flag incompatibilitiesJohannes Schindelin, Feb 2, 2023
  70. Elijah NewrenFeb 2, 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.