1GIT-REBASE(1) Git Manual GIT-REBASE(1)
2
6 git-rebase - Reapply commits on top of another base tip
7
9 git rebase [-i | --interactive] [<options>] [--exec <cmd>]
10 [--onto <newbase> | --keep-base] [<upstream> [<branch>]]
11 git rebase [-i | --interactive] [<options>] [--exec <cmd>] [--onto <newbase>]
12 --root [<branch>]
13 git rebase (--continue | --skip | --abort | --quit | --edit-todo | --show-current-patch)
14
16 If <branch> is specified, git rebase will perform an automatic git
17 switch <branch> before doing anything else. Otherwise it remains on the
18 current branch.
19 If <upstream> is not specified, the upstream configured in
21 branch.<name>.remote and branch.<name>.merge options will be used (see
22 git-config(1) for details) and the --fork-point option is assumed. If
23 you are currently not on any branch or if the current branch does not
24 have a configured upstream, the rebase will abort.
25 All changes made by commits in the current branch but that are not in
27 <upstream> are saved to a temporary area. This is the same set of
28 commits that would be shown by git log <upstream>..HEAD; or by git log
29 'fork_point'..HEAD, if --fork-point is active (see the description on
30 --fork-point below); or by git log HEAD, if the --root option is
31 specified.
32 The current branch is reset to <upstream>, or <newbase> if the --onto
34 option was supplied. This has the exact same effect as git reset --hard
35 <upstream> (or <newbase>). ORIG_HEAD is set to point at the tip of the
36 branch before the reset.
37 The commits that were previously saved into the temporary area are then
39 reapplied to the current branch, one by one, in order. Note that any
40 commits in HEAD which introduce the same textual changes as a commit in
41 HEAD..<upstream> are omitted (i.e., a patch already accepted upstream
42 with a different commit message or timestamp will be skipped).
43 It is possible that a merge failure will prevent this process from
45 being completely automatic. You will have to resolve any such merge
46 failure and run git rebase --continue. Another option is to bypass the
47 commit that caused the merge failure with git rebase --skip. To check
48 out the original <branch> and remove the .git/rebase-apply working
49 files, use the command git rebase --abort instead.
50 Assume the following history exists and the current branch is "topic":
52 A---B---C topic
54 /
55 D---E---F---G master
56 From this point, the result of either of the following commands:
58 git rebase master
60 git rebase master topic
61 would be:
63 A'--B'--C' topic
65 /
66 D---E---F---G master
67 NOTE: The latter form is just a short-hand of git checkout topic
69 followed by git rebase master. When rebase exits topic will remain the
70 checked-out branch.
71 If the upstream branch already contains a change you have made (e.g.,
73 because you mailed a patch which was applied upstream), then that
74 commit will be skipped. For example, running git rebase master on the
75 following history (in which A' and A introduce the same set of changes,
76 but have different committer information):
77 A---B---C topic
79 /
80 D---E---A'---F master
81 will result in:
83 B'---C' topic
85 /
86 D---E---A'---F master
87 Here is how you would transplant a topic branch based on one branch to
89 another, to pretend that you forked the topic branch from the latter
90 branch, using rebase --onto.
91 First let’s assume your topic is based on branch next. For example, a
93 feature developed in topic depends on some functionality which is found
94 in next.
95 o---o---o---o---o master
97 \
98 o---o---o---o---o next
99 \
100 o---o---o topic
101 We want to make topic forked from branch master; for example, because
103 the functionality on which topic depends was merged into the more
104 stable master branch. We want our tree to look like this:
105 o---o---o---o---o master
107 | \
108 | o'--o'--o' topic
109 \
110 o---o---o---o---o next
111 We can get this using the following command:
113 git rebase --onto master next topic
115 Another example of --onto option is to rebase part of a branch. If we
117 have the following situation:
118 H---I---J topicB
120 /
121 E---F---G topicA
122 /
123 A---B---C---D master
124 then the command
126 git rebase --onto master topicA topicB
128 would result in:
130 H'--I'--J' topicB
132 /
133 | E---F---G topicA
134 |/
135 A---B---C---D master
136 This is useful when topicB does not depend on topicA.
138 A range of commits could also be removed with rebase. If we have the
140 following situation:
141 E---F---G---H---I---J topicA
143 then the command
145 git rebase --onto topicA~5 topicA~3 topicA
147 would result in the removal of commits F and G:
149 E---H'---I'---J' topicA
151 This is useful if F and G were flawed in some way, or should not be
153 part of topicA. Note that the argument to --onto and the <upstream>
154 parameter can be any valid commit-ish.
155 In case of conflict, git rebase will stop at the first problematic
157 commit and leave conflict markers in the tree. You can use git diff to
158 locate the markers (<<<<<<) and make edits to resolve the conflict. For
159 each file you edit, you need to tell Git that the conflict has been
160 resolved, typically this would be done with
161 git add <filename>
163 After resolving the conflict manually and updating the index with the
165 desired resolution, you can continue the rebasing process with
166 git rebase --continue
168 Alternatively, you can undo the git rebase with
170 git rebase --abort
172
174 rebase.useBuiltin
175 Unused configuration variable. Used in Git versions 2.20 and 2.21
176 as an escape hatch to enable the legacy shellscript implementation
177 of rebase. Now the built-in rewrite of it in C is always used.
178 Setting this will emit a warning, to alert any remaining users that
179 setting this now does nothing.
180 rebase.backend
182 Default backend to use for rebasing. Possible choices are apply or
183 merge. In the future, if the merge backend gains all remaining
184 capabilities of the apply backend, this setting may become unused.
185 rebase.stat
187 Whether to show a diffstat of what changed upstream since the last
188 rebase. False by default.
189 rebase.autoSquash
191 If set to true enable --autosquash option by default.
192 rebase.autoStash
194 When set to true, automatically create a temporary stash entry
195 before the operation begins, and apply it after the operation ends.
196 This means that you can run rebase on a dirty worktree. However,
197 use with care: the final stash application after a successful
198 rebase might result in non-trivial conflicts. This option can be
199 overridden by the --no-autostash and --autostash options of git-
200 rebase(1). Defaults to false.
201 rebase.missingCommitsCheck
203 If set to "warn", git rebase -i will print a warning if some
204 commits are removed (e.g. a line was deleted), however the rebase
205 will still proceed. If set to "error", it will print the previous
206 warning and stop the rebase, git rebase --edit-todo can then be
207 used to correct the error. If set to "ignore", no checking is done.
208 To drop a commit without warning or error, use the drop command in
209 the todo list. Defaults to "ignore".
210 rebase.instructionFormat
212 A format string, as specified in git-log(1), to be used for the
213 todo list during an interactive rebase. The format will
214 automatically have the long commit hash prepended to the format.
215 rebase.abbreviateCommands
217 If set to true, git rebase will use abbreviated command names in
218 the todo list resulting in something like this:
219 p deadbee The oneline of the commit
221 p fa1afe1 The oneline of the next commit
222 ...
223 instead of:
225 pick deadbee The oneline of the commit
227 pick fa1afe1 The oneline of the next commit
228 ...
229 Defaults to false.
231 rebase.rescheduleFailedExec
233 Automatically reschedule exec commands that failed. This only makes
234 sense in interactive mode (or when an --exec option was provided).
235 This is the same as specifying the --reschedule-failed-exec option.
236
238 --onto <newbase>
239 Starting point at which to create the new commits. If the --onto
240 option is not specified, the starting point is <upstream>. May be
241 any valid commit, and not just an existing branch name.
242 As a special case, you may use "A...B" as a shortcut for the merge
244 base of A and B if there is exactly one merge base. You can leave
245 out at most one of A and B, in which case it defaults to HEAD.
246 --keep-base
248 Set the starting point at which to create the new commits to the
249 merge base of <upstream> <branch>. Running git rebase --keep-base
250 <upstream> <branch> is equivalent to running git rebase --onto
251 <upstream>... <upstream>.
252 This option is useful in the case where one is developing a feature
254 on top of an upstream branch. While the feature is being worked on,
255 the upstream branch may advance and it may not be the best idea to
256 keep rebasing on top of the upstream but to keep the base commit
257 as-is.
258 Although both this option and --fork-point find the merge base
260 between <upstream> and <branch>, this option uses the merge base as
261 the starting point on which new commits will be created, whereas
262 --fork-point uses the merge base to determine the set of commits
263 which will be rebased.
264 See also INCOMPATIBLE OPTIONS below.
266 <upstream>
268 Upstream branch to compare against. May be any valid commit, not
269 just an existing branch name. Defaults to the configured upstream
270 for the current branch.
271 <branch>
273 Working branch; defaults to HEAD.
274 --continue
276 Restart the rebasing process after having resolved a merge
277 conflict.
278 --abort
280 Abort the rebase operation and reset HEAD to the original branch.
281 If <branch> was provided when the rebase operation was started,
282 then HEAD will be reset to <branch>. Otherwise HEAD will be reset
283 to where it was when the rebase operation was started.
284 --quit
286 Abort the rebase operation but HEAD is not reset back to the
287 original branch. The index and working tree are also left unchanged
288 as a result.
289 --apply: Use applying strategies to rebase (calling git-am internally).
291 This option may become a no-op in the future once the merge backend
292 handles everything the apply one does.
293 + See also INCOMPATIBLE OPTIONS below.
295 --empty={drop,keep,ask}
297 How to handle commits that are not empty to start and are not clean
298 cherry-picks of any upstream commit, but which become empty after
299 rebasing (because they contain a subset of already upstream
300 changes). With drop (the default), commits that become empty are
301 dropped. With keep, such commits are kept. With ask (implied by
302 --interactive), the rebase will halt when an empty commit is
303 applied allowing you to choose whether to drop it, edit files more,
304 or just commit the empty changes. Other options, like --exec, will
305 use the default of drop unless -i/--interactive is explicitly
306 specified.
307 Note that commits which start empty are kept, and commits which are
309 clean cherry-picks (as determined by git log --cherry-mark ...) are
310 always dropped.
311 See also INCOMPATIBLE OPTIONS below.
313 --keep-empty
315 No-op. Rebasing commits that started empty (had no change relative
316 to their parent) used to fail and this option would override that
317 behavior, allowing commits with empty changes to be rebased. Now
318 commits with no changes do not cause rebasing to halt.
319 See also BEHAVIORAL DIFFERENCES and INCOMPATIBLE OPTIONS below.
321 --allow-empty-message
323 No-op. Rebasing commits with an empty message used to fail and this
324 option would override that behavior, allowing commits with empty
325 messages to be rebased. Now commits with an empty message do not
326 cause rebasing to halt.
327 See also INCOMPATIBLE OPTIONS below.
329 --skip
331 Restart the rebasing process by skipping the current patch.
332 --edit-todo
334 Edit the todo list during an interactive rebase.
335 --show-current-patch
337 Show the current patch in an interactive rebase or when rebase is
338 stopped because of conflicts. This is the equivalent of git show
339 REBASE_HEAD.
340 -m, --merge
342 Use merging strategies to rebase. When the recursive (default)
343 merge strategy is used, this allows rebase to be aware of renames
344 on the upstream side. This is the default.
345 Note that a rebase merge works by replaying each commit from the
347 working branch on top of the <upstream> branch. Because of this,
348 when a merge conflict happens, the side reported as ours is the
349 so-far rebased series, starting with <upstream>, and theirs is the
350 working branch. In other words, the sides are swapped.
351 See also INCOMPATIBLE OPTIONS below.
353 -s <strategy>, --strategy=<strategy>
355 Use the given merge strategy. If there is no -s option git
356 merge-recursive is used instead. This implies --merge.
357 Because git rebase replays each commit from the working branch on
359 top of the <upstream> branch using the given strategy, using the
360 ours strategy simply empties all patches from the <branch>, which
361 makes little sense.
362 See also INCOMPATIBLE OPTIONS below.
364 -X <strategy-option>, --strategy-option=<strategy-option>
366 Pass the <strategy-option> through to the merge strategy. This
367 implies --merge and, if no strategy has been specified, -s
368 recursive. Note the reversal of ours and theirs as noted above for
369 the -m option.
370 See also INCOMPATIBLE OPTIONS below.
372 --rerere-autoupdate, --no-rerere-autoupdate
374 Allow the rerere mechanism to update the index with the result of
375 auto-conflict resolution if possible.
376 -S[<keyid>], --gpg-sign[=<keyid>]
378 GPG-sign commits. The keyid argument is optional and defaults to
379 the committer identity; if specified, it must be stuck to the
380 option without a space.
381 -q, --quiet
383 Be quiet. Implies --no-stat.
384 -v, --verbose
386 Be verbose. Implies --stat.
387 --stat
389 Show a diffstat of what changed upstream since the last rebase. The
390 diffstat is also controlled by the configuration option
391 rebase.stat.
392 -n, --no-stat
394 Do not show a diffstat as part of the rebase process.
395 --no-verify
397 This option bypasses the pre-rebase hook. See also githooks(5).
398 --verify
400 Allows the pre-rebase hook to run, which is the default. This
401 option can be used to override --no-verify. See also githooks(5).
402 -C<n>
404 Ensure at least <n> lines of surrounding context match before and
405 after each change. When fewer lines of surrounding context exist
406 they all must match. By default no context is ever ignored. Implies
407 --apply.
408 See also INCOMPATIBLE OPTIONS below.
410 --no-ff, --force-rebase, -f
412 Individually replay all rebased commits instead of fast-forwarding
413 over the unchanged ones. This ensures that the entire history of
414 the rebased branch is composed of new commits.
415 You may find this helpful after reverting a topic branch merge, as
417 this option recreates the topic branch with fresh commits so it can
418 be remerged successfully without needing to "revert the reversion"
419 (see the revert-a-faulty-merge How-To[1] for details).
420 --fork-point, --no-fork-point
422 Use reflog to find a better common ancestor between <upstream> and
423 <branch> when calculating which commits have been introduced by
424 <branch>.
425 When --fork-point is active, fork_point will be used instead of
427 <upstream> to calculate the set of commits to rebase, where
428 fork_point is the result of git merge-base --fork-point <upstream>
429 <branch> command (see git-merge-base(1)). If fork_point ends up
430 being empty, the <upstream> will be used as a fallback.
431 If either <upstream> or --root is given on the command line, then
433 the default is --no-fork-point, otherwise the default is
434 --fork-point.
435 If your branch was based on <upstream> but <upstream> was rewound
437 and your branch contains commits which were dropped, this option
438 can be used with --keep-base in order to drop those commits from
439 your branch.
440 --ignore-whitespace, --whitespace=<option>
442 These flags are passed to the git apply program (see git-apply(1))
443 that applies the patch. Implies --apply.
444 See also INCOMPATIBLE OPTIONS below.
446 --committer-date-is-author-date, --ignore-date
448 These flags are passed to git am to easily change the dates of the
449 rebased commits (see git-am(1)).
450 See also INCOMPATIBLE OPTIONS below.
452 --signoff
454 Add a Signed-off-by: trailer to all the rebased commits. Note that
455 if --interactive is given then only commits marked to be picked,
456 edited or reworded will have the trailer added.
457 See also INCOMPATIBLE OPTIONS below.
459 -i, --interactive
461 Make a list of the commits which are about to be rebased. Let the
462 user edit that list before rebasing. This mode can also be used to
463 split commits (see SPLITTING COMMITS below).
464 The commit list format can be changed by setting the configuration
466 option rebase.instructionFormat. A customized instruction format
467 will automatically have the long commit hash prepended to the
468 format.
469 See also INCOMPATIBLE OPTIONS below.
471 -r, --rebase-merges[=(rebase-cousins|no-rebase-cousins)]
473 By default, a rebase will simply drop merge commits from the todo
474 list, and put the rebased commits into a single, linear branch.
475 With --rebase-merges, the rebase will instead try to preserve the
476 branching structure within the commits that are to be rebased, by
477 recreating the merge commits. Any resolved merge conflicts or
478 manual amendments in these merge commits will have to be
479 resolved/re-applied manually.
480 By default, or when no-rebase-cousins was specified, commits which
482 do not have <upstream> as direct ancestor will keep their original
483 branch point, i.e. commits that would be excluded by git-log(1)'s
484 --ancestry-path option will keep their original ancestry by
485 default. If the rebase-cousins mode is turned on, such commits are
486 instead rebased onto <upstream> (or <onto>, if specified).
487 The --rebase-merges mode is similar in spirit to the deprecated
489 --preserve-merges but works with interactive rebases, where commits
490 can be reordered, inserted and dropped at will.
491 It is currently only possible to recreate the merge commits using
493 the recursive merge strategy; Different merge strategies can be
494 used only via explicit exec git merge -s <strategy> [...]
495 commands.
496 See also REBASING MERGES and INCOMPATIBLE OPTIONS below.
498 -p, --preserve-merges
500 [DEPRECATED: use --rebase-merges instead] Recreate merge commits
501 instead of flattening the history by replaying commits a merge
502 commit introduces. Merge conflict resolutions or manual amendments
503 to merge commits are not preserved.
504 This uses the --interactive machinery internally, but combining it
506 with the --interactive option explicitly is generally not a good
507 idea unless you know what you are doing (see BUGS below).
508 See also INCOMPATIBLE OPTIONS below.
510 -x <cmd>, --exec <cmd>
512 Append "exec <cmd>" after each line creating a commit in the final
513 history. <cmd> will be interpreted as one or more shell commands.
514 Any command that fails will interrupt the rebase, with exit code 1.
515 You may execute several commands by either using one instance of
517 --exec with several commands:
518 git rebase -i --exec "cmd1 && cmd2 && ..."
520 or by giving more than one --exec:
522 git rebase -i --exec "cmd1" --exec "cmd2" --exec ...
524 If --autosquash is used, "exec" lines will not be appended for the
526 intermediate commits, and will only appear at the end of each
527 squash/fixup series.
528 This uses the --interactive machinery internally, but it can be run
530 without an explicit --interactive.
531 See also INCOMPATIBLE OPTIONS below.
533 --root
535 Rebase all commits reachable from <branch>, instead of limiting
536 them with an <upstream>. This allows you to rebase the root
537 commit(s) on a branch. When used with --onto, it will skip changes
538 already contained in <newbase> (instead of <upstream>) whereas
539 without --onto it will operate on every change. When used together
540 with both --onto and --preserve-merges, all root commits will be
541 rewritten to have <newbase> as parent instead.
542 See also INCOMPATIBLE OPTIONS below.
544 --autosquash, --no-autosquash
546 When the commit log message begins with "squash! ..." (or "fixup!
547 ..."), and there is already a commit in the todo list that matches
548 the same ..., automatically modify the todo list of rebase -i so
549 that the commit marked for squashing comes right after the commit
550 to be modified, and change the action of the moved commit from pick
551 to squash (or fixup). A commit matches the ... if the commit
552 subject matches, or if the ... refers to the commit’s hash. As a
553 fall-back, partial matches of the commit subject work, too. The
554 recommended way to create fixup/squash commits is by using the
555 --fixup/--squash options of git-commit(1).
556 If the --autosquash option is enabled by default using the
558 configuration variable rebase.autoSquash, this option can be used
559 to override and disable this setting.
560 See also INCOMPATIBLE OPTIONS below.
562 --autostash, --no-autostash
564 Automatically create a temporary stash entry before the operation
565 begins, and apply it after the operation ends. This means that you
566 can run rebase on a dirty worktree. However, use with care: the
567 final stash application after a successful rebase might result in
568 non-trivial conflicts.
569 --reschedule-failed-exec, --no-reschedule-failed-exec
571 Automatically reschedule exec commands that failed. This only makes
572 sense in interactive mode (or when an --exec option was provided).
573
575 The following options:
576 · --apply
578 · --committer-date-is-author-date
580 · --ignore-date
582 · --ignore-whitespace
584 · --whitespace
586 · -C
588 are incompatible with the following options:
590 · --merge
592 · --strategy
594 · --strategy-option
596 · --allow-empty-message
598 · --[no-]autosquash
600 · --rebase-merges
602 · --preserve-merges
604 · --interactive
606 · --exec
608 · --keep-empty
610 · --empty=
612 · --edit-todo
614 · --root when used in combination with --onto
616 In addition, the following pairs of options are incompatible:
618 · --preserve-merges and --interactive
620 · --preserve-merges and --signoff
622 · --preserve-merges and --rebase-merges
624 · --preserve-merges and --empty=
626 · --keep-base and --onto
628 · --keep-base and --root
630
632 git rebase has two primary backends: apply and merge. (The apply
633 backend used to known as the am backend, but the name led to confusion
634 as it looks like a verb instead of a noun. Also, the merge backend used
635 to be known as the interactive backend, but it is now used for
636 non-interactive cases as well. Both were renamed based on lower-level
637 functionality that underpinned each.) There are some subtle differences
638 in how these two backends behave:
639 Empty commits
641 The apply backend unfortunately drops intentionally empty commits, i.e.
642 commits that started empty, though these are rare in practice. It also
643 drops commits that become empty and has no option for controlling this
644 behavior.
645 The merge backend keeps intentionally empty commits. Similar to the
647 apply backend, by default the merge backend drops commits that become
648 empty unless -i/--interactive is specified (in which case it stops and
649 asks the user what to do). The merge backend also has an
650 --empty={drop,keep,ask} option for changing the behavior of handling
651 commits that become empty.
652 Directory rename detection
654 Due to the lack of accurate tree information (arising from constructing
655 fake ancestors with the limited information available in patches),
656 directory rename detection is disabled in the apply backend. Disabled
657 directory rename detection means that if one side of history renames a
658 directory and the other adds new files to the old directory, then the
659 new files will be left behind in the old directory without any warning
660 at the time of rebasing that you may want to move these files into the
661 new directory.
662 Directory rename detection works with the merge backend to provide you
664 warnings in such cases.
665 Context
667 The apply backend works by creating a sequence of patches (by calling
668 format-patch internally), and then applying the patches in sequence
669 (calling am internally). Patches are composed of multiple hunks, each
670 with line numbers, a context region, and the actual changes. The line
671 numbers have to be taken with some fuzz, since the other side will
672 likely have inserted or deleted lines earlier in the file. The context
673 region is meant to help find how to adjust the line numbers in order to
674 apply the changes to the right lines. However, if multiple areas of the
675 code have the same surrounding lines of context, the wrong one can be
676 picked. There are real-world cases where this has caused commits to be
677 reapplied incorrectly with no conflicts reported. Setting diff.context
678 to a larger value may prevent such types of problems, but increases the
679 chance of spurious conflicts (since it will require more lines of
680 matching context to apply).
681 The merge backend works with a full copy of each relevant file,
683 insulating it from these types of problems.
684 Labelling of conflicts markers
686 When there are content conflicts, the merge machinery tries to annotate
687 each side’s conflict markers with the commits where the content came
688 from. Since the apply backend drops the original information about the
689 rebased commits and their parents (and instead generates new fake
690 commits based off limited information in the generated patches), those
691 commits cannot be identified; instead it has to fall back to a commit
692 summary. Also, when merge.conflictStyle is set to diff3, the apply
693 backend will use "constructed merge base" to label the content from the
694 merge base, and thus provide no information about the merge base commit
695 whatsoever.
696 The merge backend works with the full commits on both sides of history
698 and thus has no such limitations.
699 Hooks
701 The apply backend has not traditionally called the post-commit hook,
702 while the merge backend has. However, this was by accident of
703 implementation rather than by design. Both backends should have the
704 same behavior, though it is not clear which one is correct.
705 Interruptability
707 The apply backend has safety problems with an ill-timed interrupt; if
708 the user presses Ctrl-C at the wrong time to try to abort the rebase,
709 the rebase can enter a state where it cannot be aborted with a
710 subsequent git rebase --abort. The merge backend does not appear to
711 suffer from the same shortcoming. (See
712 https://lore.kernel.org/git/20200207132152.GC2868@szeder.dev/ for
713 details.)
714 Commit Rewording
716 When a conflict occurs while rebasing, rebase stops and asks the user
717 to resolve. Since the user may need to make notable changes while
718 resolving conflicts, after conflicts are resolved and the user has run
719 git rebase --continue, the rebase should open an editor and ask the
720 user to update the commit message. The merge backend does this, while
721 the apply backend blindly applies the original commit message.
722 Miscellaneous differences
724 There are a few more behavioral differences that most folks would
725 probably consider inconsequential but which are mentioned for
726 completeness:
727 · Reflog: The two backends will use different wording when describing
729 the changes made in the reflog, though both will make use of the
730 word "rebase".
731 · Progress, informational, and error messages: The two backends
733 provide slightly different progress and informational messages.
734 Also, the apply backend writes error messages (such as "Your files
735 would be overwritten...") to stdout, while the merge backend writes
736 them to stderr.
737 · State directories: The two backends keep their state in different
739 directories under .git/
740
742 The merge mechanism (git merge and git pull commands) allows the
743 backend merge strategies to be chosen with -s option. Some strategies
744 can also take their own options, which can be passed by giving
745 -X<option> arguments to git merge and/or git pull.
746 resolve
748 This can only resolve two heads (i.e. the current branch and
749 another branch you pulled from) using a 3-way merge algorithm. It
750 tries to carefully detect criss-cross merge ambiguities and is
751 considered generally safe and fast.
752 recursive
754 This can only resolve two heads using a 3-way merge algorithm. When
755 there is more than one common ancestor that can be used for 3-way
756 merge, it creates a merged tree of the common ancestors and uses
757 that as the reference tree for the 3-way merge. This has been
758 reported to result in fewer merge conflicts without causing
759 mismerges by tests done on actual merge commits taken from Linux
760 2.6 kernel development history. Additionally this can detect and
761 handle merges involving renames, but currently cannot make use of
762 detected copies. This is the default merge strategy when pulling or
763 merging one branch.
764 The recursive strategy can take the following options:
766 ours
768 This option forces conflicting hunks to be auto-resolved
769 cleanly by favoring our version. Changes from the other tree
770 that do not conflict with our side are reflected in the merge
771 result. For a binary file, the entire contents are taken from
772 our side.
773 This should not be confused with the ours merge strategy, which
775 does not even look at what the other tree contains at all. It
776 discards everything the other tree did, declaring our history
777 contains all that happened in it.
778 theirs
780 This is the opposite of ours; note that, unlike ours, there is
781 no theirs merge strategy to confuse this merge option with.
782 patience
784 With this option, merge-recursive spends a little extra time to
785 avoid mismerges that sometimes occur due to unimportant
786 matching lines (e.g., braces from distinct functions). Use this
787 when the branches to be merged have diverged wildly. See also
788 git-diff(1) --patience.
789 diff-algorithm=[patience|minimal|histogram|myers]
791 Tells merge-recursive to use a different diff algorithm, which
792 can help avoid mismerges that occur due to unimportant matching
793 lines (such as braces from distinct functions). See also git-
794 diff(1) --diff-algorithm.
795 ignore-space-change, ignore-all-space, ignore-space-at-eol,
797 ignore-cr-at-eol
798 Treats lines with the indicated type of whitespace change as
799 unchanged for the sake of a three-way merge. Whitespace changes
800 mixed with other changes to a line are not ignored. See also
801 git-diff(1) -b, -w, --ignore-space-at-eol, and
802 --ignore-cr-at-eol.
803 · If their version only introduces whitespace changes to a
805 line, our version is used;
806 · If our version introduces whitespace changes but their
808 version includes a substantial change, their version is
809 used;
810 · Otherwise, the merge proceeds in the usual way.
812 renormalize
814 This runs a virtual check-out and check-in of all three stages
815 of a file when resolving a three-way merge. This option is
816 meant to be used when merging branches with different clean
817 filters or end-of-line normalization rules. See "Merging
818 branches with differing checkin/checkout attributes" in
819 gitattributes(5) for details.
820 no-renormalize
822 Disables the renormalize option. This overrides the
823 merge.renormalize configuration variable.
824 no-renames
826 Turn off rename detection. This overrides the merge.renames
827 configuration variable. See also git-diff(1) --no-renames.
828 find-renames[=<n>]
830 Turn on rename detection, optionally setting the similarity
831 threshold. This is the default. This overrides the
832 merge.renames configuration variable. See also git-diff(1)
833 --find-renames.
834 rename-threshold=<n>
836 Deprecated synonym for find-renames=<n>.
837 subtree[=<path>]
839 This option is a more advanced form of subtree strategy, where
840 the strategy makes a guess on how two trees must be shifted to
841 match with each other when merging. Instead, the specified path
842 is prefixed (or stripped from the beginning) to make the shape
843 of two trees to match.
844 octopus
846 This resolves cases with more than two heads, but refuses to do a
847 complex merge that needs manual resolution. It is primarily meant
848 to be used for bundling topic branch heads together. This is the
849 default merge strategy when pulling or merging more than one
850 branch.
851 ours
853 This resolves any number of heads, but the resulting tree of the
854 merge is always that of the current branch head, effectively
855 ignoring all changes from all other branches. It is meant to be
856 used to supersede old development history of side branches. Note
857 that this is different from the -Xours option to the recursive
858 merge strategy.
859 subtree
861 This is a modified recursive strategy. When merging trees A and B,
862 if B corresponds to a subtree of A, B is first adjusted to match
863 the tree structure of A, instead of reading the trees at the same
864 level. This adjustment is also done to the common ancestor tree.
865 With the strategies that use 3-way merge (including the default,
867 recursive), if a change is made on both branches, but later reverted on
868 one of the branches, that change will be present in the merged result;
869 some people find this behavior confusing. It occurs because only the
870 heads and the merge base are considered when performing a merge, not
871 the individual commits. The merge algorithm therefore considers the
872 reverted change as no change at all, and substitutes the changed
873 version instead.
874
876 You should understand the implications of using git rebase on a
877 repository that you share. See also RECOVERING FROM UPSTREAM REBASE
878 below.
879 When the git-rebase command is run, it will first execute a
881 "pre-rebase" hook if one exists. You can use this hook to do sanity
882 checks and reject the rebase if it isn’t appropriate. Please see the
883 template pre-rebase hook script for an example.
884 Upon completion, <branch> will be the current branch.
886
888 Rebasing interactively means that you have a chance to edit the commits
889 which are rebased. You can reorder the commits, and you can remove them
890 (weeding out bad or otherwise unwanted patches).
891 The interactive mode is meant for this type of workflow:
893 1. have a wonderful idea
895 2. hack on the code
897 3. prepare a series for submission
899 4. submit
901 where point 2. consists of several instances of
903 a) regular use
905 1. finish something worthy of a commit
907 2. commit
909 b) independent fixup
911 1. realize that something does not work
913 2. fix that
915 3. commit it
917 Sometimes the thing fixed in b.2. cannot be amended to the not-quite
919 perfect commit it fixes, because that commit is buried deeply in a
920 patch series. That is exactly what interactive rebase is for: use it
921 after plenty of "a"s and "b"s, by rearranging and editing commits, and
922 squashing multiple commits into one.
923 Start it with the last commit you want to retain as-is:
925 git rebase -i <after-this-commit>
927 An editor will be fired up with all the commits in your current branch
929 (ignoring merge commits), which come after the given commit. You can
930 reorder the commits in this list to your heart’s content, and you can
931 remove them. The list looks more or less like this:
932 pick deadbee The oneline of this commit
934 pick fa1afe1 The oneline of the next commit
935 ...
936 The oneline descriptions are purely for your pleasure; git rebase will
938 not look at them but at the commit names ("deadbee" and "fa1afe1" in
939 this example), so do not delete or edit the names.
940 By replacing the command "pick" with the command "edit", you can tell
942 git rebase to stop after applying that commit, so that you can edit the
943 files and/or the commit message, amend the commit, and continue
944 rebasing.
945 To interrupt the rebase (just like an "edit" command would do, but
947 without cherry-picking any commit first), use the "break" command.
948 If you just want to edit the commit message for a commit, replace the
950 command "pick" with the command "reword".
951 To drop a commit, replace the command "pick" with "drop", or just
953 delete the matching line.
954 If you want to fold two or more commits into one, replace the command
956 "pick" for the second and subsequent commits with "squash" or "fixup".
957 If the commits had different authors, the folded commit will be
958 attributed to the author of the first commit. The suggested commit
959 message for the folded commit is the concatenation of the commit
960 messages of the first commit and of those with the "squash" command,
961 but omits the commit messages of commits with the "fixup" command.
962 git rebase will stop when "pick" has been replaced with "edit" or when
964 a command fails due to merge errors. When you are done editing and/or
965 resolving conflicts you can continue with git rebase --continue.
966 For example, if you want to reorder the last 5 commits, such that what
968 was HEAD~4 becomes the new HEAD. To achieve that, you would call git
969 rebase like this:
970 $ git rebase -i HEAD~5
972 And move the first patch to the end of the list.
974 You might want to recreate merge commits, e.g. if you have a history
976 like this:
977 X
979 \
980 A---M---B
981 /
982 ---o---O---P---Q
983 Suppose you want to rebase the side branch starting at "A" to "Q". Make
985 sure that the current HEAD is "B", and call
986 $ git rebase -i -r --onto Q O
988 Reordering and editing commits usually creates untested intermediate
990 steps. You may want to check that your history editing did not break
991 anything by running a test, or at least recompiling at intermediate
992 points in history by using the "exec" command (shortcut "x"). You may
993 do so by creating a todo list like this one:
994 pick deadbee Implement feature XXX
996 fixup f1a5c00 Fix to feature XXX
997 exec make
998 pick c0ffeee The oneline of the next commit
999 edit deadbab The oneline of the commit after
1000 exec cd subdir; make test
1001 ...
1002 The interactive rebase will stop when a command fails (i.e. exits with
1004 non-0 status) to give you an opportunity to fix the problem. You can
1005 continue with git rebase --continue.
1006 The "exec" command launches the command in a shell (the one specified
1008 in $SHELL, or the default shell if $SHELL is not set), so you can use
1009 shell features (like "cd", ">", ";" ...). The command is run from the
1010 root of the working tree.
1011 $ git rebase -i --exec "make test"
1013 This command lets you check that intermediate commits are compilable.
1015 The todo list becomes like that:
1016 pick 5928aea one
1018 exec make test
1019 pick 04d0fda two
1020 exec make test
1021 pick ba46169 three
1022 exec make test
1023 pick f4593f9 four
1024 exec make test
1025
1027 In interactive mode, you can mark commits with the action "edit".
1028 However, this does not necessarily mean that git rebase expects the
1029 result of this edit to be exactly one commit. Indeed, you can undo the
1030 commit, or you can add other commits. This can be used to split a
1031 commit into two:
1032 · Start an interactive rebase with git rebase -i <commit>^, where
1034 <commit> is the commit you want to split. In fact, any commit range
1035 will do, as long as it contains that commit.
1036 · Mark the commit you want to split with the action "edit".
1038 · When it comes to editing that commit, execute git reset HEAD^. The
1040 effect is that the HEAD is rewound by one, and the index follows
1041 suit. However, the working tree stays the same.
1042 · Now add the changes to the index that you want to have in the first
1044 commit. You can use git add (possibly interactively) or git gui (or
1045 both) to do that.
1046 · Commit the now-current index with whatever commit message is
1048 appropriate now.
1049 · Repeat the last two steps until your working tree is clean.
1051 · Continue the rebase with git rebase --continue.
1053 If you are not absolutely sure that the intermediate revisions are
1055 consistent (they compile, pass the testsuite, etc.) you should use git
1056 stash to stash away the not-yet-committed changes after each commit,
1057 test, and amend the commit if fixes are necessary.
1058
1060 Rebasing (or any other form of rewriting) a branch that others have
1061 based work on is a bad idea: anyone downstream of it is forced to
1062 manually fix their history. This section explains how to do the fix
1063 from the downstream’s point of view. The real fix, however, would be to
1064 avoid rebasing the upstream in the first place.
1065 To illustrate, suppose you are in a situation where someone develops a
1067 subsystem branch, and you are working on a topic that is dependent on
1068 this subsystem. You might end up with a history like the following:
1069 o---o---o---o---o---o---o---o master
1071 \
1072 o---o---o---o---o subsystem
1073 \
1074 *---*---* topic
1075 If subsystem is rebased against master, the following happens:
1077 o---o---o---o---o---o---o---o master
1079 \ \
1080 o---o---o---o---o o'--o'--o'--o'--o' subsystem
1081 \
1082 *---*---* topic
1083 If you now continue development as usual, and eventually merge topic to
1085 subsystem, the commits from subsystem will remain duplicated forever:
1086 o---o---o---o---o---o---o---o master
1088 \ \
1089 o---o---o---o---o o'--o'--o'--o'--o'--M subsystem
1090 \ /
1091 *---*---*-..........-*--* topic
1092 Such duplicates are generally frowned upon because they clutter up
1094 history, making it harder to follow. To clean things up, you need to
1095 transplant the commits on topic to the new subsystem tip, i.e., rebase
1096 topic. This becomes a ripple effect: anyone downstream from topic is
1097 forced to rebase too, and so on!
1098 There are two kinds of fixes, discussed in the following subsections:
1100 Easy case: The changes are literally the same.
1102 This happens if the subsystem rebase was a simple rebase and had no
1103 conflicts.
1104 Hard case: The changes are not the same.
1106 This happens if the subsystem rebase had conflicts, or used
1107 --interactive to omit, edit, squash, or fixup commits; or if the
1108 upstream used one of commit --amend, reset, or a full history
1109 rewriting command like filter-repo[2].
1110 The easy case
1112 Only works if the changes (patch IDs based on the diff contents) on
1113 subsystem are literally the same before and after the rebase subsystem
1114 did.
1115 In that case, the fix is easy because git rebase knows to skip changes
1117 that are already present in the new upstream. So if you say (assuming
1118 you’re on topic)
1119 $ git rebase subsystem
1121 you will end up with the fixed history
1123 o---o---o---o---o---o---o---o master
1125 \
1126 o'--o'--o'--o'--o' subsystem
1127 \
1128 *---*---* topic
1129 The hard case
1131 Things get more complicated if the subsystem changes do not exactly
1132 correspond to the ones before the rebase.
1133 Note
1135 While an "easy case recovery" sometimes appears to be successful
1136 even in the hard case, it may have unintended consequences. For
1137 example, a commit that was removed via git rebase --interactive
1138 will be resurrected!
1139 The idea is to manually tell git rebase "where the old subsystem ended
1141 and your topic began", that is, what the old merge base between them
1142 was. You will have to find a way to name the last commit of the old
1143 subsystem, for example:
1144 · With the subsystem reflog: after git fetch, the old tip of
1146 subsystem is at subsystem@{1}. Subsequent fetches will increase the
1147 number. (See git-reflog(1).)
1148 · Relative to the tip of topic: knowing that your topic has three
1150 commits, the old tip of subsystem must be topic~3.
1151 You can then transplant the old subsystem..topic to the new tip by
1153 saying (for the reflog case, and assuming you are on topic already):
1154 $ git rebase --onto subsystem subsystem@{1}
1156 The ripple effect of a "hard case" recovery is especially bad: everyone
1158 downstream from topic will now have to perform a "hard case" recovery
1159 too!
1160
1162 The interactive rebase command was originally designed to handle
1163 individual patch series. As such, it makes sense to exclude merge
1164 commits from the todo list, as the developer may have merged the
1165 then-current master while working on the branch, only to rebase all the
1166 commits onto master eventually (skipping the merge commits).
1167 However, there are legitimate reasons why a developer may want to
1169 recreate merge commits: to keep the branch structure (or "commit
1170 topology") when working on multiple, inter-related branches.
1171 In the following example, the developer works on a topic branch that
1173 refactors the way buttons are defined, and on another topic branch that
1174 uses that refactoring to implement a "Report a bug" button. The output
1175 of git log --graph --format=%s -5 may look like this:
1176 * Merge branch 'report-a-bug'
1178 |\
1179 | * Add the feedback button
1180 * | Merge branch 'refactor-button'
1181 |\ \
1182 | |/
1183 | * Use the Button class for all buttons
1184 | * Extract a generic Button class from the DownloadButton one
1185 The developer might want to rebase those commits to a newer master
1187 while keeping the branch topology, for example when the first topic
1188 branch is expected to be integrated into master much earlier than the
1189 second one, say, to resolve merge conflicts with changes to the
1190 DownloadButton class that made it into master.
1191 This rebase can be performed using the --rebase-merges option. It will
1193 generate a todo list looking like this:
1194 label onto
1196 # Branch: refactor-button
1198 reset onto
1199 pick 123456 Extract a generic Button class from the DownloadButton one
1200 pick 654321 Use the Button class for all buttons
1201 label refactor-button
1202 # Branch: report-a-bug
1204 reset refactor-button # Use the Button class for all buttons
1205 pick abcdef Add the feedback button
1206 label report-a-bug
1207 reset onto
1209 merge -C a1b2c3 refactor-button # Merge 'refactor-button'
1210 merge -C 6f5e4d report-a-bug # Merge 'report-a-bug'
1211 In contrast to a regular interactive rebase, there are label, reset and
1213 merge commands in addition to pick ones.
1214 The label command associates a label with the current HEAD when that
1216 command is executed. These labels are created as worktree-local refs
1217 (refs/rewritten/<label>) that will be deleted when the rebase finishes.
1218 That way, rebase operations in multiple worktrees linked to the same
1219 repository do not interfere with one another. If the label command
1220 fails, it is rescheduled immediately, with a helpful message how to
1221 proceed.
1222 The reset command resets the HEAD, index and worktree to the specified
1224 revision. It is similar to an exec git reset --hard <label>, but
1225 refuses to overwrite untracked files. If the reset command fails, it is
1226 rescheduled immediately, with a helpful message how to edit the todo
1227 list (this typically happens when a reset command was inserted into the
1228 todo list manually and contains a typo).
1229 The merge command will merge the specified revision(s) into whatever is
1231 HEAD at that time. With -C <original-commit>, the commit message of the
1232 specified merge commit will be used. When the -C is changed to a
1233 lower-case -c, the message will be opened in an editor after a
1234 successful merge so that the user can edit the message.
1235 If a merge command fails for any reason other than merge conflicts
1237 (i.e. when the merge operation did not even start), it is rescheduled
1238 immediately.
1239 At this time, the merge command will always use the recursive merge
1241 strategy for regular merges, and octopus for octopus merges, with no
1242 way to choose a different one. To work around this, an exec command can
1243 be used to call git merge explicitly, using the fact that the labels
1244 are worktree-local refs (the ref refs/rewritten/onto would correspond
1245 to the label onto, for example).
1246 Note: the first command (label onto) labels the revision onto which the
1248 commits are rebased; The name onto is just a convention, as a nod to
1249 the --onto option.
1250 It is also possible to introduce completely new merge commits from
1252 scratch by adding a command of the form merge <merge-head>. This form
1253 will generate a tentative commit message and always open an editor to
1254 let the user edit it. This can be useful e.g. when a topic branch turns
1255 out to address more than a single concern and wants to be split into
1256 two or even more topic branches. Consider this todo list:
1257 pick 192837 Switch from GNU Makefiles to CMake
1259 pick 5a6c7e Document the switch to CMake
1260 pick 918273 Fix detection of OpenSSL in CMake
1261 pick afbecd http: add support for TLS v1.3
1262 pick fdbaec Fix detection of cURL in CMake on Windows
1263 The one commit in this list that is not related to CMake may very well
1265 have been motivated by working on fixing all those bugs introduced by
1266 switching to CMake, but it addresses a different concern. To split this
1267 branch into two topic branches, the todo list could be edited like
1268 this:
1269 label onto
1271 pick afbecd http: add support for TLS v1.3
1273 label tlsv1.3
1274 reset onto
1276 pick 192837 Switch from GNU Makefiles to CMake
1277 pick 918273 Fix detection of OpenSSL in CMake
1278 pick fdbaec Fix detection of cURL in CMake on Windows
1279 pick 5a6c7e Document the switch to CMake
1280 label cmake
1281 reset onto
1283 merge tlsv1.3
1284 merge cmake
1285
1287 The todo list presented by the deprecated --preserve-merges
1288 --interactive does not represent the topology of the revision graph
1289 (use --rebase-merges instead). Editing commits and rewording their
1290 commit messages should work fine, but attempts to reorder commits tend
1291 to produce counterintuitive results. Use --rebase-merges in such
1292 scenarios instead.
1293 For example, an attempt to rearrange
1295 1 --- 2 --- 3 --- 4 --- 5
1297 to
1299 1 --- 2 --- 4 --- 3 --- 5
1301 by moving the "pick 4" line will result in the following history:
1303 3
1305 /
1306 1 --- 2 --- 4 --- 5
1307
1309 Part of the git(1) suite
1310
1312 1. revert-a-faulty-merge How-To
1313 file:///usr/share/doc/git/howto/revert-a-faulty-merge.html
1314 2. filter-repo
1316 https://github.com/newren/git-filter-repo
1317Git 2.26.2 2020-04-20 GIT-REBASE(1)