Merge branch 'mh/release-commit-memory-fix'
[git/git.git] / Documentation / git-checkout.txt
1 git-checkout(1)
2 ===============
5 ----
6 git-checkout - Switch branches or restore working tree files
9 --------
10 [verse]
11 'git checkout' [-q] [-f] [-m] [<branch>]
12 'git checkout' [-q] [-f] [-m] --detach [<branch>]
13 'git checkout' [-q] [-f] [-m] [--detach] <commit>
14 'git checkout' [-q] [-f] [-m] [[-b|-B|--orphan] <new_branch>] [<start_point>]
15 'git checkout' [-f|--ours|--theirs|-m|--conflict=<style>] [<tree-ish>] [--] <paths>...
16 'git checkout' [<tree-ish>] [--] <pathspec>...
17 'git checkout' (-p|--patch) [<tree-ish>] [--] [<paths>...]
20 -----------
21 Updates files in the working tree to match the version in the index
22 or the specified tree. If no paths are given, 'git checkout' will
23 also update `HEAD` to set the specified branch as the current
24 branch.
26 'git checkout' [<branch>]::
27 To prepare for working on `<branch>`, switch to it by updating
28 the index and the files in the working tree, and by pointing
29 `HEAD` at the branch. Local modifications to the files in the
30 working tree are kept, so that they can be committed to the
31 `<branch>`.
32 +
33 If `<branch>` is not found but there does exist a tracking branch in
34 exactly one remote (call it `<remote>`) with a matching name and
35 `--no-guess` is not specified, treat as equivalent to
36 +
37 ------------
38 $ git checkout -b <branch> --track <remote>/<branch>
39 ------------
40 +
41 You could omit `<branch>`, in which case the command degenerates to
42 "check out the current branch", which is a glorified no-op with
43 rather expensive side-effects to show only the tracking information,
44 if exists, for the current branch.
46 'git checkout' -b|-B <new_branch> [<start point>]::
48 Specifying `-b` causes a new branch to be created as if
49 linkgit:git-branch[1] were called and then checked out. In
50 this case you can use the `--track` or `--no-track` options,
51 which will be passed to 'git branch'. As a convenience,
52 `--track` without `-b` implies branch creation; see the
53 description of `--track` below.
54 +
55 If `-B` is given, `<new_branch>` is created if it doesn't exist; otherwise, it
56 is reset. This is the transactional equivalent of
57 +
58 ------------
59 $ git branch -f <branch> [<start point>]
60 $ git checkout <branch>
61 ------------
62 +
63 that is to say, the branch is not reset/created unless "git checkout" is
64 successful.
66 'git checkout' --detach [<branch>]::
67 'git checkout' [--detach] <commit>::
69 Prepare to work on top of `<commit>`, by detaching `HEAD` at it
70 (see "DETACHED HEAD" section), and updating the index and the
71 files in the working tree. Local modifications to the files
72 in the working tree are kept, so that the resulting working
73 tree will be the state recorded in the commit plus the local
74 modifications.
75 +
76 When the `<commit>` argument is a branch name, the `--detach` option can
77 be used to detach `HEAD` at the tip of the branch (`git checkout
78 <branch>` would check out that branch without detaching `HEAD`).
79 +
80 Omitting `<branch>` detaches `HEAD` at the tip of the current branch.
82 'git checkout' [<tree-ish>] [--] <pathspec>...::
84 Overwrite paths in the working tree by replacing with the
85 contents in the index or in the `<tree-ish>` (most often a
86 commit). When a `<tree-ish>` is given, the paths that
87 match the `<pathspec>` are updated both in the index and in
88 the working tree.
89 +
90 The index may contain unmerged entries because of a previous failed merge.
91 By default, if you try to check out such an entry from the index, the
92 checkout operation will fail and nothing will be checked out.
93 Using `-f` will ignore these unmerged entries. The contents from a
94 specific side of the merge can be checked out of the index by
95 using `--ours` or `--theirs`. With `-m`, changes made to the working tree
96 file can be discarded to re-create the original conflicted merge result.
98 'git checkout' (-p|--patch) [<tree-ish>] [--] [<pathspec>...]::
99 This is similar to the "check out paths to the working tree
100 from either the index or from a tree-ish" mode described
101 above, but lets you use the interactive interface to show
102 the "diff" output and choose which hunks to use in the
103 result. See below for the description of `--patch` option.
107 -------
108 -q::
109 --quiet::
110 Quiet, suppress feedback messages.
112 --progress::
113 --no-progress::
114 Progress status is reported on the standard error stream
115 by default when it is attached to a terminal, unless `--quiet`
116 is specified. This flag enables progress reporting even if not
117 attached to a terminal, regardless of `--quiet`.
119 -f::
120 --force::
121 When switching branches, proceed even if the index or the
122 working tree differs from `HEAD`. This is used to throw away
123 local changes.
124 +
125 When checking out paths from the index, do not fail upon unmerged
126 entries; instead, unmerged entries are ignored.
128 --ours::
129 --theirs::
130 When checking out paths from the index, check out stage #2
131 ('ours') or #3 ('theirs') for unmerged paths.
132 +
133 Note that during `git rebase` and `git pull --rebase`, 'ours' and
134 'theirs' may appear swapped; `--ours` gives the version from the
135 branch the changes are rebased onto, while `--theirs` gives the
136 version from the branch that holds your work that is being rebased.
137 +
138 This is because `rebase` is used in a workflow that treats the
139 history at the remote as the shared canonical one, and treats the
140 work done on the branch you are rebasing as the third-party work to
141 be integrated, and you are temporarily assuming the role of the
142 keeper of the canonical history during the rebase. As the keeper of
143 the canonical history, you need to view the history from the remote
144 as `ours` (i.e. "our shared canonical history"), while what you did
145 on your side branch as `theirs` (i.e. "one contributor's work on top
146 of it").
148 -b <new_branch>::
149 Create a new branch named `<new_branch>` and start it at
150 `<start_point>`; see linkgit:git-branch[1] for details.
152 -B <new_branch>::
153 Creates the branch `<new_branch>` and start it at `<start_point>`;
154 if it already exists, then reset it to `<start_point>`. This is
155 equivalent to running "git branch" with "-f"; see
156 linkgit:git-branch[1] for details.
158 -t::
159 --track::
160 When creating a new branch, set up "upstream" configuration. See
161 "--track" in linkgit:git-branch[1] for details.
162 +
163 If no `-b` option is given, the name of the new branch will be
164 derived from the remote-tracking branch, by looking at the local part of
165 the refspec configured for the corresponding remote, and then stripping
166 the initial part up to the "*".
167 This would tell us to use `hack` as the local branch when branching
168 off of `origin/hack` (or `remotes/origin/hack`, or even
169 `refs/remotes/origin/hack`). If the given name has no slash, or the above
170 guessing results in an empty name, the guessing is aborted. You can
171 explicitly give a name with `-b` in such a case.
173 --no-track::
174 Do not set up "upstream" configuration, even if the
175 `branch.autoSetupMerge` configuration variable is true.
177 --guess::
178 --no-guess::
179 If `<branch>` is not found but there does exist a tracking
180 branch in exactly one remote (call it `<remote>`) with a
181 matching name, treat as equivalent to
182 +
183 ------------
184 $ git checkout -b <branch> --track <remote>/<branch>
185 ------------
186 +
187 If the branch exists in multiple remotes and one of them is named by
188 the `checkout.defaultRemote` configuration variable, we'll use that
189 one for the purposes of disambiguation, even if the `<branch>` isn't
190 unique across all remotes. Set it to
191 e.g. `checkout.defaultRemote=origin` to always checkout remote
192 branches from there if `<branch>` is ambiguous but exists on the
193 'origin' remote. See also `checkout.defaultRemote` in
194 linkgit:git-config[1].
195 +
196 Use `--no-guess` to disable this.
198 -l::
199 Create the new branch's reflog; see linkgit:git-branch[1] for
200 details.
202 --detach::
203 Rather than checking out a branch to work on it, check out a
204 commit for inspection and discardable experiments.
205 This is the default behavior of `git checkout <commit>` when
206 `<commit>` is not a branch name. See the "DETACHED HEAD" section
207 below for details.
209 --orphan <new_branch>::
210 Create a new 'orphan' branch, named `<new_branch>`, started from
211 `<start_point>` and switch to it. The first commit made on this
212 new branch will have no parents and it will be the root of a new
213 history totally disconnected from all the other branches and
214 commits.
215 +
216 The index and the working tree are adjusted as if you had previously run
217 `git checkout <start_point>`. This allows you to start a new history
218 that records a set of paths similar to `<start_point>` by easily running
219 `git commit -a` to make the root commit.
220 +
221 This can be useful when you want to publish the tree from a commit
222 without exposing its full history. You might want to do this to publish
223 an open source branch of a project whose current tree is "clean", but
224 whose full history contains proprietary or otherwise encumbered bits of
225 code.
226 +
227 If you want to start a disconnected history that records a set of paths
228 that is totally different from the one of `<start_point>`, then you should
229 clear the index and the working tree right after creating the orphan
230 branch by running `git rm -rf .` from the top level of the working tree.
231 Afterwards you will be ready to prepare your new files, repopulating the
232 working tree, by copying them from elsewhere, extracting a tarball, etc.
234 --ignore-skip-worktree-bits::
235 In sparse checkout mode, `git checkout -- <paths>` would
236 update only entries matched by `<paths>` and sparse patterns
237 in `$GIT_DIR/info/sparse-checkout`. This option ignores
238 the sparse patterns and adds back any files in `<paths>`.
240 -m::
241 --merge::
242 When switching branches,
243 if you have local modifications to one or more files that
244 are different between the current branch and the branch to
245 which you are switching, the command refuses to switch
246 branches in order to preserve your modifications in context.
247 However, with this option, a three-way merge between the current
248 branch, your working tree contents, and the new branch
249 is done, and you will be on the new branch.
250 +
251 When a merge conflict happens, the index entries for conflicting
252 paths are left unmerged, and you need to resolve the conflicts
253 and mark the resolved paths with `git add` (or `git rm` if the merge
254 should result in deletion of the path).
255 +
256 When checking out paths from the index, this option lets you recreate
257 the conflicted merge in the specified paths.
258 +
259 When switching branches with `--merge`, staged changes may be lost.
261 --conflict=<style>::
262 The same as `--merge` option above, but changes the way the
263 conflicting hunks are presented, overriding the
264 `merge.conflictStyle` configuration variable. Possible values are
265 "merge" (default) and "diff3" (in addition to what is shown by
266 "merge" style, shows the original contents).
268 -p::
269 --patch::
270 Interactively select hunks in the difference between the
271 `<tree-ish>` (or the index, if unspecified) and the working
272 tree. The chosen hunks are then applied in reverse to the
273 working tree (and if a `<tree-ish>` was specified, the index).
274 +
275 This means that you can use `git checkout -p` to selectively discard
276 edits from your current working tree. See the ``Interactive Mode''
277 section of linkgit:git-add[1] to learn how to operate the `--patch` mode.
278 +
279 Note that this option uses the no overlay mode by default (see also
280 `--overlay`), and currently doesn't support overlay mode.
282 --ignore-other-worktrees::
283 `git checkout` refuses when the wanted ref is already checked
284 out by another worktree. This option makes it check the ref
285 out anyway. In other words, the ref can be held by more than one
286 worktree.
288 --overwrite-ignore::
289 --no-overwrite-ignore::
290 Silently overwrite ignored files when switching branches. This
291 is the default behavior. Use `--no-overwrite-ignore` to abort
292 the operation when the new branch contains ignored files.
294 --recurse-submodules::
295 --no-recurse-submodules::
296 Using `--recurse-submodules` will update the content of all initialized
297 submodules according to the commit recorded in the superproject. If
298 local modifications in a submodule would be overwritten the checkout
299 will fail unless `-f` is used. If nothing (or `--no-recurse-submodules`)
300 is used, the work trees of submodules will not be updated.
301 Just like linkgit:git-submodule[1], this will detach `HEAD` of the
302 submodule.
304 --overlay::
305 --no-overlay::
306 In the default overlay mode, `git checkout` never
307 removes files from the index or the working tree. When
308 specifying `--no-overlay`, files that appear in the index and
309 working tree, but not in `<tree-ish>` are removed, to make them
310 match `<tree-ish>` exactly.
312 <branch>::
313 Branch to checkout; if it refers to a branch (i.e., a name that,
314 when prepended with "refs/heads/", is a valid ref), then that
315 branch is checked out. Otherwise, if it refers to a valid
316 commit, your `HEAD` becomes "detached" and you are no longer on
317 any branch (see below for details).
318 +
319 You can use the `@{-N}` syntax to refer to the N-th last
320 branch/commit checked out using "git checkout" operation. You may
321 also specify `-` which is synonymous to `@{-1}`.
322 +
323 As a special case, you may use `A...B` as a shortcut for the
324 merge base of `A` and `B` if there is exactly one merge base. You can
325 leave out at most one of `A` and `B`, in which case it defaults to `HEAD`.
327 <new_branch>::
328 Name for the new branch.
330 <start_point>::
331 The name of a commit at which to start the new branch; see
332 linkgit:git-branch[1] for details. Defaults to `HEAD`.
333 +
334 As a special case, you may use `"A...B"` as a shortcut for the
335 merge base of `A` and `B` if there is exactly one merge base. You can
336 leave out at most one of `A` and `B`, in which case it defaults to `HEAD`.
338 <tree-ish>::
339 Tree to checkout from (when paths are given). If not specified,
340 the index will be used.
345 -------------
346 `HEAD` normally refers to a named branch (e.g. `master`). Meanwhile, each
347 branch refers to a specific commit. Let's look at a repo with three
348 commits, one of them tagged, and with branch `master` checked out:
350 ------------
351 HEAD (refers to branch 'master')
352 |
353 v
354 a---b---c branch 'master' (refers to commit 'c')
355 ^
356 |
357 tag 'v2.0' (refers to commit 'b')
358 ------------
360 When a commit is created in this state, the branch is updated to refer to
361 the new commit. Specifically, 'git commit' creates a new commit `d`, whose
362 parent is commit `c`, and then updates branch `master` to refer to new
363 commit `d`. `HEAD` still refers to branch `master` and so indirectly now refers
364 to commit `d`:
366 ------------
367 $ edit; git add; git commit
369 HEAD (refers to branch 'master')
370 |
371 v
372 a---b---c---d branch 'master' (refers to commit 'd')
373 ^
374 |
375 tag 'v2.0' (refers to commit 'b')
376 ------------
378 It is sometimes useful to be able to checkout a commit that is not at
379 the tip of any named branch, or even to create a new commit that is not
380 referenced by a named branch. Let's look at what happens when we
381 checkout commit `b` (here we show two ways this may be done):
383 ------------
384 $ git checkout v2.0 # or
385 $ git checkout master^^
387 HEAD (refers to commit 'b')
388 |
389 v
390 a---b---c---d branch 'master' (refers to commit 'd')
391 ^
392 |
393 tag 'v2.0' (refers to commit 'b')
394 ------------
396 Notice that regardless of which checkout command we use, `HEAD` now refers
397 directly to commit `b`. This is known as being in detached `HEAD` state.
398 It means simply that `HEAD` refers to a specific commit, as opposed to
399 referring to a named branch. Let's see what happens when we create a commit:
401 ------------
402 $ edit; git add; git commit
404 HEAD (refers to commit 'e')
405 |
406 v
407 e
408 /
409 a---b---c---d branch 'master' (refers to commit 'd')
410 ^
411 |
412 tag 'v2.0' (refers to commit 'b')
413 ------------
415 There is now a new commit `e`, but it is referenced only by `HEAD`. We can
416 of course add yet another commit in this state:
418 ------------
419 $ edit; git add; git commit
421 HEAD (refers to commit 'f')
422 |
423 v
424 e---f
425 /
426 a---b---c---d branch 'master' (refers to commit 'd')
427 ^
428 |
429 tag 'v2.0' (refers to commit 'b')
430 ------------
432 In fact, we can perform all the normal Git operations. But, let's look
433 at what happens when we then checkout `master`:
435 ------------
436 $ git checkout master
438 HEAD (refers to branch 'master')
439 e---f |
440 / v
441 a---b---c---d branch 'master' (refers to commit 'd')
442 ^
443 |
444 tag 'v2.0' (refers to commit 'b')
445 ------------
447 It is important to realize that at this point nothing refers to commit
448 `f`. Eventually commit `f` (and by extension commit `e`) will be deleted
449 by the routine Git garbage collection process, unless we create a reference
450 before that happens. If we have not yet moved away from commit `f`,
451 any of these will create a reference to it:
453 ------------
454 $ git checkout -b foo <1>
455 $ git branch foo <2>
456 $ git tag foo <3>
457 ------------
459 <1> creates a new branch `foo`, which refers to commit `f`, and then
460 updates `HEAD` to refer to branch `foo`. In other words, we'll no longer
461 be in detached `HEAD` state after this command.
463 <2> similarly creates a new branch `foo`, which refers to commit `f`,
464 but leaves `HEAD` detached.
466 <3> creates a new tag `foo`, which refers to commit `f`,
467 leaving `HEAD` detached.
469 If we have moved away from commit `f`, then we must first recover its object
470 name (typically by using git reflog), and then we can create a reference to
471 it. For example, to see the last two commits to which `HEAD` referred, we
472 can use either of these commands:
474 ------------
475 $ git reflog -2 HEAD # or
476 $ git log -g -2 HEAD
477 ------------
480 -----------------------
482 When there is only one argument given and it is not `--` (e.g. `git
483 checkout abc`), and when the argument is both a valid `<tree-ish>`
484 (e.g. a branch `abc` exists) and a valid `<pathspec>` (e.g. a file
485 or a directory whose name is "abc" exists), Git would usually ask
486 you to disambiguate. Because checking out a branch is so common an
487 operation, however, `git checkout abc` takes "abc" as a `<tree-ish>`
488 in such a situation. Use `git checkout -- <pathspec>` if you want
489 to checkout these paths out of the index.
492 --------
494 . The following sequence checks out the `master` branch, reverts
495 the `Makefile` to two revisions back, deletes `hello.c` by
496 mistake, and gets it back from the index.
497 +
498 ------------
499 $ git checkout master <1>
500 $ git checkout master~2 Makefile <2>
501 $ rm -f hello.c
502 $ git checkout hello.c <3>
503 ------------
504 +
505 <1> switch branch
506 <2> take a file out of another commit
507 <3> restore `hello.c` from the index
508 +
509 If you want to check out _all_ C source files out of the index,
510 you can say
511 +
512 ------------
513 $ git checkout -- '*.c'
514 ------------
515 +
516 Note the quotes around `*.c`. The file `hello.c` will also be
517 checked out, even though it is no longer in the working tree,
518 because the file globbing is used to match entries in the index
519 (not in the working tree by the shell).
520 +
521 If you have an unfortunate branch that is named `hello.c`, this
522 step would be confused as an instruction to switch to that branch.
523 You should instead write:
524 +
525 ------------
526 $ git checkout -- hello.c
527 ------------
529 . After working in the wrong branch, switching to the correct
530 branch would be done using:
531 +
532 ------------
533 $ git checkout mytopic
534 ------------
535 +
536 However, your "wrong" branch and correct `mytopic` branch may
537 differ in files that you have modified locally, in which case
538 the above checkout would fail like this:
539 +
540 ------------
541 $ git checkout mytopic
542 error: You have local changes to 'frotz'; not switching branches.
543 ------------
544 +
545 You can give the `-m` flag to the command, which would try a
546 three-way merge:
547 +
548 ------------
549 $ git checkout -m mytopic
550 Auto-merging frotz
551 ------------
552 +
553 After this three-way merge, the local modifications are _not_
554 registered in your index file, so `git diff` would show you what
555 changes you made since the tip of the new branch.
557 . When a merge conflict happens during switching branches with
558 the `-m` option, you would see something like this:
559 +
560 ------------
561 $ git checkout -m mytopic
562 Auto-merging frotz
563 ERROR: Merge conflict in frotz
564 fatal: merge program failed
565 ------------
566 +
567 At this point, `git diff` shows the changes cleanly merged as in
568 the previous example, as well as the changes in the conflicted
569 files. Edit and resolve the conflict and mark it resolved with
570 `git add` as usual:
571 +
572 ------------
573 $ edit frotz
574 $ git add frotz
575 ------------
578 --------
579 linkgit:git-switch[1],
580 linkgit:git-restore[1]
582 GIT
583 ---
584 Part of the linkgit:git[1] suite