1GIT-CHECKOUT(1) Git Manual GIT-CHECKOUT(1)
2
6 git-checkout - Switch branches or restore working tree files
7
9 git checkout [-q] [-f] [-m] [<branch>]
10 git checkout [-q] [-f] [-m] --detach [<branch>]
11 git checkout [-q] [-f] [-m] [--detach] <commit>
12 git checkout [-q] [-f] [-m] [[-b|-B|--orphan] <new_branch>] [<start_point>]
13 git checkout [-f|--ours|--theirs|-m|--conflict=<style>] [<tree-ish>] [--] <paths>...
14 git checkout [<tree-ish>] [--] <pathspec>...
15 git checkout (-p|--patch) [<tree-ish>] [--] [<paths>...]
16
19 Updates files in the working tree to match the version in the index or
20 the specified tree. If no paths are given, git checkout will also
21 update HEAD to set the specified branch as the current branch.
22 git checkout [<branch>]
24 To prepare for working on <branch>, switch to it by updating the
25 index and the files in the working tree, and by pointing HEAD at
26 the branch. Local modifications to the files in the working tree
27 are kept, so that they can be committed to the <branch>.
28 If <branch> is not found but there does exist a tracking branch in
30 exactly one remote (call it <remote>) with a matching name and
31 --no-guess is not specified, treat as equivalent to
32 $ git checkout -b <branch> --track <remote>/<branch>
34 You could omit <branch>, in which case the command degenerates to
36 "check out the current branch", which is a glorified no-op with
37 rather expensive side-effects to show only the tracking
38 information, if exists, for the current branch.
39 git checkout -b|-B <new_branch> [<start point>]
41 Specifying -b causes a new branch to be created as if git-branch(1)
42 were called and then checked out. In this case you can use the
43 --track or --no-track options, which will be passed to git branch.
44 As a convenience, --track without -b implies branch creation; see
45 the description of --track below.
46 If -B is given, <new_branch> is created if it doesn’t exist;
48 otherwise, it is reset. This is the transactional equivalent of
49 $ git branch -f <branch> [<start point>]
51 $ git checkout <branch>
52 that is to say, the branch is not reset/created unless "git
54 checkout" is successful.
55 git checkout --detach [<branch>], git checkout [--detach] <commit>
57 Prepare to work on top of <commit>, by detaching HEAD at it (see
58 "DETACHED HEAD" section), and updating the index and the files in
59 the working tree. Local modifications to the files in the working
60 tree are kept, so that the resulting working tree will be the state
61 recorded in the commit plus the local modifications.
62 When the <commit> argument is a branch name, the --detach option
64 can be used to detach HEAD at the tip of the branch (git checkout
65 <branch> would check out that branch without detaching HEAD).
66 Omitting <branch> detaches HEAD at the tip of the current branch.
68 git checkout [<tree-ish>] [--] <pathspec>...
70 Overwrite paths in the working tree by replacing with the contents
71 in the index or in the <tree-ish> (most often a commit). When a
72 <tree-ish> is given, the paths that match the <pathspec> are
73 updated both in the index and in the working tree.
74 The index may contain unmerged entries because of a previous failed
76 merge. By default, if you try to check out such an entry from the
77 index, the checkout operation will fail and nothing will be checked
78 out. Using -f will ignore these unmerged entries. The contents from
79 a specific side of the merge can be checked out of the index by
80 using --ours or --theirs. With -m, changes made to the working tree
81 file can be discarded to re-create the original conflicted merge
82 result.
83 git checkout (-p|--patch) [<tree-ish>] [--] [<pathspec>...]
85 This is similar to the "check out paths to the working tree from
86 either the index or from a tree-ish" mode described above, but lets
87 you use the interactive interface to show the "diff" output and
88 choose which hunks to use in the result. See below for the
89 description of --patch option.
90
92 -q, --quiet
93 Quiet, suppress feedback messages.
94 --progress, --no-progress
96 Progress status is reported on the standard error stream by default
97 when it is attached to a terminal, unless --quiet is specified.
98 This flag enables progress reporting even if not attached to a
99 terminal, regardless of --quiet.
100 -f, --force
102 When switching branches, proceed even if the index or the working
103 tree differs from HEAD. This is used to throw away local changes.
104 When checking out paths from the index, do not fail upon unmerged
106 entries; instead, unmerged entries are ignored.
107 --ours, --theirs
109 When checking out paths from the index, check out stage #2 (ours)
110 or #3 (theirs) for unmerged paths.
111 Note that during git rebase and git pull --rebase, ours and theirs
113 may appear swapped; --ours gives the version from the branch the
114 changes are rebased onto, while --theirs gives the version from the
115 branch that holds your work that is being rebased.
116 This is because rebase is used in a workflow that treats the
118 history at the remote as the shared canonical one, and treats the
119 work done on the branch you are rebasing as the third-party work to
120 be integrated, and you are temporarily assuming the role of the
121 keeper of the canonical history during the rebase. As the keeper of
122 the canonical history, you need to view the history from the remote
123 as ours (i.e. "our shared canonical history"), while what you did
124 on your side branch as theirs (i.e. "one contributor’s work on top
125 of it").
126 -b <new_branch>
128 Create a new branch named <new_branch> and start it at
129 <start_point>; see git-branch(1) for details.
130 -B <new_branch>
132 Creates the branch <new_branch> and start it at <start_point>; if
133 it already exists, then reset it to <start_point>. This is
134 equivalent to running "git branch" with "-f"; see git-branch(1) for
135 details.
136 -t, --track
138 When creating a new branch, set up "upstream" configuration. See
139 "--track" in git-branch(1) for details.
140 If no -b option is given, the name of the new branch will be
142 derived from the remote-tracking branch, by looking at the local
143 part of the refspec configured for the corresponding remote, and
144 then stripping the initial part up to the "*". This would tell us
145 to use hack as the local branch when branching off of origin/hack
146 (or remotes/origin/hack, or even refs/remotes/origin/hack). If the
147 given name has no slash, or the above guessing results in an empty
148 name, the guessing is aborted. You can explicitly give a name with
149 -b in such a case.
150 --no-track
152 Do not set up "upstream" configuration, even if the
153 branch.autoSetupMerge configuration variable is true.
154 --guess, --no-guess
156 If <branch> is not found but there does exist a tracking branch in
157 exactly one remote (call it <remote>) with a matching name, treat
158 as equivalent to
159 $ git checkout -b <branch> --track <remote>/<branch>
161 If the branch exists in multiple remotes and one of them is named
163 by the checkout.defaultRemote configuration variable, we’ll use
164 that one for the purposes of disambiguation, even if the <branch>
165 isn’t unique across all remotes. Set it to e.g.
166 checkout.defaultRemote=origin to always checkout remote branches
167 from there if <branch> is ambiguous but exists on the origin
168 remote. See also checkout.defaultRemote in git-config(1).
169 Use --no-guess to disable this.
171 -l
173 Create the new branch’s reflog; see git-branch(1) for details.
174 --detach
176 Rather than checking out a branch to work on it, check out a commit
177 for inspection and discardable experiments. This is the default
178 behavior of git checkout <commit> when <commit> is not a branch
179 name. See the "DETACHED HEAD" section below for details.
180 --orphan <new_branch>
182 Create a new orphan branch, named <new_branch>, started from
183 <start_point> and switch to it. The first commit made on this new
184 branch will have no parents and it will be the root of a new
185 history totally disconnected from all the other branches and
186 commits.
187 The index and the working tree are adjusted as if you had
189 previously run git checkout <start_point>. This allows you to start
190 a new history that records a set of paths similar to <start_point>
191 by easily running git commit -a to make the root commit.
192 This can be useful when you want to publish the tree from a commit
194 without exposing its full history. You might want to do this to
195 publish an open source branch of a project whose current tree is
196 "clean", but whose full history contains proprietary or otherwise
197 encumbered bits of code.
198 If you want to start a disconnected history that records a set of
200 paths that is totally different from the one of <start_point>, then
201 you should clear the index and the working tree right after
202 creating the orphan branch by running git rm -rf . from the top
203 level of the working tree. Afterwards you will be ready to prepare
204 your new files, repopulating the working tree, by copying them from
205 elsewhere, extracting a tarball, etc.
206 --ignore-skip-worktree-bits
208 In sparse checkout mode, git checkout -- <paths> would update only
209 entries matched by <paths> and sparse patterns in
210 $GIT_DIR/info/sparse-checkout. This option ignores the sparse
211 patterns and adds back any files in <paths>.
212 -m, --merge
214 When switching branches, if you have local modifications to one or
215 more files that are different between the current branch and the
216 branch to which you are switching, the command refuses to switch
217 branches in order to preserve your modifications in context.
218 However, with this option, a three-way merge between the current
219 branch, your working tree contents, and the new branch is done, and
220 you will be on the new branch.
221 When a merge conflict happens, the index entries for conflicting
223 paths are left unmerged, and you need to resolve the conflicts and
224 mark the resolved paths with git add (or git rm if the merge should
225 result in deletion of the path).
226 When checking out paths from the index, this option lets you
228 recreate the conflicted merge in the specified paths.
229 When switching branches with --merge, staged changes may be lost.
231 --conflict=<style>
233 The same as --merge option above, but changes the way the
234 conflicting hunks are presented, overriding the merge.conflictStyle
235 configuration variable. Possible values are "merge" (default) and
236 "diff3" (in addition to what is shown by "merge" style, shows the
237 original contents).
238 -p, --patch
240 Interactively select hunks in the difference between the <tree-ish>
241 (or the index, if unspecified) and the working tree. The chosen
242 hunks are then applied in reverse to the working tree (and if a
243 <tree-ish> was specified, the index).
244 This means that you can use git checkout -p to selectively discard
246 edits from your current working tree. See the “Interactive Mode”
247 section of git-add(1) to learn how to operate the --patch mode.
248 Note that this option uses the no overlay mode by default (see also
250 --overlay), and currently doesn’t support overlay mode.
251 --ignore-other-worktrees
253 git checkout refuses when the wanted ref is already checked out by
254 another worktree. This option makes it check the ref out anyway. In
255 other words, the ref can be held by more than one worktree.
256 --overwrite-ignore, --no-overwrite-ignore
258 Silently overwrite ignored files when switching branches. This is
259 the default behavior. Use --no-overwrite-ignore to abort the
260 operation when the new branch contains ignored files.
261 --recurse-submodules, --no-recurse-submodules
263 Using --recurse-submodules will update the content of all
264 initialized submodules according to the commit recorded in the
265 superproject. If local modifications in a submodule would be
266 overwritten the checkout will fail unless -f is used. If nothing
267 (or --no-recurse-submodules) is used, the work trees of submodules
268 will not be updated. Just like git-submodule(1), this will detach
269 HEAD of the submodule.
270 --overlay, --no-overlay
272 In the default overlay mode, git checkout never removes files from
273 the index or the working tree. When specifying --no-overlay, files
274 that appear in the index and working tree, but not in <tree-ish>
275 are removed, to make them match <tree-ish> exactly.
276 <branch>
278 Branch to checkout; if it refers to a branch (i.e., a name that,
279 when prepended with "refs/heads/", is a valid ref), then that
280 branch is checked out. Otherwise, if it refers to a valid commit,
281 your HEAD becomes "detached" and you are no longer on any branch
282 (see below for details).
283 You can use the @{-N} syntax to refer to the N-th last
285 branch/commit checked out using "git checkout" operation. You may
286 also specify - which is synonymous to @{-1}.
287 As a special case, you may use A...B as a shortcut for the merge
289 base of A and B if there is exactly one merge base. You can leave
290 out at most one of A and B, in which case it defaults to HEAD.
291 <new_branch>
293 Name for the new branch.
294 <start_point>
296 The name of a commit at which to start the new branch; see git-
297 branch(1) for details. Defaults to HEAD.
298 As a special case, you may use "A...B" as a shortcut for the merge
300 base of A and B if there is exactly one merge base. You can leave
301 out at most one of A and B, in which case it defaults to HEAD.
302 <tree-ish>
304 Tree to checkout from (when paths are given). If not specified, the
305 index will be used.
306
308 HEAD normally refers to a named branch (e.g. master). Meanwhile, each
309 branch refers to a specific commit. Let’s look at a repo with three
310 commits, one of them tagged, and with branch master checked out:
311 HEAD (refers to branch 'master')
313 |
314 v
315 a---b---c branch 'master' (refers to commit 'c')
316 ^
317 |
318 tag 'v2.0' (refers to commit 'b')
319 When a commit is created in this state, the branch is updated to refer
322 to the new commit. Specifically, git commit creates a new commit d,
323 whose parent is commit c, and then updates branch master to refer to
324 new commit d. HEAD still refers to branch master and so indirectly now
325 refers to commit d:
326 $ edit; git add; git commit
328 HEAD (refers to branch 'master')
330 |
331 v
332 a---b---c---d branch 'master' (refers to commit 'd')
333 ^
334 |
335 tag 'v2.0' (refers to commit 'b')
336 It is sometimes useful to be able to checkout a commit that is not at
339 the tip of any named branch, or even to create a new commit that is not
340 referenced by a named branch. Let’s look at what happens when we
341 checkout commit b (here we show two ways this may be done):
342 $ git checkout v2.0 # or
344 $ git checkout master^^
345 HEAD (refers to commit 'b')
347 |
348 v
349 a---b---c---d branch 'master' (refers to commit 'd')
350 ^
351 |
352 tag 'v2.0' (refers to commit 'b')
353 Notice that regardless of which checkout command we use, HEAD now
356 refers directly to commit b. This is known as being in detached HEAD
357 state. It means simply that HEAD refers to a specific commit, as
358 opposed to referring to a named branch. Let’s see what happens when we
359 create a commit:
360 $ edit; git add; git commit
362 HEAD (refers to commit 'e')
364 |
365 v
366 e
367 /
368 a---b---c---d branch 'master' (refers to commit 'd')
369 ^
370 |
371 tag 'v2.0' (refers to commit 'b')
372 There is now a new commit e, but it is referenced only by HEAD. We can
375 of course add yet another commit in this state:
376 $ edit; git add; git commit
378 HEAD (refers to commit 'f')
380 |
381 v
382 e---f
383 /
384 a---b---c---d branch 'master' (refers to commit 'd')
385 ^
386 |
387 tag 'v2.0' (refers to commit 'b')
388 In fact, we can perform all the normal Git operations. But, let’s look
391 at what happens when we then checkout master:
392 $ git checkout master
394 HEAD (refers to branch 'master')
396 e---f |
397 / v
398 a---b---c---d branch 'master' (refers to commit 'd')
399 ^
400 |
401 tag 'v2.0' (refers to commit 'b')
402 It is important to realize that at this point nothing refers to commit
405 f. Eventually commit f (and by extension commit e) will be deleted by
406 the routine Git garbage collection process, unless we create a
407 reference before that happens. If we have not yet moved away from
408 commit f, any of these will create a reference to it:
409 $ git checkout -b foo [1m(1)
411 $ git branch foo [1m(2)
412 $ git tag foo [1m(3)
413 1. creates a new branch foo, which refers to commit f, and then updates
416 HEAD to refer to branch foo. In other words, we’ll no longer be in
417 detached HEAD state after this command.
418 2. similarly creates a new branch foo, which refers to commit f, but
419 leaves HEAD detached.
420 3. creates a new tag foo, which refers to commit f, leaving HEAD
421 detached.
422 If we have moved away from commit f, then we must first recover its
424 object name (typically by using git reflog), and then we can create a
425 reference to it. For example, to see the last two commits to which HEAD
426 referred, we can use either of these commands:
427 $ git reflog -2 HEAD # or
429 $ git log -g -2 HEAD
430
433 When there is only one argument given and it is not -- (e.g. git
434 checkout abc), and when the argument is both a valid <tree-ish> (e.g. a
435 branch abc exists) and a valid <pathspec> (e.g. a file or a directory
436 whose name is "abc" exists), Git would usually ask you to disambiguate.
437 Because checking out a branch is so common an operation, however, git
438 checkout abc takes "abc" as a <tree-ish> in such a situation. Use git
439 checkout -- <pathspec> if you want to checkout these paths out of the
440 index.
441
443 1. The following sequence checks out the master branch, reverts the
444 Makefile to two revisions back, deletes hello.c by mistake, and
445 gets it back from the index.
446 $ git checkout master [1m(1)
448 $ git checkout master~2 Makefile [1m(2)
449 $ rm -f hello.c
450 $ git checkout hello.c [1m(3)
451 1. switch branch
453 2. take a file out of another commit
454 3. restore hello.c from the index
455 If you want to check out all C source files out of the index, you
457 can say
458 $ git checkout -- '*.c'
460 Note the quotes around *.c. The file hello.c will also be checked
462 out, even though it is no longer in the working tree, because the
463 file globbing is used to match entries in the index (not in the
464 working tree by the shell).
465 If you have an unfortunate branch that is named hello.c, this step
467 would be confused as an instruction to switch to that branch. You
468 should instead write:
469 $ git checkout -- hello.c
471 2. After working in the wrong branch, switching to the correct branch
474 would be done using:
475 $ git checkout mytopic
477 However, your "wrong" branch and correct mytopic branch may differ
479 in files that you have modified locally, in which case the above
480 checkout would fail like this:
481 $ git checkout mytopic
483 error: You have local changes to 'frotz'; not switching branches.
484 You can give the -m flag to the command, which would try a
486 three-way merge:
487 $ git checkout -m mytopic
489 Auto-merging frotz
490 After this three-way merge, the local modifications are not
492 registered in your index file, so git diff would show you what
493 changes you made since the tip of the new branch.
494 3. When a merge conflict happens during switching branches with the -m
496 option, you would see something like this:
497 $ git checkout -m mytopic
499 Auto-merging frotz
500 ERROR: Merge conflict in frotz
501 fatal: merge program failed
502 At this point, git diff shows the changes cleanly merged as in the
504 previous example, as well as the changes in the conflicted files.
505 Edit and resolve the conflict and mark it resolved with git add as
506 usual:
507 $ edit frotz
509 $ git add frotz
510
513 git-switch(1), git-restore(1)
514
516 Part of the git(1) suite
517Git 2.24.1 12/10/2019 GIT-CHECKOUT(1)