4 This manual is designed to be readable by someone with basic unix
5 commandline skills, but no previous knowledge of git.
7 Chapter 1 gives a brief overview of git commands, without any
8 explanation; you may prefer to skip to chapter 2 on a first reading.
10 Chapters 2 and 3 explain how to fetch and study a project using
11 git--the tools you'd need to build and test a particular version of a
12 software project, to search for regressions, and so on.
14 Chapter 4 explains how to do development with git, and chapter 5 how
15 to share that development with others.
17 Further chapters cover more specialized topics.
19 Comprehensive reference documentation is available through the man
20 pages. For a command such as "git clone", just use
22 ------------------------------------------------
24 ------------------------------------------------
29 This is a quick summary of the major commands; the following chapters
30 will explain how these work in more detail.
32 Creating a new repository
33 -------------------------
37 -----------------------------------------------
38 $ tar xzf project.tar.gz
41 Initialized empty Git repository in .git/
44 -----------------------------------------------
46 From a remote repository:
48 -----------------------------------------------
49 $ git clone git://example.com/pub/project.git
51 -----------------------------------------------
56 -----------------------------------------------
57 $ git branch # list all branches in this repo
58 $ git checkout test # switch working directory to branch "test"
59 $ git branch new # create branch "new" starting at current HEAD
60 $ git branch -d new # delete branch "new"
61 -----------------------------------------------
63 Instead of basing new branch on current HEAD (the default), use:
65 -----------------------------------------------
66 $ git branch new test # branch named "test"
67 $ git branch new v2.6.15 # tag named v2.6.15
68 $ git branch new HEAD^ # commit before the most recent
69 $ git branch new HEAD^^ # commit before that
70 $ git branch new test~10 # ten commits before tip of branch "test"
71 -----------------------------------------------
73 Create and switch to a new branch at the same time:
75 -----------------------------------------------
76 $ git checkout -b new v2.6.15
77 -----------------------------------------------
79 Update and examine branches from the repository you cloned from:
81 -----------------------------------------------
83 $ git branch -r # list
87 $ git branch checkout -b masterwork origin/master
88 -----------------------------------------------
90 Fetch a branch from a different repository, and give it a new
91 name in your repository:
93 -----------------------------------------------
94 $ git fetch git://example.com/project.git theirbranch:mybranch
95 $ git fetch git://example.com/project.git v2.6.15:mybranch
96 -----------------------------------------------
98 Keep a list of repositories you work with regularly:
100 -----------------------------------------------
101 $ git remote add example git://example.com/project.git
102 $ git remote # list remote repositories
105 $ git remote show example # get details
107 URL: git://example.com/project.git
108 Tracked remote branches
110 $ git fetch example # update branches from example
111 $ git branch -r # list all remote branches
112 -----------------------------------------------
118 -----------------------------------------------
119 $ gitk # visualize and browse history
120 $ git log # list all commits
121 $ git log src/ # ...modifying src/
122 $ git log v2.6.15..v2.6.16 # ...in v2.6.16, not in v2.6.15
123 $ git log master..test # ...in branch test, not in branch master
124 $ git log test..master # ...in branch master, but not in test
125 $ git log test...master # ...in one branch, not in both
126 $ git log -S'foo()' # ...where difference contain "foo()"
127 $ git log --since="2 weeks ago"
128 $ git log -p # show patches as well
129 $ git show # most recent commit
130 $ git diff v2.6.15..v2.6.16 # diff between two tagged versions
131 $ git diff v2.6.15..HEAD # diff with current head
132 $ git grep "foo()" # search working directory for "foo()"
133 $ git grep v2.6.15 "foo()" # search old tree for "foo()"
134 $ git show v2.6.15:a.txt # look at old version of a.txt
135 -----------------------------------------------
137 Search for regressions:
139 -----------------------------------------------
141 $ git bisect bad # current version is bad
142 $ git bisect good v2.6.13-rc2 # last known good revision
143 Bisecting: 675 revisions left to test after this
145 $ git bisect good # if this revision is good, or
146 $ git bisect bad # if this revision is bad.
148 -----------------------------------------------
153 Make sure git knows who to blame:
155 ------------------------------------------------
156 $ cat >~/.gitconfig <<\EOF
158 name = Your Name Comes Here
159 email = you@yourdomain.example.com
161 ------------------------------------------------
163 Select file contents to include in the next commit, then make the
166 -----------------------------------------------
167 $ git add a.txt # updated file
168 $ git add b.txt # new file
169 $ git rm c.txt # old file
171 -----------------------------------------------
173 Or, prepare and create the commit in one step:
175 -----------------------------------------------
176 $ git commit d.txt # use latest content only of d.txt
177 $ git commit -a # use latest content of all tracked files
178 -----------------------------------------------
183 -----------------------------------------------
184 $ git merge test # merge branch "test" into the current branch
185 $ git pull git://example.com/project.git master
186 # fetch and merge in remote branch
187 $ git pull . test # equivalent to git merge test
188 -----------------------------------------------
193 Importing or exporting patches:
195 -----------------------------------------------
196 $ git format-patch origin..HEAD # format a patch for each commit
197 # in HEAD but not in origin
198 $ git-am mbox # import patches from the mailbox "mbox"
199 -----------------------------------------------
201 Fetch a branch in a different git repository, then merge into the
204 -----------------------------------------------
205 $ git pull git://example.com/project.git theirbranch
206 -----------------------------------------------
208 Store the fetched branch into a local branch before merging into the
211 -----------------------------------------------
212 $ git pull git://example.com/project.git theirbranch:mybranch
213 -----------------------------------------------
215 After creating commits on a local branch, update the remote
216 branch with your commits:
218 -----------------------------------------------
219 $ git push ssh://example.com/project.git mybranch:theirbranch
220 -----------------------------------------------
222 When remote and local branch are both named "test":
224 -----------------------------------------------
225 $ git push ssh://example.com/project.git test
226 -----------------------------------------------
228 Shortcut version for a frequently used remote repository:
230 -----------------------------------------------
231 $ git remote add example ssh://example.com/project.git
232 $ git push example test
233 -----------------------------------------------
235 Repository maintenance
236 ----------------------
238 Check for corruption:
240 -----------------------------------------------
242 -----------------------------------------------
244 Recompress, remove unused cruft:
246 -----------------------------------------------
248 -----------------------------------------------
250 Repositories and Branches
251 =========================
253 How to get a git repository
254 ---------------------------
256 It will be useful to have a git repository to experiment with as you
259 The best way to get one is by using the gitlink:git-clone[1] command
260 to download a copy of an existing repository for a project that you
261 are interested in. If you don't already have a project in mind, here
262 are some interesting examples:
264 ------------------------------------------------
265 # git itself (approx. 10MB download):
266 $ git clone git://git.kernel.org/pub/scm/git/git.git
267 # the linux kernel (approx. 150MB download):
268 $ git clone git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git
269 ------------------------------------------------
271 The initial clone may be time-consuming for a large project, but you
272 will only need to clone once.
274 The clone command creates a new directory named after the project
275 ("git" or "linux-2.6" in the examples above). After you cd into this
276 directory, you will see that it contains a copy of the project files,
277 together with a special top-level directory named ".git", which
278 contains all the information about the history of the project.
280 In most of the following, examples will be taken from one of the two
283 How to check out a different version of a project
284 -------------------------------------------------
286 Git is best thought of as a tool for storing the history of a
287 collection of files. It stores the history as a compressed
288 collection of interrelated snapshots (versions) of the project's
291 A single git repository may contain multiple branches. Each branch
292 is a bookmark referencing a particular point in the project history.
293 The gitlink:git-branch[1] command shows you the list of branches:
295 ------------------------------------------------
298 ------------------------------------------------
300 A freshly cloned repository contains a single branch, named "master",
301 and the working directory contains the version of the project
302 referred to by the master branch.
304 Most projects also use tags. Tags, like branches, are references
305 into the project's history, and can be listed using the
306 gitlink:git-tag[1] command:
308 ------------------------------------------------
320 ------------------------------------------------
322 Tags are expected to always point at the same version of a project,
323 while branches are expected to advance as development progresses.
325 Create a new branch pointing to one of these versions and check it
326 out using gitlink:git-checkout[1]:
328 ------------------------------------------------
329 $ git checkout -b new v2.6.13
330 ------------------------------------------------
332 The working directory then reflects the contents that the project had
333 when it was tagged v2.6.13, and gitlink:git-branch[1] shows two
334 branches, with an asterisk marking the currently checked-out branch:
336 ------------------------------------------------
340 ------------------------------------------------
342 If you decide that you'd rather see version 2.6.17, you can modify
343 the current branch to point at v2.6.17 instead, with
345 ------------------------------------------------
346 $ git reset --hard v2.6.17
347 ------------------------------------------------
349 Note that if the current branch was your only reference to a
350 particular point in history, then resetting that branch may leave you
351 with no way to find the history it used to point to; so use this
354 Understanding History: Commits
355 ------------------------------
357 Every change in the history of a project is represented by a commit.
358 The gitlink:git-show[1] command shows the most recent commit on the
361 ------------------------------------------------
363 commit 2b5f6dcce5bf94b9b119e9ed8d537098ec61c3d2
364 Author: Jamal Hadi Salim <hadi@cyberus.ca>
365 Date: Sat Dec 2 22:22:25 2006 -0800
367 [XFRM]: Fix aevent structuring to be more complete.
369 aevents can not uniquely identify an SA. We break the ABI with this
370 patch, but consensus is that since it is not yet utilized by any
371 (known) application then it is fine (better do it now than later).
373 Signed-off-by: Jamal Hadi Salim <hadi@cyberus.ca>
374 Signed-off-by: David S. Miller <davem@davemloft.net>
376 diff --git a/Documentation/networking/xfrm_sync.txt b/Documentation/networking/xfrm_sync.txt
377 index 8be626f..d7aac9d 100644
378 --- a/Documentation/networking/xfrm_sync.txt
379 +++ b/Documentation/networking/xfrm_sync.txt
380 @@ -47,10 +47,13 @@ aevent_id structure looks like:
382 struct xfrm_aevent_id {
383 struct xfrm_usersa_id sa_id;
384 + xfrm_address_t saddr;
389 ------------------------------------------------
391 As you can see, a commit shows who made the latest change, what they
394 Every commit has a 40-hexdigit id, sometimes called the "object name"
395 or the "SHA1 id", shown on the first line of the "git show" output.
396 You can usually refer to a commit by a shorter name, such as a tag or a
397 branch name, but this longer name can also be useful. Most
398 importantly, it is a globally unique name for this commit: so if you
399 tell somebody else the object name (for example in email), then you are
400 guaranteed that name will refer to the same commit in their repository
401 that you it does in yours (assuming their repository has that commit at
404 Understanding history: commits, parents, and reachability
405 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
407 Every commit (except the very first commit in a project) also has a
408 parent commit which shows what happened before this commit.
409 Following the chain of parents will eventually take you back to the
410 beginning of the project.
412 However, the commits do not form a simple list; git allows lines of
413 development to diverge and then reconverge, and the point where two
414 lines of development reconverge is called a "merge". The commit
415 representing a merge can therefore have more than one parent, with
416 each parent representing the most recent commit on one of the lines
417 of development leading to that point.
419 The best way to see how this works is using the gitlink:gitk[1]
420 command; running gitk now on a git repository and looking for merge
421 commits will help understand how the git organizes history.
423 In the following, we say that commit X is "reachable" from commit Y
424 if commit X is an ancestor of commit Y. Equivalently, you could say
425 that Y is a descendent of X, or that there is a chain of parents
426 leading from commit Y to commit X.
428 Undestanding history: History diagrams
429 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
431 We will sometimes represent git history using diagrams like the one
432 below. Commits are shown as "o", and the links between them with
433 lines drawn with - / and \. Time goes left to right:
441 If we need to talk about a particular commit, the character "o" may
442 be replaced with another letter or number.
444 Understanding history: What is a branch?
445 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
447 Though we've been using the word "branch" to mean a kind of reference
448 to a particular commit, the word branch is also commonly used to
449 refer to the line of commits leading up to that point. In the
450 example above, git may think of the branch named "A" as just a
451 pointer to one particular commit, but we may refer informally to the
452 line of three commits leading up to that point as all being part of
455 If we need to make it clear that we're just talking about the most
456 recent commit on the branch, we may refer to that commit as the
457 "head" of the branch.
459 Manipulating branches
460 ---------------------
462 Creating, deleting, and modifying branches is quick and easy; here's
463 a summary of the commands:
467 git branch <branch>::
468 create a new branch named <branch>, referencing the same
469 point in history as the current branch
470 git branch <branch> <start-point>::
471 create a new branch named <branch>, referencing
472 <start-point>, which may be specified any way you like,
473 including using a branch name or a tag name
474 git branch -d <branch>::
475 delete the branch <branch>; if the branch you are deleting
476 points to a commit which is not reachable from this branch,
477 this command will fail with a warning.
478 git branch -D <branch>::
479 even if the branch points to a commit not reachable
480 from the current branch, you may know that that commit
481 is still reachable from some other branch or tag. In that
482 case it is safe to use this command to force git to delete
484 git checkout <branch>::
485 make the current branch <branch>, updating the working
486 directory to reflect the version referenced by <branch>
487 git checkout -b <new> <start-point>::
488 create a new branch <new> referencing <start-point>, and
491 It is also useful to know that the special symbol "HEAD" can always
492 be used to refer to the current branch.
494 Examining branches from a remote repository
495 -------------------------------------------
497 The "master" branch that was created at the time you cloned is a copy
498 of the HEAD in the repository that you cloned from. That repository
499 may also have had other branches, though, and your local repository
500 keeps branches which track each of those remote branches, which you
501 can view using the "-r" option to gitlink:git-branch[1]:
503 ------------------------------------------------
513 ------------------------------------------------
515 You cannot check out these remote-tracking branches, but you can
516 examine them on a branch of your own, just as you would a tag:
518 ------------------------------------------------
519 $ git checkout -b my-todo-copy origin/todo
520 ------------------------------------------------
522 Note that the name "origin" is just the name that git uses by default
523 to refer to the repository that you cloned from.
525 [[how-git-stores-references]]
526 Naming branches, tags, and other references
527 -------------------------------------------
529 Branches, remote-tracking branches, and tags are all references to
530 commits. All references are named with a slash-separated path name
531 starting with "refs"; the names we've been using so far are actually
534 - The branch "test" is short for "refs/heads/test".
535 - The tag "v2.6.18" is short for "refs/tags/v2.6.18".
536 - "origin/master" is short for "refs/remotes/origin/master".
538 The full name is occasionally useful if, for example, there ever
539 exists a tag and a branch with the same name.
541 As another useful shortcut, if the repository "origin" posesses only
542 a single branch, you can refer to that branch as just "origin".
544 More generally, if you have defined a remote repository named
545 "example", you can refer to the branch in that repository as
546 "example". And for a repository with multiple branches, this will
547 refer to the branch designated as the "HEAD" branch.
549 For the complete list of paths which git checks for references, and
550 the order it uses to decide which to choose when there are multiple
551 references with the same shorthand name, see the "SPECIFYING
552 REVISIONS" section of gitlink:git-rev-parse[1].
554 [[Updating-a-repository-with-git-fetch]]
555 Updating a repository with git fetch
556 ------------------------------------
558 Eventually the developer cloned from will do additional work in her
559 repository, creating new commits and advancing the branches to point
562 The command "git fetch", with no arguments, will update all of the
563 remote-tracking branches to the latest version found in her
564 repository. It will not touch any of your own branches--not even the
565 "master" branch that was created for you on clone.
567 Fetching branches from other repositories
568 -----------------------------------------
570 You can also track branches from repositories other than the one you
571 cloned from, using gitlink:git-remote[1]:
573 -------------------------------------------------
574 $ git remote add linux-nfs git://linux-nfs.org/pub/nfs-2.6.git
576 * refs/remotes/linux-nfs/master: storing branch 'master' ...
578 -------------------------------------------------
580 New remote-tracking branches will be stored under the shorthand name
581 that you gave "git remote add", in this case linux-nfs:
583 -------------------------------------------------
587 -------------------------------------------------
589 If you run "git fetch <remote>" later, the tracking branches for the
590 named <remote> will be updated.
592 If you examine the file .git/config, you will see that git has added
595 -------------------------------------------------
599 url = git://linux-nfs.org/~bfields/git.git
600 fetch = +refs/heads/*:refs/remotes/linux-nfs-read/*
602 -------------------------------------------------
604 This is what causes git to track the remote's branches; you may
605 modify or delete these configuration options by editing .git/config
608 Exploring git history
609 =====================
611 Git is best thought of as a tool for storing the history of a
612 collection of files. It does this by storing compressed snapshots of
613 the contents of a file heirarchy, together with "commits" which show
614 the relationships between these snapshots.
616 Git provides extremely flexible and fast tools for exploring the
617 history of a project.
619 We start with one specialized tool which is useful for finding the
620 commit that introduced a bug into a project.
622 How to use bisect to find a regression
623 --------------------------------------
625 Suppose version 2.6.18 of your project worked, but the version at
626 "master" crashes. Sometimes the best way to find the cause of such a
627 regression is to perform a brute-force search through the project's
628 history to find the particular commit that caused the problem. The
629 gitlink:git-bisect[1] command can help you do this:
631 -------------------------------------------------
633 $ git bisect good v2.6.18
634 $ git bisect bad master
635 Bisecting: 3537 revisions left to test after this
636 [65934a9a028b88e83e2b0f8b36618fe503349f8e] BLOCK: Make USB storage depend on SCSI rather than selecting it [try #6]
637 -------------------------------------------------
639 If you run "git branch" at this point, you'll see that git has
640 temporarily moved you to a new branch named "bisect". This branch
641 points to a commit (with commit id 65934...) that is reachable from
642 v2.6.19 but not from v2.6.18. Compile and test it, and see whether
643 it crashes. Assume it does crash. Then:
645 -------------------------------------------------
647 Bisecting: 1769 revisions left to test after this
648 [7eff82c8b1511017ae605f0c99ac275a7e21b867] i2c-core: Drop useless bitmaskings
649 -------------------------------------------------
651 checks out an older version. Continue like this, telling git at each
652 stage whether the version it gives you is good or bad, and notice
653 that the number of revisions left to test is cut approximately in
656 After about 13 tests (in this case), it will output the commit id of
657 the guilty commit. You can then examine the commit with
658 gitlink:git-show[1], find out who wrote it, and mail them your bug
659 report with the commit id. Finally, run
661 -------------------------------------------------
663 -------------------------------------------------
665 to return you to the branch you were on before and delete the
666 temporary "bisect" branch.
668 Note that the version which git-bisect checks out for you at each
669 point is just a suggestion, and you're free to try a different
670 version if you think it would be a good idea. For example,
671 occasionally you may land on a commit that broke something unrelated;
674 -------------------------------------------------
675 $ git bisect-visualize
676 -------------------------------------------------
678 which will run gitk and label the commit it chose with a marker that
679 says "bisect". Chose a safe-looking commit nearby, note its commit
680 id, and check it out with:
682 -------------------------------------------------
683 $ git reset --hard fb47ddb2db...
684 -------------------------------------------------
686 then test, run "bisect good" or "bisect bad" as appropriate, and
692 We have seen several ways of naming commits already:
694 - 40-hexdigit SHA1 id
695 - branch name: refers to the commit at the head of the given
697 - tag name: refers to the commit pointed to by the given tag
698 (we've seen branches and tags are special cases of
699 <<how-git-stores-references,references>>).
700 - HEAD: refers to the head of the current branch
702 There are many more; see the "SPECIFYING REVISIONS" section of the
703 gitlink:git-rev-parse[1] man page for the complete list of ways to
704 name revisions. Some examples:
706 -------------------------------------------------
707 $ git show fb47ddb2 # the first few characters of the SHA1 id
708 # are usually enough to specify it uniquely
709 $ git show HEAD^ # the parent of the HEAD commit
710 $ git show HEAD^^ # the grandparent
711 $ git show HEAD~4 # the great-great-grandparent
712 -------------------------------------------------
714 Recall that merge commits may have more than one parent; by default,
715 ^ and ~ follow the first parent listed in the commit, but you can
718 -------------------------------------------------
719 $ git show HEAD^1 # show the first parent of HEAD
720 $ git show HEAD^2 # show the second parent of HEAD
721 -------------------------------------------------
723 In addition to HEAD, there are several other special names for
726 Merges (to be discussed later), as well as operations such as
727 git-reset, which change the currently checked-out commit, generally
728 set ORIG_HEAD to the value HEAD had before the current operation.
730 The git-fetch operation always stores the head of the last fetched
731 branch in FETCH_HEAD. For example, if you run git fetch without
732 specifying a local branch as the target of the operation
734 -------------------------------------------------
735 $ git fetch git://example.com/proj.git theirbranch
736 -------------------------------------------------
738 the fetched commits will still be available from FETCH_HEAD.
740 When we discuss merges we'll also see the special name MERGE_HEAD,
741 which refers to the other branch that we're merging in to the current
744 The gitlink:git-rev-parse[1] command is a low-level command that is
745 occasionally useful for translating some name for a commit to the SHA1 id for
748 -------------------------------------------------
749 $ git rev-parse origin
750 e05db0fd4f31dde7005f075a84f96b360d05984b
751 -------------------------------------------------
756 We can also create a tag to refer to a particular commit; after
759 -------------------------------------------------
760 $ git-tag stable-1 1b2e1d63ff
761 -------------------------------------------------
763 You can use stable-1 to refer to the commit 1b2e1d63ff.
765 This creates a "lightweight" tag. If the tag is a tag you wish to
766 share with others, and possibly sign cryptographically, then you
767 should create a tag object instead; see the gitlink:git-tag[1] man
773 The gitlink:git-log[1] command can show lists of commits. On its
774 own, it shows all commits reachable from the parent commit; but you
775 can also make more specific requests:
777 -------------------------------------------------
778 $ git log v2.5.. # commits since (not reachable from) v2.5
779 $ git log test..master # commits reachable from master but not test
780 $ git log master..test # ...reachable from test but not master
781 $ git log master...test # ...reachable from either test or master,
783 $ git log --since="2 weeks ago" # commits from the last 2 weeks
784 $ git log Makefile # commits which modify Makefile
785 $ git log fs/ # ... which modify any file under fs/
786 $ git log -S'foo()' # commits which add or remove any file data
787 # matching the string 'foo()'
788 -------------------------------------------------
790 And of course you can combine all of these; the following finds
791 commits since v2.5 which touch the Makefile or any file under fs:
793 -------------------------------------------------
794 $ git log v2.5.. Makefile fs/
795 -------------------------------------------------
797 You can also ask git log to show patches:
799 -------------------------------------------------
801 -------------------------------------------------
803 See the "--pretty" option in the gitlink:git-log[1] man page for more
806 Note that git log starts with the most recent commit and works
807 backwards through the parents; however, since git history can contain
808 multiple independant lines of development, the particular order that
809 commits are listed in may be somewhat arbitrary.
814 You can generate diffs between any two versions using
817 -------------------------------------------------
818 $ git diff master..test
819 -------------------------------------------------
821 Sometimes what you want instead is a set of patches:
823 -------------------------------------------------
824 $ git format-patch master..test
825 -------------------------------------------------
827 will generate a file with a patch for each commit reachable from test
828 but not from master. Note that if master also has commits which are
829 not reachable from test, then the combined result of these patches
830 will not be the same as the diff produced by the git-diff example.
832 Viewing old file versions
833 -------------------------
835 You can always view an old version of a file by just checking out the
836 correct revision first. But sometimes it is more convenient to be
837 able to view an old version of a single file without checking
838 anything out; this command does that:
840 -------------------------------------------------
841 $ git show v2.5:fs/locks.c
842 -------------------------------------------------
844 Before the colon may be anything that names a commit, and after it
845 may be any path to a file tracked by git.
850 Check whether two branches point at the same history
851 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
853 Suppose you want to check whether two branches point at the same point
856 -------------------------------------------------
857 $ git diff origin..master
858 -------------------------------------------------
860 will tell you whether the contents of the project are the same at the
861 two branches; in theory, however, it's possible that the same project
862 contents could have been arrived at by two different historical
863 routes. You could compare the SHA1 id's:
865 -------------------------------------------------
866 $ git rev-list origin
867 e05db0fd4f31dde7005f075a84f96b360d05984b
868 $ git rev-list master
869 e05db0fd4f31dde7005f075a84f96b360d05984b
870 -------------------------------------------------
872 Or you could recall that the ... operator selects all commits
873 contained reachable from either one reference or the other but not
876 -------------------------------------------------
877 $ git log origin...master
878 -------------------------------------------------
880 will return no commits when the two branches are equal.
882 Find first tagged version including a given fix
883 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
885 Suppose you know that the commit e05db0fd fixed a certain problem.
886 You'd like to find the earliest tagged release that contains that
889 Of course, there may be more than one answer--if the history branched
890 after commit e05db0fd, then there could be multiple "earliest" tagged
893 You could just visually inspect the commits since e05db0fd:
895 -------------------------------------------------
897 -------------------------------------------------
899 Or you can use gitlink:git-name-rev[1], which will give the commit a
900 name based on any tag it finds pointing to one of the commit's
903 -------------------------------------------------
904 $ git name-rev e05db0fd
905 e05db0fd tags/v1.5.0-rc1^0~23
906 -------------------------------------------------
908 The gitlink:git-describe[1] command does the opposite, naming the
909 revision using a tag on which the given commit is based:
911 -------------------------------------------------
912 $ git describe e05db0fd
914 -------------------------------------------------
916 but that may sometimes help you guess which tags might come after the
919 If you just want to verify whether a given tagged version contains a
920 given commit, you could use gitlink:git-merge-base[1]:
922 -------------------------------------------------
923 $ git merge-base e05db0fd v1.5.0-rc1
924 e05db0fd4f31dde7005f075a84f96b360d05984b
925 -------------------------------------------------
927 The merge-base command finds a common ancestor of the given commits,
928 and always returns one or the other in the case where one is a
929 descendant of the other; so the above output shows that e05db0fd
930 actually is an ancestor of v1.5.0-rc1.
932 Alternatively, note that
934 -------------------------------------------------
935 $ git log v1.5.0-rc1..305db0fd
936 -------------------------------------------------
938 will produce empty output if and only if v1.5.0-rc1 includes 305db0fd,
939 because it outputs only commits that are not reachable from v1.5.0-rc1.
944 Telling git your name
945 ---------------------
947 Before creating any commits, you should introduce yourself to git. The
948 easiest way to do so is:
950 ------------------------------------------------
951 $ cat >~/.gitconfig <<\EOF
953 name = Your Name Comes Here
954 email = you@yourdomain.example.com
956 ------------------------------------------------
959 Creating a new repository
960 -------------------------
962 Creating a new repository from scratch is very easy:
964 -------------------------------------------------
968 -------------------------------------------------
970 If you have some initial content (say, a tarball):
972 -------------------------------------------------
973 $ tar -xzvf project.tar.gz
976 $ git add . # include everything below ./ in the first commit:
978 -------------------------------------------------
980 [[how-to-make-a-commit]]
984 Creating a new commit takes three steps:
986 1. Making some changes to the working directory using your
988 2. Telling git about your changes.
989 3. Creating the commit using the content you told git about
992 In practice, you can interleave and repeat steps 1 and 2 as many
993 times as you want: in order to keep track of what you want committed
994 at step 3, git maintains a snapshot of the tree's contents in a
995 special staging area called "the index."
997 At the beginning, the content of the index will be identical to
998 that of the HEAD. The command "git diff --cached", which shows
999 the difference between the HEAD and the index, should therefore
1000 produce no output at that point.
1002 Modifying the index is easy:
1004 To update the index with the new contents of a modified file, use
1006 -------------------------------------------------
1007 $ git add path/to/file
1008 -------------------------------------------------
1010 To add the contents of a new file to the index, use
1012 -------------------------------------------------
1013 $ git add path/to/file
1014 -------------------------------------------------
1016 To remove a file from the index and from the working tree,
1018 -------------------------------------------------
1019 $ git rm path/to/file
1020 -------------------------------------------------
1022 After each step you can verify that
1024 -------------------------------------------------
1026 -------------------------------------------------
1028 always shows the difference between the HEAD and the index file--this
1029 is what you'd commit if you created the commit now--and that
1031 -------------------------------------------------
1033 -------------------------------------------------
1035 shows the difference between the working tree and the index file.
1037 Note that "git add" always adds just the current contents of a file
1038 to the index; further changes to the same file will be ignored unless
1039 you run git-add on the file again.
1041 When you're ready, just run
1043 -------------------------------------------------
1045 -------------------------------------------------
1047 and git will prompt you for a commit message and then create the new
1048 commmit. Check to make sure it looks like what you expected with
1050 -------------------------------------------------
1052 -------------------------------------------------
1054 As a special shortcut,
1056 -------------------------------------------------
1058 -------------------------------------------------
1060 will update the index with any files that you've modified or removed
1061 and create a commit, all in one step.
1063 A number of commands are useful for keeping track of what you're
1066 -------------------------------------------------
1067 $ git diff --cached # difference between HEAD and the index; what
1068 # would be commited if you ran "commit" now.
1069 $ git diff # difference between the index file and your
1070 # working directory; changes that would not
1071 # be included if you ran "commit" now.
1072 $ git status # a brief per-file summary of the above.
1073 -------------------------------------------------
1075 creating good commit messages
1076 -----------------------------
1078 Though not required, it's a good idea to begin the commit message
1079 with a single short (less than 50 character) line summarizing the
1080 change, followed by a blank line and then a more thorough
1081 description. Tools that turn commits into email, for example, use
1082 the first line on the Subject line and the rest of the commit in the
1088 You can rejoin two diverging branches of development using
1089 gitlink:git-merge[1]:
1091 -------------------------------------------------
1092 $ git merge branchname
1093 -------------------------------------------------
1095 merges the development in the branch "branchname" into the current
1096 branch. If there are conflicts--for example, if the same file is
1097 modified in two different ways in the remote branch and the local
1098 branch--then you are warned; the output may look something like this:
1100 -------------------------------------------------
1102 Trying really trivial in-index merge...
1103 fatal: Merge requires file-level merging
1105 Merging HEAD with 77976da35a11db4580b80ae27e8d65caf5208086
1109 found 1 common ancestor(s):
1111 Auto-merging file.txt
1112 CONFLICT (content): Merge conflict in file.txt
1113 Automatic merge failed; fix conflicts and then commit the result.
1114 -------------------------------------------------
1116 Conflict markers are left in the problematic files, and after
1117 you resolve the conflicts manually, you can update the index
1118 with the contents and run git commit, as you normally would when
1119 creating a new file.
1121 If you examine the resulting commit using gitk, you will see that it
1122 has two parents, one pointing to the top of the current branch, and
1123 one to the top of the other branch.
1127 [[resolving-a-merge]]
1131 When a merge isn't resolved automatically, git leaves the index and
1132 the working tree in a special state that gives you all the
1133 information you need to help resolve the merge.
1135 Files with conflicts are marked specially in the index, so until you
1136 resolve the problem and update the index, git commit will fail:
1138 -------------------------------------------------
1140 file.txt: needs merge
1141 -------------------------------------------------
1143 Also, git status will list those files as "unmerged".
1145 All of the changes that git was able to merge automatically are
1146 already added to the index file, so gitlink:git-diff[1] shows only
1147 the conflicts. Also, it uses a somewhat unusual syntax:
1149 -------------------------------------------------
1152 index 802992c,2b60207..0000000
1155 @@@ -1,1 -1,1 +1,5 @@@
1156 ++<<<<<<< HEAD:file.txt
1160 ++>>>>>>> 77976da35a11db4580b80ae27e8d65caf5208086:file.txt
1161 -------------------------------------------------
1163 Recall that the commit which will be commited after we resolve this
1164 conflict will have two parents instead of the usual one: one parent
1165 will be HEAD, the tip of the current branch; the other will be the
1166 tip of the other branch, which is stored temporarily in MERGE_HEAD.
1168 The diff above shows the differences between the working-tree version
1169 of file.txt and two previous version: one version from HEAD, and one
1170 from MERGE_HEAD. So instead of preceding each line by a single "+"
1171 or "-", it now uses two columns: the first column is used for
1172 differences between the first parent and the working directory copy,
1173 and the second for differences between the second parent and the
1174 working directory copy. Thus after resolving the conflict in the
1175 obvious way, the diff will look like:
1177 -------------------------------------------------
1180 index 802992c,2b60207..0000000
1183 @@@ -1,1 -1,1 +1,1 @@@
1187 -------------------------------------------------
1189 This shows that our resolved version deleted "Hello world" from the
1190 first parent, deleted "Goodbye" from the second parent, and added
1191 "Goodbye world", which was previously absent from both.
1193 The gitlink:git-log[1] command also provides special help for merges:
1195 -------------------------------------------------
1197 -------------------------------------------------
1199 This will list all commits which exist only on HEAD or on MERGE_HEAD,
1200 and which touch an unmerged file.
1202 We can now add the resolved version to the index and commit:
1204 -------------------------------------------------
1207 -------------------------------------------------
1209 Note that the commit message will already be filled in for you with
1210 some information about the merge. Normally you can just use this
1211 default message unchanged, but you may add additional commentary of
1212 your own if desired.
1218 If you get stuck and decide to just give up and throw the whole mess
1219 away, you can always return to the pre-merge state with
1221 -------------------------------------------------
1222 $ git reset --hard HEAD
1223 -------------------------------------------------
1225 Or, if you've already commited the merge that you want to throw away,
1227 -------------------------------------------------
1228 $ git reset --hard HEAD^
1229 -------------------------------------------------
1231 However, this last command can be dangerous in some cases--never
1232 throw away a commit you have already committed if that commit may
1233 itself have been merged into another branch, as doing so may confuse
1239 There is one special case not mentioned above, which is treated
1240 differently. Normally, a merge results in a merge commit, with two
1241 parents, one pointing at each of the two lines of development that
1244 However, if one of the two lines of development is completely
1245 contained within the other--so every commit present in the one is
1246 already contained in the other--then git just performs a
1247 <<fast-forwards,fast forward>>; the head of the current branch is
1248 moved forward to point at the head of the merged-in branch, without
1249 any new commits being created.
1254 If you've messed up the working tree, but haven't yet committed your
1255 mistake, you can return the entire working tree to the last committed
1258 -------------------------------------------------
1259 $ git reset --hard HEAD
1260 -------------------------------------------------
1262 If you make a commit that you later wish you hadn't, there are two
1263 fundamentally different ways to fix the problem:
1265 1. You can create a new commit that undoes whatever was done
1266 by the previous commit. This is the correct thing if your
1267 mistake has already been made public.
1269 2. You can go back and modify the old commit. You should
1270 never do this if you have already made the history public;
1271 git does not normally expect the "history" of a project to
1272 change, and cannot correctly perform repeated merges from
1273 a branch that has had its history changed.
1275 Fixing a mistake with a new commit
1276 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1278 Creating a new commit that reverts an earlier change is very easy;
1279 just pass the gitlink:git-revert[1] command a reference to the bad
1280 commit; for example, to revert the most recent commit:
1282 -------------------------------------------------
1284 -------------------------------------------------
1286 This will create a new commit which undoes the change in HEAD. You
1287 will be given a chance to edit the commit message for the new commit.
1289 You can also revert an earlier change, for example, the next-to-last:
1291 -------------------------------------------------
1293 -------------------------------------------------
1295 In this case git will attempt to undo the old change while leaving
1296 intact any changes made since then. If more recent changes overlap
1297 with the changes to be reverted, then you will be asked to fix
1298 conflicts manually, just as in the case of <<resolving-a-merge,
1299 resolving a merge>>.
1301 Fixing a mistake by editing history
1302 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1304 If the problematic commit is the most recent commit, and you have not
1305 yet made that commit public, then you may just
1306 <<undoing-a-merge,destroy it using git-reset>>.
1309 can edit the working directory and update the index to fix your
1310 mistake, just as if you were going to <<how-to-make-a-commit,create a
1311 new commit>>, then run
1313 -------------------------------------------------
1314 $ git commit --amend
1315 -------------------------------------------------
1317 which will replace the old commit by a new commit incorporating your
1318 changes, giving you a chance to edit the old commit message first.
1320 Again, you should never do this to a commit that may already have
1321 been merged into another branch; use gitlink:git-revert[1] instead in
1324 It is also possible to edit commits further back in the history, but
1325 this is an advanced topic to be left for
1326 <<cleaning-up-history,another chapter>>.
1328 Checking out an old version of a file
1329 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1331 In the process of undoing a previous bad change, you may find it
1332 useful to check out an older version of a particular file using
1333 gitlink:git-checkout[1]. We've used git checkout before to switch
1334 branches, but it has quite different behavior if it is given a path
1337 -------------------------------------------------
1338 $ git checkout HEAD^ path/to/file
1339 -------------------------------------------------
1341 replaces path/to/file by the contents it had in the commit HEAD^, and
1342 also updates the index to match. It does not change branches.
1344 If you just want to look at an old version of the file, without
1345 modifying the working directory, you can do that with
1346 gitlink:git-show[1]:
1348 -------------------------------------------------
1349 $ git show HEAD^ path/to/file
1350 -------------------------------------------------
1352 which will display the given version of the file.
1354 Ensuring good performance
1355 -------------------------
1357 On large repositories, git depends on compression to keep the history
1358 information from taking up to much space on disk or in memory.
1360 This compression is not performed automatically. Therefore you
1361 should occasionally run gitlink:git-gc[1]:
1363 -------------------------------------------------
1365 -------------------------------------------------
1367 to recompress the archive. This can be very time-consuming, so
1368 you may prefer to run git-gc when you are not doing other work.
1370 Ensuring reliability
1371 --------------------
1373 Checking the repository for corruption
1374 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1378 "dangling objects" explanation
1379 Brief explanation here,
1380 include forward reference to longer explanation from
1381 Linus, to be added to later chapter
1383 Recovering lost changes
1384 ~~~~~~~~~~~~~~~~~~~~~~~
1389 low-level examination of objects
1391 Sharing development with others
1392 ===============================
1394 [[getting-updates-with-git-pull]]
1395 Getting updates with git pull
1396 -----------------------------
1398 After you clone a repository and make a few changes of your own, you
1399 may wish to check the original repository for updates and merge them
1402 We have already seen <<Updating-a-repository-with-git-fetch,how to
1403 keep remote tracking branches up to date>> with gitlink:git-fetch[1],
1404 and how to merge two branches. So you can merge in changes from the
1405 original repository's master branch with:
1407 -------------------------------------------------
1409 $ git merge origin/master
1410 -------------------------------------------------
1412 However, the gitlink:git-pull[1] command provides a way to do this in
1415 -------------------------------------------------
1416 $ git pull origin master
1417 -------------------------------------------------
1419 In fact, "origin" is normally the default repository to pull from,
1420 and the default branch is normally the HEAD of the remote repository,
1421 so often you can accomplish the above with just
1423 -------------------------------------------------
1425 -------------------------------------------------
1427 See the descriptions of the branch.<name>.remote and
1428 branch.<name>.merge options in gitlink:git-repo-config[1] to learn
1429 how to control these defaults depending on the current branch.
1431 In addition to saving you keystrokes, "git pull" also helps you by
1432 producing a default commit message documenting the branch and
1433 repository that you pulled from.
1435 (But note that no such commit will be created in the case of a
1436 <<fast-forwards,fast forward>>; instead, your branch will just be
1437 updated to point to the latest commit from the upstream branch).
1439 The git-pull command can also be given "." as the "remote" repository, in
1440 which case it just merges in a branch from the current repository; so
1443 -------------------------------------------------
1446 -------------------------------------------------
1448 are roughly equivalent. The former is actually very commonly used.
1450 Submitting patches to a project
1451 -------------------------------
1453 If you just have a few changes, the simplest way to submit them may
1454 just be to send them as patches in email:
1456 First, use gitlink:git-format-patches[1]; for example:
1458 -------------------------------------------------
1459 $ git format-patch origin
1460 -------------------------------------------------
1462 will produce a numbered series of files in the current directory, one
1463 for each patch in the current branch but not in origin/HEAD.
1465 You can then import these into your mail client and send them by
1466 hand. However, if you have a lot to send at once, you may prefer to
1467 use the gitlink:git-send-email[1] script to automate the process.
1468 Consult the mailing list for your project first to determine how they
1469 prefer such patches be handled.
1471 Importing patches to a project
1472 ------------------------------
1474 Git also provides a tool called gitlink:git-am[1] (am stands for
1475 "apply mailbox"), for importing such an emailed series of patches.
1476 Just save all of the patch-containing messages, in order, into a
1477 single mailbox file, say "patches.mbox", then run
1479 -------------------------------------------------
1480 $ git am -3 patches.mbox
1481 -------------------------------------------------
1483 Git will apply each patch in order; if any conflicts are found, it
1484 will stop, and you can fix the conflicts as described in
1485 "<<resolving-a-merge,Resolving a merge>>". (The "-3" option tells
1486 git to perform a merge; if you would prefer it just to abort and
1487 leave your tree and index untouched, you may omit that option.)
1489 Once the index is updated with the results of the conflict
1490 resolution, instead of creating a new commit, just run
1492 -------------------------------------------------
1494 -------------------------------------------------
1496 and git will create the commit for you and continue applying the
1497 remaining patches from the mailbox.
1499 The final result will be a series of commits, one for each patch in
1500 the original mailbox, with authorship and commit log message each
1501 taken from the message containing each patch.
1503 [[setting-up-a-public-repository]]
1504 Setting up a public repository
1505 ------------------------------
1507 Another way to submit changes to a project is to simply tell the
1508 maintainer of that project to pull from your repository, exactly as
1509 you did in the section "<<getting-updates-with-git-pull, Getting
1510 updates with git pull>>".
1512 If you and maintainer both have accounts on the same machine, then
1513 then you can just pull changes from each other's repositories
1514 directly; note that all of the command (gitlink:git-clone[1],
1515 git-fetch[1], git-pull[1], etc.) which accept a URL as an argument
1516 will also accept a local file patch; so, for example, you can
1519 -------------------------------------------------
1520 $ git clone /path/to/repository
1521 $ git pull /path/to/other/repository
1522 -------------------------------------------------
1524 If this sort of setup is inconvenient or impossible, another (more
1525 common) option is to set up a public repository on a public server.
1526 This also allows you to cleanly separate private work in progress
1527 from publicly visible work.
1529 You will continue to do your day-to-day work in your personal
1530 repository, but periodically "push" changes from your personal
1531 repository into your public repository, allowing other developers to
1532 pull from that repository. So the flow of changes, in a situation
1533 where there is one other developer with a public repository, looks
1537 your personal repo ------------------> your public repo
1540 | you pull | they pull
1544 their public repo <------------------- their repo
1546 Now, assume your personal repository is in the directory ~/proj. We
1547 first create a new clone of the repository:
1549 -------------------------------------------------
1550 $ git clone --bare proj-clone.git
1551 -------------------------------------------------
1553 The resulting directory proj-clone.git will contains a "bare" git
1554 repository--it is just the contents of the ".git" directory, without
1555 a checked-out copy of a working directory.
1557 Next, copy proj-clone.git to the server where you plan to host the
1558 public repository. You can use scp, rsync, or whatever is most
1561 If somebody else maintains the public server, they may already have
1562 set up a git service for you, and you may skip to the section
1563 "<<pushing-changes-to-a-public-repository,Pushing changes to a public
1564 repository>>", below.
1566 Otherwise, the following sections explain how to export your newly
1567 created public repository:
1569 [[exporting-via-http]]
1570 Exporting a git repository via http
1571 -----------------------------------
1573 The git protocol gives better performance and reliability, but on a
1574 host with a web server set up, http exports may be simpler to set up.
1576 All you need to do is place the newly created bare git repository in
1577 a directory that is exported by the web server, and make some
1578 adjustments to give web clients some extra information they need:
1580 -------------------------------------------------
1581 $ mv proj.git /home/you/public_html/proj.git
1583 $ git update-server-info
1584 $ chmod a+x hooks/post-update
1585 -------------------------------------------------
1587 (For an explanation of the last two lines, see
1588 gitlink:git-update-server-info[1], and the documentation
1589 link:hooks.txt[Hooks used by git].)
1591 Advertise the url of proj.git. Anybody else should then be able to
1592 clone or pull from that url, for example with a commandline like:
1594 -------------------------------------------------
1595 $ git clone http://yourserver.com/~you/proj.git
1596 -------------------------------------------------
1599 link:howto/setup-git-server-over-http.txt[setup-git-server-over-http]
1600 for a slightly more sophisticated setup using WebDAV which also
1601 allows pushing over http.)
1603 [[exporting-via-git]]
1604 Exporting a git repository via the git protocol
1605 -----------------------------------------------
1607 This is the preferred method.
1609 For now, we refer you to the gitlink:git-daemon[1] man page for
1610 instructions. (See especially the examples section.)
1612 [[pushing-changes-to-a-public-repository]]
1613 Pushing changes to a public repository
1614 --------------------------------------
1616 Note that the two techniques outline above (exporting via
1617 <<exporting-via-http,http>> or <<exporting-via-git,git>>) allow other
1618 maintainers to fetch your latest changes, but they do not allow write
1619 access, which you will need to update the public repository with the
1620 latest changes created in your private repository.
1622 The simplest way to do this is using gitlink:git-push[1] and ssh; to
1623 update the remote branch named "master" with the latest state of your
1624 branch named "master", run
1626 -------------------------------------------------
1627 $ git push ssh://yourserver.com/~you/proj.git master:master
1628 -------------------------------------------------
1632 -------------------------------------------------
1633 $ git push ssh://yourserver.com/~you/proj.git master
1634 -------------------------------------------------
1636 As with git-fetch, git-push will complain if this does not result in
1637 a <<fast-forwards,fast forward>>. Normally this is a sign of
1638 something wrong. However, if you are sure you know what you're
1639 doing, you may force git-push to perform the update anyway by
1640 proceeding the branch name by a plus sign:
1642 -------------------------------------------------
1643 $ git push ssh://yourserver.com/~you/proj.git +master
1644 -------------------------------------------------
1646 As with git-fetch, you may also set up configuration options to
1647 save typing; so, for example, after
1649 -------------------------------------------------
1650 $ cat >.git/config <<EOF
1651 [remote "public-repo"]
1652 url = ssh://yourserver.com/~you/proj.git
1654 -------------------------------------------------
1656 you should be able to perform the above push with just
1658 -------------------------------------------------
1659 $ git push public-repo master
1660 -------------------------------------------------
1662 See the explanations of the remote.<name>.url, branch.<name>.remote,
1663 and remote.<name>.push options in gitlink:git-repo-config[1] for
1666 Setting up a shared repository
1667 ------------------------------
1669 Another way to collaborate is by using a model similar to that
1670 commonly used in CVS, where several developers with special rights
1671 all push to and pull from a single shared repository. See
1672 link:cvs-migration.txt[git for CVS users] for instructions on how to
1675 Allow web browsing of a repository
1676 ----------------------------------
1678 TODO: Brief setup-instructions for gitweb
1683 TODO: topic branches, typical roles as in everyday.txt, ?
1686 Working with other version control systems
1687 ==========================================
1689 TODO: CVS, Subversion, series-of-release-tarballs, etc.
1691 [[cleaning-up-history]]
1692 Rewriting history and maintaining patch series
1693 ==============================================
1695 Normally commits are only added to a project, never taken away or
1696 replaced. Git is designed with this assumption, and violating it will
1697 cause git's merge machinery (for example) to do the wrong thing.
1699 However, there is a situation in which it can be useful to violate this
1702 Creating the perfect patch series
1703 ---------------------------------
1705 Suppose you are a contributor to a large project, and you want to add a
1706 complicated feature, and to present it to the other developers in a way
1707 that makes it easy for them to read your changes, verify that they are
1708 correct, and understand why you made each change.
1710 If you present all of your changes as a single patch (or commit), they
1711 may find it is too much to digest all at once.
1713 If you present them with the entire history of your work, complete with
1714 mistakes, corrections, and dead ends, they may be overwhelmed.
1716 So the ideal is usually to produce a series of patches such that:
1718 1. Each patch can be applied in order.
1720 2. Each patch includes a single logical change, together with a
1721 message explaining the change.
1723 3. No patch introduces a regression: after applying any initial
1724 part of the series, the resulting project still compiles and
1725 works, and has no bugs that it didn't have before.
1727 4. The complete series produces the same end result as your own
1728 (probably much messier!) development process did.
1730 We will introduce some tools that can help you do this, explain how to
1731 use them, and then explain some of the problems that can arise because
1732 you are rewriting history.
1734 Keeping a patch series up to date using git-rebase
1735 --------------------------------------------------
1737 Suppose you have a series of commits in a branch "mywork", which
1738 originally branched off from "origin".
1740 Suppose you create a branch "mywork" on a remote-tracking branch
1741 "origin", and created some commits on top of it:
1743 -------------------------------------------------
1744 $ git checkout -b mywork origin
1750 -------------------------------------------------
1752 You have performed no merges into mywork, so it is just a simple linear
1753 sequence of patches on top of "origin":
1760 Some more interesting work has been done in the upstream project, and
1761 "origin" has advanced:
1763 o--o--O--o--o--o <-- origin
1767 At this point, you could use "pull" to merge your changes back in;
1768 the result would create a new merge commit, like this:
1771 o--o--O--o--o--o <-- origin
1773 a--b--c--m <-- mywork
1775 However, if you prefer to keep the history in mywork a simple series of
1776 commits without any merges, you may instead choose to use
1777 gitlink:git-rebase[1]:
1779 -------------------------------------------------
1780 $ git checkout mywork
1782 -------------------------------------------------
1784 This will remove each of your commits from mywork, temporarily saving
1785 them as patches (in a directory named ".dotest"), update mywork to
1786 point at the latest version of origin, then apply each of the saved
1787 patches to the new mywork. The result will look like:
1790 o--o--O--o--o--o <-- origin
1792 a'--b'--c' <-- mywork
1794 In the process, it may discover conflicts. In that case it will stop
1795 and allow you to fix the conflicts; after fixing conflicts, use "git
1796 add" to update the index with those contents, and then, instead of
1797 running git-commit, just run
1799 -------------------------------------------------
1800 $ git rebase --continue
1801 -------------------------------------------------
1803 and git will continue applying the rest of the patches.
1805 At any point you may use the --abort option to abort this process and
1806 return mywork to the state it had before you started the rebase:
1808 -------------------------------------------------
1809 $ git rebase --abort
1810 -------------------------------------------------
1812 Reordering or selecting from a patch series
1813 -------------------------------------------
1815 Given one existing commit, the gitlink:git-cherry-pick[1] command
1816 allows you to apply the change introduced by that commit and create a
1817 new commit that records it. So, for example, if "mywork" points to a
1818 series of patches on top of "origin", you might do something like:
1820 -------------------------------------------------
1821 $ git checkout -b mywork-new origin
1822 $ gitk origin..mywork &
1823 -------------------------------------------------
1825 And browse through the list of patches in the mywork branch using gitk,
1826 applying them (possibly in a different order) to mywork-new using
1827 cherry-pick, and possibly modifying them as you go using commit
1830 Another technique is to use git-format-patch to create a series of
1831 patches, then reset the state to before the patches:
1833 -------------------------------------------------
1834 $ git format-patch origin
1835 $ git reset --hard origin
1836 -------------------------------------------------
1838 Then modify, reorder, or eliminate patches as preferred before applying
1839 them again with gitlink:git-am[1].
1844 There are numerous other tools, such as stgit, which exist for the
1845 purpose of maintaining a patch series. These are out of the scope of
1848 Problems with rewriting history
1849 -------------------------------
1851 The primary problem with rewriting the history of a branch has to do
1852 with merging. Suppose somebody fetches your branch and merges it into
1853 their branch, with a result something like this:
1855 o--o--O--o--o--o <-- origin
1857 t--t--t--m <-- their branch:
1859 Then suppose you modify the last three commits:
1861 o--o--o <-- new head of origin
1863 o--o--O--o--o--o <-- old head of origin
1865 If we examined all this history together in one repository, it will
1868 o--o--o <-- new head of origin
1870 o--o--O--o--o--o <-- old head of origin
1872 t--t--t--m <-- their branch:
1874 Git has no way of knowing that the new head is an updated version of
1875 the old head; it treats this situation exactly the same as it would if
1876 two developers had independently done the work on the old and new heads
1877 in parallel. At this point, if someone attempts to merge the new head
1878 in to their branch, git will attempt to merge together the two (old and
1879 new) lines of development, instead of trying to replace the old by the
1880 new. The results are likely to be unexpected.
1882 You may still choose to publish branches whose history is rewritten,
1883 and it may be useful for others to be able to fetch those branches in
1884 order to examine or test them, but they should not attempt to pull such
1885 branches into their own work.
1887 For true distributed development that supports proper merging,
1888 published branches should never be rewritten.
1890 Advanced branch management
1891 ==========================
1893 Fetching individual branches
1894 ----------------------------
1896 Instead of using gitlink:git-remote[1], you can also choose just
1897 to update one branch at a time, and to store it locally under an
1900 -------------------------------------------------
1901 $ git fetch origin todo:my-todo-work
1902 -------------------------------------------------
1904 The first argument, "origin", just tells git to fetch from the
1905 repository you originally cloned from. The second argument tells git
1906 to fetch the branch named "todo" from the remote repository, and to
1907 store it locally under the name refs/heads/my-todo-work.
1909 You can also fetch branches from other repositories; so
1911 -------------------------------------------------
1912 $ git fetch git://example.com/proj.git master:example-master
1913 -------------------------------------------------
1915 will create a new branch named "example-master" and store in it the
1916 branch named "master" from the repository at the given URL. If you
1917 already have a branch named example-master, it will attempt to
1918 "fast-forward" to the commit given by example.com's master branch. So
1919 next we explain what a fast-forward is:
1922 Understanding git history: fast-forwards
1923 ----------------------------------------
1925 In the previous example, when updating an existing branch, "git
1926 fetch" checks to make sure that the most recent commit on the remote
1927 branch is a descendant of the most recent commit on your copy of the
1928 branch before updating your copy of the branch to point at the new
1929 commit. Git calls this process a "fast forward".
1931 A fast forward looks something like this:
1933 o--o--o--o <-- old head of the branch
1935 o--o--o <-- new head of the branch
1938 In some cases it is possible that the new head will *not* actually be
1939 a descendant of the old head. For example, the developer may have
1940 realized she made a serious mistake, and decided to backtrack,
1941 resulting in a situation like:
1943 o--o--o--o--a--b <-- old head of the branch
1945 o--o--o <-- new head of the branch
1949 In this case, "git fetch" will fail, and print out a warning.
1951 In that case, you can still force git to update to the new head, as
1952 described in the following section. However, note that in the
1953 situation above this may mean losing the commits labeled "a" and "b",
1954 unless you've already created a reference of your own pointing to
1957 Forcing git fetch to do non-fast-forward updates
1958 ------------------------------------------------
1960 If git fetch fails because the new head of a branch is not a
1961 descendant of the old head, you may force the update with:
1963 -------------------------------------------------
1964 $ git fetch git://example.com/proj.git +master:refs/remotes/example/master
1965 -------------------------------------------------
1967 Note the addition of the "+" sign. Be aware that commits which the
1968 old version of example/master pointed at may be lost, as we saw in
1969 the previous section.
1971 Configuring remote branches
1972 ---------------------------
1974 We saw above that "origin" is just a shortcut to refer to the
1975 repository which you originally cloned from. This information is
1976 stored in git configuration variables, which you can see using
1977 gitlink:git-repo-config[1]:
1979 -------------------------------------------------
1980 $ git-repo-config -l
1981 core.repositoryformatversion=0
1983 core.logallrefupdates=true
1984 remote.origin.url=git://git.kernel.org/pub/scm/git/git.git
1985 remote.origin.fetch=+refs/heads/*:refs/remotes/origin/*
1986 branch.master.remote=origin
1987 branch.master.merge=refs/heads/master
1988 -------------------------------------------------
1990 If there are other repositories that you also use frequently, you can
1991 create similar configuration options to save typing; for example,
1994 -------------------------------------------------
1995 $ git repo-config remote.example.url git://example.com/proj.git
1996 -------------------------------------------------
1998 then the following two commands will do the same thing:
2000 -------------------------------------------------
2001 $ git fetch git://example.com/proj.git master:refs/remotes/example/master
2002 $ git fetch example master:refs/remotes/example/master
2003 -------------------------------------------------
2005 Even better, if you add one more option:
2007 -------------------------------------------------
2008 $ git repo-config remote.example.fetch master:refs/remotes/example/master
2009 -------------------------------------------------
2011 then the following commands will all do the same thing:
2013 -------------------------------------------------
2014 $ git fetch git://example.com/proj.git master:ref/remotes/example/master
2015 $ git fetch example master:ref/remotes/example/master
2016 $ git fetch example example/master
2018 -------------------------------------------------
2020 You can also add a "+" to force the update each time:
2022 -------------------------------------------------
2023 $ git repo-config remote.example.fetch +master:ref/remotes/example/master
2024 -------------------------------------------------
2026 Don't do this unless you're sure you won't mind "git fetch" possibly
2027 throwing away commits on mybranch.
2029 Also note that all of the above configuration can be performed by
2030 directly editing the file .git/config instead of using
2031 gitlink:git-repo-config[1].
2033 See gitlink:git-repo-config[1] for more details on the configuration
2034 options mentioned above.
2040 There are two object abstractions: the "object database", and the
2041 "current directory cache" aka "index".
2046 The object database is literally just a content-addressable collection
2047 of objects. All objects are named by their content, which is
2048 approximated by the SHA1 hash of the object itself. Objects may refer
2049 to other objects (by referencing their SHA1 hash), and so you can
2050 build up a hierarchy of objects.
2052 All objects have a statically determined "type" aka "tag", which is
2053 determined at object creation time, and which identifies the format of
2054 the object (i.e. how it is used, and how it can refer to other
2055 objects). There are currently four different object types: "blob",
2056 "tree", "commit" and "tag".
2058 A "blob" object cannot refer to any other object, and is, like the type
2059 implies, a pure storage object containing some user data. It is used to
2060 actually store the file data, i.e. a blob object is associated with some
2061 particular version of some file.
2063 A "tree" object is an object that ties one or more "blob" objects into a
2064 directory structure. In addition, a tree object can refer to other tree
2065 objects, thus creating a directory hierarchy.
2067 A "commit" object ties such directory hierarchies together into
2068 a DAG of revisions - each "commit" is associated with exactly one tree
2069 (the directory hierarchy at the time of the commit). In addition, a
2070 "commit" refers to one or more "parent" commit objects that describe the
2071 history of how we arrived at that directory hierarchy.
2073 As a special case, a commit object with no parents is called the "root"
2074 object, and is the point of an initial project commit. Each project
2075 must have at least one root, and while you can tie several different
2076 root objects together into one project by creating a commit object which
2077 has two or more separate roots as its ultimate parents, that's probably
2078 just going to confuse people. So aim for the notion of "one root object
2079 per project", even if git itself does not enforce that.
2081 A "tag" object symbolically identifies and can be used to sign other
2082 objects. It contains the identifier and type of another object, a
2083 symbolic name (of course!) and, optionally, a signature.
2085 Regardless of object type, all objects share the following
2086 characteristics: they are all deflated with zlib, and have a header
2087 that not only specifies their type, but also provides size information
2088 about the data in the object. It's worth noting that the SHA1 hash
2089 that is used to name the object is the hash of the original data
2090 plus this header, so `sha1sum` 'file' does not match the object name
2092 (Historical note: in the dawn of the age of git the hash
2093 was the sha1 of the 'compressed' object.)
2095 As a result, the general consistency of an object can always be tested
2096 independently of the contents or the type of the object: all objects can
2097 be validated by verifying that (a) their hashes match the content of the
2098 file and (b) the object successfully inflates to a stream of bytes that
2099 forms a sequence of <ascii type without space> + <space> + <ascii decimal
2100 size> + <byte\0> + <binary object data>.
2102 The structured objects can further have their structure and
2103 connectivity to other objects verified. This is generally done with
2104 the `git-fsck-objects` program, which generates a full dependency graph
2105 of all objects, and verifies their internal consistency (in addition
2106 to just verifying their superficial consistency through the hash).
2108 The object types in some more detail:
2113 A "blob" object is nothing but a binary blob of data, and doesn't
2114 refer to anything else. There is no signature or any other
2115 verification of the data, so while the object is consistent (it 'is'
2116 indexed by its sha1 hash, so the data itself is certainly correct), it
2117 has absolutely no other attributes. No name associations, no
2118 permissions. It is purely a blob of data (i.e. normally "file
2121 In particular, since the blob is entirely defined by its data, if two
2122 files in a directory tree (or in multiple different versions of the
2123 repository) have the same contents, they will share the same blob
2124 object. The object is totally independent of its location in the
2125 directory tree, and renaming a file does not change the object that
2126 file is associated with in any way.
2128 A blob is typically created when gitlink:git-update-index[1]
2129 is run, and its data can be accessed by gitlink:git-cat-file[1].
2134 The next hierarchical object type is the "tree" object. A tree object
2135 is a list of mode/name/blob data, sorted by name. Alternatively, the
2136 mode data may specify a directory mode, in which case instead of
2137 naming a blob, that name is associated with another TREE object.
2139 Like the "blob" object, a tree object is uniquely determined by the
2140 set contents, and so two separate but identical trees will always
2141 share the exact same object. This is true at all levels, i.e. it's
2142 true for a "leaf" tree (which does not refer to any other trees, only
2143 blobs) as well as for a whole subdirectory.
2145 For that reason a "tree" object is just a pure data abstraction: it
2146 has no history, no signatures, no verification of validity, except
2147 that since the contents are again protected by the hash itself, we can
2148 trust that the tree is immutable and its contents never change.
2150 So you can trust the contents of a tree to be valid, the same way you
2151 can trust the contents of a blob, but you don't know where those
2152 contents 'came' from.
2154 Side note on trees: since a "tree" object is a sorted list of
2155 "filename+content", you can create a diff between two trees without
2156 actually having to unpack two trees. Just ignore all common parts,
2157 and your diff will look right. In other words, you can effectively
2158 (and efficiently) tell the difference between any two random trees by
2159 O(n) where "n" is the size of the difference, rather than the size of
2162 Side note 2 on trees: since the name of a "blob" depends entirely and
2163 exclusively on its contents (i.e. there are no names or permissions
2164 involved), you can see trivial renames or permission changes by
2165 noticing that the blob stayed the same. However, renames with data
2166 changes need a smarter "diff" implementation.
2168 A tree is created with gitlink:git-write-tree[1] and
2169 its data can be accessed by gitlink:git-ls-tree[1].
2170 Two trees can be compared with gitlink:git-diff-tree[1].
2175 The "commit" object is an object that introduces the notion of
2176 history into the picture. In contrast to the other objects, it
2177 doesn't just describe the physical state of a tree, it describes how
2178 we got there, and why.
2180 A "commit" is defined by the tree-object that it results in, the
2181 parent commits (zero, one or more) that led up to that point, and a
2182 comment on what happened. Again, a commit is not trusted per se:
2183 the contents are well-defined and "safe" due to the cryptographically
2184 strong signatures at all levels, but there is no reason to believe
2185 that the tree is "good" or that the merge information makes sense.
2186 The parents do not have to actually have any relationship with the
2187 result, for example.
2189 Note on commits: unlike real SCM's, commits do not contain
2190 rename information or file mode change information. All of that is
2191 implicit in the trees involved (the result tree, and the result trees
2192 of the parents), and describing that makes no sense in this idiotic
2195 A commit is created with gitlink:git-commit-tree[1] and
2196 its data can be accessed by gitlink:git-cat-file[1].
2201 An aside on the notion of "trust". Trust is really outside the scope
2202 of "git", but it's worth noting a few things. First off, since
2203 everything is hashed with SHA1, you 'can' trust that an object is
2204 intact and has not been messed with by external sources. So the name
2205 of an object uniquely identifies a known state - just not a state that
2206 you may want to trust.
2208 Furthermore, since the SHA1 signature of a commit refers to the
2209 SHA1 signatures of the tree it is associated with and the signatures
2210 of the parent, a single named commit specifies uniquely a whole set
2211 of history, with full contents. You can't later fake any step of the
2212 way once you have the name of a commit.
2214 So to introduce some real trust in the system, the only thing you need
2215 to do is to digitally sign just 'one' special note, which includes the
2216 name of a top-level commit. Your digital signature shows others
2217 that you trust that commit, and the immutability of the history of
2218 commits tells others that they can trust the whole history.
2220 In other words, you can easily validate a whole archive by just
2221 sending out a single email that tells the people the name (SHA1 hash)
2222 of the top commit, and digitally sign that email using something
2225 To assist in this, git also provides the tag object...
2230 Git provides the "tag" object to simplify creating, managing and
2231 exchanging symbolic and signed tokens. The "tag" object at its
2232 simplest simply symbolically identifies another object by containing
2233 the sha1, type and symbolic name.
2235 However it can optionally contain additional signature information
2236 (which git doesn't care about as long as there's less than 8k of
2237 it). This can then be verified externally to git.
2239 Note that despite the tag features, "git" itself only handles content
2240 integrity; the trust framework (and signature provision and
2241 verification) has to come from outside.
2243 A tag is created with gitlink:git-mktag[1],
2244 its data can be accessed by gitlink:git-cat-file[1],
2245 and the signature can be verified by
2246 gitlink:git-verify-tag[1].
2249 The "index" aka "Current Directory Cache"
2250 -----------------------------------------
2252 The index is a simple binary file, which contains an efficient
2253 representation of a virtual directory content at some random time. It
2254 does so by a simple array that associates a set of names, dates,
2255 permissions and content (aka "blob") objects together. The cache is
2256 always kept ordered by name, and names are unique (with a few very
2257 specific rules) at any point in time, but the cache has no long-term
2258 meaning, and can be partially updated at any time.
2260 In particular, the index certainly does not need to be consistent with
2261 the current directory contents (in fact, most operations will depend on
2262 different ways to make the index 'not' be consistent with the directory
2263 hierarchy), but it has three very important attributes:
2265 '(a) it can re-generate the full state it caches (not just the
2266 directory structure: it contains pointers to the "blob" objects so
2267 that it can regenerate the data too)'
2269 As a special case, there is a clear and unambiguous one-way mapping
2270 from a current directory cache to a "tree object", which can be
2271 efficiently created from just the current directory cache without
2272 actually looking at any other data. So a directory cache at any one
2273 time uniquely specifies one and only one "tree" object (but has
2274 additional data to make it easy to match up that tree object with what
2275 has happened in the directory)
2277 '(b) it has efficient methods for finding inconsistencies between that
2278 cached state ("tree object waiting to be instantiated") and the
2281 '(c) it can additionally efficiently represent information about merge
2282 conflicts between different tree objects, allowing each pathname to be
2283 associated with sufficient information about the trees involved that
2284 you can create a three-way merge between them.'
2286 Those are the three ONLY things that the directory cache does. It's a
2287 cache, and the normal operation is to re-generate it completely from a
2288 known tree object, or update/compare it with a live tree that is being
2289 developed. If you blow the directory cache away entirely, you generally
2290 haven't lost any information as long as you have the name of the tree
2293 At the same time, the index is at the same time also the
2294 staging area for creating new trees, and creating a new tree always
2295 involves a controlled modification of the index file. In particular,
2296 the index file can have the representation of an intermediate tree that
2297 has not yet been instantiated. So the index can be thought of as a
2298 write-back cache, which can contain dirty information that has not yet
2299 been written back to the backing store.
2306 Generally, all "git" operations work on the index file. Some operations
2307 work *purely* on the index file (showing the current state of the
2308 index), but most operations move data to and from the index file. Either
2309 from the database or from the working directory. Thus there are four
2312 working directory -> index
2313 ~~~~~~~~~~~~~~~~~~~~~~~~~~
2315 You update the index with information from the working directory with
2316 the gitlink:git-update-index[1] command. You
2317 generally update the index information by just specifying the filename
2318 you want to update, like so:
2320 -------------------------------------------------
2321 $ git-update-index filename
2322 -------------------------------------------------
2324 but to avoid common mistakes with filename globbing etc, the command
2325 will not normally add totally new entries or remove old entries,
2326 i.e. it will normally just update existing cache entries.
2328 To tell git that yes, you really do realize that certain files no
2329 longer exist, or that new files should be added, you
2330 should use the `--remove` and `--add` flags respectively.
2332 NOTE! A `--remove` flag does 'not' mean that subsequent filenames will
2333 necessarily be removed: if the files still exist in your directory
2334 structure, the index will be updated with their new status, not
2335 removed. The only thing `--remove` means is that update-cache will be
2336 considering a removed file to be a valid thing, and if the file really
2337 does not exist any more, it will update the index accordingly.
2339 As a special case, you can also do `git-update-index --refresh`, which
2340 will refresh the "stat" information of each index to match the current
2341 stat information. It will 'not' update the object status itself, and
2342 it will only update the fields that are used to quickly test whether
2343 an object still matches its old backing store object.
2345 index -> object database
2346 ~~~~~~~~~~~~~~~~~~~~~~~~
2348 You write your current index file to a "tree" object with the program
2350 -------------------------------------------------
2352 -------------------------------------------------
2354 that doesn't come with any options - it will just write out the
2355 current index into the set of tree objects that describe that state,
2356 and it will return the name of the resulting top-level tree. You can
2357 use that tree to re-generate the index at any time by going in the
2360 object database -> index
2361 ~~~~~~~~~~~~~~~~~~~~~~~~
2363 You read a "tree" file from the object database, and use that to
2364 populate (and overwrite - don't do this if your index contains any
2365 unsaved state that you might want to restore later!) your current
2366 index. Normal operation is just
2368 -------------------------------------------------
2369 $ git-read-tree <sha1 of tree>
2370 -------------------------------------------------
2372 and your index file will now be equivalent to the tree that you saved
2373 earlier. However, that is only your 'index' file: your working
2374 directory contents have not been modified.
2376 index -> working directory
2377 ~~~~~~~~~~~~~~~~~~~~~~~~~~
2379 You update your working directory from the index by "checking out"
2380 files. This is not a very common operation, since normally you'd just
2381 keep your files updated, and rather than write to your working
2382 directory, you'd tell the index files about the changes in your
2383 working directory (i.e. `git-update-index`).
2385 However, if you decide to jump to a new version, or check out somebody
2386 else's version, or just restore a previous tree, you'd populate your
2387 index file with read-tree, and then you need to check out the result
2390 -------------------------------------------------
2391 $ git-checkout-index filename
2392 -------------------------------------------------
2394 or, if you want to check out all of the index, use `-a`.
2396 NOTE! git-checkout-index normally refuses to overwrite old files, so
2397 if you have an old version of the tree already checked out, you will
2398 need to use the "-f" flag ('before' the "-a" flag or the filename) to
2399 'force' the checkout.
2402 Finally, there are a few odds and ends which are not purely moving
2403 from one representation to the other:
2405 Tying it all together
2406 ~~~~~~~~~~~~~~~~~~~~~
2408 To commit a tree you have instantiated with "git-write-tree", you'd
2409 create a "commit" object that refers to that tree and the history
2410 behind it - most notably the "parent" commits that preceded it in
2413 Normally a "commit" has one parent: the previous state of the tree
2414 before a certain change was made. However, sometimes it can have two
2415 or more parent commits, in which case we call it a "merge", due to the
2416 fact that such a commit brings together ("merges") two or more
2417 previous states represented by other commits.
2419 In other words, while a "tree" represents a particular directory state
2420 of a working directory, a "commit" represents that state in "time",
2421 and explains how we got there.
2423 You create a commit object by giving it the tree that describes the
2424 state at the time of the commit, and a list of parents:
2426 -------------------------------------------------
2427 $ git-commit-tree <tree> -p <parent> [-p <parent2> ..]
2428 -------------------------------------------------
2430 and then giving the reason for the commit on stdin (either through
2431 redirection from a pipe or file, or by just typing it at the tty).
2433 git-commit-tree will return the name of the object that represents
2434 that commit, and you should save it away for later use. Normally,
2435 you'd commit a new `HEAD` state, and while git doesn't care where you
2436 save the note about that state, in practice we tend to just write the
2437 result to the file pointed at by `.git/HEAD`, so that we can always see
2438 what the last committed state was.
2440 Here is an ASCII art by Jon Loeliger that illustrates how
2441 various pieces fit together.
2469 checkout-index -u | | checkout-index
2483 You can examine the data represented in the object database and the
2484 index with various helper tools. For every object, you can use
2485 gitlink:git-cat-file[1] to examine details about the
2488 -------------------------------------------------
2489 $ git-cat-file -t <objectname>
2490 -------------------------------------------------
2492 shows the type of the object, and once you have the type (which is
2493 usually implicit in where you find the object), you can use
2495 -------------------------------------------------
2496 $ git-cat-file blob|tree|commit|tag <objectname>
2497 -------------------------------------------------
2499 to show its contents. NOTE! Trees have binary content, and as a result
2500 there is a special helper for showing that content, called
2501 `git-ls-tree`, which turns the binary content into a more easily
2504 It's especially instructive to look at "commit" objects, since those
2505 tend to be small and fairly self-explanatory. In particular, if you
2506 follow the convention of having the top commit name in `.git/HEAD`,
2509 -------------------------------------------------
2510 $ git-cat-file commit HEAD
2511 -------------------------------------------------
2513 to see what the top commit was.
2515 Merging multiple trees
2516 ----------------------
2518 Git helps you do a three-way merge, which you can expand to n-way by
2519 repeating the merge procedure arbitrary times until you finally
2520 "commit" the state. The normal situation is that you'd only do one
2521 three-way merge (two parents), and commit it, but if you like to, you
2522 can do multiple parents in one go.
2524 To do a three-way merge, you need the two sets of "commit" objects
2525 that you want to merge, use those to find the closest common parent (a
2526 third "commit" object), and then use those commit objects to find the
2527 state of the directory ("tree" object) at these points.
2529 To get the "base" for the merge, you first look up the common parent
2532 -------------------------------------------------
2533 $ git-merge-base <commit1> <commit2>
2534 -------------------------------------------------
2536 which will return you the commit they are both based on. You should
2537 now look up the "tree" objects of those commits, which you can easily
2538 do with (for example)
2540 -------------------------------------------------
2541 $ git-cat-file commit <commitname> | head -1
2542 -------------------------------------------------
2544 since the tree object information is always the first line in a commit
2547 Once you know the three trees you are going to merge (the one
2548 "original" tree, aka the common case, and the two "result" trees, aka
2549 the branches you want to merge), you do a "merge" read into the
2550 index. This will complain if it has to throw away your old index contents, so you should
2551 make sure that you've committed those - in fact you would normally
2552 always do a merge against your last commit (which should thus match
2553 what you have in your current index anyway).
2557 -------------------------------------------------
2558 $ git-read-tree -m -u <origtree> <yourtree> <targettree>
2559 -------------------------------------------------
2561 which will do all trivial merge operations for you directly in the
2562 index file, and you can just write the result out with
2566 Merging multiple trees, continued
2567 ---------------------------------
2569 Sadly, many merges aren't trivial. If there are files that have
2570 been added.moved or removed, or if both branches have modified the
2571 same file, you will be left with an index tree that contains "merge
2572 entries" in it. Such an index tree can 'NOT' be written out to a tree
2573 object, and you will have to resolve any such merge clashes using
2574 other tools before you can write out the result.
2576 You can examine such index state with `git-ls-files --unmerged`
2577 command. An example:
2579 ------------------------------------------------
2580 $ git-read-tree -m $orig HEAD $target
2581 $ git-ls-files --unmerged
2582 100644 263414f423d0e4d70dae8fe53fa34614ff3e2860 1 hello.c
2583 100644 06fa6a24256dc7e560efa5687fa84b51f0263c3a 2 hello.c
2584 100644 cc44c73eb783565da5831b4d820c962954019b69 3 hello.c
2585 ------------------------------------------------
2587 Each line of the `git-ls-files --unmerged` output begins with
2588 the blob mode bits, blob SHA1, 'stage number', and the
2589 filename. The 'stage number' is git's way to say which tree it
2590 came from: stage 1 corresponds to `$orig` tree, stage 2 `HEAD`
2591 tree, and stage3 `$target` tree.
2593 Earlier we said that trivial merges are done inside
2594 `git-read-tree -m`. For example, if the file did not change
2595 from `$orig` to `HEAD` nor `$target`, or if the file changed
2596 from `$orig` to `HEAD` and `$orig` to `$target` the same way,
2597 obviously the final outcome is what is in `HEAD`. What the
2598 above example shows is that file `hello.c` was changed from
2599 `$orig` to `HEAD` and `$orig` to `$target` in a different way.
2600 You could resolve this by running your favorite 3-way merge
2601 program, e.g. `diff3` or `merge`, on the blob objects from
2602 these three stages yourself, like this:
2604 ------------------------------------------------
2605 $ git-cat-file blob 263414f... >hello.c~1
2606 $ git-cat-file blob 06fa6a2... >hello.c~2
2607 $ git-cat-file blob cc44c73... >hello.c~3
2608 $ merge hello.c~2 hello.c~1 hello.c~3
2609 ------------------------------------------------
2611 This would leave the merge result in `hello.c~2` file, along
2612 with conflict markers if there are conflicts. After verifying
2613 the merge result makes sense, you can tell git what the final
2614 merge result for this file is by:
2616 -------------------------------------------------
2617 $ mv -f hello.c~2 hello.c
2618 $ git-update-index hello.c
2619 -------------------------------------------------
2621 When a path is in unmerged state, running `git-update-index` for
2622 that path tells git to mark the path resolved.
2624 The above is the description of a git merge at the lowest level,
2625 to help you understand what conceptually happens under the hood.
2626 In practice, nobody, not even git itself, uses three `git-cat-file`
2627 for this. There is `git-merge-index` program that extracts the
2628 stages to temporary files and calls a "merge" script on it:
2630 -------------------------------------------------
2631 $ git-merge-index git-merge-one-file hello.c
2632 -------------------------------------------------
2634 and that is what higher level `git resolve` is implemented with.
2636 How git stores objects efficiently: pack files
2637 ----------------------------------------------
2639 We've seen how git stores each object in a file named after the
2642 Unfortunately this system becomes inefficient once a project has a
2643 lot of objects. Try this on an old project:
2645 ------------------------------------------------
2647 6930 objects, 47620 kilobytes
2648 ------------------------------------------------
2650 The first number is the number of objects which are kept in
2651 individual files. The second is the amount of space taken up by
2652 those "loose" objects.
2654 You can save space and make git faster by moving these loose objects in
2655 to a "pack file", which stores a group of objects in an efficient
2656 compressed format; the details of how pack files are formatted can be
2657 found in link:technical/pack-format.txt[technical/pack-format.txt].
2659 To put the loose objects into a pack, just run git repack:
2661 ------------------------------------------------
2664 Done counting 6020 objects.
2665 Deltifying 6020 objects.
2666 100% (6020/6020) done
2667 Writing 6020 objects.
2668 100% (6020/6020) done
2669 Total 6020, written 6020 (delta 4070), reused 0 (delta 0)
2670 Pack pack-3e54ad29d5b2e05838c75df582c65257b8d08e1c created.
2671 ------------------------------------------------
2675 ------------------------------------------------
2677 ------------------------------------------------
2679 to remove any of the "loose" objects that are now contained in the
2680 pack. This will also remove any unreferenced objects (which may be
2681 created when, for example, you use "git reset" to remove a commit).
2682 You can verify that the loose objects are gone by looking at the
2683 .git/objects directory or by running
2685 ------------------------------------------------
2687 0 objects, 0 kilobytes
2688 ------------------------------------------------
2690 Although the object files are gone, any commands that refer to those
2691 objects will work exactly as they did before.
2693 The gitlink:git-gc[1] command performs packing, pruning, and more for
2694 you, so is normally the only high-level command you need.
2696 Glossary of git terms
2697 =====================
2699 include::glossary.txt[]
2701 Notes and todo list for this manual
2702 ===================================
2704 This is a work in progress.
2706 The basic requirements:
2707 - It must be readable in order, from beginning to end, by
2708 someone intelligent with a basic grasp of the unix
2709 commandline, but without any special knowledge of git. If
2710 necessary, any other prerequisites should be specifically
2711 mentioned as they arise.
2712 - Whenever possible, section headings should clearly describe
2713 the task they explain how to do, in language that requires
2714 no more knowledge than necessary: for example, "importing
2715 patches into a project" rather than "the git-am command"
2717 Think about how to create a clear chapter dependency graph that will
2718 allow people to get to important topics without necessarily reading
2719 everything in between.
2721 Scan Documentation/ for other stuff left out; in particular:
2728 Scan email archives for other stuff left out
2730 Scan man pages to see if any assume more background than this manual
2733 Simplify beginning by suggesting disconnected head instead of
2734 temporary branch creation?
2736 Explain how to refer to file stages in the "how to resolve a merge"
2737 section: diff -1, -2, -3, --ours, --theirs :1:/path notation. The
2738 "git ls-files --unmerged --stage" thing is sorta useful too,
2739 actually. And note gitk --merge.
2741 Add more good examples. Entire sections of just cookbook examples
2742 might be a good idea; maybe make an "advanced examples" section a
2743 standard end-of-chapter section?
2745 Include cross-references to the glossary, where appropriate.
2748 reflogs, git reflog expire
2749 shallow clones?? See draft 1.5.0 release notes for some documentation.