1Git User Manual 2=============== 3 4Git is a fast distributed revision control system. 5 6This manual is designed to be readable by someone with basic UNIX 7command-line skills, but no previous knowledge of Git. 8 9<<repositories-and-branches>> and <<exploring-git-history>> explain how 10to fetch and study a project using git--read these chapters to learn how 11to build and test a particular version of a software project, search for 12regressions, and so on. 13 14People needing to do actual development will also want to read 15<<Developing-With-git>> and <<sharing-development>>. 16 17Further chapters cover more specialized topics. 18 19Comprehensive reference documentation is available through the man 20pages, or linkgit:git-help[1] command. For example, for the command 21`git clone <repo>`, you can either use: 22 23------------------------------------------------ 24$ man git-clone 25------------------------------------------------ 26 27or: 28 29------------------------------------------------ 30$ git help clone 31------------------------------------------------ 32 33With the latter, you can use the manual viewer of your choice; see 34linkgit:git-help[1] for more information. 35 36See also <<git-quick-start>> for a brief overview of Git commands, 37without any explanation. 38 39Finally, see <<todo>> for ways that you can help make this manual more 40complete. 41 42 43[[repositories-and-branches]] 44Repositories and Branches 45========================= 46 47[[how-to-get-a-git-repository]] 48How to get a Git repository 49--------------------------- 50 51It will be useful to have a Git repository to experiment with as you 52read this manual. 53 54The best way to get one is by using the linkgit:git-clone[1] command to 55download a copy of an existing repository. If you don't already have a 56project in mind, here are some interesting examples: 57 58------------------------------------------------ 59 # Git itself (approx. 40MB download): 60$ git clone git://git.kernel.org/pub/scm/git/git.git 61 # the Linux kernel (approx. 640MB download): 62$ git clone git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git 63------------------------------------------------ 64 65The initial clone may be time-consuming for a large project, but you 66will only need to clone once. 67 68The clone command creates a new directory named after the project 69(`git` or `linux` in the examples above). After you cd into this 70directory, you will see that it contains a copy of the project files, 71called the <<def_working_tree,working tree>>, together with a special 72top-level directory named `.git`, which contains all the information 73about the history of the project. 74 75[[how-to-check-out]] 76How to check out a different version of a project 77------------------------------------------------- 78 79Git is best thought of as a tool for storing the history of a collection 80of files. It stores the history as a compressed collection of 81interrelated snapshots of the project's contents. In Git each such 82version is called a <<def_commit,commit>>. 83 84Those snapshots aren't necessarily all arranged in a single line from 85oldest to newest; instead, work may simultaneously proceed along 86parallel lines of development, called <<def_branch,branches>>, which may 87merge and diverge. 88 89A single Git repository can track development on multiple branches. It 90does this by keeping a list of <<def_head,heads>> which reference the 91latest commit on each branch; the linkgit:git-branch[1] command shows 92you the list of branch heads: 93 94------------------------------------------------ 95$ git branch 96* master 97------------------------------------------------ 98 99A freshly cloned repository contains a single branch head, by default 100named "master", with the working directory initialized to the state of 101the project referred to by that branch head. 102 103Most projects also use <<def_tag,tags>>. Tags, like heads, are 104references into the project's history, and can be listed using the 105linkgit:git-tag[1] command: 106 107------------------------------------------------ 108$ git tag -l 109v2.6.11 110v2.6.11-tree 111v2.6.12 112v2.6.12-rc2 113v2.6.12-rc3 114v2.6.12-rc4 115v2.6.12-rc5 116v2.6.12-rc6 117v2.6.13 118... 119------------------------------------------------ 120 121Tags are expected to always point at the same version of a project, 122while heads are expected to advance as development progresses. 123 124Create a new branch head pointing to one of these versions and check it 125out using linkgit:git-checkout[1]: 126 127------------------------------------------------ 128$ git checkout -b new v2.6.13 129------------------------------------------------ 130 131The working directory then reflects the contents that the project had 132when it was tagged v2.6.13, and linkgit:git-branch[1] shows two 133branches, with an asterisk marking the currently checked-out branch: 134 135------------------------------------------------ 136$ git branch 137 master 138* new 139------------------------------------------------ 140 141If you decide that you'd rather see version 2.6.17, you can modify 142the current branch to point at v2.6.17 instead, with 143 144------------------------------------------------ 145$ git reset --hard v2.6.17 146------------------------------------------------ 147 148Note that if the current branch head was your only reference to a 149particular point in history, then resetting that branch may leave you 150with no way to find the history it used to point to; so use this command 151carefully. 152 153[[understanding-commits]] 154Understanding History: Commits 155------------------------------ 156 157Every change in the history of a project is represented by a commit. 158The linkgit:git-show[1] command shows the most recent commit on the 159current branch: 160 161------------------------------------------------ 162$ git show 163commit 17cf781661e6d38f737f15f53ab552f1e95960d7 164Author: Linus Torvalds <torvalds@ppc970.osdl.org.(none)> 165Date: Tue Apr 19 14:11:06 2005 -0700 166 167 Remove duplicate getenv(DB_ENVIRONMENT) call 168 169 Noted by Tony Luck. 170 171diff --git a/init-db.c b/init-db.c 172index 65898fa..b002dc6 100644 173--- a/init-db.c 174+++ b/init-db.c 175@@ -7,7 +7,7 @@ 176 177 int main(int argc, char **argv) 178 { 179- char *sha1_dir = getenv(DB_ENVIRONMENT), *path; 180+ char *sha1_dir, *path; 181 int len, i; 182 183 if (mkdir(".git", 0755) < 0) { 184------------------------------------------------ 185 186As you can see, a commit shows who made the latest change, what they 187did, and why. 188 189Every commit has a 40-hexdigit id, sometimes called the "object name" or the 190"SHA-1 id", shown on the first line of the `git show` output. You can usually 191refer to a commit by a shorter name, such as a tag or a branch name, but this 192longer name can also be useful. Most importantly, it is a globally unique 193name for this commit: so if you tell somebody else the object name (for 194example in email), then you are guaranteed that name will refer to the same 195commit in their repository that it does in yours (assuming their repository 196has that commit at all). Since the object name is computed as a hash over the 197contents of the commit, you are guaranteed that the commit can never change 198without its name also changing. 199 200In fact, in <<git-concepts>> we shall see that everything stored in Git 201history, including file data and directory contents, is stored in an object 202with a name that is a hash of its contents. 203 204[[understanding-reachability]] 205Understanding history: commits, parents, and reachability 206~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 207 208Every commit (except the very first commit in a project) also has a 209parent commit which shows what happened before this commit. 210Following the chain of parents will eventually take you back to the 211beginning of the project. 212 213However, the commits do not form a simple list; Git allows lines of 214development to diverge and then reconverge, and the point where two 215lines of development reconverge is called a "merge". The commit 216representing a merge can therefore have more than one parent, with 217each parent representing the most recent commit on one of the lines 218of development leading to that point. 219 220The best way to see how this works is using the linkgit:gitk[1] 221command; running gitk now on a Git repository and looking for merge 222commits will help understand how Git organizes history. 223 224In the following, we say that commit X is "reachable" from commit Y 225if commit X is an ancestor of commit Y. Equivalently, you could say 226that Y is a descendant of X, or that there is a chain of parents 227leading from commit Y to commit X. 228 229[[history-diagrams]] 230Understanding history: History diagrams 231~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 232 233We will sometimes represent Git history using diagrams like the one 234below. Commits are shown as "o", and the links between them with 235lines drawn with - / and \. Time goes left to right: 236 237 238................................................ 239 o--o--o <-- Branch A 240 / 241 o--o--o <-- master 242 \ 243 o--o--o <-- Branch B 244................................................ 245 246If we need to talk about a particular commit, the character "o" may 247be replaced with another letter or number. 248 249[[what-is-a-branch]] 250Understanding history: What is a branch? 251~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 252 253When we need to be precise, we will use the word "branch" to mean a line 254of development, and "branch head" (or just "head") to mean a reference 255to the most recent commit on a branch. In the example above, the branch 256head named "A" is a pointer to one particular commit, but we refer to 257the line of three commits leading up to that point as all being part of 258"branch A". 259 260However, when no confusion will result, we often just use the term 261"branch" both for branches and for branch heads. 262 263[[manipulating-branches]] 264Manipulating branches 265--------------------- 266 267Creating, deleting, and modifying branches is quick and easy; here's 268a summary of the commands: 269 270`git branch`:: 271 list all branches. 272`git branch <branch>`:: 273 create a new branch named `<branch>`, referencing the same 274 point in history as the current branch. 275`git branch <branch> <start-point>`:: 276 create a new branch named `<branch>`, referencing 277 `<start-point>`, which may be specified any way you like, 278 including using a branch name or a tag name. 279`git branch -d <branch>`:: 280 delete the branch `<branch>`; if the branch is not fully 281 merged in its upstream branch or contained in the current branch, 282 this command will fail with a warning. 283`git branch -D <branch>`:: 284 delete the branch `<branch>` irrespective of its merged status. 285`git checkout <branch>`:: 286 make the current branch `<branch>`, updating the working 287 directory to reflect the version referenced by `<branch>`. 288`git checkout -b <new> <start-point>`:: 289 create a new branch `<new>` referencing `<start-point>`, and 290 check it out. 291 292The special symbol "HEAD" can always be used to refer to the current 293branch. In fact, Git uses a file named `HEAD` in the `.git` directory 294to remember which branch is current: 295 296------------------------------------------------ 297$ cat .git/HEAD 298ref: refs/heads/master 299------------------------------------------------ 300 301[[detached-head]] 302Examining an old version without creating a new branch 303------------------------------------------------------ 304 305The `git checkout` command normally expects a branch head, but will also 306accept an arbitrary commit; for example, you can check out the commit 307referenced by a tag: 308 309------------------------------------------------ 310$ git checkout v2.6.17 311Note: checking out 'v2.6.17'. 312 313You are in 'detached HEAD' state. You can look around, make experimental 314changes and commit them, and you can discard any commits you make in this 315state without impacting any branches by performing another checkout. 316 317If you want to create a new branch to retain commits you create, you may 318do so (now or later) by using -b with the checkout command again. Example: 319 320 git checkout -b new_branch_name 321 322HEAD is now at 427abfa Linux v2.6.17 323------------------------------------------------ 324 325The HEAD then refers to the SHA-1 of the commit instead of to a branch, 326and git branch shows that you are no longer on a branch: 327 328------------------------------------------------ 329$ cat .git/HEAD 330427abfa28afedffadfca9dd8b067eb6d36bac53f 331$ git branch 332* (detached from v2.6.17) 333 master 334------------------------------------------------ 335 336In this case we say that the HEAD is "detached". 337 338This is an easy way to check out a particular version without having to 339make up a name for the new branch. You can still create a new branch 340(or tag) for this version later if you decide to. 341 342[[examining-remote-branches]] 343Examining branches from a remote repository 344------------------------------------------- 345 346The "master" branch that was created at the time you cloned is a copy 347of the HEAD in the repository that you cloned from. That repository 348may also have had other branches, though, and your local repository 349keeps branches which track each of those remote branches, called 350remote-tracking branches, which you 351can view using the `-r` option to linkgit:git-branch[1]: 352 353------------------------------------------------ 354$ git branch -r 355 origin/HEAD 356 origin/html 357 origin/maint 358 origin/man 359 origin/master 360 origin/next 361 origin/pu 362 origin/todo 363------------------------------------------------ 364 365In this example, "origin" is called a remote repository, or "remote" 366for short. The branches of this repository are called "remote 367branches" from our point of view. The remote-tracking branches listed 368above were created based on the remote branches at clone time and will 369be updated by `git fetch` (hence `git pull`) and `git push`. See 370<<Updating-a-repository-With-git-fetch>> for details. 371 372You might want to build on one of these remote-tracking branches 373on a branch of your own, just as you would for a tag: 374 375------------------------------------------------ 376$ git checkout -b my-todo-copy origin/todo 377------------------------------------------------ 378 379You can also check out `origin/todo` directly to examine it or 380write a one-off patch. See <<detached-head,detached head>>. 381 382Note that the name "origin" is just the name that Git uses by default 383to refer to the repository that you cloned from. 384 385[[how-git-stores-references]] 386Naming branches, tags, and other references 387------------------------------------------- 388 389Branches, remote-tracking branches, and tags are all references to 390commits. All references are named with a slash-separated path name 391starting with `refs`; the names we've been using so far are actually 392shorthand: 393 394 - The branch `test` is short for `refs/heads/test`. 395 - The tag `v2.6.18` is short for `refs/tags/v2.6.18`. 396 - `origin/master` is short for `refs/remotes/origin/master`. 397 398The full name is occasionally useful if, for example, there ever 399exists a tag and a branch with the same name. 400 401(Newly created refs are actually stored in the `.git/refs` directory, 402under the path given by their name. However, for efficiency reasons 403they may also be packed together in a single file; see 404linkgit:git-pack-refs[1]). 405 406As another useful shortcut, the "HEAD" of a repository can be referred 407to just using the name of that repository. So, for example, "origin" 408is usually a shortcut for the HEAD branch in the repository "origin". 409 410For the complete list of paths which Git checks for references, and 411the order it uses to decide which to choose when there are multiple 412references with the same shorthand name, see the "SPECIFYING 413REVISIONS" section of linkgit:gitrevisions[7]. 414 415[[Updating-a-repository-With-git-fetch]] 416Updating a repository with git fetch 417------------------------------------ 418 419After you clone a repository and commit a few changes of your own, you 420may wish to check the original repository for updates. 421 422The `git-fetch` command, with no arguments, will update all of the 423remote-tracking branches to the latest version found in the original 424repository. It will not touch any of your own branches--not even the 425"master" branch that was created for you on clone. 426 427[[fetching-branches]] 428Fetching branches from other repositories 429----------------------------------------- 430 431You can also track branches from repositories other than the one you 432cloned from, using linkgit:git-remote[1]: 433 434------------------------------------------------- 435$ git remote add staging git://git.kernel.org/.../gregkh/staging.git 436$ git fetch staging 437... 438From git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/staging 439 * [new branch] master -> staging/master 440 * [new branch] staging-linus -> staging/staging-linus 441 * [new branch] staging-next -> staging/staging-next 442------------------------------------------------- 443 444New remote-tracking branches will be stored under the shorthand name 445that you gave `git remote add`, in this case `staging`: 446 447------------------------------------------------- 448$ git branch -r 449 origin/HEAD -> origin/master 450 origin/master 451 staging/master 452 staging/staging-linus 453 staging/staging-next 454------------------------------------------------- 455 456If you run `git fetch <remote>` later, the remote-tracking branches 457for the named `<remote>` will be updated. 458 459If you examine the file `.git/config`, you will see that Git has added 460a new stanza: 461 462------------------------------------------------- 463$ cat .git/config 464... 465[remote "staging"] 466 url = git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/staging.git 467 fetch = +refs/heads/*:refs/remotes/staging/* 468... 469------------------------------------------------- 470 471This is what causes Git to track the remote's branches; you may modify 472or delete these configuration options by editing `.git/config` with a 473text editor. (See the "CONFIGURATION FILE" section of 474linkgit:git-config[1] for details.) 475 476[[exploring-git-history]] 477Exploring Git history 478===================== 479 480Git is best thought of as a tool for storing the history of a 481collection of files. It does this by storing compressed snapshots of 482the contents of a file hierarchy, together with "commits" which show 483the relationships between these snapshots. 484 485Git provides extremely flexible and fast tools for exploring the 486history of a project. 487 488We start with one specialized tool that is useful for finding the 489commit that introduced a bug into a project. 490 491[[using-bisect]] 492How to use bisect to find a regression 493-------------------------------------- 494 495Suppose version 2.6.18 of your project worked, but the version at 496"master" crashes. Sometimes the best way to find the cause of such a 497regression is to perform a brute-force search through the project's 498history to find the particular commit that caused the problem. The 499linkgit:git-bisect[1] command can help you do this: 500 501------------------------------------------------- 502$ git bisect start 503$ git bisect good v2.6.18 504$ git bisect bad master 505Bisecting: 3537 revisions left to test after this 506[65934a9a028b88e83e2b0f8b36618fe503349f8e] BLOCK: Make USB storage depend on SCSI rather than selecting it [try #6] 507------------------------------------------------- 508 509If you run `git branch` at this point, you'll see that Git has 510temporarily moved you in "(no branch)". HEAD is now detached from any 511branch and points directly to a commit (with commit id 65934) that 512is reachable from "master" but not from v2.6.18. Compile and test it, 513and see whether it crashes. Assume it does crash. Then: 514 515------------------------------------------------- 516$ git bisect bad 517Bisecting: 1769 revisions left to test after this 518[7eff82c8b1511017ae605f0c99ac275a7e21b867] i2c-core: Drop useless bitmaskings 519------------------------------------------------- 520 521checks out an older version. Continue like this, telling Git at each 522stage whether the version it gives you is good or bad, and notice 523that the number of revisions left to test is cut approximately in 524half each time. 525 526After about 13 tests (in this case), it will output the commit id of 527the guilty commit. You can then examine the commit with 528linkgit:git-show[1], find out who wrote it, and mail them your bug 529report with the commit id. Finally, run 530 531------------------------------------------------- 532$ git bisect reset 533------------------------------------------------- 534 535to return you to the branch you were on before. 536 537Note that the version which `git bisect` checks out for you at each 538point is just a suggestion, and you're free to try a different 539version if you think it would be a good idea. For example, 540occasionally you may land on a commit that broke something unrelated; 541run 542 543------------------------------------------------- 544$ git bisect visualize 545------------------------------------------------- 546 547which will run gitk and label the commit it chose with a marker that 548says "bisect". Choose a safe-looking commit nearby, note its commit 549id, and check it out with: 550 551------------------------------------------------- 552$ git reset --hard fb47ddb2db 553------------------------------------------------- 554 555then test, run `bisect good` or `bisect bad` as appropriate, and 556continue. 557 558Instead of `git bisect visualize` and then `git reset --hard 559fb47ddb2db`, you might just want to tell Git that you want to skip 560the current commit: 561 562------------------------------------------------- 563$ git bisect skip 564------------------------------------------------- 565 566In this case, though, Git may not eventually be able to tell the first 567bad one between some first skipped commits and a later bad commit. 568 569There are also ways to automate the bisecting process if you have a 570test script that can tell a good from a bad commit. See 571linkgit:git-bisect[1] for more information about this and other `git 572bisect` features. 573 574[[naming-commits]] 575Naming commits 576-------------- 577 578We have seen several ways of naming commits already: 579 580 - 40-hexdigit object name 581 - branch name: refers to the commit at the head of the given 582 branch 583 - tag name: refers to the commit pointed to by the given tag 584 (we've seen branches and tags are special cases of 585 <<how-git-stores-references,references>>). 586 - HEAD: refers to the head of the current branch 587 588There are many more; see the "SPECIFYING REVISIONS" section of the 589linkgit:gitrevisions[7] man page for the complete list of ways to 590name revisions. Some examples: 591 592------------------------------------------------- 593$ git show fb47ddb2 # the first few characters of the object name 594 # are usually enough to specify it uniquely 595$ git show HEAD^ # the parent of the HEAD commit 596$ git show HEAD^^ # the grandparent 597$ git show HEAD~4 # the great-great-grandparent 598------------------------------------------------- 599 600Recall that merge commits may have more than one parent; by default, 601`^` and `~` follow the first parent listed in the commit, but you can 602also choose: 603 604------------------------------------------------- 605$ git show HEAD^1 # show the first parent of HEAD 606$ git show HEAD^2 # show the second parent of HEAD 607------------------------------------------------- 608 609In addition to HEAD, there are several other special names for 610commits: 611 612Merges (to be discussed later), as well as operations such as 613`git reset`, which change the currently checked-out commit, generally 614set ORIG_HEAD to the value HEAD had before the current operation. 615 616The `git fetch` operation always stores the head of the last fetched 617branch in FETCH_HEAD. For example, if you run `git fetch` without 618specifying a local branch as the target of the operation 619 620------------------------------------------------- 621$ git fetch git://example.com/proj.git theirbranch 622------------------------------------------------- 623 624the fetched commits will still be available from FETCH_HEAD. 625 626When we discuss merges we'll also see the special name MERGE_HEAD, 627which refers to the other branch that we're merging in to the current 628branch. 629 630The linkgit:git-rev-parse[1] command is a low-level command that is 631occasionally useful for translating some name for a commit to the object 632name for that commit: 633 634------------------------------------------------- 635$ git rev-parse origin 636e05db0fd4f31dde7005f075a84f96b360d05984b 637------------------------------------------------- 638 639[[creating-tags]] 640Creating tags 641------------- 642 643We can also create a tag to refer to a particular commit; after 644running 645 646------------------------------------------------- 647$ git tag stable-1 1b2e1d63ff 648------------------------------------------------- 649 650You can use `stable-1` to refer to the commit 1b2e1d63ff. 651 652This creates a "lightweight" tag. If you would also like to include a 653comment with the tag, and possibly sign it cryptographically, then you 654should create a tag object instead; see the linkgit:git-tag[1] man page 655for details. 656 657[[browsing-revisions]] 658Browsing revisions 659------------------ 660 661The linkgit:git-log[1] command can show lists of commits. On its 662own, it shows all commits reachable from the parent commit; but you 663can also make more specific requests: 664 665------------------------------------------------- 666$ git log v2.5.. # commits since (not reachable from) v2.5 667$ git log test..master # commits reachable from master but not test 668$ git log master..test # ...reachable from test but not master 669$ git log master...test # ...reachable from either test or master, 670 # but not both 671$ git log --since="2 weeks ago" # commits from the last 2 weeks 672$ git log Makefile # commits which modify Makefile 673$ git log fs/ # ... which modify any file under fs/ 674$ git log -S'foo()' # commits which add or remove any file data 675 # matching the string 'foo()' 676------------------------------------------------- 677 678And of course you can combine all of these; the following finds 679commits since v2.5 which touch the `Makefile` or any file under `fs`: 680 681------------------------------------------------- 682$ git log v2.5.. Makefile fs/ 683------------------------------------------------- 684 685You can also ask git log to show patches: 686 687------------------------------------------------- 688$ git log -p 689------------------------------------------------- 690 691See the `--pretty` option in the linkgit:git-log[1] man page for more 692display options. 693 694Note that git log starts with the most recent commit and works 695backwards through the parents; however, since Git history can contain 696multiple independent lines of development, the particular order that 697commits are listed in may be somewhat arbitrary. 698 699[[generating-diffs]] 700Generating diffs 701---------------- 702 703You can generate diffs between any two versions using 704linkgit:git-diff[1]: 705 706------------------------------------------------- 707$ git diff master..test 708------------------------------------------------- 709 710That will produce the diff between the tips of the two branches. If 711you'd prefer to find the diff from their common ancestor to test, you 712can use three dots instead of two: 713 714------------------------------------------------- 715$ git diff master...test 716------------------------------------------------- 717 718Sometimes what you want instead is a set of patches; for this you can 719use linkgit:git-format-patch[1]: 720 721------------------------------------------------- 722$ git format-patch master..test 723------------------------------------------------- 724 725will generate a file with a patch for each commit reachable from test 726but not from master. 727 728[[viewing-old-file-versions]] 729Viewing old file versions 730------------------------- 731 732You can always view an old version of a file by just checking out the 733correct revision first. But sometimes it is more convenient to be 734able to view an old version of a single file without checking 735anything out; this command does that: 736 737------------------------------------------------- 738$ git show v2.5:fs/locks.c 739------------------------------------------------- 740 741Before the colon may be anything that names a commit, and after it 742may be any path to a file tracked by Git. 743 744[[history-examples]] 745Examples 746-------- 747 748[[counting-commits-on-a-branch]] 749Counting the number of commits on a branch 750~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 751 752Suppose you want to know how many commits you've made on `mybranch` 753since it diverged from `origin`: 754 755------------------------------------------------- 756$ git log --pretty=oneline origin..mybranch | wc -l 757------------------------------------------------- 758 759Alternatively, you may often see this sort of thing done with the 760lower-level command linkgit:git-rev-list[1], which just lists the SHA-1's 761of all the given commits: 762 763------------------------------------------------- 764$ git rev-list origin..mybranch | wc -l 765------------------------------------------------- 766 767[[checking-for-equal-branches]] 768Check whether two branches point at the same history 769~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 770 771Suppose you want to check whether two branches point at the same point 772in history. 773 774------------------------------------------------- 775$ git diff origin..master 776------------------------------------------------- 777 778will tell you whether the contents of the project are the same at the 779two branches; in theory, however, it's possible that the same project 780contents could have been arrived at by two different historical 781routes. You could compare the object names: 782 783------------------------------------------------- 784$ git rev-list origin 785e05db0fd4f31dde7005f075a84f96b360d05984b 786$ git rev-list master 787e05db0fd4f31dde7005f075a84f96b360d05984b 788------------------------------------------------- 789 790Or you could recall that the `...` operator selects all commits 791reachable from either one reference or the other but not 792both; so 793 794------------------------------------------------- 795$ git log origin...master 796------------------------------------------------- 797 798will return no commits when the two branches are equal. 799 800[[finding-tagged-descendants]] 801Find first tagged version including a given fix 802~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 803 804Suppose you know that the commit e05db0fd fixed a certain problem. 805You'd like to find the earliest tagged release that contains that 806fix. 807 808Of course, there may be more than one answer--if the history branched 809after commit e05db0fd, then there could be multiple "earliest" tagged 810releases. 811 812You could just visually inspect the commits since e05db0fd: 813 814------------------------------------------------- 815$ gitk e05db0fd.. 816------------------------------------------------- 817 818or you can use linkgit:git-name-rev[1], which will give the commit a 819name based on any tag it finds pointing to one of the commit's 820descendants: 821 822------------------------------------------------- 823$ git name-rev --tags e05db0fd 824e05db0fd tags/v1.5.0-rc1^0~23 825------------------------------------------------- 826 827The linkgit:git-describe[1] command does the opposite, naming the 828revision using a tag on which the given commit is based: 829 830------------------------------------------------- 831$ git describe e05db0fd 832v1.5.0-rc0-260-ge05db0f 833------------------------------------------------- 834 835but that may sometimes help you guess which tags might come after the 836given commit. 837 838If you just want to verify whether a given tagged version contains a 839given commit, you could use linkgit:git-merge-base[1]: 840 841------------------------------------------------- 842$ git merge-base e05db0fd v1.5.0-rc1 843e05db0fd4f31dde7005f075a84f96b360d05984b 844------------------------------------------------- 845 846The merge-base command finds a common ancestor of the given commits, 847and always returns one or the other in the case where one is a 848descendant of the other; so the above output shows that e05db0fd 849actually is an ancestor of v1.5.0-rc1. 850 851Alternatively, note that 852 853------------------------------------------------- 854$ git log v1.5.0-rc1..e05db0fd 855------------------------------------------------- 856 857will produce empty output if and only if v1.5.0-rc1 includes e05db0fd, 858because it outputs only commits that are not reachable from v1.5.0-rc1. 859 860As yet another alternative, the linkgit:git-show-branch[1] command lists 861the commits reachable from its arguments with a display on the left-hand 862side that indicates which arguments that commit is reachable from. 863So, if you run something like 864 865------------------------------------------------- 866$ git show-branch e05db0fd v1.5.0-rc0 v1.5.0-rc1 v1.5.0-rc2 867! [e05db0fd] Fix warnings in sha1_file.c - use C99 printf format if 868available 869 ! [v1.5.0-rc0] GIT v1.5.0 preview 870 ! [v1.5.0-rc1] GIT v1.5.0-rc1 871 ! [v1.5.0-rc2] GIT v1.5.0-rc2 872... 873------------------------------------------------- 874 875then a line like 876 877------------------------------------------------- 878+ ++ [e05db0fd] Fix warnings in sha1_file.c - use C99 printf format if 879available 880------------------------------------------------- 881 882shows that e05db0fd is reachable from itself, from v1.5.0-rc1, 883and from v1.5.0-rc2, and not from v1.5.0-rc0. 884 885[[showing-commits-unique-to-a-branch]] 886Showing commits unique to a given branch 887~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 888 889Suppose you would like to see all the commits reachable from the branch 890head named `master` but not from any other head in your repository. 891 892We can list all the heads in this repository with 893linkgit:git-show-ref[1]: 894 895------------------------------------------------- 896$ git show-ref --heads 897bf62196b5e363d73353a9dcf094c59595f3153b7 refs/heads/core-tutorial 898db768d5504c1bb46f63ee9d6e1772bd047e05bf9 refs/heads/maint 899a07157ac624b2524a059a3414e99f6f44bebc1e7 refs/heads/master 90024dbc180ea14dc1aebe09f14c8ecf32010690627 refs/heads/tutorial-2 9011e87486ae06626c2f31eaa63d26fc0fd646c8af2 refs/heads/tutorial-fixes 902------------------------------------------------- 903 904We can get just the branch-head names, and remove `master`, with 905the help of the standard utilities cut and grep: 906 907------------------------------------------------- 908$ git show-ref --heads | cut -d' ' -f2 | grep -v '^refs/heads/master' 909refs/heads/core-tutorial 910refs/heads/maint 911refs/heads/tutorial-2 912refs/heads/tutorial-fixes 913------------------------------------------------- 914 915And then we can ask to see all the commits reachable from master 916but not from these other heads: 917 918------------------------------------------------- 919$ gitk master --not $( git show-ref --heads | cut -d' ' -f2 | 920 grep -v '^refs/heads/master' ) 921------------------------------------------------- 922 923Obviously, endless variations are possible; for example, to see all 924commits reachable from some head but not from any tag in the repository: 925 926------------------------------------------------- 927$ gitk $( git show-ref --heads ) --not $( git show-ref --tags ) 928------------------------------------------------- 929 930(See linkgit:gitrevisions[7] for explanations of commit-selecting 931syntax such as `--not`.) 932 933[[making-a-release]] 934Creating a changelog and tarball for a software release 935~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 936 937The linkgit:git-archive[1] command can create a tar or zip archive from 938any version of a project; for example: 939 940------------------------------------------------- 941$ git archive -o latest.tar.gz --prefix=project/ HEAD 942------------------------------------------------- 943 944will use HEAD to produce a gzipped tar archive in which each filename 945is preceded by `project/`. The output file format is inferred from 946the output file extension if possible, see linkgit:git-archive[1] for 947details. 948 949Versions of Git older than 1.7.7 don't know about the `tar.gz` format, 950you'll need to use gzip explicitly: 951 952------------------------------------------------- 953$ git archive --format=tar --prefix=project/ HEAD | gzip >latest.tar.gz 954------------------------------------------------- 955 956If you're releasing a new version of a software project, you may want 957to simultaneously make a changelog to include in the release 958announcement. 959 960Linus Torvalds, for example, makes new kernel releases by tagging them, 961then running: 962 963------------------------------------------------- 964$ release-script 2.6.12 2.6.13-rc6 2.6.13-rc7 965------------------------------------------------- 966 967where release-script is a shell script that looks like: 968 969------------------------------------------------- 970#!/bin/sh 971stable="$1" 972last="$2" 973new="$3" 974echo "# git tag v$new" 975echo "git archive --prefix=linux-$new/ v$new | gzip -9 > ../linux-$new.tar.gz" 976echo "git diff v$stable v$new | gzip -9 > ../patch-$new.gz" 977echo "git log --no-merges v$new ^v$last > ../ChangeLog-$new" 978echo "git shortlog --no-merges v$new ^v$last > ../ShortLog" 979echo "git diff --stat --summary -M v$last v$new > ../diffstat-$new" 980------------------------------------------------- 981 982and then he just cut-and-pastes the output commands after verifying that 983they look OK. 984 985[[Finding-commits-With-given-Content]] 986Finding commits referencing a file with given content 987~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 988 989Somebody hands you a copy of a file, and asks which commits modified a 990file such that it contained the given content either before or after the 991commit. You can find out with this: 992 993------------------------------------------------- 994$ git log --raw --abbrev=40 --pretty=oneline | 995 grep -B 1 `git hash-object filename` 996------------------------------------------------- 997 998Figuring out why this works is left as an exercise to the (advanced) 999student. The linkgit:git-log[1], linkgit:git-diff-tree[1], and1000linkgit:git-hash-object[1] man pages may prove helpful.10011002[[Developing-With-git]]1003Developing with Git1004===================10051006[[telling-git-your-name]]1007Telling Git your name1008---------------------10091010Before creating any commits, you should introduce yourself to Git.1011The easiest way to do so is to use linkgit:git-config[1]:10121013------------------------------------------------1014$ git config --global user.name 'Your Name Comes Here'1015$ git config --global user.email 'you@yourdomain.example.com'1016------------------------------------------------10171018Which will add the following to a file named `.gitconfig` in your1019home directory:10201021------------------------------------------------1022[user]1023 name = Your Name Comes Here1024 email = you@yourdomain.example.com1025------------------------------------------------10261027See the "CONFIGURATION FILE" section of linkgit:git-config[1] for1028details on the configuration file. The file is plain text, so you can1029also edit it with your favorite editor.103010311032[[creating-a-new-repository]]1033Creating a new repository1034-------------------------10351036Creating a new repository from scratch is very easy:10371038-------------------------------------------------1039$ mkdir project1040$ cd project1041$ git init1042-------------------------------------------------10431044If you have some initial content (say, a tarball):10451046-------------------------------------------------1047$ tar xzvf project.tar.gz1048$ cd project1049$ git init1050$ git add . # include everything below ./ in the first commit:1051$ git commit1052-------------------------------------------------10531054[[how-to-make-a-commit]]1055How to make a commit1056--------------------10571058Creating a new commit takes three steps:10591060 1. Making some changes to the working directory using your1061 favorite editor.1062 2. Telling Git about your changes.1063 3. Creating the commit using the content you told Git about1064 in step 2.10651066In practice, you can interleave and repeat steps 1 and 2 as many1067times as you want: in order to keep track of what you want committed1068at step 3, Git maintains a snapshot of the tree's contents in a1069special staging area called "the index."10701071At the beginning, the content of the index will be identical to1072that of the HEAD. The command `git diff --cached`, which shows1073the difference between the HEAD and the index, should therefore1074produce no output at that point.10751076Modifying the index is easy:10771078To update the index with the contents of a new or modified file, use10791080-------------------------------------------------1081$ git add path/to/file1082-------------------------------------------------10831084To remove a file from the index and from the working tree, use10851086-------------------------------------------------1087$ git rm path/to/file1088-------------------------------------------------10891090After each step you can verify that10911092-------------------------------------------------1093$ git diff --cached1094-------------------------------------------------10951096always shows the difference between the HEAD and the index file--this1097is what you'd commit if you created the commit now--and that10981099-------------------------------------------------1100$ git diff1101-------------------------------------------------11021103shows the difference between the working tree and the index file.11041105Note that `git add` always adds just the current contents of a file1106to the index; further changes to the same file will be ignored unless1107you run `git add` on the file again.11081109When you're ready, just run11101111-------------------------------------------------1112$ git commit1113-------------------------------------------------11141115and Git will prompt you for a commit message and then create the new1116commit. Check to make sure it looks like what you expected with11171118-------------------------------------------------1119$ git show1120-------------------------------------------------11211122As a special shortcut,11231124-------------------------------------------------1125$ git commit -a1126-------------------------------------------------11271128will update the index with any files that you've modified or removed1129and create a commit, all in one step.11301131A number of commands are useful for keeping track of what you're1132about to commit:11331134-------------------------------------------------1135$ git diff --cached # difference between HEAD and the index; what1136 # would be committed if you ran "commit" now.1137$ git diff # difference between the index file and your1138 # working directory; changes that would not1139 # be included if you ran "commit" now.1140$ git diff HEAD # difference between HEAD and working tree; what1141 # would be committed if you ran "commit -a" now.1142$ git status # a brief per-file summary of the above.1143-------------------------------------------------11441145You can also use linkgit:git-gui[1] to create commits, view changes in1146the index and the working tree files, and individually select diff hunks1147for inclusion in the index (by right-clicking on the diff hunk and1148choosing "Stage Hunk For Commit").11491150[[creating-good-commit-messages]]1151Creating good commit messages1152-----------------------------11531154Though not required, it's a good idea to begin the commit message1155with a single short (less than 50 character) line summarizing the1156change, followed by a blank line and then a more thorough1157description. The text up to the first blank line in a commit1158message is treated as the commit title, and that title is used1159throughout Git. For example, linkgit:git-format-patch[1] turns a1160commit into email, and it uses the title on the Subject line and the1161rest of the commit in the body.116211631164[[ignoring-files]]1165Ignoring files1166--------------11671168A project will often generate files that you do 'not' want to track with Git.1169This typically includes files generated by a build process or temporary1170backup files made by your editor. Of course, 'not' tracking files with Git1171is just a matter of 'not' calling `git add` on them. But it quickly becomes1172annoying to have these untracked files lying around; e.g. they make1173`git add .` practically useless, and they keep showing up in the output of1174`git status`.11751176You can tell Git to ignore certain files by creating a file called1177`.gitignore` in the top level of your working directory, with contents1178such as:11791180-------------------------------------------------1181# Lines starting with '#' are considered comments.1182# Ignore any file named foo.txt.1183foo.txt1184# Ignore (generated) html files,1185*.html1186# except foo.html which is maintained by hand.1187!foo.html1188# Ignore objects and archives.1189*.[oa]1190-------------------------------------------------11911192See linkgit:gitignore[5] for a detailed explanation of the syntax. You can1193also place .gitignore files in other directories in your working tree, and they1194will apply to those directories and their subdirectories. The `.gitignore`1195files can be added to your repository like any other files (just run `git add1196.gitignore` and `git commit`, as usual), which is convenient when the exclude1197patterns (such as patterns matching build output files) would also make sense1198for other users who clone your repository.11991200If you wish the exclude patterns to affect only certain repositories1201(instead of every repository for a given project), you may instead put1202them in a file in your repository named `.git/info/exclude`, or in any1203file specified by the `core.excludesFile` configuration variable.1204Some Git commands can also take exclude patterns directly on the1205command line. See linkgit:gitignore[5] for the details.12061207[[how-to-merge]]1208How to merge1209------------12101211You can rejoin two diverging branches of development using1212linkgit:git-merge[1]:12131214-------------------------------------------------1215$ git merge branchname1216-------------------------------------------------12171218merges the development in the branch `branchname` into the current1219branch.12201221A merge is made by combining the changes made in `branchname` and the1222changes made up to the latest commit in your current branch since1223their histories forked. The work tree is overwritten by the result of1224the merge when this combining is done cleanly, or overwritten by a1225half-merged results when this combining results in conflicts.1226Therefore, if you have uncommitted changes touching the same files as1227the ones impacted by the merge, Git will refuse to proceed. Most of1228the time, you will want to commit your changes before you can merge,1229and if you don't, then linkgit:git-stash[1] can take these changes1230away while you're doing the merge, and reapply them afterwards.12311232If the changes are independent enough, Git will automatically complete1233the merge and commit the result (or reuse an existing commit in case1234of <<fast-forwards,fast-forward>>, see below). On the other hand,1235if there are conflicts--for example, if the same file is1236modified in two different ways in the remote branch and the local1237branch--then you are warned; the output may look something like this:12381239-------------------------------------------------1240$ git merge next1241 100% (4/4) done1242Auto-merged file.txt1243CONFLICT (content): Merge conflict in file.txt1244Automatic merge failed; fix conflicts and then commit the result.1245-------------------------------------------------12461247Conflict markers are left in the problematic files, and after1248you resolve the conflicts manually, you can update the index1249with the contents and run Git commit, as you normally would when1250creating a new file.12511252If you examine the resulting commit using gitk, you will see that it1253has two parents, one pointing to the top of the current branch, and1254one to the top of the other branch.12551256[[resolving-a-merge]]1257Resolving a merge1258-----------------12591260When a merge isn't resolved automatically, Git leaves the index and1261the working tree in a special state that gives you all the1262information you need to help resolve the merge.12631264Files with conflicts are marked specially in the index, so until you1265resolve the problem and update the index, linkgit:git-commit[1] will1266fail:12671268-------------------------------------------------1269$ git commit1270file.txt: needs merge1271-------------------------------------------------12721273Also, linkgit:git-status[1] will list those files as "unmerged", and the1274files with conflicts will have conflict markers added, like this:12751276-------------------------------------------------1277<<<<<<< HEAD:file.txt1278Hello world1279=======1280Goodbye1281>>>>>>> 77976da35a11db4580b80ae27e8d65caf5208086:file.txt1282-------------------------------------------------12831284All you need to do is edit the files to resolve the conflicts, and then12851286-------------------------------------------------1287$ git add file.txt1288$ git commit1289-------------------------------------------------12901291Note that the commit message will already be filled in for you with1292some information about the merge. Normally you can just use this1293default message unchanged, but you may add additional commentary of1294your own if desired.12951296The above is all you need to know to resolve a simple merge. But Git1297also provides more information to help resolve conflicts:12981299[[conflict-resolution]]1300Getting conflict-resolution help during a merge1301~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~13021303All of the changes that Git was able to merge automatically are1304already added to the index file, so linkgit:git-diff[1] shows only1305the conflicts. It uses an unusual syntax:13061307-------------------------------------------------1308$ git diff1309diff --cc file.txt1310index 802992c,2b60207..00000001311--- a/file.txt1312+++ b/file.txt1313@@@ -1,1 -1,1 +1,5 @@@1314++<<<<<<< HEAD:file.txt1315 +Hello world1316++=======1317+ Goodbye1318++>>>>>>> 77976da35a11db4580b80ae27e8d65caf5208086:file.txt1319-------------------------------------------------13201321Recall that the commit which will be committed after we resolve this1322conflict will have two parents instead of the usual one: one parent1323will be HEAD, the tip of the current branch; the other will be the1324tip of the other branch, which is stored temporarily in MERGE_HEAD.13251326During the merge, the index holds three versions of each file. Each of1327these three "file stages" represents a different version of the file:13281329-------------------------------------------------1330$ git show :1:file.txt # the file in a common ancestor of both branches1331$ git show :2:file.txt # the version from HEAD.1332$ git show :3:file.txt # the version from MERGE_HEAD.1333-------------------------------------------------13341335When you ask linkgit:git-diff[1] to show the conflicts, it runs a1336three-way diff between the conflicted merge results in the work tree with1337stages 2 and 3 to show only hunks whose contents come from both sides,1338mixed (in other words, when a hunk's merge results come only from stage 2,1339that part is not conflicting and is not shown. Same for stage 3).13401341The diff above shows the differences between the working-tree version of1342file.txt and the stage 2 and stage 3 versions. So instead of preceding1343each line by a single `+` or `-`, it now uses two columns: the first1344column is used for differences between the first parent and the working1345directory copy, and the second for differences between the second parent1346and the working directory copy. (See the "COMBINED DIFF FORMAT" section1347of linkgit:git-diff-files[1] for a details of the format.)13481349After resolving the conflict in the obvious way (but before updating the1350index), the diff will look like:13511352-------------------------------------------------1353$ git diff1354diff --cc file.txt1355index 802992c,2b60207..00000001356--- a/file.txt1357+++ b/file.txt1358@@@ -1,1 -1,1 +1,1 @@@1359- Hello world1360 -Goodbye1361++Goodbye world1362-------------------------------------------------13631364This shows that our resolved version deleted "Hello world" from the1365first parent, deleted "Goodbye" from the second parent, and added1366"Goodbye world", which was previously absent from both.13671368Some special diff options allow diffing the working directory against1369any of these stages:13701371-------------------------------------------------1372$ git diff -1 file.txt # diff against stage 11373$ git diff --base file.txt # same as the above1374$ git diff -2 file.txt # diff against stage 21375$ git diff --ours file.txt # same as the above1376$ git diff -3 file.txt # diff against stage 31377$ git diff --theirs file.txt # same as the above.1378-------------------------------------------------13791380The linkgit:git-log[1] and linkgit:gitk[1] commands also provide special help1381for merges:13821383-------------------------------------------------1384$ git log --merge1385$ gitk --merge1386-------------------------------------------------13871388These will display all commits which exist only on HEAD or on1389MERGE_HEAD, and which touch an unmerged file.13901391You may also use linkgit:git-mergetool[1], which lets you merge the1392unmerged files using external tools such as Emacs or kdiff3.13931394Each time you resolve the conflicts in a file and update the index:13951396-------------------------------------------------1397$ git add file.txt1398-------------------------------------------------13991400the different stages of that file will be "collapsed", after which1401`git diff` will (by default) no longer show diffs for that file.14021403[[undoing-a-merge]]1404Undoing a merge1405---------------14061407If you get stuck and decide to just give up and throw the whole mess1408away, you can always return to the pre-merge state with14091410-------------------------------------------------1411$ git reset --hard HEAD1412-------------------------------------------------14131414Or, if you've already committed the merge that you want to throw away,14151416-------------------------------------------------1417$ git reset --hard ORIG_HEAD1418-------------------------------------------------14191420However, this last command can be dangerous in some cases--never1421throw away a commit you have already committed if that commit may1422itself have been merged into another branch, as doing so may confuse1423further merges.14241425[[fast-forwards]]1426Fast-forward merges1427-------------------14281429There is one special case not mentioned above, which is treated1430differently. Normally, a merge results in a merge commit, with two1431parents, one pointing at each of the two lines of development that1432were merged.14331434However, if the current branch is an ancestor of the other--so every commit1435present in the current branch is already contained in the other branch--then Git1436just performs a "fast-forward"; the head of the current branch is moved forward1437to point at the head of the merged-in branch, without any new commits being1438created.14391440[[fixing-mistakes]]1441Fixing mistakes1442---------------14431444If you've messed up the working tree, but haven't yet committed your1445mistake, you can return the entire working tree to the last committed1446state with14471448-------------------------------------------------1449$ git reset --hard HEAD1450-------------------------------------------------14511452If you make a commit that you later wish you hadn't, there are two1453fundamentally different ways to fix the problem:14541455 1. You can create a new commit that undoes whatever was done1456 by the old commit. This is the correct thing if your1457 mistake has already been made public.14581459 2. You can go back and modify the old commit. You should1460 never do this if you have already made the history public;1461 Git does not normally expect the "history" of a project to1462 change, and cannot correctly perform repeated merges from1463 a branch that has had its history changed.14641465[[reverting-a-commit]]1466Fixing a mistake with a new commit1467~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~14681469Creating a new commit that reverts an earlier change is very easy;1470just pass the linkgit:git-revert[1] command a reference to the bad1471commit; for example, to revert the most recent commit:14721473-------------------------------------------------1474$ git revert HEAD1475-------------------------------------------------14761477This will create a new commit which undoes the change in HEAD. You1478will be given a chance to edit the commit message for the new commit.14791480You can also revert an earlier change, for example, the next-to-last:14811482-------------------------------------------------1483$ git revert HEAD^1484-------------------------------------------------14851486In this case Git will attempt to undo the old change while leaving1487intact any changes made since then. If more recent changes overlap1488with the changes to be reverted, then you will be asked to fix1489conflicts manually, just as in the case of <<resolving-a-merge,1490resolving a merge>>.14911492[[fixing-a-mistake-by-rewriting-history]]1493Fixing a mistake by rewriting history1494~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~14951496If the problematic commit is the most recent commit, and you have not1497yet made that commit public, then you may just1498<<undoing-a-merge,destroy it using `git reset`>>.14991500Alternatively, you1501can edit the working directory and update the index to fix your1502mistake, just as if you were going to <<how-to-make-a-commit,create a1503new commit>>, then run15041505-------------------------------------------------1506$ git commit --amend1507-------------------------------------------------15081509which will replace the old commit by a new commit incorporating your1510changes, giving you a chance to edit the old commit message first.15111512Again, you should never do this to a commit that may already have1513been merged into another branch; use linkgit:git-revert[1] instead in1514that case.15151516It is also possible to replace commits further back in the history, but1517this is an advanced topic to be left for1518<<cleaning-up-history,another chapter>>.15191520[[checkout-of-path]]1521Checking out an old version of a file1522~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~15231524In the process of undoing a previous bad change, you may find it1525useful to check out an older version of a particular file using1526linkgit:git-checkout[1]. We've used `git checkout` before to switch1527branches, but it has quite different behavior if it is given a path1528name: the command15291530-------------------------------------------------1531$ git checkout HEAD^ path/to/file1532-------------------------------------------------15331534replaces path/to/file by the contents it had in the commit HEAD^, and1535also updates the index to match. It does not change branches.15361537If you just want to look at an old version of the file, without1538modifying the working directory, you can do that with1539linkgit:git-show[1]:15401541-------------------------------------------------1542$ git show HEAD^:path/to/file1543-------------------------------------------------15441545which will display the given version of the file.15461547[[interrupted-work]]1548Temporarily setting aside work in progress1549~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~15501551While you are in the middle of working on something complicated, you1552find an unrelated but obvious and trivial bug. You would like to fix it1553before continuing. You can use linkgit:git-stash[1] to save the current1554state of your work, and after fixing the bug (or, optionally after doing1555so on a different branch and then coming back), unstash the1556work-in-progress changes.15571558------------------------------------------------1559$ git stash push -m "work in progress for foo feature"1560------------------------------------------------15611562This command will save your changes away to the `stash`, and1563reset your working tree and the index to match the tip of your1564current branch. Then you can make your fix as usual.15651566------------------------------------------------1567... edit and test ...1568$ git commit -a -m "blorpl: typofix"1569------------------------------------------------15701571After that, you can go back to what you were working on with1572`git stash pop`:15731574------------------------------------------------1575$ git stash pop1576------------------------------------------------157715781579[[ensuring-good-performance]]1580Ensuring good performance1581-------------------------15821583On large repositories, Git depends on compression to keep the history1584information from taking up too much space on disk or in memory. Some1585Git commands may automatically run linkgit:git-gc[1], so you don't1586have to worry about running it manually. However, compressing a large1587repository may take a while, so you may want to call `gc` explicitly1588to avoid automatic compression kicking in when it is not convenient.158915901591[[ensuring-reliability]]1592Ensuring reliability1593--------------------15941595[[checking-for-corruption]]1596Checking the repository for corruption1597~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~15981599The linkgit:git-fsck[1] command runs a number of self-consistency checks1600on the repository, and reports on any problems. This may take some1601time.16021603-------------------------------------------------1604$ git fsck1605dangling commit 7281251ddd2a61e38657c827739c57015671a6b31606dangling commit 2706a059f258c6b245f298dc4ff2ccd30ec21a631607dangling commit 13472b7c4b80851a1bc551779171dcb03655e9b51608dangling blob 218761f9d90712d37a9c5e36f406f92202db07eb1609dangling commit bf093535a34a4d35731aa2bd90fe6b176302f14f1610dangling commit 8e4bec7f2ddaa268bef999853c25755452100f8e1611dangling tree d50bb86186bf27b681d25af89d3b5b68382e40851612dangling tree b24c2473f1fd3d91352a624795be026d64c8841f1613...1614-------------------------------------------------16151616You will see informational messages on dangling objects. They are objects1617that still exist in the repository but are no longer referenced by any of1618your branches, and can (and will) be removed after a while with `gc`.1619You can run `git fsck --no-dangling` to suppress these messages, and still1620view real errors.16211622[[recovering-lost-changes]]1623Recovering lost changes1624~~~~~~~~~~~~~~~~~~~~~~~16251626[[reflogs]]1627Reflogs1628^^^^^^^16291630Say you modify a branch with <<fixing-mistakes,`git reset --hard`>>,1631and then realize that the branch was the only reference you had to1632that point in history.16331634Fortunately, Git also keeps a log, called a "reflog", of all the1635previous values of each branch. So in this case you can still find the1636old history using, for example,16371638-------------------------------------------------1639$ git log master@{1}1640-------------------------------------------------16411642This lists the commits reachable from the previous version of the1643`master` branch head. This syntax can be used with any Git command1644that accepts a commit, not just with `git log`. Some other examples:16451646-------------------------------------------------1647$ git show master@{2} # See where the branch pointed 2,1648$ git show master@{3} # 3, ... changes ago.1649$ gitk master@{yesterday} # See where it pointed yesterday,1650$ gitk master@{"1 week ago"} # ... or last week1651$ git log --walk-reflogs master # show reflog entries for master1652-------------------------------------------------16531654A separate reflog is kept for the HEAD, so16551656-------------------------------------------------1657$ git show HEAD@{"1 week ago"}1658-------------------------------------------------16591660will show what HEAD pointed to one week ago, not what the current branch1661pointed to one week ago. This allows you to see the history of what1662you've checked out.16631664The reflogs are kept by default for 30 days, after which they may be1665pruned. See linkgit:git-reflog[1] and linkgit:git-gc[1] to learn1666how to control this pruning, and see the "SPECIFYING REVISIONS"1667section of linkgit:gitrevisions[7] for details.16681669Note that the reflog history is very different from normal Git history.1670While normal history is shared by every repository that works on the1671same project, the reflog history is not shared: it tells you only about1672how the branches in your local repository have changed over time.16731674[[dangling-object-recovery]]1675Examining dangling objects1676^^^^^^^^^^^^^^^^^^^^^^^^^^16771678In some situations the reflog may not be able to save you. For example,1679suppose you delete a branch, then realize you need the history it1680contained. The reflog is also deleted; however, if you have not yet1681pruned the repository, then you may still be able to find the lost1682commits in the dangling objects that `git fsck` reports. See1683<<dangling-objects>> for the details.16841685-------------------------------------------------1686$ git fsck1687dangling commit 7281251ddd2a61e38657c827739c57015671a6b31688dangling commit 2706a059f258c6b245f298dc4ff2ccd30ec21a631689dangling commit 13472b7c4b80851a1bc551779171dcb03655e9b51690...1691-------------------------------------------------16921693You can examine1694one of those dangling commits with, for example,16951696------------------------------------------------1697$ gitk 7281251ddd --not --all1698------------------------------------------------16991700which does what it sounds like: it says that you want to see the commit1701history that is described by the dangling commit(s), but not the1702history that is described by all your existing branches and tags. Thus1703you get exactly the history reachable from that commit that is lost.1704(And notice that it might not be just one commit: we only report the1705"tip of the line" as being dangling, but there might be a whole deep1706and complex commit history that was dropped.)17071708If you decide you want the history back, you can always create a new1709reference pointing to it, for example, a new branch:17101711------------------------------------------------1712$ git branch recovered-branch 7281251ddd1713------------------------------------------------17141715Other types of dangling objects (blobs and trees) are also possible, and1716dangling objects can arise in other situations.171717181719[[sharing-development]]1720Sharing development with others1721===============================17221723[[getting-updates-With-git-pull]]1724Getting updates with git pull1725-----------------------------17261727After you clone a repository and commit a few changes of your own, you1728may wish to check the original repository for updates and merge them1729into your own work.17301731We have already seen <<Updating-a-repository-With-git-fetch,how to1732keep remote-tracking branches up to date>> with linkgit:git-fetch[1],1733and how to merge two branches. So you can merge in changes from the1734original repository's master branch with:17351736-------------------------------------------------1737$ git fetch1738$ git merge origin/master1739-------------------------------------------------17401741However, the linkgit:git-pull[1] command provides a way to do this in1742one step:17431744-------------------------------------------------1745$ git pull origin master1746-------------------------------------------------17471748In fact, if you have `master` checked out, then this branch has been1749configured by `git clone` to get changes from the HEAD branch of the1750origin repository. So often you can1751accomplish the above with just a simple17521753-------------------------------------------------1754$ git pull1755-------------------------------------------------17561757This command will fetch changes from the remote branches to your1758remote-tracking branches `origin/*`, and merge the default branch into1759the current branch.17601761More generally, a branch that is created from a remote-tracking branch1762will pull1763by default from that branch. See the descriptions of the1764`branch.<name>.remote` and `branch.<name>.merge` options in1765linkgit:git-config[1], and the discussion of the `--track` option in1766linkgit:git-checkout[1], to learn how to control these defaults.17671768In addition to saving you keystrokes, `git pull` also helps you by1769producing a default commit message documenting the branch and1770repository that you pulled from.17711772(But note that no such commit will be created in the case of a1773<<fast-forwards,fast-forward>>; instead, your branch will just be1774updated to point to the latest commit from the upstream branch.)17751776The `git pull` command can also be given `.` as the "remote" repository,1777in which case it just merges in a branch from the current repository; so1778the commands17791780-------------------------------------------------1781$ git pull . branch1782$ git merge branch1783-------------------------------------------------17841785are roughly equivalent.17861787[[submitting-patches]]1788Submitting patches to a project1789-------------------------------17901791If you just have a few changes, the simplest way to submit them may1792just be to send them as patches in email:17931794First, use linkgit:git-format-patch[1]; for example:17951796-------------------------------------------------1797$ git format-patch origin1798-------------------------------------------------17991800will produce a numbered series of files in the current directory, one1801for each patch in the current branch but not in `origin/HEAD`.18021803`git format-patch` can include an initial "cover letter". You can insert1804commentary on individual patches after the three dash line which1805`format-patch` places after the commit message but before the patch1806itself. If you use `git notes` to track your cover letter material,1807`git format-patch --notes` will include the commit's notes in a similar1808manner.18091810You can then import these into your mail client and send them by1811hand. However, if you have a lot to send at once, you may prefer to1812use the linkgit:git-send-email[1] script to automate the process.1813Consult the mailing list for your project first to determine1814their requirements for submitting patches.18151816[[importing-patches]]1817Importing patches to a project1818------------------------------18191820Git also provides a tool called linkgit:git-am[1] (am stands for1821"apply mailbox"), for importing such an emailed series of patches.1822Just save all of the patch-containing messages, in order, into a1823single mailbox file, say `patches.mbox`, then run18241825-------------------------------------------------1826$ git am -3 patches.mbox1827-------------------------------------------------18281829Git will apply each patch in order; if any conflicts are found, it1830will stop, and you can fix the conflicts as described in1831"<<resolving-a-merge,Resolving a merge>>". (The `-3` option tells1832Git to perform a merge; if you would prefer it just to abort and1833leave your tree and index untouched, you may omit that option.)18341835Once the index is updated with the results of the conflict1836resolution, instead of creating a new commit, just run18371838-------------------------------------------------1839$ git am --continue1840-------------------------------------------------18411842and Git will create the commit for you and continue applying the1843remaining patches from the mailbox.18441845The final result will be a series of commits, one for each patch in1846the original mailbox, with authorship and commit log message each1847taken from the message containing each patch.18481849[[public-repositories]]1850Public Git repositories1851-----------------------18521853Another way to submit changes to a project is to tell the maintainer1854of that project to pull the changes from your repository using1855linkgit:git-pull[1]. In the section "<<getting-updates-With-git-pull,1856Getting updates with `git pull`>>" we described this as a way to get1857updates from the "main" repository, but it works just as well in the1858other direction.18591860If you and the maintainer both have accounts on the same machine, then1861you can just pull changes from each other's repositories directly;1862commands that accept repository URLs as arguments will also accept a1863local directory name:18641865-------------------------------------------------1866$ git clone /path/to/repository1867$ git pull /path/to/other/repository1868-------------------------------------------------18691870or an ssh URL:18711872-------------------------------------------------1873$ git clone ssh://yourhost/~you/repository1874-------------------------------------------------18751876For projects with few developers, or for synchronizing a few private1877repositories, this may be all you need.18781879However, the more common way to do this is to maintain a separate public1880repository (usually on a different host) for others to pull changes1881from. This is usually more convenient, and allows you to cleanly1882separate private work in progress from publicly visible work.18831884You will continue to do your day-to-day work in your personal1885repository, but periodically "push" changes from your personal1886repository into your public repository, allowing other developers to1887pull from that repository. So the flow of changes, in a situation1888where there is one other developer with a public repository, looks1889like this:18901891 you push1892 your personal repo ------------------> your public repo1893 ^ |1894 | |1895 | you pull | they pull1896 | |1897 | |1898 | they push V1899 their public repo <------------------- their repo19001901We explain how to do this in the following sections.19021903[[setting-up-a-public-repository]]1904Setting up a public repository1905~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~19061907Assume your personal repository is in the directory `~/proj`. We1908first create a new clone of the repository and tell `git daemon` that it1909is meant to be public:19101911-------------------------------------------------1912$ git clone --bare ~/proj proj.git1913$ touch proj.git/git-daemon-export-ok1914-------------------------------------------------19151916The resulting directory proj.git contains a "bare" git repository--it is1917just the contents of the `.git` directory, without any files checked out1918around it.19191920Next, copy `proj.git` to the server where you plan to host the1921public repository. You can use scp, rsync, or whatever is most1922convenient.19231924[[exporting-via-git]]1925Exporting a Git repository via the Git protocol1926~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~19271928This is the preferred method.19291930If someone else administers the server, they should tell you what1931directory to put the repository in, and what `git://` URL it will1932appear at. You can then skip to the section1933"<<pushing-changes-to-a-public-repository,Pushing changes to a public1934repository>>", below.19351936Otherwise, all you need to do is start linkgit:git-daemon[1]; it will1937listen on port 9418. By default, it will allow access to any directory1938that looks like a Git directory and contains the magic file1939git-daemon-export-ok. Passing some directory paths as `git daemon`1940arguments will further restrict the exports to those paths.19411942You can also run `git daemon` as an inetd service; see the1943linkgit:git-daemon[1] man page for details. (See especially the1944examples section.)19451946[[exporting-via-http]]1947Exporting a git repository via HTTP1948~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~19491950The Git protocol gives better performance and reliability, but on a1951host with a web server set up, HTTP exports may be simpler to set up.19521953All you need to do is place the newly created bare Git repository in1954a directory that is exported by the web server, and make some1955adjustments to give web clients some extra information they need:19561957-------------------------------------------------1958$ mv proj.git /home/you/public_html/proj.git1959$ cd proj.git1960$ git --bare update-server-info1961$ mv hooks/post-update.sample hooks/post-update1962-------------------------------------------------19631964(For an explanation of the last two lines, see1965linkgit:git-update-server-info[1] and linkgit:githooks[5].)19661967Advertise the URL of `proj.git`. Anybody else should then be able to1968clone or pull from that URL, for example with a command line like:19691970-------------------------------------------------1971$ git clone http://yourserver.com/~you/proj.git1972-------------------------------------------------19731974(See also1975link:howto/setup-git-server-over-http.html[setup-git-server-over-http]1976for a slightly more sophisticated setup using WebDAV which also1977allows pushing over HTTP.)19781979[[pushing-changes-to-a-public-repository]]1980Pushing changes to a public repository1981~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~19821983Note that the two techniques outlined above (exporting via1984<<exporting-via-http,http>> or <<exporting-via-git,git>>) allow other1985maintainers to fetch your latest changes, but they do not allow write1986access, which you will need to update the public repository with the1987latest changes created in your private repository.19881989The simplest way to do this is using linkgit:git-push[1] and ssh; to1990update the remote branch named `master` with the latest state of your1991branch named `master`, run19921993-------------------------------------------------1994$ git push ssh://yourserver.com/~you/proj.git master:master1995-------------------------------------------------19961997or just19981999-------------------------------------------------2000$ git push ssh://yourserver.com/~you/proj.git master2001-------------------------------------------------20022003As with `git fetch`, `git push` will complain if this does not result in a2004<<fast-forwards,fast-forward>>; see the following section for details on2005handling this case.20062007Note that the target of a `push` is normally a2008<<def_bare_repository,bare>> repository. You can also push to a2009repository that has a checked-out working tree, but a push to update the2010currently checked-out branch is denied by default to prevent confusion.2011See the description of the receive.denyCurrentBranch option2012in linkgit:git-config[1] for details.20132014As with `git fetch`, you may also set up configuration options to2015save typing; so, for example:20162017-------------------------------------------------2018$ git remote add public-repo ssh://yourserver.com/~you/proj.git2019-------------------------------------------------20202021adds the following to `.git/config`:20222023-------------------------------------------------2024[remote "public-repo"]2025 url = yourserver.com:proj.git2026 fetch = +refs/heads/*:refs/remotes/example/*2027-------------------------------------------------20282029which lets you do the same push with just20302031-------------------------------------------------2032$ git push public-repo master2033-------------------------------------------------20342035See the explanations of the `remote.<name>.url`,2036`branch.<name>.remote`, and `remote.<name>.push` options in2037linkgit:git-config[1] for details.20382039[[forcing-push]]2040What to do when a push fails2041~~~~~~~~~~~~~~~~~~~~~~~~~~~~20422043If a push would not result in a <<fast-forwards,fast-forward>> of the2044remote branch, then it will fail with an error like:20452046-------------------------------------------------2047 ! [rejected] master -> master (non-fast-forward)2048error: failed to push some refs to '...'2049hint: Updates were rejected because the tip of your current branch is behind2050hint: its remote counterpart. Integrate the remote changes (e.g.2051hint: 'git pull ...') before pushing again.2052hint: See the 'Note about fast-forwards' in 'git push --help' for details.2053-------------------------------------------------20542055This can happen, for example, if you:20562057 - use `git reset --hard` to remove already-published commits, or2058 - use `git commit --amend` to replace already-published commits2059 (as in <<fixing-a-mistake-by-rewriting-history>>), or2060 - use `git rebase` to rebase any already-published commits (as2061 in <<using-git-rebase>>).20622063You may force `git push` to perform the update anyway by preceding the2064branch name with a plus sign:20652066-------------------------------------------------2067$ git push ssh://yourserver.com/~you/proj.git +master2068-------------------------------------------------20692070Note the addition of the `+` sign. Alternatively, you can use the2071`-f` flag to force the remote update, as in:20722073-------------------------------------------------2074$ git push -f ssh://yourserver.com/~you/proj.git master2075-------------------------------------------------20762077Normally whenever a branch head in a public repository is modified, it2078is modified to point to a descendant of the commit that it pointed to2079before. By forcing a push in this situation, you break that convention.2080(See <<problems-With-rewriting-history>>.)20812082Nevertheless, this is a common practice for people that need a simple2083way to publish a work-in-progress patch series, and it is an acceptable2084compromise as long as you warn other developers that this is how you2085intend to manage the branch.20862087It's also possible for a push to fail in this way when other people have2088the right to push to the same repository. In that case, the correct2089solution is to retry the push after first updating your work: either by a2090pull, or by a fetch followed by a rebase; see the2091<<setting-up-a-shared-repository,next section>> and2092linkgit:gitcvs-migration[7] for more.20932094[[setting-up-a-shared-repository]]2095Setting up a shared repository2096~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~20972098Another way to collaborate is by using a model similar to that2099commonly used in CVS, where several developers with special rights2100all push to and pull from a single shared repository. See2101linkgit:gitcvs-migration[7] for instructions on how to2102set this up.21032104However, while there is nothing wrong with Git's support for shared2105repositories, this mode of operation is not generally recommended,2106simply because the mode of collaboration that Git supports--by2107exchanging patches and pulling from public repositories--has so many2108advantages over the central shared repository:21092110 - Git's ability to quickly import and merge patches allows a2111 single maintainer to process incoming changes even at very2112 high rates. And when that becomes too much, `git pull` provides2113 an easy way for that maintainer to delegate this job to other2114 maintainers while still allowing optional review of incoming2115 changes.2116 - Since every developer's repository has the same complete copy2117 of the project history, no repository is special, and it is2118 trivial for another developer to take over maintenance of a2119 project, either by mutual agreement, or because a maintainer2120 becomes unresponsive or difficult to work with.2121 - The lack of a central group of "committers" means there is2122 less need for formal decisions about who is "in" and who is2123 "out".21242125[[setting-up-gitweb]]2126Allowing web browsing of a repository2127~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~21282129The gitweb cgi script provides users an easy way to browse your2130project's revisions, file contents and logs without having to install2131Git. Features like RSS/Atom feeds and blame/annotation details may2132optionally be enabled.21332134The linkgit:git-instaweb[1] command provides a simple way to start2135browsing the repository using gitweb. The default server when using2136instaweb is lighttpd.21372138See the file gitweb/INSTALL in the Git source tree and2139linkgit:gitweb[1] for instructions on details setting up a permanent2140installation with a CGI or Perl capable server.21412142[[how-to-get-a-git-repository-with-minimal-history]]2143How to get a Git repository with minimal history2144------------------------------------------------21452146A <<def_shallow_clone,shallow clone>>, with its truncated2147history, is useful when one is interested only in recent history2148of a project and getting full history from the upstream is2149expensive.21502151A <<def_shallow_clone,shallow clone>> is created by specifying2152the linkgit:git-clone[1] `--depth` switch. The depth can later be2153changed with the linkgit:git-fetch[1] `--depth` switch, or full2154history restored with `--unshallow`.21552156Merging inside a <<def_shallow_clone,shallow clone>> will work as long2157as a merge base is in the recent history.2158Otherwise, it will be like merging unrelated histories and may2159have to result in huge conflicts. This limitation may make such2160a repository unsuitable to be used in merge based workflows.21612162[[sharing-development-examples]]2163Examples2164--------21652166[[maintaining-topic-branches]]2167Maintaining topic branches for a Linux subsystem maintainer2168~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~21692170This describes how Tony Luck uses Git in his role as maintainer of the2171IA64 architecture for the Linux kernel.21722173He uses two public branches:21742175 - A "test" tree into which patches are initially placed so that they2176 can get some exposure when integrated with other ongoing development.2177 This tree is available to Andrew for pulling into -mm whenever he2178 wants.21792180 - A "release" tree into which tested patches are moved for final sanity2181 checking, and as a vehicle to send them upstream to Linus (by sending2182 him a "please pull" request.)21832184He also uses a set of temporary branches ("topic branches"), each2185containing a logical grouping of patches.21862187To set this up, first create your work tree by cloning Linus's public2188tree:21892190-------------------------------------------------2191$ git clone git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git work2192$ cd work2193-------------------------------------------------21942195Linus's tree will be stored in the remote-tracking branch named origin/master,2196and can be updated using linkgit:git-fetch[1]; you can track other2197public trees using linkgit:git-remote[1] to set up a "remote" and2198linkgit:git-fetch[1] to keep them up to date; see2199<<repositories-and-branches>>.22002201Now create the branches in which you are going to work; these start out2202at the current tip of origin/master branch, and should be set up (using2203the `--track` option to linkgit:git-branch[1]) to merge changes in from2204Linus by default.22052206-------------------------------------------------2207$ git branch --track test origin/master2208$ git branch --track release origin/master2209-------------------------------------------------22102211These can be easily kept up to date using linkgit:git-pull[1].22122213-------------------------------------------------2214$ git checkout test && git pull2215$ git checkout release && git pull2216-------------------------------------------------22172218Important note! If you have any local changes in these branches, then2219this merge will create a commit object in the history (with no local2220changes Git will simply do a "fast-forward" merge). Many people dislike2221the "noise" that this creates in the Linux history, so you should avoid2222doing this capriciously in the `release` branch, as these noisy commits2223will become part of the permanent history when you ask Linus to pull2224from the release branch.22252226A few configuration variables (see linkgit:git-config[1]) can2227make it easy to push both branches to your public tree. (See2228<<setting-up-a-public-repository>>.)22292230-------------------------------------------------2231$ cat >> .git/config <<EOF2232[remote "mytree"]2233 url = master.kernel.org:/pub/scm/linux/kernel/git/aegl/linux.git2234 push = release2235 push = test2236EOF2237-------------------------------------------------22382239Then you can push both the test and release trees using2240linkgit:git-push[1]:22412242-------------------------------------------------2243$ git push mytree2244-------------------------------------------------22452246or push just one of the test and release branches using:22472248-------------------------------------------------2249$ git push mytree test2250-------------------------------------------------22512252or22532254-------------------------------------------------2255$ git push mytree release2256-------------------------------------------------22572258Now to apply some patches from the community. Think of a short2259snappy name for a branch to hold this patch (or related group of2260patches), and create a new branch from a recent stable tag of2261Linus's branch. Picking a stable base for your branch will:22621) help you: by avoiding inclusion of unrelated and perhaps lightly2263tested changes22642) help future bug hunters that use `git bisect` to find problems22652266-------------------------------------------------2267$ git checkout -b speed-up-spinlocks v2.6.352268-------------------------------------------------22692270Now you apply the patch(es), run some tests, and commit the change(s). If2271the patch is a multi-part series, then you should apply each as a separate2272commit to this branch.22732274-------------------------------------------------2275$ ... patch ... test ... commit [ ... patch ... test ... commit ]*2276-------------------------------------------------22772278When you are happy with the state of this change, you can merge it into the2279"test" branch in preparation to make it public:22802281-------------------------------------------------2282$ git checkout test && git merge speed-up-spinlocks2283-------------------------------------------------22842285It is unlikely that you would have any conflicts here ... but you might if you2286spent a while on this step and had also pulled new versions from upstream.22872288Sometime later when enough time has passed and testing done, you can pull the2289same branch into the `release` tree ready to go upstream. This is where you2290see the value of keeping each patch (or patch series) in its own branch. It2291means that the patches can be moved into the `release` tree in any order.22922293-------------------------------------------------2294$ git checkout release && git merge speed-up-spinlocks2295-------------------------------------------------22962297After a while, you will have a number of branches, and despite the2298well chosen names you picked for each of them, you may forget what2299they are for, or what status they are in. To get a reminder of what2300changes are in a specific branch, use:23012302-------------------------------------------------2303$ git log linux..branchname | git shortlog2304-------------------------------------------------23052306To see whether it has already been merged into the test or release branches,2307use:23082309-------------------------------------------------2310$ git log test..branchname2311-------------------------------------------------23122313or23142315-------------------------------------------------2316$ git log release..branchname2317-------------------------------------------------23182319(If this branch has not yet been merged, you will see some log entries.2320If it has been merged, then there will be no output.)23212322Once a patch completes the great cycle (moving from test to release,2323then pulled by Linus, and finally coming back into your local2324`origin/master` branch), the branch for this change is no longer needed.2325You detect this when the output from:23262327-------------------------------------------------2328$ git log origin..branchname2329-------------------------------------------------23302331is empty. At this point the branch can be deleted:23322333-------------------------------------------------2334$ git branch -d branchname2335-------------------------------------------------23362337Some changes are so trivial that it is not necessary to create a separate2338branch and then merge into each of the test and release branches. For2339these changes, just apply directly to the `release` branch, and then2340merge that into the `test` branch.23412342After pushing your work to `mytree`, you can use2343linkgit:git-request-pull[1] to prepare a "please pull" request message2344to send to Linus:23452346-------------------------------------------------2347$ git push mytree2348$ git request-pull origin mytree release2349-------------------------------------------------23502351Here are some of the scripts that simplify all this even further.23522353-------------------------------------------------2354==== update script ====2355# Update a branch in my Git tree. If the branch to be updated2356# is origin, then pull from kernel.org. Otherwise merge2357# origin/master branch into test|release branch23582359case "$1" in2360test|release)2361 git checkout $1 && git pull . origin2362 ;;2363origin)2364 before=$(git rev-parse refs/remotes/origin/master)2365 git fetch origin2366 after=$(git rev-parse refs/remotes/origin/master)2367 if [ $before != $after ]2368 then2369 git log $before..$after | git shortlog2370 fi2371 ;;2372*)2373 echo "usage: $0 origin|test|release" 1>&22374 exit 12375 ;;2376esac2377-------------------------------------------------23782379-------------------------------------------------2380==== merge script ====2381# Merge a branch into either the test or release branch23822383pname=$023842385usage()2386{2387 echo "usage: $pname branch test|release" 1>&22388 exit 12389}23902391git show-ref -q --verify -- refs/heads/"$1" || {2392 echo "Can't see branch <$1>" 1>&22393 usage2394}23952396case "$2" in2397test|release)2398 if [ $(git log $2..$1 | wc -c) -eq 0 ]2399 then2400 echo $1 already merged into $2 1>&22401 exit 12402 fi2403 git checkout $2 && git pull . $12404 ;;2405*)2406 usage2407 ;;2408esac2409-------------------------------------------------24102411-------------------------------------------------2412==== status script ====2413# report on status of my ia64 Git tree24142415gb=$(tput setab 2)2416rb=$(tput setab 1)2417restore=$(tput setab 9)24182419if [ `git rev-list test..release | wc -c` -gt 0 ]2420then2421 echo $rb Warning: commits in release that are not in test $restore2422 git log test..release2423fi24242425for branch in `git show-ref --heads | sed 's|^.*/||'`2426do2427 if [ $branch = test -o $branch = release ]2428 then2429 continue2430 fi24312432 echo -n $gb ======= $branch ====== $restore " "2433 status=2434 for ref in test release origin/master2435 do2436 if [ `git rev-list $ref..$branch | wc -c` -gt 0 ]2437 then2438 status=$status${ref:0:1}2439 fi2440 done2441 case $status in2442 trl)2443 echo $rb Need to pull into test $restore2444 ;;2445 rl)2446 echo "In test"2447 ;;2448 l)2449 echo "Waiting for linus"2450 ;;2451 "")2452 echo $rb All done $restore2453 ;;2454 *)2455 echo $rb "<$status>" $restore2456 ;;2457 esac2458 git log origin/master..$branch | git shortlog2459done2460-------------------------------------------------246124622463[[cleaning-up-history]]2464Rewriting history and maintaining patch series2465==============================================24662467Normally commits are only added to a project, never taken away or2468replaced. Git is designed with this assumption, and violating it will2469cause Git's merge machinery (for example) to do the wrong thing.24702471However, there is a situation in which it can be useful to violate this2472assumption.24732474[[patch-series]]2475Creating the perfect patch series2476---------------------------------24772478Suppose you are a contributor to a large project, and you want to add a2479complicated feature, and to present it to the other developers in a way2480that makes it easy for them to read your changes, verify that they are2481correct, and understand why you made each change.24822483If you present all of your changes as a single patch (or commit), they2484may find that it is too much to digest all at once.24852486If you present them with the entire history of your work, complete with2487mistakes, corrections, and dead ends, they may be overwhelmed.24882489So the ideal is usually to produce a series of patches such that:24902491 1. Each patch can be applied in order.24922493 2. Each patch includes a single logical change, together with a2494 message explaining the change.24952496 3. No patch introduces a regression: after applying any initial2497 part of the series, the resulting project still compiles and2498 works, and has no bugs that it didn't have before.24992500 4. The complete series produces the same end result as your own2501 (probably much messier!) development process did.25022503We will introduce some tools that can help you do this, explain how to2504use them, and then explain some of the problems that can arise because2505you are rewriting history.25062507[[using-git-rebase]]2508Keeping a patch series up to date using git rebase2509--------------------------------------------------25102511Suppose that you create a branch `mywork` on a remote-tracking branch2512`origin`, and create some commits on top of it:25132514-------------------------------------------------2515$ git checkout -b mywork origin2516$ vi file.txt2517$ git commit2518$ vi otherfile.txt2519$ git commit2520...2521-------------------------------------------------25222523You have performed no merges into mywork, so it is just a simple linear2524sequence of patches on top of `origin`:25252526................................................2527 o--o--O <-- origin2528 \2529 a--b--c <-- mywork2530................................................25312532Some more interesting work has been done in the upstream project, and2533`origin` has advanced:25342535................................................2536 o--o--O--o--o--o <-- origin2537 \2538 a--b--c <-- mywork2539................................................25402541At this point, you could use `pull` to merge your changes back in;2542the result would create a new merge commit, like this:25432544................................................2545 o--o--O--o--o--o <-- origin2546 \ \2547 a--b--c--m <-- mywork2548................................................25492550However, if you prefer to keep the history in mywork a simple series of2551commits without any merges, you may instead choose to use2552linkgit:git-rebase[1]:25532554-------------------------------------------------2555$ git checkout mywork2556$ git rebase origin2557-------------------------------------------------25582559This will remove each of your commits from mywork, temporarily saving2560them as patches (in a directory named `.git/rebase-apply`), update mywork to2561point at the latest version of origin, then apply each of the saved2562patches to the new mywork. The result will look like:256325642565................................................2566 o--o--O--o--o--o <-- origin2567 \2568 a'--b'--c' <-- mywork2569................................................25702571In the process, it may discover conflicts. In that case it will stop2572and allow you to fix the conflicts; after fixing conflicts, use `git add`2573to update the index with those contents, and then, instead of2574running `git commit`, just run25752576-------------------------------------------------2577$ git rebase --continue2578-------------------------------------------------25792580and Git will continue applying the rest of the patches.25812582At any point you may use the `--abort` option to abort this process and2583return mywork to the state it had before you started the rebase:25842585-------------------------------------------------2586$ git rebase --abort2587-------------------------------------------------25882589If you need to reorder or edit a number of commits in a branch, it may2590be easier to use `git rebase -i`, which allows you to reorder and2591squash commits, as well as marking them for individual editing during2592the rebase. See <<interactive-rebase>> for details, and2593<<reordering-patch-series>> for alternatives.25942595[[rewriting-one-commit]]2596Rewriting a single commit2597-------------------------25982599We saw in <<fixing-a-mistake-by-rewriting-history>> that you can replace the2600most recent commit using26012602-------------------------------------------------2603$ git commit --amend2604-------------------------------------------------26052606which will replace the old commit by a new commit incorporating your2607changes, giving you a chance to edit the old commit message first.2608This is useful for fixing typos in your last commit, or for adjusting2609the patch contents of a poorly staged commit.26102611If you need to amend commits from deeper in your history, you can2612use <<interactive-rebase,interactive rebase's `edit` instruction>>.26132614[[reordering-patch-series]]2615Reordering or selecting from a patch series2616-------------------------------------------26172618Sometimes you want to edit a commit deeper in your history. One2619approach is to use `git format-patch` to create a series of patches2620and then reset the state to before the patches:26212622-------------------------------------------------2623$ git format-patch origin2624$ git reset --hard origin2625-------------------------------------------------26262627Then modify, reorder, or eliminate patches as needed before applying2628them again with linkgit:git-am[1]:26292630-------------------------------------------------2631$ git am *.patch2632-------------------------------------------------26332634[[interactive-rebase]]2635Using interactive rebases2636-------------------------26372638You can also edit a patch series with an interactive rebase. This is2639the same as <<reordering-patch-series,reordering a patch series using2640`format-patch`>>, so use whichever interface you like best.26412642Rebase your current HEAD on the last commit you want to retain as-is.2643For example, if you want to reorder the last 5 commits, use:26442645-------------------------------------------------2646$ git rebase -i HEAD~52647-------------------------------------------------26482649This will open your editor with a list of steps to be taken to perform2650your rebase.26512652-------------------------------------------------2653pick deadbee The oneline of this commit2654pick fa1afe1 The oneline of the next commit2655...26562657# Rebase c0ffeee..deadbee onto c0ffeee2658#2659# Commands:2660# p, pick = use commit2661# r, reword = use commit, but edit the commit message2662# e, edit = use commit, but stop for amending2663# s, squash = use commit, but meld into previous commit2664# f, fixup = like "squash", but discard this commit's log message2665# x, exec = run command (the rest of the line) using shell2666#2667# These lines can be re-ordered; they are executed from top to bottom.2668#2669# If you remove a line here THAT COMMIT WILL BE LOST.2670#2671# However, if you remove everything, the rebase will be aborted.2672#2673# Note that empty commits are commented out2674-------------------------------------------------26752676As explained in the comments, you can reorder commits, squash them2677together, edit commit messages, etc. by editing the list. Once you2678are satisfied, save the list and close your editor, and the rebase2679will begin.26802681The rebase will stop where `pick` has been replaced with `edit` or2682when a step in the list fails to mechanically resolve conflicts and2683needs your help. When you are done editing and/or resolving conflicts2684you can continue with `git rebase --continue`. If you decide that2685things are getting too hairy, you can always bail out with `git rebase2686--abort`. Even after the rebase is complete, you can still recover2687the original branch by using the <<reflogs,reflog>>.26882689For a more detailed discussion of the procedure and additional tips,2690see the "INTERACTIVE MODE" section of linkgit:git-rebase[1].26912692[[patch-series-tools]]2693Other tools2694-----------26952696There are numerous other tools, such as StGit, which exist for the2697purpose of maintaining a patch series. These are outside of the scope of2698this manual.26992700[[problems-With-rewriting-history]]2701Problems with rewriting history2702-------------------------------27032704The primary problem with rewriting the history of a branch has to do2705with merging. Suppose somebody fetches your branch and merges it into2706their branch, with a result something like this:27072708................................................2709 o--o--O--o--o--o <-- origin2710 \ \2711 t--t--t--m <-- their branch:2712................................................27132714Then suppose you modify the last three commits:27152716................................................2717 o--o--o <-- new head of origin2718 /2719 o--o--O--o--o--o <-- old head of origin2720................................................27212722If we examined all this history together in one repository, it will2723look like:27242725................................................2726 o--o--o <-- new head of origin2727 /2728 o--o--O--o--o--o <-- old head of origin2729 \ \2730 t--t--t--m <-- their branch:2731................................................27322733Git has no way of knowing that the new head is an updated version of2734the old head; it treats this situation exactly the same as it would if2735two developers had independently done the work on the old and new heads2736in parallel. At this point, if someone attempts to merge the new head2737in to their branch, Git will attempt to merge together the two (old and2738new) lines of development, instead of trying to replace the old by the2739new. The results are likely to be unexpected.27402741You may still choose to publish branches whose history is rewritten,2742and it may be useful for others to be able to fetch those branches in2743order to examine or test them, but they should not attempt to pull such2744branches into their own work.27452746For true distributed development that supports proper merging,2747published branches should never be rewritten.27482749[[bisect-merges]]2750Why bisecting merge commits can be harder than bisecting linear history2751-----------------------------------------------------------------------27522753The linkgit:git-bisect[1] command correctly handles history that2754includes merge commits. However, when the commit that it finds is a2755merge commit, the user may need to work harder than usual to figure out2756why that commit introduced a problem.27572758Imagine this history:27592760................................................2761 ---Z---o---X---...---o---A---C---D2762 \ /2763 o---o---Y---...---o---B2764................................................27652766Suppose that on the upper line of development, the meaning of one2767of the functions that exists at Z is changed at commit X. The2768commits from Z leading to A change both the function's2769implementation and all calling sites that exist at Z, as well2770as new calling sites they add, to be consistent. There is no2771bug at A.27722773Suppose that in the meantime on the lower line of development somebody2774adds a new calling site for that function at commit Y. The2775commits from Z leading to B all assume the old semantics of that2776function and the callers and the callee are consistent with each2777other. There is no bug at B, either.27782779Suppose further that the two development lines merge cleanly at C,2780so no conflict resolution is required.27812782Nevertheless, the code at C is broken, because the callers added2783on the lower line of development have not been converted to the new2784semantics introduced on the upper line of development. So if all2785you know is that D is bad, that Z is good, and that2786linkgit:git-bisect[1] identifies C as the culprit, how will you2787figure out that the problem is due to this change in semantics?27882789When the result of a `git bisect` is a non-merge commit, you should2790normally be able to discover the problem by examining just that commit.2791Developers can make this easy by breaking their changes into small2792self-contained commits. That won't help in the case above, however,2793because the problem isn't obvious from examination of any single2794commit; instead, a global view of the development is required. To2795make matters worse, the change in semantics in the problematic2796function may be just one small part of the changes in the upper2797line of development.27982799On the other hand, if instead of merging at C you had rebased the2800history between Z to B on top of A, you would have gotten this2801linear history:28022803................................................................2804 ---Z---o---X--...---o---A---o---o---Y*--...---o---B*--D*2805................................................................28062807Bisecting between Z and D* would hit a single culprit commit Y*,2808and understanding why Y* was broken would probably be easier.28092810Partly for this reason, many experienced Git users, even when2811working on an otherwise merge-heavy project, keep the history2812linear by rebasing against the latest upstream version before2813publishing.28142815[[advanced-branch-management]]2816Advanced branch management2817==========================28182819[[fetching-individual-branches]]2820Fetching individual branches2821----------------------------28222823Instead of using linkgit:git-remote[1], you can also choose just2824to update one branch at a time, and to store it locally under an2825arbitrary name:28262827-------------------------------------------------2828$ git fetch origin todo:my-todo-work2829-------------------------------------------------28302831The first argument, `origin`, just tells Git to fetch from the2832repository you originally cloned from. The second argument tells Git2833to fetch the branch named `todo` from the remote repository, and to2834store it locally under the name `refs/heads/my-todo-work`.28352836You can also fetch branches from other repositories; so28372838-------------------------------------------------2839$ git fetch git://example.com/proj.git master:example-master2840-------------------------------------------------28412842will create a new branch named `example-master` and store in it the2843branch named `master` from the repository at the given URL. If you2844already have a branch named example-master, it will attempt to2845<<fast-forwards,fast-forward>> to the commit given by example.com's2846master branch. In more detail:28472848[[fetch-fast-forwards]]2849git fetch and fast-forwards2850---------------------------28512852In the previous example, when updating an existing branch, `git fetch`2853checks to make sure that the most recent commit on the remote2854branch is a descendant of the most recent commit on your copy of the2855branch before updating your copy of the branch to point at the new2856commit. Git calls this process a <<fast-forwards,fast-forward>>.28572858A fast-forward looks something like this:28592860................................................2861 o--o--o--o <-- old head of the branch2862 \2863 o--o--o <-- new head of the branch2864................................................286528662867In some cases it is possible that the new head will *not* actually be2868a descendant of the old head. For example, the developer may have2869realized she made a serious mistake, and decided to backtrack,2870resulting in a situation like:28712872................................................2873 o--o--o--o--a--b <-- old head of the branch2874 \2875 o--o--o <-- new head of the branch2876................................................28772878In this case, `git fetch` will fail, and print out a warning.28792880In that case, you can still force Git to update to the new head, as2881described in the following section. However, note that in the2882situation above this may mean losing the commits labeled `a` and `b`,2883unless you've already created a reference of your own pointing to2884them.28852886[[forcing-fetch]]2887Forcing git fetch to do non-fast-forward updates2888------------------------------------------------28892890If git fetch fails because the new head of a branch is not a2891descendant of the old head, you may force the update with:28922893-------------------------------------------------2894$ git fetch git://example.com/proj.git +master:refs/remotes/example/master2895-------------------------------------------------28962897Note the addition of the `+` sign. Alternatively, you can use the `-f`2898flag to force updates of all the fetched branches, as in:28992900-------------------------------------------------2901$ git fetch -f origin2902-------------------------------------------------29032904Be aware that commits that the old version of example/master pointed at2905may be lost, as we saw in the previous section.29062907[[remote-branch-configuration]]2908Configuring remote-tracking branches2909------------------------------------29102911We saw above that `origin` is just a shortcut to refer to the2912repository that you originally cloned from. This information is2913stored in Git configuration variables, which you can see using2914linkgit:git-config[1]:29152916-------------------------------------------------2917$ git config -l2918core.repositoryformatversion=02919core.filemode=true2920core.logallrefupdates=true2921remote.origin.url=git://git.kernel.org/pub/scm/git/git.git2922remote.origin.fetch=+refs/heads/*:refs/remotes/origin/*2923branch.master.remote=origin2924branch.master.merge=refs/heads/master2925-------------------------------------------------29262927If there are other repositories that you also use frequently, you can2928create similar configuration options to save typing; for example,29292930-------------------------------------------------2931$ git remote add example git://example.com/proj.git2932-------------------------------------------------29332934adds the following to `.git/config`:29352936-------------------------------------------------2937[remote "example"]2938 url = git://example.com/proj.git2939 fetch = +refs/heads/*:refs/remotes/example/*2940-------------------------------------------------29412942Also note that the above configuration can be performed by directly2943editing the file `.git/config` instead of using linkgit:git-remote[1].29442945After configuring the remote, the following three commands will do the2946same thing:29472948-------------------------------------------------2949$ git fetch git://example.com/proj.git +refs/heads/*:refs/remotes/example/*2950$ git fetch example +refs/heads/*:refs/remotes/example/*2951$ git fetch example2952-------------------------------------------------29532954See linkgit:git-config[1] for more details on the configuration2955options mentioned above and linkgit:git-fetch[1] for more details on2956the refspec syntax.295729582959[[git-concepts]]2960Git concepts2961============29622963Git is built on a small number of simple but powerful ideas. While it2964is possible to get things done without understanding them, you will find2965Git much more intuitive if you do.29662967We start with the most important, the <<def_object_database,object2968database>> and the <<def_index,index>>.29692970[[the-object-database]]2971The Object Database2972-------------------297329742975We already saw in <<understanding-commits>> that all commits are stored2976under a 40-digit "object name". In fact, all the information needed to2977represent the history of a project is stored in objects with such names.2978In each case the name is calculated by taking the SHA-1 hash of the2979contents of the object. The SHA-1 hash is a cryptographic hash function.2980What that means to us is that it is impossible to find two different2981objects with the same name. This has a number of advantages; among2982others:29832984- Git can quickly determine whether two objects are identical or not,2985 just by comparing names.2986- Since object names are computed the same way in every repository, the2987 same content stored in two repositories will always be stored under2988 the same name.2989- Git can detect errors when it reads an object, by checking that the2990 object's name is still the SHA-1 hash of its contents.29912992(See <<object-details>> for the details of the object formatting and2993SHA-1 calculation.)29942995There are four different types of objects: "blob", "tree", "commit", and2996"tag".29972998- A <<def_blob_object,"blob" object>> is used to store file data.2999- A <<def_tree_object,"tree" object>> ties one or more3000 "blob" objects into a directory structure. In addition, a tree object3001 can refer to other tree objects, thus creating a directory hierarchy.3002- A <<def_commit_object,"commit" object>> ties such directory hierarchies3003 together into a <<def_DAG,directed acyclic graph>> of revisions--each3004 commit contains the object name of exactly one tree designating the3005 directory hierarchy at the time of the commit. In addition, a commit3006 refers to "parent" commit objects that describe the history of how we3007 arrived at that directory hierarchy.3008- A <<def_tag_object,"tag" object>> symbolically identifies and can be3009 used to sign other objects. It contains the object name and type of3010 another object, a symbolic name (of course!) and, optionally, a3011 signature.30123013The object types in some more detail:30143015[[commit-object]]3016Commit Object3017~~~~~~~~~~~~~30183019The "commit" object links a physical state of a tree with a description3020of how we got there and why. Use the `--pretty=raw` option to3021linkgit:git-show[1] or linkgit:git-log[1] to examine your favorite3022commit:30233024------------------------------------------------3025$ git show -s --pretty=raw 2be7fcb4763026commit 2be7fcb4764f2dbcee52635b91fedb1b3dcf7ab43027tree fb3a8bdd0ceddd019615af4d57a53f43d8cee2bf3028parent 257a84d9d02e90447b149af58b271c19405edb6a3029author Dave Watson <dwatson@mimvista.com> 1187576872 -04003030committer Junio C Hamano <gitster@pobox.com> 1187591163 -070030313032 Fix misspelling of 'suppress' in docs30333034 Signed-off-by: Junio C Hamano <gitster@pobox.com>3035------------------------------------------------30363037As you can see, a commit is defined by:30383039- a tree: The SHA-1 name of a tree object (as defined below), representing3040 the contents of a directory at a certain point in time.3041- parent(s): The SHA-1 name(s) of some number of commits which represent the3042 immediately previous step(s) in the history of the project. The3043 example above has one parent; merge commits may have more than3044 one. A commit with no parents is called a "root" commit, and3045 represents the initial revision of a project. Each project must have3046 at least one root. A project can also have multiple roots, though3047 that isn't common (or necessarily a good idea).3048- an author: The name of the person responsible for this change, together3049 with its date.3050- a committer: The name of the person who actually created the commit,3051 with the date it was done. This may be different from the author, for3052 example, if the author was someone who wrote a patch and emailed it3053 to the person who used it to create the commit.3054- a comment describing this commit.30553056Note that a commit does not itself contain any information about what3057actually changed; all changes are calculated by comparing the contents3058of the tree referred to by this commit with the trees associated with3059its parents. In particular, Git does not attempt to record file renames3060explicitly, though it can identify cases where the existence of the same3061file data at changing paths suggests a rename. (See, for example, the3062`-M` option to linkgit:git-diff[1]).30633064A commit is usually created by linkgit:git-commit[1], which creates a3065commit whose parent is normally the current HEAD, and whose tree is3066taken from the content currently stored in the index.30673068[[tree-object]]3069Tree Object3070~~~~~~~~~~~30713072The ever-versatile linkgit:git-show[1] command can also be used to3073examine tree objects, but linkgit:git-ls-tree[1] will give you more3074details:30753076------------------------------------------------3077$ git ls-tree fb3a8bdd0ce3078100644 blob 63c918c667fa005ff12ad89437f2fdc80926e21c .gitignore3079100644 blob 5529b198e8d14decbe4ad99db3f7fb632de0439d .mailmap3080100644 blob 6ff87c4664981e4397625791c8ea3bbb5f2279a3 COPYING3081040000 tree 2fb783e477100ce076f6bf57e4a6f026013dc745 Documentation3082100755 blob 3c0032cec592a765692234f1cba47dfdcc3a9200 GIT-VERSION-GEN3083100644 blob 289b046a443c0647624607d471289b2c7dcd470b INSTALL3084100644 blob 4eb463797adc693dc168b926b6932ff53f17d0b1 Makefile3085100644 blob 548142c327a6790ff8821d67c2ee1eff7a656b52 README3086...3087------------------------------------------------30883089As you can see, a tree object contains a list of entries, each with a3090mode, object type, SHA-1 name, and name, sorted by name. It represents3091the contents of a single directory tree.30923093The object type may be a blob, representing the contents of a file, or3094another tree, representing the contents of a subdirectory. Since trees3095and blobs, like all other objects, are named by the SHA-1 hash of their3096contents, two trees have the same SHA-1 name if and only if their3097contents (including, recursively, the contents of all subdirectories)3098are identical. This allows Git to quickly determine the differences3099between two related tree objects, since it can ignore any entries with3100identical object names.31013102(Note: in the presence of submodules, trees may also have commits as3103entries. See <<submodules>> for documentation.)31043105Note that the files all have mode 644 or 755: Git actually only pays3106attention to the executable bit.31073108[[blob-object]]3109Blob Object3110~~~~~~~~~~~31113112You can use linkgit:git-show[1] to examine the contents of a blob; take,3113for example, the blob in the entry for `COPYING` from the tree above:31143115------------------------------------------------3116$ git show 6ff87c466431173118 Note that the only valid version of the GPL as far as this project3119 is concerned is _this_ particular version of the license (ie v2, not3120 v2.2 or v3.x or whatever), unless explicitly otherwise stated.3121...3122------------------------------------------------31233124A "blob" object is nothing but a binary blob of data. It doesn't refer3125to anything else or have attributes of any kind.31263127Since the blob is entirely defined by its data, if two files in a3128directory tree (or in multiple different versions of the repository)3129have the same contents, they will share the same blob object. The object3130is totally independent of its location in the directory tree, and3131renaming a file does not change the object that file is associated with.31323133Note that any tree or blob object can be examined using3134linkgit:git-show[1] with the <revision>:<path> syntax. This can3135sometimes be useful for browsing the contents of a tree that is not3136currently checked out.31373138[[trust]]3139Trust3140~~~~~31413142If you receive the SHA-1 name of a blob from one source, and its contents3143from another (possibly untrusted) source, you can still trust that those3144contents are correct as long as the SHA-1 name agrees. This is because3145the SHA-1 is designed so that it is infeasible to find different contents3146that produce the same hash.31473148Similarly, you need only trust the SHA-1 name of a top-level tree object3149to trust the contents of the entire directory that it refers to, and if3150you receive the SHA-1 name of a commit from a trusted source, then you3151can easily verify the entire history of commits reachable through3152parents of that commit, and all of those contents of the trees referred3153to by those commits.31543155So to introduce some real trust in the system, the only thing you need3156to do is to digitally sign just 'one' special note, which includes the3157name of a top-level commit. Your digital signature shows others3158that you trust that commit, and the immutability of the history of3159commits tells others that they can trust the whole history.31603161In other words, you can easily validate a whole archive by just3162sending out a single email that tells the people the name (SHA-1 hash)3163of the top commit, and digitally sign that email using something3164like GPG/PGP.31653166To assist in this, Git also provides the tag object...31673168[[tag-object]]3169Tag Object3170~~~~~~~~~~31713172A tag object contains an object, object type, tag name, the name of the3173person ("tagger") who created the tag, and a message, which may contain3174a signature, as can be seen using linkgit:git-cat-file[1]:31753176------------------------------------------------3177$ git cat-file tag v1.5.03178object 437b1b20df4b356c9342dac8d38849f24ef44f273179type commit3180tag v1.5.03181tagger Junio C Hamano <junkio@cox.net> 1171411200 +000031823183GIT 1.5.03184-----BEGIN PGP SIGNATURE-----3185Version: GnuPG v1.4.6 (GNU/Linux)31863187iD8DBQBF0lGqwMbZpPMRm5oRAuRiAJ9ohBLd7s2kqjkKlq1qqC57SbnmzQCdG4ui3188nLE/L9aUXdWeTFPron96DLA=3189=2E+03190-----END PGP SIGNATURE-----3191------------------------------------------------31923193See the linkgit:git-tag[1] command to learn how to create and verify tag3194objects. (Note that linkgit:git-tag[1] can also be used to create3195"lightweight tags", which are not tag objects at all, but just simple3196references whose names begin with `refs/tags/`).31973198[[pack-files]]3199How Git stores objects efficiently: pack files3200~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~32013202Newly created objects are initially created in a file named after the3203object's SHA-1 hash (stored in `.git/objects`).32043205Unfortunately this system becomes inefficient once a project has a3206lot of objects. Try this on an old project:32073208------------------------------------------------3209$ git count-objects32106930 objects, 47620 kilobytes3211------------------------------------------------32123213The first number is the number of objects which are kept in3214individual files. The second is the amount of space taken up by3215those "loose" objects.32163217You can save space and make Git faster by moving these loose objects in3218to a "pack file", which stores a group of objects in an efficient3219compressed format; the details of how pack files are formatted can be3220found in link:technical/pack-format.html[pack format].32213222To put the loose objects into a pack, just run git repack:32233224------------------------------------------------3225$ git repack3226Counting objects: 6020, done.3227Delta compression using up to 4 threads.3228Compressing objects: 100% (6020/6020), done.3229Writing objects: 100% (6020/6020), done.3230Total 6020 (delta 4070), reused 0 (delta 0)3231------------------------------------------------32323233This creates a single "pack file" in .git/objects/pack/3234containing all currently unpacked objects. You can then run32353236------------------------------------------------3237$ git prune3238------------------------------------------------32393240to remove any of the "loose" objects that are now contained in the3241pack. This will also remove any unreferenced objects (which may be3242created when, for example, you use `git reset` to remove a commit).3243You can verify that the loose objects are gone by looking at the3244`.git/objects` directory or by running32453246------------------------------------------------3247$ git count-objects32480 objects, 0 kilobytes3249------------------------------------------------32503251Although the object files are gone, any commands that refer to those3252objects will work exactly as they did before.32533254The linkgit:git-gc[1] command performs packing, pruning, and more for3255you, so is normally the only high-level command you need.32563257[[dangling-objects]]3258Dangling objects3259~~~~~~~~~~~~~~~~32603261The linkgit:git-fsck[1] command will sometimes complain about dangling3262objects. They are not a problem.32633264The most common cause of dangling objects is that you've rebased a3265branch, or you have pulled from somebody else who rebased a branch--see3266<<cleaning-up-history>>. In that case, the old head of the original3267branch still exists, as does everything it pointed to. The branch3268pointer itself just doesn't, since you replaced it with another one.32693270There are also other situations that cause dangling objects. For3271example, a "dangling blob" may arise because you did a `git add` of a3272file, but then, before you actually committed it and made it part of the3273bigger picture, you changed something else in that file and committed3274that *updated* thing--the old state that you added originally ends up3275not being pointed to by any commit or tree, so it's now a dangling blob3276object.32773278Similarly, when the "recursive" merge strategy runs, and finds that3279there are criss-cross merges and thus more than one merge base (which is3280fairly unusual, but it does happen), it will generate one temporary3281midway tree (or possibly even more, if you had lots of criss-crossing3282merges and more than two merge bases) as a temporary internal merge3283base, and again, those are real objects, but the end result will not end3284up pointing to them, so they end up "dangling" in your repository.32853286Generally, dangling objects aren't anything to worry about. They can3287even be very useful: if you screw something up, the dangling objects can3288be how you recover your old tree (say, you did a rebase, and realized3289that you really didn't want to--you can look at what dangling objects3290you have, and decide to reset your head to some old dangling state).32913292For commits, you can just use:32933294------------------------------------------------3295$ gitk <dangling-commit-sha-goes-here> --not --all3296------------------------------------------------32973298This asks for all the history reachable from the given commit but not3299from any branch, tag, or other reference. If you decide it's something3300you want, you can always create a new reference to it, e.g.,33013302------------------------------------------------3303$ git branch recovered-branch <dangling-commit-sha-goes-here>3304------------------------------------------------33053306For blobs and trees, you can't do the same, but you can still examine3307them. You can just do33083309------------------------------------------------3310$ git show <dangling-blob/tree-sha-goes-here>3311------------------------------------------------33123313to show what the contents of the blob were (or, for a tree, basically3314what the `ls` for that directory was), and that may give you some idea3315of what the operation was that left that dangling object.33163317Usually, dangling blobs and trees aren't very interesting. They're3318almost always the result of either being a half-way mergebase (the blob3319will often even have the conflict markers from a merge in it, if you3320have had conflicting merges that you fixed up by hand), or simply3321because you interrupted a `git fetch` with ^C or something like that,3322leaving _some_ of the new objects in the object database, but just3323dangling and useless.33243325Anyway, once you are sure that you're not interested in any dangling3326state, you can just prune all unreachable objects:33273328------------------------------------------------3329$ git prune3330------------------------------------------------33313332and they'll be gone. (You should only run `git prune` on a quiescent3333repository--it's kind of like doing a filesystem fsck recovery: you3334don't want to do that while the filesystem is mounted.3335`git prune` is designed not to cause any harm in such cases of concurrent3336accesses to a repository but you might receive confusing or scary messages.)33373338[[recovering-from-repository-corruption]]3339Recovering from repository corruption3340~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~33413342By design, Git treats data trusted to it with caution. However, even in3343the absence of bugs in Git itself, it is still possible that hardware or3344operating system errors could corrupt data.33453346The first defense against such problems is backups. You can back up a3347Git directory using clone, or just using cp, tar, or any other backup3348mechanism.33493350As a last resort, you can search for the corrupted objects and attempt3351to replace them by hand. Back up your repository before attempting this3352in case you corrupt things even more in the process.33533354We'll assume that the problem is a single missing or corrupted blob,3355which is sometimes a solvable problem. (Recovering missing trees and3356especially commits is *much* harder).33573358Before starting, verify that there is corruption, and figure out where3359it is with linkgit:git-fsck[1]; this may be time-consuming.33603361Assume the output looks like this:33623363------------------------------------------------3364$ git fsck --full --no-dangling3365broken link from tree 2d9263c6d23595e7cb2a21e5ebbb53655278dff83366 to blob 4b9458b3786228369c63936db65827de3cc062003367missing blob 4b9458b3786228369c63936db65827de3cc062003368------------------------------------------------33693370Now you know that blob 4b9458b3 is missing, and that the tree 2d9263c63371points to it. If you could find just one copy of that missing blob3372object, possibly in some other repository, you could move it into3373`.git/objects/4b/9458b3...` and be done. Suppose you can't. You can3374still examine the tree that pointed to it with linkgit:git-ls-tree[1],3375which might output something like:33763377------------------------------------------------3378$ git ls-tree 2d9263c6d23595e7cb2a21e5ebbb53655278dff83379100644 blob 8d14531846b95bfa3564b58ccfb7913a034323b8 .gitignore3380100644 blob ebf9bf84da0aab5ed944264a5db2a65fe3a3e883 .mailmap3381100644 blob ca442d313d86dc67e0a2e5d584b465bd382cbf5c COPYING3382...3383100644 blob 4b9458b3786228369c63936db65827de3cc06200 myfile3384...3385------------------------------------------------33863387So now you know that the missing blob was the data for a file named3388`myfile`. And chances are you can also identify the directory--let's3389say it's in `somedirectory`. If you're lucky the missing copy might be3390the same as the copy you have checked out in your working tree at3391`somedirectory/myfile`; you can test whether that's right with3392linkgit:git-hash-object[1]:33933394------------------------------------------------3395$ git hash-object -w somedirectory/myfile3396------------------------------------------------33973398which will create and store a blob object with the contents of3399somedirectory/myfile, and output the SHA-1 of that object. if you're3400extremely lucky it might be 4b9458b3786228369c63936db65827de3cc06200, in3401which case you've guessed right, and the corruption is fixed!34023403Otherwise, you need more information. How do you tell which version of3404the file has been lost?34053406The easiest way to do this is with:34073408------------------------------------------------3409$ git log --raw --all --full-history -- somedirectory/myfile3410------------------------------------------------34113412Because you're asking for raw output, you'll now get something like34133414------------------------------------------------3415commit abc3416Author:3417Date:3418...3419:100644 100644 4b9458b newsha M somedirectory/myfile342034213422commit xyz3423Author:3424Date:34253426...3427:100644 100644 oldsha 4b9458b M somedirectory/myfile3428------------------------------------------------34293430This tells you that the immediately following version of the file was3431"newsha", and that the immediately preceding version was "oldsha".3432You also know the commit messages that went with the change from oldsha3433to 4b9458b and with the change from 4b9458b to newsha.34343435If you've been committing small enough changes, you may now have a good3436shot at reconstructing the contents of the in-between state 4b9458b.34373438If you can do that, you can now recreate the missing object with34393440------------------------------------------------3441$ git hash-object -w <recreated-file>3442------------------------------------------------34433444and your repository is good again!34453446(Btw, you could have ignored the `fsck`, and started with doing a34473448------------------------------------------------3449$ git log --raw --all3450------------------------------------------------34513452and just looked for the sha of the missing object (4b9458b) in that3453whole thing. It's up to you--Git does *have* a lot of information, it is3454just missing one particular blob version.34553456[[the-index]]3457The index3458---------34593460The index is a binary file (generally kept in `.git/index`) containing a3461sorted list of path names, each with permissions and the SHA-1 of a blob3462object; linkgit:git-ls-files[1] can show you the contents of the index:34633464-------------------------------------------------3465$ git ls-files --stage3466100644 63c918c667fa005ff12ad89437f2fdc80926e21c 0 .gitignore3467100644 5529b198e8d14decbe4ad99db3f7fb632de0439d 0 .mailmap3468100644 6ff87c4664981e4397625791c8ea3bbb5f2279a3 0 COPYING3469100644 a37b2152bd26be2c2289e1f57a292534a51a93c7 0 Documentation/.gitignore3470100644 fbefe9a45b00a54b58d94d06eca48b03d40a50e0 0 Documentation/Makefile3471...3472100644 2511aef8d89ab52be5ec6a5e46236b4b6bcd07ea 0 xdiff/xtypes.h3473100644 2ade97b2574a9f77e7ae4002a4e07a6a38e46d07 0 xdiff/xutils.c3474100644 d5de8292e05e7c36c4b68857c1cf9855e3d2f70a 0 xdiff/xutils.h3475-------------------------------------------------34763477Note that in older documentation you may see the index called the3478"current directory cache" or just the "cache". It has three important3479properties:348034811. The index contains all the information necessary to generate a single3482(uniquely determined) tree object.3483+3484For example, running linkgit:git-commit[1] generates this tree object3485from the index, stores it in the object database, and uses it as the3486tree object associated with the new commit.348734882. The index enables fast comparisons between the tree object it defines3489and the working tree.3490+3491It does this by storing some additional data for each entry (such as3492the last modified time). This data is not displayed above, and is not3493stored in the created tree object, but it can be used to determine3494quickly which files in the working directory differ from what was3495stored in the index, and thus save Git from having to read all of the3496data from such files to look for changes.349734983. It can efficiently represent information about merge conflicts3499between different tree objects, allowing each pathname to be3500associated with sufficient information about the trees involved that3501you can create a three-way merge between them.3502+3503We saw in <<conflict-resolution>> that during a merge the index can3504store multiple versions of a single file (called "stages"). The third3505column in the linkgit:git-ls-files[1] output above is the stage3506number, and will take on values other than 0 for files with merge3507conflicts.35083509The index is thus a sort of temporary staging area, which is filled with3510a tree which you are in the process of working on.35113512If you blow the index away entirely, you generally haven't lost any3513information as long as you have the name of the tree that it described.35143515[[submodules]]3516Submodules3517==========35183519Large projects are often composed of smaller, self-contained modules. For3520example, an embedded Linux distribution's source tree would include every3521piece of software in the distribution with some local modifications; a movie3522player might need to build against a specific, known-working version of a3523decompression library; several independent programs might all share the same3524build scripts.35253526With centralized revision control systems this is often accomplished by3527including every module in one single repository. Developers can check out3528all modules or only the modules they need to work with. They can even modify3529files across several modules in a single commit while moving things around3530or updating APIs and translations.35313532Git does not allow partial checkouts, so duplicating this approach in Git3533would force developers to keep a local copy of modules they are not3534interested in touching. Commits in an enormous checkout would be slower3535than you'd expect as Git would have to scan every directory for changes.3536If modules have a lot of local history, clones would take forever.35373538On the plus side, distributed revision control systems can much better3539integrate with external sources. In a centralized model, a single arbitrary3540snapshot of the external project is exported from its own revision control3541and then imported into the local revision control on a vendor branch. All3542the history is hidden. With distributed revision control you can clone the3543entire external history and much more easily follow development and re-merge3544local changes.35453546Git's submodule support allows a repository to contain, as a subdirectory, a3547checkout of an external project. Submodules maintain their own identity;3548the submodule support just stores the submodule repository location and3549commit ID, so other developers who clone the containing project3550("superproject") can easily clone all the submodules at the same revision.3551Partial checkouts of the superproject are possible: you can tell Git to3552clone none, some or all of the submodules.35533554The linkgit:git-submodule[1] command is available since Git 1.5.3. Users3555with Git 1.5.2 can look up the submodule commits in the repository and3556manually check them out; earlier versions won't recognize the submodules at3557all.35583559To see how submodule support works, create four example3560repositories that can be used later as a submodule:35613562-------------------------------------------------3563$ mkdir ~/git3564$ cd ~/git3565$ for i in a b c d3566do3567 mkdir $i3568 cd $i3569 git init3570 echo "module $i" > $i.txt3571 git add $i.txt3572 git commit -m "Initial commit, submodule $i"3573 cd ..3574done3575-------------------------------------------------35763577Now create the superproject and add all the submodules:35783579-------------------------------------------------3580$ mkdir super3581$ cd super3582$ git init3583$ for i in a b c d3584do3585 git submodule add ~/git/$i $i3586done3587-------------------------------------------------35883589NOTE: Do not use local URLs here if you plan to publish your superproject!35903591See what files `git submodule` created:35923593-------------------------------------------------3594$ ls -a3595. .. .git .gitmodules a b c d3596-------------------------------------------------35973598The `git submodule add <repo> <path>` command does a couple of things:35993600- It clones the submodule from `<repo>` to the given `<path>` under the3601 current directory and by default checks out the master branch.3602- It adds the submodule's clone path to the linkgit:gitmodules[5] file and3603 adds this file to the index, ready to be committed.3604- It adds the submodule's current commit ID to the index, ready to be3605 committed.36063607Commit the superproject:36083609-------------------------------------------------3610$ git commit -m "Add submodules a, b, c and d."3611-------------------------------------------------36123613Now clone the superproject:36143615-------------------------------------------------3616$ cd ..3617$ git clone super cloned3618$ cd cloned3619-------------------------------------------------36203621The submodule directories are there, but they're empty:36223623-------------------------------------------------3624$ ls -a a3625. ..3626$ git submodule status3627-d266b9873ad50488163457f025db7cdd9683d88b a3628-e81d457da15309b4fef4249aba9b50187999670d b3629-c1536a972b9affea0f16e0680ba87332dc059146 c3630-d96249ff5d57de5de093e6baff9e0aafa5276a74 d3631-------------------------------------------------36323633NOTE: The commit object names shown above would be different for you, but they3634should match the HEAD commit object names of your repositories. You can check3635it by running `git ls-remote ../a`.36363637Pulling down the submodules is a two-step process. First run `git submodule3638init` to add the submodule repository URLs to `.git/config`:36393640-------------------------------------------------3641$ git submodule init3642-------------------------------------------------36433644Now use `git submodule update` to clone the repositories and check out the3645commits specified in the superproject:36463647-------------------------------------------------3648$ git submodule update3649$ cd a3650$ ls -a3651. .. .git a.txt3652-------------------------------------------------36533654One major difference between `git submodule update` and `git submodule add` is3655that `git submodule update` checks out a specific commit, rather than the tip3656of a branch. It's like checking out a tag: the head is detached, so you're not3657working on a branch.36583659-------------------------------------------------3660$ git branch3661* (detached from d266b98)3662 master3663-------------------------------------------------36643665If you want to make a change within a submodule and you have a detached head,3666then you should create or checkout a branch, make your changes, publish the3667change within the submodule, and then update the superproject to reference the3668new commit:36693670-------------------------------------------------3671$ git checkout master3672-------------------------------------------------36733674or36753676-------------------------------------------------3677$ git checkout -b fix-up3678-------------------------------------------------36793680then36813682-------------------------------------------------3683$ echo "adding a line again" >> a.txt3684$ git commit -a -m "Updated the submodule from within the superproject."3685$ git push3686$ cd ..3687$ git diff3688diff --git a/a b/a3689index d266b98..261dfac 1600003690--- a/a3691+++ b/a3692@@ -1 +1 @@3693-Subproject commit d266b9873ad50488163457f025db7cdd9683d88b3694+Subproject commit 261dfac35cb99d380eb966e102c1197139f7fa243695$ git add a3696$ git commit -m "Updated submodule a."3697$ git push3698-------------------------------------------------36993700You have to run `git submodule update` after `git pull` if you want to update3701submodules, too.37023703Pitfalls with submodules3704------------------------37053706Always publish the submodule change before publishing the change to the3707superproject that references it. If you forget to publish the submodule change,3708others won't be able to clone the repository:37093710-------------------------------------------------3711$ cd ~/git/super/a3712$ echo i added another line to this file >> a.txt3713$ git commit -a -m "doing it wrong this time"3714$ cd ..3715$ git add a3716$ git commit -m "Updated submodule a again."3717$ git push3718$ cd ~/git/cloned3719$ git pull3720$ git submodule update3721error: pathspec '261dfac35cb99d380eb966e102c1197139f7fa24' did not match any file(s) known to git.3722Did you forget to 'git add'?3723Unable to checkout '261dfac35cb99d380eb966e102c1197139f7fa24' in submodule path 'a'3724-------------------------------------------------37253726In older Git versions it could be easily forgotten to commit new or modified3727files in a submodule, which silently leads to similar problems as not pushing3728the submodule changes. Starting with Git 1.7.0 both `git status` and `git diff`3729in the superproject show submodules as modified when they contain new or3730modified files to protect against accidentally committing such a state. `git3731diff` will also add a `-dirty` to the work tree side when generating patch3732output or used with the `--submodule` option:37333734-------------------------------------------------3735$ git diff3736diff --git a/sub b/sub3737--- a/sub3738+++ b/sub3739@@ -1 +1 @@3740-Subproject commit 3f356705649b5d566d97ff843cf193359229a4533741+Subproject commit 3f356705649b5d566d97ff843cf193359229a453-dirty3742$ git diff --submodule3743Submodule sub 3f35670..3f35670-dirty:3744-------------------------------------------------37453746You also should not rewind branches in a submodule beyond commits that were3747ever recorded in any superproject.37483749It's not safe to run `git submodule update` if you've made and committed3750changes within a submodule without checking out a branch first. They will be3751silently overwritten:37523753-------------------------------------------------3754$ cat a.txt3755module a3756$ echo line added from private2 >> a.txt3757$ git commit -a -m "line added inside private2"3758$ cd ..3759$ git submodule update3760Submodule path 'a': checked out 'd266b9873ad50488163457f025db7cdd9683d88b'3761$ cd a3762$ cat a.txt3763module a3764-------------------------------------------------37653766NOTE: The changes are still visible in the submodule's reflog.37673768If you have uncommitted changes in your submodule working tree, `git3769submodule update` will not overwrite them. Instead, you get the usual3770warning about not being able switch from a dirty branch.37713772[[low-level-operations]]3773Low-level Git operations3774========================37753776Many of the higher-level commands were originally implemented as shell3777scripts using a smaller core of low-level Git commands. These can still3778be useful when doing unusual things with Git, or just as a way to3779understand its inner workings.37803781[[object-manipulation]]3782Object access and manipulation3783------------------------------37843785The linkgit:git-cat-file[1] command can show the contents of any object,3786though the higher-level linkgit:git-show[1] is usually more useful.37873788The linkgit:git-commit-tree[1] command allows constructing commits with3789arbitrary parents and trees.37903791A tree can be created with linkgit:git-write-tree[1] and its data can be3792accessed by linkgit:git-ls-tree[1]. Two trees can be compared with3793linkgit:git-diff-tree[1].37943795A tag is created with linkgit:git-mktag[1], and the signature can be3796verified by linkgit:git-verify-tag[1], though it is normally simpler to3797use linkgit:git-tag[1] for both.37983799[[the-workflow]]3800The Workflow3801------------38023803High-level operations such as linkgit:git-commit[1],3804linkgit:git-checkout[1] and linkgit:git-reset[1] work by moving data3805between the working tree, the index, and the object database. Git3806provides low-level operations which perform each of these steps3807individually.38083809Generally, all Git operations work on the index file. Some operations3810work *purely* on the index file (showing the current state of the3811index), but most operations move data between the index file and either3812the database or the working directory. Thus there are four main3813combinations:38143815[[working-directory-to-index]]3816working directory -> index3817~~~~~~~~~~~~~~~~~~~~~~~~~~38183819The linkgit:git-update-index[1] command updates the index with3820information from the working directory. You generally update the3821index information by just specifying the filename you want to update,3822like so:38233824-------------------------------------------------3825$ git update-index filename3826-------------------------------------------------38273828but to avoid common mistakes with filename globbing etc., the command3829will not normally add totally new entries or remove old entries,3830i.e. it will normally just update existing cache entries.38313832To tell Git that yes, you really do realize that certain files no3833longer exist, or that new files should be added, you3834should use the `--remove` and `--add` flags respectively.38353836NOTE! A `--remove` flag does 'not' mean that subsequent filenames will3837necessarily be removed: if the files still exist in your directory3838structure, the index will be updated with their new status, not3839removed. The only thing `--remove` means is that update-index will be3840considering a removed file to be a valid thing, and if the file really3841does not exist any more, it will update the index accordingly.38423843As a special case, you can also do `git update-index --refresh`, which3844will refresh the "stat" information of each index to match the current3845stat information. It will 'not' update the object status itself, and3846it will only update the fields that are used to quickly test whether3847an object still matches its old backing store object.38483849The previously introduced linkgit:git-add[1] is just a wrapper for3850linkgit:git-update-index[1].38513852[[index-to-object-database]]3853index -> object database3854~~~~~~~~~~~~~~~~~~~~~~~~38553856You write your current index file to a "tree" object with the program38573858-------------------------------------------------3859$ git write-tree3860-------------------------------------------------38613862that doesn't come with any options--it will just write out the3863current index into the set of tree objects that describe that state,3864and it will return the name of the resulting top-level tree. You can3865use that tree to re-generate the index at any time by going in the3866other direction:38673868[[object-database-to-index]]3869object database -> index3870~~~~~~~~~~~~~~~~~~~~~~~~38713872You read a "tree" file from the object database, and use that to3873populate (and overwrite--don't do this if your index contains any3874unsaved state that you might want to restore later!) your current3875index. Normal operation is just38763877-------------------------------------------------3878$ git read-tree <SHA-1 of tree>3879-------------------------------------------------38803881and your index file will now be equivalent to the tree that you saved3882earlier. However, that is only your 'index' file: your working3883directory contents have not been modified.38843885[[index-to-working-directory]]3886index -> working directory3887~~~~~~~~~~~~~~~~~~~~~~~~~~38883889You update your working directory from the index by "checking out"3890files. This is not a very common operation, since normally you'd just3891keep your files updated, and rather than write to your working3892directory, you'd tell the index files about the changes in your3893working directory (i.e. `git update-index`).38943895However, if you decide to jump to a new version, or check out somebody3896else's version, or just restore a previous tree, you'd populate your3897index file with read-tree, and then you need to check out the result3898with38993900-------------------------------------------------3901$ git checkout-index filename3902-------------------------------------------------39033904or, if you want to check out all of the index, use `-a`.39053906NOTE! `git checkout-index` normally refuses to overwrite old files, so3907if you have an old version of the tree already checked out, you will3908need to use the `-f` flag ('before' the `-a` flag or the filename) to3909'force' the checkout.391039113912Finally, there are a few odds and ends which are not purely moving3913from one representation to the other:39143915[[tying-it-all-together]]3916Tying it all together3917~~~~~~~~~~~~~~~~~~~~~39183919To commit a tree you have instantiated with `git write-tree`, you'd3920create a "commit" object that refers to that tree and the history3921behind it--most notably the "parent" commits that preceded it in3922history.39233924Normally a "commit" has one parent: the previous state of the tree3925before a certain change was made. However, sometimes it can have two3926or more parent commits, in which case we call it a "merge", due to the3927fact that such a commit brings together ("merges") two or more3928previous states represented by other commits.39293930In other words, while a "tree" represents a particular directory state3931of a working directory, a "commit" represents that state in time,3932and explains how we got there.39333934You create a commit object by giving it the tree that describes the3935state at the time of the commit, and a list of parents:39363937-------------------------------------------------3938$ git commit-tree <tree> -p <parent> [(-p <parent2>)...]3939-------------------------------------------------39403941and then giving the reason for the commit on stdin (either through3942redirection from a pipe or file, or by just typing it at the tty).39433944`git commit-tree` will return the name of the object that represents3945that commit, and you should save it away for later use. Normally,3946you'd commit a new `HEAD` state, and while Git doesn't care where you3947save the note about that state, in practice we tend to just write the3948result to the file pointed at by `.git/HEAD`, so that we can always see3949what the last committed state was.39503951Here is a picture that illustrates how various pieces fit together:39523953------------39543955 commit-tree3956 commit obj3957 +----+3958 | |3959 | |3960 V V3961 +-----------+3962 | Object DB |3963 | Backing |3964 | Store |3965 +-----------+3966 ^3967 write-tree | |3968 tree obj | |3969 | | read-tree3970 | | tree obj3971 V3972 +-----------+3973 | Index |3974 | "cache" |3975 +-----------+3976 update-index ^3977 blob obj | |3978 | |3979 checkout-index -u | | checkout-index3980 stat | | blob obj3981 V3982 +-----------+3983 | Working |3984 | Directory |3985 +-----------+39863987------------398839893990[[examining-the-data]]3991Examining the data3992------------------39933994You can examine the data represented in the object database and the3995index with various helper tools. For every object, you can use3996linkgit:git-cat-file[1] to examine details about the3997object:39983999-------------------------------------------------4000$ git cat-file -t <objectname>4001-------------------------------------------------40024003shows the type of the object, and once you have the type (which is4004usually implicit in where you find the object), you can use40054006-------------------------------------------------4007$ git cat-file blob|tree|commit|tag <objectname>4008-------------------------------------------------40094010to show its contents. NOTE! Trees have binary content, and as a result4011there is a special helper for showing that content, called4012`git ls-tree`, which turns the binary content into a more easily4013readable form.40144015It's especially instructive to look at "commit" objects, since those4016tend to be small and fairly self-explanatory. In particular, if you4017follow the convention of having the top commit name in `.git/HEAD`,4018you can do40194020-------------------------------------------------4021$ git cat-file commit HEAD4022-------------------------------------------------40234024to see what the top commit was.40254026[[merging-multiple-trees]]4027Merging multiple trees4028----------------------40294030Git can help you perform a three-way merge, which can in turn be4031used for a many-way merge by repeating the merge procedure several4032times. The usual situation is that you only do one three-way merge4033(reconciling two lines of history) and commit the result, but if4034you like to, you can merge several branches in one go.40354036To perform a three-way merge, you start with the two commits you4037want to merge, find their closest common parent (a third commit),4038and compare the trees corresponding to these three commits.40394040To get the "base" for the merge, look up the common parent of two4041commits:40424043-------------------------------------------------4044$ git merge-base <commit1> <commit2>4045-------------------------------------------------40464047This prints the name of a commit they are both based on. You should4048now look up the tree objects of those commits, which you can easily4049do with40504051-------------------------------------------------4052$ git cat-file commit <commitname> | head -14053-------------------------------------------------40544055since the tree object information is always the first line in a commit4056object.40574058Once you know the three trees you are going to merge (the one "original"4059tree, aka the common tree, and the two "result" trees, aka the branches4060you want to merge), you do a "merge" read into the index. This will4061complain if it has to throw away your old index contents, so you should4062make sure that you've committed those--in fact you would normally4063always do a merge against your last commit (which should thus match what4064you have in your current index anyway).40654066To do the merge, do40674068-------------------------------------------------4069$ git read-tree -m -u <origtree> <yourtree> <targettree>4070-------------------------------------------------40714072which will do all trivial merge operations for you directly in the4073index file, and you can just write the result out with4074`git write-tree`.407540764077[[merging-multiple-trees-2]]4078Merging multiple trees, continued4079---------------------------------40804081Sadly, many merges aren't trivial. If there are files that have4082been added, moved or removed, or if both branches have modified the4083same file, you will be left with an index tree that contains "merge4084entries" in it. Such an index tree can 'NOT' be written out to a tree4085object, and you will have to resolve any such merge clashes using4086other tools before you can write out the result.40874088You can examine such index state with `git ls-files --unmerged`4089command. An example:40904091------------------------------------------------4092$ git read-tree -m $orig HEAD $target4093$ git ls-files --unmerged4094100644 263414f423d0e4d70dae8fe53fa34614ff3e2860 1 hello.c4095100644 06fa6a24256dc7e560efa5687fa84b51f0263c3a 2 hello.c4096100644 cc44c73eb783565da5831b4d820c962954019b69 3 hello.c4097------------------------------------------------40984099Each line of the `git ls-files --unmerged` output begins with4100the blob mode bits, blob SHA-1, 'stage number', and the4101filename. The 'stage number' is Git's way to say which tree it4102came from: stage 1 corresponds to the `$orig` tree, stage 2 to4103the `HEAD` tree, and stage 3 to the `$target` tree.41044105Earlier we said that trivial merges are done inside4106`git read-tree -m`. For example, if the file did not change4107from `$orig` to `HEAD` or `$target`, or if the file changed4108from `$orig` to `HEAD` and `$orig` to `$target` the same way,4109obviously the final outcome is what is in `HEAD`. What the4110above example shows is that file `hello.c` was changed from4111`$orig` to `HEAD` and `$orig` to `$target` in a different way.4112You could resolve this by running your favorite 3-way merge4113program, e.g. `diff3`, `merge`, or Git's own merge-file, on4114the blob objects from these three stages yourself, like this:41154116------------------------------------------------4117$ git cat-file blob 263414f >hello.c~14118$ git cat-file blob 06fa6a2 >hello.c~24119$ git cat-file blob cc44c73 >hello.c~34120$ git merge-file hello.c~2 hello.c~1 hello.c~34121------------------------------------------------41224123This would leave the merge result in `hello.c~2` file, along4124with conflict markers if there are conflicts. After verifying4125the merge result makes sense, you can tell Git what the final4126merge result for this file is by:41274128-------------------------------------------------4129$ mv -f hello.c~2 hello.c4130$ git update-index hello.c4131-------------------------------------------------41324133When a path is in the "unmerged" state, running `git update-index` for4134that path tells Git to mark the path resolved.41354136The above is the description of a Git merge at the lowest level,4137to help you understand what conceptually happens under the hood.4138In practice, nobody, not even Git itself, runs `git cat-file` three times4139for this. There is a `git merge-index` program that extracts the4140stages to temporary files and calls a "merge" script on it:41414142-------------------------------------------------4143$ git merge-index git-merge-one-file hello.c4144-------------------------------------------------41454146and that is what higher level `git merge -s resolve` is implemented with.41474148[[hacking-git]]4149Hacking Git4150===========41514152This chapter covers internal details of the Git implementation which4153probably only Git developers need to understand.41544155[[object-details]]4156Object storage format4157---------------------41584159All objects have a statically determined "type" which identifies the4160format of the object (i.e. how it is used, and how it can refer to other4161objects). There are currently four different object types: "blob",4162"tree", "commit", and "tag".41634164Regardless of object type, all objects share the following4165characteristics: they are all deflated with zlib, and have a header4166that not only specifies their type, but also provides size information4167about the data in the object. It's worth noting that the SHA-1 hash4168that is used to name the object is the hash of the original data4169plus this header, so `sha1sum` 'file' does not match the object name4170for 'file'.41714172As a result, the general consistency of an object can always be tested4173independently of the contents or the type of the object: all objects can4174be validated by verifying that (a) their hashes match the content of the4175file and (b) the object successfully inflates to a stream of bytes that4176forms a sequence of4177`<ascii type without space> + <space> + <ascii decimal size> +4178<byte\0> + <binary object data>`.41794180The structured objects can further have their structure and4181connectivity to other objects verified. This is generally done with4182the `git fsck` program, which generates a full dependency graph4183of all objects, and verifies their internal consistency (in addition4184to just verifying their superficial consistency through the hash).41854186[[birdview-on-the-source-code]]4187A birds-eye view of Git's source code4188-------------------------------------41894190It is not always easy for new developers to find their way through Git's4191source code. This section gives you a little guidance to show where to4192start.41934194A good place to start is with the contents of the initial commit, with:41954196----------------------------------------------------4197$ git checkout e83c51634198----------------------------------------------------41994200The initial revision lays the foundation for almost everything Git has4201today, but is small enough to read in one sitting.42024203Note that terminology has changed since that revision. For example, the4204README in that revision uses the word "changeset" to describe what we4205now call a <<def_commit_object,commit>>.42064207Also, we do not call it "cache" any more, but rather "index"; however, the4208file is still called `cache.h`. Remark: Not much reason to change it now,4209especially since there is no good single name for it anyway, because it is4210basically _the_ header file which is included by _all_ of Git's C sources.42114212If you grasp the ideas in that initial commit, you should check out a4213more recent version and skim `cache.h`, `object.h` and `commit.h`.42144215In the early days, Git (in the tradition of UNIX) was a bunch of programs4216which were extremely simple, and which you used in scripts, piping the4217output of one into another. This turned out to be good for initial4218development, since it was easier to test new things. However, recently4219many of these parts have become builtins, and some of the core has been4220"libified", i.e. put into libgit.a for performance, portability reasons,4221and to avoid code duplication.42224223By now, you know what the index is (and find the corresponding data4224structures in `cache.h`), and that there are just a couple of object types4225(blobs, trees, commits and tags) which inherit their common structure from4226`struct object`, which is their first member (and thus, you can cast e.g.4227`(struct object *)commit` to achieve the _same_ as `&commit->object`, i.e.4228get at the object name and flags).42294230Now is a good point to take a break to let this information sink in.42314232Next step: get familiar with the object naming. Read <<naming-commits>>.4233There are quite a few ways to name an object (and not only revisions!).4234All of these are handled in `sha1_name.c`. Just have a quick look at4235the function `get_sha1()`. A lot of the special handling is done by4236functions like `get_sha1_basic()` or the likes.42374238This is just to get you into the groove for the most libified part of Git:4239the revision walker.42404241Basically, the initial version of `git log` was a shell script:42424243----------------------------------------------------------------4244$ git-rev-list --pretty $(git-rev-parse --default HEAD "$@") | \4245 LESS=-S ${PAGER:-less}4246----------------------------------------------------------------42474248What does this mean?42494250`git rev-list` is the original version of the revision walker, which4251_always_ printed a list of revisions to stdout. It is still functional,4252and needs to, since most new Git commands start out as scripts using4253`git rev-list`.42544255`git rev-parse` is not as important any more; it was only used to filter out4256options that were relevant for the different plumbing commands that were4257called by the script.42584259Most of what `git rev-list` did is contained in `revision.c` and4260`revision.h`. It wraps the options in a struct named `rev_info`, which4261controls how and what revisions are walked, and more.42624263The original job of `git rev-parse` is now taken by the function4264`setup_revisions()`, which parses the revisions and the common command-line4265options for the revision walker. This information is stored in the struct4266`rev_info` for later consumption. You can do your own command-line option4267parsing after calling `setup_revisions()`. After that, you have to call4268`prepare_revision_walk()` for initialization, and then you can get the4269commits one by one with the function `get_revision()`.42704271If you are interested in more details of the revision walking process,4272just have a look at the first implementation of `cmd_log()`; call4273`git show v1.3.0~155^2~4` and scroll down to that function (note that you4274no longer need to call `setup_pager()` directly).42754276Nowadays, `git log` is a builtin, which means that it is _contained_ in the4277command `git`. The source side of a builtin is42784279- a function called `cmd_<bla>`, typically defined in `builtin/<bla.c>`4280 (note that older versions of Git used to have it in `builtin-<bla>.c`4281 instead), and declared in `builtin.h`.42824283- an entry in the `commands[]` array in `git.c`, and42844285- an entry in `BUILTIN_OBJECTS` in the `Makefile`.42864287Sometimes, more than one builtin is contained in one source file. For4288example, `cmd_whatchanged()` and `cmd_log()` both reside in `builtin/log.c`,4289since they share quite a bit of code. In that case, the commands which are4290_not_ named like the `.c` file in which they live have to be listed in4291`BUILT_INS` in the `Makefile`.42924293`git log` looks more complicated in C than it does in the original script,4294but that allows for a much greater flexibility and performance.42954296Here again it is a good point to take a pause.42974298Lesson three is: study the code. Really, it is the best way to learn about4299the organization of Git (after you know the basic concepts).43004301So, think about something which you are interested in, say, "how can I4302access a blob just knowing the object name of it?". The first step is to4303find a Git command with which you can do it. In this example, it is either4304`git show` or `git cat-file`.43054306For the sake of clarity, let's stay with `git cat-file`, because it43074308- is plumbing, and43094310- was around even in the initial commit (it literally went only through4311 some 20 revisions as `cat-file.c`, was renamed to `builtin/cat-file.c`4312 when made a builtin, and then saw less than 10 versions).43134314So, look into `builtin/cat-file.c`, search for `cmd_cat_file()` and look what4315it does.43164317------------------------------------------------------------------4318 git_config(git_default_config);4319 if (argc != 3)4320 usage("git cat-file [-t|-s|-e|-p|<type>] <sha1>");4321 if (get_sha1(argv[2], sha1))4322 die("Not a valid object name %s", argv[2]);4323------------------------------------------------------------------43244325Let's skip over the obvious details; the only really interesting part4326here is the call to `get_sha1()`. It tries to interpret `argv[2]` as an4327object name, and if it refers to an object which is present in the current4328repository, it writes the resulting SHA-1 into the variable `sha1`.43294330Two things are interesting here:43314332- `get_sha1()` returns 0 on _success_. This might surprise some new4333 Git hackers, but there is a long tradition in UNIX to return different4334 negative numbers in case of different errors--and 0 on success.43354336- the variable `sha1` in the function signature of `get_sha1()` is `unsigned4337 char *`, but is actually expected to be a pointer to `unsigned4338 char[20]`. This variable will contain the 160-bit SHA-1 of the given4339 commit. Note that whenever a SHA-1 is passed as `unsigned char *`, it4340 is the binary representation, as opposed to the ASCII representation in4341 hex characters, which is passed as `char *`.43424343You will see both of these things throughout the code.43444345Now, for the meat:43464347-----------------------------------------------------------------------------4348 case 0:4349 buf = read_object_with_reference(sha1, argv[1], &size, NULL);4350-----------------------------------------------------------------------------43514352This is how you read a blob (actually, not only a blob, but any type of4353object). To know how the function `read_object_with_reference()` actually4354works, find the source code for it (something like `git grep4355read_object_with | grep ":[a-z]"` in the Git repository), and read4356the source.43574358To find out how the result can be used, just read on in `cmd_cat_file()`:43594360-----------------------------------4361 write_or_die(1, buf, size);4362-----------------------------------43634364Sometimes, you do not know where to look for a feature. In many such cases,4365it helps to search through the output of `git log`, and then `git show` the4366corresponding commit.43674368Example: If you know that there was some test case for `git bundle`, but4369do not remember where it was (yes, you _could_ `git grep bundle t/`, but that4370does not illustrate the point!):43714372------------------------4373$ git log --no-merges t/4374------------------------43754376In the pager (`less`), just search for "bundle", go a few lines back,4377and see that it is in commit 18449ab0. Now just copy this object name,4378and paste it into the command line43794380-------------------4381$ git show 18449ab04382-------------------43834384Voila.43854386Another example: Find out what to do in order to make some script a4387builtin:43884389-------------------------------------------------4390$ git log --no-merges --diff-filter=A builtin/*.c4391-------------------------------------------------43924393You see, Git is actually the best tool to find out about the source of Git4394itself!43954396[[glossary]]4397Git Glossary4398============43994400[[git-explained]]4401Git explained4402-------------44034404include::glossary-content.txt[]44054406[[git-quick-start]]4407Appendix A: Git Quick Reference4408===============================44094410This is a quick summary of the major commands; the previous chapters4411explain how these work in more detail.44124413[[quick-creating-a-new-repository]]4414Creating a new repository4415-------------------------44164417From a tarball:44184419-----------------------------------------------4420$ tar xzf project.tar.gz4421$ cd project4422$ git init4423Initialized empty Git repository in .git/4424$ git add .4425$ git commit4426-----------------------------------------------44274428From a remote repository:44294430-----------------------------------------------4431$ git clone git://example.com/pub/project.git4432$ cd project4433-----------------------------------------------44344435[[managing-branches]]4436Managing branches4437-----------------44384439-----------------------------------------------4440$ git branch # list all local branches in this repo4441$ git checkout test # switch working directory to branch "test"4442$ git branch new # create branch "new" starting at current HEAD4443$ git branch -d new # delete branch "new"4444-----------------------------------------------44454446Instead of basing a new branch on current HEAD (the default), use:44474448-----------------------------------------------4449$ git branch new test # branch named "test"4450$ git branch new v2.6.15 # tag named v2.6.154451$ git branch new HEAD^ # commit before the most recent4452$ git branch new HEAD^^ # commit before that4453$ git branch new test~10 # ten commits before tip of branch "test"4454-----------------------------------------------44554456Create and switch to a new branch at the same time:44574458-----------------------------------------------4459$ git checkout -b new v2.6.154460-----------------------------------------------44614462Update and examine branches from the repository you cloned from:44634464-----------------------------------------------4465$ git fetch # update4466$ git branch -r # list4467 origin/master4468 origin/next4469 ...4470$ git checkout -b masterwork origin/master4471-----------------------------------------------44724473Fetch a branch from a different repository, and give it a new4474name in your repository:44754476-----------------------------------------------4477$ git fetch git://example.com/project.git theirbranch:mybranch4478$ git fetch git://example.com/project.git v2.6.15:mybranch4479-----------------------------------------------44804481Keep a list of repositories you work with regularly:44824483-----------------------------------------------4484$ git remote add example git://example.com/project.git4485$ git remote # list remote repositories4486example4487origin4488$ git remote show example # get details4489* remote example4490 URL: git://example.com/project.git4491 Tracked remote branches4492 master4493 next4494 ...4495$ git fetch example # update branches from example4496$ git branch -r # list all remote branches4497-----------------------------------------------449844994500[[exploring-history]]4501Exploring history4502-----------------45034504-----------------------------------------------4505$ gitk # visualize and browse history4506$ git log # list all commits4507$ git log src/ # ...modifying src/4508$ git log v2.6.15..v2.6.16 # ...in v2.6.16, not in v2.6.154509$ git log master..test # ...in branch test, not in branch master4510$ git log test..master # ...in branch master, but not in test4511$ git log test...master # ...in one branch, not in both4512$ git log -S'foo()' # ...where difference contain "foo()"4513$ git log --since="2 weeks ago"4514$ git log -p # show patches as well4515$ git show # most recent commit4516$ git diff v2.6.15..v2.6.16 # diff between two tagged versions4517$ git diff v2.6.15..HEAD # diff with current head4518$ git grep "foo()" # search working directory for "foo()"4519$ git grep v2.6.15 "foo()" # search old tree for "foo()"4520$ git show v2.6.15:a.txt # look at old version of a.txt4521-----------------------------------------------45224523Search for regressions:45244525-----------------------------------------------4526$ git bisect start4527$ git bisect bad # current version is bad4528$ git bisect good v2.6.13-rc2 # last known good revision4529Bisecting: 675 revisions left to test after this4530 # test here, then:4531$ git bisect good # if this revision is good, or4532$ git bisect bad # if this revision is bad.4533 # repeat until done.4534-----------------------------------------------45354536[[making-changes]]4537Making changes4538--------------45394540Make sure Git knows who to blame:45414542------------------------------------------------4543$ cat >>~/.gitconfig <<\EOF4544[user]4545 name = Your Name Comes Here4546 email = you@yourdomain.example.com4547EOF4548------------------------------------------------45494550Select file contents to include in the next commit, then make the4551commit:45524553-----------------------------------------------4554$ git add a.txt # updated file4555$ git add b.txt # new file4556$ git rm c.txt # old file4557$ git commit4558-----------------------------------------------45594560Or, prepare and create the commit in one step:45614562-----------------------------------------------4563$ git commit d.txt # use latest content only of d.txt4564$ git commit -a # use latest content of all tracked files4565-----------------------------------------------45664567[[merging]]4568Merging4569-------45704571-----------------------------------------------4572$ git merge test # merge branch "test" into the current branch4573$ git pull git://example.com/project.git master4574 # fetch and merge in remote branch4575$ git pull . test # equivalent to git merge test4576-----------------------------------------------45774578[[sharing-your-changes]]4579Sharing your changes4580--------------------45814582Importing or exporting patches:45834584-----------------------------------------------4585$ git format-patch origin..HEAD # format a patch for each commit4586 # in HEAD but not in origin4587$ git am mbox # import patches from the mailbox "mbox"4588-----------------------------------------------45894590Fetch a branch in a different Git repository, then merge into the4591current branch:45924593-----------------------------------------------4594$ git pull git://example.com/project.git theirbranch4595-----------------------------------------------45964597Store the fetched branch into a local branch before merging into the4598current branch:45994600-----------------------------------------------4601$ git pull git://example.com/project.git theirbranch:mybranch4602-----------------------------------------------46034604After creating commits on a local branch, update the remote4605branch with your commits:46064607-----------------------------------------------4608$ git push ssh://example.com/project.git mybranch:theirbranch4609-----------------------------------------------46104611When remote and local branch are both named "test":46124613-----------------------------------------------4614$ git push ssh://example.com/project.git test4615-----------------------------------------------46164617Shortcut version for a frequently used remote repository:46184619-----------------------------------------------4620$ git remote add example ssh://example.com/project.git4621$ git push example test4622-----------------------------------------------46234624[[repository-maintenance]]4625Repository maintenance4626----------------------46274628Check for corruption:46294630-----------------------------------------------4631$ git fsck4632-----------------------------------------------46334634Recompress, remove unused cruft:46354636-----------------------------------------------4637$ git gc4638-----------------------------------------------463946404641[[todo]]4642Appendix B: Notes and todo list for this manual4643===============================================46444645[[todo-list]]4646Todo list4647---------46484649This is a work in progress.46504651The basic requirements:46524653- It must be readable in order, from beginning to end, by someone4654 intelligent with a basic grasp of the UNIX command line, but without4655 any special knowledge of Git. If necessary, any other prerequisites4656 should be specifically mentioned as they arise.4657- Whenever possible, section headings should clearly describe the task4658 they explain how to do, in language that requires no more knowledge4659 than necessary: for example, "importing patches into a project" rather4660 than "the `git am` command"46614662Think about how to create a clear chapter dependency graph that will4663allow people to get to important topics without necessarily reading4664everything in between.46654666Scan `Documentation/` for other stuff left out; in particular:46674668- howto's4669- some of `technical/`?4670- hooks4671- list of commands in linkgit:git[1]46724673Scan email archives for other stuff left out46744675Scan man pages to see if any assume more background than this manual4676provides.46774678Add more good examples. Entire sections of just cookbook examples4679might be a good idea; maybe make an "advanced examples" section a4680standard end-of-chapter section?46814682Include cross-references to the glossary, where appropriate.46834684Add a section on working with other version control systems, including4685CVS, Subversion, and just imports of series of release tarballs.46864687Write a chapter on using plumbing and writing scripts.46884689Alternates, clone -reference, etc.46904691More on recovery from repository corruption. See:4692 http://marc.info/?l=git&m=117263864820799&w=24693 http://marc.info/?l=git&m=117147855503798&w=2