HOWTO 29 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413414415416417418419420421422423424425426427428429430431432433434435436437438439440441442443444445446447448449450451452453454455456457458459460461462463464465466467468469470471472473474475476477478479480481482483484485486487488489490491492493494495496497498499500501502503504505506507508509510511512513514515516517518519520521522523524525526527528529530531532533534535536537538539540541542543544545546547548549550551552553554555556557558559560561562563564565566567568569570571572573574575576577578579580581582583584585586587588589590591592593594595596597598599600601602603604605606607608609610611612613614615616617618619620621622623624625626627628629630631632633634635636637638639640641642643644645646647648649650651652653654655656657658659660661
  1. HOWTO do Linux kernel development
  2. ---------------------------------
  3. This is the be-all, end-all document on this topic. It contains
  4. instructions on how to become a Linux kernel developer and how to learn
  5. to work with the Linux kernel development community. It tries to not
  6. contain anything related to the technical aspects of kernel programming,
  7. but will help point you in the right direction for that.
  8. If anything in this document becomes out of date, please send in patches
  9. to the maintainer of this file, who is listed at the bottom of the
  10. document.
  11. Introduction
  12. ------------
  13. So, you want to learn how to become a Linux kernel developer? Or you
  14. have been told by your manager, "Go write a Linux driver for this
  15. device." This document's goal is to teach you everything you need to
  16. know to achieve this by describing the process you need to go through,
  17. and hints on how to work with the community. It will also try to
  18. explain some of the reasons why the community works like it does.
  19. The kernel is written mostly in C, with some architecture-dependent
  20. parts written in assembly. A good understanding of C is required for
  21. kernel development. Assembly (any architecture) is not required unless
  22. you plan to do low-level development for that architecture. Though they
  23. are not a good substitute for a solid C education and/or years of
  24. experience, the following books are good for, if anything, reference:
  25. - "The C Programming Language" by Kernighan and Ritchie [Prentice Hall]
  26. - "Practical C Programming" by Steve Oualline [O'Reilly]
  27. - "C: A Reference Manual" by Harbison and Steele [Prentice Hall]
  28. The kernel is written using GNU C and the GNU toolchain. While it
  29. adheres to the ISO C89 standard, it uses a number of extensions that are
  30. not featured in the standard. The kernel is a freestanding C
  31. environment, with no reliance on the standard C library, so some
  32. portions of the C standard are not supported. Arbitrary long long
  33. divisions and floating point are not allowed. It can sometimes be
  34. difficult to understand the assumptions the kernel has on the toolchain
  35. and the extensions that it uses, and unfortunately there is no
  36. definitive reference for them. Please check the gcc info pages (`info
  37. gcc`) for some information on them.
  38. Please remember that you are trying to learn how to work with the
  39. existing development community. It is a diverse group of people, with
  40. high standards for coding, style and procedure. These standards have
  41. been created over time based on what they have found to work best for
  42. such a large and geographically dispersed team. Try to learn as much as
  43. possible about these standards ahead of time, as they are well
  44. documented; do not expect people to adapt to you or your company's way
  45. of doing things.
  46. Legal Issues
  47. ------------
  48. The Linux kernel source code is released under the GPL. Please see the
  49. file, COPYING, in the main directory of the source tree, for details on
  50. the license. If you have further questions about the license, please
  51. contact a lawyer, and do not ask on the Linux kernel mailing list. The
  52. people on the mailing lists are not lawyers, and you should not rely on
  53. their statements on legal matters.
  54. For common questions and answers about the GPL, please see:
  55. http://www.gnu.org/licenses/gpl-faq.html
  56. Documentation
  57. ------------
  58. The Linux kernel source tree has a large range of documents that are
  59. invaluable for learning how to interact with the kernel community. When
  60. new features are added to the kernel, it is recommended that new
  61. documentation files are also added which explain how to use the feature.
  62. When a kernel change causes the interface that the kernel exposes to
  63. userspace to change, it is recommended that you send the information or
  64. a patch to the manual pages explaining the change to the manual pages
  65. maintainer at mtk-manpages@gmx.net.
  66. Here is a list of files that are in the kernel source tree that are
  67. required reading:
  68. README
  69. This file gives a short background on the Linux kernel and describes
  70. what is necessary to do to configure and build the kernel. People
  71. who are new to the kernel should start here.
  72. Documentation/Changes
  73. This file gives a list of the minimum levels of various software
  74. packages that are necessary to build and run the kernel
  75. successfully.
  76. Documentation/CodingStyle
  77. This describes the Linux kernel coding style, and some of the
  78. rationale behind it. All new code is expected to follow the
  79. guidelines in this document. Most maintainers will only accept
  80. patches if these rules are followed, and many people will only
  81. review code if it is in the proper style.
  82. Documentation/SubmittingPatches
  83. Documentation/SubmittingDrivers
  84. These files describe in explicit detail how to successfully create
  85. and send a patch, including (but not limited to):
  86. - Email contents
  87. - Email format
  88. - Who to send it to
  89. Following these rules will not guarantee success (as all patches are
  90. subject to scrutiny for content and style), but not following them
  91. will almost always prevent it.
  92. Other excellent descriptions of how to create patches properly are:
  93. "The Perfect Patch"
  94. http://www.zip.com.au/~akpm/linux/patches/stuff/tpp.txt
  95. "Linux kernel patch submission format"
  96. http://linux.yyz.us/patch-format.html
  97. Documentation/stable_api_nonsense.txt
  98. This file describes the rationale behind the conscious decision to
  99. not have a stable API within the kernel, including things like:
  100. - Subsystem shim-layers (for compatibility?)
  101. - Driver portability between Operating Systems.
  102. - Mitigating rapid change within the kernel source tree (or
  103. preventing rapid change)
  104. This document is crucial for understanding the Linux development
  105. philosophy and is very important for people moving to Linux from
  106. development on other Operating Systems.
  107. Documentation/SecurityBugs
  108. If you feel you have found a security problem in the Linux kernel,
  109. please follow the steps in this document to help notify the kernel
  110. developers, and help solve the issue.
  111. Documentation/ManagementStyle
  112. This document describes how Linux kernel maintainers operate and the
  113. shared ethos behind their methodologies. This is important reading
  114. for anyone new to kernel development (or anyone simply curious about
  115. it), as it resolves a lot of common misconceptions and confusion
  116. about the unique behavior of kernel maintainers.
  117. Documentation/stable_kernel_rules.txt
  118. This file describes the rules on how the stable kernel releases
  119. happen, and what to do if you want to get a change into one of these
  120. releases.
  121. Documentation/kernel-docs.txt
  122. A list of external documentation that pertains to kernel
  123. development. Please consult this list if you do not find what you
  124. are looking for within the in-kernel documentation.
  125. Documentation/applying-patches.txt
  126. A good introduction describing exactly what a patch is and how to
  127. apply it to the different development branches of the kernel.
  128. The kernel also has a large number of documents that can be
  129. automatically generated from the source code itself. This includes a
  130. full description of the in-kernel API, and rules on how to handle
  131. locking properly. The documents will be created in the
  132. Documentation/DocBook/ directory and can be generated as PDF,
  133. Postscript, HTML, and man pages by running:
  134. make pdfdocs
  135. make psdocs
  136. make htmldocs
  137. make mandocs
  138. respectively from the main kernel source directory.
  139. Becoming A Kernel Developer
  140. ---------------------------
  141. If you do not know anything about Linux kernel development, you should
  142. look at the Linux KernelNewbies project:
  143. http://kernelnewbies.org
  144. It consists of a helpful mailing list where you can ask almost any type
  145. of basic kernel development question (make sure to search the archives
  146. first, before asking something that has already been answered in the
  147. past.) It also has an IRC channel that you can use to ask questions in
  148. real-time, and a lot of helpful documentation that is useful for
  149. learning about Linux kernel development.
  150. The website has basic information about code organization, subsystems,
  151. and current projects (both in-tree and out-of-tree). It also describes
  152. some basic logistical information, like how to compile a kernel and
  153. apply a patch.
  154. If you do not know where you want to start, but you want to look for
  155. some task to start doing to join into the kernel development community,
  156. go to the Linux Kernel Janitor's project:
  157. http://janitor.kernelnewbies.org/
  158. It is a great place to start. It describes a list of relatively simple
  159. problems that need to be cleaned up and fixed within the Linux kernel
  160. source tree. Working with the developers in charge of this project, you
  161. will learn the basics of getting your patch into the Linux kernel tree,
  162. and possibly be pointed in the direction of what to go work on next, if
  163. you do not already have an idea.
  164. If you already have a chunk of code that you want to put into the kernel
  165. tree, but need some help getting it in the proper form, the
  166. kernel-mentors project was created to help you out with this. It is a
  167. mailing list, and can be found at:
  168. http://selenic.com/mailman/listinfo/kernel-mentors
  169. Before making any actual modifications to the Linux kernel code, it is
  170. imperative to understand how the code in question works. For this
  171. purpose, nothing is better than reading through it directly (most tricky
  172. bits are commented well), perhaps even with the help of specialized
  173. tools. One such tool that is particularly recommended is the Linux
  174. Cross-Reference project, which is able to present source code in a
  175. self-referential, indexed webpage format. An excellent up-to-date
  176. repository of the kernel code may be found at:
  177. http://sosdg.org/~coywolf/lxr/
  178. The development process
  179. -----------------------
  180. Linux kernel development process currently consists of a few different
  181. main kernel "branches" and lots of different subsystem-specific kernel
  182. branches. These different branches are:
  183. - main 2.6.x kernel tree
  184. - 2.6.x.y -stable kernel tree
  185. - 2.6.x -git kernel patches
  186. - 2.6.x -mm kernel patches
  187. - subsystem specific kernel trees and patches
  188. 2.6.x kernel tree
  189. -----------------
  190. 2.6.x kernels are maintained by Linus Torvalds, and can be found on
  191. kernel.org in the pub/linux/kernel/v2.6/ directory. Its development
  192. process is as follows:
  193. - As soon as a new kernel is released a two weeks window is open,
  194. during this period of time maintainers can submit big diffs to
  195. Linus, usually the patches that have already been included in the
  196. -mm kernel for a few weeks. The preferred way to submit big changes
  197. is using git (the kernel's source management tool, more information
  198. can be found at http://git.or.cz/) but plain patches are also just
  199. fine.
  200. - After two weeks a -rc1 kernel is released it is now possible to push
  201. only patches that do not include new features that could affect the
  202. stability of the whole kernel. Please note that a whole new driver
  203. (or filesystem) might be accepted after -rc1 because there is no
  204. risk of causing regressions with such a change as long as the change
  205. is self-contained and does not affect areas outside of the code that
  206. is being added. git can be used to send patches to Linus after -rc1
  207. is released, but the patches need to also be sent to a public
  208. mailing list for review.
  209. - A new -rc is released whenever Linus deems the current git tree to
  210. be in a reasonably sane state adequate for testing. The goal is to
  211. release a new -rc kernel every week.
  212. - Process continues until the kernel is considered "ready", the
  213. process should last around 6 weeks.
  214. It is worth mentioning what Andrew Morton wrote on the linux-kernel
  215. mailing list about kernel releases:
  216. "Nobody knows when a kernel will be released, because it's
  217. released according to perceived bug status, not according to a
  218. preconceived timeline."
  219. 2.6.x.y -stable kernel tree
  220. ---------------------------
  221. Kernels with 4 digit versions are -stable kernels. They contain
  222. relatively small and critical fixes for security problems or significant
  223. regressions discovered in a given 2.6.x kernel.
  224. This is the recommended branch for users who want the most recent stable
  225. kernel and are not interested in helping test development/experimental
  226. versions.
  227. If no 2.6.x.y kernel is available, then the highest numbered 2.6.x
  228. kernel is the current stable kernel.
  229. 2.6.x.y are maintained by the "stable" team <stable@kernel.org>, and are
  230. released almost every other week.
  231. The file Documentation/stable_kernel_rules.txt in the kernel tree
  232. documents what kinds of changes are acceptable for the -stable tree, and
  233. how the release process works.
  234. 2.6.x -git patches
  235. ------------------
  236. These are daily snapshots of Linus' kernel tree which are managed in a
  237. git repository (hence the name.) These patches are usually released
  238. daily and represent the current state of Linus' tree. They are more
  239. experimental than -rc kernels since they are generated automatically
  240. without even a cursory glance to see if they are sane.
  241. 2.6.x -mm kernel patches
  242. ------------------------
  243. These are experimental kernel patches released by Andrew Morton. Andrew
  244. takes all of the different subsystem kernel trees and patches and mushes
  245. them together, along with a lot of patches that have been plucked from
  246. the linux-kernel mailing list. This tree serves as a proving ground for
  247. new features and patches. Once a patch has proved its worth in -mm for
  248. a while Andrew or the subsystem maintainer pushes it on to Linus for
  249. inclusion in mainline.
  250. It is heavily encouraged that all new patches get tested in the -mm tree
  251. before they are sent to Linus for inclusion in the main kernel tree.
  252. These kernels are not appropriate for use on systems that are supposed
  253. to be stable and they are more risky to run than any of the other
  254. branches.
  255. If you wish to help out with the kernel development process, please test
  256. and use these kernel releases and provide feedback to the linux-kernel
  257. mailing list if you have any problems, and if everything works properly.
  258. In addition to all the other experimental patches, these kernels usually
  259. also contain any changes in the mainline -git kernels available at the
  260. time of release.
  261. The -mm kernels are not released on a fixed schedule, but usually a few
  262. -mm kernels are released in between each -rc kernel (1 to 3 is common).
  263. Subsystem Specific kernel trees and patches
  264. -------------------------------------------
  265. A number of the different kernel subsystem developers expose their
  266. development trees so that others can see what is happening in the
  267. different areas of the kernel. These trees are pulled into the -mm
  268. kernel releases as described above.
  269. Here is a list of some of the different kernel trees available:
  270. git trees:
  271. - Kbuild development tree, Sam Ravnborg <sam@ravnborg.org>
  272. kernel.org:/pub/scm/linux/kernel/git/sam/kbuild.git
  273. - ACPI development tree, Len Brown <len.brown@intel.com>
  274. kernel.org:/pub/scm/linux/kernel/git/lenb/linux-acpi-2.6.git
  275. - Block development tree, Jens Axboe <axboe@suse.de>
  276. kernel.org:/pub/scm/linux/kernel/git/axboe/linux-2.6-block.git
  277. - DRM development tree, Dave Airlie <airlied@linux.ie>
  278. kernel.org:/pub/scm/linux/kernel/git/airlied/drm-2.6.git
  279. - ia64 development tree, Tony Luck <tony.luck@intel.com>
  280. kernel.org:/pub/scm/linux/kernel/git/aegl/linux-2.6.git
  281. - ieee1394 development tree, Jody McIntyre <scjody@modernduck.com>
  282. kernel.org:/pub/scm/linux/kernel/git/scjody/ieee1394.git
  283. - infiniband, Roland Dreier <rolandd@cisco.com>
  284. kernel.org:/pub/scm/linux/kernel/git/roland/infiniband.git
  285. - libata, Jeff Garzik <jgarzik@pobox.com>
  286. kernel.org:/pub/scm/linux/kernel/git/jgarzik/libata-dev.git
  287. - network drivers, Jeff Garzik <jgarzik@pobox.com>
  288. kernel.org:/pub/scm/linux/kernel/git/jgarzik/netdev-2.6.git
  289. - pcmcia, Dominik Brodowski <linux@dominikbrodowski.net>
  290. kernel.org:/pub/scm/linux/kernel/git/brodo/pcmcia-2.6.git
  291. - SCSI, James Bottomley <James.Bottomley@SteelEye.com>
  292. kernel.org:/pub/scm/linux/kernel/git/jejb/scsi-misc-2.6.git
  293. Other git kernel trees can be found listed at http://kernel.org/git
  294. quilt trees:
  295. - USB, PCI, Driver Core, and I2C, Greg Kroah-Hartman <gregkh@suse.de>
  296. kernel.org/pub/linux/kernel/people/gregkh/gregkh-2.6/
  297. - x86-64, partly i386, Andi Kleen <ak@suse.de>
  298. ftp.firstfloor.org:/pub/ak/x86_64/quilt/
  299. Bug Reporting
  300. -------------
  301. bugzilla.kernel.org is where the Linux kernel developers track kernel
  302. bugs. Users are encouraged to report all bugs that they find in this
  303. tool. For details on how to use the kernel bugzilla, please see:
  304. http://test.kernel.org/bugzilla/faq.html
  305. The file REPORTING-BUGS in the main kernel source directory has a good
  306. template for how to report a possible kernel bug, and details what kind
  307. of information is needed by the kernel developers to help track down the
  308. problem.
  309. Managing bug reports
  310. --------------------
  311. One of the best ways to put into practice your hacking skills is by fixing
  312. bugs reported by other people. Not only you will help to make the kernel
  313. more stable, you'll learn to fix real world problems and you will improve
  314. your skills, and other developers will be aware of your presence. Fixing
  315. bugs is one of the best ways to earn merit amongst the developers, because
  316. not many people like wasting time fixing other people's bugs.
  317. To work in the already reported bug reports, go to http://bugzilla.kernel.org.
  318. If you want to be advised of the future bug reports, you can subscribe to the
  319. bugme-new mailing list (only new bug reports are mailed here) or to the
  320. bugme-janitor mailing list (every change in the bugzilla is mailed here)
  321. http://lists.osdl.org/mailman/listinfo/bugme-new
  322. http://lists.osdl.org/mailman/listinfo/bugme-janitors
  323. Managing bug reports
  324. --------------------
  325. One of the best ways to put into practice your hacking skills is by fixing
  326. bugs reported by other people. Not only you will help to make the kernel
  327. more stable, you'll learn to fix real world problems and you will improve
  328. your skills, and other developers will be aware of your presence. Fixing
  329. bugs is one of the best ways to get merits among other developers, because
  330. not many people like wasting time fixing other people's bugs.
  331. To work in the already reported bug reports, go to http://bugzilla.kernel.org.
  332. If you want to be advised of the future bug reports, you can subscribe to the
  333. bugme-new mailing list (only new bug reports are mailed here) or to the
  334. bugme-janitor mailing list (every change in the bugzilla is mailed here)
  335. http://lists.osdl.org/mailman/listinfo/bugme-new
  336. http://lists.osdl.org/mailman/listinfo/bugme-janitors
  337. Mailing lists
  338. -------------
  339. As some of the above documents describe, the majority of the core kernel
  340. developers participate on the Linux Kernel Mailing list. Details on how
  341. to subscribe and unsubscribe from the list can be found at:
  342. http://vger.kernel.org/vger-lists.html#linux-kernel
  343. There are archives of the mailing list on the web in many different
  344. places. Use a search engine to find these archives. For example:
  345. http://dir.gmane.org/gmane.linux.kernel
  346. It is highly recommended that you search the archives about the topic
  347. you want to bring up, before you post it to the list. A lot of things
  348. already discussed in detail are only recorded at the mailing list
  349. archives.
  350. Most of the individual kernel subsystems also have their own separate
  351. mailing list where they do their development efforts. See the
  352. MAINTAINERS file for a list of what these lists are for the different
  353. groups.
  354. Many of the lists are hosted on kernel.org. Information on them can be
  355. found at:
  356. http://vger.kernel.org/vger-lists.html
  357. Please remember to follow good behavioral habits when using the lists.
  358. Though a bit cheesy, the following URL has some simple guidelines for
  359. interacting with the list (or any list):
  360. http://www.albion.com/netiquette/
  361. If multiple people respond to your mail, the CC: list of recipients may
  362. get pretty large. Don't remove anybody from the CC: list without a good
  363. reason, or don't reply only to the list address. Get used to receiving the
  364. mail twice, one from the sender and the one from the list, and don't try
  365. to tune that by adding fancy mail-headers, people will not like it.
  366. Remember to keep the context and the attribution of your replies intact,
  367. keep the "John Kernelhacker wrote ...:" lines at the top of your reply, and
  368. add your statements between the individual quoted sections instead of
  369. writing at the top of the mail.
  370. If you add patches to your mail, make sure they are plain readable text
  371. as stated in Documentation/SubmittingPatches. Kernel developers don't
  372. want to deal with attachments or compressed patches; they may want
  373. to comment on individual lines of your patch, which works only that way.
  374. Make sure you use a mail program that does not mangle spaces and tab
  375. characters. A good first test is to send the mail to yourself and try
  376. to apply your own patch by yourself. If that doesn't work, get your
  377. mail program fixed or change it until it works.
  378. Above all, please remember to show respect to other subscribers.
  379. Working with the community
  380. --------------------------
  381. The goal of the kernel community is to provide the best possible kernel
  382. there is. When you submit a patch for acceptance, it will be reviewed
  383. on its technical merits and those alone. So, what should you be
  384. expecting?
  385. - criticism
  386. - comments
  387. - requests for change
  388. - requests for justification
  389. - silence
  390. Remember, this is part of getting your patch into the kernel. You have
  391. to be able to take criticism and comments about your patches, evaluate
  392. them at a technical level and either rework your patches or provide
  393. clear and concise reasoning as to why those changes should not be made.
  394. If there are no responses to your posting, wait a few days and try
  395. again, sometimes things get lost in the huge volume.
  396. What should you not do?
  397. - expect your patch to be accepted without question
  398. - become defensive
  399. - ignore comments
  400. - resubmit the patch without making any of the requested changes
  401. In a community that is looking for the best technical solution possible,
  402. there will always be differing opinions on how beneficial a patch is.
  403. You have to be cooperative, and willing to adapt your idea to fit within
  404. the kernel. Or at least be willing to prove your idea is worth it.
  405. Remember, being wrong is acceptable as long as you are willing to work
  406. toward a solution that is right.
  407. It is normal that the answers to your first patch might simply be a list
  408. of a dozen things you should correct. This does _not_ imply that your
  409. patch will not be accepted, and it is _not_ meant against you
  410. personally. Simply correct all issues raised against your patch and
  411. resend it.
  412. Differences between the kernel community and corporate structures
  413. -----------------------------------------------------------------
  414. The kernel community works differently than most traditional corporate
  415. development environments. Here are a list of things that you can try to
  416. do to try to avoid problems:
  417. Good things to say regarding your proposed changes:
  418. - "This solves multiple problems."
  419. - "This deletes 2000 lines of code."
  420. - "Here is a patch that explains what I am trying to describe."
  421. - "I tested it on 5 different architectures..."
  422. - "Here is a series of small patches that..."
  423. - "This increases performance on typical machines..."
  424. Bad things you should avoid saying:
  425. - "We did it this way in AIX/ptx/Solaris, so therefore it must be
  426. good..."
  427. - "I've being doing this for 20 years, so..."
  428. - "This is required for my company to make money"
  429. - "This is for our Enterprise product line."
  430. - "Here is my 1000 page design document that describes my idea"
  431. - "I've been working on this for 6 months..."
  432. - "Here's a 5000 line patch that..."
  433. - "I rewrote all of the current mess, and here it is..."
  434. - "I have a deadline, and this patch needs to be applied now."
  435. Another way the kernel community is different than most traditional
  436. software engineering work environments is the faceless nature of
  437. interaction. One benefit of using email and irc as the primary forms of
  438. communication is the lack of discrimination based on gender or race.
  439. The Linux kernel work environment is accepting of women and minorities
  440. because all you are is an email address. The international aspect also
  441. helps to level the playing field because you can't guess gender based on
  442. a person's name. A man may be named Andrea and a woman may be named Pat.
  443. Most women who have worked in the Linux kernel and have expressed an
  444. opinion have had positive experiences.
  445. The language barrier can cause problems for some people who are not
  446. comfortable with English. A good grasp of the language can be needed in
  447. order to get ideas across properly on mailing lists, so it is
  448. recommended that you check your emails to make sure they make sense in
  449. English before sending them.
  450. Break up your changes
  451. ---------------------
  452. The Linux kernel community does not gladly accept large chunks of code
  453. dropped on it all at once. The changes need to be properly introduced,
  454. discussed, and broken up into tiny, individual portions. This is almost
  455. the exact opposite of what companies are used to doing. Your proposal
  456. should also be introduced very early in the development process, so that
  457. you can receive feedback on what you are doing. It also lets the
  458. community feel that you are working with them, and not simply using them
  459. as a dumping ground for your feature. However, don't send 50 emails at
  460. one time to a mailing list, your patch series should be smaller than
  461. that almost all of the time.
  462. The reasons for breaking things up are the following:
  463. 1) Small patches increase the likelihood that your patches will be
  464. applied, since they don't take much time or effort to verify for
  465. correctness. A 5 line patch can be applied by a maintainer with
  466. barely a second glance. However, a 500 line patch may take hours to
  467. review for correctness (the time it takes is exponentially
  468. proportional to the size of the patch, or something).
  469. Small patches also make it very easy to debug when something goes
  470. wrong. It's much easier to back out patches one by one than it is
  471. to dissect a very large patch after it's been applied (and broken
  472. something).
  473. 2) It's important not only to send small patches, but also to rewrite
  474. and simplify (or simply re-order) patches before submitting them.
  475. Here is an analogy from kernel developer Al Viro:
  476. "Think of a teacher grading homework from a math student. The
  477. teacher does not want to see the student's trials and errors
  478. before they came up with the solution. They want to see the
  479. cleanest, most elegant answer. A good student knows this, and
  480. would never submit her intermediate work before the final
  481. solution."
  482. The same is true of kernel development. The maintainers and
  483. reviewers do not want to see the thought process behind the
  484. solution to the problem one is solving. They want to see a
  485. simple and elegant solution."
  486. It may be challenging to keep the balance between presenting an elegant
  487. solution and working together with the community and discussing your
  488. unfinished work. Therefore it is good to get early in the process to
  489. get feedback to improve your work, but also keep your changes in small
  490. chunks that they may get already accepted, even when your whole task is
  491. not ready for inclusion now.
  492. Also realize that it is not acceptable to send patches for inclusion
  493. that are unfinished and will be "fixed up later."
  494. Justify your change
  495. -------------------
  496. Along with breaking up your patches, it is very important for you to let
  497. the Linux community know why they should add this change. New features
  498. must be justified as being needed and useful.
  499. Document your change
  500. --------------------
  501. When sending in your patches, pay special attention to what you say in
  502. the text in your email. This information will become the ChangeLog
  503. information for the patch, and will be preserved for everyone to see for
  504. all time. It should describe the patch completely, containing:
  505. - why the change is necessary
  506. - the overall design approach in the patch
  507. - implementation details
  508. - testing results
  509. For more details on what this should all look like, please see the
  510. ChangeLog section of the document:
  511. "The Perfect Patch"
  512. http://www.zip.com.au/~akpm/linux/patches/stuff/tpp.txt
  513. All of these things are sometimes very hard to do. It can take years to
  514. perfect these practices (if at all). It's a continuous process of
  515. improvement that requires a lot of patience and determination. But
  516. don't give up, it's possible. Many have done it before, and each had to
  517. start exactly where you are now.
  518. ----------
  519. Thanks to Paolo Ciarrocchi who allowed the "Development Process"
  520. (http://linux.tar.bz/articles/2.6-development_process) section
  521. to be based on text he had written, and to Randy Dunlap and Gerrit
  522. Huizenga for some of the list of things you should and should not say.
  523. Also thanks to Pat Mochel, Hanna Linder, Randy Dunlap, Kay Sievers,
  524. Vojtech Pavlik, Jan Kara, Josh Boyer, Kees Cook, Andrew Morton, Andi
  525. Kleen, Vadim Lobanov, Jesper Juhl, Adrian Bunk, Keri Harris, Frans Pop,
  526. David A. Wheeler, Junio Hamano, Michael Kerrisk, and Alex Shepard for
  527. their review, comments, and contributions. Without their help, this
  528. document would not have been possible.
  529. Maintainer: Greg Kroah-Hartman <greg@kroah.com>