submitting-patches.rst 34 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413414415416417418419420421422423424425426427428429430431432433434435436437438439440441442443444445446447448449450451452453454455456457458459460461462463464465466467468469470471472473474475476477478479480481482483484485486487488489490491492493494495496497498499500501502503504505506507508509510511512513514515516517518519520521522523524525526527528529530531532533534535536537538539540541542543544545546547548549550551552553554555556557558559560561562563564565566567568569570571572573574575576577578579580581582583584585586587588589590591592593594595596597598599600601602603604605606607608609610611612613614615616617618619620621622623624625626627628629630631632633634635636637638639640641642643644645646647648649650651652653654655656657658659660661662663664665666667668669670671672673674675676677678679680681682683684685686687688689690691692693694695696697698699700701702703704705706707708709710711712713714715716717718719720721722723724725726727728729730731732733734735736737738739740741742743744745746747748749750751752753754755756757758759760761762763764765766767768769770
  1. .. _submittingpatches:
  2. Submitting patches: the essential guide to getting your code into the kernel
  3. ============================================================================
  4. For a person or company who wishes to submit a change to the Linux
  5. kernel, the process can sometimes be daunting if you're not familiar
  6. with "the system." This text is a collection of suggestions which
  7. can greatly increase the chances of your change being accepted.
  8. This document contains a large number of suggestions in a relatively terse
  9. format. For detailed information on how the kernel development process
  10. works, see :doc:`development-process`. Also, read :doc:`submit-checklist`
  11. for a list of items to check before submitting code. If you are submitting
  12. a driver, also read :doc:`submitting-drivers`; for device tree binding patches,
  13. read :doc:`submitting-patches`.
  14. This documentation assumes that you're using ``git`` to prepare your patches.
  15. If you're unfamiliar with ``git``, you would be well-advised to learn how to
  16. use it, it will make your life as a kernel developer and in general much
  17. easier.
  18. Obtain a current source tree
  19. ----------------------------
  20. If you do not have a repository with the current kernel source handy, use
  21. ``git`` to obtain one. You'll want to start with the mainline repository,
  22. which can be grabbed with::
  23. git clone git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
  24. Note, however, that you may not want to develop against the mainline tree
  25. directly. Most subsystem maintainers run their own trees and want to see
  26. patches prepared against those trees. See the **T:** entry for the subsystem
  27. in the MAINTAINERS file to find that tree, or simply ask the maintainer if
  28. the tree is not listed there.
  29. .. _describe_changes:
  30. Describe your changes
  31. ---------------------
  32. Describe your problem. Whether your patch is a one-line bug fix or
  33. 5000 lines of a new feature, there must be an underlying problem that
  34. motivated you to do this work. Convince the reviewer that there is a
  35. problem worth fixing and that it makes sense for them to read past the
  36. first paragraph.
  37. Describe user-visible impact. Straight up crashes and lockups are
  38. pretty convincing, but not all bugs are that blatant. Even if the
  39. problem was spotted during code review, describe the impact you think
  40. it can have on users. Keep in mind that the majority of Linux
  41. installations run kernels from secondary stable trees or
  42. vendor/product-specific trees that cherry-pick only specific patches
  43. from upstream, so include anything that could help route your change
  44. downstream: provoking circumstances, excerpts from dmesg, crash
  45. descriptions, performance regressions, latency spikes, lockups, etc.
  46. Quantify optimizations and trade-offs. If you claim improvements in
  47. performance, memory consumption, stack footprint, or binary size,
  48. include numbers that back them up. But also describe non-obvious
  49. costs. Optimizations usually aren't free but trade-offs between CPU,
  50. memory, and readability; or, when it comes to heuristics, between
  51. different workloads. Describe the expected downsides of your
  52. optimization so that the reviewer can weigh costs against benefits.
  53. Once the problem is established, describe what you are actually doing
  54. about it in technical detail. It's important to describe the change
  55. in plain English for the reviewer to verify that the code is behaving
  56. as you intend it to.
  57. The maintainer will thank you if you write your patch description in a
  58. form which can be easily pulled into Linux's source code management
  59. system, ``git``, as a "commit log". See :ref:`explicit_in_reply_to`.
  60. Solve only one problem per patch. If your description starts to get
  61. long, that's a sign that you probably need to split up your patch.
  62. See :ref:`split_changes`.
  63. When you submit or resubmit a patch or patch series, include the
  64. complete patch description and justification for it. Don't just
  65. say that this is version N of the patch (series). Don't expect the
  66. subsystem maintainer to refer back to earlier patch versions or referenced
  67. URLs to find the patch description and put that into the patch.
  68. I.e., the patch (series) and its description should be self-contained.
  69. This benefits both the maintainers and reviewers. Some reviewers
  70. probably didn't even receive earlier versions of the patch.
  71. Describe your changes in imperative mood, e.g. "make xyzzy do frotz"
  72. instead of "[This patch] makes xyzzy do frotz" or "[I] changed xyzzy
  73. to do frotz", as if you are giving orders to the codebase to change
  74. its behaviour.
  75. If the patch fixes a logged bug entry, refer to that bug entry by
  76. number and URL. If the patch follows from a mailing list discussion,
  77. give a URL to the mailing list archive; use the https://lkml.kernel.org/
  78. redirector with a ``Message-Id``, to ensure that the links cannot become
  79. stale.
  80. However, try to make your explanation understandable without external
  81. resources. In addition to giving a URL to a mailing list archive or
  82. bug, summarize the relevant points of the discussion that led to the
  83. patch as submitted.
  84. If you want to refer to a specific commit, don't just refer to the
  85. SHA-1 ID of the commit. Please also include the oneline summary of
  86. the commit, to make it easier for reviewers to know what it is about.
  87. Example::
  88. Commit e21d2170f36602ae2708 ("video: remove unnecessary
  89. platform_set_drvdata()") removed the unnecessary
  90. platform_set_drvdata(), but left the variable "dev" unused,
  91. delete it.
  92. You should also be sure to use at least the first twelve characters of the
  93. SHA-1 ID. The kernel repository holds a *lot* of objects, making
  94. collisions with shorter IDs a real possibility. Bear in mind that, even if
  95. there is no collision with your six-character ID now, that condition may
  96. change five years from now.
  97. If your patch fixes a bug in a specific commit, e.g. you found an issue using
  98. ``git bisect``, please use the 'Fixes:' tag with the first 12 characters of
  99. the SHA-1 ID, and the one line summary. Do not split the tag across multiple
  100. lines, tags are exempt from the "wrap at 75 columns" rule in order to simplify
  101. parsing scripts. For example::
  102. Fixes: 54a4f0239f2e ("KVM: MMU: make kvm_mmu_zap_page() return the number of pages it actually freed")
  103. The following ``git config`` settings can be used to add a pretty format for
  104. outputting the above style in the ``git log`` or ``git show`` commands::
  105. [core]
  106. abbrev = 12
  107. [pretty]
  108. fixes = Fixes: %h (\"%s\")
  109. An example call::
  110. $ git log -1 --pretty=fixes 54a4f0239f2e
  111. Fixes: 54a4f0239f2e ("KVM: MMU: make kvm_mmu_zap_page() return the number of pages it actually freed")
  112. .. _split_changes:
  113. Separate your changes
  114. ---------------------
  115. Separate each **logical change** into a separate patch.
  116. For example, if your changes include both bug fixes and performance
  117. enhancements for a single driver, separate those changes into two
  118. or more patches. If your changes include an API update, and a new
  119. driver which uses that new API, separate those into two patches.
  120. On the other hand, if you make a single change to numerous files,
  121. group those changes into a single patch. Thus a single logical change
  122. is contained within a single patch.
  123. The point to remember is that each patch should make an easily understood
  124. change that can be verified by reviewers. Each patch should be justifiable
  125. on its own merits.
  126. If one patch depends on another patch in order for a change to be
  127. complete, that is OK. Simply note **"this patch depends on patch X"**
  128. in your patch description.
  129. When dividing your change into a series of patches, take special care to
  130. ensure that the kernel builds and runs properly after each patch in the
  131. series. Developers using ``git bisect`` to track down a problem can end up
  132. splitting your patch series at any point; they will not thank you if you
  133. introduce bugs in the middle.
  134. If you cannot condense your patch set into a smaller set of patches,
  135. then only post say 15 or so at a time and wait for review and integration.
  136. Style-check your changes
  137. ------------------------
  138. Check your patch for basic style violations, details of which can be
  139. found in
  140. :ref:`Documentation/process/coding-style.rst <codingstyle>`.
  141. Failure to do so simply wastes
  142. the reviewers time and will get your patch rejected, probably
  143. without even being read.
  144. One significant exception is when moving code from one file to
  145. another -- in this case you should not modify the moved code at all in
  146. the same patch which moves it. This clearly delineates the act of
  147. moving the code and your changes. This greatly aids review of the
  148. actual differences and allows tools to better track the history of
  149. the code itself.
  150. Check your patches with the patch style checker prior to submission
  151. (scripts/checkpatch.pl). Note, though, that the style checker should be
  152. viewed as a guide, not as a replacement for human judgment. If your code
  153. looks better with a violation then its probably best left alone.
  154. The checker reports at three levels:
  155. - ERROR: things that are very likely to be wrong
  156. - WARNING: things requiring careful review
  157. - CHECK: things requiring thought
  158. You should be able to justify all violations that remain in your
  159. patch.
  160. Select the recipients for your patch
  161. ------------------------------------
  162. You should always copy the appropriate subsystem maintainer(s) on any patch
  163. to code that they maintain; look through the MAINTAINERS file and the
  164. source code revision history to see who those maintainers are. The
  165. script scripts/get_maintainer.pl can be very useful at this step. If you
  166. cannot find a maintainer for the subsystem you are working on, Andrew
  167. Morton (akpm@linux-foundation.org) serves as a maintainer of last resort.
  168. You should also normally choose at least one mailing list to receive a copy
  169. of your patch set. linux-kernel@vger.kernel.org functions as a list of
  170. last resort, but the volume on that list has caused a number of developers
  171. to tune it out. Look in the MAINTAINERS file for a subsystem-specific
  172. list; your patch will probably get more attention there. Please do not
  173. spam unrelated lists, though.
  174. Many kernel-related lists are hosted on vger.kernel.org; you can find a
  175. list of them at http://vger.kernel.org/vger-lists.html. There are
  176. kernel-related lists hosted elsewhere as well, though.
  177. Do not send more than 15 patches at once to the vger mailing lists!!!
  178. Linus Torvalds is the final arbiter of all changes accepted into the
  179. Linux kernel. His e-mail address is <torvalds@linux-foundation.org>.
  180. He gets a lot of e-mail, and, at this point, very few patches go through
  181. Linus directly, so typically you should do your best to -avoid-
  182. sending him e-mail.
  183. If you have a patch that fixes an exploitable security bug, send that patch
  184. to security@kernel.org. For severe bugs, a short embargo may be considered
  185. to allow distributors to get the patch out to users; in such cases,
  186. obviously, the patch should not be sent to any public lists. See also
  187. :doc:`/admin-guide/security-bugs`.
  188. Patches that fix a severe bug in a released kernel should be directed
  189. toward the stable maintainers by putting a line like this::
  190. Cc: stable@vger.kernel.org
  191. into the sign-off area of your patch (note, NOT an email recipient). You
  192. should also read
  193. :ref:`Documentation/process/stable-kernel-rules.rst <stable_kernel_rules>`
  194. in addition to this file.
  195. If changes affect userland-kernel interfaces, please send the MAN-PAGES
  196. maintainer (as listed in the MAINTAINERS file) a man-pages patch, or at
  197. least a notification of the change, so that some information makes its way
  198. into the manual pages. User-space API changes should also be copied to
  199. linux-api@vger.kernel.org.
  200. For small patches you may want to CC the Trivial Patch Monkey
  201. trivial@kernel.org which collects "trivial" patches. Have a look
  202. into the MAINTAINERS file for its current manager.
  203. Trivial patches must qualify for one of the following rules:
  204. - Spelling fixes in documentation
  205. - Spelling fixes for errors which could break :manpage:`grep(1)`
  206. - Warning fixes (cluttering with useless warnings is bad)
  207. - Compilation fixes (only if they are actually correct)
  208. - Runtime fixes (only if they actually fix things)
  209. - Removing use of deprecated functions/macros
  210. - Contact detail and documentation fixes
  211. - Non-portable code replaced by portable code (even in arch-specific,
  212. since people copy, as long as it's trivial)
  213. - Any fix by the author/maintainer of the file (ie. patch monkey
  214. in re-transmission mode)
  215. No MIME, no links, no compression, no attachments. Just plain text
  216. -------------------------------------------------------------------
  217. Linus and other kernel developers need to be able to read and comment
  218. on the changes you are submitting. It is important for a kernel
  219. developer to be able to "quote" your changes, using standard e-mail
  220. tools, so that they may comment on specific portions of your code.
  221. For this reason, all patches should be submitted by e-mail "inline". The
  222. easiest way to do this is with ``git send-email``, which is strongly
  223. recommended. An interactive tutorial for ``git send-email`` is available at
  224. https://git-send-email.io.
  225. If you choose not to use ``git send-email``:
  226. .. warning::
  227. Be wary of your editor's word-wrap corrupting your patch,
  228. if you choose to cut-n-paste your patch.
  229. Do not attach the patch as a MIME attachment, compressed or not.
  230. Many popular e-mail applications will not always transmit a MIME
  231. attachment as plain text, making it impossible to comment on your
  232. code. A MIME attachment also takes Linus a bit more time to process,
  233. decreasing the likelihood of your MIME-attached change being accepted.
  234. Exception: If your mailer is mangling patches then someone may ask
  235. you to re-send them using MIME.
  236. See :doc:`/process/email-clients` for hints about configuring your e-mail
  237. client so that it sends your patches untouched.
  238. Respond to review comments
  239. --------------------------
  240. Your patch will almost certainly get comments from reviewers on ways in
  241. which the patch can be improved, in the form of a reply to your email. You must
  242. respond to those comments; ignoring reviewers is a good way to get ignored in
  243. return. You can simply reply to their emails to answer their comments. Review
  244. comments or questions that do not lead to a code change should almost certainly
  245. bring about a comment or changelog entry so that the next reviewer better
  246. understands what is going on.
  247. Be sure to tell the reviewers what changes you are making and to thank them
  248. for their time. Code review is a tiring and time-consuming process, and
  249. reviewers sometimes get grumpy. Even in that case, though, respond
  250. politely and address the problems they have pointed out.
  251. See :doc:`email-clients` for recommendations on email
  252. clients and mailing list etiquette.
  253. Don't get discouraged - or impatient
  254. ------------------------------------
  255. After you have submitted your change, be patient and wait. Reviewers are
  256. busy people and may not get to your patch right away.
  257. Once upon a time, patches used to disappear into the void without comment,
  258. but the development process works more smoothly than that now. You should
  259. receive comments within a week or so; if that does not happen, make sure
  260. that you have sent your patches to the right place. Wait for a minimum of
  261. one week before resubmitting or pinging reviewers - possibly longer during
  262. busy times like merge windows.
  263. Include PATCH in the subject
  264. -----------------------------
  265. Due to high e-mail traffic to Linus, and to linux-kernel, it is common
  266. convention to prefix your subject line with [PATCH]. This lets Linus
  267. and other kernel developers more easily distinguish patches from other
  268. e-mail discussions.
  269. ``git send-email`` will do this for you automatically.
  270. Sign your work - the Developer's Certificate of Origin
  271. ------------------------------------------------------
  272. To improve tracking of who did what, especially with patches that can
  273. percolate to their final resting place in the kernel through several
  274. layers of maintainers, we've introduced a "sign-off" procedure on
  275. patches that are being emailed around.
  276. The sign-off is a simple line at the end of the explanation for the
  277. patch, which certifies that you wrote it or otherwise have the right to
  278. pass it on as an open-source patch. The rules are pretty simple: if you
  279. can certify the below:
  280. Developer's Certificate of Origin 1.1
  281. ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  282. By making a contribution to this project, I certify that:
  283. (a) The contribution was created in whole or in part by me and I
  284. have the right to submit it under the open source license
  285. indicated in the file; or
  286. (b) The contribution is based upon previous work that, to the best
  287. of my knowledge, is covered under an appropriate open source
  288. license and I have the right under that license to submit that
  289. work with modifications, whether created in whole or in part
  290. by me, under the same open source license (unless I am
  291. permitted to submit under a different license), as indicated
  292. in the file; or
  293. (c) The contribution was provided directly to me by some other
  294. person who certified (a), (b) or (c) and I have not modified
  295. it.
  296. (d) I understand and agree that this project and the contribution
  297. are public and that a record of the contribution (including all
  298. personal information I submit with it, including my sign-off) is
  299. maintained indefinitely and may be redistributed consistent with
  300. this project or the open source license(s) involved.
  301. then you just add a line saying::
  302. Signed-off-by: Random J Developer <random@developer.example.org>
  303. using your real name (sorry, no pseudonyms or anonymous contributions.)
  304. This will be done for you automatically if you use ``git commit -s``.
  305. Some people also put extra tags at the end. They'll just be ignored for
  306. now, but you can do this to mark internal company procedures or just
  307. point out some special detail about the sign-off.
  308. When to use Acked-by:, Cc:, and Co-developed-by:
  309. ------------------------------------------------
  310. The Signed-off-by: tag indicates that the signer was involved in the
  311. development of the patch, or that he/she was in the patch's delivery path.
  312. If a person was not directly involved in the preparation or handling of a
  313. patch but wishes to signify and record their approval of it then they can
  314. ask to have an Acked-by: line added to the patch's changelog.
  315. Acked-by: is often used by the maintainer of the affected code when that
  316. maintainer neither contributed to nor forwarded the patch.
  317. Acked-by: is not as formal as Signed-off-by:. It is a record that the acker
  318. has at least reviewed the patch and has indicated acceptance. Hence patch
  319. mergers will sometimes manually convert an acker's "yep, looks good to me"
  320. into an Acked-by: (but note that it is usually better to ask for an
  321. explicit ack).
  322. Acked-by: does not necessarily indicate acknowledgement of the entire patch.
  323. For example, if a patch affects multiple subsystems and has an Acked-by: from
  324. one subsystem maintainer then this usually indicates acknowledgement of just
  325. the part which affects that maintainer's code. Judgement should be used here.
  326. When in doubt people should refer to the original discussion in the mailing
  327. list archives.
  328. If a person has had the opportunity to comment on a patch, but has not
  329. provided such comments, you may optionally add a ``Cc:`` tag to the patch.
  330. This is the only tag which might be added without an explicit action by the
  331. person it names - but it should indicate that this person was copied on the
  332. patch. This tag documents that potentially interested parties
  333. have been included in the discussion.
  334. Co-developed-by: states that the patch was co-created by multiple developers;
  335. it is a used to give attribution to co-authors (in addition to the author
  336. attributed by the From: tag) when several people work on a single patch. Since
  337. Co-developed-by: denotes authorship, every Co-developed-by: must be immediately
  338. followed by a Signed-off-by: of the associated co-author. Standard sign-off
  339. procedure applies, i.e. the ordering of Signed-off-by: tags should reflect the
  340. chronological history of the patch insofar as possible, regardless of whether
  341. the author is attributed via From: or Co-developed-by:. Notably, the last
  342. Signed-off-by: must always be that of the developer submitting the patch.
  343. Note, the From: tag is optional when the From: author is also the person (and
  344. email) listed in the From: line of the email header.
  345. Example of a patch submitted by the From: author::
  346. <changelog>
  347. Co-developed-by: First Co-Author <first@coauthor.example.org>
  348. Signed-off-by: First Co-Author <first@coauthor.example.org>
  349. Co-developed-by: Second Co-Author <second@coauthor.example.org>
  350. Signed-off-by: Second Co-Author <second@coauthor.example.org>
  351. Signed-off-by: From Author <from@author.example.org>
  352. Example of a patch submitted by a Co-developed-by: author::
  353. From: From Author <from@author.example.org>
  354. <changelog>
  355. Co-developed-by: Random Co-Author <random@coauthor.example.org>
  356. Signed-off-by: Random Co-Author <random@coauthor.example.org>
  357. Signed-off-by: From Author <from@author.example.org>
  358. Co-developed-by: Submitting Co-Author <sub@coauthor.example.org>
  359. Signed-off-by: Submitting Co-Author <sub@coauthor.example.org>
  360. Using Reported-by:, Tested-by:, Reviewed-by:, Suggested-by: and Fixes:
  361. ----------------------------------------------------------------------
  362. The Reported-by tag gives credit to people who find bugs and report them and it
  363. hopefully inspires them to help us again in the future. Please note that if
  364. the bug was reported in private, then ask for permission first before using the
  365. Reported-by tag.
  366. A Tested-by: tag indicates that the patch has been successfully tested (in
  367. some environment) by the person named. This tag informs maintainers that
  368. some testing has been performed, provides a means to locate testers for
  369. future patches, and ensures credit for the testers.
  370. Reviewed-by:, instead, indicates that the patch has been reviewed and found
  371. acceptable according to the Reviewer's Statement:
  372. Reviewer's statement of oversight
  373. ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  374. By offering my Reviewed-by: tag, I state that:
  375. (a) I have carried out a technical review of this patch to
  376. evaluate its appropriateness and readiness for inclusion into
  377. the mainline kernel.
  378. (b) Any problems, concerns, or questions relating to the patch
  379. have been communicated back to the submitter. I am satisfied
  380. with the submitter's response to my comments.
  381. (c) While there may be things that could be improved with this
  382. submission, I believe that it is, at this time, (1) a
  383. worthwhile modification to the kernel, and (2) free of known
  384. issues which would argue against its inclusion.
  385. (d) While I have reviewed the patch and believe it to be sound, I
  386. do not (unless explicitly stated elsewhere) make any
  387. warranties or guarantees that it will achieve its stated
  388. purpose or function properly in any given situation.
  389. A Reviewed-by tag is a statement of opinion that the patch is an
  390. appropriate modification of the kernel without any remaining serious
  391. technical issues. Any interested reviewer (who has done the work) can
  392. offer a Reviewed-by tag for a patch. This tag serves to give credit to
  393. reviewers and to inform maintainers of the degree of review which has been
  394. done on the patch. Reviewed-by: tags, when supplied by reviewers known to
  395. understand the subject area and to perform thorough reviews, will normally
  396. increase the likelihood of your patch getting into the kernel.
  397. Both Tested-by and Reviewed-by tags, once received on mailing list from tester
  398. or reviewer, should be added by author to the applicable patches when sending
  399. next versions. However if the patch has changed substantially in following
  400. version, these tags might not be applicable anymore and thus should be removed.
  401. Usually removal of someone's Tested-by or Reviewed-by tags should be mentioned
  402. in the patch changelog (after the '---' separator).
  403. A Suggested-by: tag indicates that the patch idea is suggested by the person
  404. named and ensures credit to the person for the idea. Please note that this
  405. tag should not be added without the reporter's permission, especially if the
  406. idea was not posted in a public forum. That said, if we diligently credit our
  407. idea reporters, they will, hopefully, be inspired to help us again in the
  408. future.
  409. A Fixes: tag indicates that the patch fixes an issue in a previous commit. It
  410. is used to make it easy to determine where a bug originated, which can help
  411. review a bug fix. This tag also assists the stable kernel team in determining
  412. which stable kernel versions should receive your fix. This is the preferred
  413. method for indicating a bug fixed by the patch. See :ref:`describe_changes`
  414. for more details.
  415. .. _the_canonical_patch_format:
  416. The canonical patch format
  417. --------------------------
  418. This section describes how the patch itself should be formatted. Note
  419. that, if you have your patches stored in a ``git`` repository, proper patch
  420. formatting can be had with ``git format-patch``. The tools cannot create
  421. the necessary text, though, so read the instructions below anyway.
  422. The canonical patch subject line is::
  423. Subject: [PATCH 001/123] subsystem: summary phrase
  424. The canonical patch message body contains the following:
  425. - A ``from`` line specifying the patch author, followed by an empty
  426. line (only needed if the person sending the patch is not the author).
  427. - The body of the explanation, line wrapped at 75 columns, which will
  428. be copied to the permanent changelog to describe this patch.
  429. - An empty line.
  430. - The ``Signed-off-by:`` lines, described above, which will
  431. also go in the changelog.
  432. - A marker line containing simply ``---``.
  433. - Any additional comments not suitable for the changelog.
  434. - The actual patch (``diff`` output).
  435. The Subject line format makes it very easy to sort the emails
  436. alphabetically by subject line - pretty much any email reader will
  437. support that - since because the sequence number is zero-padded,
  438. the numerical and alphabetic sort is the same.
  439. The ``subsystem`` in the email's Subject should identify which
  440. area or subsystem of the kernel is being patched.
  441. The ``summary phrase`` in the email's Subject should concisely
  442. describe the patch which that email contains. The ``summary
  443. phrase`` should not be a filename. Do not use the same ``summary
  444. phrase`` for every patch in a whole patch series (where a ``patch
  445. series`` is an ordered sequence of multiple, related patches).
  446. Bear in mind that the ``summary phrase`` of your email becomes a
  447. globally-unique identifier for that patch. It propagates all the way
  448. into the ``git`` changelog. The ``summary phrase`` may later be used in
  449. developer discussions which refer to the patch. People will want to
  450. google for the ``summary phrase`` to read discussion regarding that
  451. patch. It will also be the only thing that people may quickly see
  452. when, two or three months later, they are going through perhaps
  453. thousands of patches using tools such as ``gitk`` or ``git log
  454. --oneline``.
  455. For these reasons, the ``summary`` must be no more than 70-75
  456. characters, and it must describe both what the patch changes, as well
  457. as why the patch might be necessary. It is challenging to be both
  458. succinct and descriptive, but that is what a well-written summary
  459. should do.
  460. The ``summary phrase`` may be prefixed by tags enclosed in square
  461. brackets: "Subject: [PATCH <tag>...] <summary phrase>". The tags are
  462. not considered part of the summary phrase, but describe how the patch
  463. should be treated. Common tags might include a version descriptor if
  464. the multiple versions of the patch have been sent out in response to
  465. comments (i.e., "v1, v2, v3"), or "RFC" to indicate a request for
  466. comments. If there are four patches in a patch series the individual
  467. patches may be numbered like this: 1/4, 2/4, 3/4, 4/4. This assures
  468. that developers understand the order in which the patches should be
  469. applied and that they have reviewed or applied all of the patches in
  470. the patch series.
  471. A couple of example Subjects::
  472. Subject: [PATCH 2/5] ext2: improve scalability of bitmap searching
  473. Subject: [PATCH v2 01/27] x86: fix eflags tracking
  474. The ``from`` line must be the very first line in the message body,
  475. and has the form:
  476. From: Patch Author <author@example.com>
  477. The ``from`` line specifies who will be credited as the author of the
  478. patch in the permanent changelog. If the ``from`` line is missing,
  479. then the ``From:`` line from the email header will be used to determine
  480. the patch author in the changelog.
  481. The explanation body will be committed to the permanent source
  482. changelog, so should make sense to a competent reader who has long
  483. since forgotten the immediate details of the discussion that might
  484. have led to this patch. Including symptoms of the failure which the
  485. patch addresses (kernel log messages, oops messages, etc.) is
  486. especially useful for people who might be searching the commit logs
  487. looking for the applicable patch. If a patch fixes a compile failure,
  488. it may not be necessary to include _all_ of the compile failures; just
  489. enough that it is likely that someone searching for the patch can find
  490. it. As in the ``summary phrase``, it is important to be both succinct as
  491. well as descriptive.
  492. The ``---`` marker line serves the essential purpose of marking for patch
  493. handling tools where the changelog message ends.
  494. One good use for the additional comments after the ``---`` marker is for
  495. a ``diffstat``, to show what files have changed, and the number of
  496. inserted and deleted lines per file. A ``diffstat`` is especially useful
  497. on bigger patches. Other comments relevant only to the moment or the
  498. maintainer, not suitable for the permanent changelog, should also go
  499. here. A good example of such comments might be ``patch changelogs``
  500. which describe what has changed between the v1 and v2 version of the
  501. patch.
  502. If you are going to include a ``diffstat`` after the ``---`` marker, please
  503. use ``diffstat`` options ``-p 1 -w 70`` so that filenames are listed from
  504. the top of the kernel source tree and don't use too much horizontal
  505. space (easily fit in 80 columns, maybe with some indentation). (``git``
  506. generates appropriate diffstats by default.)
  507. See more details on the proper patch format in the following
  508. references.
  509. .. _explicit_in_reply_to:
  510. Explicit In-Reply-To headers
  511. ----------------------------
  512. It can be helpful to manually add In-Reply-To: headers to a patch
  513. (e.g., when using ``git send-email``) to associate the patch with
  514. previous relevant discussion, e.g. to link a bug fix to the email with
  515. the bug report. However, for a multi-patch series, it is generally
  516. best to avoid using In-Reply-To: to link to older versions of the
  517. series. This way multiple versions of the patch don't become an
  518. unmanageable forest of references in email clients. If a link is
  519. helpful, you can use the https://lkml.kernel.org/ redirector (e.g., in
  520. the cover email text) to link to an earlier version of the patch series.
  521. Providing base tree information
  522. -------------------------------
  523. When other developers receive your patches and start the review process,
  524. it is often useful for them to know where in the tree history they
  525. should place your work. This is particularly useful for automated CI
  526. processes that attempt to run a series of tests in order to establish
  527. the quality of your submission before the maintainer starts the review.
  528. If you are using ``git format-patch`` to generate your patches, you can
  529. automatically include the base tree information in your submission by
  530. using the ``--base`` flag. The easiest and most convenient way to use
  531. this option is with topical branches::
  532. $ git checkout -t -b my-topical-branch master
  533. Branch 'my-topical-branch' set up to track local branch 'master'.
  534. Switched to a new branch 'my-topical-branch'
  535. [perform your edits and commits]
  536. $ git format-patch --base=auto --cover-letter -o outgoing/ master
  537. outgoing/0000-cover-letter.patch
  538. outgoing/0001-First-Commit.patch
  539. outgoing/...
  540. When you open ``outgoing/0000-cover-letter.patch`` for editing, you will
  541. notice that it will have the ``base-commit:`` trailer at the very
  542. bottom, which provides the reviewer and the CI tools enough information
  543. to properly perform ``git am`` without worrying about conflicts::
  544. $ git checkout -b patch-review [base-commit-id]
  545. Switched to a new branch 'patch-review'
  546. $ git am patches.mbox
  547. Applying: First Commit
  548. Applying: ...
  549. Please see ``man git-format-patch`` for more information about this
  550. option.
  551. .. note::
  552. The ``--base`` feature was introduced in git version 2.9.0.
  553. If you are not using git to format your patches, you can still include
  554. the same ``base-commit`` trailer to indicate the commit hash of the tree
  555. on which your work is based. You should add it either in the cover
  556. letter or in the first patch of the series and it should be placed
  557. either below the ``---`` line or at the very bottom of all other
  558. content, right before your email signature.
  559. References
  560. ----------
  561. Andrew Morton, "The perfect patch" (tpp).
  562. <https://www.ozlabs.org/~akpm/stuff/tpp.txt>
  563. Jeff Garzik, "Linux kernel patch submission format".
  564. <https://web.archive.org/web/20180829112450/http://linux.yyz.us/patch-format.html>
  565. Greg Kroah-Hartman, "How to piss off a kernel subsystem maintainer".
  566. <http://www.kroah.com/log/linux/maintainer.html>
  567. <http://www.kroah.com/log/linux/maintainer-02.html>
  568. <http://www.kroah.com/log/linux/maintainer-03.html>
  569. <http://www.kroah.com/log/linux/maintainer-04.html>
  570. <http://www.kroah.com/log/linux/maintainer-05.html>
  571. <http://www.kroah.com/log/linux/maintainer-06.html>
  572. NO!!!! No more huge patch bombs to linux-kernel@vger.kernel.org people!
  573. <https://lkml.org/lkml/2005/7/11/336>
  574. Kernel Documentation/process/coding-style.rst:
  575. :ref:`Documentation/process/coding-style.rst <codingstyle>`
  576. Linus Torvalds's mail on the canonical patch format:
  577. <http://lkml.org/lkml/2005/4/7/183>
  578. Andi Kleen, "On submitting kernel patches"
  579. Some strategies to get difficult or controversial changes in.
  580. http://halobates.de/on-submitting-patches.pdf