3 Submitting patches: the essential guide to getting your code into the kernel
4 ============================================================================
6 For a person or company who wishes to submit a change to the Linux
7 kernel, the process can sometimes be daunting if you're not familiar
8 with "the system." This text is a collection of suggestions which
9 can greatly increase the chances of your change being accepted.
11 This document contains a large number of suggestions in a relatively terse
12 format. For detailed information on how the kernel development process
13 works, see :doc:`development-process`. Also, read :doc:`submit-checklist`
14 for a list of items to check before submitting code. If you are submitting
15 a driver, also read :doc:`submitting-drivers`; for device tree binding patches,
16 read :doc:`submitting-patches`.
18 This documentation assumes that you're using ``git`` to prepare your patches.
19 If you're unfamiliar with ``git``, you would be well-advised to learn how to
20 use it, it will make your life as a kernel developer and in general much
23 Obtain a current source tree
24 ----------------------------
26 If you do not have a repository with the current kernel source handy, use
27 ``git`` to obtain one. You'll want to start with the mainline repository,
28 which can be grabbed with::
30 git clone git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
32 Note, however, that you may not want to develop against the mainline tree
33 directly. Most subsystem maintainers run their own trees and want to see
34 patches prepared against those trees. See the **T:** entry for the subsystem
35 in the MAINTAINERS file to find that tree, or simply ask the maintainer if
36 the tree is not listed there.
43 Describe your problem. Whether your patch is a one-line bug fix or
44 5000 lines of a new feature, there must be an underlying problem that
45 motivated you to do this work. Convince the reviewer that there is a
46 problem worth fixing and that it makes sense for them to read past the
49 Describe user-visible impact. Straight up crashes and lockups are
50 pretty convincing, but not all bugs are that blatant. Even if the
51 problem was spotted during code review, describe the impact you think
52 it can have on users. Keep in mind that the majority of Linux
53 installations run kernels from secondary stable trees or
54 vendor/product-specific trees that cherry-pick only specific patches
55 from upstream, so include anything that could help route your change
56 downstream: provoking circumstances, excerpts from dmesg, crash
57 descriptions, performance regressions, latency spikes, lockups, etc.
59 Quantify optimizations and trade-offs. If you claim improvements in
60 performance, memory consumption, stack footprint, or binary size,
61 include numbers that back them up. But also describe non-obvious
62 costs. Optimizations usually aren't free but trade-offs between CPU,
63 memory, and readability; or, when it comes to heuristics, between
64 different workloads. Describe the expected downsides of your
65 optimization so that the reviewer can weigh costs against benefits.
67 Once the problem is established, describe what you are actually doing
68 about it in technical detail. It's important to describe the change
69 in plain English for the reviewer to verify that the code is behaving
72 The maintainer will thank you if you write your patch description in a
73 form which can be easily pulled into Linux's source code management
74 system, ``git``, as a "commit log". See :ref:`explicit_in_reply_to`.
76 Solve only one problem per patch. If your description starts to get
77 long, that's a sign that you probably need to split up your patch.
78 See :ref:`split_changes`.
80 When you submit or resubmit a patch or patch series, include the
81 complete patch description and justification for it. Don't just
82 say that this is version N of the patch (series). Don't expect the
83 subsystem maintainer to refer back to earlier patch versions or referenced
84 URLs to find the patch description and put that into the patch.
85 I.e., the patch (series) and its description should be self-contained.
86 This benefits both the maintainers and reviewers. Some reviewers
87 probably didn't even receive earlier versions of the patch.
89 Describe your changes in imperative mood, e.g. "make xyzzy do frotz"
90 instead of "[This patch] makes xyzzy do frotz" or "[I] changed xyzzy
91 to do frotz", as if you are giving orders to the codebase to change
94 If the patch fixes a logged bug entry, refer to that bug entry by
95 number and URL. If the patch follows from a mailing list discussion,
96 give a URL to the mailing list archive; use the https://lkml.kernel.org/
97 redirector with a ``Message-Id``, to ensure that the links cannot become
100 However, try to make your explanation understandable without external
101 resources. In addition to giving a URL to a mailing list archive or
102 bug, summarize the relevant points of the discussion that led to the
105 If you want to refer to a specific commit, don't just refer to the
106 SHA-1 ID of the commit. Please also include the oneline summary of
107 the commit, to make it easier for reviewers to know what it is about.
110 Commit e21d2170f36602ae2708 ("video: remove unnecessary
111 platform_set_drvdata()") removed the unnecessary
112 platform_set_drvdata(), but left the variable "dev" unused,
115 You should also be sure to use at least the first twelve characters of the
116 SHA-1 ID. The kernel repository holds a *lot* of objects, making
117 collisions with shorter IDs a real possibility. Bear in mind that, even if
118 there is no collision with your six-character ID now, that condition may
119 change five years from now.
121 If your patch fixes a bug in a specific commit, e.g. you found an issue using
122 ``git bisect``, please use the 'Fixes:' tag with the first 12 characters of
123 the SHA-1 ID, and the one line summary. Do not split the tag across multiple
124 lines, tags are exempt from the "wrap at 75 columns" rule in order to simplify
125 parsing scripts. For example::
127 Fixes: 54a4f0239f2e ("KVM: MMU: make kvm_mmu_zap_page() return the number of pages it actually freed")
129 The following ``git config`` settings can be used to add a pretty format for
130 outputting the above style in the ``git log`` or ``git show`` commands::
135 fixes = Fixes: %h (\"%s\")
139 $ git log -1 --pretty=fixes 54a4f0239f2e
140 Fixes: 54a4f0239f2e ("KVM: MMU: make kvm_mmu_zap_page() return the number of pages it actually freed")
144 Separate your changes
145 ---------------------
147 Separate each **logical change** into a separate patch.
149 For example, if your changes include both bug fixes and performance
150 enhancements for a single driver, separate those changes into two
151 or more patches. If your changes include an API update, and a new
152 driver which uses that new API, separate those into two patches.
154 On the other hand, if you make a single change to numerous files,
155 group those changes into a single patch. Thus a single logical change
156 is contained within a single patch.
158 The point to remember is that each patch should make an easily understood
159 change that can be verified by reviewers. Each patch should be justifiable
162 If one patch depends on another patch in order for a change to be
163 complete, that is OK. Simply note **"this patch depends on patch X"**
164 in your patch description.
166 When dividing your change into a series of patches, take special care to
167 ensure that the kernel builds and runs properly after each patch in the
168 series. Developers using ``git bisect`` to track down a problem can end up
169 splitting your patch series at any point; they will not thank you if you
170 introduce bugs in the middle.
172 If you cannot condense your patch set into a smaller set of patches,
173 then only post say 15 or so at a time and wait for review and integration.
177 Style-check your changes
178 ------------------------
180 Check your patch for basic style violations, details of which can be
182 :ref:`Documentation/process/coding-style.rst <codingstyle>`.
183 Failure to do so simply wastes
184 the reviewers time and will get your patch rejected, probably
185 without even being read.
187 One significant exception is when moving code from one file to
188 another -- in this case you should not modify the moved code at all in
189 the same patch which moves it. This clearly delineates the act of
190 moving the code and your changes. This greatly aids review of the
191 actual differences and allows tools to better track the history of
194 Check your patches with the patch style checker prior to submission
195 (scripts/checkpatch.pl). Note, though, that the style checker should be
196 viewed as a guide, not as a replacement for human judgment. If your code
197 looks better with a violation then its probably best left alone.
199 The checker reports at three levels:
200 - ERROR: things that are very likely to be wrong
201 - WARNING: things requiring careful review
202 - CHECK: things requiring thought
204 You should be able to justify all violations that remain in your
208 Select the recipients for your patch
209 ------------------------------------
211 You should always copy the appropriate subsystem maintainer(s) on any patch
212 to code that they maintain; look through the MAINTAINERS file and the
213 source code revision history to see who those maintainers are. The
214 script scripts/get_maintainer.pl can be very useful at this step. If you
215 cannot find a maintainer for the subsystem you are working on, Andrew
216 Morton (akpm@linux-foundation.org) serves as a maintainer of last resort.
218 You should also normally choose at least one mailing list to receive a copy
219 of your patch set. linux-kernel@vger.kernel.org functions as a list of
220 last resort, but the volume on that list has caused a number of developers
221 to tune it out. Look in the MAINTAINERS file for a subsystem-specific
222 list; your patch will probably get more attention there. Please do not
223 spam unrelated lists, though.
225 Many kernel-related lists are hosted on vger.kernel.org; you can find a
226 list of them at http://vger.kernel.org/vger-lists.html. There are
227 kernel-related lists hosted elsewhere as well, though.
229 Do not send more than 15 patches at once to the vger mailing lists!!!
231 Linus Torvalds is the final arbiter of all changes accepted into the
232 Linux kernel. His e-mail address is <torvalds@linux-foundation.org>.
233 He gets a lot of e-mail, and, at this point, very few patches go through
234 Linus directly, so typically you should do your best to -avoid-
237 If you have a patch that fixes an exploitable security bug, send that patch
238 to security@kernel.org. For severe bugs, a short embargo may be considered
239 to allow distributors to get the patch out to users; in such cases,
240 obviously, the patch should not be sent to any public lists. See also
241 :doc:`/admin-guide/security-bugs`.
243 Patches that fix a severe bug in a released kernel should be directed
244 toward the stable maintainers by putting a line like this::
246 Cc: stable@vger.kernel.org
248 into the sign-off area of your patch (note, NOT an email recipient). You
250 :ref:`Documentation/process/stable-kernel-rules.rst <stable_kernel_rules>`
251 in addition to this file.
253 Note, however, that some subsystem maintainers want to come to their own
254 conclusions on which patches should go to the stable trees. The networking
255 maintainer, in particular, would rather not see individual developers
256 adding lines like the above to their patches.
258 If changes affect userland-kernel interfaces, please send the MAN-PAGES
259 maintainer (as listed in the MAINTAINERS file) a man-pages patch, or at
260 least a notification of the change, so that some information makes its way
261 into the manual pages. User-space API changes should also be copied to
262 linux-api@vger.kernel.org.
264 For small patches you may want to CC the Trivial Patch Monkey
265 trivial@kernel.org which collects "trivial" patches. Have a look
266 into the MAINTAINERS file for its current manager.
268 Trivial patches must qualify for one of the following rules:
270 - Spelling fixes in documentation
271 - Spelling fixes for errors which could break :manpage:`grep(1)`
272 - Warning fixes (cluttering with useless warnings is bad)
273 - Compilation fixes (only if they are actually correct)
274 - Runtime fixes (only if they actually fix things)
275 - Removing use of deprecated functions/macros
276 - Contact detail and documentation fixes
277 - Non-portable code replaced by portable code (even in arch-specific,
278 since people copy, as long as it's trivial)
279 - Any fix by the author/maintainer of the file (ie. patch monkey
280 in re-transmission mode)
284 No MIME, no links, no compression, no attachments. Just plain text
285 -------------------------------------------------------------------
287 Linus and other kernel developers need to be able to read and comment
288 on the changes you are submitting. It is important for a kernel
289 developer to be able to "quote" your changes, using standard e-mail
290 tools, so that they may comment on specific portions of your code.
292 For this reason, all patches should be submitted by e-mail "inline". The
293 easiest way to do this is with ``git send-email``, which is strongly
294 recommended. An interactive tutorial for ``git send-email`` is available at
295 https://git-send-email.io.
297 If you choose not to use ``git send-email``:
301 Be wary of your editor's word-wrap corrupting your patch,
302 if you choose to cut-n-paste your patch.
304 Do not attach the patch as a MIME attachment, compressed or not.
305 Many popular e-mail applications will not always transmit a MIME
306 attachment as plain text, making it impossible to comment on your
307 code. A MIME attachment also takes Linus a bit more time to process,
308 decreasing the likelihood of your MIME-attached change being accepted.
310 Exception: If your mailer is mangling patches then someone may ask
311 you to re-send them using MIME.
313 See :doc:`/process/email-clients` for hints about configuring your e-mail
314 client so that it sends your patches untouched.
316 Respond to review comments
317 --------------------------
319 Your patch will almost certainly get comments from reviewers on ways in
320 which the patch can be improved, in the form of a reply to your email. You must
321 respond to those comments; ignoring reviewers is a good way to get ignored in
322 return. You can simply reply to their emails to answer their comments. Review
323 comments or questions that do not lead to a code change should almost certainly
324 bring about a comment or changelog entry so that the next reviewer better
325 understands what is going on.
327 Be sure to tell the reviewers what changes you are making and to thank them
328 for their time. Code review is a tiring and time-consuming process, and
329 reviewers sometimes get grumpy. Even in that case, though, respond
330 politely and address the problems they have pointed out.
332 See :doc:`email-clients` for recommendations on email
333 clients and mailing list etiquette.
336 Don't get discouraged - or impatient
337 ------------------------------------
339 After you have submitted your change, be patient and wait. Reviewers are
340 busy people and may not get to your patch right away.
342 Once upon a time, patches used to disappear into the void without comment,
343 but the development process works more smoothly than that now. You should
344 receive comments within a week or so; if that does not happen, make sure
345 that you have sent your patches to the right place. Wait for a minimum of
346 one week before resubmitting or pinging reviewers - possibly longer during
347 busy times like merge windows.
350 Include PATCH in the subject
351 -----------------------------
353 Due to high e-mail traffic to Linus, and to linux-kernel, it is common
354 convention to prefix your subject line with [PATCH]. This lets Linus
355 and other kernel developers more easily distinguish patches from other
358 ``git send-email`` will do this for you automatically.
361 Sign your work - the Developer's Certificate of Origin
362 ------------------------------------------------------
364 To improve tracking of who did what, especially with patches that can
365 percolate to their final resting place in the kernel through several
366 layers of maintainers, we've introduced a "sign-off" procedure on
367 patches that are being emailed around.
369 The sign-off is a simple line at the end of the explanation for the
370 patch, which certifies that you wrote it or otherwise have the right to
371 pass it on as an open-source patch. The rules are pretty simple: if you
372 can certify the below:
374 Developer's Certificate of Origin 1.1
375 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
377 By making a contribution to this project, I certify that:
379 (a) The contribution was created in whole or in part by me and I
380 have the right to submit it under the open source license
381 indicated in the file; or
383 (b) The contribution is based upon previous work that, to the best
384 of my knowledge, is covered under an appropriate open source
385 license and I have the right under that license to submit that
386 work with modifications, whether created in whole or in part
387 by me, under the same open source license (unless I am
388 permitted to submit under a different license), as indicated
391 (c) The contribution was provided directly to me by some other
392 person who certified (a), (b) or (c) and I have not modified
395 (d) I understand and agree that this project and the contribution
396 are public and that a record of the contribution (including all
397 personal information I submit with it, including my sign-off) is
398 maintained indefinitely and may be redistributed consistent with
399 this project or the open source license(s) involved.
401 then you just add a line saying::
403 Signed-off-by: Random J Developer <random@developer.example.org>
405 using your real name (sorry, no pseudonyms or anonymous contributions.)
406 This will be done for you automatically if you use ``git commit -s``.
407 Reverts should also include "Signed-off-by". ``git revert -s`` does that
410 Some people also put extra tags at the end. They'll just be ignored for
411 now, but you can do this to mark internal company procedures or just
412 point out some special detail about the sign-off.
415 When to use Acked-by:, Cc:, and Co-developed-by:
416 ------------------------------------------------
418 The Signed-off-by: tag indicates that the signer was involved in the
419 development of the patch, or that he/she was in the patch's delivery path.
421 If a person was not directly involved in the preparation or handling of a
422 patch but wishes to signify and record their approval of it then they can
423 ask to have an Acked-by: line added to the patch's changelog.
425 Acked-by: is often used by the maintainer of the affected code when that
426 maintainer neither contributed to nor forwarded the patch.
428 Acked-by: is not as formal as Signed-off-by:. It is a record that the acker
429 has at least reviewed the patch and has indicated acceptance. Hence patch
430 mergers will sometimes manually convert an acker's "yep, looks good to me"
431 into an Acked-by: (but note that it is usually better to ask for an
434 Acked-by: does not necessarily indicate acknowledgement of the entire patch.
435 For example, if a patch affects multiple subsystems and has an Acked-by: from
436 one subsystem maintainer then this usually indicates acknowledgement of just
437 the part which affects that maintainer's code. Judgement should be used here.
438 When in doubt people should refer to the original discussion in the mailing
441 If a person has had the opportunity to comment on a patch, but has not
442 provided such comments, you may optionally add a ``Cc:`` tag to the patch.
443 This is the only tag which might be added without an explicit action by the
444 person it names - but it should indicate that this person was copied on the
445 patch. This tag documents that potentially interested parties
446 have been included in the discussion.
448 Co-developed-by: states that the patch was co-created by multiple developers;
449 it is a used to give attribution to co-authors (in addition to the author
450 attributed by the From: tag) when several people work on a single patch. Since
451 Co-developed-by: denotes authorship, every Co-developed-by: must be immediately
452 followed by a Signed-off-by: of the associated co-author. Standard sign-off
453 procedure applies, i.e. the ordering of Signed-off-by: tags should reflect the
454 chronological history of the patch insofar as possible, regardless of whether
455 the author is attributed via From: or Co-developed-by:. Notably, the last
456 Signed-off-by: must always be that of the developer submitting the patch.
458 Note, the From: tag is optional when the From: author is also the person (and
459 email) listed in the From: line of the email header.
461 Example of a patch submitted by the From: author::
465 Co-developed-by: First Co-Author <first@coauthor.example.org>
466 Signed-off-by: First Co-Author <first@coauthor.example.org>
467 Co-developed-by: Second Co-Author <second@coauthor.example.org>
468 Signed-off-by: Second Co-Author <second@coauthor.example.org>
469 Signed-off-by: From Author <from@author.example.org>
471 Example of a patch submitted by a Co-developed-by: author::
473 From: From Author <from@author.example.org>
477 Co-developed-by: Random Co-Author <random@coauthor.example.org>
478 Signed-off-by: Random Co-Author <random@coauthor.example.org>
479 Signed-off-by: From Author <from@author.example.org>
480 Co-developed-by: Submitting Co-Author <sub@coauthor.example.org>
481 Signed-off-by: Submitting Co-Author <sub@coauthor.example.org>
484 Using Reported-by:, Tested-by:, Reviewed-by:, Suggested-by: and Fixes:
485 ----------------------------------------------------------------------
487 The Reported-by tag gives credit to people who find bugs and report them and it
488 hopefully inspires them to help us again in the future. Please note that if
489 the bug was reported in private, then ask for permission first before using the
492 A Tested-by: tag indicates that the patch has been successfully tested (in
493 some environment) by the person named. This tag informs maintainers that
494 some testing has been performed, provides a means to locate testers for
495 future patches, and ensures credit for the testers.
497 Reviewed-by:, instead, indicates that the patch has been reviewed and found
498 acceptable according to the Reviewer's Statement:
500 Reviewer's statement of oversight
501 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
503 By offering my Reviewed-by: tag, I state that:
505 (a) I have carried out a technical review of this patch to
506 evaluate its appropriateness and readiness for inclusion into
509 (b) Any problems, concerns, or questions relating to the patch
510 have been communicated back to the submitter. I am satisfied
511 with the submitter's response to my comments.
513 (c) While there may be things that could be improved with this
514 submission, I believe that it is, at this time, (1) a
515 worthwhile modification to the kernel, and (2) free of known
516 issues which would argue against its inclusion.
518 (d) While I have reviewed the patch and believe it to be sound, I
519 do not (unless explicitly stated elsewhere) make any
520 warranties or guarantees that it will achieve its stated
521 purpose or function properly in any given situation.
523 A Reviewed-by tag is a statement of opinion that the patch is an
524 appropriate modification of the kernel without any remaining serious
525 technical issues. Any interested reviewer (who has done the work) can
526 offer a Reviewed-by tag for a patch. This tag serves to give credit to
527 reviewers and to inform maintainers of the degree of review which has been
528 done on the patch. Reviewed-by: tags, when supplied by reviewers known to
529 understand the subject area and to perform thorough reviews, will normally
530 increase the likelihood of your patch getting into the kernel.
532 Both Tested-by and Reviewed-by tags, once received on mailing list from tester
533 or reviewer, should be added by author to the applicable patches when sending
534 next versions. However if the patch has changed substantially in following
535 version, these tags might not be applicable anymore and thus should be removed.
536 Usually removal of someone's Tested-by or Reviewed-by tags should be mentioned
537 in the patch changelog (after the '---' separator).
539 A Suggested-by: tag indicates that the patch idea is suggested by the person
540 named and ensures credit to the person for the idea. Please note that this
541 tag should not be added without the reporter's permission, especially if the
542 idea was not posted in a public forum. That said, if we diligently credit our
543 idea reporters, they will, hopefully, be inspired to help us again in the
546 A Fixes: tag indicates that the patch fixes an issue in a previous commit. It
547 is used to make it easy to determine where a bug originated, which can help
548 review a bug fix. This tag also assists the stable kernel team in determining
549 which stable kernel versions should receive your fix. This is the preferred
550 method for indicating a bug fixed by the patch. See :ref:`describe_changes`
553 .. _the_canonical_patch_format:
555 The canonical patch format
556 --------------------------
558 This section describes how the patch itself should be formatted. Note
559 that, if you have your patches stored in a ``git`` repository, proper patch
560 formatting can be had with ``git format-patch``. The tools cannot create
561 the necessary text, though, so read the instructions below anyway.
563 The canonical patch subject line is::
565 Subject: [PATCH 001/123] subsystem: summary phrase
567 The canonical patch message body contains the following:
569 - A ``from`` line specifying the patch author, followed by an empty
570 line (only needed if the person sending the patch is not the author).
572 - The body of the explanation, line wrapped at 75 columns, which will
573 be copied to the permanent changelog to describe this patch.
577 - The ``Signed-off-by:`` lines, described above, which will
578 also go in the changelog.
580 - A marker line containing simply ``---``.
582 - Any additional comments not suitable for the changelog.
584 - The actual patch (``diff`` output).
586 The Subject line format makes it very easy to sort the emails
587 alphabetically by subject line - pretty much any email reader will
588 support that - since because the sequence number is zero-padded,
589 the numerical and alphabetic sort is the same.
591 The ``subsystem`` in the email's Subject should identify which
592 area or subsystem of the kernel is being patched.
594 The ``summary phrase`` in the email's Subject should concisely
595 describe the patch which that email contains. The ``summary
596 phrase`` should not be a filename. Do not use the same ``summary
597 phrase`` for every patch in a whole patch series (where a ``patch
598 series`` is an ordered sequence of multiple, related patches).
600 Bear in mind that the ``summary phrase`` of your email becomes a
601 globally-unique identifier for that patch. It propagates all the way
602 into the ``git`` changelog. The ``summary phrase`` may later be used in
603 developer discussions which refer to the patch. People will want to
604 google for the ``summary phrase`` to read discussion regarding that
605 patch. It will also be the only thing that people may quickly see
606 when, two or three months later, they are going through perhaps
607 thousands of patches using tools such as ``gitk`` or ``git log
610 For these reasons, the ``summary`` must be no more than 70-75
611 characters, and it must describe both what the patch changes, as well
612 as why the patch might be necessary. It is challenging to be both
613 succinct and descriptive, but that is what a well-written summary
616 The ``summary phrase`` may be prefixed by tags enclosed in square
617 brackets: "Subject: [PATCH <tag>...] <summary phrase>". The tags are
618 not considered part of the summary phrase, but describe how the patch
619 should be treated. Common tags might include a version descriptor if
620 the multiple versions of the patch have been sent out in response to
621 comments (i.e., "v1, v2, v3"), or "RFC" to indicate a request for
622 comments. If there are four patches in a patch series the individual
623 patches may be numbered like this: 1/4, 2/4, 3/4, 4/4. This assures
624 that developers understand the order in which the patches should be
625 applied and that they have reviewed or applied all of the patches in
628 A couple of example Subjects::
630 Subject: [PATCH 2/5] ext2: improve scalability of bitmap searching
631 Subject: [PATCH v2 01/27] x86: fix eflags tracking
633 The ``from`` line must be the very first line in the message body,
636 From: Patch Author <author@example.com>
638 The ``from`` line specifies who will be credited as the author of the
639 patch in the permanent changelog. If the ``from`` line is missing,
640 then the ``From:`` line from the email header will be used to determine
641 the patch author in the changelog.
643 The explanation body will be committed to the permanent source
644 changelog, so should make sense to a competent reader who has long
645 since forgotten the immediate details of the discussion that might
646 have led to this patch. Including symptoms of the failure which the
647 patch addresses (kernel log messages, oops messages, etc.) is
648 especially useful for people who might be searching the commit logs
649 looking for the applicable patch. If a patch fixes a compile failure,
650 it may not be necessary to include _all_ of the compile failures; just
651 enough that it is likely that someone searching for the patch can find
652 it. As in the ``summary phrase``, it is important to be both succinct as
655 The ``---`` marker line serves the essential purpose of marking for patch
656 handling tools where the changelog message ends.
658 One good use for the additional comments after the ``---`` marker is for
659 a ``diffstat``, to show what files have changed, and the number of
660 inserted and deleted lines per file. A ``diffstat`` is especially useful
661 on bigger patches. Other comments relevant only to the moment or the
662 maintainer, not suitable for the permanent changelog, should also go
663 here. A good example of such comments might be ``patch changelogs``
664 which describe what has changed between the v1 and v2 version of the
667 If you are going to include a ``diffstat`` after the ``---`` marker, please
668 use ``diffstat`` options ``-p 1 -w 70`` so that filenames are listed from
669 the top of the kernel source tree and don't use too much horizontal
670 space (easily fit in 80 columns, maybe with some indentation). (``git``
671 generates appropriate diffstats by default.)
673 See more details on the proper patch format in the following
676 .. _explicit_in_reply_to:
678 Explicit In-Reply-To headers
679 ----------------------------
681 It can be helpful to manually add In-Reply-To: headers to a patch
682 (e.g., when using ``git send-email``) to associate the patch with
683 previous relevant discussion, e.g. to link a bug fix to the email with
684 the bug report. However, for a multi-patch series, it is generally
685 best to avoid using In-Reply-To: to link to older versions of the
686 series. This way multiple versions of the patch don't become an
687 unmanageable forest of references in email clients. If a link is
688 helpful, you can use the https://lkml.kernel.org/ redirector (e.g., in
689 the cover email text) to link to an earlier version of the patch series.
692 Providing base tree information
693 -------------------------------
695 When other developers receive your patches and start the review process,
696 it is often useful for them to know where in the tree history they
697 should place your work. This is particularly useful for automated CI
698 processes that attempt to run a series of tests in order to establish
699 the quality of your submission before the maintainer starts the review.
701 If you are using ``git format-patch`` to generate your patches, you can
702 automatically include the base tree information in your submission by
703 using the ``--base`` flag. The easiest and most convenient way to use
704 this option is with topical branches::
706 $ git checkout -t -b my-topical-branch master
707 Branch 'my-topical-branch' set up to track local branch 'master'.
708 Switched to a new branch 'my-topical-branch'
710 [perform your edits and commits]
712 $ git format-patch --base=auto --cover-letter -o outgoing/ master
713 outgoing/0000-cover-letter.patch
714 outgoing/0001-First-Commit.patch
717 When you open ``outgoing/0000-cover-letter.patch`` for editing, you will
718 notice that it will have the ``base-commit:`` trailer at the very
719 bottom, which provides the reviewer and the CI tools enough information
720 to properly perform ``git am`` without worrying about conflicts::
722 $ git checkout -b patch-review [base-commit-id]
723 Switched to a new branch 'patch-review'
724 $ git am patches.mbox
725 Applying: First Commit
728 Please see ``man git-format-patch`` for more information about this
733 The ``--base`` feature was introduced in git version 2.9.0.
735 If you are not using git to format your patches, you can still include
736 the same ``base-commit`` trailer to indicate the commit hash of the tree
737 on which your work is based. You should add it either in the cover
738 letter or in the first patch of the series and it should be placed
739 either below the ``---`` line or at the very bottom of all other
740 content, right before your email signature.
746 Andrew Morton, "The perfect patch" (tpp).
747 <https://www.ozlabs.org/~akpm/stuff/tpp.txt>
749 Jeff Garzik, "Linux kernel patch submission format".
750 <https://web.archive.org/web/20180829112450/http://linux.yyz.us/patch-format.html>
752 Greg Kroah-Hartman, "How to piss off a kernel subsystem maintainer".
753 <http://www.kroah.com/log/linux/maintainer.html>
755 <http://www.kroah.com/log/linux/maintainer-02.html>
757 <http://www.kroah.com/log/linux/maintainer-03.html>
759 <http://www.kroah.com/log/linux/maintainer-04.html>
761 <http://www.kroah.com/log/linux/maintainer-05.html>
763 <http://www.kroah.com/log/linux/maintainer-06.html>
765 NO!!!! No more huge patch bombs to linux-kernel@vger.kernel.org people!
766 <https://lkml.org/lkml/2005/7/11/336>
768 Kernel Documentation/process/coding-style.rst:
769 :ref:`Documentation/process/coding-style.rst <codingstyle>`
771 Linus Torvalds's mail on the canonical patch format:
772 <http://lkml.org/lkml/2005/4/7/183>
774 Andi Kleen, "On submitting kernel patches"
775 Some strategies to get difficult or controversial changes in.
777 http://halobates.de/on-submitting-patches.pdf