• Home
  • Line#
  • Scopes#
  • Navigate#
  • Raw
  • Download
1.. _submittingpatches:
2
3Submitting patches: the essential guide to getting your code into the kernel
4============================================================================
5
6For a person or company who wishes to submit a change to the Linux
7kernel, the process can sometimes be daunting if you're not familiar
8with "the system."  This text is a collection of suggestions which
9can greatly increase the chances of your change being accepted.
10
11This document contains a large number of suggestions in a relatively terse
12format.  For detailed information on how the kernel development process
13works, see :doc:`development-process`. Also, read :doc:`submit-checklist`
14for a list of items to check before submitting code.  If you are submitting
15a driver, also read :doc:`submitting-drivers`; for device tree binding patches,
16read :doc:`submitting-patches`.
17
18This documentation assumes that you're using ``git`` to prepare your patches.
19If you're unfamiliar with ``git``, you would be well-advised to learn how to
20use it, it will make your life as a kernel developer and in general much
21easier.
22
23Obtain a current source tree
24----------------------------
25
26If 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,
28which can be grabbed with::
29
30  git clone git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
31
32Note, however, that you may not want to develop against the mainline tree
33directly.  Most subsystem maintainers run their own trees and want to see
34patches prepared against those trees.  See the **T:** entry for the subsystem
35in the MAINTAINERS file to find that tree, or simply ask the maintainer if
36the tree is not listed there.
37
38.. _describe_changes:
39
40Describe your changes
41---------------------
42
43Describe your problem.  Whether your patch is a one-line bug fix or
445000 lines of a new feature, there must be an underlying problem that
45motivated you to do this work.  Convince the reviewer that there is a
46problem worth fixing and that it makes sense for them to read past the
47first paragraph.
48
49Describe user-visible impact.  Straight up crashes and lockups are
50pretty convincing, but not all bugs are that blatant.  Even if the
51problem was spotted during code review, describe the impact you think
52it can have on users.  Keep in mind that the majority of Linux
53installations run kernels from secondary stable trees or
54vendor/product-specific trees that cherry-pick only specific patches
55from upstream, so include anything that could help route your change
56downstream: provoking circumstances, excerpts from dmesg, crash
57descriptions, performance regressions, latency spikes, lockups, etc.
58
59Quantify optimizations and trade-offs.  If you claim improvements in
60performance, memory consumption, stack footprint, or binary size,
61include numbers that back them up.  But also describe non-obvious
62costs.  Optimizations usually aren't free but trade-offs between CPU,
63memory, and readability; or, when it comes to heuristics, between
64different workloads.  Describe the expected downsides of your
65optimization so that the reviewer can weigh costs against benefits.
66
67Once the problem is established, describe what you are actually doing
68about it in technical detail.  It's important to describe the change
69in plain English for the reviewer to verify that the code is behaving
70as you intend it to.
71
72The maintainer will thank you if you write your patch description in a
73form which can be easily pulled into Linux's source code management
74system, ``git``, as a "commit log".  See :ref:`explicit_in_reply_to`.
75
76Solve only one problem per patch.  If your description starts to get
77long, that's a sign that you probably need to split up your patch.
78See :ref:`split_changes`.
79
80When you submit or resubmit a patch or patch series, include the
81complete patch description and justification for it.  Don't just
82say that this is version N of the patch (series).  Don't expect the
83subsystem maintainer to refer back to earlier patch versions or referenced
84URLs to find the patch description and put that into the patch.
85I.e., the patch (series) and its description should be self-contained.
86This benefits both the maintainers and reviewers.  Some reviewers
87probably didn't even receive earlier versions of the patch.
88
89Describe your changes in imperative mood, e.g. "make xyzzy do frotz"
90instead of "[This patch] makes xyzzy do frotz" or "[I] changed xyzzy
91to do frotz", as if you are giving orders to the codebase to change
92its behaviour.
93
94If the patch fixes a logged bug entry, refer to that bug entry by
95number and URL.  If the patch follows from a mailing list discussion,
96give a URL to the mailing list archive; use the https://lkml.kernel.org/
97redirector with a ``Message-Id``, to ensure that the links cannot become
98stale.
99
100However, try to make your explanation understandable without external
101resources.  In addition to giving a URL to a mailing list archive or
102bug, summarize the relevant points of the discussion that led to the
103patch as submitted.
104
105If you want to refer to a specific commit, don't just refer to the
106SHA-1 ID of the commit. Please also include the oneline summary of
107the commit, to make it easier for reviewers to know what it is about.
108Example::
109
110	Commit e21d2170f36602ae2708 ("video: remove unnecessary
111	platform_set_drvdata()") removed the unnecessary
112	platform_set_drvdata(), but left the variable "dev" unused,
113	delete it.
114
115You should also be sure to use at least the first twelve characters of the
116SHA-1 ID.  The kernel repository holds a *lot* of objects, making
117collisions with shorter IDs a real possibility.  Bear in mind that, even if
118there is no collision with your six-character ID now, that condition may
119change five years from now.
120
121If 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
123the SHA-1 ID, and the one line summary.  Do not split the tag across multiple
124lines, tags are exempt from the "wrap at 75 columns" rule in order to simplify
125parsing scripts.  For example::
126
127	Fixes: 54a4f0239f2e ("KVM: MMU: make kvm_mmu_zap_page() return the number of pages it actually freed")
128
129The following ``git config`` settings can be used to add a pretty format for
130outputting the above style in the ``git log`` or ``git show`` commands::
131
132	[core]
133		abbrev = 12
134	[pretty]
135		fixes = Fixes: %h (\"%s\")
136
137An example call::
138
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")
141
142.. _split_changes:
143
144Separate your changes
145---------------------
146
147Separate each **logical change** into a separate patch.
148
149For example, if your changes include both bug fixes and performance
150enhancements for a single driver, separate those changes into two
151or more patches.  If your changes include an API update, and a new
152driver which uses that new API, separate those into two patches.
153
154On the other hand, if you make a single change to numerous files,
155group those changes into a single patch.  Thus a single logical change
156is contained within a single patch.
157
158The point to remember is that each patch should make an easily understood
159change that can be verified by reviewers.  Each patch should be justifiable
160on its own merits.
161
162If one patch depends on another patch in order for a change to be
163complete, that is OK.  Simply note **"this patch depends on patch X"**
164in your patch description.
165
166When dividing your change into a series of patches, take special care to
167ensure that the kernel builds and runs properly after each patch in the
168series.  Developers using ``git bisect`` to track down a problem can end up
169splitting your patch series at any point; they will not thank you if you
170introduce bugs in the middle.
171
172If you cannot condense your patch set into a smaller set of patches,
173then only post say 15 or so at a time and wait for review and integration.
174
175
176
177Style-check your changes
178------------------------
179
180Check your patch for basic style violations, details of which can be
181found in
182:ref:`Documentation/process/coding-style.rst <codingstyle>`.
183Failure to do so simply wastes
184the reviewers time and will get your patch rejected, probably
185without even being read.
186
187One significant exception is when moving code from one file to
188another -- in this case you should not modify the moved code at all in
189the same patch which moves it.  This clearly delineates the act of
190moving the code and your changes.  This greatly aids review of the
191actual differences and allows tools to better track the history of
192the code itself.
193
194Check your patches with the patch style checker prior to submission
195(scripts/checkpatch.pl).  Note, though, that the style checker should be
196viewed as a guide, not as a replacement for human judgment.  If your code
197looks better with a violation then its probably best left alone.
198
199The 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
203
204You should be able to justify all violations that remain in your
205patch.
206
207
208Select the recipients for your patch
209------------------------------------
210
211You should always copy the appropriate subsystem maintainer(s) on any patch
212to code that they maintain; look through the MAINTAINERS file and the
213source code revision history to see who those maintainers are.  The
214script scripts/get_maintainer.pl can be very useful at this step.  If you
215cannot find a maintainer for the subsystem you are working on, Andrew
216Morton (akpm@linux-foundation.org) serves as a maintainer of last resort.
217
218You should also normally choose at least one mailing list to receive a copy
219of your patch set.  linux-kernel@vger.kernel.org functions as a list of
220last resort, but the volume on that list has caused a number of developers
221to tune it out.  Look in the MAINTAINERS file for a subsystem-specific
222list; your patch will probably get more attention there.  Please do not
223spam unrelated lists, though.
224
225Many kernel-related lists are hosted on vger.kernel.org; you can find a
226list of them at http://vger.kernel.org/vger-lists.html.  There are
227kernel-related lists hosted elsewhere as well, though.
228
229Do not send more than 15 patches at once to the vger mailing lists!!!
230
231Linus Torvalds is the final arbiter of all changes accepted into the
232Linux kernel.  His e-mail address is <torvalds@linux-foundation.org>.
233He gets a lot of e-mail, and, at this point, very few patches go through
234Linus directly, so typically you should do your best to -avoid-
235sending him e-mail.
236
237If you have a patch that fixes an exploitable security bug, send that patch
238to security@kernel.org.  For severe bugs, a short embargo may be considered
239to allow distributors to get the patch out to users; in such cases,
240obviously, the patch should not be sent to any public lists. See also
241:doc:`/admin-guide/security-bugs`.
242
243Patches that fix a severe bug in a released kernel should be directed
244toward the stable maintainers by putting a line like this::
245
246  Cc: stable@vger.kernel.org
247
248into the sign-off area of your patch (note, NOT an email recipient).  You
249should also read
250:ref:`Documentation/process/stable-kernel-rules.rst <stable_kernel_rules>`
251in addition to this file.
252
253If changes affect userland-kernel interfaces, please send the MAN-PAGES
254maintainer (as listed in the MAINTAINERS file) a man-pages patch, or at
255least a notification of the change, so that some information makes its way
256into the manual pages.  User-space API changes should also be copied to
257linux-api@vger.kernel.org.
258
259For small patches you may want to CC the Trivial Patch Monkey
260trivial@kernel.org which collects "trivial" patches. Have a look
261into the MAINTAINERS file for its current manager.
262
263Trivial patches must qualify for one of the following rules:
264
265- Spelling fixes in documentation
266- Spelling fixes for errors which could break :manpage:`grep(1)`
267- Warning fixes (cluttering with useless warnings is bad)
268- Compilation fixes (only if they are actually correct)
269- Runtime fixes (only if they actually fix things)
270- Removing use of deprecated functions/macros
271- Contact detail and documentation fixes
272- Non-portable code replaced by portable code (even in arch-specific,
273  since people copy, as long as it's trivial)
274- Any fix by the author/maintainer of the file (ie. patch monkey
275  in re-transmission mode)
276
277
278
279No MIME, no links, no compression, no attachments.  Just plain text
280-------------------------------------------------------------------
281
282Linus and other kernel developers need to be able to read and comment
283on the changes you are submitting.  It is important for a kernel
284developer to be able to "quote" your changes, using standard e-mail
285tools, so that they may comment on specific portions of your code.
286
287For this reason, all patches should be submitted by e-mail "inline". The
288easiest way to do this is with ``git send-email``, which is strongly
289recommended.  An interactive tutorial for ``git send-email`` is available at
290https://git-send-email.io.
291
292If you choose not to use ``git send-email``:
293
294.. warning::
295
296  Be wary of your editor's word-wrap corrupting your patch,
297  if you choose to cut-n-paste your patch.
298
299Do not attach the patch as a MIME attachment, compressed or not.
300Many popular e-mail applications will not always transmit a MIME
301attachment as plain text, making it impossible to comment on your
302code.  A MIME attachment also takes Linus a bit more time to process,
303decreasing the likelihood of your MIME-attached change being accepted.
304
305Exception:  If your mailer is mangling patches then someone may ask
306you to re-send them using MIME.
307
308See :doc:`/process/email-clients` for hints about configuring your e-mail
309client so that it sends your patches untouched.
310
311Respond to review comments
312--------------------------
313
314Your patch will almost certainly get comments from reviewers on ways in
315which the patch can be improved, in the form of a reply to your email. You must
316respond to those comments; ignoring reviewers is a good way to get ignored in
317return. You can simply reply to their emails to answer their comments. Review
318comments or questions that do not lead to a code change should almost certainly
319bring about a comment or changelog entry so that the next reviewer better
320understands what is going on.
321
322Be sure to tell the reviewers what changes you are making and to thank them
323for their time.  Code review is a tiring and time-consuming process, and
324reviewers sometimes get grumpy.  Even in that case, though, respond
325politely and address the problems they have pointed out.
326
327See :doc:`email-clients` for recommendations on email
328clients and mailing list etiquette.
329
330
331Don't get discouraged - or impatient
332------------------------------------
333
334After you have submitted your change, be patient and wait.  Reviewers are
335busy people and may not get to your patch right away.
336
337Once upon a time, patches used to disappear into the void without comment,
338but the development process works more smoothly than that now.  You should
339receive comments within a week or so; if that does not happen, make sure
340that you have sent your patches to the right place.  Wait for a minimum of
341one week before resubmitting or pinging reviewers - possibly longer during
342busy times like merge windows.
343
344
345Include PATCH in the subject
346-----------------------------
347
348Due to high e-mail traffic to Linus, and to linux-kernel, it is common
349convention to prefix your subject line with [PATCH].  This lets Linus
350and other kernel developers more easily distinguish patches from other
351e-mail discussions.
352
353``git send-email`` will do this for you automatically.
354
355
356Sign your work - the Developer's Certificate of Origin
357------------------------------------------------------
358
359To improve tracking of who did what, especially with patches that can
360percolate to their final resting place in the kernel through several
361layers of maintainers, we've introduced a "sign-off" procedure on
362patches that are being emailed around.
363
364The sign-off is a simple line at the end of the explanation for the
365patch, which certifies that you wrote it or otherwise have the right to
366pass it on as an open-source patch.  The rules are pretty simple: if you
367can certify the below:
368
369Developer's Certificate of Origin 1.1
370^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
371
372By making a contribution to this project, I certify that:
373
374        (a) The contribution was created in whole or in part by me and I
375            have the right to submit it under the open source license
376            indicated in the file; or
377
378        (b) The contribution is based upon previous work that, to the best
379            of my knowledge, is covered under an appropriate open source
380            license and I have the right under that license to submit that
381            work with modifications, whether created in whole or in part
382            by me, under the same open source license (unless I am
383            permitted to submit under a different license), as indicated
384            in the file; or
385
386        (c) The contribution was provided directly to me by some other
387            person who certified (a), (b) or (c) and I have not modified
388            it.
389
390        (d) I understand and agree that this project and the contribution
391            are public and that a record of the contribution (including all
392            personal information I submit with it, including my sign-off) is
393            maintained indefinitely and may be redistributed consistent with
394            this project or the open source license(s) involved.
395
396then you just add a line saying::
397
398	Signed-off-by: Random J Developer <random@developer.example.org>
399
400using your real name (sorry, no pseudonyms or anonymous contributions.)
401This will be done for you automatically if you use ``git commit -s``.
402
403Some people also put extra tags at the end.  They'll just be ignored for
404now, but you can do this to mark internal company procedures or just
405point out some special detail about the sign-off.
406
407
408When to use Acked-by:, Cc:, and Co-developed-by:
409------------------------------------------------
410
411The Signed-off-by: tag indicates that the signer was involved in the
412development of the patch, or that he/she was in the patch's delivery path.
413
414If a person was not directly involved in the preparation or handling of a
415patch but wishes to signify and record their approval of it then they can
416ask to have an Acked-by: line added to the patch's changelog.
417
418Acked-by: is often used by the maintainer of the affected code when that
419maintainer neither contributed to nor forwarded the patch.
420
421Acked-by: is not as formal as Signed-off-by:.  It is a record that the acker
422has at least reviewed the patch and has indicated acceptance.  Hence patch
423mergers will sometimes manually convert an acker's "yep, looks good to me"
424into an Acked-by: (but note that it is usually better to ask for an
425explicit ack).
426
427Acked-by: does not necessarily indicate acknowledgement of the entire patch.
428For example, if a patch affects multiple subsystems and has an Acked-by: from
429one subsystem maintainer then this usually indicates acknowledgement of just
430the part which affects that maintainer's code.  Judgement should be used here.
431When in doubt people should refer to the original discussion in the mailing
432list archives.
433
434If a person has had the opportunity to comment on a patch, but has not
435provided such comments, you may optionally add a ``Cc:`` tag to the patch.
436This is the only tag which might be added without an explicit action by the
437person it names - but it should indicate that this person was copied on the
438patch.  This tag documents that potentially interested parties
439have been included in the discussion.
440
441Co-developed-by: states that the patch was co-created by multiple developers;
442it is a used to give attribution to co-authors (in addition to the author
443attributed by the From: tag) when several people work on a single patch.  Since
444Co-developed-by: denotes authorship, every Co-developed-by: must be immediately
445followed by a Signed-off-by: of the associated co-author.  Standard sign-off
446procedure applies, i.e. the ordering of Signed-off-by: tags should reflect the
447chronological history of the patch insofar as possible, regardless of whether
448the author is attributed via From: or Co-developed-by:.  Notably, the last
449Signed-off-by: must always be that of the developer submitting the patch.
450
451Note, the From: tag is optional when the From: author is also the person (and
452email) listed in the From: line of the email header.
453
454Example of a patch submitted by the From: author::
455
456	<changelog>
457
458	Co-developed-by: First Co-Author <first@coauthor.example.org>
459	Signed-off-by: First Co-Author <first@coauthor.example.org>
460	Co-developed-by: Second Co-Author <second@coauthor.example.org>
461	Signed-off-by: Second Co-Author <second@coauthor.example.org>
462	Signed-off-by: From Author <from@author.example.org>
463
464Example of a patch submitted by a Co-developed-by: author::
465
466	From: From Author <from@author.example.org>
467
468	<changelog>
469
470	Co-developed-by: Random Co-Author <random@coauthor.example.org>
471	Signed-off-by: Random Co-Author <random@coauthor.example.org>
472	Signed-off-by: From Author <from@author.example.org>
473	Co-developed-by: Submitting Co-Author <sub@coauthor.example.org>
474	Signed-off-by: Submitting Co-Author <sub@coauthor.example.org>
475
476
477Using Reported-by:, Tested-by:, Reviewed-by:, Suggested-by: and Fixes:
478----------------------------------------------------------------------
479
480The Reported-by tag gives credit to people who find bugs and report them and it
481hopefully inspires them to help us again in the future.  Please note that if
482the bug was reported in private, then ask for permission first before using the
483Reported-by tag.
484
485A Tested-by: tag indicates that the patch has been successfully tested (in
486some environment) by the person named.  This tag informs maintainers that
487some testing has been performed, provides a means to locate testers for
488future patches, and ensures credit for the testers.
489
490Reviewed-by:, instead, indicates that the patch has been reviewed and found
491acceptable according to the Reviewer's Statement:
492
493Reviewer's statement of oversight
494^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
495
496By offering my Reviewed-by: tag, I state that:
497
498	 (a) I have carried out a technical review of this patch to
499	     evaluate its appropriateness and readiness for inclusion into
500	     the mainline kernel.
501
502	 (b) Any problems, concerns, or questions relating to the patch
503	     have been communicated back to the submitter.  I am satisfied
504	     with the submitter's response to my comments.
505
506	 (c) While there may be things that could be improved with this
507	     submission, I believe that it is, at this time, (1) a
508	     worthwhile modification to the kernel, and (2) free of known
509	     issues which would argue against its inclusion.
510
511	 (d) While I have reviewed the patch and believe it to be sound, I
512	     do not (unless explicitly stated elsewhere) make any
513	     warranties or guarantees that it will achieve its stated
514	     purpose or function properly in any given situation.
515
516A Reviewed-by tag is a statement of opinion that the patch is an
517appropriate modification of the kernel without any remaining serious
518technical issues.  Any interested reviewer (who has done the work) can
519offer a Reviewed-by tag for a patch.  This tag serves to give credit to
520reviewers and to inform maintainers of the degree of review which has been
521done on the patch.  Reviewed-by: tags, when supplied by reviewers known to
522understand the subject area and to perform thorough reviews, will normally
523increase the likelihood of your patch getting into the kernel.
524
525Both Tested-by and Reviewed-by tags, once received on mailing list from tester
526or reviewer, should be added by author to the applicable patches when sending
527next versions.  However if the patch has changed substantially in following
528version, these tags might not be applicable anymore and thus should be removed.
529Usually removal of someone's Tested-by or Reviewed-by tags should be mentioned
530in the patch changelog (after the '---' separator).
531
532A Suggested-by: tag indicates that the patch idea is suggested by the person
533named and ensures credit to the person for the idea. Please note that this
534tag should not be added without the reporter's permission, especially if the
535idea was not posted in a public forum. That said, if we diligently credit our
536idea reporters, they will, hopefully, be inspired to help us again in the
537future.
538
539A Fixes: tag indicates that the patch fixes an issue in a previous commit. It
540is used to make it easy to determine where a bug originated, which can help
541review a bug fix. This tag also assists the stable kernel team in determining
542which stable kernel versions should receive your fix. This is the preferred
543method for indicating a bug fixed by the patch. See :ref:`describe_changes`
544for more details.
545
546.. _the_canonical_patch_format:
547
548The canonical patch format
549--------------------------
550
551This section describes how the patch itself should be formatted.  Note
552that, if you have your patches stored in a ``git`` repository, proper patch
553formatting can be had with ``git format-patch``.  The tools cannot create
554the necessary text, though, so read the instructions below anyway.
555
556The canonical patch subject line is::
557
558    Subject: [PATCH 001/123] subsystem: summary phrase
559
560The canonical patch message body contains the following:
561
562  - A ``from`` line specifying the patch author, followed by an empty
563    line (only needed if the person sending the patch is not the author).
564
565  - The body of the explanation, line wrapped at 75 columns, which will
566    be copied to the permanent changelog to describe this patch.
567
568  - An empty line.
569
570  - The ``Signed-off-by:`` lines, described above, which will
571    also go in the changelog.
572
573  - A marker line containing simply ``---``.
574
575  - Any additional comments not suitable for the changelog.
576
577  - The actual patch (``diff`` output).
578
579The Subject line format makes it very easy to sort the emails
580alphabetically by subject line - pretty much any email reader will
581support that - since because the sequence number is zero-padded,
582the numerical and alphabetic sort is the same.
583
584The ``subsystem`` in the email's Subject should identify which
585area or subsystem of the kernel is being patched.
586
587The ``summary phrase`` in the email's Subject should concisely
588describe the patch which that email contains.  The ``summary
589phrase`` should not be a filename.  Do not use the same ``summary
590phrase`` for every patch in a whole patch series (where a ``patch
591series`` is an ordered sequence of multiple, related patches).
592
593Bear in mind that the ``summary phrase`` of your email becomes a
594globally-unique identifier for that patch.  It propagates all the way
595into the ``git`` changelog.  The ``summary phrase`` may later be used in
596developer discussions which refer to the patch.  People will want to
597google for the ``summary phrase`` to read discussion regarding that
598patch.  It will also be the only thing that people may quickly see
599when, two or three months later, they are going through perhaps
600thousands of patches using tools such as ``gitk`` or ``git log
601--oneline``.
602
603For these reasons, the ``summary`` must be no more than 70-75
604characters, and it must describe both what the patch changes, as well
605as why the patch might be necessary.  It is challenging to be both
606succinct and descriptive, but that is what a well-written summary
607should do.
608
609The ``summary phrase`` may be prefixed by tags enclosed in square
610brackets: "Subject: [PATCH <tag>...] <summary phrase>".  The tags are
611not considered part of the summary phrase, but describe how the patch
612should be treated.  Common tags might include a version descriptor if
613the multiple versions of the patch have been sent out in response to
614comments (i.e., "v1, v2, v3"), or "RFC" to indicate a request for
615comments.  If there are four patches in a patch series the individual
616patches may be numbered like this: 1/4, 2/4, 3/4, 4/4.  This assures
617that developers understand the order in which the patches should be
618applied and that they have reviewed or applied all of the patches in
619the patch series.
620
621A couple of example Subjects::
622
623    Subject: [PATCH 2/5] ext2: improve scalability of bitmap searching
624    Subject: [PATCH v2 01/27] x86: fix eflags tracking
625
626The ``from`` line must be the very first line in the message body,
627and has the form:
628
629        From: Patch Author <author@example.com>
630
631The ``from`` line specifies who will be credited as the author of the
632patch in the permanent changelog.  If the ``from`` line is missing,
633then the ``From:`` line from the email header will be used to determine
634the patch author in the changelog.
635
636The explanation body will be committed to the permanent source
637changelog, so should make sense to a competent reader who has long
638since forgotten the immediate details of the discussion that might
639have led to this patch.  Including symptoms of the failure which the
640patch addresses (kernel log messages, oops messages, etc.) is
641especially useful for people who might be searching the commit logs
642looking for the applicable patch.  If a patch fixes a compile failure,
643it may not be necessary to include _all_ of the compile failures; just
644enough that it is likely that someone searching for the patch can find
645it.  As in the ``summary phrase``, it is important to be both succinct as
646well as descriptive.
647
648The ``---`` marker line serves the essential purpose of marking for patch
649handling tools where the changelog message ends.
650
651One good use for the additional comments after the ``---`` marker is for
652a ``diffstat``, to show what files have changed, and the number of
653inserted and deleted lines per file.  A ``diffstat`` is especially useful
654on bigger patches.  Other comments relevant only to the moment or the
655maintainer, not suitable for the permanent changelog, should also go
656here.  A good example of such comments might be ``patch changelogs``
657which describe what has changed between the v1 and v2 version of the
658patch.
659
660If you are going to include a ``diffstat`` after the ``---`` marker, please
661use ``diffstat`` options ``-p 1 -w 70`` so that filenames are listed from
662the top of the kernel source tree and don't use too much horizontal
663space (easily fit in 80 columns, maybe with some indentation).  (``git``
664generates appropriate diffstats by default.)
665
666See more details on the proper patch format in the following
667references.
668
669.. _explicit_in_reply_to:
670
671Explicit In-Reply-To headers
672----------------------------
673
674It can be helpful to manually add In-Reply-To: headers to a patch
675(e.g., when using ``git send-email``) to associate the patch with
676previous relevant discussion, e.g. to link a bug fix to the email with
677the bug report.  However, for a multi-patch series, it is generally
678best to avoid using In-Reply-To: to link to older versions of the
679series.  This way multiple versions of the patch don't become an
680unmanageable forest of references in email clients.  If a link is
681helpful, you can use the https://lkml.kernel.org/ redirector (e.g., in
682the cover email text) to link to an earlier version of the patch series.
683
684
685Providing base tree information
686-------------------------------
687
688When other developers receive your patches and start the review process,
689it is often useful for them to know where in the tree history they
690should place your work. This is particularly useful for automated CI
691processes that attempt to run a series of tests in order to establish
692the quality of your submission before the maintainer starts the review.
693
694If you are using ``git format-patch`` to generate your patches, you can
695automatically include the base tree information in your submission by
696using the ``--base`` flag. The easiest and most convenient way to use
697this option is with topical branches::
698
699    $ git checkout -t -b my-topical-branch master
700    Branch 'my-topical-branch' set up to track local branch 'master'.
701    Switched to a new branch 'my-topical-branch'
702
703    [perform your edits and commits]
704
705    $ git format-patch --base=auto --cover-letter -o outgoing/ master
706    outgoing/0000-cover-letter.patch
707    outgoing/0001-First-Commit.patch
708    outgoing/...
709
710When you open ``outgoing/0000-cover-letter.patch`` for editing, you will
711notice that it will have the ``base-commit:`` trailer at the very
712bottom, which provides the reviewer and the CI tools enough information
713to properly perform ``git am`` without worrying about conflicts::
714
715    $ git checkout -b patch-review [base-commit-id]
716    Switched to a new branch 'patch-review'
717    $ git am patches.mbox
718    Applying: First Commit
719    Applying: ...
720
721Please see ``man git-format-patch`` for more information about this
722option.
723
724.. note::
725
726    The ``--base`` feature was introduced in git version 2.9.0.
727
728If you are not using git to format your patches, you can still include
729the same ``base-commit`` trailer to indicate the commit hash of the tree
730on which your work is based. You should add it either in the cover
731letter or in the first patch of the series and it should be placed
732either below the ``---`` line or at the very bottom of all other
733content, right before your email signature.
734
735
736References
737----------
738
739Andrew Morton, "The perfect patch" (tpp).
740  <https://www.ozlabs.org/~akpm/stuff/tpp.txt>
741
742Jeff Garzik, "Linux kernel patch submission format".
743  <https://web.archive.org/web/20180829112450/http://linux.yyz.us/patch-format.html>
744
745Greg Kroah-Hartman, "How to piss off a kernel subsystem maintainer".
746  <http://www.kroah.com/log/linux/maintainer.html>
747
748  <http://www.kroah.com/log/linux/maintainer-02.html>
749
750  <http://www.kroah.com/log/linux/maintainer-03.html>
751
752  <http://www.kroah.com/log/linux/maintainer-04.html>
753
754  <http://www.kroah.com/log/linux/maintainer-05.html>
755
756  <http://www.kroah.com/log/linux/maintainer-06.html>
757
758NO!!!! No more huge patch bombs to linux-kernel@vger.kernel.org people!
759  <https://lkml.org/lkml/2005/7/11/336>
760
761Kernel Documentation/process/coding-style.rst:
762  :ref:`Documentation/process/coding-style.rst <codingstyle>`
763
764Linus Torvalds's mail on the canonical patch format:
765  <http://lkml.org/lkml/2005/4/7/183>
766
767Andi Kleen, "On submitting kernel patches"
768  Some strategies to get difficult or controversial changes in.
769
770  http://halobates.de/on-submitting-patches.pdf
771