884 lines · plain
1.. _submittingpatches:2 3Submitting patches: the essential guide to getting your code into the kernel4============================================================================5 6For a person or company who wishes to submit a change to the Linux7kernel, the process can sometimes be daunting if you're not familiar8with "the system." This text is a collection of suggestions which9can greatly increase the chances of your change being accepted.10 11This document contains a large number of suggestions in a relatively terse12format. For detailed information on how the kernel development process13works, see Documentation/process/development-process.rst. Also, read14Documentation/process/submit-checklist.rst15for a list of items to check before submitting code.16For device tree binding patches, read17Documentation/devicetree/bindings/submitting-patches.rst.18 19This documentation assumes that you're using ``git`` to prepare your patches.20If you're unfamiliar with ``git``, you would be well-advised to learn how to21use it, it will make your life as a kernel developer and in general much22easier.23 24Some subsystems and maintainer trees have additional information about25their workflow and expectations, see26:ref:`Documentation/process/maintainer-handbooks.rst <maintainer_handbooks_main>`.27 28Obtain a current source tree29----------------------------30 31If you do not have a repository with the current kernel source handy, use32``git`` to obtain one. You'll want to start with the mainline repository,33which can be grabbed with::34 35 git clone git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git36 37Note, however, that you may not want to develop against the mainline tree38directly. Most subsystem maintainers run their own trees and want to see39patches prepared against those trees. See the **T:** entry for the subsystem40in the MAINTAINERS file to find that tree, or simply ask the maintainer if41the tree is not listed there.42 43.. _describe_changes:44 45Describe your changes46---------------------47 48Describe your problem. Whether your patch is a one-line bug fix or495000 lines of a new feature, there must be an underlying problem that50motivated you to do this work. Convince the reviewer that there is a51problem worth fixing and that it makes sense for them to read past the52first paragraph.53 54Describe user-visible impact. Straight up crashes and lockups are55pretty convincing, but not all bugs are that blatant. Even if the56problem was spotted during code review, describe the impact you think57it can have on users. Keep in mind that the majority of Linux58installations run kernels from secondary stable trees or59vendor/product-specific trees that cherry-pick only specific patches60from upstream, so include anything that could help route your change61downstream: provoking circumstances, excerpts from dmesg, crash62descriptions, performance regressions, latency spikes, lockups, etc.63 64Quantify optimizations and trade-offs. If you claim improvements in65performance, memory consumption, stack footprint, or binary size,66include numbers that back them up. But also describe non-obvious67costs. Optimizations usually aren't free but trade-offs between CPU,68memory, and readability; or, when it comes to heuristics, between69different workloads. Describe the expected downsides of your70optimization so that the reviewer can weigh costs against benefits.71 72Once the problem is established, describe what you are actually doing73about it in technical detail. It's important to describe the change74in plain English for the reviewer to verify that the code is behaving75as you intend it to.76 77The maintainer will thank you if you write your patch description in a78form which can be easily pulled into Linux's source code management79system, ``git``, as a "commit log". See :ref:`the_canonical_patch_format`.80 81Solve only one problem per patch. If your description starts to get82long, that's a sign that you probably need to split up your patch.83See :ref:`split_changes`.84 85When you submit or resubmit a patch or patch series, include the86complete patch description and justification for it. Don't just87say that this is version N of the patch (series). Don't expect the88subsystem maintainer to refer back to earlier patch versions or referenced89URLs to find the patch description and put that into the patch.90I.e., the patch (series) and its description should be self-contained.91This benefits both the maintainers and reviewers. Some reviewers92probably didn't even receive earlier versions of the patch.93 94Describe your changes in imperative mood, e.g. "make xyzzy do frotz"95instead of "[This patch] makes xyzzy do frotz" or "[I] changed xyzzy96to do frotz", as if you are giving orders to the codebase to change97its behaviour.98 99If you want to refer to a specific commit, don't just refer to the100SHA-1 ID of the commit. Please also include the oneline summary of101the commit, to make it easier for reviewers to know what it is about.102Example::103 104 Commit e21d2170f36602ae2708 ("video: remove unnecessary105 platform_set_drvdata()") removed the unnecessary106 platform_set_drvdata(), but left the variable "dev" unused,107 delete it.108 109You should also be sure to use at least the first twelve characters of the110SHA-1 ID. The kernel repository holds a *lot* of objects, making111collisions with shorter IDs a real possibility. Bear in mind that, even if112there is no collision with your six-character ID now, that condition may113change five years from now.114 115If related discussions or any other background information behind the change116can be found on the web, add 'Link:' tags pointing to it. If the patch is a117result of some earlier mailing list discussions or something documented on the118web, point to it.119 120When linking to mailing list archives, preferably use the lore.kernel.org121message archiver service. To create the link URL, use the contents of the122``Message-ID`` header of the message without the surrounding angle brackets.123For example::124 125 Link: https://lore.kernel.org/30th.anniversary.repost@klaava.Helsinki.FI126 127Please check the link to make sure that it is actually working and points128to the relevant message.129 130However, try to make your explanation understandable without external131resources. In addition to giving a URL to a mailing list archive or bug,132summarize the relevant points of the discussion that led to the133patch as submitted.134 135In case your patch fixes a bug, use the 'Closes:' tag with a URL referencing136the report in the mailing list archives or a public bug tracker. For example::137 138 Closes: https://example.com/issues/1234139 140Some bug trackers have the ability to close issues automatically when a141commit with such a tag is applied. Some bots monitoring mailing lists can142also track such tags and take certain actions. Private bug trackers and143invalid URLs are forbidden.144 145If your patch fixes a bug in a specific commit, e.g. you found an issue using146``git bisect``, please use the 'Fixes:' tag with the first 12 characters of147the SHA-1 ID, and the one line summary. Do not split the tag across multiple148lines, tags are exempt from the "wrap at 75 columns" rule in order to simplify149parsing scripts. For example::150 151 Fixes: 54a4f0239f2e ("KVM: MMU: make kvm_mmu_zap_page() return the number of pages it actually freed")152 153The following ``git config`` settings can be used to add a pretty format for154outputting the above style in the ``git log`` or ``git show`` commands::155 156 [core]157 abbrev = 12158 [pretty]159 fixes = Fixes: %h (\"%s\")160 161An example call::162 163 $ git log -1 --pretty=fixes 54a4f0239f2e164 Fixes: 54a4f0239f2e ("KVM: MMU: make kvm_mmu_zap_page() return the number of pages it actually freed")165 166.. _split_changes:167 168Separate your changes169---------------------170 171Separate each **logical change** into a separate patch.172 173For example, if your changes include both bug fixes and performance174enhancements for a single driver, separate those changes into two175or more patches. If your changes include an API update, and a new176driver which uses that new API, separate those into two patches.177 178On the other hand, if you make a single change to numerous files,179group those changes into a single patch. Thus a single logical change180is contained within a single patch.181 182The point to remember is that each patch should make an easily understood183change that can be verified by reviewers. Each patch should be justifiable184on its own merits.185 186If one patch depends on another patch in order for a change to be187complete, that is OK. Simply note **"this patch depends on patch X"**188in your patch description.189 190When dividing your change into a series of patches, take special care to191ensure that the kernel builds and runs properly after each patch in the192series. Developers using ``git bisect`` to track down a problem can end up193splitting your patch series at any point; they will not thank you if you194introduce bugs in the middle.195 196If you cannot condense your patch set into a smaller set of patches,197then only post say 15 or so at a time and wait for review and integration.198 199 200 201Style-check your changes202------------------------203 204Check your patch for basic style violations, details of which can be205found in Documentation/process/coding-style.rst.206Failure to do so simply wastes207the reviewers time and will get your patch rejected, probably208without even being read.209 210One significant exception is when moving code from one file to211another -- in this case you should not modify the moved code at all in212the same patch which moves it. This clearly delineates the act of213moving the code and your changes. This greatly aids review of the214actual differences and allows tools to better track the history of215the code itself.216 217Check your patches with the patch style checker prior to submission218(scripts/checkpatch.pl). Note, though, that the style checker should be219viewed as a guide, not as a replacement for human judgment. If your code220looks better with a violation then its probably best left alone.221 222The checker reports at three levels:223 - ERROR: things that are very likely to be wrong224 - WARNING: things requiring careful review225 - CHECK: things requiring thought226 227You should be able to justify all violations that remain in your228patch.229 230 231Select the recipients for your patch232------------------------------------233 234You should always copy the appropriate subsystem maintainer(s) and list(s) on235any patch to code that they maintain; look through the MAINTAINERS file and the236source code revision history to see who those maintainers are. The script237scripts/get_maintainer.pl can be very useful at this step (pass paths to your238patches as arguments to scripts/get_maintainer.pl). If you cannot find a239maintainer for the subsystem you are working on, Andrew Morton240(akpm@linux-foundation.org) serves as a maintainer of last resort.241 242linux-kernel@vger.kernel.org should be used by default for all patches, but the243volume on that list has caused a number of developers to tune it out. Please244do not spam unrelated lists and unrelated people, though.245 246Many kernel-related lists are hosted at kernel.org; you can find a list247of them at https://subspace.kernel.org. There are kernel-related lists248hosted elsewhere as well, though.249 250Linus Torvalds is the final arbiter of all changes accepted into the251Linux kernel. His e-mail address is <torvalds@linux-foundation.org>.252He gets a lot of e-mail, and, at this point, very few patches go through253Linus directly, so typically you should do your best to -avoid-254sending him e-mail.255 256If you have a patch that fixes an exploitable security bug, send that patch257to security@kernel.org. For severe bugs, a short embargo may be considered258to allow distributors to get the patch out to users; in such cases,259obviously, the patch should not be sent to any public lists. See also260Documentation/process/security-bugs.rst.261 262Patches that fix a severe bug in a released kernel should be directed263toward the stable maintainers by putting a line like this::264 265 Cc: stable@vger.kernel.org266 267into the sign-off area of your patch (note, NOT an email recipient). You268should also read Documentation/process/stable-kernel-rules.rst269in addition to this document.270 271If changes affect userland-kernel interfaces, please send the MAN-PAGES272maintainer (as listed in the MAINTAINERS file) a man-pages patch, or at273least a notification of the change, so that some information makes its way274into the manual pages. User-space API changes should also be copied to275linux-api@vger.kernel.org.276 277 278No MIME, no links, no compression, no attachments. Just plain text279-------------------------------------------------------------------280 281Linus and other kernel developers need to be able to read and comment282on the changes you are submitting. It is important for a kernel283developer to be able to "quote" your changes, using standard e-mail284tools, so that they may comment on specific portions of your code.285 286For this reason, all patches should be submitted by e-mail "inline". The287easiest way to do this is with ``git send-email``, which is strongly288recommended. An interactive tutorial for ``git send-email`` is available at289https://git-send-email.io.290 291If you choose not to use ``git send-email``:292 293.. warning::294 295 Be wary of your editor's word-wrap corrupting your patch,296 if you choose to cut-n-paste your patch.297 298Do not attach the patch as a MIME attachment, compressed or not.299Many popular e-mail applications will not always transmit a MIME300attachment as plain text, making it impossible to comment on your301code. A MIME attachment also takes Linus a bit more time to process,302decreasing the likelihood of your MIME-attached change being accepted.303 304Exception: If your mailer is mangling patches then someone may ask305you to re-send them using MIME.306 307See Documentation/process/email-clients.rst for hints about configuring308your e-mail client so that it sends your patches untouched.309 310Respond to review comments311--------------------------312 313Your patch will almost certainly get comments from reviewers on ways in314which the patch can be improved, in the form of a reply to your email. You must315respond to those comments; ignoring reviewers is a good way to get ignored in316return. You can simply reply to their emails to answer their comments. Review317comments or questions that do not lead to a code change should almost certainly318bring about a comment or changelog entry so that the next reviewer better319understands what is going on.320 321Be sure to tell the reviewers what changes you are making and to thank them322for their time. Code review is a tiring and time-consuming process, and323reviewers sometimes get grumpy. Even in that case, though, respond324politely and address the problems they have pointed out. When sending a next325version, add a ``patch changelog`` to the cover letter or to individual patches326explaining difference against previous submission (see327:ref:`the_canonical_patch_format`).328Notify people that commented on your patch about new versions by adding them to329the patches CC list.330 331See Documentation/process/email-clients.rst for recommendations on email332clients and mailing list etiquette.333 334.. _interleaved_replies:335 336Use trimmed interleaved replies in email discussions337----------------------------------------------------338Top-posting is strongly discouraged in Linux kernel development339discussions. Interleaved (or "inline") replies make conversations much340easier to follow. For more details see:341https://en.wikipedia.org/wiki/Posting_style#Interleaved_style342 343As is frequently quoted on the mailing list::344 345 A: http://en.wikipedia.org/wiki/Top_post346 Q: Were do I find info about this thing called top-posting?347 A: Because it messes up the order in which people normally read text.348 Q: Why is top-posting such a bad thing?349 A: Top-posting.350 Q: What is the most annoying thing in e-mail?351 352Similarly, please trim all unneeded quotations that aren't relevant353to your reply. This makes responses easier to find, and saves time and354space. For more details see: http://daringfireball.net/2007/07/on_top ::355 356 A: No.357 Q: Should I include quotations after my reply?358 359.. _resend_reminders:360 361Don't get discouraged - or impatient362------------------------------------363 364After you have submitted your change, be patient and wait. Reviewers are365busy people and may not get to your patch right away.366 367Once upon a time, patches used to disappear into the void without comment,368but the development process works more smoothly than that now. You should369receive comments within a few weeks (typically 2-3); if that does not370happen, make sure that you have sent your patches to the right place.371Wait for a minimum of one week before resubmitting or pinging reviewers372- possibly longer during busy times like merge windows.373 374It's also ok to resend the patch or the patch series after a couple of375weeks with the word "RESEND" added to the subject line::376 377 [PATCH Vx RESEND] sub/sys: Condensed patch summary378 379Don't add "RESEND" when you are submitting a modified version of your380patch or patch series - "RESEND" only applies to resubmission of a381patch or patch series which have not been modified in any way from the382previous submission.383 384 385Include PATCH in the subject386-----------------------------387 388Due to high e-mail traffic to Linus, and to linux-kernel, it is common389convention to prefix your subject line with [PATCH]. This lets Linus390and other kernel developers more easily distinguish patches from other391e-mail discussions.392 393``git send-email`` will do this for you automatically.394 395 396Sign your work - the Developer's Certificate of Origin397------------------------------------------------------398 399To improve tracking of who did what, especially with patches that can400percolate to their final resting place in the kernel through several401layers of maintainers, we've introduced a "sign-off" procedure on402patches that are being emailed around.403 404The sign-off is a simple line at the end of the explanation for the405patch, which certifies that you wrote it or otherwise have the right to406pass it on as an open-source patch. The rules are pretty simple: if you407can certify the below:408 409Developer's Certificate of Origin 1.1410^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^411 412By making a contribution to this project, I certify that:413 414 (a) The contribution was created in whole or in part by me and I415 have the right to submit it under the open source license416 indicated in the file; or417 418 (b) The contribution is based upon previous work that, to the best419 of my knowledge, is covered under an appropriate open source420 license and I have the right under that license to submit that421 work with modifications, whether created in whole or in part422 by me, under the same open source license (unless I am423 permitted to submit under a different license), as indicated424 in the file; or425 426 (c) The contribution was provided directly to me by some other427 person who certified (a), (b) or (c) and I have not modified428 it.429 430 (d) I understand and agree that this project and the contribution431 are public and that a record of the contribution (including all432 personal information I submit with it, including my sign-off) is433 maintained indefinitely and may be redistributed consistent with434 this project or the open source license(s) involved.435 436then you just add a line saying::437 438 Signed-off-by: Random J Developer <random@developer.example.org>439 440using a known identity (sorry, no anonymous contributions.)441This will be done for you automatically if you use ``git commit -s``.442Reverts should also include "Signed-off-by". ``git revert -s`` does that443for you.444 445Some people also put extra tags at the end. They'll just be ignored for446now, but you can do this to mark internal company procedures or just447point out some special detail about the sign-off.448 449Any further SoBs (Signed-off-by:'s) following the author's SoB are from450people handling and transporting the patch, but were not involved in its451development. SoB chains should reflect the **real** route a patch took452as it was propagated to the maintainers and ultimately to Linus, with453the first SoB entry signalling primary authorship of a single author.454 455 456When to use Acked-by:, Cc:, and Co-developed-by:457------------------------------------------------458 459The Signed-off-by: tag indicates that the signer was involved in the460development of the patch, or that he/she was in the patch's delivery path.461 462If a person was not directly involved in the preparation or handling of a463patch but wishes to signify and record their approval of it then they can464ask to have an Acked-by: line added to the patch's changelog.465 466Acked-by: is often used by the maintainer of the affected code when that467maintainer neither contributed to nor forwarded the patch.468 469Acked-by: is not as formal as Signed-off-by:. It is a record that the acker470has at least reviewed the patch and has indicated acceptance. Hence patch471mergers will sometimes manually convert an acker's "yep, looks good to me"472into an Acked-by: (but note that it is usually better to ask for an473explicit ack).474 475Acked-by: does not necessarily indicate acknowledgement of the entire patch.476For example, if a patch affects multiple subsystems and has an Acked-by: from477one subsystem maintainer then this usually indicates acknowledgement of just478the part which affects that maintainer's code. Judgement should be used here.479When in doubt people should refer to the original discussion in the mailing480list archives.481 482If a person has had the opportunity to comment on a patch, but has not483provided such comments, you may optionally add a ``Cc:`` tag to the patch.484This is the only tag which might be added without an explicit action by the485person it names - but it should indicate that this person was copied on the486patch. This tag documents that potentially interested parties487have been included in the discussion.488 489Co-developed-by: states that the patch was co-created by multiple developers;490it is used to give attribution to co-authors (in addition to the author491attributed by the From: tag) when several people work on a single patch. Since492Co-developed-by: denotes authorship, every Co-developed-by: must be immediately493followed by a Signed-off-by: of the associated co-author. Standard sign-off494procedure applies, i.e. the ordering of Signed-off-by: tags should reflect the495chronological history of the patch insofar as possible, regardless of whether496the author is attributed via From: or Co-developed-by:. Notably, the last497Signed-off-by: must always be that of the developer submitting the patch.498 499Note, the From: tag is optional when the From: author is also the person (and500email) listed in the From: line of the email header.501 502Example of a patch submitted by the From: author::503 504 <changelog>505 506 Co-developed-by: First Co-Author <first@coauthor.example.org>507 Signed-off-by: First Co-Author <first@coauthor.example.org>508 Co-developed-by: Second Co-Author <second@coauthor.example.org>509 Signed-off-by: Second Co-Author <second@coauthor.example.org>510 Signed-off-by: From Author <from@author.example.org>511 512Example of a patch submitted by a Co-developed-by: author::513 514 From: From Author <from@author.example.org>515 516 <changelog>517 518 Co-developed-by: Random Co-Author <random@coauthor.example.org>519 Signed-off-by: Random Co-Author <random@coauthor.example.org>520 Signed-off-by: From Author <from@author.example.org>521 Co-developed-by: Submitting Co-Author <sub@coauthor.example.org>522 Signed-off-by: Submitting Co-Author <sub@coauthor.example.org>523 524 525Using Reported-by:, Tested-by:, Reviewed-by:, Suggested-by: and Fixes:526----------------------------------------------------------------------527 528The Reported-by tag gives credit to people who find bugs and report them and it529hopefully inspires them to help us again in the future. The tag is intended for530bugs; please do not use it to credit feature requests. The tag should be531followed by a Closes: tag pointing to the report, unless the report is not532available on the web. The Link: tag can be used instead of Closes: if the patch533fixes a part of the issue(s) being reported. Please note that if the bug was534reported in private, then ask for permission first before using the Reported-by535tag.536 537A Tested-by: tag indicates that the patch has been successfully tested (in538some environment) by the person named. This tag informs maintainers that539some testing has been performed, provides a means to locate testers for540future patches, and ensures credit for the testers.541 542Reviewed-by:, instead, indicates that the patch has been reviewed and found543acceptable according to the Reviewer's Statement:544 545Reviewer's statement of oversight546^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^547 548By offering my Reviewed-by: tag, I state that:549 550 (a) I have carried out a technical review of this patch to551 evaluate its appropriateness and readiness for inclusion into552 the mainline kernel.553 554 (b) Any problems, concerns, or questions relating to the patch555 have been communicated back to the submitter. I am satisfied556 with the submitter's response to my comments.557 558 (c) While there may be things that could be improved with this559 submission, I believe that it is, at this time, (1) a560 worthwhile modification to the kernel, and (2) free of known561 issues which would argue against its inclusion.562 563 (d) While I have reviewed the patch and believe it to be sound, I564 do not (unless explicitly stated elsewhere) make any565 warranties or guarantees that it will achieve its stated566 purpose or function properly in any given situation.567 568A Reviewed-by tag is a statement of opinion that the patch is an569appropriate modification of the kernel without any remaining serious570technical issues. Any interested reviewer (who has done the work) can571offer a Reviewed-by tag for a patch. This tag serves to give credit to572reviewers and to inform maintainers of the degree of review which has been573done on the patch. Reviewed-by: tags, when supplied by reviewers known to574understand the subject area and to perform thorough reviews, will normally575increase the likelihood of your patch getting into the kernel.576 577Both Tested-by and Reviewed-by tags, once received on mailing list from tester578or reviewer, should be added by author to the applicable patches when sending579next versions. However if the patch has changed substantially in following580version, these tags might not be applicable anymore and thus should be removed.581Usually removal of someone's Tested-by or Reviewed-by tags should be mentioned582in the patch changelog (after the '---' separator).583 584A Suggested-by: tag indicates that the patch idea is suggested by the person585named and ensures credit to the person for the idea. Please note that this586tag should not be added without the reporter's permission, especially if the587idea was not posted in a public forum. That said, if we diligently credit our588idea reporters, they will, hopefully, be inspired to help us again in the589future.590 591A Fixes: tag indicates that the patch fixes an issue in a previous commit. It592is used to make it easy to determine where a bug originated, which can help593review a bug fix. This tag also assists the stable kernel team in determining594which stable kernel versions should receive your fix. This is the preferred595method for indicating a bug fixed by the patch. See :ref:`describe_changes`596for more details.597 598Note: Attaching a Fixes: tag does not subvert the stable kernel rules599process nor the requirement to Cc: stable@vger.kernel.org on all stable600patch candidates. For more information, please read601Documentation/process/stable-kernel-rules.rst.602 603.. _the_canonical_patch_format:604 605The canonical patch format606--------------------------607 608This section describes how the patch itself should be formatted. Note609that, if you have your patches stored in a ``git`` repository, proper patch610formatting can be had with ``git format-patch``. The tools cannot create611the necessary text, though, so read the instructions below anyway.612 613The canonical patch subject line is::614 615 Subject: [PATCH 001/123] subsystem: summary phrase616 617The canonical patch message body contains the following:618 619 - A ``from`` line specifying the patch author, followed by an empty620 line (only needed if the person sending the patch is not the author).621 622 - The body of the explanation, line wrapped at 75 columns, which will623 be copied to the permanent changelog to describe this patch.624 625 - An empty line.626 627 - The ``Signed-off-by:`` lines, described above, which will628 also go in the changelog.629 630 - A marker line containing simply ``---``.631 632 - Any additional comments not suitable for the changelog.633 634 - The actual patch (``diff`` output).635 636The Subject line format makes it very easy to sort the emails637alphabetically by subject line - pretty much any email reader will638support that - since because the sequence number is zero-padded,639the numerical and alphabetic sort is the same.640 641The ``subsystem`` in the email's Subject should identify which642area or subsystem of the kernel is being patched.643 644The ``summary phrase`` in the email's Subject should concisely645describe the patch which that email contains. The ``summary646phrase`` should not be a filename. Do not use the same ``summary647phrase`` for every patch in a whole patch series (where a ``patch648series`` is an ordered sequence of multiple, related patches).649 650Bear in mind that the ``summary phrase`` of your email becomes a651globally-unique identifier for that patch. It propagates all the way652into the ``git`` changelog. The ``summary phrase`` may later be used in653developer discussions which refer to the patch. People will want to654google for the ``summary phrase`` to read discussion regarding that655patch. It will also be the only thing that people may quickly see656when, two or three months later, they are going through perhaps657thousands of patches using tools such as ``gitk`` or ``git log658--oneline``.659 660For these reasons, the ``summary`` must be no more than 70-75661characters, and it must describe both what the patch changes, as well662as why the patch might be necessary. It is challenging to be both663succinct and descriptive, but that is what a well-written summary664should do.665 666The ``summary phrase`` may be prefixed by tags enclosed in square667brackets: "Subject: [PATCH <tag>...] <summary phrase>". The tags are668not considered part of the summary phrase, but describe how the patch669should be treated. Common tags might include a version descriptor if670the multiple versions of the patch have been sent out in response to671comments (i.e., "v1, v2, v3"), or "RFC" to indicate a request for672comments.673 674If there are four patches in a patch series the individual patches may675be numbered like this: 1/4, 2/4, 3/4, 4/4. This assures that developers676understand the order in which the patches should be applied and that677they have reviewed or applied all of the patches in the patch series.678 679Here are some good example Subjects::680 681 Subject: [PATCH 2/5] ext2: improve scalability of bitmap searching682 Subject: [PATCH v2 01/27] x86: fix eflags tracking683 Subject: [PATCH v2] sub/sys: Condensed patch summary684 Subject: [PATCH v2 M/N] sub/sys: Condensed patch summary685 686The ``from`` line must be the very first line in the message body,687and has the form:688 689 From: Patch Author <author@example.com>690 691The ``from`` line specifies who will be credited as the author of the692patch in the permanent changelog. If the ``from`` line is missing,693then the ``From:`` line from the email header will be used to determine694the patch author in the changelog.695 696The explanation body will be committed to the permanent source697changelog, so should make sense to a competent reader who has long since698forgotten the immediate details of the discussion that might have led to699this patch. Including symptoms of the failure which the patch addresses700(kernel log messages, oops messages, etc.) are especially useful for701people who might be searching the commit logs looking for the applicable702patch. The text should be written in such detail so that when read703weeks, months or even years later, it can give the reader the needed704details to grasp the reasoning for **why** the patch was created.705 706If a patch fixes a compile failure, it may not be necessary to include707_all_ of the compile failures; just enough that it is likely that708someone searching for the patch can find it. As in the ``summary709phrase``, it is important to be both succinct as well as descriptive.710 711The ``---`` marker line serves the essential purpose of marking for712patch handling tools where the changelog message ends.713 714One good use for the additional comments after the ``---`` marker is715for a ``diffstat``, to show what files have changed, and the number of716inserted and deleted lines per file. A ``diffstat`` is especially useful717on bigger patches. If you are going to include a ``diffstat`` after the718``---`` marker, please use ``diffstat`` options ``-p 1 -w 70`` so that719filenames are listed from the top of the kernel source tree and don't720use too much horizontal space (easily fit in 80 columns, maybe with some721indentation). (``git`` generates appropriate diffstats by default.)722 723Other comments relevant only to the moment or the maintainer, not724suitable for the permanent changelog, should also go here. A good725example of such comments might be ``patch changelogs`` which describe726what has changed between the v1 and v2 version of the patch.727 728Please put this information **after** the ``---`` line which separates729the changelog from the rest of the patch. The version information is730not part of the changelog which gets committed to the git tree. It is731additional information for the reviewers. If it's placed above the732commit tags, it needs manual interaction to remove it. If it is below733the separator line, it gets automatically stripped off when applying the734patch::735 736 <commit message>737 ...738 Signed-off-by: Author <author@mail>739 ---740 V2 -> V3: Removed redundant helper function741 V1 -> V2: Cleaned up coding style and addressed review comments742 743 path/to/file | 5+++--744 ...745 746See more details on the proper patch format in the following747references.748 749.. _backtraces:750 751Backtraces in commit messages752^^^^^^^^^^^^^^^^^^^^^^^^^^^^^753 754Backtraces help document the call chain leading to a problem. However,755not all backtraces are helpful. For example, early boot call chains are756unique and obvious. Copying the full dmesg output verbatim, however,757adds distracting information like timestamps, module lists, register and758stack dumps.759 760Therefore, the most useful backtraces should distill the relevant761information from the dump, which makes it easier to focus on the real762issue. Here is an example of a well-trimmed backtrace::763 764 unchecked MSR access error: WRMSR to 0xd51 (tried to write 0x0000000000000064)765 at rIP: 0xffffffffae059994 (native_write_msr+0x4/0x20)766 Call Trace:767 mba_wrmsr768 update_domains769 rdtgroup_mkdir770 771.. _explicit_in_reply_to:772 773Explicit In-Reply-To headers774----------------------------775 776It can be helpful to manually add In-Reply-To: headers to a patch777(e.g., when using ``git send-email``) to associate the patch with778previous relevant discussion, e.g. to link a bug fix to the email with779the bug report. However, for a multi-patch series, it is generally780best to avoid using In-Reply-To: to link to older versions of the781series. This way multiple versions of the patch don't become an782unmanageable forest of references in email clients. If a link is783helpful, you can use the https://lore.kernel.org/ redirector (e.g., in784the cover email text) to link to an earlier version of the patch series.785 786 787Providing base tree information788-------------------------------789 790When other developers receive your patches and start the review process,791it is absolutely necessary for them to know what is the base792commit/branch your work applies on, considering the sheer amount of793maintainer trees present nowadays. Note again the **T:** entry in the794MAINTAINERS file explained above.795 796This is even more important for automated CI processes that attempt to797run a series of tests in order to establish the quality of your798submission before the maintainer starts the review.799 800If you are using ``git format-patch`` to generate your patches, you can801automatically include the base tree information in your submission by802using the ``--base`` flag. The easiest and most convenient way to use803this option is with topical branches::804 805 $ git checkout -t -b my-topical-branch master806 Branch 'my-topical-branch' set up to track local branch 'master'.807 Switched to a new branch 'my-topical-branch'808 809 [perform your edits and commits]810 811 $ git format-patch --base=auto --cover-letter -o outgoing/ master812 outgoing/0000-cover-letter.patch813 outgoing/0001-First-Commit.patch814 outgoing/...815 816When you open ``outgoing/0000-cover-letter.patch`` for editing, you will817notice that it will have the ``base-commit:`` trailer at the very818bottom, which provides the reviewer and the CI tools enough information819to properly perform ``git am`` without worrying about conflicts::820 821 $ git checkout -b patch-review [base-commit-id]822 Switched to a new branch 'patch-review'823 $ git am patches.mbox824 Applying: First Commit825 Applying: ...826 827Please see ``man git-format-patch`` for more information about this828option.829 830.. note::831 832 The ``--base`` feature was introduced in git version 2.9.0.833 834If you are not using git to format your patches, you can still include835the same ``base-commit`` trailer to indicate the commit hash of the tree836on which your work is based. You should add it either in the cover837letter or in the first patch of the series and it should be placed838either below the ``---`` line or at the very bottom of all other839content, right before your email signature.840 841Make sure that base commit is in an official maintainer/mainline tree842and not in some internal, accessible only to you tree - otherwise it843would be worthless.844 845Tooling846-------847 848Many of the technical aspects of this process can be automated using849b4, documented at <https://b4.docs.kernel.org/en/latest/>. This can850help with things like tracking dependencies, running checkpatch and851with formatting and sending mails.852 853References854----------855 856Andrew Morton, "The perfect patch" (tpp).857 <https://www.ozlabs.org/~akpm/stuff/tpp.txt>858 859Jeff Garzik, "Linux kernel patch submission format".860 <https://web.archive.org/web/20180829112450/http://linux.yyz.us/patch-format.html>861 862Greg Kroah-Hartman, "How to piss off a kernel subsystem maintainer".863 <http://www.kroah.com/log/linux/maintainer.html>864 865 <http://www.kroah.com/log/linux/maintainer-02.html>866 867 <http://www.kroah.com/log/linux/maintainer-03.html>868 869 <http://www.kroah.com/log/linux/maintainer-04.html>870 871 <http://www.kroah.com/log/linux/maintainer-05.html>872 873 <http://www.kroah.com/log/linux/maintainer-06.html>874 875Kernel Documentation/process/coding-style.rst876 877Linus Torvalds's mail on the canonical patch format:878 <https://lore.kernel.org/r/Pine.LNX.4.58.0504071023190.28951@ppc970.osdl.org>879 880Andi Kleen, "On submitting kernel patches"881 Some strategies to get difficult or controversial changes in.882 883 http://halobates.de/on-submitting-patches.pdf884