blob: 36f1dedc944ccf846afe5c58b8b3918c94a13977 [file] [log] [blame]
Mauro Carvalho Chehab609d99a2016-09-19 08:07:56 -03001.. _submittingpatches:
Linus Torvalds1da177e2005-04-16 15:20:36 -07002
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -03003How to Get Your Change Into the Linux Kernel or Care And Operation Of Your Linus Torvalds
4=========================================================================================
Linus Torvalds1da177e2005-04-16 15:20:36 -07005
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
Jonathan Corbetd00c4552014-12-23 08:54:36 -070011This document contains a large number of suggestions in a relatively terse
12format. For detailed information on how the kernel development process
Mauro Carvalho Chehabdca22a62016-09-21 08:51:05 -030013works, see :ref:`Documentation/development-process <development_process_main>`.
14Also, read :ref:`Documentation/SubmitChecklist <submitchecklist>`
15for a list of items to check before
Jonathan Corbetd00c4552014-12-23 08:54:36 -070016submitting code. If you are submitting a driver, also read
Mauro Carvalho Chehabdca22a62016-09-21 08:51:05 -030017:ref:`Documentation/SubmittingDrivers <submittingdrivers>`;
18for device tree binding patches, read
Jonathan Corbet082bd1c2014-12-23 09:27:04 -070019Documentation/devicetree/bindings/submitting-patches.txt.
Linus Torvalds1da177e2005-04-16 15:20:36 -070020
Mauro Carvalho Chehab9b2c7672016-09-19 08:07:54 -030021Many of these steps describe the default behavior of the ``git`` version
22control system; if you use ``git`` to prepare your patches, you'll find much
Josh Triplett8e3072a2014-04-03 14:48:30 -070023of the mechanical work done for you, though you'll still need to prepare
Mauro Carvalho Chehab9b2c7672016-09-19 08:07:54 -030024and document a sensible set of patches. In general, use of ``git`` will make
Jonathan Corbetd00c4552014-12-23 08:54:36 -070025your life as a kernel developer easier.
Linus Torvalds1da177e2005-04-16 15:20:36 -070026
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -030027Creating and Sending your Change
28********************************
Linus Torvalds1da177e2005-04-16 15:20:36 -070029
30
Jonathan Corbet7994cc12014-12-23 08:43:41 -0700310) Obtain a current source tree
32-------------------------------
33
34If you do not have a repository with the current kernel source handy, use
Mauro Carvalho Chehab9b2c7672016-09-19 08:07:54 -030035``git`` to obtain one. You'll want to start with the mainline repository,
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -030036which can be grabbed with::
Jonathan Corbet7994cc12014-12-23 08:43:41 -070037
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -030038 git clone git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
Jonathan Corbet7994cc12014-12-23 08:43:41 -070039
40Note, however, that you may not want to develop against the mainline tree
41directly. Most subsystem maintainers run their own trees and want to see
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -030042patches prepared against those trees. See the **T:** entry for the subsystem
Jonathan Corbet7994cc12014-12-23 08:43:41 -070043in the MAINTAINERS file to find that tree, or simply ask the maintainer if
44the tree is not listed there.
45
46It is still possible to download kernel releases via tarballs (as described
47in the next section), but that is the hard way to do kernel development.
Linus Torvalds1da177e2005-04-16 15:20:36 -070048
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300491) ``diff -up``
50---------------
Linus Torvalds1da177e2005-04-16 15:20:36 -070051
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -030052If you must generate your patches by hand, use ``diff -up`` or ``diff -uprN``
Jonathan Corbet7994cc12014-12-23 08:43:41 -070053to create patches. Git generates patches in this form by default; if
Mauro Carvalho Chehab9b2c7672016-09-19 08:07:54 -030054you're using ``git``, you can skip this section entirely.
Linus Torvalds1da177e2005-04-16 15:20:36 -070055
56All changes to the Linux kernel occur in the form of patches, as
Mauro Carvalho Chehab9b2c7672016-09-19 08:07:54 -030057generated by :manpage:`diff(1)`. When creating your patch, make sure to
58create it in "unified diff" format, as supplied by the ``-u`` argument
59to :manpage:`diff(1)`.
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -030060Also, please use the ``-p`` argument which shows which C function each
Mauro Carvalho Chehab9b2c7672016-09-19 08:07:54 -030061change is in - that makes the resultant ``diff`` a lot easier to read.
Linus Torvalds1da177e2005-04-16 15:20:36 -070062Patches should be based in the root kernel source directory,
63not in any lower subdirectory.
64
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -030065To create a patch for a single file, it is often sufficient to do::
Linus Torvalds1da177e2005-04-16 15:20:36 -070066
Jonathan Corbetd00c4552014-12-23 08:54:36 -070067 SRCTREE= linux
Linus Torvalds1da177e2005-04-16 15:20:36 -070068 MYFILE= drivers/net/mydriver.c
69
70 cd $SRCTREE
71 cp $MYFILE $MYFILE.orig
72 vi $MYFILE # make your change
73 cd ..
74 diff -up $SRCTREE/$MYFILE{.orig,} > /tmp/patch
75
76To create a patch for multiple files, you should unpack a "vanilla",
Mauro Carvalho Chehab9b2c7672016-09-19 08:07:54 -030077or unmodified kernel source tree, and generate a ``diff`` against your
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -030078own source tree. For example::
Linus Torvalds1da177e2005-04-16 15:20:36 -070079
Jonathan Corbetd00c4552014-12-23 08:54:36 -070080 MYSRC= /devel/linux
Linus Torvalds1da177e2005-04-16 15:20:36 -070081
Jonathan Corbetd00c4552014-12-23 08:54:36 -070082 tar xvfz linux-3.19.tar.gz
83 mv linux-3.19 linux-3.19-vanilla
84 diff -uprN -X linux-3.19-vanilla/Documentation/dontdiff \
85 linux-3.19-vanilla $MYSRC > /tmp/patch
Linus Torvalds1da177e2005-04-16 15:20:36 -070086
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -030087``dontdiff`` is a list of files which are generated by the kernel during
Mauro Carvalho Chehab9b2c7672016-09-19 08:07:54 -030088the build process, and should be ignored in any :manpage:`diff(1)`-generated
Jonathan Corbetd00c4552014-12-23 08:54:36 -070089patch.
Linus Torvalds1da177e2005-04-16 15:20:36 -070090
91Make sure your patch does not include any extra files which do not
92belong in a patch submission. Make sure to review your patch -after-
Mauro Carvalho Chehab9b2c7672016-09-19 08:07:54 -030093generating it with :manpage:`diff(1)`, to ensure accuracy.
Linus Torvalds1da177e2005-04-16 15:20:36 -070094
Josh Triplett8e3072a2014-04-03 14:48:30 -070095If your changes produce a lot of deltas, you need to split them into
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -030096individual patches which modify things in logical stages; see
97:ref:`split_changes`. This will facilitate review by other kernel developers,
Josh Triplett8e3072a2014-04-03 14:48:30 -070098very important if you want your patch accepted.
Linus Torvalds1da177e2005-04-16 15:20:36 -070099
Mauro Carvalho Chehab9b2c7672016-09-19 08:07:54 -0300100If you're using ``git``, ``git rebase -i`` can help you with this process. If
101you're not using ``git``, ``quilt`` <http://savannah.nongnu.org/projects/quilt>
Josh Triplett8e3072a2014-04-03 14:48:30 -0700102is another popular alternative.
Randy Dunlap84da7c02005-06-28 20:45:30 -0700103
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300104.. _describe_changes:
Randy Dunlap84da7c02005-06-28 20:45:30 -0700105
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -03001062) Describe your changes
107------------------------
Linus Torvalds1da177e2005-04-16 15:20:36 -0700108
Johannes Weiner7b9828d2014-08-05 23:32:56 -0700109Describe your problem. Whether your patch is a one-line bug fix or
1105000 lines of a new feature, there must be an underlying problem that
111motivated you to do this work. Convince the reviewer that there is a
112problem worth fixing and that it makes sense for them to read past the
113first paragraph.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700114
Johannes Weiner7b9828d2014-08-05 23:32:56 -0700115Describe user-visible impact. Straight up crashes and lockups are
116pretty convincing, but not all bugs are that blatant. Even if the
117problem was spotted during code review, describe the impact you think
118it can have on users. Keep in mind that the majority of Linux
119installations run kernels from secondary stable trees or
120vendor/product-specific trees that cherry-pick only specific patches
121from upstream, so include anything that could help route your change
122downstream: provoking circumstances, excerpts from dmesg, crash
123descriptions, performance regressions, latency spikes, lockups, etc.
124
125Quantify optimizations and trade-offs. If you claim improvements in
126performance, memory consumption, stack footprint, or binary size,
127include numbers that back them up. But also describe non-obvious
128costs. Optimizations usually aren't free but trade-offs between CPU,
129memory, and readability; or, when it comes to heuristics, between
130different workloads. Describe the expected downsides of your
131optimization so that the reviewer can weigh costs against benefits.
132
133Once the problem is established, describe what you are actually doing
134about it in technical detail. It's important to describe the change
135in plain English for the reviewer to verify that the code is behaving
136as you intend it to.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700137
Theodore Ts'o2ae19aca2009-04-16 07:44:45 -0400138The maintainer will thank you if you write your patch description in a
139form which can be easily pulled into Linux's source code management
Mauro Carvalho Chehab9b2c7672016-09-19 08:07:54 -0300140system, ``git``, as a "commit log". See :ref:`explicit_in_reply_to`.
Theodore Ts'o2ae19aca2009-04-16 07:44:45 -0400141
Johannes Weiner7b9828d2014-08-05 23:32:56 -0700142Solve only one problem per patch. If your description starts to get
143long, that's a sign that you probably need to split up your patch.
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300144See :ref:`split_changes`.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700145
Randy Dunlapd89b1942010-08-09 17:20:21 -0700146When you submit or resubmit a patch or patch series, include the
147complete patch description and justification for it. Don't just
148say that this is version N of the patch (series). Don't expect the
Jonathan Corbetd00c4552014-12-23 08:54:36 -0700149subsystem maintainer to refer back to earlier patch versions or referenced
Randy Dunlapd89b1942010-08-09 17:20:21 -0700150URLs to find the patch description and put that into the patch.
151I.e., the patch (series) and its description should be self-contained.
Jonathan Corbetd00c4552014-12-23 08:54:36 -0700152This benefits both the maintainers and reviewers. Some reviewers
Randy Dunlapd89b1942010-08-09 17:20:21 -0700153probably didn't even receive earlier versions of the patch.
154
Josh Triplett74a475a2014-04-03 14:48:28 -0700155Describe your changes in imperative mood, e.g. "make xyzzy do frotz"
156instead of "[This patch] makes xyzzy do frotz" or "[I] changed xyzzy
157to do frotz", as if you are giving orders to the codebase to change
158its behaviour.
159
Randy Dunlapd89b1942010-08-09 17:20:21 -0700160If the patch fixes a logged bug entry, refer to that bug entry by
Josh Triplett9547c702014-04-03 14:48:29 -0700161number and URL. If the patch follows from a mailing list discussion,
162give a URL to the mailing list archive; use the https://lkml.kernel.org/
Mauro Carvalho Chehab9b2c7672016-09-19 08:07:54 -0300163redirector with a ``Message-Id``, to ensure that the links cannot become
Josh Triplett9547c702014-04-03 14:48:29 -0700164stale.
165
166However, try to make your explanation understandable without external
167resources. In addition to giving a URL to a mailing list archive or
168bug, summarize the relevant points of the discussion that led to the
169patch as submitted.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700170
Geert Uytterhoeven0af52702013-07-31 14:59:38 -0700171If you want to refer to a specific commit, don't just refer to the
172SHA-1 ID of the commit. Please also include the oneline summary of
173the commit, to make it easier for reviewers to know what it is about.
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300174Example::
Geert Uytterhoeven0af52702013-07-31 14:59:38 -0700175
176 Commit e21d2170f36602ae2708 ("video: remove unnecessary
177 platform_set_drvdata()") removed the unnecessary
178 platform_set_drvdata(), but left the variable "dev" unused,
179 delete it.
180
Jonathan Corbet7994cc12014-12-23 08:43:41 -0700181You should also be sure to use at least the first twelve characters of the
182SHA-1 ID. The kernel repository holds a *lot* of objects, making
183collisions with shorter IDs a real possibility. Bear in mind that, even if
184there is no collision with your six-character ID now, that condition may
185change five years from now.
186
Jacob Keller8401aa12014-06-06 14:36:39 -0700187If your patch fixes a bug in a specific commit, e.g. you found an issue using
Mauro Carvalho Chehab9b2c7672016-09-19 08:07:54 -0300188``git bisect``, please use the 'Fixes:' tag with the first 12 characters of
189the SHA-1 ID, and the one line summary. For example::
Jacob Keller8401aa12014-06-06 14:36:39 -0700190
191 Fixes: e21d2170f366 ("video: remove unnecessary platform_set_drvdata()")
192
Mauro Carvalho Chehab9b2c7672016-09-19 08:07:54 -0300193The following ``git config`` settings can be used to add a pretty format for
194outputting the above style in the ``git log`` or ``git show`` commands::
Jacob Keller8401aa12014-06-06 14:36:39 -0700195
196 [core]
197 abbrev = 12
198 [pretty]
199 fixes = Fixes: %h (\"%s\")
Linus Torvalds1da177e2005-04-16 15:20:36 -0700200
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300201.. _split_changes:
Linus Torvalds1da177e2005-04-16 15:20:36 -0700202
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -03002033) Separate your changes
204------------------------
205
206Separate each **logical change** into a separate patch.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700207
208For example, if your changes include both bug fixes and performance
209enhancements for a single driver, separate those changes into two
210or more patches. If your changes include an API update, and a new
211driver which uses that new API, separate those into two patches.
212
213On the other hand, if you make a single change to numerous files,
214group those changes into a single patch. Thus a single logical change
215is contained within a single patch.
216
Jonathan Corbetd00c4552014-12-23 08:54:36 -0700217The point to remember is that each patch should make an easily understood
218change that can be verified by reviewers. Each patch should be justifiable
219on its own merits.
220
Linus Torvalds1da177e2005-04-16 15:20:36 -0700221If one patch depends on another patch in order for a change to be
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300222complete, that is OK. Simply note **"this patch depends on patch X"**
Linus Torvalds1da177e2005-04-16 15:20:36 -0700223in your patch description.
224
Jonathan Corbet7994cc12014-12-23 08:43:41 -0700225When dividing your change into a series of patches, take special care to
226ensure that the kernel builds and runs properly after each patch in the
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300227series. Developers using ``git bisect`` to track down a problem can end up
Jonathan Corbet7994cc12014-12-23 08:43:41 -0700228splitting your patch series at any point; they will not thank you if you
229introduce bugs in the middle.
230
Xose Vazquez Perez5b0ed2c2006-01-08 01:02:49 -0800231If you cannot condense your patch set into a smaller set of patches,
232then only post say 15 or so at a time and wait for review and integration.
233
234
Linus Torvalds1da177e2005-04-16 15:20:36 -0700235
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -03002364) Style-check your changes
237---------------------------
Andy Whitcroft0a920b52007-06-01 00:46:48 -0700238
239Check your patch for basic style violations, details of which can be
Mauro Carvalho Chehabdca22a62016-09-21 08:51:05 -0300240found in
241:ref:`Documentation/CodingStyle <codingstyle>`.
242Failure to do so simply wastes
Linus Nilssonf56d35e2007-07-21 17:49:06 +0200243the reviewers time and will get your patch rejected, probably
Andy Whitcroft0a920b52007-06-01 00:46:48 -0700244without even being read.
245
Jonathan Corbet6de16eb2014-12-23 08:38:24 -0700246One significant exception is when moving code from one file to
247another -- in this case you should not modify the moved code at all in
248the same patch which moves it. This clearly delineates the act of
249moving the code and your changes. This greatly aids review of the
250actual differences and allows tools to better track the history of
251the code itself.
Andy Whitcroft0a920b52007-06-01 00:46:48 -0700252
Jonathan Corbet6de16eb2014-12-23 08:38:24 -0700253Check your patches with the patch style checker prior to submission
254(scripts/checkpatch.pl). Note, though, that the style checker should be
255viewed as a guide, not as a replacement for human judgment. If your code
256looks better with a violation then its probably best left alone.
257
258The checker reports at three levels:
259 - ERROR: things that are very likely to be wrong
260 - WARNING: things requiring careful review
261 - CHECK: things requiring thought
262
263You should be able to justify all violations that remain in your
264patch.
Andy Whitcroft0a920b52007-06-01 00:46:48 -0700265
266
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -03002675) Select the recipients for your patch
268---------------------------------------
Andy Whitcroft0a920b52007-06-01 00:46:48 -0700269
Jonathan Corbetccae8612014-12-23 08:49:18 -0700270You should always copy the appropriate subsystem maintainer(s) on any patch
271to code that they maintain; look through the MAINTAINERS file and the
272source code revision history to see who those maintainers are. The
273script scripts/get_maintainer.pl can be very useful at this step. If you
Sébastien Hindererd6eff072015-07-08 21:12:32 +0200274cannot find a maintainer for the subsystem you are working on, Andrew
Jonathan Corbetccae8612014-12-23 08:49:18 -0700275Morton (akpm@linux-foundation.org) serves as a maintainer of last resort.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700276
Jonathan Corbetccae8612014-12-23 08:49:18 -0700277You should also normally choose at least one mailing list to receive a copy
278of your patch set. linux-kernel@vger.kernel.org functions as a list of
279last resort, but the volume on that list has caused a number of developers
280to tune it out. Look in the MAINTAINERS file for a subsystem-specific
281list; your patch will probably get more attention there. Please do not
282spam unrelated lists, though.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700283
Jonathan Corbetccae8612014-12-23 08:49:18 -0700284Many kernel-related lists are hosted on vger.kernel.org; you can find a
285list of them at http://vger.kernel.org/vger-lists.html. There are
286kernel-related lists hosted elsewhere as well, though.
Xose Vazquez Perez5b0ed2c2006-01-08 01:02:49 -0800287
288Do not send more than 15 patches at once to the vger mailing lists!!!
289
Linus Torvalds1da177e2005-04-16 15:20:36 -0700290Linus Torvalds is the final arbiter of all changes accepted into the
Stefan Bellere00bfcb2014-12-17 17:13:56 -0800291Linux kernel. His e-mail address is <torvalds@linux-foundation.org>.
Jonathan Corbetccae8612014-12-23 08:49:18 -0700292He gets a lot of e-mail, and, at this point, very few patches go through
293Linus directly, so typically you should do your best to -avoid-
Stefan Bellere00bfcb2014-12-17 17:13:56 -0800294sending him e-mail.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700295
Jonathan Corbetccae8612014-12-23 08:49:18 -0700296If you have a patch that fixes an exploitable security bug, send that patch
297to security@kernel.org. For severe bugs, a short embargo may be considered
Nik Nyby253508c2015-06-26 12:05:39 -0400298to allow distributors to get the patch out to users; in such cases,
Jonathan Corbetccae8612014-12-23 08:49:18 -0700299obviously, the patch should not be sent to any public lists.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700300
Jonathan Corbetccae8612014-12-23 08:49:18 -0700301Patches that fix a severe bug in a released kernel should be directed
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300302toward the stable maintainers by putting a line like this::
Linus Torvalds1da177e2005-04-16 15:20:36 -0700303
Jonathan Corbetccae8612014-12-23 08:49:18 -0700304 Cc: stable@vger.kernel.org
Linus Torvalds1da177e2005-04-16 15:20:36 -0700305
Luke Dashjr8cda4c32015-05-14 18:58:01 +0000306into the sign-off area of your patch (note, NOT an email recipient). You
Mauro Carvalho Chehabdca22a62016-09-21 08:51:05 -0300307should also read
308:ref:`Documentation/stable_kernel_rules.txt <stable_kernel_rules>`
309in addition to this file.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700310
Jonathan Corbetccae8612014-12-23 08:49:18 -0700311Note, however, that some subsystem maintainers want to come to their own
312conclusions on which patches should go to the stable trees. The networking
313maintainer, in particular, would rather not see individual developers
314adding lines like the above to their patches.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700315
Jonathan Corbetccae8612014-12-23 08:49:18 -0700316If changes affect userland-kernel interfaces, please send the MAN-PAGES
317maintainer (as listed in the MAINTAINERS file) a man-pages patch, or at
318least a notification of the change, so that some information makes its way
319into the manual pages. User-space API changes should also be copied to
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300320linux-api@vger.kernel.org.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700321
322For small patches you may want to CC the Trivial Patch Monkey
Markus Heidelberg82d27b22009-06-12 01:02:34 +0200323trivial@kernel.org which collects "trivial" patches. Have a look
324into the MAINTAINERS file for its current manager.
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300325
Markus Heidelberg82d27b22009-06-12 01:02:34 +0200326Trivial patches must qualify for one of the following rules:
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300327
Mauro Carvalho Chehab9b2c7672016-09-19 08:07:54 -0300328- Spelling fixes in documentation
329- Spelling fixes for errors which could break :manpage:`grep(1)`
330- Warning fixes (cluttering with useless warnings is bad)
331- Compilation fixes (only if they are actually correct)
332- Runtime fixes (only if they actually fix things)
333- Removing use of deprecated functions/macros
334- Contact detail and documentation fixes
335- Non-portable code replaced by portable code (even in arch-specific,
336 since people copy, as long as it's trivial)
337- Any fix by the author/maintainer of the file (ie. patch monkey
338 in re-transmission mode)
Randy Dunlap84da7c02005-06-28 20:45:30 -0700339
Linus Torvalds1da177e2005-04-16 15:20:36 -0700340
341
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -03003426) No MIME, no links, no compression, no attachments. Just plain text
343----------------------------------------------------------------------
Linus Torvalds1da177e2005-04-16 15:20:36 -0700344
345Linus and other kernel developers need to be able to read and comment
346on the changes you are submitting. It is important for a kernel
347developer to be able to "quote" your changes, using standard e-mail
348tools, so that they may comment on specific portions of your code.
349
Sébastien Hindererbdc89212015-07-12 17:44:15 +0200350For this reason, all patches should be submitted by e-mail "inline".
Mauro Carvalho Chehab9b2c7672016-09-19 08:07:54 -0300351
352.. warning::
353
354 Be wary of your editor's word-wrap corrupting your patch,
355 if you choose to cut-n-paste your patch.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700356
357Do not attach the patch as a MIME attachment, compressed or not.
358Many popular e-mail applications will not always transmit a MIME
359attachment as plain text, making it impossible to comment on your
360code. A MIME attachment also takes Linus a bit more time to process,
361decreasing the likelihood of your MIME-attached change being accepted.
362
363Exception: If your mailer is mangling patches then someone may ask
364you to re-send them using MIME.
365
Mauro Carvalho Chehabdca22a62016-09-21 08:51:05 -0300366See :ref:`Documentation/email-clients.txt <email_clients>`
367for hints about configuring your e-mail client so that it sends your patches
368untouched.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700369
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -03003707) E-mail size
371--------------
Linus Torvalds1da177e2005-04-16 15:20:36 -0700372
373Large changes are not appropriate for mailing lists, and some
Randy Dunlap4932be72009-10-01 15:44:06 -0700374maintainers. If your patch, uncompressed, exceeds 300 kB in size,
Linus Torvalds1da177e2005-04-16 15:20:36 -0700375it is preferred that you store your patch on an Internet-accessible
Jonathan Corbetd00c4552014-12-23 08:54:36 -0700376server, and provide instead a URL (link) pointing to your patch. But note
377that if your patch exceeds 300 kB, it almost certainly needs to be broken up
378anyway.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700379
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -03003808) Respond to review comments
381-----------------------------
Linus Torvalds1da177e2005-04-16 15:20:36 -0700382
Jonathan Corbet0eea2312014-12-23 08:52:01 -0700383Your patch will almost certainly get comments from reviewers on ways in
384which the patch can be improved. You must respond to those comments;
385ignoring reviewers is a good way to get ignored in return. Review comments
386or questions that do not lead to a code change should almost certainly
387bring about a comment or changelog entry so that the next reviewer better
388understands what is going on.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700389
Jonathan Corbet0eea2312014-12-23 08:52:01 -0700390Be sure to tell the reviewers what changes you are making and to thank them
391for their time. Code review is a tiring and time-consuming process, and
392reviewers sometimes get grumpy. Even in that case, though, respond
393politely and address the problems they have pointed out.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700394
395
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -03003969) Don't get discouraged - or impatient
397---------------------------------------
Linus Torvalds1da177e2005-04-16 15:20:36 -0700398
Jonathan Corbet0eea2312014-12-23 08:52:01 -0700399After you have submitted your change, be patient and wait. Reviewers are
400busy people and may not get to your patch right away.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700401
Jonathan Corbet0eea2312014-12-23 08:52:01 -0700402Once upon a time, patches used to disappear into the void without comment,
403but the development process works more smoothly than that now. You should
404receive comments within a week or so; if that does not happen, make sure
405that you have sent your patches to the right place. Wait for a minimum of
406one week before resubmitting or pinging reviewers - possibly longer during
407busy times like merge windows.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700408
409
Jonathan Corbetccae8612014-12-23 08:49:18 -070041010) Include PATCH in the subject
Jonathan Corbetd00c4552014-12-23 08:54:36 -0700411--------------------------------
Linus Torvalds1da177e2005-04-16 15:20:36 -0700412
413Due to high e-mail traffic to Linus, and to linux-kernel, it is common
414convention to prefix your subject line with [PATCH]. This lets Linus
415and other kernel developers more easily distinguish patches from other
416e-mail discussions.
417
418
419
Jonathan Corbetccae8612014-12-23 08:49:18 -070042011) Sign your work
Jonathan Corbetd00c4552014-12-23 08:54:36 -0700421------------------
Linus Torvalds1da177e2005-04-16 15:20:36 -0700422
423To improve tracking of who did what, especially with patches that can
424percolate to their final resting place in the kernel through several
425layers of maintainers, we've introduced a "sign-off" procedure on
426patches that are being emailed around.
427
428The sign-off is a simple line at the end of the explanation for the
429patch, which certifies that you wrote it or otherwise have the right to
Zac Storerdb12fb82011-08-13 12:34:45 -0700430pass it on as an open-source patch. The rules are pretty simple: if you
Linus Torvalds1da177e2005-04-16 15:20:36 -0700431can certify the below:
432
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300433Developer's Certificate of Origin 1.1
434^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Linus Torvalds1da177e2005-04-16 15:20:36 -0700435
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300436By making a contribution to this project, I certify that:
Linus Torvalds1da177e2005-04-16 15:20:36 -0700437
438 (a) The contribution was created in whole or in part by me and I
439 have the right to submit it under the open source license
440 indicated in the file; or
441
442 (b) The contribution is based upon previous work that, to the best
443 of my knowledge, is covered under an appropriate open source
444 license and I have the right under that license to submit that
445 work with modifications, whether created in whole or in part
446 by me, under the same open source license (unless I am
447 permitted to submit under a different license), as indicated
448 in the file; or
449
450 (c) The contribution was provided directly to me by some other
451 person who certified (a), (b) or (c) and I have not modified
452 it.
453
Stefan Bellere00bfcb2014-12-17 17:13:56 -0800454 (d) I understand and agree that this project and the contribution
455 are public and that a record of the contribution (including all
456 personal information I submit with it, including my sign-off) is
457 maintained indefinitely and may be redistributed consistent with
458 this project or the open source license(s) involved.
Linus Torvaldscbd83da2005-06-13 17:51:55 -0700459
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300460then you just add a line saying::
Linus Torvalds1da177e2005-04-16 15:20:36 -0700461
Alexey Dobriyan9fd55592005-06-25 14:59:34 -0700462 Signed-off-by: Random J Developer <random@developer.example.org>
Linus Torvalds1da177e2005-04-16 15:20:36 -0700463
Greg KHaf45f322006-09-12 20:35:52 -0700464using your real name (sorry, no pseudonyms or anonymous contributions.)
465
Linus Torvalds1da177e2005-04-16 15:20:36 -0700466Some people also put extra tags at the end. They'll just be ignored for
467now, but you can do this to mark internal company procedures or just
Stefan Bellere00bfcb2014-12-17 17:13:56 -0800468point out some special detail about the sign-off.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700469
Willy Tarreauadbd5882008-06-03 00:20:28 +0200470If you are a subsystem or branch maintainer, sometimes you need to slightly
471modify patches you receive in order to merge them, because the code is not
472exactly the same in your tree and the submitters'. If you stick strictly to
473rule (c), you should ask the submitter to rediff, but this is a totally
474counter-productive waste of time and energy. Rule (b) allows you to adjust
475the code, but then it is very impolite to change one submitter's code and
476make him endorse your bugs. To solve this problem, it is recommended that
477you add a line between the last Signed-off-by header and yours, indicating
478the nature of your changes. While there is nothing mandatory about this, it
479seems like prepending the description with your mail and/or name, all
480enclosed in square brackets, is noticeable enough to make it obvious that
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300481you are responsible for last-minute changes. Example::
Willy Tarreauadbd5882008-06-03 00:20:28 +0200482
483 Signed-off-by: Random J Developer <random@developer.example.org>
484 [lucky@maintainer.example.org: struct foo moved from foo.c to foo.h]
485 Signed-off-by: Lucky K Maintainer <lucky@maintainer.example.org>
486
Jeremiah Mahler305af082014-05-22 00:04:26 -0700487This practice is particularly helpful if you maintain a stable branch and
Willy Tarreauadbd5882008-06-03 00:20:28 +0200488want at the same time to credit the author, track changes, merge the fix,
489and protect the submitter from complaints. Note that under no circumstances
490can you change the author's identity (the From header), as it is the one
491which appears in the changelog.
492
Jeremiah Mahler305af082014-05-22 00:04:26 -0700493Special note to back-porters: It seems to be a common and useful practice
Willy Tarreauadbd5882008-06-03 00:20:28 +0200494to insert an indication of the origin of a patch at the top of the commit
495message (just after the subject line) to facilitate tracking. For instance,
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300496here's what we see in a 3.x-stable release::
Willy Tarreauadbd5882008-06-03 00:20:28 +0200497
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300498 Date: Tue Oct 7 07:26:38 2014 -0400
Willy Tarreauadbd5882008-06-03 00:20:28 +0200499
Jonathan Corbet7994cc12014-12-23 08:43:41 -0700500 libata: Un-break ATA blacklist
Willy Tarreauadbd5882008-06-03 00:20:28 +0200501
Jonathan Corbet7994cc12014-12-23 08:43:41 -0700502 commit 1c40279960bcd7d52dbdf1d466b20d24b99176c8 upstream.
Willy Tarreauadbd5882008-06-03 00:20:28 +0200503
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300504And here's what might appear in an older kernel once a patch is backported::
Willy Tarreauadbd5882008-06-03 00:20:28 +0200505
506 Date: Tue May 13 22:12:27 2008 +0200
507
508 wireless, airo: waitbusy() won't delay
509
510 [backport of 2.6 commit b7acbdfbd1f277c1eb23f344f899cfa4cd0bf36a]
511
512Whatever the format, this information provides a valuable help to people
Jonathan Corbet7994cc12014-12-23 08:43:41 -0700513tracking your trees, and to people trying to troubleshoot bugs in your
Willy Tarreauadbd5882008-06-03 00:20:28 +0200514tree.
515
Linus Torvalds1da177e2005-04-16 15:20:36 -0700516
Jonathan Corbetccae8612014-12-23 08:49:18 -070051712) When to use Acked-by: and Cc:
Jonathan Corbetd00c4552014-12-23 08:54:36 -0700518---------------------------------
Andy Whitcroft0a920b52007-06-01 00:46:48 -0700519
Andrew Morton0f44cd22007-06-08 13:46:45 -0700520The Signed-off-by: tag indicates that the signer was involved in the
521development of the patch, or that he/she was in the patch's delivery path.
522
523If a person was not directly involved in the preparation or handling of a
524patch but wishes to signify and record their approval of it then they can
Jonathan Corbetd00c4552014-12-23 08:54:36 -0700525ask to have an Acked-by: line added to the patch's changelog.
Andrew Morton0f44cd22007-06-08 13:46:45 -0700526
527Acked-by: is often used by the maintainer of the affected code when that
528maintainer neither contributed to nor forwarded the patch.
529
530Acked-by: is not as formal as Signed-off-by:. It is a record that the acker
531has at least reviewed the patch and has indicated acceptance. Hence patch
532mergers will sometimes manually convert an acker's "yep, looks good to me"
Jonathan Corbetd00c4552014-12-23 08:54:36 -0700533into an Acked-by: (but note that it is usually better to ask for an
534explicit ack).
Andrew Morton0f44cd22007-06-08 13:46:45 -0700535
536Acked-by: does not necessarily indicate acknowledgement of the entire patch.
537For example, if a patch affects multiple subsystems and has an Acked-by: from
538one subsystem maintainer then this usually indicates acknowledgement of just
539the part which affects that maintainer's code. Judgement should be used here.
Jonathan Corbetef402032008-03-28 11:22:38 -0600540When in doubt people should refer to the original discussion in the mailing
Andrew Morton0f44cd22007-06-08 13:46:45 -0700541list archives.
542
Jonathan Corbetef402032008-03-28 11:22:38 -0600543If a person has had the opportunity to comment on a patch, but has not
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300544provided such comments, you may optionally add a ``Cc:`` tag to the patch.
Jonathan Corbetef402032008-03-28 11:22:38 -0600545This is the only tag which might be added without an explicit action by the
Jonathan Corbetd00c4552014-12-23 08:54:36 -0700546person it names - but it should indicate that this person was copied on the
547patch. This tag documents that potentially interested parties
548have been included in the discussion.
Andrew Morton0f44cd22007-06-08 13:46:45 -0700549
Jonathan Corbetef402032008-03-28 11:22:38 -0600550
Jonathan Corbetccae8612014-12-23 08:49:18 -070055113) Using Reported-by:, Tested-by:, Reviewed-by:, Suggested-by: and Fixes:
Jonathan Corbetd00c4552014-12-23 08:54:36 -0700552--------------------------------------------------------------------------
Jonathan Corbetbbb0a4242009-01-16 09:49:50 -0700553
Dan Carpenterd75ef702014-10-29 13:01:36 +0300554The Reported-by tag gives credit to people who find bugs and report them and it
555hopefully inspires them to help us again in the future. Please note that if
556the bug was reported in private, then ask for permission first before using the
557Reported-by tag.
Jonathan Corbetef402032008-03-28 11:22:38 -0600558
559A Tested-by: tag indicates that the patch has been successfully tested (in
560some environment) by the person named. This tag informs maintainers that
561some testing has been performed, provides a means to locate testers for
562future patches, and ensures credit for the testers.
563
564Reviewed-by:, instead, indicates that the patch has been reviewed and found
565acceptable according to the Reviewer's Statement:
566
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300567Reviewer's statement of oversight
568^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Jonathan Corbetef402032008-03-28 11:22:38 -0600569
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300570By offering my Reviewed-by: tag, I state that:
Jonathan Corbetef402032008-03-28 11:22:38 -0600571
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300572 (a) I have carried out a technical review of this patch to
Jonathan Corbetef402032008-03-28 11:22:38 -0600573 evaluate its appropriateness and readiness for inclusion into
574 the mainline kernel.
575
576 (b) Any problems, concerns, or questions relating to the patch
577 have been communicated back to the submitter. I am satisfied
578 with the submitter's response to my comments.
579
580 (c) While there may be things that could be improved with this
581 submission, I believe that it is, at this time, (1) a
582 worthwhile modification to the kernel, and (2) free of known
583 issues which would argue against its inclusion.
584
585 (d) While I have reviewed the patch and believe it to be sound, I
586 do not (unless explicitly stated elsewhere) make any
587 warranties or guarantees that it will achieve its stated
588 purpose or function properly in any given situation.
589
590A Reviewed-by tag is a statement of opinion that the patch is an
591appropriate modification of the kernel without any remaining serious
592technical issues. Any interested reviewer (who has done the work) can
593offer a Reviewed-by tag for a patch. This tag serves to give credit to
594reviewers and to inform maintainers of the degree of review which has been
595done on the patch. Reviewed-by: tags, when supplied by reviewers known to
596understand the subject area and to perform thorough reviews, will normally
Pavel Machek5801da12009-06-04 16:26:50 +0200597increase the likelihood of your patch getting into the kernel.
Jonathan Corbetef402032008-03-28 11:22:38 -0600598
Mugunthan V N8543ae12013-04-29 16:18:17 -0700599A Suggested-by: tag indicates that the patch idea is suggested by the person
600named and ensures credit to the person for the idea. Please note that this
601tag should not be added without the reporter's permission, especially if the
602idea was not posted in a public forum. That said, if we diligently credit our
603idea reporters, they will, hopefully, be inspired to help us again in the
604future.
605
Jacob Keller8401aa12014-06-06 14:36:39 -0700606A Fixes: tag indicates that the patch fixes an issue in a previous commit. It
607is used to make it easy to determine where a bug originated, which can help
608review a bug fix. This tag also assists the stable kernel team in determining
609which stable kernel versions should receive your fix. This is the preferred
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300610method for indicating a bug fixed by the patch. See :ref:`describe_changes`
611for more details.
Jacob Keller8401aa12014-06-06 14:36:39 -0700612
Jonathan Corbetef402032008-03-28 11:22:38 -0600613
Jonathan Corbetccae8612014-12-23 08:49:18 -070061414) The canonical patch format
Jonathan Corbet7994cc12014-12-23 08:43:41 -0700615------------------------------
616
617This section describes how the patch itself should be formatted. Note
Mauro Carvalho Chehab9b2c7672016-09-19 08:07:54 -0300618that, if you have your patches stored in a ``git`` repository, proper patch
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300619formatting can be had with ``git format-patch``. The tools cannot create
Jonathan Corbet7994cc12014-12-23 08:43:41 -0700620the necessary text, though, so read the instructions below anyway.
Randy Dunlap84da7c02005-06-28 20:45:30 -0700621
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300622The canonical patch subject line is::
Paul Jackson75f84262005-10-02 18:01:42 -0700623
Paul Jacksond6b9acc2005-10-03 00:29:10 -0700624 Subject: [PATCH 001/123] subsystem: summary phrase
Paul Jackson75f84262005-10-02 18:01:42 -0700625
626The canonical patch message body contains the following:
627
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300628 - A ``from`` line specifying the patch author (only needed if the person
Jonathan Corbetccae8612014-12-23 08:49:18 -0700629 sending the patch is not the author).
Paul Jackson75f84262005-10-02 18:01:42 -0700630
631 - An empty line.
632
Joe Perches2a076f42015-04-16 12:44:28 -0700633 - The body of the explanation, line wrapped at 75 columns, which will
634 be copied to the permanent changelog to describe this patch.
Paul Jackson75f84262005-10-02 18:01:42 -0700635
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300636 - The ``Signed-off-by:`` lines, described above, which will
Paul Jackson75f84262005-10-02 18:01:42 -0700637 also go in the changelog.
638
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300639 - A marker line containing simply ``---``.
Paul Jackson75f84262005-10-02 18:01:42 -0700640
641 - Any additional comments not suitable for the changelog.
642
Mauro Carvalho Chehab9b2c7672016-09-19 08:07:54 -0300643 - The actual patch (``diff`` output).
Paul Jackson75f84262005-10-02 18:01:42 -0700644
645The Subject line format makes it very easy to sort the emails
646alphabetically by subject line - pretty much any email reader will
647support that - since because the sequence number is zero-padded,
648the numerical and alphabetic sort is the same.
649
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300650The ``subsystem`` in the email's Subject should identify which
Paul Jacksond6b9acc2005-10-03 00:29:10 -0700651area or subsystem of the kernel is being patched.
652
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300653The ``summary phrase`` in the email's Subject should concisely
654describe the patch which that email contains. The ``summary
655phrase`` should not be a filename. Do not use the same ``summary
656phrase`` for every patch in a whole patch series (where a ``patch
657series`` is an ordered sequence of multiple, related patches).
Paul Jacksond6b9acc2005-10-03 00:29:10 -0700658
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300659Bear in mind that the ``summary phrase`` of your email becomes a
Theodore Ts'o2ae19aca2009-04-16 07:44:45 -0400660globally-unique identifier for that patch. It propagates all the way
Mauro Carvalho Chehab9b2c7672016-09-19 08:07:54 -0300661into the ``git`` changelog. The ``summary phrase`` may later be used in
Theodore Ts'o2ae19aca2009-04-16 07:44:45 -0400662developer discussions which refer to the patch. People will want to
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300663google for the ``summary phrase`` to read discussion regarding that
Theodore Ts'o2ae19aca2009-04-16 07:44:45 -0400664patch. It will also be the only thing that people may quickly see
665when, two or three months later, they are going through perhaps
Mauro Carvalho Chehab9b2c7672016-09-19 08:07:54 -0300666thousands of patches using tools such as ``gitk`` or ``git log
667--oneline``.
Theodore Ts'o2ae19aca2009-04-16 07:44:45 -0400668
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300669For these reasons, the ``summary`` must be no more than 70-75
Theodore Ts'o2ae19aca2009-04-16 07:44:45 -0400670characters, and it must describe both what the patch changes, as well
671as why the patch might be necessary. It is challenging to be both
672succinct and descriptive, but that is what a well-written summary
673should do.
674
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300675The ``summary phrase`` may be prefixed by tags enclosed in square
Alex Henriee12d7462015-09-20 14:11:19 +0200676brackets: "Subject: [PATCH <tag>...] <summary phrase>". The tags are
677not considered part of the summary phrase, but describe how the patch
Theodore Ts'o2ae19aca2009-04-16 07:44:45 -0400678should be treated. Common tags might include a version descriptor if
679the multiple versions of the patch have been sent out in response to
680comments (i.e., "v1, v2, v3"), or "RFC" to indicate a request for
681comments. If there are four patches in a patch series the individual
682patches may be numbered like this: 1/4, 2/4, 3/4, 4/4. This assures
683that developers understand the order in which the patches should be
684applied and that they have reviewed or applied all of the patches in
685the patch series.
Paul Jacksond6b9acc2005-10-03 00:29:10 -0700686
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300687A couple of example Subjects::
Paul Jacksond6b9acc2005-10-03 00:29:10 -0700688
Alex Henriee12d7462015-09-20 14:11:19 +0200689 Subject: [PATCH 2/5] ext2: improve scalability of bitmap searching
690 Subject: [PATCH v2 01/27] x86: fix eflags tracking
Paul Jackson75f84262005-10-02 18:01:42 -0700691
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300692The ``from`` line must be the very first line in the message body,
Paul Jackson75f84262005-10-02 18:01:42 -0700693and has the form:
694
695 From: Original Author <author@example.com>
696
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300697The ``from`` line specifies who will be credited as the author of the
698patch in the permanent changelog. If the ``from`` line is missing,
699then the ``From:`` line from the email header will be used to determine
Paul Jackson75f84262005-10-02 18:01:42 -0700700the patch author in the changelog.
701
702The explanation body will be committed to the permanent source
703changelog, so should make sense to a competent reader who has long
704since forgotten the immediate details of the discussion that might
Theodore Ts'o2ae19aca2009-04-16 07:44:45 -0400705have led to this patch. Including symptoms of the failure which the
706patch addresses (kernel log messages, oops messages, etc.) is
707especially useful for people who might be searching the commit logs
708looking for the applicable patch. If a patch fixes a compile failure,
709it may not be necessary to include _all_ of the compile failures; just
710enough that it is likely that someone searching for the patch can find
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300711it. As in the ``summary phrase``, it is important to be both succinct as
Theodore Ts'o2ae19aca2009-04-16 07:44:45 -0400712well as descriptive.
Paul Jackson75f84262005-10-02 18:01:42 -0700713
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300714The ``---`` marker line serves the essential purpose of marking for patch
Paul Jackson75f84262005-10-02 18:01:42 -0700715handling tools where the changelog message ends.
716
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300717One good use for the additional comments after the ``---`` marker is for
Mauro Carvalho Chehab9b2c7672016-09-19 08:07:54 -0300718a ``diffstat``, to show what files have changed, and the number of
719inserted and deleted lines per file. A ``diffstat`` is especially useful
Theodore Ts'o2ae19aca2009-04-16 07:44:45 -0400720on bigger patches. Other comments relevant only to the moment or the
721maintainer, not suitable for the permanent changelog, should also go
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300722here. A good example of such comments might be ``patch changelogs``
Theodore Ts'o2ae19aca2009-04-16 07:44:45 -0400723which describe what has changed between the v1 and v2 version of the
724patch.
725
Mauro Carvalho Chehab9b2c7672016-09-19 08:07:54 -0300726If you are going to include a ``diffstat`` after the ``---`` marker, please
727use ``diffstat`` options ``-p 1 -w 70`` so that filenames are listed from
Theodore Ts'o2ae19aca2009-04-16 07:44:45 -0400728the top of the kernel source tree and don't use too much horizontal
Mauro Carvalho Chehab9b2c7672016-09-19 08:07:54 -0300729space (easily fit in 80 columns, maybe with some indentation). (``git``
Josh Triplett8e3072a2014-04-03 14:48:30 -0700730generates appropriate diffstats by default.)
Paul Jackson75f84262005-10-02 18:01:42 -0700731
732See more details on the proper patch format in the following
733references.
734
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300735.. _explicit_in_reply_to:
736
Chris Metcalfd7ac8d82015-11-05 15:21:47 -050073715) Explicit In-Reply-To headers
738--------------------------------
Paul Jackson75f84262005-10-02 18:01:42 -0700739
Chris Metcalfd7ac8d82015-11-05 15:21:47 -0500740It can be helpful to manually add In-Reply-To: headers to a patch
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300741(e.g., when using ``git send-email``) to associate the patch with
Chris Metcalfd7ac8d82015-11-05 15:21:47 -0500742previous relevant discussion, e.g. to link a bug fix to the email with
743the bug report. However, for a multi-patch series, it is generally
744best to avoid using In-Reply-To: to link to older versions of the
745series. This way multiple versions of the patch don't become an
746unmanageable forest of references in email clients. If a link is
747helpful, you can use the https://lkml.kernel.org/ redirector (e.g., in
748the cover email text) to link to an earlier version of the patch series.
749
750
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -030075116) Sending ``git pull`` requests
752---------------------------------
Randy Dunlap84da7c02005-06-28 20:45:30 -0700753
Jonathan Corbet7994cc12014-12-23 08:43:41 -0700754If you have a series of patches, it may be most convenient to have the
755maintainer pull them directly into the subsystem repository with a
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300756``git pull`` operation. Note, however, that pulling patches from a developer
Jonathan Corbet7994cc12014-12-23 08:43:41 -0700757requires a higher degree of trust than taking patches from a mailing list.
758As a result, many subsystem maintainers are reluctant to take pull
Jonathan Corbetb792ffe2014-12-23 09:28:40 -0700759requests, especially from new, unknown developers. If in doubt you can use
760the pull request as the cover letter for a normal posting of the patch
761series, giving the maintainer the option of using either.
Randy Dunlap14863612008-07-27 20:44:24 -0700762
Jonathan Corbet7994cc12014-12-23 08:43:41 -0700763A pull request should have [GIT] or [PULL] in the subject line. The
764request itself should include the repository name and the branch of
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300765interest on a single line; it should look something like::
Randy Dunlap14863612008-07-27 20:44:24 -0700766
Jonathan Corbet7994cc12014-12-23 08:43:41 -0700767 Please pull from
Randy Dunlap14863612008-07-27 20:44:24 -0700768
Jonathan Corbet7994cc12014-12-23 08:43:41 -0700769 git://jdelvare.pck.nerim.net/jdelvare-2.6 i2c-for-linus
Randy Dunlap14863612008-07-27 20:44:24 -0700770
Jakub Wilk64e32892015-07-27 10:15:18 +0200771 to get these changes:
Randy Dunlap14863612008-07-27 20:44:24 -0700772
Jonathan Corbet7994cc12014-12-23 08:43:41 -0700773A pull request should also include an overall message saying what will be
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300774included in the request, a ``git shortlog`` listing of the patches
Mauro Carvalho Chehab9b2c7672016-09-19 08:07:54 -0300775themselves, and a ``diffstat`` showing the overall effect of the patch series.
Jonathan Corbet7994cc12014-12-23 08:43:41 -0700776The easiest way to get all this information together is, of course, to let
Mauro Carvalho Chehab9b2c7672016-09-19 08:07:54 -0300777``git`` do it for you with the ``git request-pull`` command.
Randy Dunlap14863612008-07-27 20:44:24 -0700778
Jonathan Corbet7994cc12014-12-23 08:43:41 -0700779Some maintainers (including Linus) want to see pull requests from signed
780commits; that increases their confidence that the request actually came
781from you. Linus, in particular, will not pull from public hosting sites
782like GitHub in the absence of a signed tag.
Randy Dunlap14863612008-07-27 20:44:24 -0700783
Jonathan Corbet7994cc12014-12-23 08:43:41 -0700784The first step toward creating such tags is to make a GNUPG key and get it
785signed by one or more core kernel developers. This step can be hard for
786new developers, but there is no way around it. Attending conferences can
787be a good way to find developers who can sign your key.
Randy Dunlap14863612008-07-27 20:44:24 -0700788
Mauro Carvalho Chehab9b2c7672016-09-19 08:07:54 -0300789Once you have prepared a patch series in ``git`` that you wish to have somebody
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300790pull, create a signed tag with ``git tag -s``. This will create a new tag
Jonathan Corbet7994cc12014-12-23 08:43:41 -0700791identifying the last commit in the series and containing a signature
792created with your private key. You will also have the opportunity to add a
793changelog-style message to the tag; this is an ideal place to describe the
794effects of the pull request as a whole.
Randy Dunlap84da7c02005-06-28 20:45:30 -0700795
Jonathan Corbet7994cc12014-12-23 08:43:41 -0700796If the tree the maintainer will be pulling from is not the repository you
797are working from, don't forget to push the signed tag explicitly to the
798public tree.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700799
Jonathan Corbet7994cc12014-12-23 08:43:41 -0700800When generating your pull request, use the signed tag as the target. A
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300801command like this will do the trick::
Linus Torvalds1da177e2005-04-16 15:20:36 -0700802
Jonathan Corbet7994cc12014-12-23 08:43:41 -0700803 git request-pull master git://my.public.tree/linux.git my-signed-tag
Xose Vazquez Perez5b0ed2c2006-01-08 01:02:49 -0800804
805
Mauro Carvalho Chehab59030192016-09-19 08:07:53 -0300806REFERENCES
807**********
Xose Vazquez Perez5b0ed2c2006-01-08 01:02:49 -0800808
809Andrew Morton, "The perfect patch" (tpp).
Mitchel Humpherys37c703f2014-04-03 14:50:40 -0700810 <http://www.ozlabs.org/~akpm/stuff/tpp.txt>
Xose Vazquez Perez5b0ed2c2006-01-08 01:02:49 -0800811
Pavel Machek8e9cb8f2006-09-29 02:01:29 -0700812Jeff Garzik, "Linux kernel patch submission format".
Xose Vazquez Perez5b0ed2c2006-01-08 01:02:49 -0800813 <http://linux.yyz.us/patch-format.html>
814
Pavel Machek8e9cb8f2006-09-29 02:01:29 -0700815Greg Kroah-Hartman, "How to piss off a kernel subsystem maintainer".
Vikram Narayananf5039932011-05-23 12:01:25 -0700816 <http://www.kroah.com/log/linux/maintainer.html>
Mauro Carvalho Chehab9b2c7672016-09-19 08:07:54 -0300817
Vikram Narayananf5039932011-05-23 12:01:25 -0700818 <http://www.kroah.com/log/linux/maintainer-02.html>
Mauro Carvalho Chehab9b2c7672016-09-19 08:07:54 -0300819
Vikram Narayananf5039932011-05-23 12:01:25 -0700820 <http://www.kroah.com/log/linux/maintainer-03.html>
Mauro Carvalho Chehab9b2c7672016-09-19 08:07:54 -0300821
Vikram Narayananf5039932011-05-23 12:01:25 -0700822 <http://www.kroah.com/log/linux/maintainer-04.html>
Mauro Carvalho Chehab9b2c7672016-09-19 08:07:54 -0300823
Vikram Narayananf5039932011-05-23 12:01:25 -0700824 <http://www.kroah.com/log/linux/maintainer-05.html>
Mauro Carvalho Chehab9b2c7672016-09-19 08:07:54 -0300825
Sudip Mukherjee7e0dae62014-09-07 11:26:12 -0700826 <http://www.kroah.com/log/linux/maintainer-06.html>
Xose Vazquez Perez5b0ed2c2006-01-08 01:02:49 -0800827
Randy Dunlapbc7455f2006-07-30 03:03:45 -0700828NO!!!! No more huge patch bombs to linux-kernel@vger.kernel.org people!
Mitchel Humpherys37c703f2014-04-03 14:50:40 -0700829 <https://lkml.org/lkml/2005/7/11/336>
Xose Vazquez Perez5b0ed2c2006-01-08 01:02:49 -0800830
Pavel Machek8e9cb8f2006-09-29 02:01:29 -0700831Kernel Documentation/CodingStyle:
Mauro Carvalho Chehabdca22a62016-09-21 08:51:05 -0300832 :ref:`Documentation/CodingStyle <codingstyle>`
Xose Vazquez Perez5b0ed2c2006-01-08 01:02:49 -0800833
Pavel Machek8e9cb8f2006-09-29 02:01:29 -0700834Linus Torvalds's mail on the canonical patch format:
Xose Vazquez Perez5b0ed2c2006-01-08 01:02:49 -0800835 <http://lkml.org/lkml/2005/4/7/183>
Andi Kleen95367272008-10-15 22:02:02 -0700836
837Andi Kleen, "On submitting kernel patches"
Lucas De Marchi25985ed2011-03-30 22:57:33 -0300838 Some strategies to get difficult or controversial changes in.
Mauro Carvalho Chehab9b2c7672016-09-19 08:07:54 -0300839
Andi Kleen95367272008-10-15 22:02:02 -0700840 http://halobates.de/on-submitting-patches.pdf
841