blob: 903a2546f13890b17d309554f3cbda69c2ee9ade [file] [log] [blame]
Jonathan Corbet75b02142008-09-30 15:15:56 -060015: POSTING PATCHES
2
3Sooner or later, the time comes when your work is ready to be presented to
4the community for review and, eventually, inclusion into the mainline
5kernel. Unsurprisingly, the kernel development community has evolved a set
6of conventions and procedures which are used in the posting of patches;
7following them will make life much easier for everybody involved. This
8document will attempt to cover these expectations in reasonable detail;
9more information can also be found in the files SubmittingPatches,
10SubmittingDrivers, and SubmitChecklist in the kernel documentation
11directory.
12
13
145.1: WHEN TO POST
15
16There is a constant temptation to avoid posting patches before they are
17completely "ready." For simple patches, that is not a problem. If the
18work being done is complex, though, there is a lot to be gained by getting
19feedback from the community before the work is complete. So you should
20consider posting in-progress work, or even making a git tree available so
21that interested developers can catch up with your work at any time.
22
23When posting code which is not yet considered ready for inclusion, it is a
24good idea to say so in the posting itself. Also mention any major work
25which remains to be done and any known problems. Fewer people will look at
26patches which are known to be half-baked, but those who do will come in
27with the idea that they can help you drive the work in the right direction.
28
29
305.2: BEFORE CREATING PATCHES
31
32There are a number of things which should be done before you consider
33sending patches to the development community. These include:
34
35 - Test the code to the extent that you can. Make use of the kernel's
36 debugging tools, ensure that the kernel will build with all reasonable
37 combinations of configuration options, use cross-compilers to build for
38 different architectures, etc.
39
40 - Make sure your code is compliant with the kernel coding style
41 guidelines.
42
43 - Does your change have performance implications? If so, you should run
44 benchmarks showing what the impact (or benefit) of your change is; a
45 summary of the results should be included with the patch.
46
47 - Be sure that you have the right to post the code. If this work was done
48 for an employer, the employer likely has a right to the work and must be
49 agreeable with its release under the GPL.
50
51As a general rule, putting in some extra thought before posting code almost
52always pays back the effort in short order.
53
54
555.3: PATCH PREPARATION
56
57The preparation of patches for posting can be a surprising amount of work,
58but, once again, attempting to save time here is not generally advisable
59even in the short term.
60
61Patches must be prepared against a specific version of the kernel. As a
62general rule, a patch should be based on the current mainline as found in
Jonathan Corbet5c050fb2011-03-25 12:17:53 -060063Linus's git tree. When basing on mainline, start with a well-known release
64point - a stable or -rc release - rather than branching off the mainline at
65an arbitrary spot.
66
67It may become necessary to make versions against -mm, linux-next, or a
68subsystem tree, though, to facilitate wider testing and review. Depending
69on the area of your patch and what is going on elsewhere, basing a patch
70against these other trees can require a significant amount of work
71resolving conflicts and dealing with API changes.
Jonathan Corbet75b02142008-09-30 15:15:56 -060072
73Only the most simple changes should be formatted as a single patch;
74everything else should be made as a logical series of changes. Splitting
75up patches is a bit of an art; some developers spend a long time figuring
76out how to do it in the way that the community expects. There are a few
77rules of thumb, however, which can help considerably:
78
79 - The patch series you post will almost certainly not be the series of
80 changes found in your working revision control system. Instead, the
81 changes you have made need to be considered in their final form, then
82 split apart in ways which make sense. The developers are interested in
83 discrete, self-contained changes, not the path you took to get to those
84 changes.
85
86 - Each logically independent change should be formatted as a separate
87 patch. These changes can be small ("add a field to this structure") or
88 large (adding a significant new driver, for example), but they should be
89 conceptually small and amenable to a one-line description. Each patch
90 should make a specific change which can be reviewed on its own and
91 verified to do what it says it does.
92
93 - As a way of restating the guideline above: do not mix different types of
94 changes in the same patch. If a single patch fixes a critical security
95 bug, rearranges a few structures, and reformats the code, there is a
96 good chance that it will be passed over and the important fix will be
97 lost.
98
99 - Each patch should yield a kernel which builds and runs properly; if your
100 patch series is interrupted in the middle, the result should still be a
101 working kernel. Partial application of a patch series is a common
102 scenario when the "git bisect" tool is used to find regressions; if the
103 result is a broken kernel, you will make life harder for developers and
104 users who are engaging in the noble work of tracking down problems.
105
Jonathan Corbet5c050fb2011-03-25 12:17:53 -0600106 - Do not overdo it, though. One developer once posted a set of edits
Jonathan Corbet75b02142008-09-30 15:15:56 -0600107 to a single file as 500 separate patches - an act which did not make him
108 the most popular person on the kernel mailing list. A single patch can
109 be reasonably large as long as it still contains a single *logical*
Jonathan Corbet5c050fb2011-03-25 12:17:53 -0600110 change.
Jonathan Corbet75b02142008-09-30 15:15:56 -0600111
112 - It can be tempting to add a whole new infrastructure with a series of
113 patches, but to leave that infrastructure unused until the final patch
114 in the series enables the whole thing. This temptation should be
115 avoided if possible; if that series adds regressions, bisection will
116 finger the last patch as the one which caused the problem, even though
117 the real bug is elsewhere. Whenever possible, a patch which adds new
118 code should make that code active immediately.
119
120Working to create the perfect patch series can be a frustrating process
121which takes quite a bit of time and thought after the "real work" has been
122done. When done properly, though, it is time well spent.
123
124
Jonathan Corbet5d989322009-04-21 13:33:06 -06001255.4: PATCH FORMATTING AND CHANGELOGS
Jonathan Corbet75b02142008-09-30 15:15:56 -0600126
127So now you have a perfect series of patches for posting, but the work is
128not done quite yet. Each patch needs to be formatted into a message which
129quickly and clearly communicates its purpose to the rest of the world. To
130that end, each patch will be composed of the following:
131
132 - An optional "From" line naming the author of the patch. This line is
133 only necessary if you are passing on somebody else's patch via email,
134 but it never hurts to add it when in doubt.
135
136 - A one-line description of what the patch does. This message should be
137 enough for a reader who sees it with no other context to figure out the
138 scope of the patch; it is the line that will show up in the "short form"
139 changelogs. This message is usually formatted with the relevant
140 subsystem name first, followed by the purpose of the patch. For
141 example:
142
143 gpio: fix build on CONFIG_GPIO_SYSFS=n
144
145 - A blank line followed by a detailed description of the contents of the
146 patch. This description can be as long as is required; it should say
147 what the patch does and why it should be applied to the kernel.
148
149 - One or more tag lines, with, at a minimum, one Signed-off-by: line from
150 the author of the patch. Tags will be described in more detail below.
151
Jonathan Corbet5d989322009-04-21 13:33:06 -0600152The items above, together, form the changelog for the patch. Writing good
153changelogs is a crucial but often-neglected art; it's worth spending
154another moment discussing this issue. When writing a changelog, you should
155bear in mind that a number of different people will be reading your words.
156These include subsystem maintainers and reviewers who need to decide
157whether the patch should be included, distributors and other maintainers
158trying to decide whether a patch should be backported to other kernels, bug
159hunters wondering whether the patch is responsible for a problem they are
160chasing, users who want to know how the kernel has changed, and more. A
161good changelog conveys the needed information to all of these people in the
162most direct and concise way possible.
163
164To that end, the summary line should describe the effects of and motivation
165for the change as well as possible given the one-line constraint. The
166detailed description can then amplify on those topics and provide any
167needed additional information. If the patch fixes a bug, cite the commit
Jonathan Corbet5c050fb2011-03-25 12:17:53 -0600168which introduced the bug if possible (and please provide both the commit ID
169and the title when citing commits). If a problem is associated with
Jonathan Corbet5d989322009-04-21 13:33:06 -0600170specific log or compiler output, include that output to help others
171searching for a solution to the same problem. If the change is meant to
172support other changes coming in later patch, say so. If internal APIs are
173changed, detail those changes and how other developers should respond. In
174general, the more you can put yourself into the shoes of everybody who will
175be reading your changelog, the better that changelog (and the kernel as a
176whole) will be.
177
178Needless to say, the changelog should be the text used when committing the
179change to a revision control system. It will be followed by:
Jonathan Corbet75b02142008-09-30 15:15:56 -0600180
181 - The patch itself, in the unified ("-u") patch format. Using the "-p"
182 option to diff will associate function names with changes, making the
183 resulting patch easier for others to read.
184
185You should avoid including changes to irrelevant files (those generated by
186the build process, for example, or editor backup files) in the patch. The
187file "dontdiff" in the Documentation directory can help in this regard;
188pass it to diff with the "-X" option.
189
190The tags mentioned above are used to describe how various developers have
191been associated with the development of this patch. They are described in
192detail in the SubmittingPatches document; what follows here is a brief
193summary. Each of these lines has the format:
194
195 tag: Full Name <email address> optional-other-stuff
196
197The tags in common use are:
198
199 - Signed-off-by: this is a developer's certification that he or she has
200 the right to submit the patch for inclusion into the kernel. It is an
201 agreement to the Developer's Certificate of Origin, the full text of
202 which can be found in Documentation/SubmittingPatches. Code without a
203 proper signoff cannot be merged into the mainline.
204
205 - Acked-by: indicates an agreement by another developer (often a
206 maintainer of the relevant code) that the patch is appropriate for
207 inclusion into the kernel.
208
209 - Tested-by: states that the named person has tested the patch and found
210 it to work.
211
212 - Reviewed-by: the named developer has reviewed the patch for correctness;
213 see the reviewer's statement in Documentation/SubmittingPatches for more
214 detail.
215
216 - Reported-by: names a user who reported a problem which is fixed by this
217 patch; this tag is used to give credit to the (often underappreciated)
218 people who test our code and let us know when things do not work
219 correctly.
220
221 - Cc: the named person received a copy of the patch and had the
222 opportunity to comment on it.
223
224Be careful in the addition of tags to your patches: only Cc: is appropriate
225for addition without the explicit permission of the person named.
226
227
2285.5: SENDING THE PATCH
229
230Before you mail your patches, there are a couple of other things you should
231take care of:
232
233 - Are you sure that your mailer will not corrupt the patches? Patches
234 which have had gratuitous white-space changes or line wrapping performed
235 by the mail client will not apply at the other end, and often will not
236 be examined in any detail. If there is any doubt at all, mail the patch
Jonathan Corbet5c050fb2011-03-25 12:17:53 -0600237 to yourself and convince yourself that it shows up intact.
Jonathan Corbet75b02142008-09-30 15:15:56 -0600238
239 Documentation/email-clients.txt has some helpful hints on making
240 specific mail clients work for sending patches.
241
242 - Are you sure your patch is free of silly mistakes? You should always
243 run patches through scripts/checkpatch.pl and address the complaints it
244 comes up with. Please bear in mind that checkpatch.pl, while being the
245 embodiment of a fair amount of thought about what kernel patches should
246 look like, is not smarter than you. If fixing a checkpatch.pl complaint
247 would make the code worse, don't do it.
248
249Patches should always be sent as plain text. Please do not send them as
250attachments; that makes it much harder for reviewers to quote sections of
251the patch in their replies. Instead, just put the patch directly into your
252message.
253
254When mailing patches, it is important to send copies to anybody who might
255be interested in it. Unlike some other projects, the kernel encourages
256people to err on the side of sending too many copies; don't assume that the
257relevant people will see your posting on the mailing lists. In particular,
258copies should go to:
259
260 - The maintainer(s) of the affected subsystem(s). As described earlier,
261 the MAINTAINERS file is the first place to look for these people.
262
263 - Other developers who have been working in the same area - especially
264 those who might be working there now. Using git to see who else has
265 modified the files you are working on can be helpful.
266
267 - If you are responding to a bug report or a feature request, copy the
268 original poster as well.
269
270 - Send a copy to the relevant mailing list, or, if nothing else applies,
271 the linux-kernel list.
272
273 - If you are fixing a bug, think about whether the fix should go into the
274 next stable update. If so, stable@kernel.org should get a copy of the
275 patch. Also add a "Cc: stable@kernel.org" to the tags within the patch
276 itself; that will cause the stable team to get a notification when your
277 fix goes into the mainline.
278
279When selecting recipients for a patch, it is good to have an idea of who
280you think will eventually accept the patch and get it merged. While it
281is possible to send patches directly to Linus Torvalds and have him merge
282them, things are not normally done that way. Linus is busy, and there are
283subsystem maintainers who watch over specific parts of the kernel. Usually
284you will be wanting that maintainer to merge your patches. If there is no
285obvious maintainer, Andrew Morton is often the patch target of last resort.
286
287Patches need good subject lines. The canonical format for a patch line is
288something like:
289
290 [PATCH nn/mm] subsys: one-line description of the patch
291
292where "nn" is the ordinal number of the patch, "mm" is the total number of
293patches in the series, and "subsys" is the name of the affected subsystem.
Jonathan Corbet5c050fb2011-03-25 12:17:53 -0600294Clearly, nn/mm can be omitted for a single, standalone patch.
Jonathan Corbet75b02142008-09-30 15:15:56 -0600295
296If you have a significant series of patches, it is customary to send an
297introductory description as part zero. This convention is not universally
298followed though; if you use it, remember that information in the
299introduction does not make it into the kernel changelogs. So please ensure
300that the patches, themselves, have complete changelog information.
301
302In general, the second and following parts of a multi-part patch should be
303sent as a reply to the first part so that they all thread together at the
304receiving end. Tools like git and quilt have commands to mail out a set of
305patches with the proper threading. If you have a long series, though, and
Jonathan Corbet5c050fb2011-03-25 12:17:53 -0600306are using git, please stay away from the --chain-reply-to option to avoid
Jonathan Corbet75b02142008-09-30 15:15:56 -0600307creating exceptionally deep nesting.