Commit | Line | Data |
---|---|---|
1da177e4 LT |
1 | |
2 | How to Get Your Change Into the Linux Kernel | |
3 | or | |
4 | Care And Operation Of Your Linus Torvalds | |
5 | ||
6 | ||
7 | ||
8 | For a person or company who wishes to submit a change to the Linux | |
9 | kernel, the process can sometimes be daunting if you're not familiar | |
10 | with "the system." This text is a collection of suggestions which | |
11 | can greatly increase the chances of your change being accepted. | |
12 | ||
bc7455fa RD |
13 | Read Documentation/SubmitChecklist for a list of items to check |
14 | before submitting code. If you are submitting a driver, also read | |
15 | Documentation/SubmittingDrivers. | |
1da177e4 | 16 | |
8e3072a2 JT |
17 | Many of these steps describe the default behavior of the git version |
18 | control system; if you use git to prepare your patches, you'll find much | |
19 | of the mechanical work done for you, though you'll still need to prepare | |
20 | and document a sensible set of patches. | |
1da177e4 LT |
21 | |
22 | -------------------------------------------- | |
23 | SECTION 1 - CREATING AND SENDING YOUR CHANGE | |
24 | -------------------------------------------- | |
25 | ||
26 | ||
7994cc15 JC |
27 | 0) Obtain a current source tree |
28 | ------------------------------- | |
29 | ||
30 | If you do not have a repository with the current kernel source handy, use | |
31 | git to obtain one. You'll want to start with the mainline repository, | |
32 | which can be grabbed with: | |
33 | ||
34 | git clone git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git | |
35 | ||
36 | Note, however, that you may not want to develop against the mainline tree | |
37 | directly. Most subsystem maintainers run their own trees and want to see | |
38 | patches prepared against those trees. See the "T:" entry for the subsystem | |
39 | in the MAINTAINERS file to find that tree, or simply ask the maintainer if | |
40 | the tree is not listed there. | |
41 | ||
42 | It is still possible to download kernel releases via tarballs (as described | |
43 | in the next section), but that is the hard way to do kernel development. | |
1da177e4 LT |
44 | |
45 | 1) "diff -up" | |
46 | ------------ | |
47 | ||
7994cc15 JC |
48 | If you must generate your patches by hand, use "diff -up" or "diff -uprN" |
49 | to create patches. Git generates patches in this form by default; if | |
50 | you're using git, you can skip this section entirely. | |
1da177e4 LT |
51 | |
52 | All changes to the Linux kernel occur in the form of patches, as | |
53 | generated by diff(1). When creating your patch, make sure to create it | |
54 | in "unified diff" format, as supplied by the '-u' argument to diff(1). | |
55 | Also, please use the '-p' argument which shows which C function each | |
56 | change is in - that makes the resultant diff a lot easier to read. | |
57 | Patches should be based in the root kernel source directory, | |
58 | not in any lower subdirectory. | |
59 | ||
60 | To create a patch for a single file, it is often sufficient to do: | |
61 | ||
84da7c08 | 62 | SRCTREE= linux-2.6 |
1da177e4 LT |
63 | MYFILE= drivers/net/mydriver.c |
64 | ||
65 | cd $SRCTREE | |
66 | cp $MYFILE $MYFILE.orig | |
67 | vi $MYFILE # make your change | |
68 | cd .. | |
69 | diff -up $SRCTREE/$MYFILE{.orig,} > /tmp/patch | |
70 | ||
71 | To create a patch for multiple files, you should unpack a "vanilla", | |
72 | or unmodified kernel source tree, and generate a diff against your | |
73 | own source tree. For example: | |
74 | ||
84da7c08 | 75 | MYSRC= /devel/linux-2.6 |
1da177e4 | 76 | |
84da7c08 RD |
77 | tar xvfz linux-2.6.12.tar.gz |
78 | mv linux-2.6.12 linux-2.6.12-vanilla | |
79 | diff -uprN -X linux-2.6.12-vanilla/Documentation/dontdiff \ | |
80 | linux-2.6.12-vanilla $MYSRC > /tmp/patch | |
1da177e4 LT |
81 | |
82 | "dontdiff" is a list of files which are generated by the kernel during | |
83 | the build process, and should be ignored in any diff(1)-generated | |
84da7c08 | 84 | patch. The "dontdiff" file is included in the kernel tree in |
755727b7 | 85 | 2.6.12 and later. |
1da177e4 LT |
86 | |
87 | Make sure your patch does not include any extra files which do not | |
88 | belong in a patch submission. Make sure to review your patch -after- | |
89 | generated it with diff(1), to ensure accuracy. | |
90 | ||
8e3072a2 JT |
91 | If your changes produce a lot of deltas, you need to split them into |
92 | individual patches which modify things in logical stages; see section | |
93 | #3. This will facilitate easier reviewing by other kernel developers, | |
94 | very important if you want your patch accepted. | |
1da177e4 | 95 | |
8e3072a2 JT |
96 | If you're using git, "git rebase -i" can help you with this process. If |
97 | you're not using git, quilt <http://savannah.nongnu.org/projects/quilt> | |
98 | is another popular alternative. | |
84da7c08 RD |
99 | |
100 | ||
1da177e4 LT |
101 | |
102 | 2) Describe your changes. | |
103 | ||
7b9828d4 JW |
104 | Describe your problem. Whether your patch is a one-line bug fix or |
105 | 5000 lines of a new feature, there must be an underlying problem that | |
106 | motivated you to do this work. Convince the reviewer that there is a | |
107 | problem worth fixing and that it makes sense for them to read past the | |
108 | first paragraph. | |
109 | ||
110 | Describe user-visible impact. Straight up crashes and lockups are | |
111 | pretty convincing, but not all bugs are that blatant. Even if the | |
112 | problem was spotted during code review, describe the impact you think | |
113 | it can have on users. Keep in mind that the majority of Linux | |
114 | installations run kernels from secondary stable trees or | |
115 | vendor/product-specific trees that cherry-pick only specific patches | |
116 | from upstream, so include anything that could help route your change | |
117 | downstream: provoking circumstances, excerpts from dmesg, crash | |
118 | descriptions, performance regressions, latency spikes, lockups, etc. | |
119 | ||
120 | Quantify optimizations and trade-offs. If you claim improvements in | |
121 | performance, memory consumption, stack footprint, or binary size, | |
122 | include numbers that back them up. But also describe non-obvious | |
123 | costs. Optimizations usually aren't free but trade-offs between CPU, | |
124 | memory, and readability; or, when it comes to heuristics, between | |
125 | different workloads. Describe the expected downsides of your | |
126 | optimization so that the reviewer can weigh costs against benefits. | |
127 | ||
128 | Once the problem is established, describe what you are actually doing | |
129 | about it in technical detail. It's important to describe the change | |
130 | in plain English for the reviewer to verify that the code is behaving | |
131 | as you intend it to. | |
1da177e4 | 132 | |
2ae19aca TT |
133 | The maintainer will thank you if you write your patch description in a |
134 | form which can be easily pulled into Linux's source code management | |
135 | system, git, as a "commit log". See #15, below. | |
136 | ||
7b9828d4 JW |
137 | Solve only one problem per patch. If your description starts to get |
138 | long, that's a sign that you probably need to split up your patch. | |
139 | See #3, next. | |
1da177e4 | 140 | |
d89b1945 RD |
141 | When you submit or resubmit a patch or patch series, include the |
142 | complete patch description and justification for it. Don't just | |
143 | say that this is version N of the patch (series). Don't expect the | |
144 | patch merger to refer back to earlier patch versions or referenced | |
145 | URLs to find the patch description and put that into the patch. | |
146 | I.e., the patch (series) and its description should be self-contained. | |
147 | This benefits both the patch merger(s) and reviewers. Some reviewers | |
148 | probably didn't even receive earlier versions of the patch. | |
149 | ||
74a475ac JT |
150 | Describe your changes in imperative mood, e.g. "make xyzzy do frotz" |
151 | instead of "[This patch] makes xyzzy do frotz" or "[I] changed xyzzy | |
152 | to do frotz", as if you are giving orders to the codebase to change | |
153 | its behaviour. | |
154 | ||
d89b1945 | 155 | If the patch fixes a logged bug entry, refer to that bug entry by |
9547c706 JT |
156 | number and URL. If the patch follows from a mailing list discussion, |
157 | give a URL to the mailing list archive; use the https://lkml.kernel.org/ | |
158 | redirector with a Message-Id, to ensure that the links cannot become | |
159 | stale. | |
160 | ||
161 | However, try to make your explanation understandable without external | |
162 | resources. In addition to giving a URL to a mailing list archive or | |
163 | bug, summarize the relevant points of the discussion that led to the | |
164 | patch as submitted. | |
1da177e4 | 165 | |
0af52703 GU |
166 | If you want to refer to a specific commit, don't just refer to the |
167 | SHA-1 ID of the commit. Please also include the oneline summary of | |
168 | the commit, to make it easier for reviewers to know what it is about. | |
169 | Example: | |
170 | ||
171 | Commit e21d2170f36602ae2708 ("video: remove unnecessary | |
172 | platform_set_drvdata()") removed the unnecessary | |
173 | platform_set_drvdata(), but left the variable "dev" unused, | |
174 | delete it. | |
175 | ||
7994cc15 JC |
176 | You should also be sure to use at least the first twelve characters of the |
177 | SHA-1 ID. The kernel repository holds a *lot* of objects, making | |
178 | collisions with shorter IDs a real possibility. Bear in mind that, even if | |
179 | there is no collision with your six-character ID now, that condition may | |
180 | change five years from now. | |
181 | ||
8401aa1f JK |
182 | If your patch fixes a bug in a specific commit, e.g. you found an issue using |
183 | git-bisect, please use the 'Fixes:' tag with the first 12 characters of the | |
7994cc15 | 184 | SHA-1 ID, and the one line summary. For example: |
8401aa1f JK |
185 | |
186 | Fixes: e21d2170f366 ("video: remove unnecessary platform_set_drvdata()") | |
187 | ||
188 | The following git-config settings can be used to add a pretty format for | |
189 | outputting the above style in the git log or git show commands | |
190 | ||
191 | [core] | |
192 | abbrev = 12 | |
193 | [pretty] | |
194 | fixes = Fixes: %h (\"%s\") | |
1da177e4 LT |
195 | |
196 | 3) Separate your changes. | |
197 | ||
5b0ed2c6 | 198 | Separate _logical changes_ into a single patch file. |
1da177e4 LT |
199 | |
200 | For example, if your changes include both bug fixes and performance | |
201 | enhancements for a single driver, separate those changes into two | |
202 | or more patches. If your changes include an API update, and a new | |
203 | driver which uses that new API, separate those into two patches. | |
204 | ||
205 | On the other hand, if you make a single change to numerous files, | |
206 | group those changes into a single patch. Thus a single logical change | |
207 | is contained within a single patch. | |
208 | ||
209 | If one patch depends on another patch in order for a change to be | |
210 | complete, that is OK. Simply note "this patch depends on patch X" | |
211 | in your patch description. | |
212 | ||
7994cc15 JC |
213 | When dividing your change into a series of patches, take special care to |
214 | ensure that the kernel builds and runs properly after each patch in the | |
215 | series. Developers using "git bisect" to track down a problem can end up | |
216 | splitting your patch series at any point; they will not thank you if you | |
217 | introduce bugs in the middle. | |
218 | ||
5b0ed2c6 XVP |
219 | If you cannot condense your patch set into a smaller set of patches, |
220 | then only post say 15 or so at a time and wait for review and integration. | |
221 | ||
222 | ||
1da177e4 | 223 | |
6de16eba JC |
224 | 4) Style-check your changes. |
225 | ---------------------------- | |
0a920b5b AW |
226 | |
227 | Check your patch for basic style violations, details of which can be | |
228 | found in Documentation/CodingStyle. Failure to do so simply wastes | |
f56d35e7 | 229 | the reviewers time and will get your patch rejected, probably |
0a920b5b AW |
230 | without even being read. |
231 | ||
6de16eba JC |
232 | One significant exception is when moving code from one file to |
233 | another -- in this case you should not modify the moved code at all in | |
234 | the same patch which moves it. This clearly delineates the act of | |
235 | moving the code and your changes. This greatly aids review of the | |
236 | actual differences and allows tools to better track the history of | |
237 | the code itself. | |
238 | ||
239 | Check your patches with the patch style checker prior to submission | |
240 | (scripts/checkpatch.pl). Note, though, that the style checker should be | |
241 | viewed as a guide, not as a replacement for human judgment. If your code | |
242 | looks better with a violation then its probably best left alone. | |
0a920b5b | 243 | |
6de16eba JC |
244 | The checker reports at three levels: |
245 | - ERROR: things that are very likely to be wrong | |
246 | - WARNING: things requiring careful review | |
247 | - CHECK: things requiring thought | |
248 | ||
249 | You should be able to justify all violations that remain in your | |
250 | patch. | |
0a920b5b AW |
251 | |
252 | ||
ccae8616 JC |
253 | 5) Select the recipients for your patch. |
254 | ---------------------------------------- | |
255 | ||
256 | You should always copy the appropriate subsystem maintainer(s) on any patch | |
257 | to code that they maintain; look through the MAINTAINERS file and the | |
258 | source code revision history to see who those maintainers are. The | |
259 | script scripts/get_maintainer.pl can be very useful at this step. If you | |
260 | cannot find a maintainer for the subsystem your are working on, Andrew | |
261 | Morton (akpm@linux-foundation.org) serves as a maintainer of last resort. | |
262 | ||
263 | You should also normally choose at least one mailing list to receive a copy | |
264 | of your patch set. linux-kernel@vger.kernel.org functions as a list of | |
265 | last resort, but the volume on that list has caused a number of developers | |
266 | to tune it out. Look in the MAINTAINERS file for a subsystem-specific | |
267 | list; your patch will probably get more attention there. Please do not | |
268 | spam unrelated lists, though. | |
269 | ||
270 | Many kernel-related lists are hosted on vger.kernel.org; you can find a | |
271 | list of them at http://vger.kernel.org/vger-lists.html. There are | |
272 | kernel-related lists hosted elsewhere as well, though. | |
5b0ed2c6 XVP |
273 | |
274 | Do not send more than 15 patches at once to the vger mailing lists!!! | |
275 | ||
1da177e4 | 276 | Linus Torvalds is the final arbiter of all changes accepted into the |
99ddcc7e | 277 | Linux kernel. His e-mail address is <torvalds@linux-foundation.org>. |
ccae8616 JC |
278 | He gets a lot of e-mail, and, at this point, very few patches go through |
279 | Linus directly, so typically you should do your best to -avoid- | |
280 | sending him e-mail. | |
1da177e4 | 281 | |
ccae8616 JC |
282 | If you have a patch that fixes an exploitable security bug, send that patch |
283 | to security@kernel.org. For severe bugs, a short embargo may be considered | |
284 | to allow distrbutors to get the patch out to users; in such cases, | |
285 | obviously, the patch should not be sent to any public lists. | |
1da177e4 | 286 | |
ccae8616 JC |
287 | Patches that fix a severe bug in a released kernel should be directed |
288 | toward the stable maintainers by putting a line like this: | |
1da177e4 | 289 | |
ccae8616 | 290 | Cc: stable@vger.kernel.org |
1da177e4 | 291 | |
ccae8616 | 292 | into your patch. |
5b0ed2c6 | 293 | |
ccae8616 JC |
294 | Note, however, that some subsystem maintainers want to come to their own |
295 | conclusions on which patches should go to the stable trees. The networking | |
296 | maintainer, in particular, would rather not see individual developers | |
297 | adding lines like the above to their patches. | |
1caf1f0f | 298 | |
ccae8616 JC |
299 | If changes affect userland-kernel interfaces, please send the MAN-PAGES |
300 | maintainer (as listed in the MAINTAINERS file) a man-pages patch, or at | |
301 | least a notification of the change, so that some information makes its way | |
302 | into the manual pages. User-space API changes should also be copied to | |
303 | linux-api@vger.kernel.org. | |
1da177e4 LT |
304 | |
305 | For small patches you may want to CC the Trivial Patch Monkey | |
82d27b2b MH |
306 | trivial@kernel.org which collects "trivial" patches. Have a look |
307 | into the MAINTAINERS file for its current manager. | |
308 | Trivial patches must qualify for one of the following rules: | |
1da177e4 | 309 | Spelling fixes in documentation |
ccae8616 | 310 | Spelling fixes for errors which could break grep(1) |
1da177e4 LT |
311 | Warning fixes (cluttering with useless warnings is bad) |
312 | Compilation fixes (only if they are actually correct) | |
313 | Runtime fixes (only if they actually fix things) | |
ccae8616 | 314 | Removing use of deprecated functions/macros |
1da177e4 LT |
315 | Contact detail and documentation fixes |
316 | Non-portable code replaced by portable code (even in arch-specific, | |
317 | since people copy, as long as it's trivial) | |
8e9cb8fd | 318 | Any fix by the author/maintainer of the file (ie. patch monkey |
1da177e4 | 319 | in re-transmission mode) |
84da7c08 | 320 | |
1da177e4 LT |
321 | |
322 | ||
ccae8616 | 323 | 6) No MIME, no links, no compression, no attachments. Just plain text. |
1da177e4 LT |
324 | |
325 | Linus and other kernel developers need to be able to read and comment | |
326 | on the changes you are submitting. It is important for a kernel | |
327 | developer to be able to "quote" your changes, using standard e-mail | |
328 | tools, so that they may comment on specific portions of your code. | |
329 | ||
330 | For this reason, all patches should be submitting e-mail "inline". | |
331 | WARNING: Be wary of your editor's word-wrap corrupting your patch, | |
332 | if you choose to cut-n-paste your patch. | |
333 | ||
334 | Do not attach the patch as a MIME attachment, compressed or not. | |
335 | Many popular e-mail applications will not always transmit a MIME | |
336 | attachment as plain text, making it impossible to comment on your | |
337 | code. A MIME attachment also takes Linus a bit more time to process, | |
338 | decreasing the likelihood of your MIME-attached change being accepted. | |
339 | ||
340 | Exception: If your mailer is mangling patches then someone may ask | |
341 | you to re-send them using MIME. | |
342 | ||
097091c0 MO |
343 | See Documentation/email-clients.txt for hints about configuring |
344 | your e-mail client so that it sends your patches untouched. | |
1da177e4 | 345 | |
ccae8616 | 346 | 7) E-mail size. |
1da177e4 | 347 | |
0a920b5b | 348 | When sending patches to Linus, always follow step #7. |
1da177e4 LT |
349 | |
350 | Large changes are not appropriate for mailing lists, and some | |
4932be77 | 351 | maintainers. If your patch, uncompressed, exceeds 300 kB in size, |
1da177e4 LT |
352 | it is preferred that you store your patch on an Internet-accessible |
353 | server, and provide instead a URL (link) pointing to your patch. | |
354 | ||
355 | ||
356 | ||
0eea2314 JC |
357 | 8) Respond to review comments. |
358 | ------------------------------ | |
1da177e4 | 359 | |
0eea2314 JC |
360 | Your patch will almost certainly get comments from reviewers on ways in |
361 | which the patch can be improved. You must respond to those comments; | |
362 | ignoring reviewers is a good way to get ignored in return. Review comments | |
363 | or questions that do not lead to a code change should almost certainly | |
364 | bring about a comment or changelog entry so that the next reviewer better | |
365 | understands what is going on. | |
1da177e4 | 366 | |
0eea2314 JC |
367 | Be sure to tell the reviewers what changes you are making and to thank them |
368 | for their time. Code review is a tiring and time-consuming process, and | |
369 | reviewers sometimes get grumpy. Even in that case, though, respond | |
370 | politely and address the problems they have pointed out. | |
1da177e4 LT |
371 | |
372 | ||
0eea2314 JC |
373 | 9) Don't get discouraged - or impatient. |
374 | ---------------------------------------- | |
1da177e4 | 375 | |
0eea2314 JC |
376 | After you have submitted your change, be patient and wait. Reviewers are |
377 | busy people and may not get to your patch right away. | |
1da177e4 | 378 | |
0eea2314 JC |
379 | Once upon a time, patches used to disappear into the void without comment, |
380 | but the development process works more smoothly than that now. You should | |
381 | receive comments within a week or so; if that does not happen, make sure | |
382 | that you have sent your patches to the right place. Wait for a minimum of | |
383 | one week before resubmitting or pinging reviewers - possibly longer during | |
384 | busy times like merge windows. | |
1da177e4 LT |
385 | |
386 | ||
ccae8616 | 387 | 10) Include PATCH in the subject |
1da177e4 LT |
388 | |
389 | Due to high e-mail traffic to Linus, and to linux-kernel, it is common | |
390 | convention to prefix your subject line with [PATCH]. This lets Linus | |
391 | and other kernel developers more easily distinguish patches from other | |
392 | e-mail discussions. | |
393 | ||
394 | ||
395 | ||
ccae8616 | 396 | 11) Sign your work |
1da177e4 LT |
397 | |
398 | To improve tracking of who did what, especially with patches that can | |
399 | percolate to their final resting place in the kernel through several | |
400 | layers of maintainers, we've introduced a "sign-off" procedure on | |
401 | patches that are being emailed around. | |
402 | ||
403 | The sign-off is a simple line at the end of the explanation for the | |
404 | patch, which certifies that you wrote it or otherwise have the right to | |
db12fb83 | 405 | pass it on as an open-source patch. The rules are pretty simple: if you |
1da177e4 LT |
406 | can certify the below: |
407 | ||
cbd83da8 | 408 | Developer's Certificate of Origin 1.1 |
1da177e4 LT |
409 | |
410 | By making a contribution to this project, I certify that: | |
411 | ||
412 | (a) The contribution was created in whole or in part by me and I | |
413 | have the right to submit it under the open source license | |
414 | indicated in the file; or | |
415 | ||
416 | (b) The contribution is based upon previous work that, to the best | |
417 | of my knowledge, is covered under an appropriate open source | |
418 | license and I have the right under that license to submit that | |
419 | work with modifications, whether created in whole or in part | |
420 | by me, under the same open source license (unless I am | |
421 | permitted to submit under a different license), as indicated | |
422 | in the file; or | |
423 | ||
424 | (c) The contribution was provided directly to me by some other | |
425 | person who certified (a), (b) or (c) and I have not modified | |
426 | it. | |
427 | ||
cbd83da8 LT |
428 | (d) I understand and agree that this project and the contribution |
429 | are public and that a record of the contribution (including all | |
430 | personal information I submit with it, including my sign-off) is | |
431 | maintained indefinitely and may be redistributed consistent with | |
432 | this project or the open source license(s) involved. | |
433 | ||
1da177e4 LT |
434 | then you just add a line saying |
435 | ||
9fd5559c | 436 | Signed-off-by: Random J Developer <random@developer.example.org> |
1da177e4 | 437 | |
af45f32d GK |
438 | using your real name (sorry, no pseudonyms or anonymous contributions.) |
439 | ||
1da177e4 LT |
440 | Some people also put extra tags at the end. They'll just be ignored for |
441 | now, but you can do this to mark internal company procedures or just | |
442 | point out some special detail about the sign-off. | |
443 | ||
adbd5886 WT |
444 | If you are a subsystem or branch maintainer, sometimes you need to slightly |
445 | modify patches you receive in order to merge them, because the code is not | |
446 | exactly the same in your tree and the submitters'. If you stick strictly to | |
447 | rule (c), you should ask the submitter to rediff, but this is a totally | |
448 | counter-productive waste of time and energy. Rule (b) allows you to adjust | |
449 | the code, but then it is very impolite to change one submitter's code and | |
450 | make him endorse your bugs. To solve this problem, it is recommended that | |
451 | you add a line between the last Signed-off-by header and yours, indicating | |
452 | the nature of your changes. While there is nothing mandatory about this, it | |
453 | seems like prepending the description with your mail and/or name, all | |
454 | enclosed in square brackets, is noticeable enough to make it obvious that | |
455 | you are responsible for last-minute changes. Example : | |
456 | ||
457 | Signed-off-by: Random J Developer <random@developer.example.org> | |
458 | [lucky@maintainer.example.org: struct foo moved from foo.c to foo.h] | |
459 | Signed-off-by: Lucky K Maintainer <lucky@maintainer.example.org> | |
460 | ||
305af08c | 461 | This practice is particularly helpful if you maintain a stable branch and |
adbd5886 WT |
462 | want at the same time to credit the author, track changes, merge the fix, |
463 | and protect the submitter from complaints. Note that under no circumstances | |
464 | can you change the author's identity (the From header), as it is the one | |
465 | which appears in the changelog. | |
466 | ||
305af08c | 467 | Special note to back-porters: It seems to be a common and useful practice |
adbd5886 WT |
468 | to insert an indication of the origin of a patch at the top of the commit |
469 | message (just after the subject line) to facilitate tracking. For instance, | |
7994cc15 | 470 | here's what we see in a 3.x-stable release: |
adbd5886 | 471 | |
7994cc15 | 472 | Date: Tue Oct 7 07:26:38 2014 -0400 |
adbd5886 | 473 | |
7994cc15 | 474 | libata: Un-break ATA blacklist |
adbd5886 | 475 | |
7994cc15 | 476 | commit 1c40279960bcd7d52dbdf1d466b20d24b99176c8 upstream. |
adbd5886 | 477 | |
7994cc15 | 478 | And here's what might appear in an older kernel once a patch is backported: |
adbd5886 WT |
479 | |
480 | Date: Tue May 13 22:12:27 2008 +0200 | |
481 | ||
482 | wireless, airo: waitbusy() won't delay | |
483 | ||
484 | [backport of 2.6 commit b7acbdfbd1f277c1eb23f344f899cfa4cd0bf36a] | |
485 | ||
486 | Whatever the format, this information provides a valuable help to people | |
7994cc15 | 487 | tracking your trees, and to people trying to troubleshoot bugs in your |
adbd5886 WT |
488 | tree. |
489 | ||
1da177e4 | 490 | |
ccae8616 | 491 | 12) When to use Acked-by: and Cc: |
0a920b5b | 492 | |
0f44cd23 AM |
493 | The Signed-off-by: tag indicates that the signer was involved in the |
494 | development of the patch, or that he/she was in the patch's delivery path. | |
495 | ||
496 | If a person was not directly involved in the preparation or handling of a | |
497 | patch but wishes to signify and record their approval of it then they can | |
498 | arrange to have an Acked-by: line added to the patch's changelog. | |
499 | ||
500 | Acked-by: is often used by the maintainer of the affected code when that | |
501 | maintainer neither contributed to nor forwarded the patch. | |
502 | ||
503 | Acked-by: is not as formal as Signed-off-by:. It is a record that the acker | |
504 | has at least reviewed the patch and has indicated acceptance. Hence patch | |
505 | mergers will sometimes manually convert an acker's "yep, looks good to me" | |
506 | into an Acked-by:. | |
507 | ||
508 | Acked-by: does not necessarily indicate acknowledgement of the entire patch. | |
509 | For example, if a patch affects multiple subsystems and has an Acked-by: from | |
510 | one subsystem maintainer then this usually indicates acknowledgement of just | |
511 | the part which affects that maintainer's code. Judgement should be used here. | |
ef40203a | 512 | When in doubt people should refer to the original discussion in the mailing |
0f44cd23 AM |
513 | list archives. |
514 | ||
ef40203a JC |
515 | If a person has had the opportunity to comment on a patch, but has not |
516 | provided such comments, you may optionally add a "Cc:" tag to the patch. | |
517 | This is the only tag which might be added without an explicit action by the | |
518 | person it names. This tag documents that potentially interested parties | |
519 | have been included in the discussion | |
0f44cd23 | 520 | |
ef40203a | 521 | |
ccae8616 | 522 | 13) Using Reported-by:, Tested-by:, Reviewed-by:, Suggested-by: and Fixes: |
bbb0a424 | 523 | |
d75ef707 DC |
524 | The Reported-by tag gives credit to people who find bugs and report them and it |
525 | hopefully inspires them to help us again in the future. Please note that if | |
526 | the bug was reported in private, then ask for permission first before using the | |
527 | Reported-by tag. | |
ef40203a JC |
528 | |
529 | A Tested-by: tag indicates that the patch has been successfully tested (in | |
530 | some environment) by the person named. This tag informs maintainers that | |
531 | some testing has been performed, provides a means to locate testers for | |
532 | future patches, and ensures credit for the testers. | |
533 | ||
534 | Reviewed-by:, instead, indicates that the patch has been reviewed and found | |
535 | acceptable according to the Reviewer's Statement: | |
536 | ||
537 | Reviewer's statement of oversight | |
538 | ||
539 | By offering my Reviewed-by: tag, I state that: | |
540 | ||
541 | (a) I have carried out a technical review of this patch to | |
542 | evaluate its appropriateness and readiness for inclusion into | |
543 | the mainline kernel. | |
544 | ||
545 | (b) Any problems, concerns, or questions relating to the patch | |
546 | have been communicated back to the submitter. I am satisfied | |
547 | with the submitter's response to my comments. | |
548 | ||
549 | (c) While there may be things that could be improved with this | |
550 | submission, I believe that it is, at this time, (1) a | |
551 | worthwhile modification to the kernel, and (2) free of known | |
552 | issues which would argue against its inclusion. | |
553 | ||
554 | (d) While I have reviewed the patch and believe it to be sound, I | |
555 | do not (unless explicitly stated elsewhere) make any | |
556 | warranties or guarantees that it will achieve its stated | |
557 | purpose or function properly in any given situation. | |
558 | ||
559 | A Reviewed-by tag is a statement of opinion that the patch is an | |
560 | appropriate modification of the kernel without any remaining serious | |
561 | technical issues. Any interested reviewer (who has done the work) can | |
562 | offer a Reviewed-by tag for a patch. This tag serves to give credit to | |
563 | reviewers and to inform maintainers of the degree of review which has been | |
564 | done on the patch. Reviewed-by: tags, when supplied by reviewers known to | |
565 | understand the subject area and to perform thorough reviews, will normally | |
5801da1b | 566 | increase the likelihood of your patch getting into the kernel. |
ef40203a | 567 | |
8543ae12 M |
568 | A Suggested-by: tag indicates that the patch idea is suggested by the person |
569 | named and ensures credit to the person for the idea. Please note that this | |
570 | tag should not be added without the reporter's permission, especially if the | |
571 | idea was not posted in a public forum. That said, if we diligently credit our | |
572 | idea reporters, they will, hopefully, be inspired to help us again in the | |
573 | future. | |
574 | ||
8401aa1f JK |
575 | A Fixes: tag indicates that the patch fixes an issue in a previous commit. It |
576 | is used to make it easy to determine where a bug originated, which can help | |
577 | review a bug fix. This tag also assists the stable kernel team in determining | |
578 | which stable kernel versions should receive your fix. This is the preferred | |
579 | method for indicating a bug fixed by the patch. See #2 above for more details. | |
580 | ||
ef40203a | 581 | |
ccae8616 | 582 | 14) The canonical patch format |
7994cc15 JC |
583 | ------------------------------ |
584 | ||
585 | This section describes how the patch itself should be formatted. Note | |
586 | that, if you have your patches stored in a git repository, proper patch | |
587 | formatting can be had with "git format-patch". The tools cannot create | |
588 | the necessary text, though, so read the instructions below anyway. | |
84da7c08 | 589 | |
75f8426c PJ |
590 | The canonical patch subject line is: |
591 | ||
d6b9acc0 | 592 | Subject: [PATCH 001/123] subsystem: summary phrase |
75f8426c PJ |
593 | |
594 | The canonical patch message body contains the following: | |
595 | ||
ccae8616 JC |
596 | - A "from" line specifying the patch author (only needed if the person |
597 | sending the patch is not the author). | |
75f8426c PJ |
598 | |
599 | - An empty line. | |
600 | ||
601 | - The body of the explanation, which will be copied to the | |
602 | permanent changelog to describe this patch. | |
603 | ||
604 | - The "Signed-off-by:" lines, described above, which will | |
605 | also go in the changelog. | |
606 | ||
607 | - A marker line containing simply "---". | |
608 | ||
609 | - Any additional comments not suitable for the changelog. | |
610 | ||
611 | - The actual patch (diff output). | |
612 | ||
613 | The Subject line format makes it very easy to sort the emails | |
614 | alphabetically by subject line - pretty much any email reader will | |
615 | support that - since because the sequence number is zero-padded, | |
616 | the numerical and alphabetic sort is the same. | |
617 | ||
d6b9acc0 PJ |
618 | The "subsystem" in the email's Subject should identify which |
619 | area or subsystem of the kernel is being patched. | |
620 | ||
621 | The "summary phrase" in the email's Subject should concisely | |
622 | describe the patch which that email contains. The "summary | |
623 | phrase" should not be a filename. Do not use the same "summary | |
66effdc6 RD |
624 | phrase" for every patch in a whole patch series (where a "patch |
625 | series" is an ordered sequence of multiple, related patches). | |
d6b9acc0 | 626 | |
2ae19aca TT |
627 | Bear in mind that the "summary phrase" of your email becomes a |
628 | globally-unique identifier for that patch. It propagates all the way | |
629 | into the git changelog. The "summary phrase" may later be used in | |
630 | developer discussions which refer to the patch. People will want to | |
631 | google for the "summary phrase" to read discussion regarding that | |
632 | patch. It will also be the only thing that people may quickly see | |
633 | when, two or three months later, they are going through perhaps | |
634 | thousands of patches using tools such as "gitk" or "git log | |
635 | --oneline". | |
636 | ||
637 | For these reasons, the "summary" must be no more than 70-75 | |
638 | characters, and it must describe both what the patch changes, as well | |
639 | as why the patch might be necessary. It is challenging to be both | |
640 | succinct and descriptive, but that is what a well-written summary | |
641 | should do. | |
642 | ||
643 | The "summary phrase" may be prefixed by tags enclosed in square | |
644 | brackets: "Subject: [PATCH tag] <summary phrase>". The tags are not | |
645 | considered part of the summary phrase, but describe how the patch | |
646 | should be treated. Common tags might include a version descriptor if | |
647 | the multiple versions of the patch have been sent out in response to | |
648 | comments (i.e., "v1, v2, v3"), or "RFC" to indicate a request for | |
649 | comments. If there are four patches in a patch series the individual | |
650 | patches may be numbered like this: 1/4, 2/4, 3/4, 4/4. This assures | |
651 | that developers understand the order in which the patches should be | |
652 | applied and that they have reviewed or applied all of the patches in | |
653 | the patch series. | |
d6b9acc0 PJ |
654 | |
655 | A couple of example Subjects: | |
656 | ||
657 | Subject: [patch 2/5] ext2: improve scalability of bitmap searching | |
658 | Subject: [PATCHv2 001/207] x86: fix eflags tracking | |
75f8426c PJ |
659 | |
660 | The "from" line must be the very first line in the message body, | |
661 | and has the form: | |
662 | ||
663 | From: Original Author <author@example.com> | |
664 | ||
665 | The "from" line specifies who will be credited as the author of the | |
666 | patch in the permanent changelog. If the "from" line is missing, | |
667 | then the "From:" line from the email header will be used to determine | |
668 | the patch author in the changelog. | |
669 | ||
670 | The explanation body will be committed to the permanent source | |
671 | changelog, so should make sense to a competent reader who has long | |
672 | since forgotten the immediate details of the discussion that might | |
2ae19aca TT |
673 | have led to this patch. Including symptoms of the failure which the |
674 | patch addresses (kernel log messages, oops messages, etc.) is | |
675 | especially useful for people who might be searching the commit logs | |
676 | looking for the applicable patch. If a patch fixes a compile failure, | |
677 | it may not be necessary to include _all_ of the compile failures; just | |
678 | enough that it is likely that someone searching for the patch can find | |
679 | it. As in the "summary phrase", it is important to be both succinct as | |
680 | well as descriptive. | |
75f8426c PJ |
681 | |
682 | The "---" marker line serves the essential purpose of marking for patch | |
683 | handling tools where the changelog message ends. | |
684 | ||
685 | One good use for the additional comments after the "---" marker is for | |
2ae19aca TT |
686 | a diffstat, to show what files have changed, and the number of |
687 | inserted and deleted lines per file. A diffstat is especially useful | |
688 | on bigger patches. Other comments relevant only to the moment or the | |
689 | maintainer, not suitable for the permanent changelog, should also go | |
690 | here. A good example of such comments might be "patch changelogs" | |
691 | which describe what has changed between the v1 and v2 version of the | |
692 | patch. | |
693 | ||
694 | If you are going to include a diffstat after the "---" marker, please | |
695 | use diffstat options "-p 1 -w 70" so that filenames are listed from | |
696 | the top of the kernel source tree and don't use too much horizontal | |
8e3072a2 JT |
697 | space (easily fit in 80 columns, maybe with some indentation). (git |
698 | generates appropriate diffstats by default.) | |
75f8426c PJ |
699 | |
700 | See more details on the proper patch format in the following | |
701 | references. | |
702 | ||
703 | ||
ccae8616 | 704 | 15) Sending "git pull" requests |
7994cc15 JC |
705 | ------------------------------- |
706 | ||
707 | If you have a series of patches, it may be most convenient to have the | |
708 | maintainer pull them directly into the subsystem repository with a | |
709 | "git pull" operation. Note, however, that pulling patches from a developer | |
710 | requires a higher degree of trust than taking patches from a mailing list. | |
711 | As a result, many subsystem maintainers are reluctant to take pull | |
712 | requests, especially from new, unknown developers. | |
713 | ||
714 | A pull request should have [GIT] or [PULL] in the subject line. The | |
715 | request itself should include the repository name and the branch of | |
716 | interest on a single line; it should look something like: | |
717 | ||
718 | Please pull from | |
84da7c08 | 719 | |
7994cc15 | 720 | git://jdelvare.pck.nerim.net/jdelvare-2.6 i2c-for-linus |
14863617 | 721 | |
7994cc15 | 722 | to get these changes:" |
14863617 | 723 | |
7994cc15 JC |
724 | A pull request should also include an overall message saying what will be |
725 | included in the request, a "git shortlog" listing of the patches | |
726 | themselves, and a diffstat showing the overall effect of the patch series. | |
727 | The easiest way to get all this information together is, of course, to let | |
728 | git do it for you with the "git request-pull" command. | |
14863617 | 729 | |
7994cc15 JC |
730 | Some maintainers (including Linus) want to see pull requests from signed |
731 | commits; that increases their confidence that the request actually came | |
732 | from you. Linus, in particular, will not pull from public hosting sites | |
733 | like GitHub in the absence of a signed tag. | |
14863617 | 734 | |
7994cc15 JC |
735 | The first step toward creating such tags is to make a GNUPG key and get it |
736 | signed by one or more core kernel developers. This step can be hard for | |
737 | new developers, but there is no way around it. Attending conferences can | |
738 | be a good way to find developers who can sign your key. | |
14863617 | 739 | |
7994cc15 JC |
740 | Once you have prepared a patch series in git that you wish to have somebody |
741 | pull, create a signed tag with "git tag -s". This will create a new tag | |
742 | identifying the last commit in the series and containing a signature | |
743 | created with your private key. You will also have the opportunity to add a | |
744 | changelog-style message to the tag; this is an ideal place to describe the | |
745 | effects of the pull request as a whole. | |
14863617 | 746 | |
7994cc15 JC |
747 | If the tree the maintainer will be pulling from is not the repository you |
748 | are working from, don't forget to push the signed tag explicitly to the | |
749 | public tree. | |
14863617 | 750 | |
7994cc15 JC |
751 | When generating your pull request, use the signed tag as the target. A |
752 | command like this will do the trick: | |
14863617 | 753 | |
7994cc15 | 754 | git request-pull master git://my.public.tree/linux.git my-signed-tag |
84da7c08 | 755 | |
5b0ed2c6 XVP |
756 | |
757 | ---------------------- | |
6de16eba | 758 | SECTION 2 - REFERENCES |
5b0ed2c6 XVP |
759 | ---------------------- |
760 | ||
761 | Andrew Morton, "The perfect patch" (tpp). | |
37c703f4 | 762 | <http://www.ozlabs.org/~akpm/stuff/tpp.txt> |
5b0ed2c6 | 763 | |
8e9cb8fd | 764 | Jeff Garzik, "Linux kernel patch submission format". |
5b0ed2c6 XVP |
765 | <http://linux.yyz.us/patch-format.html> |
766 | ||
8e9cb8fd | 767 | Greg Kroah-Hartman, "How to piss off a kernel subsystem maintainer". |
f5039935 VN |
768 | <http://www.kroah.com/log/linux/maintainer.html> |
769 | <http://www.kroah.com/log/linux/maintainer-02.html> | |
770 | <http://www.kroah.com/log/linux/maintainer-03.html> | |
771 | <http://www.kroah.com/log/linux/maintainer-04.html> | |
772 | <http://www.kroah.com/log/linux/maintainer-05.html> | |
7e0dae61 | 773 | <http://www.kroah.com/log/linux/maintainer-06.html> |
5b0ed2c6 | 774 | |
bc7455fa | 775 | NO!!!! No more huge patch bombs to linux-kernel@vger.kernel.org people! |
37c703f4 | 776 | <https://lkml.org/lkml/2005/7/11/336> |
5b0ed2c6 | 777 | |
8e9cb8fd | 778 | Kernel Documentation/CodingStyle: |
4db29c17 | 779 | <http://users.sosdg.org/~qiyong/lxr/source/Documentation/CodingStyle> |
5b0ed2c6 | 780 | |
8e9cb8fd | 781 | Linus Torvalds's mail on the canonical patch format: |
5b0ed2c6 | 782 | <http://lkml.org/lkml/2005/4/7/183> |
9536727e AK |
783 | |
784 | Andi Kleen, "On submitting kernel patches" | |
25985edc | 785 | Some strategies to get difficult or controversial changes in. |
9536727e AK |
786 | http://halobates.de/on-submitting-patches.pdf |
787 | ||
5b0ed2c6 | 788 | -- |