Commit | Line | Data |
---|---|---|
609d99a3 | 1 | .. _submittingpatches: |
1da177e4 | 2 | |
89edeedd JC |
3 | Submitting patches: the essential guide to getting your code into the kernel |
4 | ============================================================================ | |
1da177e4 LT |
5 | |
6 | For a person or company who wishes to submit a change to the Linux | |
7 | kernel, the process can sometimes be daunting if you're not familiar | |
8 | with "the system." This text is a collection of suggestions which | |
9 | can greatly increase the chances of your change being accepted. | |
10 | ||
d00c4559 JC |
11 | This document contains a large number of suggestions in a relatively terse |
12 | format. For detailed information on how the kernel development process | |
9912d0bb MCC |
13 | works, see Documentation/process/development-process.rst. Also, read |
14 | Documentation/process/submit-checklist.rst | |
5ff4aa70 | 15 | for a list of items to check before submitting code. If you are submitting |
9912d0bb | 16 | a driver, also read Documentation/process/submitting-drivers.rst; for device |
aa9b5e0d EE |
17 | tree binding patches, read |
18 | Documentation/devicetree/bindings/submitting-patches.rst. | |
1da177e4 | 19 | |
9f364b60 DD |
20 | This documentation assumes that you're using ``git`` to prepare your patches. |
21 | If you're unfamiliar with ``git``, you would be well-advised to learn how to | |
22 | use it, it will make your life as a kernel developer and in general much | |
23 | easier. | |
1da177e4 | 24 | |
604370e1 | 25 | Some subsystems and maintainer trees have additional information about |
b96ff02a MCC |
26 | their workflow and expectations, see |
27 | :ref:`Documentation/process/maintainer-handbooks.rst <maintainer_handbooks_main>`. | |
604370e1 | 28 | |
ef227c39 DD |
29 | Obtain a current source tree |
30 | ---------------------------- | |
7994cc15 JC |
31 | |
32 | If you do not have a repository with the current kernel source handy, use | |
9b2c7677 | 33 | ``git`` to obtain one. You'll want to start with the mainline repository, |
5903019b | 34 | which can be grabbed with:: |
7994cc15 | 35 | |
5903019b | 36 | git clone git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git |
7994cc15 JC |
37 | |
38 | Note, however, that you may not want to develop against the mainline tree | |
39 | directly. Most subsystem maintainers run their own trees and want to see | |
5903019b | 40 | patches prepared against those trees. See the **T:** entry for the subsystem |
7994cc15 JC |
41 | in the MAINTAINERS file to find that tree, or simply ask the maintainer if |
42 | the tree is not listed there. | |
43 | ||
5903019b | 44 | .. _describe_changes: |
84da7c08 | 45 | |
ef227c39 DD |
46 | Describe your changes |
47 | --------------------- | |
1da177e4 | 48 | |
7b9828d4 JW |
49 | Describe your problem. Whether your patch is a one-line bug fix or |
50 | 5000 lines of a new feature, there must be an underlying problem that | |
51 | motivated you to do this work. Convince the reviewer that there is a | |
52 | problem worth fixing and that it makes sense for them to read past the | |
53 | first paragraph. | |
54 | ||
55 | Describe user-visible impact. Straight up crashes and lockups are | |
56 | pretty convincing, but not all bugs are that blatant. Even if the | |
57 | problem was spotted during code review, describe the impact you think | |
58 | it can have on users. Keep in mind that the majority of Linux | |
59 | installations run kernels from secondary stable trees or | |
60 | vendor/product-specific trees that cherry-pick only specific patches | |
61 | from upstream, so include anything that could help route your change | |
62 | downstream: provoking circumstances, excerpts from dmesg, crash | |
63 | descriptions, performance regressions, latency spikes, lockups, etc. | |
64 | ||
65 | Quantify optimizations and trade-offs. If you claim improvements in | |
66 | performance, memory consumption, stack footprint, or binary size, | |
67 | include numbers that back them up. But also describe non-obvious | |
68 | costs. Optimizations usually aren't free but trade-offs between CPU, | |
69 | memory, and readability; or, when it comes to heuristics, between | |
70 | different workloads. Describe the expected downsides of your | |
71 | optimization so that the reviewer can weigh costs against benefits. | |
72 | ||
73 | Once the problem is established, describe what you are actually doing | |
74 | about it in technical detail. It's important to describe the change | |
75 | in plain English for the reviewer to verify that the code is behaving | |
76 | as you intend it to. | |
1da177e4 | 77 | |
2ae19aca TT |
78 | The maintainer will thank you if you write your patch description in a |
79 | form which can be easily pulled into Linux's source code management | |
6d5aa418 | 80 | system, ``git``, as a "commit log". See :ref:`the_canonical_patch_format`. |
2ae19aca | 81 | |
7b9828d4 JW |
82 | Solve only one problem per patch. If your description starts to get |
83 | long, that's a sign that you probably need to split up your patch. | |
5903019b | 84 | See :ref:`split_changes`. |
1da177e4 | 85 | |
d89b1945 RD |
86 | When you submit or resubmit a patch or patch series, include the |
87 | complete patch description and justification for it. Don't just | |
88 | say that this is version N of the patch (series). Don't expect the | |
d00c4559 | 89 | subsystem maintainer to refer back to earlier patch versions or referenced |
d89b1945 RD |
90 | URLs to find the patch description and put that into the patch. |
91 | I.e., the patch (series) and its description should be self-contained. | |
d00c4559 | 92 | This benefits both the maintainers and reviewers. Some reviewers |
d89b1945 RD |
93 | probably didn't even receive earlier versions of the patch. |
94 | ||
74a475ac JT |
95 | Describe your changes in imperative mood, e.g. "make xyzzy do frotz" |
96 | instead of "[This patch] makes xyzzy do frotz" or "[I] changed xyzzy | |
97 | to do frotz", as if you are giving orders to the codebase to change | |
98 | its behaviour. | |
99 | ||
0af52703 GU |
100 | If you want to refer to a specific commit, don't just refer to the |
101 | SHA-1 ID of the commit. Please also include the oneline summary of | |
102 | the commit, to make it easier for reviewers to know what it is about. | |
5903019b | 103 | Example:: |
0af52703 GU |
104 | |
105 | Commit e21d2170f36602ae2708 ("video: remove unnecessary | |
106 | platform_set_drvdata()") removed the unnecessary | |
107 | platform_set_drvdata(), but left the variable "dev" unused, | |
108 | delete it. | |
109 | ||
7994cc15 JC |
110 | You should also be sure to use at least the first twelve characters of the |
111 | SHA-1 ID. The kernel repository holds a *lot* of objects, making | |
112 | collisions with shorter IDs a real possibility. Bear in mind that, even if | |
113 | there is no collision with your six-character ID now, that condition may | |
114 | change five years from now. | |
115 | ||
1f57bd42 TL |
116 | If related discussions or any other background information behind the change |
117 | can be found on the web, add 'Link:' tags pointing to it. In case your patch | |
118 | fixes a bug, for example, add a tag with a URL referencing the report in the | |
119 | mailing list archives or a bug tracker; if the patch is a result of some | |
120 | earlier mailing list discussion or something documented on the web, point to | |
121 | it. | |
122 | ||
123 | When linking to mailing list archives, preferably use the lore.kernel.org | |
124 | message archiver service. To create the link URL, use the contents of the | |
125 | ``Message-Id`` header of the message without the surrounding angle brackets. | |
126 | For example:: | |
127 | ||
128 | Link: https://lore.kernel.org/r/30th.anniversary.repost@klaava.Helsinki.FI/ | |
129 | ||
130 | Please check the link to make sure that it is actually working and points | |
131 | to the relevant message. | |
132 | ||
133 | However, try to make your explanation understandable without external | |
134 | resources. In addition to giving a URL to a mailing list archive or bug, | |
135 | summarize the relevant points of the discussion that led to the | |
136 | patch as submitted. | |
137 | ||
8401aa1f | 138 | If your patch fixes a bug in a specific commit, e.g. you found an issue using |
9b2c7677 | 139 | ``git bisect``, please use the 'Fixes:' tag with the first 12 characters of |
19c3fe28 SC |
140 | the SHA-1 ID, and the one line summary. Do not split the tag across multiple |
141 | lines, tags are exempt from the "wrap at 75 columns" rule in order to simplify | |
142 | parsing scripts. For example:: | |
8401aa1f | 143 | |
19c3fe28 | 144 | Fixes: 54a4f0239f2e ("KVM: MMU: make kvm_mmu_zap_page() return the number of pages it actually freed") |
8401aa1f | 145 | |
9b2c7677 MCC |
146 | The following ``git config`` settings can be used to add a pretty format for |
147 | outputting the above style in the ``git log`` or ``git show`` commands:: | |
8401aa1f JK |
148 | |
149 | [core] | |
150 | abbrev = 12 | |
151 | [pretty] | |
152 | fixes = Fixes: %h (\"%s\") | |
1da177e4 | 153 | |
5b5bbb8c TR |
154 | An example call:: |
155 | ||
156 | $ git log -1 --pretty=fixes 54a4f0239f2e | |
157 | Fixes: 54a4f0239f2e ("KVM: MMU: make kvm_mmu_zap_page() return the number of pages it actually freed") | |
158 | ||
5903019b MCC |
159 | .. _split_changes: |
160 | ||
ef227c39 DD |
161 | Separate your changes |
162 | --------------------- | |
1da177e4 | 163 | |
5903019b | 164 | Separate each **logical change** into a separate patch. |
1da177e4 LT |
165 | |
166 | For example, if your changes include both bug fixes and performance | |
167 | enhancements for a single driver, separate those changes into two | |
168 | or more patches. If your changes include an API update, and a new | |
169 | driver which uses that new API, separate those into two patches. | |
170 | ||
171 | On the other hand, if you make a single change to numerous files, | |
172 | group those changes into a single patch. Thus a single logical change | |
173 | is contained within a single patch. | |
174 | ||
d00c4559 JC |
175 | The point to remember is that each patch should make an easily understood |
176 | change that can be verified by reviewers. Each patch should be justifiable | |
177 | on its own merits. | |
178 | ||
1da177e4 | 179 | If one patch depends on another patch in order for a change to be |
5903019b | 180 | complete, that is OK. Simply note **"this patch depends on patch X"** |
1da177e4 LT |
181 | in your patch description. |
182 | ||
7994cc15 JC |
183 | When dividing your change into a series of patches, take special care to |
184 | ensure that the kernel builds and runs properly after each patch in the | |
5903019b | 185 | series. Developers using ``git bisect`` to track down a problem can end up |
7994cc15 JC |
186 | splitting your patch series at any point; they will not thank you if you |
187 | introduce bugs in the middle. | |
188 | ||
5b0ed2c6 XVP |
189 | If you cannot condense your patch set into a smaller set of patches, |
190 | then only post say 15 or so at a time and wait for review and integration. | |
191 | ||
192 | ||
1da177e4 | 193 | |
ef227c39 DD |
194 | Style-check your changes |
195 | ------------------------ | |
0a920b5b AW |
196 | |
197 | Check your patch for basic style violations, details of which can be | |
9912d0bb | 198 | found in Documentation/process/coding-style.rst. |
dca22a63 | 199 | Failure to do so simply wastes |
f56d35e7 | 200 | the reviewers time and will get your patch rejected, probably |
0a920b5b AW |
201 | without even being read. |
202 | ||
6de16eba JC |
203 | One significant exception is when moving code from one file to |
204 | another -- in this case you should not modify the moved code at all in | |
205 | the same patch which moves it. This clearly delineates the act of | |
206 | moving the code and your changes. This greatly aids review of the | |
207 | actual differences and allows tools to better track the history of | |
208 | the code itself. | |
209 | ||
210 | Check your patches with the patch style checker prior to submission | |
211 | (scripts/checkpatch.pl). Note, though, that the style checker should be | |
212 | viewed as a guide, not as a replacement for human judgment. If your code | |
213 | looks better with a violation then its probably best left alone. | |
0a920b5b | 214 | |
6de16eba JC |
215 | The checker reports at three levels: |
216 | - ERROR: things that are very likely to be wrong | |
217 | - WARNING: things requiring careful review | |
218 | - CHECK: things requiring thought | |
0a920b5b | 219 | |
6de16eba JC |
220 | You should be able to justify all violations that remain in your |
221 | patch. | |
0a920b5b AW |
222 | |
223 | ||
ef227c39 DD |
224 | Select the recipients for your patch |
225 | ------------------------------------ | |
1da177e4 | 226 | |
ccae8616 JC |
227 | You should always copy the appropriate subsystem maintainer(s) on any patch |
228 | to code that they maintain; look through the MAINTAINERS file and the | |
229 | source code revision history to see who those maintainers are. The | |
f1a69399 KK |
230 | script scripts/get_maintainer.pl can be very useful at this step (pass paths to |
231 | your patches as arguments to scripts/get_maintainer.pl). If you cannot find a | |
232 | maintainer for the subsystem you are working on, Andrew Morton | |
233 | (akpm@linux-foundation.org) serves as a maintainer of last resort. | |
1da177e4 | 234 | |
ccae8616 | 235 | You should also normally choose at least one mailing list to receive a copy |
77167b96 HH |
236 | of your patch set. linux-kernel@vger.kernel.org should be used by default |
237 | for all patches, but the volume on that list has caused a number of | |
238 | developers to tune it out. Look in the MAINTAINERS file for a | |
239 | subsystem-specific list; your patch will probably get more attention there. | |
240 | Please do not spam unrelated lists, though. | |
1da177e4 | 241 | |
ccae8616 JC |
242 | Many kernel-related lists are hosted on vger.kernel.org; you can find a |
243 | list of them at http://vger.kernel.org/vger-lists.html. There are | |
244 | kernel-related lists hosted elsewhere as well, though. | |
5b0ed2c6 XVP |
245 | |
246 | Do not send more than 15 patches at once to the vger mailing lists!!! | |
247 | ||
1da177e4 | 248 | Linus Torvalds is the final arbiter of all changes accepted into the |
e00bfcbf | 249 | Linux kernel. His e-mail address is <torvalds@linux-foundation.org>. |
ccae8616 JC |
250 | He gets a lot of e-mail, and, at this point, very few patches go through |
251 | Linus directly, so typically you should do your best to -avoid- | |
e00bfcbf | 252 | sending him e-mail. |
1da177e4 | 253 | |
ccae8616 JC |
254 | If you have a patch that fixes an exploitable security bug, send that patch |
255 | to security@kernel.org. For severe bugs, a short embargo may be considered | |
253508ca | 256 | to allow distributors to get the patch out to users; in such cases, |
eb45fb2f | 257 | obviously, the patch should not be sent to any public lists. See also |
9912d0bb | 258 | Documentation/admin-guide/security-bugs.rst. |
1da177e4 | 259 | |
ccae8616 | 260 | Patches that fix a severe bug in a released kernel should be directed |
5903019b | 261 | toward the stable maintainers by putting a line like this:: |
1da177e4 | 262 | |
ccae8616 | 263 | Cc: stable@vger.kernel.org |
1da177e4 | 264 | |
8cda4c3a | 265 | into the sign-off area of your patch (note, NOT an email recipient). You |
9912d0bb MCC |
266 | should also read Documentation/process/stable-kernel-rules.rst |
267 | in addition to this document. | |
5b0ed2c6 | 268 | |
ccae8616 JC |
269 | If changes affect userland-kernel interfaces, please send the MAN-PAGES |
270 | maintainer (as listed in the MAINTAINERS file) a man-pages patch, or at | |
271 | least a notification of the change, so that some information makes its way | |
272 | into the manual pages. User-space API changes should also be copied to | |
5903019b | 273 | linux-api@vger.kernel.org. |
1da177e4 | 274 | |
1da177e4 | 275 | |
ef227c39 DD |
276 | No MIME, no links, no compression, no attachments. Just plain text |
277 | ------------------------------------------------------------------- | |
1da177e4 LT |
278 | |
279 | Linus and other kernel developers need to be able to read and comment | |
280 | on the changes you are submitting. It is important for a kernel | |
281 | developer to be able to "quote" your changes, using standard e-mail | |
282 | tools, so that they may comment on specific portions of your code. | |
283 | ||
9f364b60 DD |
284 | For this reason, all patches should be submitted by e-mail "inline". The |
285 | easiest way to do this is with ``git send-email``, which is strongly | |
286 | recommended. An interactive tutorial for ``git send-email`` is available at | |
287 | https://git-send-email.io. | |
288 | ||
289 | If you choose not to use ``git send-email``: | |
9b2c7677 MCC |
290 | |
291 | .. warning:: | |
292 | ||
293 | Be wary of your editor's word-wrap corrupting your patch, | |
294 | if you choose to cut-n-paste your patch. | |
1da177e4 LT |
295 | |
296 | Do not attach the patch as a MIME attachment, compressed or not. | |
297 | Many popular e-mail applications will not always transmit a MIME | |
298 | attachment as plain text, making it impossible to comment on your | |
299 | code. A MIME attachment also takes Linus a bit more time to process, | |
300 | decreasing the likelihood of your MIME-attached change being accepted. | |
301 | ||
302 | Exception: If your mailer is mangling patches then someone may ask | |
303 | you to re-send them using MIME. | |
304 | ||
9912d0bb MCC |
305 | See Documentation/process/email-clients.rst for hints about configuring |
306 | your e-mail client so that it sends your patches untouched. | |
1da177e4 | 307 | |
ef227c39 DD |
308 | Respond to review comments |
309 | -------------------------- | |
1da177e4 | 310 | |
0eea2314 | 311 | Your patch will almost certainly get comments from reviewers on ways in |
9f364b60 DD |
312 | which the patch can be improved, in the form of a reply to your email. You must |
313 | respond to those comments; ignoring reviewers is a good way to get ignored in | |
314 | return. You can simply reply to their emails to answer their comments. Review | |
315 | comments or questions that do not lead to a code change should almost certainly | |
0eea2314 JC |
316 | bring about a comment or changelog entry so that the next reviewer better |
317 | understands what is going on. | |
1da177e4 | 318 | |
0eea2314 JC |
319 | Be sure to tell the reviewers what changes you are making and to thank them |
320 | for their time. Code review is a tiring and time-consuming process, and | |
321 | reviewers sometimes get grumpy. Even in that case, though, respond | |
0c603a5c KK |
322 | politely and address the problems they have pointed out. When sending a next |
323 | version, add a ``patch changelog`` to the cover letter or to individual patches | |
324 | explaining difference aganst previous submission (see | |
325 | :ref:`the_canonical_patch_format`). | |
1da177e4 | 326 | |
9912d0bb | 327 | See Documentation/process/email-clients.rst for recommendations on email |
7433ff33 DD |
328 | clients and mailing list etiquette. |
329 | ||
31c9d7c8 | 330 | .. _resend_reminders: |
1da177e4 | 331 | |
ef227c39 DD |
332 | Don't get discouraged - or impatient |
333 | ------------------------------------ | |
1da177e4 | 334 | |
0eea2314 JC |
335 | After you have submitted your change, be patient and wait. Reviewers are |
336 | busy people and may not get to your patch right away. | |
1da177e4 | 337 | |
0eea2314 JC |
338 | Once upon a time, patches used to disappear into the void without comment, |
339 | but the development process works more smoothly than that now. You should | |
340 | receive comments within a week or so; if that does not happen, make sure | |
341 | that you have sent your patches to the right place. Wait for a minimum of | |
342 | one week before resubmitting or pinging reviewers - possibly longer during | |
343 | busy times like merge windows. | |
1da177e4 | 344 | |
6349469a BP |
345 | It's also ok to resend the patch or the patch series after a couple of |
346 | weeks with the word "RESEND" added to the subject line:: | |
347 | ||
348 | [PATCH Vx RESEND] sub/sys: Condensed patch summary | |
349 | ||
350 | Don't add "RESEND" when you are submitting a modified version of your | |
351 | patch or patch series - "RESEND" only applies to resubmission of a | |
352 | patch or patch series which have not been modified in any way from the | |
353 | previous submission. | |
1da177e4 | 354 | |
1da177e4 | 355 | |
ef227c39 DD |
356 | Include PATCH in the subject |
357 | ----------------------------- | |
1da177e4 LT |
358 | |
359 | Due to high e-mail traffic to Linus, and to linux-kernel, it is common | |
360 | convention to prefix your subject line with [PATCH]. This lets Linus | |
361 | and other kernel developers more easily distinguish patches from other | |
362 | e-mail discussions. | |
363 | ||
9f364b60 | 364 | ``git send-email`` will do this for you automatically. |
1da177e4 LT |
365 | |
366 | ||
ef227c39 DD |
367 | Sign your work - the Developer's Certificate of Origin |
368 | ------------------------------------------------------ | |
1da177e4 LT |
369 | |
370 | To improve tracking of who did what, especially with patches that can | |
371 | percolate to their final resting place in the kernel through several | |
372 | layers of maintainers, we've introduced a "sign-off" procedure on | |
373 | patches that are being emailed around. | |
374 | ||
375 | The sign-off is a simple line at the end of the explanation for the | |
376 | patch, which certifies that you wrote it or otherwise have the right to | |
db12fb83 | 377 | pass it on as an open-source patch. The rules are pretty simple: if you |
1da177e4 LT |
378 | can certify the below: |
379 | ||
5903019b MCC |
380 | Developer's Certificate of Origin 1.1 |
381 | ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ | |
1da177e4 | 382 | |
5903019b | 383 | By making a contribution to this project, I certify that: |
1da177e4 LT |
384 | |
385 | (a) The contribution was created in whole or in part by me and I | |
386 | have the right to submit it under the open source license | |
387 | indicated in the file; or | |
388 | ||
389 | (b) The contribution is based upon previous work that, to the best | |
390 | of my knowledge, is covered under an appropriate open source | |
391 | license and I have the right under that license to submit that | |
392 | work with modifications, whether created in whole or in part | |
393 | by me, under the same open source license (unless I am | |
394 | permitted to submit under a different license), as indicated | |
395 | in the file; or | |
396 | ||
397 | (c) The contribution was provided directly to me by some other | |
398 | person who certified (a), (b) or (c) and I have not modified | |
399 | it. | |
400 | ||
e00bfcbf SB |
401 | (d) I understand and agree that this project and the contribution |
402 | are public and that a record of the contribution (including all | |
403 | personal information I submit with it, including my sign-off) is | |
404 | maintained indefinitely and may be redistributed consistent with | |
405 | this project or the open source license(s) involved. | |
cbd83da8 | 406 | |
5903019b | 407 | then you just add a line saying:: |
1da177e4 | 408 | |
9fd5559c | 409 | Signed-off-by: Random J Developer <random@developer.example.org> |
1da177e4 | 410 | |
af45f32d | 411 | using your real name (sorry, no pseudonyms or anonymous contributions.) |
9f364b60 | 412 | This will be done for you automatically if you use ``git commit -s``. |
7d717887 AS |
413 | Reverts should also include "Signed-off-by". ``git revert -s`` does that |
414 | for you. | |
af45f32d | 415 | |
1da177e4 LT |
416 | Some people also put extra tags at the end. They'll just be ignored for |
417 | now, but you can do this to mark internal company procedures or just | |
e00bfcbf | 418 | point out some special detail about the sign-off. |
1da177e4 | 419 | |
9bf19b78 BP |
420 | Any further SoBs (Signed-off-by:'s) following the author's SoB are from |
421 | people handling and transporting the patch, but were not involved in its | |
422 | development. SoB chains should reflect the **real** route a patch took | |
423 | as it was propagated to the maintainers and ultimately to Linus, with | |
424 | the first SoB entry signalling primary authorship of a single author. | |
425 | ||
1da177e4 | 426 | |
ef227c39 DD |
427 | When to use Acked-by:, Cc:, and Co-developed-by: |
428 | ------------------------------------------------ | |
0a920b5b | 429 | |
0f44cd23 AM |
430 | The Signed-off-by: tag indicates that the signer was involved in the |
431 | development of the patch, or that he/she was in the patch's delivery path. | |
432 | ||
433 | If a person was not directly involved in the preparation or handling of a | |
434 | patch but wishes to signify and record their approval of it then they can | |
d00c4559 | 435 | ask to have an Acked-by: line added to the patch's changelog. |
0f44cd23 AM |
436 | |
437 | Acked-by: is often used by the maintainer of the affected code when that | |
438 | maintainer neither contributed to nor forwarded the patch. | |
439 | ||
440 | Acked-by: is not as formal as Signed-off-by:. It is a record that the acker | |
441 | has at least reviewed the patch and has indicated acceptance. Hence patch | |
442 | mergers will sometimes manually convert an acker's "yep, looks good to me" | |
d00c4559 JC |
443 | into an Acked-by: (but note that it is usually better to ask for an |
444 | explicit ack). | |
0f44cd23 AM |
445 | |
446 | Acked-by: does not necessarily indicate acknowledgement of the entire patch. | |
447 | For example, if a patch affects multiple subsystems and has an Acked-by: from | |
448 | one subsystem maintainer then this usually indicates acknowledgement of just | |
449 | the part which affects that maintainer's code. Judgement should be used here. | |
ef40203a | 450 | When in doubt people should refer to the original discussion in the mailing |
0f44cd23 AM |
451 | list archives. |
452 | ||
ef40203a | 453 | If a person has had the opportunity to comment on a patch, but has not |
5903019b | 454 | provided such comments, you may optionally add a ``Cc:`` tag to the patch. |
ef40203a | 455 | This is the only tag which might be added without an explicit action by the |
d00c4559 JC |
456 | person it names - but it should indicate that this person was copied on the |
457 | patch. This tag documents that potentially interested parties | |
458 | have been included in the discussion. | |
0f44cd23 | 459 | |
24a2bb90 | 460 | Co-developed-by: states that the patch was co-created by multiple developers; |
7e902857 | 461 | it is used to give attribution to co-authors (in addition to the author |
24a2bb90 SC |
462 | attributed by the From: tag) when several people work on a single patch. Since |
463 | Co-developed-by: denotes authorship, every Co-developed-by: must be immediately | |
464 | followed by a Signed-off-by: of the associated co-author. Standard sign-off | |
465 | procedure applies, i.e. the ordering of Signed-off-by: tags should reflect the | |
466 | chronological history of the patch insofar as possible, regardless of whether | |
467 | the author is attributed via From: or Co-developed-by:. Notably, the last | |
468 | Signed-off-by: must always be that of the developer submitting the patch. | |
469 | ||
470 | Note, the From: tag is optional when the From: author is also the person (and | |
471 | email) listed in the From: line of the email header. | |
472 | ||
473 | Example of a patch submitted by the From: author:: | |
474 | ||
475 | <changelog> | |
476 | ||
477 | Co-developed-by: First Co-Author <first@coauthor.example.org> | |
478 | Signed-off-by: First Co-Author <first@coauthor.example.org> | |
479 | Co-developed-by: Second Co-Author <second@coauthor.example.org> | |
480 | Signed-off-by: Second Co-Author <second@coauthor.example.org> | |
481 | Signed-off-by: From Author <from@author.example.org> | |
482 | ||
483 | Example of a patch submitted by a Co-developed-by: author:: | |
484 | ||
485 | From: From Author <from@author.example.org> | |
486 | ||
487 | <changelog> | |
488 | ||
489 | Co-developed-by: Random Co-Author <random@coauthor.example.org> | |
490 | Signed-off-by: Random Co-Author <random@coauthor.example.org> | |
491 | Signed-off-by: From Author <from@author.example.org> | |
492 | Co-developed-by: Submitting Co-Author <sub@coauthor.example.org> | |
493 | Signed-off-by: Submitting Co-Author <sub@coauthor.example.org> | |
82d95343 | 494 | |
ef40203a | 495 | |
ef227c39 DD |
496 | Using Reported-by:, Tested-by:, Reviewed-by:, Suggested-by: and Fixes: |
497 | ---------------------------------------------------------------------- | |
bbb0a424 | 498 | |
d75ef707 DC |
499 | The Reported-by tag gives credit to people who find bugs and report them and it |
500 | hopefully inspires them to help us again in the future. Please note that if | |
501 | the bug was reported in private, then ask for permission first before using the | |
869f496e AS |
502 | Reported-by tag. The tag is intended for bugs; please do not use it to credit |
503 | feature requests. | |
ef40203a JC |
504 | |
505 | A Tested-by: tag indicates that the patch has been successfully tested (in | |
506 | some environment) by the person named. This tag informs maintainers that | |
507 | some testing has been performed, provides a means to locate testers for | |
508 | future patches, and ensures credit for the testers. | |
509 | ||
510 | Reviewed-by:, instead, indicates that the patch has been reviewed and found | |
511 | acceptable according to the Reviewer's Statement: | |
512 | ||
5903019b MCC |
513 | Reviewer's statement of oversight |
514 | ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ | |
ef40203a | 515 | |
5903019b | 516 | By offering my Reviewed-by: tag, I state that: |
ef40203a | 517 | |
5903019b | 518 | (a) I have carried out a technical review of this patch to |
ef40203a JC |
519 | evaluate its appropriateness and readiness for inclusion into |
520 | the mainline kernel. | |
521 | ||
522 | (b) Any problems, concerns, or questions relating to the patch | |
523 | have been communicated back to the submitter. I am satisfied | |
524 | with the submitter's response to my comments. | |
525 | ||
526 | (c) While there may be things that could be improved with this | |
527 | submission, I believe that it is, at this time, (1) a | |
528 | worthwhile modification to the kernel, and (2) free of known | |
529 | issues which would argue against its inclusion. | |
530 | ||
531 | (d) While I have reviewed the patch and believe it to be sound, I | |
532 | do not (unless explicitly stated elsewhere) make any | |
533 | warranties or guarantees that it will achieve its stated | |
534 | purpose or function properly in any given situation. | |
535 | ||
536 | A Reviewed-by tag is a statement of opinion that the patch is an | |
537 | appropriate modification of the kernel without any remaining serious | |
538 | technical issues. Any interested reviewer (who has done the work) can | |
539 | offer a Reviewed-by tag for a patch. This tag serves to give credit to | |
540 | reviewers and to inform maintainers of the degree of review which has been | |
541 | done on the patch. Reviewed-by: tags, when supplied by reviewers known to | |
542 | understand the subject area and to perform thorough reviews, will normally | |
5801da1b | 543 | increase the likelihood of your patch getting into the kernel. |
ef40203a | 544 | |
030f066f KK |
545 | Both Tested-by and Reviewed-by tags, once received on mailing list from tester |
546 | or reviewer, should be added by author to the applicable patches when sending | |
547 | next versions. However if the patch has changed substantially in following | |
548 | version, these tags might not be applicable anymore and thus should be removed. | |
549 | Usually removal of someone's Tested-by or Reviewed-by tags should be mentioned | |
550 | in the patch changelog (after the '---' separator). | |
551 | ||
8543ae12 M |
552 | A Suggested-by: tag indicates that the patch idea is suggested by the person |
553 | named and ensures credit to the person for the idea. Please note that this | |
554 | tag should not be added without the reporter's permission, especially if the | |
555 | idea was not posted in a public forum. That said, if we diligently credit our | |
556 | idea reporters, they will, hopefully, be inspired to help us again in the | |
557 | future. | |
558 | ||
8401aa1f JK |
559 | A Fixes: tag indicates that the patch fixes an issue in a previous commit. It |
560 | is used to make it easy to determine where a bug originated, which can help | |
561 | review a bug fix. This tag also assists the stable kernel team in determining | |
562 | which stable kernel versions should receive your fix. This is the preferred | |
5903019b MCC |
563 | method for indicating a bug fixed by the patch. See :ref:`describe_changes` |
564 | for more details. | |
8401aa1f | 565 | |
f0ea149e | 566 | Note: Attaching a Fixes: tag does not subvert the stable kernel rules |
9912d0bb | 567 | process nor the requirement to Cc: stable@vger.kernel.org on all stable |
f0ea149e | 568 | patch candidates. For more information, please read |
9912d0bb MCC |
569 | Documentation/process/stable-kernel-rules.rst. |
570 | ||
f58252cd | 571 | .. _the_canonical_patch_format: |
ef40203a | 572 | |
ef227c39 DD |
573 | The canonical patch format |
574 | -------------------------- | |
7994cc15 JC |
575 | |
576 | This section describes how the patch itself should be formatted. Note | |
9b2c7677 | 577 | that, if you have your patches stored in a ``git`` repository, proper patch |
5903019b | 578 | formatting can be had with ``git format-patch``. The tools cannot create |
7994cc15 | 579 | the necessary text, though, so read the instructions below anyway. |
84da7c08 | 580 | |
5903019b | 581 | The canonical patch subject line is:: |
75f8426c | 582 | |
d6b9acc0 | 583 | Subject: [PATCH 001/123] subsystem: summary phrase |
75f8426c PJ |
584 | |
585 | The canonical patch message body contains the following: | |
586 | ||
d19b3e32 JH |
587 | - A ``from`` line specifying the patch author, followed by an empty |
588 | line (only needed if the person sending the patch is not the author). | |
75f8426c | 589 | |
2a076f40 JP |
590 | - The body of the explanation, line wrapped at 75 columns, which will |
591 | be copied to the permanent changelog to describe this patch. | |
75f8426c | 592 | |
d19b3e32 JH |
593 | - An empty line. |
594 | ||
5903019b | 595 | - The ``Signed-off-by:`` lines, described above, which will |
75f8426c PJ |
596 | also go in the changelog. |
597 | ||
5903019b | 598 | - A marker line containing simply ``---``. |
75f8426c PJ |
599 | |
600 | - Any additional comments not suitable for the changelog. | |
601 | ||
9b2c7677 | 602 | - The actual patch (``diff`` output). |
75f8426c PJ |
603 | |
604 | The Subject line format makes it very easy to sort the emails | |
605 | alphabetically by subject line - pretty much any email reader will | |
606 | support that - since because the sequence number is zero-padded, | |
607 | the numerical and alphabetic sort is the same. | |
608 | ||
5903019b | 609 | The ``subsystem`` in the email's Subject should identify which |
d6b9acc0 PJ |
610 | area or subsystem of the kernel is being patched. |
611 | ||
5903019b MCC |
612 | The ``summary phrase`` in the email's Subject should concisely |
613 | describe the patch which that email contains. The ``summary | |
614 | phrase`` should not be a filename. Do not use the same ``summary | |
615 | phrase`` for every patch in a whole patch series (where a ``patch | |
616 | series`` is an ordered sequence of multiple, related patches). | |
d6b9acc0 | 617 | |
5903019b | 618 | Bear in mind that the ``summary phrase`` of your email becomes a |
2ae19aca | 619 | globally-unique identifier for that patch. It propagates all the way |
9b2c7677 | 620 | into the ``git`` changelog. The ``summary phrase`` may later be used in |
2ae19aca | 621 | developer discussions which refer to the patch. People will want to |
5903019b | 622 | google for the ``summary phrase`` to read discussion regarding that |
2ae19aca TT |
623 | patch. It will also be the only thing that people may quickly see |
624 | when, two or three months later, they are going through perhaps | |
9b2c7677 MCC |
625 | thousands of patches using tools such as ``gitk`` or ``git log |
626 | --oneline``. | |
2ae19aca | 627 | |
5903019b | 628 | For these reasons, the ``summary`` must be no more than 70-75 |
2ae19aca TT |
629 | characters, and it must describe both what the patch changes, as well |
630 | as why the patch might be necessary. It is challenging to be both | |
631 | succinct and descriptive, but that is what a well-written summary | |
632 | should do. | |
633 | ||
5903019b | 634 | The ``summary phrase`` may be prefixed by tags enclosed in square |
e12d7462 AH |
635 | brackets: "Subject: [PATCH <tag>...] <summary phrase>". The tags are |
636 | not considered part of the summary phrase, but describe how the patch | |
2ae19aca TT |
637 | should be treated. Common tags might include a version descriptor if |
638 | the multiple versions of the patch have been sent out in response to | |
639 | comments (i.e., "v1, v2, v3"), or "RFC" to indicate a request for | |
875f82cb | 640 | comments. |
d6b9acc0 | 641 | |
875f82cb BP |
642 | If there are four patches in a patch series the individual patches may |
643 | be numbered like this: 1/4, 2/4, 3/4, 4/4. This assures that developers | |
644 | understand the order in which the patches should be applied and that | |
645 | they have reviewed or applied all of the patches in the patch series. | |
646 | ||
647 | Here are some good example Subjects:: | |
d6b9acc0 | 648 | |
e12d7462 AH |
649 | Subject: [PATCH 2/5] ext2: improve scalability of bitmap searching |
650 | Subject: [PATCH v2 01/27] x86: fix eflags tracking | |
875f82cb BP |
651 | Subject: [PATCH v2] sub/sys: Condensed patch summary |
652 | Subject: [PATCH v2 M/N] sub/sys: Condensed patch summary | |
75f8426c | 653 | |
5903019b | 654 | The ``from`` line must be the very first line in the message body, |
75f8426c PJ |
655 | and has the form: |
656 | ||
24a2bb90 | 657 | From: Patch Author <author@example.com> |
75f8426c | 658 | |
5903019b MCC |
659 | The ``from`` line specifies who will be credited as the author of the |
660 | patch in the permanent changelog. If the ``from`` line is missing, | |
661 | then the ``From:`` line from the email header will be used to determine | |
75f8426c PJ |
662 | the patch author in the changelog. |
663 | ||
664 | The explanation body will be committed to the permanent source | |
875f82cb BP |
665 | changelog, so should make sense to a competent reader who has long since |
666 | forgotten the immediate details of the discussion that might have led to | |
667 | this patch. Including symptoms of the failure which the patch addresses | |
668 | (kernel log messages, oops messages, etc.) are especially useful for | |
669 | people who might be searching the commit logs looking for the applicable | |
670 | patch. The text should be written in such detail so that when read | |
671 | weeks, months or even years later, it can give the reader the needed | |
672 | details to grasp the reasoning for **why** the patch was created. | |
673 | ||
674 | If a patch fixes a compile failure, it may not be necessary to include | |
675 | _all_ of the compile failures; just enough that it is likely that | |
676 | someone searching for the patch can find it. As in the ``summary | |
677 | phrase``, it is important to be both succinct as well as descriptive. | |
678 | ||
679 | The ``---`` marker line serves the essential purpose of marking for | |
680 | patch handling tools where the changelog message ends. | |
681 | ||
682 | One good use for the additional comments after the ``---`` marker is | |
683 | for a ``diffstat``, to show what files have changed, and the number of | |
684 | inserted and deleted lines per file. A ``diffstat`` is especially useful | |
685 | on bigger patches. If you are going to include a ``diffstat`` after the | |
686 | ``---`` marker, please use ``diffstat`` options ``-p 1 -w 70`` so that | |
687 | filenames are listed from the top of the kernel source tree and don't | |
688 | use too much horizontal space (easily fit in 80 columns, maybe with some | |
689 | indentation). (``git`` generates appropriate diffstats by default.) | |
690 | ||
691 | Other comments relevant only to the moment or the maintainer, not | |
692 | suitable for the permanent changelog, should also go here. A good | |
693 | example of such comments might be ``patch changelogs`` which describe | |
694 | what has changed between the v1 and v2 version of the patch. | |
695 | ||
696 | Please put this information **after** the ``---`` line which separates | |
697 | the changelog from the rest of the patch. The version information is | |
698 | not part of the changelog which gets committed to the git tree. It is | |
699 | additional information for the reviewers. If it's placed above the | |
700 | commit tags, it needs manual interaction to remove it. If it is below | |
701 | the separator line, it gets automatically stripped off when applying the | |
702 | patch:: | |
703 | ||
704 | <commit message> | |
705 | ... | |
706 | Signed-off-by: Author <author@mail> | |
707 | --- | |
708 | V2 -> V3: Removed redundant helper function | |
709 | V1 -> V2: Cleaned up coding style and addressed review comments | |
710 | ||
711 | path/to/file | 5+++-- | |
712 | ... | |
75f8426c PJ |
713 | |
714 | See more details on the proper patch format in the following | |
715 | references. | |
716 | ||
31c9d7c8 TG |
717 | .. _backtraces: |
718 | ||
78f101a1 BP |
719 | Backtraces in commit mesages |
720 | ^^^^^^^^^^^^^^^^^^^^^^^^^^^^ | |
721 | ||
722 | Backtraces help document the call chain leading to a problem. However, | |
723 | not all backtraces are helpful. For example, early boot call chains are | |
724 | unique and obvious. Copying the full dmesg output verbatim, however, | |
725 | adds distracting information like timestamps, module lists, register and | |
726 | stack dumps. | |
727 | ||
728 | Therefore, the most useful backtraces should distill the relevant | |
729 | information from the dump, which makes it easier to focus on the real | |
730 | issue. Here is an example of a well-trimmed backtrace:: | |
731 | ||
732 | unchecked MSR access error: WRMSR to 0xd51 (tried to write 0x0000000000000064) | |
733 | at rIP: 0xffffffffae059994 (native_write_msr+0x4/0x20) | |
734 | Call Trace: | |
735 | mba_wrmsr | |
736 | update_domains | |
737 | rdtgroup_mkdir | |
738 | ||
5903019b MCC |
739 | .. _explicit_in_reply_to: |
740 | ||
ef227c39 DD |
741 | Explicit In-Reply-To headers |
742 | ---------------------------- | |
d7ac8d85 CM |
743 | |
744 | It can be helpful to manually add In-Reply-To: headers to a patch | |
5903019b | 745 | (e.g., when using ``git send-email``) to associate the patch with |
d7ac8d85 CM |
746 | previous relevant discussion, e.g. to link a bug fix to the email with |
747 | the bug report. However, for a multi-patch series, it is generally | |
748 | best to avoid using In-Reply-To: to link to older versions of the | |
749 | series. This way multiple versions of the patch don't become an | |
750 | unmanageable forest of references in email clients. If a link is | |
a9d85efb | 751 | helpful, you can use the https://lore.kernel.org/ redirector (e.g., in |
d7ac8d85 CM |
752 | the cover email text) to link to an earlier version of the patch series. |
753 | ||
75f8426c | 754 | |
ef227c39 DD |
755 | Providing base tree information |
756 | ------------------------------- | |
e8686a40 KR |
757 | |
758 | When other developers receive your patches and start the review process, | |
759 | it is often useful for them to know where in the tree history they | |
760 | should place your work. This is particularly useful for automated CI | |
761 | processes that attempt to run a series of tests in order to establish | |
762 | the quality of your submission before the maintainer starts the review. | |
763 | ||
764 | If you are using ``git format-patch`` to generate your patches, you can | |
765 | automatically include the base tree information in your submission by | |
766 | using the ``--base`` flag. The easiest and most convenient way to use | |
767 | this option is with topical branches:: | |
768 | ||
769 | $ git checkout -t -b my-topical-branch master | |
770 | Branch 'my-topical-branch' set up to track local branch 'master'. | |
771 | Switched to a new branch 'my-topical-branch' | |
772 | ||
773 | [perform your edits and commits] | |
774 | ||
775 | $ git format-patch --base=auto --cover-letter -o outgoing/ master | |
776 | outgoing/0000-cover-letter.patch | |
777 | outgoing/0001-First-Commit.patch | |
778 | outgoing/... | |
779 | ||
780 | When you open ``outgoing/0000-cover-letter.patch`` for editing, you will | |
781 | notice that it will have the ``base-commit:`` trailer at the very | |
782 | bottom, which provides the reviewer and the CI tools enough information | |
783 | to properly perform ``git am`` without worrying about conflicts:: | |
784 | ||
785 | $ git checkout -b patch-review [base-commit-id] | |
786 | Switched to a new branch 'patch-review' | |
787 | $ git am patches.mbox | |
788 | Applying: First Commit | |
789 | Applying: ... | |
790 | ||
791 | Please see ``man git-format-patch`` for more information about this | |
792 | option. | |
793 | ||
794 | .. note:: | |
795 | ||
796 | The ``--base`` feature was introduced in git version 2.9.0. | |
797 | ||
798 | If you are not using git to format your patches, you can still include | |
799 | the same ``base-commit`` trailer to indicate the commit hash of the tree | |
800 | on which your work is based. You should add it either in the cover | |
801 | letter or in the first patch of the series and it should be placed | |
802 | either below the ``---`` line or at the very bottom of all other | |
803 | content, right before your email signature. | |
804 | ||
805 | ||
89edeedd JC |
806 | References |
807 | ---------- | |
5b0ed2c6 XVP |
808 | |
809 | Andrew Morton, "The perfect patch" (tpp). | |
e7b4311e | 810 | <https://www.ozlabs.org/~akpm/stuff/tpp.txt> |
5b0ed2c6 | 811 | |
8e9cb8fd | 812 | Jeff Garzik, "Linux kernel patch submission format". |
5aff7c46 | 813 | <https://web.archive.org/web/20180829112450/http://linux.yyz.us/patch-format.html> |
5b0ed2c6 | 814 | |
8e9cb8fd | 815 | Greg Kroah-Hartman, "How to piss off a kernel subsystem maintainer". |
f5039935 | 816 | <http://www.kroah.com/log/linux/maintainer.html> |
9b2c7677 | 817 | |
f5039935 | 818 | <http://www.kroah.com/log/linux/maintainer-02.html> |
9b2c7677 | 819 | |
f5039935 | 820 | <http://www.kroah.com/log/linux/maintainer-03.html> |
9b2c7677 | 821 | |
f5039935 | 822 | <http://www.kroah.com/log/linux/maintainer-04.html> |
9b2c7677 | 823 | |
f5039935 | 824 | <http://www.kroah.com/log/linux/maintainer-05.html> |
9b2c7677 | 825 | |
7e0dae61 | 826 | <http://www.kroah.com/log/linux/maintainer-06.html> |
5b0ed2c6 | 827 | |
bc7455fa | 828 | NO!!!! No more huge patch bombs to linux-kernel@vger.kernel.org people! |
05a5f51c | 829 | <https://lore.kernel.org/r/20050711.125305.08322243.davem@davemloft.net> |
5b0ed2c6 | 830 | |
9912d0bb | 831 | Kernel Documentation/process/coding-style.rst |
5b0ed2c6 | 832 | |
8e9cb8fd | 833 | Linus Torvalds's mail on the canonical patch format: |
05a5f51c | 834 | <https://lore.kernel.org/r/Pine.LNX.4.58.0504071023190.28951@ppc970.osdl.org> |
9536727e AK |
835 | |
836 | Andi Kleen, "On submitting kernel patches" | |
25985edc | 837 | Some strategies to get difficult or controversial changes in. |
9b2c7677 | 838 | |
9536727e | 839 | http://halobates.de/on-submitting-patches.pdf |