1.. _development_process:
2
3How the development process works
4=================================
5
6Linux kernel development in the early 1990's was a pretty loose affair,
7with relatively small numbers of users and developers involved.  With a
8user base in the millions and with some 2,000 developers involved over the
9course of one year, the kernel has since had to evolve a number of
10processes to keep development happening smoothly.  A solid understanding of
11how the process works is required in order to be an effective part of it.
12
13The big picture
14---------------
15
16The kernel developers use a loosely time-based release process, with a new
17major kernel release happening every two or three months.  The recent
18release history looks like this:
19
20	======  =================
21	5.0	March 3, 2019
22	5.1	May 5, 2019
23	5.2	July 7, 2019
24	5.3	September 15, 2019
25	5.4	November 24, 2019
26	5.5	January 6, 2020
27	======  =================
28
29Every 5.x release is a major kernel release with new features, internal
30API changes, and more.  A typical release can contain about 13,000
31changesets with changes to several hundred thousand lines of code.  5.x is
32the leading edge of Linux kernel development; the kernel uses a
33rolling development model which is continually integrating major changes.
34
35A relatively straightforward discipline is followed with regard to the
36merging of patches for each release.  At the beginning of each development
37cycle, the "merge window" is said to be open.  At that time, code which is
38deemed to be sufficiently stable (and which is accepted by the development
39community) is merged into the mainline kernel.  The bulk of changes for a
40new development cycle (and all of the major changes) will be merged during
41this time, at a rate approaching 1,000 changes ("patches," or "changesets")
42per day.
43
44(As an aside, it is worth noting that the changes integrated during the
45merge window do not come out of thin air; they have been collected, tested,
46and staged ahead of time.  How that process works will be described in
47detail later on).
48
49The merge window lasts for approximately two weeks.  At the end of this
50time, Linus Torvalds will declare that the window is closed and release the
51first of the "rc" kernels.  For the kernel which is destined to be 5.6,
52for example, the release which happens at the end of the merge window will
53be called 5.6-rc1.  The -rc1 release is the signal that the time to
54merge new features has passed, and that the time to stabilize the next
55kernel has begun.
56
57Over the next six to ten weeks, only patches which fix problems should be
58submitted to the mainline.  On occasion a more significant change will be
59allowed, but such occasions are rare; developers who try to merge new
60features outside of the merge window tend to get an unfriendly reception.
61As a general rule, if you miss the merge window for a given feature, the
62best thing to do is to wait for the next development cycle.  (An occasional
63exception is made for drivers for previously-unsupported hardware; if they
64touch no in-tree code, they cannot cause regressions and should be safe to
65add at any time).
66
67As fixes make their way into the mainline, the patch rate will slow over
68time.  Linus releases new -rc kernels about once a week; a normal series
69will get up to somewhere between -rc6 and -rc9 before the kernel is
70considered to be sufficiently stable and the final release is made.
71At that point the whole process starts over again.
72
73As an example, here is how the 5.4 development cycle went (all dates in
742019):
75
76	==============  ===============================
77	September 15	5.3 stable release
78	September 30	5.4-rc1, merge window closes
79	October 6	5.4-rc2
80	October 13	5.4-rc3
81	October 20	5.4-rc4
82	October 27	5.4-rc5
83	November 3	5.4-rc6
84	November 10	5.4-rc7
85	November 17	5.4-rc8
86	November 24	5.4 stable release
87	==============  ===============================
88
89How do the developers decide when to close the development cycle and create
90the stable release?  The most significant metric used is the list of
91regressions from previous releases.  No bugs are welcome, but those which
92break systems which worked in the past are considered to be especially
93serious.  For this reason, patches which cause regressions are looked upon
94unfavorably and are quite likely to be reverted during the stabilization
95period.
96
97The developers' goal is to fix all known regressions before the stable
98release is made.  In the real world, this kind of perfection is hard to
99achieve; there are just too many variables in a project of this size.
100There comes a point where delaying the final release just makes the problem
101worse; the pile of changes waiting for the next merge window will grow
102larger, creating even more regressions the next time around.  So most 5.x
103kernels go out with a handful of known regressions though, hopefully, none
104of them are serious.
105
106Once a stable release is made, its ongoing maintenance is passed off to the
107"stable team," currently Greg Kroah-Hartman. The stable team will release
108occasional updates to the stable release using the 5.x.y numbering scheme.
109To be considered for an update release, a patch must (1) fix a significant
110bug, and (2) already be merged into the mainline for the next development
111kernel. Kernels will typically receive stable updates for a little more
112than one development cycle past their initial release. So, for example, the
1135.2 kernel's history looked like this (all dates in 2019):
114
115	==============  ===============================
116	July 7		5.2 stable release
117	July 14		5.2.1
118	July 21		5.2.2
119	July 26		5.2.3
120	July 28		5.2.4
121	July 31  	5.2.5
122	...		...
123	October 11	5.2.21
124	==============  ===============================
125
1265.2.21 was the final stable update of the 5.2 release.
127
128Some kernels are designated "long term" kernels; they will receive support
129for a longer period.  Please refer to the following link for the list of active
130long term kernel versions and their maintainers:
131
132	https://www.kernel.org/category/releases.html
133
134The selection of a kernel for long-term support is purely a matter of a
135maintainer having the need and the time to maintain that release.  There
136are no known plans for long-term support for any specific upcoming
137release.
138
139
140The lifecycle of a patch
141------------------------
142
143Patches do not go directly from the developer's keyboard into the mainline
144kernel.  There is, instead, a somewhat involved (if somewhat informal)
145process designed to ensure that each patch is reviewed for quality and that
146each patch implements a change which is desirable to have in the mainline.
147This process can happen quickly for minor fixes, or, in the case of large
148and controversial changes, go on for years.  Much developer frustration
149comes from a lack of understanding of this process or from attempts to
150circumvent it.
151
152In the hopes of reducing that frustration, this document will describe how
153a patch gets into the kernel.  What follows below is an introduction which
154describes the process in a somewhat idealized way.  A much more detailed
155treatment will come in later sections.
156
157The stages that a patch goes through are, generally:
158
159 - Design.  This is where the real requirements for the patch - and the way
160   those requirements will be met - are laid out.  Design work is often
161   done without involving the community, but it is better to do this work
162   in the open if at all possible; it can save a lot of time redesigning
163   things later.
164
165 - Early review.  Patches are posted to the relevant mailing list, and
166   developers on that list reply with any comments they may have.  This
167   process should turn up any major problems with a patch if all goes
168   well.
169
170 - Wider review.  When the patch is getting close to ready for mainline
171   inclusion, it should be accepted by a relevant subsystem maintainer -
172   though this acceptance is not a guarantee that the patch will make it
173   all the way to the mainline.  The patch will show up in the maintainer's
174   subsystem tree and into the -next trees (described below).  When the
175   process works, this step leads to more extensive review of the patch and
176   the discovery of any problems resulting from the integration of this
177   patch with work being done by others.
178
179-  Please note that most maintainers also have day jobs, so merging
180   your patch may not be their highest priority.  If your patch is
181   getting feedback about changes that are needed, you should either
182   make those changes or justify why they should not be made.  If your
183   patch has no review complaints but is not being merged by its
184   appropriate subsystem or driver maintainer, you should be persistent
185   in updating the patch to the current kernel so that it applies cleanly
186   and keep sending it for review and merging.
187
188 - Merging into the mainline.  Eventually, a successful patch will be
189   merged into the mainline repository managed by Linus Torvalds.  More
190   comments and/or problems may surface at this time; it is important that
191   the developer be responsive to these and fix any issues which arise.
192
193 - Stable release.  The number of users potentially affected by the patch
194   is now large, so, once again, new problems may arise.
195
196 - Long-term maintenance.  While it is certainly possible for a developer
197   to forget about code after merging it, that sort of behavior tends to
198   leave a poor impression in the development community.  Merging code
199   eliminates some of the maintenance burden, in that others will fix
200   problems caused by API changes.  But the original developer should
201   continue to take responsibility for the code if it is to remain useful
202   in the longer term.
203
204One of the largest mistakes made by kernel developers (or their employers)
205is to try to cut the process down to a single "merging into the mainline"
206step.  This approach invariably leads to frustration for everybody
207involved.
208
209How patches get into the Kernel
210-------------------------------
211
212There is exactly one person who can merge patches into the mainline kernel
213repository: Linus Torvalds. But, for example, of the over 9,500 patches
214which went into the 2.6.38 kernel, only 112 (around 1.3%) were directly
215chosen by Linus himself. The kernel project has long since grown to a size
216where no single developer could possibly inspect and select every patch
217unassisted. The way the kernel developers have addressed this growth is
218through the use of a lieutenant system built around a chain of trust.
219
220The kernel code base is logically broken down into a set of subsystems:
221networking, specific architecture support, memory management, video
222devices, etc.  Most subsystems have a designated maintainer, a developer
223who has overall responsibility for the code within that subsystem.  These
224subsystem maintainers are the gatekeepers (in a loose way) for the portion
225of the kernel they manage; they are the ones who will (usually) accept a
226patch for inclusion into the mainline kernel.
227
228Subsystem maintainers each manage their own version of the kernel source
229tree, usually (but certainly not always) using the git source management
230tool.  Tools like git (and related tools like quilt or mercurial) allow
231maintainers to track a list of patches, including authorship information
232and other metadata.  At any given time, the maintainer can identify which
233patches in his or her repository are not found in the mainline.
234
235When the merge window opens, top-level maintainers will ask Linus to "pull"
236the patches they have selected for merging from their repositories.  If
237Linus agrees, the stream of patches will flow up into his repository,
238becoming part of the mainline kernel.  The amount of attention that Linus
239pays to specific patches received in a pull operation varies.  It is clear
240that, sometimes, he looks quite closely.  But, as a general rule, Linus
241trusts the subsystem maintainers to not send bad patches upstream.
242
243Subsystem maintainers, in turn, can pull patches from other maintainers.
244For example, the networking tree is built from patches which accumulated
245first in trees dedicated to network device drivers, wireless networking,
246etc.  This chain of repositories can be arbitrarily long, though it rarely
247exceeds two or three links.  Since each maintainer in the chain trusts
248those managing lower-level trees, this process is known as the "chain of
249trust."
250
251Clearly, in a system like this, getting patches into the kernel depends on
252finding the right maintainer.  Sending patches directly to Linus is not
253normally the right way to go.
254
255
256Next trees
257----------
258
259The chain of subsystem trees guides the flow of patches into the kernel,
260but it also raises an interesting question: what if somebody wants to look
261at all of the patches which are being prepared for the next merge window?
262Developers will be interested in what other changes are pending to see
263whether there are any conflicts to worry about; a patch which changes a
264core kernel function prototype, for example, will conflict with any other
265patches which use the older form of that function.  Reviewers and testers
266want access to the changes in their integrated form before all of those
267changes land in the mainline kernel.  One could pull changes from all of
268the interesting subsystem trees, but that would be a big and error-prone
269job.
270
271The answer comes in the form of -next trees, where subsystem trees are
272collected for testing and review.  The older of these trees, maintained by
273Andrew Morton, is called "-mm" (for memory management, which is how it got
274started).  The -mm tree integrates patches from a long list of subsystem
275trees; it also has some patches aimed at helping with debugging.
276
277Beyond that, -mm contains a significant collection of patches which have
278been selected by Andrew directly.  These patches may have been posted on a
279mailing list, or they may apply to a part of the kernel for which there is
280no designated subsystem tree.  As a result, -mm operates as a sort of
281subsystem tree of last resort; if there is no other obvious path for a
282patch into the mainline, it is likely to end up in -mm.  Miscellaneous
283patches which accumulate in -mm will eventually either be forwarded on to
284an appropriate subsystem tree or be sent directly to Linus.  In a typical
285development cycle, approximately 5-10% of the patches going into the
286mainline get there via -mm.
287
288The current -mm patch is available in the "mmotm" (-mm of the moment)
289directory at:
290
291	https://www.ozlabs.org/~akpm/mmotm/
292
293Use of the MMOTM tree is likely to be a frustrating experience, though;
294there is a definite chance that it will not even compile.
295
296The primary tree for next-cycle patch merging is linux-next, maintained by
297Stephen Rothwell.  The linux-next tree is, by design, a snapshot of what
298the mainline is expected to look like after the next merge window closes.
299Linux-next trees are announced on the linux-kernel and linux-next mailing
300lists when they are assembled; they can be downloaded from:
301
302	https://www.kernel.org/pub/linux/kernel/next/
303
304Linux-next has become an integral part of the kernel development process;
305all patches merged during a given merge window should really have found
306their way into linux-next some time before the merge window opens.
307
308
309Staging trees
310-------------
311
312The kernel source tree contains the drivers/staging/ directory, where
313many sub-directories for drivers or filesystems that are on their way to
314being added to the kernel tree live.  They remain in drivers/staging while
315they still need more work; once complete, they can be moved into the
316kernel proper.  This is a way to keep track of drivers that aren't
317up to Linux kernel coding or quality standards, but people may want to use
318them and track development.
319
320Greg Kroah-Hartman currently maintains the staging tree.  Drivers that
321still need work are sent to him, with each driver having its own
322subdirectory in drivers/staging/.  Along with the driver source files, a
323TODO file should be present in the directory as well.  The TODO file lists
324the pending work that the driver needs for acceptance into the kernel
325proper, as well as a list of people that should be Cc'd for any patches to
326the driver.  Current rules require that drivers contributed to staging
327must, at a minimum, compile properly.
328
329Staging can be a relatively easy way to get new drivers into the mainline
330where, with luck, they will come to the attention of other developers and
331improve quickly.  Entry into staging is not the end of the story, though;
332code in staging which is not seeing regular progress will eventually be
333removed.  Distributors also tend to be relatively reluctant to enable
334staging drivers.  So staging is, at best, a stop on the way toward becoming
335a proper mainline driver.
336
337
338Tools
339-----
340
341As can be seen from the above text, the kernel development process depends
342heavily on the ability to herd collections of patches in various
343directions.  The whole thing would not work anywhere near as well as it
344does without suitably powerful tools.  Tutorials on how to use these tools
345are well beyond the scope of this document, but there is space for a few
346pointers.
347
348By far the dominant source code management system used by the kernel
349community is git.  Git is one of a number of distributed version control
350systems being developed in the free software community.  It is well tuned
351for kernel development, in that it performs quite well when dealing with
352large repositories and large numbers of patches.  It also has a reputation
353for being difficult to learn and use, though it has gotten better over
354time.  Some sort of familiarity with git is almost a requirement for kernel
355developers; even if they do not use it for their own work, they'll need git
356to keep up with what other developers (and the mainline) are doing.
357
358Git is now packaged by almost all Linux distributions.  There is a home
359page at:
360
361	https://git-scm.com/
362
363That page has pointers to documentation and tutorials.
364
365Among the kernel developers who do not use git, the most popular choice is
366almost certainly Mercurial:
367
368	https://www.selenic.com/mercurial/
369
370Mercurial shares many features with git, but it provides an interface which
371many find easier to use.
372
373The other tool worth knowing about is Quilt:
374
375	https://savannah.nongnu.org/projects/quilt/
376
377Quilt is a patch management system, rather than a source code management
378system.  It does not track history over time; it is, instead, oriented
379toward tracking a specific set of changes against an evolving code base.
380Some major subsystem maintainers use quilt to manage patches intended to go
381upstream.  For the management of certain kinds of trees (-mm, for example),
382quilt is the best tool for the job.
383
384
385Mailing lists
386-------------
387
388A great deal of Linux kernel development work is done by way of mailing
389lists.  It is hard to be a fully-functioning member of the community
390without joining at least one list somewhere.  But Linux mailing lists also
391represent a potential hazard to developers, who risk getting buried under a
392load of electronic mail, running afoul of the conventions used on the Linux
393lists, or both.
394
395Most kernel mailing lists are run on vger.kernel.org; the master list can
396be found at:
397
398	http://vger.kernel.org/vger-lists.html
399
400There are lists hosted elsewhere, though; a number of them are at
401redhat.com/mailman/listinfo.
402
403The core mailing list for kernel development is, of course, linux-kernel.
404This list is an intimidating place to be; volume can reach 500 messages per
405day, the amount of noise is high, the conversation can be severely
406technical, and participants are not always concerned with showing a high
407degree of politeness.  But there is no other place where the kernel
408development community comes together as a whole; developers who avoid this
409list will miss important information.
410
411There are a few hints which can help with linux-kernel survival:
412
413- Have the list delivered to a separate folder, rather than your main
414  mailbox.  One must be able to ignore the stream for sustained periods of
415  time.
416
417- Do not try to follow every conversation - nobody else does.  It is
418  important to filter on both the topic of interest (though note that
419  long-running conversations can drift away from the original subject
420  without changing the email subject line) and the people who are
421  participating.
422
423- Do not feed the trolls.  If somebody is trying to stir up an angry
424  response, ignore them.
425
426- When responding to linux-kernel email (or that on other lists) preserve
427  the Cc: header for all involved.  In the absence of a strong reason (such
428  as an explicit request), you should never remove recipients.  Always make
429  sure that the person you are responding to is in the Cc: list.  This
430  convention also makes it unnecessary to explicitly ask to be copied on
431  replies to your postings.
432
433- Search the list archives (and the net as a whole) before asking
434  questions.  Some developers can get impatient with people who clearly
435  have not done their homework.
436
437- Avoid top-posting (the practice of putting your answer above the quoted
438  text you are responding to).  It makes your response harder to read and
439  makes a poor impression.
440
441- Ask on the correct mailing list.  Linux-kernel may be the general meeting
442  point, but it is not the best place to find developers from all
443  subsystems.
444
445The last point - finding the correct mailing list - is a common place for
446beginning developers to go wrong.  Somebody who asks a networking-related
447question on linux-kernel will almost certainly receive a polite suggestion
448to ask on the netdev list instead, as that is the list frequented by most
449networking developers.  Other lists exist for the SCSI, video4linux, IDE,
450filesystem, etc. subsystems.  The best place to look for mailing lists is
451in the MAINTAINERS file packaged with the kernel source.
452
453
454Getting started with Kernel development
455---------------------------------------
456
457Questions about how to get started with the kernel development process are
458common - from both individuals and companies.  Equally common are missteps
459which make the beginning of the relationship harder than it has to be.
460
461Companies often look to hire well-known developers to get a development
462group started.  This can, in fact, be an effective technique.  But it also
463tends to be expensive and does not do much to grow the pool of experienced
464kernel developers.  It is possible to bring in-house developers up to speed
465on Linux kernel development, given the investment of a bit of time.  Taking
466this time can endow an employer with a group of developers who understand
467the kernel and the company both, and who can help to train others as well.
468Over the medium term, this is often the more profitable approach.
469
470Individual developers are often, understandably, at a loss for a place to
471start.  Beginning with a large project can be intimidating; one often wants
472to test the waters with something smaller first.  This is the point where
473some developers jump into the creation of patches fixing spelling errors or
474minor coding style issues.  Unfortunately, such patches create a level of
475noise which is distracting for the development community as a whole, so,
476increasingly, they are looked down upon.  New developers wishing to
477introduce themselves to the community will not get the sort of reception
478they wish for by these means.
479
480Andrew Morton gives this advice for aspiring kernel developers
481
482::
483
484	The #1 project for all kernel beginners should surely be "make sure
485	that the kernel runs perfectly at all times on all machines which
486	you can lay your hands on".  Usually the way to do this is to work
487	with others on getting things fixed up (this can require
488	persistence!) but that's fine - it's a part of kernel development.
489
490(https://lwn.net/Articles/283982/).
491
492In the absence of obvious problems to fix, developers are advised to look
493at the current lists of regressions and open bugs in general.  There is
494never any shortage of issues in need of fixing; by addressing these issues,
495developers will gain experience with the process while, at the same time,
496building respect with the rest of the development community.
497