Based on kernel version 4.1. Page generated on 2015-06-28 12:08 EST.
1 2: HOW THE DEVELOPMENT PROCESS WORKS 2 3 Linux kernel development in the early 1990's was a pretty loose affair, 4 with relatively small numbers of users and developers involved. With a 5 user base in the millions and with some 2,000 developers involved over the 6 course of one year, the kernel has since had to evolve a number of 7 processes to keep development happening smoothly. A solid understanding of 8 how the process works is required in order to be an effective part of it. 9 10 11 2.1: THE BIG PICTURE 12 13 The kernel developers use a loosely time-based release process, with a new 14 major kernel release happening every two or three months. The recent 15 release history looks like this: 16 17 2.6.38 March 14, 2011 18 2.6.37 January 4, 2011 19 2.6.36 October 20, 2010 20 2.6.35 August 1, 2010 21 2.6.34 May 15, 2010 22 2.6.33 February 24, 2010 23 24 Every 2.6.x release is a major kernel release with new features, internal 25 API changes, and more. A typical 2.6 release can contain nearly 10,000 26 changesets with changes to several hundred thousand lines of code. 2.6 is 27 thus the leading edge of Linux kernel development; the kernel uses a 28 rolling development model which is continually integrating major changes. 29 30 A relatively straightforward discipline is followed with regard to the 31 merging of patches for each release. At the beginning of each development 32 cycle, the "merge window" is said to be open. At that time, code which is 33 deemed to be sufficiently stable (and which is accepted by the development 34 community) is merged into the mainline kernel. The bulk of changes for a 35 new development cycle (and all of the major changes) will be merged during 36 this time, at a rate approaching 1,000 changes ("patches," or "changesets") 37 per day. 38 39 (As an aside, it is worth noting that the changes integrated during the 40 merge window do not come out of thin air; they have been collected, tested, 41 and staged ahead of time. How that process works will be described in 42 detail later on). 43 44 The merge window lasts for approximately two weeks. At the end of this 45 time, Linus Torvalds will declare that the window is closed and release the 46 first of the "rc" kernels. For the kernel which is destined to be 2.6.40, 47 for example, the release which happens at the end of the merge window will 48 be called 2.6.40-rc1. The -rc1 release is the signal that the time to 49 merge new features has passed, and that the time to stabilize the next 50 kernel has begun. 51 52 Over the next six to ten weeks, only patches which fix problems should be 53 submitted to the mainline. On occasion a more significant change will be 54 allowed, but such occasions are rare; developers who try to merge new 55 features outside of the merge window tend to get an unfriendly reception. 56 As a general rule, if you miss the merge window for a given feature, the 57 best thing to do is to wait for the next development cycle. (An occasional 58 exception is made for drivers for previously-unsupported hardware; if they 59 touch no in-tree code, they cannot cause regressions and should be safe to 60 add at any time). 61 62 As fixes make their way into the mainline, the patch rate will slow over 63 time. Linus releases new -rc kernels about once a week; a normal series 64 will get up to somewhere between -rc6 and -rc9 before the kernel is 65 considered to be sufficiently stable and the final 2.6.x release is made. 66 At that point the whole process starts over again. 67 68 As an example, here is how the 2.6.38 development cycle went (all dates in 69 2011): 70 71 January 4 2.6.37 stable release 72 January 18 2.6.38-rc1, merge window closes 73 January 21 2.6.38-rc2 74 February 1 2.6.38-rc3 75 February 7 2.6.38-rc4 76 February 15 2.6.38-rc5 77 February 21 2.6.38-rc6 78 March 1 2.6.38-rc7 79 March 7 2.6.38-rc8 80 March 14 2.6.38 stable release 81 82 How do the developers decide when to close the development cycle and create 83 the stable release? The most significant metric used is the list of 84 regressions from previous releases. No bugs are welcome, but those which 85 break systems which worked in the past are considered to be especially 86 serious. For this reason, patches which cause regressions are looked upon 87 unfavorably and are quite likely to be reverted during the stabilization 88 period. 89 90 The developers' goal is to fix all known regressions before the stable 91 release is made. In the real world, this kind of perfection is hard to 92 achieve; there are just too many variables in a project of this size. 93 There comes a point where delaying the final release just makes the problem 94 worse; the pile of changes waiting for the next merge window will grow 95 larger, creating even more regressions the next time around. So most 2.6.x 96 kernels go out with a handful of known regressions though, hopefully, none 97 of them are serious. 98 99 Once a stable release is made, its ongoing maintenance is passed off to the 100 "stable team," currently consisting of Greg Kroah-Hartman. The stable team 101 will release occasional updates to the stable release using the 2.6.x.y 102 numbering scheme. To be considered for an update release, a patch must (1) 103 fix a significant bug, and (2) already be merged into the mainline for the 104 next development kernel. Kernels will typically receive stable updates for 105 a little more than one development cycle past their initial release. So, 106 for example, the 2.6.36 kernel's history looked like: 107 108 October 10 2.6.36 stable release 109 November 22 126.96.36.199 110 December 9 188.8.131.52 111 January 7 184.108.40.206 112 February 17 220.127.116.11 113 114 18.104.22.168 was the final stable update for the 2.6.36 release. 115 116 Some kernels are designated "long term" kernels; they will receive support 117 for a longer period. As of this writing, the current long term kernels 118 and their maintainers are: 119 120 2.6.27 Willy Tarreau (Deep-frozen stable kernel) 121 2.6.32 Greg Kroah-Hartman 122 2.6.35 Andi Kleen (Embedded flag kernel) 123 124 The selection of a kernel for long-term support is purely a matter of a 125 maintainer having the need and the time to maintain that release. There 126 are no known plans for long-term support for any specific upcoming 127 release. 128 129 130 2.2: THE LIFECYCLE OF A PATCH 131 132 Patches do not go directly from the developer's keyboard into the mainline 133 kernel. There is, instead, a somewhat involved (if somewhat informal) 134 process designed to ensure that each patch is reviewed for quality and that 135 each patch implements a change which is desirable to have in the mainline. 136 This process can happen quickly for minor fixes, or, in the case of large 137 and controversial changes, go on for years. Much developer frustration 138 comes from a lack of understanding of this process or from attempts to 139 circumvent it. 140 141 In the hopes of reducing that frustration, this document will describe how 142 a patch gets into the kernel. What follows below is an introduction which 143 describes the process in a somewhat idealized way. A much more detailed 144 treatment will come in later sections. 145 146 The stages that a patch goes through are, generally: 147 148 - Design. This is where the real requirements for the patch - and the way 149 those requirements will be met - are laid out. Design work is often 150 done without involving the community, but it is better to do this work 151 in the open if at all possible; it can save a lot of time redesigning 152 things later. 153 154 - Early review. Patches are posted to the relevant mailing list, and 155 developers on that list reply with any comments they may have. This 156 process should turn up any major problems with a patch if all goes 157 well. 158 159 - Wider review. When the patch is getting close to ready for mainline 160 inclusion, it should be accepted by a relevant subsystem maintainer - 161 though this acceptance is not a guarantee that the patch will make it 162 all the way to the mainline. The patch will show up in the maintainer's 163 subsystem tree and into the -next trees (described below). When the 164 process works, this step leads to more extensive review of the patch and 165 the discovery of any problems resulting from the integration of this 166 patch with work being done by others. 167 168 - Please note that most maintainers also have day jobs, so merging 169 your patch may not be their highest priority. If your patch is 170 getting feedback about changes that are needed, you should either 171 make those changes or justify why they should not be made. If your 172 patch has no review complaints but is not being merged by its 173 appropriate subsystem or driver maintainer, you should be persistent 174 in updating the patch to the current kernel so that it applies cleanly 175 and keep sending it for review and merging. 176 177 - Merging into the mainline. Eventually, a successful patch will be 178 merged into the mainline repository managed by Linus Torvalds. More 179 comments and/or problems may surface at this time; it is important that 180 the developer be responsive to these and fix any issues which arise. 181 182 - Stable release. The number of users potentially affected by the patch 183 is now large, so, once again, new problems may arise. 184 185 - Long-term maintenance. While it is certainly possible for a developer 186 to forget about code after merging it, that sort of behavior tends to 187 leave a poor impression in the development community. Merging code 188 eliminates some of the maintenance burden, in that others will fix 189 problems caused by API changes. But the original developer should 190 continue to take responsibility for the code if it is to remain useful 191 in the longer term. 192 193 One of the largest mistakes made by kernel developers (or their employers) 194 is to try to cut the process down to a single "merging into the mainline" 195 step. This approach invariably leads to frustration for everybody 196 involved. 197 198 199 2.3: HOW PATCHES GET INTO THE KERNEL 200 201 There is exactly one person who can merge patches into the mainline kernel 202 repository: Linus Torvalds. But, of the over 9,500 patches which went 203 into the 2.6.38 kernel, only 112 (around 1.3%) were directly chosen by Linus 204 himself. The kernel project has long since grown to a size where no single 205 developer could possibly inspect and select every patch unassisted. The 206 way the kernel developers have addressed this growth is through the use of 207 a lieutenant system built around a chain of trust. 208 209 The kernel code base is logically broken down into a set of subsystems: 210 networking, specific architecture support, memory management, video 211 devices, etc. Most subsystems have a designated maintainer, a developer 212 who has overall responsibility for the code within that subsystem. These 213 subsystem maintainers are the gatekeepers (in a loose way) for the portion 214 of the kernel they manage; they are the ones who will (usually) accept a 215 patch for inclusion into the mainline kernel. 216 217 Subsystem maintainers each manage their own version of the kernel source 218 tree, usually (but certainly not always) using the git source management 219 tool. Tools like git (and related tools like quilt or mercurial) allow 220 maintainers to track a list of patches, including authorship information 221 and other metadata. At any given time, the maintainer can identify which 222 patches in his or her repository are not found in the mainline. 223 224 When the merge window opens, top-level maintainers will ask Linus to "pull" 225 the patches they have selected for merging from their repositories. If 226 Linus agrees, the stream of patches will flow up into his repository, 227 becoming part of the mainline kernel. The amount of attention that Linus 228 pays to specific patches received in a pull operation varies. It is clear 229 that, sometimes, he looks quite closely. But, as a general rule, Linus 230 trusts the subsystem maintainers to not send bad patches upstream. 231 232 Subsystem maintainers, in turn, can pull patches from other maintainers. 233 For example, the networking tree is built from patches which accumulated 234 first in trees dedicated to network device drivers, wireless networking, 235 etc. This chain of repositories can be arbitrarily long, though it rarely 236 exceeds two or three links. Since each maintainer in the chain trusts 237 those managing lower-level trees, this process is known as the "chain of 238 trust." 239 240 Clearly, in a system like this, getting patches into the kernel depends on 241 finding the right maintainer. Sending patches directly to Linus is not 242 normally the right way to go. 243 244 245 2.4: NEXT TREES 246 247 The chain of subsystem trees guides the flow of patches into the kernel, 248 but it also raises an interesting question: what if somebody wants to look 249 at all of the patches which are being prepared for the next merge window? 250 Developers will be interested in what other changes are pending to see 251 whether there are any conflicts to worry about; a patch which changes a 252 core kernel function prototype, for example, will conflict with any other 253 patches which use the older form of that function. Reviewers and testers 254 want access to the changes in their integrated form before all of those 255 changes land in the mainline kernel. One could pull changes from all of 256 the interesting subsystem trees, but that would be a big and error-prone 257 job. 258 259 The answer comes in the form of -next trees, where subsystem trees are 260 collected for testing and review. The older of these trees, maintained by 261 Andrew Morton, is called "-mm" (for memory management, which is how it got 262 started). The -mm tree integrates patches from a long list of subsystem 263 trees; it also has some patches aimed at helping with debugging. 264 265 Beyond that, -mm contains a significant collection of patches which have 266 been selected by Andrew directly. These patches may have been posted on a 267 mailing list, or they may apply to a part of the kernel for which there is 268 no designated subsystem tree. As a result, -mm operates as a sort of 269 subsystem tree of last resort; if there is no other obvious path for a 270 patch into the mainline, it is likely to end up in -mm. Miscellaneous 271 patches which accumulate in -mm will eventually either be forwarded on to 272 an appropriate subsystem tree or be sent directly to Linus. In a typical 273 development cycle, approximately 5-10% of the patches going into the 274 mainline get there via -mm. 275 276 The current -mm patch is available in the "mmotm" (-mm of the moment) 277 directory at: 278 279 http://www.ozlabs.org/~akpm/mmotm/ 280 281 Use of the MMOTM tree is likely to be a frustrating experience, though; 282 there is a definite chance that it will not even compile. 283 284 The primary tree for next-cycle patch merging is linux-next, maintained by 285 Stephen Rothwell. The linux-next tree is, by design, a snapshot of what 286 the mainline is expected to look like after the next merge window closes. 287 Linux-next trees are announced on the linux-kernel and linux-next mailing 288 lists when they are assembled; they can be downloaded from: 289 290 http://www.kernel.org/pub/linux/kernel/next/ 291 292 Linux-next has become an integral part of the kernel development process; 293 all patches merged during a given merge window should really have found 294 their way into linux-next some time before the merge window opens. 295 296 297 2.4.1: STAGING TREES 298 299 The kernel source tree contains the drivers/staging/ directory, where 300 many sub-directories for drivers or filesystems that are on their way to 301 being added to the kernel tree live. They remain in drivers/staging while 302 they still need more work; once complete, they can be moved into the 303 kernel proper. This is a way to keep track of drivers that aren't 304 up to Linux kernel coding or quality standards, but people may want to use 305 them and track development. 306 307 Greg Kroah-Hartman currently maintains the staging tree. Drivers that 308 still need work are sent to him, with each driver having its own 309 subdirectory in drivers/staging/. Along with the driver source files, a 310 TODO file should be present in the directory as well. The TODO file lists 311 the pending work that the driver needs for acceptance into the kernel 312 proper, as well as a list of people that should be Cc'd for any patches to 313 the driver. Current rules require that drivers contributed to staging 314 must, at a minimum, compile properly. 315 316 Staging can be a relatively easy way to get new drivers into the mainline 317 where, with luck, they will come to the attention of other developers and 318 improve quickly. Entry into staging is not the end of the story, though; 319 code in staging which is not seeing regular progress will eventually be 320 removed. Distributors also tend to be relatively reluctant to enable 321 staging drivers. So staging is, at best, a stop on the way toward becoming 322 a proper mainline driver. 323 324 325 2.5: TOOLS 326 327 As can be seen from the above text, the kernel development process depends 328 heavily on the ability to herd collections of patches in various 329 directions. The whole thing would not work anywhere near as well as it 330 does without suitably powerful tools. Tutorials on how to use these tools 331 are well beyond the scope of this document, but there is space for a few 332 pointers. 333 334 By far the dominant source code management system used by the kernel 335 community is git. Git is one of a number of distributed version control 336 systems being developed in the free software community. It is well tuned 337 for kernel development, in that it performs quite well when dealing with 338 large repositories and large numbers of patches. It also has a reputation 339 for being difficult to learn and use, though it has gotten better over 340 time. Some sort of familiarity with git is almost a requirement for kernel 341 developers; even if they do not use it for their own work, they'll need git 342 to keep up with what other developers (and the mainline) are doing. 343 344 Git is now packaged by almost all Linux distributions. There is a home 345 page at: 346 347 http://git-scm.com/ 348 349 That page has pointers to documentation and tutorials. 350 351 Among the kernel developers who do not use git, the most popular choice is 352 almost certainly Mercurial: 353 354 http://www.selenic.com/mercurial/ 355 356 Mercurial shares many features with git, but it provides an interface which 357 many find easier to use. 358 359 The other tool worth knowing about is Quilt: 360 361 http://savannah.nongnu.org/projects/quilt/ 362 363 Quilt is a patch management system, rather than a source code management 364 system. It does not track history over time; it is, instead, oriented 365 toward tracking a specific set of changes against an evolving code base. 366 Some major subsystem maintainers use quilt to manage patches intended to go 367 upstream. For the management of certain kinds of trees (-mm, for example), 368 quilt is the best tool for the job. 369 370 371 2.6: MAILING LISTS 372 373 A great deal of Linux kernel development work is done by way of mailing 374 lists. It is hard to be a fully-functioning member of the community 375 without joining at least one list somewhere. But Linux mailing lists also 376 represent a potential hazard to developers, who risk getting buried under a 377 load of electronic mail, running afoul of the conventions used on the Linux 378 lists, or both. 379 380 Most kernel mailing lists are run on vger.kernel.org; the master list can 381 be found at: 382 383 http://vger.kernel.org/vger-lists.html 384 385 There are lists hosted elsewhere, though; a number of them are at 386 lists.redhat.com. 387 388 The core mailing list for kernel development is, of course, linux-kernel. 389 This list is an intimidating place to be; volume can reach 500 messages per 390 day, the amount of noise is high, the conversation can be severely 391 technical, and participants are not always concerned with showing a high 392 degree of politeness. But there is no other place where the kernel 393 development community comes together as a whole; developers who avoid this 394 list will miss important information. 395 396 There are a few hints which can help with linux-kernel survival: 397 398 - Have the list delivered to a separate folder, rather than your main 399 mailbox. One must be able to ignore the stream for sustained periods of 400 time. 401 402 - Do not try to follow every conversation - nobody else does. It is 403 important to filter on both the topic of interest (though note that 404 long-running conversations can drift away from the original subject 405 without changing the email subject line) and the people who are 406 participating. 407 408 - Do not feed the trolls. If somebody is trying to stir up an angry 409 response, ignore them. 410 411 - When responding to linux-kernel email (or that on other lists) preserve 412 the Cc: header for all involved. In the absence of a strong reason (such 413 as an explicit request), you should never remove recipients. Always make 414 sure that the person you are responding to is in the Cc: list. This 415 convention also makes it unnecessary to explicitly ask to be copied on 416 replies to your postings. 417 418 - Search the list archives (and the net as a whole) before asking 419 questions. Some developers can get impatient with people who clearly 420 have not done their homework. 421 422 - Avoid top-posting (the practice of putting your answer above the quoted 423 text you are responding to). It makes your response harder to read and 424 makes a poor impression. 425 426 - Ask on the correct mailing list. Linux-kernel may be the general meeting 427 point, but it is not the best place to find developers from all 428 subsystems. 429 430 The last point - finding the correct mailing list - is a common place for 431 beginning developers to go wrong. Somebody who asks a networking-related 432 question on linux-kernel will almost certainly receive a polite suggestion 433 to ask on the netdev list instead, as that is the list frequented by most 434 networking developers. Other lists exist for the SCSI, video4linux, IDE, 435 filesystem, etc. subsystems. The best place to look for mailing lists is 436 in the MAINTAINERS file packaged with the kernel source. 437 438 439 2.7: GETTING STARTED WITH KERNEL DEVELOPMENT 440 441 Questions about how to get started with the kernel development process are 442 common - from both individuals and companies. Equally common are missteps 443 which make the beginning of the relationship harder than it has to be. 444 445 Companies often look to hire well-known developers to get a development 446 group started. This can, in fact, be an effective technique. But it also 447 tends to be expensive and does not do much to grow the pool of experienced 448 kernel developers. It is possible to bring in-house developers up to speed 449 on Linux kernel development, given the investment of a bit of time. Taking 450 this time can endow an employer with a group of developers who understand 451 the kernel and the company both, and who can help to train others as well. 452 Over the medium term, this is often the more profitable approach. 453 454 Individual developers are often, understandably, at a loss for a place to 455 start. Beginning with a large project can be intimidating; one often wants 456 to test the waters with something smaller first. This is the point where 457 some developers jump into the creation of patches fixing spelling errors or 458 minor coding style issues. Unfortunately, such patches create a level of 459 noise which is distracting for the development community as a whole, so, 460 increasingly, they are looked down upon. New developers wishing to 461 introduce themselves to the community will not get the sort of reception 462 they wish for by these means. 463 464 Andrew Morton gives this advice for aspiring kernel developers 465 466 The #1 project for all kernel beginners should surely be "make sure 467 that the kernel runs perfectly at all times on all machines which 468 you can lay your hands on". Usually the way to do this is to work 469 with others on getting things fixed up (this can require 470 persistence!) but that's fine - it's a part of kernel development. 471 472 (http://lwn.net/Articles/283982/). 473 474 In the absence of obvious problems to fix, developers are advised to look 475 at the current lists of regressions and open bugs in general. There is 476 never any shortage of issues in need of fixing; by addressing these issues, 477 developers will gain experience with the process while, at the same time, 478 building respect with the rest of the development community.