blob: 08e47a18d9fe38933c20648d2a8143902b2d11ca [file] [log] [blame]
Reid Spencer8d0ac692007-02-13 09:06:01 +00001<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"
2 "http://www.w3.org/TR/html4/strict.dtd">
3<html>
4<head>
Bill Wendling19a8dca2009-04-05 12:38:44 +00005 <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
Reid Spencer8d0ac692007-02-13 09:06:01 +00006 <title>LLVM Developer Policy</title>
7 <link rel="stylesheet" href="llvm.css" type="text/css">
8</head>
9<body>
Reid Spencer8d0ac692007-02-13 09:06:01 +000010
NAKAMURA Takumi05d02652011-04-18 23:59:50 +000011<h1>LLVM Developer Policy</h1>
Reid Spencer8d0ac692007-02-13 09:06:01 +000012<ol>
Reid Spencer78bade22007-02-14 07:22:19 +000013 <li><a href="#introduction">Introduction</a></li>
Chris Lattner1ff20cd2007-02-19 06:05:58 +000014 <li><a href="#policies">Developer Policies</a>
Reid Spencer8d0ac692007-02-13 09:06:01 +000015 <ol>
Chris Lattner1653fec2007-02-19 05:57:29 +000016 <li><a href="#informed">Stay Informed</a></li>
17 <li><a href="#patches">Making a Patch</a></li>
Reid Spencer8d0ac692007-02-13 09:06:01 +000018 <li><a href="#reviews">Code Reviews</a></li>
Chris Lattnere268a402007-12-03 19:00:47 +000019 <li><a href="#owners">Code Owners</a></li>
Reid Spencer8d0ac692007-02-13 09:06:01 +000020 <li><a href="#testcases">Test Cases</a></li>
Chris Lattner1acdc952007-02-19 05:49:11 +000021 <li><a href="#quality">Quality</a></li>
22 <li><a href="#commitaccess">Obtaining Commit Access</a></li>
Chris Lattner1ff20cd2007-02-19 06:05:58 +000023 <li><a href="#newwork">Making a Major Change</a></li>
24 <li><a href="#incremental">Incremental Development</a></li>
Chris Lattnerc7d954e2007-02-19 05:43:04 +000025 <li><a href="#attribution">Attribution of Changes</a></li>
Reid Spencer8d0ac692007-02-13 09:06:01 +000026 </ol></li>
Chris Lattner793aa382007-02-19 06:19:16 +000027 <li><a href="#clp">Copyright, License, and Patents</a>
Reid Spencer78bade22007-02-14 07:22:19 +000028 <ol>
29 <li><a href="#copyright">Copyright</a></li>
30 <li><a href="#license">License</a></li>
Chris Lattner1ff20cd2007-02-19 06:05:58 +000031 <li><a href="#patents">Patents</a></li>
Reid Spencer78bade22007-02-14 07:22:19 +000032 </ol></li>
Reid Spencer8d0ac692007-02-13 09:06:01 +000033</ol>
Chris Lattner2ae49dd2007-02-19 06:24:23 +000034<div class="doc_author">Written by the LLVM Oversight Team</div>
Reid Spencer8d0ac692007-02-13 09:06:01 +000035
36<!--=========================================================================-->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +000037<h2><a name="introduction">Introduction</a></h2>
Reid Spencer8d0ac692007-02-13 09:06:01 +000038<!--=========================================================================-->
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +000039<div>
Bill Wendling78c8fce2009-04-05 12:37:44 +000040<p>This document contains the LLVM Developer Policy which defines the project's
41 policy towards developers and their contributions. The intent of this policy
42 is to eliminate miscommunication, rework, and confusion that might arise from
43 the distributed nature of LLVM's development. By stating the policy in clear
44 terms, we hope each developer can know ahead of time what to expect when
Chris Lattnerb742ff02010-09-02 00:09:17 +000045 making LLVM contributions. This policy covers all llvm.org subprojects,
Chris Lattner48717012012-02-08 07:58:38 +000046 including Clang, LLDB, libc++, etc.</p>
Bill Wendling78c8fce2009-04-05 12:37:44 +000047<p>This policy is also designed to accomplish the following objectives:</p>
48
49<ol>
50 <li>Attract both users and developers to the LLVM project.</li>
51
52 <li>Make life as simple and easy for contributors as possible.</li>
53
54 <li>Keep the top of Subversion trees as stable as possible.</li>
Chris Lattner48717012012-02-08 07:58:38 +000055
56 <li>Establish awareness of the project's <a href="#clp">copyright,
Chris Lattner45261952012-02-08 22:20:00 +000057 license, and patent policies</a> with contributors to the project.</li>
Bill Wendling78c8fce2009-04-05 12:37:44 +000058</ol>
Chris Lattnerc7d954e2007-02-19 05:43:04 +000059
Bill Wendling78c8fce2009-04-05 12:37:44 +000060<p>This policy is aimed at frequent contributors to LLVM. People interested in
61 contributing one-off patches can do so in an informal way by sending them to
62 the
63 <a href="http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits">llvm-commits
64 mailing list</a> and engaging another developer to see it through the
65 process.</p>
Reid Spencer8d0ac692007-02-13 09:06:01 +000066</div>
Reid Spencer8d0ac692007-02-13 09:06:01 +000067
68<!--=========================================================================-->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +000069<h2><a name="policies">Developer Policies</a></h2>
Reid Spencer8d0ac692007-02-13 09:06:01 +000070<!--=========================================================================-->
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +000071<div>
Bill Wendling78c8fce2009-04-05 12:37:44 +000072<p>This section contains policies that pertain to frequent LLVM developers. We
73 always welcome <a href="#patches">one-off patches</a> from people who do not
74 routinely contribute to LLVM, but we expect more from frequent contributors
75 to keep the system as efficient as possible for everyone. Frequent LLVM
76 contributors are expected to meet the following requirements in order for
77 LLVM to maintain a high standard of quality.<p>
Reid Spencer8d0ac692007-02-13 09:06:01 +000078
79<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +000080<h3><a name="informed">Stay Informed</a></h3>
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +000081<div>
Chris Lattnerb742ff02010-09-02 00:09:17 +000082<p>Developers should stay informed by reading at least the "dev" mailing list
83 for the projects you are interested in, such as
84 <a href="http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev">llvmdev</a> for
85 LLVM, <a href="http://lists.cs.uiuc.edu/mailman/listinfo/cfe-dev">cfe-dev</a>
86 for Clang, or <a
87 href="http://lists.cs.uiuc.edu/mailman/listinfo/lldb-dev">lldb-dev</a>
88 for LLDB. If you are doing anything more than just casual work on LLVM, it
89 is suggested that you also subscribe to the "commits" mailing list for the
90 subproject you're interested in, such as
91 <a href="http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits">llvm-commits</a>,
92 <a href="http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits">cfe-commits</a>,
93 or <a href="http://lists.cs.uiuc.edu/mailman/listinfo/lldb-commits">lldb-commits</a>.
94 Reading the "commits" list and paying attention to changes being made by
95 others is a good way to see what other people are interested in and watching
96 the flow of the project as a whole.</p>
Bill Wendling78c8fce2009-04-05 12:37:44 +000097
98<p>We recommend that active developers register an email account with
99 <a href="http://llvm.org/bugs/">LLVM Bugzilla</a> and preferably subscribe to
100 the <a href="http://lists.cs.uiuc.edu/mailman/listinfo/llvmbugs">llvm-bugs</a>
Chris Lattnerb742ff02010-09-02 00:09:17 +0000101 email list to keep track of bugs and enhancements occurring in LLVM. We
102 really appreciate people who are proactive at catching incoming bugs in their
103 components and dealing with them promptly.</p>
Reid Spencer8d0ac692007-02-13 09:06:01 +0000104</div>
105
106<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000107<h3><a name="patches">Making a Patch</a></h3>
Chris Lattner1653fec2007-02-19 05:57:29 +0000108
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000109<div>
Bill Wendling78c8fce2009-04-05 12:37:44 +0000110<p>When making a patch for review, the goal is to make it as easy for the
111 reviewer to read it as possible. As such, we recommend that you:</p>
Chris Lattner1653fec2007-02-19 05:57:29 +0000112
Bill Wendling78c8fce2009-04-05 12:37:44 +0000113<ol>
114 <li>Make your patch against the Subversion trunk, not a branch, and not an old
Chris Lattner9b96e802009-10-10 21:37:16 +0000115 version of LLVM. This makes it easy to apply the patch. For information
116 on how to check out SVN trunk, please see the <a
117 href="GettingStarted.html#checkout">Getting Started Guide</a>.</li>
Chris Lattner1653fec2007-02-19 05:57:29 +0000118
Bill Wendling78c8fce2009-04-05 12:37:44 +0000119 <li>Similarly, patches should be submitted soon after they are generated. Old
120 patches may not apply correctly if the underlying code changes between the
121 time the patch was created and the time it is applied.</li>
122
Dan Gohmanc1076ea2010-08-04 16:07:22 +0000123 <li>Patches should be made with <tt>svn diff</tt>, or similar. If you use
124 a different tool, make sure it uses the <tt>diff -u</tt> format and
125 that it doesn't contain clutter which makes it hard to read.</li>
Bill Wendling78c8fce2009-04-05 12:37:44 +0000126
Dan Gohmanc1076ea2010-08-04 16:07:22 +0000127 <li>If you are modifying generated files, such as the top-level
128 <tt>configure</tt> script, please separate out those changes into
129 a separate patch from the rest of your changes.</li>
Bill Wendling78c8fce2009-04-05 12:37:44 +0000130</ol>
Chris Lattner0cca50c2007-02-19 06:57:16 +0000131
Bill Wendling78c8fce2009-04-05 12:37:44 +0000132<p>When sending a patch to a mailing list, it is a good idea to send it as an
133 <em>attachment</em> to the message, not embedded into the text of the
134 message. This ensures that your mailer will not mangle the patch when it
135 sends it (e.g. by making whitespace changes or by wrapping lines).</p>
136
137<p><em>For Thunderbird users:</em> Before submitting a patch, please open
138 <em>Preferences &#8594; Advanced &#8594; General &#8594; Config Editor</em>,
139 find the key <tt>mail.content_disposition_type</tt>, and set its value to
140 <tt>1</tt>. Without this setting, Thunderbird sends your attachment using
141 <tt>Content-Disposition: inline</tt> rather than <tt>Content-Disposition:
142 attachment</tt>. Apple Mail gamely displays such a file inline, making it
143 difficult to work with for reviewers using that program.</p>
Chris Lattner1653fec2007-02-19 05:57:29 +0000144</div>
145
146<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000147<h3><a name="reviews">Code Reviews</a></h3>
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000148<div>
Bill Wendling78c8fce2009-04-05 12:37:44 +0000149<p>LLVM has a code review policy. Code review is one way to increase the quality
150 of software. We generally follow these policies:</p>
Chris Lattner0cca50c2007-02-19 06:57:16 +0000151
Bill Wendling78c8fce2009-04-05 12:37:44 +0000152<ol>
153 <li>All developers are required to have significant changes reviewed before
154 they are committed to the repository.</li>
155
156 <li>Code reviews are conducted by email, usually on the llvm-commits
157 list.</li>
158
159 <li>Code can be reviewed either before it is committed or after. We expect
160 major changes to be reviewed before being committed, but smaller changes
161 (or changes where the developer owns the component) can be reviewed after
162 commit.</li>
163
164 <li>The developer responsible for a code change is also responsible for making
165 all necessary review-related changes.</li>
166
167 <li>Code review can be an iterative process, which continues until the patch
168 is ready to be committed.</li>
169</ol>
170
171<p>Developers should participate in code reviews as both reviewers and
172 reviewees. If someone is kind enough to review your code, you should return
173 the favor for someone else. Note that anyone is welcome to review and give
174 feedback on a patch, but only people with Subversion write access can approve
175 it.</p>
Reid Spencer8d0ac692007-02-13 09:06:01 +0000176</div>
Reid Spencer8d0ac692007-02-13 09:06:01 +0000177
178<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000179<h3><a name="owners">Code Owners</a></h3>
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000180<div>
Chris Lattnere268a402007-12-03 19:00:47 +0000181
Bill Wendling78c8fce2009-04-05 12:37:44 +0000182<p>The LLVM Project relies on two features of its process to maintain rapid
183 development in addition to the high quality of its source base: the
184 combination of code review plus post-commit review for trusted maintainers.
185 Having both is a great way for the project to take advantage of the fact that
186 most people do the right thing most of the time, and only commit patches
187 without pre-commit review when they are confident they are right.</p>
Chris Lattnere268a402007-12-03 19:00:47 +0000188
Bill Wendling78c8fce2009-04-05 12:37:44 +0000189<p>The trick to this is that the project has to guarantee that all patches that
190 are committed are reviewed after they go in: you don't want everyone to
191 assume someone else will review it, allowing the patch to go unreviewed. To
192 solve this problem, we have a notion of an 'owner' for a piece of the code.
193 The sole responsibility of a code owner is to ensure that a commit to their
194 area of the code is appropriately reviewed, either by themself or by someone
195 else. The current code owners are:</p>
Chris Lattnere268a402007-12-03 19:00:47 +0000196
Bill Wendling78c8fce2009-04-05 12:37:44 +0000197<ol>
Bill Wendling78c8fce2009-04-05 12:37:44 +0000198 <li><b>Evan Cheng</b>: Code generator and all targets.</li>
199
Chris Lattner5ce89812010-09-23 17:27:54 +0000200 <li><b>Greg Clayton</b>: LLDB.</li>
201
202 <li><b>Doug Gregor</b>: Clang Frontend Libraries.</li>
203
204 <li><b>Howard Hinnant</b>: libc++.</li>
Chris Lattneraf5bd672009-09-16 05:36:07 +0000205
Chris Lattnerb2030432009-09-16 05:37:13 +0000206 <li><b>Anton Korobeynikov</b>: Exception handling, debug information, and
207 Windows codegen.</li>
208
Chris Lattnerf86a7782009-09-16 05:42:12 +0000209 <li><b>Ted Kremenek</b>: Clang Static Analyzer.</li>
210
Chris Lattner941f4cd2009-09-16 05:36:54 +0000211 <li><b>Chris Lattner</b>: Everything not covered by someone else.</li>
Chris Lattneraf5bd672009-09-16 05:36:07 +0000212
John McCall74699fd2011-08-02 01:38:19 +0000213 <li><b>John McCall</b>: Clang LLVM IR generation.</li>
214
Evan Chengb7676132011-10-07 17:26:38 +0000215 <li><b>Jakob Olesen</b>: Register allocators and TableGen.</li>
216
Chris Lattnerb0387f92011-09-23 22:46:43 +0000217 <li><b>Duncan Sands</b>: dragonegg and llvm-gcc 4.2.</li>
Chris Lattnerecc39de2012-02-22 19:17:20 +0000218
219 <li><b>Peter Collingbourne</b>: libclc.</li>
220
221 <li><b>Tobias Grosser</b>: polly.</li>
Bill Wendling78c8fce2009-04-05 12:37:44 +0000222</ol>
Chris Lattnere268a402007-12-03 19:00:47 +0000223
Bill Wendling78c8fce2009-04-05 12:37:44 +0000224<p>Note that code ownership is completely different than reviewers: anyone can
225 review a piece of code, and we welcome code review from anyone who is
226 interested. Code owners are the "last line of defense" to guarantee that all
227 patches that are committed are actually reviewed.</p>
Chris Lattnere268a402007-12-03 19:00:47 +0000228
Bill Wendling78c8fce2009-04-05 12:37:44 +0000229<p>Being a code owner is a somewhat unglamorous position, but it is incredibly
230 important for the ongoing success of the project. Because people get busy,
231 interests change, and unexpected things happen, code ownership is purely
232 opt-in, and anyone can choose to resign their "title" at any time. For now,
233 we do not have an official policy on how one gets elected to be a code
234 owner.</p>
Chris Lattnere268a402007-12-03 19:00:47 +0000235</div>
236
Chris Lattnere268a402007-12-03 19:00:47 +0000237<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000238<h3><a name="testcases">Test Cases</a></h3>
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000239<div>
Bill Wendling78c8fce2009-04-05 12:37:44 +0000240<p>Developers are required to create test cases for any bugs fixed and any new
241 features added. Some tips for getting your testcase approved:</p>
242
243<ol>
244 <li>All feature and regression test cases are added to the
245 <tt>llvm/test</tt> directory. The appropriate sub-directory should be
246 selected (see the <a href="TestingGuide.html">Testing Guide</a> for
247 details).</li>
248
249 <li>Test cases should be written in <a href="LangRef.html">LLVM assembly
250 language</a> unless the feature or regression being tested requires
251 another language (e.g. the bug being fixed or feature being implemented is
252 in the llvm-gcc C++ front-end, in which case it must be written in
253 C++).</li>
254
255 <li>Test cases, especially for regressions, should be reduced as much as
256 possible, by <a href="Bugpoint.html">bugpoint</a> or manually. It is
257 unacceptable to place an entire failing program into <tt>llvm/test</tt> as
258 this creates a <i>time-to-test</i> burden on all developers. Please keep
259 them short.</li>
260</ol>
Chris Lattner0cca50c2007-02-19 06:57:16 +0000261
Chris Lattnerb742ff02010-09-02 00:09:17 +0000262<p>Note that llvm/test and clang/test are designed for regression and small
263 feature tests only. More extensive test cases (e.g., entire applications,
264 benchmarks, etc)
Bill Wendling78c8fce2009-04-05 12:37:44 +0000265 should be added to the <tt>llvm-test</tt> test suite. The llvm-test suite is
266 for coverage (correctness, performance, etc) testing, not feature or
267 regression testing.</p>
Reid Spencer8d0ac692007-02-13 09:06:01 +0000268</div>
269
Chris Lattner450f40a2007-02-19 03:50:31 +0000270<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000271<h3><a name="quality">Quality</a></h3>
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000272<div>
Bill Wendling78c8fce2009-04-05 12:37:44 +0000273<p>The minimum quality standards that any change must satisfy before being
274 committed to the main development branch are:</p>
275
276<ol>
277 <li>Code must adhere to the <a href="CodingStandards.html">LLVM Coding
278 Standards</a>.</li>
279
280 <li>Code must compile cleanly (no errors, no warnings) on at least one
281 platform.</li>
282
283 <li>Bug fixes and new features should <a href="#testcases">include a
284 testcase</a> so we know if the fix/feature ever regresses in the
285 future.</li>
286
Chris Lattnerb742ff02010-09-02 00:09:17 +0000287 <li>Code must pass the <tt>llvm/test</tt> test suite.</li>
Bill Wendling78c8fce2009-04-05 12:37:44 +0000288
289 <li>The code must not cause regressions on a reasonable subset of llvm-test,
290 where "reasonable" depends on the contributor's judgement and the scope of
291 the change (more invasive changes require more testing). A reasonable
292 subset might be something like
293 "<tt>llvm-test/MultiSource/Benchmarks</tt>".</li>
294</ol>
295
296<p>Additionally, the committer is responsible for addressing any problems found
297 in the future that the change is responsible for. For example:</p>
298
299<ul>
300 <li>The code should compile cleanly on all supported platforms.</li>
301
302 <li>The changes should not cause any correctness regressions in the
303 <tt>llvm-test</tt> suite and must not cause any major performance
304 regressions.</li>
305
306 <li>The change set should not cause performance or correctness regressions for
307 the LLVM tools.</li>
308
309 <li>The changes should not cause performance or correctness regressions in
310 code compiled by LLVM on all applicable targets.</li>
311
312 <li>You are expected to address any <a href="http://llvm.org/bugs/">bugzilla
313 bugs</a> that result from your change.</li>
314</ul>
Chris Lattner39582bd2007-02-19 05:47:13 +0000315
Bill Wendling78c8fce2009-04-05 12:37:44 +0000316<p>We prefer for this to be handled before submission but understand that it
Chris Lattnerd7f10d32009-08-01 19:25:25 +0000317 isn't possible to test all of this for every submission. Our build bots and
318 nightly testing infrastructure normally finds these problems. A good rule of
319 thumb is to check the nightly testers for regressions the day after your
320 change. Build bots will directly email you if a group of commits that
321 included yours caused a failure. You are expected to check the build bot
322 messages to see if they are your fault and, if so, fix the breakage.</p>
Bill Wendling78c8fce2009-04-05 12:37:44 +0000323
324<p>Commits that violate these quality standards (e.g. are very broken) may be
325 reverted. This is necessary when the change blocks other developers from
326 making progress. The developer is welcome to re-commit the change after the
327 problem has been fixed.</p>
Chris Lattner39582bd2007-02-19 05:47:13 +0000328</div>
329
330<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000331<h3><a name="commitaccess">Obtaining Commit Access</a></h3>
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000332<div>
Chris Lattner450f40a2007-02-19 03:50:31 +0000333
Bill Wendling78c8fce2009-04-05 12:37:44 +0000334<p>We grant commit access to contributors with a track record of submitting high
335 quality patches. If you would like commit access, please send an email to
336 <a href="mailto:sabre@nondot.org">Chris</a> with the following
337 information:</p>
Chris Lattnerbebcdabd82007-12-03 00:36:20 +0000338
339<ol>
Bill Wendling78c8fce2009-04-05 12:37:44 +0000340 <li>The user name you want to commit with, e.g. "hacker".</li>
341
Chris Lattnerbebcdabd82007-12-03 00:36:20 +0000342 <li>The full name and email address you want message to llvm-commits to come
Bill Wendling78c8fce2009-04-05 12:37:44 +0000343 from, e.g. "J. Random Hacker &lt;hacker@yoyodyne.com&gt;".</li>
344
Chris Lattnerbebcdabd82007-12-03 00:36:20 +0000345 <li>A "password hash" of the password you want to use, e.g. "2ACR96qjUqsyM".
346 Note that you don't ever tell us what your password is, you just give it
347 to us in an encrypted form. To get this, run "htpasswd" (a utility that
348 comes with apache) in crypt mode (often enabled with "-d"), or find a web
349 page that will do it for you.</li>
350</ol>
351
352<p>Once you've been granted commit access, you should be able to check out an
353 LLVM tree with an SVN URL of "https://username@llvm.org/..." instead of the
354 normal anonymous URL of "http://llvm.org/...". The first time you commit
355 you'll have to type in your password. Note that you may get a warning from
356 SVN about an untrusted key, you can ignore this. To verify that your commit
357 access works, please do a test commit (e.g. change a comment or add a blank
358 line). Your first commit to a repository may require the autogenerated email
359 to be approved by a mailing list. This is normal, and will be done when
360 the mailing list owner has time.</p>
Chris Lattner450f40a2007-02-19 03:50:31 +0000361
362<p>If you have recently been granted commit access, these policies apply:</p>
Chris Lattnerbebcdabd82007-12-03 00:36:20 +0000363
Chris Lattner450f40a2007-02-19 03:50:31 +0000364<ol>
Bill Wendling78c8fce2009-04-05 12:37:44 +0000365 <li>You are granted <i>commit-after-approval</i> to all parts of LLVM. To get
366 approval, submit a <a href="#patches">patch</a> to
367 <a href="http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits">llvm-commits</a>.
368 When approved you may commit it yourself.</li>
369
Chris Lattner450f40a2007-02-19 03:50:31 +0000370 <li>You are allowed to commit patches without approval which you think are
Bill Wendling78c8fce2009-04-05 12:37:44 +0000371 obvious. This is clearly a subjective decision &mdash; we simply expect
372 you to use good judgement. Examples include: fixing build breakage,
373 reverting obviously broken patches, documentation/comment changes, any
374 other minor changes.</li>
375
376 <li>You are allowed to commit patches without approval to those portions of
377 LLVM that you have contributed or maintain (i.e., have been assigned
378 responsibility for), with the proviso that such commits must not break the
379 build. This is a "trust but verify" policy and commits of this nature are
380 reviewed after they are committed.</li>
381
382 <li>Multiple violations of these policies or a single egregious violation may
383 cause commit access to be revoked.</li>
Chris Lattner450f40a2007-02-19 03:50:31 +0000384</ol>
Chris Lattner0cca50c2007-02-19 06:57:16 +0000385
386<p>In any case, your changes are still subject to <a href="#reviews">code
Bill Wendling78c8fce2009-04-05 12:37:44 +0000387 review</a> (either before or after they are committed, depending on the
388 nature of the change). You are encouraged to review other peoples' patches
389 as well, but you aren't required to.</p>
Chris Lattner450f40a2007-02-19 03:50:31 +0000390</div>
391
392<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000393<h3><a name="newwork">Making a Major Change</a></h3>
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000394<div>
Bill Wendling78c8fce2009-04-05 12:37:44 +0000395<p>When a developer begins a major new project with the aim of contributing it
396 back to LLVM, s/he should inform the community with an email to
397 the <a href="http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev">llvmdev</a>
398 email list, to the extent possible. The reason for this is to:
399
400<ol>
401 <li>keep the community informed about future changes to LLVM, </li>
402
403 <li>avoid duplication of effort by preventing multiple parties working on the
404 same thing and not knowing about it, and</li>
405
406 <li>ensure that any technical issues around the proposed work are discussed
407 and resolved before any significant work is done.</li>
408</ol>
Chris Lattnerc7d954e2007-02-19 05:43:04 +0000409
Bill Wendling78c8fce2009-04-05 12:37:44 +0000410<p>The design of LLVM is carefully controlled to ensure that all the pieces fit
411 together well and are as consistent as possible. If you plan to make a major
412 change to the way LLVM works or want to add a major new extension, it is a
413 good idea to get consensus with the development community before you start
414 working on it.</p>
Chris Lattnerc7d954e2007-02-19 05:43:04 +0000415
Bill Wendling78c8fce2009-04-05 12:37:44 +0000416<p>Once the design of the new feature is finalized, the work itself should be
417 done as a series of <a href="#incremental">incremental changes</a>, not as a
418 long-term development branch.</p>
Chris Lattnerc7d954e2007-02-19 05:43:04 +0000419</div>
420
421<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000422<h3><a name="incremental">Incremental Development</a></h3>
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000423<div>
Bill Wendling78c8fce2009-04-05 12:37:44 +0000424<p>In the LLVM project, we do all significant changes as a series of incremental
425 patches. We have a strong dislike for huge changes or long-term development
426 branches. Long-term development branches have a number of drawbacks:</p>
427
428<ol>
429 <li>Branches must have mainline merged into them periodically. If the branch
430 development and mainline development occur in the same pieces of code,
431 resolving merge conflicts can take a lot of time.</li>
432
433 <li>Other people in the community tend to ignore work on branches.</li>
434
435 <li>Huge changes (produced when a branch is merged back onto mainline) are
436 extremely difficult to <a href="#reviews">code review</a>.</li>
437
438 <li>Branches are not routinely tested by our nightly tester
439 infrastructure.</li>
440
441 <li>Changes developed as monolithic large changes often don't work until the
442 entire set of changes is done. Breaking it down into a set of smaller
443 changes increases the odds that any of the work will be committed to the
444 main repository.</li>
445</ol>
Chris Lattner450f40a2007-02-19 03:50:31 +0000446
Bill Wendling78c8fce2009-04-05 12:37:44 +0000447<p>To address these problems, LLVM uses an incremental development style and we
448 require contributors to follow this practice when making a large/invasive
449 change. Some tips:</p>
450
451<ul>
452 <li>Large/invasive changes usually have a number of secondary changes that are
453 required before the big change can be made (e.g. API cleanup, etc). These
454 sorts of changes can often be done before the major change is done,
455 independently of that work.</li>
456
457 <li>The remaining inter-related work should be decomposed into unrelated sets
458 of changes if possible. Once this is done, define the first increment and
459 get consensus on what the end goal of the change is.</li>
460
461 <li>Each change in the set can be stand alone (e.g. to fix a bug), or part of
462 a planned series of changes that works towards the development goal.</li>
463
464 <li>Each change should be kept as small as possible. This simplifies your work
465 (into a logical progression), simplifies code review and reduces the
466 chance that you will get negative feedback on the change. Small increments
467 also facilitate the maintenance of a high quality code base.</li>
468
469 <li>Often, an independent precursor to a big change is to add a new API and
470 slowly migrate clients to use the new API. Each change to use the new API
471 is often "obvious" and can be committed without review. Once the new API
472 is in place and used, it is much easier to replace the underlying
473 implementation of the API. This implementation change is logically
474 separate from the API change.</li>
475</ul>
Chris Lattner450f40a2007-02-19 03:50:31 +0000476
Bill Wendling78c8fce2009-04-05 12:37:44 +0000477<p>If you are interested in making a large change, and this scares you, please
478 make sure to first <a href="#newwork">discuss the change/gather consensus</a>
479 then ask about the best way to go about making the change.</p>
Chris Lattner450f40a2007-02-19 03:50:31 +0000480</div>
481
482<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000483<h3><a name="attribution">Attribution of Changes</a></h3>
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000484<div>
Bill Wendling78c8fce2009-04-05 12:37:44 +0000485<p>We believe in correct attribution of contributions to their contributors.
486 However, we do not want the source code to be littered with random
487 attributions "this code written by J. Random Hacker" (this is noisy and
488 distracting). In practice, the revision control system keeps a perfect
489 history of who changed what, and the CREDITS.txt file describes higher-level
490 contributions. If you commit a patch for someone else, please say "patch
491 contributed by J. Random Hacker!" in the commit message.</p>
Chris Lattnerab2b10c2007-12-29 19:56:08 +0000492
Bill Wendling78c8fce2009-04-05 12:37:44 +0000493<p>Overall, please do not add contributor names to the source code.</p>
Chris Lattner450f40a2007-02-19 03:50:31 +0000494</div>
495
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000496</div>
497
Reid Spencer8d0ac692007-02-13 09:06:01 +0000498<!--=========================================================================-->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000499<h2>
Chris Lattner793aa382007-02-19 06:19:16 +0000500 <a name="clp">Copyright, License, and Patents</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000501</h2>
Reid Spencer78bade22007-02-14 07:22:19 +0000502<!--=========================================================================-->
Chris Lattner450f40a2007-02-19 03:50:31 +0000503
Chris Lattner48717012012-02-08 07:58:38 +0000504<div class="doc_notes">
505<p style="text-align:center;font-weight:bold">NOTE: This section deals with
506 legal matters but does not provide legal advice. We are not lawyers &mdash;
507 please seek legal counsel from an attorney.</p>
508</div>
509
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000510<div>
Bill Wendling78c8fce2009-04-05 12:37:44 +0000511<p>This section addresses the issues of copyright, license and patents for the
Chris Lattner48717012012-02-08 07:58:38 +0000512 LLVM project. The copyright for the code is held by the individual
Chris Lattner5d8f43f2011-09-22 18:54:31 +0000513 contributors of the code and the terms of its license to LLVM users and
514 developers is the
Bill Wendling78c8fce2009-04-05 12:37:44 +0000515 <a href="http://www.opensource.org/licenses/UoI-NCSA.php">University of
Chris Lattner48717012012-02-08 07:58:38 +0000516 Illinois/NCSA Open Source License</a> (with portions dual licensed under the
517 <a href="http://www.opensource.org/licenses/mit-license.php">MIT License</a>,
518 see below). As contributor to the LLVM project, you agree to allow any
519 contributions to the project to licensed under these terms.</p>
Reid Spencer78bade22007-02-14 07:22:19 +0000520
Reid Spencer78bade22007-02-14 07:22:19 +0000521
522<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000523<h3><a name="copyright">Copyright</a></h3>
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000524<div>
Chris Lattner305915b2008-05-20 20:06:53 +0000525
Chris Lattnerc226fc82010-09-27 21:17:15 +0000526<p>The LLVM project does not require copyright assignments, which means that the
527 copyright for the code in the project is held by its respective contributors
528 who have each agreed to release their contributed code under the terms of the
529 <a href="#license">LLVM License</a>.</p>
530
531<p>An implication of this is that the LLVM license is unlikely to ever change:
532 changing it would require tracking down all the contributors to LLVM and
533 getting them to agree that a license change is acceptable for their
534 contribution. Since there are no plans to change the license, this is not a
535 cause for concern.</p>
536
537<p>As a contributor to the project, this means that you (or your company) retain
538 ownership of the code you contribute, that it cannot be used in a way that
539 contradicts the license (which is a liberal BSD-style license), and that the
540 license for your contributions won't change without your approval in the
541 future.</p>
542
Reid Spencer78bade22007-02-14 07:22:19 +0000543</div>
Chris Lattner450f40a2007-02-19 03:50:31 +0000544
Reid Spencer78bade22007-02-14 07:22:19 +0000545<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000546<h3><a name="license">License</a></h3>
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000547<div>
Bill Wendling78c8fce2009-04-05 12:37:44 +0000548<p>We intend to keep LLVM perpetually open source and to use a liberal open
Chris Lattner48717012012-02-08 07:58:38 +0000549 source license. <b>As a contributor to the project, you agree that any
550 contributions be licensed under the terms of the corresponding
551 subproject.</b>
552 All of the code in LLVM is available under the
Bill Wendling78c8fce2009-04-05 12:37:44 +0000553 <a href="http://www.opensource.org/licenses/UoI-NCSA.php">University of
Dan Gohman621a4152010-02-26 20:18:32 +0000554 Illinois/NCSA Open Source License</a>, which boils down to this:</p>
Chris Lattner450f40a2007-02-19 03:50:31 +0000555
Bill Wendling78c8fce2009-04-05 12:37:44 +0000556<ul>
557 <li>You can freely distribute LLVM.</li>
Bill Wendling78c8fce2009-04-05 12:37:44 +0000558 <li>You must retain the copyright notice if you redistribute LLVM.</li>
Chris Lattnerac139f12010-11-16 21:32:53 +0000559 <li>Binaries derived from LLVM must reproduce the copyright notice (e.g. in an
560 included readme file).</li>
Bill Wendling78c8fce2009-04-05 12:37:44 +0000561 <li>You can't use our names to promote your LLVM derived products.</li>
Bill Wendling78c8fce2009-04-05 12:37:44 +0000562 <li>There's no warranty on LLVM at all.</li>
563</ul>
564
565<p>We believe this fosters the widest adoption of LLVM because it <b>allows
566 commercial products to be derived from LLVM</b> with few restrictions and
567 without a requirement for making any derived works also open source (i.e.
568 LLVM's license is not a "copyleft" license like the GPL). We suggest that you
569 read the <a href="http://www.opensource.org/licenses/UoI-NCSA.php">License</a>
570 if further clarification is needed.</p>
Chris Lattnerac139f12010-11-16 21:32:53 +0000571
572<p>In addition to the UIUC license, the runtime library components of LLVM
Chris Lattnerecc39de2012-02-22 19:17:20 +0000573 (<b>compiler_rt, libc++, and libclc</b>) are also licensed under the <a
Chris Lattnerac139f12010-11-16 21:32:53 +0000574 href="http://www.opensource.org/licenses/mit-license.php">MIT license</a>,
575 which does not contain the binary redistribution clause. As a user of these
576 runtime libraries, it means that you can choose to use the code under either
577 license (and thus don't need the binary redistribution clause), and as a
578 contributor to the code that you agree that any contributions to these
579 libraries be licensed under both licenses. We feel that this is important
580 for runtime libraries, because they are implicitly linked into applications
581 and therefore should not subject those applications to the binary
582 redistribution clause. This also means that it is ok to move code from (e.g.)
583 libc++ to the LLVM core without concern, but that code cannot be moved from
584 the LLVM core to libc++ without the copyright owner's permission.
585</p>
586
Chris Lattner48717012012-02-08 07:58:38 +0000587<p>Note that the LLVM Project does distribute llvm-gcc and dragonegg, <b>which
588 are GPL.</b>
Bill Wendling78c8fce2009-04-05 12:37:44 +0000589 This means that anything "linked" into llvm-gcc must itself be compatible
590 with the GPL, and must be releasable under the terms of the GPL. This
591 implies that <b>any code linked into llvm-gcc and distributed to others may
592 be subject to the viral aspects of the GPL</b> (for example, a proprietary
593 code generator linked into llvm-gcc must be made available under the GPL).
594 This is not a problem for code already distributed under a more liberal
Chris Lattner48717012012-02-08 07:58:38 +0000595 license (like the UIUC license), and GPL-containing subprojects are kept
596 in separate SVN repositories whose LICENSE.txt files specifically indicate
597 that they contain GPL code.</p>
Bill Wendling78c8fce2009-04-05 12:37:44 +0000598
599<p>We have no plans to change the license of LLVM. If you have questions or
600 comments about the license, please contact the
Chris Lattnerac139f12010-11-16 21:32:53 +0000601 <a href="mailto:llvmdev@cs.uiuc.edu">LLVM Developer's Mailing List</a>.</p>
Reid Spencer78bade22007-02-14 07:22:19 +0000602</div>
Chris Lattner1ff20cd2007-02-19 06:05:58 +0000603
604<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000605<h3><a name="patents">Patents</a></h3>
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000606<div>
Chris Lattnerb87fb2a2007-02-19 06:13:50 +0000607<p>To the best of our knowledge, LLVM does not infringe on any patents (we have
608 actually removed code from LLVM in the past that was found to infringe).
Bill Wendling78c8fce2009-04-05 12:37:44 +0000609 Having code in LLVM that infringes on patents would violate an important goal
610 of the project by making it hard or impossible to reuse the code for
Chris Lattnerd0742ce2007-02-19 06:15:33 +0000611 arbitrary purposes (including commercial use).</p>
Chris Lattnerb87fb2a2007-02-19 06:13:50 +0000612
613<p>When contributing code, we expect contributors to notify us of any potential
Chris Lattner48717012012-02-08 07:58:38 +0000614 for patent-related trouble with their changes (including from third parties).
615 If you or your employer own
Bill Wendling78c8fce2009-04-05 12:37:44 +0000616 the rights to a patent and would like to contribute code to LLVM that relies
617 on it, we require that the copyright owner sign an agreement that allows any
618 other user of LLVM to freely use your patent. Please contact
619 the <a href="mailto:llvm-oversight@cs.uiuc.edu">oversight group</a> for more
Chris Lattnerb87fb2a2007-02-19 06:13:50 +0000620 details.</p>
Chris Lattner1ff20cd2007-02-19 06:05:58 +0000621</div>
622
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000623</div>
624
Reid Spencer8d0ac692007-02-13 09:06:01 +0000625<!-- *********************************************************************** -->
626<hr>
627<address>
628 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
Misha Brukman44408702008-12-11 17:34:48 +0000629 src="http://jigsaw.w3.org/css-validator/images/vcss-blue" alt="Valid CSS"></a>
Reid Spencer8d0ac692007-02-13 09:06:01 +0000630 <a href="http://validator.w3.org/check/referer"><img
Misha Brukmanf00ddb02008-12-11 18:23:24 +0000631 src="http://www.w3.org/Icons/valid-html401-blue" alt="Valid HTML 4.01"></a>
Chris Lattnerd0742ce2007-02-19 06:15:33 +0000632 Written by the
Reid Spencer3eedbd32007-02-14 07:57:48 +0000633 <a href="mailto:llvm-oversight@cs.uiuc.edu">LLVM Oversight Group</a><br>
NAKAMURA Takumib9a33632011-04-09 02:13:37 +0000634 <a href="http://llvm.org/">The LLVM Compiler Infrastructure</a><br>
Reid Spencer8d0ac692007-02-13 09:06:01 +0000635 Last modified: $Date$
636</address>
637</body>
638</html>