Gitiles
Code Review Sign In
gerrit-public.fairphone.software / fp2-dev / platform / external / llvm / 01bb8b0c36ffb81b0ee09c01d14eb90386b4b2ca / . / docs / DeveloperPolicy.html
blob: 6e8f9703cc71d5d49a811f01396ab531863953de [file] [log] [blame]
Reid Spencer8d0ac692007-02-13 09:06:01 +0000[diff] [blame]1<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"
2 "http://www.w3.org/TR/html4/strict.dtd">
3<html>
4<head>
5 <title>LLVM Developer Policy</title>
6 <link rel="stylesheet" href="llvm.css" type="text/css">
7</head>
8<body>
Reid Spencer8d0ac692007-02-13 09:06:01 +0000[diff] [blame]9
10<div class="doc_title">LLVM Developer Policy</div>
Reid Spencer8d0ac692007-02-13 09:06:01 +0000[diff] [blame]11<ol>
Reid Spencer78bade22007-02-14 07:22:19 +0000[diff] [blame]12 <li><a href="#introduction">Introduction</a></li>
Chris Lattner1ff20cd2007-02-19 06:05:58 +0000[diff] [blame]13 <li><a href="#policies">Developer Policies</a>
Reid Spencer8d0ac692007-02-13 09:06:01 +0000[diff] [blame]14 <ol>
Chris Lattner1653fec2007-02-19 05:57:29 +0000[diff] [blame]15 <li><a href="#informed">Stay Informed</a></li>
16 <li><a href="#patches">Making a Patch</a></li>
Reid Spencer8d0ac692007-02-13 09:06:01 +0000[diff] [blame]17 <li><a href="#reviews">Code Reviews</a></li>
Chris Lattnere268a402007-12-03 19:00:47 +0000[diff] [blame]18 <li><a href="#owners">Code Owners</a></li>
Reid Spencer8d0ac692007-02-13 09:06:01 +0000[diff] [blame]19 <li><a href="#testcases">Test Cases</a></li>
Chris Lattner1acdc952007-02-19 05:49:11 +0000[diff] [blame]20 <li><a href="#quality">Quality</a></li>
21 <li><a href="#commitaccess">Obtaining Commit Access</a></li>
Chris Lattner1ff20cd2007-02-19 06:05:58 +0000[diff] [blame]22 <li><a href="#newwork">Making a Major Change</a></li>
23 <li><a href="#incremental">Incremental Development</a></li>
Chris Lattnerc7d954e2007-02-19 05:43:04 +0000[diff] [blame]24 <li><a href="#attribution">Attribution of Changes</a></li>
Reid Spencer8d0ac692007-02-13 09:06:01 +0000[diff] [blame]25 </ol></li>
Chris Lattner793aa382007-02-19 06:19:16 +0000[diff] [blame]26 <li><a href="#clp">Copyright, License, and Patents</a>
Reid Spencer78bade22007-02-14 07:22:19 +0000[diff] [blame]27 <ol>
28 <li><a href="#copyright">Copyright</a></li>
29 <li><a href="#license">License</a></li>
Chris Lattner1ff20cd2007-02-19 06:05:58 +0000[diff] [blame]30 <li><a href="#patents">Patents</a></li>
Reid Spencer78bade22007-02-14 07:22:19 +0000[diff] [blame]31 <li><a href="#devagree">Developer Agreements</a></li>
32 </ol></li>
Reid Spencer8d0ac692007-02-13 09:06:01 +0000[diff] [blame]33</ol>
Chris Lattner2ae49dd2007-02-19 06:24:23 +0000[diff] [blame]34<div class="doc_author">Written by the LLVM Oversight Team</div>
Reid Spencer8d0ac692007-02-13 09:06:01 +0000[diff] [blame]35
36<!--=========================================================================-->
Reid Spencere7bd7d62007-02-14 17:24:04 +0000[diff] [blame]37<div class="doc_section"><a name="introduction">Introduction</a></div>
Reid Spencer8d0ac692007-02-13 09:06:01 +0000[diff] [blame]38<!--=========================================================================-->
39<div class="doc_text">
40 <p>This document contains the LLVM Developer Policy which defines the
Reid Spencerbed92532007-02-13 17:04:09 +0000[diff] [blame]41 project's policy towards developers and their contributions. The intent of
Reid Spencer8d0ac692007-02-13 09:06:01 +0000[diff] [blame]42 this policy is to eliminate mis-communication, rework, and confusion that
43 might arise from the distributed nature of LLVM's development. By stating
44 the policy in clear terms, we hope each developer can know ahead of time
45 what to expect when making LLVM contributions.</p>
Reid Spencer5b9a6092007-02-14 22:55:40 +0000[diff] [blame]46 <p>This policy is also designed to accomplish the following objectives:</p>
47 <ol>
48 <li>Attract both users and developers to the LLVM project.</li>
49 <li>Make life as simple and easy for contributors as possible.</li>
Reid Spencer669ed452007-07-09 08:04:31 +0000[diff] [blame]50 <li>Keep the top of Subversion trees as stable as possible.</li>
Reid Spencer5b9a6092007-02-14 22:55:40 +0000[diff] [blame]51 </ol>
Chris Lattnerc7d954e2007-02-19 05:43:04 +0000[diff] [blame]52
Chris Lattner0cca50c2007-02-19 06:57:16 +0000[diff] [blame]53 <p>This policy is aimed at frequent contributors to LLVM. People interested in
Chris Lattnerc7d954e2007-02-19 05:43:04 +0000[diff] [blame]54 contributing one-off patches can do so in an informal way by sending them to
55 the <a href="http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits">
56 llvm-commits mailing list</a> and engaging another developer to see it through
57 the process.</p>
58
Reid Spencer8d0ac692007-02-13 09:06:01 +0000[diff] [blame]59</div>
Reid Spencer8d0ac692007-02-13 09:06:01 +0000[diff] [blame]60
61<!--=========================================================================-->
Chris Lattner1ff20cd2007-02-19 06:05:58 +0000[diff] [blame]62<div class="doc_section"><a name="policies">Developer Policies</a></div>
Reid Spencer8d0ac692007-02-13 09:06:01 +0000[diff] [blame]63<!--=========================================================================-->
64<div class="doc_text">
Chris Lattner0cca50c2007-02-19 06:57:16 +0000[diff] [blame]65 <p>This section contains policies that pertain to frequent LLVM
Reid Spencercb069bb2007-02-19 17:38:38 +0000[diff] [blame]66 developers. We always welcome <a href="#patches">one-off patches</a> from
67 people who do not routinely contribute to LLVM, but we expect more from
68 frequent contributors to keep the system as efficient as possible for
69 everyone.
70 Frequent LLVM contributors are expected to meet the following requirements in
Chris Lattner450f40a2007-02-19 03:50:31 +0000[diff] [blame]71 order for LLVM to maintain a high standard of quality.<p>
Reid Spencer8d0ac692007-02-13 09:06:01 +0000[diff] [blame]72</div>
73
74<!-- _______________________________________________________________________ -->
75<div class="doc_subsection"> <a name="informed">Stay Informed</a> </div>
76<div class="doc_text">
77 <p>Developers should stay informed by reading at least the
Reid Spencer01f56be2007-02-14 16:21:10 +0000[diff] [blame]78 <a href="http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev">llvmdev</a>
79 email list. If you are doing anything more than just casual work on LLVM,
80 it is suggested that you also subscribe to the
81 <a href="http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits">llvm-commits</a>
82 list and pay attention to changes being made by others.</p>
Reid Spencer78bade22007-02-14 07:22:19 +0000[diff] [blame]83 <p>We recommend that active developers register an email account with
84 <a href="http://llvm.org/bugs/">LLVM Bugzilla</a> and preferably subscribe to
85 the <a href="http://lists.cs.uiuc.edu/mailman/listinfo/llvmbugs">llvm-bugs</a>
86 email list to keep track of bugs and enhancements occurring in LLVM.</p>
Reid Spencer8d0ac692007-02-13 09:06:01 +0000[diff] [blame]87</div>
88
89<!-- _______________________________________________________________________ -->
Chris Lattner1653fec2007-02-19 05:57:29 +0000[diff] [blame]90<div class="doc_subsection"> <a name="patches">Making a Patch</a></div>
91
92<div class="doc_text">
93
Misha Brukman66286e92008-12-11 19:44:51 +0000[diff] [blame]94 <p>When making a patch for review, the goal is to make it as easy for the
95 reviewer to read it as possible. As such, we recommend that you:</p>
Chris Lattner1653fec2007-02-19 05:57:29 +0000[diff] [blame]96 <ol>
Reid Spencer669ed452007-07-09 08:04:31 +0000[diff] [blame]97 <li>Make your patch against the Subversion trunk, not a branch, and not an
98 old version of LLVM. This makes it easy to apply the patch.</li>
Chris Lattner1653fec2007-02-19 05:57:29 +0000[diff] [blame]99
100 <li>Similarly, patches should be submitted soon after they are generated.
101 Old patches may not apply correctly if the underlying code changes between
102 the time the patch was created and the time it is applied.</li>
103
104 <li>Patches should be made with this command:
Misha Brukman66286e92008-12-11 19:44:51 +0000[diff] [blame]105 <div class="doc_code"><pre>svn diff -x -u</pre></div>
106 or with the utility <tt>utils/mkpatch</tt>, which makes it easy to read the
107 diff.</li>
Chris Lattner1653fec2007-02-19 05:57:29 +0000[diff] [blame]108
109 <li>Patches should not include differences in generated code such as the
Chris Lattner01bb8b02009-01-02 07:10:51 +0000[diff] [blame^]110 code generated by <tt>autoconf</tt> or <tt>tblgen</tt>. The
Chris Lattner1653fec2007-02-19 05:57:29 +0000[diff] [blame]111 <tt>utils/mkpatch</tt> utility takes care of this for you.</li>
Chris Lattner1653fec2007-02-19 05:57:29 +0000[diff] [blame]112 </ol>
Chris Lattner0cca50c2007-02-19 06:57:16 +0000[diff] [blame]113
114 <p>When sending a patch to a mailing list, it is a good idea to send it as an
115 <em>attachment</em> to the message, not embedded into the text of the
116 message. This ensures that your mailer will not mangle the patch when it
117 sends it (e.g. by making whitespace changes or by wrapping lines).</p>
Gordon Henriksen72efde22008-06-26 22:58:37 +0000[diff] [blame]118
119 <p><em>For Thunderbird users:</em> Before submitting a patch, please open
120 <em>Preferences &#8594; Advanced &#8594; General &#8594; Config Editor</em>,
121 find the key <tt>mail.content_disposition_type</tt>, and set its value to
122 <tt>1</tt>. Without this setting, Thunderbird sends your attachment using
123 <tt>Content-Disposition: inline</tt> rather than <tt>Content-Disposition:
124 attachment</tt>. Apple Mail gamely displays such a file inline, making it
125 difficult to work with for reviewers using that program.</p>
Chris Lattner1653fec2007-02-19 05:57:29 +0000[diff] [blame]126</div>
127
128<!-- _______________________________________________________________________ -->
Reid Spencer8d0ac692007-02-13 09:06:01 +0000[diff] [blame]129<div class="doc_subsection"> <a name="reviews">Code Reviews</a></div>
130<div class="doc_text">
Chris Lattner450f40a2007-02-19 03:50:31 +0000[diff] [blame]131 <p>LLVM has a code review policy. Code review is one way to increase the
132 quality of software. We generally follow these policies:</p>
Reid Spencer8d0ac692007-02-13 09:06:01 +0000[diff] [blame]133 <ol>
Reid Spencer01f56be2007-02-14 16:21:10 +0000[diff] [blame]134 <li>All developers are required to have significant changes reviewed
Reid Spencer8d0ac692007-02-13 09:06:01 +0000[diff] [blame]135 before they are committed to the repository.</li>
Chris Lattner450f40a2007-02-19 03:50:31 +0000[diff] [blame]136 <li>Code reviews are conducted by email, usually on the llvm-commits
137 list.</li>
138 <li>Code can be reviewed either before it is committed or after. We expect
139 major changes to be reviewed before being committed, but smaller
140 changes (or changes where the developer owns the component) can be
141 reviewed after commit.</li>
Reid Spencer8d0ac692007-02-13 09:06:01 +0000[diff] [blame]142 <li>The developer responsible for a code change is also responsible for
Chris Lattner1653fec2007-02-19 05:57:29 +0000[diff] [blame]143 making all necessary review-related changes.</li>
Reid Spencercb069bb2007-02-19 17:38:38 +0000[diff] [blame]144 <li>Code review can be an iterative process, which continues until the patch
Chris Lattner1653fec2007-02-19 05:57:29 +0000[diff] [blame]145 is ready to be committed.</li>
Reid Spencer8d0ac692007-02-13 09:06:01 +0000[diff] [blame]146 </ol>
Chris Lattner0cca50c2007-02-19 06:57:16 +0000[diff] [blame]147
148 <p>Developers should participate in code reviews as both reviewers and
Reid Spencercb069bb2007-02-19 17:38:38 +0000[diff] [blame]149 reviewees. If someone is kind enough to review your code, you should
Chris Lattner0cca50c2007-02-19 06:57:16 +0000[diff] [blame]150 return the favor for someone else. Note that anyone is welcome to review
Reid Spencer669ed452007-07-09 08:04:31 +0000[diff] [blame]151 and give feedback on a patch, but only people with Subversion write access
152 can approve it.</p>
Chris Lattner0cca50c2007-02-19 06:57:16 +0000[diff] [blame]153
Reid Spencer8d0ac692007-02-13 09:06:01 +0000[diff] [blame]154</div>
Reid Spencer8d0ac692007-02-13 09:06:01 +0000[diff] [blame]155
156<!-- _______________________________________________________________________ -->
Chris Lattnere268a402007-12-03 19:00:47 +0000[diff] [blame]157<div class="doc_subsection"> <a name="owners">Code Owners</a></div>
158<div class="doc_text">
159
160 <p>The LLVM Project relies on two features of its process to maintain rapid
161 development in addition to the high quality of its source base: the
162 combination of code review plus post-commit review for trusted maintainers.
163 Having both is a great way for the project to take advantage of the fact
164 that most people do the right thing most of the time, and only commit
165 patches without pre-commit review when they are confident they are
166 right.</p>
167
168 <p>The trick to this is that the project has to guarantee that all patches
169 that are committed are reviewed after they go in: you don't want everyone
170 to assume someone else will review it, allowing the patch to go unreviewed.
171 To solve this problem, we have a notion of an 'owner' for a piece of the
172 code. The sole responsibility of a code owner is to ensure that a commit
173 to their area of the code is appropriately reviewed, either by themself or
174 by someone else. The current code owners are:</p>
175
176 <ol>
177 <li><b>Anton Korobeynikov</b>: Exception handling, debug information, and
178 Windows codegen.</li>
179 <li><b>Duncan Sands</b>: llvm-gcc 4.2.</li>
180 <li><b>Evan Cheng</b>: Code generator and all targets.</li>
181 <li><b>Chris Lattner</b>: Everything else.</li>
182 </ol>
183
184 <p>Note that code ownership is completely different than reviewers: anyone can
185 review a piece of code, and we welcome code review from anyone who is
186 interested. Code owners are the "last line of defense" to guarantee that
187 all patches that are committed are actually reviewed.</p>
188
189 <p>Being a code owner is a somewhat unglamorous position, but it is incredibly
190 important for the ongoing success of the project. Because people get busy,
191 interests change, and unexpected things happen, code ownership is purely
192 opt-in, and anyone can choose to resign their "title" at any time. For now,
193 we do not have an official policy on how one gets elected to be a code
194 owner.
195 </p>
196
197</div>
198
199
200<!-- _______________________________________________________________________ -->
Reid Spencer8d0ac692007-02-13 09:06:01 +0000[diff] [blame]201<div class="doc_subsection"> <a name="testcases">Test Cases</a></div>
202<div class="doc_text">
Chris Lattner450f40a2007-02-19 03:50:31 +0000[diff] [blame]203 <p>Developers are required to create test cases for any bugs fixed and any new
Chris Lattner0cca50c2007-02-19 06:57:16 +0000[diff] [blame]204 features added. Some tips for getting your testcase approved:</p>
Reid Spencer8d0ac692007-02-13 09:06:01 +0000[diff] [blame]205 <ol>
Chris Lattner0cca50c2007-02-19 06:57:16 +0000[diff] [blame]206 <li>All feature and regression test cases are added to the
Reid Spencer80483bb2007-02-13 09:20:14 +0000[diff] [blame]207 <tt>llvm/test</tt> directory. The appropriate sub-directory should be
Reid Spencer8d0ac692007-02-13 09:06:01 +0000[diff] [blame]208 selected (see the <a href="TestingGuide.html">Testing Guide</a> for
Reid Spencer40602962007-02-13 20:57:57 +0000[diff] [blame]209 details).</li>
Reid Spencer01f56be2007-02-14 16:21:10 +0000[diff] [blame]210 <li>Test cases should be written in
211 <a href="LangRef.html">LLVM assembly language</a> unless the
Reid Spencer3eedbd32007-02-14 07:57:48 +0000[diff] [blame]212 feature or regression being tested requires another language (e.g. the
Chris Lattner450f40a2007-02-19 03:50:31 +0000[diff] [blame]213 bug being fixed or feature being implemented is in the llvm-gcc C++
Chris Lattner0cca50c2007-02-19 06:57:16 +0000[diff] [blame]214 front-end, in which case it must be written in C++).</li>
Reid Spencer01f56be2007-02-14 16:21:10 +0000[diff] [blame]215 <li>Test cases, especially for regressions, should be reduced as much as
Duncan Sands8036ca42007-03-30 12:22:09 +0000[diff] [blame]216 possible, by <a href="Bugpoint.html">bugpoint</a> or
Reid Spencer3eedbd32007-02-14 07:57:48 +0000[diff] [blame]217 manually. It is unacceptable
Reid Spencer8d0ac692007-02-13 09:06:01 +0000[diff] [blame]218 to place an entire failing program into <tt>llvm/test</tt> as this creates
Reid Spencer01f56be2007-02-14 16:21:10 +0000[diff] [blame]219 a <i>time-to-test</i> burden on all developers. Please keep them short.</li>
Reid Spencer8d0ac692007-02-13 09:06:01 +0000[diff] [blame]220 </ol>
Chris Lattner0cca50c2007-02-19 06:57:16 +0000[diff] [blame]221
222 <p>Note that llvm/test is designed for regression and small feature tests
223 only. More extensive test cases (e.g., entire applications, benchmarks,
224 etc) should be added to the <tt>llvm-test</tt> test suite. The llvm-test
225 suite is for coverage (correctness, performance, etc) testing, not feature
Chris Lattner64113a52007-02-19 06:57:46 +0000[diff] [blame]226 or regression testing.</p>
Reid Spencer8d0ac692007-02-13 09:06:01 +0000[diff] [blame]227</div>
228
Chris Lattner450f40a2007-02-19 03:50:31 +0000[diff] [blame]229<!-- _______________________________________________________________________ -->
Chris Lattner39582bd2007-02-19 05:47:13 +0000[diff] [blame]230<div class="doc_subsection"> <a name="quality">Quality</a></div>
231<div class="doc_text">
232 <p>The minimum quality standards that any change must satisfy before being
233 committed to the main development branch are:</p>
234 <ol>
235 <li>Code must adhere to the
236 <a href="CodingStandards.html">LLVM Coding Standards</a>.</li>
237 <li>Code must compile cleanly (no errors, no warnings) on at least one
238 platform.</li>
239 <li>Bug fixes and new features should <a href="#testcases">include a
240 testcase</a> so we know if the fix/feature ever regresses in the
241 future.</li>
Chris Lattner0cca50c2007-02-19 06:57:16 +0000[diff] [blame]242 <li>Code must pass the dejagnu (<tt>llvm/test</tt>) test suite.</li>
Chris Lattner39582bd2007-02-19 05:47:13 +0000[diff] [blame]243 <li>The code must not cause regressions on a reasonable subset of llvm-test,
244 where "reasonable" depends on the contributor's judgement and the scope
245 of the change (more invasive changes require more testing). A reasonable
Dan Gohman72587372008-09-09 22:13:09 +0000[diff] [blame]246 subset might be something like
247 "<tt>llvm-test/MultiSource/Benchmarks</tt>".</li>
Chris Lattner39582bd2007-02-19 05:47:13 +0000[diff] [blame]248 </ol>
249 <p>Additionally, the committer is responsible for addressing any problems
250 found in the future that the change is responsible for. For example:</p>
251 <ul>
Chris Lattner0cca50c2007-02-19 06:57:16 +0000[diff] [blame]252 <li>The code should compile cleanly on all supported platforms.</li>
253 <li>The changes should not cause any correctness regressions in the
254 <tt>llvm-test</tt> suite and must not cause any major performance
255 regressions.</li>
Chris Lattner39582bd2007-02-19 05:47:13 +0000[diff] [blame]256 <li>The change set should not cause performance or correctness regressions
257 for the LLVM tools.</li>
258 <li>The changes should not cause performance or correctness regressions in
259 code compiled by LLVM on all applicable targets.</li>
260 <li>You are expected to address any <a href="http://llvm.org/bugs/">bugzilla
261 bugs</a> that result from your change.</li>
262 </ul>
263
Chris Lattner0cca50c2007-02-19 06:57:16 +0000[diff] [blame]264 <p>We prefer for this to be handled before submission but understand that it
265 isn't possible to test all of this for every submission. Our nightly
266 testing
Chris Lattner39582bd2007-02-19 05:47:13 +0000[diff] [blame]267 infrastructure normally finds these problems. A good rule of thumb is to
268 check the nightly testers for regressions the day after your change.</p>
269
270 <p>Commits that violate these quality standards (e.g. are very broken) may
271 be reverted. This is necessary when the change blocks other developers from
272 making progress. The developer is welcome to re-commit the change after
273 the problem has been fixed.</p>
274</div>
275
276<!-- _______________________________________________________________________ -->
Chris Lattner1acdc952007-02-19 05:49:11 +0000[diff] [blame]277<div class="doc_subsection">
278 <a name="commitaccess">Obtaining Commit Access</a></div>
Chris Lattner450f40a2007-02-19 03:50:31 +0000[diff] [blame]279<div class="doc_text">
280
281<p>
282We grant commit access to contributors with a track record of submitting high
Chris Lattnerbebcdabd82007-12-03 00:36:20 +0000[diff] [blame]283quality patches. If you would like commit access, please send an email to
284<a href="mailto:sabre@nondot.org">Chris</a> with the following information:</p>
285
286<ol>
287 <li>The user name you want to commit with, e.g. "sabre".</li>
288 <li>The full name and email address you want message to llvm-commits to come
289 from, e.g. "Chris Lattner &lt;sabre@nondot.org&gt;".</li>
290 <li>A "password hash" of the password you want to use, e.g. "2ACR96qjUqsyM".
291 Note that you don't ever tell us what your password is, you just give it
292 to us in an encrypted form. To get this, run "htpasswd" (a utility that
293 comes with apache) in crypt mode (often enabled with "-d"), or find a web
294 page that will do it for you.</li>
295</ol>
296
297<p>Once you've been granted commit access, you should be able to check out an
298 LLVM tree with an SVN URL of "https://username@llvm.org/..." instead of the
299 normal anonymous URL of "http://llvm.org/...". The first time you commit
300 you'll have to type in your password. Note that you may get a warning from
301 SVN about an untrusted key, you can ignore this. To verify that your commit
302 access works, please do a test commit (e.g. change a comment or add a blank
303 line). Your first commit to a repository may require the autogenerated email
304 to be approved by a mailing list. This is normal, and will be done when
305 the mailing list owner has time.</p>
Chris Lattner450f40a2007-02-19 03:50:31 +0000[diff] [blame]306
307<p>If you have recently been granted commit access, these policies apply:</p>
Chris Lattnerbebcdabd82007-12-03 00:36:20 +0000[diff] [blame]308
Chris Lattner450f40a2007-02-19 03:50:31 +0000[diff] [blame]309<ol>
310 <li>You are granted <i>commit-after-approval</i> to all parts of LLVM.
311 To get approval, submit a <a href="#patches">patch</a> to
312 <a href="http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits">
313 llvm-commits</a>. When approved you may commit it yourself.</li>
314 <li>You are allowed to commit patches without approval which you think are
Chris Lattner0cca50c2007-02-19 06:57:16 +0000[diff] [blame]315 obvious. This is clearly a subjective decision &mdash; we simply expect you
316 to use good judgement. Examples include: fixing build breakage, reverting
Chris Lattner450f40a2007-02-19 03:50:31 +0000[diff] [blame]317 obviously broken patches, documentation/comment changes, any other minor
318 changes.</li>
319 <li>You are allowed to commit patches without approval to those portions
Chris Lattner0cca50c2007-02-19 06:57:16 +0000[diff] [blame]320 of LLVM that you have contributed or maintain (i.e., have been assigned
Chris Lattner450f40a2007-02-19 03:50:31 +0000[diff] [blame]321 responsibility for), with the proviso that such commits must not break the
322 build. This is a "trust but verify" policy and commits of this nature are
323 reviewed after they are committed.</li>
324 <li>Multiple violations of these policies or a single egregious violation
325 may cause commit access to be revoked.</li>
326</ol>
Chris Lattner0cca50c2007-02-19 06:57:16 +0000[diff] [blame]327
328<p>In any case, your changes are still subject to <a href="#reviews">code
329review</a> (either before or after they are committed, depending on the nature
330of the change). You are encouraged to review other peoples' patches as well,
Chris Lattnerd2dceea2007-03-02 02:57:34 +0000[diff] [blame]331but you aren't required to.</p>
Chris Lattner450f40a2007-02-19 03:50:31 +0000[diff] [blame]332
333</div>
334
335<!-- _______________________________________________________________________ -->
Chris Lattnerc7d954e2007-02-19 05:43:04 +0000[diff] [blame]336<div class="doc_subsection"> <a name="newwork">Making a Major Change</a></div>
337<div class="doc_text">
338 <p>When a developer begins a major new project with the aim of contributing
339 it back to LLVM, s/he should inform the community with an email to
Chris Lattner0cca50c2007-02-19 06:57:16 +0000[diff] [blame]340 the <a href="http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev">llvmdev</a>
Chris Lattnerc7d954e2007-02-19 05:43:04 +0000[diff] [blame]341 email list, to the extent possible. The reason for this is to:
342 <ol>
343 <li>keep the community informed about future changes to LLVM, </li>
Chris Lattner0cca50c2007-02-19 06:57:16 +0000[diff] [blame]344 <li>avoid duplication of effort by preventing multiple parties working on
345 the same thing and not knowing about it, and</li>
Chris Lattnerc7d954e2007-02-19 05:43:04 +0000[diff] [blame]346 <li>ensure that any technical issues around the proposed work are
347 discussed and resolved before any significant work is done.</li>
348 </ol>
349
350 <p>The design of LLVM is carefully controlled to ensure that all the pieces
351 fit together well and are as consistent as possible. If you plan to make a
Chris Lattner0cca50c2007-02-19 06:57:16 +0000[diff] [blame]352 major change to the way LLVM works or want to add a major new extension, it
353 is a good idea to get consensus with the development
Chris Lattnerc7d954e2007-02-19 05:43:04 +0000[diff] [blame]354 community before you start working on it.</p>
355
Chris Lattner1ff20cd2007-02-19 06:05:58 +0000[diff] [blame]356 <p>Once the design of the new feature is finalized, the work itself should be
357 done as a series of <a href="#incremental">incremental changes</a>, not as
358 a long-term development branch.</p>
359
Chris Lattnerc7d954e2007-02-19 05:43:04 +0000[diff] [blame]360</div>
361
362<!-- _______________________________________________________________________ -->
Chris Lattner1ff20cd2007-02-19 06:05:58 +0000[diff] [blame]363<div class="doc_subsection"> <a name="incremental">Incremental Development</a>
Chris Lattner450f40a2007-02-19 03:50:31 +0000[diff] [blame]364</div>
365<div class="doc_text">
Chris Lattner1ff20cd2007-02-19 06:05:58 +0000[diff] [blame]366 <p>In the LLVM project, we do all significant changes as a series of
367 incremental patches. We have a strong dislike for huge changes or
368 long-term development branches. Long-term development branches have a
369 number of drawbacks:</p>
Chris Lattner450f40a2007-02-19 03:50:31 +0000[diff] [blame]370
371 <ol>
372 <li>Branches must have mainline merged into them periodically. If the branch
373 development and mainline development occur in the same pieces of code,
374 resolving merge conflicts can take a lot of time.</li>
375 <li>Other people in the community tend to ignore work on branches.</li>
376 <li>Huge changes (produced when a branch is merged back onto mainline) are
377 extremely difficult to <a href="#reviews">code review</a>.</li>
378 <li>Branches are not routinely tested by our nightly tester
379 infrastructure.</li>
380 <li>Changes developed as monolithic large changes often don't work until the
381 entire set of changes is done. Breaking it down into a set of smaller
382 changes increases the odds that any of the work will be committed to the
383 main repository.</li>
384 </ol>
385
386 <p>
387 To address these problems, LLVM uses an incremental development style and we
388 require contributors to follow this practice when making a large/invasive
389 change. Some tips:</p>
390
391 <ul>
392 <li>Large/invasive changes usually have a number of secondary changes that
393 are required before the big change can be made (e.g. API cleanup, etc).
394 These sorts of changes can often be done before the major change is done,
395 independently of that work.</li>
396 <li>The remaining inter-related work should be decomposed into unrelated
397 sets of changes if possible. Once this is done, define the first increment
398 and get consensus on what the end goal of the change is.</li>
Chris Lattner8bb16ff2007-02-19 05:59:30 +0000[diff] [blame]399
400 <li>Each change in the set can be stand alone (e.g. to fix a bug), or part
401 of a planned series of changes that works towards the development goal.</li>
402
403 <li>Each change should be kept as small as possible. This simplifies your
Chris Lattner450f40a2007-02-19 03:50:31 +0000[diff] [blame]404 work (into a logical progression), simplifies code review and reduces the
405 chance that you will get negative feedback on the change. Small increments
406 also facilitate the maintenance of a high quality code base.</li>
Chris Lattner8bb16ff2007-02-19 05:59:30 +0000[diff] [blame]407
Chris Lattner450f40a2007-02-19 03:50:31 +0000[diff] [blame]408 <li>Often, an independent precursor to a big change is to add a new API and
409 slowly migrate clients to use the new API. Each change to use the new
410 API is often "obvious" and can be committed without review. Once the
Chris Lattner0cca50c2007-02-19 06:57:16 +0000[diff] [blame]411 new API is in place and used, it is much easier to replace the
412 underlying implementation of the API. This implementation change is
413 logically separate from the API change.</li>
Chris Lattner450f40a2007-02-19 03:50:31 +0000[diff] [blame]414 </ul>
415
416 <p>If you are interested in making a large change, and this scares you, please
417 make sure to first <a href="#newwork">discuss the change/gather
Chris Lattner0cca50c2007-02-19 06:57:16 +0000[diff] [blame]418 consensus</a> then ask about the best way to go about making
Chris Lattner450f40a2007-02-19 03:50:31 +0000[diff] [blame]419 the change.</p>
420</div>
421
422<!-- _______________________________________________________________________ -->
Chris Lattnerc7d954e2007-02-19 05:43:04 +0000[diff] [blame]423<div class="doc_subsection"><a name="attribution">Attribution of
424Changes</a></div>
Chris Lattner450f40a2007-02-19 03:50:31 +0000[diff] [blame]425<div class="doc_text">
426 <p>We believe in correct attribution of contributions to
427 their contributors. However, we do not want the source code to be littered
Chris Lattnerab2b10c2007-12-29 19:56:08 +0000[diff] [blame]428 with random attributions "this code written by J Random Guy" (this is noisy
Chris Lattner37a1ebb2008-11-16 17:40:16 +0000[diff] [blame]429 and distracting). In practice, the revision control system keeps a perfect
Daniel Dunbar607d6bb2008-11-19 02:37:39 +0000[diff] [blame]430 history of who changed what, and the CREDITS.txt file describes higher-level
Chris Lattner37a1ebb2008-11-16 17:40:16 +0000[diff] [blame]431 contributions. If you commit a patch for someone else, please say "patch
432 contributed by J Random Guy!" in the commit message.</p>
Chris Lattnerab2b10c2007-12-29 19:56:08 +0000[diff] [blame]433
Chris Lattner37a1ebb2008-11-16 17:40:16 +0000[diff] [blame]434 <p>Overall, please do not add contributor names to the source code.</p>
Chris Lattner450f40a2007-02-19 03:50:31 +0000[diff] [blame]435</div>
436
437
Reid Spencer8d0ac692007-02-13 09:06:01 +0000[diff] [blame]438
Reid Spencer8d0ac692007-02-13 09:06:01 +0000[diff] [blame]439<!--=========================================================================-->
Chris Lattner793aa382007-02-19 06:19:16 +0000[diff] [blame]440<div class="doc_section">
441 <a name="clp">Copyright, License, and Patents</a>
442</div>
Reid Spencer78bade22007-02-14 07:22:19 +0000[diff] [blame]443<!--=========================================================================-->
Chris Lattner450f40a2007-02-19 03:50:31 +0000[diff] [blame]444
Reid Spencer78bade22007-02-14 07:22:19 +0000[diff] [blame]445<div class="doc_text">
Reid Spencercb069bb2007-02-19 17:38:38 +0000[diff] [blame]446 <p>This section addresses the issues of copyright, license and patents for
447 the LLVM project.
Reid Spencer78bade22007-02-14 07:22:19 +0000[diff] [blame]448 Currently, the University of Illinois is the LLVM copyright holder and the
449 terms of its license to LLVM users and developers is the
450 <a href="http://www.opensource.org/licenses/UoI-NCSA.php">University of
Chris Lattner0cca50c2007-02-19 06:57:16 +0000[diff] [blame]451 Illinois/NCSA Open Source License</a>.</p>
Reid Spencer78bade22007-02-14 07:22:19 +0000[diff] [blame]452
453<div class="doc_notes">
Misha Brukman740e3db2008-12-29 19:55:01 +0000[diff] [blame]454 <p style="text-align:center;font-weight:bold">NOTE: This section
455 deals with legal matters but does not provide legal advice. We are not
456 lawyers, please seek legal counsel from an attorney.</p>
Reid Spencer78bade22007-02-14 07:22:19 +0000[diff] [blame]457</div>
Reid Spencer78bade22007-02-14 07:22:19 +0000[diff] [blame]458</div>
459
460<!-- _______________________________________________________________________ -->
461<div class="doc_subsection"><a name="copyright">Copyright</a></div>
462<div class="doc_text">
463 <p>
Reid Spencer01f56be2007-02-14 16:21:10 +0000[diff] [blame]464 <p>For consistency and ease of management, the project requires the
Chris Lattner450f40a2007-02-19 03:50:31 +0000[diff] [blame]465 copyright for all LLVM software to be held by a single copyright holder:
466 the University of Illinois (UIUC).</p>
467
468 <p>
469 Although UIUC may eventually reassign the copyright of the software to another
Chris Lattner305915b2008-05-20 20:06:53 +0000[diff] [blame]470 entity (e.g. a dedicated non-profit "LLVM Organization")
Chris Lattner450f40a2007-02-19 03:50:31 +0000[diff] [blame]471 the intent for the project is to always have a single entity hold the
472 copyrights to LLVM at any given time.</p>
473
474 <p>We believe that having a single copyright
Reid Spencer78bade22007-02-14 07:22:19 +0000[diff] [blame]475 holder is in the best interests of all developers and users as it greatly
476 reduces the managerial burden for any kind of administrative or technical
Chris Lattner450f40a2007-02-19 03:50:31 +0000[diff] [blame]477 decisions about LLVM. The goal of the LLVM project is to always keep the code
478 open and <a href="#license">licensed under a very liberal license</a>.</p>
Chris Lattner305915b2008-05-20 20:06:53 +0000[diff] [blame]479
Reid Spencer78bade22007-02-14 07:22:19 +0000[diff] [blame]480</div>
Chris Lattner450f40a2007-02-19 03:50:31 +0000[diff] [blame]481
Reid Spencer78bade22007-02-14 07:22:19 +0000[diff] [blame]482<!-- _______________________________________________________________________ -->
483<div class="doc_subsection"><a name="license">License</a></div>
484<div class="doc_text">
Chris Lattner450f40a2007-02-19 03:50:31 +0000[diff] [blame]485 <p>We intend to keep LLVM perpetually open source
486 and to use a liberal open source license. The current license is the
Reid Spencer01f56be2007-02-14 16:21:10 +0000[diff] [blame]487 <a href="http://www.opensource.org/licenses/UoI-NCSA.php">
488 University of Illinois/NCSA Open Source License</a>, which boils
Reid Spencer78bade22007-02-14 07:22:19 +0000[diff] [blame]489 down to this:</p>
490 <ul>
491 <li>You can freely distribute LLVM.</li>
492 <li>You must retain the copyright notice if you redistribute LLVM.</li>
Chris Lattner305915b2008-05-20 20:06:53 +0000[diff] [blame]493 <li>Binaries derived from LLVM must reproduce the copyright notice (e.g.
494 in an included readme file).</li>
Reid Spencer78bade22007-02-14 07:22:19 +0000[diff] [blame]495 <li>You can't use our names to promote your LLVM derived products.</li>
496 <li>There's no warranty on LLVM at all.</li>
497 </ul>
Chris Lattner450f40a2007-02-19 03:50:31 +0000[diff] [blame]498
499 <p>We believe this fosters the widest adoption of LLVM because it <b>allows
500 commercial products to be derived from LLVM</b> with few restrictions and
Reid Spencer5b9a6092007-02-14 22:55:40 +0000[diff] [blame]501 without a requirement for making any derived works also open source (i.e.
Chris Lattner450f40a2007-02-19 03:50:31 +0000[diff] [blame]502 LLVM's license is not a "copyleft" license like the GPL). We suggest that you
503 read the <a href="http://www.opensource.org/licenses/UoI-NCSA.php">License</a>
Reid Spencer853418d2007-02-14 22:58:39 +0000[diff] [blame]504 if further clarification is needed.</p>
Chris Lattner450f40a2007-02-19 03:50:31 +0000[diff] [blame]505
Chris Lattner7d8012b2007-02-22 06:33:23 +0000[diff] [blame]506 <p>Note that the LLVM Project does distribute llvm-gcc, <b>which is GPL.</b>
Bill Wendlingd932aab2007-02-19 18:32:40 +0000[diff] [blame]507 This means that anything "linked" into llvm-gcc must itself be compatible
Chris Lattner450f40a2007-02-19 03:50:31 +0000[diff] [blame]508 with the GPL, and must be releasable under the terms of the GPL. This implies
Chris Lattner7d8012b2007-02-22 06:33:23 +0000[diff] [blame]509 that <b>any code linked into llvm-gcc and distributed to others may be subject
510 to the viral aspects of the GPL</b> (for example, a proprietary code generator
Chris Lattnerd5e2d402007-02-23 06:53:06 +0000[diff] [blame]511 linked into llvm-gcc must be made available under the GPL). This is not a
512 problem for code already distributed under a more liberal license (like the
513 UIUC license), and does not affect code generated by llvm-gcc. It may be a
514 problem if you intend to base commercial development on llvm-gcc without
Chris Lattner0cca50c2007-02-19 06:57:16 +0000[diff] [blame]515 redistributing your source code.</p>
Chris Lattner450f40a2007-02-19 03:50:31 +0000[diff] [blame]516
517 <p>We have no plans to change the license of LLVM. If you have questions
518 or comments about the license, please contact the <a
519 href="mailto:llvm-oversight@cs.uiuc.edu">LLVM Oversight Group</a>.</p>
520
Reid Spencer78bade22007-02-14 07:22:19 +0000[diff] [blame]521</div>
Chris Lattner1ff20cd2007-02-19 06:05:58 +0000[diff] [blame]522
523<!-- _______________________________________________________________________ -->
524<div class="doc_subsection"><a name="patents">Patents</a></div>
525<div class="doc_text">
526
Chris Lattnerb87fb2a2007-02-19 06:13:50 +0000[diff] [blame]527<p>To the best of our knowledge, LLVM does not infringe on any patents (we have
528 actually removed code from LLVM in the past that was found to infringe).
Chris Lattnerd0742ce2007-02-19 06:15:33 +0000[diff] [blame]529 Having code in LLVM that infringes on patents would violate an important
530 goal of the project by making it hard or impossible to reuse the code for
531 arbitrary purposes (including commercial use).</p>
Chris Lattnerb87fb2a2007-02-19 06:13:50 +0000[diff] [blame]532
533<p>When contributing code, we expect contributors to notify us of any potential
Chris Lattner64a4c112008-05-22 03:06:14 +0000[diff] [blame]534 for patent-related trouble with their changes. If you or your employer
535 own the rights to a
Chris Lattnerb87fb2a2007-02-19 06:13:50 +0000[diff] [blame]536 patent and would like to contribute code to LLVM that relies on it, we
Chris Lattner64a4c112008-05-22 03:06:14 +0000[diff] [blame]537 require that
538 the copyright owner sign an agreement that allows any other user of LLVM to
Chris Lattnerb87fb2a2007-02-19 06:13:50 +0000[diff] [blame]539 freely use your patent. Please contact the <a
540 href="mailto:llvm-oversight@cs.uiuc.edu">oversight group</a> for more
541 details.</p>
Chris Lattner1ff20cd2007-02-19 06:05:58 +0000[diff] [blame]542</div>
543
544
Reid Spencer78bade22007-02-14 07:22:19 +0000[diff] [blame]545<!-- _______________________________________________________________________ -->
546<div class="doc_subsection"><a name="devagree">Developer Agreements</a></div>
547<div class="doc_text">
Chris Lattner450f40a2007-02-19 03:50:31 +0000[diff] [blame]548 <p>With regards to the LLVM copyright and licensing, developers agree to
549 assign their copyrights to UIUC for any contribution made so that
550 the entire software base can be managed by a single copyright holder. This
551 implies that any contributions can be licensed under the license that the
Chris Lattner1acdc952007-02-19 05:49:11 +0000[diff] [blame]552 project uses.</p>
Chris Lattner64a4c112008-05-22 03:06:14 +0000[diff] [blame]553
554 <p>When contributing code, you also affirm that you are legally entitled to
555 grant this copyright, personally or on behalf of your employer. If the code
556 belongs to some other entity, please raise this issue with the oversight
557 group before the code is committed.</p>
Reid Spencer78bade22007-02-14 07:22:19 +0000[diff] [blame]558</div>
559
Reid Spencer8d0ac692007-02-13 09:06:01 +0000[diff] [blame]560<!-- *********************************************************************** -->
561<hr>
562<address>
563 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
Misha Brukman44408702008-12-11 17:34:48 +0000[diff] [blame]564 src="http://jigsaw.w3.org/css-validator/images/vcss-blue" alt="Valid CSS"></a>
Reid Spencer8d0ac692007-02-13 09:06:01 +0000[diff] [blame]565 <a href="http://validator.w3.org/check/referer"><img
Misha Brukmanf00ddb02008-12-11 18:23:24 +0000[diff] [blame]566 src="http://www.w3.org/Icons/valid-html401-blue" alt="Valid HTML 4.01"></a>
Chris Lattnerd0742ce2007-02-19 06:15:33 +0000[diff] [blame]567 Written by the
Reid Spencer3eedbd32007-02-14 07:57:48 +0000[diff] [blame]568 <a href="mailto:llvm-oversight@cs.uiuc.edu">LLVM Oversight Group</a><br>
Reid Spencer8d0ac692007-02-13 09:06:01 +0000[diff] [blame]569 <a href="http://llvm.org">The LLVM Compiler Infrastructure</a><br>
570 Last modified: $Date$
571</address>
572</body>
573</html>