blob: 6119e7cea33ac4404236cec2a4606df6291db4d1 [file] [log] [blame]
Misha Brukman500bc302003-11-22 00:38:41 +00001<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"
2 "http://www.w3.org/TR/html4/strict.dtd">
3<html>
4<head>
Reid Spencer6454ed32004-11-18 18:38:58 +00005 <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
Misha Brukman500bc302003-11-22 00:38:41 +00006 <link rel="stylesheet" href="llvm.css" type="text/css">
Chad Rosiere6291d02011-05-27 22:50:46 +00007 <title>LLVM 3.0 Release Notes</title>
Misha Brukman500bc302003-11-22 00:38:41 +00008</head>
9<body>
Chris Lattner79c3fe12003-10-02 04:57:28 +000010
Chad Rosiere6291d02011-05-27 22:50:46 +000011<h1>LLVM 3.0 Release Notes</h1>
Mikhail Glushenkovea65d7d2008-10-13 02:08:34 +000012
Chris Lattner0e464a92010-03-17 04:02:39 +000013<img align=right src="http://llvm.org/img/DragonSmall.png"
Gabor Greifee2187a2010-04-22 10:21:43 +000014 width="136" height="136" alt="LLVM Dragon Logo">
Chris Lattner0e464a92010-03-17 04:02:39 +000015
Chris Lattner79c3fe12003-10-02 04:57:28 +000016<ol>
Misha Brukman500bc302003-11-22 00:38:41 +000017 <li><a href="#intro">Introduction</a></li>
Chris Lattner96a445e2008-10-13 18:01:01 +000018 <li><a href="#subproj">Sub-project Status Update</a></li>
Chad Rosiere6291d02011-05-27 22:50:46 +000019 <li><a href="#externalproj">External Projects Using LLVM 3.0</a></li>
20 <li><a href="#whatsnew">What's New in LLVM 3.0?</a></li>
Chris Lattner4b538b92004-04-30 22:17:12 +000021 <li><a href="GettingStarted.html">Installation Instructions</a></li>
Dan Gohman44aa9212008-10-14 16:23:02 +000022 <li><a href="#knownproblems">Known Problems</a></li>
Misha Brukman500bc302003-11-22 00:38:41 +000023 <li><a href="#additionalinfo">Additional Information</a></li>
Chris Lattner79c3fe12003-10-02 04:57:28 +000024</ol>
25
Chris Lattner7911ce22004-05-23 21:07:27 +000026<div class="doc_author">
NAKAMURA Takumib9a33632011-04-09 02:13:37 +000027 <p>Written by the <a href="http://llvm.org/">LLVM Team</a></p>
Misha Brukman500bc302003-11-22 00:38:41 +000028</div>
Chris Lattner79c3fe12003-10-02 04:57:28 +000029
Chris Lattner49123fd2011-04-06 06:29:50 +000030<!--
Chad Rosiere6291d02011-05-27 22:50:46 +000031<h1 style="color:red">These are in-progress notes for the upcoming LLVM 3.0
Jeffrey Yasskinbec48772010-01-28 01:14:43 +000032release.<br>
33You may prefer the
Chad Rosiere6291d02011-05-27 22:50:46 +000034<a href="http://llvm.org/releases/2.9/docs/ReleaseNotes.html">LLVM 2.9
Dan Gohmanb44f6c62010-05-03 23:51:05 +000035Release Notes</a>.</h1>
Chris Lattner49123fd2011-04-06 06:29:50 +000036 -->
Jeffrey Yasskinbec48772010-01-28 01:14:43 +000037
Chris Lattner79c3fe12003-10-02 04:57:28 +000038<!-- *********************************************************************** -->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +000039<h2>
Misha Brukman500bc302003-11-22 00:38:41 +000040 <a name="intro">Introduction</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +000041</h2>
Chris Lattner79c3fe12003-10-02 04:57:28 +000042<!-- *********************************************************************** -->
43
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +000044<div>
Misha Brukman500bc302003-11-22 00:38:41 +000045
Chris Lattnerf6662f92008-10-13 17:57:36 +000046<p>This document contains the release notes for the LLVM Compiler
Bill Wendling7b7fa742011-10-26 18:46:16 +000047 Infrastructure, release 3.0. Here we describe the status of LLVM, including
Chris Lattner1ab8ce92011-11-27 18:47:37 +000048 major improvements from the previous release, improvements in various
49 subprojects of LLVM, and some of the current users of the code.
Bill Wendling7b7fa742011-10-26 18:46:16 +000050 All LLVM releases may be downloaded from
51 the <a href="http://llvm.org/releases/">LLVM releases web site</a>.</p>
Chris Lattner19092612003-10-02 16:38:05 +000052
Chris Lattner7506b1d2004-12-07 08:04:13 +000053<p>For more information about LLVM, including information about the latest
Bill Wendling7b7fa742011-10-26 18:46:16 +000054 release, please check out the <a href="http://llvm.org/">main LLVM web
55 site</a>. If you have questions or comments,
56 the <a href="http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev">LLVM
57 Developer's Mailing List</a> is a good place to send them.</p>
Chris Lattner79c3fe12003-10-02 04:57:28 +000058
Bill Wendling7b7fa742011-10-26 18:46:16 +000059<p>Note that if you are reading this file from a Subversion checkout or the main
60 LLVM web page, this document applies to the <i>next</i> release, not the
61 current one. To see the release notes for a specific release, please see the
62 <a href="http://llvm.org/releases/">releases page</a>.</p>
Misha Brukman500bc302003-11-22 00:38:41 +000063
64</div>
Michael J. Spencer60f790c2011-11-28 18:20:09 +000065
66
Chris Lattner96a445e2008-10-13 18:01:01 +000067<!-- *********************************************************************** -->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +000068<h2>
Chris Lattner96a445e2008-10-13 18:01:01 +000069 <a name="subproj">Sub-project Status Update</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +000070</h2>
Chris Lattner96a445e2008-10-13 18:01:01 +000071<!-- *********************************************************************** -->
Chris Lattnerea34f642008-06-08 21:34:41 +000072
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +000073<div>
Bill Wendling7b7fa742011-10-26 18:46:16 +000074
75<p>The LLVM 3.0 distribution currently consists of code from the core LLVM
76 repository (which roughly includes the LLVM optimizers, code generators and
Chris Lattnerc343e312011-11-10 20:15:40 +000077 supporting tools), and the Clang repository. In
Bill Wendling7b7fa742011-10-26 18:46:16 +000078 addition to this code, the LLVM Project includes other sub-projects that are
79 in development. Here we include updates on these subprojects.</p>
Chris Lattner96a445e2008-10-13 18:01:01 +000080
Chris Lattner8348b472008-06-08 21:58:17 +000081<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +000082<h3>
Chris Lattnerfb97b2d2008-10-13 18:11:54 +000083<a name="clang">Clang: C/C++/Objective-C Frontend Toolkit</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +000084</h3>
Chris Lattner8348b472008-06-08 21:58:17 +000085
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +000086<div>
Chris Lattner8348b472008-06-08 21:58:17 +000087
Chris Lattner095539f2010-04-26 17:42:18 +000088<p><a href="http://clang.llvm.org/">Clang</a> is an LLVM front end for the C,
Bill Wendling7b7fa742011-10-26 18:46:16 +000089 C++, and Objective-C languages. Clang aims to provide a better user
90 experience through expressive diagnostics, a high level of conformance to
91 language standards, fast compilation, and low memory use. Like LLVM, Clang
92 provides a modular, library-based architecture that makes it suitable for
93 creating or integrating with other development tools. Clang is considered a
94 production-quality compiler for C, Objective-C, C++ and Objective-C++ on x86
Chris Lattner1ab8ce92011-11-27 18:47:37 +000095 (32- and 64-bit), and for Darwin/ARM targets.</p>
Chris Lattner61358ab2009-10-13 17:48:04 +000096
Chandler Carruthcc966de2011-11-29 00:32:43 +000097<p>In the LLVM 3.0 time-frame, the Clang team has made many improvements:
Douglas Gregorba087df2011-10-15 00:48:01 +000098<ul>
Bill Wendling7b7fa742011-10-26 18:46:16 +000099 <li>Greatly improved support for building C++ applications, with greater
100 stability and better diagnostics.</li>
Michael J. Spencer60f790c2011-11-28 18:20:09 +0000101
Bill Wendling7b7fa742011-10-26 18:46:16 +0000102 <li><a href="http://clang.llvm.org/cxx_status.html">Improved support</a> for
103 the <a href="http://www.iso.org/iso/iso_catalogue/catalogue_tc/catalogue_detail.htm?csnumber=50372">C++
Chris Lattner1ab8ce92011-11-27 18:47:37 +0000104 2011</a> standard (aka "C++'0x"), including implementations of non-static data member
105 initializers, alias templates, delegating constructors, range-based
106 for loops, and implicitly-generated move constructors and move assignment
Bill Wendling7b7fa742011-10-26 18:46:16 +0000107 operators, among others.</li>
Douglas Gregorba087df2011-10-15 00:48:01 +0000108
Bill Wendling7b7fa742011-10-26 18:46:16 +0000109 <li>Implemented support for some features of the upcoming C1x standard,
110 including static assertions and generic selections.</li>
Michael J. Spencer60f790c2011-11-28 18:20:09 +0000111
Bill Wendling7b7fa742011-10-26 18:46:16 +0000112 <li>Better detection of include and linking paths for system headers and
113 libraries, especially for Linux distributions.</li>
Douglas Gregorba087df2011-10-15 00:48:01 +0000114
David Chisnall553284e2011-11-26 10:56:17 +0000115 <li>Several improvements to Objective-C support, including:
116
117 <ul>
118 <li><a href="http://clang.llvm.org/docs/AutomaticReferenceCounting.html">
119 Automatic Reference Counting</a> (ARC) and an improved memory model
120 cleanly separating object and C memory.</li>
121
Chris Lattner1ab8ce92011-11-27 18:47:37 +0000122 <li>A migration tool for moving manual retain/release code to ARC</li>
David Chisnall553284e2011-11-26 10:56:17 +0000123
124 <li>Better support for data hiding, allowing instance variables to be
125 declared in implementation contexts or class extensions</li>
Chris Lattner1ab8ce92011-11-27 18:47:37 +0000126 <li>Weak linking support for Objective-C classes</li>
David Chisnall553284e2011-11-26 10:56:17 +0000127 <li>Improved static type checking by inferring the return type of methods
128 such as +alloc and -init.</li>
129 </ul>
130
Chris Lattner1ab8ce92011-11-27 18:47:37 +0000131 Some new Objective-C features require either the Mac OS X 10.7 / iOS 5
132 Objective-C runtime, or version 1.6 or later of the GNUstep Objective-C
133 runtime version.</li>
Douglas Gregorba087df2011-10-15 00:48:01 +0000134
Bill Wendling7b7fa742011-10-26 18:46:16 +0000135 <li>Implemented a number of optimizations in <tt>libclang</tt>, the Clang C
136 interface, to improve the performance of code completion and the mapping
137 from source locations to abstract syntax tree nodes.</li>
Douglas Gregorba087df2011-10-15 00:48:01 +0000138</ul>
Chandler Carruthcc966de2011-11-29 00:32:43 +0000139For more details about the changes to Clang since the 2.9 release, see the
140<a href="http://clang.llvm.org/docs/ReleaseNotes.html">Clang release notes</a>
141</p>
Douglas Gregorba087df2011-10-15 00:48:01 +0000142
Michael J. Spencer60f790c2011-11-28 18:20:09 +0000143
Duncan Sandsf3ba7af2011-04-06 08:07:40 +0000144<p>If Clang rejects your code but another compiler accepts it, please take a
Bill Wendling7b7fa742011-10-26 18:46:16 +0000145 look at the <a href="http://clang.llvm.org/compatibility.html">language
146 compatibility</a> guide to make sure this is not intentional or a known
147 issue.</p>
Bill Wendling741748a2008-10-27 09:27:33 +0000148
Chris Lattnerfb97b2d2008-10-13 18:11:54 +0000149</div>
150
151<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000152<h3>
Duncan Sands528a5102011-04-04 11:09:08 +0000153<a name="dragonegg">DragonEgg: GCC front-ends, LLVM back-end</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000154</h3>
Chris Lattner61358ab2009-10-13 17:48:04 +0000155
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000156<div>
Bill Wendling7b7fa742011-10-26 18:46:16 +0000157<p><a href="http://dragonegg.llvm.org/">DragonEgg</a> is a
158 <a href="http://gcc.gnu.org/wiki/plugins">gcc plugin</a> that replaces GCC's
Duncan Sands77352c92011-11-10 18:44:29 +0000159 optimizers and code generators with LLVM's. It works with gcc-4.5 or gcc-4.6,
160 targets the x86-32 and x86-64 processor families, and has been successfully
161 used on the Darwin, FreeBSD, KFreeBSD, Linux and OpenBSD platforms. It fully
162 supports Ada, C, C++ and Fortran. It has partial support for Go, Java, Obj-C
163 and Obj-C++.</p>
Duncan Sands749fd832010-04-02 09:23:15 +0000164
Bill Wendling7b7fa742011-10-26 18:46:16 +0000165<p>The 3.0 release has the following notable changes:</p>
166
Chris Lattner1ab8ce92011-11-27 18:47:37 +0000167 <ul>
Duncan Sands77352c92011-11-10 18:44:29 +0000168 <li>GCC version 4.6 is now fully supported.</li>
169
170 <li>Patching and building GCC is no longer required: the plugin should work
171 with your system GCC (version 4.5 or 4.6; on Debian/Ubuntu systems the
172 gcc-4.5-plugin-dev or gcc-4.6-plugin-dev package is also needed).</li>
173
174 <li>The <tt>-fplugin-arg-dragonegg-enable-gcc-optzns</tt> option, which runs
175 GCC's optimizers as well as LLVM's, now works much better. This is the
Chris Lattner1ab8ce92011-11-27 18:47:37 +0000176 option to use if you want ultimate performance! It is still experimental
Duncan Sands0fcb4312011-11-30 08:46:05 +0000177 though: it may cause the plugin to crash. Setting the optimization level
178 to <tt>-O4</tt> when using this option will optimize even harder, though
Chris Lattner9b15ba92011-12-01 17:25:28 +0000179 this usually doesn't result in any improvement over <tt>-O3</tt>.</li>
Duncan Sands77352c92011-11-10 18:44:29 +0000180
181 <li>The type and constant conversion logic has been almost entirely rewritten,
182 fixing a multitude of obscure bugs.</li>
183
Duncan Sands4b1da2b2010-09-30 17:37:34 +0000184</ul>
Chris Lattner61358ab2009-10-13 17:48:04 +0000185
186</div>
187
Chris Lattner75547712010-10-03 23:49:06 +0000188<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000189<h3>
Chris Lattner75547712010-10-03 23:49:06 +0000190<a name="compiler-rt">compiler-rt: Compiler Runtime Library</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000191</h3>
Chris Lattner75547712010-10-03 23:49:06 +0000192
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000193<div>
Bill Wendling7b7fa742011-10-26 18:46:16 +0000194
195<p>The new LLVM <a href="http://compiler-rt.llvm.org/">compiler-rt project</a>
196 is a simple library that provides an implementation of the low-level
197 target-specific hooks required by code generation and other runtime
198 components. For example, when compiling for a 32-bit target, converting a
199 double to a 64-bit unsigned integer is compiled into a runtime call to the
200 "__fixunsdfdi" function. The compiler-rt library provides highly optimized
201 implementations of this and other low-level routines (some are 3x faster than
202 the equivalent libgcc routines).</p>
Chris Lattner75547712010-10-03 23:49:06 +0000203
Chris Lattner1ab8ce92011-11-27 18:47:37 +0000204<p>In the LLVM 3.0 timeframe, the target specific ARM code has converted to
205 "unified" assembly syntax, and several new functions have been added to the
206 library.</p>
Chris Lattner75547712010-10-03 23:49:06 +0000207
208</div>
Chris Lattner61358ab2009-10-13 17:48:04 +0000209
210<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000211<h3>
Chris Lattnere07043c2010-09-29 05:30:03 +0000212<a name="lldb">LLDB: Low Level Debugger</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000213</h3>
Chris Lattnere07043c2010-09-29 05:30:03 +0000214
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000215<div>
Michael J. Spencer60f790c2011-11-28 18:20:09 +0000216
Chris Lattner9e896712011-11-27 18:53:41 +0000217<p>LLDB is a ground-up implementation of a command line debugger, as well as a
218 debugger API that can be used from other applications. LLDB makes use of the
219 Clang parser to provide high-fidelity expression parsing (particularly for
220 C++) and uses the LLVM JIT for target support.</p>
Chris Lattnere07043c2010-09-29 05:30:03 +0000221
Bill Wendling7b7fa742011-10-26 18:46:16 +0000222<p>LLDB has advanced by leaps and bounds in the 3.0 timeframe. It is
223 dramatically more stable and useful, and includes both a
224 new <a href="http://lldb.llvm.org/tutorial.html">tutorial</a> and
225 a <a href="http://lldb.llvm.org/lldb-gdb.html">side-by-side comparison with
226 GDB</a>.</p>
Chris Lattnere07043c2010-09-29 05:30:03 +0000227
228</div>
229
230<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000231<h3>
Chris Lattnere07043c2010-09-29 05:30:03 +0000232<a name="libc++">libc++: C++ Standard Library</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000233</h3>
Chris Lattnere07043c2010-09-29 05:30:03 +0000234
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000235<div>
Chris Lattnere07043c2010-09-29 05:30:03 +0000236
Bill Wendling7b7fa742011-10-26 18:46:16 +0000237<p>Like compiler_rt, libc++ is now <a href="DeveloperPolicy.html#license">dual
238 licensed</a> under the MIT and UIUC license, allowing it to be used more
239 permissively.</p>
Chris Lattnere07043c2010-09-29 05:30:03 +0000240
David Chisnall553284e2011-11-26 10:56:17 +0000241<p>Libc++ has been ported to FreeBSD and imported into the base system. It is
242 planned to be the default STL implementation for FreeBSD 10.</p>
243
Chris Lattnere07043c2010-09-29 05:30:03 +0000244</div>
245
Chris Lattner3d6a80a2011-04-07 03:08:22 +0000246<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000247<h3>
Chris Lattner3d6a80a2011-04-07 03:08:22 +0000248<a name="vmkit">VMKit</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000249</h3>
Chris Lattner3d6a80a2011-04-07 03:08:22 +0000250
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000251<div>
Bill Wendling7b7fa742011-10-26 18:46:16 +0000252
Nicolas Geoffray54d5df92011-11-10 23:37:56 +0000253 <p>The <a href="http://vmkit.llvm.org/">VMKit project</a> is an
254 implementation of a Java Virtual Machine (Java VM or JVM) that uses LLVM for
255 static and just-in-time compilation.
256
257 <p>In the LLVM 3.0 time-frame, VMKit has had significant improvements on both
258 runtime and startup performance:</p>
259
260 <ul>
261 <li>Precompilation: by compiling ahead of time a small subset of Java's core
262 library, the startup performance have been highly optimized to the point that
263 running a 'Hello World' program takes less than 30 milliseconds.</li>
264
265 <li>Customization: by customizing virtual methods for individual classes,
266 the VM can statically determine the target of a virtual call, and decide to
267 inline it.</li>
268
269 <li>Inlining: the VM does more inlining than it did before, by allowing more
270 bytecode instructions to be inlined, and thanks to customization. It also
271 inlines GC barriers, and object allocations.</li>
272
273 <li>New exception model: the generated code for a method that does not do
274 any try/catch is not penalized anymore by the eventuality of calling a
275 method that throws an exception. Instead, the method that throws the
276 exception jumps directly to the method that could catch it.</li>
277 </ul>
Bill Wendling7b7fa742011-10-26 18:46:16 +0000278
Chris Lattner3d6a80a2011-04-07 03:08:22 +0000279</div>
Michael J. Spencer60f790c2011-11-28 18:20:09 +0000280
Chris Lattner9e896712011-11-27 18:53:41 +0000281
282<!--=========================================================================-->
283<h3>
284<a name="LLBrowse">LLBrowse: IR Browser</a>
285</h3>
286
287<div>
288
289<p><a href="http://llvm.org/svn/llvm-project/llbrowse/trunk/doc/LLBrowse.html">
290 LLBrowse</a> is an interactive viewer for LLVM modules. It can load any LLVM
291 module and displays its contents as an expandable tree view, facilitating an
292 easy way to inspect types, functions, global variables, or metadata nodes. It
293 is fully cross-platform, being based on the popular wxWidgets GUI
294 toolkit.</p>
295
296</div>
297
298
Daniel Dunbar97b01a82010-10-04 17:39:47 +0000299<!--=========================================================================-->
Chris Lattner7a8e6c52011-04-05 18:38:45 +0000300<!--
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000301<h3>
Daniel Dunbar97b01a82010-10-04 17:39:47 +0000302<a name="klee">KLEE: A Symbolic Execution Virtual Machine</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000303</h3>
Daniel Dunbar97b01a82010-10-04 17:39:47 +0000304
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000305<div>
Daniel Dunbar97b01a82010-10-04 17:39:47 +0000306<p>
307<a href="http://klee.llvm.org/">KLEE</a> is a symbolic execution framework for
308programs in LLVM bitcode form. KLEE tries to symbolically evaluate "all" paths
309through the application and records state transitions that lead to fault
310states. This allows it to construct testcases that lead to faults and can even
311be used to verify some algorithms.
312</p>
313
Chris Lattnerbe2e1b52011-03-10 07:43:44 +0000314<p>UPDATE!</p>
Chris Lattner7a8e6c52011-04-05 18:38:45 +0000315</div>-->
Daniel Dunbar97b01a82010-10-04 17:39:47 +0000316
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000317</div>
Daniel Dunbar97b01a82010-10-04 17:39:47 +0000318
Chris Lattnerab68e9e2009-02-26 22:33:38 +0000319<!-- *********************************************************************** -->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000320<h2>
Chad Rosiere6291d02011-05-27 22:50:46 +0000321 <a name="externalproj">External Open Source Projects Using LLVM 3.0</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000322</h2>
Chris Lattnerab68e9e2009-02-26 22:33:38 +0000323<!-- *********************************************************************** -->
324
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000325<div>
Chris Lattner61358ab2009-10-13 17:48:04 +0000326
327<p>An exciting aspect of LLVM is that it is used as an enabling technology for
328 a lot of other language and tools projects. This section lists some of the
Chad Rosiere6291d02011-05-27 22:50:46 +0000329 projects that have already been updated to work with LLVM 3.0.</p>
Chris Lattner75547712010-10-03 23:49:06 +0000330
Chris Lattner3bfe57e2011-04-06 01:13:49 +0000331<!--=========================================================================-->
Bill Wendling7be6bc52011-10-26 00:17:54 +0000332<h3>AddressSanitizer</h3>
Michael J. Spencer60f790c2011-11-28 18:20:09 +0000333
Bill Wendling7be6bc52011-10-26 00:17:54 +0000334<div>
335
336<p><a href="http://code.google.com/p/address-sanitizer/">AddressSanitizer</a>
337 uses compiler instrumentation and a specialized malloc library to find C/C++
338 bugs such as use-after-free and out-of-bound accesses to heap, stack, and
339 globals. The key feature of the tool is speed: the average slowdown
340 introduced by AddressSanitizer is less than 2x.</p>
341
342</div>
343
344<!--=========================================================================-->
Bill Wendling29817ea2011-10-26 00:14:36 +0000345<h3>ClamAV</h3>
Michael J. Spencer60f790c2011-11-28 18:20:09 +0000346
Bill Wendling29817ea2011-10-26 00:14:36 +0000347<div>
Bill Wendlingf2a78332011-10-25 01:01:42 +0000348
Bill Wendling29817ea2011-10-26 00:14:36 +0000349<p><a href="http://www.clamav.net">Clam AntiVirus</a> is an open source (GPL)
350 anti-virus toolkit for UNIX, designed especially for e-mail scanning on mail
351 gateways.</p>
352
353<p>Since version 0.96 it
354 has <a href="http://vrt-sourcefire.blogspot.com/2010/09/introduction-to-clamavs-low-level.html">bytecode
Chris Lattner4f0fe432011-11-27 19:26:30 +0000355 signatures</a> that allow writing detections for complex malware.
356 It uses LLVM's JIT to speed up the execution of bytecode on X86, X86-64,
Bill Wendling29817ea2011-10-26 00:14:36 +0000357 PPC32/64, falling back to its own interpreter otherwise. The git version was
358 updated to work with LLVM 3.0.</p>
359
360</div>
361
362<!--=========================================================================-->
Tobias Grosserae5a6fd2011-11-14 09:09:26 +0000363<h3>clang_complete for VIM</h3>
364
365<div>
366
367<p><a href="https://github.com/Rip-Rip/clang_complete">clang_complete</a> is a
368 VIM plugin, that provides accurate C/C++ autocompletion using the clang front
369 end. The development version of clang complete, can directly use libclang
370 which can maintain a cache to speed up auto completion.</p>
371
372</div>
373
374<!--=========================================================================-->
Bill Wendling65d1f412011-10-26 18:23:06 +0000375<h3>clReflect</h3>
376
377<div>
378
379<p><a href="https://bitbucket.org/dwilliamson/clreflect">clReflect</a> is a C++
380 parser that uses clang/LLVM to derive a light-weight reflection database
381 suitable for use in game development. It comes with a very simple runtime
382 library for loading and querying the database, requiring no external
383 dependencies (including CRT), and an additional utility library for object
384 management and serialisation.</p>
385
386</div>
387
388<!--=========================================================================-->
Bill Wendling63507d12011-10-29 01:10:01 +0000389<h3>Cling C++ Interpreter</h3>
390
391<div>
392
393<p><a href="http://cern.ch/cling">Cling</a> is an interactive compiler interface
Chris Lattner4f0fe432011-11-27 19:26:30 +0000394 (aka C++ interpreter). It supports C++ and C, and uses LLVM's JIT and the
395 Clang parser. It has a prompt interface, runs source files, calls into shared
Bill Wendling63507d12011-10-29 01:10:01 +0000396 libraries, prints the value of expressions, even does runtime lookup of
397 identifiers (dynamic scopes). And it just behaves like one would expect from
398 an interpreter.</p>
399
400</div>
401
402<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000403<h3>Crack Programming Language</h3>
Chris Lattner3bfe57e2011-04-06 01:13:49 +0000404
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000405<div>
Bill Wendling55d6e672011-11-03 20:10:01 +0000406
407<p><a href="http://code.google.com/p/crack-language/">Crack</a> aims to provide
408 the ease of development of a scripting language with the performance of a
409 compiled language. The language derives concepts from C++, Java and Python,
410 incorporating object-oriented programming, operator overloading and strong
411 typing.</p>
412
Chris Lattner3bfe57e2011-04-06 01:13:49 +0000413</div>
Michael J. Spencer60f790c2011-11-28 18:20:09 +0000414
Chris Lattner3bfe57e2011-04-06 01:13:49 +0000415<!--=========================================================================-->
Bill Wendlingb99486f2011-11-08 05:22:54 +0000416<h3>Eero</h3>
Michael J. Spencer60f790c2011-11-28 18:20:09 +0000417
Bill Wendlingb99486f2011-11-08 05:22:54 +0000418<div>
419
420<p><a href="http://eerolanguage.org/">Eero</a> is a fully
421 header-and-binary-compatible dialect of Objective-C 2.0, implemented with a
422 patched version of the Clang/LLVM compiler. It features a streamlined syntax,
423 Python-like indentation, and new operators, for improved readability and
424 reduced code clutter. It also has new features such as limited forms of
425 operator overloading and namespaces, and strict (type-and-operator-safe)
426 enumerations. It is inspired by languages such as Smalltalk, Python, and
427 Ruby.</p>
428
429</div>
430
431<!--=========================================================================-->
Chris Lattneradb417a2011-11-25 20:28:16 +0000432<h3>FAUST Real-Time Audio Signal Processing Language</h3>
Michael J. Spencer60f790c2011-11-28 18:20:09 +0000433
Chris Lattneradb417a2011-11-25 20:28:16 +0000434<div>
435
436<p><a href="http://faust.grame.fr/">FAUST</a> is a compiled language for
437 real-time audio signal processing. The name FAUST stands for Functional
438 AUdio STream. Its programming model combines two approaches: functional
439 programming and block diagram composition. In addition with the C, C++, Java
440 output formats, the Faust compiler can now generate LLVM bitcode, and works
441 with LLVM 2.7-3.0.
442 </p>
443
444</div>
445
446<!--=========================================================================-->
Bill Wendlingf9778192011-10-26 00:09:55 +0000447<h3>Glasgow Haskell Compiler (GHC)</h3>
Michael J. Spencer60f790c2011-11-28 18:20:09 +0000448
Bill Wendlingf9778192011-10-26 00:09:55 +0000449<div>
450
451<p>GHC is an open source, state-of-the-art programming suite for Haskell, a
452 standard lazy functional programming language. It includes an optimizing
453 static compiler generating good code for a variety of platforms, together
454 with an interactive system for convenient, quick development.</p>
455
456<p>GHC 7.0 and onwards include an LLVM code generator, supporting LLVM 2.8 and
457 later. Since LLVM 2.9, GHC now includes experimental support for the ARM
458 platform with LLVM 3.0.</p>
459
460</div>
461
462<!--=========================================================================-->
Bill Wendlingf62333d2011-10-25 20:35:31 +0000463<h3>gwXscript</h3>
464
465<div>
466
467<p><a href="http://botwars.tk/gwscript/">gwXscript</a> is an object oriented,
Bill Wendling7c38de22011-10-26 04:24:15 +0000468 aspect oriented programming language which can create both executables (ELF,
Bill Wendlingf62333d2011-10-25 20:35:31 +0000469 EXE) and shared libraries (DLL, SO, DYNLIB). The compiler is implemented in
470 its own language and translates scripts into LLVM-IR which can be optimized
471 and translated into native code by the LLVM framework. Source code in
472 gwScript contains definitions that expand the namespaces. So you can build
473 your project and simply 'plug out' features by removing a file. The remaining
474 project does not leave scars since you directly separate concerns by the
475 'template' feature of gwX. It is also possible to add new features to a
476 project by just adding files and without editing the original project. This
477 language is used for example to create games or content management systems
478 that should be extendable.</p>
479
480<p>gwXscript is strongly typed and offers comfort with its native types string,
481 hash and array. You can easily write new libraries in gwXscript or native
482 code. gwXscript is type safe and users should not be able to crash your
483 program or execute malicious code except code that is eating CPU time.</p>
484
485</div>
486
487<!--=========================================================================-->
Bill Wendling50cacc82011-10-26 22:55:18 +0000488<h3>include-what-you-use</h3>
489
490<div>
491
492<p><a href="http://code.google.com/p/include-what-you-use">include-what-you-use</a>
493 is a tool to ensure that a file directly <code>#include</code>s
494 all <code>.h</code> files that provide a symbol that the file uses. It also
495 removes superfluous <code>#include</code>s from source files.</p>
496
497</div>
498
499<!--=========================================================================-->
Bill Wendling32dc4d92011-11-07 22:05:17 +0000500<h3>ispc: The Intel SPMD Program Compiler</h3>
501
502<div>
503
504<p><a href="http://ispc.github.com">ispc</a> is a compiler for "single program,
505 multiple data" (SPMD) programs. It compiles a C-based SPMD programming
506 language to run on the SIMD units of CPUs; it often delivers 5-6x speedups on
507 a single core of a CPU with an 8-wide SIMD unit compared to serial code,
508 while still providing a clean and easy-to-understand programming model. For
509 an introduction to the language and its performance,
Chris Lattnerc343e312011-11-10 20:15:40 +0000510 see <a href="http://ispc.github.com/example.html">the walkthrough</a> of a short
Bill Wendling32dc4d92011-11-07 22:05:17 +0000511 example program. ispc is licensed under the BSD license.</p>
512
513</div>
Michael J. Spencer60f790c2011-11-28 18:20:09 +0000514
Chris Lattnercc089772011-11-25 20:36:17 +0000515<!--=========================================================================-->
516<h3>The Julia Programming Language</h3>
517
518<div>
519
520<p><a href="http://github.com/JuliaLang/julia">Julia</a> is a high-level,
521 high-performance dynamic language for technical
522 computing. It provides a sophisticated compiler, distributed parallel
523 execution, numerical accuracy, and an extensive mathematical function
524 library. The compiler uses type inference to generate fast code
525 without any type declarations, and uses LLVM's optimization passes and
526 JIT compiler. The language is designed around multiple dispatch,
527 giving programs a large degree of flexibility. It is ready for use on many
528 kinds of problems.</p>
529</div>
Bill Wendling32dc4d92011-11-07 22:05:17 +0000530
531<!--=========================================================================-->
Bill Wendling57fd8762011-10-26 18:20:54 +0000532<h3>LanguageKit and Pragmatic Smalltalk</h3>
533
534<div>
535
536<p><a href="http://etoileos.com/etoile/features/languagekit/">LanguageKit</a> is
537 a framework for implementing dynamic languages sharing an object model with
538 Objective-C. It provides static and JIT compilation using LLVM along with
539 its own interpreter. Pragmatic Smalltalk is a dialect of Smalltalk, built on
540 top of LanguageKit, that interfaces directly with Objective-C, sharing the
541 same object representation and message sending behaviour. These projects are
Benjamin Kramer7c5025b2011-11-25 21:26:00 +0000542 developed as part of the &Eacute;toil&eacute; desktop environment.</p>
Bill Wendling57fd8762011-10-26 18:20:54 +0000543
544</div>
545
546<!--=========================================================================-->
Bill Wendling03250532011-11-01 04:08:23 +0000547<h3>LuaAV</h3>
548
549<div>
550
551<p><a href="http://lua-av.mat.ucsb.edu/blog/">LuaAV</a> is a real-time
552 audiovisual scripting environment based around the Lua language and a
553 collection of libraries for sound, graphics, and other media protocols. LuaAV
554 uses LLVM and Clang to JIT compile efficient user-defined audio synthesis
555 routines specified in a declarative syntax.</p>
556
557</div>
558
559<!--=========================================================================-->
Bill Wendlingd4821b82011-10-26 00:16:17 +0000560<h3>Mono</h3>
561
562<div>
563
564<p>An open source, cross-platform implementation of C# and the CLR that is
565 binary compatible with Microsoft.NET. Has an optional, dynamically-loaded
566 LLVM code generation backend in Mini, the JIT compiler.</p>
567
Chris Lattner4f0fe432011-11-27 19:26:30 +0000568<p>Note that we use a Git mirror of LLVM <a
569 href="https://github.com/mono/llvm">with some patches</a>.</p>
Bill Wendlingd4821b82011-10-26 00:16:17 +0000570
571</div>
572
573<!--=========================================================================-->
Tobias Grosser093cb7e2011-11-14 09:09:23 +0000574<h3>Polly</h3>
575
576<div>
577
578<p><a href="http://polly.grosser.es">Polly</a> is an advanced data-locality
579 optimizer and automatic parallelizer. It uses an advanced, mathematical
580 model to calculate detailed data dependency information which it uses to
581 optimize the loop structure of a program. Polly can speed up sequential code
582 by improving memory locality and consequently the cache use. Furthermore,
583 Polly is able to expose different kind of parallelism which it exploits by
584 introducing (basic) OpenMP and SIMD code. A mid-term goal of Polly is to
585 automatically create optimized GPU code.</p>
586
587</div>
588
589<!--=========================================================================-->
Bill Wendlingba226272011-10-25 20:37:45 +0000590<h3>Portable OpenCL (pocl)</h3>
591
592<div>
593
594<p>Portable OpenCL is an open source implementation of the OpenCL standard which
595 can be easily adapted for new targets. One of the goals of the project is
596 improving performance portability of OpenCL programs, avoiding the need for
597 target-dependent manual optimizations. A "native" target is included, which
598 allows running OpenCL kernels on the host (CPU).</p>
599
600</div>
601
602<!--=========================================================================-->
Bill Wendling0bad98c2011-10-25 20:39:06 +0000603<h3>Pure</h3>
Michael J. Spencer60f790c2011-11-28 18:20:09 +0000604
Bill Wendling0bad98c2011-10-25 20:39:06 +0000605<div>
606<p><a href="http://pure-lang.googlecode.com/">Pure</a> is an
607 algebraic/functional programming language based on term rewriting. Programs
608 are collections of equations which are used to evaluate expressions in a
609 symbolic fashion. The interpreter uses LLVM as a backend to JIT-compile Pure
610 programs to fast native code. Pure offers dynamic typing, eager and lazy
611 evaluation, lexical closures, a hygienic macro system (also based on term
612 rewriting), built-in list and matrix support (including list and matrix
613 comprehensions) and an easy-to-use interface to C and other programming
614 languages (including the ability to load LLVM bitcode modules, and inline C,
615 C++, Fortran and Faust code in Pure programs if the corresponding LLVM-enabled
616 compilers are installed).</p>
Michael J. Spencer60f790c2011-11-28 18:20:09 +0000617
Bill Wendling0bad98c2011-10-25 20:39:06 +0000618<p>Pure version 0.48 has been tested and is known to work with LLVM 3.0
619 (and continues to work with older LLVM releases &gt;= 2.5).</p>
620
621</div>
622
623<!--=========================================================================-->
Bill Wendling537d85b2011-10-26 00:12:04 +0000624<h3>Renderscript</h3>
625
626<div>
627
628<p><a href="http://developer.android.com/guide/topics/renderscript/index.html">Renderscript</a>
629 is Android's advanced 3D graphics rendering and compute API. It provides a
630 portable C99-based language with extensions to facilitate common use cases
631 for enhancing graphics and thread level parallelism. The Renderscript
632 compiler frontend is based on Clang/LLVM. It emits a portable bitcode format
633 for the actual compiled script code, as well as reflects a Java interface for
634 developers to control the execution of the compiled bitcode. Executable
635 machine code is then generated from this bitcode by an LLVM backend on the
636 device. Renderscript is thus able to provide a mechanism by which Android
637 developers can improve performance of their applications while retaining
638 portability.</p>
639
640</div>
641
642<!--=========================================================================-->
Bill Wendling7d5b6212011-10-25 20:40:26 +0000643<h3>SAFECode</h3>
644
645<div>
646
647<p><a href="http://safecode.cs.illinois.edu">SAFECode</a> is a memory safe C/C++
648 compiler built using LLVM. It takes standard, unannotated C/C++ code,
649 analyzes the code to ensure that memory accesses and array indexing
650 operations are safe, and instruments the code with run-time checks when
651 safety cannot be proven statically. SAFECode can be used as a debugging aid
652 (like Valgrind) to find and repair memory safety bugs. It can also be used
653 to protect code from security attacks at run-time.</p>
654
655</div>
656
657<!--=========================================================================-->
Bill Wendling02b77b72011-10-26 07:38:19 +0000658<h3>The Stupid D Compiler (SDC)</h3>
659
660<div>
661
662<p><a href="https://github.com/bhelyer/SDC">The Stupid D Compiler</a> is a
663 project seeking to write a self-hosting compiler for the D programming
664 language without using the frontend of the reference compiler (DMD).</p>
665
666</div>
667
668<!--=========================================================================-->
Bill Wendling2d7b4af2011-10-25 20:24:32 +0000669<h3>TTA-based Co-design Environment (TCE)</h3>
670
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000671<div>
Bill Wendling2d7b4af2011-10-25 20:24:32 +0000672
Chris Lattner3bfe57e2011-04-06 01:13:49 +0000673<p>TCE is a toolset for designing application-specific processors (ASP) based on
Bill Wendling2d7b4af2011-10-25 20:24:32 +0000674 the Transport triggered architecture (TTA). The toolset provides a complete
675 co-design flow from C/C++ programs down to synthesizable VHDL and parallel
676 program binaries. Processor customization points include the register files,
677 function units, supported operations, and the interconnection network.</p>
Michael J. Spencer60f790c2011-11-28 18:20:09 +0000678
Chris Lattner3bfe57e2011-04-06 01:13:49 +0000679<p>TCE uses Clang and LLVM for C/C++ language support, target independent
Bill Wendling2d7b4af2011-10-25 20:24:32 +0000680 optimizations and also for parts of code generation. It generates new
Bill Wendling7b7fa742011-10-26 18:46:16 +0000681 LLVM-based code generators "on the fly" for the designed TTA processors and
682 loads them in to the compiler backend as runtime libraries to avoid
Bill Wendling2d7b4af2011-10-25 20:24:32 +0000683 per-target recompilation of larger parts of the compiler chain.</p>
Chris Lattner3bfe57e2011-04-06 01:13:49 +0000684
Bill Wendling7b7fa742011-10-26 18:46:16 +0000685</div>
Michael J. Spencer60f790c2011-11-28 18:20:09 +0000686
Chris Lattner3bfe57e2011-04-06 01:13:49 +0000687<!--=========================================================================-->
Bill Wendling628c2662011-10-25 20:27:37 +0000688<h3>Tart Programming Language</h3>
689
690<div>
691
692<p><a href="http://code.google.com/p/tart/">Tart</a> is a general-purpose,
693 strongly typed programming language designed for application
694 developers. Strongly inspired by Python and C#, Tart focuses on practical
695 solutions for the professional software developer, while avoiding the clutter
696 and boilerplate of legacy languages like Java and C++. Although Tart is still
697 in development, the current implementation supports many features expected of
698 a modern programming language, such as garbage collection, powerful
699 bidirectional type inference, a greatly simplified syntax for template
700 metaprogramming, closures and function literals, reflection, operator
701 overloading, explicit mutability and immutability, and much more. Tart is
702 flexible enough to accommodate a broad range of programming styles and
703 philosophies, while maintaining a strong commitment to simplicity, minimalism
704 and elegance in design.</p>
705
706</div>
707
708<!--=========================================================================-->
Bill Wendling644ce532011-10-26 09:25:01 +0000709<h3>ThreadSanitizer</h3>
710
711<div>
712
713<p><a href="http://code.google.com/p/data-race-test/">ThreadSanitizer</a> is a
714 data race detector for (mostly) C and C++ code, available for Linux, Mac OS
715 and Windows. On different systems, we use binary instrumentation frameworks
Bill Wendlingae8538e2011-10-29 01:11:15 +0000716 (Valgrind and Pin) as frontends that generate the program events for the race
717 detection algorithm. On Linux, there's an option of using LLVM-based
718 compile-time instrumentation.</p>
Bill Wendling644ce532011-10-26 09:25:01 +0000719
720</div>
721
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000722</div>
723
Chris Lattner8348b472008-06-08 21:58:17 +0000724<!-- *********************************************************************** -->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000725<h2>
Chad Rosiere6291d02011-05-27 22:50:46 +0000726 <a name="whatsnew">What's New in LLVM 3.0?</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000727</h2>
Chris Lattner8348b472008-06-08 21:58:17 +0000728<!-- *********************************************************************** -->
729
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000730<div>
Chris Lattnerf8e0b4e2008-06-08 22:59:35 +0000731
Chris Lattner61358ab2009-10-13 17:48:04 +0000732<p>This release includes a huge number of bug fixes, performance tweaks and
Bill Wendling7b7fa742011-10-26 18:46:16 +0000733 minor improvements. Some of the major improvements and new features are
734 listed in this section.</p>
Chris Lattner61358ab2009-10-13 17:48:04 +0000735
Chris Lattner914ce462010-04-22 06:28:20 +0000736<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000737<h3>
Chris Lattner252b83d2008-02-06 18:00:06 +0000738<a name="majorfeatures">Major New Features</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000739</h3>
Chris Lattner252b83d2008-02-06 18:00:06 +0000740
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000741<div>
Chris Lattner252b83d2008-02-06 18:00:06 +0000742
Chris Lattner1ab8ce92011-11-27 18:47:37 +0000743 <!-- Features that need text if they're finished for 3.1:
744 ARM EHABI
745 combiner-aa?
746 strong phi elim
747 loop dependence analysis
748 CorrelatedValuePropagation
749 lib/Transforms/IPO/MergeFunctions.cpp => consider for 3.1.
Chris Lattner1c80fbf2011-11-27 20:51:47 +0000750 Integrated assembler on by default for arm/thumb?
751
Chris Lattner1ab8ce92011-11-27 18:47:37 +0000752 -->
753
Chris Lattner1c80fbf2011-11-27 20:51:47 +0000754 <!-- Near dead:
Chris Lattnerdec23b62011-11-15 22:13:27 +0000755 Analysis/RegionInfo.h + Dom Frontiers
756 SparseBitVector: used in LiveVar.
Chris Lattner5a1731d2011-11-27 08:32:32 +0000757 llvm/lib/Archive - replace with lib object?
Chris Lattner1c80fbf2011-11-27 20:51:47 +0000758 -->
Chris Lattner6a007d12011-11-25 20:33:27 +0000759
Chris Lattner064caf92011-11-27 21:30:28 +0000760<p>LLVM 3.0 includes several major changes and big features:</p>
Bill Wendlingbc5f6dd2011-10-26 18:33:01 +0000761
Chris Lattner791f77b2008-06-05 06:25:56 +0000762<ul>
Michael J. Spencer60f790c2011-11-28 18:20:09 +0000763<li>llvm-gcc is no longer supported, and not included in the release. We
Chris Lattner1c80fbf2011-11-27 20:51:47 +0000764 recommend switching to <a
Michael J. Spencer60f790c2011-11-28 18:20:09 +0000765 href="http://clang.llvm.org/">Clang</a> or <a
Chris Lattner1c80fbf2011-11-27 20:51:47 +0000766 href="http://dragonegg.llvm.org/">DragonEgg</a>.</li>
Bill Wendlingbc5f6dd2011-10-26 18:33:01 +0000767
Chris Lattner1c80fbf2011-11-27 20:51:47 +0000768<li>The linear scan register allocator has been replaced with a new "greedy"
769 register allocator, enabling live range splitting and many other
770 optimizations that lead to better code quality. Please see its <a
Michael J. Spencer60f790c2011-11-28 18:20:09 +0000771 href="http://blog.llvm.org/2011/09/greedy-register-allocation-in-llvm-30.html">blog post</a> or its talk at the <a
Chris Lattner1c80fbf2011-11-27 20:51:47 +0000772 href="http://llvm.org/devmtg/2011-11/">Developer Meeting</a>
773 for more information.</li>
Chris Lattner064caf92011-11-27 21:30:28 +0000774<li>LLVM IR now includes full support for <a href="Atomics.html">atomics
775 memory operations</a> intended to support the C++'11 and C'1x memory models.
776 This includes <a href="LangRef.html#memoryops">atomic load and store,
777 compare and exchange, and read/modify/write instructions</a> as well as a
778 full set of <a href="LangRef.html#ordering">memory ordering constraints</a>.
779 Please see the <a href="Atomics.html">Atomics Guide</a> for more
780 information.
Chris Lattner1c80fbf2011-11-27 20:51:47 +0000781</li>
782<li>The LLVM IR exception handling representation has been redesigned and
783 reimplemented, making it more elegant, fixing a huge number of bugs, and
Benjamin Kramer6e7a13d2011-11-29 19:24:11 +0000784 enabling inlining and other optimizations. Please see its <a href=
785 "http://blog.llvm.org/2011/11/llvm-30-exception-handling-redesign.html">blog
786 post</a> and the <a href="ExceptionHandling.html">Exception Handling
Chris Lattner064caf92011-11-27 21:30:28 +0000787 documentation</a> for more information.</li>
Chris Lattner1c80fbf2011-11-27 20:51:47 +0000788<li>The LLVM IR Type system has been redesigned and reimplemented, making it
789 faster and solving some long-standing problems.
Michael J. Spencer60f790c2011-11-28 18:20:09 +0000790 Please see its <a
Chris Lattner1c80fbf2011-11-27 20:51:47 +0000791 href="http://blog.llvm.org/2011/11/llvm-30-type-system-rewrite.html">blog
792 post</a> for more information.</li>
Michael J. Spencer60f790c2011-11-28 18:20:09 +0000793
Chris Lattner1c80fbf2011-11-27 20:51:47 +0000794<li>The MIPS backend has made major leaps in this release, going from an
Chris Lattner064caf92011-11-27 21:30:28 +0000795 experimental target to being virtually production quality and supporting a
Chris Lattner1c80fbf2011-11-27 20:51:47 +0000796 wide variety of MIPS subtargets. See the <a href="#MIPS">MIPS section</a>
797 below for more information.</li>
Michael J. Spencer60f790c2011-11-28 18:20:09 +0000798
Chris Lattner1c80fbf2011-11-27 20:51:47 +0000799<li>The optimizer and code generator now supports gprof and gcov-style coverage
Chris Lattner064caf92011-11-27 21:30:28 +0000800 and profiling information, and includes a new llvm-cov tool (but also works
801 with gcov). Clang exposes coverage and profiling through GCC-compatible
802 command line options.</li>
Chris Lattnerf304ffc2008-02-10 08:17:19 +0000803</ul>
Mikhail Glushenkovea65d7d2008-10-13 02:08:34 +0000804
Bill Wendlingbc5f6dd2011-10-26 18:33:01 +0000805</div>
806
Michael J. Spencer60f790c2011-11-28 18:20:09 +0000807
Chris Lattner4f0fe432011-11-27 19:26:30 +0000808<!--=========================================================================-->
809<h3>
810<a name="coreimprovements">LLVM IR and Core Improvements</a>
811</h3>
812
813<div>
814
815<p>LLVM IR has several new features for better support of new targets and that
816 expose new optimization opportunities:</p>
817
Chris Lattner064caf92011-11-27 21:30:28 +0000818 <ul>
819 <li><a href="Atomics.html">Atomic memory accesses and memory ordering</a> are
820 now directly expressible in the IR.</li>
821 <li>A new <a href="LangRef.html#int_fma">llvm.fma intrinsic</a> directly
822 represents floating point multiply accumulate operations without an
823 intermediate rounding stage.</li>
Jakub Staszak93e9d3a2011-12-06 20:56:36 +0000824 <li>A new <a href="LangRef.html#int_expect">llvm.expect intrinsic</a> allows a
Jakub Staszak44762ca2011-12-06 22:31:27 +0000825 frontend to express expected control flow (and the
826 <a href="BranchWeightMetadata.html#builtin_expect">
827 <code>__builtin_expect</code></a> from GNU C).</li>
Chris Lattner064caf92011-11-27 21:30:28 +0000828 <li>The <a href="LangRef.html#int_prefetch">llvm.prefetch intrinsic</a> now
829 takes a 4th argument that specifies whether the prefetch happens from the
830 icache or dcache.</li>
831 <li>The new <a href="LangRef.html#uwtable">uwtable function attribute</a>
832 allows a frontend to control emission of unwind tables.</li>
833 <li>The new <a href="LangRef.html#fnattrs">nonlazybind function
834 attribute</a> allow optimization of Global Offset Table (GOT) accesses.</li>
835 <li>The new <a href="LangRef.html#returns_twice">returns_twice attribute</a>
836 allows better modeling of functions like setjmp.</li>
837 <li>The <a href="LangRef.html#datalayout">target datalayout</a> string can now
838 encode the natural alignment of the target's stack for better optimization.
839 </li>
840 </ul>
Andrew Trick5aab6382011-11-06 17:59:24 +0000841</div>
842
843<!--=========================================================================-->
844<h3>
Chris Lattneracce85d2008-02-10 07:46:44 +0000845<a name="optimizer">Optimizer Improvements</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000846</h3>
Chris Lattneracce85d2008-02-10 07:46:44 +0000847
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000848<div>
Chris Lattneracce85d2008-02-10 07:46:44 +0000849
Chris Lattner064caf92011-11-27 21:30:28 +0000850<p>In addition to many minor performance tweaks and bug fixes, this
Bill Wendling7b7fa742011-10-26 18:46:16 +0000851 release includes a few major enhancements and additions to the
852 optimizers:</p>
Chris Lattneracce85d2008-02-10 07:46:44 +0000853
854<ul>
Chris Lattner064caf92011-11-27 21:30:28 +0000855<li>The pass manager now has an extension API that allows front-ends and plugins
856 to insert their own optimizations in the well-known places in the standard
857 pass optimization pipeline.</li>
Chris Lattner1c80fbf2011-11-27 20:51:47 +0000858
Benjamin Kramer933a78c2011-11-26 11:14:54 +0000859<li>Information about <a href="BranchWeightMetadata.html">branch probability</a>
860 and basic block frequency is now available within LLVM, based on a
861 combination of static branch prediction heuristics and
862 <code>__builtin_expect</code> calls. That information is currently used for
863 register spill placement and if-conversion, with additional optimizations
864 planned for future releases. The same framework is intended for eventual
865 use with profile-guided optimization.</li>
Michael J. Spencer60f790c2011-11-28 18:20:09 +0000866
Chris Lattner064caf92011-11-27 21:30:28 +0000867<li>The "-indvars" induction variable simplification pass only modifies
868 induction variables when profitable. Sign and zero extension
869 elimination, linear function test replacement, loop unrolling, and
870 other simplifications that require induction variable analysis have
Michael J. Spencer60f790c2011-11-28 18:20:09 +0000871 been generalized so they no longer require loops to be rewritten into
Chris Lattner064caf92011-11-27 21:30:28 +0000872 canonical form prior to optimization. This new design
873 preserves more IR level information, avoids undoing earlier loop
874 optimizations (particularly hand-optimized loops), and no longer
875 requires the code generator to reconstruct loops into an optimal form -
876 an intractable problem.</li>
877
878<li>LLVM now includes a pass to optimize retain/release calls for the
879 <a href="http://clang.llvm.org/docs/AutomaticReferenceCounting.html">Automatic
880 Reference Counting</a> (ARC) Objective-C language feature (in
Chris Lattner2f206022011-11-27 22:03:34 +0000881 lib/Transforms/Scalar/ObjCARC.cpp). It is a decent example of implementing
882 a source-language-specific optimization in LLVM.</li>
Chris Lattner064caf92011-11-27 21:30:28 +0000883
Chris Lattner11b66112010-10-04 02:42:39 +0000884</ul>
885
Chris Lattner61358ab2009-10-13 17:48:04 +0000886</div>
Chris Lattner917cc712009-03-02 02:37:32 +0000887
Chris Lattner0b832202008-06-08 02:45:07 +0000888<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000889<h3>
Chris Lattner4ba2b652010-09-30 16:31:33 +0000890<a name="mc">MC Level Improvements</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000891</h3>
Chris Lattner4ba2b652010-09-30 16:31:33 +0000892
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000893<div>
Bill Wendling7b7fa742011-10-26 18:46:16 +0000894
895<p>The LLVM Machine Code (aka MC) subsystem was created to solve a number of
896 problems in the realm of assembly, disassembly, object file format handling,
897 and a number of other related areas that CPU instruction-set level tools work
Chris Lattner2f206022011-11-27 22:03:34 +0000898 in. For more information, please see
899 the <a href="http://blog.llvm.org/2010/04/intro-to-llvm-mc-project.html">Intro
900 to the LLVM MC Project Blog Post</a>.</p>
Chris Lattner7d9b6b42010-10-02 21:59:30 +0000901
Chris Lattner7d9b6b42010-10-02 21:59:30 +0000902<ul>
Chris Lattner2f206022011-11-27 22:03:34 +0000903 <li>The MC layer has undergone significant refactoring to eliminate layering
904 violations that caused it to pull in the LLVM compiler backend code.</li>
905 <li>The ELF object file writers are much more full featured.</li>
906 <li>The integrated assembler now supports #line directives.</li>
907 <li>An early implementation of a JIT built on top of the MC framework (known
908 as MC-JIT) has been implemented and will eventually replace the old JIT.
909 It emits object files direct to memory and uses a runtime dynamic linker to
910 resolve references and drive lazy compilation. The MC-JIT enables much
911 greater code reuse between the JIT and the static compiler and provides
912 better integration with the platform ABI as a result.
913 </li>
914 <li>The assembly printer now makes uses of assemblers instruction aliases
915 (InstAliases) to print simplified mneumonics when possible.</li>
916 <li>TableGen can now autogenerate MC expansion logic for pseudo
917 instructions that expand to multiple MC instructions (through the
918 PseudoInstExpansion class).</li>
Michael J. Spencer60f790c2011-11-28 18:20:09 +0000919 <li>A new llvm-dwarfdump tool provides a start of a drop-in
920 replacement for the corresponding tool that use LLVM libraries. As part of
Chris Lattner7c224462011-11-27 22:39:23 +0000921 this, LLVM has the beginnings of a dwarf parsing library.</li>
Michael J. Spencer60f790c2011-11-28 18:20:09 +0000922 <li>llvm-objdump has more output including, symbol by symbol disassembly,
923 inline relocations, section headers, symbol tables, and section contents.
924 Support for archive files has also been added.</li>
925 <li>llvm-nm has gained support for archives of binary files.</li>
926 <li>llvm-size has been added. This tool prints out section sizes.</li>
Chris Lattner7d9b6b42010-10-02 21:59:30 +0000927</ul>
928
NAKAMURA Takumi45c435a2011-04-05 08:24:22 +0000929</div>
Chris Lattner4ba2b652010-09-30 16:31:33 +0000930
Chris Lattner4ba2b652010-09-30 16:31:33 +0000931<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000932<h3>
Chris Lattner511433e2009-03-02 03:24:11 +0000933<a name="codegen">Target Independent Code Generator Improvements</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000934</h3>
Chris Lattner0b832202008-06-08 02:45:07 +0000935
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000936<div>
Chris Lattner0b832202008-06-08 02:45:07 +0000937
Mikhail Glushenkovf795ef02009-03-01 18:09:47 +0000938<p>We have put a significant amount of work into the code generator
Bill Wendling7b7fa742011-10-26 18:46:16 +0000939 infrastructure, which allows us to implement more aggressive algorithms and
940 make it run faster:</p>
Chris Lattner0b832202008-06-08 02:45:07 +0000941
942<ul>
Rafael Espindolabdef6fe2011-11-29 19:08:23 +0000943<li>LLVM can now produce code that works with libgcc
944 to <a href="SegmentedStacks.html">dynamically allocate stack
945 segments</a>, as opposed to allocating a worst-case chunk of
946 virtual memory for each thread.</li>
Chris Lattner2f206022011-11-27 22:03:34 +0000947<li>LLVM generates substantially better code for indirect gotos due to a new
948 tail duplication pass, which can be a substantial performance win for
949 interpreter loops that use them.</li>
Rafael Espindola7ca7b532011-11-28 23:55:49 +0000950<li>Exception handling and debug frame information is now emitted with CFI
951 directives. This lets the assembler produce more compact info as it knows
952 the final offsets, yielding <a href="http://blog.mozilla.com/respindola/2011/05/12/cfi-directives/">much smaller executables</a> for some C++ applications.
953 If the system assembler doesn't support it, MC exands the directives when
954 the integrated assembler is not used.
Chris Lattner2f206022011-11-27 22:03:34 +0000955</li>
Chris Lattner1c80fbf2011-11-27 20:51:47 +0000956
Chris Lattner2f206022011-11-27 22:03:34 +0000957<li>The code generator now supports vector "select" operations on vector
958 comparisons, turning them into various optimized code sequences (e.g.
959 using the SSE4/AVX "blend" instructions).</li>
Jakob Stoklund Olesen09e61ca2011-11-28 01:46:19 +0000960<li>The SSE execution domain fix pass and the ARM NEON move fix pass have been
Jakob Stoklund Olesen87f95dc2011-11-28 18:03:11 +0000961 merged to a target independent execution dependency fix pass. This pass is
962 used to select alternative equivalent opcodes in a way that minimizes
963 execution domain crossings. Closely connected instructions are moved to
964 the same execution domain when possible. Targets can override the
965 <code>getExecutionDomain</code> and <code>setExecutionDomain</code> hooks
966 to use the pass.</li>
Chris Lattner84977642007-09-21 03:54:09 +0000967</ul>
Chris Lattner84977642007-09-21 03:54:09 +0000968</div>
969
Chris Lattner917cc712009-03-02 02:37:32 +0000970<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000971<h3>
Chris Lattner917cc712009-03-02 02:37:32 +0000972<a name="x86">X86-32 and X86-64 Target Improvements</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000973</h3>
Chris Lattner917cc712009-03-02 02:37:32 +0000974
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000975<div>
Bill Wendling7b7fa742011-10-26 18:46:16 +0000976
977<p>New features and major changes in the X86 target include:</p>
Chris Lattner917cc712009-03-02 02:37:32 +0000978
979<ul>
Chris Lattner2f206022011-11-27 22:03:34 +0000980<li>The X86 backend, assembler and disassembler now have full support for AVX 1.
981 To enable it pass <code>-mavx</code> to the compiler. AVX2 implementation is
982 underway on mainline.</li>
983<li>The integrated assembler and disassembler now support a broad range of new
984 instructions including Atom, Ivy Bridge, <a
985 href="http://en.wikipedia.org/wiki/SSE4a">SSE4a/BMI</a> instructions, <a
986 href="http://en.wikipedia.org/wiki/RdRand">rdrand</a> and many others.</li>
987<li>The X86 backend now fully supports the <a href="http://llvm.org/PR879">X87
988 floating point stack inline assembly constraints</a>.</li>
989<li>The integrated assembler now supports the <tt>.code32</tt> and
990 <tt>.code64</tt> directives to switch between 32-bit and 64-bit
991 instructions.</li>
992<li>The X86 backend now synthesizes horizontal add/sub instructions from generic
993 vector code when the appropriate instructions are enabled.</li>
994<li>The X86-64 backend generates smaller and faster code at -O0 due to
995 improvements in fast instruction selection.</li>
996<li><a href="http://code.google.com/p/nativeclient/">Native Client</a>
997 subtarget support has been added.</li>
Bill Wendling7b7fa742011-10-26 18:46:16 +0000998
Chris Lattner2f206022011-11-27 22:03:34 +0000999<li>The CRC32 intrinsics have been renamed. The intrinsics were previously
1000 <code>@llvm.x86.sse42.crc32.[8|16|32]</code>
1001 and <code>@llvm.x86.sse42.crc64.[8|64]</code>. They have been renamed to
1002 <code>@llvm.x86.sse42.crc32.32.[8|16|32]</code> and
1003 <code>@llvm.x86.sse42.crc32.64.[8|64]</code>.</li>
Chris Lattner917cc712009-03-02 02:37:32 +00001004</ul>
1005
Chris Lattner917cc712009-03-02 02:37:32 +00001006</div>
Chris Lattner84977642007-09-21 03:54:09 +00001007
1008<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +00001009<h3>
Chris Lattner61358ab2009-10-13 17:48:04 +00001010<a name="ARM">ARM Target Improvements</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +00001011</h3>
Chris Lattnerc441fb82009-03-01 02:30:21 +00001012
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +00001013<div>
Bill Wendling7b7fa742011-10-26 18:46:16 +00001014
1015<p>New features of the ARM target include:</p>
Chris Lattnerc441fb82009-03-01 02:30:21 +00001016
1017<ul>
Chris Lattner1cc489b2011-11-27 22:12:32 +00001018<li>The ARM backend generates much faster code for Cortex-A9 chips.</li>
1019<li>The ARM backend has improved support for Cortex-M series processors.</li>
1020<li>The ARM inline assembly constraints have been implemented and are now fully
1021 supported.</li>
1022<li>NEON code produced by Clang often runs much faster due to improvements in
1023 the Scalar Replacement of Aggregates pass.</li>
1024<li>The old ARM disassembler is replaced with a new one based on autogenerated
1025 encoding information from ARM .td files.</li>
1026<li>The integrated assembler has made major leaps forward, but is still beta quality in LLVM 3.0.</li>
Bob Wilsone8472772010-09-13 17:39:35 +00001027</ul>
Chris Lattner61358ab2009-10-13 17:48:04 +00001028</div>
Chris Lattner7b95c382011-11-15 22:23:46 +00001029
Michael J. Spencer60f790c2011-11-28 18:20:09 +00001030
Chris Lattner1efe27e2011-04-06 00:45:11 +00001031<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +00001032<h3>
Akira Hatanaka5381cbf2011-11-15 21:33:05 +00001033<a name="MIPS">MIPS Target Improvements</a>
1034</h3>
1035
1036<div>
1037
Chris Lattner1cc489b2011-11-27 22:12:32 +00001038<p>This release has seen major new work on just about every aspect of the MIPS
1039 backend. Some of the major new features include:</p>
Akira Hatanaka5381cbf2011-11-15 21:33:05 +00001040
1041<ul>
1042 <li>Most MIPS32r1 and r2 instructions are now supported.</li>
1043 <li>LE/BE MIPS32r1/r2 has been tested extensively.</li>
1044 <li>O32 ABI has been fully tested.</li>
1045 <li>MIPS backend has migrated to using the MC infrastructure for assembly printing. Initial support for direct object code emission has been implemented too.</li>
Michael J. Spencer60f790c2011-11-28 18:20:09 +00001046 <li>Delay slot filler has been updated. Now it tries to fill delay slots with useful instructions instead of always filling them with NOPs.</li>
Akira Hatanaka5381cbf2011-11-15 21:33:05 +00001047 <li>Support for old-style JIT is complete.</li>
1048 <li>Support for old architectures (MIPS1 and MIPS2) has been removed.</li>
1049 <li>Initial support for MIPS64 has been added.</li>
1050</ul>
1051</div>
Chris Lattner7b95c382011-11-15 22:23:46 +00001052
1053<!--=========================================================================-->
1054<h3>
1055 <a name="PTX">PTX Target Improvements</a>
1056</h3>
1057
1058<div>
Michael J. Spencer60f790c2011-11-28 18:20:09 +00001059
Chris Lattner7b95c382011-11-15 22:23:46 +00001060 <p>
1061 The PTX back-end is still experimental, but is fairly usable for compute kernels
1062 in LLVM 3.0. Most scalar arithmetic is implemented, as well as intrinsics to
1063 access the special PTX registers and sync instructions. The major missing
1064 pieces are texture/sampler support and some vector operations.</p>
Michael J. Spencer60f790c2011-11-28 18:20:09 +00001065
Chris Lattner7b95c382011-11-15 22:23:46 +00001066 <p>That said, the backend is already being used for domain-specific languages
Peter Collingbournee6e73622011-11-29 02:04:48 +00001067 and can be used by Clang to
1068 <a href="http://clang.llvm.org/docs/ReleaseNotes.html#opencl">compile OpenCL
1069 C code</a> into PTX.</p>
Michael J. Spencer60f790c2011-11-28 18:20:09 +00001070
Chris Lattner7b95c382011-11-15 22:23:46 +00001071</div>
1072
Akira Hatanaka5381cbf2011-11-15 21:33:05 +00001073<!--=========================================================================-->
1074<h3>
Chris Lattner1efe27e2011-04-06 00:45:11 +00001075<a name="OtherTS">Other Target Specific Improvements</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +00001076</h3>
Chris Lattnerc441fb82009-03-01 02:30:21 +00001077
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +00001078<div>
Bill Wendling7b7fa742011-10-26 18:46:16 +00001079
Chris Lattner1efe27e2011-04-06 00:45:11 +00001080<ul>
Chris Lattnerd6cc2c22011-11-27 22:36:22 +00001081<li>Many PowerPC improvements have been implemented for ELF targets, including
1082 support for varargs and initial support for direct .o file emission.</li>
Michael J. Spencer60f790c2011-11-28 18:20:09 +00001083
Chris Lattnerd6cc2c22011-11-27 22:36:22 +00001084<li>MicroBlaze scheduling itineraries were added that model the
1085 3-stage and the 5-stage pipeline architectures. The 3-stage
Michael J. Spencer60f790c2011-11-28 18:20:09 +00001086 pipeline model can be selected with <code>-mcpu=mblaze3</code>
1087 and the 5-stage pipeline model can be selected with
Chris Lattnerd6cc2c22011-11-27 22:36:22 +00001088 <code>-mcpu=mblaze5</code>.</li>
1089
Chris Lattner1efe27e2011-04-06 00:45:11 +00001090</ul>
Bill Wendling7b7fa742011-10-26 18:46:16 +00001091
Chris Lattner1efe27e2011-04-06 00:45:11 +00001092</div>
Chris Lattner77d29b12008-06-05 08:02:49 +00001093
1094<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +00001095<h3>
Chris Lattnerf6662f92008-10-13 17:57:36 +00001096<a name="changes">Major Changes and Removed Features</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +00001097</h3>
Chris Lattnerf6662f92008-10-13 17:57:36 +00001098
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +00001099<div>
Chris Lattnerf6662f92008-10-13 17:57:36 +00001100
Bill Wendling2626dba2011-08-03 22:18:20 +00001101<p>If you're already an LLVM user or developer with out-of-tree changes based on
1102 LLVM 2.9, this section lists some "gotchas" that you may run into upgrading
1103 from the previous release.</p>
Chris Lattnerf6662f92008-10-13 17:57:36 +00001104
1105<ul>
Chris Lattnerd6cc2c22011-11-27 22:36:22 +00001106<li>LLVM 3.0 removes support for reading LLVM 2.8 and earlier files, and LLVM
1107 3.1 will eliminate support for reading LLVM 2.9 files. Going forward, we
Michael J. Spencer60f790c2011-11-28 18:20:09 +00001108 aim for all future versions of LLVM to read bitcode files and .ll files
Chris Lattnerd6cc2c22011-11-27 22:36:22 +00001109 produced by LLVM 3.0.</li>
1110<li>Tablegen has been split into a library, allowing the clang tblgen pieces
Peter Collingbournef0a66052011-11-29 02:04:44 +00001111 to now live in the clang tree. The llvm version has been renamed to
Chris Lattnerd6cc2c22011-11-27 22:36:22 +00001112 llvm-tblgen instead of tblgen.</li>
Chris Lattner5a1731d2011-11-27 08:32:32 +00001113 <li>The <code>LLVMC</code> meta compiler driver was removed.</li>
Chris Lattnerd6cc2c22011-11-27 22:36:22 +00001114 <li>The unused PostOrder Dominator Frontiers and LowerSetJmp passes were removed.</li>
1115
1116
Rafael Espindolaf940a1a2011-08-30 23:03:45 +00001117 <li>The old <code>TailDup</code> pass was not used in the standard pipeline
1118 and was unable to update ssa form, so it has been removed.
Eli Friedmanf03bb262011-08-12 22:50:01 +00001119 <li>The syntax of volatile loads and stores in IR has been changed to
1120 "<code>load volatile</code>"/"<code>store volatile</code>". The old
1121 syntax ("<code>volatile load</code>"/"<code>volatile store</code>")
Chris Lattnerd6cc2c22011-11-27 22:36:22 +00001122 is still accepted, but is now considered deprecated and will be removed in
1123 3.1.</li>
1124 <li>llvm-gcc's frontend tests have been removed from llvm/test/Frontend*, sunk
1125 into the clang and dragonegg testsuites.</li>
Benjamin Kramer7c5025b2011-11-25 21:26:00 +00001126 <li>The old atomic intrinsics (<code>llvm.memory.barrier</code> and
Eli Friedman526e1bb2011-10-26 00:55:23 +00001127 <code>llvm.atomic.*</code>) are now gone. Please use the new atomic
1128 instructions, described in the <a href="Atomics.html">atomics guide</a>.
Chris Lattnerd6cc2c22011-11-27 22:36:22 +00001129 <li>LLVM's configure script doesn't depend on llvm-gcc anymore, eliminating a
1130 strange circular dependence between projects.</li>
Devang Patelb34dd132008-10-14 20:03:43 +00001131</ul>
Chris Lattnerf6662f92008-10-13 17:57:36 +00001132
NAKAMURA Takumi2026de22011-08-22 23:22:05 +00001133<h4>Windows (32-bit)</h4>
1134<div>
Bill Wendling7b7fa742011-10-26 18:46:16 +00001135
NAKAMURA Takumi2026de22011-08-22 23:22:05 +00001136<ul>
1137 <li>On Win32(MinGW32 and MSVC), Windows 2000 will not be supported.
1138 Windows XP or higher is required.</li>
1139</ul>
Bill Wendling7b7fa742011-10-26 18:46:16 +00001140
NAKAMURA Takumi2026de22011-08-22 23:22:05 +00001141</div>
1142
Chris Lattnerf6662f92008-10-13 17:57:36 +00001143</div>
1144
Daniel Dunbarf0233c62010-10-04 20:11:41 +00001145<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +00001146<h3>
Chris Lattner1efe27e2011-04-06 00:45:11 +00001147<a name="api_changes">Internal API Changes</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +00001148</h3>
Daniel Dunbarf0233c62010-10-04 20:11:41 +00001149
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +00001150<div>
Daniel Dunbarf0233c62010-10-04 20:11:41 +00001151
Chris Lattner1efe27e2011-04-06 00:45:11 +00001152<p>In addition, many APIs have changed in this release. Some of the major
Bill Wendling16005252011-08-02 06:20:17 +00001153 LLVM API changes are:</p>
Daniel Dunbarf0233c62010-10-04 20:11:41 +00001154
1155<ul>
Joe Abbey54d3b832011-11-28 22:07:12 +00001156 <li>The biggest and most pervasive change is that the type system has been
1157 rewritten: <code>PATypeHolder</code> and <code>OpaqueType</code> are gone,
1158 and all APIs deal with <code>Type*</code> instead of <code>const
1159 Type*</code>. If you need to create recursive structures, then create a
1160 named structure, and use <code>setBody()</code> when all its elements are
1161 built. Type merging and refining is gone too: named structures are not
1162 merged with other structures, even if their layout is identical. (of
1163 course anonymous structures are still uniqued by layout).</li>
Michael J. Spencer60f790c2011-11-28 18:20:09 +00001164
Bill Wendling7b7fa742011-10-26 18:46:16 +00001165 <li><code>PHINode::reserveOperandSpace</code> has been removed. Instead, you
1166 must specify how many operands to reserve space for when you create the
1167 PHINode, by passing an extra argument
1168 into <code>PHINode::Create</code>.</li>
Jay Foad558d3762011-07-14 09:19:05 +00001169
Bill Wendling7b7fa742011-10-26 18:46:16 +00001170 <li>PHINodes no longer store their incoming BasicBlocks as operands. Instead,
1171 the list of incoming BasicBlocks is stored separately, and can be accessed
1172 with new functions <code>PHINode::block_begin</code>
1173 and <code>PHINode::block_end</code>.</li>
Jay Foad558d3762011-07-14 09:19:05 +00001174
Bill Wendling7b7fa742011-10-26 18:46:16 +00001175 <li>Various functions now take an <code>ArrayRef</code> instead of either a
1176 pair of pointers (or iterators) to the beginning and end of a range, or a
1177 pointer and a length. Others now return an <code>ArrayRef</code> instead
1178 of a reference to a <code>SmallVector</code>
1179 or <code>std::vector</code>. These include:
Jay Foad558d3762011-07-14 09:19:05 +00001180<ul>
1181<!-- Please keep this list sorted. -->
Jay Foada3efbb12011-07-15 08:37:34 +00001182<li><code>CallInst::Create</code></li>
Jay Foad558d3762011-07-14 09:19:05 +00001183<li><code>ComputeLinearIndex</code> (in <code>llvm/CodeGen/Analysis.h</code>)</li>
1184<li><code>ConstantArray::get</code></li>
1185<li><code>ConstantExpr::getExtractElement</code></li>
Jay Foaddab3d292011-07-21 14:31:17 +00001186<li><code>ConstantExpr::getGetElementPtr</code></li>
1187<li><code>ConstantExpr::getInBoundsGetElementPtr</code></li>
Jay Foad558d3762011-07-14 09:19:05 +00001188<li><code>ConstantExpr::getIndices</code></li>
1189<li><code>ConstantExpr::getInsertElement</code></li>
1190<li><code>ConstantExpr::getWithOperands</code></li>
Jay Foad1d2f5692011-07-19 13:32:40 +00001191<li><code>ConstantFoldCall</code> (in <code>llvm/Analysis/ConstantFolding.h</code>)</li>
1192<li><code>ConstantFoldInstOperands</code> (in <code>llvm/Analysis/ConstantFolding.h</code>)</li>
Jay Foad558d3762011-07-14 09:19:05 +00001193<li><code>ConstantVector::get</code></li>
1194<li><code>DIBuilder::createComplexVariable</code></li>
1195<li><code>DIBuilder::getOrCreateArray</code></li>
1196<li><code>ExtractValueInst::Create</code></li>
1197<li><code>ExtractValueInst::getIndexedType</code></li>
1198<li><code>ExtractValueInst::getIndices</code></li>
1199<li><code>FindInsertedValue</code> (in <code>llvm/Analysis/ValueTracking.h</code>)</li>
Jay Foadca12a212011-07-19 14:42:50 +00001200<li><code>gep_type_begin</code> (in <code>llvm/Support/GetElementPtrTypeIterator.h</code>)</li>
1201<li><code>gep_type_end</code> (in <code>llvm/Support/GetElementPtrTypeIterator.h</code>)</li>
Jay Foada9203102011-07-25 09:48:08 +00001202<li><code>GetElementPtrInst::Create</code></li>
1203<li><code>GetElementPtrInst::CreateInBounds</code></li>
1204<li><code>GetElementPtrInst::getIndexedType</code></li>
Jay Foadb60e8512011-07-21 14:42:51 +00001205<li><code>InsertValueInst::Create</code></li>
1206<li><code>InsertValueInst::getIndices</code></li>
1207<li><code>InvokeInst::Create</code></li>
Jay Foada3efbb12011-07-15 08:37:34 +00001208<li><code>IRBuilder::CreateCall</code></li>
Jay Foad558d3762011-07-14 09:19:05 +00001209<li><code>IRBuilder::CreateExtractValue</code></li>
Jay Foad0a2a60a2011-07-22 08:16:57 +00001210<li><code>IRBuilder::CreateGEP</code></li>
1211<li><code>IRBuilder::CreateInBoundsGEP</code></li>
Jay Foad558d3762011-07-14 09:19:05 +00001212<li><code>IRBuilder::CreateInsertValue</code></li>
Jay Foada3efbb12011-07-15 08:37:34 +00001213<li><code>IRBuilder::CreateInvoke</code></li>
Jay Foad558d3762011-07-14 09:19:05 +00001214<li><code>MDNode::get</code></li>
1215<li><code>MDNode::getIfExists</code></li>
1216<li><code>MDNode::getTemporary</code></li>
1217<li><code>MDNode::getWhenValsUnresolved</code></li>
Jay Foadb9b54eb2011-07-19 15:07:52 +00001218<li><code>SimplifyGEPInst</code> (in <code>llvm/Analysis/InstructionSimplify.h</code>)</li>
Jay Foad8fbbb392011-07-19 14:01:37 +00001219<li><code>TargetData::getIndexedOffset</code></li>
Jay Foad558d3762011-07-14 09:19:05 +00001220</ul></li>
1221
Bill Wendling7b7fa742011-10-26 18:46:16 +00001222 <li>All forms of <code>StringMap::getOrCreateValue</code> have been remove
1223 except for the one which takes a <code>StringRef</code>.</li>
Jay Foad558d3762011-07-14 09:19:05 +00001224
Bill Wendling7b7fa742011-10-26 18:46:16 +00001225 <li>The <code>LLVMBuildUnwind</code> function from the C API was removed. The
1226 LLVM <code>unwind</code> instruction has been deprecated for a long time
1227 and isn't used by the current front-ends. So this was removed during the
1228 exception handling rewrite.</li>
Bill Wendling16005252011-08-02 06:20:17 +00001229
Bill Wendling7b7fa742011-10-26 18:46:16 +00001230 <li>The <code>LLVMAddLowerSetJmpPass</code> function from the C API was
1231 removed because the <code>LowerSetJmp</code> pass was removed.</li>
Bill Wendling2626dba2011-08-03 22:18:20 +00001232
Bill Wendling7b7fa742011-10-26 18:46:16 +00001233 <li>The <code>DIBuilder</code> interface used by front ends to encode
1234 debugging information in the LLVM IR now expects clients to
1235 use <code>DIBuilder::finalize()</code> at the end of translation unit to
1236 complete debugging information encoding.</li>
Devang Patel6326a422011-08-15 23:00:00 +00001237
Bill Wendling7b7fa742011-10-26 18:46:16 +00001238 <li>TargetSelect.h moved to Support/ from Target/</li>
Torok Edwinf16e2d42011-09-30 13:07:52 +00001239
Bill Wendling7b7fa742011-10-26 18:46:16 +00001240 <li>UpgradeIntrinsicCall no longer upgrades pre-2.9 intrinsic calls (for
1241 example <code>llvm.memset.i32</code>).</li>
Torok Edwinf16e2d42011-09-30 13:07:52 +00001242
Bill Wendling7b7fa742011-10-26 18:46:16 +00001243 <li>It is mandatory to initialize all out-of-tree passes too and their dependencies now with
1244 <code>INITIALIZE_PASS{BEGIN,END,}</code>
1245 and <code>INITIALIZE_{PASS,AG}_DEPENDENCY</code>.</li>
Torok Edwinf16e2d42011-09-30 13:07:52 +00001246
Bill Wendling7b7fa742011-10-26 18:46:16 +00001247 <li>The interface for MemDepResult in MemoryDependenceAnalysis has been
1248 enhanced with new return types Unknown and NonFuncLocal, in addition to
1249 the existing types Clobber, Def, and NonLocal.</li>
Daniel Dunbarf0233c62010-10-04 20:11:41 +00001250</ul>
Bill Wendling7b7fa742011-10-26 18:46:16 +00001251
Daniel Dunbarf0233c62010-10-04 20:11:41 +00001252</div>
Chris Lattnerf6662f92008-10-13 17:57:36 +00001253
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +00001254</div>
1255
Chris Lattner79c3fe12003-10-02 04:57:28 +00001256<!-- *********************************************************************** -->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +00001257<h2>
Misha Brukman500bc302003-11-22 00:38:41 +00001258 <a name="knownproblems">Known Problems</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +00001259</h2>
Chris Lattner79c3fe12003-10-02 04:57:28 +00001260<!-- *********************************************************************** -->
1261
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +00001262<div>
Misha Brukman500bc302003-11-22 00:38:41 +00001263
Chris Lattner70e22012011-11-27 19:38:20 +00001264<p>LLVM is generally a production quality compiler, and is used by a broad range
1265 of applications and shipping in many products. That said, not every
1266 subsystem is as mature as the aggregate, particularly the more obscure
1267 targets. If you run into a problem, please check the <a
1268 href="http://llvm.org/bugs/">LLVM bug database</a> and submit a bug if
1269 there isn't already one or ask on the <a
1270 href="http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev">LLVMdev
1271 list</a>.</p>
Michael J. Spencer60f790c2011-11-28 18:20:09 +00001272
Chris Lattner70e22012011-11-27 19:38:20 +00001273 <p>Known problem areas include:</p>
Michael J. Spencer60f790c2011-11-28 18:20:09 +00001274
Chris Lattnerf5ee1702004-03-14 02:03:02 +00001275<ul>
Chris Lattner70e22012011-11-27 19:38:20 +00001276 <li>The Alpha, Blackfin, CellSPU, MSP430, PTX, SystemZ and
Chris Lattner1c80fbf2011-11-27 20:51:47 +00001277 XCore backends are experimental, and the Alpha, Blackfin and SystemZ
1278 targets have already been removed from mainline.</li>
Bill Wendling7b7fa742011-10-26 18:46:16 +00001279
Chris Lattner70e22012011-11-27 19:38:20 +00001280 <li>The integrated assembler, disassembler, and JIT is not supported by
1281 several targets. If an integrated assembler is not supported, then a
1282 system assembler is required. For more details, see the <a
1283 href="CodeGenerator.html#targetfeatures">Target Features Matrix</a>.
1284 </li>
Michael J. Spencer60f790c2011-11-28 18:20:09 +00001285
Chris Lattner70e22012011-11-27 19:38:20 +00001286 <li>The C backend has numerous problems and is not being actively maintained.
1287 Depending on it for anything serious is not advised.</li>
Chris Lattner26299222006-11-18 07:51:14 +00001288</ul>
1289
1290</div>
John Criswellc0c186d2005-11-08 21:11:33 +00001291
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +00001292</div>
1293
Chris Lattner79c3fe12003-10-02 04:57:28 +00001294<!-- *********************************************************************** -->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +00001295<h2>
Misha Brukman500bc302003-11-22 00:38:41 +00001296 <a name="additionalinfo">Additional Information</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +00001297</h2>
Chris Lattner79c3fe12003-10-02 04:57:28 +00001298<!-- *********************************************************************** -->
1299
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +00001300<div>
Misha Brukman500bc302003-11-22 00:38:41 +00001301
Bill Wendling7b7fa742011-10-26 18:46:16 +00001302<p>A wide variety of additional information is available on
1303 the <a href="http://llvm.org/">LLVM web page</a>, in particular in
1304 the <a href="http://llvm.org/docs/">documentation</a> section. The web page
1305 also contains versions of the API documentation which is up-to-date with the
1306 Subversion version of the source code. You can access versions of these
1307 documents specific to this release by going into the "<tt>llvm/doc/</tt>"
1308 directory in the LLVM tree.</p>
Chris Lattner79c3fe12003-10-02 04:57:28 +00001309
Misha Brukman500bc302003-11-22 00:38:41 +00001310<p>If you have any questions or comments about LLVM, please feel free to contact
Bill Wendling7b7fa742011-10-26 18:46:16 +00001311 us via the <a href="http://llvm.org/docs/#maillist"> mailing lists</a>.</p>
Chris Lattner79c3fe12003-10-02 04:57:28 +00001312
Misha Brukman500bc302003-11-22 00:38:41 +00001313</div>
Chris Lattner79c3fe12003-10-02 04:57:28 +00001314
Chris Lattner1c80fbf2011-11-27 20:51:47 +00001315<!--=========================================================================-->
1316
1317<!-- EH details: to be moved to a blog post:
Chris Lattner1c80fbf2011-11-27 20:51:47 +00001318
Michael J. Spencer60f790c2011-11-28 18:20:09 +00001319
1320
1321
Chris Lattner1c80fbf2011-11-27 20:51:47 +00001322<p>One of the biggest changes is that 3.0 has a new exception handling
1323 system. The old system used LLVM intrinsics to convey the exception handling
1324 information to the code generator. It worked in most cases, but not
1325 all. Inlining was especially difficult to get right. Also, the intrinsics
1326 could be moved away from the <code>invoke</code> instruction, making it hard
1327 to recover that information.</p>
1328
1329<p>The new EH system makes exception handling a first-class member of the IR. It
1330 adds two new instructions:</p>
1331
1332<ul>
1333 <li><a href="LangRef.html#i_landingpad"><code>landingpad</code></a> &mdash;
1334 this instruction defines a landing pad basic block. It contains all of the
1335 information that's needed by the code generator. It's also required to be
1336 the first non-PHI instruction in the landing pad. In addition, a landing
1337 pad may be jumped to only by the unwind edge of an <code>invoke</code>
1338 instruction.</li>
1339
1340 <li><a href="LangRef.html#i_resume"><code>resume</code></a> &mdash; this
1341 instruction causes the current exception to resume traveling up the
1342 stack. It replaces the <code>@llvm.eh.resume</code> intrinsic.</li>
1343</ul>
1344
1345<p>Converting from the old EH API to the new EH API is rather simple, because a
1346 lot of complexity has been removed. The two intrinsics,
1347 <code>@llvm.eh.exception</code> and <code>@llvm.eh.selector</code> have been
1348 superseded by the <code>landingpad</code> instruction. Instead of generating
1349 a call to <code>@llvm.eh.exception</code> and <code>@llvm.eh.selector</code>:
1350
1351<div class="doc_code">
1352<pre>
1353Function *ExcIntr = Intrinsic::getDeclaration(TheModule,
1354 Intrinsic::eh_exception);
1355Function *SlctrIntr = Intrinsic::getDeclaration(TheModule,
1356 Intrinsic::eh_selector);
1357
1358// The exception pointer.
1359Value *ExnPtr = Builder.CreateCall(ExcIntr, "exc_ptr");
1360
1361std::vector&lt;Value*&gt; Args;
1362Args.push_back(ExnPtr);
1363Args.push_back(Builder.CreateBitCast(Personality,
1364 Type::getInt8PtrTy(Context)));
1365
1366<i>// Add selector clauses to Args.</i>
1367
1368// The selector call.
1369Builder.CreateCall(SlctrIntr, Args, "exc_sel");
1370</pre>
1371</div>
1372
1373<p>You should instead generate a <code>landingpad</code> instruction, that
1374 returns an exception object and selector value:</p>
1375
1376<div class="doc_code">
1377<pre>
1378LandingPadInst *LPadInst =
1379 Builder.CreateLandingPad(StructType::get(Int8PtrTy, Int32Ty, NULL),
1380 Personality, 0);
1381
1382Value *LPadExn = Builder.CreateExtractValue(LPadInst, 0);
1383Builder.CreateStore(LPadExn, getExceptionSlot());
1384
1385Value *LPadSel = Builder.CreateExtractValue(LPadInst, 1);
1386Builder.CreateStore(LPadSel, getEHSelectorSlot());
1387</pre>
1388</div>
1389
1390<p>It's now trivial to add the individual clauses to the <code>landingpad</code>
1391 instruction.</p>
1392
1393<div class="doc_code">
1394<pre>
1395<i><b>// Adding a catch clause</b></i>
1396Constant *TypeInfo = getTypeInfo();
1397LPadInst-&gt;addClause(TypeInfo);
1398
1399<i><b>// Adding a C++ catch-all</b></i>
1400LPadInst-&gt;addClause(Constant::getNullValue(Builder.getInt8PtrTy()));
1401
1402<i><b>// Adding a cleanup</b></i>
1403LPadInst-&gt;setCleanup(true);
1404
1405<i><b>// Adding a filter clause</b></i>
1406std::vector&lt;Constant*&gt; TypeInfos;
1407Constant *TypeInfo = getFilterTypeInfo();
1408TypeInfos.push_back(Builder.CreateBitCast(TypeInfo, Builder.getInt8PtrTy()));
1409
1410ArrayType *FilterTy = ArrayType::get(Int8PtrTy, TypeInfos.size());
1411LPadInst-&gt;addClause(ConstantArray::get(FilterTy, TypeInfos));
1412</pre>
1413</div>
1414
1415<p>Converting from using the <code>@llvm.eh.resume</code> intrinsic to
1416 the <code>resume</code> instruction is trivial. It takes the exception
1417 pointer and exception selector values returned by
1418 the <code>landingpad</code> instruction:</p>
1419
1420<div class="doc_code">
1421<pre>
1422Type *UnwindDataTy = StructType::get(Builder.getInt8PtrTy(),
1423 Builder.getInt32Ty(), NULL);
1424Value *UnwindData = UndefValue::get(UnwindDataTy);
1425Value *ExcPtr = Builder.CreateLoad(getExceptionObjSlot());
1426Value *ExcSel = Builder.CreateLoad(getExceptionSelSlot());
1427UnwindData = Builder.CreateInsertValue(UnwindData, ExcPtr, 0, "exc_ptr");
1428UnwindData = Builder.CreateInsertValue(UnwindData, ExcSel, 1, "exc_sel");
1429Builder.CreateResume(UnwindData);
1430</pre>
1431</div>
1432
1433
Michael J. Spencer60f790c2011-11-28 18:20:09 +00001434
1435
Chris Lattner1c80fbf2011-11-27 20:51:47 +00001436 -->
1437
Michael J. Spencer60f790c2011-11-28 18:20:09 +00001438
Chris Lattner79c3fe12003-10-02 04:57:28 +00001439<!-- *********************************************************************** -->
Chris Lattner79c3fe12003-10-02 04:57:28 +00001440
Misha Brukman500bc302003-11-22 00:38:41 +00001441<hr>
Misha Brukman2061e892003-11-22 01:23:39 +00001442<address>
Misha Brukman38847d52003-12-21 22:53:21 +00001443 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
Misha Brukman44408702008-12-11 17:34:48 +00001444 src="http://jigsaw.w3.org/css-validator/images/vcss-blue" alt="Valid CSS"></a>
Misha Brukman2061e892003-11-22 01:23:39 +00001445 <a href="http://validator.w3.org/check/referer"><img
Misha Brukman44408702008-12-11 17:34:48 +00001446 src="http://www.w3.org/Icons/valid-html401-blue" alt="Valid HTML 4.01"></a>
Misha Brukman2061e892003-11-22 01:23:39 +00001447
Chris Lattnerb4b0ce72007-05-18 00:44:29 +00001448 <a href="http://llvm.org/">LLVM Compiler Infrastructure</a><br>
Misha Brukman500bc302003-11-22 00:38:41 +00001449 Last modified: $Date$
Misha Brukman2061e892003-11-22 01:23:39 +00001450</address>
Chris Lattner79c3fe12003-10-02 04:57:28 +00001451
Misha Brukman500bc302003-11-22 00:38:41 +00001452</body>
1453</html>