blob: 333d82985d15aad25997be93b284963db145acad [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
48 major improvements from the previous release and significant known problems.
49 All LLVM releases may be downloaded from
50 the <a href="http://llvm.org/releases/">LLVM releases web site</a>.</p>
Chris Lattner19092612003-10-02 16:38:05 +000051
Chris Lattner7506b1d2004-12-07 08:04:13 +000052<p>For more information about LLVM, including information about the latest
Bill Wendling7b7fa742011-10-26 18:46:16 +000053 release, please check out the <a href="http://llvm.org/">main LLVM web
54 site</a>. If you have questions or comments,
55 the <a href="http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev">LLVM
56 Developer's Mailing List</a> is a good place to send them.</p>
Chris Lattner79c3fe12003-10-02 04:57:28 +000057
Bill Wendling7b7fa742011-10-26 18:46:16 +000058<p>Note that if you are reading this file from a Subversion checkout or the main
59 LLVM web page, this document applies to the <i>next</i> release, not the
60 current one. To see the release notes for a specific release, please see the
61 <a href="http://llvm.org/releases/">releases page</a>.</p>
Misha Brukman500bc302003-11-22 00:38:41 +000062
63</div>
Chris Lattner61358ab2009-10-13 17:48:04 +000064
Chris Lattnere4dc1962011-04-05 23:22:33 +000065<!-- Features that need text if they're finished for 3.1:
66 ARM EHABI
Chris Lattner914ce462010-04-22 06:28:20 +000067 combiner-aa?
Chris Lattner61358ab2009-10-13 17:48:04 +000068 strong phi elim
Chris Lattner61358ab2009-10-13 17:48:04 +000069 loop dependence analysis
Chris Lattnere07043c2010-09-29 05:30:03 +000070 CorrelatedValuePropagation
Chris Lattnere4dc1962011-04-05 23:22:33 +000071 lib/Transforms/IPO/MergeFunctions.cpp => consider for 3.1.
Chris Lattneracce85d2008-02-10 07:46:44 +000072 -->
Chris Lattnere07043c2010-09-29 05:30:03 +000073
Chris Lattner96a445e2008-10-13 18:01:01 +000074<!-- *********************************************************************** -->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +000075<h2>
Chris Lattner96a445e2008-10-13 18:01:01 +000076 <a name="subproj">Sub-project Status Update</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +000077</h2>
Chris Lattner96a445e2008-10-13 18:01:01 +000078<!-- *********************************************************************** -->
Chris Lattnerea34f642008-06-08 21:34:41 +000079
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +000080<div>
Bill Wendling7b7fa742011-10-26 18:46:16 +000081
82<p>The LLVM 3.0 distribution currently consists of code from the core LLVM
83 repository (which roughly includes the LLVM optimizers, code generators and
Chris Lattnerc343e312011-11-10 20:15:40 +000084 supporting tools), and the Clang repository. In
Bill Wendling7b7fa742011-10-26 18:46:16 +000085 addition to this code, the LLVM Project includes other sub-projects that are
86 in development. Here we include updates on these subprojects.</p>
Chris Lattner96a445e2008-10-13 18:01:01 +000087
Chris Lattner8348b472008-06-08 21:58:17 +000088<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +000089<h3>
Chris Lattnerfb97b2d2008-10-13 18:11:54 +000090<a name="clang">Clang: C/C++/Objective-C Frontend Toolkit</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +000091</h3>
Chris Lattner8348b472008-06-08 21:58:17 +000092
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +000093<div>
Chris Lattner8348b472008-06-08 21:58:17 +000094
Chris Lattner095539f2010-04-26 17:42:18 +000095<p><a href="http://clang.llvm.org/">Clang</a> is an LLVM front end for the C,
Bill Wendling7b7fa742011-10-26 18:46:16 +000096 C++, and Objective-C languages. Clang aims to provide a better user
97 experience through expressive diagnostics, a high level of conformance to
98 language standards, fast compilation, and low memory use. Like LLVM, Clang
99 provides a modular, library-based architecture that makes it suitable for
100 creating or integrating with other development tools. Clang is considered a
101 production-quality compiler for C, Objective-C, C++ and Objective-C++ on x86
102 (32- and 64-bit), and for darwin/arm targets.</p>
Chris Lattner61358ab2009-10-13 17:48:04 +0000103
Chad Rosiere6291d02011-05-27 22:50:46 +0000104<p>In the LLVM 3.0 time-frame, the Clang team has made many improvements:</p>
Douglas Gregorba087df2011-10-15 00:48:01 +0000105
106<ul>
Bill Wendling7b7fa742011-10-26 18:46:16 +0000107 <li>Greatly improved support for building C++ applications, with greater
108 stability and better diagnostics.</li>
Douglas Gregorba087df2011-10-15 00:48:01 +0000109
Bill Wendling7b7fa742011-10-26 18:46:16 +0000110 <li><a href="http://clang.llvm.org/cxx_status.html">Improved support</a> for
111 the <a href="http://www.iso.org/iso/iso_catalogue/catalogue_tc/catalogue_detail.htm?csnumber=50372">C++
112 2011</a> standard, including implementations of non-static data member
113 initializers, alias templates, delegating constructors, the range-based
114 for loop, and implicitly-generated move constructors and move assignment
115 operators, among others.</li>
Douglas Gregorba087df2011-10-15 00:48:01 +0000116
Bill Wendling7b7fa742011-10-26 18:46:16 +0000117 <li>Implemented support for some features of the upcoming C1x standard,
118 including static assertions and generic selections.</li>
Douglas Gregorba087df2011-10-15 00:48:01 +0000119
Bill Wendling7b7fa742011-10-26 18:46:16 +0000120 <li>Better detection of include and linking paths for system headers and
121 libraries, especially for Linux distributions.</li>
Douglas Gregorba087df2011-10-15 00:48:01 +0000122
Bill Wendling7b7fa742011-10-26 18:46:16 +0000123 <li>Implemented support
124 for <a href="http://clang.llvm.org/docs/AutomaticReferenceCounting.html">Automatic
125 Reference Counting</a> for Objective-C.</li>
Douglas Gregorba087df2011-10-15 00:48:01 +0000126
Bill Wendling7b7fa742011-10-26 18:46:16 +0000127 <li>Implemented a number of optimizations in <tt>libclang</tt>, the Clang C
128 interface, to improve the performance of code completion and the mapping
129 from source locations to abstract syntax tree nodes.</li>
Douglas Gregorba087df2011-10-15 00:48:01 +0000130</ul>
131
Chris Lattner0a6f6d52011-04-05 07:19:28 +0000132
Duncan Sandsf3ba7af2011-04-06 08:07:40 +0000133<p>If Clang rejects your code but another compiler accepts it, please take a
Bill Wendling7b7fa742011-10-26 18:46:16 +0000134 look at the <a href="http://clang.llvm.org/compatibility.html">language
135 compatibility</a> guide to make sure this is not intentional or a known
136 issue.</p>
Bill Wendling741748a2008-10-27 09:27:33 +0000137
Chris Lattnerfb97b2d2008-10-13 18:11:54 +0000138</div>
139
140<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000141<h3>
Duncan Sands528a5102011-04-04 11:09:08 +0000142<a name="dragonegg">DragonEgg: GCC front-ends, LLVM back-end</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000143</h3>
Chris Lattner61358ab2009-10-13 17:48:04 +0000144
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000145<div>
Bill Wendling7b7fa742011-10-26 18:46:16 +0000146<p><a href="http://dragonegg.llvm.org/">DragonEgg</a> is a
147 <a href="http://gcc.gnu.org/wiki/plugins">gcc plugin</a> that replaces GCC's
Duncan Sands77352c92011-11-10 18:44:29 +0000148 optimizers and code generators with LLVM's. It works with gcc-4.5 or gcc-4.6,
149 targets the x86-32 and x86-64 processor families, and has been successfully
150 used on the Darwin, FreeBSD, KFreeBSD, Linux and OpenBSD platforms. It fully
151 supports Ada, C, C++ and Fortran. It has partial support for Go, Java, Obj-C
152 and Obj-C++.</p>
Duncan Sands749fd832010-04-02 09:23:15 +0000153
Bill Wendling7b7fa742011-10-26 18:46:16 +0000154<p>The 3.0 release has the following notable changes:</p>
155
Duncan Sands77352c92011-11-10 18:44:29 +0000156 <li>GCC version 4.6 is now fully supported.</li>
157
158 <li>Patching and building GCC is no longer required: the plugin should work
159 with your system GCC (version 4.5 or 4.6; on Debian/Ubuntu systems the
160 gcc-4.5-plugin-dev or gcc-4.6-plugin-dev package is also needed).</li>
161
162 <li>The <tt>-fplugin-arg-dragonegg-enable-gcc-optzns</tt> option, which runs
163 GCC's optimizers as well as LLVM's, now works much better. This is the
164 option to use if you want ultimate performance! It not yet completely
165 stable: it may cause the plugin to crash.</li>
166
167 <li>The type and constant conversion logic has been almost entirely rewritten,
168 fixing a multitude of obscure bugs.</li>
169
Duncan Sands4b1da2b2010-09-30 17:37:34 +0000170<ul>
Chad Rosiere6291d02011-05-27 22:50:46 +0000171<!--
172<li></li>
173-->
Duncan Sands4b1da2b2010-09-30 17:37:34 +0000174</ul>
Chris Lattner61358ab2009-10-13 17:48:04 +0000175
176</div>
177
Chris Lattner75547712010-10-03 23:49:06 +0000178<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000179<h3>
Chris Lattner75547712010-10-03 23:49:06 +0000180<a name="compiler-rt">compiler-rt: Compiler Runtime Library</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000181</h3>
Chris Lattner75547712010-10-03 23:49:06 +0000182
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000183<div>
Bill Wendling7b7fa742011-10-26 18:46:16 +0000184
185<p>The new LLVM <a href="http://compiler-rt.llvm.org/">compiler-rt project</a>
186 is a simple library that provides an implementation of the low-level
187 target-specific hooks required by code generation and other runtime
188 components. For example, when compiling for a 32-bit target, converting a
189 double to a 64-bit unsigned integer is compiled into a runtime call to the
190 "__fixunsdfdi" function. The compiler-rt library provides highly optimized
191 implementations of this and other low-level routines (some are 3x faster than
192 the equivalent libgcc routines).</p>
Chris Lattner75547712010-10-03 23:49:06 +0000193
Chad Rosiere6291d02011-05-27 22:50:46 +0000194<p>In the LLVM 3.0 timeframe,</p>
Chris Lattner75547712010-10-03 23:49:06 +0000195
196</div>
Chris Lattner61358ab2009-10-13 17:48:04 +0000197
198<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000199<h3>
Chris Lattnere07043c2010-09-29 05:30:03 +0000200<a name="lldb">LLDB: Low Level Debugger</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000201</h3>
Chris Lattnere07043c2010-09-29 05:30:03 +0000202
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000203<div>
Chris Lattnere07043c2010-09-29 05:30:03 +0000204
Bill Wendling7b7fa742011-10-26 18:46:16 +0000205<p>LLDB has advanced by leaps and bounds in the 3.0 timeframe. It is
206 dramatically more stable and useful, and includes both a
207 new <a href="http://lldb.llvm.org/tutorial.html">tutorial</a> and
208 a <a href="http://lldb.llvm.org/lldb-gdb.html">side-by-side comparison with
209 GDB</a>.</p>
Chris Lattnere07043c2010-09-29 05:30:03 +0000210
211</div>
212
213<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000214<h3>
Chris Lattnere07043c2010-09-29 05:30:03 +0000215<a name="libc++">libc++: C++ Standard Library</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000216</h3>
Chris Lattnere07043c2010-09-29 05:30:03 +0000217
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000218<div>
Chris Lattnere07043c2010-09-29 05:30:03 +0000219
Bill Wendling7b7fa742011-10-26 18:46:16 +0000220<p>Like compiler_rt, libc++ is now <a href="DeveloperPolicy.html#license">dual
221 licensed</a> under the MIT and UIUC license, allowing it to be used more
222 permissively.</p>
Chris Lattnere07043c2010-09-29 05:30:03 +0000223
224</div>
225
Chris Lattner61358ab2009-10-13 17:48:04 +0000226
Chris Lattner3d6a80a2011-04-07 03:08:22 +0000227<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000228<h3>
Chris Lattner3d6a80a2011-04-07 03:08:22 +0000229<a name="LLBrowse">LLBrowse: IR Browser</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000230</h3>
Daniel Dunbar97b01a82010-10-04 17:39:47 +0000231
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000232<div>
Bill Wendling7b7fa742011-10-26 18:46:16 +0000233
234<p><a href="http://llvm.org/svn/llvm-project/llbrowse/trunk/doc/LLBrowse.html">
235 LLBrowse</a> is an interactive viewer for LLVM modules. It can load any LLVM
236 module and displays its contents as an expandable tree view, facilitating an
237 easy way to inspect types, functions, global variables, or metadata nodes. It
238 is fully cross-platform, being based on the popular wxWidgets GUI
239 toolkit.</p>
240
Chris Lattner3d6a80a2011-04-07 03:08:22 +0000241</div>
242
243<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000244<h3>
Chris Lattner3d6a80a2011-04-07 03:08:22 +0000245<a name="vmkit">VMKit</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000246</h3>
Chris Lattner3d6a80a2011-04-07 03:08:22 +0000247
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000248<div>
Bill Wendling7b7fa742011-10-26 18:46:16 +0000249
Nicolas Geoffray54d5df92011-11-10 23:37:56 +0000250 <p>The <a href="http://vmkit.llvm.org/">VMKit project</a> is an
251 implementation of a Java Virtual Machine (Java VM or JVM) that uses LLVM for
252 static and just-in-time compilation.
253
254 <p>In the LLVM 3.0 time-frame, VMKit has had significant improvements on both
255 runtime and startup performance:</p>
256
257 <ul>
258 <li>Precompilation: by compiling ahead of time a small subset of Java's core
259 library, the startup performance have been highly optimized to the point that
260 running a 'Hello World' program takes less than 30 milliseconds.</li>
261
262 <li>Customization: by customizing virtual methods for individual classes,
263 the VM can statically determine the target of a virtual call, and decide to
264 inline it.</li>
265
266 <li>Inlining: the VM does more inlining than it did before, by allowing more
267 bytecode instructions to be inlined, and thanks to customization. It also
268 inlines GC barriers, and object allocations.</li>
269
270 <li>New exception model: the generated code for a method that does not do
271 any try/catch is not penalized anymore by the eventuality of calling a
272 method that throws an exception. Instead, the method that throws the
273 exception jumps directly to the method that could catch it.</li>
274 </ul>
Bill Wendling7b7fa742011-10-26 18:46:16 +0000275
Chris Lattner3d6a80a2011-04-07 03:08:22 +0000276</div>
277
278
Daniel Dunbar97b01a82010-10-04 17:39:47 +0000279<!--=========================================================================-->
Chris Lattner7a8e6c52011-04-05 18:38:45 +0000280<!--
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000281<h3>
Daniel Dunbar97b01a82010-10-04 17:39:47 +0000282<a name="klee">KLEE: A Symbolic Execution Virtual Machine</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000283</h3>
Daniel Dunbar97b01a82010-10-04 17:39:47 +0000284
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000285<div>
Daniel Dunbar97b01a82010-10-04 17:39:47 +0000286<p>
287<a href="http://klee.llvm.org/">KLEE</a> is a symbolic execution framework for
288programs in LLVM bitcode form. KLEE tries to symbolically evaluate "all" paths
289through the application and records state transitions that lead to fault
290states. This allows it to construct testcases that lead to faults and can even
291be used to verify some algorithms.
292</p>
293
Chris Lattnerbe2e1b52011-03-10 07:43:44 +0000294<p>UPDATE!</p>
Chris Lattner7a8e6c52011-04-05 18:38:45 +0000295</div>-->
Daniel Dunbar97b01a82010-10-04 17:39:47 +0000296
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000297</div>
Daniel Dunbar97b01a82010-10-04 17:39:47 +0000298
Chris Lattnerab68e9e2009-02-26 22:33:38 +0000299<!-- *********************************************************************** -->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000300<h2>
Chad Rosiere6291d02011-05-27 22:50:46 +0000301 <a name="externalproj">External Open Source Projects Using LLVM 3.0</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000302</h2>
Chris Lattnerab68e9e2009-02-26 22:33:38 +0000303<!-- *********************************************************************** -->
304
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000305<div>
Chris Lattner61358ab2009-10-13 17:48:04 +0000306
307<p>An exciting aspect of LLVM is that it is used as an enabling technology for
308 a lot of other language and tools projects. This section lists some of the
Chad Rosiere6291d02011-05-27 22:50:46 +0000309 projects that have already been updated to work with LLVM 3.0.</p>
Chris Lattner75547712010-10-03 23:49:06 +0000310
Chris Lattner3bfe57e2011-04-06 01:13:49 +0000311<!--=========================================================================-->
Bill Wendling7be6bc52011-10-26 00:17:54 +0000312<h3>AddressSanitizer</h3>
313
314<div>
315
316<p><a href="http://code.google.com/p/address-sanitizer/">AddressSanitizer</a>
317 uses compiler instrumentation and a specialized malloc library to find C/C++
318 bugs such as use-after-free and out-of-bound accesses to heap, stack, and
319 globals. The key feature of the tool is speed: the average slowdown
320 introduced by AddressSanitizer is less than 2x.</p>
321
322</div>
323
324<!--=========================================================================-->
Bill Wendling29817ea2011-10-26 00:14:36 +0000325<h3>ClamAV</h3>
326
327<div>
Bill Wendlingf2a78332011-10-25 01:01:42 +0000328
Bill Wendling29817ea2011-10-26 00:14:36 +0000329<p><a href="http://www.clamav.net">Clam AntiVirus</a> is an open source (GPL)
330 anti-virus toolkit for UNIX, designed especially for e-mail scanning on mail
331 gateways.</p>
332
333<p>Since version 0.96 it
334 has <a href="http://vrt-sourcefire.blogspot.com/2010/09/introduction-to-clamavs-low-level.html">bytecode
335 signatures</a> that allow writing detections for complex malware.</p>
336
337<p>It uses LLVM's JIT to speed up the execution of bytecode on X86, X86-64,
338 PPC32/64, falling back to its own interpreter otherwise. The git version was
339 updated to work with LLVM 3.0.</p>
340
341</div>
342
343<!--=========================================================================-->
Tobias Grosserae5a6fd2011-11-14 09:09:26 +0000344<h3>clang_complete for VIM</h3>
345
346<div>
347
348<p><a href="https://github.com/Rip-Rip/clang_complete">clang_complete</a> is a
349 VIM plugin, that provides accurate C/C++ autocompletion using the clang front
350 end. The development version of clang complete, can directly use libclang
351 which can maintain a cache to speed up auto completion.</p>
352
353</div>
354
355<!--=========================================================================-->
Bill Wendling65d1f412011-10-26 18:23:06 +0000356<h3>clReflect</h3>
357
358<div>
359
360<p><a href="https://bitbucket.org/dwilliamson/clreflect">clReflect</a> is a C++
361 parser that uses clang/LLVM to derive a light-weight reflection database
362 suitable for use in game development. It comes with a very simple runtime
363 library for loading and querying the database, requiring no external
364 dependencies (including CRT), and an additional utility library for object
365 management and serialisation.</p>
366
367</div>
368
369<!--=========================================================================-->
Bill Wendling63507d12011-10-29 01:10:01 +0000370<h3>Cling C++ Interpreter</h3>
371
372<div>
373
374<p><a href="http://cern.ch/cling">Cling</a> is an interactive compiler interface
375 (aka C++ interpreter). It uses LLVM's JIT and clang; it currently supports
376 C++ and C. It has a prompt interface, runs source files, calls into shared
377 libraries, prints the value of expressions, even does runtime lookup of
378 identifiers (dynamic scopes). And it just behaves like one would expect from
379 an interpreter.</p>
380
381</div>
382
383<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000384<h3>Crack Programming Language</h3>
Chris Lattner3bfe57e2011-04-06 01:13:49 +0000385
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000386<div>
Bill Wendling55d6e672011-11-03 20:10:01 +0000387
388<p><a href="http://code.google.com/p/crack-language/">Crack</a> aims to provide
389 the ease of development of a scripting language with the performance of a
390 compiled language. The language derives concepts from C++, Java and Python,
391 incorporating object-oriented programming, operator overloading and strong
392 typing.</p>
393
Chris Lattner3bfe57e2011-04-06 01:13:49 +0000394</div>
Chris Lattner3bfe57e2011-04-06 01:13:49 +0000395
396<!--=========================================================================-->
Bill Wendlingb99486f2011-11-08 05:22:54 +0000397<h3>Eero</h3>
398
399<div>
400
401<p><a href="http://eerolanguage.org/">Eero</a> is a fully
402 header-and-binary-compatible dialect of Objective-C 2.0, implemented with a
403 patched version of the Clang/LLVM compiler. It features a streamlined syntax,
404 Python-like indentation, and new operators, for improved readability and
405 reduced code clutter. It also has new features such as limited forms of
406 operator overloading and namespaces, and strict (type-and-operator-safe)
407 enumerations. It is inspired by languages such as Smalltalk, Python, and
408 Ruby.</p>
409
410</div>
411
412<!--=========================================================================-->
Chris Lattneradb417a2011-11-25 20:28:16 +0000413<h3>FAUST Real-Time Audio Signal Processing Language</h3>
414
415<div>
416
417<p><a href="http://faust.grame.fr/">FAUST</a> is a compiled language for
418 real-time audio signal processing. The name FAUST stands for Functional
419 AUdio STream. Its programming model combines two approaches: functional
420 programming and block diagram composition. In addition with the C, C++, Java
421 output formats, the Faust compiler can now generate LLVM bitcode, and works
422 with LLVM 2.7-3.0.
423 </p>
424
425</div>
426
427<!--=========================================================================-->
Bill Wendlingf9778192011-10-26 00:09:55 +0000428<h3>Glasgow Haskell Compiler (GHC)</h3>
429
430<div>
431
432<p>GHC is an open source, state-of-the-art programming suite for Haskell, a
433 standard lazy functional programming language. It includes an optimizing
434 static compiler generating good code for a variety of platforms, together
435 with an interactive system for convenient, quick development.</p>
436
437<p>GHC 7.0 and onwards include an LLVM code generator, supporting LLVM 2.8 and
438 later. Since LLVM 2.9, GHC now includes experimental support for the ARM
439 platform with LLVM 3.0.</p>
440
441</div>
442
443<!--=========================================================================-->
Bill Wendlingf62333d2011-10-25 20:35:31 +0000444<h3>gwXscript</h3>
445
446<div>
447
448<p><a href="http://botwars.tk/gwscript/">gwXscript</a> is an object oriented,
Bill Wendling7c38de22011-10-26 04:24:15 +0000449 aspect oriented programming language which can create both executables (ELF,
Bill Wendlingf62333d2011-10-25 20:35:31 +0000450 EXE) and shared libraries (DLL, SO, DYNLIB). The compiler is implemented in
451 its own language and translates scripts into LLVM-IR which can be optimized
452 and translated into native code by the LLVM framework. Source code in
453 gwScript contains definitions that expand the namespaces. So you can build
454 your project and simply 'plug out' features by removing a file. The remaining
455 project does not leave scars since you directly separate concerns by the
456 'template' feature of gwX. It is also possible to add new features to a
457 project by just adding files and without editing the original project. This
458 language is used for example to create games or content management systems
459 that should be extendable.</p>
460
461<p>gwXscript is strongly typed and offers comfort with its native types string,
462 hash and array. You can easily write new libraries in gwXscript or native
463 code. gwXscript is type safe and users should not be able to crash your
464 program or execute malicious code except code that is eating CPU time.</p>
465
466</div>
467
468<!--=========================================================================-->
Bill Wendling50cacc82011-10-26 22:55:18 +0000469<h3>include-what-you-use</h3>
470
471<div>
472
473<p><a href="http://code.google.com/p/include-what-you-use">include-what-you-use</a>
474 is a tool to ensure that a file directly <code>#include</code>s
475 all <code>.h</code> files that provide a symbol that the file uses. It also
476 removes superfluous <code>#include</code>s from source files.</p>
477
478</div>
479
480<!--=========================================================================-->
Bill Wendling32dc4d92011-11-07 22:05:17 +0000481<h3>ispc: The Intel SPMD Program Compiler</h3>
482
483<div>
484
485<p><a href="http://ispc.github.com">ispc</a> is a compiler for "single program,
486 multiple data" (SPMD) programs. It compiles a C-based SPMD programming
487 language to run on the SIMD units of CPUs; it often delivers 5-6x speedups on
488 a single core of a CPU with an 8-wide SIMD unit compared to serial code,
489 while still providing a clean and easy-to-understand programming model. For
490 an introduction to the language and its performance,
Chris Lattnerc343e312011-11-10 20:15:40 +0000491 see <a href="http://ispc.github.com/example.html">the walkthrough</a> of a short
Bill Wendling32dc4d92011-11-07 22:05:17 +0000492 example program. ispc is licensed under the BSD license.</p>
493
494</div>
Chris Lattnercc089772011-11-25 20:36:17 +0000495
496<!--=========================================================================-->
497<h3>The Julia Programming Language</h3>
498
499<div>
500
501<p><a href="http://github.com/JuliaLang/julia">Julia</a> is a high-level,
502 high-performance dynamic language for technical
503 computing. It provides a sophisticated compiler, distributed parallel
504 execution, numerical accuracy, and an extensive mathematical function
505 library. The compiler uses type inference to generate fast code
506 without any type declarations, and uses LLVM's optimization passes and
507 JIT compiler. The language is designed around multiple dispatch,
508 giving programs a large degree of flexibility. It is ready for use on many
509 kinds of problems.</p>
510</div>
Bill Wendling32dc4d92011-11-07 22:05:17 +0000511
512<!--=========================================================================-->
Bill Wendling57fd8762011-10-26 18:20:54 +0000513<h3>LanguageKit and Pragmatic Smalltalk</h3>
514
515<div>
516
517<p><a href="http://etoileos.com/etoile/features/languagekit/">LanguageKit</a> is
518 a framework for implementing dynamic languages sharing an object model with
519 Objective-C. It provides static and JIT compilation using LLVM along with
520 its own interpreter. Pragmatic Smalltalk is a dialect of Smalltalk, built on
521 top of LanguageKit, that interfaces directly with Objective-C, sharing the
522 same object representation and message sending behaviour. These projects are
523 developed as part of the &Eacute;toi&eacute; desktop environment.</p>
524
525</div>
526
527<!--=========================================================================-->
Bill Wendling03250532011-11-01 04:08:23 +0000528<h3>LuaAV</h3>
529
530<div>
531
532<p><a href="http://lua-av.mat.ucsb.edu/blog/">LuaAV</a> is a real-time
533 audiovisual scripting environment based around the Lua language and a
534 collection of libraries for sound, graphics, and other media protocols. LuaAV
535 uses LLVM and Clang to JIT compile efficient user-defined audio synthesis
536 routines specified in a declarative syntax.</p>
537
538</div>
539
540<!--=========================================================================-->
Bill Wendlingd4821b82011-10-26 00:16:17 +0000541<h3>Mono</h3>
542
543<div>
544
545<p>An open source, cross-platform implementation of C# and the CLR that is
546 binary compatible with Microsoft.NET. Has an optional, dynamically-loaded
547 LLVM code generation backend in Mini, the JIT compiler.</p>
548
549<p>Note that we use a Git mirror of LLVM with some patches. See:
550 https://github.com/mono/llvm</p>
551
552</div>
553
554<!--=========================================================================-->
Tobias Grosser093cb7e2011-11-14 09:09:23 +0000555<h3>Polly</h3>
556
557<div>
558
559<p><a href="http://polly.grosser.es">Polly</a> is an advanced data-locality
560 optimizer and automatic parallelizer. It uses an advanced, mathematical
561 model to calculate detailed data dependency information which it uses to
562 optimize the loop structure of a program. Polly can speed up sequential code
563 by improving memory locality and consequently the cache use. Furthermore,
564 Polly is able to expose different kind of parallelism which it exploits by
565 introducing (basic) OpenMP and SIMD code. A mid-term goal of Polly is to
566 automatically create optimized GPU code.</p>
567
568</div>
569
570<!--=========================================================================-->
Bill Wendlingba226272011-10-25 20:37:45 +0000571<h3>Portable OpenCL (pocl)</h3>
572
573<div>
574
575<p>Portable OpenCL is an open source implementation of the OpenCL standard which
576 can be easily adapted for new targets. One of the goals of the project is
577 improving performance portability of OpenCL programs, avoiding the need for
578 target-dependent manual optimizations. A "native" target is included, which
579 allows running OpenCL kernels on the host (CPU).</p>
580
581</div>
582
583<!--=========================================================================-->
Bill Wendling0bad98c2011-10-25 20:39:06 +0000584<h3>Pure</h3>
585
586<div>
587<p><a href="http://pure-lang.googlecode.com/">Pure</a> is an
588 algebraic/functional programming language based on term rewriting. Programs
589 are collections of equations which are used to evaluate expressions in a
590 symbolic fashion. The interpreter uses LLVM as a backend to JIT-compile Pure
591 programs to fast native code. Pure offers dynamic typing, eager and lazy
592 evaluation, lexical closures, a hygienic macro system (also based on term
593 rewriting), built-in list and matrix support (including list and matrix
594 comprehensions) and an easy-to-use interface to C and other programming
595 languages (including the ability to load LLVM bitcode modules, and inline C,
596 C++, Fortran and Faust code in Pure programs if the corresponding LLVM-enabled
597 compilers are installed).</p>
598
599<p>Pure version 0.48 has been tested and is known to work with LLVM 3.0
600 (and continues to work with older LLVM releases &gt;= 2.5).</p>
601
602</div>
603
604<!--=========================================================================-->
Bill Wendling537d85b2011-10-26 00:12:04 +0000605<h3>Renderscript</h3>
606
607<div>
608
609<p><a href="http://developer.android.com/guide/topics/renderscript/index.html">Renderscript</a>
610 is Android's advanced 3D graphics rendering and compute API. It provides a
611 portable C99-based language with extensions to facilitate common use cases
612 for enhancing graphics and thread level parallelism. The Renderscript
613 compiler frontend is based on Clang/LLVM. It emits a portable bitcode format
614 for the actual compiled script code, as well as reflects a Java interface for
615 developers to control the execution of the compiled bitcode. Executable
616 machine code is then generated from this bitcode by an LLVM backend on the
617 device. Renderscript is thus able to provide a mechanism by which Android
618 developers can improve performance of their applications while retaining
619 portability.</p>
620
621</div>
622
623<!--=========================================================================-->
Bill Wendling7d5b6212011-10-25 20:40:26 +0000624<h3>SAFECode</h3>
625
626<div>
627
628<p><a href="http://safecode.cs.illinois.edu">SAFECode</a> is a memory safe C/C++
629 compiler built using LLVM. It takes standard, unannotated C/C++ code,
630 analyzes the code to ensure that memory accesses and array indexing
631 operations are safe, and instruments the code with run-time checks when
632 safety cannot be proven statically. SAFECode can be used as a debugging aid
633 (like Valgrind) to find and repair memory safety bugs. It can also be used
634 to protect code from security attacks at run-time.</p>
635
636</div>
637
638<!--=========================================================================-->
Bill Wendling02b77b72011-10-26 07:38:19 +0000639<h3>The Stupid D Compiler (SDC)</h3>
640
641<div>
642
643<p><a href="https://github.com/bhelyer/SDC">The Stupid D Compiler</a> is a
644 project seeking to write a self-hosting compiler for the D programming
645 language without using the frontend of the reference compiler (DMD).</p>
646
647</div>
648
649<!--=========================================================================-->
Bill Wendling2d7b4af2011-10-25 20:24:32 +0000650<h3>TTA-based Co-design Environment (TCE)</h3>
651
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000652<div>
Bill Wendling2d7b4af2011-10-25 20:24:32 +0000653
Chris Lattner3bfe57e2011-04-06 01:13:49 +0000654<p>TCE is a toolset for designing application-specific processors (ASP) based on
Bill Wendling2d7b4af2011-10-25 20:24:32 +0000655 the Transport triggered architecture (TTA). The toolset provides a complete
656 co-design flow from C/C++ programs down to synthesizable VHDL and parallel
657 program binaries. Processor customization points include the register files,
658 function units, supported operations, and the interconnection network.</p>
Bill Wendling7b7fa742011-10-26 18:46:16 +0000659
Chris Lattner3bfe57e2011-04-06 01:13:49 +0000660<p>TCE uses Clang and LLVM for C/C++ language support, target independent
Bill Wendling2d7b4af2011-10-25 20:24:32 +0000661 optimizations and also for parts of code generation. It generates new
Bill Wendling7b7fa742011-10-26 18:46:16 +0000662 LLVM-based code generators "on the fly" for the designed TTA processors and
663 loads them in to the compiler backend as runtime libraries to avoid
Bill Wendling2d7b4af2011-10-25 20:24:32 +0000664 per-target recompilation of larger parts of the compiler chain.</p>
Chris Lattner3bfe57e2011-04-06 01:13:49 +0000665
Bill Wendling7b7fa742011-10-26 18:46:16 +0000666</div>
Chris Lattner3bfe57e2011-04-06 01:13:49 +0000667
668<!--=========================================================================-->
Bill Wendling628c2662011-10-25 20:27:37 +0000669<h3>Tart Programming Language</h3>
670
671<div>
672
673<p><a href="http://code.google.com/p/tart/">Tart</a> is a general-purpose,
674 strongly typed programming language designed for application
675 developers. Strongly inspired by Python and C#, Tart focuses on practical
676 solutions for the professional software developer, while avoiding the clutter
677 and boilerplate of legacy languages like Java and C++. Although Tart is still
678 in development, the current implementation supports many features expected of
679 a modern programming language, such as garbage collection, powerful
680 bidirectional type inference, a greatly simplified syntax for template
681 metaprogramming, closures and function literals, reflection, operator
682 overloading, explicit mutability and immutability, and much more. Tart is
683 flexible enough to accommodate a broad range of programming styles and
684 philosophies, while maintaining a strong commitment to simplicity, minimalism
685 and elegance in design.</p>
686
687</div>
688
689<!--=========================================================================-->
Bill Wendling644ce532011-10-26 09:25:01 +0000690<h3>ThreadSanitizer</h3>
691
692<div>
693
694<p><a href="http://code.google.com/p/data-race-test/">ThreadSanitizer</a> is a
695 data race detector for (mostly) C and C++ code, available for Linux, Mac OS
696 and Windows. On different systems, we use binary instrumentation frameworks
Bill Wendlingae8538e2011-10-29 01:11:15 +0000697 (Valgrind and Pin) as frontends that generate the program events for the race
698 detection algorithm. On Linux, there's an option of using LLVM-based
699 compile-time instrumentation.</p>
Bill Wendling644ce532011-10-26 09:25:01 +0000700
701</div>
702
703<!--=========================================================================-->
Bill Wendling8a924c62011-10-26 07:42:45 +0000704<h3>The ZooLib C++ Cross-Platform Application Framework</h3>
705
706<div>
707
708<p><a href="http://www.zoolib.org/">ZooLib</a> is Open Source under the MIT
709 License. It provides GUI, filesystem access, TCP networking, thread-safe
710 memory management, threading and locking for Mac OS X, Classic Mac OS,
711 Microsoft Windows, POSIX operating systems with X11, BeOS, Haiku, Apple's iOS
712 and Research in Motion's BlackBerry.</p>
713
714<p>My current work is to use CLang's static analyzer to improve ZooLib's code
715 quality. I also plan to set up LLVM compiles of the demo programs and test
716 programs using CLang and LLVM on all the platforms that CLang, LLVM and
717 ZooLib all support.</p>
718
719</div>
Chris Lattnera844a3e2011-04-07 03:09:21 +0000720
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000721</div>
722
Chris Lattner8348b472008-06-08 21:58:17 +0000723<!-- *********************************************************************** -->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000724<h2>
Chad Rosiere6291d02011-05-27 22:50:46 +0000725 <a name="whatsnew">What's New in LLVM 3.0?</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000726</h2>
Chris Lattner8348b472008-06-08 21:58:17 +0000727<!-- *********************************************************************** -->
728
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000729<div>
Chris Lattnerf8e0b4e2008-06-08 22:59:35 +0000730
Chris Lattner61358ab2009-10-13 17:48:04 +0000731<p>This release includes a huge number of bug fixes, performance tweaks and
Bill Wendling7b7fa742011-10-26 18:46:16 +0000732 minor improvements. Some of the major improvements and new features are
733 listed in this section.</p>
Chris Lattner61358ab2009-10-13 17:48:04 +0000734
Chris Lattner914ce462010-04-22 06:28:20 +0000735<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000736<h3>
Chris Lattner252b83d2008-02-06 18:00:06 +0000737<a name="majorfeatures">Major New Features</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000738</h3>
Chris Lattner252b83d2008-02-06 18:00:06 +0000739
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000740<div>
Chris Lattner252b83d2008-02-06 18:00:06 +0000741
Chris Lattnerc343e312011-11-10 20:15:40 +0000742<p><b>llvm-gcc is gone</b></p>
Chris Lattnerf3e5bc62007-05-14 06:56:09 +0000743
Chris Lattnerc343e312011-11-10 20:15:40 +0000744<p>LLVM 3.0 includes several major new capabilities:</p>
745
Chris Lattnerdec23b62011-11-15 22:13:27 +0000746<!-- Near dead:
747 Analysis/RegionInfo.h + Dom Frontiers
748 SparseBitVector: used in LiveVar.
749
750 -->
751
752<!--
753 Type system rewrite.
754 Better performance for Neon code in clang due to SRoA improvements.
755 New regalloc on by default. Lin scan going away in 3.1
756 PGO / builtin_expect improvements (summary needed)
757 Big EH rewrite.
758 AVX support, assembler, compiler and disassembler.
759 IndVar improvements: andy
760 PTX backend improvements: Justin
761 llvm-rtdyld & MC JIT: JimG
762 InstAliases now automatically used in the asmprinter where they are shorter.
763 Integrated assembler on by default for arm/thumb?
764 PostOrder Dominator frontiers were removed.
765 Line Profiling / gcov support
766 EH and debug information produced with CFI directives, yielding smaller executables: http://blog.mozilla.com/respindola/2011/05/12/cfi-directives/
767 X86-64 generates smaller and faster code at -O0 (fast isel improvements)
768 Better code generation for Cortex-A9
769 Many APIs take ArrayRef's now.
770 Pass manager extension API.
Chris Lattner6a007d12011-11-25 20:33:27 +0000771
772
773Information about branch probability and basic block frequency is now available within LLVM based on a combination of static branch prediction heuristics and __builtin_expect calls. That information is currently used for register spill placement and if-conversion, with additional optimizations planned for future releases. The same frameworks are intended for eventual use with profile-guided optimization, but that is not yet implemented.
Chris Lattnerdec23b62011-11-15 22:13:27 +0000774
775 -->
776
Chris Lattnerf3e5bc62007-05-14 06:56:09 +0000777<ul>
Chris Lattner1efe27e2011-04-06 00:45:11 +0000778
Chad Rosiere6291d02011-05-27 22:50:46 +0000779<!--
780<li></li>
781-->
Chris Lattner7a8e6c52011-04-05 18:38:45 +0000782
Chris Lattner8170c102008-02-10 08:18:42 +0000783</ul>
Chris Lattner0a6f6d52011-04-05 07:19:28 +0000784
Chris Lattnerf3e5bc62007-05-14 06:56:09 +0000785</div>
786
Chris Lattner84977642007-09-21 03:54:09 +0000787<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000788<h3>
Chris Lattner917cc712009-03-02 02:37:32 +0000789<a name="coreimprovements">LLVM IR and Core Improvements</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000790</h3>
Chris Lattnerf304ffc2008-02-10 08:17:19 +0000791
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000792<div>
Bill Wendling7b7fa742011-10-26 18:46:16 +0000793
Chris Lattner61358ab2009-10-13 17:48:04 +0000794<p>LLVM IR has several new features for better support of new targets and that
Bill Wendling7b7fa742011-10-26 18:46:16 +0000795 expose new optimization opportunities:</p>
Chris Lattnerf304ffc2008-02-10 08:17:19 +0000796
Bill Wendlingbc5f6dd2011-10-26 18:33:01 +0000797<p>One of the biggest changes is that 3.0 has a new exception handling
798 system. The old system used LLVM intrinsics to convey the exception handling
799 information to the code generator. It worked in most cases, but not
800 all. Inlining was especially difficult to get right. Also, the intrinsics
801 could be moved away from the <code>invoke</code> instruction, making it hard
802 to recover that information.</p>
803
804<p>The new EH system makes exception handling a first-class member of the IR. It
805 adds two new instructions:</p>
806
Chris Lattner791f77b2008-06-05 06:25:56 +0000807<ul>
Bill Wendlingbc5f6dd2011-10-26 18:33:01 +0000808 <li><a href="LangRef.html#i_landingpad"><code>landingpad</code></a> &mdash;
809 this instruction defines a landing pad basic block. It contains all of the
810 information that's needed by the code generator. It's also required to be
811 the first non-PHI instruction in the landing pad. In addition, a landing
812 pad may be jumped to only by the unwind edge of an <code>invoke</code>
813 instruction.</li>
814
815 <li><a href="LangRef.html#i_resume"><code>resume</code></a> &mdash; this
816 instruction causes the current exception to resume traveling up the
817 stack. It replaces the <code>@llvm.eh.resume</code> intrinsic.</li>
Chris Lattnerf304ffc2008-02-10 08:17:19 +0000818</ul>
Mikhail Glushenkovea65d7d2008-10-13 02:08:34 +0000819
Bill Wendlingbc5f6dd2011-10-26 18:33:01 +0000820<p>Converting from the old EH API to the new EH API is rather simple, because a
821 lot of complexity has been removed. The two intrinsics,
822 <code>@llvm.eh.exception</code> and <code>@llvm.eh.selector</code> have been
823 superceded by the <code>landingpad</code> instruction. Instead of generating
824 a call to <code>@llvm.eh.exception</code> and <code>@llvm.eh.selector</code>:
825
826<div class="doc_code">
827<pre>
828Function *ExcIntr = Intrinsic::getDeclaration(TheModule,
829 Intrinsic::eh_exception);
830Function *SlctrIntr = Intrinsic::getDeclaration(TheModule,
831 Intrinsic::eh_selector);
832
833// The exception pointer.
834Value *ExnPtr = Builder.CreateCall(ExcIntr, "exc_ptr");
835
836std::vector&lt;Value*&gt; Args;
837Args.push_back(ExnPtr);
838Args.push_back(Builder.CreateBitCast(Personality,
839 Type::getInt8PtrTy(Context)));
840
841<i>// Add selector clauses to Args.</i>
842
843// The selector call.
844Builder.CreateCall(SlctrIntr, Args, "exc_sel");
845</pre>
846</div>
847
848<p>You should instead generate a <code>landingpad</code> instruction, that
849 returns an exception object and selector value:</p>
850
851<div class="doc_code">
852<pre>
853LandingPadInst *LPadInst =
854 Builder.CreateLandingPad(StructType::get(Int8PtrTy, Int32Ty, NULL),
855 Personality, 0);
856
857Value *LPadExn = Builder.CreateExtractValue(LPadInst, 0);
858Builder.CreateStore(LPadExn, getExceptionSlot());
859
860Value *LPadSel = Builder.CreateExtractValue(LPadInst, 1);
861Builder.CreateStore(LPadSel, getEHSelectorSlot());
862</pre>
863</div>
864
865<p>It's now trivial to add the individual clauses to the <code>landingpad</code>
866 instruction.</p>
867
868<div class="doc_code">
869<pre>
870<i><b>// Adding a catch clause</b></i>
871Constant *TypeInfo = getTypeInfo();
872LPadInst-&gt;addClause(TypeInfo);
873
874<i><b>// Adding a C++ catch-all</b></i>
875LPadInst-&gt;addClause(Constant::getNullValue(Builder.getInt8PtrTy()));
876
877<i><b>// Adding a cleanup</b></i>
878LPadInst-&gt;setCleanup(true);
879
880<i><b>// Adding a filter clause</b></i>
881std::vector&lt;Constant*&gt; TypeInfos;
882Constant *TypeInfo = getFilterTypeInfo();
883TypeInfos.push_back(Builder.CreateBitCast(TypeInfo, Builder.getInt8PtrTy()));
884
885ArrayType *FilterTy = ArrayType::get(Int8PtrTy, TypeInfos.size());
886LPadInst-&gt;addClause(ConstantArray::get(FilterTy, TypeInfos));
887</pre>
888</div>
889
890<p>Converting from using the <code>@llvm.eh.resume</code> intrinsic to
891 the <code>resume</code> instruction is trivial. It takes the exception
892 pointer and exception selector values returned by
893 the <code>landingpad</code> instruction:</p>
894
895<div class="doc_code">
896<pre>
897Type *UnwindDataTy = StructType::get(Builder.getInt8PtrTy(),
898 Builder.getInt32Ty(), NULL);
899Value *UnwindData = UndefValue::get(UnwindDataTy);
900Value *ExcPtr = Builder.CreateLoad(getExceptionObjSlot());
901Value *ExcSel = Builder.CreateLoad(getExceptionSelSlot());
902UnwindData = Builder.CreateInsertValue(UnwindData, ExcPtr, 0, "exc_ptr");
903UnwindData = Builder.CreateInsertValue(UnwindData, ExcSel, 1, "exc_sel");
904Builder.CreateResume(UnwindData);
905</pre>
906</div>
907
Chris Lattnerf304ffc2008-02-10 08:17:19 +0000908</div>
909
910<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000911<h3>
Andrew Trick5aab6382011-11-06 17:59:24 +0000912<a name="loopoptimization">Loop Optimization Improvements</a>
913</h3>
914
915<div>
916<p>The induction variable simplification pass in 3.0 only modifies
917 induction variables when profitable. Sign and zero extension
918 elimination, linear function test replacement, loop unrolling, and
919 other simplifications that require induction variable analysis have
920 been generalized so they no longer require loops to be rewritten in a
921 typically suboptimal form prior to optimization. This new design
922 preserves more IR level information, avoids undoing earlier loop
923 optimizations (particularly hand-optimized loops), and no longer
924 strongly depends on the code generator rewriting loops a second time
925 in a now optimal form--an intractable problem.</p>
926
927<p>The original behavior can be restored with -mllvm -enable-iv-rewrite;
928 however, support for this mode will be short lived. As such, bug
929 reports should be filed for any significant performance regressions
930 when moving from -mllvm -enable-iv-rewrite to the 3.0 default mode.</p>
931</div>
932
933<!--=========================================================================-->
934<h3>
Chris Lattneracce85d2008-02-10 07:46:44 +0000935<a name="optimizer">Optimizer Improvements</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000936</h3>
Chris Lattneracce85d2008-02-10 07:46:44 +0000937
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000938<div>
Chris Lattneracce85d2008-02-10 07:46:44 +0000939
Chris Lattner61358ab2009-10-13 17:48:04 +0000940<p>In addition to a large array of minor performance tweaks and bug fixes, this
Bill Wendling7b7fa742011-10-26 18:46:16 +0000941 release includes a few major enhancements and additions to the
942 optimizers:</p>
Chris Lattneracce85d2008-02-10 07:46:44 +0000943
944<ul>
Chad Rosiere6291d02011-05-27 22:50:46 +0000945<!--
946<li></li>
947-->
Chris Lattnerc5ac61d2011-04-06 05:50:04 +0000948</li>
Chris Lattner1efe27e2011-04-06 00:45:11 +0000949
Chris Lattner11b66112010-10-04 02:42:39 +0000950</ul>
951
Chris Lattner61358ab2009-10-13 17:48:04 +0000952</div>
Chris Lattner917cc712009-03-02 02:37:32 +0000953
Chris Lattner0b832202008-06-08 02:45:07 +0000954<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000955<h3>
Chris Lattner4ba2b652010-09-30 16:31:33 +0000956<a name="mc">MC Level Improvements</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000957</h3>
Chris Lattner4ba2b652010-09-30 16:31:33 +0000958
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000959<div>
Bill Wendling7b7fa742011-10-26 18:46:16 +0000960
961<p>The LLVM Machine Code (aka MC) subsystem was created to solve a number of
962 problems in the realm of assembly, disassembly, object file format handling,
963 and a number of other related areas that CPU instruction-set level tools work
964 in.</p>
Chris Lattner7d9b6b42010-10-02 21:59:30 +0000965
Chris Lattner7d9b6b42010-10-02 21:59:30 +0000966<ul>
Jim Grosbach2552de02011-11-24 00:49:21 +0000967 <li>The ELF object streamers are much more full featured.</li>
968 <li>Target dependent relocation handling has been refactored into the Targets.</li>
969 <li>Early stage MC-JIT infrastructure has been implemented.</li>
Chris Lattner7d9b6b42010-10-02 21:59:30 +0000970</ul>
971
Jim Grosbach2552de02011-11-24 00:49:21 +0000972<p>The MC-JIT is a major new feature for MC, and will eventually grow to replace
973the current JIT implementation. It emits object files direct to memory and
974uses a runtime dynamic linker to resolve references and drive lazy compilation.
975The MC-JIT enables much greater code reuse between the JIT and the static
976compiler and provides better integration with the platform ABI as a result.</p>
977
Bill Wendling7b7fa742011-10-26 18:46:16 +0000978<p>For more information, please see
979 the <a href="http://blog.llvm.org/2010/04/intro-to-llvm-mc-project.html">Intro
980 to the LLVM MC Project Blog Post</a>.</p>
Chris Lattner4ba2b652010-09-30 16:31:33 +0000981
NAKAMURA Takumi45c435a2011-04-05 08:24:22 +0000982</div>
Chris Lattner4ba2b652010-09-30 16:31:33 +0000983
Chris Lattner4ba2b652010-09-30 16:31:33 +0000984<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000985<h3>
Chris Lattner511433e2009-03-02 03:24:11 +0000986<a name="codegen">Target Independent Code Generator Improvements</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000987</h3>
Chris Lattner0b832202008-06-08 02:45:07 +0000988
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000989<div>
Chris Lattner0b832202008-06-08 02:45:07 +0000990
Mikhail Glushenkovf795ef02009-03-01 18:09:47 +0000991<p>We have put a significant amount of work into the code generator
Bill Wendling7b7fa742011-10-26 18:46:16 +0000992 infrastructure, which allows us to implement more aggressive algorithms and
993 make it run faster:</p>
Chris Lattner0b832202008-06-08 02:45:07 +0000994
995<ul>
Chad Rosiere6291d02011-05-27 22:50:46 +0000996<!--
997<li></li>
998-->
Chris Lattner84977642007-09-21 03:54:09 +0000999</ul>
Chris Lattner84977642007-09-21 03:54:09 +00001000</div>
1001
Chris Lattner917cc712009-03-02 02:37:32 +00001002<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +00001003<h3>
Chris Lattner917cc712009-03-02 02:37:32 +00001004<a name="x86">X86-32 and X86-64 Target Improvements</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +00001005</h3>
Chris Lattner917cc712009-03-02 02:37:32 +00001006
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +00001007<div>
Bill Wendling7b7fa742011-10-26 18:46:16 +00001008
1009<p>New features and major changes in the X86 target include:</p>
Chris Lattner917cc712009-03-02 02:37:32 +00001010
1011<ul>
Bill Wendling7b7fa742011-10-26 18:46:16 +00001012
Chris Lattner62f009a2011-11-15 22:48:24 +00001013 <li>The X86 backend now supports
1014 all <a href="http://llvm.org/PR879">inline assembly that uses the X86
1015 floating point stack</a>.</li>
1016
Bill Wendling7b7fa742011-10-26 18:46:16 +00001017 <li>The CRC32 intrinsics have been renamed. The intrinsics were previously
1018 <code>@llvm.x86.sse42.crc32.[8|16|32]</code>
1019 and <code>@llvm.x86.sse42.crc64.[8|64]</code>. They have been renamed to
1020 <code>@llvm.x86.sse42.crc32.32.[8|16|32]</code> and
1021 <code>@llvm.x86.sse42.crc32.64.[8|64]</code>.</li>
Chad Rosierf94c9c12011-05-27 20:13:10 +00001022
Chris Lattner917cc712009-03-02 02:37:32 +00001023</ul>
1024
Chris Lattner917cc712009-03-02 02:37:32 +00001025</div>
Chris Lattner84977642007-09-21 03:54:09 +00001026
1027<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +00001028<h3>
Chris Lattner61358ab2009-10-13 17:48:04 +00001029<a name="ARM">ARM Target Improvements</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +00001030</h3>
Chris Lattnerc441fb82009-03-01 02:30:21 +00001031
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +00001032<div>
Bill Wendling7b7fa742011-10-26 18:46:16 +00001033
1034<p>New features of the ARM target include:</p>
Chris Lattnerc441fb82009-03-01 02:30:21 +00001035
1036<ul>
Jim Grosbach2552de02011-11-24 00:49:21 +00001037 <li>Reworked Set Jump Long Jump EH Lowering,</li>
1038 <li>improved support for Cortex-M series processors, and</li>
1039 <li>beta quality integrated assembler support.</li>
Bob Wilsone8472772010-09-13 17:39:35 +00001040</ul>
Chris Lattner61358ab2009-10-13 17:48:04 +00001041</div>
Chris Lattner7b95c382011-11-15 22:23:46 +00001042
Chris Lattner1efe27e2011-04-06 00:45:11 +00001043
1044<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +00001045<h3>
Akira Hatanaka5381cbf2011-11-15 21:33:05 +00001046<a name="MIPS">MIPS Target Improvements</a>
1047</h3>
1048
1049<div>
1050
1051<p>New features and major changes in the MIPS target include:</p>
1052
1053<ul>
1054 <li>Most MIPS32r1 and r2 instructions are now supported.</li>
1055 <li>LE/BE MIPS32r1/r2 has been tested extensively.</li>
1056 <li>O32 ABI has been fully tested.</li>
1057 <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>
1058 <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>
1059 <li>Support for old-style JIT is complete.</li>
1060 <li>Support for old architectures (MIPS1 and MIPS2) has been removed.</li>
1061 <li>Initial support for MIPS64 has been added.</li>
1062</ul>
1063</div>
Chris Lattner7b95c382011-11-15 22:23:46 +00001064
1065<!--=========================================================================-->
1066<h3>
1067 <a name="PTX">PTX Target Improvements</a>
1068</h3>
1069
1070<div>
Akira Hatanaka5381cbf2011-11-15 21:33:05 +00001071
Chris Lattner7b95c382011-11-15 22:23:46 +00001072 <p>
1073 The PTX back-end is still experimental, but is fairly usable for compute kernels
1074 in LLVM 3.0. Most scalar arithmetic is implemented, as well as intrinsics to
1075 access the special PTX registers and sync instructions. The major missing
1076 pieces are texture/sampler support and some vector operations.</p>
1077
1078 <p>That said, the backend is already being used for domain-specific languages
1079 and works well with the <a href="http://www.pcc.me.uk/~peter/libclc/">libclc
1080 library</a> to supply OpenCL built-ins. With it, you can use Clang to compile
1081 OpenCL code into PTX and execute it by loading the resulting PTX as a binary
1082 blob using the nVidia OpenCL library. It has been tested with several OpenCL
1083 programs, including some from the nVidia GPU Computing SDK, and the performance
1084 is on par with the nVidia compiler.</p>
1085
1086</div>
1087
Akira Hatanaka5381cbf2011-11-15 21:33:05 +00001088<!--=========================================================================-->
1089<h3>
Chris Lattner1efe27e2011-04-06 00:45:11 +00001090<a name="OtherTS">Other Target Specific Improvements</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +00001091</h3>
Chris Lattnerc441fb82009-03-01 02:30:21 +00001092
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +00001093<div>
Bill Wendling7b7fa742011-10-26 18:46:16 +00001094
Chris Lattnerc343e312011-11-10 20:15:40 +00001095 <p>PPC32/ELF va_arg was implemented.</p>
1096 <p>PPC32 initial support for .o file writing was implemented.</p>
Wesley Peck3ff16db2011-11-14 18:56:41 +00001097 <p>MicroBlaze scheduling itineraries were added that model the
1098 3-stage and the 5-stage pipeline architectures. The 3-stage
1099 pipeline model can be selected with <code>-mcpu=mblaze3</code>
1100 and the 5-stage pipeline model can be selected with
1101 <code>-mcpu=mblaze5</code>.</p>
Chris Lattnerc343e312011-11-10 20:15:40 +00001102
Chris Lattner1efe27e2011-04-06 00:45:11 +00001103<ul>
Chad Rosiere6291d02011-05-27 22:50:46 +00001104<!--
1105<li></li>
1106-->
Chris Lattner1efe27e2011-04-06 00:45:11 +00001107</ul>
Bill Wendling7b7fa742011-10-26 18:46:16 +00001108
Chris Lattner1efe27e2011-04-06 00:45:11 +00001109</div>
Chris Lattner77d29b12008-06-05 08:02:49 +00001110
1111<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +00001112<h3>
Chris Lattnerf6662f92008-10-13 17:57:36 +00001113<a name="changes">Major Changes and Removed Features</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +00001114</h3>
Chris Lattnerf6662f92008-10-13 17:57:36 +00001115
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +00001116<div>
Chris Lattnerf6662f92008-10-13 17:57:36 +00001117
Bill Wendling2626dba2011-08-03 22:18:20 +00001118<p>If you're already an LLVM user or developer with out-of-tree changes based on
1119 LLVM 2.9, this section lists some "gotchas" that you may run into upgrading
1120 from the previous release.</p>
Chris Lattnerf6662f92008-10-13 17:57:36 +00001121
1122<ul>
Eric Christopher90d6ec52011-09-28 19:47:28 +00001123 <li>The <code>LLVMC</code> front end code was removed while separating
1124 out language independence.</li>
Jay Foadf42e9b22011-08-04 10:43:43 +00001125 <li>The <code>LowerSetJmp</code> pass wasn't used effectively by any
1126 target and has been removed.</li>
Rafael Espindolaf940a1a2011-08-30 23:03:45 +00001127 <li>The old <code>TailDup</code> pass was not used in the standard pipeline
1128 and was unable to update ssa form, so it has been removed.
Eli Friedmanf03bb262011-08-12 22:50:01 +00001129 <li>The syntax of volatile loads and stores in IR has been changed to
1130 "<code>load volatile</code>"/"<code>store volatile</code>". The old
1131 syntax ("<code>volatile load</code>"/"<code>volatile store</code>")
1132 is still accepted, but is now considered deprecated.</li>
Eli Friedman526e1bb2011-10-26 00:55:23 +00001133 <li>The old atomic intrinscs (<code>llvm.memory.barrier</code> and
1134 <code>llvm.atomic.*</code>) are now gone. Please use the new atomic
1135 instructions, described in the <a href="Atomics.html">atomics guide</a>.
Devang Patelb34dd132008-10-14 20:03:43 +00001136</ul>
Chris Lattnerf6662f92008-10-13 17:57:36 +00001137
NAKAMURA Takumi2026de22011-08-22 23:22:05 +00001138<h4>Windows (32-bit)</h4>
1139<div>
Bill Wendling7b7fa742011-10-26 18:46:16 +00001140
NAKAMURA Takumi2026de22011-08-22 23:22:05 +00001141<ul>
1142 <li>On Win32(MinGW32 and MSVC), Windows 2000 will not be supported.
1143 Windows XP or higher is required.</li>
1144</ul>
Bill Wendling7b7fa742011-10-26 18:46:16 +00001145
NAKAMURA Takumi2026de22011-08-22 23:22:05 +00001146</div>
1147
Chris Lattnerf6662f92008-10-13 17:57:36 +00001148</div>
1149
Daniel Dunbarf0233c62010-10-04 20:11:41 +00001150<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +00001151<h3>
Chris Lattner1efe27e2011-04-06 00:45:11 +00001152<a name="api_changes">Internal API Changes</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +00001153</h3>
Daniel Dunbarf0233c62010-10-04 20:11:41 +00001154
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +00001155<div>
Daniel Dunbarf0233c62010-10-04 20:11:41 +00001156
Chris Lattner1efe27e2011-04-06 00:45:11 +00001157<p>In addition, many APIs have changed in this release. Some of the major
Bill Wendling16005252011-08-02 06:20:17 +00001158 LLVM API changes are:</p>
Daniel Dunbarf0233c62010-10-04 20:11:41 +00001159
1160<ul>
Bill Wendling7b7fa742011-10-26 18:46:16 +00001161 <li>The biggest and most pervasive change is that llvm::Type's are no longer
1162 returned or accepted as 'const' values. Instead, just pass around
1163 non-const Type's.</li>
Chris Lattnerd1324302011-07-18 04:56:02 +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>The way the type system works has been
1239 rewritten: <code>PATypeHolder</code> and <code>OpaqueType</code> are gone,
1240 and all APIs deal with <code>Type*</code> instead of <code>const
1241 Type*</code>. If you need to create recursive structures, then create a
1242 named structure, and use <code>setBody()</code> when all its elements are
1243 built. Type merging and refining is gone too: named structures are not
1244 merged with other structures, even if their layout is identical. (of
1245 course anonymous structures are still uniqued by layout).</li>
Torok Edwinf16e2d42011-09-30 13:07:52 +00001246
Bill Wendling7b7fa742011-10-26 18:46:16 +00001247 <li>TargetSelect.h moved to Support/ from Target/</li>
Torok Edwinf16e2d42011-09-30 13:07:52 +00001248
Bill Wendling7b7fa742011-10-26 18:46:16 +00001249 <li>UpgradeIntrinsicCall no longer upgrades pre-2.9 intrinsic calls (for
1250 example <code>llvm.memset.i32</code>).</li>
Torok Edwinf16e2d42011-09-30 13:07:52 +00001251
Bill Wendling7b7fa742011-10-26 18:46:16 +00001252 <li>It is mandatory to initialize all out-of-tree passes too and their dependencies now with
1253 <code>INITIALIZE_PASS{BEGIN,END,}</code>
1254 and <code>INITIALIZE_{PASS,AG}_DEPENDENCY</code>.</li>
Torok Edwinf16e2d42011-09-30 13:07:52 +00001255
Bill Wendling7b7fa742011-10-26 18:46:16 +00001256 <li>The interface for MemDepResult in MemoryDependenceAnalysis has been
1257 enhanced with new return types Unknown and NonFuncLocal, in addition to
1258 the existing types Clobber, Def, and NonLocal.</li>
Daniel Dunbarf0233c62010-10-04 20:11:41 +00001259</ul>
Bill Wendling7b7fa742011-10-26 18:46:16 +00001260
Daniel Dunbarf0233c62010-10-04 20:11:41 +00001261</div>
Chris Lattnerf6662f92008-10-13 17:57:36 +00001262
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +00001263</div>
1264
Chris Lattner79c3fe12003-10-02 04:57:28 +00001265<!-- *********************************************************************** -->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +00001266<h2>
Misha Brukman500bc302003-11-22 00:38:41 +00001267 <a name="knownproblems">Known Problems</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +00001268</h2>
Chris Lattner79c3fe12003-10-02 04:57:28 +00001269<!-- *********************************************************************** -->
1270
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +00001271<div>
Misha Brukman500bc302003-11-22 00:38:41 +00001272
Bill Wendling7b7fa742011-10-26 18:46:16 +00001273<p>This section contains significant known problems with the LLVM system, listed
1274 by component. If you run into a problem, please check
1275 the <a href="http://llvm.org/bugs/">LLVM bug database</a> and submit a bug if
1276 there isn't already one.</p>
Chris Lattner79c3fe12003-10-02 04:57:28 +00001277
Chris Lattnerf5ee1702004-03-14 02:03:02 +00001278<!-- ======================================================================= -->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +00001279<h3>
Chris Lattnerf5ee1702004-03-14 02:03:02 +00001280 <a name="experimental">Experimental features included with this release</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +00001281</h3>
Chris Lattner79c3fe12003-10-02 04:57:28 +00001282
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +00001283<div>
Chris Lattnerf5ee1702004-03-14 02:03:02 +00001284
Misha Brukman6df9e2c2004-05-12 21:46:05 +00001285<p>The following components of this LLVM release are either untested, known to
Bill Wendling7b7fa742011-10-26 18:46:16 +00001286 be broken or unreliable, or are in early development. These components
1287 should not be relied on, and bugs should not be filed against them, but they
1288 may be useful to some people. In particular, if you would like to work on
1289 one of these components, please contact us on
1290 the <a href="http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev">LLVMdev
1291 list</a>.</p>
Chris Lattnerf5ee1702004-03-14 02:03:02 +00001292
1293<ul>
Bill Wendling7b7fa742011-10-26 18:46:16 +00001294 <li>The Alpha, Blackfin, CellSPU, MicroBlaze, MSP430, MIPS, PTX, SystemZ and
1295 XCore backends are experimental.</li>
1296
1297 <li><tt>llc</tt> "<tt>-filetype=obj</tt>" is experimental on all targets other
1298 than darwin and ELF X86 systems.</li>
Chris Lattnerf5ee1702004-03-14 02:03:02 +00001299</ul>
1300
1301</div>
Chris Lattner79c3fe12003-10-02 04:57:28 +00001302
Misha Brukman500bc302003-11-22 00:38:41 +00001303<!-- ======================================================================= -->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +00001304<h3>
Chris Lattner26299222006-11-18 07:51:14 +00001305 <a name="x86-be">Known problems with the X86 back-end</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +00001306</h3>
John Criswellc0c186d2005-11-08 21:11:33 +00001307
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +00001308<div>
John Criswellc0c186d2005-11-08 21:11:33 +00001309
1310<ul>
Chris Lattnerc78daaf2011-11-17 01:42:23 +00001311 <li>The X86-64 backend <a href="http://llvm.org/PR1740">does not yet support
1312 the <tt>va_arg</tt> LLVM IR instruction</a>. Currently, front-ends support
1313 variadic argument constructs on X86-64 by lowering them manually.</li>
Chris Lattner26299222006-11-18 07:51:14 +00001314</ul>
1315
1316</div>
1317
1318<!-- ======================================================================= -->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +00001319<h3>
Chris Lattner26299222006-11-18 07:51:14 +00001320 <a name="ppc-be">Known problems with the PowerPC back-end</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +00001321</h3>
Chris Lattner26299222006-11-18 07:51:14 +00001322
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +00001323<div>
Chris Lattner26299222006-11-18 07:51:14 +00001324
1325<ul>
Roman Divacky223764c2011-10-30 07:49:04 +00001326 <li>The PPC32/ELF support lacks PIC support.</li>
Chris Lattner26299222006-11-18 07:51:14 +00001327</ul>
1328
1329</div>
1330
1331<!-- ======================================================================= -->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +00001332<h3>
Chris Lattnerf3e5bc62007-05-14 06:56:09 +00001333 <a name="arm-be">Known problems with the ARM back-end</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +00001334</h3>
Chris Lattnerf3e5bc62007-05-14 06:56:09 +00001335
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +00001336<div>
Chris Lattnerf3e5bc62007-05-14 06:56:09 +00001337
1338<ul>
Bill Wendling7b7fa742011-10-26 18:46:16 +00001339 <li>Thumb mode works only on ARMv6 or higher processors. On sub-ARMv6
1340 processors, thumb programs can crash or produce wrong results
1341 (<a href="http://llvm.org/PR1388">PR1388</a>).</li>
1342
1343 <li>Compilation for ARM Linux OABI (old ABI) is supported but not fully
1344 tested.</li>
Chris Lattnerf3e5bc62007-05-14 06:56:09 +00001345</ul>
1346
1347</div>
1348
1349<!-- ======================================================================= -->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +00001350<h3>
Chris Lattner26299222006-11-18 07:51:14 +00001351 <a name="sparc-be">Known problems with the SPARC back-end</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +00001352</h3>
Chris Lattner26299222006-11-18 07:51:14 +00001353
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +00001354<div>
Chris Lattner26299222006-11-18 07:51:14 +00001355
1356<ul>
Bill Wendling7b7fa742011-10-26 18:46:16 +00001357 <li>The SPARC backend only supports the 32-bit SPARC ABI (-m32); it does not
1358 support the 64-bit SPARC ABI (-m64).</li>
Chris Lattner26299222006-11-18 07:51:14 +00001359</ul>
1360
1361</div>
1362
1363<!-- ======================================================================= -->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +00001364<h3>
Bruno Cardoso Lopesb7e1a4f2008-10-25 14:56:26 +00001365 <a name="mips-be">Known problems with the MIPS back-end</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +00001366</h3>
Bruno Cardoso Lopesb7e1a4f2008-10-25 14:56:26 +00001367
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +00001368<div>
Bruno Cardoso Lopesb7e1a4f2008-10-25 14:56:26 +00001369
1370<ul>
Bill Wendling7b7fa742011-10-26 18:46:16 +00001371 <li>64-bit MIPS targets are not supported yet.</li>
Bruno Cardoso Lopesb7e1a4f2008-10-25 14:56:26 +00001372</ul>
1373
1374</div>
1375
1376<!-- ======================================================================= -->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +00001377<h3>
Chris Lattner26299222006-11-18 07:51:14 +00001378 <a name="alpha-be">Known problems with the Alpha back-end</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +00001379</h3>
Chris Lattner26299222006-11-18 07:51:14 +00001380
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +00001381<div>
Chris Lattner26299222006-11-18 07:51:14 +00001382
1383<ul>
Bill Wendling7b7fa742011-10-26 18:46:16 +00001384 <li>On 21164s, some rare FP arithmetic sequences which may trap do not have
1385 the appropriate nops inserted to ensure restartability.</li>
John Criswellc0c186d2005-11-08 21:11:33 +00001386</ul>
Bill Wendling7b7fa742011-10-26 18:46:16 +00001387
John Criswellc0c186d2005-11-08 21:11:33 +00001388</div>
1389
Chris Lattner26299222006-11-18 07:51:14 +00001390<!-- ======================================================================= -->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +00001391<h3>
Chris Lattnerf3e5bc62007-05-14 06:56:09 +00001392 <a name="c-be">Known problems with the C back-end</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +00001393</h3>
Chris Lattner26299222006-11-18 07:51:14 +00001394
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +00001395<div>
Chris Lattner26299222006-11-18 07:51:14 +00001396
Chris Lattner3016ee92010-09-29 05:34:42 +00001397<p>The C backend has numerous problems and is not being actively maintained.
Bill Wendling7b7fa742011-10-26 18:46:16 +00001398 Depending on it for anything serious is not advised.</p>
Chris Lattner3016ee92010-09-29 05:34:42 +00001399
Chris Lattner26299222006-11-18 07:51:14 +00001400<ul>
Bill Wendling7b7fa742011-10-26 18:46:16 +00001401 <li><a href="http://llvm.org/PR802">The C backend has only basic support for
1402 inline assembly code</a>.</li>
1403
1404 <li><a href="http://llvm.org/PR1658">The C backend violates the ABI of common
1405 C++ programs</a>, preventing intermixing between C++ compiled by the CBE
1406 and C++ code compiled with <tt>llc</tt> or native compilers.</li>
1407
1408 <li>The C backend does not support all exception handling constructs.</li>
1409
1410 <li>The C backend does not support arbitrary precision integers.</li>
Chris Lattner26299222006-11-18 07:51:14 +00001411</ul>
1412
1413</div>
John Criswellc0c186d2005-11-08 21:11:33 +00001414
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +00001415</div>
1416
Chris Lattner79c3fe12003-10-02 04:57:28 +00001417<!-- *********************************************************************** -->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +00001418<h2>
Misha Brukman500bc302003-11-22 00:38:41 +00001419 <a name="additionalinfo">Additional Information</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +00001420</h2>
Chris Lattner79c3fe12003-10-02 04:57:28 +00001421<!-- *********************************************************************** -->
1422
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +00001423<div>
Misha Brukman500bc302003-11-22 00:38:41 +00001424
Bill Wendling7b7fa742011-10-26 18:46:16 +00001425<p>A wide variety of additional information is available on
1426 the <a href="http://llvm.org/">LLVM web page</a>, in particular in
1427 the <a href="http://llvm.org/docs/">documentation</a> section. The web page
1428 also contains versions of the API documentation which is up-to-date with the
1429 Subversion version of the source code. You can access versions of these
1430 documents specific to this release by going into the "<tt>llvm/doc/</tt>"
1431 directory in the LLVM tree.</p>
Chris Lattner79c3fe12003-10-02 04:57:28 +00001432
Misha Brukman500bc302003-11-22 00:38:41 +00001433<p>If you have any questions or comments about LLVM, please feel free to contact
Bill Wendling7b7fa742011-10-26 18:46:16 +00001434 us via the <a href="http://llvm.org/docs/#maillist"> mailing lists</a>.</p>
Chris Lattner79c3fe12003-10-02 04:57:28 +00001435
Misha Brukman500bc302003-11-22 00:38:41 +00001436</div>
Chris Lattner79c3fe12003-10-02 04:57:28 +00001437
1438<!-- *********************************************************************** -->
Chris Lattner79c3fe12003-10-02 04:57:28 +00001439
Misha Brukman500bc302003-11-22 00:38:41 +00001440<hr>
Misha Brukman2061e892003-11-22 01:23:39 +00001441<address>
Misha Brukman38847d52003-12-21 22:53:21 +00001442 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
Misha Brukman44408702008-12-11 17:34:48 +00001443 src="http://jigsaw.w3.org/css-validator/images/vcss-blue" alt="Valid CSS"></a>
Misha Brukman2061e892003-11-22 01:23:39 +00001444 <a href="http://validator.w3.org/check/referer"><img
Misha Brukman44408702008-12-11 17:34:48 +00001445 src="http://www.w3.org/Icons/valid-html401-blue" alt="Valid HTML 4.01"></a>
Misha Brukman2061e892003-11-22 01:23:39 +00001446
Chris Lattnerb4b0ce72007-05-18 00:44:29 +00001447 <a href="http://llvm.org/">LLVM Compiler Infrastructure</a><br>
Misha Brukman500bc302003-11-22 00:38:41 +00001448 Last modified: $Date$
Misha Brukman2061e892003-11-22 01:23:39 +00001449</address>
Chris Lattner79c3fe12003-10-02 04:57:28 +00001450
Misha Brukman500bc302003-11-22 00:38:41 +00001451</body>
1452</html>