Gitiles
Code Review Sign In
gerrit-public.fairphone.software / fp2-dev / platform / external / llvm / f94c9c136b9f67bcab147a68dd65132d64d358fa / . / docs / ReleaseNotes.html
blob: 40adbc60efec6ed7955018e3c8fb859d777273cc [file] [log] [blame]
Misha Brukman500bc302003-11-22 00:38:41 +0000[diff] [blame]1<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"
2 "http://www.w3.org/TR/html4/strict.dtd">
3<html>
4<head>
Reid Spencer6454ed32004-11-18 18:38:58 +0000[diff] [blame]5 <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
Misha Brukman500bc302003-11-22 00:38:41 +0000[diff] [blame]6 <link rel="stylesheet" href="llvm.css" type="text/css">
Chris Lattnerbe2e1b52011-03-10 07:43:44 +0000[diff] [blame]7 <title>LLVM 2.9 Release Notes</title>
Misha Brukman500bc302003-11-22 00:38:41 +0000[diff] [blame]8</head>
9<body>
Chris Lattner79c3fe12003-10-02 04:57:28 +0000[diff] [blame]10
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]11<h1>LLVM 2.9 Release Notes</h1>
Mikhail Glushenkovea65d7d2008-10-13 02:08:34 +0000[diff] [blame]12
Chris Lattner0e464a92010-03-17 04:02:39 +0000[diff] [blame]13<img align=right src="http://llvm.org/img/DragonSmall.png"
Gabor Greifee2187a2010-04-22 10:21:43 +0000[diff] [blame]14 width="136" height="136" alt="LLVM Dragon Logo">
Chris Lattner0e464a92010-03-17 04:02:39 +0000[diff] [blame]15
Chris Lattner79c3fe12003-10-02 04:57:28 +0000[diff] [blame]16<ol>
Misha Brukman500bc302003-11-22 00:38:41 +0000[diff] [blame]17 <li><a href="#intro">Introduction</a></li>
Chris Lattner96a445e2008-10-13 18:01:01 +0000[diff] [blame]18 <li><a href="#subproj">Sub-project Status Update</a></li>
Chris Lattnerbe2e1b52011-03-10 07:43:44 +0000[diff] [blame]19 <li><a href="#externalproj">External Projects Using LLVM 2.9</a></li>
20 <li><a href="#whatsnew">What's New in LLVM 2.9?</a></li>
Chris Lattner4b538b92004-04-30 22:17:12 +0000[diff] [blame]21 <li><a href="GettingStarted.html">Installation Instructions</a></li>
Dan Gohman44aa9212008-10-14 16:23:02 +0000[diff] [blame]22 <li><a href="#knownproblems">Known Problems</a></li>
Misha Brukman500bc302003-11-22 00:38:41 +0000[diff] [blame]23 <li><a href="#additionalinfo">Additional Information</a></li>
Chris Lattner79c3fe12003-10-02 04:57:28 +0000[diff] [blame]24</ol>
25
Chris Lattner7911ce22004-05-23 21:07:27 +0000[diff] [blame]26<div class="doc_author">
NAKAMURA Takumib9a33632011-04-09 02:13:37 +0000[diff] [blame]27 <p>Written by the <a href="http://llvm.org/">LLVM Team</a></p>
Misha Brukman500bc302003-11-22 00:38:41 +0000[diff] [blame]28</div>
Chris Lattner79c3fe12003-10-02 04:57:28 +0000[diff] [blame]29
Chris Lattner49123fd2011-04-06 06:29:50 +0000[diff] [blame]30<!--
Chris Lattnerbe2e1b52011-03-10 07:43:44 +0000[diff] [blame]31<h1 style="color:red">These are in-progress notes for the upcoming LLVM 2.9
Jeffrey Yasskinbec48772010-01-28 01:14:43 +0000[diff] [blame]32release.<br>
33You may prefer the
Chris Lattnerbe2e1b52011-03-10 07:43:44 +0000[diff] [blame]34<a href="http://llvm.org/releases/2.8/docs/ReleaseNotes.html">LLVM 2.8
Dan Gohmanb44f6c62010-05-03 23:51:05 +0000[diff] [blame]35Release Notes</a>.</h1>
Chris Lattner49123fd2011-04-06 06:29:50 +0000[diff] [blame]36 -->
Jeffrey Yasskinbec48772010-01-28 01:14:43 +0000[diff] [blame]37
Chris Lattner79c3fe12003-10-02 04:57:28 +0000[diff] [blame]38<!-- *********************************************************************** -->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]39<h2>
Misha Brukman500bc302003-11-22 00:38:41 +0000[diff] [blame]40 <a name="intro">Introduction</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]41</h2>
Chris Lattner79c3fe12003-10-02 04:57:28 +0000[diff] [blame]42<!-- *********************************************************************** -->
43
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]44<div>
Misha Brukman500bc302003-11-22 00:38:41 +0000[diff] [blame]45
Chris Lattnerf6662f92008-10-13 17:57:36 +0000[diff] [blame]46<p>This document contains the release notes for the LLVM Compiler
Chris Lattnerbe2e1b52011-03-10 07:43:44 +0000[diff] [blame]47Infrastructure, release 2.9. Here we describe the status of LLVM, including
Chris Lattnerf6662f92008-10-13 17:57:36 +0000[diff] [blame]48major improvements from the previous release and significant known problems.
Mikhail Glushenkovf795ef02009-03-01 18:09:47 +0000[diff] [blame]49All LLVM releases may be downloaded from the <a
Chris Lattnerf6662f92008-10-13 17:57:36 +0000[diff] [blame]50href="http://llvm.org/releases/">LLVM releases web site</a>.</p>
Chris Lattner19092612003-10-02 16:38:05 +0000[diff] [blame]51
Chris Lattner7506b1d2004-12-07 08:04:13 +0000[diff] [blame]52<p>For more information about LLVM, including information about the latest
Chris Lattnerc463b272005-10-29 07:07:09 +0000[diff] [blame]53release, please check out the <a href="http://llvm.org/">main LLVM
Chris Lattner47ad72c2003-10-07 21:38:31 +0000[diff] [blame]54web site</a>. If you have questions or comments, the <a
Chris Lattnerc66bfef2010-03-17 04:41:49 +0000[diff] [blame]55href="http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev">LLVM Developer's
56Mailing List</a> is a good place to send them.</p>
Chris Lattner79c3fe12003-10-02 04:57:28 +0000[diff] [blame]57
Mikhail Glushenkovea65d7d2008-10-13 02:08:34 +0000[diff] [blame]58<p>Note that if you are reading this file from a Subversion checkout or the
Chris Lattner84977642007-09-21 03:54:09 +0000[diff] [blame]59main LLVM web page, this document applies to the <i>next</i> release, not the
Gabor Greiffa933f82008-10-14 11:00:32 +0000[diff] [blame]60current one. To see the release notes for a specific release, please see the
Chris Lattner84977642007-09-21 03:54:09 +0000[diff] [blame]61<a href="http://llvm.org/releases/">releases page</a>.</p>
Misha Brukman500bc302003-11-22 00:38:41 +0000[diff] [blame]62
63</div>
Chris Lattner61358ab2009-10-13 17:48:04 +0000[diff] [blame]64
Chris Lattnere4dc1962011-04-05 23:22:33 +0000[diff] [blame]65<!-- Features that need text if they're finished for 3.1:
66 ARM EHABI
Chris Lattner914ce462010-04-22 06:28:20 +0000[diff] [blame]67 combiner-aa?
Chris Lattner61358ab2009-10-13 17:48:04 +0000[diff] [blame]68 strong phi elim
Chris Lattner61358ab2009-10-13 17:48:04 +0000[diff] [blame]69 loop dependence analysis
Chris Lattnere07043c2010-09-29 05:30:03 +0000[diff] [blame]70 CorrelatedValuePropagation
Chris Lattnere4dc1962011-04-05 23:22:33 +0000[diff] [blame]71 lib/Transforms/IPO/MergeFunctions.cpp => consider for 3.1.
Chris Lattneracce85d2008-02-10 07:46:44 +0000[diff] [blame]72 -->
Chris Lattnere07043c2010-09-29 05:30:03 +0000[diff] [blame]73
Chris Lattner96a445e2008-10-13 18:01:01 +0000[diff] [blame]74<!-- *********************************************************************** -->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]75<h2>
Chris Lattner96a445e2008-10-13 18:01:01 +0000[diff] [blame]76 <a name="subproj">Sub-project Status Update</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]77</h2>
Chris Lattner96a445e2008-10-13 18:01:01 +0000[diff] [blame]78<!-- *********************************************************************** -->
Chris Lattnerea34f642008-06-08 21:34:41 +0000[diff] [blame]79
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]80<div>
Chris Lattner8348b472008-06-08 21:58:17 +0000[diff] [blame]81<p>
Chris Lattnerbe2e1b52011-03-10 07:43:44 +0000[diff] [blame]82The LLVM 2.9 distribution currently consists of code from the core LLVM
Chris Lattner61358ab2009-10-13 17:48:04 +0000[diff] [blame]83repository (which roughly includes the LLVM optimizers, code generators
84and supporting tools), the Clang repository and the llvm-gcc repository. In
85addition to this code, the LLVM Project includes other sub-projects that are in
86development. Here we include updates on these subprojects.
Bill Wendling63d8c552009-03-02 04:28:57 +0000[diff] [blame]87</p>
Chris Lattner96a445e2008-10-13 18:01:01 +0000[diff] [blame]88
Chris Lattner8348b472008-06-08 21:58:17 +0000[diff] [blame]89<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]90<h3>
Chris Lattnerfb97b2d2008-10-13 18:11:54 +0000[diff] [blame]91<a name="clang">Clang: C/C++/Objective-C Frontend Toolkit</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]92</h3>
Chris Lattner8348b472008-06-08 21:58:17 +0000[diff] [blame]93
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]94<div>
Chris Lattner8348b472008-06-08 21:58:17 +0000[diff] [blame]95
Chris Lattner095539f2010-04-26 17:42:18 +0000[diff] [blame]96<p><a href="http://clang.llvm.org/">Clang</a> is an LLVM front end for the C,
97C++, and Objective-C languages. Clang aims to provide a better user experience
98through expressive diagnostics, a high level of conformance to language
99standards, fast compilation, and low memory use. Like LLVM, Clang provides a
100modular, library-based architecture that makes it suitable for creating or
101integrating with other development tools. Clang is considered a
Chris Lattnere07043c2010-09-29 05:30:03 +0000[diff] [blame]102production-quality compiler for C, Objective-C, C++ and Objective-C++ on x86
Chris Lattner1efe27e2011-04-06 00:45:11 +0000[diff] [blame]103(32- and 64-bit), and for darwin/arm targets.</p>
Chris Lattner61358ab2009-10-13 17:48:04 +0000[diff] [blame]104
Chris Lattner0a6f6d52011-04-05 07:19:28 +0000[diff] [blame]105<p>In the LLVM 2.9 time-frame, the Clang team has made many improvements in C,
106C++ and Objective-C support. C++ support is now generally rock solid, has
Chris Lattnerdf448a32011-04-06 00:56:12 +0000[diff] [blame]107been exercised on a broad variety of code, and has several new <a
108href="http://clang.llvm.org/cxx_status.html#cxx0x">C++'0x features</a>
Chris Lattner0a6f6d52011-04-05 07:19:28 +0000[diff] [blame]109implemented (such as rvalue references and variadic templates). LLVM 2.9 has
110also brought in a large range of bug fixes and minor features (e.g. __label__
111support), and is much more compatible with the Linux Kernel.</p>
112
Duncan Sandsf3ba7af2011-04-06 08:07:40 +0000[diff] [blame]113<p>If Clang rejects your code but another compiler accepts it, please take a
Chris Lattner0a6f6d52011-04-05 07:19:28 +0000[diff] [blame]114look at the <a href="http://clang.llvm.org/compatibility.html">language
Duncan Sandsf3ba7af2011-04-06 08:07:40 +0000[diff] [blame]115compatibility</a> guide to make sure this is not intentional or a known issue.
Chris Lattner0a6f6d52011-04-05 07:19:28 +0000[diff] [blame]116</p>
Bill Wendling741748a2008-10-27 09:27:33 +0000[diff] [blame]117
Chris Lattnerfb97b2d2008-10-13 18:11:54 +0000[diff] [blame]118</div>
119
120<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]121<h3>
Duncan Sands528a5102011-04-04 11:09:08 +0000[diff] [blame]122<a name="dragonegg">DragonEgg: GCC front-ends, LLVM back-end</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]123</h3>
Chris Lattner61358ab2009-10-13 17:48:04 +0000[diff] [blame]124
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]125<div>
Chris Lattner61358ab2009-10-13 17:48:04 +0000[diff] [blame]126<p>
Duncan Sands528a5102011-04-04 11:09:08 +0000[diff] [blame]127<a href="http://dragonegg.llvm.org/">DragonEgg</a> is a
128<a href="http://gcc.gnu.org/wiki/plugins">gcc plugin</a> that replaces GCC's
129optimizers and code generators with LLVM's.
130Currently it requires a patched version of gcc-4.5.
131The plugin can target the x86-32 and x86-64 processor families and has been
132used successfully on the Darwin, FreeBSD and Linux platforms.
133The Ada, C, C++ and Fortran languages work well.
134The plugin is capable of compiling plenty of Obj-C, Obj-C++ and Java but it is
135not known whether the compiled code actually works or not!
Duncan Sands749fd832010-04-02 09:23:15 +0000[diff] [blame]136</p>
137
138<p>
Chris Lattnerbe2e1b52011-03-10 07:43:44 +0000[diff] [blame]139The 2.9 release has the following notable changes:
Duncan Sands4b1da2b2010-09-30 17:37:34 +0000[diff] [blame]140<ul>
Duncan Sands528a5102011-04-04 11:09:08 +0000[diff] [blame]141<li>The plugin is much more stable when compiling Fortran.</li>
Chris Lattnerdf448a32011-04-06 00:56:12 +0000[diff] [blame]142<li>Inline assembly where an asm output is tied to an input of a different size
143is now supported in many more cases.</li>
Duncan Sands528a5102011-04-04 11:09:08 +0000[diff] [blame]144<li>Basic support for the __float128 type was added. It is now possible to
145generate LLVM IR from programs using __float128 but code generation does not
146work yet.</li>
147<li>Compiling Java programs no longer systematically crashes the plugin.</li>
Duncan Sands4b1da2b2010-09-30 17:37:34 +0000[diff] [blame]148</ul>
Chris Lattner61358ab2009-10-13 17:48:04 +0000[diff] [blame]149
150</div>
151
Chris Lattner75547712010-10-03 23:49:06 +0000[diff] [blame]152<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]153<h3>
Chris Lattner75547712010-10-03 23:49:06 +0000[diff] [blame]154<a name="compiler-rt">compiler-rt: Compiler Runtime Library</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]155</h3>
Chris Lattner75547712010-10-03 23:49:06 +0000[diff] [blame]156
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]157<div>
Chris Lattner75547712010-10-03 23:49:06 +0000[diff] [blame]158<p>
159The new LLVM <a href="http://compiler-rt.llvm.org/">compiler-rt project</a>
160is a simple library that provides an implementation of the low-level
161target-specific hooks required by code generation and other runtime components.
162For example, when compiling for a 32-bit target, converting a double to a 64-bit
163unsigned integer is compiled into a runtime call to the "__fixunsdfdi"
164function. The compiler-rt library provides highly optimized implementations of
165this and other low-level routines (some are 3x faster than the equivalent
166libgcc routines).</p>
167
Chris Lattnerdf448a32011-04-06 00:56:12 +0000[diff] [blame]168<p>In the LLVM 2.9 timeframe, compiler_rt has had several minor changes for
169 better ARM support, and a fairly major license change. All of the code in the
170 compiler-rt project is now <a href="DeveloperPolicy.html#license">dual
171 licensed</a> under MIT and UIUC license, which allows you to use compiler-rt
172 in applications without the binary copyright reproduction clause. If you
173 prefer the LLVM/UIUC license, you are free to continue using it under that
174 license as well.</p>
Chris Lattner75547712010-10-03 23:49:06 +0000[diff] [blame]175
176</div>
Chris Lattner61358ab2009-10-13 17:48:04 +0000[diff] [blame]177
178<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]179<h3>
Chris Lattnere07043c2010-09-29 05:30:03 +0000[diff] [blame]180<a name="lldb">LLDB: Low Level Debugger</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]181</h3>
Chris Lattnere07043c2010-09-29 05:30:03 +0000[diff] [blame]182
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]183<div>
Chris Lattnere07043c2010-09-29 05:30:03 +0000[diff] [blame]184<p>
Chris Lattner75547712010-10-03 23:49:06 +0000[diff] [blame]185<a href="http://lldb.llvm.org/">LLDB</a> is a brand new member of the LLVM
186umbrella of projects. LLDB is a next generation, high-performance debugger. It
187is built as a set of reusable components which highly leverage existing
188libraries in the larger LLVM Project, such as the Clang expression parser, the
189LLVM disassembler and the LLVM JIT.</p>
Chris Lattnere07043c2010-09-29 05:30:03 +0000[diff] [blame]190
191<p>
Chris Lattnerdf448a32011-04-06 00:56:12 +0000[diff] [blame]192LLDB is has advanced by leaps and bounds in the 2.9 timeframe. It is
193dramatically more stable and useful, and includes both a new <a
194href="http://lldb.llvm.org/tutorial.html">tutorial</a> and a <a
195href="http://lldb.llvm.org/lldb-gdb.html">side-by-side comparison with
196GDB</a>.</p>
Chris Lattnere07043c2010-09-29 05:30:03 +0000[diff] [blame]197
198</div>
199
200<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]201<h3>
Chris Lattnere07043c2010-09-29 05:30:03 +0000[diff] [blame]202<a name="libc++">libc++: C++ Standard Library</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]203</h3>
Chris Lattnere07043c2010-09-29 05:30:03 +0000[diff] [blame]204
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]205<div>
Chris Lattnere07043c2010-09-29 05:30:03 +0000[diff] [blame]206<p>
Tobias Grossercdce44b2010-10-06 21:07:30 +0000[diff] [blame]207<a href="http://libcxx.llvm.org/">libc++</a> is another new member of the LLVM
Chris Lattner75547712010-10-03 23:49:06 +0000[diff] [blame]208family. It is an implementation of the C++ standard library, written from the
209ground up to specifically target the forthcoming C++'0X standard and focus on
210delivering great performance.</p>
Chris Lattnere07043c2010-09-29 05:30:03 +0000[diff] [blame]211
212<p>
Chris Lattner2009c492011-04-06 00:59:18 +0000[diff] [blame]213In the LLVM 2.9 timeframe, libc++ has had numerous bugs fixed, and is now being
214co-developed with Clang's C++'0x mode.</p>
Chris Lattner0a6f6d52011-04-05 07:19:28 +0000[diff] [blame]215
Chris Lattner2009c492011-04-06 00:59:18 +0000[diff] [blame]216<p>
217Like compiler_rt, libc++ is now <a href="DeveloperPolicy.html#license">dual
218 licensed</a> under the MIT and UIUC license, allowing it to be used more
219 permissively.
Chris Lattnere07043c2010-09-29 05:30:03 +0000[diff] [blame]220</p>
221
222</div>
223
Chris Lattner61358ab2009-10-13 17:48:04 +0000[diff] [blame]224
Chris Lattner3d6a80a2011-04-07 03:08:22 +0000[diff] [blame]225<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]226<h3>
Chris Lattner3d6a80a2011-04-07 03:08:22 +0000[diff] [blame]227<a name="LLBrowse">LLBrowse: IR Browser</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]228</h3>
Daniel Dunbar97b01a82010-10-04 17:39:47 +0000[diff] [blame]229
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]230<div>
Chris Lattner3d6a80a2011-04-07 03:08:22 +0000[diff] [blame]231<p>
232<a href="http://llvm.org/svn/llvm-project/llbrowse/trunk/doc/LLBrowse.html">
233 LLBrowse</a> is an interactive viewer for LLVM modules. It can load any LLVM
234 module and displays its contents as an expandable tree view, facilitating an
235 easy way to inspect types, functions, global variables, or metadata nodes. It
236 is fully cross-platform, being based on the popular wxWidgets GUI toolkit.
237</p>
238</div>
239
240<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]241<h3>
Chris Lattner3d6a80a2011-04-07 03:08:22 +0000[diff] [blame]242<a name="vmkit">VMKit</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]243</h3>
Chris Lattner3d6a80a2011-04-07 03:08:22 +0000[diff] [blame]244
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]245<div>
Chris Lattner3d6a80a2011-04-07 03:08:22 +0000[diff] [blame]246<p>The <a href="http://vmkit.llvm.org/">VMKit project</a> is an implementation
247 of a Java Virtual Machine (Java VM or JVM) that uses LLVM for static and
248 just-in-time compilation. As of LLVM 2.9, VMKit now supports generational
249 garbage collectors. The garbage collectors are provided by the MMTk framework,
250 and VMKit can be configured to use one of the numerous implemented collectors
251 of MMTk.
252</p>
253</div>
254
255
Daniel Dunbar97b01a82010-10-04 17:39:47 +0000[diff] [blame]256<!--=========================================================================-->
Chris Lattner7a8e6c52011-04-05 18:38:45 +0000[diff] [blame]257<!--
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]258<h3>
Daniel Dunbar97b01a82010-10-04 17:39:47 +0000[diff] [blame]259<a name="klee">KLEE: A Symbolic Execution Virtual Machine</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]260</h3>
Daniel Dunbar97b01a82010-10-04 17:39:47 +0000[diff] [blame]261
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]262<div>
Daniel Dunbar97b01a82010-10-04 17:39:47 +0000[diff] [blame]263<p>
264<a href="http://klee.llvm.org/">KLEE</a> is a symbolic execution framework for
265programs in LLVM bitcode form. KLEE tries to symbolically evaluate "all" paths
266through the application and records state transitions that lead to fault
267states. This allows it to construct testcases that lead to faults and can even
268be used to verify some algorithms.
269</p>
270
Chris Lattnerbe2e1b52011-03-10 07:43:44 +0000[diff] [blame]271<p>UPDATE!</p>
Chris Lattner7a8e6c52011-04-05 18:38:45 +0000[diff] [blame]272</div>-->
Daniel Dunbar97b01a82010-10-04 17:39:47 +0000[diff] [blame]273
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]274</div>
Daniel Dunbar97b01a82010-10-04 17:39:47 +0000[diff] [blame]275
Chris Lattnerab68e9e2009-02-26 22:33:38 +0000[diff] [blame]276<!-- *********************************************************************** -->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]277<h2>
Chris Lattnerbe2e1b52011-03-10 07:43:44 +0000[diff] [blame]278 <a name="externalproj">External Open Source Projects Using LLVM 2.9</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]279</h2>
Chris Lattnerab68e9e2009-02-26 22:33:38 +0000[diff] [blame]280<!-- *********************************************************************** -->
281
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]282<div>
Chris Lattner61358ab2009-10-13 17:48:04 +0000[diff] [blame]283
284<p>An exciting aspect of LLVM is that it is used as an enabling technology for
285 a lot of other language and tools projects. This section lists some of the
Chris Lattnerbe2e1b52011-03-10 07:43:44 +0000[diff] [blame]286 projects that have already been updated to work with LLVM 2.9.</p>
Chris Lattner75547712010-10-03 23:49:06 +0000[diff] [blame]287
Chris Lattner3bfe57e2011-04-06 01:13:49 +0000[diff] [blame]288<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]289<h3>Crack Programming Language</h3>
Chris Lattner3bfe57e2011-04-06 01:13:49 +0000[diff] [blame]290
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]291<div>
Chris Lattner3bfe57e2011-04-06 01:13:49 +0000[diff] [blame]292<p>
293<a href="http://code.google.com/p/crack-language/">Crack</a> aims to provide the
294ease of development of a scripting language with the performance of a compiled
295language. The language derives concepts from C++, Java and Python, incorporating
296object-oriented programming, operator overloading and strong typing.</p>
297</div>
298
299
300<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]301<h3>TTA-based Codesign Environment (TCE)</h3>
Chris Lattner3bfe57e2011-04-06 01:13:49 +0000[diff] [blame]302
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]303<div>
Chris Lattner3bfe57e2011-04-06 01:13:49 +0000[diff] [blame]304<p>TCE is a toolset for designing application-specific processors (ASP) based on
305the Transport triggered architecture (TTA). The toolset provides a complete
306co-design flow from C/C++ programs down to synthesizable VHDL and parallel
307program binaries. Processor customization points include the register files,
308function units, supported operations, and the interconnection network.</p>
309
310<p>TCE uses Clang and LLVM for C/C++ language support, target independent
311optimizations and also for parts of code generation. It generates new LLVM-based
312code generators "on the fly" for the designed TTA processors and loads them in
313to the compiler backend as runtime libraries to avoid per-target recompilation
314of larger parts of the compiler chain.</p>
315</div>
316
317
318
319<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]320<h3>PinaVM</h3>
Chris Lattner3bfe57e2011-04-06 01:13:49 +0000[diff] [blame]321
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]322<div>
Chris Lattner3bfe57e2011-04-06 01:13:49 +0000[diff] [blame]323<p><a href="http://gitorious.org/pinavm/pages/Home">PinaVM</a> is an open
324source, <a href="http://www.systemc.org/">SystemC</a> front-end. Unlike many
325other front-ends, PinaVM actually executes the elaboration of the
326program analyzed using LLVM's JIT infrastructure. It later enriches the
327bitcode with SystemC-specific information.</p>
328</div>
329
330<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]331<h3>Pure</h3>
Chris Lattner3bfe57e2011-04-06 01:13:49 +0000[diff] [blame]332
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]333<div>
Chris Lattner3bfe57e2011-04-06 01:13:49 +0000[diff] [blame]334<p><a href="http://pure-lang.googlecode.com/">Pure</a> is an
335 algebraic/functional
336 programming language based on term rewriting. Programs are collections
337 of equations which are used to evaluate expressions in a symbolic
338 fashion. The interpreter uses LLVM as a backend to JIT-compile Pure
339 programs to fast native code. Pure offers dynamic typing, eager and lazy
340 evaluation, lexical closures, a hygienic macro system (also based on
341 term rewriting), built-in list and matrix support (including list and
342 matrix comprehensions) and an easy-to-use interface to C and other
343 programming languages (including the ability to load LLVM bitcode
344 modules, and inline C, C++, Fortran and Faust code in Pure programs if
345 the corresponding LLVM-enabled compilers are installed).</p>
346
347<p>Pure version 0.47 has been tested and is known to work with LLVM 2.9
348 (and continues to work with older LLVM releases &gt;= 2.5).</p>
349</div>
350
351<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]352<h3 id="icedtea">IcedTea Java Virtual Machine Implementation</h3>
Chris Lattner3bfe57e2011-04-06 01:13:49 +0000[diff] [blame]353
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]354<div>
Chris Lattner3bfe57e2011-04-06 01:13:49 +0000[diff] [blame]355<p>
356<a href="http://icedtea.classpath.org/wiki/Main_Page">IcedTea</a> provides a
357harness to build OpenJDK using only free software build tools and to provide
358replacements for the not-yet free parts of OpenJDK. One of the extensions that
359IcedTea provides is a new JIT compiler named <a
360href="http://icedtea.classpath.org/wiki/ZeroSharkFaq">Shark</a> which uses LLVM
361to provide native code generation without introducing processor-dependent
362code.
363</p>
364
365<p> OpenJDK 7 b112, IcedTea6 1.9 and IcedTea7 1.13 and later have been tested
366and are known to work with LLVM 2.9 (and continue to work with older LLVM
367releases &gt;= 2.6 as well).</p>
368</div>
369
370<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]371<h3>Glasgow Haskell Compiler (GHC)</h3>
Chris Lattner3bfe57e2011-04-06 01:13:49 +0000[diff] [blame]372
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]373<div>
Chris Lattner3bfe57e2011-04-06 01:13:49 +0000[diff] [blame]374<p>GHC is an open source, state-of-the-art programming suite for Haskell,
375a standard lazy functional programming language. It includes an
376optimizing static compiler generating good code for a variety of
377platforms, together with an interactive system for convenient, quick
378development.</p>
379
380<p>In addition to the existing C and native code generators, GHC 7.0 now
381supports an LLVM code generator. GHC supports LLVM 2.7 and later.</p>
382</div>
383
384<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]385<h3>Polly - Polyhedral optimizations for LLVM</h3>
Chris Lattner3bfe57e2011-04-06 01:13:49 +0000[diff] [blame]386
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]387<div>
Chris Lattner3bfe57e2011-04-06 01:13:49 +0000[diff] [blame]388<p>Polly is a project that aims to provide advanced memory access optimizations
389to better take advantage of SIMD units, cache hierarchies, multiple cores or
390even vector accelerators for LLVM. Built around an abstract mathematical
391description based on Z-polyhedra, it provides the infrastructure to develop
392advanced optimizations in LLVM and to connect complex external optimizers. In
393its first year of existence Polly already provides an exact value-based
394dependency analysis as well as basic SIMD and OpenMP code generation support.
395Furthermore, Polly can use PoCC(Pluto) an advanced optimizer for data-locality
396and parallelism.</p>
397</div>
Chris Lattner75547712010-10-03 23:49:06 +0000[diff] [blame]398
Chris Lattner0fa5da92011-04-06 16:14:25 +0000[diff] [blame]399<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]400<h3>Rubinius</h3>
Chris Lattner0fa5da92011-04-06 16:14:25 +0000[diff] [blame]401
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]402<div>
Chris Lattner0fa5da92011-04-06 16:14:25 +0000[diff] [blame]403 <p><a href="http://github.com/evanphx/rubinius">Rubinius</a> is an environment
404 for running Ruby code which strives to write as much of the implementation in
405 Ruby as possible. Combined with a bytecode interpreting VM, it uses LLVM to
406 optimize and compile ruby code down to machine code. Techniques such as type
407 feedback, method inlining, and deoptimization are all used to remove dynamism
408 from ruby execution and increase performance.</p>
409</div>
410
411
Chris Lattnera844a3e2011-04-07 03:09:21 +0000[diff] [blame]412<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]413<h3>
Chris Lattnera844a3e2011-04-07 03:09:21 +0000[diff] [blame]414<a name="FAUST">FAUST Real-Time Audio Signal Processing Language</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]415</h3>
Chris Lattnere0518442010-10-01 06:34:49 +0000[diff] [blame]416
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]417<div>
Chris Lattnera844a3e2011-04-07 03:09:21 +0000[diff] [blame]418<p>
419<a href="http://faust.grame.fr">FAUST</a> is a compiled language for real-time
420audio signal processing. The name FAUST stands for Functional AUdio STream. Its
421programming model combines two approaches: functional programming and block
422diagram composition. In addition with the C, C++, JAVA output formats, the
423Faust compiler can now generate LLVM bitcode, and works with LLVM 2.7-2.9.</p>
424
425</div>
426
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]427</div>
428
Chris Lattner8348b472008-06-08 21:58:17 +0000[diff] [blame]429<!-- *********************************************************************** -->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]430<h2>
Chris Lattnerbe2e1b52011-03-10 07:43:44 +0000[diff] [blame]431 <a name="whatsnew">What's New in LLVM 2.9?</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]432</h2>
Chris Lattner8348b472008-06-08 21:58:17 +0000[diff] [blame]433<!-- *********************************************************************** -->
434
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]435<div>
Chris Lattnerf8e0b4e2008-06-08 22:59:35 +0000[diff] [blame]436
Chris Lattner61358ab2009-10-13 17:48:04 +0000[diff] [blame]437<p>This release includes a huge number of bug fixes, performance tweaks and
Chris Lattnerf6662f92008-10-13 17:57:36 +0000[diff] [blame]438minor improvements. Some of the major improvements and new features are listed
439in this section.
Chris Lattnerf8e0b4e2008-06-08 22:59:35 +0000[diff] [blame]440</p>
Chris Lattner61358ab2009-10-13 17:48:04 +0000[diff] [blame]441
Chris Lattner914ce462010-04-22 06:28:20 +0000[diff] [blame]442<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]443<h3>
Chris Lattner252b83d2008-02-06 18:00:06 +0000[diff] [blame]444<a name="majorfeatures">Major New Features</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]445</h3>
Chris Lattner252b83d2008-02-06 18:00:06 +0000[diff] [blame]446
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]447<div>
Chris Lattner252b83d2008-02-06 18:00:06 +0000[diff] [blame]448
Chris Lattnerbe2e1b52011-03-10 07:43:44 +0000[diff] [blame]449<p>LLVM 2.9 includes several major new capabilities:</p>
Chris Lattnerf3e5bc62007-05-14 06:56:09 +0000[diff] [blame]450
451<ul>
Chris Lattner1efe27e2011-04-06 00:45:11 +0000[diff] [blame]452
Chris Lattnerc5ac61d2011-04-06 05:50:04 +0000[diff] [blame]453<li>Type Based Alias Analysis (TBAA) is now implemented and turned on by default
454 in Clang. This allows substantially better load/store optimization in some
455 cases. TBAA can be disabled by passing -fno-strict-aliasing.
Chris Lattner1efe27e2011-04-06 00:45:11 +0000[diff] [blame]456</li>
457
Chris Lattnerc5ac61d2011-04-06 05:50:04 +0000[diff] [blame]458<li>This release has seen a continued focus on quality of debug information.
459 LLVM now generates much higher fidelity debug information, particularly when
460 debugging optimized code.</li>
461
462<li>Inline assembly now supports multiple alternative constraints.</li>
463
464<li>A new backend for the NVIDIA PTX virtual ISA (used to target its GPUs) is
465 under rapid development. It is not generally useful in 2.9, but is making
466 rapid progress.</li>
Chris Lattner7a8e6c52011-04-05 18:38:45 +0000[diff] [blame]467
Chris Lattner8170c102008-02-10 08:18:42 +0000[diff] [blame]468</ul>
Chris Lattner0a6f6d52011-04-05 07:19:28 +0000[diff] [blame]469
Chris Lattnerf3e5bc62007-05-14 06:56:09 +0000[diff] [blame]470</div>
471
Chris Lattner84977642007-09-21 03:54:09 +0000[diff] [blame]472<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]473<h3>
Chris Lattner917cc712009-03-02 02:37:32 +0000[diff] [blame]474<a name="coreimprovements">LLVM IR and Core Improvements</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]475</h3>
Chris Lattnerf304ffc2008-02-10 08:17:19 +0000[diff] [blame]476
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]477<div>
Chris Lattner61358ab2009-10-13 17:48:04 +0000[diff] [blame]478<p>LLVM IR has several new features for better support of new targets and that
479expose new optimization opportunities:</p>
Chris Lattnerf304ffc2008-02-10 08:17:19 +0000[diff] [blame]480
Chris Lattner791f77b2008-06-05 06:25:56 +0000[diff] [blame]481<ul>
Chris Lattnerc5ac61d2011-04-06 05:50:04 +0000[diff] [blame]482<li>The <a href="LangRef.html#bitwiseops">udiv, ashr, lshr, and shl</a>
483 instructions now have support exact and nuw/nsw bits to indicate that they
484 don't overflow or shift out bits. This is useful for optimization of <a
485 href="http://llvm.org/PR8862">pointer differences</a> and other cases.</li>
Chris Lattner1efe27e2011-04-06 00:45:11 +0000[diff] [blame]486
Chris Lattnerc5ac61d2011-04-06 05:50:04 +0000[diff] [blame]487<li>LLVM IR now supports the <a href="LangRef.html#globalvars">unnamed_addr</a>
488 attribute to indicate that constant global variables with identical
489 initializers can be merged. This fixed <a href="http://llvm.org/PR8927">an
490 issue</a> where LLVM would incorrectly merge two globals which were supposed
491 to have distinct addresses.</li>
492
493<li>The new <a href="LangRef.html#fnattrs">hotpatch attribute</a> has been added
494 to allow runtime patching of functions.</li>
Chris Lattnerf304ffc2008-02-10 08:17:19 +0000[diff] [blame]495</ul>
Mikhail Glushenkovea65d7d2008-10-13 02:08:34 +0000[diff] [blame]496
Chris Lattnerf304ffc2008-02-10 08:17:19 +0000[diff] [blame]497</div>
498
499<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]500<h3>
Chris Lattneracce85d2008-02-10 07:46:44 +0000[diff] [blame]501<a name="optimizer">Optimizer Improvements</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]502</h3>
Chris Lattneracce85d2008-02-10 07:46:44 +0000[diff] [blame]503
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]504<div>
Chris Lattneracce85d2008-02-10 07:46:44 +0000[diff] [blame]505
Chris Lattner61358ab2009-10-13 17:48:04 +0000[diff] [blame]506<p>In addition to a large array of minor performance tweaks and bug fixes, this
Chris Lattnerf3013872008-10-13 21:50:36 +0000[diff] [blame]507release includes a few major enhancements and additions to the optimizers:</p>
Chris Lattneracce85d2008-02-10 07:46:44 +0000[diff] [blame]508
509<ul>
Chris Lattnerc5ac61d2011-04-06 05:50:04 +0000[diff] [blame]510<li>Link Time Optimization (LTO) has been improved to use MC for parsing inline
511 assembly and now can build large programs like Firefox 4 on both Mac OS X and
512 Linux.</li>
Chris Lattner1efe27e2011-04-06 00:45:11 +0000[diff] [blame]513
Chris Lattnerc5ac61d2011-04-06 05:50:04 +0000[diff] [blame]514<li>The new -loop-idiom pass recognizes memset/memcpy loops (and memset_pattern
515 on darwin), turning them into library calls, which are typically better
516 optimized than inline code. If you are building a libc and notice that your
517 memcpy and memset functions are compiled into infinite recursion, please build
518 with -ffreestanding or -fno-builtin to disable this pass.</li>
Chris Lattner1efe27e2011-04-06 00:45:11 +0000[diff] [blame]519
Chris Lattnerc5ac61d2011-04-06 05:50:04 +0000[diff] [blame]520<li>A new -early-cse pass does a fast pass over functions to fold constants,
521 simplify expressions, perform simple dead store elimination, and perform
522 common subexpression elimination. It does a good job at catching some of the
523 trivial redundancies that exist in unoptimized code, making later passes more
Roman Divacky77b42e92011-04-06 19:12:21 +0000[diff] [blame]524 effective.</li>
Chris Lattner1efe27e2011-04-06 00:45:11 +0000[diff] [blame]525
Chris Lattnerc5ac61d2011-04-06 05:50:04 +0000[diff] [blame]526<li>A new -loop-instsimplify pass is used to clean up loop bodies in the loop
527 optimizer.</li>
Chris Lattner1efe27e2011-04-06 00:45:11 +0000[diff] [blame]528
Chris Lattnerc5ac61d2011-04-06 05:50:04 +0000[diff] [blame]529<li>The new TargetLibraryInfo interface allows mid-level optimizations to know
530 whether the current target's runtime library has certain functions. For
531 example, the optimizer can now transform integer-only printf calls to call
532 iprintf, allowing reduced code size for embedded C libraries (e.g. newlib).
533</li>
534
535<li>LLVM has a new <a href="WritingAnLLVMPass.html#RegionPass">RegionPass</a>
536 infrastructure for region-based optimizations.</li>
537
538<li>Several optimizer passes have been substantially sped up:
539 GVN is much faster on functions with deep dominator trees and lots of basic
540 blocks. The dominator tree and dominance frontier passes are much faster to
541 compute, and preserved by more passes (so they are computed less often). The
542 -scalar-repl pass is also much faster and doesn't use DominanceFrontier.
543</li>
544
545<li>The Dead Store Elimination pass is more aggressive optimizing stores of
546 different types: e.g. a large store following a small one to the same address.
547 The MemCpyOptimizer pass handles several new forms of memcpy elimination.</li>
Chris Lattner1efe27e2011-04-06 00:45:11 +0000[diff] [blame]548
Chris Lattnerc5ac61d2011-04-06 05:50:04 +0000[diff] [blame]549<li>LLVM now optimizes various idioms for overflow detection into check of the
550 flag register on various CPUs. For example, we now compile:
Chris Lattner1efe27e2011-04-06 00:45:11 +0000[diff] [blame]551
Chris Lattnerc5ac61d2011-04-06 05:50:04 +0000[diff] [blame]552 <pre>
Chris Lattner1efe27e2011-04-06 00:45:11 +0000[diff] [blame]553 unsigned long t = a+b;
554 if (t &lt; a) ...
Chris Lattnerc5ac61d2011-04-06 05:50:04 +0000[diff] [blame]555 </pre>
Chris Lattner1efe27e2011-04-06 00:45:11 +0000[diff] [blame]556 into:
Chris Lattnerc5ac61d2011-04-06 05:50:04 +0000[diff] [blame]557 <pre>
558 addq %rdi, %rbx
559 jno LBB0_2
560 </pre>
561</li>
Chris Lattner1efe27e2011-04-06 00:45:11 +0000[diff] [blame]562
Chris Lattner11b66112010-10-04 02:42:39 +0000[diff] [blame]563</ul>
564
Chris Lattner61358ab2009-10-13 17:48:04 +0000[diff] [blame]565</div>
Chris Lattner917cc712009-03-02 02:37:32 +0000[diff] [blame]566
Chris Lattner0b832202008-06-08 02:45:07 +0000[diff] [blame]567<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]568<h3>
Chris Lattner4ba2b652010-09-30 16:31:33 +0000[diff] [blame]569<a name="mc">MC Level Improvements</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]570</h3>
Chris Lattner4ba2b652010-09-30 16:31:33 +0000[diff] [blame]571
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]572<div>
Chris Lattner4ba2b652010-09-30 16:31:33 +0000[diff] [blame]573<p>
Chris Lattner7d9b6b42010-10-02 21:59:30 +0000[diff] [blame]574The LLVM Machine Code (aka MC) subsystem was created to solve a number
Chris Lattner4ba2b652010-09-30 16:31:33 +0000[diff] [blame]575of problems in the realm of assembly, disassembly, object file format handling,
576and a number of other related areas that CPU instruction-set level tools work
Chris Lattner7d9b6b42010-10-02 21:59:30 +0000[diff] [blame]577in.</p>
578
Chris Lattner7d9b6b42010-10-02 21:59:30 +0000[diff] [blame]579<ul>
Chris Lattnerc5ac61d2011-04-06 05:50:04 +0000[diff] [blame]580<li>ELF MC support has matured enough for the integrated assembler to be turned
581 on by default in Clang on X86-32 and X86-64 ELF systems.</li>
582
583<li>MC supports and CodeGen uses the <tt>.file</tt> and <tt>.loc</tt> directives
584 for producing line number debug info. This produces more compact line
585 tables and easier to read .s files.</li>
586
587<li>MC supports the <tt>.cfi_*</tt> directives for producing DWARF
Rafael Espindolaa26f36c2011-03-18 04:07:44 +0000[diff] [blame]588 frame information, but it is still not used by CodeGen by default.</li>
Chris Lattnerc5ac61d2011-04-06 05:50:04 +0000[diff] [blame]589
Chris Lattner1efe27e2011-04-06 00:45:11 +0000[diff] [blame]590
Chris Lattnerc5ac61d2011-04-06 05:50:04 +0000[diff] [blame]591<li>The MC assembler now generates much better diagnostics for common errors,
Chris Lattner1efe27e2011-04-06 00:45:11 +0000[diff] [blame]592 is much faster at matching instructions, is much more bug-compatible with
593 the GAS assembler, and is now generally useful for a broad range of X86
Chris Lattnerc5ac61d2011-04-06 05:50:04 +0000[diff] [blame]594 assembly.</li>
Chris Lattner1efe27e2011-04-06 00:45:11 +0000[diff] [blame]595
Chris Lattnerc5ac61d2011-04-06 05:50:04 +0000[diff] [blame]596<li>We now have some basic <a href="CodeGenerator.html#mc">internals
597 documentation</a> for MC.</li>
Chris Lattner1efe27e2011-04-06 00:45:11 +0000[diff] [blame]598
Chris Lattnerc5ac61d2011-04-06 05:50:04 +0000[diff] [blame]599<li>.td files can now specify assembler aliases directly with the <a
600 href="CodeGenerator.html#na_instparsing">MnemonicAlias and InstAlias</a>
601 tblgen classes.</li>
Chris Lattner1efe27e2011-04-06 00:45:11 +0000[diff] [blame]602
Chris Lattnerc5ac61d2011-04-06 05:50:04 +0000[diff] [blame]603<li>LLVM now has an experimental format-independent object file manipulation
604 library (lib/Object). It supports both PE/COFF and ELF. The llvm-nm tool has
605 been extended to work with native object files, and the new llvm-objdump tool
606 supports disassembly of object files (but no relocations are displayed yet).
607</li>
Chris Lattner1efe27e2011-04-06 00:45:11 +0000[diff] [blame]608
Chris Lattnerc5ac61d2011-04-06 05:50:04 +0000[diff] [blame]609<li>Win32 PE-COFF support in the MC assembler has made a lot of progress in the
610 2.9 timeframe, but is still not generally useful.</li>
Chris Lattner1efe27e2011-04-06 00:45:11 +0000[diff] [blame]611
Chris Lattner7d9b6b42010-10-02 21:59:30 +0000[diff] [blame]612</ul>
613
614<p>For more information, please see the <a
Chris Lattner4ba2b652010-09-30 16:31:33 +0000[diff] [blame]615href="http://blog.llvm.org/2010/04/intro-to-llvm-mc-project.html">Intro to the
616LLVM MC Project Blog Post</a>.
617</p>
618
NAKAMURA Takumi45c435a2011-04-05 08:24:22 +0000[diff] [blame]619</div>
Chris Lattner4ba2b652010-09-30 16:31:33 +0000[diff] [blame]620
Chris Lattner4ba2b652010-09-30 16:31:33 +0000[diff] [blame]621<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]622<h3>
Chris Lattner511433e2009-03-02 03:24:11 +0000[diff] [blame]623<a name="codegen">Target Independent Code Generator Improvements</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]624</h3>
Chris Lattner0b832202008-06-08 02:45:07 +0000[diff] [blame]625
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]626<div>
Chris Lattner0b832202008-06-08 02:45:07 +0000[diff] [blame]627
Mikhail Glushenkovf795ef02009-03-01 18:09:47 +0000[diff] [blame]628<p>We have put a significant amount of work into the code generator
629infrastructure, which allows us to implement more aggressive algorithms and make
630it run faster:</p>
Chris Lattner0b832202008-06-08 02:45:07 +0000[diff] [blame]631
632<ul>
Chris Lattner1efe27e2011-04-06 00:45:11 +0000[diff] [blame]633<li>The pre-register-allocation (preRA) instruction scheduler models register
634 pressure much more accurately in some cases. This allows the adoption of more
Chris Lattner49123fd2011-04-06 06:29:50 +0000[diff] [blame]635 aggressive scheduling heuristics without causing spills to be generated.
Chris Lattner1efe27e2011-04-06 00:45:11 +0000[diff] [blame]636</li>
637
Chris Lattner49123fd2011-04-06 06:29:50 +0000[diff] [blame]638<li>LiveDebugVariables is a new pass that keeps track of debugging information
639 for user variables that are promoted to registers in optimized builds.</li>
Chris Lattner1efe27e2011-04-06 00:45:11 +0000[diff] [blame]640
Chris Lattner49123fd2011-04-06 06:29:50 +0000[diff] [blame]641<li>The scheduler now models operand latency and pipeline forwarding.</li>
Chris Lattner1efe27e2011-04-06 00:45:11 +0000[diff] [blame]642
Chris Lattner49123fd2011-04-06 06:29:50 +0000[diff] [blame]643<li>A major register allocator infrastructure rewrite is underway. It is not on
644 by default for 2.9 and you are not advised to use it, but it has made
645 substantial progress in the 2.9 timeframe:
646 <ul>
647 <li>A new -regalloc=basic "basic" register allocator can be used as a simple
648 fallback when debugging. It uses the new infrastructure.</li>
649 <li>New infrastructure is in place for live range splitting. "SplitKit" can
650 break a live interval into smaller pieces while preserving SSA form, and
651 SpillPlacement can help find the best split points. This is a work in
652 progress so the API is changing quickly.</li>
653 <li>The inline spiller has learned to clean up after live range splitting. It
654 can hoist spills out of loops, and it can eliminate redundant spills.</li>
655 <li>Rematerialization works with live range splitting.</li>
656 <li>The new "greedy" register allocator using live range splitting. This will
657 be the default register allocator in the next LLVM release, but it is not
658 turned on by default in 2.9.</li>
659 </ul>
660</li>
Chris Lattner84977642007-09-21 03:54:09 +0000[diff] [blame]661</ul>
Chris Lattner84977642007-09-21 03:54:09 +0000[diff] [blame]662</div>
663
Chris Lattner917cc712009-03-02 02:37:32 +0000[diff] [blame]664<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]665<h3>
Chris Lattner917cc712009-03-02 02:37:32 +0000[diff] [blame]666<a name="x86">X86-32 and X86-64 Target Improvements</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]667</h3>
Chris Lattner917cc712009-03-02 02:37:32 +0000[diff] [blame]668
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]669<div>
Chris Lattner7714c912010-10-04 04:39:25 +0000[diff] [blame]670<p>New features and major changes in the X86 target include:
Chris Lattner917cc712009-03-02 02:37:32 +0000[diff] [blame]671</p>
672
673<ul>
Chris Lattner49123fd2011-04-06 06:29:50 +0000[diff] [blame]674<li>LLVM 2.9 includes a complete reimplementation of the MMX instruction set.
675 The reimplementation uses a new LLVM IR <a
676 href="LangRef.html#t_x86mmx">x86_mmx</a> type to ensure that MMX operations
677 are <em>only</em> generated from source that uses MMX builtin operations. With
Duncan Sandsf3ba7af2011-04-06 08:07:40 +0000[diff] [blame]678 this, random types like &lt;2 x i32&gt; are not turned into MMX operations
Chris Lattner49123fd2011-04-06 06:29:50 +0000[diff] [blame]679 (which can be catastrophic without proper "emms" insertion). Because the X86
680 code generator always generates reliable code, the -disable-mmx flag is now
681 removed.
Chris Lattner1efe27e2011-04-06 00:45:11 +0000[diff] [blame]682</li>
683
Chris Lattner49123fd2011-04-06 06:29:50 +0000[diff] [blame]684<li>X86 support for FS/GS relative loads and stores using <a
Jay Foadcb88ec32011-04-06 07:55:30 +0000[diff] [blame]685 href="CodeGenerator.html#x86_memory">address space 256/257</a> works reliably
Chris Lattner49123fd2011-04-06 06:29:50 +0000[diff] [blame]686 now.</li>
Chris Lattner1efe27e2011-04-06 00:45:11 +0000[diff] [blame]687
Chris Lattner49123fd2011-04-06 06:29:50 +0000[diff] [blame]688<li>LLVM 2.9 generates much better code in several cases by using adc/sbb to
689 avoid generation of conditional move instructions for conditional increment
690 and other idioms.</li>
Chris Lattner1efe27e2011-04-06 00:45:11 +0000[diff] [blame]691
Chris Lattner49123fd2011-04-06 06:29:50 +0000[diff] [blame]692<li>The X86 backend has adopted a new preRA scheduling mode, "list-ilp", to
693 shorten the height of instruction schedules without inducing register spills.
694</li>
Chris Lattner1efe27e2011-04-06 00:45:11 +0000[diff] [blame]695
Jay Foadcb88ec32011-04-06 07:55:30 +0000[diff] [blame]696<li>The MC assembler supports 3dNow! and 3DNowA instructions.</li>
Chris Lattner1efe27e2011-04-06 00:45:11 +0000[diff] [blame]697
Chris Lattner49123fd2011-04-06 06:29:50 +0000[diff] [blame]698<li>Several bugs have been fixed for Windows x64 code generator.</li>
Chad Rosierf94c9c12011-05-27 20:13:10 +0000[diff] [blame^]699
700<li>The CRC32 intrinsics have been renamed. The intrinsics were previously
701 @llvm.x86.sse42.crc32.[8|16|32] and @llvm.x86.sse42.crc64.[8|64]. They have
702 been renamed to @llvm.x86.sse42.crc32.32.[8|16|32] and
703 @llvm.x86.sse42.crc32.64.[8|64].</li>
704
Chris Lattner917cc712009-03-02 02:37:32 +0000[diff] [blame]705</ul>
706
Chris Lattner917cc712009-03-02 02:37:32 +0000[diff] [blame]707</div>
Chris Lattner84977642007-09-21 03:54:09 +0000[diff] [blame]708
709<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]710<h3>
Chris Lattner61358ab2009-10-13 17:48:04 +0000[diff] [blame]711<a name="ARM">ARM Target Improvements</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]712</h3>
Chris Lattnerc441fb82009-03-01 02:30:21 +0000[diff] [blame]713
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]714<div>
Chris Lattner61358ab2009-10-13 17:48:04 +0000[diff] [blame]715<p>New features of the ARM target include:
716</p>
Chris Lattnerc441fb82009-03-01 02:30:21 +0000[diff] [blame]717
718<ul>
Chris Lattner49123fd2011-04-06 06:29:50 +0000[diff] [blame]719<li>The ARM backend now has a fast instruction selector, which dramatically
720 improves -O0 compile times.</li>
721<li>The ARM backend has new tuning for Cortex-A8 and Cortex-A9 CPUs.</li>
722<li>The __builtin_prefetch builtin (and llvm.prefetch intrinsic) is compiled
723 into prefetch instructions instead of being discarded.</li>
Chris Lattner1efe27e2011-04-06 00:45:11 +0000[diff] [blame]724
725<li> The ARM backend preRA scheduler now models machine resources at cycle
726 granularity. This allows the scheduler to both accurately model
727 instruction latency and avoid overcommitting functional units.</li>
728
Chris Lattner49123fd2011-04-06 06:29:50 +0000[diff] [blame]729<li>Countless ARM microoptimizations have landed in LLVM 2.9.</li>
Bob Wilsone8472772010-09-13 17:39:35 +0000[diff] [blame]730</ul>
Chris Lattner61358ab2009-10-13 17:48:04 +0000[diff] [blame]731</div>
Chris Lattner1efe27e2011-04-06 00:45:11 +0000[diff] [blame]732
733<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]734<h3>
Chris Lattner1efe27e2011-04-06 00:45:11 +0000[diff] [blame]735<a name="OtherTS">Other Target Specific Improvements</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]736</h3>
Chris Lattnerc441fb82009-03-01 02:30:21 +0000[diff] [blame]737
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]738<div>
Chris Lattner1efe27e2011-04-06 00:45:11 +0000[diff] [blame]739<ul>
Chris Lattner49123fd2011-04-06 06:29:50 +0000[diff] [blame]740<li>MicroBlaze: major updates for aggressive delay slot filler, MC-based
741 assembly printing, assembly instruction parsing, ELF .o file emission, and MC
742 instruction disassembler have landed.</li>
743
744<li>SPARC: Many improvements, including using the Y registers for
745 multiplications and addition of a simple delay slot filler.</li>
746
747<li>PowerPC: The backend has been largely MC'ized and is ready to support
Duncan Sandsf3ba7af2011-04-06 08:07:40 +0000[diff] [blame]748 directly writing out mach-o object files. No one seems interested in finishing
Chris Lattner49123fd2011-04-06 06:29:50 +0000[diff] [blame]749 this final step though.</li>
Bruno Cardoso Lopes3dcac712011-04-08 03:06:22 +0000[diff] [blame]750
751<li>Mips: Improved o32 ABI support, including better varags handling.
752More instructions supported in codegen: madd, msub, rotr, rotrv and clo.
753It also now supports lowering block addresses.</li>
754
Chris Lattner1efe27e2011-04-06 00:45:11 +0000[diff] [blame]755</ul>
756</div>
Chris Lattner77d29b12008-06-05 08:02:49 +0000[diff] [blame]757
758<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]759<h3>
Chris Lattnerf6662f92008-10-13 17:57:36 +0000[diff] [blame]760<a name="changes">Major Changes and Removed Features</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]761</h3>
Chris Lattnerf6662f92008-10-13 17:57:36 +0000[diff] [blame]762
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]763<div>
Chris Lattnerf6662f92008-10-13 17:57:36 +0000[diff] [blame]764
Chris Lattnereeb4da02008-10-13 22:06:31 +0000[diff] [blame]765<p>If you're already an LLVM user or developer with out-of-tree changes based
Chris Lattnerbe2e1b52011-03-10 07:43:44 +0000[diff] [blame]766on LLVM 2.8, this section lists some "gotchas" that you may run into upgrading
Chris Lattnereeb4da02008-10-13 22:06:31 +0000[diff] [blame]767from the previous release.</p>
Chris Lattnerf6662f92008-10-13 17:57:36 +0000[diff] [blame]768
769<ul>
Chris Lattner49123fd2011-04-06 06:29:50 +0000[diff] [blame]770<li><b>This is the last release to support the llvm-gcc frontend.</b></li>
Chris Lattner1efe27e2011-04-06 00:45:11 +0000[diff] [blame]771
Chris Lattner49123fd2011-04-06 06:29:50 +0000[diff] [blame]772<li>LLVM has a new <a href="CodingStandards.html#ll_naming">naming
773 convention standard</a>, though the codebase hasn't fully adopted it yet.</li>
Chris Lattnerc5ac61d2011-04-06 05:50:04 +0000[diff] [blame]774
Chris Lattner49123fd2011-04-06 06:29:50 +0000[diff] [blame]775<li>The new DIBuilder class provides a simpler interface for front ends to
776 encode debug info in LLVM IR, and has replaced DIFactory.</li>
Chris Lattnerc5ac61d2011-04-06 05:50:04 +0000[diff] [blame]777
Chris Lattner49123fd2011-04-06 06:29:50 +0000[diff] [blame]778<li>LLVM IR and other tools always work on normalized target triples (which have
779 been run through <tt>Triple::normalize</tt>).</li>
780
781<li>The target triple x86_64--mingw64 is obsoleted. Use x86_64--mingw32
782 instead.</li>
783
784<li>The PointerTracking pass has been removed from mainline, and moved to The
785 ClamAV project (its only client).</li>
Chris Lattner1efe27e2011-04-06 00:45:11 +0000[diff] [blame]786
Chris Lattner49123fd2011-04-06 06:29:50 +0000[diff] [blame]787<li>The LoopIndexSplit, LiveValues, SimplifyHalfPowrLibCalls, GEPSplitter, and
788 PartialSpecialization passes were removed. They were unmaintained,
Duncan Sandsf3ba7af2011-04-06 08:07:40 +0000[diff] [blame]789 buggy, or deemed to be a bad idea.</li>
Devang Patelb34dd132008-10-14 20:03:43 +0000[diff] [blame]790</ul>
Chris Lattnerf6662f92008-10-13 17:57:36 +0000[diff] [blame]791
Chris Lattnerf6662f92008-10-13 17:57:36 +0000[diff] [blame]792</div>
793
Daniel Dunbarf0233c62010-10-04 20:11:41 +0000[diff] [blame]794<!--=========================================================================-->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]795<h3>
Chris Lattner1efe27e2011-04-06 00:45:11 +0000[diff] [blame]796<a name="api_changes">Internal API Changes</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]797</h3>
Daniel Dunbarf0233c62010-10-04 20:11:41 +0000[diff] [blame]798
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]799<div>
Daniel Dunbarf0233c62010-10-04 20:11:41 +0000[diff] [blame]800
Chris Lattner1efe27e2011-04-06 00:45:11 +0000[diff] [blame]801<p>In addition, many APIs have changed in this release. Some of the major
802 LLVM API changes are:</p>
Daniel Dunbarf0233c62010-10-04 20:11:41 +0000[diff] [blame]803
804<ul>
Chris Lattner49123fd2011-04-06 06:29:50 +0000[diff] [blame]805<li>include/llvm/System merged into include/llvm/Support.</li>
806<li>The <a href="http://llvm.org/PR5207">llvm::APInt API</a> was significantly
807 cleaned up.</li>
Chris Lattner1efe27e2011-04-06 00:45:11 +0000[diff] [blame]808
Chris Lattner49123fd2011-04-06 06:29:50 +0000[diff] [blame]809<li>In the code generator, MVT::Flag was renamed to MVT::Glue to more accurately
810 describe its behavior.</li>
811
812<li>The system_error header from C++0x was added, and is now pervasively used to
813 capture and handle i/o and other errors in LLVM.</li>
814
815<li>The old sys::Path API has been deprecated in favor of the new PathV2 API,
816 which is more efficient and flexible.</li>
Daniel Dunbarf0233c62010-10-04 20:11:41 +0000[diff] [blame]817</ul>
818</div>
Chris Lattnerf6662f92008-10-13 17:57:36 +0000[diff] [blame]819
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]820</div>
821
Chris Lattner79c3fe12003-10-02 04:57:28 +0000[diff] [blame]822<!-- *********************************************************************** -->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]823<h2>
Misha Brukman500bc302003-11-22 00:38:41 +0000[diff] [blame]824 <a name="knownproblems">Known Problems</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]825</h2>
Chris Lattner79c3fe12003-10-02 04:57:28 +0000[diff] [blame]826<!-- *********************************************************************** -->
827
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]828<div>
Misha Brukman500bc302003-11-22 00:38:41 +0000[diff] [blame]829
Mikhail Glushenkovf795ef02009-03-01 18:09:47 +0000[diff] [blame]830<p>This section contains significant known problems with the LLVM system,
Chris Lattnere18b32e2008-11-10 05:40:34 +0000[diff] [blame]831listed by component. If you run into a problem, please check the <a
Chris Lattnerc463b272005-10-29 07:07:09 +0000[diff] [blame]832href="http://llvm.org/bugs/">LLVM bug database</a> and submit a bug if
Chris Lattner5eccca42003-12-12 21:22:16 +0000[diff] [blame]833there isn't already one.</p>
Chris Lattner79c3fe12003-10-02 04:57:28 +0000[diff] [blame]834
Chris Lattnerf5ee1702004-03-14 02:03:02 +0000[diff] [blame]835<!-- ======================================================================= -->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]836<h3>
Chris Lattnerf5ee1702004-03-14 02:03:02 +0000[diff] [blame]837 <a name="experimental">Experimental features included with this release</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]838</h3>
Chris Lattner79c3fe12003-10-02 04:57:28 +0000[diff] [blame]839
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]840<div>
Chris Lattnerf5ee1702004-03-14 02:03:02 +0000[diff] [blame]841
Misha Brukman6df9e2c2004-05-12 21:46:05 +0000[diff] [blame]842<p>The following components of this LLVM release are either untested, known to
843be broken or unreliable, or are in early development. These components should
844not be relied on, and bugs should not be filed against them, but they may be
845useful to some people. In particular, if you would like to work on one of these
Chris Lattner2b659ef2008-02-12 06:29:45 +0000[diff] [blame]846components, please contact us on the <a
847href="http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev">LLVMdev list</a>.</p>
Chris Lattnerf5ee1702004-03-14 02:03:02 +0000[diff] [blame]848
849<ul>
Chris Lattnerbe2e1b52011-03-10 07:43:44 +0000[diff] [blame]850<li>The Alpha, Blackfin, CellSPU, MicroBlaze, MSP430, MIPS, PTX, SystemZ
Chris Lattnerbb117712010-10-04 01:29:06 +0000[diff] [blame]851 and XCore backends are experimental.</li>
Chris Lattner7d9b6b42010-10-02 21:59:30 +0000[diff] [blame]852<li><tt>llc</tt> "<tt>-filetype=obj</tt>" is experimental on all targets
Chris Lattner49123fd2011-04-06 06:29:50 +0000[diff] [blame]853 other than darwin and ELF X86 systems.</li>
Chris Lattnerbe2e1b52011-03-10 07:43:44 +0000[diff] [blame]854
Chris Lattnerf5ee1702004-03-14 02:03:02 +0000[diff] [blame]855</ul>
856
857</div>
Chris Lattner79c3fe12003-10-02 04:57:28 +0000[diff] [blame]858
Misha Brukman500bc302003-11-22 00:38:41 +0000[diff] [blame]859<!-- ======================================================================= -->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]860<h3>
Chris Lattner26299222006-11-18 07:51:14 +0000[diff] [blame]861 <a name="x86-be">Known problems with the X86 back-end</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]862</h3>
John Criswellc0c186d2005-11-08 21:11:33 +0000[diff] [blame]863
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]864<div>
John Criswellc0c186d2005-11-08 21:11:33 +0000[diff] [blame]865
866<ul>
Anton Korobeynikova6094be2008-06-08 10:24:13 +0000[diff] [blame]867 <li>The X86 backend does not yet support
868 all <a href="http://llvm.org/PR879">inline assembly that uses the X86
869 floating point stack</a>. It supports the 'f' and 't' constraints, but not
870 'u'.</li>
Dan Gohman8207ba92008-06-08 23:05:11 +0000[diff] [blame]871 <li>The X86-64 backend does not yet support the LLVM IR instruction
Chris Lattner914ce462010-04-22 06:28:20 +0000[diff] [blame]872 <tt>va_arg</tt>. Currently, front-ends support variadic
Dan Gohman8207ba92008-06-08 23:05:11 +0000[diff] [blame]873 argument constructs on X86-64 by lowering them manually.</li>
NAKAMURA Takumi45c435a2011-04-05 08:24:22 +0000[diff] [blame]874 <li>Windows x64 (aka Win64) code generator has a few issues.
875 <ul>
876 <li>llvm-gcc cannot build the mingw-w64 runtime currently
877 due to lack of support for the 'u' inline assembly
878 constraint and for X87 floating point inline assembly.</li>
879 <li>On mingw-w64, you will see unresolved symbol <tt>__chkstk</tt>
880 due to <a href="http://llvm.org/bugs/show_bug.cgi?id=8919">Bug 8919</a>.
881 It is fixed in <a href="http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20110321/118499.html">r128206</a>.</li>
882 <li>Miss-aligned MOVDQA might crash your program. It is due to
883 <a href="http://llvm.org/bugs/show_bug.cgi?id=9483">Bug 9483</a>,
884 lack of handling aligned internal globals.</li>
885 </ul>
886 </li>
887
Chris Lattner26299222006-11-18 07:51:14 +0000[diff] [blame]888</ul>
889
890</div>
891
892<!-- ======================================================================= -->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]893<h3>
Chris Lattner26299222006-11-18 07:51:14 +0000[diff] [blame]894 <a name="ppc-be">Known problems with the PowerPC back-end</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]895</h3>
Chris Lattner26299222006-11-18 07:51:14 +0000[diff] [blame]896
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]897<div>
Chris Lattner26299222006-11-18 07:51:14 +0000[diff] [blame]898
899<ul>
Nicolas Geoffraye4285dc2007-05-15 09:21:28 +0000[diff] [blame]900<li>The Linux PPC32/ABI support needs testing for the interpreter and static
Chris Lattner57a460e2007-05-23 04:39:32 +0000[diff] [blame]901compilation, and lacks support for debug information.</li>
Chris Lattner26299222006-11-18 07:51:14 +0000[diff] [blame]902</ul>
903
904</div>
905
906<!-- ======================================================================= -->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]907<h3>
Chris Lattnerf3e5bc62007-05-14 06:56:09 +0000[diff] [blame]908 <a name="arm-be">Known problems with the ARM back-end</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]909</h3>
Chris Lattnerf3e5bc62007-05-14 06:56:09 +0000[diff] [blame]910
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]911<div>
Chris Lattnerf3e5bc62007-05-14 06:56:09 +0000[diff] [blame]912
913<ul>
Chris Lattner57a460e2007-05-23 04:39:32 +0000[diff] [blame]914<li>Thumb mode works only on ARMv6 or higher processors. On sub-ARMv6
Duncan Sandsc90d68b2007-09-26 15:59:54 +0000[diff] [blame]915processors, thumb programs can crash or produce wrong
Chris Lattner57a460e2007-05-23 04:39:32 +0000[diff] [blame]916results (<a href="http://llvm.org/PR1388">PR1388</a>).</li>
John Criswell524a5dd2009-03-02 15:28:15 +0000[diff] [blame]917<li>Compilation for ARM Linux OABI (old ABI) is supported but not fully tested.
Chris Lattnerf3e5bc62007-05-14 06:56:09 +0000[diff] [blame]918</li>
Chris Lattnerf3e5bc62007-05-14 06:56:09 +0000[diff] [blame]919</ul>
920
921</div>
922
923<!-- ======================================================================= -->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]924<h3>
Chris Lattner26299222006-11-18 07:51:14 +0000[diff] [blame]925 <a name="sparc-be">Known problems with the SPARC back-end</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]926</h3>
Chris Lattner26299222006-11-18 07:51:14 +0000[diff] [blame]927
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]928<div>
Chris Lattner26299222006-11-18 07:51:14 +0000[diff] [blame]929
930<ul>
John Criswell524a5dd2009-03-02 15:28:15 +0000[diff] [blame]931<li>The SPARC backend only supports the 32-bit SPARC ABI (-m32); it does not
Chris Lattner26299222006-11-18 07:51:14 +0000[diff] [blame]932 support the 64-bit SPARC ABI (-m64).</li>
933</ul>
934
935</div>
936
937<!-- ======================================================================= -->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]938<h3>
Bruno Cardoso Lopesb7e1a4f2008-10-25 14:56:26 +0000[diff] [blame]939 <a name="mips-be">Known problems with the MIPS back-end</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]940</h3>
Bruno Cardoso Lopesb7e1a4f2008-10-25 14:56:26 +0000[diff] [blame]941
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]942<div>
Bruno Cardoso Lopesb7e1a4f2008-10-25 14:56:26 +0000[diff] [blame]943
944<ul>
Bruno Cardoso Lopesb7e1a4f2008-10-25 14:56:26 +0000[diff] [blame]945<li>64-bit MIPS targets are not supported yet.</li>
946</ul>
947
948</div>
949
950<!-- ======================================================================= -->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]951<h3>
Chris Lattner26299222006-11-18 07:51:14 +0000[diff] [blame]952 <a name="alpha-be">Known problems with the Alpha back-end</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]953</h3>
Chris Lattner26299222006-11-18 07:51:14 +0000[diff] [blame]954
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]955<div>
Chris Lattner26299222006-11-18 07:51:14 +0000[diff] [blame]956
957<ul>
958
959<li>On 21164s, some rare FP arithmetic sequences which may trap do not have the
960appropriate nops inserted to ensure restartability.</li>
961
John Criswellc0c186d2005-11-08 21:11:33 +0000[diff] [blame]962</ul>
963</div>
964
Chris Lattner26299222006-11-18 07:51:14 +0000[diff] [blame]965<!-- ======================================================================= -->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]966<h3>
Chris Lattnerf3e5bc62007-05-14 06:56:09 +0000[diff] [blame]967 <a name="c-be">Known problems with the C back-end</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]968</h3>
Chris Lattner26299222006-11-18 07:51:14 +0000[diff] [blame]969
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]970<div>
Chris Lattner26299222006-11-18 07:51:14 +0000[diff] [blame]971
Chris Lattner3016ee92010-09-29 05:34:42 +0000[diff] [blame]972<p>The C backend has numerous problems and is not being actively maintained.
973Depending on it for anything serious is not advised.</p>
974
Chris Lattner26299222006-11-18 07:51:14 +0000[diff] [blame]975<ul>
Chris Lattner5733b272008-06-05 06:35:40 +0000[diff] [blame]976<li><a href="http://llvm.org/PR802">The C backend has only basic support for
977 inline assembly code</a>.</li>
Chris Lattner725a0d82007-09-26 06:01:35 +0000[diff] [blame]978<li><a href="http://llvm.org/PR1658">The C backend violates the ABI of common
979 C++ programs</a>, preventing intermixing between C++ compiled by the CBE and
Gabor Greif4906abe2009-03-02 12:02:51 +0000[diff] [blame]980 C++ code compiled with <tt>llc</tt> or native compilers.</li>
Duncan Sandsf74c0cc2008-02-10 13:40:55 +0000[diff] [blame]981<li>The C backend does not support all exception handling constructs.</li>
Duncan Sands50723a92009-02-25 11:51:54 +0000[diff] [blame]982<li>The C backend does not support arbitrary precision integers.</li>
Chris Lattner26299222006-11-18 07:51:14 +0000[diff] [blame]983</ul>
984
985</div>
John Criswellc0c186d2005-11-08 21:11:33 +0000[diff] [blame]986
Misha Brukman500bc302003-11-22 00:38:41 +0000[diff] [blame]987
988<!-- ======================================================================= -->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]989<h3>
Chris Lattner3016ee92010-09-29 05:34:42 +0000[diff] [blame]990 <a name="llvm-gcc">Known problems with the llvm-gcc front-end</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]991</h3>
Chris Lattner47588f92003-10-02 05:07:23 +0000[diff] [blame]992
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]993<div>
Chris Lattnerc5d658a2006-03-03 00:34:26 +0000[diff] [blame]994
Chris Lattner49123fd2011-04-06 06:29:50 +0000[diff] [blame]995<p><b>LLVM 2.9 will be the last release of llvm-gcc.</b></p>
996
Chris Lattner3016ee92010-09-29 05:34:42 +0000[diff] [blame]997<p>llvm-gcc is generally very stable for the C family of languages. The only
998 major language feature of GCC not supported by llvm-gcc is the
999 <tt>__builtin_apply</tt> family of builtins. However, some extensions
1000 are only supported on some targets. For example, trampolines are only
1001 supported on some targets (these are used when you take the address of a
1002 nested function).</p>
Misha Brukman500bc302003-11-22 00:38:41 +0000[diff] [blame]1003
Chris Lattner3016ee92010-09-29 05:34:42 +0000[diff] [blame]1004<p>Fortran support generally works, but there are still several unresolved bugs
1005 in <a href="http://llvm.org/bugs/">Bugzilla</a>. Please see the
1006 tools/gfortran component for details. Note that llvm-gcc is missing major
1007 Fortran performance work in the frontend and library that went into GCC after
1008 4.2. If you are interested in Fortran, we recommend that you consider using
1009 <a href="#dragonegg">dragonegg</a> instead.</p>
Chris Lattner79c3fe12003-10-02 04:57:28 +0000[diff] [blame]1010
Duncan Sands3af96332010-10-04 10:06:56 +0000[diff] [blame]1011<p>The llvm-gcc 4.2 Ada compiler has basic functionality, but is no longer being
1012actively maintained. If you are interested in Ada, we recommend that you
1013consider using <a href="#dragonegg">dragonegg</a> instead.</p>
Chris Lattner2b659ef2008-02-12 06:29:45 +0000[diff] [blame]1014</div>
1015
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]1016</div>
1017
Chris Lattner79c3fe12003-10-02 04:57:28 +0000[diff] [blame]1018<!-- *********************************************************************** -->
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]1019<h2>
Misha Brukman500bc302003-11-22 00:38:41 +0000[diff] [blame]1020 <a name="additionalinfo">Additional Information</a>
NAKAMURA Takumi06c6d9a2011-04-18 01:17:51 +0000[diff] [blame]1021</h2>
Chris Lattner79c3fe12003-10-02 04:57:28 +0000[diff] [blame]1022<!-- *********************************************************************** -->
1023
NAKAMURA Takumi074eeaa2011-04-21 01:52:00 +0000[diff] [blame]1024<div>
Misha Brukman500bc302003-11-22 00:38:41 +0000[diff] [blame]1025
Chris Lattner416db102005-05-16 17:13:10 +0000[diff] [blame]1026<p>A wide variety of additional information is available on the <a
NAKAMURA Takumib9a33632011-04-09 02:13:37 +0000[diff] [blame]1027href="http://llvm.org/">LLVM web page</a>, in particular in the <a
Chris Lattnerb4b0ce72007-05-18 00:44:29 +0000[diff] [blame]1028href="http://llvm.org/docs/">documentation</a> section. The web page also
Reid Spencer669ed452007-07-09 08:04:31 +0000[diff] [blame]1029contains versions of the API documentation which is up-to-date with the
1030Subversion version of the source code.
Misha Brukman109d9e82005-03-30 19:14:24 +0000[diff] [blame]1031You can access versions of these documents specific to this release by going
1032into the "<tt>llvm/doc/</tt>" directory in the LLVM tree.</p>
Chris Lattner79c3fe12003-10-02 04:57:28 +0000[diff] [blame]1033
Misha Brukman500bc302003-11-22 00:38:41 +0000[diff] [blame]1034<p>If you have any questions or comments about LLVM, please feel free to contact
Chris Lattnerc463b272005-10-29 07:07:09 +0000[diff] [blame]1035us via the <a href="http://llvm.org/docs/#maillist"> mailing
Chris Lattner5eccca42003-12-12 21:22:16 +0000[diff] [blame]1036lists</a>.</p>
Chris Lattner79c3fe12003-10-02 04:57:28 +0000[diff] [blame]1037
Misha Brukman500bc302003-11-22 00:38:41 +0000[diff] [blame]1038</div>
Chris Lattner79c3fe12003-10-02 04:57:28 +0000[diff] [blame]1039
1040<!-- *********************************************************************** -->
Chris Lattner79c3fe12003-10-02 04:57:28 +0000[diff] [blame]1041
Misha Brukman500bc302003-11-22 00:38:41 +0000[diff] [blame]1042<hr>
Misha Brukman2061e892003-11-22 01:23:39 +0000[diff] [blame]1043<address>
Misha Brukman38847d52003-12-21 22:53:21 +0000[diff] [blame]1044 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
Misha Brukman44408702008-12-11 17:34:48 +0000[diff] [blame]1045 src="http://jigsaw.w3.org/css-validator/images/vcss-blue" alt="Valid CSS"></a>
Misha Brukman2061e892003-11-22 01:23:39 +0000[diff] [blame]1046 <a href="http://validator.w3.org/check/referer"><img
Misha Brukman44408702008-12-11 17:34:48 +0000[diff] [blame]1047 src="http://www.w3.org/Icons/valid-html401-blue" alt="Valid HTML 4.01"></a>
Misha Brukman2061e892003-11-22 01:23:39 +0000[diff] [blame]1048
Chris Lattnerb4b0ce72007-05-18 00:44:29 +0000[diff] [blame]1049 <a href="http://llvm.org/">LLVM Compiler Infrastructure</a><br>
Misha Brukman500bc302003-11-22 00:38:41 +0000[diff] [blame]1050 Last modified: $Date$
Misha Brukman2061e892003-11-22 01:23:39 +0000[diff] [blame]1051</address>
Chris Lattner79c3fe12003-10-02 04:57:28 +0000[diff] [blame]1052
Misha Brukman500bc302003-11-22 00:38:41 +0000[diff] [blame]1053</body>
1054</html>