Gitiles
Code Review Sign In
gerrit-public.fairphone.software / toolchain / llvm-project / 137d90c763200b85018e0cde4986d7306b8a15b0 / . / llvm / docs / ReleaseNotes.html
blob: 19b92f1d03761be874a371573ed8f8d46a267560 [file] [log] [blame]
Misha Brukman80731b92003-11-22 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 Spencere4190fa2004-11-18 18:38:58 +0000[diff] [blame]5 <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
Misha Brukman80731b92003-11-22 00:38:41 +0000[diff] [blame]6 <link rel="stylesheet" href="llvm.css" type="text/css">
Jeffrey Yasskin0830b972010-01-28 01:14:43 +0000[diff] [blame]7 <title>LLVM 2.7 Release Notes</title>
Misha Brukman80731b92003-11-22 00:38:41 +0000[diff] [blame]8</head>
9<body>
Chris Lattner3d482502003-10-02 04:57:28 +0000[diff] [blame]10
Jeffrey Yasskin0830b972010-01-28 01:14:43 +0000[diff] [blame]11<div class="doc_title">LLVM 2.7 Release Notes</div>
Mikhail Glushenkov024f7cf2008-10-13 02:08:34 +0000[diff] [blame]12
Chris Lattnerc871bac2010-03-17 04:02:39 +0000[diff] [blame]13<img align=right src="http://llvm.org/img/DragonSmall.png"
Gabor Greif27b166352010-04-22 10:21:43 +0000[diff] [blame]14 width="136" height="136" alt="LLVM Dragon Logo">
Chris Lattnerc871bac2010-03-17 04:02:39 +0000[diff] [blame]15
Chris Lattner3d482502003-10-02 04:57:28 +0000[diff] [blame]16<ol>
Misha Brukman80731b92003-11-22 00:38:41 +0000[diff] [blame]17 <li><a href="#intro">Introduction</a></li>
Chris Lattnerf5cd9862008-10-13 18:01:01 +0000[diff] [blame]18 <li><a href="#subproj">Sub-project Status Update</a></li>
Jeffrey Yasskin0830b972010-01-28 01:14:43 +0000[diff] [blame]19 <li><a href="#externalproj">External Projects Using LLVM 2.7</a></li>
20 <li><a href="#whatsnew">What's New in LLVM 2.7?</a></li>
Chris Lattner77a51732004-04-30 22:17:12 +0000[diff] [blame]21 <li><a href="GettingStarted.html">Installation Instructions</a></li>
Misha Brukman80731b92003-11-22 00:38:41 +0000[diff] [blame]22 <li><a href="#portability">Portability and Supported Platforms</a></li>
Dan Gohmanad888912008-10-14 16:23:02 +0000[diff] [blame]23 <li><a href="#knownproblems">Known Problems</a></li>
Misha Brukman80731b92003-11-22 00:38:41 +0000[diff] [blame]24 <li><a href="#additionalinfo">Additional Information</a></li>
Chris Lattner3d482502003-10-02 04:57:28 +0000[diff] [blame]25</ol>
26
Chris Lattner020e1fc2004-05-23 21:07:27 +0000[diff] [blame]27<div class="doc_author">
Dan Gohmanad888912008-10-14 16:23:02 +0000[diff] [blame]28 <p>Written by the <a href="http://llvm.org">LLVM Team</a></p>
Misha Brukman80731b92003-11-22 00:38:41 +0000[diff] [blame]29</div>
Chris Lattner3d482502003-10-02 04:57:28 +0000[diff] [blame]30
Chris Lattner06971182010-04-13 06:37:00 +0000[diff] [blame]31<!--
Chris Lattner656db162010-04-22 05:41:35 +0000[diff] [blame]32<h1 style="color:red">These are in-progress notes for the upcoming LLVM 2.8
Jeffrey Yasskin0830b972010-01-28 01:14:43 +0000[diff] [blame]33release.<br>
34You may prefer the
Chris Lattner656db162010-04-22 05:41:35 +0000[diff] [blame]35<a href="http://llvm.org/releases/2.6/docs/ReleaseNotes.html">LLVM 2.7
Chris Lattner06971182010-04-13 06:37:00 +0000[diff] [blame]36Release Notes</a>.</h1>-->
Jeffrey Yasskin0830b972010-01-28 01:14:43 +0000[diff] [blame]37
Chris Lattner3d482502003-10-02 04:57:28 +0000[diff] [blame]38<!-- *********************************************************************** -->
Misha Brukman80731b92003-11-22 00:38:41 +0000[diff] [blame]39<div class="doc_section">
40 <a name="intro">Introduction</a>
41</div>
Chris Lattner3d482502003-10-02 04:57:28 +0000[diff] [blame]42<!-- *********************************************************************** -->
43
Misha Brukman80731b92003-11-22 00:38:41 +0000[diff] [blame]44<div class="doc_text">
45
Chris Lattner1e4d5bc2008-10-13 17:57:36 +0000[diff] [blame]46<p>This document contains the release notes for the LLVM Compiler
Jeffrey Yasskin0830b972010-01-28 01:14:43 +0000[diff] [blame]47Infrastructure, release 2.7. Here we describe the status of LLVM, including
Chris Lattner1e4d5bc2008-10-13 17:57:36 +0000[diff] [blame]48major improvements from the previous release and significant known problems.
Mikhail Glushenkov25422542009-03-01 18:09:47 +0000[diff] [blame]49All LLVM releases may be downloaded from the <a
Chris Lattner1e4d5bc2008-10-13 17:57:36 +0000[diff] [blame]50href="http://llvm.org/releases/">LLVM releases web site</a>.</p>
Chris Lattner62495762003-10-02 16:38:05 +0000[diff] [blame]51
Chris Lattnerb5bb5972004-12-07 08:04:13 +0000[diff] [blame]52<p>For more information about LLVM, including information about the latest
Chris Lattnera69595e2005-10-29 07:07:09 +0000[diff] [blame]53release, please check out the <a href="http://llvm.org/">main LLVM
Chris Lattnere7525b52003-10-07 21:38:31 +0000[diff] [blame]54web site</a>. If you have questions or comments, the <a
Chris Lattner0b1c9a52010-03-17 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 Lattner3d482502003-10-02 04:57:28 +0000[diff] [blame]57
Mikhail Glushenkov024f7cf2008-10-13 02:08:34 +0000[diff] [blame]58<p>Note that if you are reading this file from a Subversion checkout or the
Chris Lattner0a1fd102007-09-21 03:54:09 +0000[diff] [blame]59main LLVM web page, this document applies to the <i>next</i> release, not the
Gabor Greif355f81c2008-10-14 11:00:32 +0000[diff] [blame]60current one. To see the release notes for a specific release, please see the
Chris Lattner0a1fd102007-09-21 03:54:09 +0000[diff] [blame]61<a href="http://llvm.org/releases/">releases page</a>.</p>
Misha Brukman80731b92003-11-22 00:38:41 +0000[diff] [blame]62
63</div>
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]64
Chris Lattner3d482502003-10-02 04:57:28 +0000[diff] [blame]65
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]66<!--
67Almost dead code.
68 include/llvm/Analysis/LiveValues.h => Dan
69 lib/Transforms/IPO/MergeFunctions.cpp => consider for 2.8.
70 llvm/Analysis/PointerTracking.h => Edwin wants this, consider for 2.8.
Chris Lattner06971182010-04-13 06:37:00 +0000[diff] [blame]71 ABCD, GEPSplitterPass
Chris Lattner89ea59d2010-01-16 21:25:13 +0000[diff] [blame]72 MSIL backend?
Chris Lattner10a3cc42010-03-17 06:41:58 +0000[diff] [blame]73 lib/Transforms/Utils/SSI.cpp -> ABCD depends on it.
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]74-->
75
76
Jeffrey Yasskin0830b972010-01-28 01:14:43 +0000[diff] [blame]77<!-- Features that need text if they're finished for 2.7:
Chris Lattnera67df2d2010-04-22 06:28:20 +0000[diff] [blame]78 combiner-aa?
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]79 strong phi elim
Chris Lattner87e04952010-04-21 05:17:40 +0000[diff] [blame]80 llvm.dbg.value: variable debug info for optimized code
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]81 loop dependence analysis
Chris Lattner2b8a52e2008-02-10 07:46:44 +0000[diff] [blame]82 -->
Mikhail Glushenkov024f7cf2008-10-13 02:08:34 +0000[diff] [blame]83
Chris Lattnerc4d68542008-10-12 19:47:48 +0000[diff] [blame]84 <!-- for announcement email:
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]85 Logo web page.
86 llvm devmtg
87 compiler_rt
88 KLEE web page at klee.llvm.org
89 Many new papers added to /pubs/
Chris Lattner10a3cc42010-03-17 06:41:58 +0000[diff] [blame]90 Mention gcc plugin.
Chris Lattnerd02da042009-02-25 06:34:50 +0000[diff] [blame]91 -->
Chris Lattner2b8a52e2008-02-10 07:46:44 +0000[diff] [blame]92
Chris Lattnerf5cd9862008-10-13 18:01:01 +0000[diff] [blame]93<!-- *********************************************************************** -->
94<div class="doc_section">
95 <a name="subproj">Sub-project Status Update</a>
Chris Lattner625a3d82008-06-08 21:34:41 +0000[diff] [blame]96</div>
Chris Lattnerf5cd9862008-10-13 18:01:01 +0000[diff] [blame]97<!-- *********************************************************************** -->
Chris Lattner625a3d82008-06-08 21:34:41 +0000[diff] [blame]98
99<div class="doc_text">
Chris Lattnerc75fd522008-06-08 21:58:17 +0000[diff] [blame]100<p>
Jeffrey Yasskin0830b972010-01-28 01:14:43 +0000[diff] [blame]101The LLVM 2.7 distribution currently consists of code from the core LLVM
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]102repository (which roughly includes the LLVM optimizers, code generators
103and supporting tools), the Clang repository and the llvm-gcc repository. In
104addition to this code, the LLVM Project includes other sub-projects that are in
105development. Here we include updates on these subprojects.
Bill Wendlingf170d2e2009-03-02 04:28:57 +0000[diff] [blame]106</p>
Chris Lattnerf5cd9862008-10-13 18:01:01 +0000[diff] [blame]107
Chris Lattnerc75fd522008-06-08 21:58:17 +0000[diff] [blame]108</div>
109
Chris Lattnerf5cd9862008-10-13 18:01:01 +0000[diff] [blame]110
Chris Lattnerc75fd522008-06-08 21:58:17 +0000[diff] [blame]111<!--=========================================================================-->
Chris Lattnerf5cd9862008-10-13 18:01:01 +0000[diff] [blame]112<div class="doc_subsection">
Chris Lattner44c09cd2008-10-13 18:11:54 +0000[diff] [blame]113<a name="clang">Clang: C/C++/Objective-C Frontend Toolkit</a>
Chris Lattnerc75fd522008-06-08 21:58:17 +0000[diff] [blame]114</div>
115
116<div class="doc_text">
117
Chris Lattner5de7f6e2010-04-26 17:42:18 +0000[diff] [blame]118<p><a href="http://clang.llvm.org/">Clang</a> is an LLVM front end for the C,
119C++, and Objective-C languages. Clang aims to provide a better user experience
120through expressive diagnostics, a high level of conformance to language
121standards, fast compilation, and low memory use. Like LLVM, Clang provides a
122modular, library-based architecture that makes it suitable for creating or
123integrating with other development tools. Clang is considered a
124production-quality compiler for C and Objective-C on x86 (32- and 64-bit).</p>
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]125
Jeffrey Yasskin0830b972010-01-28 01:14:43 +0000[diff] [blame]126<p>In the LLVM 2.7 time-frame, the Clang team has made many improvements:</p>
Bill Wendlingef362462008-10-27 09:27:33 +0000[diff] [blame]127
Daniel Dunbarc11b40f2008-10-14 23:25:09 +0000[diff] [blame]128<ul>
Douglas Gregorc5b5c052010-04-22 20:42:40 +0000[diff] [blame]129
Chris Lattner5de7f6e2010-04-26 17:42:18 +0000[diff] [blame]130<li>C++ Support: Clang is now capable of self-hosting! While still
131alpha-quality, Clang's C++ support has matured enough to build LLVM and Clang,
132and C++ is now enabled by default. See the <a
133href="http://clang.llvm.org/cxx_compatibility.html">Clang C++ compatibility
134page</a> for common C++ migration issues.</li>
Douglas Gregorc5b5c052010-04-22 20:42:40 +0000[diff] [blame]135
Chris Lattner5de7f6e2010-04-26 17:42:18 +0000[diff] [blame]136<li>Objective-C: Clang now includes experimental support for an updated
137Objective-C ABI on non-Darwin platforms. This includes support for non-fragile
138instance variables and accelerated proxies, as well as greater potential for
139future optimisations. The new ABI is used when compiling with the
140-fobjc-nonfragile-abi and -fgnu-runtime options. Code compiled with these
141options may be mixed with code compiled with GCC or clang using the old GNU ABI,
142but requires the libobjc2 runtime from the GNUstep project.</li>
David Chisnall17dff472010-04-25 19:13:33 +0000[diff] [blame]143
Chris Lattner5de7f6e2010-04-26 17:42:18 +0000[diff] [blame]144<li>New warnings: Clang contains a number of new warnings, including
145control-flow warnings (unreachable code, missing return statements in a
146non-<code>void</code> function, etc.), sign-comparison warnings, and improved
147format-string warnings.</li>
Daniel Dunbareb4bc7f2010-03-25 16:09:18 +0000[diff] [blame]148
Daniel Dunbareb4bc7f2010-03-25 16:09:18 +0000[diff] [blame]149<li>CIndex API and Python bindings: Clang now includes a C API as part of the
Gabor Greifc1670312010-04-25 21:27:54 +0000[diff] [blame]150CIndex library. Although we may make some changes to the API in the future, it
Daniel Dunbareb4bc7f2010-03-25 16:09:18 +0000[diff] [blame]151is intended to be stable and has been designed for use by external projects. See
152the Clang
153doxygen <a href="http://clang.llvm.org/doxygen/group__CINDEX.html">CIndex</a>
Wesley Peck9bc5e8f2010-04-22 14:19:00 +0000[diff] [blame]154documentation for more details. The CIndex API also includes a preliminary
Daniel Dunbareb4bc7f2010-03-25 16:09:18 +0000[diff] [blame]155set of Python bindings.</li>
156
157<li>ARM Support: Clang now has ABI support for both the Darwin and Linux ARM
158ABIs. Coupled with many improvements to the LLVM ARM backend, Clang is now
Gabor Greifc1670312010-04-25 21:27:54 +0000[diff] [blame]159suitable for use as a beta quality ARM compiler.</li>
David Chisnall17dff472010-04-25 19:13:33 +0000[diff] [blame]160
Bill Wendling87adbe42009-03-02 04:28:18 +0000[diff] [blame]161</ul>
Chris Lattner44c09cd2008-10-13 18:11:54 +0000[diff] [blame]162</div>
163
164<!--=========================================================================-->
165<div class="doc_subsection">
166<a name="clangsa">Clang Static Analyzer</a>
167</div>
168
169<div class="doc_text">
170
Chris Lattner2284b6a2010-04-22 06:38:11 +0000[diff] [blame]171<p>The <a href="http://clang-analyzer.llvm.org/">Clang Static Analyzer</a>
172 project is an effort to use static source code analysis techniques to
173 automatically find bugs in C and Objective-C programs (and hopefully <a
174 href="http://clang-analyzer.llvm.org/dev_cxx.html">C++ in the
175 future</a>!). The tool is very good at finding bugs that occur on specific
176 paths through code, such as on error conditions.</p>
Chris Lattnerbcffa5a2008-10-14 00:52:49 +0000[diff] [blame]177
Chris Lattner2284b6a2010-04-22 06:38:11 +0000[diff] [blame]178<p>In the LLVM 2.7 time-frame, the analyzer core has made several major and
179 minor improvements, including better support for tracking the fields of
180 structures, initial support (not enabled by default yet) for doing
181 interprocedural (cross-function) analysis, and new checks have been added.
182</p>
Chris Lattnerc75fd522008-06-08 21:58:17 +0000[diff] [blame]183
184</div>
185
Chris Lattnerf5cd9862008-10-13 18:01:01 +0000[diff] [blame]186<!--=========================================================================-->
187<div class="doc_subsection">
Nicolas Geoffraye5ec9322008-10-14 19:23:04 +0000[diff] [blame]188<a name="vmkit">VMKit: JVM/CLI Virtual Machine Implementation</a>
Chris Lattnerf5cd9862008-10-13 18:01:01 +0000[diff] [blame]189</div>
190
191<div class="doc_text">
192<p>
Nicolas Geoffraye5ec9322008-10-14 19:23:04 +0000[diff] [blame]193The <a href="http://vmkit.llvm.org/">VMKit project</a> is an implementation of
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]194a JVM and a CLI Virtual Machine (Microsoft .NET is an
195implementation of the CLI) using LLVM for static and just-in-time
196compilation.</p>
Chris Lattnerf5cd9862008-10-13 18:01:01 +0000[diff] [blame]197
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]198<p>
Chris Lattneref2e5732010-03-29 18:34:13 +0000[diff] [blame]199With the release of LLVM 2.7, VMKit has shifted to a great framework for writing
200virtual machines. VMKit now offers precise and efficient garbage collection with
201multi-threading support, thanks to the MMTk memory management toolkit, as well
202as just in time and ahead of time compilation with LLVM. The major changes in
203VMKit 0.27 are:</p>
Chris Lattnerf5cd9862008-10-13 18:01:01 +0000[diff] [blame]204
Nicolas Geoffraye5ec9322008-10-14 19:23:04 +0000[diff] [blame]205<ul>
206
Chris Lattneref2e5732010-03-29 18:34:13 +0000[diff] [blame]207<li>Garbage collection: VMKit now uses the MMTk toolkit for garbage collectors.
208 The first collector to be ported is the MarkSweep collector, which is precise,
209 and drastically improves the performance of VMKit.</li>
210<li>Line number information in the JVM: by using the debug metadata of LLVM, the
211 JVM now supports precise line number information, useful when printing a stack
212 trace.</li>
213<li>Interface calls in the JVM: we implemented a variant of the Interface Method
214 Table technique for interface calls in the JVM.
215</li>
Nicolas Geoffraye5ec9322008-10-14 19:23:04 +0000[diff] [blame]216
217</ul>
Chris Lattnerf5cd9862008-10-13 18:01:01 +0000[diff] [blame]218</div>
219
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]220
221<!--=========================================================================-->
222<div class="doc_subsection">
223<a name="compiler-rt">compiler-rt: Compiler Runtime Library</a>
224</div>
225
226<div class="doc_text">
227<p>
228The new LLVM <a href="http://compiler-rt.llvm.org/">compiler-rt project</a>
229is a simple library that provides an implementation of the low-level
230target-specific hooks required by code generation and other runtime components.
231For example, when compiling for a 32-bit target, converting a double to a 64-bit
232unsigned integer is compiled into a runtime call to the "__fixunsdfdi"
233function. The compiler-rt library provides highly optimized implementations of
234this and other low-level routines (some are 3x faster than the equivalent
235libgcc routines).</p>
236
237<p>
238All of the code in the compiler-rt project is available under the standard LLVM
Chris Lattner87e04952010-04-21 05:17:40 +0000[diff] [blame]239License, a "BSD-style" license. New in LLVM 2.7: compiler_rt now
240supports ARM targets.</p>
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]241
242</div>
243
244<!--=========================================================================-->
245<div class="doc_subsection">
Duncan Sands92452b92010-04-02 09:23:15 +0000[diff] [blame]246<a name="dragonegg">DragonEgg: llvm-gcc ported to gcc-4.5</a>
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]247</div>
248
249<div class="doc_text">
250<p>
Duncan Sands92452b92010-04-02 09:23:15 +0000[diff] [blame]251<a href="http://dragonegg.llvm.org/">DragonEgg</a> is a port of llvm-gcc to
252gcc-4.5. Unlike llvm-gcc, which makes many intrusive changes to the underlying
253gcc-4.2 code, dragonegg in theory does not require any gcc-4.5 modifications
254whatsoever (currently one small patch is needed). This is thanks to the new
255<a href="http://gcc.gnu.org/wiki/plugins">gcc plugin architecture</a>, which
256makes it possible to modify the behaviour of gcc at runtime by loading a plugin,
257which is nothing more than a dynamic library which conforms to the gcc plugin
258interface. DragonEgg is a gcc plugin that causes the LLVM optimizers to be run
259instead of the gcc optimizers, and the LLVM code generators instead of the gcc
260code generators, just like llvm-gcc. To use it, you add
261"-fplugin=path/dragonegg.so" to the gcc-4.5 command line, and gcc-4.5 magically
262becomes llvm-gcc-4.5!
263</p>
264
265<p>
266DragonEgg is still a work in progress. Currently C works very well, while C++,
267Ada and Fortran work fairly well. All other languages either don't work at all,
268or only work poorly. For the moment only the x86-32 and x86-64 targets are
Duncan Sands4dd125e2010-04-20 19:40:58 +0000[diff] [blame]269supported, and only on linux and darwin (darwin needs an additional gcc patch).
Duncan Sands92452b92010-04-02 09:23:15 +0000[diff] [blame]270</p>
271
272<p>
Duncan Sandsfe291172010-04-21 13:51:48 +0000[diff] [blame]273DragonEgg is a new project which is seeing its first release with llvm-2.7.
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]274</p>
275
276</div>
277
278
279<!--=========================================================================-->
280<div class="doc_subsection">
281<a name="mc">llvm-mc: Machine Code Toolkit</a>
282</div>
283
284<div class="doc_text">
285<p>
Chris Lattner2284b6a2010-04-22 06:38:11 +0000[diff] [blame]286The LLVM Machine Code (aka MC) sub-project of LLVM was created to solve a number
287of problems in the realm of assembly, disassembly, object file format handling,
288and a number of other related areas that CPU instruction-set level tools work
289in. It is a sub-project of LLVM which provides it with a number of advantages
290over other compilers that do not have tightly integrated assembly-level tools.
291For a gentle introduction, please see the <a
292href="http://blog.llvm.org/2010/04/intro-to-llvm-mc-project.html">Intro to the
293LLVM MC Project Blog Post</a>.
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]294</p>
295
Chris Lattner2284b6a2010-04-22 06:38:11 +0000[diff] [blame]296<p>2.7 includes major parts of the work required by the new MC Project. A few
297 targets have been refactored to support it, and work is underway to support a
Gabor Greif60084b12010-04-25 21:30:22 +0000[diff] [blame]298 native assembler in LLVM. This work is not complete in LLVM 2.7, but it has
Chris Lattner2284b6a2010-04-22 06:38:11 +0000[diff] [blame]299 made substantially more progress on LLVM mainline.</p>
300
301<p>One minor example of what MC can do is to transcode an AT&amp;T syntax
Gabor Greifd80517c2010-04-22 10:25:23 +0000[diff] [blame]302 X86 .s file into intel syntax. You can do this with something like:</p>
Chris Lattner2284b6a2010-04-22 06:38:11 +0000[diff] [blame]303<pre>
304 llvm-mc foo.s -output-asm-variant=1 -o foo-intel.s
305</pre>
Chris Lattner1dcca8b2010-04-21 06:23:40 +0000[diff] [blame]306
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]307</div>
308
309
Chris Lattner53e06f92009-02-26 22:33:38 +0000[diff] [blame]310<!-- *********************************************************************** -->
311<div class="doc_section">
Jeffrey Yasskin0830b972010-01-28 01:14:43 +0000[diff] [blame]312 <a name="externalproj">External Open Source Projects Using LLVM 2.7</a>
Chris Lattner53e06f92009-02-26 22:33:38 +0000[diff] [blame]313</div>
314<!-- *********************************************************************** -->
315
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]316<div class="doc_text">
317
318<p>An exciting aspect of LLVM is that it is used as an enabling technology for
319 a lot of other language and tools projects. This section lists some of the
Jeffrey Yasskin0830b972010-01-28 01:14:43 +0000[diff] [blame]320 projects that have already been updated to work with LLVM 2.7.</p>
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]321</div>
322
Chris Lattner53e06f92009-02-26 22:33:38 +0000[diff] [blame]323<!--=========================================================================-->
324<div class="doc_subsection">
325<a name="pure">Pure</a>
326</div>
327
328<div class="doc_text">
329<p>
Chris Lattnerd1094e02009-03-02 02:37:32 +0000[diff] [blame]330<a href="http://pure-lang.googlecode.com/">Pure</a>
331is an algebraic/functional programming language based on term rewriting.
Chris Lattner53e06f92009-02-26 22:33:38 +0000[diff] [blame]332Programs are collections of equations which are used to evaluate expressions in
333a symbolic fashion. Pure offers dynamic typing, eager and lazy evaluation,
Mikhail Glushenkov25422542009-03-01 18:09:47 +0000[diff] [blame]334lexical closures, a hygienic macro system (also based on term rewriting),
Chris Lattner53e06f92009-02-26 22:33:38 +0000[diff] [blame]335built-in list and matrix support (including list and matrix comprehensions) and
336an easy-to-use C interface. The interpreter uses LLVM as a backend to
337 JIT-compile Pure programs to fast native code.</p>
338
Chris Lattner6aba4862010-03-17 17:25:49 +0000[diff] [blame]339<p>Pure versions 0.43 and later have been tested and are known to work with
340LLVM 2.7 (and continue to work with older LLVM releases >= 2.5).</p>
341
Chris Lattner53e06f92009-02-26 22:33:38 +0000[diff] [blame]342</div>
343
Chris Lattnerfb23a962009-02-28 18:58:01 +0000[diff] [blame]344<!--=========================================================================-->
345<div class="doc_subsection">
346<a name="RoadsendPHP">Roadsend PHP</a>
347</div>
348
349<div class="doc_text">
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]350<p>
351<a href="http://code.roadsend.com/rphp">Roadsend PHP</a> (rphp) is an open
Chris Lattner0c72d8f2009-03-02 19:07:24 +0000[diff] [blame]352source implementation of the PHP programming
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]353language that uses LLVM for its optimizer, JIT and static compiler. This is a
Chris Lattnerc871bac2010-03-17 04:02:39 +0000[diff] [blame]354reimplementation of an earlier project that is now based on LLVM.
Chris Lattner6aba4862010-03-17 17:25:49 +0000[diff] [blame]355</p>
Chris Lattnerfb23a962009-02-28 18:58:01 +0000[diff] [blame]356</div>
Chris Lattner53e06f92009-02-26 22:33:38 +0000[diff] [blame]357
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]358<!--=========================================================================-->
359<div class="doc_subsection">
360<a name="UnladenSwallow">Unladen Swallow</a>
361</div>
362
363<div class="doc_text">
364<p>
365<a href="http://code.google.com/p/unladen-swallow/">Unladen Swallow</a> is a
366branch of <a href="http://python.org/">Python</a> intended to be fully
367compatible and significantly faster. It uses LLVM's optimization passes and JIT
Chris Lattner5b59a732010-03-18 06:52:15 +0000[diff] [blame]368compiler.
Chris Lattnerc871bac2010-03-17 04:02:39 +0000[diff] [blame]369</p>
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]370</div>
371
372<!--=========================================================================-->
373<div class="doc_subsection">
Chris Lattnerb4a02792010-03-29 17:50:39 +0000[diff] [blame]374<a name="tce">TTA-based Codesign Environment (TCE)</a>
375</div>
376
377<div class="doc_text">
378<p>
379<a href="http://tce.cs.tut.fi/">TCE</a> is a toolset for designing
380application-specific processors (ASP) based on the Transport triggered
381architecture (TTA). The toolset provides a complete co-design flow from C/C++
382programs down to synthesizable VHDL and parallel program binaries. Processor
383customization points include the register files, function units, supported
384operations, and the interconnection network.</p>
385
386<p>TCE uses llvm-gcc/Clang and LLVM for C/C++ language support, target
387independent optimizations and also for parts of code generation. It generates
388new LLVM-based code generators "on the fly" for the designed TTA processors and
389loads them in to the compiler backend as runtime libraries to avoid per-target
390recompilation of larger parts of the compiler chain.</p>
391
392</div>
Chris Lattnerc75fd522008-06-08 21:58:17 +0000[diff] [blame]393
John Criswell9003e0f2010-04-06 14:52:14 +0000[diff] [blame]394<!--=========================================================================-->
395<div class="doc_subsection">
396<a name="safecode">SAFECode Compiler</a>
397</div>
398
399<div class="doc_text">
400<p>
401<a href="http://safecode.cs.illinois.edu">SAFECode</a> is a memory safe C
402compiler built using LLVM. It takes standard, unannotated C code, analyzes the
403code to ensure that memory accesses and array indexing operations are safe, and
404instruments the code with run-time checks when safety cannot be proven
405statically.
406</p>
407</div>
408
Chris Lattnerad280412010-04-22 17:28:36 +0000[diff] [blame]409<!--=========================================================================-->
410<div class="doc_subsection">
411<a name="icedtea">IcedTea Java Virtual Machine Implementation</a>
412</div>
John Criswell9003e0f2010-04-06 14:52:14 +0000[diff] [blame]413
Chris Lattnerad280412010-04-22 17:28:36 +0000[diff] [blame]414<div class="doc_text">
415<p>
416<a href="http://icedtea.classpath.org/wiki/Main_Page">IcedTea</a> provides a
417harness to build OpenJDK using only free software build tools and to provide
418replacements for the not-yet free parts of OpenJDK. One of the extensions that
419IcedTea provides is a new JIT compiler named <a
420href="http://icedtea.classpath.org/wiki/ZeroSharkFaq">Shark</a> which uses LLVM
421to provide native code generation without introducing processor-dependent
422code.
423</p>
424<p>Icedtea6 1.8 and later have been tested and are known to work with
425LLVM 2.7 (and continue to work with older LLVM releases >= 2.6 as well).
426</p>
427</div>
428
429<!--=========================================================================-->
430<div class="doc_subsection">
431<a name="llvm-lua">LLVM-Lua</a>
432</div>
433
434<div class="doc_text">
435<p>
436<a href="http://code.google.com/p/llvm-lua/">LLVM-Lua</a> uses LLVM
437 to add JIT and static compiling support to the Lua VM. Lua
438bytecode is analyzed to remove type checks, then LLVM is used to compile the
439bytecode down to machine code.
440</p>
441<p>LLVM-Lua 1.2.0 have been tested and is known to work with LLVM 2.7.
442</p>
443</div>
Chris Lattner64694df2010-04-22 21:34:16 +0000[diff] [blame]444
445<!--=========================================================================-->
446<div class="doc_subsection">
447<a name="MacRuby">MacRuby</a>
448</div>
449
450<div class="doc_text">
451<p>
452<a href="http://macruby.org">MacRuby</a> is an implementation of Ruby based on
453core Mac OS technologies, sponsored by Apple Inc. It uses LLVM at runtime for
454optimization passes, JIT compilation and exception handling. It also allows
455static (ahead-of-time) compilation of Ruby code straight to machine code.
456</p>
457<p>The upcoming MacRuby 0.6 release works with LLVM 2.7.
458</p>
459</div>
460
Chris Lattnerdf85c892010-04-26 17:38:10 +0000[diff] [blame]461<!--=========================================================================-->
462<div class="doc_subsection">
463<a name="GHC">Glasgow Haskell Compiler (GHC)</a>
464</div>
465
466<div class="doc_text">
467<p>
468<a href="http://www.haskell.org/ghc/">GHC</a> is an open source,
469state-of-the-art programming suite for Haskell, a standard lazy
470functional programming language. It includes an optimizing static
471compiler generating good code for a variety of platforms, together
472with an interactive system for convenient, quick development.</p>
473
474<p>In addition to the existing C and native code generators, GHC now
475supports an <a
476href="http://hackage.haskell.org/trac/ghc/wiki/Commentary/Compiler/Backends/LLVM">LLVM
477code generator</a>. GHC supports LLVM 2.7.</p>
478
479</div>
480
Chris Lattner64694df2010-04-22 21:34:16 +0000[diff] [blame]481
Chris Lattnerc75fd522008-06-08 21:58:17 +0000[diff] [blame]482<!-- *********************************************************************** -->
483<div class="doc_section">
Jeffrey Yasskin0830b972010-01-28 01:14:43 +0000[diff] [blame]484 <a name="whatsnew">What's New in LLVM 2.7?</a>
Chris Lattnerc75fd522008-06-08 21:58:17 +0000[diff] [blame]485</div>
486<!-- *********************************************************************** -->
487
Chris Lattnerb7bc2aa2008-06-08 22:59:35 +0000[diff] [blame]488<div class="doc_text">
489
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]490<p>This release includes a huge number of bug fixes, performance tweaks and
Chris Lattner1e4d5bc2008-10-13 17:57:36 +0000[diff] [blame]491minor improvements. Some of the major improvements and new features are listed
492in this section.
Chris Lattnerb7bc2aa2008-06-08 22:59:35 +0000[diff] [blame]493</p>
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]494
Chris Lattnera67df2d2010-04-22 06:28:20 +0000[diff] [blame]495</div>
496
497<!--=========================================================================-->
498<div class="doc_subsection">
499<a name="orgchanges">LLVM Community Changes</a>
500</div>
501
502<div class="doc_text">
503
Chris Lattner87e04952010-04-21 05:17:40 +0000[diff] [blame]504<p>In addition to changes to the code, between LLVM 2.6 and 2.7, a number of
505organization changes have happened:
506</p>
507
508<ul>
Chris Lattner656db162010-04-22 05:41:35 +0000[diff] [blame]509<li>LLVM has a new <a href="http://llvm.org/Logo.html">official logo</a>!</li>
Chris Lattner87e04952010-04-21 05:17:40 +0000[diff] [blame]510
Chris Lattner87e04952010-04-21 05:17:40 +0000[diff] [blame]511<li>Ted Kremenek and Doug Gregor have stepped forward as <a
512 href="http://llvm.org/docs/DeveloperPolicy.html#owners">Code Owners</a> of the
Chris Lattnera67df2d2010-04-22 06:28:20 +0000[diff] [blame]513 Clang static analyzer and the Clang frontend, respectively.</li>
Chris Lattner1dcca8b2010-04-21 06:23:40 +0000[diff] [blame]514
515<li>LLVM now has an <a href="http://blog.llvm.org">official Blog</a> at
516 <a href="http://blog.llvm.org">http://blog.llvm.org</a>. This is a great way
517 to learn about new LLVM-related features as they are implemented. Several
518 features in this release are already explained on the blog.</li>
519
Chris Lattner87e04952010-04-21 05:17:40 +0000[diff] [blame]520<li>The LLVM web pages are now checked into the SVN server, in the "www",
521 "www-pubs" and "www-releases" SVN modules. Previously they were hidden in a
Chris Lattner656db162010-04-22 05:41:35 +0000[diff] [blame]522 largely inaccessible old CVS server.</li>
Chris Lattner1dcca8b2010-04-21 06:23:40 +0000[diff] [blame]523
524<li><a href="http://llvm.org">llvm.org</a> is now hosted on a new (and much
525 faster) server. It is still graciously hosted at the University of Illinois
526 of Urbana Champaign.</li>
Chris Lattner87e04952010-04-21 05:17:40 +0000[diff] [blame]527</ul>
Chris Lattnerb7bc2aa2008-06-08 22:59:35 +0000[diff] [blame]528</div>
529
Chris Lattnerc75fd522008-06-08 21:58:17 +0000[diff] [blame]530<!--=========================================================================-->
Chris Lattner625a3d82008-06-08 21:34:41 +0000[diff] [blame]531<div class="doc_subsection">
Chris Lattnercdc44ed2008-02-06 18:00:06 +0000[diff] [blame]532<a name="majorfeatures">Major New Features</a>
533</div>
534
535<div class="doc_text">
536
Jeffrey Yasskin0830b972010-01-28 01:14:43 +0000[diff] [blame]537<p>LLVM 2.7 includes several major new capabilities:</p>
Chris Lattner97beb512007-05-14 06:56:09 +0000[diff] [blame]538
539<ul>
Chris Lattner656db162010-04-22 05:41:35 +0000[diff] [blame]540<li>2.7 includes initial support for the <a
Chris Lattner1dcca8b2010-04-21 06:23:40 +0000[diff] [blame]541 href="http://en.wikipedia.org/wiki/MicroBlaze">MicroBlaze</a> target.
542 MicroBlaze is a soft processor core designed for Xilinx FPGAs.</li>
Chris Lattner87e04952010-04-21 05:17:40 +0000[diff] [blame]543
Chris Lattner656db162010-04-22 05:41:35 +0000[diff] [blame]544<li>2.7 includes a new LLVM IR "extensible metadata" feature. This feature
Chris Lattner1dcca8b2010-04-21 06:23:40 +0000[diff] [blame]545 supports many different use cases, including allowing front-end authors to
546 encode source level information into LLVM IR, which is consumed by later
547 language-specific passes. This is a great way to do high-level optimizations
548 like devirtualization, type-based alias analysis, etc. See the <a
549 href="http://blog.llvm.org/2010/04/extensible-metadata-in-llvm-ir.html">
550 Extensible Metadata Blog Post</a> for more information.</li>
551
Chris Lattner656db162010-04-22 05:41:35 +0000[diff] [blame]552<li>2.7 encodes <a href="SourceLevelDebugging.html">debug information</a>
Chris Lattner1dcca8b2010-04-21 06:23:40 +0000[diff] [blame]553in a completely new way, built on extensible metadata. The new implementation
554is much more memory efficient and paves the way for improvements to optimized
555code debugging experience.</li>
Chris Lattner87e04952010-04-21 05:17:40 +0000[diff] [blame]556
Chris Lattner656db162010-04-22 05:41:35 +0000[diff] [blame]557<li>2.7 now directly supports taking the address of a label and doing an
Chris Lattner1dcca8b2010-04-21 06:23:40 +0000[diff] [blame]558 indirect branch through a pointer. This is particularly useful for
559 interpreter loops, and is used to implement the GCC "address of label"
560 extension. For more information, see the <a
561href="http://blog.llvm.org/2010/01/address-of-label-and-indirect-branches.html">
562Address of Label and Indirect Branches in LLVM IR Blog Post</a>.
Chris Lattner87e04952010-04-21 05:17:40 +0000[diff] [blame]563
Chris Lattner656db162010-04-22 05:41:35 +0000[diff] [blame]564<li>2.7 is the first release to start supporting APIs for assembling and
Chris Lattner1dcca8b2010-04-21 06:23:40 +0000[diff] [blame]565 disassembling target machine code. These APIs are useful for a variety of
566 low level clients, and are surfaced in the new "enhanced disassembly" API.
567 For more information see the <a
568 href="http://blog.llvm.org/2010/01/x86-disassembler.html">The X86
569 Disassembler Blog Post</a> for more information.</li>
Chris Lattner87e04952010-04-21 05:17:40 +0000[diff] [blame]570
Chris Lattner2284b6a2010-04-22 06:38:11 +0000[diff] [blame]571<li>2.7 includes major parts of the work required by the new MC Project,
572 see the <a href="#mc">MC update above</a> for more information.</li>
Chris Lattner87e04952010-04-21 05:17:40 +0000[diff] [blame]573
Chris Lattner458e79f2008-02-10 08:18:42 +0000[diff] [blame]574</ul>
Chris Lattner10a3cc42010-03-17 06:41:58 +0000[diff] [blame]575
Chris Lattner97beb512007-05-14 06:56:09 +0000[diff] [blame]576</div>
577
Chris Lattner0a1fd102007-09-21 03:54:09 +0000[diff] [blame]578<!--=========================================================================-->
579<div class="doc_subsection">
Chris Lattnerd1094e02009-03-02 02:37:32 +0000[diff] [blame]580<a name="coreimprovements">LLVM IR and Core Improvements</a>
Chris Lattnerdd6acc02008-02-10 08:17:19 +0000[diff] [blame]581</div>
582
583<div class="doc_text">
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]584<p>LLVM IR has several new features for better support of new targets and that
585expose new optimization opportunities:</p>
Chris Lattnerdd6acc02008-02-10 08:17:19 +0000[diff] [blame]586
Chris Lattnerb7112222008-06-05 06:25:56 +0000[diff] [blame]587<ul>
Chris Lattner656db162010-04-22 05:41:35 +0000[diff] [blame]588<li>LLVM IR now supports a 16-bit "half float" data type through <a
589 href="LangRef.html#int_fp16">two new intrinsics</a> and APFloat support.</li>
590<li>LLVM IR supports two new <a href="LangRef.html#fnattrs">function
Chris Lattner1dcca8b2010-04-21 06:23:40 +0000[diff] [blame]591 attributes</a>: inlinehint and alignstack(n). The former is a hint to the
Chris Lattner656db162010-04-22 05:41:35 +0000[diff] [blame]592 optimizer that a function was declared 'inline' and thus the inliner should
593 weight it higher when considering inlining it. The later
Chris Lattner1dcca8b2010-04-21 06:23:40 +0000[diff] [blame]594 indicates to the code generator that the function diverges from the platform
595 ABI on stack alignment.</li>
596<li>The new <a href="LangRef.html#int_objectsize">llvm.objectsize</a> intrinsic
597 allows the optimizer to infer the sizes of memory objects in some cases.
Chris Lattner656db162010-04-22 05:41:35 +0000[diff] [blame]598 This intrinsic is used to implement the GCC <tt>__builtin_object_size</tt>
Chris Lattner1dcca8b2010-04-21 06:23:40 +0000[diff] [blame]599 extension.</li>
Chris Lattnera67df2d2010-04-22 06:28:20 +0000[diff] [blame]600<li>LLVM IR now supports marking load and store instructions with <a
601 href="LangRef.html#i_load">"non-temporal" hints</a> (building on the new
602 metadata feature). This hint encourages the code
603 generator to generate non-temporal accesses when possible, which are useful
604 for code that is carefully managing cache behavior. Currently, only the
605 X86 backend provides target support for this feature.</li>
606
Chris Lattner1dcca8b2010-04-21 06:23:40 +0000[diff] [blame]607<li>LLVM 2.7 has pre-alpha support for <a
Gabor Greif862b3532010-04-22 10:11:24 +0000[diff] [blame]608 href="LangRef.html#t_union">unions in LLVM IR</a>.
Chris Lattnera67df2d2010-04-22 06:28:20 +0000[diff] [blame]609 Unfortunately, this support is not really usable in 2.7, so if you're
Chris Lattner1dcca8b2010-04-21 06:23:40 +0000[diff] [blame]610 interested in pushing it forward, please help contribute to LLVM mainline.</li>
Gabor Greif27b166352010-04-22 10:21:43 +0000[diff] [blame]611
Chris Lattnerdd6acc02008-02-10 08:17:19 +0000[diff] [blame]612</ul>
Mikhail Glushenkov024f7cf2008-10-13 02:08:34 +0000[diff] [blame]613
Chris Lattnerdd6acc02008-02-10 08:17:19 +0000[diff] [blame]614</div>
615
616<!--=========================================================================-->
617<div class="doc_subsection">
Chris Lattner2b8a52e2008-02-10 07:46:44 +0000[diff] [blame]618<a name="optimizer">Optimizer Improvements</a>
619</div>
620
621<div class="doc_text">
622
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]623<p>In addition to a large array of minor performance tweaks and bug fixes, this
Chris Lattner25879d72008-10-13 21:50:36 +0000[diff] [blame]624release includes a few major enhancements and additions to the optimizers:</p>
Chris Lattner2b8a52e2008-02-10 07:46:44 +0000[diff] [blame]625
626<ul>
627
Chris Lattner656db162010-04-22 05:41:35 +0000[diff] [blame]628<li>The inliner reuses now merges arrays stack objects in different callees when
629 inlining multiple call sites into one function. This reduces the stack size
630 of the resultant function.</li>
631<li>The -basicaa alias analysis pass (which is the default) has been improved to
632 be less dependent on "type safe" pointers. It can now look through bitcasts
633 and other constructs more aggressively, allowing better load/store
634 optimization.</li>
635<li>The load elimination optimization in the GVN Pass [<a
636href="http://blog.llvm.org/2009/12/introduction-to-load-elimination-in-gvn.html">intro
637 blog post</a>] has been substantially improved to be more aggressive about
638 partial redundancy elimination and do more aggressive phi translation. Please
639 see the <a
640 href="http://blog.llvm.org/2009/12/advanced-topics-in-redundant-load.html">
641 Advanced Topics in Redundant Load Elimination with a Focus on PHI Translation
642 Blog Post</a> for more details.</li>
643<li>The module <a href="LangRef.html#datalayout">target data string</a> now
Wesley Peckc8459162010-04-22 13:28:34 +0000[diff] [blame]644 includes a notion of 'native' integer data types for the target. This
645 helps mid-level optimizations avoid promoting complex sequences of
646 operations to data types that are not natively supported (e.g. converting
647 i32 operations to i64 on 32-bit chips).</li>
Chris Lattner656db162010-04-22 05:41:35 +0000[diff] [blame]648<li>The mid-level optimizer is now conservative when operating on a module with
649 no target data. Previously, it would default to SparcV9 settings, which is
650 not what most people expected.</li>
Chris Lattnera668b452010-04-21 06:42:24 +0000[diff] [blame]651<li>Jump threading is now much more aggressive at simplifying correlated
Chris Lattner656db162010-04-22 05:41:35 +0000[diff] [blame]652 conditionals and threading blocks with otherwise complex logic. It has
653 subsumed the old "Conditional Propagation" pass, and -condprop has been
654 removed from LLVM 2.7.</li>
655<li>The -instcombine pass has been refactored from being one huge file to being
656 a library of its own. Internally, it uses a customized IRBuilder to clean
657 it up and simplify it.</li>
Chris Lattner87e04952010-04-21 05:17:40 +0000[diff] [blame]658
Chris Lattner656db162010-04-22 05:41:35 +0000[diff] [blame]659<li>The optimal edge profiling pass is reliable and much more complete than in
660 2.6. It can be used with the llvm-prof tool but isn't wired up to the
661 llvm-gcc and clang command line options yet.</li>
Chris Lattnerd1094e02009-03-02 02:37:32 +0000[diff] [blame]662
Chris Lattner656db162010-04-22 05:41:35 +0000[diff] [blame]663<li>A new experimental alias analysis implementation, -scev-aa, has been added.
664 It uses LLVM's Scalar Evolution implementation to do symbolic analysis of
665 pointer offset expressions to disambiguate pointers. It can catch a few
666 cases that basicaa cannot, particularly in complex loop nests.</li>
Chris Lattnera668b452010-04-21 06:42:24 +0000[diff] [blame]667
Chris Lattnera67df2d2010-04-22 06:28:20 +0000[diff] [blame]668<li>The default pass ordering has been tweaked for improved optimization
Chris Lattner656db162010-04-22 05:41:35 +0000[diff] [blame]669 effectiveness.</li>
Chris Lattnera668b452010-04-21 06:42:24 +0000[diff] [blame]670
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]671</ul>
Chris Lattnerd1094e02009-03-02 02:37:32 +0000[diff] [blame]672
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]673</div>
Chris Lattnerd1094e02009-03-02 02:37:32 +0000[diff] [blame]674
Chris Lattnerd1094e02009-03-02 02:37:32 +0000[diff] [blame]675
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]676<!--=========================================================================-->
677<div class="doc_subsection">
678<a name="executionengine">Interpreter and JIT Improvements</a>
679</div>
Chris Lattner25879d72008-10-13 21:50:36 +0000[diff] [blame]680
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]681<div class="doc_text">
682
683<ul>
Wesley Peckba1ad202010-04-22 13:36:27 +0000[diff] [blame]684<li>The JIT now supports generating debug information and is compatible with
685the new GDB 7.0 (and later) interfaces for registering dynamically generated
686debug info.</li>
Chris Lattner87e04952010-04-21 05:17:40 +0000[diff] [blame]687
Chris Lattnera67df2d2010-04-22 06:28:20 +0000[diff] [blame]688<li>The JIT now <a href="http://llvm.org/PR5184">defaults
Jeffrey Yasskin9fb8ce82010-01-29 19:10:38 +0000[diff] [blame]689to compiling eagerly</a> to avoid a race condition in the lazy JIT.
690Clients that still want the lazy JIT can switch it on by calling
691<tt>ExecutionEngine::DisableLazyCompilation(false)</tt>.</li>
Chris Lattner87e04952010-04-21 05:17:40 +0000[diff] [blame]692
Jeffrey Yasskine0913882010-02-11 01:07:39 +0000[diff] [blame]693<li>It is now possible to create more than one JIT instance in the same process.
694These JITs can generate machine code in parallel,
695although <a href="http://llvm.org/docs/ProgrammersManual.html#jitthreading">you
696still have to obey the other threading restrictions</a>.</li>
Chris Lattner87e04952010-04-21 05:17:40 +0000[diff] [blame]697
Chris Lattner7795ea92008-06-08 02:45:07 +0000[diff] [blame]698</ul>
699
700</div>
701
702<!--=========================================================================-->
703<div class="doc_subsection">
Chris Lattnerd434bfb2009-03-02 03:24:11 +0000[diff] [blame]704<a name="codegen">Target Independent Code Generator Improvements</a>
Chris Lattner7795ea92008-06-08 02:45:07 +0000[diff] [blame]705</div>
706
707<div class="doc_text">
708
Mikhail Glushenkov25422542009-03-01 18:09:47 +0000[diff] [blame]709<p>We have put a significant amount of work into the code generator
710infrastructure, which allows us to implement more aggressive algorithms and make
711it run faster:</p>
Chris Lattner7795ea92008-06-08 02:45:07 +0000[diff] [blame]712
713<ul>
Chris Lattnera67df2d2010-04-22 06:28:20 +0000[diff] [blame]714<li>The 'llc -asm-verbose' option (which is now the default) has been enhanced
715 to emit many useful comments to .s files indicating information about spill
716 slots and loop nest structure. This should make it much easier to read and
717 understand assembly files. This is wired up in llvm-gcc and clang to
718 the <tt>-fverbose-asm</tt> option.</li>
719
Dan Gohman00365f92010-04-22 20:50:43 +0000[diff] [blame]720<li>New LSR with "full strength reduction" mode, which can reduce address
721 register pressure in loops where address generation is important.</li>
Chris Lattnera67df2d2010-04-22 06:28:20 +0000[diff] [blame]722
723<li>A new codegen level Common Subexpression Elimination pass (MachineCSE)
724 is available and enabled by default. It catches redundancies exposed by
725 lowering.</li>
726<li>A new pre-register-allocation tail duplication pass is available and enabled
727 by default, it can substantially improve branch prediction quality in some
728 cases.</li>
729<li>A new sign and zero extension optimization pass (OptimizeExtsPass)
730 is available and enabled by default. This pass can takes advantage
731 architecture features like x86-64 implicit zero extension behavior and
732 sub-registers.</li>
733<li>The code generator now supports a mode where it attempts to preserve the
734 order of instructions in the input code. This is important for source that
735 is hand scheduled and extremely sensitive to scheduling. It is compatible
736 with the GCC <tt>-fno-schedule-insns</tt> option.</li>
737<li>The target-independent code generator now supports generating code with
738 arbitrary numbers of result values. Returning more values than was
739 previously supported is handled by returning through a hidden pointer. In
740 2.7, only the X86 and XCore targets have adopted support for this
741 though.</li>
742<li>The code generator now supports generating code that follows the
743 <a href="LangRef.html#callingconv">Glasgow Haskell Compiler Calling
744 Convention</a> and ABI.</li>
745<li>The "<a href="CodeGenerator.html#selectiondag_select">DAG instruction
746 selection</a>" phase of the code generator has been largely rewritten for
747 2.7. Previously, tblgen spit out tons of C++ code which was compiled and
748 linked into the target to do the pattern matching, now it emits a much
749 smaller table which is read by the target-independent code. The primary
750 advantages of this approach is that the size and compile time of various
751 targets is much improved. The X86 code generator shrunk by 1.5MB of code,
752 for example.</li>
753<li>Almost the entire code generator has switched to emitting code through the
754 MC interfaces instead of printing textually to the .s file. This led to a
755 number of cleanups and speedups. In 2.7, debug an exception handling
756 information does not go through MC yet.</li>
Chris Lattner0a1fd102007-09-21 03:54:09 +0000[diff] [blame]757</ul>
Chris Lattner0a1fd102007-09-21 03:54:09 +0000[diff] [blame]758</div>
759
Chris Lattnerd1094e02009-03-02 02:37:32 +0000[diff] [blame]760<!--=========================================================================-->
761<div class="doc_subsection">
762<a name="x86">X86-32 and X86-64 Target Improvements</a>
763</div>
764
765<div class="doc_text">
766<p>New features of the X86 target include:
767</p>
768
769<ul>
Chris Lattnera668b452010-04-21 06:42:24 +0000[diff] [blame]770<li>The X86 backend now optimizes tails calls much more aggressively for
771 functions that use the standard C calling convention.</li>
772<li>The X86 backend now models scalar SSE registers as subregs of the SSE vector
773 registers, making the code generator more aggressive in cases where scalars
774 and vector types are mixed.</li>
Chris Lattnerd434bfb2009-03-02 03:24:11 +0000[diff] [blame]775
Chris Lattnerd1094e02009-03-02 02:37:32 +0000[diff] [blame]776</ul>
777
Chris Lattnerd1094e02009-03-02 02:37:32 +0000[diff] [blame]778</div>
Chris Lattner0a1fd102007-09-21 03:54:09 +0000[diff] [blame]779
780<!--=========================================================================-->
781<div class="doc_subsection">
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]782<a name="ARM">ARM Target Improvements</a>
Chris Lattnerc92d7692009-03-01 02:30:21 +0000[diff] [blame]783</div>
784
785<div class="doc_text">
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]786<p>New features of the ARM target include:
787</p>
Chris Lattnerc92d7692009-03-01 02:30:21 +0000[diff] [blame]788
789<ul>
Chris Lattnerc92d7692009-03-01 02:30:21 +0000[diff] [blame]790
Chris Lattner1dcca8b2010-04-21 06:23:40 +0000[diff] [blame]791<li>The ARM backend now generates instructions in unified assembly syntax.</li>
792
793<li>llvm-gcc now has complete support for the ARM v7 NEON instruction set. This
794 support differs slightly from the GCC implementation. Please see the
795 <a
796href="http://blog.llvm.org/2010/04/arm-advanced-simd-neon-intrinsics-and.html">
797 ARM Advanced SIMD (NEON) Intrinsics and Types in LLVM Blog Post</a> for
798 helpful information if migrating code from GCC to LLVM-GCC.</li>
799
Wesley Peck16144fb2010-04-22 13:43:14 +0000[diff] [blame]800<li>The ARM and Thumb code generators now use register scavenging for stack
Jim Grosbach77f0f5b2010-04-22 18:28:43 +0000[diff] [blame]801 object address materialization. This allows the use of R3 as a general
802 purpose register in Thumb1 code, as it was previous reserved for use in
803 stack address materialization. Secondly, sequential uses of the same
804 value will now re-use the materialized constant.</li>
805
Wesley Peck16144fb2010-04-22 13:43:14 +0000[diff] [blame]806<li>The ARM backend now has good support for ARMv4 targets and has been tested
Chris Lattner1dcca8b2010-04-21 06:23:40 +0000[diff] [blame]807 on StrongARM hardware. Previously, LLVM only supported ARMv4T and
808 newer chips.</li>
Jim Grosbach96284682010-04-22 18:33:31 +0000[diff] [blame]809
810<li>Atomic builtins are now supported for ARMv6 and ARMv7 (__sync_synchronize,
811 __sync_fetch_and_add, etc.).</li>
812
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]813</ul>
Chris Lattnerc92d7692009-03-01 02:30:21 +0000[diff] [blame]814
Chris Lattnerc92d7692009-03-01 02:30:21 +0000[diff] [blame]815
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]816</div>
Chris Lattnerc92d7692009-03-01 02:30:21 +0000[diff] [blame]817
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]818<!--=========================================================================-->
819<div class="doc_subsection">
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]820<a name="newapis">New Useful APIs</a>
821</div>
822
823<div class="doc_text">
824
825<p>This release includes a number of new APIs that are used internally, which
826 may also be useful for external clients.
827</p>
828
829<ul>
Chris Lattner656db162010-04-22 05:41:35 +0000[diff] [blame]830<li>The optimizer uses the new CodeMetrics class to measure the size of code.
Chris Lattner77a23802010-04-22 17:39:38 +0000[diff] [blame]831 Various passes (like the inliner, loop unswitcher, etc) all use this to make
832 more accurate estimates of the code size impact of various
Chris Lattner656db162010-04-22 05:41:35 +0000[diff] [blame]833 optimizations.</li>
834<li>A new <a href="http://llvm.org/doxygen/InstructionSimplify_8h-source.html">
Wesley Peck3edc1682010-04-22 13:50:46 +0000[diff] [blame]835 llvm/Analysis/InstructionSimplify.h</a> interface is available for doing
Chris Lattner656db162010-04-22 05:41:35 +0000[diff] [blame]836 symbolic simplification of instructions (e.g. <tt>a+0</tt> -&gt; <tt>a</tt>)
837 without requiring the instruction to exist. This centralizes a lot of
838 ad-hoc symbolic manipulation code scattered in various passes.</li>
839<li>The optimizer now uses a new <a
840 href="http://llvm.org/doxygen/SSAUpdater_8h-source.html">SSAUpdater</a>
841 class which efficiently supports
842 doing unstructured SSA update operations. This centralized a bunch of code
Wesley Peck3edc1682010-04-22 13:50:46 +0000[diff] [blame]843 scattered throughout various passes (e.g. jump threading, lcssa,
844 loop rotate, etc) for doing this sort of thing. The code generator has a
845 similar <a href="http://llvm.org/doxygen/MachineSSAUpdater_8h-source.html">
Chris Lattner656db162010-04-22 05:41:35 +0000[diff] [blame]846 MachineSSAUpdater</a> class.</li>
847<li>The <a href="http://llvm.org/doxygen/Regex_8h-source.html">
848 llvm/Support/Regex.h</a> header exposes a platform independent regular
Chris Lattner1dcca8b2010-04-21 06:23:40 +0000[diff] [blame]849 expression API. Building on this, the <a
850 href="TestingGuide.html#FileCheck">FileCheck</a> utility now supports
851 regular exressions.</li>
852<li>raw_ostream now supports a circular "debug stream" accessed with "dbgs()".
853 By default, this stream works the same way as "errs()", but if you pass
854 <tt>-debug-buffer-size=1000</tt> to opt, the debug stream is capped to a
855 fixed sized circular buffer and the output is printed at the end of the
856 program's execution. This is helpful if you have a long lived compiler
857 process and you're interested in seeing snapshots in time.</li>
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]858</ul>
859
860
861</div>
862
863<!--=========================================================================-->
864<div class="doc_subsection">
865<a name="otherimprovements">Other Improvements and New Features</a>
866</div>
867
868<div class="doc_text">
869<p>Other miscellaneous features include:</p>
870
871<ul>
Chris Lattner1dcca8b2010-04-21 06:23:40 +0000[diff] [blame]872<li>You can now build LLVM as a big dynamic library (e.g. "libllvm2.7.so"). To
873 get this, configure LLVM with the --enable-shared option.</li>
874
Wesley Peckcad613d2010-04-22 13:53:54 +0000[diff] [blame]875<li>LLVM command line tools now overwrite their output by default. Previously,
876 they would only do this with -f. This makes them more convenient to use, and
Chris Lattner87e04952010-04-21 05:17:40 +0000[diff] [blame]877 behave more like standard unix tools.</li>
878
879<li>The opt and llc tools now autodetect whether their input is a .ll or .bc
880 file, and automatically do the right thing. This means you don't need to
881 explicitly use the llvm-as tool for most things.</li>
Chris Lattnerc92d7692009-03-01 02:30:21 +0000[diff] [blame]882</ul>
883
884</div>
885
Chris Lattner6cb64032008-06-05 08:02:49 +0000[diff] [blame]886
887<!--=========================================================================-->
888<div class="doc_subsection">
Chris Lattner1e4d5bc2008-10-13 17:57:36 +0000[diff] [blame]889<a name="changes">Major Changes and Removed Features</a>
890</div>
891
892<div class="doc_text">
893
Chris Lattner934e2d42008-10-13 22:06:31 +0000[diff] [blame]894<p>If you're already an LLVM user or developer with out-of-tree changes based
Jeffrey Yasskin0830b972010-01-28 01:14:43 +0000[diff] [blame]895on LLVM 2.6, this section lists some "gotchas" that you may run into upgrading
Chris Lattner934e2d42008-10-13 22:06:31 +0000[diff] [blame]896from the previous release.</p>
Chris Lattner1e4d5bc2008-10-13 17:57:36 +0000[diff] [blame]897
898<ul>
Chris Lattner87e04952010-04-21 05:17:40 +0000[diff] [blame]899
900<li>
901The Andersen's alias analysis ("anders-aa") pass, the Predicate Simplifier
902("predsimplify") pass, the LoopVR pass, the GVNPRE pass, and the random sampling
903profiling ("rsprofiling") passes have all been removed. They were not being
904actively maintained and had substantial problems. If you are interested in
905these components, you are welcome to ressurect them from SVN, fix the
906correctness problems, and resubmit them to mainline.</li>
907
908<li>LLVM now defaults to building most libraries with RTTI turned off, providing
909a code size reduction. Packagers who are interested in building LLVM to support
910plugins that require RTTI information should build with "make REQUIRE_RTTI=1"
911and should read the new <a href="Packaging.html">Advice on Packaging LLVM</a>
912document.</li>
913
Jeffrey Yasskin914050b2010-02-09 23:03:44 +0000[diff] [blame]914<li>The LLVM interpreter now defaults to <em>not</em> using <tt>libffi</tt> even
915if you have it installed. This makes it more likely that an LLVM built on one
916system will work when copied to a similar system. To use <tt>libffi</tt>,
Chris Lattnera67df2d2010-04-22 06:28:20 +0000[diff] [blame]917configure with <tt>--enable-libffi</tt>.</li>
Chris Lattner934e2d42008-10-13 22:06:31 +0000[diff] [blame]918
Chris Lattner87e04952010-04-21 05:17:40 +0000[diff] [blame]919<li>Debug information uses a completely different representation, an LLVM 2.6
920.bc file should work with LLVM 2.7, but debug info won't come forward.</li>
921
922<li>The LLVM 2.6 (and earlier) "malloc" and "free" instructions got removed,
923 along with LowerAllocations pass. Now you should just use a call to the
924 malloc and free functions in libc. These calls are optimized as well as
925 the old instructions were.</li>
926</ul>
Chris Lattner934e2d42008-10-13 22:06:31 +0000[diff] [blame]927
928<p>In addition, many APIs have changed in this release. Some of the major LLVM
929API changes are:</p>
930
931<ul>
Gabor Greif137d90c2010-04-26 20:46:03 +0000[diff] [blame^]932<li>Just about everything has been converted to use <tt>raw_ostream</tt> instead of
933 <tt>std::ostream</tt>.</li>
934<li><tt>llvm/ADT/iterator.h</tt> has been removed, just use <tt>&lt;iterator&gt;</tt>
Chris Lattner87e04952010-04-21 05:17:40 +0000[diff] [blame]935 instead.</li>
Gabor Greif137d90c2010-04-26 20:46:03 +0000[diff] [blame^]936<li>The <tt>Streams.h</tt> file and <tt>DOUT</tt> got removed, use <tt>DEBUG(errs() &lt;&lt; ...);</tt>
Chris Lattner87e04952010-04-21 05:17:40 +0000[diff] [blame]937 instead.</li>
Gabor Greif137d90c2010-04-26 20:46:03 +0000[diff] [blame^]938<li>The <tt>TargetAsmInfo</tt> interface was renamed to <tt>MCAsmInfo</tt>.</li>
Jeffrey Yasskin4a8b1da2010-01-28 01:41:20 +0000[diff] [blame]939<li><tt>ModuleProvider</tt> has been <a
Gabor Greif27b166352010-04-22 10:21:43 +0000[diff] [blame]940href="http://llvm.org/viewvc/llvm-project?view=rev&amp;revision=94686">removed</a>
Jeffrey Yasskin4a8b1da2010-01-28 01:41:20 +0000[diff] [blame]941and its methods moved to <tt>Module</tt> and <tt>GlobalValue</tt>.
942Most clients can remove uses of <tt>ExistingModuleProvider</tt>,
943replace <tt>getBitcodeModuleProvider</tt> with
944<tt>getLazyBitcodeModule</tt>, and pass their <tt>Module</tt> to
945functions that used to accept <tt>ModuleProvider</tt>. Clients who
946wrote their own <tt>ModuleProvider</tt>s will need to derive from
947<tt>GVMaterializer</tt> instead and use
948<tt>Module::setMaterializer</tt> to attach it to a
949<tt>Module</tt>.</li>
950
951<li><tt>GhostLinkage</tt> has given up the ghost.
952<tt>GlobalValue</tt>s that have not yet been read from their backing
953storage have the same linkage they will have after being read in.
954Clients must replace calls to
955<tt>GlobalValue::hasNotBeenReadFromBitcode</tt> with
956<tt>GlobalValue::isMaterializable</tt>.</li>
Daniel Dunbar1e96fca2010-02-10 04:09:52 +0000[diff] [blame]957
Duncan Sands80c430a2010-02-17 17:20:17 +0000[diff] [blame]958<li>The <tt>isInteger</tt>, <tt>isIntOrIntVector</tt>, <tt>isFloatingPoint</tt>,
959<tt>isFPOrFPVector</tt> and <tt>isFPOrFPVector</tt> methods have been renamed
960<tt>isIntegerTy</tt>, <tt>isIntOrIntVectorTy</tt>, <tt>isFloatingPointTy</tt>,
961<tt>isFPOrFPVectorTy</tt> and <tt>isFPOrFPVectorTy</tt> respectively.</li>
Chris Lattnercecb9232010-04-22 17:25:00 +0000[diff] [blame]962
963<li><tt>llvm::Instruction::clone()</tt> no longer takes argument.</li>
964<li><tt>raw_fd_ostream</tt>'s constructor now takes a flag argument, not individual
965 booleans (see <tt>include/llvm/Support/raw_ostream.h</tt> for details).</li>
966<li>Some header files have been renamed:
967<ul>
968 <li><tt>llvm/Support/AIXDataTypesFix.h</tt> to
969 <tt>llvm/System/AIXDataTypesFix.h</tt></li>
970 <li><tt>llvm/Support/DataTypes.h</tt> to <tt>llvm/System/DataTypes.h</tt></li>
971 <li><tt>llvm/Transforms/Utils/InlineCost.h</tt> to
972 <tt>llvm/Analysis/InlineCost.h</tt></li>
973 <li><tt>llvm/Support/Mangler.h</tt> to <tt>llvm/Target/Mangler.h</tt></li>
974 <li><tt>llvm/Analysis/Passes.h</tt> to <tt>llvm/CodeGen/Passes.h</tt></li>
975</ul></li>
Devang Pateldbf83832008-10-14 20:03:43 +0000[diff] [blame]976</ul>
Chris Lattner1e4d5bc2008-10-13 17:57:36 +0000[diff] [blame]977
Chris Lattner1e4d5bc2008-10-13 17:57:36 +0000[diff] [blame]978</div>
979
980
981
Chris Lattner62495762003-10-02 16:38:05 +0000[diff] [blame]982<!-- *********************************************************************** -->
Misha Brukman80731b92003-11-22 00:38:41 +0000[diff] [blame]983<div class="doc_section">
984 <a name="portability">Portability and Supported Platforms</a>
985</div>
Chris Lattner62495762003-10-02 16:38:05 +0000[diff] [blame]986<!-- *********************************************************************** -->
987
Misha Brukman80731b92003-11-22 00:38:41 +0000[diff] [blame]988<div class="doc_text">
989
John Criswell64307722004-12-08 20:35:47 +0000[diff] [blame]990<p>LLVM is known to work on the following platforms:</p>
Chris Lattner2c7b0872004-06-01 18:22:41 +0000[diff] [blame]991
992<ul>
Mikhail Glushenkov25422542009-03-01 18:09:47 +0000[diff] [blame]993<li>Intel and AMD machines (IA32, X86-64, AMD64, EMT-64) running Red Hat
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]994 Linux, Fedora Core, FreeBSD and AuroraUX (and probably other unix-like
995 systems).</li>
Chris Lattnera67df2d2010-04-22 06:28:20 +0000[diff] [blame]996<li>PowerPC and X86-based Mac OS X systems, running 10.4 and above in 32-bit
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]997 and 64-bit modes.</li>
Chris Lattnercdc44ed2008-02-06 18:00:06 +0000[diff] [blame]998<li>Intel and AMD machines running on Win32 using MinGW libraries (native).</li>
Chris Lattnerbee7b322007-05-23 04:39:32 +0000[diff] [blame]999<li>Intel and AMD machines running on Win32 with the Cygwin libraries (limited
1000 support is available for native builds with Visual C++).</li>
Chris Lattner51948622009-10-16 16:30:58 +0000[diff] [blame]1001<li>Sun x86 and AMD64 machines running Solaris 10, OpenSolaris 0906.</li>
John Criswelld5234922005-05-13 20:28:15 +0000[diff] [blame]1002<li>Alpha-based machines running Debian GNU/Linux.</li>
Chris Lattner2c7b0872004-06-01 18:22:41 +0000[diff] [blame]1003</ul>
1004
Chris Lattner618a99f2008-06-05 06:57:39 +0000[diff] [blame]1005<p>The core LLVM infrastructure uses GNU autoconf to adapt itself
Brian Gaekee3e557e2004-05-09 05:28:35 +0000[diff] [blame]1006to the machine and operating system on which it is built. However, minor
1007porting may be required to get LLVM to work on new platforms. We welcome your
1008portability patches and reports of successful builds or error messages.</p>
Chris Lattner3d482502003-10-02 04:57:28 +0000[diff] [blame]1009
Misha Brukman80731b92003-11-22 00:38:41 +0000[diff] [blame]1010</div>
Chris Lattner3d482502003-10-02 04:57:28 +0000[diff] [blame]1011
1012<!-- *********************************************************************** -->
Misha Brukman80731b92003-11-22 00:38:41 +0000[diff] [blame]1013<div class="doc_section">
1014 <a name="knownproblems">Known Problems</a>
1015</div>
Chris Lattner3d482502003-10-02 04:57:28 +0000[diff] [blame]1016<!-- *********************************************************************** -->
1017
Misha Brukman80731b92003-11-22 00:38:41 +0000[diff] [blame]1018<div class="doc_text">
1019
Mikhail Glushenkov25422542009-03-01 18:09:47 +0000[diff] [blame]1020<p>This section contains significant known problems with the LLVM system,
Chris Lattner2a092392008-11-10 05:40:34 +0000[diff] [blame]1021listed by component. If you run into a problem, please check the <a
Chris Lattnera69595e2005-10-29 07:07:09 +0000[diff] [blame]1022href="http://llvm.org/bugs/">LLVM bug database</a> and submit a bug if
Chris Lattnerb84f3322003-12-12 21:22:16 +0000[diff] [blame]1023there isn't already one.</p>
Chris Lattner3d482502003-10-02 04:57:28 +0000[diff] [blame]1024
Chris Lattner6aba4862010-03-17 17:25:49 +0000[diff] [blame]1025<ul>
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]1026<li>LLVM will not correctly compile on Solaris and/or OpenSolaris
1027using the stock GCC 3.x.x series 'out the box',
Chris Lattner6e41c792009-11-03 21:50:09 +0000[diff] [blame]1028See: <a href="GettingStarted.html#brokengcc">Broken versions of GCC and other tools</a>.
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]1029However, A <a href="http://pkg.auroraux.org/GCC">Modern GCC Build</a>
1030for x86/x86-64 has been made available from the third party AuroraUX Project
1031that has been meticulously tested for bootstrapping LLVM &amp; Clang.</li>
1032</ul>
1033
Misha Brukman80731b92003-11-22 00:38:41 +0000[diff] [blame]1034</div>
Chris Lattner3d482502003-10-02 04:57:28 +0000[diff] [blame]1035
Chris Lattnerb911de42004-03-14 02:03:02 +0000[diff] [blame]1036<!-- ======================================================================= -->
1037<div class="doc_subsection">
1038 <a name="experimental">Experimental features included with this release</a>
1039</div>
Chris Lattner3d482502003-10-02 04:57:28 +0000[diff] [blame]1040
Chris Lattnerb911de42004-03-14 02:03:02 +0000[diff] [blame]1041<div class="doc_text">
1042
Misha Brukmanfa50a222004-05-12 21:46:05 +0000[diff] [blame]1043<p>The following components of this LLVM release are either untested, known to
1044be broken or unreliable, or are in early development. These components should
1045not be relied on, and bugs should not be filed against them, but they may be
1046useful to some people. In particular, if you would like to work on one of these
Chris Lattnere38ac152008-02-12 06:29:45 +0000[diff] [blame]1047components, please contact us on the <a
1048href="http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev">LLVMdev list</a>.</p>
Chris Lattnerb911de42004-03-14 02:03:02 +0000[diff] [blame]1049
1050<ul>
Wesley Peck3aa84e52010-03-18 14:31:30 +0000[diff] [blame]1051<li>The MSIL, Alpha, SPU, MIPS, PIC16, Blackfin, MSP430, SystemZ and MicroBlaze
1052 backends are experimental.</li>
Chris Lattnera67df2d2010-04-22 06:28:20 +0000[diff] [blame]1053<li><tt>llc</tt> "<tt>-filetype=asm</tt>" (the default) is the only
Chris Lattner0b1c9a52010-03-17 04:41:49 +0000[diff] [blame]1054 supported value for this option. The MachO writer is experimental, and
1055 works much better in mainline SVN.</li>
Chris Lattnerb911de42004-03-14 02:03:02 +0000[diff] [blame]1056</ul>
1057
1058</div>
Chris Lattner3d482502003-10-02 04:57:28 +0000[diff] [blame]1059
Misha Brukman80731b92003-11-22 00:38:41 +0000[diff] [blame]1060<!-- ======================================================================= -->
1061<div class="doc_subsection">
Chris Lattnerb81f10e2006-11-18 07:51:14 +0000[diff] [blame]1062 <a name="x86-be">Known problems with the X86 back-end</a>
John Criswell3bdbd302005-11-08 21:11:33 +0000[diff] [blame]1063</div>
1064
1065<div class="doc_text">
1066
1067<ul>
Anton Korobeynikov486c7d32008-06-08 10:24:13 +0000[diff] [blame]1068 <li>The X86 backend does not yet support
1069 all <a href="http://llvm.org/PR879">inline assembly that uses the X86
1070 floating point stack</a>. It supports the 'f' and 't' constraints, but not
1071 'u'.</li>
Chris Lattnerb076b002008-06-08 21:19:07 +0000[diff] [blame]1072 <li>The X86 backend generates inefficient floating point code when configured
1073 to generate code for systems that don't have SSE2.</li>
Duncan Sands0bc15262008-06-08 19:38:43 +0000[diff] [blame]1074 <li>Win64 code generation wasn't widely tested. Everything should work, but we
John Criswellea03c9d2009-03-02 15:28:15 +0000[diff] [blame]1075 expect small issues to happen. Also, llvm-gcc cannot build the mingw64
Chris Lattnera67df2d2010-04-22 06:28:20 +0000[diff] [blame]1076 runtime currently due to lack of support for the 'u' inline assembly
1077 constraint and for X87 floating point inline assembly.</li>
Dan Gohman721b3722008-06-08 23:05:11 +0000[diff] [blame]1078 <li>The X86-64 backend does not yet support the LLVM IR instruction
Chris Lattnera67df2d2010-04-22 06:28:20 +0000[diff] [blame]1079 <tt>va_arg</tt>. Currently, front-ends support variadic
Dan Gohman721b3722008-06-08 23:05:11 +0000[diff] [blame]1080 argument constructs on X86-64 by lowering them manually.</li>
Chris Lattnerb81f10e2006-11-18 07:51:14 +0000[diff] [blame]1081</ul>
1082
1083</div>
1084
1085<!-- ======================================================================= -->
1086<div class="doc_subsection">
1087 <a name="ppc-be">Known problems with the PowerPC back-end</a>
1088</div>
1089
1090<div class="doc_text">
1091
1092<ul>
Nicolas Geoffray77d99502007-05-15 09:21:28 +0000[diff] [blame]1093<li>The Linux PPC32/ABI support needs testing for the interpreter and static
Chris Lattnerbee7b322007-05-23 04:39:32 +0000[diff] [blame]1094compilation, and lacks support for debug information.</li>
Chris Lattnerb81f10e2006-11-18 07:51:14 +0000[diff] [blame]1095</ul>
1096
1097</div>
1098
1099<!-- ======================================================================= -->
1100<div class="doc_subsection">
Chris Lattner97beb512007-05-14 06:56:09 +0000[diff] [blame]1101 <a name="arm-be">Known problems with the ARM back-end</a>
1102</div>
1103
1104<div class="doc_text">
1105
1106<ul>
Chris Lattnerbee7b322007-05-23 04:39:32 +0000[diff] [blame]1107<li>Thumb mode works only on ARMv6 or higher processors. On sub-ARMv6
Duncan Sands47fc0a22007-09-26 15:59:54 +0000[diff] [blame]1108processors, thumb programs can crash or produce wrong
Chris Lattnerbee7b322007-05-23 04:39:32 +0000[diff] [blame]1109results (<a href="http://llvm.org/PR1388">PR1388</a>).</li>
John Criswellea03c9d2009-03-02 15:28:15 +0000[diff] [blame]1110<li>Compilation for ARM Linux OABI (old ABI) is supported but not fully tested.
Chris Lattner97beb512007-05-14 06:56:09 +0000[diff] [blame]1111</li>
Chris Lattner97beb512007-05-14 06:56:09 +0000[diff] [blame]1112</ul>
1113
1114</div>
1115
1116<!-- ======================================================================= -->
1117<div class="doc_subsection">
Chris Lattnerb81f10e2006-11-18 07:51:14 +0000[diff] [blame]1118 <a name="sparc-be">Known problems with the SPARC back-end</a>
1119</div>
1120
1121<div class="doc_text">
1122
1123<ul>
John Criswellea03c9d2009-03-02 15:28:15 +0000[diff] [blame]1124<li>The SPARC backend only supports the 32-bit SPARC ABI (-m32); it does not
Chris Lattnerb81f10e2006-11-18 07:51:14 +0000[diff] [blame]1125 support the 64-bit SPARC ABI (-m64).</li>
1126</ul>
1127
1128</div>
1129
1130<!-- ======================================================================= -->
1131<div class="doc_subsection">
Bruno Cardoso Lopes24eb3de2008-10-25 14:56:26 +0000[diff] [blame]1132 <a name="mips-be">Known problems with the MIPS back-end</a>
1133</div>
1134
1135<div class="doc_text">
1136
1137<ul>
Bruno Cardoso Lopes24eb3de2008-10-25 14:56:26 +0000[diff] [blame]1138<li>64-bit MIPS targets are not supported yet.</li>
1139</ul>
1140
1141</div>
1142
1143<!-- ======================================================================= -->
1144<div class="doc_subsection">
Chris Lattnerb81f10e2006-11-18 07:51:14 +0000[diff] [blame]1145 <a name="alpha-be">Known problems with the Alpha back-end</a>
1146</div>
1147
1148<div class="doc_text">
1149
1150<ul>
1151
1152<li>On 21164s, some rare FP arithmetic sequences which may trap do not have the
1153appropriate nops inserted to ensure restartability.</li>
1154
John Criswell3bdbd302005-11-08 21:11:33 +0000[diff] [blame]1155</ul>
1156</div>
1157
Chris Lattnerb81f10e2006-11-18 07:51:14 +0000[diff] [blame]1158<!-- ======================================================================= -->
1159<div class="doc_subsection">
Chris Lattner97beb512007-05-14 06:56:09 +0000[diff] [blame]1160 <a name="c-be">Known problems with the C back-end</a>
Chris Lattnerb81f10e2006-11-18 07:51:14 +0000[diff] [blame]1161</div>
1162
1163<div class="doc_text">
1164
1165<ul>
Chris Lattnera1a4c9a2008-06-05 06:35:40 +0000[diff] [blame]1166<li><a href="http://llvm.org/PR802">The C backend has only basic support for
1167 inline assembly code</a>.</li>
Chris Lattner8e061162007-09-26 06:01:35 +0000[diff] [blame]1168<li><a href="http://llvm.org/PR1658">The C backend violates the ABI of common
1169 C++ programs</a>, preventing intermixing between C++ compiled by the CBE and
Gabor Greif75b2f762009-03-02 12:02:51 +0000[diff] [blame]1170 C++ code compiled with <tt>llc</tt> or native compilers.</li>
Duncan Sandse09506a2008-02-10 13:40:55 +0000[diff] [blame]1171<li>The C backend does not support all exception handling constructs.</li>
Duncan Sands3aa36732009-02-25 11:51:54 +0000[diff] [blame]1172<li>The C backend does not support arbitrary precision integers.</li>
Chris Lattnerb81f10e2006-11-18 07:51:14 +0000[diff] [blame]1173</ul>
1174
1175</div>
John Criswell3bdbd302005-11-08 21:11:33 +0000[diff] [blame]1176
Misha Brukman80731b92003-11-22 00:38:41 +0000[diff] [blame]1177
1178<!-- ======================================================================= -->
1179<div class="doc_subsection">
Chris Lattner0b1c9a52010-03-17 04:41:49 +0000[diff] [blame]1180 <a name="c-fe">Known problems with the llvm-gcc C and C++ front-end</a>
Misha Brukman80731b92003-11-22 00:38:41 +0000[diff] [blame]1181</div>
Chris Lattner178f3db2003-10-02 05:07:23 +0000[diff] [blame]1182
Misha Brukman80731b92003-11-22 00:38:41 +0000[diff] [blame]1183<div class="doc_text">
Chris Lattner72a269f2006-03-03 00:34:26 +0000[diff] [blame]1184
Chris Lattnera1a4c9a2008-06-05 06:35:40 +0000[diff] [blame]1185<p>The only major language feature of GCC not supported by llvm-gcc is
1186 the <tt>__builtin_apply</tt> family of builtins. However, some extensions
1187 are only supported on some targets. For example, trampolines are only
Duncan Sands56448462008-06-08 20:18:35 +0000[diff] [blame]1188 supported on some targets (these are used when you take the address of a
1189 nested function).</p>
Misha Brukman80731b92003-11-22 00:38:41 +0000[diff] [blame]1190
Misha Brukman80731b92003-11-22 00:38:41 +0000[diff] [blame]1191</div>
Chris Lattner3d482502003-10-02 04:57:28 +0000[diff] [blame]1192
Misha Brukman80731b92003-11-22 00:38:41 +0000[diff] [blame]1193<!-- ======================================================================= -->
1194<div class="doc_subsection">
Chris Lattner7292fb12008-10-30 03:58:13 +0000[diff] [blame]1195 <a name="fortran-fe">Known problems with the llvm-gcc Fortran front-end</a>
1196</div>
1197
1198<div class="doc_text">
Gabor Greife5f9d672008-11-04 21:50:59 +0000[diff] [blame]1199<ul>
Chris Lattner7292fb12008-10-30 03:58:13 +0000[diff] [blame]1200<li>Fortran support generally works, but there are still several unresolved bugs
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]1201 in <a href="http://llvm.org/bugs/">Bugzilla</a>. Please see the
1202 tools/gfortran component for details.</li>
Gabor Greife5f9d672008-11-04 21:50:59 +0000[diff] [blame]1203</ul>
Chris Lattner7292fb12008-10-30 03:58:13 +0000[diff] [blame]1204</div>
Misha Brukman80731b92003-11-22 00:38:41 +0000[diff] [blame]1205
Chris Lattnere38ac152008-02-12 06:29:45 +0000[diff] [blame]1206<!-- ======================================================================= -->
1207<div class="doc_subsection">
Chris Lattnera1a4c9a2008-06-05 06:35:40 +0000[diff] [blame]1208 <a name="ada-fe">Known problems with the llvm-gcc Ada front-end</a>
Chris Lattnere38ac152008-02-12 06:29:45 +0000[diff] [blame]1209</div>
1210
1211<div class="doc_text">
John Criswellea03c9d2009-03-02 15:28:15 +0000[diff] [blame]1212The llvm-gcc 4.2 Ada compiler works fairly well; however, this is not a mature
1213technology, and problems should be expected.
Chris Lattnere38ac152008-02-12 06:29:45 +0000[diff] [blame]1214<ul>
Duncan Sands56448462008-06-08 20:18:35 +0000[diff] [blame]1215<li>The Ada front-end currently only builds on X86-32. This is mainly due
John Criswellea03c9d2009-03-02 15:28:15 +0000[diff] [blame]1216to lack of trampoline support (pointers to nested functions) on other platforms.
1217However, it <a href="http://llvm.org/PR2006">also fails to build on X86-64</a>
Chris Lattnere38ac152008-02-12 06:29:45 +0000[diff] [blame]1218which does support trampolines.</li>
1219<li>The Ada front-end <a href="http://llvm.org/PR2007">fails to bootstrap</a>.
Duncan Sandsd34bf762009-02-25 11:59:06 +0000[diff] [blame]1220This is due to lack of LLVM support for <tt>setjmp</tt>/<tt>longjmp</tt> style
1221exception handling, which is used internally by the compiler.
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]1222Workaround: configure with <tt>--disable-bootstrap</tt>.</li>
Duncan Sandsaef747f2008-10-13 17:27:23 +0000[diff] [blame]1223<li>The c380004, <a href="http://llvm.org/PR2010">c393010</a>
1224and <a href="http://llvm.org/PR2421">cxg2021</a> ACATS tests fail
Duncan Sandsd34bf762009-02-25 11:59:06 +0000[diff] [blame]1225(c380004 also fails with gcc-4.2 mainline).
1226If the compiler is built with checks disabled then <a href="http://llvm.org/PR2010">c393010</a>
1227causes the compiler to go into an infinite loop, using up all system memory.</li>
Duncan Sands2dbb00e2009-03-02 16:35:57 +0000[diff] [blame]1228<li>Some GCC specific Ada tests continue to crash the compiler.</li>
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]1229<li>The <tt>-E</tt> binder option (exception backtraces)
Chris Lattnere38ac152008-02-12 06:29:45 +0000[diff] [blame]1230<a href="http://llvm.org/PR1982">does not work</a> and will result in programs
Chris Lattneraa61f412009-10-13 17:48:04 +0000[diff] [blame]1231crashing if an exception is raised. Workaround: do not use <tt>-E</tt>.</li>
Chris Lattnere38ac152008-02-12 06:29:45 +0000[diff] [blame]1232<li>Only discrete types <a href="http://llvm.org/PR1981">are allowed to start
1233or finish at a non-byte offset</a> in a record. Workaround: do not pack records
1234or use representation clauses that result in a field of a non-discrete type
1235starting or finishing in the middle of a byte.</li>
Chris Lattnerb076b002008-06-08 21:19:07 +0000[diff] [blame]1236<li>The <tt>lli</tt> interpreter <a href="http://llvm.org/PR2009">considers
1237'main' as generated by the Ada binder to be invalid</a>.
1238Workaround: hand edit the file to use pointers for <tt>argv</tt> and
1239<tt>envp</tt> rather than integers.</li>
1240<li>The <tt>-fstack-check</tt> option <a href="http://llvm.org/PR2008">is
1241ignored</a>.</li>
Chris Lattnere38ac152008-02-12 06:29:45 +0000[diff] [blame]1242</ul>
1243</div>
1244
Chris Lattner3d482502003-10-02 04:57:28 +0000[diff] [blame]1245<!-- *********************************************************************** -->
Misha Brukman80731b92003-11-22 00:38:41 +0000[diff] [blame]1246<div class="doc_section">
1247 <a name="additionalinfo">Additional Information</a>
1248</div>
Chris Lattner3d482502003-10-02 04:57:28 +0000[diff] [blame]1249<!-- *********************************************************************** -->
1250
Misha Brukman80731b92003-11-22 00:38:41 +0000[diff] [blame]1251<div class="doc_text">
1252
Chris Lattnercb5596d2005-05-16 17:13:10 +0000[diff] [blame]1253<p>A wide variety of additional information is available on the <a
Chris Lattnere0c1df42007-05-18 00:44:29 +0000[diff] [blame]1254href="http://llvm.org">LLVM web page</a>, in particular in the <a
1255href="http://llvm.org/docs/">documentation</a> section. The web page also
Reid Spencerc7f87f22007-07-09 08:04:31 +0000[diff] [blame]1256contains versions of the API documentation which is up-to-date with the
1257Subversion version of the source code.
Misha Brukman96158092005-03-30 19:14:24 +0000[diff] [blame]1258You can access versions of these documents specific to this release by going
1259into the "<tt>llvm/doc/</tt>" directory in the LLVM tree.</p>
Chris Lattner3d482502003-10-02 04:57:28 +0000[diff] [blame]1260
Misha Brukman80731b92003-11-22 00:38:41 +0000[diff] [blame]1261<p>If you have any questions or comments about LLVM, please feel free to contact
Chris Lattnera69595e2005-10-29 07:07:09 +0000[diff] [blame]1262us via the <a href="http://llvm.org/docs/#maillist"> mailing
Chris Lattnerb84f3322003-12-12 21:22:16 +0000[diff] [blame]1263lists</a>.</p>
Chris Lattner3d482502003-10-02 04:57:28 +0000[diff] [blame]1264
Misha Brukman80731b92003-11-22 00:38:41 +0000[diff] [blame]1265</div>
Chris Lattner3d482502003-10-02 04:57:28 +0000[diff] [blame]1266
1267<!-- *********************************************************************** -->
Chris Lattner3d482502003-10-02 04:57:28 +0000[diff] [blame]1268
Misha Brukman80731b92003-11-22 00:38:41 +0000[diff] [blame]1269<hr>
Misha Brukman68aab3b2003-11-22 01:23:39 +0000[diff] [blame]1270<address>
Misha Brukman8834af92003-12-21 22:53:21 +0000[diff] [blame]1271 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
Misha Brukman86242e12008-12-11 17:34:48 +0000[diff] [blame]1272 src="http://jigsaw.w3.org/css-validator/images/vcss-blue" alt="Valid CSS"></a>
Misha Brukman68aab3b2003-11-22 01:23:39 +0000[diff] [blame]1273 <a href="http://validator.w3.org/check/referer"><img
Misha Brukman86242e12008-12-11 17:34:48 +0000[diff] [blame]1274 src="http://www.w3.org/Icons/valid-html401-blue" alt="Valid HTML 4.01"></a>
Misha Brukman68aab3b2003-11-22 01:23:39 +0000[diff] [blame]1275
Chris Lattnere0c1df42007-05-18 00:44:29 +0000[diff] [blame]1276 <a href="http://llvm.org/">LLVM Compiler Infrastructure</a><br>
Misha Brukman80731b92003-11-22 00:38:41 +0000[diff] [blame]1277 Last modified: $Date$
Misha Brukman68aab3b2003-11-22 01:23:39 +0000[diff] [blame]1278</address>
Chris Lattner3d482502003-10-02 04:57:28 +0000[diff] [blame]1279
Misha Brukman80731b92003-11-22 00:38:41 +0000[diff] [blame]1280</body>
1281</html>