Gitiles
Code Review Sign In
gerrit-public.fairphone.software / fp2-dev / platform / external / llvm / 7b2e71b53bf33a072fea725aa23a5dc6ee7d57db / . / docs / ReleaseNotes-2.6.html
blob: 4f1d85b802cadc32029017893342aab6a5c4a05d [file] [log] [blame]
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]1<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"
2 "http://www.w3.org/TR/html4/strict.dtd">
3<html>
4<head>
5 <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
6 <link rel="stylesheet" href="llvm.css" type="text/css">
7 <title>LLVM 2.6 Release Notes</title>
8</head>
9<body>
10
11<div class="doc_title">LLVM 2.6 Release Notes</div>
12
13<ol>
14 <li><a href="#intro">Introduction</a></li>
15 <li><a href="#subproj">Sub-project Status Update</a></li>
16 <li><a href="#externalproj">External Projects Using LLVM 2.6</a></li>
17 <li><a href="#whatsnew">What's New in LLVM 2.6?</a></li>
18 <li><a href="GettingStarted.html">Installation Instructions</a></li>
19 <li><a href="#portability">Portability and Supported Platforms</a></li>
20 <li><a href="#knownproblems">Known Problems</a></li>
21 <li><a href="#additionalinfo">Additional Information</a></li>
22</ol>
23
24<div class="doc_author">
25 <p>Written by the <a href="http://llvm.org">LLVM Team</a></p>
26</div>
27
28<!-- *********************************************************************** -->
29<div class="doc_section">
30 <a name="intro">Introduction</a>
31</div>
32<!-- *********************************************************************** -->
33
34<div class="doc_text">
35
36<p>This document contains the release notes for the LLVM Compiler
37Infrastructure, release 2.6. Here we describe the status of LLVM, including
38major improvements from the previous release and significant known problems.
39All LLVM releases may be downloaded from the <a
40href="http://llvm.org/releases/">LLVM releases web site</a>.</p>
41
42<p>For more information about LLVM, including information about the latest
43release, please check out the <a href="http://llvm.org/">main LLVM
44web site</a>. If you have questions or comments, the <a
45href="http://mail.cs.uiuc.edu/mailman/listinfo/llvmdev">LLVM Developer's Mailing
46List</a> is a good place to send them.</p>
47
48<p>Note that if you are reading this file from a Subversion checkout or the
49main LLVM web page, this document applies to the <i>next</i> release, not the
50current one. To see the release notes for a specific release, please see the
51<a href="http://llvm.org/releases/">releases page</a>.</p>
52
53</div>
Chris Lattner80ed2552009-10-08 07:01:46 +0000[diff] [blame]54
Chris Lattnerc758fec2009-10-05 02:12:39 +0000[diff] [blame]55
Chris Lattner7b0a6812009-10-08 06:27:53 +0000[diff] [blame]56<!--
57Almost dead code.
Chris Lattner80ed2552009-10-08 07:01:46 +0000[diff] [blame]58 include/llvm/Analysis/LiveValues.h => Dan
59 lib/Transforms/IPO/MergeFunctions.cpp => consider for 2.8.
60 llvm/Analysis/PointerTracking.h => Edwin wants this, consider for 2.8.
Chris Lattner7b0a6812009-10-08 06:27:53 +0000[diff] [blame]61-->
Chris Lattnerc758fec2009-10-05 02:12:39 +0000[diff] [blame]62
Chris Lattnerc758fec2009-10-05 02:12:39 +0000[diff] [blame]63
64<!-- Unfinished features in 2.6:
Chris Lattner80ed2552009-10-08 07:01:46 +0000[diff] [blame]65 gcc plugin.
Chris Lattnerc758fec2009-10-05 02:12:39 +0000[diff] [blame]66 strong phi elim
67 variable debug info for optimized code
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]68 postalloc scheduler: anti dependence breaking, hazard recognizer?
Chris Lattnerc758fec2009-10-05 02:12:39 +0000[diff] [blame]69 metadata
70 loop dependence analysis
Chris Lattner80ed2552009-10-08 07:01:46 +0000[diff] [blame]71 ELF Writer? How stable?
72 <li>PostRA scheduler improvements, ARM adoption (David Goodwin).</li>
Chris Lattner45178502009-10-09 06:24:25 +0000[diff] [blame]73 2.7 supports the GDB 7.0 jit interfaces for debug info.
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]74 -->
75
76 <!-- for announcement email:
Chris Lattnerc758fec2009-10-05 02:12:39 +0000[diff] [blame]77 Logo web page.
78 llvm devmtg
79 compiler_rt
Chris Lattner7b0a6812009-10-08 06:27:53 +0000[diff] [blame]80 klee web page at klee.llvm.org
81 Many new papers added to /pubs/
82 Mention gcc plugin.
83
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]84 -->
85
86<!-- *********************************************************************** -->
87<div class="doc_section">
88 <a name="subproj">Sub-project Status Update</a>
89</div>
90<!-- *********************************************************************** -->
91
92<div class="doc_text">
93<p>
94The LLVM 2.6 distribution currently consists of code from the core LLVM
95repository &mdash;which roughly includes the LLVM optimizers, code generators
96and supporting tools &mdash; and the llvm-gcc repository. In addition to this
97code, the LLVM Project includes other sub-projects that are in development. The
98two which are the most actively developed are the <a href="#clang">Clang
99Project</a> and the <a href="#vmkit">VMKit Project</a>.
100</p>
101
102</div>
103
104
105<!--=========================================================================-->
106<div class="doc_subsection">
107<a name="clang">Clang: C/C++/Objective-C Frontend Toolkit</a>
108</div>
109
110<div class="doc_text">
111
Chris Lattner28d02742009-10-09 05:01:15 +0000[diff] [blame]112<p>The <a href="http://clang.llvm.org/">Clang project</a> is an effort to build
113a set of new 'LLVM native' front-end technologies for the C family of languages.
114LLVM 2.6 is the first release to officially include Clang, and it provides a
115production quality C and Objective-C compiler. If you are interested in fast
116compiles and good diagnostics, we encourage you to try it out.</p>
117
118<p>In addition to supporting these languages, C++ support is also <a
119href="http://clang.llvm.org/cxx_status.html">well under way</a>, and mainline
120Clang is able to parse the libstdc++ 4.2 headers and even codegen simple apps.
121If you are interested in Clang C++ support or any other Clang feature, we
122strongly encourage you to get involved on the <a
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]123href="http://lists.cs.uiuc.edu/mailman/listinfo/cfe-dev">Clang front-end mailing
124list</a>.</p>
125
126<p>In the LLVM 2.6 time-frame, the Clang team has made many improvements:</p>
127
128<ul>
Chris Lattner28d02742009-10-09 05:01:15 +0000[diff] [blame]129<li>C and Objective-C support are now considered production quality.</li>
Chris Lattnerc32a5322009-09-30 06:27:22 +0000[diff] [blame]130<li>AuroraUX / FreeBSD &amp; OpenBSD Toolchain support.</li>
Chris Lattner28d02742009-10-09 05:01:15 +0000[diff] [blame]131<li>Most of Objective-C 2.0 is now supported with the GNU runtime.</li>
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]132<li>Many many bugs are fixed and many features have been added.</li>
133</ul>
134</div>
135
136<!--=========================================================================-->
137<div class="doc_subsection">
138<a name="clangsa">Clang Static Analyzer</a>
139</div>
140
141<div class="doc_text">
142
Chris Lattner80ed2552009-10-08 07:01:46 +0000[diff] [blame]143<p><b>UPDATE!</b> Previously announced in the 2.4 and 2.5 LLVM releases, the Clang project also
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]144includes an early stage static source code analysis tool for <a
145href="http://clang.llvm.org/StaticAnalysis.html">automatically finding bugs</a>
146in C and Objective-C programs. The tool performs a growing set of checks to find
147bugs that occur on a specific path within a program.</p>
148
149<p>In the LLVM 2.6 time-frame there have been many significant improvements to
150XYZ.</p>
151
152<p>The set of checks performed by the static analyzer continues to expand, and
153future plans for the tool include full source-level inter-procedural analysis
154and deeper checks such as buffer overrun detection. There are many opportunities
155to extend and enhance the static analyzer, and anyone interested in working on
156this project is encouraged to get involved!</p>
157
158</div>
159
160<!--=========================================================================-->
161<div class="doc_subsection">
162<a name="vmkit">VMKit: JVM/CLI Virtual Machine Implementation</a>
163</div>
164
165<div class="doc_text">
166<p>
167The <a href="http://vmkit.llvm.org/">VMKit project</a> is an implementation of
Nicolas Geoffraye83ae232009-10-09 10:13:08 +0000[diff] [blame]168a JVM and a CLI Virtual Machine (Microsoft .NET is an
Nicolas Geoffray7b2e71b2009-10-09 10:17:14 +0000[diff] [blame^]169implementation of the CLI) using LLVM for static and just-in-time
170compilation.</p>
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]171
Nicolas Geoffray7b2e71b2009-10-09 10:17:14 +0000[diff] [blame^]172<p>
173VMKit version 0.26 builds with LLVM 2.6 and you can find it on its
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]174<a href="http://vmkit.llvm.org/releases/">webpage</a>. The release includes
175bug fixes, cleanup and new features. The major changes are:</p>
176
177<ul>
178
Nicolas Geoffray7b2e71b2009-10-09 10:17:14 +0000[diff] [blame^]179<li>A new llcj tool to generate shared libraries or executables of Java
180 files.</li>
181<li>A rewriting of the JVM C++ codebase to call LLVM garbage collection (GC)
182 intrinsics when it manipulates GC objects. </li>
183<li>Cooperative GC and precise stack scanning using the GC framework of
184 LLVM. </li>
Nicolas Geoffraye83ae232009-10-09 10:13:08 +0000[diff] [blame]185<li>Fast subtype checking (paper from Click et al [JGI'02]). </li>
Nicolas Geoffray7b2e71b2009-10-09 10:17:14 +0000[diff] [blame^]186<li>Implementation of a two-word header for Java objects instead of the orginal
187 three-word header. </li>
188<li>Better Java specification-compliance: division by zero checks, stack
189 overflow checks, finalization and references support. </li>
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]190
191</ul>
192</div>
193
Chris Lattner6ee62f82009-10-09 05:55:04 +0000[diff] [blame]194
195<!--=========================================================================-->
196<div class="doc_subsection">
197<a name="compiler-rt">compiler-rt: Compiler Runtime Library</a>
198</div>
199
200<div class="doc_text">
201<p>
202The new LLVM <a href="http://compiler-rt.llvm.org/">compiler-rt project</a>
203is a simple library that provides an implementation of the low-level
204target-specific hooks required by code generation and other runtime components.
205For example, when compiling for a 32-bit target, converting a double to a 64-bit
206unsigned integer is compiling into a runtime call to the "__fixunsdfdi"
207function. The compiler-rt library provides optimized implementations of this and
208other low-level routines.</p>
209
210<p>
211All of the code in the compiler-rt project is available under the standard LLVM
212License, a "BSD-style" license.</p>
213
214</div>
215
216<!--=========================================================================-->
217<div class="doc_subsection">
218<a name="klee">klee: Symbolic Execution and Automatic Test Case Generator</a>
219</div>
220
221<div class="doc_text">
222<p>
223The new LLVM <a href="http://klee.llvm.org/">klee project</a> is a symbolic
224execution framework for programs in LLVM bitcode form. Klee tries to
225symbolically evaluate "all" paths through the application and records state
226transitions that lead to fault states. This allows it to construct testcases
227that lead to faults and can even be used to verify algorithms. For more
228details, please see the <a
229href="http://llvm.org/pubs/2008-12-OSDI-KLEE.html">OSDI 2008 paper</a> about
230Klee.</p>
231
232</div>
233
234<!--=========================================================================-->
235<div class="doc_subsection">
236<a name="dragonegg">Dragon Egg: An LLVM backend plugin for GCC</a>
237</div>
238
239<div class="doc_text">
240<p>
241<b>Duncan needs to write me</b>.
242</p>
243
244</div>
245
246
Chris Lattner80ed2552009-10-08 07:01:46 +0000[diff] [blame]247<!--=========================================================================-->
248<div class="doc_subsection">
249<a name="mc">llvm-mc: Machine Code Toolkit</a>
250</div>
251
252<div class="doc_text">
253<p>
Chris Lattner45178502009-10-09 06:24:25 +0000[diff] [blame]254The LLVM Machine Code (MC) Toolkit project is a (very early) effort to build
255better tools for dealing with machine code, object file formats, etc. The idea
256is to be able to generate most of the target specific details of assemblers and
257disassemblers from existing LLVM target .td files (with suitable enhancements),
258and to build infrastructure for reading and writing common object file formats.
259One of the first deliverables is to build a full assembler and integrate it into
260the compiler, which is predicted to substantially reduce compile time in some
261scenarios.
262</p>
Chris Lattner80ed2552009-10-08 07:01:46 +0000[diff] [blame]263
Chris Lattner45178502009-10-09 06:24:25 +0000[diff] [blame]264<p>In the LLVM 2.6 timeframe, the MC framework has grown to the point where it
265can reliably parse and pretty print (with some encoding information) a
266darwin/x86 .s file successfully, and has the very early phases of a Mach-O
267assembler in progress. Beyond the MC framework itself, major refactoring of the
268LLVM code generator has started. The idea is to make the code generator reason
269about the code it is producing in a much more semantic way, rather than a
270textual way. For example, the code generator now uses MCSection objects to
271represent section assignments, instead of text strings that print to .section
272directives.</p>
273
274<p>MC is an early and ongoing project that will hopefully continue to lead to
275many improvements in the code generator and build infrastructure useful for many
276other situations.
Chris Lattner80ed2552009-10-08 07:01:46 +0000[diff] [blame]277</p>
278
279</div>
280
281
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]282<!-- *********************************************************************** -->
283<div class="doc_section">
284 <a name="externalproj">External Projects Using LLVM 2.6</a>
285</div>
286<!-- *********************************************************************** -->
287
Chris Lattner80ed2552009-10-08 07:01:46 +0000[diff] [blame]288<!--=========================================================================-->
289<div class="doc_subsection">
290<a name="Rubinius">Rubinius</a>
291</div>
292
293<div class="doc_text">
294<p><a href="http://github.com/evanphx/rubinius">Rubinius</a> is an environment
295for running Ruby code which strives to write as much of the core class
296implementation in Ruby as possible. Combined with a bytecode interpreting VM, it
297uses LLVM to optimize and compile ruby code down to machine code. Techniques
298such as type feedback, method inlining, and uncommon traps are all used to
299remove dynamism from ruby execution and increase performance.</p>
Chris Lattnerd4a537b2009-10-08 16:01:33 +0000[diff] [blame]300
301<p>Since LLVM 2.5, Rubinius has made several major leaps forward, implementing
302a counter based JIT, type feedback, and speculative method inlining.
303</p>
304
Chris Lattner80ed2552009-10-08 07:01:46 +0000[diff] [blame]305</div>
Chris Lattnerc32a5322009-09-30 06:27:22 +0000[diff] [blame]306
307<!--=========================================================================-->
308<div class="doc_subsection">
309<a name="macruby">MacRuby</a>
310</div>
311
312<div class="doc_text">
313
314<p>
315<a href="http://macruby.org">MacRuby</a> is an implementation of Ruby on top of
316core Mac OS X technologies, such as the Objective-C common runtime and garbage
317collector, and the CoreFoundation framework. It is principally developed by
318Apple and aims at enabling the creation of full-fledged Mac OS X applications.
319</p>
320
321<p>
322MacRuby uses LLVM for optimization passes, JIT and AOT compilation of Ruby
323expressions. It also uses zero-cost DWARF exceptions to implement Ruby exception
324handling.</p>
325
326</div>
327
328
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]329<!--=========================================================================-->
330<div class="doc_subsection">
331<a name="pure">Pure</a>
332</div>
333
334<div class="doc_text">
335<p>
336<a href="http://pure-lang.googlecode.com/">Pure</a>
337is an algebraic/functional programming language based on term rewriting.
338Programs are collections of equations which are used to evaluate expressions in
339a symbolic fashion. Pure offers dynamic typing, eager and lazy evaluation,
340lexical closures, a hygienic macro system (also based on term rewriting),
341built-in list and matrix support (including list and matrix comprehensions) and
342an easy-to-use C interface. The interpreter uses LLVM as a backend to
343 JIT-compile Pure programs to fast native code.</p>
344
Chris Lattnerd4a537b2009-10-08 16:01:33 +0000[diff] [blame]345<p>Pure versions 0.31 and later have been tested and are known to work with
346LLVM 2.6 (and continue to work with older LLVM releases >= 2.3 as well).
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]347</p>
348</div>
349
350
351<!--=========================================================================-->
352<div class="doc_subsection">
353<a name="ldc">LLVM D Compiler</a>
354</div>
355
356<div class="doc_text">
357<p>
358<a href="http://www.dsource.org/projects/ldc">LDC</a> is an implementation of
359the D Programming Language using the LLVM optimizer and code generator.
360The LDC project works great with the LLVM 2.6 release. General improvements in
361this
362cycle have included new inline asm constraint handling, better debug info
363support, general bugfixes, and better x86-64 support. This has allowed
364some major improvements in LDC, getting us much closer to being as
365fully featured as the original DMD compiler from DigitalMars.
366</p>
367</div>
368
369<!--=========================================================================-->
370<div class="doc_subsection">
371<a name="RoadsendPHP">Roadsend PHP</a>
372</div>
373
374<div class="doc_text">
Chris Lattner80ed2552009-10-08 07:01:46 +0000[diff] [blame]375<p>
Chris Lattner80ed2552009-10-08 07:01:46 +0000[diff] [blame]376<a href="http://code.roadsend.com/rphp">Roadsend PHP</a> (rphp) is an open
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]377source implementation of the PHP programming
378language that uses LLVM for its optimizer, JIT, and static compiler. This is a
379reimplementation of an earlier project that is now based on LLVM.</p>
380</div>
381
Jeffrey Yasskinba2aa782009-06-24 21:09:13 +0000[diff] [blame]382<!--=========================================================================-->
383<div class="doc_subsection">
Chris Lattner6ee62f82009-10-09 05:55:04 +0000[diff] [blame]384<a name="UnladenSwallow">Unladen Swallow</a>
Jeffrey Yasskinba2aa782009-06-24 21:09:13 +0000[diff] [blame]385</div>
386
387<div class="doc_text">
Chris Lattner80ed2552009-10-08 07:01:46 +0000[diff] [blame]388<p>
Chris Lattner80ed2552009-10-08 07:01:46 +0000[diff] [blame]389<a href="http://code.google.com/p/unladen-swallow/">Unladen Swallow</a> is a
Jeffrey Yasskinba2aa782009-06-24 21:09:13 +0000[diff] [blame]390branch of <a href="http://python.org/">Python</a> intended to be fully
391compatible and significantly faster. It uses LLVM's optimization passes and JIT
392compiler.</p>
393</div>
394
Chris Lattner6ee62f82009-10-09 05:55:04 +0000[diff] [blame]395<!--=========================================================================-->
396<div class="doc_subsection">
397<a name="llvm-lua">llvm-lua</a>
398</div>
399
400<div class="doc_text">
401<p>
402<a href="http://code.google.com/p/llvm-lua/">LLVM-Lua</a> uses LLVM to add JIT
403&amp; static compiling support to the Lua VM. Lua bytecode is analyzed to
404remove type checks, then LLVM is used to compile those bytecodes down to machine
405code.</p>
406</div>
407
Jeffrey Yasskin5e98cb72009-06-24 21:26:42 +0000[diff] [blame]408
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]409
410<!-- *********************************************************************** -->
411<div class="doc_section">
412 <a name="whatsnew">What's New in LLVM 2.6?</a>
413</div>
414<!-- *********************************************************************** -->
415
416<div class="doc_text">
417
418<p>This release includes a huge number of bug fixes, performance tweaks, and
419minor improvements. Some of the major improvements and new features are listed
420in this section.
421</p>
Chris Lattner7b0a6812009-10-08 06:27:53 +0000[diff] [blame]422
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]423</div>
424
425<!--=========================================================================-->
426<div class="doc_subsection">
427<a name="majorfeatures">Major New Features</a>
428</div>
429
430<div class="doc_text">
431
432<p>LLVM 2.6 includes several major new capabilities:</p>
433
434<ul>
Chris Lattner6ee62f82009-10-09 05:55:04 +0000[diff] [blame]435<li>New <a href="#compiler-rt">compiler-rt</a>, <A href="#klee">klee</a>,
Chris Lattner45178502009-10-09 06:24:25 +0000[diff] [blame]436 and <a href="#mc">machine code toolkit</a> sub-projects.</li>
437<li>Debug information now includes line numbers when optimizations are enabled.
438 This allows statistical sampling tools like oprofile and Shark to map
439 samples back to source lines.</li>
440<li>LLVM now includes new experimental backends to support the MSP430, SystemZ,
Chris Lattner6ee62f82009-10-09 05:55:04 +0000[diff] [blame]441 and BlackFin architectures.</li>
442<li>LLVM supports a new <a href="GoldPlugin.html">Gold Linker Plugin</a> which
443 enables support for <a href="LinkTimeOptimization.html">transparent
444 link-time optimization</a> on ELF targets when used with the Gold binutils
445 linker.</li>
Chris Lattner45178502009-10-09 06:24:25 +0000[diff] [blame]446<li>LLVM now supports doing optimization and code generation on multiple threads
447 by allowing multiple "LLVMContext" objects to exist. Please see the <a
448 href="ProgrammersManual.html#threading">threading entry in the Programmer's
449 Manual</a> for more information.</li>
Chris Lattnera8f42142009-10-09 06:36:25 +0000[diff] [blame]450<li>LLVM now has experimental support for <a
451 href="http://nondot.org/~sabre/LLVMNotes/EmbeddedMetadata.txt">embedded
452 metadata</a> in LLVM IR, though the implementation is not guaranteed to be
453 final and the .bc file format may change in future releases. Debug info
454 does not yet use this format in LLVM 2.6.</p>
455
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]456</ul>
457
458</div>
459
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]460<!--=========================================================================-->
461<div class="doc_subsection">
462<a name="coreimprovements">LLVM IR and Core Improvements</a>
463</div>
464
465<div class="doc_text">
466<p>LLVM IR has several new features that are used by our existing front-ends and
467can be useful if you are writing a front-end for LLVM:</p>
468
469<ul>
Chris Lattner7b0a6812009-10-08 06:27:53 +0000[diff] [blame]470<li>Getelementpr instruction now allows any integer type for array/pointer indexes.</li>
471<li>Inbounds for GEP</li>
472<li>NSW/NUW/exact div</li>
473<li>LSR promotes int induction variables to 64-bit on 64-bit targets, major perf boost for numerical code.</li>
474<li>LSR now analyzes pointer expressions (e.g. getelementptrs), not just integers.</li>
475<li>new linkage types linkonce_odr, weak_odr, linker_private, and available_externally.</li>
476<li>New fadd, fsub, fmul instructions and classes. </li>
477<li>Target intrinsics can now return multiple results.</li>
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]478</ul>
479
480</div>
481
482<!--=========================================================================-->
483<div class="doc_subsection">
484<a name="optimizer">Optimizer Improvements</a>
485</div>
486
487<div class="doc_text">
488
489<p>In addition to a large array of bug fixes and minor performance tweaks, this
490release includes a few major enhancements and additions to the optimizers:</p>
491
492<ul>
493
Chris Lattner7b0a6812009-10-08 06:27:53 +0000[diff] [blame]494<li>SRoA improvements for vector unions, memset, arbitrary weird bitfield accesses etc. It now produces "strange" sized integers.</li>
495<li>Inliner reuse stack space when inlining arrays?</li>
496<li>Enabled GVN Load PRE.</li>
497<li>New Static Single Information (SSI) construction pass (not used by anything yet, experimental).</li>
498</li>
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]499
500</ul>
501
502</div>
503
504<!--=========================================================================-->
505<div class="doc_subsection">
506<a name="codegen">Target Independent Code Generator Improvements</a>
507</div>
508
509<div class="doc_text">
510
511<p>We have put a significant amount of work into the code generator
512infrastructure, which allows us to implement more aggressive algorithms and make
513it run faster:</p>
514
515<ul>
516
Chris Lattner7b0a6812009-10-08 06:27:53 +0000[diff] [blame]517<li> -asm-verbose now prints location info (with -g) and loop nest info.</li>
518<li>Tblgen now supports multiclass inheritance and a number of new string and
519 list operations like !(subst), !(foreach), !car, !cdr, !null, !if, !cast.
520 These make the .td files more expressive and allow more aggressive factoring
521 of duplication across instruction patterns.</li>
522<li>New MachineVerifier pass.</li>
523<li>Machine LICM, hoists things like constant pool loads, loads from readonly stubs, vector constant synthesization code, etc.</li>
524<li>Machine Sinking</li>
525<li>target-specific intrinsics (r63765)</li>
526<li>Regalloc improvements for commuting, various spiller peephole optimizations, cross-class coalescing.</li>
527<li><tt>llc -enable-value-prop</tt>, propagation of value info (sign/zero ext info) from one MBB to another</li>
528<li>Regalloc hints for allocation stuff: Evan r73381/r73671. Finished/enabled?</li>
529<li>Stack slot coloring for register spills (denser stack frames)</li>
530<li>SelectionDAGS: New BuildVectorSDNode (r65296), and ISD::VECTOR_SHUFFLE (r69952 / PR2957)</li>
Chris Lattner80ed2552009-10-08 07:01:46 +0000[diff] [blame]531<li>Experimental support for shrink wrapping support in PEI.</li>
Chris Lattner6ee62f82009-10-09 05:55:04 +0000[diff] [blame]532<li>Experimental support for writing ELF .o files directly from the compiler,
533 it works well for many simple C testcases, but doesn't support exception
534 handling, debug info, inline assembly, etc.</li>
535<li>Targets can now specify register allocation hints through
536 MachineRegisterInfo:: setRegAllocationHint. A regalloc hint consists 1) hint
537 type, 2) physical register number. A hint type of zero specifies a register
538 allocation preference. Other hint type values are target specific which are
539 resolved by TargetRegisterInfo::ResolveRegAllocHint. An example of which is
540 the ARM target can uses register hint to request that the register allocator
541 provide an even / odd register pair to two virtual registers. It is
542 important to note the register allocation hints are just hints. There is no
543 guarantee the register allocators will be able to satisfy the hints.</li>
544
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]545</ul>
546</div>
547
548<!--=========================================================================-->
549<div class="doc_subsection">
550<a name="x86">X86-32 and X86-64 Target Improvements</a>
551</div>
552
553<div class="doc_text">
554<p>New features of the X86 target include:
555</p>
556
557<ul>
558
Chris Lattner7b0a6812009-10-08 06:27:53 +0000[diff] [blame]559<li>Preliminary support for addrspace 256 -> GS, 257 -> FS, known problems: CodeGenerator.html#x86_memory</li>
560<li>Support for softfloat modes, typically used by OS kernels.</li>
561
562<li>X86-64: better modeling of implicit zero extensions, eliminates a lot of redundant zexts</li>
563<li>X86-64 TLS support for local exec and initial exec.</li>
564<li>Better modeling of H registerts as subregs.</li>
565<li>Vector icmp/fcmp now work with SSE codegen.</li>
566<li>SSE 4.2 support.</li>
567<li>all global variable reference logic is now in ClassifyGlobalReference.</li>
568</li>
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]569</ul>
570
571</div>
572
573<!--=========================================================================-->
574<div class="doc_subsection">
575<a name="pic16">PIC16 Target Improvements</a>
576</div>
577
578<div class="doc_text">
579<p>New features of the PIC16 target include:
580</p>
581
582<ul>
Chris Lattner7b0a6812009-10-08 06:27:53 +0000[diff] [blame]583<li>Support for floating-point, indirect function calls, and
584 passing/returning aggregate types to functions.
585<li>The code generator is able to generate debug info into output COFF files.
586<li>Support for placing an object into a specific section or at a specific
587 address in memory.</li>
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]588</ul>
589
590<p>Things not yet supported:</p>
591
592<ul>
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]593<li>Variable arguments.</li>
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]594<li>Interrupts/programs.</li>
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]595</ul>
596
597</div>
598
Bob Wilson755cbe02009-08-12 21:19:49 +0000[diff] [blame]599<!--=========================================================================-->
600<div class="doc_subsection">
601<a name="ARM">ARM Target Improvements</a>
602</div>
603
604<div class="doc_text">
605<p>New features of the ARM target include:
606</p>
607
608<ul>
609
610<li>Preliminary support for processors, such as the Cortex-A8 and Cortex-A9,
Sandeep Patel64e39612009-08-20 15:01:16 +0000[diff] [blame]611that implement version v7-A of the ARM architecture. The ARM backend now
612supports both the Thumb2 and Advanced SIMD (Neon) instruction sets. The
613AAPCS-VFP "hard float" calling conventions are also supported with the
614<tt>-float-abi=hard</tt> flag. These features are still somewhat experimental
615and subject to change. The Neon intrinsics, in particular, may change in future
616releases of LLVM.
Bob Wilson755cbe02009-08-12 21:19:49 +0000[diff] [blame]617</li>
Chris Lattner7b0a6812009-10-08 06:27:53 +0000[diff] [blame]618
619 ARM AAPCS-VFP hard float ABI is supported.
620 ARM calling convention code is now tblgen generated instead of manual.
621 ARM: NEON support. neonfp for doing single precision fp with neon instead of VFP.
622
Bob Wilson755cbe02009-08-12 21:19:49 +0000[diff] [blame]623</ul>
624
625</div>
626
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]627<!--=========================================================================-->
628<div class="doc_subsection">
Chris Lattner7b0a6812009-10-08 06:27:53 +0000[diff] [blame]629<a name="OtherTarget">Other Target Specific Improvements</a>
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]630</div>
631
632<div class="doc_text">
Chris Lattner7b0a6812009-10-08 06:27:53 +0000[diff] [blame]633<p>New features of other targets include:
634</p>
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]635
636<ul>
Chris Lattner7b0a6812009-10-08 06:27:53 +0000[diff] [blame]637<li>Mips now supports O32 Calling Convention.</li>
Chris Lattnera8f42142009-10-09 06:36:25 +0000[diff] [blame]638<li>Many improvements to the 32-bit PowerPC SVR4 ABI (used on powerpc-linux)
639 support, lots of bugs fixed.</li>
640<li>Added support for the 64-bit PowerPC SVR4 ABI (used on powerpc64-linux).
641 Needs more testing.</li>
Chris Lattner7b0a6812009-10-08 06:27:53 +0000[diff] [blame]642</ul>
643
644</div>
645
646<!--=========================================================================-->
647<div class="doc_subsection">
Chris Lattner45178502009-10-09 06:24:25 +0000[diff] [blame]648<a name="executionengine">Interpreter and JIT Improvements</a>
649</div>
650
651<div class="doc_text">
652
653<ul>
654<li>The JIT now supports generating more than 16M of code.</li>
655<li>When configured with --with-oprofile, the JIT can now inform oprofile about
656 JIT'd code, allowing oprofile to get line number and function name
657 information for JIT'd functions.</li>
658<li>When "libffi" is available, the LLVM interpreter now uses it, which supports
659 calling almost arbitrary external (natively compiled) functions.</li>
660<li>Clients of the JIT can now register a 'JITEventListener' object to receive
661 callbacks when the JIT emits or frees machine code. The OProfile support
662 uses this mechanism.</li>
663</ul>
664
665</div>
666
667
668<!--=========================================================================-->
669<div class="doc_subsection">
Chris Lattner7b0a6812009-10-08 06:27:53 +0000[diff] [blame]670<a name="newapis">New Useful APIs</a>
671</div>
672
673<div class="doc_text">
674
675<ul>
676<li>New EngineBuilder class for creating JITs: r76276</li>
677 New PrettyStackTrace, crashes of llvm tools should give some indication of what the compiler was doing at the time of the crash (e.g. running a pass), and print out command line arguments.
678 StringRef class, Twine class.
679 New WeakVH and AssertingVH and CallbackVH classes.
680 New llvm/ADT/Triple class.
681 llvm_report_error() error handling API (llvm/Support/ErrorHandling.h)
682 New llvm/System/Atomic.h, llvm/System/RWMutex.h for portable atomic ops, rw locks.
683 New SourceMgr, SMLoc classes for simple parsers with caret diagnostics and #include support, (used by
684 tablegen, llvm-mc, the .ll parser, FileCheck, etc)
685
Chris Lattner45178502009-10-09 06:24:25 +0000[diff] [blame]686</ul>
Chris Lattner7b0a6812009-10-08 06:27:53 +0000[diff] [blame]687
688
689</div>
690
691<!--=========================================================================-->
692<div class="doc_subsection">
693<a name="otherimprovements">Other Improvements and New Features</a>
694</div>
695
696<div class="doc_text">
697<p>Other miscellaneous features include:</p>
698
699<ul>
Chris Lattner45178502009-10-09 06:24:25 +0000[diff] [blame]700<li>LLVM now includes a new internal '<a
701 href="http://llvm.org/cmds/FileCheck.html">FileCheck</a>' tool which allows
702 writing much more accurate regression tests that run faster. Please see the
703 <a href="TestingGuide.html#FileCheck">FileCheck section of the Testing
704 Guide</a> for more information.</li>
Chris Lattner6ee62f82009-10-09 05:55:04 +0000[diff] [blame]705<li>LLVM profile information support has been significantly improved to produce
706correct use counts, and has support for edge profiling with reduced runtime
707overhead. Combined, the generated profile information is both more correct and
708imposes about half as much overhead (2.6. from 12% to 6% overhead on SPEC
709CPU2000).</li>
Chris Lattner80ed2552009-10-08 07:01:46 +0000[diff] [blame]710<li>Many extensions to the C APIs.</li>
Chris Lattner45178502009-10-09 06:24:25 +0000[diff] [blame]711<li>LLVM 2.6 includes a brand new experimental LLVM bindings to the Ada2005
712programming language.</li>
Chris Lattner80ed2552009-10-08 07:01:46 +0000[diff] [blame]713
714<li>LLVMC:
715
716* Dynamic plugins now work on Windows.
717* New option property: init. Makes possible to provide default values for
718 options defined in plugins (interface to cl::init).
719* New example: Skeleton, shows how to create a standalone LLVMC-based driver.
720* New example: mcc16, a driver for the PIC16 toolchain.</li>
721
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]722</ul>
723
724</div>
725
726
727<!--=========================================================================-->
728<div class="doc_subsection">
729<a name="changes">Major Changes and Removed Features</a>
730</div>
731
732<div class="doc_text">
733
734<p>If you're already an LLVM user or developer with out-of-tree changes based
Owen Andersone27be3a2009-07-02 16:48:38 +0000[diff] [blame]735on LLVM 2.5, this section lists some "gotchas" that you may run into upgrading
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]736from the previous release.</p>
737
738<ul>
739
Chris Lattner7b0a6812009-10-08 06:27:53 +0000[diff] [blame]740<li>The Itanium (IA64) backend has been removed. It was not supported and
741 bitrotted.</li>
742<li>The BigBlock register allocator has been removed, it also bitrotted.</li>
743<li>The C Backend (-march=c) is no longer considered part of the LLVM release
744criteria. We still want it to work, but no one is maintaining it and it lacks
745support for arbitrary precision integers and other important IR features.</li>
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]746</ul>
747
Chris Lattner80ed2552009-10-08 07:01:46 +0000[diff] [blame]748 LLVM build now builds all libraries as .a files instead of some
749 libraries as relinked .o files. This requires some APIs like
750 InitializeAllTargets.h. TargetRegistry!
751
752
753
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]754
755<p>In addition, many APIs have changed in this release. Some of the major LLVM
756API changes are:</p>
757
Chris Lattner7b0a6812009-10-08 06:27:53 +0000[diff] [blame]758
759 API Cleanup:
760 no use of hash_set/hash_map, no more llvm::OStream
761 Use raw_ostream for everything, killed off llvm/Streams.h and DOUT
762
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]763<ul>
Owen Andersone27be3a2009-07-02 16:48:38 +0000[diff] [blame]764<li>LLVM's global uniquing tables for <tt>Type</tt>s and <tt>Constant</tt>s have
765 been privatized into members of an <tt>LLVMContext</tt>. A number of APIs
766 now take an <tt>LLVMContext</tt> as a parameter. To smooth the transition
767 for clients that will only ever use a single context, the new
768 <tt>getGlobalContext()</tt> API can be used to access a default global
769 context which can be passed in any and all cases where a context is
770 required.
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]771<li>The <tt>getABITypeSize</tt> methods are now called <tt>getAllocSize</tt>.</li>
Dan Gohmanf667e7e2009-07-07 20:05:15 +0000[diff] [blame]772<li>The <tt>Add</tt>, <tt>Sub</tt>, and <tt>Mul</tt> operators are no longer
773 overloaded for floating-point types. Floating-point addition, subtraction,
774 and multiplication are now represented with new operators <tt>FAdd</tt>,
775 <tt>FSub</tt>, and <tt>FMul</tt>. In the <tt>IRBuilder</tt> API,
776 <tt>CreateAdd</tt>, <tt>CreateSub</tt>, <tt>CreateMul</tt>, and
777 <tt>CreateNeg</tt> should only be used for integer arithmetic now;
778 <tt>CreateFAdd</tt>, <tt>CreateFSub</tt>, <tt>CreateFMul</tt>, and
779 <tt>CreateFNeg</tt> should now be used for floating-point arithmetic.</li>
Daniel Dunbar02e73132009-07-12 20:41:27 +0000[diff] [blame]780<li>The DynamicLibrary class can no longer be constructed, its functionality has
781 moved to static member functions.</li>
Dan Gohmane4d54d72009-07-15 19:59:19 +0000[diff] [blame]782<li><tt>raw_fd_ostream</tt>'s constructor for opening a given filename now
783 takes an extra <tt>Force</tt> argument. If <tt>Force</tt> is set to
784 <tt>false</tt>, an error will be reported if a file with the given name
785 already exists. If <tt>Force</tt> is set to <tt>true</tt>, the file will
786 be silently truncated (which is the behavior before this flag was
787 added).</li>
Torok Edwinf6fa8ae2009-07-21 20:27:10 +0000[diff] [blame]788<li><tt>SCEVHandle</tt> no longer exists, because reference counting is no
789longer done for <tt>SCEV*</tt> objects, instead <tt>const SCEV*</tt> should be
790used.</li>
Daniel Dunbar8b5ee822009-07-25 05:26:53 +0000[diff] [blame]791
Daniel Dunbar6e0d1cb2009-07-25 04:41:11 +0000[diff] [blame]792<li>Many APIs, notably <tt>llvm::Value</tt>, now use the <tt>StringRef</tt>
793and <tt>Twine</tt> classes instead of passing <tt>const char*</tt>
794or <tt>std::string</tt>, as described in
795the <a href="ProgrammersManual.html#string_apis">Programmer's Manual</a>. Most
Daniel Dunbar8b5ee822009-07-25 05:26:53 +0000[diff] [blame]796clients should be unaffected by this transition, unless they are used to <tt>Value::getName()</tt> returning a string. Here are some tips on updating to 2.6:
797 <ul>
798 <li><tt>getNameStr()</tt> is still available, and matches the old
799 behavior. Replacing <tt>getName()</tt> calls with this is an safe option,
800 although more efficient alternatives are now possible.</li>
801
802 <li>If you were just relying on <tt>getName()</tt> being able to be sent to
803 a <tt>std::ostream</tt>, consider migrating
804 to <tt>llvm::raw_ostream</tt>.</li>
805
806 <li>If you were using <tt>getName().c_str()</tt> to get a <tt>const
807 char*</tt> pointer to the name, you can use <tt>getName().data()</tt>.
808 Note that this string (as before), may not be the entire name if the
809 name containts embedded null characters.</li>
810
811 <li>If you were using operator plus on the result of <tt>getName()</tt> and
812 treating the result as an <tt>std::string</tt>, you can either
813 uses <tt>Twine::str</tt> to get the result as an <tt>std::string</tt>, or
814 could move to a <tt>Twine</tt> based design.</li>
Daniel Dunbar03d76512009-07-25 23:55:21 +0000[diff] [blame]815
816 <li><tt>isName()</tt> should be replaced with comparison
Benjamin Kramere15192b2009-08-05 15:42:44 +0000[diff] [blame]817 against <tt>getName()</tt> (this is now efficient).
Daniel Dunbar8b5ee822009-07-25 05:26:53 +0000[diff] [blame]818 </ul>
819</li>
820
Daniel Dunbara5881e32009-07-26 02:12:58 +0000[diff] [blame]821<li>The registration interfaces for backend Targets has changed (what was
Daniel Dunbard6b06b12009-07-26 05:41:39 +0000[diff] [blame]822previously TargetMachineRegistry). For backend authors, see the <a href="WritingAnLLVMBackend.html#TargetRegistration">Writing An LLVM Backend</a> guide. For clients, the notable API changes are:
823 <ul>
824 <li><tt>TargetMachineRegistry</tt> has been renamed
825 to <tt>TargetRegistry</tt>.</li>
826
827 <li>Clients should move to using the <tt>TargetRegistry::lookupTarget()</tt>
828 function to find targets.</li>
829 </ul>
830</li>
Daniel Dunbar8b5ee822009-07-25 05:26:53 +0000[diff] [blame]831
Torok Edwinf6fa8ae2009-07-21 20:27:10 +0000[diff] [blame]832<li>llvm-dis now fails if output file exists, instead of dumping to stdout.
Daniel Dunbar6e0d1cb2009-07-25 04:41:11 +0000[diff] [blame]833FIXME: describe any other tool changes due to the raw_fd_ostream change. FIXME:
834This is not an API change, maybe there should be a tool changes section?</li>
Torok Edwinf6fa8ae2009-07-21 20:27:10 +0000[diff] [blame]835<li>temporarely due to Context API change passes should call doInitialization()
836method of the pass they inherit from, otherwise Context is NULL.
837FIXME: remove this entry when this is no longer needed.<li>
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]838</ul>
839
840</div>
841
842
843
844<!-- *********************************************************************** -->
845<div class="doc_section">
846 <a name="portability">Portability and Supported Platforms</a>
847</div>
848<!-- *********************************************************************** -->
849
850<div class="doc_text">
851
852<p>LLVM is known to work on the following platforms:</p>
853
854<ul>
855<li>Intel and AMD machines (IA32, X86-64, AMD64, EMT-64) running Red Hat
Chris Lattner3e7b5ca2009-07-21 22:47:03 +0000[diff] [blame]856Linux, Fedora Core, FreeBSD and AuroraUX (and probably other unix-like systems).</li>
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]857<li>PowerPC and X86-based Mac OS X systems, running 10.3 and above in 32-bit
858and 64-bit modes.</li>
859<li>Intel and AMD machines running on Win32 using MinGW libraries (native).</li>
860<li>Intel and AMD machines running on Win32 with the Cygwin libraries (limited
861 support is available for native builds with Visual C++).</li>
862<li>Sun UltraSPARC workstations running Solaris 10.</li>
863<li>Alpha-based machines running Debian GNU/Linux.</li>
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]864</ul>
865
866<p>The core LLVM infrastructure uses GNU autoconf to adapt itself
867to the machine and operating system on which it is built. However, minor
868porting may be required to get LLVM to work on new platforms. We welcome your
869portability patches and reports of successful builds or error messages.</p>
870
871</div>
872
873<!-- *********************************************************************** -->
874<div class="doc_section">
875 <a name="knownproblems">Known Problems</a>
876</div>
877<!-- *********************************************************************** -->
878
879<div class="doc_text">
880
881<p>This section contains significant known problems with the LLVM system,
882listed by component. If you run into a problem, please check the <a
883href="http://llvm.org/bugs/">LLVM bug database</a> and submit a bug if
884there isn't already one.</p>
885
Chris Lattner3e7b5ca2009-07-21 22:47:03 +0000[diff] [blame]886<ul>
Chris Lattneraedb59a2009-07-21 23:17:26 +0000[diff] [blame]887<li>LLVM will not correctly compile on Solaris and/or OpenSolaris
Chris Lattner3e7b5ca2009-07-21 22:47:03 +0000[diff] [blame]888using the stock GCC 3.x.x series 'out the box',
889See: <a href="#brokengcc">Broken versions of GCC and other tools</a>.
890However, A <a href="http://pkg.auroraux.org/GCC">Modern GCC Build</a>
891for x86/x64 has been made available from the third party AuroraUX Project
Chris Lattner7b0a6812009-10-08 06:27:53 +0000[diff] [blame]892that has been meticulously tested for bootstrapping LLVM &amp; Clang.</li>
Chris Lattner3e7b5ca2009-07-21 22:47:03 +0000[diff] [blame]893</ul>
894
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]895</div>
896
897<!-- ======================================================================= -->
898<div class="doc_subsection">
899 <a name="experimental">Experimental features included with this release</a>
900</div>
901
902<div class="doc_text">
903
904<p>The following components of this LLVM release are either untested, known to
905be broken or unreliable, or are in early development. These components should
906not be relied on, and bugs should not be filed against them, but they may be
907useful to some people. In particular, if you would like to work on one of these
908components, please contact us on the <a
909href="http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev">LLVMdev list</a>.</p>
910
911<ul>
Dan Gohmand2cb3d22009-07-24 00:30:09 +0000[diff] [blame]912<li>The MSIL, Alpha, SPU, MIPS, and PIC16 backends are experimental.</li>
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]913<li>The <tt>llc</tt> "<tt>-filetype=asm</tt>" (the default) is the only
914 supported value for this option.</li>
915</ul>
916
917</div>
918
919<!-- ======================================================================= -->
920<div class="doc_subsection">
921 <a name="x86-be">Known problems with the X86 back-end</a>
922</div>
923
924<div class="doc_text">
925
926<ul>
927 <li>The X86 backend does not yet support
928 all <a href="http://llvm.org/PR879">inline assembly that uses the X86
929 floating point stack</a>. It supports the 'f' and 't' constraints, but not
930 'u'.</li>
931 <li>The X86 backend generates inefficient floating point code when configured
932 to generate code for systems that don't have SSE2.</li>
933 <li>Win64 code generation wasn't widely tested. Everything should work, but we
934 expect small issues to happen. Also, llvm-gcc cannot build the mingw64
935 runtime currently due
936 to <a href="http://llvm.org/PR2255">several</a>
937 <a href="http://llvm.org/PR2257">bugs</a> and due to lack of support for
938 the
939 'u' inline assembly constraint and for X87 floating point inline assembly.</li>
940 <li>The X86-64 backend does not yet support the LLVM IR instruction
941 <tt>va_arg</tt>. Currently, the llvm-gcc and front-ends support variadic
942 argument constructs on X86-64 by lowering them manually.</li>
943</ul>
944
945</div>
946
947<!-- ======================================================================= -->
948<div class="doc_subsection">
949 <a name="ppc-be">Known problems with the PowerPC back-end</a>
950</div>
951
952<div class="doc_text">
953
954<ul>
955<li>The Linux PPC32/ABI support needs testing for the interpreter and static
956compilation, and lacks support for debug information.</li>
957</ul>
958
959</div>
960
961<!-- ======================================================================= -->
962<div class="doc_subsection">
963 <a name="arm-be">Known problems with the ARM back-end</a>
964</div>
965
966<div class="doc_text">
967
968<ul>
Bob Wilson755cbe02009-08-12 21:19:49 +0000[diff] [blame]969<li>Support for the Advanced SIMD (Neon) instruction set is still incomplete
970and not well tested. Some features may not work at all, and the code quality
971may be poor in some cases.</li>
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]972<li>Thumb mode works only on ARMv6 or higher processors. On sub-ARMv6
973processors, thumb programs can crash or produce wrong
974results (<a href="http://llvm.org/PR1388">PR1388</a>).</li>
975<li>Compilation for ARM Linux OABI (old ABI) is supported but not fully tested.
976</li>
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]977</ul>
978
979</div>
980
981<!-- ======================================================================= -->
982<div class="doc_subsection">
983 <a name="sparc-be">Known problems with the SPARC back-end</a>
984</div>
985
986<div class="doc_text">
987
988<ul>
989<li>The SPARC backend only supports the 32-bit SPARC ABI (-m32); it does not
990 support the 64-bit SPARC ABI (-m64).</li>
991</ul>
992
993</div>
994
995<!-- ======================================================================= -->
996<div class="doc_subsection">
997 <a name="mips-be">Known problems with the MIPS back-end</a>
998</div>
999
1000<div class="doc_text">
1001
1002<ul>
1003<li>The O32 ABI is not fully supported.</li>
1004<li>64-bit MIPS targets are not supported yet.</li>
1005</ul>
1006
1007</div>
1008
1009<!-- ======================================================================= -->
1010<div class="doc_subsection">
1011 <a name="alpha-be">Known problems with the Alpha back-end</a>
1012</div>
1013
1014<div class="doc_text">
1015
1016<ul>
1017
1018<li>On 21164s, some rare FP arithmetic sequences which may trap do not have the
1019appropriate nops inserted to ensure restartability.</li>
1020
1021</ul>
1022</div>
1023
1024<!-- ======================================================================= -->
1025<div class="doc_subsection">
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]1026 <a name="c-be">Known problems with the C back-end</a>
1027</div>
1028
1029<div class="doc_text">
1030
1031<ul>
1032<li><a href="http://llvm.org/PR802">The C backend has only basic support for
1033 inline assembly code</a>.</li>
1034<li><a href="http://llvm.org/PR1658">The C backend violates the ABI of common
1035 C++ programs</a>, preventing intermixing between C++ compiled by the CBE and
1036 C++ code compiled with <tt>llc</tt> or native compilers.</li>
1037<li>The C backend does not support all exception handling constructs.</li>
1038<li>The C backend does not support arbitrary precision integers.</li>
1039</ul>
1040
1041</div>
1042
1043
1044<!-- ======================================================================= -->
1045<div class="doc_subsection">
1046 <a name="c-fe">Known problems with the llvm-gcc C front-end</a>
1047</div>
1048
1049<div class="doc_text">
1050
1051<p>llvm-gcc does not currently support <a href="http://llvm.org/PR869">Link-Time
1052Optimization</a> on most platforms "out-of-the-box". Please inquire on the
1053LLVMdev mailing list if you are interested.</p>
1054
1055<p>The only major language feature of GCC not supported by llvm-gcc is
1056 the <tt>__builtin_apply</tt> family of builtins. However, some extensions
1057 are only supported on some targets. For example, trampolines are only
1058 supported on some targets (these are used when you take the address of a
1059 nested function).</p>
1060
1061<p>If you run into GCC extensions which are not supported, please let us know.
1062</p>
1063
1064</div>
1065
1066<!-- ======================================================================= -->
1067<div class="doc_subsection">
1068 <a name="c++-fe">Known problems with the llvm-gcc C++ front-end</a>
1069</div>
1070
1071<div class="doc_text">
1072
1073<p>The C++ front-end is considered to be fully
1074tested and works for a number of non-trivial programs, including LLVM
1075itself, Qt, Mozilla, etc.</p>
1076
1077<ul>
1078<li>Exception handling works well on the X86 and PowerPC targets. Currently
1079 only Linux and Darwin targets are supported (both 32 and 64 bit).</li>
1080</ul>
1081
1082</div>
1083
1084<!-- ======================================================================= -->
1085<div class="doc_subsection">
1086 <a name="fortran-fe">Known problems with the llvm-gcc Fortran front-end</a>
1087</div>
1088
1089<div class="doc_text">
1090<ul>
1091<li>Fortran support generally works, but there are still several unresolved bugs
1092 in Bugzilla. Please see the tools/gfortran component for details.</li>
1093</ul>
1094</div>
1095
1096<!-- ======================================================================= -->
1097<div class="doc_subsection">
1098 <a name="ada-fe">Known problems with the llvm-gcc Ada front-end</a>
1099</div>
1100
1101<div class="doc_text">
1102The llvm-gcc 4.2 Ada compiler works fairly well; however, this is not a mature
1103technology, and problems should be expected.
1104<ul>
1105<li>The Ada front-end currently only builds on X86-32. This is mainly due
1106to lack of trampoline support (pointers to nested functions) on other platforms.
1107However, it <a href="http://llvm.org/PR2006">also fails to build on X86-64</a>
1108which does support trampolines.</li>
1109<li>The Ada front-end <a href="http://llvm.org/PR2007">fails to bootstrap</a>.
1110This is due to lack of LLVM support for <tt>setjmp</tt>/<tt>longjmp</tt> style
1111exception handling, which is used internally by the compiler.
1112Workaround: configure with --disable-bootstrap.</li>
1113<li>The c380004, <a href="http://llvm.org/PR2010">c393010</a>
1114and <a href="http://llvm.org/PR2421">cxg2021</a> ACATS tests fail
1115(c380004 also fails with gcc-4.2 mainline).
1116If the compiler is built with checks disabled then <a href="http://llvm.org/PR2010">c393010</a>
1117causes the compiler to go into an infinite loop, using up all system memory.</li>
1118<li>Some GCC specific Ada tests continue to crash the compiler.</li>
1119<li>The -E binder option (exception backtraces)
1120<a href="http://llvm.org/PR1982">does not work</a> and will result in programs
1121crashing if an exception is raised. Workaround: do not use -E.</li>
1122<li>Only discrete types <a href="http://llvm.org/PR1981">are allowed to start
1123or finish at a non-byte offset</a> in a record. Workaround: do not pack records
1124or use representation clauses that result in a field of a non-discrete type
1125starting or finishing in the middle of a byte.</li>
1126<li>The <tt>lli</tt> interpreter <a href="http://llvm.org/PR2009">considers
1127'main' as generated by the Ada binder to be invalid</a>.
1128Workaround: hand edit the file to use pointers for <tt>argv</tt> and
1129<tt>envp</tt> rather than integers.</li>
1130<li>The <tt>-fstack-check</tt> option <a href="http://llvm.org/PR2008">is
1131ignored</a>.</li>
1132</ul>
1133</div>
1134
Erick Tryzelaar17167be2009-09-28 04:42:55 +0000[diff] [blame]1135<!-- ======================================================================= -->
1136<div class="doc_subsection">
1137 <a name="ocaml-bindingse">Known problems with the O'Caml bindings</a>
1138</div>
1139
1140<div class="doc_text">
1141
1142<p>The Llvm.Linkage module is broken, and has incorrect values. Only
1143Llvm.Linkage.External, Llvm.Linkage.Available_externally, and
1144Llvm.Linkage.Link_once will be correct. If you need any of the other linkage
1145modes, you'll have to write an external C library in order to expose the
1146functionality. This has been fixed in the trunk.</p>
1147</div>
1148
Duncan Sandse2e4a892009-06-24 08:38:48 +0000[diff] [blame]1149<!-- *********************************************************************** -->
1150<div class="doc_section">
1151 <a name="additionalinfo">Additional Information</a>
1152</div>
1153<!-- *********************************************************************** -->
1154
1155<div class="doc_text">
1156
1157<p>A wide variety of additional information is available on the <a
1158href="http://llvm.org">LLVM web page</a>, in particular in the <a
1159href="http://llvm.org/docs/">documentation</a> section. The web page also
1160contains versions of the API documentation which is up-to-date with the
1161Subversion version of the source code.
1162You can access versions of these documents specific to this release by going
1163into the "<tt>llvm/doc/</tt>" directory in the LLVM tree.</p>
1164
1165<p>If you have any questions or comments about LLVM, please feel free to contact
1166us via the <a href="http://llvm.org/docs/#maillist"> mailing
1167lists</a>.</p>
1168
1169</div>
1170
1171<!-- *********************************************************************** -->
1172
1173<hr>
1174<address>
1175 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
1176 src="http://jigsaw.w3.org/css-validator/images/vcss-blue" alt="Valid CSS"></a>
1177 <a href="http://validator.w3.org/check/referer"><img
1178 src="http://www.w3.org/Icons/valid-html401-blue" alt="Valid HTML 4.01"></a>
1179
1180 <a href="http://llvm.org/">LLVM Compiler Infrastructure</a><br>
1181 Last modified: $Date$
1182</address>
1183
1184</body>
1185</html>