Gitiles
Code Review Sign In
gerrit-public.fairphone.software / fp2-dev / platform / external / llvm / 0640ac251534f7e3cf16a79a5f999e61552959f6 / . / docs / ReleaseNotes.html
blob: 2c7e232e7b7273186c4a81c0e54a912228bb6ecb [file] [log] [blame]
Dan Gohmanf17a25c2007-07-18 16:29:46 +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">
Chris Lattner2c3ef982008-10-11 18:18:28 +0000[diff] [blame]7 <title>LLVM 2.4 Release Notes</title>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]8</head>
9<body>
10
Chris Lattner2c3ef982008-10-11 18:18:28 +0000[diff] [blame]11<div class="doc_title">LLVM 2.4 Release Notes</div>
Mikhail Glushenkov808cd3e2008-10-13 02:08:34 +0000[diff] [blame]12
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]13<ol>
14 <li><a href="#intro">Introduction</a></li>
Chris Lattner78a130b2008-10-13 18:01:01 +0000[diff] [blame]15 <li><a href="#subproj">Sub-project Status Update</a></li>
16 <li><a href="#whatsnew">What's New in LLVM?</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]17 <li><a href="GettingStarted.html">Installation Instructions</a></li>
18 <li><a href="#portability">Portability and Supported Platforms</a></li>
Dan Gohman0de549c2008-10-14 16:23:02 +0000[diff] [blame]19 <li><a href="#knownproblems">Known Problems</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]20 <li><a href="#additionalinfo">Additional Information</a></li>
21</ol>
22
23<div class="doc_author">
Dan Gohman0de549c2008-10-14 16:23:02 +0000[diff] [blame]24 <p>Written by the <a href="http://llvm.org">LLVM Team</a></p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]25</div>
26
27<!-- *********************************************************************** -->
28<div class="doc_section">
29 <a name="intro">Introduction</a>
30</div>
31<!-- *********************************************************************** -->
32
33<div class="doc_text">
34
Chris Lattner92eacf82008-10-13 17:57:36 +0000[diff] [blame]35<p>This document contains the release notes for the LLVM Compiler
36Infrastructure, release 2.4. Here we describe the status of LLVM, including
37major improvements from the previous release and significant known problems.
38All LLVM releases may be downloaded from the <a
39href="http://llvm.org/releases/">LLVM releases web site</a>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]40
41<p>For more information about LLVM, including information about the latest
42release, please check out the <a href="http://llvm.org/">main LLVM
43web site</a>. If you have questions or comments, the <a
Chris Lattner92eacf82008-10-13 17:57:36 +0000[diff] [blame]44href="http://mail.cs.uiuc.edu/mailman/listinfo/llvmdev">LLVM Developer's Mailing
45List</a> is a good place to send them.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]46
Mikhail Glushenkov808cd3e2008-10-13 02:08:34 +0000[diff] [blame]47<p>Note that if you are reading this file from a Subversion checkout or the
Chris Lattnerb0d6c502007-09-21 03:54:09 +0000[diff] [blame]48main LLVM web page, this document applies to the <i>next</i> release, not the
Gabor Greif7c65c4f2008-10-14 11:00:32 +0000[diff] [blame]49current one. To see the release notes for a specific release, please see the
Chris Lattnerb0d6c502007-09-21 03:54:09 +0000[diff] [blame]50<a href="http://llvm.org/releases/">releases page</a>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]51
52</div>
53
Chris Lattner2c3ef982008-10-11 18:18:28 +0000[diff] [blame]54<!-- Unfinished features in 2.4:
Chris Lattner8fda4952008-02-10 07:46:44 +0000[diff] [blame]55 Machine LICM
56 Machine Sinking
57 LegalizeDAGTypes
Chris Lattner4f076402008-10-12 19:47:48 +0000[diff] [blame]58 llc -enable-value-prop, propagation of value info (sign/zero ext info) from
59 one MBB to another
Chris Lattner8fda4952008-02-10 07:46:44 +0000[diff] [blame]60 -->
Mikhail Glushenkov808cd3e2008-10-13 02:08:34 +0000[diff] [blame]61
Chris Lattner4f076402008-10-12 19:47:48 +0000[diff] [blame]62 <!-- for announcement email:
63 mention dev mtg
Chris Lattner548493b2008-10-12 22:57:58 +0000[diff] [blame]64 Xcode 3.1 and 3.1.1.
Chris Lattner4f076402008-10-12 19:47:48 +0000[diff] [blame]65 -->
Chris Lattner8fda4952008-02-10 07:46:44 +0000[diff] [blame]66
Chris Lattner78a130b2008-10-13 18:01:01 +0000[diff] [blame]67<!-- *********************************************************************** -->
68<div class="doc_section">
69 <a name="subproj">Sub-project Status Update</a>
Chris Lattnerf65224f2008-06-08 21:34:41 +0000[diff] [blame]70</div>
Chris Lattner78a130b2008-10-13 18:01:01 +0000[diff] [blame]71<!-- *********************************************************************** -->
Chris Lattnerf65224f2008-06-08 21:34:41 +0000[diff] [blame]72
73<div class="doc_text">
Chris Lattnerf6a0d382008-06-08 21:58:17 +0000[diff] [blame]74<p>
Chris Lattner78a130b2008-10-13 18:01:01 +0000[diff] [blame]75The LLVM 2.4 distribution currently consists of code from the core LLVM
Chris Lattnera3c683b2008-10-13 18:11:54 +0000[diff] [blame]76repository (which roughly includes the LLVM optimizers, code generators and
Chris Lattnerf6a0d382008-06-08 21:58:17 +0000[diff] [blame]77supporting tools) and the llvm-gcc repository. In addition to this code, the
78LLVM Project includes other sub-projects that are in development. The two which
Chris Lattner2c3ef982008-10-11 18:18:28 +0000[diff] [blame]79are the most actively developed are the <a href="#clang">Clang Project</a> and
Nicolas Geoffray2b139382008-10-14 19:23:04 +0000[diff] [blame]80the <a href="#vmkit">VMKit Project</a>.
Chris Lattnerf6a0d382008-06-08 21:58:17 +0000[diff] [blame]81</p>
Chris Lattner78a130b2008-10-13 18:01:01 +0000[diff] [blame]82
Chris Lattnerf6a0d382008-06-08 21:58:17 +0000[diff] [blame]83</div>
84
Chris Lattner78a130b2008-10-13 18:01:01 +0000[diff] [blame]85
Chris Lattnerf6a0d382008-06-08 21:58:17 +0000[diff] [blame]86<!--=========================================================================-->
Chris Lattner78a130b2008-10-13 18:01:01 +0000[diff] [blame]87<div class="doc_subsection">
Chris Lattnera3c683b2008-10-13 18:11:54 +0000[diff] [blame]88<a name="clang">Clang: C/C++/Objective-C Frontend Toolkit</a>
Chris Lattnerf6a0d382008-06-08 21:58:17 +0000[diff] [blame]89</div>
90
91<div class="doc_text">
92
93<p>The <a href="http://clang.llvm.org/">Clang project</a> is an effort to build
94a set of new 'LLVM native' front-end technologies for the LLVM optimizer
95and code generator. Clang is continuing to make major strides forward in all
96areas. Its C and Objective-C parsing support is very solid, and the code
97generation support is far enough along to build many C applications. While not
98yet production quality, it is progressing very nicely. In addition, C++
99front-end work has started to make significant progress.</p>
100
Bill Wendling385b0d32008-10-27 09:27:33 +0000[diff] [blame]101<p>Clang, in conjunction with the <tt>ccc</tt> driver, is now usable as a
102replacement for gcc for building some small- to medium-sized C applications.
Daniel Dunbarf3e35782008-10-14 23:25:09 +0000[diff] [blame]103Additionally, Clang now has code generation support for Objective-C on Mac OS X
Bill Wendling385b0d32008-10-27 09:27:33 +0000[diff] [blame]104platform. Major highlights include:</p>
105
Daniel Dunbarf3e35782008-10-14 23:25:09 +0000[diff] [blame]106<ul>
107 <li> Clang/ccc pass almost all of the LLVM test suite on Mac OS X and Linux
108on the 32-bit x86 architecture. This includes significant C
109applications such as <a href="http://www.sqlite.org">sqlite3</a>,
110<a href="http://www.lua.org">lua</a>, and
Bill Wendling385b0d32008-10-27 09:27:33 +0000[diff] [blame]111<a href="http://www.clamav.net">Clam AntiVirus</a>. </li>
Daniel Dunbarf3e35782008-10-14 23:25:09 +0000[diff] [blame]112
113 <li> Clang can build the majority of Objective-C examples shipped with the
Bill Wendling385b0d32008-10-27 09:27:33 +0000[diff] [blame]114Mac OS X Developer Tools. </li>
Daniel Dunbarf3e35782008-10-14 23:25:09 +0000[diff] [blame]115</ul>
116
Bill Wendling385b0d32008-10-27 09:27:33 +0000[diff] [blame]117<p>Clang code generation still needs considerable testing and development,
118however. Some areas under active development include:</p>
119
Daniel Dunbarf3e35782008-10-14 23:25:09 +0000[diff] [blame]120<ul>
Bill Wendling385b0d32008-10-27 09:27:33 +0000[diff] [blame]121 <li> Improved support for C and Objective-C features, for example
122 variable-length arrays, va_arg, exception handling (Obj-C), and garbage
123 collection (Obj-C). </li>
124 <li> ABI compatibility, especially for platforms other than 32-bit
125 x86. </li>
Daniel Dunbarf3e35782008-10-14 23:25:09 +0000[diff] [blame]126</ul>
Chris Lattner2c3ef982008-10-11 18:18:28 +0000[diff] [blame]127
Chris Lattnera3c683b2008-10-13 18:11:54 +0000[diff] [blame]128</div>
129
130<!--=========================================================================-->
131<div class="doc_subsection">
132<a name="clangsa">Clang Static Analyzer</a>
133</div>
134
135<div class="doc_text">
136
Ted Kremenek0437b2c2008-10-14 05:14:21 +0000[diff] [blame]137<p>The Clang project also includes an early stage static source code analysis
138tool for <a href="http://clang.llvm.org/StaticAnalysis.html">automatically
139finding bugs</a> in C and Objective-C programs. The tool performs a growing set
Chris Lattnerff170152008-10-14 06:02:29 +0000[diff] [blame]140of checks to find bugs that occur on a specific path within a program. Examples
Ted Kremenek0437b2c2008-10-14 05:14:21 +0000[diff] [blame]141of bugs the tool finds include logic errors such as null dereferences,
142violations of various API rules, dead code, and potential memory leaks in
Chris Lattnerff170152008-10-14 06:02:29 +0000[diff] [blame]143Objective-C programs. Since its inception, public feedback on the tool has been
144extremely positive, and conservative estimates put the number of real bugs it
145has found in industrial-quality software on the order of thousands.</p>
Chris Lattnerc2d84672008-10-14 00:52:49 +0000[diff] [blame]146
Ted Kremenek0437b2c2008-10-14 05:14:21 +0000[diff] [blame]147<p>The tool also provides a simple web GUI to inspect potential bugs found by
Chris Lattnerff170152008-10-14 06:02:29 +0000[diff] [blame]148the tool. While still early in development, the GUI illustrates some of the key
Ted Kremenek0437b2c2008-10-14 05:14:21 +0000[diff] [blame]149features of Clang: accurate source location information, which is used by the
150GUI to highlight specific code expressions that relate to a bug (including those
Bill Wendling385b0d32008-10-27 09:27:33 +0000[diff] [blame]151that span multiple lines); and built-in knowledge of macros, which is used to
Ted Kremenek0437b2c2008-10-14 05:14:21 +0000[diff] [blame]152perform inline expansion of macros within the GUI itself.</p>
153
Chris Lattnerff170152008-10-14 06:02:29 +0000[diff] [blame]154<p>The set of checks performed by the static analyzer is gradually expanding,
Bill Wendling385b0d32008-10-27 09:27:33 +0000[diff] [blame]155and future plans for the tool include full source-level inter-procedural
156analysis and deeper checks such as buffer overrun detection. There are many
157opportunities to extend and enhance the static analyzer, and anyone interested
158in working on this project is encouraged to get involved!</p>
Chris Lattnerf6a0d382008-06-08 21:58:17 +0000[diff] [blame]159
160</div>
161
Chris Lattner78a130b2008-10-13 18:01:01 +0000[diff] [blame]162<!--=========================================================================-->
163<div class="doc_subsection">
Nicolas Geoffray2b139382008-10-14 19:23:04 +0000[diff] [blame]164<a name="vmkit">VMKit: JVM/CLI Virtual Machine Implementation</a>
Chris Lattner78a130b2008-10-13 18:01:01 +0000[diff] [blame]165</div>
166
167<div class="doc_text">
168<p>
Nicolas Geoffray2b139382008-10-14 19:23:04 +0000[diff] [blame]169The <a href="http://vmkit.llvm.org/">VMKit project</a> is an implementation of
Chris Lattner78a130b2008-10-13 18:01:01 +0000[diff] [blame]170a JVM and a CLI Virtual Machines (Microsoft .NET is an
171implementation of the CLI) using the Just-In-Time compiler of LLVM.</p>
172
Nicolas Geoffray20cfdd42008-10-15 20:25:04 +0000[diff] [blame]173<p>Following LLVM 2.4, VMKit has its first release 0.24 that you can find on its
174<a href="http://vmkit.llvm.org/releases/">webpage</a>. The release includes
175bug fixes, cleanup and new features. The major changes are:</p>
Chris Lattner78a130b2008-10-13 18:01:01 +0000[diff] [blame]176
Nicolas Geoffray2b139382008-10-14 19:23:04 +0000[diff] [blame]177<ul>
178
Bill Wendling385b0d32008-10-27 09:27:33 +0000[diff] [blame]179<li> Support for generics in the .Net virtual machine.</li>
180<li> Initial support for the Mono class libraries. </li>
Gabor Greif8ad68f12008-10-15 10:29:51 +0000[diff] [blame]181<li> Support for MacOSX/x86, following LLVM's support for exceptions in
Bill Wendling385b0d32008-10-27 09:27:33 +0000[diff] [blame]182JIT on MacOSX/x86. </li>
183<li> A new vmkit driver: a program to run java or .net applications. The driver
184supports llvm command line arguments including the new "-fast" option. </li>
Gabor Greif8ad68f12008-10-15 10:29:51 +0000[diff] [blame]185<li> A new memory allocation scheme in the JVM that makes unloading a
Bill Wendling385b0d32008-10-27 09:27:33 +0000[diff] [blame]186class loader very fast. </li>
187<li> VMKit now follows the LLVM Makefile machinery. </li>
Nicolas Geoffray2b139382008-10-14 19:23:04 +0000[diff] [blame]188
189</ul>
Chris Lattner78a130b2008-10-13 18:01:01 +0000[diff] [blame]190</div>
191
Chris Lattnerf6a0d382008-06-08 21:58:17 +0000[diff] [blame]192
193<!-- *********************************************************************** -->
194<div class="doc_section">
Chris Lattner78a130b2008-10-13 18:01:01 +0000[diff] [blame]195 <a name="whatsnew">What's New in LLVM?</a>
Chris Lattnerf6a0d382008-06-08 21:58:17 +0000[diff] [blame]196</div>
197<!-- *********************************************************************** -->
198
Chris Lattner5cce7a52008-06-08 22:59:35 +0000[diff] [blame]199<div class="doc_text">
200
Bill Wendling385b0d32008-10-27 09:27:33 +0000[diff] [blame]201<p>This release includes a huge number of bug fixes, performance tweaks, and
Chris Lattner92eacf82008-10-13 17:57:36 +0000[diff] [blame]202minor improvements. Some of the major improvements and new features are listed
203in this section.
Chris Lattner5cce7a52008-06-08 22:59:35 +0000[diff] [blame]204</p>
205</div>
206
Chris Lattnerf6a0d382008-06-08 21:58:17 +0000[diff] [blame]207<!--=========================================================================-->
Chris Lattnerf65224f2008-06-08 21:34:41 +0000[diff] [blame]208<div class="doc_subsection">
Chris Lattnere0798082008-02-06 18:00:06 +0000[diff] [blame]209<a name="majorfeatures">Major New Features</a>
210</div>
211
212<div class="doc_text">
213
Chris Lattner2c3ef982008-10-11 18:18:28 +0000[diff] [blame]214<p>LLVM 2.4 includes several major new capabilities:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]215
216<ul>
Chris Lattner4352cc82008-10-13 20:47:20 +0000[diff] [blame]217<li><p>The most visible end-user change in LLVM 2.4 is that it includes many
Chris Lattnerea9c2242008-10-13 21:50:36 +0000[diff] [blame]218optimizations and changes to make -O0 compile times much faster. You should see
Bill Wendling385b0d32008-10-27 09:27:33 +0000[diff] [blame]219improvements in speed on the order of 30% (or more) than in LLVM 2.3. There are
220many pieces to this change described in more detail below. The speedups and new
221components can also be used for JIT compilers that want fast
222compilation.</p></li>
Chris Lattner2c3ef982008-10-11 18:18:28 +0000[diff] [blame]223
Chris Lattner4352cc82008-10-13 20:47:20 +0000[diff] [blame]224<li><p>The biggest change to the LLVM IR is that Multiple Return Values (which
225were introduced in LLVM 2.3) have been generalized to full support for "First
226Class Aggregate" values in LLVM 2.4. This means that LLVM IR supports using
227structs and arrays as values in a function. This capability is mostly useful
228for front-end authors, who prefer to treat things like complex numbers, simple
Bill Wendling385b0d32008-10-27 09:27:33 +0000[diff] [blame]229tuples, dope vectors, etc., as Value*'s instead of as a tuple of Value*'s or as
Gabor Greif7c65c4f2008-10-14 11:00:32 +0000[diff] [blame]230memory values. Bitcode files from LLVM 2.3 will automatically migrate to the
231general representation.</p></li>
Mikhail Glushenkov808cd3e2008-10-13 02:08:34 +0000[diff] [blame]232
Chris Lattner4352cc82008-10-13 20:47:20 +0000[diff] [blame]233<li><p>LLVM 2.4 also includes an initial port for the PIC16 microprocessor. This
Bill Wendling385b0d32008-10-27 09:27:33 +0000[diff] [blame]234target only has support for 8 bit registers, and a number of other crazy
235constraints. While the port is still in early development stages, it shows some
236interesting things you can do with LLVM.</p></li>
Chris Lattnerba948522008-06-05 06:57:39 +0000[diff] [blame]237
Chris Lattner5ba16862008-02-10 08:18:42 +0000[diff] [blame]238</ul>
239
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]240</div>
241
Chris Lattnerc5a4e312008-06-05 06:25:56 +0000[diff] [blame]242
243<!--=========================================================================-->
244<div class="doc_subsection">
Chris Lattnerf65224f2008-06-08 21:34:41 +0000[diff] [blame]245<a name="llvm-gcc">llvm-gcc 4.2 Improvements</a>
Chris Lattnerc5a4e312008-06-05 06:25:56 +0000[diff] [blame]246</div>
247
248<div class="doc_text">
249
Chris Lattner4352cc82008-10-13 20:47:20 +0000[diff] [blame]250<p>LLVM fully supports the llvm-gcc 4.2 front-end, which marries the GCC
251front-ends and driver with the LLVM optimizer and code generator. It currently
252includes support for the C, C++, Objective-C, Ada, and Fortran front-ends.</p>
Chris Lattnerc5a4e312008-06-05 06:25:56 +0000[diff] [blame]253
Chris Lattner5cce7a52008-06-08 22:59:35 +0000[diff] [blame]254<ul>
Chris Lattnerf7bae6d2008-10-13 20:59:32 +0000[diff] [blame]255<li>LLVM 2.4 supports the full set of atomic <tt>__sync_*</tt> builtins. LLVM
Bill Wendling385b0d32008-10-27 09:27:33 +0000[diff] [blame]2562.3 only supported those used by OpenMP, but 2.4 supports them all. Note that
257while llvm-gcc supports all of these builtins, not all targets do. X86 support
258them all in both 32-bit and 64-bit mode and PowerPC supports them all except for
259the 64-bit operations when in 32-bit mode.</li>
Chris Lattner5cce7a52008-06-08 22:59:35 +0000[diff] [blame]260
Chris Lattnerf7bae6d2008-10-13 20:59:32 +0000[diff] [blame]261<li>llvm-gcc now supports an <tt>-flimited-precision</tt> option, which tells
Bill Wendling385b0d32008-10-27 09:27:33 +0000[diff] [blame]262the compiler that it is okay to use low-precision approximations of certain libm
263functions (like <tt>exp</tt>, <tt>log</tt>, etc). This allows you to get high
264performance if you only need (say) 12-bits of precision.</li>
Chris Lattner4352cc82008-10-13 20:47:20 +0000[diff] [blame]265
266<li>llvm-gcc now supports a C language extension known as "<a
Gabor Greif3babf012008-10-17 14:43:58 +0000[diff] [blame]267href="http://lists.cs.uiuc.edu/pipermail/cfe-dev/2008-August/002670.html">Blocks</a>".
268This feature is similar to nested functions and closures, but does not
Bill Wendling385b0d32008-10-27 09:27:33 +0000[diff] [blame]269require stack trampolines (with most ABIs), and supports returning closures
Chris Lattner4352cc82008-10-13 20:47:20 +0000[diff] [blame]270from functions that define them. Note that actually <em>using</em> Blocks
271requires a small runtime that is not included with llvm-gcc.</li>
272
273<li>llvm-gcc now supports a new <tt>-flto</tt> option. On systems that support
274transparent Link Time Optimization (currently Darwin systems with Xcode 3.1 and
275later) this allows the use of LTO with other optimization levels like -Os.
276Previously, LTO could only be used with -O4, which implied optimizations in
277-O3 that can increase code size.</li>
Gabor Greif0e49cc82008-06-09 06:06:18 +0000[diff] [blame]278</ul>
Chris Lattnerc5a4e312008-06-05 06:25:56 +0000[diff] [blame]279
Chris Lattnerc5a4e312008-06-05 06:25:56 +0000[diff] [blame]280</div>
281
282
Chris Lattnerb0d6c502007-09-21 03:54:09 +0000[diff] [blame]283<!--=========================================================================-->
284<div class="doc_subsection">
Chris Lattnerc3649452008-02-10 08:17:19 +0000[diff] [blame]285<a name="coreimprovements">LLVM Core Improvements</a>
286</div>
287
288<div class="doc_text">
Bill Wendling385b0d32008-10-27 09:27:33 +0000[diff] [blame]289<p>New features include:</p>
Chris Lattnerc3649452008-02-10 08:17:19 +0000[diff] [blame]290
Chris Lattnerc5a4e312008-06-05 06:25:56 +0000[diff] [blame]291<ul>
Gabor Greif7c65c4f2008-10-14 11:00:32 +0000[diff] [blame]292<li>A major change to the <tt>Use</tt> class landed, which shrank it by 25%. Since
Chris Lattnerf7bae6d2008-10-13 20:59:32 +0000[diff] [blame]293this is a pervasive part of the LLVM, it ended up reducing the memory use of
294LLVM IR in general by 15% for most programs.</li>
295
296<li>Values with no names are now pretty printed by <tt>llvm-dis</tt> more
297nicely. They now print as "<tt>%3 = add i32 %A, 4</tt>" instead of
298"<tt>add i32 %A, 4 ; &lt;i32&gt;:3</tt>", which makes it much easier to read.
299</li>
300
301<li>LLVM 2.4 includes some changes for better vector support. First, the shift
Bill Wendling385b0d32008-10-27 09:27:33 +0000[diff] [blame]302operations (<tt>shl</tt>, <tt>ashr</tt>, and <tt>lshr</tt>) now all support
303vectors and do an element-by-element shift (shifts of the whole vector can be
304accomplished by bitcasting the vector to &lt;1 x i128&gt;, for example). Second,
305there is initial support in development for vector comparisons with the
306<tt><a href="LangRef.html#i_fcmp">fcmp</a>/<a href="LangRef.html#i_icmp">icmp</a></tt>
Chris Lattnerc2d84672008-10-14 00:52:49 +0000[diff] [blame]307instructions. These instructions compare two vectors and return a vector of
Bill Wendling385b0d32008-10-27 09:27:33 +0000[diff] [blame]308<tt>i1</tt>'s for each result. Note that there is very little codegen support
309available for any of these IR features though.</li>
Chris Lattnerf7bae6d2008-10-13 20:59:32 +0000[diff] [blame]310
311<li>A new <tt>DebugInfoBuilder</tt> class is available, which makes it much
312easier for front-ends to create debug info descriptors, similar to the way that
Gabor Greifaa53e562008-10-14 11:31:14 +0000[diff] [blame]313<tt>IRBuilder</tt> makes it easier to create LLVM IR.</li>
Chris Lattnerf7bae6d2008-10-13 20:59:32 +0000[diff] [blame]314
Dan Gohmand74867f2008-10-14 16:13:59 +0000[diff] [blame]315<li>The <tt>IRBuilder</tt> class is now parameterized by a class responsible
Gabor Greif7c65c4f2008-10-14 11:00:32 +0000[diff] [blame]316for constant folding. The default <tt>ConstantFolder</tt> class does target independent
317constant folding. The <tt>NoFolder</tt> class does no constant folding at all, which is
318useful when learning how LLVM works. The <tt>TargetFolder</tt> class folds the most,
Duncan Sandsc16c9242008-10-14 07:06:37 +0000[diff] [blame]319doing target dependent constant folding.</li>
320
Bill Wendling385b0d32008-10-27 09:27:33 +0000[diff] [blame]321<li>LLVM now supports "function attributes", which allow us to separate return
Chris Lattnerf7bae6d2008-10-13 20:59:32 +0000[diff] [blame]322value attributes from function attributes. LLVM now supports attributes on a
323function itself, a return value, and its parameters. New supported function
Bill Wendling385b0d32008-10-27 09:27:33 +0000[diff] [blame]324attributes include <tt>noinline/alwaysinline</tt> and the <tt>opt-size</tt> flag,
325which says the function should be optimized for code size.</li>
Chris Lattnerf7bae6d2008-10-13 20:59:32 +0000[diff] [blame]326
Chris Lattnera5a49382008-10-12 18:30:33 +0000[diff] [blame]327<li>LLVM IR now directly represents "common" linkage, instead of
328 representing it as a form of weak linkage.</li>
Chris Lattnerf7bae6d2008-10-13 20:59:32 +0000[diff] [blame]329
Chris Lattnerc3649452008-02-10 08:17:19 +0000[diff] [blame]330</ul>
Mikhail Glushenkov808cd3e2008-10-13 02:08:34 +0000[diff] [blame]331
Chris Lattnerc3649452008-02-10 08:17:19 +0000[diff] [blame]332</div>
333
334<!--=========================================================================-->
335<div class="doc_subsection">
Chris Lattner8fda4952008-02-10 07:46:44 +0000[diff] [blame]336<a name="optimizer">Optimizer Improvements</a>
337</div>
338
339<div class="doc_text">
340
Chris Lattnerea9c2242008-10-13 21:50:36 +0000[diff] [blame]341<p>In addition to a huge array of bug fixes and minor performance tweaks, this
342release includes a few major enhancements and additions to the optimizers:</p>
Chris Lattner8fda4952008-02-10 07:46:44 +0000[diff] [blame]343
344<ul>
345
Chris Lattnerea9c2242008-10-13 21:50:36 +0000[diff] [blame]346<li>The Global Value Numbering (GVN) pass now does local Partial Redundancy
347Elimination (PRE) to eliminate some partially redundant expressions in cases
348where doing so won't grow code size.</li>
Chris Lattner4f076402008-10-12 19:47:48 +0000[diff] [blame]349
Chris Lattnerea9c2242008-10-13 21:50:36 +0000[diff] [blame]350<li>LLVM 2.4 includes a new loop deletion pass (which removes output-free
351provably-finite loops) and a rewritten Aggressive Dead Code Elimination (ADCE)
352pass that no longer uses control dependence information. These changes speed up
Gabor Greifaa53e562008-10-14 11:31:14 +0000[diff] [blame]353the optimizer and also prevent it from deleting output-free infinite
Chris Lattnerea9c2242008-10-13 21:50:36 +0000[diff] [blame]354loops.</li>
Chris Lattner4f076402008-10-12 19:47:48 +0000[diff] [blame]355
Duncan Sands2aecc3f2008-10-13 18:42:43 +0000[diff] [blame]356<li>The new AddReadAttrs pass works out which functions are read-only or
Gabor Greifaa53e562008-10-14 11:31:14 +0000[diff] [blame]357read-none (these correspond to 'pure' and 'const' in GCC) and marks them
Duncan Sands2aecc3f2008-10-13 18:42:43 +0000[diff] [blame]358with the appropriate attribute.</li>
Chris Lattnerea9c2242008-10-13 21:50:36 +0000[diff] [blame]359
360<li>LLVM 2.4 now includes a new SparsePropagation framework, which makes it
361trivial to build lattice-based dataflow solvers that operate over LLVM IR. Using
362this interface means that you just define objects to represent your lattice
363values and the transfer functions that operate on them. It handles the
364mechanics of worklist processing, liveness tracking, handling PHI nodes,
365etc.</li>
366
Chris Lattnerff170152008-10-14 06:02:29 +0000[diff] [blame]367<li>The Loop Strength Reduction and induction variable optimization passes have
368several improvements to avoid inserting MAX expressions, to optimize simple
369floating point induction variables and to analyze trip counts of more
370loops.</li>
371
Chris Lattnerea9c2242008-10-13 21:50:36 +0000[diff] [blame]372<li>Various helper functions (ComputeMaskedBits, ComputeNumSignBits, etc) were
373pulled out of the Instruction Combining pass and put into a new
374<tt>ValueTracking.h</tt> header, where they can be reused by other passes.</li>
375
376<li>The tail duplication pass has been removed from the standard optimizer
377sequence used by llvm-gcc. This pass still exists, but the benefits it once
378provided are now achieved by other passes.</li>
379
Chris Lattner0eef6e42008-06-08 02:45:07 +0000[diff] [blame]380</ul>
381
382</div>
383
384<!--=========================================================================-->
385<div class="doc_subsection">
386<a name="codegen">Code Generator Improvements</a>
387</div>
388
389<div class="doc_text">
390
Gabor Greifaa53e562008-10-14 11:31:14 +0000[diff] [blame]391<p>We have put a significant amount of work into the code generator infrastructure,
Chris Lattner0eef6e42008-06-08 02:45:07 +0000[diff] [blame]392which allows us to implement more aggressive algorithms and make it run
393faster:</p>
394
395<ul>
Chris Lattnerea9c2242008-10-13 21:50:36 +0000[diff] [blame]396<li>The target-independent code generator supports (and the X86 backend
397 currently implements) a new interface for "fast" instruction selection. This
398 interface is optimized to produce code as quickly as possible, sacrificing
399 code quality to do it. This is used by default at -O0 or when using
400 "llc -fast" on X86. It is straight-forward to add support for
401 other targets if faster -O0 compilation is desired.</li>
402
403<li>In addition to the new 'fast' instruction selection path, many existing
404 pieces of the code generator have been optimized in significant ways.
405 SelectionDAG's are now pool allocated and use better algorithms in many
Gabor Greifaa53e562008-10-14 11:31:14 +0000[diff] [blame]406 places, the ".s" file printers now use <tt>raw_ostream</tt> to emit text much faster,
Chris Lattnerea9c2242008-10-13 21:50:36 +0000[diff] [blame]407 etc. The end result of these improvements is that the compiler also takes
408 substantially less time to generate code that is just as good (and often
409 better) than before.</li>
410
Gabor Greifaa53e562008-10-14 11:31:14 +0000[diff] [blame]411<li>Each target has been split to separate the ".s" file printing logic from the
Chris Lattnerea9c2242008-10-13 21:50:36 +0000[diff] [blame]412 rest of the target. This enables JIT compilers that don't link in the
Gabor Greifaa53e562008-10-14 11:31:14 +0000[diff] [blame]413 (somewhat large) code and data tables used for printing a ".s" file.</li>
Chris Lattnerea9c2242008-10-13 21:50:36 +0000[diff] [blame]414
415<li>The code generator now includes a "stack slot coloring" pass, which packs
416 together individual spilled values into common stack slots. This reduces
417 the size of stack frames with many spills, which tends to increase L1 cache
418 effectiveness.</li>
419
420<li>Various pieces of the register allocator (e.g. the coalescer and two-address
421 operation elimination pass) now know how to rematerialize trivial operations
422 to avoid copies and include several other optimizations.</li>
423
424<li>The <a href="CodeGenerator.html#selectiondag_process">graphs</a> produced by
425 the <tt>llc -view-*-dags</tt> options are now significantly prettier and
426 easier to read.</li>
427
428<li>LLVM 2.4 includes a new register allocator based on Partitioned Boolean
429 Quadratic Programming (PBQP). This register allocator is still in
430 development, but is very simple and clean.</li>
Chris Lattner0eef6e42008-06-08 02:45:07 +0000[diff] [blame]431
Chris Lattnerb0d6c502007-09-21 03:54:09 +0000[diff] [blame]432</ul>
433
434</div>
435
436
437<!--=========================================================================-->
438<div class="doc_subsection">
Chris Lattnercc64a322008-10-13 22:06:31 +0000[diff] [blame]439<a name="targetspecific">Target Specific Improvements</a>
Chris Lattnerb0d6c502007-09-21 03:54:09 +0000[diff] [blame]440</div>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]441
442<div class="doc_text">
Chris Lattnerc3649452008-02-10 08:17:19 +0000[diff] [blame]443<p>New target-specific features include:
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]444</p>
445
446<ul>
Chris Lattner4f076402008-10-12 19:47:48 +0000[diff] [blame]447<li>Exception handling is supported by default on Linux/x86-64.</li>
Duncan Sandsd8745a92008-10-13 17:32:15 +0000[diff] [blame]448<li>Position Independent Code (PIC) is now supported on Linux/x86-64.</li>
Gabor Greif0abe3232008-10-27 10:17:24 +0000[diff] [blame]449<li><tt>@llvm.frameaddress</tt> now supports getting the frame address of stack frames
Chris Lattnerff170152008-10-14 06:02:29 +0000[diff] [blame]450 &gt; 0 on x86/x86-64.</li>
Bruno Cardoso Lopes62d45d72008-10-25 14:56:26 +0000[diff] [blame]451<li>MIPS has improved a lot since last release, the most important changes
452 are: Little endian support, floating point support, allegrex core and
453 intrinsics support. O32 ABI is improved but isn't complete. The EABI
454 was implemented and is fully supported. We also have support for small
455 sections and gp_rel relocation for its access, a threshold in bytes can be
456 specified through command line.</li>
Chris Lattnercc64a322008-10-13 22:06:31 +0000[diff] [blame]457<li>The PowerPC backend now supports trampolines.</li>
Chris Lattnere45f18b2008-06-05 08:02:49 +0000[diff] [blame]458</ul>
Chris Lattnere45f18b2008-06-05 08:02:49 +0000[diff] [blame]459
Mikhail Glushenkov808cd3e2008-10-13 02:08:34 +0000[diff] [blame]460</div>
Chris Lattnere45f18b2008-06-05 08:02:49 +0000[diff] [blame]461
462
463<!--=========================================================================-->
464<div class="doc_subsection">
Chris Lattnerb0d6c502007-09-21 03:54:09 +0000[diff] [blame]465<a name="otherimprovements">Other Improvements</a>
466</div>
467
468<div class="doc_text">
469<p>New features include:
470</p>
471
472<ul>
Mikhail Glushenkovcc2d3012008-10-13 02:46:01 +0000[diff] [blame]473<li><tt>llvmc2</tt> (the generic compiler driver) gained plugin
Chris Lattnercc64a322008-10-13 22:06:31 +0000[diff] [blame]474 support. It is now easier to experiment with <tt>llvmc2</tt> and
475 build your own tools based on it.</li>
476
477<li>LLVM 2.4 includes a number of new generic algorithms and data structures,
Dan Gohman8d9af692008-10-14 15:14:55 +0000[diff] [blame]478 include a scoped hash table, 'immutable' data structures, a simple
479 free-list manager, and a <tt>raw_ostream</tt> class.
480 The <tt>raw_ostream</tt> class and
Chris Lattnercc64a322008-10-13 22:06:31 +0000[diff] [blame]481 <tt>format</tt> allow for efficient file output, and various pieces of LLVM
482 have switched over to use it. The eventual goal is to eliminate
483 std::ostream in favor of it.</li>
484
Gabor Greif0640ac22008-11-04 21:48:10 +0000[diff] [blame^]485<li>LLVM 2.4 includes an optional build system based on CMake. It
486 still is in its early stages but can be useful for Visual C++
487 users who can not use the Visual Studio IDE.</li>
Tanya Lattnerc79f7542008-11-04 18:40:27 +0000[diff] [blame]488
Chris Lattnerb0d6c502007-09-21 03:54:09 +0000[diff] [blame]489</ul>
Mikhail Glushenkov808cd3e2008-10-13 02:08:34 +0000[diff] [blame]490
Chris Lattnerb0d6c502007-09-21 03:54:09 +0000[diff] [blame]491</div>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]492
Chris Lattner92eacf82008-10-13 17:57:36 +0000[diff] [blame]493<!--=========================================================================-->
494<div class="doc_subsection">
495<a name="changes">Major Changes and Removed Features</a>
496</div>
497
498<div class="doc_text">
499
Chris Lattnercc64a322008-10-13 22:06:31 +0000[diff] [blame]500<p>If you're already an LLVM user or developer with out-of-tree changes based
501on LLVM 2.3, this section lists some "gotchas" that you may run into upgrading
502from the previous release.</p>
Chris Lattner92eacf82008-10-13 17:57:36 +0000[diff] [blame]503
504<ul>
505
Chris Lattnercc64a322008-10-13 22:06:31 +0000[diff] [blame]506<li>The LLVM IR generated by llvm-gcc no longer names all instructions. This
507 makes it run faster, but may be more confusing to some people. If you
508 prefer to have names, the '<tt>opt -instnamer</tt>' pass will add names to
509 all instructions.</li>
510
511<li>The LoadVN and GCSE passes have been removed from the tree. They are
512 obsolete and have been replaced with the GVN and MemoryDependence passes.
513 </li>
514</ul>
515
516
517<p>In addition, many APIs have changed in this release. Some of the major LLVM
518API changes are:</p>
519
520<ul>
521
Devang Patel0e3da1a2008-10-14 20:03:43 +0000[diff] [blame]522<li>Now, function attributes and return value attributes are managed
Gabor Greif45fc8b12008-10-15 10:47:24 +0000[diff] [blame]523separately. Interface exported by <tt>ParameterAttributes.h</tt> header is now
Devang Patel0e3da1a2008-10-14 20:03:43 +0000[diff] [blame]524experted by <tt>Attributes.h</tt> header. The new attributes interface changes are:
525<ul>
526<li><tt>getParamAttrs</tt> method is now replaced by
527<tt>getParamAttributes</tt>, <tt>getRetAttributes</tt> and
528<tt>getFnAttributes</tt> methods.</li>
529<li> Return value attributes are stored at index 0. Function attributes are
530stored at index ~0U. Parameter attributes are stored at index that matches
531parameter number.</li>
532<li> <tt>ParamAttr</tt> namespace is now renamed as <tt>Attribute</tt>.</li>
Gabor Greif8ad68f12008-10-15 10:29:51 +0000[diff] [blame]533<li> The name of the class that manages reference count of opaque
Devang Patel0e3da1a2008-10-14 20:03:43 +0000[diff] [blame]534attributes is changed from <tt>PAListPtr</tt> to <tt>AttrListPtr</tt>.</li>
535<li> <tt>ParamAttrsWithIndex</tt> is now renamed as <tt>AttributeWithIndex</tt>.
536</li>
537</ul>
538</li>
Chris Lattner92eacf82008-10-13 17:57:36 +0000[diff] [blame]539
540<li>The <tt>DbgStopPointInst</tt> methods <tt>getDirectory</tt> and
541<tt>getFileName</tt> now return <tt>Value*</tt> instead of strings. These can be
542converted to strings using <tt>llvm::GetConstantStringInfo</tt> defined via
Gabor Greifaa53e562008-10-14 11:31:14 +0000[diff] [blame]543"<tt>llvm/Analysis/ValueTracking.h</tt>".</li>
Chris Lattner92eacf82008-10-13 17:57:36 +0000[diff] [blame]544
Chris Lattnercc64a322008-10-13 22:06:31 +0000[diff] [blame]545<li>The APIs to create various instructions have changed from lower case
546 "create" methods to upper case "Create" methods (e.g.
547 <tt>BinaryOperator::create</tt>). LLVM 2.4 includes both cases, but the
548 lower case ones are removed in mainline, please migrate.</li>
549
Gabor Greifaa53e562008-10-14 11:31:14 +0000[diff] [blame]550<li>Various header files like "<tt>llvm/ADT/iterator</tt>" were given a ".h" suffix.
551 Change your code to #include "<tt>llvm/ADT/iterator.h</tt>" instead.</li>
Chris Lattner92eacf82008-10-13 17:57:36 +0000[diff] [blame]552
Chris Lattner64e49952008-10-27 04:39:23 +0000[diff] [blame]553<li>The <tt>getresult</tt> instruction has been removed and replaced with the
554 <tt>extractvalue</tt> instruction. This is part of support for first class
555 aggregates.</li>
556
Gabor Greifaa53e562008-10-14 11:31:14 +0000[diff] [blame]557<li>In the code generator, many <tt>MachineOperand</tt> predicates were renamed to be
Chris Lattnerb7b69332008-10-14 06:37:11 +0000[diff] [blame]558 shorter (e.g. <tt>isFrameIndex()</tt> -&gt; <tt>isFI()</tt>),
559 <tt>SDOperand</tt> was renamed to <tt>SDValue</tt> (and the "<tt>Val</tt>"
560 member was changed to be the <tt>getNode()</tt> accessor), and the
561 <tt>MVT::ValueType</tt> enum has been replaced with an "<tt>MVT</tt>"
Dan Gohmanc16fad22008-10-14 17:06:44 +0000[diff] [blame]562 struct. The <tt>getSignExtended</tt> and <tt>getValue</tt> methods in the
563 ConstantSDNode class were renamed to <tt>getSExtValue</tt> and
564 <tt>getZExtValue</tt> respectively, to be more consistent with
565 the <tt>ConstantInt</tt> class.</li>
Chris Lattner92eacf82008-10-13 17:57:36 +0000[diff] [blame]566</ul>
567
568</div>
569
570
571
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]572<!-- *********************************************************************** -->
573<div class="doc_section">
574 <a name="portability">Portability and Supported Platforms</a>
575</div>
576<!-- *********************************************************************** -->
577
578<div class="doc_text">
579
580<p>LLVM is known to work on the following platforms:</p>
581
582<ul>
Mikhail Glushenkov808cd3e2008-10-13 02:08:34 +0000[diff] [blame]583<li>Intel and AMD machines (IA32) running Red Hat Linux, Fedora Core and FreeBSD
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]584 (and probably other unix-like systems).</li>
Chris Lattnerbd633de92008-02-06 06:30:34 +0000[diff] [blame]585<li>PowerPC and X86-based Mac OS X systems, running 10.3 and above in 32-bit and
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]586 64-bit modes.</li>
Chris Lattnere0798082008-02-06 18:00:06 +0000[diff] [blame]587<li>Intel and AMD machines running on Win32 using MinGW libraries (native).</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]588<li>Intel and AMD machines running on Win32 with the Cygwin libraries (limited
589 support is available for native builds with Visual C++).</li>
Gabor Greif5e75e652008-06-05 18:39:01 +0000[diff] [blame]590<li>Sun UltraSPARC workstations running Solaris 10.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]591<li>Alpha-based machines running Debian GNU/Linux.</li>
Gabor Greif5e75e652008-06-05 18:39:01 +0000[diff] [blame]592<li>Itanium-based (IA64) machines running Linux and HP-UX.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]593</ul>
594
Chris Lattnerba948522008-06-05 06:57:39 +0000[diff] [blame]595<p>The core LLVM infrastructure uses GNU autoconf to adapt itself
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]596to the machine and operating system on which it is built. However, minor
597porting may be required to get LLVM to work on new platforms. We welcome your
598portability patches and reports of successful builds or error messages.</p>
599
600</div>
601
602<!-- *********************************************************************** -->
603<div class="doc_section">
604 <a name="knownproblems">Known Problems</a>
605</div>
606<!-- *********************************************************************** -->
607
608<div class="doc_text">
609
610<p>This section contains all known problems with the LLVM system, listed by
611component. As new problems are discovered, they will be added to these
612sections. If you run into a problem, please check the <a
613href="http://llvm.org/bugs/">LLVM bug database</a> and submit a bug if
614there isn't already one.</p>
615
616</div>
617
618<!-- ======================================================================= -->
619<div class="doc_subsection">
620 <a name="experimental">Experimental features included with this release</a>
621</div>
622
623<div class="doc_text">
624
625<p>The following components of this LLVM release are either untested, known to
626be broken or unreliable, or are in early development. These components should
627not be relied on, and bugs should not be filed against them, but they may be
628useful to some people. In particular, if you would like to work on one of these
Chris Lattnerdaeb3232008-02-12 06:29:45 +0000[diff] [blame]629components, please contact us on the <a
630href="http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev">LLVMdev list</a>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]631
632<ul>
Gabor Greifaa53e562008-10-14 11:31:14 +0000[diff] [blame]633<li>The MSIL, IA64, Alpha, SPU, MIPS, and PIC16 backends are experimental.</li>
Duncan Sandsa73561c2008-06-08 20:18:35 +0000[diff] [blame]634<li>The llc "<tt>-filetype=asm</tt>" (the default) is the only supported
Chris Lattnerbd633de92008-02-06 06:30:34 +0000[diff] [blame]635 value for this option.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]636</ul>
637
638</div>
639
640<!-- ======================================================================= -->
641<div class="doc_subsection">
642 <a name="x86-be">Known problems with the X86 back-end</a>
643</div>
644
645<div class="doc_text">
646
647<ul>
Anton Korobeynikov77d15e22008-06-08 10:24:13 +0000[diff] [blame]648 <li>The X86 backend does not yet support
649 all <a href="http://llvm.org/PR879">inline assembly that uses the X86
650 floating point stack</a>. It supports the 'f' and 't' constraints, but not
651 'u'.</li>
Chris Lattnerd95b5962008-06-08 21:19:07 +0000[diff] [blame]652 <li>The X86 backend generates inefficient floating point code when configured
653 to generate code for systems that don't have SSE2.</li>
Duncan Sands8b4f3772008-06-08 19:38:43 +0000[diff] [blame]654 <li>Win64 code generation wasn't widely tested. Everything should work, but we
Anton Korobeynikov77d15e22008-06-08 10:24:13 +0000[diff] [blame]655 expect small issues to happen. Also, llvm-gcc cannot build mingw64 runtime
656 currently due
657 to <a href="http://llvm.org/PR2255">several</a>
Chris Lattnerf06141a2008-06-08 23:12:47 +0000[diff] [blame]658 <a href="http://llvm.org/PR2257">bugs</a> due to lack of support for the
659 'u' inline assembly constraint and X87 floating point inline assembly.</li>
Dan Gohman3882ecb2008-06-08 23:05:11 +0000[diff] [blame]660 <li>The X86-64 backend does not yet support the LLVM IR instruction
661 <tt>va_arg</tt>. Currently, the llvm-gcc front-end supports variadic
662 argument constructs on X86-64 by lowering them manually.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]663</ul>
664
665</div>
666
667<!-- ======================================================================= -->
668<div class="doc_subsection">
669 <a name="ppc-be">Known problems with the PowerPC back-end</a>
670</div>
671
672<div class="doc_text">
673
674<ul>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]675<li>The Linux PPC32/ABI support needs testing for the interpreter and static
676compilation, and lacks support for debug information.</li>
677</ul>
678
679</div>
680
681<!-- ======================================================================= -->
682<div class="doc_subsection">
683 <a name="arm-be">Known problems with the ARM back-end</a>
684</div>
685
686<div class="doc_text">
687
688<ul>
689<li>Thumb mode works only on ARMv6 or higher processors. On sub-ARMv6
Duncan Sands1827fa62007-09-26 15:59:54 +0000[diff] [blame]690processors, thumb programs can crash or produce wrong
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]691results (<a href="http://llvm.org/PR1388">PR1388</a>).</li>
692<li>Compilation for ARM Linux OABI (old ABI) is supported, but not fully tested.
693</li>
Chris Lattnerd95b5962008-06-08 21:19:07 +0000[diff] [blame]694<li>There is a bug in QEMU-ARM (&lt;= 0.9.0) which causes it to incorrectly
695 execute
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]696programs compiled with LLVM. Please use more recent versions of QEMU.</li>
697</ul>
698
699</div>
700
701<!-- ======================================================================= -->
702<div class="doc_subsection">
703 <a name="sparc-be">Known problems with the SPARC back-end</a>
704</div>
705
706<div class="doc_text">
707
708<ul>
709<li>The SPARC backend only supports the 32-bit SPARC ABI (-m32), it does not
710 support the 64-bit SPARC ABI (-m64).</li>
711</ul>
712
713</div>
714
715<!-- ======================================================================= -->
716<div class="doc_subsection">
Bruno Cardoso Lopes62d45d72008-10-25 14:56:26 +0000[diff] [blame]717 <a name="mips-be">Known problems with the MIPS back-end</a>
718</div>
719
720<div class="doc_text">
721
722<ul>
723<li>The O32 ABI is not fully supported.</li>
724<li>64-bit MIPS targets are not supported yet.</li>
725</ul>
726
727</div>
728
729<!-- ======================================================================= -->
730<div class="doc_subsection">
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]731 <a name="alpha-be">Known problems with the Alpha back-end</a>
732</div>
733
734<div class="doc_text">
735
736<ul>
737
738<li>On 21164s, some rare FP arithmetic sequences which may trap do not have the
739appropriate nops inserted to ensure restartability.</li>
740
741</ul>
742</div>
743
744<!-- ======================================================================= -->
745<div class="doc_subsection">
746 <a name="ia64-be">Known problems with the IA64 back-end</a>
747</div>
748
749<div class="doc_text">
750
751<ul>
Chris Lattnerf06141a2008-06-08 23:12:47 +0000[diff] [blame]752<li>The Itanium backend is highly experimental, and has a number of known
753 issues. We are looking for a maintainer for the Itanium backend. If you
Gabor Greif45fc8b12008-10-15 10:47:24 +0000[diff] [blame]754 are interested, please contact the LLVMdev mailing list.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]755</ul>
756
757</div>
758
759<!-- ======================================================================= -->
760<div class="doc_subsection">
761 <a name="c-be">Known problems with the C back-end</a>
762</div>
763
764<div class="doc_text">
765
766<ul>
Chris Lattner499c07c2008-06-05 06:35:40 +0000[diff] [blame]767<li><a href="http://llvm.org/PR802">The C backend has only basic support for
768 inline assembly code</a>.</li>
Chris Lattnere0194462007-09-26 06:01:35 +0000[diff] [blame]769<li><a href="http://llvm.org/PR1658">The C backend violates the ABI of common
770 C++ programs</a>, preventing intermixing between C++ compiled by the CBE and
Duncan Sandsa73561c2008-06-08 20:18:35 +0000[diff] [blame]771 C++ code compiled with llc or native compilers.</li>
Duncan Sandsbd0ad6e2008-02-10 13:40:55 +0000[diff] [blame]772<li>The C backend does not support all exception handling constructs.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]773</ul>
774
775</div>
776
777
778<!-- ======================================================================= -->
779<div class="doc_subsection">
Chris Lattner499c07c2008-06-05 06:35:40 +0000[diff] [blame]780 <a name="c-fe">Known problems with the llvm-gcc C front-end</a>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]781</div>
782
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]783<div class="doc_text">
784
Chris Lattnere0798082008-02-06 18:00:06 +0000[diff] [blame]785<p>llvm-gcc does not currently support <a href="http://llvm.org/PR869">Link-Time
786Optimization</a> on most platforms "out-of-the-box". Please inquire on the
Gabor Greif45fc8b12008-10-15 10:47:24 +0000[diff] [blame]787LLVMdev mailing list if you are interested.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]788
Chris Lattner499c07c2008-06-05 06:35:40 +0000[diff] [blame]789<p>The only major language feature of GCC not supported by llvm-gcc is
790 the <tt>__builtin_apply</tt> family of builtins. However, some extensions
791 are only supported on some targets. For example, trampolines are only
Duncan Sandsa73561c2008-06-08 20:18:35 +0000[diff] [blame]792 supported on some targets (these are used when you take the address of a
793 nested function).</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]794
Chris Lattner499c07c2008-06-05 06:35:40 +0000[diff] [blame]795<p>If you run into GCC extensions which are not supported, please let us know.
796</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]797
798</div>
799
800<!-- ======================================================================= -->
801<div class="doc_subsection">
Chris Lattner499c07c2008-06-05 06:35:40 +0000[diff] [blame]802 <a name="c++-fe">Known problems with the llvm-gcc C++ front-end</a>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]803</div>
804
805<div class="doc_text">
806
807<p>The C++ front-end is considered to be fully
808tested and works for a number of non-trivial programs, including LLVM
809itself, Qt, Mozilla, etc.</p>
810
811<ul>
Anton Korobeynikovd9a17932008-10-11 18:27:16 +0000[diff] [blame]812<li>Exception handling works well on the X86 and PowerPC targets. Currently
813 only linux and darwin targets are supported (both 32 and 64 bit).</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]814</ul>
815
816</div>
817
Chris Lattner6c8a1b82008-10-30 03:58:13 +0000[diff] [blame]818<!-- ======================================================================= -->
819<div class="doc_subsection">
820 <a name="fortran-fe">Known problems with the llvm-gcc Fortran front-end</a>
821</div>
822
823<div class="doc_text">
824<li>Fortran support generally works, but there are still several unresolved bugs
825 in Bugzilla. Please see the tools/gfortran component for details.</li>
826
827<li>The Fortran front-end currently does not build on Darwin (without tweaks)
828 due to unresolved dependencies on the C front-end.</li>
829
830</div>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]831
Chris Lattnerdaeb3232008-02-12 06:29:45 +0000[diff] [blame]832<!-- ======================================================================= -->
833<div class="doc_subsection">
Chris Lattner499c07c2008-06-05 06:35:40 +0000[diff] [blame]834 <a name="ada-fe">Known problems with the llvm-gcc Ada front-end</a>
Chris Lattnerdaeb3232008-02-12 06:29:45 +0000[diff] [blame]835</div>
836
837<div class="doc_text">
838The llvm-gcc 4.2 Ada compiler works fairly well, however this is not a mature
839technology and problems should be expected.
840<ul>
Duncan Sandsa73561c2008-06-08 20:18:35 +0000[diff] [blame]841<li>The Ada front-end currently only builds on X86-32. This is mainly due
Chris Lattnerdaeb3232008-02-12 06:29:45 +0000[diff] [blame]842to lack of trampoline support (pointers to nested functions) on other platforms,
Duncan Sandsa73561c2008-06-08 20:18:35 +0000[diff] [blame]843however it <a href="http://llvm.org/PR2006">also fails to build on X86-64</a>
Chris Lattnerdaeb3232008-02-12 06:29:45 +0000[diff] [blame]844which does support trampolines.</li>
845<li>The Ada front-end <a href="http://llvm.org/PR2007">fails to bootstrap</a>.
846Workaround: configure with --disable-bootstrap.</li>
Duncan Sands2dbc2e82008-10-13 17:27:23 +0000[diff] [blame]847<li>The c380004, <a href="http://llvm.org/PR2010">c393010</a>
848and <a href="http://llvm.org/PR2421">cxg2021</a> ACATS tests fail
849(c380004 also fails with gcc-4.2 mainline).</li>
850<li>Some gcc specific Ada tests continue to crash the compiler.</li>
Chris Lattnerdaeb3232008-02-12 06:29:45 +0000[diff] [blame]851<li>The -E binder option (exception backtraces)
852<a href="http://llvm.org/PR1982">does not work</a> and will result in programs
853crashing if an exception is raised. Workaround: do not use -E.</li>
854<li>Only discrete types <a href="http://llvm.org/PR1981">are allowed to start
855or finish at a non-byte offset</a> in a record. Workaround: do not pack records
856or use representation clauses that result in a field of a non-discrete type
857starting or finishing in the middle of a byte.</li>
Chris Lattnerd95b5962008-06-08 21:19:07 +0000[diff] [blame]858<li>The <tt>lli</tt> interpreter <a href="http://llvm.org/PR2009">considers
859'main' as generated by the Ada binder to be invalid</a>.
860Workaround: hand edit the file to use pointers for <tt>argv</tt> and
861<tt>envp</tt> rather than integers.</li>
862<li>The <tt>-fstack-check</tt> option <a href="http://llvm.org/PR2008">is
863ignored</a>.</li>
Chris Lattnerdaeb3232008-02-12 06:29:45 +0000[diff] [blame]864</ul>
865</div>
866
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]867<!-- *********************************************************************** -->
868<div class="doc_section">
869 <a name="additionalinfo">Additional Information</a>
870</div>
871<!-- *********************************************************************** -->
872
873<div class="doc_text">
874
875<p>A wide variety of additional information is available on the <a
876href="http://llvm.org">LLVM web page</a>, in particular in the <a
877href="http://llvm.org/docs/">documentation</a> section. The web page also
878contains versions of the API documentation which is up-to-date with the
879Subversion version of the source code.
880You can access versions of these documents specific to this release by going
881into the "<tt>llvm/doc/</tt>" directory in the LLVM tree.</p>
882
883<p>If you have any questions or comments about LLVM, please feel free to contact
884us via the <a href="http://llvm.org/docs/#maillist"> mailing
885lists</a>.</p>
886
887</div>
888
889<!-- *********************************************************************** -->
890
891<hr>
892<address>
893 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
894 src="http://jigsaw.w3.org/css-validator/images/vcss" alt="Valid CSS!"></a>
895 <a href="http://validator.w3.org/check/referer"><img
Chris Lattner5ba16862008-02-10 08:18:42 +0000[diff] [blame]896 src="http://www.w3.org/Icons/valid-html401" alt="Valid HTML 4.01!"></a>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]897
898 <a href="http://llvm.org/">LLVM Compiler Infrastructure</a><br>
899 Last modified: $Date$
900</address>
901
902</body>
903</html>