Blame - docs/ReleaseNotes.html - fp2-dev/platform/external/llvm

blob: edaecd1f0d0ec2999b4be18a48889c6bfd1d7746 [file] [log] [blame]

Dan Gohman	f17a25c	2007-07-18 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 Lattner	95d2507	2009-02-25 04:41:31 +0000	[diff] [blame]	7	<title>LLVM 2.5 Release Notes</title>
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	8	</head>
				9	<body>
				10
Chris Lattner	95d2507	2009-02-25 04:41:31 +0000	[diff] [blame]	11	<div class="doc_title">LLVM 2.5 Release Notes</div>
Mikhail Glushenkov	808cd3e	2008-10-13 02:08:34 +0000	[diff] [blame]	12
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	13	<ol>
				14	<li><a href="#intro">Introduction</a></li>
Chris Lattner	78a130b	2008-10-13 18:01:01 +0000	[diff] [blame]	15	<li><a href="#subproj">Sub-project Status Update</a></li>
Chris Lattner	b2cc502	2009-02-26 22:33:38 +0000	[diff] [blame]	16	<li><a href="#externalproj">External Projects Using LLVM 2.5</a></li>
Chris Lattner	47202c3	2009-03-02 02:37:32 +0000	[diff] [blame]	17	<li><a href="#whatsnew">What's New in LLVM 2.5?</a></li>
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	18	<li><a href="GettingStarted.html">Installation Instructions</a></li>
				19	<li><a href="#portability">Portability and Supported Platforms</a></li>
Dan Gohman	0de549c	2008-10-14 16:23:02 +0000	[diff] [blame]	20	<li><a href="#knownproblems">Known Problems</a></li>
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	21	<li><a href="#additionalinfo">Additional Information</a></li>
				22	</ol>
				23
				24	<div class="doc_author">
Dan Gohman	0de549c	2008-10-14 16:23:02 +0000	[diff] [blame]	25	<p>Written by the <a href="http://llvm.org">LLVM Team</a></p>
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	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
Chris Lattner	92eacf8	2008-10-13 17:57:36 +0000	[diff] [blame]	36	<p>This document contains the release notes for the LLVM Compiler
Chris Lattner	95d2507	2009-02-25 04:41:31 +0000	[diff] [blame]	37	Infrastructure, release 2.5. Here we describe the status of LLVM, including
Chris Lattner	92eacf8	2008-10-13 17:57:36 +0000	[diff] [blame]	38	major improvements from the previous release and significant known problems.
Mikhail Glushenkov	f61066a	2009-03-01 18:09:47 +0000	[diff] [blame]	39	All LLVM releases may be downloaded from the <a
Chris Lattner	92eacf8	2008-10-13 17:57:36 +0000	[diff] [blame]	40	href="http://llvm.org/releases/">LLVM releases web site</a>.</p>
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	41
				42	<p>For more information about LLVM, including information about the latest
				43	release, please check out the <a href="http://llvm.org/">main LLVM
				44	web site</a>. If you have questions or comments, the <a
Chris Lattner	92eacf8	2008-10-13 17:57:36 +0000	[diff] [blame]	45	href="http://mail.cs.uiuc.edu/mailman/listinfo/llvmdev">LLVM Developer's Mailing
				46	List</a> is a good place to send them.</p>
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	47
Mikhail Glushenkov	808cd3e	2008-10-13 02:08:34 +0000	[diff] [blame]	48	<p>Note that if you are reading this file from a Subversion checkout or the
Chris Lattner	b0d6c50	2007-09-21 03:54:09 +0000	[diff] [blame]	49	main LLVM web page, this document applies to the <i>next</i> release, not the
Gabor Greif	7c65c4f	2008-10-14 11:00:32 +0000	[diff] [blame]	50	current one. To see the release notes for a specific release, please see the
Chris Lattner	b0d6c50	2007-09-21 03:54:09 +0000	[diff] [blame]	51	<a href="http://llvm.org/releases/">releases page</a>.</p>
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	52
				53	</div>
				54
Chris Lattner	95d2507	2009-02-25 04:41:31 +0000	[diff] [blame]	55	<!-- Unfinished features in 2.5:
Chris Lattner	8fda495	2008-02-10 07:46:44 +0000	[diff] [blame]	56	Machine LICM
				57	Machine Sinking
Chris Lattner	a53f497	2009-02-25 06:34:50 +0000	[diff] [blame]	58	target-specific intrinsics
				59	gold lto plugin
Chris Lattner	56b6470	2009-02-25 07:32:53 +0000	[diff] [blame]	60	pre-alloc splitter, strong phi elim
Gabor Greif	e33c532	2009-03-02 12:02:51 +0000	[diff] [blame]	61	<tt>llc -enable-value-prop</tt>, propagation of value info
				62	(sign/zero ext info) from one MBB to another
Chris Lattner	a53f497	2009-02-25 06:34:50 +0000	[diff] [blame]	63	debug info for optimized code
Chris Lattner	5c10993	2009-02-26 07:29:17 +0000	[diff] [blame]	64	interpreter + libffi
Chris Lattner	b2cc502	2009-02-26 22:33:38 +0000	[diff] [blame]	65	postalloc scheduler: anti dependence breaking, hazard recognizer?
				66
Chris Lattner	47202c3	2009-03-02 02:37:32 +0000	[diff] [blame]	67	initial support for debug line numbers when optimization enabled, not useful in
				68	2.5 but will be for 2.6.
				69
Chris Lattner	8fda495	2008-02-10 07:46:44 +0000	[diff] [blame]	70	-->
Mikhail Glushenkov	808cd3e	2008-10-13 02:08:34 +0000	[diff] [blame]	71
Chris Lattner	4f07640	2008-10-12 19:47:48 +0000	[diff] [blame]	72	<!-- for announcement email:
Chris Lattner	a53f497	2009-02-25 06:34:50 +0000	[diff] [blame]	73	-->
Chris Lattner	8fda495	2008-02-10 07:46:44 +0000	[diff] [blame]	74
Chris Lattner	78a130b	2008-10-13 18:01:01 +0000	[diff] [blame]	75	<!-- *********************************************************************** -->
				76	<div class="doc_section">
				77	<a name="subproj">Sub-project Status Update</a>
Chris Lattner	f65224f	2008-06-08 21:34:41 +0000	[diff] [blame]	78	</div>
Chris Lattner	78a130b	2008-10-13 18:01:01 +0000	[diff] [blame]	79	<!-- *********************************************************************** -->
Chris Lattner	f65224f	2008-06-08 21:34:41 +0000	[diff] [blame]	80
				81	<div class="doc_text">
Chris Lattner	f6a0d38	2008-06-08 21:58:17 +0000	[diff] [blame]	82	<p>
Chris Lattner	95d2507	2009-02-25 04:41:31 +0000	[diff] [blame]	83	The LLVM 2.5 distribution currently consists of code from the core LLVM
Bill Wendling	a8fb81d	2009-03-02 04:28:18 +0000	[diff] [blame]	84	repository —which roughly includes the LLVM optimizers, code generators
				85	and supporting tools — and the llvm-gcc repository. In addition to this
				86	code, the LLVM Project includes other sub-projects that are in development. The
				87	two which are the most actively developed are the <a href="#clang">Clang
				88	Project</a> and the <a href="#vmkit">VMKit Project</a>.
Bill Wendling	dde41b8	2009-03-02 04:28:57 +0000	[diff] [blame]	89	</p>
Chris Lattner	78a130b	2008-10-13 18:01:01 +0000	[diff] [blame]	90
Chris Lattner	f6a0d38	2008-06-08 21:58:17 +0000	[diff] [blame]	91	</div>
				92
Chris Lattner	78a130b	2008-10-13 18:01:01 +0000	[diff] [blame]	93
Chris Lattner	f6a0d38	2008-06-08 21:58:17 +0000	[diff] [blame]	94	<!--=========================================================================-->
Chris Lattner	78a130b	2008-10-13 18:01:01 +0000	[diff] [blame]	95	<div class="doc_subsection">
Chris Lattner	a3c683b	2008-10-13 18:11:54 +0000	[diff] [blame]	96	<a name="clang">Clang: C/C++/Objective-C Frontend Toolkit</a>
Chris Lattner	f6a0d38	2008-06-08 21:58:17 +0000	[diff] [blame]	97	</div>
				98
				99	<div class="doc_text">
				100
				101	<p>The <a href="http://clang.llvm.org/">Clang project</a> is an effort to build
Bill Wendling	a8fb81d	2009-03-02 04:28:18 +0000	[diff] [blame]	102	a set of new 'LLVM native' front-end technologies for the LLVM optimizer and
				103	code generator. While Clang is not included in the LLVM 2.5 release, it is
				104	continuing to make major strides forward in all areas. Its C and Objective-C
Chris Lattner	c97507b	2009-02-25 05:09:54 +0000	[diff] [blame]	105	parsing and code generation support is now very solid. For example, it is
Bill Wendling	a8fb81d	2009-03-02 04:28:18 +0000	[diff] [blame]	106	capable of successfully building many real-world applications for X86-32
				107	andX86-64,
Chris Lattner	b2cc502	2009-02-26 22:33:38 +0000	[diff] [blame]	108	including <a href="http://wiki.freebsd.org/BuildingFreeBSDWithClang">the FreeBSD
Bill Wendling	a8fb81d	2009-03-02 04:28:18 +0000	[diff] [blame]	109	kernel</a>. C++ is also
				110	making <a href="http://clang.llvm.org/cxx_status.html">incredible progress</a>,
				111	and work on templates has recently started.</p>
Chris Lattner	f6a0d38	2008-06-08 21:58:17 +0000	[diff] [blame]	112
Chris Lattner	c97507b	2009-02-25 05:09:54 +0000	[diff] [blame]	113	<p>While Clang is not yet production quality, it is progressing very nicely and
				114	is quite usable for building many C and Objective-C applications. If you are
				115	interested in fast compiles and good diagnostics, we encourage you to try it out
Chris Lattner	a53f497	2009-02-25 06:34:50 +0000	[diff] [blame]	116	by <a href="http://clang.llvm.org/get_started.html">building from mainline</a>
Mikhail Glushenkov	f61066a	2009-03-01 18:09:47 +0000	[diff] [blame]	117	and reporting any issues you hit to the <a
Chris Lattner	c97507b	2009-02-25 05:09:54 +0000	[diff] [blame]	118	href="http://lists.cs.uiuc.edu/mailman/listinfo/cfe-dev">Clang front-end mailing
				119	list</a>.</p>
				120
				121	<p>In the LLVM 2.5 time-frame, the Clang team has made many improvements:</p>
Bill Wendling	385b0d3	2008-10-27 09:27:33 +0000	[diff] [blame]	122
Daniel Dunbar	f3e3578	2008-10-14 23:25:09 +0000	[diff] [blame]	123	<ul>
Chris Lattner	c97507b	2009-02-25 05:09:54 +0000	[diff] [blame]	124	<li>Clang now has a new driver, which is focused on providing a GCC-compatible
Mikhail Glushenkov	f61066a	2009-03-01 18:09:47 +0000	[diff] [blame]	125	interface.</li>
				126	<li>The X86-64 ABI is now supported.</li>
				127	<li>Precompiled header support is now implemented.</li>
Chris Lattner	c97507b	2009-02-25 05:09:54 +0000	[diff] [blame]	128	<li>Objective-C support is significantly improved beyond LLVM 2.4, supporting
				129	many features, such as Objective-C Garbage Collection.</li>
Chris Lattner	47202c3	2009-03-02 02:37:32 +0000	[diff] [blame]	130	<li>Many many bugs are fixed and many features have been added.</li>
Bill Wendling	a8fb81d	2009-03-02 04:28:18 +0000	[diff] [blame]	131	</ul>
Chris Lattner	a3c683b	2008-10-13 18:11:54 +0000	[diff] [blame]	132	</div>
				133
				134	<!--=========================================================================-->
				135	<div class="doc_subsection">
				136	<a name="clangsa">Clang Static Analyzer</a>
				137	</div>
				138
				139	<div class="doc_text">
				140
Ted Kremenek	a0d77d7	2009-02-27 07:01:32 +0000	[diff] [blame]	141	<p>Previously announced in the last LLVM release, the Clang project also
				142	includes an early stage static source code analysis tool for <a
				143	href="http://clang.llvm.org/StaticAnalysis.html">automatically finding bugs</a>
				144	in C and Objective-C programs. The tool performs a growing set of checks to find
				145	bugs that occur on a specific path within a program.</p>
Chris Lattner	c2d8467	2008-10-14 00:52:49 +0000	[diff] [blame]	146
Ted Kremenek	a0d77d7	2009-02-27 07:01:32 +0000	[diff] [blame]	147	<p>In the LLVM 2.5 time-frame there have been many significant improvements to
				148	the analyzer's core path simulation engine and machinery for generating
				149	path-based bug reports to end-users. Particularly noteworthy improvements
				150	include experimental support for full field-sensitivity and reasoning about heap
				151	objects as well as an improved value-constraints subengine that does a much
				152	better job of reasoning about inequality relationships (e.g., <tt>x > 2</tt>)
				153	between variables and constants.
Ted Kremenek	0437b2c	2008-10-14 05:14:21 +0000	[diff] [blame]	154
John Criswell	aea8c75	2009-03-02 15:28:15 +0000	[diff] [blame]	155	<p>The set of checks performed by the static analyzer continues to expand, and
Ted Kremenek	a0d77d7	2009-02-27 07:01:32 +0000	[diff] [blame]	156	future plans for the tool include full source-level inter-procedural analysis
				157	and deeper checks such as buffer overrun detection. There are many opportunities
				158	to extend and enhance the static analyzer, and anyone interested in working on
				159	this project is encouraged to get involved!</p>
Chris Lattner	f6a0d38	2008-06-08 21:58:17 +0000	[diff] [blame]	160
				161	</div>
				162
Chris Lattner	78a130b	2008-10-13 18:01:01 +0000	[diff] [blame]	163	<!--=========================================================================-->
				164	<div class="doc_subsection">
Nicolas Geoffray	2b13938	2008-10-14 19:23:04 +0000	[diff] [blame]	165	<a name="vmkit">VMKit: JVM/CLI Virtual Machine Implementation</a>
Chris Lattner	78a130b	2008-10-13 18:01:01 +0000	[diff] [blame]	166	</div>
				167
				168	<div class="doc_text">
				169	<p>
Nicolas Geoffray	2b13938	2008-10-14 19:23:04 +0000	[diff] [blame]	170	The <a href="http://vmkit.llvm.org/">VMKit project</a> is an implementation of
Chris Lattner	78a130b	2008-10-13 18:01:01 +0000	[diff] [blame]	171	a JVM and a CLI Virtual Machines (Microsoft .NET is an
				172	implementation of the CLI) using the Just-In-Time compiler of LLVM.</p>
				173
Nicolas Geoffray	8c3565e	2009-03-02 09:25:38 +0000	[diff] [blame]	174	<p>Following LLVM 2.5, VMKit has its second release that you can find on its
Nicolas Geoffray	20cfdd4	2008-10-15 20:25:04 +0000	[diff] [blame]	175	<a href="http://vmkit.llvm.org/releases/">webpage</a>. The release includes
				176	bug fixes, cleanup and new features. The major changes are:</p>
Chris Lattner	78a130b	2008-10-13 18:01:01 +0000	[diff] [blame]	177
Nicolas Geoffray	2b13938	2008-10-14 19:23:04 +0000	[diff] [blame]	178	<ul>
				179
Chris Lattner	47bcf61	2009-02-26 05:20:36 +0000	[diff] [blame]	180	<li>Ahead of Time compiler: compiles .class files to llvm .bc. VMKit uses this
Gabor Greif	ed585c6	2009-03-02 11:34:51 +0000	[diff] [blame]	181	functionality to native compile the standard classes (e.g. java.lang.String).
Gabor Greif	e33c532	2009-03-02 12:02:51 +0000	[diff] [blame]	182	Users can compile AoT .class files into dynamic libraries and run them with the
Chris Lattner	47bcf61	2009-02-26 05:20:36 +0000	[diff] [blame]	183	help of VMKit.</li>
				184
				185	<li>New exception model: the dwarf exception model is very slow for
				186	exception-intensive applications, so the JVM has had a new implementation of
				187	exceptions which check at each function call if an exception happened. There is
				188	a low performance penalty on applications without exceptions, but it is a big
				189	gain for exception-intensive applications. For example the jack benchmark in
				190	Spec JVM98 is 6x faster (performance gain of 83%).</li>
				191
Nicolas Geoffray	8c3565e	2009-03-02 09:25:38 +0000	[diff] [blame]	192	<li>User-level management of thread stacks, so that thread local data access
				193	at runtime is fast and portable. </li>
				194
				195	<li>Implementation of biased locking for faster object synchronizations at
				196	runtime.</li>
				197
				198	<li>New support for OSX/X64, Linux/X64 (with the Boehm GC) and Linux/ppc32.</li>
Nicolas Geoffray	2b13938	2008-10-14 19:23:04 +0000	[diff] [blame]	199
				200	</ul>
Chris Lattner	78a130b	2008-10-13 18:01:01 +0000	[diff] [blame]	201	</div>
				202
Chris Lattner	b2cc502	2009-02-26 22:33:38 +0000	[diff] [blame]	203	<!-- *********************************************************************** -->
				204	<div class="doc_section">
				205	<a name="externalproj">External Projects Using LLVM 2.5</a>
				206	</div>
				207	<!-- *********************************************************************** -->
				208
				209	<!--=========================================================================-->
				210	<div class="doc_subsection">
				211	<a name="pure">Pure</a>
				212	</div>
				213
				214	<div class="doc_text">
				215	<p>
Chris Lattner	47202c3	2009-03-02 02:37:32 +0000	[diff] [blame]	216	<a href="http://pure-lang.googlecode.com/">Pure</a>
				217	is an algebraic/functional programming language based on term rewriting.
Chris Lattner	b2cc502	2009-02-26 22:33:38 +0000	[diff] [blame]	218	Programs are collections of equations which are used to evaluate expressions in
				219	a symbolic fashion. Pure offers dynamic typing, eager and lazy evaluation,
Mikhail Glushenkov	f61066a	2009-03-01 18:09:47 +0000	[diff] [blame]	220	lexical closures, a hygienic macro system (also based on term rewriting),
Chris Lattner	b2cc502	2009-02-26 22:33:38 +0000	[diff] [blame]	221	built-in list and matrix support (including list and matrix comprehensions) and
				222	an easy-to-use C interface. The interpreter uses LLVM as a backend to
				223	JIT-compile Pure programs to fast native code.</p>
				224
				225	<p>In addition to the usual algebraic data structures, Pure also has
				226	MATLAB-style matrices in order to support numeric computations and signal
				227	processing in an efficient way. Pure is mainly aimed at mathematical
				228	applications right now, but it has been designed as a general purpose language.
				229	The dynamic interpreter environment and the C interface make it possible to use
				230	it as a kind of functional scripting language for many application areas.
				231	</p>
				232	</div>
				233
				234
				235	<!--=========================================================================-->
				236	<div class="doc_subsection">
				237	<a name="ldc">LLVM D Compiler</a>
				238	</div>
				239
				240	<div class="doc_text">
				241	<p>
Chris Lattner	47202c3	2009-03-02 02:37:32 +0000	[diff] [blame]	242	<a href="http://www.dsource.org/projects/ldc">LDC</a> is an implementation of
				243	the D Programming Language using the LLVM optimizer and code generator.
John Criswell	aea8c75	2009-03-02 15:28:15 +0000	[diff] [blame]	244	The LDC project works great with the LLVM 2.5 release. General improvements in
				245	this
Chris Lattner	47202c3	2009-03-02 02:37:32 +0000	[diff] [blame]	246	cycle have included new inline asm constraint handling, better debug info
				247	support, general bugfixes, and better x86-64 support. This has allowed
Chris Lattner	b2cc502	2009-02-26 22:33:38 +0000	[diff] [blame]	248	some major improvements in LDC, getting us much closer to being as
				249	fully featured as the original DMD compiler from DigitalMars.
				250	</p>
				251	</div>
				252
Chris Lattner	1ad599a	2009-02-28 18:58:01 +0000	[diff] [blame]	253	<!--=========================================================================-->
				254	<div class="doc_subsection">
				255	<a name="RoadsendPHP">Roadsend PHP</a>
				256	</div>
				257
				258	<div class="doc_text">
Chris Lattner	47202c3	2009-03-02 02:37:32 +0000	[diff] [blame]	259	<p><a href="http://code.roadsend.com/rphp">Roadsend PHP</a> (rphp) is an open
Chris Lattner	9e08de1	2009-03-02 19:07:24 +0000	[diff] [blame^]	260	source implementation of the PHP programming
				261	language that uses LLVM for its optimizer, JIT, and static compiler. This is a
				262	reimplementation of an earlier project that is now based on LLVM.</p>
Chris Lattner	1ad599a	2009-02-28 18:58:01 +0000	[diff] [blame]	263	</div>
Chris Lattner	b2cc502	2009-02-26 22:33:38 +0000	[diff] [blame]	264
Chris Lattner	f6a0d38	2008-06-08 21:58:17 +0000	[diff] [blame]	265
				266	<!-- *********************************************************************** -->
				267	<div class="doc_section">
Chris Lattner	47202c3	2009-03-02 02:37:32 +0000	[diff] [blame]	268	<a name="whatsnew">What's New in LLVM 2.5?</a>
Chris Lattner	f6a0d38	2008-06-08 21:58:17 +0000	[diff] [blame]	269	</div>
				270	<!-- *********************************************************************** -->
				271
Chris Lattner	5cce7a5	2008-06-08 22:59:35 +0000	[diff] [blame]	272	<div class="doc_text">
				273
Bill Wendling	385b0d3	2008-10-27 09:27:33 +0000	[diff] [blame]	274	<p>This release includes a huge number of bug fixes, performance tweaks, and
Chris Lattner	92eacf8	2008-10-13 17:57:36 +0000	[diff] [blame]	275	minor improvements. Some of the major improvements and new features are listed
				276	in this section.
Chris Lattner	5cce7a5	2008-06-08 22:59:35 +0000	[diff] [blame]	277	</p>
				278	</div>
				279
Chris Lattner	f6a0d38	2008-06-08 21:58:17 +0000	[diff] [blame]	280	<!--=========================================================================-->
Chris Lattner	f65224f	2008-06-08 21:34:41 +0000	[diff] [blame]	281	<div class="doc_subsection">
Chris Lattner	e079808	2008-02-06 18:00:06 +0000	[diff] [blame]	282	<a name="majorfeatures">Major New Features</a>
				283	</div>
				284
				285	<div class="doc_text">
				286
Chris Lattner	95d2507	2009-02-25 04:41:31 +0000	[diff] [blame]	287	<p>LLVM 2.5 includes several major new capabilities:</p>
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	288
				289	<ul>
Chris Lattner	47202c3	2009-03-02 02:37:32 +0000	[diff] [blame]	290	<li>LLVM 2.5 includes a brand new <a
				291	href="http://en.wikipedia.org/wiki/XCore">XCore</a> backend.</li>
Chris Lattner	a53f497	2009-02-25 06:34:50 +0000	[diff] [blame]	292
Bill Wendling	69afba8	2009-03-02 07:42:15 +0000	[diff] [blame]	293	<li>llvm-gcc now generally supports the GFortran front-end, and the precompiled
Chris Lattner	47202c3	2009-03-02 02:37:32 +0000	[diff] [blame]	294	release binaries now support Fortran, even on Mac OS/X.</li>
Chris Lattner	a53f497	2009-02-25 06:34:50 +0000	[diff] [blame]	295
Chris Lattner	47202c3	2009-03-02 02:37:32 +0000	[diff] [blame]	296	<li>CMake is now used by the <a href="GettingStartedVS.html">LLVM build process
				297	on Windows</a>. It automatically generates Visual Studio project files (and
				298	more) from a set of simple text files. This makes it much easier to
				299	maintain. In time, we'd like to standardize on CMake for everything.</li>
Chris Lattner	5c10993	2009-02-26 07:29:17 +0000	[diff] [blame]	300
Chris Lattner	47202c3	2009-03-02 02:37:32 +0000	[diff] [blame]	301	<li>LLVM 2.5 now uses (and includes) Google Test for unit testing.</li>
Chris Lattner	5c10993	2009-02-26 07:29:17 +0000	[diff] [blame]	302
Chris Lattner	47202c3	2009-03-02 02:37:32 +0000	[diff] [blame]	303	<li>The LLVM native code generator now supports arbitrary precision integers.
				304	Types like <tt>i33</tt> have long been valid in the LLVM IR, but were previously
				305	only supported by the interpreter. Note that the C backend still does not
				306	support these.</li>
Chris Lattner	56b6470	2009-02-25 07:32:53 +0000	[diff] [blame]	307
John Criswell	aea8c75	2009-03-02 15:28:15 +0000	[diff] [blame]	308	<li>LLVM 2.5 no longer uses 'bison,' so it is easier to build on Windows.</li>
Chris Lattner	5ba1686	2008-02-10 08:18:42 +0000	[diff] [blame]	309	</ul>
				310
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	311	</div>
				312
Chris Lattner	c5a4e31	2008-06-05 06:25:56 +0000	[diff] [blame]	313
				314	<!--=========================================================================-->
				315	<div class="doc_subsection">
Chris Lattner	f65224f	2008-06-08 21:34:41 +0000	[diff] [blame]	316	<a name="llvm-gcc">llvm-gcc 4.2 Improvements</a>
Chris Lattner	c5a4e31	2008-06-05 06:25:56 +0000	[diff] [blame]	317	</div>
				318
				319	<div class="doc_text">
				320
Chris Lattner	4352cc8	2008-10-13 20:47:20 +0000	[diff] [blame]	321	<p>LLVM fully supports the llvm-gcc 4.2 front-end, which marries the GCC
				322	front-ends and driver with the LLVM optimizer and code generator. It currently
				323	includes support for the C, C++, Objective-C, Ada, and Fortran front-ends.</p>
Chris Lattner	c5a4e31	2008-06-05 06:25:56 +0000	[diff] [blame]	324
Chris Lattner	5cce7a5	2008-06-08 22:59:35 +0000	[diff] [blame]	325	<ul>
Chris Lattner	47202c3	2009-03-02 02:37:32 +0000	[diff] [blame]	326	<li>In this release, the GCC inliner is completely disabled. Previously the GCC
				327	inliner was used to handle always-inline functions and other cases. This caused
				328	problems with code size growth, and it is completely disabled in this
				329	release.</li>
				330
				331	<li>llvm-gcc (and LLVM in general) now support code generation for stack
				332	canaries, which is an effective form of <a
				333	href="http://en.wikipedia.org/wiki/Stack-smashing_protection">buffer overflow
				334	protection</a>. llvm-gcc supports this with the <tt>-fstack-protector</tt>
				335	command line option (just like GCC). In LLVM IR, you can request code
				336	generation for stack canaries with function attributes.
				337	</li>
Gabor Greif	0e49cc8	2008-06-09 06:06:18 +0000	[diff] [blame]	338	</ul>
Chris Lattner	c5a4e31	2008-06-05 06:25:56 +0000	[diff] [blame]	339
Chris Lattner	c5a4e31	2008-06-05 06:25:56 +0000	[diff] [blame]	340	</div>
				341
				342
Chris Lattner	b0d6c50	2007-09-21 03:54:09 +0000	[diff] [blame]	343	<!--=========================================================================-->
				344	<div class="doc_subsection">
Chris Lattner	47202c3	2009-03-02 02:37:32 +0000	[diff] [blame]	345	<a name="coreimprovements">LLVM IR and Core Improvements</a>
Chris Lattner	c364945	2008-02-10 08:17:19 +0000	[diff] [blame]	346	</div>
				347
				348	<div class="doc_text">
Chris Lattner	47202c3	2009-03-02 02:37:32 +0000	[diff] [blame]	349	<p>LLVM IR has several new features that are used by our existing front-ends and
				350	can be useful if you are writing a front-end for LLVM:</p>
Chris Lattner	c364945	2008-02-10 08:17:19 +0000	[diff] [blame]	351
Chris Lattner	c5a4e31	2008-06-05 06:25:56 +0000	[diff] [blame]	352	<ul>
Chris Lattner	47202c3	2009-03-02 02:37:32 +0000	[diff] [blame]	353	<li>The <a href="LangRef.html#i_shufflevector">shufflevector</a> instruction
				354	has been generalized to allow different shuffle mask width than its input
				355	vectors. This allows you to use shufflevector to combine two
				356	"<4 x float>" vectors into a "<8 x float>" for example.</li>
				357
				358	<li>LLVM IR now supports new intrinsics for computing and acting on <a
				359	href="LangRef.html#int_overflow">overflow of integer operations</a>. This allows
				360	efficient code generation for languages that must trap or throw an exception on
				361	overflow. While these intrinsics work on all targets, they only generate
				362	efficient code on X86 so far.</li>
				363
				364	<li>LLVM IR now supports a new <a href="LangRef.html#linkage">private
				365	linkage</a> type to produce labels that are stripped by the assembler before it
				366	produces a .o file (thus they are invisible to the linker).</li>
				367
				368	<li>LLVM IR supports two new attributes for better alias analysis. The <a
				369	href="LangRef.html#paramattrs">noalias</a> attribute can now be used on the
				370	return value of a function to indicate that it returns new memory (e.g.
Duncan Sands	dd8d267	2009-03-02 16:58:00 +0000	[diff] [blame]	371	'malloc', 'calloc', etc).
				372	The new <a href="LangRef.html#paramattrs">nocapture</a> attribute can be used
				373	on pointer arguments to indicate that the function does not return the pointer,
				374	store it in an object that outlives the call, or let the value of the pointer
				375	escape from the function in any other way.
				376	Note that it is the pointer itself that must not escape, not the value it
				377	points to: loading a value out of the pointer is perfectly fine.
				378	Many standard library functions (e.g. 'strlen', 'memcpy') have this property.
				379	<!-- The simplifylibcalls pass applies these attributes to standard libc functions. -->
				380	</li>
Chris Lattner	47202c3	2009-03-02 02:37:32 +0000	[diff] [blame]	381
				382	<li>The parser for ".ll" files in lib/AsmParser is now completely rewritten as a
				383	recursive descent parser. This parser produces better error messages (including
John Criswell	aea8c75	2009-03-02 15:28:15 +0000	[diff] [blame]	384	caret diagnostics), is less fragile (less likely to crash on strange things),
				385	does not leak memory, is more efficient, and eliminates LLVM's last use of the
Chris Lattner	47202c3	2009-03-02 02:37:32 +0000	[diff] [blame]	386	'bison' tool.</li>
				387
				388	<li>Debug information representation and manipulation internals have been
				389	consolidated to use a new set of classes in
John Criswell	aea8c75	2009-03-02 15:28:15 +0000	[diff] [blame]	390	<tt>llvm/Analysis/DebugInfo.h</tt>. These routines are more
Chris Lattner	47202c3	2009-03-02 02:37:32 +0000	[diff] [blame]	391	efficient, robust, and extensible and replace the older mechanisms.
				392	llvm-gcc, clang, and the code generator now use them to create and process
				393	debug information.</li>
Mikhail Glushenkov	f61066a	2009-03-01 18:09:47 +0000	[diff] [blame]	394
Chris Lattner	c364945	2008-02-10 08:17:19 +0000	[diff] [blame]	395	</ul>
Mikhail Glushenkov	808cd3e	2008-10-13 02:08:34 +0000	[diff] [blame]	396
Chris Lattner	c364945	2008-02-10 08:17:19 +0000	[diff] [blame]	397	</div>
				398
				399	<!--=========================================================================-->
				400	<div class="doc_subsection">
Chris Lattner	8fda495	2008-02-10 07:46:44 +0000	[diff] [blame]	401	<a name="optimizer">Optimizer Improvements</a>
				402	</div>
				403
				404	<div class="doc_text">
				405
John Criswell	aea8c75	2009-03-02 15:28:15 +0000	[diff] [blame]	406	<p>In addition to a large array of bug fixes and minor performance tweaks, this
Chris Lattner	ea9c224	2008-10-13 21:50:36 +0000	[diff] [blame]	407	release includes a few major enhancements and additions to the optimizers:</p>
Chris Lattner	8fda495	2008-02-10 07:46:44 +0000	[diff] [blame]	408
				409	<ul>
				410
Chris Lattner	47202c3	2009-03-02 02:37:32 +0000	[diff] [blame]	411	<li>The loop optimizer now improves floating point induction variables in
				412	several ways, including adding shadow induction variables to avoid
				413	"integer <-> floating point" conversions in loops when safe.</li>
				414
John Criswell	aea8c75	2009-03-02 15:28:15 +0000	[diff] [blame]	415	<li>The "-mem2reg" pass is now much faster on code with large basic blocks.</li>
Chris Lattner	47202c3	2009-03-02 02:37:32 +0000	[diff] [blame]	416
				417	<li>The "-jump-threading" pass is more powerful: it is iterative
				418	and handles threading based on values with fully and partially redundant
				419	loads.</li>
				420
				421	<li>The "-memdep" memory dependence analysis pass (used by GVN and memcpyopt) is
				422	both faster and more aggressive.</li>
				423
				424	<li>The "-scalarrepl" scalar replacement of aggregates pass is more aggressive
				425	about promoting unions to registers.</li>
Chris Lattner	ea9c224	2008-10-13 21:50:36 +0000	[diff] [blame]	426
Chris Lattner	0eef6e4	2008-06-08 02:45:07 +0000	[diff] [blame]	427	</ul>
				428
				429	</div>
				430
				431	<!--=========================================================================-->
				432	<div class="doc_subsection">
Chris Lattner	5f39354	2009-03-02 03:24:11 +0000	[diff] [blame]	433	<a name="codegen">Target Independent Code Generator Improvements</a>
Chris Lattner	0eef6e4	2008-06-08 02:45:07 +0000	[diff] [blame]	434	</div>
				435
				436	<div class="doc_text">
				437
Mikhail Glushenkov	f61066a	2009-03-01 18:09:47 +0000	[diff] [blame]	438	<p>We have put a significant amount of work into the code generator
				439	infrastructure, which allows us to implement more aggressive algorithms and make
				440	it run faster:</p>
Chris Lattner	0eef6e4	2008-06-08 02:45:07 +0000	[diff] [blame]	441
				442	<ul>
Chris Lattner	5f39354	2009-03-02 03:24:11 +0000	[diff] [blame]	443	<li>The <a href="WritingAnLLVMBackend.html">Writing an LLVM Compiler
				444	Backend</a> document has been greatly expanded and is substantially more
				445	complete.</li>
				446
				447	<li>The SelectionDAG type legalization logic has been completely rewritten, is
				448	now more powerful (it supports arbitrary precision integer types for example),
John Criswell	aea8c75	2009-03-02 15:28:15 +0000	[diff] [blame]	449	and is more correct in several corner cases. The type legalizer converts
Chris Lattner	5f39354	2009-03-02 03:24:11 +0000	[diff] [blame]	450	operations on types that are not natively supported by the target machine into
				451	equivalent code sequences that only use natively supported types. The old type
				452	legalizer is still available (for now) and will be used if
				453	<tt>-disable-legalize-types</tt> is passed to the code generator.
Duncan Sands	335d5e4	2009-02-25 11:50:08 +0000	[diff] [blame]	454	</li>
Chris Lattner	0eef6e4	2008-06-08 02:45:07 +0000	[diff] [blame]	455
Chris Lattner	5f39354	2009-03-02 03:24:11 +0000	[diff] [blame]	456	<li>The code generator now supports widening illegal vectors to larger legal
				457	ones (for example, converting operations on <3 x float> to work on
				458	<4 x float>) which is very important for common graphics
				459	applications.</li>
Chris Lattner	47202c3	2009-03-02 02:37:32 +0000	[diff] [blame]	460
Chris Lattner	5f39354	2009-03-02 03:24:11 +0000	[diff] [blame]	461	<li>The assembly printers for each target are now split out into their own
				462	libraries that are separate from the main code generation logic. This reduces
John Criswell	aea8c75	2009-03-02 15:28:15 +0000	[diff] [blame]	463	the code size of JIT compilers by not requiring them to be linked in.</li>
Chris Lattner	47202c3	2009-03-02 02:37:32 +0000	[diff] [blame]	464
Chris Lattner	5f39354	2009-03-02 03:24:11 +0000	[diff] [blame]	465	<li>The 'fast' instruction selection path (used at -O0 and for fast JIT
				466	compilers) now supports accelerating codegen for code that uses exception
				467	handling constructs.</li>
				468
				469	<li>The optional PBQP register allocator now supports register coalescing.</li>
Chris Lattner	b0d6c50	2007-09-21 03:54:09 +0000	[diff] [blame]	470	</ul>
Chris Lattner	b0d6c50	2007-09-21 03:54:09 +0000	[diff] [blame]	471	</div>
				472
Chris Lattner	47202c3	2009-03-02 02:37:32 +0000	[diff] [blame]	473	<!--=========================================================================-->
				474	<div class="doc_subsection">
				475	<a name="x86">X86-32 and X86-64 Target Improvements</a>
				476	</div>
				477
				478	<div class="doc_text">
				479	<p>New features of the X86 target include:
				480	</p>
				481
				482	<ul>
Bill Wendling	9c7c438	2009-03-02 07:54:14 +0000	[diff] [blame]	483	<li>The <tt><a href="LangRef.html#int_returnaddress">llvm.returnaddress</a></tt>
				484	intrinsic (which is used to implement <tt>__builtin_return_address</tt>) now
				485	supports non-zero stack depths on X86.</li>
Chris Lattner	5f39354	2009-03-02 03:24:11 +0000	[diff] [blame]	486
				487	<li>The X86 backend now supports code generation of vector shift operations
				488	using SSE instructions.</li>
				489
Bill Wendling	9c7c438	2009-03-02 07:54:14 +0000	[diff] [blame]	490	<li>X86-64 code generation now takes advantage of red zone, unless the
				491	<tt>-mno-red-zone</tt> option is specified.</li>
Chris Lattner	5f39354	2009-03-02 03:24:11 +0000	[diff] [blame]	492
				493	<li>The X86 backend now supports using address space #256 in LLVM IR as a way of
				494	performing memory references off the GS segment register. This allows a
				495	front-end to take advantage of very low-level programming techniques when
Gabor Greif	ed585c6	2009-03-02 11:34:51 +0000	[diff] [blame]	496	targeting X86 CPUs. See <tt>test/CodeGen/X86/movgs.ll</tt> for a simple
Bill Wendling	9c7c438	2009-03-02 07:54:14 +0000	[diff] [blame]	497	example.</li>
Chris Lattner	5f39354	2009-03-02 03:24:11 +0000	[diff] [blame]	498
				499	<li>The X86 backend now supports a <tt>-disable-mmx</tt> command line option to
				500	prevent use of MMX even on chips that support it. This is important for cases
Bill Wendling	9c7c438	2009-03-02 07:54:14 +0000	[diff] [blame]	501	where code does not contain the proper <tt>llvm.x86.mmx.emms</tt>
				502	intrinsics.</li>
Chris Lattner	5f39354	2009-03-02 03:24:11 +0000	[diff] [blame]	503
Bill Wendling	9c7c438	2009-03-02 07:54:14 +0000	[diff] [blame]	504	<li>The X86 JIT now detects the new Intel <a
				505	href="http://en.wikipedia.org/wiki/Intel_Core_i7">Core i7</a> and <a
John Criswell	aea8c75	2009-03-02 15:28:15 +0000	[diff] [blame]	506	href="http://en.wikipedia.org/wiki/Intel_Atom">Atom</a> chips and
				507	auto-configures itself appropriately for the features of these chips.</li>
Chris Lattner	5f39354	2009-03-02 03:24:11 +0000	[diff] [blame]	508
				509	<li>The JIT now supports exception handling constructs on Linux/X86-64 and
				510	Darwin/x86-64.</li>
Bill Wendling	9c7c438	2009-03-02 07:54:14 +0000	[diff] [blame]	511
Chris Lattner	5f39354	2009-03-02 03:24:11 +0000	[diff] [blame]	512	<li>The JIT supports Thread Local Storage (TLS) on Linux/X86-32 but not yet on
				513	X86-64.</li>
Chris Lattner	47202c3	2009-03-02 02:37:32 +0000	[diff] [blame]	514	</ul>
				515
Chris Lattner	47202c3	2009-03-02 02:37:32 +0000	[diff] [blame]	516	</div>
Chris Lattner	b0d6c50	2007-09-21 03:54:09 +0000	[diff] [blame]	517
				518	<!--=========================================================================-->
				519	<div class="doc_subsection">
Chris Lattner	e1fbd75	2009-02-26 07:32:11 +0000	[diff] [blame]	520	<a name="pic16">PIC16 Target Improvements</a>
				521	</div>
				522
				523	<div class="doc_text">
				524	<p>New features of the PIC16 target include:
				525	</p>
				526
				527	<ul>
				528	<li>Both direct and indirect load/stores work now.</li>
				529	<li>Logical, bitwise and conditional operations now work for integer data
				530	types.</li>
				531	<li>Function calls involving basic types work now.</li>
				532	<li>Support for integer arrays.</li>
John Criswell	aea8c75	2009-03-02 15:28:15 +0000	[diff] [blame]	533	<li>The compiler can now emit libcalls for operations not supported by m/c
Bill Wendling	9c7c438	2009-03-02 07:54:14 +0000	[diff] [blame]	534	instructions.</li>
				535	<li>Support for both data and ROM address spaces.</li>
Chris Lattner	e1fbd75	2009-02-26 07:32:11 +0000	[diff] [blame]	536	</ul>
				537
				538	<p>Things not yet supported:</p>
				539
				540	<ul>
				541	<li>Floating point.</li>
Bill Wendling	9c7c438	2009-03-02 07:54:14 +0000	[diff] [blame]	542	<li>Passing/returning aggregate types to and from functions.</li>
Chris Lattner	e1fbd75	2009-02-26 07:32:11 +0000	[diff] [blame]	543	<li>Variable arguments.</li>
				544	<li>Indirect function calls.</li>
Chris Lattner	5f39354	2009-03-02 03:24:11 +0000	[diff] [blame]	545	<li>Interrupts/programs.</li>
Chris Lattner	e1fbd75	2009-02-26 07:32:11 +0000	[diff] [blame]	546	<li>Debug info.</li>
Chris Lattner	e1fbd75	2009-02-26 07:32:11 +0000	[diff] [blame]	547	</ul>
				548
				549	</div>
				550
				551
				552	<!--=========================================================================-->
				553	<div class="doc_subsection">
Chris Lattner	7ccc40d	2009-03-01 02:30:21 +0000	[diff] [blame]	554	<a name="llvmc">Improvements in LLVMC</a>
				555	</div>
				556
				557	<div class="doc_text">
				558	<p>New features include:</p>
				559
				560	<ul>
Mikhail Glushenkov	f61066a	2009-03-01 18:09:47 +0000	[diff] [blame]	561	<li>Beginning with LLVM 2.5, <tt>llvmc2</tt> is known as
				562	just <tt>llvmc</tt>. The old <tt>llvmc</tt> driver was removed.</li>
Chris Lattner	7ccc40d	2009-03-01 02:30:21 +0000	[diff] [blame]	563
				564	<li>The Clang plugin was substantially improved and is now enabled
Mikhail Glushenkov	f61066a	2009-03-01 18:09:47 +0000	[diff] [blame]	565	by default. The command <tt>llvmc --clang</tt> can be now used as a
				566	synonym to <tt>ccc</tt>.</li>
Chris Lattner	7ccc40d	2009-03-01 02:30:21 +0000	[diff] [blame]	567
Bill Wendling	9c7c438	2009-03-02 07:54:14 +0000	[diff] [blame]	568	<li>There is now a <tt>--check-graph</tt> option, which is supposed to catch
Mikhail Glushenkov	f61066a	2009-03-01 18:09:47 +0000	[diff] [blame]	569	common errors like multiple default edges, mismatched output/input language
				570	names and cycles. In general, these checks can't be done at compile-time
				571	because of the need to support plugins.</li>
Chris Lattner	7ccc40d	2009-03-01 02:30:21 +0000	[diff] [blame]	572
Mikhail Glushenkov	f61066a	2009-03-01 18:09:47 +0000	[diff] [blame]	573	<li>Plugins are now more flexible and can refer to compilation graph nodes and
				574	options defined in other plugins. To manage dependencies, a priority-sorting
Bill Wendling	9c7c438	2009-03-02 07:54:14 +0000	[diff] [blame]	575	mechanism was introduced. This change affects the TableGen file syntax. See the
Mikhail Glushenkov	f61066a	2009-03-01 18:09:47 +0000	[diff] [blame]	576	documentation for details.</li>
Chris Lattner	7ccc40d	2009-03-01 02:30:21 +0000	[diff] [blame]	577
Mikhail Glushenkov	f61066a	2009-03-01 18:09:47 +0000	[diff] [blame]	578	<li>Hooks can now be provided with arguments. The syntax is "<tt>$CALL(MyHook,
Bill Wendling	9c7c438	2009-03-02 07:54:14 +0000	[diff] [blame]	579	'Arg1', 'Arg2', 'Arg3')</tt>".</li>
Chris Lattner	7ccc40d	2009-03-01 02:30:21 +0000	[diff] [blame]	580
Mikhail Glushenkov	f61066a	2009-03-01 18:09:47 +0000	[diff] [blame]	581	<li>A new option type: multi-valued option, for options that take more than one
				582	argument (for example, "<tt>-foo a b c</tt>").</li>
Chris Lattner	7ccc40d	2009-03-01 02:30:21 +0000	[diff] [blame]	583
Mikhail Glushenkov	f61066a	2009-03-01 18:09:47 +0000	[diff] [blame]	584	<li>New option properties: '<tt>one_or_more</tt>', '<tt>zero_or_more</tt>',
				585	'<tt>hidden</tt>' and '<tt>really_hidden</tt>'.</li>
Chris Lattner	7ccc40d	2009-03-01 02:30:21 +0000	[diff] [blame]	586
Mikhail Glushenkov	f61066a	2009-03-01 18:09:47 +0000	[diff] [blame]	587	<li>The '<tt>case</tt>' expression gained an '<tt>error</tt>' action and
				588	an '<tt>empty</tt>' test (equivalent to "<tt>(not (not_empty ...))</tt>").</li>
Chris Lattner	7ccc40d	2009-03-01 02:30:21 +0000	[diff] [blame]	589
				590	<li>Documentation now looks more consistent to the rest of the LLVM
				591	docs. There is also a man page now.</li>
Mikhail Glushenkov	f61066a	2009-03-01 18:09:47 +0000	[diff] [blame]	592
Chris Lattner	7ccc40d	2009-03-01 02:30:21 +0000	[diff] [blame]	593	</ul>
				594
				595	</div>
				596
Chris Lattner	e45f18b	2008-06-05 08:02:49 +0000	[diff] [blame]	597
				598	<!--=========================================================================-->
				599	<div class="doc_subsection">
Chris Lattner	92eacf8	2008-10-13 17:57:36 +0000	[diff] [blame]	600	<a name="changes">Major Changes and Removed Features</a>
				601	</div>
				602
				603	<div class="doc_text">
				604
Chris Lattner	cc64a32	2008-10-13 22:06:31 +0000	[diff] [blame]	605	<p>If you're already an LLVM user or developer with out-of-tree changes based
Chris Lattner	95d2507	2009-02-25 04:41:31 +0000	[diff] [blame]	606	on LLVM 2.4, this section lists some "gotchas" that you may run into upgrading
Chris Lattner	cc64a32	2008-10-13 22:06:31 +0000	[diff] [blame]	607	from the previous release.</p>
Chris Lattner	92eacf8	2008-10-13 17:57:36 +0000	[diff] [blame]	608
				609	<ul>
				610
Bill Wendling	9c7c438	2009-03-02 07:54:14 +0000	[diff] [blame]	611	<li>llvm-gcc defaults to <tt>-fno-math-errno</tt> on all X86 targets.</li>
				612
Chris Lattner	cc64a32	2008-10-13 22:06:31 +0000	[diff] [blame]	613	</ul>
				614
				615
				616	<p>In addition, many APIs have changed in this release. Some of the major LLVM
				617	API changes are:</p>
				618
				619	<ul>
Gabor Greif	f7ea27c	2009-03-02 12:23:55 +0000	[diff] [blame]	620	<li>Some deprecated interfaces to create <tt>Instruction</tt> subclasses, that
John Criswell	aea8c75	2009-03-02 15:28:15 +0000	[diff] [blame]	621	were spelled with lower case "create," have been removed.</li>
Devang Patel	0e3da1a	2008-10-14 20:03:43 +0000	[diff] [blame]	622	</ul>
Chris Lattner	92eacf8	2008-10-13 17:57:36 +0000	[diff] [blame]	623
Chris Lattner	92eacf8	2008-10-13 17:57:36 +0000	[diff] [blame]	624	</div>
				625
				626
				627
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	628	<!-- *********************************************************************** -->
				629	<div class="doc_section">
				630	<a name="portability">Portability and Supported Platforms</a>
				631	</div>
				632	<!-- *********************************************************************** -->
				633
				634	<div class="doc_text">
				635
				636	<p>LLVM is known to work on the following platforms:</p>
				637
				638	<ul>
Mikhail Glushenkov	f61066a	2009-03-01 18:09:47 +0000	[diff] [blame]	639	<li>Intel and AMD machines (IA32, X86-64, AMD64, EMT-64) running Red Hat
Chris Lattner	b1ad47c	2008-11-10 05:40:34 +0000	[diff] [blame]	640	Linux, Fedora Core and FreeBSD (and probably other unix-like systems).</li>
Mikhail Glushenkov	f61066a	2009-03-01 18:09:47 +0000	[diff] [blame]	641	<li>PowerPC and X86-based Mac OS X systems, running 10.3 and above in 32-bit
Chris Lattner	b1ad47c	2008-11-10 05:40:34 +0000	[diff] [blame]	642	and 64-bit modes.</li>
Chris Lattner	e079808	2008-02-06 18:00:06 +0000	[diff] [blame]	643	<li>Intel and AMD machines running on Win32 using MinGW libraries (native).</li>
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	644	<li>Intel and AMD machines running on Win32 with the Cygwin libraries (limited
				645	support is available for native builds with Visual C++).</li>
Gabor Greif	5e75e65	2008-06-05 18:39:01 +0000	[diff] [blame]	646	<li>Sun UltraSPARC workstations running Solaris 10.</li>
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	647	<li>Alpha-based machines running Debian GNU/Linux.</li>
Gabor Greif	5e75e65	2008-06-05 18:39:01 +0000	[diff] [blame]	648	<li>Itanium-based (IA64) machines running Linux and HP-UX.</li>
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	649	</ul>
				650
Chris Lattner	ba94852	2008-06-05 06:57:39 +0000	[diff] [blame]	651	<p>The core LLVM infrastructure uses GNU autoconf to adapt itself
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	652	to the machine and operating system on which it is built. However, minor
				653	porting may be required to get LLVM to work on new platforms. We welcome your
				654	portability patches and reports of successful builds or error messages.</p>
				655
				656	</div>
				657
				658	<!-- *********************************************************************** -->
				659	<div class="doc_section">
				660	<a name="knownproblems">Known Problems</a>
				661	</div>
				662	<!-- *********************************************************************** -->
				663
				664	<div class="doc_text">
				665
Mikhail Glushenkov	f61066a	2009-03-01 18:09:47 +0000	[diff] [blame]	666	<p>This section contains significant known problems with the LLVM system,
Chris Lattner	b1ad47c	2008-11-10 05:40:34 +0000	[diff] [blame]	667	listed by component. If you run into a problem, please check the <a
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	668	href="http://llvm.org/bugs/">LLVM bug database</a> and submit a bug if
				669	there isn't already one.</p>
				670
				671	</div>
				672
				673	<!-- ======================================================================= -->
				674	<div class="doc_subsection">
				675	<a name="experimental">Experimental features included with this release</a>
				676	</div>
				677
				678	<div class="doc_text">
				679
				680	<p>The following components of this LLVM release are either untested, known to
				681	be broken or unreliable, or are in early development. These components should
				682	not be relied on, and bugs should not be filed against them, but they may be
				683	useful to some people. In particular, if you would like to work on one of these
Chris Lattner	daeb323	2008-02-12 06:29:45 +0000	[diff] [blame]	684	components, please contact us on the <a
				685	href="http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev">LLVMdev list</a>.</p>
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	686
				687	<ul>
Gabor Greif	aa53e56	2008-10-14 11:31:14 +0000	[diff] [blame]	688	<li>The MSIL, IA64, Alpha, SPU, MIPS, and PIC16 backends are experimental.</li>
Bill Wendling	9c7c438	2009-03-02 07:54:14 +0000	[diff] [blame]	689	<li>The <tt>llc</tt> "<tt>-filetype=asm</tt>" (the default) is the only
				690	supported value for this option.</li>
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	691	</ul>
				692
				693	</div>
				694
				695	<!-- ======================================================================= -->
				696	<div class="doc_subsection">
				697	<a name="x86-be">Known problems with the X86 back-end</a>
				698	</div>
				699
				700	<div class="doc_text">
				701
				702	<ul>
Anton Korobeynikov	77d15e2	2008-06-08 10:24:13 +0000	[diff] [blame]	703	<li>The X86 backend does not yet support
				704	all <a href="http://llvm.org/PR879">inline assembly that uses the X86
				705	floating point stack</a>. It supports the 'f' and 't' constraints, but not
				706	'u'.</li>
Chris Lattner	d95b596	2008-06-08 21:19:07 +0000	[diff] [blame]	707	<li>The X86 backend generates inefficient floating point code when configured
				708	to generate code for systems that don't have SSE2.</li>
Duncan Sands	8b4f377	2008-06-08 19:38:43 +0000	[diff] [blame]	709	<li>Win64 code generation wasn't widely tested. Everything should work, but we
John Criswell	aea8c75	2009-03-02 15:28:15 +0000	[diff] [blame]	710	expect small issues to happen. Also, llvm-gcc cannot build the mingw64
				711	runtime currently due
Anton Korobeynikov	77d15e2	2008-06-08 10:24:13 +0000	[diff] [blame]	712	to <a href="http://llvm.org/PR2255">several</a>
John Criswell	aea8c75	2009-03-02 15:28:15 +0000	[diff] [blame]	713	<a href="http://llvm.org/PR2257">bugs</a> and due to lack of support for
				714	the
				715	'u' inline assembly constraint and for X87 floating point inline assembly.</li>
Dan Gohman	3882ecb	2008-06-08 23:05:11 +0000	[diff] [blame]	716	<li>The X86-64 backend does not yet support the LLVM IR instruction
Chris Lattner	47202c3	2009-03-02 02:37:32 +0000	[diff] [blame]	717	<tt>va_arg</tt>. Currently, the llvm-gcc and front-ends support variadic
Dan Gohman	3882ecb	2008-06-08 23:05:11 +0000	[diff] [blame]	718	argument constructs on X86-64 by lowering them manually.</li>
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	719	</ul>
				720
				721	</div>
				722
				723	<!-- ======================================================================= -->
				724	<div class="doc_subsection">
				725	<a name="ppc-be">Known problems with the PowerPC back-end</a>
				726	</div>
				727
				728	<div class="doc_text">
				729
				730	<ul>
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	731	<li>The Linux PPC32/ABI support needs testing for the interpreter and static
				732	compilation, and lacks support for debug information.</li>
				733	</ul>
				734
				735	</div>
				736
				737	<!-- ======================================================================= -->
				738	<div class="doc_subsection">
				739	<a name="arm-be">Known problems with the ARM back-end</a>
				740	</div>
				741
				742	<div class="doc_text">
				743
				744	<ul>
				745	<li>Thumb mode works only on ARMv6 or higher processors. On sub-ARMv6
Duncan Sands	1827fa6	2007-09-26 15:59:54 +0000	[diff] [blame]	746	processors, thumb programs can crash or produce wrong
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	747	results (<a href="http://llvm.org/PR1388">PR1388</a>).</li>
John Criswell	aea8c75	2009-03-02 15:28:15 +0000	[diff] [blame]	748	<li>Compilation for ARM Linux OABI (old ABI) is supported but not fully tested.
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	749	</li>
Chris Lattner	d95b596	2008-06-08 21:19:07 +0000	[diff] [blame]	750	<li>There is a bug in QEMU-ARM (<= 0.9.0) which causes it to incorrectly
				751	execute
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	752	programs compiled with LLVM. Please use more recent versions of QEMU.</li>
				753	</ul>
				754
				755	</div>
				756
				757	<!-- ======================================================================= -->
				758	<div class="doc_subsection">
				759	<a name="sparc-be">Known problems with the SPARC back-end</a>
				760	</div>
				761
				762	<div class="doc_text">
				763
				764	<ul>
John Criswell	aea8c75	2009-03-02 15:28:15 +0000	[diff] [blame]	765	<li>The SPARC backend only supports the 32-bit SPARC ABI (-m32); it does not
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	766	support the 64-bit SPARC ABI (-m64).</li>
				767	</ul>
				768
				769	</div>
				770
				771	<!-- ======================================================================= -->
				772	<div class="doc_subsection">
Bruno Cardoso Lopes	62d45d7	2008-10-25 14:56:26 +0000	[diff] [blame]	773	<a name="mips-be">Known problems with the MIPS back-end</a>
				774	</div>
				775
				776	<div class="doc_text">
				777
				778	<ul>
				779	<li>The O32 ABI is not fully supported.</li>
				780	<li>64-bit MIPS targets are not supported yet.</li>
				781	</ul>
				782
				783	</div>
				784
				785	<!-- ======================================================================= -->
				786	<div class="doc_subsection">
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	787	<a name="alpha-be">Known problems with the Alpha back-end</a>
				788	</div>
				789
				790	<div class="doc_text">
				791
				792	<ul>
				793
				794	<li>On 21164s, some rare FP arithmetic sequences which may trap do not have the
				795	appropriate nops inserted to ensure restartability.</li>
				796
				797	</ul>
				798	</div>
				799
				800	<!-- ======================================================================= -->
				801	<div class="doc_subsection">
				802	<a name="ia64-be">Known problems with the IA64 back-end</a>
				803	</div>
				804
				805	<div class="doc_text">
				806
				807	<ul>
John Criswell	aea8c75	2009-03-02 15:28:15 +0000	[diff] [blame]	808	<li>The Itanium backend is highly experimental and has a number of known
Chris Lattner	f06141a	2008-06-08 23:12:47 +0000	[diff] [blame]	809	issues. We are looking for a maintainer for the Itanium backend. If you
Gabor Greif	45fc8b1	2008-10-15 10:47:24 +0000	[diff] [blame]	810	are interested, please contact the LLVMdev mailing list.</li>
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	811	</ul>
				812
				813	</div>
				814
				815	<!-- ======================================================================= -->
				816	<div class="doc_subsection">
				817	<a name="c-be">Known problems with the C back-end</a>
				818	</div>
				819
				820	<div class="doc_text">
				821
				822	<ul>
Chris Lattner	499c07c	2008-06-05 06:35:40 +0000	[diff] [blame]	823	<li><a href="http://llvm.org/PR802">The C backend has only basic support for
				824	inline assembly code</a>.</li>
Chris Lattner	e019446	2007-09-26 06:01:35 +0000	[diff] [blame]	825	<li><a href="http://llvm.org/PR1658">The C backend violates the ABI of common
				826	C++ programs</a>, preventing intermixing between C++ compiled by the CBE and
Gabor Greif	e33c532	2009-03-02 12:02:51 +0000	[diff] [blame]	827	C++ code compiled with <tt>llc</tt> or native compilers.</li>
Duncan Sands	bd0ad6e	2008-02-10 13:40:55 +0000	[diff] [blame]	828	<li>The C backend does not support all exception handling constructs.</li>
Duncan Sands	e6e264c	2009-02-25 11:51:54 +0000	[diff] [blame]	829	<li>The C backend does not support arbitrary precision integers.</li>
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	830	</ul>
				831
				832	</div>
				833
				834
				835	<!-- ======================================================================= -->
				836	<div class="doc_subsection">
Chris Lattner	499c07c	2008-06-05 06:35:40 +0000	[diff] [blame]	837	<a name="c-fe">Known problems with the llvm-gcc C front-end</a>
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	838	</div>
				839
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	840	<div class="doc_text">
				841
Chris Lattner	e079808	2008-02-06 18:00:06 +0000	[diff] [blame]	842	<p>llvm-gcc does not currently support <a href="http://llvm.org/PR869">Link-Time
				843	Optimization</a> on most platforms "out-of-the-box". Please inquire on the
Gabor Greif	45fc8b1	2008-10-15 10:47:24 +0000	[diff] [blame]	844	LLVMdev mailing list if you are interested.</p>
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	845
Chris Lattner	499c07c	2008-06-05 06:35:40 +0000	[diff] [blame]	846	<p>The only major language feature of GCC not supported by llvm-gcc is
				847	the <tt>__builtin_apply</tt> family of builtins. However, some extensions
				848	are only supported on some targets. For example, trampolines are only
Duncan Sands	a73561c	2008-06-08 20:18:35 +0000	[diff] [blame]	849	supported on some targets (these are used when you take the address of a
				850	nested function).</p>
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	851
Chris Lattner	499c07c	2008-06-05 06:35:40 +0000	[diff] [blame]	852	<p>If you run into GCC extensions which are not supported, please let us know.
				853	</p>
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	854
				855	</div>
				856
				857	<!-- ======================================================================= -->
				858	<div class="doc_subsection">
Chris Lattner	499c07c	2008-06-05 06:35:40 +0000	[diff] [blame]	859	<a name="c++-fe">Known problems with the llvm-gcc C++ front-end</a>
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	860	</div>
				861
				862	<div class="doc_text">
				863
				864	<p>The C++ front-end is considered to be fully
				865	tested and works for a number of non-trivial programs, including LLVM
				866	itself, Qt, Mozilla, etc.</p>
				867
				868	<ul>
Anton Korobeynikov	d9a1793	2008-10-11 18:27:16 +0000	[diff] [blame]	869	<li>Exception handling works well on the X86 and PowerPC targets. Currently
Chris Lattner	b1ad47c	2008-11-10 05:40:34 +0000	[diff] [blame]	870	only Linux and Darwin targets are supported (both 32 and 64 bit).</li>
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	871	</ul>
				872
				873	</div>
				874
Chris Lattner	6c8a1b8	2008-10-30 03:58:13 +0000	[diff] [blame]	875	<!-- ======================================================================= -->
				876	<div class="doc_subsection">
				877	<a name="fortran-fe">Known problems with the llvm-gcc Fortran front-end</a>
				878	</div>
				879
				880	<div class="doc_text">
Gabor Greif	1fc8fbc	2008-11-04 21:50:59 +0000	[diff] [blame]	881	<ul>
Chris Lattner	6c8a1b8	2008-10-30 03:58:13 +0000	[diff] [blame]	882	<li>Fortran support generally works, but there are still several unresolved bugs
				883	in Bugzilla. Please see the tools/gfortran component for details.</li>
Mikhail Glushenkov	f61066a	2009-03-01 18:09:47 +0000	[diff] [blame]	884
Chris Lattner	6c8a1b8	2008-10-30 03:58:13 +0000	[diff] [blame]	885	<li>The Fortran front-end currently does not build on Darwin (without tweaks)
				886	due to unresolved dependencies on the C front-end.</li>
Gabor Greif	1fc8fbc	2008-11-04 21:50:59 +0000	[diff] [blame]	887	</ul>
Chris Lattner	6c8a1b8	2008-10-30 03:58:13 +0000	[diff] [blame]	888	</div>
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	889
Chris Lattner	daeb323	2008-02-12 06:29:45 +0000	[diff] [blame]	890	<!-- ======================================================================= -->
				891	<div class="doc_subsection">
Chris Lattner	499c07c	2008-06-05 06:35:40 +0000	[diff] [blame]	892	<a name="ada-fe">Known problems with the llvm-gcc Ada front-end</a>
Chris Lattner	daeb323	2008-02-12 06:29:45 +0000	[diff] [blame]	893	</div>
				894
				895	<div class="doc_text">
John Criswell	aea8c75	2009-03-02 15:28:15 +0000	[diff] [blame]	896	The llvm-gcc 4.2 Ada compiler works fairly well; however, this is not a mature
				897	technology, and problems should be expected.
Chris Lattner	daeb323	2008-02-12 06:29:45 +0000	[diff] [blame]	898	<ul>
Duncan Sands	a73561c	2008-06-08 20:18:35 +0000	[diff] [blame]	899	<li>The Ada front-end currently only builds on X86-32. This is mainly due
John Criswell	aea8c75	2009-03-02 15:28:15 +0000	[diff] [blame]	900	to lack of trampoline support (pointers to nested functions) on other platforms.
				901	However, it <a href="http://llvm.org/PR2006">also fails to build on X86-64</a>
Chris Lattner	daeb323	2008-02-12 06:29:45 +0000	[diff] [blame]	902	which does support trampolines.</li>
				903	<li>The Ada front-end <a href="http://llvm.org/PR2007">fails to bootstrap</a>.
Duncan Sands	f73473e	2009-02-25 11:59:06 +0000	[diff] [blame]	904	This is due to lack of LLVM support for <tt>setjmp</tt>/<tt>longjmp</tt> style
				905	exception handling, which is used internally by the compiler.
Chris Lattner	daeb323	2008-02-12 06:29:45 +0000	[diff] [blame]	906	Workaround: configure with --disable-bootstrap.</li>
Duncan Sands	2dbc2e8	2008-10-13 17:27:23 +0000	[diff] [blame]	907	<li>The c380004, <a href="http://llvm.org/PR2010">c393010</a>
				908	and <a href="http://llvm.org/PR2421">cxg2021</a> ACATS tests fail
Duncan Sands	f73473e	2009-02-25 11:59:06 +0000	[diff] [blame]	909	(c380004 also fails with gcc-4.2 mainline).
				910	If the compiler is built with checks disabled then <a href="http://llvm.org/PR2010">c393010</a>
				911	causes the compiler to go into an infinite loop, using up all system memory.</li>
Duncan Sands	c2a29a0	2009-03-02 16:35:57 +0000	[diff] [blame]	912	<li>Some GCC specific Ada tests continue to crash the compiler.</li>
Chris Lattner	daeb323	2008-02-12 06:29:45 +0000	[diff] [blame]	913	<li>The -E binder option (exception backtraces)
				914	<a href="http://llvm.org/PR1982">does not work</a> and will result in programs
				915	crashing if an exception is raised. Workaround: do not use -E.</li>
				916	<li>Only discrete types <a href="http://llvm.org/PR1981">are allowed to start
				917	or finish at a non-byte offset</a> in a record. Workaround: do not pack records
				918	or use representation clauses that result in a field of a non-discrete type
				919	starting or finishing in the middle of a byte.</li>
Chris Lattner	d95b596	2008-06-08 21:19:07 +0000	[diff] [blame]	920	<li>The <tt>lli</tt> interpreter <a href="http://llvm.org/PR2009">considers
				921	'main' as generated by the Ada binder to be invalid</a>.
				922	Workaround: hand edit the file to use pointers for <tt>argv</tt> and
				923	<tt>envp</tt> rather than integers.</li>
				924	<li>The <tt>-fstack-check</tt> option <a href="http://llvm.org/PR2008">is
				925	ignored</a>.</li>
Chris Lattner	daeb323	2008-02-12 06:29:45 +0000	[diff] [blame]	926	</ul>
				927	</div>
				928
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	929	<!-- *********************************************************************** -->
				930	<div class="doc_section">
				931	<a name="additionalinfo">Additional Information</a>
				932	</div>
				933	<!-- *********************************************************************** -->
				934
				935	<div class="doc_text">
				936
				937	<p>A wide variety of additional information is available on the <a
				938	href="http://llvm.org">LLVM web page</a>, in particular in the <a
				939	href="http://llvm.org/docs/">documentation</a> section. The web page also
				940	contains versions of the API documentation which is up-to-date with the
				941	Subversion version of the source code.
				942	You can access versions of these documents specific to this release by going
				943	into the "<tt>llvm/doc/</tt>" directory in the LLVM tree.</p>
				944
				945	<p>If you have any questions or comments about LLVM, please feel free to contact
				946	us via the <a href="http://llvm.org/docs/#maillist"> mailing
				947	lists</a>.</p>
				948
				949	</div>
				950
				951	<!-- *********************************************************************** -->
				952
				953	<hr>
				954	<address>
				955	<a href="http://jigsaw.w3.org/css-validator/check/referer"><img
Misha Brukman	947321d	2008-12-11 17:34:48 +0000	[diff] [blame]	956	src="http://jigsaw.w3.org/css-validator/images/vcss-blue" alt="Valid CSS"></a>
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	957	<a href="http://validator.w3.org/check/referer"><img
Misha Brukman	947321d	2008-12-11 17:34:48 +0000	[diff] [blame]	958	src="http://www.w3.org/Icons/valid-html401-blue" alt="Valid HTML 4.01"></a>
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	959
				960	<a href="http://llvm.org/">LLVM Compiler Infrastructure</a><br>
				961	Last modified: $Date$
				962	</address>
				963
				964	</body>
				965	</html>