Gitiles
Code Review Sign In
gerrit-public.fairphone.software / fp2-dev / platform / external / llvm / c2d846779d42cba68ed8bf31aec4768da1977b9d / . / docs / ReleaseNotes.html
blob: 02974bba84171f8b34a588e22e1bf17e5cfa542f [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>
19 <li><a href="#knownproblems">Known Problems</a>
20 <li><a href="#additionalinfo">Additional Information</a></li>
21</ol>
22
23<div class="doc_author">
24 <p>Written by the <a href="http://llvm.org">LLVM Team</a><p>
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
49current one. To see the release notes for a specific releases, please see the
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
Duncan Sands275d1e02008-10-13 18:33:28 +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
Chris Lattnerc2d84672008-10-14 00:52:49 +0000[diff] [blame^]101<p>Codegen progress/state [DANIEL]</p>
Chris Lattner2c3ef982008-10-11 18:18:28 +0000[diff] [blame]102
Chris Lattnera3c683b2008-10-13 18:11:54 +0000[diff] [blame]103</div>
104
105<!--=========================================================================-->
106<div class="doc_subsection">
107<a name="clangsa">Clang Static Analyzer</a>
108</div>
109
110<div class="doc_text">
111
112<p>The
Chris Lattner2c3ef982008-10-11 18:18:28 +0000[diff] [blame]113<a href="http://clang.llvm.org/StaticAnalysis.html">static analysis tool</a>
Chris Lattnerc2d84672008-10-14 00:52:49 +0000[diff] [blame^]114
115[TED]
Chris Lattnera3c683b2008-10-13 18:11:54 +0000[diff] [blame]116.</p>
Chris Lattnerf6a0d382008-06-08 21:58:17 +0000[diff] [blame]117
118</div>
119
Chris Lattner78a130b2008-10-13 18:01:01 +0000[diff] [blame]120<!--=========================================================================-->
121<div class="doc_subsection">
Chris Lattnera3c683b2008-10-13 18:11:54 +0000[diff] [blame]122<a name="vmkit">vmkit: JVM/CLI Virtual Machine Implementation</a>
Chris Lattner78a130b2008-10-13 18:01:01 +0000[diff] [blame]123</div>
124
125<div class="doc_text">
126<p>
Chris Lattnera3c683b2008-10-13 18:11:54 +0000[diff] [blame]127The <a href="http://vmkit.llvm.org/">vmkit project</a> is an implementation of
Chris Lattner78a130b2008-10-13 18:01:01 +0000[diff] [blame]128a JVM and a CLI Virtual Machines (Microsoft .NET is an
129implementation of the CLI) using the Just-In-Time compiler of LLVM.</p>
130
Chris Lattnerc2d84672008-10-14 00:52:49 +0000[diff] [blame^]131<p>[NICOLAS]</p>
Chris Lattner78a130b2008-10-13 18:01:01 +0000[diff] [blame]132
133</div>
134
Chris Lattnerf6a0d382008-06-08 21:58:17 +0000[diff] [blame]135
136<!-- *********************************************************************** -->
137<div class="doc_section">
Chris Lattner78a130b2008-10-13 18:01:01 +0000[diff] [blame]138 <a name="whatsnew">What's New in LLVM?</a>
Chris Lattnerf6a0d382008-06-08 21:58:17 +0000[diff] [blame]139</div>
140<!-- *********************************************************************** -->
141
Chris Lattner5cce7a52008-06-08 22:59:35 +0000[diff] [blame]142<div class="doc_text">
143
Chris Lattner92eacf82008-10-13 17:57:36 +0000[diff] [blame]144<p>This release includes a huge number of bug fixes, performance tweaks and
145minor improvements. Some of the major improvements and new features are listed
146in this section.
Chris Lattner5cce7a52008-06-08 22:59:35 +0000[diff] [blame]147</p>
148</div>
149
Chris Lattnerf6a0d382008-06-08 21:58:17 +0000[diff] [blame]150<!--=========================================================================-->
Chris Lattnerf65224f2008-06-08 21:34:41 +0000[diff] [blame]151<div class="doc_subsection">
Chris Lattnere0798082008-02-06 18:00:06 +0000[diff] [blame]152<a name="majorfeatures">Major New Features</a>
153</div>
154
155<div class="doc_text">
156
Chris Lattner2c3ef982008-10-11 18:18:28 +0000[diff] [blame]157<p>LLVM 2.4 includes several major new capabilities:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]158
159<ul>
Chris Lattner4352cc82008-10-13 20:47:20 +0000[diff] [blame]160<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]161optimizations and changes to make -O0 compile times much faster. You should see
Chris Lattner4352cc82008-10-13 20:47:20 +0000[diff] [blame]162improvements on the order of 30% or more faster than LLVM 2.3. There are many
163pieces to this change, described in more detail below. The speedups and new
164components can also be used for JIT compilers that want fast compilation as
165well.</p></li>
Chris Lattner2c3ef982008-10-11 18:18:28 +0000[diff] [blame]166
Chris Lattner4352cc82008-10-13 20:47:20 +0000[diff] [blame]167<li><p>The biggest change to the LLVM IR is that Multiple Return Values (which
168were introduced in LLVM 2.3) have been generalized to full support for "First
169Class Aggregate" values in LLVM 2.4. This means that LLVM IR supports using
170structs and arrays as values in a function. This capability is mostly useful
171for front-end authors, who prefer to treat things like complex numbers, simple
172tuples, dope vectors, etc as Value*'s instead of as a tuple of Value*'s or as
173memory values.</p></li>
Mikhail Glushenkov808cd3e2008-10-13 02:08:34 +0000[diff] [blame]174
Chris Lattner4352cc82008-10-13 20:47:20 +0000[diff] [blame]175<li><p>LLVM 2.4 also includes an initial port for the PIC16 microprocessor. This
176is the LLVM targer that only has support for 8 bit registers, and a number of
177other crazy constraints. While the port is still in early development stages,
178it shows some interesting things you can do with LLVM.</p></li>
Chris Lattnerba948522008-06-05 06:57:39 +0000[diff] [blame]179
Chris Lattner5ba16862008-02-10 08:18:42 +0000[diff] [blame]180</ul>
181
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]182</div>
183
Chris Lattnerc5a4e312008-06-05 06:25:56 +0000[diff] [blame]184
185<!--=========================================================================-->
186<div class="doc_subsection">
Chris Lattnerf65224f2008-06-08 21:34:41 +0000[diff] [blame]187<a name="llvm-gcc">llvm-gcc 4.2 Improvements</a>
Chris Lattnerc5a4e312008-06-05 06:25:56 +0000[diff] [blame]188</div>
189
190<div class="doc_text">
191
Chris Lattner4352cc82008-10-13 20:47:20 +0000[diff] [blame]192<p>LLVM fully supports the llvm-gcc 4.2 front-end, which marries the GCC
193front-ends and driver with the LLVM optimizer and code generator. It currently
194includes support for the C, C++, Objective-C, Ada, and Fortran front-ends.</p>
Chris Lattnerc5a4e312008-06-05 06:25:56 +0000[diff] [blame]195
Chris Lattner5cce7a52008-06-08 22:59:35 +0000[diff] [blame]196<ul>
Chris Lattnerf7bae6d2008-10-13 20:59:32 +0000[diff] [blame]197<li>LLVM 2.4 supports the full set of atomic <tt>__sync_*</tt> builtins. LLVM
1982.3 only supported those used by OpenMP, but 2.4 supports them all. While
Chris Lattnerea9c2242008-10-13 21:50:36 +0000[diff] [blame]199llvm-gcc supports all of these builtins, note that not all targets do. X86
200support them all in both 32-bit and 64-bit mode and PowerPC supports them all
201except for the 64-bit operations when in 32-bit mode.</li>
Chris Lattner5cce7a52008-06-08 22:59:35 +0000[diff] [blame]202
Chris Lattnerf7bae6d2008-10-13 20:59:32 +0000[diff] [blame]203<li>llvm-gcc now supports an <tt>-flimited-precision</tt> option, which tells
204the compiler that it is ok to use low-precision approximations of certain libm
Chris Lattner4352cc82008-10-13 20:47:20 +0000[diff] [blame]205functions (like tan, log, etc). This allows you to get high performance if you
206only need (say) 14-bits of precision.</li>
207
208<li>llvm-gcc now supports a C language extension known as "<a
209href="http://lists.cs.uiuc.edu/pipermail/cfe-dev/2008-August/002670.html">Blocks
210</a>. This feature is similar to nested functions and closures, but does not
211require stack trampolines (with most ABIs) and supports returning closures
212from functions that define them. Note that actually <em>using</em> Blocks
213requires a small runtime that is not included with llvm-gcc.</li>
214
215<li>llvm-gcc now supports a new <tt>-flto</tt> option. On systems that support
216transparent Link Time Optimization (currently Darwin systems with Xcode 3.1 and
217later) this allows the use of LTO with other optimization levels like -Os.
218Previously, LTO could only be used with -O4, which implied optimizations in
219-O3 that can increase code size.</li>
Gabor Greif0e49cc82008-06-09 06:06:18 +0000[diff] [blame]220</ul>
Chris Lattnerc5a4e312008-06-05 06:25:56 +0000[diff] [blame]221
Chris Lattnerc5a4e312008-06-05 06:25:56 +0000[diff] [blame]222</div>
223
224
Chris Lattnerb0d6c502007-09-21 03:54:09 +0000[diff] [blame]225<!--=========================================================================-->
226<div class="doc_subsection">
Chris Lattnerc3649452008-02-10 08:17:19 +0000[diff] [blame]227<a name="coreimprovements">LLVM Core Improvements</a>
228</div>
229
230<div class="doc_text">
231<p>New features include:
232</p>
233
Chris Lattnerc5a4e312008-06-05 06:25:56 +0000[diff] [blame]234<ul>
Chris Lattnerf7bae6d2008-10-13 20:59:32 +0000[diff] [blame]235<li>A major change to the "Use" class landed, which shrank it by 25%. Since
236this is a pervasive part of the LLVM, it ended up reducing the memory use of
237LLVM IR in general by 15% for most programs.</li>
238
239<li>Values with no names are now pretty printed by <tt>llvm-dis</tt> more
240nicely. They now print as "<tt>%3 = add i32 %A, 4</tt>" instead of
241"<tt>add i32 %A, 4 ; &lt;i32&gt;:3</tt>", which makes it much easier to read.
242</li>
243
244<li>LLVM 2.4 includes some changes for better vector support. First, the shift
245operations (<tt>shl</tt>, <tt>ashr</tt>, <tt>lshr</tt>) now all support vectors
246and do an element-by-element shift (shifts of the whole vector can be
247accomplished by bitcasting the vector to &lt;1 x i128&gt; for example). Second,
Chris Lattnerc2d84672008-10-14 00:52:49 +0000[diff] [blame^]248there is initial support in development for vector comparisons with the
249<a href="LangRef.html#i_fcmp">fcmp</a>/<a href="LangRef.html#i_icmp">icmp</a>
250instructions. These instructions compare two vectors and return a vector of
251i1's for each result. Note that there is very little codegen support available
252for any of these IR features though.</li>
Chris Lattnerf7bae6d2008-10-13 20:59:32 +0000[diff] [blame]253
254<li>A new <tt>DebugInfoBuilder</tt> class is available, which makes it much
255easier for front-ends to create debug info descriptors, similar to the way that
256IRBuilder makes it easier to create LLVM IR.</li>
257
258<li>LLVM now supports "function attributes", which allows us to seperate return
259value attributes from function attributes. LLVM now supports attributes on a
260function itself, a return value, and its parameters. New supported function
261attributes include noinline/alwaysinline and the "opt-size" flag which says the
262function should be optimized for code size.</li>
263
Chris Lattnera5a49382008-10-12 18:30:33 +0000[diff] [blame]264<li>LLVM IR now directly represents "common" linkage, instead of
265 representing it as a form of weak linkage.</li>
Chris Lattnerf7bae6d2008-10-13 20:59:32 +0000[diff] [blame]266
Chris Lattnerc3649452008-02-10 08:17:19 +0000[diff] [blame]267</ul>
Mikhail Glushenkov808cd3e2008-10-13 02:08:34 +0000[diff] [blame]268
Chris Lattnerc3649452008-02-10 08:17:19 +0000[diff] [blame]269</div>
270
271<!--=========================================================================-->
272<div class="doc_subsection">
Chris Lattner8fda4952008-02-10 07:46:44 +0000[diff] [blame]273<a name="optimizer">Optimizer Improvements</a>
274</div>
275
276<div class="doc_text">
277
Chris Lattnerea9c2242008-10-13 21:50:36 +0000[diff] [blame]278<p>In addition to a huge array of bug fixes and minor performance tweaks, this
279release includes a few major enhancements and additions to the optimizers:</p>
Chris Lattner8fda4952008-02-10 07:46:44 +0000[diff] [blame]280
281<ul>
282
Chris Lattnerea9c2242008-10-13 21:50:36 +0000[diff] [blame]283<li>The Global Value Numbering (GVN) pass now does local Partial Redundancy
284Elimination (PRE) to eliminate some partially redundant expressions in cases
285where doing so won't grow code size.</li>
Chris Lattner4f076402008-10-12 19:47:48 +0000[diff] [blame]286
Chris Lattnerea9c2242008-10-13 21:50:36 +0000[diff] [blame]287<li>LLVM 2.4 includes a new loop deletion pass (which removes output-free
288provably-finite loops) and a rewritten Aggressive Dead Code Elimination (ADCE)
289pass that no longer uses control dependence information. These changes speed up
290the optimizer and also prevents it from deleting output-free infinite
291loops.</li>
Chris Lattner4f076402008-10-12 19:47:48 +0000[diff] [blame]292
Duncan Sands2aecc3f2008-10-13 18:42:43 +0000[diff] [blame]293<li>The new AddReadAttrs pass works out which functions are read-only or
294read-none (these correspond to 'pure' and 'const' in C) and marks them
295with the appropriate attribute.</li>
Chris Lattnerea9c2242008-10-13 21:50:36 +0000[diff] [blame]296
297<li>LLVM 2.4 now includes a new SparsePropagation framework, which makes it
298trivial to build lattice-based dataflow solvers that operate over LLVM IR. Using
299this interface means that you just define objects to represent your lattice
300values and the transfer functions that operate on them. It handles the
301mechanics of worklist processing, liveness tracking, handling PHI nodes,
302etc.</li>
303
304<li>Various helper functions (ComputeMaskedBits, ComputeNumSignBits, etc) were
305pulled out of the Instruction Combining pass and put into a new
306<tt>ValueTracking.h</tt> header, where they can be reused by other passes.</li>
307
308<li>The tail duplication pass has been removed from the standard optimizer
309sequence used by llvm-gcc. This pass still exists, but the benefits it once
310provided are now achieved by other passes.</li>
311
Chris Lattner0eef6e42008-06-08 02:45:07 +0000[diff] [blame]312</ul>
313
314</div>
315
316<!--=========================================================================-->
317<div class="doc_subsection">
318<a name="codegen">Code Generator Improvements</a>
319</div>
320
321<div class="doc_text">
322
323<p>We put a significant amount of work into the code generator infrastructure,
324which allows us to implement more aggressive algorithms and make it run
325faster:</p>
326
327<ul>
Chris Lattnerea9c2242008-10-13 21:50:36 +0000[diff] [blame]328<li>The target-independent code generator supports (and the X86 backend
329 currently implements) a new interface for "fast" instruction selection. This
330 interface is optimized to produce code as quickly as possible, sacrificing
331 code quality to do it. This is used by default at -O0 or when using
332 "llc -fast" on X86. It is straight-forward to add support for
333 other targets if faster -O0 compilation is desired.</li>
334
335<li>In addition to the new 'fast' instruction selection path, many existing
336 pieces of the code generator have been optimized in significant ways.
337 SelectionDAG's are now pool allocated and use better algorithms in many
338 places, the ".s" file printers now use raw_ostream to emit text much faster,
339 etc. The end result of these improvements is that the compiler also takes
340 substantially less time to generate code that is just as good (and often
341 better) than before.</li>
342
343<li>Each target has been split to separate the .s file printing logic from the
344 rest of the target. This enables JIT compilers that don't link in the
345 (somewhat large) code and data tables used for printing a .s file.</li>
346
347<li>The code generator now includes a "stack slot coloring" pass, which packs
348 together individual spilled values into common stack slots. This reduces
349 the size of stack frames with many spills, which tends to increase L1 cache
350 effectiveness.</li>
351
352<li>Various pieces of the register allocator (e.g. the coalescer and two-address
353 operation elimination pass) now know how to rematerialize trivial operations
354 to avoid copies and include several other optimizations.</li>
355
356<li>The <a href="CodeGenerator.html#selectiondag_process">graphs</a> produced by
357 the <tt>llc -view-*-dags</tt> options are now significantly prettier and
358 easier to read.</li>
359
360<li>LLVM 2.4 includes a new register allocator based on Partitioned Boolean
361 Quadratic Programming (PBQP). This register allocator is still in
362 development, but is very simple and clean.</li>
Chris Lattner0eef6e42008-06-08 02:45:07 +0000[diff] [blame]363
Chris Lattnerb0d6c502007-09-21 03:54:09 +0000[diff] [blame]364</ul>
365
366</div>
367
368
369<!--=========================================================================-->
370<div class="doc_subsection">
Chris Lattnercc64a322008-10-13 22:06:31 +0000[diff] [blame]371<a name="targetspecific">Target Specific Improvements</a>
Chris Lattnerb0d6c502007-09-21 03:54:09 +0000[diff] [blame]372</div>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]373
374<div class="doc_text">
Chris Lattnerc3649452008-02-10 08:17:19 +0000[diff] [blame]375<p>New target-specific features include:
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]376</p>
377
378<ul>
Chris Lattner4f076402008-10-12 19:47:48 +0000[diff] [blame]379<li>Exception handling is supported by default on Linux/x86-64.</li>
Duncan Sandsd8745a92008-10-13 17:32:15 +0000[diff] [blame]380<li>Position Independent Code (PIC) is now supported on Linux/x86-64.</li>
Chris Lattnercc64a322008-10-13 22:06:31 +0000[diff] [blame]381<li>MIPS floating point support? [BRUNO]</li>
382<li>The PowerPC backend now supports trampolines.</li>
Chris Lattnere45f18b2008-06-05 08:02:49 +0000[diff] [blame]383</ul>
Chris Lattnere45f18b2008-06-05 08:02:49 +0000[diff] [blame]384
Mikhail Glushenkov808cd3e2008-10-13 02:08:34 +0000[diff] [blame]385</div>
Chris Lattnere45f18b2008-06-05 08:02:49 +0000[diff] [blame]386
387
388<!--=========================================================================-->
389<div class="doc_subsection">
Chris Lattnerb0d6c502007-09-21 03:54:09 +0000[diff] [blame]390<a name="otherimprovements">Other Improvements</a>
391</div>
392
393<div class="doc_text">
394<p>New features include:
395</p>
396
397<ul>
Mikhail Glushenkovcc2d3012008-10-13 02:46:01 +0000[diff] [blame]398<li><tt>llvmc2</tt> (the generic compiler driver) gained plugin
Chris Lattnercc64a322008-10-13 22:06:31 +0000[diff] [blame]399 support. It is now easier to experiment with <tt>llvmc2</tt> and
400 build your own tools based on it.</li>
401
402<li>LLVM 2.4 includes a number of new generic algorithms and data structures,
403 include a scoped hash table, 'immutable' data structures, and a
404 <tt>raw_ostream</tt> class. The <tt>raw_ostream</tt> class and
405 <tt>format</tt> allow for efficient file output, and various pieces of LLVM
406 have switched over to use it. The eventual goal is to eliminate
407 std::ostream in favor of it.</li>
408
409<li>Recycler + pool allocation stuff? [DAN]</li>
Chris Lattnerb0d6c502007-09-21 03:54:09 +0000[diff] [blame]410</ul>
Mikhail Glushenkov808cd3e2008-10-13 02:08:34 +0000[diff] [blame]411
Chris Lattnerb0d6c502007-09-21 03:54:09 +0000[diff] [blame]412</div>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]413
Chris Lattner92eacf82008-10-13 17:57:36 +0000[diff] [blame]414<!--=========================================================================-->
415<div class="doc_subsection">
416<a name="changes">Major Changes and Removed Features</a>
417</div>
418
419<div class="doc_text">
420
Chris Lattnercc64a322008-10-13 22:06:31 +0000[diff] [blame]421<p>If you're already an LLVM user or developer with out-of-tree changes based
422on LLVM 2.3, this section lists some "gotchas" that you may run into upgrading
423from the previous release.</p>
Chris Lattner92eacf82008-10-13 17:57:36 +0000[diff] [blame]424
425<ul>
426
Chris Lattnercc64a322008-10-13 22:06:31 +0000[diff] [blame]427<li>The LLVM IR generated by llvm-gcc no longer names all instructions. This
428 makes it run faster, but may be more confusing to some people. If you
429 prefer to have names, the '<tt>opt -instnamer</tt>' pass will add names to
430 all instructions.</li>
431
432<li>The LoadVN and GCSE passes have been removed from the tree. They are
433 obsolete and have been replaced with the GVN and MemoryDependence passes.
434 </li>
435</ul>
436
437
438<p>In addition, many APIs have changed in this release. Some of the major LLVM
439API changes are:</p>
440
441<ul>
442
443<li>Attributes changes [DEVANG] </li>
Chris Lattner92eacf82008-10-13 17:57:36 +0000[diff] [blame]444
445<li>The <tt>DbgStopPointInst</tt> methods <tt>getDirectory</tt> and
446<tt>getFileName</tt> now return <tt>Value*</tt> instead of strings. These can be
447converted to strings using <tt>llvm::GetConstantStringInfo</tt> defined via
448"llvm/Analysis/ValueTracking.h".</li>
449
Chris Lattnercc64a322008-10-13 22:06:31 +0000[diff] [blame]450<li>The APIs to create various instructions have changed from lower case
451 "create" methods to upper case "Create" methods (e.g.
452 <tt>BinaryOperator::create</tt>). LLVM 2.4 includes both cases, but the
453 lower case ones are removed in mainline, please migrate.</li>
454
Chris Lattner92eacf82008-10-13 17:57:36 +0000[diff] [blame]455<li>Various header files like "llvm/ADT/iterator" were given a .h suffix.
456 Change your code to #include "llvm/ADT/iterator.h" instead.</li>
457
458</ul>
459
460</div>
461
462
463
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]464<!-- *********************************************************************** -->
465<div class="doc_section">
466 <a name="portability">Portability and Supported Platforms</a>
467</div>
468<!-- *********************************************************************** -->
469
470<div class="doc_text">
471
472<p>LLVM is known to work on the following platforms:</p>
473
474<ul>
Mikhail Glushenkov808cd3e2008-10-13 02:08:34 +0000[diff] [blame]475<li>Intel and AMD machines (IA32) running Red Hat Linux, Fedora Core and FreeBSD
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]476 (and probably other unix-like systems).</li>
Chris Lattnerbd633de92008-02-06 06:30:34 +0000[diff] [blame]477<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]478 64-bit modes.</li>
Chris Lattnere0798082008-02-06 18:00:06 +0000[diff] [blame]479<li>Intel and AMD machines running on Win32 using MinGW libraries (native).</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]480<li>Intel and AMD machines running on Win32 with the Cygwin libraries (limited
481 support is available for native builds with Visual C++).</li>
Gabor Greif5e75e652008-06-05 18:39:01 +0000[diff] [blame]482<li>Sun UltraSPARC workstations running Solaris 10.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]483<li>Alpha-based machines running Debian GNU/Linux.</li>
Gabor Greif5e75e652008-06-05 18:39:01 +0000[diff] [blame]484<li>Itanium-based (IA64) machines running Linux and HP-UX.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]485</ul>
486
Chris Lattnerba948522008-06-05 06:57:39 +0000[diff] [blame]487<p>The core LLVM infrastructure uses GNU autoconf to adapt itself
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]488to the machine and operating system on which it is built. However, minor
489porting may be required to get LLVM to work on new platforms. We welcome your
490portability patches and reports of successful builds or error messages.</p>
491
492</div>
493
494<!-- *********************************************************************** -->
495<div class="doc_section">
496 <a name="knownproblems">Known Problems</a>
497</div>
498<!-- *********************************************************************** -->
499
500<div class="doc_text">
501
502<p>This section contains all known problems with the LLVM system, listed by
503component. As new problems are discovered, they will be added to these
504sections. If you run into a problem, please check the <a
505href="http://llvm.org/bugs/">LLVM bug database</a> and submit a bug if
506there isn't already one.</p>
507
508</div>
509
510<!-- ======================================================================= -->
511<div class="doc_subsection">
512 <a name="experimental">Experimental features included with this release</a>
513</div>
514
515<div class="doc_text">
516
517<p>The following components of this LLVM release are either untested, known to
518be broken or unreliable, or are in early development. These components should
519not be relied on, and bugs should not be filed against them, but they may be
520useful 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]521components, please contact us on the <a
522href="http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev">LLVMdev list</a>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]523
524<ul>
Chris Lattnerdaeb3232008-02-12 06:29:45 +0000[diff] [blame]525<li>The MSIL, IA64, Alpha, SPU, and MIPS backends are experimental.</li>
Duncan Sandsa73561c2008-06-08 20:18:35 +0000[diff] [blame]526<li>The llc "<tt>-filetype=asm</tt>" (the default) is the only supported
Chris Lattnerbd633de92008-02-06 06:30:34 +0000[diff] [blame]527 value for this option.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]528</ul>
529
530</div>
531
532<!-- ======================================================================= -->
533<div class="doc_subsection">
534 <a name="x86-be">Known problems with the X86 back-end</a>
535</div>
536
537<div class="doc_text">
538
539<ul>
Anton Korobeynikov77d15e22008-06-08 10:24:13 +0000[diff] [blame]540 <li>The X86 backend does not yet support
541 all <a href="http://llvm.org/PR879">inline assembly that uses the X86
542 floating point stack</a>. It supports the 'f' and 't' constraints, but not
543 'u'.</li>
Chris Lattnerd95b5962008-06-08 21:19:07 +0000[diff] [blame]544 <li>The X86 backend generates inefficient floating point code when configured
545 to generate code for systems that don't have SSE2.</li>
Duncan Sands8b4f3772008-06-08 19:38:43 +0000[diff] [blame]546 <li>Win64 code generation wasn't widely tested. Everything should work, but we
Anton Korobeynikov77d15e22008-06-08 10:24:13 +0000[diff] [blame]547 expect small issues to happen. Also, llvm-gcc cannot build mingw64 runtime
548 currently due
549 to <a href="http://llvm.org/PR2255">several</a>
Chris Lattnerf06141a2008-06-08 23:12:47 +0000[diff] [blame]550 <a href="http://llvm.org/PR2257">bugs</a> due to lack of support for the
551 'u' inline assembly constraint and X87 floating point inline assembly.</li>
Dan Gohman3882ecb2008-06-08 23:05:11 +0000[diff] [blame]552 <li>The X86-64 backend does not yet support the LLVM IR instruction
553 <tt>va_arg</tt>. Currently, the llvm-gcc front-end supports variadic
554 argument constructs on X86-64 by lowering them manually.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]555</ul>
556
557</div>
558
559<!-- ======================================================================= -->
560<div class="doc_subsection">
561 <a name="ppc-be">Known problems with the PowerPC back-end</a>
562</div>
563
564<div class="doc_text">
565
566<ul>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]567<li>The Linux PPC32/ABI support needs testing for the interpreter and static
568compilation, and lacks support for debug information.</li>
569</ul>
570
571</div>
572
573<!-- ======================================================================= -->
574<div class="doc_subsection">
575 <a name="arm-be">Known problems with the ARM back-end</a>
576</div>
577
578<div class="doc_text">
579
580<ul>
581<li>Thumb mode works only on ARMv6 or higher processors. On sub-ARMv6
Duncan Sands1827fa62007-09-26 15:59:54 +0000[diff] [blame]582processors, thumb programs can crash or produce wrong
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]583results (<a href="http://llvm.org/PR1388">PR1388</a>).</li>
584<li>Compilation for ARM Linux OABI (old ABI) is supported, but not fully tested.
585</li>
Chris Lattnerd95b5962008-06-08 21:19:07 +0000[diff] [blame]586<li>There is a bug in QEMU-ARM (&lt;= 0.9.0) which causes it to incorrectly
587 execute
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]588programs compiled with LLVM. Please use more recent versions of QEMU.</li>
589</ul>
590
591</div>
592
593<!-- ======================================================================= -->
594<div class="doc_subsection">
595 <a name="sparc-be">Known problems with the SPARC back-end</a>
596</div>
597
598<div class="doc_text">
599
600<ul>
601<li>The SPARC backend only supports the 32-bit SPARC ABI (-m32), it does not
602 support the 64-bit SPARC ABI (-m64).</li>
603</ul>
604
605</div>
606
607<!-- ======================================================================= -->
608<div class="doc_subsection">
609 <a name="alpha-be">Known problems with the Alpha back-end</a>
610</div>
611
612<div class="doc_text">
613
614<ul>
615
616<li>On 21164s, some rare FP arithmetic sequences which may trap do not have the
617appropriate nops inserted to ensure restartability.</li>
618
619</ul>
620</div>
621
622<!-- ======================================================================= -->
623<div class="doc_subsection">
624 <a name="ia64-be">Known problems with the IA64 back-end</a>
625</div>
626
627<div class="doc_text">
628
629<ul>
Chris Lattnerf06141a2008-06-08 23:12:47 +0000[diff] [blame]630<li>The Itanium backend is highly experimental, and has a number of known
631 issues. We are looking for a maintainer for the Itanium backend. If you
632 are interested, please contact the llvmdev mailing list.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]633</ul>
634
635</div>
636
637<!-- ======================================================================= -->
638<div class="doc_subsection">
639 <a name="c-be">Known problems with the C back-end</a>
640</div>
641
642<div class="doc_text">
643
644<ul>
Chris Lattner499c07c2008-06-05 06:35:40 +0000[diff] [blame]645<li><a href="http://llvm.org/PR802">The C backend has only basic support for
646 inline assembly code</a>.</li>
Chris Lattnere0194462007-09-26 06:01:35 +0000[diff] [blame]647<li><a href="http://llvm.org/PR1658">The C backend violates the ABI of common
648 C++ programs</a>, preventing intermixing between C++ compiled by the CBE and
Duncan Sandsa73561c2008-06-08 20:18:35 +0000[diff] [blame]649 C++ code compiled with llc or native compilers.</li>
Duncan Sandsbd0ad6e2008-02-10 13:40:55 +0000[diff] [blame]650<li>The C backend does not support all exception handling constructs.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]651</ul>
652
653</div>
654
655
656<!-- ======================================================================= -->
657<div class="doc_subsection">
Chris Lattner499c07c2008-06-05 06:35:40 +0000[diff] [blame]658 <a name="c-fe">Known problems with the llvm-gcc C front-end</a>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]659</div>
660
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]661<div class="doc_text">
662
Chris Lattnere0798082008-02-06 18:00:06 +0000[diff] [blame]663<p>llvm-gcc does not currently support <a href="http://llvm.org/PR869">Link-Time
664Optimization</a> on most platforms "out-of-the-box". Please inquire on the
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]665llvmdev mailing list if you are interested.</p>
666
Chris Lattner499c07c2008-06-05 06:35:40 +0000[diff] [blame]667<p>The only major language feature of GCC not supported by llvm-gcc is
668 the <tt>__builtin_apply</tt> family of builtins. However, some extensions
669 are only supported on some targets. For example, trampolines are only
Duncan Sandsa73561c2008-06-08 20:18:35 +0000[diff] [blame]670 supported on some targets (these are used when you take the address of a
671 nested function).</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]672
Chris Lattner499c07c2008-06-05 06:35:40 +0000[diff] [blame]673<p>If you run into GCC extensions which are not supported, please let us know.
674</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]675
676</div>
677
678<!-- ======================================================================= -->
679<div class="doc_subsection">
Chris Lattner499c07c2008-06-05 06:35:40 +0000[diff] [blame]680 <a name="c++-fe">Known problems with the llvm-gcc C++ front-end</a>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]681</div>
682
683<div class="doc_text">
684
685<p>The C++ front-end is considered to be fully
686tested and works for a number of non-trivial programs, including LLVM
687itself, Qt, Mozilla, etc.</p>
688
689<ul>
Anton Korobeynikovd9a17932008-10-11 18:27:16 +0000[diff] [blame]690<li>Exception handling works well on the X86 and PowerPC targets. Currently
691 only linux and darwin targets are supported (both 32 and 64 bit).</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]692</ul>
693
694</div>
695
696
Chris Lattnerdaeb3232008-02-12 06:29:45 +0000[diff] [blame]697<!-- ======================================================================= -->
698<div class="doc_subsection">
Chris Lattner499c07c2008-06-05 06:35:40 +0000[diff] [blame]699 <a name="ada-fe">Known problems with the llvm-gcc Ada front-end</a>
Chris Lattnerdaeb3232008-02-12 06:29:45 +0000[diff] [blame]700</div>
701
702<div class="doc_text">
703The llvm-gcc 4.2 Ada compiler works fairly well, however this is not a mature
704technology and problems should be expected.
705<ul>
Duncan Sandsa73561c2008-06-08 20:18:35 +0000[diff] [blame]706<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]707to lack of trampoline support (pointers to nested functions) on other platforms,
Duncan Sandsa73561c2008-06-08 20:18:35 +0000[diff] [blame]708however 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]709which does support trampolines.</li>
710<li>The Ada front-end <a href="http://llvm.org/PR2007">fails to bootstrap</a>.
711Workaround: configure with --disable-bootstrap.</li>
Duncan Sands2dbc2e82008-10-13 17:27:23 +0000[diff] [blame]712<li>The c380004, <a href="http://llvm.org/PR2010">c393010</a>
713and <a href="http://llvm.org/PR2421">cxg2021</a> ACATS tests fail
714(c380004 also fails with gcc-4.2 mainline).</li>
715<li>Some gcc specific Ada tests continue to crash the compiler.</li>
Chris Lattnerdaeb3232008-02-12 06:29:45 +0000[diff] [blame]716<li>The -E binder option (exception backtraces)
717<a href="http://llvm.org/PR1982">does not work</a> and will result in programs
718crashing if an exception is raised. Workaround: do not use -E.</li>
719<li>Only discrete types <a href="http://llvm.org/PR1981">are allowed to start
720or finish at a non-byte offset</a> in a record. Workaround: do not pack records
721or use representation clauses that result in a field of a non-discrete type
722starting or finishing in the middle of a byte.</li>
Chris Lattnerd95b5962008-06-08 21:19:07 +0000[diff] [blame]723<li>The <tt>lli</tt> interpreter <a href="http://llvm.org/PR2009">considers
724'main' as generated by the Ada binder to be invalid</a>.
725Workaround: hand edit the file to use pointers for <tt>argv</tt> and
726<tt>envp</tt> rather than integers.</li>
727<li>The <tt>-fstack-check</tt> option <a href="http://llvm.org/PR2008">is
728ignored</a>.</li>
Chris Lattnerdaeb3232008-02-12 06:29:45 +0000[diff] [blame]729</ul>
730</div>
731
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]732<!-- *********************************************************************** -->
733<div class="doc_section">
734 <a name="additionalinfo">Additional Information</a>
735</div>
736<!-- *********************************************************************** -->
737
738<div class="doc_text">
739
740<p>A wide variety of additional information is available on the <a
741href="http://llvm.org">LLVM web page</a>, in particular in the <a
742href="http://llvm.org/docs/">documentation</a> section. The web page also
743contains versions of the API documentation which is up-to-date with the
744Subversion version of the source code.
745You can access versions of these documents specific to this release by going
746into the "<tt>llvm/doc/</tt>" directory in the LLVM tree.</p>
747
748<p>If you have any questions or comments about LLVM, please feel free to contact
749us via the <a href="http://llvm.org/docs/#maillist"> mailing
750lists</a>.</p>
751
752</div>
753
754<!-- *********************************************************************** -->
755
756<hr>
757<address>
758 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
759 src="http://jigsaw.w3.org/css-validator/images/vcss" alt="Valid CSS!"></a>
760 <a href="http://validator.w3.org/check/referer"><img
Chris Lattner5ba16862008-02-10 08:18:42 +0000[diff] [blame]761 src="http://www.w3.org/Icons/valid-html401" alt="Valid HTML 4.01!"></a>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]762
763 <a href="http://llvm.org/">LLVM Compiler Infrastructure</a><br>
764 Last modified: $Date$
765</address>
766
767</body>
768</html>