1 <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"
2 "http://www.w3.org/TR/html4/strict.dtd">
5 <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
6 <link rel="stylesheet" href="llvm.css" type="text/css">
7 <title>LLVM 2.6 Release Notes</title>
11 <div class="doc_title">LLVM 2.6 Release Notes</div>
14 <li><a href="#intro">Introduction</a></li>
15 <li><a href="#subproj">Sub-project Status Update</a></li>
16 <li><a href="#externalproj">External Projects Using LLVM 2.6</a></li>
17 <li><a href="#whatsnew">What's New in LLVM 2.6?</a></li>
18 <li><a href="GettingStarted.html">Installation Instructions</a></li>
19 <li><a href="#portability">Portability and Supported Platforms</a></li>
20 <li><a href="#knownproblems">Known Problems</a></li>
21 <li><a href="#additionalinfo">Additional Information</a></li>
24 <div class="doc_author">
25 <p>Written by the <a href="http://llvm.org">LLVM Team</a></p>
28 <!-- *********************************************************************** -->
29 <div class="doc_section">
30 <a name="intro">Introduction</a>
32 <!-- *********************************************************************** -->
34 <div class="doc_text">
36 <p>This document contains the release notes for the LLVM Compiler
37 Infrastructure, release 2.6. Here we describe the status of LLVM, including
38 major improvements from the previous release and significant known problems.
39 All LLVM releases may be downloaded from the <a
40 href="http://llvm.org/releases/">LLVM releases web site</a>.</p>
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
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>
48 <p>Note that if you are reading this file from a Subversion checkout or the
49 main LLVM web page, this document applies to the <i>next</i> release, not the
50 current one. To see the release notes for a specific release, please see the
51 <a href="http://llvm.org/releases/">releases page</a>.</p>
58 include/llvm/Analysis/LiveValues.h => Dan
59 lib/Transforms/IPO/MergeFunctions.cpp => consider for 2.8.
60 llvm/Analysis/PointerTracking.h => Edwin wants this, consider for 2.8.
64 <!-- Unfinished features in 2.6:
67 variable debug info for optimized code
68 postalloc scheduler: anti dependence breaking, hazard recognizer?
70 loop dependence analysis
71 ELF Writer? How stable?
72 <li>PostRA scheduler improvements, ARM adoption (David Goodwin).</li>
73 2.7 supports the GDB 7.0 jit interfaces for debug info.
76 <!-- for announcement email:
80 klee web page at klee.llvm.org
81 Many new papers added to /pubs/
86 <!-- *********************************************************************** -->
87 <div class="doc_section">
88 <a name="subproj">Sub-project Status Update</a>
90 <!-- *********************************************************************** -->
92 <div class="doc_text">
94 The LLVM 2.6 distribution currently consists of code from the core LLVM
95 repository —which roughly includes the LLVM optimizers, code generators
96 and supporting tools — and the llvm-gcc repository. In addition to this
97 code, the LLVM Project includes other sub-projects that are in development. The
98 two which are the most actively developed are the <a href="#clang">Clang
99 Project</a> and the <a href="#vmkit">VMKit Project</a>.
105 <!--=========================================================================-->
106 <div class="doc_subsection">
107 <a name="clang">Clang: C/C++/Objective-C Frontend Toolkit</a>
110 <div class="doc_text">
112 <p>The <a href="http://clang.llvm.org/">Clang project</a> is an effort to build
113 a set of new 'LLVM native' front-end technologies for the C family of languages.
114 LLVM 2.6 is the first release to officially include Clang, and it provides a
115 production quality C and Objective-C compiler. If you are interested in fast
116 compiles and good diagnostics, we encourage you to try it out.</p>
118 <p>In addition to supporting these languages, C++ support is also <a
119 href="http://clang.llvm.org/cxx_status.html">well under way</a>, and mainline
120 Clang is able to parse the libstdc++ 4.2 headers and even codegen simple apps.
121 If you are interested in Clang C++ support or any other Clang feature, we
122 strongly encourage you to get involved on the <a
123 href="http://lists.cs.uiuc.edu/mailman/listinfo/cfe-dev">Clang front-end mailing
126 <p>In the LLVM 2.6 time-frame, the Clang team has made many improvements:</p>
129 <li>C and Objective-C support are now considered production quality.</li>
130 <li>AuroraUX / FreeBSD & OpenBSD Toolchain support.</li>
131 <li>Most of Objective-C 2.0 is now supported with the GNU runtime.</li>
132 <li>Many many bugs are fixed and many features have been added.</li>
136 <!--=========================================================================-->
137 <div class="doc_subsection">
138 <a name="clangsa">Clang Static Analyzer</a>
141 <div class="doc_text">
143 <p><b>UPDATE!</b> Previously announced in the 2.4 and 2.5 LLVM releases, the Clang project also
144 includes an early stage static source code analysis tool for <a
145 href="http://clang.llvm.org/StaticAnalysis.html">automatically finding bugs</a>
146 in C and Objective-C programs. The tool performs a growing set of checks to find
147 bugs that occur on a specific path within a program.</p>
149 <p>In the LLVM 2.6 time-frame there have been many significant improvements to
152 <p>The set of checks performed by the static analyzer continues to expand, and
153 future plans for the tool include full source-level inter-procedural analysis
154 and deeper checks such as buffer overrun detection. There are many opportunities
155 to extend and enhance the static analyzer, and anyone interested in working on
156 this project is encouraged to get involved!</p>
160 <!--=========================================================================-->
161 <div class="doc_subsection">
162 <a name="vmkit">VMKit: JVM/CLI Virtual Machine Implementation</a>
165 <div class="doc_text">
168 The <a href="http://vmkit.llvm.org/">VMKit project</a> is an implementation of
169 a JVM and a CLI Virtual Machines (Microsoft .NET is an
170 implementation of the CLI) using the Just-In-Time compiler of LLVM.</p>
172 <p>Following LLVM 2.6, VMKit has its XYZ release that you can find on its
173 <a href="http://vmkit.llvm.org/releases/">webpage</a>. The release includes
174 bug fixes, cleanup and new features. The major changes are:</p>
178 <li>Something wonderful!</li>
184 <!--=========================================================================-->
185 <div class="doc_subsection">
186 <a name="compiler-rt">compiler-rt: Compiler Runtime Library</a>
189 <div class="doc_text">
191 The new LLVM <a href="http://compiler-rt.llvm.org/">compiler-rt project</a>
192 is a simple library that provides an implementation of the low-level
193 target-specific hooks required by code generation and other runtime components.
194 For example, when compiling for a 32-bit target, converting a double to a 64-bit
195 unsigned integer is compiling into a runtime call to the "__fixunsdfdi"
196 function. The compiler-rt library provides optimized implementations of this and
197 other low-level routines.</p>
200 All of the code in the compiler-rt project is available under the standard LLVM
201 License, a "BSD-style" license.</p>
205 <!--=========================================================================-->
206 <div class="doc_subsection">
207 <a name="klee">klee: Symbolic Execution and Automatic Test Case Generator</a>
210 <div class="doc_text">
212 The new LLVM <a href="http://klee.llvm.org/">klee project</a> is a symbolic
213 execution framework for programs in LLVM bitcode form. Klee tries to
214 symbolically evaluate "all" paths through the application and records state
215 transitions that lead to fault states. This allows it to construct testcases
216 that lead to faults and can even be used to verify algorithms. For more
217 details, please see the <a
218 href="http://llvm.org/pubs/2008-12-OSDI-KLEE.html">OSDI 2008 paper</a> about
223 <!--=========================================================================-->
224 <div class="doc_subsection">
225 <a name="dragonegg">Dragon Egg: An LLVM backend plugin for GCC</a>
228 <div class="doc_text">
230 <b>Duncan needs to write me</b>.
236 <!--=========================================================================-->
237 <div class="doc_subsection">
238 <a name="mc">llvm-mc: Machine Code Toolkit</a>
241 <div class="doc_text">
243 The LLVM Machine Code (MC) Toolkit project is a (very early) effort to build
244 better tools for dealing with machine code, object file formats, etc. The idea
245 is to be able to generate most of the target specific details of assemblers and
246 disassemblers from existing LLVM target .td files (with suitable enhancements),
247 and to build infrastructure for reading and writing common object file formats.
248 One of the first deliverables is to build a full assembler and integrate it into
249 the compiler, which is predicted to substantially reduce compile time in some
253 <p>In the LLVM 2.6 timeframe, the MC framework has grown to the point where it
254 can reliably parse and pretty print (with some encoding information) a
255 darwin/x86 .s file successfully, and has the very early phases of a Mach-O
256 assembler in progress. Beyond the MC framework itself, major refactoring of the
257 LLVM code generator has started. The idea is to make the code generator reason
258 about the code it is producing in a much more semantic way, rather than a
259 textual way. For example, the code generator now uses MCSection objects to
260 represent section assignments, instead of text strings that print to .section
263 <p>MC is an early and ongoing project that will hopefully continue to lead to
264 many improvements in the code generator and build infrastructure useful for many
271 <!-- *********************************************************************** -->
272 <div class="doc_section">
273 <a name="externalproj">External Projects Using LLVM 2.6</a>
275 <!-- *********************************************************************** -->
277 <!--=========================================================================-->
278 <div class="doc_subsection">
279 <a name="Rubinius">Rubinius</a>
282 <div class="doc_text">
283 <p><a href="http://github.com/evanphx/rubinius">Rubinius</a> is an environment
284 for running Ruby code which strives to write as much of the core class
285 implementation in Ruby as possible. Combined with a bytecode interpreting VM, it
286 uses LLVM to optimize and compile ruby code down to machine code. Techniques
287 such as type feedback, method inlining, and uncommon traps are all used to
288 remove dynamism from ruby execution and increase performance.</p>
290 <p>Since LLVM 2.5, Rubinius has made several major leaps forward, implementing
291 a counter based JIT, type feedback, and speculative method inlining.
296 <!--=========================================================================-->
297 <div class="doc_subsection">
298 <a name="macruby">MacRuby</a>
301 <div class="doc_text">
304 <a href="http://macruby.org">MacRuby</a> is an implementation of Ruby on top of
305 core Mac OS X technologies, such as the Objective-C common runtime and garbage
306 collector, and the CoreFoundation framework. It is principally developed by
307 Apple and aims at enabling the creation of full-fledged Mac OS X applications.
311 MacRuby uses LLVM for optimization passes, JIT and AOT compilation of Ruby
312 expressions. It also uses zero-cost DWARF exceptions to implement Ruby exception
318 <!--=========================================================================-->
319 <div class="doc_subsection">
320 <a name="pure">Pure</a>
323 <div class="doc_text">
325 <a href="http://pure-lang.googlecode.com/">Pure</a>
326 is an algebraic/functional programming language based on term rewriting.
327 Programs are collections of equations which are used to evaluate expressions in
328 a symbolic fashion. Pure offers dynamic typing, eager and lazy evaluation,
329 lexical closures, a hygienic macro system (also based on term rewriting),
330 built-in list and matrix support (including list and matrix comprehensions) and
331 an easy-to-use C interface. The interpreter uses LLVM as a backend to
332 JIT-compile Pure programs to fast native code.</p>
334 <p>Pure versions 0.31 and later have been tested and are known to work with
335 LLVM 2.6 (and continue to work with older LLVM releases >= 2.3 as well).
340 <!--=========================================================================-->
341 <div class="doc_subsection">
342 <a name="ldc">LLVM D Compiler</a>
345 <div class="doc_text">
347 <a href="http://www.dsource.org/projects/ldc">LDC</a> is an implementation of
348 the D Programming Language using the LLVM optimizer and code generator.
349 The LDC project works great with the LLVM 2.6 release. General improvements in
351 cycle have included new inline asm constraint handling, better debug info
352 support, general bugfixes, and better x86-64 support. This has allowed
353 some major improvements in LDC, getting us much closer to being as
354 fully featured as the original DMD compiler from DigitalMars.
358 <!--=========================================================================-->
359 <div class="doc_subsection">
360 <a name="RoadsendPHP">Roadsend PHP</a>
363 <div class="doc_text">
365 <a href="http://code.roadsend.com/rphp">Roadsend PHP</a> (rphp) is an open
366 source implementation of the PHP programming
367 language that uses LLVM for its optimizer, JIT, and static compiler. This is a
368 reimplementation of an earlier project that is now based on LLVM.</p>
371 <!--=========================================================================-->
372 <div class="doc_subsection">
373 <a name="UnladenSwallow">Unladen Swallow</a>
376 <div class="doc_text">
378 <a href="http://code.google.com/p/unladen-swallow/">Unladen Swallow</a> is a
379 branch of <a href="http://python.org/">Python</a> intended to be fully
380 compatible and significantly faster. It uses LLVM's optimization passes and JIT
384 <!--=========================================================================-->
385 <div class="doc_subsection">
386 <a name="llvm-lua">llvm-lua</a>
389 <div class="doc_text">
391 <a href="http://code.google.com/p/llvm-lua/">LLVM-Lua</a> uses LLVM to add JIT
392 & static compiling support to the Lua VM. Lua bytecode is analyzed to
393 remove type checks, then LLVM is used to compile those bytecodes down to machine
399 <!-- *********************************************************************** -->
400 <div class="doc_section">
401 <a name="whatsnew">What's New in LLVM 2.6?</a>
403 <!-- *********************************************************************** -->
405 <div class="doc_text">
407 <p>This release includes a huge number of bug fixes, performance tweaks, and
408 minor improvements. Some of the major improvements and new features are listed
414 <!--=========================================================================-->
415 <div class="doc_subsection">
416 <a name="majorfeatures">Major New Features</a>
419 <div class="doc_text">
421 <p>LLVM 2.6 includes several major new capabilities:</p>
424 <li>New <a href="#compiler-rt">compiler-rt</a>, <A href="#klee">klee</a>,
425 and <a href="#mc">machine code toolkit</a> sub-projects.</li>
426 <li>Debug information now includes line numbers when optimizations are enabled.
427 This allows statistical sampling tools like oprofile and Shark to map
428 samples back to source lines.</li>
429 <li>LLVM now includes new experimental backends to support the MSP430, SystemZ,
430 and BlackFin architectures.</li>
431 <li>LLVM supports a new <a href="GoldPlugin.html">Gold Linker Plugin</a> which
432 enables support for <a href="LinkTimeOptimization.html">transparent
433 link-time optimization</a> on ELF targets when used with the Gold binutils
435 <li>LLVM now supports doing optimization and code generation on multiple threads
436 by allowing multiple "LLVMContext" objects to exist. Please see the <a
437 href="ProgrammersManual.html#threading">threading entry in the Programmer's
438 Manual</a> for more information.</li>
443 <!--=========================================================================-->
444 <div class="doc_subsection">
445 <a name="coreimprovements">LLVM IR and Core Improvements</a>
448 <div class="doc_text">
449 <p>LLVM IR has several new features that are used by our existing front-ends and
450 can be useful if you are writing a front-end for LLVM:</p>
453 <li>Getelementpr instruction now allows any integer type for array/pointer indexes.</li>
454 <li>Inbounds for GEP</li>
455 <li>NSW/NUW/exact div</li>
456 <li>LSR promotes int induction variables to 64-bit on 64-bit targets, major perf boost for numerical code.</li>
457 <li>LSR now analyzes pointer expressions (e.g. getelementptrs), not just integers.</li>
458 <li>new linkage types linkonce_odr, weak_odr, linker_private, and available_externally.</li>
459 <li>New fadd, fsub, fmul instructions and classes. </li>
460 <li>Target intrinsics can now return multiple results.</li>
465 <!--=========================================================================-->
466 <div class="doc_subsection">
467 <a name="optimizer">Optimizer Improvements</a>
470 <div class="doc_text">
472 <p>In addition to a large array of bug fixes and minor performance tweaks, this
473 release includes a few major enhancements and additions to the optimizers:</p>
477 <li>SRoA improvements for vector unions, memset, arbitrary weird bitfield accesses etc. It now produces "strange" sized integers.</li>
478 <li>Inliner reuse stack space when inlining arrays?</li>
479 <li>Enabled GVN Load PRE.</li>
480 <li>New Static Single Information (SSI) construction pass (not used by anything yet, experimental).</li>
487 <!--=========================================================================-->
488 <div class="doc_subsection">
489 <a name="codegen">Target Independent Code Generator Improvements</a>
492 <div class="doc_text">
494 <p>We have put a significant amount of work into the code generator
495 infrastructure, which allows us to implement more aggressive algorithms and make
500 <li> -asm-verbose now prints location info (with -g) and loop nest info.</li>
501 <li>Tblgen now supports multiclass inheritance and a number of new string and
502 list operations like !(subst), !(foreach), !car, !cdr, !null, !if, !cast.
503 These make the .td files more expressive and allow more aggressive factoring
504 of duplication across instruction patterns.</li>
505 <li>New MachineVerifier pass.</li>
506 <li>Machine LICM, hoists things like constant pool loads, loads from readonly stubs, vector constant synthesization code, etc.</li>
507 <li>Machine Sinking</li>
508 <li>target-specific intrinsics (r63765)</li>
509 <li>Regalloc improvements for commuting, various spiller peephole optimizations, cross-class coalescing.</li>
510 <li><tt>llc -enable-value-prop</tt>, propagation of value info (sign/zero ext info) from one MBB to another</li>
511 <li>Regalloc hints for allocation stuff: Evan r73381/r73671. Finished/enabled?</li>
512 <li>Stack slot coloring for register spills (denser stack frames)</li>
513 <li>SelectionDAGS: New BuildVectorSDNode (r65296), and ISD::VECTOR_SHUFFLE (r69952 / PR2957)</li>
514 <li>Experimental support for shrink wrapping support in PEI.</li>
515 <li>Experimental support for writing ELF .o files directly from the compiler,
516 it works well for many simple C testcases, but doesn't support exception
517 handling, debug info, inline assembly, etc.</li>
518 <li>Targets can now specify register allocation hints through
519 MachineRegisterInfo:: setRegAllocationHint. A regalloc hint consists 1) hint
520 type, 2) physical register number. A hint type of zero specifies a register
521 allocation preference. Other hint type values are target specific which are
522 resolved by TargetRegisterInfo::ResolveRegAllocHint. An example of which is
523 the ARM target can uses register hint to request that the register allocator
524 provide an even / odd register pair to two virtual registers. It is
525 important to note the register allocation hints are just hints. There is no
526 guarantee the register allocators will be able to satisfy the hints.</li>
531 <!--=========================================================================-->
532 <div class="doc_subsection">
533 <a name="x86">X86-32 and X86-64 Target Improvements</a>
536 <div class="doc_text">
537 <p>New features of the X86 target include:
542 <li>Preliminary support for addrspace 256 -> GS, 257 -> FS, known problems: CodeGenerator.html#x86_memory</li>
543 <li>Support for softfloat modes, typically used by OS kernels.</li>
545 <li>X86-64: better modeling of implicit zero extensions, eliminates a lot of redundant zexts</li>
546 <li>X86-64 TLS support for local exec and initial exec.</li>
547 <li>Better modeling of H registerts as subregs.</li>
548 <li>Vector icmp/fcmp now work with SSE codegen.</li>
549 <li>SSE 4.2 support.</li>
550 <li>all global variable reference logic is now in ClassifyGlobalReference.</li>
556 <!--=========================================================================-->
557 <div class="doc_subsection">
558 <a name="pic16">PIC16 Target Improvements</a>
561 <div class="doc_text">
562 <p>New features of the PIC16 target include:
566 <li>Support for floating-point, indirect function calls, and
567 passing/returning aggregate types to functions.
568 <li>The code generator is able to generate debug info into output COFF files.
569 <li>Support for placing an object into a specific section or at a specific
570 address in memory.</li>
573 <p>Things not yet supported:</p>
576 <li>Variable arguments.</li>
577 <li>Interrupts/programs.</li>
582 <!--=========================================================================-->
583 <div class="doc_subsection">
584 <a name="ARM">ARM Target Improvements</a>
587 <div class="doc_text">
588 <p>New features of the ARM target include:
593 <li>Preliminary support for processors, such as the Cortex-A8 and Cortex-A9,
594 that implement version v7-A of the ARM architecture. The ARM backend now
595 supports both the Thumb2 and Advanced SIMD (Neon) instruction sets. The
596 AAPCS-VFP "hard float" calling conventions are also supported with the
597 <tt>-float-abi=hard</tt> flag. These features are still somewhat experimental
598 and subject to change. The Neon intrinsics, in particular, may change in future
602 ARM AAPCS-VFP hard float ABI is supported.
603 ARM calling convention code is now tblgen generated instead of manual.
604 ARM: NEON support. neonfp for doing single precision fp with neon instead of VFP.
610 <!--=========================================================================-->
611 <div class="doc_subsection">
612 <a name="OtherTarget">Other Target Specific Improvements</a>
615 <div class="doc_text">
616 <p>New features of other targets include:
620 <li>Add support for the PowerPC 64-bit SVR4 ABI.</li>
621 <li>Mips now supports O32 Calling Convention.</li>
626 <!--=========================================================================-->
627 <div class="doc_subsection">
628 <a name="executionengine">Interpreter and JIT Improvements</a>
631 <div class="doc_text">
634 <li>The JIT now supports generating more than 16M of code.</li>
635 <li>When configured with --with-oprofile, the JIT can now inform oprofile about
636 JIT'd code, allowing oprofile to get line number and function name
637 information for JIT'd functions.</li>
638 <li>When "libffi" is available, the LLVM interpreter now uses it, which supports
639 calling almost arbitrary external (natively compiled) functions.</li>
640 <li>Clients of the JIT can now register a 'JITEventListener' object to receive
641 callbacks when the JIT emits or frees machine code. The OProfile support
642 uses this mechanism.</li>
648 <!--=========================================================================-->
649 <div class="doc_subsection">
650 <a name="newapis">New Useful APIs</a>
653 <div class="doc_text">
656 <li>New EngineBuilder class for creating JITs: r76276</li>
657 New PrettyStackTrace, crashes of llvm tools should give some indication of what the compiler was doing at the time of the crash (e.g. running a pass), and print out command line arguments.
658 StringRef class, Twine class.
659 New WeakVH and AssertingVH and CallbackVH classes.
660 New llvm/ADT/Triple class.
661 llvm_report_error() error handling API (llvm/Support/ErrorHandling.h)
662 New llvm/System/Atomic.h, llvm/System/RWMutex.h for portable atomic ops, rw locks.
663 New SourceMgr, SMLoc classes for simple parsers with caret diagnostics and #include support, (used by
664 tablegen, llvm-mc, the .ll parser, FileCheck, etc)
671 <!--=========================================================================-->
672 <div class="doc_subsection">
673 <a name="otherimprovements">Other Improvements and New Features</a>
676 <div class="doc_text">
677 <p>Other miscellaneous features include:</p>
680 <li>LLVM now includes a new internal '<a
681 href="http://llvm.org/cmds/FileCheck.html">FileCheck</a>' tool which allows
682 writing much more accurate regression tests that run faster. Please see the
683 <a href="TestingGuide.html#FileCheck">FileCheck section of the Testing
684 Guide</a> for more information.</li>
685 <li>LLVM profile information support has been significantly improved to produce
686 correct use counts, and has support for edge profiling with reduced runtime
687 overhead. Combined, the generated profile information is both more correct and
688 imposes about half as much overhead (2.6. from 12% to 6% overhead on SPEC
690 <li>Many extensions to the C APIs.</li>
691 <li>LLVM 2.6 includes a brand new experimental LLVM bindings to the Ada2005
692 programming language.</li>
696 * Dynamic plugins now work on Windows.
697 * New option property: init. Makes possible to provide default values for
698 options defined in plugins (interface to cl::init).
699 * New example: Skeleton, shows how to create a standalone LLVMC-based driver.
700 * New example: mcc16, a driver for the PIC16 toolchain.</li>
707 <!--=========================================================================-->
708 <div class="doc_subsection">
709 <a name="changes">Major Changes and Removed Features</a>
712 <div class="doc_text">
714 <p>If you're already an LLVM user or developer with out-of-tree changes based
715 on LLVM 2.5, this section lists some "gotchas" that you may run into upgrading
716 from the previous release.</p>
720 <li>The Itanium (IA64) backend has been removed. It was not supported and
722 <li>The BigBlock register allocator has been removed, it also bitrotted.</li>
723 <li>The C Backend (-march=c) is no longer considered part of the LLVM release
724 criteria. We still want it to work, but no one is maintaining it and it lacks
725 support for arbitrary precision integers and other important IR features.</li>
728 LLVM build now builds all libraries as .a files instead of some
729 libraries as relinked .o files. This requires some APIs like
730 InitializeAllTargets.h. TargetRegistry!
735 <p>In addition, many APIs have changed in this release. Some of the major LLVM
740 no use of hash_set/hash_map, no more llvm::OStream
741 Use raw_ostream for everything, killed off llvm/Streams.h and DOUT
744 <li>LLVM's global uniquing tables for <tt>Type</tt>s and <tt>Constant</tt>s have
745 been privatized into members of an <tt>LLVMContext</tt>. A number of APIs
746 now take an <tt>LLVMContext</tt> as a parameter. To smooth the transition
747 for clients that will only ever use a single context, the new
748 <tt>getGlobalContext()</tt> API can be used to access a default global
749 context which can be passed in any and all cases where a context is
751 <li>The <tt>getABITypeSize</tt> methods are now called <tt>getAllocSize</tt>.</li>
752 <li>The <tt>Add</tt>, <tt>Sub</tt>, and <tt>Mul</tt> operators are no longer
753 overloaded for floating-point types. Floating-point addition, subtraction,
754 and multiplication are now represented with new operators <tt>FAdd</tt>,
755 <tt>FSub</tt>, and <tt>FMul</tt>. In the <tt>IRBuilder</tt> API,
756 <tt>CreateAdd</tt>, <tt>CreateSub</tt>, <tt>CreateMul</tt>, and
757 <tt>CreateNeg</tt> should only be used for integer arithmetic now;
758 <tt>CreateFAdd</tt>, <tt>CreateFSub</tt>, <tt>CreateFMul</tt>, and
759 <tt>CreateFNeg</tt> should now be used for floating-point arithmetic.</li>
760 <li>The DynamicLibrary class can no longer be constructed, its functionality has
761 moved to static member functions.</li>
762 <li><tt>raw_fd_ostream</tt>'s constructor for opening a given filename now
763 takes an extra <tt>Force</tt> argument. If <tt>Force</tt> is set to
764 <tt>false</tt>, an error will be reported if a file with the given name
765 already exists. If <tt>Force</tt> is set to <tt>true</tt>, the file will
766 be silently truncated (which is the behavior before this flag was
768 <li><tt>SCEVHandle</tt> no longer exists, because reference counting is no
769 longer done for <tt>SCEV*</tt> objects, instead <tt>const SCEV*</tt> should be
772 <li>Many APIs, notably <tt>llvm::Value</tt>, now use the <tt>StringRef</tt>
773 and <tt>Twine</tt> classes instead of passing <tt>const char*</tt>
774 or <tt>std::string</tt>, as described in
775 the <a href="ProgrammersManual.html#string_apis">Programmer's Manual</a>. Most
776 clients should be unaffected by this transition, unless they are used to <tt>Value::getName()</tt> returning a string. Here are some tips on updating to 2.6:
778 <li><tt>getNameStr()</tt> is still available, and matches the old
779 behavior. Replacing <tt>getName()</tt> calls with this is an safe option,
780 although more efficient alternatives are now possible.</li>
782 <li>If you were just relying on <tt>getName()</tt> being able to be sent to
783 a <tt>std::ostream</tt>, consider migrating
784 to <tt>llvm::raw_ostream</tt>.</li>
786 <li>If you were using <tt>getName().c_str()</tt> to get a <tt>const
787 char*</tt> pointer to the name, you can use <tt>getName().data()</tt>.
788 Note that this string (as before), may not be the entire name if the
789 name containts embedded null characters.</li>
791 <li>If you were using operator plus on the result of <tt>getName()</tt> and
792 treating the result as an <tt>std::string</tt>, you can either
793 uses <tt>Twine::str</tt> to get the result as an <tt>std::string</tt>, or
794 could move to a <tt>Twine</tt> based design.</li>
796 <li><tt>isName()</tt> should be replaced with comparison
797 against <tt>getName()</tt> (this is now efficient).
801 <li>The registration interfaces for backend Targets has changed (what was
802 previously TargetMachineRegistry). For backend authors, see the <a href="WritingAnLLVMBackend.html#TargetRegistration">Writing An LLVM Backend</a> guide. For clients, the notable API changes are:
804 <li><tt>TargetMachineRegistry</tt> has been renamed
805 to <tt>TargetRegistry</tt>.</li>
807 <li>Clients should move to using the <tt>TargetRegistry::lookupTarget()</tt>
808 function to find targets.</li>
812 <li>llvm-dis now fails if output file exists, instead of dumping to stdout.
813 FIXME: describe any other tool changes due to the raw_fd_ostream change. FIXME:
814 This is not an API change, maybe there should be a tool changes section?</li>
815 <li>temporarely due to Context API change passes should call doInitialization()
816 method of the pass they inherit from, otherwise Context is NULL.
817 FIXME: remove this entry when this is no longer needed.<li>
824 <!-- *********************************************************************** -->
825 <div class="doc_section">
826 <a name="portability">Portability and Supported Platforms</a>
828 <!-- *********************************************************************** -->
830 <div class="doc_text">
832 <p>LLVM is known to work on the following platforms:</p>
835 <li>Intel and AMD machines (IA32, X86-64, AMD64, EMT-64) running Red Hat
836 Linux, Fedora Core, FreeBSD and AuroraUX (and probably other unix-like systems).</li>
837 <li>PowerPC and X86-based Mac OS X systems, running 10.3 and above in 32-bit
838 and 64-bit modes.</li>
839 <li>Intel and AMD machines running on Win32 using MinGW libraries (native).</li>
840 <li>Intel and AMD machines running on Win32 with the Cygwin libraries (limited
841 support is available for native builds with Visual C++).</li>
842 <li>Sun UltraSPARC workstations running Solaris 10.</li>
843 <li>Alpha-based machines running Debian GNU/Linux.</li>
846 <p>The core LLVM infrastructure uses GNU autoconf to adapt itself
847 to the machine and operating system on which it is built. However, minor
848 porting may be required to get LLVM to work on new platforms. We welcome your
849 portability patches and reports of successful builds or error messages.</p>
853 <!-- *********************************************************************** -->
854 <div class="doc_section">
855 <a name="knownproblems">Known Problems</a>
857 <!-- *********************************************************************** -->
859 <div class="doc_text">
861 <p>This section contains significant known problems with the LLVM system,
862 listed by component. If you run into a problem, please check the <a
863 href="http://llvm.org/bugs/">LLVM bug database</a> and submit a bug if
864 there isn't already one.</p>
867 <li>LLVM will not correctly compile on Solaris and/or OpenSolaris
868 using the stock GCC 3.x.x series 'out the box',
869 See: <a href="#brokengcc">Broken versions of GCC and other tools</a>.
870 However, A <a href="http://pkg.auroraux.org/GCC">Modern GCC Build</a>
871 for x86/x64 has been made available from the third party AuroraUX Project
872 that has been meticulously tested for bootstrapping LLVM & Clang.</li>
877 <!-- ======================================================================= -->
878 <div class="doc_subsection">
879 <a name="experimental">Experimental features included with this release</a>
882 <div class="doc_text">
884 <p>The following components of this LLVM release are either untested, known to
885 be broken or unreliable, or are in early development. These components should
886 not be relied on, and bugs should not be filed against them, but they may be
887 useful to some people. In particular, if you would like to work on one of these
888 components, please contact us on the <a
889 href="http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev">LLVMdev list</a>.</p>
892 <li>The MSIL, Alpha, SPU, MIPS, and PIC16 backends are experimental.</li>
893 <li>The <tt>llc</tt> "<tt>-filetype=asm</tt>" (the default) is the only
894 supported value for this option.</li>
899 <!-- ======================================================================= -->
900 <div class="doc_subsection">
901 <a name="x86-be">Known problems with the X86 back-end</a>
904 <div class="doc_text">
907 <li>The X86 backend does not yet support
908 all <a href="http://llvm.org/PR879">inline assembly that uses the X86
909 floating point stack</a>. It supports the 'f' and 't' constraints, but not
911 <li>The X86 backend generates inefficient floating point code when configured
912 to generate code for systems that don't have SSE2.</li>
913 <li>Win64 code generation wasn't widely tested. Everything should work, but we
914 expect small issues to happen. Also, llvm-gcc cannot build the mingw64
915 runtime currently due
916 to <a href="http://llvm.org/PR2255">several</a>
917 <a href="http://llvm.org/PR2257">bugs</a> and due to lack of support for
919 'u' inline assembly constraint and for X87 floating point inline assembly.</li>
920 <li>The X86-64 backend does not yet support the LLVM IR instruction
921 <tt>va_arg</tt>. Currently, the llvm-gcc and front-ends support variadic
922 argument constructs on X86-64 by lowering them manually.</li>
927 <!-- ======================================================================= -->
928 <div class="doc_subsection">
929 <a name="ppc-be">Known problems with the PowerPC back-end</a>
932 <div class="doc_text">
935 <li>The Linux PPC32/ABI support needs testing for the interpreter and static
936 compilation, and lacks support for debug information.</li>
941 <!-- ======================================================================= -->
942 <div class="doc_subsection">
943 <a name="arm-be">Known problems with the ARM back-end</a>
946 <div class="doc_text">
949 <li>Support for the Advanced SIMD (Neon) instruction set is still incomplete
950 and not well tested. Some features may not work at all, and the code quality
951 may be poor in some cases.</li>
952 <li>Thumb mode works only on ARMv6 or higher processors. On sub-ARMv6
953 processors, thumb programs can crash or produce wrong
954 results (<a href="http://llvm.org/PR1388">PR1388</a>).</li>
955 <li>Compilation for ARM Linux OABI (old ABI) is supported but not fully tested.
961 <!-- ======================================================================= -->
962 <div class="doc_subsection">
963 <a name="sparc-be">Known problems with the SPARC back-end</a>
966 <div class="doc_text">
969 <li>The SPARC backend only supports the 32-bit SPARC ABI (-m32); it does not
970 support the 64-bit SPARC ABI (-m64).</li>
975 <!-- ======================================================================= -->
976 <div class="doc_subsection">
977 <a name="mips-be">Known problems with the MIPS back-end</a>
980 <div class="doc_text">
983 <li>The O32 ABI is not fully supported.</li>
984 <li>64-bit MIPS targets are not supported yet.</li>
989 <!-- ======================================================================= -->
990 <div class="doc_subsection">
991 <a name="alpha-be">Known problems with the Alpha back-end</a>
994 <div class="doc_text">
998 <li>On 21164s, some rare FP arithmetic sequences which may trap do not have the
999 appropriate nops inserted to ensure restartability.</li>
1004 <!-- ======================================================================= -->
1005 <div class="doc_subsection">
1006 <a name="c-be">Known problems with the C back-end</a>
1009 <div class="doc_text">
1012 <li><a href="http://llvm.org/PR802">The C backend has only basic support for
1013 inline assembly code</a>.</li>
1014 <li><a href="http://llvm.org/PR1658">The C backend violates the ABI of common
1015 C++ programs</a>, preventing intermixing between C++ compiled by the CBE and
1016 C++ code compiled with <tt>llc</tt> or native compilers.</li>
1017 <li>The C backend does not support all exception handling constructs.</li>
1018 <li>The C backend does not support arbitrary precision integers.</li>
1024 <!-- ======================================================================= -->
1025 <div class="doc_subsection">
1026 <a name="c-fe">Known problems with the llvm-gcc C front-end</a>
1029 <div class="doc_text">
1031 <p>llvm-gcc does not currently support <a href="http://llvm.org/PR869">Link-Time
1032 Optimization</a> on most platforms "out-of-the-box". Please inquire on the
1033 LLVMdev mailing list if you are interested.</p>
1035 <p>The only major language feature of GCC not supported by llvm-gcc is
1036 the <tt>__builtin_apply</tt> family of builtins. However, some extensions
1037 are only supported on some targets. For example, trampolines are only
1038 supported on some targets (these are used when you take the address of a
1039 nested function).</p>
1041 <p>If you run into GCC extensions which are not supported, please let us know.
1046 <!-- ======================================================================= -->
1047 <div class="doc_subsection">
1048 <a name="c++-fe">Known problems with the llvm-gcc C++ front-end</a>
1051 <div class="doc_text">
1053 <p>The C++ front-end is considered to be fully
1054 tested and works for a number of non-trivial programs, including LLVM
1055 itself, Qt, Mozilla, etc.</p>
1058 <li>Exception handling works well on the X86 and PowerPC targets. Currently
1059 only Linux and Darwin targets are supported (both 32 and 64 bit).</li>
1064 <!-- ======================================================================= -->
1065 <div class="doc_subsection">
1066 <a name="fortran-fe">Known problems with the llvm-gcc Fortran front-end</a>
1069 <div class="doc_text">
1071 <li>Fortran support generally works, but there are still several unresolved bugs
1072 in Bugzilla. Please see the tools/gfortran component for details.</li>
1076 <!-- ======================================================================= -->
1077 <div class="doc_subsection">
1078 <a name="ada-fe">Known problems with the llvm-gcc Ada front-end</a>
1081 <div class="doc_text">
1082 The llvm-gcc 4.2 Ada compiler works fairly well; however, this is not a mature
1083 technology, and problems should be expected.
1085 <li>The Ada front-end currently only builds on X86-32. This is mainly due
1086 to lack of trampoline support (pointers to nested functions) on other platforms.
1087 However, it <a href="http://llvm.org/PR2006">also fails to build on X86-64</a>
1088 which does support trampolines.</li>
1089 <li>The Ada front-end <a href="http://llvm.org/PR2007">fails to bootstrap</a>.
1090 This is due to lack of LLVM support for <tt>setjmp</tt>/<tt>longjmp</tt> style
1091 exception handling, which is used internally by the compiler.
1092 Workaround: configure with --disable-bootstrap.</li>
1093 <li>The c380004, <a href="http://llvm.org/PR2010">c393010</a>
1094 and <a href="http://llvm.org/PR2421">cxg2021</a> ACATS tests fail
1095 (c380004 also fails with gcc-4.2 mainline).
1096 If the compiler is built with checks disabled then <a href="http://llvm.org/PR2010">c393010</a>
1097 causes the compiler to go into an infinite loop, using up all system memory.</li>
1098 <li>Some GCC specific Ada tests continue to crash the compiler.</li>
1099 <li>The -E binder option (exception backtraces)
1100 <a href="http://llvm.org/PR1982">does not work</a> and will result in programs
1101 crashing if an exception is raised. Workaround: do not use -E.</li>
1102 <li>Only discrete types <a href="http://llvm.org/PR1981">are allowed to start
1103 or finish at a non-byte offset</a> in a record. Workaround: do not pack records
1104 or use representation clauses that result in a field of a non-discrete type
1105 starting or finishing in the middle of a byte.</li>
1106 <li>The <tt>lli</tt> interpreter <a href="http://llvm.org/PR2009">considers
1107 'main' as generated by the Ada binder to be invalid</a>.
1108 Workaround: hand edit the file to use pointers for <tt>argv</tt> and
1109 <tt>envp</tt> rather than integers.</li>
1110 <li>The <tt>-fstack-check</tt> option <a href="http://llvm.org/PR2008">is
1115 <!-- ======================================================================= -->
1116 <div class="doc_subsection">
1117 <a name="ocaml-bindingse">Known problems with the O'Caml bindings</a>
1120 <div class="doc_text">
1122 <p>The Llvm.Linkage module is broken, and has incorrect values. Only
1123 Llvm.Linkage.External, Llvm.Linkage.Available_externally, and
1124 Llvm.Linkage.Link_once will be correct. If you need any of the other linkage
1125 modes, you'll have to write an external C library in order to expose the
1126 functionality. This has been fixed in the trunk.</p>
1129 <!-- *********************************************************************** -->
1130 <div class="doc_section">
1131 <a name="additionalinfo">Additional Information</a>
1133 <!-- *********************************************************************** -->
1135 <div class="doc_text">
1137 <p>A wide variety of additional information is available on the <a
1138 href="http://llvm.org">LLVM web page</a>, in particular in the <a
1139 href="http://llvm.org/docs/">documentation</a> section. The web page also
1140 contains versions of the API documentation which is up-to-date with the
1141 Subversion version of the source code.
1142 You can access versions of these documents specific to this release by going
1143 into the "<tt>llvm/doc/</tt>" directory in the LLVM tree.</p>
1145 <p>If you have any questions or comments about LLVM, please feel free to contact
1146 us via the <a href="http://llvm.org/docs/#maillist"> mailing
1151 <!-- *********************************************************************** -->
1155 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
1156 src="http://jigsaw.w3.org/css-validator/images/vcss-blue" alt="Valid CSS"></a>
1157 <a href="http://validator.w3.org/check/referer"><img
1158 src="http://www.w3.org/Icons/valid-html401-blue" alt="Valid HTML 4.01"></a>
1160 <a href="http://llvm.org/">LLVM Compiler Infrastructure</a><br>
1161 Last modified: $Date$