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 1.7 Release Notes</title>
11 <div class="doc_title">LLVM 1.7 Release Notes</div>
14 <li><a href="#intro">Introduction</a></li>
15 <li><a href="#whatsnew">What's New?</a></li>
16 <li><a href="GettingStarted.html">Installation Instructions</a></li>
17 <li><a href="#portability">Portability and Supported Platforms</a></li>
18 <li><a href="#knownproblems">Known Problems</a>
19 <li><a href="#additionalinfo">Additional Information</a></li>
22 <div class="doc_author">
23 <p>Written by the <a href="http://llvm.org">LLVM Team</a><p>
26 <!-- *********************************************************************** -->
27 <div class="doc_section">
28 <a name="intro">Introduction</a>
30 <!-- *********************************************************************** -->
32 <div class="doc_text">
34 <p>This document contains the release notes for the LLVM compiler
35 infrastructure, release 1.7. Here we describe the status of LLVM, including any
36 known problems and major improvements from the previous release. The most
37 up-to-date version of this document can be found on the <a
38 href="http://llvm.org/releases/">LLVM releases web site</a>. If you are
39 not reading this on the LLVM web pages, you should probably go there because
40 this document may be updated after the release.</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 CVS or the main LLVM web page,
49 this document applies to the <i>next</i> release, not the current one. To see
50 the release notes for the current or previous releases, see the <a
51 href="http://llvm.org/releases/">releases page</a>.</p>
55 <!-- *********************************************************************** -->
56 <div class="doc_section">
57 <a name="whatsnew">What's New?</a>
59 <!-- *********************************************************************** -->
61 <div class="doc_text">
63 <p>This is the eighth public release of the LLVM Compiler Infrastructure. This
64 release incorporates a large number of enhancements and new features,
65 including vector support (Intel SSE and Altivec), a new GCC4.0-based
66 C/C++ front-end, Objective C/C++ support, inline assembly support, and many
72 <!--=========================================================================-->
73 <div class="doc_subsection">
74 <a name="newfeatures">New Features in LLVM 1.7</a>
77 <!--_________________________________________________________________________-->
78 <div class="doc_subsubsection"><a name="llvmgcc4">GCC4.0-based llvm-gcc
81 <div class="doc_text">
83 <p>LLVM 1.7 includes a brand new llvm-gcc, based on GCC 4.0.1. This version
84 of llvm-gcc solves many serious long-standing problems with llvm-gcc, including
85 all of those blocked by the <a href="http://llvm.org/PR498">llvm-gcc 4 meta
86 bug</a>. In addition, llvm-gcc4 implements support for many new features,
87 including GCC inline assembly, generic vector support, SSE and Altivec
88 intrinsics, and several new GCC attributes. Finally, llvm-gcc4 is
89 significantly faster than llvm-gcc3, respects -O options, its -c/-S options
90 correspond to GCC's (they emit native code), supports Objective C/C++, and
91 it has debugging support well underway.</p>
93 <p>If you can use it, llvm-gcc4 offers significant new functionality, and we
94 hope that it will replace llvm-gcc3 completely in a future release.
95 Unfortunately, it does not currently support C++ exception handling at all, and
96 it only works on Apple Mac OS/X machines with X86 or PowerPC processors.
101 <!--_________________________________________________________________________-->
102 <div class="doc_subsubsection"><a name="inlineasm">Inline Assembly
105 <div class="doc_text">
107 <p>The LLVM IR and llvm-gcc4 front-end now fully support arbitrary GCC <a
108 href="LangRef.html#inlineasm">inline assembly</a>. The LLVM X86 and PowerPC
109 code generators have initial support for it,
110 being able to compile basic statements, but are missing some features. Please
111 report any inline asm statements that crash the compiler or that are miscompiled
116 <!--_________________________________________________________________________-->
117 <div class="doc_subsubsection"><a name="newsparc">New SPARC backend</a></div>
119 <div class="doc_text">
121 <p>LLVM 1.7 includes a new, fully functional, SPARC backend built in the
122 target-independent code generator. This SPARC backend includes support for
123 SPARC V8 and SPARC V9 subtargets (controlling whether V9 features can be used),
124 and targets the 32-bit SPARC ABI.</p>
126 <p>The LLVM 1.7 release is the last release that will include the LLVM "SparcV9"
127 backend, which was the very first LLVM native code generator. It will
128 be removed in LLVM 1.8, being replaced with the new SPARC backend.</p>
132 <!--_________________________________________________________________________-->
133 <div class="doc_subsubsection"><a name="genvector">Generic Vector Support
136 <div class="doc_text">
138 <p>LLVM now includes significantly extended support for SIMD vectors in its
139 core instruction set. It now includes three new instructions for manipulating
140 vectors: <a href="LangRef.html#i_extractelement"><tt>extractelement</tt></a>,
141 <a href="LangRef.html#i_insertelement"><tt>insertelement</tt></a>, and
142 <a href="LangRef.html#i_shufflevector"><tt>shufflevector</tt></a>. Further,
143 many bugs in vector handling have been fixed, and vectors are now supported by
144 the target-independent code generator. For example, if a vector operation is
145 not supported by a particular target, it will be correctly broken down and
146 executed as scalar operations.</p>
148 <p>Because llvm-gcc3 does not support GCC generic vectors or vector intrinsics,
149 llvm-gcc4 must be used.</p>
153 <!--_________________________________________________________________________-->
154 <div class="doc_subsubsection"><a name="ssealtivec">Intel SSE and PowerPC
158 <div class="doc_text">
160 <p>The LLVM X86 backend now supports Intel SSE 1, 2, and 3, and now uses scalar
161 SSE operations to implement scalar floating point math when the target supports
162 SSE1 (for floats) or SSE2 (for doubles). Vector SSE instructions are generated
163 by llvm-gcc4 when the generic vector mechanism or specific SSE intrinsics are
167 <p>The LLVM PowerPC backend now supports the Altivec instruction set, including
168 both GCC -maltivec and -faltivec modes. Altivec instructions are generated
169 by llvm-gcc4 when the generic vector mechanism or specific Altivec intrinsics
175 <!--_________________________________________________________________________-->
176 <div class="doc_subsubsection"><a name="optimizernew">Optimizer
177 Improvements</a></div>
179 <div class="doc_text">
181 <li>The Loop Unswitching pass (<tt>-loop-unswitch</tt>) has had several bugs
182 fixed, has several new features, and is enabled by default in llvmgcc3
184 <li>The Loop Strength Reduction pass (<tt>-loop-reduce</tt>) is now enabled for
185 the X86 and Alpha backends.</li>
186 <li>The Instruction Combining pass (<tt>-instcombine</tt>) now includes a
187 framework and implementation for simplifying code based on whether computed
188 bits are demanded or not.</li>
189 <li>The Scalar Replacement of Aggregates pass (<tt>-scalarrepl</tt>) can now
190 promote simple unions to registers.</li>
191 <li>The Reassociation pass (<tt>-reassociate</tt>) can now
192 factor expressions, e.g. turning "A*A+A*B" into "A*(A+B)".</li>
193 <li>Several LLVM passes are <a href="http://llvm.org/PR681">significantly
198 <!--_________________________________________________________________________-->
199 <div class="doc_subsubsection"><a name="codgennew">Code Generator
200 Improvements</a></div>
202 <div class="doc_text">
204 <li>LLVM has a new prepass (before register allocation) list scheduler, which
205 supports bottom-up and top-down scheduling, pluggable priority functions and
206 pluggable hazard recognizers. The X86 backend uses this to reduce register
207 pressure and RISC targets schedule based on operation latency.</li>
208 <li>The tblgen-based target description framework introduced in LLVM 1.6 has
209 several new features, useful for targets that can fold loads and stores into
210 operations, and features that make the .td files more expressive.</li>
211 <li>The instruction selector is significantly faster in 1.7 than in 1.6.</li>
212 <li>The X86, Alpha and Itanium backends use new DAG-DAG instruction selectors,
213 making them easier to maintain and generate slightly better code.</li>
214 <li>The X86 backend now supports generation of Scalar SSE code for scalar FP
215 expressions. LLVM provides significantly better performance with Scalar SSE
216 instructions than it does with the Intel floating point stack
218 <li>The Itanium backend now has a bundling pass, which improves performance
219 by ~10% and reduces code size (previously it unconditionally inserted a stop
220 bit after every instruction).</li>
224 <!--_________________________________________________________________________-->
225 <div class="doc_subsubsection"><a name="othernew">Other New Features</a></div>
227 <div class="doc_text">
229 <li>The Mac OS/X PowerPC and X86 backends now have initial support for
231 debugging information, however, debug info generation has been disabled for
232 the 1.7 release in llvmgcc4.</li>
233 <li>LLVM includes the new <a href="docs/CommandGuide/html/llvm-config.html">
234 llvm-config</a> utility, which makes it easier to build and link programs
235 against the LLVM libraries when not using the LLVM makefiles.</li>
236 <li>LLVM now supports first class global ctor/dtor initialization lists, no
237 longer forcing targets to use "__main".</li>
238 <li>LLVM supports assigning globals and functions to a particular section
239 in the result executable using the GCC section attribute.</li>
240 <li><a href="ExtendingLLVM.html">Adding intrinsics to LLVM</a> is now
241 significantly easier.</li>
242 <li>llvmgcc4 now fully supports C99 Variable Length Arrays, including dynamic
243 stack deallocation.</li>
249 <!--=========================================================================-->
250 <div class="doc_subsection">
251 <a name="changes">Significant Changes in LLVM 1.7</a>
254 <div class="doc_text">
256 <li>The official LLVM URL is now <a href="http://llvm.org/">
257 http://llvm.org/</a>.</li>
258 <li>The LLVM intrinsics used to be overloaded based on type: for example,
259 <a href="LangRef.html#int_ctpop"><tt>llvm.ctpop</tt></a> could work with any
260 integer datatype. They are now separated into different intrinsics with
261 suffixes to denote their argument type (e.g. <tt>llvm.ctpop.i32</tt>)). Old
262 LLVM .ll and .bc files that use these intrinsics will continue to work with
263 new LLVM versions (they are transparently upgraded by the parsers), but will
264 cause a warning to be emitted.</li>
265 <li>The <tt>llvm.readport</tt>, <tt>llvm.writeport</tt>, <tt>llvm.readio</tt>,
266 and <tt>llvm.writeio</tt> intrinsics have been removed. The first two
267 were ever only supported by the X86 backend, the last two were never
268 correctly supported by any target, and none were accessible through the
269 C front-end. Inline assembly support can now be used to
270 implement these operations.</li>
271 <li>The <tt>llvm-db</tt> tool had basic support for stepping through code, which
272 used the JIT. This code has been removed, and DWARF emission support added
273 instead. <tt>llvm-db</tt> still exists in CVS if someone wanted to write a
274 <tt>ptrace</tt> backend for it.</li>
279 <!-- *********************************************************************** -->
280 <div class="doc_section">
281 <a name="portability">Portability and Supported Platforms</a>
283 <!-- *********************************************************************** -->
285 <div class="doc_text">
287 <p>LLVM is known to work on the following platforms:</p>
290 <li>Intel and AMD machines running Red Hat Linux, Fedora Core and FreeBSD
291 (and probably other unix-like systems).</li>
292 <li>Sun UltraSPARC workstations running Solaris 8.</li>
293 <li>Intel and AMD machines running on Win32 with the Cygwin libraries (limited
294 support is available for native builds with Visual C++).</li>
295 <li>PowerPC and X86-based Mac OS X systems, running 10.2 and above.</li>
296 <li>Alpha-based machines running Debian GNU/Linux.</li>
297 <li>Itanium-based machines running Linux and HP-UX.</li>
300 <p>The core LLVM infrastructure uses
301 <a href="http://www.gnu.org/software/autoconf/">GNU autoconf</a> to adapt itself
302 to the machine and operating system on which it is built. However, minor
303 porting may be required to get LLVM to work on new platforms. We welcome your
304 portability patches and reports of successful builds or error messages.</p>
308 <!-- *********************************************************************** -->
309 <div class="doc_section">
310 <a name="knownproblems">Known Problems</a>
312 <!-- *********************************************************************** -->
314 <div class="doc_text">
316 <p>This section contains all known problems with the LLVM system, listed by
317 component. As new problems are discovered, they will be added to these
318 sections. If you run into a problem, please check the <a
319 href="http://llvm.org/bugs/">LLVM bug database</a> and submit a bug if
320 there isn't already one.</p>
324 <!-- ======================================================================= -->
325 <div class="doc_subsection">
326 <a name="experimental">Experimental features included with this release</a>
329 <div class="doc_text">
331 <p>The following components of this LLVM release are either untested, known to
332 be broken or unreliable, or are in early development. These components should
333 not be relied on, and bugs should not be filed against them, but they may be
334 useful to some people. In particular, if you would like to work on one of these
335 components, please contact us on the llvmdev list.</p>
338 <li>The <tt>-cee</tt> pass is known to be buggy, and may be removed in in a
340 <li>The IA64 code generator is experimental.</li>
341 <li>The Alpha JIT is experimental.</li>
342 <li>"<tt>-filetype=asm</tt>" (the default) is the only supported value for the
343 <tt>-filetype</tt> llc option.</li>
348 <!-- ======================================================================= -->
349 <div class="doc_subsection">
350 <a name="build">Known problems with the Build System</a>
353 <div class="doc_text">
361 <!-- ======================================================================= -->
362 <div class="doc_subsection">
363 <a name="core">Known problems with the LLVM Core</a>
366 <div class="doc_text">
369 <li>In the JIT, <tt>dlsym()</tt> on a symbol compiled by the JIT will not
374 <!-- ======================================================================= -->
375 <div class="doc_subsection">
376 <a name="c-fe">Known problems with the C front-end</a>
379 <!-- _______________________________________________________________________ -->
380 <div class="doc_subsubsection">Bugs</div>
382 <div class="doc_text">
385 llvm-gcc3 has many significant problems that are fixed by llvm-gcc4. See
386 those blocked on the <a href="http://llvm.org/PR498">llvm-gcc4 meta bug</a>.
387 Two major ones include:</p>
391 C99 variable sized arrays do not release stack memory when they go out of
392 scope. Thus, the following program may run out of stack space:
394 for (i = 0; i != 1000000; ++i) {
400 <li>With llvm-gcc3, Initialization of global union variables can only be done <a
401 href="http://llvm.org/PR162">with the largest union member</a>.</li>
406 <!-- _______________________________________________________________________ -->
407 <div class="doc_subsubsection">
411 <div class="doc_text">
415 <li>"long double" is transformed by the front-end into "double". There is no
416 support for floating point data types of any size other than 32 and 64
419 <li>The following Unix system functionality has not been tested and may not
422 <li><tt>sigsetjmp</tt>, <tt>siglongjmp</tt> - These are not turned into the
423 appropriate <tt>invoke</tt>/<tt>unwind</tt> instructions. Note that
424 <tt>setjmp</tt> and <tt>longjmp</tt> <em>are</em> compiled correctly.
425 <li><tt>getcontext</tt>, <tt>setcontext</tt>, <tt>makecontext</tt>
426 - These functions have not been tested.
429 <li>Although many GCC extensions are supported, some are not. In particular,
430 the following extensions are known to <b>not be</b> supported:
432 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Local-Labels.html#Local%20Labels">Local Labels</a>: Labels local to a block.</li>
433 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Nested-Functions.html#Nested%20Functions">Nested Functions</a>: As in Algol and Pascal, lexical scoping of functions.</li>
434 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Constructing-Calls.html#Constructing%20Calls">Constructing Calls</a>: Dispatching a call to another function.</li>
435 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Extended-Asm.html#Extended%20Asm">Extended Asm</a>: Assembler instructions with C expressions as operands.</li>
436 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Constraints.html#Constraints">Constraints</a>: Constraints for asm operands.</li>
437 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Asm-Labels.html#Asm%20Labels">Asm Labels</a>: Specifying the assembler name to use for a C symbol.</li>
438 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Explicit-Reg-Vars.html#Explicit%20Reg%20Vars">Explicit Reg Vars</a>: Defining variables residing in specified registers.</li>
439 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Vector-Extensions.html#Vector%20Extensions">Vector Extensions</a>: Using vector instructions through built-in functions.</li>
440 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Target-Builtins.html#Target%20Builtins">Target Builtins</a>: Built-in functions specific to particular targets.</li>
441 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Thread_002dLocal.html">Thread-Local</a>: Per-thread variables.</li>
442 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Pragmas.html#Pragmas">Pragmas</a>: Pragmas accepted by GCC.</li>
445 <p>The following GCC extensions are <b>partially</b> supported. An ignored
446 attribute means that the LLVM compiler ignores the presence of the attribute,
447 but the code should still work. An unsupported attribute is one which is
448 ignored by the LLVM compiler and will cause a different interpretation of
452 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Variable-Length.html#Variable%20Length">Variable Length</a>:
453 Arrays whose length is computed at run time.<br>
454 Supported, but allocated stack space is not freed until the function returns (noted above).</li>
456 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Function-Attributes.html#Function%20Attributes">Function Attributes</a>:
458 Declaring that functions have no side effects or that they can never
461 <b>Supported:</b> <tt>format</tt>, <tt>format_arg</tt>, <tt>non_null</tt>,
462 <tt>noreturn</tt>, <tt>constructor</tt>, <tt>destructor</tt>,
463 <tt>unused</tt>, <tt>used</tt>,
464 <tt>deprecated</tt>, <tt>warn_unused_result</tt>, <tt>weak</tt><br>
466 <b>Ignored:</b> <tt>noinline</tt>,
467 <tt>always_inline</tt>, <tt>pure</tt>, <tt>const</tt>, <tt>nothrow</tt>,
468 <tt>malloc</tt>, <tt>no_instrument_function</tt>, <tt>cdecl</tt><br>
470 <b>Unsupported:</b> <tt>section</tt>, <tt>alias</tt>,
471 <tt>visibility</tt>, <tt>regparm</tt>, <tt>stdcall</tt>,
472 <tt>fastcall</tt>, all other target specific attributes</li>
474 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Variable-Attributes.html#Variable%20Attributes">Variable Attributes</a>:
475 Specifying attributes of variables.<br>
476 <b>Supported:</b> <tt>cleanup</tt>, <tt>common</tt>, <tt>nocommon</tt>,
477 <tt>deprecated</tt>, <tt>transparent_union</tt>,
478 <tt>unused</tt>, <tt>used</tt>, <tt>weak</tt><br>
480 <b>Unsupported:</b> <tt>aligned</tt>, <tt>mode</tt>, <tt>packed</tt>,
481 <tt>section</tt>, <tt>shared</tt>, <tt>tls_model</tt>,
482 <tt>vector_size</tt>, <tt>dllimport</tt>,
483 <tt>dllexport</tt>, all target specific attributes.</li>
485 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Type-Attributes.html#Type%20Attributes">Type Attributes</a>: Specifying attributes of types.<br>
486 <b>Supported:</b> <tt>transparent_union</tt>, <tt>unused</tt>,
487 <tt>deprecated</tt>, <tt>may_alias</tt><br>
489 <b>Unsupported:</b> <tt>aligned</tt>, <tt>packed</tt>,
490 all target specific attributes.</li>
492 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Other-Builtins.html#Other%20Builtins">Other Builtins</a>:
493 Other built-in functions.<br>
494 We support all builtins which have a C language equivalent (e.g.,
495 <tt>__builtin_cos</tt>), <tt>__builtin_alloca</tt>,
496 <tt>__builtin_types_compatible_p</tt>, <tt>__builtin_choose_expr</tt>,
497 <tt>__builtin_constant_p</tt>, and <tt>__builtin_expect</tt>
498 (currently ignored). We also support builtins for ISO C99 floating
499 point comparison macros (e.g., <tt>__builtin_islessequal</tt>),
500 <tt>__builtin_prefetch</tt>, <tt>__builtin_popcount[ll]</tt>,
501 <tt>__builtin_clz[ll]</tt>, and <tt>__builtin_ctz[ll]</tt>.</li>
504 <p>The following extensions <b>are</b> known to be supported:</p>
507 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Labels-as-Values.html#Labels%20as%20Values">Labels as Values</a>: Getting pointers to labels and computed gotos.</li>
508 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Statement-Exprs.html#Statement%20Exprs">Statement Exprs</a>: Putting statements and declarations inside expressions.</li>
509 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Typeof.html#Typeof">Typeof</a>: <code>typeof</code>: referring to the type of an expression.</li>
510 <li><a href="http://gcc.gnu.org/onlinedocs/gcc-3.4.0/gcc/Lvalues.html#Lvalues">Lvalues</a>: Using <code>?:</code>, "<code>,</code>" and casts in lvalues.</li>
511 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Conditionals.html#Conditionals">Conditionals</a>: Omitting the middle operand of a <code>?:</code> expression.</li>
512 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Long-Long.html#Long%20Long">Long Long</a>: Double-word integers.</li>
513 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Complex.html#Complex">Complex</a>: Data types for complex numbers.</li>
514 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Hex-Floats.html#Hex%20Floats">Hex Floats</a>:Hexadecimal floating-point constants.</li>
515 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html#Zero%20Length">Zero Length</a>: Zero-length arrays.</li>
516 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Empty-Structures.html#Empty%20Structures">Empty Structures</a>: Structures with no members.</li>
517 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Variadic-Macros.html#Variadic%20Macros">Variadic Macros</a>: Macros with a variable number of arguments.</li>
518 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Escaped-Newlines.html#Escaped%20Newlines">Escaped Newlines</a>: Slightly looser rules for escaped newlines.</li>
519 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Subscripting.html#Subscripting">Subscripting</a>: Any array can be subscripted, even if not an lvalue.</li>
520 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Pointer-Arith.html#Pointer%20Arith">Pointer Arith</a>: Arithmetic on <code>void</code>-pointers and function pointers.</li>
521 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Initializers.html#Initializers">Initializers</a>: Non-constant initializers.</li>
522 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Compound-Literals.html#Compound%20Literals">Compound Literals</a>: Compound literals give structures, unions,
523 or arrays as values.</li>
524 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Designated-Inits.html#Designated%20Inits">Designated Inits</a>: Labeling elements of initializers.</li>
525 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Cast-to-Union.html#Cast%20to%20Union">Cast to Union</a>: Casting to union type from any member of the union.</li>
526 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Case-Ranges.html#Case%20Ranges">Case Ranges</a>: `case 1 ... 9' and such.</li>
527 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Mixed-Declarations.html#Mixed%20Declarations">Mixed Declarations</a>: Mixing declarations and code.</li>
528 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Function-Prototypes.html#Function%20Prototypes">Function Prototypes</a>: Prototype declarations and old-style definitions.</li>
529 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/C_002b_002b-Comments.html#C_002b_002b-Comments">C++ Comments</a>: C++ comments are recognized.</li>
530 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Dollar-Signs.html#Dollar%20Signs">Dollar Signs</a>: Dollar sign is allowed in identifiers.</li>
531 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Character-Escapes.html#Character%20Escapes">Character Escapes</a>: <code>\e</code> stands for the character <ESC>.</li>
532 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Alignment.html#Alignment">Alignment</a>: Inquiring about the alignment of a type or variable.</li>
533 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Inline.html#Inline">Inline</a>: Defining inline functions (as fast as macros).</li>
534 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Alternate-Keywords.html#Alternate%20Keywords">Alternate Keywords</a>:<code>__const__</code>, <code>__asm__</code>, etc., for header files.</li>
535 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Incomplete-Enums.html#Incomplete%20Enums">Incomplete Enums</a>: <code>enum foo;</code>, with details to follow.</li>
536 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Function-Names.html#Function%20Names">Function Names</a>: Printable strings which are the name of the current function.</li>
537 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Return-Address.html#Return%20Address">Return Address</a>: Getting the return or frame address of a function.</li>
538 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Unnamed-Fields.html#Unnamed%20Fields">Unnamed Fields</a>: Unnamed struct/union fields within structs/unions.</li>
539 <li><a href="http://gcc.gnu.org/onlinedocs/gcc/Attribute-Syntax.html#Attribute%20Syntax">Attribute Syntax</a>: Formal syntax for attributes.</li>
544 <p>If you run into GCC extensions which have not been included in any of these
545 lists, please let us know (also including whether or not they work).</p>
549 <!-- ======================================================================= -->
550 <div class="doc_subsection">
551 <a name="c++-fe">Known problems with the C++ front-end</a>
554 <div class="doc_text">
556 <p>For this release, the C++ front-end is considered to be fully
557 tested and works for a number of non-trivial programs, including LLVM
562 <!-- _______________________________________________________________________ -->
563 <div class="doc_subsubsection">Bugs</div>
565 <div class="doc_text">
568 <li>The C++ front-end inherits all problems afflicting the <a href="#c-fe">C
575 <!-- _______________________________________________________________________ -->
576 <div class="doc_subsubsection">
580 <div class="doc_text">
584 <li>Destructors for local objects are not always run when a <tt>longjmp</tt> is
585 performed. In particular, destructors for objects in the <tt>longjmp</tt>ing
586 function and in the <tt>setjmp</tt> receiver function may not be run.
587 Objects in intervening stack frames will be destroyed, however (which is
588 better than most compilers).</li>
590 <li>The LLVM C++ front-end follows the <a
591 href="http://www.codesourcery.com/cxx-abi">Itanium C++ ABI</a>.
592 This document, which is not Itanium specific, specifies a standard for name
593 mangling, class layout, v-table layout, RTTI formats, and other C++
594 representation issues. Because we use this API, code generated by the LLVM
595 compilers should be binary compatible with machine code generated by other
596 Itanium ABI C++ compilers (such as G++, the Intel and HP compilers, etc).
597 <i>However</i>, the exception handling mechanism used by LLVM is very
598 different from the model used in the Itanium ABI, so <b>exceptions will not
599 interact correctly</b>. </li>
605 <!-- ======================================================================= -->
606 <div class="doc_subsection">
607 <a name="c-be">Known problems with the C back-end</a>
610 <div class="doc_text">
614 <li>The C back-end produces code that violates the ANSI C Type-Based Alias
615 Analysis rules. As such, special options may be necessary to compile the code
616 (for example, GCC requires the <tt>-fno-strict-aliasing</tt> option). This
617 problem probably cannot be fixed.</li>
619 <li><a href="http://llvm.org/PR56">Zero arg vararg functions are not
620 supported</a>. This should not affect LLVM produced by the C or C++
623 <li>The C backend does not correctly implement the <a
624 href="LangRef.html#i_stacksave"><tt>llvm.stacksave</tt></a> or
625 <a href="LangRef.html#i_stackrestore"><tt>llvm.stackrestore</tt></a>
626 intrinsics. This means that some code compiled by it can run out of stack
627 space if they depend on these (e.g. C99 varargs).</li>
633 <!-- ======================================================================= -->
634 <div class="doc_subsection">
635 <a name="x86-be">Known problems with the X86 back-end</a>
638 <div class="doc_text">
641 <li><a href="http://llvm.org/PR736">Indirect calls crash JIT on
647 <!-- ======================================================================= -->
648 <div class="doc_subsection">
649 <a name="ppc-be">Known problems with the PowerPC back-end</a>
652 <div class="doc_text">
655 <li><a href="http://llvm.org/PR642">PowerPC backend does not correctly
656 implement ordered FP comparisons</a>.</li>
661 <!-- ======================================================================= -->
662 <div class="doc_subsection">
663 <a name="alpha-be">Known problems with the Alpha back-end</a>
666 <div class="doc_text">
670 <li>On 21164s, some rare FP arithmetic sequences which may trap do not have the
671 appropriate nops inserted to ensure restartability.</li>
677 <!-- ======================================================================= -->
678 <div class="doc_subsection">
679 <a name="ia64-be">Known problems with the IA64 back-end</a>
682 <div class="doc_text">
686 <li>C++ programs are likely to fail on IA64, as calls to <tt>setjmp</tt> are
687 made where the argument is not 16-byte aligned, as required on IA64. (Strictly
688 speaking this is not a bug in the IA64 back-end; it will also be encountered
689 when building C++ programs using the C back-end.)</li>
691 <li>The C++ front-end does not use <a href="http://llvm.org/PR406">IA64
692 ABI compliant layout of v-tables</a>. In particular, it just stores function
693 pointers instead of function descriptors in the vtable. This bug prevents
694 mixing C++ code compiled with LLVM with C++ objects compiled by other C++
697 <li>There are a few ABI violations which will lead to problems when mixing LLVM
698 output with code built with other compilers, particularly for floating-point
701 <li>Defining vararg functions is not supported (but calling them is ok).</li>
707 <!-- ======================================================================= -->
708 <div class="doc_subsection">
709 <a name="sparc-be">Known problems with the SPARC back-end</a>
712 <div class="doc_text">
715 <li>The SPARC backend only supports the 32-bit SPARC ABI (-m32), it does not
716 support the 64-bit SPARC ABI (-m64).</li>
722 <!-- ======================================================================= -->
723 <div class="doc_subsection">
724 <a name="sparcv9-be">Known problems with the SparcV9 back-end</a>
727 <div class="doc_text">
730 <li><a href="http://llvm.org/PR60">[sparcv9] SparcV9 backend miscompiles
731 several programs in the LLVM test suite</a></li>
732 <li>The SparcV9 backend is slated to be removed before the LLVM 1.8
738 <!-- *********************************************************************** -->
739 <div class="doc_section">
740 <a name="additionalinfo">Additional Information</a>
742 <!-- *********************************************************************** -->
744 <div class="doc_text">
746 <p>A wide variety of additional information is available on the <a
747 href="http://llvm.org">LLVM web page</a>, including <a
748 href="http://llvm.org/docs/">documentation</a> and <a
749 href="http://llvm.org/pubs/">publications describing algorithms and
750 components implemented in LLVM</a>. The web page also contains versions of the
751 API documentation which is up-to-date with the CVS version of the source code.
752 You can access versions of these documents specific to this release by going
753 into the "<tt>llvm/doc/</tt>" directory in the LLVM tree.</p>
755 <p>If you have any questions or comments about LLVM, please feel free to contact
756 us via the <a href="http://llvm.org/docs/#maillist"> mailing
761 <!-- *********************************************************************** -->
765 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
766 src="http://jigsaw.w3.org/css-validator/images/vcss" alt="Valid CSS!"></a>
767 <a href="http://validator.w3.org/check/referer"><img
768 src="http://www.w3.org/Icons/valid-html401" alt="Valid HTML 4.01!" /></a>
770 <a href="http://llvm.org/">The LLVM Compiler Infrastructure</a><br>
771 Last modified: $Date$