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 <title>LLVM Programmer's Manual</title>
7 <link rel="stylesheet" href="llvm.css" type="text/css">
12 LLVM Programmer's Manual
16 <li><a href="#introduction">Introduction</a></li>
17 <li><a href="#general">General Information</a>
19 <li><a href="#stl">The C++ Standard Template Library</a></li>
21 <li>The <tt>-time-passes</tt> option</li>
22 <li>How to use the LLVM Makefile system</li>
23 <li>How to write a regression test</li>
28 <li><a href="#apis">Important and useful LLVM APIs</a>
30 <li><a href="#isa">The <tt>isa<></tt>, <tt>cast<></tt>
31 and <tt>dyn_cast<></tt> templates</a> </li>
32 <li><a href="#string_apis">Passing strings (the <tt>StringRef</tt>
33 and <tt>Twine</tt> classes)</a>
35 <li><a href="#StringRef">The <tt>StringRef</tt> class</a> </li>
36 <li><a href="#Twine">The <tt>Twine</tt> class</a> </li>
39 <li><a href="#DEBUG">The <tt>DEBUG()</tt> macro and <tt>-debug</tt>
42 <li><a href="#DEBUG_TYPE">Fine grained debug info with <tt>DEBUG_TYPE</tt>
43 and the <tt>-debug-only</tt> option</a> </li>
46 <li><a href="#Statistic">The <tt>Statistic</tt> class & <tt>-stats</tt>
49 <li>The <tt>InstVisitor</tt> template
50 <li>The general graph API
52 <li><a href="#ViewGraph">Viewing graphs while debugging code</a></li>
55 <li><a href="#datastructure">Picking the Right Data Structure for a Task</a>
57 <li><a href="#ds_sequential">Sequential Containers (std::vector, std::list, etc)</a>
59 <li><a href="#dss_arrayref">llvm/ADT/ArrayRef.h</a></li>
60 <li><a href="#dss_fixedarrays">Fixed Size Arrays</a></li>
61 <li><a href="#dss_heaparrays">Heap Allocated Arrays</a></li>
62 <li><a href="#dss_tinyptrvector">"llvm/ADT/TinyPtrVector.h"</a></li>
63 <li><a href="#dss_smallvector">"llvm/ADT/SmallVector.h"</a></li>
64 <li><a href="#dss_vector"><vector></a></li>
65 <li><a href="#dss_deque"><deque></a></li>
66 <li><a href="#dss_list"><list></a></li>
67 <li><a href="#dss_ilist">llvm/ADT/ilist.h</a></li>
68 <li><a href="#dss_packedvector">llvm/ADT/PackedVector.h</a></li>
69 <li><a href="#dss_other">Other Sequential Container Options</a></li>
71 <li><a href="#ds_set">Set-Like Containers (std::set, SmallSet, SetVector, etc)</a>
73 <li><a href="#dss_sortedvectorset">A sorted 'vector'</a></li>
74 <li><a href="#dss_smallset">"llvm/ADT/SmallSet.h"</a></li>
75 <li><a href="#dss_smallptrset">"llvm/ADT/SmallPtrSet.h"</a></li>
76 <li><a href="#dss_denseset">"llvm/ADT/DenseSet.h"</a></li>
77 <li><a href="#dss_FoldingSet">"llvm/ADT/FoldingSet.h"</a></li>
78 <li><a href="#dss_set"><set></a></li>
79 <li><a href="#dss_setvector">"llvm/ADT/SetVector.h"</a></li>
80 <li><a href="#dss_uniquevector">"llvm/ADT/UniqueVector.h"</a></li>
81 <li><a href="#dss_otherset">Other Set-Like ContainerOptions</a></li>
83 <li><a href="#ds_map">Map-Like Containers (std::map, DenseMap, etc)</a>
85 <li><a href="#dss_sortedvectormap">A sorted 'vector'</a></li>
86 <li><a href="#dss_stringmap">"llvm/ADT/StringMap.h"</a></li>
87 <li><a href="#dss_indexedmap">"llvm/ADT/IndexedMap.h"</a></li>
88 <li><a href="#dss_densemap">"llvm/ADT/DenseMap.h"</a></li>
89 <li><a href="#dss_valuemap">"llvm/ADT/ValueMap.h"</a></li>
90 <li><a href="#dss_intervalmap">"llvm/ADT/IntervalMap.h"</a></li>
91 <li><a href="#dss_map"><map></a></li>
92 <li><a href="#dss_inteqclasses">"llvm/ADT/IntEqClasses.h"</a></li>
93 <li><a href="#dss_othermap">Other Map-Like Container Options</a></li>
95 <li><a href="#ds_string">String-like containers</a>
99 <li><a href="#ds_bit">BitVector-like containers</a>
101 <li><a href="#dss_bitvector">A dense bitvector</a></li>
102 <li><a href="#dss_smallbitvector">A "small" dense bitvector</a></li>
103 <li><a href="#dss_sparsebitvector">A sparse bitvector</a></li>
107 <li><a href="#common">Helpful Hints for Common Operations</a>
109 <li><a href="#inspection">Basic Inspection and Traversal Routines</a>
111 <li><a href="#iterate_function">Iterating over the <tt>BasicBlock</tt>s
112 in a <tt>Function</tt></a> </li>
113 <li><a href="#iterate_basicblock">Iterating over the <tt>Instruction</tt>s
114 in a <tt>BasicBlock</tt></a> </li>
115 <li><a href="#iterate_institer">Iterating over the <tt>Instruction</tt>s
116 in a <tt>Function</tt></a> </li>
117 <li><a href="#iterate_convert">Turning an iterator into a
118 class pointer</a> </li>
119 <li><a href="#iterate_complex">Finding call sites: a more
120 complex example</a> </li>
121 <li><a href="#calls_and_invokes">Treating calls and invokes
122 the same way</a> </li>
123 <li><a href="#iterate_chains">Iterating over def-use &
124 use-def chains</a> </li>
125 <li><a href="#iterate_preds">Iterating over predecessors &
126 successors of blocks</a></li>
129 <li><a href="#simplechanges">Making simple changes</a>
131 <li><a href="#schanges_creating">Creating and inserting new
132 <tt>Instruction</tt>s</a> </li>
133 <li><a href="#schanges_deleting">Deleting <tt>Instruction</tt>s</a> </li>
134 <li><a href="#schanges_replacing">Replacing an <tt>Instruction</tt>
135 with another <tt>Value</tt></a> </li>
136 <li><a href="#schanges_deletingGV">Deleting <tt>GlobalVariable</tt>s</a> </li>
139 <li><a href="#create_types">How to Create Types</a></li>
141 <li>Working with the Control Flow Graph
143 <li>Accessing predecessors and successors of a <tt>BasicBlock</tt>
151 <li><a href="#threading">Threads and LLVM</a>
153 <li><a href="#startmultithreaded">Entering and Exiting Multithreaded Mode
155 <li><a href="#shutdown">Ending execution with <tt>llvm_shutdown()</tt></a></li>
156 <li><a href="#managedstatic">Lazy initialization with <tt>ManagedStatic</tt></a></li>
157 <li><a href="#llvmcontext">Achieving Isolation with <tt>LLVMContext</tt></a></li>
158 <li><a href="#jitthreading">Threads and the JIT</a></li>
162 <li><a href="#advanced">Advanced Topics</a>
165 <li><a href="#SymbolTable">The <tt>ValueSymbolTable</tt> class</a></li>
166 <li><a href="#UserLayout">The <tt>User</tt> and owned <tt>Use</tt> classes' memory layout</a></li>
169 <li><a href="#coreclasses">The Core LLVM Class Hierarchy Reference</a>
171 <li><a href="#Type">The <tt>Type</tt> class</a> </li>
172 <li><a href="#Module">The <tt>Module</tt> class</a></li>
173 <li><a href="#Value">The <tt>Value</tt> class</a>
175 <li><a href="#User">The <tt>User</tt> class</a>
177 <li><a href="#Instruction">The <tt>Instruction</tt> class</a></li>
178 <li><a href="#Constant">The <tt>Constant</tt> class</a>
180 <li><a href="#GlobalValue">The <tt>GlobalValue</tt> class</a>
182 <li><a href="#Function">The <tt>Function</tt> class</a></li>
183 <li><a href="#GlobalVariable">The <tt>GlobalVariable</tt> class</a></li>
190 <li><a href="#BasicBlock">The <tt>BasicBlock</tt> class</a></li>
191 <li><a href="#Argument">The <tt>Argument</tt> class</a></li>
198 <div class="doc_author">
199 <p>Written by <a href="mailto:sabre@nondot.org">Chris Lattner</a>,
200 <a href="mailto:dhurjati@cs.uiuc.edu">Dinakar Dhurjati</a>,
201 <a href="mailto:ggreif@gmail.com">Gabor Greif</a>,
202 <a href="mailto:jstanley@cs.uiuc.edu">Joel Stanley</a>,
203 <a href="mailto:rspencer@x10sys.com">Reid Spencer</a> and
204 <a href="mailto:owen@apple.com">Owen Anderson</a></p>
207 <!-- *********************************************************************** -->
209 <a name="introduction">Introduction </a>
211 <!-- *********************************************************************** -->
215 <p>This document is meant to highlight some of the important classes and
216 interfaces available in the LLVM source-base. This manual is not
217 intended to explain what LLVM is, how it works, and what LLVM code looks
218 like. It assumes that you know the basics of LLVM and are interested
219 in writing transformations or otherwise analyzing or manipulating the
222 <p>This document should get you oriented so that you can find your
223 way in the continuously growing source code that makes up the LLVM
224 infrastructure. Note that this manual is not intended to serve as a
225 replacement for reading the source code, so if you think there should be
226 a method in one of these classes to do something, but it's not listed,
227 check the source. Links to the <a href="/doxygen/">doxygen</a> sources
228 are provided to make this as easy as possible.</p>
230 <p>The first section of this document describes general information that is
231 useful to know when working in the LLVM infrastructure, and the second describes
232 the Core LLVM classes. In the future this manual will be extended with
233 information describing how to use extension libraries, such as dominator
234 information, CFG traversal routines, and useful utilities like the <tt><a
235 href="/doxygen/InstVisitor_8h-source.html">InstVisitor</a></tt> template.</p>
239 <!-- *********************************************************************** -->
241 <a name="general">General Information</a>
243 <!-- *********************************************************************** -->
247 <p>This section contains general information that is useful if you are working
248 in the LLVM source-base, but that isn't specific to any particular API.</p>
250 <!-- ======================================================================= -->
252 <a name="stl">The C++ Standard Template Library</a>
257 <p>LLVM makes heavy use of the C++ Standard Template Library (STL),
258 perhaps much more than you are used to, or have seen before. Because of
259 this, you might want to do a little background reading in the
260 techniques used and capabilities of the library. There are many good
261 pages that discuss the STL, and several books on the subject that you
262 can get, so it will not be discussed in this document.</p>
264 <p>Here are some useful links:</p>
268 <li><a href="http://www.dinkumware.com/manuals/#Standard C++ Library">Dinkumware
269 C++ Library reference</a> - an excellent reference for the STL and other parts
270 of the standard C++ library.</li>
272 <li><a href="http://www.tempest-sw.com/cpp/">C++ In a Nutshell</a> - This is an
273 O'Reilly book in the making. It has a decent Standard Library
274 Reference that rivals Dinkumware's, and is unfortunately no longer free since the
275 book has been published.</li>
277 <li><a href="http://www.parashift.com/c++-faq-lite/">C++ Frequently Asked
280 <li><a href="http://www.sgi.com/tech/stl/">SGI's STL Programmer's Guide</a> -
282 href="http://www.sgi.com/tech/stl/stl_introduction.html">Introduction to the
285 <li><a href="http://www.research.att.com/%7Ebs/C++.html">Bjarne Stroustrup's C++
288 <li><a href="http://64.78.49.204/">
289 Bruce Eckel's Thinking in C++, 2nd ed. Volume 2 Revision 4.0 (even better, get
294 <p>You are also encouraged to take a look at the <a
295 href="CodingStandards.html">LLVM Coding Standards</a> guide which focuses on how
296 to write maintainable code more than where to put your curly braces.</p>
300 <!-- ======================================================================= -->
302 <a name="stl">Other useful references</a>
308 <li><a href="http://www.fortran-2000.com/ArnaudRecipes/sharedlib.html">Using
309 static and shared libraries across platforms</a></li>
316 <!-- *********************************************************************** -->
318 <a name="apis">Important and useful LLVM APIs</a>
320 <!-- *********************************************************************** -->
324 <p>Here we highlight some LLVM APIs that are generally useful and good to
325 know about when writing transformations.</p>
327 <!-- ======================================================================= -->
329 <a name="isa">The <tt>isa<></tt>, <tt>cast<></tt> and
330 <tt>dyn_cast<></tt> templates</a>
335 <p>The LLVM source-base makes extensive use of a custom form of RTTI.
336 These templates have many similarities to the C++ <tt>dynamic_cast<></tt>
337 operator, but they don't have some drawbacks (primarily stemming from
338 the fact that <tt>dynamic_cast<></tt> only works on classes that
339 have a v-table). Because they are used so often, you must know what they
340 do and how they work. All of these templates are defined in the <a
341 href="/doxygen/Casting_8h-source.html"><tt>llvm/Support/Casting.h</tt></a>
342 file (note that you very rarely have to include this file directly).</p>
345 <dt><tt>isa<></tt>: </dt>
347 <dd><p>The <tt>isa<></tt> operator works exactly like the Java
348 "<tt>instanceof</tt>" operator. It returns true or false depending on whether
349 a reference or pointer points to an instance of the specified class. This can
350 be very useful for constraint checking of various sorts (example below).</p>
353 <dt><tt>cast<></tt>: </dt>
355 <dd><p>The <tt>cast<></tt> operator is a "checked cast" operation. It
356 converts a pointer or reference from a base class to a derived class, causing
357 an assertion failure if it is not really an instance of the right type. This
358 should be used in cases where you have some information that makes you believe
359 that something is of the right type. An example of the <tt>isa<></tt>
360 and <tt>cast<></tt> template is:</p>
362 <div class="doc_code">
364 static bool isLoopInvariant(const <a href="#Value">Value</a> *V, const Loop *L) {
365 if (isa<<a href="#Constant">Constant</a>>(V) || isa<<a href="#Argument">Argument</a>>(V) || isa<<a href="#GlobalValue">GlobalValue</a>>(V))
368 // <i>Otherwise, it must be an instruction...</i>
369 return !L->contains(cast<<a href="#Instruction">Instruction</a>>(V)->getParent());
374 <p>Note that you should <b>not</b> use an <tt>isa<></tt> test followed
375 by a <tt>cast<></tt>, for that use the <tt>dyn_cast<></tt>
380 <dt><tt>dyn_cast<></tt>:</dt>
382 <dd><p>The <tt>dyn_cast<></tt> operator is a "checking cast" operation.
383 It checks to see if the operand is of the specified type, and if so, returns a
384 pointer to it (this operator does not work with references). If the operand is
385 not of the correct type, a null pointer is returned. Thus, this works very
386 much like the <tt>dynamic_cast<></tt> operator in C++, and should be
387 used in the same circumstances. Typically, the <tt>dyn_cast<></tt>
388 operator is used in an <tt>if</tt> statement or some other flow control
389 statement like this:</p>
391 <div class="doc_code">
393 if (<a href="#AllocationInst">AllocationInst</a> *AI = dyn_cast<<a href="#AllocationInst">AllocationInst</a>>(Val)) {
399 <p>This form of the <tt>if</tt> statement effectively combines together a call
400 to <tt>isa<></tt> and a call to <tt>cast<></tt> into one
401 statement, which is very convenient.</p>
403 <p>Note that the <tt>dyn_cast<></tt> operator, like C++'s
404 <tt>dynamic_cast<></tt> or Java's <tt>instanceof</tt> operator, can be
405 abused. In particular, you should not use big chained <tt>if/then/else</tt>
406 blocks to check for lots of different variants of classes. If you find
407 yourself wanting to do this, it is much cleaner and more efficient to use the
408 <tt>InstVisitor</tt> class to dispatch over the instruction type directly.</p>
412 <dt><tt>cast_or_null<></tt>: </dt>
414 <dd><p>The <tt>cast_or_null<></tt> operator works just like the
415 <tt>cast<></tt> operator, except that it allows for a null pointer as an
416 argument (which it then propagates). This can sometimes be useful, allowing
417 you to combine several null checks into one.</p></dd>
419 <dt><tt>dyn_cast_or_null<></tt>: </dt>
421 <dd><p>The <tt>dyn_cast_or_null<></tt> operator works just like the
422 <tt>dyn_cast<></tt> operator, except that it allows for a null pointer
423 as an argument (which it then propagates). This can sometimes be useful,
424 allowing you to combine several null checks into one.</p></dd>
428 <p>These five templates can be used with any classes, whether they have a
429 v-table or not. To add support for these templates, you simply need to add
430 <tt>classof</tt> static methods to the class you are interested casting
431 to. Describing this is currently outside the scope of this document, but there
432 are lots of examples in the LLVM source base.</p>
437 <!-- ======================================================================= -->
439 <a name="string_apis">Passing strings (the <tt>StringRef</tt>
440 and <tt>Twine</tt> classes)</a>
445 <p>Although LLVM generally does not do much string manipulation, we do have
446 several important APIs which take strings. Two important examples are the
447 Value class -- which has names for instructions, functions, etc. -- and the
448 StringMap class which is used extensively in LLVM and Clang.</p>
450 <p>These are generic classes, and they need to be able to accept strings which
451 may have embedded null characters. Therefore, they cannot simply take
452 a <tt>const char *</tt>, and taking a <tt>const std::string&</tt> requires
453 clients to perform a heap allocation which is usually unnecessary. Instead,
454 many LLVM APIs use a <tt>StringRef</tt> or a <tt>const Twine&</tt> for
455 passing strings efficiently.</p>
457 <!-- _______________________________________________________________________ -->
459 <a name="StringRef">The <tt>StringRef</tt> class</a>
464 <p>The <tt>StringRef</tt> data type represents a reference to a constant string
465 (a character array and a length) and supports the common operations available
466 on <tt>std:string</tt>, but does not require heap allocation.</p>
468 <p>It can be implicitly constructed using a C style null-terminated string,
469 an <tt>std::string</tt>, or explicitly with a character pointer and length.
470 For example, the <tt>StringRef</tt> find function is declared as:</p>
472 <pre class="doc_code">
473 iterator find(StringRef Key);
476 <p>and clients can call it using any one of:</p>
478 <pre class="doc_code">
479 Map.find("foo"); <i>// Lookup "foo"</i>
480 Map.find(std::string("bar")); <i>// Lookup "bar"</i>
481 Map.find(StringRef("\0baz", 4)); <i>// Lookup "\0baz"</i>
484 <p>Similarly, APIs which need to return a string may return a <tt>StringRef</tt>
485 instance, which can be used directly or converted to an <tt>std::string</tt>
486 using the <tt>str</tt> member function. See
487 "<tt><a href="/doxygen/classllvm_1_1StringRef_8h-source.html">llvm/ADT/StringRef.h</a></tt>"
488 for more information.</p>
490 <p>You should rarely use the <tt>StringRef</tt> class directly, because it contains
491 pointers to external memory it is not generally safe to store an instance of the
492 class (unless you know that the external storage will not be freed). StringRef is
493 small and pervasive enough in LLVM that it should always be passed by value.</p>
497 <!-- _______________________________________________________________________ -->
499 <a name="Twine">The <tt>Twine</tt> class</a>
504 <p>The <tt>Twine</tt> class is an efficient way for APIs to accept concatenated
505 strings. For example, a common LLVM paradigm is to name one instruction based on
506 the name of another instruction with a suffix, for example:</p>
508 <div class="doc_code">
510 New = CmpInst::Create(<i>...</i>, SO->getName() + ".cmp");
514 <p>The <tt>Twine</tt> class is effectively a
515 lightweight <a href="http://en.wikipedia.org/wiki/Rope_(computer_science)">rope</a>
516 which points to temporary (stack allocated) objects. Twines can be implicitly
517 constructed as the result of the plus operator applied to strings (i.e., a C
518 strings, an <tt>std::string</tt>, or a <tt>StringRef</tt>). The twine delays the
519 actual concatenation of strings until it is actually required, at which point
520 it can be efficiently rendered directly into a character array. This avoids
521 unnecessary heap allocation involved in constructing the temporary results of
522 string concatenation. See
523 "<tt><a href="/doxygen/classllvm_1_1Twine_8h-source.html">llvm/ADT/Twine.h</a></tt>"
524 for more information.</p>
526 <p>As with a <tt>StringRef</tt>, <tt>Twine</tt> objects point to external memory
527 and should almost never be stored or mentioned directly. They are intended
528 solely for use when defining a function which should be able to efficiently
529 accept concatenated strings.</p>
535 <!-- ======================================================================= -->
537 <a name="DEBUG">The <tt>DEBUG()</tt> macro and <tt>-debug</tt> option</a>
542 <p>Often when working on your pass you will put a bunch of debugging printouts
543 and other code into your pass. After you get it working, you want to remove
544 it, but you may need it again in the future (to work out new bugs that you run
547 <p> Naturally, because of this, you don't want to delete the debug printouts,
548 but you don't want them to always be noisy. A standard compromise is to comment
549 them out, allowing you to enable them if you need them in the future.</p>
551 <p>The "<tt><a href="/doxygen/Debug_8h-source.html">llvm/Support/Debug.h</a></tt>"
552 file provides a macro named <tt>DEBUG()</tt> that is a much nicer solution to
553 this problem. Basically, you can put arbitrary code into the argument of the
554 <tt>DEBUG</tt> macro, and it is only executed if '<tt>opt</tt>' (or any other
555 tool) is run with the '<tt>-debug</tt>' command line argument:</p>
557 <div class="doc_code">
559 DEBUG(errs() << "I am here!\n");
563 <p>Then you can run your pass like this:</p>
565 <div class="doc_code">
567 $ opt < a.bc > /dev/null -mypass
568 <i><no output></i>
569 $ opt < a.bc > /dev/null -mypass -debug
574 <p>Using the <tt>DEBUG()</tt> macro instead of a home-brewed solution allows you
575 to not have to create "yet another" command line option for the debug output for
576 your pass. Note that <tt>DEBUG()</tt> macros are disabled for optimized builds,
577 so they do not cause a performance impact at all (for the same reason, they
578 should also not contain side-effects!).</p>
580 <p>One additional nice thing about the <tt>DEBUG()</tt> macro is that you can
581 enable or disable it directly in gdb. Just use "<tt>set DebugFlag=0</tt>" or
582 "<tt>set DebugFlag=1</tt>" from the gdb if the program is running. If the
583 program hasn't been started yet, you can always just run it with
586 <!-- _______________________________________________________________________ -->
588 <a name="DEBUG_TYPE">Fine grained debug info with <tt>DEBUG_TYPE</tt> and
589 the <tt>-debug-only</tt> option</a>
594 <p>Sometimes you may find yourself in a situation where enabling <tt>-debug</tt>
595 just turns on <b>too much</b> information (such as when working on the code
596 generator). If you want to enable debug information with more fine-grained
597 control, you define the <tt>DEBUG_TYPE</tt> macro and the <tt>-debug</tt> only
598 option as follows:</p>
600 <div class="doc_code">
603 DEBUG(errs() << "No debug type\n");
604 #define DEBUG_TYPE "foo"
605 DEBUG(errs() << "'foo' debug type\n");
607 #define DEBUG_TYPE "bar"
608 DEBUG(errs() << "'bar' debug type\n"));
610 #define DEBUG_TYPE ""
611 DEBUG(errs() << "No debug type (2)\n");
615 <p>Then you can run your pass like this:</p>
617 <div class="doc_code">
619 $ opt < a.bc > /dev/null -mypass
620 <i><no output></i>
621 $ opt < a.bc > /dev/null -mypass -debug
626 $ opt < a.bc > /dev/null -mypass -debug-only=foo
628 $ opt < a.bc > /dev/null -mypass -debug-only=bar
633 <p>Of course, in practice, you should only set <tt>DEBUG_TYPE</tt> at the top of
634 a file, to specify the debug type for the entire module (if you do this before
635 you <tt>#include "llvm/Support/Debug.h"</tt>, you don't have to insert the ugly
636 <tt>#undef</tt>'s). Also, you should use names more meaningful than "foo" and
637 "bar", because there is no system in place to ensure that names do not
638 conflict. If two different modules use the same string, they will all be turned
639 on when the name is specified. This allows, for example, all debug information
640 for instruction scheduling to be enabled with <tt>-debug-type=InstrSched</tt>,
641 even if the source lives in multiple files.</p>
643 <p>The <tt>DEBUG_WITH_TYPE</tt> macro is also available for situations where you
644 would like to set <tt>DEBUG_TYPE</tt>, but only for one specific <tt>DEBUG</tt>
645 statement. It takes an additional first parameter, which is the type to use. For
646 example, the preceding example could be written as:</p>
649 <div class="doc_code">
651 DEBUG_WITH_TYPE("", errs() << "No debug type\n");
652 DEBUG_WITH_TYPE("foo", errs() << "'foo' debug type\n");
653 DEBUG_WITH_TYPE("bar", errs() << "'bar' debug type\n"));
654 DEBUG_WITH_TYPE("", errs() << "No debug type (2)\n");
662 <!-- ======================================================================= -->
664 <a name="Statistic">The <tt>Statistic</tt> class & <tt>-stats</tt>
671 href="/doxygen/Statistic_8h-source.html">llvm/ADT/Statistic.h</a></tt>" file
672 provides a class named <tt>Statistic</tt> that is used as a unified way to
673 keep track of what the LLVM compiler is doing and how effective various
674 optimizations are. It is useful to see what optimizations are contributing to
675 making a particular program run faster.</p>
677 <p>Often you may run your pass on some big program, and you're interested to see
678 how many times it makes a certain transformation. Although you can do this with
679 hand inspection, or some ad-hoc method, this is a real pain and not very useful
680 for big programs. Using the <tt>Statistic</tt> class makes it very easy to
681 keep track of this information, and the calculated information is presented in a
682 uniform manner with the rest of the passes being executed.</p>
684 <p>There are many examples of <tt>Statistic</tt> uses, but the basics of using
685 it are as follows:</p>
688 <li><p>Define your statistic like this:</p>
690 <div class="doc_code">
692 #define <a href="#DEBUG_TYPE">DEBUG_TYPE</a> "mypassname" <i>// This goes before any #includes.</i>
693 STATISTIC(NumXForms, "The # of times I did stuff");
697 <p>The <tt>STATISTIC</tt> macro defines a static variable, whose name is
698 specified by the first argument. The pass name is taken from the DEBUG_TYPE
699 macro, and the description is taken from the second argument. The variable
700 defined ("NumXForms" in this case) acts like an unsigned integer.</p></li>
702 <li><p>Whenever you make a transformation, bump the counter:</p>
704 <div class="doc_code">
706 ++NumXForms; // <i>I did stuff!</i>
713 <p>That's all you have to do. To get '<tt>opt</tt>' to print out the
714 statistics gathered, use the '<tt>-stats</tt>' option:</p>
716 <div class="doc_code">
718 $ opt -stats -mypassname < program.bc > /dev/null
719 <i>... statistics output ...</i>
723 <p> When running <tt>opt</tt> on a C file from the SPEC benchmark
724 suite, it gives a report that looks like this:</p>
726 <div class="doc_code">
728 7646 bitcodewriter - Number of normal instructions
729 725 bitcodewriter - Number of oversized instructions
730 129996 bitcodewriter - Number of bitcode bytes written
731 2817 raise - Number of insts DCEd or constprop'd
732 3213 raise - Number of cast-of-self removed
733 5046 raise - Number of expression trees converted
734 75 raise - Number of other getelementptr's formed
735 138 raise - Number of load/store peepholes
736 42 deadtypeelim - Number of unused typenames removed from symtab
737 392 funcresolve - Number of varargs functions resolved
738 27 globaldce - Number of global variables removed
739 2 adce - Number of basic blocks removed
740 134 cee - Number of branches revectored
741 49 cee - Number of setcc instruction eliminated
742 532 gcse - Number of loads removed
743 2919 gcse - Number of instructions removed
744 86 indvars - Number of canonical indvars added
745 87 indvars - Number of aux indvars removed
746 25 instcombine - Number of dead inst eliminate
747 434 instcombine - Number of insts combined
748 248 licm - Number of load insts hoisted
749 1298 licm - Number of insts hoisted to a loop pre-header
750 3 licm - Number of insts hoisted to multiple loop preds (bad, no loop pre-header)
751 75 mem2reg - Number of alloca's promoted
752 1444 cfgsimplify - Number of blocks simplified
756 <p>Obviously, with so many optimizations, having a unified framework for this
757 stuff is very nice. Making your pass fit well into the framework makes it more
758 maintainable and useful.</p>
762 <!-- ======================================================================= -->
764 <a name="ViewGraph">Viewing graphs while debugging code</a>
769 <p>Several of the important data structures in LLVM are graphs: for example
770 CFGs made out of LLVM <a href="#BasicBlock">BasicBlock</a>s, CFGs made out of
771 LLVM <a href="CodeGenerator.html#machinebasicblock">MachineBasicBlock</a>s, and
772 <a href="CodeGenerator.html#selectiondag_intro">Instruction Selection
773 DAGs</a>. In many cases, while debugging various parts of the compiler, it is
774 nice to instantly visualize these graphs.</p>
776 <p>LLVM provides several callbacks that are available in a debug build to do
777 exactly that. If you call the <tt>Function::viewCFG()</tt> method, for example,
778 the current LLVM tool will pop up a window containing the CFG for the function
779 where each basic block is a node in the graph, and each node contains the
780 instructions in the block. Similarly, there also exists
781 <tt>Function::viewCFGOnly()</tt> (does not include the instructions), the
782 <tt>MachineFunction::viewCFG()</tt> and <tt>MachineFunction::viewCFGOnly()</tt>,
783 and the <tt>SelectionDAG::viewGraph()</tt> methods. Within GDB, for example,
784 you can usually use something like <tt>call DAG.viewGraph()</tt> to pop
785 up a window. Alternatively, you can sprinkle calls to these functions in your
786 code in places you want to debug.</p>
788 <p>Getting this to work requires a small amount of configuration. On Unix
789 systems with X11, install the <a href="http://www.graphviz.org">graphviz</a>
790 toolkit, and make sure 'dot' and 'gv' are in your path. If you are running on
791 Mac OS/X, download and install the Mac OS/X <a
792 href="http://www.pixelglow.com/graphviz/">Graphviz program</a>, and add
793 <tt>/Applications/Graphviz.app/Contents/MacOS/</tt> (or wherever you install
794 it) to your path. Once in your system and path are set up, rerun the LLVM
795 configure script and rebuild LLVM to enable this functionality.</p>
797 <p><tt>SelectionDAG</tt> has been extended to make it easier to locate
798 <i>interesting</i> nodes in large complex graphs. From gdb, if you
799 <tt>call DAG.setGraphColor(<i>node</i>, "<i>color</i>")</tt>, then the
800 next <tt>call DAG.viewGraph()</tt> would highlight the node in the
801 specified color (choices of colors can be found at <a
802 href="http://www.graphviz.org/doc/info/colors.html">colors</a>.) More
803 complex node attributes can be provided with <tt>call
804 DAG.setGraphAttrs(<i>node</i>, "<i>attributes</i>")</tt> (choices can be
805 found at <a href="http://www.graphviz.org/doc/info/attrs.html">Graph
806 Attributes</a>.) If you want to restart and clear all the current graph
807 attributes, then you can <tt>call DAG.clearGraphAttrs()</tt>. </p>
809 <p>Note that graph visualization features are compiled out of Release builds
810 to reduce file size. This means that you need a Debug+Asserts or
811 Release+Asserts build to use these features.</p>
817 <!-- *********************************************************************** -->
819 <a name="datastructure">Picking the Right Data Structure for a Task</a>
821 <!-- *********************************************************************** -->
825 <p>LLVM has a plethora of data structures in the <tt>llvm/ADT/</tt> directory,
826 and we commonly use STL data structures. This section describes the trade-offs
827 you should consider when you pick one.</p>
830 The first step is a choose your own adventure: do you want a sequential
831 container, a set-like container, or a map-like container? The most important
832 thing when choosing a container is the algorithmic properties of how you plan to
833 access the container. Based on that, you should use:</p>
836 <li>a <a href="#ds_map">map-like</a> container if you need efficient look-up
837 of an value based on another value. Map-like containers also support
838 efficient queries for containment (whether a key is in the map). Map-like
839 containers generally do not support efficient reverse mapping (values to
840 keys). If you need that, use two maps. Some map-like containers also
841 support efficient iteration through the keys in sorted order. Map-like
842 containers are the most expensive sort, only use them if you need one of
843 these capabilities.</li>
845 <li>a <a href="#ds_set">set-like</a> container if you need to put a bunch of
846 stuff into a container that automatically eliminates duplicates. Some
847 set-like containers support efficient iteration through the elements in
848 sorted order. Set-like containers are more expensive than sequential
852 <li>a <a href="#ds_sequential">sequential</a> container provides
853 the most efficient way to add elements and keeps track of the order they are
854 added to the collection. They permit duplicates and support efficient
855 iteration, but do not support efficient look-up based on a key.
858 <li>a <a href="#ds_string">string</a> container is a specialized sequential
859 container or reference structure that is used for character or byte
862 <li>a <a href="#ds_bit">bit</a> container provides an efficient way to store and
863 perform set operations on sets of numeric id's, while automatically
864 eliminating duplicates. Bit containers require a maximum of 1 bit for each
865 identifier you want to store.
870 Once the proper category of container is determined, you can fine tune the
871 memory use, constant factors, and cache behaviors of access by intelligently
872 picking a member of the category. Note that constant factors and cache behavior
873 can be a big deal. If you have a vector that usually only contains a few
874 elements (but could contain many), for example, it's much better to use
875 <a href="#dss_smallvector">SmallVector</a> than <a href="#dss_vector">vector</a>
876 . Doing so avoids (relatively) expensive malloc/free calls, which dwarf the
877 cost of adding the elements to the container. </p>
879 <!-- ======================================================================= -->
881 <a name="ds_sequential">Sequential Containers (std::vector, std::list, etc)</a>
885 There are a variety of sequential containers available for you, based on your
886 needs. Pick the first in this section that will do what you want.
888 <!-- _______________________________________________________________________ -->
890 <a name="dss_arrayref">llvm/ADT/ArrayRef.h</a>
894 <p>The llvm::ArrayRef class is the preferred class to use in an interface that
895 accepts a sequential list of elements in memory and just reads from them. By
896 taking an ArrayRef, the API can be passed a fixed size array, an std::vector,
897 an llvm::SmallVector and anything else that is contiguous in memory.
903 <!-- _______________________________________________________________________ -->
905 <a name="dss_fixedarrays">Fixed Size Arrays</a>
909 <p>Fixed size arrays are very simple and very fast. They are good if you know
910 exactly how many elements you have, or you have a (low) upper bound on how many
914 <!-- _______________________________________________________________________ -->
916 <a name="dss_heaparrays">Heap Allocated Arrays</a>
920 <p>Heap allocated arrays (new[] + delete[]) are also simple. They are good if
921 the number of elements is variable, if you know how many elements you will need
922 before the array is allocated, and if the array is usually large (if not,
923 consider a <a href="#dss_smallvector">SmallVector</a>). The cost of a heap
924 allocated array is the cost of the new/delete (aka malloc/free). Also note that
925 if you are allocating an array of a type with a constructor, the constructor and
926 destructors will be run for every element in the array (re-sizable vectors only
927 construct those elements actually used).</p>
930 <!-- _______________________________________________________________________ -->
932 <a name="dss_tinyptrvector">"llvm/ADT/TinyPtrVector.h"</a>
937 <p><tt>TinyPtrVector<Type></tt> is a highly specialized collection class
938 that is optimized to avoid allocation in the case when a vector has zero or one
939 elements. It has two major restrictions: 1) it can only hold values of pointer
940 type, and 2) it cannot hold a null pointer.</p>
942 <p>Since this container is highly specialized, it is rarely used.</p>
948 <!-- _______________________________________________________________________ -->
950 <a name="dss_smallvector">"llvm/ADT/SmallVector.h"</a>
954 <p><tt>SmallVector<Type, N></tt> is a simple class that looks and smells
955 just like <tt>vector<Type></tt>:
956 it supports efficient iteration, lays out elements in memory order (so you can
957 do pointer arithmetic between elements), supports efficient push_back/pop_back
958 operations, supports efficient random access to its elements, etc.</p>
960 <p>The advantage of SmallVector is that it allocates space for
961 some number of elements (N) <b>in the object itself</b>. Because of this, if
962 the SmallVector is dynamically smaller than N, no malloc is performed. This can
963 be a big win in cases where the malloc/free call is far more expensive than the
964 code that fiddles around with the elements.</p>
966 <p>This is good for vectors that are "usually small" (e.g. the number of
967 predecessors/successors of a block is usually less than 8). On the other hand,
968 this makes the size of the SmallVector itself large, so you don't want to
969 allocate lots of them (doing so will waste a lot of space). As such,
970 SmallVectors are most useful when on the stack.</p>
972 <p>SmallVector also provides a nice portable and efficient replacement for
977 <!-- _______________________________________________________________________ -->
979 <a name="dss_vector"><vector></a>
984 std::vector is well loved and respected. It is useful when SmallVector isn't:
985 when the size of the vector is often large (thus the small optimization will
986 rarely be a benefit) or if you will be allocating many instances of the vector
987 itself (which would waste space for elements that aren't in the container).
988 vector is also useful when interfacing with code that expects vectors :).
991 <p>One worthwhile note about std::vector: avoid code like this:</p>
993 <div class="doc_code">
996 std::vector<foo> V;
1002 <p>Instead, write this as:</p>
1004 <div class="doc_code">
1006 std::vector<foo> V;
1014 <p>Doing so will save (at least) one heap allocation and free per iteration of
1019 <!-- _______________________________________________________________________ -->
1021 <a name="dss_deque"><deque></a>
1025 <p>std::deque is, in some senses, a generalized version of std::vector. Like
1026 std::vector, it provides constant time random access and other similar
1027 properties, but it also provides efficient access to the front of the list. It
1028 does not guarantee continuity of elements within memory.</p>
1030 <p>In exchange for this extra flexibility, std::deque has significantly higher
1031 constant factor costs than std::vector. If possible, use std::vector or
1032 something cheaper.</p>
1035 <!-- _______________________________________________________________________ -->
1037 <a name="dss_list"><list></a>
1041 <p>std::list is an extremely inefficient class that is rarely useful.
1042 It performs a heap allocation for every element inserted into it, thus having an
1043 extremely high constant factor, particularly for small data types. std::list
1044 also only supports bidirectional iteration, not random access iteration.</p>
1046 <p>In exchange for this high cost, std::list supports efficient access to both
1047 ends of the list (like std::deque, but unlike std::vector or SmallVector). In
1048 addition, the iterator invalidation characteristics of std::list are stronger
1049 than that of a vector class: inserting or removing an element into the list does
1050 not invalidate iterator or pointers to other elements in the list.</p>
1053 <!-- _______________________________________________________________________ -->
1055 <a name="dss_ilist">llvm/ADT/ilist.h</a>
1059 <p><tt>ilist<T></tt> implements an 'intrusive' doubly-linked list. It is
1060 intrusive, because it requires the element to store and provide access to the
1061 prev/next pointers for the list.</p>
1063 <p><tt>ilist</tt> has the same drawbacks as <tt>std::list</tt>, and additionally
1064 requires an <tt>ilist_traits</tt> implementation for the element type, but it
1065 provides some novel characteristics. In particular, it can efficiently store
1066 polymorphic objects, the traits class is informed when an element is inserted or
1067 removed from the list, and <tt>ilist</tt>s are guaranteed to support a
1068 constant-time splice operation.</p>
1070 <p>These properties are exactly what we want for things like
1071 <tt>Instruction</tt>s and basic blocks, which is why these are implemented with
1072 <tt>ilist</tt>s.</p>
1074 Related classes of interest are explained in the following subsections:
1076 <li><a href="#dss_ilist_traits">ilist_traits</a></li>
1077 <li><a href="#dss_iplist">iplist</a></li>
1078 <li><a href="#dss_ilist_node">llvm/ADT/ilist_node.h</a></li>
1079 <li><a href="#dss_ilist_sentinel">Sentinels</a></li>
1083 <!-- _______________________________________________________________________ -->
1085 <a name="dss_packedvector">llvm/ADT/PackedVector.h</a>
1090 Useful for storing a vector of values using only a few number of bits for each
1091 value. Apart from the standard operations of a vector-like container, it can
1092 also perform an 'or' set operation.
1097 <div class="doc_code">
1101 FirstCondition = 0x1,
1102 SecondCondition = 0x2,
1107 PackedVector<State, 2> Vec1;
1108 Vec1.push_back(FirstCondition);
1110 PackedVector<State, 2> Vec2;
1111 Vec2.push_back(SecondCondition);
1114 return Vec1[0]; // returns 'Both'.
1121 <!-- _______________________________________________________________________ -->
1123 <a name="dss_ilist_traits">ilist_traits</a>
1127 <p><tt>ilist_traits<T></tt> is <tt>ilist<T></tt>'s customization
1128 mechanism. <tt>iplist<T></tt> (and consequently <tt>ilist<T></tt>)
1129 publicly derive from this traits class.</p>
1132 <!-- _______________________________________________________________________ -->
1134 <a name="dss_iplist">iplist</a>
1138 <p><tt>iplist<T></tt> is <tt>ilist<T></tt>'s base and as such
1139 supports a slightly narrower interface. Notably, inserters from
1140 <tt>T&</tt> are absent.</p>
1142 <p><tt>ilist_traits<T></tt> is a public base of this class and can be
1143 used for a wide variety of customizations.</p>
1146 <!-- _______________________________________________________________________ -->
1148 <a name="dss_ilist_node">llvm/ADT/ilist_node.h</a>
1152 <p><tt>ilist_node<T></tt> implements a the forward and backward links
1153 that are expected by the <tt>ilist<T></tt> (and analogous containers)
1154 in the default manner.</p>
1156 <p><tt>ilist_node<T></tt>s are meant to be embedded in the node type
1157 <tt>T</tt>, usually <tt>T</tt> publicly derives from
1158 <tt>ilist_node<T></tt>.</p>
1161 <!-- _______________________________________________________________________ -->
1163 <a name="dss_ilist_sentinel">Sentinels</a>
1167 <p><tt>ilist</tt>s have another specialty that must be considered. To be a good
1168 citizen in the C++ ecosystem, it needs to support the standard container
1169 operations, such as <tt>begin</tt> and <tt>end</tt> iterators, etc. Also, the
1170 <tt>operator--</tt> must work correctly on the <tt>end</tt> iterator in the
1171 case of non-empty <tt>ilist</tt>s.</p>
1173 <p>The only sensible solution to this problem is to allocate a so-called
1174 <i>sentinel</i> along with the intrusive list, which serves as the <tt>end</tt>
1175 iterator, providing the back-link to the last element. However conforming to the
1176 C++ convention it is illegal to <tt>operator++</tt> beyond the sentinel and it
1177 also must not be dereferenced.</p>
1179 <p>These constraints allow for some implementation freedom to the <tt>ilist</tt>
1180 how to allocate and store the sentinel. The corresponding policy is dictated
1181 by <tt>ilist_traits<T></tt>. By default a <tt>T</tt> gets heap-allocated
1182 whenever the need for a sentinel arises.</p>
1184 <p>While the default policy is sufficient in most cases, it may break down when
1185 <tt>T</tt> does not provide a default constructor. Also, in the case of many
1186 instances of <tt>ilist</tt>s, the memory overhead of the associated sentinels
1187 is wasted. To alleviate the situation with numerous and voluminous
1188 <tt>T</tt>-sentinels, sometimes a trick is employed, leading to <i>ghostly
1191 <p>Ghostly sentinels are obtained by specially-crafted <tt>ilist_traits<T></tt>
1192 which superpose the sentinel with the <tt>ilist</tt> instance in memory. Pointer
1193 arithmetic is used to obtain the sentinel, which is relative to the
1194 <tt>ilist</tt>'s <tt>this</tt> pointer. The <tt>ilist</tt> is augmented by an
1195 extra pointer, which serves as the back-link of the sentinel. This is the only
1196 field in the ghostly sentinel which can be legally accessed.</p>
1199 <!-- _______________________________________________________________________ -->
1201 <a name="dss_other">Other Sequential Container options</a>
1205 <p>Other STL containers are available, such as std::string.</p>
1207 <p>There are also various STL adapter classes such as std::queue,
1208 std::priority_queue, std::stack, etc. These provide simplified access to an
1209 underlying container but don't affect the cost of the container itself.</p>
1215 <!-- ======================================================================= -->
1217 <a name="ds_set">Set-Like Containers (std::set, SmallSet, SetVector, etc)</a>
1222 <p>Set-like containers are useful when you need to canonicalize multiple values
1223 into a single representation. There are several different choices for how to do
1224 this, providing various trade-offs.</p>
1226 <!-- _______________________________________________________________________ -->
1228 <a name="dss_sortedvectorset">A sorted 'vector'</a>
1233 <p>If you intend to insert a lot of elements, then do a lot of queries, a
1234 great approach is to use a vector (or other sequential container) with
1235 std::sort+std::unique to remove duplicates. This approach works really well if
1236 your usage pattern has these two distinct phases (insert then query), and can be
1237 coupled with a good choice of <a href="#ds_sequential">sequential container</a>.
1241 This combination provides the several nice properties: the result data is
1242 contiguous in memory (good for cache locality), has few allocations, is easy to
1243 address (iterators in the final vector are just indices or pointers), and can be
1244 efficiently queried with a standard binary or radix search.</p>
1248 <!-- _______________________________________________________________________ -->
1250 <a name="dss_smallset">"llvm/ADT/SmallSet.h"</a>
1255 <p>If you have a set-like data structure that is usually small and whose elements
1256 are reasonably small, a <tt>SmallSet<Type, N></tt> is a good choice. This set
1257 has space for N elements in place (thus, if the set is dynamically smaller than
1258 N, no malloc traffic is required) and accesses them with a simple linear search.
1259 When the set grows beyond 'N' elements, it allocates a more expensive representation that
1260 guarantees efficient access (for most types, it falls back to std::set, but for
1261 pointers it uses something far better, <a
1262 href="#dss_smallptrset">SmallPtrSet</a>).</p>
1264 <p>The magic of this class is that it handles small sets extremely efficiently,
1265 but gracefully handles extremely large sets without loss of efficiency. The
1266 drawback is that the interface is quite small: it supports insertion, queries
1267 and erasing, but does not support iteration.</p>
1271 <!-- _______________________________________________________________________ -->
1273 <a name="dss_smallptrset">"llvm/ADT/SmallPtrSet.h"</a>
1278 <p>SmallPtrSet has all the advantages of <tt>SmallSet</tt> (and a <tt>SmallSet</tt> of pointers is
1279 transparently implemented with a <tt>SmallPtrSet</tt>), but also supports iterators. If
1280 more than 'N' insertions are performed, a single quadratically
1281 probed hash table is allocated and grows as needed, providing extremely
1282 efficient access (constant time insertion/deleting/queries with low constant
1283 factors) and is very stingy with malloc traffic.</p>
1285 <p>Note that, unlike <tt>std::set</tt>, the iterators of <tt>SmallPtrSet</tt> are invalidated
1286 whenever an insertion occurs. Also, the values visited by the iterators are not
1287 visited in sorted order.</p>
1291 <!-- _______________________________________________________________________ -->
1293 <a name="dss_denseset">"llvm/ADT/DenseSet.h"</a>
1299 DenseSet is a simple quadratically probed hash table. It excels at supporting
1300 small values: it uses a single allocation to hold all of the pairs that
1301 are currently inserted in the set. DenseSet is a great way to unique small
1302 values that are not simple pointers (use <a
1303 href="#dss_smallptrset">SmallPtrSet</a> for pointers). Note that DenseSet has
1304 the same requirements for the value type that <a
1305 href="#dss_densemap">DenseMap</a> has.
1310 <!-- _______________________________________________________________________ -->
1312 <a name="dss_FoldingSet">"llvm/ADT/FoldingSet.h"</a>
1318 FoldingSet is an aggregate class that is really good at uniquing
1319 expensive-to-create or polymorphic objects. It is a combination of a chained
1320 hash table with intrusive links (uniqued objects are required to inherit from
1321 FoldingSetNode) that uses <a href="#dss_smallvector">SmallVector</a> as part of
1324 <p>Consider a case where you want to implement a "getOrCreateFoo" method for
1325 a complex object (for example, a node in the code generator). The client has a
1326 description of *what* it wants to generate (it knows the opcode and all the
1327 operands), but we don't want to 'new' a node, then try inserting it into a set
1328 only to find out it already exists, at which point we would have to delete it
1329 and return the node that already exists.
1332 <p>To support this style of client, FoldingSet perform a query with a
1333 FoldingSetNodeID (which wraps SmallVector) that can be used to describe the
1334 element that we want to query for. The query either returns the element
1335 matching the ID or it returns an opaque ID that indicates where insertion should
1336 take place. Construction of the ID usually does not require heap traffic.</p>
1338 <p>Because FoldingSet uses intrusive links, it can support polymorphic objects
1339 in the set (for example, you can have SDNode instances mixed with LoadSDNodes).
1340 Because the elements are individually allocated, pointers to the elements are
1341 stable: inserting or removing elements does not invalidate any pointers to other
1347 <!-- _______________________________________________________________________ -->
1349 <a name="dss_set"><set></a>
1354 <p><tt>std::set</tt> is a reasonable all-around set class, which is decent at
1355 many things but great at nothing. std::set allocates memory for each element
1356 inserted (thus it is very malloc intensive) and typically stores three pointers
1357 per element in the set (thus adding a large amount of per-element space
1358 overhead). It offers guaranteed log(n) performance, which is not particularly
1359 fast from a complexity standpoint (particularly if the elements of the set are
1360 expensive to compare, like strings), and has extremely high constant factors for
1361 lookup, insertion and removal.</p>
1363 <p>The advantages of std::set are that its iterators are stable (deleting or
1364 inserting an element from the set does not affect iterators or pointers to other
1365 elements) and that iteration over the set is guaranteed to be in sorted order.
1366 If the elements in the set are large, then the relative overhead of the pointers
1367 and malloc traffic is not a big deal, but if the elements of the set are small,
1368 std::set is almost never a good choice.</p>
1372 <!-- _______________________________________________________________________ -->
1374 <a name="dss_setvector">"llvm/ADT/SetVector.h"</a>
1378 <p>LLVM's SetVector<Type> is an adapter class that combines your choice of
1379 a set-like container along with a <a href="#ds_sequential">Sequential
1380 Container</a>. The important property
1381 that this provides is efficient insertion with uniquing (duplicate elements are
1382 ignored) with iteration support. It implements this by inserting elements into
1383 both a set-like container and the sequential container, using the set-like
1384 container for uniquing and the sequential container for iteration.
1387 <p>The difference between SetVector and other sets is that the order of
1388 iteration is guaranteed to match the order of insertion into the SetVector.
1389 This property is really important for things like sets of pointers. Because
1390 pointer values are non-deterministic (e.g. vary across runs of the program on
1391 different machines), iterating over the pointers in the set will
1392 not be in a well-defined order.</p>
1395 The drawback of SetVector is that it requires twice as much space as a normal
1396 set and has the sum of constant factors from the set-like container and the
1397 sequential container that it uses. Use it *only* if you need to iterate over
1398 the elements in a deterministic order. SetVector is also expensive to delete
1399 elements out of (linear time), unless you use it's "pop_back" method, which is
1403 <p>SetVector is an adapter class that defaults to using std::vector and std::set
1404 for the underlying containers, so it is quite expensive. However,
1405 <tt>"llvm/ADT/SetVector.h"</tt> also provides a SmallSetVector class, which
1406 defaults to using a SmallVector and SmallSet of a specified size. If you use
1407 this, and if your sets are dynamically smaller than N, you will save a lot of
1412 <!-- _______________________________________________________________________ -->
1414 <a name="dss_uniquevector">"llvm/ADT/UniqueVector.h"</a>
1420 UniqueVector is similar to <a href="#dss_setvector">SetVector</a>, but it
1421 retains a unique ID for each element inserted into the set. It internally
1422 contains a map and a vector, and it assigns a unique ID for each value inserted
1425 <p>UniqueVector is very expensive: its cost is the sum of the cost of
1426 maintaining both the map and vector, it has high complexity, high constant
1427 factors, and produces a lot of malloc traffic. It should be avoided.</p>
1432 <!-- _______________________________________________________________________ -->
1434 <a name="dss_otherset">Other Set-Like Container Options</a>
1440 The STL provides several other options, such as std::multiset and the various
1441 "hash_set" like containers (whether from C++ TR1 or from the SGI library). We
1442 never use hash_set and unordered_set because they are generally very expensive
1443 (each insertion requires a malloc) and very non-portable.
1446 <p>std::multiset is useful if you're not interested in elimination of
1447 duplicates, but has all the drawbacks of std::set. A sorted vector (where you
1448 don't delete duplicate entries) or some other approach is almost always
1455 <!-- ======================================================================= -->
1457 <a name="ds_map">Map-Like Containers (std::map, DenseMap, etc)</a>
1461 Map-like containers are useful when you want to associate data to a key. As
1462 usual, there are a lot of different ways to do this. :)
1464 <!-- _______________________________________________________________________ -->
1466 <a name="dss_sortedvectormap">A sorted 'vector'</a>
1472 If your usage pattern follows a strict insert-then-query approach, you can
1473 trivially use the same approach as <a href="#dss_sortedvectorset">sorted vectors
1474 for set-like containers</a>. The only difference is that your query function
1475 (which uses std::lower_bound to get efficient log(n) lookup) should only compare
1476 the key, not both the key and value. This yields the same advantages as sorted
1481 <!-- _______________________________________________________________________ -->
1483 <a name="dss_stringmap">"llvm/ADT/StringMap.h"</a>
1489 Strings are commonly used as keys in maps, and they are difficult to support
1490 efficiently: they are variable length, inefficient to hash and compare when
1491 long, expensive to copy, etc. StringMap is a specialized container designed to
1492 cope with these issues. It supports mapping an arbitrary range of bytes to an
1493 arbitrary other object.</p>
1495 <p>The StringMap implementation uses a quadratically-probed hash table, where
1496 the buckets store a pointer to the heap allocated entries (and some other
1497 stuff). The entries in the map must be heap allocated because the strings are
1498 variable length. The string data (key) and the element object (value) are
1499 stored in the same allocation with the string data immediately after the element
1500 object. This container guarantees the "<tt>(char*)(&Value+1)</tt>" points
1501 to the key string for a value.</p>
1503 <p>The StringMap is very fast for several reasons: quadratic probing is very
1504 cache efficient for lookups, the hash value of strings in buckets is not
1505 recomputed when looking up an element, StringMap rarely has to touch the
1506 memory for unrelated objects when looking up a value (even when hash collisions
1507 happen), hash table growth does not recompute the hash values for strings
1508 already in the table, and each pair in the map is store in a single allocation
1509 (the string data is stored in the same allocation as the Value of a pair).</p>
1511 <p>StringMap also provides query methods that take byte ranges, so it only ever
1512 copies a string if a value is inserted into the table.</p>
1515 <!-- _______________________________________________________________________ -->
1517 <a name="dss_indexedmap">"llvm/ADT/IndexedMap.h"</a>
1522 IndexedMap is a specialized container for mapping small dense integers (or
1523 values that can be mapped to small dense integers) to some other type. It is
1524 internally implemented as a vector with a mapping function that maps the keys to
1525 the dense integer range.
1529 This is useful for cases like virtual registers in the LLVM code generator: they
1530 have a dense mapping that is offset by a compile-time constant (the first
1531 virtual register ID).</p>
1535 <!-- _______________________________________________________________________ -->
1537 <a name="dss_densemap">"llvm/ADT/DenseMap.h"</a>
1543 DenseMap is a simple quadratically probed hash table. It excels at supporting
1544 small keys and values: it uses a single allocation to hold all of the pairs that
1545 are currently inserted in the map. DenseMap is a great way to map pointers to
1546 pointers, or map other small types to each other.
1550 There are several aspects of DenseMap that you should be aware of, however. The
1551 iterators in a densemap are invalidated whenever an insertion occurs, unlike
1552 map. Also, because DenseMap allocates space for a large number of key/value
1553 pairs (it starts with 64 by default), it will waste a lot of space if your keys
1554 or values are large. Finally, you must implement a partial specialization of
1555 DenseMapInfo for the key that you want, if it isn't already supported. This
1556 is required to tell DenseMap about two special marker values (which can never be
1557 inserted into the map) that it needs internally.</p>
1561 <!-- _______________________________________________________________________ -->
1563 <a name="dss_valuemap">"llvm/ADT/ValueMap.h"</a>
1569 ValueMap is a wrapper around a <a href="#dss_densemap">DenseMap</a> mapping
1570 Value*s (or subclasses) to another type. When a Value is deleted or RAUW'ed,
1571 ValueMap will update itself so the new version of the key is mapped to the same
1572 value, just as if the key were a WeakVH. You can configure exactly how this
1573 happens, and what else happens on these two events, by passing
1574 a <code>Config</code> parameter to the ValueMap template.</p>
1578 <!-- _______________________________________________________________________ -->
1580 <a name="dss_intervalmap">"llvm/ADT/IntervalMap.h"</a>
1585 <p> IntervalMap is a compact map for small keys and values. It maps key
1586 intervals instead of single keys, and it will automatically coalesce adjacent
1587 intervals. When then map only contains a few intervals, they are stored in the
1588 map object itself to avoid allocations.</p>
1590 <p> The IntervalMap iterators are quite big, so they should not be passed around
1591 as STL iterators. The heavyweight iterators allow a smaller data structure.</p>
1595 <!-- _______________________________________________________________________ -->
1597 <a name="dss_map"><map></a>
1603 std::map has similar characteristics to <a href="#dss_set">std::set</a>: it uses
1604 a single allocation per pair inserted into the map, it offers log(n) lookup with
1605 an extremely large constant factor, imposes a space penalty of 3 pointers per
1606 pair in the map, etc.</p>
1608 <p>std::map is most useful when your keys or values are very large, if you need
1609 to iterate over the collection in sorted order, or if you need stable iterators
1610 into the map (i.e. they don't get invalidated if an insertion or deletion of
1611 another element takes place).</p>
1615 <!-- _______________________________________________________________________ -->
1617 <a name="dss_inteqclasses">"llvm/ADT/IntEqClasses.h"</a>
1622 <p>IntEqClasses provides a compact representation of equivalence classes of
1623 small integers. Initially, each integer in the range 0..n-1 has its own
1624 equivalence class. Classes can be joined by passing two class representatives to
1625 the join(a, b) method. Two integers are in the same class when findLeader()
1626 returns the same representative.</p>
1628 <p>Once all equivalence classes are formed, the map can be compressed so each
1629 integer 0..n-1 maps to an equivalence class number in the range 0..m-1, where m
1630 is the total number of equivalence classes. The map must be uncompressed before
1631 it can be edited again.</p>
1635 <!-- _______________________________________________________________________ -->
1637 <a name="dss_othermap">Other Map-Like Container Options</a>
1643 The STL provides several other options, such as std::multimap and the various
1644 "hash_map" like containers (whether from C++ TR1 or from the SGI library). We
1645 never use hash_set and unordered_set because they are generally very expensive
1646 (each insertion requires a malloc) and very non-portable.</p>
1648 <p>std::multimap is useful if you want to map a key to multiple values, but has
1649 all the drawbacks of std::map. A sorted vector or some other approach is almost
1656 <!-- ======================================================================= -->
1658 <a name="ds_string">String-like containers</a>
1664 TODO: const char* vs stringref vs smallstring vs std::string. Describe twine,
1665 xref to #string_apis.
1670 <!-- ======================================================================= -->
1672 <a name="ds_bit">Bit storage containers (BitVector, SparseBitVector)</a>
1676 <p>Unlike the other containers, there are only two bit storage containers, and
1677 choosing when to use each is relatively straightforward.</p>
1679 <p>One additional option is
1680 <tt>std::vector<bool></tt>: we discourage its use for two reasons 1) the
1681 implementation in many common compilers (e.g. commonly available versions of
1682 GCC) is extremely inefficient and 2) the C++ standards committee is likely to
1683 deprecate this container and/or change it significantly somehow. In any case,
1684 please don't use it.</p>
1686 <!-- _______________________________________________________________________ -->
1688 <a name="dss_bitvector">BitVector</a>
1692 <p> The BitVector container provides a dynamic size set of bits for manipulation.
1693 It supports individual bit setting/testing, as well as set operations. The set
1694 operations take time O(size of bitvector), but operations are performed one word
1695 at a time, instead of one bit at a time. This makes the BitVector very fast for
1696 set operations compared to other containers. Use the BitVector when you expect
1697 the number of set bits to be high (IE a dense set).
1701 <!-- _______________________________________________________________________ -->
1703 <a name="dss_smallbitvector">SmallBitVector</a>
1707 <p> The SmallBitVector container provides the same interface as BitVector, but
1708 it is optimized for the case where only a small number of bits, less than
1709 25 or so, are needed. It also transparently supports larger bit counts, but
1710 slightly less efficiently than a plain BitVector, so SmallBitVector should
1711 only be used when larger counts are rare.
1715 At this time, SmallBitVector does not support set operations (and, or, xor),
1716 and its operator[] does not provide an assignable lvalue.
1720 <!-- _______________________________________________________________________ -->
1722 <a name="dss_sparsebitvector">SparseBitVector</a>
1726 <p> The SparseBitVector container is much like BitVector, with one major
1727 difference: Only the bits that are set, are stored. This makes the
1728 SparseBitVector much more space efficient than BitVector when the set is sparse,
1729 as well as making set operations O(number of set bits) instead of O(size of
1730 universe). The downside to the SparseBitVector is that setting and testing of random bits is O(N), and on large SparseBitVectors, this can be slower than BitVector. In our implementation, setting or testing bits in sorted order
1731 (either forwards or reverse) is O(1) worst case. Testing and setting bits within 128 bits (depends on size) of the current bit is also O(1). As a general statement, testing/setting bits in a SparseBitVector is O(distance away from last set bit).
1739 <!-- *********************************************************************** -->
1741 <a name="common">Helpful Hints for Common Operations</a>
1743 <!-- *********************************************************************** -->
1747 <p>This section describes how to perform some very simple transformations of
1748 LLVM code. This is meant to give examples of common idioms used, showing the
1749 practical side of LLVM transformations. <p> Because this is a "how-to" section,
1750 you should also read about the main classes that you will be working with. The
1751 <a href="#coreclasses">Core LLVM Class Hierarchy Reference</a> contains details
1752 and descriptions of the main classes that you should know about.</p>
1754 <!-- NOTE: this section should be heavy on example code -->
1755 <!-- ======================================================================= -->
1757 <a name="inspection">Basic Inspection and Traversal Routines</a>
1762 <p>The LLVM compiler infrastructure have many different data structures that may
1763 be traversed. Following the example of the C++ standard template library, the
1764 techniques used to traverse these various data structures are all basically the
1765 same. For a enumerable sequence of values, the <tt>XXXbegin()</tt> function (or
1766 method) returns an iterator to the start of the sequence, the <tt>XXXend()</tt>
1767 function returns an iterator pointing to one past the last valid element of the
1768 sequence, and there is some <tt>XXXiterator</tt> data type that is common
1769 between the two operations.</p>
1771 <p>Because the pattern for iteration is common across many different aspects of
1772 the program representation, the standard template library algorithms may be used
1773 on them, and it is easier to remember how to iterate. First we show a few common
1774 examples of the data structures that need to be traversed. Other data
1775 structures are traversed in very similar ways.</p>
1777 <!-- _______________________________________________________________________ -->
1779 <a name="iterate_function">Iterating over the </a><a
1780 href="#BasicBlock"><tt>BasicBlock</tt></a>s in a <a
1781 href="#Function"><tt>Function</tt></a>
1786 <p>It's quite common to have a <tt>Function</tt> instance that you'd like to
1787 transform in some way; in particular, you'd like to manipulate its
1788 <tt>BasicBlock</tt>s. To facilitate this, you'll need to iterate over all of
1789 the <tt>BasicBlock</tt>s that constitute the <tt>Function</tt>. The following is
1790 an example that prints the name of a <tt>BasicBlock</tt> and the number of
1791 <tt>Instruction</tt>s it contains:</p>
1793 <div class="doc_code">
1795 // <i>func is a pointer to a Function instance</i>
1796 for (Function::iterator i = func->begin(), e = func->end(); i != e; ++i)
1797 // <i>Print out the name of the basic block if it has one, and then the</i>
1798 // <i>number of instructions that it contains</i>
1799 errs() << "Basic block (name=" << i->getName() << ") has "
1800 << i->size() << " instructions.\n";
1804 <p>Note that i can be used as if it were a pointer for the purposes of
1805 invoking member functions of the <tt>Instruction</tt> class. This is
1806 because the indirection operator is overloaded for the iterator
1807 classes. In the above code, the expression <tt>i->size()</tt> is
1808 exactly equivalent to <tt>(*i).size()</tt> just like you'd expect.</p>
1812 <!-- _______________________________________________________________________ -->
1814 <a name="iterate_basicblock">Iterating over the </a><a
1815 href="#Instruction"><tt>Instruction</tt></a>s in a <a
1816 href="#BasicBlock"><tt>BasicBlock</tt></a>
1821 <p>Just like when dealing with <tt>BasicBlock</tt>s in <tt>Function</tt>s, it's
1822 easy to iterate over the individual instructions that make up
1823 <tt>BasicBlock</tt>s. Here's a code snippet that prints out each instruction in
1824 a <tt>BasicBlock</tt>:</p>
1826 <div class="doc_code">
1828 // <i>blk is a pointer to a BasicBlock instance</i>
1829 for (BasicBlock::iterator i = blk->begin(), e = blk->end(); i != e; ++i)
1830 // <i>The next statement works since operator<<(ostream&,...)</i>
1831 // <i>is overloaded for Instruction&</i>
1832 errs() << *i << "\n";
1836 <p>However, this isn't really the best way to print out the contents of a
1837 <tt>BasicBlock</tt>! Since the ostream operators are overloaded for virtually
1838 anything you'll care about, you could have just invoked the print routine on the
1839 basic block itself: <tt>errs() << *blk << "\n";</tt>.</p>
1843 <!-- _______________________________________________________________________ -->
1845 <a name="iterate_institer">Iterating over the </a><a
1846 href="#Instruction"><tt>Instruction</tt></a>s in a <a
1847 href="#Function"><tt>Function</tt></a>
1852 <p>If you're finding that you commonly iterate over a <tt>Function</tt>'s
1853 <tt>BasicBlock</tt>s and then that <tt>BasicBlock</tt>'s <tt>Instruction</tt>s,
1854 <tt>InstIterator</tt> should be used instead. You'll need to include <a
1855 href="/doxygen/InstIterator_8h-source.html"><tt>llvm/Support/InstIterator.h</tt></a>,
1856 and then instantiate <tt>InstIterator</tt>s explicitly in your code. Here's a
1857 small example that shows how to dump all instructions in a function to the standard error stream:<p>
1859 <div class="doc_code">
1861 #include "<a href="/doxygen/InstIterator_8h-source.html">llvm/Support/InstIterator.h</a>"
1863 // <i>F is a pointer to a Function instance</i>
1864 for (inst_iterator I = inst_begin(F), E = inst_end(F); I != E; ++I)
1865 errs() << *I << "\n";
1869 <p>Easy, isn't it? You can also use <tt>InstIterator</tt>s to fill a
1870 work list with its initial contents. For example, if you wanted to
1871 initialize a work list to contain all instructions in a <tt>Function</tt>
1872 F, all you would need to do is something like:</p>
1874 <div class="doc_code">
1876 std::set<Instruction*> worklist;
1877 // or better yet, SmallPtrSet<Instruction*, 64> worklist;
1879 for (inst_iterator I = inst_begin(F), E = inst_end(F); I != E; ++I)
1880 worklist.insert(&*I);
1884 <p>The STL set <tt>worklist</tt> would now contain all instructions in the
1885 <tt>Function</tt> pointed to by F.</p>
1889 <!-- _______________________________________________________________________ -->
1891 <a name="iterate_convert">Turning an iterator into a class pointer (and
1897 <p>Sometimes, it'll be useful to grab a reference (or pointer) to a class
1898 instance when all you've got at hand is an iterator. Well, extracting
1899 a reference or a pointer from an iterator is very straight-forward.
1900 Assuming that <tt>i</tt> is a <tt>BasicBlock::iterator</tt> and <tt>j</tt>
1901 is a <tt>BasicBlock::const_iterator</tt>:</p>
1903 <div class="doc_code">
1905 Instruction& inst = *i; // <i>Grab reference to instruction reference</i>
1906 Instruction* pinst = &*i; // <i>Grab pointer to instruction reference</i>
1907 const Instruction& inst = *j;
1911 <p>However, the iterators you'll be working with in the LLVM framework are
1912 special: they will automatically convert to a ptr-to-instance type whenever they
1913 need to. Instead of dereferencing the iterator and then taking the address of
1914 the result, you can simply assign the iterator to the proper pointer type and
1915 you get the dereference and address-of operation as a result of the assignment
1916 (behind the scenes, this is a result of overloading casting mechanisms). Thus
1917 the last line of the last example,</p>
1919 <div class="doc_code">
1921 Instruction *pinst = &*i;
1925 <p>is semantically equivalent to</p>
1927 <div class="doc_code">
1929 Instruction *pinst = i;
1933 <p>It's also possible to turn a class pointer into the corresponding iterator,
1934 and this is a constant time operation (very efficient). The following code
1935 snippet illustrates use of the conversion constructors provided by LLVM
1936 iterators. By using these, you can explicitly grab the iterator of something
1937 without actually obtaining it via iteration over some structure:</p>
1939 <div class="doc_code">
1941 void printNextInstruction(Instruction* inst) {
1942 BasicBlock::iterator it(inst);
1943 ++it; // <i>After this line, it refers to the instruction after *inst</i>
1944 if (it != inst->getParent()->end()) errs() << *it << "\n";
1949 <p>Unfortunately, these implicit conversions come at a cost; they prevent
1950 these iterators from conforming to standard iterator conventions, and thus
1951 from being usable with standard algorithms and containers. For example, they
1952 prevent the following code, where <tt>B</tt> is a <tt>BasicBlock</tt>,
1955 <div class="doc_code">
1957 llvm::SmallVector<llvm::Instruction *, 16>(B->begin(), B->end());
1961 <p>Because of this, these implicit conversions may be removed some day,
1962 and <tt>operator*</tt> changed to return a pointer instead of a reference.</p>
1966 <!--_______________________________________________________________________-->
1968 <a name="iterate_complex">Finding call sites: a slightly more complex
1974 <p>Say that you're writing a FunctionPass and would like to count all the
1975 locations in the entire module (that is, across every <tt>Function</tt>) where a
1976 certain function (i.e., some <tt>Function</tt>*) is already in scope. As you'll
1977 learn later, you may want to use an <tt>InstVisitor</tt> to accomplish this in a
1978 much more straight-forward manner, but this example will allow us to explore how
1979 you'd do it if you didn't have <tt>InstVisitor</tt> around. In pseudo-code, this
1980 is what we want to do:</p>
1982 <div class="doc_code">
1984 initialize callCounter to zero
1985 for each Function f in the Module
1986 for each BasicBlock b in f
1987 for each Instruction i in b
1988 if (i is a CallInst and calls the given function)
1989 increment callCounter
1993 <p>And the actual code is (remember, because we're writing a
1994 <tt>FunctionPass</tt>, our <tt>FunctionPass</tt>-derived class simply has to
1995 override the <tt>runOnFunction</tt> method):</p>
1997 <div class="doc_code">
1999 Function* targetFunc = ...;
2001 class OurFunctionPass : public FunctionPass {
2003 OurFunctionPass(): callCounter(0) { }
2005 virtual runOnFunction(Function& F) {
2006 for (Function::iterator b = F.begin(), be = F.end(); b != be; ++b) {
2007 for (BasicBlock::iterator i = b->begin(), ie = b->end(); i != ie; ++i) {
2008 if (<a href="#CallInst">CallInst</a>* callInst = <a href="#isa">dyn_cast</a><<a
2009 href="#CallInst">CallInst</a>>(&*i)) {
2010 // <i>We know we've encountered a call instruction, so we</i>
2011 // <i>need to determine if it's a call to the</i>
2012 // <i>function pointed to by m_func or not.</i>
2013 if (callInst->getCalledFunction() == targetFunc)
2021 unsigned callCounter;
2028 <!--_______________________________________________________________________-->
2030 <a name="calls_and_invokes">Treating calls and invokes the same way</a>
2035 <p>You may have noticed that the previous example was a bit oversimplified in
2036 that it did not deal with call sites generated by 'invoke' instructions. In
2037 this, and in other situations, you may find that you want to treat
2038 <tt>CallInst</tt>s and <tt>InvokeInst</tt>s the same way, even though their
2039 most-specific common base class is <tt>Instruction</tt>, which includes lots of
2040 less closely-related things. For these cases, LLVM provides a handy wrapper
2042 href="http://llvm.org/doxygen/classllvm_1_1CallSite.html"><tt>CallSite</tt></a>.
2043 It is essentially a wrapper around an <tt>Instruction</tt> pointer, with some
2044 methods that provide functionality common to <tt>CallInst</tt>s and
2045 <tt>InvokeInst</tt>s.</p>
2047 <p>This class has "value semantics": it should be passed by value, not by
2048 reference and it should not be dynamically allocated or deallocated using
2049 <tt>operator new</tt> or <tt>operator delete</tt>. It is efficiently copyable,
2050 assignable and constructable, with costs equivalents to that of a bare pointer.
2051 If you look at its definition, it has only a single pointer member.</p>
2055 <!--_______________________________________________________________________-->
2057 <a name="iterate_chains">Iterating over def-use & use-def chains</a>
2062 <p>Frequently, we might have an instance of the <a
2063 href="/doxygen/classllvm_1_1Value.html">Value Class</a> and we want to
2064 determine which <tt>User</tt>s use the <tt>Value</tt>. The list of all
2065 <tt>User</tt>s of a particular <tt>Value</tt> is called a <i>def-use</i> chain.
2066 For example, let's say we have a <tt>Function*</tt> named <tt>F</tt> to a
2067 particular function <tt>foo</tt>. Finding all of the instructions that
2068 <i>use</i> <tt>foo</tt> is as simple as iterating over the <i>def-use</i> chain
2071 <div class="doc_code">
2075 for (Value::use_iterator i = F->use_begin(), e = F->use_end(); i != e; ++i)
2076 if (Instruction *Inst = dyn_cast<Instruction>(*i)) {
2077 errs() << "F is used in instruction:\n";
2078 errs() << *Inst << "\n";
2083 <p>Note that dereferencing a <tt>Value::use_iterator</tt> is not a very cheap
2084 operation. Instead of performing <tt>*i</tt> above several times, consider
2085 doing it only once in the loop body and reusing its result.</p>
2087 <p>Alternatively, it's common to have an instance of the <a
2088 href="/doxygen/classllvm_1_1User.html">User Class</a> and need to know what
2089 <tt>Value</tt>s are used by it. The list of all <tt>Value</tt>s used by a
2090 <tt>User</tt> is known as a <i>use-def</i> chain. Instances of class
2091 <tt>Instruction</tt> are common <tt>User</tt>s, so we might want to iterate over
2092 all of the values that a particular instruction uses (that is, the operands of
2093 the particular <tt>Instruction</tt>):</p>
2095 <div class="doc_code">
2097 Instruction *pi = ...;
2099 for (User::op_iterator i = pi->op_begin(), e = pi->op_end(); i != e; ++i) {
2106 <p>Declaring objects as <tt>const</tt> is an important tool of enforcing
2107 mutation free algorithms (such as analyses, etc.). For this purpose above
2108 iterators come in constant flavors as <tt>Value::const_use_iterator</tt>
2109 and <tt>Value::const_op_iterator</tt>. They automatically arise when
2110 calling <tt>use/op_begin()</tt> on <tt>const Value*</tt>s or
2111 <tt>const User*</tt>s respectively. Upon dereferencing, they return
2112 <tt>const Use*</tt>s. Otherwise the above patterns remain unchanged.</p>
2116 <!--_______________________________________________________________________-->
2118 <a name="iterate_preds">Iterating over predecessors &
2119 successors of blocks</a>
2124 <p>Iterating over the predecessors and successors of a block is quite easy
2125 with the routines defined in <tt>"llvm/Support/CFG.h"</tt>. Just use code like
2126 this to iterate over all predecessors of BB:</p>
2128 <div class="doc_code">
2130 #include "llvm/Support/CFG.h"
2131 BasicBlock *BB = ...;
2133 for (pred_iterator PI = pred_begin(BB), E = pred_end(BB); PI != E; ++PI) {
2134 BasicBlock *Pred = *PI;
2140 <p>Similarly, to iterate over successors use
2141 succ_iterator/succ_begin/succ_end.</p>
2147 <!-- ======================================================================= -->
2149 <a name="simplechanges">Making simple changes</a>
2154 <p>There are some primitive transformation operations present in the LLVM
2155 infrastructure that are worth knowing about. When performing
2156 transformations, it's fairly common to manipulate the contents of basic
2157 blocks. This section describes some of the common methods for doing so
2158 and gives example code.</p>
2160 <!--_______________________________________________________________________-->
2162 <a name="schanges_creating">Creating and inserting new
2163 <tt>Instruction</tt>s</a>
2168 <p><i>Instantiating Instructions</i></p>
2170 <p>Creation of <tt>Instruction</tt>s is straight-forward: simply call the
2171 constructor for the kind of instruction to instantiate and provide the necessary
2172 parameters. For example, an <tt>AllocaInst</tt> only <i>requires</i> a
2173 (const-ptr-to) <tt>Type</tt>. Thus:</p>
2175 <div class="doc_code">
2177 AllocaInst* ai = new AllocaInst(Type::Int32Ty);
2181 <p>will create an <tt>AllocaInst</tt> instance that represents the allocation of
2182 one integer in the current stack frame, at run time. Each <tt>Instruction</tt>
2183 subclass is likely to have varying default parameters which change the semantics
2184 of the instruction, so refer to the <a
2185 href="/doxygen/classllvm_1_1Instruction.html">doxygen documentation for the subclass of
2186 Instruction</a> that you're interested in instantiating.</p>
2188 <p><i>Naming values</i></p>
2190 <p>It is very useful to name the values of instructions when you're able to, as
2191 this facilitates the debugging of your transformations. If you end up looking
2192 at generated LLVM machine code, you definitely want to have logical names
2193 associated with the results of instructions! By supplying a value for the
2194 <tt>Name</tt> (default) parameter of the <tt>Instruction</tt> constructor, you
2195 associate a logical name with the result of the instruction's execution at
2196 run time. For example, say that I'm writing a transformation that dynamically
2197 allocates space for an integer on the stack, and that integer is going to be
2198 used as some kind of index by some other code. To accomplish this, I place an
2199 <tt>AllocaInst</tt> at the first point in the first <tt>BasicBlock</tt> of some
2200 <tt>Function</tt>, and I'm intending to use it within the same
2201 <tt>Function</tt>. I might do:</p>
2203 <div class="doc_code">
2205 AllocaInst* pa = new AllocaInst(Type::Int32Ty, 0, "indexLoc");
2209 <p>where <tt>indexLoc</tt> is now the logical name of the instruction's
2210 execution value, which is a pointer to an integer on the run time stack.</p>
2212 <p><i>Inserting instructions</i></p>
2214 <p>There are essentially two ways to insert an <tt>Instruction</tt>
2215 into an existing sequence of instructions that form a <tt>BasicBlock</tt>:</p>
2218 <li>Insertion into an explicit instruction list
2220 <p>Given a <tt>BasicBlock* pb</tt>, an <tt>Instruction* pi</tt> within that
2221 <tt>BasicBlock</tt>, and a newly-created instruction we wish to insert
2222 before <tt>*pi</tt>, we do the following: </p>
2224 <div class="doc_code">
2226 BasicBlock *pb = ...;
2227 Instruction *pi = ...;
2228 Instruction *newInst = new Instruction(...);
2230 pb->getInstList().insert(pi, newInst); // <i>Inserts newInst before pi in pb</i>
2234 <p>Appending to the end of a <tt>BasicBlock</tt> is so common that
2235 the <tt>Instruction</tt> class and <tt>Instruction</tt>-derived
2236 classes provide constructors which take a pointer to a
2237 <tt>BasicBlock</tt> to be appended to. For example code that
2240 <div class="doc_code">
2242 BasicBlock *pb = ...;
2243 Instruction *newInst = new Instruction(...);
2245 pb->getInstList().push_back(newInst); // <i>Appends newInst to pb</i>
2251 <div class="doc_code">
2253 BasicBlock *pb = ...;
2254 Instruction *newInst = new Instruction(..., pb);
2258 <p>which is much cleaner, especially if you are creating
2259 long instruction streams.</p></li>
2261 <li>Insertion into an implicit instruction list
2263 <p><tt>Instruction</tt> instances that are already in <tt>BasicBlock</tt>s
2264 are implicitly associated with an existing instruction list: the instruction
2265 list of the enclosing basic block. Thus, we could have accomplished the same
2266 thing as the above code without being given a <tt>BasicBlock</tt> by doing:
2269 <div class="doc_code">
2271 Instruction *pi = ...;
2272 Instruction *newInst = new Instruction(...);
2274 pi->getParent()->getInstList().insert(pi, newInst);
2278 <p>In fact, this sequence of steps occurs so frequently that the
2279 <tt>Instruction</tt> class and <tt>Instruction</tt>-derived classes provide
2280 constructors which take (as a default parameter) a pointer to an
2281 <tt>Instruction</tt> which the newly-created <tt>Instruction</tt> should
2282 precede. That is, <tt>Instruction</tt> constructors are capable of
2283 inserting the newly-created instance into the <tt>BasicBlock</tt> of a
2284 provided instruction, immediately before that instruction. Using an
2285 <tt>Instruction</tt> constructor with a <tt>insertBefore</tt> (default)
2286 parameter, the above code becomes:</p>
2288 <div class="doc_code">
2290 Instruction* pi = ...;
2291 Instruction* newInst = new Instruction(..., pi);
2295 <p>which is much cleaner, especially if you're creating a lot of
2296 instructions and adding them to <tt>BasicBlock</tt>s.</p></li>
2301 <!--_______________________________________________________________________-->
2303 <a name="schanges_deleting">Deleting <tt>Instruction</tt>s</a>
2308 <p>Deleting an instruction from an existing sequence of instructions that form a
2309 <a href="#BasicBlock"><tt>BasicBlock</tt></a> is very straight-forward: just
2310 call the instruction's eraseFromParent() method. For example:</p>
2312 <div class="doc_code">
2314 <a href="#Instruction">Instruction</a> *I = .. ;
2315 I->eraseFromParent();
2319 <p>This unlinks the instruction from its containing basic block and deletes
2320 it. If you'd just like to unlink the instruction from its containing basic
2321 block but not delete it, you can use the <tt>removeFromParent()</tt> method.</p>
2325 <!--_______________________________________________________________________-->
2327 <a name="schanges_replacing">Replacing an <tt>Instruction</tt> with another
2333 <p><i>Replacing individual instructions</i></p>
2335 <p>Including "<a href="/doxygen/BasicBlockUtils_8h-source.html">llvm/Transforms/Utils/BasicBlockUtils.h</a>"
2336 permits use of two very useful replace functions: <tt>ReplaceInstWithValue</tt>
2337 and <tt>ReplaceInstWithInst</tt>.</p>
2339 <h5><a name="schanges_deleting">Deleting <tt>Instruction</tt>s</a></h5>
2342 <li><tt>ReplaceInstWithValue</tt>
2344 <p>This function replaces all uses of a given instruction with a value,
2345 and then removes the original instruction. The following example
2346 illustrates the replacement of the result of a particular
2347 <tt>AllocaInst</tt> that allocates memory for a single integer with a null
2348 pointer to an integer.</p>
2350 <div class="doc_code">
2352 AllocaInst* instToReplace = ...;
2353 BasicBlock::iterator ii(instToReplace);
2355 ReplaceInstWithValue(instToReplace->getParent()->getInstList(), ii,
2356 Constant::getNullValue(PointerType::getUnqual(Type::Int32Ty)));
2359 <li><tt>ReplaceInstWithInst</tt>
2361 <p>This function replaces a particular instruction with another
2362 instruction, inserting the new instruction into the basic block at the
2363 location where the old instruction was, and replacing any uses of the old
2364 instruction with the new instruction. The following example illustrates
2365 the replacement of one <tt>AllocaInst</tt> with another.</p>
2367 <div class="doc_code">
2369 AllocaInst* instToReplace = ...;
2370 BasicBlock::iterator ii(instToReplace);
2372 ReplaceInstWithInst(instToReplace->getParent()->getInstList(), ii,
2373 new AllocaInst(Type::Int32Ty, 0, "ptrToReplacedInt"));
2377 <p><i>Replacing multiple uses of <tt>User</tt>s and <tt>Value</tt>s</i></p>
2379 <p>You can use <tt>Value::replaceAllUsesWith</tt> and
2380 <tt>User::replaceUsesOfWith</tt> to change more than one use at a time. See the
2381 doxygen documentation for the <a href="/doxygen/classllvm_1_1Value.html">Value Class</a>
2382 and <a href="/doxygen/classllvm_1_1User.html">User Class</a>, respectively, for more
2385 <!-- Value::replaceAllUsesWith User::replaceUsesOfWith Point out:
2386 include/llvm/Transforms/Utils/ especially BasicBlockUtils.h with:
2387 ReplaceInstWithValue, ReplaceInstWithInst -->
2391 <!--_______________________________________________________________________-->
2393 <a name="schanges_deletingGV">Deleting <tt>GlobalVariable</tt>s</a>
2398 <p>Deleting a global variable from a module is just as easy as deleting an
2399 Instruction. First, you must have a pointer to the global variable that you wish
2400 to delete. You use this pointer to erase it from its parent, the module.
2403 <div class="doc_code">
2405 <a href="#GlobalVariable">GlobalVariable</a> *GV = .. ;
2407 GV->eraseFromParent();
2415 <!-- ======================================================================= -->
2417 <a name="create_types">How to Create Types</a>
2422 <p>In generating IR, you may need some complex types. If you know these types
2423 statically, you can use <tt>TypeBuilder<...>::get()</tt>, defined
2424 in <tt>llvm/Support/TypeBuilder.h</tt>, to retrieve them. <tt>TypeBuilder</tt>
2425 has two forms depending on whether you're building types for cross-compilation
2426 or native library use. <tt>TypeBuilder<T, true></tt> requires
2427 that <tt>T</tt> be independent of the host environment, meaning that it's built
2429 the <a href="/doxygen/namespacellvm_1_1types.html"><tt>llvm::types</tt></a>
2430 namespace and pointers, functions, arrays, etc. built of
2431 those. <tt>TypeBuilder<T, false></tt> additionally allows native C types
2432 whose size may depend on the host compiler. For example,</p>
2434 <div class="doc_code">
2436 FunctionType *ft = TypeBuilder<types::i<8>(types::i<32>*), true>::get();
2440 <p>is easier to read and write than the equivalent</p>
2442 <div class="doc_code">
2444 std::vector<const Type*> params;
2445 params.push_back(PointerType::getUnqual(Type::Int32Ty));
2446 FunctionType *ft = FunctionType::get(Type::Int8Ty, params, false);
2450 <p>See the <a href="/doxygen/TypeBuilder_8h-source.html#l00001">class
2451 comment</a> for more details.</p>
2457 <!-- *********************************************************************** -->
2459 <a name="threading">Threads and LLVM</a>
2461 <!-- *********************************************************************** -->
2465 This section describes the interaction of the LLVM APIs with multithreading,
2466 both on the part of client applications, and in the JIT, in the hosted
2471 Note that LLVM's support for multithreading is still relatively young. Up
2472 through version 2.5, the execution of threaded hosted applications was
2473 supported, but not threaded client access to the APIs. While this use case is
2474 now supported, clients <em>must</em> adhere to the guidelines specified below to
2475 ensure proper operation in multithreaded mode.
2479 Note that, on Unix-like platforms, LLVM requires the presence of GCC's atomic
2480 intrinsics in order to support threaded operation. If you need a
2481 multhreading-capable LLVM on a platform without a suitably modern system
2482 compiler, consider compiling LLVM and LLVM-GCC in single-threaded mode, and
2483 using the resultant compiler to build a copy of LLVM with multithreading
2487 <!-- ======================================================================= -->
2489 <a name="startmultithreaded">Entering and Exiting Multithreaded Mode</a>
2495 In order to properly protect its internal data structures while avoiding
2496 excessive locking overhead in the single-threaded case, the LLVM must intialize
2497 certain data structures necessary to provide guards around its internals. To do
2498 so, the client program must invoke <tt>llvm_start_multithreaded()</tt> before
2499 making any concurrent LLVM API calls. To subsequently tear down these
2500 structures, use the <tt>llvm_stop_multithreaded()</tt> call. You can also use
2501 the <tt>llvm_is_multithreaded()</tt> call to check the status of multithreaded
2506 Note that both of these calls must be made <em>in isolation</em>. That is to
2507 say that no other LLVM API calls may be executing at any time during the
2508 execution of <tt>llvm_start_multithreaded()</tt> or <tt>llvm_stop_multithreaded
2509 </tt>. It's is the client's responsibility to enforce this isolation.
2513 The return value of <tt>llvm_start_multithreaded()</tt> indicates the success or
2514 failure of the initialization. Failure typically indicates that your copy of
2515 LLVM was built without multithreading support, typically because GCC atomic
2516 intrinsics were not found in your system compiler. In this case, the LLVM API
2517 will not be safe for concurrent calls. However, it <em>will</em> be safe for
2518 hosting threaded applications in the JIT, though <a href="#jitthreading">care
2519 must be taken</a> to ensure that side exits and the like do not accidentally
2520 result in concurrent LLVM API calls.
2524 <!-- ======================================================================= -->
2526 <a name="shutdown">Ending Execution with <tt>llvm_shutdown()</tt></a>
2531 When you are done using the LLVM APIs, you should call <tt>llvm_shutdown()</tt>
2532 to deallocate memory used for internal structures. This will also invoke
2533 <tt>llvm_stop_multithreaded()</tt> if LLVM is operating in multithreaded mode.
2534 As such, <tt>llvm_shutdown()</tt> requires the same isolation guarantees as
2535 <tt>llvm_stop_multithreaded()</tt>.
2539 Note that, if you use scope-based shutdown, you can use the
2540 <tt>llvm_shutdown_obj</tt> class, which calls <tt>llvm_shutdown()</tt> in its
2544 <!-- ======================================================================= -->
2546 <a name="managedstatic">Lazy Initialization with <tt>ManagedStatic</tt></a>
2551 <tt>ManagedStatic</tt> is a utility class in LLVM used to implement static
2552 initialization of static resources, such as the global type tables. Before the
2553 invocation of <tt>llvm_shutdown()</tt>, it implements a simple lazy
2554 initialization scheme. Once <tt>llvm_start_multithreaded()</tt> returns,
2555 however, it uses double-checked locking to implement thread-safe lazy
2560 Note that, because no other threads are allowed to issue LLVM API calls before
2561 <tt>llvm_start_multithreaded()</tt> returns, it is possible to have
2562 <tt>ManagedStatic</tt>s of <tt>llvm::sys::Mutex</tt>s.
2566 The <tt>llvm_acquire_global_lock()</tt> and <tt>llvm_release_global_lock</tt>
2567 APIs provide access to the global lock used to implement the double-checked
2568 locking for lazy initialization. These should only be used internally to LLVM,
2569 and only if you know what you're doing!
2573 <!-- ======================================================================= -->
2575 <a name="llvmcontext">Achieving Isolation with <tt>LLVMContext</tt></a>
2580 <tt>LLVMContext</tt> is an opaque class in the LLVM API which clients can use
2581 to operate multiple, isolated instances of LLVM concurrently within the same
2582 address space. For instance, in a hypothetical compile-server, the compilation
2583 of an individual translation unit is conceptually independent from all the
2584 others, and it would be desirable to be able to compile incoming translation
2585 units concurrently on independent server threads. Fortunately,
2586 <tt>LLVMContext</tt> exists to enable just this kind of scenario!
2590 Conceptually, <tt>LLVMContext</tt> provides isolation. Every LLVM entity
2591 (<tt>Module</tt>s, <tt>Value</tt>s, <tt>Type</tt>s, <tt>Constant</tt>s, etc.)
2592 in LLVM's in-memory IR belongs to an <tt>LLVMContext</tt>. Entities in
2593 different contexts <em>cannot</em> interact with each other: <tt>Module</tt>s in
2594 different contexts cannot be linked together, <tt>Function</tt>s cannot be added
2595 to <tt>Module</tt>s in different contexts, etc. What this means is that is is
2596 safe to compile on multiple threads simultaneously, as long as no two threads
2597 operate on entities within the same context.
2601 In practice, very few places in the API require the explicit specification of a
2602 <tt>LLVMContext</tt>, other than the <tt>Type</tt> creation/lookup APIs.
2603 Because every <tt>Type</tt> carries a reference to its owning context, most
2604 other entities can determine what context they belong to by looking at their
2605 own <tt>Type</tt>. If you are adding new entities to LLVM IR, please try to
2606 maintain this interface design.
2610 For clients that do <em>not</em> require the benefits of isolation, LLVM
2611 provides a convenience API <tt>getGlobalContext()</tt>. This returns a global,
2612 lazily initialized <tt>LLVMContext</tt> that may be used in situations where
2613 isolation is not a concern.
2617 <!-- ======================================================================= -->
2619 <a name="jitthreading">Threads and the JIT</a>
2624 LLVM's "eager" JIT compiler is safe to use in threaded programs. Multiple
2625 threads can call <tt>ExecutionEngine::getPointerToFunction()</tt> or
2626 <tt>ExecutionEngine::runFunction()</tt> concurrently, and multiple threads can
2627 run code output by the JIT concurrently. The user must still ensure that only
2628 one thread accesses IR in a given <tt>LLVMContext</tt> while another thread
2629 might be modifying it. One way to do that is to always hold the JIT lock while
2630 accessing IR outside the JIT (the JIT <em>modifies</em> the IR by adding
2631 <tt>CallbackVH</tt>s). Another way is to only
2632 call <tt>getPointerToFunction()</tt> from the <tt>LLVMContext</tt>'s thread.
2635 <p>When the JIT is configured to compile lazily (using
2636 <tt>ExecutionEngine::DisableLazyCompilation(false)</tt>), there is currently a
2637 <a href="http://llvm.org/bugs/show_bug.cgi?id=5184">race condition</a> in
2638 updating call sites after a function is lazily-jitted. It's still possible to
2639 use the lazy JIT in a threaded program if you ensure that only one thread at a
2640 time can call any particular lazy stub and that the JIT lock guards any IR
2641 access, but we suggest using only the eager JIT in threaded programs.
2647 <!-- *********************************************************************** -->
2649 <a name="advanced">Advanced Topics</a>
2651 <!-- *********************************************************************** -->
2655 This section describes some of the advanced or obscure API's that most clients
2656 do not need to be aware of. These API's tend manage the inner workings of the
2657 LLVM system, and only need to be accessed in unusual circumstances.
2661 <!-- ======================================================================= -->
2663 <a name="SymbolTable">The <tt>ValueSymbolTable</tt> class</a>
2667 <p>The <tt><a href="http://llvm.org/doxygen/classllvm_1_1ValueSymbolTable.html">
2668 ValueSymbolTable</a></tt> class provides a symbol table that the <a
2669 href="#Function"><tt>Function</tt></a> and <a href="#Module">
2670 <tt>Module</tt></a> classes use for naming value definitions. The symbol table
2671 can provide a name for any <a href="#Value"><tt>Value</tt></a>.
2674 <p>Note that the <tt>SymbolTable</tt> class should not be directly accessed
2675 by most clients. It should only be used when iteration over the symbol table
2676 names themselves are required, which is very special purpose. Note that not
2678 <tt><a href="#Value">Value</a></tt>s have names, and those without names (i.e. they have
2679 an empty name) do not exist in the symbol table.
2682 <p>Symbol tables support iteration over the values in the symbol
2683 table with <tt>begin/end/iterator</tt> and supports querying to see if a
2684 specific name is in the symbol table (with <tt>lookup</tt>). The
2685 <tt>ValueSymbolTable</tt> class exposes no public mutator methods, instead,
2686 simply call <tt>setName</tt> on a value, which will autoinsert it into the
2687 appropriate symbol table.</p>
2693 <!-- ======================================================================= -->
2695 <a name="UserLayout">The <tt>User</tt> and owned <tt>Use</tt> classes' memory layout</a>
2699 <p>The <tt><a href="http://llvm.org/doxygen/classllvm_1_1User.html">
2700 User</a></tt> class provides a basis for expressing the ownership of <tt>User</tt>
2701 towards other <tt><a href="http://llvm.org/doxygen/classllvm_1_1Value.html">
2702 Value</a></tt>s. The <tt><a href="http://llvm.org/doxygen/classllvm_1_1Use.html">
2703 Use</a></tt> helper class is employed to do the bookkeeping and to facilitate <i>O(1)</i>
2704 addition and removal.</p>
2706 <!-- ______________________________________________________________________ -->
2709 Interaction and relationship between <tt>User</tt> and <tt>Use</tt> objects
2715 A subclass of <tt>User</tt> can choose between incorporating its <tt>Use</tt> objects
2716 or refer to them out-of-line by means of a pointer. A mixed variant
2717 (some <tt>Use</tt>s inline others hung off) is impractical and breaks the invariant
2718 that the <tt>Use</tt> objects belonging to the same <tt>User</tt> form a contiguous array.
2722 We have 2 different layouts in the <tt>User</tt> (sub)classes:
2725 The <tt>Use</tt> object(s) are inside (resp. at fixed offset) of the <tt>User</tt>
2726 object and there are a fixed number of them.</p>
2729 The <tt>Use</tt> object(s) are referenced by a pointer to an
2730 array from the <tt>User</tt> object and there may be a variable
2734 As of v2.4 each layout still possesses a direct pointer to the
2735 start of the array of <tt>Use</tt>s. Though not mandatory for layout a),
2736 we stick to this redundancy for the sake of simplicity.
2737 The <tt>User</tt> object also stores the number of <tt>Use</tt> objects it
2738 has. (Theoretically this information can also be calculated
2739 given the scheme presented below.)</p>
2741 Special forms of allocation operators (<tt>operator new</tt>)
2742 enforce the following memory layouts:</p>
2745 <li><p>Layout a) is modelled by prepending the <tt>User</tt> object by the <tt>Use[]</tt> array.</p>
2748 ...---.---.---.---.-------...
2749 | P | P | P | P | User
2750 '''---'---'---'---'-------'''
2753 <li><p>Layout b) is modelled by pointing at the <tt>Use[]</tt> array.</p>
2765 <i>(In the above figures '<tt>P</tt>' stands for the <tt>Use**</tt> that
2766 is stored in each <tt>Use</tt> object in the member <tt>Use::Prev</tt>)</i>
2770 <!-- ______________________________________________________________________ -->
2772 <a name="Waymarking">The waymarking algorithm</a>
2777 Since the <tt>Use</tt> objects are deprived of the direct (back)pointer to
2778 their <tt>User</tt> objects, there must be a fast and exact method to
2779 recover it. This is accomplished by the following scheme:</p>
2781 A bit-encoding in the 2 LSBits (least significant bits) of the <tt>Use::Prev</tt> allows to find the
2782 start of the <tt>User</tt> object:
2784 <li><tt>00</tt> —> binary digit 0</li>
2785 <li><tt>01</tt> —> binary digit 1</li>
2786 <li><tt>10</tt> —> stop and calculate (<tt>s</tt>)</li>
2787 <li><tt>11</tt> —> full stop (<tt>S</tt>)</li>
2790 Given a <tt>Use*</tt>, all we have to do is to walk till we get
2791 a stop and we either have a <tt>User</tt> immediately behind or
2792 we have to walk to the next stop picking up digits
2793 and calculating the offset:</p>
2795 .---.---.---.---.---.---.---.---.---.---.---.---.---.---.---.---.----------------
2796 | 1 | s | 1 | 0 | 1 | 0 | s | 1 | 1 | 0 | s | 1 | 1 | s | 1 | S | User (or User*)
2797 '---'---'---'---'---'---'---'---'---'---'---'---'---'---'---'---'----------------
2798 |+15 |+10 |+6 |+3 |+1
2801 | | |______________________>
2802 | |______________________________________>
2803 |__________________________________________________________>
2806 Only the significant number of bits need to be stored between the
2807 stops, so that the <i>worst case is 20 memory accesses</i> when there are
2808 1000 <tt>Use</tt> objects associated with a <tt>User</tt>.</p>
2812 <!-- ______________________________________________________________________ -->
2814 <a name="ReferenceImpl">Reference implementation</a>
2819 The following literate Haskell fragment demonstrates the concept:</p>
2821 <div class="doc_code">
2823 > import Test.QuickCheck
2825 > digits :: Int -> [Char] -> [Char]
2826 > digits 0 acc = '0' : acc
2827 > digits 1 acc = '1' : acc
2828 > digits n acc = digits (n `div` 2) $ digits (n `mod` 2) acc
2830 > dist :: Int -> [Char] -> [Char]
2833 > dist 1 acc = let r = dist 0 acc in 's' : digits (length r) r
2834 > dist n acc = dist (n - 1) $ dist 1 acc
2836 > takeLast n ss = reverse $ take n $ reverse ss
2838 > test = takeLast 40 $ dist 20 []
2843 Printing <test> gives: <tt>"1s100000s11010s10100s1111s1010s110s11s1S"</tt></p>
2845 The reverse algorithm computes the length of the string just by examining
2846 a certain prefix:</p>
2848 <div class="doc_code">
2850 > pref :: [Char] -> Int
2852 > pref ('s':'1':rest) = decode 2 1 rest
2853 > pref (_:rest) = 1 + pref rest
2855 > decode walk acc ('0':rest) = decode (walk + 1) (acc * 2) rest
2856 > decode walk acc ('1':rest) = decode (walk + 1) (acc * 2 + 1) rest
2857 > decode walk acc _ = walk + acc
2862 Now, as expected, printing <pref test> gives <tt>40</tt>.</p>
2864 We can <i>quickCheck</i> this with following property:</p>
2866 <div class="doc_code">
2868 > testcase = dist 2000 []
2869 > testcaseLength = length testcase
2871 > identityProp n = n > 0 && n <= testcaseLength ==> length arr == pref arr
2872 > where arr = takeLast n testcase
2877 As expected <quickCheck identityProp> gives:</p>
2880 *Main> quickCheck identityProp
2881 OK, passed 100 tests.
2884 Let's be a bit more exhaustive:</p>
2886 <div class="doc_code">
2889 > deepCheck p = check (defaultConfig { configMaxTest = 500 }) p
2894 And here is the result of <deepCheck identityProp>:</p>
2897 *Main> deepCheck identityProp
2898 OK, passed 500 tests.
2903 <!-- ______________________________________________________________________ -->
2905 <a name="Tagging">Tagging considerations</a>
2911 To maintain the invariant that the 2 LSBits of each <tt>Use**</tt> in <tt>Use</tt>
2912 never change after being set up, setters of <tt>Use::Prev</tt> must re-tag the
2913 new <tt>Use**</tt> on every modification. Accordingly getters must strip the
2916 For layout b) instead of the <tt>User</tt> we find a pointer (<tt>User*</tt> with LSBit set).
2917 Following this pointer brings us to the <tt>User</tt>. A portable trick ensures
2918 that the first bytes of <tt>User</tt> (if interpreted as a pointer) never has
2919 the LSBit set. (Portability is relying on the fact that all known compilers place the
2920 <tt>vptr</tt> in the first word of the instances.)</p>
2928 <!-- *********************************************************************** -->
2930 <a name="coreclasses">The Core LLVM Class Hierarchy Reference </a>
2932 <!-- *********************************************************************** -->
2935 <p><tt>#include "<a href="/doxygen/Type_8h-source.html">llvm/Type.h</a>"</tt>
2936 <br>doxygen info: <a href="/doxygen/classllvm_1_1Type.html">Type Class</a></p>
2938 <p>The Core LLVM classes are the primary means of representing the program
2939 being inspected or transformed. The core LLVM classes are defined in
2940 header files in the <tt>include/llvm/</tt> directory, and implemented in
2941 the <tt>lib/VMCore</tt> directory.</p>
2943 <!-- ======================================================================= -->
2945 <a name="Type">The <tt>Type</tt> class and Derived Types</a>
2950 <p><tt>Type</tt> is a superclass of all type classes. Every <tt>Value</tt> has
2951 a <tt>Type</tt>. <tt>Type</tt> cannot be instantiated directly but only
2952 through its subclasses. Certain primitive types (<tt>VoidType</tt>,
2953 <tt>LabelType</tt>, <tt>FloatType</tt> and <tt>DoubleType</tt>) have hidden
2954 subclasses. They are hidden because they offer no useful functionality beyond
2955 what the <tt>Type</tt> class offers except to distinguish themselves from
2956 other subclasses of <tt>Type</tt>.</p>
2957 <p>All other types are subclasses of <tt>DerivedType</tt>. Types can be
2958 named, but this is not a requirement. There exists exactly
2959 one instance of a given shape at any one time. This allows type equality to
2960 be performed with address equality of the Type Instance. That is, given two
2961 <tt>Type*</tt> values, the types are identical if the pointers are identical.
2964 <!-- _______________________________________________________________________ -->
2966 <a name="m_Type">Important Public Methods</a>
2972 <li><tt>bool isIntegerTy() const</tt>: Returns true for any integer type.</li>
2974 <li><tt>bool isFloatingPointTy()</tt>: Return true if this is one of the five
2975 floating point types.</li>
2977 <li><tt>bool isSized()</tt>: Return true if the type has known size. Things
2978 that don't have a size are abstract types, labels and void.</li>
2983 <!-- _______________________________________________________________________ -->
2985 <a name="derivedtypes">Important Derived Types</a>
2989 <dt><tt>IntegerType</tt></dt>
2990 <dd>Subclass of DerivedType that represents integer types of any bit width.
2991 Any bit width between <tt>IntegerType::MIN_INT_BITS</tt> (1) and
2992 <tt>IntegerType::MAX_INT_BITS</tt> (~8 million) can be represented.
2994 <li><tt>static const IntegerType* get(unsigned NumBits)</tt>: get an integer
2995 type of a specific bit width.</li>
2996 <li><tt>unsigned getBitWidth() const</tt>: Get the bit width of an integer
3000 <dt><tt>SequentialType</tt></dt>
3001 <dd>This is subclassed by ArrayType, PointerType and VectorType.
3003 <li><tt>const Type * getElementType() const</tt>: Returns the type of each
3004 of the elements in the sequential type. </li>
3007 <dt><tt>ArrayType</tt></dt>
3008 <dd>This is a subclass of SequentialType and defines the interface for array
3011 <li><tt>unsigned getNumElements() const</tt>: Returns the number of
3012 elements in the array. </li>
3015 <dt><tt>PointerType</tt></dt>
3016 <dd>Subclass of SequentialType for pointer types.</dd>
3017 <dt><tt>VectorType</tt></dt>
3018 <dd>Subclass of SequentialType for vector types. A
3019 vector type is similar to an ArrayType but is distinguished because it is
3020 a first class type whereas ArrayType is not. Vector types are used for
3021 vector operations and are usually small vectors of of an integer or floating
3023 <dt><tt>StructType</tt></dt>
3024 <dd>Subclass of DerivedTypes for struct types.</dd>
3025 <dt><tt><a name="FunctionType">FunctionType</a></tt></dt>
3026 <dd>Subclass of DerivedTypes for function types.
3028 <li><tt>bool isVarArg() const</tt>: Returns true if it's a vararg
3030 <li><tt> const Type * getReturnType() const</tt>: Returns the
3031 return type of the function.</li>
3032 <li><tt>const Type * getParamType (unsigned i)</tt>: Returns
3033 the type of the ith parameter.</li>
3034 <li><tt> const unsigned getNumParams() const</tt>: Returns the
3035 number of formal parameters.</li>
3043 <!-- ======================================================================= -->
3045 <a name="Module">The <tt>Module</tt> class</a>
3051 href="/doxygen/Module_8h-source.html">llvm/Module.h</a>"</tt><br> doxygen info:
3052 <a href="/doxygen/classllvm_1_1Module.html">Module Class</a></p>
3054 <p>The <tt>Module</tt> class represents the top level structure present in LLVM
3055 programs. An LLVM module is effectively either a translation unit of the
3056 original program or a combination of several translation units merged by the
3057 linker. The <tt>Module</tt> class keeps track of a list of <a
3058 href="#Function"><tt>Function</tt></a>s, a list of <a
3059 href="#GlobalVariable"><tt>GlobalVariable</tt></a>s, and a <a
3060 href="#SymbolTable"><tt>SymbolTable</tt></a>. Additionally, it contains a few
3061 helpful member functions that try to make common operations easy.</p>
3063 <!-- _______________________________________________________________________ -->
3065 <a name="m_Module">Important Public Members of the <tt>Module</tt> class</a>
3071 <li><tt>Module::Module(std::string name = "")</tt></li>
3074 <p>Constructing a <a href="#Module">Module</a> is easy. You can optionally
3075 provide a name for it (probably based on the name of the translation unit).</p>
3078 <li><tt>Module::iterator</tt> - Typedef for function list iterator<br>
3079 <tt>Module::const_iterator</tt> - Typedef for const_iterator.<br>
3081 <tt>begin()</tt>, <tt>end()</tt>
3082 <tt>size()</tt>, <tt>empty()</tt>
3084 <p>These are forwarding methods that make it easy to access the contents of
3085 a <tt>Module</tt> object's <a href="#Function"><tt>Function</tt></a>
3088 <li><tt>Module::FunctionListType &getFunctionList()</tt>
3090 <p> Returns the list of <a href="#Function"><tt>Function</tt></a>s. This is
3091 necessary to use when you need to update the list or perform a complex
3092 action that doesn't have a forwarding method.</p>
3094 <p><!-- Global Variable --></p></li>
3100 <li><tt>Module::global_iterator</tt> - Typedef for global variable list iterator<br>
3102 <tt>Module::const_global_iterator</tt> - Typedef for const_iterator.<br>
3104 <tt>global_begin()</tt>, <tt>global_end()</tt>
3105 <tt>global_size()</tt>, <tt>global_empty()</tt>
3107 <p> These are forwarding methods that make it easy to access the contents of
3108 a <tt>Module</tt> object's <a
3109 href="#GlobalVariable"><tt>GlobalVariable</tt></a> list.</p></li>
3111 <li><tt>Module::GlobalListType &getGlobalList()</tt>
3113 <p>Returns the list of <a
3114 href="#GlobalVariable"><tt>GlobalVariable</tt></a>s. This is necessary to
3115 use when you need to update the list or perform a complex action that
3116 doesn't have a forwarding method.</p>
3118 <p><!-- Symbol table stuff --> </p></li>
3124 <li><tt><a href="#SymbolTable">SymbolTable</a> *getSymbolTable()</tt>
3126 <p>Return a reference to the <a href="#SymbolTable"><tt>SymbolTable</tt></a>
3127 for this <tt>Module</tt>.</p>
3129 <p><!-- Convenience methods --></p></li>
3135 <li><tt><a href="#Function">Function</a> *getFunction(const std::string
3136 &Name, const <a href="#FunctionType">FunctionType</a> *Ty)</tt>
3138 <p>Look up the specified function in the <tt>Module</tt> <a
3139 href="#SymbolTable"><tt>SymbolTable</tt></a>. If it does not exist, return
3140 <tt>null</tt>.</p></li>
3142 <li><tt><a href="#Function">Function</a> *getOrInsertFunction(const
3143 std::string &Name, const <a href="#FunctionType">FunctionType</a> *T)</tt>
3145 <p>Look up the specified function in the <tt>Module</tt> <a
3146 href="#SymbolTable"><tt>SymbolTable</tt></a>. If it does not exist, add an
3147 external declaration for the function and return it.</p></li>
3149 <li><tt>std::string getTypeName(const <a href="#Type">Type</a> *Ty)</tt>
3151 <p>If there is at least one entry in the <a
3152 href="#SymbolTable"><tt>SymbolTable</tt></a> for the specified <a
3153 href="#Type"><tt>Type</tt></a>, return it. Otherwise return the empty
3156 <li><tt>bool addTypeName(const std::string &Name, const <a
3157 href="#Type">Type</a> *Ty)</tt>
3159 <p>Insert an entry in the <a href="#SymbolTable"><tt>SymbolTable</tt></a>
3160 mapping <tt>Name</tt> to <tt>Ty</tt>. If there is already an entry for this
3161 name, true is returned and the <a
3162 href="#SymbolTable"><tt>SymbolTable</tt></a> is not modified.</p></li>
3169 <!-- ======================================================================= -->
3171 <a name="Value">The <tt>Value</tt> class</a>
3176 <p><tt>#include "<a href="/doxygen/Value_8h-source.html">llvm/Value.h</a>"</tt>
3178 doxygen info: <a href="/doxygen/classllvm_1_1Value.html">Value Class</a></p>
3180 <p>The <tt>Value</tt> class is the most important class in the LLVM Source
3181 base. It represents a typed value that may be used (among other things) as an
3182 operand to an instruction. There are many different types of <tt>Value</tt>s,
3183 such as <a href="#Constant"><tt>Constant</tt></a>s,<a
3184 href="#Argument"><tt>Argument</tt></a>s. Even <a
3185 href="#Instruction"><tt>Instruction</tt></a>s and <a
3186 href="#Function"><tt>Function</tt></a>s are <tt>Value</tt>s.</p>
3188 <p>A particular <tt>Value</tt> may be used many times in the LLVM representation
3189 for a program. For example, an incoming argument to a function (represented
3190 with an instance of the <a href="#Argument">Argument</a> class) is "used" by
3191 every instruction in the function that references the argument. To keep track
3192 of this relationship, the <tt>Value</tt> class keeps a list of all of the <a
3193 href="#User"><tt>User</tt></a>s that is using it (the <a
3194 href="#User"><tt>User</tt></a> class is a base class for all nodes in the LLVM
3195 graph that can refer to <tt>Value</tt>s). This use list is how LLVM represents
3196 def-use information in the program, and is accessible through the <tt>use_</tt>*
3197 methods, shown below.</p>
3199 <p>Because LLVM is a typed representation, every LLVM <tt>Value</tt> is typed,
3200 and this <a href="#Type">Type</a> is available through the <tt>getType()</tt>
3201 method. In addition, all LLVM values can be named. The "name" of the
3202 <tt>Value</tt> is a symbolic string printed in the LLVM code:</p>
3204 <div class="doc_code">
3206 %<b>foo</b> = add i32 1, 2
3210 <p><a name="nameWarning">The name of this instruction is "foo".</a> <b>NOTE</b>
3211 that the name of any value may be missing (an empty string), so names should
3212 <b>ONLY</b> be used for debugging (making the source code easier to read,
3213 debugging printouts), they should not be used to keep track of values or map
3214 between them. For this purpose, use a <tt>std::map</tt> of pointers to the
3215 <tt>Value</tt> itself instead.</p>
3217 <p>One important aspect of LLVM is that there is no distinction between an SSA
3218 variable and the operation that produces it. Because of this, any reference to
3219 the value produced by an instruction (or the value available as an incoming
3220 argument, for example) is represented as a direct pointer to the instance of
3222 represents this value. Although this may take some getting used to, it
3223 simplifies the representation and makes it easier to manipulate.</p>
3225 <!-- _______________________________________________________________________ -->
3227 <a name="m_Value">Important Public Members of the <tt>Value</tt> class</a>
3233 <li><tt>Value::use_iterator</tt> - Typedef for iterator over the
3235 <tt>Value::const_use_iterator</tt> - Typedef for const_iterator over
3237 <tt>unsigned use_size()</tt> - Returns the number of users of the
3239 <tt>bool use_empty()</tt> - Returns true if there are no users.<br>
3240 <tt>use_iterator use_begin()</tt> - Get an iterator to the start of
3242 <tt>use_iterator use_end()</tt> - Get an iterator to the end of the
3244 <tt><a href="#User">User</a> *use_back()</tt> - Returns the last
3245 element in the list.
3246 <p> These methods are the interface to access the def-use
3247 information in LLVM. As with all other iterators in LLVM, the naming
3248 conventions follow the conventions defined by the <a href="#stl">STL</a>.</p>
3250 <li><tt><a href="#Type">Type</a> *getType() const</tt>
3251 <p>This method returns the Type of the Value.</p>
3253 <li><tt>bool hasName() const</tt><br>
3254 <tt>std::string getName() const</tt><br>
3255 <tt>void setName(const std::string &Name)</tt>
3256 <p> This family of methods is used to access and assign a name to a <tt>Value</tt>,
3257 be aware of the <a href="#nameWarning">precaution above</a>.</p>
3259 <li><tt>void replaceAllUsesWith(Value *V)</tt>
3261 <p>This method traverses the use list of a <tt>Value</tt> changing all <a
3262 href="#User"><tt>User</tt>s</a> of the current value to refer to
3263 "<tt>V</tt>" instead. For example, if you detect that an instruction always
3264 produces a constant value (for example through constant folding), you can
3265 replace all uses of the instruction with the constant like this:</p>
3267 <div class="doc_code">
3269 Inst->replaceAllUsesWith(ConstVal);
3279 <!-- ======================================================================= -->
3281 <a name="User">The <tt>User</tt> class</a>
3287 <tt>#include "<a href="/doxygen/User_8h-source.html">llvm/User.h</a>"</tt><br>
3288 doxygen info: <a href="/doxygen/classllvm_1_1User.html">User Class</a><br>
3289 Superclass: <a href="#Value"><tt>Value</tt></a></p>
3291 <p>The <tt>User</tt> class is the common base class of all LLVM nodes that may
3292 refer to <a href="#Value"><tt>Value</tt></a>s. It exposes a list of "Operands"
3293 that are all of the <a href="#Value"><tt>Value</tt></a>s that the User is
3294 referring to. The <tt>User</tt> class itself is a subclass of
3297 <p>The operands of a <tt>User</tt> point directly to the LLVM <a
3298 href="#Value"><tt>Value</tt></a> that it refers to. Because LLVM uses Static
3299 Single Assignment (SSA) form, there can only be one definition referred to,
3300 allowing this direct connection. This connection provides the use-def
3301 information in LLVM.</p>
3303 <!-- _______________________________________________________________________ -->
3305 <a name="m_User">Important Public Members of the <tt>User</tt> class</a>
3310 <p>The <tt>User</tt> class exposes the operand list in two ways: through
3311 an index access interface and through an iterator based interface.</p>
3314 <li><tt>Value *getOperand(unsigned i)</tt><br>
3315 <tt>unsigned getNumOperands()</tt>
3316 <p> These two methods expose the operands of the <tt>User</tt> in a
3317 convenient form for direct access.</p></li>
3319 <li><tt>User::op_iterator</tt> - Typedef for iterator over the operand
3321 <tt>op_iterator op_begin()</tt> - Get an iterator to the start of
3322 the operand list.<br>
3323 <tt>op_iterator op_end()</tt> - Get an iterator to the end of the
3325 <p> Together, these methods make up the iterator based interface to
3326 the operands of a <tt>User</tt>.</p></li>
3333 <!-- ======================================================================= -->
3335 <a name="Instruction">The <tt>Instruction</tt> class</a>
3340 <p><tt>#include "</tt><tt><a
3341 href="/doxygen/Instruction_8h-source.html">llvm/Instruction.h</a>"</tt><br>
3342 doxygen info: <a href="/doxygen/classllvm_1_1Instruction.html">Instruction Class</a><br>
3343 Superclasses: <a href="#User"><tt>User</tt></a>, <a
3344 href="#Value"><tt>Value</tt></a></p>
3346 <p>The <tt>Instruction</tt> class is the common base class for all LLVM
3347 instructions. It provides only a few methods, but is a very commonly used
3348 class. The primary data tracked by the <tt>Instruction</tt> class itself is the
3349 opcode (instruction type) and the parent <a
3350 href="#BasicBlock"><tt>BasicBlock</tt></a> the <tt>Instruction</tt> is embedded
3351 into. To represent a specific type of instruction, one of many subclasses of
3352 <tt>Instruction</tt> are used.</p>
3354 <p> Because the <tt>Instruction</tt> class subclasses the <a
3355 href="#User"><tt>User</tt></a> class, its operands can be accessed in the same
3356 way as for other <a href="#User"><tt>User</tt></a>s (with the
3357 <tt>getOperand()</tt>/<tt>getNumOperands()</tt> and
3358 <tt>op_begin()</tt>/<tt>op_end()</tt> methods).</p> <p> An important file for
3359 the <tt>Instruction</tt> class is the <tt>llvm/Instruction.def</tt> file. This
3360 file contains some meta-data about the various different types of instructions
3361 in LLVM. It describes the enum values that are used as opcodes (for example
3362 <tt>Instruction::Add</tt> and <tt>Instruction::ICmp</tt>), as well as the
3363 concrete sub-classes of <tt>Instruction</tt> that implement the instruction (for
3364 example <tt><a href="#BinaryOperator">BinaryOperator</a></tt> and <tt><a
3365 href="#CmpInst">CmpInst</a></tt>). Unfortunately, the use of macros in
3366 this file confuses doxygen, so these enum values don't show up correctly in the
3367 <a href="/doxygen/classllvm_1_1Instruction.html">doxygen output</a>.</p>
3369 <!-- _______________________________________________________________________ -->
3371 <a name="s_Instruction">
3372 Important Subclasses of the <tt>Instruction</tt> class
3377 <li><tt><a name="BinaryOperator">BinaryOperator</a></tt>
3378 <p>This subclasses represents all two operand instructions whose operands
3379 must be the same type, except for the comparison instructions.</p></li>
3380 <li><tt><a name="CastInst">CastInst</a></tt>
3381 <p>This subclass is the parent of the 12 casting instructions. It provides
3382 common operations on cast instructions.</p>
3383 <li><tt><a name="CmpInst">CmpInst</a></tt>
3384 <p>This subclass respresents the two comparison instructions,
3385 <a href="LangRef.html#i_icmp">ICmpInst</a> (integer opreands), and
3386 <a href="LangRef.html#i_fcmp">FCmpInst</a> (floating point operands).</p>
3387 <li><tt><a name="TerminatorInst">TerminatorInst</a></tt>
3388 <p>This subclass is the parent of all terminator instructions (those which
3389 can terminate a block).</p>
3393 <!-- _______________________________________________________________________ -->
3395 <a name="m_Instruction">
3396 Important Public Members of the <tt>Instruction</tt> class
3403 <li><tt><a href="#BasicBlock">BasicBlock</a> *getParent()</tt>
3404 <p>Returns the <a href="#BasicBlock"><tt>BasicBlock</tt></a> that
3405 this <tt>Instruction</tt> is embedded into.</p></li>
3406 <li><tt>bool mayWriteToMemory()</tt>
3407 <p>Returns true if the instruction writes to memory, i.e. it is a
3408 <tt>call</tt>,<tt>free</tt>,<tt>invoke</tt>, or <tt>store</tt>.</p></li>
3409 <li><tt>unsigned getOpcode()</tt>
3410 <p>Returns the opcode for the <tt>Instruction</tt>.</p></li>
3411 <li><tt><a href="#Instruction">Instruction</a> *clone() const</tt>
3412 <p>Returns another instance of the specified instruction, identical
3413 in all ways to the original except that the instruction has no parent
3414 (ie it's not embedded into a <a href="#BasicBlock"><tt>BasicBlock</tt></a>),
3415 and it has no name</p></li>
3422 <!-- ======================================================================= -->
3424 <a name="Constant">The <tt>Constant</tt> class and subclasses</a>
3429 <p>Constant represents a base class for different types of constants. It
3430 is subclassed by ConstantInt, ConstantArray, etc. for representing
3431 the various types of Constants. <a href="#GlobalValue">GlobalValue</a> is also
3432 a subclass, which represents the address of a global variable or function.
3435 <!-- _______________________________________________________________________ -->
3436 <h4>Important Subclasses of Constant</h4>
3439 <li>ConstantInt : This subclass of Constant represents an integer constant of
3442 <li><tt>const APInt& getValue() const</tt>: Returns the underlying
3443 value of this constant, an APInt value.</li>
3444 <li><tt>int64_t getSExtValue() const</tt>: Converts the underlying APInt
3445 value to an int64_t via sign extension. If the value (not the bit width)
3446 of the APInt is too large to fit in an int64_t, an assertion will result.
3447 For this reason, use of this method is discouraged.</li>
3448 <li><tt>uint64_t getZExtValue() const</tt>: Converts the underlying APInt
3449 value to a uint64_t via zero extension. IF the value (not the bit width)
3450 of the APInt is too large to fit in a uint64_t, an assertion will result.
3451 For this reason, use of this method is discouraged.</li>
3452 <li><tt>static ConstantInt* get(const APInt& Val)</tt>: Returns the
3453 ConstantInt object that represents the value provided by <tt>Val</tt>.
3454 The type is implied as the IntegerType that corresponds to the bit width
3455 of <tt>Val</tt>.</li>
3456 <li><tt>static ConstantInt* get(const Type *Ty, uint64_t Val)</tt>:
3457 Returns the ConstantInt object that represents the value provided by
3458 <tt>Val</tt> for integer type <tt>Ty</tt>.</li>
3461 <li>ConstantFP : This class represents a floating point constant.
3463 <li><tt>double getValue() const</tt>: Returns the underlying value of
3464 this constant. </li>
3467 <li>ConstantArray : This represents a constant array.
3469 <li><tt>const std::vector<Use> &getValues() const</tt>: Returns
3470 a vector of component constants that makeup this array. </li>
3473 <li>ConstantStruct : This represents a constant struct.
3475 <li><tt>const std::vector<Use> &getValues() const</tt>: Returns
3476 a vector of component constants that makeup this array. </li>
3479 <li>GlobalValue : This represents either a global variable or a function. In
3480 either case, the value is a constant fixed address (after linking).
3487 <!-- ======================================================================= -->
3489 <a name="GlobalValue">The <tt>GlobalValue</tt> class</a>
3495 href="/doxygen/GlobalValue_8h-source.html">llvm/GlobalValue.h</a>"</tt><br>
3496 doxygen info: <a href="/doxygen/classllvm_1_1GlobalValue.html">GlobalValue
3498 Superclasses: <a href="#Constant"><tt>Constant</tt></a>,
3499 <a href="#User"><tt>User</tt></a>, <a href="#Value"><tt>Value</tt></a></p>
3501 <p>Global values (<a href="#GlobalVariable"><tt>GlobalVariable</tt></a>s or <a
3502 href="#Function"><tt>Function</tt></a>s) are the only LLVM values that are
3503 visible in the bodies of all <a href="#Function"><tt>Function</tt></a>s.
3504 Because they are visible at global scope, they are also subject to linking with
3505 other globals defined in different translation units. To control the linking
3506 process, <tt>GlobalValue</tt>s know their linkage rules. Specifically,
3507 <tt>GlobalValue</tt>s know whether they have internal or external linkage, as
3508 defined by the <tt>LinkageTypes</tt> enumeration.</p>
3510 <p>If a <tt>GlobalValue</tt> has internal linkage (equivalent to being
3511 <tt>static</tt> in C), it is not visible to code outside the current translation
3512 unit, and does not participate in linking. If it has external linkage, it is
3513 visible to external code, and does participate in linking. In addition to
3514 linkage information, <tt>GlobalValue</tt>s keep track of which <a
3515 href="#Module"><tt>Module</tt></a> they are currently part of.</p>
3517 <p>Because <tt>GlobalValue</tt>s are memory objects, they are always referred to
3518 by their <b>address</b>. As such, the <a href="#Type"><tt>Type</tt></a> of a
3519 global is always a pointer to its contents. It is important to remember this
3520 when using the <tt>GetElementPtrInst</tt> instruction because this pointer must
3521 be dereferenced first. For example, if you have a <tt>GlobalVariable</tt> (a
3522 subclass of <tt>GlobalValue)</tt> that is an array of 24 ints, type <tt>[24 x
3523 i32]</tt>, then the <tt>GlobalVariable</tt> is a pointer to that array. Although
3524 the address of the first element of this array and the value of the
3525 <tt>GlobalVariable</tt> are the same, they have different types. The
3526 <tt>GlobalVariable</tt>'s type is <tt>[24 x i32]</tt>. The first element's type
3527 is <tt>i32.</tt> Because of this, accessing a global value requires you to
3528 dereference the pointer with <tt>GetElementPtrInst</tt> first, then its elements
3529 can be accessed. This is explained in the <a href="LangRef.html#globalvars">LLVM
3530 Language Reference Manual</a>.</p>
3532 <!-- _______________________________________________________________________ -->
3534 <a name="m_GlobalValue">
3535 Important Public Members of the <tt>GlobalValue</tt> class
3542 <li><tt>bool hasInternalLinkage() const</tt><br>
3543 <tt>bool hasExternalLinkage() const</tt><br>
3544 <tt>void setInternalLinkage(bool HasInternalLinkage)</tt>
3545 <p> These methods manipulate the linkage characteristics of the <tt>GlobalValue</tt>.</p>
3548 <li><tt><a href="#Module">Module</a> *getParent()</tt>
3549 <p> This returns the <a href="#Module"><tt>Module</tt></a> that the
3550 GlobalValue is currently embedded into.</p></li>
3557 <!-- ======================================================================= -->
3559 <a name="Function">The <tt>Function</tt> class</a>
3565 href="/doxygen/Function_8h-source.html">llvm/Function.h</a>"</tt><br> doxygen
3566 info: <a href="/doxygen/classllvm_1_1Function.html">Function Class</a><br>
3567 Superclasses: <a href="#GlobalValue"><tt>GlobalValue</tt></a>,
3568 <a href="#Constant"><tt>Constant</tt></a>,
3569 <a href="#User"><tt>User</tt></a>,
3570 <a href="#Value"><tt>Value</tt></a></p>
3572 <p>The <tt>Function</tt> class represents a single procedure in LLVM. It is
3573 actually one of the more complex classes in the LLVM hierarchy because it must
3574 keep track of a large amount of data. The <tt>Function</tt> class keeps track
3575 of a list of <a href="#BasicBlock"><tt>BasicBlock</tt></a>s, a list of formal
3576 <a href="#Argument"><tt>Argument</tt></a>s, and a
3577 <a href="#SymbolTable"><tt>SymbolTable</tt></a>.</p>
3579 <p>The list of <a href="#BasicBlock"><tt>BasicBlock</tt></a>s is the most
3580 commonly used part of <tt>Function</tt> objects. The list imposes an implicit
3581 ordering of the blocks in the function, which indicate how the code will be
3582 laid out by the backend. Additionally, the first <a
3583 href="#BasicBlock"><tt>BasicBlock</tt></a> is the implicit entry node for the
3584 <tt>Function</tt>. It is not legal in LLVM to explicitly branch to this initial
3585 block. There are no implicit exit nodes, and in fact there may be multiple exit
3586 nodes from a single <tt>Function</tt>. If the <a
3587 href="#BasicBlock"><tt>BasicBlock</tt></a> list is empty, this indicates that
3588 the <tt>Function</tt> is actually a function declaration: the actual body of the
3589 function hasn't been linked in yet.</p>
3591 <p>In addition to a list of <a href="#BasicBlock"><tt>BasicBlock</tt></a>s, the
3592 <tt>Function</tt> class also keeps track of the list of formal <a
3593 href="#Argument"><tt>Argument</tt></a>s that the function receives. This
3594 container manages the lifetime of the <a href="#Argument"><tt>Argument</tt></a>
3595 nodes, just like the <a href="#BasicBlock"><tt>BasicBlock</tt></a> list does for
3596 the <a href="#BasicBlock"><tt>BasicBlock</tt></a>s.</p>
3598 <p>The <a href="#SymbolTable"><tt>SymbolTable</tt></a> is a very rarely used
3599 LLVM feature that is only used when you have to look up a value by name. Aside
3600 from that, the <a href="#SymbolTable"><tt>SymbolTable</tt></a> is used
3601 internally to make sure that there are not conflicts between the names of <a
3602 href="#Instruction"><tt>Instruction</tt></a>s, <a
3603 href="#BasicBlock"><tt>BasicBlock</tt></a>s, or <a
3604 href="#Argument"><tt>Argument</tt></a>s in the function body.</p>
3606 <p>Note that <tt>Function</tt> is a <a href="#GlobalValue">GlobalValue</a>
3607 and therefore also a <a href="#Constant">Constant</a>. The value of the function
3608 is its address (after linking) which is guaranteed to be constant.</p>
3610 <!-- _______________________________________________________________________ -->
3612 <a name="m_Function">
3613 Important Public Members of the <tt>Function</tt> class
3620 <li><tt>Function(const </tt><tt><a href="#FunctionType">FunctionType</a>
3621 *Ty, LinkageTypes Linkage, const std::string &N = "", Module* Parent = 0)</tt>
3623 <p>Constructor used when you need to create new <tt>Function</tt>s to add
3624 the the program. The constructor must specify the type of the function to
3625 create and what type of linkage the function should have. The <a
3626 href="#FunctionType"><tt>FunctionType</tt></a> argument
3627 specifies the formal arguments and return value for the function. The same
3628 <a href="#FunctionType"><tt>FunctionType</tt></a> value can be used to
3629 create multiple functions. The <tt>Parent</tt> argument specifies the Module
3630 in which the function is defined. If this argument is provided, the function
3631 will automatically be inserted into that module's list of
3634 <li><tt>bool isDeclaration()</tt>
3636 <p>Return whether or not the <tt>Function</tt> has a body defined. If the
3637 function is "external", it does not have a body, and thus must be resolved
3638 by linking with a function defined in a different translation unit.</p></li>
3640 <li><tt>Function::iterator</tt> - Typedef for basic block list iterator<br>
3641 <tt>Function::const_iterator</tt> - Typedef for const_iterator.<br>
3643 <tt>begin()</tt>, <tt>end()</tt>
3644 <tt>size()</tt>, <tt>empty()</tt>
3646 <p>These are forwarding methods that make it easy to access the contents of
3647 a <tt>Function</tt> object's <a href="#BasicBlock"><tt>BasicBlock</tt></a>
3650 <li><tt>Function::BasicBlockListType &getBasicBlockList()</tt>
3652 <p>Returns the list of <a href="#BasicBlock"><tt>BasicBlock</tt></a>s. This
3653 is necessary to use when you need to update the list or perform a complex
3654 action that doesn't have a forwarding method.</p></li>
3656 <li><tt>Function::arg_iterator</tt> - Typedef for the argument list
3658 <tt>Function::const_arg_iterator</tt> - Typedef for const_iterator.<br>
3660 <tt>arg_begin()</tt>, <tt>arg_end()</tt>
3661 <tt>arg_size()</tt>, <tt>arg_empty()</tt>
3663 <p>These are forwarding methods that make it easy to access the contents of
3664 a <tt>Function</tt> object's <a href="#Argument"><tt>Argument</tt></a>
3667 <li><tt>Function::ArgumentListType &getArgumentList()</tt>
3669 <p>Returns the list of <a href="#Argument"><tt>Argument</tt></a>s. This is
3670 necessary to use when you need to update the list or perform a complex
3671 action that doesn't have a forwarding method.</p></li>
3673 <li><tt><a href="#BasicBlock">BasicBlock</a> &getEntryBlock()</tt>
3675 <p>Returns the entry <a href="#BasicBlock"><tt>BasicBlock</tt></a> for the
3676 function. Because the entry block for the function is always the first
3677 block, this returns the first block of the <tt>Function</tt>.</p></li>
3679 <li><tt><a href="#Type">Type</a> *getReturnType()</tt><br>
3680 <tt><a href="#FunctionType">FunctionType</a> *getFunctionType()</tt>
3682 <p>This traverses the <a href="#Type"><tt>Type</tt></a> of the
3683 <tt>Function</tt> and returns the return type of the function, or the <a
3684 href="#FunctionType"><tt>FunctionType</tt></a> of the actual
3687 <li><tt><a href="#SymbolTable">SymbolTable</a> *getSymbolTable()</tt>
3689 <p> Return a pointer to the <a href="#SymbolTable"><tt>SymbolTable</tt></a>
3690 for this <tt>Function</tt>.</p></li>
3697 <!-- ======================================================================= -->
3699 <a name="GlobalVariable">The <tt>GlobalVariable</tt> class</a>
3705 href="/doxygen/GlobalVariable_8h-source.html">llvm/GlobalVariable.h</a>"</tt>
3707 doxygen info: <a href="/doxygen/classllvm_1_1GlobalVariable.html">GlobalVariable
3709 Superclasses: <a href="#GlobalValue"><tt>GlobalValue</tt></a>,
3710 <a href="#Constant"><tt>Constant</tt></a>,
3711 <a href="#User"><tt>User</tt></a>,
3712 <a href="#Value"><tt>Value</tt></a></p>
3714 <p>Global variables are represented with the (surprise surprise)
3715 <tt>GlobalVariable</tt> class. Like functions, <tt>GlobalVariable</tt>s are also
3716 subclasses of <a href="#GlobalValue"><tt>GlobalValue</tt></a>, and as such are
3717 always referenced by their address (global values must live in memory, so their
3718 "name" refers to their constant address). See
3719 <a href="#GlobalValue"><tt>GlobalValue</tt></a> for more on this. Global
3720 variables may have an initial value (which must be a
3721 <a href="#Constant"><tt>Constant</tt></a>), and if they have an initializer,
3722 they may be marked as "constant" themselves (indicating that their contents
3723 never change at runtime).</p>
3725 <!-- _______________________________________________________________________ -->
3727 <a name="m_GlobalVariable">
3728 Important Public Members of the <tt>GlobalVariable</tt> class
3735 <li><tt>GlobalVariable(const </tt><tt><a href="#Type">Type</a> *Ty, bool
3736 isConstant, LinkageTypes& Linkage, <a href="#Constant">Constant</a>
3737 *Initializer = 0, const std::string &Name = "", Module* Parent = 0)</tt>
3739 <p>Create a new global variable of the specified type. If
3740 <tt>isConstant</tt> is true then the global variable will be marked as
3741 unchanging for the program. The Linkage parameter specifies the type of
3742 linkage (internal, external, weak, linkonce, appending) for the variable.
3743 If the linkage is InternalLinkage, WeakAnyLinkage, WeakODRLinkage,
3744 LinkOnceAnyLinkage or LinkOnceODRLinkage, then the resultant
3745 global variable will have internal linkage. AppendingLinkage concatenates
3746 together all instances (in different translation units) of the variable
3747 into a single variable but is only applicable to arrays. See
3748 the <a href="LangRef.html#modulestructure">LLVM Language Reference</a> for
3749 further details on linkage types. Optionally an initializer, a name, and the
3750 module to put the variable into may be specified for the global variable as
3753 <li><tt>bool isConstant() const</tt>
3755 <p>Returns true if this is a global variable that is known not to
3756 be modified at runtime.</p></li>
3758 <li><tt>bool hasInitializer()</tt>
3760 <p>Returns true if this <tt>GlobalVariable</tt> has an intializer.</p></li>
3762 <li><tt><a href="#Constant">Constant</a> *getInitializer()</tt>
3764 <p>Returns the initial value for a <tt>GlobalVariable</tt>. It is not legal
3765 to call this method if there is no initializer.</p></li>
3772 <!-- ======================================================================= -->
3774 <a name="BasicBlock">The <tt>BasicBlock</tt> class</a>
3780 href="/doxygen/BasicBlock_8h-source.html">llvm/BasicBlock.h</a>"</tt><br>
3781 doxygen info: <a href="/doxygen/classllvm_1_1BasicBlock.html">BasicBlock
3783 Superclass: <a href="#Value"><tt>Value</tt></a></p>
3785 <p>This class represents a single entry single exit section of the code,
3786 commonly known as a basic block by the compiler community. The
3787 <tt>BasicBlock</tt> class maintains a list of <a
3788 href="#Instruction"><tt>Instruction</tt></a>s, which form the body of the block.
3789 Matching the language definition, the last element of this list of instructions
3790 is always a terminator instruction (a subclass of the <a
3791 href="#TerminatorInst"><tt>TerminatorInst</tt></a> class).</p>
3793 <p>In addition to tracking the list of instructions that make up the block, the
3794 <tt>BasicBlock</tt> class also keeps track of the <a
3795 href="#Function"><tt>Function</tt></a> that it is embedded into.</p>
3797 <p>Note that <tt>BasicBlock</tt>s themselves are <a
3798 href="#Value"><tt>Value</tt></a>s, because they are referenced by instructions
3799 like branches and can go in the switch tables. <tt>BasicBlock</tt>s have type
3802 <!-- _______________________________________________________________________ -->
3804 <a name="m_BasicBlock">
3805 Important Public Members of the <tt>BasicBlock</tt> class
3812 <li><tt>BasicBlock(const std::string &Name = "", </tt><tt><a
3813 href="#Function">Function</a> *Parent = 0)</tt>
3815 <p>The <tt>BasicBlock</tt> constructor is used to create new basic blocks for
3816 insertion into a function. The constructor optionally takes a name for the new
3817 block, and a <a href="#Function"><tt>Function</tt></a> to insert it into. If
3818 the <tt>Parent</tt> parameter is specified, the new <tt>BasicBlock</tt> is
3819 automatically inserted at the end of the specified <a
3820 href="#Function"><tt>Function</tt></a>, if not specified, the BasicBlock must be
3821 manually inserted into the <a href="#Function"><tt>Function</tt></a>.</p></li>
3823 <li><tt>BasicBlock::iterator</tt> - Typedef for instruction list iterator<br>
3824 <tt>BasicBlock::const_iterator</tt> - Typedef for const_iterator.<br>
3825 <tt>begin()</tt>, <tt>end()</tt>, <tt>front()</tt>, <tt>back()</tt>,
3826 <tt>size()</tt>, <tt>empty()</tt>
3827 STL-style functions for accessing the instruction list.
3829 <p>These methods and typedefs are forwarding functions that have the same
3830 semantics as the standard library methods of the same names. These methods
3831 expose the underlying instruction list of a basic block in a way that is easy to
3832 manipulate. To get the full complement of container operations (including
3833 operations to update the list), you must use the <tt>getInstList()</tt>
3836 <li><tt>BasicBlock::InstListType &getInstList()</tt>
3838 <p>This method is used to get access to the underlying container that actually
3839 holds the Instructions. This method must be used when there isn't a forwarding
3840 function in the <tt>BasicBlock</tt> class for the operation that you would like
3841 to perform. Because there are no forwarding functions for "updating"
3842 operations, you need to use this if you want to update the contents of a
3843 <tt>BasicBlock</tt>.</p></li>
3845 <li><tt><a href="#Function">Function</a> *getParent()</tt>
3847 <p> Returns a pointer to <a href="#Function"><tt>Function</tt></a> the block is
3848 embedded into, or a null pointer if it is homeless.</p></li>
3850 <li><tt><a href="#TerminatorInst">TerminatorInst</a> *getTerminator()</tt>
3852 <p> Returns a pointer to the terminator instruction that appears at the end of
3853 the <tt>BasicBlock</tt>. If there is no terminator instruction, or if the last
3854 instruction in the block is not a terminator, then a null pointer is
3863 <!-- ======================================================================= -->
3865 <a name="Argument">The <tt>Argument</tt> class</a>
3870 <p>This subclass of Value defines the interface for incoming formal
3871 arguments to a function. A Function maintains a list of its formal
3872 arguments. An argument has a pointer to the parent Function.</p>
3878 <!-- *********************************************************************** -->
3881 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
3882 src="http://jigsaw.w3.org/css-validator/images/vcss-blue" alt="Valid CSS"></a>
3883 <a href="http://validator.w3.org/check/referer"><img
3884 src="http://www.w3.org/Icons/valid-html401" alt="Valid HTML 4.01 Strict"></a>
3886 <a href="mailto:dhurjati@cs.uiuc.edu">Dinakar Dhurjati</a> and
3887 <a href="mailto:sabre@nondot.org">Chris Lattner</a><br>
3888 <a href="http://llvm.org/">The LLVM Compiler Infrastructure</a><br>
3889 Last modified: $Date$