1 <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
2 <html><head><title>LLVM Programmer's Manual</title></head>
6 <table width="100%" bgcolor="#330077" border=0 cellpadding=4 cellspacing=0>
7 <tr><td> <font size=+3 color="#EEEEFF" face="Georgia,Palatino,Times,Roman"><b>LLVM Programmer's Manual</b></font></td>
11 <li><a href="#introduction">Introduction</a>
12 <li><a href="#general">General Information</a>
14 <li><a href="#stl">The C++ Standard Template Library</a>
16 <li>The <tt>-time-passes</tt> option
17 <li>How to use the LLVM Makefile system
18 <li>How to write a regression test
21 <li><a href="#apis">Important and useful LLVM APIs</a>
23 <li><a href="#isa">The <tt>isa<></tt>, <tt>cast<></tt> and
24 <tt>dyn_cast<></tt> templates</a>
25 <li><a href="#DEBUG">The <tt>DEBUG()</tt> macro &
26 <tt>-debug</tt> option</a>
27 <li><a href="#Statistic">The <tt>Statistic</tt> template &
28 <tt>-stats</tt> option</a>
30 <li>The <tt>InstVisitor</tt> template
31 <li>The general graph API
34 <li><a href="#common">Helpful Hints for Common Operations</a>
36 <li><a href="#inspection">Basic Inspection and Traversal Routines</a>
38 <li><a href="#iterate_function">Iterating over the <tt>BasicBlock</tt>s
39 in a <tt>Function</tt></a>
40 <li><a href="#iterate_basicblock">Iterating over the <tt>Instruction</tt>s
41 in a <tt>BasicBlock</tt></a>
42 <li><a href="#iterate_institer">Iterating over the <tt>Instruction</tt>s
43 in a <tt>Function</tt></a>
44 <li><a href="#iterate_convert">Turning an iterator into a class
46 <li><a href="#iterate_complex">Finding call sites: a more complex
48 <li><a href="#iterate_chains">Iterating over def-use & use-def
51 <li><a href="#simplechanges">Making simple changes</a>
53 <li><a href="#schanges_creating">Creating and inserting new
54 <tt>Instruction</tt>s</a>
55 <li><a href="#schanges_deleting">Deleting
56 <tt>Instruction</tt>s</a>
57 <li><a href="#schanges_replacing">Replacing an
58 <tt>Instruction</tt> with another <tt>Value</tt></a>
61 <li>Working with the Control Flow Graph
63 <li>Accessing predecessors and successors of a <tt>BasicBlock</tt>
69 <li><a href="#coreclasses">The Core LLVM Class Hierarchy Reference</a>
71 <li><a href="#Value">The <tt>Value</tt> class</a>
73 <li><a href="#User">The <tt>User</tt> class</a>
75 <li><a href="#Instruction">The <tt>Instruction</tt> class</a>
79 <li><a href="#GlobalValue">The <tt>GlobalValue</tt> class</a>
81 <li><a href="#BasicBlock">The <tt>BasicBlock</tt> class</a>
82 <li><a href="#Function">The <tt>Function</tt> class</a>
83 <li><a href="#GlobalVariable">The <tt>GlobalVariable</tt> class</a>
85 <li><a href="#Module">The <tt>Module</tt> class</a>
86 <li><a href="#Constant">The <tt>Constant</tt> class</a>
92 <li><a href="#Type">The <tt>Type</tt> class</a>
93 <li><a href="#Argument">The <tt>Argument</tt> class</a>
95 <li>The <tt>SymbolTable</tt> class
96 <li>The <tt>ilist</tt> and <tt>iplist</tt> classes
98 <li>Creating, inserting, moving and deleting from LLVM lists
100 <li>Important iterator invalidation semantics to be aware of
103 <p><b>Written by <a href="mailto:sabre@nondot.org">Chris Lattner</a>,
104 <a href="mailto:dhurjati@cs.uiuc.edu">Dinakar Dhurjati</a>, and
105 <a href="mailto:jstanley@cs.uiuc.edu">Joel Stanley</a></b><p>
109 <!-- *********************************************************************** -->
110 <table width="100%" bgcolor="#330077" border=0 cellpadding=4 cellspacing=0>
111 <tr><td align=center><font color="#EEEEFF" size=+2 face="Georgia,Palatino"><b>
112 <a name="introduction">Introduction
113 </b></font></td></tr></table><ul>
114 <!-- *********************************************************************** -->
116 This document is meant to highlight some of the important classes and interfaces
117 available in the LLVM source-base. This manual is not intended to explain what
118 LLVM is, how it works, and what LLVM code looks like. It assumes that you know
119 the basics of LLVM and are interested in writing transformations or otherwise
120 analyzing or manipulating the code.<p>
122 This document should get you oriented so that you can find your way in the
123 continuously growing source code that makes up the LLVM infrastructure. Note
124 that this manual is not intended to serve as a replacement for reading the
125 source code, so if you think there should be a method in one of these classes to
126 do something, but it's not listed, check the source. Links to the <a
127 href="/doxygen/">doxygen</a> sources are provided to make this as easy as
130 The first section of this document describes general information that is useful
131 to know when working in the LLVM infrastructure, and the second describes the
132 Core LLVM classes. In the future this manual will be extended with information
133 describing how to use extension libraries, such as dominator information, CFG
134 traversal routines, and useful utilities like the <tt><a
135 href="/doxygen/InstVisitor_8h-source.html">InstVisitor</a></tt> template.<p>
138 <!-- *********************************************************************** -->
139 </ul><table width="100%" bgcolor="#330077" border=0 cellpadding=4 cellspacing=0>
140 <tr><td align=center><font color="#EEEEFF" size=+2 face="Georgia,Palatino"><b>
141 <a name="general">General Information
142 </b></font></td></tr></table><ul>
143 <!-- *********************************************************************** -->
145 This section contains general information that is useful if you are working in
146 the LLVM source-base, but that isn't specific to any particular API.<p>
149 <!-- ======================================================================= -->
150 </ul><table width="100%" bgcolor="#441188" border=0 cellpadding=4 cellspacing=0>
151 <tr><td> </td><td width="100%">
152 <font color="#EEEEFF" face="Georgia,Palatino"><b>
153 <a name="stl">The C++ Standard Template Library</a>
154 </b></font></td></tr></table><ul>
156 LLVM makes heavy use of the C++ Standard Template Library (STL), perhaps much
157 more than you are used to, or have seen before. Because of this, you might want
158 to do a little background reading in the techniques used and capabilities of the
159 library. There are many good pages that discuss the STL, and several books on
160 the subject that you can get, so it will not be discussed in this document.<p>
162 Here are some useful links:<p>
164 <li><a href="http://www.dinkumware.com/refxcpp.html">Dinkumware C++
165 Library reference</a> - an excellent reference for the STL and other parts of
166 the standard C++ library.
168 <li><a href="http://www.tempest-sw.com/cpp/">C++ In a Nutshell</a> - This is an
169 O'Reilly book in the making. It has a decent <a
170 href="http://www.tempest-sw.com/cpp/ch13-libref.html">Standard Library
171 Reference</a> that rivals Dinkumware's, and is actually free until the book is
174 <li><a href="http://www.parashift.com/c++-faq-lite/">C++ Frequently Asked
177 <li><a href="http://www.sgi.com/tech/stl/">SGI's STL Programmer's Guide</a> -
179 href="http://www.sgi.com/tech/stl/stl_introduction.html">Introduction to the
182 <li><a href="http://www.research.att.com/~bs/C++.html">Bjarne Stroustrup's C++
187 You are also encouraged to take a look at the <a
188 href="CodingStandards.html">LLVM Coding Standards</a> guide which focuses on how
189 to write maintainable code more than where to put your curly braces.<p>
192 <!-- *********************************************************************** -->
193 </ul><table width="100%" bgcolor="#330077" border=0 cellpadding=4 cellspacing=0>
194 <tr><td align=center><font color="#EEEEFF" size=+2 face="Georgia,Palatino"><b>
195 <a name="apis">Important and useful LLVM APIs
196 </b></font></td></tr></table><ul>
197 <!-- *********************************************************************** -->
199 Here we highlight some LLVM APIs that are generally useful and good to know
200 about when writing transformations.<p>
202 <!-- ======================================================================= -->
203 </ul><table width="100%" bgcolor="#441188" border=0 cellpadding=4 cellspacing=0>
204 <tr><td> </td><td width="100%">
205 <font color="#EEEEFF" face="Georgia,Palatino"><b>
206 <a name="isa">The isa<>, cast<> and dyn_cast<> templates</a>
207 </b></font></td></tr></table><ul>
209 The LLVM source-base makes extensive use of a custom form of RTTI. These
210 templates have many similarities to the C++ <tt>dynamic_cast<></tt>
211 operator, but they don't have some drawbacks (primarily stemming from the fact
212 that <tt>dynamic_cast<></tt> only works on classes that have a v-table).
213 Because they are used so often, you must know what they do and how they work.
214 All of these templates are defined in the <a
215 href="/doxygen/Casting_8h-source.html"><tt>Support/Casting.h</tt></a> file (note
216 that you very rarely have to include this file directly).<p>
220 <dt><tt>isa<></tt>:
222 <dd>The <tt>isa<></tt> operator works exactly like the Java
223 "<tt>instanceof</tt>" operator. It returns true or false depending on whether a
224 reference or pointer points to an instance of the specified class. This can be
225 very useful for constraint checking of various sorts (example below).<p>
228 <dt><tt>cast<></tt>:
230 <dd>The <tt>cast<></tt> operator is a "checked cast" operation. It
231 converts a pointer or reference from a base class to a derived cast, causing an
232 assertion failure if it is not really an instance of the right type. This
233 should be used in cases where you have some information that makes you believe
234 that something is of the right type. An example of the <tt>isa<></tt> and
235 <tt>cast<></tt> template is:<p>
238 static bool isLoopInvariant(const <a href="#Value">Value</a> *V, const Loop *L) {
239 if (isa<<a href="#Constant">Constant</a>>(V) || isa<<a href="#Argument">Argument</a>>(V) || isa<<a href="#GlobalValue">GlobalValue</a>>(V))
242 <i>// Otherwise, it must be an instruction...</i>
243 return !L->contains(cast<<a href="#Instruction">Instruction</a>>(V)->getParent());
246 Note that you should <b>not</b> use an <tt>isa<></tt> test followed by a
247 <tt>cast<></tt>, for that use the <tt>dyn_cast<></tt> operator.<p>
250 <dt><tt>dyn_cast<></tt>:
252 <dd>The <tt>dyn_cast<></tt> operator is a "checking cast" operation. It
253 checks to see if the operand is of the specified type, and if so, returns a
254 pointer to it (this operator does not work with references). If the operand is
255 not of the correct type, a null pointer is returned. Thus, this works very much
256 like the <tt>dynamic_cast</tt> operator in C++, and should be used in the same
257 circumstances. Typically, the <tt>dyn_cast<></tt> operator is used in an
258 <tt>if</tt> statement or some other flow control statement like this:<p>
261 if (<a href="#AllocationInst">AllocationInst</a> *AI = dyn_cast<<a href="#AllocationInst">AllocationInst</a>>(Val)) {
266 This form of the <tt>if</tt> statement effectively combines together a call to
267 <tt>isa<></tt> and a call to <tt>cast<></tt> into one statement,
268 which is very convenient.<p>
270 Another common example is:<p>
273 <i>// Loop over all of the phi nodes in a basic block</i>
274 BasicBlock::iterator BBI = BB->begin();
275 for (; <a href="#PhiNode">PHINode</a> *PN = dyn_cast<<a href="#PHINode">PHINode</a>>(&*BBI); ++BBI)
279 Note that the <tt>dyn_cast<></tt> operator, like C++'s
280 <tt>dynamic_cast</tt> or Java's <tt>instanceof</tt> operator, can be abused. In
281 particular you should not use big chained <tt>if/then/else</tt> blocks to check
282 for lots of different variants of classes. If you find yourself wanting to do
283 this, it is much cleaner and more efficient to use the InstVisitor class to
284 dispatch over the instruction type directly.<p>
287 <dt><tt>cast_or_null<></tt>:
289 <dd>The <tt>cast_or_null<></tt> operator works just like the
290 <tt>cast<></tt> operator, except that it allows for a null pointer as an
291 argument (which it then propagates). This can sometimes be useful, allowing you
292 to combine several null checks into one.<p>
295 <dt><tt>dyn_cast_or_null<></tt>:
297 <dd>The <tt>dyn_cast_or_null<></tt> operator works just like the
298 <tt>dyn_cast<></tt> operator, except that it allows for a null pointer as
299 an argument (which it then propagates). This can sometimes be useful, allowing
300 you to combine several null checks into one.<p>
304 These five templates can be used with any classes, whether they have a v-table
305 or not. To add support for these templates, you simply need to add
306 <tt>classof</tt> static methods to the class you are interested casting to.
307 Describing this is currently outside the scope of this document, but there are
308 lots of examples in the LLVM source base.<p>
311 <!-- ======================================================================= -->
312 </ul><table width="100%" bgcolor="#441188" border=0 cellpadding=4 cellspacing=0>
313 <tr><td> </td><td width="100%">
314 <font color="#EEEEFF" face="Georgia,Palatino"><b>
315 <a name="DEBUG">The <tt>DEBUG()</tt> macro & <tt>-debug</tt> option</a>
316 </b></font></td></tr></table><ul>
318 Often when working on your pass you will put a bunch of debugging printouts and
319 other code into your pass. After you get it working, you want to remove
320 it... but you may need it again in the future (to work out new bugs that you run
323 Naturally, because of this, you don't want to delete the debug printouts, but
324 you don't want them to always be noisy. A standard compromise is to comment
325 them out, allowing you to enable them if you need them in the future.<p>
328 href="/doxygen/Statistic_8h-source.html">Support/Statistic.h</a></tt>"
329 file provides a macro named <tt>DEBUG()</tt> that is a much nicer solution to
330 this problem. Basically, you can put arbitrary code into the argument of the
331 <tt>DEBUG</tt> macro, and it is only executed if '<tt>opt</tt>' is run with the
332 '<tt>-debug</tt>' command line argument:
336 DEBUG(std::cerr << "I am here!\n");
340 Then you can run your pass like this:<p>
343 $ opt < a.bc > /dev/null -mypass
345 $ opt < a.bc > /dev/null -mypass -debug
350 Using the <tt>DEBUG()</tt> macro instead of a home brewed solution allows you to
351 now have to create "yet another" command line option for the debug output for
352 your pass. Note that <tt>DEBUG()</tt> macros are disabled for optimized
353 builds, so they do not cause a performance impact at all.<p>
356 <!-- ======================================================================= -->
357 </ul><table width="100%" bgcolor="#441188" border=0 cellpadding=4 cellspacing=0>
358 <tr><td> </td><td width="100%">
359 <font color="#EEEEFF" face="Georgia,Palatino"><b>
360 <a name="Statistic">The <tt>Statistic</tt> template & <tt>-stats</tt>
362 </b></font></td></tr></table><ul>
365 href="/doxygen/Statistic_8h-source.html">Support/Statistic.h</a></tt>"
366 file provides a template named <tt>Statistic</tt> that is used as a unified way
367 to keeping track of what the LLVM compiler is doing and how effective various
368 optimizations are. It is useful to see what optimizations are contributing to
369 making a particular program run faster.<p>
371 Often you may run your pass on some big program, and you're interested to see
372 how many times it makes a certain transformation. Although you can do this with
373 hand inspection, or some ad-hoc method, this is a real pain and not very useful
374 for big programs. Using the <tt>Statistic</tt> template makes it very easy to
375 keep track of this information, and the calculated information is presented in a
376 uniform manner with the rest of the passes being executed.<p>
378 There are many examples of <tt>Statistic</tt> users, but this basics of using it
382 <li>Define your statistic like this:<p>
385 static Statistic<> NumXForms("mypassname", "The # of times I did stuff");
388 The <tt>Statistic</tt> template can emulate just about any data-type, but if you
389 do not specify a template argument, it defaults to acting like an unsigned int
390 counter (this is usually what you want).<p>
392 <li>Whenever you make a transformation, bump the counter:<p>
395 ++NumXForms; // I did stuff
400 That's all you have to do. To get '<tt>opt</tt>' to print out the statistics
401 gathered, use the '<tt>-stats</tt>' option:<p>
404 $ opt -stats -mypassname < program.bc > /dev/null
405 ... statistic output ...
408 When running <tt>gccas</tt> on a C file from the SPEC benchmark suite, it gives
409 a report that looks like this:<p>
412 7646 bytecodewriter - Number of normal instructions
413 725 bytecodewriter - Number of oversized instructions
414 129996 bytecodewriter - Number of bytecode bytes written
415 2817 raise - Number of insts DCEd or constprop'd
416 3213 raise - Number of cast-of-self removed
417 5046 raise - Number of expression trees converted
418 75 raise - Number of other getelementptr's formed
419 138 raise - Number of load/store peepholes
420 42 deadtypeelim - Number of unused typenames removed from symtab
421 392 funcresolve - Number of varargs functions resolved
422 27 globaldce - Number of global variables removed
423 2 adce - Number of basic blocks removed
424 134 cee - Number of branches revectored
425 49 cee - Number of setcc instruction eliminated
426 532 gcse - Number of loads removed
427 2919 gcse - Number of instructions removed
428 86 indvars - Number of cannonical indvars added
429 87 indvars - Number of aux indvars removed
430 25 instcombine - Number of dead inst eliminate
431 434 instcombine - Number of insts combined
432 248 licm - Number of load insts hoisted
433 1298 licm - Number of insts hoisted to a loop pre-header
434 3 licm - Number of insts hoisted to multiple loop preds (bad, no loop pre-header)
435 75 mem2reg - Number of alloca's promoted
436 1444 cfgsimplify - Number of blocks simplified
439 Obviously, with so many optimizations, having a unified framework for this stuff
440 is very nice. Making your pass fit well into the framework makes it more
441 maintainable and useful.<p>
444 <!-- *********************************************************************** -->
445 </ul><table width="100%" bgcolor="#330077" border=0 cellpadding=4 cellspacing=0>
446 <tr><td align=center><font color="#EEEEFF" size=+2 face="Georgia,Palatino"><b>
447 <a name="common">Helpful Hints for Common Operations
448 </b></font></td></tr></table><ul> <!--
449 *********************************************************************** -->
451 This section describes how to perform some very simple transformations of LLVM
452 code. This is meant to give examples of common idioms used, showing the
453 practical side of LLVM transformations.<p>
455 Because this is a "how-to" section, you should also read about the main classes
456 that you will be working with. The <a href="#coreclasses">Core LLVM Class
457 Hierarchy Reference</a> contains details and descriptions of the main classes
458 that you should know about.<p>
460 <!-- NOTE: this section should be heavy on example code -->
463 <!-- ======================================================================= -->
464 </ul><table width="100%" bgcolor="#441188" border=0 cellpadding=4 cellspacing=0>
465 <tr><td> </td><td width="100%">
466 <font color="#EEEEFF" face="Georgia,Palatino"><b>
467 <a name="inspection">Basic Inspection and Traversal Routines</a>
468 </b></font></td></tr></table><ul>
470 The LLVM compiler infrastructure have many different data structures that may be
471 traversed. Following the example of the C++ standard template library, the
472 techniques used to traverse these various data structures are all basically the
473 same. For a enumerable sequence of values, the <tt>XXXbegin()</tt> function (or
474 method) returns an iterator to the start of the sequence, the <tt>XXXend()</tt>
475 function returns an iterator pointing to one past the last valid element of the
476 sequence, and there is some <tt>XXXiterator</tt> data type that is common
477 between the two operations.<p>
479 Because the pattern for iteration is common across many different aspects of the
480 program representation, the standard template library algorithms may be used on
481 them, and it is easier to remember how to iterate. First we show a few common
482 examples of the data structures that need to be traversed. Other data
483 structures are traversed in very similar ways.<p>
486 <!-- _______________________________________________________________________ -->
487 </ul><h4><a name="iterate_function"><hr size=0>Iterating over the <a
488 href="#BasicBlock"><tt>BasicBlock</tt></a>s in a <a
489 href="#Function"><tt>Function</tt></a> </h4><ul>
491 It's quite common to have a <tt>Function</tt> instance that you'd like
492 to transform in some way; in particular, you'd like to manipulate its
493 <tt>BasicBlock</tt>s. To facilitate this, you'll need to iterate over
494 all of the <tt>BasicBlock</tt>s that constitute the <tt>Function</tt>.
495 The following is an example that prints the name of a
496 <tt>BasicBlock</tt> and the number of <tt>Instruction</tt>s it
500 // func is a pointer to a Function instance
501 for(Function::iterator i = func->begin(), e = func->end(); i != e; ++i) {
503 // print out the name of the basic block if it has one, and then the
504 // number of instructions that it contains
506 cerr << "Basic block (name=" << i->getName() << ") has "
507 << i->size() << " instructions.\n";
511 Note that i can be used as if it were a pointer for the purposes of
512 invoking member functions of the <tt>Instruction</tt> class. This is
513 because the indirection operator is overloaded for the iterator
514 classes. In the above code, the expression <tt>i->size()</tt> is
515 exactly equivalent to <tt>(*i).size()</tt> just like you'd expect.
517 <!-- _______________________________________________________________________ -->
518 </ul><h4><a name="iterate_basicblock"><hr size=0>Iterating over the <a
519 href="#Instruction"><tt>Instruction</tt></a>s in a <a
520 href="#BasicBlock"><tt>BasicBlock</tt></a> </h4><ul>
522 Just like when dealing with <tt>BasicBlock</tt>s in
523 <tt>Function</tt>s, it's easy to iterate over the individual
524 instructions that make up <tt>BasicBlock</tt>s. Here's a code snippet
525 that prints out each instruction in a <tt>BasicBlock</tt>:
528 // blk is a pointer to a BasicBlock instance
529 for(BasicBlock::iterator i = blk->begin(), e = blk->end(); i != e; ++i)
530 // the next statement works since operator<<(ostream&,...)
531 // is overloaded for Instruction&
532 cerr << *i << "\n";
535 However, this isn't really the best way to print out the contents of a
536 <tt>BasicBlock</tt>! Since the ostream operators are overloaded for
537 virtually anything you'll care about, you could have just invoked the
538 print routine on the basic block itself: <tt>cerr << *blk <<
541 Note that currently operator<< is implemented for <tt>Value*</tt>, so it
542 will print out the contents of the pointer, instead of
543 the pointer value you might expect. This is a deprecated interface that will
544 be removed in the future, so it's best not to depend on it. To print out the
545 pointer value for now, you must cast to <tt>void*</tt>.<p>
548 <!-- _______________________________________________________________________ -->
549 </ul><h4><a name="iterate_institer"><hr size=0>Iterating over the <a
550 href="#Instruction"><tt>Instruction</tt></a>s in a <a
551 href="#Function"><tt>Function</tt></a></h4><ul>
553 If you're finding that you commonly iterate over a <tt>Function</tt>'s
554 <tt>BasicBlock</tt>s and then that <tt>BasicBlock</tt>'s
555 <tt>Instruction</tt>s, <tt>InstIterator</tt> should be used instead.
556 You'll need to include <a href="/doxygen/InstIterator_8h-source.html"><tt>llvm/Support/InstIterator.h</tt></a>, and then
557 instantiate <tt>InstIterator</tt>s explicitly in your code. Here's a
558 small example that shows how to dump all instructions in a function to
559 stderr (<b>Note:</b> Dereferencing an <tt>InstIterator</tt> yields an
560 <tt>Instruction*</tt>, <i>not</i> an <tt>Instruction&</tt>!):
563 #include "<a href="/doxygen/InstIterator_8h-source.html">llvm/Support/InstIterator.h</a>"
565 // Suppose F is a ptr to a function
566 for(inst_iterator i = inst_begin(F), e = inst_end(F); i != e; ++i)
567 cerr << **i << "\n";
570 Easy, isn't it? You can also use <tt>InstIterator</tt>s to fill a
571 worklist with its initial contents. For example, if you wanted to
572 initialize a worklist to contain all instructions in a
573 <tt>Function</tt> F, all you would need to do is something like:
576 std::set<Instruction*> worklist;
577 worklist.insert(inst_begin(F), inst_end(F));
580 The STL set <tt>worklist</tt> would now contain all instructions in
581 the <tt>Function</tt> pointed to by F.
583 <!-- _______________________________________________________________________ -->
584 </ul><h4><a name="iterate_convert"><hr size=0>Turning an iterator into a class
585 pointer (and vice-versa) </h4><ul>
587 Sometimes, it'll be useful to grab a reference (or pointer) to a class
588 instance when all you've got at hand is an iterator. Well, extracting
589 a reference or a pointer from an iterator is very straightforward.
590 Assuming that <tt>i</tt> is a <tt>BasicBlock::iterator</tt> and
591 <tt>j</tt> is a <tt>BasicBlock::const_iterator</tt>:
594 Instruction& inst = *i; // grab reference to instruction reference
595 Instruction* pinst = &*i; // grab pointer to instruction reference
596 const Instruction& inst = *j;
598 However, the iterators you'll be working with in the LLVM framework
599 are special: they will automatically convert to a ptr-to-instance type
600 whenever they need to. Instead of dereferencing the iterator and then
601 taking the address of the result, you can simply assign the iterator
602 to the proper pointer type and you get the dereference and address-of
603 operation as a result of the assignment (behind the scenes, this is a
604 result of overloading casting mechanisms). Thus the last line of the
607 <pre>Instruction* pinst = &*i;</pre>
609 is semantically equivalent to
611 <pre>Instruction* pinst = i;</pre>
613 <b>Caveat emptor</b>: The above syntax works <i>only</i> when you're <i>not</i>
614 working with <tt>dyn_cast</tt>. The template definition of <tt><a
615 href="#isa">dyn_cast</a></tt> isn't implemented to handle this yet, so you'll
616 still need the following in order for things to work properly:
619 BasicBlock::iterator bbi = ...;
620 <a href="#BranchInst">BranchInst</a>* b = <a href="#isa">dyn_cast</a><<a href="#BranchInst">BranchInst</a>>(&*bbi);
623 It's also possible to turn a class pointer into the corresponding
624 iterator. Usually, this conversion is quite inexpensive. The
625 following code snippet illustrates use of the conversion constructors
626 provided by LLVM iterators. By using these, you can explicitly grab
627 the iterator of something without actually obtaining it via iteration
631 void printNextInstruction(Instruction* inst) {
632 BasicBlock::iterator it(inst);
633 ++it; // after this line, it refers to the instruction after *inst.
634 if(it != inst->getParent()->end()) cerr << *it << "\n";
637 Of course, this example is strictly pedagogical, because it'd be much
638 better to explicitly grab the next instruction directly from inst.
641 <!--_______________________________________________________________________-->
642 </ul><h4><a name="iterate_complex"><hr size=0>Finding call sites: a slightly
643 more complex example </h4><ul>
645 Say that you're writing a FunctionPass and would like to count all the
646 locations in the entire module (that is, across every
647 <tt>Function</tt>) where a certain function (i.e. some
648 <tt>Function</tt>*) already in scope. As you'll learn later, you may
649 want to use an <tt>InstVisitor</tt> to accomplish this in a much more
650 straightforward manner, but this example will allow us to explore how
651 you'd do it if you didn't have <tt>InstVisitor</tt> around. In
652 pseudocode, this is what we want to do:
655 initialize callCounter to zero
656 for each Function f in the Module
657 for each BasicBlock b in f
658 for each Instruction i in b
659 if(i is a CallInst and calls the given function)
660 increment callCounter
663 And the actual code is (remember, since we're writing a
664 <tt>FunctionPass</tt>, our <tt>FunctionPass</tt>-derived class simply
665 has to override the <tt>runOnFunction</tt> method...):
668 Function* targetFunc = ...;
670 class OurFunctionPass : public FunctionPass {
672 OurFunctionPass(): callCounter(0) { }
674 virtual runOnFunction(Function& F) {
675 for(Function::iterator b = F.begin(), be = F.end(); b != be; ++b) {
676 for(BasicBlock::iterator i = b->begin(); ie = b->end(); i != ie; ++i) {
677 if (<a href="#CallInst">CallInst</a>* callInst = <a href="#isa">dyn_cast</a><<a href="#CallInst">CallInst</a>>(&*i)) {
678 // we know we've encountered a call instruction, so we
679 // need to determine if it's a call to the
680 // function pointed to by m_func or not.
682 if(callInst->getCalledFunction() == targetFunc)
689 unsigned callCounter;
693 <!--_______________________________________________________________________-->
694 </ul><h4><a name="iterate_chains"><hr size=0>Iterating over def-use &
695 use-def chains</h4><ul>
697 Frequently, we might have an instance of the <a
698 href="/doxygen/classValue.html">Value Class</a> and we want to
699 determine which <tt>User</tt>s use the <tt>Value</tt>. The list of
700 all <tt>User</tt>s of a particular <tt>Value</tt> is called a
701 <i>def-use</i> chain. For example, let's say we have a
702 <tt>Function*</tt> named <tt>F</tt> to a particular function
703 <tt>foo</tt>. Finding all of the instructions that <i>use</i>
704 <tt>foo</tt> is as simple as iterating over the <i>def-use</i> chain of
710 for(Value::use_iterator i = F->use_begin(), e = F->use_end(); i != e; ++i) {
711 if(Instruction* Inst = dyn_cast<Instruction>(*i)) {
712 cerr << "F is used in instruction:\n";
713 cerr << *Inst << "\n";
718 Alternately, it's common to have an instance of the <a
719 href="/doxygen/classUser.html">User Class</a> and need to know what
720 <tt>Value</tt>s are used by it. The list of all <tt>Value</tt>s used
721 by a <tt>User</tt> is known as a <i>use-def</i> chain. Instances of
722 class <tt>Instruction</tt> are common <tt>User</tt>s, so we might want
723 to iterate over all of the values that a particular instruction uses
724 (that is, the operands of the particular <tt>Instruction</tt>):
727 Instruction* pi = ...;
729 for(User::op_iterator i = pi->op_begin(), e = pi->op_end(); i != e; ++i) {
737 def-use chains ("finding all users of"): Value::use_begin/use_end
738 use-def chains ("finding all values used"): User::op_begin/op_end [op=operand]
741 <!-- ======================================================================= -->
742 </ul><table width="100%" bgcolor="#441188" border=0 cellpadding=4 cellspacing=0>
743 <tr><td> </td><td width="100%">
744 <font color="#EEEEFF" face="Georgia,Palatino"><b>
745 <a name="simplechanges">Making simple changes</a>
746 </b></font></td></tr></table><ul>
748 There are some primitive transformation operations present in the LLVM
749 infrastructure that are worth knowing about. When performing
750 transformations, it's fairly common to manipulate the contents of
751 basic blocks. This section describes some of the common methods for
752 doing so and gives example code.
754 <!--_______________________________________________________________________-->
755 </ul><h4><a name="schanges_creating"><hr size=0>Creating and inserting
756 new <tt>Instruction</tt>s</h4><ul>
758 <i>Instantiating Instructions</i>
760 <p>Creation of <tt>Instruction</tt>s is straightforward: simply call the
761 constructor for the kind of instruction to instantiate and provide the
762 necessary parameters. For example, an <tt>AllocaInst</tt> only
763 <i>requires</i> a (const-ptr-to) <tt>Type</tt>. Thus:
765 <pre>AllocaInst* ai = new AllocaInst(Type::IntTy);</pre>
767 will create an <tt>AllocaInst</tt> instance that represents the
768 allocation of one integer in the current stack frame, at runtime.
769 Each <tt>Instruction</tt> subclass is likely to have varying default
770 parameters which change the semantics of the instruction, so refer to
771 the <a href="/doxygen/classInstruction.html">doxygen documentation for
772 the subclass of Instruction</a> that you're interested in
775 <p><i>Naming values</i></p>
778 It is very useful to name the values of instructions when you're able
779 to, as this facilitates the debugging of your transformations. If you
780 end up looking at generated LLVM machine code, you definitely want to
781 have logical names associated with the results of instructions! By
782 supplying a value for the <tt>Name</tt> (default) parameter of the
783 <tt>Instruction</tt> constructor, you associate a logical name with
784 the result of the instruction's execution at runtime. For example,
785 say that I'm writing a transformation that dynamically allocates space
786 for an integer on the stack, and that integer is going to be used as
787 some kind of index by some other code. To accomplish this, I place an
788 <tt>AllocaInst</tt> at the first point in the first
789 <tt>BasicBlock</tt> of some <tt>Function</tt>, and I'm intending to
790 use it within the same <tt>Function</tt>. I might do:
792 <pre>AllocaInst* pa = new AllocaInst(Type::IntTy, 0, "indexLoc");</pre>
794 where <tt>indexLoc</tt> is now the logical name of the instruction's
795 execution value, which is a pointer to an integer on the runtime
799 <p><i>Inserting instructions</i></p>
802 There are essentially two ways to insert an <tt>Instruction</tt> into
803 an existing sequence of instructions that form a <tt>BasicBlock</tt>:
805 <li>Insertion into an explicit instruction list
807 <p>Given a <tt>BasicBlock* pb</tt>, an <tt>Instruction* pi</tt> within
808 that <tt>BasicBlock</tt>, and a newly-created instruction
809 we wish to insert before <tt>*pi</tt>, we do the following:
812 BasicBlock* pb = ...;
813 Instruction* pi = ...;
814 Instruction* newInst = new Instruction(...);
815 pb->getInstList().insert(pi, newInst); // inserts newInst before pi in pb
819 <li>Insertion into an implicit instruction list
820 <p><tt>Instruction</tt> instances that are already in
821 <tt>BasicBlock</tt>s are implicitly associated with an existing
822 instruction list: the instruction list of the enclosing basic block.
823 Thus, we could have accomplished the same thing as the above code
824 without being given a <tt>BasicBlock</tt> by doing:
826 Instruction* pi = ...;
827 Instruction* newInst = new Instruction(...);
828 pi->getParent()->getInstList().insert(pi, newInst);
830 In fact, this sequence of steps occurs so frequently that the
831 <tt>Instruction</tt> class and <tt>Instruction</tt>-derived classes
832 provide constructors which take (as a default parameter) a pointer to
833 an <tt>Instruction</tt> which the newly-created <tt>Instruction</tt>
834 should precede. That is, <tt>Instruction</tt> constructors are
835 capable of inserting the newly-created instance into the
836 <tt>BasicBlock</tt> of a provided instruction, immediately before that
837 instruction. Using an <tt>Instruction</tt> constructor with a
838 <tt>insertBefore</tt> (default) parameter, the above code becomes:
840 Instruction* pi = ...;
841 Instruction* newInst = new Instruction(..., pi);
843 which is much cleaner, especially if you're creating a lot of
844 instructions and adding them to <tt>BasicBlock</tt>s.
849 <!--_______________________________________________________________________-->
850 </ul><h4><a name="schanges_deleting"><hr size=0>Deleting
851 <tt>Instruction</tt>s</h4><ul>
853 Deleting an instruction from an existing sequence of instructions that form a <a
854 href="#BasicBlock"><tt>BasicBlock</tt></a> is very straightforward. First, you
855 must have a pointer to the instruction that you wish to delete. Second, you
856 need to obtain the pointer to that instruction's basic block. You use the
857 pointer to the basic block to get its list of instructions and then use the
858 erase function to remove your instruction.<p>
863 <a href="#Instruction">Instruction</a> *I = .. ;
864 <a href="#BasicBlock">BasicBlock</a> *BB = I->getParent();
865 BB->getInstList().erase(I);
868 <!--_______________________________________________________________________-->
869 </ul><h4><a name="schanges_replacing"><hr size=0>Replacing an
870 <tt>Instruction</tt> with another <tt>Value</tt></h4><ul>
872 <p><i>Replacing individual instructions</i></p>
875 href="/doxygen/BasicBlockUtils_8h-source.html">llvm/Transforms/Utils/BasicBlockUtils.h
876 </a>" permits use of two very useful replace functions:
877 <tt>ReplaceInstWithValue</tt> and <tt>ReplaceInstWithInst</tt>.
881 <li><tt>ReplaceInstWithValue</tt>
883 <p>This function replaces all uses (within a basic block) of a given
884 instruction with a value, and then removes the original instruction.
885 The following example illustrates the replacement of the result of a
886 particular <tt>AllocaInst</tt> that allocates memory for a single
887 integer with an null pointer to an integer.</p>
890 AllocaInst* instToReplace = ...;
891 BasicBlock::iterator ii(instToReplace);
892 ReplaceInstWithValue(instToReplace->getParent()->getInstList(), ii,
893 Constant::getNullValue(PointerType::get(Type::IntTy)));
896 <li><tt>ReplaceInstWithInst</tt>
898 <p>This function replaces a particular instruction with another
899 instruction. The following example illustrates the replacement of one
900 <tt>AllocaInst</tt> with another.<p>
903 AllocaInst* instToReplace = ...;
904 BasicBlock::iterator ii(instToReplace);
905 ReplaceInstWithInst(instToReplace->getParent()->getInstList(), ii,
906 new AllocaInst(Type::IntTy, 0, "ptrToReplacedInt");
910 <p><i>Replacing multiple uses of <tt>User</tt>s and
911 <tt>Value</tt>s</i></p>
913 You can use <tt>Value::replaceAllUsesWith</tt> and
914 <tt>User::replaceUsesOfWith</tt> to change more than one use at a
915 time. See the doxygen documentation for the <a
916 href="/doxygen/classValue.html">Value Class</a> and <a
917 href="/doxygen/classUser.html">User Class</a>, respectively, for more
920 <!-- Value::replaceAllUsesWith User::replaceUsesOfWith Point out:
921 include/llvm/Transforms/Utils/ especially BasicBlockUtils.h with:
922 ReplaceInstWithValue, ReplaceInstWithInst
925 <!-- *********************************************************************** -->
926 </ul><table width="100%" bgcolor="#330077" border=0 cellpadding=4 cellspacing=0>
927 <tr><td align=center><font color="#EEEEFF" size=+2 face="Georgia,Palatino"><b>
928 <a name="coreclasses">The Core LLVM Class Hierarchy Reference
929 </b></font></td></tr></table><ul>
930 <!-- *********************************************************************** -->
932 The Core LLVM classes are the primary means of representing the program being
933 inspected or transformed. The core LLVM classes are defined in header files in
934 the <tt>include/llvm/</tt> directory, and implemented in the <tt>lib/VMCore</tt>
938 <!-- ======================================================================= -->
939 </ul><table width="100%" bgcolor="#441188" border=0 cellpadding=4 cellspacing=0>
940 <tr><td> </td><td width="100%">
941 <font color="#EEEEFF" face="Georgia,Palatino"><b>
942 <a name="Value">The <tt>Value</tt> class</a>
943 </b></font></td></tr></table><ul>
945 <tt>#include "<a href="/doxygen/Value_8h-source.html">llvm/Value.h</a>"</tt></b><br>
946 doxygen info: <a href="/doxygen/classValue.html">Value Class</a><p>
949 The <tt>Value</tt> class is the most important class in LLVM Source base. It
950 represents a typed value that may be used (among other things) as an operand to
951 an instruction. There are many different types of <tt>Value</tt>s, such as <a
952 href="#Constant"><tt>Constant</tt></a>s, <a
953 href="#Argument"><tt>Argument</tt></a>s, and even <a
954 href="#Instruction"><tt>Instruction</tt></a>s and <a
955 href="#Function"><tt>Function</tt></a>s are <tt>Value</tt>s.<p>
957 A particular <tt>Value</tt> may be used many times in the LLVM representation
958 for a program. For example, an incoming argument to a function (represented
959 with an instance of the <a href="#Argument">Argument</a> class) is "used" by
960 every instruction in the function that references the argument. To keep track
961 of this relationship, the <tt>Value</tt> class keeps a list of all of the <a
962 href="#User"><tt>User</tt></a>s that is using it (the <a
963 href="#User"><tt>User</tt></a> class is a base class for all nodes in the LLVM
964 graph that can refer to <tt>Value</tt>s). This use list is how LLVM represents
965 def-use information in the program, and is accessible through the <tt>use_</tt>*
966 methods, shown below.<p>
968 Because LLVM is a typed representation, every LLVM <tt>Value</tt> is typed, and
969 this <a href="#Type">Type</a> is available through the <tt>getType()</tt>
970 method. <a name="#nameWarning">In addition, all LLVM values can be named. The
971 "name" of the <tt>Value</tt> is symbolic string printed in the LLVM code:<p>
974 %<b>foo</b> = add int 1, 2
977 The name of this instruction is "foo". <b>NOTE</b> that the name of any value
978 may be missing (an empty string), so names should <b>ONLY</b> be used for
979 debugging (making the source code easier to read, debugging printouts), they
980 should not be used to keep track of values or map between them. For this
981 purpose, use a <tt>std::map</tt> of pointers to the <tt>Value</tt> itself
984 One important aspect of LLVM is that there is no distinction between an SSA
985 variable and the operation that produces it. Because of this, any reference to
986 the value produced by an instruction (or the value available as an incoming
987 argument, for example) is represented as a direct pointer to the class that
988 represents this value. Although this may take some getting used to, it
989 simplifies the representation and makes it easier to manipulate.<p>
992 <!-- _______________________________________________________________________ -->
993 </ul><h4><a name="m_Value"><hr size=0>Important Public Members of
994 the <tt>Value</tt> class</h4><ul>
996 <li><tt>Value::use_iterator</tt> - Typedef for iterator over the use-list<br>
997 <tt>Value::use_const_iterator</tt>
998 - Typedef for const_iterator over the use-list<br>
999 <tt>unsigned use_size()</tt> - Returns the number of users of the value.<br>
1000 <tt>bool use_empty()</tt> - Returns true if there are no users.<br>
1001 <tt>use_iterator use_begin()</tt>
1002 - Get an iterator to the start of the use-list.<br>
1003 <tt>use_iterator use_end()</tt>
1004 - Get an iterator to the end of the use-list.<br>
1005 <tt><a href="#User">User</a> *use_back()</tt>
1006 - Returns the last element in the list.<p>
1008 These methods are the interface to access the def-use information in LLVM. As with all other iterators in LLVM, the naming conventions follow the conventions defined by the <a href="#stl">STL</a>.<p>
1010 <li><tt><a href="#Type">Type</a> *getType() const</tt><p>
1011 This method returns the Type of the Value.
1013 <li><tt>bool hasName() const</tt><br>
1014 <tt>std::string getName() const</tt><br>
1015 <tt>void setName(const std::string &Name)</tt><p>
1017 This family of methods is used to access and assign a name to a <tt>Value</tt>,
1018 be aware of the <a href="#nameWarning">precaution above</a>.<p>
1021 <li><tt>void replaceAllUsesWith(Value *V)</tt><p>
1023 This method traverses the use list of a <tt>Value</tt> changing all <a
1024 href="#User"><tt>User</tt>s</a> of the current value to refer to "<tt>V</tt>"
1025 instead. For example, if you detect that an instruction always produces a
1026 constant value (for example through constant folding), you can replace all uses
1027 of the instruction with the constant like this:<p>
1030 Inst->replaceAllUsesWith(ConstVal);
1035 <!-- ======================================================================= -->
1036 </ul><table width="100%" bgcolor="#441188" border=0 cellpadding=4 cellspacing=0>
1037 <tr><td> </td><td width="100%">
1038 <font color="#EEEEFF" face="Georgia,Palatino"><b>
1039 <a name="User">The <tt>User</tt> class</a>
1040 </b></font></td></tr></table><ul>
1042 <tt>#include "<a href="/doxygen/User_8h-source.html">llvm/User.h</a>"</tt></b><br>
1043 doxygen info: <a href="/doxygen/classUser.html">User Class</a><br>
1044 Superclass: <a href="#Value"><tt>Value</tt></a><p>
1047 The <tt>User</tt> class is the common base class of all LLVM nodes that may
1048 refer to <a href="#Value"><tt>Value</tt></a>s. It exposes a list of "Operands"
1049 that are all of the <a href="#Value"><tt>Value</tt></a>s that the User is
1050 referring to. The <tt>User</tt> class itself is a subclass of
1053 The operands of a <tt>User</tt> point directly to the LLVM <a
1054 href="#Value"><tt>Value</tt></a> that it refers to. Because LLVM uses Static
1055 Single Assignment (SSA) form, there can only be one definition referred to,
1056 allowing this direct connection. This connection provides the use-def
1057 information in LLVM.<p>
1059 <!-- _______________________________________________________________________ -->
1060 </ul><h4><a name="m_User"><hr size=0>Important Public Members of
1061 the <tt>User</tt> class</h4><ul>
1063 The <tt>User</tt> class exposes the operand list in two ways: through an index
1064 access interface and through an iterator based interface.<p>
1066 <li><tt>Value *getOperand(unsigned i)</tt><br>
1067 <tt>unsigned getNumOperands()</tt><p>
1069 These two methods expose the operands of the <tt>User</tt> in a convenient form
1070 for direct access.<p>
1072 <li><tt>User::op_iterator</tt> - Typedef for iterator over the operand list<br>
1073 <tt>User::op_const_iterator</tt>
1074 <tt>use_iterator op_begin()</tt>
1075 - Get an iterator to the start of the operand list.<br>
1076 <tt>use_iterator op_end()</tt>
1077 - Get an iterator to the end of the operand list.<p>
1079 Together, these methods make up the iterator based interface to the operands of
1084 <!-- ======================================================================= -->
1085 </ul><table width="100%" bgcolor="#441188" border=0 cellpadding=4 cellspacing=0>
1086 <tr><td> </td><td width="100%">
1087 <font color="#EEEEFF" face="Georgia,Palatino"><b>
1088 <a name="Instruction">The <tt>Instruction</tt> class</a>
1089 </b></font></td></tr></table><ul>
1092 href="/doxygen/Instruction_8h-source.html">llvm/Instruction.h</a>"</tt></b><br>
1093 doxygen info: <a href="/doxygen/classInstruction.html">Instruction Class</a><br>
1094 Superclasses: <a href="#User"><tt>User</tt></a>, <a
1095 href="#Value"><tt>Value</tt></a><p>
1097 The <tt>Instruction</tt> class is the common base class for all LLVM
1098 instructions. It provides only a few methods, but is a very commonly used
1099 class. The primary data tracked by the <tt>Instruction</tt> class itself is the
1100 opcode (instruction type) and the parent <a
1101 href="#BasicBlock"><tt>BasicBlock</tt></a> the <tt>Instruction</tt> is embedded
1102 into. To represent a specific type of instruction, one of many subclasses of
1103 <tt>Instruction</tt> are used.<p>
1105 Because the <tt>Instruction</tt> class subclasses the <a
1106 href="#User"><tt>User</tt></a> class, its operands can be accessed in the same
1107 way as for other <a href="#User"><tt>User</tt></a>s (with the
1108 <tt>getOperand()</tt>/<tt>getNumOperands()</tt> and
1109 <tt>op_begin()</tt>/<tt>op_end()</tt> methods).<p>
1111 An important file for the <tt>Instruction</tt> class is the
1112 <tt>llvm/Instruction.def</tt> file. This file contains some meta-data about the
1113 various different types of instructions in LLVM. It describes the enum values
1114 that are used as opcodes (for example <tt>Instruction::Add</tt> and
1115 <tt>Instruction::SetLE</tt>), as well as the concrete sub-classes of
1116 <tt>Instruction</tt> that implement the instruction (for example <tt><a
1117 href="#BinaryOperator">BinaryOperator</a></tt> and <tt><a
1118 href="#SetCondInst">SetCondInst</a></tt>). Unfortunately, the use of macros in
1119 this file confused doxygen, so these enum values don't show up correctly in the
1120 <a href="/doxygen/classInstruction.html">doxygen output</a>.<p>
1123 <!-- _______________________________________________________________________ -->
1124 </ul><h4><a name="m_Instruction"><hr size=0>Important Public Members of
1125 the <tt>Instruction</tt> class</h4><ul>
1127 <li><tt><a href="#BasicBlock">BasicBlock</a> *getParent()</tt><p>
1129 Returns the <a href="#BasicBlock"><tt>BasicBlock</tt></a> that this
1130 <tt>Instruction</tt> is embedded into.<p>
1132 <li><tt>bool hasSideEffects()</tt><p>
1134 Returns true if the instruction has side effects, i.e. it is a <tt>call</tt>,
1135 <tt>free</tt>, <tt>invoke</tt>, or <tt>store</tt>.<p>
1137 <li><tt>unsigned getOpcode()</tt><p>
1139 Returns the opcode for the <tt>Instruction</tt>.<p>
1141 <li><tt><a href="#Instruction">Instruction</a> *clone() const</tt><p>
1143 Returns another instance of the specified instruction, identical in all ways to
1144 the original except that the instruction has no parent (ie it's not embedded
1145 into a <a href="#BasicBlock"><tt>BasicBlock</tt></a>), and it has no name.<p>
1151 \subsection{Subclasses of Instruction :}
1153 <li>BinaryOperator : This subclass of Instruction defines a general interface to the all the instructions involvong binary operators in LLVM.
1155 <li><tt>bool swapOperands()</tt>: Exchange the two operands to this instruction. If the instruction cannot be reversed (i.e. if it's a Div), it returns true.
1157 <li>TerminatorInst : This subclass of Instructions defines an interface for all instructions that can terminate a BasicBlock.
1159 <li> <tt>unsigned getNumSuccessors()</tt>: Returns the number of successors for this terminator instruction.
1160 <li><tt>BasicBlock *getSuccessor(unsigned i)</tt>: As the name suggests returns the ith successor BasicBlock.
1161 <li><tt>void setSuccessor(unsigned i, BasicBlock *B)</tt>: sets BasicBlock B as the ith succesor to this terminator instruction.
1164 <li>PHINode : This represents the PHI instructions in the SSA form.
1166 <li><tt> unsigned getNumIncomingValues()</tt>: Returns the number of incoming edges to this PHI node.
1167 <li><tt> Value *getIncomingValue(unsigned i)</tt>: Returns the ith incoming Value.
1168 <li><tt>void setIncomingValue(unsigned i, Value *V)</tt>: Sets the ith incoming Value as V
1169 <li><tt>BasicBlock *getIncomingBlock(unsigned i)</tt>: Returns the Basic Block corresponding to the ith incoming Value.
1170 <li><tt> void addIncoming(Value *D, BasicBlock *BB)</tt>:
1171 Add an incoming value to the end of the PHI list
1172 <li><tt> int getBasicBlockIndex(const BasicBlock *BB) const</tt>:
1173 Returns the first index of the specified basic block in the value list for this PHI. Returns -1 if no instance.
1175 <li>CastInst : In LLVM all casts have to be done through explicit cast instructions. CastInst defines the interface to the cast instructions.
1176 <li>CallInst : This defines an interface to the call instruction in LLVM. ARguments to the function are nothing but operands of the instruction.
1178 <li>: <tt>Function *getCalledFunction()</tt>: Returns a handle to the function that is being called by this Function.
1180 <li>LoadInst, StoreInst, GetElemPtrInst : These subclasses represent load, store and getelementptr instructions in LLVM.
1182 <li><tt>Value * getPointerOperand()</tt>: Returns the Pointer Operand which is typically the 0th operand.
1184 <li>BranchInst : This is a subclass of TerminatorInst and defines the interface for conditional and unconditional branches in LLVM.
1186 <li><tt>bool isConditional()</tt>: Returns true if the branch is a conditional branch else returns false
1187 <li> <tt>Value *getCondition()</tt>: Returns the condition if it is a conditional branch else returns null.
1188 <li> <tt>void setUnconditionalDest(BasicBlock *Dest)</tt>: Changes the current branch to an unconditional one targetting the specified block.
1196 <!-- ======================================================================= -->
1197 </ul><table width="100%" bgcolor="#441188" border=0 cellpadding=4 cellspacing=0>
1198 <tr><td> </td><td width="100%">
1199 <font color="#EEEEFF" face="Georgia,Palatino"><b>
1200 <a name="BasicBlock">The <tt>BasicBlock</tt> class</a>
1201 </b></font></td></tr></table><ul>
1204 href="/doxygen/BasicBlock_8h-source.html">llvm/BasicBlock.h</a>"</tt></b><br>
1205 doxygen info: <a href="/doxygen/classBasicBlock.html">BasicBlock Class</a><br>
1206 Superclass: <a href="#Value"><tt>Value</tt></a><p>
1209 This class represents a single entry multiple exit section of the code, commonly
1210 known as a basic block by the compiler community. The <tt>BasicBlock</tt> class
1211 maintains a list of <a href="#Instruction"><tt>Instruction</tt></a>s, which form
1212 the body of the block. Matching the language definition, the last element of
1213 this list of instructions is always a terminator instruction (a subclass of the
1214 <a href="#TerminatorInst"><tt>TerminatorInst</tt></a> class).<p>
1216 In addition to tracking the list of instructions that make up the block, the
1217 <tt>BasicBlock</tt> class also keeps track of the <a
1218 href="#Function"><tt>Function</tt></a> that it is embedded into.<p>
1220 Note that <tt>BasicBlock</tt>s themselves are <a
1221 href="#Value"><tt>Value</tt></a>s, because they are referenced by instructions
1222 like branches and can go in the switch tables. <tt>BasicBlock</tt>s have type
1226 <!-- _______________________________________________________________________ -->
1227 </ul><h4><a name="m_BasicBlock"><hr size=0>Important Public Members of
1228 the <tt>BasicBlock</tt> class</h4><ul>
1230 <li><tt>BasicBlock(const std::string &Name = "", <a
1231 href="#Function">Function</a> *Parent = 0)</tt><p>
1233 The <tt>BasicBlock</tt> constructor is used to create new basic blocks for
1234 insertion into a function. The constructor simply takes a name for the new
1235 block, and optionally a <a href="#Function"><tt>Function</tt></a> to insert it
1236 into. If the <tt>Parent</tt> parameter is specified, the new
1237 <tt>BasicBlock</tt> is automatically inserted at the end of the specified <a
1238 href="#Function"><tt>Function</tt></a>, if not specified, the BasicBlock must be
1239 manually inserted into the <a href="#Function"><tt>Function</tt></a>.<p>
1241 <li><tt>BasicBlock::iterator</tt> - Typedef for instruction list iterator<br>
1242 <tt>BasicBlock::const_iterator</tt> - Typedef for const_iterator.<br>
1243 <tt>begin()</tt>, <tt>end()</tt>, <tt>front()</tt>, <tt>back()</tt>,
1244 <tt>size()</tt>, <tt>empty()</tt>, <tt>rbegin()</tt>, <tt>rend()</tt><p>
1246 These methods and typedefs are forwarding functions that have the same semantics
1247 as the standard library methods of the same names. These methods expose the
1248 underlying instruction list of a basic block in a way that is easy to
1249 manipulate. To get the full complement of container operations (including
1250 operations to update the list), you must use the <tt>getInstList()</tt>
1253 <li><tt>BasicBlock::InstListType &getInstList()</tt><p>
1255 This method is used to get access to the underlying container that actually
1256 holds the Instructions. This method must be used when there isn't a forwarding
1257 function in the <tt>BasicBlock</tt> class for the operation that you would like
1258 to perform. Because there are no forwarding functions for "updating"
1259 operations, you need to use this if you want to update the contents of a
1260 <tt>BasicBlock</tt>.<p>
1262 <li><tt><A href="#Function">Function</a> *getParent()</tt><p>
1264 Returns a pointer to <a href="#Function"><tt>Function</tt></a> the block is
1265 embedded into, or a null pointer if it is homeless.<p>
1267 <li><tt><a href="#TerminatorInst">TerminatorInst</a> *getTerminator()</tt><p>
1269 Returns a pointer to the terminator instruction that appears at the end of the
1270 <tt>BasicBlock</tt>. If there is no terminator instruction, or if the last
1271 instruction in the block is not a terminator, then a null pointer is
1275 <!-- ======================================================================= -->
1276 </ul><table width="100%" bgcolor="#441188" border=0 cellpadding=4 cellspacing=0>
1277 <tr><td> </td><td width="100%">
1278 <font color="#EEEEFF" face="Georgia,Palatino"><b>
1279 <a name="GlobalValue">The <tt>GlobalValue</tt> class</a>
1280 </b></font></td></tr></table><ul>
1283 href="/doxygen/GlobalValue_8h-source.html">llvm/GlobalValue.h</a>"</tt></b><br>
1284 doxygen info: <a href="/doxygen/classGlobalValue.html">GlobalValue Class</a><br>
1285 Superclasses: <a href="#User"><tt>User</tt></a>, <a
1286 href="#Value"><tt>Value</tt></a><p>
1288 Global values (<A href="#GlobalVariable"><tt>GlobalVariable</tt></a>s or <a
1289 href="#Function"><tt>Function</tt></a>s) are the only LLVM values that are
1290 visible in the bodies of all <a href="#Function"><tt>Function</tt></a>s.
1291 Because they are visible at global scope, they are also subject to linking with
1292 other globals defined in different translation units. To control the linking
1293 process, <tt>GlobalValue</tt>s know their linkage rules. Specifically,
1294 <tt>GlobalValue</tt>s know whether they have internal or external linkage.<p>
1296 If a <tt>GlobalValue</tt> has internal linkage (equivalent to being
1297 <tt>static</tt> in C), it is not visible to code outside the current translation
1298 unit, and does not participate in linking. If it has external linkage, it is
1299 visible to external code, and does participate in linking. In addition to
1300 linkage information, <tt>GlobalValue</tt>s keep track of which <a
1301 href="#Module"><tt>Module</tt></a> they are currently part of.<p>
1303 Because <tt>GlobalValue</tt>s are memory objects, they are always referred to by
1304 their address. As such, the <a href="#Type"><tt>Type</tt></a> of a global is
1305 always a pointer to its contents. This is explained in the LLVM Language
1306 Reference Manual.<p>
1309 <!-- _______________________________________________________________________ -->
1310 </ul><h4><a name="m_GlobalValue"><hr size=0>Important Public Members of
1311 the <tt>GlobalValue</tt> class</h4><ul>
1313 <li><tt>bool hasInternalLinkage() const</tt><br>
1314 <tt>bool hasExternalLinkage() const</tt><br>
1315 <tt>void setInternalLinkage(bool HasInternalLinkage)</tt><p>
1317 These methods manipulate the linkage characteristics of the
1318 <tt>GlobalValue</tt>.<p>
1320 <li><tt><a href="#Module">Module</a> *getParent()</tt><p>
1322 This returns the <a href="#Module"><tt>Module</tt></a> that the GlobalValue is
1323 currently embedded into.<p>
1327 <!-- ======================================================================= -->
1328 </ul><table width="100%" bgcolor="#441188" border=0 cellpadding=4 cellspacing=0>
1329 <tr><td> </td><td width="100%">
1330 <font color="#EEEEFF" face="Georgia,Palatino"><b>
1331 <a name="Function">The <tt>Function</tt> class</a>
1332 </b></font></td></tr></table><ul>
1335 href="/doxygen/Function_8h-source.html">llvm/Function.h</a>"</tt></b><br>
1336 doxygen info: <a href="/doxygen/classFunction.html">Function Class</a><br>
1337 Superclasses: <a href="#GlobalValue"><tt>GlobalValue</tt></a>, <a
1338 href="#User"><tt>User</tt></a>, <a href="#Value"><tt>Value</tt></a><p>
1340 The <tt>Function</tt> class represents a single procedure in LLVM. It is
1341 actually one of the more complex classes in the LLVM heirarchy because it must
1342 keep track of a large amount of data. The <tt>Function</tt> class keeps track
1343 of a list of <a href="#BasicBlock"><tt>BasicBlock</tt></a>s, a list of formal <a
1344 href="#Argument"><tt>Argument</tt></a>s, and a <a
1345 href="#SymbolTable"><tt>SymbolTable</tt></a>.<p>
1347 The list of <a href="#BasicBlock"><tt>BasicBlock</tt></a>s is the most commonly
1348 used part of <tt>Function</tt> objects. The list imposes an implicit ordering
1349 of the blocks in the function, which indicate how the code will be layed out by
1350 the backend. Additionally, the first <a
1351 href="#BasicBlock"><tt>BasicBlock</tt></a> is the implicit entry node for the
1352 <tt>Function</tt>. It is not legal in LLVM explicitly branch to this initial
1353 block. There are no implicit exit nodes, and in fact there may be multiple exit
1354 nodes from a single <tt>Function</tt>. If the <a
1355 href="#BasicBlock"><tt>BasicBlock</tt></a> list is empty, this indicates that
1356 the <tt>Function</tt> is actually a function declaration: the actual body of the
1357 function hasn't been linked in yet.<p>
1359 In addition to a list of <a href="#BasicBlock"><tt>BasicBlock</tt></a>s, the
1360 <tt>Function</tt> class also keeps track of the list of formal <a
1361 href="#Argument"><tt>Argument</tt></a>s that the function receives. This
1362 container manages the lifetime of the <a href="#Argument"><tt>Argument</tt></a>
1363 nodes, just like the <a href="#BasicBlock"><tt>BasicBlock</tt></a> list does for
1364 the <a href="#BasicBlock"><tt>BasicBlock</tt></a>s.<p>
1366 The <a href="#SymbolTable"><tt>SymbolTable</tt></a> is a very rarely used LLVM
1367 feature that is only used when you have to look up a value by name. Aside from
1368 that, the <a href="#SymbolTable"><tt>SymbolTable</tt></a> is used internally to
1369 make sure that there are not conflicts between the names of <a
1370 href="#Instruction"><tt>Instruction</tt></a>s, <a
1371 href="#BasicBlock"><tt>BasicBlock</tt></a>s, or <a
1372 href="#Argument"><tt>Argument</tt></a>s in the function body.<p>
1375 <!-- _______________________________________________________________________ -->
1376 </ul><h4><a name="m_Function"><hr size=0>Important Public Members of
1377 the <tt>Function</tt> class</h4><ul>
1379 <li><tt>Function(const <a href="#FunctionType">FunctionType</a> *Ty, bool isInternal, const std::string &N = "")</tt><p>
1381 Constructor used when you need to create new <tt>Function</tt>s to add the the
1382 program. The constructor must specify the type of the function to create and
1383 whether or not it should start out with internal or external linkage.<p>
1385 <li><tt>bool isExternal()</tt><p>
1387 Return whether or not the <tt>Function</tt> has a body defined. If the function
1388 is "external", it does not have a body, and thus must be resolved by linking
1389 with a function defined in a different translation unit.<p>
1392 <li><tt>Function::iterator</tt> - Typedef for basic block list iterator<br>
1393 <tt>Function::const_iterator</tt> - Typedef for const_iterator.<br>
1394 <tt>begin()</tt>, <tt>end()</tt>, <tt>front()</tt>, <tt>back()</tt>,
1395 <tt>size()</tt>, <tt>empty()</tt>, <tt>rbegin()</tt>, <tt>rend()</tt><p>
1397 These are forwarding methods that make it easy to access the contents of a
1398 <tt>Function</tt> object's <a href="#BasicBlock"><tt>BasicBlock</tt></a>
1401 <li><tt>Function::BasicBlockListType &getBasicBlockList()</tt><p>
1403 Returns the list of <a href="#BasicBlock"><tt>BasicBlock</tt></a>s. This is
1404 neccesary to use when you need to update the list or perform a complex action
1405 that doesn't have a forwarding method.<p>
1408 <li><tt>Function::aiterator</tt> - Typedef for the argument list iterator<br>
1409 <tt>Function::const_aiterator</tt> - Typedef for const_iterator.<br>
1410 <tt>abegin()</tt>, <tt>aend()</tt>, <tt>afront()</tt>, <tt>aback()</tt>,
1411 <tt>asize()</tt>, <tt>aempty()</tt>, <tt>arbegin()</tt>, <tt>arend()</tt><p>
1413 These are forwarding methods that make it easy to access the contents of a
1414 <tt>Function</tt> object's <a href="#Argument"><tt>Argument</tt></a> list.<p>
1416 <li><tt>Function::ArgumentListType &getArgumentList()</tt><p>
1418 Returns the list of <a href="#Argument"><tt>Argument</tt></a>s. This is
1419 neccesary to use when you need to update the list or perform a complex action
1420 that doesn't have a forwarding method.<p>
1424 <li><tt><a href="#BasicBlock">BasicBlock</a> &getEntryNode()</tt><p>
1426 Returns the entry <a href="#BasicBlock"><tt>BasicBlock</tt></a> for the
1427 function. Because the entry block for the function is always the first block,
1428 this returns the first block of the <tt>Function</tt>.<p>
1430 <li><tt><a href="#Type">Type</a> *getReturnType()</tt><br>
1431 <tt><a href="#FunctionType">FunctionType</a> *getFunctionType()</tt><p>
1433 This traverses the <a href="#Type"><tt>Type</tt></a> of the <tt>Function</tt>
1434 and returns the return type of the function, or the <a
1435 href="#FunctionType"><tt>FunctionType</tt></a> of the actual function.<p>
1438 <li><tt>bool hasSymbolTable() const</tt><p>
1440 Return true if the <tt>Function</tt> has a symbol table allocated to it and if
1441 there is at least one entry in it.<p>
1443 <li><tt><a href="#SymbolTable">SymbolTable</a> *getSymbolTable()</tt><p>
1445 Return a pointer to the <a href="#SymbolTable"><tt>SymbolTable</tt></a> for this
1446 <tt>Function</tt> or a null pointer if one has not been allocated (because there
1447 are no named values in the function).<p>
1449 <li><tt><a href="#SymbolTable">SymbolTable</a> *getSymbolTableSure()</tt><p>
1451 Return a pointer to the <a href="#SymbolTable"><tt>SymbolTable</tt></a> for this
1452 <tt>Function</tt> or allocate a new <a
1453 href="#SymbolTable"><tt>SymbolTable</tt></a> if one is not already around. This
1454 should only be used when adding elements to the <a
1455 href="#SymbolTable"><tt>SymbolTable</tt></a>, so that empty symbol tables are
1456 not left laying around.<p>
1460 <!-- ======================================================================= -->
1461 </ul><table width="100%" bgcolor="#441188" border=0 cellpadding=4 cellspacing=0>
1462 <tr><td> </td><td width="100%">
1463 <font color="#EEEEFF" face="Georgia,Palatino"><b>
1464 <a name="GlobalVariable">The <tt>GlobalVariable</tt> class</a>
1465 </b></font></td></tr></table><ul>
1468 href="/doxygen/GlobalVariable_8h-source.html">llvm/GlobalVariable.h</a>"</tt></b><br>
1469 doxygen info: <a href="/doxygen/classGlobalVariable.html">GlobalVariable Class</a><br>
1470 Superclasses: <a href="#GlobalValue"><tt>GlobalValue</tt></a>, <a
1471 href="#User"><tt>User</tt></a>, <a href="#Value"><tt>Value</tt></a><p>
1473 Global variables are represented with the (suprise suprise)
1474 <tt>GlobalVariable</tt> class. Like functions, <tt>GlobalVariable</tt>s are
1475 also subclasses of <a href="#GlobalValue"><tt>GlobalValue</tt></a>, and as such
1476 are always referenced by their address (global values must live in memory, so
1477 their "name" refers to their address). Global variables may have an initial
1478 value (which must be a <a href="#Constant"><tt>Constant</tt></a>), and if they
1479 have an initializer, they may be marked as "constant" themselves (indicating
1480 that their contents never change at runtime).<p>
1483 <!-- _______________________________________________________________________ -->
1484 </ul><h4><a name="m_GlobalVariable"><hr size=0>Important Public Members of the
1485 <tt>GlobalVariable</tt> class</h4><ul>
1487 <li><tt>GlobalVariable(const <a href="#Type">Type</a> *Ty, bool isConstant, bool
1488 isInternal, <a href="#Constant">Constant</a> *Initializer = 0, const std::string
1489 &Name = "")</tt><p>
1491 Create a new global variable of the specified type. If <tt>isConstant</tt> is
1492 true then the global variable will be marked as unchanging for the program, and
1493 if <tt>isInternal</tt> is true the resultant global variable will have internal
1494 linkage. Optionally an initializer and name may be specified for the global variable as well.<p>
1497 <li><tt>bool isConstant() const</tt><p>
1499 Returns true if this is a global variable is known not to be modified at
1503 <li><tt>bool hasInitializer()</tt><p>
1505 Returns true if this <tt>GlobalVariable</tt> has an intializer.<p>
1508 <li><tt><a href="#Constant">Constant</a> *getInitializer()</tt><p>
1510 Returns the intial value for a <tt>GlobalVariable</tt>. It is not legal to call
1511 this method if there is no initializer.<p>
1514 <!-- ======================================================================= -->
1515 </ul><table width="100%" bgcolor="#441188" border=0 cellpadding=4 cellspacing=0>
1516 <tr><td> </td><td width="100%">
1517 <font color="#EEEEFF" face="Georgia,Palatino"><b>
1518 <a name="Module">The <tt>Module</tt> class</a>
1519 </b></font></td></tr></table><ul>
1522 href="/doxygen/Module_8h-source.html">llvm/Module.h</a>"</tt></b><br>
1523 doxygen info: <a href="/doxygen/classModule.html">Module Class</a><p>
1525 The <tt>Module</tt> class represents the top level structure present in LLVM
1526 programs. An LLVM module is effectively either a translation unit of the
1527 original program or a combination of several translation units merged by the
1528 linker. The <tt>Module</tt> class keeps track of a list of <a
1529 href="#Function"><tt>Function</tt></a>s, a list of <a
1530 href="#GlobalVariable"><tt>GlobalVariable</tt></a>s, and a <a
1531 href="#SymbolTable"><tt>SymbolTable</tt></a>. Additionally, it contains a few
1532 helpful member functions that try to make common operations easy.<p>
1535 <!-- _______________________________________________________________________ -->
1536 </ul><h4><a name="m_Module"><hr size=0>Important Public Members of the
1537 <tt>Module</tt> class</h4><ul>
1539 <li><tt>Module::iterator</tt> - Typedef for function list iterator<br>
1540 <tt>Module::const_iterator</tt> - Typedef for const_iterator.<br>
1541 <tt>begin()</tt>, <tt>end()</tt>, <tt>front()</tt>, <tt>back()</tt>,
1542 <tt>size()</tt>, <tt>empty()</tt>, <tt>rbegin()</tt>, <tt>rend()</tt><p>
1544 These are forwarding methods that make it easy to access the contents of a
1545 <tt>Module</tt> object's <a href="#Function"><tt>Function</tt></a>
1548 <li><tt>Module::FunctionListType &getFunctionList()</tt><p>
1550 Returns the list of <a href="#Function"><tt>Function</tt></a>s. This is
1551 neccesary to use when you need to update the list or perform a complex action
1552 that doesn't have a forwarding method.<p>
1554 <!-- Global Variable -->
1557 <li><tt>Module::giterator</tt> - Typedef for global variable list iterator<br>
1558 <tt>Module::const_giterator</tt> - Typedef for const_iterator.<br>
1559 <tt>gbegin()</tt>, <tt>gend()</tt>, <tt>gfront()</tt>, <tt>gback()</tt>,
1560 <tt>gsize()</tt>, <tt>gempty()</tt>, <tt>grbegin()</tt>, <tt>grend()</tt><p>
1562 These are forwarding methods that make it easy to access the contents of a
1563 <tt>Module</tt> object's <a href="#GlobalVariable"><tt>GlobalVariable</tt></a>
1566 <li><tt>Module::GlobalListType &getGlobalList()</tt><p>
1568 Returns the list of <a href="#GlobalVariable"><tt>GlobalVariable</tt></a>s.
1569 This is neccesary to use when you need to update the list or perform a complex
1570 action that doesn't have a forwarding method.<p>
1573 <!-- Symbol table stuff -->
1576 <li><tt>bool hasSymbolTable() const</tt><p>
1578 Return true if the <tt>Module</tt> has a symbol table allocated to it and if
1579 there is at least one entry in it.<p>
1581 <li><tt><a href="#SymbolTable">SymbolTable</a> *getSymbolTable()</tt><p>
1583 Return a pointer to the <a href="#SymbolTable"><tt>SymbolTable</tt></a> for this
1584 <tt>Module</tt> or a null pointer if one has not been allocated (because there
1585 are no named values in the function).<p>
1587 <li><tt><a href="#SymbolTable">SymbolTable</a> *getSymbolTableSure()</tt><p>
1589 Return a pointer to the <a href="#SymbolTable"><tt>SymbolTable</tt></a> for this
1590 <tt>Module</tt> or allocate a new <a
1591 href="#SymbolTable"><tt>SymbolTable</tt></a> if one is not already around. This
1592 should only be used when adding elements to the <a
1593 href="#SymbolTable"><tt>SymbolTable</tt></a>, so that empty symbol tables are
1594 not left laying around.<p>
1597 <!-- Convenience methods -->
1600 <li><tt><a href="#Function">Function</a> *getFunction(const std::string &Name, const <a href="#FunctionType">FunctionType</a> *Ty)</tt><p>
1602 Look up the specified function in the <tt>Module</tt> <a
1603 href="#SymbolTable"><tt>SymbolTable</tt></a>. If it does not exist, return
1607 <li><tt><a href="#Function">Function</a> *getOrInsertFunction(const std::string
1608 &Name, const <a href="#FunctionType">FunctionType</a> *T)</tt><p>
1610 Look up the specified function in the <tt>Module</tt> <a
1611 href="#SymbolTable"><tt>SymbolTable</tt></a>. If it does not exist, add an
1612 external declaration for the function and return it.<p>
1615 <li><tt>std::string getTypeName(const <a href="#Type">Type</a> *Ty)</tt><p>
1617 If there is at least one entry in the <a
1618 href="#SymbolTable"><tt>SymbolTable</tt></a> for the specified <a
1619 href="#Type"><tt>Type</tt></a>, return it. Otherwise return the empty
1623 <li><tt>bool addTypeName(const std::string &Name, const <a href="#Type">Type</a>
1626 Insert an entry in the <a href="#SymbolTable"><tt>SymbolTable</tt></a> mapping
1627 <tt>Name</tt> to <tt>Ty</tt>. If there is already an entry for this name, true
1628 is returned and the <a href="#SymbolTable"><tt>SymbolTable</tt></a> is not
1632 <!-- ======================================================================= -->
1633 </ul><table width="100%" bgcolor="#441188" border=0 cellpadding=4 cellspacing=0>
1634 <tr><td> </td><td width="100%">
1635 <font color="#EEEEFF" face="Georgia,Palatino"><b>
1636 <a name="Constant">The <tt>Constant</tt> class and subclasses</a>
1637 </b></font></td></tr></table><ul>
1639 Constant represents a base class for different types of constants. It is
1640 subclassed by ConstantBool, ConstantInt, ConstantSInt, ConstantUInt,
1641 ConstantArray etc for representing the various types of Constants.<p>
1644 <!-- _______________________________________________________________________ -->
1645 </ul><h4><a name="m_Value"><hr size=0>Important Public Methods</h4><ul>
1647 <li><tt>bool isConstantExpr()</tt>: Returns true if it is a ConstantExpr
1651 Important Subclasses of Constant<p>
1654 <li>ConstantSInt : This subclass of Constant represents a signed integer constant.
1656 <li><tt>int64_t getValue() const</tt>: Returns the underlying value of this constant.
1658 <li>ConstantUInt : This class represents an unsigned integer.
1660 <li><tt>uint64_t getValue() const</tt>: Returns the underlying value of this constant.
1662 <li>ConstantFP : This class represents a floating point constant.
1664 <li><tt>double getValue() const</tt>: Returns the underlying value of this constant.
1666 <li>ConstantBool : This represents a boolean constant.
1668 <li><tt>bool getValue() const</tt>: Returns the underlying value of this constant.
1670 <li>ConstantArray : This represents a constant array.
1672 <li><tt>const std::vector<Use> &getValues() const</tt>: Returns a Vecotr of component constants that makeup this array.
1674 <li>ConstantStruct : This represents a constant struct.
1676 <li><tt>const std::vector<Use> &getValues() const</tt>: Returns a Vecotr of component constants that makeup this array.
1678 <li>ConstantPointerRef : This represents a constant pointer value that is initialized to point to a global value, which lies at a constant fixed address.
1680 <li><tt>GlobalValue *getValue()</tt>: Returns the global value to which this pointer is pointing to.
1685 <!-- ======================================================================= -->
1686 </ul><table width="100%" bgcolor="#441188" border=0 cellpadding=4 cellspacing=0>
1687 <tr><td> </td><td width="100%">
1688 <font color="#EEEEFF" face="Georgia,Palatino"><b>
1689 <a name="Type">The <tt>Type</tt> class and Derived Types</a>
1690 </b></font></td></tr></table><ul>
1692 Type as noted earlier is also a subclass of a Value class. Any primitive
1693 type (like int, short etc) in LLVM is an instance of Type Class. All
1694 other types are instances of subclasses of type like FunctionType,
1695 ArrayType etc. DerivedType is the interface for all such dervied types
1696 including FunctionType, ArrayType, PointerType, StructType. Types can have
1697 names. They can be recursive (StructType). There exists exactly one instance
1698 of any type structure at a time. This allows using pointer equality of Type *s for comparing types.
1700 <!-- _______________________________________________________________________ -->
1701 </ul><h4><a name="m_Value"><hr size=0>Important Public Methods</h4><ul>
1703 <li><tt>PrimitiveID getPrimitiveID() const</tt>: Returns the base type of the type.
1704 <li><tt> bool isSigned() const</tt>: Returns whether an integral numeric type is signed. This is true for SByteTy, ShortTy, IntTy, LongTy. Note that this is not true for Float and Double.
1705 <li><tt>bool isUnsigned() const</tt>: Returns whether a numeric type is unsigned. This is not quite the complement of isSigned... nonnumeric types return false as they do with isSigned. This returns true for UByteTy, UShortTy, UIntTy, and ULongTy.
1706 <li><tt> bool isInteger() const</tt>: Equilivent to isSigned() || isUnsigned(), but with only a single virtual function invocation.
1707 <li><tt>bool isIntegral() const</tt>: Returns true if this is an integral type, which is either Bool type or one of the Integer types.
1709 <li><tt>bool isFloatingPoint()</tt>: Return true if this is one of the two floating point types.
1710 <li><tt>bool isRecursive() const</tt>: Returns rue if the type graph contains a cycle.
1711 <li><tt>isLosslesslyConvertableTo (const Type *Ty) const</tt>: Return true if this type can be converted to 'Ty' without any reinterpretation of bits. For example, uint to int.
1712 <li><tt>bool isPrimitiveType() const</tt>: Returns true if it is a primitive type.
1713 <li><tt>bool isDerivedType() const</tt>: Returns true if it is a derived type.
1714 <li><tt>const Type * getContainedType (unsigned i) const</tt>:
1715 This method is used to implement the type iterator. For derived types, this returns the types 'contained' in the derived type, returning 0 when 'i' becomes invalid. This allows the user to iterate over the types in a struct, for example, really easily.
1716 <li><tt>unsigned getNumContainedTypes() const</tt>: Return the number of types in the derived type.
1724 <li>SequentialType : This is subclassed by ArrayType and PointerType
1726 <li><tt>const Type * getElementType() const</tt>: Returns the type of each of the elements in the sequential type.
1728 <li>ArrayType : This is a subclass of SequentialType and defines interface for array types.
1730 <li><tt>unsigned getNumElements() const</tt>: Returns the number of elements in the array.
1732 <li>PointerType : Subclass of SequentialType for pointer types.
1733 <li>StructType : subclass of DerivedTypes for struct types
1734 <li>FunctionType : subclass of DerivedTypes for function types.
1738 <li><tt>bool isVarArg() const</tt>: Returns true if its a vararg function
1739 <li><tt> const Type * getReturnType() const</tt>: Returns the return type of the function.
1740 <li><tt> const ParamTypes &getParamTypes() const</tt>: Returns a vector of parameter types.
1741 <li><tt>const Type * getParamType (unsigned i)</tt>: Returns the type of the ith parameter.
1742 <li><tt> const unsigned getNumParams() const</tt>: Returns the number of formal parameters.
1749 <!-- ======================================================================= -->
1750 </ul><table width="100%" bgcolor="#441188" border=0 cellpadding=4 cellspacing=0>
1751 <tr><td> </td><td width="100%">
1752 <font color="#EEEEFF" face="Georgia,Palatino"><b>
1753 <a name="Argument">The <tt>Argument</tt> class</a>
1754 </b></font></td></tr></table><ul>
1756 This subclass of Value defines the interface for incoming formal arguments to a
1757 function. A Function maitanis a list of its formal arguments. An argument has a
1758 pointer to the parent Function.
1763 <!-- *********************************************************************** -->
1765 <!-- *********************************************************************** -->
1768 <address>By: <a href="mailto:dhurjati@cs.uiuc.edu">Dinakar Dhurjati</a> and
1769 <a href="mailto:sabre@nondot.org">Chris Lattner</a></address>
1770 <!-- Created: Tue Aug 6 15:00:33 CDT 2002 -->
1771 <!-- hhmts start -->
1772 Last modified: Sun Oct 20 21:37:06 CDT 2002
1774 </font></body></html>