1 <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"
2 "http://www.w3.org/TR/html4/strict.dtd">
5 <title>LLVM Programmer's Manual</title>
6 <link rel="stylesheet" href="llvm.css" type="text/css">
10 <div class="doc_title">
11 LLVM Programmer's Manual
15 <li><a href="#introduction">Introduction</a></li>
16 <li><a href="#general">General Information</a>
18 <li><a href="#stl">The C++ Standard Template Library</a></li>
20 <li>The <tt>-time-passes</tt> option</li>
21 <li>How to use the LLVM Makefile system</li>
22 <li>How to write a regression test</li>
27 <li><a href="#apis">Important and useful LLVM APIs</a>
29 <li><a href="#isa">The <tt>isa<></tt>, <tt>cast<></tt>
30 and <tt>dyn_cast<></tt> templates</a> </li>
31 <li><a href="#DEBUG">The <tt>DEBUG()</tt> macro and <tt>-debug</tt>
34 <li><a href="#DEBUG_TYPE">Fine grained debug info with <tt>DEBUG_TYPE</tt>
35 and the <tt>-debug-only</tt> option</a> </li>
38 <li><a href="#Statistic">The <tt>Statistic</tt> class & <tt>-stats</tt>
41 <li>The <tt>InstVisitor</tt> template
42 <li>The general graph API
44 <li><a href="#ViewGraph">Viewing graphs while debugging code</a></li>
47 <li><a href="#datastructure">Picking the Right Data Structure for a Task</a>
49 <li><a href="#ds_sequential">Sequential Containers (std::vector, std::list, etc)</a>
51 <li><a href="#dss_fixedarrays">Fixed Size Arrays</a></li>
52 <li><a href="#dss_heaparrays">Heap Allocated Arrays</a></li>
53 <li><a href="#dss_smallvector">"llvm/ADT/SmallVector.h"</a></li>
54 <li><a href="#dss_vector"><vector></a></li>
55 <li><a href="#dss_deque"><deque></a></li>
56 <li><a href="#dss_list"><list></a></li>
57 <li><a href="#dss_ilist">llvm/ADT/ilist</a></li>
58 <li><a href="#dss_other">Other Sequential Container Options</a></li>
60 <li><a href="#ds_set">Set-Like Containers (std::set, SmallSet, SetVector, etc)</a>
62 <li><a href="#dss_sortedvectorset">A sorted 'vector'</a></li>
63 <li><a href="#dss_smallset">"llvm/ADT/SmallSet.h"</a></li>
64 <li><a href="#dss_smallptrset">"llvm/ADT/SmallPtrSet.h"</a></li>
65 <li><a href="#dss_FoldingSet">"llvm/ADT/FoldingSet.h"</a></li>
66 <li><a href="#dss_set"><set></a></li>
67 <li><a href="#dss_setvector">"llvm/ADT/SetVector.h"</a></li>
68 <li><a href="#dss_uniquevector">"llvm/ADT/UniqueVector.h"</a></li>
69 <li><a href="#dss_otherset">Other Set-Like ContainerOptions</a></li>
71 <li><a href="#ds_map">Map-Like Containers (std::map, DenseMap, etc)</a>
73 <li><a href="#dss_sortedvectormap">A sorted 'vector'</a></li>
74 <li><a href="#dss_cstringmap">"llvm/ADT/CStringMap.h"</a></li>
75 <li><a href="#dss_indexedmap">"llvm/ADT/IndexedMap.h"</a></li>
76 <li><a href="#dss_densemap">"llvm/ADT/DenseMap.h"</a></li>
77 <li><a href="#dss_map"><map></a></li>
78 <li><a href="#dss_othermap">Other Map-Like Container Options</a></li>
82 <li><a href="#common">Helpful Hints for Common Operations</a>
84 <li><a href="#inspection">Basic Inspection and Traversal Routines</a>
86 <li><a href="#iterate_function">Iterating over the <tt>BasicBlock</tt>s
87 in a <tt>Function</tt></a> </li>
88 <li><a href="#iterate_basicblock">Iterating over the <tt>Instruction</tt>s
89 in a <tt>BasicBlock</tt></a> </li>
90 <li><a href="#iterate_institer">Iterating over the <tt>Instruction</tt>s
91 in a <tt>Function</tt></a> </li>
92 <li><a href="#iterate_convert">Turning an iterator into a
93 class pointer</a> </li>
94 <li><a href="#iterate_complex">Finding call sites: a more
95 complex example</a> </li>
96 <li><a href="#calls_and_invokes">Treating calls and invokes
97 the same way</a> </li>
98 <li><a href="#iterate_chains">Iterating over def-use &
99 use-def chains</a> </li>
102 <li><a href="#simplechanges">Making simple changes</a>
104 <li><a href="#schanges_creating">Creating and inserting new
105 <tt>Instruction</tt>s</a> </li>
106 <li><a href="#schanges_deleting">Deleting <tt>Instruction</tt>s</a> </li>
107 <li><a href="#schanges_replacing">Replacing an <tt>Instruction</tt>
108 with another <tt>Value</tt></a> </li>
112 <li>Working with the Control Flow Graph
114 <li>Accessing predecessors and successors of a <tt>BasicBlock</tt>
122 <li><a href="#advanced">Advanced Topics</a>
124 <li><a href="#TypeResolve">LLVM Type Resolution</a>
126 <li><a href="#BuildRecType">Basic Recursive Type Construction</a></li>
127 <li><a href="#refineAbstractTypeTo">The <tt>refineAbstractTypeTo</tt> method</a></li>
128 <li><a href="#PATypeHolder">The PATypeHolder Class</a></li>
129 <li><a href="#AbstractTypeUser">The AbstractTypeUser Class</a></li>
132 <li><a href="#SymbolTable">The <tt>SymbolTable</tt> class </a></li>
135 <li><a href="#coreclasses">The Core LLVM Class Hierarchy Reference</a>
137 <li><a href="#Type">The <tt>Type</tt> class</a> </li>
138 <li><a href="#Module">The <tt>Module</tt> class</a></li>
139 <li><a href="#Value">The <tt>Value</tt> class</a>
141 <li><a href="#User">The <tt>User</tt> class</a>
143 <li><a href="#Instruction">The <tt>Instruction</tt> class</a></li>
144 <li><a href="#Constant">The <tt>Constant</tt> class</a>
146 <li><a href="#GlobalValue">The <tt>GlobalValue</tt> class</a>
148 <li><a href="#Function">The <tt>Function</tt> class</a></li>
149 <li><a href="#GlobalVariable">The <tt>GlobalVariable</tt> class</a></li>
156 <li><a href="#BasicBlock">The <tt>BasicBlock</tt> class</a></li>
157 <li><a href="#Argument">The <tt>Argument</tt> class</a></li>
164 <div class="doc_author">
165 <p>Written by <a href="mailto:sabre@nondot.org">Chris Lattner</a>,
166 <a href="mailto:dhurjati@cs.uiuc.edu">Dinakar Dhurjati</a>,
167 <a href="mailto:jstanley@cs.uiuc.edu">Joel Stanley</a>, and
168 <a href="mailto:rspencer@x10sys.com">Reid Spencer</a></p>
171 <!-- *********************************************************************** -->
172 <div class="doc_section">
173 <a name="introduction">Introduction </a>
175 <!-- *********************************************************************** -->
177 <div class="doc_text">
179 <p>This document is meant to highlight some of the important classes and
180 interfaces available in the LLVM source-base. This manual is not
181 intended to explain what LLVM is, how it works, and what LLVM code looks
182 like. It assumes that you know the basics of LLVM and are interested
183 in writing transformations or otherwise analyzing or manipulating the
186 <p>This document should get you oriented so that you can find your
187 way in the continuously growing source code that makes up the LLVM
188 infrastructure. Note that this manual is not intended to serve as a
189 replacement for reading the source code, so if you think there should be
190 a method in one of these classes to do something, but it's not listed,
191 check the source. Links to the <a href="/doxygen/">doxygen</a> sources
192 are provided to make this as easy as possible.</p>
194 <p>The first section of this document describes general information that is
195 useful to know when working in the LLVM infrastructure, and the second describes
196 the Core LLVM classes. In the future this manual will be extended with
197 information describing how to use extension libraries, such as dominator
198 information, CFG traversal routines, and useful utilities like the <tt><a
199 href="/doxygen/InstVisitor_8h-source.html">InstVisitor</a></tt> template.</p>
203 <!-- *********************************************************************** -->
204 <div class="doc_section">
205 <a name="general">General Information</a>
207 <!-- *********************************************************************** -->
209 <div class="doc_text">
211 <p>This section contains general information that is useful if you are working
212 in the LLVM source-base, but that isn't specific to any particular API.</p>
216 <!-- ======================================================================= -->
217 <div class="doc_subsection">
218 <a name="stl">The C++ Standard Template Library</a>
221 <div class="doc_text">
223 <p>LLVM makes heavy use of the C++ Standard Template Library (STL),
224 perhaps much more than you are used to, or have seen before. Because of
225 this, you might want to do a little background reading in the
226 techniques used and capabilities of the library. There are many good
227 pages that discuss the STL, and several books on the subject that you
228 can get, so it will not be discussed in this document.</p>
230 <p>Here are some useful links:</p>
234 <li><a href="http://www.dinkumware.com/refxcpp.html">Dinkumware C++ Library
235 reference</a> - an excellent reference for the STL and other parts of the
236 standard C++ library.</li>
238 <li><a href="http://www.tempest-sw.com/cpp/">C++ In a Nutshell</a> - This is an
239 O'Reilly book in the making. It has a decent
241 Reference that rivals Dinkumware's, and is unfortunately no longer free since the book has been
244 <li><a href="http://www.parashift.com/c++-faq-lite/">C++ Frequently Asked
247 <li><a href="http://www.sgi.com/tech/stl/">SGI's STL Programmer's Guide</a> -
249 href="http://www.sgi.com/tech/stl/stl_introduction.html">Introduction to the
252 <li><a href="http://www.research.att.com/%7Ebs/C++.html">Bjarne Stroustrup's C++
255 <li><a href="http://64.78.49.204/">
256 Bruce Eckel's Thinking in C++, 2nd ed. Volume 2 Revision 4.0 (even better, get
261 <p>You are also encouraged to take a look at the <a
262 href="CodingStandards.html">LLVM Coding Standards</a> guide which focuses on how
263 to write maintainable code more than where to put your curly braces.</p>
267 <!-- ======================================================================= -->
268 <div class="doc_subsection">
269 <a name="stl">Other useful references</a>
272 <div class="doc_text">
275 <li><a href="http://www.psc.edu/%7Esemke/cvs_branches.html">CVS
276 Branch and Tag Primer</a></li>
277 <li><a href="http://www.fortran-2000.com/ArnaudRecipes/sharedlib.html">Using
278 static and shared libraries across platforms</a></li>
283 <!-- *********************************************************************** -->
284 <div class="doc_section">
285 <a name="apis">Important and useful LLVM APIs</a>
287 <!-- *********************************************************************** -->
289 <div class="doc_text">
291 <p>Here we highlight some LLVM APIs that are generally useful and good to
292 know about when writing transformations.</p>
296 <!-- ======================================================================= -->
297 <div class="doc_subsection">
298 <a name="isa">The <tt>isa<></tt>, <tt>cast<></tt> and
299 <tt>dyn_cast<></tt> templates</a>
302 <div class="doc_text">
304 <p>The LLVM source-base makes extensive use of a custom form of RTTI.
305 These templates have many similarities to the C++ <tt>dynamic_cast<></tt>
306 operator, but they don't have some drawbacks (primarily stemming from
307 the fact that <tt>dynamic_cast<></tt> only works on classes that
308 have a v-table). Because they are used so often, you must know what they
309 do and how they work. All of these templates are defined in the <a
310 href="/doxygen/Casting_8h-source.html"><tt>llvm/Support/Casting.h</tt></a>
311 file (note that you very rarely have to include this file directly).</p>
314 <dt><tt>isa<></tt>: </dt>
316 <dd><p>The <tt>isa<></tt> operator works exactly like the Java
317 "<tt>instanceof</tt>" operator. It returns true or false depending on whether
318 a reference or pointer points to an instance of the specified class. This can
319 be very useful for constraint checking of various sorts (example below).</p>
322 <dt><tt>cast<></tt>: </dt>
324 <dd><p>The <tt>cast<></tt> operator is a "checked cast" operation. It
325 converts a pointer or reference from a base class to a derived cast, causing
326 an assertion failure if it is not really an instance of the right type. This
327 should be used in cases where you have some information that makes you believe
328 that something is of the right type. An example of the <tt>isa<></tt>
329 and <tt>cast<></tt> template is:</p>
331 <div class="doc_code">
333 static bool isLoopInvariant(const <a href="#Value">Value</a> *V, const Loop *L) {
334 if (isa<<a href="#Constant">Constant</a>>(V) || isa<<a href="#Argument">Argument</a>>(V) || isa<<a href="#GlobalValue">GlobalValue</a>>(V))
337 // <i>Otherwise, it must be an instruction...</i>
338 return !L->contains(cast<<a href="#Instruction">Instruction</a>>(V)->getParent());
343 <p>Note that you should <b>not</b> use an <tt>isa<></tt> test followed
344 by a <tt>cast<></tt>, for that use the <tt>dyn_cast<></tt>
349 <dt><tt>dyn_cast<></tt>:</dt>
351 <dd><p>The <tt>dyn_cast<></tt> operator is a "checking cast" operation.
352 It checks to see if the operand is of the specified type, and if so, returns a
353 pointer to it (this operator does not work with references). If the operand is
354 not of the correct type, a null pointer is returned. Thus, this works very
355 much like the <tt>dynamic_cast<></tt> operator in C++, and should be
356 used in the same circumstances. Typically, the <tt>dyn_cast<></tt>
357 operator is used in an <tt>if</tt> statement or some other flow control
358 statement like this:</p>
360 <div class="doc_code">
362 if (<a href="#AllocationInst">AllocationInst</a> *AI = dyn_cast<<a href="#AllocationInst">AllocationInst</a>>(Val)) {
368 <p>This form of the <tt>if</tt> statement effectively combines together a call
369 to <tt>isa<></tt> and a call to <tt>cast<></tt> into one
370 statement, which is very convenient.</p>
372 <p>Note that the <tt>dyn_cast<></tt> operator, like C++'s
373 <tt>dynamic_cast<></tt> or Java's <tt>instanceof</tt> operator, can be
374 abused. In particular, you should not use big chained <tt>if/then/else</tt>
375 blocks to check for lots of different variants of classes. If you find
376 yourself wanting to do this, it is much cleaner and more efficient to use the
377 <tt>InstVisitor</tt> class to dispatch over the instruction type directly.</p>
381 <dt><tt>cast_or_null<></tt>: </dt>
383 <dd><p>The <tt>cast_or_null<></tt> operator works just like the
384 <tt>cast<></tt> operator, except that it allows for a null pointer as an
385 argument (which it then propagates). This can sometimes be useful, allowing
386 you to combine several null checks into one.</p></dd>
388 <dt><tt>dyn_cast_or_null<></tt>: </dt>
390 <dd><p>The <tt>dyn_cast_or_null<></tt> operator works just like the
391 <tt>dyn_cast<></tt> operator, except that it allows for a null pointer
392 as an argument (which it then propagates). This can sometimes be useful,
393 allowing you to combine several null checks into one.</p></dd>
397 <p>These five templates can be used with any classes, whether they have a
398 v-table or not. To add support for these templates, you simply need to add
399 <tt>classof</tt> static methods to the class you are interested casting
400 to. Describing this is currently outside the scope of this document, but there
401 are lots of examples in the LLVM source base.</p>
405 <!-- ======================================================================= -->
406 <div class="doc_subsection">
407 <a name="DEBUG">The <tt>DEBUG()</tt> macro and <tt>-debug</tt> option</a>
410 <div class="doc_text">
412 <p>Often when working on your pass you will put a bunch of debugging printouts
413 and other code into your pass. After you get it working, you want to remove
414 it, but you may need it again in the future (to work out new bugs that you run
417 <p> Naturally, because of this, you don't want to delete the debug printouts,
418 but you don't want them to always be noisy. A standard compromise is to comment
419 them out, allowing you to enable them if you need them in the future.</p>
421 <p>The "<tt><a href="/doxygen/Debug_8h-source.html">llvm/Support/Debug.h</a></tt>"
422 file provides a macro named <tt>DEBUG()</tt> that is a much nicer solution to
423 this problem. Basically, you can put arbitrary code into the argument of the
424 <tt>DEBUG</tt> macro, and it is only executed if '<tt>opt</tt>' (or any other
425 tool) is run with the '<tt>-debug</tt>' command line argument:</p>
427 <div class="doc_code">
429 DOUT << "I am here!\n";
433 <p>Then you can run your pass like this:</p>
435 <div class="doc_code">
437 $ opt < a.bc > /dev/null -mypass
438 <i><no output></i>
439 $ opt < a.bc > /dev/null -mypass -debug
444 <p>Using the <tt>DEBUG()</tt> macro instead of a home-brewed solution allows you
445 to not have to create "yet another" command line option for the debug output for
446 your pass. Note that <tt>DEBUG()</tt> macros are disabled for optimized builds,
447 so they do not cause a performance impact at all (for the same reason, they
448 should also not contain side-effects!).</p>
450 <p>One additional nice thing about the <tt>DEBUG()</tt> macro is that you can
451 enable or disable it directly in gdb. Just use "<tt>set DebugFlag=0</tt>" or
452 "<tt>set DebugFlag=1</tt>" from the gdb if the program is running. If the
453 program hasn't been started yet, you can always just run it with
458 <!-- _______________________________________________________________________ -->
459 <div class="doc_subsubsection">
460 <a name="DEBUG_TYPE">Fine grained debug info with <tt>DEBUG_TYPE</tt> and
461 the <tt>-debug-only</tt> option</a>
464 <div class="doc_text">
466 <p>Sometimes you may find yourself in a situation where enabling <tt>-debug</tt>
467 just turns on <b>too much</b> information (such as when working on the code
468 generator). If you want to enable debug information with more fine-grained
469 control, you define the <tt>DEBUG_TYPE</tt> macro and the <tt>-debug</tt> only
470 option as follows:</p>
472 <div class="doc_code">
474 DOUT << "No debug type\n";
476 #define DEBUG_TYPE "foo"
477 DOUT << "'foo' debug type\n";
479 #define DEBUG_TYPE "bar"
480 DOUT << "'bar' debug type\n";
482 #define DEBUG_TYPE ""
483 DOUT << "No debug type (2)\n";
487 <p>Then you can run your pass like this:</p>
489 <div class="doc_code">
491 $ opt < a.bc > /dev/null -mypass
492 <i><no output></i>
493 $ opt < a.bc > /dev/null -mypass -debug
498 $ opt < a.bc > /dev/null -mypass -debug-only=foo
500 $ opt < a.bc > /dev/null -mypass -debug-only=bar
505 <p>Of course, in practice, you should only set <tt>DEBUG_TYPE</tt> at the top of
506 a file, to specify the debug type for the entire module (if you do this before
507 you <tt>#include "llvm/Support/Debug.h"</tt>, you don't have to insert the ugly
508 <tt>#undef</tt>'s). Also, you should use names more meaningful than "foo" and
509 "bar", because there is no system in place to ensure that names do not
510 conflict. If two different modules use the same string, they will all be turned
511 on when the name is specified. This allows, for example, all debug information
512 for instruction scheduling to be enabled with <tt>-debug-type=InstrSched</tt>,
513 even if the source lives in multiple files.</p>
517 <!-- ======================================================================= -->
518 <div class="doc_subsection">
519 <a name="Statistic">The <tt>Statistic</tt> class & <tt>-stats</tt>
523 <div class="doc_text">
526 href="/doxygen/Statistic_8h-source.html">llvm/ADT/Statistic.h</a></tt>" file
527 provides a class named <tt>Statistic</tt> that is used as a unified way to
528 keep track of what the LLVM compiler is doing and how effective various
529 optimizations are. It is useful to see what optimizations are contributing to
530 making a particular program run faster.</p>
532 <p>Often you may run your pass on some big program, and you're interested to see
533 how many times it makes a certain transformation. Although you can do this with
534 hand inspection, or some ad-hoc method, this is a real pain and not very useful
535 for big programs. Using the <tt>Statistic</tt> class makes it very easy to
536 keep track of this information, and the calculated information is presented in a
537 uniform manner with the rest of the passes being executed.</p>
539 <p>There are many examples of <tt>Statistic</tt> uses, but the basics of using
540 it are as follows:</p>
543 <li><p>Define your statistic like this:</p>
545 <div class="doc_code">
547 #define <a href="#DEBUG_TYPE">DEBUG_TYPE</a> "mypassname" <i>// This goes before any #includes.</i>
548 STATISTIC(NumXForms, "The # of times I did stuff");
552 <p>The <tt>STATISTIC</tt> macro defines a static variable, whose name is
553 specified by the first argument. The pass name is taken from the DEBUG_TYPE
554 macro, and the description is taken from the second argument. The variable
555 defined ("NumXForms" in this case) acts like an unsigned integer.</p></li>
557 <li><p>Whenever you make a transformation, bump the counter:</p>
559 <div class="doc_code">
561 ++NumXForms; // <i>I did stuff!</i>
568 <p>That's all you have to do. To get '<tt>opt</tt>' to print out the
569 statistics gathered, use the '<tt>-stats</tt>' option:</p>
571 <div class="doc_code">
573 $ opt -stats -mypassname < program.bc > /dev/null
574 <i>... statistics output ...</i>
578 <p> When running <tt>gccas</tt> on a C file from the SPEC benchmark
579 suite, it gives a report that looks like this:</p>
581 <div class="doc_code">
583 7646 bytecodewriter - Number of normal instructions
584 725 bytecodewriter - Number of oversized instructions
585 129996 bytecodewriter - Number of bytecode bytes written
586 2817 raise - Number of insts DCEd or constprop'd
587 3213 raise - Number of cast-of-self removed
588 5046 raise - Number of expression trees converted
589 75 raise - Number of other getelementptr's formed
590 138 raise - Number of load/store peepholes
591 42 deadtypeelim - Number of unused typenames removed from symtab
592 392 funcresolve - Number of varargs functions resolved
593 27 globaldce - Number of global variables removed
594 2 adce - Number of basic blocks removed
595 134 cee - Number of branches revectored
596 49 cee - Number of setcc instruction eliminated
597 532 gcse - Number of loads removed
598 2919 gcse - Number of instructions removed
599 86 indvars - Number of canonical indvars added
600 87 indvars - Number of aux indvars removed
601 25 instcombine - Number of dead inst eliminate
602 434 instcombine - Number of insts combined
603 248 licm - Number of load insts hoisted
604 1298 licm - Number of insts hoisted to a loop pre-header
605 3 licm - Number of insts hoisted to multiple loop preds (bad, no loop pre-header)
606 75 mem2reg - Number of alloca's promoted
607 1444 cfgsimplify - Number of blocks simplified
611 <p>Obviously, with so many optimizations, having a unified framework for this
612 stuff is very nice. Making your pass fit well into the framework makes it more
613 maintainable and useful.</p>
617 <!-- ======================================================================= -->
618 <div class="doc_subsection">
619 <a name="ViewGraph">Viewing graphs while debugging code</a>
622 <div class="doc_text">
624 <p>Several of the important data structures in LLVM are graphs: for example
625 CFGs made out of LLVM <a href="#BasicBlock">BasicBlock</a>s, CFGs made out of
626 LLVM <a href="CodeGenerator.html#machinebasicblock">MachineBasicBlock</a>s, and
627 <a href="CodeGenerator.html#selectiondag_intro">Instruction Selection
628 DAGs</a>. In many cases, while debugging various parts of the compiler, it is
629 nice to instantly visualize these graphs.</p>
631 <p>LLVM provides several callbacks that are available in a debug build to do
632 exactly that. If you call the <tt>Function::viewCFG()</tt> method, for example,
633 the current LLVM tool will pop up a window containing the CFG for the function
634 where each basic block is a node in the graph, and each node contains the
635 instructions in the block. Similarly, there also exists
636 <tt>Function::viewCFGOnly()</tt> (does not include the instructions), the
637 <tt>MachineFunction::viewCFG()</tt> and <tt>MachineFunction::viewCFGOnly()</tt>,
638 and the <tt>SelectionDAG::viewGraph()</tt> methods. Within GDB, for example,
639 you can usually use something like <tt>call DAG.viewGraph()</tt> to pop
640 up a window. Alternatively, you can sprinkle calls to these functions in your
641 code in places you want to debug.</p>
643 <p>Getting this to work requires a small amount of configuration. On Unix
644 systems with X11, install the <a href="http://www.graphviz.org">graphviz</a>
645 toolkit, and make sure 'dot' and 'gv' are in your path. If you are running on
646 Mac OS/X, download and install the Mac OS/X <a
647 href="http://www.pixelglow.com/graphviz/">Graphviz program</a>, and add
648 <tt>/Applications/Graphviz.app/Contents/MacOS/</tt> (or wherever you install
649 it) to your path. Once in your system and path are set up, rerun the LLVM
650 configure script and rebuild LLVM to enable this functionality.</p>
652 <p><tt>SelectionDAG</tt> has been extended to make it easier to locate
653 <i>interesting</i> nodes in large complex graphs. From gdb, if you
654 <tt>call DAG.setGraphColor(<i>node</i>, "<i>color</i>")</tt>, then the
655 next <tt>call DAG.viewGraph()</tt> would highlight the node in the
656 specified color (choices of colors can be found at <a
657 href="http://www.graphviz.org/doc/info/colors.html">colors</a>.) More
658 complex node attributes can be provided with <tt>call
659 DAG.setGraphAttrs(<i>node</i>, "<i>attributes</i>")</tt> (choices can be
660 found at <a href="http://www.graphviz.org/doc/info/attrs.html">Graph
661 Attributes</a>.) If you want to restart and clear all the current graph
662 attributes, then you can <tt>call DAG.clearGraphAttrs()</tt>. </p>
666 <!-- *********************************************************************** -->
667 <div class="doc_section">
668 <a name="datastructure">Picking the Right Data Structure for a Task</a>
670 <!-- *********************************************************************** -->
672 <div class="doc_text">
674 <p>LLVM has a plethora of data structures in the <tt>llvm/ADT/</tt> directory,
675 and we commonly use STL data structures. This section describes the trade-offs
676 you should consider when you pick one.</p>
679 The first step is a choose your own adventure: do you want a sequential
680 container, a set-like container, or a map-like container? The most important
681 thing when choosing a container is the algorithmic properties of how you plan to
682 access the container. Based on that, you should use:</p>
685 <li>a <a href="#ds_map">map-like</a> container if you need efficient look-up
686 of an value based on another value. Map-like containers also support
687 efficient queries for containment (whether a key is in the map). Map-like
688 containers generally do not support efficient reverse mapping (values to
689 keys). If you need that, use two maps. Some map-like containers also
690 support efficient iteration through the keys in sorted order. Map-like
691 containers are the most expensive sort, only use them if you need one of
692 these capabilities.</li>
694 <li>a <a href="#ds_set">set-like</a> container if you need to put a bunch of
695 stuff into a container that automatically eliminates duplicates. Some
696 set-like containers support efficient iteration through the elements in
697 sorted order. Set-like containers are more expensive than sequential
701 <li>a <a href="#ds_sequential">sequential</a> container provides
702 the most efficient way to add elements and keeps track of the order they are
703 added to the collection. They permit duplicates and support efficient
704 iteration, but do not support efficient look-up based on a key.
710 Once the proper category of container is determined, you can fine tune the
711 memory use, constant factors, and cache behaviors of access by intelligently
712 picking a member of the category. Note that constant factors and cache behavior
713 can be a big deal. If you have a vector that usually only contains a few
714 elements (but could contain many), for example, it's much better to use
715 <a href="#dss_smallvector">SmallVector</a> than <a href="#dss_vector">vector</a>
716 . Doing so avoids (relatively) expensive malloc/free calls, which dwarf the
717 cost of adding the elements to the container. </p>
721 <!-- ======================================================================= -->
722 <div class="doc_subsection">
723 <a name="ds_sequential">Sequential Containers (std::vector, std::list, etc)</a>
726 <div class="doc_text">
727 There are a variety of sequential containers available for you, based on your
728 needs. Pick the first in this section that will do what you want.
731 <!-- _______________________________________________________________________ -->
732 <div class="doc_subsubsection">
733 <a name="dss_fixedarrays">Fixed Size Arrays</a>
736 <div class="doc_text">
737 <p>Fixed size arrays are very simple and very fast. They are good if you know
738 exactly how many elements you have, or you have a (low) upper bound on how many
742 <!-- _______________________________________________________________________ -->
743 <div class="doc_subsubsection">
744 <a name="dss_heaparrays">Heap Allocated Arrays</a>
747 <div class="doc_text">
748 <p>Heap allocated arrays (new[] + delete[]) are also simple. They are good if
749 the number of elements is variable, if you know how many elements you will need
750 before the array is allocated, and if the array is usually large (if not,
751 consider a <a href="#dss_smallvector">SmallVector</a>). The cost of a heap
752 allocated array is the cost of the new/delete (aka malloc/free). Also note that
753 if you are allocating an array of a type with a constructor, the constructor and
754 destructors will be run for every element in the array (re-sizable vectors only
755 construct those elements actually used).</p>
758 <!-- _______________________________________________________________________ -->
759 <div class="doc_subsubsection">
760 <a name="dss_smallvector">"llvm/ADT/SmallVector.h"</a>
763 <div class="doc_text">
764 <p><tt>SmallVector<Type, N></tt> is a simple class that looks and smells
765 just like <tt>vector<Type></tt>:
766 it supports efficient iteration, lays out elements in memory order (so you can
767 do pointer arithmetic between elements), supports efficient push_back/pop_back
768 operations, supports efficient random access to its elements, etc.</p>
770 <p>The advantage of SmallVector is that it allocates space for
771 some number of elements (N) <b>in the object itself</b>. Because of this, if
772 the SmallVector is dynamically smaller than N, no malloc is performed. This can
773 be a big win in cases where the malloc/free call is far more expensive than the
774 code that fiddles around with the elements.</p>
776 <p>This is good for vectors that are "usually small" (e.g. the number of
777 predecessors/successors of a block is usually less than 8). On the other hand,
778 this makes the size of the SmallVector itself large, so you don't want to
779 allocate lots of them (doing so will waste a lot of space). As such,
780 SmallVectors are most useful when on the stack.</p>
782 <p>SmallVector also provides a nice portable and efficient replacement for
787 <!-- _______________________________________________________________________ -->
788 <div class="doc_subsubsection">
789 <a name="dss_vector"><vector></a>
792 <div class="doc_text">
794 std::vector is well loved and respected. It is useful when SmallVector isn't:
795 when the size of the vector is often large (thus the small optimization will
796 rarely be a benefit) or if you will be allocating many instances of the vector
797 itself (which would waste space for elements that aren't in the container).
798 vector is also useful when interfacing with code that expects vectors :).
802 <!-- _______________________________________________________________________ -->
803 <div class="doc_subsubsection">
804 <a name="dss_deque"><deque></a>
807 <div class="doc_text">
808 <p>std::deque is, in some senses, a generalized version of std::vector. Like
809 std::vector, it provides constant time random access and other similar
810 properties, but it also provides efficient access to the front of the list. It
811 does not guarantee continuity of elements within memory.</p>
813 <p>In exchange for this extra flexibility, std::deque has significantly higher
814 constant factor costs than std::vector. If possible, use std::vector or
815 something cheaper.</p>
818 <!-- _______________________________________________________________________ -->
819 <div class="doc_subsubsection">
820 <a name="dss_list"><list></a>
823 <div class="doc_text">
824 <p>std::list is an extremely inefficient class that is rarely useful.
825 It performs a heap allocation for every element inserted into it, thus having an
826 extremely high constant factor, particularly for small data types. std::list
827 also only supports bidirectional iteration, not random access iteration.</p>
829 <p>In exchange for this high cost, std::list supports efficient access to both
830 ends of the list (like std::deque, but unlike std::vector or SmallVector). In
831 addition, the iterator invalidation characteristics of std::list are stronger
832 than that of a vector class: inserting or removing an element into the list does
833 not invalidate iterator or pointers to other elements in the list.</p>
836 <!-- _______________________________________________________________________ -->
837 <div class="doc_subsubsection">
838 <a name="dss_ilist">llvm/ADT/ilist</a>
841 <div class="doc_text">
842 <p><tt>ilist<T></tt> implements an 'intrusive' doubly-linked list. It is
843 intrusive, because it requires the element to store and provide access to the
844 prev/next pointers for the list.</p>
846 <p>ilist has the same drawbacks as std::list, and additionally requires an
847 ilist_traits implementation for the element type, but it provides some novel
848 characteristics. In particular, it can efficiently store polymorphic objects,
849 the traits class is informed when an element is inserted or removed from the
850 list, and ilists are guaranteed to support a constant-time splice operation.
853 <p>These properties are exactly what we want for things like Instructions and
854 basic blocks, which is why these are implemented with ilists.</p>
857 <!-- _______________________________________________________________________ -->
858 <div class="doc_subsubsection">
859 <a name="dss_other">Other Sequential Container options</a>
862 <div class="doc_text">
863 <p>Other STL containers are available, such as std::string.</p>
865 <p>There are also various STL adapter classes such as std::queue,
866 std::priority_queue, std::stack, etc. These provide simplified access to an
867 underlying container but don't affect the cost of the container itself.</p>
872 <!-- ======================================================================= -->
873 <div class="doc_subsection">
874 <a name="ds_set">Set-Like Containers (std::set, SmallSet, SetVector, etc)</a>
877 <div class="doc_text">
879 <p>Set-like containers are useful when you need to canonicalize multiple values
880 into a single representation. There are several different choices for how to do
881 this, providing various trade-offs.</p>
886 <!-- _______________________________________________________________________ -->
887 <div class="doc_subsubsection">
888 <a name="dss_sortedvectorset">A sorted 'vector'</a>
891 <div class="doc_text">
893 <p>If you intend to insert a lot of elements, then do a lot of queries, a
894 great approach is to use a vector (or other sequential container) with
895 std::sort+std::unique to remove duplicates. This approach works really well if
896 your usage pattern has these two distinct phases (insert then query), and can be
897 coupled with a good choice of <a href="#ds_sequential">sequential container</a>.
901 This combination provides the several nice properties: the result data is
902 contiguous in memory (good for cache locality), has few allocations, is easy to
903 address (iterators in the final vector are just indices or pointers), and can be
904 efficiently queried with a standard binary or radix search.</p>
908 <!-- _______________________________________________________________________ -->
909 <div class="doc_subsubsection">
910 <a name="dss_smallset">"llvm/ADT/SmallSet.h"</a>
913 <div class="doc_text">
915 <p>If you have a set-like data structure that is usually small and whose elements
916 are reasonably small, a <tt>SmallSet<Type, N></tt> is a good choice. This set
917 has space for N elements in place (thus, if the set is dynamically smaller than
918 N, no malloc traffic is required) and accesses them with a simple linear search.
919 When the set grows beyond 'N' elements, it allocates a more expensive representation that
920 guarantees efficient access (for most types, it falls back to std::set, but for
921 pointers it uses something far better, <a
922 href="#dss_smallptrset">SmallPtrSet</a>).</p>
924 <p>The magic of this class is that it handles small sets extremely efficiently,
925 but gracefully handles extremely large sets without loss of efficiency. The
926 drawback is that the interface is quite small: it supports insertion, queries
927 and erasing, but does not support iteration.</p>
931 <!-- _______________________________________________________________________ -->
932 <div class="doc_subsubsection">
933 <a name="dss_smallptrset">"llvm/ADT/SmallPtrSet.h"</a>
936 <div class="doc_text">
938 <p>SmallPtrSet has all the advantages of SmallSet (and a SmallSet of pointers is
939 transparently implemented with a SmallPtrSet), but also supports iterators. If
940 more than 'N' insertions are performed, a single quadratically
941 probed hash table is allocated and grows as needed, providing extremely
942 efficient access (constant time insertion/deleting/queries with low constant
943 factors) and is very stingy with malloc traffic.</p>
945 <p>Note that, unlike std::set, the iterators of SmallPtrSet are invalidated
946 whenever an insertion occurs. Also, the values visited by the iterators are not
947 visited in sorted order.</p>
951 <!-- _______________________________________________________________________ -->
952 <div class="doc_subsubsection">
953 <a name="dss_FoldingSet">"llvm/ADT/FoldingSet.h"</a>
956 <div class="doc_text">
959 FoldingSet is an aggregate class that is really good at uniquing
960 expensive-to-create or polymorphic objects. It is a combination of a chained
961 hash table with intrusive links (uniqued objects are required to inherit from
962 FoldingSetNode) that uses <a href="#dss_smallvector">SmallVector</a> as part of
965 <p>Consider a case where you want to implement a "getOrCreateFoo" method for
966 a complex object (for example, a node in the code generator). The client has a
967 description of *what* it wants to generate (it knows the opcode and all the
968 operands), but we don't want to 'new' a node, then try inserting it into a set
969 only to find out it already exists, at which point we would have to delete it
970 and return the node that already exists.
973 <p>To support this style of client, FoldingSet perform a query with a
974 FoldingSetNodeID (which wraps SmallVector) that can be used to describe the
975 element that we want to query for. The query either returns the element
976 matching the ID or it returns an opaque ID that indicates where insertion should
977 take place. Construction of the ID usually does not require heap traffic.</p>
979 <p>Because FoldingSet uses intrusive links, it can support polymorphic objects
980 in the set (for example, you can have SDNode instances mixed with LoadSDNodes).
981 Because the elements are individually allocated, pointers to the elements are
982 stable: inserting or removing elements does not invalidate any pointers to other
988 <!-- _______________________________________________________________________ -->
989 <div class="doc_subsubsection">
990 <a name="dss_set"><set></a>
993 <div class="doc_text">
995 <p><tt>std::set</tt> is a reasonable all-around set class, which is decent at
996 many things but great at nothing. std::set allocates memory for each element
997 inserted (thus it is very malloc intensive) and typically stores three pointers
998 per element in the set (thus adding a large amount of per-element space
999 overhead). It offers guaranteed log(n) performance, which is not particularly
1000 fast from a complexity standpoint (particularly if the elements of the set are
1001 expensive to compare, like strings), and has extremely high constant factors for
1002 lookup, insertion and removal.</p>
1004 <p>The advantages of std::set are that its iterators are stable (deleting or
1005 inserting an element from the set does not affect iterators or pointers to other
1006 elements) and that iteration over the set is guaranteed to be in sorted order.
1007 If the elements in the set are large, then the relative overhead of the pointers
1008 and malloc traffic is not a big deal, but if the elements of the set are small,
1009 std::set is almost never a good choice.</p>
1013 <!-- _______________________________________________________________________ -->
1014 <div class="doc_subsubsection">
1015 <a name="dss_setvector">"llvm/ADT/SetVector.h"</a>
1018 <div class="doc_text">
1019 <p>LLVM's SetVector<Type> is actually a combination of a set along with
1020 a <a href="#ds_sequential">Sequential Container</a>. The important property
1021 that this provides is efficient insertion with uniquing (duplicate elements are
1022 ignored) with iteration support. It implements this by inserting elements into
1023 both a set-like container and the sequential container, using the set-like
1024 container for uniquing and the sequential container for iteration.
1027 <p>The difference between SetVector and other sets is that the order of
1028 iteration is guaranteed to match the order of insertion into the SetVector.
1029 This property is really important for things like sets of pointers. Because
1030 pointer values are non-deterministic (e.g. vary across runs of the program on
1031 different machines), iterating over the pointers in a std::set or other set will
1032 not be in a well-defined order.</p>
1035 The drawback of SetVector is that it requires twice as much space as a normal
1036 set and has the sum of constant factors from the set-like container and the
1037 sequential container that it uses. Use it *only* if you need to iterate over
1038 the elements in a deterministic order. SetVector is also expensive to delete
1039 elements out of (linear time).
1044 <!-- _______________________________________________________________________ -->
1045 <div class="doc_subsubsection">
1046 <a name="dss_uniquevector">"llvm/ADT/UniqueVector.h"</a>
1049 <div class="doc_text">
1052 UniqueVector is similar to <a href="#dss_setvector">SetVector</a>, but it
1053 retains a unique ID for each element inserted into the set. It internally
1054 contains a map and a vector, and it assigns a unique ID for each value inserted
1057 <p>UniqueVector is very expensive: its cost is the sum of the cost of
1058 maintaining both the map and vector, it has high complexity, high constant
1059 factors, and produces a lot of malloc traffic. It should be avoided.</p>
1064 <!-- _______________________________________________________________________ -->
1065 <div class="doc_subsubsection">
1066 <a name="dss_otherset">Other Set-Like Container Options</a>
1069 <div class="doc_text">
1072 The STL provides several other options, such as std::multiset and the various
1073 "hash_set" like containers (whether from C++ TR1 or from the SGI library).</p>
1075 <p>std::multiset is useful if you're not interested in elimination of
1076 duplicates, but has all the drawbacks of std::set. A sorted vector (where you
1077 don't delete duplicate entries) or some other approach is almost always
1080 <p>The various hash_set implementations (exposed portably by
1081 "llvm/ADT/hash_set") is a simple chained hashtable. This algorithm is as malloc
1082 intensive as std::set (performing an allocation for each element inserted,
1083 thus having really high constant factors) but (usually) provides O(1)
1084 insertion/deletion of elements. This can be useful if your elements are large
1085 (thus making the constant-factor cost relatively low) or if comparisons are
1086 expensive. Element iteration does not visit elements in a useful order.</p>
1090 <!-- ======================================================================= -->
1091 <div class="doc_subsection">
1092 <a name="ds_map">Map-Like Containers (std::map, DenseMap, etc)</a>
1095 <div class="doc_text">
1096 Map-like containers are useful when you want to associate data to a key. As
1097 usual, there are a lot of different ways to do this. :)
1100 <!-- _______________________________________________________________________ -->
1101 <div class="doc_subsubsection">
1102 <a name="dss_sortedvectormap">A sorted 'vector'</a>
1105 <div class="doc_text">
1108 If your usage pattern follows a strict insert-then-query approach, you can
1109 trivially use the same approach as <a href="#dss_sortedvectorset">sorted vectors
1110 for set-like containers</a>. The only difference is that your query function
1111 (which uses std::lower_bound to get efficient log(n) lookup) should only compare
1112 the key, not both the key and value. This yields the same advantages as sorted
1117 <!-- _______________________________________________________________________ -->
1118 <div class="doc_subsubsection">
1119 <a name="dss_cstringmap">"llvm/ADT/CStringMap.h"</a>
1122 <div class="doc_text">
1125 Strings are commonly used as keys in maps, and they are difficult to support
1126 efficiently: they are variable length, inefficient to hash and compare when
1127 long, expensive to copy, etc. CStringMap is a specialized container designed to
1128 cope with these issues. It supports mapping an arbitrary range of bytes that
1129 does not have an embedded nul character in it ("C strings") to an arbitrary
1132 <p>The CStringMap implementation uses a quadratically-probed hash table, where
1133 the buckets store a pointer to the heap allocated entries (and some other
1134 stuff). The entries in the map must be heap allocated because the strings are
1135 variable length. The string data (key) and the element object (value) are
1136 stored in the same allocation with the string data immediately after the element
1137 object. This container guarantees the "<tt>(char*)(&Value+1)</tt>" points
1138 to the key string for a value.</p>
1140 <p>The CStringMap is very fast for several reasons: quadratic probing is very
1141 cache efficient for lookups, the hash value of strings in buckets is not
1142 recomputed when lookup up an element, CStringMap rarely has to touch the
1143 memory for unrelated objects when looking up a value (even when hash collisions
1144 happen), hash table growth does not recompute the hash values for strings
1145 already in the table, and each pair in the map is store in a single allocation
1146 (the string data is stored in the same allocation as the Value of a pair).</p>
1148 <p>CStringMap also provides query methods that take byte ranges, so it only ever
1149 copies a string if a value is inserted into the table.</p>
1152 <!-- _______________________________________________________________________ -->
1153 <div class="doc_subsubsection">
1154 <a name="dss_indexedmap">"llvm/ADT/IndexedMap.h"</a>
1157 <div class="doc_text">
1159 IndexedMap is a specialized container for mapping small dense integers (or
1160 values that can be mapped to small dense integers) to some other type. It is
1161 internally implemented as a vector with a mapping function that maps the keys to
1162 the dense integer range.
1166 This is useful for cases like virtual registers in the LLVM code generator: they
1167 have a dense mapping that is offset by a compile-time constant (the first
1168 virtual register ID).</p>
1172 <!-- _______________________________________________________________________ -->
1173 <div class="doc_subsubsection">
1174 <a name="dss_densemap">"llvm/ADT/DenseMap.h"</a>
1177 <div class="doc_text">
1180 DenseMap is a simple quadratically probed hash table. It excels at supporting
1181 small keys and values: it uses a single allocation to hold all of the pairs that
1182 are currently inserted in the map. DenseMap is a great way to map pointers to
1183 pointers, or map other small types to each other.
1187 There are several aspects of DenseMap that you should be aware of, however. The
1188 iterators in a densemap are invalidated whenever an insertion occurs, unlike
1189 map. Also, because DenseMap allocates space for a large number of key/value
1190 pairs (it starts with 64 by default), it will waste a lot of space if your keys
1191 or values are large. Finally, you must implement a partial specialization of
1192 DenseMapKeyInfo for the key that you want, if it isn't already supported. This
1193 is required to tell DenseMap about two special marker values (which can never be
1194 inserted into the map) that it needs internally.</p>
1198 <!-- _______________________________________________________________________ -->
1199 <div class="doc_subsubsection">
1200 <a name="dss_map"><map></a>
1203 <div class="doc_text">
1206 std::map has similar characteristics to <a href="#dss_set">std::set</a>: it uses
1207 a single allocation per pair inserted into the map, it offers log(n) lookup with
1208 an extremely large constant factor, imposes a space penalty of 3 pointers per
1209 pair in the map, etc.</p>
1211 <p>std::map is most useful when your keys or values are very large, if you need
1212 to iterate over the collection in sorted order, or if you need stable iterators
1213 into the map (i.e. they don't get invalidated if an insertion or deletion of
1214 another element takes place).</p>
1218 <!-- _______________________________________________________________________ -->
1219 <div class="doc_subsubsection">
1220 <a name="dss_othermap">Other Map-Like Container Options</a>
1223 <div class="doc_text">
1226 The STL provides several other options, such as std::multimap and the various
1227 "hash_map" like containers (whether from C++ TR1 or from the SGI library).</p>
1229 <p>std::multimap is useful if you want to map a key to multiple values, but has
1230 all the drawbacks of std::map. A sorted vector or some other approach is almost
1233 <p>The various hash_map implementations (exposed portably by
1234 "llvm/ADT/hash_map") are simple chained hash tables. This algorithm is as
1235 malloc intensive as std::map (performing an allocation for each element
1236 inserted, thus having really high constant factors) but (usually) provides O(1)
1237 insertion/deletion of elements. This can be useful if your elements are large
1238 (thus making the constant-factor cost relatively low) or if comparisons are
1239 expensive. Element iteration does not visit elements in a useful order.</p>
1244 <!-- *********************************************************************** -->
1245 <div class="doc_section">
1246 <a name="common">Helpful Hints for Common Operations</a>
1248 <!-- *********************************************************************** -->
1250 <div class="doc_text">
1252 <p>This section describes how to perform some very simple transformations of
1253 LLVM code. This is meant to give examples of common idioms used, showing the
1254 practical side of LLVM transformations. <p> Because this is a "how-to" section,
1255 you should also read about the main classes that you will be working with. The
1256 <a href="#coreclasses">Core LLVM Class Hierarchy Reference</a> contains details
1257 and descriptions of the main classes that you should know about.</p>
1261 <!-- NOTE: this section should be heavy on example code -->
1262 <!-- ======================================================================= -->
1263 <div class="doc_subsection">
1264 <a name="inspection">Basic Inspection and Traversal Routines</a>
1267 <div class="doc_text">
1269 <p>The LLVM compiler infrastructure have many different data structures that may
1270 be traversed. Following the example of the C++ standard template library, the
1271 techniques used to traverse these various data structures are all basically the
1272 same. For a enumerable sequence of values, the <tt>XXXbegin()</tt> function (or
1273 method) returns an iterator to the start of the sequence, the <tt>XXXend()</tt>
1274 function returns an iterator pointing to one past the last valid element of the
1275 sequence, and there is some <tt>XXXiterator</tt> data type that is common
1276 between the two operations.</p>
1278 <p>Because the pattern for iteration is common across many different aspects of
1279 the program representation, the standard template library algorithms may be used
1280 on them, and it is easier to remember how to iterate. First we show a few common
1281 examples of the data structures that need to be traversed. Other data
1282 structures are traversed in very similar ways.</p>
1286 <!-- _______________________________________________________________________ -->
1287 <div class="doc_subsubsection">
1288 <a name="iterate_function">Iterating over the </a><a
1289 href="#BasicBlock"><tt>BasicBlock</tt></a>s in a <a
1290 href="#Function"><tt>Function</tt></a>
1293 <div class="doc_text">
1295 <p>It's quite common to have a <tt>Function</tt> instance that you'd like to
1296 transform in some way; in particular, you'd like to manipulate its
1297 <tt>BasicBlock</tt>s. To facilitate this, you'll need to iterate over all of
1298 the <tt>BasicBlock</tt>s that constitute the <tt>Function</tt>. The following is
1299 an example that prints the name of a <tt>BasicBlock</tt> and the number of
1300 <tt>Instruction</tt>s it contains:</p>
1302 <div class="doc_code">
1304 // <i>func is a pointer to a Function instance</i>
1305 for (Function::iterator i = func->begin(), e = func->end(); i != e; ++i)
1306 // <i>Print out the name of the basic block if it has one, and then the</i>
1307 // <i>number of instructions that it contains</i>
1308 llvm::cerr << "Basic block (name=" << i->getName() << ") has "
1309 << i->size() << " instructions.\n";
1313 <p>Note that i can be used as if it were a pointer for the purposes of
1314 invoking member functions of the <tt>Instruction</tt> class. This is
1315 because the indirection operator is overloaded for the iterator
1316 classes. In the above code, the expression <tt>i->size()</tt> is
1317 exactly equivalent to <tt>(*i).size()</tt> just like you'd expect.</p>
1321 <!-- _______________________________________________________________________ -->
1322 <div class="doc_subsubsection">
1323 <a name="iterate_basicblock">Iterating over the </a><a
1324 href="#Instruction"><tt>Instruction</tt></a>s in a <a
1325 href="#BasicBlock"><tt>BasicBlock</tt></a>
1328 <div class="doc_text">
1330 <p>Just like when dealing with <tt>BasicBlock</tt>s in <tt>Function</tt>s, it's
1331 easy to iterate over the individual instructions that make up
1332 <tt>BasicBlock</tt>s. Here's a code snippet that prints out each instruction in
1333 a <tt>BasicBlock</tt>:</p>
1335 <div class="doc_code">
1337 // <i>blk is a pointer to a BasicBlock instance</i>
1338 for (BasicBlock::iterator i = blk->begin(), e = blk->end(); i != e; ++i)
1339 // <i>The next statement works since operator<<(ostream&,...)</i>
1340 // <i>is overloaded for Instruction&</i>
1341 llvm::cerr << *i << "\n";
1345 <p>However, this isn't really the best way to print out the contents of a
1346 <tt>BasicBlock</tt>! Since the ostream operators are overloaded for virtually
1347 anything you'll care about, you could have just invoked the print routine on the
1348 basic block itself: <tt>llvm::cerr << *blk << "\n";</tt>.</p>
1352 <!-- _______________________________________________________________________ -->
1353 <div class="doc_subsubsection">
1354 <a name="iterate_institer">Iterating over the </a><a
1355 href="#Instruction"><tt>Instruction</tt></a>s in a <a
1356 href="#Function"><tt>Function</tt></a>
1359 <div class="doc_text">
1361 <p>If you're finding that you commonly iterate over a <tt>Function</tt>'s
1362 <tt>BasicBlock</tt>s and then that <tt>BasicBlock</tt>'s <tt>Instruction</tt>s,
1363 <tt>InstIterator</tt> should be used instead. You'll need to include <a
1364 href="/doxygen/InstIterator_8h-source.html"><tt>llvm/Support/InstIterator.h</tt></a>,
1365 and then instantiate <tt>InstIterator</tt>s explicitly in your code. Here's a
1366 small example that shows how to dump all instructions in a function to the standard error stream:<p>
1368 <div class="doc_code">
1370 #include "<a href="/doxygen/InstIterator_8h-source.html">llvm/Support/InstIterator.h</a>"
1372 // <i>F is a pointer to a Function instance</i>
1373 for (inst_iterator i = inst_begin(F), e = inst_end(F); i != e; ++i)
1374 llvm::cerr << *i << "\n";
1378 <p>Easy, isn't it? You can also use <tt>InstIterator</tt>s to fill a
1379 work list with its initial contents. For example, if you wanted to
1380 initialize a work list to contain all instructions in a <tt>Function</tt>
1381 F, all you would need to do is something like:</p>
1383 <div class="doc_code">
1385 std::set<Instruction*> worklist;
1386 worklist.insert(inst_begin(F), inst_end(F));
1390 <p>The STL set <tt>worklist</tt> would now contain all instructions in the
1391 <tt>Function</tt> pointed to by F.</p>
1395 <!-- _______________________________________________________________________ -->
1396 <div class="doc_subsubsection">
1397 <a name="iterate_convert">Turning an iterator into a class pointer (and
1401 <div class="doc_text">
1403 <p>Sometimes, it'll be useful to grab a reference (or pointer) to a class
1404 instance when all you've got at hand is an iterator. Well, extracting
1405 a reference or a pointer from an iterator is very straight-forward.
1406 Assuming that <tt>i</tt> is a <tt>BasicBlock::iterator</tt> and <tt>j</tt>
1407 is a <tt>BasicBlock::const_iterator</tt>:</p>
1409 <div class="doc_code">
1411 Instruction& inst = *i; // <i>Grab reference to instruction reference</i>
1412 Instruction* pinst = &*i; // <i>Grab pointer to instruction reference</i>
1413 const Instruction& inst = *j;
1417 <p>However, the iterators you'll be working with in the LLVM framework are
1418 special: they will automatically convert to a ptr-to-instance type whenever they
1419 need to. Instead of dereferencing the iterator and then taking the address of
1420 the result, you can simply assign the iterator to the proper pointer type and
1421 you get the dereference and address-of operation as a result of the assignment
1422 (behind the scenes, this is a result of overloading casting mechanisms). Thus
1423 the last line of the last example,</p>
1425 <div class="doc_code">
1427 Instruction* pinst = &*i;
1431 <p>is semantically equivalent to</p>
1433 <div class="doc_code">
1435 Instruction* pinst = i;
1439 <p>It's also possible to turn a class pointer into the corresponding iterator,
1440 and this is a constant time operation (very efficient). The following code
1441 snippet illustrates use of the conversion constructors provided by LLVM
1442 iterators. By using these, you can explicitly grab the iterator of something
1443 without actually obtaining it via iteration over some structure:</p>
1445 <div class="doc_code">
1447 void printNextInstruction(Instruction* inst) {
1448 BasicBlock::iterator it(inst);
1449 ++it; // <i>After this line, it refers to the instruction after *inst</i>
1450 if (it != inst->getParent()->end()) llvm::cerr << *it << "\n";
1457 <!--_______________________________________________________________________-->
1458 <div class="doc_subsubsection">
1459 <a name="iterate_complex">Finding call sites: a slightly more complex
1463 <div class="doc_text">
1465 <p>Say that you're writing a FunctionPass and would like to count all the
1466 locations in the entire module (that is, across every <tt>Function</tt>) where a
1467 certain function (i.e., some <tt>Function</tt>*) is already in scope. As you'll
1468 learn later, you may want to use an <tt>InstVisitor</tt> to accomplish this in a
1469 much more straight-forward manner, but this example will allow us to explore how
1470 you'd do it if you didn't have <tt>InstVisitor</tt> around. In pseudo-code, this
1471 is what we want to do:</p>
1473 <div class="doc_code">
1475 initialize callCounter to zero
1476 for each Function f in the Module
1477 for each BasicBlock b in f
1478 for each Instruction i in b
1479 if (i is a CallInst and calls the given function)
1480 increment callCounter
1484 <p>And the actual code is (remember, because we're writing a
1485 <tt>FunctionPass</tt>, our <tt>FunctionPass</tt>-derived class simply has to
1486 override the <tt>runOnFunction</tt> method):</p>
1488 <div class="doc_code">
1490 Function* targetFunc = ...;
1492 class OurFunctionPass : public FunctionPass {
1494 OurFunctionPass(): callCounter(0) { }
1496 virtual runOnFunction(Function& F) {
1497 for (Function::iterator b = F.begin(), be = F.end(); b != be; ++b) {
1498 for (BasicBlock::iterator i = b->begin(); ie = b->end(); i != ie; ++i) {
1499 if (<a href="#CallInst">CallInst</a>* callInst = <a href="#isa">dyn_cast</a><<a
1500 href="#CallInst">CallInst</a>>(&*i)) {
1501 // <i>We know we've encountered a call instruction, so we</i>
1502 // <i>need to determine if it's a call to the</i>
1503 // <i>function pointed to by m_func or not</i>
1505 if (callInst->getCalledFunction() == targetFunc)
1513 unsigned callCounter;
1520 <!--_______________________________________________________________________-->
1521 <div class="doc_subsubsection">
1522 <a name="calls_and_invokes">Treating calls and invokes the same way</a>
1525 <div class="doc_text">
1527 <p>You may have noticed that the previous example was a bit oversimplified in
1528 that it did not deal with call sites generated by 'invoke' instructions. In
1529 this, and in other situations, you may find that you want to treat
1530 <tt>CallInst</tt>s and <tt>InvokeInst</tt>s the same way, even though their
1531 most-specific common base class is <tt>Instruction</tt>, which includes lots of
1532 less closely-related things. For these cases, LLVM provides a handy wrapper
1534 href="http://llvm.org/doxygen/classllvm_1_1CallSite.html"><tt>CallSite</tt></a>.
1535 It is essentially a wrapper around an <tt>Instruction</tt> pointer, with some
1536 methods that provide functionality common to <tt>CallInst</tt>s and
1537 <tt>InvokeInst</tt>s.</p>
1539 <p>This class has "value semantics": it should be passed by value, not by
1540 reference and it should not be dynamically allocated or deallocated using
1541 <tt>operator new</tt> or <tt>operator delete</tt>. It is efficiently copyable,
1542 assignable and constructable, with costs equivalents to that of a bare pointer.
1543 If you look at its definition, it has only a single pointer member.</p>
1547 <!--_______________________________________________________________________-->
1548 <div class="doc_subsubsection">
1549 <a name="iterate_chains">Iterating over def-use & use-def chains</a>
1552 <div class="doc_text">
1554 <p>Frequently, we might have an instance of the <a
1555 href="/doxygen/classllvm_1_1Value.html">Value Class</a> and we want to
1556 determine which <tt>User</tt>s use the <tt>Value</tt>. The list of all
1557 <tt>User</tt>s of a particular <tt>Value</tt> is called a <i>def-use</i> chain.
1558 For example, let's say we have a <tt>Function*</tt> named <tt>F</tt> to a
1559 particular function <tt>foo</tt>. Finding all of the instructions that
1560 <i>use</i> <tt>foo</tt> is as simple as iterating over the <i>def-use</i> chain
1563 <div class="doc_code">
1567 for (Value::use_iterator i = F->use_begin(), e = F->use_end(); i != e; ++i)
1568 if (Instruction *Inst = dyn_cast<Instruction>(*i)) {
1569 llvm::cerr << "F is used in instruction:\n";
1570 llvm::cerr << *Inst << "\n";
1575 <p>Alternately, it's common to have an instance of the <a
1576 href="/doxygen/classllvm_1_1User.html">User Class</a> and need to know what
1577 <tt>Value</tt>s are used by it. The list of all <tt>Value</tt>s used by a
1578 <tt>User</tt> is known as a <i>use-def</i> chain. Instances of class
1579 <tt>Instruction</tt> are common <tt>User</tt>s, so we might want to iterate over
1580 all of the values that a particular instruction uses (that is, the operands of
1581 the particular <tt>Instruction</tt>):</p>
1583 <div class="doc_code">
1585 Instruction* pi = ...;
1587 for (User::op_iterator i = pi->op_begin(), e = pi->op_end(); i != e; ++i) {
1595 def-use chains ("finding all users of"): Value::use_begin/use_end
1596 use-def chains ("finding all values used"): User::op_begin/op_end [op=operand]
1601 <!-- ======================================================================= -->
1602 <div class="doc_subsection">
1603 <a name="simplechanges">Making simple changes</a>
1606 <div class="doc_text">
1608 <p>There are some primitive transformation operations present in the LLVM
1609 infrastructure that are worth knowing about. When performing
1610 transformations, it's fairly common to manipulate the contents of basic
1611 blocks. This section describes some of the common methods for doing so
1612 and gives example code.</p>
1616 <!--_______________________________________________________________________-->
1617 <div class="doc_subsubsection">
1618 <a name="schanges_creating">Creating and inserting new
1619 <tt>Instruction</tt>s</a>
1622 <div class="doc_text">
1624 <p><i>Instantiating Instructions</i></p>
1626 <p>Creation of <tt>Instruction</tt>s is straight-forward: simply call the
1627 constructor for the kind of instruction to instantiate and provide the necessary
1628 parameters. For example, an <tt>AllocaInst</tt> only <i>requires</i> a
1629 (const-ptr-to) <tt>Type</tt>. Thus:</p>
1631 <div class="doc_code">
1633 AllocaInst* ai = new AllocaInst(Type::IntTy);
1637 <p>will create an <tt>AllocaInst</tt> instance that represents the allocation of
1638 one integer in the current stack frame, at run time. Each <tt>Instruction</tt>
1639 subclass is likely to have varying default parameters which change the semantics
1640 of the instruction, so refer to the <a
1641 href="/doxygen/classllvm_1_1Instruction.html">doxygen documentation for the subclass of
1642 Instruction</a> that you're interested in instantiating.</p>
1644 <p><i>Naming values</i></p>
1646 <p>It is very useful to name the values of instructions when you're able to, as
1647 this facilitates the debugging of your transformations. If you end up looking
1648 at generated LLVM machine code, you definitely want to have logical names
1649 associated with the results of instructions! By supplying a value for the
1650 <tt>Name</tt> (default) parameter of the <tt>Instruction</tt> constructor, you
1651 associate a logical name with the result of the instruction's execution at
1652 run time. For example, say that I'm writing a transformation that dynamically
1653 allocates space for an integer on the stack, and that integer is going to be
1654 used as some kind of index by some other code. To accomplish this, I place an
1655 <tt>AllocaInst</tt> at the first point in the first <tt>BasicBlock</tt> of some
1656 <tt>Function</tt>, and I'm intending to use it within the same
1657 <tt>Function</tt>. I might do:</p>
1659 <div class="doc_code">
1661 AllocaInst* pa = new AllocaInst(Type::IntTy, 0, "indexLoc");
1665 <p>where <tt>indexLoc</tt> is now the logical name of the instruction's
1666 execution value, which is a pointer to an integer on the run time stack.</p>
1668 <p><i>Inserting instructions</i></p>
1670 <p>There are essentially two ways to insert an <tt>Instruction</tt>
1671 into an existing sequence of instructions that form a <tt>BasicBlock</tt>:</p>
1674 <li>Insertion into an explicit instruction list
1676 <p>Given a <tt>BasicBlock* pb</tt>, an <tt>Instruction* pi</tt> within that
1677 <tt>BasicBlock</tt>, and a newly-created instruction we wish to insert
1678 before <tt>*pi</tt>, we do the following: </p>
1680 <div class="doc_code">
1682 BasicBlock *pb = ...;
1683 Instruction *pi = ...;
1684 Instruction *newInst = new Instruction(...);
1686 pb->getInstList().insert(pi, newInst); // <i>Inserts newInst before pi in pb</i>
1690 <p>Appending to the end of a <tt>BasicBlock</tt> is so common that
1691 the <tt>Instruction</tt> class and <tt>Instruction</tt>-derived
1692 classes provide constructors which take a pointer to a
1693 <tt>BasicBlock</tt> to be appended to. For example code that
1696 <div class="doc_code">
1698 BasicBlock *pb = ...;
1699 Instruction *newInst = new Instruction(...);
1701 pb->getInstList().push_back(newInst); // <i>Appends newInst to pb</i>
1707 <div class="doc_code">
1709 BasicBlock *pb = ...;
1710 Instruction *newInst = new Instruction(..., pb);
1714 <p>which is much cleaner, especially if you are creating
1715 long instruction streams.</p></li>
1717 <li>Insertion into an implicit instruction list
1719 <p><tt>Instruction</tt> instances that are already in <tt>BasicBlock</tt>s
1720 are implicitly associated with an existing instruction list: the instruction
1721 list of the enclosing basic block. Thus, we could have accomplished the same
1722 thing as the above code without being given a <tt>BasicBlock</tt> by doing:
1725 <div class="doc_code">
1727 Instruction *pi = ...;
1728 Instruction *newInst = new Instruction(...);
1730 pi->getParent()->getInstList().insert(pi, newInst);
1734 <p>In fact, this sequence of steps occurs so frequently that the
1735 <tt>Instruction</tt> class and <tt>Instruction</tt>-derived classes provide
1736 constructors which take (as a default parameter) a pointer to an
1737 <tt>Instruction</tt> which the newly-created <tt>Instruction</tt> should
1738 precede. That is, <tt>Instruction</tt> constructors are capable of
1739 inserting the newly-created instance into the <tt>BasicBlock</tt> of a
1740 provided instruction, immediately before that instruction. Using an
1741 <tt>Instruction</tt> constructor with a <tt>insertBefore</tt> (default)
1742 parameter, the above code becomes:</p>
1744 <div class="doc_code">
1746 Instruction* pi = ...;
1747 Instruction* newInst = new Instruction(..., pi);
1751 <p>which is much cleaner, especially if you're creating a lot of
1752 instructions and adding them to <tt>BasicBlock</tt>s.</p></li>
1757 <!--_______________________________________________________________________-->
1758 <div class="doc_subsubsection">
1759 <a name="schanges_deleting">Deleting <tt>Instruction</tt>s</a>
1762 <div class="doc_text">
1764 <p>Deleting an instruction from an existing sequence of instructions that form a
1765 <a href="#BasicBlock"><tt>BasicBlock</tt></a> is very straight-forward. First,
1766 you must have a pointer to the instruction that you wish to delete. Second, you
1767 need to obtain the pointer to that instruction's basic block. You use the
1768 pointer to the basic block to get its list of instructions and then use the
1769 erase function to remove your instruction. For example:</p>
1771 <div class="doc_code">
1773 <a href="#Instruction">Instruction</a> *I = .. ;
1774 <a href="#BasicBlock">BasicBlock</a> *BB = I->getParent();
1776 BB->getInstList().erase(I);
1782 <!--_______________________________________________________________________-->
1783 <div class="doc_subsubsection">
1784 <a name="schanges_replacing">Replacing an <tt>Instruction</tt> with another
1788 <div class="doc_text">
1790 <p><i>Replacing individual instructions</i></p>
1792 <p>Including "<a href="/doxygen/BasicBlockUtils_8h-source.html">llvm/Transforms/Utils/BasicBlockUtils.h</a>"
1793 permits use of two very useful replace functions: <tt>ReplaceInstWithValue</tt>
1794 and <tt>ReplaceInstWithInst</tt>.</p>
1796 <h4><a name="schanges_deleting">Deleting <tt>Instruction</tt>s</a></h4>
1799 <li><tt>ReplaceInstWithValue</tt>
1801 <p>This function replaces all uses (within a basic block) of a given
1802 instruction with a value, and then removes the original instruction. The
1803 following example illustrates the replacement of the result of a particular
1804 <tt>AllocaInst</tt> that allocates memory for a single integer with a null
1805 pointer to an integer.</p>
1807 <div class="doc_code">
1809 AllocaInst* instToReplace = ...;
1810 BasicBlock::iterator ii(instToReplace);
1812 ReplaceInstWithValue(instToReplace->getParent()->getInstList(), ii,
1813 Constant::getNullValue(PointerType::get(Type::IntTy)));
1816 <li><tt>ReplaceInstWithInst</tt>
1818 <p>This function replaces a particular instruction with another
1819 instruction. The following example illustrates the replacement of one
1820 <tt>AllocaInst</tt> with another.</p>
1822 <div class="doc_code">
1824 AllocaInst* instToReplace = ...;
1825 BasicBlock::iterator ii(instToReplace);
1827 ReplaceInstWithInst(instToReplace->getParent()->getInstList(), ii,
1828 new AllocaInst(Type::IntTy, 0, "ptrToReplacedInt"));
1832 <p><i>Replacing multiple uses of <tt>User</tt>s and <tt>Value</tt>s</i></p>
1834 <p>You can use <tt>Value::replaceAllUsesWith</tt> and
1835 <tt>User::replaceUsesOfWith</tt> to change more than one use at a time. See the
1836 doxygen documentation for the <a href="/doxygen/classllvm_1_1Value.html">Value Class</a>
1837 and <a href="/doxygen/classllvm_1_1User.html">User Class</a>, respectively, for more
1840 <!-- Value::replaceAllUsesWith User::replaceUsesOfWith Point out:
1841 include/llvm/Transforms/Utils/ especially BasicBlockUtils.h with:
1842 ReplaceInstWithValue, ReplaceInstWithInst -->
1846 <!-- *********************************************************************** -->
1847 <div class="doc_section">
1848 <a name="advanced">Advanced Topics</a>
1850 <!-- *********************************************************************** -->
1852 <div class="doc_text">
1854 This section describes some of the advanced or obscure API's that most clients
1855 do not need to be aware of. These API's tend manage the inner workings of the
1856 LLVM system, and only need to be accessed in unusual circumstances.
1860 <!-- ======================================================================= -->
1861 <div class="doc_subsection">
1862 <a name="TypeResolve">LLVM Type Resolution</a>
1865 <div class="doc_text">
1868 The LLVM type system has a very simple goal: allow clients to compare types for
1869 structural equality with a simple pointer comparison (aka a shallow compare).
1870 This goal makes clients much simpler and faster, and is used throughout the LLVM
1875 Unfortunately achieving this goal is not a simple matter. In particular,
1876 recursive types and late resolution of opaque types makes the situation very
1877 difficult to handle. Fortunately, for the most part, our implementation makes
1878 most clients able to be completely unaware of the nasty internal details. The
1879 primary case where clients are exposed to the inner workings of it are when
1880 building a recursive type. In addition to this case, the LLVM bytecode reader,
1881 assembly parser, and linker also have to be aware of the inner workings of this
1886 For our purposes below, we need three concepts. First, an "Opaque Type" is
1887 exactly as defined in the <a href="LangRef.html#t_opaque">language
1888 reference</a>. Second an "Abstract Type" is any type which includes an
1889 opaque type as part of its type graph (for example "<tt>{ opaque, i32 }</tt>").
1890 Third, a concrete type is a type that is not an abstract type (e.g. "<tt>{ i32,
1896 <!-- ______________________________________________________________________ -->
1897 <div class="doc_subsubsection">
1898 <a name="BuildRecType">Basic Recursive Type Construction</a>
1901 <div class="doc_text">
1904 Because the most common question is "how do I build a recursive type with LLVM",
1905 we answer it now and explain it as we go. Here we include enough to cause this
1906 to be emitted to an output .ll file:
1909 <div class="doc_code">
1911 %mylist = type { %mylist*, i32 }
1916 To build this, use the following LLVM APIs:
1919 <div class="doc_code">
1921 // <i>Create the initial outer struct</i>
1922 <a href="#PATypeHolder">PATypeHolder</a> StructTy = OpaqueType::get();
1923 std::vector<const Type*> Elts;
1924 Elts.push_back(PointerType::get(StructTy));
1925 Elts.push_back(Type::IntTy);
1926 StructType *NewSTy = StructType::get(Elts);
1928 // <i>At this point, NewSTy = "{ opaque*, i32 }". Tell VMCore that</i>
1929 // <i>the struct and the opaque type are actually the same.</i>
1930 cast<OpaqueType>(StructTy.get())-><a href="#refineAbstractTypeTo">refineAbstractTypeTo</a>(NewSTy);
1932 // <i>NewSTy is potentially invalidated, but StructTy (a <a href="#PATypeHolder">PATypeHolder</a>) is</i>
1933 // <i>kept up-to-date</i>
1934 NewSTy = cast<StructType>(StructTy.get());
1936 // <i>Add a name for the type to the module symbol table (optional)</i>
1937 MyModule->addTypeName("mylist", NewSTy);
1942 This code shows the basic approach used to build recursive types: build a
1943 non-recursive type using 'opaque', then use type unification to close the cycle.
1944 The type unification step is performed by the <tt><a
1945 href="#refineAbstractTypeTo">refineAbstractTypeTo</a></tt> method, which is
1946 described next. After that, we describe the <a
1947 href="#PATypeHolder">PATypeHolder class</a>.
1952 <!-- ______________________________________________________________________ -->
1953 <div class="doc_subsubsection">
1954 <a name="refineAbstractTypeTo">The <tt>refineAbstractTypeTo</tt> method</a>
1957 <div class="doc_text">
1959 The <tt>refineAbstractTypeTo</tt> method starts the type unification process.
1960 While this method is actually a member of the DerivedType class, it is most
1961 often used on OpaqueType instances. Type unification is actually a recursive
1962 process. After unification, types can become structurally isomorphic to
1963 existing types, and all duplicates are deleted (to preserve pointer equality).
1967 In the example above, the OpaqueType object is definitely deleted.
1968 Additionally, if there is an "{ \2*, i32}" type already created in the system,
1969 the pointer and struct type created are <b>also</b> deleted. Obviously whenever
1970 a type is deleted, any "Type*" pointers in the program are invalidated. As
1971 such, it is safest to avoid having <i>any</i> "Type*" pointers to abstract types
1972 live across a call to <tt>refineAbstractTypeTo</tt> (note that non-abstract
1973 types can never move or be deleted). To deal with this, the <a
1974 href="#PATypeHolder">PATypeHolder</a> class is used to maintain a stable
1975 reference to a possibly refined type, and the <a
1976 href="#AbstractTypeUser">AbstractTypeUser</a> class is used to update more
1977 complex datastructures.
1982 <!-- ______________________________________________________________________ -->
1983 <div class="doc_subsubsection">
1984 <a name="PATypeHolder">The PATypeHolder Class</a>
1987 <div class="doc_text">
1989 PATypeHolder is a form of a "smart pointer" for Type objects. When VMCore
1990 happily goes about nuking types that become isomorphic to existing types, it
1991 automatically updates all PATypeHolder objects to point to the new type. In the
1992 example above, this allows the code to maintain a pointer to the resultant
1993 resolved recursive type, even though the Type*'s are potentially invalidated.
1997 PATypeHolder is an extremely light-weight object that uses a lazy union-find
1998 implementation to update pointers. For example the pointer from a Value to its
1999 Type is maintained by PATypeHolder objects.
2004 <!-- ______________________________________________________________________ -->
2005 <div class="doc_subsubsection">
2006 <a name="AbstractTypeUser">The AbstractTypeUser Class</a>
2009 <div class="doc_text">
2012 Some data structures need more to perform more complex updates when types get
2013 resolved. The <a href="#SymbolTable">SymbolTable</a> class, for example, needs
2014 move and potentially merge type planes in its representation when a pointer
2018 To support this, a class can derive from the AbstractTypeUser class. This class
2019 allows it to get callbacks when certain types are resolved. To register to get
2020 callbacks for a particular type, the DerivedType::{add/remove}AbstractTypeUser
2021 methods can be called on a type. Note that these methods only work for <i>
2022 abstract</i> types. Concrete types (those that do not include any opaque
2023 objects) can never be refined.
2028 <!-- ======================================================================= -->
2029 <div class="doc_subsection">
2030 <a name="SymbolTable">The <tt>SymbolTable</tt> class</a>
2033 <div class="doc_text">
2034 <p>This class provides a symbol table that the <a
2035 href="#Function"><tt>Function</tt></a> and <a href="#Module">
2036 <tt>Module</tt></a> classes use for naming definitions. The symbol table can
2037 provide a name for any <a href="#Value"><tt>Value</tt></a>.
2038 <tt>SymbolTable</tt> is an abstract data type. It hides the data it contains
2039 and provides access to it through a controlled interface.</p>
2041 <p>Note that the <tt>SymbolTable</tt> class should not be directly accessed
2042 by most clients. It should only be used when iteration over the symbol table
2043 names themselves are required, which is very special purpose. Note that not
2045 <a href="#Value">Value</a>s have names, and those without names (i.e. they have
2046 an empty name) do not exist in the symbol table.
2049 <p>To use the <tt>SymbolTable</tt> well, you need to understand the
2050 structure of the information it holds. The class contains two
2051 <tt>std::map</tt> objects. The first, <tt>pmap</tt>, is a map of
2052 <tt>Type*</tt> to maps of name (<tt>std::string</tt>) to <tt>Value*</tt>.
2053 Thus, Values are stored in two-dimensions and accessed by <tt>Type</tt> and
2056 <p>The interface of this class provides three basic types of operations:
2058 <li><em>Accessors</em>. Accessors provide read-only access to information
2059 such as finding a value for a name with the
2060 <a href="#SymbolTable_lookup">lookup</a> method.</li>
2061 <li><em>Mutators</em>. Mutators allow the user to add information to the
2062 <tt>SymbolTable</tt> with methods like
2063 <a href="#SymbolTable_insert"><tt>insert</tt></a>.</li>
2064 <li><em>Iterators</em>. Iterators allow the user to traverse the content
2065 of the symbol table in well defined ways, such as the method
2066 <a href="#SymbolTable_plane_begin"><tt>plane_begin</tt></a>.</li>
2071 <dt><tt>Value* lookup(const Type* Ty, const std::string& name) const</tt>:
2073 <dd>The <tt>lookup</tt> method searches the type plane given by the
2074 <tt>Ty</tt> parameter for a <tt>Value</tt> with the provided <tt>name</tt>.
2075 If a suitable <tt>Value</tt> is not found, null is returned.</dd>
2077 <dt><tt>bool isEmpty() const</tt>:</dt>
2078 <dd>This function returns true if both the value and types maps are
2084 <dt><tt>void insert(Value *Val)</tt>:</dt>
2085 <dd>This method adds the provided value to the symbol table. The Value must
2086 have both a name and a type which are extracted and used to place the value
2087 in the correct type plane under the value's name.</dd>
2089 <dt><tt>void insert(const std::string& Name, Value *Val)</tt>:</dt>
2090 <dd> Inserts a constant or type into the symbol table with the specified
2091 name. There can be a many to one mapping between names and constants
2094 <dt><tt>void remove(Value* Val)</tt>:</dt>
2095 <dd> This method removes a named value from the symbol table. The
2096 type and name of the Value are extracted from \p N and used to
2097 lookup the Value in the correct type plane. If the Value is
2098 not in the symbol table, this method silently ignores the
2101 <dt><tt>Value* remove(const std::string& Name, Value *Val)</tt>:</dt>
2102 <dd> Remove a constant or type with the specified name from the
2105 <dt><tt>Value *remove(const value_iterator& It)</tt>:</dt>
2106 <dd> Removes a specific value from the symbol table.
2107 Returns the removed value.</dd>
2109 <dt><tt>bool strip()</tt>:</dt>
2110 <dd> This method will strip the symbol table of its names leaving
2111 the type and values. </dd>
2113 <dt><tt>void clear()</tt>:</dt>
2114 <dd>Empty the symbol table completely.</dd>
2118 <p>The following functions describe three types of iterators you can obtain
2119 the beginning or end of the sequence for both const and non-const. It is
2120 important to keep track of the different kinds of iterators. There are
2121 three idioms worth pointing out:</p>
2124 <tr><th>Units</th><th>Iterator</th><th>Idiom</th></tr>
2126 <td align="left">Planes Of name/Value maps</td><td>PI</td>
2127 <td align="left"><pre><tt>
2128 for (SymbolTable::plane_const_iterator PI = ST.plane_begin(),
2129 PE = ST.plane_end(); PI != PE; ++PI ) {
2130 PI->first // <i>This is the Type* of the plane</i>
2131 PI->second // <i>This is the SymbolTable::ValueMap of name/Value pairs</i>
2136 <td align="left">name/Value pairs in a plane</td><td>VI</td>
2137 <td align="left"><pre><tt>
2138 for (SymbolTable::value_const_iterator VI = ST.value_begin(SomeType),
2139 VE = ST.value_end(SomeType); VI != VE; ++VI ) {
2140 VI->first // <i>This is the name of the Value</i>
2141 VI->second // <i>This is the Value* value associated with the name</i>
2147 <p>Using the recommended iterator names and idioms will help you avoid
2148 making mistakes. Of particular note, make sure that whenever you use
2149 value_begin(SomeType) that you always compare the resulting iterator
2150 with value_end(SomeType) not value_end(SomeOtherType) or else you
2151 will loop infinitely.</p>
2155 <dt><tt>plane_iterator plane_begin()</tt>:</dt>
2156 <dd>Get an iterator that starts at the beginning of the type planes.
2157 The iterator will iterate over the Type/ValueMap pairs in the
2160 <dt><tt>plane_const_iterator plane_begin() const</tt>:</dt>
2161 <dd>Get a const_iterator that starts at the beginning of the type
2162 planes. The iterator will iterate over the Type/ValueMap pairs
2163 in the type planes. </dd>
2165 <dt><tt>plane_iterator plane_end()</tt>:</dt>
2166 <dd>Get an iterator at the end of the type planes. This serves as
2167 the marker for end of iteration over the type planes.</dd>
2169 <dt><tt>plane_const_iterator plane_end() const</tt>:</dt>
2170 <dd>Get a const_iterator at the end of the type planes. This serves as
2171 the marker for end of iteration over the type planes.</dd>
2173 <dt><tt>value_iterator value_begin(const Type *Typ)</tt>:</dt>
2174 <dd>Get an iterator that starts at the beginning of a type plane.
2175 The iterator will iterate over the name/value pairs in the type plane.
2176 Note: The type plane must already exist before using this.</dd>
2178 <dt><tt>value_const_iterator value_begin(const Type *Typ) const</tt>:</dt>
2179 <dd>Get a const_iterator that starts at the beginning of a type plane.
2180 The iterator will iterate over the name/value pairs in the type plane.
2181 Note: The type plane must already exist before using this.</dd>
2183 <dt><tt>value_iterator value_end(const Type *Typ)</tt>:</dt>
2184 <dd>Get an iterator to the end of a type plane. This serves as the marker
2185 for end of iteration of the type plane.
2186 Note: The type plane must already exist before using this.</dd>
2188 <dt><tt>value_const_iterator value_end(const Type *Typ) const</tt>:</dt>
2189 <dd>Get a const_iterator to the end of a type plane. This serves as the
2190 marker for end of iteration of the type plane.
2191 Note: the type plane must already exist before using this.</dd>
2193 <dt><tt>plane_const_iterator find(const Type* Typ ) const</tt>:</dt>
2194 <dd>This method returns a plane_const_iterator for iteration over
2195 the type planes starting at a specific plane, given by \p Ty.</dd>
2197 <dt><tt>plane_iterator find( const Type* Typ </tt>:</dt>
2198 <dd>This method returns a plane_iterator for iteration over the
2199 type planes starting at a specific plane, given by \p Ty.</dd>
2206 <!-- *********************************************************************** -->
2207 <div class="doc_section">
2208 <a name="coreclasses">The Core LLVM Class Hierarchy Reference </a>
2210 <!-- *********************************************************************** -->
2212 <div class="doc_text">
2213 <p><tt>#include "<a href="/doxygen/Type_8h-source.html">llvm/Type.h</a>"</tt>
2214 <br>doxygen info: <a href="/doxygen/classllvm_1_1Type.html">Type Class</a></p>
2216 <p>The Core LLVM classes are the primary means of representing the program
2217 being inspected or transformed. The core LLVM classes are defined in
2218 header files in the <tt>include/llvm/</tt> directory, and implemented in
2219 the <tt>lib/VMCore</tt> directory.</p>
2223 <!-- ======================================================================= -->
2224 <div class="doc_subsection">
2225 <a name="Type">The <tt>Type</tt> class and Derived Types</a>
2228 <div class="doc_text">
2230 <p><tt>Type</tt> is a superclass of all type classes. Every <tt>Value</tt> has
2231 a <tt>Type</tt>. <tt>Type</tt> cannot be instantiated directly but only
2232 through its subclasses. Certain primitive types (<tt>VoidType</tt>,
2233 <tt>LabelType</tt>, <tt>FloatType</tt> and <tt>DoubleType</tt>) have hidden
2234 subclasses. They are hidden because they offer no useful functionality beyond
2235 what the <tt>Type</tt> class offers except to distinguish themselves from
2236 other subclasses of <tt>Type</tt>.</p>
2237 <p>All other types are subclasses of <tt>DerivedType</tt>. Types can be
2238 named, but this is not a requirement. There exists exactly
2239 one instance of a given shape at any one time. This allows type equality to
2240 be performed with address equality of the Type Instance. That is, given two
2241 <tt>Type*</tt> values, the types are identical if the pointers are identical.
2245 <!-- _______________________________________________________________________ -->
2246 <div class="doc_subsubsection">
2247 <a name="m_Value">Important Public Methods</a>
2250 <div class="doc_text">
2253 <li><tt>bool isInteger() const</tt>: Returns true for any integer type.</li>
2255 <li><tt>bool isFloatingPoint()</tt>: Return true if this is one of the two
2256 floating point types.</li>
2258 <li><tt>bool isAbstract()</tt>: Return true if the type is abstract (contains
2259 an OpaqueType anywhere in its definition).</li>
2261 <li><tt>bool isSized()</tt>: Return true if the type has known size. Things
2262 that don't have a size are abstract types, labels and void.</li>
2267 <!-- _______________________________________________________________________ -->
2268 <div class="doc_subsubsection">
2269 <a name="m_Value">Important Derived Types</a>
2271 <div class="doc_text">
2273 <dt><tt>IntegerType</tt></dt>
2274 <dd>Subclass of DerivedType that represents integer types of any bit width.
2275 Any bit width between <tt>IntegerType::MIN_INT_BITS</tt> (1) and
2276 <tt>IntegerType::MAX_INT_BITS</tt> (~8 million) can be represented.
2278 <li><tt>static const IntegerType* get(unsigned NumBits)</tt>: get an integer
2279 type of a specific bit width.</li>
2280 <li><tt>unsigned getBitWidth() const</tt>: Get the bit width of an integer
2284 <dt><tt>SequentialType</tt></dt>
2285 <dd>This is subclassed by ArrayType and PointerType
2287 <li><tt>const Type * getElementType() const</tt>: Returns the type of each
2288 of the elements in the sequential type. </li>
2291 <dt><tt>ArrayType</tt></dt>
2292 <dd>This is a subclass of SequentialType and defines the interface for array
2295 <li><tt>unsigned getNumElements() const</tt>: Returns the number of
2296 elements in the array. </li>
2299 <dt><tt>PointerType</tt></dt>
2300 <dd>Subclass of SequentialType for pointer types.</dd>
2301 <dt><tt>PackedType</tt></dt>
2302 <dd>Subclass of SequentialType for packed (vector) types. A
2303 packed type is similar to an ArrayType but is distinguished because it is
2304 a first class type wherease ArrayType is not. Packed types are used for
2305 vector operations and are usually small vectors of of an integer or floating
2307 <dt><tt>StructType</tt></dt>
2308 <dd>Subclass of DerivedTypes for struct types.</dd>
2309 <dt><tt>FunctionType</tt></dt>
2310 <dd>Subclass of DerivedTypes for function types.
2312 <li><tt>bool isVarArg() const</tt>: Returns true if its a vararg
2314 <li><tt> const Type * getReturnType() const</tt>: Returns the
2315 return type of the function.</li>
2316 <li><tt>const Type * getParamType (unsigned i)</tt>: Returns
2317 the type of the ith parameter.</li>
2318 <li><tt> const unsigned getNumParams() const</tt>: Returns the
2319 number of formal parameters.</li>
2322 <dt><tt>OpaqueType</tt></dt>
2323 <dd>Sublcass of DerivedType for abstract types. This class
2324 defines no content and is used as a placeholder for some other type. Note
2325 that OpaqueType is used (temporarily) during type resolution for forward
2326 references of types. Once the referenced type is resolved, the OpaqueType
2327 is replaced with the actual type. OpaqueType can also be used for data
2328 abstraction. At link time opaque types can be resolved to actual types
2329 of the same name.</dd>
2335 <!-- ======================================================================= -->
2336 <div class="doc_subsection">
2337 <a name="Module">The <tt>Module</tt> class</a>
2340 <div class="doc_text">
2343 href="/doxygen/Module_8h-source.html">llvm/Module.h</a>"</tt><br> doxygen info:
2344 <a href="/doxygen/classllvm_1_1Module.html">Module Class</a></p>
2346 <p>The <tt>Module</tt> class represents the top level structure present in LLVM
2347 programs. An LLVM module is effectively either a translation unit of the
2348 original program or a combination of several translation units merged by the
2349 linker. The <tt>Module</tt> class keeps track of a list of <a
2350 href="#Function"><tt>Function</tt></a>s, a list of <a
2351 href="#GlobalVariable"><tt>GlobalVariable</tt></a>s, and a <a
2352 href="#SymbolTable"><tt>SymbolTable</tt></a>. Additionally, it contains a few
2353 helpful member functions that try to make common operations easy.</p>
2357 <!-- _______________________________________________________________________ -->
2358 <div class="doc_subsubsection">
2359 <a name="m_Module">Important Public Members of the <tt>Module</tt> class</a>
2362 <div class="doc_text">
2365 <li><tt>Module::Module(std::string name = "")</tt></li>
2368 <p>Constructing a <a href="#Module">Module</a> is easy. You can optionally
2369 provide a name for it (probably based on the name of the translation unit).</p>
2372 <li><tt>Module::iterator</tt> - Typedef for function list iterator<br>
2373 <tt>Module::const_iterator</tt> - Typedef for const_iterator.<br>
2375 <tt>begin()</tt>, <tt>end()</tt>
2376 <tt>size()</tt>, <tt>empty()</tt>
2378 <p>These are forwarding methods that make it easy to access the contents of
2379 a <tt>Module</tt> object's <a href="#Function"><tt>Function</tt></a>
2382 <li><tt>Module::FunctionListType &getFunctionList()</tt>
2384 <p> Returns the list of <a href="#Function"><tt>Function</tt></a>s. This is
2385 necessary to use when you need to update the list or perform a complex
2386 action that doesn't have a forwarding method.</p>
2388 <p><!-- Global Variable --></p></li>
2394 <li><tt>Module::global_iterator</tt> - Typedef for global variable list iterator<br>
2396 <tt>Module::const_global_iterator</tt> - Typedef for const_iterator.<br>
2398 <tt>global_begin()</tt>, <tt>global_end()</tt>
2399 <tt>global_size()</tt>, <tt>global_empty()</tt>
2401 <p> These are forwarding methods that make it easy to access the contents of
2402 a <tt>Module</tt> object's <a
2403 href="#GlobalVariable"><tt>GlobalVariable</tt></a> list.</p></li>
2405 <li><tt>Module::GlobalListType &getGlobalList()</tt>
2407 <p>Returns the list of <a
2408 href="#GlobalVariable"><tt>GlobalVariable</tt></a>s. This is necessary to
2409 use when you need to update the list or perform a complex action that
2410 doesn't have a forwarding method.</p>
2412 <p><!-- Symbol table stuff --> </p></li>
2418 <li><tt><a href="#SymbolTable">SymbolTable</a> *getSymbolTable()</tt>
2420 <p>Return a reference to the <a href="#SymbolTable"><tt>SymbolTable</tt></a>
2421 for this <tt>Module</tt>.</p>
2423 <p><!-- Convenience methods --></p></li>
2429 <li><tt><a href="#Function">Function</a> *getFunction(const std::string
2430 &Name, const <a href="#FunctionType">FunctionType</a> *Ty)</tt>
2432 <p>Look up the specified function in the <tt>Module</tt> <a
2433 href="#SymbolTable"><tt>SymbolTable</tt></a>. If it does not exist, return
2434 <tt>null</tt>.</p></li>
2436 <li><tt><a href="#Function">Function</a> *getOrInsertFunction(const
2437 std::string &Name, const <a href="#FunctionType">FunctionType</a> *T)</tt>
2439 <p>Look up the specified function in the <tt>Module</tt> <a
2440 href="#SymbolTable"><tt>SymbolTable</tt></a>. If it does not exist, add an
2441 external declaration for the function and return it.</p></li>
2443 <li><tt>std::string getTypeName(const <a href="#Type">Type</a> *Ty)</tt>
2445 <p>If there is at least one entry in the <a
2446 href="#SymbolTable"><tt>SymbolTable</tt></a> for the specified <a
2447 href="#Type"><tt>Type</tt></a>, return it. Otherwise return the empty
2450 <li><tt>bool addTypeName(const std::string &Name, const <a
2451 href="#Type">Type</a> *Ty)</tt>
2453 <p>Insert an entry in the <a href="#SymbolTable"><tt>SymbolTable</tt></a>
2454 mapping <tt>Name</tt> to <tt>Ty</tt>. If there is already an entry for this
2455 name, true is returned and the <a
2456 href="#SymbolTable"><tt>SymbolTable</tt></a> is not modified.</p></li>
2462 <!-- ======================================================================= -->
2463 <div class="doc_subsection">
2464 <a name="Value">The <tt>Value</tt> class</a>
2467 <div class="doc_text">
2469 <p><tt>#include "<a href="/doxygen/Value_8h-source.html">llvm/Value.h</a>"</tt>
2471 doxygen info: <a href="/doxygen/classllvm_1_1Value.html">Value Class</a></p>
2473 <p>The <tt>Value</tt> class is the most important class in the LLVM Source
2474 base. It represents a typed value that may be used (among other things) as an
2475 operand to an instruction. There are many different types of <tt>Value</tt>s,
2476 such as <a href="#Constant"><tt>Constant</tt></a>s,<a
2477 href="#Argument"><tt>Argument</tt></a>s. Even <a
2478 href="#Instruction"><tt>Instruction</tt></a>s and <a
2479 href="#Function"><tt>Function</tt></a>s are <tt>Value</tt>s.</p>
2481 <p>A particular <tt>Value</tt> may be used many times in the LLVM representation
2482 for a program. For example, an incoming argument to a function (represented
2483 with an instance of the <a href="#Argument">Argument</a> class) is "used" by
2484 every instruction in the function that references the argument. To keep track
2485 of this relationship, the <tt>Value</tt> class keeps a list of all of the <a
2486 href="#User"><tt>User</tt></a>s that is using it (the <a
2487 href="#User"><tt>User</tt></a> class is a base class for all nodes in the LLVM
2488 graph that can refer to <tt>Value</tt>s). This use list is how LLVM represents
2489 def-use information in the program, and is accessible through the <tt>use_</tt>*
2490 methods, shown below.</p>
2492 <p>Because LLVM is a typed representation, every LLVM <tt>Value</tt> is typed,
2493 and this <a href="#Type">Type</a> is available through the <tt>getType()</tt>
2494 method. In addition, all LLVM values can be named. The "name" of the
2495 <tt>Value</tt> is a symbolic string printed in the LLVM code:</p>
2497 <div class="doc_code">
2499 %<b>foo</b> = add i32 1, 2
2503 <p><a name="#nameWarning">The name of this instruction is "foo".</a> <b>NOTE</b>
2504 that the name of any value may be missing (an empty string), so names should
2505 <b>ONLY</b> be used for debugging (making the source code easier to read,
2506 debugging printouts), they should not be used to keep track of values or map
2507 between them. For this purpose, use a <tt>std::map</tt> of pointers to the
2508 <tt>Value</tt> itself instead.</p>
2510 <p>One important aspect of LLVM is that there is no distinction between an SSA
2511 variable and the operation that produces it. Because of this, any reference to
2512 the value produced by an instruction (or the value available as an incoming
2513 argument, for example) is represented as a direct pointer to the instance of
2515 represents this value. Although this may take some getting used to, it
2516 simplifies the representation and makes it easier to manipulate.</p>
2520 <!-- _______________________________________________________________________ -->
2521 <div class="doc_subsubsection">
2522 <a name="m_Value">Important Public Members of the <tt>Value</tt> class</a>
2525 <div class="doc_text">
2528 <li><tt>Value::use_iterator</tt> - Typedef for iterator over the
2530 <tt>Value::use_const_iterator</tt> - Typedef for const_iterator over
2532 <tt>unsigned use_size()</tt> - Returns the number of users of the
2534 <tt>bool use_empty()</tt> - Returns true if there are no users.<br>
2535 <tt>use_iterator use_begin()</tt> - Get an iterator to the start of
2537 <tt>use_iterator use_end()</tt> - Get an iterator to the end of the
2539 <tt><a href="#User">User</a> *use_back()</tt> - Returns the last
2540 element in the list.
2541 <p> These methods are the interface to access the def-use
2542 information in LLVM. As with all other iterators in LLVM, the naming
2543 conventions follow the conventions defined by the <a href="#stl">STL</a>.</p>
2545 <li><tt><a href="#Type">Type</a> *getType() const</tt>
2546 <p>This method returns the Type of the Value.</p>
2548 <li><tt>bool hasName() const</tt><br>
2549 <tt>std::string getName() const</tt><br>
2550 <tt>void setName(const std::string &Name)</tt>
2551 <p> This family of methods is used to access and assign a name to a <tt>Value</tt>,
2552 be aware of the <a href="#nameWarning">precaution above</a>.</p>
2554 <li><tt>void replaceAllUsesWith(Value *V)</tt>
2556 <p>This method traverses the use list of a <tt>Value</tt> changing all <a
2557 href="#User"><tt>User</tt>s</a> of the current value to refer to
2558 "<tt>V</tt>" instead. For example, if you detect that an instruction always
2559 produces a constant value (for example through constant folding), you can
2560 replace all uses of the instruction with the constant like this:</p>
2562 <div class="doc_code">
2564 Inst->replaceAllUsesWith(ConstVal);
2572 <!-- ======================================================================= -->
2573 <div class="doc_subsection">
2574 <a name="User">The <tt>User</tt> class</a>
2577 <div class="doc_text">
2580 <tt>#include "<a href="/doxygen/User_8h-source.html">llvm/User.h</a>"</tt><br>
2581 doxygen info: <a href="/doxygen/classllvm_1_1User.html">User Class</a><br>
2582 Superclass: <a href="#Value"><tt>Value</tt></a></p>
2584 <p>The <tt>User</tt> class is the common base class of all LLVM nodes that may
2585 refer to <a href="#Value"><tt>Value</tt></a>s. It exposes a list of "Operands"
2586 that are all of the <a href="#Value"><tt>Value</tt></a>s that the User is
2587 referring to. The <tt>User</tt> class itself is a subclass of
2590 <p>The operands of a <tt>User</tt> point directly to the LLVM <a
2591 href="#Value"><tt>Value</tt></a> that it refers to. Because LLVM uses Static
2592 Single Assignment (SSA) form, there can only be one definition referred to,
2593 allowing this direct connection. This connection provides the use-def
2594 information in LLVM.</p>
2598 <!-- _______________________________________________________________________ -->
2599 <div class="doc_subsubsection">
2600 <a name="m_User">Important Public Members of the <tt>User</tt> class</a>
2603 <div class="doc_text">
2605 <p>The <tt>User</tt> class exposes the operand list in two ways: through
2606 an index access interface and through an iterator based interface.</p>
2609 <li><tt>Value *getOperand(unsigned i)</tt><br>
2610 <tt>unsigned getNumOperands()</tt>
2611 <p> These two methods expose the operands of the <tt>User</tt> in a
2612 convenient form for direct access.</p></li>
2614 <li><tt>User::op_iterator</tt> - Typedef for iterator over the operand
2616 <tt>op_iterator op_begin()</tt> - Get an iterator to the start of
2617 the operand list.<br>
2618 <tt>op_iterator op_end()</tt> - Get an iterator to the end of the
2620 <p> Together, these methods make up the iterator based interface to
2621 the operands of a <tt>User</tt>.</p></li>
2626 <!-- ======================================================================= -->
2627 <div class="doc_subsection">
2628 <a name="Instruction">The <tt>Instruction</tt> class</a>
2631 <div class="doc_text">
2633 <p><tt>#include "</tt><tt><a
2634 href="/doxygen/Instruction_8h-source.html">llvm/Instruction.h</a>"</tt><br>
2635 doxygen info: <a href="/doxygen/classllvm_1_1Instruction.html">Instruction Class</a><br>
2636 Superclasses: <a href="#User"><tt>User</tt></a>, <a
2637 href="#Value"><tt>Value</tt></a></p>
2639 <p>The <tt>Instruction</tt> class is the common base class for all LLVM
2640 instructions. It provides only a few methods, but is a very commonly used
2641 class. The primary data tracked by the <tt>Instruction</tt> class itself is the
2642 opcode (instruction type) and the parent <a
2643 href="#BasicBlock"><tt>BasicBlock</tt></a> the <tt>Instruction</tt> is embedded
2644 into. To represent a specific type of instruction, one of many subclasses of
2645 <tt>Instruction</tt> are used.</p>
2647 <p> Because the <tt>Instruction</tt> class subclasses the <a
2648 href="#User"><tt>User</tt></a> class, its operands can be accessed in the same
2649 way as for other <a href="#User"><tt>User</tt></a>s (with the
2650 <tt>getOperand()</tt>/<tt>getNumOperands()</tt> and
2651 <tt>op_begin()</tt>/<tt>op_end()</tt> methods).</p> <p> An important file for
2652 the <tt>Instruction</tt> class is the <tt>llvm/Instruction.def</tt> file. This
2653 file contains some meta-data about the various different types of instructions
2654 in LLVM. It describes the enum values that are used as opcodes (for example
2655 <tt>Instruction::Add</tt> and <tt>Instruction::ICmp</tt>), as well as the
2656 concrete sub-classes of <tt>Instruction</tt> that implement the instruction (for
2657 example <tt><a href="#BinaryOperator">BinaryOperator</a></tt> and <tt><a
2658 href="#CmpInst">CmpInst</a></tt>). Unfortunately, the use of macros in
2659 this file confuses doxygen, so these enum values don't show up correctly in the
2660 <a href="/doxygen/classllvm_1_1Instruction.html">doxygen output</a>.</p>
2664 <!-- _______________________________________________________________________ -->
2665 <div class="doc_subsubsection">
2666 <a name="s_Instruction">Important Subclasses of the <tt>Instruction</tt>
2669 <div class="doc_text">
2671 <li><tt><a name="BinaryOperator">BinaryOperator</a></tt>
2672 <p>This subclasses represents all two operand instructions whose operands
2673 must be the same type, except for the comparison instructions.</p></li>
2674 <li><tt><a name="CastInst">CastInst</a></tt>
2675 <p>This subclass is the parent of the 12 casting instructions. It provides
2676 common operations on cast instructions.</p>
2677 <li><tt><a name="CmpInst">CmpInst</a></tt>
2678 <p>This subclass respresents the two comparison instructions,
2679 <a href="LangRef.html#i_icmp">ICmpInst</a> (integer opreands), and
2680 <a href="LangRef.html#i_fcmp">FCmpInst</a> (floating point operands).</p>
2681 <li><tt><a name="TerminatorInst">TerminatorInst</a></tt>
2682 <p>This subclass is the parent of all terminator instructions (those which
2683 can terminate a block).</p>
2687 <!-- _______________________________________________________________________ -->
2688 <div class="doc_subsubsection">
2689 <a name="m_Instruction">Important Public Members of the <tt>Instruction</tt>
2693 <div class="doc_text">
2696 <li><tt><a href="#BasicBlock">BasicBlock</a> *getParent()</tt>
2697 <p>Returns the <a href="#BasicBlock"><tt>BasicBlock</tt></a> that
2698 this <tt>Instruction</tt> is embedded into.</p></li>
2699 <li><tt>bool mayWriteToMemory()</tt>
2700 <p>Returns true if the instruction writes to memory, i.e. it is a
2701 <tt>call</tt>,<tt>free</tt>,<tt>invoke</tt>, or <tt>store</tt>.</p></li>
2702 <li><tt>unsigned getOpcode()</tt>
2703 <p>Returns the opcode for the <tt>Instruction</tt>.</p></li>
2704 <li><tt><a href="#Instruction">Instruction</a> *clone() const</tt>
2705 <p>Returns another instance of the specified instruction, identical
2706 in all ways to the original except that the instruction has no parent
2707 (ie it's not embedded into a <a href="#BasicBlock"><tt>BasicBlock</tt></a>),
2708 and it has no name</p></li>
2713 <!-- ======================================================================= -->
2714 <div class="doc_subsection">
2715 <a name="Constant">The <tt>Constant</tt> class and subclasses</a>
2718 <div class="doc_text">
2720 <p>Constant represents a base class for different types of constants. It
2721 is subclassed by ConstantInt, ConstantArray, etc. for representing
2722 the various types of Constants. <a href="#GlobalValue">GlobalValue</a> is also
2723 a subclass, which represents the address of a global variable or function.
2728 <!-- _______________________________________________________________________ -->
2729 <div class="doc_subsubsection">Important Subclasses of Constant </div>
2730 <div class="doc_text">
2732 <li>ConstantInt : This subclass of Constant represents an integer constant of
2735 <li><tt>int64_t getSExtValue() const</tt>: Returns the underlying value of
2736 this constant as a sign extended signed integer value.</li>
2737 <li><tt>uint64_t getZExtValue() const</tt>: Returns the underlying value
2738 of this constant as a zero extended unsigned integer value.</li>
2739 <li><tt>static ConstantInt* get(const Type *Ty, uint64_t Val)</tt>:
2740 Returns the ConstantInt object that represents the value provided by
2741 <tt>Val</tt> for integer type <tt>Ty</tt>.</li>
2744 <li>ConstantFP : This class represents a floating point constant.
2746 <li><tt>double getValue() const</tt>: Returns the underlying value of
2747 this constant. </li>
2750 <li>ConstantArray : This represents a constant array.
2752 <li><tt>const std::vector<Use> &getValues() const</tt>: Returns
2753 a vector of component constants that makeup this array. </li>
2756 <li>ConstantStruct : This represents a constant struct.
2758 <li><tt>const std::vector<Use> &getValues() const</tt>: Returns
2759 a vector of component constants that makeup this array. </li>
2762 <li>GlobalValue : This represents either a global variable or a function. In
2763 either case, the value is a constant fixed address (after linking).
2769 <!-- ======================================================================= -->
2770 <div class="doc_subsection">
2771 <a name="GlobalValue">The <tt>GlobalValue</tt> class</a>
2774 <div class="doc_text">
2777 href="/doxygen/GlobalValue_8h-source.html">llvm/GlobalValue.h</a>"</tt><br>
2778 doxygen info: <a href="/doxygen/classllvm_1_1GlobalValue.html">GlobalValue
2780 Superclasses: <a href="#Constant"><tt>Constant</tt></a>,
2781 <a href="#User"><tt>User</tt></a>, <a href="#Value"><tt>Value</tt></a></p>
2783 <p>Global values (<a href="#GlobalVariable"><tt>GlobalVariable</tt></a>s or <a
2784 href="#Function"><tt>Function</tt></a>s) are the only LLVM values that are
2785 visible in the bodies of all <a href="#Function"><tt>Function</tt></a>s.
2786 Because they are visible at global scope, they are also subject to linking with
2787 other globals defined in different translation units. To control the linking
2788 process, <tt>GlobalValue</tt>s know their linkage rules. Specifically,
2789 <tt>GlobalValue</tt>s know whether they have internal or external linkage, as
2790 defined by the <tt>LinkageTypes</tt> enumeration.</p>
2792 <p>If a <tt>GlobalValue</tt> has internal linkage (equivalent to being
2793 <tt>static</tt> in C), it is not visible to code outside the current translation
2794 unit, and does not participate in linking. If it has external linkage, it is
2795 visible to external code, and does participate in linking. In addition to
2796 linkage information, <tt>GlobalValue</tt>s keep track of which <a
2797 href="#Module"><tt>Module</tt></a> they are currently part of.</p>
2799 <p>Because <tt>GlobalValue</tt>s are memory objects, they are always referred to
2800 by their <b>address</b>. As such, the <a href="#Type"><tt>Type</tt></a> of a
2801 global is always a pointer to its contents. It is important to remember this
2802 when using the <tt>GetElementPtrInst</tt> instruction because this pointer must
2803 be dereferenced first. For example, if you have a <tt>GlobalVariable</tt> (a
2804 subclass of <tt>GlobalValue)</tt> that is an array of 24 ints, type <tt>[24 x
2805 i32]</tt>, then the <tt>GlobalVariable</tt> is a pointer to that array. Although
2806 the address of the first element of this array and the value of the
2807 <tt>GlobalVariable</tt> are the same, they have different types. The
2808 <tt>GlobalVariable</tt>'s type is <tt>[24 x i32]</tt>. The first element's type
2809 is <tt>i32.</tt> Because of this, accessing a global value requires you to
2810 dereference the pointer with <tt>GetElementPtrInst</tt> first, then its elements
2811 can be accessed. This is explained in the <a href="LangRef.html#globalvars">LLVM
2812 Language Reference Manual</a>.</p>
2816 <!-- _______________________________________________________________________ -->
2817 <div class="doc_subsubsection">
2818 <a name="m_GlobalValue">Important Public Members of the <tt>GlobalValue</tt>
2822 <div class="doc_text">
2825 <li><tt>bool hasInternalLinkage() const</tt><br>
2826 <tt>bool hasExternalLinkage() const</tt><br>
2827 <tt>void setInternalLinkage(bool HasInternalLinkage)</tt>
2828 <p> These methods manipulate the linkage characteristics of the <tt>GlobalValue</tt>.</p>
2831 <li><tt><a href="#Module">Module</a> *getParent()</tt>
2832 <p> This returns the <a href="#Module"><tt>Module</tt></a> that the
2833 GlobalValue is currently embedded into.</p></li>
2838 <!-- ======================================================================= -->
2839 <div class="doc_subsection">
2840 <a name="Function">The <tt>Function</tt> class</a>
2843 <div class="doc_text">
2846 href="/doxygen/Function_8h-source.html">llvm/Function.h</a>"</tt><br> doxygen
2847 info: <a href="/doxygen/classllvm_1_1Function.html">Function Class</a><br>
2848 Superclasses: <a href="#GlobalValue"><tt>GlobalValue</tt></a>,
2849 <a href="#Constant"><tt>Constant</tt></a>,
2850 <a href="#User"><tt>User</tt></a>,
2851 <a href="#Value"><tt>Value</tt></a></p>
2853 <p>The <tt>Function</tt> class represents a single procedure in LLVM. It is
2854 actually one of the more complex classes in the LLVM heirarchy because it must
2855 keep track of a large amount of data. The <tt>Function</tt> class keeps track
2856 of a list of <a href="#BasicBlock"><tt>BasicBlock</tt></a>s, a list of formal
2857 <a href="#Argument"><tt>Argument</tt></a>s, and a
2858 <a href="#SymbolTable"><tt>SymbolTable</tt></a>.</p>
2860 <p>The list of <a href="#BasicBlock"><tt>BasicBlock</tt></a>s is the most
2861 commonly used part of <tt>Function</tt> objects. The list imposes an implicit
2862 ordering of the blocks in the function, which indicate how the code will be
2863 layed out by the backend. Additionally, the first <a
2864 href="#BasicBlock"><tt>BasicBlock</tt></a> is the implicit entry node for the
2865 <tt>Function</tt>. It is not legal in LLVM to explicitly branch to this initial
2866 block. There are no implicit exit nodes, and in fact there may be multiple exit
2867 nodes from a single <tt>Function</tt>. If the <a
2868 href="#BasicBlock"><tt>BasicBlock</tt></a> list is empty, this indicates that
2869 the <tt>Function</tt> is actually a function declaration: the actual body of the
2870 function hasn't been linked in yet.</p>
2872 <p>In addition to a list of <a href="#BasicBlock"><tt>BasicBlock</tt></a>s, the
2873 <tt>Function</tt> class also keeps track of the list of formal <a
2874 href="#Argument"><tt>Argument</tt></a>s that the function receives. This
2875 container manages the lifetime of the <a href="#Argument"><tt>Argument</tt></a>
2876 nodes, just like the <a href="#BasicBlock"><tt>BasicBlock</tt></a> list does for
2877 the <a href="#BasicBlock"><tt>BasicBlock</tt></a>s.</p>
2879 <p>The <a href="#SymbolTable"><tt>SymbolTable</tt></a> is a very rarely used
2880 LLVM feature that is only used when you have to look up a value by name. Aside
2881 from that, the <a href="#SymbolTable"><tt>SymbolTable</tt></a> is used
2882 internally to make sure that there are not conflicts between the names of <a
2883 href="#Instruction"><tt>Instruction</tt></a>s, <a
2884 href="#BasicBlock"><tt>BasicBlock</tt></a>s, or <a
2885 href="#Argument"><tt>Argument</tt></a>s in the function body.</p>
2887 <p>Note that <tt>Function</tt> is a <a href="#GlobalValue">GlobalValue</a>
2888 and therefore also a <a href="#Constant">Constant</a>. The value of the function
2889 is its address (after linking) which is guaranteed to be constant.</p>
2892 <!-- _______________________________________________________________________ -->
2893 <div class="doc_subsubsection">
2894 <a name="m_Function">Important Public Members of the <tt>Function</tt>
2898 <div class="doc_text">
2901 <li><tt>Function(const </tt><tt><a href="#FunctionType">FunctionType</a>
2902 *Ty, LinkageTypes Linkage, const std::string &N = "", Module* Parent = 0)</tt>
2904 <p>Constructor used when you need to create new <tt>Function</tt>s to add
2905 the the program. The constructor must specify the type of the function to
2906 create and what type of linkage the function should have. The <a
2907 href="#FunctionType"><tt>FunctionType</tt></a> argument
2908 specifies the formal arguments and return value for the function. The same
2909 <a href="#FunctionTypel"><tt>FunctionType</tt></a> value can be used to
2910 create multiple functions. The <tt>Parent</tt> argument specifies the Module
2911 in which the function is defined. If this argument is provided, the function
2912 will automatically be inserted into that module's list of
2915 <li><tt>bool isExternal()</tt>
2917 <p>Return whether or not the <tt>Function</tt> has a body defined. If the
2918 function is "external", it does not have a body, and thus must be resolved
2919 by linking with a function defined in a different translation unit.</p></li>
2921 <li><tt>Function::iterator</tt> - Typedef for basic block list iterator<br>
2922 <tt>Function::const_iterator</tt> - Typedef for const_iterator.<br>
2924 <tt>begin()</tt>, <tt>end()</tt>
2925 <tt>size()</tt>, <tt>empty()</tt>
2927 <p>These are forwarding methods that make it easy to access the contents of
2928 a <tt>Function</tt> object's <a href="#BasicBlock"><tt>BasicBlock</tt></a>
2931 <li><tt>Function::BasicBlockListType &getBasicBlockList()</tt>
2933 <p>Returns the list of <a href="#BasicBlock"><tt>BasicBlock</tt></a>s. This
2934 is necessary to use when you need to update the list or perform a complex
2935 action that doesn't have a forwarding method.</p></li>
2937 <li><tt>Function::arg_iterator</tt> - Typedef for the argument list
2939 <tt>Function::const_arg_iterator</tt> - Typedef for const_iterator.<br>
2941 <tt>arg_begin()</tt>, <tt>arg_end()</tt>
2942 <tt>arg_size()</tt>, <tt>arg_empty()</tt>
2944 <p>These are forwarding methods that make it easy to access the contents of
2945 a <tt>Function</tt> object's <a href="#Argument"><tt>Argument</tt></a>
2948 <li><tt>Function::ArgumentListType &getArgumentList()</tt>
2950 <p>Returns the list of <a href="#Argument"><tt>Argument</tt></a>s. This is
2951 necessary to use when you need to update the list or perform a complex
2952 action that doesn't have a forwarding method.</p></li>
2954 <li><tt><a href="#BasicBlock">BasicBlock</a> &getEntryBlock()</tt>
2956 <p>Returns the entry <a href="#BasicBlock"><tt>BasicBlock</tt></a> for the
2957 function. Because the entry block for the function is always the first
2958 block, this returns the first block of the <tt>Function</tt>.</p></li>
2960 <li><tt><a href="#Type">Type</a> *getReturnType()</tt><br>
2961 <tt><a href="#FunctionType">FunctionType</a> *getFunctionType()</tt>
2963 <p>This traverses the <a href="#Type"><tt>Type</tt></a> of the
2964 <tt>Function</tt> and returns the return type of the function, or the <a
2965 href="#FunctionType"><tt>FunctionType</tt></a> of the actual
2968 <li><tt><a href="#SymbolTable">SymbolTable</a> *getSymbolTable()</tt>
2970 <p> Return a pointer to the <a href="#SymbolTable"><tt>SymbolTable</tt></a>
2971 for this <tt>Function</tt>.</p></li>
2976 <!-- ======================================================================= -->
2977 <div class="doc_subsection">
2978 <a name="GlobalVariable">The <tt>GlobalVariable</tt> class</a>
2981 <div class="doc_text">
2984 href="/doxygen/GlobalVariable_8h-source.html">llvm/GlobalVariable.h</a>"</tt>
2986 doxygen info: <a href="/doxygen/classllvm_1_1GlobalVariable.html">GlobalVariable
2988 Superclasses: <a href="#GlobalValue"><tt>GlobalValue</tt></a>,
2989 <a href="#Constant"><tt>Constant</tt></a>,
2990 <a href="#User"><tt>User</tt></a>,
2991 <a href="#Value"><tt>Value</tt></a></p>
2993 <p>Global variables are represented with the (suprise suprise)
2994 <tt>GlobalVariable</tt> class. Like functions, <tt>GlobalVariable</tt>s are also
2995 subclasses of <a href="#GlobalValue"><tt>GlobalValue</tt></a>, and as such are
2996 always referenced by their address (global values must live in memory, so their
2997 "name" refers to their constant address). See
2998 <a href="#GlobalValue"><tt>GlobalValue</tt></a> for more on this. Global
2999 variables may have an initial value (which must be a
3000 <a href="#Constant"><tt>Constant</tt></a>), and if they have an initializer,
3001 they may be marked as "constant" themselves (indicating that their contents
3002 never change at runtime).</p>
3005 <!-- _______________________________________________________________________ -->
3006 <div class="doc_subsubsection">
3007 <a name="m_GlobalVariable">Important Public Members of the
3008 <tt>GlobalVariable</tt> class</a>
3011 <div class="doc_text">
3014 <li><tt>GlobalVariable(const </tt><tt><a href="#Type">Type</a> *Ty, bool
3015 isConstant, LinkageTypes& Linkage, <a href="#Constant">Constant</a>
3016 *Initializer = 0, const std::string &Name = "", Module* Parent = 0)</tt>
3018 <p>Create a new global variable of the specified type. If
3019 <tt>isConstant</tt> is true then the global variable will be marked as
3020 unchanging for the program. The Linkage parameter specifies the type of
3021 linkage (internal, external, weak, linkonce, appending) for the variable. If
3022 the linkage is InternalLinkage, WeakLinkage, or LinkOnceLinkage, then
3023 the resultant global variable will have internal linkage. AppendingLinkage
3024 concatenates together all instances (in different translation units) of the
3025 variable into a single variable but is only applicable to arrays. See
3026 the <a href="LangRef.html#modulestructure">LLVM Language Reference</a> for
3027 further details on linkage types. Optionally an initializer, a name, and the
3028 module to put the variable into may be specified for the global variable as
3031 <li><tt>bool isConstant() const</tt>
3033 <p>Returns true if this is a global variable that is known not to
3034 be modified at runtime.</p></li>
3036 <li><tt>bool hasInitializer()</tt>
3038 <p>Returns true if this <tt>GlobalVariable</tt> has an intializer.</p></li>
3040 <li><tt><a href="#Constant">Constant</a> *getInitializer()</tt>
3042 <p>Returns the intial value for a <tt>GlobalVariable</tt>. It is not legal
3043 to call this method if there is no initializer.</p></li>
3049 <!-- ======================================================================= -->
3050 <div class="doc_subsection">
3051 <a name="BasicBlock">The <tt>BasicBlock</tt> class</a>
3054 <div class="doc_text">
3057 href="/doxygen/BasicBlock_8h-source.html">llvm/BasicBlock.h</a>"</tt><br>
3058 doxygen info: <a href="/doxygen/structllvm_1_1BasicBlock.html">BasicBlock
3060 Superclass: <a href="#Value"><tt>Value</tt></a></p>
3062 <p>This class represents a single entry multiple exit section of the code,
3063 commonly known as a basic block by the compiler community. The
3064 <tt>BasicBlock</tt> class maintains a list of <a
3065 href="#Instruction"><tt>Instruction</tt></a>s, which form the body of the block.
3066 Matching the language definition, the last element of this list of instructions
3067 is always a terminator instruction (a subclass of the <a
3068 href="#TerminatorInst"><tt>TerminatorInst</tt></a> class).</p>
3070 <p>In addition to tracking the list of instructions that make up the block, the
3071 <tt>BasicBlock</tt> class also keeps track of the <a
3072 href="#Function"><tt>Function</tt></a> that it is embedded into.</p>
3074 <p>Note that <tt>BasicBlock</tt>s themselves are <a
3075 href="#Value"><tt>Value</tt></a>s, because they are referenced by instructions
3076 like branches and can go in the switch tables. <tt>BasicBlock</tt>s have type
3081 <!-- _______________________________________________________________________ -->
3082 <div class="doc_subsubsection">
3083 <a name="m_BasicBlock">Important Public Members of the <tt>BasicBlock</tt>
3087 <div class="doc_text">
3090 <li><tt>BasicBlock(const std::string &Name = "", </tt><tt><a
3091 href="#Function">Function</a> *Parent = 0)</tt>
3093 <p>The <tt>BasicBlock</tt> constructor is used to create new basic blocks for
3094 insertion into a function. The constructor optionally takes a name for the new
3095 block, and a <a href="#Function"><tt>Function</tt></a> to insert it into. If
3096 the <tt>Parent</tt> parameter is specified, the new <tt>BasicBlock</tt> is
3097 automatically inserted at the end of the specified <a
3098 href="#Function"><tt>Function</tt></a>, if not specified, the BasicBlock must be
3099 manually inserted into the <a href="#Function"><tt>Function</tt></a>.</p></li>
3101 <li><tt>BasicBlock::iterator</tt> - Typedef for instruction list iterator<br>
3102 <tt>BasicBlock::const_iterator</tt> - Typedef for const_iterator.<br>
3103 <tt>begin()</tt>, <tt>end()</tt>, <tt>front()</tt>, <tt>back()</tt>,
3104 <tt>size()</tt>, <tt>empty()</tt>
3105 STL-style functions for accessing the instruction list.
3107 <p>These methods and typedefs are forwarding functions that have the same
3108 semantics as the standard library methods of the same names. These methods
3109 expose the underlying instruction list of a basic block in a way that is easy to
3110 manipulate. To get the full complement of container operations (including
3111 operations to update the list), you must use the <tt>getInstList()</tt>
3114 <li><tt>BasicBlock::InstListType &getInstList()</tt>
3116 <p>This method is used to get access to the underlying container that actually
3117 holds the Instructions. This method must be used when there isn't a forwarding
3118 function in the <tt>BasicBlock</tt> class for the operation that you would like
3119 to perform. Because there are no forwarding functions for "updating"
3120 operations, you need to use this if you want to update the contents of a
3121 <tt>BasicBlock</tt>.</p></li>
3123 <li><tt><a href="#Function">Function</a> *getParent()</tt>
3125 <p> Returns a pointer to <a href="#Function"><tt>Function</tt></a> the block is
3126 embedded into, or a null pointer if it is homeless.</p></li>
3128 <li><tt><a href="#TerminatorInst">TerminatorInst</a> *getTerminator()</tt>
3130 <p> Returns a pointer to the terminator instruction that appears at the end of
3131 the <tt>BasicBlock</tt>. If there is no terminator instruction, or if the last
3132 instruction in the block is not a terminator, then a null pointer is
3140 <!-- ======================================================================= -->
3141 <div class="doc_subsection">
3142 <a name="Argument">The <tt>Argument</tt> class</a>
3145 <div class="doc_text">
3147 <p>This subclass of Value defines the interface for incoming formal
3148 arguments to a function. A Function maintains a list of its formal
3149 arguments. An argument has a pointer to the parent Function.</p>
3153 <!-- *********************************************************************** -->
3156 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
3157 src="http://jigsaw.w3.org/css-validator/images/vcss" alt="Valid CSS!"></a>
3158 <a href="http://validator.w3.org/check/referer"><img
3159 src="http://www.w3.org/Icons/valid-html401" alt="Valid HTML 4.01!" /></a>
3161 <a href="mailto:dhurjati@cs.uiuc.edu">Dinakar Dhurjati</a> and
3162 <a href="mailto:sabre@nondot.org">Chris Lattner</a><br>
3163 <a href="http://llvm.org">The LLVM Compiler Infrastructure</a><br>
3164 Last modified: $Date$