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_stringmap">"llvm/ADT/StringMap.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>opt</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 :).
801 <p>One worthwhile note about std::vector: avoid code like this:</p>
803 <div class="doc_code">
812 <p>Instead, write this as:</p>
814 <div class="doc_code">
824 <p>Doing so will save (at least) one heap allocation and free per iteration of
829 <!-- _______________________________________________________________________ -->
830 <div class="doc_subsubsection">
831 <a name="dss_deque"><deque></a>
834 <div class="doc_text">
835 <p>std::deque is, in some senses, a generalized version of std::vector. Like
836 std::vector, it provides constant time random access and other similar
837 properties, but it also provides efficient access to the front of the list. It
838 does not guarantee continuity of elements within memory.</p>
840 <p>In exchange for this extra flexibility, std::deque has significantly higher
841 constant factor costs than std::vector. If possible, use std::vector or
842 something cheaper.</p>
845 <!-- _______________________________________________________________________ -->
846 <div class="doc_subsubsection">
847 <a name="dss_list"><list></a>
850 <div class="doc_text">
851 <p>std::list is an extremely inefficient class that is rarely useful.
852 It performs a heap allocation for every element inserted into it, thus having an
853 extremely high constant factor, particularly for small data types. std::list
854 also only supports bidirectional iteration, not random access iteration.</p>
856 <p>In exchange for this high cost, std::list supports efficient access to both
857 ends of the list (like std::deque, but unlike std::vector or SmallVector). In
858 addition, the iterator invalidation characteristics of std::list are stronger
859 than that of a vector class: inserting or removing an element into the list does
860 not invalidate iterator or pointers to other elements in the list.</p>
863 <!-- _______________________________________________________________________ -->
864 <div class="doc_subsubsection">
865 <a name="dss_ilist">llvm/ADT/ilist</a>
868 <div class="doc_text">
869 <p><tt>ilist<T></tt> implements an 'intrusive' doubly-linked list. It is
870 intrusive, because it requires the element to store and provide access to the
871 prev/next pointers for the list.</p>
873 <p>ilist has the same drawbacks as std::list, and additionally requires an
874 ilist_traits implementation for the element type, but it provides some novel
875 characteristics. In particular, it can efficiently store polymorphic objects,
876 the traits class is informed when an element is inserted or removed from the
877 list, and ilists are guaranteed to support a constant-time splice operation.
880 <p>These properties are exactly what we want for things like Instructions and
881 basic blocks, which is why these are implemented with ilists.</p>
884 <!-- _______________________________________________________________________ -->
885 <div class="doc_subsubsection">
886 <a name="dss_other">Other Sequential Container options</a>
889 <div class="doc_text">
890 <p>Other STL containers are available, such as std::string.</p>
892 <p>There are also various STL adapter classes such as std::queue,
893 std::priority_queue, std::stack, etc. These provide simplified access to an
894 underlying container but don't affect the cost of the container itself.</p>
899 <!-- ======================================================================= -->
900 <div class="doc_subsection">
901 <a name="ds_set">Set-Like Containers (std::set, SmallSet, SetVector, etc)</a>
904 <div class="doc_text">
906 <p>Set-like containers are useful when you need to canonicalize multiple values
907 into a single representation. There are several different choices for how to do
908 this, providing various trade-offs.</p>
913 <!-- _______________________________________________________________________ -->
914 <div class="doc_subsubsection">
915 <a name="dss_sortedvectorset">A sorted 'vector'</a>
918 <div class="doc_text">
920 <p>If you intend to insert a lot of elements, then do a lot of queries, a
921 great approach is to use a vector (or other sequential container) with
922 std::sort+std::unique to remove duplicates. This approach works really well if
923 your usage pattern has these two distinct phases (insert then query), and can be
924 coupled with a good choice of <a href="#ds_sequential">sequential container</a>.
928 This combination provides the several nice properties: the result data is
929 contiguous in memory (good for cache locality), has few allocations, is easy to
930 address (iterators in the final vector are just indices or pointers), and can be
931 efficiently queried with a standard binary or radix search.</p>
935 <!-- _______________________________________________________________________ -->
936 <div class="doc_subsubsection">
937 <a name="dss_smallset">"llvm/ADT/SmallSet.h"</a>
940 <div class="doc_text">
942 <p>If you have a set-like data structure that is usually small and whose elements
943 are reasonably small, a <tt>SmallSet<Type, N></tt> is a good choice. This set
944 has space for N elements in place (thus, if the set is dynamically smaller than
945 N, no malloc traffic is required) and accesses them with a simple linear search.
946 When the set grows beyond 'N' elements, it allocates a more expensive representation that
947 guarantees efficient access (for most types, it falls back to std::set, but for
948 pointers it uses something far better, <a
949 href="#dss_smallptrset">SmallPtrSet</a>).</p>
951 <p>The magic of this class is that it handles small sets extremely efficiently,
952 but gracefully handles extremely large sets without loss of efficiency. The
953 drawback is that the interface is quite small: it supports insertion, queries
954 and erasing, but does not support iteration.</p>
958 <!-- _______________________________________________________________________ -->
959 <div class="doc_subsubsection">
960 <a name="dss_smallptrset">"llvm/ADT/SmallPtrSet.h"</a>
963 <div class="doc_text">
965 <p>SmallPtrSet has all the advantages of SmallSet (and a SmallSet of pointers is
966 transparently implemented with a SmallPtrSet), but also supports iterators. If
967 more than 'N' insertions are performed, a single quadratically
968 probed hash table is allocated and grows as needed, providing extremely
969 efficient access (constant time insertion/deleting/queries with low constant
970 factors) and is very stingy with malloc traffic.</p>
972 <p>Note that, unlike std::set, the iterators of SmallPtrSet are invalidated
973 whenever an insertion occurs. Also, the values visited by the iterators are not
974 visited in sorted order.</p>
978 <!-- _______________________________________________________________________ -->
979 <div class="doc_subsubsection">
980 <a name="dss_FoldingSet">"llvm/ADT/FoldingSet.h"</a>
983 <div class="doc_text">
986 FoldingSet is an aggregate class that is really good at uniquing
987 expensive-to-create or polymorphic objects. It is a combination of a chained
988 hash table with intrusive links (uniqued objects are required to inherit from
989 FoldingSetNode) that uses <a href="#dss_smallvector">SmallVector</a> as part of
992 <p>Consider a case where you want to implement a "getOrCreateFoo" method for
993 a complex object (for example, a node in the code generator). The client has a
994 description of *what* it wants to generate (it knows the opcode and all the
995 operands), but we don't want to 'new' a node, then try inserting it into a set
996 only to find out it already exists, at which point we would have to delete it
997 and return the node that already exists.
1000 <p>To support this style of client, FoldingSet perform a query with a
1001 FoldingSetNodeID (which wraps SmallVector) that can be used to describe the
1002 element that we want to query for. The query either returns the element
1003 matching the ID or it returns an opaque ID that indicates where insertion should
1004 take place. Construction of the ID usually does not require heap traffic.</p>
1006 <p>Because FoldingSet uses intrusive links, it can support polymorphic objects
1007 in the set (for example, you can have SDNode instances mixed with LoadSDNodes).
1008 Because the elements are individually allocated, pointers to the elements are
1009 stable: inserting or removing elements does not invalidate any pointers to other
1015 <!-- _______________________________________________________________________ -->
1016 <div class="doc_subsubsection">
1017 <a name="dss_set"><set></a>
1020 <div class="doc_text">
1022 <p><tt>std::set</tt> is a reasonable all-around set class, which is decent at
1023 many things but great at nothing. std::set allocates memory for each element
1024 inserted (thus it is very malloc intensive) and typically stores three pointers
1025 per element in the set (thus adding a large amount of per-element space
1026 overhead). It offers guaranteed log(n) performance, which is not particularly
1027 fast from a complexity standpoint (particularly if the elements of the set are
1028 expensive to compare, like strings), and has extremely high constant factors for
1029 lookup, insertion and removal.</p>
1031 <p>The advantages of std::set are that its iterators are stable (deleting or
1032 inserting an element from the set does not affect iterators or pointers to other
1033 elements) and that iteration over the set is guaranteed to be in sorted order.
1034 If the elements in the set are large, then the relative overhead of the pointers
1035 and malloc traffic is not a big deal, but if the elements of the set are small,
1036 std::set is almost never a good choice.</p>
1040 <!-- _______________________________________________________________________ -->
1041 <div class="doc_subsubsection">
1042 <a name="dss_setvector">"llvm/ADT/SetVector.h"</a>
1045 <div class="doc_text">
1046 <p>LLVM's SetVector<Type> is an adapter class that combines your choice of
1047 a set-like container along with a <a href="#ds_sequential">Sequential
1048 Container</a>. The important property
1049 that this provides is efficient insertion with uniquing (duplicate elements are
1050 ignored) with iteration support. It implements this by inserting elements into
1051 both a set-like container and the sequential container, using the set-like
1052 container for uniquing and the sequential container for iteration.
1055 <p>The difference between SetVector and other sets is that the order of
1056 iteration is guaranteed to match the order of insertion into the SetVector.
1057 This property is really important for things like sets of pointers. Because
1058 pointer values are non-deterministic (e.g. vary across runs of the program on
1059 different machines), iterating over the pointers in the set will
1060 not be in a well-defined order.</p>
1063 The drawback of SetVector is that it requires twice as much space as a normal
1064 set and has the sum of constant factors from the set-like container and the
1065 sequential container that it uses. Use it *only* if you need to iterate over
1066 the elements in a deterministic order. SetVector is also expensive to delete
1067 elements out of (linear time), unless you use it's "pop_back" method, which is
1071 <p>SetVector is an adapter class that defaults to using std::vector and std::set
1072 for the underlying containers, so it is quite expensive. However,
1073 <tt>"llvm/ADT/SetVector.h"</tt> also provides a SmallSetVector class, which
1074 defaults to using a SmallVector and SmallSet of a specified size. If you use
1075 this, and if your sets are dynamically smaller than N, you will save a lot of
1080 <!-- _______________________________________________________________________ -->
1081 <div class="doc_subsubsection">
1082 <a name="dss_uniquevector">"llvm/ADT/UniqueVector.h"</a>
1085 <div class="doc_text">
1088 UniqueVector is similar to <a href="#dss_setvector">SetVector</a>, but it
1089 retains a unique ID for each element inserted into the set. It internally
1090 contains a map and a vector, and it assigns a unique ID for each value inserted
1093 <p>UniqueVector is very expensive: its cost is the sum of the cost of
1094 maintaining both the map and vector, it has high complexity, high constant
1095 factors, and produces a lot of malloc traffic. It should be avoided.</p>
1100 <!-- _______________________________________________________________________ -->
1101 <div class="doc_subsubsection">
1102 <a name="dss_otherset">Other Set-Like Container Options</a>
1105 <div class="doc_text">
1108 The STL provides several other options, such as std::multiset and the various
1109 "hash_set" like containers (whether from C++ TR1 or from the SGI library).</p>
1111 <p>std::multiset is useful if you're not interested in elimination of
1112 duplicates, but has all the drawbacks of std::set. A sorted vector (where you
1113 don't delete duplicate entries) or some other approach is almost always
1116 <p>The various hash_set implementations (exposed portably by
1117 "llvm/ADT/hash_set") is a simple chained hashtable. This algorithm is as malloc
1118 intensive as std::set (performing an allocation for each element inserted,
1119 thus having really high constant factors) but (usually) provides O(1)
1120 insertion/deletion of elements. This can be useful if your elements are large
1121 (thus making the constant-factor cost relatively low) or if comparisons are
1122 expensive. Element iteration does not visit elements in a useful order.</p>
1126 <!-- ======================================================================= -->
1127 <div class="doc_subsection">
1128 <a name="ds_map">Map-Like Containers (std::map, DenseMap, etc)</a>
1131 <div class="doc_text">
1132 Map-like containers are useful when you want to associate data to a key. As
1133 usual, there are a lot of different ways to do this. :)
1136 <!-- _______________________________________________________________________ -->
1137 <div class="doc_subsubsection">
1138 <a name="dss_sortedvectormap">A sorted 'vector'</a>
1141 <div class="doc_text">
1144 If your usage pattern follows a strict insert-then-query approach, you can
1145 trivially use the same approach as <a href="#dss_sortedvectorset">sorted vectors
1146 for set-like containers</a>. The only difference is that your query function
1147 (which uses std::lower_bound to get efficient log(n) lookup) should only compare
1148 the key, not both the key and value. This yields the same advantages as sorted
1153 <!-- _______________________________________________________________________ -->
1154 <div class="doc_subsubsection">
1155 <a name="dss_stringmap">"llvm/ADT/StringMap.h"</a>
1158 <div class="doc_text">
1161 Strings are commonly used as keys in maps, and they are difficult to support
1162 efficiently: they are variable length, inefficient to hash and compare when
1163 long, expensive to copy, etc. StringMap is a specialized container designed to
1164 cope with these issues. It supports mapping an arbitrary range of bytes to an
1165 arbitrary other object.</p>
1167 <p>The StringMap implementation uses a quadratically-probed hash table, where
1168 the buckets store a pointer to the heap allocated entries (and some other
1169 stuff). The entries in the map must be heap allocated because the strings are
1170 variable length. The string data (key) and the element object (value) are
1171 stored in the same allocation with the string data immediately after the element
1172 object. This container guarantees the "<tt>(char*)(&Value+1)</tt>" points
1173 to the key string for a value.</p>
1175 <p>The StringMap is very fast for several reasons: quadratic probing is very
1176 cache efficient for lookups, the hash value of strings in buckets is not
1177 recomputed when lookup up an element, StringMap rarely has to touch the
1178 memory for unrelated objects when looking up a value (even when hash collisions
1179 happen), hash table growth does not recompute the hash values for strings
1180 already in the table, and each pair in the map is store in a single allocation
1181 (the string data is stored in the same allocation as the Value of a pair).</p>
1183 <p>StringMap also provides query methods that take byte ranges, so it only ever
1184 copies a string if a value is inserted into the table.</p>
1187 <!-- _______________________________________________________________________ -->
1188 <div class="doc_subsubsection">
1189 <a name="dss_indexedmap">"llvm/ADT/IndexedMap.h"</a>
1192 <div class="doc_text">
1194 IndexedMap is a specialized container for mapping small dense integers (or
1195 values that can be mapped to small dense integers) to some other type. It is
1196 internally implemented as a vector with a mapping function that maps the keys to
1197 the dense integer range.
1201 This is useful for cases like virtual registers in the LLVM code generator: they
1202 have a dense mapping that is offset by a compile-time constant (the first
1203 virtual register ID).</p>
1207 <!-- _______________________________________________________________________ -->
1208 <div class="doc_subsubsection">
1209 <a name="dss_densemap">"llvm/ADT/DenseMap.h"</a>
1212 <div class="doc_text">
1215 DenseMap is a simple quadratically probed hash table. It excels at supporting
1216 small keys and values: it uses a single allocation to hold all of the pairs that
1217 are currently inserted in the map. DenseMap is a great way to map pointers to
1218 pointers, or map other small types to each other.
1222 There are several aspects of DenseMap that you should be aware of, however. The
1223 iterators in a densemap are invalidated whenever an insertion occurs, unlike
1224 map. Also, because DenseMap allocates space for a large number of key/value
1225 pairs (it starts with 64 by default), it will waste a lot of space if your keys
1226 or values are large. Finally, you must implement a partial specialization of
1227 DenseMapKeyInfo for the key that you want, if it isn't already supported. This
1228 is required to tell DenseMap about two special marker values (which can never be
1229 inserted into the map) that it needs internally.</p>
1233 <!-- _______________________________________________________________________ -->
1234 <div class="doc_subsubsection">
1235 <a name="dss_map"><map></a>
1238 <div class="doc_text">
1241 std::map has similar characteristics to <a href="#dss_set">std::set</a>: it uses
1242 a single allocation per pair inserted into the map, it offers log(n) lookup with
1243 an extremely large constant factor, imposes a space penalty of 3 pointers per
1244 pair in the map, etc.</p>
1246 <p>std::map is most useful when your keys or values are very large, if you need
1247 to iterate over the collection in sorted order, or if you need stable iterators
1248 into the map (i.e. they don't get invalidated if an insertion or deletion of
1249 another element takes place).</p>
1253 <!-- _______________________________________________________________________ -->
1254 <div class="doc_subsubsection">
1255 <a name="dss_othermap">Other Map-Like Container Options</a>
1258 <div class="doc_text">
1261 The STL provides several other options, such as std::multimap and the various
1262 "hash_map" like containers (whether from C++ TR1 or from the SGI library).</p>
1264 <p>std::multimap is useful if you want to map a key to multiple values, but has
1265 all the drawbacks of std::map. A sorted vector or some other approach is almost
1268 <p>The various hash_map implementations (exposed portably by
1269 "llvm/ADT/hash_map") are simple chained hash tables. This algorithm is as
1270 malloc intensive as std::map (performing an allocation for each element
1271 inserted, thus having really high constant factors) but (usually) provides O(1)
1272 insertion/deletion of elements. This can be useful if your elements are large
1273 (thus making the constant-factor cost relatively low) or if comparisons are
1274 expensive. Element iteration does not visit elements in a useful order.</p>
1279 <!-- *********************************************************************** -->
1280 <div class="doc_section">
1281 <a name="common">Helpful Hints for Common Operations</a>
1283 <!-- *********************************************************************** -->
1285 <div class="doc_text">
1287 <p>This section describes how to perform some very simple transformations of
1288 LLVM code. This is meant to give examples of common idioms used, showing the
1289 practical side of LLVM transformations. <p> Because this is a "how-to" section,
1290 you should also read about the main classes that you will be working with. The
1291 <a href="#coreclasses">Core LLVM Class Hierarchy Reference</a> contains details
1292 and descriptions of the main classes that you should know about.</p>
1296 <!-- NOTE: this section should be heavy on example code -->
1297 <!-- ======================================================================= -->
1298 <div class="doc_subsection">
1299 <a name="inspection">Basic Inspection and Traversal Routines</a>
1302 <div class="doc_text">
1304 <p>The LLVM compiler infrastructure have many different data structures that may
1305 be traversed. Following the example of the C++ standard template library, the
1306 techniques used to traverse these various data structures are all basically the
1307 same. For a enumerable sequence of values, the <tt>XXXbegin()</tt> function (or
1308 method) returns an iterator to the start of the sequence, the <tt>XXXend()</tt>
1309 function returns an iterator pointing to one past the last valid element of the
1310 sequence, and there is some <tt>XXXiterator</tt> data type that is common
1311 between the two operations.</p>
1313 <p>Because the pattern for iteration is common across many different aspects of
1314 the program representation, the standard template library algorithms may be used
1315 on them, and it is easier to remember how to iterate. First we show a few common
1316 examples of the data structures that need to be traversed. Other data
1317 structures are traversed in very similar ways.</p>
1321 <!-- _______________________________________________________________________ -->
1322 <div class="doc_subsubsection">
1323 <a name="iterate_function">Iterating over the </a><a
1324 href="#BasicBlock"><tt>BasicBlock</tt></a>s in a <a
1325 href="#Function"><tt>Function</tt></a>
1328 <div class="doc_text">
1330 <p>It's quite common to have a <tt>Function</tt> instance that you'd like to
1331 transform in some way; in particular, you'd like to manipulate its
1332 <tt>BasicBlock</tt>s. To facilitate this, you'll need to iterate over all of
1333 the <tt>BasicBlock</tt>s that constitute the <tt>Function</tt>. The following is
1334 an example that prints the name of a <tt>BasicBlock</tt> and the number of
1335 <tt>Instruction</tt>s it contains:</p>
1337 <div class="doc_code">
1339 // <i>func is a pointer to a Function instance</i>
1340 for (Function::iterator i = func->begin(), e = func->end(); i != e; ++i)
1341 // <i>Print out the name of the basic block if it has one, and then the</i>
1342 // <i>number of instructions that it contains</i>
1343 llvm::cerr << "Basic block (name=" << i->getName() << ") has "
1344 << i->size() << " instructions.\n";
1348 <p>Note that i can be used as if it were a pointer for the purposes of
1349 invoking member functions of the <tt>Instruction</tt> class. This is
1350 because the indirection operator is overloaded for the iterator
1351 classes. In the above code, the expression <tt>i->size()</tt> is
1352 exactly equivalent to <tt>(*i).size()</tt> just like you'd expect.</p>
1356 <!-- _______________________________________________________________________ -->
1357 <div class="doc_subsubsection">
1358 <a name="iterate_basicblock">Iterating over the </a><a
1359 href="#Instruction"><tt>Instruction</tt></a>s in a <a
1360 href="#BasicBlock"><tt>BasicBlock</tt></a>
1363 <div class="doc_text">
1365 <p>Just like when dealing with <tt>BasicBlock</tt>s in <tt>Function</tt>s, it's
1366 easy to iterate over the individual instructions that make up
1367 <tt>BasicBlock</tt>s. Here's a code snippet that prints out each instruction in
1368 a <tt>BasicBlock</tt>:</p>
1370 <div class="doc_code">
1372 // <i>blk is a pointer to a BasicBlock instance</i>
1373 for (BasicBlock::iterator i = blk->begin(), e = blk->end(); i != e; ++i)
1374 // <i>The next statement works since operator<<(ostream&,...)</i>
1375 // <i>is overloaded for Instruction&</i>
1376 llvm::cerr << *i << "\n";
1380 <p>However, this isn't really the best way to print out the contents of a
1381 <tt>BasicBlock</tt>! Since the ostream operators are overloaded for virtually
1382 anything you'll care about, you could have just invoked the print routine on the
1383 basic block itself: <tt>llvm::cerr << *blk << "\n";</tt>.</p>
1387 <!-- _______________________________________________________________________ -->
1388 <div class="doc_subsubsection">
1389 <a name="iterate_institer">Iterating over the </a><a
1390 href="#Instruction"><tt>Instruction</tt></a>s in a <a
1391 href="#Function"><tt>Function</tt></a>
1394 <div class="doc_text">
1396 <p>If you're finding that you commonly iterate over a <tt>Function</tt>'s
1397 <tt>BasicBlock</tt>s and then that <tt>BasicBlock</tt>'s <tt>Instruction</tt>s,
1398 <tt>InstIterator</tt> should be used instead. You'll need to include <a
1399 href="/doxygen/InstIterator_8h-source.html"><tt>llvm/Support/InstIterator.h</tt></a>,
1400 and then instantiate <tt>InstIterator</tt>s explicitly in your code. Here's a
1401 small example that shows how to dump all instructions in a function to the standard error stream:<p>
1403 <div class="doc_code">
1405 #include "<a href="/doxygen/InstIterator_8h-source.html">llvm/Support/InstIterator.h</a>"
1407 // <i>F is a pointer to a Function instance</i>
1408 for (inst_iterator i = inst_begin(F), e = inst_end(F); i != e; ++i)
1409 llvm::cerr << *i << "\n";
1413 <p>Easy, isn't it? You can also use <tt>InstIterator</tt>s to fill a
1414 work list with its initial contents. For example, if you wanted to
1415 initialize a work list to contain all instructions in a <tt>Function</tt>
1416 F, all you would need to do is something like:</p>
1418 <div class="doc_code">
1420 std::set<Instruction*> worklist;
1421 worklist.insert(inst_begin(F), inst_end(F));
1425 <p>The STL set <tt>worklist</tt> would now contain all instructions in the
1426 <tt>Function</tt> pointed to by F.</p>
1430 <!-- _______________________________________________________________________ -->
1431 <div class="doc_subsubsection">
1432 <a name="iterate_convert">Turning an iterator into a class pointer (and
1436 <div class="doc_text">
1438 <p>Sometimes, it'll be useful to grab a reference (or pointer) to a class
1439 instance when all you've got at hand is an iterator. Well, extracting
1440 a reference or a pointer from an iterator is very straight-forward.
1441 Assuming that <tt>i</tt> is a <tt>BasicBlock::iterator</tt> and <tt>j</tt>
1442 is a <tt>BasicBlock::const_iterator</tt>:</p>
1444 <div class="doc_code">
1446 Instruction& inst = *i; // <i>Grab reference to instruction reference</i>
1447 Instruction* pinst = &*i; // <i>Grab pointer to instruction reference</i>
1448 const Instruction& inst = *j;
1452 <p>However, the iterators you'll be working with in the LLVM framework are
1453 special: they will automatically convert to a ptr-to-instance type whenever they
1454 need to. Instead of dereferencing the iterator and then taking the address of
1455 the result, you can simply assign the iterator to the proper pointer type and
1456 you get the dereference and address-of operation as a result of the assignment
1457 (behind the scenes, this is a result of overloading casting mechanisms). Thus
1458 the last line of the last example,</p>
1460 <div class="doc_code">
1462 Instruction* pinst = &*i;
1466 <p>is semantically equivalent to</p>
1468 <div class="doc_code">
1470 Instruction* pinst = i;
1474 <p>It's also possible to turn a class pointer into the corresponding iterator,
1475 and this is a constant time operation (very efficient). The following code
1476 snippet illustrates use of the conversion constructors provided by LLVM
1477 iterators. By using these, you can explicitly grab the iterator of something
1478 without actually obtaining it via iteration over some structure:</p>
1480 <div class="doc_code">
1482 void printNextInstruction(Instruction* inst) {
1483 BasicBlock::iterator it(inst);
1484 ++it; // <i>After this line, it refers to the instruction after *inst</i>
1485 if (it != inst->getParent()->end()) llvm::cerr << *it << "\n";
1492 <!--_______________________________________________________________________-->
1493 <div class="doc_subsubsection">
1494 <a name="iterate_complex">Finding call sites: a slightly more complex
1498 <div class="doc_text">
1500 <p>Say that you're writing a FunctionPass and would like to count all the
1501 locations in the entire module (that is, across every <tt>Function</tt>) where a
1502 certain function (i.e., some <tt>Function</tt>*) is already in scope. As you'll
1503 learn later, you may want to use an <tt>InstVisitor</tt> to accomplish this in a
1504 much more straight-forward manner, but this example will allow us to explore how
1505 you'd do it if you didn't have <tt>InstVisitor</tt> around. In pseudo-code, this
1506 is what we want to do:</p>
1508 <div class="doc_code">
1510 initialize callCounter to zero
1511 for each Function f in the Module
1512 for each BasicBlock b in f
1513 for each Instruction i in b
1514 if (i is a CallInst and calls the given function)
1515 increment callCounter
1519 <p>And the actual code is (remember, because we're writing a
1520 <tt>FunctionPass</tt>, our <tt>FunctionPass</tt>-derived class simply has to
1521 override the <tt>runOnFunction</tt> method):</p>
1523 <div class="doc_code">
1525 Function* targetFunc = ...;
1527 class OurFunctionPass : public FunctionPass {
1529 OurFunctionPass(): callCounter(0) { }
1531 virtual runOnFunction(Function& F) {
1532 for (Function::iterator b = F.begin(), be = F.end(); b != be; ++b) {
1533 for (BasicBlock::iterator i = b->begin(); ie = b->end(); i != ie; ++i) {
1534 if (<a href="#CallInst">CallInst</a>* callInst = <a href="#isa">dyn_cast</a><<a
1535 href="#CallInst">CallInst</a>>(&*i)) {
1536 // <i>We know we've encountered a call instruction, so we</i>
1537 // <i>need to determine if it's a call to the</i>
1538 // <i>function pointed to by m_func or not</i>
1540 if (callInst->getCalledFunction() == targetFunc)
1548 unsigned callCounter;
1555 <!--_______________________________________________________________________-->
1556 <div class="doc_subsubsection">
1557 <a name="calls_and_invokes">Treating calls and invokes the same way</a>
1560 <div class="doc_text">
1562 <p>You may have noticed that the previous example was a bit oversimplified in
1563 that it did not deal with call sites generated by 'invoke' instructions. In
1564 this, and in other situations, you may find that you want to treat
1565 <tt>CallInst</tt>s and <tt>InvokeInst</tt>s the same way, even though their
1566 most-specific common base class is <tt>Instruction</tt>, which includes lots of
1567 less closely-related things. For these cases, LLVM provides a handy wrapper
1569 href="http://llvm.org/doxygen/classllvm_1_1CallSite.html"><tt>CallSite</tt></a>.
1570 It is essentially a wrapper around an <tt>Instruction</tt> pointer, with some
1571 methods that provide functionality common to <tt>CallInst</tt>s and
1572 <tt>InvokeInst</tt>s.</p>
1574 <p>This class has "value semantics": it should be passed by value, not by
1575 reference and it should not be dynamically allocated or deallocated using
1576 <tt>operator new</tt> or <tt>operator delete</tt>. It is efficiently copyable,
1577 assignable and constructable, with costs equivalents to that of a bare pointer.
1578 If you look at its definition, it has only a single pointer member.</p>
1582 <!--_______________________________________________________________________-->
1583 <div class="doc_subsubsection">
1584 <a name="iterate_chains">Iterating over def-use & use-def chains</a>
1587 <div class="doc_text">
1589 <p>Frequently, we might have an instance of the <a
1590 href="/doxygen/classllvm_1_1Value.html">Value Class</a> and we want to
1591 determine which <tt>User</tt>s use the <tt>Value</tt>. The list of all
1592 <tt>User</tt>s of a particular <tt>Value</tt> is called a <i>def-use</i> chain.
1593 For example, let's say we have a <tt>Function*</tt> named <tt>F</tt> to a
1594 particular function <tt>foo</tt>. Finding all of the instructions that
1595 <i>use</i> <tt>foo</tt> is as simple as iterating over the <i>def-use</i> chain
1598 <div class="doc_code">
1602 for (Value::use_iterator i = F->use_begin(), e = F->use_end(); i != e; ++i)
1603 if (Instruction *Inst = dyn_cast<Instruction>(*i)) {
1604 llvm::cerr << "F is used in instruction:\n";
1605 llvm::cerr << *Inst << "\n";
1610 <p>Alternately, it's common to have an instance of the <a
1611 href="/doxygen/classllvm_1_1User.html">User Class</a> and need to know what
1612 <tt>Value</tt>s are used by it. The list of all <tt>Value</tt>s used by a
1613 <tt>User</tt> is known as a <i>use-def</i> chain. Instances of class
1614 <tt>Instruction</tt> are common <tt>User</tt>s, so we might want to iterate over
1615 all of the values that a particular instruction uses (that is, the operands of
1616 the particular <tt>Instruction</tt>):</p>
1618 <div class="doc_code">
1620 Instruction* pi = ...;
1622 for (User::op_iterator i = pi->op_begin(), e = pi->op_end(); i != e; ++i) {
1630 def-use chains ("finding all users of"): Value::use_begin/use_end
1631 use-def chains ("finding all values used"): User::op_begin/op_end [op=operand]
1636 <!-- ======================================================================= -->
1637 <div class="doc_subsection">
1638 <a name="simplechanges">Making simple changes</a>
1641 <div class="doc_text">
1643 <p>There are some primitive transformation operations present in the LLVM
1644 infrastructure that are worth knowing about. When performing
1645 transformations, it's fairly common to manipulate the contents of basic
1646 blocks. This section describes some of the common methods for doing so
1647 and gives example code.</p>
1651 <!--_______________________________________________________________________-->
1652 <div class="doc_subsubsection">
1653 <a name="schanges_creating">Creating and inserting new
1654 <tt>Instruction</tt>s</a>
1657 <div class="doc_text">
1659 <p><i>Instantiating Instructions</i></p>
1661 <p>Creation of <tt>Instruction</tt>s is straight-forward: simply call the
1662 constructor for the kind of instruction to instantiate and provide the necessary
1663 parameters. For example, an <tt>AllocaInst</tt> only <i>requires</i> a
1664 (const-ptr-to) <tt>Type</tt>. Thus:</p>
1666 <div class="doc_code">
1668 AllocaInst* ai = new AllocaInst(Type::IntTy);
1672 <p>will create an <tt>AllocaInst</tt> instance that represents the allocation of
1673 one integer in the current stack frame, at run time. Each <tt>Instruction</tt>
1674 subclass is likely to have varying default parameters which change the semantics
1675 of the instruction, so refer to the <a
1676 href="/doxygen/classllvm_1_1Instruction.html">doxygen documentation for the subclass of
1677 Instruction</a> that you're interested in instantiating.</p>
1679 <p><i>Naming values</i></p>
1681 <p>It is very useful to name the values of instructions when you're able to, as
1682 this facilitates the debugging of your transformations. If you end up looking
1683 at generated LLVM machine code, you definitely want to have logical names
1684 associated with the results of instructions! By supplying a value for the
1685 <tt>Name</tt> (default) parameter of the <tt>Instruction</tt> constructor, you
1686 associate a logical name with the result of the instruction's execution at
1687 run time. For example, say that I'm writing a transformation that dynamically
1688 allocates space for an integer on the stack, and that integer is going to be
1689 used as some kind of index by some other code. To accomplish this, I place an
1690 <tt>AllocaInst</tt> at the first point in the first <tt>BasicBlock</tt> of some
1691 <tt>Function</tt>, and I'm intending to use it within the same
1692 <tt>Function</tt>. I might do:</p>
1694 <div class="doc_code">
1696 AllocaInst* pa = new AllocaInst(Type::IntTy, 0, "indexLoc");
1700 <p>where <tt>indexLoc</tt> is now the logical name of the instruction's
1701 execution value, which is a pointer to an integer on the run time stack.</p>
1703 <p><i>Inserting instructions</i></p>
1705 <p>There are essentially two ways to insert an <tt>Instruction</tt>
1706 into an existing sequence of instructions that form a <tt>BasicBlock</tt>:</p>
1709 <li>Insertion into an explicit instruction list
1711 <p>Given a <tt>BasicBlock* pb</tt>, an <tt>Instruction* pi</tt> within that
1712 <tt>BasicBlock</tt>, and a newly-created instruction we wish to insert
1713 before <tt>*pi</tt>, we do the following: </p>
1715 <div class="doc_code">
1717 BasicBlock *pb = ...;
1718 Instruction *pi = ...;
1719 Instruction *newInst = new Instruction(...);
1721 pb->getInstList().insert(pi, newInst); // <i>Inserts newInst before pi in pb</i>
1725 <p>Appending to the end of a <tt>BasicBlock</tt> is so common that
1726 the <tt>Instruction</tt> class and <tt>Instruction</tt>-derived
1727 classes provide constructors which take a pointer to a
1728 <tt>BasicBlock</tt> to be appended to. For example code that
1731 <div class="doc_code">
1733 BasicBlock *pb = ...;
1734 Instruction *newInst = new Instruction(...);
1736 pb->getInstList().push_back(newInst); // <i>Appends newInst to pb</i>
1742 <div class="doc_code">
1744 BasicBlock *pb = ...;
1745 Instruction *newInst = new Instruction(..., pb);
1749 <p>which is much cleaner, especially if you are creating
1750 long instruction streams.</p></li>
1752 <li>Insertion into an implicit instruction list
1754 <p><tt>Instruction</tt> instances that are already in <tt>BasicBlock</tt>s
1755 are implicitly associated with an existing instruction list: the instruction
1756 list of the enclosing basic block. Thus, we could have accomplished the same
1757 thing as the above code without being given a <tt>BasicBlock</tt> by doing:
1760 <div class="doc_code">
1762 Instruction *pi = ...;
1763 Instruction *newInst = new Instruction(...);
1765 pi->getParent()->getInstList().insert(pi, newInst);
1769 <p>In fact, this sequence of steps occurs so frequently that the
1770 <tt>Instruction</tt> class and <tt>Instruction</tt>-derived classes provide
1771 constructors which take (as a default parameter) a pointer to an
1772 <tt>Instruction</tt> which the newly-created <tt>Instruction</tt> should
1773 precede. That is, <tt>Instruction</tt> constructors are capable of
1774 inserting the newly-created instance into the <tt>BasicBlock</tt> of a
1775 provided instruction, immediately before that instruction. Using an
1776 <tt>Instruction</tt> constructor with a <tt>insertBefore</tt> (default)
1777 parameter, the above code becomes:</p>
1779 <div class="doc_code">
1781 Instruction* pi = ...;
1782 Instruction* newInst = new Instruction(..., pi);
1786 <p>which is much cleaner, especially if you're creating a lot of
1787 instructions and adding them to <tt>BasicBlock</tt>s.</p></li>
1792 <!--_______________________________________________________________________-->
1793 <div class="doc_subsubsection">
1794 <a name="schanges_deleting">Deleting <tt>Instruction</tt>s</a>
1797 <div class="doc_text">
1799 <p>Deleting an instruction from an existing sequence of instructions that form a
1800 <a href="#BasicBlock"><tt>BasicBlock</tt></a> is very straight-forward. First,
1801 you must have a pointer to the instruction that you wish to delete. Second, you
1802 need to obtain the pointer to that instruction's basic block. You use the
1803 pointer to the basic block to get its list of instructions and then use the
1804 erase function to remove your instruction. For example:</p>
1806 <div class="doc_code">
1808 <a href="#Instruction">Instruction</a> *I = .. ;
1809 <a href="#BasicBlock">BasicBlock</a> *BB = I->getParent();
1811 BB->getInstList().erase(I);
1817 <!--_______________________________________________________________________-->
1818 <div class="doc_subsubsection">
1819 <a name="schanges_replacing">Replacing an <tt>Instruction</tt> with another
1823 <div class="doc_text">
1825 <p><i>Replacing individual instructions</i></p>
1827 <p>Including "<a href="/doxygen/BasicBlockUtils_8h-source.html">llvm/Transforms/Utils/BasicBlockUtils.h</a>"
1828 permits use of two very useful replace functions: <tt>ReplaceInstWithValue</tt>
1829 and <tt>ReplaceInstWithInst</tt>.</p>
1831 <h4><a name="schanges_deleting">Deleting <tt>Instruction</tt>s</a></h4>
1834 <li><tt>ReplaceInstWithValue</tt>
1836 <p>This function replaces all uses (within a basic block) of a given
1837 instruction with a value, and then removes the original instruction. The
1838 following example illustrates the replacement of the result of a particular
1839 <tt>AllocaInst</tt> that allocates memory for a single integer with a null
1840 pointer to an integer.</p>
1842 <div class="doc_code">
1844 AllocaInst* instToReplace = ...;
1845 BasicBlock::iterator ii(instToReplace);
1847 ReplaceInstWithValue(instToReplace->getParent()->getInstList(), ii,
1848 Constant::getNullValue(PointerType::get(Type::IntTy)));
1851 <li><tt>ReplaceInstWithInst</tt>
1853 <p>This function replaces a particular instruction with another
1854 instruction. The following example illustrates the replacement of one
1855 <tt>AllocaInst</tt> with another.</p>
1857 <div class="doc_code">
1859 AllocaInst* instToReplace = ...;
1860 BasicBlock::iterator ii(instToReplace);
1862 ReplaceInstWithInst(instToReplace->getParent()->getInstList(), ii,
1863 new AllocaInst(Type::IntTy, 0, "ptrToReplacedInt"));
1867 <p><i>Replacing multiple uses of <tt>User</tt>s and <tt>Value</tt>s</i></p>
1869 <p>You can use <tt>Value::replaceAllUsesWith</tt> and
1870 <tt>User::replaceUsesOfWith</tt> to change more than one use at a time. See the
1871 doxygen documentation for the <a href="/doxygen/classllvm_1_1Value.html">Value Class</a>
1872 and <a href="/doxygen/classllvm_1_1User.html">User Class</a>, respectively, for more
1875 <!-- Value::replaceAllUsesWith User::replaceUsesOfWith Point out:
1876 include/llvm/Transforms/Utils/ especially BasicBlockUtils.h with:
1877 ReplaceInstWithValue, ReplaceInstWithInst -->
1881 <!-- *********************************************************************** -->
1882 <div class="doc_section">
1883 <a name="advanced">Advanced Topics</a>
1885 <!-- *********************************************************************** -->
1887 <div class="doc_text">
1889 This section describes some of the advanced or obscure API's that most clients
1890 do not need to be aware of. These API's tend manage the inner workings of the
1891 LLVM system, and only need to be accessed in unusual circumstances.
1895 <!-- ======================================================================= -->
1896 <div class="doc_subsection">
1897 <a name="TypeResolve">LLVM Type Resolution</a>
1900 <div class="doc_text">
1903 The LLVM type system has a very simple goal: allow clients to compare types for
1904 structural equality with a simple pointer comparison (aka a shallow compare).
1905 This goal makes clients much simpler and faster, and is used throughout the LLVM
1910 Unfortunately achieving this goal is not a simple matter. In particular,
1911 recursive types and late resolution of opaque types makes the situation very
1912 difficult to handle. Fortunately, for the most part, our implementation makes
1913 most clients able to be completely unaware of the nasty internal details. The
1914 primary case where clients are exposed to the inner workings of it are when
1915 building a recursive type. In addition to this case, the LLVM bytecode reader,
1916 assembly parser, and linker also have to be aware of the inner workings of this
1921 For our purposes below, we need three concepts. First, an "Opaque Type" is
1922 exactly as defined in the <a href="LangRef.html#t_opaque">language
1923 reference</a>. Second an "Abstract Type" is any type which includes an
1924 opaque type as part of its type graph (for example "<tt>{ opaque, i32 }</tt>").
1925 Third, a concrete type is a type that is not an abstract type (e.g. "<tt>{ i32,
1931 <!-- ______________________________________________________________________ -->
1932 <div class="doc_subsubsection">
1933 <a name="BuildRecType">Basic Recursive Type Construction</a>
1936 <div class="doc_text">
1939 Because the most common question is "how do I build a recursive type with LLVM",
1940 we answer it now and explain it as we go. Here we include enough to cause this
1941 to be emitted to an output .ll file:
1944 <div class="doc_code">
1946 %mylist = type { %mylist*, i32 }
1951 To build this, use the following LLVM APIs:
1954 <div class="doc_code">
1956 // <i>Create the initial outer struct</i>
1957 <a href="#PATypeHolder">PATypeHolder</a> StructTy = OpaqueType::get();
1958 std::vector<const Type*> Elts;
1959 Elts.push_back(PointerType::get(StructTy));
1960 Elts.push_back(Type::IntTy);
1961 StructType *NewSTy = StructType::get(Elts);
1963 // <i>At this point, NewSTy = "{ opaque*, i32 }". Tell VMCore that</i>
1964 // <i>the struct and the opaque type are actually the same.</i>
1965 cast<OpaqueType>(StructTy.get())-><a href="#refineAbstractTypeTo">refineAbstractTypeTo</a>(NewSTy);
1967 // <i>NewSTy is potentially invalidated, but StructTy (a <a href="#PATypeHolder">PATypeHolder</a>) is</i>
1968 // <i>kept up-to-date</i>
1969 NewSTy = cast<StructType>(StructTy.get());
1971 // <i>Add a name for the type to the module symbol table (optional)</i>
1972 MyModule->addTypeName("mylist", NewSTy);
1977 This code shows the basic approach used to build recursive types: build a
1978 non-recursive type using 'opaque', then use type unification to close the cycle.
1979 The type unification step is performed by the <tt><a
1980 href="#refineAbstractTypeTo">refineAbstractTypeTo</a></tt> method, which is
1981 described next. After that, we describe the <a
1982 href="#PATypeHolder">PATypeHolder class</a>.
1987 <!-- ______________________________________________________________________ -->
1988 <div class="doc_subsubsection">
1989 <a name="refineAbstractTypeTo">The <tt>refineAbstractTypeTo</tt> method</a>
1992 <div class="doc_text">
1994 The <tt>refineAbstractTypeTo</tt> method starts the type unification process.
1995 While this method is actually a member of the DerivedType class, it is most
1996 often used on OpaqueType instances. Type unification is actually a recursive
1997 process. After unification, types can become structurally isomorphic to
1998 existing types, and all duplicates are deleted (to preserve pointer equality).
2002 In the example above, the OpaqueType object is definitely deleted.
2003 Additionally, if there is an "{ \2*, i32}" type already created in the system,
2004 the pointer and struct type created are <b>also</b> deleted. Obviously whenever
2005 a type is deleted, any "Type*" pointers in the program are invalidated. As
2006 such, it is safest to avoid having <i>any</i> "Type*" pointers to abstract types
2007 live across a call to <tt>refineAbstractTypeTo</tt> (note that non-abstract
2008 types can never move or be deleted). To deal with this, the <a
2009 href="#PATypeHolder">PATypeHolder</a> class is used to maintain a stable
2010 reference to a possibly refined type, and the <a
2011 href="#AbstractTypeUser">AbstractTypeUser</a> class is used to update more
2012 complex datastructures.
2017 <!-- ______________________________________________________________________ -->
2018 <div class="doc_subsubsection">
2019 <a name="PATypeHolder">The PATypeHolder Class</a>
2022 <div class="doc_text">
2024 PATypeHolder is a form of a "smart pointer" for Type objects. When VMCore
2025 happily goes about nuking types that become isomorphic to existing types, it
2026 automatically updates all PATypeHolder objects to point to the new type. In the
2027 example above, this allows the code to maintain a pointer to the resultant
2028 resolved recursive type, even though the Type*'s are potentially invalidated.
2032 PATypeHolder is an extremely light-weight object that uses a lazy union-find
2033 implementation to update pointers. For example the pointer from a Value to its
2034 Type is maintained by PATypeHolder objects.
2039 <!-- ______________________________________________________________________ -->
2040 <div class="doc_subsubsection">
2041 <a name="AbstractTypeUser">The AbstractTypeUser Class</a>
2044 <div class="doc_text">
2047 Some data structures need more to perform more complex updates when types get
2048 resolved. The <a href="#SymbolTable">SymbolTable</a> class, for example, needs
2049 move and potentially merge type planes in its representation when a pointer
2053 To support this, a class can derive from the AbstractTypeUser class. This class
2054 allows it to get callbacks when certain types are resolved. To register to get
2055 callbacks for a particular type, the DerivedType::{add/remove}AbstractTypeUser
2056 methods can be called on a type. Note that these methods only work for <i>
2057 abstract</i> types. Concrete types (those that do not include any opaque
2058 objects) can never be refined.
2063 <!-- ======================================================================= -->
2064 <div class="doc_subsection">
2065 <a name="SymbolTable">The <tt>SymbolTable</tt> class</a>
2068 <div class="doc_text">
2069 <p>This class provides a symbol table that the <a
2070 href="#Function"><tt>Function</tt></a> and <a href="#Module">
2071 <tt>Module</tt></a> classes use for naming definitions. The symbol table can
2072 provide a name for any <a href="#Value"><tt>Value</tt></a>.
2073 <tt>SymbolTable</tt> is an abstract data type. It hides the data it contains
2074 and provides access to it through a controlled interface.</p>
2076 <p>Note that the <tt>SymbolTable</tt> class should not be directly accessed
2077 by most clients. It should only be used when iteration over the symbol table
2078 names themselves are required, which is very special purpose. Note that not
2080 <a href="#Value">Value</a>s have names, and those without names (i.e. they have
2081 an empty name) do not exist in the symbol table.
2084 <p>To use the <tt>SymbolTable</tt> well, you need to understand the
2085 structure of the information it holds. The class contains two
2086 <tt>std::map</tt> objects. The first, <tt>pmap</tt>, is a map of
2087 <tt>Type*</tt> to maps of name (<tt>std::string</tt>) to <tt>Value*</tt>.
2088 Thus, Values are stored in two-dimensions and accessed by <tt>Type</tt> and
2091 <p>The interface of this class provides three basic types of operations:
2093 <li><em>Accessors</em>. Accessors provide read-only access to information
2094 such as finding a value for a name with the
2095 <a href="#SymbolTable_lookup">lookup</a> method.</li>
2096 <li><em>Mutators</em>. Mutators allow the user to add information to the
2097 <tt>SymbolTable</tt> with methods like
2098 <a href="#SymbolTable_insert"><tt>insert</tt></a>.</li>
2099 <li><em>Iterators</em>. Iterators allow the user to traverse the content
2100 of the symbol table in well defined ways, such as the method
2101 <a href="#SymbolTable_plane_begin"><tt>plane_begin</tt></a>.</li>
2106 <dt><tt>Value* lookup(const Type* Ty, const std::string& name) const</tt>:
2108 <dd>The <tt>lookup</tt> method searches the type plane given by the
2109 <tt>Ty</tt> parameter for a <tt>Value</tt> with the provided <tt>name</tt>.
2110 If a suitable <tt>Value</tt> is not found, null is returned.</dd>
2112 <dt><tt>bool isEmpty() const</tt>:</dt>
2113 <dd>This function returns true if both the value and types maps are
2119 <dt><tt>void insert(Value *Val)</tt>:</dt>
2120 <dd>This method adds the provided value to the symbol table. The Value must
2121 have both a name and a type which are extracted and used to place the value
2122 in the correct type plane under the value's name.</dd>
2124 <dt><tt>void remove(Value* Val)</tt>:</dt>
2125 <dd> This method removes a named value from the symbol table. The
2126 type and name of the Value are extracted from \p N and used to
2127 lookup the Value in the correct type plane. If the Value is
2128 not in the symbol table, this method silently ignores the
2134 <p>The following functions describe three types of iterators you can obtain
2135 the beginning or end of the sequence for both const and non-const. It is
2136 important to keep track of the different kinds of iterators. There are
2137 three idioms worth pointing out:</p>
2140 <tr><th>Units</th><th>Iterator</th><th>Idiom</th></tr>
2142 <td align="left">Planes Of name/Value maps</td><td>PI</td>
2143 <td align="left"><pre><tt>
2144 for (SymbolTable::plane_const_iterator PI = ST.plane_begin(),
2145 PE = ST.plane_end(); PI != PE; ++PI ) {
2146 PI->first // <i>This is the Type* of the plane</i>
2147 PI->second // <i>This is the SymbolTable::ValueMap of name/Value pairs</i>
2152 <td align="left">name/Value pairs in a plane</td><td>VI</td>
2153 <td align="left"><pre><tt>
2154 for (SymbolTable::value_const_iterator VI = ST.value_begin(SomeType),
2155 VE = ST.value_end(SomeType); VI != VE; ++VI ) {
2156 VI->first // <i>This is the name of the Value</i>
2157 VI->second // <i>This is the Value* value associated with the name</i>
2163 <p>Using the recommended iterator names and idioms will help you avoid
2164 making mistakes. Of particular note, make sure that whenever you use
2165 value_begin(SomeType) that you always compare the resulting iterator
2166 with value_end(SomeType) not value_end(SomeOtherType) or else you
2167 will loop infinitely.</p>
2171 <dt><tt>plane_iterator plane_begin()</tt>:</dt>
2172 <dd>Get an iterator that starts at the beginning of the type planes.
2173 The iterator will iterate over the Type/ValueMap pairs in the
2176 <dt><tt>plane_const_iterator plane_begin() const</tt>:</dt>
2177 <dd>Get a const_iterator that starts at the beginning of the type
2178 planes. The iterator will iterate over the Type/ValueMap pairs
2179 in the type planes. </dd>
2181 <dt><tt>plane_iterator plane_end()</tt>:</dt>
2182 <dd>Get an iterator at the end of the type planes. This serves as
2183 the marker for end of iteration over the type planes.</dd>
2185 <dt><tt>plane_const_iterator plane_end() const</tt>:</dt>
2186 <dd>Get a const_iterator at the end of the type planes. This serves as
2187 the marker for end of iteration over the type planes.</dd>
2189 <dt><tt>value_iterator value_begin(const Type *Typ)</tt>:</dt>
2190 <dd>Get an iterator that starts at the beginning of a type plane.
2191 The iterator will iterate over the name/value pairs in the type plane.
2192 Note: The type plane must already exist before using this.</dd>
2194 <dt><tt>value_const_iterator value_begin(const Type *Typ) const</tt>:</dt>
2195 <dd>Get a const_iterator that starts at the beginning of a type plane.
2196 The iterator will iterate over the name/value pairs in the type plane.
2197 Note: The type plane must already exist before using this.</dd>
2199 <dt><tt>value_iterator value_end(const Type *Typ)</tt>:</dt>
2200 <dd>Get an iterator to the end of a type plane. This serves as the marker
2201 for end of iteration of the type plane.
2202 Note: The type plane must already exist before using this.</dd>
2204 <dt><tt>value_const_iterator value_end(const Type *Typ) const</tt>:</dt>
2205 <dd>Get a const_iterator to the end of a type plane. This serves as the
2206 marker for end of iteration of the type plane.
2207 Note: the type plane must already exist before using this.</dd>
2209 <dt><tt>plane_const_iterator find(const Type* Typ ) const</tt>:</dt>
2210 <dd>This method returns a plane_const_iterator for iteration over
2211 the type planes starting at a specific plane, given by \p Ty.</dd>
2213 <dt><tt>plane_iterator find( const Type* Typ </tt>:</dt>
2214 <dd>This method returns a plane_iterator for iteration over the
2215 type planes starting at a specific plane, given by \p Ty.</dd>
2222 <!-- *********************************************************************** -->
2223 <div class="doc_section">
2224 <a name="coreclasses">The Core LLVM Class Hierarchy Reference </a>
2226 <!-- *********************************************************************** -->
2228 <div class="doc_text">
2229 <p><tt>#include "<a href="/doxygen/Type_8h-source.html">llvm/Type.h</a>"</tt>
2230 <br>doxygen info: <a href="/doxygen/classllvm_1_1Type.html">Type Class</a></p>
2232 <p>The Core LLVM classes are the primary means of representing the program
2233 being inspected or transformed. The core LLVM classes are defined in
2234 header files in the <tt>include/llvm/</tt> directory, and implemented in
2235 the <tt>lib/VMCore</tt> directory.</p>
2239 <!-- ======================================================================= -->
2240 <div class="doc_subsection">
2241 <a name="Type">The <tt>Type</tt> class and Derived Types</a>
2244 <div class="doc_text">
2246 <p><tt>Type</tt> is a superclass of all type classes. Every <tt>Value</tt> has
2247 a <tt>Type</tt>. <tt>Type</tt> cannot be instantiated directly but only
2248 through its subclasses. Certain primitive types (<tt>VoidType</tt>,
2249 <tt>LabelType</tt>, <tt>FloatType</tt> and <tt>DoubleType</tt>) have hidden
2250 subclasses. They are hidden because they offer no useful functionality beyond
2251 what the <tt>Type</tt> class offers except to distinguish themselves from
2252 other subclasses of <tt>Type</tt>.</p>
2253 <p>All other types are subclasses of <tt>DerivedType</tt>. Types can be
2254 named, but this is not a requirement. There exists exactly
2255 one instance of a given shape at any one time. This allows type equality to
2256 be performed with address equality of the Type Instance. That is, given two
2257 <tt>Type*</tt> values, the types are identical if the pointers are identical.
2261 <!-- _______________________________________________________________________ -->
2262 <div class="doc_subsubsection">
2263 <a name="m_Value">Important Public Methods</a>
2266 <div class="doc_text">
2269 <li><tt>bool isInteger() const</tt>: Returns true for any integer type.</li>
2271 <li><tt>bool isFloatingPoint()</tt>: Return true if this is one of the two
2272 floating point types.</li>
2274 <li><tt>bool isAbstract()</tt>: Return true if the type is abstract (contains
2275 an OpaqueType anywhere in its definition).</li>
2277 <li><tt>bool isSized()</tt>: Return true if the type has known size. Things
2278 that don't have a size are abstract types, labels and void.</li>
2283 <!-- _______________________________________________________________________ -->
2284 <div class="doc_subsubsection">
2285 <a name="m_Value">Important Derived Types</a>
2287 <div class="doc_text">
2289 <dt><tt>IntegerType</tt></dt>
2290 <dd>Subclass of DerivedType that represents integer types of any bit width.
2291 Any bit width between <tt>IntegerType::MIN_INT_BITS</tt> (1) and
2292 <tt>IntegerType::MAX_INT_BITS</tt> (~8 million) can be represented.
2294 <li><tt>static const IntegerType* get(unsigned NumBits)</tt>: get an integer
2295 type of a specific bit width.</li>
2296 <li><tt>unsigned getBitWidth() const</tt>: Get the bit width of an integer
2300 <dt><tt>SequentialType</tt></dt>
2301 <dd>This is subclassed by ArrayType and PointerType
2303 <li><tt>const Type * getElementType() const</tt>: Returns the type of each
2304 of the elements in the sequential type. </li>
2307 <dt><tt>ArrayType</tt></dt>
2308 <dd>This is a subclass of SequentialType and defines the interface for array
2311 <li><tt>unsigned getNumElements() const</tt>: Returns the number of
2312 elements in the array. </li>
2315 <dt><tt>PointerType</tt></dt>
2316 <dd>Subclass of SequentialType for pointer types.</dd>
2317 <dt><tt>PackedType</tt></dt>
2318 <dd>Subclass of SequentialType for packed (vector) types. A
2319 packed type is similar to an ArrayType but is distinguished because it is
2320 a first class type wherease ArrayType is not. Packed types are used for
2321 vector operations and are usually small vectors of of an integer or floating
2323 <dt><tt>StructType</tt></dt>
2324 <dd>Subclass of DerivedTypes for struct types.</dd>
2325 <dt><tt>FunctionType</tt></dt>
2326 <dd>Subclass of DerivedTypes for function types.
2328 <li><tt>bool isVarArg() const</tt>: Returns true if its a vararg
2330 <li><tt> const Type * getReturnType() const</tt>: Returns the
2331 return type of the function.</li>
2332 <li><tt>const Type * getParamType (unsigned i)</tt>: Returns
2333 the type of the ith parameter.</li>
2334 <li><tt> const unsigned getNumParams() const</tt>: Returns the
2335 number of formal parameters.</li>
2338 <dt><tt>OpaqueType</tt></dt>
2339 <dd>Sublcass of DerivedType for abstract types. This class
2340 defines no content and is used as a placeholder for some other type. Note
2341 that OpaqueType is used (temporarily) during type resolution for forward
2342 references of types. Once the referenced type is resolved, the OpaqueType
2343 is replaced with the actual type. OpaqueType can also be used for data
2344 abstraction. At link time opaque types can be resolved to actual types
2345 of the same name.</dd>
2351 <!-- ======================================================================= -->
2352 <div class="doc_subsection">
2353 <a name="Module">The <tt>Module</tt> class</a>
2356 <div class="doc_text">
2359 href="/doxygen/Module_8h-source.html">llvm/Module.h</a>"</tt><br> doxygen info:
2360 <a href="/doxygen/classllvm_1_1Module.html">Module Class</a></p>
2362 <p>The <tt>Module</tt> class represents the top level structure present in LLVM
2363 programs. An LLVM module is effectively either a translation unit of the
2364 original program or a combination of several translation units merged by the
2365 linker. The <tt>Module</tt> class keeps track of a list of <a
2366 href="#Function"><tt>Function</tt></a>s, a list of <a
2367 href="#GlobalVariable"><tt>GlobalVariable</tt></a>s, and a <a
2368 href="#SymbolTable"><tt>SymbolTable</tt></a>. Additionally, it contains a few
2369 helpful member functions that try to make common operations easy.</p>
2373 <!-- _______________________________________________________________________ -->
2374 <div class="doc_subsubsection">
2375 <a name="m_Module">Important Public Members of the <tt>Module</tt> class</a>
2378 <div class="doc_text">
2381 <li><tt>Module::Module(std::string name = "")</tt></li>
2384 <p>Constructing a <a href="#Module">Module</a> is easy. You can optionally
2385 provide a name for it (probably based on the name of the translation unit).</p>
2388 <li><tt>Module::iterator</tt> - Typedef for function list iterator<br>
2389 <tt>Module::const_iterator</tt> - Typedef for const_iterator.<br>
2391 <tt>begin()</tt>, <tt>end()</tt>
2392 <tt>size()</tt>, <tt>empty()</tt>
2394 <p>These are forwarding methods that make it easy to access the contents of
2395 a <tt>Module</tt> object's <a href="#Function"><tt>Function</tt></a>
2398 <li><tt>Module::FunctionListType &getFunctionList()</tt>
2400 <p> Returns the list of <a href="#Function"><tt>Function</tt></a>s. This is
2401 necessary to use when you need to update the list or perform a complex
2402 action that doesn't have a forwarding method.</p>
2404 <p><!-- Global Variable --></p></li>
2410 <li><tt>Module::global_iterator</tt> - Typedef for global variable list iterator<br>
2412 <tt>Module::const_global_iterator</tt> - Typedef for const_iterator.<br>
2414 <tt>global_begin()</tt>, <tt>global_end()</tt>
2415 <tt>global_size()</tt>, <tt>global_empty()</tt>
2417 <p> These are forwarding methods that make it easy to access the contents of
2418 a <tt>Module</tt> object's <a
2419 href="#GlobalVariable"><tt>GlobalVariable</tt></a> list.</p></li>
2421 <li><tt>Module::GlobalListType &getGlobalList()</tt>
2423 <p>Returns the list of <a
2424 href="#GlobalVariable"><tt>GlobalVariable</tt></a>s. This is necessary to
2425 use when you need to update the list or perform a complex action that
2426 doesn't have a forwarding method.</p>
2428 <p><!-- Symbol table stuff --> </p></li>
2434 <li><tt><a href="#SymbolTable">SymbolTable</a> *getSymbolTable()</tt>
2436 <p>Return a reference to the <a href="#SymbolTable"><tt>SymbolTable</tt></a>
2437 for this <tt>Module</tt>.</p>
2439 <p><!-- Convenience methods --></p></li>
2445 <li><tt><a href="#Function">Function</a> *getFunction(const std::string
2446 &Name, const <a href="#FunctionType">FunctionType</a> *Ty)</tt>
2448 <p>Look up the specified function in the <tt>Module</tt> <a
2449 href="#SymbolTable"><tt>SymbolTable</tt></a>. If it does not exist, return
2450 <tt>null</tt>.</p></li>
2452 <li><tt><a href="#Function">Function</a> *getOrInsertFunction(const
2453 std::string &Name, const <a href="#FunctionType">FunctionType</a> *T)</tt>
2455 <p>Look up the specified function in the <tt>Module</tt> <a
2456 href="#SymbolTable"><tt>SymbolTable</tt></a>. If it does not exist, add an
2457 external declaration for the function and return it.</p></li>
2459 <li><tt>std::string getTypeName(const <a href="#Type">Type</a> *Ty)</tt>
2461 <p>If there is at least one entry in the <a
2462 href="#SymbolTable"><tt>SymbolTable</tt></a> for the specified <a
2463 href="#Type"><tt>Type</tt></a>, return it. Otherwise return the empty
2466 <li><tt>bool addTypeName(const std::string &Name, const <a
2467 href="#Type">Type</a> *Ty)</tt>
2469 <p>Insert an entry in the <a href="#SymbolTable"><tt>SymbolTable</tt></a>
2470 mapping <tt>Name</tt> to <tt>Ty</tt>. If there is already an entry for this
2471 name, true is returned and the <a
2472 href="#SymbolTable"><tt>SymbolTable</tt></a> is not modified.</p></li>
2478 <!-- ======================================================================= -->
2479 <div class="doc_subsection">
2480 <a name="Value">The <tt>Value</tt> class</a>
2483 <div class="doc_text">
2485 <p><tt>#include "<a href="/doxygen/Value_8h-source.html">llvm/Value.h</a>"</tt>
2487 doxygen info: <a href="/doxygen/classllvm_1_1Value.html">Value Class</a></p>
2489 <p>The <tt>Value</tt> class is the most important class in the LLVM Source
2490 base. It represents a typed value that may be used (among other things) as an
2491 operand to an instruction. There are many different types of <tt>Value</tt>s,
2492 such as <a href="#Constant"><tt>Constant</tt></a>s,<a
2493 href="#Argument"><tt>Argument</tt></a>s. Even <a
2494 href="#Instruction"><tt>Instruction</tt></a>s and <a
2495 href="#Function"><tt>Function</tt></a>s are <tt>Value</tt>s.</p>
2497 <p>A particular <tt>Value</tt> may be used many times in the LLVM representation
2498 for a program. For example, an incoming argument to a function (represented
2499 with an instance of the <a href="#Argument">Argument</a> class) is "used" by
2500 every instruction in the function that references the argument. To keep track
2501 of this relationship, the <tt>Value</tt> class keeps a list of all of the <a
2502 href="#User"><tt>User</tt></a>s that is using it (the <a
2503 href="#User"><tt>User</tt></a> class is a base class for all nodes in the LLVM
2504 graph that can refer to <tt>Value</tt>s). This use list is how LLVM represents
2505 def-use information in the program, and is accessible through the <tt>use_</tt>*
2506 methods, shown below.</p>
2508 <p>Because LLVM is a typed representation, every LLVM <tt>Value</tt> is typed,
2509 and this <a href="#Type">Type</a> is available through the <tt>getType()</tt>
2510 method. In addition, all LLVM values can be named. The "name" of the
2511 <tt>Value</tt> is a symbolic string printed in the LLVM code:</p>
2513 <div class="doc_code">
2515 %<b>foo</b> = add i32 1, 2
2519 <p><a name="#nameWarning">The name of this instruction is "foo".</a> <b>NOTE</b>
2520 that the name of any value may be missing (an empty string), so names should
2521 <b>ONLY</b> be used for debugging (making the source code easier to read,
2522 debugging printouts), they should not be used to keep track of values or map
2523 between them. For this purpose, use a <tt>std::map</tt> of pointers to the
2524 <tt>Value</tt> itself instead.</p>
2526 <p>One important aspect of LLVM is that there is no distinction between an SSA
2527 variable and the operation that produces it. Because of this, any reference to
2528 the value produced by an instruction (or the value available as an incoming
2529 argument, for example) is represented as a direct pointer to the instance of
2531 represents this value. Although this may take some getting used to, it
2532 simplifies the representation and makes it easier to manipulate.</p>
2536 <!-- _______________________________________________________________________ -->
2537 <div class="doc_subsubsection">
2538 <a name="m_Value">Important Public Members of the <tt>Value</tt> class</a>
2541 <div class="doc_text">
2544 <li><tt>Value::use_iterator</tt> - Typedef for iterator over the
2546 <tt>Value::use_const_iterator</tt> - Typedef for const_iterator over
2548 <tt>unsigned use_size()</tt> - Returns the number of users of the
2550 <tt>bool use_empty()</tt> - Returns true if there are no users.<br>
2551 <tt>use_iterator use_begin()</tt> - Get an iterator to the start of
2553 <tt>use_iterator use_end()</tt> - Get an iterator to the end of the
2555 <tt><a href="#User">User</a> *use_back()</tt> - Returns the last
2556 element in the list.
2557 <p> These methods are the interface to access the def-use
2558 information in LLVM. As with all other iterators in LLVM, the naming
2559 conventions follow the conventions defined by the <a href="#stl">STL</a>.</p>
2561 <li><tt><a href="#Type">Type</a> *getType() const</tt>
2562 <p>This method returns the Type of the Value.</p>
2564 <li><tt>bool hasName() const</tt><br>
2565 <tt>std::string getName() const</tt><br>
2566 <tt>void setName(const std::string &Name)</tt>
2567 <p> This family of methods is used to access and assign a name to a <tt>Value</tt>,
2568 be aware of the <a href="#nameWarning">precaution above</a>.</p>
2570 <li><tt>void replaceAllUsesWith(Value *V)</tt>
2572 <p>This method traverses the use list of a <tt>Value</tt> changing all <a
2573 href="#User"><tt>User</tt>s</a> of the current value to refer to
2574 "<tt>V</tt>" instead. For example, if you detect that an instruction always
2575 produces a constant value (for example through constant folding), you can
2576 replace all uses of the instruction with the constant like this:</p>
2578 <div class="doc_code">
2580 Inst->replaceAllUsesWith(ConstVal);
2588 <!-- ======================================================================= -->
2589 <div class="doc_subsection">
2590 <a name="User">The <tt>User</tt> class</a>
2593 <div class="doc_text">
2596 <tt>#include "<a href="/doxygen/User_8h-source.html">llvm/User.h</a>"</tt><br>
2597 doxygen info: <a href="/doxygen/classllvm_1_1User.html">User Class</a><br>
2598 Superclass: <a href="#Value"><tt>Value</tt></a></p>
2600 <p>The <tt>User</tt> class is the common base class of all LLVM nodes that may
2601 refer to <a href="#Value"><tt>Value</tt></a>s. It exposes a list of "Operands"
2602 that are all of the <a href="#Value"><tt>Value</tt></a>s that the User is
2603 referring to. The <tt>User</tt> class itself is a subclass of
2606 <p>The operands of a <tt>User</tt> point directly to the LLVM <a
2607 href="#Value"><tt>Value</tt></a> that it refers to. Because LLVM uses Static
2608 Single Assignment (SSA) form, there can only be one definition referred to,
2609 allowing this direct connection. This connection provides the use-def
2610 information in LLVM.</p>
2614 <!-- _______________________________________________________________________ -->
2615 <div class="doc_subsubsection">
2616 <a name="m_User">Important Public Members of the <tt>User</tt> class</a>
2619 <div class="doc_text">
2621 <p>The <tt>User</tt> class exposes the operand list in two ways: through
2622 an index access interface and through an iterator based interface.</p>
2625 <li><tt>Value *getOperand(unsigned i)</tt><br>
2626 <tt>unsigned getNumOperands()</tt>
2627 <p> These two methods expose the operands of the <tt>User</tt> in a
2628 convenient form for direct access.</p></li>
2630 <li><tt>User::op_iterator</tt> - Typedef for iterator over the operand
2632 <tt>op_iterator op_begin()</tt> - Get an iterator to the start of
2633 the operand list.<br>
2634 <tt>op_iterator op_end()</tt> - Get an iterator to the end of the
2636 <p> Together, these methods make up the iterator based interface to
2637 the operands of a <tt>User</tt>.</p></li>
2642 <!-- ======================================================================= -->
2643 <div class="doc_subsection">
2644 <a name="Instruction">The <tt>Instruction</tt> class</a>
2647 <div class="doc_text">
2649 <p><tt>#include "</tt><tt><a
2650 href="/doxygen/Instruction_8h-source.html">llvm/Instruction.h</a>"</tt><br>
2651 doxygen info: <a href="/doxygen/classllvm_1_1Instruction.html">Instruction Class</a><br>
2652 Superclasses: <a href="#User"><tt>User</tt></a>, <a
2653 href="#Value"><tt>Value</tt></a></p>
2655 <p>The <tt>Instruction</tt> class is the common base class for all LLVM
2656 instructions. It provides only a few methods, but is a very commonly used
2657 class. The primary data tracked by the <tt>Instruction</tt> class itself is the
2658 opcode (instruction type) and the parent <a
2659 href="#BasicBlock"><tt>BasicBlock</tt></a> the <tt>Instruction</tt> is embedded
2660 into. To represent a specific type of instruction, one of many subclasses of
2661 <tt>Instruction</tt> are used.</p>
2663 <p> Because the <tt>Instruction</tt> class subclasses the <a
2664 href="#User"><tt>User</tt></a> class, its operands can be accessed in the same
2665 way as for other <a href="#User"><tt>User</tt></a>s (with the
2666 <tt>getOperand()</tt>/<tt>getNumOperands()</tt> and
2667 <tt>op_begin()</tt>/<tt>op_end()</tt> methods).</p> <p> An important file for
2668 the <tt>Instruction</tt> class is the <tt>llvm/Instruction.def</tt> file. This
2669 file contains some meta-data about the various different types of instructions
2670 in LLVM. It describes the enum values that are used as opcodes (for example
2671 <tt>Instruction::Add</tt> and <tt>Instruction::ICmp</tt>), as well as the
2672 concrete sub-classes of <tt>Instruction</tt> that implement the instruction (for
2673 example <tt><a href="#BinaryOperator">BinaryOperator</a></tt> and <tt><a
2674 href="#CmpInst">CmpInst</a></tt>). Unfortunately, the use of macros in
2675 this file confuses doxygen, so these enum values don't show up correctly in the
2676 <a href="/doxygen/classllvm_1_1Instruction.html">doxygen output</a>.</p>
2680 <!-- _______________________________________________________________________ -->
2681 <div class="doc_subsubsection">
2682 <a name="s_Instruction">Important Subclasses of the <tt>Instruction</tt>
2685 <div class="doc_text">
2687 <li><tt><a name="BinaryOperator">BinaryOperator</a></tt>
2688 <p>This subclasses represents all two operand instructions whose operands
2689 must be the same type, except for the comparison instructions.</p></li>
2690 <li><tt><a name="CastInst">CastInst</a></tt>
2691 <p>This subclass is the parent of the 12 casting instructions. It provides
2692 common operations on cast instructions.</p>
2693 <li><tt><a name="CmpInst">CmpInst</a></tt>
2694 <p>This subclass respresents the two comparison instructions,
2695 <a href="LangRef.html#i_icmp">ICmpInst</a> (integer opreands), and
2696 <a href="LangRef.html#i_fcmp">FCmpInst</a> (floating point operands).</p>
2697 <li><tt><a name="TerminatorInst">TerminatorInst</a></tt>
2698 <p>This subclass is the parent of all terminator instructions (those which
2699 can terminate a block).</p>
2703 <!-- _______________________________________________________________________ -->
2704 <div class="doc_subsubsection">
2705 <a name="m_Instruction">Important Public Members of the <tt>Instruction</tt>
2709 <div class="doc_text">
2712 <li><tt><a href="#BasicBlock">BasicBlock</a> *getParent()</tt>
2713 <p>Returns the <a href="#BasicBlock"><tt>BasicBlock</tt></a> that
2714 this <tt>Instruction</tt> is embedded into.</p></li>
2715 <li><tt>bool mayWriteToMemory()</tt>
2716 <p>Returns true if the instruction writes to memory, i.e. it is a
2717 <tt>call</tt>,<tt>free</tt>,<tt>invoke</tt>, or <tt>store</tt>.</p></li>
2718 <li><tt>unsigned getOpcode()</tt>
2719 <p>Returns the opcode for the <tt>Instruction</tt>.</p></li>
2720 <li><tt><a href="#Instruction">Instruction</a> *clone() const</tt>
2721 <p>Returns another instance of the specified instruction, identical
2722 in all ways to the original except that the instruction has no parent
2723 (ie it's not embedded into a <a href="#BasicBlock"><tt>BasicBlock</tt></a>),
2724 and it has no name</p></li>
2729 <!-- ======================================================================= -->
2730 <div class="doc_subsection">
2731 <a name="Constant">The <tt>Constant</tt> class and subclasses</a>
2734 <div class="doc_text">
2736 <p>Constant represents a base class for different types of constants. It
2737 is subclassed by ConstantInt, ConstantArray, etc. for representing
2738 the various types of Constants. <a href="#GlobalValue">GlobalValue</a> is also
2739 a subclass, which represents the address of a global variable or function.
2744 <!-- _______________________________________________________________________ -->
2745 <div class="doc_subsubsection">Important Subclasses of Constant </div>
2746 <div class="doc_text">
2748 <li>ConstantInt : This subclass of Constant represents an integer constant of
2751 <li><tt>int64_t getSExtValue() const</tt>: Returns the underlying value of
2752 this constant as a sign extended signed integer value.</li>
2753 <li><tt>uint64_t getZExtValue() const</tt>: Returns the underlying value
2754 of this constant as a zero extended unsigned integer value.</li>
2755 <li><tt>static ConstantInt* get(const Type *Ty, uint64_t Val)</tt>:
2756 Returns the ConstantInt object that represents the value provided by
2757 <tt>Val</tt> for integer type <tt>Ty</tt>.</li>
2760 <li>ConstantFP : This class represents a floating point constant.
2762 <li><tt>double getValue() const</tt>: Returns the underlying value of
2763 this constant. </li>
2766 <li>ConstantArray : This represents a constant array.
2768 <li><tt>const std::vector<Use> &getValues() const</tt>: Returns
2769 a vector of component constants that makeup this array. </li>
2772 <li>ConstantStruct : This represents a constant struct.
2774 <li><tt>const std::vector<Use> &getValues() const</tt>: Returns
2775 a vector of component constants that makeup this array. </li>
2778 <li>GlobalValue : This represents either a global variable or a function. In
2779 either case, the value is a constant fixed address (after linking).
2785 <!-- ======================================================================= -->
2786 <div class="doc_subsection">
2787 <a name="GlobalValue">The <tt>GlobalValue</tt> class</a>
2790 <div class="doc_text">
2793 href="/doxygen/GlobalValue_8h-source.html">llvm/GlobalValue.h</a>"</tt><br>
2794 doxygen info: <a href="/doxygen/classllvm_1_1GlobalValue.html">GlobalValue
2796 Superclasses: <a href="#Constant"><tt>Constant</tt></a>,
2797 <a href="#User"><tt>User</tt></a>, <a href="#Value"><tt>Value</tt></a></p>
2799 <p>Global values (<a href="#GlobalVariable"><tt>GlobalVariable</tt></a>s or <a
2800 href="#Function"><tt>Function</tt></a>s) are the only LLVM values that are
2801 visible in the bodies of all <a href="#Function"><tt>Function</tt></a>s.
2802 Because they are visible at global scope, they are also subject to linking with
2803 other globals defined in different translation units. To control the linking
2804 process, <tt>GlobalValue</tt>s know their linkage rules. Specifically,
2805 <tt>GlobalValue</tt>s know whether they have internal or external linkage, as
2806 defined by the <tt>LinkageTypes</tt> enumeration.</p>
2808 <p>If a <tt>GlobalValue</tt> has internal linkage (equivalent to being
2809 <tt>static</tt> in C), it is not visible to code outside the current translation
2810 unit, and does not participate in linking. If it has external linkage, it is
2811 visible to external code, and does participate in linking. In addition to
2812 linkage information, <tt>GlobalValue</tt>s keep track of which <a
2813 href="#Module"><tt>Module</tt></a> they are currently part of.</p>
2815 <p>Because <tt>GlobalValue</tt>s are memory objects, they are always referred to
2816 by their <b>address</b>. As such, the <a href="#Type"><tt>Type</tt></a> of a
2817 global is always a pointer to its contents. It is important to remember this
2818 when using the <tt>GetElementPtrInst</tt> instruction because this pointer must
2819 be dereferenced first. For example, if you have a <tt>GlobalVariable</tt> (a
2820 subclass of <tt>GlobalValue)</tt> that is an array of 24 ints, type <tt>[24 x
2821 i32]</tt>, then the <tt>GlobalVariable</tt> is a pointer to that array. Although
2822 the address of the first element of this array and the value of the
2823 <tt>GlobalVariable</tt> are the same, they have different types. The
2824 <tt>GlobalVariable</tt>'s type is <tt>[24 x i32]</tt>. The first element's type
2825 is <tt>i32.</tt> Because of this, accessing a global value requires you to
2826 dereference the pointer with <tt>GetElementPtrInst</tt> first, then its elements
2827 can be accessed. This is explained in the <a href="LangRef.html#globalvars">LLVM
2828 Language Reference Manual</a>.</p>
2832 <!-- _______________________________________________________________________ -->
2833 <div class="doc_subsubsection">
2834 <a name="m_GlobalValue">Important Public Members of the <tt>GlobalValue</tt>
2838 <div class="doc_text">
2841 <li><tt>bool hasInternalLinkage() const</tt><br>
2842 <tt>bool hasExternalLinkage() const</tt><br>
2843 <tt>void setInternalLinkage(bool HasInternalLinkage)</tt>
2844 <p> These methods manipulate the linkage characteristics of the <tt>GlobalValue</tt>.</p>
2847 <li><tt><a href="#Module">Module</a> *getParent()</tt>
2848 <p> This returns the <a href="#Module"><tt>Module</tt></a> that the
2849 GlobalValue is currently embedded into.</p></li>
2854 <!-- ======================================================================= -->
2855 <div class="doc_subsection">
2856 <a name="Function">The <tt>Function</tt> class</a>
2859 <div class="doc_text">
2862 href="/doxygen/Function_8h-source.html">llvm/Function.h</a>"</tt><br> doxygen
2863 info: <a href="/doxygen/classllvm_1_1Function.html">Function Class</a><br>
2864 Superclasses: <a href="#GlobalValue"><tt>GlobalValue</tt></a>,
2865 <a href="#Constant"><tt>Constant</tt></a>,
2866 <a href="#User"><tt>User</tt></a>,
2867 <a href="#Value"><tt>Value</tt></a></p>
2869 <p>The <tt>Function</tt> class represents a single procedure in LLVM. It is
2870 actually one of the more complex classes in the LLVM heirarchy because it must
2871 keep track of a large amount of data. The <tt>Function</tt> class keeps track
2872 of a list of <a href="#BasicBlock"><tt>BasicBlock</tt></a>s, a list of formal
2873 <a href="#Argument"><tt>Argument</tt></a>s, and a
2874 <a href="#SymbolTable"><tt>SymbolTable</tt></a>.</p>
2876 <p>The list of <a href="#BasicBlock"><tt>BasicBlock</tt></a>s is the most
2877 commonly used part of <tt>Function</tt> objects. The list imposes an implicit
2878 ordering of the blocks in the function, which indicate how the code will be
2879 layed out by the backend. Additionally, the first <a
2880 href="#BasicBlock"><tt>BasicBlock</tt></a> is the implicit entry node for the
2881 <tt>Function</tt>. It is not legal in LLVM to explicitly branch to this initial
2882 block. There are no implicit exit nodes, and in fact there may be multiple exit
2883 nodes from a single <tt>Function</tt>. If the <a
2884 href="#BasicBlock"><tt>BasicBlock</tt></a> list is empty, this indicates that
2885 the <tt>Function</tt> is actually a function declaration: the actual body of the
2886 function hasn't been linked in yet.</p>
2888 <p>In addition to a list of <a href="#BasicBlock"><tt>BasicBlock</tt></a>s, the
2889 <tt>Function</tt> class also keeps track of the list of formal <a
2890 href="#Argument"><tt>Argument</tt></a>s that the function receives. This
2891 container manages the lifetime of the <a href="#Argument"><tt>Argument</tt></a>
2892 nodes, just like the <a href="#BasicBlock"><tt>BasicBlock</tt></a> list does for
2893 the <a href="#BasicBlock"><tt>BasicBlock</tt></a>s.</p>
2895 <p>The <a href="#SymbolTable"><tt>SymbolTable</tt></a> is a very rarely used
2896 LLVM feature that is only used when you have to look up a value by name. Aside
2897 from that, the <a href="#SymbolTable"><tt>SymbolTable</tt></a> is used
2898 internally to make sure that there are not conflicts between the names of <a
2899 href="#Instruction"><tt>Instruction</tt></a>s, <a
2900 href="#BasicBlock"><tt>BasicBlock</tt></a>s, or <a
2901 href="#Argument"><tt>Argument</tt></a>s in the function body.</p>
2903 <p>Note that <tt>Function</tt> is a <a href="#GlobalValue">GlobalValue</a>
2904 and therefore also a <a href="#Constant">Constant</a>. The value of the function
2905 is its address (after linking) which is guaranteed to be constant.</p>
2908 <!-- _______________________________________________________________________ -->
2909 <div class="doc_subsubsection">
2910 <a name="m_Function">Important Public Members of the <tt>Function</tt>
2914 <div class="doc_text">
2917 <li><tt>Function(const </tt><tt><a href="#FunctionType">FunctionType</a>
2918 *Ty, LinkageTypes Linkage, const std::string &N = "", Module* Parent = 0)</tt>
2920 <p>Constructor used when you need to create new <tt>Function</tt>s to add
2921 the the program. The constructor must specify the type of the function to
2922 create and what type of linkage the function should have. The <a
2923 href="#FunctionType"><tt>FunctionType</tt></a> argument
2924 specifies the formal arguments and return value for the function. The same
2925 <a href="#FunctionTypel"><tt>FunctionType</tt></a> value can be used to
2926 create multiple functions. The <tt>Parent</tt> argument specifies the Module
2927 in which the function is defined. If this argument is provided, the function
2928 will automatically be inserted into that module's list of
2931 <li><tt>bool isExternal()</tt>
2933 <p>Return whether or not the <tt>Function</tt> has a body defined. If the
2934 function is "external", it does not have a body, and thus must be resolved
2935 by linking with a function defined in a different translation unit.</p></li>
2937 <li><tt>Function::iterator</tt> - Typedef for basic block list iterator<br>
2938 <tt>Function::const_iterator</tt> - Typedef for const_iterator.<br>
2940 <tt>begin()</tt>, <tt>end()</tt>
2941 <tt>size()</tt>, <tt>empty()</tt>
2943 <p>These are forwarding methods that make it easy to access the contents of
2944 a <tt>Function</tt> object's <a href="#BasicBlock"><tt>BasicBlock</tt></a>
2947 <li><tt>Function::BasicBlockListType &getBasicBlockList()</tt>
2949 <p>Returns the list of <a href="#BasicBlock"><tt>BasicBlock</tt></a>s. This
2950 is necessary to use when you need to update the list or perform a complex
2951 action that doesn't have a forwarding method.</p></li>
2953 <li><tt>Function::arg_iterator</tt> - Typedef for the argument list
2955 <tt>Function::const_arg_iterator</tt> - Typedef for const_iterator.<br>
2957 <tt>arg_begin()</tt>, <tt>arg_end()</tt>
2958 <tt>arg_size()</tt>, <tt>arg_empty()</tt>
2960 <p>These are forwarding methods that make it easy to access the contents of
2961 a <tt>Function</tt> object's <a href="#Argument"><tt>Argument</tt></a>
2964 <li><tt>Function::ArgumentListType &getArgumentList()</tt>
2966 <p>Returns the list of <a href="#Argument"><tt>Argument</tt></a>s. This is
2967 necessary to use when you need to update the list or perform a complex
2968 action that doesn't have a forwarding method.</p></li>
2970 <li><tt><a href="#BasicBlock">BasicBlock</a> &getEntryBlock()</tt>
2972 <p>Returns the entry <a href="#BasicBlock"><tt>BasicBlock</tt></a> for the
2973 function. Because the entry block for the function is always the first
2974 block, this returns the first block of the <tt>Function</tt>.</p></li>
2976 <li><tt><a href="#Type">Type</a> *getReturnType()</tt><br>
2977 <tt><a href="#FunctionType">FunctionType</a> *getFunctionType()</tt>
2979 <p>This traverses the <a href="#Type"><tt>Type</tt></a> of the
2980 <tt>Function</tt> and returns the return type of the function, or the <a
2981 href="#FunctionType"><tt>FunctionType</tt></a> of the actual
2984 <li><tt><a href="#SymbolTable">SymbolTable</a> *getSymbolTable()</tt>
2986 <p> Return a pointer to the <a href="#SymbolTable"><tt>SymbolTable</tt></a>
2987 for this <tt>Function</tt>.</p></li>
2992 <!-- ======================================================================= -->
2993 <div class="doc_subsection">
2994 <a name="GlobalVariable">The <tt>GlobalVariable</tt> class</a>
2997 <div class="doc_text">
3000 href="/doxygen/GlobalVariable_8h-source.html">llvm/GlobalVariable.h</a>"</tt>
3002 doxygen info: <a href="/doxygen/classllvm_1_1GlobalVariable.html">GlobalVariable
3004 Superclasses: <a href="#GlobalValue"><tt>GlobalValue</tt></a>,
3005 <a href="#Constant"><tt>Constant</tt></a>,
3006 <a href="#User"><tt>User</tt></a>,
3007 <a href="#Value"><tt>Value</tt></a></p>
3009 <p>Global variables are represented with the (suprise suprise)
3010 <tt>GlobalVariable</tt> class. Like functions, <tt>GlobalVariable</tt>s are also
3011 subclasses of <a href="#GlobalValue"><tt>GlobalValue</tt></a>, and as such are
3012 always referenced by their address (global values must live in memory, so their
3013 "name" refers to their constant address). See
3014 <a href="#GlobalValue"><tt>GlobalValue</tt></a> for more on this. Global
3015 variables may have an initial value (which must be a
3016 <a href="#Constant"><tt>Constant</tt></a>), and if they have an initializer,
3017 they may be marked as "constant" themselves (indicating that their contents
3018 never change at runtime).</p>
3021 <!-- _______________________________________________________________________ -->
3022 <div class="doc_subsubsection">
3023 <a name="m_GlobalVariable">Important Public Members of the
3024 <tt>GlobalVariable</tt> class</a>
3027 <div class="doc_text">
3030 <li><tt>GlobalVariable(const </tt><tt><a href="#Type">Type</a> *Ty, bool
3031 isConstant, LinkageTypes& Linkage, <a href="#Constant">Constant</a>
3032 *Initializer = 0, const std::string &Name = "", Module* Parent = 0)</tt>
3034 <p>Create a new global variable of the specified type. If
3035 <tt>isConstant</tt> is true then the global variable will be marked as
3036 unchanging for the program. The Linkage parameter specifies the type of
3037 linkage (internal, external, weak, linkonce, appending) for the variable. If
3038 the linkage is InternalLinkage, WeakLinkage, or LinkOnceLinkage, then
3039 the resultant global variable will have internal linkage. AppendingLinkage
3040 concatenates together all instances (in different translation units) of the
3041 variable into a single variable but is only applicable to arrays. See
3042 the <a href="LangRef.html#modulestructure">LLVM Language Reference</a> for
3043 further details on linkage types. Optionally an initializer, a name, and the
3044 module to put the variable into may be specified for the global variable as
3047 <li><tt>bool isConstant() const</tt>
3049 <p>Returns true if this is a global variable that is known not to
3050 be modified at runtime.</p></li>
3052 <li><tt>bool hasInitializer()</tt>
3054 <p>Returns true if this <tt>GlobalVariable</tt> has an intializer.</p></li>
3056 <li><tt><a href="#Constant">Constant</a> *getInitializer()</tt>
3058 <p>Returns the intial value for a <tt>GlobalVariable</tt>. It is not legal
3059 to call this method if there is no initializer.</p></li>
3065 <!-- ======================================================================= -->
3066 <div class="doc_subsection">
3067 <a name="BasicBlock">The <tt>BasicBlock</tt> class</a>
3070 <div class="doc_text">
3073 href="/doxygen/BasicBlock_8h-source.html">llvm/BasicBlock.h</a>"</tt><br>
3074 doxygen info: <a href="/doxygen/structllvm_1_1BasicBlock.html">BasicBlock
3076 Superclass: <a href="#Value"><tt>Value</tt></a></p>
3078 <p>This class represents a single entry multiple exit section of the code,
3079 commonly known as a basic block by the compiler community. The
3080 <tt>BasicBlock</tt> class maintains a list of <a
3081 href="#Instruction"><tt>Instruction</tt></a>s, which form the body of the block.
3082 Matching the language definition, the last element of this list of instructions
3083 is always a terminator instruction (a subclass of the <a
3084 href="#TerminatorInst"><tt>TerminatorInst</tt></a> class).</p>
3086 <p>In addition to tracking the list of instructions that make up the block, the
3087 <tt>BasicBlock</tt> class also keeps track of the <a
3088 href="#Function"><tt>Function</tt></a> that it is embedded into.</p>
3090 <p>Note that <tt>BasicBlock</tt>s themselves are <a
3091 href="#Value"><tt>Value</tt></a>s, because they are referenced by instructions
3092 like branches and can go in the switch tables. <tt>BasicBlock</tt>s have type
3097 <!-- _______________________________________________________________________ -->
3098 <div class="doc_subsubsection">
3099 <a name="m_BasicBlock">Important Public Members of the <tt>BasicBlock</tt>
3103 <div class="doc_text">
3106 <li><tt>BasicBlock(const std::string &Name = "", </tt><tt><a
3107 href="#Function">Function</a> *Parent = 0)</tt>
3109 <p>The <tt>BasicBlock</tt> constructor is used to create new basic blocks for
3110 insertion into a function. The constructor optionally takes a name for the new
3111 block, and a <a href="#Function"><tt>Function</tt></a> to insert it into. If
3112 the <tt>Parent</tt> parameter is specified, the new <tt>BasicBlock</tt> is
3113 automatically inserted at the end of the specified <a
3114 href="#Function"><tt>Function</tt></a>, if not specified, the BasicBlock must be
3115 manually inserted into the <a href="#Function"><tt>Function</tt></a>.</p></li>
3117 <li><tt>BasicBlock::iterator</tt> - Typedef for instruction list iterator<br>
3118 <tt>BasicBlock::const_iterator</tt> - Typedef for const_iterator.<br>
3119 <tt>begin()</tt>, <tt>end()</tt>, <tt>front()</tt>, <tt>back()</tt>,
3120 <tt>size()</tt>, <tt>empty()</tt>
3121 STL-style functions for accessing the instruction list.
3123 <p>These methods and typedefs are forwarding functions that have the same
3124 semantics as the standard library methods of the same names. These methods
3125 expose the underlying instruction list of a basic block in a way that is easy to
3126 manipulate. To get the full complement of container operations (including
3127 operations to update the list), you must use the <tt>getInstList()</tt>
3130 <li><tt>BasicBlock::InstListType &getInstList()</tt>
3132 <p>This method is used to get access to the underlying container that actually
3133 holds the Instructions. This method must be used when there isn't a forwarding
3134 function in the <tt>BasicBlock</tt> class for the operation that you would like
3135 to perform. Because there are no forwarding functions for "updating"
3136 operations, you need to use this if you want to update the contents of a
3137 <tt>BasicBlock</tt>.</p></li>
3139 <li><tt><a href="#Function">Function</a> *getParent()</tt>
3141 <p> Returns a pointer to <a href="#Function"><tt>Function</tt></a> the block is
3142 embedded into, or a null pointer if it is homeless.</p></li>
3144 <li><tt><a href="#TerminatorInst">TerminatorInst</a> *getTerminator()</tt>
3146 <p> Returns a pointer to the terminator instruction that appears at the end of
3147 the <tt>BasicBlock</tt>. If there is no terminator instruction, or if the last
3148 instruction in the block is not a terminator, then a null pointer is
3156 <!-- ======================================================================= -->
3157 <div class="doc_subsection">
3158 <a name="Argument">The <tt>Argument</tt> class</a>
3161 <div class="doc_text">
3163 <p>This subclass of Value defines the interface for incoming formal
3164 arguments to a function. A Function maintains a list of its formal
3165 arguments. An argument has a pointer to the parent Function.</p>
3169 <!-- *********************************************************************** -->
3172 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
3173 src="http://jigsaw.w3.org/css-validator/images/vcss" alt="Valid CSS!"></a>
3174 <a href="http://validator.w3.org/check/referer"><img
3175 src="http://www.w3.org/Icons/valid-html401" alt="Valid HTML 4.01!" /></a>
3177 <a href="mailto:dhurjati@cs.uiuc.edu">Dinakar Dhurjati</a> and
3178 <a href="mailto:sabre@nondot.org">Chris Lattner</a><br>
3179 <a href="http://llvm.org">The LLVM Compiler Infrastructure</a><br>
3180 Last modified: $Date$