Blame - docs/ProgrammersManual.html - platform/external/llvm

blob: cf46a97e68869ff388bfc603262b7d117db90bc3 [file] [log] [blame]

Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	1	<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"
				2	"http://www.w3.org/TR/html4/strict.dtd">
				3	<html>
				4	<head>
				5	<title>LLVM Programmer's Manual</title>
				6	<link rel="stylesheet" href="llvm.css" type="text/css">
				7	</head>
				8	<body>
				9
				10	<div class="doc_title">
				11	LLVM Programmer's Manual
				12	</div>
				13
				14	<ol>
				15	<li><a href="#introduction">Introduction</a></li>
				16	<li><a href="#general">General Information</a>
				17	<ul>
				18	<li><a href="#stl">The C++ Standard Template Library</a></li>
				19	<!--
				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>
				23
				24	-->
				25	</ul>
				26	</li>
				27	<li><a href="#apis">Important and useful LLVM APIs</a>
				28	<ul>
				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>
				32	option</a>
				33	<ul>
				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>
				36	</ul>
				37	</li>
				38	<li><a href="#Statistic">The <tt>Statistic</tt> class & <tt>-stats</tt>
				39	option</a></li>
				40	<!--
				41	<li>The <tt>InstVisitor</tt> template
				42	<li>The general graph API
				43	-->
				44	<li><a href="#ViewGraph">Viewing graphs while debugging code</a></li>
				45	</ul>
				46	</li>
				47	<li><a href="#datastructure">Picking the Right Data Structure for a Task</a>
				48	<ul>
				49	<li><a href="#ds_sequential">Sequential Containers (std::vector, std::list, etc)</a>
				50	<ul>
				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>
Gabor Greif	bb17f65	2009-02-27 11:37:41 +0000	[diff] [blame]	57	<li><a href="#dss_ilist">llvm/ADT/ilist.h</a></li>
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	58	<li><a href="#dss_other">Other Sequential Container Options</a></li>
				59	</ul></li>
				60	<li><a href="#ds_set">Set-Like Containers (std::set, SmallSet, SetVector, etc)</a>
				61	<ul>
				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>
Chris Lattner	77ab46d	2007-09-30 00:58:59 +0000	[diff] [blame]	65	<li><a href="#dss_denseset">"llvm/ADT/DenseSet.h"</a></li>
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	66	<li><a href="#dss_FoldingSet">"llvm/ADT/FoldingSet.h"</a></li>
				67	<li><a href="#dss_set"><set></a></li>
				68	<li><a href="#dss_setvector">"llvm/ADT/SetVector.h"</a></li>
				69	<li><a href="#dss_uniquevector">"llvm/ADT/UniqueVector.h"</a></li>
				70	<li><a href="#dss_otherset">Other Set-Like ContainerOptions</a></li>
				71	</ul></li>
				72	<li><a href="#ds_map">Map-Like Containers (std::map, DenseMap, etc)</a>
				73	<ul>
				74	<li><a href="#dss_sortedvectormap">A sorted 'vector'</a></li>
				75	<li><a href="#dss_stringmap">"llvm/ADT/StringMap.h"</a></li>
				76	<li><a href="#dss_indexedmap">"llvm/ADT/IndexedMap.h"</a></li>
				77	<li><a href="#dss_densemap">"llvm/ADT/DenseMap.h"</a></li>
				78	<li><a href="#dss_map"><map></a></li>
				79	<li><a href="#dss_othermap">Other Map-Like Container Options</a></li>
				80	</ul></li>
Daniel Berlin	7ea44dc	2007-09-24 17:52:25 +0000	[diff] [blame]	81	<li><a href="#ds_bit">BitVector-like containers</a>
				82	<ul>
				83	<li><a href="#dss_bitvector">A dense bitvector</a></li>
				84	<li><a href="#dss_sparsebitvector">A sparse bitvector</a></li>
				85	</ul></li>
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	86	</ul>
				87	</li>
				88	<li><a href="#common">Helpful Hints for Common Operations</a>
				89	<ul>
				90	<li><a href="#inspection">Basic Inspection and Traversal Routines</a>
				91	<ul>
				92	<li><a href="#iterate_function">Iterating over the <tt>BasicBlock</tt>s
				93	in a <tt>Function</tt></a> </li>
				94	<li><a href="#iterate_basicblock">Iterating over the <tt>Instruction</tt>s
				95	in a <tt>BasicBlock</tt></a> </li>
				96	<li><a href="#iterate_institer">Iterating over the <tt>Instruction</tt>s
				97	in a <tt>Function</tt></a> </li>
				98	<li><a href="#iterate_convert">Turning an iterator into a
				99	class pointer</a> </li>
				100	<li><a href="#iterate_complex">Finding call sites: a more
				101	complex example</a> </li>
				102	<li><a href="#calls_and_invokes">Treating calls and invokes
				103	the same way</a> </li>
				104	<li><a href="#iterate_chains">Iterating over def-use &
				105	use-def chains</a> </li>
Chris Lattner	0665e1f	2008-01-03 16:56:04 +0000	[diff] [blame]	106	<li><a href="#iterate_preds">Iterating over predecessors &
				107	successors of blocks</a></li>
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	108	</ul>
				109	</li>
				110	<li><a href="#simplechanges">Making simple changes</a>
				111	<ul>
				112	<li><a href="#schanges_creating">Creating and inserting new
				113	<tt>Instruction</tt>s</a> </li>
				114	<li><a href="#schanges_deleting">Deleting <tt>Instruction</tt>s</a> </li>
				115	<li><a href="#schanges_replacing">Replacing an <tt>Instruction</tt>
				116	with another <tt>Value</tt></a> </li>
				117	<li><a href="#schanges_deletingGV">Deleting <tt>GlobalVariable</tt>s</a> </li>
				118	</ul>
				119	</li>
				120	<!--
				121	<li>Working with the Control Flow Graph
				122	<ul>
				123	<li>Accessing predecessors and successors of a <tt>BasicBlock</tt>
				124	<li>
				125	<li>
				126	</ul>
				127	-->
				128	</ul>
				129	</li>
				130
				131	<li><a href="#advanced">Advanced Topics</a>
				132	<ul>
				133	<li><a href="#TypeResolve">LLVM Type Resolution</a>
				134	<ul>
				135	<li><a href="#BuildRecType">Basic Recursive Type Construction</a></li>
				136	<li><a href="#refineAbstractTypeTo">The <tt>refineAbstractTypeTo</tt> method</a></li>
				137	<li><a href="#PATypeHolder">The PATypeHolder Class</a></li>
				138	<li><a href="#AbstractTypeUser">The AbstractTypeUser Class</a></li>
				139	</ul></li>
				140
Gabor Greif	92e8776	2008-06-16 21:06:12 +0000	[diff] [blame]	141	<li><a href="#SymbolTable">The <tt>ValueSymbolTable</tt> and <tt>TypeSymbolTable</tt> classes</a></li>
				142	<li><a href="#UserLayout">The <tt>User</tt> and owned <tt>Use</tt> classes' memory layout</a></li>
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	143	</ul></li>
				144
				145	<li><a href="#coreclasses">The Core LLVM Class Hierarchy Reference</a>
				146	<ul>
				147	<li><a href="#Type">The <tt>Type</tt> class</a> </li>
				148	<li><a href="#Module">The <tt>Module</tt> class</a></li>
				149	<li><a href="#Value">The <tt>Value</tt> class</a>
				150	<ul>
				151	<li><a href="#User">The <tt>User</tt> class</a>
				152	<ul>
				153	<li><a href="#Instruction">The <tt>Instruction</tt> class</a></li>
				154	<li><a href="#Constant">The <tt>Constant</tt> class</a>
				155	<ul>
				156	<li><a href="#GlobalValue">The <tt>GlobalValue</tt> class</a>
				157	<ul>
				158	<li><a href="#Function">The <tt>Function</tt> class</a></li>
				159	<li><a href="#GlobalVariable">The <tt>GlobalVariable</tt> class</a></li>
				160	</ul>
				161	</li>
				162	</ul>
				163	</li>
				164	</ul>
				165	</li>
				166	<li><a href="#BasicBlock">The <tt>BasicBlock</tt> class</a></li>
				167	<li><a href="#Argument">The <tt>Argument</tt> class</a></li>
				168	</ul>
				169	</li>
				170	</ul>
				171	</li>
				172	</ol>
				173
				174	<div class="doc_author">
				175	<p>Written by <a href="mailto:sabre@nondot.org">Chris Lattner</a>,
				176	<a href="mailto:dhurjati@cs.uiuc.edu">Dinakar Dhurjati</a>,
Gabor Greif	92e8776	2008-06-16 21:06:12 +0000	[diff] [blame]	177	<a href="mailto:ggreif@gmail.com">Gabor Greif</a>,
				178	<a href="mailto:jstanley@cs.uiuc.edu">Joel Stanley</a> and
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	179	<a href="mailto:rspencer@x10sys.com">Reid Spencer</a></p>
				180	</div>
				181
				182	<!-- *********************************************************************** -->
				183	<div class="doc_section">
				184	<a name="introduction">Introduction </a>
				185	</div>
				186	<!-- *********************************************************************** -->
				187
				188	<div class="doc_text">
				189
				190	<p>This document is meant to highlight some of the important classes and
				191	interfaces available in the LLVM source-base. This manual is not
				192	intended to explain what LLVM is, how it works, and what LLVM code looks
				193	like. It assumes that you know the basics of LLVM and are interested
				194	in writing transformations or otherwise analyzing or manipulating the
				195	code.</p>
				196
				197	<p>This document should get you oriented so that you can find your
				198	way in the continuously growing source code that makes up the LLVM
				199	infrastructure. Note that this manual is not intended to serve as a
				200	replacement for reading the source code, so if you think there should be
				201	a method in one of these classes to do something, but it's not listed,
				202	check the source. Links to the <a href="/doxygen/">doxygen</a> sources
				203	are provided to make this as easy as possible.</p>
				204
				205	<p>The first section of this document describes general information that is
				206	useful to know when working in the LLVM infrastructure, and the second describes
				207	the Core LLVM classes. In the future this manual will be extended with
				208	information describing how to use extension libraries, such as dominator
				209	information, CFG traversal routines, and useful utilities like the <tt><a
				210	href="/doxygen/InstVisitor_8h-source.html">InstVisitor</a></tt> template.</p>
				211
				212	</div>
				213
				214	<!-- *********************************************************************** -->
				215	<div class="doc_section">
				216	<a name="general">General Information</a>
				217	</div>
				218	<!-- *********************************************************************** -->
				219
				220	<div class="doc_text">
				221
				222	<p>This section contains general information that is useful if you are working
				223	in the LLVM source-base, but that isn't specific to any particular API.</p>
				224
				225	</div>
				226
				227	<!-- ======================================================================= -->
				228	<div class="doc_subsection">
				229	<a name="stl">The C++ Standard Template Library</a>
				230	</div>
				231
				232	<div class="doc_text">
				233
				234	<p>LLVM makes heavy use of the C++ Standard Template Library (STL),
				235	perhaps much more than you are used to, or have seen before. Because of
				236	this, you might want to do a little background reading in the
				237	techniques used and capabilities of the library. There are many good
				238	pages that discuss the STL, and several books on the subject that you
				239	can get, so it will not be discussed in this document.</p>
				240
				241	<p>Here are some useful links:</p>
				242
				243	<ol>
				244
				245	<li><a href="http://www.dinkumware.com/refxcpp.html">Dinkumware C++ Library
				246	reference</a> - an excellent reference for the STL and other parts of the
				247	standard C++ library.</li>
				248
				249	<li><a href="http://www.tempest-sw.com/cpp/">C++ In a Nutshell</a> - This is an
				250	O'Reilly book in the making. It has a decent
				251	Standard Library
				252	Reference that rivals Dinkumware's, and is unfortunately no longer free since the book has been
				253	published.</li>
				254
				255	<li><a href="http://www.parashift.com/c++-faq-lite/">C++ Frequently Asked
				256	Questions</a></li>
				257
				258	<li><a href="http://www.sgi.com/tech/stl/">SGI's STL Programmer's Guide</a> -
				259	Contains a useful <a
				260	href="http://www.sgi.com/tech/stl/stl_introduction.html">Introduction to the
				261	STL</a>.</li>
				262
				263	<li><a href="http://www.research.att.com/%7Ebs/C++.html">Bjarne Stroustrup's C++
				264	Page</a></li>
				265
				266	<li><a href="http://64.78.49.204/">
				267	Bruce Eckel's Thinking in C++, 2nd ed. Volume 2 Revision 4.0 (even better, get
				268	the book).</a></li>
				269
				270	</ol>
				271
				272	<p>You are also encouraged to take a look at the <a
				273	href="CodingStandards.html">LLVM Coding Standards</a> guide which focuses on how
				274	to write maintainable code more than where to put your curly braces.</p>
				275
				276	</div>
				277
				278	<!-- ======================================================================= -->
				279	<div class="doc_subsection">
				280	<a name="stl">Other useful references</a>
				281	</div>
				282
				283	<div class="doc_text">
				284
				285	<ol>
				286	<li><a href="http://www.psc.edu/%7Esemke/cvs_branches.html">CVS
				287	Branch and Tag Primer</a></li>
				288	<li><a href="http://www.fortran-2000.com/ArnaudRecipes/sharedlib.html">Using
				289	static and shared libraries across platforms</a></li>
				290	</ol>
				291
				292	</div>
				293
				294	<!-- *********************************************************************** -->
				295	<div class="doc_section">
				296	<a name="apis">Important and useful LLVM APIs</a>
				297	</div>
				298	<!-- *********************************************************************** -->
				299
				300	<div class="doc_text">
				301
				302	<p>Here we highlight some LLVM APIs that are generally useful and good to
				303	know about when writing transformations.</p>
				304
				305	</div>
				306
				307	<!-- ======================================================================= -->
				308	<div class="doc_subsection">
				309	<a name="isa">The <tt>isa<></tt>, <tt>cast<></tt> and
				310	<tt>dyn_cast<></tt> templates</a>
				311	</div>
				312
				313	<div class="doc_text">
				314
				315	<p>The LLVM source-base makes extensive use of a custom form of RTTI.
				316	These templates have many similarities to the C++ <tt>dynamic_cast<></tt>
				317	operator, but they don't have some drawbacks (primarily stemming from
				318	the fact that <tt>dynamic_cast<></tt> only works on classes that
				319	have a v-table). Because they are used so often, you must know what they
				320	do and how they work. All of these templates are defined in the <a
				321	href="/doxygen/Casting_8h-source.html"><tt>llvm/Support/Casting.h</tt></a>
				322	file (note that you very rarely have to include this file directly).</p>
				323
				324	<dl>
				325	<dt><tt>isa<></tt>: </dt>
				326
				327	<dd><p>The <tt>isa<></tt> operator works exactly like the Java
				328	"<tt>instanceof</tt>" operator. It returns true or false depending on whether
				329	a reference or pointer points to an instance of the specified class. This can
				330	be very useful for constraint checking of various sorts (example below).</p>
				331	</dd>
				332
				333	<dt><tt>cast<></tt>: </dt>
				334
				335	<dd><p>The <tt>cast<></tt> operator is a "checked cast" operation. It
Chris Lattner	1d5610a	2008-06-20 05:03:17 +0000	[diff] [blame]	336	converts a pointer or reference from a base class to a derived class, causing
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	337	an assertion failure if it is not really an instance of the right type. This
				338	should be used in cases where you have some information that makes you believe
				339	that something is of the right type. An example of the <tt>isa<></tt>
				340	and <tt>cast<></tt> template is:</p>
				341
				342	<div class="doc_code">
				343	<pre>
				344	static bool isLoopInvariant(const <a href="#Value">Value</a> V, const Loop L) {
				345	if (isa<<a href="#Constant">Constant</a>>(V) \|\| isa<<a href="#Argument">Argument</a>>(V) \|\| isa<<a href="#GlobalValue">GlobalValue</a>>(V))
				346	return true;
				347
				348	// <i>Otherwise, it must be an instruction...</i>
				349	return !L->contains(cast<<a href="#Instruction">Instruction</a>>(V)->getParent());
				350	}
				351	</pre>
				352	</div>
				353
				354	<p>Note that you should <b>not</b> use an <tt>isa<></tt> test followed
				355	by a <tt>cast<></tt>, for that use the <tt>dyn_cast<></tt>
				356	operator.</p>
				357
				358	</dd>
				359
				360	<dt><tt>dyn_cast<></tt>:</dt>
				361
				362	<dd><p>The <tt>dyn_cast<></tt> operator is a "checking cast" operation.
				363	It checks to see if the operand is of the specified type, and if so, returns a
				364	pointer to it (this operator does not work with references). If the operand is
				365	not of the correct type, a null pointer is returned. Thus, this works very
				366	much like the <tt>dynamic_cast<></tt> operator in C++, and should be
				367	used in the same circumstances. Typically, the <tt>dyn_cast<></tt>
				368	operator is used in an <tt>if</tt> statement or some other flow control
				369	statement like this:</p>
				370
				371	<div class="doc_code">
				372	<pre>
				373	if (<a href="#AllocationInst">AllocationInst</a> *AI = dyn_cast<<a href="#AllocationInst">AllocationInst</a>>(Val)) {
				374	// <i>...</i>
				375	}
				376	</pre>
				377	</div>
				378
				379	<p>This form of the <tt>if</tt> statement effectively combines together a call
				380	to <tt>isa<></tt> and a call to <tt>cast<></tt> into one
				381	statement, which is very convenient.</p>
				382
				383	<p>Note that the <tt>dyn_cast<></tt> operator, like C++'s
				384	<tt>dynamic_cast<></tt> or Java's <tt>instanceof</tt> operator, can be
				385	abused. In particular, you should not use big chained <tt>if/then/else</tt>
				386	blocks to check for lots of different variants of classes. If you find
				387	yourself wanting to do this, it is much cleaner and more efficient to use the
				388	<tt>InstVisitor</tt> class to dispatch over the instruction type directly.</p>
				389
				390	</dd>
				391
				392	<dt><tt>cast_or_null<></tt>: </dt>
				393
				394	<dd><p>The <tt>cast_or_null<></tt> operator works just like the
				395	<tt>cast<></tt> operator, except that it allows for a null pointer as an
				396	argument (which it then propagates). This can sometimes be useful, allowing
				397	you to combine several null checks into one.</p></dd>
				398
				399	<dt><tt>dyn_cast_or_null<></tt>: </dt>
				400
				401	<dd><p>The <tt>dyn_cast_or_null<></tt> operator works just like the
				402	<tt>dyn_cast<></tt> operator, except that it allows for a null pointer
				403	as an argument (which it then propagates). This can sometimes be useful,
				404	allowing you to combine several null checks into one.</p></dd>
				405
				406	</dl>
				407
				408	<p>These five templates can be used with any classes, whether they have a
				409	v-table or not. To add support for these templates, you simply need to add
				410	<tt>classof</tt> static methods to the class you are interested casting
				411	to. Describing this is currently outside the scope of this document, but there
				412	are lots of examples in the LLVM source base.</p>
				413
				414	</div>
				415
				416	<!-- ======================================================================= -->
				417	<div class="doc_subsection">
				418	<a name="DEBUG">The <tt>DEBUG()</tt> macro and <tt>-debug</tt> option</a>
				419	</div>
				420
				421	<div class="doc_text">
				422
				423	<p>Often when working on your pass you will put a bunch of debugging printouts
				424	and other code into your pass. After you get it working, you want to remove
				425	it, but you may need it again in the future (to work out new bugs that you run
				426	across).</p>
				427
				428	<p> Naturally, because of this, you don't want to delete the debug printouts,
				429	but you don't want them to always be noisy. A standard compromise is to comment
				430	them out, allowing you to enable them if you need them in the future.</p>
				431
				432	<p>The "<tt><a href="/doxygen/Debug_8h-source.html">llvm/Support/Debug.h</a></tt>"
				433	file provides a macro named <tt>DEBUG()</tt> that is a much nicer solution to
				434	this problem. Basically, you can put arbitrary code into the argument of the
				435	<tt>DEBUG</tt> macro, and it is only executed if '<tt>opt</tt>' (or any other
				436	tool) is run with the '<tt>-debug</tt>' command line argument:</p>
				437
				438	<div class="doc_code">
				439	<pre>
				440	DOUT << "I am here!\n";
				441	</pre>
				442	</div>
				443
				444	<p>Then you can run your pass like this:</p>
				445
				446	<div class="doc_code">
				447	<pre>
				448	$ opt < a.bc > /dev/null -mypass
				449	<i><no output></i>
				450	$ opt < a.bc > /dev/null -mypass -debug
				451	I am here!
				452	</pre>
				453	</div>
				454
				455	<p>Using the <tt>DEBUG()</tt> macro instead of a home-brewed solution allows you
				456	to not have to create "yet another" command line option for the debug output for
				457	your pass. Note that <tt>DEBUG()</tt> macros are disabled for optimized builds,
				458	so they do not cause a performance impact at all (for the same reason, they
				459	should also not contain side-effects!).</p>
				460
				461	<p>One additional nice thing about the <tt>DEBUG()</tt> macro is that you can
				462	enable or disable it directly in gdb. Just use "<tt>set DebugFlag=0</tt>" or
				463	"<tt>set DebugFlag=1</tt>" from the gdb if the program is running. If the
				464	program hasn't been started yet, you can always just run it with
				465	<tt>-debug</tt>.</p>
				466
				467	</div>
				468
				469	<!-- _______________________________________________________________________ -->
				470	<div class="doc_subsubsection">
				471	<a name="DEBUG_TYPE">Fine grained debug info with <tt>DEBUG_TYPE</tt> and
				472	the <tt>-debug-only</tt> option</a>
				473	</div>
				474
				475	<div class="doc_text">
				476
				477	<p>Sometimes you may find yourself in a situation where enabling <tt>-debug</tt>
				478	just turns on <b>too much</b> information (such as when working on the code
				479	generator). If you want to enable debug information with more fine-grained
				480	control, you define the <tt>DEBUG_TYPE</tt> macro and the <tt>-debug</tt> only
				481	option as follows:</p>
				482
				483	<div class="doc_code">
				484	<pre>
				485	DOUT << "No debug type\n";
				486	#undef DEBUG_TYPE
				487	#define DEBUG_TYPE "foo"
				488	DOUT << "'foo' debug type\n";
				489	#undef DEBUG_TYPE
				490	#define DEBUG_TYPE "bar"
				491	DOUT << "'bar' debug type\n";
				492	#undef DEBUG_TYPE
				493	#define DEBUG_TYPE ""
				494	DOUT << "No debug type (2)\n";
				495	</pre>
				496	</div>
				497
				498	<p>Then you can run your pass like this:</p>
				499
				500	<div class="doc_code">
				501	<pre>
				502	$ opt < a.bc > /dev/null -mypass
				503	<i><no output></i>
				504	$ opt < a.bc > /dev/null -mypass -debug
				505	No debug type
				506	'foo' debug type
				507	'bar' debug type
				508	No debug type (2)
				509	$ opt < a.bc > /dev/null -mypass -debug-only=foo
				510	'foo' debug type
				511	$ opt < a.bc > /dev/null -mypass -debug-only=bar
				512	'bar' debug type
				513	</pre>
				514	</div>
				515
				516	<p>Of course, in practice, you should only set <tt>DEBUG_TYPE</tt> at the top of
				517	a file, to specify the debug type for the entire module (if you do this before
				518	you <tt>#include "llvm/Support/Debug.h"</tt>, you don't have to insert the ugly
				519	<tt>#undef</tt>'s). Also, you should use names more meaningful than "foo" and
				520	"bar", because there is no system in place to ensure that names do not
				521	conflict. If two different modules use the same string, they will all be turned
				522	on when the name is specified. This allows, for example, all debug information
				523	for instruction scheduling to be enabled with <tt>-debug-type=InstrSched</tt>,
				524	even if the source lives in multiple files.</p>
				525
				526	</div>
				527
				528	<!-- ======================================================================= -->
				529	<div class="doc_subsection">
				530	<a name="Statistic">The <tt>Statistic</tt> class & <tt>-stats</tt>
				531	option</a>
				532	</div>
				533
				534	<div class="doc_text">
				535
				536	<p>The "<tt><a
				537	href="/doxygen/Statistic_8h-source.html">llvm/ADT/Statistic.h</a></tt>" file
				538	provides a class named <tt>Statistic</tt> that is used as a unified way to
				539	keep track of what the LLVM compiler is doing and how effective various
				540	optimizations are. It is useful to see what optimizations are contributing to
				541	making a particular program run faster.</p>
				542
				543	<p>Often you may run your pass on some big program, and you're interested to see
				544	how many times it makes a certain transformation. Although you can do this with
				545	hand inspection, or some ad-hoc method, this is a real pain and not very useful
				546	for big programs. Using the <tt>Statistic</tt> class makes it very easy to
				547	keep track of this information, and the calculated information is presented in a
				548	uniform manner with the rest of the passes being executed.</p>
				549
				550	<p>There are many examples of <tt>Statistic</tt> uses, but the basics of using
				551	it are as follows:</p>
				552
				553	<ol>
				554	<li><p>Define your statistic like this:</p>
				555
				556	<div class="doc_code">
				557	<pre>
				558	#define <a href="#DEBUG_TYPE">DEBUG_TYPE</a> "mypassname" <i>// This goes before any #includes.</i>
				559	STATISTIC(NumXForms, "The # of times I did stuff");
				560	</pre>
				561	</div>
				562
				563	<p>The <tt>STATISTIC</tt> macro defines a static variable, whose name is
				564	specified by the first argument. The pass name is taken from the DEBUG_TYPE
				565	macro, and the description is taken from the second argument. The variable
				566	defined ("NumXForms" in this case) acts like an unsigned integer.</p></li>
				567
				568	<li><p>Whenever you make a transformation, bump the counter:</p>
				569
				570	<div class="doc_code">
				571	<pre>
				572	++NumXForms; // <i>I did stuff!</i>
				573	</pre>
				574	</div>
				575
				576	</li>
				577	</ol>
				578
				579	<p>That's all you have to do. To get '<tt>opt</tt>' to print out the
				580	statistics gathered, use the '<tt>-stats</tt>' option:</p>
				581
				582	<div class="doc_code">
				583	<pre>
				584	$ opt -stats -mypassname < program.bc > /dev/null
				585	<i>... statistics output ...</i>
				586	</pre>
				587	</div>
				588
				589	<p> When running <tt>opt</tt> on a C file from the SPEC benchmark
				590	suite, it gives a report that looks like this:</p>
				591
				592	<div class="doc_code">
				593	<pre>
				594	7646 bitcodewriter - Number of normal instructions
				595	725 bitcodewriter - Number of oversized instructions
				596	129996 bitcodewriter - Number of bitcode bytes written
				597	2817 raise - Number of insts DCEd or constprop'd
				598	3213 raise - Number of cast-of-self removed
				599	5046 raise - Number of expression trees converted
				600	75 raise - Number of other getelementptr's formed
				601	138 raise - Number of load/store peepholes
				602	42 deadtypeelim - Number of unused typenames removed from symtab
				603	392 funcresolve - Number of varargs functions resolved
				604	27 globaldce - Number of global variables removed
				605	2 adce - Number of basic blocks removed
				606	134 cee - Number of branches revectored
				607	49 cee - Number of setcc instruction eliminated
				608	532 gcse - Number of loads removed
				609	2919 gcse - Number of instructions removed
				610	86 indvars - Number of canonical indvars added
				611	87 indvars - Number of aux indvars removed
				612	25 instcombine - Number of dead inst eliminate
				613	434 instcombine - Number of insts combined
				614	248 licm - Number of load insts hoisted
				615	1298 licm - Number of insts hoisted to a loop pre-header
				616	3 licm - Number of insts hoisted to multiple loop preds (bad, no loop pre-header)
				617	75 mem2reg - Number of alloca's promoted
				618	1444 cfgsimplify - Number of blocks simplified
				619	</pre>
				620	</div>
				621
				622	<p>Obviously, with so many optimizations, having a unified framework for this
				623	stuff is very nice. Making your pass fit well into the framework makes it more
				624	maintainable and useful.</p>
				625
				626	</div>
				627
				628	<!-- ======================================================================= -->
				629	<div class="doc_subsection">
				630	<a name="ViewGraph">Viewing graphs while debugging code</a>
				631	</div>
				632
				633	<div class="doc_text">
				634
				635	<p>Several of the important data structures in LLVM are graphs: for example
				636	CFGs made out of LLVM <a href="#BasicBlock">BasicBlock</a>s, CFGs made out of
				637	LLVM <a href="CodeGenerator.html#machinebasicblock">MachineBasicBlock</a>s, and
				638	<a href="CodeGenerator.html#selectiondag_intro">Instruction Selection
				639	DAGs</a>. In many cases, while debugging various parts of the compiler, it is
				640	nice to instantly visualize these graphs.</p>
				641
				642	<p>LLVM provides several callbacks that are available in a debug build to do
				643	exactly that. If you call the <tt>Function::viewCFG()</tt> method, for example,
				644	the current LLVM tool will pop up a window containing the CFG for the function
				645	where each basic block is a node in the graph, and each node contains the
				646	instructions in the block. Similarly, there also exists
				647	<tt>Function::viewCFGOnly()</tt> (does not include the instructions), the
				648	<tt>MachineFunction::viewCFG()</tt> and <tt>MachineFunction::viewCFGOnly()</tt>,
				649	and the <tt>SelectionDAG::viewGraph()</tt> methods. Within GDB, for example,
				650	you can usually use something like <tt>call DAG.viewGraph()</tt> to pop
				651	up a window. Alternatively, you can sprinkle calls to these functions in your
				652	code in places you want to debug.</p>
				653
				654	<p>Getting this to work requires a small amount of configuration. On Unix
				655	systems with X11, install the <a href="http://www.graphviz.org">graphviz</a>
				656	toolkit, and make sure 'dot' and 'gv' are in your path. If you are running on
				657	Mac OS/X, download and install the Mac OS/X <a
				658	href="http://www.pixelglow.com/graphviz/">Graphviz program</a>, and add
				659	<tt>/Applications/Graphviz.app/Contents/MacOS/</tt> (or wherever you install
				660	it) to your path. Once in your system and path are set up, rerun the LLVM
				661	configure script and rebuild LLVM to enable this functionality.</p>
				662
				663	<p><tt>SelectionDAG</tt> has been extended to make it easier to locate
				664	<i>interesting</i> nodes in large complex graphs. From gdb, if you
				665	<tt>call DAG.setGraphColor(<i>node</i>, "<i>color</i>")</tt>, then the
				666	next <tt>call DAG.viewGraph()</tt> would highlight the node in the
				667	specified color (choices of colors can be found at <a
				668	href="http://www.graphviz.org/doc/info/colors.html">colors</a>.) More
				669	complex node attributes can be provided with <tt>call
				670	DAG.setGraphAttrs(<i>node</i>, "<i>attributes</i>")</tt> (choices can be
				671	found at <a href="http://www.graphviz.org/doc/info/attrs.html">Graph
				672	Attributes</a>.) If you want to restart and clear all the current graph
				673	attributes, then you can <tt>call DAG.clearGraphAttrs()</tt>. </p>
				674
				675	</div>
				676
				677	<!-- *********************************************************************** -->
				678	<div class="doc_section">
				679	<a name="datastructure">Picking the Right Data Structure for a Task</a>
				680	</div>
				681	<!-- *********************************************************************** -->
				682
				683	<div class="doc_text">
				684
				685	<p>LLVM has a plethora of data structures in the <tt>llvm/ADT/</tt> directory,
				686	and we commonly use STL data structures. This section describes the trade-offs
				687	you should consider when you pick one.</p>
				688
				689	<p>
				690	The first step is a choose your own adventure: do you want a sequential
				691	container, a set-like container, or a map-like container? The most important
				692	thing when choosing a container is the algorithmic properties of how you plan to
				693	access the container. Based on that, you should use:</p>
				694
				695	<ul>
				696	<li>a <a href="#ds_map">map-like</a> container if you need efficient look-up
				697	of an value based on another value. Map-like containers also support
				698	efficient queries for containment (whether a key is in the map). Map-like
				699	containers generally do not support efficient reverse mapping (values to
				700	keys). If you need that, use two maps. Some map-like containers also
				701	support efficient iteration through the keys in sorted order. Map-like
				702	containers are the most expensive sort, only use them if you need one of
				703	these capabilities.</li>
				704
				705	<li>a <a href="#ds_set">set-like</a> container if you need to put a bunch of
				706	stuff into a container that automatically eliminates duplicates. Some
				707	set-like containers support efficient iteration through the elements in
				708	sorted order. Set-like containers are more expensive than sequential
				709	containers.
				710	</li>
				711
				712	<li>a <a href="#ds_sequential">sequential</a> container provides
				713	the most efficient way to add elements and keeps track of the order they are
				714	added to the collection. They permit duplicates and support efficient
				715	iteration, but do not support efficient look-up based on a key.
				716	</li>
				717
Daniel Berlin	7ea44dc	2007-09-24 17:52:25 +0000	[diff] [blame]	718	<li>a <a href="#ds_bit">bit</a> container provides an efficient way to store and
				719	perform set operations on sets of numeric id's, while automatically
				720	eliminating duplicates. Bit containers require a maximum of 1 bit for each
				721	identifier you want to store.
				722	</li>
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	723	</ul>
				724
				725	<p>
				726	Once the proper category of container is determined, you can fine tune the
				727	memory use, constant factors, and cache behaviors of access by intelligently
				728	picking a member of the category. Note that constant factors and cache behavior
				729	can be a big deal. If you have a vector that usually only contains a few
				730	elements (but could contain many), for example, it's much better to use
				731	<a href="#dss_smallvector">SmallVector</a> than <a href="#dss_vector">vector</a>
				732	. Doing so avoids (relatively) expensive malloc/free calls, which dwarf the
				733	cost of adding the elements to the container. </p>
				734
				735	</div>
				736
				737	<!-- ======================================================================= -->
				738	<div class="doc_subsection">
				739	<a name="ds_sequential">Sequential Containers (std::vector, std::list, etc)</a>
				740	</div>
				741
				742	<div class="doc_text">
				743	There are a variety of sequential containers available for you, based on your
				744	needs. Pick the first in this section that will do what you want.
				745	</div>
				746
				747	<!-- _______________________________________________________________________ -->
				748	<div class="doc_subsubsection">
				749	<a name="dss_fixedarrays">Fixed Size Arrays</a>
				750	</div>
				751
				752	<div class="doc_text">
				753	<p>Fixed size arrays are very simple and very fast. They are good if you know
				754	exactly how many elements you have, or you have a (low) upper bound on how many
				755	you have.</p>
				756	</div>
				757
				758	<!-- _______________________________________________________________________ -->
				759	<div class="doc_subsubsection">
				760	<a name="dss_heaparrays">Heap Allocated Arrays</a>
				761	</div>
				762
				763	<div class="doc_text">
				764	<p>Heap allocated arrays (new[] + delete[]) are also simple. They are good if
				765	the number of elements is variable, if you know how many elements you will need
				766	before the array is allocated, and if the array is usually large (if not,
				767	consider a <a href="#dss_smallvector">SmallVector</a>). The cost of a heap
				768	allocated array is the cost of the new/delete (aka malloc/free). Also note that
				769	if you are allocating an array of a type with a constructor, the constructor and
				770	destructors will be run for every element in the array (re-sizable vectors only
				771	construct those elements actually used).</p>
				772	</div>
				773
				774	<!-- _______________________________________________________________________ -->
				775	<div class="doc_subsubsection">
				776	<a name="dss_smallvector">"llvm/ADT/SmallVector.h"</a>
				777	</div>
				778
				779	<div class="doc_text">
				780	<p><tt>SmallVector<Type, N></tt> is a simple class that looks and smells
				781	just like <tt>vector<Type></tt>:
				782	it supports efficient iteration, lays out elements in memory order (so you can
				783	do pointer arithmetic between elements), supports efficient push_back/pop_back
				784	operations, supports efficient random access to its elements, etc.</p>
				785
				786	<p>The advantage of SmallVector is that it allocates space for
				787	some number of elements (N) <b>in the object itself</b>. Because of this, if
				788	the SmallVector is dynamically smaller than N, no malloc is performed. This can
				789	be a big win in cases where the malloc/free call is far more expensive than the
				790	code that fiddles around with the elements.</p>
				791
				792	<p>This is good for vectors that are "usually small" (e.g. the number of
				793	predecessors/successors of a block is usually less than 8). On the other hand,
				794	this makes the size of the SmallVector itself large, so you don't want to
				795	allocate lots of them (doing so will waste a lot of space). As such,
				796	SmallVectors are most useful when on the stack.</p>
				797
				798	<p>SmallVector also provides a nice portable and efficient replacement for
				799	<tt>alloca</tt>.</p>
				800
				801	</div>
				802
				803	<!-- _______________________________________________________________________ -->
				804	<div class="doc_subsubsection">
				805	<a name="dss_vector"><vector></a>
				806	</div>
				807
				808	<div class="doc_text">
				809	<p>
				810	std::vector is well loved and respected. It is useful when SmallVector isn't:
				811	when the size of the vector is often large (thus the small optimization will
				812	rarely be a benefit) or if you will be allocating many instances of the vector
				813	itself (which would waste space for elements that aren't in the container).
				814	vector is also useful when interfacing with code that expects vectors :).
				815	</p>
				816
				817	<p>One worthwhile note about std::vector: avoid code like this:</p>
				818
				819	<div class="doc_code">
				820	<pre>
				821	for ( ... ) {
				822	std::vector<foo> V;
				823	use V;
				824	}
				825	</pre>
				826	</div>
				827
				828	<p>Instead, write this as:</p>
				829
				830	<div class="doc_code">
				831	<pre>
				832	std::vector<foo> V;
				833	for ( ... ) {
				834	use V;
				835	V.clear();
				836	}
				837	</pre>
				838	</div>
				839
				840	<p>Doing so will save (at least) one heap allocation and free per iteration of
				841	the loop.</p>
				842
				843	</div>
				844
				845	<!-- _______________________________________________________________________ -->
				846	<div class="doc_subsubsection">
				847	<a name="dss_deque"><deque></a>
				848	</div>
				849
				850	<div class="doc_text">
				851	<p>std::deque is, in some senses, a generalized version of std::vector. Like
				852	std::vector, it provides constant time random access and other similar
				853	properties, but it also provides efficient access to the front of the list. It
				854	does not guarantee continuity of elements within memory.</p>
				855
				856	<p>In exchange for this extra flexibility, std::deque has significantly higher
				857	constant factor costs than std::vector. If possible, use std::vector or
				858	something cheaper.</p>
				859	</div>
				860
				861	<!-- _______________________________________________________________________ -->
				862	<div class="doc_subsubsection">
				863	<a name="dss_list"><list></a>
				864	</div>
				865
				866	<div class="doc_text">
				867	<p>std::list is an extremely inefficient class that is rarely useful.
				868	It performs a heap allocation for every element inserted into it, thus having an
				869	extremely high constant factor, particularly for small data types. std::list
				870	also only supports bidirectional iteration, not random access iteration.</p>
				871
				872	<p>In exchange for this high cost, std::list supports efficient access to both
				873	ends of the list (like std::deque, but unlike std::vector or SmallVector). In
				874	addition, the iterator invalidation characteristics of std::list are stronger
				875	than that of a vector class: inserting or removing an element into the list does
				876	not invalidate iterator or pointers to other elements in the list.</p>
				877	</div>
				878
				879	<!-- _______________________________________________________________________ -->
				880	<div class="doc_subsubsection">
Gabor Greif	bb17f65	2009-02-27 11:37:41 +0000	[diff] [blame]	881	<a name="dss_ilist">llvm/ADT/ilist.h</a>
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	882	</div>
				883
				884	<div class="doc_text">
				885	<p><tt>ilist<T></tt> implements an 'intrusive' doubly-linked list. It is
				886	intrusive, because it requires the element to store and provide access to the
				887	prev/next pointers for the list.</p>
				888
Gabor Greif	b6b21ec	2009-02-27 12:02:19 +0000	[diff] [blame]	889	<p><tt>ilist</tt> has the same drawbacks as <tt>std::list</tt>, and additionally
				890	requires an <tt>ilist_traits</tt> implementation for the element type, but it
				891	provides some novel characteristics. In particular, it can efficiently store
				892	polymorphic objects, the traits class is informed when an element is inserted or
				893	removed from the list, and ilists are guaranteed to support a constant-time splice
				894	operation.</p>
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	895
Gabor Greif	b6b21ec	2009-02-27 12:02:19 +0000	[diff] [blame]	896	<p>These properties are exactly what we want for things like <tt>Instruction</tt>s
				897	and basic blocks, which is why these are implemented with <tt>ilist</tt>s.</p>
Gabor Greif	bb17f65	2009-02-27 11:37:41 +0000	[diff] [blame]	898
				899	Related classes of interest are explained in the following subsections:
				900	<ul>
Gabor Greif	a17abe1	2009-02-27 13:28:07 +0000	[diff] [blame]	901	<li><a href="#dss_ilist_traits">ilist_traits</a></li>
Gabor Greif	b6b21ec	2009-02-27 12:02:19 +0000	[diff] [blame]	902	<li><a href="#dss_iplist">iplist</a></li>
Gabor Greif	bb17f65	2009-02-27 11:37:41 +0000	[diff] [blame]	903	<li><a href="#dss_ilist_node">llvm/ADT/ilist_node.h</a></li>
				904	</ul>
				905	</div>
				906
				907	<!-- _______________________________________________________________________ -->
				908	<div class="doc_subsubsection">
Gabor Greif	a17abe1	2009-02-27 13:28:07 +0000	[diff] [blame]	909	<a name="dss_ilist_traits">ilist_traits</a>
				910	</div>
				911
				912	<div class="doc_text">
				913	<p><tt>ilist_traits<T></tt> is <tt>ilist<T></tt>'s customization
				914	mechanism. <tt>iplist<T></tt> (and consequently <tt>ilist<T></tt>)
				915	publicly derive from this traits class.</p>
				916	</div>
				917
				918	<!-- _______________________________________________________________________ -->
				919	<div class="doc_subsubsection">
Gabor Greif	b6b21ec	2009-02-27 12:02:19 +0000	[diff] [blame]	920	<a name="dss_iplist">iplist</a>
				921	</div>
				922
				923	<div class="doc_text">
				924	<p><tt>iplist<T></tt> is <tt>ilist<T></tt>'s base and as such
				925	supports a slightly narrower interface. Notably, inserters from <tt>T&</tt>
				926	are absent.</p>
Gabor Greif	a17abe1	2009-02-27 13:28:07 +0000	[diff] [blame]	927
				928	<p><tt>ilist_traits<T></tt> is a public base of this class and can be
				929	used for a wide variety of customizations.</p>
Gabor Greif	b6b21ec	2009-02-27 12:02:19 +0000	[diff] [blame]	930	</div>
				931
				932	<!-- _______________________________________________________________________ -->
				933	<div class="doc_subsubsection">
Gabor Greif	bb17f65	2009-02-27 11:37:41 +0000	[diff] [blame]	934	<a name="dss_ilist_node">llvm/ADT/ilist_node.h</a>
				935	</div>
				936
				937	<div class="doc_text">
				938	<p><tt>ilist_node<T></tt> implements a the forward and backward links
				939	that are expected by the <tt>ilist<T></tt> (and analogous containers)
				940	in the default manner.</p>
				941
				942	<p><tt>ilist_node<T></tt>s are meant to be embedded in the node type
				943	<tt>T</tt>.</p>
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	944	</div>
				945
				946	<!-- _______________________________________________________________________ -->
				947	<div class="doc_subsubsection">
				948	<a name="dss_other">Other Sequential Container options</a>
				949	</div>
				950
				951	<div class="doc_text">
				952	<p>Other STL containers are available, such as std::string.</p>
				953
				954	<p>There are also various STL adapter classes such as std::queue,
				955	std::priority_queue, std::stack, etc. These provide simplified access to an
				956	underlying container but don't affect the cost of the container itself.</p>
				957
				958	</div>
				959
				960
				961	<!-- ======================================================================= -->
				962	<div class="doc_subsection">
				963	<a name="ds_set">Set-Like Containers (std::set, SmallSet, SetVector, etc)</a>
				964	</div>
				965
				966	<div class="doc_text">
				967
				968	<p>Set-like containers are useful when you need to canonicalize multiple values
				969	into a single representation. There are several different choices for how to do
				970	this, providing various trade-offs.</p>
				971
				972	</div>
				973
				974
				975	<!-- _______________________________________________________________________ -->
				976	<div class="doc_subsubsection">
				977	<a name="dss_sortedvectorset">A sorted 'vector'</a>
				978	</div>
				979
				980	<div class="doc_text">
				981
				982	<p>If you intend to insert a lot of elements, then do a lot of queries, a
				983	great approach is to use a vector (or other sequential container) with
				984	std::sort+std::unique to remove duplicates. This approach works really well if
				985	your usage pattern has these two distinct phases (insert then query), and can be
				986	coupled with a good choice of <a href="#ds_sequential">sequential container</a>.
				987	</p>
				988
				989	<p>
				990	This combination provides the several nice properties: the result data is
				991	contiguous in memory (good for cache locality), has few allocations, is easy to
				992	address (iterators in the final vector are just indices or pointers), and can be
				993	efficiently queried with a standard binary or radix search.</p>
				994
				995	</div>
				996
				997	<!-- _______________________________________________________________________ -->
				998	<div class="doc_subsubsection">
				999	<a name="dss_smallset">"llvm/ADT/SmallSet.h"</a>
				1000	</div>
				1001
				1002	<div class="doc_text">
				1003
				1004	<p>If you have a set-like data structure that is usually small and whose elements
				1005	are reasonably small, a <tt>SmallSet<Type, N></tt> is a good choice. This set
				1006	has space for N elements in place (thus, if the set is dynamically smaller than
				1007	N, no malloc traffic is required) and accesses them with a simple linear search.
				1008	When the set grows beyond 'N' elements, it allocates a more expensive representation that
				1009	guarantees efficient access (for most types, it falls back to std::set, but for
				1010	pointers it uses something far better, <a
				1011	href="#dss_smallptrset">SmallPtrSet</a>).</p>
				1012
				1013	<p>The magic of this class is that it handles small sets extremely efficiently,
				1014	but gracefully handles extremely large sets without loss of efficiency. The
				1015	drawback is that the interface is quite small: it supports insertion, queries
				1016	and erasing, but does not support iteration.</p>
				1017
				1018	</div>
				1019
				1020	<!-- _______________________________________________________________________ -->
				1021	<div class="doc_subsubsection">
				1022	<a name="dss_smallptrset">"llvm/ADT/SmallPtrSet.h"</a>
				1023	</div>
				1024
				1025	<div class="doc_text">
				1026
				1027	<p>SmallPtrSet has all the advantages of SmallSet (and a SmallSet of pointers is
				1028	transparently implemented with a SmallPtrSet), but also supports iterators. If
				1029	more than 'N' insertions are performed, a single quadratically
				1030	probed hash table is allocated and grows as needed, providing extremely
				1031	efficient access (constant time insertion/deleting/queries with low constant
				1032	factors) and is very stingy with malloc traffic.</p>
				1033
				1034	<p>Note that, unlike std::set, the iterators of SmallPtrSet are invalidated
				1035	whenever an insertion occurs. Also, the values visited by the iterators are not
				1036	visited in sorted order.</p>
				1037
				1038	</div>
				1039
				1040	<!-- _______________________________________________________________________ -->
				1041	<div class="doc_subsubsection">
Chris Lattner	77ab46d	2007-09-30 00:58:59 +0000	[diff] [blame]	1042	<a name="dss_denseset">"llvm/ADT/DenseSet.h"</a>
				1043	</div>
				1044
				1045	<div class="doc_text">
				1046
				1047	<p>
				1048	DenseSet is a simple quadratically probed hash table. It excels at supporting
				1049	small values: it uses a single allocation to hold all of the pairs that
				1050	are currently inserted in the set. DenseSet is a great way to unique small
				1051	values that are not simple pointers (use <a
				1052	href="#dss_smallptrset">SmallPtrSet</a> for pointers). Note that DenseSet has
				1053	the same requirements for the value type that <a
				1054	href="#dss_densemap">DenseMap</a> has.
				1055	</p>
				1056
				1057	</div>
				1058
				1059	<!-- _______________________________________________________________________ -->
				1060	<div class="doc_subsubsection">
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	1061	<a name="dss_FoldingSet">"llvm/ADT/FoldingSet.h"</a>
				1062	</div>
				1063
				1064	<div class="doc_text">
				1065
				1066	<p>
				1067	FoldingSet is an aggregate class that is really good at uniquing
				1068	expensive-to-create or polymorphic objects. It is a combination of a chained
				1069	hash table with intrusive links (uniqued objects are required to inherit from
				1070	FoldingSetNode) that uses <a href="#dss_smallvector">SmallVector</a> as part of
				1071	its ID process.</p>
				1072
				1073	<p>Consider a case where you want to implement a "getOrCreateFoo" method for
				1074	a complex object (for example, a node in the code generator). The client has a
				1075	description of what it wants to generate (it knows the opcode and all the
				1076	operands), but we don't want to 'new' a node, then try inserting it into a set
				1077	only to find out it already exists, at which point we would have to delete it
				1078	and return the node that already exists.
				1079	</p>
				1080
				1081	<p>To support this style of client, FoldingSet perform a query with a
				1082	FoldingSetNodeID (which wraps SmallVector) that can be used to describe the
				1083	element that we want to query for. The query either returns the element
				1084	matching the ID or it returns an opaque ID that indicates where insertion should
				1085	take place. Construction of the ID usually does not require heap traffic.</p>
				1086
				1087	<p>Because FoldingSet uses intrusive links, it can support polymorphic objects
				1088	in the set (for example, you can have SDNode instances mixed with LoadSDNodes).
				1089	Because the elements are individually allocated, pointers to the elements are
				1090	stable: inserting or removing elements does not invalidate any pointers to other
				1091	elements.
				1092	</p>
				1093
				1094	</div>
				1095
				1096	<!-- _______________________________________________________________________ -->
				1097	<div class="doc_subsubsection">
				1098	<a name="dss_set"><set></a>
				1099	</div>
				1100
				1101	<div class="doc_text">
				1102
				1103	<p><tt>std::set</tt> is a reasonable all-around set class, which is decent at
				1104	many things but great at nothing. std::set allocates memory for each element
				1105	inserted (thus it is very malloc intensive) and typically stores three pointers
				1106	per element in the set (thus adding a large amount of per-element space
				1107	overhead). It offers guaranteed log(n) performance, which is not particularly
				1108	fast from a complexity standpoint (particularly if the elements of the set are
				1109	expensive to compare, like strings), and has extremely high constant factors for
				1110	lookup, insertion and removal.</p>
				1111
				1112	<p>The advantages of std::set are that its iterators are stable (deleting or
				1113	inserting an element from the set does not affect iterators or pointers to other
				1114	elements) and that iteration over the set is guaranteed to be in sorted order.
				1115	If the elements in the set are large, then the relative overhead of the pointers
				1116	and malloc traffic is not a big deal, but if the elements of the set are small,
				1117	std::set is almost never a good choice.</p>
				1118
				1119	</div>
				1120
				1121	<!-- _______________________________________________________________________ -->
				1122	<div class="doc_subsubsection">
				1123	<a name="dss_setvector">"llvm/ADT/SetVector.h"</a>
				1124	</div>
				1125
				1126	<div class="doc_text">
				1127	<p>LLVM's SetVector<Type> is an adapter class that combines your choice of
				1128	a set-like container along with a <a href="#ds_sequential">Sequential
				1129	Container</a>. The important property
				1130	that this provides is efficient insertion with uniquing (duplicate elements are
				1131	ignored) with iteration support. It implements this by inserting elements into
				1132	both a set-like container and the sequential container, using the set-like
				1133	container for uniquing and the sequential container for iteration.
				1134	</p>
				1135
				1136	<p>The difference between SetVector and other sets is that the order of
				1137	iteration is guaranteed to match the order of insertion into the SetVector.
				1138	This property is really important for things like sets of pointers. Because
				1139	pointer values are non-deterministic (e.g. vary across runs of the program on
				1140	different machines), iterating over the pointers in the set will
				1141	not be in a well-defined order.</p>
				1142
				1143	<p>
				1144	The drawback of SetVector is that it requires twice as much space as a normal
				1145	set and has the sum of constant factors from the set-like container and the
				1146	sequential container that it uses. Use it only if you need to iterate over
				1147	the elements in a deterministic order. SetVector is also expensive to delete
				1148	elements out of (linear time), unless you use it's "pop_back" method, which is
				1149	faster.
				1150	</p>
				1151
				1152	<p>SetVector is an adapter class that defaults to using std::vector and std::set
				1153	for the underlying containers, so it is quite expensive. However,
				1154	<tt>"llvm/ADT/SetVector.h"</tt> also provides a SmallSetVector class, which
				1155	defaults to using a SmallVector and SmallSet of a specified size. If you use
				1156	this, and if your sets are dynamically smaller than N, you will save a lot of
				1157	heap traffic.</p>
				1158
				1159	</div>
				1160
				1161	<!-- _______________________________________________________________________ -->
				1162	<div class="doc_subsubsection">
				1163	<a name="dss_uniquevector">"llvm/ADT/UniqueVector.h"</a>
				1164	</div>
				1165
				1166	<div class="doc_text">
				1167
				1168	<p>
				1169	UniqueVector is similar to <a href="#dss_setvector">SetVector</a>, but it
				1170	retains a unique ID for each element inserted into the set. It internally
				1171	contains a map and a vector, and it assigns a unique ID for each value inserted
				1172	into the set.</p>
				1173
				1174	<p>UniqueVector is very expensive: its cost is the sum of the cost of
				1175	maintaining both the map and vector, it has high complexity, high constant
				1176	factors, and produces a lot of malloc traffic. It should be avoided.</p>
				1177
				1178	</div>
				1179
				1180
				1181	<!-- _______________________________________________________________________ -->
				1182	<div class="doc_subsubsection">
				1183	<a name="dss_otherset">Other Set-Like Container Options</a>
				1184	</div>
				1185
				1186	<div class="doc_text">
				1187
				1188	<p>
				1189	The STL provides several other options, such as std::multiset and the various
				1190	"hash_set" like containers (whether from C++ TR1 or from the SGI library).</p>
				1191
				1192	<p>std::multiset is useful if you're not interested in elimination of
				1193	duplicates, but has all the drawbacks of std::set. A sorted vector (where you
				1194	don't delete duplicate entries) or some other approach is almost always
				1195	better.</p>
				1196
				1197	<p>The various hash_set implementations (exposed portably by
				1198	"llvm/ADT/hash_set") is a simple chained hashtable. This algorithm is as malloc
				1199	intensive as std::set (performing an allocation for each element inserted,
				1200	thus having really high constant factors) but (usually) provides O(1)
				1201	insertion/deletion of elements. This can be useful if your elements are large
				1202	(thus making the constant-factor cost relatively low) or if comparisons are
				1203	expensive. Element iteration does not visit elements in a useful order.</p>
				1204
				1205	</div>
				1206
				1207	<!-- ======================================================================= -->
				1208	<div class="doc_subsection">
				1209	<a name="ds_map">Map-Like Containers (std::map, DenseMap, etc)</a>
				1210	</div>
				1211
				1212	<div class="doc_text">
				1213	Map-like containers are useful when you want to associate data to a key. As
				1214	usual, there are a lot of different ways to do this. :)
				1215	</div>
				1216
				1217	<!-- _______________________________________________________________________ -->
				1218	<div class="doc_subsubsection">
				1219	<a name="dss_sortedvectormap">A sorted 'vector'</a>
				1220	</div>
				1221
				1222	<div class="doc_text">
				1223
				1224	<p>
				1225	If your usage pattern follows a strict insert-then-query approach, you can
				1226	trivially use the same approach as <a href="#dss_sortedvectorset">sorted vectors
				1227	for set-like containers</a>. The only difference is that your query function
				1228	(which uses std::lower_bound to get efficient log(n) lookup) should only compare
				1229	the key, not both the key and value. This yields the same advantages as sorted
				1230	vectors for sets.
				1231	</p>
				1232	</div>
				1233
				1234	<!-- _______________________________________________________________________ -->
				1235	<div class="doc_subsubsection">
				1236	<a name="dss_stringmap">"llvm/ADT/StringMap.h"</a>
				1237	</div>
				1238
				1239	<div class="doc_text">
				1240
				1241	<p>
				1242	Strings are commonly used as keys in maps, and they are difficult to support
				1243	efficiently: they are variable length, inefficient to hash and compare when
				1244	long, expensive to copy, etc. StringMap is a specialized container designed to
				1245	cope with these issues. It supports mapping an arbitrary range of bytes to an
				1246	arbitrary other object.</p>
				1247
				1248	<p>The StringMap implementation uses a quadratically-probed hash table, where
				1249	the buckets store a pointer to the heap allocated entries (and some other
				1250	stuff). The entries in the map must be heap allocated because the strings are
				1251	variable length. The string data (key) and the element object (value) are
				1252	stored in the same allocation with the string data immediately after the element
				1253	object. This container guarantees the "<tt>(char*)(&Value+1)</tt>" points
				1254	to the key string for a value.</p>
				1255
				1256	<p>The StringMap is very fast for several reasons: quadratic probing is very
				1257	cache efficient for lookups, the hash value of strings in buckets is not
				1258	recomputed when lookup up an element, StringMap rarely has to touch the
				1259	memory for unrelated objects when looking up a value (even when hash collisions
				1260	happen), hash table growth does not recompute the hash values for strings
				1261	already in the table, and each pair in the map is store in a single allocation
				1262	(the string data is stored in the same allocation as the Value of a pair).</p>
				1263
				1264	<p>StringMap also provides query methods that take byte ranges, so it only ever
				1265	copies a string if a value is inserted into the table.</p>
				1266	</div>
				1267
				1268	<!-- _______________________________________________________________________ -->
				1269	<div class="doc_subsubsection">
				1270	<a name="dss_indexedmap">"llvm/ADT/IndexedMap.h"</a>
				1271	</div>
				1272
				1273	<div class="doc_text">
				1274	<p>
				1275	IndexedMap is a specialized container for mapping small dense integers (or
				1276	values that can be mapped to small dense integers) to some other type. It is
				1277	internally implemented as a vector with a mapping function that maps the keys to
				1278	the dense integer range.
				1279	</p>
				1280
				1281	<p>
				1282	This is useful for cases like virtual registers in the LLVM code generator: they
				1283	have a dense mapping that is offset by a compile-time constant (the first
				1284	virtual register ID).</p>
				1285
				1286	</div>
				1287
				1288	<!-- _______________________________________________________________________ -->
				1289	<div class="doc_subsubsection">
				1290	<a name="dss_densemap">"llvm/ADT/DenseMap.h"</a>
				1291	</div>
				1292
				1293	<div class="doc_text">
				1294
				1295	<p>
				1296	DenseMap is a simple quadratically probed hash table. It excels at supporting
				1297	small keys and values: it uses a single allocation to hold all of the pairs that
				1298	are currently inserted in the map. DenseMap is a great way to map pointers to
				1299	pointers, or map other small types to each other.
				1300	</p>
				1301
				1302	<p>
				1303	There are several aspects of DenseMap that you should be aware of, however. The
				1304	iterators in a densemap are invalidated whenever an insertion occurs, unlike
				1305	map. Also, because DenseMap allocates space for a large number of key/value
				1306	pairs (it starts with 64 by default), it will waste a lot of space if your keys
				1307	or values are large. Finally, you must implement a partial specialization of
Chris Lattner	92eea07	2007-09-17 18:34:04 +0000	[diff] [blame]	1308	DenseMapInfo for the key that you want, if it isn't already supported. This
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	1309	is required to tell DenseMap about two special marker values (which can never be
				1310	inserted into the map) that it needs internally.</p>
				1311
				1312	</div>
				1313
				1314	<!-- _______________________________________________________________________ -->
				1315	<div class="doc_subsubsection">
				1316	<a name="dss_map"><map></a>
				1317	</div>
				1318
				1319	<div class="doc_text">
				1320
				1321	<p>
				1322	std::map has similar characteristics to <a href="#dss_set">std::set</a>: it uses
				1323	a single allocation per pair inserted into the map, it offers log(n) lookup with
				1324	an extremely large constant factor, imposes a space penalty of 3 pointers per
				1325	pair in the map, etc.</p>
				1326
				1327	<p>std::map is most useful when your keys or values are very large, if you need
				1328	to iterate over the collection in sorted order, or if you need stable iterators
				1329	into the map (i.e. they don't get invalidated if an insertion or deletion of
				1330	another element takes place).</p>
				1331
				1332	</div>
				1333
				1334	<!-- _______________________________________________________________________ -->
				1335	<div class="doc_subsubsection">
				1336	<a name="dss_othermap">Other Map-Like Container Options</a>
				1337	</div>
				1338
				1339	<div class="doc_text">
				1340
				1341	<p>
				1342	The STL provides several other options, such as std::multimap and the various
				1343	"hash_map" like containers (whether from C++ TR1 or from the SGI library).</p>
				1344
				1345	<p>std::multimap is useful if you want to map a key to multiple values, but has
				1346	all the drawbacks of std::map. A sorted vector or some other approach is almost
				1347	always better.</p>
				1348
				1349	<p>The various hash_map implementations (exposed portably by
				1350	"llvm/ADT/hash_map") are simple chained hash tables. This algorithm is as
				1351	malloc intensive as std::map (performing an allocation for each element
				1352	inserted, thus having really high constant factors) but (usually) provides O(1)
				1353	insertion/deletion of elements. This can be useful if your elements are large
				1354	(thus making the constant-factor cost relatively low) or if comparisons are
				1355	expensive. Element iteration does not visit elements in a useful order.</p>
				1356
				1357	</div>
				1358
Daniel Berlin	7ea44dc	2007-09-24 17:52:25 +0000	[diff] [blame]	1359	<!-- ======================================================================= -->
				1360	<div class="doc_subsection">
				1361	<a name="ds_bit">Bit storage containers (BitVector, SparseBitVector)</a>
				1362	</div>
				1363
				1364	<div class="doc_text">
Chris Lattner	62a4eae	2007-09-25 22:37:50 +0000	[diff] [blame]	1365	<p>Unlike the other containers, there are only two bit storage containers, and
				1366	choosing when to use each is relatively straightforward.</p>
				1367
				1368	<p>One additional option is
				1369	<tt>std::vector<bool></tt>: we discourage its use for two reasons 1) the
				1370	implementation in many common compilers (e.g. commonly available versions of
				1371	GCC) is extremely inefficient and 2) the C++ standards committee is likely to
				1372	deprecate this container and/or change it significantly somehow. In any case,
				1373	please don't use it.</p>
Daniel Berlin	7ea44dc	2007-09-24 17:52:25 +0000	[diff] [blame]	1374	</div>
				1375
				1376	<!-- _______________________________________________________________________ -->
				1377	<div class="doc_subsubsection">
				1378	<a name="dss_bitvector">BitVector</a>
				1379	</div>
				1380
				1381	<div class="doc_text">
				1382	<p> The BitVector container provides a fixed size set of bits for manipulation.
				1383	It supports individual bit setting/testing, as well as set operations. The set
				1384	operations take time O(size of bitvector), but operations are performed one word
				1385	at a time, instead of one bit at a time. This makes the BitVector very fast for
				1386	set operations compared to other containers. Use the BitVector when you expect
				1387	the number of set bits to be high (IE a dense set).
				1388	</p>
				1389	</div>
				1390
				1391	<!-- _______________________________________________________________________ -->
				1392	<div class="doc_subsubsection">
				1393	<a name="dss_sparsebitvector">SparseBitVector</a>
				1394	</div>
				1395
				1396	<div class="doc_text">
				1397	<p> The SparseBitVector container is much like BitVector, with one major
				1398	difference: Only the bits that are set, are stored. This makes the
				1399	SparseBitVector much more space efficient than BitVector when the set is sparse,
				1400	as well as making set operations O(number of set bits) instead of O(size of
				1401	universe). The downside to the SparseBitVector is that setting and testing of random bits is O(N), and on large SparseBitVectors, this can be slower than BitVector. In our implementation, setting or testing bits in sorted order
				1402	(either forwards or reverse) is O(1) worst case. Testing and setting bits within 128 bits (depends on size) of the current bit is also O(1). As a general statement, testing/setting bits in a SparseBitVector is O(distance away from last set bit).
				1403	</p>
				1404	</div>
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	1405
				1406	<!-- *********************************************************************** -->
				1407	<div class="doc_section">
				1408	<a name="common">Helpful Hints for Common Operations</a>
				1409	</div>
				1410	<!-- *********************************************************************** -->
				1411
				1412	<div class="doc_text">
				1413
				1414	<p>This section describes how to perform some very simple transformations of
				1415	LLVM code. This is meant to give examples of common idioms used, showing the
				1416	practical side of LLVM transformations. <p> Because this is a "how-to" section,
				1417	you should also read about the main classes that you will be working with. The
				1418	<a href="#coreclasses">Core LLVM Class Hierarchy Reference</a> contains details
				1419	and descriptions of the main classes that you should know about.</p>
				1420
				1421	</div>
				1422
				1423	<!-- NOTE: this section should be heavy on example code -->
				1424	<!-- ======================================================================= -->
				1425	<div class="doc_subsection">
				1426	<a name="inspection">Basic Inspection and Traversal Routines</a>
				1427	</div>
				1428
				1429	<div class="doc_text">
				1430
				1431	<p>The LLVM compiler infrastructure have many different data structures that may
				1432	be traversed. Following the example of the C++ standard template library, the
				1433	techniques used to traverse these various data structures are all basically the
				1434	same. For a enumerable sequence of values, the <tt>XXXbegin()</tt> function (or
				1435	method) returns an iterator to the start of the sequence, the <tt>XXXend()</tt>
				1436	function returns an iterator pointing to one past the last valid element of the
				1437	sequence, and there is some <tt>XXXiterator</tt> data type that is common
				1438	between the two operations.</p>
				1439
				1440	<p>Because the pattern for iteration is common across many different aspects of
				1441	the program representation, the standard template library algorithms may be used
				1442	on them, and it is easier to remember how to iterate. First we show a few common
				1443	examples of the data structures that need to be traversed. Other data
				1444	structures are traversed in very similar ways.</p>
				1445
				1446	</div>
				1447
				1448	<!-- _______________________________________________________________________ -->
				1449	<div class="doc_subsubsection">
				1450	<a name="iterate_function">Iterating over the </a><a
				1451	href="#BasicBlock"><tt>BasicBlock</tt></a>s in a <a
				1452	href="#Function"><tt>Function</tt></a>
				1453	</div>
				1454
				1455	<div class="doc_text">
				1456
				1457	<p>It's quite common to have a <tt>Function</tt> instance that you'd like to
				1458	transform in some way; in particular, you'd like to manipulate its
				1459	<tt>BasicBlock</tt>s. To facilitate this, you'll need to iterate over all of
				1460	the <tt>BasicBlock</tt>s that constitute the <tt>Function</tt>. The following is
				1461	an example that prints the name of a <tt>BasicBlock</tt> and the number of
				1462	<tt>Instruction</tt>s it contains:</p>
				1463
				1464	<div class="doc_code">
				1465	<pre>
				1466	// <i>func is a pointer to a Function instance</i>
				1467	for (Function::iterator i = func->begin(), e = func->end(); i != e; ++i)
				1468	// <i>Print out the name of the basic block if it has one, and then the</i>
				1469	// <i>number of instructions that it contains</i>
				1470	llvm::cerr << "Basic block (name=" << i->getName() << ") has "
				1471	<< i->size() << " instructions.\n";
				1472	</pre>
				1473	</div>
				1474
				1475	<p>Note that i can be used as if it were a pointer for the purposes of
				1476	invoking member functions of the <tt>Instruction</tt> class. This is
				1477	because the indirection operator is overloaded for the iterator
				1478	classes. In the above code, the expression <tt>i->size()</tt> is
				1479	exactly equivalent to <tt>(*i).size()</tt> just like you'd expect.</p>
				1480
				1481	</div>
				1482
				1483	<!-- _______________________________________________________________________ -->
				1484	<div class="doc_subsubsection">
				1485	<a name="iterate_basicblock">Iterating over the </a><a
				1486	href="#Instruction"><tt>Instruction</tt></a>s in a <a
				1487	href="#BasicBlock"><tt>BasicBlock</tt></a>
				1488	</div>
				1489
				1490	<div class="doc_text">
				1491
				1492	<p>Just like when dealing with <tt>BasicBlock</tt>s in <tt>Function</tt>s, it's
				1493	easy to iterate over the individual instructions that make up
				1494	<tt>BasicBlock</tt>s. Here's a code snippet that prints out each instruction in
				1495	a <tt>BasicBlock</tt>:</p>
				1496
				1497	<div class="doc_code">
				1498	<pre>
				1499	// <i>blk is a pointer to a BasicBlock instance</i>
				1500	for (BasicBlock::iterator i = blk->begin(), e = blk->end(); i != e; ++i)
				1501	// <i>The next statement works since operator<<(ostream&,...)</i>
				1502	// <i>is overloaded for Instruction&</i>
				1503	llvm::cerr << *i << "\n";
				1504	</pre>
				1505	</div>
				1506
				1507	<p>However, this isn't really the best way to print out the contents of a
				1508	<tt>BasicBlock</tt>! Since the ostream operators are overloaded for virtually
				1509	anything you'll care about, you could have just invoked the print routine on the
				1510	basic block itself: <tt>llvm::cerr << *blk << "\n";</tt>.</p>
				1511
				1512	</div>
				1513
				1514	<!-- _______________________________________________________________________ -->
				1515	<div class="doc_subsubsection">
				1516	<a name="iterate_institer">Iterating over the </a><a
				1517	href="#Instruction"><tt>Instruction</tt></a>s in a <a
				1518	href="#Function"><tt>Function</tt></a>
				1519	</div>
				1520
				1521	<div class="doc_text">
				1522
				1523	<p>If you're finding that you commonly iterate over a <tt>Function</tt>'s
				1524	<tt>BasicBlock</tt>s and then that <tt>BasicBlock</tt>'s <tt>Instruction</tt>s,
				1525	<tt>InstIterator</tt> should be used instead. You'll need to include <a
				1526	href="/doxygen/InstIterator_8h-source.html"><tt>llvm/Support/InstIterator.h</tt></a>,
				1527	and then instantiate <tt>InstIterator</tt>s explicitly in your code. Here's a
				1528	small example that shows how to dump all instructions in a function to the standard error stream:<p>
				1529
				1530	<div class="doc_code">
				1531	<pre>
				1532	#include "<a href="/doxygen/InstIterator_8h-source.html">llvm/Support/InstIterator.h</a>"
				1533
				1534	// <i>F is a pointer to a Function instance</i>
Chris Lattner	fd62ee7	2008-06-04 18:20:42 +0000	[diff] [blame]	1535	for (inst_iterator I = inst_begin(F), E = inst_end(F); I != E; ++I)
				1536	llvm::cerr << *I << "\n";
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	1537	</pre>
				1538	</div>
				1539
				1540	<p>Easy, isn't it? You can also use <tt>InstIterator</tt>s to fill a
				1541	work list with its initial contents. For example, if you wanted to
				1542	initialize a work list to contain all instructions in a <tt>Function</tt>
				1543	F, all you would need to do is something like:</p>
				1544
				1545	<div class="doc_code">
				1546	<pre>
				1547	std::set<Instruction*> worklist;
Chris Lattner	fd62ee7	2008-06-04 18:20:42 +0000	[diff] [blame]	1548	// or better yet, SmallPtrSet<Instruction*, 64> worklist;
				1549
				1550	for (inst_iterator I = inst_begin(F), E = inst_end(F); I != E; ++I)
				1551	worklist.insert(&*I);
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	1552	</pre>
				1553	</div>
				1554
				1555	<p>The STL set <tt>worklist</tt> would now contain all instructions in the
				1556	<tt>Function</tt> pointed to by F.</p>
				1557
				1558	</div>
				1559
				1560	<!-- _______________________________________________________________________ -->
				1561	<div class="doc_subsubsection">
				1562	<a name="iterate_convert">Turning an iterator into a class pointer (and
				1563	vice-versa)</a>
				1564	</div>
				1565
				1566	<div class="doc_text">
				1567
				1568	<p>Sometimes, it'll be useful to grab a reference (or pointer) to a class
				1569	instance when all you've got at hand is an iterator. Well, extracting
				1570	a reference or a pointer from an iterator is very straight-forward.
				1571	Assuming that <tt>i</tt> is a <tt>BasicBlock::iterator</tt> and <tt>j</tt>
				1572	is a <tt>BasicBlock::const_iterator</tt>:</p>
				1573
				1574	<div class="doc_code">
				1575	<pre>
				1576	Instruction& inst = *i; // <i>Grab reference to instruction reference</i>
				1577	Instruction* pinst = &*i; // <i>Grab pointer to instruction reference</i>
				1578	const Instruction& inst = *j;
				1579	</pre>
				1580	</div>
				1581
				1582	<p>However, the iterators you'll be working with in the LLVM framework are
				1583	special: they will automatically convert to a ptr-to-instance type whenever they
				1584	need to. Instead of dereferencing the iterator and then taking the address of
				1585	the result, you can simply assign the iterator to the proper pointer type and
				1586	you get the dereference and address-of operation as a result of the assignment
				1587	(behind the scenes, this is a result of overloading casting mechanisms). Thus
				1588	the last line of the last example,</p>
				1589
				1590	<div class="doc_code">
				1591	<pre>
Chris Lattner	0665e1f	2008-01-03 16:56:04 +0000	[diff] [blame]	1592	Instruction pinst = &i;
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	1593	</pre>
				1594	</div>
				1595
				1596	<p>is semantically equivalent to</p>
				1597
				1598	<div class="doc_code">
				1599	<pre>
Chris Lattner	0665e1f	2008-01-03 16:56:04 +0000	[diff] [blame]	1600	Instruction *pinst = i;
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	1601	</pre>
				1602	</div>
				1603
				1604	<p>It's also possible to turn a class pointer into the corresponding iterator,
				1605	and this is a constant time operation (very efficient). The following code
				1606	snippet illustrates use of the conversion constructors provided by LLVM
				1607	iterators. By using these, you can explicitly grab the iterator of something
				1608	without actually obtaining it via iteration over some structure:</p>
				1609
				1610	<div class="doc_code">
				1611	<pre>
				1612	void printNextInstruction(Instruction* inst) {
				1613	BasicBlock::iterator it(inst);
				1614	++it; // <i>After this line, it refers to the instruction after *inst</i>
				1615	if (it != inst->getParent()->end()) llvm::cerr << *it << "\n";
				1616	}
				1617	</pre>
				1618	</div>
				1619
				1620	</div>
				1621
				1622	<!--_______________________________________________________________________-->
				1623	<div class="doc_subsubsection">
				1624	<a name="iterate_complex">Finding call sites: a slightly more complex
				1625	example</a>
				1626	</div>
				1627
				1628	<div class="doc_text">
				1629
				1630	<p>Say that you're writing a FunctionPass and would like to count all the
				1631	locations in the entire module (that is, across every <tt>Function</tt>) where a
				1632	certain function (i.e., some <tt>Function</tt>*) is already in scope. As you'll
				1633	learn later, you may want to use an <tt>InstVisitor</tt> to accomplish this in a
				1634	much more straight-forward manner, but this example will allow us to explore how
				1635	you'd do it if you didn't have <tt>InstVisitor</tt> around. In pseudo-code, this
				1636	is what we want to do:</p>
				1637
				1638	<div class="doc_code">
				1639	<pre>
				1640	initialize callCounter to zero
				1641	for each Function f in the Module
				1642	for each BasicBlock b in f
				1643	for each Instruction i in b
				1644	if (i is a CallInst and calls the given function)
				1645	increment callCounter
				1646	</pre>
				1647	</div>
				1648
				1649	<p>And the actual code is (remember, because we're writing a
				1650	<tt>FunctionPass</tt>, our <tt>FunctionPass</tt>-derived class simply has to
				1651	override the <tt>runOnFunction</tt> method):</p>
				1652
				1653	<div class="doc_code">
				1654	<pre>
				1655	Function* targetFunc = ...;
				1656
				1657	class OurFunctionPass : public FunctionPass {
				1658	public:
				1659	OurFunctionPass(): callCounter(0) { }
				1660
				1661	virtual runOnFunction(Function& F) {
				1662	for (Function::iterator b = F.begin(), be = F.end(); b != be; ++b) {
Eric Christopher	5fbecf7	2008-11-08 08:20:49 +0000	[diff] [blame]	1663	for (BasicBlock::iterator i = b->begin(), ie = b->end(); i != ie; ++i) {
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	1664	if (<a href="#CallInst">CallInst</a>* callInst = <a href="#isa">dyn_cast</a><<a
				1665	href="#CallInst">CallInst</a>>(&*i)) {
				1666	// <i>We know we've encountered a call instruction, so we</i>
				1667	// <i>need to determine if it's a call to the</i>
Chris Lattner	0665e1f	2008-01-03 16:56:04 +0000	[diff] [blame]	1668	// <i>function pointed to by m_func or not.</i>
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	1669	if (callInst->getCalledFunction() == targetFunc)
				1670	++callCounter;
				1671	}
				1672	}
				1673	}
				1674	}
				1675
				1676	private:
Chris Lattner	0665e1f	2008-01-03 16:56:04 +0000	[diff] [blame]	1677	unsigned callCounter;
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	1678	};
				1679	</pre>
				1680	</div>
				1681
				1682	</div>
				1683
				1684	<!--_______________________________________________________________________-->
				1685	<div class="doc_subsubsection">
				1686	<a name="calls_and_invokes">Treating calls and invokes the same way</a>
				1687	</div>
				1688
				1689	<div class="doc_text">
				1690
				1691	<p>You may have noticed that the previous example was a bit oversimplified in
				1692	that it did not deal with call sites generated by 'invoke' instructions. In
				1693	this, and in other situations, you may find that you want to treat
				1694	<tt>CallInst</tt>s and <tt>InvokeInst</tt>s the same way, even though their
				1695	most-specific common base class is <tt>Instruction</tt>, which includes lots of
				1696	less closely-related things. For these cases, LLVM provides a handy wrapper
				1697	class called <a
				1698	href="http://llvm.org/doxygen/classllvm_1_1CallSite.html"><tt>CallSite</tt></a>.
				1699	It is essentially a wrapper around an <tt>Instruction</tt> pointer, with some
				1700	methods that provide functionality common to <tt>CallInst</tt>s and
				1701	<tt>InvokeInst</tt>s.</p>
				1702
				1703	<p>This class has "value semantics": it should be passed by value, not by
				1704	reference and it should not be dynamically allocated or deallocated using
				1705	<tt>operator new</tt> or <tt>operator delete</tt>. It is efficiently copyable,
				1706	assignable and constructable, with costs equivalents to that of a bare pointer.
				1707	If you look at its definition, it has only a single pointer member.</p>
				1708
				1709	</div>
				1710
				1711	<!--_______________________________________________________________________-->
				1712	<div class="doc_subsubsection">
				1713	<a name="iterate_chains">Iterating over def-use & use-def chains</a>
				1714	</div>
				1715
				1716	<div class="doc_text">
				1717
				1718	<p>Frequently, we might have an instance of the <a
				1719	href="/doxygen/classllvm_1_1Value.html">Value Class</a> and we want to
				1720	determine which <tt>User</tt>s use the <tt>Value</tt>. The list of all
				1721	<tt>User</tt>s of a particular <tt>Value</tt> is called a <i>def-use</i> chain.
				1722	For example, let's say we have a <tt>Function*</tt> named <tt>F</tt> to a
				1723	particular function <tt>foo</tt>. Finding all of the instructions that
				1724	<i>use</i> <tt>foo</tt> is as simple as iterating over the <i>def-use</i> chain
				1725	of <tt>F</tt>:</p>
				1726
				1727	<div class="doc_code">
				1728	<pre>
Chris Lattner	0665e1f	2008-01-03 16:56:04 +0000	[diff] [blame]	1729	Function *F = ...;
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	1730
				1731	for (Value::use_iterator i = F->use_begin(), e = F->use_end(); i != e; ++i)
				1732	if (Instruction Inst = dyn_cast<Instruction>(i)) {
				1733	llvm::cerr << "F is used in instruction:\n";
				1734	llvm::cerr << *Inst << "\n";
				1735	}
				1736	</pre>
				1737	</div>
				1738
				1739	<p>Alternately, it's common to have an instance of the <a
				1740	href="/doxygen/classllvm_1_1User.html">User Class</a> and need to know what
				1741	<tt>Value</tt>s are used by it. The list of all <tt>Value</tt>s used by a
				1742	<tt>User</tt> is known as a <i>use-def</i> chain. Instances of class
				1743	<tt>Instruction</tt> are common <tt>User</tt>s, so we might want to iterate over
				1744	all of the values that a particular instruction uses (that is, the operands of
				1745	the particular <tt>Instruction</tt>):</p>
				1746
				1747	<div class="doc_code">
				1748	<pre>
Chris Lattner	0665e1f	2008-01-03 16:56:04 +0000	[diff] [blame]	1749	Instruction *pi = ...;
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	1750
				1751	for (User::op_iterator i = pi->op_begin(), e = pi->op_end(); i != e; ++i) {
Chris Lattner	0665e1f	2008-01-03 16:56:04 +0000	[diff] [blame]	1752	Value v = i;
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	1753	// <i>...</i>
				1754	}
				1755	</pre>
				1756	</div>
				1757
				1758	<!--
				1759	def-use chains ("finding all users of"): Value::use_begin/use_end
				1760	use-def chains ("finding all values used"): User::op_begin/op_end [op=operand]
				1761	-->
				1762
				1763	</div>
				1764
Chris Lattner	0665e1f	2008-01-03 16:56:04 +0000	[diff] [blame]	1765	<!--_______________________________________________________________________-->
				1766	<div class="doc_subsubsection">
				1767	<a name="iterate_preds">Iterating over predecessors &
				1768	successors of blocks</a>
				1769	</div>
				1770
				1771	<div class="doc_text">
				1772
				1773	<p>Iterating over the predecessors and successors of a block is quite easy
				1774	with the routines defined in <tt>"llvm/Support/CFG.h"</tt>. Just use code like
				1775	this to iterate over all predecessors of BB:</p>
				1776
				1777	<div class="doc_code">
				1778	<pre>
				1779	#include "llvm/Support/CFG.h"
				1780	BasicBlock *BB = ...;
				1781
				1782	for (pred_iterator PI = pred_begin(BB), E = pred_end(BB); PI != E; ++PI) {
				1783	BasicBlock Pred = PI;
				1784	// <i>...</i>
				1785	}
				1786	</pre>
				1787	</div>
				1788
				1789	<p>Similarly, to iterate over successors use
				1790	succ_iterator/succ_begin/succ_end.</p>
				1791
				1792	</div>
				1793
				1794
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	1795	<!-- ======================================================================= -->
				1796	<div class="doc_subsection">
				1797	<a name="simplechanges">Making simple changes</a>
				1798	</div>
				1799
				1800	<div class="doc_text">
				1801
				1802	<p>There are some primitive transformation operations present in the LLVM
				1803	infrastructure that are worth knowing about. When performing
				1804	transformations, it's fairly common to manipulate the contents of basic
				1805	blocks. This section describes some of the common methods for doing so
				1806	and gives example code.</p>
				1807
				1808	</div>
				1809
				1810	<!--_______________________________________________________________________-->
				1811	<div class="doc_subsubsection">
				1812	<a name="schanges_creating">Creating and inserting new
				1813	<tt>Instruction</tt>s</a>
				1814	</div>
				1815
				1816	<div class="doc_text">
				1817
				1818	<p><i>Instantiating Instructions</i></p>
				1819
				1820	<p>Creation of <tt>Instruction</tt>s is straight-forward: simply call the
				1821	constructor for the kind of instruction to instantiate and provide the necessary
				1822	parameters. For example, an <tt>AllocaInst</tt> only <i>requires</i> a
				1823	(const-ptr-to) <tt>Type</tt>. Thus:</p>
				1824
				1825	<div class="doc_code">
				1826	<pre>
Nick Lewycky	c004e5b	2007-12-03 01:52:52 +0000	[diff] [blame]	1827	AllocaInst* ai = new AllocaInst(Type::Int32Ty);
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	1828	</pre>
				1829	</div>
				1830
				1831	<p>will create an <tt>AllocaInst</tt> instance that represents the allocation of
				1832	one integer in the current stack frame, at run time. Each <tt>Instruction</tt>
				1833	subclass is likely to have varying default parameters which change the semantics
				1834	of the instruction, so refer to the <a
				1835	href="/doxygen/classllvm_1_1Instruction.html">doxygen documentation for the subclass of
				1836	Instruction</a> that you're interested in instantiating.</p>
				1837
				1838	<p><i>Naming values</i></p>
				1839
				1840	<p>It is very useful to name the values of instructions when you're able to, as
				1841	this facilitates the debugging of your transformations. If you end up looking
				1842	at generated LLVM machine code, you definitely want to have logical names
				1843	associated with the results of instructions! By supplying a value for the
				1844	<tt>Name</tt> (default) parameter of the <tt>Instruction</tt> constructor, you
				1845	associate a logical name with the result of the instruction's execution at
				1846	run time. For example, say that I'm writing a transformation that dynamically
				1847	allocates space for an integer on the stack, and that integer is going to be
				1848	used as some kind of index by some other code. To accomplish this, I place an
				1849	<tt>AllocaInst</tt> at the first point in the first <tt>BasicBlock</tt> of some
				1850	<tt>Function</tt>, and I'm intending to use it within the same
				1851	<tt>Function</tt>. I might do:</p>
				1852
				1853	<div class="doc_code">
				1854	<pre>
Nick Lewycky	c004e5b	2007-12-03 01:52:52 +0000	[diff] [blame]	1855	AllocaInst* pa = new AllocaInst(Type::Int32Ty, 0, "indexLoc");
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	1856	</pre>
				1857	</div>
				1858
				1859	<p>where <tt>indexLoc</tt> is now the logical name of the instruction's
				1860	execution value, which is a pointer to an integer on the run time stack.</p>
				1861
				1862	<p><i>Inserting instructions</i></p>
				1863
				1864	<p>There are essentially two ways to insert an <tt>Instruction</tt>
				1865	into an existing sequence of instructions that form a <tt>BasicBlock</tt>:</p>
				1866
				1867	<ul>
				1868	<li>Insertion into an explicit instruction list
				1869
				1870	<p>Given a <tt>BasicBlock* pb</tt>, an <tt>Instruction* pi</tt> within that
				1871	<tt>BasicBlock</tt>, and a newly-created instruction we wish to insert
				1872	before <tt>*pi</tt>, we do the following: </p>
				1873
				1874	<div class="doc_code">
				1875	<pre>
				1876	BasicBlock *pb = ...;
				1877	Instruction *pi = ...;
				1878	Instruction *newInst = new Instruction(...);
				1879
				1880	pb->getInstList().insert(pi, newInst); // <i>Inserts newInst before pi in pb</i>
				1881	</pre>
				1882	</div>
				1883
				1884	<p>Appending to the end of a <tt>BasicBlock</tt> is so common that
				1885	the <tt>Instruction</tt> class and <tt>Instruction</tt>-derived
				1886	classes provide constructors which take a pointer to a
				1887	<tt>BasicBlock</tt> to be appended to. For example code that
				1888	looked like: </p>
				1889
				1890	<div class="doc_code">
				1891	<pre>
				1892	BasicBlock *pb = ...;
				1893	Instruction *newInst = new Instruction(...);
				1894
				1895	pb->getInstList().push_back(newInst); // <i>Appends newInst to pb</i>
				1896	</pre>
				1897	</div>
				1898
				1899	<p>becomes: </p>
				1900
				1901	<div class="doc_code">
				1902	<pre>
				1903	BasicBlock *pb = ...;
				1904	Instruction *newInst = new Instruction(..., pb);
				1905	</pre>
				1906	</div>
				1907
				1908	<p>which is much cleaner, especially if you are creating
				1909	long instruction streams.</p></li>
				1910
				1911	<li>Insertion into an implicit instruction list
				1912
				1913	<p><tt>Instruction</tt> instances that are already in <tt>BasicBlock</tt>s
				1914	are implicitly associated with an existing instruction list: the instruction
				1915	list of the enclosing basic block. Thus, we could have accomplished the same
				1916	thing as the above code without being given a <tt>BasicBlock</tt> by doing:
				1917	</p>
				1918
				1919	<div class="doc_code">
				1920	<pre>
				1921	Instruction *pi = ...;
				1922	Instruction *newInst = new Instruction(...);
				1923
				1924	pi->getParent()->getInstList().insert(pi, newInst);
				1925	</pre>
				1926	</div>
				1927
				1928	<p>In fact, this sequence of steps occurs so frequently that the
				1929	<tt>Instruction</tt> class and <tt>Instruction</tt>-derived classes provide
				1930	constructors which take (as a default parameter) a pointer to an
				1931	<tt>Instruction</tt> which the newly-created <tt>Instruction</tt> should
				1932	precede. That is, <tt>Instruction</tt> constructors are capable of
				1933	inserting the newly-created instance into the <tt>BasicBlock</tt> of a
				1934	provided instruction, immediately before that instruction. Using an
				1935	<tt>Instruction</tt> constructor with a <tt>insertBefore</tt> (default)
				1936	parameter, the above code becomes:</p>
				1937
				1938	<div class="doc_code">
				1939	<pre>
				1940	Instruction* pi = ...;
				1941	Instruction* newInst = new Instruction(..., pi);
				1942	</pre>
				1943	</div>
				1944
				1945	<p>which is much cleaner, especially if you're creating a lot of
				1946	instructions and adding them to <tt>BasicBlock</tt>s.</p></li>
				1947	</ul>
				1948
				1949	</div>
				1950
				1951	<!--_______________________________________________________________________-->
				1952	<div class="doc_subsubsection">
				1953	<a name="schanges_deleting">Deleting <tt>Instruction</tt>s</a>
				1954	</div>
				1955
				1956	<div class="doc_text">
				1957
				1958	<p>Deleting an instruction from an existing sequence of instructions that form a
				1959	<a href="#BasicBlock"><tt>BasicBlock</tt></a> is very straight-forward. First,
				1960	you must have a pointer to the instruction that you wish to delete. Second, you
				1961	need to obtain the pointer to that instruction's basic block. You use the
				1962	pointer to the basic block to get its list of instructions and then use the
				1963	erase function to remove your instruction. For example:</p>
				1964
				1965	<div class="doc_code">
				1966	<pre>
				1967	<a href="#Instruction">Instruction</a> *I = .. ;
Chris Lattner	3db8f77	2008-02-15 22:57:17 +0000	[diff] [blame]	1968	I->eraseFromParent();
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	1969	</pre>
				1970	</div>
				1971
				1972	</div>
				1973
				1974	<!--_______________________________________________________________________-->
				1975	<div class="doc_subsubsection">
				1976	<a name="schanges_replacing">Replacing an <tt>Instruction</tt> with another
				1977	<tt>Value</tt></a>
				1978	</div>
				1979
				1980	<div class="doc_text">
				1981
				1982	<p><i>Replacing individual instructions</i></p>
				1983
				1984	<p>Including "<a href="/doxygen/BasicBlockUtils_8h-source.html">llvm/Transforms/Utils/BasicBlockUtils.h</a>"
				1985	permits use of two very useful replace functions: <tt>ReplaceInstWithValue</tt>
				1986	and <tt>ReplaceInstWithInst</tt>.</p>
				1987
				1988	<h4><a name="schanges_deleting">Deleting <tt>Instruction</tt>s</a></h4>
				1989
				1990	<ul>
				1991	<li><tt>ReplaceInstWithValue</tt>
				1992
Nick Lewycky	48d4b03	2008-09-15 06:31:52 +0000	[diff] [blame]	1993	<p>This function replaces all uses of a given instruction with a value,
				1994	and then removes the original instruction. The following example
				1995	illustrates the replacement of the result of a particular
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	1996	<tt>AllocaInst</tt> that allocates memory for a single integer with a null
				1997	pointer to an integer.</p>
				1998
				1999	<div class="doc_code">
				2000	<pre>
				2001	AllocaInst* instToReplace = ...;
				2002	BasicBlock::iterator ii(instToReplace);
				2003
				2004	ReplaceInstWithValue(instToReplace->getParent()->getInstList(), ii,
Daniel Dunbar	8ce7962	2008-10-03 22:17:25 +0000	[diff] [blame]	2005	Constant::getNullValue(PointerType::getUnqual(Type::Int32Ty)));
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	2006	</pre></div></li>
				2007
				2008	<li><tt>ReplaceInstWithInst</tt>
				2009
				2010	<p>This function replaces a particular instruction with another
Nick Lewycky	48d4b03	2008-09-15 06:31:52 +0000	[diff] [blame]	2011	instruction, inserting the new instruction into the basic block at the
				2012	location where the old instruction was, and replacing any uses of the old
				2013	instruction with the new instruction. The following example illustrates
				2014	the replacement of one <tt>AllocaInst</tt> with another.</p>
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	2015
				2016	<div class="doc_code">
				2017	<pre>
				2018	AllocaInst* instToReplace = ...;
				2019	BasicBlock::iterator ii(instToReplace);
				2020
				2021	ReplaceInstWithInst(instToReplace->getParent()->getInstList(), ii,
Nick Lewycky	c004e5b	2007-12-03 01:52:52 +0000	[diff] [blame]	2022	new AllocaInst(Type::Int32Ty, 0, "ptrToReplacedInt"));
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	2023	</pre></div></li>
				2024	</ul>
				2025
				2026	<p><i>Replacing multiple uses of <tt>User</tt>s and <tt>Value</tt>s</i></p>
				2027
				2028	<p>You can use <tt>Value::replaceAllUsesWith</tt> and
				2029	<tt>User::replaceUsesOfWith</tt> to change more than one use at a time. See the
				2030	doxygen documentation for the <a href="/doxygen/classllvm_1_1Value.html">Value Class</a>
				2031	and <a href="/doxygen/classllvm_1_1User.html">User Class</a>, respectively, for more
				2032	information.</p>
				2033
				2034	<!-- Value::replaceAllUsesWith User::replaceUsesOfWith Point out:
				2035	include/llvm/Transforms/Utils/ especially BasicBlockUtils.h with:
				2036	ReplaceInstWithValue, ReplaceInstWithInst -->
				2037
				2038	</div>
				2039
				2040	<!--_______________________________________________________________________-->
				2041	<div class="doc_subsubsection">
				2042	<a name="schanges_deletingGV">Deleting <tt>GlobalVariable</tt>s</a>
				2043	</div>
				2044
				2045	<div class="doc_text">
				2046
				2047	<p>Deleting a global variable from a module is just as easy as deleting an
				2048	Instruction. First, you must have a pointer to the global variable that you wish
				2049	to delete. You use this pointer to erase it from its parent, the module.
				2050	For example:</p>
				2051
				2052	<div class="doc_code">
				2053	<pre>
				2054	<a href="#GlobalVariable">GlobalVariable</a> *GV = .. ;
				2055
				2056	GV->eraseFromParent();
				2057	</pre>
				2058	</div>
				2059
				2060	</div>
				2061
				2062	<!-- *********************************************************************** -->
				2063	<div class="doc_section">
				2064	<a name="advanced">Advanced Topics</a>
				2065	</div>
				2066	<!-- *********************************************************************** -->
				2067
				2068	<div class="doc_text">
				2069	<p>
				2070	This section describes some of the advanced or obscure API's that most clients
				2071	do not need to be aware of. These API's tend manage the inner workings of the
				2072	LLVM system, and only need to be accessed in unusual circumstances.
				2073	</p>
				2074	</div>
				2075
				2076	<!-- ======================================================================= -->
				2077	<div class="doc_subsection">
				2078	<a name="TypeResolve">LLVM Type Resolution</a>
				2079	</div>
				2080
				2081	<div class="doc_text">
				2082
				2083	<p>
				2084	The LLVM type system has a very simple goal: allow clients to compare types for
				2085	structural equality with a simple pointer comparison (aka a shallow compare).
				2086	This goal makes clients much simpler and faster, and is used throughout the LLVM
				2087	system.
				2088	</p>
				2089
				2090	<p>
				2091	Unfortunately achieving this goal is not a simple matter. In particular,
				2092	recursive types and late resolution of opaque types makes the situation very
				2093	difficult to handle. Fortunately, for the most part, our implementation makes
				2094	most clients able to be completely unaware of the nasty internal details. The
				2095	primary case where clients are exposed to the inner workings of it are when
				2096	building a recursive type. In addition to this case, the LLVM bitcode reader,
				2097	assembly parser, and linker also have to be aware of the inner workings of this
				2098	system.
				2099	</p>
				2100
				2101	<p>
				2102	For our purposes below, we need three concepts. First, an "Opaque Type" is
				2103	exactly as defined in the <a href="LangRef.html#t_opaque">language
				2104	reference</a>. Second an "Abstract Type" is any type which includes an
				2105	opaque type as part of its type graph (for example "<tt>{ opaque, i32 }</tt>").
				2106	Third, a concrete type is a type that is not an abstract type (e.g. "<tt>{ i32,
				2107	float }</tt>").
				2108	</p>
				2109
				2110	</div>
				2111
				2112	<!-- ______________________________________________________________________ -->
				2113	<div class="doc_subsubsection">
				2114	<a name="BuildRecType">Basic Recursive Type Construction</a>
				2115	</div>
				2116
				2117	<div class="doc_text">
				2118
				2119	<p>
				2120	Because the most common question is "how do I build a recursive type with LLVM",
				2121	we answer it now and explain it as we go. Here we include enough to cause this
				2122	to be emitted to an output .ll file:
				2123	</p>
				2124
				2125	<div class="doc_code">
				2126	<pre>
				2127	%mylist = type { %mylist*, i32 }
				2128	</pre>
				2129	</div>
				2130
				2131	<p>
				2132	To build this, use the following LLVM APIs:
				2133	</p>
				2134
				2135	<div class="doc_code">
				2136	<pre>
				2137	// <i>Create the initial outer struct</i>
				2138	<a href="#PATypeHolder">PATypeHolder</a> StructTy = OpaqueType::get();
				2139	std::vector<const Type*> Elts;
Daniel Dunbar	8ce7962	2008-10-03 22:17:25 +0000	[diff] [blame]	2140	Elts.push_back(PointerType::getUnqual(StructTy));
Nick Lewycky	c004e5b	2007-12-03 01:52:52 +0000	[diff] [blame]	2141	Elts.push_back(Type::Int32Ty);
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	2142	StructType *NewSTy = StructType::get(Elts);
				2143
				2144	// <i>At this point, NewSTy = "{ opaque*, i32 }". Tell VMCore that</i>
				2145	// <i>the struct and the opaque type are actually the same.</i>
				2146	cast<OpaqueType>(StructTy.get())-><a href="#refineAbstractTypeTo">refineAbstractTypeTo</a>(NewSTy);
				2147
				2148	// <i>NewSTy is potentially invalidated, but StructTy (a <a href="#PATypeHolder">PATypeHolder</a>) is</i>
				2149	// <i>kept up-to-date</i>
				2150	NewSTy = cast<StructType>(StructTy.get());
				2151
				2152	// <i>Add a name for the type to the module symbol table (optional)</i>
				2153	MyModule->addTypeName("mylist", NewSTy);
				2154	</pre>
				2155	</div>
				2156
				2157	<p>
				2158	This code shows the basic approach used to build recursive types: build a
				2159	non-recursive type using 'opaque', then use type unification to close the cycle.
				2160	The type unification step is performed by the <tt><a
				2161	href="#refineAbstractTypeTo">refineAbstractTypeTo</a></tt> method, which is
				2162	described next. After that, we describe the <a
				2163	href="#PATypeHolder">PATypeHolder class</a>.
				2164	</p>
				2165
				2166	</div>
				2167
				2168	<!-- ______________________________________________________________________ -->
				2169	<div class="doc_subsubsection">
				2170	<a name="refineAbstractTypeTo">The <tt>refineAbstractTypeTo</tt> method</a>
				2171	</div>
				2172
				2173	<div class="doc_text">
				2174	<p>
				2175	The <tt>refineAbstractTypeTo</tt> method starts the type unification process.
				2176	While this method is actually a member of the DerivedType class, it is most
				2177	often used on OpaqueType instances. Type unification is actually a recursive
				2178	process. After unification, types can become structurally isomorphic to
				2179	existing types, and all duplicates are deleted (to preserve pointer equality).
				2180	</p>
				2181
				2182	<p>
				2183	In the example above, the OpaqueType object is definitely deleted.
				2184	Additionally, if there is an "{ \2*, i32}" type already created in the system,
				2185	the pointer and struct type created are <b>also</b> deleted. Obviously whenever
				2186	a type is deleted, any "Type*" pointers in the program are invalidated. As
				2187	such, it is safest to avoid having <i>any</i> "Type*" pointers to abstract types
				2188	live across a call to <tt>refineAbstractTypeTo</tt> (note that non-abstract
				2189	types can never move or be deleted). To deal with this, the <a
				2190	href="#PATypeHolder">PATypeHolder</a> class is used to maintain a stable
				2191	reference to a possibly refined type, and the <a
				2192	href="#AbstractTypeUser">AbstractTypeUser</a> class is used to update more
				2193	complex datastructures.
				2194	</p>
				2195
				2196	</div>
				2197
				2198	<!-- ______________________________________________________________________ -->
				2199	<div class="doc_subsubsection">
				2200	<a name="PATypeHolder">The PATypeHolder Class</a>
				2201	</div>
				2202
				2203	<div class="doc_text">
				2204	<p>
				2205	PATypeHolder is a form of a "smart pointer" for Type objects. When VMCore
				2206	happily goes about nuking types that become isomorphic to existing types, it
				2207	automatically updates all PATypeHolder objects to point to the new type. In the
				2208	example above, this allows the code to maintain a pointer to the resultant
				2209	resolved recursive type, even though the Type*'s are potentially invalidated.
				2210	</p>
				2211
				2212	<p>
				2213	PATypeHolder is an extremely light-weight object that uses a lazy union-find
				2214	implementation to update pointers. For example the pointer from a Value to its
				2215	Type is maintained by PATypeHolder objects.
				2216	</p>
				2217
				2218	</div>
				2219
				2220	<!-- ______________________________________________________________________ -->
				2221	<div class="doc_subsubsection">
				2222	<a name="AbstractTypeUser">The AbstractTypeUser Class</a>
				2223	</div>
				2224
				2225	<div class="doc_text">
				2226
				2227	<p>
				2228	Some data structures need more to perform more complex updates when types get
				2229	resolved. To support this, a class can derive from the AbstractTypeUser class.
				2230	This class
				2231	allows it to get callbacks when certain types are resolved. To register to get
				2232	callbacks for a particular type, the DerivedType::{add/remove}AbstractTypeUser
				2233	methods can be called on a type. Note that these methods only work for <i>
				2234	abstract</i> types. Concrete types (those that do not include any opaque
				2235	objects) can never be refined.
				2236	</p>
				2237	</div>
				2238
				2239
				2240	<!-- ======================================================================= -->
				2241	<div class="doc_subsection">
				2242	<a name="SymbolTable">The <tt>ValueSymbolTable</tt> and
				2243	<tt>TypeSymbolTable</tt> classes</a>
				2244	</div>
				2245
				2246	<div class="doc_text">
				2247	<p>The <tt><a href="http://llvm.org/doxygen/classllvm_1_1ValueSymbolTable.html">
				2248	ValueSymbolTable</a></tt> class provides a symbol table that the <a
				2249	href="#Function"><tt>Function</tt></a> and <a href="#Module">
				2250	<tt>Module</tt></a> classes use for naming value definitions. The symbol table
				2251	can provide a name for any <a href="#Value"><tt>Value</tt></a>.
				2252	The <tt><a href="http://llvm.org/doxygen/classllvm_1_1TypeSymbolTable.html">
				2253	TypeSymbolTable</a></tt> class is used by the <tt>Module</tt> class to store
				2254	names for types.</p>
				2255
				2256	<p>Note that the <tt>SymbolTable</tt> class should not be directly accessed
				2257	by most clients. It should only be used when iteration over the symbol table
				2258	names themselves are required, which is very special purpose. Note that not
				2259	all LLVM
Gabor Greif	92e8776	2008-06-16 21:06:12 +0000	[diff] [blame]	2260	<tt><a href="#Value">Value</a></tt>s have names, and those without names (i.e. they have
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	2261	an empty name) do not exist in the symbol table.
				2262	</p>
				2263
				2264	<p>These symbol tables support iteration over the values/types in the symbol
				2265	table with <tt>begin/end/iterator</tt> and supports querying to see if a
				2266	specific name is in the symbol table (with <tt>lookup</tt>). The
				2267	<tt>ValueSymbolTable</tt> class exposes no public mutator methods, instead,
				2268	simply call <tt>setName</tt> on a value, which will autoinsert it into the
				2269	appropriate symbol table. For types, use the Module::addTypeName method to
				2270	insert entries into the symbol table.</p>
				2271
				2272	</div>
				2273
				2274
				2275
Gabor Greif	92e8776	2008-06-16 21:06:12 +0000	[diff] [blame]	2276	<!-- ======================================================================= -->
				2277	<div class="doc_subsection">
				2278	<a name="UserLayout">The <tt>User</tt> and owned <tt>Use</tt> classes' memory layout</a>
				2279	</div>
				2280
				2281	<div class="doc_text">
				2282	<p>The <tt><a href="http://llvm.org/doxygen/classllvm_1_1User.html">
Gabor Greif	50626fc	2009-01-05 16:05:32 +0000	[diff] [blame]	2283	User</a></tt> class provides a basis for expressing the ownership of <tt>User</tt>
Gabor Greif	92e8776	2008-06-16 21:06:12 +0000	[diff] [blame]	2284	towards other <tt><a href="http://llvm.org/doxygen/classllvm_1_1Value.html">
				2285	Value</a></tt>s. The <tt><a href="http://llvm.org/doxygen/classllvm_1_1Use.html">
Gabor Greif	93b462b	2008-06-18 13:44:57 +0000	[diff] [blame]	2286	Use</a></tt> helper class is employed to do the bookkeeping and to facilitate <i>O(1)</i>
Gabor Greif	92e8776	2008-06-16 21:06:12 +0000	[diff] [blame]	2287	addition and removal.</p>
				2288
Gabor Greif	93b462b	2008-06-18 13:44:57 +0000	[diff] [blame]	2289	<!-- ______________________________________________________________________ -->
				2290	<div class="doc_subsubsection">
Gabor Greif	50626fc	2009-01-05 16:05:32 +0000	[diff] [blame]	2291	<a name="Use2User">Interaction and relationship between <tt>User</tt> and <tt>Use</tt> objects</a>
Gabor Greif	93b462b	2008-06-18 13:44:57 +0000	[diff] [blame]	2292	</div>
Gabor Greif	92e8776	2008-06-16 21:06:12 +0000	[diff] [blame]	2293
Gabor Greif	93b462b	2008-06-18 13:44:57 +0000	[diff] [blame]	2294	<div class="doc_text">
				2295	<p>
				2296	A subclass of <tt>User</tt> can choose between incorporating its <tt>Use</tt> objects
Gabor Greif	92e8776	2008-06-16 21:06:12 +0000	[diff] [blame]	2297	or refer to them out-of-line by means of a pointer. A mixed variant
Gabor Greif	93b462b	2008-06-18 13:44:57 +0000	[diff] [blame]	2298	(some <tt>Use</tt>s inline others hung off) is impractical and breaks the invariant
				2299	that the <tt>Use</tt> objects belonging to the same <tt>User</tt> form a contiguous array.
				2300	</p>
				2301	</div>
Gabor Greif	92e8776	2008-06-16 21:06:12 +0000	[diff] [blame]	2302
Gabor Greif	93b462b	2008-06-18 13:44:57 +0000	[diff] [blame]	2303	<p>
				2304	We have 2 different layouts in the <tt>User</tt> (sub)classes:
				2305	<ul>
				2306	<li><p>Layout a)
				2307	The <tt>Use</tt> object(s) are inside (resp. at fixed offset) of the <tt>User</tt>
				2308	object and there are a fixed number of them.</p>
Gabor Greif	92e8776	2008-06-16 21:06:12 +0000	[diff] [blame]	2309
Gabor Greif	93b462b	2008-06-18 13:44:57 +0000	[diff] [blame]	2310	<li><p>Layout b)
				2311	The <tt>Use</tt> object(s) are referenced by a pointer to an
				2312	array from the <tt>User</tt> object and there may be a variable
				2313	number of them.</p>
				2314	</ul>
				2315	<p>
Gabor Greif	e247e36	2008-06-25 00:10:22 +0000	[diff] [blame]	2316	As of v2.4 each layout still possesses a direct pointer to the
Gabor Greif	93b462b	2008-06-18 13:44:57 +0000	[diff] [blame]	2317	start of the array of <tt>Use</tt>s. Though not mandatory for layout a),
Gabor Greif	92e8776	2008-06-16 21:06:12 +0000	[diff] [blame]	2318	we stick to this redundancy for the sake of simplicity.
Gabor Greif	e247e36	2008-06-25 00:10:22 +0000	[diff] [blame]	2319	The <tt>User</tt> object also stores the number of <tt>Use</tt> objects it
Gabor Greif	92e8776	2008-06-16 21:06:12 +0000	[diff] [blame]	2320	has. (Theoretically this information can also be calculated
Gabor Greif	93b462b	2008-06-18 13:44:57 +0000	[diff] [blame]	2321	given the scheme presented below.)</p>
				2322	<p>
				2323	Special forms of allocation operators (<tt>operator new</tt>)
Gabor Greif	e247e36	2008-06-25 00:10:22 +0000	[diff] [blame]	2324	enforce the following memory layouts:</p>
Gabor Greif	92e8776	2008-06-16 21:06:12 +0000	[diff] [blame]	2325
Gabor Greif	93b462b	2008-06-18 13:44:57 +0000	[diff] [blame]	2326	<ul>
Gabor Greif	e247e36	2008-06-25 00:10:22 +0000	[diff] [blame]	2327	<li><p>Layout a) is modelled by prepending the <tt>User</tt> object by the <tt>Use[]</tt> array.</p>
Gabor Greif	92e8776	2008-06-16 21:06:12 +0000	[diff] [blame]	2328
Gabor Greif	93b462b	2008-06-18 13:44:57 +0000	[diff] [blame]	2329	<pre>
				2330	...---.---.---.---.-------...
				2331	\| P \| P \| P \| P \| User
				2332	'''---'---'---'---'-------'''
				2333	</pre>
Gabor Greif	92e8776	2008-06-16 21:06:12 +0000	[diff] [blame]	2334
Gabor Greif	e247e36	2008-06-25 00:10:22 +0000	[diff] [blame]	2335	<li><p>Layout b) is modelled by pointing at the <tt>Use[]</tt> array.</p>
Gabor Greif	93b462b	2008-06-18 13:44:57 +0000	[diff] [blame]	2336	<pre>
				2337	.-------...
				2338	\| User
				2339	'-------'''
				2340	\|
				2341	v
				2342	.---.---.---.---...
				2343	\| P \| P \| P \| P \|
				2344	'---'---'---'---'''
				2345	</pre>
				2346	</ul>
				2347	<i>(In the above figures '<tt>P</tt>' stands for the <tt>Use**</tt> that
				2348	is stored in each <tt>Use</tt> object in the member <tt>Use::Prev</tt>)</i>
Gabor Greif	92e8776	2008-06-16 21:06:12 +0000	[diff] [blame]	2349
Gabor Greif	93b462b	2008-06-18 13:44:57 +0000	[diff] [blame]	2350	<!-- ______________________________________________________________________ -->
				2351	<div class="doc_subsubsection">
Gabor Greif	50626fc	2009-01-05 16:05:32 +0000	[diff] [blame]	2352	<a name="Waymarking">The waymarking algorithm</a>
Gabor Greif	93b462b	2008-06-18 13:44:57 +0000	[diff] [blame]	2353	</div>
Gabor Greif	92e8776	2008-06-16 21:06:12 +0000	[diff] [blame]	2354
Gabor Greif	93b462b	2008-06-18 13:44:57 +0000	[diff] [blame]	2355	<div class="doc_text">
				2356	<p>
Gabor Greif	e247e36	2008-06-25 00:10:22 +0000	[diff] [blame]	2357	Since the <tt>Use</tt> objects are deprived of the direct (back)pointer to
Gabor Greif	93b462b	2008-06-18 13:44:57 +0000	[diff] [blame]	2358	their <tt>User</tt> objects, there must be a fast and exact method to
				2359	recover it. This is accomplished by the following scheme:</p>
				2360	</div>
Gabor Greif	92e8776	2008-06-16 21:06:12 +0000	[diff] [blame]	2361
Gabor Greif	e247e36	2008-06-25 00:10:22 +0000	[diff] [blame]	2362	A bit-encoding in the 2 LSBits (least significant bits) of the <tt>Use::Prev</tt> allows to find the
Gabor Greif	93b462b	2008-06-18 13:44:57 +0000	[diff] [blame]	2363	start of the <tt>User</tt> object:
				2364	<ul>
				2365	<li><tt>00</tt> —> binary digit 0</li>
				2366	<li><tt>01</tt> —> binary digit 1</li>
				2367	<li><tt>10</tt> —> stop and calculate (<tt>s</tt>)</li>
				2368	<li><tt>11</tt> —> full stop (<tt>S</tt>)</li>
				2369	</ul>
				2370	<p>
				2371	Given a <tt>Use*</tt>, all we have to do is to walk till we get
				2372	a stop and we either have a <tt>User</tt> immediately behind or
Gabor Greif	92e8776	2008-06-16 21:06:12 +0000	[diff] [blame]	2373	we have to walk to the next stop picking up digits
Gabor Greif	93b462b	2008-06-18 13:44:57 +0000	[diff] [blame]	2374	and calculating the offset:</p>
				2375	<pre>
Gabor Greif	92e8776	2008-06-16 21:06:12 +0000	[diff] [blame]	2376	.---.---.---.---.---.---.---.---.---.---.---.---.---.---.---.---.----------------
				2377	\| 1 \| s \| 1 \| 0 \| 1 \| 0 \| s \| 1 \| 1 \| 0 \| s \| 1 \| 1 \| s \| 1 \| S \| User (or User*)
				2378	'---'---'---'---'---'---'---'---'---'---'---'---'---'---'---'---'----------------
				2379	\|+15 \|+10 \|+6 \|+3 \|+1
				2380	\| \| \| \| \|__>
				2381	\| \| \| \|__________>
				2382	\| \| \|______________________>
				2383	\| \|______________________________________>
				2384	\|__________________________________________________________>
Gabor Greif	93b462b	2008-06-18 13:44:57 +0000	[diff] [blame]	2385	</pre>
				2386	<p>
Gabor Greif	92e8776	2008-06-16 21:06:12 +0000	[diff] [blame]	2387	Only the significant number of bits need to be stored between the
Gabor Greif	93b462b	2008-06-18 13:44:57 +0000	[diff] [blame]	2388	stops, so that the <i>worst case is 20 memory accesses</i> when there are
				2389	1000 <tt>Use</tt> objects associated with a <tt>User</tt>.</p>
Gabor Greif	92e8776	2008-06-16 21:06:12 +0000	[diff] [blame]	2390
Gabor Greif	93b462b	2008-06-18 13:44:57 +0000	[diff] [blame]	2391	<!-- ______________________________________________________________________ -->
				2392	<div class="doc_subsubsection">
Gabor Greif	50626fc	2009-01-05 16:05:32 +0000	[diff] [blame]	2393	<a name="ReferenceImpl">Reference implementation</a>
Gabor Greif	93b462b	2008-06-18 13:44:57 +0000	[diff] [blame]	2394	</div>
Gabor Greif	92e8776	2008-06-16 21:06:12 +0000	[diff] [blame]	2395
Gabor Greif	93b462b	2008-06-18 13:44:57 +0000	[diff] [blame]	2396	<div class="doc_text">
				2397	<p>
				2398	The following literate Haskell fragment demonstrates the concept:</p>
				2399	</div>
				2400
				2401	<div class="doc_code">
				2402	<pre>
Gabor Greif	92e8776	2008-06-16 21:06:12 +0000	[diff] [blame]	2403	> import Test.QuickCheck
				2404	>
				2405	> digits :: Int -> [Char] -> [Char]
				2406	> digits 0 acc = '0' : acc
				2407	> digits 1 acc = '1' : acc
				2408	> digits n acc = digits (n `div` 2) $ digits (n `mod` 2) acc
				2409	>
				2410	> dist :: Int -> [Char] -> [Char]
				2411	> dist 0 [] = ['S']
				2412	> dist 0 acc = acc
				2413	> dist 1 acc = let r = dist 0 acc in 's' : digits (length r) r
				2414	> dist n acc = dist (n - 1) $ dist 1 acc
				2415	>
				2416	> takeLast n ss = reverse $ take n $ reverse ss
				2417	>
				2418	> test = takeLast 40 $ dist 20 []
				2419	>
Gabor Greif	93b462b	2008-06-18 13:44:57 +0000	[diff] [blame]	2420	</pre>
				2421	</div>
				2422	<p>
				2423	Printing <test> gives: <tt>"1s100000s11010s10100s1111s1010s110s11s1S"</tt></p>
				2424	<p>
				2425	The reverse algorithm computes the length of the string just by examining
				2426	a certain prefix:</p>
Gabor Greif	92e8776	2008-06-16 21:06:12 +0000	[diff] [blame]	2427
Gabor Greif	93b462b	2008-06-18 13:44:57 +0000	[diff] [blame]	2428	<div class="doc_code">
				2429	<pre>
Gabor Greif	92e8776	2008-06-16 21:06:12 +0000	[diff] [blame]	2430	> pref :: [Char] -> Int
				2431	> pref "S" = 1
				2432	> pref ('s':'1':rest) = decode 2 1 rest
				2433	> pref (_:rest) = 1 + pref rest
				2434	>
				2435	> decode walk acc ('0':rest) = decode (walk + 1) (acc * 2) rest
				2436	> decode walk acc ('1':rest) = decode (walk + 1) (acc * 2 + 1) rest
				2437	> decode walk acc _ = walk + acc
				2438	>
Gabor Greif	93b462b	2008-06-18 13:44:57 +0000	[diff] [blame]	2439	</pre>
				2440	</div>
				2441	<p>
				2442	Now, as expected, printing <pref test> gives <tt>40</tt>.</p>
				2443	<p>
				2444	We can <i>quickCheck</i> this with following property:</p>
Gabor Greif	92e8776	2008-06-16 21:06:12 +0000	[diff] [blame]	2445
Gabor Greif	93b462b	2008-06-18 13:44:57 +0000	[diff] [blame]	2446	<div class="doc_code">
				2447	<pre>
Gabor Greif	92e8776	2008-06-16 21:06:12 +0000	[diff] [blame]	2448	> testcase = dist 2000 []
				2449	> testcaseLength = length testcase
				2450	>
				2451	> identityProp n = n > 0 && n <= testcaseLength ==> length arr == pref arr
				2452	> where arr = takeLast n testcase
Gabor Greif	93b462b	2008-06-18 13:44:57 +0000	[diff] [blame]	2453	>
				2454	</pre>
				2455	</div>
				2456	<p>
				2457	As expected <quickCheck identityProp> gives:</p>
Gabor Greif	92e8776	2008-06-16 21:06:12 +0000	[diff] [blame]	2458
Gabor Greif	93b462b	2008-06-18 13:44:57 +0000	[diff] [blame]	2459	<pre>
Gabor Greif	92e8776	2008-06-16 21:06:12 +0000	[diff] [blame]	2460	*Main> quickCheck identityProp
				2461	OK, passed 100 tests.
Gabor Greif	93b462b	2008-06-18 13:44:57 +0000	[diff] [blame]	2462	</pre>
				2463	<p>
				2464	Let's be a bit more exhaustive:</p>
Gabor Greif	92e8776	2008-06-16 21:06:12 +0000	[diff] [blame]	2465
Gabor Greif	93b462b	2008-06-18 13:44:57 +0000	[diff] [blame]	2466	<div class="doc_code">
				2467	<pre>
Gabor Greif	92e8776	2008-06-16 21:06:12 +0000	[diff] [blame]	2468	>
				2469	> deepCheck p = check (defaultConfig { configMaxTest = 500 }) p
				2470	>
Gabor Greif	93b462b	2008-06-18 13:44:57 +0000	[diff] [blame]	2471	</pre>
				2472	</div>
				2473	<p>
				2474	And here is the result of <deepCheck identityProp>:</p>
Gabor Greif	92e8776	2008-06-16 21:06:12 +0000	[diff] [blame]	2475
Gabor Greif	93b462b	2008-06-18 13:44:57 +0000	[diff] [blame]	2476	<pre>
Gabor Greif	92e8776	2008-06-16 21:06:12 +0000	[diff] [blame]	2477	*Main> deepCheck identityProp
				2478	OK, passed 500 tests.
Gabor Greif	92e8776	2008-06-16 21:06:12 +0000	[diff] [blame]	2479	</pre>
				2480
Gabor Greif	93b462b	2008-06-18 13:44:57 +0000	[diff] [blame]	2481	<!-- ______________________________________________________________________ -->
				2482	<div class="doc_subsubsection">
Gabor Greif	50626fc	2009-01-05 16:05:32 +0000	[diff] [blame]	2483	<a name="Tagging">Tagging considerations</a>
Gabor Greif	93b462b	2008-06-18 13:44:57 +0000	[diff] [blame]	2484	</div>
				2485
				2486	<p>
				2487	To maintain the invariant that the 2 LSBits of each <tt>Use**</tt> in <tt>Use</tt>
				2488	never change after being set up, setters of <tt>Use::Prev</tt> must re-tag the
				2489	new <tt>Use**</tt> on every modification. Accordingly getters must strip the
				2490	tag bits.</p>
				2491	<p>
Gabor Greif	e247e36	2008-06-25 00:10:22 +0000	[diff] [blame]	2492	For layout b) instead of the <tt>User</tt> we find a pointer (<tt>User*</tt> with LSBit set).
				2493	Following this pointer brings us to the <tt>User</tt>. A portable trick ensures
				2494	that the first bytes of <tt>User</tt> (if interpreted as a pointer) never has
Gabor Greif	50626fc	2009-01-05 16:05:32 +0000	[diff] [blame]	2495	the LSBit set. (Portability is relying on the fact that all known compilers place the
				2496	<tt>vptr</tt> in the first word of the instances.)</p>
Gabor Greif	93b462b	2008-06-18 13:44:57 +0000	[diff] [blame]	2497
Gabor Greif	92e8776	2008-06-16 21:06:12 +0000	[diff] [blame]	2498	</div>
				2499
				2500	<!-- *********************************************************************** -->
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	2501	<div class="doc_section">
				2502	<a name="coreclasses">The Core LLVM Class Hierarchy Reference </a>
				2503	</div>
				2504	<!-- *********************************************************************** -->
				2505
				2506	<div class="doc_text">
				2507	<p><tt>#include "<a href="/doxygen/Type_8h-source.html">llvm/Type.h</a>"</tt>
				2508	<br>doxygen info: <a href="/doxygen/classllvm_1_1Type.html">Type Class</a></p>
				2509
				2510	<p>The Core LLVM classes are the primary means of representing the program
				2511	being inspected or transformed. The core LLVM classes are defined in
				2512	header files in the <tt>include/llvm/</tt> directory, and implemented in
				2513	the <tt>lib/VMCore</tt> directory.</p>
				2514
				2515	</div>
				2516
				2517	<!-- ======================================================================= -->
				2518	<div class="doc_subsection">
				2519	<a name="Type">The <tt>Type</tt> class and Derived Types</a>
				2520	</div>
				2521
				2522	<div class="doc_text">
				2523
				2524	<p><tt>Type</tt> is a superclass of all type classes. Every <tt>Value</tt> has
				2525	a <tt>Type</tt>. <tt>Type</tt> cannot be instantiated directly but only
				2526	through its subclasses. Certain primitive types (<tt>VoidType</tt>,
				2527	<tt>LabelType</tt>, <tt>FloatType</tt> and <tt>DoubleType</tt>) have hidden
				2528	subclasses. They are hidden because they offer no useful functionality beyond
				2529	what the <tt>Type</tt> class offers except to distinguish themselves from
				2530	other subclasses of <tt>Type</tt>.</p>
				2531	<p>All other types are subclasses of <tt>DerivedType</tt>. Types can be
				2532	named, but this is not a requirement. There exists exactly
				2533	one instance of a given shape at any one time. This allows type equality to
				2534	be performed with address equality of the Type Instance. That is, given two
				2535	<tt>Type*</tt> values, the types are identical if the pointers are identical.
				2536	</p>
				2537	</div>
				2538
				2539	<!-- _______________________________________________________________________ -->
				2540	<div class="doc_subsubsection">
Gabor Greif	50626fc	2009-01-05 16:05:32 +0000	[diff] [blame]	2541	<a name="m_Type">Important Public Methods</a>
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	2542	</div>
				2543
				2544	<div class="doc_text">
				2545
				2546	<ul>
				2547	<li><tt>bool isInteger() const</tt>: Returns true for any integer type.</li>
				2548
				2549	<li><tt>bool isFloatingPoint()</tt>: Return true if this is one of the two
				2550	floating point types.</li>
				2551
				2552	<li><tt>bool isAbstract()</tt>: Return true if the type is abstract (contains
				2553	an OpaqueType anywhere in its definition).</li>
				2554
				2555	<li><tt>bool isSized()</tt>: Return true if the type has known size. Things
				2556	that don't have a size are abstract types, labels and void.</li>
				2557
				2558	</ul>
				2559	</div>
				2560
				2561	<!-- _______________________________________________________________________ -->
				2562	<div class="doc_subsubsection">
Gabor Greif	50626fc	2009-01-05 16:05:32 +0000	[diff] [blame]	2563	<a name="derivedtypes">Important Derived Types</a>
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	2564	</div>
				2565	<div class="doc_text">
				2566	<dl>
				2567	<dt><tt>IntegerType</tt></dt>
				2568	<dd>Subclass of DerivedType that represents integer types of any bit width.
				2569	Any bit width between <tt>IntegerType::MIN_INT_BITS</tt> (1) and
				2570	<tt>IntegerType::MAX_INT_BITS</tt> (~8 million) can be represented.
				2571	<ul>
				2572	<li><tt>static const IntegerType* get(unsigned NumBits)</tt>: get an integer
				2573	type of a specific bit width.</li>
				2574	<li><tt>unsigned getBitWidth() const</tt>: Get the bit width of an integer
				2575	type.</li>
				2576	</ul>
				2577	</dd>
				2578	<dt><tt>SequentialType</tt></dt>
				2579	<dd>This is subclassed by ArrayType and PointerType
				2580	<ul>
				2581	<li><tt>const Type * getElementType() const</tt>: Returns the type of each
				2582	of the elements in the sequential type. </li>
				2583	</ul>
				2584	</dd>
				2585	<dt><tt>ArrayType</tt></dt>
				2586	<dd>This is a subclass of SequentialType and defines the interface for array
				2587	types.
				2588	<ul>
				2589	<li><tt>unsigned getNumElements() const</tt>: Returns the number of
				2590	elements in the array. </li>
				2591	</ul>
				2592	</dd>
				2593	<dt><tt>PointerType</tt></dt>
				2594	<dd>Subclass of SequentialType for pointer types.</dd>
				2595	<dt><tt>VectorType</tt></dt>
				2596	<dd>Subclass of SequentialType for vector types. A
				2597	vector type is similar to an ArrayType but is distinguished because it is
				2598	a first class type wherease ArrayType is not. Vector types are used for
				2599	vector operations and are usually small vectors of of an integer or floating
				2600	point type.</dd>
				2601	<dt><tt>StructType</tt></dt>
				2602	<dd>Subclass of DerivedTypes for struct types.</dd>
				2603	<dt><tt><a name="FunctionType">FunctionType</a></tt></dt>
				2604	<dd>Subclass of DerivedTypes for function types.
				2605	<ul>
				2606	<li><tt>bool isVarArg() const</tt>: Returns true if its a vararg
				2607	function</li>
				2608	<li><tt> const Type * getReturnType() const</tt>: Returns the
				2609	return type of the function.</li>
				2610	<li><tt>const Type * getParamType (unsigned i)</tt>: Returns
				2611	the type of the ith parameter.</li>
				2612	<li><tt> const unsigned getNumParams() const</tt>: Returns the
				2613	number of formal parameters.</li>
				2614	</ul>
				2615	</dd>
				2616	<dt><tt>OpaqueType</tt></dt>
				2617	<dd>Sublcass of DerivedType for abstract types. This class
				2618	defines no content and is used as a placeholder for some other type. Note
				2619	that OpaqueType is used (temporarily) during type resolution for forward
				2620	references of types. Once the referenced type is resolved, the OpaqueType
				2621	is replaced with the actual type. OpaqueType can also be used for data
				2622	abstraction. At link time opaque types can be resolved to actual types
				2623	of the same name.</dd>
				2624	</dl>
				2625	</div>
				2626
				2627
				2628
				2629	<!-- ======================================================================= -->
				2630	<div class="doc_subsection">
				2631	<a name="Module">The <tt>Module</tt> class</a>
				2632	</div>
				2633
				2634	<div class="doc_text">
				2635
				2636	<p><tt>#include "<a
				2637	href="/doxygen/Module_8h-source.html">llvm/Module.h</a>"</tt><br> doxygen info:
				2638	<a href="/doxygen/classllvm_1_1Module.html">Module Class</a></p>
				2639
				2640	<p>The <tt>Module</tt> class represents the top level structure present in LLVM
				2641	programs. An LLVM module is effectively either a translation unit of the
				2642	original program or a combination of several translation units merged by the
				2643	linker. The <tt>Module</tt> class keeps track of a list of <a
				2644	href="#Function"><tt>Function</tt></a>s, a list of <a
				2645	href="#GlobalVariable"><tt>GlobalVariable</tt></a>s, and a <a
				2646	href="#SymbolTable"><tt>SymbolTable</tt></a>. Additionally, it contains a few
				2647	helpful member functions that try to make common operations easy.</p>
				2648
				2649	</div>
				2650
				2651	<!-- _______________________________________________________________________ -->
				2652	<div class="doc_subsubsection">
				2653	<a name="m_Module">Important Public Members of the <tt>Module</tt> class</a>
				2654	</div>
				2655
				2656	<div class="doc_text">
				2657
				2658	<ul>
				2659	<li><tt>Module::Module(std::string name = "")</tt></li>
				2660	</ul>
				2661
				2662	<p>Constructing a <a href="#Module">Module</a> is easy. You can optionally
				2663	provide a name for it (probably based on the name of the translation unit).</p>
				2664
				2665	<ul>
				2666	<li><tt>Module::iterator</tt> - Typedef for function list iterator<br>
				2667	<tt>Module::const_iterator</tt> - Typedef for const_iterator.<br>
				2668
				2669	<tt>begin()</tt>, <tt>end()</tt>
				2670	<tt>size()</tt>, <tt>empty()</tt>
				2671
				2672	<p>These are forwarding methods that make it easy to access the contents of
				2673	a <tt>Module</tt> object's <a href="#Function"><tt>Function</tt></a>
				2674	list.</p></li>
				2675
				2676	<li><tt>Module::FunctionListType &getFunctionList()</tt>
				2677
				2678	<p> Returns the list of <a href="#Function"><tt>Function</tt></a>s. This is
				2679	necessary to use when you need to update the list or perform a complex
				2680	action that doesn't have a forwarding method.</p>
				2681
				2682	<p><!-- Global Variable --></p></li>
				2683	</ul>
				2684
				2685	<hr>
				2686
				2687	<ul>
				2688	<li><tt>Module::global_iterator</tt> - Typedef for global variable list iterator<br>
				2689
				2690	<tt>Module::const_global_iterator</tt> - Typedef for const_iterator.<br>
				2691
				2692	<tt>global_begin()</tt>, <tt>global_end()</tt>
				2693	<tt>global_size()</tt>, <tt>global_empty()</tt>
				2694
				2695	<p> These are forwarding methods that make it easy to access the contents of
				2696	a <tt>Module</tt> object's <a
				2697	href="#GlobalVariable"><tt>GlobalVariable</tt></a> list.</p></li>
				2698
				2699	<li><tt>Module::GlobalListType &getGlobalList()</tt>
				2700
				2701	<p>Returns the list of <a
				2702	href="#GlobalVariable"><tt>GlobalVariable</tt></a>s. This is necessary to
				2703	use when you need to update the list or perform a complex action that
				2704	doesn't have a forwarding method.</p>
				2705
				2706	<p><!-- Symbol table stuff --> </p></li>
				2707	</ul>
				2708
				2709	<hr>
				2710
				2711	<ul>
				2712	<li><tt><a href="#SymbolTable">SymbolTable</a> *getSymbolTable()</tt>
				2713
				2714	<p>Return a reference to the <a href="#SymbolTable"><tt>SymbolTable</tt></a>
				2715	for this <tt>Module</tt>.</p>
				2716
				2717	<p><!-- Convenience methods --></p></li>
				2718	</ul>
				2719
				2720	<hr>
				2721
				2722	<ul>
				2723	<li><tt><a href="#Function">Function</a> *getFunction(const std::string
				2724	&Name, const <a href="#FunctionType">FunctionType</a> *Ty)</tt>
				2725
				2726	<p>Look up the specified function in the <tt>Module</tt> <a
				2727	href="#SymbolTable"><tt>SymbolTable</tt></a>. If it does not exist, return
				2728	<tt>null</tt>.</p></li>
				2729
				2730	<li><tt><a href="#Function">Function</a> *getOrInsertFunction(const
				2731	std::string &Name, const <a href="#FunctionType">FunctionType</a> *T)</tt>
				2732
				2733	<p>Look up the specified function in the <tt>Module</tt> <a
				2734	href="#SymbolTable"><tt>SymbolTable</tt></a>. If it does not exist, add an
				2735	external declaration for the function and return it.</p></li>
				2736
				2737	<li><tt>std::string getTypeName(const <a href="#Type">Type</a> *Ty)</tt>
				2738
				2739	<p>If there is at least one entry in the <a
				2740	href="#SymbolTable"><tt>SymbolTable</tt></a> for the specified <a
				2741	href="#Type"><tt>Type</tt></a>, return it. Otherwise return the empty
				2742	string.</p></li>
				2743
				2744	<li><tt>bool addTypeName(const std::string &Name, const <a
				2745	href="#Type">Type</a> *Ty)</tt>
				2746
				2747	<p>Insert an entry in the <a href="#SymbolTable"><tt>SymbolTable</tt></a>
				2748	mapping <tt>Name</tt> to <tt>Ty</tt>. If there is already an entry for this
				2749	name, true is returned and the <a
				2750	href="#SymbolTable"><tt>SymbolTable</tt></a> is not modified.</p></li>
				2751	</ul>
				2752
				2753	</div>
				2754
				2755
				2756	<!-- ======================================================================= -->
				2757	<div class="doc_subsection">
				2758	<a name="Value">The <tt>Value</tt> class</a>
				2759	</div>
				2760
				2761	<div class="doc_text">
				2762
				2763	<p><tt>#include "<a href="/doxygen/Value_8h-source.html">llvm/Value.h</a>"</tt>
				2764	<br>
				2765	doxygen info: <a href="/doxygen/classllvm_1_1Value.html">Value Class</a></p>
				2766
				2767	<p>The <tt>Value</tt> class is the most important class in the LLVM Source
				2768	base. It represents a typed value that may be used (among other things) as an
				2769	operand to an instruction. There are many different types of <tt>Value</tt>s,
				2770	such as <a href="#Constant"><tt>Constant</tt></a>s,<a
				2771	href="#Argument"><tt>Argument</tt></a>s. Even <a
				2772	href="#Instruction"><tt>Instruction</tt></a>s and <a
				2773	href="#Function"><tt>Function</tt></a>s are <tt>Value</tt>s.</p>
				2774
				2775	<p>A particular <tt>Value</tt> may be used many times in the LLVM representation
				2776	for a program. For example, an incoming argument to a function (represented
				2777	with an instance of the <a href="#Argument">Argument</a> class) is "used" by
				2778	every instruction in the function that references the argument. To keep track
				2779	of this relationship, the <tt>Value</tt> class keeps a list of all of the <a
				2780	href="#User"><tt>User</tt></a>s that is using it (the <a
				2781	href="#User"><tt>User</tt></a> class is a base class for all nodes in the LLVM
				2782	graph that can refer to <tt>Value</tt>s). This use list is how LLVM represents
				2783	def-use information in the program, and is accessible through the <tt>use_</tt>*
				2784	methods, shown below.</p>
				2785
				2786	<p>Because LLVM is a typed representation, every LLVM <tt>Value</tt> is typed,
				2787	and this <a href="#Type">Type</a> is available through the <tt>getType()</tt>
				2788	method. In addition, all LLVM values can be named. The "name" of the
				2789	<tt>Value</tt> is a symbolic string printed in the LLVM code:</p>
				2790
				2791	<div class="doc_code">
				2792	<pre>
				2793	%<b>foo</b> = add i32 1, 2
				2794	</pre>
				2795	</div>
				2796
				2797	<p><a name="nameWarning">The name of this instruction is "foo".</a> <b>NOTE</b>
				2798	that the name of any value may be missing (an empty string), so names should
				2799	<b>ONLY</b> be used for debugging (making the source code easier to read,
				2800	debugging printouts), they should not be used to keep track of values or map
				2801	between them. For this purpose, use a <tt>std::map</tt> of pointers to the
				2802	<tt>Value</tt> itself instead.</p>
				2803
				2804	<p>One important aspect of LLVM is that there is no distinction between an SSA
				2805	variable and the operation that produces it. Because of this, any reference to
				2806	the value produced by an instruction (or the value available as an incoming
				2807	argument, for example) is represented as a direct pointer to the instance of
				2808	the class that
				2809	represents this value. Although this may take some getting used to, it
				2810	simplifies the representation and makes it easier to manipulate.</p>
				2811
				2812	</div>
				2813
				2814	<!-- _______________________________________________________________________ -->
				2815	<div class="doc_subsubsection">
				2816	<a name="m_Value">Important Public Members of the <tt>Value</tt> class</a>
				2817	</div>
				2818
				2819	<div class="doc_text">
				2820
				2821	<ul>
				2822	<li><tt>Value::use_iterator</tt> - Typedef for iterator over the
				2823	use-list<br>
				2824	<tt>Value::use_const_iterator</tt> - Typedef for const_iterator over
				2825	the use-list<br>
				2826	<tt>unsigned use_size()</tt> - Returns the number of users of the
				2827	value.<br>
				2828	<tt>bool use_empty()</tt> - Returns true if there are no users.<br>
				2829	<tt>use_iterator use_begin()</tt> - Get an iterator to the start of
				2830	the use-list.<br>
				2831	<tt>use_iterator use_end()</tt> - Get an iterator to the end of the
				2832	use-list.<br>
				2833	<tt><a href="#User">User</a> *use_back()</tt> - Returns the last
				2834	element in the list.
				2835	<p> These methods are the interface to access the def-use
				2836	information in LLVM. As with all other iterators in LLVM, the naming
				2837	conventions follow the conventions defined by the <a href="#stl">STL</a>.</p>
				2838	</li>
				2839	<li><tt><a href="#Type">Type</a> *getType() const</tt>
				2840	<p>This method returns the Type of the Value.</p>
				2841	</li>
				2842	<li><tt>bool hasName() const</tt><br>
				2843	<tt>std::string getName() const</tt><br>
				2844	<tt>void setName(const std::string &Name)</tt>
				2845	<p> This family of methods is used to access and assign a name to a <tt>Value</tt>,
				2846	be aware of the <a href="#nameWarning">precaution above</a>.</p>
				2847	</li>
				2848	<li><tt>void replaceAllUsesWith(Value *V)</tt>
				2849
				2850	<p>This method traverses the use list of a <tt>Value</tt> changing all <a
				2851	href="#User"><tt>User</tt>s</a> of the current value to refer to
				2852	"<tt>V</tt>" instead. For example, if you detect that an instruction always
				2853	produces a constant value (for example through constant folding), you can
				2854	replace all uses of the instruction with the constant like this:</p>
				2855
				2856	<div class="doc_code">
				2857	<pre>
				2858	Inst->replaceAllUsesWith(ConstVal);
				2859	</pre>
				2860	</div>
				2861
				2862	</ul>
				2863
				2864	</div>
				2865
				2866	<!-- ======================================================================= -->
				2867	<div class="doc_subsection">
				2868	<a name="User">The <tt>User</tt> class</a>
				2869	</div>
				2870
				2871	<div class="doc_text">
				2872
				2873	<p>
				2874	<tt>#include "<a href="/doxygen/User_8h-source.html">llvm/User.h</a>"</tt><br>
				2875	doxygen info: <a href="/doxygen/classllvm_1_1User.html">User Class</a><br>
				2876	Superclass: <a href="#Value"><tt>Value</tt></a></p>
				2877
				2878	<p>The <tt>User</tt> class is the common base class of all LLVM nodes that may
				2879	refer to <a href="#Value"><tt>Value</tt></a>s. It exposes a list of "Operands"
				2880	that are all of the <a href="#Value"><tt>Value</tt></a>s that the User is
				2881	referring to. The <tt>User</tt> class itself is a subclass of
				2882	<tt>Value</tt>.</p>
				2883
				2884	<p>The operands of a <tt>User</tt> point directly to the LLVM <a
				2885	href="#Value"><tt>Value</tt></a> that it refers to. Because LLVM uses Static
				2886	Single Assignment (SSA) form, there can only be one definition referred to,
				2887	allowing this direct connection. This connection provides the use-def
				2888	information in LLVM.</p>
				2889
				2890	</div>
				2891
				2892	<!-- _______________________________________________________________________ -->
				2893	<div class="doc_subsubsection">
				2894	<a name="m_User">Important Public Members of the <tt>User</tt> class</a>
				2895	</div>
				2896
				2897	<div class="doc_text">
				2898
				2899	<p>The <tt>User</tt> class exposes the operand list in two ways: through
				2900	an index access interface and through an iterator based interface.</p>
				2901
				2902	<ul>
				2903	<li><tt>Value *getOperand(unsigned i)</tt><br>
				2904	<tt>unsigned getNumOperands()</tt>
				2905	<p> These two methods expose the operands of the <tt>User</tt> in a
				2906	convenient form for direct access.</p></li>
				2907
				2908	<li><tt>User::op_iterator</tt> - Typedef for iterator over the operand
				2909	list<br>
				2910	<tt>op_iterator op_begin()</tt> - Get an iterator to the start of
				2911	the operand list.<br>
				2912	<tt>op_iterator op_end()</tt> - Get an iterator to the end of the
				2913	operand list.
				2914	<p> Together, these methods make up the iterator based interface to
				2915	the operands of a <tt>User</tt>.</p></li>
				2916	</ul>
				2917
				2918	</div>
				2919
				2920	<!-- ======================================================================= -->
				2921	<div class="doc_subsection">
				2922	<a name="Instruction">The <tt>Instruction</tt> class</a>
				2923	</div>
				2924
				2925	<div class="doc_text">
				2926
				2927	<p><tt>#include "</tt><tt><a
				2928	href="/doxygen/Instruction_8h-source.html">llvm/Instruction.h</a>"</tt><br>
				2929	doxygen info: <a href="/doxygen/classllvm_1_1Instruction.html">Instruction Class</a><br>
				2930	Superclasses: <a href="#User"><tt>User</tt></a>, <a
				2931	href="#Value"><tt>Value</tt></a></p>
				2932
				2933	<p>The <tt>Instruction</tt> class is the common base class for all LLVM
				2934	instructions. It provides only a few methods, but is a very commonly used
				2935	class. The primary data tracked by the <tt>Instruction</tt> class itself is the
				2936	opcode (instruction type) and the parent <a
				2937	href="#BasicBlock"><tt>BasicBlock</tt></a> the <tt>Instruction</tt> is embedded
				2938	into. To represent a specific type of instruction, one of many subclasses of
				2939	<tt>Instruction</tt> are used.</p>
				2940
				2941	<p> Because the <tt>Instruction</tt> class subclasses the <a
				2942	href="#User"><tt>User</tt></a> class, its operands can be accessed in the same
				2943	way as for other <a href="#User"><tt>User</tt></a>s (with the
				2944	<tt>getOperand()</tt>/<tt>getNumOperands()</tt> and
				2945	<tt>op_begin()</tt>/<tt>op_end()</tt> methods).</p> <p> An important file for
				2946	the <tt>Instruction</tt> class is the <tt>llvm/Instruction.def</tt> file. This
				2947	file contains some meta-data about the various different types of instructions
				2948	in LLVM. It describes the enum values that are used as opcodes (for example
				2949	<tt>Instruction::Add</tt> and <tt>Instruction::ICmp</tt>), as well as the
				2950	concrete sub-classes of <tt>Instruction</tt> that implement the instruction (for
				2951	example <tt><a href="#BinaryOperator">BinaryOperator</a></tt> and <tt><a
				2952	href="#CmpInst">CmpInst</a></tt>). Unfortunately, the use of macros in
				2953	this file confuses doxygen, so these enum values don't show up correctly in the
				2954	<a href="/doxygen/classllvm_1_1Instruction.html">doxygen output</a>.</p>
				2955
				2956	</div>
				2957
				2958	<!-- _______________________________________________________________________ -->
				2959	<div class="doc_subsubsection">
				2960	<a name="s_Instruction">Important Subclasses of the <tt>Instruction</tt>
				2961	class</a>
				2962	</div>
				2963	<div class="doc_text">
				2964	<ul>
				2965	<li><tt><a name="BinaryOperator">BinaryOperator</a></tt>
				2966	<p>This subclasses represents all two operand instructions whose operands
				2967	must be the same type, except for the comparison instructions.</p></li>
				2968	<li><tt><a name="CastInst">CastInst</a></tt>
				2969	<p>This subclass is the parent of the 12 casting instructions. It provides
				2970	common operations on cast instructions.</p>
				2971	<li><tt><a name="CmpInst">CmpInst</a></tt>
				2972	<p>This subclass respresents the two comparison instructions,
				2973	<a href="LangRef.html#i_icmp">ICmpInst</a> (integer opreands), and
				2974	<a href="LangRef.html#i_fcmp">FCmpInst</a> (floating point operands).</p>
				2975	<li><tt><a name="TerminatorInst">TerminatorInst</a></tt>
				2976	<p>This subclass is the parent of all terminator instructions (those which
				2977	can terminate a block).</p>
				2978	</ul>
				2979	</div>
				2980
				2981	<!-- _______________________________________________________________________ -->
				2982	<div class="doc_subsubsection">
				2983	<a name="m_Instruction">Important Public Members of the <tt>Instruction</tt>
				2984	class</a>
				2985	</div>
				2986
				2987	<div class="doc_text">
				2988
				2989	<ul>
				2990	<li><tt><a href="#BasicBlock">BasicBlock</a> *getParent()</tt>
				2991	<p>Returns the <a href="#BasicBlock"><tt>BasicBlock</tt></a> that
				2992	this <tt>Instruction</tt> is embedded into.</p></li>
				2993	<li><tt>bool mayWriteToMemory()</tt>
				2994	<p>Returns true if the instruction writes to memory, i.e. it is a
				2995	<tt>call</tt>,<tt>free</tt>,<tt>invoke</tt>, or <tt>store</tt>.</p></li>
				2996	<li><tt>unsigned getOpcode()</tt>
				2997	<p>Returns the opcode for the <tt>Instruction</tt>.</p></li>
				2998	<li><tt><a href="#Instruction">Instruction</a> *clone() const</tt>
				2999	<p>Returns another instance of the specified instruction, identical
				3000	in all ways to the original except that the instruction has no parent
				3001	(ie it's not embedded into a <a href="#BasicBlock"><tt>BasicBlock</tt></a>),
				3002	and it has no name</p></li>
				3003	</ul>
				3004
				3005	</div>
				3006
				3007	<!-- ======================================================================= -->
				3008	<div class="doc_subsection">
				3009	<a name="Constant">The <tt>Constant</tt> class and subclasses</a>
				3010	</div>
				3011
				3012	<div class="doc_text">
				3013
				3014	<p>Constant represents a base class for different types of constants. It
				3015	is subclassed by ConstantInt, ConstantArray, etc. for representing
				3016	the various types of Constants. <a href="#GlobalValue">GlobalValue</a> is also
				3017	a subclass, which represents the address of a global variable or function.
				3018	</p>
				3019
				3020	</div>
				3021
				3022	<!-- _______________________________________________________________________ -->
				3023	<div class="doc_subsubsection">Important Subclasses of Constant </div>
				3024	<div class="doc_text">
				3025	<ul>
				3026	<li>ConstantInt : This subclass of Constant represents an integer constant of
				3027	any width.
				3028	<ul>
				3029	<li><tt>const APInt& getValue() const</tt>: Returns the underlying
				3030	value of this constant, an APInt value.</li>
				3031	<li><tt>int64_t getSExtValue() const</tt>: Converts the underlying APInt
				3032	value to an int64_t via sign extension. If the value (not the bit width)
				3033	of the APInt is too large to fit in an int64_t, an assertion will result.
				3034	For this reason, use of this method is discouraged.</li>
				3035	<li><tt>uint64_t getZExtValue() const</tt>: Converts the underlying APInt
				3036	value to a uint64_t via zero extension. IF the value (not the bit width)
				3037	of the APInt is too large to fit in a uint64_t, an assertion will result.
				3038	For this reason, use of this method is discouraged.</li>
				3039	<li><tt>static ConstantInt* get(const APInt& Val)</tt>: Returns the
				3040	ConstantInt object that represents the value provided by <tt>Val</tt>.
				3041	The type is implied as the IntegerType that corresponds to the bit width
				3042	of <tt>Val</tt>.</li>
				3043	<li><tt>static ConstantInt* get(const Type *Ty, uint64_t Val)</tt>:
				3044	Returns the ConstantInt object that represents the value provided by
				3045	<tt>Val</tt> for integer type <tt>Ty</tt>.</li>
				3046	</ul>
				3047	</li>
				3048	<li>ConstantFP : This class represents a floating point constant.
				3049	<ul>
				3050	<li><tt>double getValue() const</tt>: Returns the underlying value of
				3051	this constant. </li>
				3052	</ul>
				3053	</li>
				3054	<li>ConstantArray : This represents a constant array.
				3055	<ul>
				3056	<li><tt>const std::vector<Use> &getValues() const</tt>: Returns
				3057	a vector of component constants that makeup this array. </li>
				3058	</ul>
				3059	</li>
				3060	<li>ConstantStruct : This represents a constant struct.
				3061	<ul>
				3062	<li><tt>const std::vector<Use> &getValues() const</tt>: Returns
				3063	a vector of component constants that makeup this array. </li>
				3064	</ul>
				3065	</li>
				3066	<li>GlobalValue : This represents either a global variable or a function. In
				3067	either case, the value is a constant fixed address (after linking).
				3068	</li>
				3069	</ul>
				3070	</div>
				3071
				3072
				3073	<!-- ======================================================================= -->
				3074	<div class="doc_subsection">
				3075	<a name="GlobalValue">The <tt>GlobalValue</tt> class</a>
				3076	</div>
				3077
				3078	<div class="doc_text">
				3079
				3080	<p><tt>#include "<a
				3081	href="/doxygen/GlobalValue_8h-source.html">llvm/GlobalValue.h</a>"</tt><br>
				3082	doxygen info: <a href="/doxygen/classllvm_1_1GlobalValue.html">GlobalValue
				3083	Class</a><br>
				3084	Superclasses: <a href="#Constant"><tt>Constant</tt></a>,
				3085	<a href="#User"><tt>User</tt></a>, <a href="#Value"><tt>Value</tt></a></p>
				3086
				3087	<p>Global values (<a href="#GlobalVariable"><tt>GlobalVariable</tt></a>s or <a
				3088	href="#Function"><tt>Function</tt></a>s) are the only LLVM values that are
				3089	visible in the bodies of all <a href="#Function"><tt>Function</tt></a>s.
				3090	Because they are visible at global scope, they are also subject to linking with
				3091	other globals defined in different translation units. To control the linking
				3092	process, <tt>GlobalValue</tt>s know their linkage rules. Specifically,
				3093	<tt>GlobalValue</tt>s know whether they have internal or external linkage, as
				3094	defined by the <tt>LinkageTypes</tt> enumeration.</p>
				3095
				3096	<p>If a <tt>GlobalValue</tt> has internal linkage (equivalent to being
				3097	<tt>static</tt> in C), it is not visible to code outside the current translation
				3098	unit, and does not participate in linking. If it has external linkage, it is
				3099	visible to external code, and does participate in linking. In addition to
				3100	linkage information, <tt>GlobalValue</tt>s keep track of which <a
				3101	href="#Module"><tt>Module</tt></a> they are currently part of.</p>
				3102
				3103	<p>Because <tt>GlobalValue</tt>s are memory objects, they are always referred to
				3104	by their <b>address</b>. As such, the <a href="#Type"><tt>Type</tt></a> of a
				3105	global is always a pointer to its contents. It is important to remember this
				3106	when using the <tt>GetElementPtrInst</tt> instruction because this pointer must
				3107	be dereferenced first. For example, if you have a <tt>GlobalVariable</tt> (a
				3108	subclass of <tt>GlobalValue)</tt> that is an array of 24 ints, type <tt>[24 x
				3109	i32]</tt>, then the <tt>GlobalVariable</tt> is a pointer to that array. Although
				3110	the address of the first element of this array and the value of the
				3111	<tt>GlobalVariable</tt> are the same, they have different types. The
				3112	<tt>GlobalVariable</tt>'s type is <tt>[24 x i32]</tt>. The first element's type
				3113	is <tt>i32.</tt> Because of this, accessing a global value requires you to
				3114	dereference the pointer with <tt>GetElementPtrInst</tt> first, then its elements
				3115	can be accessed. This is explained in the <a href="LangRef.html#globalvars">LLVM
				3116	Language Reference Manual</a>.</p>
				3117
				3118	</div>
				3119
				3120	<!-- _______________________________________________________________________ -->
				3121	<div class="doc_subsubsection">
				3122	<a name="m_GlobalValue">Important Public Members of the <tt>GlobalValue</tt>
				3123	class</a>
				3124	</div>
				3125
				3126	<div class="doc_text">
				3127
				3128	<ul>
				3129	<li><tt>bool hasInternalLinkage() const</tt><br>
				3130	<tt>bool hasExternalLinkage() const</tt><br>
				3131	<tt>void setInternalLinkage(bool HasInternalLinkage)</tt>
				3132	<p> These methods manipulate the linkage characteristics of the <tt>GlobalValue</tt>.</p>
				3133	<p> </p>
				3134	</li>
				3135	<li><tt><a href="#Module">Module</a> *getParent()</tt>
				3136	<p> This returns the <a href="#Module"><tt>Module</tt></a> that the
				3137	GlobalValue is currently embedded into.</p></li>
				3138	</ul>
				3139
				3140	</div>
				3141
				3142	<!-- ======================================================================= -->
				3143	<div class="doc_subsection">
				3144	<a name="Function">The <tt>Function</tt> class</a>
				3145	</div>
				3146
				3147	<div class="doc_text">
				3148
				3149	<p><tt>#include "<a
				3150	href="/doxygen/Function_8h-source.html">llvm/Function.h</a>"</tt><br> doxygen
				3151	info: <a href="/doxygen/classllvm_1_1Function.html">Function Class</a><br>
				3152	Superclasses: <a href="#GlobalValue"><tt>GlobalValue</tt></a>,
				3153	<a href="#Constant"><tt>Constant</tt></a>,
				3154	<a href="#User"><tt>User</tt></a>,
				3155	<a href="#Value"><tt>Value</tt></a></p>
				3156
				3157	<p>The <tt>Function</tt> class represents a single procedure in LLVM. It is
				3158	actually one of the more complex classes in the LLVM heirarchy because it must
				3159	keep track of a large amount of data. The <tt>Function</tt> class keeps track
				3160	of a list of <a href="#BasicBlock"><tt>BasicBlock</tt></a>s, a list of formal
				3161	<a href="#Argument"><tt>Argument</tt></a>s, and a
				3162	<a href="#SymbolTable"><tt>SymbolTable</tt></a>.</p>
				3163
				3164	<p>The list of <a href="#BasicBlock"><tt>BasicBlock</tt></a>s is the most
				3165	commonly used part of <tt>Function</tt> objects. The list imposes an implicit
				3166	ordering of the blocks in the function, which indicate how the code will be
				3167	layed out by the backend. Additionally, the first <a
				3168	href="#BasicBlock"><tt>BasicBlock</tt></a> is the implicit entry node for the
				3169	<tt>Function</tt>. It is not legal in LLVM to explicitly branch to this initial
				3170	block. There are no implicit exit nodes, and in fact there may be multiple exit
				3171	nodes from a single <tt>Function</tt>. If the <a
				3172	href="#BasicBlock"><tt>BasicBlock</tt></a> list is empty, this indicates that
				3173	the <tt>Function</tt> is actually a function declaration: the actual body of the
				3174	function hasn't been linked in yet.</p>
				3175
				3176	<p>In addition to a list of <a href="#BasicBlock"><tt>BasicBlock</tt></a>s, the
				3177	<tt>Function</tt> class also keeps track of the list of formal <a
				3178	href="#Argument"><tt>Argument</tt></a>s that the function receives. This
				3179	container manages the lifetime of the <a href="#Argument"><tt>Argument</tt></a>
				3180	nodes, just like the <a href="#BasicBlock"><tt>BasicBlock</tt></a> list does for
				3181	the <a href="#BasicBlock"><tt>BasicBlock</tt></a>s.</p>
				3182
				3183	<p>The <a href="#SymbolTable"><tt>SymbolTable</tt></a> is a very rarely used
				3184	LLVM feature that is only used when you have to look up a value by name. Aside
				3185	from that, the <a href="#SymbolTable"><tt>SymbolTable</tt></a> is used
				3186	internally to make sure that there are not conflicts between the names of <a
				3187	href="#Instruction"><tt>Instruction</tt></a>s, <a
				3188	href="#BasicBlock"><tt>BasicBlock</tt></a>s, or <a
				3189	href="#Argument"><tt>Argument</tt></a>s in the function body.</p>
				3190
				3191	<p>Note that <tt>Function</tt> is a <a href="#GlobalValue">GlobalValue</a>
				3192	and therefore also a <a href="#Constant">Constant</a>. The value of the function
				3193	is its address (after linking) which is guaranteed to be constant.</p>
				3194	</div>
				3195
				3196	<!-- _______________________________________________________________________ -->
				3197	<div class="doc_subsubsection">
				3198	<a name="m_Function">Important Public Members of the <tt>Function</tt>
				3199	class</a>
				3200	</div>
				3201
				3202	<div class="doc_text">
				3203
				3204	<ul>
				3205	<li><tt>Function(const </tt><tt><a href="#FunctionType">FunctionType</a>
				3206	Ty, LinkageTypes Linkage, const std::string &N = "", Module Parent = 0)</tt>
				3207
				3208	<p>Constructor used when you need to create new <tt>Function</tt>s to add
				3209	the the program. The constructor must specify the type of the function to
				3210	create and what type of linkage the function should have. The <a
				3211	href="#FunctionType"><tt>FunctionType</tt></a> argument
				3212	specifies the formal arguments and return value for the function. The same
				3213	<a href="#FunctionType"><tt>FunctionType</tt></a> value can be used to
				3214	create multiple functions. The <tt>Parent</tt> argument specifies the Module
				3215	in which the function is defined. If this argument is provided, the function
				3216	will automatically be inserted into that module's list of
				3217	functions.</p></li>
				3218
Chris Lattner	5e70957	2008-11-25 18:34:50 +0000	[diff] [blame]	3219	<li><tt>bool isDeclaration()</tt>
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	3220
				3221	<p>Return whether or not the <tt>Function</tt> has a body defined. If the
				3222	function is "external", it does not have a body, and thus must be resolved
				3223	by linking with a function defined in a different translation unit.</p></li>
				3224
				3225	<li><tt>Function::iterator</tt> - Typedef for basic block list iterator<br>
				3226	<tt>Function::const_iterator</tt> - Typedef for const_iterator.<br>
				3227
				3228	<tt>begin()</tt>, <tt>end()</tt>
				3229	<tt>size()</tt>, <tt>empty()</tt>
				3230
				3231	<p>These are forwarding methods that make it easy to access the contents of
				3232	a <tt>Function</tt> object's <a href="#BasicBlock"><tt>BasicBlock</tt></a>
				3233	list.</p></li>
				3234
				3235	<li><tt>Function::BasicBlockListType &getBasicBlockList()</tt>
				3236
				3237	<p>Returns the list of <a href="#BasicBlock"><tt>BasicBlock</tt></a>s. This
				3238	is necessary to use when you need to update the list or perform a complex
				3239	action that doesn't have a forwarding method.</p></li>
				3240
				3241	<li><tt>Function::arg_iterator</tt> - Typedef for the argument list
				3242	iterator<br>
				3243	<tt>Function::const_arg_iterator</tt> - Typedef for const_iterator.<br>
				3244
				3245	<tt>arg_begin()</tt>, <tt>arg_end()</tt>
				3246	<tt>arg_size()</tt>, <tt>arg_empty()</tt>
				3247
				3248	<p>These are forwarding methods that make it easy to access the contents of
				3249	a <tt>Function</tt> object's <a href="#Argument"><tt>Argument</tt></a>
				3250	list.</p></li>
				3251
				3252	<li><tt>Function::ArgumentListType &getArgumentList()</tt>
				3253
				3254	<p>Returns the list of <a href="#Argument"><tt>Argument</tt></a>s. This is
				3255	necessary to use when you need to update the list or perform a complex
				3256	action that doesn't have a forwarding method.</p></li>
				3257
				3258	<li><tt><a href="#BasicBlock">BasicBlock</a> &getEntryBlock()</tt>
				3259
				3260	<p>Returns the entry <a href="#BasicBlock"><tt>BasicBlock</tt></a> for the
				3261	function. Because the entry block for the function is always the first
				3262	block, this returns the first block of the <tt>Function</tt>.</p></li>
				3263
				3264	<li><tt><a href="#Type">Type</a> *getReturnType()</tt><br>
				3265	<tt><a href="#FunctionType">FunctionType</a> *getFunctionType()</tt>
				3266
				3267	<p>This traverses the <a href="#Type"><tt>Type</tt></a> of the
				3268	<tt>Function</tt> and returns the return type of the function, or the <a
				3269	href="#FunctionType"><tt>FunctionType</tt></a> of the actual
				3270	function.</p></li>
				3271
				3272	<li><tt><a href="#SymbolTable">SymbolTable</a> *getSymbolTable()</tt>
				3273
				3274	<p> Return a pointer to the <a href="#SymbolTable"><tt>SymbolTable</tt></a>
				3275	for this <tt>Function</tt>.</p></li>
				3276	</ul>
				3277
				3278	</div>
				3279
				3280	<!-- ======================================================================= -->
				3281	<div class="doc_subsection">
				3282	<a name="GlobalVariable">The <tt>GlobalVariable</tt> class</a>
				3283	</div>
				3284
				3285	<div class="doc_text">
				3286
				3287	<p><tt>#include "<a
				3288	href="/doxygen/GlobalVariable_8h-source.html">llvm/GlobalVariable.h</a>"</tt>
				3289	<br>
				3290	doxygen info: <a href="/doxygen/classllvm_1_1GlobalVariable.html">GlobalVariable
				3291	Class</a><br>
				3292	Superclasses: <a href="#GlobalValue"><tt>GlobalValue</tt></a>,
				3293	<a href="#Constant"><tt>Constant</tt></a>,
				3294	<a href="#User"><tt>User</tt></a>,
				3295	<a href="#Value"><tt>Value</tt></a></p>
				3296
				3297	<p>Global variables are represented with the (suprise suprise)
				3298	<tt>GlobalVariable</tt> class. Like functions, <tt>GlobalVariable</tt>s are also
				3299	subclasses of <a href="#GlobalValue"><tt>GlobalValue</tt></a>, and as such are
				3300	always referenced by their address (global values must live in memory, so their
				3301	"name" refers to their constant address). See
				3302	<a href="#GlobalValue"><tt>GlobalValue</tt></a> for more on this. Global
				3303	variables may have an initial value (which must be a
				3304	<a href="#Constant"><tt>Constant</tt></a>), and if they have an initializer,
				3305	they may be marked as "constant" themselves (indicating that their contents
				3306	never change at runtime).</p>
				3307	</div>
				3308
				3309	<!-- _______________________________________________________________________ -->
				3310	<div class="doc_subsubsection">
				3311	<a name="m_GlobalVariable">Important Public Members of the
				3312	<tt>GlobalVariable</tt> class</a>
				3313	</div>
				3314
				3315	<div class="doc_text">
				3316
				3317	<ul>
				3318	<li><tt>GlobalVariable(const </tt><tt><a href="#Type">Type</a> *Ty, bool
				3319	isConstant, LinkageTypes& Linkage, <a href="#Constant">Constant</a>
				3320	Initializer = 0, const std::string &Name = "", Module Parent = 0)</tt>
				3321
				3322	<p>Create a new global variable of the specified type. If
				3323	<tt>isConstant</tt> is true then the global variable will be marked as
				3324	unchanging for the program. The Linkage parameter specifies the type of
Duncan Sands	19d161f	2009-03-07 15:45:40 +0000	[diff] [blame^]	3325	linkage (internal, external, weak, linkonce, appending) for the variable.
				3326	If the linkage is InternalLinkage, WeakAnyLinkage, WeakODRLinkage,
				3327	LinkOnceAnyLinkage or LinkOnceODRLinkage,  then the resultant
				3328	global variable will have internal linkage. AppendingLinkage concatenates
				3329	together all instances (in different translation units) of the variable
				3330	into a single variable but is only applicable to arrays.  See
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	3331	the <a href="LangRef.html#modulestructure">LLVM Language Reference</a> for
				3332	further details on linkage types. Optionally an initializer, a name, and the
				3333	module to put the variable into may be specified for the global variable as
				3334	well.</p></li>
				3335
				3336	<li><tt>bool isConstant() const</tt>
				3337
				3338	<p>Returns true if this is a global variable that is known not to
				3339	be modified at runtime.</p></li>
				3340
				3341	<li><tt>bool hasInitializer()</tt>
				3342
				3343	<p>Returns true if this <tt>GlobalVariable</tt> has an intializer.</p></li>
				3344
				3345	<li><tt><a href="#Constant">Constant</a> *getInitializer()</tt>
				3346
				3347	<p>Returns the intial value for a <tt>GlobalVariable</tt>. It is not legal
				3348	to call this method if there is no initializer.</p></li>
				3349	</ul>
				3350
				3351	</div>
				3352
				3353
				3354	<!-- ======================================================================= -->
				3355	<div class="doc_subsection">
				3356	<a name="BasicBlock">The <tt>BasicBlock</tt> class</a>
				3357	</div>
				3358
				3359	<div class="doc_text">
				3360
				3361	<p><tt>#include "<a
				3362	href="/doxygen/BasicBlock_8h-source.html">llvm/BasicBlock.h</a>"</tt><br>
				3363	doxygen info: <a href="/doxygen/structllvm_1_1BasicBlock.html">BasicBlock
				3364	Class</a><br>
				3365	Superclass: <a href="#Value"><tt>Value</tt></a></p>
				3366
				3367	<p>This class represents a single entry multiple exit section of the code,
				3368	commonly known as a basic block by the compiler community. The
				3369	<tt>BasicBlock</tt> class maintains a list of <a
				3370	href="#Instruction"><tt>Instruction</tt></a>s, which form the body of the block.
				3371	Matching the language definition, the last element of this list of instructions
				3372	is always a terminator instruction (a subclass of the <a
				3373	href="#TerminatorInst"><tt>TerminatorInst</tt></a> class).</p>
				3374
				3375	<p>In addition to tracking the list of instructions that make up the block, the
				3376	<tt>BasicBlock</tt> class also keeps track of the <a
				3377	href="#Function"><tt>Function</tt></a> that it is embedded into.</p>
				3378
				3379	<p>Note that <tt>BasicBlock</tt>s themselves are <a
				3380	href="#Value"><tt>Value</tt></a>s, because they are referenced by instructions
				3381	like branches and can go in the switch tables. <tt>BasicBlock</tt>s have type
				3382	<tt>label</tt>.</p>
				3383
				3384	</div>
				3385
				3386	<!-- _______________________________________________________________________ -->
				3387	<div class="doc_subsubsection">
				3388	<a name="m_BasicBlock">Important Public Members of the <tt>BasicBlock</tt>
				3389	class</a>
				3390	</div>
				3391
				3392	<div class="doc_text">
				3393	<ul>
				3394
				3395	<li><tt>BasicBlock(const std::string &Name = "", </tt><tt><a
				3396	href="#Function">Function</a> *Parent = 0)</tt>
				3397
				3398	<p>The <tt>BasicBlock</tt> constructor is used to create new basic blocks for
				3399	insertion into a function. The constructor optionally takes a name for the new
				3400	block, and a <a href="#Function"><tt>Function</tt></a> to insert it into. If
				3401	the <tt>Parent</tt> parameter is specified, the new <tt>BasicBlock</tt> is
				3402	automatically inserted at the end of the specified <a
				3403	href="#Function"><tt>Function</tt></a>, if not specified, the BasicBlock must be
				3404	manually inserted into the <a href="#Function"><tt>Function</tt></a>.</p></li>
				3405
				3406	<li><tt>BasicBlock::iterator</tt> - Typedef for instruction list iterator<br>
				3407	<tt>BasicBlock::const_iterator</tt> - Typedef for const_iterator.<br>
				3408	<tt>begin()</tt>, <tt>end()</tt>, <tt>front()</tt>, <tt>back()</tt>,
				3409	<tt>size()</tt>, <tt>empty()</tt>
				3410	STL-style functions for accessing the instruction list.
				3411
				3412	<p>These methods and typedefs are forwarding functions that have the same
				3413	semantics as the standard library methods of the same names. These methods
				3414	expose the underlying instruction list of a basic block in a way that is easy to
				3415	manipulate. To get the full complement of container operations (including
				3416	operations to update the list), you must use the <tt>getInstList()</tt>
				3417	method.</p></li>
				3418
				3419	<li><tt>BasicBlock::InstListType &getInstList()</tt>
				3420
				3421	<p>This method is used to get access to the underlying container that actually
				3422	holds the Instructions. This method must be used when there isn't a forwarding
				3423	function in the <tt>BasicBlock</tt> class for the operation that you would like
				3424	to perform. Because there are no forwarding functions for "updating"
				3425	operations, you need to use this if you want to update the contents of a
				3426	<tt>BasicBlock</tt>.</p></li>
				3427
				3428	<li><tt><a href="#Function">Function</a> *getParent()</tt>
				3429
				3430	<p> Returns a pointer to <a href="#Function"><tt>Function</tt></a> the block is
				3431	embedded into, or a null pointer if it is homeless.</p></li>
				3432
				3433	<li><tt><a href="#TerminatorInst">TerminatorInst</a> *getTerminator()</tt>
				3434
				3435	<p> Returns a pointer to the terminator instruction that appears at the end of
				3436	the <tt>BasicBlock</tt>. If there is no terminator instruction, or if the last
				3437	instruction in the block is not a terminator, then a null pointer is
				3438	returned.</p></li>
				3439
				3440	</ul>
				3441
				3442	</div>
				3443
				3444
				3445	<!-- ======================================================================= -->
				3446	<div class="doc_subsection">
				3447	<a name="Argument">The <tt>Argument</tt> class</a>
				3448	</div>
				3449
				3450	<div class="doc_text">
				3451
				3452	<p>This subclass of Value defines the interface for incoming formal
				3453	arguments to a function. A Function maintains a list of its formal
				3454	arguments. An argument has a pointer to the parent Function.</p>
				3455
				3456	</div>
				3457
				3458	<!-- *********************************************************************** -->
				3459	<hr>
				3460	<address>
				3461	<a href="http://jigsaw.w3.org/css-validator/check/referer"><img
Misha Brukman	947321d	2008-12-11 17:34:48 +0000	[diff] [blame]	3462	src="http://jigsaw.w3.org/css-validator/images/vcss-blue" alt="Valid CSS"></a>
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	3463	<a href="http://validator.w3.org/check/referer"><img
Gabor Greif	7dabe38	2008-06-18 14:05:31 +0000	[diff] [blame]	3464	src="http://www.w3.org/Icons/valid-html401" alt="Valid HTML 4.01 Strict"></a>
Dan Gohman	f17a25c	2007-07-18 16:29:46 +0000	[diff] [blame]	3465
				3466	<a href="mailto:dhurjati@cs.uiuc.edu">Dinakar Dhurjati</a> and
				3467	<a href="mailto:sabre@nondot.org">Chris Lattner</a><br>
				3468	<a href="http://llvm.org">The LLVM Compiler Infrastructure</a><br>
				3469	Last modified: $Date$
				3470	</address>
				3471
				3472	</body>
				3473	</html>