Blame - docs/WritingAnLLVMPass.html - fp2-dev/platform/external/llvm

blob: 3e2f167ccbb2b9f016fd9749fb9e0c0c3ba80dc6 [file] [log] [blame]

Chris Lattner	c6bb824	2002-08-08 20:11:18 +0000	[diff] [blame]	1	<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
				2	<html><head><title>Writing an LLVM Pass</title></head>
				3
				4	<!--
				5	I. General Structure of an LLVM Program
				6
				7	I.1 "What is a 'Value'?": Value & User class
				8	I.2 Type & Derived Types
				9	I.3 GlobalVariable, Function
				10	I.4 BasicBlock
				11	I.5 Instruction & Subclasses
				12	1.6 Argument
				13	1.7 Constants
				14
				15	III. Useful things to know about the LLVM source base:
				16
				17	III.1 Useful links that introduce the STL
				18	III.2 isa<>, cast<>, dyn_cast<>
				19	III.3 Makefiles, useful options
				20	III.4 How to use opt & analyze to debug stuff
				21	III.5 How to write a regression test
				22	III.6 DEBUG() and Statistics (-debug & -stats)
				23	III.7 The -time-passes option
				24	III.8 ... more as needed ...
				25
				26	I think that writing Section #1 would be very helpful and that's the most
				27	stable portion of the sourcebase. #3 can be started on, but will probably
Chris Lattner	7991070	2002-08-22 19:21:04 +0000	[diff] [blame^]	28	just grow as time goes on.
Chris Lattner	c6bb824	2002-08-08 20:11:18 +0000	[diff] [blame]	29
				30	-->
				31
				32	<body bgcolor=white>
				33
				34	<table width="100%" bgcolor="#330077" border=0 cellpadding=4 cellspacing=0>
				35	<tr><td>  <font size=+3 color="#EEEEFF" face="Georgia,Palatino,Times,Roman"><b>Writing an LLVM Pass</b></font></td>
				36	</tr></table>
				37
				38
				39	<ol>
				40	<li><a href="#introduction">Introduction - What is a pass?</a>
				41	<li><a href="#quickstart">Quick Start - Writing hello world</a>
				42	<ul>
				43	<li><a href="#makefile">Setting up the build environment</a>
				44	<li><a href="#basiccode">Basic code required</a>
				45	<li><a href="#running">Running a pass with <tt>opt</tt>
				46	or <tt>analyze</tt></a>
				47	</ul>
				48	<li><a href="#passtype">Pass classes and requirements</a>
				49	<ul>
				50	<li><a href="#Pass">The <tt>Pass</tt> class</a>
				51	<ul>
				52	<li><a href="#run">The <tt>run</tt> method</a>
				53	</ul>
				54	<li><a href="#FunctionPass">The <tt>FunctionPass</tt> class</a>
				55	<ul>
				56	<li><a href="#doInitialization">The <tt>doInitialization</tt> method</a>
				57	<li><a href="#runOnFunction">The <tt>runOnFunction</tt> method</a>
				58	<li><a href="#doFinalization">The <tt>doFinalization</tt> method</a>
				59	</ul>
				60	<li><a href="#BasicBlockPass">The <tt>BasicBlockPass</tt> class</a>
				61	<ul>
				62	<li><a href="#runOnBasicBlock">The <tt>runOnBasicBlock</tt> method</a>
				63	</ul>
				64	</ul>
				65	<li><a href="#registration">Pass Registration</a>
				66	<ul>
				67	<li><a href="#print">The <tt>print</tt> method</a>
				68	</ul>
				69	<li><a href="#interaction">Specifying interactions between passes</a>
				70	<ul>
				71	<li><a href="#getAnalysisUsage">The <tt>getAnalysisUsage</tt> method</a>
				72	<li><a href="#getAnalysis">The <tt>getAnalysis</tt> method</a>
				73	</ul>
Chris Lattner	7991070	2002-08-22 19:21:04 +0000	[diff] [blame^]	74	<li><a href="#analysisgroup">Implementing Analysis Groups</a>
				75	<ul>
				76	<li><a href="#agconcepts">Analysis Group Concepts</a>
				77	<li><a href="#registerag">Using <tt>RegisterAnalysisGroup</tt></a>
				78	</ul>
Chris Lattner	c6bb824	2002-08-08 20:11:18 +0000	[diff] [blame]	79	<li><a href="#passmanager">What PassManager does</a>
				80	<ul>
				81	<li><a href="#releaseMemory">The <tt>releaseMemory</tt> method</a>
				82	</ul>
				83	<li><a href="#future">Future extensions planned</a>
				84	<ul>
				85	<li><a href="#SMP">Multithreaded LLVM</a>
				86	<li><a href="#ModuleSource">A new <tt>ModuleSource</tt> interface</a>
				87	<li><a href="#PassFunctionPass"><tt>Pass</tt>'s requiring <tt>FunctionPass</tt>'s</a>
				88	</ul>
Chris Lattner	38c633d	2002-08-08 20:23:41 +0000	[diff] [blame]	89
				90	<p><b>Written by <a href="mailto:sabre@nondot.org">Chris Lattner</a></b><p>
Chris Lattner	c6bb824	2002-08-08 20:11:18 +0000	[diff] [blame]	91	</ol><p>
				92
				93
				94
				95	<!-- *********************************************************************** -->
				96	<table width="100%" bgcolor="#330077" border=0 cellpadding=4 cellspacing=0>
				97	<tr><td align=center><font color="#EEEEFF" size=+2 face="Georgia,Palatino"><b>
				98	<a name="introduction">Introduction - What is a pass?
				99	</b></font></td></tr></table><ul>
				100	<!-- *********************************************************************** -->
				101
				102	The LLVM Pass Framework is an important part of the LLVM system, because LLVM
				103	passes are where the interesting parts of the compiler exist. Passes perform
				104	the transformations and optimizations that make up the compiler, they build
				105	the analysis results that are used by these transformations, and they are, above
				106	all, a structuring technique for compiler code.<p>
				107
				108	All LLVM passes are subclasses of the <tt><a
				109	href="http://llvm.cs.uiuc.edu/doxygen/classPass.html">Pass</a></tt> class, which
				110	implement functionality by overriding virtual methods inherited from
				111	<tt>Pass</tt>. Depending on how your pass works, you may be able to inherit
				112	from the <tt><a
				113	href="http://llvm.cs.uiuc.edu/doxygen/structFunctionPass.html">FunctionPass</a></tt>
				114	or <tt><a
				115	href="http://llvm.cs.uiuc.edu/doxygen/structBasicBlockPass.html">BasicBlockPass</a></tt>,
				116	which gives the system more information about what your pass does, and how it
				117	can be combined with other passes. One of the main features of the LLVM Pass
				118	Framework is that it schedules passes to run in an efficient way based on the
				119	constraints that your pass has.<p>
				120
				121	We start by showing you how to construct a pass, everything from setting up the
				122	code, to compiling, loading, and executing it. After the basics are down, more
				123	advanced features are discussed.<p>
				124
				125
				126	<!-- *********************************************************************** -->
				127	</ul><table width="100%" bgcolor="#330077" border=0 cellpadding=4 cellspacing=0>
				128	<tr><td align=center><font color="#EEEEFF" size=+2 face="Georgia,Palatino"><b>
				129	<a name="quickstart">Quick Start - Writing hello world
				130	</b></font></td></tr></table><ul>
				131	<!-- *********************************************************************** -->
				132
				133	Here we describe how to write the "hello world" of passes. The "Hello" pass is
				134	designed to simply print out the name of non-external functions that exist in
				135	the program being compiled. It does not modify the program at all, just
				136	inspects it. The source code and files for this pass are available in the LLVM
				137	source tree in the <tt>lib/Transforms/Hello</tt> directory.<p>
				138
				139
				140	<!-- ======================================================================= -->
				141	</ul><table width="100%" bgcolor="#441188" border=0 cellpadding=4 cellspacing=0>
				142	<tr><td> </td><td width="100%">
				143	<font color="#EEEEFF" face="Georgia,Palatino"><b>
				144	<a name="makefile">Setting up the build environment
				145	</b></font></td></tr></table><ul>
				146
				147	First thing you need to do is create a new directory somewhere in the LLVM
				148	source base. For this example, we'll assume that you made
				149	"<tt>lib/Transforms/Hello</tt>". The first thing you must do is set up a build
				150	script (Makefile) that will compile the source code for the new pass. To do
Chris Lattner	17a4c3e	2002-08-14 20:06:13 +0000	[diff] [blame]	151	this, copy this into "<tt>Makefile</tt>" (be very careful that there are no
				152	extra space characters at the end of the lines though... that seems to confuse
				153	<tt>gmake</tt>):<p>
Chris Lattner	c6bb824	2002-08-08 20:11:18 +0000	[diff] [blame]	154
				155	</ul><hr><ul><pre>
				156	# Makefile for hello pass
Chris Lattner	c6bb824	2002-08-08 20:11:18 +0000	[diff] [blame]	157
Chris Lattner	17a4c3e	2002-08-14 20:06:13 +0000	[diff] [blame]	158	# Path to top level of LLVM heirarchy
Chris Lattner	7ce83e5	2002-08-14 20:07:01 +0000	[diff] [blame]	159	LEVEL = ../../..
Chris Lattner	17a4c3e	2002-08-14 20:06:13 +0000	[diff] [blame]	160
				161	# Name of the library to build
Chris Lattner	7ce83e5	2002-08-14 20:07:01 +0000	[diff] [blame]	162	LIBRARYNAME = hello
Chris Lattner	17a4c3e	2002-08-14 20:06:13 +0000	[diff] [blame]	163
				164	# Build a dynamically loadable shared object
				165	SHARED_LIBRARY = 1
				166
				167	# Include the makefile implementation stuff
				168	include $(LEVEL)/Makefile.common
Chris Lattner	c6bb824	2002-08-08 20:11:18 +0000	[diff] [blame]	169	</pre></ul><hr><ul><p>
				170
				171	This makefile specifies that all of the <tt>.cpp</tt> files in the current
				172	directory are to be compiled and linked together into a
				173	<tt>lib/Debug/libhello.so</tt> shared object that can be dynamically loaded by
				174	the <tt>opt</tt> or <tt>analyze</tt> tools.<p>
				175
				176	Now that we have the build scripts set up, we just need to write the code for
				177	the pass itself.<p>
				178
				179
				180	<!-- ======================================================================= -->
				181	</ul><table width="100%" bgcolor="#441188" border=0 cellpadding=4 cellspacing=0>
				182	<tr><td> </td><td width="100%">
				183	<font color="#EEEEFF" face="Georgia,Palatino"><b>
				184	<a name="basiccode">Basic code required
				185	</b></font></td></tr></table><ul>
				186
				187	Now that we have a way to compile our new pass, we just have to write it. Start
				188	out with:<p>
				189
				190	<pre>
				191	<b>#include</b> "<a href="http://llvm.cs.uiuc.edu/doxygen/Pass_8h-source.html">llvm/Pass.h</a>"
				192	<b>#include</b> "<a href="http://llvm.cs.uiuc.edu/doxygen/Function_8h-source.html">llvm/Function.h</a>"
				193	</pre>
				194
				195	Which are needed because we are writing a <tt><a
				196	href="http://llvm.cs.uiuc.edu/doxygen/classPass.html">Pass</a></tt>, and we are
				197	operating on <tt><a
				198	href="http://llvm.cs.uiuc.edu/doxygen/classFunction.html">Function</a></tt>'s.<p>
				199
				200	Next we have:<p>
				201
				202	<pre>
				203	<b>namespace</b> {
				204	</pre><p>
				205
				206	... which starts out an anonymous namespace. Anonymous namespaces are to C++
				207	what the "<tt>static</tt>" keyword is to C (at global scope). It makes the
				208	things declared inside of the anonymous namespace only visible to the current
				209	file. If you're not familiar with them, consult a decent C++ book for more
				210	information.<p>
				211
				212	Next, we declare our pass itself:<p>
				213
				214	<pre>
				215	<b>struct</b> Hello : <b>public</b> <a href="#FunctionPass">FunctionPass</a> {
				216	</pre><p>
				217
				218	This declares a "<tt>Hello</tt>" class that is a subclass of <tt><a
				219	href="http://llvm.cs.uiuc.edu/doxygen/structFunctionPass.html">FunctionPass</a></tt>.
				220	The different builting pass subclasses are described in detail <a
				221	href="#passtype">later</a>, but for now, know that <a href="#FunctionPass"><tt>FunctionPass</tt></a>'s
				222	operate a function at a time.<p>
				223
				224	<pre>
				225	<b>virtual bool</b> <a href="#runOnFunction">runOnFunction</a>(Function &F) {
				226	std::cerr << "<i>Hello: </i>" << F.getName() << "\n";
				227	<b>return false</b>;
				228	}
				229	}; <i>// end of struct Hello</i>
				230	</pre>
				231
				232	We declare a "<a href="#runOnFunction"><tt>runOnFunction</tt></a>" method, which
				233	overloads an abstract virtual method inherited from <a
				234	href="#FunctionPass"><tt>FunctionPass</tt></a>. This is where we are supposed
				235	to do our thing, so we just print out our message with the name of each
				236	function.<p>
				237
				238	<pre>
				239	RegisterOpt<Hello> X("<i>hello</i>", "<i>Hello World Pass</i>");
				240	} <i>// end of anonymous namespace</i>
				241	</pre><p>
				242
				243	Lastly, we register our class <tt>Hello</tt>, giving it a command line argument
				244	"<tt>hello</tt>", and a name "<tt>Hello World Pass</tt>". There are several
				245	different ways of <a href="#registration">registering your pass</a>, depending
				246	on what it is to be used for. For "optimizations" we use the
				247	<tt>RegisterOpt</tt> template.<p>
				248
				249	As a whole, the <tt>.cpp</tt> file looks like:<p>
				250
				251	<pre>
				252	<b>#include</b> "<a href="http://llvm.cs.uiuc.edu/doxygen/Pass_8h-source.html">llvm/Pass.h</a>"
				253	<b>#include</b> "<a href="http://llvm.cs.uiuc.edu/doxygen/Function_8h-source.html">llvm/Function.h</a>"
				254
				255	<b>namespace</b> {
				256	<b>struct Hello</b> : <b>public</b> <a href="#FunctionPass">FunctionPass</a> {
				257	<b>virtual bool</b> <a href="#runOnFunction">runOnFunction</a>(Function &F) {
				258	std::cerr << "<i>Hello: </i>" << F.getName() << "\n";
				259	<b>return false</b>;
				260	}
				261	};
				262
				263	RegisterOpt<Hello> X("<i>hello</i>", "<i>Hello World Pass</i>");
				264	}
				265	</pre><p>
				266
				267	Now that it's all together, compile the file with a simple "<tt>gmake</tt>"
				268	command in the local directory and you should get a new
				269	"<tt>lib/Debug/libhello.so</tt> file. Note that everything in this file is
				270	contained in an anonymous namespace: this reflects the fact that passes are self
				271	contained units that do not need external interfaces (although they can have
				272	them) to be useful.<p>
				273
				274
				275	<!-- ======================================================================= -->
				276	</ul><table width="100%" bgcolor="#441188" border=0 cellpadding=4 cellspacing=0>
				277	<tr><td> </td><td width="100%">
				278	<font color="#EEEEFF" face="Georgia,Palatino"><b>
				279	<a name="running">Running a pass with <tt>opt</tt> or <tt>analyze</tt>
				280	</b></font></td></tr></table><ul>
				281
				282	Now that you have a brand new shiny <tt>.so</tt> file, we can use the
				283	<tt>opt</tt> command to run an LLVM program through your pass. Because you
				284	registered your pass with the <tt>RegisterOpt</tt> template, you will be able to
				285	use the <tt>opt</tt> tool to access it, once loaded.<p>
				286
				287	To test it, follow the example at the end of the <a
				288	href="GettingStarted.html">Getting Started Guide</a> to compile "Hello World" to
				289	LLVM. We can now run the bytecode file (<tt>hello.bc</tt>) for the program
				290	through our transformation like this (or course, any bytecode file will
				291	work):<p>
				292
				293	<pre>
				294	$ opt -load ../../../lib/Debug/libhello.so -hello < hello.bc > /dev/null
				295	Hello: __main
				296	Hello: puts
				297	Hello: main
				298	</pre><p>
				299
				300	The '<tt>-load</tt>' option specifies that '<tt>opt</tt>' should load your pass
				301	as a shared object, which makes '<tt>-hello</tt>' a valid command line argument
				302	(which is one reason you need to <a href="#registration">register your
				303	pass</a>). Because the hello pass does not modify the program in any
				304	interesting way, we just throw away the result of <tt>opt</tt> (sending it to
				305	<tt>/dev/null</tt>).<p>
				306
				307	To see what happened to the other string you registered, try running
				308	<tt>opt</tt> with the <tt>--help</tt> option:<p>
				309
				310	<pre>
				311	$ opt -load ../../../lib/Debug/libhello.so --help
				312	OVERVIEW: llvm .bc -> .bc modular optimizer
				313
				314	USAGE: opt [options] <input bytecode>
				315
				316	OPTIONS:
				317	Optimizations available:
				318	...
				319	-funcresolve - Resolve Functions
				320	-gcse - Global Common Subexpression Elimination
				321	-globaldce - Dead Global Elimination
				322	<b>-hello - Hello World Pass</b>
				323	-indvars - Cannonicalize Induction Variables
				324	-inline - Function Integration/Inlining
				325	-instcombine - Combine redundant instructions
				326	...
				327	</pre><p>
				328
				329	The pass name get added as the information string for your pass, giving some
				330	documentation to users of <tt>opt</tt>. Now that you have a working pass, you
				331	would go ahead and make it do the cool transformations you want. Once you get
				332	it all working and tested, it may become useful to find out how fast your pass
				333	is. The <a href="#passManager"><tt>PassManager</tt></a> provides a nice command
				334	line option (<tt>--time-passes</tt>) that allows you to get information about
				335	the execution time of your pass along with the other passes you queue up. For
				336	example:<p>
				337
				338	<pre>
				339	$ opt -load ../../../lib/Debug/libhello.so -hello -time-passes < hello.bc > /dev/null
				340	Hello: __main
				341	Hello: puts
				342	Hello: main
				343	===============================================================================
				344	... Pass execution timing report ...
				345	===============================================================================
				346	Total Execution Time: 0.02 seconds (0.0479059 wall clock)
				347
				348	---User Time--- --System Time-- --User+System-- ---Wall Time--- --- Pass Name ---
				349	0.0100 (100.0%) 0.0000 ( 0.0%) 0.0100 ( 50.0%) 0.0402 ( 84.0%) Bytecode Writer
				350	0.0000 ( 0.0%) 0.0100 (100.0%) 0.0100 ( 50.0%) 0.0031 ( 6.4%) Dominator Set Construction
				351	0.0000 ( 0.0%) 0.0000 ( 0.0%) 0.0000 ( 0.0%) 0.0013 ( 2.7%) Module Verifier
				352	<b> 0.0000 ( 0.0%) 0.0000 ( 0.0%) 0.0000 ( 0.0%) 0.0033 ( 6.9%) Hello World Pass</b>
				353	0.0100 (100.0%) 0.0100 (100.0%) 0.0200 (100.0%) 0.0479 (100.0%) TOTAL
				354	</pre><p>
				355
				356	As you can see, our implementation above is pretty fast :). The additional
				357	passes listed are automatically inserted by the '<tt>opt</tt>' tool to verify
				358	that the LLVM emitted by your pass is still valid and well formed LLVM, which
				359	hasn't been broken somehow.
				360
				361	Now that you have seen the basics of the mechanics behind passes, we can talk
				362	about some more details of how they work and how to use them.<p>
				363
				364
				365
				366	<!-- *********************************************************************** -->
				367	</ul><table width="100%" bgcolor="#330077" border=0 cellpadding=4 cellspacing=0>
				368	<tr><td align=center><font color="#EEEEFF" size=+2 face="Georgia,Palatino"><b>
				369	<a name="passtype">Pass classes and requirements
				370	</b></font></td></tr></table><ul>
				371	<!-- *********************************************************************** -->
				372
				373	One of the first things that you should do when designing a new pass is to
				374	decide what class you should subclass for your pass. The <a
				375	href="#basiccode">Hello World</a> example uses the <tt><a
				376	href="#FunctionPass">FunctionPass</a></tt> class for its implementation, but we
				377	did not discuss why or when this should occur. Here we talk about the classes
				378	available, from the most general to the most specific.<p>
				379
Chris Lattner	7991070	2002-08-22 19:21:04 +0000	[diff] [blame^]	380	When choosing a superclass for your Pass, you should choose the <b>most
				381	specific</b> class possible, while still being able to meet the requirements
				382	listed. This gives the LLVM Pass Infrastructure information neccesary to
				383	optimize how passes are run, so that the resultant compiler isn't unneccesarily
				384	slow.<p>
Chris Lattner	c6bb824	2002-08-08 20:11:18 +0000	[diff] [blame]	385
				386
				387
				388	<!-- ======================================================================= -->
				389	</ul><table width="100%" bgcolor="#441188" border=0 cellpadding=4 cellspacing=0>
				390	<tr><td> </td><td width="100%">
				391	<font color="#EEEEFF" face="Georgia,Palatino"><b>
				392	<a name="Pass">The <tt>Pass</tt> class
				393	</b></font></td></tr></table><ul>
				394
				395	The "<tt><a href="http://llvm.cs.uiuc.edu/doxygen/classPass.html">Pass</a></tt>"
				396	class is the most general of all superclasses that you can use. Deriving from
				397	<tt>Pass</tt> indicates that your pass uses the entire program as a unit,
				398	refering to function bodies in no predictable order, or adding and removing
				399	functions. Because nothing is known about the behavior of direct <tt>Pass</tt>
				400	subclasses, no optimization can be done for their execution.<p>
				401
				402	To write a correct <tt>Pass</tt> subclass, derive from <tt>Pass</tt> and
				403	overload the <tt>run</tt> method with the following signature:<p>
				404
				405	<!-- _______________________________________________________________________ -->
				406	</ul><h4><a name="run"><hr size=0>The <tt>run</tt> method</h4><ul>
				407
				408
				409	<pre>
				410	<b>virtual bool</b> run(Module &M) = 0;
				411	</pre><p>
				412
				413	The <tt>run</tt> method performs the interesting work of the pass, and should
				414	return true if the module was modified by the transformation, false
				415	otherwise.<p>
				416
				417
				418
				419	<!-- ======================================================================= -->
				420	</ul><table width="100%" bgcolor="#441188" border=0 cellpadding=4 cellspacing=0>
				421	<tr><td> </td><td width="100%">
				422	<font color="#EEEEFF" face="Georgia,Palatino"><b>
				423	<a name="FunctionPass">The <tt>FunctionPass</tt> class
				424	</b></font></td></tr></table><ul>
				425
				426	In contrast to direct <tt>Pass</tt> subclasses, direct <tt><a
				427	href="http://llvm.cs.uiuc.edu/doxygen/classPass.html">FunctionPass</a></tt>
				428	subclasses do have a predictable, local behavior that can be expected by the
				429	system. All <tt>FunctionPass</tt> execute on each function in the program
				430	independant of all of the other functions in the program.
				431	<tt>FunctionPass</tt>'s do not require that they are executed in a particular
				432	order, and <tt>FunctionPass</tt>'s do not modify external functions.<p>
				433
				434	To be explicit, <tt>FunctionPass</tt> subclasses are not allowed to:<p>
				435
				436	<ol>
				437	<li>Modify a Function other than the one currently being processed.
				438	<li>Add or remove Function's from the current Module.
				439	<li>Add or remove global variables from the current Module.
				440	<li>Maintain state across invocations of
				441	<a href="#runOnFunction"><tt>runOnFunction</tt></a> (including global data)
				442	</ol><p>
				443
				444	Implementing a <tt>FunctionPass</tt> is usually straightforward (See the <a
				445	href="#basiccode">Hello World</a> pass for example). <tt>FunctionPass</tt>'s
				446	may overload three virtual methods to do their work. All of these methods
				447	should return true if they modified the program, or false if they didn't.<p>
				448
				449	<!-- _______________________________________________________________________ -->
				450	</ul><h4><a name="doInitialization"><hr size=0>The <tt>doInitialization</tt>
				451	method</h4><ul>
				452
				453	<pre>
				454	<b>virtual bool</b> doInitialization(Module &M);
				455	</pre><p>
				456
				457	The <tt>doIninitialize</tt> method is allowed to do most of the things that
				458	<tt>FunctionPass</tt>'s are not allowed to do. They can add and remove
				459	functions, get pointers to functions, etc. The <tt>doInitialize</tt> method is
				460	designed to do simple initialization type of stuff that does not depend on the
				461	functions being processed. The <tt>doInitialization</tt> function call is not
				462	scheduled to overlap with any other pass executions.<p>
				463
				464	A good example of how this method should be used is the <a
				465	href="http://llvm.cs.uiuc.edu/doxygen/LowerAllocations_8cpp-source.html">LowerAllocations</a>
				466	pass. This pass converts <tt>malloc</tt> and <tt>free</tt> instructions into
				467	platform dependant <tt>malloc()</tt> and <tt>free()</tt> function calls. It
				468	uses the <tt>doInitialization</tt> method to get a reference to the malloc and
				469	free functions that it needs, adding prototypes to the module if neccesary.<p>
				470
				471	<!-- _______________________________________________________________________ -->
				472	</ul><h4><a name="runOnFunction"><hr size=0>The <tt>runOnFunction</tt> method</h4><ul>
				473
				474	<pre>
				475	<b>virtual bool</b> runOnFunction(Function &F) = 0;
				476	</pre><p>
				477
				478	The <tt>runOnFunction</tt> method must be implemented by your subclass to do the
				479	transformation or analysis work of your pass. As usual, a true value should be
				480	returned if the function is modified.<p>
				481
				482	<!-- _______________________________________________________________________ -->
				483	</ul><h4><a name="doFinalization"><hr size=0>The <tt>doFinalization</tt> method</h4><ul>
				484
				485	<pre>
				486	<b>virtual bool</b> doFinalization(Module &M);
				487	</pre</p>
				488
				489	The <tt>doFinalization</tt> method is an infrequently used method that is called
				490	when the pass framework has finished calling <a
				491	href="#runOnFunction"><tt>runOnFunction</tt></a> for every function in the
				492	program being compiled.<p>
				493
				494
				495
				496	<!-- ======================================================================= -->
				497	</ul><table width="100%" bgcolor="#441188" border=0 cellpadding=4 cellspacing=0>
				498	<tr><td> </td><td width="100%">
				499	<font color="#EEEEFF" face="Georgia,Palatino"><b>
				500	<a name="BasicBlockPass">The <tt>BasicBlockPass</tt> class</a>
				501	</b></font></td></tr></table><ul>
				502
				503	<tt>BasicBlockPass</tt>'s are just like <a
				504	href="#FunctionPass"><tt>FunctionPass</tt></a>'s, except that they must limit
				505	their scope of inspection and modification to a single basic block at a time.
				506	As such, they are <b>not</b> allowed to do any of the following:<p>
				507
				508	<ol>
				509	<li>Modify or inspect any basic blocks outside of the current one
				510	<li>Maintain state across invocations of
				511	<a href="#runOnBasicBlock"><tt>runOnBasicBlock</tt></a>
				512	<li>Modify the constrol flow graph (by altering terminator instructions)
				513	<li>Any of the things verboten for
				514	<a href="#FunctionPass"><tt>FunctionPass</tt></a>'s.
				515	</ol><p>
				516
				517	<tt>BasicBlockPass</tt>'s are useful for traditional local and "peephole"
				518	optimizations. They may override the same <a
				519	href="#doInitialization"><tt>doInitialization</tt></a> and <a
				520	href="#doFinalization"><tt>doFinalization</tt></a> methods that <a
				521	href="#FunctionPass"><tt>FunctionPass</tt></a>'s have, but also have a
				522	<tt>runOnBasicBlock</tt> method:<p>
				523
				524	<!-- _______________________________________________________________________ -->
				525	</ul><h4><a name="runOnBasicBlock"><hr size=0>The <tt>runOnBasicBlock</tt> method</h4><ul>
				526
				527	<pre>
				528	<b>virtual bool</b> runOnBasicBlock(BasicBlock &BB) = 0;
				529	</pre><p>
				530
				531	Override this function to do the work of the <tt>BasicBlockPass</tt>. This
				532	function is not allowed to inspect or modify basic blocks other than the
				533	parameter, and are not allowed to modify the CFG. A true value must be returned
				534	if the basic block is modified.<p>
				535
				536
				537	<!-- *********************************************************************** -->
				538	</ul><table width="100%" bgcolor="#330077" border=0 cellpadding=4 cellspacing=0>
				539	<tr><td align=center><font color="#EEEEFF" size=+2 face="Georgia,Palatino"><b>
				540	<a name="registration">Pass registration
				541	</b></font></td></tr></table><ul>
				542	<!-- *********************************************************************** -->
				543
				544	In the <a href="#basiccode">Hello World</a> example pass we illustrated how pass
				545	registration works, and discussed some of the reasons that it is used and what
				546	it does. Here we discuss how and why passes are registered.<p>
				547
				548	Passes can be registered in several different ways. Depending on the general
				549	classification of the pass, you should use one of the following templates to
				550	register the pass:<p>
				551
				552	<ul>
				553	<li><b><tt>RegisterOpt</tt></b> - This template should be used when you are
				554	registering a pass that logically should be available for use in the
				555	'<tt>opt</tt>' utility.<p>
				556
				557	<li><b><tt>RegisterAnalysis</tt></b> - This template should be used when you are
				558	registering a pass that logically should be available for use in the
				559	'<tt>analysis</tt>' utility.<p>
				560
				561	<li><b><tt>RegisterLLC</tt></b> - This template should be used when you are
				562	registering a pass that logically should be available for use in the
				563	'<tt>llc</tt>' utility.<p>
				564
				565	<li><b><tt>RegisterPass</tt></b> - This is the generic form of the
				566	<tt>Register*</tt> templates that should be used if you want your pass listed by
				567	multiple or no utilities. This template takes an extra third argument that
				568	specifies which tools it should be listed in. See the <a
				569	href="http://llvm.cs.uiuc.edu/doxygen/PassSupport_8h-source.html">PassSupport.h</a>
				570	file for more information.<p>
				571	</ul><p>
				572
				573	Regardless of how you register your pass, you must specify at least two
				574	parameters. The first parameter is the name of the pass that is to be used on
				575	the command line to specify that the pass should be added to a program (for
				576	example <tt>opt</tt> or <tt>analyze</tt>). The second argument is the name of
				577	the pass, which is to be used for the <tt>--help</tt> output of programs, as
				578	well as for debug output generated by the <tt>--debug-pass</tt> option.<p>
				579
				580	If you pass is constructed by its default constructor, you only ever have to
				581	pass these two arguments. If, on the other hand, you require other information
				582	(like target specific information), you must pass an additional argument. This
				583	argument is a pointer to a function used to create the pass. For an example of
				584	how this works, look at the <a
				585	href="http://llvm.cs.uiuc.edu/doxygen/LowerAllocations_8cpp-source.html">LowerAllocations.cpp</a>
				586	file.<p>
				587
				588	If a pass is registered to be used by the <tt>analyze</tt> utility, you should
				589	implement the virtual <tt>print</tt> method:<p>
				590
				591	<!-- _______________________________________________________________________ -->
				592	</ul><h4><a name="print"><hr size=0>The <tt>print</tt> method</h4><ul>
				593
				594	<pre>
				595	<b>virtual void</b> print(std::ostream &O, <b>const</b> Module *M) <b>const</b>;
				596	</pre><p>
				597
				598	The <tt>print</tt> method must be implemented by "analyses" in order to print a
				599	human readable version of the analysis results. This is useful for debugging an
				600	analysis itself, as well as for other people to figure out how an analysis
				601	works. The <tt>analyze</tt> tool uses this method to generate its output.<p>
				602
				603	The <tt>ostream</tt> parameter specifies the stream to write the results on, and
				604	the <tt>Module</tt> parameter gives a pointer to the top level module of the
				605	program that has been analyzed. Note however that this pointer may be null in
				606	certain circumstances (such as calling the <tt>Pass::dump()</tt> from a
				607	debugger), so it should only be used to enhance debug output, it should not be
				608	depended on.<p>
				609
				610
				611	<!-- *********************************************************************** -->
				612	</ul><table width="100%" bgcolor="#330077" border=0 cellpadding=4 cellspacing=0>
				613	<tr><td align=center><font color="#EEEEFF" size=+2 face="Georgia,Palatino"><b>
				614	<a name="interaction">Specifying interactions between passes
				615	</b></font></td></tr></table><ul>
				616	<!-- *********************************************************************** -->
				617
				618	One of the main responsibilities of the <tt>PassManager</tt> is the make sure
				619	that passes interact with each other correctly. Because <tt>PassManager</tt>
				620	tries to <a href="#passmanager">optimize the execution of passes</a> it must
				621	know how the passes interact with each other and what dependencies exist between
				622	the various passes. To track this, each pass can declare the set of passes that
				623	are required to be executed before the current pass, and the passes which are
				624	invalidated by the current pass.<p>
				625
				626	Typically this functionality is used to require that analysis results are
				627	computed before your pass is run. Running arbitrary transformation passes can
				628	invalidate the computed analysis results, which is what the invalidation set
				629	specifies. If a pass does not implement the <tt><a
				630	href="#getAnalysisUsage">getAnalysisUsage</a></tt> method, it defaults to not
				631	having any prerequisite passes, and invalidating <b>all</b> other passes.<p>
				632
				633
				634	<!-- _______________________________________________________________________ -->
				635	</ul><h4><a name="getAnalysisUsage"><hr size=0>The <tt>getAnalysisUsage</tt> method</h4><ul>
				636
				637	<pre>
				638	<b>virtual void</b> getAnalysisUsage(AnalysisUsage &Info) <b>const</b>;
				639	</pre><p>
				640
				641	By implementing the <tt>getAnalysisUsage</tt> method, the required and
				642	invalidated sets may be specified for your transformation. The implementation
				643	should fill in the <tt><a
				644	href="http://llvm.cs.uiuc.edu/doxygen/classAnalysisUsage.html">AnalysisUsage</a></tt>
				645	object with information about which passes are required and not invalidated. To do this, the following set methods are provided by the <tt><a
				646	href="http://llvm.cs.uiuc.edu/doxygen/classAnalysisUsage.html">AnalysisUsage</a></tt> class:<p>
				647
				648	<pre>
				649	<i>// addRequires - Add the specified pass to the required set for your pass.</i>
				650	<b>template</b><<b>class</b> PassClass>
				651	AnalysisUsage &AnalysisUsage::addRequired();
				652
				653	<i>// addPreserved - Add the specified pass to the set of analyses preserved by
				654	// this pass</i>
				655	<b>template</b><<b>class</b> PassClass>
				656	AnalysisUsage &AnalysisUsage::addPreserved();
				657
				658	<i>// setPreservesAll - Call this if the pass does not modify its input at all</i>
				659	<b>void</b> AnalysisUsage::setPreservesAll();
				660
				661	<i>// preservesCFG - This function should be called by the pass, iff they do not:
				662	//
				663	// 1. Add or remove basic blocks from the function
				664	// 2. Modify terminator instructions in any way.
				665	//
				666	// This is automatically implied for <a href="#BasicBlockPass">BasicBlockPass</a>'s
				667	//</i>
				668	<b>void</b> AnalysisUsage::preservesCFG();
				669	</pre><p>
				670
				671	Some examples of how to use these methods are:<p>
				672
				673	<pre>
				674	<i>// This is an example implementation from an analysis, which does not modify
				675	// the program at all, yet has a prerequisite.</i>
				676	<b>void</b> <a href="http://llvm.cs.uiuc.edu/doxygen/structPostDominanceFrontier.html">PostDominanceFrontier</a>::getAnalysisUsage(AnalysisUsage &AU) <b>const</b> {
				677	AU.setPreservesAll();
				678	AU.addRequired<<a href="http://llvm.cs.uiuc.edu/doxygen/structPostDominatorTree.html">PostDominatorTree</a>>();
				679	}
				680	</pre><p>
				681
				682	and:<p>
				683
				684	<pre>
				685	<i>// This example modifies the program, but does not modify the CFG</i>
				686	<b>void</b> <a href="http://llvm.cs.uiuc.edu/doxygen/structLICM.html">LICM</a>::getAnalysisUsage(AnalysisUsage &AU) <b>const</b> {
				687	AU.preservesCFG();
				688	AU.addRequired<<a href="http://llvm.cs.uiuc.edu/doxygen/classLoopInfo.html">LoopInfo</a>>();
				689	}
				690	</pre><p>
				691
				692	<!-- _______________________________________________________________________ -->
				693	</ul><h4><a name="getAnalysis"><hr size=0>The <tt>getAnalysis<></tt> method</h4><ul>
				694
				695	The <tt>Pass::getAnalysis<></tt> method is inherited by your class,
				696	providing you with access to the passes that you declared that you required with
				697	the <a href="#getAnalysisUsage"><tt>getAnalysisUsage</tt></a> method. It takes
				698	a single template argument that specifies which pass class you want, and returns
				699	a reference to that pass.<p>
				700
				701	<pre>
				702	<b>template</b><<b>typename</b> PassClass>
				703	AnalysisType &getAnalysis();
				704	</pre><p>
				705
				706	This method call returns a reference to the pass desired. You may get a runtime
				707	assertion failure if you attempt to get an analysis that you did not declare as
				708	required in your <a href="#getAnalysisUsage"><tt>getAnalysisUsage</tt></a>
				709	implementation. This method can be called by your <tt>run*</tt> method
				710	implementation, or by any other local method invoked by your <tt>run*</tt>
				711	method.<p>
				712
Chris Lattner	7991070	2002-08-22 19:21:04 +0000	[diff] [blame^]	713	<!-- *********************************************************************** -->
				714	</ul><table width="100%" bgcolor="#330077" border=0 cellpadding=4 cellspacing=0>
				715	<tr><td align=center><font color="#EEEEFF" size=+2 face="Georgia,Palatino"><b>
				716	<a name="analysisgroup">Implementing Analysis Groups
				717	</b></font></td></tr></table><ul>
				718	<!-- *********************************************************************** -->
				719
				720	Now that we understand the basics of how passes are defined, how the are used,
				721	and how they are required from other passes, it's time to get a little bit
				722	fancier. All of the pass relationships that we have seen so far are very
				723	simple: one pass depends on one other specific pass to be run before it can run.
				724	For many applications, this is great, for others, more flexibility is
				725	required.<p>
				726
				727	In particular, some analyses are defined such that there is a single simple
				728	interface to the analysis results, but multiple ways of calculating them.
				729	Consider alias analysis for example. The most trivial alias analysis returns
				730	"may alias" for any alias query. The most sophisticated analysis a
				731	flow-sensitive, context-sensitive interprocedural analysis that can take a
				732	significant amount of time to execute (and obviously, there is a lot of room
				733	between these two extremes for other implementations). To cleanly support
				734	situations like this, the LLVM Pass Infrastructure supports the notion of
				735	Analysis Groups.<p>
				736
				737	<!-- _______________________________________________________________________ -->
				738	</ul><h4><a name="agconcepts"><hr size=0>Analysis Group Concepts</h4><ul>
				739
				740	An Analysis Group is a single simple interface that may be implemented by
				741	multiple different passes. Analysis Groups can be given human readable names
				742	just like passes, but unlike passes, they need not derive from the <tt>Pass</tt>
				743	class. An analysis group may have one or more implementations, one of which is
				744	the "default" implementation.<p>
				745
				746	Analysis groups are used by client passes just like other passes are: the
				747	<tt>AnalysisUsage::addRequired()</tt> and <tt>Pass::getAnalysis()</tt> methods.
				748	In order to resolve this requirement, the <a href="#passmanager">PassManager</a>
				749	scans the available passes to see if any implementations of the analysis group
				750	are available. If none is available, the default implementation is created for
				751	the pass to use. All standard rules for <A href="#interaction">interaction
				752	between passes</a> still apply.<p>
				753
				754	Although <a href="#registration">Pass Registration</a> is optional for normal
				755	passes, all analysis group implementations must be registered, and must use the
				756	<A href="#registerag"><tt>RegisterAnalysisGroup</tt></a> template to join the
				757	implementation pool. Also, a default implementation of the interface
				758	<b>must</b> be registered with <A
				759	href="#registerag"><tt>RegisterAnalysisGroup</tt></a>.<p>
				760
				761	As a concrete example of an Analysis Group in action, consider the <a
				762	href="http://llvm.cs.uiuc.edu/doxygen/structAliasAnalysis.html">AliasAnalysis</a>
				763	analysis group. The default implementation of the alias analysis interface (the
				764	<tt><a
				765	href="http://llvm.cs.uiuc.edu/doxygen/structBasicAliasAnalysis.html">basicaa</a></tt>
				766	pass) just does a few simple checks that don't require significant analysis to
				767	compute (such as: two different globals can never alias each other, etc).
				768	Passes that use the <tt><a
				769	href="http://llvm.cs.uiuc.edu/doxygen/structAliasAnalysis.html">AliasAnalysis</a></tt>
				770	interface (for example the <tt><a
				771	href="http://llvm.cs.uiuc.edu/doxygen/classGCSE.html">gcse</a></tt> pass), do not care which implementation
				772	of alias analysis is actually provided, they just use the designated
				773	interface.<p>
				774
				775	From the user's perspective, commands work just like normal. Issuing the
				776	command '<tt>opt -gcse ...</tt>' will cause the <tt>basicaa</tt> class to be
				777	instantiated and added to the pass sequence. Issuing the command '<tt>opt
				778	-somefancyaa -gcse ...</tt>' will cause the <tt>gcse</tt> pass to use the
				779	<tt>somefancyaa</tt> alias analysis (which doesn't actually exist, it's just a
				780	hypothetical example) instead.<p>
				781
				782
				783	<!-- _______________________________________________________________________ -->
				784	</ul><h4><a name="registerag"><hr size=0>Using <tt>RegisterAnalysisGroup</tt></h4><ul>
				785
				786	The <tt>RegisterAnalysisGroup</tt> template is used to register the analysis
				787	group itself as well as add pass implementations to the analysis group. First,
				788	an analysis should be registered, with a human readable name provided for it.
				789	Unlike registration of passes, there is no command line argument to be specified
				790	for the Analysis Group Interface itself, because it is "abstract":<p>
				791
				792	<pre>
				793	<b>static</b> RegisterAnalysisGroup<<a href="http://llvm.cs.uiuc.edu/doxygen/structAliasAnalysis.html">AliasAnalysis</a>> A("<i>Alias Analysis</i>");
				794	</pre><p>
				795
				796	Once the analysis is registered, passes can declare that they are valid
				797	implementations of the interface by using the following code:<p>
				798
				799	<pre>
				800	<b>namespace</b> {
				801	//<i> Analysis Group implementations <b>must</b> be registered normally...</i>
				802	RegisterOpt<FancyAA>
				803	B("<i>somefancyaa</i>", "<i>A more complex alias analysis implementation</i>");
				804
				805	//<i> Declare that we implement the AliasAnalysis interface</i>
				806	RegisterAnalysisGroup<<a href="http://llvm.cs.uiuc.edu/doxygen/structAliasAnalysis.html">AliasAnalysis</a>, FancyAA> C;
				807	}
				808	</pre><p>
				809
				810	This just shows a class <tt>FancyAA</tt> that is registered normally, then uses
				811	the <tt>RegisterAnalysisGroup</tt> template to "join" the <tt><a
				812	href="http://llvm.cs.uiuc.edu/doxygen/structAliasAnalysis.html">AliasAnalysis</a></tt>
				813	analysis group. Every implementation of an analysis group should join using
				814	this template. A single pass may join multiple different analysis groups with
				815	no problem.<p>
				816
				817	<pre>
				818	<b>namespace</b> {
				819	//<i> Analysis Group implementations <b>must</b> be registered normally...</i>
				820	RegisterOpt<<a href="http://llvm.cs.uiuc.edu/doxygen/structBasicAliasAnalysis.html">BasicAliasAnalysis</a>>
				821	D("<i>basicaa</i>", "<i>Basic Alias Analysis (default AA impl)</i>");
				822
				823	//<i> Declare that we implement the AliasAnalysis interface</i>
				824	RegisterAnalysisGroup<<a href="http://llvm.cs.uiuc.edu/doxygen/structAliasAnalysis.html">AliasAnalysis</a>, <a href="http://llvm.cs.uiuc.edu/doxygen/structBasicAliasAnalysis.html">BasicAliasAnalysis</a>, <b>true</b>> E;
				825	}
				826	</pre><p>
				827
				828	Here we show how the default implementation is specified (using the extra
				829	argument to the <tt>RegisterAnalysisGroup</tt> template). There must be exactly
				830	one default implementation available at all times for an Analysis Group to be
				831	used. Here we declare that the <tt><a
				832	href="http://llvm.cs.uiuc.edu/doxygen/structBasicAliasAnalysis.html">BasicAliasAnalysis</a></tt>
				833	pass is the default implementation for the interface.<p>
Chris Lattner	c6bb824	2002-08-08 20:11:18 +0000	[diff] [blame]	834
				835
				836	<!-- *********************************************************************** -->
				837	</ul><table width="100%" bgcolor="#330077" border=0 cellpadding=4 cellspacing=0>
				838	<tr><td align=center><font color="#EEEEFF" size=+2 face="Georgia,Palatino"><b>
				839	<a name="passmanager">What PassManager does
				840	</b></font></td></tr></table><ul>
				841	<!-- *********************************************************************** -->
				842
				843	The <a
				844	href="http://llvm.cs.uiuc.edu/doxygen/PassManager_8h-source.html"><tt>PassManager</tt></a>
				845	<a href="http://llvm.cs.uiuc.edu/doxygen/classPassManager.html">class</a> takes
				846	a list of passes, ensures their <a href="#interaction">prerequisites</a> are set
				847	up correctly, and then schedules passes to run efficiently. All of the LLVM
				848	tools that run passes use the <tt>PassManager</tt> for execution of these
				849	passes.<p>
				850
				851	The <tt>PassManager</tt> does two main things to try to reduce the execution
				852	time of a series of passes:<p>
				853
				854	<ol>
				855	<li><b>Share analysis results</b> - The PassManager attempts to avoid
				856	recomputing analysis results as much as possible. This means keeping track of
				857	which analyses are available already, which analyses get invalidated, and which
				858	analyses are needed to be run for a pass. An important part of work is that the
				859	<tt>PassManager</tt> tracks the exact lifetime of all analysis results, allowing
				860	it to <a href="#releaseMemory">free memory</a> allocated to holding analysis
				861	results as soon as they are no longer needed.<p>
				862
				863	<li><b>Pipeline the execution of passes on the program</b> - The
				864	<tt>PassManager</tt> attempts to get better cache and memory usage behavior out
				865	of a series of passes by pipelining the passes together. This means that, given
				866	a series of consequtive <a href="#FunctionPass"><tt>FunctionPass</tt></a>'s, it
				867	will execute all of the <a href="#FunctionPass"><tt>FunctionPass</tt></a>'s on
				868	the first function, then all of the <a
				869	href="#FunctionPass"><tt>FunctionPass</tt></a>'s on the second function,
				870	etc... until the entire program has been run through the passes.<p>
				871
				872	This improves the cache behavior of the compiler, because it is only touching
				873	the LLVM program representation for a single function at a time, instead of
				874	traversing the entire program. It reduces the memory consumption of compiler,
				875	because, for example, only one <a
				876	href="http://llvm.cs.uiuc.edu/doxygen/structDominatorSet.html"><tt>DominatorSet</tt></a>
				877	needs to be calculated at a time. This also makes it possible some <a
				878	href="#SMP">interesting enhancements</a> in the future.<p>
				879
				880	</ol><p>
				881
				882	The effectiveness of the <tt>PassManager</tt> is influenced directly by how much
				883	information it has about the behaviors of the passes it is scheduling. For
				884	example, the "preserved" set is intentionally conservative in the face of an
				885	unimplemented <a href="#getAnalysisUsage"><tt>getAnalysisUsage</tt></a> method.
				886	Not implementing when it should be implemented will have the effect of not
				887	allowing any analysis results to live across the execution of your pass.<p>
				888
				889	The <tt>PassManager</tt> class exposes a <tt>--debug-pass</tt> command line
				890	options that is useful for debugging pass execution, seeing how things work, and
				891	diagnosing when you should be preserving more analyses than you currently are
				892	(To get information about all of the variants of the <tt>--debug-pass</tt>
				893	option, just type '<tt>opt --help-hidden</tt>').<p>
				894
				895	By using the <tt>--debug-pass=Structure</tt> option, for example, we can see how
				896	our <a href="#basiccode">Hello World</a> pass interacts with other passes. Lets
				897	try it out with the <tt>gcse</tt> and <tt>licm</tt> passes:<p>
				898
				899	<pre>
				900	$ opt -load ../../../lib/Debug/libhello.so -gcse -licm --debug-pass=Structure < hello.bc > /dev/null
				901	Module Pass Manager
				902	Function Pass Manager
				903	Dominator Set Construction
				904	Immediate Dominators Construction
				905	Global Common Subexpression Elimination
				906	-- Immediate Dominators Construction
				907	-- Global Common Subexpression Elimination
				908	Natural Loop Construction
				909	Loop Invariant Code Motion
				910	-- Natural Loop Construction
				911	-- Loop Invariant Code Motion
				912	Module Verifier
				913	-- Dominator Set Construction
				914	-- Module Verifier
				915	Bytecode Writer
				916	--Bytecode Writer
				917	</pre><p>
				918
				919	This output shows us when passes are constructed and when the analysis results
				920	are known to be dead (prefixed with '<tt>--</tt>'). Here we see that GCSE uses
				921	dominator and immediate dominator information to do its job. The LICM pass uses
				922	natural loop information, which uses dominator sets, but not immediate
				923	dominators. Because immediate dominators are no longer useful after the GCSE
				924	pass, it is immediately destroyed. The dominator sets are then reused to
				925	compute natural loop information, which is then used by the LICM pass.<p>
				926
				927	After the LICM pass, the module verifier runs (which is automatically added by
				928	the '<tt>opt</tt>' tool), which uses the dominator set to check that the
				929	resultant LLVM code is well formed. After it finishes, the dominator set
				930	information is destroyed, after being computed once, and shared by three
				931	passes.<p>
				932
				933	Lets see how this changes when we run the <a href="#basiccode">Hello World</a>
				934	pass in between the two passes:<p>
				935
				936	<pre>
				937	$ opt -load ../../../lib/Debug/libhello.so -gcse -hello -licm --debug-pass=Structure < hello.bc > /dev/null
				938	Module Pass Manager
				939	Function Pass Manager
				940	Dominator Set Construction
				941	Immediate Dominators Construction
				942	Global Common Subexpression Elimination
				943	<b>-- Dominator Set Construction</b>
				944	-- Immediate Dominators Construction
				945	-- Global Common Subexpression Elimination
				946	<b> Hello World Pass
				947	-- Hello World Pass
				948	Dominator Set Construction</b>
				949	Natural Loop Construction
				950	Loop Invariant Code Motion
				951	-- Natural Loop Construction
				952	-- Loop Invariant Code Motion
				953	Module Verifier
				954	-- Dominator Set Construction
				955	-- Module Verifier
				956	Bytecode Writer
				957	--Bytecode Writer
				958	Hello: __main
				959	Hello: puts
				960	Hello: main
				961	</pre><p>
				962
				963	Here we see that the <a href="#basiccode">Hello World</a> pass has killed the
				964	Dominator Set pass, even though it doesn't modify the code at all! To fix this,
				965	we need to add the following <a
				966	href="#getAnalysisUsage"><tt>getAnalysisUsage</tt></a> method to our pass:<p>
				967
				968	<pre>
				969	<i>// We don't modify the program, so we preserve all analyses</i>
				970	<b>virtual void</b> getAnalysisUsage(AnalysisUsage &AU) <b>const</b> {
				971	AU.setPreservesAll();
				972	}
				973	</pre><p>
				974
				975	Now when we run our pass, we get this output:<p>
				976
				977	<pre>
				978	$ opt -load ../../../lib/Debug/libhello.so -gcse -hello -licm --debug-pass=Structure < hello.bc > /dev/null
				979	Pass Arguments: -gcse -hello -licm
				980	Module Pass Manager
				981	Function Pass Manager
				982	Dominator Set Construction
				983	Immediate Dominators Construction
				984	Global Common Subexpression Elimination
				985	-- Immediate Dominators Construction
				986	-- Global Common Subexpression Elimination
				987	Hello World Pass
				988	-- Hello World Pass
				989	Natural Loop Construction
				990	Loop Invariant Code Motion
				991	-- Loop Invariant Code Motion
				992	-- Natural Loop Construction
				993	Module Verifier
				994	-- Dominator Set Construction
				995	-- Module Verifier
				996	Bytecode Writer
				997	--Bytecode Writer
				998	Hello: __main
				999	Hello: puts
				1000	Hello: main
				1001	</pre><p>
				1002
				1003	Which shows that we don't accidentally invalidate dominator information
				1004	anymore, and therefore do not have to compute it twice.<p>
				1005
				1006
				1007	<!-- _______________________________________________________________________ -->
				1008	</ul><h4><a name="releaseMemory"><hr size=0>The <tt>releaseMemory</tt> method</h4><ul>
				1009
				1010	<pre>
				1011	<b>virtual void</b> releaseMemory();
				1012	</pre><p>
				1013
				1014	The <tt>PassManager</tt> automatically determines when to compute analysis
				1015	results, and how long to keep them around for. Because the lifetime of the pass
				1016	object itself is effectively the entire duration of the compilation process, we
				1017	need some way to free analysis results when they are no longer useful. The
				1018	<tt>releaseMemory</tt> virtual method is the way to do this.<p>
				1019
				1020	If you are writing an analysis or any other pass that retains a significant
				1021	amount of state (for use by another pass which "requires" your pass and uses the
				1022	<a href="#getAnalysis">getAnalysis</a> method) you should implement
				1023	<tt>releaseMEmory</tt> to, well, release the memory allocated to maintain this
				1024	internal state. This method is called after the <tt>run*</tt> method for the
				1025	class, before the next call of <tt>run*</tt> in your pass.<p>
				1026
				1027
				1028	<!-- *********************************************************************** -->
				1029	</ul><table width="100%" bgcolor="#330077" border=0 cellpadding=4 cellspacing=0>
				1030	<tr><td align=center><font color="#EEEEFF" size=+2 face="Georgia,Palatino"><b>
				1031	<a name="future">Future extensions planned
				1032	</b></font></td></tr></table><ul>
				1033	<!-- *********************************************************************** -->
				1034
				1035	Although the LLVM Pass Infrastructure is very capable as it stands, and does
				1036	some nifty stuff, there are things we'd like to add in the future. Here is
				1037	where we are going:<p>
				1038
				1039	<!-- _______________________________________________________________________ -->
				1040	</ul><h4><a name="SMP"><hr size=0>Multithreaded LLVM</h4><ul>
				1041
				1042	Multiple CPU machines are becoming more command and compilation can never be
				1043	fast enough: obviously we should allow for a multithreaded compiler. Because of
				1044	the semantics defined for passes above (specifically they cannot maintain state
				1045	across invocations of their <tt>run*</tt> methods), a nice clean way to
				1046	implement a multithreaded compiler would be for the <tt>PassManager</tt> class
				1047	to create multiple instances of each pass object, and allow the seperate
				1048	instances to be hacking on different parts of the program at the same time.<p>
				1049
				1050	This implementation would prevent each of the passes from having to implement
				1051	multithreaded constructs, requiring only the LLVM core to have locking in a few
				1052	places (for global resources). Although this is a simple extension, we simply
				1053	haven't had time (or multiprocessor machines, thus a reason) to implement this.
				1054	Despite that, we have kept the LLVM passes SMP ready, and you should too.<p>
				1055
				1056
				1057	<!-- _______________________________________________________________________ -->
				1058	</ul><h4><a name="ModuleSource"><hr size=0>A new <tt>ModuleSource</tt> interface</h4><ul>
				1059
				1060	Currently, the <tt>PassManager</tt>'s <tt>run</tt> method takes a <tt><a
				1061	href="http://llvm.cs.uiuc.edu/doxygen/classModule.html">Module</a></tt> as
				1062	input, and runs all of the passes on this module. The problem with this
				1063	approach is that none of the <tt>PassManager</tt> features can be used for
				1064	timing and debugging the actual <b>loading</b> of the module from disk or
				1065	standard input.<p>
				1066
				1067	To solve this problem, eventually the <tt>PassManger</tt> class will accept a
				1068	<tt>ModuleSource</tt> object instead of a Module itself. When complete, this
				1069	will also allow for streaming of functions out of the bytecode representation,
				1070	allowing us to avoid holding the entire program in memory at once if we only are
				1071	dealing with <a href="#FunctionPass">FunctionPass</a>'s.<p>
				1072
				1073	As part of a different issue, eventually the bytecode loader will be extended to
				1074	allow on-demand loading of functions from the bytecode representation, in order
				1075	to better support the runtime reoptimizer. The bytecode format is already
				1076	capable of this, the loader just needs to be reworked a bit.<p>
				1077
				1078
				1079	<!-- _______________________________________________________________________ -->
				1080	</ul><h4><a name="PassFunctionPass"><hr size=0><tt>Pass</tt>'s requiring <tt>FunctionPass</tt>'s</h4><ul>
				1081
				1082	Currently it is illegal for a <a href="#Pass"><tt>Pass</tt></a> to require a <a
				1083	href="#FunctionPass"><tt>FunctionPass</tt></a>. This is because there is only
				1084	one instance of the <a href="#FunctionPass"><tt>FunctionPass</tt></a> object
				1085	ever created, thus nowhere to store information for all of the functions in the
				1086	program at the same time. Although this has come up a couple of times before,
				1087	this has always been worked around by factoring one big complicated pass into a
				1088	global and an interprocedural part, both of which are distinct. In the future,
				1089	it would be nice to have this though.<p>
				1090
				1091	Note that it is no problem for a <a
				1092	href="#FunctionPass"><tt>FunctionPass</tt></a> to require the results of a <a
				1093	href="#Pass"><tt>Pass</tt></a>, only the other way around.<p>
				1094
				1095
				1096	<!-- *********************************************************************** -->
				1097	</ul>
				1098	<!-- *********************************************************************** -->
				1099
				1100	<hr><font size-1>
				1101	<address><a href="mailto:sabre@nondot.org">Christopher Lattner</a></address>
				1102	<!-- Created: Tue Aug 6 15:00:33 CDT 2002 -->
				1103	<!-- hhmts start -->
Chris Lattner	7991070	2002-08-22 19:21:04 +0000	[diff] [blame^]	1104	Last modified: Thu Aug 22 14:19:43 CDT 2002
Chris Lattner	c6bb824	2002-08-08 20:11:18 +0000	[diff] [blame]	1105	<!-- hhmts end -->
				1106	</font></body></html>