Blame - docs/InternalsManual.html - platform/external/clang

blob: 6434a910e132bfcbf6af83f518fb8c202281ce75 [file] [log] [blame]

Ted Kremenek	17a295d	2008-06-11 06:19:49 +0000	[diff] [blame]	1	<html>
				2	<head>
Chris Lattner	86920d3	2007-07-31 05:42:17 +0000	[diff] [blame]	3	<title>"clang" CFE Internals Manual</title>
Ted Kremenek	17a295d	2008-06-11 06:19:49 +0000	[diff] [blame]	4	<link type="text/css" rel="stylesheet" href="../menu.css" />
				5	<link type="text/css" rel="stylesheet" href="../content.css" />
Sebastian Redl	6816856	2008-11-22 22:16:45 +0000	[diff] [blame^]	6	<style type="text/css">
				7	td {
				8	vertical-align: top;
				9	}
				10	</style>
Ted Kremenek	17a295d	2008-06-11 06:19:49 +0000	[diff] [blame]	11	</head>
				12	<body>
				13
				14	<!--#include virtual="../menu.html.incl"-->
				15
				16	<div id="content">
Chris Lattner	86920d3	2007-07-31 05:42:17 +0000	[diff] [blame]	17
				18	<h1>"clang" CFE Internals Manual</h1>
				19
				20	<ul>
				21	<li><a href="#intro">Introduction</a></li>
				22	<li><a href="#libsystem">LLVM System and Support Libraries</a></li>
				23	<li><a href="#libbasic">The clang 'Basic' Library</a>
				24	<ul>
Chris Lattner	62fd278	2008-11-22 21:41:31 +0000	[diff] [blame]	25	<li><a href="#Diagnostics">The Diagnostics Subsystem</a></li>
Chris Lattner	86920d3	2007-07-31 05:42:17 +0000	[diff] [blame]	26	<li><a href="#SourceLocation">The SourceLocation and SourceManager
				27	classes</a></li>
				28	</ul>
				29	</li>
				30	<li><a href="#liblex">The Lexer and Preprocessor Library</a>
				31	<ul>
				32	<li><a href="#Token">The Token class</a></li>
				33	<li><a href="#Lexer">The Lexer class</a></li>
Chris Lattner	7928125	2008-03-09 02:27:26 +0000	[diff] [blame]	34	<li><a href="#TokenLexer">The TokenLexer class</a></li>
Chris Lattner	86920d3	2007-07-31 05:42:17 +0000	[diff] [blame]	35	<li><a href="#MultipleIncludeOpt">The MultipleIncludeOpt class</a></li>
				36	</ul>
				37	</li>
				38	<li><a href="#libparse">The Parser Library</a>
				39	<ul>
				40	</ul>
				41	</li>
				42	<li><a href="#libast">The AST Library</a>
				43	<ul>
				44	<li><a href="#Type">The Type class and its subclasses</a></li>
				45	<li><a href="#QualType">The QualType class</a></li>
Douglas Gregor	2e1cd42	2008-11-17 14:58:09 +0000	[diff] [blame]	46	<li><a href="#DeclarationName">Declaration names</a></li>
Ted Kremenek	8bc0571	2007-10-10 23:01:43 +0000	[diff] [blame]	47	<li><a href="#CFG">The CFG class</a></li>
Chris Lattner	7bad199	2008-11-16 21:48:07 +0000	[diff] [blame]	48	<li><a href="#Constants">Constant Folding in the Clang AST</a></li>
Chris Lattner	86920d3	2007-07-31 05:42:17 +0000	[diff] [blame]	49	</ul>
				50	</li>
				51	</ul>
				52
				53
				54	<!-- ======================================================================= -->
				55	<h2 id="intro">Introduction</h2>
				56	<!-- ======================================================================= -->
				57
				58	<p>This document describes some of the more important APIs and internal design
				59	decisions made in the clang C front-end. The purpose of this document is to
				60	both capture some of this high level information and also describe some of the
				61	design decisions behind it. This is meant for people interested in hacking on
				62	clang, not for end-users. The description below is categorized by
				63	libraries, and does not describe any of the clients of the libraries.</p>
				64
				65	<!-- ======================================================================= -->
				66	<h2 id="libsystem">LLVM System and Support Libraries</h2>
				67	<!-- ======================================================================= -->
				68
				69	<p>The LLVM libsystem library provides the basic clang system abstraction layer,
				70	which is used for file system access. The LLVM libsupport library provides many
				71	underlying libraries and <a
				72	href="http://llvm.org/docs/ProgrammersManual.html">data-structures</a>,
				73	including command line option
				74	processing and various containers.</p>
				75
				76	<!-- ======================================================================= -->
				77	<h2 id="libbasic">The clang 'Basic' Library</h2>
				78	<!-- ======================================================================= -->
				79
				80	<p>This library certainly needs a better name. The 'basic' library contains a
				81	number of low-level utilities for tracking and manipulating source buffers,
				82	locations within the source buffers, diagnostics, tokens, target abstraction,
				83	and information about the subset of the language being compiled for.</p>
				84
				85	<p>Part of this infrastructure is specific to C (such as the TargetInfo class),
				86	other parts could be reused for other non-C-based languages (SourceLocation,
				87	SourceManager, Diagnostics, FileManager). When and if there is future demand
				88	we can figure out if it makes sense to introduce a new library, move the general
				89	classes somewhere else, or introduce some other solution.</p>
				90
				91	<p>We describe the roles of these classes in order of their dependencies.</p>
				92
Chris Lattner	62fd278	2008-11-22 21:41:31 +0000	[diff] [blame]	93
				94	<!-- ======================================================================= -->
				95	<h3 id="Diagnostics">The Diagnostics Subsystem</h3>
				96	<!-- ======================================================================= -->
				97
				98	<p>The Clang Diagnostics subsystem is an important part of how the compiler
				99	communicates with the human. Diagnostics are the warnings and errors produced
				100	when the code is incorrect or dubious. In Clang, each diagnostic produced has
				101	(at the minimum) a unique ID, a <a href="#SourceLocation">SourceLocation</a> to
				102	"put the caret", an English translation associated with it, and a severity (e.g.
				103	<tt>WARNING</tt> or <tt>ERROR</tt>). They can also optionally include a number
				104	of arguments to the dianostic (which fill in "%0"'s in the string) as well as a
				105	number of source ranges that related to the diagnostic.</p>
				106
				107	<p>In this section, we'll be giving examples produced by the clang command line
				108	driver, but diagnostics can be <a href="#DiagnosticClient">rendered in many
				109	different ways</a> depending on how the DiagnosticClient interface is
				110	implemented. A representative example of a diagonstic is:</p>
				111
				112	<pre>
				113	t.c:38:15: error: invalid operands to binary expression ('int *' and '_Complex float')
				114	<font color="darkgreen">P = (P-42) + Gamma*4;</font>
				115	<font color="blue">~~~~~~ ^ ~~~~~~~</font>
				116	</pre>
				117
				118	<p>In this example, you can see the English translation, the severity (error),
				119	you can see the source location (the caret ("^") and file/line/column info),
				120	the source ranges "~~~~", arguments to the diagnostic ("int*" and "_Complex
				121	float"). You'll have to believe me that there is a unique ID backing the
				122	diagnostic :).</p>
				123
				124	<p>Getting all of this to happen has several steps and involves many moving
				125	pieces, this section describes them and talks about best practices when adding
				126	a new diagnostic.</p>
				127
				128	<!-- ============================ -->
				129	<h4>The DiagnosticKinds.def file</h4>
				130	<!-- ============================ -->
				131
				132	<p>Diagnostics are created by adding an entry to the <tt><a
				133	href="http://llvm.org/svn/llvm-project/cfe/trunk/include/clang/Basic/DiagnosticKinds.def"
				134	>DiagnosticKinds.def</a></tt> file. This file encodes the unique ID of the
				135	diagnostic (as an enum, the first argument), the severity of the diagnostic
				136	(second argument) and the English translation + format string.</p>
				137
				138	<p>There is little sanity with the naming of the unique ID's right now. Some
				139	start with err_, warn_, ext_ to encode the severity into the name. Since the
				140	enum is referenced in the C++ code that produces the diagnostic, it is somewhat
				141	useful for it to be reasonably short.</p>
				142
				143	<p>The severity of the diagnostic comes from the set {<tt>NOTE</tt>,
				144	<tt>WARNING</tt>, <tt>EXTENSION</tt>, <tt>EXTWARN</tt>, <tt>ERROR</tt>}. The
				145	<tt>ERROR</tt> severity is used for diagnostics indicating the program is never
				146	acceptable under any circumstances. When an error is emitted, the AST for the
				147	input code may not be fully built. The <tt>EXTENSION</tt> and <tt>EXTWARN</tt>
				148	severities are used for extensions to the language that Clang accepts. This
				149	means that Clang fully understands and can represent them in the AST, but we
				150	produce diagnostics to tell the user their code is non-portable. The difference
				151	is that the former are ignored by default, and the later warn by default. The
				152	<tt>WARNING</tt> severity is used for constructs that are valid in the currently
				153	selected source language but that are dubious in some way. The <tt>NOTE</tt>
				154	level is used to staple more information onto a previous diagnostics.</p>
				155
				156	<p>These <em>severities</em> are mapped into a smaller set (the
				157	Diagnostic::Level enum, {<tt>Ignored</tt>, <tt>Note</tt>, <tt>Warning</tt>,
				158	<tt>Error</tt> }) of output <em>levels</em> by the diagnostics subsystem based
				159	on various configuration options. For example, if the user specifies
				160	<tt>-pedantic</tt>, <tt>EXTENSION</tt> maps to <tt>Warning</tt>, if they specify
				161	<tt>-pedantic-errors</tt>, it turns into <tt>Error</tt>. Clang also internally
				162	supports a fully fine grained mapping mechanism that allows you to map any
				163	diagnostic that doesn't have <tt>ERRROR</tt> severity to any output level that
				164	you want. This is used to implement options like <tt>-Wunused_macros</tt>,
				165	<tt>-Wundef</tt> etc.</p>
				166
				167	<!-- ================= -->
				168	<h4>The Format String</h4>
				169	<!-- ================= -->
				170
				171	<p>The format string for the diagnostic is very simple, but it has some power.
				172	It takes the form of a string in English with markers that indicate where and
				173	how arguments to the diagnostic are inserted and formatted. For example, here
				174	are some simple format strings:</p>
				175
				176	<pre>
				177	"binary integer literals are an extension"
				178	"format string contains '\\0' within the string body"
				179	"more '<b>%%</b>' conversions than data arguments"
				180	"invalid operands to binary expression ('<b>%0</b>' and '<b>%1</b>')"
				181	"overloaded '<b>%0</b>' must be a <b>%select{unary\|binary\|unary or binary}2</b> operator"
				182	" (has <b>%1</b> parameter<b>%s1</b>)"
				183	</pre>
				184
				185	<p>These examples show some important points of format strings. You can use any
				186	plain ASCII character in the diagnostic string except "%" without a problem,
				187	but these are C strings, so you have to use and be aware of all the C escape
				188	sequences (as in the second example). If you want to produce a "%" in the
				189	output, use the "%%" escape sequence, like the third diagnostic. Finally,
				190	clang uses the "%...[digit]" sequences to specify where and how arguments to
				191	the diagnostic are formatted.</p>
				192
				193	<p>Arguments to the diagnostic are numbered according to how they are specified
				194	by the C++ code that <a href="#producingdiag">produces them</a>, and are
				195	referenced by <tt>%0</tt> .. <tt>%9</tt>. If you have more than 10 arguments
				196	to your diagnostic, you are doing something wrong. :). Unlike printf, there
				197	is no requirement that arguments to the diagnostic end up in the output in
				198	the same order as they are specified, you could have a format string with
				199	<tt>"%1 %0"</tt> that swaps them, for example. The text in between the
				200	percent and digit are formatting instructions. If there are no instructions,
				201	the argument is just turned into a string and substituted in.</p>
				202
				203	<p>Here are some "best practices" for writing the English format string:</p>
				204
				205	<ul>
				206	<li>Keep the string short. It should ideally fit in the 80 column limit of the
				207	<tt>DiagnosticKinds.def</tt> file. This avoids the diagnostic wrapping when
				208	printed, and forces you to think about the important point you are conveying
				209	with the diagnostic.</li>
				210	<li>Take advantage of location information. The user will be able to see the
				211	line and location of the caret, so you don't need to tell them that the
				212	problem is with the 4th argument to the function: just point to it.</li>
				213	<li>Do not capitalize the diagnostic string, and do not end it with a
				214	period.</li>
				215	<li>If you need to quote something in the diagnostic string, use single
				216	quotes.</li>
				217	</ul>
				218
				219	<p>Diagnostics should never take random English strings as arguments: you
				220	shouldn't use <tt>"you have a problem with %0"</tt> and pass in things like
				221	<tt>"your argument"</tt> or <tt>"your return value"</tt> as arguments. Doing
				222	this prevents <a href="translation">translating</a> the Clang diagnostics to
				223	other languages (because they'll get random English words in their otherwise
				224	localized diagnostic). The exceptions to this are C/C++ language keywords
				225	(e.g. auto, const, mutable, etc) and C/C++ operators (<tt>/=</tt>). Note
				226	that things like "pointer" and "reference" are not keywords. On the other
				227	hand, you <em>can</em> include anything that comes from the user's source code,
				228	including variable names, types, labels, etc.</p>
				229
				230	<!-- ==================================== -->
				231	<h4>Formatting a Diagnostic Argument</a></h4>
				232	<!-- ==================================== -->
				233
				234	<p>Arguments to diagnostics are fully typed internally, and come from a couple
				235	different classes: integers, types, names, and random strings. Depending on
				236	the class of the argument, it can be optionally formatted in different ways.
				237	This gives the DiagnosticClient information about what the argument means
				238	without requiring it to use a specific presentation (consider this MVC for
				239	Clang :).</p>
				240
				241	<p>Here are the different diagnostic argument formats currently supported by
				242	Clang:</p>
				243
				244	<table>
				245	<tr><td colspan="2"><b>"s" format</b></td></tr>
				246	<tr><td>Example:</td><td><tt>"requires %1 parameter%s1"</tt></td></tr>
				247	<tr><td>Classes:</td><td>Integers</td></tr>
				248	<tr><td>Description:</td><td>This is a simple formatter for integers that is
				249	useful when producing English diagnostics. When the integer is 1, it prints
				250	as nothing. When the integer is not 1, it prints as "s". This allows some
				251	simple grammar to be to be handled correctly, and eliminates the need to use
				252	gross things like <tt>"rewrite %1 parameter(s)"</tt>.</td></tr>
				253
				254	<tr><td colspan="2"><b>"select" format</b></td></tr>
				255	<tr><td>Example:</td><td><tt>"must be a %select{unary\|binary\|unary or binary}2
				256	operator"</tt></td></tr>
				257	<tr><td>Classes:</td><td>Integers</td></tr>
				258	<tr><td>Description:</td><td>...</td></tr>
				259
				260	<tr><td colspan="2"><b>"plural" format</b></td></tr>
Sebastian Redl	6816856	2008-11-22 22:16:45 +0000	[diff] [blame^]	261	<tr><td>Example:</td><td><tt>"you have %1 %plural{1:mouse\|:mice}1 connected to
				262	your computer"</tt></td></tr>
Chris Lattner	62fd278	2008-11-22 21:41:31 +0000	[diff] [blame]	263	<tr><td>Classes:</td><td>Integers</td></tr>
Sebastian Redl	6816856	2008-11-22 22:16:45 +0000	[diff] [blame^]	264	<tr><td>Description:</td><td><p>This is a formatter for complex plural forms.
				265	It is designed to handle even the requirements of languages with very
				266	complex plural forms, as many Baltic languages have. The argument consists
				267	of a series of expression/form pairs, separated by ':', where the first form
				268	whose expression evaluates to true is the result of the modifier.</p>
				269	<p>An expression can be empty, in which case it is always true. See the
				270	example at the top. Otherwise, it is a series of one or more numeric
				271	conditions, separated by ','. If any condition matches, the expression
				272	matches. Each numeric condition can take one of three forms.</p>
				273	<ul>
				274	<li>number: A simple decimal number matches if the argument is the same
				275	as the number. Example: <tt>"{1:mouse\|:mice}"</tt></li>
				276	<li>range: A range in square brackets matches if the argument is within
				277	the range. Then range is inclusive both ends. Example:
				278	<tt>"{0:none\|1:one\|[2,5]:some\|:many}"</tt></li>
				279	<li>modulo: A modulo operator is followed by a number, and equals sign
				280	and either a number or a range. The tests are the same as for plain
				281	numbers and ranges, but the argument is taken modulo the number first.
				282	Example: <tt>"{%100=0:even hundred\|%100=[1,50]:lower half\|:everything
				283	else}"</tt></li>
				284	</ul>
				285	<p>The parser is very unforgiving. A syntax error, even whitespace, will
				286	abort, as will a failure to match the argument against any
				287	expression.</p></td></tr>
Chris Lattner	62fd278	2008-11-22 21:41:31 +0000	[diff] [blame]	288
				289
				290	</table>
				291
				292
				293
				294
				295	<!-- ===================================================== -->
				296	<h4><a name="#producingdiag">Producing the Diagnostic</a></h4>
				297	<!-- ===================================================== -->
				298
				299	<p>SemaExpr.cpp example</p>
				300
				301
				302	<!-- ============================================================= -->
				303	<h4><a name="DiagnosticClient">The DiagnosticClient Interface</a></h4>
				304	<!-- ============================================================= -->
				305
				306	<p>Clang command line, buffering, HTMLizing, etc.</p>
				307
				308	<!-- ====================================================== -->
				309	<h4><a name="translation">Adding Translations to Clang</a></h4>
				310	<!-- ====================================================== -->
				311
				312	<p>Not possible yet!</p>
				313
				314
Chris Lattner	86920d3	2007-07-31 05:42:17 +0000	[diff] [blame]	315	<!-- ======================================================================= -->
				316	<h3 id="SourceLocation">The SourceLocation and SourceManager classes</h3>
				317	<!-- ======================================================================= -->
				318
				319	<p>Strangely enough, the SourceLocation class represents a location within the
				320	source code of the program. Important design points include:</p>
				321
				322	<ol>
				323	<li>sizeof(SourceLocation) must be extremely small, as these are embedded into
				324	many AST nodes and are passed around often. Currently it is 32 bits.</li>
				325	<li>SourceLocation must be a simple value object that can be efficiently
				326	copied.</li>
				327	<li>We should be able to represent a source location for any byte of any input
				328	file. This includes in the middle of tokens, in whitespace, in trigraphs,
				329	etc.</li>
				330	<li>A SourceLocation must encode the current #include stack that was active when
				331	the location was processed. For example, if the location corresponds to a
				332	token, it should contain the set of #includes active when the token was
				333	lexed. This allows us to print the #include stack for a diagnostic.</li>
				334	<li>SourceLocation must be able to describe macro expansions, capturing both
				335	the ultimate instantiation point and the source of the original character
				336	data.</li>
				337	</ol>
				338
				339	<p>In practice, the SourceLocation works together with the SourceManager class
				340	to encode two pieces of information about a location: it's physical location
				341	and it's virtual location. For most tokens, these will be the same. However,
				342	for a macro expansion (or tokens that came from a _Pragma directive) these will
				343	describe the location of the characters corresponding to the token and the
				344	location where the token was used (i.e. the macro instantiation point or the
				345	location of the _Pragma itself).</p>
				346
				347	<p>For efficiency, we only track one level of macro instantions: if a token was
				348	produced by multiple instantiations, we only track the source and ultimate
				349	destination. Though we could track the intermediate instantiation points, this
				350	would require extra bookkeeping and no known client would benefit substantially
				351	from this.</p>
				352
				353	<p>The clang front-end inherently depends on the location of a token being
				354	tracked correctly. If it is ever incorrect, the front-end may get confused and
				355	die. The reason for this is that the notion of the 'spelling' of a Token in
				356	clang depends on being able to find the original input characters for the token.
				357	This concept maps directly to the "physical" location for the token.</p>
				358
				359	<!-- ======================================================================= -->
				360	<h2 id="liblex">The Lexer and Preprocessor Library</h2>
				361	<!-- ======================================================================= -->
				362
				363	<p>The Lexer library contains several tightly-connected classes that are involved
				364	with the nasty process of lexing and preprocessing C source code. The main
				365	interface to this library for outside clients is the large <a
				366	href="#Preprocessor">Preprocessor</a> class.
				367	It contains the various pieces of state that are required to coherently read
				368	tokens out of a translation unit.</p>
				369
				370	<p>The core interface to the Preprocessor object (once it is set up) is the
				371	Preprocessor::Lex method, which returns the next <a href="#Token">Token</a> from
				372	the preprocessor stream. There are two types of token providers that the
				373	preprocessor is capable of reading from: a buffer lexer (provided by the <a
				374	href="#Lexer">Lexer</a> class) and a buffered token stream (provided by the <a
Chris Lattner	7928125	2008-03-09 02:27:26 +0000	[diff] [blame]	375	href="#TokenLexer">TokenLexer</a> class).
Chris Lattner	86920d3	2007-07-31 05:42:17 +0000	[diff] [blame]	376
				377
				378	<!-- ======================================================================= -->
				379	<h3 id="Token">The Token class</h3>
				380	<!-- ======================================================================= -->
				381
				382	<p>The Token class is used to represent a single lexed token. Tokens are
				383	intended to be used by the lexer/preprocess and parser libraries, but are not
				384	intended to live beyond them (for example, they should not live in the ASTs).<p>
				385
				386	<p>Tokens most often live on the stack (or some other location that is efficient
				387	to access) as the parser is running, but occasionally do get buffered up. For
				388	example, macro definitions are stored as a series of tokens, and the C++
				389	front-end will eventually need to buffer tokens up for tentative parsing and
				390	various pieces of look-ahead. As such, the size of a Token matter. On a 32-bit
				391	system, sizeof(Token) is currently 16 bytes.</p>
				392
				393	<p>Tokens contain the following information:</p>
				394
				395	<ul>
				396	<li><b>A SourceLocation</b> - This indicates the location of the start of the
				397	token.</li>
				398
				399	<li><b>A length</b> - This stores the length of the token as stored in the
				400	SourceBuffer. For tokens that include them, this length includes trigraphs and
				401	escaped newlines which are ignored by later phases of the compiler. By pointing
				402	into the original source buffer, it is always possible to get the original
				403	spelling of a token completely accurately.</li>
				404
				405	<li><b>IdentifierInfo</b> - If a token takes the form of an identifier, and if
				406	identifier lookup was enabled when the token was lexed (e.g. the lexer was not
				407	reading in 'raw' mode) this contains a pointer to the unique hash value for the
				408	identifier. Because the lookup happens before keyword identification, this
				409	field is set even for language keywords like 'for'.</li>
				410
				411	<li><b>TokenKind</b> - This indicates the kind of token as classified by the
				412	lexer. This includes things like <tt>tok::starequal</tt> (for the "*="
				413	operator), <tt>tok::ampamp</tt> for the "&&" token, and keyword values
				414	(e.g. <tt>tok::kw_for</tt>) for identifiers that correspond to keywords. Note
				415	that some tokens can be spelled multiple ways. For example, C++ supports
				416	"operator keywords", where things like "and" are treated exactly like the
				417	"&&" operator. In these cases, the kind value is set to
				418	<tt>tok::ampamp</tt>, which is good for the parser, which doesn't have to
				419	consider both forms. For something that cares about which form is used (e.g.
				420	the preprocessor 'stringize' operator) the spelling indicates the original
				421	form.</li>
				422
				423	<li><b>Flags</b> - There are currently four flags tracked by the
				424	lexer/preprocessor system on a per-token basis:
				425
				426	<ol>
				427	<li><b>StartOfLine</b> - This was the first token that occurred on its input
				428	source line.</li>
				429	<li><b>LeadingSpace</b> - There was a space character either immediately
				430	before the token or transitively before the token as it was expanded
				431	through a macro. The definition of this flag is very closely defined by
				432	the stringizing requirements of the preprocessor.</li>
				433	<li><b>DisableExpand</b> - This flag is used internally to the preprocessor to
				434	represent identifier tokens which have macro expansion disabled. This
				435	prevents them from being considered as candidates for macro expansion ever
				436	in the future.</li>
				437	<li><b>NeedsCleaning</b> - This flag is set if the original spelling for the
				438	token includes a trigraph or escaped newline. Since this is uncommon,
				439	many pieces of code can fast-path on tokens that did not need cleaning.
				440	</p>
				441	</ol>
				442	</li>
				443	</ul>
				444
				445	<p>One interesting (and somewhat unusual) aspect of tokens is that they don't
				446	contain any semantic information about the lexed value. For example, if the
				447	token was a pp-number token, we do not represent the value of the number that
				448	was lexed (this is left for later pieces of code to decide). Additionally, the
				449	lexer library has no notion of typedef names vs variable names: both are
				450	returned as identifiers, and the parser is left to decide whether a specific
				451	identifier is a typedef or a variable (tracking this requires scope information
				452	among other things).</p>
				453
				454	<!-- ======================================================================= -->
				455	<h3 id="Lexer">The Lexer class</h3>
				456	<!-- ======================================================================= -->
				457
				458	<p>The Lexer class provides the mechanics of lexing tokens out of a source
				459	buffer and deciding what they mean. The Lexer is complicated by the fact that
				460	it operates on raw buffers that have not had spelling eliminated (this is a
				461	necessity to get decent performance), but this is countered with careful coding
				462	as well as standard performance techniques (for example, the comment handling
				463	code is vectorized on X86 and PowerPC hosts).</p>
				464
				465	<p>The lexer has a couple of interesting modal features:</p>
				466
				467	<ul>
				468	<li>The lexer can operate in 'raw' mode. This mode has several features that
				469	make it possible to quickly lex the file (e.g. it stops identifier lookup,
				470	doesn't specially handle preprocessor tokens, handles EOF differently, etc).
				471	This mode is used for lexing within an "<tt>#if 0</tt>" block, for
				472	example.</li>
				473	<li>The lexer can capture and return comments as tokens. This is required to
				474	support the -C preprocessor mode, which passes comments through, and is
				475	used by the diagnostic checker to identifier expect-error annotations.</li>
				476	<li>The lexer can be in ParsingFilename mode, which happens when preprocessing
Chris Lattner	8438624	2007-09-16 19:25:23 +0000	[diff] [blame]	477	after reading a #include directive. This mode changes the parsing of '<'
Chris Lattner	86920d3	2007-07-31 05:42:17 +0000	[diff] [blame]	478	to return an "angled string" instead of a bunch of tokens for each thing
				479	within the filename.</li>
				480	<li>When parsing a preprocessor directive (after "<tt>#</tt>") the
				481	ParsingPreprocessorDirective mode is entered. This changes the parser to
				482	return EOM at a newline.</li>
				483	<li>The Lexer uses a LangOptions object to know whether trigraphs are enabled,
				484	whether C++ or ObjC keywords are recognized, etc.</li>
				485	</ul>
				486
				487	<p>In addition to these modes, the lexer keeps track of a couple of other
				488	features that are local to a lexed buffer, which change as the buffer is
				489	lexed:</p>
				490
				491	<ul>
				492	<li>The Lexer uses BufferPtr to keep track of the current character being
				493	lexed.</li>
				494	<li>The Lexer uses IsAtStartOfLine to keep track of whether the next lexed token
				495	will start with its "start of line" bit set.</li>
				496	<li>The Lexer keeps track of the current #if directives that are active (which
				497	can be nested).</li>
				498	<li>The Lexer keeps track of an <a href="#MultipleIncludeOpt">
				499	MultipleIncludeOpt</a> object, which is used to
				500	detect whether the buffer uses the standard "<tt>#ifndef XX</tt> /
				501	<tt>#define XX</tt>" idiom to prevent multiple inclusion. If a buffer does,
				502	subsequent includes can be ignored if the XX macro is defined.</li>
				503	</ul>
				504
				505	<!-- ======================================================================= -->
Chris Lattner	7928125	2008-03-09 02:27:26 +0000	[diff] [blame]	506	<h3 id="TokenLexer">The TokenLexer class</h3>
Chris Lattner	86920d3	2007-07-31 05:42:17 +0000	[diff] [blame]	507	<!-- ======================================================================= -->
				508
Chris Lattner	7928125	2008-03-09 02:27:26 +0000	[diff] [blame]	509	<p>The TokenLexer class is a token provider that returns tokens from a list
Chris Lattner	86920d3	2007-07-31 05:42:17 +0000	[diff] [blame]	510	of tokens that came from somewhere else. It typically used for two things: 1)
				511	returning tokens from a macro definition as it is being expanded 2) returning
				512	tokens from an arbitrary buffer of tokens. The later use is used by _Pragma and
				513	will most likely be used to handle unbounded look-ahead for the C++ parser.</p>
				514
				515	<!-- ======================================================================= -->
				516	<h3 id="MultipleIncludeOpt">The MultipleIncludeOpt class</h3>
				517	<!-- ======================================================================= -->
				518
				519	<p>The MultipleIncludeOpt class implements a really simple little state machine
				520	that is used to detect the standard "<tt>#ifndef XX</tt> / <tt>#define XX</tt>"
				521	idiom that people typically use to prevent multiple inclusion of headers. If a
				522	buffer uses this idiom and is subsequently #include'd, the preprocessor can
				523	simply check to see whether the guarding condition is defined or not. If so,
				524	the preprocessor can completely ignore the include of the header.</p>
				525
				526
				527
				528	<!-- ======================================================================= -->
				529	<h2 id="libparse">The Parser Library</h2>
				530	<!-- ======================================================================= -->
				531
				532	<!-- ======================================================================= -->
				533	<h2 id="libast">The AST Library</h2>
				534	<!-- ======================================================================= -->
				535
				536	<!-- ======================================================================= -->
				537	<h3 id="Type">The Type class and its subclasses</h3>
				538	<!-- ======================================================================= -->
				539
				540	<p>The Type class (and its subclasses) are an important part of the AST. Types
				541	are accessed through the ASTContext class, which implicitly creates and uniques
				542	them as they are needed. Types have a couple of non-obvious features: 1) they
				543	do not capture type qualifiers like const or volatile (See
				544	<a href="#QualType">QualType</a>), and 2) they implicitly capture typedef
Chris Lattner	8a2bc62	2007-07-31 06:37:39 +0000	[diff] [blame]	545	information. Once created, types are immutable (unlike decls).</p>
Chris Lattner	86920d3	2007-07-31 05:42:17 +0000	[diff] [blame]	546
				547	<p>Typedefs in C make semantic analysis a bit more complex than it would
				548	be without them. The issue is that we want to capture typedef information
				549	and represent it in the AST perfectly, but the semantics of operations need to
				550	"see through" typedefs. For example, consider this code:</p>
				551
				552	<code>
				553	void func() {<br>
Bill Wendling	30d1775	2007-10-06 01:56:01 +0000	[diff] [blame]	554	typedef int foo;<br>
				555	foo X, *Y;<br>
				556	typedef foo* bar;<br>
				557	bar Z;<br>
				558	*X; <i>// error</i><br>
				559	**Y; <i>// error</i><br>
				560	**Z; <i>// error</i><br>
Chris Lattner	86920d3	2007-07-31 05:42:17 +0000	[diff] [blame]	561	}<br>
				562	</code>
				563
				564	<p>The code above is illegal, and thus we expect there to be diagnostics emitted
				565	on the annotated lines. In this example, we expect to get:</p>
				566
				567	<pre>
Chris Lattner	8a2bc62	2007-07-31 06:37:39 +0000	[diff] [blame]	568	<b>test.c:6:1: error: indirection requires pointer operand ('foo' invalid)</b>
Chris Lattner	86920d3	2007-07-31 05:42:17 +0000	[diff] [blame]	569	*X; // error
				570	<font color="blue">^~</font>
Chris Lattner	8a2bc62	2007-07-31 06:37:39 +0000	[diff] [blame]	571	<b>test.c:7:1: error: indirection requires pointer operand ('foo' invalid)</b>
Chris Lattner	86920d3	2007-07-31 05:42:17 +0000	[diff] [blame]	572	**Y; // error
				573	<font color="blue">^~~</font>
Chris Lattner	8a2bc62	2007-07-31 06:37:39 +0000	[diff] [blame]	574	<b>test.c:8:1: error: indirection requires pointer operand ('foo' invalid)</b>
				575	**Z; // error
				576	<font color="blue">^~~</font>
Chris Lattner	86920d3	2007-07-31 05:42:17 +0000	[diff] [blame]	577	</pre>
				578
				579	<p>While this example is somewhat silly, it illustrates the point: we want to
				580	retain typedef information where possible, so that we can emit errors about
				581	"<tt>std::string</tt>" instead of "<tt>std::basic_string<char, std:...</tt>".
				582	Doing this requires properly keeping typedef information (for example, the type
				583	of "X" is "foo", not "int"), and requires properly propagating it through the
Chris Lattner	8a2bc62	2007-07-31 06:37:39 +0000	[diff] [blame]	584	various operators (for example, the type of *Y is "foo", not "int"). In order
				585	to retain this information, the type of these expressions is an instance of the
				586	TypedefType class, which indicates that the type of these expressions is a
				587	typedef for foo.
				588	</p>
Chris Lattner	86920d3	2007-07-31 05:42:17 +0000	[diff] [blame]	589
Chris Lattner	8a2bc62	2007-07-31 06:37:39 +0000	[diff] [blame]	590	<p>Representing types like this is great for diagnostics, because the
				591	user-specified type is always immediately available. There are two problems
				592	with this: first, various semantic checks need to make judgements about the
Chris Lattner	33fc68a	2007-07-31 18:54:50 +0000	[diff] [blame]	593	<em>actual structure</em> of a type, ignoring typdefs. Second, we need an
				594	efficient way to query whether two types are structurally identical to each
				595	other, ignoring typedefs. The solution to both of these problems is the idea of
Chris Lattner	8a2bc62	2007-07-31 06:37:39 +0000	[diff] [blame]	596	canonical types.</p>
Chris Lattner	86920d3	2007-07-31 05:42:17 +0000	[diff] [blame]	597
Chris Lattner	62fd278	2008-11-22 21:41:31 +0000	[diff] [blame]	598	<!-- =============== -->
Chris Lattner	8a2bc62	2007-07-31 06:37:39 +0000	[diff] [blame]	599	<h4>Canonical Types</h4>
Chris Lattner	62fd278	2008-11-22 21:41:31 +0000	[diff] [blame]	600	<!-- =============== -->
Chris Lattner	86920d3	2007-07-31 05:42:17 +0000	[diff] [blame]	601
Chris Lattner	8a2bc62	2007-07-31 06:37:39 +0000	[diff] [blame]	602	<p>Every instance of the Type class contains a canonical type pointer. For
				603	simple types with no typedefs involved (e.g. "<tt>int</tt>", "<tt>int*</tt>",
				604	"<tt>int**</tt>"), the type just points to itself. For types that have a
				605	typedef somewhere in their structure (e.g. "<tt>foo</tt>", "<tt>foo*</tt>",
				606	"<tt>foo**</tt>", "<tt>bar</tt>"), the canonical type pointer points to their
				607	structurally equivalent type without any typedefs (e.g. "<tt>int</tt>",
				608	"<tt>int</tt>", "<tt>int</tt>", and "<tt>int</tt>" respectively).</p>
Chris Lattner	86920d3	2007-07-31 05:42:17 +0000	[diff] [blame]	609
Chris Lattner	8a2bc62	2007-07-31 06:37:39 +0000	[diff] [blame]	610	<p>This design provides a constant time operation (dereferencing the canonical
				611	type pointer) that gives us access to the structure of types. For example,
				612	we can trivially tell that "bar" and "foo*" are the same type by dereferencing
				613	their canonical type pointers and doing a pointer comparison (they both point
				614	to the single "<tt>int*</tt>" type).</p>
				615
				616	<p>Canonical types and typedef types bring up some complexities that must be
				617	carefully managed. Specifically, the "isa/cast/dyncast" operators generally
				618	shouldn't be used in code that is inspecting the AST. For example, when type
				619	checking the indirection operator (unary '*' on a pointer), the type checker
				620	must verify that the operand has a pointer type. It would not be correct to
				621	check that with "<tt>isa<PointerType>(SubExpr->getType())</tt>",
				622	because this predicate would fail if the subexpression had a typedef type.</p>
				623
				624	<p>The solution to this problem are a set of helper methods on Type, used to
				625	check their properties. In this case, it would be correct to use
				626	"<tt>SubExpr->getType()->isPointerType()</tt>" to do the check. This
				627	predicate will return true if the <em>canonical type is a pointer</em>, which is
				628	true any time the type is structurally a pointer type. The only hard part here
				629	is remembering not to use the <tt>isa/cast/dyncast</tt> operations.</p>
				630
				631	<p>The second problem we face is how to get access to the pointer type once we
				632	know it exists. To continue the example, the result type of the indirection
				633	operator is the pointee type of the subexpression. In order to determine the
				634	type, we need to get the instance of PointerType that best captures the typedef
				635	information in the program. If the type of the expression is literally a
				636	PointerType, we can return that, otherwise we have to dig through the
				637	typedefs to find the pointer type. For example, if the subexpression had type
				638	"<tt>foo*</tt>", we could return that type as the result. If the subexpression
				639	had type "<tt>bar</tt>", we want to return "<tt>foo*</tt>" (note that we do
				640	<em>not</em> want "<tt>int*</tt>"). In order to provide all of this, Type has
Chris Lattner	11406c1	2007-07-31 16:50:51 +0000	[diff] [blame]	641	a getAsPointerType() method that checks whether the type is structurally a
Chris Lattner	8a2bc62	2007-07-31 06:37:39 +0000	[diff] [blame]	642	PointerType and, if so, returns the best one. If not, it returns a null
				643	pointer.</p>
				644
				645	<p>This structure is somewhat mystical, but after meditating on it, it will
				646	make sense to you :).</p>
Chris Lattner	86920d3	2007-07-31 05:42:17 +0000	[diff] [blame]	647
				648	<!-- ======================================================================= -->
				649	<h3 id="QualType">The QualType class</h3>
				650	<!-- ======================================================================= -->
				651
				652	<p>The QualType class is designed as a trivial value class that is small,
				653	passed by-value and is efficient to query. The idea of QualType is that it
				654	stores the type qualifiers (const, volatile, restrict) separately from the types
				655	themselves: QualType is conceptually a pair of "Type*" and bits for the type
				656	qualifiers.</p>
				657
				658	<p>By storing the type qualifiers as bits in the conceptual pair, it is
				659	extremely efficient to get the set of qualifiers on a QualType (just return the
				660	field of the pair), add a type qualifier (which is a trivial constant-time
				661	operation that sets a bit), and remove one or more type qualifiers (just return
				662	a QualType with the bitfield set to empty).</p>
				663
				664	<p>Further, because the bits are stored outside of the type itself, we do not
				665	need to create duplicates of types with different sets of qualifiers (i.e. there
				666	is only a single heap allocated "int" type: "const int" and "volatile const int"
				667	both point to the same heap allocated "int" type). This reduces the heap size
				668	used to represent bits and also means we do not have to consider qualifiers when
				669	uniquing types (<a href="#Type">Type</a> does not even contain qualifiers).</p>
				670
				671	<p>In practice, on hosts where it is safe, the 3 type qualifiers are stored in
				672	the low bit of the pointer to the Type object. This means that QualType is
				673	exactly the same size as a pointer, and this works fine on any system where
				674	malloc'd objects are at least 8 byte aligned.</p>
Ted Kremenek	8bc0571	2007-10-10 23:01:43 +0000	[diff] [blame]	675
				676	<!-- ======================================================================= -->
Douglas Gregor	2e1cd42	2008-11-17 14:58:09 +0000	[diff] [blame]	677	<h3 id="DeclarationName">Declaration names</h3>
				678	<!-- ======================================================================= -->
				679
				680	<p>The <tt>DeclarationName</tt> class represents the name of a
				681	declaration in Clang. Declarations in the C family of languages can
				682	take several different forms. Most declarations are named by are
				683	simple identifiers, e.g., "<code>f</code>" and "<code>x</code>" in
				684	the function declaration <code>f(int x)</code>. In C++, declaration
				685	names can also name class constructors ("<code>Class</code>"
				686	in <code>struct Class { Class(); }</code>), class destructors
				687	("<code>~Class</code>"), overloaded operator names ("operator+"),
				688	and conversion functions ("<code>operator void const *</code>"). In
				689	Objective-C, declaration names can refer to the names of Objective-C
				690	methods, which involve the method name and the parameters,
				691	collectively called a <i>selector</i>, e.g..,
				692	"<code>setWidth:height:</code>". Since all of these kinds of
				693	entities--variables, functions, Objective-C methods, C++
				694	constructors, destructors, and operators---are represented as
				695	subclasses of Clang's common <code>NamedDecl</code>
				696	class, <code>DeclarationName</code> is designed to efficiently
				697	represent any kind of name.</p>
				698
				699	<p>Given
				700	a <code>DeclarationName</code> <code>N</code>, <code>N.getNameKind()</code>
Douglas Gregor	2def483	2008-11-17 20:34:05 +0000	[diff] [blame]	701	will produce a value that describes what kind of name <code>N</code>
Douglas Gregor	e94ca9e4	2008-11-18 14:39:36 +0000	[diff] [blame]	702	stores. There are 8 options (all of the names are inside
Douglas Gregor	2e1cd42	2008-11-17 14:58:09 +0000	[diff] [blame]	703	the <code>DeclarationName</code> class)</p>
				704	<dl>
				705	<dt>Identifier</dt>
				706	<dd>The name is a simple
				707	identifier. Use <code>N.getAsIdentifierInfo()</code> to retrieve the
				708	corresponding <code>IdentifierInfo*</code> pointing to the actual
				709	identifier. Note that C++ overloaded operators (e.g.,
				710	"<code>operator+</code>") are represented as special kinds of
				711	identifiers. Use <code>IdentifierInfo</code>'s <code>getOverloadedOperatorID</code>
				712	function to determine whether an identifier is an overloaded
				713	operator name.</dd>
				714
				715	<dt>ObjCZeroArgSelector, ObjCOneArgSelector,
				716	ObjCMultiArgSelector</dt>
				717	<dd>The name is an Objective-C selector, which can be retrieved as a
				718	<code>Selector</code> instance
				719	via <code>N.getObjCSelector()</code>. The three possible name
				720	kinds for Objective-C reflect an optimization within
				721	the <code>DeclarationName</code> class: both zero- and
				722	one-argument selectors are stored as a
				723	masked <code>IdentifierInfo</code> pointer, and therefore require
				724	very little space, since zero- and one-argument selectors are far
				725	more common than multi-argument selectors (which use a different
				726	structure).</dd>
				727
				728	<dt>CXXConstructorName</dt>
				729	<dd>The name is a C++ constructor
				730	name. Use <code>N.getCXXNameType()</code> to retrieve
				731	the <a href="#QualType">type</a> that this constructor is meant to
				732	construct. The type is always the canonical type, since all
				733	constructors for a given type have the same name.</dd>
				734
				735	<dt>CXXDestructorName</dt>
				736	<dd>The name is a C++ destructor
				737	name. Use <code>N.getCXXNameType()</code> to retrieve
				738	the <a href="#QualType">type</a> whose destructor is being
				739	named. This type is always a canonical type.</dd>
				740
				741	<dt>CXXConversionFunctionName</dt>
				742	<dd>The name is a C++ conversion function. Conversion functions are
				743	named according to the type they convert to, e.g., "<code>operator void
				744	const *</code>". Use <code>N.getCXXNameType()</code> to retrieve
				745	the type that this conversion function converts to. This type is
				746	always a canonical type.</dd>
Douglas Gregor	e94ca9e4	2008-11-18 14:39:36 +0000	[diff] [blame]	747
				748	<dt>CXXOperatorName</dt>
				749	<dd>The name is a C++ overloaded operator name. Overloaded operators
				750	are named according to their spelling, e.g.,
				751	"<code>operator+</code>" or "<code>operator new
				752	[]</code>". Use <code>N.getCXXOverloadedOperator()</code> to
				753	retrieve the overloaded operator (a value of
				754	type <code>OverloadedOperatorKind</code>).</dd>
Douglas Gregor	2e1cd42	2008-11-17 14:58:09 +0000	[diff] [blame]	755	</dl>
				756
				757	<p><code>DeclarationName</code>s are cheap to create, copy, and
				758	compare. They require only a single pointer's worth of storage in
Douglas Gregor	e94ca9e4	2008-11-18 14:39:36 +0000	[diff] [blame]	759	the common cases (identifiers, zero-
Douglas Gregor	2e1cd42	2008-11-17 14:58:09 +0000	[diff] [blame]	760	and one-argument Objective-C selectors) and use dense, uniqued
				761	storage for the other kinds of
				762	names. Two <code>DeclarationName</code>s can be compared for
				763	equality (<code>==</code>, <code>!=</code>) using a simple bitwise
				764	comparison, can be ordered
				765	with <code><</code>, <code>></code>, <code><=</code>,
				766	and <code>>=</code> (which provide a lexicographical ordering for
				767	normal identifiers but an unspecified ordering for other kinds of
				768	names), and can be placed into LLVM <code>DenseMap</code>s
				769	and <code>DenseSet</code>s.</p>
				770
				771	<p><code>DeclarationName</code> instances can be created in different
				772	ways depending on what kind of name the instance will store. Normal
Douglas Gregor	e94ca9e4	2008-11-18 14:39:36 +0000	[diff] [blame]	773	identifiers (<code>IdentifierInfo</code> pointers) and Objective-C selectors
Douglas Gregor	2e1cd42	2008-11-17 14:58:09 +0000	[diff] [blame]	774	(<code>Selector</code>) can be implicitly converted
				775	to <code>DeclarationName</code>s. Names for C++ constructors,
Douglas Gregor	e94ca9e4	2008-11-18 14:39:36 +0000	[diff] [blame]	776	destructors, conversion functions, and overloaded operators can be retrieved from
Douglas Gregor	2e1cd42	2008-11-17 14:58:09 +0000	[diff] [blame]	777	the <code>DeclarationNameTable</code>, an instance of which is
				778	available as <code>ASTContext::DeclarationNames</code>. The member
				779	functions <code>getCXXConstructorName</code>, <code>getCXXDestructorName</code>,
Douglas Gregor	e94ca9e4	2008-11-18 14:39:36 +0000	[diff] [blame]	780	<code>getCXXConversionFunctionName</code>, and <code>getCXXOperatorName</code>, respectively,
				781	return <code>DeclarationName</code> instances for the four kinds of
Douglas Gregor	2e1cd42	2008-11-17 14:58:09 +0000	[diff] [blame]	782	C++ special function names.</p>
				783
				784	<!-- ======================================================================= -->
Ted Kremenek	8bc0571	2007-10-10 23:01:43 +0000	[diff] [blame]	785	<h3 id="CFG">The <tt>CFG</tt> class</h3>
				786	<!-- ======================================================================= -->
				787
				788	<p>The <tt>CFG</tt> class is designed to represent a source-level
				789	control-flow graph for a single statement (<tt>Stmt*</tt>). Typically
				790	instances of <tt>CFG</tt> are constructed for function bodies (usually
				791	an instance of <tt>CompoundStmt</tt>), but can also be instantiated to
				792	represent the control-flow of any class that subclasses <tt>Stmt</tt>,
				793	which includes simple expressions. Control-flow graphs are especially
				794	useful for performing
				795	<a href="http://en.wikipedia.org/wiki/Data_flow_analysis#Sensitivities">flow-
				796	or path-sensitive</a> program analyses on a given function.</p>
				797
Chris Lattner	62fd278	2008-11-22 21:41:31 +0000	[diff] [blame]	798	<!-- ============ -->
Ted Kremenek	8bc0571	2007-10-10 23:01:43 +0000	[diff] [blame]	799	<h4>Basic Blocks</h4>
Chris Lattner	62fd278	2008-11-22 21:41:31 +0000	[diff] [blame]	800	<!-- ============ -->
Ted Kremenek	8bc0571	2007-10-10 23:01:43 +0000	[diff] [blame]	801
				802	<p>Concretely, an instance of <tt>CFG</tt> is a collection of basic
				803	blocks. Each basic block is an instance of <tt>CFGBlock</tt>, which
				804	simply contains an ordered sequence of <tt>Stmt*</tt> (each referring
				805	to statements in the AST). The ordering of statements within a block
				806	indicates unconditional flow of control from one statement to the
				807	next. <a href="#ConditionalControlFlow">Conditional control-flow</a>
				808	is represented using edges between basic blocks. The statements
				809	within a given <tt>CFGBlock</tt> can be traversed using
				810	the <tt>CFGBlock::*iterator</tt> interface.</p>
				811
				812	<p>
Ted Kremenek	18e17e7	2007-10-18 22:50:52 +0000	[diff] [blame]	813	A <tt>CFG</tt> object owns the instances of <tt>CFGBlock</tt> within
Ted Kremenek	8bc0571	2007-10-10 23:01:43 +0000	[diff] [blame]	814	the control-flow graph it represents. Each <tt>CFGBlock</tt> within a
				815	CFG is also uniquely numbered (accessible
				816	via <tt>CFGBlock::getBlockID()</tt>). Currently the number is
				817	based on the ordering the blocks were created, but no assumptions
				818	should be made on how <tt>CFGBlock</tt>s are numbered other than their
				819	numbers are unique and that they are numbered from 0..N-1 (where N is
				820	the number of basic blocks in the CFG).</p>
				821
Chris Lattner	62fd278	2008-11-22 21:41:31 +0000	[diff] [blame]	822	<!-- ===================== -->
Ted Kremenek	8bc0571	2007-10-10 23:01:43 +0000	[diff] [blame]	823	<h4>Entry and Exit Blocks</h4>
Chris Lattner	62fd278	2008-11-22 21:41:31 +0000	[diff] [blame]	824	<!-- ===================== -->
Ted Kremenek	8bc0571	2007-10-10 23:01:43 +0000	[diff] [blame]	825
				826	Each instance of <tt>CFG</tt> contains two special blocks:
				827	an <i>entry</i> block (accessible via <tt>CFG::getEntry()</tt>), which
				828	has no incoming edges, and an <i>exit</i> block (accessible
				829	via <tt>CFG::getExit()</tt>), which has no outgoing edges. Neither
				830	block contains any statements, and they serve the role of providing a
				831	clear entrance and exit for a body of code such as a function body.
				832	The presence of these empty blocks greatly simplifies the
				833	implementation of many analyses built on top of CFGs.
				834
Chris Lattner	62fd278	2008-11-22 21:41:31 +0000	[diff] [blame]	835	<!-- ===================================================== -->
Ted Kremenek	8bc0571	2007-10-10 23:01:43 +0000	[diff] [blame]	836	<h4 id ="ConditionalControlFlow">Conditional Control-Flow</h4>
Chris Lattner	62fd278	2008-11-22 21:41:31 +0000	[diff] [blame]	837	<!-- ===================================================== -->
Ted Kremenek	8bc0571	2007-10-10 23:01:43 +0000	[diff] [blame]	838
				839	<p>Conditional control-flow (such as those induced by if-statements
				840	and loops) is represented as edges between <tt>CFGBlock</tt>s.
				841	Because different C language constructs can induce control-flow,
				842	each <tt>CFGBlock</tt> also records an extra <tt>Stmt*</tt> that
				843	represents the <i>terminator</i> of the block. A terminator is simply
				844	the statement that caused the control-flow, and is used to identify
				845	the nature of the conditional control-flow between blocks. For
				846	example, in the case of an if-statement, the terminator refers to
				847	the <tt>IfStmt</tt> object in the AST that represented the given
				848	branch.</p>
				849
				850	<p>To illustrate, consider the following code example:</p>
				851
				852	<code>
				853	int foo(int x) {<br>
				854	x = x + 1;<br>
				855	<br>
				856	if (x > 2) x++;<br>
				857	else {<br>
				858	x += 2;<br>
				859	x *= 2;<br>
				860	}<br>
				861	<br>
				862	return x;<br>
				863	}
				864	</code>
				865
				866	<p>After invoking the parser+semantic analyzer on this code fragment,
				867	the AST of the body of <tt>foo</tt> is referenced by a
				868	single <tt>Stmt*</tt>. We can then construct an instance
				869	of <tt>CFG</tt> representing the control-flow graph of this function
				870	body by single call to a static class method:</p>
				871
				872	<code>
				873	Stmt* FooBody = ...<br>
				874	CFG* FooCFG = <b>CFG::buildCFG</b>(FooBody);
				875	</code>
				876
				877	<p>It is the responsibility of the caller of <tt>CFG::buildCFG</tt>
				878	to <tt>delete</tt> the returned <tt>CFG*</tt> when the CFG is no
				879	longer needed.</p>
				880
				881	<p>Along with providing an interface to iterate over
				882	its <tt>CFGBlock</tt>s, the <tt>CFG</tt> class also provides methods
				883	that are useful for debugging and visualizing CFGs. For example, the
				884	method
				885	<tt>CFG::dump()</tt> dumps a pretty-printed version of the CFG to
				886	standard error. This is especially useful when one is using a
				887	debugger such as gdb. For example, here is the output
				888	of <tt>FooCFG->dump()</tt>:</p>
				889
				890	<code>
				891	[ B5 (ENTRY) ]<br>
				892	Predecessors (0):<br>
				893	Successors (1): B4<br>
				894	<br>
				895	[ B4 ]<br>
				896	1: x = x + 1<br>
				897	2: (x > 2)<br>
				898	<b>T: if [B4.2]</b><br>
				899	Predecessors (1): B5<br>
				900	Successors (2): B3 B2<br>
				901	<br>
				902	[ B3 ]<br>
				903	1: x++<br>
				904	Predecessors (1): B4<br>
				905	Successors (1): B1<br>
				906	<br>
				907	[ B2 ]<br>
				908	1: x += 2<br>
				909	2: x *= 2<br>
				910	Predecessors (1): B4<br>
				911	Successors (1): B1<br>
				912	<br>
				913	[ B1 ]<br>
				914	1: return x;<br>
				915	Predecessors (2): B2 B3<br>
				916	Successors (1): B0<br>
				917	<br>
				918	[ B0 (EXIT) ]<br>
				919	Predecessors (1): B1<br>
				920	Successors (0):
				921	</code>
				922
				923	<p>For each block, the pretty-printed output displays for each block
				924	the number of <i>predecessor</i> blocks (blocks that have outgoing
				925	control-flow to the given block) and <i>successor</i> blocks (blocks
				926	that have control-flow that have incoming control-flow from the given
				927	block). We can also clearly see the special entry and exit blocks at
				928	the beginning and end of the pretty-printed output. For the entry
				929	block (block B5), the number of predecessor blocks is 0, while for the
				930	exit block (block B0) the number of successor blocks is 0.</p>
				931
				932	<p>The most interesting block here is B4, whose outgoing control-flow
				933	represents the branching caused by the sole if-statement
				934	in <tt>foo</tt>. Of particular interest is the second statement in
				935	the block, <b><tt>(x > 2)</tt></b>, and the terminator, printed
				936	as <b><tt>if [B4.2]</tt></b>. The second statement represents the
				937	evaluation of the condition of the if-statement, which occurs before
				938	the actual branching of control-flow. Within the <tt>CFGBlock</tt>
				939	for B4, the <tt>Stmt*</tt> for the second statement refers to the
				940	actual expression in the AST for <b><tt>(x > 2)</tt></b>. Thus
				941	pointers to subclasses of <tt>Expr</tt> can appear in the list of
				942	statements in a block, and not just subclasses of <tt>Stmt</tt> that
				943	refer to proper C statements.</p>
				944
				945	<p>The terminator of block B4 is a pointer to the <tt>IfStmt</tt>
				946	object in the AST. The pretty-printer outputs <b><tt>if
				947	[B4.2]</tt></b> because the condition expression of the if-statement
				948	has an actual place in the basic block, and thus the terminator is
				949	essentially
				950	<i>referring</i> to the expression that is the second statement of
				951	block B4 (i.e., B4.2). In this manner, conditions for control-flow
				952	(which also includes conditions for loops and switch statements) are
				953	hoisted into the actual basic block.</p>
				954
Chris Lattner	62fd278	2008-11-22 21:41:31 +0000	[diff] [blame]	955	<!-- ===================== -->
				956	<!-- <h4>Implicit Control-Flow</h4> -->
				957	<!-- ===================== -->
Ted Kremenek	8bc0571	2007-10-10 23:01:43 +0000	[diff] [blame]	958
				959	<!--
				960	<p>A key design principle of the <tt>CFG</tt> class was to not require
				961	any transformations to the AST in order to represent control-flow.
				962	Thus the <tt>CFG</tt> does not perform any "lowering" of the
				963	statements in an AST: loops are not transformed into guarded gotos,
				964	short-circuit operations are not converted to a set of if-statements,
				965	and so on.</p>
				966	-->
Ted Kremenek	17a295d	2008-06-11 06:19:49 +0000	[diff] [blame]	967
Chris Lattner	7bad199	2008-11-16 21:48:07 +0000	[diff] [blame]	968
				969	<!-- ======================================================================= -->
				970	<h3 id="Constants">Constant Folding in the Clang AST</h3>
				971	<!-- ======================================================================= -->
				972
				973	<p>There are several places where constants and constant folding matter a lot to
				974	the Clang front-end. First, in general, we prefer the AST to retain the source
				975	code as close to how the user wrote it as possible. This means that if they
				976	wrote "5+4", we want to keep the addition and two constants in the AST, we don't
				977	want to fold to "9". This means that constant folding in various ways turns
				978	into a tree walk that needs to handle the various cases.</p>
				979
				980	<p>However, there are places in both C and C++ that require constants to be
				981	folded. For example, the C standard defines what an "integer constant
				982	expression" (i-c-e) is with very precise and specific requirements. The
				983	language then requires i-c-e's in a lot of places (for example, the size of a
				984	bitfield, the value for a case statement, etc). For these, we have to be able
				985	to constant fold the constants, to do semantic checks (e.g. verify bitfield size
				986	is non-negative and that case statements aren't duplicated). We aim for Clang
				987	to be very pedantic about this, diagnosing cases when the code does not use an
				988	i-c-e where one is required, but accepting the code unless running with
				989	<tt>-pedantic-errors</tt>.</p>
				990
				991	<p>Things get a little bit more tricky when it comes to compatibility with
				992	real-world source code. Specifically, GCC has historically accepted a huge
				993	superset of expressions as i-c-e's, and a lot of real world code depends on this
				994	unfortuate accident of history (including, e.g., the glibc system headers). GCC
				995	accepts anything its "fold" optimizer is capable of reducing to an integer
				996	constant, which means that the definition of what it accepts changes as its
				997	optimizer does. One example is that GCC accepts things like "case X-X:" even
				998	when X is a variable, because it can fold this to 0.</p>
				999
				1000	<p>Another issue are how constants interact with the extensions we support, such
				1001	as __builtin_constant_p, __builtin_inf, __extension__ and many others. C99
				1002	obviously does not specify the semantics of any of these extensions, and the
				1003	definition of i-c-e does not include them. However, these extensions are often
				1004	used in real code, and we have to have a way to reason about them.</p>
				1005
				1006	<p>Finally, this is not just a problem for semantic analysis. The code
				1007	generator and other clients have to be able to fold constants (e.g. to
				1008	initialize global variables) and has to handle a superset of what C99 allows.
				1009	Further, these clients can benefit from extended information. For example, we
				1010	know that "foo()\|\|1" always evaluates to true, but we can't replace the
				1011	expression with true because it has side effects.</p>
				1012
				1013	<!-- ======================= -->
				1014	<h4>Implementation Approach</h4>
				1015	<!-- ======================= -->
				1016
				1017	<p>After trying several different approaches, we've finally converged on a
				1018	design (Note, at the time of this writing, not all of this has been implemented,
				1019	consider this a design goal!). Our basic approach is to define a single
				1020	recursive method evaluation method (<tt>Expr::Evaluate</tt>), which is
				1021	implemented in <tt>AST/ExprConstant.cpp</tt>. Given an expression with 'scalar'
				1022	type (integer, fp, complex, or pointer) this method returns the following
				1023	information:</p>
				1024
				1025	<ul>
				1026	<li>Whether the expression is an integer constant expression, a general
				1027	constant that was folded but has no side effects, a general constant that
				1028	was folded but that does have side effects, or an uncomputable/unfoldable
				1029	value.
				1030	</li>
				1031	<li>If the expression was computable in any way, this method returns the APValue
				1032	for the result of the expression.</li>
				1033	<li>If the expression is not evaluatable at all, this method returns
				1034	information on one of the problems with the expression. This includes a
				1035	SourceLocation for where the problem is, and a diagnostic ID that explains
				1036	the problem. The diagnostic should be have ERROR type.</li>
				1037	<li>If the expression is not an integer constant expression, this method returns
				1038	information on one of the problems with the expression. This includes a
				1039	SourceLocation for where the problem is, and a diagnostic ID that explains
				1040	the problem. The diagnostic should be have EXTENSION type.</li>
				1041	</ul>
				1042
				1043	<p>This information gives various clients the flexibility that they want, and we
				1044	will eventually have some helper methods for various extensions. For example,
				1045	Sema should have a <tt>Sema::VerifyIntegerConstantExpression</tt> method, which
				1046	calls Evaluate on the expression. If the expression is not foldable, the error
				1047	is emitted, and it would return true. If the expression is not an i-c-e, the
				1048	EXTENSION diagnostic is emitted. Finally it would return false to indicate that
				1049	the AST is ok.</p>
				1050
				1051	<p>Other clients can use the information in other ways, for example, codegen can
				1052	just use expressions that are foldable in any way.</p>
				1053
				1054	<!-- ========== -->
				1055	<h4>Extensions</h4>
				1056	<!-- ========== -->
				1057
				1058	<p>This section describes how some of the various extensions clang supports
				1059	interacts with constant evaluation:</p>
				1060
				1061	<ul>
				1062	<li><b><tt>__extension__</tt></b>: The expression form of this extension causes
				1063	any evaluatable subexpression to be accepted as an integer constant
				1064	expression.</li>
				1065	<li><b><tt>__builtin_constant_p</tt></b>: This returns true (as a integer
				1066	constant expression) if the operand is any evaluatable constant.</li>
				1067	<li><b><tt>__builtin_choose_expr</tt></b>: The condition is required to be an
				1068	integer constant expression, but we accept any constant as an "extension of
				1069	an extension". This only evaluates one operand depending on which way the
				1070	condition evaluates.</li>
				1071	<li><b><tt>__builtin_classify_type</tt></b>: This always returns an integer
				1072	constant expression.</li>
				1073	<li><b><tt>__builtin_inf,nan,..</tt></b>: These are treated just like a
				1074	floating-point literal.</li>
				1075	<li><b><tt>__builtin_abs,copysign,..</tt></b>: These are constant folded as
				1076	general constant expressions.</li>
				1077	</ul>
				1078
				1079
				1080
				1081
Ted Kremenek	17a295d	2008-06-11 06:19:49 +0000	[diff] [blame]	1082	</div>
				1083	</body>
Douglas Gregor	2e1cd42	2008-11-17 14:58:09 +0000	[diff] [blame]	1084	</html>