Blame - llvm/docs/CodingStandards.html - toolchain/llvm-project

blob: 44a55a55857548133d99fa24c2cd8346a65dda72 [file] [log] [blame]

Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	1	<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"
				2	"http://www.w3.org/TR/html4/strict.dtd">
				3	<html>
				4	<head>
				5	<link rel="stylesheet" href="llvm.css" type="text/css">
Misha Brukman	6c2522a	2008-12-10 23:07:02 +0000	[diff] [blame]	6	<title>LLVM Coding Standards</title>
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	7	</head>
				8	<body>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	9
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	10	<div class="doc_title">
Misha Brukman	6c2522a	2008-12-10 23:07:02 +0000	[diff] [blame]	11	LLVM Coding Standards
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	12	</div>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	13
				14	<ol>
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	15	<li><a href="#introduction">Introduction</a></li>
Chris Lattner	85014f4	2001-07-23 20:40:41 +0000	[diff] [blame]	16	<li><a href="#mechanicalissues">Mechanical Source Issues</a>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	17	<ol>
				18	<li><a href="#sourceformating">Source Code Formatting</a>
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	19	<ol>
				20	<li><a href="#scf_commenting">Commenting</a></li>
				21	<li><a href="#scf_commentformat">Comment Formatting</a></li>
Misha Brukman	03f87d5	2004-10-26 16:18:43 +0000	[diff] [blame]	22	<li><a href="#scf_includes"><tt>#include</tt> Style</a></li>
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	23	<li><a href="#scf_codewidth">Source Code Width</a></li>
				24	<li><a href="#scf_spacestabs">Use Spaces Instead of Tabs</a></li>
				25	<li><a href="#scf_indentation">Indent Code Consistently</a></li>
				26	</ol></li>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	27	<li><a href="#compilerissues">Compiler Issues</a>
				28	<ol>
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	29	<li><a href="#ci_warningerrors">Treat Compiler Warnings Like
				30	Errors</a></li>
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	31	<li><a href="#ci_portable_code">Write Portable Code</a></li>
Reid Spencer	bf6439f	2004-09-23 16:03:48 +0000	[diff] [blame]	32	<li><a href="#ci_class_struct">Use of class/struct Keywords</a></li>
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	33	</ol></li>
				34	</ol></li>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	35	<li><a href="#styleissues">Style Issues</a>
				36	<ol>
				37	<li><a href="#macro">The High Level Issues</a>
				38	<ol>
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	39	<li><a href="#hl_module">A Public Header File <b>is</b> a
				40	Module</a></li>
				41	<li><a href="#hl_dontinclude">#include as Little as Possible</a></li>
				42	<li><a href="#hl_privateheaders">Keep "internal" Headers
				43	Private</a></li>
Chris Lattner	6720d7e	2009-07-22 05:40:54 +0000	[diff] [blame]	44	<li><a href="#hl_earlyexit">Use Early Exits and 'continue' to Simplify
				45	Code</a></li>
Chris Lattner	14abb83	2009-07-28 22:54:04 +0000	[diff] [blame]	46	<li><a href="#hl_else_after_return">Don't use "else" after a
				47	return</a></li>
Chris Lattner	6720d7e	2009-07-22 05:40:54 +0000	[diff] [blame]	48	<li><a href="#hl_predicateloops">Turn Predicate Loops into Predicate
				49	Functions</a></li>
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	50	</ol></li>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	51	<li><a href="#micro">The Low Level Issues</a>
				52	<ol>
Chris Lattner	85ea83e	2006-07-27 04:24:14 +0000	[diff] [blame]	53	<li><a href="#ll_assert">Assert Liberally</a></li>
				54	<li><a href="#ll_ns_std">Do not use 'using namespace std'</a></li>
Bill Wendling	2b52dc1	2006-12-09 01:20:34 +0000	[diff] [blame]	55	<li><a href="#ll_virtual_anch">Provide a virtual method anchor for
Bill Wendling	b062948	2006-12-09 01:35:43 +0000	[diff] [blame]	56	classes in headers</a></li>
Chris Lattner	56c1094	2009-06-30 06:13:23 +0000	[diff] [blame]	57	<li><a href="#ll_end">Don't evaluate end() every time through a
				58	loop</a></li>
Chris Lattner	6720d7e	2009-07-22 05:40:54 +0000	[diff] [blame]	59	<li><a href="#ll_iostream"><tt>#include <iostream></tt> is
				60	<em>forbidden</em></a></li>
Chris Lattner	85ea83e	2006-07-27 04:24:14 +0000	[diff] [blame]	61	<li><a href="#ll_avoidendl">Avoid <tt>std::endl</tt></a></li>
Daniel Dunbar	fcd54f7	2009-07-24 23:04:51 +0000	[diff] [blame]	62	<li><a href="#ll_raw_ostream">Use <tt>raw_ostream</tt></a</li>
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	63	</ol></li>
Chris Lattner	6720d7e	2009-07-22 05:40:54 +0000	[diff] [blame]	64
				65	<li><a href="#nano">Microscopic Details</a>
				66	<ol>
				67	<li><a href="#micro_spaceparen">Spaces Before Parentheses</a></li>
				68	<li><a href="#micro_preincrement">Prefer Preincrement</a></li>
				69	<li><a href="#micro_namespaceindent">Namespace Indentation</a></li>
				70	<li><a href="#micro_anonns">Anonymous Namespaces</a></li>
				71	</ol></li>
				72
				73
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	74	</ol></li>
				75	<li><a href="#seealso">See Also</a></li>
				76	</ol>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	77
Chris Lattner	1dab192	2004-05-23 21:05:07 +0000	[diff] [blame]	78	<div class="doc_author">
Chris Lattner	9567fff	2009-07-22 05:43:01 +0000	[diff] [blame]	79	<p>Written by <a href="mailto:sabre@nondot.org">Chris Lattner</a></p>
Chris Lattner	1dab192	2004-05-23 21:05:07 +0000	[diff] [blame]	80	</div>
				81
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	82
				83	<!-- *********************************************************************** -->
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	84	<div class="doc_section">
				85	<a name="introduction">Introduction</a>
				86	</div>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	87	<!-- *********************************************************************** -->
				88
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	89	<div class="doc_text">
				90
				91	<p>This document attempts to describe a few coding standards that are being used
				92	in the LLVM source tree. Although no coding standards should be regarded as
Misha Brukman	0cedb1f	2003-10-06 19:26:00 +0000	[diff] [blame]	93	absolute requirements to be followed in all instances, coding standards can be
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	94	useful.</p>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	95
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	96	<p>This document intentionally does not prescribe fixed standards for religious
Misha Brukman	0cedb1f	2003-10-06 19:26:00 +0000	[diff] [blame]	97	issues such as brace placement and space usage. For issues like this, follow
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	98	the golden rule:</p>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	99
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	100	<blockquote>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	101
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	102	<p><b><a name="goldenrule">If you are adding a significant body of source to a
				103	project, feel free to use whatever style you are most comfortable with. If you
				104	are extending, enhancing, or bug fixing already implemented code, use the style
				105	that is already being used so that the source is uniform and easy to
				106	follow.</a></b></p>
				107
				108	</blockquote>
				109
				110	<p>The ultimate goal of these guidelines is the increase readability and
Misha Brukman	0cedb1f	2003-10-06 19:26:00 +0000	[diff] [blame]	111	maintainability of our common source base. If you have suggestions for topics to
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	112	be included, please mail them to <a
				113	href="mailto:sabre@nondot.org">Chris</a>.</p>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	114
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	115	</div>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	116
				117	<!-- *********************************************************************** -->
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	118	<div class="doc_section">
				119	<a name="mechanicalissues">Mechanical Source Issues</a>
				120	</div>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	121	<!-- *********************************************************************** -->
				122
				123	<!-- ======================================================================= -->
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	124	<div class="doc_subsection">
				125	<a name="sourceformating">Source Code Formatting</a>
				126	</div>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	127
				128	<!-- _______________________________________________________________________ -->
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	129	<div class="doc_subsubsection">
				130	<a name="scf_commenting">Commenting</a>
				131	</div>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	132
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	133	<div class="doc_text">
				134
				135	<p>Comments are one critical part of readability and maintainability. Everyone
Chris Lattner	bbe228b	2009-07-12 00:10:24 +0000	[diff] [blame]	136	knows they should comment, so should you. When writing comments, write them as
				137	English prose, which means they should use proper capitalization, punctuation,
				138	etc. Although we all should probably
Misha Brukman	0cedb1f	2003-10-06 19:26:00 +0000	[diff] [blame]	139	comment our code more than we do, there are a few very critical places that
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	140	documentation is very useful:</p>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	141
Chris Lattner	1dab192	2004-05-23 21:05:07 +0000	[diff] [blame]	142	<b>File Headers</b>
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	143
Reid Spencer	c7f87f2	2007-07-09 08:04:31 +0000	[diff] [blame]	144	<p>Every source file should have a header on it that describes the basic
				145	purpose of the file. If a file does not have a header, it should not be
				146	checked into Subversion. Most source trees will probably have a standard
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	147	file header format. The standard format for the LLVM source tree looks like
				148	this:</p>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	149
Misha Brukman	b224615	2004-07-28 22:31:54 +0000	[diff] [blame]	150	<div class="doc_code">
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	151	<pre>
Chris Lattner	e6f4e07	2003-10-13 14:58:11 +0000	[diff] [blame]	152	//===-- llvm/Instruction.h - Instruction class definition -------- C++ --===//
Misha Brukman	02805a6	2009-01-02 16:58:42 +0000	[diff] [blame]	153	//
Chris Lattner	1dab192	2004-05-23 21:05:07 +0000	[diff] [blame]	154	// The LLVM Compiler Infrastructure
				155	//
Chris Lattner	ebf5666	2007-12-29 19:56:08 +0000	[diff] [blame]	156	// This file is distributed under the University of Illinois Open Source
				157	// License. See LICENSE.TXT for details.
Misha Brukman	02805a6	2009-01-02 16:58:42 +0000	[diff] [blame]	158	//
Chris Lattner	1dab192	2004-05-23 21:05:07 +0000	[diff] [blame]	159	//===----------------------------------------------------------------------===//
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	160	//
				161	// This file contains the declaration of the Instruction class, which is the
				162	// base class for all of the VM instructions.
				163	//
				164	//===----------------------------------------------------------------------===//
				165	</pre>
Misha Brukman	b224615	2004-07-28 22:31:54 +0000	[diff] [blame]	166	</div>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	167
Chris Lattner	ebf5666	2007-12-29 19:56:08 +0000	[diff] [blame]	168	<p>A few things to note about this particular format: The "<tt>-*- C++
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	169	-*-</tt>" string on the first line is there to tell Emacs that the source file
Misha Brukman	0d640e8	2004-07-28 22:37:57 +0000	[diff] [blame]	170	is a C++ file, not a C file (Emacs assumes .h files are C files by default).
				171	Note that this tag is not necessary in .cpp files. The name of the file is also
				172	on the first line, along with a very short description of the purpose of the
				173	file. This is important when printing out code and flipping though lots of
				174	pages.</p>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	175
Chris Lattner	ebf5666	2007-12-29 19:56:08 +0000	[diff] [blame]	176	<p>The next section in the file is a concise note that defines the license
				177	that the file is released under. This makes it perfectly clear what terms the
				178	source code can be distributed under and should not be modified in any way.</p>
Chris Lattner	1dab192	2004-05-23 21:05:07 +0000	[diff] [blame]	179
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	180	<p>The main body of the description does not have to be very long in most cases.
Misha Brukman	0cedb1f	2003-10-06 19:26:00 +0000	[diff] [blame]	181	Here it's only two lines. If an algorithm is being implemented or something
				182	tricky is going on, a reference to the paper where it is published should be
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	183	included, as well as any notes or "gotchas" in the code to watch out for.</p>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	184
Chris Lattner	1dab192	2004-05-23 21:05:07 +0000	[diff] [blame]	185	<b>Class overviews</b>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	186
Jim Laskey	88b5e79	2006-07-31 20:18:49 +0000	[diff] [blame]	187	<p>Classes are one fundamental part of a good object oriented design. As such,
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	188	a class definition should have a comment block that explains what the class is
Misha Brukman	0cedb1f	2003-10-06 19:26:00 +0000	[diff] [blame]	189	used for... if it's not obvious. If it's so completely obvious your grandma
				190	could figure it out, it's probably safe to leave it out. Naming classes
Misha Brukman	6d1686c	2004-12-04 00:32:12 +0000	[diff] [blame]	191	something sane goes a long ways towards avoiding writing documentation.</p>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	192
				193
Chris Lattner	1dab192	2004-05-23 21:05:07 +0000	[diff] [blame]	194	<b>Method information</b>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	195
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	196	<p>Methods defined in a class (as well as any global functions) should also be
Dan Gohman	42f5691	2010-02-26 20:18:32 +0000	[diff] [blame]	197	documented properly. A quick note about what it does and a description of the
Misha Brukman	0cedb1f	2003-10-06 19:26:00 +0000	[diff] [blame]	198	borderline behaviour is all that is necessary here (unless something
Dan Gohman	42f5691	2010-02-26 20:18:32 +0000	[diff] [blame]	199	particularly tricky or insidious is going on). The hope is that people can
Misha Brukman	0cedb1f	2003-10-06 19:26:00 +0000	[diff] [blame]	200	figure out how to use your interfaces without reading the code itself... that is
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	201	the goal metric.</p>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	202
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	203	<p>Good things to talk about here are what happens when something unexpected
				204	happens: does the method return null? Abort? Format your hard disk?</p>
				205
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	206	</div>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	207
				208	<!-- _______________________________________________________________________ -->
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	209	<div class="doc_subsubsection">
				210	<a name="scf_commentformat">Comment Formatting</a>
				211	</div>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	212
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	213	<div class="doc_text">
				214
				215	<p>In general, prefer C++ style (<tt>//</tt>) comments. They take less space,
Misha Brukman	0cedb1f	2003-10-06 19:26:00 +0000	[diff] [blame]	216	require less typing, don't have nesting problems, etc. There are a few cases
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	217	when it is useful to use C style (<tt>/* */</tt>) comments however:</p>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	218
				219	<ol>
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	220	<li>When writing a C code: Obviously if you are writing C code, use C style
Misha Brukman	6d1686c	2004-12-04 00:32:12 +0000	[diff] [blame]	221	comments.</li>
Misha Brukman	03f87d5	2004-10-26 16:18:43 +0000	[diff] [blame]	222	<li>When writing a header file that may be <tt>#include</tt>d by a C source
				223	file.</li>
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	224	<li>When writing a source file that is used by a tool that only accepts C
				225	style comments.</li>
				226	</ol>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	227
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	228	<p>To comment out a large block of code, use <tt>#if 0</tt> and <tt>#endif</tt>.
				229	These nest properly and are better behaved in general than C style comments.</p>
				230
				231	</div>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	232
Chris Lattner	245b525	2003-08-07 21:45:47 +0000	[diff] [blame]	233	<!-- _______________________________________________________________________ -->
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	234	<div class="doc_subsubsection">
Misha Brukman	03f87d5	2004-10-26 16:18:43 +0000	[diff] [blame]	235	<a name="scf_includes"><tt>#include</tt> Style</a>
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	236	</div>
Chris Lattner	245b525	2003-08-07 21:45:47 +0000	[diff] [blame]	237
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	238	<div class="doc_text">
				239
				240	<p>Immediately after the <a href="#scf_commenting">header file comment</a> (and
Chris Lattner	245b525	2003-08-07 21:45:47 +0000	[diff] [blame]	241	include guards if working on a header file), the <a
Misha Brukman	b224615	2004-07-28 22:31:54 +0000	[diff] [blame]	242	href="#hl_dontinclude">minimal</a> list of <tt>#include</tt>s required by the
				243	file should be listed. We prefer these <tt>#include</tt>s to be listed in this
				244	order:</p>
Chris Lattner	245b525	2003-08-07 21:45:47 +0000	[diff] [blame]	245
				246	<ol>
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	247	<li><a href="#mmheader">Main Module header</a></li>
				248	<li><a href="#hl_privateheaders">Local/Private Headers</a></li>
Misha Brukman	b224615	2004-07-28 22:31:54 +0000	[diff] [blame]	249	<li><tt>llvm/*</tt></li>
				250	<li><tt>llvm/Analysis/*</tt></li>
				251	<li><tt>llvm/Assembly/*</tt></li>
Duncan Sands	e9da6db	2010-09-29 20:09:55 +0000	[diff] [blame]	252	<li><tt>llvm/Bitcode/*</tt></li>
Misha Brukman	b224615	2004-07-28 22:31:54 +0000	[diff] [blame]	253	<li><tt>llvm/CodeGen/*</tt></li>
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	254	<li>...</li>
Misha Brukman	b224615	2004-07-28 22:31:54 +0000	[diff] [blame]	255	<li><tt>Support/*</tt></li>
				256	<li><tt>Config/*</tt></li>
				257	<li>System <tt>#includes</tt></li>
Chris Lattner	245b525	2003-08-07 21:45:47 +0000	[diff] [blame]	258	</ol>
				259
Nick Lewycky	3584780	2008-11-29 20:13:25 +0000	[diff] [blame]	260	<p>... and each category should be sorted by name.</p>
Chris Lattner	245b525	2003-08-07 21:45:47 +0000	[diff] [blame]	261
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	262	<p><a name="mmheader">The "Main Module Header"</a> file applies to .cpp file
Misha Brukman	03f87d5	2004-10-26 16:18:43 +0000	[diff] [blame]	263	which implement an interface defined by a .h file. This <tt>#include</tt>
				264	should always be included <b>first</b> regardless of where it lives on the file
				265	system. By including a header file first in the .cpp files that implement the
				266	interfaces, we ensure that the header does not have any hidden dependencies
				267	which are not explicitly #included in the header, but should be. It is also a
				268	form of documentation in the .cpp file to indicate where the interfaces it
				269	implements are defined.</p>
Chris Lattner	245b525	2003-08-07 21:45:47 +0000	[diff] [blame]	270
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	271	</div>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	272
				273	<!-- _______________________________________________________________________ -->
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	274	<div class="doc_subsubsection">
				275	<a name="scf_codewidth">Source Code Width</a>
				276	</div>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	277
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	278	<div class="doc_text">
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	279
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	280	<p>Write your code to fit within 80 columns of text. This helps those of us who
				281	like to print out code and look at your code in an xterm without resizing
				282	it.</p>
				283
Chris Lattner	fc12d2e	2008-07-08 05:12:37 +0000	[diff] [blame]	284	<p>The longer answer is that there must be some limit to the width of the code
				285	in order to reasonably allow developers to have multiple files side-by-side in
				286	windows on a modest display. If you are going to pick a width limit, it is
				287	somewhat arbitrary but you might as well pick something standard. Going with
				288	90 columns (for example) instead of 80 columns wouldn't add any significant
				289	value and would be detrimental to printing out code. Also many other projects
				290	have standardized on 80 columns, so some people have already configured their
				291	editors for it (vs something else, like 90 columns).</p>
				292
				293	<p>This is one of many contentious issues in coding standards, but is not up
				294	for debate.</p>
				295
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	296	</div>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	297
				298	<!-- _______________________________________________________________________ -->
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	299	<div class="doc_subsubsection">
				300	<a name="scf_spacestabs">Use Spaces Instead of Tabs</a>
				301	</div>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	302
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	303	<div class="doc_text">
				304
				305	<p>In all cases, prefer spaces to tabs in source files. People have different
Benjamin Kramer	0f42038	2009-10-12 14:46:08 +0000	[diff] [blame]	306	preferred indentation levels, and different styles of indentation that they
Misha Brukman	0cedb1f	2003-10-06 19:26:00 +0000	[diff] [blame]	307	like... this is fine. What isn't is that different editors/viewers expand tabs
				308	out to different tab stops. This can cause your code to look completely
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	309	unreadable, and it is not worth dealing with.</p>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	310
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	311	<p>As always, follow the <a href="#goldenrule">Golden Rule</a> above: follow the
Misha Brukman	0cedb1f	2003-10-06 19:26:00 +0000	[diff] [blame]	312	style of existing code if your are modifying and extending it. If you like four
				313	spaces of indentation, <b>DO NOT</b> do that in the middle of a chunk of code
				314	with two spaces of indentation. Also, do not reindent a whole source file: it
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	315	makes for incredible diffs that are absolutely worthless.</p>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	316
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	317	</div>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	318
				319	<!-- _______________________________________________________________________ -->
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	320	<div class="doc_subsubsection">
				321	<a name="scf_indentation">Indent Code Consistently</a>
				322	</div>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	323
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	324	<div class="doc_text">
				325
				326	<p>Okay, your first year of programming you were told that indentation is
Misha Brukman	0cedb1f	2003-10-06 19:26:00 +0000	[diff] [blame]	327	important. If you didn't believe and internalize this then, now is the time.
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	328	Just do it.</p>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	329
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	330	</div>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	331
				332
				333	<!-- ======================================================================= -->
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	334	<div class="doc_subsection">
				335	<a name="compilerissues">Compiler Issues</a>
				336	</div>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	337
				338
				339	<!-- _______________________________________________________________________ -->
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	340	<div class="doc_subsubsection">
				341	<a name="ci_warningerrors">Treat Compiler Warnings Like Errors</a>
				342	</div>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	343
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	344	<div class="doc_text">
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	345
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	346	<p>If your code has compiler warnings in it, something is wrong: you aren't
				347	casting values correctly, your have "questionable" constructs in your code, or
				348	you are doing something legitimately wrong. Compiler warnings can cover up
				349	legitimate errors in output and make dealing with a translation unit
				350	difficult.</p>
				351
				352	<p>It is not possible to prevent all warnings from all compilers, nor is it
Misha Brukman	0cedb1f	2003-10-06 19:26:00 +0000	[diff] [blame]	353	desirable. Instead, pick a standard compiler (like <tt>gcc</tt>) that provides
				354	a good thorough set of warnings, and stick to them. At least in the case of
				355	<tt>gcc</tt>, it is possible to work around any spurious errors by changing the
				356	syntax of the code slightly. For example, an warning that annoys me occurs when
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	357	I write code like this:</p>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	358
Misha Brukman	b224615	2004-07-28 22:31:54 +0000	[diff] [blame]	359	<div class="doc_code">
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	360	<pre>
Misha Brukman	b224615	2004-07-28 22:31:54 +0000	[diff] [blame]	361	if (V = getValue()) {
				362	...
				363	}
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	364	</pre>
Misha Brukman	b224615	2004-07-28 22:31:54 +0000	[diff] [blame]	365	</div>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	366
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	367	<p><tt>gcc</tt> will warn me that I probably want to use the <tt>==</tt>
				368	operator, and that I probably mistyped it. In most cases, I haven't, and I
				369	really don't want the spurious errors. To fix this particular problem, I
				370	rewrite the code like this:</p>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	371
Misha Brukman	b224615	2004-07-28 22:31:54 +0000	[diff] [blame]	372	<div class="doc_code">
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	373	<pre>
Misha Brukman	b224615	2004-07-28 22:31:54 +0000	[diff] [blame]	374	if ((V = getValue())) {
				375	...
				376	}
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	377	</pre>
Misha Brukman	b224615	2004-07-28 22:31:54 +0000	[diff] [blame]	378	</div>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	379
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	380	<p>...which shuts <tt>gcc</tt> up. Any <tt>gcc</tt> warning that annoys you can
				381	be fixed by massaging the code appropriately.</p>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	382
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	383	<p>These are the <tt>gcc</tt> warnings that I prefer to enable: <tt>-Wall
				384	-Winline -W -Wwrite-strings -Wno-unused</tt></p>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	385
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	386	</div>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	387
				388	<!-- _______________________________________________________________________ -->
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	389	<div class="doc_subsubsection">
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	390	<a name="ci_portable_code">Write Portable Code</a>
				391	</div>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	392
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	393	<div class="doc_text">
				394
				395	<p>In almost all cases, it is possible and within reason to write completely
Misha Brukman	c3e7893	2003-07-28 21:57:18 +0000	[diff] [blame]	396	portable code. If there are cases where it isn't possible to write portable
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	397	code, isolate it behind a well defined (and well documented) interface.</p>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	398
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	399	<p>In practice, this means that you shouldn't assume much about the host
Chris Lattner	44c2241	2010-11-16 22:19:06 +0000	[diff] [blame]	400	compiler, and Visual Studio tends to be the lowest common denominator.
				401	If advanced features are used, they should only be an implementation detail of
				402	a library which has a simple exposed API, and preferably be buried in
				403	libSystem.</p>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	404
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	405	</div>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	406
Reid Spencer	bf6439f	2004-09-23 16:03:48 +0000	[diff] [blame]	407	<!-- _______________________________________________________________________ -->
				408	<div class="doc_subsubsection">
				409	<a name="ci_class_struct">Use of <tt>class</tt> and <tt>struct</tt> Keywords</a>
				410	</div>
				411	<div class="doc_text">
Misha Brukman	f249913	2004-10-26 15:45:13 +0000	[diff] [blame]	412
				413	<p>In C++, the <tt>class</tt> and <tt>struct</tt> keywords can be used almost
				414	interchangeably. The only difference is when they are used to declare a class:
				415	<tt>class</tt> makes all members private by default while <tt>struct</tt> makes
				416	all members public by default.</p>
				417
				418	<p>Unfortunately, not all compilers follow the rules and some will generate
				419	different symbols based on whether <tt>class</tt> or <tt>struct</tt> was used to
				420	declare the symbol. This can lead to problems at link time.</p>
				421
				422	<p>So, the rule for LLVM is to always use the <tt>class</tt> keyword, unless
Chris Lattner	73e9f8c	2009-08-11 17:38:15 +0000	[diff] [blame]	423	<b>all</b> members are public and the type is a C++ "POD" type, in which case
				424	<tt>struct</tt> is allowed.</p>
Misha Brukman	f249913	2004-10-26 15:45:13 +0000	[diff] [blame]	425
Reid Spencer	bf6439f	2004-09-23 16:03:48 +0000	[diff] [blame]	426	</div>
				427
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	428	<!-- *********************************************************************** -->
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	429	<div class="doc_section">
				430	<a name="styleissues">Style Issues</a>
				431	</div>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	432	<!-- *********************************************************************** -->
				433
				434
				435	<!-- ======================================================================= -->
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	436	<div class="doc_subsection">
				437	<a name="macro">The High Level Issues</a>
				438	</div>
Chris Lattner	6720d7e	2009-07-22 05:40:54 +0000	[diff] [blame]	439	<!-- ======================================================================= -->
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	440
				441
				442	<!-- _______________________________________________________________________ -->
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	443	<div class="doc_subsubsection">
				444	<a name="hl_module">A Public Header File <b>is</b> a Module</a>
				445	</div>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	446
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	447	<div class="doc_text">
				448
				449	<p>C++ doesn't do too well in the modularity department. There is no real
Misha Brukman	c3e7893	2003-07-28 21:57:18 +0000	[diff] [blame]	450	encapsulation or data hiding (unless you use expensive protocol classes), but it
				451	is what we have to work with. When you write a public header file (in the LLVM
				452	source tree, they live in the top level "include" directory), you are defining a
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	453	module of functionality.</p>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	454
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	455	<p>Ideally, modules should be completely independent of each other, and their
Misha Brukman	c3e7893	2003-07-28 21:57:18 +0000	[diff] [blame]	456	header files should only include the absolute minimum number of headers
				457	possible. A module is not just a class, a function, or a namespace: <a
				458	href="http://www.cuj.com/articles/2000/0002/0002c/0002c.htm">it's a collection
				459	of these</a> that defines an interface. This interface may be several
				460	functions, classes or data structures, but the important issue is how they work
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	461	together.</p>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	462
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	463	<p>In general, a module should be implemented with one or more <tt>.cpp</tt>
				464	files. Each of these <tt>.cpp</tt> files should include the header that defines
				465	their interface first. This ensure that all of the dependences of the module
				466	header have been properly added to the module header itself, and are not
				467	implicit. System headers should be included after user headers for a
				468	translation unit.</p>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	469
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	470	</div>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	471
				472	<!-- _______________________________________________________________________ -->
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	473	<div class="doc_subsubsection">
Misha Brukman	03f87d5	2004-10-26 16:18:43 +0000	[diff] [blame]	474	<a name="hl_dontinclude"><tt>#include</tt> as Little as Possible</a>
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	475	</div>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	476
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	477	<div class="doc_text">
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	478
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	479	<p><tt>#include</tt> hurts compile time performance. Don't do it unless you
				480	have to, especially in header files.</p>
				481
				482	<p>But wait, sometimes you need to have the definition of a class to use it, or
Misha Brukman	03f87d5	2004-10-26 16:18:43 +0000	[diff] [blame]	483	to inherit from it. In these cases go ahead and <tt>#include</tt> that header
				484	file. Be aware however that there are many cases where you don't need to have
				485	the full definition of a class. If you are using a pointer or reference to a
				486	class, you don't need the header file. If you are simply returning a class
				487	instance from a prototyped function or method, you don't need it. In fact, for
				488	most cases, you simply don't need the definition of a class... and not
				489	<tt>#include</tt>'ing speeds up compilation.</p>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	490
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	491	<p>It is easy to try to go too overboard on this recommendation, however. You
Chris Lattner	756f83f	2007-02-10 18:35:31 +0000	[diff] [blame]	492	<b>must</b> include all of the header files that you are using -- you can
				493	include them either directly
Misha Brukman	c3e7893	2003-07-28 21:57:18 +0000	[diff] [blame]	494	or indirectly (through another header file). To make sure that you don't
Benjamin Kramer	0f42038	2009-10-12 14:46:08 +0000	[diff] [blame]	495	accidentally forget to include a header file in your module header, make sure to
Misha Brukman	c3e7893	2003-07-28 21:57:18 +0000	[diff] [blame]	496	include your module header <b>first</b> in the implementation file (as mentioned
				497	above). This way there won't be any hidden dependencies that you'll find out
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	498	about later...</p>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	499
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	500	</div>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	501
				502	<!-- _______________________________________________________________________ -->
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	503	<div class="doc_subsubsection">
				504	<a name="hl_privateheaders">Keep "internal" Headers Private</a>
				505	</div>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	506
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	507	<div class="doc_text">
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	508
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	509	<p>Many modules have a complex implementation that causes them to use more than
				510	one implementation (<tt>.cpp</tt>) file. It is often tempting to put the
				511	internal communication interface (helper classes, extra functions, etc) in the
Misha Brukman	6d1686c	2004-12-04 00:32:12 +0000	[diff] [blame]	512	public module header file. Don't do this.</p>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	513
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	514	<p>If you really need to do something like this, put a private header file in
				515	the same directory as the source files, and include it locally. This ensures
				516	that your private interface remains private and undisturbed by outsiders.</p>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	517
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	518	<p>Note however, that it's okay to put extra implementation methods a public
				519	class itself... just make them private (or protected), and all is well.</p>
				520
				521	</div>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	522
Bill Wendling	2b52dc1	2006-12-09 01:20:34 +0000	[diff] [blame]	523	<!-- _______________________________________________________________________ -->
				524	<div class="doc_subsubsection">
Chris Lattner	6720d7e	2009-07-22 05:40:54 +0000	[diff] [blame]	525	<a name="hl_earlyexit">Use Early Exits and 'continue' to Simplify Code</a>
Bill Wendling	2b52dc1	2006-12-09 01:20:34 +0000	[diff] [blame]	526	</div>
				527
				528	<div class="doc_text">
				529
Chris Lattner	6720d7e	2009-07-22 05:40:54 +0000	[diff] [blame]	530	<p>When reading code, keep in mind how much state and how many previous
				531	decisions have to be remembered by the reader to understand a block of code.
				532	Aim to reduce indentation where possible when it doesn't make it more difficult
				533	to understand the code. One great way to do this is by making use of early
				534	exits and the 'continue' keyword in long loops. As an example of using an early
				535	exit from a function, consider this "bad" code:</p>
Bill Wendling	2b52dc1	2006-12-09 01:20:34 +0000	[diff] [blame]	536
Chris Lattner	6720d7e	2009-07-22 05:40:54 +0000	[diff] [blame]	537	<div class="doc_code">
				538	<pre>
				539	Value DoSomething(Instruction I) {
				540	if (!isa<TerminatorInst>(I) &&
				541	I->hasOneUse() && SomeOtherThing(I)) {
				542	... some long code ....
				543	}
				544
				545	return 0;
				546	}
				547	</pre>
				548	</div>
Bill Wendling	2b52dc1	2006-12-09 01:20:34 +0000	[diff] [blame]	549
Chris Lattner	6720d7e	2009-07-22 05:40:54 +0000	[diff] [blame]	550	<p>This code has several problems if the body of the 'if' is large. When you're
				551	looking at the top of the function, it isn't immediately clear that this
				552	<em>only</em> does interesting things with non-terminator instructions, and only
				553	applies to things with the other predicates. Second, it is relatively difficult
				554	to describe (in comments) why these predicates are important because the if
				555	statement makes it difficult to lay out the comments. Third, when you're deep
				556	within the body of the code, it is indented an extra level. Finally, when
				557	reading the top of the function, it isn't clear what the result is if the
				558	predicate isn't true, you have to read to the end of the function to know that
				559	it returns null.</p>
Matthijs Kooijman	ca4ffaa	2008-07-30 12:14:10 +0000	[diff] [blame]	560
Chris Lattner	6720d7e	2009-07-22 05:40:54 +0000	[diff] [blame]	561	<p>It is much preferred to format the code like this:</p>
				562
				563	<div class="doc_code">
				564	<pre>
				565	Value DoSomething(Instruction I) {
				566	// Terminators never need 'something' done to them because, ...
				567	if (isa<TerminatorInst>(I))
				568	return 0;
				569
				570	// We conservatively avoid transforming instructions with multiple uses
				571	// because goats like cheese.
				572	if (!I->hasOneUse())
				573	return 0;
				574
				575	// This is really just here for example.
				576	if (!SomeOtherThing(I))
				577	return 0;
				578
				579	... some long code ....
				580	}
				581	</pre>
				582	</div>
				583
				584	<p>This fixes these problems. A similar problem frequently happens in for
				585	loops. A silly example is something like this:</p>
				586
				587	<div class="doc_code">
				588	<pre>
				589	for (BasicBlock::iterator II = BB->begin(), E = BB->end(); II != E; ++II) {
				590	if (BinaryOperator *BO = dyn_cast<BinaryOperator>(II)) {
				591	Value *LHS = BO->getOperand(0);
				592	Value *RHS = BO->getOperand(1);
				593	if (LHS != RHS) {
				594	...
				595	}
				596	}
				597	}
				598	</pre>
				599	</div>
				600
				601	<p>When you have very very small loops, this sort of structure is fine, but if
				602	it exceeds more than 10-15 lines, it becomes difficult for people to read and
				603	understand at a glance.
				604	The problem with this sort of code is that it gets very nested very quickly,
				605	meaning that the reader of the code has to keep a lot of context in their brain
				606	to remember what is going immediately on in the loop, because they don't know
				607	if/when the if conditions will have elses etc. It is strongly preferred to
				608	structure the loop like this:</p>
				609
				610	<div class="doc_code">
				611	<pre>
				612	for (BasicBlock::iterator II = BB->begin(), E = BB->end(); II != E; ++II) {
				613	BinaryOperator *BO = dyn_cast<BinaryOperator>(II);
				614	if (!BO) continue;
				615
				616	Value *LHS = BO->getOperand(0);
				617	Value *RHS = BO->getOperand(1);
				618	if (LHS == RHS) continue;
				619	}
				620	</pre>
				621	</div>
				622
				623	<p>This has all the benefits of using early exits from functions: it reduces
				624	nesting of the loop, it makes it easier to describe why the conditions are true,
				625	and it makes it obvious to the reader that there is no "else" coming up that
				626	they have to push context into their brain for. If a loop is large, this can
				627	be a big understandability win.</p>
				628
				629	</div>
				630
Chris Lattner	14abb83	2009-07-28 22:54:04 +0000	[diff] [blame]	631	<!-- _______________________________________________________________________ -->
				632	<div class="doc_subsubsection">
				633	<a name="hl_else_after_return">Don't use "else" after a return</a>
				634	</div>
				635
				636	<div class="doc_text">
				637
				638	<p>For similar reasons above (reduction of indentation and easier reading),
				639	please do not use "else" or "else if" after something that interrupts
				640	control flow like return, break, continue, goto, etc. For example, this is
				641	"bad":</p>
				642
				643	<div class="doc_code">
				644	<pre>
				645	case 'J': {
				646	if (Signed) {
				647	Type = Context.getsigjmp_bufType();
				648	if (Type.isNull()) {
				649	Error = ASTContext::GE_Missing_sigjmp_buf;
				650	return QualType();
				651	} else {
				652	break;
				653	}
				654	} else {
				655	Type = Context.getjmp_bufType();
				656	if (Type.isNull()) {
				657	Error = ASTContext::GE_Missing_jmp_buf;
				658	return QualType();
				659	} else {
				660	break;
				661	}
				662	}
				663	}
				664	}
				665	</pre>
				666	</div>
				667
				668	<p>It is better to write this something like:</p>
				669
				670	<div class="doc_code">
				671	<pre>
				672	case 'J':
				673	if (Signed) {
				674	Type = Context.getsigjmp_bufType();
				675	if (Type.isNull()) {
				676	Error = ASTContext::GE_Missing_sigjmp_buf;
				677	return QualType();
				678	}
				679	} else {
				680	Type = Context.getjmp_bufType();
				681	if (Type.isNull()) {
				682	Error = ASTContext::GE_Missing_jmp_buf;
				683	return QualType();
				684	}
				685	}
				686	break;
				687	</pre>
				688	</div>
				689
				690	<p>Or better yet (in this case), as:</p>
				691
				692	<div class="doc_code">
				693	<pre>
				694	case 'J':
				695	if (Signed)
				696	Type = Context.getsigjmp_bufType();
				697	else
				698	Type = Context.getjmp_bufType();
				699
				700	if (Type.isNull()) {
				701	Error = Signed ? ASTContext::GE_Missing_sigjmp_buf :
				702	ASTContext::GE_Missing_jmp_buf;
				703	return QualType();
				704	}
				705	break;
				706	</pre>
				707	</div>
				708
				709	<p>The idea is to reduce indentation and the amount of code you have to keep
				710	track of when reading the code.</p>
				711
				712	</div>
Chris Lattner	6720d7e	2009-07-22 05:40:54 +0000	[diff] [blame]	713
				714	<!-- _______________________________________________________________________ -->
				715	<div class="doc_subsubsection">
				716	<a name="hl_predicateloops">Turn Predicate Loops into Predicate Functions</a>
				717	</div>
				718
				719	<div class="doc_text">
				720
Chris Lattner	d63f1237	2009-07-22 16:54:14 +0000	[diff] [blame]	721	<p>It is very common to write small loops that just compute a boolean
Chris Lattner	6720d7e	2009-07-22 05:40:54 +0000	[diff] [blame]	722	value. There are a number of ways that people commonly write these, but an
				723	example of this sort of thing is:</p>
				724
				725	<div class="doc_code">
				726	<pre>
				727	<b>bool FoundFoo = false;</b>
				728	for (unsigned i = 0, e = BarList.size(); i != e; ++i)
				729	if (BarList[i]->isFoo()) {
				730	<b>FoundFoo = true;</b>
				731	break;
				732	}
				733
				734	<b>if (FoundFoo) {</b>
				735	...
				736	}
				737	</pre>
				738	</div>
				739
				740	<p>This sort of code is awkward to write, and is almost always a bad sign.
				741	Instead of this sort of loop, we strongly prefer to use a predicate function
				742	(which may be <a href="#micro_anonns">static</a>) that uses
Chris Lattner	d63f1237	2009-07-22 16:54:14 +0000	[diff] [blame]	743	<a href="#hl_earlyexit">early exits</a> to compute the predicate. We prefer
				744	the code to be structured like this:
Chris Lattner	6720d7e	2009-07-22 05:40:54 +0000	[diff] [blame]	745	</p>
				746
				747
				748	<div class="doc_code">
				749	<pre>
				750	/// ListContainsFoo - Return true if the specified list has an element that is
				751	/// a foo.
				752	static bool ListContainsFoo(const std::vector<Bar*> &List) {
				753	for (unsigned i = 0, e = List.size(); i != e; ++i)
				754	if (List[i]->isFoo())
				755	return true;
				756	return false;
				757	}
				758	...
				759
				760	<b>if (ListContainsFoo(BarList)) {</b>
				761	...
				762	}
				763	</pre>
				764	</div>
				765
				766	<p>There are many reasons for doing this: it reduces indentation and factors out
				767	code which can often be shared by other code that checks for the same predicate.
				768	More importantly, it <em>forces you to pick a name</em> for the function, and
				769	forces you to write a comment for it. In this silly example, this doesn't add
				770	much value. However, if the condition is complex, this can make it a lot easier
				771	for the reader to understand the code that queries for this predicate. Instead
Chris Lattner	611a188	2009-07-22 16:30:39 +0000	[diff] [blame]	772	of being faced with the in-line details of how we check to see if the BarList
Chris Lattner	6720d7e	2009-07-22 05:40:54 +0000	[diff] [blame]	773	contains a foo, we can trust the function name and continue reading with better
				774	locality.</p>
Bill Wendling	2b52dc1	2006-12-09 01:20:34 +0000	[diff] [blame]	775
				776	</div>
				777
				778
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	779	<!-- ======================================================================= -->
Chris Lattner	1dab192	2004-05-23 21:05:07 +0000	[diff] [blame]	780	<div class="doc_subsection">
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	781	<a name="micro">The Low Level Issues</a>
				782	</div>
Chris Lattner	6720d7e	2009-07-22 05:40:54 +0000	[diff] [blame]	783	<!-- ======================================================================= -->
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	784
				785
				786	<!-- _______________________________________________________________________ -->
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	787	<div class="doc_subsubsection">
Chris Lattner	85ea83e	2006-07-27 04:24:14 +0000	[diff] [blame]	788	<a name="ll_assert">Assert Liberally</a>
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	789	</div>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	790
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	791	<div class="doc_text">
				792
				793	<p>Use the "<tt>assert</tt>" function to its fullest. Check all of your
Benjamin Kramer	0f42038	2009-10-12 14:46:08 +0000	[diff] [blame]	794	preconditions and assumptions, you never know when a bug (not necessarily even
Misha Brukman	c3e7893	2003-07-28 21:57:18 +0000	[diff] [blame]	795	yours) might be caught early by an assertion, which reduces debugging time
				796	dramatically. The "<tt><cassert></tt>" header file is probably already
				797	included by the header files you are using, so it doesn't cost anything to use
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	798	it.</p>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	799
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	800	<p>To further assist with debugging, make sure to put some kind of error message
				801	in the assertion statement (which is printed if the assertion is tripped). This
Misha Brukman	c3e7893	2003-07-28 21:57:18 +0000	[diff] [blame]	802	helps the poor debugging make sense of why an assertion is being made and
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	803	enforced, and hopefully what to do about it. Here is one complete example:</p>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	804
Misha Brukman	b224615	2004-07-28 22:31:54 +0000	[diff] [blame]	805	<div class="doc_code">
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	806	<pre>
Misha Brukman	b224615	2004-07-28 22:31:54 +0000	[diff] [blame]	807	inline Value *getOperand(unsigned i) {
				808	assert(i < Operands.size() && "getOperand() out of range!");
				809	return Operands[i];
				810	}
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	811	</pre>
Misha Brukman	b224615	2004-07-28 22:31:54 +0000	[diff] [blame]	812	</div>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	813
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	814	<p>Here are some examples:</p>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	815
Misha Brukman	b224615	2004-07-28 22:31:54 +0000	[diff] [blame]	816	<div class="doc_code">
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	817	<pre>
Misha Brukman	b224615	2004-07-28 22:31:54 +0000	[diff] [blame]	818	assert(Ty->isPointerType() && "Can't allocate a non pointer type!");
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	819
Misha Brukman	b224615	2004-07-28 22:31:54 +0000	[diff] [blame]	820	assert((Opcode == Shl \|\| Opcode == Shr) && "ShiftInst Opcode invalid!");
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	821
Misha Brukman	b224615	2004-07-28 22:31:54 +0000	[diff] [blame]	822	assert(idx < getNumSuccessors() && "Successor # out of range!");
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	823
Misha Brukman	b224615	2004-07-28 22:31:54 +0000	[diff] [blame]	824	assert(V1.getType() == V2.getType() && "Constant types must be identical!");
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	825
Misha Brukman	b224615	2004-07-28 22:31:54 +0000	[diff] [blame]	826	assert(isa<PHINode>(Succ->front()) && "Only works on PHId BBs!");
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	827	</pre>
Misha Brukman	b224615	2004-07-28 22:31:54 +0000	[diff] [blame]	828	</div>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	829
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	830	<p>You get the idea...</p>
				831
Nick Lewycky	fb75d42	2008-05-31 23:54:55 +0000	[diff] [blame]	832	<p>Please be aware when adding assert statements that not all compilers are aware of
				833	the semantics of the assert. In some places, asserts are used to indicate a piece of
				834	code that should not be reached. These are typically of the form:</p>
				835
				836	<div class="doc_code">
				837	<pre>
Dan Gohman	8ef4498	2008-11-24 17:18:39 +0000	[diff] [blame]	838	assert(0 && "Some helpful error message");
Nick Lewycky	fb75d42	2008-05-31 23:54:55 +0000	[diff] [blame]	839	</pre>
				840	</div>
				841
				842	<p>When used in a function that returns a value, they should be followed with a return
				843	statement and a comment indicating that this line is never reached. This will prevent
				844	a compiler which is unable to deduce that the assert statement never returns from
				845	generating a warning.</p>
				846
				847	<div class="doc_code">
				848	<pre>
Dan Gohman	8ef4498	2008-11-24 17:18:39 +0000	[diff] [blame]	849	assert(0 && "Some helpful error message");
Nick Lewycky	fb75d42	2008-05-31 23:54:55 +0000	[diff] [blame]	850	// Not reached
				851	return 0;
				852	</pre>
				853	</div>
				854
Chris Lattner	8a7f4da	2010-11-12 00:19:41 +0000	[diff] [blame]	855	<p>Another issue is that values used only by assertions will produce an "unused
				856	value" warning when assertions are disabled. For example, this code will warn:
				857	</p>
				858
				859	<div class="doc_code">
				860	<pre>
				861	unsigned Size = V.size();
				862	assert(Size > 42 && "Vector smaller than it should be");
				863
				864	bool NewToSet = Myset.insert(Value);
				865	assert(NewToSet && "The value shouldn't be in the set yet");
				866	</pre>
				867	</div>
				868
				869	<p>These are two interesting different cases: in the first case, the call to
				870	V.size() is only useful for the assert, and we don't want it executed when
				871	assertions are disabled. Code like this should move the call into the assert
				872	itself. In the second case, the side effects of the call must happen whether
				873	the assert is enabled or not. In this case, the value should be cast to void
				874	to disable the warning. To be specific, it is preferred to write the code
				875	like this:</p>
				876
				877	<div class="doc_code">
				878	<pre>
				879	assert(V.size() > 42 && "Vector smaller than it should be");
				880
				881	bool NewToSet = Myset.insert(Value); (void)NewToSet;
				882	assert(NewToSet && "The value shouldn't be in the set yet");
				883	</pre>
				884	</div>
				885
				886
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	887	</div>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	888
Chris Lattner	01e81e6	2006-01-01 21:59:22 +0000	[diff] [blame]	889	<!-- _______________________________________________________________________ -->
				890	<div class="doc_subsubsection">
Bill Wendling	b062948	2006-12-09 01:35:43 +0000	[diff] [blame]	891	<a name="ll_ns_std">Do not use '<tt>using namespace std</tt>'</a>
Chris Lattner	01e81e6	2006-01-01 21:59:22 +0000	[diff] [blame]	892	</div>
				893
				894	<div class="doc_text">
				895	<p>In LLVM, we prefer to explicitly prefix all identifiers from the standard
Bill Wendling	b062948	2006-12-09 01:35:43 +0000	[diff] [blame]	896	namespace with an "<tt>std::</tt>" prefix, rather than rely on
				897	"<tt>using namespace std;</tt>".</p>
Chris Lattner	01e81e6	2006-01-01 21:59:22 +0000	[diff] [blame]	898
Bill Wendling	b062948	2006-12-09 01:35:43 +0000	[diff] [blame]	899	<p> In header files, adding a '<tt>using namespace XXX</tt>' directive pollutes
Misha Brukman	586a15c	2008-12-11 19:37:04 +0000	[diff] [blame]	900	the namespace of any source file that <tt>#include</tt>s the header. This is
				901	clearly a bad thing.</p>
Chris Lattner	01e81e6	2006-01-01 21:59:22 +0000	[diff] [blame]	902
Bill Wendling	b062948	2006-12-09 01:35:43 +0000	[diff] [blame]	903	<p>In implementation files (e.g. .cpp files), the rule is more of a stylistic
				904	rule, but is still important. Basically, using explicit namespace prefixes
				905	makes the code <b>clearer</b>, because it is immediately obvious what facilities
				906	are being used and where they are coming from, and <b>more portable</b>, because
				907	namespace clashes cannot occur between LLVM code and other namespaces. The
				908	portability rule is important because different standard library implementations
				909	expose different symbols (potentially ones they shouldn't), and future revisions
				910	to the C++ standard will add more symbols to the <tt>std</tt> namespace. As
				911	such, we never use '<tt>using namespace std;</tt>' in LLVM.</p>
Chris Lattner	01e81e6	2006-01-01 21:59:22 +0000	[diff] [blame]	912
Bill Wendling	b062948	2006-12-09 01:35:43 +0000	[diff] [blame]	913	<p>The exception to the general rule (i.e. it's not an exception for
				914	the <tt>std</tt> namespace) is for implementation files. For example, all of
				915	the code in the LLVM project implements code that lives in the 'llvm' namespace.
				916	As such, it is ok, and actually clearer, for the .cpp files to have a '<tt>using
				917	namespace llvm</tt>' directive at their top, after the <tt>#include</tt>s. The
				918	general form of this rule is that any .cpp file that implements code in any
				919	namespace may use that namespace (and its parents'), but should not use any
				920	others.</p>
Chris Lattner	01e81e6	2006-01-01 21:59:22 +0000	[diff] [blame]	921
				922	</div>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	923
				924	<!-- _______________________________________________________________________ -->
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	925	<div class="doc_subsubsection">
Bill Wendling	b062948	2006-12-09 01:35:43 +0000	[diff] [blame]	926	<a name="ll_virtual_anch">Provide a virtual method anchor for classes
				927	in headers</a>
Chris Lattner	85ea83e	2006-07-27 04:24:14 +0000	[diff] [blame]	928	</div>
				929
				930	<div class="doc_text">
				931
				932	<p>If a class is defined in a header file and has a v-table (either it has
				933	virtual methods or it derives from classes with virtual methods), it must
				934	always have at least one out-of-line virtual method in the class. Without
Misha Brukman	586a15c	2008-12-11 19:37:04 +0000	[diff] [blame]	935	this, the compiler will copy the vtable and RTTI into every <tt>.o</tt> file
				936	that <tt>#include</tt>s the header, bloating <tt>.o</tt> file sizes and
				937	increasing link times.</p>
Chris Lattner	85ea83e	2006-07-27 04:24:14 +0000	[diff] [blame]	938
				939	</div>
				940
Chris Lattner	56c1094	2009-06-30 06:13:23 +0000	[diff] [blame]	941	<!-- _______________________________________________________________________ -->
				942	<div class="doc_subsubsection">
				943	<a name="ll_end">Don't evaluate end() every time through a loop</a>
				944	</div>
				945
				946	<div class="doc_text">
				947
				948	<p>Because C++ doesn't have a standard "foreach" loop (though it can be emulated
				949	with macros and may be coming in C++'0x) we end up writing a lot of loops that
				950	manually iterate from begin to end on a variety of containers or through other
				951	data structures. One common mistake is to write a loop in this style:</p>
				952
				953	<div class="doc_code">
				954	<pre>
				955	BasicBlock *BB = ...
				956	for (BasicBlock::iterator I = BB->begin(); I != <b>BB->end()</b>; ++I)
				957	... use I ...
				958	</pre>
				959	</div>
				960
				961	<p>The problem with this construct is that it evaluates "<tt>BB->end()</tt>"
				962	every time through the loop. Instead of writing the loop like this, we strongly
				963	prefer loops to be written so that they evaluate it once before the loop starts.
				964	A convenient way to do this is like so:</p>
				965
				966	<div class="doc_code">
				967	<pre>
				968	BasicBlock *BB = ...
				969	for (BasicBlock::iterator I = BB->begin(), E = <b>BB->end()</b>; I != E; ++I)
				970	... use I ...
				971	</pre>
				972	</div>
				973
				974	<p>The observant may quickly point out that these two loops may have different
				975	semantics: if the container (a basic block in this case) is being mutated, then
				976	"<tt>BB->end()</tt>" may change its value every time through the loop and the
				977	second loop may not in fact be correct. If you actually do depend on this
Chris Lattner	6ad089c	2009-06-30 06:20:03 +0000	[diff] [blame]	978	behavior, please write the loop in the first form and add a comment indicating
Chris Lattner	56c1094	2009-06-30 06:13:23 +0000	[diff] [blame]	979	that you did it intentionally.</p>
				980
				981	<p>Why do we prefer the second form (when correct)? Writing the loop in the
				982	first form has two problems: First it may be less efficient than evaluating it
				983	at the start of the loop. In this case, the cost is probably minor: a few extra
				984	loads every time through the loop. However, if the base expression is more
				985	complex, then the cost can rise quickly. I've seen loops where the end
				986	expression was actually something like: "<tt>SomeMap[x]->end()</tt>" and map
Chris Lattner	67f2a4ac	2009-06-30 06:27:54 +0000	[diff] [blame]	987	lookups really aren't cheap. By writing it in the second form consistently, you
Chris Lattner	56c1094	2009-06-30 06:13:23 +0000	[diff] [blame]	988	eliminate the issue entirely and don't even have to think about it.</p>
				989
Chris Lattner	6ad089c	2009-06-30 06:20:03 +0000	[diff] [blame]	990	<p>The second (even bigger) issue is that writing the loop in the first form
Chris Lattner	56c1094	2009-06-30 06:13:23 +0000	[diff] [blame]	991	hints to the reader that the loop is mutating the container (a fact that a
				992	comment would handily confirm!). If you write the loop in the second form, it
				993	is immediately obvious without even looking at the body of the loop that the
				994	container isn't being modified, which makes it easier to read the code and
				995	understand what it does.</p>
				996
				997	<p>While the second form of the loop is a few extra keystrokes, we do strongly
				998	prefer it.</p>
				999
				1000	</div>
				1001
Chris Lattner	85ea83e	2006-07-27 04:24:14 +0000	[diff] [blame]	1002	<!-- _______________________________________________________________________ -->
				1003	<div class="doc_subsubsection">
Chris Lattner	6720d7e	2009-07-22 05:40:54 +0000	[diff] [blame]	1004	<a name="ll_iostream"><tt>#include <iostream></tt> is forbidden</a>
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	1005	</div>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	1006
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	1007	<div class="doc_text">
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	1008
Chris Lattner	6720d7e	2009-07-22 05:40:54 +0000	[diff] [blame]	1009	<p>The use of <tt>#include <iostream></tt> in library files is
				1010	hereby <b><em>forbidden</em></b>. The primary reason for doing this is to
				1011	support clients using LLVM libraries as part of larger systems. In particular,
				1012	we statically link LLVM into some dynamic libraries. Even if LLVM isn't used,
				1013	the static c'tors are run whenever an application start up that uses the dynamic
				1014	library. There are two problems with this:</p>
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	1015
Chris Lattner	6720d7e	2009-07-22 05:40:54 +0000	[diff] [blame]	1016	<ol>
				1017	<li>The time to run the static c'tors impacts startup time of
				1018	applications—a critical time for GUI apps.</li>
				1019	<li>The static c'tors cause the app to pull many extra pages of memory off the
				1020	disk: both the code for the static c'tors in each <tt>.o</tt> file and the
				1021	small amount of data that gets touched. In addition, touched/dirty pages
				1022	put more pressure on the VM system on low-memory machines.</li>
				1023	</ol>
				1024
				1025	<p>Note that using the other stream headers (<tt><sstream></tt> for
Chris Lattner	5db2ff5	2009-08-23 21:53:47 +0000	[diff] [blame]	1026	example) is not problematic in this regard (just <tt><iostream></tt>).
				1027	However, raw_ostream provides various APIs that are better performing for almost
				1028	every use than std::ostream style APIs, so you should just use it for new
				1029	code.</p>
Chris Lattner	6720d7e	2009-07-22 05:40:54 +0000	[diff] [blame]	1030
Chris Lattner	5db2ff5	2009-08-23 21:53:47 +0000	[diff] [blame]	1031	<p><b>New code should always
				1032	use <a href="#ll_raw_ostream"><tt>raw_ostream</tt></a> for writing, or
				1033	the <tt>llvm::MemoryBuffer</tt> API for reading files.</b></p>
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	1034
				1035	</div>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	1036
Chris Lattner	6720d7e	2009-07-22 05:40:54 +0000	[diff] [blame]	1037
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	1038	<!-- _______________________________________________________________________ -->
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	1039	<div class="doc_subsubsection">
Chris Lattner	85ea83e	2006-07-27 04:24:14 +0000	[diff] [blame]	1040	<a name="ll_avoidendl">Avoid <tt>std::endl</tt></a>
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	1041	</div>
Chris Lattner	850d4f6	2002-01-20 19:01:26 +0000	[diff] [blame]	1042
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	1043	<div class="doc_text">
				1044
Chris Lattner	1dab192	2004-05-23 21:05:07 +0000	[diff] [blame]	1045	<p>The <tt>std::endl</tt> modifier, when used with iostreams outputs a newline
				1046	to the output stream specified. In addition to doing this, however, it also
				1047	flushes the output stream. In other words, these are equivalent:</p>
Chris Lattner	850d4f6	2002-01-20 19:01:26 +0000	[diff] [blame]	1048
Misha Brukman	b224615	2004-07-28 22:31:54 +0000	[diff] [blame]	1049	<div class="doc_code">
Chris Lattner	850d4f6	2002-01-20 19:01:26 +0000	[diff] [blame]	1050	<pre>
Misha Brukman	b224615	2004-07-28 22:31:54 +0000	[diff] [blame]	1051	std::cout << std::endl;
				1052	std::cout << '\n' << std::flush;
Chris Lattner	850d4f6	2002-01-20 19:01:26 +0000	[diff] [blame]	1053	</pre>
Misha Brukman	b224615	2004-07-28 22:31:54 +0000	[diff] [blame]	1054	</div>
Chris Lattner	850d4f6	2002-01-20 19:01:26 +0000	[diff] [blame]	1055
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	1056	<p>Most of the time, you probably have no reason to flush the output stream, so
Misha Brukman	b224615	2004-07-28 22:31:54 +0000	[diff] [blame]	1057	it's better to use a literal <tt>'\n'</tt>.</p>
Chris Lattner	850d4f6	2002-01-20 19:01:26 +0000	[diff] [blame]	1058
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	1059	</div>
Chris Lattner	850d4f6	2002-01-20 19:01:26 +0000	[diff] [blame]	1060
Bill Wendling	2b52dc1	2006-12-09 01:20:34 +0000	[diff] [blame]	1061
Daniel Dunbar	fcd54f7	2009-07-24 23:04:51 +0000	[diff] [blame]	1062	<!-- _______________________________________________________________________ -->
				1063	<div class="doc_subsubsection">
				1064	<a name="ll_raw_ostream">Use <tt>raw_ostream</tt></a>
				1065	</div>
				1066
				1067	<div class="doc_text">
				1068
				1069	<p>LLVM includes a lightweight, simple, and efficient stream implementation
				1070	in <tt>llvm/Support/raw_ostream.h</tt> which provides all of the common features
Daniel Dunbar	4bd53c6	2009-07-24 23:54:34 +0000	[diff] [blame]	1071	of <tt>std::ostream</tt>. All new code should use <tt>raw_ostream</tt> instead
				1072	of <tt>ostream</tt>.</p>
Daniel Dunbar	fcd54f7	2009-07-24 23:04:51 +0000	[diff] [blame]	1073
Daniel Dunbar	4bd53c6	2009-07-24 23:54:34 +0000	[diff] [blame]	1074	<p>Unlike <tt>std::ostream</tt>, <tt>raw_ostream</tt> is not a template and can
Daniel Dunbar	fcd54f7	2009-07-24 23:04:51 +0000	[diff] [blame]	1075	be forward declared as <tt>class raw_ostream</tt>. Public headers should
				1076	generally not include the <tt>raw_ostream</tt> header, but use forward
				1077	declarations and constant references to <tt>raw_ostream</tt> instances.</p>
				1078
				1079	</div>
				1080
				1081
Chris Lattner	6720d7e	2009-07-22 05:40:54 +0000	[diff] [blame]	1082	<!-- ======================================================================= -->
				1083	<div class="doc_subsection">
				1084	<a name="nano">Microscopic Details</a>
				1085	</div>
				1086	<!-- ======================================================================= -->
				1087
				1088	<p>This section describes preferred low-level formatting guidelines along with
				1089	reasoning on why we prefer them.</p>
				1090
				1091	<!-- _______________________________________________________________________ -->
				1092	<div class="doc_subsubsection">
				1093	<a name="micro_spaceparen">Spaces Before Parentheses</a>
				1094	</div>
				1095
				1096	<div class="doc_text">
				1097
				1098	<p>We prefer to put a space before a parentheses only in control flow
				1099	statements, but not in normal function call expressions and function-like
				1100	macros. For example, this is good:</p>
				1101
				1102	<div class="doc_code">
				1103	<pre>
				1104	<b>if (</b>x) ...
				1105	<b>for (</b>i = 0; i != 100; ++i) ...
				1106	<b>while (</b>llvm_rocks) ...
				1107
				1108	<b>somefunc(</b>42);
				1109	<b><a href="#ll_assert">assert</a>(</b>3 != 4 && "laws of math are failing me");
				1110
				1111	a = <b>foo(</b>42, 92) + <b>bar(</b>x);
				1112	</pre>
				1113	</div>
				1114
				1115	<p>... and this is bad:</p>
				1116
				1117	<div class="doc_code">
				1118	<pre>
				1119	<b>if(</b>x) ...
				1120	<b>for(</b>i = 0; i != 100; ++i) ...
				1121	<b>while(</b>llvm_rocks) ...
				1122
				1123	<b>somefunc (</b>42);
				1124	<b><a href="#ll_assert">assert</a> (</b>3 != 4 && "laws of math are failing me");
				1125
				1126	a = <b>foo (</b>42, 92) + <b>bar (</b>x);
				1127	</pre>
				1128	</div>
				1129
				1130	<p>The reason for doing this is not completely arbitrary. This style makes
				1131	control flow operators stand out more, and makes expressions flow better. The
				1132	function call operator binds very tightly as a postfix operator. Putting
				1133	a space after a function name (as in the last example) makes it appear that
				1134	the code might bind the arguments of the left-hand-side of a binary operator
				1135	with the argument list of a function and the name of the right side. More
				1136	specifically, it is easy to misread the "a" example as:</p>
				1137
				1138	<div class="doc_code">
				1139	<pre>
				1140	a = foo <b>(</b>(42, 92) + bar<b>)</b> (x);
				1141	</pre>
				1142	</div>
				1143
				1144	<p>... when skimming through the code. By avoiding a space in a function, we
				1145	avoid this misinterpretation.</p>
				1146
				1147	</div>
				1148
				1149	<!-- _______________________________________________________________________ -->
				1150	<div class="doc_subsubsection">
				1151	<a name="micro_preincrement">Prefer Preincrement</a>
				1152	</div>
				1153
				1154	<div class="doc_text">
				1155
				1156	<p>Hard fast rule: Preincrement (<tt>++X</tt>) may be no slower than
				1157	postincrement (<tt>X++</tt>) and could very well be a lot faster than it. Use
				1158	preincrementation whenever possible.</p>
				1159
				1160	<p>The semantics of postincrement include making a copy of the value being
				1161	incremented, returning it, and then preincrementing the "work value". For
				1162	primitive types, this isn't a big deal... but for iterators, it can be a huge
				1163	issue (for example, some iterators contains stack and set objects in them...
				1164	copying an iterator could invoke the copy ctor's of these as well). In general,
				1165	get in the habit of always using preincrement, and you won't have a problem.</p>
				1166
				1167	</div>
				1168
				1169	<!-- _______________________________________________________________________ -->
				1170	<div class="doc_subsubsection">
				1171	<a name="micro_namespaceindent">Namespace Indentation</a>
				1172	</div>
				1173
				1174	<div class="doc_text">
				1175
				1176	<p>
				1177	In general, we strive to reduce indentation where ever possible. This is useful
				1178	because we want code to <a href="#scf_codewidth">fit into 80 columns</a> without
				1179	wrapping horribly, but also because it makes it easier to understand the code.
				1180	Namespaces are a funny thing: they are often large, and we often desire to put
				1181	lots of stuff into them (so they can be large). Other times they are tiny,
				1182	because they just hold an enum or something similar. In order to balance this,
				1183	we use different approaches for small versus large namespaces.
				1184	</p>
				1185
				1186	<p>
				1187	If a namespace definition is small and <em>easily</em> fits on a screen (say,
				1188	less than 35 lines of code), then you should indent its body. Here's an
				1189	example:
				1190	</p>
				1191
				1192	<div class="doc_code">
				1193	<pre>
Chris Lattner	d63f1237	2009-07-22 16:54:14 +0000	[diff] [blame]	1194	namespace llvm {
				1195	namespace X86 {
				1196	/// RelocationType - An enum for the x86 relocation codes. Note that
				1197	/// the terminology here doesn't follow x86 convention - word means
				1198	/// 32-bit and dword means 64-bit.
				1199	enum RelocationType {
				1200	/// reloc_pcrel_word - PC relative relocation, add the relocated value to
				1201	/// the value already in memory, after we adjust it for where the PC is.
				1202	reloc_pcrel_word = 0,
				1203
				1204	/// reloc_picrel_word - PIC base relative relocation, add the relocated
				1205	/// value to the value already in memory, after we adjust it for where the
				1206	/// PIC base is.
				1207	reloc_picrel_word = 1,
				1208
				1209	/// reloc_absolute_word, reloc_absolute_dword - Absolute relocation, just
				1210	/// add the relocated value to the value already in memory.
				1211	reloc_absolute_word = 2,
				1212	reloc_absolute_dword = 3
				1213	};
				1214	}
Chris Lattner	6720d7e	2009-07-22 05:40:54 +0000	[diff] [blame]	1215	}
				1216	</pre>
				1217	</div>
				1218
				1219	<p>Since the body is small, indenting adds value because it makes it very clear
				1220	where the namespace starts and ends, and it is easy to take the whole thing in
				1221	in one "gulp" when reading the code. If the blob of code in the namespace is
				1222	larger (as it typically is in a header in the llvm or clang namespaces), do not
				1223	indent the code, and add a comment indicating what namespace is being closed.
				1224	For example:</p>
				1225
				1226	<div class="doc_code">
				1227	<pre>
				1228	namespace llvm {
				1229	namespace knowledge {
				1230
				1231	/// Grokable - This class represents things that Smith can have an intimate
				1232	/// understanding of and contains the data associated with it.
				1233	class Grokable {
				1234	...
				1235	public:
				1236	explicit Grokable() { ... }
				1237	virtual ~Grokable() = 0;
				1238
				1239	...
				1240
				1241	};
				1242
				1243	} // end namespace knowledge
				1244	} // end namespace llvm
				1245	</pre>
				1246	</div>
				1247
				1248	<p>Because the class is large, we don't expect that the reader can easily
				1249	understand the entire concept in a glance, and the end of the file (where the
				1250	namespaces end) may be a long ways away from the place they open. As such,
				1251	indenting the contents of the namespace doesn't add any value, and detracts from
				1252	the readability of the class. In these cases it is best to <em>not</em> indent
				1253	the contents of the namespace.</p>
				1254
				1255	</div>
				1256
				1257	<!-- _______________________________________________________________________ -->
				1258	<div class="doc_subsubsection">
				1259	<a name="micro_anonns">Anonymous Namespaces</a>
				1260	</div>
				1261
				1262	<div class="doc_text">
				1263
Chris Lattner	d63f1237	2009-07-22 16:54:14 +0000	[diff] [blame]	1264	<p>After talking about namespaces in general, you may be wondering about
				1265	anonymous namespaces in particular.
Chris Lattner	6720d7e	2009-07-22 05:40:54 +0000	[diff] [blame]	1266	Anonymous namespaces are a great language feature that tells the C++ compiler
				1267	that the contents of the namespace are only visible within the current
				1268	translation unit, allowing more aggressive optimization and eliminating the
				1269	possibility of symbol name collisions. Anonymous namespaces are to C++ as
				1270	"static" is to C functions and global variables. While "static" is available
				1271	in C++, anonymous namespaces are more general: they can make entire classes
				1272	private to a file.</p>
				1273
				1274	<p>The problem with anonymous namespaces is that they naturally want to
				1275	encourage indentation of their body, and they reduce locality of reference: if
				1276	you see a random function definition in a C++ file, it is easy to see if it is
				1277	marked static, but seeing if it is in an anonymous namespace requires scanning
				1278	a big chunk of the file.</p>
				1279
				1280	<p>Because of this, we have a simple guideline: make anonymous namespaces as
				1281	small as possible, and only use them for class declarations. For example, this
				1282	is good:</p>
				1283
				1284	<div class="doc_code">
				1285	<pre>
				1286	<b>namespace {</b>
				1287	class StringSort {
				1288	...
				1289	public:
				1290	StringSort(...)
				1291	bool operator<(const char *RHS) const;
				1292	};
				1293	<b>} // end anonymous namespace</b>
				1294
				1295	static void Helper() {
				1296	...
				1297	}
				1298
				1299	bool StringSort::operator<(const char *RHS) const {
				1300	...
				1301	}
				1302
				1303	</pre>
				1304	</div>
				1305
				1306	<p>This is bad:</p>
				1307
				1308
				1309	<div class="doc_code">
				1310	<pre>
				1311	<b>namespace {</b>
				1312	class StringSort {
				1313	...
				1314	public:
				1315	StringSort(...)
				1316	bool operator<(const char *RHS) const;
				1317	};
				1318
				1319	void Helper() {
				1320	...
				1321	}
				1322
				1323	bool StringSort::operator<(const char *RHS) const {
				1324	...
				1325	}
				1326
				1327	<b>} // end anonymous namespace</b>
				1328
				1329	</pre>
				1330	</div>
				1331
				1332
				1333	<p>This is bad specifically because if you're looking at "Helper" in the middle
				1334	of a large C++ file, that you have no immediate way to tell if it is local to
				1335	the file. When it is marked static explicitly, this is immediately obvious.
				1336	Also, there is no reason to enclose the definition of "operator<" in the
Chris Lattner	d63f1237	2009-07-22 16:54:14 +0000	[diff] [blame]	1337	namespace just because it was declared there.
Chris Lattner	6720d7e	2009-07-22 05:40:54 +0000	[diff] [blame]	1338	</p>
				1339
				1340	</div>
				1341
				1342
				1343
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	1344	<!-- *********************************************************************** -->
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	1345	<div class="doc_section">
				1346	<a name="seealso">See Also</a>
				1347	</div>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	1348	<!-- *********************************************************************** -->
				1349
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	1350	<div class="doc_text">
				1351
				1352	<p>A lot of these comments and recommendations have been culled for other
				1353	sources. Two particularly important books for our work are:</p>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	1354
				1355	<ol>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	1356
Chris Lattner	f9f08bf	2007-11-09 21:49:08 +0000	[diff] [blame]	1357	<li><a href="http://www.amazon.com/Effective-Specific-Addison-Wesley-Professional-Computing/dp/0321334876">Effective
				1358	C++</a> by Scott Meyers. Also
Chris Lattner	1dab192	2004-05-23 21:05:07 +0000	[diff] [blame]	1359	interesting and useful are "More Effective C++" and "Effective STL" by the same
				1360	author.</li>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	1361
Chris Lattner	f9f08bf	2007-11-09 21:49:08 +0000	[diff] [blame]	1362	<li>Large-Scale C++ Software Design by John Lakos</li>
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	1363
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	1364	</ol>
				1365
				1366	<p>If you get some free time, and you haven't read them: do so, you might learn
Misha Brukman	6d1686c	2004-12-04 00:32:12 +0000	[diff] [blame]	1367	something.</p>
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	1368
				1369	</div>
				1370
Chris Lattner	ac457c4	2001-07-09 03:27:08 +0000	[diff] [blame]	1371	<!-- *********************************************************************** -->
				1372
				1373	<hr>
Misha Brukman	d4f290a	2004-05-12 18:37:22 +0000	[diff] [blame]	1374	<address>
				1375	<a href="http://jigsaw.w3.org/css-validator/check/referer"><img
Misha Brukman	86242e1	2008-12-11 17:34:48 +0000	[diff] [blame]	1376	src="http://jigsaw.w3.org/css-validator/images/vcss-blue" alt="Valid CSS"></a>
Misha Brukman	d4f290a	2004-05-12 18:37:22 +0000	[diff] [blame]	1377	<a href="http://validator.w3.org/check/referer"><img
Misha Brukman	86242e1	2008-12-11 17:34:48 +0000	[diff] [blame]	1378	src="http://www.w3.org/Icons/valid-html401-blue" alt="Valid HTML 4.01"></a>
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	1379
Misha Brukman	d4f290a	2004-05-12 18:37:22 +0000	[diff] [blame]	1380	<a href="mailto:sabre@nondot.org">Chris Lattner</a><br>
Reid Spencer	ca05854	2006-03-14 05:39:39 +0000	[diff] [blame]	1381	<a href="http://llvm.org">LLVM Compiler Infrastructure</a><br>
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	1382	Last modified: $Date$
Misha Brukman	d4f290a	2004-05-12 18:37:22 +0000	[diff] [blame]	1383	</address>
Misha Brukman	f196dbb	2003-10-24 17:57:33 +0000	[diff] [blame]	1384
				1385	</body>
				1386	</html>