Gitiles
Code Review Sign In
gerrit-public.fairphone.software / fp2-dev / platform / external / llvm / 5c6e4df71323e3c3b6eb23bfd4b1d908ab5b149a / . / docs / CodingStandards.html
blob: 7b05573703bca3531ae3a8432b6c9879e82fa4a9 [file] [log] [blame]
Misha Brukman75f258e2003-10-24 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>
NAKAMURA Takumi5c6e4df2011-10-31 11:21:59 +0000[diff] [blame^]5 <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]6 <link rel="stylesheet" href="llvm.css" type="text/css">
Misha Brukmanc1449e62008-12-10 23:07:02 +0000[diff] [blame]7 <title>LLVM Coding Standards</title>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]8</head>
9<body>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]10
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]11<h1>
Misha Brukmanc1449e62008-12-10 23:07:02 +0000[diff] [blame]12 LLVM Coding Standards
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]13</h1>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]14
15<ol>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]16 <li><a href="#introduction">Introduction</a></li>
Chris Lattner7ae36bb2001-07-23 20:40:41 +0000[diff] [blame]17 <li><a href="#mechanicalissues">Mechanical Source Issues</a>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]18 <ol>
19 <li><a href="#sourceformating">Source Code Formatting</a>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]20 <ol>
21 <li><a href="#scf_commenting">Commenting</a></li>
22 <li><a href="#scf_commentformat">Comment Formatting</a></li>
Misha Brukman91d64af2004-10-26 16:18:43 +0000[diff] [blame]23 <li><a href="#scf_includes"><tt>#include</tt> Style</a></li>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]24 <li><a href="#scf_codewidth">Source Code Width</a></li>
25 <li><a href="#scf_spacestabs">Use Spaces Instead of Tabs</a></li>
26 <li><a href="#scf_indentation">Indent Code Consistently</a></li>
27 </ol></li>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]28 <li><a href="#compilerissues">Compiler Issues</a>
29 <ol>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]30 <li><a href="#ci_warningerrors">Treat Compiler Warnings Like
31 Errors</a></li>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]32 <li><a href="#ci_portable_code">Write Portable Code</a></li>
Chris Lattner558f5d22010-11-30 19:20:40 +0000[diff] [blame]33 <li><a href="#ci_rtti_exceptions">Do not use RTTI or Exceptions</a></li>
Zhanyong Wancafe0b42010-11-23 05:03:07 +0000[diff] [blame]34 <li><a href="#ci_class_struct">Use of <tt>class</tt>/<tt>struct</tt> Keywords</a></li>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]35 </ol></li>
36 </ol></li>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]37 <li><a href="#styleissues">Style Issues</a>
38 <ol>
Chris Lattner684cf812010-11-17 17:14:55 +0000[diff] [blame]39 <li><a href="#macro">The High-Level Issues</a>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]40 <ol>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]41 <li><a href="#hl_module">A Public Header File <b>is</b> a
42 Module</a></li>
Zhanyong Wancafe0b42010-11-23 05:03:07 +0000[diff] [blame]43 <li><a href="#hl_dontinclude"><tt>#include</tt> as Little as Possible</a></li>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]44 <li><a href="#hl_privateheaders">Keep "internal" Headers
45 Private</a></li>
Zhanyong Wancafe0b42010-11-23 05:03:07 +0000[diff] [blame]46 <li><a href="#hl_earlyexit">Use Early Exits and <tt>continue</tt> to Simplify
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]47 Code</a></li>
Zhanyong Wancafe0b42010-11-23 05:03:07 +0000[diff] [blame]48 <li><a href="#hl_else_after_return">Don't use <tt>else</tt> after a
49 <tt>return</tt></a></li>
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]50 <li><a href="#hl_predicateloops">Turn Predicate Loops into Predicate
51 Functions</a></li>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]52 </ol></li>
Chris Lattner684cf812010-11-17 17:14:55 +0000[diff] [blame]53 <li><a href="#micro">The Low-Level Issues</a>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]54 <ol>
Zhanyong Wan7fcd4dc2010-12-02 05:10:07 +0000[diff] [blame]55 <li><a href="#ll_naming">Name Types, Functions, Variables, and Enumerators Properly</a></li>
Chris Lattner6883a882006-07-27 04:24:14 +0000[diff] [blame]56 <li><a href="#ll_assert">Assert Liberally</a></li>
Zhanyong Wancafe0b42010-11-23 05:03:07 +0000[diff] [blame]57 <li><a href="#ll_ns_std">Do not use '<tt>using namespace std</tt>'</a></li>
Bill Wendling7f564c02006-12-09 01:20:34 +0000[diff] [blame]58 <li><a href="#ll_virtual_anch">Provide a virtual method anchor for
Bill Wendling4ad6d612006-12-09 01:35:43 +0000[diff] [blame]59 classes in headers</a></li>
Zhanyong Wancafe0b42010-11-23 05:03:07 +0000[diff] [blame]60 <li><a href="#ll_end">Don't evaluate <tt>end()</tt> every time through a
Chris Lattner0af39ea2009-06-30 06:13:23 +0000[diff] [blame]61 loop</a></li>
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]62 <li><a href="#ll_iostream"><tt>#include &lt;iostream&gt;</tt> is
63 <em>forbidden</em></a></li>
Chris Lattner78ceb3a2010-12-10 00:54:03 +0000[diff] [blame]64 <li><a href="#ll_raw_ostream">Use <tt>raw_ostream</tt></a></li>
Chris Lattner684cf812010-11-17 17:14:55 +0000[diff] [blame]65 <li><a href="#ll_avoidendl">Avoid <tt>std::endl</tt></a></li>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]66 </ol></li>
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]67
68 <li><a href="#nano">Microscopic Details</a>
69 <ol>
70 <li><a href="#micro_spaceparen">Spaces Before Parentheses</a></li>
71 <li><a href="#micro_preincrement">Prefer Preincrement</a></li>
72 <li><a href="#micro_namespaceindent">Namespace Indentation</a></li>
73 <li><a href="#micro_anonns">Anonymous Namespaces</a></li>
74 </ol></li>
75
76
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]77 </ol></li>
78 <li><a href="#seealso">See Also</a></li>
79</ol>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]80
Chris Lattner999cf092004-05-23 21:05:07 +0000[diff] [blame]81<div class="doc_author">
Chris Lattner219bd292009-07-22 05:43:01 +0000[diff] [blame]82 <p>Written by <a href="mailto:sabre@nondot.org">Chris Lattner</a></p>
Chris Lattner999cf092004-05-23 21:05:07 +0000[diff] [blame]83</div>
84
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]85
86<!-- *********************************************************************** -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]87<h2>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]88 <a name="introduction">Introduction</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]89</h2>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]90<!-- *********************************************************************** -->
91
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]92<div>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]93
94<p>This document attempts to describe a few coding standards that are being used
95in the LLVM source tree. Although no coding standards should be regarded as
Misha Brukman1b36d672003-10-06 19:26:00 +0000[diff] [blame]96absolute requirements to be followed in all instances, coding standards can be
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]97useful.</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]98
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]99<p>This document intentionally does not prescribe fixed standards for religious
Misha Brukman1b36d672003-10-06 19:26:00 +0000[diff] [blame]100issues such as brace placement and space usage. For issues like this, follow
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]101the golden rule:</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]102
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]103<blockquote>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]104
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]105<p><b><a name="goldenrule">If you are adding a significant body of source to a
106project, feel free to use whatever style you are most comfortable with. If you
107are extending, enhancing, or bug fixing already implemented code, use the style
108that is already being used so that the source is uniform and easy to
109follow.</a></b></p>
110
111</blockquote>
112
113<p>The ultimate goal of these guidelines is the increase readability and
Misha Brukman1b36d672003-10-06 19:26:00 +0000[diff] [blame]114maintainability of our common source base. If you have suggestions for topics to
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]115be included, please mail them to <a
116href="mailto:sabre@nondot.org">Chris</a>.</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]117
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]118</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]119
120<!-- *********************************************************************** -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]121<h2>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]122 <a name="mechanicalissues">Mechanical Source Issues</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]123</h2>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]124<!-- *********************************************************************** -->
125
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]126<div>
127
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]128<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]129<h3>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]130 <a name="sourceformating">Source Code Formatting</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]131</h3>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]132
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]133<div>
134
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]135<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]136<h4>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]137 <a name="scf_commenting">Commenting</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]138</h4>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]139
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]140<div>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]141
142<p>Comments are one critical part of readability and maintainability. Everyone
Chris Lattnerb927ca82009-07-12 00:10:24 +0000[diff] [blame]143knows they should comment, so should you. When writing comments, write them as
144English prose, which means they should use proper capitalization, punctuation,
145etc. Although we all should probably
Misha Brukman1b36d672003-10-06 19:26:00 +0000[diff] [blame]146comment our code more than we do, there are a few very critical places that
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]147documentation is very useful:</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]148
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]149<h5>File Headers</h5>
150
151<div>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]152
Reid Spencer669ed452007-07-09 08:04:31 +0000[diff] [blame]153<p>Every source file should have a header on it that describes the basic
154purpose of the file. If a file does not have a header, it should not be
155checked into Subversion. Most source trees will probably have a standard
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]156file header format. The standard format for the LLVM source tree looks like
157this:</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]158
Misha Brukmana8c760a2004-07-28 22:31:54 +0000[diff] [blame]159<div class="doc_code">
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]160<pre>
Chris Lattnerc49fcd12003-10-13 14:58:11 +0000[diff] [blame]161//===-- llvm/Instruction.h - Instruction class definition -------*- C++ -*-===//
Misha Brukman98117cd2009-01-02 16:58:42 +0000[diff] [blame]162//
Chris Lattner999cf092004-05-23 21:05:07 +0000[diff] [blame]163// The LLVM Compiler Infrastructure
164//
Chris Lattnerab2b10c2007-12-29 19:56:08 +0000[diff] [blame]165// This file is distributed under the University of Illinois Open Source
166// License. See LICENSE.TXT for details.
Misha Brukman98117cd2009-01-02 16:58:42 +0000[diff] [blame]167//
Chris Lattner999cf092004-05-23 21:05:07 +0000[diff] [blame]168//===----------------------------------------------------------------------===//
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]169//
170// This file contains the declaration of the Instruction class, which is the
171// base class for all of the VM instructions.
172//
173//===----------------------------------------------------------------------===//
174</pre>
Misha Brukmana8c760a2004-07-28 22:31:54 +0000[diff] [blame]175</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]176
Chris Lattnerab2b10c2007-12-29 19:56:08 +0000[diff] [blame]177<p>A few things to note about this particular format: The "<tt>-*- C++
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]178-*-</tt>" string on the first line is there to tell Emacs that the source file
Chris Lattner684cf812010-11-17 17:14:55 +0000[diff] [blame]179is a C++ file, not a C file (Emacs assumes <tt>.h</tt> files are C files by default).
180Note that this tag is not necessary in <tt>.cpp</tt> files. The name of the file is also
Misha Brukman69c58892004-07-28 22:37:57 +0000[diff] [blame]181on the first line, along with a very short description of the purpose of the
182file. This is important when printing out code and flipping though lots of
183pages.</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]184
Chris Lattnerab2b10c2007-12-29 19:56:08 +0000[diff] [blame]185<p>The next section in the file is a concise note that defines the license
186that the file is released under. This makes it perfectly clear what terms the
187source code can be distributed under and should not be modified in any way.</p>
Chris Lattner999cf092004-05-23 21:05:07 +0000[diff] [blame]188
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]189<p>The main body of the description does not have to be very long in most cases.
Misha Brukman1b36d672003-10-06 19:26:00 +0000[diff] [blame]190Here it's only two lines. If an algorithm is being implemented or something
191tricky is going on, a reference to the paper where it is published should be
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]192included, as well as any notes or "gotchas" in the code to watch out for.</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]193
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]194</div>
195
196<h5>Class overviews</h5>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]197
Jim Laskeyf55914a2006-07-31 20:18:49 +0000[diff] [blame]198<p>Classes are one fundamental part of a good object oriented design. As such,
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]199a class definition should have a comment block that explains what the class is
Misha Brukman1b36d672003-10-06 19:26:00 +0000[diff] [blame]200used for... if it's not obvious. If it's so completely obvious your grandma
201could figure it out, it's probably safe to leave it out. Naming classes
Misha Brukman3b53a262004-12-04 00:32:12 +0000[diff] [blame]202something sane goes a long ways towards avoiding writing documentation.</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]203
204
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]205<h5>Method information</h5>
206
207<div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]208
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]209<p>Methods defined in a class (as well as any global functions) should also be
Dan Gohman621a4152010-02-26 20:18:32 +0000[diff] [blame]210documented properly. A quick note about what it does and a description of the
Misha Brukman1b36d672003-10-06 19:26:00 +0000[diff] [blame]211borderline behaviour is all that is necessary here (unless something
Dan Gohman621a4152010-02-26 20:18:32 +0000[diff] [blame]212particularly tricky or insidious is going on). The hope is that people can
Misha Brukman1b36d672003-10-06 19:26:00 +0000[diff] [blame]213figure out how to use your interfaces without reading the code itself... that is
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]214the goal metric.</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]215
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]216<p>Good things to talk about here are what happens when something unexpected
217happens: does the method return null? Abort? Format your hard disk?</p>
218
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]219</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]220
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]221</div>
222
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]223<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]224<h4>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]225 <a name="scf_commentformat">Comment Formatting</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]226</h4>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]227
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]228<div>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]229
230<p>In general, prefer C++ style (<tt>//</tt>) comments. They take less space,
Misha Brukman1b36d672003-10-06 19:26:00 +0000[diff] [blame]231require less typing, don't have nesting problems, etc. There are a few cases
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]232when it is useful to use C style (<tt>/* */</tt>) comments however:</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]233
234<ol>
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]235 <li>When writing C code: Obviously if you are writing C code, use C style
Misha Brukman3b53a262004-12-04 00:32:12 +0000[diff] [blame]236 comments.</li>
Misha Brukman91d64af2004-10-26 16:18:43 +0000[diff] [blame]237 <li>When writing a header file that may be <tt>#include</tt>d by a C source
238 file.</li>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]239 <li>When writing a source file that is used by a tool that only accepts C
240 style comments.</li>
241</ol>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]242
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]243<p>To comment out a large block of code, use <tt>#if 0</tt> and <tt>#endif</tt>.
244These nest properly and are better behaved in general than C style comments.</p>
245
246</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]247
Chris Lattner2e5cbc22003-08-07 21:45:47 +0000[diff] [blame]248<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]249<h4>
Misha Brukman91d64af2004-10-26 16:18:43 +0000[diff] [blame]250 <a name="scf_includes"><tt>#include</tt> Style</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]251</h4>
Chris Lattner2e5cbc22003-08-07 21:45:47 +0000[diff] [blame]252
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]253<div>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]254
255<p>Immediately after the <a href="#scf_commenting">header file comment</a> (and
Chris Lattner2e5cbc22003-08-07 21:45:47 +0000[diff] [blame]256include guards if working on a header file), the <a
Misha Brukmana8c760a2004-07-28 22:31:54 +0000[diff] [blame]257href="#hl_dontinclude">minimal</a> list of <tt>#include</tt>s required by the
258file should be listed. We prefer these <tt>#include</tt>s to be listed in this
259order:</p>
Chris Lattner2e5cbc22003-08-07 21:45:47 +0000[diff] [blame]260
261<ol>
Chris Lattner684cf812010-11-17 17:14:55 +0000[diff] [blame]262 <li><a href="#mmheader">Main Module Header</a></li>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]263 <li><a href="#hl_privateheaders">Local/Private Headers</a></li>
Misha Brukmana8c760a2004-07-28 22:31:54 +0000[diff] [blame]264 <li><tt>llvm/*</tt></li>
265 <li><tt>llvm/Analysis/*</tt></li>
266 <li><tt>llvm/Assembly/*</tt></li>
Duncan Sands18d52f22010-09-29 20:09:55 +0000[diff] [blame]267 <li><tt>llvm/Bitcode/*</tt></li>
Misha Brukmana8c760a2004-07-28 22:31:54 +0000[diff] [blame]268 <li><tt>llvm/CodeGen/*</tt></li>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]269 <li>...</li>
Misha Brukmana8c760a2004-07-28 22:31:54 +0000[diff] [blame]270 <li><tt>Support/*</tt></li>
271 <li><tt>Config/*</tt></li>
272 <li>System <tt>#includes</tt></li>
Chris Lattner2e5cbc22003-08-07 21:45:47 +0000[diff] [blame]273</ol>
274
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]275<p>and each category should be sorted by name.</p>
Chris Lattner2e5cbc22003-08-07 21:45:47 +0000[diff] [blame]276
Chris Lattner684cf812010-11-17 17:14:55 +0000[diff] [blame]277<p><a name="mmheader">The "Main Module Header"</a> file applies to <tt>.cpp</tt> files
278which implement an interface defined by a <tt>.h</tt> file. This <tt>#include</tt>
Misha Brukman91d64af2004-10-26 16:18:43 +0000[diff] [blame]279should always be included <b>first</b> regardless of where it lives on the file
Chris Lattner684cf812010-11-17 17:14:55 +0000[diff] [blame]280system. By including a header file first in the <tt>.cpp</tt> files that implement the
Misha Brukman91d64af2004-10-26 16:18:43 +0000[diff] [blame]281interfaces, we ensure that the header does not have any hidden dependencies
282which are not explicitly #included in the header, but should be. It is also a
Chris Lattner684cf812010-11-17 17:14:55 +0000[diff] [blame]283form of documentation in the <tt>.cpp</tt> file to indicate where the interfaces it
Misha Brukman91d64af2004-10-26 16:18:43 +0000[diff] [blame]284implements are defined.</p>
Chris Lattner2e5cbc22003-08-07 21:45:47 +0000[diff] [blame]285
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]286</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]287
288<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]289<h4>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]290 <a name="scf_codewidth">Source Code Width</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]291</h4>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]292
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]293<div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]294
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]295<p>Write your code to fit within 80 columns of text. This helps those of us who
296like to print out code and look at your code in an xterm without resizing
297it.</p>
298
Chris Lattner5822e9e2008-07-08 05:12:37 +0000[diff] [blame]299<p>The longer answer is that there must be some limit to the width of the code
300in order to reasonably allow developers to have multiple files side-by-side in
301windows on a modest display. If you are going to pick a width limit, it is
302somewhat arbitrary but you might as well pick something standard. Going with
30390 columns (for example) instead of 80 columns wouldn't add any significant
304value and would be detrimental to printing out code. Also many other projects
305have standardized on 80 columns, so some people have already configured their
306editors for it (vs something else, like 90 columns).</p>
307
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]308<p>This is one of many contentious issues in coding standards, but it is not up
Chris Lattner5822e9e2008-07-08 05:12:37 +0000[diff] [blame]309for debate.</p>
310
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]311</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]312
313<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]314<h4>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]315 <a name="scf_spacestabs">Use Spaces Instead of Tabs</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]316</h4>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]317
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]318<div>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]319
320<p>In all cases, prefer spaces to tabs in source files. People have different
Benjamin Kramer8040cd32009-10-12 14:46:08 +0000[diff] [blame]321preferred indentation levels, and different styles of indentation that they
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]322like; this is fine. What isn't fine is that different editors/viewers expand
323tabs out to different tab stops. This can cause your code to look completely
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]324unreadable, and it is not worth dealing with.</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]325
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]326<p>As always, follow the <a href="#goldenrule">Golden Rule</a> above: follow the
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]327style of existing code if you are modifying and extending it. If you like four
Misha Brukman1b36d672003-10-06 19:26:00 +0000[diff] [blame]328spaces of indentation, <b>DO NOT</b> do that in the middle of a chunk of code
329with two spaces of indentation. Also, do not reindent a whole source file: it
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]330makes for incredible diffs that are absolutely worthless.</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]331
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]332</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]333
334<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]335<h4>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]336 <a name="scf_indentation">Indent Code Consistently</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]337</h4>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]338
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]339<div>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]340
Chris Lattner684cf812010-11-17 17:14:55 +0000[diff] [blame]341<p>Okay, in your first year of programming you were told that indentation is
Misha Brukman1b36d672003-10-06 19:26:00 +0000[diff] [blame]342important. If you didn't believe and internalize this then, now is the time.
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]343Just do it.</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]344
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]345</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]346
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]347</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]348
349<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]350<h3>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]351 <a name="compilerissues">Compiler Issues</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]352</h3>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]353
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]354<div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]355
356<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]357<h4>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]358 <a name="ci_warningerrors">Treat Compiler Warnings Like Errors</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]359</h4>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]360
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]361<div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]362
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]363<p>If your code has compiler warnings in it, something is wrong &mdash; you
364aren't casting values correctly, your have "questionable" constructs in your
365code, or you are doing something legitimately wrong. Compiler warnings can
366cover up legitimate errors in output and make dealing with a translation unit
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]367difficult.</p>
368
369<p>It is not possible to prevent all warnings from all compilers, nor is it
Misha Brukman1b36d672003-10-06 19:26:00 +0000[diff] [blame]370desirable. Instead, pick a standard compiler (like <tt>gcc</tt>) that provides
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]371a good thorough set of warnings, and stick to it. At least in the case of
Misha Brukman1b36d672003-10-06 19:26:00 +0000[diff] [blame]372<tt>gcc</tt>, it is possible to work around any spurious errors by changing the
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]373syntax of the code slightly. For example, a warning that annoys me occurs when
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]374I write code like this:</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]375
Misha Brukmana8c760a2004-07-28 22:31:54 +0000[diff] [blame]376<div class="doc_code">
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]377<pre>
Misha Brukmana8c760a2004-07-28 22:31:54 +0000[diff] [blame]378if (V = getValue()) {
379 ...
380}
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]381</pre>
Misha Brukmana8c760a2004-07-28 22:31:54 +0000[diff] [blame]382</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]383
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]384<p><tt>gcc</tt> will warn me that I probably want to use the <tt>==</tt>
385operator, and that I probably mistyped it. In most cases, I haven't, and I
386really don't want the spurious errors. To fix this particular problem, I
387rewrite the code like this:</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]388
Misha Brukmana8c760a2004-07-28 22:31:54 +0000[diff] [blame]389<div class="doc_code">
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]390<pre>
Misha Brukmana8c760a2004-07-28 22:31:54 +0000[diff] [blame]391if ((V = getValue())) {
392 ...
393}
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]394</pre>
Misha Brukmana8c760a2004-07-28 22:31:54 +0000[diff] [blame]395</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]396
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]397<p>which shuts <tt>gcc</tt> up. Any <tt>gcc</tt> warning that annoys you can
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]398be fixed by massaging the code appropriately.</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]399
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]400<p>These are the <tt>gcc</tt> warnings that I prefer to enable:</p>
401
402<div class="doc_code">
403<pre>
404-Wall -Winline -W -Wwrite-strings -Wno-unused
405</pre>
406</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]407
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]408</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]409
410<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]411<h4>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]412 <a name="ci_portable_code">Write Portable Code</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]413</h4>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]414
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]415<div>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]416
417<p>In almost all cases, it is possible and within reason to write completely
Misha Brukman2d10d752003-07-28 21:57:18 +0000[diff] [blame]418portable code. If there are cases where it isn't possible to write portable
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]419code, isolate it behind a well defined (and well documented) interface.</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]420
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]421<p>In practice, this means that you shouldn't assume much about the host
Chris Lattner221b2392010-11-16 22:19:06 +0000[diff] [blame]422compiler, and Visual Studio tends to be the lowest common denominator.
423If advanced features are used, they should only be an implementation detail of
424a library which has a simple exposed API, and preferably be buried in
425libSystem.</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]426
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]427</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]428
Reid Spencerb40df562004-09-23 16:03:48 +0000[diff] [blame]429<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]430<h4>
Chris Lattner558f5d22010-11-30 19:20:40 +0000[diff] [blame]431<a name="ci_rtti_exceptions">Do not use RTTI or Exceptions</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]432</h4>
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]433<div>
Chris Lattner558f5d22010-11-30 19:20:40 +0000[diff] [blame]434
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]435<p>In an effort to reduce code and executable size, LLVM does not use RTTI
436(e.g. <tt>dynamic_cast&lt;&gt;</tt>) or exceptions. These two language features
437violate the general C++ principle of <i>"you only pay for what you use"</i>,
438causing executable bloat even if exceptions are never used in the code base, or
439if RTTI is never used for a class. Because of this, we turn them off globally
440in the code.</p>
Chris Lattner558f5d22010-11-30 19:20:40 +0000[diff] [blame]441
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]442<p>That said, LLVM does make extensive use of a hand-rolled form of RTTI that
443use templates like <a href="ProgrammersManual.html#isa"><tt>isa&lt;&gt;</tt>,
444<tt>cast&lt;&gt;</tt>, and <tt>dyn_cast&lt;&gt;</tt></a>. This form of RTTI is
445opt-in and can be added to any class. It is also substantially more efficient
446than <tt>dynamic_cast&lt;&gt;</tt>.</p>
Chris Lattner558f5d22010-11-30 19:20:40 +0000[diff] [blame]447
448</div>
449
450<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]451<h4>
Reid Spencerb40df562004-09-23 16:03:48 +0000[diff] [blame]452<a name="ci_class_struct">Use of <tt>class</tt> and <tt>struct</tt> Keywords</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]453</h4>
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]454<div>
Misha Brukmanedcc78a2004-10-26 15:45:13 +0000[diff] [blame]455
456<p>In C++, the <tt>class</tt> and <tt>struct</tt> keywords can be used almost
457interchangeably. The only difference is when they are used to declare a class:
458<tt>class</tt> makes all members private by default while <tt>struct</tt> makes
459all members public by default.</p>
460
461<p>Unfortunately, not all compilers follow the rules and some will generate
462different symbols based on whether <tt>class</tt> or <tt>struct</tt> was used to
463declare the symbol. This can lead to problems at link time.</p>
464
465<p>So, the rule for LLVM is to always use the <tt>class</tt> keyword, unless
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]466<b>all</b> members are public and the type is a C++
467<a href="http://en.wikipedia.org/wiki/Plain_old_data_structure">POD</a> type, in
468which case <tt>struct</tt> is allowed.</p>
Misha Brukmanedcc78a2004-10-26 15:45:13 +0000[diff] [blame]469
Reid Spencerb40df562004-09-23 16:03:48 +0000[diff] [blame]470</div>
471
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]472</div>
473
474</div>
475
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]476<!-- *********************************************************************** -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]477<h2>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]478 <a name="styleissues">Style Issues</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]479</h2>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]480<!-- *********************************************************************** -->
481
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]482<div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]483
484<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]485<h3>
Chris Lattner684cf812010-11-17 17:14:55 +0000[diff] [blame]486 <a name="macro">The High-Level Issues</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]487</h3>
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]488<!-- ======================================================================= -->
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]489
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]490<div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]491
492<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]493<h4>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]494 <a name="hl_module">A Public Header File <b>is</b> a Module</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]495</h4>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]496
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]497<div>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]498
499<p>C++ doesn't do too well in the modularity department. There is no real
Misha Brukman2d10d752003-07-28 21:57:18 +0000[diff] [blame]500encapsulation or data hiding (unless you use expensive protocol classes), but it
501is what we have to work with. When you write a public header file (in the LLVM
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]502source tree, they live in the top level "<tt>include</tt>" directory), you are
503defining a module of functionality.</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]504
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]505<p>Ideally, modules should be completely independent of each other, and their
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]506header files should only <tt>#include</tt> the absolute minimum number of
507headers possible. A module is not just a class, a function, or a
508namespace: <a href="http://www.cuj.com/articles/2000/0002/0002c/0002c.htm">it's
509a collection of these</a> that defines an interface. This interface may be
510several functions, classes, or data structures, but the important issue is how
511they work together.</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]512
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]513<p>In general, a module should be implemented by one or more <tt>.cpp</tt>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]514files. Each of these <tt>.cpp</tt> files should include the header that defines
Chris Lattner684cf812010-11-17 17:14:55 +0000[diff] [blame]515their interface first. This ensures that all of the dependences of the module
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]516header have been properly added to the module header itself, and are not
517implicit. System headers should be included after user headers for a
518translation unit.</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]519
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]520</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]521
522<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]523<h4>
Misha Brukman91d64af2004-10-26 16:18:43 +0000[diff] [blame]524 <a name="hl_dontinclude"><tt>#include</tt> as Little as Possible</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]525</h4>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]526
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]527<div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]528
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]529<p><tt>#include</tt> hurts compile time performance. Don't do it unless you
530have to, especially in header files.</p>
531
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]532<p>But wait! Sometimes you need to have the definition of a class to use it, or
Misha Brukman91d64af2004-10-26 16:18:43 +0000[diff] [blame]533to inherit from it. In these cases go ahead and <tt>#include</tt> that header
534file. Be aware however that there are many cases where you don't need to have
535the full definition of a class. If you are using a pointer or reference to a
536class, you don't need the header file. If you are simply returning a class
537instance from a prototyped function or method, you don't need it. In fact, for
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]538most cases, you simply don't need the definition of a class. And not
Misha Brukman91d64af2004-10-26 16:18:43 +0000[diff] [blame]539<tt>#include</tt>'ing speeds up compilation.</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]540
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]541<p>It is easy to try to go too overboard on this recommendation, however. You
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]542<b>must</b> include all of the header files that you are using &mdash; you can
543include them either directly or indirectly (through another header file). To
544make sure that you don't accidentally forget to include a header file in your
545module header, make sure to include your module header <b>first</b> in the
546implementation file (as mentioned above). This way there won't be any hidden
547dependencies that you'll find out about later.</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]548
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]549</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]550
551<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]552<h4>
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]553 <a name="hl_privateheaders">Keep "Internal" Headers Private</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]554</h4>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]555
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]556<div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]557
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]558<p>Many modules have a complex implementation that causes them to use more than
559one implementation (<tt>.cpp</tt>) file. It is often tempting to put the
560internal communication interface (helper classes, extra functions, etc) in the
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]561public module header file. Don't do this!</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]562
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]563<p>If you really need to do something like this, put a private header file in
564the same directory as the source files, and include it locally. This ensures
565that your private interface remains private and undisturbed by outsiders.</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]566
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]567<p>Note however, that it's okay to put extra implementation methods in a public
568class itself. Just make them private (or protected) and all is well.</p>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]569
570</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]571
Bill Wendling7f564c02006-12-09 01:20:34 +0000[diff] [blame]572<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]573<h4>
Zhanyong Wancafe0b42010-11-23 05:03:07 +0000[diff] [blame]574 <a name="hl_earlyexit">Use Early Exits and <tt>continue</tt> to Simplify Code</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]575</h4>
Bill Wendling7f564c02006-12-09 01:20:34 +0000[diff] [blame]576
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]577<div>
Bill Wendling7f564c02006-12-09 01:20:34 +0000[diff] [blame]578
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]579<p>When reading code, keep in mind how much state and how many previous
580decisions have to be remembered by the reader to understand a block of code.
581Aim to reduce indentation where possible when it doesn't make it more difficult
582to understand the code. One great way to do this is by making use of early
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]583exits and the <tt>continue</tt> keyword in long loops. As an example of using
584an early exit from a function, consider this "bad" code:</p>
Bill Wendling7f564c02006-12-09 01:20:34 +0000[diff] [blame]585
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]586<div class="doc_code">
587<pre>
588Value *DoSomething(Instruction *I) {
589 if (!isa&lt;TerminatorInst&gt;(I) &amp;&amp;
590 I-&gt;hasOneUse() &amp;&amp; SomeOtherThing(I)) {
591 ... some long code ....
592 }
593
594 return 0;
595}
596</pre>
597</div>
Bill Wendling7f564c02006-12-09 01:20:34 +0000[diff] [blame]598
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]599<p>This code has several problems if the body of the '<tt>if</tt>' is large.
600When you're looking at the top of the function, it isn't immediately clear that
601this <em>only</em> does interesting things with non-terminator instructions, and
602only applies to things with the other predicates. Second, it is relatively
603difficult to describe (in comments) why these predicates are important because
604the <tt>if</tt> statement makes it difficult to lay out the comments. Third,
605when you're deep within the body of the code, it is indented an extra level.
606Finally, when reading the top of the function, it isn't clear what the result is
607if the predicate isn't true; you have to read to the end of the function to know
608that it returns null.</p>
Matthijs Kooijmaned58a972008-07-30 12:14:10 +0000[diff] [blame]609
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]610<p>It is much preferred to format the code like this:</p>
611
612<div class="doc_code">
613<pre>
614Value *DoSomething(Instruction *I) {
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]615 // Terminators never need 'something' done to them because ...
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]616 if (isa&lt;TerminatorInst&gt;(I))
617 return 0;
618
619 // We conservatively avoid transforming instructions with multiple uses
620 // because goats like cheese.
621 if (!I-&gt;hasOneUse())
622 return 0;
623
624 // This is really just here for example.
625 if (!SomeOtherThing(I))
626 return 0;
627
628 ... some long code ....
629}
630</pre>
631</div>
632
Chris Lattner684cf812010-11-17 17:14:55 +0000[diff] [blame]633<p>This fixes these problems. A similar problem frequently happens in <tt>for</tt>
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]634loops. A silly example is something like this:</p>
635
636<div class="doc_code">
637<pre>
638 for (BasicBlock::iterator II = BB-&gt;begin(), E = BB-&gt;end(); II != E; ++II) {
639 if (BinaryOperator *BO = dyn_cast&lt;BinaryOperator&gt;(II)) {
640 Value *LHS = BO-&gt;getOperand(0);
641 Value *RHS = BO-&gt;getOperand(1);
642 if (LHS != RHS) {
643 ...
644 }
645 }
646 }
647</pre>
648</div>
649
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]650<p>When you have very, very small loops, this sort of structure is fine. But if
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]651it exceeds more than 10-15 lines, it becomes difficult for people to read and
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]652understand at a glance. The problem with this sort of code is that it gets very
653nested very quickly. Meaning that the reader of the code has to keep a lot of
654context in their brain to remember what is going immediately on in the loop,
655because they don't know if/when the <tt>if</tt> conditions will have elses etc.
656It is strongly preferred to structure the loop like this:</p>
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]657
658<div class="doc_code">
659<pre>
660 for (BasicBlock::iterator II = BB-&gt;begin(), E = BB-&gt;end(); II != E; ++II) {
661 BinaryOperator *BO = dyn_cast&lt;BinaryOperator&gt;(II);
662 if (!BO) continue;
663
664 Value *LHS = BO-&gt;getOperand(0);
665 Value *RHS = BO-&gt;getOperand(1);
666 if (LHS == RHS) continue;
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]667
Chris Lattner684cf812010-11-17 17:14:55 +0000[diff] [blame]668 ...
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]669 }
670</pre>
671</div>
672
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]673<p>This has all the benefits of using early exits for functions: it reduces
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]674nesting of the loop, it makes it easier to describe why the conditions are true,
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]675and it makes it obvious to the reader that there is no <tt>else</tt> coming up
676that they have to push context into their brain for. If a loop is large, this
677can be a big understandability win.</p>
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]678
679</div>
680
Chris Lattner9eb7e0a2009-07-28 22:54:04 +0000[diff] [blame]681<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]682<h4>
Zhanyong Wancafe0b42010-11-23 05:03:07 +0000[diff] [blame]683 <a name="hl_else_after_return">Don't use <tt>else</tt> after a <tt>return</tt></a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]684</h4>
Chris Lattner9eb7e0a2009-07-28 22:54:04 +0000[diff] [blame]685
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]686<div>
Chris Lattner9eb7e0a2009-07-28 22:54:04 +0000[diff] [blame]687
688<p>For similar reasons above (reduction of indentation and easier reading),
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]689please do not use '<tt>else</tt>' or '<tt>else if</tt>' after something that
690interrupts control flow &mdash; like <tt>return</tt>, <tt>break</tt>,
691<tt>continue</tt>, <tt>goto</tt>, etc. For example, this is <em>bad</em>:</p>
692
Chris Lattner9eb7e0a2009-07-28 22:54:04 +0000[diff] [blame]693<div class="doc_code">
694<pre>
695 case 'J': {
696 if (Signed) {
697 Type = Context.getsigjmp_bufType();
698 if (Type.isNull()) {
699 Error = ASTContext::GE_Missing_sigjmp_buf;
700 return QualType();
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]701 <b>} else {
Chris Lattner9eb7e0a2009-07-28 22:54:04 +0000[diff] [blame]702 break;
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]703 }</b>
Chris Lattner9eb7e0a2009-07-28 22:54:04 +0000[diff] [blame]704 } else {
705 Type = Context.getjmp_bufType();
706 if (Type.isNull()) {
707 Error = ASTContext::GE_Missing_jmp_buf;
708 return QualType();
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]709 <b>} else {
Chris Lattner9eb7e0a2009-07-28 22:54:04 +0000[diff] [blame]710 break;
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]711 }</b>
Chris Lattner9eb7e0a2009-07-28 22:54:04 +0000[diff] [blame]712 }
713 }
714 }
715</pre>
716</div>
717
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]718<p>It is better to write it like this:</p>
Chris Lattner9eb7e0a2009-07-28 22:54:04 +0000[diff] [blame]719
720<div class="doc_code">
721<pre>
722 case 'J':
723 if (Signed) {
724 Type = Context.getsigjmp_bufType();
725 if (Type.isNull()) {
726 Error = ASTContext::GE_Missing_sigjmp_buf;
727 return QualType();
728 }
729 } else {
730 Type = Context.getjmp_bufType();
731 if (Type.isNull()) {
732 Error = ASTContext::GE_Missing_jmp_buf;
733 return QualType();
734 }
735 }
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]736 <b>break;</b>
Chris Lattner9eb7e0a2009-07-28 22:54:04 +0000[diff] [blame]737</pre>
738</div>
739
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]740<p>Or better yet (in this case) as:</p>
Chris Lattner9eb7e0a2009-07-28 22:54:04 +0000[diff] [blame]741
742<div class="doc_code">
743<pre>
744 case 'J':
745 if (Signed)
746 Type = Context.getsigjmp_bufType();
747 else
748 Type = Context.getjmp_bufType();
749
750 if (Type.isNull()) {
751 Error = Signed ? ASTContext::GE_Missing_sigjmp_buf :
752 ASTContext::GE_Missing_jmp_buf;
753 return QualType();
754 }
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]755 <b>break;</b>
Chris Lattner9eb7e0a2009-07-28 22:54:04 +0000[diff] [blame]756</pre>
757</div>
758
759<p>The idea is to reduce indentation and the amount of code you have to keep
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]760track of when reading the code.</p>
Chris Lattner9eb7e0a2009-07-28 22:54:04 +0000[diff] [blame]761
762</div>
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]763
764<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]765<h4>
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]766 <a name="hl_predicateloops">Turn Predicate Loops into Predicate Functions</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]767</h4>
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]768
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]769<div>
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]770
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]771<p>It is very common to write small loops that just compute a boolean value.
772There are a number of ways that people commonly write these, but an example of
773this sort of thing is:</p>
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]774
775<div class="doc_code">
776<pre>
777 <b>bool FoundFoo = false;</b>
778 for (unsigned i = 0, e = BarList.size(); i != e; ++i)
779 if (BarList[i]-&gt;isFoo()) {
780 <b>FoundFoo = true;</b>
781 break;
782 }
783
784 <b>if (FoundFoo) {</b>
785 ...
786 }
787</pre>
788</div>
789
790<p>This sort of code is awkward to write, and is almost always a bad sign.
791Instead of this sort of loop, we strongly prefer to use a predicate function
792(which may be <a href="#micro_anonns">static</a>) that uses
Chris Lattner59fec6a2009-07-22 16:54:14 +0000[diff] [blame]793<a href="#hl_earlyexit">early exits</a> to compute the predicate. We prefer
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]794the code to be structured like this:</p>
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]795
796<div class="doc_code">
797<pre>
798/// ListContainsFoo - Return true if the specified list has an element that is
799/// a foo.
800static bool ListContainsFoo(const std::vector&lt;Bar*&gt; &amp;List) {
801 for (unsigned i = 0, e = List.size(); i != e; ++i)
802 if (List[i]-&gt;isFoo())
803 return true;
804 return false;
805}
806...
807
808 <b>if (ListContainsFoo(BarList)) {</b>
809 ...
810 }
811</pre>
812</div>
813
814<p>There are many reasons for doing this: it reduces indentation and factors out
815code which can often be shared by other code that checks for the same predicate.
816More importantly, it <em>forces you to pick a name</em> for the function, and
817forces you to write a comment for it. In this silly example, this doesn't add
818much value. However, if the condition is complex, this can make it a lot easier
819for the reader to understand the code that queries for this predicate. Instead
Chris Lattner71d8f3b2009-07-22 16:30:39 +0000[diff] [blame]820of being faced with the in-line details of how we check to see if the BarList
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]821contains a foo, we can trust the function name and continue reading with better
822locality.</p>
Bill Wendling7f564c02006-12-09 01:20:34 +0000[diff] [blame]823
824</div>
825
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]826</div>
Bill Wendling7f564c02006-12-09 01:20:34 +0000[diff] [blame]827
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]828<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]829<h3>
Chris Lattner684cf812010-11-17 17:14:55 +0000[diff] [blame]830 <a name="micro">The Low-Level Issues</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]831</h3>
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]832<!-- ======================================================================= -->
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]833
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]834<div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]835
836<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]837<h4>
838 <a name="ll_naming">
839 Name Types, Functions, Variables, and Enumerators Properly
840 </a>
841</h4>
Zhanyong Wan7fcd4dc2010-12-02 05:10:07 +0000[diff] [blame]842
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]843<div>
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]844
845<p>Poorly-chosen names can mislead the reader and cause bugs. We cannot stress
846enough how important it is to use <em>descriptive</em> names. Pick names that
847match the semantics and role of the underlying entities, within reason. Avoid
848abbreviations unless they are well known. After picking a good name, make sure
849to use consistent capitalization for the name, as inconsistency requires clients
850to either memorize the APIs or to look it up to find the exact spelling.</p>
Zhanyong Wan7fcd4dc2010-12-02 05:10:07 +0000[diff] [blame]851
Chris Lattner69ad7972010-12-10 00:52:35 +0000[diff] [blame]852<p>In general, names should be in camel case (e.g. <tt>TextFileReader</tt>
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]853and <tt>isLValue()</tt>). Different kinds of declarations have different
854rules:</p>
Zhanyong Wan7fcd4dc2010-12-02 05:10:07 +0000[diff] [blame]855
Chris Lattner69ad7972010-12-10 00:52:35 +0000[diff] [blame]856<ul>
Chris Lattner78ceb3a2010-12-10 00:54:03 +0000[diff] [blame]857<li><p><b>Type names</b> (including classes, structs, enums, typedefs, etc)
Chris Lattnerc5133092011-08-12 19:49:16 +0000[diff] [blame]858 should be nouns and start with an upper-case letter (e.g.
859 <tt>TextFileReader</tt>).</p></li>
860
861<li><p><b>Variable names</b> should be nouns (as they represent state). The
862 name should be camel case, and start with an upper case letter (e.g.
863 <tt>Leader</tt> or <tt>Boats</tt>).</p></li>
Chris Lattner69ad7972010-12-10 00:52:35 +0000[diff] [blame]864
Chris Lattner78ceb3a2010-12-10 00:54:03 +0000[diff] [blame]865<li><p><b>Function names</b> should be verb phrases (as they represent
Chris Lattner69ad7972010-12-10 00:52:35 +0000[diff] [blame]866 actions), and command-like function should be imperative. The name should
867 be camel case, and start with a lower case letter (e.g. <tt>openFile()</tt>
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]868 or <tt>isFoo()</tt>).</p></li>
Chris Lattner69ad7972010-12-10 00:52:35 +0000[diff] [blame]869
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]870<li><p><b>Enum declarations</b> (e.g. <tt>enum Foo {...}</tt>) are types, so
871 they should follow the naming conventions for types. A common use for enums
872 is as a discriminator for a union, or an indicator of a subclass. When an
873 enum is used for something like this, it should have a <tt>Kind</tt> suffix
874 (e.g. <tt>ValueKind</tt>).</p></li>
Chris Lattner69ad7972010-12-10 00:52:35 +0000[diff] [blame]875
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]876<li><p><b>Enumerators</b> (e.g. <tt>enum { Foo, Bar }</tt>) and <b>public member
877 variables</b> should start with an upper-case letter, just like types.
878 Unless the enumerators are defined in their own small namespace or inside a
879 class, enumerators should have a prefix corresponding to the enum
880 declaration name. For example, <tt>enum ValueKind { ... };</tt> may contain
881 enumerators like <tt>VK_Argument</tt>, <tt>VK_BasicBlock</tt>, etc.
882 Enumerators that are just convenience constants are exempt from the
883 requirement for a prefix. For instance:</p>
884
Zhanyong Wan7fcd4dc2010-12-02 05:10:07 +0000[diff] [blame]885<div class="doc_code">
886<pre>
887enum {
888 MaxSize = 42,
889 Density = 12
890};
891</pre>
892</div>
Chris Lattner69ad7972010-12-10 00:52:35 +0000[diff] [blame]893</li>
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]894
Chris Lattner69ad7972010-12-10 00:52:35 +0000[diff] [blame]895</ul>
896
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]897<p>As an exception, classes that mimic STL classes can have member names in
898STL's style of lower-case words separated by underscores (e.g. <tt>begin()</tt>,
899<tt>push_back()</tt>, and <tt>empty()</tt>).</p>
Chris Lattner69ad7972010-12-10 00:52:35 +0000[diff] [blame]900
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]901<p>Here are some examples of good and bad names:</p>
Zhanyong Wan7fcd4dc2010-12-02 05:10:07 +0000[diff] [blame]902
Zhanyong Wan7fcd4dc2010-12-02 05:10:07 +0000[diff] [blame]903<div class="doc_code">
904<pre>
905class VehicleMaker {
906 ...
Chris Lattner69ad7972010-12-10 00:52:35 +0000[diff] [blame]907 Factory&lt;Tire&gt; F; // Bad -- abbreviation and non-descriptive.
908 Factory&lt;Tire&gt; Factory; // Better.
909 Factory&lt;Tire&gt; TireFactory; // Even better -- if VehicleMaker has more than one
Zhanyong Wan7fcd4dc2010-12-02 05:10:07 +0000[diff] [blame]910 // kind of factories.
911};
912
913Vehicle MakeVehicle(VehicleType Type) {
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]914 VehicleMaker M; // Might be OK if having a short life-span.
915 Tire tmp1 = M.makeTire(); // Bad -- 'tmp1' provides no information.
Chris Lattner69ad7972010-12-10 00:52:35 +0000[diff] [blame]916 Light headlight = M.makeLight("head"); // Good -- descriptive.
Zhanyong Wan7fcd4dc2010-12-02 05:10:07 +0000[diff] [blame]917 ...
918}
919</pre>
920</div>
921
922</div>
923
924
925<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]926<h4>
Chris Lattner6883a882006-07-27 04:24:14 +0000[diff] [blame]927 <a name="ll_assert">Assert Liberally</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]928</h4>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]929
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]930<div>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]931
Chris Lattner684cf812010-11-17 17:14:55 +0000[diff] [blame]932<p>Use the "<tt>assert</tt>" macro to its fullest. Check all of your
Benjamin Kramer8040cd32009-10-12 14:46:08 +0000[diff] [blame]933preconditions and assumptions, you never know when a bug (not necessarily even
Misha Brukman2d10d752003-07-28 21:57:18 +0000[diff] [blame]934yours) might be caught early by an assertion, which reduces debugging time
935dramatically. The "<tt>&lt;cassert&gt;</tt>" header file is probably already
936included by the header files you are using, so it doesn't cost anything to use
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]937it.</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]938
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]939<p>To further assist with debugging, make sure to put some kind of error message
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]940in the assertion statement, which is printed if the assertion is tripped. This
Chris Lattner684cf812010-11-17 17:14:55 +0000[diff] [blame]941helps the poor debugger make sense of why an assertion is being made and
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]942enforced, and hopefully what to do about it. Here is one complete example:</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]943
Misha Brukmana8c760a2004-07-28 22:31:54 +0000[diff] [blame]944<div class="doc_code">
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]945<pre>
Misha Brukmana8c760a2004-07-28 22:31:54 +0000[diff] [blame]946inline Value *getOperand(unsigned i) {
947 assert(i &lt; Operands.size() &amp;&amp; "getOperand() out of range!");
948 return Operands[i];
949}
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]950</pre>
Misha Brukmana8c760a2004-07-28 22:31:54 +0000[diff] [blame]951</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]952
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]953<p>Here are more examples:</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]954
Misha Brukmana8c760a2004-07-28 22:31:54 +0000[diff] [blame]955<div class="doc_code">
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]956<pre>
Misha Brukmana8c760a2004-07-28 22:31:54 +0000[diff] [blame]957assert(Ty-&gt;isPointerType() &amp;&amp; "Can't allocate a non pointer type!");
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]958
Misha Brukmana8c760a2004-07-28 22:31:54 +0000[diff] [blame]959assert((Opcode == Shl || Opcode == Shr) &amp;&amp; "ShiftInst Opcode invalid!");
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]960
Misha Brukmana8c760a2004-07-28 22:31:54 +0000[diff] [blame]961assert(idx &lt; getNumSuccessors() &amp;&amp; "Successor # out of range!");
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]962
Misha Brukmana8c760a2004-07-28 22:31:54 +0000[diff] [blame]963assert(V1.getType() == V2.getType() &amp;&amp; "Constant types must be identical!");
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]964
Misha Brukmana8c760a2004-07-28 22:31:54 +0000[diff] [blame]965assert(isa&lt;PHINode&gt;(Succ-&gt;front()) &amp;&amp; "Only works on PHId BBs!");
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]966</pre>
Misha Brukmana8c760a2004-07-28 22:31:54 +0000[diff] [blame]967</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]968
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]969<p>You get the idea.</p>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]970
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]971<p>Please be aware that, when adding assert statements, not all compilers are aware of
Nick Lewyckyf4408e82008-05-31 23:54:55 +0000[diff] [blame]972the semantics of the assert. In some places, asserts are used to indicate a piece of
973code that should not be reached. These are typically of the form:</p>
974
975<div class="doc_code">
976<pre>
Dan Gohmanf0032762008-11-24 17:18:39 +0000[diff] [blame]977assert(0 &amp;&amp; "Some helpful error message");
Nick Lewyckyf4408e82008-05-31 23:54:55 +0000[diff] [blame]978</pre>
979</div>
980
981<p>When used in a function that returns a value, they should be followed with a return
982statement and a comment indicating that this line is never reached. This will prevent
983a compiler which is unable to deduce that the assert statement never returns from
984generating a warning.</p>
985
986<div class="doc_code">
987<pre>
Dan Gohmanf0032762008-11-24 17:18:39 +0000[diff] [blame]988assert(0 &amp;&amp; "Some helpful error message");
Nick Lewyckyf4408e82008-05-31 23:54:55 +0000[diff] [blame]989// Not reached
990return 0;
991</pre>
992</div>
993
Chris Lattner97928d12010-11-12 00:19:41 +0000[diff] [blame]994<p>Another issue is that values used only by assertions will produce an "unused
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]995value" warning when assertions are disabled. For example, this code will
996warn:</p>
Chris Lattner97928d12010-11-12 00:19:41 +0000[diff] [blame]997
998<div class="doc_code">
999<pre>
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]1000unsigned Size = V.size();
1001assert(Size &gt; 42 &amp;&amp; "Vector smaller than it should be");
Chris Lattner97928d12010-11-12 00:19:41 +0000[diff] [blame]1002
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]1003bool NewToSet = Myset.insert(Value);
1004assert(NewToSet &amp;&amp; "The value shouldn't be in the set yet");
Chris Lattner97928d12010-11-12 00:19:41 +0000[diff] [blame]1005</pre>
1006</div>
1007
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]1008<p>These are two interesting different cases. In the first case, the call to
Chris Lattner97928d12010-11-12 00:19:41 +0000[diff] [blame]1009V.size() is only useful for the assert, and we don't want it executed when
1010assertions are disabled. Code like this should move the call into the assert
1011itself. In the second case, the side effects of the call must happen whether
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]1012the assert is enabled or not. In this case, the value should be cast to void to
1013disable the warning. To be specific, it is preferred to write the code like
1014this:</p>
Chris Lattner97928d12010-11-12 00:19:41 +0000[diff] [blame]1015
1016<div class="doc_code">
1017<pre>
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]1018assert(V.size() &gt; 42 &amp;&amp; "Vector smaller than it should be");
Chris Lattner97928d12010-11-12 00:19:41 +0000[diff] [blame]1019
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]1020bool NewToSet = Myset.insert(Value); (void)NewToSet;
1021assert(NewToSet &amp;&amp; "The value shouldn't be in the set yet");
Chris Lattner97928d12010-11-12 00:19:41 +0000[diff] [blame]1022</pre>
1023</div>
1024
1025
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]1026</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]1027
Chris Lattnerb0bff9e2006-01-01 21:59:22 +0000[diff] [blame]1028<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1029<h4>
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]1030 <a name="ll_ns_std">Do Not Use '<tt>using namespace std</tt>'</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1031</h4>
Chris Lattnerb0bff9e2006-01-01 21:59:22 +0000[diff] [blame]1032
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1033<div>
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]1034
Chris Lattnerb0bff9e2006-01-01 21:59:22 +0000[diff] [blame]1035<p>In LLVM, we prefer to explicitly prefix all identifiers from the standard
Bill Wendling4ad6d612006-12-09 01:35:43 +0000[diff] [blame]1036namespace with an "<tt>std::</tt>" prefix, rather than rely on
1037"<tt>using namespace std;</tt>".</p>
Chris Lattnerb0bff9e2006-01-01 21:59:22 +0000[diff] [blame]1038
Bill Wendling4ad6d612006-12-09 01:35:43 +0000[diff] [blame]1039<p> In header files, adding a '<tt>using namespace XXX</tt>' directive pollutes
Misha Brukman2d33c542008-12-11 19:37:04 +0000[diff] [blame]1040the namespace of any source file that <tt>#include</tt>s the header. This is
1041clearly a bad thing.</p>
Chris Lattnerb0bff9e2006-01-01 21:59:22 +0000[diff] [blame]1042
Chris Lattner684cf812010-11-17 17:14:55 +0000[diff] [blame]1043<p>In implementation files (e.g. <tt>.cpp</tt> files), the rule is more of a stylistic
Bill Wendling4ad6d612006-12-09 01:35:43 +0000[diff] [blame]1044rule, but is still important. Basically, using explicit namespace prefixes
1045makes the code <b>clearer</b>, because it is immediately obvious what facilities
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]1046are being used and where they are coming from. And <b>more portable</b>, because
Bill Wendling4ad6d612006-12-09 01:35:43 +0000[diff] [blame]1047namespace clashes cannot occur between LLVM code and other namespaces. The
1048portability rule is important because different standard library implementations
1049expose different symbols (potentially ones they shouldn't), and future revisions
1050to the C++ standard will add more symbols to the <tt>std</tt> namespace. As
1051such, we never use '<tt>using namespace std;</tt>' in LLVM.</p>
Chris Lattnerb0bff9e2006-01-01 21:59:22 +0000[diff] [blame]1052
Bill Wendling4ad6d612006-12-09 01:35:43 +0000[diff] [blame]1053<p>The exception to the general rule (i.e. it's not an exception for
1054the <tt>std</tt> namespace) is for implementation files. For example, all of
1055the code in the LLVM project implements code that lives in the 'llvm' namespace.
Chris Lattnereddd9692010-11-17 19:47:20 +0000[diff] [blame]1056As such, it is ok, and actually clearer, for the <tt>.cpp</tt> files to have a
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]1057'<tt>using namespace llvm;</tt>' directive at the top, after the
Chris Lattnereddd9692010-11-17 19:47:20 +0000[diff] [blame]1058<tt>#include</tt>s. This reduces indentation in the body of the file for source
1059editors that indent based on braces, and keeps the conceptual context cleaner.
1060The general form of this rule is that any <tt>.cpp</tt> file that implements
1061code in any namespace may use that namespace (and its parents'), but should not
1062use any others.</p>
Chris Lattnerb0bff9e2006-01-01 21:59:22 +0000[diff] [blame]1063
1064</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]1065
1066<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1067<h4>
1068 <a name="ll_virtual_anch">
1069 Provide a Virtual Method Anchor for Classes in Headers
1070 </a>
1071</h4>
Chris Lattner6883a882006-07-27 04:24:14 +0000[diff] [blame]1072
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1073<div>
Chris Lattner6883a882006-07-27 04:24:14 +0000[diff] [blame]1074
1075<p>If a class is defined in a header file and has a v-table (either it has
1076virtual methods or it derives from classes with virtual methods), it must
1077always have at least one out-of-line virtual method in the class. Without
Misha Brukman2d33c542008-12-11 19:37:04 +0000[diff] [blame]1078this, the compiler will copy the vtable and RTTI into every <tt>.o</tt> file
1079that <tt>#include</tt>s the header, bloating <tt>.o</tt> file sizes and
1080increasing link times.</p>
Chris Lattner6883a882006-07-27 04:24:14 +0000[diff] [blame]1081
1082</div>
1083
Chris Lattner0af39ea2009-06-30 06:13:23 +0000[diff] [blame]1084<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1085<h4>
Nick Lewycky7ac01992011-02-20 02:03:04 +0000[diff] [blame]1086 <a name="ll_end">Don't evaluate <tt>end()</tt> every time through a loop</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1087</h4>
Chris Lattner0af39ea2009-06-30 06:13:23 +0000[diff] [blame]1088
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1089<div>
Chris Lattner0af39ea2009-06-30 06:13:23 +0000[diff] [blame]1090
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]1091<p>Because C++ doesn't have a standard "<tt>foreach</tt>" loop (though it can be
1092emulated with macros and may be coming in C++'0x) we end up writing a lot of
1093loops that manually iterate from begin to end on a variety of containers or
1094through other data structures. One common mistake is to write a loop in this
1095style:</p>
Chris Lattner0af39ea2009-06-30 06:13:23 +0000[diff] [blame]1096
1097<div class="doc_code">
1098<pre>
1099 BasicBlock *BB = ...
1100 for (BasicBlock::iterator I = BB->begin(); I != <b>BB->end()</b>; ++I)
1101 ... use I ...
1102</pre>
1103</div>
1104
1105<p>The problem with this construct is that it evaluates "<tt>BB->end()</tt>"
1106every time through the loop. Instead of writing the loop like this, we strongly
1107prefer loops to be written so that they evaluate it once before the loop starts.
1108A convenient way to do this is like so:</p>
1109
1110<div class="doc_code">
1111<pre>
1112 BasicBlock *BB = ...
1113 for (BasicBlock::iterator I = BB->begin(), E = <b>BB->end()</b>; I != E; ++I)
1114 ... use I ...
1115</pre>
1116</div>
1117
1118<p>The observant may quickly point out that these two loops may have different
1119semantics: if the container (a basic block in this case) is being mutated, then
1120"<tt>BB->end()</tt>" may change its value every time through the loop and the
1121second loop may not in fact be correct. If you actually do depend on this
Chris Lattner09bc65e2009-06-30 06:20:03 +0000[diff] [blame]1122behavior, please write the loop in the first form and add a comment indicating
Chris Lattner0af39ea2009-06-30 06:13:23 +0000[diff] [blame]1123that you did it intentionally.</p>
1124
1125<p>Why do we prefer the second form (when correct)? Writing the loop in the
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]1126first form has two problems. First it may be less efficient than evaluating it
1127at the start of the loop. In this case, the cost is probably minor &mdash; a
1128few extra loads every time through the loop. However, if the base expression is
1129more complex, then the cost can rise quickly. I've seen loops where the end
Chris Lattner0af39ea2009-06-30 06:13:23 +0000[diff] [blame]1130expression was actually something like: "<tt>SomeMap[x]->end()</tt>" and map
Chris Lattnera44f87f2009-06-30 06:27:54 +0000[diff] [blame]1131lookups really aren't cheap. By writing it in the second form consistently, you
Chris Lattner0af39ea2009-06-30 06:13:23 +0000[diff] [blame]1132eliminate the issue entirely and don't even have to think about it.</p>
1133
Chris Lattner09bc65e2009-06-30 06:20:03 +0000[diff] [blame]1134<p>The second (even bigger) issue is that writing the loop in the first form
Chris Lattner0af39ea2009-06-30 06:13:23 +0000[diff] [blame]1135hints to the reader that the loop is mutating the container (a fact that a
1136comment would handily confirm!). If you write the loop in the second form, it
1137is immediately obvious without even looking at the body of the loop that the
1138container isn't being modified, which makes it easier to read the code and
1139understand what it does.</p>
1140
1141<p>While the second form of the loop is a few extra keystrokes, we do strongly
1142prefer it.</p>
1143
1144</div>
1145
Chris Lattner6883a882006-07-27 04:24:14 +0000[diff] [blame]1146<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1147<h4>
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]1148 <a name="ll_iostream"><tt>#include &lt;iostream&gt;</tt> is Forbidden</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1149</h4>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]1150
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1151<div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]1152
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]1153<p>The use of <tt>#include &lt;iostream&gt;</tt> in library files is
1154hereby <b><em>forbidden</em></b>. The primary reason for doing this is to
1155support clients using LLVM libraries as part of larger systems. In particular,
1156we statically link LLVM into some dynamic libraries. Even if LLVM isn't used,
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]1157the static constructors are run whenever an application starts up that uses the
1158dynamic library. There are two problems with this:</p>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]1159
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]1160<ol>
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]1161 <li>The time to run the static c'tors impacts startup time of applications
1162 &mdash; a critical time for GUI apps.</li>
1163
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]1164 <li>The static c'tors cause the app to pull many extra pages of memory off the
1165 disk: both the code for the static c'tors in each <tt>.o</tt> file and the
1166 small amount of data that gets touched. In addition, touched/dirty pages
1167 put more pressure on the VM system on low-memory machines.</li>
1168</ol>
1169
1170<p>Note that using the other stream headers (<tt>&lt;sstream&gt;</tt> for
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]1171example) is not problematic in this regard &mdash;
1172just <tt>&lt;iostream&gt;</tt>. However, <tt>raw_ostream</tt> provides various
1173APIs that are better performing for almost every use than <tt>std::ostream</tt>
1174style APIs. <b>Therefore new code should always
Chris Lattner983c5922009-08-23 21:53:47 +0000[diff] [blame]1175use <a href="#ll_raw_ostream"><tt>raw_ostream</tt></a> for writing, or
1176the <tt>llvm::MemoryBuffer</tt> API for reading files.</b></p>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]1177
1178</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]1179
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]1180
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]1181<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1182<h4>
Chris Lattner684cf812010-11-17 17:14:55 +0000[diff] [blame]1183 <a name="ll_raw_ostream">Use <tt>raw_ostream</tt></a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1184</h4>
Chris Lattner684cf812010-11-17 17:14:55 +0000[diff] [blame]1185
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1186<div>
Chris Lattner684cf812010-11-17 17:14:55 +0000[diff] [blame]1187
1188<p>LLVM includes a lightweight, simple, and efficient stream implementation
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]1189in <tt>llvm/Support/raw_ostream.h</tt>, which provides all of the common
1190features of <tt>std::ostream</tt>. All new code should use <tt>raw_ostream</tt>
1191instead of <tt>ostream</tt>.</p>
Chris Lattner684cf812010-11-17 17:14:55 +0000[diff] [blame]1192
1193<p>Unlike <tt>std::ostream</tt>, <tt>raw_ostream</tt> is not a template and can
1194be forward declared as <tt>class raw_ostream</tt>. Public headers should
1195generally not include the <tt>raw_ostream</tt> header, but use forward
1196declarations and constant references to <tt>raw_ostream</tt> instances.</p>
1197
1198</div>
1199
1200
1201<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1202<h4>
Chris Lattner6883a882006-07-27 04:24:14 +0000[diff] [blame]1203 <a name="ll_avoidendl">Avoid <tt>std::endl</tt></a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1204</h4>
Chris Lattner13c46592002-01-20 19:01:26 +0000[diff] [blame]1205
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1206<div>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]1207
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]1208<p>The <tt>std::endl</tt> modifier, when used with <tt>iostreams</tt> outputs a
1209newline to the output stream specified. In addition to doing this, however, it
1210also flushes the output stream. In other words, these are equivalent:</p>
Chris Lattner13c46592002-01-20 19:01:26 +0000[diff] [blame]1211
Misha Brukmana8c760a2004-07-28 22:31:54 +0000[diff] [blame]1212<div class="doc_code">
Chris Lattner13c46592002-01-20 19:01:26 +0000[diff] [blame]1213<pre>
Misha Brukmana8c760a2004-07-28 22:31:54 +0000[diff] [blame]1214std::cout &lt;&lt; std::endl;
1215std::cout &lt;&lt; '\n' &lt;&lt; std::flush;
Chris Lattner13c46592002-01-20 19:01:26 +0000[diff] [blame]1216</pre>
Misha Brukmana8c760a2004-07-28 22:31:54 +0000[diff] [blame]1217</div>
Chris Lattner13c46592002-01-20 19:01:26 +0000[diff] [blame]1218
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]1219<p>Most of the time, you probably have no reason to flush the output stream, so
Misha Brukmana8c760a2004-07-28 22:31:54 +0000[diff] [blame]1220it's better to use a literal <tt>'\n'</tt>.</p>
Chris Lattner13c46592002-01-20 19:01:26 +0000[diff] [blame]1221
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]1222</div>
Chris Lattner13c46592002-01-20 19:01:26 +0000[diff] [blame]1223
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1224</div>
Bill Wendling7f564c02006-12-09 01:20:34 +0000[diff] [blame]1225
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]1226<!-- ======================================================================= -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1227<h3>
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]1228 <a name="nano">Microscopic Details</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1229</h3>
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]1230<!-- ======================================================================= -->
1231
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1232<div>
1233
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]1234<p>This section describes preferred low-level formatting guidelines along with
1235reasoning on why we prefer them.</p>
1236
1237<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1238<h4>
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]1239 <a name="micro_spaceparen">Spaces Before Parentheses</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1240</h4>
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]1241
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1242<div>
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]1243
Chris Lattner684cf812010-11-17 17:14:55 +0000[diff] [blame]1244<p>We prefer to put a space before an open parenthesis only in control flow
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]1245statements, but not in normal function call expressions and function-like
1246macros. For example, this is good:</p>
1247
1248<div class="doc_code">
1249<pre>
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]1250<b>if (</b>x) ...
1251<b>for (</b>i = 0; i != 100; ++i) ...
1252<b>while (</b>llvm_rocks) ...
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]1253
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]1254<b>somefunc(</b>42);
1255<b><a href="#ll_assert">assert</a>(</b>3 != 4 &amp;&amp; "laws of math are failing me");
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]1256
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]1257a = <b>foo(</b>42, 92) + <b>bar(</b>x);
1258</pre>
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]1259</div>
1260
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]1261<p>and this is bad:</p>
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]1262
1263<div class="doc_code">
1264<pre>
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]1265<b>if(</b>x) ...
1266<b>for(</b>i = 0; i != 100; ++i) ...
1267<b>while(</b>llvm_rocks) ...
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]1268
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]1269<b>somefunc (</b>42);
1270<b><a href="#ll_assert">assert</a> (</b>3 != 4 &amp;&amp; "laws of math are failing me");
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]1271
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]1272a = <b>foo (</b>42, 92) + <b>bar (</b>x);
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]1273</pre>
1274</div>
1275
1276<p>The reason for doing this is not completely arbitrary. This style makes
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]1277control flow operators stand out more, and makes expressions flow better. The
1278function call operator binds very tightly as a postfix operator. Putting a
1279space after a function name (as in the last example) makes it appear that the
1280code might bind the arguments of the left-hand-side of a binary operator with
1281the argument list of a function and the name of the right side. More
1282specifically, it is easy to misread the "a" example as:</p>
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]1283
1284<div class="doc_code">
1285<pre>
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]1286a = foo <b>(</b>(42, 92) + bar<b>)</b> (x);
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]1287</pre>
1288</div>
1289
Bill Wendlingb8602032010-12-21 03:31:05 +0000[diff] [blame]1290<p>when skimming through the code. By avoiding a space in a function, we avoid
1291this misinterpretation.</p>
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]1292
1293</div>
1294
1295<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1296<h4>
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]1297 <a name="micro_preincrement">Prefer Preincrement</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1298</h4>
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]1299
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1300<div>
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]1301
1302<p>Hard fast rule: Preincrement (<tt>++X</tt>) may be no slower than
1303postincrement (<tt>X++</tt>) and could very well be a lot faster than it. Use
1304preincrementation whenever possible.</p>
1305
1306<p>The semantics of postincrement include making a copy of the value being
1307incremented, returning it, and then preincrementing the "work value". For
1308primitive types, this isn't a big deal... but for iterators, it can be a huge
1309issue (for example, some iterators contains stack and set objects in them...
1310copying an iterator could invoke the copy ctor's of these as well). In general,
1311get in the habit of always using preincrement, and you won't have a problem.</p>
1312
1313</div>
1314
1315<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1316<h4>
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]1317 <a name="micro_namespaceindent">Namespace Indentation</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1318</h4>
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]1319
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1320<div>
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]1321
1322<p>
Chris Lattner684cf812010-11-17 17:14:55 +0000[diff] [blame]1323In general, we strive to reduce indentation wherever possible. This is useful
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]1324because we want code to <a href="#scf_codewidth">fit into 80 columns</a> without
1325wrapping horribly, but also because it makes it easier to understand the code.
1326Namespaces are a funny thing: they are often large, and we often desire to put
1327lots of stuff into them (so they can be large). Other times they are tiny,
1328because they just hold an enum or something similar. In order to balance this,
1329we use different approaches for small versus large namespaces.
1330</p>
1331
1332<p>
1333If a namespace definition is small and <em>easily</em> fits on a screen (say,
1334less than 35 lines of code), then you should indent its body. Here's an
1335example:
1336</p>
1337
1338<div class="doc_code">
1339<pre>
Chris Lattner59fec6a2009-07-22 16:54:14 +0000[diff] [blame]1340namespace llvm {
1341 namespace X86 {
1342 /// RelocationType - An enum for the x86 relocation codes. Note that
1343 /// the terminology here doesn't follow x86 convention - word means
1344 /// 32-bit and dword means 64-bit.
1345 enum RelocationType {
1346 /// reloc_pcrel_word - PC relative relocation, add the relocated value to
1347 /// the value already in memory, after we adjust it for where the PC is.
1348 reloc_pcrel_word = 0,
1349
1350 /// reloc_picrel_word - PIC base relative relocation, add the relocated
1351 /// value to the value already in memory, after we adjust it for where the
1352 /// PIC base is.
1353 reloc_picrel_word = 1,
1354
1355 /// reloc_absolute_word, reloc_absolute_dword - Absolute relocation, just
1356 /// add the relocated value to the value already in memory.
1357 reloc_absolute_word = 2,
1358 reloc_absolute_dword = 3
1359 };
1360 }
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]1361}
1362</pre>
1363</div>
1364
1365<p>Since the body is small, indenting adds value because it makes it very clear
1366where the namespace starts and ends, and it is easy to take the whole thing in
1367in one "gulp" when reading the code. If the blob of code in the namespace is
Chris Lattner684cf812010-11-17 17:14:55 +0000[diff] [blame]1368larger (as it typically is in a header in the <tt>llvm</tt> or <tt>clang</tt> namespaces), do not
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]1369indent the code, and add a comment indicating what namespace is being closed.
1370For example:</p>
1371
1372<div class="doc_code">
1373<pre>
1374namespace llvm {
1375namespace knowledge {
1376
1377/// Grokable - This class represents things that Smith can have an intimate
1378/// understanding of and contains the data associated with it.
1379class Grokable {
1380...
1381public:
1382 explicit Grokable() { ... }
1383 virtual ~Grokable() = 0;
1384
1385 ...
1386
1387};
1388
1389} // end namespace knowledge
1390} // end namespace llvm
1391</pre>
1392</div>
1393
1394<p>Because the class is large, we don't expect that the reader can easily
1395understand the entire concept in a glance, and the end of the file (where the
1396namespaces end) may be a long ways away from the place they open. As such,
1397indenting the contents of the namespace doesn't add any value, and detracts from
1398the readability of the class. In these cases it is best to <em>not</em> indent
1399the contents of the namespace.</p>
1400
1401</div>
1402
1403<!-- _______________________________________________________________________ -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1404<h4>
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]1405 <a name="micro_anonns">Anonymous Namespaces</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1406</h4>
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]1407
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1408<div>
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]1409
Chris Lattner59fec6a2009-07-22 16:54:14 +0000[diff] [blame]1410<p>After talking about namespaces in general, you may be wondering about
1411anonymous namespaces in particular.
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]1412Anonymous namespaces are a great language feature that tells the C++ compiler
1413that the contents of the namespace are only visible within the current
1414translation unit, allowing more aggressive optimization and eliminating the
1415possibility of symbol name collisions. Anonymous namespaces are to C++ as
1416"static" is to C functions and global variables. While "static" is available
1417in C++, anonymous namespaces are more general: they can make entire classes
1418private to a file.</p>
1419
1420<p>The problem with anonymous namespaces is that they naturally want to
1421encourage indentation of their body, and they reduce locality of reference: if
1422you see a random function definition in a C++ file, it is easy to see if it is
1423marked static, but seeing if it is in an anonymous namespace requires scanning
1424a big chunk of the file.</p>
1425
1426<p>Because of this, we have a simple guideline: make anonymous namespaces as
1427small as possible, and only use them for class declarations. For example, this
1428is good:</p>
1429
1430<div class="doc_code">
1431<pre>
1432<b>namespace {</b>
1433 class StringSort {
1434 ...
1435 public:
1436 StringSort(...)
1437 bool operator&lt;(const char *RHS) const;
1438 };
1439<b>} // end anonymous namespace</b>
1440
1441static void Helper() {
1442 ...
1443}
1444
1445bool StringSort::operator&lt;(const char *RHS) const {
1446 ...
1447}
1448
1449</pre>
1450</div>
1451
1452<p>This is bad:</p>
1453
1454
1455<div class="doc_code">
1456<pre>
1457<b>namespace {</b>
1458class StringSort {
1459...
1460public:
1461 StringSort(...)
1462 bool operator&lt;(const char *RHS) const;
1463};
1464
1465void Helper() {
1466 ...
1467}
1468
1469bool StringSort::operator&lt;(const char *RHS) const {
1470 ...
1471}
1472
1473<b>} // end anonymous namespace</b>
1474
1475</pre>
1476</div>
1477
1478
1479<p>This is bad specifically because if you're looking at "Helper" in the middle
1480of a large C++ file, that you have no immediate way to tell if it is local to
1481the file. When it is marked static explicitly, this is immediately obvious.
1482Also, there is no reason to enclose the definition of "operator&lt;" in the
Chris Lattner59fec6a2009-07-22 16:54:14 +0000[diff] [blame]1483namespace just because it was declared there.
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]1484</p>
1485
1486</div>
1487
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1488</div>
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]1489
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1490</div>
Chris Lattnerd283cb12009-07-22 05:40:54 +0000[diff] [blame]1491
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]1492<!-- *********************************************************************** -->
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1493<h2>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]1494 <a name="seealso">See Also</a>
NAKAMURA Takumi05d02652011-04-18 23:59:50 +0000[diff] [blame]1495</h2>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]1496<!-- *********************************************************************** -->
1497
NAKAMURA Takumif5af6ad2011-04-23 00:30:22 +0000[diff] [blame]1498<div>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]1499
1500<p>A lot of these comments and recommendations have been culled for other
1501sources. Two particularly important books for our work are:</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]1502
1503<ol>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]1504
Chris Lattner788ee282007-11-09 21:49:08 +0000[diff] [blame]1505<li><a href="http://www.amazon.com/Effective-Specific-Addison-Wesley-Professional-Computing/dp/0321334876">Effective
1506C++</a> by Scott Meyers. Also
Chris Lattner999cf092004-05-23 21:05:07 +0000[diff] [blame]1507interesting and useful are "More Effective C++" and "Effective STL" by the same
1508author.</li>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]1509
Chris Lattner788ee282007-11-09 21:49:08 +0000[diff] [blame]1510<li>Large-Scale C++ Software Design by John Lakos</li>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]1511
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]1512</ol>
1513
1514<p>If you get some free time, and you haven't read them: do so, you might learn
Misha Brukman3b53a262004-12-04 00:32:12 +0000[diff] [blame]1515something.</p>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]1516
1517</div>
1518
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]1519<!-- *********************************************************************** -->
1520
1521<hr>
Misha Brukman6c16c4b2004-05-12 18:37:22 +0000[diff] [blame]1522<address>
1523 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
Misha Brukman44408702008-12-11 17:34:48 +0000[diff] [blame]1524 src="http://jigsaw.w3.org/css-validator/images/vcss-blue" alt="Valid CSS"></a>
Misha Brukman6c16c4b2004-05-12 18:37:22 +0000[diff] [blame]1525 <a href="http://validator.w3.org/check/referer"><img
Misha Brukman44408702008-12-11 17:34:48 +0000[diff] [blame]1526 src="http://www.w3.org/Icons/valid-html401-blue" alt="Valid HTML 4.01"></a>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]1527
Misha Brukman6c16c4b2004-05-12 18:37:22 +0000[diff] [blame]1528 <a href="mailto:sabre@nondot.org">Chris Lattner</a><br>
NAKAMURA Takumib9a33632011-04-09 02:13:37 +0000[diff] [blame]1529 <a href="http://llvm.org/">LLVM Compiler Infrastructure</a><br>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]1530 Last modified: $Date$
Misha Brukman6c16c4b2004-05-12 18:37:22 +0000[diff] [blame]1531</address>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]1532
1533</body>
1534</html>