Gitiles
Code Review Sign In
gerrit-public.fairphone.software / platform / external / llvm / 999cf0984eb22f4f69da96eb7d0c8e46a3d8b7e5 / . / docs / CodingStandards.html
blob: 51a45583fa121ca9bceef1b8894381cd847132fc [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>
5 <link rel="stylesheet" href="llvm.css" type="text/css">
6 <title>A Few Coding Standards</title>
7</head>
8<body>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]9
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]10<div class="doc_title">
11 A Few Coding Standards
12</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]13
14<ol>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]15 <li><a href="#introduction">Introduction</a></li>
Chris Lattner7ae36bb2001-07-23 20:40:41 +0000[diff] [blame]16 <li><a href="#mechanicalissues">Mechanical Source Issues</a>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]17 <ol>
18 <li><a href="#sourceformating">Source Code Formatting</a>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]19 <ol>
20 <li><a href="#scf_commenting">Commenting</a></li>
21 <li><a href="#scf_commentformat">Comment Formatting</a></li>
22 <li><a href="#scf_includes">#include Style</a></li>
23 <li><a href="#scf_codewidth">Source Code Width</a></li>
24 <li><a href="#scf_spacestabs">Use Spaces Instead of Tabs</a></li>
25 <li><a href="#scf_indentation">Indent Code Consistently</a></li>
26 </ol></li>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]27 <li><a href="#compilerissues">Compiler Issues</a>
28 <ol>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]29 <li><a href="#ci_warningerrors">Treat Compiler Warnings Like
30 Errors</a></li>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]31 <li><a href="#ci_portable_code">Write Portable Code</a></li>
32 </ol></li>
33 </ol></li>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]34 <li><a href="#styleissues">Style Issues</a>
35 <ol>
36 <li><a href="#macro">The High Level Issues</a>
37 <ol>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]38 <li><a href="#hl_module">A Public Header File <b>is</b> a
39 Module</a></li>
40 <li><a href="#hl_dontinclude">#include as Little as Possible</a></li>
41 <li><a href="#hl_privateheaders">Keep "internal" Headers
42 Private</a></li>
43 </ol></li>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]44 <li><a href="#micro">The Low Level Issues</a>
45 <ol>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]46 <li><a href="#hl_assert">Assert Liberally</a></li>
47 <li><a href="#hl_preincrement">Prefer Preincrement</a></li>
Chris Lattner999cf092004-05-23 21:05:07 +0000[diff] [blame^]48 <li><a href="#hl_avoidendl">Avoid std::endl</a></li>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]49 <li><a href="#hl_exploitcpp">Exploit C++ to its Fullest</a></li>
50 </ol></li>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]51 </ol></li>
52 <li><a href="#seealso">See Also</a></li>
53</ol>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]54
Chris Lattner999cf092004-05-23 21:05:07 +0000[diff] [blame^]55<div class="doc_author">
56 <p>Written by <a href="mailto:sabre@nondot.org">Chris Lattner</a></p>
57</div>
58
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]59
60<!-- *********************************************************************** -->
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]61<div class="doc_section">
62 <a name="introduction">Introduction</a>
63</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]64<!-- *********************************************************************** -->
65
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]66<div class="doc_text">
67
68<p>This document attempts to describe a few coding standards that are being used
69in the LLVM source tree. Although no coding standards should be regarded as
Misha Brukman1b36d672003-10-06 19:26:00 +0000[diff] [blame]70absolute requirements to be followed in all instances, coding standards can be
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]71useful.</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]72
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]73<p>This document intentionally does not prescribe fixed standards for religious
Misha Brukman1b36d672003-10-06 19:26:00 +0000[diff] [blame]74issues such as brace placement and space usage. For issues like this, follow
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]75the golden rule:</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]76
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]77<blockquote>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]78
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]79<p><b><a name="goldenrule">If you are adding a significant body of source to a
80project, feel free to use whatever style you are most comfortable with. If you
81are extending, enhancing, or bug fixing already implemented code, use the style
82that is already being used so that the source is uniform and easy to
83follow.</a></b></p>
84
85</blockquote>
86
87<p>The ultimate goal of these guidelines is the increase readability and
Misha Brukman1b36d672003-10-06 19:26:00 +0000[diff] [blame]88maintainability of our common source base. If you have suggestions for topics to
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]89be included, please mail them to <a
90href="mailto:sabre@nondot.org">Chris</a>.</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]91
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]92</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]93
94<!-- *********************************************************************** -->
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]95<div class="doc_section">
96 <a name="mechanicalissues">Mechanical Source Issues</a>
97</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]98<!-- *********************************************************************** -->
99
100<!-- ======================================================================= -->
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]101<div class="doc_subsection">
102 <a name="sourceformating">Source Code Formatting</a>
103</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]104
105<!-- _______________________________________________________________________ -->
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]106<div class="doc_subsubsection">
107 <a name="scf_commenting">Commenting</a>
108</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]109
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]110<div class="doc_text">
111
112<p>Comments are one critical part of readability and maintainability. Everyone
Misha Brukman1b36d672003-10-06 19:26:00 +0000[diff] [blame]113knows they should comment, so should you. :) Although we all should probably
114comment our code more than we do, there are a few very critical places that
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]115documentation is very useful:</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]116
Chris Lattner999cf092004-05-23 21:05:07 +0000[diff] [blame^]117<b>File Headers</b>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]118
119<p>Every source file should have a header on it that
120describes the basic purpose of the file. If a file does not have a header, it
121should not be checked into CVS. Most source trees will probably have a standard
122file header format. The standard format for the LLVM source tree looks like
123this:</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]124
125<pre>
Chris Lattnerc49fcd12003-10-13 14:58:11 +0000[diff] [blame]126//===-- llvm/Instruction.h - Instruction class definition -------*- C++ -*-===//
Chris Lattner999cf092004-05-23 21:05:07 +0000[diff] [blame^]127//
128// The LLVM Compiler Infrastructure
129//
130// This file was developed by the LLVM research group and is distributed under
131// the University of Illinois Open Source License. See LICENSE.TXT for details.
132//
133//===----------------------------------------------------------------------===//
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]134//
135// This file contains the declaration of the Instruction class, which is the
136// base class for all of the VM instructions.
137//
138//===----------------------------------------------------------------------===//
Chris Lattner999cf092004-05-23 21:05:07 +0000[diff] [blame^]139
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]140</pre>
141
Chris Lattner999cf092004-05-23 21:05:07 +0000[diff] [blame^]142<p>A few things to note about this particular format: The "<tt>-*- C++
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]143-*-</tt>" string on the first line is there to tell Emacs that the source file
144is a C++ file, not a C file (Emacs assumes .h files are C files by default [Note
145that tag this is not necessary in .cpp files]). The name of the file is also on
146the first line, along with a very short description of the purpose of the file.
147This is important when printing out code and flipping though lots of pages.</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]148
Chris Lattner999cf092004-05-23 21:05:07 +0000[diff] [blame^]149<p>The next section in the file is a concise note that defines the license that
150the file is released under. This makes it perfectly clear what terms the source
151code can be distributed under.</p>
152
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]153<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]154Here it's only two lines. If an algorithm is being implemented or something
155tricky is going on, a reference to the paper where it is published should be
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]156included, 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]157
Chris Lattner999cf092004-05-23 21:05:07 +0000[diff] [blame^]158<b>Class overviews</b>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]159
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]160<p>Classes are one fundemental part of a good object oriented design. As such,
161a class definition should have a comment block that explains what the class is
Misha Brukman1b36d672003-10-06 19:26:00 +0000[diff] [blame]162used for... if it's not obvious. If it's so completely obvious your grandma
163could figure it out, it's probably safe to leave it out. Naming classes
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]164something sane goes a long ways towards avoiding writing documentation. :)</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]165
166
Chris Lattner999cf092004-05-23 21:05:07 +0000[diff] [blame^]167<b>Method information</b>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]168
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]169<p>Methods defined in a class (as well as any global functions) should also be
Misha Brukman1b36d672003-10-06 19:26:00 +0000[diff] [blame]170documented properly. A quick note about what it does any a description of the
171borderline behaviour is all that is necessary here (unless something
172particularly tricky or insideous is going on). The hope is that people can
173figure out how to use your interfaces without reading the code itself... that is
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]174the goal metric.</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]175
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]176<p>Good things to talk about here are what happens when something unexpected
177happens: does the method return null? Abort? Format your hard disk?</p>
178
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]179</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]180
181<!-- _______________________________________________________________________ -->
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]182<div class="doc_subsubsection">
183 <a name="scf_commentformat">Comment Formatting</a>
184</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]185
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]186<div class="doc_text">
187
188<p>In general, prefer C++ style (<tt>//</tt>) comments. They take less space,
Misha Brukman1b36d672003-10-06 19:26:00 +0000[diff] [blame]189require less typing, don't have nesting problems, etc. There are a few cases
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]190when it is useful to use C style (<tt>/* */</tt>) comments however:</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]191
192<ol>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]193 <li>When writing a C code: Obviously if you are writing C code, use C style
194 comments. :)</li>
195 <li>When writing a header file that may be #included by a C source file.</li>
196 <li>When writing a source file that is used by a tool that only accepts C
197 style comments.</li>
198</ol>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]199
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]200<p>To comment out a large block of code, use <tt>#if 0</tt> and <tt>#endif</tt>.
201These nest properly and are better behaved in general than C style comments.</p>
202
203</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]204
Chris Lattner2e5cbc22003-08-07 21:45:47 +0000[diff] [blame]205<!-- _______________________________________________________________________ -->
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]206<div class="doc_subsubsection">
207 <a name="scf_includes">#include Style</a>
208</div>
Chris Lattner2e5cbc22003-08-07 21:45:47 +0000[diff] [blame]209
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]210<div class="doc_text">
211
212<p>Immediately after the <a href="#scf_commenting">header file comment</a> (and
Chris Lattner2e5cbc22003-08-07 21:45:47 +0000[diff] [blame]213include guards if working on a header file), the <a
214href="hl_dontinclude">minimal</a> list of #includes required by the file should
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]215be listed. We prefer these #includes to be listed in this order:</p>
Chris Lattner2e5cbc22003-08-07 21:45:47 +0000[diff] [blame]216
217<ol>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]218 <li><a href="#mmheader">Main Module header</a></li>
219 <li><a href="#hl_privateheaders">Local/Private Headers</a></li>
220 <li>llvm/*</li>
221 <li>llvm/Analysis/*</li>
222 <li>llvm/Assembly/*</li>
223 <li>llvm/Bytecode/*</li>
224 <li>llvm/CodeGen/*</li>
225 <li>...</li>
226 <li>Support/*</li>
227 <li>Config/*</li>
228 <li>System #includes</li>
Chris Lattner2e5cbc22003-08-07 21:45:47 +0000[diff] [blame]229</ol>
230
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]231<p>... and each catagory should be sorted by name.</p>
Chris Lattner2e5cbc22003-08-07 21:45:47 +0000[diff] [blame]232
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]233<p><a name="mmheader">The "Main Module Header"</a> file applies to .cpp file
234which implement an interface defined by a .h file. This #include should always
235be included <b>first</b> regardless of where it lives on the file system. By
Chris Lattner2e5cbc22003-08-07 21:45:47 +0000[diff] [blame]236including a header file first in the .cpp files that implement the interfaces,
237we ensure that the header does not have any hidden dependencies which are not
238explicitly #included in the header, but should be. It is also a form of
239documentation in the .cpp file to indicate where the interfaces it implements
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]240are defined.</p>
Chris Lattner2e5cbc22003-08-07 21:45:47 +0000[diff] [blame]241
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]242</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]243
244<!-- _______________________________________________________________________ -->
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]245<div class="doc_subsubsection">
246 <a name="scf_codewidth">Source Code Width</a>
247</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]248
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]249<div class="doc_text">
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]250
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]251<p>Write your code to fit within 80 columns of text. This helps those of us who
252like to print out code and look at your code in an xterm without resizing
253it.</p>
254
255</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]256
257<!-- _______________________________________________________________________ -->
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]258<div class="doc_subsubsection">
259 <a name="scf_spacestabs">Use Spaces Instead of Tabs</a>
260</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]261
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]262<div class="doc_text">
263
264<p>In all cases, prefer spaces to tabs in source files. People have different
Misha Brukman1b36d672003-10-06 19:26:00 +0000[diff] [blame]265prefered indentation levels, and different styles of indentation that they
266like... this is fine. What isn't is that different editors/viewers expand tabs
267out to different tab stops. This can cause your code to look completely
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]268unreadable, and it is not worth dealing with.</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]269
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]270<p>As always, follow the <a href="#goldenrule">Golden Rule</a> above: follow the
Misha Brukman1b36d672003-10-06 19:26:00 +0000[diff] [blame]271style of existing code if your are modifying and extending it. If you like four
272spaces of indentation, <b>DO NOT</b> do that in the middle of a chunk of code
273with two spaces of indentation. Also, do not reindent a whole source file: it
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]274makes for incredible diffs that are absolutely worthless.</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]275
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]276</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]277
278<!-- _______________________________________________________________________ -->
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]279<div class="doc_subsubsection">
280 <a name="scf_indentation">Indent Code Consistently</a>
281</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]282
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]283<div class="doc_text">
284
285<p>Okay, your first year of programming you were told that indentation is
Misha Brukman1b36d672003-10-06 19:26:00 +0000[diff] [blame]286important. If you didn't believe and internalize this then, now is the time.
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]287Just do it.</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]288
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]289</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]290
291
292<!-- ======================================================================= -->
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]293<div class="doc_subsection">
294 <a name="compilerissues">Compiler Issues</a>
295</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]296
297
298<!-- _______________________________________________________________________ -->
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]299<div class="doc_subsubsection">
300 <a name="ci_warningerrors">Treat Compiler Warnings Like Errors</a>
301</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]302
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]303<div class="doc_text">
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]304
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]305<p>If your code has compiler warnings in it, something is wrong: you aren't
306casting values correctly, your have "questionable" constructs in your code, or
307you are doing something legitimately wrong. Compiler warnings can cover up
308legitimate errors in output and make dealing with a translation unit
309difficult.</p>
310
311<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]312desirable. Instead, pick a standard compiler (like <tt>gcc</tt>) that provides
313a good thorough set of warnings, and stick to them. At least in the case of
314<tt>gcc</tt>, it is possible to work around any spurious errors by changing the
315syntax of the code slightly. For example, an warning that annoys me occurs when
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]316I write code like this:</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]317
318<pre>
319 if (V = getValue()) {
320 ..
321 }
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]322</pre>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]323
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]324<p><tt>gcc</tt> will warn me that I probably want to use the <tt>==</tt>
325operator, and that I probably mistyped it. In most cases, I haven't, and I
326really don't want the spurious errors. To fix this particular problem, I
327rewrite the code like this:</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]328
329<pre>
330 if ((V = getValue())) {
331 ..
332 }
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]333</pre>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]334
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]335<p>...which shuts <tt>gcc</tt> up. Any <tt>gcc</tt> warning that annoys you can
336be fixed by massaging the code appropriately.</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]337
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]338<p>These are the <tt>gcc</tt> warnings that I prefer to enable: <tt>-Wall
339-Winline -W -Wwrite-strings -Wno-unused</tt></p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]340
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]341</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]342
343<!-- _______________________________________________________________________ -->
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]344<div class="doc_subsubsection">
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]345 <a name="ci_portable_code">Write Portable Code</a>
346</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]347
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]348<div class="doc_text">
349
350<p>In almost all cases, it is possible and within reason to write completely
Misha Brukman2d10d752003-07-28 21:57:18 +0000[diff] [blame]351portable code. If there are cases where it isn't possible to write portable
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]352code, isolate it behind a well defined (and well documented) interface.</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]353
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]354<p>In practice, this means that you shouldn't assume much about the host
355compiler, including its support for "high tech" features like partial
356specialization of templates. In fact, Visual C++ 6 could be an important target
357for our work in the future, and we don't want to have to rewrite all of our code
358to support it.</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]359
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]360</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]361
362<!-- *********************************************************************** -->
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]363<div class="doc_section">
364 <a name="styleissues">Style Issues</a>
365</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]366<!-- *********************************************************************** -->
367
368
369<!-- ======================================================================= -->
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]370<div class="doc_subsection">
371 <a name="macro">The High Level Issues</a>
372</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]373
374
375<!-- _______________________________________________________________________ -->
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]376<div class="doc_subsubsection">
377 <a name="hl_module">A Public Header File <b>is</b> a Module</a>
378</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]379
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]380<div class="doc_text">
381
382<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]383encapsulation or data hiding (unless you use expensive protocol classes), but it
384is what we have to work with. When you write a public header file (in the LLVM
385source tree, they live in the top level "include" directory), you are defining a
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]386module of functionality.</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]387
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]388<p>Ideally, modules should be completely independent of each other, and their
Misha Brukman2d10d752003-07-28 21:57:18 +0000[diff] [blame]389header files should only include the absolute minimum number of headers
390possible. A module is not just a class, a function, or a namespace: <a
391href="http://www.cuj.com/articles/2000/0002/0002c/0002c.htm">it's a collection
392of these</a> that defines an interface. This interface may be several
393functions, classes or data structures, but the important issue is how they work
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]394together.</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]395
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]396<p>In general, a module should be implemented with one or more <tt>.cpp</tt>
397files. Each of these <tt>.cpp</tt> files should include the header that defines
398their interface first. This ensure that all of the dependences of the module
399header have been properly added to the module header itself, and are not
400implicit. System headers should be included after user headers for a
401translation unit.</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]402
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]403</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]404
405<!-- _______________________________________________________________________ -->
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]406<div class="doc_subsubsection">
407 <a name="hl_dontinclude">#include as Little as Possible</a>
408</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]409
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]410<div class="doc_text">
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]411
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]412<p><tt>#include</tt> hurts compile time performance. Don't do it unless you
413have to, especially in header files.</p>
414
415<p>But wait, sometimes you need to have the definition of a class to use it, or
416to inherit from it. In these cases go ahead and #include that header file. Be
Misha Brukman2d10d752003-07-28 21:57:18 +0000[diff] [blame]417aware however that there are many cases where you don't need to have the full
418definition of a class. If you are using a pointer or reference to a class, you
419don't need the header file. If you are simply returning a class instance from a
420prototyped function or method, you don't need it. In fact, for most cases, you
421simply don't need the definition of a class... and not <tt>#include</tt>'ing
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]422speeds up compilation.</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]423
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]424<p>It is easy to try to go too overboard on this recommendation, however. You
Misha Brukman2d10d752003-07-28 21:57:18 +0000[diff] [blame]425<b>must</b> include all of the header files that you are using, either directly
426or indirectly (through another header file). To make sure that you don't
427accidently forget to include a header file in your module header, make sure to
428include your module header <b>first</b> in the implementation file (as mentioned
429above). This way there won't be any hidden dependencies that you'll find out
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]430about later...</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]431
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]432</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]433
434<!-- _______________________________________________________________________ -->
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]435<div class="doc_subsubsection">
436 <a name="hl_privateheaders">Keep "internal" Headers Private</a>
437</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]438
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]439<div class="doc_text">
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]440
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]441<p>Many modules have a complex implementation that causes them to use more than
442one implementation (<tt>.cpp</tt>) file. It is often tempting to put the
443internal communication interface (helper classes, extra functions, etc) in the
444public module header file. Don't do this. :)</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]445
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]446<p>If you really need to do something like this, put a private header file in
447the same directory as the source files, and include it locally. This ensures
448that your private interface remains private and undisturbed by outsiders.</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]449
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]450<p>Note however, that it's okay to put extra implementation methods a public
451class itself... just make them private (or protected), and all is well.</p>
452
453</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]454
455<!-- ======================================================================= -->
Chris Lattner999cf092004-05-23 21:05:07 +0000[diff] [blame^]456<div class="doc_subsection">
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]457 <a name="micro">The Low Level Issues</a>
458</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]459
460
461<!-- _______________________________________________________________________ -->
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]462<div class="doc_subsubsection">
463 <a name="hl_assert">Assert Liberally</a>
464</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]465
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]466<div class="doc_text">
467
468<p>Use the "<tt>assert</tt>" function to its fullest. Check all of your
Misha Brukman2d10d752003-07-28 21:57:18 +0000[diff] [blame]469preconditions and assumptions, you never know when a bug (not neccesarily even
470yours) might be caught early by an assertion, which reduces debugging time
471dramatically. The "<tt>&lt;cassert&gt;</tt>" header file is probably already
472included 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]473it.</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]474
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]475<p>To further assist with debugging, make sure to put some kind of error message
476in the assertion statement (which is printed if the assertion is tripped). This
Misha Brukman2d10d752003-07-28 21:57:18 +0000[diff] [blame]477helps the poor debugging make sense of why an assertion is being made and
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]478enforced, and hopefully what to do about it. Here is one complete example:</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]479
480<pre>
481 inline Value *getOperand(unsigned i) {
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]482 assert(i &lt; Operands.size() &amp;&amp; "getOperand() out of range!");
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]483 return Operands[i];
484 }
485</pre>
486
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]487<p>Here are some examples:</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]488
489<pre>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]490 assert(Ty-&gt;isPointerType() &amp;&amp; "Can't allocate a non pointer type!");
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]491
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]492 assert((Opcode == Shl || Opcode == Shr) &amp;&amp; "ShiftInst Opcode invalid!");
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]493
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]494 assert(idx &lt; getNumSuccessors() &amp;&amp; "Successor # out of range!");
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]495
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]496 assert(V1.getType() == V2.getType() &amp;&amp; "Constant types must be identical!");
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]497
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]498 assert(isa&lt;PHINode&gt;(Succ-&gt;front()) &amp;&amp; "Only works on PHId BBs!");
499</pre>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]500
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]501<p>You get the idea...</p>
502
503</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]504
505
506<!-- _______________________________________________________________________ -->
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]507<div class="doc_subsubsection">
508 <a name="hl_preincrement">Prefer Preincrement</a>
509</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]510
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]511<div class="doc_text">
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]512
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]513<p>Hard fast rule: Preincrement (++X) may be no slower than postincrement (X++)
514and could very well be a lot faster than it. Use preincrementation whenever
515possible.</p>
516
517<p>The semantics of postincrement include making a copy of the value being
Misha Brukman2d10d752003-07-28 21:57:18 +0000[diff] [blame]518incremented, returning it, and then preincrementing the "work value". For
519primitive types, this isn't a big deal... but for iterators, it can be a huge
520issue (for example, some iterators contains stack and set objects in them...
521copying an iterator could invoke the copy ctor's of these as well). In general,
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]522get in the habit of always using preincrement, and you won't have a problem.</p>
523
524</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]525
526
527<!-- _______________________________________________________________________ -->
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]528<div class="doc_subsubsection">
Chris Lattner999cf092004-05-23 21:05:07 +0000[diff] [blame^]529 <a name="hl_avoidendl">Avoid std::endl</a>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]530</div>
Chris Lattner13c46592002-01-20 19:01:26 +0000[diff] [blame]531
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]532<div class="doc_text">
533
Chris Lattner999cf092004-05-23 21:05:07 +0000[diff] [blame^]534<p>The <tt>std::endl</tt> modifier, when used with iostreams outputs a newline
535to the output stream specified. In addition to doing this, however, it also
536flushes the output stream. In other words, these are equivalent:</p>
Chris Lattner13c46592002-01-20 19:01:26 +0000[diff] [blame]537
538<pre>
Chris Lattner999cf092004-05-23 21:05:07 +0000[diff] [blame^]539 std::cout &lt;&lt; std::endl;
540 std::cout &lt;&lt; "\n" &lt;&lt; std::flush;
Chris Lattner13c46592002-01-20 19:01:26 +0000[diff] [blame]541</pre>
542
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]543<p>Most of the time, you probably have no reason to flush the output stream, so
544it's better to use a literal <tt>"\n"</tt>.</p>
Chris Lattner13c46592002-01-20 19:01:26 +0000[diff] [blame]545
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]546</div>
Chris Lattner13c46592002-01-20 19:01:26 +0000[diff] [blame]547
548<!-- _______________________________________________________________________ -->
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]549<div class="doc_subsubsection">
550 <a name="hl_exploitcpp">Exploit C++ to its Fullest</a>
551</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]552
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]553<div class="doc_text">
554
555<p>C++ is a powerful language. With a firm grasp on its capabilities, you can make
Misha Brukman2d10d752003-07-28 21:57:18 +0000[diff] [blame]556write effective, consise, readable and maintainable code all at the same time.
557By staying consistent, you reduce the amount of special cases that need to be
558remembered. Reducing the total number of lines of code you write is a good way
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]559to avoid documentation, and avoid giving bugs a place to hide.</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]560
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]561<p>For these reasons, come to know and love the contents of your local
Misha Brukman2d10d752003-07-28 21:57:18 +0000[diff] [blame]562&lt;algorithm&gt; header file. Know about &lt;functional&gt; and what it can do
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]563for you. C++ is just a tool that wants you to master it. :)</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]564
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]565</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]566
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]567<!-- *********************************************************************** -->
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]568<div class="doc_section">
569 <a name="seealso">See Also</a>
570</div>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]571<!-- *********************************************************************** -->
572
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]573<div class="doc_text">
574
575<p>A lot of these comments and recommendations have been culled for other
576sources. Two particularly important books for our work are:</p>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]577
578<ol>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]579
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]580<li><a href="http://www.aw.com/product/0,2627,0201924889,00.html">Effective
581C++</a> by Scott Meyers. There is an online version of the book (only some
582chapters though) <a
Chris Lattner999cf092004-05-23 21:05:07 +0000[diff] [blame^]583href="http://www.awlonline.com/cseng/meyerscddemo/">available as well</a>. Also
584interesting and useful are "More Effective C++" and "Effective STL" by the same
585author.</li>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]586
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]587<li><a href="http://cseng.aw.com/book/0,3828,0201633620,00.html">Large-Scale C++
588Software Design</a> by John Lakos</li>
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]589
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]590</ol>
591
592<p>If you get some free time, and you haven't read them: do so, you might learn
593something. :)</p>
594
595</div>
596
Chris Lattnerd80c43c2001-07-09 03:27:08 +0000[diff] [blame]597<!-- *********************************************************************** -->
598
599<hr>
Misha Brukman6c16c4b2004-05-12 18:37:22 +0000[diff] [blame]600<address>
601 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
602 src="http://jigsaw.w3.org/css-validator/images/vcss" alt="Valid CSS!"></a>
603 <a href="http://validator.w3.org/check/referer"><img
604 src="http://www.w3.org/Icons/valid-html401" alt="Valid HTML 4.01!"></a>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]605
Misha Brukman6c16c4b2004-05-12 18:37:22 +0000[diff] [blame]606 <a href="mailto:sabre@nondot.org">Chris Lattner</a><br>
607 <a href="http://llvm.cs.uiuc.edu">LLVM Compiler Infrastructure</a><br>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]608 Last modified: $Date$
Misha Brukman6c16c4b2004-05-12 18:37:22 +0000[diff] [blame]609</address>
Misha Brukman75f258e2003-10-24 17:57:33 +0000[diff] [blame]610
611</body>
612</html>