Please note: This project is no longer active. The website is kept online for historic purposes only.
If you´re looking for a Linux driver for your Atheros WLAN device, you should continue here .

DevDocs/PreparePatches: CodingStyle

File CodingStyle, 15.0 kB (added by mrenzmann, 13 years ago)

The CodingStyle? document that comes with current 2.6 kernels.

Line 
1
2                 Linux kernel coding style
3
4 This is a short document describing the preferred coding style for the
5 linux kernel.  Coding style is very personal, and I won't _force_ my
6 views on anybody, but this is what goes for anything that I have to be
7 able to maintain, and I'd prefer it for most other things too.  Please
8 at least consider the points made here.
9
10 First off, I'd suggest printing out a copy of the GNU coding standards,
11 and NOT read it.  Burn them, it's a great symbolic gesture.
12
13 Anyway, here goes:
14
15
16                 Chapter 1: Indentation
17
18 Tabs are 8 characters, and thus indentations are also 8 characters.
19 There are heretic movements that try to make indentations 4 (or even 2!)
20 characters deep, and that is akin to trying to define the value of PI to
21 be 3.
22
23 Rationale: The whole idea behind indentation is to clearly define where
24 a block of control starts and ends.  Especially when you've been looking
25 at your screen for 20 straight hours, you'll find it a lot easier to see
26 how the indentation works if you have large indentations.
27
28 Now, some people will claim that having 8-character indentations makes
29 the code move too far to the right, and makes it hard to read on a
30 80-character terminal screen.  The answer to that is that if you need
31 more than 3 levels of indentation, you're screwed anyway, and should fix
32 your program.
33
34 In short, 8-char indents make things easier to read, and have the added
35 benefit of warning you when you're nesting your functions too deep.
36 Heed that warning.
37
38 Don't put multiple statements on a single line unless you have
39 something to hide:
40
41         if (condition) do_this;
42           do_something_everytime;
43
44 Outside of comments, documentation and except in Kconfig, spaces are never
45 used for indentation, and the above example is deliberately broken.
46
47 Get a decent editor and don't leave whitespace at the end of lines.
48
49
50                 Chapter 2: Breaking long lines and strings
51
52 Coding style is all about readability and maintainability using commonly
53 available tools.
54
55 The limit on the length of lines is 80 columns and this is a hard limit.
56
57 Statements longer than 80 columns will be broken into sensible chunks.
58 Descendants are always substantially shorter than the parent and are placed
59 substantially to the right. The same applies to function headers with a long
60 argument list. Long strings are as well broken into shorter strings.
61
62 void fun(int a, int b, int c)
63 {
64         if (condition)
65                 printk(KERN_WARNING "Warning this is a long printk with "
66                                                 "3 parameters a: %u b: %u "
67                                                 "c: %u \n", a, b, c);
68         else
69                 next_statement;
70 }
71
72                 Chapter 3: Placing Braces
73
74 The other issue that always comes up in C styling is the placement of
75 braces.  Unlike the indent size, there are few technical reasons to
76 choose one placement strategy over the other, but the preferred way, as
77 shown to us by the prophets Kernighan and Ritchie, is to put the opening
78 brace last on the line, and put the closing brace first, thusly:
79
80         if (x is true) {
81                 we do y
82         }
83
84 However, there is one special case, namely functions: they have the
85 opening brace at the beginning of the next line, thus:
86
87         int function(int x)
88         {
89                 body of function
90         }
91
92 Heretic people all over the world have claimed that this inconsistency
93 is ...  well ...  inconsistent, but all right-thinking people know that
94 (a) K&R are _right_ and (b) K&R are right.  Besides, functions are
95 special anyway (you can't nest them in C).
96
97 Note that the closing brace is empty on a line of its own, _except_ in
98 the cases where it is followed by a continuation of the same statement,
99 ie a "while" in a do-statement or an "else" in an if-statement, like
100 this:
101
102         do {
103                 body of do-loop
104         } while (condition);
105
106 and
107
108         if (x == y) {
109                 ..
110         } else if (x > y) {
111                 ...
112         } else {
113                 ....
114         }
115
116 Rationale: K&R.
117
118 Also, note that this brace-placement also minimizes the number of empty
119 (or almost empty) lines, without any loss of readability.  Thus, as the
120 supply of new-lines on your screen is not a renewable resource (think
121 25-line terminal screens here), you have more empty lines to put
122 comments on.
123
124
125                 Chapter 4: Naming
126
127 C is a Spartan language, and so should your naming be.  Unlike Modula-2
128 and Pascal programmers, C programmers do not use cute names like
129 ThisVariableIsATemporaryCounter.  A C programmer would call that
130 variable "tmp", which is much easier to write, and not the least more
131 difficult to understand.
132
133 HOWEVER, while mixed-case names are frowned upon, descriptive names for
134 global variables are a must.  To call a global function "foo" is a
135 shooting offense.
136
137 GLOBAL variables (to be used only if you _really_ need them) need to
138 have descriptive names, as do global functions.  If you have a function
139 that counts the number of active users, you should call that
140 "count_active_users()" or similar, you should _not_ call it "cntusr()".
141
142 Encoding the type of a function into the name (so-called Hungarian
143 notation) is brain damaged - the compiler knows the types anyway and can
144 check those, and it only confuses the programmer.  No wonder MicroSoft
145 makes buggy programs.
146
147 LOCAL variable names should be short, and to the point.  If you have
148 some random integer loop counter, it should probably be called "i".
149 Calling it "loop_counter" is non-productive, if there is no chance of it
150 being mis-understood.  Similarly, "tmp" can be just about any type of
151 variable that is used to hold a temporary value.
152
153 If you are afraid to mix up your local variable names, you have another
154 problem, which is called the function-growth-hormone-imbalance syndrome.
155 See next chapter.
156
157
158                 Chapter 5: Functions
159
160 Functions should be short and sweet, and do just one thing.  They should
161 fit on one or two screenfuls of text (the ISO/ANSI screen size is 80x24,
162 as we all know), and do one thing and do that well.
163
164 The maximum length of a function is inversely proportional to the
165 complexity and indentation level of that function.  So, if you have a
166 conceptually simple function that is just one long (but simple)
167 case-statement, where you have to do lots of small things for a lot of
168 different cases, it's OK to have a longer function.
169
170 However, if you have a complex function, and you suspect that a
171 less-than-gifted first-year high-school student might not even
172 understand what the function is all about, you should adhere to the
173 maximum limits all the more closely.  Use helper functions with
174 descriptive names (you can ask the compiler to in-line them if you think
175 it's performance-critical, and it will probably do a better job of it
176 than you would have done).
177
178 Another measure of the function is the number of local variables.  They
179 shouldn't exceed 5-10, or you're doing something wrong.  Re-think the
180 function, and split it into smaller pieces.  A human brain can
181 generally easily keep track of about 7 different things, anything more
182 and it gets confused.  You know you're brilliant, but maybe you'd like
183 to understand what you did 2 weeks from now.
184
185
186                 Chapter 6: Centralized exiting of functions
187
188 Albeit deprecated by some people, the equivalent of the goto statement is
189 used frequently by compilers in form of the unconditional jump instruction.
190
191 The goto statement comes in handy when a function exits from multiple
192 locations and some common work such as cleanup has to be done.
193
194 The rationale is:
195
196 - unconditional statements are easier to understand and follow
197 - nesting is reduced
198 - errors by not updating individual exit points when making
199     modifications are prevented
200 - saves the compiler work to optimize redundant code away ;)
201
202 int fun(int )
203 {
204         int result = 0;
205         char *buffer = kmalloc(SIZE);
206
207         if (buffer == NULL)
208                 return -ENOMEM;
209
210         if (condition1) {
211                 while (loop1) {
212                         ...
213                 }
214                 result = 1;
215                 goto out;
216         }
217         ...
218 out:
219         kfree(buffer);
220         return result;
221 }
222
223                 Chapter 7: Commenting
224
225 Comments are good, but there is also a danger of over-commenting.  NEVER
226 try to explain HOW your code works in a comment: it's much better to
227 write the code so that the _working_ is obvious, and it's a waste of
228 time to explain badly written code.
229
230 Generally, you want your comments to tell WHAT your code does, not HOW.
231 Also, try to avoid putting comments inside a function body: if the
232 function is so complex that you need to separately comment parts of it,
233 you should probably go back to chapter 5 for a while.  You can make
234 small comments to note or warn about something particularly clever (or
235 ugly), but try to avoid excess.  Instead, put the comments at the head
236 of the function, telling people what it does, and possibly WHY it does
237 it.
238
239
240                 Chapter 8: You've made a mess of it
241
242 That's OK, we all do.  You've probably been told by your long-time Unix
243 user helper that "GNU emacs" automatically formats the C sources for
244 you, and you've noticed that yes, it does do that, but the defaults it
245 uses are less than desirable (in fact, they are worse than random
246 typing - an infinite number of monkeys typing into GNU emacs would never
247 make a good program).
248
249 So, you can either get rid of GNU emacs, or change it to use saner
250 values.  To do the latter, you can stick the following in your .emacs file:
251
252 (defun linux-c-mode ()
253   "C mode with adjusted defaults for use with the Linux kernel."
254   (interactive)
255   (c-mode)
256   (c-set-style "K&R")
257   (setq tab-width 8)
258   (setq indent-tabs-mode t)
259   (setq c-basic-offset 8))
260
261 This will define the M-x linux-c-mode command.  When hacking on a
262 module, if you put the string -*- linux-c -*- somewhere on the first
263 two lines, this mode will be automatically invoked. Also, you may want
264 to add
265
266 (setq auto-mode-alist (cons '("/usr/src/linux.*/.*\\.[ch]$" . linux-c-mode)
267                         auto-mode-alist))
268
269 to your .emacs file if you want to have linux-c-mode switched on
270 automagically when you edit source files under /usr/src/linux.
271
272 But even if you fail in getting emacs to do sane formatting, not
273 everything is lost: use "indent".
274
275 Now, again, GNU indent has the same brain-dead settings that GNU emacs
276 has, which is why you need to give it a few command line options.
277 However, that's not too bad, because even the makers of GNU indent
278 recognize the authority of K&R (the GNU people aren't evil, they are
279 just severely misguided in this matter), so you just give indent the
280 options "-kr -i8" (stands for "K&R, 8 character indents"), or use
281 "scripts/Lindent", which indents in the latest style.
282
283 "indent" has a lot of options, and especially when it comes to comment
284 re-formatting you may want to take a look at the man page.  But
285 remember: "indent" is not a fix for bad programming.
286
287
288                 Chapter 9: Configuration-files
289
290 For configuration options (arch/xxx/Kconfig, and all the Kconfig files),
291 somewhat different indentation is used.
292
293 Help text is indented with 2 spaces.
294
295 if CONFIG_EXPERIMENTAL
296         tristate CONFIG_BOOM
297         default n
298         help
299           Apply nitroglycerine inside the keyboard (DANGEROUS)
300         bool CONFIG_CHEER
301         depends on CONFIG_BOOM
302         default y
303         help
304           Output nice messages when you explode
305 endif
306
307 Generally, CONFIG_EXPERIMENTAL should surround all options not considered
308 stable. All options that are known to trash data (experimental write-
309 support for file-systems, for instance) should be denoted (DANGEROUS), other
310 experimental options should be denoted (EXPERIMENTAL).
311
312
313                 Chapter 10: Data structures
314
315 Data structures that have visibility outside the single-threaded
316 environment they are created and destroyed in should always have
317 reference counts.  In the kernel, garbage collection doesn't exist (and
318 outside the kernel garbage collection is slow and inefficient), which
319 means that you absolutely _have_ to reference count all your uses.
320
321 Reference counting means that you can avoid locking, and allows multiple
322 users to have access to the data structure in parallel - and not having
323 to worry about the structure suddenly going away from under them just
324 because they slept or did something else for a while.
325
326 Note that locking is _not_ a replacement for reference counting.
327 Locking is used to keep data structures coherent, while reference
328 counting is a memory management technique.  Usually both are needed, and
329 they are not to be confused with each other.
330
331 Many data structures can indeed have two levels of reference counting,
332 when there are users of different "classes".  The subclass count counts
333 the number of subclass users, and decrements the global count just once
334 when the subclass count goes to zero.
335
336 Examples of this kind of "multi-level-reference-counting" can be found in
337 memory management ("struct mm_struct": mm_users and mm_count), and in
338 filesystem code ("struct super_block": s_count and s_active).
339
340 Remember: if another thread can find your data structure, and you don't
341 have a reference count on it, you almost certainly have a bug.
342
343
344                 Chapter 11: Macros, Enums, Inline functions and RTL
345
346 Names of macros defining constants and labels in enums are capitalized.
347
348 #define CONSTANT 0x12345
349
350 Enums are preferred when defining several related constants.
351
352 CAPITALIZED macro names are appreciated but macros resembling functions
353 may be named in lower case.
354
355 Generally, inline functions are preferable to macros resembling functions.
356
357 Macros with multiple statements should be enclosed in a do - while block:
358
359 #define macrofun(a, b, c)                       \
360         do {                                    \
361                 if (a == 5)                     \
362                         do_this(b, c);          \
363         } while (0)
364
365 Things to avoid when using macros:
366
367 1) macros that affect control flow:
368
369 #define FOO(x)                                  \
370         do {                                    \
371                 if (blah(x) < 0)                \
372                         return -EBUGGERED;      \
373         } while(0)
374
375 is a _very_ bad idea.  It looks like a function call but exits the "calling"
376 function; don't break the internal parsers of those who will read the code.
377
378 2) macros that depend on having a local variable with a magic name:
379
380 #define FOO(val) bar(index, val)
381
382 might look like a good thing, but it's confusing as hell when one reads the
383 code and it's prone to breakage from seemingly innocent changes.
384
385 3) macros with arguments that are used as l-values: FOO(x) = y; will
386 bite you if somebody e.g. turns FOO into an inline function.
387
388 4) forgetting about precedence: macros defining constants using expressions
389 must enclose the expression in parentheses. Beware of similar issues with
390 macros using parameters.
391
392 #define CONSTANT 0x4000
393 #define CONSTEXP (CONSTANT | 3)
394
395 The cpp manual deals with macros exhaustively. The gcc internals manual also
396 covers RTL which is used frequently with assembly language in the kernel.
397
398
399                 Chapter 12: Printing kernel messages
400
401 Kernel developers like to be seen as literate. Do mind the spelling
402 of kernel messages to make a good impression. Do not use crippled
403 words like "dont" and use "do not" or "don't" instead.
404
405 Kernel messages do not have to be terminated with a period.
406
407 Printing numbers in parentheses (%d) adds no value and should be avoided.
408
409
410                 Chapter 13: References
411
412 The C Programming Language, Second Edition
413 by Brian W. Kernighan and Dennis M. Ritchie.
414 Prentice Hall, Inc., 1988.
415 ISBN 0-13-110362-8 (paperback), 0-13-110370-9 (hardback).
416 URL: http://cm.bell-labs.com/cm/cs/cbook/
417
418 The Practice of Programming
419 by Brian W. Kernighan and Rob Pike.
420 Addison-Wesley, Inc., 1999.
421 ISBN 0-201-61586-X.
422 URL: http://cm.bell-labs.com/cm/cs/tpop/
423
424 GNU manuals - where in compliance with K&R and this text - for cpp, gcc,
425 gcc internals and indent, all available from http://www.gnu.org
426
427 WG14 is the international standardization working group for the programming
428 language C, URL: http://std.dkuug.dk/JTC1/SC22/WG14/
429
430 --
431 Last updated on 16 February 2004 by a community effort on LKML.