1 | /** @file
|
---|
2 | *
|
---|
3 | * VBox - Coding Guidelines.
|
---|
4 | */
|
---|
5 |
|
---|
6 | /*
|
---|
7 | * Copyright (C) 2006-2007 innotek GmbH
|
---|
8 | *
|
---|
9 | * innotek GmbH confidential
|
---|
10 | * All rights reserved
|
---|
11 | */
|
---|
12 |
|
---|
13 | /** @page pg_vbox_guideline VBox Coding Guidelines
|
---|
14 | *
|
---|
15 | * The VBox Coding guidelines are followed by all of VBox with the exception of
|
---|
16 | * the GUI and qemu. The GUI is using something close to the Qt style. Qemu is
|
---|
17 | * using whatever the frenchman does.
|
---|
18 | *
|
---|
19 | * There are a few compulsory rules and a bunch of optional ones. The following
|
---|
20 | * sections will describe these in details. In addition there is a section of
|
---|
21 | * Subversion 'rules'.
|
---|
22 | *
|
---|
23 | *
|
---|
24 | *
|
---|
25 | * @section sec_vbox_guideline_compulsory Compulsory
|
---|
26 | *
|
---|
27 | *
|
---|
28 | * - Use RT and VBOX types.
|
---|
29 | *
|
---|
30 | * - Use Runtime functions.
|
---|
31 | *
|
---|
32 | * - Use the standard bool, uintptr_t, intptr_t and [u]int[1-9+]_t types.
|
---|
33 | *
|
---|
34 | * - Avoid using plain unsigned and int.
|
---|
35 | *
|
---|
36 | * - Use static wherever possible. This makes the namespace less polluted
|
---|
37 | * and avoid nasty name clash problems which can occure, especially on
|
---|
38 | * Unix like systems. (1)
|
---|
39 | *
|
---|
40 | * - Public names are on the form Domain[Subdomain[]]Method using mixed
|
---|
41 | * casing to mark the words. The main domain is all uppercase.
|
---|
42 | * (Think like java, mapping domain and subdomain to packages/classes.)
|
---|
43 | *
|
---|
44 | * - Public names are always declared using the appropriate DECL macro. (2)
|
---|
45 | *
|
---|
46 | * - Internal names starts with a lowercased main domain.
|
---|
47 | *
|
---|
48 | * - Defines are all uppercase and separate words with underscore.
|
---|
49 | * This applies to enum values too.
|
---|
50 | *
|
---|
51 | * - Typedefs are all uppercase and contain no underscores to distinguish
|
---|
52 | * them from defines.
|
---|
53 | *
|
---|
54 | * - Pointer typedefs start with 'P'. If pointer to const value then 'PC'.
|
---|
55 | *
|
---|
56 | * - Function typedefs start with 'FN'. If pointer to one then 'PFN'.
|
---|
57 | *
|
---|
58 | * - All files are case sensitive.
|
---|
59 | *
|
---|
60 | * - Slashes are unix slashes ('/') runtime converts when necessary.
|
---|
61 | *
|
---|
62 | * - char strings are UTF-8.
|
---|
63 | *
|
---|
64 | * - All functions returns VBox status codes. There are three general exceptions
|
---|
65 | * from this:
|
---|
66 | * -# Predicate functions. These are function which are boolean in nature
|
---|
67 | * and usage. They return bool. The function name will include
|
---|
68 | * 'Has', 'Is' or similar.
|
---|
69 | * -# Functions which by nature cannot possibly fail.
|
---|
70 | * These return void.
|
---|
71 | * -# "Get"-functions which return what they ask for.
|
---|
72 | * A get function becomes a "Query" function if there is any
|
---|
73 | * doubt about getting what is ask for.
|
---|
74 | *
|
---|
75 | * - VBox status codes have three subdivisions:
|
---|
76 | * -# Errors, which are VERR_ prefixed and negative.
|
---|
77 | * -# Warnings, which are VWRN_ prefixed and positive.
|
---|
78 | * -# Informational, which are VINF_ prefixed and positive.
|
---|
79 | *
|
---|
80 | * - Platform/OS operation are generalized and put in the IPRT.
|
---|
81 | *
|
---|
82 | * - Other useful constructs are also put in the IPRT.
|
---|
83 | *
|
---|
84 | * - The code shall not cause compiler warnings. Check this with on ALL
|
---|
85 | * the platforms.
|
---|
86 | *
|
---|
87 | * - All files have file headers with $Id and a file tag which describes
|
---|
88 | * the file in a sentence or two.
|
---|
89 | * Note: Remember to enable keyword expansion when adding files to svn.
|
---|
90 | *
|
---|
91 | * - All public functions are fully documented in Doxygen style using the
|
---|
92 | * javadoc dialect (using the 'at' insdead of the 'slash' as commandprefix.)
|
---|
93 | *
|
---|
94 | * - All structures in header files are described, including all of their
|
---|
95 | * members.
|
---|
96 | *
|
---|
97 | * - All modules have a documentation 'page' in the main source file which
|
---|
98 | * describes the intent and actual implementation.
|
---|
99 | *
|
---|
100 | * - Code which is doing things that are not immediatly comprehendable
|
---|
101 | * shall include explanatory comments!
|
---|
102 | *
|
---|
103 | * - Documentation and comments are kept up to date.
|
---|
104 | *
|
---|
105 | * - Headers in /include/VBox shall not contain any slash-slash C++ comments,
|
---|
106 | * only ansi C comments!
|
---|
107 | *
|
---|
108 | *
|
---|
109 | * (1) It is common practice on Unix to have a single symbol namespace for an
|
---|
110 | * entire process. If one is careless symbols might be resolved in a
|
---|
111 | * different way that one expects, leading to weird problems.
|
---|
112 | *
|
---|
113 | * (2) This is common practice among most projects dealing with modules in
|
---|
114 | * shared libraries. The Windows / PE __declspect(import) and
|
---|
115 | * __declspect(export) constructs are the main reason for this.
|
---|
116 | * OTH, we do perhaps have a bit too detailed graining of this in VMM...
|
---|
117 | *
|
---|
118 | *
|
---|
119 | *
|
---|
120 | * @subsection sec_vbox_guideline_compulsory_sub64 64-bit and 32-bit
|
---|
121 | *
|
---|
122 | * Here are some amendments which address 64-bit vs. 32-bit portability issues.
|
---|
123 | *
|
---|
124 | * Some facts first:
|
---|
125 | *
|
---|
126 | * - On 64-bit Windows the type long remains 32-bit. On nearly all other
|
---|
127 | * 64-bit platforms long is 64-bit.
|
---|
128 | *
|
---|
129 | * - On all 64-bit platforms we care about, int is 32-bit, short is 16 bit
|
---|
130 | * and char is 8-bit.
|
---|
131 | * (I don't know about any platforms yet where this isn't true.)
|
---|
132 | *
|
---|
133 | * - size_t, ssize_t, uintptr_t, ptrdiff_t and similar are all 64-bit on
|
---|
134 | * 64-bit platforms. (These are 32-bit on 32-bit platforms.)
|
---|
135 | *
|
---|
136 | * - There is no inline assembly support in the 64-bit Microsoft compilers.
|
---|
137 | *
|
---|
138 | *
|
---|
139 | * Now for the guidelines:
|
---|
140 | *
|
---|
141 | * - Never, ever, use int, long, ULONG, LONG, DWORD or similar to cast a
|
---|
142 | * pointer to integer. Use uintptr_t or intptr_t. If you have to use
|
---|
143 | * NT/Windows types, there is the choice of ULONG_PTR and DWORD_PTR.
|
---|
144 | *
|
---|
145 | * - RT_OS_WINDOWS is defined to indicate Windows. Do not use __WIN32__,
|
---|
146 | * __WIN64__ and __WIN__ because they are all deprecated and schedule
|
---|
147 | * for removal (if not removed already). Do not use the compiler
|
---|
148 | * defined _WIN32, _WIN64, or similar either. The bitness can be
|
---|
149 | * determined by testing ARCH_BITS.
|
---|
150 | * Example:
|
---|
151 | * @code
|
---|
152 | * #ifdef RT_OS_WINDOWS
|
---|
153 | * // call win32/64 api.
|
---|
154 | * #endif
|
---|
155 | * #ifdef RT_OS_WINDOWS
|
---|
156 | * # if ARCH_BITS == 64
|
---|
157 | * // call win64 api.
|
---|
158 | * # else // ARCH_BITS == 32
|
---|
159 | * // call win32 api.
|
---|
160 | * # endif // ARCH_BITS == 32
|
---|
161 | * #else // !RT_OS_WINDOWS
|
---|
162 | * // call posix api
|
---|
163 | * #endif // !RT_OS_WINDOWS
|
---|
164 | * @endcode
|
---|
165 | *
|
---|
166 | * - There are RT_OS_xxx defines for each OS, just like RT_OS_WINDOWS
|
---|
167 | * mentioned above. Use these defines instead of any predefined
|
---|
168 | * compiler stuff or defines from system headers.
|
---|
169 | *
|
---|
170 | * - RT_ARCH_X86 is defined when compiling for the x86 the architecture.
|
---|
171 | * Do not use __x86__, __X86__, __[Ii]386__, __[Ii]586__, or similar
|
---|
172 | * for this purpose.
|
---|
173 | *
|
---|
174 | * - RT_ARCH_AMD64 is defined when compiling for the AMD64 the architecture.
|
---|
175 | * Do not use __AMD64__, __amd64__ or __x64_86__.
|
---|
176 | *
|
---|
177 | * - Take care and use size_t when you have to, esp. when passing a pointer
|
---|
178 | * to a size_t as a parameter.
|
---|
179 | *
|
---|
180 | *
|
---|
181 | *
|
---|
182 | * @section sec_vbox_guideline_optional Optional
|
---|
183 | *
|
---|
184 | * First part is the actual coding style and all the prefixes the second part is
|
---|
185 | * the a bunch of good advice.
|
---|
186 | *
|
---|
187 | *
|
---|
188 | * @subsection sec_vbox_guideline_optional_layout The code layout
|
---|
189 | *
|
---|
190 | * - Curly brackets are not indented.
|
---|
191 | *
|
---|
192 | * - Space before the parenthesis when it comes after a C keyword.
|
---|
193 | *
|
---|
194 | * - No space between argument and parenthesis. Exception for complex
|
---|
195 | * expression.
|
---|
196 | * Example:
|
---|
197 | * @code
|
---|
198 | * if (PATMR3IsPatchGCAddr(pVM, GCPtr))
|
---|
199 | * @endcode
|
---|
200 | *
|
---|
201 | * - The else of an if is always first statement on a line. (No curly
|
---|
202 | * stuff before it!)
|
---|
203 | *
|
---|
204 | * - else and if goes on the same line if no curly stuff is needed around the if.
|
---|
205 | * Example:
|
---|
206 | * @code
|
---|
207 | * if (fFlags & MYFLAGS_1)
|
---|
208 | * fFlags &= ~MYFLAGS_10;
|
---|
209 | * else if (fFlags & MYFLAGS_2)
|
---|
210 | * {
|
---|
211 | * fFlags &= ~MYFLAGS_MASK;
|
---|
212 | * fFlags |= MYFLAGS_5;
|
---|
213 | * }
|
---|
214 | * else if (fFlags & MYFLAGS_3)
|
---|
215 | * @endcode
|
---|
216 | *
|
---|
217 | * - The case is indented from the switch.
|
---|
218 | *
|
---|
219 | * - If a case needs curly brackets they contain the entire case, are not
|
---|
220 | * indented from the case, and the break or return is placed inside them.
|
---|
221 | * Example:
|
---|
222 | * @code
|
---|
223 | * switch (pCur->eType)
|
---|
224 | * {
|
---|
225 | * case PGMMAPPINGTYPE_PAGETABLES:
|
---|
226 | * {
|
---|
227 | * unsigned iPDE = pCur->GCPtr >> PGDIR_SHIFT;
|
---|
228 | * unsigned iPT = (pCur->GCPtrEnd - pCur->GCPtr) >> PGDIR_SHIFT;
|
---|
229 | * while (iPT-- > 0)
|
---|
230 | * if (pPD->a[iPDE + iPT].n.u1Present)
|
---|
231 | * return VERR_HYPERVISOR_CONFLICT;
|
---|
232 | * break;
|
---|
233 | * }
|
---|
234 | * }
|
---|
235 | * @endcode
|
---|
236 | *
|
---|
237 | * - In a do while construction, the while is on the same line as the
|
---|
238 | * closing bracket if any are used.
|
---|
239 | * Example:
|
---|
240 | * @code
|
---|
241 | * do
|
---|
242 | * {
|
---|
243 | * stuff;
|
---|
244 | * i--;
|
---|
245 | * } while (i > 0);
|
---|
246 | * @endcode
|
---|
247 | *
|
---|
248 | * - Comments are in C style. C++ style comments are used for temporary
|
---|
249 | * disabling a few lines of code.
|
---|
250 | *
|
---|
251 | * - Sligtly complex boolean expressions are splitt into multiple lines,
|
---|
252 | * putting the operators first on the line and indenting it all according
|
---|
253 | * to the nesting of the expression. The purpose is to make it as easy as
|
---|
254 | * possible to read.
|
---|
255 | * Example:
|
---|
256 | * @code
|
---|
257 | * if ( RT_SUCCESS(rc)
|
---|
258 | * || (fFlags & SOME_FLAG))
|
---|
259 | * @endcode
|
---|
260 | *
|
---|
261 | * - No unnecessary parentheses in expressions (just don't over do this
|
---|
262 | * so that gcc / msc starts bitching). Find a correct C/C++ operator
|
---|
263 | * precedence table if needed.
|
---|
264 | *
|
---|
265 | *
|
---|
266 | * @subsection sec_vbox_guideline_optional_prefix Variable / Member Prefixes
|
---|
267 | *
|
---|
268 | * - The 'g_' (or 'g') prefix means a global variable, either on file or module level.
|
---|
269 | *
|
---|
270 | * - The 's_' (or 's') prefix means a static variable inside a function or class.
|
---|
271 | *
|
---|
272 | * - The 'm_' (or 'm') prefix means a class data member.
|
---|
273 | *
|
---|
274 | * - The 'p' prefix means pointer. For instance 'pVM' is pointer to VM.
|
---|
275 | *
|
---|
276 | * - The 'a' prefix means array. For instance 'aPages' could be read as array
|
---|
277 | * of pages.
|
---|
278 | *
|
---|
279 | * - The 'c' prefix means count. For instance 'cbBlock' could be read, count
|
---|
280 | * of bytes in block.
|
---|
281 | *
|
---|
282 | * - The 'off' prefix means offset.
|
---|
283 | *
|
---|
284 | * - The 'i' or 'idx' prefixes usually means index. Although the 'i' one can
|
---|
285 | * sometimes just mean signed integer.
|
---|
286 | *
|
---|
287 | * - The 'e' (or 'enm') prefix means enum.
|
---|
288 | *
|
---|
289 | * - The 'u' prefix usually means unsigned integer. Exceptions follows.
|
---|
290 | *
|
---|
291 | * - The 'u[1-9]+' prefix means a fixed bit size variable. Frequently used
|
---|
292 | * with the uint[1-9]+_t types and with bitfields.
|
---|
293 | *
|
---|
294 | * - The 'b' prefix means byte or bytes.
|
---|
295 | *
|
---|
296 | * - The 'f' prefix means flags. Flags are unsigned integers of some kind or bools.
|
---|
297 | *
|
---|
298 | * - The 'ch' prefix means a char, the (signed) char type.
|
---|
299 | *
|
---|
300 | * - The 'wc' prefix means a wide/windows char, the RTUTF16 type.
|
---|
301 | *
|
---|
302 | * - The 'uc' prefix means a Unicode Code point, the RTUNICP type.
|
---|
303 | *
|
---|
304 | * - The 'uch' prefix means unsigned char. It's rarely used.
|
---|
305 | *
|
---|
306 | * - The 'sz' prefix means zero terminated character string (array of chars). (UTF-8)
|
---|
307 | *
|
---|
308 | * - The 'wsz' prefix means zero terminated wide/windows character string (array of RTUTF16).
|
---|
309 | *
|
---|
310 | * - The 'usz' prefix means zero terminated Unicode string (array of RTUNICP).
|
---|
311 | *
|
---|
312 | * - The 'pfn' prefix means pointer to function. Common usage is 'pfnCallback'
|
---|
313 | * and such like.
|
---|
314 | *
|
---|
315 | *
|
---|
316 | * @subsection sec_vbox_guideline_optional_misc Misc / Advice / Stuff
|
---|
317 | *
|
---|
318 | * - When writing code think as the reader.
|
---|
319 | *
|
---|
320 | * - When writing code think as the compiler.
|
---|
321 | *
|
---|
322 | * - When reading code think as that it's fully of bugs - find them and fix them.
|
---|
323 | *
|
---|
324 | * - Pointer within range tests like:
|
---|
325 | * @code
|
---|
326 | * if ((uintptr_t)pv >= (uintptr_t)pvBase && (uintptr_t)pv < (uintptr_t)pvBase + cbRange)
|
---|
327 | * @endcode
|
---|
328 | * Can also be written as (assuming cbRange unsigned):
|
---|
329 | * @code
|
---|
330 | * if ((uintptr_t)pv - (uintptr_t)pvBase < cbRange)
|
---|
331 | * @endcode
|
---|
332 | * Which is shorter and potentially faster. (1)
|
---|
333 | *
|
---|
334 | * - Avoid unnecessary casting. All pointers automatically casts down to void *,
|
---|
335 | * at least for non class instance pointers.
|
---|
336 | *
|
---|
337 | * - It's very very bad practise to write a function larger than a
|
---|
338 | * screen full (1024x768) without any comprehendable and explaining comments.
|
---|
339 | *
|
---|
340 | * - More to come....
|
---|
341 | *
|
---|
342 | *
|
---|
343 | * (1) Important, be very careful with the casting. In particular, note that
|
---|
344 | * a compiler might treat pointers as signed (IIRC).
|
---|
345 | *
|
---|
346 | *
|
---|
347 | *
|
---|
348 | *
|
---|
349 | * @section sec_vbox_guideline_warnings Compiler Warnings
|
---|
350 | *
|
---|
351 | * The code should when possible compile on all platforms and compilers without any
|
---|
352 | * warnings. That's a nice idea, however, if it means making the code harder to read,
|
---|
353 | * less portable, unreliable or similar, the warning should not be fixed.
|
---|
354 | *
|
---|
355 | * Some of the warnings can seem kind of innocent at first glance. So, let's take the
|
---|
356 | * most common ones and explain them.
|
---|
357 | *
|
---|
358 | * @subsection sec_vbox_guideline_warnings_signed_unsigned_compare Signed / Unsigned Compare
|
---|
359 | *
|
---|
360 | * GCC says: "warning: comparison between signed and unsigned integer expressions"
|
---|
361 | * MSC says: "warning C4018: '<|<=|==|>=|>' : signed/unsigned mismatch"
|
---|
362 | *
|
---|
363 | * The following example will not output what you expect:
|
---|
364 | @code
|
---|
365 | #include <stdio.h>
|
---|
366 | int main()
|
---|
367 | {
|
---|
368 | signed long a = -1;
|
---|
369 | unsigned long b = 2294967295;
|
---|
370 | if (a < b)
|
---|
371 | printf("%ld < %lu: true\n", a, b);
|
---|
372 | else
|
---|
373 | printf("%ld < %lu: false\n", a, b);
|
---|
374 | return 0;
|
---|
375 | }
|
---|
376 | @endcode
|
---|
377 | * If I understood it correctly, the compiler will convert a to an unsigned long before
|
---|
378 | * doing the compare.
|
---|
379 | *
|
---|
380 | *
|
---|
381 | * @section sec_vbox_guideline_svn Subversion Commit Rules
|
---|
382 | *
|
---|
383 | *
|
---|
384 | * Before checking in:
|
---|
385 | *
|
---|
386 | * - Check Tinderbox and make sure the tree is green across all platforms. If it's
|
---|
387 | * red on a platform, don't check in. If you want, warn in the \#vbox channel and
|
---|
388 | * help make the responsible person fix it.
|
---|
389 | * NEVER CHECK IN TO A BROKEN BUILD.
|
---|
390 | *
|
---|
391 | * - When checking in keep in mind that a commit is atomical and that the Tinderbox and
|
---|
392 | * developers are constantly checking out the tree. Therefore do not split up the
|
---|
393 | * commit unless it's into 100% indepentant parts. If you need to split it up in order
|
---|
394 | * to have sensible commit comments, make the sub-commits as rapid as possible.
|
---|
395 | *
|
---|
396 | * - Make sure you add an entry to the ChangeLog file.
|
---|
397 | *
|
---|
398 | *
|
---|
399 | * After checking in:
|
---|
400 | *
|
---|
401 | * - After checking-in, you watch Tinderbox until your check-ins clear. You do not
|
---|
402 | * go home. You do not sleep. You do not log out or experiment with drugs. You do
|
---|
403 | * not become unavailable. If you break the tree, add a comment saying that you're
|
---|
404 | * fixing it. If you can't fix it and need help, ask in the \#innotek channel or back
|
---|
405 | * out the change.
|
---|
406 | *
|
---|
407 | * (Inspired by mozilla tree rules.)
|
---|
408 | */
|
---|
409 |
|
---|