VBox-CodingGuidelines.cpp@ 80070

Last change on this file since 80070 was 79974, checked in by vboxsync, 5 years ago
VBox-CodingGuidelines.cpp: Added alternative for typedefs.
Property svn:eol-style set to `native` Property svn:keywords set to `Author Date Id Revision`
File size: 36.5 KB

Line
1	/* $Id: VBox-CodingGuidelines.cpp 79974 2019-07-25 09:01:39Z vboxsync $ */
2	/** @file
3	* VBox - Coding Guidelines.
4	*/
5
6	/*
7	* Copyright (C) 2006-2019 Oracle Corporation
8	*
9	* This file is part of VirtualBox Open Source Edition (OSE), as
10	* available from http://www.virtualbox.org. This file is free software;
11	* you can redistribute it and/or modify it under the terms of the GNU
12	* General Public License (GPL) as published by the Free Software
13	* Foundation, in version 2 as it comes in the "COPYING" file of the
14	* VirtualBox OSE distribution. VirtualBox OSE is distributed in the
15	* hope that it will be useful, but WITHOUT ANY WARRANTY of any kind.
16	*/
17
18	/** @page pg_vbox_guideline VBox Coding Guidelines
19	*
20	* The compulsory sections of these guidelines are to be followed in all of the
21	* VBox sources. Please note that local guidelines in parts of the VBox source
22	* tree may promote the optional ones to compulsory status. The VBox tree also
23	* contains some 3rd party sources where it is good to follow the local coding
24	* style while keeping these guidelines in mind.
25	*
26	* Contents:
27	* - @ref sec_vbox_guideline_compulsory
28	* - @ref sec_vbox_guideline_compulsory_sub64
29	* - @ref sec_vbox_guideline_compulsory_cppmain
30	* - @ref sec_vbox_guideline_compulsory_cppqtgui
31	* - @ref sec_vbox_guideline_compulsory_xslt
32	* - @ref sec_vbox_guideline_compulsory_doxygen
33	* - @ref sec_vbox_guideline_compulsory_guest
34	* - @ref sec_vbox_guideline_optional
35	* - @ref sec_vbox_guideline_optional_layout
36	* - @ref sec_vbox_guideline_optional_prefix
37	* - @ref sec_vbox_guideline_optional_misc
38	* - @ref sec_vbox_guideline_warnings
39	* - @ref sec_vbox_guideline_warnings_signed_unsigned_compare
40	* - @ref sec_vbox_guideline_svn
41	*
42	* Local guidelines overrides:
43	* - src/VBox/VMM/: @ref pg_vmm_guideline (src/VBox/VMM/Docs-CodingGuidelines.cpp)
44	* - src/VBox/ValidationKit/: @ref pg_validationkit_guideline (src/VBox/ValidationKit/ValidationKitCodingGuidelines.cpp)
45	* - src/VBox/Runtime/: All of @ref sec_vbox_guideline_optional is mandatory.
46	* - src/VBox/Main/: @ref sec_vbox_guideline_compulsory_cppmain
47	* - src/VBox/Frontends/VirtualBox/: @ref sec_vbox_guideline_compulsory_cppqtgui
48	*
49	*
50	* @section sec_vbox_guideline_compulsory Compulsory
51	*
52	* <ul>
53	*
54	* <li> The indentation size is 4 chars.
55	*
56	* <li> Tabs are only ever used in makefiles.
57	*
58	* <li> Use RT and VBOX types.
59	*
60	* <li> Use Runtime functions.
61	*
62	* <li> Use the standard bool, uintptr_t, intptr_t and [u]int[1-9+]_t types.
63	*
64	* <li> Avoid using plain unsigned and int.
65	*
66	* <li> Use static wherever possible. This makes the namespace less polluted
67	* and avoids nasty name clash problems which can occur, especially on
68	* Unix-like systems. (1) It also simplifies locating callers when
69	* changing it (single source file vs entire VBox tree).
70	*
71	* <li> Public names are of the form Domain[Subdomain[]]Method, using mixed
72	* casing to mark the words. The main domain is all uppercase.
73	* (Think like java, mapping domain and subdomain to packages/classes.)
74	*
75	* <li> Public names are always declared using the appropriate DECL macro. (2)
76	*
77	* <li> Internal names starts with a lowercased main domain.
78	*
79	* <li> Defines are all uppercase and separate words with underscore.
80	* This applies to enum values too.
81	*
82	* <li> Typedefs are all uppercase and contain no underscores to distinguish
83	* them from defines. Alternatively, all uppercase, separate words with
84	* underscores and ending with '_T'. The latter is not allowed in IPRT.
85	*
86	* <li> Pointer typedefs start with 'P'. If pointer to const then 'PC'.
87	*
88	* <li> Function typedefs start with 'FN'. If pointer to FN then 'PFN'.
89	*
90	* <li> All files are case sensitive.
91	*
92	* <li> Slashes are unix slashes ('/') runtime converts when necessary.
93	*
94	* <li> char strings are UTF-8.
95	*
96	* <li> Strings from any external source must be treated with utmost care as
97	* they do not have to be valid UTF-8. Only trust internal strings.
98	*
99	* <li> All functions return VBox status codes. There are three general
100	* exceptions from this:
101	*
102	* <ol>
103	* <li>Predicate functions. These are function which are boolean in
104	* nature and usage. They return bool. The function name will
105	* include 'Has', 'Is' or similar.
106	* <li>Functions which by nature cannot possibly fail.
107	* These return void.
108	* <li>"Get"-functions which return what they ask for.
109	* A get function becomes a "Query" function if there is any
110	* doubt about getting what is ask for.
111	* </ol>
112	*
113	* <li> VBox status codes have three subdivisions:
114	* <ol>
115	* <li> Errors, which are VERR_ prefixed and negative.
116	* <li> Warnings, which are VWRN_ prefixed and positive.
117	* <li> Informational, which are VINF_ prefixed and positive.
118	* </ol>
119	*
120	* <li> Platform/OS operation are generalized and put in the IPRT.
121	*
122	* <li> Other useful constructs are also put in the IPRT.
123	*
124	* <li> The code shall not cause compiler warnings. Check this on ALL
125	* the platforms.
126	*
127	* <li> The use of symbols leading with single or double underscores is
128	* forbidden as that intrudes on reserved compiler/system namespace. (3)
129	*
130	* <li> All files have file headers with $Id and a file tag which describes
131	* the file in a sentence or two.
132	* Note: Use the svn-ps.cmd/svn-ps.sh utility with the -a option to add
133	* new sources with keyword expansion and exporting correctly
134	* configured.
135	*
136	* <li> All public functions are fully documented in Doxygen style using the
137	* javadoc dialect (using the 'at' instead of the 'slash' as
138	* commandprefix.)
139	*
140	* <li> All structures in header files are described, including all their
141	* members. (Doxygen style, of course.)
142	*
143	* <li> All modules have a documentation '\@page' in the main source file
144	* which describes the intent and actual implementation.
145	*
146	* <li> Code which is doing things that are not immediately comprehensible
147	* shall include explanatory comments.
148	*
149	* <li> Documentation and comments are kept up to date.
150	*
151	* <li> Headers in /include/VBox shall not contain any slash-slash C++
152	* comments, only ANSI C comments!
153	*
154	* <li> Comments on \#else indicates what begins while the comment on a
155	* \#endif indicates what ended. Only add these when there are more than
156	* a few lines (6-10) of \#ifdef'ed code, otherwise they're just clutter.
157	*
158	* <li> \#ifdefs around a single function shall be tight, i.e. no empty
159	* lines between it and the function documentation and body.
160	*
161	* <li> \#ifdefs around more than one function shall be relaxed, i.e. leave at
162	* least one line before the first function's documentation comment and
163	* one line after the end of the last function.
164	*
165	* <li> No 'else' after if block ending with 'return', 'break', or 'continue'.
166	*
167	* <li> The term 'last' is inclusive, whereas the term 'end' is exclusive.
168	*
169	* <li> Go through all of this: https://www.slideshare.net/olvemaudal/deep-c/
170	*
171	* <li> Avoid throwing exceptions, always prefer returning statuses.
172	* Crappy exception handling is rewared by a glass of water in the face.
173	*
174	* </ul>
175	*
176	* (1) It is common practice on Unix to have a single symbol namespace for an
177	* entire process. If one is careless symbols might be resolved in a
178	* different way that one expects, leading to weird problems.
179	*
180	* (2) This is common practice among most projects dealing with modules in
181	* shared libraries. The Windows / PE __declspect(import) and
182	* __declspect(export) constructs are the main reason for this.
183	* OTOH, we do perhaps have a bit too detailed graining of this in VMM...
184	*
185	* (3) There are guys out there grepping public sources for symbols leading with
186	* single and double underscores as well as gotos and other things
187	* considered bad practice. They'll post statistics on how bad our sources
188	* are on some mailing list, forum or similar.
189	*
190	*
191	* @subsection sec_vbox_guideline_compulsory_sub64 64-bit and 32-bit
192	*
193	* Here are some amendments which address 64-bit vs. 32-bit portability issues.
194	*
195	* Some facts first:
196	*
197	* <ul>
198	*
199	* <li> On 64-bit Windows the type long remains 32-bit. On nearly all other
200	* 64-bit platforms long is 64-bit.
201	*
202	* <li> On all 64-bit platforms we care about, int is 32-bit, short is 16 bit
203	* and char is 8-bit.
204	* (I don't know about any platforms yet where this isn't true.)
205	*
206	* <li> size_t, ssize_t, uintptr_t, ptrdiff_t and similar are all 64-bit on
207	* 64-bit platforms. (These are 32-bit on 32-bit platforms.)
208	*
209	* <li> There is no inline assembly support in the 64-bit Microsoft compilers.
210	*
211	* </ul>
212	*
213	* Now for the guidelines:
214	*
215	* <ul>
216	*
217	* <li> Never, ever, use int, long, ULONG, LONG, DWORD or similar to cast a
218	* pointer to integer. Use uintptr_t or intptr_t. If you have to use
219	* NT/Windows types, there is the choice of ULONG_PTR and DWORD_PTR.
220	*
221	* <li> Avoid where ever possible the use of the types 'long' and 'unsigned
222	* long' as these differs in size between windows and the other hosts
223	* (see above).
224	*
225	* <li> RT_OS_WINDOWS is defined to indicate Windows. Do not use __WIN32__,
226	* __WIN64__ and __WIN__ because they are all deprecated and scheduled
227	* for removal (if not removed already). Do not use the compiler
228	* defined _WIN32, _WIN64, or similar either. The bitness can be
229	* determined by testing ARCH_BITS.
230	* Example:
231	* @code
232	* #ifdef RT_OS_WINDOWS
233	* // call win32/64 api.
234	* #endif
235	* #ifdef RT_OS_WINDOWS
236	* # if ARCH_BITS == 64
237	* // call win64 api.
238	* # else // ARCH_BITS == 32
239	* // call win32 api.
240	* # endif // ARCH_BITS == 32
241	* #else // !RT_OS_WINDOWS
242	* // call posix api
243	* #endif // !RT_OS_WINDOWS
244	* @endcode
245	*
246	* <li> There are RT_OS_xxx defines for each OS, just like RT_OS_WINDOWS
247	* mentioned above. Use these defines instead of any predefined
248	* compiler stuff or defines from system headers.
249	*
250	* <li> RT_ARCH_X86 is defined when compiling for the x86 the architecture.
251	* Do not use __x86__, __X86__, __[Ii]386__, __[Ii]586__, or similar
252	* for this purpose.
253	*
254	* <li> RT_ARCH_AMD64 is defined when compiling for the AMD64 architecture.
255	* Do not use __AMD64__, __amd64__ or __x64_86__.
256	*
257	* <li> Take care and use size_t when you have to, esp. when passing a pointer
258	* to a size_t as a parameter.
259	*
260	* <li> Be wary of type promotion to (signed) integer. For example the
261	* following will cause u8 to be promoted to int in the shift, and then
262	* sign extended in the assignment 64-bit:
263	* @code
264	* uint8_t u8 = 0xfe;
265	* uint64_t u64 = u8 << 24;
266	* // u64 == 0xfffffffffe000000
267	* @endcode
268	*
269	* </ul>
270	*
271	* @subsection sec_vbox_guideline_compulsory_cppmain C++ guidelines for Main
272	*
273	* Since the Main API code is a large amount of C++ code, it is allowed but
274	* not required to use C++ style comments (as permanent comments, beyond the
275	* temporary use allowed by the general coding guideline). This is a weak
276	* preference, i.e. large scale comment style changes are not encouraged.
277	*
278	* Main is currently (2009) full of hard-to-maintain code that uses complicated
279	* templates. The new mid-term goal for Main is to have less custom templates
280	* instead of more for the following reasons:
281	*
282	* <ul>
283	*
284	* <li> Template code is harder to read and understand. Custom templates create
285	* territories which only the code writer understands.
286	*
287	* <li> Errors in using templates create terrible C++ compiler messages.
288	*
289	* <li> Template code is really hard to look at in a debugger.
290	*
291	* <li> Templates slow down the compiler a lot.
292	*
293	* </ul>
294	*
295	* In particular, the following bits should be considered deprecated and should
296	* NOT be used in new code:
297	*
298	* <ul>
299	*
300	* <li> everything in include/iprt/cpputils.h (auto_ref_ptr, exception_trap_base,
301	* char_auto_ptr and friends)
302	*
303	* </ul>
304	*
305	* Generally, in many cases, a simple class with a proper destructor can achieve
306	* the same effect as a 1,000-line template include file, and the code is
307	* much more accessible that way.
308	*
309	* Using standard STL templates like std::list, std::vector and std::map is OK.
310	* Exceptions are:
311	*
312	* <ul>
313	*
314	* <li> Guest Additions because we don't want to link against libstdc++ there.
315	*
316	* <li> std::string should not be used because we have iprt::MiniString and
317	* com::Utf8Str which can convert efficiently with COM's UTF-16 strings.
318	*
319	* <li> std::auto_ptr<> in general; that part of the C++ standard is just broken.
320	* Write a destructor that calls delete.
321	*
322	* </ul>
323	*
324	* @subsection sec_vbox_guideline_compulsory_cppqtgui C++ guidelines for the Qt GUI
325	*
326	* The Qt GUI is currently (2010) on its way to become more compatible to the
327	* rest of VirtualBox coding style wise. From now on, all the coding style
328	* rules described in this file are also mandatory for the Qt GUI. Additionally
329	* the following rules should be respected:
330	*
331	* <ul>
332	*
333	* <li> GUI classes which correspond to GUI tasks should be prefixed by UI (no VBox anymore)
334	*
335	* <li> Classes which extents some of the Qt classes should be prefix by QI
336	*
337	* <li> General task classes should be prefixed by C
338	*
339	* <li> Slots are prefixed by slt -> sltName
340	*
341	* <li> Signals are prefixed by sig -> sigName
342	*
343	* <li> Use Qt classes for lists, strings and so on, the use of STL classes should
344	* be avoided
345	*
346	* <li> All files like .cpp, .h, .ui, which belong together are located in the
347	* same directory and named the same
348	*
349	* </ul>
350	*
351	*
352	* @subsection sec_vbox_guideline_compulsory_xslt XSLT
353	*
354	* XSLT (eXtensible Stylesheet Language Transformations) is used quite a bit in
355	* the Main API area of VirtualBox to generate sources and bindings to that API.
356	* There are a couple of common pitfalls worth mentioning:
357	*
358	* <ul>
359	*
360	* <li> Never do repeated //interface[\@name=...] and //enum[\@name=...] lookups
361	* because they are expensive. Instead delcare xsl:key elements for these
362	* searches and do the lookup using the key() function. xsltproc uses
363	* (per current document) hash tables for each xsl:key, i.e. very fast.
364	*
365	* <li> When output type is 'text' make sure to call xsltprocNewlineOutputHack
366	* from typemap-shared.inc.xsl every few KB of output, or xsltproc will
367	* end up wasting all the time reallocating the output buffer.
368	*
369	* </ul>
370	*
371	*
372	* @subsection sec_vbox_guideline_compulsory_doxygen Doxygen Comments
373	*
374	* As mentioned above, we shall use doxygen/javadoc style commenting of public
375	* functions, typedefs, classes and such. It is mandatory to use this style
376	* everywhere!
377	*
378	* A couple of hints on how to best write doxygen comments:
379	*
380	* <ul>
381	*
382	* <li> A good class, method, function, structure or enum doxygen comment
383	* starts with a one line sentence giving a brief description of the
384	* item. Details comes in a new paragraph (after blank line).
385	*
386	* <li> Except for list generators like \@todo, \@cfgm, \@gcfgm and others,
387	* all doxygen comments are related to things in the code. So, for
388	* instance you DO NOT add a doxygen \@note comment in the middle of a
389	* because you've got something important to note, you add a normal
390	* comment like 'Note! blah, very importan blah!'
391	*
392	* <li> We do NOT use TODO/XXX/BUGBUG or similar markers in the code to flag
393	* things needing fixing later, we always use \@todo doxygen comments.
394	*
395	* <li> There is no colon after the \@todo. And it is ALWAYS in a doxygen
396	* comment.
397	*
398	* <li> The \@retval tag is used to explain status codes a method/function may
399	* returns. It is not used to describe output parameters, that is done
400	* using the \@param or \@param[out] tag.
401	*
402	* </ul>
403	*
404	* See https://www.stack.nl/~dimitri/doxygen/manual/index.html for the official
405	* doxygen documention.
406	*
407	*
408	*
409	* @subsection sec_vbox_guideline_compulsory_guest Handling of guest input
410	*
411	* First, guest input should ALWAYS be consider to be TOXIC and constructed with
412	* MALICIOUS intent! Max paranoia level!
413	*
414	* Second, when getting inputs from memory shared with the guest, be EXTREMELY
415	* careful to not re-read input from shared memory after validating it, because
416	* that will create TOCTOU problems. So, after reading input from shared memory
417	* always use the RT_UNTRUSTED_NONVOLATILE_COPY_FENCE() macro. For more details
418	* on TOCTOU: https://en.wikipedia.org/wiki/Time_of_check_to_time_of_use
419	*
420	* Thirdly, considering the recent speculation side channel issues, spectre v1
421	* in particular, we would like to be ready for future screwups. This means
422	* having input validation in a separate block of code that ends with one (or
423	* more) RT_UNTRUSTED_VALIDATED_FENCE().
424	*
425	* So the rules:
426	*
427	* <ul>
428	*
429	* <li> Mark all pointers to shared memory with RT_UNTRUSTED_VOLATILE_GUEST.
430	*
431	* <li> Copy volatile data into local variables or heap before validating
432	* them (see RT_COPY_VOLATILE() and RT_BCOPY_VOLATILE().
433	*
434	* <li> Place RT_UNTRUSTED_NONVOLATILE_COPY_FENCE() after a block copying
435	* volatile data.
436	*
437	* <li> Always validate untrusted inputs in a block ending with a
438	* RT_UNTRUSTED_VALIDATED_FENCE().
439	*
440	* <li> Use the ASSERT_GUEST_XXXX macros from VBox/AssertGuest.h to validate
441	* guest input. (Do NOT use iprt/assert.h macros.)
442	*
443	* <li> Validation of an input B may require using another input A to look up
444	* some data, in which case its necessary to insert an
445	* RT_UNTRUSTED_VALIDATED_FENCE() after validating A and before A is used
446	* for the lookup.
447	*
448	* For example A is a view identifier, idView, and B is an offset into
449	* the view's framebuffer area, offView. To validate offView (B) it is
450	* necessary to get the size of the views framebuffer region:
451	* @code
452	* uint32_t const idView = pReq->idView; // A
453	* uint32_t const offView = pReq->offView; // B
454	* RT_UNTRUSTED_NONVOLATILE_COPY_FENCE();
455	*
456	* ASSERT_GUEST_RETURN(idView < pThis->cView,
457	* VERR_INVALID_PARAMETER);
458	* RT_UNTRUSTED_VALIDATED_FENCE();
459	* const MYVIEW *pView = &pThis->aViews[idView];
460	* ASSERT_GUEST_RETURN(offView < pView->cbFramebufferArea,
461	* VERR_OUT_OF_RANGE);
462	* RT_UNTRUSTED_VALIDATED_FENCE();
463	* @endcode
464	*
465	* <li> Take care to make sure input check are not subject to integer overflow problems.
466	*
467	* For instance when validating an area, you must not just add cbDst + offDst
468	* and check against pThis->offEnd or something like that. Rather do:
469	* @code
470	* uint32_t const offDst = pReq->offDst;
471	* uint32_t const cbDst = pReq->cbDst;
472	* RT_UNTRUSTED_NONVOLATILE_COPY_FENCE();
473	*
474	* ASSERT_GUEST_RETURN( cbDst <= pThis->cbSrc
475	* && offDst < pThis->cbSrc - cbDst,
476	* VERR_OUT_OF_RANGE);
477	* RT_UNTRUSTED_VALIDATED_FENCE();
478	* @endcode
479	*
480	* <li> Input validation does not only apply to shared data cases, but also to
481	* I/O port and MMIO handlers.
482	*
483	* <li> Ditto for kernel drivers working with usermode inputs.
484	*
485	* </ul>
486	*
487	*
488	* Problem patterns:
489	* - https://en.wikipedia.org/wiki/Time_of_check_to_time_of_use
490	* - https://googleprojectzero.blogspot.de/2018/01/reading-privileged-memory-with-side.html
491	* (Variant 1 only).
492	* - https://en.wikipedia.org/wiki/Integer_overflow
493	*
494	*
495	*
496	* @section sec_vbox_guideline_optional Optional
497	*
498	* First part is the actual coding style and all the prefixes. The second part
499	* is a bunch of good advice.
500	*
501	*
502	* @subsection sec_vbox_guideline_optional_layout The code layout
503	*
504	* <ul>
505	*
506	* <li> Max line length is 130 chars. Exceptions are table-like
507	* code/initializers and Log*() statements (don't waste unnecessary
508	* vertical space on debug logging).
509	*
510	* <li> Comments should try stay within the usual 80 columns as these are
511	* denser and too long lines may be harder to read.
512	*
513	* <li> Curly brackets are not indented. Example:
514	* @code
515	* if (true)
516	* {
517	* Something1();
518	* Something2();
519	* }
520	* else
521	* {
522	* SomethingElse1().
523	* SomethingElse2().
524	* }
525	* @endcode
526	*
527	* <li> Space before the parentheses when it comes after a C keyword.
528	*
529	* <li> No space between argument and parentheses. Exception for complex
530	* expression. Example:
531	* @code
532	* if (PATMR3IsPatchGCAddr(pVM, GCPtr))
533	* @endcode
534	*
535	* <li> The else of an if is always the first statement on a line. (No curly
536	* stuff before it!)
537	*
538	* <li> else and if go on the same line if no { compound statement }
539	* follows the if. Example:
540	* @code
541	* if (fFlags & MYFLAGS_1)
542	* fFlags &= ~MYFLAGS_10;
543	* else if (fFlags & MYFLAGS_2)
544	* {
545	* fFlags &= ~MYFLAGS_MASK;
546	* fFlags \|= MYFLAGS_5;
547	* }
548	* else if (fFlags & MYFLAGS_3)
549	* @endcode
550	*
551	* <li> Slightly complex boolean expressions are split into multiple lines,
552	* putting the operators first on the line and indenting it all according
553	* to the nesting of the expression. The purpose is to make it as easy as
554	* possible to read. Example:
555	* @code
556	* if ( RT_SUCCESS(rc)
557	* \|\| (fFlags & SOME_FLAG))
558	* @endcode
559	*
560	* <li> When 'if' or 'while' statements gets long, the closing parentheses
561	* goes right below the opening parentheses. This may be applied to
562	* sub-expression. Example:
563	* @code
564	* if ( RT_SUCCESS(rc)
565	* \|\| ( fSomeStuff
566	* && fSomeOtherStuff
567	* && fEvenMoreStuff
568	* )
569	* \|\| SomePredicateFunction()
570	* )
571	* {
572	* ...
573	* }
574	* @endcode
575	*
576	* <li> The case is indented from the switch (to avoid having the braces for
577	* the 'case' at the same level as the 'switch' statement).
578	*
579	* <li> If a case needs curly brackets they contain the entire case, are not
580	* indented from the case, and the break or return is placed inside them.
581	* Example:
582	* @code
583	* switch (pCur->eType)
584	* {
585	* case PGMMAPPINGTYPE_PAGETABLES:
586	* {
587	* unsigned iPDE = pCur->GCPtr >> PGDIR_SHIFT;
588	* unsigned iPT = (pCur->GCPtrEnd - pCur->GCPtr) >> PGDIR_SHIFT;
589	* while (iPT-- > 0)
590	* if (pPD->a[iPDE + iPT].n.u1Present)
591	* return VERR_HYPERVISOR_CONFLICT;
592	* break;
593	* }
594	* }
595	* @endcode
596	*
597	* <li> In a do while construction, the while is on the same line as the
598	* closing "}" if any are used.
599	* Example:
600	* @code
601	* do
602	* {
603	* stuff;
604	* i--;
605	* } while (i > 0);
606	* @endcode
607	*
608	* <li> Comments are in C style. C++ style comments are used for temporary
609	* disabling a few lines of code.
610	*
611	* <li> No unnecessary parentheses in expressions (just don't over do this
612	* so that gcc / msc starts bitching). Find a correct C/C++ operator
613	* precedence table if needed.
614	*
615	* <li> 'for (;;)' is preferred over 'while (true)' and 'while (1)'.
616	*
617	* <li> Parameters are indented to the start parentheses when breaking up
618	* function calls, declarations or prototypes. (This is in line with
619	* how 'if', 'for' and 'while' statements are done as well.) Example:
620	* @code
621	* RTPROCESS hProcess;
622	* int rc = RTProcCreateEx(papszArgs[0],
623	* papszArgs,
624	* RTENV_DEFAULT,
625	* fFlags,
626	* NULL, // phStdIn
627	* NULL, // phStdOut
628	* NULL, // phStdErr
629	* NULL, // pszAsUser
630	* NULL, // pszPassword
631	* &hProcess);
632	* @endcode
633	*
634	* <li> That Dijkstra is dead is no excuse for using gotos.
635	*
636	* <li> Using do-while-false loops to avoid gotos is considered very bad form.
637	* They create hard to read code. They tend to be either too short (i.e.
638	* pointless) or way to long (split up the function already), making
639	* tracking the state is difficult and prone to bugs. Also, they cause
640	* the compiler to generate suboptimal code, because the break branches
641	* are by preferred over the main code flow (MSC has no branch hinting!).
642	* Instead, do make use the 130 columns (i.e. nested ifs) and split
643	* the code up into more functions!
644	*
645	* <li> Avoid code like
646	* @code
647	* int foo;
648	* int rc;
649	* ...
650	* rc = FooBar();
651	* if (RT_SUCCESS(rc))
652	* {
653	* foo = getFoo();
654	* ...
655	* pvBar = RTMemAlloc(sizeof(*pvBar));
656	* if (!pvBar)
657	* rc = VERR_NO_MEMORY;
658	* }
659	* if (RT_SUCCESS(rc))
660	* {
661	* buzz = foo;
662	* ...
663	* }
664	* @endcode
665	* The intention of such code is probably to save some horizontal space
666	* but unfortunately it's hard to read and the scope of certain varables
667	* (e.g. foo in this example) is not optimal. Better use the following
668	* style:
669	* @code
670	* int rc;
671	* ...
672	* rc = FooBar();
673	* if (RT_SUCCESS(rc))
674	* {
675	* int foo = getFoo();
676	* ...
677	* pvBar = RTMemAlloc(sizeof(*pvBar));
678	* if (pvBar)
679	* {
680	* buzz = foo;
681	* ...
682	* }
683	* else
684	* rc = VERR_NO_MEMORY;
685	* }
686	* @endcode
687	*
688	* </ul>
689	*
690	* @subsection sec_vbox_guideline_optional_prefix Variable / Member Prefixes
691	*
692	* Prefixes are meant to provide extra context clues to a variable/member, we
693	* therefore avoid using prefixes that just indicating the type if a better
694	* choice is available.
695	*
696	*
697	* The prefixes:
698	*
699	* <ul>
700	*
701	* <li> The 'g_' (or 'g') prefix means a global variable, either on file or module level.
702	*
703	* <li> The 's_' (or 's') prefix means a static variable inside a function or
704	* class. This is not used for static variables on file level, use 'g_'
705	* for those (logical, right).
706	*
707	* <li> The 'm_' (or 'm') prefix means a class data member.
708	*
709	* In new code in Main, use "m_" (and common sense). As an exception,
710	* in Main, if a class encapsulates its member variables in an anonymous
711	* structure which is declared in the class, but defined only in the
712	* implementation (like this: 'class X { struct Data; Data *m; }'), then
713	* the pointer to that struct is called 'm' itself and its members then
714	* need no prefix, because the members are accessed with 'm->member'
715	* already which is clear enough.
716	*
717	* <li> The 'a_' prefix means a parameter (argument) variable. This is
718	* sometimes written 'a' in parts of the source code that does not use
719	* the array prefix.
720	*
721	* <li> The 'p' prefix means pointer. For instance 'pVM' is pointer to VM.
722	*
723	* <li> The 'r' prefix means that something is passed by reference.
724	*
725	* <li> The 'k' prefix means that something is a constant. For instance
726	* 'enum { kStuff };'. This is usually not used in combination with
727	* 'p', 'r' or any such thing, it's main main use is to make enums
728	* easily identifiable.
729	*
730	* <li> The 'a' prefix means array. For instance 'aPages' could be read as
731	* array of pages.
732	*
733	* <li> The 'c' prefix means count. For instance 'cbBlock' could be read,
734	* count of bytes in block. (1)
735	*
736	* <li> The 'cx' prefix means width (count of 'x' units).
737	*
738	* <li> The 'cy' prefix means height (count of 'y' units).
739	*
740	* <li> The 'x', 'y' and 'z' prefix refers to the x-, y- , and z-axis
741	* respectively.
742	*
743	* <li> The 'off' prefix means offset.
744	*
745	* <li> The 'i' or 'idx' prefixes usually means index. Although the 'i' one
746	* can sometimes just mean signed integer.
747	*
748	* <li> The 'i[1-9]+' prefix means a fixed bit size variable. Frequently
749	* used with the int[1-9]+_t types where the width is really important.
750	* In most cases 'i' is more appropriate. [type]
751	*
752	* <li> The 'e' (or 'enm') prefix means enum.
753	*
754	* <li> The 'u' prefix usually means unsigned integer. Exceptions follows.
755	*
756	* <li> The 'u[1-9]+' prefix means a fixed bit size variable. Frequently
757	* used with the uint[1-9]+_t types and with bitfields where the width is
758	* really important. In most cases 'u' or 'b' (byte) would be more
759	* appropriate. [type]
760	*
761	* <li> The 'b' prefix means byte or bytes. [type]
762	*
763	* <li> The 'f' prefix means flags. Flags are unsigned integers of some kind
764	* or booleans.
765	*
766	* <li> TODO: need prefix for real float. [type]
767	*
768	* <li> The 'rd' prefix means real double and is used for 'double' variables.
769	* [type]
770	*
771	* <li> The 'lrd' prefix means long real double and is used for 'long double'
772	* variables. [type]
773	*
774	* <li> The 'ch' prefix means a char, the (signed) char type. [type]
775	*
776	* <li> The 'wc' prefix means a wide/windows char, the RTUTF16 type. [type]
777	*
778	* <li> The 'uc' prefix means a Unicode Code point, the RTUNICP type. [type]
779	*
780	* <li> The 'uch' prefix means unsigned char. It's rarely used. [type]
781	*
782	* <li> The 'sz' prefix means zero terminated character string (array of
783	* chars). (UTF-8)
784	*
785	* <li> The 'wsz' prefix means zero terminated wide/windows character string
786	* (array of RTUTF16).
787	*
788	* <li> The 'usz' prefix means zero terminated Unicode string (array of
789	* RTUNICP).
790	*
791	* <li> The 'str' prefix means C++ string; either a std::string or, in Main,
792	* a Utf8Str or, in Qt, a QString. When used with 'p', 'r', 'a' or 'c'
793	* the first letter should be capitalized.
794	*
795	* <li> The 'bstr' prefix, in Main, means a UTF-16 Bstr. When used with 'p',
796	* 'r', 'a' or 'c' the first letter should be capitalized.
797	*
798	* <li> The 'pfn' prefix means pointer to function. Common usage is 'pfnCallback'
799	* and such like.
800	*
801	* <li> The 'psz' prefix is a combination of 'p' and 'sz' and thus means
802	* pointer to a zero terminated character string. (UTF-8)
803	*
804	* <li> The 'pcsz' prefix is used to indicate constant string pointers in
805	* parts of the code. Most code uses 'psz' for const and non-const
806	* string pointers, so please ignore this one.
807	*
808	* <li> The 'l' prefix means (signed) long. We try avoid using this,
809	* expecially with the 'LONG' types in Main as these are not 'long' on
810	* 64-bit non-Windows platforms and can cause confusion. Alternatives:
811	* 'i' or 'i32'. [type]
812	*
813	* <li> The 'ul' prefix means unsigned long. We try avoid using this,
814	* expecially with the 'ULONG' types in Main as these are not 'unsigned
815	* long' on 64-bit non-Windows platforms and can cause confusion.
816	* Alternatives: 'u' or 'u32'. [type]
817	*
818	* </ul>
819	*
820	* (1) Except in the occasional 'pcsz' prefix, the 'c' prefix is never ever
821	* used in the meaning 'const'.
822	*
823	*
824	* @subsection sec_vbox_guideline_optional_misc Misc / Advice / Stuff
825	*
826	* <ul>
827	*
828	* <li> When writing code think as the reader.
829	*
830	* <li> When writing code think as the compiler. (2)
831	*
832	* <li> When reading code think as if it's full of bugs - find them and fix them.
833	*
834	* <li> Pointer within range tests like:
835	* @code
836	* if ((uintptr_t)pv >= (uintptr_t)pvBase && (uintptr_t)pv < (uintptr_t)pvBase + cbRange)
837	* @endcode
838	* Can also be written as (assuming cbRange unsigned):
839	* @code
840	* if ((uintptr_t)pv - (uintptr_t)pvBase < cbRange)
841	* @endcode
842	* Which is shorter and potentially faster. (1)
843	*
844	* <li> Avoid unnecessary casting. All pointers automatically cast down to
845	* void *, at least for non class instance pointers.
846	*
847	* <li> It's very very bad practise to write a function larger than a
848	* screen full (1024x768) without any comprehensibility and explaining
849	* comments.
850	*
851	* <li> More to come....
852	*
853	* </ul>
854	*
855	* (1) Important, be very careful with the casting. In particular, note that
856	* a compiler might treat pointers as signed (IIRC).
857	*
858	* (2) "A really advanced hacker comes to understand the true inner workings of
859	* the machine - he sees through the language he's working in and glimpses
860	* the secret functioning of the binary code - becomes a Ba'al Shem of
861	* sorts." (Neal Stephenson "Snow Crash")
862	*
863	*
864	*
865	* @section sec_vbox_guideline_warnings Compiler Warnings
866	*
867	* The code should when possible compile on all platforms and compilers without any
868	* warnings. That's a nice idea, however, if it means making the code harder to read,
869	* less portable, unreliable or similar, the warning should not be fixed.
870	*
871	* Some of the warnings can seem kind of innocent at first glance. So, let's take the
872	* most common ones and explain them.
873	*
874	*
875	* @subsection sec_vbox_guideline_warnings_signed_unsigned_compare Signed / Unsigned Compare
876	*
877	* GCC says: "warning: comparison between signed and unsigned integer expressions"
878	* MSC says: "warning C4018: '<\|<=\|==\|>=\|>' : signed/unsigned mismatch"
879	*
880	* The following example will not output what you expect:
881	@code
882	#include <stdio.h>
883	int main()
884	{
885	signed long a = -1;
886	unsigned long b = 2294967295;
887	if (a < b)
888	printf("%ld < %lu: true\n", a, b);
889	else
890	printf("%ld < %lu: false\n", a, b);
891	return 0;
892	}
893	@endcode
894	* If I understood it correctly, the compiler will convert a to an
895	* unsigned long before doing the compare.
896	*
897	*
898	*
899	* @section sec_vbox_guideline_svn Subversion Commit Rules
900	*
901	*
902	* Before checking in:
903	*
904	* <ul>
905	*
906	* <li> Check Tinderbox and make sure the tree is green across all platforms. If it's
907	* red on a platform, don't check in. If you want, warn in the \#vbox channel and
908	* help make the responsible person fix it.
909	* NEVER CHECK IN TO A BROKEN BUILD.
910	*
911	* <li> When checking in keep in mind that a commit is atomic and that the Tinderbox and
912	* developers are constantly checking out the tree. Therefore do not split up the
913	* commit unless it's into 100% independent parts. If you need to split it up in order
914	* to have sensible commit comments, make the sub-commits as rapid as possible.
915	*
916	* <li> If you make a user visible change, such as fixing a reported bug,
917	* make sure you add an entry to doc/manual/user_ChangeLogImpl.xml.
918	*
919	* <li> If you are adding files make sure set the right attributes.
920	* svn-ps.sh/cmd was created for this purpose, please make use of it.
921	*
922	* </ul>
923	*
924	* After checking in:
925	*
926	* <ul>
927	*
928	* <li> After checking-in, you watch Tinderbox until your check-ins clear. You do not
929	* go home. You do not sleep. You do not log out or experiment with drugs. You do
930	* not become unavailable. If you break the tree, add a comment saying that you're
931	* fixing it. If you can't fix it and need help, ask in the \#innotek channel or back
932	* out the change.
933	*
934	* </ul>
935	*
936	* (Inspired by mozilla tree rules.)
937	*
938	*
939	*/
940

Note: See TracBrowser for help on using the repository browser.

source: vbox/trunk/doc/VBox-CodingGuidelines.cpp@ 80070

Download in other formats: