4 Please use the script checkpatch.pl in the scripts directory to check
5 patches before submitting.
9 Of course, the most important aspect in any coding style is whitespace.
10 Crusty old coders who have trouble spotting the glasses on their noses
11 can tell the difference between a tab and eight spaces from a distance
12 of approximately fifteen parsecs. Many a flamewar has been fought and
15 QEMU indents are four spaces. Tabs are never used, except in Makefiles
16 where they have been irreversibly coded into the syntax.
17 Spaces of course are superior to tabs because:
19 - You have just one way to specify whitespace, not two. Ambiguity breeds
21 - The confusion surrounding 'use tabs to indent, spaces to justify' is gone.
22 - Tab indents push your code to the right, making your screen seriously
24 - Tabs will be rendered incorrectly on editors who are misconfigured not
25 to use tab stops of eight positions.
26 - Tabs are rendered badly in patches, causing off-by-one errors in almost
28 - It is the QEMU coding style.
30 Do not leave whitespace dangling off the ends of lines.
34 Lines should be 80 characters; try not to make them longer.
36 Sometimes it is hard to do, especially when dealing with QEMU subsystems
37 that use long function or symbol names. Even in that case, do not make
38 lines much longer than 80 characters.
41 - Some people like to tile their 24" screens with a 6x4 matrix of 80x24
42 xterms and use vi in all of them. The best way to punish them is to
43 let them keep doing it.
44 - Code and especially patches is much more readable if limited to a sane
45 line length. Eighty is traditional.
46 - The four-space indentation makes the most common excuse ("But look
47 at all that white space on the left!") moot.
48 - It is the QEMU coding style.
52 Variables are lower_case_with_underscores; easy to type and read. Structured
53 type names are in CamelCase; harder to type but standing out. Enum type
54 names and function type names should also be in CamelCase. Scalar type
55 names are lower_case_with_underscores_ending_with_a_t, like the POSIX
56 uint64_t and family. Note that this last convention contradicts POSIX
57 and is therefore likely to be changed.
59 When wrapping standard library functions, use the prefix qemu_ to alert
60 readers that they are seeing a wrapped version; otherwise avoid this prefix.
64 Every indented statement is braced; even if the block contains just one
65 statement. The opening brace is on the line that contains the control
66 flow statement that introduces the new block; the closing brace is on the
67 same line as the else keyword, or on a line by itself if there is no else
75 printf("a was something else entirely.\n");
78 Note that 'else if' is considered a single statement; otherwise a long if/
79 else if/else if/.../else sequence would need an indent for every else
82 An exception is the opening brace for a function; for reasons of tradition
83 and clarity it comes on a line by itself:
90 Rationale: a consistent (except for functions...) bracing style reduces
91 ambiguity and avoids needless churn when lines are added or removed.
92 Furthermore, it is the QEMU coding style.
96 Mixed declarations (interleaving statements and declarations within
97 blocks) are generally not allowed; declarations should be at the beginning
100 Every now and then, an exception is made for declarations inside a
101 #ifdef or #ifndef block: if the code looks nicer, such declarations can
102 be placed at the top of the block even if there are statements above.
103 On the other hand, however, it's often best to move that #ifdef/#ifndef
104 block to a separate function altogether.
106 6. Conditional statements
108 When comparing a variable for (in)equality with a constant, list the
109 constant on the right, as in:
112 /* Reads like: "If a equals 1" */
116 Rationale: Yoda conditions (as in 'if (1 == a)') are awkward to read.
117 Besides, good compilers already warn users when '==' is mis-typed as '=',
118 even when the constant is on the right.
122 We use traditional C-style /* */ comments and avoid // comments.
124 Rationale: The // form is valid in C99, so this is purely a matter of
125 consistency of style. The checkpatch script will warn you about this.
127 Multiline comment blocks should have a row of stars on the left,
128 and the initial /* and terminating */ both on their own lines:
133 This is the same format required by the Linux kernel coding style.
135 (Some of the existing comments in the codebase use the GNU Coding
136 Standards form which does not have stars on the left, or other
137 variations; avoid these when writing new comments, but don't worry
138 about converting to the preferred form unless you're editing that
141 Rationale: Consistency, and ease of visually picking out a multiline
142 comment from the surrounding code.
144 8. trace-events style
148 In trace-events files, use a '0x' prefix to specify hex numbers, as in:
150 some_trace(unsigned x, uint64_t y) "x 0x%x y 0x" PRIx64
152 An exception is made for groups of numbers that are hexadecimal by
153 convention and separated by the symbols '.', '/', ':', or ' ' (such as
156 another_trace(int cssid, int ssid, int dev_num) "bus id: %x.%x.%04x"
158 However, you can use '0x' for such groups if you want. Anyway, be sure that
159 it is obvious that numbers are in hex, ex.:
161 data_dump(uint8_t c1, uint8_t c2, uint8_t c3) "bytes (in hex): %02x %02x %02x"
163 Rationale: hex numbers are hard to read in logs when there is no 0x prefix,
164 especially when (occasionally) the representation doesn't contain any letters
165 and especially in one line with other decimal numbers. Number groups are allowed
166 to not use '0x' because for some things notations like %x.%x.%x are used not
167 only in Qemu. Also dumping raw data bytes with '0x' is less readable.
171 Do not use printf flag '#', like '%#x'.
173 Rationale: there are two ways to add a '0x' prefix to printed number: '0x%...'
174 and '%#...'. For consistency the only one way should be used. Arguments for
177 - '%#' omits the 0x for the value 0 which makes output inconsistent