Linux内核编码风格(编程代码风格推荐)
谈谈为 Linux 内核写驱动的编码规范
谈谈为 Linux 内核写驱动的编码规范最近在向Linux内核提交一些驱动程序，在提交的过程中，发现自己的代码离Linux内核的coding style要求还是差很多。
当初自己对内核文档里的CodingStyle一文只是粗略的浏览，真正写代码的时候在很多细节上会照顾不周。
不过，在不遵守规则的程序员队伍里，我并不是孤独的。
如果去看drivers/staging下的代码，就会发现很多驱动程序都没有严格遵守内核的coding style，而且在很多驱动程序的TODO文件里，都会把"checkpatch.pl fixes"作为自己的目标之一（checkpatch.pl是用来检查代码是否符合coding style的脚本）。
不可否认，coding style是仁者见仁、智者见智的事情。
比如Microsoft所推崇的匈牙利命名法，在Linus看来就是及其脑残(brain damaged)的做法。
也许您并不赞成Linus制定的coding style，但在提交内核驱动这件事上，最好还是以大局为重。
对于这么一个庞大的集市式的开发来说，随意书写代码必将带来严重的可维护性的灾难。
（题图来自：mota.ru）一些辅助工具当代码量达到一定程度时，手动去检查和修改coding style是非常繁琐的工作，幸好，我们还有一些工具可以使用。
scripts/checkpatch.pl这是一个检查代码是否符合内核编码规范的的脚本。
顾名思义，checkpatch是用来检查patch的，默认的调用也确实如此。
如果用来检查原文件，需要加上“-f”的选项。
我们来看一段无聊的代码(文件名为print_msg.c)：1.void print_msg(int a)2.{3. switch (a) {4. case 1:5. printf("a == 1\n");6. break;7.8. case 2:9. printf("a == 2\n");10. break;11. }12.}这段代码的coding style是否有问题呢？用checkpatch.pl来检查一下：1.scripts/checkpatch.pl -f print_msg.c检查的结果是：1.ERROR: switch and case should be at the same indent2.#3: FILE: switch.c:3:3.+ switch (a) {4.+ case 1:5.[...]6.+ case 2:7.8.total: 1 errors, 0 warnings, 12 lines checked9.10.switch.c has style problems, please review. If any of theseerrors11.are false positives report them to the maintainer, see12.CHECKPATCH in MAINTAINERS.在Linux内核的coding style里，switch和case要求有相同的缩进。
LINUX内核模块编程指南
LINUX内核模块编程指南Linux内核模块编程是一种在Linux操作系统中向内核添加新功能的方法。
内核模块是一段特殊的代码，可以被编译并加载到内核中，从而实现对操作系统底层的修改和扩展。
下面将介绍一些关键的步骤和注意事项，以帮助初学者入门Linux内核模块编程。
首先，要编写一个内核模块，需要有合适的开发环境。
一般来说，需要一台运行Linux的计算机，并安装有相应的开发工具链（如GCC编译器）和Linux内核源代码。
编写内核模块的第一步是定义模块的入口函数。
在C语言中，入口函数的原型通常是`int init_module(void)`。
在该函数中，可以注册一些回调函数来处理特定的事件，或者进行其他必要的初始化工作。
接下来，可以添加自定义的函数和数据结构，以实现模块的功能。
这些函数和数据结构可以被其他模块或内核中的其他代码调用和使用。
由于内核是一个复杂的系统，因此在编写内核模块时需要特别关注数据的正确性和同步性。
在编写模块的过程中，还需要了解一些内核编程的特殊机制，如内核的锁机制、中断处理和内存管理等。
对这些内容的学习需要一定的时间和精力，但是它们是实现高性能和高可靠性内核模块的关键。
完成模块的编写后，需要对其进行编译和加载。
一般来说，可以使用Makefile来编译模块，并且可以使用insmod命令来加载模块。
加载模块时，内核将会调用模块的入口函数，从而完成初始化工作。
在模块编写和调试的过程中，需要注意一些常见的问题。
例如，内核模块应该遵守内核的编程规范和风格，避免对内核进行不必要的修改或侵入。
此外，要时刻注意内存管理和锁机制，以防止出现内存泄漏或并发访问的问题。
最后，需要强调的是，Linux内核模块编程是一项复杂而底层的技术。
对于初学者来说，学习和掌握这些知识需要一定的时间和耐心，并且需要有一定的计算机基础和编程经验。
但是，一旦熟练掌握了这些技术，就能够更加深入地了解和使用Linux操作系统，并且为其添加新的功能。
LINUX编程风格
Linux编程风格0.1一、GNU风格1.函数返回类型说明和函数名分两行放置，函数起始字符和函数开头左花括号放到最左边，例如：static char *main (argc, argv)int argc;char *argv[];{......}或者是用标准C：static char *main (int argc, char *argv[]){......}如果参数太长不能放到一行，请在每行参数开头处对齐：intnet_connect (struct sockaddr_in *cs, char *server, unsigned short int port,char *sourceip, unsigned short int sourceport, int sec) 对于函数体，我们应该按照如何方式排版：在左括号之前、逗号之后，以及运算符号前后添加空格使程序便于阅读，例如：if (x < foo (y, z))haha = bar[4] + 5;else{while (z){haha += foo (z, z);z--;}return ++x + bar ();}当一个表达式需要分成多行书写的时候，应该在操作符之前分割。
例如：if (foo_this_is_long && bar > win (x, y, z)&& remaining_condition)2. 尽量不要让两个不同优先级的操作符出现在相同的对齐方式中,应该附加额外的括号使得代码缩进可以表示出嵌套。
例如：错误的对齐：mode = (inmode[j] == VOIDmode|| GET_MODE_SIZE (outmode[j]) > GET_MODE_SIZE (inmode[j])? outmode[j] : inmode[j];正确的对齐：mode = ((inmode[j] == VOIDmode|| (GET_MODE_SIZE (outmode[j]) > GET_MODE_SIZE (inmode[j]))) ? outmode[j] : inmode[j];3. 按照如下方式排版do-while语句：do{a = foo (a);}while (a > 0);4. 每个程序都应该以一段简短的说明其功能的注释开头。
linux 的编码格式
linux 的编码格式**Linux 编码格式简介**在Linux 系统中，编码格式起着至关重要的作用。
它是一种将字符转换为二进制的方式，以便在计算机内部进行存储和传输。
了解Linux 编码格式对于更好地使用这个强大的操作系统至关重要。
**Linux 常用编码格式对比**Linux 系统中存在多种编码格式，其中最常用的有ASCII、UTF-8、GBK 等。
1.ASCII：这是一种最基本的编码格式，主要应用于英文和数字。
然而，它在处理中文等复杂字符时存在局限性。
2.UTF-8：这是一种广泛应用于Linux 的编码格式，它可以表示几乎所有的Unicode 字符。
UTF-8 具有良好的兼容性和跨平台性，是目前Linux 系统中最为推荐的编码格式。
3.GBK：这是一种针对中文设计的编码格式，适用于简体中文环境。
在Linux 系统中，GBK 编码可以满足大部分中文文本的处理需求。
**如何选择合适的编码格式**在选择编码格式时，需要考虑以下几个方面：1.目标用户：根据您的读者群体，选择他们熟悉的编码格式。
例如，面向国内用户，可以选择GBK 编码；面向国际用户，可以选择UTF-8 编码。
2.文件兼容性：确保所选编码格式与其他操作系统和软件的兼容性。
例如，使用UTF-8 编码可以确保在各种操作系统和设备上都能正常显示文本。
3.存储空间：较小的编码格式如ASCII 和GBK 占用存储空间较少，但处理字符范围有限；UTF-8 编码虽然兼容性较好，但占用存储空间相对较大。
4.文本处理需求：根据您的文本处理需求，选择合适的编码格式。
例如，如果您需要处理大量特殊字符或Unicode 字符，可以选择UTF-8 编码。
**编码格式在Linux 中的应用场景**编码格式在Linux 系统中有着广泛的应用场景，如：1.文本编辑器：不同的编码格式可以在文本编辑器中显示和编辑不同类型的文本。
2.编程语言：编程语言中的变量名、注释和字符串常使用编码格式进行存储和处理。
Linux底层驱动开发的最佳实践让你的代码更优雅
Linux底层驱动开发的最佳实践让你的代码更优雅在Linux系统中，底层驱动扮演着连接硬件设备与操作系统的重要角色。
良好的驱动开发实践不仅能够确保系统的稳定性和性能，还能提高代码的可读性和可维护性。
本文将介绍一些Linux底层驱动开发的最佳实践，帮助你编写更优雅的驱动代码。
1. 遵循Linux内核开发规范Linux内核有一套自己的开发规范，严格遵循这些规范能够确保你的驱动代码与内核的其他模块相协调。
比如，驱动应该使用适当的内核API来访问硬件，而不是直接与硬件进行交互。
此外，在注册驱动时，应该明确指定驱动支持的设备ID，以避免与其他驱动产生冲突。
2. 使用合适的数据结构在编写驱动代码时，选择合适的数据结构对于代码的可读性和可维护性非常重要。
Linux内核提供了许多常用的数据结构，如链表、哈希表和位图等，驱动开发者可以根据实际需求选择合适的数据结构来组织和管理设备数据。
3. 使用适当的设备驱动模型Linux内核提供了多种设备驱动模型，如Platform Driver、Character Driver和Network Driver等。
选择合适的设备驱动模型有助于简化驱动开发流程，并提高代码的可维护性。
例如，如果你的驱动只需要与特定硬件平台进行通信，那么Platform Driver模型可能是最合适的选择。
4. 优化中断处理中断处理是驱动开发中一个关键的环节。
合理处理中断可以提高系统的响应性能。
在编写中断处理程序时，应该尽量减少对共享资源的访问，使用适当的同步机制来保护共享资源的访问。
5. 调试和错误处理在驱动开发过程中，合理的调试和错误处理机制是非常重要的。
你可以使用调试工具和技术来跟踪代码执行流程、打印调试信息和监测性能瓶颈。
此外，你还应该为驱动添加适当的错误处理代码，以应对意外情况和异常情况，提高系统的稳定性。
6. 编写适当的文档为你的驱动代码编写适当的文档是一种良好的开发实践。
文档应该包括驱动的功能、接口、使用示例和注意事项等信息，以方便其他开发者使用和理解你的代码。
linux操作系统内核版表达方式
Linux操作系统内核源代码的表达方式Linux操作系统内核源代码的表达方式是指内核代码中所使用的语法和结构。
它决定了内核代码的可读性、可维护性和可移植性。
内核源代码的表达方式主要包括以下几个方面：语法：Linux内核源代码使用C语言编写。
C语言是一种广泛使用的通用编程语言，具有丰富的库和工具支持，并且很容易移植到不同的平台上。
结构：Linux内核源代码由许多相互关联的文件组成，这些文件包含了内核的各种功能和模块。
内核源代码的结构通常是分层的，每一层都包含了不同的功能。
命名约定：Linux内核源代码使用了严格的命名约定，这使得内核代码更容易阅读和维护。
例如，内核中的函数通常以“sys_”或“do_”开头，而结构体通常以“struct_”开头。
注释：Linux内核源代码中包含了大量的注释，这些注释解释了代码的功能和用法。
注释对于理解内核代码非常有用，尤其是对于新内核开发者来说。
Linux内核源代码的表达方式的特点Linux内核源代码的表达方式具有以下几个特点：简洁：Linux内核源代码非常简洁，代码中没有多余的注释和代码段。
这种简洁性使得内核代码更容易阅读和维护。
模块化：Linux内核源代码是模块化的，内核中的各个模块可以独立编译和加载。
这种模块化设计使得内核更容易扩展和维护。
可移植性：Linux内核源代码具有很强的可移植性，它可以移植到不同的硬件平台上。
这种可移植性使得Linux内核成为世界上最受欢迎的操作系统之一。
Linux内核源代码的表达方式的优点Linux内核源代码的表达方式具有以下几个优点：可读性：Linux内核源代码非常可读，代码中的注释和命名约定使得内核代码很容易理解。
这种可读性对于内核开发者来说非常重要，它可以帮助开发者快速理解和修改内核代码。
可维护性：Linux内核源代码也非常可维护，代码中的模块化设计使得内核更容易扩展和维护。
这种可维护性对于内核的长期发展非常重要，它可以确保内核能够不断更新和改进。
Linux内核编码风格
Linux内核编码风格 像其他⼤型软件⼀样，Linux制订了⼀套编码风格，对代码的格式、风格和布局做出了规定。
我写这篇的⽬的也就是希望⼤家能够从中借鉴，有利于⼤家提⾼编程效率。
像Linux内核这样⼤型软件中，涉及许许多多的开发者，故它的编码风格也很有参考价值。
1、左括号紧跟在语句的最后，与语句在相同的⼀⾏。
⽽右括号要另起⼀⾏，作为该⾏的第⼀个字符。
2、如果接下来的部分是相同语句的⼀部分，那么右括号就不单独占⼀⾏。
3、还有 4、函数采⽤以下的书写⽅式： 5、最后不需要⼀定使⽤括号的语句可以忽略它： 要尽可能地保证代码长度不超过80个字符，如果代码⾏超过80应该折到下⼀⾏。
将参数分⾏输⼊，在开头简单地加⼊两个标准tab： 名称中不允许使⽤混合的⼤⼩写字符。
局部变量如果能够清楚地表明它的⽤途，那么选取idx甚⾄是i这样的名称都是可⾏的。
⽽像theLoopIndex这样冗长反复的名字不在接受之列。
——匈⽛利命名法（在变量名称中加⼊变量的类别）危害极⼤。
函数 根据经验函数的代码长度不应该超过两屏，局部变量不应该超过⼗个。
1、⼀个函数应该功能单⼀并且实现精准。
2、将⼀个函数分解成⼀些更短⼩的函数的组合不会带来危害。
——如果你担⼼函数调⽤导致的开销，可以使⽤inline关键字。
⼀般情况下，注释的⽬的是描述你的代码要做什么和为什么要做，⽽不是具体通过什么⽅式实现的。
怎么实现应该由代码本⾝展现。
注释不应该包含谁写了那个函数，修改⽇期和其他那些琐碎⽽⽆实际意义的内容。
这些信息应该集中在⽂件最开头地⽅。
内核中⼀条注释看起来如下： 重要信息常常以“XXX：”开头，⽽bug通常以“FIXME"开头，就像：总结 希望这篇博客对⼤家有所帮助！更详尽的内容，请看"Linux 内核代码规范原⽂"Linus 内部代码规范原⽂1 Linux kernel coding style23 This is a short document describing the preferred coding style for the4 linux kernel. Coding style is very personal, and I won't _force_ my5 views on anybody, but this is what goes for anything that I have to be6 able to maintain, and I'd prefer it for most other things too. Please7 at least consider the points made here.89 First off, I'd suggest printing out a copy of the GNU coding standards,10 and NOT read it. Burn them, it's a great symbolic gesture.1112 Anyway, here goes:131415 Chapter 1: Indentation1617 Tabs are 8 characters, and thus indentations are also 8 characters.18 There are heretic movements that try to make indentations 4 (or even 2!)19 characters deep, and that is akin to trying to define the value of PI to20 be 3.2122 Rationale: The whole idea behind indentation is to clearly define where23 a block of control starts and ends. Especially when you've been looking24 at your screen for20 straight hours, you'll find it a lot easier to see25 how the indentation works if you have large indentations.2627 Now, some people will claim that having 8-character indentations makes28 the code move too far to the right, and makes it hard to read on a2980-character terminal screen. The answer to that is that if you need30 more than 3 levels of indentation, you're screwed anyway, and should fix31 your program.3233 In short, 8-char indents make things easier to read, and have the added34 benefit of warning you when you're nesting your functions too deep.35 Heed that warning.3637 The preferred way to ease multiple indentation levels in a switch statement is38 to align the "switch" and its subordinate "case" labels in the same column39 instead of "double-indenting" the "case" labels. E.g.:4041switch (suffix) {42case'G':43case'g':44 mem <<= 30;45break;46case'M':47case'm':48 mem <<= 20;49break;51case'k':52 mem <<= 10;53/* fall through */54default:55break;56 }575859 Don't put multiple statements on a single line unless you have60 something to hide:6162if (condition) do_this;63 do_something_everytime;6465 Don't put multiple assignments on a single line either. Kernel coding style66is super simple. Avoid tricky expressions.6768 Outside of comments, documentation and except in Kconfig, spaces are never69 used for indentation, and the above example is deliberately broken.7071 Get a decent editor and don't leave whitespace at the end of lines.727374 Chapter 2: Breaking long lines and strings7576 Coding style is all about readability and maintainability using commonly77 available tools.7879 The limit on the length of lines is80 columns and this is a strongly80 preferred limit.8182 Statements longer than 80 columns will be broken into sensible chunks, unless83 exceeding 80 columns significantly increases readability and does not hide84 information. Descendants are always substantially shorter than the parent and85 are placed substantially to the right. The same applies to function headers86 with a long argument list. However, never break user-visible strings such as87 printk messages, because that breaks the ability to grep for them.888990 Chapter 3: Placing Braces and Spaces9192 The other issue that always comes up in C styling is the placement of93 braces. Unlike the indent size, there are few technical reasons to94 choose one placement strategy over the other, but the preferred way, as95 shown to us by the prophets Kernighan and Ritchie, is to put the opening96 brace last on the line, and put the closing brace first, thusly:9798if (x is true) {99 we do y100 }101102 This applies to all non-function statement blocks (if, switch, for,103while, do). E.g.:104105switch (action) {106case KOBJ_ADD:107return"add";108case KOBJ_REMOVE:109return"remove";110case KOBJ_CHANGE:111return"change";112default:113return NULL;114 }115116 However, there is one special case, namely functions: they have the117 opening brace at the beginning of the next line, thus:118119int function(int x)120 {121 body of function122 }123124 Heretic people all over the world have claimed that this inconsistency125is ... well ... inconsistent, but all right-thinking people know that126 (a) K&R are _right_ and (b) K&R are right. Besides, functions are127 special anyway (you can't nest them in C).128129 Note that the closing brace is empty on a line of its own, _except_ in130 the cases where it is followed by a continuation of the same statement,131 ie a "while"in a do-statement or an "else"in an if-statement, like132this:133135 body of do-loop136 } while (condition);137138 and139140if (x == y) {141 ..142 } else if (x > y) {143 ...144 } else {145 ....146 }147148 Rationale: K&R.149150 Also, note that this brace-placement also minimizes the number of empty151 (or almost empty) lines, without any loss of readability. Thus, as the152 supply of new-lines on your screen is not a renewable resource (think15325-line terminal screens here), you have more empty lines to put154 comments on.155156 Do not unnecessarily use braces where a single statement will do.157158if (condition)159 action();160161 and162163if (condition)164 do_this();165else166 do_that();167168 This does not apply if only one branch of a conditional statement is a single 169 statement; in the latter case use braces in both branches:170171if (condition) {172 do_this();173 do_that();174 } else {175 otherwise();176 }177178 3.1: Spaces179180 Linux kernel style for use of spaces depends (mostly) on181 function-versus-keyword usage. Use a space after (most) keywords. The 182 notable exceptions are sizeof, typeof, alignof, and __attribute__, which look 183 somewhat like functions (and are usually used with parentheses in Linux,184 although they are not required in the language, as in: "sizeof info" after 185"struct fileinfo info;"is declared).186187 So use a space after these keywords:188if, switch, case, for, do, while189 but not with sizeof, typeof, alignof, or __attribute__. E.g.,190 s = sizeof(struct file);191192 Do not add spaces around (inside) parenthesized expressions. This example is 193 *bad*:194195 s = sizeof( struct file );196197 When declaring pointer data or a function that returns a pointer type, the198 preferred use of '*'is adjacent to the data name or function name and not199 adjacent to the type name. Examples:200201char *linux_banner;202 unsigned long long memparse(char *ptr, char **retptr);203char *match_strdup(substring_t *s);204205 Use one space around (on each side of) most binary and ternary operators, 206 such as any of these:207208 = + - < > * / % | & ^ <= >= == != ? :209210 but no space after unary operators:211 & * + - ~ ! sizeof typeof alignof __attribute__ defined212213 no space before the postfix increment & decrement unary operators:214 ++ --215216 no space after the prefix increment & decrement unary operators:217 ++ --219 and no space around the '.' and "->" structure member operators.220221 Do not leave trailing whitespace at the ends of lines. Some editors with 222"smart" indentation will insert whitespace at the beginning of new lines as 223 appropriate, so you can start typing the next line of code right away.224 However, some such editors do not remove the whitespace if you end up not 225 putting a line of code there, such as if you leave a blank line. As a result, 226 you end up with lines containing trailing whitespace.227228 Git will warn you about patches that introduce trailing whitespace, and can 229 optionally strip the trailing whitespace for you; however, if applying a series 230 of patches, this may make later patches in the series fail by changing their 231 context lines.232233234 Chapter 4: Naming235236 C is a Spartan language, and so should your naming be. Unlike Modula-2 237 and Pascal programmers, C programmers do not use cute names like238 ThisVariableIsATemporaryCounter. A C programmer would call that239 variable "tmp", which is much easier to write, and not the least more240 difficult to understand.241242 HOWEVER, while mixed-case names are frowned upon, descriptive names for 243global variables are a must. To call a global function "foo"is a244 shooting offense.245246 GLOBAL variables (to be used only if you _really_ need them) need to247 have descriptive names, as do global functions. If you have a function248 that counts the number of active users, you should call that249"count_active_users()" or similar, you should _not_ call it "cntusr()".250251 Encoding the type of a function into the name (so-called Hungarian252 notation) is brain damaged - the compiler knows the types anyway and can 253 check those, and it only confuses the programmer. No wonder MicroSoft 254 makes buggy programs.255256 LOCAL variable names should be short, and to the point. If you have257 some random integer loop counter, it should probably be called "i".258 Calling it "loop_counter"is non-productive, if there is no chance of it259 being mis-understood. Similarly, "tmp" can be just about any type of260 variable that is used to hold a temporary value.261262 If you are afraid to mix up your local variable names, you have another263 problem, which is called the function-growth-hormone-imbalance syndrome. 264 See chapter 6 (Functions).265266267 Chapter 5: Typedefs268269 Please don't use things like "vps_t".270271 It's a _mistake_ to use typedef for structures and pointers. When you see a 272273 vps_t a;274275in the source, what does it mean?276277 In contrast, if it says278279struct virtual_container *a;280281 you can actually tell what "a"is.282283 Lots of people think that typedefs "help readability". Not so. They are284 useful only for:285286 (a) totally opaque objects (where the typedef is actively used to _hide_287 what the object is).288289 Example: "pte_t" etc. opaque objects that you can only access using290 the proper accessor functions.291292 NOTE! Opaqueness and "accessor functions" are not good in themselves. 293 The reason we have them for things like pte_t etc. is that there294 really is absolutely _zero_ portably accessible information there.295296 (b) Clear integer types, where the abstraction _helps_ avoid confusion297 whether it is"int" or "long".298299 u8/u16/u32 are perfectly fine typedefs, although they fit into300 category (d) better than here.301303"unsigned long", then there's no reason to do304305 typedef unsigned long myflags_t;306307 but if there is a clear reason for why it under certain circumstances308 might be an "unsigned int" and under other configurations might be 309"unsigned long", then by all means go ahead and use a typedef.310311 (c) when you use sparse to literally create a _new_ type for312 type-checking.313314 (d) New types which are identical to standard C99 types, in certain315 exceptional circumstances.316317 Although it would only take a short amount of time for the eyes and318 brain to become accustomed to the standard types like 'uint32_t',319 some people object to their use anyway.320321 Therefore, the Linux-specific 'u8/u16/u32/u64' types and their322 signed equivalents which are identical to standard types are323 permitted -- although they are not mandatory in new code of your324 own.325326 When editing existing code which already uses one or the other set327 of types, you should conform to the existing choices in that code.328329 (e) Types safe for use in userspace.330331 In certain structures which are visible to userspace, we cannot332 require C99 types and cannot use the 'u32' form above. Thus, we333 use __u32 and similar types in all structures which are shared334 with userspace.335336 Maybe there are other cases too, but the rule should basically be to NEVER 337 EVER use a typedef unless you can clearly match one of those rules.338339 In general, a pointer, or a struct that has elements that can reasonably340 be directly accessed should _never_ be a typedef.341342343 Chapter 6: Functions344345 Functions should be short and sweet, and do just one thing. They should346 fit on one or two screenfuls of text (the ISO/ANSI screen size is 80x24,347as we all know), and do one thing and do that well.348349 The maximum length of a function is inversely proportional to the350 complexity and indentation level of that function. So, if you have a351 conceptually simple function that is just one long (but simple)352case-statement, where you have to do lots of small things for a lot of353 different cases, it's OK to have a longer function.354355 However, if you have a complex function, and you suspect that a356 less-than-gifted first-year high-school student might not even357 understand what the function is all about, you should adhere to the358 maximum limits all the more closely. Use helper functions with359 descriptive names (you can ask the compiler to in-line them if you think360 it's performance-critical, and it will probably do a better job of it361 than you would have done).362363 Another measure of the function is the number of local variables. They364 shouldn't exceed 5-10, or you're doing something wrong. Re-think the365 function, and split it into smaller pieces. A human brain can366 generally easily keep track of about 7 different things, anything more367 and it gets confused. You know you're brilliant, but maybe you'd like368 to understand what you did 2 weeks from now.369370 In source files, separate functions with one blank line. If the function is371 exported, the EXPORT* macro for it should follow immediately after the closing 372 function brace line. E.g.:373374int system_is_up(void)375 {376return system_state == SYSTEM_RUNNING;377 }378 EXPORT_SYMBOL(system_is_up);379380 In function prototypes, include parameter names with their data types.381 Although this is not required by the C language, it is preferred in Linux382 because it is a simple way to add valuable information for the reader.383384385 Chapter 7: Centralized exiting of functions387 Albeit deprecated by some people, the equivalent of the goto statement is388 used frequently by compilers in form of the unconditional jump instruction.389390 The goto statement comes in handy when a function exits from multiple391 locations and some common work such as cleanup has to be done.392393 The rationale is:394395 - unconditional statements are easier to understand and follow396 - nesting is reduced397 - errors by not updating individual exit points when making398 modifications are prevented399 - saves the compiler work to optimize redundant code away ;)400401int fun(int a)402 {403int result = 0;404char *buffer = kmalloc(SIZE);405406if (buffer == NULL)407return -ENOMEM;408409if (condition1) {410while (loop1) {411 ...412 }413 result = 1;414goto out;415 }416 ...417out:418 kfree(buffer);419return result;420 }421422 Chapter 8: Commenting423424 Comments are good, but there is also a danger of over-commenting. NEVER 425try to explain HOW your code works in a comment: it's much better to426 write the code so that the _working_ is obvious, and it's a waste of427 time to explain badly written code.428429 Generally, you want your comments to tell WHAT your code does, not HOW. 430 Also, try to avoid putting comments inside a function body: if the431 function is so complex that you need to separately comment parts of it,432 you should probably go back to chapter 6for a while. You can make433 small comments to note or warn about something particularly clever (or434 ugly), but try to avoid excess. Instead, put the comments at the head435 of the function, telling people what it does, and possibly WHY it does436 it.437438 When commenting the kernel API functions, please use the kernel-doc format. 439 See the files Documentation/kernel-doc-nano-HOWTO.txt and scripts/kernel-doc 440for details.441442 Linux style for comments is the C89 "/* ... */" style.443 Don't use C99-style "// ..." comments.444445 The preferred style for long (multi-line) comments is:446447/*448 * This is the preferred style for multi-line449 * comments in the Linux kernel source code.450 * Please use it consistently.451 *452 * Description: A column of asterisks on the left side,453 * with beginning and ending almost-blank lines.454*/455456 For files in net/ and drivers/net/ the preferred style for long (multi-line)457 comments is a little different.458459/* The preferred comment style for files in net/ and drivers/net460 * looks like this.461 *462 * It is nearly the same as the generally preferred comment style,463 * but there is no initial almost-blank line.464*/465466 It's also important to comment data, whether they are basic types or derived 467 types. To this end, use just one data declaration per line (no commas for468 multiple data declarations). This leaves you room for a small comment on each 469 item, explaining its use.472 Chapter 9: You've made a mess of it473474 That's OK, we all do. You've probably been told by your long-time Unix 475 user helper that "GNU emacs" automatically formats the C sources for476 you, and you've noticed that yes, it does do that, but the defaults it477 uses are less than desirable (in fact, they are worse than random478 typing - an infinite number of monkeys typing into GNU emacs would never 479 make a good program).480481 So, you can either get rid of GNU emacs, or change it to use saner482 values. To do the latter, you can stick the following in your .emacs file: 483484 (defun c-lineup-arglist-tabs-only (ignored)485"Line up argument lists by tabs, not spaces"486 (let* ((anchor (c-langelem-pos c-syntactic-element))487 (column (c-langelem-2nd-pos c-syntactic-element))488 (offset (- (1+ column) anchor))489 (steps (floor offset c-basic-offset)))490 (* (max steps 1)491 c-basic-offset)))492493 (add-hook 'c-mode-common-hook494 (lambda ()495 ;; Add kernel style496 (c-add-style497"linux-tabs-only"498'("linux" (c-offsets-alist499 (arglist-cont-nonempty500 c-lineup-gcc-asm-reg501 c-lineup-arglist-tabs-only))))))502503 (add-hook 'c-mode-hook504 (lambda ()505 (let ((filename (buffer-file-name)))506 ;; Enable kernel mode for the appropriate files507 (when (and filename508 (string-match (expand-file-name "~/src/linux-trees")509 filename))510 (setq indent-tabs-mode t)511 (c-set-style "linux-tabs-only")))))512513 This will make emacs go better with the kernel coding style for C514 files below ~/src/linux-trees.515516 But even if you fail in getting emacs to do sane formatting, not517 everything is lost: use "indent".518519 Now, again, GNU indent has the same brain-dead settings that GNU emacs 520 has, which is why you need to give it a few command line options.521 However, that's not too bad, because even the makers of GNU indent522 recognize the authority of K&R (the GNU people aren't evil, they are523 just severely misguided in this matter), so you just give indent the524 options "-kr -i8" (stands for"K&R, 8 character indents"), or use 525"scripts/Lindent", which indents in the latest style.526527"indent" has a lot of options, and especially when it comes to comment 528 re-formatting you may want to take a look at the man page. But529 remember: "indent"is not a fix for bad programming.530531532 Chapter 10: Kconfig configuration files533534 For all of the Kconfig* configuration files throughout the source tree,535 the indentation is somewhat different. Lines under a "config" definition536 are indented with one tab, while help text is indented an additional two537 spaces. Example:538539 config AUDIT540bool"Auditing support"541 depends on NET542 help543 Enable auditing infrastructure that can be used with another544 kernel subsystem, such as SELinux (which requires this for545 logging of avc messages output). Does not do system-call546 auditing without CONFIG_AUDITSYSCALL.547548 Features that might still be considered unstable should be defined as549 dependent on "EXPERIMENTAL":550551 config SLUB552 depends on EXPERIMENTAL && !ARCH_USES_SLAB_PAGE_STRUCT 553bool"SLUB (Unqueued Allocator)"556while seriously dangerous features (such as write support for certain557 filesystems) should advertise this prominently in their prompt string:558559 config ADFS_FS_RW560bool"ADFS write support (DANGEROUS)"561 depends on ADFS_FS562 ...563564 For full documentation on the configuration files, see the file565 Documentation/kbuild/kconfig-language.txt.566567568 Chapter 11: Data structures569570 Data structures that have visibility outside the single-threaded571 environment they are created and destroyed in should always have572 reference counts. In the kernel, garbage collection doesn't exist (and573 outside the kernel garbage collection is slow and inefficient), which574 means that you absolutely _have_ to reference count all your uses.575576 Reference counting means that you can avoid locking, and allows multiple 577 users to have access to the data structure in parallel - and not having578 to worry about the structure suddenly going away from under them just579 because they slept or did something else for a while.580581 Note that locking is _not_ a replacement for reference counting.582 Locking is used to keep data structures coherent, while reference583 counting is a memory management technique. Usually both are needed, and 584 they are not to be confused with each other.585586 Many data structures can indeed have two levels of reference counting,587 when there are users of different "classes". The subclass count counts588 the number of subclass users, and decrements the global count just once 589 when the subclass count goes to zero.590591 Examples of this kind of "multi-level-reference-counting" can be found in592 memory management ("struct mm_struct": mm_users and mm_count), and in 593 filesystem code ("struct super_block": s_count and s_active).594595 Remember: if another thread can find your data structure, and you don't596 have a reference count on it, you almost certainly have a bug.597598599 Chapter 12: Macros, Enums and RTL600601 Names of macros defining constants and labels in enums are capitalized. 602603#define CONSTANT 0x12345604605 Enums are preferred when defining several related constants.606607 CAPITALIZED macro names are appreciated but macros resembling functions 608 may be named in lower case.609610 Generally, inline functions are preferable to macros resembling functions. 611612 Macros with multiple statements should be enclosed in a do - while block: 613614#define macrofun(a, b, c) \615do { \616if (a == 5) \617 do_this(b, c); \618 } while (0)619620 Things to avoid when using macros:6216221) macros that affect control flow:623624#define FOO(x) \625do { \626if (blah(x) < 0) \627return -EBUGGERED; \628 } while(0)629630is a _very_ bad idea. It looks like a function call but exits the "calling"631 function; don't break the internal parsers of those who will read the code.6326332) macros that depend on having a local variable with a magic name:634635#define FOO(val) bar(index, val)636637 might look like a good thing, but it's confusing as hell when one reads the6403) macros with arguments that are used as l-values: FOO(x) = y; will641 bite you if somebody e.g. turns FOO into an inline function.6426434) forgetting about precedence: macros defining constants using expressions644 must enclose the expression in parentheses. Beware of similar issues with645 macros using parameters.646647#define CONSTANT 0x4000648#define CONSTEXP (CONSTANT | 3)649650 The cpp manual deals with macros exhaustively. The gcc internals manual also 651 covers RTL which is used frequently with assembly language in the kernel.652653654 Chapter 13: Printing kernel messages655656 Kernel developers like to be seen as literate. Do mind the spelling657 of kernel messages to make a good impression. Do not use crippled658 words like "dont"; use "do not" or "don't" instead. Make the messages659 concise, clear, and unambiguous.660661 Kernel messages do not have to be terminated with a period.662663 Printing numbers in parentheses (%d) adds no value and should be avoided.664665 There are a number of driver model diagnostic macros in <linux/device.h>666 which you should use to make sure messages are matched to the right device 667 and driver, and are tagged with the right level: dev_err(), dev_warn(),668 dev_info(), and so forth. For messages that aren't associated with a669 particular device, <linux/printk.h> defines pr_debug() and pr_info().670671 Coming up with good debugging messages can be quite a challenge; and once 672 you have them, they can be a huge help for remote troubleshooting. Such673 messages should be compiled out when the DEBUG symbol is not defined (that 674is, by default they are not included). When you use dev_dbg() or pr_debug(),675 that's automatic. Many subsystems have Kconfig options to turn on -DDEBUG. 676 A related convention uses VERBOSE_DEBUG to add dev_vdbg() messages to the 677 ones already enabled by DEBUG.678679680 Chapter 14: Allocating memory681682 The kernel provides the following general purpose memory allocators:683 kmalloc(), kzalloc(), kmalloc_array(), kcalloc(), vmalloc(), and684 vzalloc(). Please refer to the API documentation for further information685 about them.686687 The preferred form for passing a size of a struct is the following:688689 p = kmalloc(sizeof(*p), ...);690691 The alternative form where struct name is spelled out hurts readability and692 introduces an opportunity for a bug when the pointer variable type is changed693 but the corresponding sizeof that is passed to a memory allocator is not.694695 Casting the return value which is a void pointer is redundant. The conversion696from void pointer to any other pointer type is guaranteed by the C programming 697 language.698699 The preferred form for allocating an array is the following:700701 p = kmalloc_array(n, sizeof(...), ...);702703 The preferred form for allocating a zeroed array is the following:704705 p = kcalloc(n, sizeof(...), ...);706707 Both forms check for overflow on the allocation size n * sizeof(...),708 and return NULL if that occurred.709710711 Chapter 15: The inline disease712713 There appears to be a common misperception that gcc has a magic "make me 714 faster" speedup option called "inline". While the use of inlines can be715 appropriate (for example as a means of replacing macros, see Chapter 12), it716 very often is not. Abundant use of the inline keyword leads to a much bigger717 kernel, which in turn slows the system as a whole down, due to a bigger718 icache footprint for the CPU and simply because there is less memory719 available for the pagecache. Just think about it; a pagecache miss causes a720 disk seek, which easily takes 5 milliseconds. There are a LOT of cpu cycles721 that can go into these 5 milliseconds.。
linux 的编码格式
linux 的编码格式Linux的编码格式简介：Linux是一种开源操作系统，广泛应用于各个领域。
在Linux系统中，编码格式是一个关键的概念，它决定了如何表示和存储文本数据。
本文将介绍常见的Linux编码格式及其特点，以及在使用Linux系统时应该注意的问题。
一、ASCII编码ASCII（American Standard Code for Information Interchange）是最早的字符编码标准，在计算机中使用7位二进制数来表示字符。
它定义了128个字符，包括大写和小写字母、数字以及一些常见的符号，如英文标点符号和控制字符。
由于ASCII编码只能表示有限的字符集合，不能满足国际化需求，因此逐渐被更全面的编码格式所取代。
二、UTF-8编码UTF-8（Unicode Transformation Format-8bit）是一种可变长度的Unicode编码标准，支持全球范围内的字符集。
UTF-8编码在Linux系统中被广泛使用，因为它能够兼容ASCII编码。
UTF-8使用1至4个字节来表示一个字符，具备良好的兼容性和可扩展性。
它可以表示几乎所有的字符，包括汉字、拉丁字母、希腊字母等。
UTF-8编码是Linux系统默认的字符编码格式。
三、GBK编码GBK编码是中国国家标准的字符集，它是对GB2312编码的扩充，能够覆盖更多的中文字符。
GBK编码使用2个字节表示一个字符，其中包含了21003个汉字和符号以及部分非汉字字符。
在Linux系统中，如果需要处理中文字符的文本数据，可以选择使用GBK编码来存储和表示。
四、其他编码格式除了ASCII、UTF-8和GBK编码外，Linux还支持其他一些编码格式，如UTF-16、ISO-8859等。
这些编码格式通常用于特定的情况和需求，比如在处理特定语言或特定文件格式时使用。
五、在Linux系统中选择编码格式的注意事项1. 考虑文本数据的应用场景和需求。
- 1、下载文档前请自行甄别文档内容的完整性，平台不提供额外的编辑、内容补充、找答案等附加服务。
- 2、"仅部分预览"的文档,不可在线预览部分如存在完整性等问题,可反馈申请退款(可完整预览的文档不适用该条件!)。
- 3、如文档侵犯您的权益，请联系客服反馈,我们会尽快为您处理(人工客服工作时间：9:00-18:30)。
这是翻译版本，英文原版是linux源码Documentation文件夹下的CodingStyle一个良好风格的程序看起来直观、美观，便于阅读，还能有助于对程序的理解，特别在代码量比较大情况下更显现编码素质的重要性。
相反没有良好的风格的代码读起来难看、晦涩，甚至有时候一个括号没对齐就能造成对程序的曲解或者不理解。
我曾经就遇见过这样的情况，花费了很多不必要的时间在程序的上下文对照上，还debug了半天没理解的程序。
后来直接用indent -kr -i8给他转换格式来看了。
特此转过来一个关于代码风格的帖子分享一下～Linux内核编码风格这是一份简短的，描述linux内核首选编码风格的文档。
编码风格是很个人化的东西，而且我也不愿意把我的观点强加给任何人，不过这里所讲述的是我必须要维护的代码所遵守的风格，并且我也希望绝大多数其他代码也能遵守这个风格。
所以请至少考虑一下本文所述的观点。
首先，我建议你打印一份GNU的编码规范，然后不要读它。
烧掉它，这是一个很高调的具有象征意义的姿态。
Anyway, here goes:第一章：缩进制表符是8个字符，所以缩进也是8个字符。
有些异端运动试图将缩进变为4（乃至2）个字符深，这跟尝试着将圆周率PI的值定义为3没什么两样。
理由：缩进的全部意义就在于清楚的定义一个控制块起止于何处。
尤其是当你盯着你的屏幕连续看了20小时之后，你将会发现大一点的缩进将会使你更容易分辨缩进。
现在，有些人会抱怨8个字符的缩进会使代码向右边移动的太远，在80个字符的终端屏幕上就很难读这样的代码。
这个问题的答案是，如果你需要3级以上的缩进，不管缩进深度如何你的代码已经有问题了，应该修正你的程序。
简而言之，8个字符的缩进可以让代码更容易阅读，还有一个好处是当你的函数嵌套太深的时候可以向你提出告警。
请留意这个警告。
在switch语句中消除多级缩进的首选的方式是让“switch”和从属于它的“case”标签对齐于同一列，而不要“两次缩进”“case”标签。
比如：switch (suffix) {case 'G':case 'g':mem <<= 30;break;case 'M':case 'm':mem <<= 20;break;case 'K':case 'k':mem <<= 10;/* fall through */default:break;}不要把多个语句放在一行里，除非你有什么东西要隐藏：if (condition) do_this;do_something_everytime;也不要在一行里放多个赋值语句。
内核编码风格超级简单。
就是请避免使用怪异的表达式。
除了注释、文档和Kconfig之外，不要使用空格来缩进，前面的例子是例外，是有意为之。
选用一个好的编辑器，不要在行尾留空格。
第二章：把长的行和字符串打散编码风格的意义就在于使用平常使用的工具来维持代码的可读性和可维护性。
每一行的长度的限制是80列，我们强烈建议您遵守这个惯例。
长于80列的语句要打散成有意义的片段。
每个片段要明显短于原来的语句，而且放置的位置也明显的靠右。
同样的规则也适用于有很长参数列表的函数头。
长字符串也要打散成较短的字符串。
唯一的例外是超过80列可以大幅度提高可读性并且不会隐藏信息的情况。
void fun(int a, int b, int c){if (condition)printk(KERN_WARNING "Warning this is a long printk with ""3 parameters a: %u b: %u ""c: %u \n", a, b, c);elsenext_statement;}第三章：大括号和空格的放置C语言风格中另外一个常见问题是大括号的放置。
和缩进大小不同，选择或弃用某种放置策略并没有多少技术上的原因，不过首选的方式，就像Kernighan和Ritchie展示给我们的，是把起始大括号放在行尾，而把结束大括号放在行首，所以：if (x is true) {we do y}这适用于所有的非函数语句块（if、switch、for、while、do）。
比如：switch (action) {case KOBJ_ADD:return "add";case KOBJ_REMOVE:return "remove";case KOBJ_CHANGE:return "change";default:return NULL;}不过，有一种特殊情况，命名函数：它们的起始大括号放置于下一行的开头，这样：int function(int x){body of function}全世界的异端可能会抱怨这个不一致性，呃…确实是不一致的，不过所有思维健全的人都知道（a）K&R是正确的，并且（b）K&R是正确的。
另外，不管怎样函数都是特殊的（在C语言中，函数是不能嵌套的）。
注意结束大括号独自占据一行，除非它后面跟着同一个语句的剩余部分，比如说do语句中的“while”或者if语句中的“else”，像这样：do {body of do-loop} while (condition);和if (x == y) {..} else if (x > y) {...} else {....}理由：K&R。
也请注意这种大括号的放置方式还能使空（或者差不多空的）行的数量最小化，同时不失可读性。
因此，由于你的屏幕上的新行的供应不是可回收的资源（想想25行的终端屏幕），你将会有更多的空行来放置注释。
仅有一个单独的语句时，不用加不必要的大括号。
if (condition)action();这点不适用于本身为某个条件语句的一个分支的单独语句。
这时应该两个分支里都使用大括号。
if (condition) {do_this();do_that();} else {otherwise();}3.1：空格Linux内核的空格使用方格（主要）取决于它是用于函数还是关键字。
（大多数）关键字后要加一个空格。
值得注意的例外是sizeof、typeof、alignof和__attribute__，这些关键字在一定程度上看起来更像函数（它们在Linux里也常常伴随小括号使用，尽管在C语言里这样的小括号不是必需的，就像“struct fileinfo info”声明过后的“sizeof info”）所以在这些关键字之后放一个空格：if, switch, case, for, do, while但是不在sizeof、typeof、alignof或者__attribute__这些关键字之后放空格。
例如， s = sizeof(struct file);不要在小括号里的表达式两侧加空格。
这是一个反例：s = sizeof( struct file );当声明指针类型或者返回指针类型的函数时，“*”的首选使用方式是使之靠近变量名或者函数名，而不是靠近类型名。
例子：char *linux_banner;unsigned long long memparse(char *ptr, char **retptr);char *match_strdup(substring_t *s);在大多数二元和三元操作符两侧使用一个空格，例如下面所有这些操作符：= + - < > * / % | & ^ <= >= == != ? :但是一元操作符后不要加空格：& * + - ~ ! sizeof typeof alignof __attribute__ defined后缀自增和自减一元操作符前不加空格：++ --前缀自增和自减一元操作符后不加空格：++ --“.”和“->”结构体成员操作符前后不加空格。
不要在行尾留空白。
有些可以自动缩进的编辑器会在新行的行首加入适量的空白，然后你就可以直接在那一行输入代码。
不过假如你最后没有在那一行输入代码，有些编辑器就不会移除已经加入的空白，就像你故意留下一个只有空白的行。
包含行尾空白的行就这样产生了。
当Git发现补丁包含了行尾空白的时候会警告你，并且可以应你的要求去掉行尾空白；不过如果你是正在打一系列补丁，这样做会导致后面的补丁失败，因为你改变了补丁的上下文。
第四章：命名C 是一个简朴的语言，你的命名也应该这样。
和Modula-2和Pascal程序员不同，C程序员不使用类似ThisVariableIsATemporaryCounter这样华丽的名字。
C程序员会称那个变量为“tmp”，这样写起来会更容易，而且至少不会令其难于理解。
不过，虽然混用大小写的名字是不提倡使用的，但是全局变量还是需要一个具描述性的名字。
称一个全局函数为“foo”是一个难以饶恕的错误。
全局变量（只有当你真正需要它们的时候再用它）需要有一个具描述性的名字，就像全局函数。
如果你有一个可以计算活动用户数量的函数，你应该叫它“count_ac tive_users()”或者类似的名字，你不应该叫它“cntuser()”。
在函数名中包含函数类型（所谓的匈牙利命名法）是脑子出了问题——编译器知道那些类型而且能够检查那些类型，这样做只能把程序员弄糊涂了。
难怪微软总是制造出有问题的程序。
本地变量名应该简短，而且能够表达相关的含义。
如果你有一些随机的整数型的循环计数器，它应该被称为“i”。
叫它“loop_counter”并无益处，如果它没有可能被误解的话。
类似的“tmp”可以用来称呼任意类型的临时变量。
如果你怕混淆了你的本地变量名，你就遇到另一个问题了，叫做函数增长荷尔蒙失衡综合症。
请看第六章（函数）。
第五章：Typedef不要使用类似“vps_t”之类的东西。
对结构体和指针使用typedef是一个错误。
当你在代码里看到：vps_t a;这代表什么意思呢？相反，如果是这样struct virtual_container *a;你就知道“a”是什么了。
很多人认为typedef“能提高可读性”。
实际不是这样的。
它们只在下列情况下有用：(a) 完全不透明的对象（这种情况下要主动使用typedef来隐藏这个对象实际上是什么）。
例如：“pte_t”等不透明对象，你只能用合适的访问函数来访问它们。
注意！不透明性和“访问函数本身”是不好的。
我们使用pte_t等类型的原因在于真的是完全没有任何共用的可访问信息。
(b) 清楚的整数类型，这样抽象层就可以帮助我们消除到底是"int"还是"long"的混淆。