Using LD - the GNU linker

Free Software Foundation, ld version 2, last updated : November 07,1998. Copyright(C) 1991~1998 Free Software Foundation.

1 Overview

LD是编译工作的最后一步，利用BFD库将目标文件(*.o)、归档文件(*.a/*.so)整合，以重定位数据(Relocates data)和绑定符号引用(Ties up symbol references)。GNU LD提供命令行和脚本(*.ld/*.lds)环境来配置运行参数，并提供尽可能多的调试信息。

Note

整合分配将输入文件的所有相同SECTION集中归纳并重定位SYMBOL在存储器的位置（绑定符号的绝对引用），最终服务于编译器的执行决策。链接器脚本的实现与命令行功能部分重合，指定使用链接器脚本通过命令行-T选项，未指定链接器脚本时将使用编译器的默认链接库。

2 Command Line

Note

Command Line and Environment Variables are ignored. For more info, click here。

3 Linker Script

链接器脚本由Command和Expressions组成，支持C语言的运算符、符号命名规则、表达式、赋值和注释（/*...*/）等。

3.0 Expressions

所有表达式的整型都为long或unsigned long类型、所有常量都是整型、支持所有C运算符、可以引用/定义/创建全局变量、提供内置函数。整型数值的使用与C相同，并支持以K、M、G结尾的缩略数值格式。符号（变量）类似解释型语言的弱类型变量，但不可以声明，需直接定义使用，符号作为全局变量定义于SECTIONS内/外，由点.、字母、下划线、数字和连字符-组成，并由点、字母或下划线开头，符号名不能与关键字冲突。赋值表达式需以分号结尾。

符号赋值表达式同C（支持所有C的赋值运算符，表达式以分号结尾）。SECTIONS中使用的特殊变量.是定位计数器，表示所引用位置的当前地址值，通过赋值（只能增加）来改变偏移量。

dot .

特殊链接器变量.用于SECTIONS中记录当前地址偏移量，且只能执行累加操作。

3 Linker Script

3.1 Command

3.1.1 ENTRY

    ENTRY(*symbol*) /**< 这里的symbol来自输入文件，通常是函数名（即入口函数）. */

Note

入口地址优先级为（降序）：命令行-e选项、链接脚本ENTRY(*symbol*)命令、链接脚本start符号、.text段的第一个字节的地址、地址0。

3.1.2 MEMORY

    MEMORY
    {
        *name [(attr)]* : ORIGIN = *origin*, LENGTH = *len* /**< name自定义，attr可选（内容选自`ALIRWX`），ORIGIN表起始地址。*/
        ...
    }

链接器默认存储器的所有连续地址都可用，通过MEMORY命令可配置存储器实际利用范围，最终通过name服务于SECTIONS（通过> name的方式将section定向于name区域）。可选的属性值有：a（Allocated sections）、l（Initialzed sections）、i（Initialzed sections）、r（Read-only sections）、w（Read/write sections）、x（Sections containing executable code）。

3.1.3 SECTIONS

SECTIONS用于将输入文件（*.o、*.a、*.so）各段（.text、.data、.bss ...）按序整合并定位到存储器中，如果不定义SECTIONS，则链接器将每个输入部分按第一次在输入文件中遇到的顺序放置到同名的输出部分中。

    SECTIONS
    {
        ...    /**< Symbol或section定义。*/

        *secname [start] [BLOCK(align)] [(NOLOAD)] : [AT ( ldadr )]*  /**< 冒号前后必须有空格。[]内容表可选。*/
        {
            *contents* 
        } *[>region :phdr=fill]*

        ...    /**< Symbol或section定义。*/
    }

Note

链接（运行）地址也是输出段的存储（加载）地址，当使用AT关键字时可将输出段的存储地址重定位，即该输出段在运行时将执行从存储地址到链接地址的拷贝操作（由编译器crt0库来完成）。

4 Others

MRI Compatible Script Files、BFD、Machine Dependent Features。

Appendix-A : X

2 Invocation

2.1 Command Line Options

2.2 Environment Variables

The behavior of ld can be changed with the environment variables GNUTARGET and LDEMULATION.

GNUTARGET determines the input-file object format if you don't use -b (or its synonym --format). Its value should be one of the BFD names for an input format (see section BFD). If there is no GNUTARGET in the environment, ld uses the natural format of the target. If GNUTARGET is set to default then BFD attempts to discover the input format by examining binary input files.

LDEMULATION determines the default emulation if you don't use the -m option. The emulation can affect various aspects of linker behaviour, particularly the default linker script. You can list the available emulations with the --verbose or -V options. If the -m option is not used, and the LDEMULATION environment variable is not defined, the default emulation depends upon how the linker was configured

3 Command Language

The command language provides explicit control over the link process, describing the mapping between the linker's input files and its output. It controls : input files, file formats, output file layout, address of sections, placement of common blocks.

A command file (also known as a linker script) SHOULD be supplied to linker through the -T option.

3.1 Linker Scripts

Every meaningful command script must have a SECTIONS command, it specifies a "picture" of the output file's layout, in varying degrees of detail. No other command is required in all cases.

The optional command MEMORY complements SECTIONS by describing the available memory in the target architecture. if a MEMORY command is not used, ld assumes sufficient memory is available in a contiguous block for all output

Comments /* */ can be used as in C/C++ programming language.

3.2 Expressions

The syntax for expressions in the command language is identical to that of C expressions, with the features:

• All expressions evaluated as integers and are of "long" or "unsigned long" type.
• All constants are integers.
• All of the C arithmetic operators are provided.
• Global variables can be referenced, defined and created.
• Special purpose built-in functions can be called.

3.2.1 Integers

The table refer to the same quantity.

3.2.2 Symbol Names

Symbol names start with a letter, underscore or point and may include any letters, underscore, digits, points and hyphens. Unquoted symbol names must not conflict with any keywords.

3.2.3 The Location Counter

The special linker dot . (NOT .xxx) refers to the current output location counter, and MUST always appear in an expression within a SECTIONS command. Assigning a value to the . symbol will cause the location counter to be moved, and a holes is created in the output section. The location counter may never be moved backwards.

    SECTIONS
    {
        output:
        {
            fileA(.text)
            . =. + 1000    /**< Offset 1000 Bytes.            */
            fileB(.text)
            . =. + 1000    /**< Offset 1000 Bytes.            */
            fileC(.text)

        } = 0X1234;        /**< Fill 0x1234 in two gaps.    */
    }

3.2.4 Operators

The linker recognizes the standard C set of arithmetic operators, with the standard bindings and precedence levels.

3.2.5 Evaluation

The linker uses "lazy evaluation" for expressions, it only calculates an expression when absolutely necessary. The linker needs the value of the start address, and the lengths of memory regions, in order to do any linking at all, these values are computed as soon as possible when the linker reads in the command file. However, other values (such as symbol values) are not known or needed until after storage allocation. Such values are evaluated later, when other information (such as the sizes of output sections) is available for use in the symbol assignment expression.

3.2.6 Assignment: Defining Symbols

3.2.7 Arithmetic Functions

3.2.8 Semicolons

Semicolons ; must appear at the end of assignment expressions and PHDRS statement.

3.3 Memory Layout

3.4 Specifying Output Sections

3.4.1 Section Definitions

3.4.2 Section Placement

3.4.3 Section Data Expressions

3.4.4 Optional Section Attributes

3.4.5 Overlays

3.5 ELF Program Headers

3.6 The Entry Point

3.7 Version Script

3.8 Option Commands

4 Machine Dependent Features

4.1 H8/300

4.2 Intel 960 family

5 BFD

组成：

1. command

   ENTRY, MEMORY, SECTIONS, PHDRS

   SECTIONS {

    secname : { //secname 不是随意的，它一定是跟目标文件相关的格式，比如如果目标文件是a.out，那么secname只能选择.text，.data，.bss，
                这怎么确定呢？到http://www.zap.org.au/elec2041-cdrom/gnutools/doc/gnu-assembler.pdf（gcc汇编）找就可以了
   
                contents
              }

   }

   在SECTION中使用的command

    PROVIDE:https://www.cnblogs.com/tureno/articles/3741291.html
   

1. symbol

   可以独立在command之外，也可以在command之内，类似C语言的变量，只不过没有类型，这类似于脚本语言的变量。

OLD GNU

https://ftp.gnu.org/old-gnu/Manuals/

Linke (ld)

https://ftp.gnu.org/old-gnu/Manuals/ld-2.9.1/html_mono/ld.html

散装知识点

1 EABI

EABI ： Embedded Application Binary Interface ，嵌入式应用程序二进制接口

任何符合EABI标准的编译器，编译得到的二进制文件是彼此兼容的。

ARM EABI，= AEABI，即是ARM平台下的嵌入式应用程序二进制接口

GNU EABI，或GNU AEABI，其实就是指EABI、AEABI，因为这些编译器都是以GNU编译器为基础制作的

arm-cortex_a9-linux-gnueabi

之所以gnu与eabi连写，是因为有些GNU工具要求编译器的配置名称在三个连字符及以内，所以实际上所有的GNU编译器的命名都是A-B-C， 有些GNU编译器也没有按照此方法命名，有可能出现不兼容的情况

ABI for the ARM architecture是很多规范的合集：

AADWARF DWARF for the ARM Architecture

AAELF ELF for the ARM Architecture

AAPCS Procedure Call Standard for the ARM Architecture

ADDENDA Addenda to, and errata in, the ABI for the ARM Architecture

BPABI Base Platform ABI for the ARM Architecture

BSABI ABI for the ARM Architecture (Base Standard)

CLIBABI C Library ABI for the ARM Architecture

CPPABI C++ ABI for the ARM Architecture

EHABI Exception Handling ABI for the ARM Architecture

EHEGI Exception handling component specimen implementations

RTABI Run-time ABI for the ARM Architecture

2 大多数库，编译支持的选项

./configure --prefix=绝对路径 --host=平台

注意，这里的平台是编译器的前缀，如编译器为arm-cortex_a9-linux-gnueabi-gcc，则平台为arm-cortex_a9-linux-gnueabi

3 路径的指定

• 头文件路径

  • C++ 头文件路径
    • 命令行 : export CPLUS_INCLUDE_PATH=绝对路径:$(CPLUS_INCLUDE_PATH)；环境变量名还可能为'CPLUSPLUS_INCLUDE_PATH'、'CXX_INCLUDE_PATH'
    • 编译器选项 ：-I绝对路径
  • C 头文件路径
    • 命令行 ：export C_INCLUDE_PATH=绝对路径:$(C_INCLUDE_PATH)；
    • 编译器选项 ：-I绝对路径

• 库路径

  • 动态库路径

    • 命令行 ：export LD_LIBRARY_PATH=绝对路径:$(LD_LIBRARY_PATH)；环境变量名还可能为'LIBRARY_PATH'
    • 编译器选项 ：-L绝对路径 -l库名；库名是去掉lib和.so的库名称
    • 配置文件 ：echo 绝对路径 > /etc/ld.so.conf, 然后执行 ldconfig

  • 静态库路径 静态库路径的搜索方式与动态库相同

Tips : 对于动态库的编译器选项 -Wl,rpath=xxx，-Wl选项用于将逗号后面的参数传递给连接器ld，-rpath用于指定运行时动态库的搜索路径，而-L只能用于编译时动态库的搜索路径

3.1 调用动态库or静态库？

静态库和动态库的搜索方式是相同的，如果静态库和动态库放在一个目录，那么编译器会调用哪一个呢？

Linux系统下，默认先调用动态库，如果动态库不存在则调用静态库，如果强制使用静态库，可使用编译选项 : -static -l库名

3.2 混合链接

静态链接的库 使用-Wl,-dn -L静态库绝对路径 -l静态库名, 动态链接的库使用-Wl,-dy -L动态库绝对路径 -l动态库名：-Wl,-dn -Lxxx -lyyy -Wl,-dy -Lnnn, -lmmm ，其中-dn可以替换为-non_shared、-Bstatic、-static，其中-dy可以替换为-Bdynamic、-call_shared