Details on this package are located in Section 8.18.2, “Contents of Binutils.”
Copyright © 1999-2022 Gerard Beekmans
Copyright © 1999-2022, Gerard Beekmans
All rights reserved.
This book is licensed under a Creative Commons License.
Computer instructions may be extracted from the book under the MIT License.
Linux® is a registered trademark of Linus Torvalds.
My journey to learn and better understand Linux began back in 1998. I had just installed my first Linux distribution and had quickly become intrigued with the whole concept and philosophy behind Linux.
There are always many ways to accomplish a single task. The same can be said about Linux distributions. A great many have existed over the years. Some still exist, some have morphed into something else, yet others have been relegated to our memories. They all do things differently to suit the needs of their target audience. Because so many different ways to accomplish the same end goal exist, I began to realize I no longer had to be limited by any one implementation. Prior to discovering Linux, we simply put up with issues in other Operating Systems as you had no choice. It was what it was, whether you liked it or not. With Linux, the concept of choice began to emerge. If you didn't like something, you were free, even encouraged, to change it.
I tried a number of distributions and could not decide on any one. They were great systems in their own right. It wasn't a matter of right and wrong anymore. It had become a matter of personal taste. With all that choice available, it became apparent that there would not be a single system that would be perfect for me. So I set out to create my own Linux system that would fully conform to my personal preferences.
To truly make it my own system, I resolved to compile everything from source code instead of using pre-compiled binary packages. This “perfect” Linux system would have the strengths of various systems without their perceived weaknesses. At first, the idea was rather daunting. I remained committed to the idea that such a system could be built.
After sorting through issues such as circular dependencies and compile-time errors, I finally built a custom-built Linux system. It was fully operational and perfectly usable like any of the other Linux systems out there at the time. But it was my own creation. It was very satisfying to have put together such a system myself. The only thing better would have been to create each piece of software myself. This was the next best thing.
As I shared my goals and experiences with other members of the Linux community, it became apparent that there was a sustained interest in these ideas. It quickly became plain that such custom-built Linux systems serve not only to meet user specific requirements, but also serve as an ideal learning opportunity for programmers and system administrators to enhance their (existing) Linux skills. Out of this broadened interest, the Linux From Scratch Project was born.
This Linux From Scratch book is the central core around that project. It provides the background and instructions necessary for you to design and build your own system. While this book provides a template that will result in a correctly working system, you are free to alter the instructions to suit yourself, which is, in part, an important part of this project. You remain in control; we just lend a helping hand to get you started on your own journey.
I sincerely hope you will have a great time working on your own Linux From Scratch system and enjoy the numerous benefits of having a system that is truly your own.
--
Gerard Beekmans
gerard AT linuxfromscratch D0T org
There are many reasons why you would want to read this book. One of the questions many people raise is, “why go through all the hassle of manually building a Linux system from scratch when you can just download and install an existing one?”
One important reason for this project's existence is to help you learn how a Linux system works from the inside out. Building an LFS system helps demonstrate what makes Linux tick, and how things work together and depend on each other. One of the best things that this learning experience can provide is the ability to customize a Linux system to suit your own unique needs.
Another key benefit of LFS is that it allows you to have more control over the system without relying on someone else's Linux implementation. With LFS, you are in the driver's seat and dictate every aspect of the system.
LFS allows you to create very compact Linux systems. When installing regular distributions, you are often forced to install a great many programs which are probably never used or understood. These programs waste resources. You may argue that with today's hard drive and CPUs, such resources are no longer a consideration. Sometimes, however, you are still constrained by size considerations if nothing else. Think about bootable CDs, USB sticks, and embedded systems. Those are areas where LFS can be beneficial.
Another advantage of a custom built Linux system is security. By compiling the entire system from source code, you are empowered to audit everything and apply all the security patches desired. It is no longer necessary to wait for somebody else to compile binary packages that fix a security hole. Unless you examine the patch and implement it yourself, you have no guarantee that the new binary package was built correctly and adequately fixes the problem.
The goal of Linux From Scratch is to build a complete and usable foundation-level system. If you do not wish to build your own Linux system from scratch, you may nevertheless benefit from the information in this book.
There are too many other good reasons to build your own LFS system to list them all here. In the end, education is by far the most powerful of reasons. As you continue in your LFS experience, you will discover the power that information and knowledge truly bring.
The primary target architectures of LFS are the AMD/Intel x86 (32-bit) and x86_64 (64-bit) CPUs. On the other hand, the instructions in this book are also known to work, with some modifications, with the Power PC and ARM CPUs. To build a system that utilizes one of these CPUs, the main prerequisite, in addition to those on the next page, is an existing Linux system such as an earlier LFS installation, Ubuntu, Red Hat/Fedora, SuSE, or other distribution that targets the architecture that you have. Also note that a 32-bit distribution can be installed and used as a host system on a 64-bit AMD/Intel computer.
For building LFS, the gain of building on a 64-bit system compared to a 32-bit system is minimal. For example, in a test build of LFS-9.1 on a Core i7-4790 CPU based system, using 4 cores, the following statistics were measured:
Architecture Build Time Build Size
32-bit 239.9 minutes 3.6 GB
64-bit 233.2 minutes 4.4 GB
As you can see, on the same hardware, the 64-bit build is only 3% faster and is 22% larger than the 32-bit build. If you plan to use LFS as a LAMP server, or a firewall, a 32-bit CPU may be largely sufficient. On the other hand, several packages in BLFS now need more than 4GB of RAM to be built and/or to run, so that if you plan to use LFS as a desktop, the LFS authors recommend building on a 64-bit system.
The default 64-bit build that results from LFS is considered a “pure” 64-bit system. That is, it supports 64-bit executables only. Building a “multi-lib” system requires compiling many applications twice, once for a 32-bit system and once for a 64-bit system. This is not directly supported in LFS because it would interfere with the educational objective of providing the instructions needed for a straightforward base Linux system. Some LFS/BLFS editors maintain a fork of LFS for multilib, which is accessible at https://www.linuxfromscratch.org/~thomas/multilib/index.html. But it is an advanced topic.
Building an LFS system is not a simple task. It requires a certain level of existing knowledge of Unix system administration in order to resolve problems and correctly execute the commands listed. In particular, as an absolute minimum, you should already have the ability to use the command line (shell) to copy or move files and directories, list directory and file contents, and change the current directory. It is also expected that you have a reasonable knowledge of using and installing Linux software.
Because the LFS book assumes at least this basic level of skill, the various LFS support forums are unlikely to be able to provide you with much assistance in these areas. You will find that your questions regarding such basic knowledge will likely go unanswered or you will simply be referred to the LFS essential pre-reading list.
Before building an LFS system, we recommend reading the following:
Software-Building-HOWTO http://www.tldp.org/HOWTO/Software-Building-HOWTO.html
This is a comprehensive guide to building and installing “generic” Unix software packages under Linux. Although it was written some time ago, it still provides a good summary of the basic techniques needed to build and install software.
Beginner's Guide to Installing from Source http://moi.vonos.net/linux/beginners-installing-from-source/
This guide provides a good summary of basic skills and techniques needed to build software from source code.
The structure of LFS follows Linux standards as closely as possible. The primary standards are:
Linux Standard Base (LSB) Version 5.0 (2015)
The LSB has four separate standards: Core, Desktop, Runtime Languages, and Imaging. In addition to generic requirements there are also architecture specific requirements. There are also two areas for trial use: Gtk3 and Graphics. LFS attempts to conform to the architectures discussed in the previous section.
Many people do not agree with the requirements of the LSB. The main purpose of defining it is to ensure that proprietary software will be able to be installed and run properly on a compliant system. Since LFS is source based, the user has complete control over what packages are desired and many choose not to install some packages that are specified by the LSB.
Creating a complete LFS system capable of passing the LSB certifications tests is possible, but not without many additional packages that are beyond the scope of LFS. These additional packages have installation instructions in BLFS.
LSB Core: |
Bash, Bc, Binutils, Coreutils, Diffutils, File, Findutils, Gawk, Grep, Gzip, M4, Man-DB, Ncurses, Procps, Psmisc, Sed, Shadow, Tar, Util-linux, Zlib |
LSB Desktop: |
None |
LSB Runtime Languages: |
Perl, Python |
LSB Imaging: |
None |
LSB Gtk3 and LSB Graphics (Trial Use): |
None |
LSB Core: |
At, Batch (a part of At), Cpio, Ed, Fcrontab, LSB-Tools, NSPR, NSS, PAM, Pax, Sendmail (or Postfix or Exim), time |
LSB Desktop: |
Alsa, ATK, Cairo, Desktop-file-utils, Freetype, Fontconfig, Gdk-pixbuf, Glib2, GTK+2, Icon-naming-utils, Libjpeg-turbo, Libpng, Libtiff, Libxml2, MesaLib, Pango, Xdg-utils, Xorg |
LSB Runtime Languages: |
Libxml2, Libxslt |
LSB Imaging: |
CUPS, Cups-filters, Ghostscript, SANE |
LSB Gtk3 and LSB Graphics (Trial Use): |
GTK+3 |
As stated earlier, the goal of LFS is to build a complete and usable foundation-level system. This includes all packages needed to replicate itself while providing a relatively minimal base from which to customize a more complete system based on the choices of the user. This does not mean that LFS is the smallest system possible. Several important packages are included that are not strictly required. The lists below document the rationale for each package in the book.
Acl
This package contains utilities to administer Access Control Lists, which are used to define more fine-grained discretionary access rights for files and directories.
Attr
This package contains programs for administering extended attributes on filesystem objects.
Autoconf
This package contains programs for producing shell scripts that can automatically configure source code from a developer's template. It is often needed to rebuild a package after updates to the build procedures.
Automake
This package contains programs for generating Make files from a template. It is often needed to rebuild a package after updates to the build procedures.
Bash
This package satisfies an LSB core requirement to provide a Bourne Shell interface to the system. It was chosen over other shell packages because of its common usage and extensive capabilities beyond basic shell functions.
Bc
This package provides an arbitrary precision numeric processing language. It satisfies a requirement needed when building the Linux kernel.
Binutils
This package contains a linker, an assembler, and other tools for handling object files. The programs in this package are needed to compile most of the packages in an LFS system and beyond.
Bison
This package contains the GNU version of yacc (Yet Another Compiler Compiler) needed to build several other LFS programs.
Bzip2
This package contains programs for compressing and decompressing files. It is required to decompress many LFS packages.
Check
This package contains a test harness for other programs.
Coreutils
This package contains a number of essential programs for viewing and manipulating files and directories. These programs are needed for command line file management, and are necessary for the installation procedures of every package in LFS.
DejaGNU
This package contains a framework for testing other programs.
Diffutils
This package contains programs that show the differences between files or directories. These programs can be used to create patches, and are also used in many packages' build procedures.
E2fsprogs
This package contains the utilities for handling the ext2, ext3 and ext4 file systems. These are the most common and thoroughly tested file systems that Linux supports.
Eudev
This package is a device manager. It dynamically controls the ownership, permissions, names, and symbolic links of devices in the /dev directory as devices are added or removed from the system.
Expat
This package contains a relatively small XML parsing library. It is required by the XML::Parser Perl module.
Expect
This package contains a program for carrying out scripted dialogues with other interactive programs. It is commonly used for testing other packages.
File
This package contains a utility for determining the type of a given file or files. A few packages need it in their build scripts.
Findutils
This package contains programs to find files in a file system. It is used in many packages' build scripts.
Flex
This package contains a utility for generating programs that recognize patterns in text. It is the GNU version of the lex (lexical analyzer) program. It is required to build several LFS packages.
Gawk
This package contains programs for manipulating text files. It is the GNU version of awk (Aho-Weinberg-Kernighan). It is used in many other packages' build scripts.
GCC
This package is the Gnu Compiler Collection. It contains the C and C++ compilers as well as several others not built by LFS.
GDBM
This package contains the GNU Database Manager library. It is used by one other LFS package, Man-DB.
Gettext
This package contains utilities and libraries for internationalization and localization of numerous packages.
Glibc
This package contains the main C library. Linux programs will not run without it.
GMP
This package contains math libraries that provide useful functions for arbitrary precision arithmetic. It is required to build GCC.
Gperf
This package contains a program that generates a perfect hash function from a key set. It is required for Eudev.
Grep
This package contains programs for searching through files. These programs are used by most packages' build scripts.
Groff
This package contains programs for processing and formatting text. One important function of these programs is to format man pages.
GRUB
This package is the Grand Unified Boot Loader. It is one of several boot loaders available, but is the most flexible.
Gzip
This package contains programs for compressing and decompressing files. It is needed to decompress many packages in LFS and beyond.
Iana-etc
This package provides data for network services and protocols. It is needed to enable proper networking capabilities.
Inetutils
This package contains programs for basic network administration.
Intltool
This package contains tools for extracting translatable strings from source files.
IProute2
This package contains programs for basic and advanced IPv4 and IPv6 networking. It was chosen over the other common network tools package (net-tools) for its IPv6 capabilities.
Kbd
This package contains key-table files, keyboard utilities for non-US keyboards, and a number of console fonts.
Kmod
This package contains programs needed to administer Linux kernel modules.
Less
This package contains a very nice text file viewer that allows scrolling up or down when viewing a file. It is also used by Man-DB for viewing manpages.
Libcap
This package implements the user-space interfaces to the POSIX 1003.1e capabilities available in Linux kernels.
Libelf
The elfutils project provides libraries and tools for ELF files and DWARF data. Most utilities in this package are available in other packages, but the library is needed to build the Linux kernel using the default (and most efficient) configuration.
Libffi
This package implements a portable, high level programming interface to various calling conventions. Some programs may not know at the time of compilation what arguments are to be passed to a function. For instance, an interpreter may be told at run-time about the number and types of arguments used to call a given function. Libffi can be used in such programs to provide a bridge from the interpreter program to compiled code.
Libpipeline
The Libpipeline package contains a library for manipulating pipelines of subprocesses in a flexible and convenient way. It is required by the Man-DB package.
Libtool
This package contains the GNU generic library support script. It wraps the complexity of using shared libraries in a consistent, portable interface. It is needed by the test suites in other LFS packages.
Linux Kernel
This package is the Operating System. It is the Linux in the GNU/Linux environment.
M4
This package contains a general text macro processor useful as a build tool for other programs.
Make
This package contains a program for directing the building of packages. It is required by almost every package in LFS.
Man-DB
This package contains programs for finding and viewing man pages. It was chosen instead of the man package due to superior internationalization capabilities. It supplies the man program.
Man-pages
This package contains the actual contents of the basic Linux man pages.
Meson
This package provides a software tool for automating the building of software. The main goal for Meson is to minimize the amount of time that software developers need to spend configuring their build system. It's required to build Systemd, as well as many BLFS packages.
MPC
This package contains functions for the arithmetic of complex numbers. It is required by GCC.
MPFR
This package contains functions for multiple precision arithmetic. It is required by GCC.
Ninja
This package contains a small build system with a focus on speed. It is designed to have its input files generated by a higher-level build system, and to run builds as fast as possible. This package is required by Meson.
Ncurses
This package contains libraries for terminal-independent handling of character screens. It is often used to provide cursor control for a menuing system. It is needed by a number of packages in LFS.
Openssl
This package provides management tools and libraries relating to cryptography. These are useful for providing cryptographic functions to other packages, including the Linux kernel.
Patch
This package contains a program for modifying or creating files by applying a patch file typically created by the diff program. It is needed by the build procedure for several LFS packages.
Perl
This package is an interpreter for the runtime language PERL. It is needed for the installation and test suites of several LFS packages.
Pkg-config
This package provides a program that returns meta-data about an installed library or package.
Procps-NG
This package contains programs for monitoring processes. These programs are useful for system administration, and are also used by the LFS Bootscripts.
Psmisc
This package contains programs for displaying information about running processes. These programs are useful for system administration.
Python 3
This package provides an interpreted language that has a design philosophy that emphasizes code readability.
Readline
This package is a set of libraries that offers command-line editing and history capabilities. It is used by Bash.
Sed
This package allows editing of text without opening it in a text editor. It is also needed by most LFS packages' configure scripts.
Shadow
This package contains programs for handling passwords in a secure way.
Sysklogd
This package contains programs for logging system messages, such as those given by the kernel or daemon processes when unusual events occur.
Sysvinit
This package provides the init program, which is the parent of all other processes on the Linux system.
Tar
This package provides archiving and extraction capabilities of virtually all packages used in LFS.
Tcl
This package contains the Tool Command Language used in many test suites in LFS packages.
Texinfo
This package contains programs for reading, writing, and converting info pages. It is used in the installation procedures of many LFS packages.
Util-linux
This package contains miscellaneous utility programs. Among them are utilities for handling file systems, consoles, partitions, and messages.
Vim
This package contains an editor. It was chosen because of its compatibility with the classic vi editor and its huge number of powerful capabilities. An editor is a very personal choice for many users and any other editor could be substituted if desired.
XML::Parser
This package is a Perl module that interfaces with Expat.
XZ Utils
This package contains programs for compressing and decompressing files. It provides the highest compression generally available and is useful for decompressing packages in XZ or LZMA format.
Zlib
This package contains compression and decompression routines used by some programs.
Zstd
This package contains compression and decompression routines used by some programs. It provides high compression ratios and a very wide range of compression / speed trade-offs.
To make things easier to follow, there are a few typographical conventions used throughout this book. This section contains some examples of the typographical format found throughout Linux From Scratch.
./configure --prefix=/usr
This form of text is designed to be typed exactly as seen unless otherwise noted in the surrounding text. It is also used in the explanation sections to identify which of the commands is being referenced.
In some cases, a logical line is extended to two or more physical lines with a backslash at the end of the line.
CC="gcc -B/usr/bin/" ../binutils-2.18/configure \ --prefix=/tools --disable-nls --disable-werror
Note that the backslash must be followed by an immediate return. Other whitespace characters like spaces or tab characters will create incorrect results.
install-info: unknown option '--dir-file=/mnt/lfs/usr/info/dir'
This form of text (fixed-width text) shows screen output, usually
as the result of commands issued. This format is also used to
show filenames, such as /etc/ld.so.conf
.
Emphasis
This form of text is used for several purposes in the book. Its main purpose is to emphasize important points or items.
https://www.linuxfromscratch.org/
This format is used for hyperlinks both within the LFS community and to external pages. It includes HOWTOs, download locations, and websites.
cat > $LFS/etc/group << "EOF"
root:x:0:
bin:x:1:
......
EOF
This format is used when creating configuration files. The first
command tells the system to create the file $LFS/etc/group
from whatever is typed on the
following lines until the sequence End Of File (EOF) is
encountered. Therefore, this entire section is generally typed as
seen.
<REPLACED TEXT>
This format is used to encapsulate text that is not to be typed as seen or for copy-and-paste operations.
[OPTIONAL TEXT]
This format is used to encapsulate text that is optional.
passwd(5)
This format is used to refer to a specific manual (man) page. The
number inside parentheses indicates a specific section inside the
manuals. For example, passwd has two man pages. Per
LFS installation instructions, those two man pages will be
located at /usr/share/man/man1/passwd.1
and /usr/share/man/man5/passwd.5
. When the book
uses passwd(5)
it is specifically
referring to /usr/share/man/man5/passwd.5
. man passwd will print the first
man page it finds that matches “passwd”,
which will be /usr/share/man/man1/passwd.1
. For this example,
you will need to run man 5
passwd in order to read the page being specified.
Note that most man pages do not have duplicate page names in
different sections. Therefore, man
<program
name>
is generally sufficient.
This book is divided into the following parts.
Part I explains a few important notes on how to proceed with the LFS installation. This section also provides meta-information about the book.
Part II describes how to prepare for the building process—making a partition, downloading the packages, and compiling temporary tools.
Part III provides instructions for building the tools needed for constructing the final LFS system.
Part IV guides the reader through the building of the LFS system—compiling and installing all the packages one by one, setting up the boot scripts, and installing the kernel. The resulting Linux system is the foundation on which other software can be built to expand the system as desired. At the end of this book, there is an easy to use reference listing all of the programs, libraries, and important files that have been installed.
Part V provides information about the book itself including acronyms and terms, acknowledgments, package dependencies, a listing of LFS boot scripts, licenses for the distribution of the book, and a comprehensive index of packages, programs, libraries, and scripts.
The software used to create an LFS system is constantly being updated and enhanced. Security warnings and bug fixes may become available after the LFS book has been released. To check whether the package versions or instructions in this release of LFS need any modifications to accommodate security vulnerabilities or other bug fixes, please visit https://www.linuxfromscratch.org/lfs/errata/11.1/ before proceeding with your build. You should note any changes shown and apply them to the relevant section of the book as you progress with building the LFS system.
In addition, the Linux From Scratch editors maintain a list of security vulnerabilities discovered after a book was released. To check whether there are any active security vulnerabilities, please visit https://www.linuxfromscratch.org/lfs/advisories/ prior to proceeding with your build. You should note any advisories and perform the steps to fix any security vulnerabilities as you progress with building the LFS system.
The LFS system will be built by using an already installed Linux distribution (such as Debian, OpenMandriva, Fedora, or openSUSE). This existing Linux system (the host) will be used as a starting point to provide necessary programs, including a compiler, linker, and shell, to build the new system. Select the “development” option during the distribution installation to be able to access these tools.
As an alternative to installing a separate distribution onto your machine, you may wish to use a LiveCD from a commercial distribution.
Chapter 2 of this book describes how to create a new Linux native partition and file system. This is the place where the new LFS system will be compiled and installed. Chapter 3 explains which packages and patches need to be downloaded to build an LFS system and how to store them on the new file system. Chapter 4 discusses the setup of an appropriate working environment. Please read Chapter 4 carefully as it explains several important issues you need be aware of before beginning to work your way through Chapter 5 and beyond.
Chapter 5, explains the installation of the initial tool chain, (binutils, gcc, and glibc) using cross compilation techniques to isolate the new tools from the host system.
Chapter 6 shows you how to cross-compile basic utilities using the just built cross-toolchain.
Chapter 7 then enters a "chroot" environment and uses the previously built tools to build the additional tools needed to build and test the final system.
This effort to isolate the new system from the host distribution may seem excessive. A full technical explanation as to why this is done is provided in Toolchain Technical Notes.
In Chapter 8, The full LFS system is built. Another advantage provided by the chroot environment is that it allows you to continue using the host system while LFS is being built. While waiting for package compilations to complete, you can continue using your computer as normal.
To finish the installation, the basic system configuration is set up in Chapter 9, and the kernel and boot loader are set up in Chapter 10. Chapter 11 contains information on continuing the LFS experience beyond this book. After the steps in this book have been implemented, the computer will be ready to reboot into the new LFS system.
This is the process in a nutshell. Detailed information on each step is discussed in the following chapters and package descriptions. Items that may seem complicated will be clarified, and everything will fall into place as you embark on the LFS adventure.
In this version of LFS, there has been a major reorganization of the book using techniques that avoid changing the host system and provides a more straight forward build process.
Below is a list of package updates made since the previous release of the book.
Upgraded to:
Automake-1.16.5
Bash 5.1.16
Bc 5.2.2
Binutils-2.38
Bison-3.8.2
Coreutils-9.0
E2fsprogs-1.46.5
Eudev-3.2.11
Expat-2.4.6
File-5.41
Findutils-4.9.0
Gawk-5.1.1
GDBM-1.23
Glibc-2.35
Gzip-1.11
IANA-Etc-20220207
Inetutils-2.2
IPRoute2-5.16.0
Libcap-2.63
Libelf-0.186 (from elfutils)
Libpipeline-1.5.5
Linux-5.16.9
Man-DB-2.10.1
Meson-0.61.1
Ncurses-6.3
Openssl-3.0.1
Python-3.10.2
Readline-8.1.2
Shadow-4.11.1
SysVinit-3.01
Tcl-8.6.12
Tzdata-2021e
Util-Linux-2.37.4
Vim-8.2.4383
Zstd-1.5.2
Added:
binutils-2.38-lto_fix-1.patch
coreutils-9.0-chmod_fix-1.patch
file-5.40-upstream_fixes-1.patch
shadow-4.10-useradd_segfault-1.patch
sysvinit-3.01-consolidated-1.patch
Removed:
This is version 11.1 of the Linux From Scratch book, dated March 1st, 2022. If this book is more than six months old, a newer and better version is probably already available. To find out, please check one of the mirrors via https://www.linuxfromscratch.org/mirrors.html.
Below is a list of changes made since the previous release of the book.
Changelog Entries:
2022-03-01
[bdubbs] - LFS-11.1 released.
2022-02-23
[bdubbs] - Update to expat-2.4.6 (security fix). Fixes #5011.
2022-02-15
[bdubbs] - LFS-11.1-rc1 released.
[bdubbs] - Add binutils-2.38 LTO patch. Fixes #5011.
[bdubbs] - Update to util-linux-2.37.4. Fixes #5010.
[bdubbs] - Update to man-db-2.10.1. Fixes #5009.
[bdubbs] - Update to linux-5.16.9. Fixes #5008.
[bdubbs] - Update to vim-8.2.4383 (Security Update). Addresses #4500.
[bdubbs] - Update to iana-etc-20220207. Addresses #5006.
2022-02-10
[xry111] - Workaround the issue causing binaries link to libraries from the host distro for pass 2 binutils. It's now unnecessary to build zlib in chapter 6.
2022-02-09
2022-02-04
[xry111] - Remove bash +h
directives in chroot. Fixes
#4998.
[xry111] - Update to man-db-2.10.0. Fixes #5002.
[xry111] - Move OpenSSL before Kmod and enable OpenSSL for Kmod build.
[xry111] - Update to gdbm-1.23. Fixes #5000.
[xry111] - Update to tcl-8.6.12. Fixes #5001.
[thomas] - Remove sed from glibc instructions in chapter 8. It has been applied upstream.
2022-02-03
2022-01-29
[bdubbs] - Update to linux-5.16.4. Fixes #4991.
2022-01-27
[bdubbs] - Update to vim-8.2.4236. Addresses #4500.
[bdubbs] - Update to zstd-1.5.2. Fixes #4988.
[bdubbs] - Update to util-linux-2.37.3 (security fix). Fixes #4989.
[bdubbs] - Update to Python-3.10.2. Fixes #4987.
[bdubbs] - Update to linux-5.16.2. Fixes #4979.
[bdubbs] - Update to libcap-2.63. Fixes #4990.
[bdubbs] - Update to iproute2-5.16.0. Fixes #4982.
[bdubbs] - Update to iana-etc-20220120. Fixes #4975.
2022-01-20
2022-01-17
[thomas] - Added a fix of a typo to the meson-0.61.0 patch.
2022-01-15
2022-01-01
[bdubbs] - Update to e2fsprogs-1.46.5. Fixes #4974.
[bdubbs] - Update to zstd-1.5.1. Fixes #4972.
[bdubbs] - Update to expat-2.4.2. Fixes #4970.
[bdubbs] - Update to shadow-4.10. Fixes #4969.
[bdubbs] - Update to sysvinit-3.01. Fixes #4968.
[bdubbs] - Update to linux-5.15.12. Fixes #4967.
[bdubbs] - Update to iana-etc-20211224. Fixes #4962.
[bdubbs] - Update to openssl-3.0.1. Fixes #4922.
[bdubbs] - Update to eudev-3.2.11. Fixes #4914.
2021-12-30
[renodr] - Update to meson-0.60.3. Fixes #4973.
2021-12-15
2021-12-14
[thomas] - Allow building findutils on 32 bit systems. Cherry-picked from multilib branch by [pierre].
2021-12-01
2021-11-15
[bdubbs] - Update to iana-etc-20211112. Fixes #4955.
[bdubbs] - Update to elfutils-0.186. Fixes #4954.
[bdubbs] - Update to bc-5.2.0. Fixes #4952.
[bdubbs] - Update to ncurses-6.3. Fixes #4951.
[bdubbs] - Update to libpipeline-1.5.4. Fixes #4950.
[bdubbs] - Update to meson-0.60.1. Fixes #4949.
[bdubbs] - Update to iproute2-5.15.0. Fixes #4948.
[bdubbs] - Update to linux-5.15.2. Fixes #4947.
2021-11-01
[bdubbs] - Update to gawk-5.1.1. Fixes #4946.
[bdubbs] - Update to meson-0.60.0. Fixes #4945.
[bdubbs] - Update to libcap-2.60. Fixes #4944.
[bdubbs] - Update to gdbm-1.22. Fixes #4943.
[bdubbs] - Update to file-5.41. Fixes #4942.
[bdubbs] - Update to linux-5.14.15. Fixes #4941.
[bdubbs] - Update to iana-etc-20211025. Fixes #4940.
[bdubbs] - Update to tzdata-2021e. Fixes #4939.
2021-10-15
[bdubbs] - Update to vim-8.2.3508. Addresses #4500.
[bdubbs] - Update to tzdata-2021c. Fixes #4934.
[bdubbs] - Update to Python-3.10.0. Fixes #4938.
[bdubbs] - Update to Jinja2-3.0.2. Fixes #4937.
[bdubbs] - Update to linux-5.14.12. Fixes #4932.
[bdubbs] - Update to iana-etc-20211004. Fixes #4933.
[bdubbs] - Update to bc-5.1.1. Fixes #4936.
[bdubbs] - Update to automake-1.16.5. Fixes #4935.
2021-10-01
[bdubbs] - Update to vim-8.2.3458. Addresses #4500.
[bdubbs] - Update to iana-etc-20210924. Addresses #4722.
[bdubbs] - Update to tzdata-2021b. Fixes #4929.
[bdubbs] - Update to sysvinit-3.0.0. Fixes #4927.
[bdubbs] - Update to meson-0.59.2. Fixes #4931.
[bdubbs] - Update to linux-5.14.8. Fixes #4925.
[bdubbs] - Update to libcap-2.59. Fixes #4926.
[bdubbs] - Update to coreutils-9.0. Fixes #4928.
[bdubbs] - Update to bison-3.8.2. Fixes #4930.
2021-09-15
[bdubbs] - Ensure tcl documentation instructions are present. Fixes #4923.
[bdubbs] - Update to Python3-3.9.7. Fixes #4916.
[bdubbs] - Update to linux-5.14.3. Fixes #4913.
[bdubbs] - Update to libcap-2.57. Fixes #4912.
[bdubbs] - Update to iproute2-5.14.0. Fixes #4917.
[bdubbs] - Update to inetutils-2.2. Fixes #4918.
[bdubbs] - Update to gzip-1.11. Fixes #4920.
[bdubbs] - Update to gdbm-1.21. Fixes #4919.
[bdubbs] - Update to bison-3.8.1. Fixes #4921.
[bdubbs] - Update to bc-5.0.2. Fixes #4905.
2021-09-08
[renodr] - Fix regressions in File that result in improper detection of text and XZ files.
2021-09-06
[bdubbs] - Text clarifications in the backup/restore section of Chapter 7. Thanks to Kevin Buckley for the patch.
2021-09-01
[bdubbs] - LFS-11.0 released.
If during the building of the LFS system you encounter any errors, have any questions, or think there is a typo in the book, please start by consulting the Frequently Asked Questions (FAQ) that is located at https://www.linuxfromscratch.org/faq/.
The linuxfromscratch.org
server
hosts a number of mailing lists used for the development of
the LFS project. These lists include the main development and
support lists, among others. If the FAQ does not solve the
problem you are having, the next step would be to search the
mailing lists at https://www.linuxfromscratch.org/search.html.
For information on the different lists, how to subscribe, archive locations, and additional information, visit https://www.linuxfromscratch.org/mail.html.
Several members of the LFS community offer assistance on
Internet Relay Chat (IRC). Before using this support, please
make sure that your question is not already answered in the
LFS FAQ or the mailing list archives. You can find the IRC
network at irc.libera.chat
. The
support channel is named #lfs-support.
The LFS project has a number of world-wide mirrors to make accessing the website and downloading the required packages more convenient. Please visit the LFS website at https://www.linuxfromscratch.org/mirrors.html for a list of current mirrors.
If an issue or a question is encountered while working through this book, please check the FAQ page at https://www.linuxfromscratch.org/faq/#generalfaq. Questions are often already answered there. If your question is not answered on this page, try to find the source of the problem. The following hint will give you some guidance for troubleshooting: https://www.linuxfromscratch.org/hints/downloads/files/errors.txt.
If you cannot find your problem listed in the FAQ, search the mailing lists at https://www.linuxfromscratch.org/search.html.
We also have a wonderful LFS community that is willing to offer assistance through the mailing lists and IRC (see the Section 1.4, “Resources” section of this book). However, we get several support questions every day and many of them can be easily answered by going to the FAQ and by searching the mailing lists first. So, for us to offer the best assistance possible, you need to do some research on your own first. That allows us to focus on the more unusual support needs. If your searches do not produce a solution, please include all relevant information (mentioned below) in your request for help.
Apart from a brief explanation of the problem being experienced, the essential things to include in any request for help are:
The version of the book being used (in this case 11.1)
The host distribution and version being used to create LFS
The output from the Host System Requirements script
The package or section the problem was encountered in
The exact error message or symptom being received
Note whether you have deviated from the book at all
Deviating from this book does not mean that we will not help you. After all, LFS is about personal preference. Being upfront about any changes to the established procedure helps us evaluate and determine possible causes of your problem.
If something goes wrong while running the configure script, review
the config.log
file. This file
may contain errors encountered during configure which were not
printed to the screen. Include the relevant lines if you need to ask
for help.
Both the screen output and the contents of various files are useful in determining the cause of compilation problems. The screen output from the configure script and the make run can be helpful. It is not necessary to include the entire output, but do include enough of the relevant information. Below is an example of the type of information to include from the screen output from make:
gcc -DALIASPATH=\"/mnt/lfs/usr/share/locale:.\"
-DLOCALEDIR=\"/mnt/lfs/usr/share/locale\"
-DLIBDIR=\"/mnt/lfs/usr/lib\"
-DINCLUDEDIR=\"/mnt/lfs/usr/include\" -DHAVE_CONFIG_H -I. -I.
-g -O2 -c getopt1.c
gcc -g -O2 -static -o make ar.o arscan.o commands.o dir.o
expand.o file.o function.o getopt.o implicit.o job.o main.o
misc.o read.o remake.o rule.o signame.o variable.o vpath.o
default.o remote-stub.o version.o opt1.o
-lutil job.o: In function `load_too_high':
/lfs/tmp/make-3.79.1/job.c:1565: undefined reference
to `getloadavg'
collect2: ld returned 1 exit status
make[2]: *** [make] Error 1
make[2]: Leaving directory `/lfs/tmp/make-3.79.1'
make[1]: *** [all-recursive] Error 1
make[1]: Leaving directory `/lfs/tmp/make-3.79.1'
make: *** [all-recursive-am] Error 2
In this case, many people would just include the bottom section:
make [2]: *** [make] Error 1
This is not enough information to properly diagnose the problem because it only notes that something went wrong, not what went wrong. The entire section, as in the example above, is what should be saved because it includes the command that was executed and the associated error message(s).
An excellent article about asking for help on the Internet is available online at http://catb.org/~esr/faqs/smart-questions.html. Read and follow the hints in this document to increase the likelihood of getting the help you need.
In this chapter, the host tools needed for building LFS are checked and, if necessary, installed. Then a partition which will host the LFS system is prepared. We will create the partition itself, create a file system on it, and mount it.
Your host system should have the following software with the minimum versions indicated. This should not be an issue for most modern Linux distributions. Also note that many distributions will place software headers into separate packages, often in the form of “<package-name>-devel” or “<package-name>-dev”. Be sure to install those if your distribution provides them.
Earlier versions of the listed software packages may work, but have not been tested.
Bash-3.2 (/bin/sh should be a symbolic or hard link to bash)
Binutils-2.13.1 (Versions greater than 2.38 are not recommended as they have not been tested)
Bison-2.7 (/usr/bin/yacc should be a link to bison or small script that executes bison)
Coreutils-6.9
Diffutils-2.8.1
Findutils-4.2.31
Gawk-4.0.1 (/usr/bin/awk should be a link to gawk)
GCC-4.8 including the C++ compiler, g++ (Versions greater than 11.2.0 are not recommended as they have not been tested). C and C++ standard libraries (with headers) must also be present so the C++ compiler can build hosted programs
Grep-2.5.1a
Gzip-1.3.12
Linux Kernel-3.2
The reason for the kernel version requirement is that we specify that version when building glibc in Chapter 5 and Chapter 8, at the recommendation of the developers. It is also required by udev.
If the host kernel is earlier than 3.2 you will need to replace the kernel with a more up to date version. There are two ways you can go about this. First, see if your Linux vendor provides a 3.2 or later kernel package. If so, you may wish to install it. If your vendor doesn't offer an acceptable kernel package, or you would prefer not to install it, you can compile a kernel yourself. Instructions for compiling the kernel and configuring the boot loader (assuming the host uses GRUB) are located in Chapter 10.
M4-1.4.10
Make-4.0
Patch-2.5.4
Perl-5.8.8
Python-3.4
Sed-4.1.5
Tar-1.22
Texinfo-4.7
Xz-5.0.0
Note that the symlinks mentioned above are required to build an LFS system using the instructions contained within this book. Symlinks that point to other software (such as dash, mawk, etc.) may work, but are not tested or supported by the LFS development team, and may require either deviation from the instructions or additional patches to some packages.
To see whether your host system has all the appropriate versions, and the ability to compile programs, run the following:
cat > version-check.sh << "EOF"
#!/bin/bash
# Simple script to list version numbers of critical development tools
export LC_ALL=C
bash --version | head -n1 | cut -d" " -f2-4
MYSH=$(readlink -f /bin/sh)
echo "/bin/sh -> $MYSH"
echo $MYSH | grep -q bash || echo "ERROR: /bin/sh does not point to bash"
unset MYSH
echo -n "Binutils: "; ld --version | head -n1 | cut -d" " -f3-
bison --version | head -n1
if [ -h /usr/bin/yacc ]; then
echo "/usr/bin/yacc -> `readlink -f /usr/bin/yacc`";
elif [ -x /usr/bin/yacc ]; then
echo yacc is `/usr/bin/yacc --version | head -n1`
else
echo "yacc not found"
fi
echo -n "Coreutils: "; chown --version | head -n1 | cut -d")" -f2
diff --version | head -n1
find --version | head -n1
gawk --version | head -n1
if [ -h /usr/bin/awk ]; then
echo "/usr/bin/awk -> `readlink -f /usr/bin/awk`";
elif [ -x /usr/bin/awk ]; then
echo awk is `/usr/bin/awk --version | head -n1`
else
echo "awk not found"
fi
gcc --version | head -n1
g++ --version | head -n1
grep --version | head -n1
gzip --version | head -n1
cat /proc/version
m4 --version | head -n1
make --version | head -n1
patch --version | head -n1
echo Perl `perl -V:version`
python3 --version
sed --version | head -n1
tar --version | head -n1
makeinfo --version | head -n1 # texinfo version
xz --version | head -n1
echo 'int main(){}' > dummy.c && g++ -o dummy dummy.c
if [ -x dummy ]
then echo "g++ compilation OK";
else echo "g++ compilation failed"; fi
rm -f dummy.c dummy
EOF
bash version-check.sh
LFS is designed to be built in one session. That is, the instructions assume that the system will not be shut down during the process. That does not mean that the system has to be done in one sitting. The issue is that certain procedures have to be re-accomplished after a reboot if resuming LFS at different points.
These chapters are accomplished on the host system. When restarting, be careful of the following:
Procedures done as the root
user after Section 2.4 need to
have the LFS environment variable set FOR THE ROOT USER.
The /mnt/lfs partition must be mounted.
These two chapters must be done as user
lfs
. A su - lfs needs to be
done before any task in these chapters. Failing to do
that, you are at risk of installing packages to the
host, and potentially rendering it unusable.
The procedures in General Compilation Instructions are critical. If there is any doubt about installing a package, ensure any previously expanded tarballs are removed, then re-extract the package files, and complete all instructions in that section.
The /mnt/lfs partition must be mounted.
A few operations, from “Changing
Ownership” to “Entering the Chroot
Environment” must be done as the
root
user, with the LFS
environment variable set for the root
user.
When entering chroot, the LFS environment variable must
be set for root
. The
LFS variable is not used afterwards.
The virtual file systems must be mounted. This can be
done before or after entering chroot by changing to a
host virtual terminal and, as root
, running the commands in
Section 7.3.2,
“Mounting and Populating /dev” and
Section 7.3.3,
“Mounting Virtual Kernel File Systems”.
Like most other operating systems, LFS is usually installed on a dedicated partition. The recommended approach to building an LFS system is to use an available empty partition or, if you have enough unpartitioned space, to create one.
A minimal system requires a partition of around 10 gigabytes (GB). This is enough to store all the source tarballs and compile the packages. However, if the LFS system is intended to be the primary Linux system, additional software will probably be installed which will require additional space. A 30 GB partition is a reasonable size to provide for growth. The LFS system itself will not take up this much room. A large portion of this requirement is to provide sufficient free temporary storage as well as for adding additional capabilities after LFS is complete. Additionally, compiling packages can require a lot of disk space which will be reclaimed after the package is installed.
Because there is not always enough Random Access Memory (RAM)
available for compilation processes, it is a good idea to use a
small disk partition as swap
space. This is used by the kernel to store seldom-used data and
leave more memory available for active processes. The
swap
partition for an LFS
system can be the same as the one used by the host system, in
which case it is not necessary to create another one.
Start a disk partitioning program such as cfdisk or fdisk with a command line
option naming the hard disk on which the new partition will be
created—for example /dev/sda
for the primary disk drive. Create a
Linux native partition and a swap
partition, if needed. Please refer to
cfdisk(8)
or fdisk(8)
if you do not yet know how to use
the programs.
For experienced users, other partitioning schemes are possible. The new LFS system can be on a software RAID array or an LVM logical volume. However, some of these options require an initramfs, which is an advanced topic. These partitioning methodologies are not recommended for first time LFS users.
Remember the designation of the new partition (e.g.,
sda5
). This book will refer to
this as the LFS partition. Also remember the designation of the
swap
partition. These names
will be needed later for the /etc/fstab
file.
Requests for advice on system partitioning are often posted on the LFS mailing lists. This is a highly subjective topic. The default for most distributions is to use the entire drive with the exception of one small swap partition. This is not optimal for LFS for several reasons. It reduces flexibility, makes sharing of data across multiple distributions or LFS builds more difficult, makes backups more time consuming, and can waste disk space through inefficient allocation of file system structures.
A root LFS partition (not to be confused with the
/root
directory) of twenty
gigabytes is a good compromise for most systems. It
provides enough space to build LFS and most of BLFS, but is
small enough so that multiple partitions can be easily
created for experimentation.
Most distributions automatically create a swap partition. Generally the recommended size of the swap partition is about twice the amount of physical RAM, however this is rarely needed. If disk space is limited, hold the swap partition to two gigabytes and monitor the amount of disk swapping.
If you want to use the hibernation feature (suspend-to-disk) of Linux, it writes out the contents of RAM to the swap partition before turning off the machine. In this case the size of the swap partition should be at least as large as the system's installed RAM.
Swapping is never good. For mechanical hard drives you can generally tell if a system is swapping by just listening to disk activity and observing how the system reacts to commands. For an SSD drive you will not be able to hear swapping but you can tell how much swap space is being used by the top or free programs. Use of an SSD drive for a swap partition should be avoided if possible. The first reaction to swapping should be to check for an unreasonable command such as trying to edit a five gigabyte file. If swapping becomes a normal occurrence, the best solution is to purchase more RAM for your system.
If the boot disk has been partitioned with a GUID Partition Table (GPT), then a small, typically 1 MB, partition must be created if it does not already exist. This partition is not formatted, but must be available for GRUB to use during installation of the boot loader. This partition will normally be labeled 'BIOS Boot' if using fdisk or have a code of EF02 if using gdisk.
The Grub Bios partition must be on the drive that the BIOS uses to boot the system. This is not necessarily the same drive where the LFS root partition is located. Disks on a system may use different partition table types. The requirement for this partition depends only on the partition table type of the boot disk.
There are several other partitions that are not required, but should be considered when designing a disk layout. The following list is not comprehensive, but is meant as a guide.
/boot – Highly recommended. Use this partition to store kernels and other booting information. To minimize potential boot problems with larger disks, make this the first physical partition on your first disk drive. A partition size of 200 megabytes is quite adequate.
/home – Highly recommended. Share your home directory and user customization across multiple distributions or LFS builds. The size is generally fairly large and depends on available disk space.
/usr – In LFS, /bin
, /lib
, and /sbin
are symlinks to their
counterpart in /usr
. So
/usr
contains all
binaries needed for the system to run. For LFS a
separate partition for /usr
is normally not needed. If you
need it anyway, you should make a partition large
enough to fit all programs and libraries in the
system. The root partition can be very small (maybe
just one gigabyte) in this configuration, so it's
suitable for a thin client or diskless workstation
(where /usr
is mounted
from a remote server). However you should take care
that an initramfs (not covered by LFS) will be needed
to boot a system with separate /usr
partition.
/opt – This directory is most useful for BLFS where multiple installations of large packages like Gnome or KDE can be installed without embedding the files in the /usr hierarchy. If used, 5 to 10 gigabytes is generally adequate.
/tmp – A separate /tmp directory is rare, but useful if configuring a thin client. This partition, if used, will usually not need to exceed a couple of gigabytes.
/usr/src – This partition is very useful for providing a location to store BLFS source files and share them across LFS builds. It can also be used as a location for building BLFS packages. A reasonably large partition of 30-50 gigabytes allows plenty of room.
Any separate partition that you want automatically mounted
upon boot needs to be specified in the /etc/fstab
. Details about how to specify
partitions will be discussed in Section 10.2,
“Creating the /etc/fstab File”.
Now that a blank partition has been set up, the file system can be created. LFS can use any file system recognized by the Linux kernel, but the most common types are ext3 and ext4. The choice of file system can be complex and depends on the characteristics of the files and the size of the partition. For example:
is suitable for small partitions that are updated infrequently such as /boot.
is an upgrade to ext2 that includes a journal to help recover the partition's status in the case of an unclean shutdown. It is commonly used as a general purpose file system.
is the latest version of the ext file system family of partition types. It provides several new capabilities including nano-second timestamps, creation and use of very large files (16 TB), and speed improvements.
Other file systems, including FAT32, NTFS, ReiserFS, JFS, and XFS are useful for specialized purposes. More information about these file systems can be found at http://en.wikipedia.org/wiki/Comparison_of_file_systems.
LFS assumes that the root file system (/) is of type ext4. To
create an ext4
file system on
the LFS partition, run the following:
mkfs -v -t ext4 /dev/<xxx>
Replace <xxx>
with the name of the LFS partition.
If you are using an existing swap
partition, there is no need to format
it. If a new swap
partition was
created, it will need to be initialized with this command:
mkswap /dev/<yyy>
Replace <yyy>
with the name of the swap
partition.
Throughout this book, the environment variable LFS
will be used several times. You should
ensure that this variable is always defined throughout the LFS
build process. It should be set to the name of the directory
where you will be building your LFS system - we will use
/mnt/lfs
as an example, but the
directory choice is up to you. If you are building LFS on a
separate partition, this directory will be the mount point for
the partition. Choose a directory location and set the variable
with the following command:
export LFS=/mnt/lfs
Having this variable set is beneficial in that commands such as mkdir -v $LFS/tools can be typed literally. The shell will automatically replace “$LFS” with “/mnt/lfs” (or whatever the variable was set to) when it processes the command line.
Do not forget to check that LFS
is
set whenever you leave and reenter the current working
environment (such as when doing a su to root
or another user). Check that the
LFS
variable is set up properly
with:
echo $LFS
Make sure the output shows the path to your LFS system's
build location, which is /mnt/lfs
if the provided example was
followed. If the output is incorrect, use the command given
earlier on this page to set $LFS
to the correct directory name.
One way to ensure that the LFS
variable is always set is to edit the .bash_profile
file in both your personal
home directory and in /root/.bash_profile
and enter the export
command above. In addition, the shell specified in the
/etc/passwd
file for all users
that need the LFS
variable needs
to be bash to ensure that the /root/.bash_profile
file is incorporated as
a part of the login process.
Another consideration is the method that is used to log into
the host system. If logging in through a graphical display
manager, the user's .bash_profile
is not normally used when a
virtual terminal is started. In this case, add the export
command to the .bashrc
file for
the user and root
. In
addition, some distributions have instructions to not run the
.bashrc
instructions in a
non-interactive bash invocation. Be sure to add the export
command before the test for non-interactive use.
Now that a file system has been created, the partition needs to
be made accessible. In order to do this, the partition needs to
be mounted at a chosen mount point. For the purposes of this
book, it is assumed that the file system is mounted under the
directory specified by the LFS
environment variable as described in the previous section.
Create the mount point and mount the LFS file system by running:
mkdir -pv $LFS
mount -v -t ext4 /dev/<xxx>
$LFS
Replace <xxx>
with the designation of the LFS partition.
If using multiple partitions for LFS (e.g., one for
/
and another for /home
), mount them using:
mkdir -pv $LFS mount -v -t ext4 /dev/<xxx>
$LFS mkdir -v $LFS/home mount -v -t ext4 /dev/<yyy>
$LFS/home
Replace <xxx>
and <yyy>
with
the appropriate partition names.
Ensure that this new partition is not mounted with permissions
that are too restrictive (such as the nosuid
or nodev
options). Run the mount command without any
parameters to see what options are set for the mounted LFS
partition. If nosuid
and/or
nodev
are set, the partition will
need to be remounted.
The above instructions assume that you will not be restarting your computer throughout the LFS process. If you shut down your system, you will either need to remount the LFS partition each time you restart the build process or modify your host system's /etc/fstab file to automatically remount it upon boot. For example:
/dev/<xxx>
/mnt/lfs ext4 defaults 1 1
If you use additional optional partitions, be sure to add them also.
If you are using a swap
partition, ensure that it is enabled using the swapon command:
/sbin/swapon -v /dev/<zzz>
Replace <zzz>
with the name of the swap
partition.
Now that there is an established place to work, it is time to download the packages.
This chapter includes a list of packages that need to be downloaded in order to build a basic Linux system. The listed version numbers correspond to versions of the software that are known to work, and this book is based on their use. We highly recommend against using different versions because the build commands for one version may not work with a different version, unless the different version is specified by a LFS errata or security advisory. The newest package versions may also have problems that require work-arounds. These work-arounds will be developed and stabilized in the development version of the book.
For some packages, the release tarball and the (Git or SVN) repository snapshot tarball for this release may be published with similar file name. A release tarball contains generated files (for example, configure script generated by autoconf), in addition to the contents of the corresponding repository snapshot. The book uses release tarballs whenever possible. Using a repository snapshot instead of a release tarball specified by the book will cause problems.
Download locations may not always be accessible. If a download location has changed since this book was published, Google (http://www.google.com/) provides a useful search engine for most packages. If this search is unsuccessful, try one of the alternative means of downloading at https://www.linuxfromscratch.org/lfs/mirrors.html#files.
Downloaded packages and patches will need to be stored
somewhere that is conveniently available throughout the entire
build. A working directory is also required to unpack the
sources and build them. $LFS/sources
can be used both as the place to
store the tarballs and patches and as a working directory. By
using this directory, the required elements will be located on
the LFS partition and will be available during all stages of
the building process.
To create this directory, execute the following command, as
user root
, before starting the
download session:
mkdir -v $LFS/sources
Make this directory writable and sticky. “Sticky” means that even if multiple users have write permission on a directory, only the owner of a file can delete the file within a sticky directory. The following command will enable the write and sticky modes:
chmod -v a+wt $LFS/sources
There are several ways to obtain all the necessary packages and patches to build LFS:
The files can be downloaded individually as described in the next two sections.
For stable versions of the book, a tarball of all the needed files can be downloaded from one of the LFS files mirrors listed at https://www.linuxfromscratch.org/mirrors.html#files.
The files can be downloaded using wget and a wget-list as described below.
To download all of the packages and patches by using wget-list as an input to the wget command, use:
wget --input-file=wget-list --continue --directory-prefix=$LFS/sources
The wget-list
file mentioned
above retrieves all packages for both the sysV and systemd
versions of LFS. There are a total of five additional small
packages not needed for the currrent book. The md5sums
file mentioned below is specific to
the current book.
Additionally, starting with LFS-7.0, there is a separate file,
md5sums, which can be
used to verify that all the correct packages are available
before proceeding. Place that file in $LFS/sources
and run:
pushd $LFS/sources md5sum -c md5sums popd
This check can be used after retrieving the needed files with any of the methods listed above.
Download or otherwise obtain the following packages:
Home page: https://savannah.nongnu.org/projects/acl
Download: https://download.savannah.gnu.org/releases/acl/acl-2.3.1.tar.xz
MD5 sum: 95ce715fe09acca7c12d3306d0f076b2
Home page: https://savannah.nongnu.org/projects/attr
Download: https://download.savannah.gnu.org/releases/attr/attr-2.5.1.tar.gz
MD5 sum: ac1c5a7a084f0f83b8cace34211f64d8
Home page: https://www.gnu.org/software/autoconf/
Download: https://ftp.gnu.org/gnu/autoconf/autoconf-2.71.tar.xz
MD5 sum: 12cfa1687ffa2606337efe1a64416106
Home page: https://www.gnu.org/software/automake/
Download: https://ftp.gnu.org/gnu/automake/automake-1.16.5.tar.xz
MD5 sum: 4017e96f89fca45ca946f1c5db6be714
SHA256 sum: 80facc09885a57e6d49d06972c0ae1089c5fa8f4d4c7cfe5baea58e5085f136d
Home page: https://www.gnu.org/software/bash/
Download: https://ftp.gnu.org/gnu/bash/bash-5.1.16.tar.gz
MD5 sum: c17b20a09fc38d67fb303aeb6c130b4e
Home page: https://git.yzena.com/gavin/bc
Download: https://github.com/gavinhoward/bc/releases/download/5.2.2/bc-5.2.2.tar.xz
MD5 sum: 632344cdb052af0e06087bd3b0034126
Home page: https://www.gnu.org/software/binutils/
Download: https://ftp.gnu.org/gnu/binutils/binutils-2.38.tar.xz
MD5 sum: 6e39cad1bb414add02b5b1169c18fdc5
Home page: https://www.gnu.org/software/bison/
Download: https://ftp.gnu.org/gnu/bison/bison-3.8.2.tar.xz
MD5 sum: c28f119f405a2304ff0a7ccdcc629713
Download: https://www.sourceware.org/pub/bzip2/bzip2-1.0.8.tar.gz
MD5 sum: 67e051268d0c475ea773822f7500d0e5
Home page: https://libcheck.github.io/check
Download: https://github.com/libcheck/check/releases/download/0.15.2/check-0.15.2.tar.gz
MD5 sum: 50fcafcecde5a380415b12e9c574e0b2
Home page: https://www.gnu.org/software/coreutils/
Download: https://ftp.gnu.org/gnu/coreutils/coreutils-9.0.tar.xz
MD5 sum: 0d79ae8a6124546e3b94171375e5e5d0
Home page: https://www.gnu.org/software/dejagnu/
Download: https://ftp.gnu.org/gnu/dejagnu/dejagnu-1.6.3.tar.gz
MD5 sum: 68c5208c58236eba447d7d6d1326b821
Home page: https://www.gnu.org/software/diffutils/
Download: https://ftp.gnu.org/gnu/diffutils/diffutils-3.8.tar.xz
MD5 sum: 6a6b0fdc72acfe3f2829aab477876fbc
Home page: http://e2fsprogs.sourceforge.net/
Download: https://downloads.sourceforge.net/project/e2fsprogs/e2fsprogs/v1.46.5/e2fsprogs-1.46.5.tar.gz
MD5 sum: 3da91854c960ad8a819b48b2a404eb43
Home page: https://sourceware.org/elfutils/
Download: https://sourceware.org/ftp/elfutils/0.186/elfutils-0.186.tar.bz2
MD5 sum: 2c095e31e35d6be7b3718477b6d52702
Download: https://github.com/eudev-project/eudev/releases/download/v3.2.11/eudev-3.2.11.tar.gz
MD5 sum: 417ba948335736d4d81874fba47a30f7
Home page: https://libexpat.github.io/
Download: https://prdownloads.sourceforge.net/expat/expat-2.4.6.tar.xz
MD5 sum: 22a30c888752fdda9f8dd1b7281c54b0
The upstream may remove tarballs of the specific releases of Expat when these releases contain a security vulnerability. You should refer to LFS security advisories to figure out which version (with the vulnerability fixed) should be used. You may download the vulnerable version from a mirror, but it's not recommended.
Home page: https://core.tcl.tk/expect/
Download: https://prdownloads.sourceforge.net/expect/expect5.45.4.tar.gz
MD5 sum: 00fce8de158422f5ccd2666512329bd2
Home page: https://www.darwinsys.com/file/
Download: https://astron.com/pub/file/file-5.41.tar.gz
MD5 sum: 18233bb0a0089dfdc7dfbc93b96f231b
Home page: https://www.gnu.org/software/findutils/
Download: https://ftp.gnu.org/gnu/findutils/findutils-4.9.0.tar.xz
MD5 sum: 4a4a547e888a944b2f3af31d789a1137
Home page: https://github.com/westes/flex
Download: https://github.com/westes/flex/releases/download/v2.6.4/flex-2.6.4.tar.gz
MD5 sum: 2882e3179748cc9f9c23ec593d6adc8d
Home page: https://www.gnu.org/software/gawk/
Download: https://ftp.gnu.org/gnu/gawk/gawk-5.1.1.tar.xz
MD5 sum: 83650aa943ff2fd519b2abedf8506ace
Home page: https://gcc.gnu.org/
Download: https://ftp.gnu.org/gnu/gcc/gcc-11.2.0/gcc-11.2.0.tar.xz
MD5 sum: 31c86f2ced76acac66992eeedce2fce2
SHA256 sum: d08edc536b54c372a1010ff6619dd274c0f1603aa49212ba20f7aa2cda36fa8b
Home page: https://www.gnu.org/software/gdbm/
Download: https://ftp.gnu.org/gnu/gdbm/gdbm-1.23.tar.gz
MD5 sum: 8551961e36bf8c70b7500d255d3658ec
Home page: https://www.gnu.org/software/gettext/
Download: https://ftp.gnu.org/gnu/gettext/gettext-0.21.tar.xz
MD5 sum: 40996bbaf7d1356d3c22e33a8b255b31
Home page: https://www.gnu.org/software/libc/
Download: https://ftp.gnu.org/gnu/glibc/glibc-2.35.tar.xz
MD5 sum: dd571c67d85d89d7f60b854a4e207423
Home page: https://www.gnu.org/software/gmp/
Download: https://ftp.gnu.org/gnu/gmp/gmp-6.2.1.tar.xz
MD5 sum: 0b82665c4a92fd2ade7440c13fcaa42b
Home page: https://www.gnu.org/software/gperf/
Download: https://ftp.gnu.org/gnu/gperf/gperf-3.1.tar.gz
MD5 sum: 9e251c0a618ad0824b51117d5d9db87e
Home page: https://www.gnu.org/software/grep/
Download: https://ftp.gnu.org/gnu/grep/grep-3.7.tar.xz
MD5 sum: 7c9cca97fa18670a21e72638c3e1dabf
Home page: https://www.gnu.org/software/groff/
Download: https://ftp.gnu.org/gnu/groff/groff-1.22.4.tar.gz
MD5 sum: 08fb04335e2f5e73f23ea4c3adbf0c5f
Home page: https://www.gnu.org/software/grub/
Download: https://ftp.gnu.org/gnu/grub/grub-2.06.tar.xz
MD5 sum: cf0fd928b1e5479c8108ee52cb114363
Home page: https://www.gnu.org/software/gzip/
Download: https://ftp.gnu.org/gnu/gzip/gzip-1.11.tar.xz
MD5 sum: d1e93996dba00cab0caa7903cd01d454
Home page: https://www.iana.org/protocols
Download: https://github.com/Mic92/iana-etc/releases/download/20220207/iana-etc-20220207.tar.gz
MD5 sum: 81d865ce7fe4240d5abed48c3ca5fa9f
Home page: https://www.gnu.org/software/inetutils/
Download: https://ftp.gnu.org/gnu/inetutils/inetutils-2.2.tar.xz
MD5 sum: de8c1b49cbde2b30e481c61c65357ad4
SHA256 sum: 01b9a4bc73a47e63f6e8a07b76122d9ad2a2e46ebf14870e9c91d660b5647a22
Home page: https://freedesktop.org/wiki/Software/intltool
Download: https://launchpad.net/intltool/trunk/0.51.0/+download/intltool-0.51.0.tar.gz
MD5 sum: 12e517cac2b57a0121cda351570f1e63
Home page: https://www.kernel.org/pub/linux/utils/net/iproute2/
Download: https://www.kernel.org/pub/linux/utils/net/iproute2/iproute2-5.16.0.tar.xz
MD5 sum: 994c1bad2a24aa9d70e89670c5b5dfcb
Home page: https://kbd-project.org/
Download: https://www.kernel.org/pub/linux/utils/kbd/kbd-2.4.0.tar.xz
MD5 sum: 3cac5be0096fcf7b32dcbd3c53831380
Download: https://www.kernel.org/pub/linux/utils/kernel/kmod/kmod-29.tar.xz
MD5 sum: e81e63acd80697d001c8d85c1acb38a0
Home page: https://www.greenwoodsoftware.com/less/
Download: https://www.greenwoodsoftware.com/less/less-590.tar.gz
MD5 sum: f029087448357812fba450091a1172ab
Download: https://www.linuxfromscratch.org/lfs/downloads/11.1/lfs-bootscripts-20210608.tar.xz
MD5 sum: 0f51a074cc4faaff93b3c80e9ab27b0c
Home page: https://sites.google.com/site/fullycapable/
Download: https://www.kernel.org/pub/linux/libs/security/linux-privs/libcap2/libcap-2.63.tar.xz
MD5 sum: 18410cec436f827e698ee9ea16ada9b7
Home page: https://sourceware.org/libffi/
Download: https://github.com/libffi/libffi/releases/download/v3.4.2/libffi-3.4.2.tar.gz
MD5 sum: 294b921e6cf9ab0fbaea4b639f8fdbe8
Home page: http://libpipeline.nongnu.org/
Download: https://download.savannah.gnu.org/releases/libpipeline/libpipeline-1.5.5.tar.gz
MD5 sum: 3e725c76bfea1985e87e851ee50c2e29
Home page: https://www.gnu.org/software/libtool/
Download: https://ftp.gnu.org/gnu/libtool/libtool-2.4.6.tar.xz
MD5 sum: 1bfb9b923f2c1339b4d2ce1807064aa5
Home page: https://www.kernel.org/
Download: https://www.kernel.org/pub/linux/kernel/v5.x/linux-5.16.9.tar.xz
MD5 sum: 4d6a704bf3e249ef6189b6f17457084b
The Linux kernel is updated relatively often, many times due to discoveries of security vulnerabilities. The latest available stable kernel version may be used, unless the errata page says otherwise.
For users with limited speed or expensive bandwidth who wish to update the Linux kernel, a baseline version of the package and patches can be downloaded separately. This may save some time or cost for a subsequent patch level upgrade within a minor release.
Home page: https://www.gnu.org/software/m4/
Download: https://ftp.gnu.org/gnu/m4/m4-1.4.19.tar.xz
MD5 sum: 0d90823e1426f1da2fd872df0311298d
Home page: https://www.gnu.org/software/make/
Download: https://ftp.gnu.org/gnu/make/make-4.3.tar.gz
MD5 sum: fc7a67ea86ace13195b0bce683fd4469
Home page: https://www.nongnu.org/man-db/
Download: https://download.savannah.gnu.org/releases/man-db/man-db-2.10.1.tar.xz
MD5 sum: b03b76a9a00d0d6b2299b823fba4f579
Home page: https://www.kernel.org/doc/man-pages/
Download: https://www.kernel.org/pub/linux/docs/man-pages/man-pages-5.13.tar.xz
MD5 sum: 3ac24e8c6fae26b801cb87ceb63c0a30
Home page: https://mesonbuild.com
Download: https://github.com/mesonbuild/meson/releases/download/0.61.1/meson-0.61.1.tar.gz
MD5 sum: 8ed66d5537275df3defffb66d1fb897f
Home page: http://www.multiprecision.org/
Download: https://ftp.gnu.org/gnu/mpc/mpc-1.2.1.tar.gz
MD5 sum: 9f16c976c25bb0f76b50be749cd7a3a8
Home page: https://www.mpfr.org/
Download: https://www.mpfr.org/mpfr-4.1.0/mpfr-4.1.0.tar.xz
MD5 sum: bdd3d5efba9c17da8d83a35ec552baef
Home page: https://www.gnu.org/software/ncurses/
Download: https://invisible-mirror.net/archives/ncurses/ncurses-6.3.tar.gz
MD5 sum: a2736befde5fee7d2b7eb45eb281cdbe
Home page: https://ninja-build.org/
Download: https://github.com/ninja-build/ninja/archive/v1.10.2/ninja-1.10.2.tar.gz
MD5 sum: 639f75bc2e3b19ab893eaf2c810d4eb4
Home page: https://www.openssl.org/
Download: https://www.openssl.org/source/openssl-3.0.1.tar.gz
MD5 sum: 7d07e849d77d276891edd579a8832bb3
Home page: https://savannah.gnu.org/projects/patch/
Download: https://ftp.gnu.org/gnu/patch/patch-2.7.6.tar.xz
MD5 sum: 78ad9937e4caadcba1526ef1853730d5
Home page: https://www.perl.org/
Download: https://www.cpan.org/src/5.0/perl-5.34.0.tar.xz
MD5 sum: df7ecb0653440b26dc951ad9dbfab517
Home page: https://www.freedesktop.org/wiki/Software/pkg-config
Download: https://pkg-config.freedesktop.org/releases/pkg-config-0.29.2.tar.gz
MD5 sum: f6e931e319531b736fadc017f470e68a
Home page: https://sourceforge.net/projects/procps-ng
Download: https://sourceforge.net/projects/procps-ng/files/Production/procps-ng-3.3.17.tar.xz
MD5 sum: d60613e88c2f442ebd462b5a75313d56
Home page: https://gitlab.com/psmisc/psmisc
Download: https://sourceforge.net/projects/psmisc/files/psmisc/psmisc-23.4.tar.xz
MD5 sum: 8114cd4489b95308efe2509c3a406bbf
Home page: https://www.python.org/
Download: https://www.python.org/ftp/python/3.10.2/Python-3.10.2.tar.xz
MD5 sum: 14e8c22458ed7779a1957b26cde01db9
Download: https://www.python.org/ftp/python/doc/3.10.2/python-3.10.2-docs-html.tar.bz2
MD5 sum: ffa52f0017baf72df9d32dec785fd6ab
Home page: https://tiswww.case.edu/php/chet/readline/rltop.html
Download: https://ftp.gnu.org/gnu/readline/readline-8.1.2.tar.gz
MD5 sum: 12819fa739a78a6172400f399ab34f81
Home page: https://www.gnu.org/software/sed/
Download: https://ftp.gnu.org/gnu/sed/sed-4.8.tar.xz
MD5 sum: 6d906edfdb3202304059233f51f9a71d
Home page: https://shadow-maint.github.io/shadow/
Download: https://github.com/shadow-maint/shadow/releases/download/v4.11.1/shadow-4.11.1.tar.xz
MD5 sum: 5a95ec069aa91508167d02fecafaa912
Home page: https://www.infodrom.org/projects/sysklogd/
Download: https://www.infodrom.org/projects/sysklogd/download/sysklogd-1.5.1.tar.gz
MD5 sum: c70599ab0d037fde724f7210c2c8d7f8
Home page: https://savannah.nongnu.org/projects/sysvinit
Download: https://download.savannah.gnu.org/releases/sysvinit/sysvinit-3.01.tar.xz
MD5 sum: dc14f92af715bcfa33cc25341730452e
Home page: https://www.gnu.org/software/tar/
Download: https://ftp.gnu.org/gnu/tar/tar-1.34.tar.xz
MD5 sum: 9a08d29a9ac4727130b5708347c0f5cf
Home page: http://tcl.sourceforge.net/
Download: https://downloads.sourceforge.net/tcl/tcl8.6.12-src.tar.gz
MD5 sum: 87ea890821d2221f2ab5157bc5eb885f
Download: https://downloads.sourceforge.net/tcl/tcl8.6.12-html.tar.gz
MD5 sum: a0d1a5b60bbb68f2f0bd3066a19c527a
Home page: https://www.gnu.org/software/texinfo/
Download: https://ftp.gnu.org/gnu/texinfo/texinfo-6.8.tar.xz
MD5 sum: a91b404e30561a5df803e6eb3a53be71
Home page: https://www.iana.org/time-zones
Download: https://www.iana.org/time-zones/repository/releases/tzdata2021e.tar.gz
MD5 sum: 4fdfad906ebc85fef30221c10964cce9
Download: https://anduin.linuxfromscratch.org/LFS/udev-lfs-20171102.tar.xz
MD5 sum: 27cd82f9a61422e186b9d6759ddf1634
Home page: https://git.kernel.org/pub/scm/utils/util-linux/util-linux.git/
Download: https://www.kernel.org/pub/linux/utils/util-linux/v2.37/util-linux-2.37.4.tar.xz
MD5 sum: 755919e658c349cad9e1c7c771742d48
Home page: https://www.vim.org
Download: https://anduin.linuxfromscratch.org/LFS/vim-8.2.4383.tar.gz
MD5 sum: 3168ff48e382a1201bd0cbd0209bd3e0
The version of vim changes daily. To get the latest version, go to https://github.com/vim/vim/tags.
Home page: https://github.com/chorny/XML-Parser
Download: https://cpan.metacpan.org/authors/id/T/TO/TODDR/XML-Parser-2.46.tar.gz
MD5 sum: 80bb18a8e6240fcf7ec2f7b57601c170
Home page: https://tukaani.org/xz
Download: https://tukaani.org/xz/xz-5.2.5.tar.xz
MD5 sum: aa1621ec7013a19abab52a8aff04fe5b
Home page: https://www.zlib.net/
Download: https://zlib.net/zlib-1.2.11.tar.xz
MD5 sum: 85adef240c5f370b308da8c938951a68
Home page: https://facebook.github.io/zstd/
Download: https://github.com/facebook/zstd/releases/download/v1.5.2/zstd-1.5.2.tar.gz
MD5 sum: 072b10f71f5820c24761a65f31f43e73
Total size of these packages: about 446 MB
In addition to the packages, several patches are also required. These patches correct any mistakes in the packages that should be fixed by the maintainer. The patches also make small modifications to make the packages easier to work with. The following patches will be needed to build an LFS system:
Download: https://www.linuxfromscratch.org/patches/lfs/11.1/binutils-2.38-lto_fix-1.patch
MD5 sum: 3df11b6123d5bbdb0fc83862a003827a
Download: https://www.linuxfromscratch.org/patches/lfs/11.1/bzip2-1.0.8-install_docs-1.patch
MD5 sum: 6a5ac7e89b791aae556de0f745916f7f
Download: https://www.linuxfromscratch.org/patches/lfs/11.1/coreutils-9.0-i18n-1.patch
MD5 sum: 1eeba2736dfea013509f9975365e4e32
Download: https://www.linuxfromscratch.org/patches/lfs/11.1/coreutils-9.0-chmod_fix-1.patch
MD5 sum: 4709df88e68279e6ef357aa819ba5b1a
Download: https://www.linuxfromscratch.org/patches/lfs/11.1/glibc-2.35-fhs-1.patch
MD5 sum: 9a5997c3452909b1769918c759eff8a2
Download: https://www.linuxfromscratch.org/patches/lfs/11.1/kbd-2.4.0-backspace-1.patch
MD5 sum: f75cca16a38da6caa7d52151f7136895
Download: https://www.linuxfromscratch.org/patches/lfs/11.1/perl-5.34.0-upstream_fixes-1.patch
MD5 sum: fb42558b59ed95ee00eb9f1c1c9b8056
Download: https://www.linuxfromscratch.org/patches/lfs/11.1/sysvinit-3.01-consolidated-1.patch
MD5 sum: 4900322141d493e74020c9cf437b2cdc
Total size of these patches: about 193.7 KB
In addition to the above required patches, there exist a number of optional patches created by the LFS community. These optional patches solve minor problems or enable functionality that is not enabled by default. Feel free to peruse the patches database located at https://www.linuxfromscratch.org/patches/downloads/ and acquire any additional patches to suit your system needs.
In this chapter, we will perform a few additional tasks to
prepare for building the temporary system. We will create a set
of directories in $LFS
for the
installation of the temporary tools, add an unprivileged user
to reduce risk, and create an appropriate build environment for
that user. We will also explain the unit of time we use to
measure how long LFS packages take to build, or “SBUs”,
and give some information about package test suites.
The first task performed in the LFS partition is to create a limited directory hierarchy so that programs compiled in Chapter 6 (as well as glibc and libstdc++ in Chapter 5) may be installed in their final location. This is needed so that those temporary programs be overwritten when rebuilding them in Chapter 8.
Create the required directory layout by running the following
as root
:
mkdir -pv $LFS/{etc,var} $LFS/usr/{bin,lib,sbin} for i in bin lib sbin; do ln -sv usr/$i $LFS/$i done case $(uname -m) in x86_64) mkdir -pv $LFS/lib64 ;; esac
Programs in Chapter 6 will be compiled with a cross-compiler (more details in section Toolchain Technical Notes). In order to separate this cross-compiler from the other programs, it will be installed in a special directory. Create this directory with:
mkdir -pv $LFS/tools
When logged in as user root
,
making a single mistake can damage or destroy a system.
Therefore, the packages in the next two chapters are built as
an unprivileged user. You could use your own user name, but to
make it easier to set up a clean working environment, create a
new user called lfs
as a member
of a new group (also named lfs
)
and use this user during the installation process. As
root
, issue the following
commands to add the new user:
groupadd lfs useradd -s /bin/bash -g lfs -m -k /dev/null lfs
The meaning of the command line options:
-s
/bin/bash
This makes bash the default shell
for user lfs
.
-g
lfs
This option adds user lfs
to group lfs
.
-m
This creates a home directory for lfs
.
-k
/dev/null
This parameter prevents possible copying of files from a
skeleton directory (default is /etc/skel
) by changing the input
location to the special null device.
lfs
This is the actual name for the created user.
To log in as lfs
(as opposed to
switching to user lfs
when
logged in as root
, which does
not require the lfs
user to
have a password), give lfs
a
password:
passwd lfs
Grant lfs
full access to all
directories under $LFS
by making
lfs
the directory owner:
chown -v lfs $LFS/{usr{,/*},lib,var,etc,bin,sbin,tools} case $(uname -m) in x86_64) chown -v lfs $LFS/lib64 ;; esac
If a separate working directory was created as suggested, give
user lfs
ownership of this
directory:
chown -v lfs $LFS/sources
In some host systems, the following command does not complete properly and suspends the login to the lfs user to the background. If the prompt "lfs:~$" does not appear immediately, entering the fg command will fix the issue.
Next, login as user lfs
. This
can be done via a virtual console, through a display manager,
or with the following substitute/switch user command:
su - lfs
The “-
” instructs
su to start a
login shell as opposed to a non-login shell. The difference
between these two types of shells can be found in detail in
bash(1)
and info bash.
Set up a good working environment by creating two new startup
files for the bash shell. While logged in
as user lfs
, issue the
following command to create a new .bash_profile
:
cat > ~/.bash_profile << "EOF"
exec env -i HOME=$HOME TERM=$TERM PS1='\u:\w\$ ' /bin/bash
EOF
When logged on as user lfs
, the
initial shell is usually a login shell which reads the
/etc/profile
of the host
(probably containing some settings and environment variables)
and then .bash_profile
. The
exec env
-i.../bin/bash command in the .bash_profile
file replaces the running shell
with a new one with a completely empty environment, except for
the HOME
, TERM
, and PS1
variables. This ensures that no unwanted and potentially
hazardous environment variables from the host system leak into
the build environment. The technique used here achieves the
goal of ensuring a clean environment.
The new instance of the shell is a non-login shell, which does not
read, and execute, the contents of /etc/profile
or .bash_profile
files, but rather reads, and
executes, the .bashrc
file
instead. Create the .bashrc
file
now:
cat > ~/.bashrc << "EOF"
set +h
umask 022
LFS=/mnt/lfs
LC_ALL=POSIX
LFS_TGT=$(uname -m)-lfs-linux-gnu
PATH=/usr/bin
if [ ! -L /bin ]; then PATH=/bin:$PATH; fi
PATH=$LFS/tools/bin:$PATH
CONFIG_SITE=$LFS/usr/share/config.site
export LFS LC_ALL LFS_TGT PATH CONFIG_SITE
EOF
The meaning of the settings in .bashrc
set
+h
The set +h
command turns off bash's hash function.
Hashing is ordinarily a useful feature—bash uses a hash table
to remember the full path of executable files to avoid
searching the PATH
time and
again to find the same executable. However, the new tools
should be used as soon as they are installed. By
switching off the hash function, the shell will always
search the PATH
when a program
is to be run. As such, the shell will find the newly
compiled tools in $LFS/tools/bin
as soon as they are
available without remembering a previous version of the
same program provided by the host distro, in /usr/bin
or /bin
.
umask
022
Setting the user file-creation mask (umask) to 022
ensures that newly created files and directories are only
writable by their owner, but are readable and executable
by anyone (assuming default modes are used by the
open(2)
system call, new
files will end up with permission mode 644 and
directories with mode 755).
LFS=/mnt/lfs
The LFS
variable should be set
to the chosen mount point.
LC_ALL=POSIX
The LC_ALL
variable controls
the localization of certain programs, making their
messages follow the conventions of a specified country.
Setting LC_ALL
to “POSIX” or “C”
(the two are equivalent) ensures that everything will
work as expected in the chroot environment.
LFS_TGT=(uname
-m)-lfs-linux-gnu
The LFS_TGT
variable sets a
non-default, but compatible machine description for use
when building our cross compiler and linker and when
cross compiling our temporary toolchain. More information
is contained in Toolchain Technical
Notes.
PATH=/usr/bin
Many modern linux distributions have merged /bin
and /usr/bin
. When this is the case, the
standard PATH
variable needs
just to be set to /usr/bin/
for the Chapter 6
environment. When this is not the case, the following
line adds /bin
to the path.
if [ ! -L
/bin ]; then PATH=/bin:$PATH; fi
If /bin
is not a symbolic
link, then it has to be added to the PATH
variable.
PATH=$LFS/tools/bin:$PATH
By putting $LFS/tools/bin
ahead of the standard PATH
,
the cross-compiler installed at the beginning of
Chapter 5
is picked up by the shell immediately after its
installation. This, combined with turning off hashing,
limits the risk that the compiler from the host be used
instead of the cross-compiler.
CONFIG_SITE=$LFS/usr/share/config.site
In Chapter 5
and
Chapter 6, if this variable is not set,
configure
scripts may attempt to load configuration items specific
to some distributions from /usr/share/config.site
on the host
system. Override it to prevent potential contamination
from the host.
export
...
While the above commands have set some variables, in order to make them visible within any sub-shells, we export them.
Several commercial distributions add a non-documented
instantiation of /etc/bash.bashrc
to the initialization of
bash. This file
has the potential to modify the lfs
user's environment in ways that can
affect the building of critical LFS packages. To make sure
the lfs
user's environment is
clean, check for the presence of /etc/bash.bashrc
and, if present, move it
out of the way. As the root
user, run:
[ ! -e /etc/bash.bashrc ] || mv -v /etc/bash.bashrc /etc/bash.bashrc.NOUSE
After use of the lfs
user is
finished at the beginning of
Chapter 7, you can restore /etc/bash.bashrc
(if desired).
Note that the LFS Bash package we will build in Section 8.34,
“Bash-5.1.16” is not configured to load or
execute /etc/bash.bashrc
, so
this file is useless on a completed LFS system.
Finally, to have the environment fully prepared for building the temporary tools, source the just-created user profile:
source ~/.bash_profile
Many people would like to know beforehand approximately how long it takes to compile and install each package. Because Linux From Scratch can be built on many different systems, it is impossible to provide accurate time estimates. The biggest package (Glibc) will take approximately 20 minutes on the fastest systems, but could take up to three days on slower systems! Instead of providing actual times, the Standard Build Unit (SBU) measure will be used instead.
The SBU measure works as follows. The first package to be compiled from this book is binutils in Chapter 5. The time it takes to compile this package is what will be referred to as the Standard Build Unit or SBU. All other compile times will be expressed relative to this time.
For example, consider a package whose compilation time is 4.5 SBUs. This means that if a system took 10 minutes to compile and install the first pass of binutils, it will take approximately 45 minutes to build this example package. Fortunately, most build times are shorter than the one for binutils.
In general, SBUs are not entirely accurate because they depend on many factors, including the host system's version of GCC. They are provided here to give an estimate of how long it might take to install a package, but the numbers can vary by as much as dozens of minutes in some cases.
For many modern systems with multiple processors (or cores) the compilation time for a package can be reduced by performing a "parallel make" by either setting an environment variable or telling the make program how many processors are available. For instance, an Intel i5-6500 CPU can support four simultaneous processes with:
export MAKEFLAGS='-j4'
or just building with:
make -j4
When multiple processors are used in this way, the SBU units in the book will vary even more than they normally would. In some cases, the make step will simply fail. Analyzing the output of the build process will also be more difficult because the lines of different processes will be interleaved. If you run into a problem with a build step, revert back to a single processor build to properly analyze the error messages.
Most packages provide a test suite. Running the test suite for a newly built package is a good idea because it can provide a “sanity check” indicating that everything compiled correctly. A test suite that passes its set of checks usually proves that the package is functioning as the developer intended. It does not, however, guarantee that the package is totally bug free.
Some test suites are more important than others. For example, the test suites for the core toolchain packages—GCC, binutils, and glibc—are of the utmost importance due to their central role in a properly functioning system. The test suites for GCC and glibc can take a very long time to complete, especially on slower hardware, but are strongly recommended.
Running the test suites in Chapter 5 and Chapter 6 is impossible, since the programs are compiled with a cross-compiler, so are not supposed to be able to run on the build host.
A common issue with running the test suites for binutils and
GCC is running out of pseudo terminals (PTYs). This can result
in a high number of failing tests. This may happen for several
reasons, but the most likely cause is that the host system does
not have the devpts
file system
set up correctly. This issue is discussed in greater detail at
https://www.linuxfromscratch.org/lfs/faq.html#no-ptys.
Sometimes package test suites will fail, but for reasons which the developers are aware of and have deemed non-critical. Consult the logs located at https://www.linuxfromscratch.org/lfs/build-logs/11.1/ to verify whether or not these failures are expected. This site is valid for all tests throughout this book.
This part is divided into three stages: first building a cross compiler and its associated libraries; second, use this cross toolchain to build several utilities in a way that isolates them from the host distribution; third, enter the chroot environment, which further improves host isolation, and build the remaining tools needed to build the final system.
With this part begins the real work of building a new system. It requires much care in ensuring that the instructions are followed exactly as the book shows them. You should try to understand what they do, and whatever your eagerness to finish your build, you should refrain from blindly type them as shown, but rather read documentation when there is something you do not understand. Also, keep track of your typing and of the output of commands, by sending them to a file, using the tee utility. This allows for better diagnosing if something gets wrong.
The next section gives a technical introduction to the build process, while the following one contains very important general instructions.
This section explains some of the rationale and technical details behind the overall build method. It is not essential to immediately understand everything in this section. Most of this information will be clearer after performing an actual build. This section can be referred to at any time during the process.
The overall goal of Chapter 5 and Chapter 6 is to produce a temporary area that contains a known-good set of tools that can be isolated from the host system. By using chroot, the commands in the remaining chapters will be contained within that environment, ensuring a clean, trouble-free build of the target LFS system. The build process has been designed to minimize the risks for new readers and to provide the most educational value at the same time.
The build process is based on the process of cross-compilation. Cross-compilation is normally used for building a compiler and its toolchain for a machine different from the one that is used for the build. This is not strictly needed for LFS, since the machine where the new system will run is the same as the one used for the build. But cross-compilation has the great advantage that anything that is cross-compiled cannot depend on the host environment.
The LFS book is not, and does not contain a general tutorial to build a cross (or native) toolchain. Don't use the command in the book for a cross toolchain which will be used for some purpose other than building LFS, unless you really understand what you are doing.
Cross-compilation involves some concepts that deserve a section on their own. Although this section may be omitted in a first reading, coming back to it later will be beneficial to your full understanding of the process.
Let us first define some terms used in this context:
is the machine where we build programs. Note that this machine is referred to as the “host” in other sections.
is the machine/system where the built programs will run. Note that this use of “host” is not the same as in other sections.
is only used for compilers. It is the machine the compiler produces code for. It may be different from both build and host.
As an example, let us imagine the following scenario (sometimes referred to as “Canadian Cross”): we may have a compiler on a slow machine only, let's call it machine A, and the compiler ccA. We may have also a fast machine (B), but with no compiler, and we may want to produce code for another slow machine (C). To build a compiler for machine C, we would have three stages:
Stage | Build | Host | Target | Action |
---|---|---|---|---|
1 | A | A | B | build cross-compiler cc1 using ccA on machine A |
2 | A | B | C | build cross-compiler cc2 using cc1 on machine A |
3 | B | C | C | build compiler ccC using cc2 on machine B |
Then, all the other programs needed by machine C can be compiled using cc2 on the fast machine B. Note that unless B can run programs produced for C, there is no way to test the built programs until machine C itself is running. For example, for testing ccC, we may want to add a fourth stage:
Stage | Build | Host | Target | Action |
---|---|---|---|---|
4 | C | C | C | rebuild and test ccC using itself on machine C |
In the example above, only cc1 and cc2 are cross-compilers, that is, they produce code for a machine different from the one they are run on. The other compilers ccA and ccC produce code for the machine they are run on. Such compilers are called native compilers.
Almost all the build systems use names of the form
cpu-vendor-kernel-os referred to as the machine triplet. An
astute reader may wonder why a “triplet” refers to a four
component name. The reason is history: initially, three
component names were enough to designate a machine
unambiguously, but with new machines and systems appearing,
that proved insufficient. The word “triplet” remained. A simple way
to determine your machine triplet is to run the
config.guess
script that comes with the source for many packages. Unpack
the binutils sources and run the script: ./config.guess
and note
the output. For example, for a 32-bit Intel processor the
output will be i686-pc-linux-gnu. On a 64-bit
system it will be x86_64-pc-linux-gnu.
Also be aware of the name of the platform's dynamic linker,
often referred to as the dynamic loader (not to be confused
with the standard linker ld that is part of
binutils). The dynamic linker provided by Glibc finds and
loads the shared libraries needed by a program, prepares
the program to run, and then runs it. The name of the
dynamic linker for a 32-bit Intel machine is ld-linux.so.2
and is ld-linux-x86-64.so.2
for 64-bit systems.
A sure-fire way to determine the name of the dynamic linker
is to inspect a random binary from the host system by
running: readelf -l
<name of binary> | grep interpreter
and noting the output. The authoritative reference covering
all platforms is in the shlib-versions
file in the root of the
Glibc source tree.
In order to fake a cross compilation in LFS, the name of the
host triplet is slightly adjusted by changing the "vendor"
field in the LFS_TGT
variable. We
also use the --with-sysroot
option when
building the cross linker and cross compiler to tell them
where to find the needed host files. This ensures that none
of the other programs built in Chapter 6
can link to libraries on the build machine. Only two stages
are mandatory, and one more for tests:
Stage | Build | Host | Target | Action |
---|---|---|---|---|
1 | pc | pc | lfs | build cross-compiler cc1 using cc-pc on pc |
2 | pc | lfs | lfs | build compiler cc-lfs using cc1 on pc |
3 | lfs | lfs | lfs | rebuild and test cc-lfs using itself on lfs |
In the above table, “on pc” means the commands are run on a machine using the already installed distribution. “On lfs” means the commands are run in a chrooted environment.
Now, there is more about cross-compiling: the C language is not just a compiler, but also defines a standard library. In this book, the GNU C library, named glibc, is used. This library must be compiled for the lfs machine, that is, using the cross compiler cc1. But the compiler itself uses an internal library implementing complex instructions not available in the assembler instruction set. This internal library is named libgcc, and must be linked to the glibc library to be fully functional! Furthermore, the standard library for C++ (libstdc++) also needs being linked to glibc. The solution to this chicken and egg problem is to first build a degraded cc1 based libgcc, lacking some functionalities such as threads and exception handling, then build glibc using this degraded compiler (glibc itself is not degraded), then build libstdc++. But this last library will lack the same functionalities as libgcc.
This is not the end of the story: the conclusion of the preceding paragraph is that cc1 is unable to build a fully functional libstdc++, but this is the only compiler available for building the C/C++ libraries during stage 2! Of course, the compiler built during stage 2, cc-lfs, would be able to build those libraries, but (1) the build system of GCC does not know that it is usable on pc, and (2) using it on pc would be at risk of linking to the pc libraries, since cc-lfs is a native compiler. So we have to build libstdc++ later, in chroot.
The cross-compiler will be installed in a separate
$LFS/tools
directory, since it
will not be part of the final system.
Binutils is installed first because the configure runs of both GCC and Glibc perform various feature tests on the assembler and linker to determine which software features to enable or disable. This is more important than one might first realize. An incorrectly configured GCC or Glibc can result in a subtly broken toolchain, where the impact of such breakage might not show up until near the end of the build of an entire distribution. A test suite failure will usually highlight this error before too much additional work is performed.
Binutils installs its assembler and linker in two locations,
$LFS/tools/bin
and $LFS/tools/$LFS_TGT/bin
. The tools in one
location are hard linked to the other. An important facet of
the linker is its library search order. Detailed information
can be obtained from ld by passing it the
--verbose
flag. For
example, $LFS_TGT-ld --verbose
| grep SEARCH will illustrate the current
search paths and their order. It shows which files are linked
by ld by
compiling a dummy program and passing the --verbose
switch to the linker.
For example, $LFS_TGT-gcc
dummy.c -Wl,--verbose 2>&1 | grep
succeeded will show all the files
successfully opened during the linking.
The next package installed is GCC. An example of what can be seen during its run of configure is:
checking what assembler to use... /mnt/lfs/tools/i686-lfs-linux-gnu/bin/as
checking what linker to use... /mnt/lfs/tools/i686-lfs-linux-gnu/bin/ld
This is important for the reasons mentioned above. It also demonstrates that GCC's configure script does not search the PATH directories to find which tools to use. However, during the actual operation of gcc itself, the same search paths are not necessarily used. To find out which standard linker gcc will use, run: $LFS_TGT-gcc -print-prog-name=ld.
Detailed information can be obtained from gcc by passing it the
-v
command line
option while compiling a dummy program. For example,
gcc -v dummy.c
will show detailed information about the preprocessor,
compilation, and assembly stages, including gcc's included search paths
and their order.
Next installed are sanitized Linux API headers. These allow the standard C library (Glibc) to interface with features that the Linux kernel will provide.
The next package installed is Glibc. The most important
considerations for building Glibc are the compiler, binary
tools, and kernel headers. The compiler is generally not an
issue since Glibc will always use the compiler relating to
the --host
parameter
passed to its configure script; e.g. in our case, the
compiler will be $LFS_TGT-gcc. The binary
tools and kernel headers can be a bit more complicated.
Therefore, we take no risks and use the available configure
switches to enforce the correct selections. After the run of
configure,
check the contents of the config.make
file in the build
directory for all important details.
Note the use of CC="$LFS_TGT-gcc"
(with
$LFS_TGT
expanded) to control
which binary tools are used and the use of the -nostdinc
and -isystem
flags to control the
compiler's include search path. These items highlight an
important aspect of the Glibc package—it is very
self-sufficient in terms of its build machinery and generally
does not rely on toolchain defaults.
As said above, the standard C++ library is compiled next,
followed in Chapter 6
by all the programs that need themselves to be built. The
install step of all those packages uses the DESTDIR
variable to have the programs land
into the LFS filesystem.
At the end of Chapter 6
the native lfs compiler is installed. First binutils-pass2 is
built, with the same DESTDIR
install as the other programs, then the second pass of GCC is
constructed, omitting libstdc++ and other non-important
libraries. Due to some weird logic in GCC's configure script,
CC_FOR_TARGET
ends up as
cc when the
host is the same as the target, but is different from the
build system. This is why CC_FOR_TARGET=$LFS_TGT-gcc
is
put explicitly into the configure options.
Upon entering the chroot environment in Chapter 7, the first task is to install libstdc++. Then temporary installations of programs needed for the proper operation of the toolchain are performed. From this point onwards, the core toolchain is self-contained and self-hosted. In Chapter 8, final versions of all the packages needed for a fully functional system are built, tested and installed.
When building packages there are several assumptions made within the instructions:
Several of the packages are patched before compilation, but only when the patch is needed to circumvent a problem. A patch is often needed in both this and the following chapters, but sometimes in only one location. Therefore, do not be concerned if instructions for a downloaded patch seem to be missing. Warning messages about offset or fuzz may also be encountered when applying a patch. Do not worry about these warnings, as the patch was still successfully applied.
During the compilation of most packages, there will be several warnings that scroll by on the screen. These are normal and can safely be ignored. These warnings are as they appear—warnings about deprecated, but not invalid, use of the C or C++ syntax. C standards change fairly often, and some packages still use the older standard. This is not a problem, but does prompt the warning.
Check one last time that the LFS
environment variable is set up
properly:
echo $LFS
Make sure the output shows the path to the LFS
partition's mount point, which is /mnt/lfs
, using our example.
Finally, two important items must be emphasized:
The build instructions assume that the Host System Requirements, including symbolic links, have been set properly:
bash is the shell in use.
sh is a symbolic link to bash.
/usr/bin/awk is a symbolic link to gawk.
/usr/bin/yacc is a symbolic link to bison or a small script that executes bison.
To re-emphasize the build process:
Place all the sources and patches in a directory
that will be accessible from the chroot
environment such as /mnt/lfs/sources/
.
Change to the sources directory.
Using the tar program, extract the package to be built. In Chapter 5 and Chapter 6, ensure you are the lfs user when extracting the package.
All methods to get the source code tree being built in-position, except extracting the package tarball, are not supported. Notably, using cp -R to copy the source code tree somewhere else can destroy links and timestamps in the sources tree and cause building failure.
Change to the directory created when the package was extracted.
Follow the book's instructions for building the package.
Change back to the sources directory.
Delete the extracted source directory unless instructed otherwise.
This chapter shows how to build a cross-compiler and its associated tools. Although here cross-compilation is faked, the principles are the same as for a real cross-toolchain.
The programs compiled in this chapter will be installed under
the $LFS/tools
directory to keep
them separate from the files installed in the following
chapters. The libraries, on the other hand, are installed into
their final place, since they pertain to the system we want to
build.
The Binutils package contains a linker, an assembler, and other tools for handling object files.
Go back and re-read the notes in the section titled General Compilation Instructions. Understanding the notes labeled important can save you a lot of problems later.
It is important that Binutils be the first package compiled because both Glibc and GCC perform various tests on the available linker and assembler to determine which of their own features to enable.
The Binutils documentation recommends building Binutils in a dedicated build directory:
mkdir -v build cd build
In order for the SBU values listed in the rest of the book
to be of any use, measure the time it takes to build this
package from the configuration, up to and including the
first install. To achieve this easily, wrap the commands in
a time
command like this: time {
../configure ... && make && make install;
}
.
Now prepare Binutils for compilation:
../configure --prefix=$LFS/tools \ --with-sysroot=$LFS \ --target=$LFS_TGT \ --disable-nls \ --disable-werror
The meaning of the configure options:
--prefix=$LFS/tools
This tells the configure script to prepare to install
the binutils programs in the $LFS/tools
directory.
--with-sysroot=$LFS
For cross compilation, this tells the build system to look in $LFS for the target system libraries as needed.
--target=$LFS_TGT
Because the machine description in the LFS_TGT
variable is slightly different
than the value returned by the config.guess script,
this switch will tell the configure script to
adjust binutil's build system for building a cross
linker.
--disable-nls
This disables internationalization as i18n is not needed for the temporary tools.
--disable-werror
This prevents the build from stopping in the event that there are warnings from the host's compiler.
Continue with compiling the package:
make
Install the package:
make install
Details on this package are located in Section 8.18.2, “Contents of Binutils.”
The GCC package contains the GNU compiler collection, which includes the C and C++ compilers.
GCC requires the GMP, MPFR and MPC packages. As these packages may not be included in your host distribution, they will be built with GCC. Unpack each package into the GCC source directory and rename the resulting directories so the GCC build procedures will automatically use them:
There are frequent misunderstandings about this chapter. The procedures are the same as every other chapter as explained earlier (Package build instructions). First extract the gcc tarball from the sources directory and then change to the directory created. Only then should you proceed with the instructions below.
tar -xf ../mpfr-4.1.0.tar.xz mv -v mpfr-4.1.0 mpfr tar -xf ../gmp-6.2.1.tar.xz mv -v gmp-6.2.1 gmp tar -xf ../mpc-1.2.1.tar.gz mv -v mpc-1.2.1 mpc
On x86_64 hosts, set the default directory name for 64-bit libraries to “lib”:
case $(uname -m) in x86_64) sed -e '/m64=/s/lib64/lib/' \ -i.orig gcc/config/i386/t-linux64 ;; esac
The GCC documentation recommends building GCC in a dedicated build directory:
mkdir -v build cd build
Prepare GCC for compilation:
../configure \ --target=$LFS_TGT \ --prefix=$LFS/tools \ --with-glibc-version=2.35 \ --with-sysroot=$LFS \ --with-newlib \ --without-headers \ --enable-initfini-array \ --disable-nls \ --disable-shared \ --disable-multilib \ --disable-decimal-float \ --disable-threads \ --disable-libatomic \ --disable-libgomp \ --disable-libquadmath \ --disable-libssp \ --disable-libvtv \ --disable-libstdcxx \ --enable-languages=c,c++
The meaning of the configure options:
--with-glibc-version=2.35
This option specifies the version of glibc which will be used on the target. It is not relevant to the libc of the host distro because everything compiled by pass1 gcc will run in the chroot environment, which is isolated from libc of the host distro.
--with-newlib
Since a working C library is not yet available, this ensures that the inhibit_libc constant is defined when building libgcc. This prevents the compiling of any code that requires libc support.
--without-headers
When creating a complete cross-compiler, GCC requires standard headers compatible with the target system. For our purposes these headers will not be needed. This switch prevents GCC from looking for them.
--enable-initfini-array
This switch forces the use of some internal data structures that are needed but cannot be detected when building a cross compiler.
--disable-shared
This switch forces GCC to link its internal libraries statically. We need this because the shared libraries require glibc, which is not yet installed on the target system.
--disable-multilib
On x86_64, LFS does not support a multilib configuration. This switch is harmless for x86.
--disable-decimal-float,
--disable-threads, --disable-libatomic,
--disable-libgomp, --disable-libquadmath,
--disable-libssp, --disable-libvtv,
--disable-libstdcxx
These switches disable support for the decimal floating point extension, threading, libatomic, libgomp, libquadmath, libssp, libvtv, and the C++ standard library respectively. These features will fail to compile when building a cross-compiler and are not necessary for the task of cross-compiling the temporary libc.
--enable-languages=c,c++
This option ensures that only the C and C++ compilers are built. These are the only languages needed now.
Compile GCC by running:
make
Install the package:
make install
This build of GCC has installed a couple of internal system
headers. Normally one of them, limits.h
, would in turn include the
corresponding system limits.h
header, in this case, $LFS/usr/include/limits.h
. However, at the
time of this build of GCC $LFS/usr/include/limits.h
does not exist,
so the internal header that has just been installed is a
partial, self-contained file and does not include the
extended features of the system header. This is adequate for
building glibc, but the full internal header will be needed
later. Create a full version of the internal header using a
command that is identical to what the GCC build system does
in normal circumstances:
cd .. cat gcc/limitx.h gcc/glimits.h gcc/limity.h > \ `dirname $($LFS_TGT-gcc -print-libgcc-file-name)`/install-tools/include/limits.h
Details on this package are located in Section 8.26.2, “Contents of GCC.”
The Linux API Headers (in linux-5.16.9.tar.xz) expose the kernel's API for use by Glibc.
The Linux kernel needs to expose an Application Programming Interface (API) for the system's C library (Glibc in LFS) to use. This is done by way of sanitizing various C header files that are shipped in the Linux kernel source tarball.
Make sure there are no stale files embedded in the package:
make mrproper
Now extract the user-visible kernel headers from the source.
The recommended make target “headers_install” cannot be used,
because it requires rsync,
which may not be available. The headers are first placed in
./usr
, then copied to the
needed location.
make headers find usr/include -name '.*' -delete rm usr/include/Makefile cp -rv usr/include $LFS/usr
The Linux API ASM Headers |
|
The Linux API ASM Generic Headers |
|
The Linux API DRM Headers |
|
The Linux API Linux Headers |
|
The Linux API Miscellaneous Headers |
|
The Linux API MTD Headers |
|
The Linux API RDMA Headers |
|
The Linux API SCSI Headers |
|
The Linux API Sound Headers |
|
The Linux API Video Headers |
|
The Linux API Xen Headers |
The Glibc package contains the main C library. This library provides the basic routines for allocating memory, searching directories, opening and closing files, reading and writing files, string handling, pattern matching, arithmetic, and so on.
First, create a symbolic link for LSB compliance. Additionally, for x86_64, create a compatibility symbolic link required for proper operation of the dynamic library loader:
case $(uname -m) in i?86) ln -sfv ld-linux.so.2 $LFS/lib/ld-lsb.so.3 ;; x86_64) ln -sfv ../lib/ld-linux-x86-64.so.2 $LFS/lib64 ln -sfv ../lib/ld-linux-x86-64.so.2 $LFS/lib64/ld-lsb-x86-64.so.3 ;; esac
The above command is correct. The ln command has a few
syntactic versions, so be sure to check info coreutils ln and
ln(1)
before reporting what
you may think is an error.
Some of the Glibc programs use the non-FHS compliant
/var/db
directory to store
their runtime data. Apply the following patch to make such
programs store their runtime data in the FHS-compliant
locations:
patch -Np1 -i ../glibc-2.35-fhs-1.patch
The Glibc documentation recommends building Glibc in a dedicated build directory:
mkdir -v build cd build
Ensure that the ldconfig and sln utilites are installed
into /usr/sbin
:
echo "rootsbindir=/usr/sbin" > configparms
Next, prepare Glibc for compilation:
../configure \ --prefix=/usr \ --host=$LFS_TGT \ --build=$(../scripts/config.guess) \ --enable-kernel=3.2 \ --with-headers=$LFS/usr/include \ libc_cv_slibdir=/usr/lib
The meaning of the configure options:
--host=$LFS_TGT,
--build=$(../scripts/config.guess)
The combined effect of these switches is that Glibc's
build system configures itself to be cross-compiled,
using the cross-linker and cross-compiler in
$LFS/tools
.
--enable-kernel=3.2
This tells Glibc to compile the library with support for 3.2 and later Linux kernels. Workarounds for older kernels are not enabled.
--with-headers=$LFS/usr/include
This tells Glibc to compile itself against the headers recently installed to the $LFS/usr/include directory, so that it knows exactly what features the kernel has and can optimize itself accordingly.
libc_cv_slibdir=/usr/lib
This ensures that the library is installed in /usr/lib instead of the default /lib64 on 64 bit machines.
During this stage the following warning might appear:
configure: WARNING: *** These auxiliary programs are missing or *** incompatible versions: msgfmt *** some features will be disabled. *** Check the INSTALL file for required versions.
The missing or incompatible msgfmt program is generally harmless. This msgfmt program is part of the Gettext package which the host distribution should provide.
There have been reports that this package may fail when building as a "parallel make". If this occurs, rerun the make command with a "-j1" option.
Compile the package:
make
Install the package:
If LFS
is not properly set, and
despite the recommendations, you are building as
root
, the next command will
install the newly built glibc to your host system, which
most likely will render it unusable. So double check that
the environment is correctly set, before running the
following command.
make DESTDIR=$LFS install
The meaning of the make install option:
DESTDIR=$LFS
The DESTDIR
make variable is
used by almost all packages to define the location
where the package should be installed. If it is not
set, it defaults to the root (/
) directory. Here we specify that
the package be installed in $LFS
, which will become the root
after Section 7.4,
“Entering the Chroot Environment”.
Fix hardcoded path to the executable loader in ldd script:
sed '/RTLDLIST=/s@/usr@@g' -i $LFS/usr/bin/ldd
At this point, it is imperative to stop and ensure that the basic functions (compiling and linking) of the new toolchain are working as expected. To perform a sanity check, run the following commands:
echo 'int main(){}' > dummy.c $LFS_TGT-gcc dummy.c readelf -l a.out | grep '/ld-linux'
If everything is working correctly, there should be no errors, and the output of the last command will be of the form:
[Requesting program interpreter: /lib64/ld-linux-x86-64.so.2]
Note that for 32-bit machines, the interpreter name will be
/lib/ld-linux.so.2
.
If the output is not shown as above or there was no output at all, then something is wrong. Investigate and retrace the steps to find out where the problem is and correct it. This issue must be resolved before continuing on.
Once all is well, clean up the test files:
rm -v dummy.c a.out
Building packages in the next chapter will serve as an additional check that the toolchain has been built properly. If some package, especially binutils-pass2 or gcc-pass2, fails to build, it is an indication that something has gone wrong with the previous Binutils, GCC, or Glibc installations.
Now that our cross-toolchain is complete, finalize the installation of the limits.h header. For doing so, run a utility provided by the GCC developers:
$LFS/tools/libexec/gcc/$LFS_TGT/11.2.0/install-tools/mkheaders
Details on this package are located in Section 8.5.3, “Contents of Glibc.”
Libstdc++ is the standard C++ library. It is needed to compile C++ code (part of GCC is written in C++), but we had to defer its installation when we built gcc-pass1 because it depends on glibc, which was not yet available in the target directory.
Libstdc++ is part of the
GCC sources. You should first unpack the GCC tarball and
change to the gcc-11.2.0
directory.
Create a separate build directory for libstdc++ and enter it:
mkdir -v build cd build
Prepare libstdc++ for compilation:
../libstdc++-v3/configure \ --host=$LFS_TGT \ --build=$(../config.guess) \ --prefix=/usr \ --disable-multilib \ --disable-nls \ --disable-libstdcxx-pch \ --with-gxx-include-dir=/tools/$LFS_TGT/include/c++/11.2.0
The meaning of the configure options:
--host=...
Specifies that the cross compiler we have just built
should be used instead of the one in /usr/bin
.
--disable-libstdcxx-pch
This switch prevents the installation of precompiled include files, which are not needed at this stage.
--with-gxx-include-dir=/tools/$LFS_TGT/include/c++/11.2.0
This specifies the installation directory for include
files. Because libstdc++ is the standard C++ library
for LFS, this directory should match the location where
the C++ compiler ($LFS_TGT-g++) would
search for the standard C++ include files. In a normal
build, this information is automatically passed to the
libstdc++ configure options
from the top level directory. In our case, this
information must be explicitly given. The C++ compiler
will prepend the sysroot path $LFS
(specified building GCC pass 1)
to the include file search path, so it will actually
search in $LFS/tools/$LFS_TGT/include/c++/11.2.0
.
The combination of the DESTDIR
variable (in the
make
install command below) and this switch
ensures to install the headers there.
Compile libstdc++ by running:
make
Install the library:
make DESTDIR=$LFS install
Details on this package are located in Section 8.26.2, “Contents of GCC.”
This chapter shows how to cross-compile basic utilities using the just built cross-toolchain. Those utilities are installed into their final location, but cannot be used yet. Basic tasks still rely on the host's tools. Nevertheless, the installed libraries are used when linking.
Using the utilities will be possible in next chapter after entering the “chroot” environment. But all the packages built in the present chapter need to be built before we do that. Therefore we cannot be independent of the host system yet.
Once again, let us recall that improper setting of LFS
together with building as root
, may render your computer unusable.
This whole chapter must be done as user lfs
, with the environment as described in
Section 4.4,
“Setting Up the Environment”.
The M4 package contains a macro processor.
Prepare M4 for compilation:
./configure --prefix=/usr \ --host=$LFS_TGT \ --build=$(build-aux/config.guess)
Compile the package:
make
Install the package:
make DESTDIR=$LFS install
Details on this package are located in Section 8.12.2, “Contents of M4.”
The Ncurses package contains libraries for terminal-independent handling of character screens.
First, ensure that gawk is found first during configuration:
sed -i s/mawk// configure
Then, run the following commands to build the “tic” program on the build host:
mkdir build pushd build ../configure make -C include make -C progs tic popd
Prepare Ncurses for compilation:
./configure --prefix=/usr \ --host=$LFS_TGT \ --build=$(./config.guess) \ --mandir=/usr/share/man \ --with-manpage-format=normal \ --with-shared \ --without-debug \ --without-ada \ --without-normal \ --disable-stripping \ --enable-widec
The meaning of the new configure options:
--with-manpage-format=normal
This prevents Ncurses installing compressed manual pages, which may happen if the host distribution itself has compressed manual pages.
--without-ada
This ensures that Ncurses does not build support for the Ada compiler which may be present on the host but will not be available once we enter the chroot environment.
--disable-stripping
This switch prevents the building system from stripping the programs using strip program from the host. Using host tools on cross-compiled program can cause failure.
--enable-widec
This switch causes wide-character libraries (e.g.,
libncursesw.so.6.3
) to be
built instead of normal ones (e.g., libncurses.so.6.3
). These
wide-character libraries are usable in both multibyte
and traditional 8-bit locales, while normal libraries
work properly only in 8-bit locales. Wide-character and
normal libraries are source-compatible, but not
binary-compatible.
--without-normal
This switch disables building and installing most static libraries.
Compile the package:
make
Install the package:
make DESTDIR=$LFS TIC_PATH=$(pwd)/build/progs/tic install echo "INPUT(-lncursesw)" > $LFS/usr/lib/libncurses.so
The meaning of the install options:
TIC_PATH=$(pwd)/build/progs/tic
We need to pass the path of the just built tic able to run on the building machine, so that the terminal database can be created without errors.
The libncurses.so
library
is needed by a few packages we will build soon. We
create this small linker script, as this is what is
done in
Chapter 8.
Details on this package are located in Section 8.28.2, “Contents of Ncurses.”
The Bash package contains the Bourne-Again SHell.
Prepare Bash for compilation:
./configure --prefix=/usr \ --build=$(support/config.guess) \ --host=$LFS_TGT \ --without-bash-malloc
The meaning of the configure options:
--without-bash-malloc
This option turns off the use of Bash's memory
allocation (malloc
)
function which is known to cause segmentation faults.
By turning this option off, Bash will use the
malloc
functions from
Glibc which are more stable.
Compile the package:
make
Install the package:
make DESTDIR=$LFS install
Make a link for the programs that use sh for a shell:
ln -sv bash $LFS/bin/sh
Details on this package are located in Section 8.34.2, “Contents of Bash.”
The Coreutils package contains utilities for showing and setting the basic system characteristics.
Prepare Coreutils for compilation:
./configure --prefix=/usr \ --host=$LFS_TGT \ --build=$(build-aux/config.guess) \ --enable-install-program=hostname \ --enable-no-install-program=kill,uptime
The meaning of the configure options:
--enable-install-program=hostname
This enables the hostname binary to be built and installed – it is disabled by default but is required by the Perl test suite.
Compile the package:
make
Install the package:
make DESTDIR=$LFS install
Move programs to their final expected locations. Although this is not necessary in this temporary environment, we must do so because some programs harcode executable locations:
mv -v $LFS/usr/bin/chroot $LFS/usr/sbin mkdir -pv $LFS/usr/share/man/man8 mv -v $LFS/usr/share/man/man1/chroot.1 $LFS/usr/share/man/man8/chroot.8 sed -i 's/"1"/"8"/' $LFS/usr/share/man/man8/chroot.8
Details on this package are located in Section 8.53.2, “Contents of Coreutils.”
The Diffutils package contains programs that show the differences between files or directories.
Prepare Diffutils for compilation:
./configure --prefix=/usr --host=$LFS_TGT
Compile the package:
make
Install the package:
make DESTDIR=$LFS install
Details on this package are located in Section 8.55.2, “Contents of Diffutils.”
The File package contains a utility for determining the type of a given file or files.
The file command on the build host needs to be same version as the one we are building in order to create the signature file. Run the following commands to build it:
mkdir build pushd build ../configure --disable-bzlib \ --disable-libseccomp \ --disable-xzlib \ --disable-zlib make popd
The meaning of the new configure option:
--disable-*
The configuration script attempts to use some packages from the host distribution if the corresponding library files exist. It may cause compilation failure if a library file exists, but the corresponding header files do not. These options prevent using these unneeded capabilities from the host.
Prepare File for compilation:
./configure --prefix=/usr --host=$LFS_TGT --build=$(./config.guess)
Compile the package:
make FILE_COMPILE=$(pwd)/build/src/file
Install the package:
make DESTDIR=$LFS install
Details on this package are located in Section 8.10.2, “Contents of File.”
The Findutils package contains programs to find files. These programs are provided to recursively search through a directory tree and to create, maintain, and search a database (often faster than the recursive find, but is unreliable if the database has not been recently updated).
Prepare Findutils for compilation:
./configure --prefix=/usr \ --localstatedir=/var/lib/locate \ --host=$LFS_TGT \ --build=$(build-aux/config.guess)
Compile the package:
make
Install the package:
make DESTDIR=$LFS install
Details on this package are located in Section 8.57.2, “Contents of Findutils.”
The Gawk package contains programs for manipulating text files.
First, ensure some unneeded files are not installed:
sed -i 's/extras//' Makefile.in
Prepare Gawk for compilation:
./configure --prefix=/usr \ --host=$LFS_TGT \ --build=$(build-aux/config.guess)
Compile the package:
make
Install the package:
make DESTDIR=$LFS install
Details on this package are located in Section 8.56.2, “Contents of Gawk.”
The Grep package contains programs for searching through the contents of files.
Prepare Grep for compilation:
./configure --prefix=/usr \ --host=$LFS_TGT
Compile the package:
make
Install the package:
make DESTDIR=$LFS install
Details on this package are located in Section 8.33.2, “Contents of Grep.”
The Gzip package contains programs for compressing and decompressing files.
Prepare Gzip for compilation:
./configure --prefix=/usr --host=$LFS_TGT
Compile the package:
make
Install the package:
make DESTDIR=$LFS install
Details on this package are located in Section 8.60.2, “Contents of Gzip.”
The Make package contains a program for controlling the generation of executables and other non-source files of a package from source files.
Prepare Make for compilation:
./configure --prefix=/usr \ --without-guile \ --host=$LFS_TGT \ --build=$(build-aux/config.guess)
The meaning of the new configure option:
--without-guile
Although we are cross-compiling, configure tries to use guile from the build host if it finds it. This makes compilation fail, so this switch prevents using it.
Compile the package:
make
Install the package:
make DESTDIR=$LFS install
Details on this package are located in Section 8.64.2, “Contents of Make.”
The Patch package contains a program for modifying or creating files by applying a “patch” file typically created by the diff program.
Prepare Patch for compilation:
./configure --prefix=/usr \ --host=$LFS_TGT \ --build=$(build-aux/config.guess)
Compile the package:
make
Install the package:
make DESTDIR=$LFS install
Details on this package are located in Section 8.65.2, “Contents of Patch.”
The Sed package contains a stream editor.
Prepare Sed for compilation:
./configure --prefix=/usr \ --host=$LFS_TGT
Compile the package:
make
Install the package:
make DESTDIR=$LFS install
Details on this package are located in Section 8.29.2, “Contents of Sed.”
The Tar package provides the ability to create tar archives as well as perform various other kinds of archive manipulation. Tar can be used on previously created archives to extract files, to store additional files, or to update or list files which were already stored.
Prepare Tar for compilation:
./configure --prefix=/usr \ --host=$LFS_TGT \ --build=$(build-aux/config.guess)
Compile the package:
make
Install the package:
make DESTDIR=$LFS install
Details on this package are located in Section 8.66.2, “Contents of Tar.”
The Xz package contains programs for compressing and decompressing files. It provides capabilities for the lzma and the newer xz compression formats. Compressing text files with xz yields a better compression percentage than with the traditional gzip or bzip2 commands.
Prepare Xz for compilation:
./configure --prefix=/usr \ --host=$LFS_TGT \ --build=$(build-aux/config.guess) \ --disable-static \ --docdir=/usr/share/doc/xz-5.2.5
Compile the package:
make
Install the package:
make DESTDIR=$LFS install
Details on this package are located in Section 8.8.2, “Contents of Xz.”
The Binutils package contains a linker, an assembler, and other tools for handling object files.
Binutils ships an outdated libtool copy in the tarball. It lacks sysroot support so the produced binaries will be mistakenly linked to libraries from the host distro. Workaround this issue:
sed '6009s/$add_dir//' -i ltmain.sh
Create a separate build directory again:
mkdir -v build cd build
Prepare Binutils for compilation:
../configure \ --prefix=/usr \ --build=$(../config.guess) \ --host=$LFS_TGT \ --disable-nls \ --enable-shared \ --disable-werror \ --enable-64-bit-bfd
The meaning of the new configure options:
--enable-shared
Builds libbfd
as a shared
library.
--enable-64-bit-bfd
Enables 64-bit support (on hosts with narrower word sizes). May not be needed on 64-bit systems, but does no harm.
Compile the package:
make
Install the package:
make DESTDIR=$LFS install
Details on this package are located in Section 8.18.2, “Contents of Binutils.”
The GCC package contains the GNU compiler collection, which includes the C and C++ compilers.
As in the first build of GCC, the GMP, MPFR, and MPC packages are required. Unpack the tarballs and move them into the required directory names:
tar -xf ../mpfr-4.1.0.tar.xz mv -v mpfr-4.1.0 mpfr tar -xf ../gmp-6.2.1.tar.xz mv -v gmp-6.2.1 gmp tar -xf ../mpc-1.2.1.tar.gz mv -v mpc-1.2.1 mpc
If building on x86_64, change the default directory name for 64-bit libraries to “lib”:
case $(uname -m) in x86_64) sed -e '/m64=/s/lib64/lib/' -i.orig gcc/config/i386/t-linux64 ;; esac
Create a separate build directory again:
mkdir -v build cd build
Create a symlink that allows libgcc to be built with posix threads support:
mkdir -pv $LFS_TGT/libgcc ln -s ../../../libgcc/gthr-posix.h $LFS_TGT/libgcc/gthr-default.h
Before starting to build GCC, remember to unset any environment variables that override the default optimization flags.
Now prepare GCC for compilation:
../configure \ --build=$(../config.guess) \ --host=$LFS_TGT \ --prefix=/usr \ CC_FOR_TARGET=$LFS_TGT-gcc \ --with-build-sysroot=$LFS \ --enable-initfini-array \ --disable-nls \ --disable-multilib \ --disable-decimal-float \ --disable-libatomic \ --disable-libgomp \ --disable-libquadmath \ --disable-libssp \ --disable-libvtv \ --disable-libstdcxx \ --enable-languages=c,c++
The meaning of the new configure options:
-with-build-sysroot=$LFS
Normally, using --host
ensures that a
cross-compiler is used for building GCC, and that
compiler knows that it has to look for headers and
libraries in $LFS
. But
the build system of GCC uses other tools, which are not
aware of this location. This switch is needed to have
them find the needed files in $LFS
, and not on the host.
--enable-initfini-array
This option is automatically enabled when building a native compiler with a native compiler on x86. But here, we build with a cross compiler, so we need to explicitly set this option.
Compile the package:
make
Install the package:
make DESTDIR=$LFS install
As a finishing touch, create a utility symlink. Many programs and scripts run cc instead of gcc, which is used to keep programs generic and therefore usable on all kinds of UNIX systems where the GNU C compiler is not always installed. Running cc leaves the system administrator free to decide which C compiler to install:
ln -sv gcc $LFS/usr/bin/cc
Details on this package are located in Section 8.26.2, “Contents of GCC.”
This chapter shows how to build the last missing bits of the temporary system: the tools needed by the build machinery of various packages. Now that all circular dependencies have been resolved, a “chroot” environment, completely isolated from the host operating system (except for the running kernel), can be used for the build.
For proper operation of the isolated environment, some communication with the running kernel must be established. This is done through the so-called Virtual Kernel File Systems, which must be mounted when entering the chroot environment. You may want to check that they are mounted by issuing findmnt.
Until Section 7.4,
“Entering the Chroot Environment”, the commands
must be run as root
, with the
LFS
variable set. After entering
chroot, all commands are run as root, fortunately without
access to the OS of the computer you built LFS on. Be careful
anyway, as it is easy to destroy the whole LFS system with
badly formed commands.
The commands in the remainder of this book must be performed
while logged in as user root
and no longer as user lfs
.
Also, double check that $LFS
is
set in root
's environment.
Currently, the whole directory hierarchy in $LFS
is owned by the user lfs
, a user that exists only on the host
system. If the directories and files under $LFS
are kept as they are, they will be owned
by a user ID without a corresponding account. This is dangerous
because a user account created later could get this same user
ID and would own all the files under $LFS
, thus exposing these files to possible
malicious manipulation.
To address this issue, change the ownership of the $LFS/*
directories to user root
by running the following command:
chown -R root:root $LFS/{usr,lib,var,etc,bin,sbin,tools} case $(uname -m) in x86_64) chown -R root:root $LFS/lib64 ;; esac
Various file systems exported by the kernel are used to communicate to and from the kernel itself. These file systems are virtual in that no disk space is used for them. The content of the file systems resides in memory.
Begin by creating directories onto which the file systems will be mounted:
mkdir -pv $LFS/{dev,proc,sys,run}
When the kernel boots the system, it requires the presence of
a few device nodes, in particular the console
and null
devices. The device nodes must be
created on the hard disk so that they are available before
the kernel populates /dev
),
and additionally when Linux is started with init=/bin/bash
. Create the
devices by running the following commands:
mknod -m 600 $LFS/dev/console c 5 1 mknod -m 666 $LFS/dev/null c 1 3
The recommended method of populating the /dev
directory with devices is to mount a
virtual filesystem (such as tmpfs
) on the /dev
directory, and allow the devices to be
created dynamically on that virtual filesystem as they are
detected or accessed. Device creation is generally done
during the boot process by Udev. Since this new system does
not yet have Udev and has not yet been booted, it is
necessary to mount and populate /dev
manually. This is accomplished by bind
mounting the host system's /dev
directory. A bind mount is a special type of mount that
allows you to create a mirror of a directory or mount point
to some other location. Use the following command to achieve
this:
mount -v --bind /dev $LFS/dev
Now mount the remaining virtual kernel filesystems:
mount -v --bind /dev/pts $LFS/dev/pts mount -vt proc proc $LFS/proc mount -vt sysfs sysfs $LFS/sys mount -vt tmpfs tmpfs $LFS/run
In some host systems, /dev/shm
is a symbolic link to /run/shm
.
The /run tmpfs was mounted above so in this case only a
directory needs to be created.
if [ -h $LFS/dev/shm ]; then mkdir -pv $LFS/$(readlink $LFS/dev/shm) fi
Now that all the packages which are required to build the rest
of the needed tools are on the system, it is time to enter the
chroot environment to finish installing the remaining temporary
tools. This environment will be in use also for installing the
final system. As user root
, run
the following command to enter the environment that is, at the
moment, populated with only the temporary tools:
chroot "$LFS" /usr/bin/env -i \ HOME=/root \ TERM="$TERM" \ PS1='(lfs chroot) \u:\w\$ ' \ PATH=/usr/bin:/usr/sbin \ /bin/bash --login
The -i
option given to
the env command
will clear all variables of the chroot environment. After that,
only the HOME
, TERM
, PS1
, and
PATH
variables are set again. The
TERM=$TERM
construct
will set the TERM
variable inside
chroot to the same value as outside chroot. This variable is
needed for programs like vim and less to operate properly. If
other variables are desired, such as CFLAGS
or CXXFLAGS
,
this is a good place to set them again.
From this point on, there is no need to use the LFS
variable anymore because all work will be
restricted to the LFS file system. This is because the Bash
shell is told that $LFS
is now
the root (/
) directory.
Notice that /tools/bin
is not in
the PATH
. This means that the cross
toolchain will no longer be used in the chroot environment.
Note that the bash prompt will say
I have no name!
This is
normal because the /etc/passwd
file has not been created yet.
It is important that all the commands throughout the remainder of this chapter and the following chapters are run from within the chroot environment. If you leave this environment for any reason (rebooting for example), ensure that the virtual kernel filesystems are mounted as explained in Section 7.3.2, “Mounting and Populating /dev” and Section 7.3.3, “Mounting Virtual Kernel File Systems” and enter chroot again before continuing with the installation.
It is time to create the full structure in the LFS file system.
Create some root-level directories that are not in the limited set required in the previous chapters by issuing the following command:
Some of the directories below have already been created earlier with explicit instructions or when installing some packages. They are repeated below for completeness.
mkdir -pv /{boot,home,mnt,opt,srv}
Create the required set of subdirectories below the root-level by issuing the following commands:
mkdir -pv /etc/{opt,sysconfig} mkdir -pv /lib/firmware mkdir -pv /media/{floppy,cdrom} mkdir -pv /usr/{,local/}{include,src} mkdir -pv /usr/local/{bin,lib,sbin} mkdir -pv /usr/{,local/}share/{color,dict,doc,info,locale,man} mkdir -pv /usr/{,local/}share/{misc,terminfo,zoneinfo} mkdir -pv /usr/{,local/}share/man/man{1..8} mkdir -pv /var/{cache,local,log,mail,opt,spool} mkdir -pv /var/lib/{color,misc,locate} ln -sfv /run /var/run ln -sfv /run/lock /var/lock install -dv -m 0750 /root install -dv -m 1777 /tmp /var/tmp
Directories are, by default, created with permission mode 755,
but this is not desirable for all directories. In the commands
above, two changes are made—one to the home directory of
user root
, and another to the
directories for temporary files.
The first mode change ensures that not just anybody can enter
the /root
directory—the
same as a normal user would do with his or her home directory.
The second mode change makes sure that any user can write to
the /tmp
and /var/tmp
directories, but cannot remove
another user's files from them. The latter is prohibited by the
so-called “sticky
bit,” the highest bit (1) in the 1777 bit
mask.
The directory tree is based on the Filesystem Hierarchy
Standard (FHS) (available at https://refspecs.linuxfoundation.org/fhs.shtml).
The FHS also specifies the optional existence of some
directories such as /usr/local/games
and /usr/share/games
. We create only the
directories that are needed. However, feel free to create
these directories.
Historically, Linux maintains a list of the mounted file
systems in the file /etc/mtab
.
Modern kernels maintain this list internally and exposes it to
the user via the /proc
filesystem. To satisfy utilities that expect the presence of
/etc/mtab
, create the following
symbolic link:
ln -sv /proc/self/mounts /etc/mtab
Create a basic /etc/hosts
file to
be referenced in some test suites, and in one of Perl's
configuration files as well:
cat > /etc/hosts << EOF 127.0.0.1 localhost $(hostname) ::1 localhost EOF
In order for user root
to be
able to login and for the name “root” to
be recognized, there must be relevant entries in the
/etc/passwd
and /etc/group
files.
Create the /etc/passwd
file by
running the following command:
cat > /etc/passwd << "EOF"
root:x:0:0:root:/root:/bin/bash
bin:x:1:1:bin:/dev/null:/usr/bin/false
daemon:x:6:6:Daemon User:/dev/null:/usr/bin/false
messagebus:x:18:18:D-Bus Message Daemon User:/run/dbus:/usr/bin/false
uuidd:x:80:80:UUID Generation Daemon User:/dev/null:/usr/bin/false
nobody:x:99:99:Unprivileged User:/dev/null:/usr/bin/false
EOF
The actual password for root
will be set later.
Create the /etc/group
file by
running the following command:
cat > /etc/group << "EOF"
root:x:0:
bin:x:1:daemon
sys:x:2:
kmem:x:3:
tape:x:4:
tty:x:5:
daemon:x:6:
floppy:x:7:
disk:x:8:
lp:x:9:
dialout:x:10:
audio:x:11:
video:x:12:
utmp:x:13:
usb:x:14:
cdrom:x:15:
adm:x:16:
messagebus:x:18:
input:x:24:
mail:x:34:
kvm:x:61:
uuidd:x:80:
wheel:x:97:
nogroup:x:99:
users:x:999:
EOF
The created groups are not part of any standard—they are
groups decided on in part by the requirements of the Udev
configuration in Chapter 9, and in part by common convention
employed by a number of existing Linux distributions. In
addition, some test suites rely on specific users or groups.
The Linux Standard Base (LSB, available at http://refspecs.linuxfoundation.org/lsb.shtml)
only recommends that, besides the group root
with a Group ID (GID) of 0, a group
bin
with a GID of 1 be present.
All other group names and GIDs can be chosen freely by the
system administrator since well-written programs do not depend
on GID numbers, but rather use the group's name.
Some tests in Chapter 8 need a regular user. We add this user here and delete this account at the end of that chapter.
echo "tester:x:101:101::/home/tester:/bin/bash" >> /etc/passwd echo "tester:x:101:" >> /etc/group install -o tester -d /home/tester
To remove the “I
have no name!” prompt, start a new shell.
Since the /etc/passwd
and
/etc/group
files have been
created, user name and group name resolution will now work:
exec /usr/bin/bash --login
The login, agetty, and init programs (and others) use a number of log files to record information such as who was logged into the system and when. However, these programs will not write to the log files if they do not already exist. Initialize the log files and give them proper permissions:
touch /var/log/{btmp,lastlog,faillog,wtmp} chgrp -v utmp /var/log/lastlog chmod -v 664 /var/log/lastlog chmod -v 600 /var/log/btmp
The /var/log/wtmp
file records
all logins and logouts. The /var/log/lastlog
file records when each user
last logged in. The /var/log/faillog
file records failed login
attempts. The /var/log/btmp
file
records the bad login attempts.
The /run/utmp
file records the
users that are currently logged in. This file is created
dynamically in the boot scripts.
When building gcc-pass2 we had to defer the installation of the C++ standard library because no suitable compiler was available to compile it. We could not use the compiler built in that section because it is a native compiler and should not be used outside of chroot and risks polluting the libraries with some host components.
Libstdc++ is part of the
GCC sources. You should first unpack the GCC tarball and
change to the gcc-11.2.0
directory.
Create a link which exists when building libstdc++ in the gcc tree:
ln -s gthr-posix.h libgcc/gthr-default.h
Create a separate build directory for libstdc++ and enter it:
mkdir -v build cd build
Prepare libstdc++ for compilation:
../libstdc++-v3/configure \ CXXFLAGS="-g -O2 -D_GNU_SOURCE" \ --prefix=/usr \ --disable-multilib \ --disable-nls \ --host=$(uname -m)-lfs-linux-gnu \ --disable-libstdcxx-pch
The meaning of the configure options:
CXXFLAGS="-g -O2
-D_GNU_SOURCE"
These flags are passed by the top level Makefile when doing a full build of GCC.
--host=$(uname
-m)-lfs-linux-gnu
We have to mimic what would happen if this package were built as part of a full compiler build. This switch would be passed to configure by GCC's build machinery.
--disable-libstdcxx-pch
This switch prevents the installation of precompiled include files, which are not needed at this stage.
Compile libstdc++ by running:
make
Install the library:
make install
Details on this package are located in Section 8.26.2, “Contents of GCC.”
The Gettext package contains utilities for internationalization and localization. These allow programs to be compiled with NLS (Native Language Support), enabling them to output messages in the user's native language.
For our temporary set of tools, we only need to install three programs from Gettext.
Prepare Gettext for compilation:
./configure --disable-shared
The meaning of the configure option:
--disable-shared
We do not need to install any of the shared Gettext libraries at this time, therefore there is no need to build them.
Compile the package:
make
Install the msgfmt, msgmerge, and xgettext programs:
cp -v gettext-tools/src/{msgfmt,msgmerge,xgettext} /usr/bin
Details on this package are located in Section 8.31.2, “Contents of Gettext.”
The Bison package contains a parser generator.
Prepare Bison for compilation:
./configure --prefix=/usr \ --docdir=/usr/share/doc/bison-3.8.2
The meaning of the new configure option:
--docdir=/usr/share/doc/bison-3.8.2
This tells the build system to install bison documentation into a versioned directory.
Compile the package:
make
Install the package:
make install
Details on this package are located in Section 8.32.2, “Contents of Bison.”
The Perl package contains the Practical Extraction and Report Language.
Prepare Perl for compilation:
sh Configure -des \ -Dprefix=/usr \ -Dvendorprefix=/usr \ -Dprivlib=/usr/lib/perl5/5.34/core_perl \ -Darchlib=/usr/lib/perl5/5.34/core_perl \ -Dsitelib=/usr/lib/perl5/5.34/site_perl \ -Dsitearch=/usr/lib/perl5/5.34/site_perl \ -Dvendorlib=/usr/lib/perl5/5.34/vendor_perl \ -Dvendorarch=/usr/lib/perl5/5.34/vendor_perl
The meaning of the new Configure options:
-des
This is a combination of three options: -d uses defaults for all items; -e ensures completion of all tasks; -s silences non-essential output.
Compile the package:
make
Install the package:
make install
Details on this package are located in Section 8.41.2, “Contents of Perl.”
The Python 3 package contains the Python development environment. It is useful for object-oriented programming, writing scripts, prototyping large programs, or developing entire applications.
There are two package files whose name starts with
“python”. The one to extract
from is Python-3.10.2.tar.xz
(notice the uppercase first letter).
Prepare Python for compilation:
./configure --prefix=/usr \ --enable-shared \ --without-ensurepip
The meaning of the configure option:
--enable-shared
This switch prevents installation of static libraries.
--without-ensurepip
This switch disables the Python package installer, which is not needed at this stage.
Compile the package:
make
Some Python 3 modules can't be built now because the dependencies are not installed yet. The building system still attempts to build them however, so the compilation of some files will fail and the compiler message may seem to indicate “fatal error”. The message should be ignored. Just make sure the toplevel make command has not failed. The optional modules are not needed now and they will be built in Chapter 8.
Install the package:
make install
Details on this package are located in Section 8.50.2, “Contents of Python 3.”
The Texinfo package contains programs for reading, writing, and converting info pages.
First, fix an issue building the package with Glibc-2.34 or later:
sed -e 's/__attribute_nonnull__/__nonnull/' \ -i gnulib/lib/malloc/dynarray-skeleton.c
Prepare Texinfo for compilation:
./configure --prefix=/usr
Compile the package:
make
Install the package:
make install
Details on this package are located in Section 8.67.2, “Contents of Texinfo.”
The Util-linux package contains miscellaneous utility programs.
The FHS recommends using the /var/lib/hwclock
directory instead of the
usual /etc
directory as the
location for the adjtime
file.
Create this directory with:
mkdir -pv /var/lib/hwclock
Prepare Util-linux for compilation:
./configure ADJTIME_PATH=/var/lib/hwclock/adjtime \ --libdir=/usr/lib \ --docdir=/usr/share/doc/util-linux-2.37.4 \ --disable-chfn-chsh \ --disable-login \ --disable-nologin \ --disable-su \ --disable-setpriv \ --disable-runuser \ --disable-pylibmount \ --disable-static \ --without-python \ runstatedir=/run
The meaning of the configure options:
ADJTIME_PATH=/var/lib/hwclock/adjtime
This sets the location of the file recording information about the hardware clock in accordance to the FHS. This is not stricly needed for this temporary tool, but it prevents creating a file at another location, which would not be overwritten or removed when building the final util-linux package.
--libdir=/usr/lib
This switch ensures the .so
symlinks targeting the shared
library file in the same directory (/usr/lib
) directly.
--disable-*
These switches prevent warnings about building components that require packages not in LFS or not installed yet.
--without-python
This switch disables using Python. It avoids trying to build unneeded bindings.
runstatedir=/run
This switch sets the location of the socket used by
uuidd and
libuuid
correctly.
Compile the package:
make
Install the package:
make install
Details on this package are located in Section 8.72.2, “Contents of Util-linux.”
First, remove the currently installed documentation to prevent them from ending up in the final system, and to save about 35 MB:
rm -rf /usr/share/{info,man,doc}/*
Second, the libtool .la files are only useful when linking with static libraries. They are unneeded and potentially harmful when using dynamic shared libraries, especially when using non-autotools build systems. While still in chroot, remove those files now:
find /usr/{lib,libexec} -name \*.la -delete
The current system size is now about 3 GB, however the /tools directory is no longer needed. It uses about 1 GB of disk space. Delete it now:
rm -rf /tools
At this point the essential programs and libraries have been created and your current LFS system is in a good state. Your system can now be backed up for later reuse. In case of fatal failures in the subsequent chapters, it often turns out that removing everything and starting over (more carefully) is the best option to recover. Unfortunately, all the temporary files will be removed, too. To avoid spending extra time to redo something which has been built successfully, creating a backup of the current LFS system may prove useful.
All the remaining steps in this section are optional. Nevertheless, as soon as you begin installing packages in Chapter 8, the temporary files will be overwritten. So it may be a good idea to do a backup of the current system as described below.
The following steps are performed from outside the chroot
environment. That means, you have to leave the chroot
environment first before continuing. The reason for that is
to get access to file system locations outside of the chroot
environment to store/read the backup archive which should not
be placed within the $LFS
hierarchy for safety reasons.
If you have decided to make a backup, leave the chroot environment:
exit
All of the following instructions are executed by
root
on your host system.
Take extra care about the commands you're going to run as
mistakes here can modify your host system. Be aware that
the environment variable LFS
is
set for user lfs
by default
but may not be set
for root
.
Whenever commands are to be executed by root
, make sure you have set
LFS
.
This has been discussed in Section 2.6, “Setting The $LFS Variable”.
Before making a backup, unmount the virtual file systems:
umount $LFS/dev/pts umount $LFS/{sys,proc,run,dev}
Make sure you have at least 1 GB free disk space (the source tarballs will be included in the backup archive) on the filesystem containing directory where you create the backup archive.
Note that the instructions below specify the home directory
of the host system's root
user, which is typically found on the root filesystem.
Replace $HOME
by a directory of
your choice if you do not want to have the backup stored in
root
's home directory.
Create the backup archive by running the following command:
Because the backup archive is compressed, it takes a relatively long time (over 10 minutes) even on a resonably fast system.
cd $LFS tar -cJpf $HOME/lfs-temp-tools-11.1.tar.xz .
If continuing to chapter 8, don't forget to reenter the chroot environment as explained in the “Important” box below.
In case some mistakes have been made and you need to start
over, you can use this backup to restore the system and save
some recovery time. Since the sources are located under
$LFS
, they are included in the
backup archive as well, so they do not need to be downloaded
again. After checking that $LFS
is
set properly, restore the backup by executing the following
commands:
The following commands are extremely dangerous. If you run
rm -rf ./* as
the root user and you do not change to the $LFS directory
or the LFS
environment variable
is not set for the root user, it will destroy your entire
host system. YOU ARE WARNED.
cd $LFS
rm -rf ./*
tar -xpf $HOME/lfs-temp-tools-11.1.tar.xz
Again, double check that the environment has been setup properly and continue building the rest of the system.
If you left the chroot environment to create a backup or restart building using a restore, remember to check that the virtual filesystems are still mounted (findmnt | grep $LFS). If they are not mounted, remount them now as described in Section 7.3, “Preparing Virtual Kernel File Systems” and re-enter the chroot environment (see Section 7.4, “Entering the Chroot Environment”) before continuing.
In this chapter, we start constructing the LFS system in earnest.
The installation of this software is straightforward. Although in many cases the installation instructions could be made shorter and more generic, we have opted to provide the full instructions for every package to minimize the possibilities for mistakes. The key to learning what makes a Linux system work is to know what each package is used for and why you (or the system) may need it.
We do not recommend using optimizations. They can make a
program run slightly faster, but they may also cause
compilation difficulties and problems when running the program.
If a package refuses to compile when using optimization, try to
compile it without optimization and see if that fixes the
problem. Even if the package does compile when using
optimization, there is the risk it may have been compiled
incorrectly because of the complex interactions between the
code and build tools. Also note that the -march
and -mtune
options using values not specified in the book have not been
tested. This may cause problems with the toolchain packages
(Binutils, GCC and Glibc). The small potential gains achieved
in using compiler optimizations are often outweighed by the
risks. First-time builders of LFS are encouraged to build
without custom optimizations. The subsequent system will still
run very fast and be stable at the same time.
Before the installation instructions, each installation page provides information about the package, including a concise description of what it contains, approximately how long it will take to build, and how much disk space is required during this building process. Following the installation instructions, there is a list of programs and libraries (along with brief descriptions) that the package installs.
The SBU values and required disk space includes test suite data for all applicable packages in Chapter 8. SBU values have been calculated using a single CPU core (-j1) for all operations.
In general, the LFS editors discourage building and installing static libraries. The original purpose for most static libraries has been made obsolete in a modern Linux system. In addition, linking a static library into a program can be detrimental. If an update to the library is needed to remove a security problem, all programs that use the static library will need to be relinked to the new library. Since the use of static libraries is not always obvious, the relevant programs (and the procedures needed to do the linking) may not even be known.
In the procedures in this chapter, we remove or disable
installation of most static libraries. Usually this is done
by passing a --disable-static
option to configure. In other cases,
alternate means are needed. In a few cases, especially glibc
and gcc, the use of static libraries remains essential to the
general package building process.
For a more complete discussion of libraries, see the discussion Libraries: Static or shared? in the BLFS book.
Package Management is an often requested addition to the LFS Book. A Package Manager allows tracking the installation of files making it easy to remove and upgrade packages. As well as the binary and library files, a package manager will handle the installation of configuration files. Before you begin to wonder, NO—this section will not talk about nor recommend any particular package manager. What it provides is a roundup of the more popular techniques and how they work. The perfect package manager for you may be among these techniques or may be a combination of two or more of these techniques. This section briefly mentions issues that may arise when upgrading packages.
Some reasons why no package manager is mentioned in LFS or BLFS include:
Dealing with package management takes the focus away from the goals of these books—teaching how a Linux system is built.
There are multiple solutions for package management, each having its strengths and drawbacks. Including one that satisfies all audiences is difficult.
There are some hints written on the topic of package management. Visit the Hints Project and see if one of them fits your need.
A Package Manager makes it easy to upgrade to newer versions when they are released. Generally the instructions in the LFS and BLFS books can be used to upgrade to the newer versions. Here are some points that you should be aware of when upgrading packages, especially on a running system.
If Linux kernel needs to be upgraded (for example, from 5.10.17 to 5.10.18 or 5.11.1), nothing else need to be rebuilt. The system will keep working fine thanks to the well-defined border between kernel and userspace. Specifically, Linux API headers need not to be (and should not be, see the next item) upgraded alongside the kernel. You'll need to reboot your system to use the upgraded kernel.
If Linux API headers or Glibc needs to be upgraded to a newer version, (e.g. from glibc-2.31 to glibc-2.32), it is safer to rebuild LFS. Though you may be able to rebuild all the packages in their dependency order, we do not recommend it.
If a package containing a shared library is updated,
and if the name of the library changes, then any
packages dynamically linked to the library need to be
recompiled in order to link against the newer library.
(Note that there is no correlation between the package
version and the name of the library.) For example,
consider a package foo-1.2.3 that installs a shared
library with name libfoo.so.1
. If you upgrade the
package to a newer version foo-1.2.4 that installs a
shared library with name libfoo.so.2
. In this case, any
packages that are dynamically linked to libfoo.so.1
need to be recompiled to
link against libfoo.so.2
in order to use the new library version. You should not
remove the previous libraries unless all the dependent
packages are recompiled.
If a package containing a shared library is updated,
and the name of library doesn't change, but the version
number of the library file decreases (for
example, the name of the library is kept named
libfoo.so.1
, but the name
of library file is changed from libfoo.so.1.25
to libfoo.so.1.24
), you should remove
the library file from the previously installed version
(libfoo.so.1.25
in the
case). Or, a ldconfig run (by
yourself using a command line, or by the installation
of some package) will reset the symlink libfoo.so.1
to point to the old
library file because it seems having a “newer” version, as its
version number is larger. This situation may happen if
you have to downgrade a package, or the package changes
the versioning scheme of library files suddenly.
If a package containing a shared library is updated,
and the name of library doesn't change, but a severe
issue (especially, a security vulnerability) is fixed,
all running programs linked to the shared library
should be restarted. The following command, run as
root
after updating,
will list what is using the old versions of those
libraries (replace libfoo
with the name of
the library):
grep -l -e 'libfoo
.*deleted' /proc/*/maps |
tr -cd 0-9\\n | xargs -r ps u
If OpenSSH is being used for accessing the system and it is linked to the updated library, you need to restart sshd service, then logout, login again, and rerun that command to confirm nothing is still using the deleted libraries.
If a binary or a shared library is overwritten, the processes using the code or data in the binary or library may crash. The correct way to update a binary or a shared library without causing the process to crash is to remove it first, then install the new version into position. The install command provided by Coreutils has already implemented this and most packages use it to install binaries and libraries. This means that you won't be troubled by this issue most of the time. However, the install process of some packages (notably Mozilla JS in BLFS) just overwrites the file if it exists and causes a crash, so it's safer to save your work and close unneeded running processes before updating a package.
The following are some common package management techniques. Before making a decision on a package manager, do some research on the various techniques, particularly the drawbacks of the particular scheme.
Yes, this is a package management technique. Some folks do not find the need for a package manager because they know the packages intimately and know what files are installed by each package. Some users also do not need any package management because they plan on rebuilding the entire system when a package is changed.
This is a simplistic package management that does not need
any extra package to manage the installations. Each package
is installed in a separate directory. For example, package
foo-1.1 is installed in /usr/pkg/foo-1.1
and a symlink is made
from /usr/pkg/foo
to
/usr/pkg/foo-1.1
. When
installing a new version foo-1.2, it is installed in
/usr/pkg/foo-1.2
and the
previous symlink is replaced by a symlink to the new
version.
Environment variables such as PATH
, LD_LIBRARY_PATH
, MANPATH
, INFOPATH
and CPPFLAGS
need to be expanded
to include /usr/pkg/foo
. For
more than a few packages, this scheme becomes unmanageable.
This is a variation of the previous package management
technique. Each package is installed similar to the
previous scheme. But instead of making the symlink, each
file is symlinked into the /usr
hierarchy. This removes the need to
expand the environment variables. Though the symlinks can
be created by the user to automate the creation, many
package managers have been written using this approach. A
few of the popular ones include Stow, Epkg, Graft, and
Depot.
The installation needs to be faked, so that the package
thinks that it is installed in /usr
though in reality it is installed in
the /usr/pkg
hierarchy.
Installing in this manner is not usually a trivial task.
For example, consider that you are installing a package
libfoo-1.1. The following instructions may not install the
package properly:
./configure --prefix=/usr/pkg/libfoo/1.1 make make install
The installation will work, but the dependent packages may
not link to libfoo as you would expect. If you compile a
package that links against libfoo, you may notice that it
is linked to /usr/pkg/libfoo/1.1/lib/libfoo.so.1
instead of /usr/lib/libfoo.so.1
as you would expect.
The correct approach is to use the DESTDIR
strategy to fake installation of the
package. This approach works as follows:
./configure --prefix=/usr make make DESTDIR=/usr/pkg/libfoo/1.1 install
Most packages support this approach, but there are some
which do not. For the non-compliant packages, you may
either need to manually install the package, or you may
find that it is easier to install some problematic packages
into /opt
.
In this technique, a file is timestamped before the installation of the package. After the installation, a simple use of the find command with the appropriate options can generate a log of all the files installed after the timestamp file was created. A package manager written with this approach is install-log.
Though this scheme has the advantage of being simple, it has two drawbacks. If, during installation, the files are installed with any timestamp other than the current time, those files will not be tracked by the package manager. Also, this scheme can only be used when one package is installed at a time. The logs are not reliable if two packages are being installed on two different consoles.
In this approach, the commands that the installation scripts perform are recorded. There are two techniques that one can use:
The LD_PRELOAD
environment
variable can be set to point to a library to be preloaded
before installation. During installation, this library
tracks the packages that are being installed by attaching
itself to various executables such as cp, install, mv and tracking the
system calls that modify the filesystem. For this approach
to work, all the executables need to be dynamically linked
without the suid or sgid bit. Preloading the library may
cause some unwanted side-effects during installation.
Therefore, it is advised that one performs some tests to
ensure that the package manager does not break anything and
logs all the appropriate files.
The second technique is to use strace, which logs all system calls made during the execution of the installation scripts.
In this scheme, the package installation is faked into a separate tree as described in the Symlink style package management. After the installation, a package archive is created using the installed files. This archive is then used to install the package either on the local machine or can even be used to install the package on other machines.
This approach is used by most of the package managers found in the commercial distributions. Examples of package managers that follow this approach are RPM (which, incidentally, is required by the Linux Standard Base Specification), pkg-utils, Debian's apt, and Gentoo's Portage system. A hint describing how to adopt this style of package management for LFS systems is located at https://www.linuxfromscratch.org/hints/downloads/files/fakeroot.txt.
Creation of package files that include dependency information is complex and is beyond the scope of LFS.
Slackware uses a tar based system for package archives. This system purposely does not handle package dependencies as more complex package managers do. For details of Slackware package management, see http://www.slackbook.org/html/package-management.html.
This scheme, unique to LFS, was devised by Matthias Benkmann, and is available from the Hints Project. In this scheme, each package is installed as a separate user into the standard locations. Files belonging to a package are easily identified by checking the user ID. The features and shortcomings of this approach are too complex to describe in this section. For the details please see the hint at https://www.linuxfromscratch.org/hints/downloads/files/more_control_and_pkg_man.txt.
One of the advantages of an LFS system is that there are no
files that depend on the position of files on a disk system.
Cloning an LFS build to another computer with the same
architecture as the base system is as simple as using
tar on the LFS
partition that contains the root directory (about 250MB
uncompressed for a base LFS build), copying that file via
network transfer or CD-ROM to the new system and expanding
it. From that point, a few configuration files will have to
be changed. Configuration files that may need to be updated
include: /etc/hosts
,
/etc/fstab
, /etc/passwd
, /etc/group
, /etc/shadow
,
/etc/ld.so.conf
, /etc/sysconfig/rc.site
, /etc/sysconfig/network
, and /etc/sysconfig/ifconfig.eth0
.
A custom kernel may need to be built for the new system depending on differences in system hardware and the original kernel configuration.
There have been some reports of issues when copying between similar but not identical architectures. For instance, the instruction set for an Intel system is not identical with an AMD processor and later versions of some processors may have instructions that are unavailable in earlier versions.
Finally the new system has to be made bootable via Section 10.4, “Using GRUB to Set Up the Boot Process”.
The Man-pages package contains over 2,200 man pages.
Install Man-pages by running:
make prefix=/usr install
The Iana-Etc package provides data for network services and protocols.
For this package, we only need to copy the files into place:
cp services protocols /etc
The Glibc package contains the main C library. This library provides the basic routines for allocating memory, searching directories, opening and closing files, reading and writing files, string handling, pattern matching, arithmetic, and so on.
Some of the Glibc programs use the non-FHS compliant
/var/db
directory to store
their runtime data. Apply the following patch to make such
programs store their runtime data in the FHS-compliant
locations:
patch -Np1 -i ../glibc-2.35-fhs-1.patch
The Glibc documentation recommends building Glibc in a dedicated build directory:
mkdir -v build cd build
Ensure that the ldconfig and sln utilites will be
installed into /usr/sbin
:
echo "rootsbindir=/usr/sbin" > configparms
Prepare Glibc for compilation:
../configure --prefix=/usr \ --disable-werror \ --enable-kernel=3.2 \ --enable-stack-protector=strong \ --with-headers=/usr/include \ libc_cv_slibdir=/usr/lib
The meaning of the configure options:
--disable-werror
This option disables the -Werror option passed to GCC. This is necessary for running the test suite.
--enable-kernel=3.2
This option tells the build system that this glibc may be used with kernels as old as 3.2. This means generating workarounds in case a system call introduced in a later version cannot be used.
--enable-stack-protector=strong
This option increases system security by adding extra code to check for buffer overflows, such as stack smashing attacks.
--with-headers=/usr/include
This option tells the build system where to find the kernel API headers.
libc_cv_slibdir=/usr/lib
This variable sets the correct library for all systems. We do not want lib64 to be used.
Compile the package:
make
In this section, the test suite for Glibc is considered critical. Do not skip it under any circumstance.
Generally a few tests do not pass. The test failures listed below are usually safe to ignore.
make check
You may see some test failures. The Glibc test suite is somewhat dependent on the host system. A few failures out of over 4200 tests can generally be ignored. This is a list of the most common issues seen for recent versions of LFS:
io/tst-lchmod is known to fail in the LFS chroot environment.
misc/tst-ttyname is known to fail in the LFS chroot environment.
The nss/tst-nss-files-hosts-multi test is known to fail if the system has no non-loopback IP addresses.
Though it is a harmless message, the install stage of Glibc
will complain about the absence of /etc/ld.so.conf
. Prevent this warning with:
touch /etc/ld.so.conf
Fix the Makefile to skip an unneeded sanity check that fails in the LFS partial environment:
sed '/test-installation/s@$(PERL)@echo not running@' -i ../Makefile
Install the package:
make install
Fix hardcoded path to the executable loader in ldd script:
sed '/RTLDLIST=/s@/usr@@g' -i /usr/bin/ldd
Install the configuration file and runtime directory for nscd:
cp -v ../nscd/nscd.conf /etc/nscd.conf mkdir -pv /var/cache/nscd
Next, install the locales that can make the system respond in a different language. None of the locales are required, but if some of them are missing, the test suites of future packages would skip important testcases.
Individual locales can be installed using the localedef program. E.g.,
the second localedef command below
combines the /usr/share/i18n/locales/cs_CZ
charset-independent locale definition with the /usr/share/i18n/charmaps/UTF-8.gz
charmap
definition and appends the result to the /usr/lib/locale/locale-archive
file. The
following instructions will install the minimum set of
locales necessary for the optimal coverage of tests:
mkdir -pv /usr/lib/locale localedef -i POSIX -f UTF-8 C.UTF-8 2> /dev/null || true localedef -i cs_CZ -f UTF-8 cs_CZ.UTF-8 localedef -i de_DE -f ISO-8859-1 de_DE localedef -i de_DE@euro -f ISO-8859-15 de_DE@euro localedef -i de_DE -f UTF-8 de_DE.UTF-8 localedef -i el_GR -f ISO-8859-7 el_GR localedef -i en_GB -f ISO-8859-1 en_GB localedef -i en_GB -f UTF-8 en_GB.UTF-8 localedef -i en_HK -f ISO-8859-1 en_HK localedef -i en_PH -f ISO-8859-1 en_PH localedef -i en_US -f ISO-8859-1 en_US localedef -i en_US -f UTF-8 en_US.UTF-8 localedef -i es_ES -f ISO-8859-15 es_ES@euro localedef -i es_MX -f ISO-8859-1 es_MX localedef -i fa_IR -f UTF-8 fa_IR localedef -i fr_FR -f ISO-8859-1 fr_FR localedef -i fr_FR@euro -f ISO-8859-15 fr_FR@euro localedef -i fr_FR -f UTF-8 fr_FR.UTF-8 localedef -i is_IS -f ISO-8859-1 is_IS localedef -i is_IS -f UTF-8 is_IS.UTF-8 localedef -i it_IT -f ISO-8859-1 it_IT localedef -i it_IT -f ISO-8859-15 it_IT@euro localedef -i it_IT -f UTF-8 it_IT.UTF-8 localedef -i ja_JP -f EUC-JP ja_JP localedef -i ja_JP -f SHIFT_JIS ja_JP.SJIS 2> /dev/null || true localedef -i ja_JP -f UTF-8 ja_JP.UTF-8 localedef -i nl_NL@euro -f ISO-8859-15 nl_NL@euro localedef -i ru_RU -f KOI8-R ru_RU.KOI8-R localedef -i ru_RU -f UTF-8 ru_RU.UTF-8 localedef -i se_NO -f UTF-8 se_NO.UTF-8 localedef -i ta_IN -f UTF-8 ta_IN.UTF-8 localedef -i tr_TR -f UTF-8 tr_TR.UTF-8 localedef -i zh_CN -f GB18030 zh_CN.GB18030 localedef -i zh_HK -f BIG5-HKSCS zh_HK.BIG5-HKSCS localedef -i zh_TW -f UTF-8 zh_TW.UTF-8
In addition, install the locale for your own country, language and character set.
Alternatively, install all locales listed in the glibc-2.35/localedata/SUPPORTED
file (it
includes every locale listed above and many more) at once
with the following time-consuming command:
make localedata/install-locales
Then use the localedef command to create
and install locales not listed in the glibc-2.35/localedata/SUPPORTED
file when
you need them. For instance, the following two locales are
needed for some tests later in this chapter:
localedef -i POSIX -f UTF-8 C.UTF-8 2> /dev/null || true localedef -i ja_JP -f SHIFT_JIS ja_JP.SJIS 2> /dev/null || true
Glibc now uses libidn2 when resolving internationalized domain names. This is a run time dependency. If this capability is needed, the instructions for installing libidn2 are in the BLFS libidn2 page.
The /etc/nsswitch.conf
file
needs to be created because the Glibc defaults do not work
well in a networked environment.
Create a new file /etc/nsswitch.conf
by running the
following:
cat > /etc/nsswitch.conf << "EOF"
# Begin /etc/nsswitch.conf
passwd: files
group: files
shadow: files
hosts: files dns
networks: files
protocols: files
services: files
ethers: files
rpc: files
# End /etc/nsswitch.conf
EOF
Install and set up the time zone data with the following:
tar -xf ../../tzdata2021e.tar.gz ZONEINFO=/usr/share/zoneinfo mkdir -pv $ZONEINFO/{posix,right} for tz in etcetera southamerica northamerica europe africa antarctica \ asia australasia backward; do zic -L /dev/null -d $ZONEINFO ${tz} zic -L /dev/null -d $ZONEINFO/posix ${tz} zic -L leapseconds -d $ZONEINFO/right ${tz} done cp -v zone.tab zone1970.tab iso3166.tab $ZONEINFO zic -d $ZONEINFO -p America/New_York unset ZONEINFO
The meaning of the zic commands:
zic -L
/dev/null ...
This creates posix time zones without any leap
seconds. It is conventional to put these in both
zoneinfo
and
zoneinfo/posix
. It is
necessary to put the POSIX time zones in zoneinfo
, otherwise various
test-suites will report errors. On an embedded
system, where space is tight and you do not intend to
ever update the time zones, you could save 1.9 MB by
not using the posix
directory, but some applications or test-suites might
produce some failures.
zic -L
leapseconds ...
This creates right time zones, including leap
seconds. On an embedded system, where space is tight
and you do not intend to ever update the time zones,
or care about the correct time, you could save 1.9MB
by omitting the right
directory.
zic ...
-p ...
This creates the posixrules
file. We use New York
because POSIX requires the daylight savings time
rules to be in accordance with US rules.
One way to determine the local time zone is to run the following script:
tzselect
After answering a few questions about the location, the
script will output the name of the time zone (e.g.,
America/Edmonton).
There are also some other possible time zones listed in
/usr/share/zoneinfo
such as
Canada/Eastern or
EST5EDT that are not
identified by the script but can be used.
Then create the /etc/localtime
file by running:
ln -sfv /usr/share/zoneinfo/<xxx>
/etc/localtime
Replace <xxx>
with the name
of the time zone selected (e.g., Canada/Eastern).
By default, the dynamic loader (/lib/ld-linux.so.2
) searches through
/lib
and /usr/lib
for dynamic libraries that are
needed by programs as they are run. However, if there are
libraries in directories other than /lib
and /usr/lib
, these need to be added to the
/etc/ld.so.conf
file in order
for the dynamic loader to find them. Two directories that
are commonly known to contain additional libraries are
/usr/local/lib
and
/opt/lib
, so add those
directories to the dynamic loader's search path.
Create a new file /etc/ld.so.conf
by running the following:
cat > /etc/ld.so.conf << "EOF"
# Begin /etc/ld.so.conf
/usr/local/lib
/opt/lib
EOF
If desired, the dynamic loader can also search a directory and include the contents of files found there. Generally the files in this include directory are one line specifying the desired library path. To add this capability run the following commands:
cat >> /etc/ld.so.conf << "EOF"
# Add an include directory
include /etc/ld.so.conf.d/*.conf
EOF
mkdir -pv /etc/ld.so.conf.d
Generates message catalogues |
|
Displays the system configuration values for file system specific variables |
|
Gets entries from an administrative database |
|
Performs character set conversion |
|
Creates fastloading iconv module configuration files |
|
Configures the dynamic linker runtime bindings |
|
Reports which shared libraries are required by each given program or shared library |
|
Assists ldd with object files. It does not exist on newer architectures like x86_64 |
|
Prints various information about the current locale |
|
Compiles locale specifications |
|
Creates a simple database from textual input |
|
Reads and interprets a memory trace file and displays a summary in human-readable format |
|
A daemon that provides a cache for the most common name service requests |
|
Dump information generated by PC profiling |
|
Lists dynamic shared objects used by running processes |
|
A statically linked ln program |
|
Traces shared library procedure calls of a specified command |
|
Reads and displays shared object profiling data |
|
Asks the user about the location of the system and reports the corresponding time zone description |
|
Traces the execution of a program by printing the currently executed function |
|
The time zone dumper |
|
The time zone compiler |
|
The helper program for shared library executables |
|
Used internally by Glibc as a gross hack to get
broken programs (e.g., some Motif applications)
running. See comments in |
|
An asynchronous name lookup library |
|
The main C library |
|
Turns on memory allocation checking when preloaded |
|
The cryptography library |
|
Dummy library containing no functions. Previously
was the dynamic linking interface library, whose
functions are now in |
|
Dummy library containing no functions. Previously was a runtime library for g++ |
|
The mathematical library |
|
The vector math library, linked in as needed when
|
|
Turns on memory allocation checking when linked to |
|
Used by memusage to help collect information about the memory usage of a program |
|
The network services library, now deprecated |
|
The Name Service Switch modules, containing
functions for resolving host names, user names,
group names, aliases, services, protocols, etc.
Loaded by |
|
Can be preloaded to PC profile an executable |
|
Dummy library containing no functions. Previously
contained functions providing most of the
interfaces specified by the POSIX.1b Realtime
Extension, now the functions are in |
|
Contains functions for creating, sending, and interpreting packets to the Internet domain name servers |
|
Contains functions providing most of the interfaces specified by the POSIX.1b Realtime Extension |
|
Contains functions useful for building debuggers for multi-threaded programs |
|
Dummy library containing no functions. Previously
contained code for “standard” functions
used in many different Unix utilities. These
functions are now in |
The Zlib package contains compression and decompression routines used by some programs.
Prepare Zlib for compilation:
./configure --prefix=/usr
Compile the package:
make
To test the results, issue:
make check
Install the package:
make install
Remove a useless static library:
rm -fv /usr/lib/libz.a
The Bzip2 package contains programs for compressing and decompressing files. Compressing text files with bzip2 yields a much better compression percentage than with the traditional gzip.
Apply a patch that will install the documentation for this package:
patch -Np1 -i ../bzip2-1.0.8-install_docs-1.patch
The following command ensures installation of symbolic links are relative:
sed -i 's@\(ln -s -f \)$(PREFIX)/bin/@\1@' Makefile
Ensure the man pages are installed into the correct location:
sed -i "s@(PREFIX)/man@(PREFIX)/share/man@g" Makefile
Prepare Bzip2 for compilation with:
make -f Makefile-libbz2_so make clean
The meaning of the make parameter:
-f
Makefile-libbz2_so
This will cause Bzip2 to be built using a different
Makefile
file, in this
case the Makefile-libbz2_so
file, which
creates a dynamic libbz2.so
library and links the Bzip2
utilities against it.
Compile and test the package:
make
Install the programs:
make PREFIX=/usr install
Install the shared library:
cp -av libbz2.so.* /usr/lib ln -sv libbz2.so.1.0.8 /usr/lib/libbz2.so
Install the shared bzip2 binary into the
/usr/bin
directory, and replace
two copies of bzip2 with symlinks:
cp -v bzip2-shared /usr/bin/bzip2 for i in /usr/bin/{bzcat,bunzip2}; do ln -sfv bzip2 $i done
Remove a useless static library:
rm -fv /usr/lib/libbz2.a
Decompresses bzipped files |
|
Decompresses to standard output |
|
Runs cmp on bzipped files |
|
Runs diff on bzipped files |
|
Runs egrep on bzipped files |
|
Runs fgrep on bzipped files |
|
Runs grep on bzipped files |
|
Compresses files using the Burrows-Wheeler block sorting text compression algorithm with Huffman coding; the compression rate is better than that achieved by more conventional compressors using “Lempel-Ziv” algorithms, like gzip |
|
Tries to recover data from damaged bzipped files |
|
Runs less on bzipped files |
|
Runs more on bzipped files |
|
The library implementing lossless, block-sorting data compression, using the Burrows-Wheeler algorithm |
The Xz package contains programs for compressing and decompressing files. It provides capabilities for the lzma and the newer xz compression formats. Compressing text files with xz yields a better compression percentage than with the traditional gzip or bzip2 commands.
Prepare Xz for compilation with:
./configure --prefix=/usr \ --disable-static \ --docdir=/usr/share/doc/xz-5.2.5
Compile the package:
make
To test the results, issue:
make check
Install the package:
make install
Decompresses to standard output |
|
Runs cmp on LZMA compressed files |
|
Runs diff on LZMA compressed files |
|
Runs egrep on LZMA compressed files |
|
Runs fgrep on LZMA compressed files |
|
Runs grep on LZMA compressed files |
|
Runs less on LZMA compressed files |
|
Compresses or decompresses files using the LZMA format |
|
A small and fast decoder for LZMA compressed files |
|
Shows information stored in the LZMA compressed file header |
|
Runs more on LZMA compressed files |
|
Decompresses files using the LZMA format |
|
Decompresses files using the XZ format |
|
Compresses or decompresses files using the XZ format |
|
Decompresses to standard output |
|
Runs cmp on XZ compressed files |
|
A small and fast decoder for XZ compressed files |
|
Runs diff on XZ compressed files |
|
Runs egrep on XZ compressed files |
|
Runs fgrep on XZ compressed files |
|
Runs grep on XZ compressed files |
|
Runs less on XZ compressed files |
|
Runs more on XZ compressed files |
|
The library implementing lossless, block-sorting data compression, using the Lempel-Ziv-Markov chain algorithm |
Zstandard is a real-time compression algorithm, providing high compression ratios. It offers a very wide range of compression / speed trade-offs, while being backed by a very fast decoder.
Compile the package:
make
In the test output there are several places that indicate 'failed'. These are expected and only 'FAIL' is an actual test failure. There should be no test failures.
To test the results, issue:
make check
Install the package:
make prefix=/usr install
Remove the static library:
rm -v /usr/lib/libzstd.a
The File package contains a utility for determining the type of a given file or files.
Prepare File for compilation:
./configure --prefix=/usr
Compile the package:
make
To test the results, issue:
make check
Install the package:
make install
The Readline package is a set of libraries that offers command-line editing and history capabilities.
Reinstalling Readline will cause the old libraries to be moved to <libraryname>.old. While this is normally not a problem, in some cases it can trigger a linking bug in ldconfig. This can be avoided by issuing the following two seds:
sed -i '/MV.*old/d' Makefile.in sed -i '/{OLDSUFF}/c:' support/shlib-install
Prepare Readline for compilation:
./configure --prefix=/usr \ --disable-static \ --with-curses \ --docdir=/usr/share/doc/readline-8.1.2
The meaning of the configure option:
--with-curses
This option tells Readline that it can find the termcap
library functions in the curses library, rather than a
separate termcap library. It allows generating a
correct readline.pc
file.
Compile the package:
make SHLIB_LIBS="-lncursesw"
The meaning of the make option:
SHLIB_LIBS="-lncursesw"
This option forces Readline to link against the
libncursesw
library.
This package does not come with a test suite.
Install the package:
make SHLIB_LIBS="-lncursesw" install
If desired, install the documentation:
install -v -m644 doc/*.{ps,pdf,html,dvi} /usr/share/doc/readline-8.1.2
The M4 package contains a macro processor.
Prepare M4 for compilation:
./configure --prefix=/usr
Compile the package:
make
To test the results, issue:
make check
Install the package:
make install
Copies the given files while expanding the macros that they contain. These macros are either built-in or user-defined and can take any number of arguments. Besides performing macro expansion, m4 has built-in functions for including named files, running Unix commands, performing integer arithmetic, manipulating text, recursion, etc. The m4 program can be used either as a front-end to a compiler or as a macro processor in its own right |
The Bc package contains an arbitrary precision numeric processing language.
Prepare Bc for compilation:
CC=gcc ./configure --prefix=/usr -G -O3
The meaning of the configure options:
CC=gcc
This parameter specifies the compiler to use.
-O3
Specify the optimization to use.
-G
Omit parts of the test suite that won't work without a GNU bc present.
Compile the package:
make
To test bc, run:
make test
Install the package:
make install
The Flex package contains a utility for generating programs that recognize patterns in text.
Prepare Flex for compilation:
./configure --prefix=/usr \ --docdir=/usr/share/doc/flex-2.6.4 \ --disable-static
Compile the package:
make
To test the results (about 0.5 SBU), issue:
make check
Install the package:
make install
A few programs do not know about flex yet and try to run its
predecessor, lex. To support those
programs, create a symbolic link named lex
that runs flex
in lex emulation mode:
ln -sv flex /usr/bin/lex
A tool for generating programs that recognize patterns in text; it allows for the versatility to specify the rules for pattern-finding, eradicating the need to develop a specialized program |
|
An extension of flex, is used for generating C++ code and classes. It is a symbolic link to flex |
|
A symbolic link that runs flex in lex emulation mode |
|
The |
The Tcl package contains the Tool Command Language, a robust general-purpose scripting language. The Expect package is written in the Tcl language.
This package and the next two (Expect and DejaGNU) are installed to support running the test suites for binutils and GCC and other packages. Installing three packages for testing purposes may seem excessive, but it is very reassuring, if not essential, to know that the most important tools are working properly.
First, unpack the documentation by issuing the following command:
tar -xf ../tcl8.6.12-html.tar.gz --strip-components=1
Prepare Tcl for compilation:
SRCDIR=$(pwd) cd unix ./configure --prefix=/usr \ --mandir=/usr/share/man \ $([ "$(uname -m)" = x86_64 ] && echo --enable-64bit)
The meaning of the configure options:
$([
"$(uname -m)" = x86_64 ] && echo
--enable-64bit)
The construct $(<shell
command>)
is replaced by the output of
the shell command. Here this output is empty if running
on a 32 bit machine, and is --enable-64bit
if running
on a 64 bit machine.
Build the package:
make sed -e "s|$SRCDIR/unix|/usr/lib|" \ -e "s|$SRCDIR|/usr/include|" \ -i tclConfig.sh sed -e "s|$SRCDIR/unix/pkgs/tdbc1.1.3|/usr/lib/tdbc1.1.3|" \ -e "s|$SRCDIR/pkgs/tdbc1.1.3/generic|/usr/include|" \ -e "s|$SRCDIR/pkgs/tdbc1.1.3/library|/usr/lib/tcl8.6|" \ -e "s|$SRCDIR/pkgs/tdbc1.1.3|/usr/include|" \ -i pkgs/tdbc1.1.3/tdbcConfig.sh sed -e "s|$SRCDIR/unix/pkgs/itcl4.2.2|/usr/lib/itcl4.2.2|" \ -e "s|$SRCDIR/pkgs/itcl4.2.2/generic|/usr/include|" \ -e "s|$SRCDIR/pkgs/itcl4.2.2|/usr/include|" \ -i pkgs/itcl4.2.2/itclConfig.sh unset SRCDIR
The various “sed” instructions after the “make” command removes references to the build directory from the configuration files and replaces them with the install directory. This is not mandatory for the remainder of LFS, but may be needed in case a package built later uses Tcl.
To test the results, issue:
make test
Install the package:
make install
Make the installed library writable so debugging symbols can be removed later:
chmod -v u+w /usr/lib/libtcl8.6.so
Install Tcl's headers. The next package, Expect, requires them.
make install-private-headers
Now make a necessary symbolic link:
ln -sfv tclsh8.6 /usr/bin/tclsh
Rename a man page that conflicts with a Perl man page:
mv /usr/share/man/man3/{Thread,Tcl_Thread}.3
If you downloaded the optional documentation, install it by issuing the following commands:
mkdir -v -p /usr/share/doc/tcl-8.6.12 cp -v -r ../html/* /usr/share/doc/tcl-8.6.12
The Expect package contains tools for automating, via scripted dialogues, interactive applications such as telnet, ftp, passwd, fsck, rlogin, and tip. Expect is also useful for testing these same applications as well as easing all sorts of tasks that are prohibitively difficult with anything else. The DejaGnu framework is written in Expect.
Prepare Expect for compilation:
./configure --prefix=/usr \ --with-tcl=/usr/lib \ --enable-shared \ --mandir=/usr/share/man \ --with-tclinclude=/usr/include
The meaning of the configure options:
--with-tcl=/usr/lib
This parameter is needed to tell configure where the tclConfig.sh script is located.
--with-tclinclude=/usr/include
This explicitly tells Expect where to find Tcl's internal headers.
Build the package:
make
To test the results, issue:
make test
Install the package:
make install ln -svf expect5.45.4/libexpect5.45.4.so /usr/lib
The DejaGnu package contains a framework for running test suites on GNU tools. It is written in expect, which itself uses Tcl (Tool Command Language).
The upstream recommends building DejaGNU in a dedicated build directory:
mkdir -v build cd build
Prepare DejaGNU for compilation:
../configure --prefix=/usr makeinfo --html --no-split -o doc/dejagnu.html ../doc/dejagnu.texi makeinfo --plaintext -o doc/dejagnu.txt ../doc/dejagnu.texi
Build and install the package:
make install install -v -dm755 /usr/share/doc/dejagnu-1.6.3 install -v -m644 doc/dejagnu.{html,txt} /usr/share/doc/dejagnu-1.6.3
To test the results, issue:
make check
The Binutils package contains a linker, an assembler, and other tools for handling object files.
Verify that the PTYs are working properly inside the chroot environment by performing a simple test:
expect -c "spawn ls"
This command should output the following:
spawn ls
If, instead, the output includes the message below, then the environment is not set up for proper PTY operation. This issue needs to be resolved before running the test suites for Binutils and GCC:
The system has no more ptys.
Ask your system administrator to create more.
Upstream has made a patch to fix a problem when using binutils for building some BLFS packages with Link Time Optimization (LTO) enabled. Apply that now:
patch -Np1 -i ../binutils-2.38-lto_fix-1.patch
Now make a fix identified upstream that affects building some packages:
sed -e '/R_386_TLS_LE /i \ || (TYPE) == R_386_TLS_IE \\' \ -i ./bfd/elfxx-x86.h
The Binutils documentation recommends building Binutils in a dedicated build directory:
mkdir -v build cd build
Prepare Binutils for compilation:
../configure --prefix=/usr \ --enable-gold \ --enable-ld=default \ --enable-plugins \ --enable-shared \ --disable-werror \ --enable-64-bit-bfd \ --with-system-zlib
The meaning of the configure parameters:
--enable-gold
Build the gold linker and install it as ld.gold (along side the default linker).
--enable-ld=default
Build the original bfd linker and install it as both ld (the default linker) and ld.bfd.
--enable-plugins
Enables plugin support for the linker.
--enable-64-bit-bfd
Enables 64-bit support (on hosts with narrower word sizes). May not be needed on 64-bit systems, but does no harm.
--with-system-zlib
Use the installed zlib library rather than building the included version.
Compile the package:
make tooldir=/usr
The meaning of the make parameter:
tooldir=/usr
Normally, the tooldir (the directory where the
executables will ultimately be located) is set to
$(exec_prefix)/$(target_alias)
. For
example, x86_64 machines would expand that to
/usr/x86_64-pc-linux-gnu
.
Because this is a custom system, this target-specific
directory in /usr
is not
required. $(exec_prefix)/$(target_alias)
would
be used if the system was used to cross-compile (for
example, compiling a package on an Intel machine that
generates code that can be executed on PowerPC
machines).
The test suite for Binutils in this section is considered critical. Do not skip it under any circumstances.
Test the results:
make -k check
Install the package:
make tooldir=/usr install
Remove useless static libraries:
rm -fv /usr/lib/lib{bfd,ctf,ctf-nobfd,opcodes}.a
Translates program addresses to file names and line numbers; given an address and the name of an executable, it uses the debugging information in the executable to determine which source file and line number are associated with the address |
|
Creates, modifies, and extracts from archives |
|
An assembler that assembles the output of gcc into object files |
|
Used by the linker to de-mangle C++ and Java symbols and to keep overloaded functions from clashing |
|
The DWARF packaging utility |
|
Updates the ELF header of ELF files |
|
Displays call graph profile data |
|
A linker that combines a number of object and archive files into a single file, relocating their data and tying up symbol references |
|
A cut down version of ld that only supports the elf object file format |
|
Hard link to ld |
|
Lists the symbols occurring in a given object file |
|
Translates one type of object file into another |
|
Displays information about the given object file, with options controlling the particular information to display; the information shown is useful to programmers who are working on the compilation tools |
|
Generates an index of the contents of an archive and stores it in the archive; the index lists all of the symbols defined by archive members that are relocatable object files |
|
Displays information about ELF type binaries |
|
Lists the section sizes and the total size for the given object files |
|
Outputs, for each given file, the sequences of printable characters that are of at least the specified length (defaulting to four); for object files, it prints, by default, only the strings from the initializing and loading sections while for other types of files, it scans the entire file |
|
Discards symbols from object files |
|
The Binary File Descriptor library |
|
The Compat ANSI-C Type Format debugging support library |
|
A libctf variant which does not use libbfd functionality |
|
A library for dealing with opcodes—the “readable text” versions of instructions for the processor; it is used for building utilities like objdump |
The GMP package contains math libraries. These have useful functions for arbitrary precision arithmetic.
If you are building for 32-bit x86, but you have a CPU
which is capable of running 64-bit code and you have specified
CFLAGS
in the environment, the
configure script will attempt to configure for 64-bits and
fail. Avoid this by invoking the configure command below
with
ABI=32
./configure ...
The default settings of GMP produce libraries optimized for the host processor. If libraries suitable for processors less capable than the host's CPU are desired, generic libraries can be created by running the following:
cp -v configfsf.guess config.guess cp -v configfsf.sub config.sub
Prepare GMP for compilation:
./configure --prefix=/usr \ --enable-cxx \ --disable-static \ --docdir=/usr/share/doc/gmp-6.2.1
The meaning of the new configure options:
--enable-cxx
This parameter enables C++ support
--docdir=/usr/share/doc/gmp-6.2.1
This variable specifies the correct place for the documentation.
Compile the package and generate the HTML documentation:
make make html
The test suite for GMP in this section is considered critical. Do not skip it under any circumstances.
Test the results:
make check 2>&1 | tee gmp-check-log
The code in gmp is highly optimized for the processor where it is built. Occasionally, the code that detects the processor misidentifies the system capabilities and there will be errors in the tests or other applications using the gmp libraries with the message "Illegal instruction". In this case, gmp should be reconfigured with the option --build=x86_64-pc-linux-gnu and rebuilt.
Ensure that all 197 tests in the test suite passed. Check the results by issuing the following command:
awk '/# PASS:/{total+=$3} ; END{print total}' gmp-check-log
Install the package and its documentation:
make install make install-html
The MPFR package contains functions for multiple precision math.
Prepare MPFR for compilation:
./configure --prefix=/usr \ --disable-static \ --enable-thread-safe \ --docdir=/usr/share/doc/mpfr-4.1.0
Compile the package and generate the HTML documentation:
make make html
The test suite for MPFR in this section is considered critical. Do not skip it under any circumstances.
Test the results and ensure that all tests passed:
make check
Install the package and its documentation:
make install make install-html
The MPC package contains a library for the arithmetic of complex numbers with arbitrarily high precision and correct rounding of the result.
Prepare MPC for compilation:
./configure --prefix=/usr \ --disable-static \ --docdir=/usr/share/doc/mpc-1.2.1
Compile the package and generate the HTML documentation:
make make html
To test the results, issue:
make check
Install the package and its documentation:
make install make install-html
The attr package contains utilities to administer the extended attributes on filesystem objects.
Prepare Attr for compilation:
./configure --prefix=/usr \ --disable-static \ --sysconfdir=/etc \ --docdir=/usr/share/doc/attr-2.5.1
Compile the package:
make
The tests need to be run on a filesystem that supports extended attributes such as the ext2, ext3, or ext4 filesystems. To test the results, issue:
make check
Install the package:
make install
The Acl package contains utilities to administer Access Control Lists, which are used to define more fine-grained discretionary access rights for files and directories.
Prepare Acl for compilation:
./configure --prefix=/usr \ --disable-static \ --docdir=/usr/share/doc/acl-2.3.1
Compile the package:
make
The Acl tests need to be run on a filesystem that supports access controls after Coreutils has been built with the Acl libraries. If desired, return to this package and run make check after Coreutils has been built later in this chapter.
Install the package:
make install
The Libcap package implements the user-space interfaces to the POSIX 1003.1e capabilities available in Linux kernels. These capabilities are a partitioning of the all powerful root privilege into a set of distinct privileges.
Prevent static libraries from being installed:
sed -i '/install -m.*STA/d' libcap/Makefile
Compile the package:
make prefix=/usr lib=lib
The meaning of the make option:
lib=lib
This parameter sets the library directory to
/usr/lib
rather than
/usr/lib64
on x86_64. It
has no effect on x86.
To test the results, issue:
make test
Install the package:
make prefix=/usr lib=lib install
A shell wrapper to explore and constrain capability support |
|
Examines file capabilities |
|
Displays the capabilities on the queried process(es) |
|
Sets file capabilities |
|
Contains the library functions for manipulating POSIX 1003.1e capabilities |
|
Contains functions to support POSIX semantics for syscalls associated with the pthread library |
The Shadow package contains programs for handling passwords in a secure way.
If you would like to enforce the use of strong passwords,
refer to
https://www.linuxfromscratch.org/blfs/view/11.1/postlfs/cracklib.html
for installing CrackLib prior to building Shadow. Then add
--with-libcrack
to
the configure
command below.
Disable the installation of the groups program and its man pages, as Coreutils provides a better version. Also, prevent the installation of manual pages that were already installed in Section 8.3, “Man-pages-5.13”:
sed -i 's/groups$(EXEEXT) //' src/Makefile.in find man -name Makefile.in -exec sed -i 's/groups\.1 / /' {} \; find man -name Makefile.in -exec sed -i 's/getspnam\.3 / /' {} \; find man -name Makefile.in -exec sed -i 's/passwd\.5 / /' {} \;
Instead of using the default
crypt method, use the
more secure SHA-512
method of password encryption, which also allows passwords
longer than 8 characters. It is also necessary to change the
obsolete /var/spool/mail
location for user mailboxes that Shadow uses by default to
the /var/mail
location used
currently. And, get rid of /bin
and /sbin
from PATH
, since they are simply symlinks to their
counterpart in /usr
.
If /bin
and/or /sbin
are preferred to be left over in
PATH
for some reason, modify
PATH
in .bashrc
after LFS is built.
sed -e 's:#ENCRYPT_METHOD DES:ENCRYPT_METHOD SHA512:' \ -e 's:/var/spool/mail:/var/mail:' \ -e '/PATH=/{s@/sbin:@@;s@/bin:@@}' \ -i etc/login.defs
If you chose to build Shadow with Cracklib support, run the following:
sed -i 's:DICTPATH.*:DICTPATH\t/lib/cracklib/pw_dict:' etc/login.defs
Prepare Shadow for compilation:
touch /usr/bin/passwd ./configure --sysconfdir=/etc \ --disable-static \ --with-group-name-max-length=32
The meaning of the configure option:
The file /usr/bin/passwd
needs to exist because its location is hardcoded in
some programs, and if it does not exist, the default
location is not right.
--with-group-name-max-length=32
The maximum user name is 32 characters. Make the maximum group name the same.
Compile the package:
make
This package does not come with a test suite.
Install the package:
make exec_prefix=/usr install make -C man install-man
This package contains utilities to add, modify, and delete
users and groups; set and change their passwords; and perform
other administrative tasks. For a full explanation of what
password shadowing
means, see the doc/HOWTO
file
within the unpacked source tree. If using Shadow support,
keep in mind that programs which need to verify passwords
(display managers, FTP programs, pop3 daemons, etc.) must be
Shadow-compliant. That is, they need to be able to work with
shadowed passwords.
To enable shadowed passwords, run the following command:
pwconv
To enable shadowed group passwords, run:
grpconv
Shadow's default configuration for the useradd utility has a few
caveats that need some explanation. First, the default action
for the useradd
utility is to create the user and a group of the same name as
the user. By default the user ID (UID) and group ID (GID)
numbers will begin with 1000. This means if you don't pass
parameters to useradd, each user will be
a member of a unique group on the system. If this behavior is
undesirable, you'll need to pass one of the -g
or -N
parameter to useradd or to change the
setting of USERGROUPS_ENAB
in /etc/login.defs
. See useradd(8)
for more information.
Second, to change the default parameters, the file
/etc/default/useradd
needs to
be created and tailored to suit your particular needs. Create
it with:
mkdir -p /etc/default useradd -D --gid 999
/etc/default/useradd
Parameter Explanations
GROUP=999
This parameter sets the beginning of the group numbers
used in the /etc/group
file. The particular value 999 comes from the
--gid
parameter
above. You can modify it to anything you desire. Note
that useradd will never
reuse a UID or GID. If the number identified in this
parameter is used, it will use the next available
number. Note also that if you don't have a group with
an ID equal to this number on your system the first
time you use useradd without the
-g
parameter,
you will get a message displayed on the terminal that
says: useradd: unknown GID
999
, although the account is correctly created.
That is why we have created the group users
with this group ID in
Section 7.6,
“Creating Essential Files and
Symlinks”.
CREATE_MAIL_SPOOL=yes
This parameter causes useradd to create a
mailbox file for the newly created user. useradd will make the
group ownership of this file to the mail
group with 0660 permissions.
If you would prefer that these mailbox files are not
created by useradd, issue the
following command:
sed -i '/MAIL/s/yes/no/' /etc/default/useradd
Choose a password for user root and set it by running:
passwd root
Used to change the maximum number of days between obligatory password changes |
|
Used to change a user's full name and other information |
|
Used to update group passwords in batch mode |
|
Used to update user passwords in batch mode |
|
Used to change a user's default login shell |
|
Checks and enforces the current password expiration policy |
|
Is used to examine the log of login failures, to set a maximum number of failures before an account is blocked, or to reset the failure count |
|
Is used to list the subordinate id ranges for a user |
|
Is used to add and delete members and administrators to groups |
|
Creates a group with the given name |
|
Deletes the group with the given name |
|
Allows a user to administer his/her own group membership list without the requirement of super user privileges. |
|
Is used to modify the given group's name or GID |
|
Verifies the integrity of the group files
|
|
Creates or updates the shadow group file from the normal group file |
|
Updates |
|
Reports the most recent login of all users or of a given user |
|
Is used by the system to let users sign on |
|
Is a daemon used to enforce restrictions on log-on time and ports |
|
Is used to set the gid mapping of a user namespace |
|
Is used to change the current GID during a login session |
|
Is used to set the uid mapping of a user namespace |
|
Is used to create or update an entire series of user accounts |
|
Displays a message that an account is not available; it is designed to be used as the default shell for accounts that have been disabled |
|
Is used to change the password for a user or group account |
|
Verifies the integrity of the password files
|
|
Creates or updates the shadow password file from the normal password file |
|
Updates |
|
Executes a given command while the user's GID is set to that of the given group |
|
Runs a shell with substitute user and group IDs |
|
Creates a new user with the given name, or updates the default new-user information |
|
Deletes the given user account |
|
Is used to modify the given user's login name, User Identification (UID), shell, initial group, home directory, etc. |
|
Edits the |
|
Edits the |
|
library for process subordinate id ranges for users |
The GCC package contains the GNU compiler collection, which includes the C and C++ compilers.
At first, fix an issue breaking libasan.a
building this package with
Glibc-2.34 or later:
sed -e '/static.*SIGSTKSZ/d' \ -e 's/return kAltStackSize/return SIGSTKSZ * 4/' \ -i libsanitizer/sanitizer_common/sanitizer_posix_libcdep.cpp
If building on x86_64, change the default directory name for 64-bit libraries to “lib”:
case $(uname -m) in x86_64) sed -e '/m64=/s/lib64/lib/' \ -i.orig gcc/config/i386/t-linux64 ;; esac
The GCC documentation recommends building GCC in a dedicated build directory:
mkdir -v build cd build
Prepare GCC for compilation:
../configure --prefix=/usr \ LD=ld \ --enable-languages=c,c++ \ --disable-multilib \ --disable-bootstrap \ --with-system-zlib
Note that for other programming languages there are some prerequisites that are not yet available. See the BLFS Book GCC page for instructions on how to build all of GCC's supported languages.
The meaning of the new configure parameters:
LD=ld
This parameter makes the configure script use the ld installed by the binutils built earlier in this chapter, rather than the cross-built version which would otherwise be used.
--with-system-zlib
This switch tells GCC to link to the system installed copy of the zlib library, rather than its own internal copy.
Compile the package:
make
In this section, the test suite for GCC is considered important, but it takes a long time. First time builders are encouraged to not skip it. The time to run the tests can be reduced significantly by adding -jx to the make command below where x is the number of cores on your system.
One set of tests in the GCC test suite is known to exhaust the default stack, so increase the stack size prior to running the tests:
ulimit -s 32768
Test the results as a non-privileged user, but do not stop at errors:
chown -Rv tester . su tester -c "PATH=$PATH make -k check"
To receive a summary of the test suite results, run:
../contrib/test_summary
For only the summaries, pipe the output through
grep -A7
Summ
.
Results can be compared with those located at https://www.linuxfromscratch.org/lfs/build-logs/11.1/ and https://gcc.gnu.org/ml/gcc-testresults/.
Eight tests related to analyzer are known to fail.
One test named asan_test.C
is
known to fail.
In libstdc++, one test named 49745.cc
is known to fail because the
header dependencies in glibc have changed.
In libstdc++, one numpunct test and six tests related to get_time are known to fail. These are all because the locale definitions in glibc have changed but libstdc++ does not currently support those changes.
A few unexpected failures cannot always be avoided. The GCC developers are usually aware of these issues, but have not resolved them yet. Unless the test results are vastly different from those at the above URL, it is safe to continue.
Install the package and remove an unneeded directory:
make install rm -rf /usr/lib/gcc/$(gcc -dumpmachine)/11.2.0/include-fixed/bits/
The GCC build directory is owned by tester
now and the ownership of the
installed header directory (and its content) will be
incorrect. Change the ownership to root
user and group:
chown -v -R root:root \ /usr/lib/gcc/*linux-gnu/11.2.0/include{,-fixed}
Create a symlink required by the FHS for "historical" reasons.
ln -svr /usr/bin/cpp /usr/lib
Add a compatibility symlink to enable building programs with Link Time Optimization (LTO):
ln -sfv ../../libexec/gcc/$(gcc -dumpmachine)/11.2.0/liblto_plugin.so \ /usr/lib/bfd-plugins/
Now that our final toolchain is in place, it is important to again ensure that compiling and linking will work as expected. We do this by performing some sanity checks:
echo 'int main(){}' > dummy.c cc dummy.c -v -Wl,--verbose &> dummy.log readelf -l a.out | grep ': /lib'
There should be no errors, and the output of the last command will be (allowing for platform-specific differences in the dynamic linker name):
[Requesting program interpreter: /lib64/ld-linux-x86-64.so.2]
Now make sure that we're setup to use the correct start files:
grep -o '/usr/lib.*/crt[1in].*succeeded' dummy.log
The output of the last command should be:
/usr/lib/gcc/x86_64-pc-linux-gnu/11.2.0/../../../../lib/crt1.o succeeded
/usr/lib/gcc/x86_64-pc-linux-gnu/11.2.0/../../../../lib/crti.o succeeded
/usr/lib/gcc/x86_64-pc-linux-gnu/11.2.0/../../../../lib/crtn.o succeeded
Depending on your machine architecture, the above may differ
slightly. The difference will be the name of the directory
after /usr/lib/gcc
. The
important thing to look for here is that gcc has found all three
crt*.o
files under the
/usr/lib
directory.
Verify that the compiler is searching for the correct header files:
grep -B4 '^ /usr/include' dummy.log
This command should return the following output:
#include <...> search starts here:
/usr/lib/gcc/x86_64-pc-linux-gnu/11.2.0/include
/usr/local/include
/usr/lib/gcc/x86_64-pc-linux-gnu/11.2.0/include-fixed
/usr/include
Again, the directory named after your target triplet may be different than the above, depending on your system architecture.
Next, verify that the new linker is being used with the correct search paths:
grep 'SEARCH.*/usr/lib' dummy.log |sed 's|; |\n|g'
References to paths that have components with '-linux-gnu' should be ignored, but otherwise the output of the last command should be:
SEARCH_DIR("/usr/x86_64-pc-linux-gnu/lib64")
SEARCH_DIR("/usr/local/lib64")
SEARCH_DIR("/lib64")
SEARCH_DIR("/usr/lib64")
SEARCH_DIR("/usr/x86_64-pc-linux-gnu/lib")
SEARCH_DIR("/usr/local/lib")
SEARCH_DIR("/lib")
SEARCH_DIR("/usr/lib");
A 32-bit system may see a few different directories. For example, here is the output from an i686 machine:
SEARCH_DIR("/usr/i686-pc-linux-gnu/lib32")
SEARCH_DIR("/usr/local/lib32")
SEARCH_DIR("/lib32")
SEARCH_DIR("/usr/lib32")
SEARCH_DIR("/usr/i686-pc-linux-gnu/lib")
SEARCH_DIR("/usr/local/lib")
SEARCH_DIR("/lib")
SEARCH_DIR("/usr/lib");
Next make sure that we're using the correct libc:
grep "/lib.*/libc.so.6 " dummy.log
The output of the last command should be:
attempt to open /usr/lib/libc.so.6 succeeded
Make sure GCC is using the correct dynamic linker:
grep found dummy.log
The output of the last command should be (allowing for platform-specific differences in dynamic linker name):
found ld-linux-x86-64.so.2 at /usr/lib/ld-linux-x86-64.so.2
If the output does not appear as shown above or is not received at all, then something is seriously wrong. Investigate and retrace the steps to find out where the problem is and correct it. Any issues will need to be resolved before continuing with the process.
Once everything is working correctly, clean up the test files:
rm -v dummy.c a.out dummy.log
Finally, move a misplaced file:
mkdir -pv /usr/share/gdb/auto-load/usr/lib mv -v /usr/lib/*gdb.py /usr/share/gdb/auto-load/usr/lib
The C++ compiler |
|
The C compiler |
|
The C preprocessor; it is used by the compiler to expand the #include, #define, and similar statements in the source files |
|
The C++ compiler |
|
The C compiler |
|
A wrapper around ar that adds a plugin to the command line. This program is only used to add "link time optimization" and is not useful with the default build options |
|
A wrapper around nm that adds a plugin to the command line. This program is only used to add "link time optimization" and is not useful with the default build options |
|
A wrapper around ranlib that adds a plugin to the command line. This program is only used to add "link time optimization" and is not useful with the default build options |
|
A coverage testing tool; it is used to analyze programs to determine where optimizations will have the most effect |
|
Offline gcda and gcno profile dump tool |
|
Offline gcda profile processing tool |
|
Tool for dumping object files produced by GCC with LTO enabled |
|
The Address Sanitizer runtime library |
|
GCC atomic built-in runtime library |
|
The C preprocessing library |
|
Contains run-time support for gcc |
|
This library is linked in to a program when GCC is instructed to enable profiling |
|
GNU implementation of the OpenMP API for multi-platform shared-memory parallel programming in C/C++ and Fortran |
|
The GNU transactional memory library |
|
The Leak Sanitizer runtime library |
|
GCC's LTO plugin allows binutils to process object files produced by GCC with LTO enabled |
|
GCC Quad Precision Math Library API |
|
Contains routines supporting GCC's stack-smashing protection functionality |
|
The standard C++ library |
|
ISO/IEC TS 18822:2015 Filesystem library |
|
Provides supporting routines for the C++ programming language |
|
The Thread Sanitizer runtime library |
|
The Undefined Behavior Sanitizer runtime library |
The pkg-config package contains a tool for passing the include path and/or library paths to build tools during the configure and make phases of package installations.
Prepare Pkg-config for compilation:
./configure --prefix=/usr \ --with-internal-glib \ --disable-host-tool \ --docdir=/usr/share/doc/pkg-config-0.29.2
The meaning of the new configure options:
--with-internal-glib
This will allow pkg-config to use its internal version of Glib because an external version is not available in LFS.
--disable-host-tool
This option disables the creation of an undesired hard link to the pkg-config program.
Compile the package:
make
To test the results, issue:
make check
Install the package:
make install
The Ncurses package contains libraries for terminal-independent handling of character screens.
Prepare Ncurses for compilation:
./configure --prefix=/usr \ --mandir=/usr/share/man \ --with-shared \ --without-debug \ --without-normal \ --enable-pc-files \ --enable-widec \ --with-pkg-config-libdir=/usr/lib/pkgconfig
The meaning of the new configure options:
--enable-widec
This switch causes wide-character libraries (e.g.,
libncursesw.so.6.3
) to be
built instead of normal ones (e.g., libncurses.so.6.3
). These
wide-character libraries are usable in both multibyte
and traditional 8-bit locales, while normal libraries
work properly only in 8-bit locales. Wide-character and
normal libraries are source-compatible, but not
binary-compatible.
--enable-pc-files
This switch generates and installs .pc files for pkg-config.
--without-normal
This switch disables building and installing most static libraries.
Compile the package:
make
This package has a test suite, but it can only be run after
the package has been installed. The tests reside in the
test/
directory. See the
README
file in that directory
for further details.
The installation of this package will overwrite libncursesw.so.6.3
in-place. It may crash
the shell process which is using code and data from the
library file. Install the package with DESTDIR
, and replace the library file
correctly using install command. A useless
static archive which is not handled by configure is also removed:
make DESTDIR=$PWD/dest install install -vm755 dest/usr/lib/libncursesw.so.6.3 /usr/lib rm -v dest/usr/lib/{libncursesw.so.6.3,libncurses++w.a} cp -av dest/* /
Many applications still expect the linker to be able to find non-wide-character Ncurses libraries. Trick such applications into linking with wide-character libraries by means of symlinks and linker scripts:
for lib in ncurses form panel menu ; do rm -vf /usr/lib/lib${lib}.so echo "INPUT(-l${lib}w)" > /usr/lib/lib${lib}.so ln -sfv ${lib}w.pc /usr/lib/pkgconfig/${lib}.pc done
Finally, make sure that old applications that look for
-lcurses
at build time are
still buildable:
rm -vf /usr/lib/libcursesw.so echo "INPUT(-lncursesw)" > /usr/lib/libcursesw.so ln -sfv libncurses.so /usr/lib/libcurses.so
If desired, install the Ncurses documentation:
mkdir -pv /usr/share/doc/ncurses-6.3 cp -v -R doc/* /usr/share/doc/ncurses-6.3
The instructions above don't create non-wide-character Ncurses libraries since no package installed by compiling from sources would link against them at runtime. However, the only known binary-only applications that link against non-wide-character Ncurses libraries require version 5. If you must have such libraries because of some binary-only application or to be compliant with LSB, build the package again with the following commands:
make distclean ./configure --prefix=/usr \ --with-shared \ --without-normal \ --without-debug \ --without-cxx-binding \ --with-abi-version=5 make sources libs cp -av lib/lib*.so.5* /usr/lib
Converts a termcap description into a terminfo description |
|
Clears the screen, if possible |
|
Compares or prints out terminfo descriptions |
|
Converts a terminfo description into a termcap description |
|
Provides configuration information for ncurses |
|
Reinitializes a terminal to its default values |
|
Clears and sets tab stops on a terminal |
|
The terminfo entry-description compiler that translates a terminfo file from source format into the binary format needed for the ncurses library routines [A terminfo file contains information on the capabilities of a certain terminal.] |
|
Lists all available terminal types, giving the primary name and description for each |
|
Makes the values of terminal-dependent capabilities available to the shell; it can also be used to reset or initialize a terminal or report its long name |
|
Can be used to initialize terminals |
|
A link to |
|
Contains functions to display text in many complex ways on a terminal screen; a good example of the use of these functions is the menu displayed during the kernel's make menuconfig |
|
Contains functions to implement forms |
|
Contains functions to implement menus |
|
Contains functions to implement panels |
The Sed package contains a stream editor.
Prepare Sed for compilation:
./configure --prefix=/usr
Compile the package and generate the HTML documentation:
make make html
To test the results, issue:
chown -Rv tester . su tester -c "PATH=$PATH make check"
Install the package and its documentation:
make install install -d -m755 /usr/share/doc/sed-4.8 install -m644 doc/sed.html /usr/share/doc/sed-4.8
The Psmisc package contains programs for displaying information about running processes.
Prepare Psmisc for compilation:
./configure --prefix=/usr
Compile the package:
make
This package does not come with a test suite.
Install the package:
make install
Reports the Process IDs (PIDs) of processes that use the given files or file systems |
|
Kills processes by name; it sends a signal to all processes running any of the given commands |
|
Peek at file descriptors of a running process, given its PID |
|
Prints information about a process |
|
Reports current logs path of a process |
|
Displays running processes as a tree |
|
Same as pstree, except that it waits for confirmation before exiting |
The Gettext package contains utilities for internationalization and localization. These allow programs to be compiled with NLS (Native Language Support), enabling them to output messages in the user's native language.
Prepare Gettext for compilation:
./configure --prefix=/usr \ --disable-static \ --docdir=/usr/share/doc/gettext-0.21
Compile the package:
make
To test the results (this takes a long time, around 3 SBUs), issue:
make check
Install the package:
make install chmod -v 0755 /usr/lib/preloadable_libintl.so
Copies standard Gettext infrastructure files into a source package |
|
Substitutes environment variables in shell format strings |
|
Translates a natural language message into the user's language by looking up the translation in a message catalog |
|
Primarily serves as a shell function library for gettext |
|
Copies all standard Gettext files into the given top-level directory of a package to begin internationalizing it |
|
Filters the messages of a translation catalog according to their attributes and manipulates the attributes |
|
Concatenates and merges the given |
|
Compares two |
|
Finds the messages that are common to the given
|
|
Converts a translation catalog to a different character encoding |
|
Creates an English translation catalog |
|
Applies a command to all translations of a translation catalog |
|
Applies a filter to all translations of a translation catalog |
|
Generates a binary message catalog from a translation catalog |
|
Extracts all messages of a translation catalog that match a given pattern or belong to some given source files |
|
Creates a new |
|
Combines two raw translations into a single file |
|
Decompiles a binary message catalog into raw translation text |
|
Unifies duplicate translations in a translation catalog |
|
Displays native language translations of a textual message whose grammatical form depends on a number |
|
Recodes Serbian text from Cyrillic to Latin script |
|
Extracts the translatable message lines from the given source files to make the first translation template |
|
defines the autosprintf class, which makes C formatted output routines usable in C++ programs, for use with the <string> strings and the <iostream> streams |
|
a private library containing common routines used by the various Gettext programs; these are not intended for general use |
|
Used to write specialized programs that process
|
|
A private library containing common routines used by the various Gettext programs; these are not intended for general use |
|
Text styling library |
|
A library, intended to be used by LD_PRELOAD that
assists |
The Bison package contains a parser generator.
Prepare Bison for compilation:
./configure --prefix=/usr --docdir=/usr/share/doc/bison-3.8.2
Compile the package:
make
To test the results (about 5.5 SBU), issue:
make check
Install the package:
make install
Generates, from a series of rules, a program for analyzing the structure of text files; Bison is a replacement for Yacc (Yet Another Compiler Compiler) |
|
A wrapper for bison, meant for
programs that still call yacc instead of
bison; it calls
bison
with the |
|
The Yacc library containing implementations of
Yacc-compatible |
The Grep package contains programs for searching through the contents of files.
Prepare Grep for compilation:
./configure --prefix=/usr
Compile the package:
make
To test the results, issue:
make check
Install the package:
make install
The Bash package contains the Bourne-Again SHell.
Prepare Bash for compilation:
./configure --prefix=/usr \ --docdir=/usr/share/doc/bash-5.1.16 \ --without-bash-malloc \ --with-installed-readline
The meaning of the new configure option:
--with-installed-readline
This option tells Bash to use the readline
library that is already
installed on the system rather than using its own
readline version.
Compile the package:
make
Skip down to “Install the package” if not running the test suite.
To prepare the tests, ensure that the tester
user can write to the sources
tree:
chown -Rv tester .
The testsuite of the package is designed to be run as a
non-root user that owns the terminal connected to standard
input. To satisfy the requirement, spawn a new pseudo
terminal using Expect and
run the tests as the tester
user:
su -s /usr/bin/expect tester << EOF set timeout -1 spawn make tests expect eof lassign [wait] _ _ _ value exit $value EOF
Install the package:
make install
Run the newly compiled bash program (replacing the one that is currently being executed):
exec /usr/bin/bash --login
A widely-used command interpreter; it performs many types of expansions and substitutions on a given command line before executing it, thus making this interpreter a powerful tool |
|
A shell script to help the user compose and mail standard formatted bug reports concerning bash |
|
A symlink to the bash program; when invoked as sh, bash tries to mimic the startup behavior of historical versions of sh as closely as possible, while conforming to the POSIX standard as well |
The Libtool package contains the GNU generic library support script. It wraps the complexity of using shared libraries in a consistent, portable interface.
Prepare Libtool for compilation:
./configure --prefix=/usr
Compile the package:
make
To test the results, issue:
make check
The test time for libtool can be reduced significantly on a system with multiple cores. To do this, append TESTSUITEFLAGS=-j<N> to the line above. For instance, using -j4 can reduce the test time by over 60 percent.
Five tests are known to fail in the LFS build environment due to a circular dependency, but all tests pass if rechecked after automake is installed.
Install the package:
make install
Remove a useless static library:
rm -fv /usr/lib/libltdl.a
The GDBM package contains the GNU Database Manager. It is a library of database functions that use extensible hashing and works similar to the standard UNIX dbm. The library provides primitives for storing key/data pairs, searching and retrieving the data by its key and deleting a key along with its data.
Prepare GDBM for compilation:
./configure --prefix=/usr \ --disable-static \ --enable-libgdbm-compat
The meaning of the configure option:
--enable-libgdbm-compat
This switch enables building the libgdbm compatibility library. Some packages outside of LFS may require the older DBM routines it provides.
Compile the package:
make
To test the results, issue:
make check
Install the package:
make install
Gperf generates a perfect hash function from a key set.
Prepare Gperf for compilation:
./configure --prefix=/usr --docdir=/usr/share/doc/gperf-3.1
Compile the package:
make
The tests are known to fail if running multiple simultaneous tests (-j option greater than 1). To test the results, issue:
make -j1 check
Install the package:
make install
The Expat package contains a stream oriented C library for parsing XML.
Prepare Expat for compilation:
./configure --prefix=/usr \ --disable-static \ --docdir=/usr/share/doc/expat-2.4.6
Compile the package:
make
To test the results, issue:
make check
Install the package:
make install
If desired, install the documentation:
install -v -m644 doc/*.{html,css} /usr/share/doc/expat-2.4.6
The Inetutils package contains programs for basic networking.
Prepare Inetutils for compilation:
./configure --prefix=/usr \ --bindir=/usr/bin \ --localstatedir=/var \ --disable-logger \ --disable-whois \ --disable-rcp \ --disable-rexec \ --disable-rlogin \ --disable-rsh \ --disable-servers
The meaning of the configure options:
--disable-logger
This option prevents Inetutils from installing the logger program, which is used by scripts to pass messages to the System Log Daemon. Do not install it because Util-linux installs a more recent version.
--disable-whois
This option disables the building of the Inetutils whois client, which is out of date. Instructions for a better whois client are in the BLFS book.
--disable-r*
These parameters disable building obsolete programs that should not be used due to security issues. The functions provided by these programs can be provided by the openssh package in the BLFS book.
--disable-servers
This disables the installation of the various network servers included as part of the Inetutils package. These servers are deemed not appropriate in a basic LFS system. Some are insecure by nature and are only considered safe on trusted networks. Note that better replacements are available for many of these servers.
Compile the package:
make
To test the results, issue:
make check
Install the package:
make install
Move a program to the proper location:
mv -v /usr/{,s}bin/ifconfig
Show the system's DNS domain name |
|
Is the file transfer protocol program |
|
Reports or sets the name of the host |
|
Manages network interfaces |
|
Sends echo-request packets and reports how long the replies take |
|
A version of ping for IPv6 networks |
|
Is used to chat with another user |
|
An interface to the TELNET protocol |
|
A trivial file transfer program |
|
Traces the route your packets take from the host you are working on to another host on a network, showing all the intermediate hops (gateways) along the way |
The Less package contains a text file viewer.
Prepare Less for compilation:
./configure --prefix=/usr --sysconfdir=/etc
The meaning of the configure options:
--sysconfdir=/etc
This option tells the programs created by the package
to look in /etc
for the
configuration files.
Compile the package:
make
This package does not come with a test suite.
Install the package:
make install
The Perl package contains the Practical Extraction and Report Language.
First, apply a patch that fixes an issue highlighted by recent versions of gdbm:
patch -Np1 -i ../perl-5.34.0-upstream_fixes-1.patch
This version of Perl now builds the Compress::Raw::Zlib and Compress::Raw::BZip2 modules. By default Perl will use an internal copy of the sources for the build. Issue the following command so that Perl will use the libraries installed on the system:
export BUILD_ZLIB=False export BUILD_BZIP2=0
To have full control over the way Perl is set up, you can remove the “-des” options from the following command and hand-pick the way this package is built. Alternatively, use the command exactly as below to use the defaults that Perl auto-detects:
sh Configure -des \ -Dprefix=/usr \ -Dvendorprefix=/usr \ -Dprivlib=/usr/lib/perl5/5.34/core_perl \ -Darchlib=/usr/lib/perl5/5.34/core_perl \ -Dsitelib=/usr/lib/perl5/5.34/site_perl \ -Dsitearch=/usr/lib/perl5/5.34/site_perl \ -Dvendorlib=/usr/lib/perl5/5.34/vendor_perl \ -Dvendorarch=/usr/lib/perl5/5.34/vendor_perl \ -Dman1dir=/usr/share/man/man1 \ -Dman3dir=/usr/share/man/man3 \ -Dpager="/usr/bin/less -isR" \ -Duseshrplib \ -Dusethreads
The meaning of the configure options:
-Dvendorprefix=/usr
This ensures perl knows how to tell packages where they should install their perl modules.
-Dpager="/usr/bin/less
-isR"
This ensures that less
is used instead
of more
.
-Dman1dir=/usr/share/man/man1
-Dman3dir=/usr/share/man/man3
Since Groff is not installed yet, Configure thinks that we do not want man pages for Perl. Issuing these parameters overrides this decision.
-Duseshrplib
Build a shared libperl needed by some perl modules.
-Dusethreads
Build perl with support for threads.
-Dprivlib,-Darchlib,-Dsitelib,...
These settings define where Perl is looking for installed modules. The LFS editors chose to put them in a directory structure based on the Major.Minor version of Perl (5.34) which allows upgrading Perl to newer Patch levels (5.34.0) without the need to reinstall all of the modules again.
Compile the package:
make
To test the results (approximately 11 SBU), issue:
make test
Install the package and clean up:
make install unset BUILD_ZLIB BUILD_BZIP2
A commandline frontend to Module::CoreList |
|
Interact with the Comprehensive Perl Archive Network (CPAN) from the command line |
|
Builds a Perl extension for the Encode module from either Unicode Character Mappings or Tcl Encoding Files |
|
Guess the encoding type of one or several files |
|
Converts |
|
Converts |
|
Shell script for examining installed Perl modules, and can create a tarball from an installed module |
|
Converts data between certain input and output formats |
|
Can be used to configure the |
|
Combines some of the best features of C, sed, awk and sh into a single swiss-army language |
|
A hard link to perl |
|
Used to generate bug reports about Perl, or the modules that come with it, and mail them |
|
Displays a piece of documentation in pod format that is embedded in the Perl installation tree or in a Perl script |
|
The Perl Installation Verification Procedure; it can be used to verify that Perl and its libraries have been installed correctly |
|
Used to generate thank you messages to mail to the Perl developers |
|
A Perl version of the character encoding converter iconv |
|
A rough tool for converting Perl4 |
|
Converts files from pod format to HTML format |
|
Converts pod data to formatted *roff input |
|
Converts pod data to formatted ASCII text |
|
Prints usage messages from embedded pod docs in files |
|
Checks the syntax of pod format documentation files |
|
Displays selected sections of pod documentation |
|
Command line tool for running tests against the Test::Harness module |
|
A tar-like program written in Perl |
|
A Perl program that compares an extracted archive with an unextracted one |
|
A Perl program that applies pattern matching to the contents of files in a tar archive |
|
Prints or checks SHA checksums |
|
Is used to force verbose warning diagnostics in Perl |
|
Converts Perl XS code into C code |
|
Displays details about the internal structure of a Zip file |
The XML::Parser module is a Perl interface to James Clark's XML parser, Expat.
Prepare XML::Parser for compilation:
perl Makefile.PL
Compile the package:
make
To test the results, issue:
make test
Install the package:
make install
The Intltool is an internationalization tool used for extracting translatable strings from source files.
First fix a warning that is caused by perl-5.22 and later:
sed -i 's:\\\${:\\\$\\{:' intltool-update.in
The above regular expression looks unusual because of all the backslashes. What it does is add a backslash before the right brace character in the sequence '\${' resulting in '\$\{'.
Prepare Intltool for compilation:
./configure --prefix=/usr
Compile the package:
make
To test the results, issue:
make check
Install the package:
make install install -v -Dm644 doc/I18N-HOWTO /usr/share/doc/intltool-0.51.0/I18N-HOWTO
The Autoconf package contains programs for producing shell scripts that can automatically configure source code.
Prepare Autoconf for compilation:
./configure --prefix=/usr
Compile the package:
make
To test the results, issue:
make check
The test time for autoconf can be reduced significantly on a system with multiple cores. To do this, append TESTSUITEFLAGS=-j<N> to the line above. For instance, using -j4 can reduce the test time by over 60 percent.
Install the package:
make install
Produces shell scripts that automatically configure software source code packages to adapt to many kinds of Unix-like systems; the configuration scripts it produces are independent—running them does not require the autoconf program |
|
A tool for creating template files of C #define statements for configure to use |
|
A wrapper for the M4 macro processor |
|
Automatically runs autoconf, autoheader, aclocal, automake, gettextize, and libtoolize in the correct order to save time when changes are made to autoconf and automake template files |
|
Helps to create a |
|
Modifies a |
|
Helps when writing |
The Automake package contains programs for generating Makefiles for use with Autoconf.
Prepare Automake for compilation:
./configure --prefix=/usr --docdir=/usr/share/doc/automake-1.16.5
Compile the package:
make
Using the -j4 make option speeds up the tests, even on systems with only one processor, due to internal delays in individual tests. To test the results, issue:
make -j4 check
The test t/subobj.sh is known to fail.
Install the package:
make install
Generates |
|
A hard link to aclocal |
|
A tool for automatically generating |
|
A hard link to automake |
The OpenSSL package contains management tools and libraries relating to cryptography. These are useful for providing cryptographic functions to other packages, such as OpenSSH, email applications, and web browsers (for accessing HTTPS sites).
Prepare OpenSSL for compilation:
./config --prefix=/usr \ --openssldir=/etc/ssl \ --libdir=lib \ shared \ zlib-dynamic
Compile the package:
make
To test the results, issue:
make test
One test, 30-test_afalg.t, is known to fail on some kernel configurations (depending on inconsistent values of CONFIG_CRYPTO_USER_API* settings.) If it fails, it can safely be ignored.
Install the package:
sed -i '/INSTALL_LIBS/s/libcrypto.a libssl.a//' Makefile make MANSUFFIX=ssl install
Add the version to the documentation directory name, to be consistent with other packages:
mv -v /usr/share/doc/openssl /usr/share/doc/openssl-3.0.1
If desired, install some additional documentation:
cp -vfr doc/* /usr/share/doc/openssl-3.0.1
You should update OpenSSL when a new version which fixes
vulnerabilities is announced. The releases run in series,
with a letter for each release after the initial release
(e.g. 1.1.1, 1.1.1a, 1.1.1b, etc). Because LFS installs
only the shared libraries, there is no need to recompile
packages which link to libcrypto.so
or libssl.so
when
upgrading in the same series.
However, any running programs linked to those libraries need to be stopped and restarted. Read the related entries in Section 8.2.1, “Upgrade Issues” for details.
is a Perl script that scans all files in a directory and adds symbolic links to their hash values |
|
is a command-line tool for using the various cryptography functions of OpenSSL's crypto library from the shell. It can be used for various functions which are documented in man 1 openssl |
|
implements a wide range of cryptographic algorithms used in various Internet standards. The services provided by this library are used by the OpenSSL implementations of SSL, TLS and S/MIME, and they have also been used to implement OpenSSH, OpenPGP, and other cryptographic standards |
|
implements the Transport Layer Security (TLS v1) protocol. It provides a rich API, documentation on which can be found by running man 3 ssl |
The Kmod package contains libraries and utilities for loading kernel modules
Prepare Kmod for compilation:
./configure --prefix=/usr \ --sysconfdir=/etc \ --with-openssl \ --with-xz \ --with-zstd \ --with-zlib
The meaning of the configure options:
--with-openssl
This option enables Kmod to handle PKCS7 signatures for kernel modules.
--with-xz
, --with-zlib
, and --with-zstd
These options enable Kmod to handle compressed kernel modules.
Compile the package:
make
The test suite of this package requires raw kernel headers (not the “sanitized” kernel headers installed earlier), which are beyond the scope of LFS.
Install the package and create symlinks for compatibility with Module-Init-Tools (the package that previously handled Linux kernel modules):
make install for target in depmod insmod modinfo modprobe rmmod; do ln -sfv ../bin/kmod /usr/sbin/$target done ln -sfv kmod /usr/bin/lsmod
Creates a dependency file based on the symbols it finds in the existing set of modules; this dependency file is used by modprobe to automatically load the required modules |
|
Installs a loadable module in the running kernel |
|
Loads and unloads kernel modules |
|
Lists currently loaded modules |
|
Examines an object file associated with a kernel module and displays any information that it can glean |
|
Uses a dependency file, created by depmod, to automatically load relevant modules |
|
Unloads modules from the running kernel |
|
This library is used by other programs to load and unload kernel modules |
Libelf is a library for handling ELF (Executable and Linkable Format) files.
Libelf is part of elfutils-0.186 package. Use the elfutils-0.186.tar.bz2 as the source tarball.
Prepare Libelf for compilation:
./configure --prefix=/usr \ --disable-debuginfod \ --enable-libdebuginfod=dummy
Compile the package:
make
To test the results, issue:
make check
Install only Libelf:
make -C libelf install install -vm644 config/libelf.pc /usr/lib/pkgconfig rm /usr/lib/libelf.a
The Libffi library provides a portable, high level programming interface to various calling conventions. This allows a programmer to call any function specified by a call interface description at run time.
Similar to GMP, libffi builds with optimizations specific to the proccesor in use. If building for another system, export CFLAGS and CXXFLAGS to specify a generic build for your architecture. If this is not done, all applications that link to libffi will trigger Illegal Operation Errors.
Prepare libffi for compilation:
./configure --prefix=/usr \ --disable-static \ --with-gcc-arch=native \ --disable-exec-static-tramp
The meaning of the configure option:
--with-gcc-arch=native
Ensure GCC optimizes for the current system. If this is not specified, the system is guessed and the code generated may not be correct for some systems. If the generated code will be copied from the native system to a less capable system, use the less capable system as a parameter. For details about alternative system types, see the x86 options in the GCC manual.
--disable-exec-static-tramp
Disable static trampoline support. It's a new security feature in libffi, but some BLFS packages (notably GJS and gobject-introspection) have not been adapted for it.
Compile the package:
make
To test the results, issue:
make check
Install the package:
make install
The Python 3 package contains the Python development environment. It is useful for object-oriented programming, writing scripts, prototyping large programs, or developing entire applications.
Prepare Python for compilation:
./configure --prefix=/usr \ --enable-shared \ --with-system-expat \ --with-system-ffi \ --with-ensurepip=yes \ --enable-optimizations
The meaning of the configure options:
--with-system-expat
This switch enables linking against system version of Expat.
--with-system-ffi
This switch enables linking against system version of libffi.
--with-ensurepip=yes
This switch enables building pip and setuptools packaging programs.
--enable-optimizations
This switch enables stable, but expensive, optimizations.
Compile the package:
make
Running the tests at this point is not recommended. The tests are known to hang indefinitely in the partial LFS environment. If desired, the tests can be rerun at the end of this chapter or when Python 3 is reinstalled in BLFS. To run the tests anyway, issue make test.
Install the package:
make install
If desired, install the preformatted documentation:
install -v -dm755 /usr/share/doc/python-3.10.2/html tar --strip-components=1 \ --no-same-owner \ --no-same-permissions \ -C /usr/share/doc/python-3.10.2/html \ -xvf ../python-3.10.2-docs-html.tar.bz2
The meaning of the documentation install commands:
--no-same-owner
and --no-same-permissions
Ensure the installed files have the correct ownership and permissions. Without these options, using tar will install the package files with the upstream creator's values.