1 =======================
2 Early userspace support
3 =======================
5 Last update: 2004-12-20 tlh
8 "Early userspace" is a set of libraries and programs that provide
9 various pieces of functionality that are important enough to be
10 available while a Linux kernel is coming up, but that don't need to be
11 run inside the kernel itself.
13 It consists of several major infrastructure components:
15 - gen_init_cpio, a program that builds a cpio-format archive
16 containing a root filesystem image. This archive is compressed, and
17 the compressed image is linked into the kernel image.
18 - initramfs, a chunk of code that unpacks the compressed cpio image
19 midway through the kernel boot process.
20 - klibc, a userspace C library, currently packaged separately, that is
21 optimized for correctness and small size.
23 The cpio file format used by initramfs is the "newc" (aka "cpio -H newc")
24 format, and is documented in the file "buffer-format.txt". There are
25 two ways to add an early userspace image: specify an existing cpio
26 archive to be used as the image or have the kernel build process build
27 the image from specifications.
32 You can create a cpio archive that contains the early userspace image.
33 Your cpio archive should be specified in CONFIG_INITRAMFS_SOURCE and it
34 will be used directly. Only a single cpio file may be specified in
35 CONFIG_INITRAMFS_SOURCE and directory and file names are not allowed in
36 combination with a cpio archive.
41 The kernel build process can also build an early userspace image from
42 source parts rather than supplying a cpio archive. This method provides
43 a way to create images with root-owned files even though the image was
44 built by an unprivileged user.
46 The image is specified as one or more sources in
47 CONFIG_INITRAMFS_SOURCE. Sources can be either directories or files -
48 cpio archives are *not* allowed when building from sources.
50 A source directory will have it and all of its contents packaged. The
51 specified directory name will be mapped to '/'. When packaging a
52 directory, limited user and group ID translation can be performed.
53 INITRAMFS_ROOT_UID can be set to a user ID that needs to be mapped to
54 user root (0). INITRAMFS_ROOT_GID can be set to a group ID that needs
55 to be mapped to group root (0).
57 A source file must be directives in the format required by the
58 usr/gen_init_cpio utility (run 'usr/gen_init_cpio -h' to get the
59 file format). The directives in the file will be passed directly to
62 When a combination of directories and files are specified then the
63 initramfs image will be an aggregate of all of them. In this way a user
64 can create a 'root-image' directory and install all files into it.
65 Because device-special files cannot be created by a unprivileged user,
66 special files can be listed in a 'root-files' file. Both 'root-image'
67 and 'root-files' can be listed in CONFIG_INITRAMFS_SOURCE and a complete
68 early userspace image can be built by an unprivileged user.
70 As a technical note, when directories and files are specified, the
71 entire CONFIG_INITRAMFS_SOURCE is passed to
72 usr/gen_initramfs_list.sh. This means that CONFIG_INITRAMFS_SOURCE
73 can really be interpreted as any legal argument to
74 gen_initramfs_list.sh. If a directory is specified as an argument then
75 the contents are scanned, uid/gid translation is performed, and
76 usr/gen_init_cpio file directives are output. If a directory is
77 specified as an argument to usr/gen_initramfs_list.sh then the
78 contents of the file are simply copied to the output. All of the output
79 directives from directory scanning and file contents copying are
80 processed by usr/gen_init_cpio.
82 See also 'usr/gen_initramfs_list.sh -h'.
84 Where's this all leading?
85 =========================
87 The klibc distribution contains some of the necessary software to make
88 early userspace useful. The klibc distribution is currently
89 maintained separately from the kernel.
91 You can obtain somewhat infrequent snapshots of klibc from
92 https://www.kernel.org/pub/linux/libs/klibc/
94 For active users, you are better off using the klibc git
95 repository, at https://git.kernel.org/?p=libs/klibc/klibc.git
97 The standalone klibc distribution currently provides three components,
98 in addition to the klibc library:
100 - ipconfig, a program that configures network interfaces. It can
101 configure them statically, or use DHCP to obtain information
102 dynamically (aka "IP autoconfiguration").
103 - nfsmount, a program that can mount an NFS filesystem.
104 - kinit, the "glue" that uses ipconfig and nfsmount to replace the old
105 support for IP autoconfig, mount a filesystem over NFS, and continue
106 system boot using that filesystem as root.
108 kinit is built as a single statically linked binary to save space.
110 Eventually, several more chunks of kernel functionality will hopefully
111 move to early userspace:
113 - Almost all of init/do_mounts* (the beginning of this is already in
116 - Insert unwieldy subsystem that doesn't really need to be in kernel
119 If kinit doesn't meet your current needs and you've got bytes to burn,
120 the klibc distribution includes a small Bourne-compatible shell (ash)
121 and a number of other utilities, so you can replace kinit and build
122 custom initramfs images that meet your needs exactly.
124 For questions and help, you can sign up for the early userspace
125 mailing list at https://www.zytor.com/mailman/listinfo/klibc
130 The kernel has currently 3 ways to mount the root filesystem:
132 a) all required device and filesystem drivers compiled into the kernel, no
133 initrd. init/main.c:init() will call prepare_namespace() to mount the
134 final root filesystem, based on the root= option and optional init= to run
135 some other init binary than listed at the end of init/main.c:init().
137 b) some device and filesystem drivers built as modules and stored in an
138 initrd. The initrd must contain a binary '/linuxrc' which is supposed to
139 load these driver modules. It is also possible to mount the final root
140 filesystem via linuxrc and use the pivot_root syscall. The initrd is
141 mounted and executed via prepare_namespace().
143 c) using initramfs. The call to prepare_namespace() must be skipped.
144 This means that a binary must do all the work. Said binary can be stored
145 into initramfs either via modifying usr/gen_init_cpio.c or via the new
146 initrd format, an cpio archive. It must be called "/init". This binary
147 is responsible to do all the things prepare_namespace() would do.
149 To maintain backwards compatibility, the /init binary will only run if it
150 comes via an initramfs cpio archive. If this is not the case,
151 init/main.c:init() will run prepare_namespace() to mount the final root
152 and exec one of the predefined init binaries.
154 Bryan O'Sullivan <bos@serpentine.com>