1INITRD(4) Linux Programmer's Manual INITRD(4)
2
6 initrd - boot loader initialized RAM disk
7
9 The special file /dev/initrd is a read-only block device. Device
10 /dev/initrd is a RAM disk that is initialized (e.g. loaded) by the boot
11 loader before the kernel is started. The kernel then can use the block
12 device /dev/initrd's contents for a two phased system boot-up.
13 In the first boot-up phase, the kernel starts up and mounts an initial
15 root file-system from the contents of /dev/initrd (e.g. RAM disk ini‐
16 tialized by the boot loader). In the second phase, additional drivers
17 or other modules are loaded from the initial root device's contents.
18 After loading the additional modules, a new root file system (i.e. the
19 normal root file system) is mounted from a different device.
20
22 When booting up with initrd, the system boots as follows:
23 1. The boot loader loads the kernel program and /dev/initrd's con‐
25 tents into memory.
26 2. On kernel startup, the kernel uncompresses and copies the contents
28 of the device /dev/initrd onto device /dev/ram0 and then frees the
29 memory used by /dev/initrd.
30 3. The kernel then read-write mounts device /dev/ram0 as the initial
32 root file system.
33 4. If the indicated normal root file system is also the initial root
35 file-system (e.g. /dev/ram0 ) then the kernel skips to the last step
36 for the usual boot sequence.
37 5. If the executable file /linuxrc is present in the initial root
39 file-system, /linuxrc is executed with UID 0. (The file /linuxrc
40 must have executable permission. The file /linuxrc can be any valid
41 executable, including a shell script.)
42 6. If /linuxrc is not executed or when /linuxrc terminates, the nor‐
44 mal root file system is mounted. (If /linuxrc exits with any file-
45 systems mounted on the initial root file-system, then the behavior of
46 the kernel is UNSPECIFIED. See the NOTES section for the current
47 kernel behavior.)
48 7. If the normal root file has directory /initrd, device /dev/ram0 is
50 moved from / to /initrd. Otherwise if directory /initrd does not
51 exist device /dev/ram0 is unmounted. (When moved from / to /initrd,
52 /dev/ram0 is not unmounted and therefore processes can remain running
53 from /dev/ram0. If directory /initrd does not exist on the normal
54 root file-system and any processes remain running from /dev/ram0 when
55 /linuxrc exits, the behavior of the kernel is UNSPECIFIED. See the
56 NOTES section for the current kernel behavior.)
57 8. The usual boot sequence (e.g. invocation of /sbin/init) is per‐
59 formed on the normal root file system.
60
62 The following boot loader options when used with initrd, affect the
63 kernel's boot-up operation:
64 initrd=filename
66 Specifies the file to load as the contents of /dev/initrd. For
67 LOADLIN this is a command line option. For LILO you have to use
68 this command in the LILO configuration file /etc/lilo.config.
69 The filename specified with this option will typically be a
70 gzipped file-system image.
71 noinitrd
73 This boot time option disables the two phase boot-up operation.
74 The kernel performs the usual boot sequence as if /dev/initrd
75 was not initialized. With this option, any contents of
76 /dev/initrd loaded into memory by the boot loader contents are
77 preserved. This option permits the contents of /dev/initrd to
78 be any data and need not be limited to a file system image.
79 However, device /dev/initrd is read-only and can be read only
80 one time after system startup.
81 root=device-name
83 Specifies the device to be used as the normal root file system.
84 For LOADLIN this is a command line option. For LILO this is a
85 boot time option or can be used as an option line in the LILO
86 configuration file /etc/lilo.config. The device specified by
87 the this option must be a mountable device having a suitable
88 root file-system.
89
91 By default, the kernel's settings (e.g. set in the kernel file with
92 rdev(8) or compiled into the kernel file), or the boot loader option
93 setting is used for the normal root file systems. For a NFS-mounted
94 normal root file system, one has to use the nfs_root_name and
95 nfs_root_addrs boot options to give the NFS settings. For more infor‐
96 mation on NFS-mounted root see the kernel documentation file nfs‐
97 root.txt. For more information on setting the root file system also
98 see the LILO and LOADLIN documentation.
99 It is also possible for the /linuxrc executable to change the normal
101 root device. For /linuxrc to change the normal root device, /proc must
102 be mounted. After mounting /proc, /linuxrc changes the normal root
103 device by writing into the proc files /proc/sys/kernel/real-root-dev,
104 /proc/sys/kernel/nfs-root-name, and /proc/sys/kernel/nfs-root-addrs.
105 For a physical root device, the root device is changed by having /lin‐
106 uxrc write the new root file system device number into /proc/sys/ker‐
107 nel/real-root-dev. For a NFS root file system, the root device is
108 changed by having /linuxrc write the NFS setting into files
109 /proc/sys/kernel/nfs-root-name and /proc/sys/kernel/nfs-root-addrs and
110 then writing 0xff (e.g. the pseudo-NFS-device number) into file
111 /proc/sys/kernel/real-root-dev. For example, the following shell com‐
112 mand line would change the normal root device to /dev/hdb1:
113 echo 0x365 >/proc/sys/kernel/real-root-dev
114 For a NFS example, the following shell command lines would change the
115 normal root device to the NFS directory /var/nfsroot on a local net‐
116 worked NFS server with IP number 193.8.232.7 for a system with IP num‐
117 ber 193.8.232.7 and named 'idefix':
118 echo /var/nfsroot >/proc/sys/kernel/nfs-root-name
119 echo 193.8.232.2:193.8.232.7::255.255.255.0:idefix \
120 >/proc/sys/kernel/nfs-root-addrs
121 echo 255 >/proc/sys/kernel/real-root-dev
122 Note: The use of /proc/sys/kernel/real-root-dev to change the root file
124 system is obsolete. See the kernel source file Documentation/ini‐
125 trd.txt as well as pivot_root(2) and pivot_root(8) for information on
126 the modern method of changing the root file system.
127
129 The main motivation for implementing initrd was to allow for modular
130 kernel configuration at system installation.
131 A possible system installation scenario is as follows:
133 1. The loader program boots from floppy or other media with a minimal
135 kernel (e.g. support for /dev/ram, /dev/initrd, and the ext2 file-
136 system) and loads /dev/initrd with a gzipped version of the initial
137 file-system.
138 2. The executable /linuxrc determines what is needed to (1) mount the
140 normal root file-system (i.e. device type, device drivers, file sys‐
141 tem) and (2) the distribution media (e.g. CD-ROM, network, tape,
142 ...). This can be done by asking the user, by auto-probing, or by
143 using a hybrid approach.
144 3. The executable /linuxrc loads the necessary modules from the ini‐
146 tial root file-system.
147 4. The executable /linuxrc creates and populates the root file sys‐
149 tem. (At this stage the normal root file system does not have to be
150 a completed system yet.)
151 5. The executable /linuxrc sets /proc/sys/kernel/real-root-dev,
153 unmount /proc, the normal root file system and any other file systems
154 it has mounted, and then terminates.
155 6. The kernel then mounts the normal root file system.
157 7. Now that the file system is accessible and intact, the boot loader
159 can be installed.
160 8. The boot loader is configured to load into /dev/initrd a file sys‐
162 tem with the set of modules that was used to bring up the system.
163 (e.g. Device /dev/ram0 can be modified, then unmounted, and finally,
164 the image is written from /dev/ram0 to a file.)
165 9. The system is now bootable and additional installation tasks can
167 be performed.
168 The key role of /dev/initrd in the above is to re-use the configuration
170 data during normal system operation without requiring initial kernel
171 selection, a large generic kernel or, recompiling the kernel.
172 A second scenario is for installations where Linux runs on systems with
174 different hardware configurations in a single administrative network.
175 In such cases, it may be desirable to use only a small set of kernels
176 (ideally only one) and to keep the system-specific part of configura‐
177 tion information as small as possible. In this case, create a common
178 file with all needed modules. Then, only the /linuxrc file or a file
179 executed by /linuxrc would be different.
180 A third scenario is more convenient recovery disks. Because informa‐
182 tion like the location of the root file-system partition is not needed
183 at boot time, the system loaded from /dev/initrd can use a dialog
184 and/or auto-detection followed by a possible sanity check.
185 Last but not least, Linux distributions on CD-ROM may use initrd for
187 easy installation from the CD-ROM. The distribution can use LOADLIN to
188 directly load /dev/initrd from CD-ROM without the need of any floppies.
189 The distribution could also use a LILO boot floppy and then bootstrap a
190 bigger ram disk via /dev/initrd from the CD-ROM.
191
193 The /dev/initrd is a read-only block device assigned major number 1 and
194 minor number 250. Typically /dev/initrd is owned by root.disk with
195 mode 0400 (read access by root only). If the Linux system does not
196 have /dev/initrd already created, it can be created with the following
197 commands:
198 mknod -m 400 /dev/initrd b 1 250
199 chown root:disk /dev/initrd
200 Also, support for both "RAM disk" and "Initial RAM disk" (e.g. CON‐
201 FIG_BLK_DEV_RAM=y and CONFIG_BLK_DEV_INITRD=y ) support must be com‐
202 piled directly into the Linux kernel to use /dev/initrd. When using
203 /dev/initrd, the RAM disk driver cannot be loaded as a module.
204
206 /dev/initrd
207 /dev/ram0
208 /linuxrc
209 /initrd
210
212 chown(1), mknod(1), ram(4), freeramdisk(8), rdev(8)
213 The documentation file initrd.txt in the kernel source package, the
215 LILO documentation, the LOADLIN documentation, the SYSLINUX documenta‐
216 tion.
217
219 1. With the current kernel, any file systems that remain mounted when
220 /dev/ram0 is moved from / to /initrd continue to be accessible. How‐
221 ever, the /proc/mounts entries are not updated.
222 2. With the current kernel, if directory /initrd does not exist, then
224 /dev/ram0 will NOT be fully unmounted if /dev/ram0 is used by any
225 process or has any file-system mounted on it. If /dev/ram0 is NOT
226 fully unmounted, then /dev/ram0 will remain in memory.
227 3. Users of /dev/initrd should not depend on the behavior give in the
229 above notes. The behavior may change in future versions of the Linux
230 kernel.
231
233 The kernel code for device initrd was written by Werner Almesberger
234 <almesber@lrc.epfl.ch> and Hans Lermen <lermen@elserv.ffm.fgan.de>.
235 The code for initrd was added to the baseline Linux kernel in develop‐
236 ment version 1.3.73.
237Linux 2.0 1997-11-06 INITRD(4)