blob: 339c6a4f220ec036a7bc681d512ef65982dcdcef [file] [log] [blame]
Rob Landley7f46a242005-11-07 01:01:09 -08001ramfs, rootfs and initramfs
2October 17, 2005
3Rob Landley <rob@landley.net>
4=============================
5
6What is ramfs?
7--------------
8
9Ramfs is a very simple filesystem that exports Linux's disk caching
10mechanisms (the page cache and dentry cache) as a dynamically resizable
Randy Dunlap18107322007-10-16 23:29:29 -070011RAM-based filesystem.
Rob Landley7f46a242005-11-07 01:01:09 -080012
13Normally all files are cached in memory by Linux. Pages of data read from
14backing store (usually the block device the filesystem is mounted on) are kept
15around in case it's needed again, but marked as clean (freeable) in case the
16Virtual Memory system needs the memory for something else. Similarly, data
17written to files is marked clean as soon as it has been written to backing
18store, but kept around for caching purposes until the VM reallocates the
19memory. A similar mechanism (the dentry cache) greatly speeds up access to
20directories.
21
22With ramfs, there is no backing store. Files written into ramfs allocate
23dentries and page cache as usual, but there's nowhere to write them to.
24This means the pages are never marked clean, so they can't be freed by the
25VM when it's looking to recycle memory.
26
27The amount of code required to implement ramfs is tiny, because all the
28work is done by the existing Linux caching infrastructure. Basically,
29you're mounting the disk cache as a filesystem. Because of this, ramfs is not
30an optional component removable via menuconfig, since there would be negligible
31space savings.
32
33ramfs and ramdisk:
34------------------
35
36The older "ram disk" mechanism created a synthetic block device out of
Randy Dunlap18107322007-10-16 23:29:29 -070037an area of RAM and used it as backing store for a filesystem. This block
Rob Landley7f46a242005-11-07 01:01:09 -080038device was of fixed size, so the filesystem mounted on it was of fixed
39size. Using a ram disk also required unnecessarily copying memory from the
40fake block device into the page cache (and copying changes back out), as well
41as creating and destroying dentries. Plus it needed a filesystem driver
42(such as ext2) to format and interpret this data.
43
44Compared to ramfs, this wastes memory (and memory bus bandwidth), creates
45unnecessary work for the CPU, and pollutes the CPU caches. (There are tricks
46to avoid this copying by playing with the page tables, but they're unpleasantly
47complicated and turn out to be about as expensive as the copying anyway.)
48More to the point, all the work ramfs is doing has to happen _anyway_,
Randy Dunlap18107322007-10-16 23:29:29 -070049since all file access goes through the page and dentry caches. The RAM
50disk is simply unnecessary; ramfs is internally much simpler.
Rob Landley7f46a242005-11-07 01:01:09 -080051
52Another reason ramdisks are semi-obsolete is that the introduction of
53loopback devices offered a more flexible and convenient way to create
54synthetic block devices, now from files instead of from chunks of memory.
55See losetup (8) for details.
56
57ramfs and tmpfs:
58----------------
59
60One downside of ramfs is you can keep writing data into it until you fill
61up all memory, and the VM can't free it because the VM thinks that files
62should get written to backing store (rather than swap space), but ramfs hasn't
63got any backing store. Because of this, only root (or a trusted user) should
64be allowed write access to a ramfs mount.
65
66A ramfs derivative called tmpfs was created to add size limits, and the ability
67to write the data to swap space. Normal users can be allowed write access to
68tmpfs mounts. See Documentation/filesystems/tmpfs.txt for more information.
69
70What is rootfs?
71---------------
72
Rob Landleye7b69052006-06-25 05:49:00 -070073Rootfs is a special instance of ramfs (or tmpfs, if that's enabled), which is
74always present in 2.6 systems. You can't unmount rootfs for approximately the
75same reason you can't kill the init process; rather than having special code
76to check for and handle an empty list, it's smaller and simpler for the kernel
77to just make sure certain lists can't become empty.
Rob Landley7f46a242005-11-07 01:01:09 -080078
Rob Landleye7b69052006-06-25 05:49:00 -070079Most systems just mount another filesystem over rootfs and ignore it. The
Rob Landley7f46a242005-11-07 01:01:09 -080080amount of space an empty instance of ramfs takes up is tiny.
81
82What is initramfs?
83------------------
84
85All 2.6 Linux kernels contain a gzipped "cpio" format archive, which is
86extracted into rootfs when the kernel boots up. After extracting, the kernel
87checks to see if rootfs contains a file "init", and if so it executes it as PID
881. If found, this init process is responsible for bringing the system the
89rest of the way up, including locating and mounting the real root device (if
90any). If rootfs does not contain an init program after the embedded cpio
91archive is extracted into it, the kernel will fall through to the older code
92to locate and mount a root partition, then exec some variant of /sbin/init
93out of that.
94
95All this differs from the old initrd in several ways:
96
Rob Landleye7b69052006-06-25 05:49:00 -070097 - The old initrd was always a separate file, while the initramfs archive is
98 linked into the linux kernel image. (The directory linux-*/usr is devoted
99 to generating this archive during the build.)
Rob Landley7f46a242005-11-07 01:01:09 -0800100
101 - The old initrd file was a gzipped filesystem image (in some file format,
Rob Landleye7b69052006-06-25 05:49:00 -0700102 such as ext2, that needed a driver built into the kernel), while the new
Rob Landley7f46a242005-11-07 01:01:09 -0800103 initramfs archive is a gzipped cpio archive (like tar only simpler,
Rob Landleye7b69052006-06-25 05:49:00 -0700104 see cpio(1) and Documentation/early-userspace/buffer-format.txt). The
105 kernel's cpio extraction code is not only extremely small, it's also
Randy Dunlap18107322007-10-16 23:29:29 -0700106 __init text and data that can be discarded during the boot process.
Rob Landley7f46a242005-11-07 01:01:09 -0800107
108 - The program run by the old initrd (which was called /initrd, not /init) did
109 some setup and then returned to the kernel, while the init program from
110 initramfs is not expected to return to the kernel. (If /init needs to hand
111 off control it can overmount / with a new root device and exec another init
112 program. See the switch_root utility, below.)
113
114 - When switching another root device, initrd would pivot_root and then
115 umount the ramdisk. But initramfs is rootfs: you can neither pivot_root
116 rootfs, nor unmount it. Instead delete everything out of rootfs to
117 free up the space (find -xdev / -exec rm '{}' ';'), overmount rootfs
118 with the new root (cd /newmount; mount --move . /; chroot .), attach
119 stdin/stdout/stderr to the new /dev/console, and exec the new init.
120
121 Since this is a remarkably persnickity process (and involves deleting
122 commands before you can run them), the klibc package introduced a helper
123 program (utils/run_init.c) to do all this for you. Most other packages
124 (such as busybox) have named this command "switch_root".
125
126Populating initramfs:
127---------------------
128
129The 2.6 kernel build process always creates a gzipped cpio format initramfs
130archive and links it into the resulting kernel binary. By default, this
Rob Landleye7b69052006-06-25 05:49:00 -0700131archive is empty (consuming 134 bytes on x86).
132
133The config option CONFIG_INITRAMFS_SOURCE (for some reason buried under
134devices->block devices in menuconfig, and living in usr/Kconfig) can be used
135to specify a source for the initramfs archive, which will automatically be
136incorporated into the resulting binary. This option can point to an existing
137gzipped cpio archive, a directory containing files to be archived, or a text
138file specification such as the following example:
Rob Landley7f46a242005-11-07 01:01:09 -0800139
140 dir /dev 755 0 0
141 nod /dev/console 644 0 0 c 5 1
142 nod /dev/loop0 644 0 0 b 7 0
143 dir /bin 755 1000 1000
144 slink /bin/sh busybox 777 0 0
145 file /bin/busybox initramfs/busybox 755 0 0
146 dir /proc 755 0 0
147 dir /sys 755 0 0
148 dir /mnt 755 0 0
149 file /init initramfs/init.sh 755 0 0
150
Rob Landley99aef422006-01-08 01:03:43 -0800151Run "usr/gen_init_cpio" (after the kernel build) to get a usage message
152documenting the above file format.
153
Rob Landleye7b69052006-06-25 05:49:00 -0700154One advantage of the configuration file is that root access is not required to
Rob Landley7f46a242005-11-07 01:01:09 -0800155set permissions or create device nodes in the new archive. (Note that those
156two example "file" entries expect to find files named "init.sh" and "busybox" in
157a directory called "initramfs", under the linux-2.6.* directory. See
158Documentation/early-userspace/README for more details.)
159
Rob Landleye7b69052006-06-25 05:49:00 -0700160The kernel does not depend on external cpio tools. If you specify a
161directory instead of a configuration file, the kernel's build infrastructure
162creates a configuration file from that directory (usr/Makefile calls
163scripts/gen_initramfs_list.sh), and proceeds to package up that directory
164using the config file (by feeding it to usr/gen_init_cpio, which is created
165from usr/gen_init_cpio.c). The kernel's build-time cpio creation code is
166entirely self-contained, and the kernel's boot-time extractor is also
167(obviously) self-contained.
168
169The one thing you might need external cpio utilities installed for is creating
170or extracting your own preprepared cpio files to feed to the kernel build
171(instead of a config file or directory).
172
173The following command line can extract a cpio image (either by the above script
174or by the kernel build) back into its component files:
Rob Landley99aef422006-01-08 01:03:43 -0800175
176 cpio -i -d -H newc -F initramfs_data.cpio --no-absolute-filenames
177
Rob Landleye7b69052006-06-25 05:49:00 -0700178The following shell script can create a prebuilt cpio archive you can
179use in place of the above config file:
180
181 #!/bin/sh
182
183 # Copyright 2006 Rob Landley <rob@landley.net> and TimeSys Corporation.
184 # Licensed under GPL version 2
185
186 if [ $# -ne 2 ]
187 then
188 echo "usage: mkinitramfs directory imagename.cpio.gz"
189 exit 1
190 fi
191
192 if [ -d "$1" ]
193 then
194 echo "creating $2 from $1"
195 (cd "$1"; find . | cpio -o -H newc | gzip) > "$2"
196 else
197 echo "First argument must be a directory"
198 exit 1
199 fi
200
201Note: The cpio man page contains some bad advice that will break your initramfs
202archive if you follow it. It says "A typical way to generate the list
203of filenames is with the find command; you should give find the -depth option
204to minimize problems with permissions on directories that are unwritable or not
205searchable." Don't do this when creating initramfs.cpio.gz images, it won't
206work. The Linux kernel cpio extractor won't create files in a directory that
207doesn't exist, so the directory entries must go before the files that go in
208those directories. The above script gets them in the right order.
209
210External initramfs images:
211--------------------------
212
213If the kernel has initrd support enabled, an external cpio.gz archive can also
214be passed into a 2.6 kernel in place of an initrd. In this case, the kernel
215will autodetect the type (initramfs, not initrd) and extract the external cpio
216archive into rootfs before trying to run /init.
217
218This has the memory efficiency advantages of initramfs (no ramdisk block
219device) but the separate packaging of initrd (which is nice if you have
220non-GPL code you'd like to run from initramfs, without conflating it with
221the GPL licensed Linux kernel binary).
222
Randy Dunlap18107322007-10-16 23:29:29 -0700223It can also be used to supplement the kernel's built-in initramfs image. The
Rob Landleye7b69052006-06-25 05:49:00 -0700224files in the external archive will overwrite any conflicting files in
225the built-in initramfs archive. Some distributors also prefer to customize
226a single kernel image with task-specific initramfs images, without recompiling.
227
Rob Landley99aef422006-01-08 01:03:43 -0800228Contents of initramfs:
229----------------------
230
Rob Landleye7b69052006-06-25 05:49:00 -0700231An initramfs archive is a complete self-contained root filesystem for Linux.
Rob Landley7f46a242005-11-07 01:01:09 -0800232If you don't already understand what shared libraries, devices, and paths
233you need to get a minimal root filesystem up and running, here are some
234references:
235http://www.tldp.org/HOWTO/Bootdisk-HOWTO/
236http://www.tldp.org/HOWTO/From-PowerUp-To-Bash-Prompt-HOWTO.html
237http://www.linuxfromscratch.org/lfs/view/stable/
238
239The "klibc" package (http://www.kernel.org/pub/linux/libs/klibc) is
240designed to be a tiny C library to statically link early userspace
241code against, along with some related utilities. It is BSD licensed.
242
243I use uClibc (http://www.uclibc.org) and busybox (http://www.busybox.net)
Rob Landley99aef422006-01-08 01:03:43 -0800244myself. These are LGPL and GPL, respectively. (A self-contained initramfs
Rob Landleye7b69052006-06-25 05:49:00 -0700245package is planned for the busybox 1.3 release.)
Rob Landley7f46a242005-11-07 01:01:09 -0800246
247In theory you could use glibc, but that's not well suited for small embedded
248uses like this. (A "hello world" program statically linked against glibc is
249over 400k. With uClibc it's 7k. Also note that glibc dlopens libnss to do
250name lookups, even when otherwise statically linked.)
251
Rob Landleye7b69052006-06-25 05:49:00 -0700252A good first step is to get initramfs to run a statically linked "hello world"
253program as init, and test it under an emulator like qemu (www.qemu.org) or
254User Mode Linux, like so:
255
256 cat > hello.c << EOF
257 #include <stdio.h>
258 #include <unistd.h>
259
260 int main(int argc, char *argv[])
261 {
262 printf("Hello world!\n");
263 sleep(999999999);
264 }
265 EOF
266 gcc -static hello2.c -o init
267 echo init | cpio -o -H newc | gzip > test.cpio.gz
268 # Testing external initramfs using the initrd loading mechanism.
269 qemu -kernel /boot/vmlinuz -initrd test.cpio.gz /dev/zero
270
271When debugging a normal root filesystem, it's nice to be able to boot with
272"init=/bin/sh". The initramfs equivalent is "rdinit=/bin/sh", and it's
273just as useful.
274
Rob Landley99aef422006-01-08 01:03:43 -0800275Why cpio rather than tar?
276-------------------------
277
278This decision was made back in December, 2001. The discussion started here:
279
280 http://www.uwsg.iu.edu/hypermail/linux/kernel/0112.2/1538.html
281
282And spawned a second thread (specifically on tar vs cpio), starting here:
283
284 http://www.uwsg.iu.edu/hypermail/linux/kernel/0112.2/1587.html
285
286The quick and dirty summary version (which is no substitute for reading
287the above threads) is:
288
2891) cpio is a standard. It's decades old (from the AT&T days), and already
290 widely used on Linux (inside RPM, Red Hat's device driver disks). Here's
291 a Linux Journal article about it from 1996:
292
293 http://www.linuxjournal.com/article/1213
294
295 It's not as popular as tar because the traditional cpio command line tools
296 require _truly_hideous_ command line arguments. But that says nothing
297 either way about the archive format, and there are alternative tools,
298 such as:
299
300 http://freshmeat.net/projects/afio/
301
3022) The cpio archive format chosen by the kernel is simpler and cleaner (and
303 thus easier to create and parse) than any of the (literally dozens of)
304 various tar archive formats. The complete initramfs archive format is
305 explained in buffer-format.txt, created in usr/gen_init_cpio.c, and
306 extracted in init/initramfs.c. All three together come to less than 26k
307 total of human-readable text.
308
3093) The GNU project standardizing on tar is approximately as relevant as
310 Windows standardizing on zip. Linux is not part of either, and is free
311 to make its own technical decisions.
312
3134) Since this is a kernel internal format, it could easily have been
314 something brand new. The kernel provides its own tools to create and
315 extract this format anyway. Using an existing standard was preferable,
316 but not essential.
317
3185) Al Viro made the decision (quote: "tar is ugly as hell and not going to be
319 supported on the kernel side"):
320
321 http://www.uwsg.iu.edu/hypermail/linux/kernel/0112.2/1540.html
322
323 explained his reasoning:
324
325 http://www.uwsg.iu.edu/hypermail/linux/kernel/0112.2/1550.html
326 http://www.uwsg.iu.edu/hypermail/linux/kernel/0112.2/1638.html
327
328 and, most importantly, designed and implemented the initramfs code.
329
Rob Landley7f46a242005-11-07 01:01:09 -0800330Future directions:
331------------------
332
Rob Landleye7b69052006-06-25 05:49:00 -0700333Today (2.6.16), initramfs is always compiled in, but not always used. The
Rob Landley7f46a242005-11-07 01:01:09 -0800334kernel falls back to legacy boot code that is reached only if initramfs does
335not contain an /init program. The fallback is legacy code, there to ensure a
336smooth transition and allowing early boot functionality to gradually move to
337"early userspace" (I.E. initramfs).
338
339The move to early userspace is necessary because finding and mounting the real
340root device is complex. Root partitions can span multiple devices (raid or
341separate journal). They can be out on the network (requiring dhcp, setting a
Randy Dunlap18107322007-10-16 23:29:29 -0700342specific MAC address, logging into a server, etc). They can live on removable
Rob Landley7f46a242005-11-07 01:01:09 -0800343media, with dynamically allocated major/minor numbers and persistent naming
344issues requiring a full udev implementation to sort out. They can be
345compressed, encrypted, copy-on-write, loopback mounted, strangely partitioned,
346and so on.
347
348This kind of complexity (which inevitably includes policy) is rightly handled
349in userspace. Both klibc and busybox/uClibc are working on simple initramfs
Rob Landleye7b69052006-06-25 05:49:00 -0700350packages to drop into a kernel build.
Rob Landley7f46a242005-11-07 01:01:09 -0800351
Rob Landleye7b69052006-06-25 05:49:00 -0700352The klibc package has now been accepted into Andrew Morton's 2.6.17-mm tree.
353The kernel's current early boot code (partition detection, etc) will probably
354be migrated into a default initramfs, automatically created and used by the
355kernel build.