From: Olaf Hering <olh@suse.de>
initramfs can not be used in current 2.6 kernels, the files will never be
executed because prepare_namespace doesn't care about them. The only way to
workaround that limitation is a root=0:0 cmdline option to force rootfs as
root filesystem. This will break further booting because rootfs is not the
final root filesystem.
This patch checks for the presence of /sbin/init which comes from the cpio
archive (and thats the only way to store files into the rootfs). This
binary/script has to do all the work of prepare_namespace(). One can specify
a different binary from the cpio archive via the kinit= option. It does not
break current setups because initrd will not touch rootfs.
Any why would one want to do that? After some thought, only 2 reasons come
to mind: the cpio archive can be built as user (no loop mount required), and
it should simplify migration to 2.7 (I hope the stuff behind
prepare_namespace will disappear).
---
Documentation/early-userspace/README | 28 ++++++++++++++++++++++++++++
Documentation/kernel-parameters.txt | 3 +++
init/main.c | 17 +++++++++++++++++
3 files changed, 48 insertions(+)
diff -puN Documentation/early-userspace/README~initramfs-kinit_command Documentation/early-userspace/README
--- 25/Documentation/early-userspace/README~initramfs-kinit_command 2004-02-20 23:37:11.000000000 -0800
+++ 25-akpm/Documentation/early-userspace/README 2004-02-20 23:37:11.000000000 -0800
@@ -72,4 +72,32 @@ For questions and help, you can sign up
mailing list at http://www.zytor.com/mailman/listinfo/klibc
+How does it work?
+=================
+
+The kernel has currently 3 ways to mount the root filesystem:
+a) all required device and filesystem drivers compiled into the
+ kernel, no initrd. init/main.c:init() will call prepare_namespace()
+ to mount the final root filesystem, based on the root= option
+ and optional init= to run some other init binary than listed
+ at the end of init/main.c:init().
+b) some device and filesystem drivers built as modules and stored in
+ an initrd. The initrd must contain a binary '/linuxrc' which is
+ supposed to load these driver modules. It is also possible to
+ mount the final root filesystem via linuxrc and use the pivot_root
+ syscall. The initrd is mounted and executed via prepare_namespace().
+c) using initramfs. The call to prepare_namespace() must be skipped. This
+ means that a binary must do all the work. Said binary can be stored
+ into initramfs either via modifying usr/gen_init_cpio.c or via the new
+ initrd format, an cpio archive. It must be called "/sbin/init".
+ A different binary can be specified with the kinit= boot option.
+ This binary is responsible to do all the things prepare_namespace()
+ would do.
+ To remain backwards compatibility, the kinit binary will only run if
+ it comes via an initramfs cpio archive, or if the kinit= option is
+ specified. If this is not the case, init/main.c:init() will run
+ prepare_namespace() to mount the final root and exec one of the
+ predefined init binaries.
+
+
Bryan O'Sullivan <bos@serpentine.com>
diff -puN Documentation/kernel-parameters.txt~initramfs-kinit_command Documentation/kernel-parameters.txt
--- 25/Documentation/kernel-parameters.txt~initramfs-kinit_command 2004-02-20 23:37:11.000000000 -0800
+++ 25-akpm/Documentation/kernel-parameters.txt 2004-02-20 23:37:11.000000000 -0800
@@ -499,6 +499,9 @@ running once the system is up.
keepinitrd [HW,ARM]
+ kinit= Specify the path to the init executable to replace
+ prepare_namespace() (for initramfs archives)
+
l2cr= [PPC]
lapic [IA-32,APIC] Enable the local APIC even if BIOS disabled it.
diff -puN init/main.c~initramfs-kinit_command init/main.c
--- 25/init/main.c~initramfs-kinit_command 2004-02-20 23:37:11.000000000 -0800
+++ 25-akpm/init/main.c 2004-02-20 23:37:11.000000000 -0800
@@ -299,6 +299,15 @@ static int __init init_setup(char *str)
}
__setup("init=", init_setup);
+static char *kinit_command;
+
+static int __init kinit_setup(char *str)
+{
+ kinit_command = str;
+ return 1;
+}
+__setup("kinit=", kinit_setup);
+
extern void setup_arch(char **);
extern void cpu_idle(void);
@@ -534,6 +543,7 @@ static void __init do_initcalls(void)
flush_scheduled_work();
}
+asmlinkage long sys_access(const char __user * filename, int mode);
/*
* Ok, the machine is now initialized. None of the devices
* have been touched yet, but the CPU subsystem is up and
@@ -597,6 +607,13 @@ static int init(void * unused)
sched_init_smp();
do_basic_setup();
+ /*
+ * check if there is an early userspace init, if yes
+ * let it do all the work
+ */
+ if (kinit_command || sys_access("/sbin/init", 0) == 0)
+ execute_command = kinit_command ? kinit_command : 0;
+ else
prepare_namespace();
/*
_