Age | Commit message (Collapse) | Author |
|
ltsp in ubuntu started using NBD+unionfs+squashfs instead of NFS, and
debian-live uses has a network boot methods using cifs in addition to a
different way of using NFS (i think it also uses unionfs and maybe
squashfs, not just a plain NFS/cifs mount)...
so splitting out the networking related code into a separate function
would move towards not having forked code for all of these different
network boot methods. at least, that's my hope.
|
|
Fixes bootfailure on bogus init bootarg too.
Plus there may be a rootfs without init, but with sh.
|
|
|
|
you can only have up to 32 fb devices going from fb0 to fb31.
|
|
|
|
just make sure we really copy into a dir and not a file.
|
|
|
|
|
|
only warn on cases where we know that busybox is needed.
|
|
* add vcs fields
* mv busybox from depends to recommneds,
as should be found in all but unusual installations
* tighten dep on latest klibc
|
|
|
|
|
|
|
|
|
|
add it as a function of it's own
|
|
assume a s/ /\// was meant ;)
|
|
push out enough pending stuff.
|
|
allow admin to set that an update-initramfs postinst call
should update any known initramfs.
|
|
|
|
don't create /etc/initramfs-tools/conf.d/resume
in chroot for debian-live and other custom debian distributions
which build their systems in a chroot.
|
|
Strange forgot about the interface almost as not seen in manpage..
|
|
* add the error messages
* fix double call to set_initlist
* skip empty dirs at start
|
|
a find might be a bit to verbose, so be happy for now with a q&d list.
|
|
a big share of bugreports concern encrypted usage,
that output is useful for /etc/fstab parsing.
|
|
a better shot and try for finding the underlying root devices
in the case of MODULES=dep and /sys walk.
dep_add_modules() should be splitted..
|
|
recover func from previous bogus change..
|
|
sleep the passed time and then call reboot.
works as arg can only be numeric.
|
|
bonus minor coding style cleanup
|
|
we need the etch version at least, so bump.
|
|
kill dangling link.
changes are imported into git.
|
|
|
|
|
|
|
|
acked by aurel32:
11:07 <aurel32> maks: I ack the fact to always add libgcc to arm,
init needs libgcc1 for _Unwind_* functions
|
|
this setting is known to be bootable,
currently we would just create an unbootable initramfs. *bad*
|
|
maybe this will reduce ration of bug reports,
as further device renaming is in the pipes. :)
|
|
apply latest patches from bts.
|
|
compatible with etch released klibc
|
|
only indirectly referenced in mkinitramfs.8
|
|
|
|
|
|
|
|
some device node will exist before they can be safely used
|
|
newer modinfo outputs firmware files corresponding to the
modules MODULE_FIRMWARE() field.
Take care to add them automaticaly.
|
|
they cause funny sleep error messages:
aka you can't sleep "bar" time ;)
take into account that subsecond sleeps are ok.
thanks david for the idea, gone for the q&d regex check.
|
|
* Change copy_exec to use the same source and
destination path if only one argument is given.
* document how copy_exec determines the target path.
This also fixes a minor bug, these two commands:
copy_exec /some/thing /thing
copy_exec /some/other/thing /thing
Would previously print (note the incorrect path):
W:copy_exec: Not copying /some/other/thing to $DESTDIR/thing/thing...
|
|
the cryptsetup scripts parse /proc/cmdline themselves
(even in the Etch version).
|
|
* add example section
* be more explicit that update-initramfs is a wrapper around mkinitramfs
and what it does
|
|
* add a first shoot that survives ide, scsi boot testing
nfs still needs work
let's see what other testers have to say
* next step is to cut on the added libs for smaller initramfs
enhances boot speed due to smaller initramfs for tuning freaks,
others might still be better off with safe MODULES=most setting.
|
|
- missing colon in closes
|