Age | Commit message (Collapse) | Author |
|
|
|
|
|
|
|
|
|
|
|
|
|
- While scanning for a filesystem on a disk, the klibc utility can
return a null string if it not finds a filesystem (e.g. on a container
partition), and grepping for a null string either reports true if
escaped or hangs grep if not escaped. Both conditions must be avoided
in order to report that the kernel supports the filesystem named as
the string.
|
|
|
|
|
|
- The plain list of "should be a supported filesystem" finally is fixed,
now it scans /proc to see if the kernel is able to mount it, if not it
modprobes it and as a failover tries to insmod it from the real rootfs
/lib/modules directory if it can.
- This should fix - en passant - the "drop to busybox shell if a ntfs
filesystem is found while scanning for persistence media" bug as well.
|
|
|
|
"lostup"
|
|
place.
|
|
|
|
using old commands when beeing on etch. Also handling udevtrigger with udevadm if available.
|
|
|
|
|
|
|
|
* New udev (>= 1.24-1) replaced some binaries (here udevtrigger) with
udevadm command line options, e.g. "udevtrigger" -> "udevadm trigger".
|
|
|
|
* Now it should work even if not using klibc's "read".
|
|
* Included support for specifing a list of files/dirs to be explicitly
kept between reboots, enabled only for the "cpio" snapshot type.
Look at "/usr/share/doc/live-initramfs/examples/live-snapshot.list"
for hints.
|
|
This is to retain compatibility with Etch's non-loop-aes-utils' losetup.
|
|
|
|
|
|
find_livefs.
|
|
* This should re-enable using of live-snapshot with targets differents
from static list cpio's.
|
|
|
|
|
|
Pass -r option to losetup when setting up loop device for root image
so that readonly files can be set up as well.
|
|
Thanks to ChoBolT and taytay on irc
|
|
|
|
|
|
show its messages during boot (Closes: #440238).
|
|
In Debian, this package doesn't exist as a .deb, so we must maintain our
own local copy of the languagelist file.
(This patch also adds code to synchronise this file to the Makefile's
"update:" target so this isn't too annoying.)
|
|
|
|
|
|
|
|
|
|
|
|
This patch goes some way towards fixing USB-backed persistence by
configuring the live initramfs to contain the "sd_mod" module and adding
a hacky workaround to a udev-related race condition.
For some reason, the usual udevtrigger/udevsettle pairing does not block
after we have loaded our USB modules and /dev/sda1 (for example) is ready.
Because of this, we sleep for a little while and break if a new block
device appears.
This seems to be a bug either in udev or in my understanding of how the
tool is meant to work. Some information may be gained from the casper bug:
https://bugs.launchpad.net/ubuntu/+source/casper/+bug/84591
Sleep-based hacks like this are really ugly, but I take some comfort in
finding them elsewhere in the source.
|
|
We are hardly likely to find persistent-related devices on RAM disks, so
we ignore them whilst searching.
|
|
This patch extends the find_files and find_cow_device utilities to search
known-working partition types (removing two "FIXME"s) and reworks the
syntax to use somewhat more readable case-statements.
|
|
This patch applies two fixes intended for when "scripts/live-helpers" is
sourced by a "set -e" script such as live-snapshot.
|
|
|
|
|
|
|
|
We must configure a valid hostname in the target before 10adduser, or sudo
fails with "unable to resolve host (none)" - before 18hostname is called
/root/etc/hostname is empty.
This was resulting in the "su as sudo" modifications not being applied in
the target which was causing X configuration applications to show different
behaviours when attempting elevating their priviledges.
Reordering the hostname configuration seems somewhat preferable to
applying a hack inside 10adduser to use 'su' or similar, as other
pre-XXhostname calls may incorporate calls to sudo in the future.
|
|
|
|
|