| Age | Commit message (Collapse) | Author |
|---|
|
Ferenc seems to be using the xvda1=img1,xvda2=img2 scheme of Xen disks
(this is what xentools gets you) rather than the whole disk with
partition scheme.
In the xvda1/xvda2 scheme /sys/block contains:
# ls /sys/block/
ram0 ram1 ram10 ram11 ram12 ram13 ram14 ram15 ram2 ram3
ram4 ram5 ram6 ram7 ram8 ram9 xvda1 xvda2
In the whole disk scheme it contains xvda as you would expect:
# ls /sys/block/
loop0 loop1 loop2 loop3 loop4 loop5 loop6 loop7 ram0
ram1 ram10 ram11 ram12 ram13 ram14 ram15 ram2 ram3 ram4
ram5 ram6 ram7 ram8 ram9 xvda
I've attached a patch which takes care of this difference.
[ remove largest suffix pattern, so that we don't stumble on big minors -maks ]
Signed-off-by: maximilian attems <maks@debian.org>
|
|
do the same dance here, same code, just slightly renamed variables.
|
|
|
|
As there is no /dev/disk/by-label//, we need to follow udev
on escape the '/' character with \x2f
don't rely on sed, as we might not have it in initramfs.
(closes: #489008)
Reported-by: Josh Triplett <josh@freedesktop.org>
Tested-by: Andres Salomon <dilinger@debian.org>
thanks for dilinger to pushing the first easy solution to a complete,
so we don't have to revisit the case.
|
|
Documentation currently lacked a list of exported variables
for boot or hook scripts.
(closes: #512453)
Reported-by: Arnaud Fontaine <arnau@debian.org>
|
|
the supported-host long call feature was already dropped
before lenny, thus just kick it.
no other long calling getopt strings yet..
|
|
..via /etc/kernel
Signed-off-by: Andres Salomon <dilinger@debian.org>
(closes: #504551)
[ fixed update-initramfs postinst call s/-u/-c/ -maks ]
Signed-off-by: maximilian attems <maks@debian.org>
|
|
The attached patch is needed for booting with root=mtd0.
Basically, the init scripts assume the root device should have a special
device file somewhere, and if not will wait for it to appear. This patch
changes it so that it only waits for a device file to appear if $ROOT starts
with "/dev". In the case of things like ROOT=LABEL=foo, the init scripts
will translate that to a device. With ROOT=mtd0, the init scripts will not
check for a device, and will successfully mount the root filesystem.
Signed-off-by: Andres Salomon <dilinger@debian.org>
|
|
always remove the largest possible suffix pattern on the block device.
(closes: #513958)
Reported-by: Nicola Canepa <canne74@gmail.com>
|
|
Since "exit 2" has a special meaning, we must "exit 1" when the
find/cpio/gzip pipe fails, otherwise "update-initramfs" will potentially
ignore failures (it ignores return code "2" from mkinitramfs).
Signed-off-by: Kees Cook <kees@debian.org>
|
|
since some keyboards will not be usable at the initramfs/busybox
prompt without them.
[ Ubuntu sync for 0.92bubuntu19 port it to Debian - maks ]
|
|
(closes: #514997)
Reported-by: Michael Prokop <mika@grml.org>
|
|
trivial fix on top of previous change.
afais cpio doesn't really give sensible return errors,
so this fall through the cracks.
|
|
The final stage of mkinitramfs that builds the image does not verify the
exit codes of find or cpio:
(cd "${DESTDIR}" && find . | cpio --quiet --dereference -o -H newc | gzip
>"${outfile}") || exit 1
Once bug 514936 is solved, this will be even more important, since cpio
will actually return errors.
Attached is a gross alternative to depending on bash...
Current behavior:
$ find /fail | cpio --quiet --dereference -o -H newc | gzip > /tmp/archive.gz
find: `/fail': No such file or directory
$ echo $?
0
Desired behavior:
$ set -o pipefail
$ find /fail | cpio --quiet --dereference -o -H newc | gzip > /tmp/archive.gz
find: `/fail': No such file or directory
$ echo $?
1
(closes: #514938)
[ already useful in finding stupid mkinitramfs path errors -maks ]
|
|
uncovered due to tests with harden cpio failures.
bug got introduced postlenny in
d2791a8fa458d473ff92d0710fe4d71487c76d6b.
|
|
not needed anymore..
|
|
no need to support callers that were unofficial and never released
as such in Lenny.
|
|
no need to deprecate it in initramfs-tools.
Lenny ships due to userspace with it and Juju firewire not yet
feature complete..
|
|
update-initramfs is the desired caller.
|
|
since Lenny lvm2 ships it's own hooks scripts, no need to dup.
|
|
clears debhelper-but-no-misc-depends lintian warning.
don't know if we really need it, but better be safe and
referenced in debhelper(7) manpage.
|
|
makes it respect "do_bootloader = no" if both lilo and grub2 are
installed.
simplify grub check as both grub2 and grub have an update-grub script.
based on patch and report by Clint Adams <schizo@debian.org>
(closes: #511514)
|
|
we have boot scripts that search for root if none is passed.
usage by grml.org and debian-live.
This reverts commit 90ad6ff31f66112e863ea86c9960b9770a0a245e.
Reported-by: Michael Prokop <mika@grml.org>
|
|
|
|
|
|
|
|
in 0.92m, the timeout for ipconfig in scripts/functions was set to 60
seconds. This broke our iscsi-rootfs setups.
It took me a while to debug this, but it seems that 60 seconds is not
enough for switches which have spanning-tree enabled, they need some
more time before they take up a port, in our case between 62 and 65
seconds.
(closes: #511085)
|
|
in 0.92m, the timeout for ipconfig in scripts/functions was set to 60
seconds. This broke our iscsi-rootfs setups.
It took me a while to debug this, but it seems that 60 seconds is not
enough for switches which have spanning-tree enabled, they need some
more time before they take up a port, in our case between 62 and 65
seconds.
(closes: #511085)
|
|
|
|
The cmdline parameter `break=init' failed to work. It looks like the variable
`break' is unset before `maybe_break init' is evaluated.
(closes: #509637)
|
|
should be just a bit faster and fixes relative path for -d option.
(closes: 472409)
|
|
with bad luck that might still be on hda1,
but then you get a warning about missing root boot param.
|
|
adds FILES section.
|
|
and no we can't write into /var/log itself.
(closes: #488804)
|
|
thanks to jidanni@jidanni.org for pointing out the missed field.
|
|
d-i fix, thus high urgency.
|
|
-d was missing in synopsis.
-c or -d or -u are not optional.
|
|
i've set up an encrypted loop-AES device as the root partition (boot
paramter dev=/dev/loopX) and update-initramfs fails with :
mkinitramfs: missing loop root /dev/loop2 /sys entry
|
|
NFSOPTS is no longer sourced in initramfs.conf,
so it should only be documented at the place of use aka boot cmdline.
(closes: #502927)
Co-Author: Gerfried Fuchs <rhonda@debian.at>
|
|
d-i uses to pass it with the value 1
Reported-by: Frans Pop <fjp@debian.org>
|
|
The root fs is located on LVM and is described in
/etc/fstab with LABEL= notation, which leads
to the following mount output for root filesystem:
mount | grep "on / "
/dev/dm-0 on / type ext3 (rw,noatime,errors=remount-ro)
The /dev/dm-0 itself isn't a symlink to /dev/mapper/, therefore root var
remains /dev/dm-0 after readlink -f attempt.
fix the consequent root detection failure.
(closes: #508906)
|
|
|
|
no need to check if the string of the corresponding variable exists,
just unset all of them, but 2 that we need to pass on for calling init.
|
|
Fix buglet in parse_numeric where *:* would match mtd:root. We only
want to match numbers. This fixes redboot partition support.
Signed-off-by: Andres Salomon <dilinger@debian.org>
|
|
|
|
udev isn't started at this point and therefore can't create framebuffer
devices. This causes usplash not to run on PS3.
set sane permissions will making the char files.
This reverts commit 0aec8b0c22b7622841c4ab7a3b492b4d2657456f.
|
|
rather than /tmp/initramfs.debug, so that it can be retrieved after boot.
|
|
Uvesafb framebuffer driver needs v86d userspace program
but when fb driver is modprobed at init-top stage of initrd,
/dev/zero and /dev/mem are missing because udev have not
been run yet.
|
|
harmless as the postrm is small, but better be clean.
|
|
|
