Bug#824394: using dpkg --force-all is problematic

2016-05-15 Thread Niall Walsh
Package: cdebootstrap Version: 0.7.3 Severity: normal Tags: patch Hi, since dpkg 1.18.5 reached sid and testing cdeboostrap fails with (if run with --debug): dpkg (subprocess): failed to chdir to '': No such file or directory cdebootstrap is calling dpkg with --force-all which enables --for

Bug#762032: cryptsetup: cryptroot fails on lvm if device isn't ready in time

2014-09-17 Thread Niall Walsh
Package: cryptsetup Version: 2:1.6.6-1 Severity: normal With the rootfs on lvm on crypt on lvm on a usb device the system can (will) fail to boot if the usb device has not appeared in time. The lvm initramfs-tools script will just call lvm lvchange for the root device, but in this case the root

Bug#752919: networking scripts runs ifup for hotplugged interfaces, races with net.agent from udev

2014-08-08 Thread Niall Walsh
On 08/08/14 20:22, Michael Biebl wrote: Am 08.08.2014 21:07, schrieb Niall Walsh: Hi I think a better thing would be to block ifup requests from udev script until networking initscript has finished Why? That was a quote from the last mail from Andrew. Yikes, I certainly won't add a

Bug#752919: networking scripts runs ifup for hotplugged interfaces, races with net.agent from udev

2014-08-08 Thread Niall Walsh
Hi I think a better thing would be to block ifup requests from udev script until networking initscript has finished The following diff seems to work. --- net.agent.orig 2014-08-08 15:41:17.389440075 +0100 +++ net.agent 2014-08-08 16:26:46.869439749 +0100 @@ -49,6 +49,12 @@ wait_f

Bug#714016: kexec-tools: use_grub_config=true loads wrong kernel with grub>=2.0 grub.cfg

2013-06-24 Thread Niall Walsh
Package: kexec-tools Version: 1:2.0.3-2 Severity: important Tags: patch security Hi, if use_grub_config=true is set in /etc/default/kexec, grub>=2 is installed and more then one kernel is installed the wrong kernel will be loaded. The problem is the newer grub places all but the latest kernel i

Bug#709567: Undefined subroutine &conffile::abs_path called at /usr/bin/ucfq line 529, line 13.

2013-05-24 Thread Niall Walsh
use Cwd qw[abs_path]; is on line 30 of /usr/bin/ucfq and hence in the ucf class Placing it on line 505 and hence in the conffile class corrects this error and makes the abs_path function available in the conffile class. abs_path is only used in ucfq in this one place so this seems ok. I'll happ

Bug#632196: libc6: unable to deboostrap as preinst thinks extracted files are copy

2011-06-30 Thread Niall Walsh
Package: libc6 Version: 2.13-8 Severity: normal Hi, attempting to (c)deboostrap sid with with 2.13-8 fails with the following: O: Unpacking libc6 (from .../libc6_2.13-8_amd64.deb) ... P: Unpacking package libc6 O: O: A copy of the C library was found in an unexpected directory: O: '/lib/x86_6

Bug#628175: upower: uploaded amd64 packages depend on libimobiledevice2 which is not in sid

2011-05-27 Thread Niall Walsh
Package: upower Version: 0.9.11-1 Severity: important Hi, the recently uploaded upower depends on libimobiledevice2 which is not available in sid but is only available from experimental. This renders the package uninstallable on sid. The package rebuilds fine and at a quick glance appears to w

Bug#625204: Missing attachment

2011-05-02 Thread Niall Walsh
Oooops I forgot the patch attachment! Sorry, here it is. Niall--- ifupdown.orig 2011-04-30 19:47:20.826097001 +0100 +++ ifupdown.new 2011-04-29 19:29:08.743097001 +0100 @@ -42,7 +42,7 @@ fi fi -while [ ! -e "$dest" ]; do +while [ ! -e "$dest" ] && [ ! -L "$dest" ]; do

Bug#625204: ifupdown initscripts myreadlink fails to handle nested symlinks (/run)

2011-05-02 Thread Niall Walsh
Package: ifupdown Version: 0.6.10 Tags: patch Hi, if a system is using /run then /dev/shm is a symlink pointing to /run/shm and ifupdowns initscript will fail to create /etc/network/run What happens is /etc/network/run is a symlink to /dev/shm/network so the initscript will try to use its myread

Bug#591275: nbd-client: initramfs script unconditionally calls configure_networking

2010-08-01 Thread Niall Walsh
Package: nbd-client Version: 1:2.9.16-2 Severity: normal Hi, Sorry to probably inspire another sigh, but the configure_networking fix has problems so I think the nbd-client.initrd needs another tweak. The nbd-client initrd script now always calls configure_networking This means that if you hav

Bug#553022: nbd-client: initrd 3 argument nbdroot sed function can fail

2009-10-29 Thread Niall Walsh
Package: nbd-client Version: 1:2.9.14-1 Severity: wishlist Tags: patch Hi again, in a "live" busybox+klibc based initramfs the sed function to extract the port from the 3 argument form of nbdroot fails. The parts of the regex work but the total regexp fails to do anything, I didn't really dig

Bug#553019: nbd-client: eth0 unalterably hardcoded in initramfs script

2009-10-29 Thread Niall Walsh
Package: nbd-client Version: 1:2.9.14-1 Severity: normal Tags: patch Hi, As it stands, and as mentioned in the README.Debian which points to bug #460569 which has been closed, it will currently only work if the nbd-server is connected to a working eth0. initramfs-tools closed their bug as they

Bug#524538: cdebootstrap fails semi-randomly due to change in tar behaviour

2009-04-17 Thread Niall Walsh
Package: cdebootstrap Version: 0.5.4 Tags: patch Hi, If tar 1.22-1 is installed, cdebootstrap (0.5.4) fails regularly, but not always, complaining about a broken pipe. tar 1.22-1 introduced a new behaviour where a SIGPIPE can be sent "early". This manifested itself in bug #522858 which i

Bug#420799: libc6: postinst fails on awk during cdebootstrap (i386, sparc)

2007-04-24 Thread Niall Walsh
Package: libc6 Version: 2.5-4 Severity: normal If you attempt to cdebootstrap sid on i386 it fails on the libc6 postinst. O: Selecting previously deselected package mawk. O: dpkg: regarding .../mawk_1.3.3-11_i386.deb containing mawk, pre-dependency problem: O: mawk pre-depends on libc6 (>= 2.3.

Bug#410920: apt-cacher: documentation of rate limit is incorrect

2007-02-14 Thread Niall Walsh
Package: apt-cacher Version: 1.5.3 Severity: minor Both the default apt-cacher.conf and the apt-cacher man page suggest you can use either k or m as arguments to the limit option. The man page goes as far as to suggest using man wget for the full syntax and claims the default value is bytes/sec

Bug#405081: kernel-package: postinst link_in_boot value set by image_in_boot

2006-12-30 Thread Niall Walsh
Package: kernel-package Version: 10.065 Severity: minor In kernel/pkg/image/postinst the link_in_boot variable is set by incorrectly checking for image_in_boot from the config file. 119c119 < $link_in_boot= "" if /image_in_boot\s*=\s*(no|false|0)\s*$/ig; --- > $link_in_boot=