Binary package hint: update-manager
TEST CASE:
1. install hardy
2. open a terminal
3. run "sudo update-manager --dist-ugprade"
4. watch failure in the terminal window
5. install update-manager from hardy-propose
6. repeat step 3
7. verify that there is no failure this time
I am already running the Hardy beta, and decided to try updating my system to the latest version running 'kdesu "adept_manager --dist-upgrade-devel"'. After detecting there are no upgrades, update manager crashes:
Traceback (most recent call last):
File "/tmp/kde-root/adept_managerHJMRla.tmp-extract/dist-upgrade.py", line 60, in
app.run()
File "/tmp/kde-root/adept_managerHJMRla.tmp-extract/DistUpgradeController.py", line 1550, in run
self.fullUpgrade()
File "/tmp/kde-root/adept_managerHJMRla.tmp-extract/DistUpgradeController.py", line 1513, in fullUpgrade
if not self.askDistUpgrade():
File "/tmp/kde-root/adept_managerHJMRla.tmp-extract/DistUpgradeController.py", line 818, in askDistUpgrade
self.cache.requiredDownload)
File "/tmp/kde-root/adept_managerHJMRla.tmp-extract/DistUpgradeViewKDE.py", line 631, in confirmChanges
msg = unicode(self.confirmChangesMessage, 'UTF-8')
AttributeError: 'DistUpgradeViewKDE' object has no attribute 'confirmChangesMessage'
Binary package hint: pulseaudio
I had been using PulseAudio in Gutsy for the past several months, so all
the mundane stuff (groups, permissions, config files) is configured
correctly. Since the upgrade to Hardy, ALSA is unable to use PulseAudio
anymore. Details follow:
- I am in the groups pulse-rt and pulse-access
- ALSA without PulseAudio works just fine
- PulseAudio seems to work fine with GStreamer
Here's what my .asoundrc looks like:
=======================
pcm.pulse {
type pulse
}
ctl.pulse {
type pulse
}
pcm.!default { type pulse }
ctl.!default { type pulse }
=======================
I'm attaching my default.pa with the bug.
Here's the output of 'pulseaudio -vvv' when I try to run speaker-test (only relevant portion here, full output attached):
I: client.c: Created 1 "Native client (UNIX socket client)"
I: protocol-native.c: Got credentials: uid=1000 gid=1000 success=1
I: protocol-native.c: Enabled SHM for new connection
I: client.c: Client 1 changed name from "Native client (UNIX socket client)" to "ALSA plug-in [speaker-test]"
I: client.c: Freed 1 "ALSA plug-in [speaker-test]"
I: protocol-native.c: connection died.
speaker-test fails with the following:
WAV file(s)
Rate set to 48000Hz (requested 48000Hz)
Buffer size range from 43 to 349525
Period size range from 21 to 87382
Using max buffer size 349524
Periods = 4
was set period_size = 87381
was set buffer_size = 349524
*** PULSEAUDIO: Unable to create stream.
Unable to set hw params for playback: Input/output error
Setting of hwparams failed: Input/output error
speaker-test: pcm_pulse.c:115: pulse_stop: Assertion `pcm->stream' failed.
Aborted
TEST CASE:
1. Create an .asoundrc file in your home directory to direct alsa to use the pulseaudio plugin by default for all sound output. This is done with the command "asoundconf set-pulseaudio"
2. Attempt to run speaker-test with the following command: "speaker-test -twav -c2"
3. You should get the following error:
*** PULSEAUDIO: Unable to create stream.
speaker-test: pcm_pulse.c:115: pulse_stop: Assertion `pcm->stream' failed.
speaker-test 1.0.15
Playback device is default
Stream parameters are 48000Hz, S16_LE, 2 channels
WAV file(s)
Rate set to 48000Hz (requested 48000Hz)
Buffer size range from 128 to 1048576
Period size range from 64 to 262144
Using max buffer size 1048576
Periods = 4
was set period_size = 262144
was set buffer_size = 1048576
Unable to set hw params for playback: Input/output error
Setting of hwparams failed: Input/output error
NOTE: This SRU is in two parts, one for alsa-lib, and the other for alsa-plugins. The new revision of alsa-plugins requires the new revision of alsa-lib in order to function properly.
Intrepid status:
This bug is fixed with alsa-lib 1.0.16 and alsa-plugins 1.0.16 which are both in intrepid.
Regression potential:
alsa-lib 1.0.16 introduces new symbols related to plugins, but doesn't change any of the API for existing applications that use alsa. Thus the regression potential is minimal.
Binary package hint: pulseaudio
I had been using PulseAudio in Gutsy for the past several months, so all
the mundane stuff (groups, permissions, config files) is configured
correctly. Since the upgrade to Hardy, ALSA is unable to use PulseAudio
anymore. Details follow:
- I am in the groups pulse-rt and pulse-access
- ALSA without PulseAudio works just fine
- PulseAudio seems to work fine with GStreamer
Here's what my .asoundrc looks like:
=======================
pcm.pulse {
type pulse
}
ctl.pulse {
type pulse
}
pcm.!default { type pulse }
ctl.!default { type pulse }
=======================
I'm attaching my default.pa with the bug.
Here's the output of 'pulseaudio -vvv' when I try to run speaker-test (only relevant portion here, full output attached):
I: client.c: Created 1 "Native client (UNIX socket client)"
I: protocol-native.c: Got credentials: uid=1000 gid=1000 success=1
I: protocol-native.c: Enabled SHM for new connection
I: client.c: Client 1 changed name from "Native client (UNIX socket client)" to "ALSA plug-in [speaker-test]"
I: client.c: Freed 1 "ALSA plug-in [speaker-test]"
I: protocol-native.c: connection died.
speaker-test fails with the following:
WAV file(s)
Rate set to 48000Hz (requested 48000Hz)
Buffer size range from 43 to 349525
Period size range from 21 to 87382
Using max buffer size 349524
Periods = 4
was set period_size = 87381
was set buffer_size = 349524
*** PULSEAUDIO: Unable to create stream.
Unable to set hw params for playback: Input/output error
Setting of hwparams failed: Input/output error
speaker-test: pcm_pulse.c:115: pulse_stop: Assertion `pcm->stream' failed.
Aborted
TEST CASE:
1. Create an .asoundrc file in your home directory to direct alsa to use the pulseaudio plugin by default for all sound output. This is done with the command "asoundconf set-pulseaudio"
2. Attempt to run speaker-test with the following command: "speaker-test -twav -c2"
3. You should get the following error:
*** PULSEAUDIO: Unable to create stream.
speaker-test: pcm_pulse.c:115: pulse_stop: Assertion `pcm->stream' failed.
speaker-test 1.0.15
Playback device is default
Stream parameters are 48000Hz, S16_LE, 2 channels
WAV file(s)
Rate set to 48000Hz (requested 48000Hz)
Buffer size range from 128 to 1048576
Period size range from 64 to 262144
Using max buffer size 1048576
Periods = 4
was set period_size = 262144
was set buffer_size = 1048576
Unable to set hw params for playback: Input/output error
Setting of hwparams failed: Input/output error
NOTE: This SRU is in two parts, one for alsa-lib, and the other for alsa-plugins. The new revision of alsa-plugins requires the new revision of alsa-lib in order to function properly.
Intrepid status:
This bug is fixed with alsa-lib 1.0.16 and alsa-plugins 1.0.16 which are both in intrepid.
Regression potential:
alsa-lib 1.0.16 introduces new symbols related to plugins, but doesn't change any of the API for existing applications that use alsa. Thus the regression potential is minimal.
Binary package hint: bacula
Actually this is the last point of the original report of bug #207527.
I am opening this as a new bug report, since the other issues of bug #207527 appear to be fixed.
When running the BackupCatalog job, I get the following error:
06-Mai 18:27 baculatestsrv-dir JobId 17: BeforeJob: run command "/usr/bin/awk -f /etc/bacula/scripts/make_catalog_backup_awk -v cat1=MyCatalog /etc/bacula/bacula-dir.conf"
06-Mai 18:27 baculatestsrv-dir JobId 17: BeforeJob: mysqldump: Got error: 1045: Access denied for user 'bacula'@'localhost' (using password: NO) when trying to connect
It seems that .my.cnf is not read, since the command is not executed in
a login shell and HOME is not set.
I was able to get the backup running by specifying HOME in the mysqldump
command (see attached patch).
Binary package hint: bacula
bacula-fd has a know bug/crash when the strippath option is used.
more information here: http://bugs.bacula.org/view.php?id=1047
The upstream bug has not been fixed in 2.2.8!!
After upgrade to Hardy I've noticed that redrawing in Evolution was very
slow. A quick google search led me to
http://thread.gmane.org/gmane.os.freebsd.devel.gnome/25373, so I checked
if the Ubuntu package contains the same patch. According to
http://cairographics.org/news/cairo-1.5.18/ the bug affects only xorg
1.3 and older (Hardy has xorg 1.4), and the workaround should be already
in cairo 1.6. So I tried to recompile the package without 03
-turn_on_buggy-repeat_handling.dpatch and it indeed fixed the slowness.
Since Hardy no longer needs the patch, I suggest dropping it from the
package.
Bug 207137, Wubi's ISO image creation from a CD process failing, seems
to be biting a fair number of users. A way around having to use the
cd2iso tool and thus avoid the bug would be to copy only the squashfs.
This change to casper implements the necessary code on the Ubuntu side
for that.
check_dev in scripts/casper could be made more flexible if it was
possible to look for a casper fs (is_casper_path) within an arbitrary
path. At the moment check_dev supports scanning either devices or ISOs.
This could be implemented with the following snippet:
if [ -d "${devname}" ]; then
mount -o bind $devname $mountpoint
if is_casper_path $mountpoint; then
echo $mountpoint
return 0
endif
fi
TEST CASE:
"On my test system I couldn't get Wubi to install using the cd-rw drive in the machine. Since then I've added a dvd-rw and wubi installs flawlessly.
The cd installs to the part at the end of creating the image file at
which point it throws up the error "Could not access cd, please make
sure other applications are not using it and try again.""
As reported by davmor2 of the QA team.
A debdiff of the uploaded patch can be found here:
http://launchpadlibrarian.net/14822820/casper_1.132ubuntu0.1.debdiff
dpkg fails
ProblemType: Package
Architecture: amd64
Date: Fri May 9 21:16:56 2008
DistroRelease: Ubuntu 8.04
ErrorMessage: subprocess post-installation script returned error exit status 1
NonfreeKernelModules: fglrx
Package: dkim-filter 2.5.4.dfsg-0ubuntu2
PackageArchitecture: amd64
SourcePackage: dkim-milter
Title: package dkim-filter 2.5.4.dfsg-0ubuntu2 failed to install/upgrade: subprocess post-installation script returned error exit status 1
Uname: Linux 2.6.24-16-server x86_64
ERROR:
Setting up dkim-filter (2.5.4.dfsg-0ubuntu2) ...
adduser: Warning: The home directory `/var/run/dkim-filter' does not belong to the user you are currently creating.
Starting DKIM Filter: dkim-filter: /etc/dkim-filter.conf: at least one selector and key required for signing mode
invoke-rc.d: initscript dkim-filter, action "start" failed.
dpkg: error processing dkim-filter (--configure):
subprocess post-installation script returned error exit status 78
TEST CASE for Hardy SRU:
Install dkim-filter package from hardy-release on a system that has not
had it configured before. See the installation failure as above.
Install the dkim-filter package from hardy-proposed and then get:
Preparing to replace dkim-filter 2.5.4.dfsg-0ubuntu2 (using dkim-filter_2.5.4.dfsg-0ubuntu2.1_i386.deb) ...
Stopping DKIM Filter: No /usr/sbin/dkim-filter found running; none killed.
dkim-filter.
Unpacking replacement dkim-filter ...
Setting up dkim-filter (2.5.4.dfsg-0ubuntu2.1) ...
Installing new version of config file /etc/init.d/dkim-filter ...
Starting DKIM Filter: dkim-filter: /etc/dkim-filter.conf: at least one selector and key required for signing mode
See /usr/share/doc/dkim-filter/README.Debian for help
Starting for DKIM verification only
The card ID lists in envyng don't contain any reference to the nVidia
GeForce 7100 / NVIDIA nForce 630i card, even though it's on the list of
supported cards at http://www.nvidia.com/object/IO_18897.html with ID
0x07E1.
This is envyng 1.1.1 on Ubuntu 8.04 (release).
Binary package hint: evolution-data-server
there was some discussion upstream about wrong timezone between
displayed, the change on
http://svn.gnome.org/viewvc/libical?view=revision&revision=645 fixes the
issue
TESTCASE:
- open the calendar and tasks properties dialo
- open the map to select a timezone
- mouseover locations, some have weird timezones informations, Guadalcanal is +10:39:48 for example, the change fixes the issue
Binary package hint: gcc-snapshot
the current gcc-snapshot package in hardy is taken from the
gcc-4_3-branch. Proposing to update this package to the final gcc-4.3.1
when it is released.
Binary package hint: genpo
/usr/share/menu/genpo missing backslash at end of line 5
The package is non-functional this way.
The bug manifests itself by making sure no applications installed after
this package appear in the Applications menu
Test Case:
Install genpo and menu packages. In Debian menu see if genpo entry is in the Applications menu.
Install another application that places a menu item after genpo. It should not appear.
This is due to bug 8497 but I am keeping it separate because I would like to provide a wubi specific workaround.
One way is to add an extra boot menu option with kernel argument edd=on to address the issue.
See https://bugs.launchpad.net/ubuntu/+source/grub/+bug/8497
TEST CASE: On some systems with multiple disks, grub can set the root device incorrectly, causing a Wubi install to fail.
As suggested by bean123 and tinybit it is possible to modify the grub-
installer line so that root is not set at all within menu.lst. This is
because another menu.lst is used beforehand which will set root and then
call configfile.
http://launchpadlibrarian.net/14092458/grub-installer-wubi.debdiff is a
patch for hardy-proposed. Please note that ubiquity needs to be
uploaded afterwards in order for this to be tested.
Potential regressions introduced by this patch should only be limited to
Wubi installs as it is currently the only case in which grub-
installer/bootdev_directory would be set.
When I try to listen to the Real Player stream or old shows on BBC Radio
1, I get totally garbled output in my headphones.
The Real Player stream is affected:
http://www.bbc.co.uk/radio/aod/networks/radio1/live.shtml
But the Windows media stream is fine:
http://www.bbc.co.uk/radio/aod/networks/radio1/wmp.shtml?9all
I'm using Hardy Beta, 64-bit, all updates applied as of today.
TEST CASE
make sure you have gstreamer0.10-plugins-ugly installed, and dont have realplayer installed (or anything else that may try to play realplayer streams)
visit http://www.bbc.co.uk/radio/aod/networks/radio1/live.shtml or http://www.bbc.co.uk/radio/aod/radio4.shtml
observe how the audio stream sounds horribly jumbled (no that not just what modern music sounds like)
now with the patched version, go to the same page, and the audio will be
clear.
Binary package hint: hal-info
According to the recently approved general SRU exception for hal-info
[1] I would like to update hardy's hal-info to a newer snapshot for
better hardware/suspend/music player support.
[1] https://wiki.ubuntu.com/StableReleaseUpdates#head-
83429308e152341202aac4f793e9a501cb742e4f
There has been much confusion surrounding the user getting dropped in
the initramfs shell when the host filesystem cannot be mounted because
it is dirty.
TEST CASE:
When using Ubuntu (Hardy) under wubi, if the NTFS dirty flag is set, the error message is missing or confusing.
This situation can and will occur when you restart and are not paying
attention, so you miss the NT Boot loader prompt and start booting
Windows, then hard-power-off to save time (can you guess how I know
this?), or if/when Windows crashes. I bet this is going to be a really
F.A.Q. with Wubi.
To reproduce:
1) Either shut Windows down ungracefully, or let it start half-way up, then hard-power-off to kill it
2) Reboot and attempt to go into Ubuntu
Case 1: default startup: You end up at a completely useless and scary BusyBox/(initramfs) prompt with no indication why you are there or what to do about it.
Case 2: recovery mode (verbose) startup: You still end up at a completely useless and scary BusyBox/(initramfs) prompt, but there are log lines on the screen as follows. But they are scary and confusing ("mount: ... failed: Success"? WTH?):
----- cut here (very top of my screen) -----
$LogFile indicates unclean shutdown (0, 0)
Failed to mount '/dev/sda2/: Operation not supported
Mount is denied because NTFS is marked to be in use. Choose one action:
Choice 1: If you have Windows then disconnect the external devices by
clicking on the 'Safely Remove Hardware' icon in the Windows
taskbar then shutdown Windows cleanly.
Choice 2: If you don't have Windows then you can use the 'force' option for
your own responsibility. For example type on the command line:
mount -t ntfs-3g /dev/sda2 /root -o force
Or add the option to the relevant row in the /etc/fstab file:
/dev/sdc1 /mnt/external/ ntfs-3g defaults,force 0 0
mount: Mounting /dev/disk/by_uuid/... on /root failed: Success
mount: Mounting /root on /host failed: invalid argument
ALERT! /host/ubuntu/disks/root.disk does not exist. Dropping to a shell!
BusyBox v1.1.3 ... (ash)
Enter 'help' ....
(initramfs)
----- cut here -----
A diff of the required changes can be found here:
http://launchpadlibrarian.net/14795872/lupin-226622.debdiff
A debdiff of the entire 0.19 upload, including the fix for bug 136682, can be found here:
http://launchpadlibrarian.net/14795586/lupin_0.16-0.19.debdiff
A debdiff of the initramfs-tools changes can be found here:
http://launchpadlibrarian.net/14795732/initramfstools_0.85eubuntu36-0.85eubuntu38.debdiff
The daemon rpc.gssd uses the clnttcp_create function from
librpcsecgss.so to build authentication connections NFS servers
providing Kerberized NFS. If a connect() call is taking a long time,
then there is a good chance that a change to the pipefs will cause
rpc.gssd to get a signal.
We've provided a patch to upstream, which they've accepted. 0.18
contains the fix. I'll also add the patch here, and look at providing a
debdiff that incorporates the patch, if I can wrap my head around CDBS.
This bug impacts users of systems that utilise Kerberised NFS in an
environment with automounting. On sufficiently loaded systems, the
signal from a new mount may interrupt a connect() in progress for an
existing mount being made. This can render NFS in such environments
ineffective.
A patch has been provided upstream, which they have incorporated into
the released 0.18 version of the software. 0.18-1 is in Debian, and A
sync has been requested in #234297
A patch and debdiff are attached to this bug report.
TEST CASE:
Probably a difficult one, you'd need a Kerberised NFS environment to
start with, then you'd have put the client mounting shares under load,
and make subsequent mount attempts. I think it's fairly non-
deterministic, but I'm not the original internal bug submitter, so I
don't have first-hand details.
The ThinkPad R31 (really an Acer) requires passing 'i8042.nomux=1' on the kernel
commandline to stop erratic pointer movement caused by the buggy embedded
controller/keyboard combp chip.
This movement occurs approximately every 60seconds if the pointer is being used
at the same time as `batt-applet` queries state. With ACPI in use, this occurs
approximately every 60 seconds. With ACPI disabled, the erratic movement is
continous, rendering the machine completely unusable with no way to switch to a
virtual-terminal.
This occurs on the TP R30 and other Acer-based machines, as that was where the
solution was pin-pointed from. See other laptops requiring this parameter:
http://www.google.co.uk/search?q=i8042.nomux+required
As well as machines from Sharp and HP.
http://wiki.ubuntu.com/LaptopTestingTeam/ThinkpadR31:
http://wiki.ubuntu.com/LaptopTestingTeam/ThinkpadR31
Binary package hint: linux-source-2.6.20
02:00.1 Generic system peripheral [0805]: Ricoh Co Ltd R5C822 SD/SDIO/MMC/MS/MSPro Host Adapter (rev 13)
Subsystem: IBM Thinkpad X40
Flags: bus master, medium devsel, latency 64, IRQ 17
Memory at d0221000 (32-bit, non-prefetchable) [size=256]
The device does not work, modprobing mmc_block and inserting an MMC card makes the kernel crash. This is a regression from Edgy.
Binary package hint: linux-source-2.6.20
I have a dvb device (ASUS My Cinema U3000 Mini) which is not detected.
Adding Vendor ID and Product ID to driver and putting the firmware in
/lib/firmware/`uname -r` the driver is loaded and the device is working.
My system:
Dell Vostro 1400
Intel Core 2 Duo @ 2.0 GHz
3 GB Ram
1 GB Swap
I recently got suspend & hibernate working on my system by modifying the
hal scripts to use acpi-support sleep.sh and hibernate.sh instead of
powersaved. Now, however, I have found a bug regarding the suspend and
resume scripts:
When I resume from a suspend (suspend to ram, in particular),
/proc/cpuinfo reports that CPU1 is running @ 1994 MHz. Further
investigation reveals that /sys/devices/system/cpu/cpu1/cpufreq is gone
entirely (It normally is a symlink to cpu0's cpufreq directory).
If there's any extra info I can provide, I am more than willing to.
There is a driver for the Hauppage WinTV HVR 900 with USB id 2040:6502
(the B3C0 revision.)
linux-ubuntu-modules-2.6.24-8 doesn't support this device. It does
support the Hauppage WinTV HVR 900, but not the one with this USB id.
I found some sort of patch here:
http://www.kernel.org/pub/linux/kernel/people/jejb/merge-
tree/patch-2.6.25-rc2-mc2.diff
Can this device be supported by hardy?
Kernel version 2.6.24 provide new support of acpi for libata by default.
This options make some problems with suspend on system with brocken DSDT
like ASUS P5LD2-VM . Workaround is it to disbale it by setting
"libata.noacpi=1" or "options libata noacpi=1"
Link to meta-bug for ASUS P5LD2-VM: Bug #202804
hardware: macbook pro generation 4,1 (feb 2008)
problem: fn key not detected
test: run xev. press fn key. nothing happens.
consequences: home/end/pgup/pgdown don't work.
i was able to fix this issue by compiling my own kernel.
i downloaded linux-source from ubuntu hardy
(VERSION = 2,PATCHLEVEL = 6,SUBLEVEL = 24,EXTRAVERSION = .3)
and revised this file:
drivers/hid/usbhid/hid-quirks.c
used lsusb to grab the id of the internal keyboard device:
Bus 005 Device 003: ID 05ac:0230 Apple Computer, Inc.
and revised the c file accordingly:
/* next line commented out by eitan and replaced with line below;
changing address for device from 0x021b to 0x0230 which is what lsusb tells me
my internal keyboard device id actually is
#define USB_DEVICE_ID_APPLE_GEYSER4_ISO 0x021b
*/
#define USB_DEVICE_ID_APPLE_GEYSER4_ISO 0x0230
rebooting with the new kernel, the fn key now works.
i don't know if this breaks other versions of the macbook pro
and am otherwise ignorant of the linux kernel source code.
thanks,
/ eitan
Trying 2.6.24-13-server doesn't help. Since 2.6.23 somewhere, sata_nv
has gotten improved exception handling, but there's a serious bug fixed
in 2.6.25-pre6. On a server, to guarantee data integrity, I run #
hdparm -W 0 /dev/sda ; hdparm -W 0 /dev/sdb. On Gutsy, this gives
periodic kernel messages, but on Hardy it fails terribly. XFS
filesystem shutdown due to error, and most times system reboots without
any messages to serial console.
The fix is in the following commit in git linux-2.6-stable tree:
author Robert Hancock
Wed, 30 Jan 2008 01:53:19 +0000 (19:53 -0600)
committer Jeff Garzik
Fri, 1 Feb 2008 17:26:38 +0000 (12:26 -0500)
commit a1fe782414b7122d4c0501d3a0988b7302fa586f
.....
This patch is based on an original patch from Kuan Luo of NVIDIA,
posted under subject "fixed a bug of adma in rhel4u5 with HDS7250SASUN500G".
His description follows. I've reworked it a bit to avoid some unnecessary
repeated checks but it should be functionally identical.
"The patch is to solve the error message "ata1: CPB flags CMD err,
flags=0x11" when testing HDS7250SASUN500G in rhel4u5.
I tested this hd in 2.6.24-rc7 which needed to remove the mask in
blacklist to run the ncq and the same error also showed up...
.....
Binary package hint: linux-source-2.6.24
If two prefixes are advertized on a link, one with a preferred_lifetime of 0 (usually the case after a renumbering), then ipv6 temporary addresses won't work.
The reason is that when router advertisement is processed, they update all the temporary addresses and not just the relevant one.
Furthermore, no temporary addresses should be created with a preferred lifetime of 0.
patches are available here:
http://git.kernel.org/?p=linux/kernel/git/davem/net-2.6.git;a=commit;h=c6fbfac2e61c9a8617f64b93e8c990b8d864bce5
http://git.kernel.org/?p=linux/kernel/git/davem/net-2.6.git;a=commit;h=eac55bf97094f6b64116426864cf4666ef7587bc
please backport it, this configuration affects the second biggest french
ISP (more than 1M subscribers) and without it temporary address is
useless.
Binary package hint: linux-source-2.6.24
I was going to give Hardy a shot on my PA Semi Electra board. I see that
the kernel is built with PPC_PASEMI=y, which is a very good start.
However, it lacks the ethernet driver for the on-chip ethernet
interfaces. Care to apply the attached patch?
Thanks!
When I boot the system with Ubuntu 8.04 Live CD/DVD, I found system abort to busybox with udevd-event "sbin/modprobe abort abnormal" error message.
By looking at the dmesg log, I found kernel abort by the time of running rtl8169_init_one function in rtl8169.c. Actually, the NIC driver on my system is RTL8102EL.
And I have tried Ubuntu 7.10 LiveCD/DVD and can't find this problem and the NIC driver can be loaded.
Dmesg file is attached.
I have tried several other kernel version with different daily build of
8.04 LiveCD including 2.6.24-12, 2.6.24-15, 2.6.24-16. It happens with
all the kernels.
The rpcb_getport_async function in the sunrpc module copies larger memory area than the allocated on Ubuntu Hardy.
This can cause oops.
This bug is derived from Linux 2.6.24 in kernel.org.
It seems to be fixed on Linux 2.6.25 in kernel.org by the commit 86d61d8638ddf9cdf87df26c7fa69b2804425fbe.
http://lkml.org/lkml/2008/5/2/205
On Fri, May 02, 2008 at 02:19:34AM -0700, Yinghai Lu wrote:
> On Fri, May 2, 2008 at 12:44 AM, Ingo Molnar wrote:
> > * Matt Domsch wrote:
> >
> > > fix x86 DMI checks for PCI quirks
> > >
> > > http://bugzilla.kernel.org/show_bug.cgi?id=10583
> > > https://bugzilla.redhat.com/show_bug.cgi?id=444791
> > >
> > > Since git commit 08f1c192c3c32797068bfe97738babb3295bbf42 (between
> > > kernels 2.6.22 and 2.6.23), arch/x86/pci/acpi.c has not called
> > > pcibios_scan_root(), which would have called
> > > arch/x86/pci/common.c:dmi_check_system(). This has prevented the
> > > quirks listed in pciprobe_dmi_table[] from being checked and
> > > appropriate action taken.
> >
> > ugh ...
> >
> >
> > > This manifests itself in several Dell and HP servers not automatically
> > > having the pci=bfsort option be applied, as well as Samsung X20 and
> > > Compaq EVO N800c systems needing pci=assign-all-busses was no longer
> > > automatically applied.
> >
> > Jesse Barnes (new PCI maintainer) Cc:-ed.
>
> please check the patch in x86.git, it should do the same thing, but
> put the call in pci_access_init...
>
> commit 9817aa147000086bc11b571620ecc1c73a4a614b
> Author: Yinghai Lu
> Date: Mon Apr 14 15:40:37 2008 -0700
Indeed it does (boot tested on one of the affected systems), and is a
simpler patch. I'd be quite happy with this. Bonus that it's already
in the x86.git tree. :-)
Ingo, is this ready to go to Linus?
Now to get it backported to -stable...
Adding the folks from HP who have lots of systems listed.
Thanks,
Matt
--
Matt Domsch
Linux Technology Strategist, Dell Office of the CTO
linux.dell.com & www.dell.com/linux
Recently i have worked on just upgraded Hardy systems.I have noticed
that there is a don't clearly visible regression about rt kernel
flavour. It seems that on 2.6.24 the Dynamic Ticks features is
incompatible with CONFIR_PREEMPT_RT. This cause huge latency and it is a
regression respect Gutsy!
When using the OpenVZ kernel, some relevant net entries are missing in
/proc.
Upstream has fixed this already, patches are available at
http://bugzilla.openvz.org/show_bug.cgi?id=860.
Example output (from Ben in
https://bugs.edge.launchpad.net/ubuntu/+source/linux/+bug/210672/comments/62):
root@hardy-pkgbuilder:/# netstat -aun
Active Internet connections (servers and established)
Proto Recv-Q Send-Q Local Address Foreign Address State
netstat: no support for `AF INET (udp)' on this system.
I'm already building the kernel with this patches to verify it and
I had upstart crashes in an OpenVZ container, when upstart was reading new/changed event files.
(This caused the whole container to crash!)
Scott has debugged this to be caused by non-POSIX behavior in mmap
related functions in the kernel.
There are two patches in the OpenVZ patchset, which cause this:
1. "Return addr from mmap(2) when len is zero"
http://git.openvz.org/?p=linux-2.6.24-openvz;a=commit;h=8d7a5ca70e9913a71ec0ac9dccdb6ce85c23ac4b
2. "IA64: mmap returns EINVAL if len==0"
http://git.openvz.org/?p=linux-2.6.24-openvz;a=commit;h=8ddefba8a401d78484899d05c0f5d3b45a4e8e0b
I've added a patch to debian/binary-custom.d/openvz/patchset/, which
reverts those patches from the OpenVZ patchset and the upstart crashes
are fixed now.
This is with 2.6.24-18-openvz (from Git) on Ubuntu Hardy. The host
system is amd64, the containers x86.
I am running Ubuntu 8.04's kernel and when I resume from S3 suspend, a single
message is displayed on the console about drm_sysfs_suspend, which seems to
come from the function:
static int drm_sysfs_suspend(struct device *dev, pm_message_t state)
This doesn't appear to be an error, so presumably this should be KERN_INFO, not
KERN_ERR?
(less technical users often express surprise and concern when their laptop
comes out of suspend and prints scary messages at them, even if it does
actually work fine)
Binary package hint: linux-source-2.6.20
I have a dvb device (ASUS My Cinema U3000 Mini) which is not detected.
Adding Vendor ID and Product ID to driver and putting the firmware in
/lib/firmware/`uname -r` the driver is loaded and the device is working.
I just upgraded envyng-core to version 1.1.1ubuntu15 because of the
problems the previous version had. Then I tried to install the drivers
of an ATI Radeon Xpress 1250 on both kernels 2.6.24-16 and 2.6.24-17 but
the installation fails because package xorg-driver-fglrx-dev-envy
couldn't be installed. It says it creates conflicts with xorg-driver-
fglrx which isn't even installed! Is there a solution?
The NVIDIA driver in the lrm-envy (169.12) doesn't support Geforce 9xxx
cards and certain 8xxx cards.
Here are the release highlights for driver 173.14.05:
* Added support for the following new GPUs:
o Quadro FX 3600M
o GeForce 9800 GX2
o GeForce 9800 GTX
o GeForce 9600 GT
o GeForce 9600 GSO
o GeForce 9500M GS
o GeForce 8400
o GeForce 8400 GS
* Added support for Quadro FX 5600/4600 SDI and Quadro G-Sync II.
* Resolved a bug causing left and right eyes to be reversed in stereo mode on some Quadro FX GPUs.
* Fixed a problem that caused OpenGL to stop rendering to windows with origins at or beyond 4096 pixels (horizontally) on GeForce 8 and 9 GPUs.
* Fixed a bug causing some Quadro FX 4500 SDI configurations to take a long time to achieve synchronization.
* Added preliminary support for X.Org server 1.5.
* Addressed a problem causing visual corruption when using GeForce 8 GPUs to drive Chi Mei 56" displays.
* Addressed visual corruption when driving Cintiq 20WSX wide format tables with some GeForce 6 and 7 GPUs.
* Fixed OpenGL rendering corruption with textures compressed using the DXT5 compression algorithm.
* Fixed a regression that caused invalid EDIDs to be detected for the internal display device on some notebooks.
* Improved hotkey switching and power management support on some GeForce 8 notebooks.
* Fixed a regression causing some GeForce 6100/6150 systems to fail to restore the screen after DPMS cycles.
* Fixed a bug that prevented the console from being restored correctly in SLI mode on GeForce 6 and 7 GPUs.
* Fixed interlaced modes on GeForce 8 GPUs.
* Fixed a problem that caused the synchronization signal polarity to always be positive for DVI devices on GeForce 8 and 9 GPUs.
* Resolved a problem resulting in X startup to fail on some GeForce 8 and 9 systems without swap space.
* Fixed a bug resulting in X crashes when using GeForce 8 and 9 GPUs in SLI to drive X screens at depth 8.
* Fixed a problem that caused TV output on secondary TVs to be black and white on some GPUs.
* Restored compatibility with recent Linux 2.6 kernels.
See ALSA bugtracker for details:
https://bugtrack.alsa-project.org/alsa-bug/view.php?id=3775
This bug is fixed in ALSA but is still present in Ubuntu 8.04.
Binary package hint: ltsp-client-core
Many of the options available in /var/lib/tftpboot/ltsp/i386/lts.conf
don't work properly.
There are five different bugs here; forgive me for lumping them
all together, but they all come to down to "configure-x.sh is
pretty broken".
1: When I specify
XKBOPTIONS = "ctrl:nocaps"
it is ignored, because it is overridden by
/opt/ltsp/i386/etc/default/console-setup.
The priority of those should go the other way.
2: When I specify
X_MOUSE_DEVICE = "/dev/gpmdata"
X_MOUSE_PROTOCOL = "MouseSystems"
I get a "Mouse1" that does that, but "Mouse0" is still using
/dev/input/mice! The reason that I'm using gpm is so that I
can re-map the mouse buttons, and this doesn't help at all.
When I specify X_MOUSE_DEVICE, it must *override* the default
mouse, not turn into an *additional* mouse.
3: Furthermore, it creates an xorg.conf file with a syntax error.
It writes:
Option Device /dev/gpmdata
but that causes an error unless those strings are quoted, like:
Option "Device" "/dev/gpmdata"
4: There is no way to turn off DPMS. I need a way to add the line
Option "DPMS" "off"
5: There is no way to specify that the DRI module must not be loaded.
I need a way to *delete* the line
Load "dri"
DISTRIB_ID=Ubuntu
DISTRIB_RELEASE=7.10
DISTRIB_CODENAME=gutsy
DISTRIB_DESCRIPTION="Ubuntu 7.10"
ltsp-client-core:
Installed: (none)
Candidate: 5.0.39
Version table:
5.0.39 0
500 http://archive.ubuntu.com gutsy/main Packages
When trying to install ltsp-server-standalone from and upgraded gutsy
but a clean install of ltsp-server-standalone it fails saying LTSP
client installation ended abnormally due to the package xubuntu-
artwork-usplash either being missing or as a guess it maybe miss
labeled.
I tried the main server and several mirrors and go the same result.
I also tried the exclude option ltsp-build-client --exclude xubuntu-
artwork-usplash. But got the same result. Is the exclude option working
correctly?
due to not working on any of the mirrors I have a strong feeling is due
to an incorrect package name being supplied in ltsp-server-stanalone.
Any one fixed this or had the same problem please.
[08:53] hmmm ltsp-update-image wants a 64bit /opt/ltsp....
[08:53] Error: chroot /opt/ltsp/amd64 doesn't exist.
[08:53] altough I Installed 32-bit clients
[08:53] it defaults to the host arch, use -a with it
[08:53] on a 64-bit server?
[08:54] -a i386 should do
[08:54] right
[08:54] good to know
[08:54] that's not in the man page :-P
[08:54] bug the man page on LP :)
[08:55] yeah
[08:55] its in the help output
[08:55] seems the manpages are behind
Binary package hint: ltsp-client-core
The Minimal printserver script /usr/sbin/jetpipe is missing in the Hardy
package. This is needed to create a socket on the LTSP-Client and
provice a local printer.
As workaround you can use the /usr/sbin/jetpipe file from ltsp-client-
core for Gutsy, it woks fine with Hardy too. Simply copy jetpipe to
/opt/ltsp/i386/usr/sbin/jetpipe and recreate squasfs with ltsp-update-
image command.
When thin-client are configured to use NFS root, then the option
LDM_DIRECTX causes a login failure.
I discovered that the file /root/.Xauthority and the directory
/root/.ssh are not created with NFS root setup, because /root is read-
only.
It simple to fix, /root need to be put in the rw directory list for
mount bind in $CHROOT/etc/default/ltsp-client-setup
--- ltsp-client-setup 2008-05-22 09:09:29.000000000 -0400
+++ ltsp-client-setup.ldm-directx-nfs-fix 2008-05-22 09:09:15.000000000 -0400
@@ -13,7 +13,7 @@
# tmpfs/bind directions that get mounted with only directory structure
# preserved
-rw_dirs="/var/lib/xkb /var/log /var/spool /var/tmp /tmp /var/lib/discover /etc/console-setup"
+rw_dirs="/var/lib/xkb /var/log /var/spool /var/tmp /tmp /var/lib/discover /etc/console-setup /root"
# tmpfs/bind directions that get mounted with directory structure and data
# copied
Fresh i386 Ubuntu Hardy install, not upgrade
Description: Ubuntu 8.04
Release: 8.04
/opt/ltsp/i386/etc/init.d/ltsp-client-setup has
NBD_PORT=${NBD_PORT:-"9210"}
but
/etc/inetd.conf has:
9572 stream tcp nowait nobody /usr/sbin/tcpd /usr/sbin/nbdswapd
And, in lts.conf, NBD_SWAP=Y won't work unless you also specify NBD_PORT=9572
NBD_PORT should default to 9572 in ltsp-client-setup
(As per the edubuntu handbook I did cp /etc/apt/sources.list
/opt/ltsp/i386/etc/apt/ and update;safe-upgrade -- This may be related)
TEST CASE: When I start Ubuntu (installed with wubi), it is almost
prefect that the host drive (the drive "D" in my case with Wubi virtual
hard disk installed) is mounted automatically. Subsequently, I found
that all the files and directory with "Traditional Chinese" name is
disappeared in the host drive. I have checked this out by reboot my
computer, starting Windows XP and found that those files and directory
are still there.
Locale support is needed in the initramfs when the ntfs host filesystem
is mounted in order for proper i18n file support. As this is a patch to
lupin, it only affects Wubi users.
A debdiff of the proposed change, written by Tormod Volden, can be found here:
http://launchpadlibrarian.net/14399869/lupin-locale.debdiff
A debdiff of the entire 0.19 upload, including the fix for bug 226622, can be found here:
http://launchpadlibrarian.net/14795525/lupin_0.16-0.19.debdiff
There has been much confusion surrounding the user getting dropped in
the initramfs shell when the host filesystem cannot be mounted because
it is dirty.
TEST CASE:
When using Ubuntu (Hardy) under wubi, if the NTFS dirty flag is set, the error message is missing or confusing.
This situation can and will occur when you restart and are not paying
attention, so you miss the NT Boot loader prompt and start booting
Windows, then hard-power-off to save time (can you guess how I know
this?), or if/when Windows crashes. I bet this is going to be a really
F.A.Q. with Wubi.
To reproduce:
1) Either shut Windows down ungracefully, or let it start half-way up, then hard-power-off to kill it
2) Reboot and attempt to go into Ubuntu
Case 1: default startup: You end up at a completely useless and scary BusyBox/(initramfs) prompt with no indication why you are there or what to do about it.
Case 2: recovery mode (verbose) startup: You still end up at a completely useless and scary BusyBox/(initramfs) prompt, but there are log lines on the screen as follows. But they are scary and confusing ("mount: ... failed: Success"? WTH?):
----- cut here (very top of my screen) -----
$LogFile indicates unclean shutdown (0, 0)
Failed to mount '/dev/sda2/: Operation not supported
Mount is denied because NTFS is marked to be in use. Choose one action:
Choice 1: If you have Windows then disconnect the external devices by
clicking on the 'Safely Remove Hardware' icon in the Windows
taskbar then shutdown Windows cleanly.
Choice 2: If you don't have Windows then you can use the 'force' option for
your own responsibility. For example type on the command line:
mount -t ntfs-3g /dev/sda2 /root -o force
Or add the option to the relevant row in the /etc/fstab file:
/dev/sdc1 /mnt/external/ ntfs-3g defaults,force 0 0
mount: Mounting /dev/disk/by_uuid/... on /root failed: Success
mount: Mounting /root on /host failed: invalid argument
ALERT! /host/ubuntu/disks/root.disk does not exist. Dropping to a shell!
BusyBox v1.1.3 ... (ash)
Enter 'help' ....
(initramfs)
----- cut here -----
A diff of the required changes can be found here:
http://launchpadlibrarian.net/14795872/lupin-226622.debdiff
A debdiff of the entire 0.19 upload, including the fix for bug 136682, can be found here:
http://launchpadlibrarian.net/14795586/lupin_0.16-0.19.debdiff
A debdiff of the initramfs-tools changes can be found here:
http://launchpadlibrarian.net/14795732/initramfstools_0.85eubuntu36-0.85eubuntu38.debdiff
Binary package hint: mailscanner
UBUNTU 8.04 HARDY - ALPHA 5 (amd64)
AMD64
MailScanner version 4.58.9-2-ubuntu1 - from the Hardy Alpha 5 Repo's
won't start due to the following error:
Variable "$FIELD_NAME" is not imported at /usr/share/MailScanner/MailScanner/Message.pm line 6368.
Variable "$FIELD_NAME" is not imported at /usr/share/MailScanner/MailScanner/Message.pm line 6371.
Global symbol "$FIELD_NAME" requires explicit package name at /usr/share/MailScanner/MailScanner/Message.pm line 6368.
Global symbol "$FIELD_NAME" requires explicit package name at /usr/share/MailScanner/MailScanner/Message.pm line 6371.
Compilation failed in require at /usr/sbin/MailScanner line 79.
BEGIN failed--compilation aborted at /usr/sbin/MailScanner line 79.
I managed to fix this issue by adding:
6356: our $FIELD_NAME = '[^\x00-\x1f\x7f-\xff :]+:';
line 6355 reads: package Mail::Header;
This was taken from the MailScanner 4.66.5-3 tarball from the
MailScanner website.
This works as a quick-fix for me - but has messed up my apt. Recommend
upgrading repo's to include 4.66.5-3.
Jason
A bug in migration-assistant is preventing French users coming from
Windows XP Home from installing Ubuntu via Wubi.
The problem occurs because migration-assistant is fed non-UTF data. The
fix is to force the locale to C when using expr. There should be
minimal to no possibility for regressions here, given the simplicity of
the change.
TEST CASE:
Install Ubuntu using Wubi from a French copy of Windows XP Home Edition ("Windows XP Édition familiale"). The install will fail at the migration-assistant apply phase ("Import documents and settings" at 88%).
A patch for the uploaded change can be found here:
http://launchpadlibrarian.net/14796991/migration-assistant-0.6.1-0.6.3.debdiff
Binary package hint: mythbuntu-control-centre
When creating a diskless image in the Control Centre, the progress bar
finishes before ltsp-build-client run in the background is finished. The
usually will usually exit the control centre before ltsp-build-client
has finished, resulting in unbootable clients. This leads to lots of
questions in the Mythbuntu forums, which makes me sad.
This upload should have gone in before Archives were hard frozen, but it
didn't make it so we need an SRU now. There's no risk for regression as
it only touches the progress bar code.
Debdiff will be attached soon.
Binary package hint: mythgallery, mythmusic, mythvideo
When /var/lib/mythtv is mounted over NFS, the following MythTV plug-in
packages will fail to configure: mythgallery, mythmusic, mythvideo.
These packages attempt to change ownership/permissions on subdirectories
of /var/lib/mythtv, which is not allowed with a typical server-side NFS
configuration. The failures will occur even if the permissions are
already set correctly on these directories, and even if completely
different directories are configured in the database settings.
Setting up mythgallery (0.21.0-0ubuntu2~gutsy1) ...
chown: changing ownership of `/var/lib/mythtv/pictures': Operation not permitted
dpkg: error processing mythgallery (--configure):
subprocess post-installation script returned error exit status 1
Setting up mythvideo (0.21.0-0ubuntu2~gutsy1) ...
chown: changing ownership of `/var/lib/mythtv/videos': Operation not permitted
dpkg: error processing mythvideo (--configure):
subprocess post-installation script returned error exit status 1
Setting up mythmusic (0.21.0-0ubuntu2~gutsy1) ...
chown: changing ownership of `/var/lib/mythtv/music': Operation not permitted
dpkg: error processing mythmusic (--configure):
subprocess post-installation script returned error exit status 1
This issue is likely to occur for those who use NFS when upgrading from
version 0.20 to 0.21 in gutsy-backports or hardy. A possible workaround
is to temporarily unmount the NFS, then completely reinstall all three
packages. Note that simply retrying the configure after unmounting does
not necessarily work, since those directories would not have been
created during package install (empty mount point directory eclipsed by
NFS mount).
Binary package hint: mythtv
Trying to install mythtv-frontend on powerpc, I get the following
failure:
The following packages have unmet dependencies:
mythtv-frontend: Depends: mythtv-common (= 0.21.0-0ubuntu2) but 0.21.0+fixes16838-0ubuntu1 is to be installed
E: Broken packages
Looking at the repository, I see that there is no powerpc build for the equivalent i386/amd64 builds:
mythtv-frontend_0.21.0+fixes16838-0ubuntu1_amd64.deb 30-Mar-2008 09:03 1.8M
mythtv-frontend_0.21.0+fixes16838-0ubuntu1_i386.deb 30-Mar-2008 08:03 1.8M
Note:
https://edge.launchpad.net/ubuntu/+source/mythtv/0.21.0+fixes16838-0ubuntu1/+build/549956
Seems that it needs building.
----
Bug Impact: All users on the port to powerpc are unable to install MythTV.
This has been addressed in Intrepid by modifying build-dependencies.
Test Case:
1) Install Ubuntu on PPC
2) Attempt to install MythTV
Binary package hint: nautilus
Nautilus 2.20.0, Ubuntu 7.10, AMD64 machine
I have looked and don't see any other bug reports about this yet, though
I have seen queries about it on foums.
I use Nautilus in "Browser" mode normally, rather than in "Spatial"
mode.
I have been trying to make sense of the Nautilus "--geometry" command
line argument. Regrettably the manual doesn't tell you the format of
the argument - A terse "--geometry=GEOMETRY" is all it says.
By experimenting (and a little snippet from forums) I deduced that the
--geometry argument is in the format WxH+HO+VO meaning dialog box width
x height plus horizontal offset (start point) and vertical offset (start
point).
The reason I care about the geometry argument is that I wanted to have a
button (script) that launches 2 Nautilus browser windows in split-window
fashion, to facilitate drag-and-drop operations. Here's my little test
script for it:
[CODE]
#!/bin/sh
nautilus --geometry=1280x512+0+0
nautilus --geometry=1280x512+0+512
[/CODE]
This should open two full-width windows (my screen res is 1280x1024),
one at top of screen and other at lower half of screen. Sometimes it
does. But Nautilus evidently has a memory. If you have clicked the
"maximise" icon near top-right of the previous Nautilus session, then
when the above script launches you get two full-screen windows instead -
so it takes no notice of the geometry argument. You have to un-maximise
each window manually to get what is wanted.
Evidently Nautilus is programmed to ignore command line arguments some
of the time :(
It would be preferable for Nautilus to have a built-in toolbar button
that switched to a dual-window mode. But I had hoped that this simple
script would do the job and am disappointed that Nautilus won't respect
its own command-line argument.
I realise that I could switch to Spatial mode as the default Nautilus
behaviour and then open two windows. But that is not as useful to me as
having two browser windows docked together.
ProblemType: Bug
Architecture: amd64
Date: Sat Oct 20 07:06:40 2007
DistroRelease: Ubuntu 7.10
ExecutablePath: /usr/bin/nautilus
Package: nautilus 1:2.20.0-0ubuntu7
PackageArchitecture: amd64
ProcCmdline: nautilus --no-default-window --sm-client-id default2
ProcCwd: /home/tonywhelan
ProcEnviron:
PATH=/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/games
LANG=en_AU.UTF-8
SHELL=/bin/bash
SourcePackage: nautilus
Uname: Linux HERCULES 2.6.22-14-generic #1 SMP Sun Oct 14 21:45:15 GMT 2007 x86_64 GNU/Linux
The netboot installer's network package retriever fetches Packages files
from the main archive it's given and from security.ubuntu.com (or the
value of apt-setup/security_host if set). However, it only actually
fetches packages from the main archive. This means that if the main
archive does not contain security updates - either because they have not
been copied from -security to -updates yet, or because it's a limited
local mirror that doesn't contain -security rather than
*.archive.ubuntu.com - then netboot installations will fail as soon as
they need to retrieve any security updates.
The fix is simple and safe: fall back to trying to retrieve packages
from the security archive if fetching them from the main archive fails.
This has been fixed in net-retriever 1.21ubuntu2 in intrepid.
The attached patch fixes this bug. Changelog entry:
* Fetch packages from security host if trying to fetch them from the main
host fails (LP: #94398).
TEST CASE: Fetch the netboot mini.iso from the archive
(http://archive.ubuntu.com/ubuntu/dists/hardy-
proposed/main/installer-i386/current/images/netboot/mini.iso; replace
hardy-proposed with hardy to test the failing version), type 'cli
anna/choose_modules=network-console' at the boot: prompt, and confirm
that it gets as far as asking you for an SSH password. (This test case
also covers bug 234486.) I suggest also confirming that a normal netboot
installation (so just type 'cli' at the boot: prompt) works.
Regression potential: netboot installations might end up broken in
different ways, though I think the risk is low. CD installations will be
unaffected.
: jm 325...; ps auxww | grep 6276
jm 6276 0.0 1.0 237004 10976 ? S Oct14 0:47 /usr/lib/notification-daemon/notification-daemon
yep, 237004 KB. looks like a leak.
The process is still running; let me know if you want maps files, etc.
TEST CASE:
1. Run in a terminal in hardy:
$ notify-send lala ; pmap $(pidof notification-daemon) | grep total
$ for i in $(seq 200); do notify-send lala; done
[wait a couple of seconds]
$ pmap $(pidof notification-daemon) | grep total
2. verify that the second total is much higher than the first
3. Instal upgraded notification-daemon
4. run "killall notification-daemon"
5. repeat the steps in 1
6. verify that the number is much smaller
Binary package hint: nvidia-settings
When nvidia-settings is opened, the window has to always be resized to do anything useful.
See attached screenshot.
-----
The bug has been fixed in Intrepid, and NVIDIA notified about the fix. It should hopefully end up in a future (undetermined) NVIDIA driver release.
As this is a voluntary universe package, the risk for regression is very
low.
Test Case:
1) Install nvidia-settings on a machine with an nvidia card
2) Open it up and observe that it opens in a very small window
3) Install the update from -proposed
4) Open nvidia-settings and verify that the window is actually a functional size.
After closing the configuration wizard by clicking on the X on the upper
right window border the background image is displayed. For a better user
experience either the system or the wizard should be restarted.
This is an easy way for a user to screw up their system during initial
configuration. As such, it seems worth clearing it up for Ubuntu 8.04.1.
The patch is a one-liner, and is attached.
This is fixed in oem-config 1.38 in intrepid.
After doing a OEM installation, the /home/oem/Desktop folder is owned by
root, not the "oem" user. The proposed patch is very small and should
pose no risk for regression.
TEST CASE:
Install Ubuntu using the OEM option. When the install reboots, check to see if the Desktop folder is owned by oem or root:
snippets from ls -al
drwxr-xr-x 24 oem oem 4096 2008-03-31 15:41 .
drwxr-xr-x 2 root root 4096 2008-03-31 15:39 Desktop
The proposed patch that has been included in the 1.8.8 ubiquity upload can be found here:
http://launchpadlibrarian.net/14905300/209683-desktop-ownership.diff
If the locale selected during oem-config is different from that selected
during initial installation (for example, if the initial installation
has language English and location New York, but oem-config is given
language English and location London), then the default keymap is not
properly recalculated; it will default to the one selected during
initial installation rather than the one appropriate for the selections
during oem-config.
Fixing this means that oem-config's keymap is not preseedable, reverting
the fix for bug 188492; however, preseeding oem-config is very poorly
supported anyway, so this is not too big a loss in the context of the
whole. For most users it will probably be better to pick a reasonable
default keymap for their locale.
This is fixed in oem-config 1.38 in intrepid. Changelog:
* Don't fail to set the locale if the mirror/country question doesn't
exist.
* Disable support for console-setup preseeding again, as it interferes
with the more important goal of inferring a default keyboard from the
selected locale (LP: #219209).
In kubuntu 8.04 I installed oggconvert using adept. The program once
installed wouldn't load and reported a error about not knowing what the
svg file format was. I fixed it by installing librsvg2-common
Therefore apt-get should list that package as a depends.
TEST CASE:
* Boot a Kubuntu system
* Install oggconvert
* Run oggconvert
* oggconvert will crash with gobject.GError exception
Binary package hint: opencryptoki
Package: opencryptoki
Version: 2.2.5+dfsg-1ubuntu1
Why did you remove /usr/lib/opencryptoki/stdll/* from the package?
$ ls -al debian/tmp/usr/lib/opencryptoki/stdll/
Total 616
drwxr-xr-x 2 dds dds 4096 2008-03-04 14:43 .
drwxr-xr-x 3 dds dds 4096 2008-03-04 14:43 ..
lrwxrwxrwx 1 dds dds 15 2008-03-04 14:43 PKCS11_SW.so -> libpkcs11_sw.so
-rwxr-xr-x 1 dds dds 830 2008-03-04 14:43 libpkcs11_sw.la
lrwxrwxrwx 1 dds dds 21 2008-03-04 14:43 libpkcs11_sw.so -> libpkcs11_sw.so.0.0.0
lrwxrwxrwx 1 dds dds 21 2008-03-04 14:43 libpkcs11_sw.so.0 -> libpkcs11_sw.so.0.0.0
-rwxr-xr-x 1 dds dds 301238 2008-03-04 14:43 libpkcs11_sw.so.0.0.0
-rwxr-xr-x 1 dds dds 852 2008-03-04 14:43 libpkcs11_tpm.la
lrwxrwxrwx 1 dds dds 22 2008-03-04 14:43 libpkcs11_tpm.so -> libpkcs11_tpm.so.0.0.0
lrwxrwxrwx 1 dds dds 22 2008-03-04 14:43 libpkcs11_tpm.so.0 -> libpkcs11_tpm.so.0.0.0
-rwxr-xr-x 1 dds dds 299061 2008-03-04 14:43 libpkcs11_tpm.so.0.0.0
Those are the backends (software and TPM) that opencryptoki requires to do anything useful. Please add them back-in!
The whole point of this package is to be able to use the TPM chip as a PKCS11 smartcard and without that libpkcs11_tpm.so.0.0.0 it's impossible.
Also, please be careful with the contents of /usr/lib/pkcs11; the
symlink "methods -> ../../sbin" is dangerous and the PKCS11_API.so
symlink is not managed by alternatives so it will become contentious
with other PCKS11 libraries. Better to just not set it in the first
place.
Cheers,
- dds
/etc/lsb-release:
DISTRIB_ID=Ubuntu
DISTRIB_RELEASE=8.04
DISTRIB_CODENAME=hardy
DISTRIB_DESCRIPTION="Ubuntu hardy (development branch)"
Binary package hint: opencryptoki
Clone of debian bug #481453. Copying the intro text here:
After installing opencryptoki, libopencryptoki0 I run the command
pkcs11_startup and then tried to start the daemon. The daemon complains
about not finding the config file under
/var/lib/opencryptoki/pk_config_data
This file should be created by pkcs11_startup, so I did take a look at
it. pkcs11_startup expects the libraries for the modules to have the
extension '.so', but the ones installed by libopencryptoki0 do have
'.so.0'. After adding the symlinks manually the nescessary files under
/var/lib/opencryptoki are created and the daemon starts.
Binary package hint: openoffice.org2
I mistakenly submitted this as question 28478. I'm adding it here as
well since I can include the file.
Sorry, but I'm not sure how to explain this problem. In short, the
spreadsheet is not working correctly. I have a sheet with data on dice
rolls for a project my daughter is working on. Column B and C, for
example contain data on dice rolls. Column D is the sum of these two
dice rolls using the sum function. She needed to count the frequencies
of each sum (2, 3, 4...12). Using the frequency function this worked
fine. After upgrading to hardy, however, the data is wrong.
In the attached file, columns P and Q contain the frequencies of the
sums in columns D and G. When opened with Open Office in Hardy, the
values in these columns are not the same as those generated using Open
Office in Windows (sorry, I had to test somewhere). For comparison, the
values in cells P3-P13 should be:
2, 16, 31, 24, 30, 50, 33, 29, 16, 13, 8
For cells Q3-Q13:
5, 11, 24, 17, 35, 36, 33, 39, 32, 13, 7
When I open this file in Hardy I get the following (note that these seem
to change with each save and open)
P3-P13: 30, 16, 31, 24, 33, 50, 2, 29, 16, 13, 8
Q3-Q13: 33, 11, 24, 17, 36, 35, 5, 39, 32, 13, 7
I hope someone can confirm this as it's very odd and very frustrating. I
have confirmed similar behavior on two separate installs of Hardy.
Binary package hint: openoffice.org-calc
I'm using Ubuntu 8.04 hardy and openoffice.org-calc 1:2.4.0-3ubuntu6.
When in any cell I press CTRL+X to cut it then press right cursor button
in keyboard, the selected cell goes to final or any other place, not to
the right one.
Something like that happens with CTRL + X then left.
Binary package hint: openoffice.org-calc
After upgrading to Ubuntu 8.04 I have problems loading existing ods
files. When there are multiple fonts mixed in a singe cell, the
formatting is lost and all characters are displayed in the same font
upon reload of the document.
In the attached document the cell A2 is a mix between Arial and Open
Symbol, the PDF shows it correctly, but it only shows in open symbol on
reload of the document.
ProblemType: Bug
Architecture: amd64
Date: Tue May 6 23:02:20 2008
DistroRelease: Ubuntu 8.04
NonfreeKernelModules: nvidia ath_hal
Package: openoffice.org-core 1:2.4.0-3ubuntu6
PackageArchitecture: amd64
ProcEnviron:
PATH=/usr/lib/openoffice/program:/home/username/bin:/home/username/bin/maven-2.0.6/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/games
LANG=de_DE.UTF-8
SHELL=/bin/bash
SourcePackage: openoffice.org
Uname: Linux 2.6.24-16-rt x86_64
Binary package hint: openoffice.org
Update OpenOffice.org to 2.4.1 which is a bugfix release. There are many
bug fixes in the release including changes in ooo-build. I will be
documenting all the bugs it fixes, that I know of, in a follow-up
comment.
Openoffice does only ship the "ooo-" icons in 48x48 size
which means the launchers look blurry inside the menu or on the panel.
The same icons are available in 16, 22, 24, 32 and scalable sizes, so
all you have to do is add them to the package.
I can attach a tar with the missing sizes if you want.
Impact: Unable to compile the ov51x-jpeg package against the default
hardy kernel.
Fix in Intrepid: The fix is cherry-picked from a new upstream version that has already
been synced from Debian.
Fix: A debdiff is attached to the bug report.
TEST CASE:
1. Install the current ov51x-jpeg-source package.
2. Attempt to compile it against a .24 kernel (default in hardy)
using module-assistant (m-a build ov51x-jpeg). It should fail with a compile error.
3. Install the fixed ov51x-jpeg-source package from -proposed.
4. Attempt the build again, it should finish without error.
Regression potential: Low: simple ifdef for symbols that have been removed in later kernels.
Fix is taken from upstream.
Binary package hint: ov511-source
Hi
I can't compile ov511 for skype usage.
I tried with the instructions from http://gingedas.net/?q=node/170 but it failes with a weird error.
I tried the ov511-source and it failed as well. with the same error
make[1]: Entering directory `/usr/src/linux-headers-2.6.24-7-generic'
CC [M] /home/amitai/Desktop/ov511-2.32/ov511_core.o
/home/amitai/Desktop/ov511-2.32/ov511_core.c:1716: error: unknown field ‘algo_control’ specified in initializer
/home/amitai/Desktop/ov511-2.32/ov511_core.c:1716: warning: initialization from incompatible pointer type
/home/amitai/Desktop/ov511-2.32/ov511_core.c:1726: error: unknown field ‘algo_control’ specified in initializer
/home/amitai/Desktop/ov511-2.32/ov511_core.c:1726: warning: initialization from incompatible pointer type
/home/amitai/Desktop/ov511-2.32/ov511_core.c: In function ‘ov51x_init_isoc’:
/home/amitai/Desktop/ov511-2.32/ov511_core.c:3581: warning: assignment from incompatible pointer type
/home/amitai/Desktop/ov511-2.32/ov511_core.c: At top level:
/home/amitai/Desktop/ov511-2.32/ov511_core.c:5011: error: unknown field ‘hardware’ specified in initializer
/home/amitai/Desktop/ov511-2.32/ov511_core.c:5011: error: ‘VID_HARDWARE_OV511’ undeclared here (not in a function)
Thanks for the help
Amitai
Binary package hint: partman-auto-loop
Bindmounts are used in #230716 (in turn required by #207137). In such
situation parsing /proc/mounts for a given device will return multiple
entries which in turn makes autopartition-loop fail.
TEST CASE: This is a change in support of another bug. While partman-
auto-loop will not fail under these circumstances in hardy as it does
not have the fix for 230716, it will fail in hardy-proposed. Therefor,
the test should be that a Wubi install does not fail at partman-auto-
loop in hardy-proposed.
The proposed patch, as uploaded in partman-auto-loop 0ubuntu14, can be found here:
http://launchpadlibrarian.net/15240795/238701-bind-mounts.diff
Impact: Xfce users cannot access polkit-gnome-authorizations, and so cannot configure
policykit authorizations graphically.
Intrepid fix: the last merge included the change to the .desktop file.
TEST CASE:
* Run an Xfce session, look in the menus for "Authorizations", it shouldn't be
present.
* Install the fixed package
* Look in the menus again, "Authorizations" should be present (Applications > System > Authorizations).
Regression potential: very low, change is a simple change to a .desktop
file.
=== Original report:
Binary package hint: policykit-gnome
Hi,
The Authorizations entry in the System menu is not visible
in Xfce, as it has Only-Show-In: GNOME in its .desktop file.
Cody, I'll fix this bug in the upcoming Intrepid upload, and
then we can start the SRU procedures if that is what you would
like.
Thanks,
James
Binary package hint: postgresql-8.2
New PostgreSQL bug fix releases are about to be released. Once they get
public, I'll post details about changes and announcements.
Binary package hint: python-aptsources
It looks like python-distutils-extra has changed a lot since this
package was last built for Feisty. This should either be updated if
it's still useful or removed.
Binary package hint: update-manager
$ lsb_release -rd
Description: Ubuntu 8.04
Release: 8.04
$ apt-cache policy update-manager
update-manager:
Installed: 1:0.87.24
Candidate: 1:0.87.24
Version table:
*** 1:0.87.24 0
500 http://mirrors.kernel.org hardy/main Packages
100 /var/lib/dpkg/status
appears to be same error reported by me from bug #225913 . Don't know why this crash error keeps coming up. It comes up right after it tries to report a nautlus crash(which fails)
Attached are the dist-upgrade logs
ProblemType: Package
Architecture: i386
Date: Fri May 2 13:21:47 2008
DistroRelease: Ubuntu 8.04
ErrorMessage: ErrorMessage: SystemError in cache.commit(): E:Sub-process /usr/bin/dpkg returned an error code (1)
NonfreeKernelModules: nvidia
Package: update-manager 1:0.87.24
PackageArchitecture: all
SourcePackage: update-manager
Title: package update-manager 1:0.87.24 failed to install/upgrade: ErrorMessage: SystemError in cache.commit(): E:Sub-process /usr/bin/dpkg returned an error code (1)
Uname: Linux 2.6.24-16-generic i686
Binary package hint: python2.5
this lets developers use the same flags that were used for python2.4
builds, getting consistant compilation flags for python2.4 and
python2.5.
Description of the problem:
When holding down enter on a particular track (presumably causing it to start and stop a lot) rhythmbox will either segfault or freeze.
Steps to reproduce:
1. Start
rhythmbox .
2. Go to Music -> Import File...
3. In the window that appears press Ctrl-L so the location bar is visible and in the text area put /usr/share/example-content/fables_01_01_aesop.spx . Click the Open button at the bottom right.
4. Click on fables_01_01_ae... so it is highlighted.
5. Press and hold enter for five or so seconds.
Expected results:
Music to start playing, interface to remain responsive.
Actual results:
No music plays, interface will become unresponsive and stop drawing itself or program will segfault and crash.
How reproducible is the problem?
The problem is reproducible every time.
Version information:
Ubuntu gutsy (development branch)
rhythmbox 0.11.2-0ubuntu2
gstreamer0.10-alsa 0.10.14-1ubuntu3
gstreamer0.10-esd 0.10.6-0ubuntu1
gstreamer0.10-gnomevfs 0.10.14-1ubuntu3
gstreamer0.10-plugins-base 0.10.14-1ubuntu3
gstreamer0.10-plugins-base-apps 0.10.14-1ubuntu3
gstreamer0.10-plugins-good 0.10.6-0ubuntu1
gstreamer0.10-plugins-ugly 0.10.6-0ubuntu2
gstreamer0.10-tools 0.10.14-1ubuntu1
gstreamer0.10-x 0.10.14-1ubuntu3
When files are transferred to an iPod, the 'year' part of the tag is
copied incorrectly onto the device. It appears to use a fragment of the
unix time instead. For example, 2007 becomes 11781, 2008 becomes 12146,
etc.
This issue has been experienced both on Ubuntu 8.04 with rhythmbox
0.11.5-0ubuntu7 and an iPod shuffle 1GB, as well as Ubuntu 7.04 with
rhythmbox 0.10.0 and a 4G iPod 40GB.
Binary package hint: rhythmbox
When I attach my iPod and drag an AAC song to it on Rhythmbox, it
transfers very slowly and with high CPU usage, as if it were transcoding
the file. I confirmed that the file transferred to the iPod was
different from the original stored in my music library folder using md5
sums, and also listened to the song on my computer from the iPod's
drive. The song sounded pretty bad. When the iPod was ejected and
disconnected, the songs wouldn't play at all... the total time indicator
showed the correct time for a moment, but it would instantly go to the
next song without playing at all.
Binary package hint: tomboy
Trying to sync over ssh to a remote server and it's erroring with:-
"Error connecting :(
An error ocurred while connecting to the specified server:
fuse: mountpoint isnot empty
fuse: if you are sure this is safe use the 'nonempty' mount option"
I have had syncing working intermittently, but now it refuses to work. I
have it working to the same server, same user, same key on another PC,
but on this one, it no longer syncs.
15/10/2007 16:42:57 [DEBUG]: Error calling /usr/bin/sshfs
^^ that's all I get in the .tomboy.log
When installing the torque-server package, the installation fails when
the init script is being executed. The installer complains about missing
directories.
When removing the package, apt-get also fails, because the init script
fails. This can only be fixed by editing the torque-server init script,
so it always returns 0.
The installer presents a progress window with a close button when it
begins to partition and copy files. If the user presses this button,
the window will disappear but the process will continue in the
background. The fix is a one-liner to disable the close button entirely
as it's more dangerous to let them cancel the install halfway through
(leaving their system unbootable) than it is to force them to sit
through the rest of it.
TEST CASE:
It should not be possible to close the main install progress window. There should be no close button on the window.
The proposed patch as included in the ubiquity 1.8.9 upload can be found here:
http://launchpadlibrarian.net/14910016/107686-disable-close.patch
Original report:
When installing kubuntu(daily-live, 20070417), i clicked cancel after the beginning of the installation process, the small window with the progress bar closed, but the installation continued in the background.
Additional annoying usability bugs were discovered in ubiquity and were
unable to be fixed in time for 8.04. These proposed changes should
improve the experience of selecting a timezone location on the map for
all users with minor risk of regression.
TEST CASE:
Run ubiquity and navigate to the time zone page. Check to see if the following problems occur:
- the zoom starts too quickly, it's not easy to move the mouse pointer over a country in less than one second which means it zoom on the wrong area while you are still moving.
- the zoom out effect needs a delay too, it's too easy to get the mouse pointer out of the widget while scrolling.
- the zoom is not correctly centered, putting the mouse pointer over the France make it zoom in the north of England for example.
The proposed patch as uploaded as part of ubiquity 1.8.8 can be found here:
http://launchpadlibrarian.net/14906817/203423-zoommap-usability.diff
Original report:
Ubuntu Hardy, alpha 6
When entering the map area to select a city, the map zooms in very
quickly and jerkily. Combined with the high zoom level, this can make it
difficult to see where you are in the map.
I suggest that the map (a) pause for one second after you enter, before
it starts zooming in; (b) takes one second (or maybe even two) to zoom
in; (c) pauses for one second after you exit, before it starts moving
out; and (d) takes one second (or maybe even two) to zoom out.
After doing a OEM installation, the /home/oem/Desktop folder is owned by
root, not the "oem" user. The proposed patch is very small and should
pose no risk for regression.
TEST CASE:
Install Ubuntu using the OEM option. When the install reboots, check to see if the Desktop folder is owned by oem or root:
snippets from ls -al
drwxr-xr-x 24 oem oem 4096 2008-03-31 15:41 .
drwxr-xr-x 2 root root 4096 2008-03-31 15:39 Desktop
The proposed patch that has been included in the 1.8.8 ubiquity upload can be found here:
http://launchpadlibrarian.net/14905300/209683-desktop-ownership.diff
The installer fails to reset the variable that causes it to stay on a
page when an error is triggered when returning to a previous page,
causing it to stay on the previous page that it just returned to despite
any attempts to proceed.
This will occur if a user enters an invalid password, then presses back
to change their partition layout. If they try to proceed forward to the
user setup page again, they will remain on the partitioning page.
This is a one line change that should have no regressions.
A diff of the proposed change as included in the ubiquity 1.8.9 upload
can be found here:
TEST CASE:
1) Run the installer, click "Forward" until you get to Step 5, Who are you?
2) Enter all data correctly except for the "What is the name of this computer?" field - enter a name with an underscore (or presumably any other non-valid character) and press "Forward"
3) The installer will complain that the computer name is invalid.
4) Click "Back" to go back to Step 4, the partition editor
5) Press "Forward" - partition editor rescans the disk and step 4 starts again (this will happen every time you press "Forward", and happens regardless of what settings you choose for the partitions.)
Original report:
There is a sequence of events that can cause the Ubiquity installer in Hardy Herron (8.04) to loop at step 4, the partitioning step.
Binary package hint: ubiquity
Filed on behalf of Colin Watson. From email:
I noticed this in an upgrade:
ubiquity (1.8.9) hardy-proposed; urgency=low
[ Colin Watson ]
* Make use of RAWMINSIZE, RAWCURSIZE, and RAWMAXSIZE variables from
partman-partitioning 59ubuntu2 to set accurate bounds on the resize
widget.
This wasn't meant for hardy-proposed for now, only for intrepid; those
variables won't be present unless you have also backported the
partman-partitioning changes. (Backporting those might be a good idea,
but this would need much more testing than I gave it.)
Binary package hint: ubuntu-vm-builder
By default ubuntu-vm-builder (e.g. in hardy) doesn't set up a locale
other than "C". When I log in to the virtual machine via ssh,
LANG=en_US.UTF-8 is picked up from my environment on the host.
Installing things from within the vm like postgresql then leads to
problems like this:
Setting up postgresql-8.3 (8.3.1-1) ...
perl: warning: Setting locale failed.
perl: warning: Please check that your locale settings:
LANGUAGE = (unset),
LC_ALL = (unset),
LANG = "en_US.UTF-8"
are supported and installed on your system.
perl: warning: Falling back to the standard locale ("C").
Error: The locale requested by the environment is invalid.
Error: could not create default cluster. Please create it manually with
pg_createcluster 8.3 main --start
ubuntu-vm-builder should set up the proper locale, as well as a
preferred timezone and any other similar settings that users are used to
from the standard ubuntu installation methods. Defaults could be set
based on the build environment, with options to override them.
Binary package hint: ubuntu-vm-builder
It looks like /usr/share/ubuntu-vm-builder/vms/qemu-common didn't get
updated fully when ubuntu-vm-builder was updated at some point. It does
not properly handle indicating it has consumed two command-line
arguments in the case of "--templates".
The attached patch seems to correct the problem.
-Andy
Binary package hint: ubuntu-vm-builder
The current ubuntu-vm-builder can not install dbus. The attached patch
the way do_avoid_starting_daemons() to write a policy-rc.d file in the
same way as pbuilder does. This fixes the dbus install issue for me.
Binary package hint: ubuntu-vm-builder
ubuntu-vm-builder assumes that $TMPDIR allows suid and device files.
For the case where a user has mounted both /home and /tmp with -o
nodev,nosuid, this is potentially problematic. While there are a number
of ways to test for this, and verify that the executing user has the
necessary permissions to create the files required by the image, an easy
workaround would be to mount a tmpfs at $WORKINGDIR to ensure that the
target temporary location supports the needed functionality.
Binary package hint: ubuntu-vm-builder
u-v-b 0.4 generates a host file that do not contain the proper IPv6
entries, thus prenventing the VM to run on IPv6 network.
Binary package hint: ubuntu-vm-builder
Implement the set -u in u-v-b for intrepid, I found out that the
"letters" variable was not initialized properly before if could be used.
Binary package hint: ubuntu-vm-builder
$ lsb_release -rd
Description: Ubuntu 8.04
Release: 8.04
$ dpkg --status ubuntu-vm-builder | grep -e Status -e Version
Status: install ok installed
Version: 0.4-0ubuntu0.1
Problem:
Specifying --net on the command line is ignored by the script. Instead, the script auto-calculates a value based on an assumed Class C (24 bit) IP address. For instance, the command line options --ip 10.5.20.57 --mask 255.255.0.0 --net 10.5.0.0 --bcast 10.5.255.255 will produce a response of "setting network to 10.5.20.0" indicating that the --net option/value from the command line was not detected and the NET value was auto-calculated (incorrectly).
Cause:
Everywhere within the script, the Network address is referenced as $NET; however, line 287 sets the value of $NETWORK when the --net command line option is passed, not $NET (as it should).
Solution:
Remove the characters WORK from line 287 of the script.
Binary package hint: ubuntu-vm-builder
Due to the way the security update locked the root account, ssh root
logins are now not working. It uses "chpasswd -l", but the way the
installer does it is using chpasswd to just set root's password to an
invalid one, but without actually locking the account.
To test:
1. Create a vm adding "--ssh-key ~/.ssh/id_rsa.pub" to the command line.
2. Start the vm.
3. ssh root@virtualmachine
If it works, it works.
Here is the test-case for the update-manager SRU it has no "test-the-
bug" section as I have no way to reproduce it.
TEST CASE:
1. install gutsy
2. run "update-manager --proposed"
3. if you are affected by this bug, verify that it works now (if not, verify that there are no regressions)
I tried upgrading to Hardy using "update-manager -d". After checking
for updates, I get a button to upgrade to 8.04, which I pressed. This
brings up the distribution upgrade dialog, showing the various steps its
going to do. This dialog shows the progress bar advancing a bit, and
then it becomes totally unresponsive, not even repainting its own
window.
see also http://ubuntuforums.org/showthread.php?t=676090
The terminal shows
mbp@lithe% sudo update-manager -d
warning: could not initiate dbus
extracting '/tmp/tmpT6yx7G/hardy.tar.gz'
authenticate '/tmp/tmpT6yx7G/hardy.tar.gz' against '/tmp/tmpT6yx7G/hardy.tar.gz.gpg'
zsh: quit (core dumped) sudo update-manager -d
/var/log/dist-upgrade/main.log has
2008-01-28 10:00:50,223 INFO release-upgrader version '0.87.4' started
2008-01-28 10:00:50,762 DEBUG Using 'DistUpgradeViewGtk' view
2008-01-28 10:00:50,948 DEBUG lsb-release: 'gutsy'
2008-01-28 10:00:50,949 DEBUG _pythonSymlinkCheck run
2008-01-28 10:00:52,279 DEBUG checkViewDepends()
Binary package hint: update-manager
TEST CASE:
1. install hardy
2. open a terminal
3. run "sudo update-manager --dist-ugprade"
4. watch failure in the terminal window
5. install update-manager from hardy-propose
6. repeat step 3
7. verify that there is no failure this time
I am already running the Hardy beta, and decided to try updating my system to the latest version running 'kdesu "adept_manager --dist-upgrade-devel"'. After detecting there are no upgrades, update manager crashes:
Traceback (most recent call last):
File "/tmp/kde-root/adept_managerHJMRla.tmp-extract/dist-upgrade.py", line 60, in
app.run()
File "/tmp/kde-root/adept_managerHJMRla.tmp-extract/DistUpgradeController.py", line 1550, in run
self.fullUpgrade()
File "/tmp/kde-root/adept_managerHJMRla.tmp-extract/DistUpgradeController.py", line 1513, in fullUpgrade
if not self.askDistUpgrade():
File "/tmp/kde-root/adept_managerHJMRla.tmp-extract/DistUpgradeController.py", line 818, in askDistUpgrade
self.cache.requiredDownload)
File "/tmp/kde-root/adept_managerHJMRla.tmp-extract/DistUpgradeViewKDE.py", line 631, in confirmChanges
msg = unicode(self.confirmChangesMessage, 'UTF-8')
AttributeError: 'DistUpgradeViewKDE' object has no attribute 'confirmChangesMessage'
Binary package hint: update-manager
TEST CASE:
1. install gutsy server
2. insert a 8.04 alternate server CD
3. mount /cdrom
4. sh /cdrom/cdromupgrade
5. watch it fail
6. umount /cdrom
6. get a new daily alternate 8.04.1 server CD
7. repeat step 3, 4
8. watch is succeed
downloaded 8.04-server-i386 cd image, mount it on a 7.10 server and run sudo /cdrom/cdromupgrade.
It seems that the upgrade does not support ubuntu-server?
--- This is the message given by cdromupgrade:
A unresolvable problem occurred while calculating the upgrade.
This can be caused by:
* Upgrading to a pre-release version of Ubuntu
* Running the current pre-release version of Ubuntu
* Unofficial software packages not provided by Ubuntu
If none of this applies, then please report this bug against the
'update-manager' package and include the files in
/var/log/dist-upgrade/ in the bugreport.
--- this is the result of tail /var/log/dist-upgrade/apt.log:
Can't guess meta-package
Your system does not contain a ubuntu-desktop, kubuntu-desktop,
xubuntu-desktop or edubuntu-desktop package and it was not possible
to detect which version of Ubuntu you are running.
Please install one of the packages above first using synaptic or
apt-get before proceeding.
Log time: 2008-04-27 11:25:26.289873
--- this is the result of tail /var/log/dist-upgrade/main.log:
2008-04-27 11:25:17,452 DEBUG none of the '['ubuntu-desktop', 'kubuntu-desktop',
'edubuntu-desktop', 'xubuntu-desktop', 'ubuntustudio-desktop', 'ichthux-desktop
', 'mythbuntu-desktop', 'gobuntu-desktop']' meta-pkgs installed
2008-04-27 11:25:17,472 ERROR Dist-upgrade failed: 'Can't upgrade required meta-
packages'
Binary package hint: update-manager
When a package fails to install, the upgrader logs are not automatically
installed.
There is a typo in the directory name:
- LOGDIR="/var/log/dist-upgrader/"
+ LOGDIR="/var/log/dist-upgrade/"
Binary package hint: update-manager
TEST CASE:
1. install gutsy
2. switch to Ukrainian as default langauge
3. run update-manager click on "upgrade"
4. watch it crash
5. run update-manager --proposed
6. confirm that it does not crash and show the expected "upgrade" dialog
Hello!
I'm using "sudo do-release-upgrade -m desktop" for upgrading. But I get error:
Hit http://ubuntu.intergenia.de hardy-backports/main Sources
Hit http://ubuntu.intergenia.de hardy-backports/multiverse Sources
Hit http://ubuntu.intergenia.de hardy-backports/restricted Sources
Done downloading
Перевірка менеджера пакунків
Reading package lists: Done
Reading state information: Done
Reading state information: Done
Reading state information: Done
Прийняття змін...
Виникла невиправна помилка
Please report this as a bug and include the files
/var/log/dist-upgrade/main.log and /var/log/dist-upgrade/apt.log in
your report. The upgrade aborts now.
Your original sources.list was saved in
/etc/apt/sources.list.distUpgrade.
Traceback (most recent call last):
File "/tmp/tmpXilql0/hardy", line 60, in
app.run()
File "/tmp/tmpXilql0/DistUpgradeController.py", line 1550, in run
self.fullUpgrade()
File "/tmp/tmpXilql0/DistUpgradeController.py", line 1513, in
fullUpgrade
if not self.askDistUpgrade():
File "/tmp/tmpXilql0/DistUpgradeController.py", line 818, in
askDistUpgrade
self.cache.requiredDownload)
File "/tmp/tmpXilql0/DistUpgradeViewText.py", line 167, in
confirmChanges
DistUpgradeView.confirmChanges(self, summary, changes, downloadSize,
actions)
File "/tmp/tmpXilql0/DistUpgradeView.py", line 200, in confirmChanges
pkgs_remove) % pkgs_remove
TypeError: not all arguments converted during string formatting
I include compressed logfiles: /var/log/dist-upgrade/apt.log & /var/log
/dist-upgrade/main.log
Thanks.
P.S. Sorry for my English, I'm Ukrainian
Binary package hint: update-manager-core
The official Landscape repositories are considered third party ones and
are disabled during a distribution upgrade.
TEST CASE:
0. install gutsy
1. add "deb http://landscape.canonical.com/packages/gutsy ./" to /etc/apt/sources.list
2. run update-manager and click on "upgrade"
3. verify that update-manager tells you that third party repository have been disabled, check /etc/apt/sources.list when the dialog is displayed and confirm that a "#" is in front of the landscape.canonical.com line
4. cancel the upgrade
5. run "update-manager --proposed" and click on "upgrade"
6. verify that no message about third party sources is displayed and check that /etc/apt/sources.list contains a "http://landscape.canonical.com/packages/hardy" line instead
7. ignore the error about the authentication failure, that is because this test here has not added the required repository key
Their format is:
deb http://landscape.canonical.com/packages/ ./
We currently support all distros from dapper to hardy, so the possibilities for sources.list entries are:
deb http://landscape.canonical.com/packages/dapper ./
deb http://landscape.canonical.com/packages/edgy ./
deb http://landscape.canonical.com/packages/feisty ./
deb http://landscape.canonical.com/packages/gutsy ./
deb http://landscape.canonical.com/packages/hardy ./
So, after a gutsy -> hardy upgrade, this line:
deb http://landscape.canonical.com/packages/gutsy ./
should be changed to:
deb http://landscape.canonical.com/packages/hardy ./
and the packages from that repo upgraded too as usual so we end up with
landscape related packages built for hardy.
Binary package hint: vino
TESTCASE:
that's a stable update, just make sure that the software is still working:
- try connection to the vnc server using an another computer
- set a password and make sure you can connect using this one
- use the option to ask confirmation and verify the question is correctly displayed when the client connect to the server
the news entry:
Vino 2.22.2
============
Fixes
+ Bug #534262 – cyclic clipboard propagation.
Binary package hint: xorg
"I sent this patch in to the mailing list last Monday. dexconf wasn't
detecting the ps3fb correctly on hardy. Because of this it misses out a
couple of options that are meant to be explicitly set for PS3. It's a
very very small change.
https://lists.ubuntu.com/archives/ubuntu-x/2008-May/000157.html
From: Dan Munckton
The output of cat /proc/fb on 2.6.24 is "0 ps3fb". Dexconf was searching for "PS3" and was therefore failing to detect the platform correctly.
---
debian/local/dexconf | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/debian/local/dexconf b/debian/local/dexconf
index 50a2f4e..596a4e2 100644
--- a/debian/local/dexconf
+++ b/debian/local/dexconf
@@ -281,7 +281,7 @@ SECTION
if [ -n "$DEVICE_DRIVER" ]; then
printf "\tDriver\t\t\"$DEVICE_DRIVER\"\n" >&4
fi
-PS3_FB=$(grep PS3 /proc/fb 2>/dev/null || true)
+PS3_FB=$(grep -i PS3 /proc/fb 2>/dev/null || true)
if [ -n "$PS3_FB" ]; then
printf "\tOption\t\t\"ShadowFB\"\t\t\"false\"\n" >&4
fi
Binary package hint: xserver-xorg-video-amd
VGA Controller is an AMD Geode LX (should be using installed xserver-
xorg-video-amd 2.7.7.7-1)
Not sure what video driver is being used, please include instructions to
check this in Hardy.
Maximum resolution is 800x600
Binary package hint: xserver-xorg-core
To complete auto-configuration of the -geode driver in X.org, "geode"
must be either added to hw/xfree86/common/xf86AutoConfig.c or made to
replace the deprecated "amd". An ideal compromise would perhaps be to
put "geode" right before "amd" in the list, keeping "amd" in the list as
a transitional measure, for those upgrading from older releases to
Hardy.
I haven't found any bug exactly like this although some seem to be very similar.
Using an up-to-date hardy installation I've had this problem plaguing me
for some time now. Every now and then during VT-switch, everything turns
black, the hard drive led keeps blinking and nothing works, not even the
magic sysrq combinations. Only thing left is to shutdown the laptop.
At first, I thought this was related to suspend, cause that's normally
when I see it, but after seeing the same problem when trying to reboot
or log out I noticed it had to do with any form of VT-switching. It
seems to happen about one out of three times.
It's been very hard to acquire any relevant debugging information. All I
have is an xorg.log. I hope it helps.
The graphics chip is an intel 855GM, I'm using the intel driver.
Binary package hint: xserver-xorg-core
To complete auto-configuration of the -geode driver in X.org, "geode"
must be either added to hw/xfree86/common/xf86AutoConfig.c or made to
replace the deprecated "amd". An ideal compromise would perhaps be to
put "geode" right before "amd" in the list, keeping "amd" in the list as
a transitional measure, for those upgrading from older releases to
Hardy.
Binary package hint: xubuntu-docs
The translations of the xubuntu-docs material done at
https://translations.launchpad.net/ubuntu/hardy/+source/xubuntu-docs are
not imported into the repository/source package.
my sister has been using her shiny new ubuntu system
i'm not sure what she does with it exactly, but after a while she gets
"too many open files!" messages from nautilus
i tracked the problem down. pitti made an ubuntu vendor patch to esound
to have it check for non-existent /usr/bin/esd and bypass a bunch of
spawning code if it's not there.
unfortunately he also had it bypass the code to close the pipe file
descriptors that had already been opened by this point.....
it's a pretty simple fix. will attach.
TEST CASE:
1) ensure that "Enable software sound mixing (ESD)" checkbox is enabled
in "Sound Preferences" (this is the default)
2) ensure that the 'esound' package is not installed (this is also the
default)
3) ensure "preview sound files" is enabled in nautilus preferences (also
the default)
3) open a terminal
4) $ cd /proc/`pidof nautilus`/fd
5) 'ls' and notice a small number of 'pipe' entries
6) open a folder with a bunch of ogg vorbis files in it and move the
mouse around a lot, moving over top of various sound files (as if you
would when you are trying to preview them).
7) 'ls' again and notice that the number of 'pipe' entries has increased
8) notice that once nautilus has 1024 open file descriptors further
attempts to open folders (or anything else) in nautilus will fail.
Install gnome-schedule.
Run it.
Help->About, try to close it --> dialog frozen
Help->Manual --> missing: Unable to find the help files in either /usr/share/gnome/help/ or /usr/share/gnome/help/. Please check your installation
--
Gutsy SRU Notes:
1. The Help About menu and Help topics are not available to anyone using
Gutsy. This fix is not critical, but it's small, should not cause
regressions, and it fixes a problem in the release.
2. The python method of calling up the Manual was modified.
The About Box now closes properly because the dialog is destroyed after it is opened.
Single package modified:
gnome-schedule 1.0.0-2 --> gnome-schedule 1.0.0-2ubuntu1
3. See the Gutsy patch attached to this report. This patches against
Gutsy version 1.0.0-2.
4. TEST CASE:
$ sudo apt-get install gnome-schedule
$ gnome-schedule
Follow the steps in the report: Try Help->About, and try to close the
about box. It freezes. Try Help->Manual. There is an error in
bringing up the manual.
Expected results with the fix: About Box should close, the manual should
come up without errors.
5. There should be no regressions with a fix like this unless the user
has somehow modified their package to a point where this fix would be
unsuitable (e.g. for some reason moving the help files).
Hi!
Can not print anything with my USB HP DeskJet D1360 in latest Gutsy. Used to print OK recently.
Job appears in queue as "Stopped: job-stopped".
/var/log/cups/error_log contains multiple errors (see attachment).
dmesg does not contain any relevant info.
I am NOT running apparmor (no related packages installed).
I am able to add/remove printers OK. Printer auto-installation works OK.
Reinstalling hpijs, cupsys does not work.
hpijs 2.7.7+2.7.7.dfsg.1-0ubuntu4
cupsys-* 1.3.2-1ubuntu5
Might be related to:
https://bugs.launchpad.net/ubuntu/+source/cupsys/+bug/131470
SRU request below: https://bugs.edge.launchpad.net/ubuntu/+source
/ubuntu-meta/+bug/149511/comments/19
TEST CASE:
Update; No printer needed for the test. Simply skip the printing test or
let the queue point into a file (Set "FileDevice yes" in
/etc/cups/cupsd.conf, restart CUPS, use URI "file:/tmp/out.prn"). You
see only with the "sudo dpkg -r hplip" and "sudo dpkg -r hplip" whether
everything behaves correctly.
On a Gutsy with a print queue for an HP inkjet or a PCL 5c/e laser
printer (with HPIJS as driver) do
sudo dpkg -r hplip
and then try to print. You will not be able to do so, even after calling
system-config-printer and modifying the print queue so that it does not
use the "hp" CUPS backend.
It is due to libraries in th hplip package which are needed by the HPIJS
driver. You see it by running
ldd /usr/bin/hpijs | grep libhp
To get your system back to a working state after doing this test, do
sudo apt-get install hplip
Binary package hint: postgresql-8.2
New PostgreSQL bug fix releases are about to be released. Once they get
public, I'll post details about changes and announcements.
Binary package hint: postgresql-8.2
New PostgreSQL bug fix releases are about to be released. Once they get
public, I'll post details about changes and announcements.
libsasl2 in Dapper is configured to use /dev/random. As systems can run
out of entropy, this can cause applications that use libsasl2 to block.
Subsequent versions of libsasl2 (since 2.1.22-0~pre04) have been
configured to use /dev/urandom instead.
TEST CASE:
Dapper:
apollock@apollock:~$ strings /usr/lib/libsasl2.so | grep random
/dev/random
Hardy:
apollock@procrastination:~$ strings /usr/lib/libsasl2.so | grep random
/dev/urandom
>From http://bugs.php.net/bug.php?id=37790
This is really annoying and prevents us from developing Webservices with Ubuntu 6.06 LTS!
Description:
------------
When upgrading to Kubuntu or Ubuntu 6.06 (or even previous
Dapper Dragon pre-releases), SOAP extension in PHP 5.1.4 stops
functioning the way it used to regarding the XML responses.
Associative arrays no longer return a XML response that
specifies type.
Reproduce code:
---------------
class MyService {
function getLists() {
$out = array(
array(
'id' => 12,
'name' => 'My New List'
)
);
return $out;
}
}
In a client page:
$result = $client->getLists();
$result will be stdClass stuff on 'broken' setup, with dapper.
Expected result:
----------------
Snippet of XML response from a $client->getLists() call:
id12
name
My New
List
Result variable should be:
Array
(
[0] => Array
(
[id] => 12
[name] => My New List
)
)
Actual result:
--------------
Actual result, missing type attributes:
id12nameMy New List
Actual $result variable on dapper:
stdClass Object
(
[item] => stdClass Object
(
[item] => Array
(
[0] => stdClass Object
(
[key] => id
[value] => 20
)
[1] => stdClass Object
(
[key] => name
[value] => Test Site List
)
)
)
)
Binary package hint: postgresql-8.2
New PostgreSQL bug fix releases are about to be released. Once they get
public, I'll post details about changes and announcements.
As done previously (see #80569), here is another upstream update for
sun-java5 for dapper.
The list of all changes can be found at
http://java.sun.com/j2se/1.5.0/ReleaseNotes.html
The debdiff for the packaging is attached.
