You have been subscribed to a public bug by Kees Cook (kees):
Hi there, this is the same bug I experienced several days ago and it is
NOT A DUPLICATE !!!!!!!!!!!!
The error message on apport is:
console-kit-daemon crashed with SIGSEGV in g_str_hash()
I just chose virtualbox as an option, since I do not want this bug
report to be autodeleted also.
Due to a very bad policy in the bug reporting system, I file this bug
with a different name. Since any other way, it gets deleted and the bug
squad does not even look at it. Now I know they are doing a good job on
average, but due to the testing season they are overworked and it is not
necessarily very bright to autodelete any bug with an old title and not
dedicate some time to writing a decent manual for novices about apport-
gtk and gdb. I already reported a similar crash a few days ago, but the
attachments were deleted and it is indicated as fix commited, confirmed
etc. (e.g. #204213 ; #259639, I also reported a Desktop crash through
bugzilla, they replied with a standard message that it was fixed.). Let
me assure you: it is NOT FIXED. BTW, was this developed by Red Hat? I
traced back error reports months ago and it is still not fixed. I think
this should receive absolute priority, you can not just build further on
a corrupt program, unless you want to end up with a UNIX Vista version.
Anyway, this bug happened about an hour ago. I was enjoying the alfa 5
which worked flawlessly for more than a day, I could not get VBox 2.0 to
work though, which is weird, since before I upgraded from alfa 4 to alfa
5, the virtualbox-2.0 worked. So, I needed to install some new stuff,
also isdnvboxclient. So vbox3 was replaced by vbox and so on. Anyway,
when I tried to restart VirtualBox, (I used sudo VirtualBox), it froze
up. I had earlier freeze ups with Virtual Box, but I was able to sudo
kill -9 'pidof virtualbox' and regain functionality. Not this time,
everything froze and alt-ctrl-F2, alt-ctrl-backspace and alt-ctrl-del
were unresponsive, so I was down to a soft reboot. I have reported this
pattern in several earlier bug reports. Somehow I think all of these are
related. I will describe again: It is a three stage phenomenon. If
abrupted early, by shutting down the application or killing in terminal,
the xsession is fully restored, without a problem. If you wait a little
longer, you will encounter a freeze up, after which you have to restore
your system with alt-ctrl-F2 or alt-ctrl-backspace and go from there. If
you wait longer still it will freeze your Desktop permanently and none
of the fast keys will work any longer. The last thing to go out is your
cursor (in my case directed solely by the touchpad). You could call this
stage three B. I have had countless repeats of this pattern. Sometimes I
had to reinstall the system, because I made things worse in the
terminal, being a UNIX novice without ICT background. I mind you that it
is not so that you can enter this process only in stage one, you can
also directly jump to stage two or three, due to reasons I cannot
understand. I don't always get an error report on reboot. Sometimes I
do, sometimes I don't. I have not discovered a pattern in it yet. So I
keep reporting different bugs, but I suspect the core issue could be
something else occasionally, not always, some bugs will still be solved,
but I don't know which ones. I just keep on reading posts about how
hardy is the worst ubuntu release ever. Could it be that something was
implemented back then, high up the three of core applications, which was
not properly tested? Maybe you should go from there. What I have so far:
something is wrong with the 'OS 2'. I reported this in the memtest86
bug, but I have gotten no response yet. I stated there that I suspect
Ubuntu has the ability to modify the OS 2. This is because I had a
problem with my Acer laptop, the single machine I have or have had for
the last two years. I bought it on eBay, through a seller who is no
longer listed. He pumped up the RAM from 1 GB, to 4 GB. When I wanted to
upgrade the system from Windows Vista Home Basic, to Windows Vista
Ultimate, I ran into a problem. The 32-bit version of Vista only
supports 3 GB max. At least back then, maybe still. Anyway, I left this
behind. The Acer manual does allow for 4 gigs of RAM. But this is meant
for the 64-bit processors only I guess. So I suspect the eBay seller had
to adept the OS 2 thing somehow. I returned the (refurbished) eBay PC to
Acer in Germany. I had warranty still. They said it was working fine,
but the memory was not 'Acer'. I called Microsoft, since I had a starter
support contract on my Windows Vista Ultimate upgrade DVD. It comes with
both a 32- and 64-bit disc, within an oversized plastic box to higher
the mailing costs, but since I only have a single processor I don't see
the point. I am sure MS does, price discrimination. So when I called MS
support, they advised me to use a little program ... called memtest86. I
still have the exact version on a backup drive if you want. Anyway, they
confirmed that the inability to upgrade to Ultimate was caused by the
memory. Back then (about two years ago) the memory test also failed,
but later then it does under Ubuntu. In ubuntu it crashes nearly
immediately (memory check stage two?). Anyway, the memory itself is just
fine, since virtualbox-2.0 let's me dedicate memory over 3 giga bytes. I
also remember that two years ago I tryed out both memory chips and they
worked both fine if installed as single. It could also be a memory
socket issue, but as the problem behaves differently under Windows, than
it does under ubuntu, I think it must be indeed something related to OS
2 and not to the hardware. Therefor again: can you check that all
affected users use OS2? Does Macbook have something else? Now I do
realize that what I have depicted here is more indicative of ubuntu not
being able to alter OS 2, but rather to restore the defaults from when
the PC was delivered to me, that is with Windows Vista Home Basic
preinstalled. Now I have noticed some different clues that contradict
this hypothesis: the ubuntu live CD's do not respond normally any
longer, after I had to reinstall. I actually did call in (anonymous)
professional support for this, since I was without any options. Now the
funny thing was that the guy urged me to try and boot the live CD in
safe graphics mode. This made sense, since I had problems with starting
the Display, and xinit, or startx failed to find a screen. However, when
I had previously killed my xsession, the root terminal responded that an
application was already using the screen, another indication of a
permissions issue 'typo console kit'. I could not do this, start in safe
graphics mode, since I lacked the option under the F4 menu. We thought
at the time that this was due to the fact that I used a hardy beta
release on DVD-RW, that I had somehow not updated, as well as an alfa-3
8.10, and maybe the test releases did not have this capability, although
I was previously able to boot from this DVD-RW/CDROM respectively. When
I tried the 7.10 and 7.04 official live CD's they did not work either.
The guy said I was probably experiencing hardware problems instead of
software problems. I did manage to somehow to use dpkg-reconfigure
xserver-xorg as an alternative to xfix. Now I tried it out afterwards,
and the CD's had this capability again. Now I mind you that none of the
live CD's/DVD's have the capability to use safe graphics mode with the
boot from first hard disk option; what is different here is that the F4
menu also lacked the safe graphics mode option in the install and try
without changes option. This was restored when I reinstalled the system
without formatting, thus preserving my personal files. So here we are:
the live CDROM did not change. It does not use the hard disk, so the
problem is not ubuntu OS. Somehow the OS 2 must have changed. Is this an
open source program? Is it written in an obsolete programming language?
Now a last clue to close my tirade. I had previously run sudo dpkg-
reconfigure -a, and I had accidentally denied the option 'let xserver
use the core video frame blablabla' or something option. I always run
this command with priorities=low. This means: display every little
detail. Now I noticed that VirtualBox (2.0) seemed to somehow freeze up
my system (see above). I had read in the community HOWTO about the
taskset -c option for dual core CPU's. I had therefor previously applied
the CPU-clock set by root as 'yes'. But this may be used in security
attacks. The sudo ufw enable command made sense to me because of this,
and I went for it. Now after I fully set up the ufw, I still got the
warning: 'Warning: '/' is world and group writeable !' Now being drunk
and an absolute UNIX newbie, I could not come up with anything smarter
than: '$ sudo chmod -cR 700 .' in the root terminal. An obvious double
mistake with a predictable outcome: reinstall #85. It did however give
me the final piece of the puzzle. As a newbie I nearly always run every
command with sudo. This prevents the annoying: root privileges are
required messages. I don't have a clue about when you actually need to
call a command as root. Therefore I am more susceptible to the
consolekit bugs than more experienced users. However, only a small
portion of the time this bug message can be reported through apport.
When I have to reinstall, any bug information is lost. When I have to
reboot, not always do I get a bug report icon. On many another occasion
I get an apport bug message about another bug. All of this makes sense.
Something high up the hierarchy makes the system misbehave, with a
variety of symptoms. That something must be OS2. The consolekit daemon
is only the ideal medium to expose this inherent flaw, are there is a
mutual incompatibility issue. Please delve into this. If you need more
info from me, the safe bet is a secure shell session, I do not mind. I
am posting this bug from the alfa-3 live CD.
I will now post you my previous bug. It has a possibly helpful slide
show attached about the symptomatology of the previous bug. Then again,
these symptoms are all very different from what I have experienced now.
or from what I did experience before. I have all crash reports backed up
from the last two weeks, since reinstalling without repartitioning
deletes these. Cheers, Thomas.
Later, or look for it yourself. My mobile broadband is too slow.
ProblemType: Crash
Architecture: i386
DistroRelease: Ubuntu 8.10
ExecutablePath: /usr/sbin/console-kit-daemon
Package: consolekit 0.2.10-1ubuntu3
ProcAttrCurrent: unconfined
ProcCmdline: /usr/sbin/console-kit-daemon
ProcEnviron:
Signal: 11
SourcePackage: consolekit
StacktraceTop:
g_str_hash () from /usr/lib/libglib-2.0.so.0
?? () from /usr/lib/libglib-2.0.so.0
?? ()
?? ()
?? ()
Title: console-kit-daemon crashed with SIGSEGV in g_str_hash()
Uname: Linux 2.6.27-2-generic i686
UserGroups:
** Affects: consolekit (Ubuntu)
Importance: Undecided
Status: Incomplete
** Tags: apport-crash apport-failed-retrace
--
console-kit-daemon crashed with SIGSEGV in g_str_hash() (not #259639)
https://bugs.edge.launchpad.net/bugs/267328
You received this bug notification because you are a member of Ubuntu Bugs,
which is a direct subscriber.
--
ubuntu-bugs mailing list
[email protected]
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
