Bug#388568: Acknowledgement (xserver-xorg: postinst can hang)
On Wed, Oct 04 2006, at 04:26 +0200, Wolfgang Pfeiffer wrote:
> Hi All
>
> On Tue, Oct 03 2006, at 22:12 +0200, Denis Barbier wrote:
> > On Tue, Oct 03, 2006 at 06:44:53PM +0200, Wolfgang Pfeiffer wrote:
> > > > The easiest solution is to remove the debug_echo. If it is needed for some
> > > > reason, a construct like this should be made (note that this happens in a set
> > > > +e block):
> > >
> > > > db_input "$PRIORITY" "$TEMPLATE"
> > > > retval=$?
> > > > if [ $retval -ne 0 ] ; then debug_echo "v_s_d_i/db_input $PRIORITY $TEMPLATE"
> > > > ; fi
> > > > if [ $? -eq 30 ]; then
> > >
> > > Thanks for finding the solution ... :)
> > >
> > > But could you please post the diff for the postinst script (provided
> > > the code above is about this script) .. the snippets so far posted
> > > here do not seem being self-explaining very much to me ... :)
> >
> > db_input "$PRIORITY" "$TEMPLATE"
> > if [ $? -eq 30 ]; then
> >
> > The patch to fix 389506, 388568 and 386191 is attached.
>
>
> Yes. Seems the patch works ... Thanks a lot, Denis. I upgraded nearly
> the entire unstable system on that alu-PowerBook 5.8, including
> xserver-xorg* packages (I just held back openssh, IIRC),
I think 4 or 5 more packages are on hold via aptitude here; didn't
check that right now ...
> and it seems, even after rebooting the machine, everything's fine
> here ..
No quite so any more ... please see below ...
>
>
> Jan, as to #379480:
>
> ... perhaps the patch helps even for what was described at
> http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=379480
>
No. Seems to be wrong now, at least on this PowerBook5,8:
"UseFBDev" still has to be disabled, as it seems now: I rebooted my
machine a few minutes ago, and starting X from the console
("etc/init.d/kdm start") rendered my machine into an unresponsive an
useless status. I didn't dare to put it into sleep mode, all I tried
then was to ssh into that crashed system with another computer, to
shut it down cleanly ...
To get an ssh connection to the crashed system was difficult: To log
in and then to 'su' to root took much longer than usually. Shutting
down x11-common/xserver-xorg was not possible anymore, as it seems,
and IIRC ...
Perhaps an interesting detail: Having an ssh connection to the crashed
system made it difficult to "shutdown -r now" the system: even after 2
or 3 minutes the system didn't seem to shut down ... a "top" then
revealed that pbbuttonsd also periodically (alternating with "xorg",
IIRC) took about 98% of the CPU. And shortly after stopping pbbuttonsd
got me a reboot of the machine with the system being crashed ...
>
> Even the frame-buffer in xorg.conf is enabled ... :)
Not any more now .... :)
But with it being disabled X seems to work here ..
HTH
Best Regards
Wolfgang
--
Wolfgang Pfeiffer: /ICQ: 286585973/ + + + /AIM: crashinglinux/
http://profiles.yahoo.com/wolfgangpfeiffer
Key ID: E3037113
http://keyserver.mine.nu/pks/lookup?search=0xE3037113&fingerprint=on
Reply to: