[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#420382: installation-reports: Install hangs when 30GB ipod connected via USB

reassign 420382 partman-base

On Sunday 22 April 2007 08:14, Christian Perrier wrote:
> > During "Starting up the partitioner.."
> >        "Scanning disks..."
> > I get the message
> >
> > "Warning!
> >  Device /dev/scsi/host2/bus0/target0/lun0/disc has logical sector
> > size of 2048.  Not all parts of GNU parted support this at the
> > moment, and the working code is highly experimental."
> That message obviously seem to come from gparted. How is it displayed
> on screen?

Not gparted, but libparted.

I read this message as a warning that devices with a sector size of 2048 
are not really supported by libparted and that if you continue you run a 
risk of losing the data on the device and possibly even the device (if it 
cannot be formatted again). Personally I probably would not continue even 
if the partitioner did not hang.

AFAIK, normal sector sizes are 512 and 1024.

The only people who can tell if this warning is still valid or not and 
what the status of support for 2048 size sectors is, are the maintainers 
(and possibly only the upstream maintainers) of libparted.

> My understanding is that this failure is badly handled by partman. But
> this needs to be confirmed.

It is correct that partman should not hang, and the fact that it does is 
probably an indication that it really is not supported. AFAIK such 
messages from partman are normally fairly well handled.

Is the installer hanging completely, or can you still switch to VT2?
If you can, it would be interesting if you could add a line "set -x" near 
the top of /lib/partman/definitions.sh _before_ partman is run (using 
nano as editor), and send us /var/log/syslog (gzipped) after the hang. 
That would give us the information on where exactly the warning is 
displayed and the hang occurs. The option "Save debug logs" in the main 
menu may help as well.

As you are the first person to report this issue, I don't think we'll add 
it in the errata just yet, but we'll keep the issue in mind.


Reply to: