#778361#5
Date:
2015-02-14 01:15:32 UTC
From:
To:
Partitioning froze up during LVM creation in the following scenario.

Using a virtualbox VM I attempted an installation of Debian sid amd64.
For this I used the installer in a live CD I built with a dev copy of
live-build (I have been working on adding EFI support to live-build).
This is a copy of the proper installer (not the "live" installer), so
somewhat essentially a jessie install disc equivalent, which uses the
'current' version (20150107) of d-i. All install components uses are
unmodified copies taken from the archives and work correctly in other
scenarios.

The VM was set to use EFI.

I firstly initialised the two virtual HDDs with GPT, then ran the
installation in expert mode.

Selecting manual partitioning, I created the following layout:

sda1: ESP
sda2: /boot (ext4)
sda3: [LUKS] <unused at this point>
sdb1: [LUKS] /home (ext4)

I then attempted to create an LVM group within the sda3 LUKS container.

I was prompted for a group name, for which I supplied 'vg1'.

It is currently now sat there with an unmoving progress bar at about
~80% completion, seemingly stuck. The progress bar states 'starting up
the partitioner', and underneath it states 'detecting file systems...'.

It's been stuck like this for at least 10mins now, so I'm giving up on it.

#778361#10
Date:
2015-02-14 03:23:37 UTC
From:
To:
...

Ugh. Sounds like a crash in parted, maybe. If it's repeatable, can you
get the syslog copied off please?

#778361#15
Date:
2015-02-20 02:58:00 UTC
From:
To:
from scratch and repeated the same procedure, resulting in the same
crash.

I struggled a little figuring out how to get at the logs at first (Ctrl
+ Alt + F2 switched my host instead of the VM, so had to repeat in
text-expert mode, then I was trying Ctrl + C in tty4, which some docs
had pointed me to, to try and get a usable prompt, to no avail, before
trying a different tty), but I got there in the end. Apologies for any
mess at the end of the syslog from my messing around figuring this out.
Log files are attached.

To get them out of the VM I had added an additional small ext4 formatted
virtual HDD beforehand on which to copy them to (which I then loaded in
another VM which has a shared folder, allowing me to copy them to the
host). This may show up as /dev/sdc in the logs.

Taking a quick look at the syslog, I see this towards the end:
Feb 20 02:00:11 kernel: [  276.039512] parted_server[14004]: segfault at
8 ip 000000000040b4b1 sp 00007fff44b01430 error 4 in
parted_server[400000+12000]

Everything after that point in the syslog I believe is just from me
figuring out how to get at the logs.