[uae] Important: Is using a hardfile partition beyond the first 4 GB of a MBR partition safe?
- From: Martin Steigerwald <Martin@xxxxxxxxxxxx>
- To: uae@xxxxxxxxxxxxx
- Date: Sat, 31 Jul 2004 17:29:48 +0200
Hello,
Is using a hardfile partition beyond the first 4 GB of a MBR partition safe? I
always just thought about the size of the hardfile / partition, not the
location on the disk.
Today I just used such a hardfile partition (partition 9, /dev/hda9) and
copied about 80MB of YAM e-mail files to it.
Then I wanted to get me e-mail with KMail again and it complained it cannot
start the "pop" process. This did not go away even after uninstalling and
re-installing KMail. Thus I rebooted.
After reboot the X11 server refused to come up. It couldn't find some basic
libraries. On accessing various files I could a "unknown error 990".
That's not good I thought, inserted a Knoppix CD and run "xfs_check" on the
partition. Indead it found problems. Well what should I do? So I run
"xfs_repair". This seemed to have fixed most of the problems at the costs of
loosing track about 100 files now in "lost+found". The output of the tool
looked quite scary to me.
Anyway the system now works quite OK again and I think I re-installed packages
whose files might have been affected. I am not 100% sure tough, but at least
I can use the KDE desktop again.
I am now puzzled about the cause of the problem. The harddisk is just a few
weeks old and I considered XFS to be rather stable.
Now could it be that UAE wrote onto my XFS partition (partition 5, /dev/hda5)?
I thought not, cause it just has access to the device files of partition 9
(/dev/hda9) and 10 (/dev/hda10). But before I do any further action on the
hardfile partition I want to be absolutely sure. Right now I will remove the
hardfile entries relating to those 2 partitions completely from the UAE
config.
It just happened at the same time that symptoms of a corrupt root filesystem
appeared just after or even while copying the YAM files within UAE. Maybe
that's more than a co-incidence.
Otherwise I really have no idea what could have caused the corruption on the
XFS filesystem.
Here is my partitioning data:
(parted) print
Plattengeometrie für /dev/hda: 0.000-57231,562 Megabytes
Disk-Label-Typ: msdos
Minor Start End Type Dateisystem Flags
1 0,031 2863,146 primär boot
2 2863,147 6675,446 primär
3 6675,447 9538,593 primär
4 9538,594 57231,562 erweitert
5 9538,625 19077,187 logisch xfs
6 19077,218 28615,781 logisch reiserfs
7 28615,812 29572,778 logisch linux-swap
8 29572,809 48649,965 logisch xfs
9 55348,976 56290,253 logisch
10 56290,285 57231,562 logisch
UAE copied data to partition 9. My question is, whether it could have written
(parts of it) to partition 5? I still doubt it, but I have no other
explaination for the filesystem corruption on that partition. Especially as
already the short SMART self test of the drive (about 2 minutes) completed
without failure. Now running the long one.
Luckily my /home filing system (partition 8) seems to be OK, as well as the
SuSE Linux 9.1 partition (partition 6).
Regards,
--
Martin 'Helios' Steigerwald - http://www.Lichtvoll.de
Other related posts:
