[2.6.12.2] unexpected block device behaviour ...

[Herbert Poetzl]

Hi Folks! Andrew!

accidentially stumbled over the following issue:

creating a filesystem (in userspace) requires a period
of quiescence after that (about 5 seconds) until you
can reliably mount the new filesystem ...

this can be observed on 2.6.11 as well as 2.6.12.2
(didn't test older kernels yet) on x86_64 (SMP) with
loopback or dm (lvm2) based block devices (probably
also with other block devices) for different file-
systems (ext2,ext3,xfs,jfs,reiserfs) ...

Rik and I agreed that it is at least unexpected
behaviour, especially as doing an fsync() on the
block device doesn't remedy the deficiency.


so the sequence to 'trigger' this basically is:

  mkfs.* $DEV
  mount $DEV $MNT

and the mount fails sometimes with:
mount: wrong fs type, bad option, bad superblock on /dev/*,
       missing codepage or other error
(but works quite fine if you wait 10 seconds)

here are some scripts (and example output) I used 
for testing this ...

http://vserver.13thfloor.at/Stuff/x.sh
http://vserver.13thfloor.at/Stuff/x.sh.txt

http://vserver.13thfloor.at/Stuff/x2.sh
http://vserver.13thfloor.at/Stuff/x2.sh.txt


best,
Herbert

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/