As another data point for this issue:
The system affected by this issue is a production system performing an
important task for us, so we have been considering our options to work around
this problem.
I tried reverting our kernel revision back to the last Lucid kernel that we
have, 2.6.32, installing it from a disk archive.
On rebooting, the system takes much longer to boot, then when the bacula
storage daemon is started, and tries to access our tape autochanger, it
stalls, and the same warning appears in dmesg. After that, the drive was
accessible normally, and (apparently) reliably.
>From memory, after the upgrade to 'Precise', our system boots seemed
typically faster, maybe because the no_probe option was passed to
aic7xxx, maybe not. What now seems apparent is that the task blocked
issue was probably present with our 'Lucid' system, but the scsi
interface to the autochanges was operable afterwards. With our current
('Precise') system, the scsi interface to the autochange is not
operable.
This seems to me to suggest that the aic7xxx module problem has been
present for some time, but that the systems response to it has changed
so that it has a more noticeable effect. This accords with my web
research, which shows reports of this problem (or similar) for some 10
years.
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1016010
Title:
Adaptec SCSI controller causes modprobe hang
To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1016010/+subscriptions
--
ubuntu-bugs mailing list
[email protected]
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
