Please do NOT do this. That is not the correct fix. The bug is with the
USB/SCSI retry logic (which changed between 2.6.26 and 2.6.27 and causes an
infinite retry loop in this case).
I tried the adding ENV{DEVTYPE}=partition workaround a while ago and rendered
my system unbootable and I got to spend a merry time with the live CD undoing
the change. If you use dmraid (fakeraid) support, then the vol_id line NEEDS to
occur for the whole disk, not for a partition only. If your udev changes don't
break dm/fakeraid, then that's much better, but it's still papering over the
real underlying problem.
The kernel retry code needs to be fixed, and to avoid getting any I/O
error at all from vol_id, all the broken devices which return
(last_sector + 1) from the call to find the capacity, instead of
last_sector need to have an UNUSUAL_DEV entry with US_FL_FIX_CAPACITY in
the flags. Sadly, this seems to be a very large number of devices.
The kernel fix that Alan Stern provided should prevent the infinite
retry logic and so you should see a handful of failures and then
everything should carry on. I haven't yet downloaded the Ubuntu -11
kernel source to see why this isn't happening.
--
USB Hard Drive Not Accessible, vol_id hangs
https://bugs.launchpad.net/bugs/264789
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
--
ubuntu-bugs mailing list
[email protected]
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
