linux 2.6.15 and tape problems with aic79xx driver

I have one machine that uses the aic7xxx driver of linux 2.6.15
and amanda and the tape driver are working fine. There is a fix in
2.6.15 to this driver that allowed this to happen that has been a bug
since the 2.6.12.x kernels (starting at 2.6.13, the bug showed up)

A similar bug on a second machine has been going on with the aic79xx
driver (for 320MB/s scsi) that is also supposed to have been fixed with
this kernel 2.6.15. I can tell the tape interface is much improved and the
system no longer hangs with tape access like it used to, but nevertheless
I cannot use the tape drive at the present time.

The tape drive is a Seagate 9sp40_000 (DSS-4) drive. Under 2.6.12.6 and
prior kernels, it works fine with AMANDA. The tape configuration shown
from mt status is:

SCSI 2 tape drive:
File number=0, block number=0, partition=0.
Tape block size 0 bytes. Density code 0x26 (DDS-4 or QIC-4GB).
Soft error count since last status=0
General status bits on (41010000):
BOT ONLINE IM_REP_EN

Now when I boot up under 2.6.15, the tape is found as expected and the output
of "mt status" is identical.

SCSI 2 tape drive:
File number=0, block number=0, partition=0.
Tape block size 0 bytes. Density code 0x26 (DDS-4 or QIC-4GB).
Soft error count since last status=0
General status bits on (41010000):
BOT ONLINE IM_REP_EN

However, if I try to write anything to the tape drive, it returns an error. For
example:

amlabel -f fea fea12

will spit out "input/output error".

I believe there is something missing in the tape configuration that is causing this
or it could still be a scsi driver bug (but I doubt it now).

Here is the input to the tape drive initialization:

# This file contains example definitions for different kinds of tape
# devices.
#
# You can find some examples in /usr/share/doc/mt-st/examples.
#
# A compressing DAT (DDS-4])
manufacturer = "SEAGATE" model = "DAT 9SP40-000" revision = "9100" {
scsi2logical=0
density=0x26
can-bsr=1
can-partitions=1
auto-lock=1
async-writes=1
read-ahead=1
buffering=1
mode1 blocksize=0 compression=0
#mode2 blocksize=0 compression=1
#mode3 blocksize=32768 compression=0
#mode4 blocksize=32768 compression=1
}

Does anyone have an idea as to what could be wrong ?

Are there others having problems with aic79xx driver and their tape drives ?

------------------------------
James D. Freels, Ph.D.
Oak Ridge National Laboratory
freelsjd AT ornl DOT gov
http://www.comsol.com/stories/hfir/