[EMAIL PROTECTED] wrote:
> My I/O exerciser will occassionally "hang" under certain subsystem error
> inject scenarios. It can run OK for anywhere between 0 and 38 attempts
> before failing.
>
> The other question is what is going on w/ host2, channel0, id6... I'll put
> a SCSI bus analyzer on that and see if I can find the cause of the 2603's
> (I assume that is the ASC/ASCQ values, correct???) although they appear on
> the successfully recovered error injects as well as the bad ones.
>
> Thanks,
> Steve
>
Please keep up the good work!
I appreciate to see someone with SCSI bus analyzer tackling
linux scsi problems.
(Just curious. Is your I/O exerciser available in binary
or source format to general public?)
I think Linux scsi drivers need to be tested
with such heavy duty I/O exerciser (not just Bonnie and friends)
to iron out the subtle remaining bugs.
My 7 CD Nakamichi changer drive (MBR-7) has turned out to be
a good LUN handling tester in the past.
Oh, before I forgot.
Mounting one CD in the Nakamichi MBR-7 drive under /mnt and
running "ls -lR" on it, and then trying to mount ANOTHER CD
in the MBR-7 under /mnt1 no longer hangs the system as it used to.
I checked this with 2.2.16-pre2.
It is still true that the mouse cursor got stuck, but I can successfully
use the different VTY such as vty0 and login as root and
kill "ls" processes umount the /mnt.
This is with Tekram DC390 2.0d28 version and 2.2.16-pre2.
(However, as I mentioned previously I still get I/O errors
and CD read from the first CD returns bogus result while
the second mount attemps are made.)
After umounting the CD, the system comes back to normal and
I can play with it until I shut it off for the day.
I think the reason that the mouse cursor seems to get stuck is
that there is a portion in the SCSI subsystem where the
INTERRUPT SEEMS TO BE DISABLED DURING PROLONGED ERROR HANDLING
(maybe some busy loop to wait for some events to happen?)
and then enabled for just a very short duration of time.
I have no idea where to look.
So, the situation is still not quite good, but not as
fatal as it used to be, and I feel more courageous to
test the double mounting excercise more often.
Sorry, I have not tested 2.3.99-preXX series. My disk
is multi-boot (DOS, OS/2) and I use
loadlin to boot linux.
It looks to me that there is a problem using
loadlin with the latest development kernel when I tested
it several weeks ago.
(Does it support initrd? Loadlin seems to require it.)
Ishikawa
-
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to [EMAIL PROTECTED]
