>Number:         1580
>Category:       kern
>Synopsis:       Timeout of SCSI cmd is insufficiant for Magneto Optical drive.
>Confidential:   no
>Severity:       serious
>Priority:       medium
>Responsible:    kern-bug-people (Kernel Bug People)
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Fri Oct  6 02:05:03 1995
>Last-Modified:
>Originator:     enami tsugutomo
>Organization:
Sony Corporation.
>Release:        NetBSD-current 1995 Sep 30
>Environment:
System: NetBSD king-show 1.0A NetBSD 1.0A (KINGSHOW) #0: Mon Oct 2 22:49:42 JST 1995 root@king-show:/usr/src/sys/arch/i386/compile/KINGSHOW i386
	Intel 486DX4/100, AIR 486EI, Buslogic BT-445C, NumberNine GXE Level 12,
	16MB of memory, 1GB of harddisk (Seagate HAWK IILP) scsi id 0.
		or
	Intel Pentium 100, Micronics M54Pi, Buslogic BT-946C, NumberNine GXE64,
	16MB of memory, 1GB of harddisk (Fujitsu M1606S) scsi id 0.

	And magneto optical disk drive, Elecom EMO 2300S (the internal
	drive is DELTIS MOS321), 3.5" 230MB Magneto Optical disk drive.

>Description:
	When using MO drive above (heavily?), the system gets timeout twice
	and panic with following message:

sd1(bt0:2:0): timed out
sd1(bt0:2:0): timed out AGAIN
bt0: exiting but not in use

>How-To-Repeat:
	1. make file system on the MO disk (disklabeling and newfs).
	2. mount it (ex, mount /dev/sd1c /mnt).
	3. use a bit heavily. (ex. cd /mnt; dump 0f - / | restore xf -).
	   tar also cause the problem, but not so dump.
>Fix:
	The work around is to increase the timeout of sdstart (for example
	from 10000 to 50000).  I don't know if 50000 is sufficiant or whether
	introduce quirk for it.
>Audit-Trail:
>Unformatted: