Dump Card State Begins ...

Mon Jul 28 21:40:23 PDT 2003

Hi SCSI gurus!

Justin T. Gibbs wrote:

> > neptune# grep "timed out" /var/log/messages
> > Jul 25 03:52:51 neptune /kernel: (da2:ahd1:0:2:0): SCB 0x40 - timed out
> > Jul 25 03:57:22 neptune /kernel: (da2:ahd1:0:2:0): SCB 0x18 - timed out
> > Jul 25 03:58:53 neptune /kernel: (da1:ahd1:0:1:0): SCB 0x1e - timed out
> > Jul 26 10:55:46 neptune /kernel: (da2:ahd1:0:2:0): SCB 0x39 - timed out
> > 
> >   The drives are all U320 Seagate Cheetah 70G ... no RAID involved, its
> > just straight drives using the motherboard's onboard SCSI controller ...
> > the motherboard is the Intel SE7501, in the SR2300 chassis ...
> 
> I need the exact model number and firmware for these drives.  There are
> at least three different Cheetah 70G U320 drives.

I am experiencing very similar problems with Seagate drives
and the mpt driver:

mpt0: <LSILogic 1030 Ultra4 Adapter> port 0xc400-0xc4ff mem 0xfe8c0000-0xfe8cffff,0xfe8d0000-0xfe8dffff irq 10 at device 8.0 on pci3
pcib3: device mpt0 requested decoded memory range 0xfe8d0000-0xfe8dffff
pcib1: device mpt0 requested decoded memory range 0xfe8d0000-0xfe8dffff
mpt0: soft reset
mpt1: <LSILogic 1030 Ultra4 Adapter> port 0xc800-0xc8ff mem 0xfe8e0000-0xfe8effff,0xfe8f0000-0xfe8fffff irq 10 at device 8.1 on pci3
mpt1: linking with peer (mpt0)
pcib3: device mpt1 requested decoded memory range 0xfe8f0000-0xfe8fffff
pcib1: device mpt1 requested decoded memory range 0xfe8f0000-0xfe8fffff
mpt1: soft reset
da0 at mpt0 bus 0 target 0 lun 0
da0: <SEAGATE ST336607LC 0005> Fixed Direct Access SCSI-3 device
da0: Serial Number 3JA1W6WK00007348H2U2
da0: 320.000MB/s transfers (160.000MHz, offset 63, 16bit), Tagged Queueing Enabled
da0: 35003MB (71687372 512 byte sectors: 255H 63S/T 4462C)
da1 at mpt1 bus 0 target 0 lun 0
da1: <SEAGATE ST336607LC 0005> Fixed Direct Access SCSI-3 device
da1: Serial Number 3JA1W17R00007347B6UR
da1: 320.000MB/s transfers (160.000MHz, offset 63, 16bit), Tagged Queueing Enabled
da1: 35003MB (71687372 512 byte sectors: 255H 63S/T 4462C)

The panic I get looks like this:

TargetID	0
SenseBufferLength	32
LUN:	0x0
Control	0x01000000 WRITE SIMPLEQ
DataLength	0x00004000
SenseBufAddr	0x009adfe0
CDB[0:10]	2a 00 00 d8 6d 1f 00 00 20 00
SE32	0xe2d35e30: Addr=0x62c64000 FlagsLength=0x14001000 HOST_TO_IOC
SE32	0xe2d35e38: Addr=0x62be5000 FlagsLength=0x94001000 HOST_TO_IOC LAST_ELEMENT
CE32	0xe2d35e40: Addr=0x9ade48 NxtChnO=0x0 Flgas=0x30 Len=0x10
SE32	0xe2d35e48: Addr=0x62d06000 FlagsLength=0x14001000 HOST_TO_IOC
SE32	0xe2d35e50: Addr=0x62a07000 FlagsLength=0xd5001000 HOST_TO_IOC LAST_ELEMENT END_OF_BUFFER END_OF_LIST
panic: initiate_write_inodeblock_ufs2: already started
syncing disks, buffers remaining...panic: bremfree: removing a buffer not on a queue
Uptime 21h33m23s
(da0:mpt0:0:0:0): Synchronize cache failed, status == 0xb, scsi status == 0x0
(da1:mpt0:0:1:0): Synchronize cache failed, status == 0xb, scsi status == 0x0
mpt0: soft reset failed: ack timeout

I was not able to reproduce the problem with drives with firmware
revision 0006 - so maybe it's fixed. OTOH the panic occurs rather
occasionally, so I cannot be sure.

You suggested disabling packetized protocol. I stared at the camcontrol
manpage and the mpt driver source for an hour - how _do_ I disable
packetized protocol?

I contacted Seagate at their published disk support mail address,
hopefully a human being will read that ;-)

Thanks for any hints,
Patrick
-- 
punkt.de GmbH         Internet - Dienstleistungen - Beratung
Vorholzstr. 25        Tel. 0721 9109 -0 Fax: -100
76137 Karlsruhe       http://punkt.de