[RELENG_5_3] Strange timeout of mpt(4) with disk array

(bcc'ed to -CURRENT)
Dear folks,

I got a strange problem with mpt(4) when connecting with a disk array.
The disk array works well when working under 160M/s mode, and 320M/s mode
for RELENG_4 and Linux 2.4.20-8(?) from a stock RedHat 9 installation.

I am trying to figure out what is happening in our mpt(4) driver which
has caused this.  Any help is appreciated.  More information upon request.

The message I got from the console (hand-copyed and typed)
mpt1: timeout on request index=0xf0 sequence=0x00000084
mpt1: status 00000001; mask 00000001; Doorbell 24000000
request state on chip

SCSI IO Request _at_ 0xe4ad5c5c
	Chain Offset	0x00
	MsgFlags	0x00
	MsgContext	0x000000fe
	Bus:		0
	TargetID:	0
	SenseBufferLength:	18
	LUN:		0x2
	Control:	0x02000500 READ UNTAGGED
	DataLength	0x00000024
	SenseBufAddr	0x3e64ade0
	CDB[0:6]	12 40 00 00 24 00
	SE32		0xe4da9030; Addr=0x854884 FlagLength=0xd1000024
		LAST_ELEMENT END_OF_BUFFER END_OF_LIST

Thanks in advance!

Cheers,
-- 
Xin LI <delphij frontfree net>	http://www.delphij.net/
See complete headers for GPG key and other information.