RTFM pointer for meaning of "offset" and "period".
Brett Smith
Bsmith at Miltope.com
Wed Jun 23 06:58:03 PDT 1999
Transfer period factor field in SCSI (taken from ANSI SCSI Parallel
Interface-2 (SPI-2) ANSI 1142D-T10)
00h-09h Reserved for higher future data rates
0ah 25.0ns (Typical of FAST-40)
0bh 30.3ns (Typical of FAST-40)
0ch 50.0ns (Typical of FAST-20)
The following is the period factor * 4ns
0dh-18h 50- 96ns (Typical of FAST-20)
19h-31h 100-196ns (Typical of FAST-10)
32h-ffh 200-1020ns (Typical of FAST-5 and the very slow....)
The REQ/ACK offset is the maximum number of REQ assertions allowed to be
outstanding before a corresponding ACK assertion is received at the target.
The Seagate Cheetah ST118202LC should give you RAW Sustained throughput of
15 Mbytes/sec.
The maximum I/O data rate on this drive is FAST-20 in Single-ended mode and
FAST-40 in LVD mode.
If your SCSI transfer period is set to 43. 2bh then something needs to be
tweak, your system is not running at it peak performance.
-----Original Message-----
From: Piete Brooks [mailto:Piete.Brooks at cl.cam.ac.uk]
Sent: Tuesday, June 22, 1999 4:37 PM
To: aic7xxx at freebsd.org
Cc: Piete.Brooks at cl.cam.ac.uk
Subject: RTFM pointer for meaning of "offset" and "period".
Can somebody point me at something which describes:
Device using Wide/Sync transfers at 11.4 MByte/sec, offset 15
Transinfo settings: current(43/15/1), goal(43/127/1), user(43/127/1)
[ I've worked out that the "/1" means "wide" (0 for narrow),
but I don't know what the meanings of `offset' and `period' are.
Ultra Enable Flags: 0x0000
is also a concern, as the 7896 and SEAGATE ST118202LC are LVD
It seems that "period" is 250/Mbit of the bus speed,
but what is offset (0, 8 or 15) ? Most Wide are 8, but not this one.
As "SEEPROM found and used", I guess I need to tweak the SCSI BIOS - yes ?
]
To Unsubscribe: send mail to majordomo at FreeBSD.org
with "unsubscribe aic7xxx" in the body of the message
To Unsubscribe: send mail to majordomo at FreeBSD.org
with "unsubscribe aic7xxx" in the body of the message
More information about the aic7xxx
mailing list