[1283] in linux-scsi channel archive
Re: AHA 1542 driver performance questions
daemon@ATHENA.MIT.EDU (David A Willmore)
Mon Jan 20 00:01:05 1997
Date: Sun, 19 Jan 1997 22:56:54 -0600
From: David A Willmore <willmore@cig.mot.com>
In-Reply-To: Jim Van Zandt <jrv@vanzandt.mv.com>
"Re: AHA 1542 driver performance questions" (Jan 18, 9:27pm)
To: Jim Van Zandt <jrv@vanzandt.mv.com>,
David A Willmore <willmore@cig.mot.com>
Cc: linux-scsi@vger.rutgers.edu
On Jan 18, 9:27pm, Jim Van Zandt wrote:
> >From my Adaptec 1542A manual:
>
> "High performance Bus Master DMA with selectable or programmable data
> rates of up to 10 MBytes/second.
>
> "Maximum synchronous SCSI transfer rate of up to 5.0 MBytes/second
>
> "The AHA-1540A/1542A supports multithreaded SCSI initiator operation
> through a simple mailbox protocol. The firmware accepts as many
> Command Control Blocks (CCB's) as required and executes them from its
> local RAM. The firmware controls all of the SCSI activity that a task
> may require This includes: Arbitration, Selection, Disconnection,
> Reconnection, and Command completion.
>
> "Jumper block J1 is the large vertically oriented block of jumpers
> located near the center of the host adaptor. Pin pair 1 is the
> topmost pair of pins.
>
>
> "Pin-pair 1 of jumper block J1 is the synchronous negotiation enable
> jumper. The AHA-1540A/1542A will initiate SCSI synchronous
> negotiation during initialization, or after a SCSI reset if this
> jumper is installed. If the jumper is not installed, the
> AHA-1540A/1542A will still support synchronous SCSI transfers, but a
> different SCSI device must initiate the negotiation. Default is
> jumper removed, synchronous negotiation disabled. If any attached
> SCSI devices support synchronous transfer, the initiation of
> negotiation should be enabled by both the AHA-154xA and the attached
> devices to fully implement the SCSI synchronous negotiation protocol."
>
> The fellow I bought my AHA1542A from had installed a replacement ROM
> which made it equivalent to an AHA1542B (SCSI-2, if I remember
> correctly). You might want to look into that.
Thank you! I'm going to make the leap of logic that not all of the performance
characteristics of this card are controlable via software and, instead, must be
communicated, via jumper settings, with the little micro (in my case it's an
early model 1542A with an 8085) that runs things. I'll review the manual and
give the little guy an overhaul. I know for a fact that the drives where I
have connected will do sync.
Again, thank you very much.
Cheers,
David
