From owner-freebsd-scsi Mon May 27 14:29:07 1996 Return-Path: owner-freebsd-scsi Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id OAA15088 for freebsd-scsi-outgoing; Mon, 27 May 1996 14:29:07 -0700 (PDT) Received: from phaeton.artisoft.com (phaeton.Artisoft.COM [198.17.250.211]) by freefall.freebsd.org (8.7.5/8.7.3) with SMTP id OAA15061; Mon, 27 May 1996 14:29:01 -0700 (PDT) Received: (from terry@localhost) by phaeton.artisoft.com (8.6.11/8.6.9) id OAA09109; Mon, 27 May 1996 14:26:59 -0700 From: Terry Lambert Message-Id: <199605272126.OAA09109@phaeton.artisoft.com> Subject: Re: Re(2): SCSI hostadapter To: Andrew.Gordon@net-tel.co.uk Date: Mon, 27 May 1996 14:26:59 -0700 (MST) Cc: freebsd-scsi@FreeBSD.org, j@uriah.heep.sax.de, freebsd-hackers@FreeBSD.org In-Reply-To: <"1443-960526225554-A900*/G=Andrew/S=Gordon/O=NET-TEL Computer from "Andrew.Gordon@net-tel.co.uk" at May 27, 96 01:30:12 am X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-scsi@FreeBSD.org X-Loop: FreeBSD.org Precedence: bulk > > Reminds me: does anybody have any information about the 53C400? > > > > I've got a couple of these boards sitting on a shelf, they used to > > accompany HP ScanJets. I don't think they will be anything that can > > be called performant, but just out of curiosity... Perhaps they are > > good enough to recommend them to someone who needs an adapter for an > > Archive Viper 150 or so (which i'd recommend rather than those floppy > > tape crap). > > As it happens, I was trying to put a system together yesterday, using mostly old junk I had lying around. Since I didn't care about disc performance, and I had an old SCSI drive plus a 53C400 adapter handy, I thought I would use those (booting off floppy with some suitably hacked bootblocks). > > The adapter originally came with a scanner, (but not an HP one) and contains just the 53C400 plus an LS245 buffering the databus and some SCSI termination resnets. Curiously, the nca driver probes the card as a NCR-5380 despite the fact that the chip on the card is clearly labelled as a 53C400A. However, hacking the driver to probe only for 53C400 (it normally probes 5380 first) caused the card not to be probed at all. > > This particular card does not support interrupts (it doesn't even have any fingers on the connector for any of the IRQ lines). > > For the hard drive, it seems to work reliably, but is _very_ slow - here are some bonnie results (and results for the same motherboard/drive but with a 2940 instead, for comparison): > > -------Sequential Output-------- ---Sequential Input-- --Random-- > -Per Char- --Block--- -Rewrite-- -Per Char- --Block--- --Seeks--- > MB K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU /sec %CPU > ncr 16 105 89.0 110 10.8 55 1.3 109 8.4 111 5.5 9.6 8.7 > 2940 16 560 28.4 553 7.7 266 6.5 612 28.7 605 8.3 24.8 3.8 > > (the drive is an old SCSI-1 device, async transfers only. CPU is AMD 486/100). This is typically an artifact of having the interrupt set incorrectly. I'd caution you that it's also possible that it's just a dog-slow interface, so there may be nothing to be done about it. Terry Lambert terry@lambert.org --- Any opinions in this posting are my own and not those of my present or previous employers.