From owner-freebsd-hackers Mon May 22 17:17:20 1995 Return-Path: hackers-owner Received: (from majordom@localhost) by freefall.cdrom.com (8.6.10/8.6.6) id RAA02373 for hackers-outgoing; Mon, 22 May 1995 17:17:20 -0700 Received: from gndrsh.aac.dev.com (gndrsh.aac.dev.com [198.145.92.241]) by freefall.cdrom.com (8.6.10/8.6.6) with ESMTP id RAA02367 for ; Mon, 22 May 1995 17:17:15 -0700 Received: (from rgrimes@localhost) by gndrsh.aac.dev.com (8.6.11/8.6.9) id RAA06371; Mon, 22 May 1995 17:16:50 -0700 From: "Rodney W. Grimes" Message-Id: <199505230016.RAA06371@gndrsh.aac.dev.com> Subject: Re: Review this hardware please To: dufault@hda.com (Peter Dufault) Date: Mon, 22 May 1995 17:16:50 -0700 (PDT) Cc: hackers@FreeBSD.org In-Reply-To: <199505221944.PAA21749@hda.com> from "Peter Dufault" at May 22, 95 03:44:41 pm X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 4512 Sender: hackers-owner@FreeBSD.org Precedence: bulk > > Our ISP, Intuitive Information (iii.net) is installing their new > Pentium system. Up to now they've been using a 486DX2-66 running > FreeBSD 1.1, and now they are upgrading to a PCI Pentium running > 1.1.5.1. > > With the recent comments about Buslogic controllers and broken > motherboards I'd like to bounce the configuration off the group so > that we can help head off some problems. The system will apparently > be going into production tonight. > > This will be their login system and will be a heavily used system. > Some of the subscribers have been complaining about FreeBSD 1.1 > stability (even when it is broken hardware) and posting "why don't > you get a real system like BSDI" comments. See jkh's comments, I concur with him on this. BSDI can be made to fall over. FreeBSD 1.1 was known to be riddled with bugs, 1.1.5.1 should be rock solid stable for this application as long as you don't try to push the limits to much (the PCI code in 1.1.5.1 was early, so don't expect to be able to add anything else to the PCI bus in this box. The bt946 was well supported in that release, since it really just looks like an ISA bus card to FreeBSD. > So here is their description of the new system, a few more details, > and the dmesg output. > > > iii2 is a Pentium P90 (Intel SX957) with an Intel Neptune ^^^^^^^^^^^ Family 5, Model 2 Stepping 4 Manufacturing stepping B5, STD type part, known to be a good working Pentium Chip. (FYI, current production part is SX968, stepping 5, mfg stepping C2 as of my April 19th spec sheet list) > > motherboard with the Neptune PCI chipset. We've got 64 Meg of memory in > > it. The 32-bit, fast SCSI-2 controller (to which we'll migrate the SCSI > > disks > > soon) can transfer 132 MBytes/sec in burst mode. It's a BusLogic BT-946C. >From the dmesg this is a rev 4.2 BIOS, I have 4.21/4.84, you may want to check the firmware revision, there was some problems in 4.82 and 4.83. If the thing boots and runs FreeBSD then I would probably just leave it alone, often upgrading the BIOS will infact break it :-(. If you have problems with certail types of devices I would try to get the 4.21/4.84 BIOS/Firmware EPROMS for it. I don't have the right sockets for my prom burner or I would blow you a set :-(. > I got this additional info from them: > > > Neptune-ISA Rev 1.A MBD-P90NG2 Don't recognize this, who makes the MB?? > > PCI chipset: Intel S82433NX Okay. > > Award ISA/PCI 586 Bios Best PCI/Pentium BIOS I have seen! Should be at version 4.50 or later. > > Intel SX957 P90 CPU > > BT-946C Rev Ae > > Here is their dmesg output: > > > FreeBSD 1.1.5.1(RELEASE) (III1) #0: Thu May 18 00:49:10 EDT 1995 > > wink@nic.iii.net:/usr/src/sys/compile/III1-04 > > CPU: i586 (586-class CPU) Id = 0x524 Origin = "GenuineIntel" > > real memory = 66715648 (16288 pages) > > avail memory = 64745472 (15807 pages) > > using 819 buffers containing 6709248 bytes of memory > > Probing for devices on the ISA bus: > > pc0 at 0x60-0x6f irq 1 on motherboard > > pc0: type color > > sio0 at 0x3f8-0x3ff irq 4 on isa > > sio0: type 16450 > > fdc0 at 0x3f0-0x3f7 irq 6 drq 2 on isa > > fdc0: [0: fd0: 1.44MB 3.5in] > > wdc0 at 0x1f0-0x1f7 irq 14 on isa > > wdc0: unit 0 (wd0): > > wd0: 515MB (1056384 total sec), 1048 cyl, 16 head, 63 sec, bytes/sec 512 > > wdc0: unit 1 (wd1): > > wd1: 516MB (1058400 total sec), 1050 cyl, 16 head, 63 sec, bytes/sec 512 > > bt0: PCI/EISA/VLB(32bit) bus detected > > bt0: reading board settings, eisa dma, int=11 > > bt0: version 4.2, sync, parity, 32 mbxs, 32 ccbs > > bt0: targ 0 offset=08, period=200nsec > > bt0: Enabling Round robin scheme > > bt0 at 0x330 irq 11 on isa > > bt0 targ 0 lun 0: type 0(direct) fixed SCSI1 > > bt0 targ 0 lun 0: EEEKKKSSSS... I hope that was just for testing, that is an ancient SCSI-I disk that is about as slow as you can go :-(. What happens when they hang a fast SCSI-II drive on that bt946, they may need firmware updates. > > sd0: 100MB (205075 total sec), 1019 cyl, 6 head, 33 sec, bytes/sec 512 > > ed0 at 0x280-0x29f irq 5 on isa > > ed0: address 00:20:c5:00:fa:8d, type NE2000 (16 bit) > > npx0 on motherboard > > Questions? Comments? Criticisms? Other than the disk it all seems okay to me... -- Rod Grimes rgrimes@gndrsh.aac.dev.com Accurate Automation Company Custom computers for FreeBSD