From owner-freebsd-hackers Tue May 30 00:25:49 1995 Return-Path: hackers-owner Received: (from majordom@localhost) by freefall.cdrom.com (8.6.10/8.6.6) id AAA25863 for hackers-outgoing; Tue, 30 May 1995 00:25:49 -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 AAA25857 for ; Tue, 30 May 1995 00:25:45 -0700 Received: (from rgrimes@localhost) by gndrsh.aac.dev.com (8.6.11/8.6.9) id AAA02866; Tue, 30 May 1995 00:12:27 -0700 From: "Rodney W. Grimes" Message-Id: <199505300712.AAA02866@gndrsh.aac.dev.com> Subject: Re: 950412 hangs on ncr0 probing: To: agl@mac.glas.apc.org (Anthony Graphics) Date: Tue, 30 May 1995 00:12:27 -0700 (PDT) Cc: freebsd-hackers@FreeBSD.org In-Reply-To: from "Anthony Graphics" at May 30, 95 11:00:49 am X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 2900 Sender: hackers-owner@FreeBSD.org Precedence: bulk ... > > I am quite sure the bug is not fixed in Saturn II, aka Saturn >= Rev 4., > > If you had a Saturn I based board I would tell you to turn your external > > cache off as the bug in it is very serious and prevents any type of > > external cache from being used with bus mastered devices. > > > > What is the CPUID value for this CPU chip?? Does this happen to be > > a stepping 0 Intel 486DX4/100?? Please send me the lines from > ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ yes. Anything wrong with it Possibly, I have 2 other sites reporting strange signals and or crashes and both of them have stepping 0 Intel 486DX4/100 CPU chips. Can you please send me ALL of the markings on both the top and the bottom of the chip. I have an open incident number with Intel on this and need that information to add to the report. If possible get your hands on a stepping 1 CPU and see if the problem just goes away. > > dmesg that are similiar to this: > > > > FreeBSD 2.0-BUILT-19950529 #0: Mon May 29 07:40:06 1995 > > rgrimes@capo1.aac.dev.com:/var/src/sys/compile/GNDRSH > > CPU: 90-MHz Pentium 735\\90 (Pentium-class CPU) > > Origin = "GenuineIntel" Id = 0x524 Stepping=4 > > Features=0x1bf > > real memory = 66715648 (16288 pages) > > avail memory = 64196608 (15673 pages) > > FreeBSD 2.0.950412-SNAP #0: Thu Apr 13 14:11:34 1995 > jkh@time.cdrom.com:/usr/src/sys/compile/GENERIC > CPU: i486 DX4 (486-class CPU) > Origin = "GenuineIntel" Id = 0x480 Stepping=0 Features=0x3 Yep, it is a stepping 0 :-(. Intel won't confirm the problem, but with 3 sites now reporting this I am getting mighty suspecious!! > > CPU: > ser # A4212608CA I need *ALL* the numbers, both top and bottom. If there is a white tag on the bottom of the CPU that says warranty void if removed do not remove it. If you have a warranty on the CPU chip take it back and get a stepping 1 CPU from who ever you got it from. > Here, motherboard settings: > JP17,18 Open > JP22 1&2 > JP21 2&3 > JP26 1&2 > JP27 2&3 > JP28 1&2 > JP29 2&3 > > JP18,20 1&2 - Ext CPU Speed 33 MHz > JP3[012] 1&2 - SL CPU > JP33 open (x3 Internal clock) > JP34 2&3 > JP35 1&2 > JP36 2&3 > JP38 Open > > I don't have a MB or the book for it here so I can't look these up. But from your comments it looks right. ... > > If you can please do the voltage measurement to make sure the power > > supply is putting out 4.95 to 5.05V at the motherboard power supply > > connector. > > > Ok, will do, but tell me first if anything wrong with dmesg output. > (stepping 0 486 in particular) I suspect, but do not have confirmation that this stepping of the chip erroniously asserts a privilidge instruction fault under certain conditions. -- Rod Grimes rgrimes@gndrsh.aac.dev.com Accurate Automation Company Custom computers for FreeBSD