Date: Thu, 20 Mar 1997 17:48:26 +1000 (EST) From: Stephen McKay <syssgm@dtir.qld.gov.au> To: Michael Smith <msmith@atrad.adelaide.edu.au> Cc: freebsd-hackers@freebsd.org, syssgm@dtir.qld.gov.au Subject: Re: Anybody using Tyan S1672 (Tacoma) motherboard with parity RAM? Message-ID: <199703200748.RAA00482@ogre.dtir.qld.gov.au>
next in thread | raw e-mail | index | archive | help
Michael Smith <msmith@atrad.adelaide.edu.au> wrote: >sthaug@nethelp.no stands accused of saying: >> Is anybody using the Tyan S1672 (Tacoma) motherboard with parity RAM >> with FreeBSD? > >No, but I've seen some oddities with another HX board. > >> We have a couple of these here, and they work fine as long as we only >> configure normal parity checking in the BIOS. As soon as we turn on >> ECC, we start getting mysterious crashes etc. We'd like to hear about >> other experiences with this board. Thanks in advance! > >With the Tekram P5H30 board, using an Award BIOS, there are several >different parity/ECC settings. At least one of these results in the >board bombing out with an ECC error long before anything resembling >an OS is loaded. I've been meaning to bring this up in -hardware, but now that the topic is open... I have an Octek Rhino 9 rev 1.4 motherboard, which, according to the PCI stuff at boot has rev 1 of both the TXC and PIIX3 chips. If rev 1 equates to A1 then ECC is supposed to not work (according to a chap here who watches the PC chipset groups). However, all the advertising, and the packaging, and the BIOS claim that it works. Plus, if I boot in parity mode, then use pciconf to flip it into ECC mode, then I get an instant NMI like I expect. So, maybe ECC is working. However, ECC only works when I run at 120Mhz (60Mhz bus speed). At 133Mhz (66Mhz bus speed) I can run parity mode at the most aggressive settings. At 120Mhz, I can run either parity or ECC mode at the most aggressive settings. At 133Mhz with the LEAST aggressive settings I can find, ECC will cause a lockup (not an NMI) about one minute into a kernel compilation. Sometimes it doesn't make it through booting. By the way, I'm using Micron 60ns FPM 16Mb (x2) ram. I have ordered Intel's errata on these chips through their web site, but that was weeks ago. If they ever turn up, it should shed some light on this. If anyone else has this info (maybe Rodney?), I'd appreciate some enlightenment! Stephen.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199703200748.RAA00482>