Date: Thu, 21 Dec 2000 15:44:00 -0500 (EST) From: Kenneth W Cochran <kwc@world.std.com> To: Mike Tancsa <mike@sentex.net> Cc: freebsd-stable@freebsd.org Subject: Re: RELENG_3 buildworld problems Message-ID: <200012212044.PAA18831@world.std.com> References: <200012210615.eBL6FJw62370@crotchety.newsbastards.org> <E148rPa-0005py-00@ratbert.oucs.ox.ac.uk> <200012210615.eBL6FJw62370@crotchety.newsbastards.org>
next in thread | previous in thread | raw e-mail | index | archive | help
>From owner-freebsd-stable@FreeBSD.ORG Thu Dec 21 13:28:59 2000 >Date: Thu, 21 Dec 2000 13:20:08 -0500 >To: Kris Kennaway <kris@FreeBSD.ORG>, freebsd-user@netscum.dk >From: Mike Tancsa <mike@sentex.net> >Subject: Re: RELENG_3 buildworld problems >Cc: Neil Long <neil.long@computing-services.oxford.ac.uk>, > freebsd-stable@FreeBSD.ORG > >At 07:45 AM 12/21/00 -0800, Kris Kennaway wrote: >>On Thu, Dec 21, 2000 at 07:15:19AM +0100, Reddy Crashalott wrote: >> >> > :: Anyone else seeing such weirdness? >> > >> > I've seen it with flaky memory. Try swapping bits of it out, or >> > doing a memory test, until you isolate the defective part(s) >> >>I have a dodgy IDE controller which also likes to flip bits. > >Interesting you mention that. With a 3ware 6400 RAID card, >setting the strip size to 128K causes all sorts of memory >errors using DOS and UNIX memtest programs. Putting it >back to the default 64K makes the problem go away. > > ---Mike Hmmmmm... (digging into brain-cobwebs) Could this have any relation to the "segments" in the x86 addressing range? IIRC (well, with '386 & '486 CPUs on ISA/EISA-bus machines), any "segment" (example 64k or 128k) had to start on an address boundary that was an integer multiple of that segment-size. (?) But I've never (yet :) seen how this relates to PCI... -kc To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200012212044.PAA18831>