Return to BSD News archive
Path: sserve!newshost.anu.edu.au!harbinger.cc.monash.edu.au!msuinfo!caen!math.ohio-state.edu!magnus.acs.ohio-state.edu!csn!col.hp.com!gag.com!arpad!fls From: fls@k0yum.ampr.org (Fred L. Schneider) Newsgroups: comp.os.386bsd.bugs Subject: re: FreeBSD 1.1.4.1 and Conner SCSI disk Date: 3 Nov 1994 01:04:24 GMT Organization: Amateur Radio Station K0YUM Lines: 31 Message-ID: <399cuo$oo@arpad.k0yum.ampr.org> Reply-To: fls@arpad.k0yum.ampr.org NNTP-Posting-Host: 44.20.3.1 X-Newsreader: TIN [version 1.2 PL0] >From: gfm@werple.apana.org.au (Graham Menhennitt) >Newsgroups: comp.os.386bsd.bugs >Subject: FreeBSD 1.1.5.1 and Conner SCSI disk >Date: 25 Oct 1994 01:49:39 GMT > >There is some incompatibility between FreeBSD 1.1.5.1 and certain Conner >SCSI disks via an Adaptec 1542B. The problem maybe more widespread than this >but it certainly exists with the 1.05Gb and 1.35Gb models. >There appears to be corruption of multi-sector reads. This most obviously >manifests itself during fsck on boot. Errors are reported when none really >exists. The error is repeatable and always occurs at the same place in the >filesystem. By that I mean that if you run fsck a number of times, it always >gives the same error at the same inode. If you modify the filesystem the >error may change. For what it is worth: I have been seeing this behaviour on a 212 Meg Conner CP30200 under NetBSD-0.9. Having recently upgraded to NetBSD-1.0 BETA, I'd hoped never to be plagued again and things have been fairly clean on a seldom-used system since last Sunday. One glitch occured after installing my custom-compiled kernel, but after that fix all seems well for now. I also run an AHA-1542(B), BTW. >My short term solution is to modify /etc/rc fsck -p [stuff deleted]... As implied above, a more permanent and proper fix is highly desireable. I can't guarantee this problem and Graham's are related, but thought it worthy of note that NetBSD and smaller Conner drives have been observed with similar behaviour. //