*BSD News Article 6648


Return to BSD News archive

Newsgroups: comp.unix.bsd
Path: sserve!manuel.anu.edu.au!munnari.oz.au!sgiblab!swrinde!gatech!concert!sas!mozart.unx.sas.com!torpid.unx.sas.com!sastdr
From: sastdr@torpid.unx.sas.com (Thomas David Rivers)
Subject: Re: Corrupted directory
Sender: news@unx.sas.com (Noter of Newsworthy Events)
Message-ID: <Bw8853.BE8@unx.sas.com>
Date: Fri, 16 Oct 1992 18:17:27 GMT
References: <Bw4Fv7.M4L@unx.sas.com> <5232.9210151353@thor.cf.ac.uk> <1992Oct16.033843.9805@fcom.cc.utah.edu>
Nntp-Posting-Host: torpid.unx.sas.com
Organization: SAS Institute Inc.
Lines: 43

In article <1992Oct16.033843.9805@fcom.cc.utah.edu> terry@cs.weber.edu (A Wizard of Earth C) writes:
>In article <5232.9210151353@thor.cf.ac.uk> spedpr@thor.cf.ac.uk (Paul Richards) writes:
>>In article <Bw4Fv7.M4L@unx.sas.com> sastdr@torpid.unx.sas.com (Thomas David Rivers) writes:
>>|
>>|Regarding this infinite reboot cycle; I recently applied the entire

	A lot of stuff deleted... (about how I noticed the infinite
 fsck/reboot problem on two disks at home after I applied the patchkit)

>I -think- the offending patch is patch00038, which updates the error
>reporting mechanism for bad sector reporting for read errors being reported
>to user programs (with this patch in place, trying to install the bin01
>distribution from files located on a bad sector actually fails where you
>would want it to, rather than in the decompression some time later).
>
>Backing out this patch does not require deinstallation of other patches.  It
>should be noted that the correct fix is to adjust the fsck code (or to return
>the expected value to the fsck code) so that the checking is not aborted by
>the error.
>
>Generally, this error will not occur on disks installed with a dist.fs with
>the patches applied; instead, it will occur on disks installed on top of
>bad sectors (without the error being caught) and to which the patches are
>later applied.
>
>I don't know how the author of the patch (Frank Maclachlan) handles this on
>his drives, but I suspect that he does one of the two approaches above.
>

  Well, one of my machines is a SCSI, which should have no bad sectors reported
by the controller  and the other is an IDE  which, according to the 
INSTALLATION document should have no bad sectors either - although I'm
skeptical/confused about that one.

  So, while patch0038 could be the problem; I don't think it's because
fsck is actually running into read errors.  Could it be that the mechanism
is incorrectly reporting such errors?

	- Dave Rivers -
	(rivers@ponds.uucp (home))
	(sastdr@unx.sas.com (work))
-- 
UPDATE ALL INFORMATION AND POD INTO COSMOS - Federal Express