*BSD News Article 97483


Return to BSD News archive

Path: euryale.cc.adfa.oz.au!newshost.carno.net.au!harbinger.cc.monash.edu.au!news.mel.connect.com.au!munnari.OZ.AU!news.ecn.uoknor.edu!feed1.news.erols.com!chi-news.cic.net!su-news-hub1.bbnplanet.com!news.bbnplanet.com!venus.sun.com!news2me.EBay.Sun.COM!newsworthy.West.Sun.COM!news70.West.Sun.COM!adler
From: adler@west.sun.com.NOSPAM.COM (Bruce Adler [Contractor])
Newsgroups: comp.os.linux.hardware,comp.unix.bsd.freebsd.misc
Subject: Re: adaptec 2940. broken. details inside
Date: 10 Jun 1997 13:35:29 GMT
Organization: Sunsoft Inc., Los Angeles, Ca.
Lines: 31
Message-ID: <5njl71$t5b@abyss.West.Sun.COM>
References: <5micnb$ljl$1@ssw.filoli.com>
NNTP-Posting-Host: lucia.west.sun.com
Originator: adler@West.Sun.COM
Xref: euryale.cc.adfa.oz.au comp.os.linux.hardware:76947 comp.unix.bsd.freebsd.misc:42701

In article <5micnb$ljl$1@ssw.filoli.com>,
Steven Bjork <sbjork@filoli.com> wrote:

> ...
>Now, I'm curious why many freebsd folks claim to not have
>seen this 2940 problem, since it seems to be completely
>hardware related. If freebsd has a workaround in their
>driver, please share the wealth! 

I think you may have misinterpreted or read way too much into the
Solaris x86 notice regarding problems with the 2940U and 2940UW
adapters.  As far as I know there are not now and have never been any
*hardware* problems on any of the various flavors of AHA-2940
adapters.

The lockup problem on Solaris x86 was resolved long ago (I think at
least a year ago, but my memory of the exact dates are rather dim).
The lockup problem turned out to be due to a rather obscure *software*
bug in the HIM code supplied by Adaptec, not a hardware problem. The
symptom was that the adapter's HIM code would in certain unusual
situations get slightly confused and spontaneously turn on the
synchronous data transfer double-clock mode which caused it to try to
transfer data twice as fast as the target expected.

Given that the Linux and BSD drivers don't seem to be based directly on
the Adaptec HIM code, it's highly unlikely that any versions of the
drivers for those OSes have the exact same problem Solaris x86 used to
have. Nor is it likely the fix for the Solaris x86 driver is at all
useful in the Linux or BSD driver.

BA