[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: i386/1521: Adaptec 2940 bug in 2.8 not in 2.7



The following reply was made to PR i386/1521; it has been noted by GNATS.

From: Kenneth Ingham <ingham_(_at_)_i-pi_(_dot_)_com>
To: Peter Galbavy <peter_(_at_)_wonderland_(_dot_)_org>
Cc: Kenneth Ingham <ingham_(_at_)_i-pi_(_dot_)_com>, gnats_(_at_)_openbsd_(_dot_)_org
Subject: Re: i386/1521: Adaptec 2940 bug in 2.8 not in 2.7
Date: Fri, 17 Nov 2000 16:25:09 -0700

 On Fri, Nov 17, 2000 at 02:17:47PM +0000, Peter Galbavy wrote:
 > He will be double checking physical termination etc later, but as of
 > yesterday, it appeared to be some sort of bus timing issue.
 Why would it be a bus timing issue when 2.7 runs fine?  Also, FreeBSD
 4.1 and  4.1.1 run fine on identical hardware (we have five identical
 machines, currently two are hot spares, soon to be only one as a
 space).  For what it's worth, the internal cable connecting the
 drives to the controller is terminated with an active terminator.
 
 > I haven't currently got remote access to the machine, so no dmesg, but
 > the config of AH2940UW with 3x SONY SDT-9000 drives was weird. The two
 > tape drives that did no have a tape loaded consistently failed to
 > report (according to dmesg) their parameters. The name, model and
 > version stuff was basically nil. The drive with a tape loaded
 > occassionally (about 50%) reported correctly, othertimes not.
 >
 > When they misreported themselves, the scsiconf stuff saw them as
 > "direct fixed" (0,0 I guess) and tried to config them up as sd
 > devices.
 In this case, the disk drives are not being misreported, and they are
 the only hardware on the SCSI bus.
 
 Kenneth