Not sure if this made it to the list, so here it is again.

Also, I have tried the NCRSCSI kernel - while it does not crash, it
drops bytes every few 100k, stuffing the CRCs.

------------- Begin Forwarded Message -------------

>From weripp Mon Dec  8 14:53:59 1997
To: port-mac68k@NetBSD.ORG
Subject: Crash reading from /dev/rst0
Content-Transfer-Encoding: 7bit
Content-MD5: O27NOkNGoctqiVmi6tM4Ng==

People,

When trying to read a tape (4mm DAT DDS1) using

dd if=/dev/rst0 ibs=1k

on a homegrown SBC kernel of 04/12/97 vintage, I received the following
crash (arguments/details available on request):

Bus error trap
pid=11,pc=cfe56,ps=2204,sfc=1,dfc=1

trace:

_Debugger
_panic
faultstkadj
_ncr5380_data_xfer
_ncr5380_data_machine
_ncr5380_data_sched
_ncr5380_data_scsi_cmd
_scsi_scsipi_cmd
_ststart
_ststrategy
_physio
_stread
_spec_read
_ufsspec_read
_vn_read
_sys_read
_syscall
_trap0

Looks like some code somewhere can't handle the 1024k blocksize AIX
uses on tapes... Fairly often (ie, 3 out of 5 times) this problem
results in a hang, but the above was generated from single-user
mode panic.

I'm going to sup tonight (the recent SBC changes - could they make
any sort of difference??) and also build a ncrscsi kernel.
FYI, I'm using an SBC kernel as my SyQuest 200 is much happier with
it.

Thanks,
--
Paul Ripke
BHP Information Technology
OpenVMS, & UNIX (AIX, DG/UX, SCO, IRIX, Digital, SunOS...) Sysadmin
 Computer Centre,  Five Islands Rd,  Port Kembla,  NSW 2505,  AUSTRALIA
  stixpjr@ozemail.com.au   weripp@itwol.bhp.com.au   pjr02@uow.edu.au
         Anyone wishing to lay claim to the opinions expressed
                   herein, do so at their own risk.

------------- End Forwarded Message -------------