gmp@rayssd.UUCP (Gregory M. Paris) (06/10/85)
We have recently connected an SI 9753 tri-density tape drive to two of our VAX 11/780's running 4.2BSD. Switch the rocker switch to connect to either machine then power up and voila, the VAX hangs with "too many zero vectors." One of people at SI's California office tells us that this happens because we have the tape drive connected to TR 12 (on both machines). He claims that UNIX cannot handle TR interrupt levels above 11. We tried running VMS on one of the VAX'es and it was able to access the drive successfully. Later this week we are going to reconnect the drive at TR 10 for experimentation, but we won't be able to leave it that way. Does anyone know whether or not this really is a UNIX limitation? Does anyone have a fix for it? Please respond by mail and I'll summarize replies to the net. Thanks! -- ++--------------------------------------------------------------------------++ || Gregory M. Paris || || ...!{allegra,linus,raybed2,ccice5,brunix}!rayssd!gmp || ++--------------------------------------------------------------------------++
gmp@rayssd.UUCP (Gregory M. Paris) (06/14/85)
> We have recently connected an SI 9753 tri-density tape drive to > two of our VAX 11/780's running 4.2BSD. Switch the rocker switch > to connect to either machine then power up and voila, the VAX > hangs with "too many zero vectors." > > One of people at SI's California office tells us that this happens > because we have the tape drive connected to TR 12 (on both machines). > He claims that UNIX cannot handle TR interrupt levels above 11. As a follow-up to my original article (abbreviated above), the following new information: 1) The SI 9753 DOES work at TR 10. 2) Our SI 9900 controllers also fail at TR 12, with the error "uba0: too many zero vectors" "uba0: reset dh0 dh1 dh2 lp0 dmc0" but only upon actual access of the drives connected to that controller (three SI 9751 eagles). This would seem to indicate that the SI people are correct. UNIX cannot handle TR 12 and higher. Again I ask (as I got no response to my first posting), "Does anyone have a fix?" Is anyone interested in a fix if WE come up with one? -- ++--------------------------------------------------------------------------++ || Gregory M. Paris || || ...!{allegra,linus,raybed2,ccice5,brunix}!rayssd!gmp || ++--------------------------------------------------------------------------++