Project generate RL02.

Tim Shoppa shoppa at
Wed Feb 25 02:29:38 AEST 1998

> <I added an RLV-controller for those RL02 drives into the system and
> I take it that is the rlv12 controller as the rlv11 twoboard set is not 
> usable. The rlv11 (m8014/8014) must be plugged into a h9273 backplane 
> only!

Actually, if you use an expansion BA23 it is possible to hook a
RLV11 set to a uVax.  Woe to the person that tries this, though, as there
are bad things in store when the system begins doing any Q-bus
transactions past the lower 248 Kbytes!  The 18-bit-ness of the
RLV11 is a very nasty form of 18-bit-ness, and not easily overcome
like, say, the RXV21.

> <What the heck is happening? Do I need to have the RL02 connected to get t
> <darn thing to interrupt or do I have a CSR / Interrupt conflict along the
> <line?
> likely yes, the drive must be connected.

Certainly yes, if the Ultrix autoconfigure logic is anything like
2.11 BSD's was before I got fed up with it insisting that I have
my RL units #0 when booting.  To quote from 2.11BSD Patch #380:

Subject: RL driver update, setvbuf(3) arrives in 2BSD, rdist fix (#380)

	1) 'autoconfig' only recognizes the RL controller if drive 0
	   is connected to the system at boot time.

	1) Boot a system with multiple RL drives, but with drive
	   0 not present.  'autoconfig' will not see an interrupt
	   from the RL subsystem during its probe of drive 0, and as
	   a result the rl driver will not be attached.

	1) Modify /sys/autoconfig/rlauto.c so it tests only for the
	   presence of the RL controller CSR, and doesn't wait for
	   an interrupt.  This is the same thing which is done in the TS11
	   probe routine because TS controllers can not be made to interrupt
	   reliably if a tape is not at the load point and the drive is not 


Received: (from major at localhost)
	by (8.8.5/8.8.5) id JAA04419
	for pups-liszt; Wed, 25 Feb 1998 09:16:19 +1100 (EST)
X-Authentication-Warning: major set sender to owner-pups at using -f

More information about the TUHS mailing list