IDE Qbus controller (was TU-58s)

[Allison]

>
>Subject: IDE Qbus controller (was TU-58s)
>   From: "Jerome H. Fine" <jhfinedp3k at compsys.to>
>   Date: Tue, 17 Apr 2007 22:19:15 -0400
>     To: General Discussion: On-Topic and Off-Topic Posts <cctalk at classiccmp.org>
>
> >Allison wrote:
>
>>I have a few contollers (dual width) that are Both MFM and 
>>SCSI that sound like those.  
>>
>>I keep putting it on my list of projects to do a simple IDE
>>for QBUS.  the design goals would be dual width, boot rom on 
>>board and uses a 2.5" drive on the card.  So far I've only 
>>seen one Qbus IDE and it was lacking for software. Software 
>>driver for that hardware is for RT11 alone is a bit of a 
>>project as I'd need both the FB and SJ versions of the 
>>driver.
>>
>Jerome Fine replies:
>
>Device drivers for RT-11 are identical for FB and SJ
>(or SB) monitors.  The XM (RT11XZ monitors use the same
>device drivers as XM) device drives are a bit different.

I know that.  That was just a typo.

>If you feel that MSCP emulation (probably OK now that
>the patent has expired) is too much trouble, then
>perhaps the HD(X).SYS protocol from E11 would be
>easier.  I suspect that the protocol is so basic,
>the concept might be included within other interface
>such as for RK05 or even a floppy.

Devices that arent MSCP like DD(tu58), DK(RX02), DY(RX02), 
DL(RL02) might make for examples. I'm fortunate to have 
the uncut sources on RL02.  Unfortunately I'm not an 
experienced PDP-11 programmer. The upside is I have the 
RT-11 docset.

>Whatever interface protocol you use, if you want to
>extend the number of RT-11 devices that are allowed,
>I would be very interested in looking at allowing
>up to 256 devices using MSCP under RT-11.  Naming
>might be the problem:  D00: => D77: where the
>numbers seem to be octal allows up to 64 devices.
>Using D00: => DFF: where the second and third character
>seem to be hex would allow 256 devices.  Alternatively,
>using D00: => D7V: where the third character has
>32 values including 0 => 9 and A => V.

Thats beyond me.

>The other possibility is to use multiple sets of
>hardware registers that look like multiple controllers
>under RT-11.  And since even 256 RT-11 devices of
>32 MBytes each covers only 8 GBytes, multiple controllers
>may be required in addition to allowing 65536 RT-11
>partitions per drive by changing the table that
>holds the RT-11 partition number to a 16 bit word
>from an 8 bit byte.  The later should not really
>be a problem since the unit number is already limited
>to a single byte even though a 16 bit word is available.
>Swapping the unit number word with the partition
>number byte should be reasonable and quite simple.
>
>Producing RT-11 bug fixes and enhancements is on
>my list of priorities.  Y3K is at the top of the
>list.  Does anyone else want to participate?  There
>are only 92 more years left.  If one additional
>word is allowed for the date, that would mean that
>23 bits are available for the year.  Likely that
>should be enough for a while since the CE use of
>97 leap days out of 400 will certainly need to
>change before 8 million years.

I should live so long.;)


Allison