> On Nov 1, 2022, at 3:38 PM, Ethan Dicks via cctalk <cctalk@classiccmp.org>
> wrote:
>
> On Tue, Nov 1, 2022 at 12:01 PM Paul Koning via cctalk
> <cctalk@classiccmp.org> wrote:
>>> On Oct 30, 2022, at 2:49 PM, Wayne S via cctalk <cctalk@classiccmp.org>
>>> wrote:
>>>
>>> The difference between dz and dh interfaces is that the dh used dma instead
>>> of interrupts to get characters to the cpu...
>>
>> No, it doesn't. I was confused about this but was recently corrected.
>>
>> The DH11 does DMA output, but not DMA input. I don't know any DEC serial
>> port devices that have DMA input; it would make very little sense to do that
>> since input generally is one character at a time.
>
> Yes. I was going to mention this in my previous reply but got sidetracked.
>
> The big benefit for DH11 and DMF32 and 3rd-party DH11 work-alikes
> (Emulex CS-21...) is that since under normal workflow, many times more
> chars go out than come in so DMA-out saves a lot of overhead when
> blasting screens of stuff (like refreshing your page in EDT...) and
> people don't type all that fast by comparison.
>
> Where we used to have problems is having multiple Kermit sessions on
> our serial ports. Those hammer both ways.
>
> Fortunately, I wasn't trying to support PDP-11s with split baud rates
> like 1200/150 that were used to _really_ reduce input interrupt
> frequency. Our machines kept up at 9600/9600.
A DEC product that shows this sort of scenario is Typeset-11 (TMS-11) which has
VT71 block transfer terminals at 9600 bps. I'm 98% sure those were on DH11s.
So with those, when the user hit the key to send the completed file back to the
host, a bulk transfer of whatever the file size was (a complete newspaper
article, in the typical case) would occur. I'm not sure if DDCMP was used for
that, probably yes. It certainly adds up to a pretty significant workload. I
suppose there was an occasional FIFO overflow under peak load (reporters
frantically finishing articles as the deadline approaches...)
paul
