> On Aug 10, 2017, at 12:24 PM, Ulrich Tagge <u.ta...@gmx.de> wrote:
>
>
>
> Am 10.08.2017 um 17:19 schrieb Paul Koning:
>>> On Aug 9, 2017, at 2:58 PM, Ulrich Tagge via cctalk <cctalk@classiccmp.org>
>>> wrote:
>>>
>>>
>>> Type a command then press the RETURN key: B MS0
>>> Trying MS0
>>> 001120
>>> @
>> You're getting an I/O error halt. That case is supposed to go to the HALT
>> at 1114, which would mean you'd get a PC of 1116 shown in the message.
>>
>> However... there's a bug in the bootstrap as documented in the manual. It
>> shows BMI HALT at 1050, but the machine code is wrong. The displacement
>> from the PC is 1114-1052 = 42, so the assembled instruction would be 100421.
>> As written, the code is a BMI MES.
>>
>> You should examine the TS controller CSR at 172522 (TSSR) to see what error
>> flags are set.
>>
>> paul
>>
>>
>
> Hi Paul,
>
> Many thanks for you Help!
> Here is the content of 172522.
>
> Trying MS0
> 001120
> @
> @172522/052525
Does the console on your machine do 22 bit addressing? You may need to ask for
the CSR by its full 22 bit address, so 17772522. The number you showed looks
suspiciously like a powerup selftest memory check pattern.
> Please excuse, maybe my understanding is wrong: Does this mean content of
> address 001050 should be 100421 instead of 100422?
Yes, that's what I meant. Otherwise it branches to the first word of the
message buffer, and if that's non-zero it wouldn't halt but do something
different and probably rather strange.
paul
