> -----Original Message-----
> From: IBM Mainframe Assembler List <ASSEMBLER-
> [email protected]> On Behalf Of Seymour J Metz
> Sent: 20 December 2020 17:12
> To: [email protected]
> Subject: Re: S/360 emulation in PC/370
> 
> The 360/30, 360/40 and 360/50 had integrated channels, cycle stealing from
> the CPU microcode. The channel-control unit-device structure was very
> similar to that of the 7000 series and the 7030. When the S/360 came out,
> channel I/O was the norm, and Interrupts go back to the 1950s.
> 

But  its quite alien to those of a certain age brought up on 8 and 16-bit
Microprocessors. The 6502 and 6800/09 only have memory IO. 
The 8080 and other family chips do have a separate IO space, but it uses the
same data bus. Essentially its memory mapped IO.

The fact that on S/360 a single IO can store information in multiple
disparate locations.
Things like the serial lines where you can see data arriving....

> Small machines like the 650 and 1401 had simpler I/O, but I don't recall
> anything older than the PDP-11 that had memory-mapped I/O.
> 

The Ferranti Pegasus I, a valve/tube computer dating from 1956 had memory
mapped IO 

https://en.wikipedia.org/wiki/Ferranti_Pegasus

there were two memory locations for input, and two for output. 
This allows "normal" binary IO where you send the binary value of the
character, and numeric IO where the bottom 4 bits are interpreted as an
integer and the corresponding characters 0-9 output or input 
The numeric output was necessary as they used a an odd character set for the
teleprinters, not IA2. 

> 
> --
> Shmuel (Seymour J.) Metz
> http://mason.gmu.edu/~smetz3
> 

Dave
G4UGM

Reply via email to