old microcode listings
|
Prev: Software vs Hardware
Next: Searching for the PDP-3
From: Eric P. on 4 Jan 2007 21:54 Peter \"Firefly\" Lund wrote: > > Which had an asynchronous protocol for memory and I/O access. Happilly, > they didn't implement it fully, so one could apparently get away with just > ground /DTACK instead of implementing Motorola's somewhat complicated > scheme. Of course, if you have a custom chip or enough PLAs, then it > doesn't matter. > > Please google "DTACK grounded" and read the first few paragraphs of the > first newsletter. Now it has been 30 years, but as far as I remember all microprocessor bus protocols had some form of NOT_READY signal to handle slow devices. If you know the device is fast enough you just ground the signal. How is this different? Eric
From: Eric Smith on 5 Jan 2007 02:14 "Peter \"Firefly\" Lund" <firefly(a)diku.dk> writes: > Does anybody know where I can find microcode listings for old Data > General, IBM S/360, or Norsk Data machines? Microcode listings for the System/360 and System/370 seem to be really hard to find. Someone does have a 360/30 microcode listing, but unfortunately is not willing to make it publicly available. If anyone does have microcode for any of those, and is willing to make it available, please speak up! The widest architecturally compatible families of processors for which microcode listings are readily available seem to be the PDP-10, PDP-11, and VAX. PDP-10: KL10 processor - many microcode versions available KS10 processor - several versions available Minnow processor (never entered production) PDP-11: PDP-11/04 PDP-11/05 PDP-11/34 PDP-11/40 PDP-11/44 PDP-11/45 PDP-11/60 PDP-11/70 DCJ11 chip set (two chips on a ceramic hybrid) VAX: VAX-11/780 MicroVAX II CVAX Rigel Mariah Raven NVAX > Microcode for the T-11 ("Tiny") PDP-11 microcontroller would also be > greatly appreciated. It would be interesting to see listings of the T11, F11, and 11/03 microcode, but no one seems to have them. The F11 CIS option (Commerical Instruction Set, for Dibol) listings still exist, but the person who has it does not have the base instruction set microcode. The WCS development software for the 11/03 apparently included the CIS microcode sources for that processor. I've considered trying to extract the microcode from the 11/03 and F11 chip sets. For the 11/03, it should be easy to dump the contents of the MICROM chips, but there are also two PLAs in the control chip that can't be easily extracted and affect microinstruction sequencing, so the MICROM contents alone would be nearly worthless. The best approach to extracting the PLAs is probably to photomicrograph the die. I don't currently know of anywhere that I can get a high- resolution photomicrograph on a hobby (<$500) budget. For the F11 chip set, things are more complicated, because the microcode is distributed among multiple control chips, each of which has its own microsequencer. It should be possible to extract the portions of microcode that are transferred on the bus to the data path chip, but the parts of the microcode word used for microsequencer control are not available externally to the control chip. Again, a photomicrograph may be the best way to dump them. Peter Monta has provided an existence proof of the feasibility of optically dumping microcode from 1970s-era chips with masked ROM: http://www.pmonta.com/calculators/hp-35/index.html He dumped the HP-35 calculator ROMs optically, and go them running on a simulator that I'd previously written for the HP-45 and HP-55.
From: Eric Smith on 5 Jan 2007 02:43 ChrisQuayle <nospam(a)devnul.co.uk> writes: > Yes it was, but probably the best that could be done at the time in > commodity devices - ok, bit slice in intent ?. I think the more > complex AMD 29xx series came quite a bit later than 1972. Sure, but there were bit slice components before the Am2900 series. The earliest I'm aware of were from Fairchild in 1968-1969. Year Vendor P/N description ---- --------- ---- --------------------------- 1968 Fairchild 3800 8-bit data path 1968 Fairchild 3804 4-bit data path 1972 National MM5750 4-bit data path 1972 National MM5751 sequencer and microcode ROM 1974 MMI 6701 4-bit data path, very similar to later Am2901 1974 MMI 6700 4-bit sequencer 1975 Intel 3002 2-bit data path 1975 Intel 3001 9-bit sequencer 1975 AMD 2901 4-bit data path 1975 AMD 2909 4-bit sequencer 1975 AMD 2911 4-bit sequencer 1976 TI 74S481 4-bit data path 1976 TI 74S482 sequencer 1977 AMD 2903 4-bit data path 1977 AMD 2910 10-bit sequencer 1977 Motorola 10800 4-bit data path, ECL 1977 Motorola 10801 sequencer, ECL ? MMI 67110 sequencer ? AMD 29203 4-bit data path ? TI 74AS888 8-bit data path ? TI 74AS890 sequencer ? TI SBP0400 4-bit data path, I2L ? AMD 29116 16-bit data path I'm missing information on the Fairchild Macrologic bitslice parts, which were available in both TTL and CMOS. There are probably some others I'm not aware of. I've seen conflicting reports over whether the Four-Phase AL1 (1970) should be considered a bit slice design. I need to track down a copy of the paper "Four-phase LSI logic offers new approach to computer designer" by L. Boysel and J. Murphy from the April 1970 issue of Computer Design.
From: "Peter "Firefly" Lund" on 5 Jan 2007 07:49 On Thu, 4 Jan 2007, Del Cecchi wrote: > You are going to build a VAX out of LSTTL just like back in the day? Yes. -Peter
From: "Peter "Firefly" Lund" on 5 Jan 2007 07:52
On Thu, 4 Jan 2007, Eric P. wrote: > Now it has been 30 years, but as far as I remember all microprocessor > bus protocols had some form of NOT_READY signal to handle slow devices. > If you know the device is fast enough you just ground the signal. > How is this different? Please google. -Peter |