GatoROM Z8 Format

[travisgoodspeed]

Close this issue when we have support for decoding ROMs from Z8 chips of multiple ROM sizes.

[Lord-Nightmare]

I figured this out using a number of guesses, some hints from the micrograph in the July 1981 issue of BYTE magazine, and a Log table at the end of the rom as a rosetta stone for the rather weird column order.

The Z8611 rom, if arranged so that the rom is at the bottom of the die (which is technically upside-down compared to the image in the july 1981 BYTE article), is arranged as such:

ROM OVERVIEW:

1 vs 0 bits: an open link (blue dot in maskromtool) is a 0 bit, a closed link (red dot in maskromtool) is a 1 bit

| D0   D1   D2   D3 || D4   D5   D6   D7 |
|.... .... .... ....||.... .... .... ....|
|.... .... .... ....||.... .... .... ....|
|.... .... .... ....||.... .... .... ....|
|.... .... .... ....||.... .... .... ....|
|.... .... .... ....||.... .... .... ....|
|.... .... .... ....||.... .... .... ....|
|.... .... .... ....||.... .... .... ....|
|.... .... .... ....||.... .... .... ....|
|.... .... .... ....||.... .... .... ....|
|.... .... .... ....||.... .... .... ....|
|.... .... .... ....||.... .... .... ....|
|.... .... .... ....||.... .... .... ....|
|.... .... .... ....||.... .... .... ....|
|.... .... .... ....||.... .... .... ....|
|.... .... .... ....||.... .... .... ....|
|.... .... .... ....||.... .... .... ....|

Column order:

Detail of the D0-D3 columns:
|                                                D0                                              ||    D1   ...
#  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f 10 11 12 13 14 15 16 17 18 19 1a 1b 1c 1d 1e 1f
c 0f 1f 07 17 0b 1b 03 13 0d 1d 05 15 09 19 01 11 0e 1e 06 16 0a 1a 02 12 0c 1c 04 14 08 18 00 10 


Detail of the D4-D7 columns: (same as D0-D3 but mirrored left to right)
...  D6    ||                                             D7                                                |
           #  0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f 10 11 12 13 14 15 16 17 18 19 1a 1b 1c 1d 1e 1f
           c 10 00 18 08 14 04 1c 0c 12 02 1a 0a 16 06 1e 0e 11 01 19 09 15 05 1d 0d 13 03 1b 0b 17 07 1f 0f|


ROM address is column 'number' plus 32 * row number

Row order from top: (T0 and T1 are the two rows of the 64-byte test rom)
T1
T0
00
01
03
02
04
05
07
06
08
09
0b
0a
...
7b
7d
7c
7e
7f

[Lord-Nightmare]

Effectively this means, to find a specific bit in the array given its rom address, you'd do something like:
D0-D3: final_column = BITSWAP8((column^0x1E), 7, 6, 5, 0, 1, 2, 3, 4);
D4-D7: final_column = BITSWAP8((column^0x01), 7, 6, 5, 0, 1, 2, 3, 4);

The BITSWAP8 function is:

#define BITSWAP8(val,B7,B6,B5,B4,B3,B2,B1,B0) \
		((BIT(val, B7) <<  7) | \
		 (BIT(val, B6) <<  6) | \
		 (BIT(val, B5) <<  5) | \
		 (BIT(val, B4) <<  4) | \
		 (BIT(val, B3) <<  3) | \
		 (BIT(val, B2) <<  2) | \
		 (BIT(val, B1) <<  1) | \
		 (BIT(val, B0) <<  0))

[Lord-Nightmare]

I think it is likely that the Z8601 ROM on the die is in the same order as the Z8611, with the sole difference that the "bottom-most" row is 3f instead of 7f, no other significant difference.

[travisgoodspeed]

Sean Riddle just published his notes on the Z8.

[travisgoodspeed]

#77 tracks a patch that adds support.

[travisgoodspeed]

We still lack support for exporting the test ROM, but this seems to work for everything else.