Files for my build of Ben Eater's excellent 6502 breadboard computer. More from him at https://eater.net/6502.
With reference to https://www.reddit.com/r/beneater/comments/doytpo/6502_project_memory_map/
| Chip | Range | Description |
|---|---|---|
| RAM | $0000 - $00ff |
Zero page |
| RAM | $0100 - $01ff |
Stack |
| RAM | $0200 - $3fff |
General RAM |
| Open | $4000 - $5fff |
Not mapped to a device |
| 6522 | $6000 |
I/O Register B |
| 6522 | $6001 |
I/O Register A |
| 6522 | $6002 |
Data Direction Register B |
| 6522 | $6003 |
Data Direction Register A |
| 6522 | $6004 |
T1 Low Order Latches/Counter |
| 6522 | $6005 |
T1 High Order Counter |
| 6522 | $6006 |
T1 Low Order Latches |
| 6522 | $6007 |
T1 High Order Latches |
| 6522 | $6008 |
T2 Low Order Latches/Counter |
| 6522 | $6009 |
T2 High Order Counter |
| 6522 | $600a |
Shift Register |
| 6522 | $600b |
Auxiliary Control Register |
| 6522 | $600c |
Peripheral Control Register |
| 6522 | $600d |
Interrupt Flag Register |
| 6522 | $600e |
Interrupt Enable Register |
| 6522 | $600f |
I/O Register A sans Handshake |
| 6522 | $6010 - 7fff |
Mirrors of the sixteen VIA registers |
| 28C256 | $8000 - $ffff |
ROM |
The machine code is put on the ROM starting at $0 (relative to the actual ROM address)
The .org directive tells the assembler where the CPU thinks the address will be.
-
Starting the assembly code with
.org $8000tells the assemlber that the CPU address space for the ROM starts at$8000 -
Subsequent
.orgdirectives will place data into the ROM at the specified address (in CPU address space)
Here is an annotated version of blink.s:
.org $8000 ; Though written to the first address in the ROM, $0,
; this code will appear to the CPU to be at $8000
reset: ; This label marks the first position in the ROM for
; the CPU is $8000
lda #$ff
sta $6002
lda #$50
sta $6000
loop:
ror
sta $6000
jmp loop
.org $fffc ; Specify that the following code will go at the position
; that appears to the CPU to be at $fffc. $fffc is the
; location of the reset vector. The value at this vector
; is loaded into the program counter after a CPU reset.
.word reset ; Place the value of the label reset, ie $8000, at this
; position
.word $0000 ; Pad out the last couple bytes
There are a couple tools needed to build the binaries and get them on the ROM. Here is how to install them on linux.
To install vasm: wget http://sun.hasenbraten.de/vasm/release/vasm.tar.gz tar xvzf vasm.tar.gz cd vasm make CPU=6502 SYNTAX=oldstyle cp vasm6502_oldstyle /usr/bin cp vobjdump /usr/bin
To install minipro: git clone https://gitlab.com/DavidGriffith/minipro.git cd minipro make make install
I've slightly modified some of the commands and code to have incremental versions of some of the programs he iterates on.
There are no commands for video 1. All the software is hardwired resistors.
"My favorite programming language is solder." - Todd Whitehurst (or Bob Pease?)
Commands for video 2.
First, the bin with just ea values.
python makerom_ea.py
hexdump -C rom_ea.bin
minipro -p AT28C256 -w rom_ea.bin
Now the full binary.
python makerom.py
hexdump -C rom.bin
minipro -p AT28C256 -w rom.bin
Commands for video 3.
vasm6502_oldstyle -Fbin -dotdir blink.s
hexdump -C a.out
minipro -p AT28C256 -w a.out
Commands for video 4.
vasm6502_oldstyle -Fbin -dotdir hello-world_vid04.s
hexdump -C a.out
minipro -p AT28C256 -w a.out
Commands for video 5.
vasm6502_oldstyle -Fbin -dotdir hello-world_vid05.s
hexdump -C a.out
minipro -p AT28C256 -w a.out
Commands for video 7.
vasm6502_oldstyle -Fbin -dotdir hello-world_vid07.s
hexdump -C a.out
minipro -p AT28C256 -w a.out
Why build an entire computer on breadboards?
No new code.
How assembly language loops work
hello-world_vid09.s
Binary to decimal can’t be that hard, right?
At about 1:31 in the video, he shows a file I here call vid10_42.s.
Then at 19:10 in the video, he starts making edits for what I call vid10_bin2dec.s. In the video, it is called bin2dec.s.
The file he writes in this video is referred to here as vid11_interrupts.s.
Code is in vid12_interrupts.s.