This project aims to implement the UL8-CPU (useless 8-bit central processing unit) with corresponding assembler. Inspired by a cumputer-model I first was introducted to at uni, this CPU, together with its custom ISA, shows what an extremly simple programmable computer could look like. It gets its name from beeing, altough theoretically functional, completely useless in any practical scenario.
It is designed to be used on the Tang Nano 9k FPGA-board (featuring a GOWIN GW1NR-9 FPGA) paired with the 1.14inch LCD. Currently, its only way of beeing programmed is by inculding the program in the synthesis and PnR for the FPGA. The program is stored in text in the program file, which currenlty holds a demo program that calculates 122 mod 7.
- 32 Bytes of RAM
- 8 Bit Data-Word length
- fixed 8-Bit Instruction length
An instruction consists of the op-code in the three most significant bits and a modifier a in the others bits.
| Instruction | Op-Code | Meaning (with memory-adress a) |
|---|---|---|
| LDA a | 000 | <a> -> Akku |
| STV a | 001 | Akku -> a |
| ADD a | 010 | Akku + <a> -> Akku |
| NAND a | 011 | Akku NAND <a> -> Akku |
| JMP a | 100 | <a> -> PC |
| JMN a | 101 | if MSB of Akku = 1: <a> -> PC |
| NOT | 110 | NOT Akku -> Akku |
| HLT | 111 | halts the machine |
An assembler for the UL8-ISA as defined above.
The assembler only accepts files with the .ul8 extension. Use it by calling
./ul8_asm <file>
in it's directory, where <file> is the path to the .ul8-file you want to compile.
A UL8-assembly-file is made up of 2 distinct sections: .data and .start.
In .data, the user is able to declare identifiers, wich the user can use to refer to memory-addresses without explicitly declaring said address in the code. The .data section is also used to initialize the content at these memory-addresses. The initialization-values are read as decimal numbers. Each Identifier defines exactly one byte.
.data:
<identifier> <value>
Example:
.data:
foo 0
bar 42
Here, identifier foo is initialized to 0 and identifier bar is initialized to 42.
This is where the code will start executing. It also defines the first (and maybe only) block of code. Blocks of code are defined by labels. A label is made of the following syntax:
<label>:
They can be used as arguments for unconditional or conditional jump-instructions. The block of code belonging to a label begins at the instruction following the label and ends at the next label or the end of the file. When jumped to, for example at start of at the .start-label, execution will resume on the label's block of code.
An instruction is made from the following syntax:
<instruction-code> <argument>
There is exactly one instruction per line. A complete listing of instructions can be found above.
This program calculates 42 mod 7.
.data:
foo 42
bar 7
baz 1
.start:
LDA bar
NOT
ADD baz
STV baz
LDA foo
loop:
JMN exit
ADD baz
JMP loop
exit:
ADD bar
HLT