A successor to an older project called 'Asmpy', meant to be actually in line with RISC-V.
I made Asmpy when I was less experienced with making interpreters and Python in general, so it's badly written.
RISCY is much better written since I have way more experience now!(if you're a Python expert reading this, I'm so sorry)
And not only that, it is(or will be, at least) be written in multiple languages.
The only things that are different from RISC-V are the 'n'(Next) register
# adds x0 + 10 to the next register that is 0
ADDI n, x0, 10
# adds x1 + 22 to the next register that is 20
ADDI n20, x1, 22
# if the given number to check for doesn't exist in any register, then it will default to zero and print a warning to the console
# using the 'n' register is completely optional, of course
And that commas nor capitalization of instructions are needed.
# valid
ADDI x0, x0, 10
# also valid
addi x1 x1 10
# things like this are also valid but that's just disgusting
# aDdI, x2 X2, 10
| Inst | Name | Summary |
|---|---|---|
| add | ADD | rd = rs1 + rs2 |
| sub | SUB | rd = rs1 - rs2 |
| and | AND | rd = rs1 & rs2 |
| or | OR | rd = rs1 |
| xor | XOR | rd = rs1 ^ rs2 |
| sll | Shift Left Logical | rd = rs1 << rs2 |
| srl | Shift Right Logical | rd = rs1 >> rs2 |
| slt | Set Less Than | rd = (rs1 < rs2)?1:0 |
| addi | ADD Immediate | rd = rs1 + imm |
| andi | AND Immediate | rd = rs1 & imm |
| ori | OR Immediate | rd = rs1 |
| xori | XOR Immediate | rd = rs1 ^ imm |
| slli | Shift Left Logical Imm | rd = rs1 << imm |
| srli | Shift Right Logical Imm | rd = rs1 >> imm |
| slti | Set Less Than Imm | rd = (rs1 < imm)?1:0 |
| beq | Branch == | if(rs1 == rs2) PC += imm or label |
| bne | Branch != | if(rs1 != rs2) PC += imm or label |
| blt | Branch < | if(rs1 < rs2) PC += imm or label |
| bqe | Branch >= | if(rs1 >= rs2) PC += imm or label |
Currently, the only interpreter I've made is in Python, but I'll be making one in Go later.
I might also make some in other languages(maybe in Rust? 😜).
The CRISP(no, I'm not British) format is a bytecode format for the RISCY interpreters.
While yes, it is entirely useless, it does have some perks.
It's smaller than a normal .s file because it forfeits all those useless things like 'spaces' and 'comments'(ew), allowing for less space used.
TL;DR, it's like Java bytecode, but it's not Java so it's automatically superior.
Currently the only CRISP compiler I've made is in Python, etc etc.
I honestly don't really understand a lot of stuff.
Like, why does the SAR_example not use any registers lower than x18?
And what does the LUI instruction do, exactly?
so if you have the explaination, I would appreciate if you made an issue about it!