stupidcomputer
/
desmos-computer
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
desmos-computer/cli/data/computer.py

116 lines
3.1 KiB
Python
Raw Normal View History

massive project restructure relocated everything in its own module, `cli`, and split the server code such that it's more component based, if you will also, added a working test suite, which is somewhat exciting.
2024-06-16 18:18:28 -05:00
payload = """
make the isa transpiler into a module there's some other stuff as well
2024-02-01 17:00:28 -06:00
ticker 1 : main
add main files used before version control
2023-11-14 22:33:34 -06:00
# Main memory structure
massive project restructure relocated everything in its own module, `cli`, and split the server code such that it's more component based, if you will also, added a working test suite, which is somewhat exciting.
2024-06-16 18:18:28 -05:00
# This should be filled in with an override
id testing : B = []
add main files used before version control
2023-11-14 22:33:34 -06:00
# Operations on memory
port over the entire computer into computer.desmos
2023-11-15 20:01:35 -06:00
: q(l, v, i)=ifval(l, i, [1...length(l)], v)
: setlistval(index, v) = B -> q(B, v, index)
: incx(index, v) = setlistval(index, B[index] + v)
: ifval(l, index, c, v) = {c = index : v, l[c]}
# Operations
: oadd(x, y) = x + y
: osub(x, y) = x - y
: odiv(x, y) = x / y
: omul(x, y) = x * y
# Instruction implementation
: ijmp(a, b, c) = ip -> a, jumped -> 1
: iadd(a, b, t_o) = setlistval(t_o, oadd(B[a], B[b]))
: isub(a, b, t_o) = setlistval(t_o, osub(B[a], B[b]))
: idiv(a, b, t_o) = setlistval(t_o, odiv(B[a], B[b]))
: imul(a, b, t_o) = setlistval(t_o, omul(B[a], B[b]))
add more tests for the ISA; ISA bugfix added more tests to the ISA -- covers all the instructions except RST (reset). fixed a bug in the ISA -- when comparing two numbers, we really compared their pointers. in C, it's the difference between ``` int a = 10; int b = 12; int *ap = &a; int *bp = &b; if (ap < bp) { /* do stuff */ } ``` the problem is that you're comparing the pointers, *not* the values they refer to. it instead should be like if (*ap < *bp) { /* do stuff */ } because you need to *dereference* the value before comparing. that's what happens in cli/data/computer.py now.
2024-06-16 19:05:58 -05:00
: icmp(a, b, z) = { \\
B[a] = B[b] : equals -> 1 , \\
B[a] > B[b] : greater -> 1 , \\
B[a] < B[b] : less -> 1 \\
make the isa transpiler into a module there's some other stuff as well
2024-02-01 17:00:28 -06:00
}
port over the entire computer into computer.desmos
2023-11-15 20:01:35 -06:00
: irst(a, b, c) = equals -> 0, greater -> 0, less -> 0
: ield(addr, b) = setlistval(addr, equals)
: igld(addr, b) = setlistval(addr, greater)
: illd(addr, b) = setlistval(addr, less)
make the isa transpiler into a module there's some other stuff as well
2024-02-01 17:00:28 -06:00
port over the entire computer into computer.desmos
2023-11-15 20:01:35 -06:00
: ibe(addr, b, c) = {equals = 1 : ijmp(addr, 0, 0), jumped -> 1}
: ibne(addr, b, c) = {equals = 0 : ijmp(addr, 0, 0), jumped -> 1}
: ibg(addr, b, c) = {greater = 1 : ijmp(addr, 0, 0), jumped -> 1}
: ibl(addr, b, c) = {less = 1 : ijmp(addr, 0, 0), jumped -> 1}
: isto(v, addr) = setlistval(addr, v)
temp commit 1; needs to be merged
2024-03-24 18:50:55 -05:00
: imov(from, target) = setlistval(target, B[from])
port over the entire computer into computer.desmos
2023-11-15 20:01:35 -06:00
# registers
# instruction pointer
: ip = 1
# is the result of icmp equal?
: equals = 0
# ditto for greater than
: greater = 0
# ditto for less than
: less = 0
# instruction loading areas
: inst = 0
# next three values after the instruction
: paramone = 0
: paramtwo = 0
: paramthree = 0
# main execution flows
add more tests for the ISA; ISA bugfix added more tests to the ISA -- covers all the instructions except RST (reset). fixed a bug in the ISA -- when comparing two numbers, we really compared their pointers. in C, it's the difference between ``` int a = 10; int b = 12; int *ap = &a; int *bp = &b; if (ap < bp) { /* do stuff */ } ``` the problem is that you're comparing the pointers, *not* the values they refer to. it instead should be like if (*ap < *bp) { /* do stuff */ } because you need to *dereference* the value before comparing. that's what happens in cli/data/computer.py now.
2024-06-16 19:05:58 -05:00
: load(addr) = jumped -> 0, inst -> B[addr], \\
paramone -> B[addr + 1], \\
paramtwo -> B[addr + 2], \\
port over the entire computer into computer.desmos
2023-11-15 20:01:35 -06:00
paramthree -> B[addr + 3]
add more tests for the ISA; ISA bugfix added more tests to the ISA -- covers all the instructions except RST (reset). fixed a bug in the ISA -- when comparing two numbers, we really compared their pointers. in C, it's the difference between ``` int a = 10; int b = 12; int *ap = &a; int *bp = &b; if (ap < bp) { /* do stuff */ } ``` the problem is that you're comparing the pointers, *not* the values they refer to. it instead should be like if (*ap < *bp) { /* do stuff */ } because you need to *dereference* the value before comparing. that's what happens in cli/data/computer.py now.
2024-06-16 19:05:58 -05:00
: exec = { \\
inst = sto : isto(paramone, paramtwo), \\
inst = mov : imov(paramone, paramtwo), \\
inst = add : iadd(paramone, paramtwo, paramthree), \\
inst = cmp : icmp(paramone, paramtwo, paramthree), \\
inst = eld : ield(paramone, paramtwo), \\
inst = gld : igld(paramone, paramtwo), \\
inst = lld : illd(paramone, paramtwo), \\
inst = jmp : ijmp(paramone, paramtwo, paramthree), \\
inst = be : ibe(paramone, paramtwo, paramthree), \\
inst = bne : ibne(paramone, paramtwo, paramthree), \\
inst = bg : ibg(paramone, paramtwo, paramthree), \\
inst = bl : ibl(paramone, paramtwo, paramthree), \\
inst = sub : isub(paramone, paramtwo, paramthree), \\
inst = mul : imul(paramone, paramtwo, paramthree), \\
inst = div : idiv(paramone, paramtwo, paramthree) \\
port over the entire computer into computer.desmos
2023-11-15 20:01:35 -06:00
}
: incip = {jumped = 0 : ip -> ip + instwidth[inst] + 1}
# execution occurs here
: execution = 0
: jumped = 0
: loop = {execution = 0 : execution -> 1, execution = 1 : execution -> 2, execution = 2 : execution -> 0}
: loopaction = {execution = 0 : load(ip), execution = 1 : exec, execution = 2 : incip}
make the isa transpiler into a module there's some other stuff as well
2024-02-01 17:00:28 -06:00
: main = loopaction, loop
port over the entire computer into computer.desmos
2023-11-15 20:01:35 -06:00
: sto = 1
temp commit 1; needs to be merged
2024-03-24 18:50:55 -05:00
: mov = 17
port over the entire computer into computer.desmos
2023-11-15 20:01:35 -06:00
: add = 2
: cmp = 3
: eld = 4
: gld = 5
: lld = 6
: jmp = 7
: be = 8
: bne = 9
: bg = 10
: bl = 11
: sub = 12
: mul = 13
: div = 14
: rst = 15
add more tests for the ISA; ISA bugfix added more tests to the ISA -- covers all the instructions except RST (reset). fixed a bug in the ISA -- when comparing two numbers, we really compared their pointers. in C, it's the difference between ``` int a = 10; int b = 12; int *ap = &a; int *bp = &b; if (ap < bp) { /* do stuff */ } ``` the problem is that you're comparing the pointers, *not* the values they refer to. it instead should be like if (*ap < *bp) { /* do stuff */ } because you need to *dereference* the value before comparing. that's what happens in cli/data/computer.py now.
2024-06-16 19:05:58 -05:00
: instwidth = [2,3,2,1,1,1,1,1,1,1,3,3,3,3,0,2,2]
massive project restructure relocated everything in its own module, `cli`, and split the server code such that it's more component based, if you will also, added a working test suite, which is somewhat exciting.
2024-06-16 18:18:28 -05:00
"""