Created
June 12, 2016 17:50
-
-
Save daeken/b1d037684ab3f85377e350eba0fd7d8a to your computer and use it in GitHub Desktop.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
import re | |
def regex(pattern, flags=0): | |
def sub(text): | |
match = pattern.match(text) | |
if match == None: | |
return False | |
return match.groups() | |
pattern = re.compile(pattern, flags) | |
return sub | |
labels = {} | |
mlabel = regex(r'^\s*([a-zA-Z0-9.]+):\s*$') | |
mint = regex(r'^(-?[0-9]+)$') | |
mref = regex(r'^(-?[0-9]+)\(\$([0-9]+)\)$') | |
def parseOperand(op): | |
if op in labels: | |
return [labels[op]] | |
elif op.startswith('$'): | |
return [int(op[1:])] | |
elif op.startswith('0x'): | |
return [int(op[2:], 16)] | |
elif op.startswith('-0x'): | |
return [-int(op[3:], 16)] | |
elif mint(op): | |
return [int(op)] | |
elif mref(op): | |
return list(map(int, mref(op))) | |
else: | |
print 'unknown', op | |
def processInsn(mnem, ops): | |
if mnem == 'nop': | |
return [('sll', ('0', '0', '0'))] | |
else: | |
return [(mnem, ops)] | |
def assemble(base, code): | |
insns = [] | |
for line in code.split('\n'): | |
line = line.split('#', 1)[0] | |
match = mlabel(line) | |
if match: | |
label = match[0] | |
labels[label] = base + len(insns) * 4 | |
elif line.strip(): | |
line = line.strip().split(' ', 1) | |
if len(line) == 1: | |
mnem, rest = line[0], '' | |
else: | |
mnem, rest = line | |
ops = [x.strip() for x in rest.split(',') if x.strip()] | |
insns += processInsn(mnem.lower(), ops) | |
out = [] | |
for i, (mnem, ops) in enumerate(insns): | |
ops = reduce(lambda a, b: a+b, map(parseOperand, ops)) | |
out.append(buildInsn(base + i * 4, mnem, ops)) | |
return out | |
def unsigned(val, size): | |
if val >= 0: | |
return val | |
else: | |
return (1 << size) + val | |
def buildInsn(addr, mnem, ops): | |
if mnem not in indefs: | |
print 'Unknown instruction:', mnem, ops | |
return 0 | |
inst = 0 | |
type, odef, params = indefs[mnem] | |
if type == IType: | |
inst |= odef << 26 | |
for i, name in enumerate(params): | |
val = ops[i] | |
if name == rs: | |
inst |= unsigned(val, 5) << 21 | |
elif name == rt: | |
inst |= unsigned(val, 5) << 16 | |
elif name == imm: | |
inst |= unsigned(val, 16) | |
elif type == JType: | |
inst |= odef << 26 | |
for i, name in enumerate(params): | |
val = ops[i] | |
if name == target: | |
inst |= (val & 0x0FFFFFFF) >> 2 | |
elif type == RType: | |
inst |= odef[0] << 26 | |
inst |= odef[1] | |
for i, name in enumerate(params): | |
val = ops[i] | |
if name == rs: | |
inst |= unsigned(val, 5) << 21 | |
elif name == rt: | |
inst |= unsigned(val, 5) << 16 | |
elif name == rd: | |
inst |= unsigned(val, 5) << 11 | |
elif name == shamt: | |
inst |= unsigned(val, 5) << 6 | |
else: | |
print 'Unknown type:', type | |
return inst | |
IType, JType, RType = 'itype', 'jtype', 'rtype' | |
imm, offset, rd, rs, rt, shamt, target = 'imm', 'offset', 'rd', 'rs', 'rt', 'shamt', 'target' | |
indefs = dict( | |
add=(RType, (0b000000, 0b100000), (rd, rs, rt)), | |
addi=(IType, 0b001000, (rt, rs, imm)), | |
j=(JType, 0b000010, (target, )), | |
lb=(IType, 0b100000, (rt, offset, rs)), | |
lbu=(IType, 0b100100, (rt, offset, rs)), | |
sll=(RType, (0b000000, 0b000000), (rd, rt, shamt)), | |
) | |
import disasm | |
def disassemble(base, insns): | |
for i, insn in enumerate(insns): | |
print '%08x %s' % (base + i * 4, disasm.disassemble(base + i * 4, insn)) | |
print disassemble(0x80100000, assemble(0x80100000, ''' | |
j .part1 | |
j .part2 | |
addi $3, $0, 1 | |
.part1: | |
addi $1, $0, 1 | |
j end | |
nop | |
.part2: | |
addi $2, $0, 1 | |
end: | |
nop | |
''')) | |
print disassemble(0x80100000, assemble(0x80100000, ''' | |
lb $1, 0($30) | |
lbu $2, 0($30) | |
#lh $3, 0($30) | |
#lw $4, 0($30) | |
''')) |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment