box86/rebuild_printer.py

1994 lines
65 KiB
Python
Raw Normal View History

#!/usr/bin/env python3
import os
import re
import sys
from math import ceil
"""
Generates src/dynarec/arm_printer.c
===
See src/dynarec/arm_instructions.txt (the input file) for the syntax documentation.
"""
# Helper class to avoid displaying '\x1b[' on errors
class string(str):
def __repr__(self):
return str(self)
def nextAvailable(num):
return 8 if num <= 8 else (16 if num <= 16 else (32 if num <= 32 else 64))
def int2hex(i, finsz=-1):
if (finsz == -1) and (i == 0): return "0x0"
elif i == 0: return "0x" + ("0" * finsz)
ret = ""
while i != 0:
ret += str(i % 16) if i % 16 < 10 else chr(ord('A') + (i % 16) - 10)
i = i // 16
rl = len(ret)
if rl < finsz:
ret = ret + ("0" * (finsz - rl))
return "0x" + ''.join(reversed(ret))
def arr2hex(a, forceBin=False):
if forceBin:
return "0b" + ''.join(map(str, a))
else:
al = len(a)
return int2hex(sum(v * 2**(al - i - 1) for i, v in enumerate(a)), ceil(al / 4))
def sz2str(sz, forceBin=False):
return int2hex(2 ** sz - 1) if not forceBin else ("0b" + "1" * sz)
def main(root, ver, __debug_forceAllDebugging=False):
# Initialize variables
output = ""
# Debugging variable
invalidationCount = 0
tabCount = 1
def append(strg):
assert("\t" not in strg)
nonlocal output, tabCount
strg = strg.split("\n")
for s in strg[:-1]:
if s.endswith("{"):
tabCount = tabCount + 1
if s.startswith("}") and output.endswith("\t"):
tabCount = tabCount - 1
output = output[:-1]
output = output + s + "\n" + "\t" * tabCount
if strg[-1].startswith("}") and output.endswith("\t"):
tabCount = tabCount - 1
output = output[:-1]
output = output + strg[-1]
insts = None
# Read the instruction and exit if nothing changed since last run
with open(os.path.join(root, "src", "dynarec", "arm_instructions.txt"), 'r') as file:
insts = file.read()
# Get all actual instructions
# Ignore white lines and lines beginning with either !, ; or #
insts = list(filter(lambda l: not re.match("^\s*$", l) and not re.match("^([!;#])", l), insts.split('\n')))
try:
# Do not open with `with` to be able to open it in writing mode
last = open(os.path.join(root, "src", "dynarec", "last_run.txt"), 'r')
if '\n'.join(insts) == last.read():
last.close()
with open(os.path.join(root, "src", "dynarec", "last_run.txt"), 'w') as f:
f.write('\n'.join(insts))
return 0
last.close()
except OSError:
# No last run file
pass
for lnno, line in enumerate(insts):
ln = line.strip()
def fail(errType, reas, allow_use_curSplt=True):
"""
Throw an error of type `errType`, with `reas` as the reason.
Appends the line number and the erroring line.
`allow_use_curSplt` is a boolean, set to False if you want to ignore `curSplt` (no colored line)
"""
try:
nonlocal curSplt
if allow_use_curSplt and (curSplt >= 0):
# Get a colorized line
# (Blah blah [CSI-color change][Here is the error][CSI-color change] blah)
line = ""
alreadyChanged = 0
csp = 0
while csp < curSplt:
if spltln[csp] == '\x00':
alreadyChanged = 1
curSplt = curSplt + 1
else:
if alreadyChanged == 0:
line = line + spltln[csp] + " "
elif alreadyChanged == 1:
line = line + spltln[csp] + "<"
alreadyChanged = 2
else:
line = line + spltln[csp] + ">"
alreadyChanged = 1
csp = csp + 1
if spltln[csp] == '\x00':
alreadyChanged = 1
curSplt = curSplt + 1
csp = csp + 1
line = line + "\033[31m[\033[91m" + spltln[csp] + "\033[31m]\033[m"
if alreadyChanged == 0:
line = line + " "
elif alreadyChanged == 1:
line = line + "<"
alreadyChanged = 2
else:
line = line + ">"
alreadyChanged = 1
csp = csp + 1
while csp < len(spltln):
if spltln[csp] == '\x00':
alreadyChanged = 1
curSplt = curSplt + 1
else:
if alreadyChanged == 0:
line = line + spltln[csp] + " "
elif alreadyChanged == 1:
line = line + spltln[csp] + "<"
alreadyChanged = 2
else:
line = line + spltln[csp] + ">"
alreadyChanged = 1
csp = csp + 1
raise errType(string(str(reas) + " (" + str(lnno + 1) + ": " + line[:-1] + ")"))
else:
# No colored line
raise errType(str(reas) + " (" + str(lnno + 1) + ": " + ln + ")")
except errType as e:
# Raise a BaseException as an error wrapper
# (otherwise it will be caught and the line will be re-appended)
raise BaseException("[Error wrapper]") from e
spltln = ln.split(' ')
curSplt = -1
mask = [0] * 32
correctBits = [0] * 32
curBit = 32
ocurBit = -1
req = ''
imms = [] # Immediates
parm = [] # Custom variable length parameters
variables = {}
def generate_bin_test(positions=[], specifics=[]):
"""
Generates the if statement at the beginning.
You may use positions and specifics to implement a "multiple choice if":
- positions is an array that contains a bit position (MSB = 0)
- specifics is an array of arrays the same length as positions that contains a tuple (mask, correctBit)
that is at position positions[current_pos]
"""
if specifics == []:
append("if ((opcode & " + arr2hex(mask) + ") == " + arr2hex(correctBits) + ") {\n")
else:
l = len(positions)
if any(map(lambda v: (v < 0) or (v > 31), positions)):
fail(
AssertionError,
"generate_bin_tests requires a valid positions ({}) and specifics ({})!".format(
len(positions), len(specifics)
)
)
if any(map(lambda s: len(s) != l, specifics)):
fail(
AssertionError,
"generate_bin_tests requires the same length for positions ({}) and each element "
"of the specifics array ({})!".format(
len(positions), [len(s) for s in specifics]
)
)
inner = []
for specific in specifics:
for i, (m, c) in zip(positions, specific):
mask[i] = m
correctBits[i] = c
inner.append("((opcode & " + arr2hex(mask) + ") == " + arr2hex(correctBits) + ")")
append("if (" + " || ".join(inner) + ") {\n")
def parse_var_requirements():
"""
Parse the `/` restrictions on the bit fields
"""
nonlocal req
while req != '':
key = req[0]
req = req[1:]
if key == '=':
if len(req) < ocurBit - curBit:
fail(KeyError, "Not enough data in constraint value (type '=' for val " + val + ")")
for i in range(ocurBit - curBit):
if req[0] == 'x':
req = req[1:]
continue
elif (req[0] != '0') and (req[0] != '1'):
fail(KeyError, "Invalid constraint '" + key + "..." + req[0] + "'")
mask[32 - ocurBit + i] = 1
correctBits[32 - ocurBit + i] = ord(req[0]) - ord('0')
req = req[1:]
elif (ord(key) >= ord('0')) and (ord(key) <= ord('9')):
if req == '':
fail(KeyError, "Not enough data in constraint value (type '" + key + "' for val " + \
val + ")")
elif (req[0] != '0') and (req[0] != '1'):
fail(KeyError, "Invalid constraint '" + key + req[0] + "'")
mask[31 - curBit + ord('0') - ord(key)] = 1
correctBits[31 - curBit + ord('0') - ord(key)] = ord(req[0]) - ord('0')
req = req[1:]
elif (ord(key) >= ord('A')) and (ord(key) <= ord('F')):
if req == '':
fail(KeyError, "Not enough data in constraint value (type '" + key + "' for val " + \
val + ")")
elif (req[0] != '0') and (req[0] != '1'):
fail(KeyError, "Invalid constraint '" + key + req[0] + "'")
mask[31 - curBit + ord('A') + 10 - ord(key)] = 1
correctBits[31 - curBit + ord('A') + 10 - ord(key)] = ord(req[0]) - ord('0')
req = req[1:]
def add_custom_variables():
nonlocal curSplt
# Check for any custom variables
if len(parm) == 1:
# One parameter, name it param
append("int param = (opcode >> " + str(parm[0][0]) + ") & " + sz2str(parm[0][1]) + ";\n")
else:
# Multiple parameters, name them "param" paramNr "_" paramBitsSize
for i, p in enumerate(parm):
append(
"int param" + str(i + 1) + "_" + str(p[1]) + " = (opcode >> " + \
str(p[0]) + ") & " + sz2str(p[1]) + ";\n"
)
if spltln[curSplt] == '@':
# Additional custom modifications
append("\n")
curSplt = curSplt + 1
while spltln[curSplt] != '@':
if '=' in spltln[curSplt]:
# Add an `if` statement
eq = spltln[curSplt]
# Read initialization statement...
curSplt = curSplt + 1
oldSplt = curSplt
while spltln[curSplt][-1] != '@':
curSplt = curSplt + 1
if len(spltln) == curSplt:
fail(KeyError, "End of '=' switch not found!")
# Always initialize if necessary
if oldSplt != curSplt:
append(' '.join(spltln[oldSplt:curSplt]) + ";\n")
ifs = [""]
if spltln[curSplt] == '@@':
# Custom ifs
# Also, requires only a single '='
if eq != "=":
fail(ValueError, "Too many '=' switches (@@ modifier)")
# Extract the statements
statements = []
while spltln[curSplt] == '@@':
curSplt = curSplt + 1
statement = spltln[curSplt]
while spltln[curSplt][-1] != '@':
curSplt = curSplt + 1
statement = statement + " " + spltln[curSplt]
if len(spltln) == curSplt:
fail(KeyError, "End of '=' switch (@@ modifier) not found!")
statements.append(statement[:-1])
curSplt = curSplt + 1
# Unified eq length
eq = statements + [0]
# Make the ifs array
for stmt in statements:
ifs[-1] = ifs[-1] + "if (" + stmt + ") {\n"
ifs.append("} else ")
ifs[-1] = ifs[-1] + "{\n"
else:
# Standard if
eq = eq.split('=')
for e in eq[1:]:
ifs[-1] = ifs[-1] + "if (" + eq[0] + " == " + e + ") {\n"
ifs.append("} else ")
ifs[-1] = ifs[-1] + "{\n"
if spltln[curSplt] == '!@':
# Custom statements
curSplt = curSplt + 1
# Extract the common value
commonPart = spltln[curSplt]
while spltln[curSplt][-1] != '@':
curSplt = curSplt + 1
commonPart = commonPart + " " + spltln[curSplt]
if len(spltln) == curSplt:
fail(KeyError, "End of '=' switch (!@ modifier) not found!")
commonPart = commonPart[:-1].replace("\\n", "\n")
curSplt = curSplt + 1
# Extract the parts
commonPart = commonPart.split('%')
if len(commonPart) < 2:
fail(ValueError, "No replacement place!")
for if_ in ifs[:-1]:
append(if_)
# For each '%', append the preceding part and the variable part
for common in commonPart[:-1]:
insert = spltln[curSplt]
while spltln[curSplt][-1] != '@':
curSplt = curSplt + 1
insert = insert + " " + spltln[curSplt]
if len(spltln) == curSplt:
fail(
KeyError,
"End of '=' switch (!@ modifier, repl1 i=" + str(i) + \
" part) not found!"
)
insert = insert[:-1].replace("\\n", "\n")
append(common + insert)
curSplt = curSplt + 1
append(commonPart[-1] + ";\n")
# Finish with the `else` part
append(ifs[-1])
# For each '%', append the preceding part and the variable part
for common in commonPart[:-1]:
insert = spltln[curSplt]
while spltln[curSplt][-1] != '@':
curSplt = curSplt + 1
insert = insert + " " + spltln[curSplt]
if len(spltln) == curSplt:
fail(KeyError, "End of '=' switch (!@ modifier, repl2 part) not found!")
insert = insert[:-1].replace("\\n", "\n")
append(common + insert)
curSplt = curSplt + 1
append(commonPart[-1] + ";\n}\n")
else:
# Standard statements
dynvars = [dynvar.split(',') for dynvar in spltln[curSplt].split(';')]
dynvars[-1][-1] = dynvars[-1][-1][:-1]
curSplt = curSplt + 1
if any(len(dynvar) != len(eq) + 1 for dynvar in dynvars):
fail(ValueError, "Not enough/too many possibilities in switch")
# Reorganize dynvars so it matches [[variables], [set-if-A-true], ..., [set-else]]
dynvars = list(map(lambda i: [dv[i] for dv in dynvars], range(len(dynvars[0]))))
for if_, vals in zip(ifs[:-1], dynvars[1:]):
append(if_)
for var, val in zip(dynvars[0], vals):
append(var + " = " + val + ";\n")
# Else
append(ifs[-1])
for var, val in zip(dynvars[0], dynvars[-1]):
append(var + " = " + val + ";\n")
append("}\n")
elif spltln[curSplt] == "set":
# Set a (new?) variable
curSplt = curSplt + 1
oldSplt = curSplt
while spltln[curSplt][-1] != '@':
curSplt = curSplt + 1
if len(spltln) == curSplt:
fail(KeyError, "End of '=' switch not found!")
# If the statement is empty just add a blank line
if (oldSplt == curSplt) and (spltln[curSplt] == "@"):
append("\n")
else:
append(' '.join(spltln[oldSplt:curSplt + 1])[:-1] + ";\n")
curSplt = curSplt + 1
else:
fail(KeyError, "Unknown custom statement type '" + spltln[curSplt] + "'")
curSplt = curSplt + 1
try:
if spltln[0] == "ARM_":
# ARM instruction
variables["cond"] = -1
variables["registers"] = {
"d": [-1, -1, -1, -1, -1], # Register: #, ##, Rd, RdLo, RdHi
"t": [-1, -1, -1], # Register: #, ##, Rt
"n": [-1, -1, -1], # Register: #, ##, Rn
"m": [-1, -1, -1], # Register: #, ##, Rm
"a": [-1, -1, -1] # Register: #, ##, Ra
}
variables["reglist16"] = -1 # Register list (16-bits)
variables["s"] = -1 # Set flags
variables["u"] = -1 # Unsigned
variables["r"] = -1 # Rotate
variables["sb"] = [-1, -1, -1] # lsb, msb
variables["w"] = -1 # wback
for i, val in enumerate(spltln[1:]):
if '/' in val:
ocurBit = curBit
req = val.split('/')
val, req = req[0], '/'.join(req[1:])
if val == '0':
curBit = curBit - 1
mask[31 - curBit] = 1
elif val == '1':
curBit = curBit - 1
mask[31 - curBit] = 1
correctBits[31 - curBit] = 1
elif (val == '(0)') or (val == '(1)'):
# Ignore, even though the result should be undefined...
curBit = curBit - 1
elif val == 'cond':
curBit = curBit - 4
variables["cond"] = curBit
elif val == 'Rd':
curBit = curBit - 4
variables["registers"]["d"][2] = curBit
elif val == 'RdLo':
curBit = curBit - 4
variables["registers"]["d"][3] = curBit
elif val == 'RdHi':
curBit = curBit - 4
variables["registers"]["d"][4] = curBit
elif val == 'Rt':
curBit = curBit - 4
variables["registers"]["t"][2] = curBit
elif val == 'Rn':
curBit = curBit - 4
variables["registers"]["n"][2] = curBit
elif val == 'Rm':
curBit = curBit - 4
variables["registers"]["m"][2] = curBit
elif val == 'Ra':
curBit = curBit - 4
variables["registers"]["a"][2] = curBit
elif val == 'S':
curBit = curBit - 1
variables["s"] = curBit
elif val == 'U':
curBit = curBit - 1
variables["u"] = curBit
elif val == 'W':
curBit = curBit - 1
variables["w"] = curBit
elif val == 'rotate':
curBit = curBit - 2
variables["r"] = curBit
elif val == 'lsb':
curBit = curBit - 5
variables["sb"][0] = curBit
elif val == 'msb':
curBit = curBit - 5
variables["sb"][1] = curBit
elif val == 'widthm1':
curBit = curBit - 5
variables["sb"][2] = curBit
elif val == 'register_list':
curBit = curBit - 16
variables["reglist16"] = curBit
elif val == 'sat_imm':
curBit = curBit - 4
imms.append([curBit, 4])
elif val.startswith('@<') and val.endswith('>'):
parmlen = int(val[2:-1])
curBit = curBit - parmlen
parm.append([curBit, parmlen])
elif val.startswith('imm'):
if val.endswith('H') or val.endswith('L'): val = val[:-1]
immsz = int(val[3:])
curBit = curBit - immsz
imms.append([curBit, immsz])
else:
fail(KeyError, "Unknown value '" + val + "'")
curSplt = i + 2
parse_var_requirements()
if curBit == 0:
break
elif curBit < 0:
fail(KeyError, "Current bit too low (" + str(curBit) + ")")
if curSplt == -1:
fail(KeyError, "Not enough arguments")
generate_bin_test()
# Add C variables
if variables["s"] != -1:
append("int s = (opcode >> " + str(variables["s"]) + ") & 1;\n")
if variables["cond"] != -1:
if mask[31 - variables["cond"]] == 0:
append("const char* cond = conds[(opcode >> " + str(variables["cond"]) + ") & 0xF];\n")
else:
variables["cond"] = -2
if variables["registers"]["d"][2] != -1:
append("int rd = (opcode >> " + str(variables["registers"]["d"][2]) + ") & 0xF;\n")
if variables["registers"]["d"][3] != -1:
assert(variables["registers"]["d"][4] != -1)
append("int rdlo = (opcode >> " + str(variables["registers"]["d"][3]) + ") & 0xF;\n")
append("int rdhi = (opcode >> " + str(variables["registers"]["d"][4]) + ") & 0xF;\n")
if variables["registers"]["t"][2] != -1:
append("int rt = (opcode >> " + str(variables["registers"]["t"][2]) + ") & 0xF;\n")
if variables["registers"]["n"][2] != -1:
append("int rn = (opcode >> " + str(variables["registers"]["n"][2]) + ") & 0xF;\n")
if variables["registers"]["m"][2] != -1:
append("int rm = (opcode >> " + str(variables["registers"]["m"][2]) + ") & 0xF;\n")
if variables["registers"]["a"][2] != -1:
append("int ra = (opcode >> " + str(variables["registers"]["a"][2]) + ") & 0xF;\n")
if variables["u"] != -1:
append("int u = (opcode >> " + str(variables["u"]) + ") & 1;\n")
if variables["w"] != -1:
append("int w = (opcode >> " + str(variables["w"]) + ") & 1;\n")
if variables["r"] != -1:
append("int rot = (opcode >> " + str(variables["r"]) + ") & 3;\nchar tmprot[8] = {0};\n")
append("if (rot) {\nsprintf(tmprot, \" ror %d\", rot * 8);\n}\n")
if variables["sb"][0] != -1:
append("int lsb = (opcode >> " + str(variables["sb"][0]) + ") & 0x1F;\n")
if variables["sb"][1] != -1:
append("int msb = (opcode >> " + str(variables["sb"][1]) + ") & 0x1F;\n")
if variables["sb"][2] != -1:
append("int widthm1 = (opcode >> " + str(variables["sb"][2]) + ") & 0x1F;\n")
if variables["reglist16"] != -1:
append("int reglist = (opcode >> " + str(variables["reglist16"]) + ") & 0xFFFF;\n")
if imms != []:
immssz = sum(map(lambda v: v[1], imms))
tmp = "(" * len(imms)
for immpos, immsz in imms:
if tmp[-1] != '(': tmp = tmp + " << " + str(immsz) + ") | "
tmp = tmp + "((opcode >> " + str(immpos) + ") & " + sz2str(immsz) + ")"
tmp = tmp[1:]
append("uint" + str(nextAvailable(immssz)) + "_t imm" + str(immssz) + " = " + tmp + ";\n")
# Destroy imms since we don't need it anymore, but we do need immssz
imms = immssz
add_custom_variables()
append("\nsprintf(ret, \"")
# Assemble the variables into the printf
instText = ' '.join(spltln[curSplt:]).split('<')
instText = [instTextPart.split('>') for instTextPart in instText]
instText = [itp for itp2 in instText for itp in itp2]
# Make the failures nicer
spltln = spltln[:curSplt] + ["\x00"] + instText
printf_args = ""
for idx, text in enumerate(instText):
text = text.replace("&l", "<").replace("&g", ">")
if "+/-" in text:
text = text.replace("+/-", "%s")
printf_args = printf_args + ", (u ? \"\" : \"-\")"
if "{!}" in text:
text = text.replace("{!}", "%s")
printf_args = printf_args + ", (w ? \"!\" : \"\")"
skiprbr = False
if idx % 2:
if text == "c":
2021-11-04 10:31:51 +00:00
if variables["cond"] >= 0:
append("%s")
printf_args = printf_args + ", cond"
elif text == "Rd":
append("%s")
printf_args = printf_args + ", regname[rd]"
elif text == "RdLo":
append("%s")
printf_args = printf_args + ", regname[rdlo]"
elif text == "RdHi":
append("%s")
printf_args = printf_args + ", regname[rdhi]"
elif text == "Rt":
append("%s")
printf_args = printf_args + ", regname[rt]"
elif text == "Rt2":
append("%s")
printf_args = printf_args + ", regname[rt + 1]"
elif text == "Rn":
append("%s")
printf_args = printf_args + ", regname[rn]"
elif text == "Rm":
append("%s")
printf_args = printf_args + ", regname[rm]"
elif text == "Ra":
append("%s")
printf_args = printf_args + ", regname[ra]"
elif text == "const":
assert(imms == 12)
append("0x%x")
printf_args = printf_args + ", print_modified_imm_ARM(imm12)"
elif text == "rotation":
if output.endswith("{, "):
output = output[:-3]
skiprbr = True
append("%s")
printf_args = printf_args + ", tmprot"
elif text == "label":
append("%+d")
printf_args = printf_args + ", (imm24 & 0x800000 ? imm24 | 0xFF000000 : imm24) + 2"
elif text == "lsb":
append("%d")
printf_args = printf_args + ", lsb"
elif text == "width":
append("%d")
printf_args = printf_args + (", msb - lsb + 1" if variables["sb"][1] != -1 else ", widthm1 + 1")
elif text == "registers":
append("%s")
printf_args = printf_args + ", print_register_list(reglist, 16)"
elif text == "imm":
#append("0x%0" + str(ceil(imms / 4)) + "x")
append("0x%x")
printf_args = printf_args + ", imm" + str(imms)
elif text.startswith("imm"):
if variables["u"] != -1:
append("%d")
else:
#append("0x%0" + str(ceil(int(text[3:]) / 4)) + "x")
append("0x%x")
printf_args = printf_args + ", " + text
else:
fail(KeyError, "Unknown variable " + text)
elif text.endswith("{S}"):
append(text[:-3] + "%s")
printf_args = printf_args + ", (s ? \"S\" : \"\")"
else:
if text == "":
continue
text = text.split('\\')
while len(text) > 1:
if text[1][0] == '%':
modifier, text[1] = '%', text[1][1:]
while (len(text[1]) > 1) \
and (text[1][0] in ['0', '1', '2', '3', '4', '5', '6', '7', '8', '9', '+']):
modifier, text[1] = modifier + text[1][0], text[1][1:]
modifier, text[1] = modifier + text[1][0], text[1][1:]
append(text[0] + modifier)
else:
append(text[0] + "%s")
printf_args = printf_args + ", " + text[1]
text = text[2:]
if len(text) == 0:
fail(AssertionError, "Substitution not finished")
if skiprbr and (text[0][0] == "}"):
text[0] = text[0][1:]
append(text[0])
curSplt = curSplt + 1
append("\"" + printf_args + ");\n} else ")
elif (spltln[0].upper() == "ARMS") or (spltln[0] == "ARM$"):
# ARM Shift instruction
# ARMs is only with immediate, ARM$ is only with register
variables["cond"] = -1
variables["registers"] = {
"d": [-1, -1, -1], # Register: #, ##, Rd
"t": [-1, -1, -1], # Register: #, ##, Rt
"n": [-1, -1, -1], # Register: #, ##, Rn
"m": [-1, -1, -1] # Register: #, ##, Rm
}
variables["s"] = -1 # Set flags
variables["u"] = -1 # Unsigned shift
variables["w"] = -1 # wback (used in adressing)
variables["shift"] = False # Shift already detected?
for i, val in enumerate(spltln[1:]):
if '/' in val:
ocurBit = curBit
req = val.split('/')
val, req = req[0], '/'.join(req[1:])
if val == '0':
curBit = curBit - 1
mask[31 - curBit] = 1
elif val == '1':
curBit = curBit - 1
mask[31 - curBit] = 1
correctBits[31 - curBit] = 1
elif (val == '(0)') or (val == '(1)'):
# Ignore, even though the result should be undefined...
curBit = curBit - 1
elif val == 'cond':
curBit = curBit - 4
variables["cond"] = curBit
elif val == 'Rd':
curBit = curBit - 4
variables["registers"]["d"][2] = curBit
elif val == 'Rt':
curBit = curBit - 4
variables["registers"]["t"][2] = curBit
elif val == 'Rn':
curBit = curBit - 4
variables["registers"]["n"][2] = curBit
elif val == 'Rm':
curBit = curBit - 4
variables["registers"]["m"][2] = curBit
elif val == 'S':
curBit = curBit - 1
variables["s"] = curBit
elif val == 'U':
curBit = curBit - 1
variables["u"] = curBit
elif val == 'W':
curBit = curBit - 1
variables["w"] = curBit
elif val == 'type':
curBit = curBit - 2
imms.append([curBit, 2])
variables["shift"] = True
elif val == 'sat_imm':
curBit = curBit - 5
imms.append([curBit, 5, True])
elif val == 'sh':
curBit = curBit - 1
elif val.startswith('@<') and val.endswith('>'):
parmlen = int(val[2:-1])
curBit = curBit - parmlen
parm.append([curBit, parmlen])
elif val.startswith('imm'):
if variables["shift"]: fail(ValueError, "Immediate after a 'type' value detected")
immsz = int(val[3:])
curBit = curBit - immsz
imms.append([curBit, immsz])
variables["shift"] = immsz == 12
else:
fail(KeyError, "Unknown value '" + val + "'")
curSplt = i + 2
parse_var_requirements()
if curBit == 0:
break
elif curBit < 0:
fail(KeyError, "Current bit too low (" + str(curBit) + ")")
if curSplt == -1:
fail(KeyError, "Not enough arguments")
# There is not necessarily an explicit shift when using an ARMs
elif not variables["shift"] and (spltln[0] == "ARMS"):
fail(KeyError, "No shift detected")
# Assumption:
# bytes 11-4 are [imm5 type 0] (then auto-complete for [Rs 0 type 1 Rm])
if (spltln[0] == "ARMS") and ((mask[20:24] + mask[25:28] != [0, 0, 0, 0, 0, 0, 1]) \
or ((correctBits[27] != 0) or (len(imms) != 2) or (imms[0] != [7, 5]))):
fail(NotImplementedError, "Unknown case with shift")
if (spltln[0] == "ARMS"):
generate_bin_test([24, 27], [[(0, 0), (1, 0)], [(1, 0), (1, 1)]])
else:
generate_bin_test()
# Add C variables
if variables["s"] != -1:
append("int s = (opcode >> " + str(variables["s"]) + ") & 1;\n")
if variables["cond"] != -1:
if mask[31 - variables["cond"]] == 0:
append("const char* cond = conds[(opcode >> " + str(variables["cond"]) + ") & 0xF];\n")
else:
variables["cond"] = -2
if variables["registers"]["d"][2] != -1:
append("int rd = (opcode >> " + str(variables["registers"]["d"][2]) + ") & 0xF;\n")
if variables["registers"]["t"][2] != -1:
append("int rt = (opcode >> " + str(variables["registers"]["t"][2]) + ") & 0xF;\n")
if variables["registers"]["n"][2] != -1:
append("int rn = (opcode >> " + str(variables["registers"]["n"][2]) + ") & 0xF;\n")
if variables["registers"]["m"][2] != -1:
append("int rm = (opcode >> " + str(variables["registers"]["m"][2]) + ") & 0xF;\n")
if variables["u"] != -1:
append("int u = (opcode >> " + str(variables["u"]) + ") & 1;\n")
if variables["w"] != -1:
append("int w = (opcode >> " + str(variables["w"]) + ") & 1;\n")
if (len(imms) == 2) and (len(imms[0]) == 3):
append("uint8_t imm = ((opcode >> " + str(imms[0][0]) + ") & 0x1F) + 1;\n")
append("uint8_t shift = ((opcode >> 4) & " + \
("0xFF)" if (spltln[0] == "ARMS") else ("0xFE)" if (spltln[0] == "ARMs") else "0xFF) | 0x01")) + \
";\n")
add_custom_variables()
append("\nsprintf(ret, \"")
# Assemble the variables into the printf
instText = ' '.join(spltln[curSplt:]).split('<')
instText = [instTextPart.split('>') for instTextPart in instText]
instText = [itp for itp2 in instText for itp in itp2]
# Make the failures nicer
spltln = spltln[:curSplt] + ["\x00"] + instText
printf_args = ""
for idx, text in enumerate(instText):
text = text.replace("&l", "<").replace("&g", ">")
if "+/-" in text:
text = text.replace("+/-", "%s")
printf_args = printf_args + ", (u ? \"\" : \"-\")"
if "{!}" in text:
text = text.replace("{!}", "%s")
printf_args = printf_args + ", (w ? \"!\" : \"\")"
if idx % 2:
if text == "c":
if variables["cond"] != -1:
append("%s")
printf_args = printf_args + ", cond"
elif text == "Rd":
append("%s")
printf_args = printf_args + ", regname[rd]"
elif text == "Rt":
append("%s")
printf_args = printf_args + ", regname[rt]"
elif text == "Rt2":
append("%s")
printf_args = printf_args + ", regname[rt + 1]"
elif text == "Rn":
append("%s")
printf_args = printf_args + ", regname[rn]"
elif text == "Rm":
append("%s")
printf_args = printf_args + ", regname[rm]"
elif text == "imm":
append("%d")
printf_args = printf_args + ", imm"
elif text == "shift":
# Tricky, since it is optional, but I want to have a pure copy-paste of the official doc
# so we remove the end of the output if necessary
# Otherwise it is as usual, however we use the print_shift method
comma = "0"
if output.endswith("{, "):
output = output[:-3]
comma = "1"
append("%s")
printf_args = printf_args + ", print_shift(shift, " + comma + ")"
else:
fail(KeyError, "Unknown variable " + text)
elif text.endswith("{S}"):
append(text[:-3] + "%s")
printf_args = printf_args + ", (s ? \"S\" : \"\")"
elif text.startswith("}"):
append(text[1:])
else:
if text == "":
continue
text = text.split('\\')
while len(text) > 1:
if text[1][0] == '%':
modifier, text[1] = '%', text[1][1:]
while (len(text[1]) > 1) \
and (text[1][0] in ['0', '1', '2', '3', '4', '5', '6', '7', '8', '9', '+']):
modifier, text[1] = modifier + text[1][0], text[1][1:]
modifier, text[1] = modifier + text[1][0], text[1][1:]
append(text[0] + modifier)
else:
append(text[0] + "%s")
printf_args = printf_args + ", " + text[1]
text = text[2:]
if len(text) == 0:
fail(AssertionError, "Substitution not finished")
append(text[0])
curSplt = curSplt + 1
append("\"" + printf_args + ");\n} else ")
elif spltln[0] == "NEON":
# NEON (advanced SIMD) instruction
variables["registers"] = {
"d": [-1, -1, -1], # Register: D, Vd, ##
"t": [-1, -1, -1], # Register: #, ##, Rt
"n": [-1, -1, -1], # Register: N, Vn, ##
"m": [-1, -1, -1] # Register: M, Vm, ##
}
variables["op"] = -1 # Operation
variables["u"] = -1 # Unsigned
variables["f"] = -1 # Floating-point
variables["sz"] = [-1, 0] # Operation size: pos, len
variables["cmode"] = -1 # Used with immediates
variables["q"] = -1 # Quadword
for i, val in enumerate(spltln[1:]):
if '/' in val:
ocurBit = curBit
req = val.split('/')
val, req = req[0], '/'.join(req[1:])
if val == '0':
curBit = curBit - 1
mask[31 - curBit] = 1
elif val == '1':
curBit = curBit - 1
mask[31 - curBit] = 1
correctBits[31 - curBit] = 1
elif (val == '(0)') or (val == '(1)'):
# Ignore, even though the result should be undefined...
curBit = curBit - 1
elif val == 'D':
curBit = curBit - 1
variables["registers"]["d"][0] = curBit
elif val == 'Vd':
curBit = curBit - 4
variables["registers"]["d"][1] = curBit
elif val == 'Rt':
curBit = curBit - 4
variables["registers"]["t"][2] = curBit
elif val == 'N':
curBit = curBit - 1
variables["registers"]["n"][0] = curBit
elif val == 'Vn':
curBit = curBit - 4
variables["registers"]["n"][1] = curBit
elif val == 'Rn':
curBit = curBit - 4
variables["registers"]["n"][2] = curBit
elif val == 'M':
curBit = curBit - 1
variables["registers"]["m"][0] = curBit
elif val == 'Vm':
curBit = curBit - 4
variables["registers"]["m"][1] = curBit
elif val == 'Rm':
curBit = curBit - 4
variables["registers"]["m"][2] = curBit
elif val == 'U':
curBit = curBit - 1
variables["u"] = curBit
elif val == 'F':
curBit = curBit - 1
variables["f"] = curBit
elif val == 'op':
curBit = curBit - 1
variables["op"] = curBit
elif val == 'cmode':
curBit = curBit - 4
variables["cmode"] = curBit
elif val == 'size':
curBit = curBit - 2
variables["sz"] = [curBit, 2]
elif val == 'sz':
curBit = curBit - 1
variables["sz"] = [curBit, 1]
elif val == 'Q':
curBit = curBit - 1
variables["q"] = curBit
elif val == 'i':
curBit = curBit - 1
imms.append([curBit, 1, val])
elif val == 'L':
curBit = curBit - 1
imms.append([curBit, 1, val])
elif val.startswith('@<') and val.endswith('>'):
parmlen = int(val[2:-1])
curBit = curBit - parmlen
parm.append([curBit, parmlen])
elif val.lower().startswith('imm'):
immsz = int(val[3:])
curBit = curBit - immsz
imms.append([curBit, immsz, val])
else:
fail(KeyError, "Unknown value '" + val + "'")
curSplt = i + 2
parse_var_requirements()
if curBit == 0:
break
elif curBit < 0:
fail(KeyError, "Current bit too low (" + str(curBit) + ")")
if curSplt == -1:
fail(KeyError, "Not enough arguments")
generate_bin_test()
# Add C variables
if variables["op"] != -1:
append("int op = (opcode >> " + str(variables["op"]) + ") & 1;\n")
if (variables["u"] != -1) and (variables["f"] != -1):
fail(ValueError, "Cannot have both unsigned and floating-point", False)
if variables["u"] != -1:
append("int u = (opcode >> " + str(variables["u"]) + ") & 1;\n")
if variables["f"] != -1:
append("int f = (opcode >> " + str(variables["f"]) + ") & 1;\n")
if variables["sz"][0] != -1:
if (mask[31 - variables["sz"][0]] == 0) and ((variables["sz"][1] == 1) or (mask[30 - variables["sz"][0]] == 0)):
append("int size = (opcode >> " + str(variables["sz"][0]) + ") & " + sz2str(variables["sz"][1]) + ";\n")
else:
variables["sz"][0] = -2
if variables["q"] != -1:
if mask[31 - variables["q"]] == 0:
append("int q = (opcode >> " + str(variables["q"]) + ") & 1;\n")
else:
variables["q"] = -2
if variables["registers"]["d"][0] != -1:
assert(variables["registers"]["d"][1] != -1)
append("int d = ((opcode >> " + str(variables["registers"]["d"][0]) + ") & 1) << 4 " + \
"| ((opcode >> " + str(variables["registers"]["d"][1]) + ") & 0xF);\n")
if variables["registers"]["t"][2] != -1:
append("int rt = (opcode >> " + str(variables["registers"]["t"][2]) + ") & 0xF;\n")
if variables["registers"]["n"][0] != -1:
assert(variables["registers"]["n"][1] != -1)
append("int n = ((opcode >> " + str(variables["registers"]["n"][0]) + ") & 1) << 4 " + \
"| ((opcode >> " + str(variables["registers"]["n"][1]) + ") & 0xF);\n")
if variables["registers"]["n"][2] != -1:
append("int rn = (opcode >> " + str(variables["registers"]["n"][2]) + ") & 0xF;\n")
if variables["registers"]["m"][0] != -1:
assert(variables["registers"]["m"][1] != -1)
append("int m = ((opcode >> " + str(variables["registers"]["m"][0]) + ") & 1) << 4 " + \
"| ((opcode >> " + str(variables["registers"]["m"][1]) + ") & 0xF);\n")
if variables["registers"]["m"][2] != -1:
append("int rm = (opcode >> " + str(variables["registers"]["m"][2]) + ") & 0xF;\n")
if variables["cmode"] != -1:
append("int cmode = (opcode >> " + str(variables["cmode"]) + ") & 0xF;\n")
if imms != []:
if (variables["cmode"] != -1) and (variables["op"] != -1):
# Modified immediate value
# Assert unique imms pattern since this is a NEON instruction, reduce the result's length
assert(imms == [[24, 1, 'i'], [16, 3, 'imm3'], [0, 4, 'imm4']])
append("uint8_t imm8 = " + \
"(((opcode >> 24) & 0x1) << 7) | (((opcode >> 16) & 0x7) << 4) | (opcode & 0xF);" + \
"\nconst char* decodedImm = print_modified_imm_NEON((op << 4) + cmode, imm8);\n")
imms = [0, 8] + imms
elif imms[0][1] in [6]:
# Shift value
assert((len(imms) == 1) or ((len(imms) == 2) and (imms[1][2] == 'L')))
l = len(imms) == 2
if l:
append("uint8_t l = (opcode >> " + str(imms[1][0]) + ") & 0x1;\n")
immssz = 6
append("uint8_t imm6 = (opcode >> " + str(imms[0][0]) + ") & 0x3F;\n")
if imms[0][2][:3] != "IMM":
append("uint8_t decodedImm = 0;\nuint8_t size = 0;\n")
if imms[0][2][:3] == "imm":
append(
("if (l == 1) {\ndecodedImm = 64 - imm6;\nsize = 3;\n} else " if l else "") + \
"if (imm6 & 0x20) {\ndecodedImm = 64 - imm6;\nsize = 2;\n} else " + \
"if (imm6 & 0x10) {\ndecodedImm = 32 - imm6;\nsize = 1;\n} else " + \
"{\ndecodedImm = 16 - imm6;\n}\n"
)
else:
append(
("if (l == 1) {\ndecodedImm = imm6;\nsize = 3;\n} else " if l else "") + \
"if (imm6 & 0x20) {\ndecodedImm = imm6 - 32;\nsize = 2;\n" + \
"} else if (imm6 & 0x10) {\ndecodedImm = imm6 - 16;\nsize = 1;\n" + \
"} else {\ndecodedImm = imm6 - 8;\n}\n"
)
imms = [1, immssz] + imms
elif imms[0][1] == 4:
# Weird shift...
assert(len(imms) == 1)
immssz = 4
append("uint8_t imm4 = (opcode >> " + str(imms[0][0]) + ") & 0xF;\n")
if imms[0][2][:3] != "IMM":
append(
"uint8_t decodedImm = 0;\nuint8_t size = 0;\n" + \
"if (imm4 & 0b0001) {\ndecodedImm = imm4 >> 1;\nsize = 8;\n} else " + \
"if (imm4 & 0b0011) {\ndecodedImm = imm4 >> 2;\nsize = 16;\n} else " + \
"if (imm4 & 0b0111) {\ndecodedImm = imm4 >> 3;\nsize = 32;\n}\n"
)
imms = [2, immssz] + imms
add_custom_variables()
append("\nsprintf(ret, \"")
# Assemble the variables into the printf
instText = ' '.join(spltln[curSplt:]).split('<')
instText = [instTextPart.split('>') for instTextPart in instText]
instText = [itp for itp2 in instText for itp in itp2]
# Make the failures nicer
spltln = spltln[:curSplt] + ["\x00"] + instText
printf_args = ""
for idx, text in enumerate(instText):
text = text.replace("&l", "<").replace("&g", ">")
if "+/-" in text:
text = text.replace("+/-", "%s")
printf_args = printf_args + ", (u ? \"\" : \"-\")"
if idx % 2:
if text == "c":
pass
elif text == "dt":
append("%s")
if imms == []:
printf_args = printf_args + ", dts[" + ("0x8 + (f << 2) + " if variables["f"] != -1 else "")
printf_args = printf_args + ("(u << 2) + " if variables["u"] != -1 else "") + "size]"
elif imms[0] == 0:
printf_args = printf_args + ", decodedImm"
elif imms[0] == 1:
printf_args = printf_args + ", dts[" + ("0x8 + (f << 2) + " if variables["f"] != -1 else "")
printf_args = printf_args + ("(u << 2) + " if variables["u"] != -1 else "") + "size]"
else:
fail(ValueError, "Unsupported immediate type {} (on dt)".format(imms[0]))
elif text == "size":
if imms[0] == 1:
append("%d")
printf_args = printf_args + ", 8 << size"
else:
fail(ValueError, "Unsupported immediate type {} (on size)".format(imms[0]))
elif text == "Qd":
append("%s")
printf_args = printf_args + ", vecname[" + ("(q << 5) + 0x20" if variables["q"] != -1 else "0x40") + " + d]"
elif text == "qd":
append("%s")
printf_args = printf_args + ", vecname[0x40 + d]"
elif text == "Dd":
append("%s")
printf_args = printf_args + ", vecname[0x20 + d]"
elif text == "Rt":
append("%s")
printf_args = printf_args + ", regname[rt]"
elif text == "Qn":
append("%s")
printf_args = printf_args + ", vecname[" + ("(q << 5) + 0x20" if variables["q"] != -1 else "0x40") + " + n]"
elif text == "Dn":
append("%s")
printf_args = printf_args + ", vecname[0x20 + n]"
elif text == "Rn":
append("%s")
printf_args = printf_args + ", regname[rn]"
elif text == "Qm":
append("%s")
printf_args = printf_args + ", vecname[" + ("(q << 5) + 0x20" if variables["q"] != -1 else "0x40") + " + m]"
elif text == "Dm":
append("%s")
printf_args = printf_args + ", vecname[0x20 + m]"
elif text == "Dm[x]":
append("%s[%d]")
printf_args = printf_args + ", vecname[0x20 + m], decodedImm"
elif text == "Rm":
append("%s")
printf_args = printf_args + ", regname[rm]"
elif text == "imm":
if imms[0] == 0:
append("%s")
printf_args = printf_args + ", decodedImm + 4"
elif imms[0] == 1:
append("%d")
printf_args = printf_args + ", decodedImm"
elif imms[0] == 2:
append("%d")
printf_args = printf_args + ", imm4"
else:
fail(ValueError, "Unsupported immediate type {}".format(imms[0]))
elif text == "fbits":
if (imms[0] == 1) and (imms[2][2][:3] == "IMM"):
append("%d")
printf_args = printf_args + ", 64 - imm6"
else:
fail(ValueError, "Unsupported immediate type {} (first imm = {})".format(imms[0], imms[2][2]))
else:
fail(KeyError, "Unknown variable " + text)
else:
if text == "":
continue
text = text.split('\\')
while len(text) > 1:
if text[1][0] == '%':
modifier, text[1] = '%', text[1][1:]
while (len(text[1]) > 1) \
and (text[1][0] in ['0', '1', '2', '3', '4', '5', '6', '7', '8', '9', '+']):
modifier, text[1] = modifier + text[1][0], text[1][1:]
modifier, text[1] = modifier + text[1][0], text[1][1:]
append(text[0] + modifier)
else:
append(text[0] + "%s")
printf_args = printf_args + ", " + text[1]
text = text[2:]
if len(text) == 0:
fail(AssertionError, "Substitution not finished")
append(text[0])
curSplt = curSplt + 1
append("\"" + printf_args + ");\n} else ")
elif spltln[0] == "VFPU":
# vFPU instruction
variables["cond"] = -1
variables["registers"] = {
"d": [-1, -1, -1], # Register: D, Vd, ##
"t": [-1, -1, -1], # Register: #, ##, Rt
"n": [-1, -1, -1], # Register: N, Vn, Rn
"m": [-1, -1, -1] # Register: M, Vm, ##
}
variables["op"] = -1 # Operation
variables["u"] = -1 # Unsigned
variables["f"] = -1 # Floating-point
variables["sz"] = [-1, 0] # Operation size: pos, len
variables["cmode"] = -1 # Used with immediates
variables["q"] = -1 # Quadword
variables["w"] = -1 # wback
for i, val in enumerate(spltln[1:]):
if '/' in val:
ocurBit = curBit
req = val.split('/')
val, req = req[0], '/'.join(req[1:])
if val == '0':
curBit = curBit - 1
mask[31 - curBit] = 1
elif val == '1':
curBit = curBit - 1
mask[31 - curBit] = 1
correctBits[31 - curBit] = 1
elif (val == '(0)') or (val == '(1)'):
# Ignore, even though the result should be undefined...
curBit = curBit - 1
elif val == 'cond':
curBit = curBit - 4
variables["cond"] = curBit
elif val == 'D':
curBit = curBit - 1
variables["registers"]["d"][0] = curBit
elif val == 'Vd':
curBit = curBit - 4
variables["registers"]["d"][1] = curBit
elif val == 'Rt':
curBit = curBit - 4
variables["registers"]["t"][2] = curBit
elif val == 'N':
curBit = curBit - 1
variables["registers"]["n"][0] = curBit
elif val == 'Vn':
curBit = curBit - 4
variables["registers"]["n"][1] = curBit
elif val == 'Rn':
curBit = curBit - 4
variables["registers"]["n"][2] = curBit
elif val == 'M':
curBit = curBit - 1
variables["registers"]["m"][0] = curBit
elif val == 'Vm':
curBit = curBit - 4
variables["registers"]["m"][1] = curBit
elif val == 'Rm':
curBit = curBit - 4
variables["registers"]["m"][2] = curBit
elif val == 'U':
curBit = curBit - 1
variables["u"] = curBit
elif val == 'F':
curBit = curBit - 1
variables["f"] = curBit
elif val == 'W':
curBit = curBit - 1
variables["w"] = curBit
elif val == 'op':
curBit = curBit - 1
variables["op"] = curBit
elif val == 'cmode':
curBit = curBit - 4
variables["cmode"] = curBit
elif val == 'size':
curBit = curBit - 2
variables["sz"] = [curBit, 2]
elif val == 'sz':
curBit = curBit - 1
variables["sz"] = [curBit, 1]
elif val == 'Q':
curBit = curBit - 1
variables["q"] = curBit
elif val == 'i':
curBit = curBit - 1
imms.append([curBit, 1, val])
elif val == 'L':
curBit = curBit - 1
imms.append([curBit, 1, val])
elif val.startswith('@<') and val.endswith('>'):
parmlen = int(val[2:-1])
curBit = curBit - parmlen
parm.append([curBit, parmlen])
elif val.lower().startswith('imm'):
immsz = int(val[3:])
curBit = curBit - immsz
imms.append([curBit, immsz, val])
else:
fail(KeyError, "Unknown value '" + val + "'")
curSplt = i + 2
parse_var_requirements()
if curBit == 0:
break
elif curBit < 0:
fail(KeyError, "Current bit too low (" + str(curBit) + ")")
if curSplt == -1:
fail(KeyError, "Not enough arguments")
generate_bin_test()
# Add C variables
if variables["cond"] != -1:
if mask[31 - variables["cond"]] == 0:
append("const char* cond = conds[(opcode >> " + str(variables["cond"]) + ") & 0xF];\n")
else:
variables["cond"] = -2
if variables["op"] != -1:
append("int op = (opcode >> " + str(variables["op"]) + ") & 1;\n")
if (variables["u"] != -1) and (variables["f"] != -1):
fail(ValueError, "Cannot have both unsigned and floating-point", False)
if variables["u"] != -1:
append("int u = (opcode >> " + str(variables["u"]) + ") & 1;\n")
if variables["f"] != -1:
append("int f = (opcode >> " + str(variables["f"]) + ") & 1;\n")
if variables["sz"][0] != -1:
if (mask[31 - variables["sz"][0]] == 0) and ((variables["sz"][1] == 1) or (mask[30 - variables["sz"][0]] == 0)):
append("int size = (opcode >> " + str(variables["sz"][0]) + ") & " + sz2str(variables["sz"][1]) + ";\n")
else:
variables["sz"][0] = -2
if variables["q"] != -1:
if mask[31 - variables["q"]] == 0:
append("int q = (opcode >> " + str(variables["q"]) + ") & 1;\n")
else:
variables["q"] = -2
if variables["registers"]["d"][0] != -1:
assert(variables["registers"]["d"][1] != -1)
append("int d = ((opcode >> " + str(variables["registers"]["d"][0]) + ") & 1) << 4 " + \
"| ((opcode >> " + str(variables["registers"]["d"][1]) + ") & 0xF);\n")
if variables["registers"]["t"][2] != -1:
append("int rt = (opcode >> " + str(variables["registers"]["t"][2]) + ") & 0xF;\n")
if variables["registers"]["n"][0] != -1:
assert(variables["registers"]["n"][1] != -1)
append("int n = ((opcode >> " + str(variables["registers"]["n"][0]) + ") & 1) << 4 " + \
"| ((opcode >> " + str(variables["registers"]["n"][1]) + ") & 0xF);\n")
if variables["registers"]["n"][2] != -1:
append("int rn = (opcode >> " + str(variables["registers"]["n"][2]) + ") & 0xF;\n")
if variables["registers"]["m"][0] != -1:
assert(variables["registers"]["m"][1] != -1)
append("int m = ((opcode >> " + str(variables["registers"]["m"][0]) + ") & 1) << 4 " + \
"| ((opcode >> " + str(variables["registers"]["m"][1]) + ") & 0xF);\n")
if variables["registers"]["m"][2] != -1:
append("int rm = (opcode >> " + str(variables["registers"]["m"][2]) + ") & 0xF;\n")
if variables["w"] != -1:
append("int w = (opcode >> " + str(variables["w"]) + ") & 1;\n")
if variables["cmode"] != -1:
append("int cmode = (opcode >> " + str(variables["cmode"]) + ") & 0xF;\n")
if imms != []:
immssz = sum(map(lambda v: v[1], imms))
tmp = "(" * len(imms)
for immpos, immsz, _ in imms:
if tmp[-1] != '(': tmp = tmp + " << " + str(immsz) + ") | "
tmp = tmp + "((opcode >> " + str(immpos) + ") & " + sz2str(immsz) + ")"
tmp = tmp[1:]
append("uint" + str(nextAvailable(immssz)) + "_t imm" + str(immssz) + " = " + tmp + ";\n")
# Destroy imms since we don't need it anymore, but we do need immssz
imms = [-1, immssz] + imms
add_custom_variables()
append("\nsprintf(ret, \"")
# Assemble the variables into the printf
instText = ' '.join(spltln[curSplt:]).split('<')
instText = [instTextPart.split('>') for instTextPart in instText]
instText = [itp for itp2 in instText for itp in itp2]
# Make the failures nicer
spltln = spltln[:curSplt] + ["\x00"] + instText
# inside '{, '...'}'?
hasOpened = False
printf_args = ""
for idx, text in enumerate(instText):
text = text.replace("&l", "<").replace("&g", ">")
if "+/-" in text:
text = text.replace("+/-", "%s")
printf_args = printf_args + ", (u ? \"\" : \"-\")"
if "{!}" in text:
text = text.replace("{!}", "%s")
printf_args = printf_args + ", (w ? \"!\" : \"\")"
if "{, " in text:
if hasOpened:
fail(AssertionError, "Cannot have an inner curly brace")
text = text.replace("{, ", ", ", 1)
hasOpened = True
if "}" in text:
if not hasOpened:
fail(AssertionError, "Cannot close an inexistant curly brace")
text = text.replace("}", "", 1)
hasOpened = False
if idx % 2:
if text == "c":
if variables["cond"] != -1:
append("%s")
printf_args = printf_args + ", cond"
elif text == "dt":
append("%s")
if imms == []:
printf_args = printf_args + ", dts[" + ("0x8 + (f << 2) + " if variables["f"] != -1 else "")
printf_args = printf_args + ("(u << 2) + " if variables["u"] != -1 else "") + "size]"
else:
fail(ValueError, "Unsupported immediate type {} (on dt)".format(imms[0]))
elif text == "size":
if imms[0] == 1:
append("%d")
printf_args = printf_args + ", 8 << size"
else:
fail(ValueError, "Unsupported immediate type {} (on size)".format(imms[0]))
elif text == "Qd":
# VDUP (ARM core register) only
assert(variables["q"] != -1)
assert((variables["registers"]["t"][0:2] == [-1, -1]) and (variables["registers"]["t"][2] != -1))
append("%s")
printf_args = printf_args + ", vecname[(q << 5) + 0x20 + d]"
elif text == "Dd":
append("%s")
printf_args = printf_args + ", vecname[" + ("(size << 5)" if variables["sz"][0] != -1 else "0x20")
printf_args = printf_args + " + d]"
elif text == "Dd[x]":
append("%s[%d]")
printf_args = printf_args + ", vecname[" + ("(size << 5)" if variables["sz"][0] != -1 else "0x20")
printf_args = printf_args + " + d], decodedImm"
elif text == "dd":
append("%s")
printf_args = printf_args + ", vecname[0x20 + d]"
elif text == "Sd":
append("%s")
printf_args = printf_args + ", vecname[d]"
elif text == "Rt":
append("%s")
printf_args = printf_args + ", regname[rt]"
elif text == "Dn":
append("%s")
printf_args = printf_args + ", vecname[" + ("(size << 5)" if variables["sz"][0] != -1 else "0x20")
printf_args = printf_args + " + n]"
elif text == "Dn[x]":
append("%s[%d]")
printf_args = printf_args + ", vecname[" + ("(size << 5)" if variables["sz"][0] != -1 else "0x20")
printf_args = printf_args + " + n], decodedImm"
elif text == "Sn":
append("%s")
printf_args = printf_args + ", vecname[n]"
elif text == "Rn":
append("%s")
printf_args = printf_args + ", regname[rn]"
elif text == "Dm":
append("%s")
printf_args = printf_args + ", vecname[" + ("(size << 5)" if variables["sz"][0] != -1 else "0x20")
printf_args = printf_args + " + m]"
elif text == "Sm":
append("%s")
printf_args = printf_args + ", vecname[m]"
elif text == "Sm1":
append("%s")
printf_args = printf_args + ", vecname[m + 1]"
elif text == "Rm":
append("%s")
printf_args = printf_args + ", regname[rm]"
elif text == "imm":
if imms[0] == -1:
append("%d")
printf_args = printf_args + ", imm" + str(imms[1])
else:
fail(ValueError, "Unsupported immediate type {}".format(imms[0]))
elif text == "label":
if (imms[0] != -1) or ((imms[1] != 8) and (imms[1] != 24)):
fail(AssertionError, "Unsupported case (label A)")
if imms[1] == 8:
if variables["u"] == -1:
fail(AssertionError, "Unsupported case (label 1.B)")
append("%s%d")
printf_args = printf_args + ", (u ? \"\" : \"-\"), imm8 + 2"
elif imms[1] == 24:
append("%+d")
printf_args = printf_args + ", (imm24 & 0x800000 ? imm24 | 0xFF000000 : imm24) + 2"
elif text.lower() == "list":
if (text == "list") and (variables["sz"] == -1):
fail(AssertionError, "Unsupported case (list A)")
if (imms[0] != -1) or (imms[1] != 8):
fail(AssertionError, "Unsupported case (list B)")
if variables["registers"]["d"][0] + variables["registers"]["d"][1] == -2:
fail(AssertionError, "Unsupported case (list C)")
append("%s")
printf_args = printf_args + ", print_register_list_fpu(d, imm8, "
printf_args = printf_args + ("0" if text == "List" else ("1" if text == "LIst" else "size"))
printf_args = printf_args + ")"
else:
fail(KeyError, "Unknown variable " + text)
else:
if text == "":
continue
text = text.split('\\')
while len(text) > 1:
if text[1][0] == '%':
modifier, text[1] = '%', text[1][1:]
while (len(text[1]) > 1) \
and (text[1][0] in ['0', '1', '2', '3', '4', '5', '6', '7', '8', '9', '+']):
modifier, text[1] = modifier + text[1][0], text[1][1:]
modifier, text[1] = modifier + text[1][0], text[1][1:]
append(text[0] + modifier)
else:
append(text[0] + "%s")
printf_args = printf_args + ", " + text[1]
text = text[2:]
if len(text) == 0:
fail(AssertionError, "Substitution not finished")
append(text[0])
curSplt = curSplt + 1
append("\"" + printf_args + ");\n} else ")
elif spltln[0] == "INVALIDATE":
# All instructions matching left are invalid
# Constant -- set to True when debugging output
# Appends a `#n` (with n being the number of INVALIDATEs before + 1) after `???`
numberInvalids = __debug_forceAllDebugging
for i, val in enumerate(spltln[1:]):
if '/' in val:
ocurBit = curBit
req = val.split('/')
val, req = req[0], '/'.join(req[1:])
if val == '0':
curBit = curBit - 1
mask[31 - curBit] = 1
elif val == '1':
curBit = curBit - 1
mask[31 - curBit] = 1
correctBits[31 - curBit] = 1
elif (val == '(0)') or (val == '(1)'):
# Ignore, even though the result should be undefined...
curBit = curBit - 1
elif val.startswith('<') and val.endswith('>'):
blanklen = int(val[1:-1])
curBit = curBit - blanklen
else:
fail(KeyError, "Unknown value '" + val + "'")
curSplt = i + 2
parse_var_requirements()
if curBit == 0:
break
elif curBit < 0:
fail(KeyError, "Current bit too low (" + str(curBit) + ")")
if curSplt == -1:
fail(KeyError, "Not enough arguments")
elif len(spltln) != curSplt:
fail(KeyError, "Too many arguments")
if 1 in mask:
generate_bin_test()
# No C variable since we're invalidating!
# Now print the invalidation, number if debugging
if numberInvalids:
append("strcpy(ret, \"??? #" + str(invalidationCount) + "\");\n} else ")
invalidationCount = invalidationCount + 1
else:
append("strcpy(ret, \"???\");\n} else ")
else:
# No more input, we invalidated everything!
break
elif spltln[0] == "TODO":
print("\033[96mNote:\033[m todo '{}' (line {})".format(' '.join(spltln[1:]), lnno))
else:
fail(NotImplementedError, spltln[0] + " is not implemented")
except Exception as e:
# Add the failing line in case of error (BaseException is not caught!)
raise Exception(str(lnno + 1) + ": " + ln) from e
# Now the files rebuilding part
# File header and guard
files_header = {
"arm_printer.c": """
#include <stdint.h>
#include <stddef.h>
#include <string.h>
#include <stdio.h>
#include "arm_printer.h"
// conditions
#define cEQ (0b0000<<28)
#define cNE (0b0001<<28)
#define cCS (0b0010<<28)
#define cCC (0b0011<<28)
#define cMI (0b0100<<28)
#define cPL (0b0101<<28)
#define cVS (0b0110<<28)
#define cVC (0b0111<<28)
#define cHI (0b1000<<28)
#define cLS (0b1001<<28)
#define cGE (0b1010<<28)
#define cLT (0b1011<<28)
#define cGT (0b1100<<28)
#define cLE (0b1101<<28)
#define c__ (0b1110<<28) // means all
static const char* conds[16] = {
"EQ", "NE", "CS", "CC",
"MI", "PL", "VS", "VC",
"HI", "LS", "GE", "LT",
"GT", "LE", "", "##"
};
static const char* regname[16] = {
"r0", "r1", "r2", "r3", "r4",
"r5", "r6", "r7", "r8", "r9",
"r10", "r11", "r12", "SP", "LR", "PC"
};
// Single precision are V_:_
// Double and quad precision are _:V_
// So we always use _:V_ and invert the corresponding single-precision (cleaner? code)
static const char* vecname[96] = {
"S0", "S2", "S4", "S6", "S8", "S10", "S12", "S14",
"S16", "S18", "S20", "S22", "S24", "S26", "S28", "S30",
"S1", "S3", "S5", "S7", "S9", "S11", "S13", "S15",
"S17", "S19", "S21", "S23", "S25", "S27", "S29", "S31",
"D0", "D1", "D2", "D3", "D4", "D5", "D6", "D7",
"D8", "D9", "D10", "D11", "D12", "D13", "D14", "D15",
"D16", "D17", "D18", "D19", "D20", "D21", "D22", "D23",
"D24", "D25", "D26", "D27", "D28", "D29", "D30", "D31",
"Q0", "!!!", "Q1", "!!!", "Q2", "!!!", "Q3", "!!!",
"Q4", "!!!", "Q5", "!!!", "Q6", "!!!", "Q7", "!!!",
"Q8", "!!!", "Q9", "!!!", "Q10", "!!!", "Q11", "!!!",
"Q12", "!!!", "Q13", "!!!", "Q14", "!!!", "Q15", "!!!"
};
static const char* dts[16] = {
"S8", "S16", "S32", "S64",
"U8", "U16", "U32", "U64",
"I8", "I16", "I32", "I64",
"F8", "F16", "F32", "F64"
};
static const char* shift_type[4] = {
"lsl ", "lsr ", "asr ", "ror "
};
const char* print_shift(int shift, int comma) {
static __thread char ret[20];
ret[0] = '\\0';
int sh_op = (shift >> 1) & 3;
if(shift & 1) {
int rs = shift >> 4;
sprintf(ret, "%s%s%s", (comma ? ", " : ""), shift_type[sh_op], regname[rs]);
} else {
uint8_t amount = shift >> 3;
if (!amount) {
switch (sh_op) {
case 0b00:
break;
case 0b01:
case 0b10:
sprintf(ret, "%s%s#32", (comma ? ", " : ""), shift_type[sh_op]);
break;
case 0b11:
sprintf(ret, "%srrx", (comma ? ", " : ""));
break;
}
} else {
sprintf(ret, "%s%s#%u", (comma ? ", " : ""), shift_type[sh_op], amount);
}
}
return ret;
}
#define print_modified_imm_ARM(imm12) (((imm12 & 0xFF) >> (2*(imm12 >> 8))) | ((imm12 & 0xFF) << (32 - 2*(imm12 >> 8))))
const char* print_register_list(int list, int size) {
int last = -2;
int cnt = 0;
static __thread char ret[68];
ret[0] = '{';
ret[1] = '\\0';
for (int i = 0; i < size; ++i) {
if (list & (1 << i)) {
if (last >= 0) {
++cnt;
} else {
if (last == -1)
strcat(ret, ", ");
strcat(ret, regname[i]);
last = i;
}
} else {
if (cnt) {
strcat(ret, "-");
strcat(ret, regname[i - 1]);
}
if (last != -2) last = -1;
cnt = 0;
}
}
if (cnt) {
strcat(ret, "-");
strcat(ret, regname[size - 1]);
}
strcat(ret, "}");
return ret;
}
const char* print_register_list_fpu(uint8_t start, uint8_t size, uint8_t double_prec) {
static __thread char ret[68];
// Assume VFPSmallRegisterBank(), so no D16 and no S32 (and above)
if (start + (double_prec + 1) * size > 32) {
return "!!!";
}
if (size == 0) {
return "!!!";
}
if (double_prec && (size % 2 == 1)) {
strcpy(ret, "!! {");
} else {
ret[0] = '{';
ret[1] = '\\0';
}
char regChr;
if (double_prec) {
regChr = 'D';
size /= 2;
} else {
regChr = 'S';
start = (start >> 4) + ((start & 0xF) << 1);
}
char tmp[6];
for (int cur = start; cur < start + size - 1; ++cur) {
sprintf(tmp, "%c%d, ", regChr, cur);
strcat(ret, tmp);
}
sprintf(tmp, "%c%d}", regChr, start + size - 1);
strcat(ret, tmp);
return ret;
}
static const char* _print_modified_imm_I64_bytes[2] = {"00", "FF"};
const char* print_modified_imm_NEON(int op_cmode, int imm) {
static __thread char ret[34] = "???\\0-"; // [dt: 4][imm: 30]
strcpy(&ret[4], "UNPREDICTABLE");
switch (op_cmode) {
case 0b00000: // VMOV.I32
case 0b00001: // VORR.I32
case 0b10000: // VMVN.I32
case 0b10001: // VBIC.I32
strcpy(ret, "I32");
sprintf(&ret[4], "0x%02X", imm);
break;
case 0b00010: // VMOV.I32
case 0b00011: // VORR.I32
case 0b10010: // VMVN.I32
case 0b10011: // VBIC.I32
if (imm) {
strcpy(ret, "I32");
sprintf(&ret[4], "0x%02X00", imm);
}
break;
case 0b00100: // VMOV.I32
case 0b00101: // VORR.I32
case 0b10100: // VMVN.I32
case 0b10101: // VBIC.I32
if (imm) {
strcpy(ret, "I32");
sprintf(&ret[4], "0x%02X0000", imm);
}
break;
case 0b00110: // VMOV.I32
case 0b00111: // VORR.I32
case 0b10110: // VMVN.I32
case 0b10111: // VBIC.I32
if (imm) {
strcpy(ret, "I32");
sprintf(&ret[4], "0x%02X000000", imm);
}
break;
case 0b01000: // VMOV.I16
case 0b01001: // VORR.I16
case 0b11000: // VMVN.I16
case 0b11001: // VBIC.I16
strcpy(ret, "I16");
sprintf(&ret[4], "0x%02X", imm);
break;
case 0b01010: // VMOV.I16
case 0b01011: // VORR.I16
case 0b11010: // VMVN.I16
case 0b11011: // VBIC.I16
if (imm) {
strcpy(ret, "I16");
sprintf(&ret[4], "0x%02X00", imm);
}
break;
case 0b01100: // VMOV.I32
case 0b11100: // VMVN.I32
if (imm) {
strcpy(ret, "I32");
sprintf(&ret[4], "0x%02XFF", imm);
}
break;
case 0b01101: // VMOV.I32
case 0b11101: // VMVN.I32
if (imm) {
strcpy(ret, "I32");
sprintf(&ret[4], "0x%02XFFFF", imm);
}
break;
case 0b01110: // VMOV.I8
strcpy(ret, "I8");
sprintf(&ret[4], "0x%02X", imm);
break;
case 0b01111: // VMOV.F32
strcpy(ret, "F32");
sprintf(&ret[4], "%e", (imm & 0x80 ? -1 : 1) * (1 << (((imm >> 4) & 0x7) ^ 0b100)) * (float) (16 + (imm & 0xF)) / 128.);
break;
case 0b11110: // VMOV.I64
{
strcpy(ret, "I64");
sprintf(&ret[4], "#0x%s%s%s%s%s%s%s%s",
_print_modified_imm_I64_bytes[(imm >> 7) & 1], _print_modified_imm_I64_bytes[(imm >> 6) & 1],
_print_modified_imm_I64_bytes[(imm >> 5) & 1], _print_modified_imm_I64_bytes[(imm >> 4) & 1],
_print_modified_imm_I64_bytes[(imm >> 3) & 1], _print_modified_imm_I64_bytes[(imm >> 2) & 1],
_print_modified_imm_I64_bytes[(imm >> 1) & 1], _print_modified_imm_I64_bytes[(imm >> 0) & 1]);
}
break;
case 0b11111: // V???.-
default:
strcpy(ret, "-");
strcpy(&ret[4], "UNDEFINED");
break;
}
return ret;
}
const char* arm_print(uint32_t opcode) {
static __thread char ret[100];
memset(ret, 0, sizeof(ret));
"""
}
files_guard = {
"arm_printer.c": """
{
strcpy(ret, "???");
}
return ret;
}
"""
}
banner = "/*******************************************************" + ('*'*len(ver)) + "***\n" \
" * File automatically generated by rebuild_printer.py (v" + ver + ") *\n" \
" *******************************************************" + ('*'*len(ver)) + "***/\n"
trim = lambda string: '\n'.join(line[2:] for line in string.splitlines())[1:]
# Yes, the for loops are inverted. This is because both dicts should have the same keys.
for fhdr in files_guard:
files_header[fhdr] = banner + trim(files_header[fhdr])
for fhdr in files_header:
files_guard[fhdr] = trim(files_guard[fhdr])
for f in files_header:
with open(os.path.join(root, "src", "dynarec", f), 'w') as file:
file.write(files_header[f])
file.write(output)
file.write(files_guard[f])
# Save the string for the next iteration, writing was successful
with open(os.path.join(root, "src", "dynarec", "last_run.txt"), 'w') as file:
file.write('\n'.join(insts))
return 0
if __name__ == '__main__':
limit = []
for i, v in enumerate(sys.argv):
if v == "--":
limit.append(i)
if main(sys.argv[1], "1.2.0.05") != 0:
exit(2)
exit(0)