hackern64/ultrasm64-2
0
0
Fork 0
mirror of https://github.com/HackerN64/ultrasm64-2.git synced 2026-01-21 10:38:08 -08:00
Code Issues Packages Projects Releases Wiki Activity
Files
master
ultrasm64-2/lib/hackerlibultra/tools/print_mdebug.py
bicyclesoda 13313f3e31 [WIP] GCC Building + libhackerultra (#17) 
Co-authored-by: a <a>
2025-06-19 21:25:47 -04:00

589 lines
31 KiB
Python
Executable File
Raw Permalink Blame History

#!/usr/bin/env python3
'''
Resources:
http://www.cs.unibo.it/~solmi/teaching/arch_2002-2003/AssemblyLanguageProgDoc.pdf
https://github.com/pathscale/binutils/blob/5c2c133020e41fc4aadd80a99156d2cea4754b96/include/coff/sym.h
https://github.com/pathscale/binutils/blob/5c2c133020e41fc4aadd80a99156d2cea4754b96/include/coff/symconst.h
https://github.com/pathscale/binutils/blob/5c2c133020e41fc4aadd80a99156d2cea4754b96/gas/ecoff.c
https://github.com/pathscale/binutils/blob/5c2c133020e41fc4aadd80a99156d2cea4754b96/bfd/ecoff.c
https://github.com/pathscale/absoft/blob/master/svn/trunk/ekopath-gcc/ekopath-gcc-4.2.0/gcc/mips-tdump.c
https://chromium.googlesource.com/native_client/nacl-toolchain/+/refs/tags/gcc-4.4.3/binutils/gas/ecoff.c
https://android.googlesource.com/toolchain/binutils/+/refs/heads/donut/binutils-2.17/bfd/ecofflink.c
https://kernel.googlesource.com/pub/scm/linux/kernel/git/hjl/binutils/+/hjl/secondary/include/coff/symconst.h
https://c0de.pw/bg/binutils-gdb/blob/cdbf20f73486c66e24f322400eba877eb534ae51/gdb/mdebugread.c
'''
import os
import struct
import collections
import sys
OFFSET = 0 # TODO why are the offsets in the symbolic header off by some amount?
indent_level = 0
is_comment = False
symbol_type_list = [
'stNil', 'stGlobal', 'stStatic', 'stParam', 'stLocal', 'stLabel', 'stProc', 'stBlock',
'stEnd', 'stMember', 'stTypedef', 'stFile', 'INVALID', 'INVALID', 'stStaticProc', 'stConstant',
'stStaParam', 'INVALID', 'INVALID', 'INVALID', 'INVALID', 'INVALID', 'INVALID', 'INVALID',
'INVALID', 'INVALID', 'stStruct', 'stUnion', 'stEnum', 'INVALID', 'INVALID', 'INVALID',
'INVALID', 'INVALID', 'stIndirect']
storage_class_list = ['scNil', 'scText', 'scData', 'scBss', 'scRegister', 'scAbs', 'scUndefined', 'reserved',
'scBits', 'scDbx', 'scRegImage', 'scInfo', 'scUserStruct', 'scSData', 'scSBss', 'scRData',
'scVar', 'scCommon', 'scSCommon', 'scVarRegister', 'scVariant', 'scUndefined', 'scInit']
basic_type_c_list = ['nil', 'addr', 'signed char', 'unsigned char', 'short', 'unsigned short', 'int', 'unsigned int',
'long', 'unsigned long', 'float', 'double', 'struct', 'union', 'enum', 'typedef',
'range', 'set', 'complex', 'double complex', 'indirect', 'fixed decimal', 'float decimal', 'string',
'bit', 'picture', 'void', 'long long', 'unsigned long long', 'INVALID', 'long', 'unsigned long',
'long long', 'unsigned long long', 'addr', 'int64', 'unsigned int64']
def increase_indent():
global indent_level
indent_level += 1
def decrease_indent():
global indent_level
indent_level -= 1
def set_is_comment(set_to):
global is_comment
old = is_comment
is_comment = set_to
return old
def get_indent():
global indent_level
global is_comment
ret = '//' if is_comment else ''
for i in range(indent_level):
ret += ' '
return ret
def check_indent():
global indent_level
assert indent_level == 0
def read_uint32_be(file_data, offset):
return struct.unpack('>I', file_data[offset:offset+4])[0]
def read_uint16_be(file_data, offset):
return struct.unpack('>H', file_data[offset:offset+2])[0]
def read_uint8_be(file_data, offset):
return struct.unpack('>B', file_data[offset:offset+1])[0]
def read_elf_header(file_data, offset):
Header = collections.namedtuple('ElfHeader',
'''e_magic e_class e_data e_version e_osabi e_abiversion e_pad
e_type e_machine e_version2 e_entry e_phoff e_shoff e_flags
e_ehsize e_phentsize e_phnum e_shentsize e_shnum e_shstrndx''')
return Header._make(struct.unpack('>I5B7s2H5I6H', file_data[offset:offset+52]))
def read_elf_section_header(file_data, offset):
Header = collections.namedtuple('SectionHeader',
'''sh_name sh_type sh_flags sh_addr sh_offset sh_size sh_link
sh_info sh_addralign sh_entsize''')
return Header._make(struct.unpack('>10I', file_data[offset:offset+40]))
def read_symbolic_header(file_data, offset):
Header = collections.namedtuple('SymbolicHeader',
'''magic vstamp ilineMax cbLine cbLineOffset idnMax cbDnOffset
ipdMax cbPdOffset isymMax cbSymOffset ioptMax cbOptOffset
iauxMax cbAuxOffset issMax cbSsOffset issExtMax cbSsExtOffset
ifdMax cbFdOffset crfd cbRfdOffset iextMax cbExtOffset''')
return Header._make(struct.unpack('>2H23I', file_data[offset:offset+96]))
# TODO find a better solution for the bitfield
def read_file_descriptor(file_data, offset):
if 'init' not in read_file_descriptor.__dict__:
read_file_descriptor.cache = {}
read_file_descriptor.header = collections.namedtuple('FileDescriptor',
'''adr rss issBase cbSs isymBase csym ilineBase cline ioptBase
copt ipdFirst cpd iauxBase caux rfdBase crfd XXX_bitfield
cbLineOffset cbLine''')
read_file_descriptor.init = True
if offset in read_file_descriptor.cache:
return read_file_descriptor.cache[offset]
read_file_descriptor.cache[offset] = read_file_descriptor.header._make(
struct.unpack('>I2iI6iHh4iI2I', file_data[offset:offset+72]))
return read_file_descriptor.cache[offset]
def read_procedure_descriptor(file_data, offset):
Header = collections.namedtuple('ProcedureDescriptor',
'''adr isym iline regmask regoffset iopt fregmask fregoffset
frameoffset framereg pcreg lnLow lnHigh cbLineOffset''')
return Header._make(struct.unpack('>I8i2h2iI', file_data[offset:offset+52]))
def read_symbol(file_data, offset):
if 'init' not in read_symbol.__dict__:
read_symbol.cache = {}
read_symbol.header = collections.namedtuple('Symbol', '''iss value st sc index''')
read_symbol.init = True
if offset in read_symbol.cache:
return read_symbol.cache[offset]
(word0, word1, word2) = struct.unpack('>iiI', file_data[offset:offset+12])
read_symbol.cache[offset] = read_symbol.header._make((
word0, word1, (word2 >> 26) & 0x3F, (word2 >> 21) & 0x1F, word2 & 0xFFFFF))
return read_symbol.cache[offset]
def read_auxiliary_symbol(file_data, offset):
if 'init' not in read_auxiliary_symbol.__dict__:
read_auxiliary_symbol.cache = {}
read_auxiliary_symbol.header = collections.namedtuple('AuxSymbol',
'''ti rndx dnLow dnHigh isym iss width count''')
read_auxiliary_symbol.type_info = collections.namedtuple('TypeInfo',
'''fBitfield continued bt tq4 tq5 tq0 tq1 tq2 tq3''')
read_auxiliary_symbol.rel_sym = collections.namedtuple('RelativeSymbol', '''rfd index''')
read_auxiliary_symbol.init = True
if offset in read_auxiliary_symbol.cache:
return read_auxiliary_symbol.cache[offset]
word0 = struct.unpack('>I', file_data[offset:offset+4])[0]
read_auxiliary_symbol.cache[offset] = read_auxiliary_symbol.header._make((
read_auxiliary_symbol.type_info._make(((word0 >> 31) & 1, (word0 >> 30) & 1, (word0 >> 24) & 0x3F, (word0 >> 20) & 0xF, (word0 >> 16) & 0xF, (word0 >> 12) & 0xF, (word0 >> 8) & 0xF, (word0 >> 4) & 0xF, word0 & 0xF)),
read_auxiliary_symbol.rel_sym._make(((word0 >> 20) & 0xFFF, word0 & 0xFFFFF)),
word0, word0, word0, word0, word0, word0))
return read_auxiliary_symbol.cache[offset]
def read_string(file_data, offset):
current_offset = 0
current_string = b''
while True:
char = struct.unpack('c', file_data[offset+current_offset:offset+current_offset+1])[0]
if char == b'\0':
return current_string.decode('ascii')
else:
current_string += char
current_offset += 1
def map_relative_file_descriptor(file_data, fd, symbolic_header, rfd_num):
if fd.crfd == 0:
return rfd_num
offset = symbolic_header.cbRfdOffset - OFFSET + (fd.rfdBase + rfd_num)*4
return struct.unpack('>I', file_data[offset:offset+4])[0]
def get_symbol_name_from_aux(file_data, fd, symbolic_header, aux_num, search_for_typedef):
aux = read_auxiliary_symbol(file_data, symbolic_header.cbAuxOffset - OFFSET + (fd.iauxBase + aux_num)*4)
fd_num = aux.rndx.rfd
next_aux = aux_num+1
if fd_num == 4095:
aux2 = read_auxiliary_symbol(file_data, symbolic_header.cbAuxOffset - OFFSET + (fd.iauxBase + next_aux)*4)
fd_num = aux2.isym
next_aux = next_aux+1;
fd_num = map_relative_file_descriptor(file_data, fd, symbolic_header, fd_num)
fd2 = read_file_descriptor(file_data, symbolic_header.cbFdOffset - OFFSET + fd_num*72)
sym = read_symbol(file_data, symbolic_header.cbSymOffset - OFFSET + (fd2.isymBase + aux.rndx.index)*12)
ret = ''
#print('%r' % (aux,));
#print('%r' % (aux2,));
#print('%r' % (sym,));
if sym.st == 26 or sym.st == 27: #stStruct, stunion
ret = get_struct_or_union_string(file_data, fd2, symbolic_header, fd2.isymBase + aux.rndx.index, search_for_typedef)
elif sym.st == 28: #stEnum:
ret = get_enum_string(file_data, fd2, symbolic_header, fd2.isymBase + aux.rndx.index)
else:
ret = read_string(file_data, symbolic_header.cbSsOffset - OFFSET + fd2.issBase + sym.iss)
return (ret, next_aux)
def get_type_string(file_data, fd, symbolic_header, aux_num, name, search_for_typedef):
ret = ''
aux = read_auxiliary_symbol(file_data, symbolic_header.cbAuxOffset - OFFSET + (fd.iauxBase + aux_num)*4)
#print('');
#print('%r' % (aux,));
next_aux = aux_num+1
has_bitfield = aux.ti.fBitfield == 1
bitwidth = 0
if has_bitfield:
bit_aux = read_auxiliary_symbol(file_data, symbolic_header.cbAuxOffset - OFFSET + (fd.iauxBase + next_aux)*4)
bitwidth = bit_aux.isym
next_aux = next_aux+1
if aux.ti.bt == 12: # btStruct
(ret, next_aux) = get_symbol_name_from_aux(file_data, fd, symbolic_header, next_aux, search_for_typedef)
elif aux.ti.bt == 13: # btUnion
(ret, next_aux) = get_symbol_name_from_aux(file_data, fd, symbolic_header, next_aux, search_for_typedef)
elif aux.ti.bt == 15: # btTypedef
(ret, next_aux) = get_symbol_name_from_aux(file_data, fd, symbolic_header, next_aux, search_for_typedef)
elif aux.ti.bt == 14: # btEnum
(ret, next_aux) = get_symbol_name_from_aux(file_data, fd, symbolic_header, next_aux, search_for_typedef)
else:
if aux.ti.bt >= 36:
print('Error unknow bt: %d' % (aux.ti.bt))
ret = basic_type_c_list[aux.ti.bt]
tq_list = (aux.ti.tq0, aux.ti.tq1, aux.ti.tq2, aux.ti.tq3, aux.ti.tq4, aux.ti.tq5)
# TODO this is very naive and probably does not work in a large amount of cases
last_was_proc = False # if we see a tqProc, assume the next will be a tqPtr
for tq in tq_list:
if tq == 0: # tqNil
break;
elif tq == 1: # tqPtr
if last_was_proc:
last_was_proc = False
continue
ret += '*'
elif tq == 2: # tqProc
last_was_proc = True
name = '(*%s)(/* ECOFF does not store param types */)' % name
elif tq == 3: # tqArray
next_aux += 2 # todo what does this skip over? (Apparantly the type of the index, so always int for C)
array_low_aux = read_auxiliary_symbol(file_data, symbolic_header.cbAuxOffset - OFFSET + (fd.iauxBase + next_aux)*4)
array_high_aux = read_auxiliary_symbol(file_data, symbolic_header.cbAuxOffset - OFFSET + (fd.iauxBase + next_aux+1)*4)
stride_aux = read_auxiliary_symbol(file_data, symbolic_header.cbAuxOffset - OFFSET + (fd.iauxBase + next_aux+2)*4)
next_aux += 3
if array_high_aux.dnHigh == 0xFFFFFFFF:
name += '[]'
else:
name += '[%d]' % (array_high_aux.dnHigh + 1)
elif tq == 4: # tqFar
print('ERROR tqFar in get_type_name')
elif tq == 5: # tqVol
ret = 'volatile ' + ret
elif tq == 6: # tqConst
ret = 'const ' + ret
if has_bitfield:
name += ' : %d' % bitwidth
return ret + ' ' + name
def get_enum_string(file_data, fd, symbolic_header, enum_sym_num):
ret = ''
start_sym = read_symbol(file_data, symbolic_header.cbSymOffset - OFFSET + enum_sym_num*12)
if start_sym.st != 28:
print('ERROR unknown type in get_enum_string start:%d' % start_sym.st)
return ret
name = read_string(file_data, symbolic_header.cbSsOffset - OFFSET + fd.issBase + start_sym.iss)
if name != '':
name += ' '
ret += 'enum %s{\n' % name
increase_indent()
sym_num = enum_sym_num + 1
while sym_num < fd.isymBase + start_sym.index:
sym = read_symbol(file_data, symbolic_header.cbSymOffset - OFFSET + sym_num*12)
if sym.st == 8: # stEnd
decrease_indent()
ret += get_indent()
ret += '}'
elif sym.st == 9: # stMember
name = read_string(file_data, symbolic_header.cbSsOffset - OFFSET + fd.issBase + sym.iss)
ret += get_indent()
ret += '%s = %d,\n' % (name, sym.value)
else:
print('ERROR unknown type in get_enum_string:%d' % sym.st)
break
sym_num += 1
return ret
def get_struct_or_union_string(file_data, fd, symbolic_header, union_sym_num, search_for_typedef):
ret = ''
start_sym = read_symbol(file_data, symbolic_header.cbSymOffset - OFFSET + union_sym_num*12)
if search_for_typedef:
typedef_sym = read_symbol(file_data, symbolic_header.cbSymOffset - OFFSET + (fd.isymBase + start_sym.index)*12)
if typedef_sym.st == 10: # stTypedef
return read_string(file_data, symbolic_header.cbSsOffset - OFFSET + fd.issBase + typedef_sym.iss)
else:
name = read_string(file_data, symbolic_header.cbSsOffset - OFFSET + fd.issBase + start_sym.iss)
if name != '':
return name
name = read_string(file_data, symbolic_header.cbSsOffset - OFFSET + fd.issBase + start_sym.iss)
if name != '':
name += ' '
if start_sym.st == 26: # stStruct
ret += 'struct %s{\n' % name
increase_indent()
elif start_sym.st == 27: # stUnion
ret += 'union %s{\n' % name
increase_indent()
else:
print('ERROR unknown type in get_struct_or_union_string start:%d' % start_sym.st)
return ret
sym_num = union_sym_num + 1
while sym_num < fd.isymBase + start_sym.index:
sym = read_symbol(file_data, symbolic_header.cbSymOffset - OFFSET + sym_num*12)
if sym.st == 8: # stEnd
decrease_indent()
ret += get_indent()
ret += '}'
elif sym.st == 9: # stMember
name = read_string(file_data, symbolic_header.cbSsOffset - OFFSET + fd.issBase + sym.iss)
ret += get_indent()
ret += '/* 0x%X */ %s;\n' % (sym.value // 8, get_type_string(file_data, fd, symbolic_header, sym.index, name, True))
elif sym.st == 26 or sym.st == 27: #stStruct, stUnion
sym_num = fd.isymBase + sym.index
continue
elif sym.st == 34: # stIndirect
# TODO what even is a stIndirect?
sym_num += 1
else:
print('ERROR unknown type in get_struct_or_union_string:%d' % sym.st)
break
sym_num += 1
return ret
def print_typedef_symbols(file_data, fd, symbolic_header, typedef_sym_num):
typedef_sym = read_symbol(file_data, symbolic_header.cbSymOffset - OFFSET + typedef_sym_num*12)
if typedef_sym.st != 10: # stTypedef
print('ERROR expected stTypedef symbol in print_typedef_symbols, found:%d' % typedef_sym.st)
return
name = read_string(file_data, symbolic_header.cbSsOffset - OFFSET + fd.issBase + typedef_sym.iss)
print('typedef %s;' % get_type_string(file_data, fd, symbolic_header, typedef_sym.index, name, False))
def print_procedure(file_data, fd, symbolic_header, proc_sym_num):
proc_sym = read_symbol(file_data, symbolic_header.cbSymOffset - OFFSET + proc_sym_num*12)
proc_name = read_string(file_data, symbolic_header.cbSsOffset - OFFSET + fd.issBase + proc_sym.iss)
print('%s(' % get_type_string(file_data, fd, symbolic_header, proc_sym.index+1, proc_name, True), end='')
sym_num = proc_sym_num+1
param_sym = read_symbol(file_data, symbolic_header.cbSymOffset - OFFSET + sym_num*12)
first = True
while param_sym.st == 3: # stParam
param_name = read_string(file_data, symbolic_header.cbSsOffset - OFFSET + fd.issBase + param_sym.iss)
print('%s%s' % ('' if first else ', ',
get_type_string(file_data, fd, symbolic_header, param_sym.index, param_name, True)),
end='')
sym_num += 1
param_sym = read_symbol(file_data, symbolic_header.cbSymOffset - OFFSET + sym_num*12)
first = False
print(');')
check_indent();
comment_old = set_is_comment(True)
while sym_num < fd.isymBase + fd.csym:
sym = read_symbol(file_data, symbolic_header.cbSymOffset - OFFSET + sym_num*12)
sym_num += 1
if sym.st == 7: # stBlock
print('%s{' % get_indent())
increase_indent()
elif sym.st == 8: # stEnd
end_name = read_string(file_data, symbolic_header.cbSsOffset - OFFSET + fd.issBase + sym.iss)
if end_name == proc_name:
break
if end_name != '':
# this is a stEnd for something other than the function. Let's back out and return
sym_num -= 1
break
decrease_indent()
print('%s}' % get_indent())
elif sym.st == 4: # stLocal
local_name = read_string(file_data, symbolic_header.cbSsOffset - OFFSET + fd.issBase + sym.iss)
is_reg = sym.sc == 4 # scRegister
print('%s%s%s;' % (get_indent(),
'register ' if is_reg else '',
get_type_string(file_data, fd, symbolic_header, sym.index, local_name, True)))
elif sym.st == 2: # stStatic
static_name = read_string(file_data, symbolic_header.cbSsOffset - OFFSET + fd.issBase + sym.iss)
if sym.index == 0xFFFFF:
print('%sstatic %s; // no type symbol' % (get_indent(),static_name))
else:
print('%sstatic %s;' % (get_indent(),get_type_string(file_data, fd, symbolic_header, sym.index, static_name, True)))
elif sym.st == 5: # stLabel
label_name = read_string(file_data, symbolic_header.cbSsOffset - OFFSET + fd.issBase + sym.iss)
print('%sLabel: %s @ %d;' % (get_indent(), label_name, sym.value))
elif sym.st == 6: # stProc
# multiple name for function?
sym_num = print_procedure(file_data, fd, symbolic_header, sym_num-1)
elif sym.st == 26 or sym.st == 27: # stStruct, stUnion
print('%s%s;' % (get_indent(), get_struct_or_union_string(file_data, fd, symbolic_header, sym_num-1, False)))
sym_num = fd.isymBase + sym.index
elif sym.st == 28: # stEnum
print('%s%s;' % (get_indent(), get_enum_string(file_data, fd, symbolic_header, sym_num-1)))
sym_num = fd.isymBase + sym.index
elif sym.st == 34: # stIndirect
# TODO what even is a stIndirect?
indirect_name = read_string(file_data, symbolic_header.cbSsOffset - OFFSET + fd.issBase + sym.iss)
print('%sTODO Indirect: %s;' % (get_indent(), indirect_name))
else:
print('ERROR unknown st in print_procedure: %d' % sym.st)
set_is_comment(comment_old)
check_indent();
return sym_num
def print_symbols(file_data, fd, symbolic_header):
sym_num = fd.isymBase
indirects = []
typedefs = []
while sym_num < fd.isymBase + fd.csym:
root_sym = read_symbol(file_data, symbolic_header.cbSymOffset - OFFSET + sym_num*12)
if root_sym.st == 10: # stTypedef
aux = read_auxiliary_symbol(file_data, symbolic_header.cbAuxOffset - OFFSET + (fd.iauxBase + root_sym.index)*4)
offset = 0
if aux.ti.fBitfield == 1:
offset = 1
if aux.ti.bt == 12 or aux.ti.bt == 13 or aux.ti.bt == 14 or aux.ti.bt == 15: # btStruct, btUnion, btEnum, btTypedef
aux2 = read_auxiliary_symbol(file_data, symbolic_header.cbAuxOffset - OFFSET + (fd.iauxBase + root_sym.index + 1 + offset)*4)
fd_num = aux2.rndx.rfd
if fd_num == 4095:
fd_num = read_auxiliary_symbol(file_data, symbolic_header.cbAuxOffset - OFFSET + (fd.iauxBase + root_sym.index + 2 + offset)*4).isym
fd_num = map_relative_file_descriptor(file_data, fd, symbolic_header, fd_num)
fd2 = read_file_descriptor(file_data, symbolic_header.cbFdOffset - OFFSET + fd_num*72)
sym2 = read_symbol(file_data, symbolic_header.cbSymOffset - OFFSET + (fd2.isymBase + aux2.rndx.index)*12)
name = read_string(file_data, symbolic_header.cbSsOffset - OFFSET + fd2.issBase + sym2.iss)
if name != '':
typedefs.append(name)
elif root_sym.st == 34: # stIndirect
name = read_string(file_data, symbolic_header.cbSsOffset - OFFSET + fd.issBase + root_sym.iss)
indirects.append(name);
sym_num += 1
sym_num = fd.isymBase
while sym_num < fd.isymBase + fd.csym:
root_sym = read_symbol(file_data, symbolic_header.cbSymOffset - OFFSET + sym_num*12)
if root_sym.st == 11: # stFile
file_name = read_string(file_data, symbolic_header.cbSsOffset - OFFSET + fd.issBase + root_sym.iss)
print('// begin file %s\n' % file_name)
sym_num += 1
leaf_sym = read_symbol(file_data, symbolic_header.cbSymOffset - OFFSET + sym_num*12)
while leaf_sym.st != 8: # stEnd
if leaf_sym.st == 26 or leaf_sym.st == 27: # stStruct, stUnion
name = read_string(file_data, symbolic_header.cbSsOffset - OFFSET + fd.issBase + leaf_sym.iss)
if (name != '') and ((name in indirects) or (name not in typedefs)): # TODO
print('%s;\n' % get_struct_or_union_string(file_data, fd, symbolic_header, sym_num, False))
sym_num = fd.isymBase + leaf_sym.index
elif leaf_sym.st == 28: # stEnum
name = read_string(file_data, symbolic_header.cbSsOffset - OFFSET + fd.issBase + leaf_sym.iss)
if (name != '') and (name not in typedefs): # TODO
print('%s;\n' % get_enum_string(file_data, fd, symbolic_header, sym_num))
sym_num = fd.isymBase + leaf_sym.index
elif leaf_sym.st == 10: # stTypedef
# TODO typedef for stIndirect shoulf print the keyword i.e. typdef >struct< THING thing
print_typedef_symbols(file_data, fd, symbolic_header, sym_num)
sym_num += 1
print('')
elif leaf_sym.st == 6 or leaf_sym.st == 14: # stProc, stStaticProc
# TODO how do stProc and stStaticProc differ? stStaticProc isn't exported?
sym_num = print_procedure(file_data, fd, symbolic_header, sym_num)
print('')
elif leaf_sym.st == 2: # stStatic
static_name = read_string(file_data, symbolic_header.cbSsOffset - OFFSET + fd.issBase + leaf_sym.iss)
if leaf_sym.sc == 2 or leaf_sym.sc == 3 or leaf_sym.sc == 5 or leaf_sym.sc == 15: # scData, scBss, scAbsolute, scRData
if leaf_sym.index != 0xFFFFF: # looks like it's an invalid value for .s files
print('static %s;\n' % get_type_string(file_data, fd, symbolic_header, leaf_sym.index, static_name, True))
else:
print('static %s;\n' % static_name)
else:
print('ERROR unknown sc for stStatic in print_symbols: %d' % leaf_sym.sc)
sym_num += 1
elif leaf_sym.st == 34: # stIndirect
# stIndirect is put out when the compiler sees a struct when it is not yet defined
# TODO more info
sym_num += 1
elif leaf_sym.st == 5: # stLabel
print('// label: %s' % read_string(file_data, symbolic_header.cbSsOffset - OFFSET + fd.issBase + leaf_sym.iss))
sym_num += 1
elif leaf_sym.st == 0: # stNil
print('// nil: %s' % read_string(file_data, symbolic_header.cbSsOffset - OFFSET + fd.issBase + leaf_sym.iss))
sym_num += 1
else:
print('ERROR unknown st in leaf_sym in print_symbols: %d' % leaf_sym.st)
sym_num += 1
leaf_sym = read_symbol(file_data, symbolic_header.cbSymOffset - OFFSET + sym_num*12)
sym_num = fd.isymBase + root_sym.index
print('// end file %s' % file_name)
else:
print('ERROR expected st of stFile as only root type in print_symbols:%d' % root_sym.st)
return
def main():
global OFFSET
if len(sys.argv) < 2:
return # TODO print usage
filename = sys.argv[1]
try:
with open(filename, 'rb') as f:
file_data = f.read()
except IOError:
print('failed to read file ' + filename)
return
elf_header = read_elf_header(file_data, 0)
section_headers = []
debug_index = 0xFFFFFFFF
#print('%r' % (elf_header,))
for i in range(elf_header.e_shnum):
section_headers.append(read_elf_section_header(file_data, elf_header.e_shoff + i*40))
#print('%r' % (section_headers[i],))
if section_headers[i].sh_type == 0x70000005:
debug_index = i
if debug_index != 0xFFFFFFFF:
symbolic_header = read_symbolic_header(file_data, section_headers[debug_index].sh_offset)
file_descriptors = []
print('%r' % (symbolic_header,))
# Set offset by assuming that there are no optimization symbols so cbOptOffset points to the start of the symbolic header
#OFFSET = symbolic_header.cbOptOffset - section_headers[debug_index].sh_offset
#print('Using OFFSET of %d' % OFFSET)
#for sym_num in range(symbolic_header.isymMax):
#sym = read_symbol(file_data, symbolic_header.cbSymOffset - OFFSET + sym_num*12)
#print('%d:%r' % (sym_num, (sym,)));
#for aux_num in range(symbolic_header.iauxMax):
#aux = read_auxiliary_symbol(file_data, symbolic_header.cbAuxOffset - OFFSET + aux_num*4)
#print('%d:%r' % (aux_num, (aux,)));
for file_num in range(symbolic_header.ifdMax):
fd = read_file_descriptor(file_data, symbolic_header.cbFdOffset - OFFSET + file_num*72)
file_descriptors.append(fd)
for file_num in range(symbolic_header.ifdMax):
fd = read_file_descriptor(file_data, symbolic_header.cbFdOffset - OFFSET + file_num*72)
print('%r' % (fd,))
print(' name:%s' % read_string(file_data, symbolic_header.cbSsOffset - OFFSET + fd.issBase + fd.rss))
'''
print(' Relative File Descriptors:')
for rfd_num in range(fd.rfdBase, fd.rfdBase + fd.crfd):
offset = symbolic_header.cbRfdOffset - OFFSET + rfd_num*4
rfd_index = struct.unpack('>I', file_data[offset:offset+4])[0]
rfd = read_file_descriptor(file_data, symbolic_header.cbFdOffset - OFFSET + rfd_index*72)
print(' %d:%r' % (rfd_index, (rfd,)))
'''
'''
print(' procedures:')
for proc_num in range(fd.ipdFirst, fd.ipdFirst + fd.cpd):
pd = read_procedure_descriptor(file_data, symbolic_header.cbPdOffset - OFFSET + proc_num*52)
print(' %r' % ((pd,)))
'''
'''
print(' symbols:')
for sym_num in range(fd.isymBase, fd.isymBase + fd.csym):
sym = read_symbol(file_data, symbolic_header.cbSymOffset - OFFSET + sym_num*12)
print(' %r' % ((sym,)))
print(' name:%s' % read_string(file_data, symbolic_header.cbSsOffset - OFFSET + fd.issBase + sym.iss))
print(' type:%s(%d)' % (symbol_type_list[sym.st], sym.st))
print(' storage class:%s(%d)' % (storage_class_list[sym.sc], sym.sc))
if sym.st == 3 or sym.st == 4 or sym.st == 9 or sym.st == 10 or sym.st == 28: # stParam, stLocal, stMember, stTypedef, stEnum
aux = read_auxiliary_symbol(file_data, symbolic_header.cbAuxOffset - OFFSET + (fd.iauxBase + sym.index)*4)
print(' %r' % ((aux,)))
offset = 0
if aux.ti.fBitfield == 1:
bitfield_aux = read_auxiliary_symbol(file_data, symbolic_header.cbAuxOffset - OFFSET + (fd.iauxBase + sym.index + 1)*4)
print(' %r' % ((bitfield_aux,)))
offset = 1
if aux.ti.bt == 12 or aux.ti.bt == 13 or aux.ti.bt == 14 or aux.ti.bt == 15: # btStruct, btUnion, btEnum, btTypedef
aux2 = read_auxiliary_symbol(file_data, symbolic_header.cbAuxOffset - OFFSET + (fd.iauxBase + sym.index + 1 + offset)*4)
print(' %r' % ((aux2,)))
fd_num = aux2.rndx.rfd
if fd_num == 4095:
aux3 = read_auxiliary_symbol(file_data, symbolic_header.cbAuxOffset - OFFSET + (fd.iauxBase + sym.index + 2 + offset)*4)
print(' %r' % ((aux3,)))
fd_num = aux3.isym
fd_num = map_relative_file_descriptor(file_data, fd, symbolic_header, fd_num)
sym2 = read_symbol(file_data, symbolic_header.cbSymOffset - OFFSET + (file_descriptors[fd_num].isymBase + aux2.rndx.index)*12)
print(' %r' % (sym2,))
print(' name:%s' % read_string(file_data, symbolic_header.cbSsOffset - OFFSET + file_descriptors[aux3.isym].issBase + sym2.iss))
if sym.st == 6 or sym.st == 14: # stProc, stStaticProc
# TODO what is the first aux symbol for?
aux = read_auxiliary_symbol(file_data, symbolic_header.cbAuxOffset - OFFSET + (fd.iauxBase + sym.index)*4)
type_aux = read_auxiliary_symbol(file_data, symbolic_header.cbAuxOffset - OFFSET + (fd.iauxBase + sym.index+1)*4)
print(' %r' % ((aux,)))
print(' %r' % ((type_aux,)))
'''
print(' pretty print:')
print_symbols(file_data, fd, symbolic_header)
main()
Reference in New Issue View Git Blame Copy Permalink