vitadump_2x.py

gistfile1.txt

#!/usr/bin/env python3

import os.path
import struct

from idautils import *
from idc import *

INFO_SIZE = 0x5c
NAME_OFF = 4
NAME_LEN = 27
ENT_TOP_OFF = 0x24
ENT_END_OFF = 0x28
STUB_TOP_OFF = 0x2c
STUB_END_OFF = 0x30

EXPORT_NUM_FUNCS_OFF = 0x6
EXPORT_NID_OFF = 0x10
EXPORT_LIBNAME_OFF = 0x14
EXPORT_NID_TABLE_OFF = 0x18
EXPORT_ENTRY_TABLE_OFF = 0x1c

IMPORT_NUM_FUNCS_OFF = 0x6
IMPORT_LIBNAME_OFF = 0x14
IMPORT_LIBNAME_OFF2 = 0x10
IMPORT_NID_TABLE_OFF = 0x1c
IMPORT_NID_TABLE_OFF2 = 0x14
IMPORT_ENTRY_TABLE_OFF = 0x20
IMPORT_ENTRY_TABLE_OFF2 = 0x18

NORETURN_FUNCS = [0xB997493D, 0x391B5B74, 0x00CCE39C, 0x37691BF8, 0x2f2c6046]


def u32(bytes, start=0):
    return struct.unpack("<I", bytes[start:start + 4])[0]


def u16(bytes, start=0):
    return struct.unpack("<H", bytes[start:start + 2])[0]


def u8(bytes, start=0):
    return struct.unpack("<B", bytes[start:start + 2])[0]


def read_cstring(addr, max_len=0):
    s = ""
    ea = addr
    while True:
        c = Byte(ea)
        if c == 0:
            break
        ea += 1
        s += chr(c)
        if max_len and len(s) > max_len:
            break
    return s


def chunk(s, l):
    """
        Chunks S into strings of length L, for example:
        >>> chunk("abcd", 2)
        ["ab", "cd"]
        >>> chunk("abcde", 2)
        ['ab', 'cd', 'e']
    """
    return [s[i:i + l] for i in range(0, len(s), l)]


nid_table = dict()

def load_nids(filename):
    if not os.path.exists(filename):
        print "cannot find nids.txt, NIDs won't be resolved"
        return
    fin = open(filename, "r")
    for line in fin.readlines():
        line = line.split()
        nid_table[int(line[0], 16)] = line[1]
    fin.close()
    print "Loaded {0} NIDs".format(len(nid_table))


def resolve_nid(nid):
    if nid in nid_table:
        return nid_table[nid]
    return ""

used_names = dict()

def rename_function(ea, name, suffix):
    """
        Renames a function, optionally adding a _XX suffix to make sure
        all names are unique.
    """
    name = name + suffix
    if name in used_names:
        used_names[name] += 1
        name += "_{}".format(used_names[name])
    else:
        used_names[name] = 0

    MakeName(ea, name)

def process_nid_table(nid_table_addr, entry_table_addr, num_funcs, libname, name_suffix=""):
    if num_funcs == 0:
        return

    nids = GetManyBytes(nid_table_addr, 4 * num_funcs)
    funcs = GetManyBytes(entry_table_addr, 4 * num_funcs)

    if not nids or not funcs:
        print "NID table at 0x{0:x} is not supported, bailing out!".format(nid_table_addr)
        return

    for nid, func in zip(chunk(nids, 4), chunk(funcs, 4)):
        nid = u32(nid)
        func = u32(func)
        print("nid {0} => func {1}".format(hex(nid), hex(func)))
        t_reg = func & 1  # 0 = ARM, 1 = THUMB
        func -= t_reg
        for i in range(4):
            SetReg(func + i, "T", t_reg)
        MakeFunction(func)

        actual_name = name = resolve_nid(nid)
        if not name:
            name = "{0}_{1:08X}".format(libname, nid)

        rename_function(func, name, name_suffix)

        if nid in NORETURN_FUNCS:
            SetFunctionFlags(func, FUNC_NORET)

        # add a comment to mangled functions with demangled name, but only for imports
        # or otherwise when ida wouldn't do it itself because of non empty suffix
        if actual_name.startswith("_Z") and name_suffix:
            demangled = Demangle(actual_name, GetLongPrm(INF_LONG_DN))
            if demangled != "":
                SetFunctionCmt(func, demangled, 1)


def process_export(exp, libname):
    num_funcs = u16(exp, EXPORT_NUM_FUNCS_OFF)
    nid_table = u32(exp, EXPORT_NID_TABLE_OFF)
    entry_table = u32(exp, EXPORT_ENTRY_TABLE_OFF)
    libname_addr = u32(exp, EXPORT_LIBNAME_OFF)
    nid = u32(exp, EXPORT_NID_OFF)
    libname = ""
    if libname_addr:
        libname = read_cstring(libname_addr, 255)

    print "{0} with NID 0x{1:x}".format(libname, nid)

    process_nid_table(nid_table, entry_table, num_funcs, libname)


def process_import(imp):
    num_funcs = u16(imp, IMPORT_NUM_FUNCS_OFF)
    nid_table = u32(imp, IMPORT_NID_TABLE_OFF if len(imp) == 0x34 else IMPORT_NID_TABLE_OFF2)
    entry_table = u32(imp, IMPORT_ENTRY_TABLE_OFF if len(imp) == 0x34 else IMPORT_ENTRY_TABLE_OFF2)
    libname_addr = u32(imp, IMPORT_LIBNAME_OFF if len(imp) == 0x34 else IMPORT_LIBNAME_OFF2)

    if not libname_addr:
        return

    libname = read_cstring(libname_addr, 255)
    process_nid_table(nid_table, entry_table, num_funcs, libname, "_imp_")


def process_module(module_info_addr):
    module_info = GetManyBytes(module_info_addr, INFO_SIZE)
    name = module_info[NAME_OFF:NAME_OFF+NAME_LEN].strip("\x00")
    ent_top = u32(module_info, ENT_TOP_OFF)
    ent_end = u32(module_info, ENT_END_OFF)
    ent_len = ent_end - ent_top
    stub_top = u32(module_info, STUB_TOP_OFF)
    stub_end = u32(module_info, STUB_END_OFF)
    stub_len = stub_end - stub_top
    print "Library {0} {1}".format(name, hex(module_info_addr))

    exports = []
    base_addr = addr = module_info_addr + INFO_SIZE
    while addr - base_addr < ent_end - ent_top:
        size = u8(GetManyBytes(addr, 1))
        exports.append((addr, size))
        addr += size

    imports = []
    base_addr = addr
    while addr - base_addr < stub_end - stub_top:
        size = u8(GetManyBytes(addr, 1))
        imports.append((addr, size))
        addr += size

    # We need to process imports first so that noreturn functions are found
    for addr, size in imports:
        process_import(GetManyBytes(addr, size))
    for addr, size in exports:
        process_export(GetManyBytes(addr, size), name)


def find_modules_with_string(haystack, off):
    ea = 0
    c = " ".join(chunk(haystack.encode("hex"), 2))
    while ea != BADADDR:
        ea = FindBinary(ea, SEARCH_DOWN | SEARCH_CASE, c)
        if ea != BADADDR:
            process_module(ea + off)
        ea = NextAddr(ea)

def find_modules():
    module_heur = [
        ("\x00\x00\x01Sce", -1),
        ("\x00\x01\x01Sce", -1),
        ("\x00\x02\x01Sce", -1),
        ("\x00\x06\x01Sce", -1),
        ("\x00\x00\x00UnityPlayer", -1),
    ]
    for haystack, off in module_heur:
        find_modules_with_string(haystack, off)

def find_strings():
    seg_start = seg_end = 0

    while seg_start != BADADDR:
        seg_start = NextSeg(seg_start)
        seg_end = SegEnd(seg_start)

        bytes = GetManyBytes(seg_start, seg_end - seg_start)

        if not bytes:
            continue

        start = 0
        while start < len(bytes):
            end = start
            while end < len(bytes) and ord(bytes[end]) >= 0x20 and ord(bytes[end]) <= 0x7e:
                end += 1
            if end - start > 8 and not isCode(GetFlags(seg_start + start)):
                MakeStr(seg_start + start, BADADDR)
            start = end + 1


def add_xrefs():
    """
        Searches for MOV / MOVT pair, probably separated by few instructions,
        and adds xrefs to things that look like addresses
    """
    addr = 0
    funcCalls = []
    while addr != BADADDR:
        addr = NextHead(addr, 0xFFFFFFFF)
        if GetMnem(addr) == "MOV":
            reg = GetOpnd(addr, 0)
            if GetOpnd(addr, 1)[0] != "#":
                continue
            val = GetOperandValue(addr, 1)
            found = False
            next_addr = addr
            for x in range(16):
                next_addr = NextHead(next_addr, 0xFFFFFFFF)
                if GetMnem(next_addr) in ["B", "BX", "BL", "BLX"]:
                    break
                if GetMnem(next_addr) == "MOVT" and GetOpnd(next_addr, 0) == reg:
                    if GetOpnd(next_addr, 1)[0] == "#":
                        found = True
                        val += GetOperandValue(next_addr, 1) * (2 ** 16)
                    break
                if GetOpnd(next_addr, 0) == reg or GetOpnd(next_addr, 1) == reg:
                    break
            if val & 0xFFFF0000 == 0:
                continue
            if found:
                # pair of MOV/MOVT
                OpOffEx(next_addr, 1, REF_HIGH16, val, 0, 0)
            else:
                # a single MOV instruction
                OpOff(addr, 1, 0)


def remove_chunks(ea):
    """
        Remove chunks from imported functions because they make no sense.
    """
    chunks = list(Chunks(ea))
    if len(chunks) > 1:
        for chunk in chunks:
            if chunk[0] != ea:
                RemoveFchunk(ea, chunk[0])
                MakeFunction(chunk[0])
        Wait()


def resolve_local_nids():
    """
        Finds resolved imported functions and renames them to actual names,
        if the module that provides that function is available and loaded.
        Only works for user-level imports.
    """
    ea = NextFunction(NextAddr(0))
    while ea != BADADDR:
        next = NextHead(ea, 0xFFFFFFFF)
        if GetMnem(ea) == "MOV" and GetMnem(next) == "BX" and GetOpnd(ea, 0) == "R12":
            remove_chunks(ea)
            faddr = GetOperandValue(ea, 1) & 0xFFFFFFFE
            actual_name = GetFunctionName(faddr)
            if actual_name and not actual_name.startswith("sub_"):
                rename_function(ea, actual_name, "_imp")
        ea = NextFunction(ea)


def main():
    path = os.path.dirname(os.path.realpath(__file__))
    load_nids(os.path.join(path, "sharedfb_nids.txt"))

    print("Finding modules")
    find_modules()
    print("Waiting")
    Wait()
    print("Finding strings")
    find_strings()
    print("Adding xrefs")
    add_xrefs()
    resolve_local_nids()


if __name__ == "__main__":
    main()