-
-
[原创]2021 KCTF 春季赛 第六题 寻回宝剑 wp
-
发表于: 2021-5-18 13:20
-
题目是windows64位应用程序.
观察PE文件的Section, 发现有8个节, 最后一个节.cop有可执行权限, 说明应该有加壳或者其他处理.
载入调试器看看EntryPoint代码.
好家伙!往EntryPoint上下翻翻都是清一色的代码块.
简单单步跟踪了一下, 总结混淆方案共有4种.
整个文件大小将近1M了, 可见混淆数量之多.
如果想靠肉眼顶着混淆看代码, 属实不太现实.
看来必须要先写代码对PE文件的混淆做优化, 再来盘程序逻辑了.
针对题目里的4种混淆方案, 需要给出对应的去混淆方法.
不需要优化得很彻底, 只需要能不太费力地使用肉眼看代码逻辑即可.
干到这里留了个心眼.
既然咱们需要patch题目文件优化混淆, 题目会不会有CRC之类的校验代码段, 如果有的话还得先要patch代码校验.
先简单手动修改一处混淆(随便找了个混淆方案1的地方改成jmp指令), 保存文件, 运行.
yes, 运行是没问题, 看来可以放心patch题目文件了.
要patch文件, 首先需要把文件读入到内存中来, Section展开不展开都行.
这里需要注意的是, 题目的ImageBase是默认0x140000000, 申请的地址要满足后面几位也最好都是0, 要不然混淆方案中的xor计算会出错.
混淆方案1等价于jmp指令.
jmp到的地址等于call $+5下一行的地址异或上后面的异或值常量.
这种混淆方案在反汇编代码中随便拉都是, 看来可以先进行一波特征搜索patch掉这些"肉眼可见"的jmp变异.
这里需要注意的是, .text节和.cop节都是代码段, 两个代码段之间会有很多来回的跳转. 没有展开Section的话, 需要注意FA和VA的转换.
混淆方案2等价于jmp指令.
但是混淆方案2的代码在内存中并不是连续的, 不好直接搜索特征来去除.
这里想到个办法, 使用x64dbg的TraceInto/TraceOver功能先跑出代码来, 再针对输出的log文件进行文本分析得到混淆方案2的地址.
x64dbg的TraceInto/TraceOver功能, 要输出到log的内容填上0x{p:cip} {i:cip}, 选择好输出log文件的路径, 设置步数500000(只能多不能少, 至少要跑完输出错误提示吧), 不用设置停止条件, 即可开始trace.
trace出来的文件大概是这样.
一眼望去, 代码里全是"去混淆方案1"中patch的jmp和成对出现的垃圾指令"push reg"/"pop reg".
写个python脚本优化掉(方法见"去混淆方案3").
去掉jmp和push reg/popreg的log文本文件中, 可以见到混淆方案2的代码已经连续出现了.
这里使用notepad++的插件python script写个脚本, 输出针对混淆方案2优化需要用到的几个重要参数(起始地址,callnext返回地址,异或运算常量,加法运算常量).
输出的地址特别多, 整理填入C代码中.
修复混淆方案2的代码(注意FA和RVA的转换):
混淆方案3是垃圾指令, 成对的push reg和pop reg.
写python脚本在log文本中搜索几次, 即可清理干净.
这里内存中就不patch了, 不是很影响看代码.
对python脚本输出的文件可以进行"去混淆方案2"和"去混淆方案4"操作.
混淆方案4等价于call指令.
混淆方案4和混淆方案2的差别, 就是最后一行一个是ret, 一个是jmp.
push addr+ret = jmp addr
push addr1+jmp addr2= call addr2 (addr1是call的返回地址)
在对trace的log进行完"去混淆指令3"和"去混淆指令2"操作后, 混淆方案4的代码就会"浮出水面", 在文本中连续.
继续写python脚本提取输出要处理的信息.
输出的数据填入C代码中, 进行修复.
优化混淆方案4的方法:
1.起始地址jmp到返回地址-5
2.返回地址-5的地址,填入call指令
经过优化代码混淆后, 看一下EntryPoint处代码.
这代码看着就眼熟多了, 一股浓浓的VS201X编译器的味道.
main函数给定位到了:0x0000000140083814
下次trace就可以从main函数开始.
暂时想不到什么很好的办法能把代码全部按顺序修复好让ida能F5, 不过这反汇编代码已经看起来非常舒适了.
又多再看了几眼, 程序编译的时候没有开优化, 反汇编代码看起来又2更亲切了.
多亏笔者在武汉科锐培训三阶段的时候, 练就了看汇编代码还原C代码的"神功", 就当成是培训时间一次正常的作业吧.
^_^
首先是输出字符串到控制台和输入flag.
输入输出使用的std::cout和std::cin.
输入之后首先判断了输入字符串的长度(0x54=84).
接着为一块大小为0x150的局部变量数组初始化数值为0.
观察对改数组的访问得到数组成员size为8字节, 算出总共42个成员, 数组定义如下.
循环往数组内写入数据.
写入数值=取一个字符转数字*0x2A+再取一个字符转数字.
仔细分析字符转数字函数, 得出输入总共有42个字符, 分别是:
0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ+-*/%=
下标就是这个字符转换出来的数值.
接着循环判断数组QWORD ary[42]中以从小到大顺序排序.
这里说明flag偶数下标的数值是限制死的.
只能是"0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ+-*/%="依次排序.
接着一个check, 循环判断了输入的偶数位和奇数位字符都不一样.
这里可以得到信息,奇数位的中每个字符都只出现一次.
接着下一个check, 嵌套的循环.
大概逻辑是, 相邻的偶数位和奇数位为一组, 对于每两组数据差得到的一个坐标都不能相同(这里形容得不太准确,没有太理解,直接把反汇编代码翻译成C代码写个check函数即可).
这里check逻辑虽然看得不是太懂, 但是突然产生个脑洞.
有点像n皇后问题, 使用回溯法写代码可以解决.
但是看这规模, 跑出flag来估计也要个很长时间, 不太现实.
继续看后面有没有check可以缩小穷举范围.
一看果然有发现:
[rsp+0x1010]是输入的flag
[0x0000000140004521]是"02152S3X4Z5Q6C7T819/ADB%C*DL"
也就是说,flag的前面28个字符已经给定.
范围已经缩到最小, 可以写代码进行穷举了.
百度抄下个n皇后回溯法求解的代码, 修改下check函数和输入输出.
运行出结果需要一段时间, 一两分钟吧, 不算太久.
02152S3X4Z5Q6C7T819/ADB%CDLEIFUG3HRIHJ6K7L0MBNKOJPPQ=RNS+TEUOVWWGXYYMZ9+4-8F/-%V=A
题目主打还是代码流, 很合笔者胃口.
题目算法验证部分最后一个循环check中, 有数组越界访问, 这个很坑, 不按题目算法这样写代码得不到flag.
混淆方法太少, 使得攻击者要恢复代码工作量不算太大.
附件文件:
crack_login.cpp是修复混淆和最后回溯穷举的代码.
KCTF_patch8.exe是优化部分混淆过后的题目可执行文件.
python script.rar是notepad++插件的python脚本.
本人目前待业中, 主要擅长windows的逆向, 研究最多的是代码流的防御和攻击手段, 有公司需要人嘛.
有意的大佬可以在CTF群里联系我, 群备注就是论坛名.
push rax push rax pushfq call $+5 pop rax xor rax,D1E15 mov qword ptr ss:[rsp+10],rax
popfq pop rax retpush rax push rax pushfq call $+5 pop rax xor rax,D1E15 mov qword ptr ss:[rsp+10],rax
popfq pop rax retpush raxpush raxpushfqpush raxpush raxpushfqcall $+5 pop raxxor rax, 0xE4EB3
mov qword ptr ss:[rsp+0x10], rax
popfqpop raxpop raxadd rax, 0x8242E
mov qword ptr ss:[rsp+0x10], rax
popfqpop raxretpush raxpush raxpushfqpush raxpush raxpushfqcall $+5 pop raxxor rax, 0xE4EB3
mov qword ptr ss:[rsp+0x10], rax
popfqpop raxpop raxadd rax, 0x8242E
mov qword ptr ss:[rsp+0x10], rax
popfqpop raxretpush rdxpush rdxpop rdxpop rdxpush rcxpush rdxpop rdxpush rbxpop rbxpop rcxpush rdxpush rdxpop rdxpop rdxpush rcxpush rdxpop rdxpush rbxpop rbxpop rcxpush raxpush raxpushfqpush raxpush raxpushfqcall $+5 pop raxxor rax, 0xDC85F
mov qword ptr ss:[rsp+0x10], rax
popfqpop raxpop raxadd rax, 0x51438
mov qword ptr ss:[rsp+0x10], rax
popfqpop raxjmp 0x00000001400979D6
push raxpush raxpushfqpush raxpush raxpushfqcall $+5 pop raxxor rax, 0xDC85F
mov qword ptr ss:[rsp+0x10], rax
popfqpop raxpop raxadd rax, 0x51438
mov qword ptr ss:[rsp+0x10], rax
popfqpop raxjmp 0x00000001400979D6
static uintptr_t g_nFileSize = 0;
static LPBYTE g_lpFileBuffer = 0;
void ReadPEFile(){ HANDLE hFile = NULL;
HANDLE hMapFile = NULL;
LPBYTE lpFileBuffer = NULL;
hFile = CreateFileA("D:\\Projects\\KCTF2021Spring\\CrackNote\\6\\KCTF.exe", // open MYFILE.TXT
GENERIC_READ, // open for reading
FILE_SHARE_READ, // share for reading
NULL, // no security
OPEN_EXISTING, // existing file only
FILE_ATTRIBUTE_NORMAL, // normal file
NULL);
if (hFile == INVALID_HANDLE_VALUE)
{
std::cout << "Could not create file." << std::endl;
goto EXIT_PROC;
}
g_nFileSize = GetFileSize(hFile, NULL);
if (g_nFileSize == INVALID_FILE_SIZE)
{
std::cout << "GetFileSize error." << std::endl;
goto EXIT_PROC;
}
hMapFile = CreateFileMappingA(hFile, NULL, PAGE_READONLY,
0, 0, NULL);
if (hMapFile == NULL)
{
std::cout << "Could not create file-mapping object." << std::endl;
goto EXIT_PROC;
}
lpFileBuffer = (LPBYTE)MapViewOfFile(hMapFile, // Handle to mapping object.
FILE_MAP_READ, // Read/write permission
0, // Max. object size.
0, // Size of hFile.
0); // Map entire file.
if (lpFileBuffer == NULL)
{
std::cout << "MapViewOfFile error." << std::endl;
goto EXIT_PROC;
}
g_lpFileBuffer = new BYTE[g_nFileSize + 0x100000000];
g_lpFileBuffer = (LPBYTE)(((QWORD)g_lpFileBuffer & 0xFFFFFFFF00000000) + 0x100000000);
if (g_lpFileBuffer == NULL)
{
std::cout << "Alloc error." << std::endl;
goto EXIT_PROC;
}
memcpy(g_lpFileBuffer, lpFileBuffer, g_nFileSize);
EXIT_PROC: if (lpFileBuffer != NULL)
{
UnmapViewOfFile(lpFileBuffer);
lpFileBuffer = NULL;
}
if (hMapFile != NULL)
{
CloseHandle(hMapFile);
hMapFile = NULL;
}
if (hFile != INVALID_HANDLE_VALUE)
{
CloseHandle(hFile);
hFile = INVALID_HANDLE_VALUE;
}
}static uintptr_t g_nFileSize = 0;
static LPBYTE g_lpFileBuffer = 0;
void ReadPEFile(){ HANDLE hFile = NULL;
HANDLE hMapFile = NULL;
LPBYTE lpFileBuffer = NULL;
hFile = CreateFileA("D:\\Projects\\KCTF2021Spring\\CrackNote\\6\\KCTF.exe", // open MYFILE.TXT
GENERIC_READ, // open for reading
FILE_SHARE_READ, // share for reading
NULL, // no security
OPEN_EXISTING, // existing file only
FILE_ATTRIBUTE_NORMAL, // normal file
NULL);
if (hFile == INVALID_HANDLE_VALUE)
{
std::cout << "Could not create file." << std::endl;
goto EXIT_PROC;
}
g_nFileSize = GetFileSize(hFile, NULL);
if (g_nFileSize == INVALID_FILE_SIZE)
{
std::cout << "GetFileSize error." << std::endl;
goto EXIT_PROC;
}
hMapFile = CreateFileMappingA(hFile, NULL, PAGE_READONLY,
0, 0, NULL);
if (hMapFile == NULL)
{
std::cout << "Could not create file-mapping object." << std::endl;
goto EXIT_PROC;
}
lpFileBuffer = (LPBYTE)MapViewOfFile(hMapFile, // Handle to mapping object.
FILE_MAP_READ, // Read/write permission
0, // Max. object size.
0, // Size of hFile.
0); // Map entire file.
if (lpFileBuffer == NULL)
{
std::cout << "MapViewOfFile error." << std::endl;
goto EXIT_PROC;
}
g_lpFileBuffer = new BYTE[g_nFileSize + 0x100000000];
g_lpFileBuffer = (LPBYTE)(((QWORD)g_lpFileBuffer & 0xFFFFFFFF00000000) + 0x100000000);
if (g_lpFileBuffer == NULL)
{
std::cout << "Alloc error." << std::endl;
goto EXIT_PROC;
}
memcpy(g_lpFileBuffer, lpFileBuffer, g_nFileSize);
EXIT_PROC: if (lpFileBuffer != NULL)
{
UnmapViewOfFile(lpFileBuffer);
lpFileBuffer = NULL;
}
if (hMapFile != NULL)
{
CloseHandle(hMapFile);
hMapFile = NULL;
}
if (hFile != INVALID_HANDLE_VALUE)
{
CloseHandle(hFile);
hFile = INVALID_HANDLE_VALUE;
}
}void AntiObfuscation1(){ LPBYTE lpStart = 0;
LPBYTE lpEnd = 0;
DWORD dwCount = 0;
//.text节
lpStart = g_lpFileBuffer+0x400;
lpEnd = lpStart + 0x3000;
dwCount = 0;
while (true)
{
//50509CE800000000584835????????48894424109D58C3
LPBYTE lpResult = (LPBYTE)FindHexBytes(lpStart, lpEnd, "50509CE800000000584835????????48894424109D58C3");
if (lpResult == (LPVOID)-1) {
break;
}
else {
dwCount++;
lpStart = lpResult + 1;
QWORD dwFindFA = (QWORD)lpResult;
QWORD dwFindVA = dwFindFA - 0x400 + 0x1000;
QWORD dwCallNextFA = (QWORD)lpResult + 8;
QWORD dwCallNextVA = dwCallNextFA - 0x400 + 0x1000 ;// 修复内存地址
QWORD dwXor = Mem_RDword(lpResult + 11);
QWORD dwNextAddrVA = dwXor ^ dwCallNextVA;
QWORD dwNextAddrFA = 0;
BYTE aryNop[25] = {
0x90, 0x90, 0x90, 0x90, 0x90,
0x90, 0x90, 0x90, 0x90, 0x90,
0x90, 0x90, 0x90, 0x90, 0x90,
0x90, 0x90, 0x90, 0x90, 0x90,
0x90, 0x90, 0x90, 0x90, 0x90,
};
//混淆方案1的opcode size 固定是23
Mem_WBytes2((LPVOID)dwFindFA, aryNop, 23);
BYTE bHook[] = { 0xE9, 00, 00, 00, 00 };
Mem_WDword(bHook + 1, (DWORD)(dwNextAddrVA - dwFindVA - 5));
Mem_WBytes2((LPVOID)dwFindFA, bHook, 5);
}
}
D("patch .text %d addr", dwCount);
//.cop节
lpStart = g_lpFileBuffer + 0x6200;
lpEnd = lpStart + 0xF0000;
dwCount = 0;
while (true)
{
//50509CE800000000584835????????48894424109D58C3
LPBYTE lpResult = (LPBYTE)FindHexBytes(lpStart, lpEnd, "50509CE800000000584835????????48894424109D58C3");
if (lpResult == (LPVOID)-1) {
break;
}
else {
dwCount++;
lpStart = lpResult + 1;
QWORD dwFindFA = (QWORD)lpResult;
QWORD dwFindVA = dwFindFA - 0x6200 + 0xB000;
QWORD dwCallNextFA = (QWORD)lpResult + 8;
QWORD dwCallNextVA = dwCallNextFA - 0x6200 + 0xB000;// 修复内存地址
QWORD dwXor = Mem_RDword(lpResult + 11);
QWORD dwNextAddrVA = dwXor ^ dwCallNextVA;
QWORD dwNextAddrFA = 0;
BYTE aryNop[25] = {
0x90, 0x90, 0x90, 0x90, 0x90,
0x90, 0x90, 0x90, 0x90, 0x90,
0x90, 0x90, 0x90, 0x90, 0x90,
0x90, 0x90, 0x90, 0x90, 0x90,
0x90, 0x90, 0x90, 0x90, 0x90,
};
//混淆方案1的opcode size 固定是23
Mem_WBytes2((LPVOID)dwFindFA, aryNop, 23);
BYTE bHook[] = { 0xE9, 00, 00, 00, 00 };
Mem_WDword(bHook + 1, (DWORD)(dwNextAddrVA - dwFindVA - 5));
Mem_WBytes2((LPVOID)dwFindFA, bHook, 5);
}
}
D("patch .cop %d addr", dwCount);
}void AntiObfuscation1(){ LPBYTE lpStart = 0;
LPBYTE lpEnd = 0;
DWORD dwCount = 0;
//.text节
lpStart = g_lpFileBuffer+0x400;
lpEnd = lpStart + 0x3000;
dwCount = 0;
while (true)
{
//50509CE800000000584835????????48894424109D58C3
LPBYTE lpResult = (LPBYTE)FindHexBytes(lpStart, lpEnd, "50509CE800000000584835????????48894424109D58C3");
if (lpResult == (LPVOID)-1) {
break;
}
else {
dwCount++;
lpStart = lpResult + 1;
QWORD dwFindFA = (QWORD)lpResult;
QWORD dwFindVA = dwFindFA - 0x400 + 0x1000;
QWORD dwCallNextFA = (QWORD)lpResult + 8;
QWORD dwCallNextVA = dwCallNextFA - 0x400 + 0x1000 ;// 修复内存地址
QWORD dwXor = Mem_RDword(lpResult + 11);
QWORD dwNextAddrVA = dwXor ^ dwCallNextVA;
QWORD dwNextAddrFA = 0;
BYTE aryNop[25] = {
0x90, 0x90, 0x90, 0x90, 0x90,
0x90, 0x90, 0x90, 0x90, 0x90,
0x90, 0x90, 0x90, 0x90, 0x90,
0x90, 0x90, 0x90, 0x90, 0x90,
0x90, 0x90, 0x90, 0x90, 0x90,
};
//混淆方案1的opcode size 固定是23
Mem_WBytes2((LPVOID)dwFindFA, aryNop, 23);
BYTE bHook[] = { 0xE9, 00, 00, 00, 00 };
Mem_WDword(bHook + 1, (DWORD)(dwNextAddrVA - dwFindVA - 5));
Mem_WBytes2((LPVOID)dwFindFA, bHook, 5);
}
}
D("patch .text %d addr", dwCount);
//.cop节
lpStart = g_lpFileBuffer + 0x6200;
lpEnd = lpStart + 0xF0000;
dwCount = 0;
while (true)
{
//50509CE800000000584835????????48894424109D58C3
LPBYTE lpResult = (LPBYTE)FindHexBytes(lpStart, lpEnd, "50509CE800000000584835????????48894424109D58C3");
if (lpResult == (LPVOID)-1) {
break;
}
else {
dwCount++;
lpStart = lpResult + 1;
QWORD dwFindFA = (QWORD)lpResult;
QWORD dwFindVA = dwFindFA - 0x6200 + 0xB000;
QWORD dwCallNextFA = (QWORD)lpResult + 8;
QWORD dwCallNextVA = dwCallNextFA - 0x6200 + 0xB000;// 修复内存地址
QWORD dwXor = Mem_RDword(lpResult + 11);
QWORD dwNextAddrVA = dwXor ^ dwCallNextVA;
QWORD dwNextAddrFA = 0;
BYTE aryNop[25] = {
0x90, 0x90, 0x90, 0x90, 0x90,
0x90, 0x90, 0x90, 0x90, 0x90,
0x90, 0x90, 0x90, 0x90, 0x90,
0x90, 0x90, 0x90, 0x90, 0x90,
0x90, 0x90, 0x90, 0x90, 0x90,
};
//混淆方案1的opcode size 固定是23
Mem_WBytes2((LPVOID)dwFindFA, aryNop, 23);
BYTE bHook[] = { 0xE9, 00, 00, 00, 00 };
Mem_WDword(bHook + 1, (DWORD)(dwNextAddrVA - dwFindVA - 5));
Mem_WBytes2((LPVOID)dwFindFA, bHook, 5);
}
}
D("patch .cop %d addr", dwCount);
}a = editor.getText()
lst=a.split("\n")
lstdel=[]
for index in range(len(lst)-18):
#for index in range(20): i1=lst[index][-8:]
i2=lst[index+1][-8:]
i3=lst[index+2][-6:]
i17=lst[index+17][-3:]
#print(i1,i2,i3,i17)
if(i1=="push rax" and i2=="push rax" and i3=="pushfq" and i17=="ret"):
eva=lst[index][:18]
cnva=lst[index+7][:18]
xi=lst[index+8].rfind(" ");
x=lst[index+8][xi:]
ai=lst[index+13].rfind(" ");
a=lst[index+13][ai:]
print(eva, cnva, x, a)
print("finish.")
a = editor.getText()
lst=a.split("\n")
lstdel=[]
for index in range(len(lst)-18):
#for index in range(20): i1=lst[index][-8:]
i2=lst[index+1][-8:]
i3=lst[index+2][-6:]
i17=lst[index+17][-3:]
#print(i1,i2,i3,i17)
if(i1=="push rax" and i2=="push rax" and i3=="pushfq" and i17=="ret"):
eva=lst[index][:18]
cnva=lst[index+7][:18]
xi=lst[index+8].rfind(" ");
x=lst[index+8][xi:]
ai=lst[index+13].rfind(" ");
a=lst[index+13][ai:]
print(eva, cnva, x, a)
print("finish.")
LPCSTR g_aryFix[] = {
"0x00000001400B5E68", "0x000000014007FBD8", "0x6665", "0x36F74",
"0x000000014005AF49", "0x000000014002BB97", "0x28376", "0xCFCD7",
"0x00000001400E05AD", "0x0000000140040B81", "0xD7FA9", "0xFFFFFFFFFFFE178B",
"0x000000014000151A", "0x000000014008F26F", "0xDB257", "0x78C45",
"0x0000000140029C72", "0x000000014008F92C", "0xEB504", "0x328B9",
"0x000000014001E0A6", "0x00000001400F1BA1", "0x6545B", "0x28976",
"0x0000000140052153", "0x0000000140012134", "0x485D5", "0xFFFFFFFFFFFF07DE",
"0x0000000140083D5A", "0x000000014000F6CA", "0x43749", "0x46ACE",
"0x000000014002C47E", "0x00000001400C2EFA", "0xA5AD1", "0x89CB0",
"0x00000001400E7AEC", "0x00000001400370EC", "0xFEA16", "0xFFFFFFFFFFFD4BAC",
"0x00000001400C0C09", "0x00000001400F44A9", "0x2F746", "0xFFFFFFFFFFF323E7",
"0x000000014001533A", "0x0000000140020B70", "0x15CD7", "0x5E0A2",
"0x00000001400572A6", "0x0000000140059016", "0xFBD1", "0xFFFFFFFFFFFDF0F0",
"0x0000000140022F8A", "0x00000001400767A4", "0x4BE4C", "0x88F4",
//... ...
};LPCSTR g_aryFix[] = {
"0x00000001400B5E68", "0x000000014007FBD8", "0x6665", "0x36F74",
"0x000000014005AF49", "0x000000014002BB97", "0x28376", "0xCFCD7",
"0x00000001400E05AD", "0x0000000140040B81", "0xD7FA9", "0xFFFFFFFFFFFE178B",
"0x000000014000151A", "0x000000014008F26F", "0xDB257", "0x78C45",
"0x0000000140029C72", "0x000000014008F92C", "0xEB504", "0x328B9",
"0x000000014001E0A6", "0x00000001400F1BA1", "0x6545B", "0x28976",
"0x0000000140052153", "0x0000000140012134", "0x485D5", "0xFFFFFFFFFFFF07DE",
"0x0000000140083D5A", "0x000000014000F6CA", "0x43749", "0x46ACE",
"0x000000014002C47E", "0x00000001400C2EFA", "0xA5AD1", "0x89CB0",
"0x00000001400E7AEC", "0x00000001400370EC", "0xFEA16", "0xFFFFFFFFFFFD4BAC",
"0x00000001400C0C09", "0x00000001400F44A9", "0x2F746", "0xFFFFFFFFFFF323E7",
"0x000000014001533A", "0x0000000140020B70", "0x15CD7", "0x5E0A2",
"0x00000001400572A6", "0x0000000140059016", "0xFBD1", "0xFFFFFFFFFFFDF0F0",
"0x0000000140022F8A", "0x00000001400767A4", "0x4BE4C", "0x88F4",
//... ...
};void AntiObfuscation2(){ for (int i = 0; i < sizeof(g_aryFix) / sizeof(g_aryFix[0]) / 4 ;i++)
{
QWORD dwFindRVA = strtoull(g_aryFix[i * 4 ], 0, 16) - 0x140000000;
QWORD dwFindFA = 0;
if (dwFindRVA < 0x3400)
{
dwFindFA = (QWORD)g_lpFileBuffer + dwFindRVA - 0x1000 + 0x400;
}
else
{
dwFindFA = (QWORD)g_lpFileBuffer + dwFindRVA - 0xB000 + 0x6200;
}
QWORD dwCallNextRVA= strtoull(g_aryFix[i * 4+1], 0, 16) - 0x140000000;
QWORD dwXor = strtoull(g_aryFix[i * 4 + 2], 0, 16);
QWORD dwAdd = strtoull(g_aryFix[i * 4 + 3], 0, 16);
QWORD dwNextAddrRVA = (dwXor ^ dwCallNextRVA) + dwAdd;
BYTE bHook[] = { 0xE9, 00, 00, 00, 00 };
Mem_WDword(bHook + 1, (DWORD)(dwNextAddrRVA - dwFindRVA - 5));
Mem_WBytes2((LPVOID)dwFindFA, bHook, 5);
}
}void AntiObfuscation2(){ for (int i = 0; i < sizeof(g_aryFix) / sizeof(g_aryFix[0]) / 4 ;i++)
{
QWORD dwFindRVA = strtoull(g_aryFix[i * 4 ], 0, 16) - 0x140000000;
QWORD dwFindFA = 0;
if (dwFindRVA < 0x3400)
{
dwFindFA = (QWORD)g_lpFileBuffer + dwFindRVA - 0x1000 + 0x400;
}
else
{
dwFindFA = (QWORD)g_lpFileBuffer + dwFindRVA - 0xB000 + 0x6200;
}
QWORD dwCallNextRVA= strtoull(g_aryFix[i * 4+1], 0, 16) - 0x140000000;
QWORD dwXor = strtoull(g_aryFix[i * 4 + 2], 0, 16);
QWORD dwAdd = strtoull(g_aryFix[i * 4 + 3], 0, 16);
QWORD dwNextAddrRVA = (dwXor ^ dwCallNextRVA) + dwAdd;
BYTE bHook[] = { 0xE9, 00, 00, 00, 00 };
Mem_WDword(bHook + 1, (DWORD)(dwNextAddrRVA - dwFindRVA - 5));
Mem_WBytes2((LPVOID)dwFindFA, bHook, 5);
}
}a = editor.getText()
lst=a.split("\n")
print(len(lst));
lstdel=[]
for index in range(len(lst)):
if(lst[index][-22:-19] == "jmp"):
lstdel.append(index)
lstdel.reverse()print(len(lstdel))
for index in range(len(lstdel)):
del lst[lstdel[index]]
for fxcktime in range(10):
lstdel=[]
for index in range(len(lst)-1):
i1=lst[index][-8:-4]
i2=lst[index+1][-7:-4]
reg1=lst[index][-3:]
reg2=lst[index+1][-3:]
if(i1=="push" and i2=="pop" and reg1==reg2):
#print(i1,reg1,i2,reg2)
lstdel.append(index)
lstdel.append(index+1)
lstdel.reverse()
for index in range(len(lstdel)):
del lst[lstdel[index]]
fo=open("D:\\Projects\\KCTF2021Spring\\CrackNote\\6\\crack\\crack\\pyoutput.txt","w")
for index in range(len(lst)):
fo.write(lst[index]+"\n")
fo.close()print("finish.");
a = editor.getText()
lst=a.split("\n")
print(len(lst));
lstdel=[]
for index in range(len(lst)):
if(lst[index][-22:-19] == "jmp"):
lstdel.append(index)
lstdel.reverse()print(len(lstdel))
for index in range(len(lstdel)):
del lst[lstdel[index]]
for fxcktime in range(10):
lstdel=[]
for index in range(len(lst)-1):
i1=lst[index][-8:-4]
i2=lst[index+1][-7:-4]
reg1=lst[index][-3:]
reg2=lst[index+1][-3:]
if(i1=="push" and i2=="pop" and reg1==reg2):
#print(i1,reg1,i2,reg2)
lstdel.append(index)
lstdel.append(index+1)
lstdel.reverse()
for index in range(len(lstdel)):
del lst[lstdel[index]]
fo=open("D:\\Projects\\KCTF2021Spring\\CrackNote\\6\\crack\\crack\\pyoutput.txt","w")
for index in range(len(lst)):
fo.write(lst[index]+"\n")
fo.close()print("finish.");
a = editor.getText()
lst=a.split("\n")
lstdel=[]
for index in range(len(lst)-18):
i1=lst[index][-8:]
i2=lst[index+1][-8:]
i3=lst[index+2][-6:]
i17=lst[index+16][-7:]
#print(i1,i2,i3,i17)
if(i1=="push rax" and i2=="push rax" and i3=="pushfq" and i17=="pop rax"):
eva=lst[index][:18]
cnva=lst[index+7][:18]
callva=lst[index+17][:18]
xi=lst[index+8].rfind(" ");
x=lst[index+8][xi:]
ai=lst[index+13].rfind(" ");
a=lst[index+13][ai:]
print(eva, cnva, x, a, callva)
print("finish.")
a = editor.getText()
lst=a.split("\n")
lstdel=[]
for index in range(len(lst)-18):
i1=lst[index][-8:]
i2=lst[index+1][-8:]
i3=lst[index+2][-6:]
i17=lst[index+16][-7:]
#print(i1,i2,i3,i17)
if(i1=="push rax" and i2=="push rax" and i3=="pushfq" and i17=="pop rax"):
eva=lst[index][:18]
cnva=lst[index+7][:18]
callva=lst[index+17][:18]
xi=lst[index+8].rfind(" ");
x=lst[index+8][xi:]
ai=lst[index+13].rfind(" ");
a=lst[index+13][ai:]
print(eva, cnva, x, a, callva)
print("finish.")
LPCSTR g_aryFixC[] = {
"0x000000014007A4C4", "0x0000000140089E51", "0xDC85F", "0x51438", "0x00000001400979D8",
"0x00000001400ADE90", "0x00000001400D4AEB", "0x4E2ED", "0xFFFFFFFFFFFD199D", "0x0000000140060882",
"0x00000001400C4678", "0x0000000140020DE5", "0xD2FE3", "0xFFFFFFFFFFFBE04E", "0x000000014005CC6E",
"0x00000001400C673D", "0x00000001400E7B10", "0xEB97C", "0x6DDA4", "0x000000014003F7D0",
"0x000000014003AF3E", "0x00000001400DDF75", "0x9704D", "0xFFFFFFFFFFFF5B08", "0x000000014003F7D0",
"0x00000001400839E3", "0x000000014007BECA", "0x36372", "0xA777F", "0x00000001400F90F3",
"0x00000001400543F0", "0x00000001400A164B", "0xAF7C5", "0x5CDEC", "0x0000000140078E09",
"0x00000001400998CB", "0x00000001400B928A", "0x2C92F", "0xE4AC", "0x00000001400D031B",
"0x000000014007C0AC", "0x000000014009B414", "0xC2AD8", "0xFFFFFFFFFFFC7598", "0x00000001400D42F1",
"0x000000014008FBE2", "0x0000000140047E12", "0x56408", "0x65F02", "0x00000001400728A0",
"0x000000014006B5D5", "0x00000001400194CA", "0x78E10", "0x3996", "0x0000000140078E09",
"0x0000000140077920", "0x00000001400A2374", "0x4FE21", "0xFFFFFFFFFFFCEA1A", "0x0000000140039068",
"0x00000001400943F5", "0x000000014008595F", "0xF3ADF", "0xFFFFFFFFFFFA3AAE", "0x00000001400DFCB3",
"0x000000014003768E", "0x000000014005087F", "0x45B11", "0xAB445", "0x000000014002650B",
//... ...
};LPCSTR g_aryFixC[] = {
"0x000000014007A4C4", "0x0000000140089E51", "0xDC85F", "0x51438", "0x00000001400979D8",
"0x00000001400ADE90", "0x00000001400D4AEB", "0x4E2ED", "0xFFFFFFFFFFFD199D", "0x0000000140060882",
"0x00000001400C4678", "0x0000000140020DE5", "0xD2FE3", "0xFFFFFFFFFFFBE04E", "0x000000014005CC6E",
"0x00000001400C673D", "0x00000001400E7B10", "0xEB97C", "0x6DDA4", "0x000000014003F7D0",
"0x000000014003AF3E", "0x00000001400DDF75", "0x9704D", "0xFFFFFFFFFFFF5B08", "0x000000014003F7D0",
"0x00000001400839E3", "0x000000014007BECA", "0x36372", "0xA777F", "0x00000001400F90F3",
"0x00000001400543F0", "0x00000001400A164B", "0xAF7C5", "0x5CDEC", "0x0000000140078E09",
"0x00000001400998CB", "0x00000001400B928A", "0x2C92F", "0xE4AC", "0x00000001400D031B",
"0x000000014007C0AC", "0x000000014009B414", "0xC2AD8", "0xFFFFFFFFFFFC7598", "0x00000001400D42F1",
"0x000000014008FBE2", "0x0000000140047E12", "0x56408", "0x65F02", "0x00000001400728A0",
"0x000000014006B5D5", "0x00000001400194CA", "0x78E10", "0x3996", "0x0000000140078E09",
"0x0000000140077920", "0x00000001400A2374", "0x4FE21", "0xFFFFFFFFFFFCEA1A", "0x0000000140039068",
"0x00000001400943F5", "0x000000014008595F", "0xF3ADF", "0xFFFFFFFFFFFA3AAE", "0x00000001400DFCB3",
"0x000000014003768E", "0x000000014005087F", "0x45B11", "0xAB445", "0x000000014002650B",
//... ...
};void AntiObfuscation3(){ for (int i = 0; i < sizeof(g_aryFixC) / sizeof(g_aryFixC[0]) / 5; i++)
{
QWORD dwFindRVA = strtoull(g_aryFixC[i * 5], 0, 16) - 0x140000000;
QWORD dwFindFA = 0;
if (dwFindRVA < 0x3400)
{
dwFindFA = (QWORD)g_lpFileBuffer + dwFindRVA - 0x1000 + 0x400;
}
else
{
dwFindFA = (QWORD)g_lpFileBuffer + dwFindRVA - 0xB000 + 0x6200;
}
QWORD dwCallNextRVA = strtoull(g_aryFixC[i * 5 + 1], 0, 16) - 0x140000000;
QWORD dwXor = strtoull(g_aryFixC[i * 5 + 2], 0, 16);
QWORD dwAdd = strtoull(g_aryFixC[i * 5 + 3], 0, 16);
QWORD dwReturnRVA = (dwXor ^ dwCallNextRVA) + dwAdd;
QWORD dwCallRVA= strtoull(g_aryFixC[i * 5+4], 0, 16) - 0x140000000;
QWORD dwHookRVA = dwReturnRVA - 5 ;
QWORD dwHookFA = 0;
if (dwHookRVA < 0x3400)
{
dwHookFA = (QWORD)g_lpFileBuffer + dwHookRVA - 0x1000 + 0x400;
}
else
{
dwHookFA = (QWORD)g_lpFileBuffer + dwHookRVA - 0xB000 + 0x6200;
}
//BYTE nop[] = { 0x90,0x90,0x90,0x90,0x90 };
//if (memcmp((LPBYTE)dwHookFA, nop, 5) != 0)
//{
// printf("5 nops is not found.\r\n");
//}
BYTE bHook[] = { 0xE9, 00, 00, 00, 00 };
Mem_WDword(bHook + 1, (DWORD)(dwHookRVA - dwFindRVA - 5));
Mem_WBytes2((LPVOID)dwFindFA, bHook, 5);
BYTE bHook2[] = { 0xE8, 00, 00, 00, 00 };
Mem_WDword(bHook2 + 1, (DWORD)(dwCallRVA - dwHookRVA - 5));
Mem_WBytes2((LPVOID)dwHookFA, bHook2, 5);
}
}void AntiObfuscation3(){ for (int i = 0; i < sizeof(g_aryFixC) / sizeof(g_aryFixC[0]) / 5; i++)
{
QWORD dwFindRVA = strtoull(g_aryFixC[i * 5], 0, 16) - 0x140000000;
QWORD dwFindFA = 0;
if (dwFindRVA < 0x3400)
{
dwFindFA = (QWORD)g_lpFileBuffer + dwFindRVA - 0x1000 + 0x400;
}
else
{
dwFindFA = (QWORD)g_lpFileBuffer + dwFindRVA - 0xB000 + 0x6200;
}
QWORD dwCallNextRVA = strtoull(g_aryFixC[i * 5 + 1], 0, 16) - 0x140000000;
QWORD dwXor = strtoull(g_aryFixC[i * 5 + 2], 0, 16);
QWORD dwAdd = strtoull(g_aryFixC[i * 5 + 3], 0, 16);
QWORD dwReturnRVA = (dwXor ^ dwCallNextRVA) + dwAdd;
QWORD dwCallRVA= strtoull(g_aryFixC[i * 5+4], 0, 16) - 0x140000000;
QWORD dwHookRVA = dwReturnRVA - 5 ;
QWORD dwHookFA = 0;
if (dwHookRVA < 0x3400)
{
dwHookFA = (QWORD)g_lpFileBuffer + dwHookRVA - 0x1000 + 0x400;
}
else
{
[招生]科锐逆向工程师培训(2024年11月15日实地，远程教学同时开班, 第51期)
最后于 2021-5-19 16:16
被KuCha128编辑
，原因：
|
|
|---|---|
