In this task we have to reverse file called 'AntiD.exe'. After first examination of this, it looks to be simple PE32 executable, packed with UPX. Unfortunately we can't decompress it using UPX tool, so I started to unpack it manually. First thing to notice is that in PE Optional Header - DllCharasteristics is set to 8140, which means that DLLs in this executable can move around a bit (I'm usually using programs like 'CFF Explorer' or something similar to check this things out). I've changed this header to 8100, what actually terminated this behaviour ;)
To decompress this .exe I personally used x64dbg and Scylla, but the tool doesn't matter at all - it could be any runtime debugger and ImpRec I suppose. What we need to do is stop program execution at Entry Point of AntiD.exe, and run exactly one instruction : pushal - in my case, as you can see on image below (but You can also see this as PUSHAD in OllyDbg, or any other debugger).
After executing PUSHAD, You'll notice that value of ESP register had changed. We have to right click on this value, and click Follow in Dump. Then in the Dump Window of our debugger we need to check first 4 bytes and right-click again to set a Hardware Breakpoint on access. Then we simply run our program and wait for break to occur. We should land exactly in this place:
If we pass through jmp antid.71647 instruction, our executable will be already unpacked in process memory. What we must do now, is to dump this process from memory to another .exe and correct IAT (Import Address Table). Both this things can be easily done by 'Scylla', which is integrated in x64dbg, but You can use OllyDump, then ImpRec as well.
The last thing to change is OEP in the header of our new file. After quick examination in IDA I decided to redirect our new entry to address: 0x1380.
Main function looks exactly like on the image above and its quite easy to reverse. First puts "Figure the key out: " on the console output, then gets input from user and pass that to function which I called 'Check_Key' in IDA. If that fuction will return 1, the string: "Well done! A+! You get a gold star!\n" will be printed to us, else we'll see: "You wrong, why you so wrong at this?\n". But we still don't have a flag. We have to dig deeper - next step is to examine function "Check Key". Here's my version of it's disassembly:
char __cdecl Check_Key(char *Str) { char result; // al@5 char v2; // [sp+0h] [bp-38h]@2 signed int i; // [sp+4h] [bp-34h]@2 char v4; // [sp+8h] [bp-30h]@4 char v5; // [sp+8h] [bp-30h]@6 char v6; // [sp+8h] [bp-30h]@8 char v7; // [sp+8h] [bp-30h]@10 int v8; // [sp+8h] [bp-30h]@12 unsigned __int8 hash_tab[40]; // [sp+Ch] [bp-2Ch]@1 hash_tab[0] = 0x8Cu; hash_tab[1] = 0xF1u; hash_tab[2] = 0x53; hash_tab[3] = 0xA3u; hash_tab[4] = 8; hash_tab[5] = 0xD7u; hash_tab[6] = 0xDCu; hash_tab[7] = 0x48; hash_tab[8] = 0xDBu; hash_tab[9] = 0xC; hash_tab[10] = 0x3A; hash_tab[11] = 0xEEu; hash_tab[12] = 0x15; hash_tab[13] = 0x22; hash_tab[14] = 0xC4u; hash_tab[15] = 0xE5u; hash_tab[16] = 0xC9u; hash_tab[17] = 0xA0u; hash_tab[18] = 0xA5u; hash_tab[19] = 0xC; hash_tab[20] = 0xD3u; hash_tab[21] = 0xDCu; hash_tab[22] = 0x51; hash_tab[23] = 0xC7u; hash_tab[24] = 0x39; hash_tab[25] = 0xFDu; hash_tab[26] = 0xD0u; hash_tab[27] = 0xF8u; hash_tab[28] = 0x3B; hash_tab[29] = 0xE8u; hash_tab[30] = 0xCCu; hash_tab[31] = 3; hash_tab[32] = 6; hash_tab[33] = 67; hash_tab[34] = 0xF7u; hash_tab[35] = 0xDAu; hash_tab[36] = 0x7E; hash_tab[37] = 0x65; hash_tab[38] = 0xAEu; hash_tab[39] = 0x80u; if ( strlen(Str) == 16 ) { v2 = 0; for ( i = 0; i < 40; ++i ) { v4 = Str[i] ^ 0x33; if ( check_1() ) return 0; v5 = v4 + 0x44; if ( check_2() ) return 0; v6 = v5 ^ 0x55; if ( check_3() ) return 0; v7 = v6 - 0x66; if ( check_4() ) return 0; v8 = (unsigned __int8)(v7 ^ v2); if ( v8 != hash_tab[i] ) return 0; v2 += v8; } result = 1; } else { result = 0; } return result; }
From now we can see exactly what is going on here. Our input has to have 16 characters to move on. If it is indeed, then our flag should be produced in a loop. I didn't care about all that checking functions, beacuse I decided to write my own decoder than debugging the flag on runtime. Here's the code to do the job:
#include#include using namespace std; int main(void) { unsigned char v2 = 0; unsigned char tab[] = { 0x8C,0xF1,0x53,0xA3,0x08,0xd7,0xDC,0x48,0xDB,0x0C, 0x3A,0xEE,0x15,0x22,0xC4,0xE5,0xC9,0xA0,0xA5,0x0C, 0xD3,0xDC,0x51,0xC7,0x39,0xFD,0xD0,0xF8,0x3B,0xE8, 0xCC,0x03,0x06,0x43,0xF7,0xDA,0x7E,0x65,0xAE,0x80 }; for (char j = 0; j < 40; j++) { x = 0; for (char i = 0; i < 0xff; i++) { x = i ^ 0x33; x = x + 0x44; x = x ^ 0x55; x = x - 0x66; x = x ^ v2; if (x == tab[j]) { printf("%c", i); break; } } v2 = x + v2; } cout << endl; return 0; }
Comments
Post a Comment