Cracking 2/Cracking II..iso/Texty/crackme/CRACKME2.ENG

Text File | 1999-01-28 | 17.0 KB | 499 lines

           The complete solution of UCF crackme2

            (c) 1999 Ak Kort [SOS group]


Step 1. Decrypting.
~~~~~~~~~~~~~~~~~~~

    Decrypting procedure uses 256-byte key, which answers on
    interrupt 3F and looks like simple dongle.

    I spent much time on analisys decrypting procedure in a
    tries of finding holes in crypting or the way of fast
    reversing. In a real that thing is practically unneeded.
    Also there is no needs to search key recursively as
    i did once in a sober foolish head...

    Look in a following fragment of cipher:

 000000B0: D1 00 6D D9│6D 6D 6D 6D│2E 3F B8 DA│17 B4 50 09  ╤.m┘mmmm.?╕┌.┤P.

    There are 4 identical bytes. Doesn't looks like coincidence.
    Look at the decrypting formula:

    Text[i+2]=Crypt[i+2] xor Key[Text[i]]

    In other words, odd and even bytes are crypted independently and
    decrypting mask for the current byte is the byte from table
    indexed by previous decoded byte.

    Well, what is the meaning of that 4 bytes? It means, that there
    are 6 identical bytes. First two bytes are crypted by previous
    bytes and we cannot resolve crypting mask. But we can get
    next two bytes if we gues several logical cases. 6 identical
    bytes can be zero, spaces and border symbols like '═' or so.
    Other cases seem to be mostly unreal.

    Lets assume that they are zeros. Then the mask of decrypting
    followed bytes is 6D 6D. Check it:

    2E 3F xor 6D 6D = 43 52 = 'CR'

    Looks like true. it can be CRACKER, CRACKED, CRACKME and so on.

    But if they are spaces then mask will be 6D 6D xor 20 20 = 4D 4D,
    and crypted text will be:

    2E 3F xor 4D 4D = 63 72 = 'cr'

    Matches too. What of that cases we can select? Lets check another
    fragment:

 00000110: 55 A2 A6 8C│AB 58 6D 6D│FB 5D 4E 94│83 05 E9 33  Uóªî½Xmm√]Nöâ.Θ3

    There are again two identical bytes 6D 6D. Lets decrypt the followed
    bytes:

    FB 5D xor 6D 6D = 96 30

    At this time every real cracker should spring up from chair,
    clear his eyes, convulsively swallow a beer and scratch his head.
    Cose every real cracker knows, that bytes 00 00 96 30 seems to be
    in a CRC32 calculation table.

    Check our idea - programm STEP-1\DECODE.BAS
    First it builds 256-byte table using file CRC32.DAT as a plain text.
    Then it decodes whole file into OK.COM. Let me skip other comments,
    i think this programm whould be clean for every schoolchild, who
    finished 3 monthes of programming course.

    Fluently look on OK.COM and satisfy youself that we had a right idea.


Step 2. Searching of keyfile.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

    After decrypting programm needs a keyfile, wich contains
    username. Before decoding this name programm check checksum
    of file by the following way:

     0000002B: B90001                    mov cx,0100
     0000002E: BE4906                    mov si,0649
     00000031: E8E004                    call 00000514
     00000034: 66A3B201                  mov [01B2],eax

     There is crc32 calculation using that table wich help us
     to dectypt code.

    Next, there are 0FD iterations at the same text, wich use
    crc32 as a starting value for cipher:

     00000038: B9FD00                    mov cx,00FD
     0000003B: BE4906                    mov si,0649
     0000003E: 668B1EB201                mov ebx,dword ptr [01B2]
     00000043: 51                        push cx
     00000044: AC                        lodsb al,ds:[si]
     00000045: 0FBEC8                    movsx cx,al
     00000048: AD                        lodsw ax,ds:[si]
     00000049: F7E1                      mul cx
     0000004B: 8A0C                      mov cl,byte ptr [si]
     0000004D: 66D3C0                    rol eax,cl
     00000050: 6633D8                    xor ebx,eax
     00000053: 51                        push cx
     00000054: 0FBEC8                    movsx cx,al
     00000057: 66D3C3                    rol ebx,cl
     0000005A: 59                        pop cx
     0000005B: 66F7D0                    not eax
     0000005E: E2ED                      loop 0000004D
     00000060: 59                        pop cx
     00000061: 83EE02                    sub si,0002
     00000064: E2DD                      loop 00000043
     00000066: 6681FB4EC29EA6            cmp ebx,A69EC24E

    Comparing ebx and A69EC24E is a chechsum of keyfile. Looking
    on programm under debugger we can see that this fragment a
    quite slow and brute force on whole 256-byte file is an extremelly
    long task.

    But we can see, that if we fill keyfile with zeros and geting crc32
    also as 0, then on each iteration ebx is equal to 0.
    It is very usefull for search - it is enought to change one
    dword as a counter and fix last dword to fit crc32 to 0.
    Then we will don't needed to perform all 0FD iterations
    but only last 8 for changed bytes.

    But that algorithm will search 256 variants in a 3 seconds or
    about 15 minutes for 65536. It is easy to calucate that it will
    take more than year to search 2^32 variants. But if we use
    several hundreds computers this time will a quite short.
    That way of attack was performed in a last days of Dec 1998.
    Unfortunatly, the way of distributing packets for search
    was'n automatical and i have to send them by hands, that quite
    hamper search.

    So, under unstopped critic of Mikhail Kondakov I perform
    second way to reverse the sum. As I notice sometimes, my
    drunk condition intensifies my cracking skills so I can
    crack very hard things. So one day I've returned to home
    and start the assembler...


Step 3. How to find a keyfile rather fast.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

    We can notice, that the time of sum's calculation is proportional
    to bytes in keyfile. So, to minimize this time we need to
    minimize the value of bytes in files. First we should not use
    crc32 fittingm this free us from big interval of values but
    we'll have to calculate crc32 on every step and use this
    new value from begining of iterations.

    But we can change only last 8 bytes of file and leave 248 bytes
    unchanged, it allow us to precaluclate crc32 for the first 248
    bytes and calculate only the rest of file on each iteration.

    Second, look at this optimized part of cipher:

    lea si, Key
    mov ebx, crc32
    mov eax, crc32
    REPT 0FDh
        movsx cl,[si]
        mov ax,[si+1]
        inc si
        mul cx
        mov cl,[si+2]
    @@1:    rol eax,cl
        xor ebx,eax
        rol ebx,al    <- there is no such command,
                   it is just for undertanding
        not eax
        loop @@1
    ENDM

    Now it is more understandable. This algorithm needs only 4 bytes
    from keyfile and 2 (!) high bytes from eax. After all iterations
    this cipher may be represented as:

        rol ebx, X
        xor ebx, Y

    It is quite simple to check, that X and Y depend only on high
    word of eax and keyfile but does not depend of ebx.
    So if we use static part of key we can precalulate all 65536
    cases of eax and change this very slow algorithm by the
    table. This speed up our search more than 200 times.
    X takes 1 byte, Y takes 4 bytes, so we need 5*65536=327680 bytes,
    wich fits on DOS memory.

    Third improovement is that we don't search bytes in a 0-255
    interval but in 1-16 so we getting rid of corner very slow cases.

    Using this three point we write CRACK.ASM, wich find the key
    in a several hours.
    Also there is 32-bit DOS4GW version wich about 20% faster but
    needs several files from Watcom C for compiling.


Step 4. Resident dongle emulator.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

    Having decoding table it is easy to write emulator.
    I have two version of programm wich are different only
    by the number of interrupt - 3F and FD, so our emulator
    will answer on both of them.

Step 5. Writting keygenerator.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

    After we find the forst 256 bytes of keyfile it is very
    easy to encrypt username by the following algorithm:

     00000066: 6681FB4EC29EA6            cmp ebx,A69EC24E
     0000006D: 752F                      jnz 0000009E
     00000071: BE4906                    mov si,0649
     00000074: 81C60001                  add si,0100
     00000078: 56                        push si
     00000079: 1E                        push ds
     0000007A: 07                        pop es
     0000007B: 8BFE                      mov di,si
     0000007D: B90800                    mov cx,0008
     00000080: 66AD                      lodsd eax,ds:[si]
     00000082: 6633C3                    xor eax,ebx
     00000085: 66D1C3                    rol ebx,1
     00000088: 66AB                      stosd es:[di],eax
     0000008A: E2F4                      loop 00000080

    Very easy and takes about 3 minutes. Source is in STEP-5 directory,
    also there is the valid keyfile.




Thanks and greetings.
~~~~~~~~~~~~~~~~~~~~~

    Mikhail Kondakov ([e-mail]) found another fast
    way for searching keyfile. I was too lazy to resolve algorithm
    of search by the right file and cann't figure wich case is
    better, but I think you should look at his keyfile in
    a STEP-X directory.

    Also I wish thank the following people for help me with a
    brute force by the first variant and give me some good
    ideas and help:

    A.D. Smolenkov, [e-mail]
    ATS, 2:5049/69
    Alecs Kuznetsov (Dark1st), [e-mail]
    Alek Akhmatov, 2:5005/28.3
    Aleksey A Pestunovich, 2:5015/111
    Aleksey Larioncev, 2:5020/1531.5
    Aleksey Timofeyev, 2:4600/96.2
    Alex Devyatykh, 2:5004/530.3
    Alex Johnson, 2:5030/181.141
    Alex Kreshchuk, 2:5063/10.58
    Alex Kuriatnikov, 2:5020/617.3
    Alex Ostapenko, 2:5020/685.21
    Alex Rozov, 2:5011/60
    Alex Shupikov, 2:5040/61.13
    Alex Stepanov, 2:5025/2.122
    Alex V.Gribov, 2:5011/45.13
    Alexander A.Kuznetsov, 2:5054/28.12
    Alexander Hokhlov, 2:5025/38.201
    Alexander Kobrin, 2:5020/620.13, [e-mail]
    Alexander Kozlov, 2:5093/11
    Alexander Moukhortov, [e-mail]
    Alexander Novikov, 2:450/9.25
    Alexander Onishchenko, 2:463/586.34
    Alexander S.Bulochnikov, [e-mail]
    Alexander Smok, 2:452/999.2
    Alexander Soutchkov, 2:5011/66.16
    Alexander Sukhochov, 2:5030/638.15
    Alexander Volok, 2:464/130.14, [e-mail]
    Alexander Wischnewski, [e-mail]
    Alexander Zolotnikov, 2:453/7.8
    Alexandr Demidneko aka CodeGlider, 2:5077/22.10, [e-mail]
    Alexandr Orlenko (Eagle), 2:5079/30.37, [e-mail]
    Alexandr Pikunow, 2:5022/27.40
    Alexey Andreev, 2:5030/142.65
    Alexey Barylnikov aka 1exa, 2:5000/72.42
    Alexey Chukanov, 2:5070/136.15
    Alexey Dobriaev, 2:5020/1822.75
    Alexey Kulinich, 2:5005/26.48
    Alexey Saenko, 2:5011/50.11
    Alexey Uzikov, 2:5011/66.9
    Alexx Valikov, 2:5011/34
    Alvils Pierhurovichs, 2:51/10
    Alvils, [e-mail]
    Anatoly Shishkoff AKA Shturman, 2:5020/390.12
    Ander Vill, 2:5080/500.333
    Andrew Lebedenko, 2:5061/4
    Andrew Petrov, 2:5020/993.139
    Andrew Shipinsky, [e-mail]
    Andrey Baramzin, 2:5056/40.12
    Andrey Bublikov, 2:5030/752.24
    Andrey Grishin, 2:5020/989.19
    Andrey Ivankov, [e-mail]
    Andrey Nikitin, 2:4642/19.33
    Andrey Novikoff, 2:5030/743.31
    Andrey Smolenkov, [e-mail]
    Andriey Korzhun, 2:5020/1662
    Andy Malyshev, 2:5020/362.55, [e-mail], http://i.am/crk
    Andy Sadovnik, 2:5030/610
    Anthony Repin, 2:5020/753.6
    Anton A. Arapov, 2:5045/27.39
    Anton Fedorov, 2:5000/111.27
    Anton Mineev, 2:5011/53
    Anton Morozov, 2:5020/1355.22
    Anton Shmigel, 2:5050/15.22
    Arkadiy Kapustinskiy (kab), 2:5000/33
    Armen Stepanian, 2:5022/83
    Art Snaker
    Artem, 2:5030/525.62
    Artur Hairullin, 2:5011/45.15
    Boris A. Lozovsky, [e-mail]
    Boris Tcherkashin, 2:5080/65.13
    Bulat Kutov, 2:5011/41
    Constantine Vetlov, 2:5020/995
    DAO, [e-mail]
    DK, [e-mail]
    DaRk dEAMon, 2:463/9797.57
    Denis Andrianov, 2:5010/120
    Denis Kozlov, 2:5012/15.26
    Denis Lebedev, 2:450/83.4
    Denis Litvinov, 2:5080/60.16
    Denis Mingulov, 2:5004/35.16
    Denis Ozerov, 2:5030/649.34
    Denis S. Kotov aka Deniska, [e-mail]
    Denis Zwerew, 2:5030/521
    Dmitriy Polyakov, 2:463/308.22
    Dmitry Bragin, 2:5011/79
    Dmitry Deniskin, 2:5020/1294.12
    Dmitry Dudarev, [e-mail]
    Dmitry Nikolaev, 2:5011/22.8
    Dmitry Novak, 2:5020/1432.445
    Dmitry Ostroushko, 2:5036/9.29
    Dmitry Sapojnikov, 2:5066/13.9
    Dmitry Shilin, 2:5020/150.23
    Dmitry Shumsky, 2:5023/11.54
    Dmitry Sokolov, 2:5077/25
    Edward Zinatullin, 2:5011/59.26
    Eliseev Victor A., 2:5020/718.3
    Eugene Konovalov, 2:5055/30.30
    Eugene Morozov, 2:5020/1437.7
    Eugene Nikolaev, [e-mail]
    Eugene Radyuk, 2:5077/24
    Eugene Shepelev, 2:5020/294.32
    Eugene Yurchik, 2:5010/126.37
    Eugeny Starikov, 2:464/500
    Evgeniy Skidan, 2:4521/2.18
    Evgeny Malyshev, 2:5011/13.14
    Expert of Nothing, http://www.chat.ru/~stking
    Gevorck Madzharyan, 2:5057/43
    Grigory Berezowsky, [e-mail]
    HaCkMAN AKA Vladimir Frolov, 2:5078/15.7
    Igor Derevenskyh, 2:5080/97.26
    Igor Kovalenko, 2:463/563
    Igor Latyshev, 2:5045/34.20
    Ivan Elin, 2:5011/89
    Ivan Redkin, 2:5011/78
    Jaffar, 2:4666/13
    Jura Marichev, 2:5011/62.19
    KPNC, [e-mail]
    Kirill Silaev, [e-mail]
    Konstantin Isakov, 2:5000/111.29, [e-mail], www.nsk.su/~bcompany
    Konstantin Norvatoff, 2:5030/163.44
    Kosh, 2:5020/1286.9
    Max Dukov, 2:5030/413.15
    Max Litvinenko, 2:5011/85
    Maxim Kovtun, 2:5020/1175.23
    Maxim Stafeev, 2:5080/157.7
    Maxim Zykov, [e-mail]
    Mega G. Baitoff, 2:5011/60.16
    MegaBit, [e-mail]
    Mickel Glumov, 2:5061/15.46
    Mihail Filichkin, 2:4631/14.1
    Mihail Karasev, 2:5095/34
    Mihail Voidenko, 2:5022/19.56
    Mikhail Vlasov, 2:5011/22.18
    Misha Myasnikov, 2:450/137.21
    Murad Urmancheev, 2:5085/1.101
    Nicholas Mednov, 2:5011/68
    Nick Klochkov, 2:463/253.222
    Nicky Shubeykin, 2:469/43.33
    Nikita Shvetsov, 2:5096/11
    Nikita Zolotarev, 2:5030/789.666
    Nikolai Lobezki, 2:5030/450.23
    Nikolay Mikolenko, 2:4642/1111
    OFF (Andrew Velder), 2:5010/146.111, [e-mail]
    OffSpEEd, [e-mail]
    Oleg Akulov, 2:5002/45.36
    Oleg Bashlykov (Olejonok), [e-mail]
    Oleg Boshkov, 2:467/73
    Oleg Gatcenko, 2:464/207.9
    Oleg Krivoshatkin, 2:5067/3.1
    Oleg Loa, [e-mail]
    Oleg Pashko, 2:4657/9
    Oleg Sharipov, 2:5011/42.33
    Ostroushko Dmitry, [e-mail]
    Pasha Soumartchenkov, 2:5057/36.7
    Pavel Chernov, 2:5057/10.30
    Pavel Dovgalenko, The CHEMIST, 2:5071/17.9, [e-mail]
    Pavel Piganov, 2:5057/25.18
    Pavel Shklovsky, 2:5011/18
    Peter Evdokimov, 2:5011/250, [e-mail], www.chat.ru/~peterb
    Ramil Garipov, 2:5011/27.11, 2:5011/36.106, [e-mail]
    Roman Khomchenko, 2:5040/57.27
    Ruslan Sokolnikov, [e-mail]
    R├\inMaN, [e-mail], http://fido.omsk.edu/~rainman
    SWW aka Slava Rusakov, 2:5030/175.76
    Saboteur, [e-mail]
    Sasha Mkrt, [e-mail]
    Serg Agarkoff, 2:5041/1
    Serg Sidorov, 2:4655/3.66
    Serge Kontarev, 2:5020/53.22
    Serge Lazo, [e-mail]
    Sergei Seperovitch, 2:5020/1133
    Sergey E. Tikhonenko, [e-mail]
    Sergey Guculiak, 2:5011/51
    Sergey Isakov, 2:5080/105.7
    Sergey Korenivskiy, 2:5020/239.9
    Sergey Korowkin, 2:5033/27
    Sergey M Litvinov, 2:4600/126.33
    Sergey Potashev, 2:5003/70, [e-mail]
    Sergey Safarov, 2:451/2.22
    Sergey Shliakhota, [e-mail]
    Serzh Shumikhin, 2:5004/18.8
    Seva Goncharov, 2:5080/31.20
    Shaman //UCL
    Shurik Mudrak, 2:5011/45.14
    Sirosh Evgeni, 2:5036/5.2
    Slava Vvedensky, 2:5004/32.24
    Sonic Mike, 2:50/520.4
    Sp0Raw //[rAN] //[dTG], 2:5030/664.409, 345:817/2, [e-mail]
    Sp0t //UCL
    Spetznaz, [e-mail]
    Stas Schekin, 2:5025/150.49
    System brain, 2:5011/87.2, [e-mail]
    Tim Komyakov, 2:5080/78.2
    Timur Kulmukhametov /Drago73, 2:5011/73
    Vadim Gromov, 2:5030/614.111
    Vadim Samoilov, 2:5003/34.15
    Valentin Varenik, 2:4642/3.31
    Valera Gavrilovets, 2:5030/389.7
    Valeri Bourak, 2:450/107
    Valery Shabaev, 2:5077/60
    Valery V. Bastrikov, 2:5011/27.18
    Victor Zamoulko, 2:4521/4
    Vilen Kamalov, 2:5085/44.7
    Vinkov Alexander, 2:5053/20.30
    Vitaly V Bochkaryov, 2:5090/28.111
    Vizit0r //UCL
    Vlad Bikkulov, 2:5011/36.105
    Vlad Mogilevsky, 2:5020/118.75
    Vladimir Eremin, 2:5020/1626.78
    Vladimir Kozlov, 2:5002/45.11
    Vladimir Sadvinsky, 2:5049/36.24
    Vladimir Timonin, 2:5023/11.44
    Vladimir Zaytcev, 2:5015/116
    Vladislav Bulochnikov, 2:5037/12.39
    Vladislav Zlobin, 2:5011/79.33
    Vsevolod Sokovikov, 2:5004/33.12
    Vyacheslav V Garonin, 2:465/167
    Wowik Mikolenko, 2:5031/1.23
    Xak, [e-mail], xak.ipc.ru
    Yaroslav Pecheritsa, 2:4635/34.19
    Yaroslav Sverdlikov, 2:450/84.54
    Yura Kropelnytskiy, 2:4623/55
    Yura Tkachenko, 2:5030/655.31, [e-mail]
    Yura V. Zaburdaev, 2:450/134.8
    Yuriy Limakov, 2:5062/1.15
    Yuriy Macintosh, 2:5020/1469.777
    Zdanevich Alexander, 2:5020/755.35
    Zerg, 2:5080/130.11
    alex ivanov, 2:5030/773
    andrew karimov (the doctor), 2:4641/223.3
    andyst, [e-mail]
    [e-mail]
    blshkv
    geront, [e-mail]
    [e-mail]
    hijaq, 2:5030/460.2
    joker, 2:4666/10.21
    [e-mail]
    max a sazonoff, 2:5004/32.17
    [e-mail]
    [e-mail]

    And many others who wish to stay anonymous ;)

Copyright
~~~~~~~~~
    (c) 25 Jan 1999 Ak Kort [SOS group]
    http://www.sos.nanko.ru
    e-mail: [e-mail]
    FiDo: 2:5011/45