Datasets:
intent
stringlengths 4
313
| snippet
stringlengths 2
271
|
---|---|
divide eax by ebx | div ebx |
move address of request variable into ecx | mov ecx, request |
decrement the eax register by one | dec eax |
declare the exit_on_error label | exit_on_error: |
push esi onto the stack | push esi |
move the 4 bytes of data at address esi+4*ebx into edx | mov edx, [esi+4*ebx] |
declare msg string containing 'hello world!' | msg db 'hello, world!', 0xa |
push 0xa to the stack | push 0xa |
restore eax from the value we pushed onto the stack at the start | pop eax |
push the double word 0x313a303a onto the stack | push dword 0x313a303a |
move the remainder into eax | mov eax, edx |
create a new process | mov eax, 2 |
move readbuffer into ecx | mov ecx, readbuffer |
syscall 63 | mov al, 0x3f |
move cl into the address [esp + 1] | mov [esp + 1], cl |
push the dword 0x6b2f7379 onto the stack | push dword 0x6b2f7379 |
push the dword 0x7665642f onto the stack | push dword 0x7665642f |
define shellcode label | shellcode: |
decrement the ecx register and jump to the L1 label if the contents of the ecx register is not zero and the zero flag is zero else jump to the shellcode label | loopnz L1 \n jmp shellcode |
define download function | download: |
define the numeric label 2 | 2: |
move exit_call into al | mov al, exit_call |
declare a byte with no label containing the value 10 | db 10 |
if zero jump to the child label | jz child |
push the contents of the edx register onto the stack and point ebx to the stack register | push edx \n mov ebx, esp |
set the contents of edx to zero | xor edx, edx |
move decimal 43 into edx | mov edx, 43 |
push zero_reg onto the stack | push zero_reg |
restore the original value of eax | pop eax |
push byte 6 onto the stack | push byte 6 |
allocate memory for 30 bytes uninitialized name variable | name db 30 dup |
define egg equal to 'egg ' | egg equ 'egg ' |
decrement ebp | dec ebp |
put the syscall 0x04 into the eax register | push byte 0x04 \n pop eax |
declare the priv_setgid label | priv_setgid: |
pop the value on the stack back into ecx | pop ecx |
jump to the shell label | jmp shell |
push the word 0x0a to the stack | push word 0x0a |
declare the here label | here: |
move 5 into eax | mov eax, 0x5 |
getppid | mov eax, 64 |
move decimal number 255 into edx | mov edx, 255 |
if the contents of the bl register is equal to the contents of the al register then jump to the loop_2 label else jump short to the not_found label | cmp bl, al \n je loop_2 \n jmp short not_found |
call the two function | call two |
move al into ebx+7 | mov [ebx+7], al |
zero out zero_reg | xor zero_reg, zero_reg |
move 0x3f into al | mov al, 0x3f |
define constant total_students equal to 50 | total_students equ 50 |
move the contents of the esi register into the dword at the memory location [esp-4] | mov dword [esp-4], esi |
clear ebx register | xor ebx, ebx |
push double word 0x39396e2d onto the stack and point the edi register to the stack register | push dword 0x39396e2d \n mov edi, esp |
declare code section .bss | section .bss |
move 0x1e into cl | mov cl, 0x1e |
add one to the doubleword integer stored at memory location value | inc dword [value] |
push 0x0a206873 onto the stack | push 0x0a206873 |
push the value 0x69622f2f onto the stack and point the ecx register to the stack register | push 0x69622f2f \n mov ecx, esp |
zero out ax register | xor ax, ax |
push the byte 0xb onto the stack | push byte 0xb |
push the byte 0x0 onto the stack | push byte 0x0 |
add 4 to the contents of the esi register | add esi, 4 |
move 0xf into al | mov al, 0xf |
put the byte my_value into the eax register | push byte my_value \n pop eax |
move value at top of the stack to ebx | mov ebx, [esp] |
move bx to count | mov count, bx |
declare the incaddr label | incaddr: |
define the _exit label | _exit: |
zero out the eax register and jump to the l3 label if the contents of the edx register is not equal to the contents of the eax register | xor eax, eax \n cmp edx, eax \n jne l3 |
push the esp onto the stack | push esp |
push 0x16 onto the stack | push 0x16 |
move 0x21 into al | mov al, 0x21 |
push the 0x3170762d onto the stack | push 0x3170762d |
call the sprint function | call sprint |
push the value 0x2f2f2f2f onto the stack and point eax to the stack register | push 0x2f2f2f2f \n mov eax, esp |
jump to the IncAddr label if the value in the eax register is not equal to the doubleword addressed by edi else jump to the edi register | scasd \n jnz IncAddr \n jmp edi |
left shift eax by two bits | shl eax,2 |
compare the byte in esi with 0x7 | cmp byte [esi], 0x7 |
move the byte at the address 3h into al | mov al, byte 3h |
add 2 to the contents of the ebx register | add ebx, 2 |
push the contents of the esi register onto the stack. | push esi |
move 0x06 into al | mov al, 0x06 |
define the _setsockopt label | _setsockopt: |
move 0x30 into al | mov al, 0x30 |
make the socketcall | mov al, 0x66 |
push the 0x6e onto the stack | push 0x6e |
push the word 0x02 onto the stack | push word 0x02 |
load the effective address of ebx+8 into ecx | lea ecx, [ebx+8] |
jump to the wrap_around label if the byte starting at the address contained in the esi register is lower than the byte value 0xD else subtract the byte value 0xD from byte starting at the address contained in the esi register | cmp byte [esi], 0xD \n jl wrap_around \n sub byte [esi], 0xD |
declare the done label | done: |
push 0x5 onto the stack | push 0x5 |
declare retry label | retry: |
declare buffersize to be a word containing 1024 | buffersize: dw 1024 |
push the word 0x3930 onto the stack | push word 0x3930 |
push 0x682f2f2f onto the stack | push 0x682f2f2f |
exchange ebx with ecx | xchg ebx, ecx |
call the strlen function | call strlen |
move the doubleword 0x65676760 into edx | mov edx, dword 0x65676760 |
clear ebx | xor ebx, ebx |
section data | section .data |
push the contents of the edi register onto the stack | push edi |
put the syscall 0x66 into the eax register | push byte 0x66 \n pop eax |
Shellcode_IA32
Shellcode_IA32 is a dataset containing 20 years of shellcodes from a variety of sources is the largest collection of shellcodes in assembly available to date.
This dataset consists of 3,200 examples of instructions in assembly language for IA-32 (the 32-bit version of the x86 Intel Architecture) from publicly available security exploits. We collected assembly programs used to generate shellcode from exploit-db and from shell-storm. We enriched the dataset by adding examples of assembly programs for the IA-32 architecture from popular tutorials and books. This allowed us to understand how different authors and assembly experts comment and, thus, how to deal with the ambiguity of natural language in this specific context. Our dataset consists of 10% of instructions collected from books and guidelines, and the rest from real shellcodes.
Our focus is on Linux, the most common OS for security-critical network services. Accordingly, we added assembly instructions written with Netwide Assembler (NASM) for Linux.
Each line of Shellcode_IA32 dataset represents a snippet - intent pair. The snippet is a line or a combination of multiple lines of assembly code, built by following the NASM syntax. The intent is a comment in the English language.
Further statistics on the dataset and a set of preliminary experiments performed with a neural machine translation (NMT) model are described in the following paper: Shellcode_IA32: A Dataset for Automatic Shellcode Generation.
Note: This work was done in collaboration with the DESSERT Lab.
The dataset is also hosted on the DESSERT Lab Github.
Please consider citing our work:
@inproceedings{liguori-etal-2021-shellcode,
title = "{S}hellcode{\_}{IA}32: A Dataset for Automatic Shellcode Generation",
author = "Liguori, Pietro and
Al-Hossami, Erfan and
Cotroneo, Domenico and
Natella, Roberto and
Cukic, Bojan and
Shaikh, Samira",
booktitle = "Proceedings of the 1st Workshop on Natural Language Processing for Programming (NLP4Prog 2021)",
month = aug,
year = "2021",
address = "Online",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2021.nlp4prog-1.7",
doi = "10.18653/v1/2021.nlp4prog-1.7",
pages = "58--64",
abstract = "We take the first step to address the task of automatically generating shellcodes, i.e., small pieces of code used as a payload in the exploitation of a software vulnerability, starting from natural language comments. We assemble and release a novel dataset (Shellcode{\_}IA32), consisting of challenging but common assembly instructions with their natural language descriptions. We experiment with standard methods in neural machine translation (NMT) to establish baseline performance levels on this task.",
}
- Downloads last month
- 125