this is just a bad joke between hell and elf :(
The aim of this project is to provide a python library for patching ELF binary file. It only supports for the moment x86 and x86_64 architecture.
pip install Hellf
from Hellf import ELF
from pwn import shellcraft, asm, context
context.arch = "amd64"
e = ELF("/bin/ls")
e.get_section_by_name(".text").data = asm(shellcraft.amd64.sh())
e.Elf64_Ehdr.e_entry = e.get_section_by_name(".text").sh_addr
e.save("/tmp/not_really_ls_anymore")
Hellf allows you to modify each par of the ELF.
- Program Header (segments definitions)
- Section Header (sections definitions)
- Section data (.text, .plt, .data, .bss, ...)
You can change every single byte of a given binary.
new_section = Elf64_Shdr_ST()
# creating our shellcode (exit(14);)
asm = """
mov rax, 0x3c
mov rdi, 14
syscall
"""
ks = Ks(KS_ARCH_X86, KS_MODE_64)
ks.syntax = KS_OPT_SYNTAX_NASM
shellcode = ks.asm(asm)[0]
shellcode = (c_char * len(shellcode)).from_buffer(bytearray(shellcode)).raw
new_section.data = shellcode
new_section.sh_name = 1 # choose a random name
new_section.sh_offset = e.Elf64_Ehdr.e_shoff
new_section.sh_size = len(new_section.data)
new_section.sh_addralign = 16
new_section.sh_type = 0x1 # PROGBITS
# adding our sections in the Section Header table
e.add_section(new_section)# need to extend the last segment to englobe our new section
segment = e.Elf64_Phdr[5]
segment.p_filesz = new_section.sh_offset + new_section.sh_size - segment.p_offset
segment.p_memsz = new_section.sh_offset + new_section.sh_size - segment.p_offset
segment.p_flags = 7 # RWX
# modifying our entrypoint to point to the shellcode
e.Elf64_Ehdr.e_entry = segment.p_vaddr - segment.p_offset + new_section.sh_offset
e.save("/tmp/exit")Hellf is a part of Embuche, a lot of Hellf scripts are used to obfuscate ELF for the purpose of this project. The scripts used are described here : File Format Hacks
For example it used to create a fake .dynsym section or to hide the entry point link to script
But Hellf is also the corner stone of the Embuche metamorphic packer.
A metamorphic packer is available in Embuche. This packer will load your binary and cipher it (AES 256 bits CBC).
If you decide to use the packer, your program will be ciphered and stored in a section of our packer. When you will execute your program the packer will copy itself in memory, unciphered your program and write it on the disk for execution.
Beside cipher your binary, the packer will also ensure its integrity. The encryption keys used for the encryption are based on the SHA sum of the
.textsection, so if the packer or your program is being debugged the SHA sum will be different of the one used for decryption.The ELF of the packer can be modified with the
packer_embucheoptions.
There is almost no documentation for the moment although the code is a bit commented, you should be lucky.
Here a list of function of the ELF object.
get_section_number(name)
get_section_by_name(name)
get_section_name(shstrtab_index)
add_section(custom_section)For the remaining, you should just interact with the object attributes them self. About the names of variables I just used the official one provided by the ABI.
- /usr/include/elf.h
- http://www.sco.com/developers/gabi/latest/contents.html
For example, the ELF Header :
typedef struct
{
unsigned char e_ident[EI_NIDENT]; /* Magic number and other info */
Elf64_Half e_type; /* Object file type */
Elf64_Half e_machine; /* Architecture */
Elf64_Word e_version; /* Object file version */
Elf64_Addr e_entry; /* Entry point virtual address */
Elf64_Off e_phoff; /* Program header table file offset */
Elf64_Off e_shoff; /* Section header table file offset */
Elf64_Word e_flags; /* Processor-specific flags */
Elf64_Half e_ehsize; /* ELF header size in bytes */
Elf64_Half e_phentsize; /* Program header table entry size */
Elf64_Half e_phnum; /* Program header table entry count */
Elf64_Half e_shentsize; /* Section header table entry size */
Elf64_Half e_shnum; /* Section header table entry count */
Elf64_Half e_shstrndx; /* Section header string table index */
} Elf64_Ehdr;
So you can just do this to interact with the section or section header :
> e.Elf64_Shdr[26]
ELF Sections header struct
sh_name: 0x1
sh_type: 0x3
sh_flags: 0x0
sh_addr: 0x0
sh_offset: 0x222b4
sh_size: 0xf7
sh_link: 0x0
sh_info: 0x0
sh_addralign: 0x1
sh_entsize: 0x0
> e.Elf64_Shdr[26].data
b'\x00.shstrtab\x00.interp\x00.note.gnu.build-id\x00.note.ABI-tag\x00.gnu.hash\x00.dynsym\x00.dynstr\x00.gnu.version\x00.gnu.version_r\x00.rela.dyn\x00.rela.plt\x00.init\x00.text\x00.fini\x00.rodata\x00.eh_frame_hdr\x00.eh_frame\x00.init_array\x00.fini_array\x00.data.rel.ro\x00.dynamic\x00.got\x00.data\x00.bss\x00.comment\x00'
> e.get_section_name(e.Elf64_Shdr[26].sh_name)
'.shstrtab'Or if you want to interact with the header itself.
> e.Elf64_Ehdr
ELF Header struct
e_ident: 0x7f 0x45 0x4c 0x46 0x2 0x1 0x1 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0 0x0
e_type: 0x3
e_machine: 0x3e
e_version: 0x1
e_entry: 0x5b20
e_phoff: 0x40
e_shoff: 0x2234b
e_flags: 0x0
e_ehsize: 0x40
e_phentsize: 0x38
e_phnum: 0xb
e_shentsize: 0x40
e_shnum: 0x1b
e_shstrndx: 0x1a
> hex(e.Elf64_Ehdr.e_entry)
'0x5b20'
> hex(e.Elf64_Ehdr.e_shnum)
'0x1b'
> hex(e.Elf64_Ehdr.e_shoff)
'0x2234b'
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent