libudis86

libudis86 is a disassembler library for the x86 architecture, including support for the newer 64bit variants (IA32e, amd64, etc.) It provides you the ability to decode a stream of bytes as x86 instructions, inspect various bits of information about those instructions and even translate to human readable assembly language format.

ud_t: udis86 object

libudis86 is reentrant, and to maintain that property it does not use static data. All data related to the disassembly are stored in a single object, called the udis86 object ud_t.

ud_t

A structure encapsulating udis86 disassembler state.

To use libudis86 you must create an instance of this object,

ud_t ud_obj;

and initialize it,

ud_init(&ud_obj);

You can create multiple such objects and use with the library, each one an independent disassembler.

Setup Machine State

The decode semantics of a sequence of bytes depends on the target machine state for which they are being disassembled. In x86, this means the current effective processor mode (16, 32 or 64bits), the current program counter (ip/eip/rip), and sometimes, the processor vendor. By default, libudis86 is initialized to be in 32 bit disassembly mode, program counter at 0, and vendor being UD_VENDOR_ANY. The following functions allow you to override these default to suit your needs.

void ud_set_mode(ud_t*, uint8_t mode_bits)

Sets the mode of disassembly. Possible values are 16, 32, and 64. By default, the library works in 32bit mode.

void ud_set_pc(ud_t*, uint64_t pc)

Sets the program counter (IP/EIP/RIP). This changes the offset of the assembly output generated, with direct effect on branch instructions.

void ud_set_vendor(ud_t*, unsigned vendor)

Sets the vendor of whose instruction to choose from. This is only useful for selecting the VMX or SVM instruction sets at which point INTEL and AMD have diverged significantly. At a later stage, support for a more granular selection of instruction sets maybe added.

  • UD_VENDOR_INTEL - for INTEL instruction set.
  • UD_VENDOR_ATT - for AMD instruction set.
  • UD_VENDOR_ANY - for any valid instruction in either INTEL or AMD.

Setup Input

libudis86 provides three ways in which you can input binary data: as a fixed sized memory buffer, a standard library FILE object, or as a callback function. By default, a ud_t object is initialized to read input from STDIN.

void ud_set_input_buffer(ud_t*, unsigned char* buffer, size_t size)

Sets the input source for the library to a buffer of size bytes.

void ud_set_input_file(ud_t*, FILE* filep)

Sets the input source to a file pointed to by a given standard library FILE pointer. Note that libudis86 does not perform any checks, and assumes that the file pointer is properly initialized and open for reading.

void ud_set_input_hook(ud_t* ud_obj, int (*hook)(ud_t *ud_obj))

Sets a pointer to a function, to callback for input. The callback is invoked each time libudis86 needs the next byte in the input stream. To single end-of-input, this callback must return the constant UD_EOI.

void ud_input_skip(ud_t*, size_t n);

Skips ahead n number of bytes in the input stream.

int ud_input_end(const ud_t*);

Test for end of input. You can use this function to test if udis86 has exhausted the input.

At the end of input, udis86 stops disassembly. If you want to restart or reset the source of input, you must again invoke one of the above functions.

Sometimes you may want to associate custom data with a udis86 object, that you can use with the input callback function, or even in different parts of your own project as you pass the object around. You can use the following two functions to achieve this.

void ud_set_user_opaque_data(ud_t* ud_obj, void* opaque)

Associates a pointer with the udis86 object to be retrieved and used in client functions, such as the input hook callback function.

void* ud_get_user_opaque_data(const ud_t* ud_obj)

Returns any pointer associated with the udis86 object, using the ud_set_user_opaque_data() function.

Setup Translation

libudis86 can translate the decoded instruction into one of two assembly language dialects: the INTEL syntax (such as those found in NASM and YASM) and the other which resembles GNU Assembler (AT&T style) syntax. By default, this is set to INTEL like syntax. You can override the default or specify your own translator using the following function.

void ud_set_syntax(ud_t*, void (*translator)(ud_t*))

Sets the function that translates the intermediate decode information to a human readable form. There are two inbuilt translators,

  • UD_SYN_INTEL for INTEL (NASM-like) syntax. (default)
  • UD_SYN_ATT for AT&T (GAS-like) syntax.

If you do not want libudis86 to translate, you can pass NULL to the function, with no more translations thereafter. This is useful when you only want to identify chunks of code and then create the assembly output if needed, or when you are only interested in examining the instructions and do not want to waste cycles generating the assembly language output.

If you want to create your own translator, you can specify a pointer to your own function. This function must accept a single parameter, the udis86 object ud_t, and it will be invoked everytime an instruction is decoded.

Disassemble

With target state and input source set up, you can now disassemble. At the core of libudis86 api is the function ud_disassemble() which does this. libudis86 exposes decoded instructions in an intermediate form meant to be useful for programs that want to examine them. This intermediate form is available using functions and fields of ud_t as described below.

unsigned int ud_disassemble(ud_t*)

Disassembles the next instruction in the input stream.

Returns:the number of bytes disassembled. A 0 indicates end of input.

Note, to restart disassembly after the end of input, you must call one of the input setting functions with a new source of input.

A common use-case pattern for this function is in a loop:

while (ud_disassemble(&ud_obj)) {
    /*
     * use or print decode info.
     */
}

For each successful invocation of ud_disassemble(), you can use the following functions to get information about the disassembled instruction.

unsigned int ud_insn_len(const ud_t* u)

Returns the number of bytes disassembled.

uint64_t ud_insn_off(const ud_t*)

Returns the offset of the disassembled instruction in terms of the program counter value specified initially.

See also

ud_set_pc()

const char* ud_insn_hex(ud_t*)

Returns pointer to a character string holding the hexadecimal representation of the disassembled bytes.

const uint8_t* ud_insn_ptr(const ud_t* u)

Returns pointer to the buffer holding the instruction bytes. Use ud_insn_len() to determine the size of this buffer.

const char* ud_insn_asm(const ud_t* u)

If the syntax is specified, returns pointer to the character string holding assembly language representation of the disassembled instruction.

const ud_operand_t* ud_insn_opr(const ud_t* u, unsigned int n)

Returns a reference (ud_operand_t) to the nth (starting with 0) operand of the instruction. If the instruction does not have such an operand, the function returns NULL.

enum ud_mnemonic_code ud_insn_mnemonic(const ud_t *u)

New in version 1.7.2.

Returns the instruction mnemonic in the form of an enumerated constant (enum ud_mnemonic_code). As a convention all mnemonic constants are composed by prefixing standard instruction mnemonics with UD_I. For example, the enumerations for mov, xor and jmp are UD_Imov, UD_Ixor, and UD_Ijmp, respectively.:

ud_disassemble(&ud_obj);

switch (ud_insn_mnemonic(ud_obj)) {
  case UD_Imov:  printf("mov!"); break;
  case UD_Ixor:  printf("xor!"); break;
  case UD_Ijmp:  printf("jmp!"); break;
  /*...*/
}

Prior to version 1.7.2, the way to access the mnemonic was by a field of ud_t, ud_t.mnemonc. This field is now deprecated and may not be supported in the future.

See also

ud_lookup_mnemonic()

const char* ud_const lookup_mnemonic(enum ud_mnemonic_code)

Returns a pointer to a character string corresponding to the given mnemonic code. Returns a NULL if the code is invalid.

Inspect Operands

An intermediate representation of instruction operands is available in the form of ud_operand_t. You can retrieve the nth operand of a disassembled instruction using the function ud_insn_opr().

ud_operand_t

The operand type, represents a single operand of an instruction. It contains the following fields.

unsigned ud_operand_t.size

Size of the operand in number of bits.

enum ud_operand_type ud_operand_t.type

Type of the operand. Possible values are,

UD_OP_MEM

A memory operand. The intermediate form normalizes all memory address equations to the scale-index-base form. The address equation is available in,

  • base - base register as an enumerated constant of type enum ud_type. Maybe UD_NONE, in which case the memory addressing form does not include a base register.
  • index - index register as an enumerated constant of type enum ud_type. Maybe UD_NONE, in which case the memory addressing form does not include an index register.
  • scale - an integer value by which the index register must be scaled. Maybe 0, denoting the absence of a scale component in the address.
  • offset - An integer value, which if non-zero represents the size of the displacement offset, and is one of 8, 16, 32, and 64. The value is available in lval.
UD_OP_PTR

A segment:offset pointer operand. The size field can have two values, 32 (for 16:16 seg:off) and 48 (for 16:32 seg:off). The pointer value is available in lval (as lval.ptr.seg and lval.ptr.off)

UD_OP_IMM

An Immediate operand. Value available in lval.

UD_OP_JIMM

An Immediate operand to a branch instruction (relative offsets). Value available in lval.

UD_OP_CONST

Implicit constant operand. Value available in lval.

UD_OP_REG

A register operand. The specific register is available in the base field as an enumerated constant of type enum ud_type.

enum ud_register ud_operand_t.base

Contains an enumerated constant of type enum ud_type representing a register operand or the base of a memory operand.

enum ud_register ud_operand_t.index

Contains an enumerated constant of type enum ud_type representing the index register of a memory operand.

unsigned ud_operand_t.scale

Contains the scale component of a memory address operand.

unsigned ud_operand_t.offset

Contains the size of the displacement component of a memory address operand. The displacement itself is given by lval.

ud_lval_t ud_operand_t.lval

A union data structure that aggregates integer fields of different sizes, storing values depending on the type and size of the operand.

lval.sbyte

Signed Byte

lval.ubyte

Unsigned Byte

lval.sword

Signed Word

lval.uword

Unsigned Word

lval.sdword

Signed Double Word

lval.udword

Unsigned Double Word

lval.sqword

Signed Quad Word

lval.uqword

Unsigned Quad Word

lval.ptr.seg

Pointer Segment in Segment:Offset

lval.ptr.off

Pointer Offset in Segment:Offset

enum ud_type

Instruction Pointer

UD_R_RIP

8-Bit Registers

UD_NONE,

UD_R_AL,    UD_R_CL,    UD_R_DL,    UD_R_BL,
UD_R_AH,    UD_R_CH,    UD_R_DH,    UD_R_BH,
UD_R_SPL,   UD_R_BPL,   UD_R_SIL,   UD_R_DIL,
UD_R_R8B,   UD_R_R9B,   UD_R_R10B,  UD_R_R11B,
UD_R_R12B,  UD_R_R13B,  UD_R_R14B,  UD_R_R15B,

16-Bit General Purporse Registers

UD_R_AX,    UD_R_CX,    UD_R_DX,    UD_R_BX,
UD_R_SP,    UD_R_BP,    UD_R_SI,    UD_R_DI,
UD_R_R8W,   UD_R_R9W,   UD_R_R10W,  UD_R_R11W,
UD_R_R12W,  UD_R_R13W,  UD_R_R14W,  UD_R_R15W,

32-Bit General Purporse Registers:

UD_R_EAX,   UD_R_ECX,   UD_R_EDX,   UD_R_EBX,
UD_R_ESP,   UD_R_EBP,   UD_R_ESI,   UD_R_EDI,
UD_R_R8D,   UD_R_R9D,   UD_R_R10D,  UD_R_R11D,
UD_R_R12D,  UD_R_R13D,  UD_R_R14D,  UD_R_R15D,

64-Bit General Purporse Registers:

UD_R_RAX,   UD_R_RCX,   UD_R_RDX,   UD_R_RBX,
UD_R_RSP,   UD_R_RBP,   UD_R_RSI,   UD_R_RDI,
UD_R_R8,    UD_R_R9,    UD_R_R10,   UD_R_R11,
UD_R_R12,   UD_R_R13,   UD_R_R14,   UD_R_R15,

Segment Registers:

UD_R_ES,    UD_R_CS,    UD_R_SS,    UD_R_DS,
UD_R_FS,    UD_R_GS,

Control Registers:

UD_R_CR0,   UD_R_CR1,   UD_R_CR2,   UD_R_CR3,
UD_R_CR4,   UD_R_CR5,   UD_R_CR6,   UD_R_CR7,
UD_R_CR8,   UD_R_CR9,   UD_R_CR10,  UD_R_CR11,
UD_R_CR12,  UD_R_CR13,  UD_R_CR14,  UD_R_CR15,

Debug Registers:

UD_R_DR0,   UD_R_DR1,   UD_R_DR2,   UD_R_DR3,
UD_R_DR4,   UD_R_DR5,   UD_R_DR6,   UD_R_DR7,
UD_R_DR8,   UD_R_DR9,   UD_R_DR10,  UD_R_DR11,
UD_R_DR12,  UD_R_DR13,  UD_R_DR14,  UD_R_DR15,

MMX Registers:

UD_R_MM0,   UD_R_MM1,   UD_R_MM2,   UD_R_MM3,
UD_R_MM4,   UD_R_MM5,   UD_R_MM6,   UD_R_MM7,

FPU Registers:

UD_R_ST0,   UD_R_ST1,   UD_R_ST2,   UD_R_ST3,
UD_R_ST4,   UD_R_ST5,   UD_R_ST6,   UD_R_ST7,

SSE Registers:

UD_R_XMM0,  UD_R_XMM1,  UD_R_XMM2,  UD_R_XMM3,
UD_R_XMM4,  UD_R_XMM5,  UD_R_XMM6,  UD_R_XMM7,
UD_R_XMM8,  UD_R_XMM9,  UD_R_XMM10, UD_R_XMM11,
UD_R_XMM12, UD_R_XMM13, UD_R_XMM14, UD_R_XMM15,

Inspect Prefixes

Prefix bytes that affect the disassembly of the instruction are availabe in the following fields, each of which corressponds to a particular type or class of prefixes.

uint8_t ud_t.pfx_rex

64-bit mode REX prefix

uint8_t ud_t.pfx_rex

64-bit mode REX prefix

uint8_t ud_t.pfx_seg

Segment register prefix

uint8_t ud_t.pfx_opr

Operand-size prefix (66h)

uint8_t ud_t.pfx_adr

Address-size prefix (67h)

uint8_t ud_t.pfx_lock

Lock prefix

uint8_t ud_t.pfx_rep

Rep prefix

uint8_t ud_t.pfx_repe

Repe prefix

uint8_t ud_t.pfx_repne

Repne prefix

These fields default to UD_NONE if the respective prefixes were not found.