/* OpenIDA - Outil d'analyse de fichiers binaires
* processor.c - gestion de l'architecture x86
*
* Copyright (C) 2008 Cyrille Bagard
*
* This file is part of OpenIDA.
*
* OpenIDA is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation; either version 3 of the License, or
* (at your option) any later version.
*
* OpenIDA is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with Foobar. If not, see .
*/
#include "processor.h"
#include
#include
#include "../processor-int.h"
#include "instruction.h"
#include "opcodes.h"
#include "operand.h"
typedef asm_x86_instr * (* read_instr) (const uint8_t *, off_t *, off_t, uint64_t, const asm_x86_processor *);
/* Carte d'identité d'un opcode */
typedef struct _x86_opcode
{
uint8_t prefix; /* préfixe eventuel */
uint8_t opcode; /* Opcode seul */
uint8_t op_ext; /* Extension de l'opcode */
bool opt_prefix; /* Préfixe optionnel ? */
bool has_op_ext; /* Ext. à prendre en compte ? */
const char *name; /* Désignation humaine */
read_instr read; /* Décodage de l'instruction */
} x86_opcode;
#define EXT_OPCODE_MASK 0x38
#define register_opcode(target, _prefix, _opcode, _name, _read) \
do { \
target.prefix = _prefix; \
target.opcode = _opcode; \
target.opt_prefix = true; \
target.has_op_ext = false; \
target.name = _name; \
target.read = _read; \
} while (0)
#define register_opcode_with_ext(target, _prefix, _opcode, _ext, _name, _read) \
do { \
target.prefix = _prefix; \
target.opcode = _opcode; \
target.op_ext = _ext << 3; \
target.opt_prefix = true; \
target.has_op_ext = true; \
target.name = _name; \
target.read = _read; \
} while (0)
/* Définition générique d'une architecture */
struct _asm_x86_processor
{
asm_processor base; /* A laisser en premier... */
AsmOperandSize operand_size; /* Taille par défaut */
x86_opcode opcodes[X86_OP_COUNT]; /* Liste des opcodes supportés */
};
/* Enregistre toutes les instructions reconnues pour x86. */
void x86_register_instructions(asm_x86_processor *);
/* Décode une instruction dans un flux de données. */
asm_instr *x86_fetch_instruction(const asm_x86_processor *, const uint8_t *, off_t *, off_t, uint64_t);
/* Traduit une instruction en version humainement lisible. */
void x86_print_instruction(const asm_x86_processor *, const asm_x86_instr *, char *, size_t, AsmSyntax);
/******************************************************************************
* *
* Paramètres : - *
* *
* Description : Crée le support de l'architecture x86. *
* *
* Retour : Architecture mise en place. *
* *
* Remarques : - *
* *
******************************************************************************/
asm_processor *create_x86_processor(void)
{
asm_x86_processor *result; /* Architecture à retourner */
result = (asm_x86_processor *)calloc(1, sizeof(asm_x86_processor));
result->operand_size = AOS_32_BITS;
x86_register_instructions(result);
ASM_PROCESSOR(result)->fetch_instr = (fetch_instruction)x86_fetch_instruction;
ASM_PROCESSOR(result)->print_instr = (print_instruction)x86_print_instruction;
return ASM_PROCESSOR(result);
}
/******************************************************************************
* *
* Paramètres : proc = architecture visée par la consultation. *
* *
* Description : Fournit la taille courante des opérandes pour x86. *
* *
* Retour : Taille d'opérande (16 ou 32 bits). *
* *
* Remarques : - *
* *
******************************************************************************/
AsmOperandSize get_x86_current_operand_size(const asm_x86_processor *proc)
{
return proc->operand_size;
}
/******************************************************************************
* *
* Paramètres : proc = architecture visée par la consultation. *
* *
* Description : Fournit la taille supplantée des opérandes pour x86. *
* *
* Retour : Taille d'opérande (16 ou 32 bits). *
* *
* Remarques : - *
* *
******************************************************************************/
AsmOperandSize switch_x86_operand_size(const asm_x86_processor *proc)
{
return (proc->operand_size == AOS_32_BITS ? AOS_16_BITS : AOS_32_BITS);
}
/******************************************************************************
* *
* Paramètres : proc = architecture visée par la procédure. *
* *
* Description : Enregistre toutes les instructions reconnues pour x86. *
* *
* Retour : - *
* *
* Remarques : - *
* *
******************************************************************************/
void x86_register_instructions(asm_x86_processor *proc)
{
register_opcode(proc->opcodes[X86_OP_XOR_REG1632], 0x00/*0x66*/, 0x31, "xor", read_instr_xor_with_reg1632);
register_opcode(proc->opcodes[X86_OP_INC_E_AX], 0x66, 0x40, "inc", read_instr_inc_1632);
register_opcode(proc->opcodes[X86_OP_INC_E_CX], 0x66, 0x41, "inc", read_instr_inc_1632);
register_opcode(proc->opcodes[X86_OP_INC_E_DX], 0x66, 0x42, "inc", read_instr_inc_1632);
register_opcode(proc->opcodes[X86_OP_INC_E_BX], 0x66, 0x43, "inc", read_instr_inc_1632);
register_opcode(proc->opcodes[X86_OP_INC_E_SP], 0x66, 0x44, "inc", read_instr_inc_1632);
register_opcode(proc->opcodes[X86_OP_INC_E_BP], 0x66, 0x45, "inc", read_instr_inc_1632);
register_opcode(proc->opcodes[X86_OP_INC_E_SI], 0x66, 0x46, "inc", read_instr_inc_1632);
register_opcode(proc->opcodes[X86_OP_INC_E_DI], 0x66, 0x47, "inc", read_instr_inc_1632);
register_opcode(proc->opcodes[X86_OP_DEC_E_AX], 0x66, 0x48, "dec", read_instr_dec_1632);
register_opcode(proc->opcodes[X86_OP_DEC_E_CX], 0x66, 0x49, "dec", read_instr_dec_1632);
register_opcode(proc->opcodes[X86_OP_DEC_E_DX], 0x66, 0x4a, "dec", read_instr_dec_1632);
register_opcode(proc->opcodes[X86_OP_DEC_E_BX], 0x66, 0x4b, "dec", read_instr_dec_1632);
register_opcode(proc->opcodes[X86_OP_DEC_E_SP], 0x66, 0x4c, "dec", read_instr_dec_1632);
register_opcode(proc->opcodes[X86_OP_DEC_E_BP], 0x66, 0x4d, "dec", read_instr_dec_1632);
register_opcode(proc->opcodes[X86_OP_DEC_E_SI], 0x66, 0x4e, "dec", read_instr_dec_1632);
register_opcode(proc->opcodes[X86_OP_DEC_E_DI], 0x66, 0x4f, "dec", read_instr_dec_1632);
register_opcode(proc->opcodes[X86_OP_PUSH_E_AX], 0x66, 0x50, "push", read_instr_push_reg1632);
register_opcode(proc->opcodes[X86_OP_PUSH_E_CX], 0x66, 0x51, "push", read_instr_push_reg1632);
register_opcode(proc->opcodes[X86_OP_PUSH_E_DX], 0x66, 0x52, "push", read_instr_push_reg1632);
register_opcode(proc->opcodes[X86_OP_PUSH_E_BX], 0x66, 0x53, "push", read_instr_push_reg1632);
register_opcode(proc->opcodes[X86_OP_PUSH_E_SP], 0x66, 0x54, "push", read_instr_push_reg1632);
register_opcode(proc->opcodes[X86_OP_PUSH_E_BP], 0x66, 0x55, "push", read_instr_push_reg1632);
register_opcode(proc->opcodes[X86_OP_PUSH_E_SI], 0x66, 0x56, "push", read_instr_push_reg1632);
register_opcode(proc->opcodes[X86_OP_PUSH_E_DI], 0x66, 0x57, "push", read_instr_push_reg1632);
register_opcode(proc->opcodes[X86_OP_POP_E_AX], 0x66, 0x58, "pop", read_instr_pop_reg1632);
register_opcode(proc->opcodes[X86_OP_POP_E_CX], 0x66, 0x59, "pop", read_instr_pop_reg1632);
register_opcode(proc->opcodes[X86_OP_POP_E_DX], 0x66, 0x5a, "pop", read_instr_pop_reg1632);
register_opcode(proc->opcodes[X86_OP_POP_E_BX], 0x66, 0x5b, "pop", read_instr_pop_reg1632);
register_opcode(proc->opcodes[X86_OP_POP_E_SP], 0x66, 0x5c, "pop", read_instr_pop_reg1632);
register_opcode(proc->opcodes[X86_OP_POP_E_BP], 0x66, 0x5d, "pop", read_instr_pop_reg1632);
register_opcode(proc->opcodes[X86_OP_POP_E_SI], 0x66, 0x5e, "pop", read_instr_pop_reg1632);
register_opcode(proc->opcodes[X86_OP_POP_E_DI], 0x66, 0x5f, "pop", read_instr_pop_reg1632);
register_opcode(proc->opcodes[X86_OP_PUSH_IMM1632], 0x66, 0x68, "push", read_instr_push_imm1632);
register_opcode(proc->opcodes[X86_OP_MOV_REG1632], 0x66, 0x89, "mov", read_instr_mov_with_reg1632);
register_opcode_with_ext(proc->opcodes[X86_OP_ADD8_REG1632], 0x66, 0x83, 0, "add", read_instr_add8_with_reg1632);
register_opcode_with_ext(proc->opcodes[X86_OP_OR8_REG1632], 0x66, 0x83, 1, "or", read_instr_or8_with_reg1632);
register_opcode_with_ext(proc->opcodes[X86_OP_ADC8_REG1632], 0x66, 0x83, 2, "adc", read_instr_adc8_with_reg1632);
register_opcode_with_ext(proc->opcodes[X86_OP_SBB8_REG1632], 0x66, 0x83, 3, "sbb", read_instr_sbb8_with_reg1632);
register_opcode_with_ext(proc->opcodes[X86_OP_AND8_REG1632], 0x66, 0x83, 4, "and", read_instr_and8_with_reg1632);
register_opcode_with_ext(proc->opcodes[X86_OP_SUB8_REG1632], 0x66, 0x83, 5, "sub", read_instr_sub8_with_reg1632);
register_opcode_with_ext(proc->opcodes[X86_OP_XOR8_REG1632], 0x66, 0x83, 6, "xor", read_instr_xor8_with_reg1632);
register_opcode(proc->opcodes[X86_OP_NOP], 0x00, 0x90, "nop", read_instr_nop);
register_opcode(proc->opcodes[X86_OP_MOV_E_AX], 0x66, 0xb8, "mov", read_instr_mov_to_1632);
register_opcode(proc->opcodes[X86_OP_MOV_E_CX], 0x66, 0xb9, "mov", read_instr_mov_to_1632);
register_opcode(proc->opcodes[X86_OP_MOV_E_DX], 0x66, 0xba, "mov", read_instr_mov_to_1632);
register_opcode(proc->opcodes[X86_OP_MOV_E_BX], 0x66, 0xbb, "mov", read_instr_mov_to_1632);
register_opcode(proc->opcodes[X86_OP_MOV_E_SP], 0x66, 0xbc, "mov", read_instr_mov_to_1632);
register_opcode(proc->opcodes[X86_OP_MOV_E_BP], 0x66, 0xbd, "mov", read_instr_mov_to_1632);
register_opcode(proc->opcodes[X86_OP_MOV_E_SI], 0x66, 0xbe, "mov", read_instr_mov_to_1632);
register_opcode(proc->opcodes[X86_OP_MOV_E_DI], 0x66, 0xbf, "mov", read_instr_mov_to_1632);
register_opcode(proc->opcodes[X86_OP_RET], 0x00, 0xc3, "ret", read_instr_ret);
register_opcode(proc->opcodes[X86_OP_LEAVE], 0x00, 0xc9, "leave", read_instr_leave);
register_opcode(proc->opcodes[X86_OP_INT], 0x00, 0xcd, "int", read_instr_int);
register_opcode(proc->opcodes[X86_OP_CALL], 0x00, 0xe8, "call", read_instr_call);
register_opcode(proc->opcodes[X86_OP_HLT], 0x00, 0xf4, "hlt", read_instr_hlt);
}
/******************************************************************************
* *
* Paramètres : proc = architecture visée par la procédure. *
* data = flux de données à analyser. *
* pos = position courante dans ce flux. [OUT] *
* len = taille totale des données à analyser. *
* offset = adresse virtuelle de l'instruction. *
* *
* Description : Décode une instruction dans un flux de données. *
* *
* Retour : Instruction mise en place ou NULL. *
* *
* Remarques : - *
* *
******************************************************************************/
asm_instr *x86_fetch_instruction(const asm_x86_processor *proc, const uint8_t *data, off_t *pos, off_t len, uint64_t offset)
{
asm_x86_instr *result; /* Résultat à faire remonter */
X86Opcodes i; /* Boucle de parcours */
off_t tmp; /* Tête de lecture */
result = NULL;
for (i = 0; i < X86_OP_COUNT; i++)
{
if (proc->opcodes[i].prefix > 0 && data[*pos] == proc->opcodes[i].prefix && data[*pos + 1] == proc->opcodes[i].opcode)
{
tmp = *pos + 2;
goto find_instr;
}
if (proc->opcodes[i].opt_prefix && data[*pos] == proc->opcodes[i].opcode)
{
tmp = *pos + 1;
goto find_instr;
}
continue;
find_instr:
if (proc->opcodes[i].has_op_ext && (data[tmp] & EXT_OPCODE_MASK) != proc->opcodes[i].op_ext)
continue;
result = proc->opcodes[i].read(data, pos, len, offset, proc);
if (result != NULL) result->type = i;
else printf("err while decoding :: [0x%02hhx] 0x%02hhx\n", proc->opcodes[i].prefix, proc->opcodes[i].opcode);
break;
}
return ASM_INSTRUCTION(result);
}
/******************************************************************************
* *
* Paramètres : proc = architecture visée par la procédure. *
* instr = instruction à traiter. *
* buffer = tampon de sortie mis à disposition. [OUT] *
* len = taille de ce tampon. *
* syntax = type de représentation demandée. *
* *
* Description : Traduit une instruction en version humainement lisible. *
* *
* Retour : - *
* *
* Remarques : - *
* *
******************************************************************************/
void x86_print_instruction(const asm_x86_processor *proc, const asm_x86_instr *instr, char *buffer, size_t len, AsmSyntax syntax)
{
size_t i; /* Boucle de parcours */
char opbuffer[3][64]; /* Tampon pour les textes */
/* Impression des opérandes */
for (i = 0; i < ASM_INSTRUCTION(instr)->operands_count; i++)
switch (ASM_OPERAND(ASM_INSTRUCTION(instr)->operands[i])->type)
{
case AOT_NONE:
print_db_operand(ASM_OPERAND(ASM_INSTRUCTION(instr)->operands[i]), opbuffer[i], 64, syntax);
break;
case AOT_IMM:
print_imm_operand(ASM_OPERAND(ASM_INSTRUCTION(instr)->operands[i]), opbuffer[i], 64, syntax);
break;
case AOT_REG:
x86_print_reg_operand(ASM_INSTRUCTION(instr)->operands[i], opbuffer[i], 64, syntax);
break;
case AOT_MEM:
break;
}
/* Impression globale finale */
if (ASM_INSTRUCTION(instr)->opcode == DB_OPCODE)
snprintf(buffer, len, "db\t%s", opbuffer[0]);
else
switch (ASM_INSTRUCTION(instr)->operands_count)
{
case 0:
snprintf(buffer, len, "%s", proc->opcodes[instr->type].name);
break;
case 1:
snprintf(buffer, len, "%s\t%s", proc->opcodes[instr->type].name, opbuffer[0]);
break;
case 2:
switch (syntax)
{
case ASX_INTEL:
snprintf(buffer, len, "%s\t%s, %s", proc->opcodes[instr->type].name, opbuffer[0], opbuffer[1]);
break;
case ASX_ATT:
snprintf(buffer, len, "%s\t%s, %s", proc->opcodes[instr->type].name, opbuffer[1], opbuffer[0]);
break;
}
break;
}
}