blob: 17d3c557108efc947fe44efe501d32314edd89e9 [file] [log] [blame]
/* Capstone Disassembly Engine */
/* BPF Backend by david942j <david942j@gmail.com>, 2019 */
#include <string.h>
#include "BPFConstants.h"
#include "BPFMapping.h"
#include "../../Mapping.h"
#include "../../utils.h"
#ifndef CAPSTONE_DIET
static const name_map group_name_maps[] = {
{ BPF_GRP_INVALID, NULL },
{ BPF_GRP_LOAD, "load" },
{ BPF_GRP_STORE, "store" },
{ BPF_GRP_ALU, "alu" },
{ BPF_GRP_JUMP, "jump" },
{ BPF_GRP_CALL, "call" },
{ BPF_GRP_RETURN, "return" },
{ BPF_GRP_MISC, "misc" },
};
#endif
const char *BPF_group_name(csh handle, unsigned int id)
{
#ifndef CAPSTONE_DIET
return id2name(group_name_maps, ARR_SIZE(group_name_maps), id);
#else
return NULL;
#endif
}
#ifndef CAPSTONE_DIET
static const name_map insn_name_maps[BPF_INS_ENDING] = {
{ BPF_INS_INVALID, NULL },
{ BPF_INS_ADD, "add" },
{ BPF_INS_SUB, "sub" },
{ BPF_INS_MUL, "mul" },
{ BPF_INS_DIV, "div" },
{ BPF_INS_OR, "or" },
{ BPF_INS_AND, "and" },
{ BPF_INS_LSH, "lsh" },
{ BPF_INS_RSH, "rsh" },
{ BPF_INS_NEG, "neg" },
{ BPF_INS_MOD, "mod" },
{ BPF_INS_XOR, "xor" },
{ BPF_INS_MOV, "mov" },
{ BPF_INS_ARSH, "arsh" },
{ BPF_INS_ADD64, "add64" },
{ BPF_INS_SUB64, "sub64" },
{ BPF_INS_MUL64, "mul64" },
{ BPF_INS_DIV64, "div64" },
{ BPF_INS_OR64, "or64" },
{ BPF_INS_AND64, "and64" },
{ BPF_INS_LSH64, "lsh64" },
{ BPF_INS_RSH64, "rsh64" },
{ BPF_INS_NEG64, "neg64" },
{ BPF_INS_MOD64, "mod64" },
{ BPF_INS_XOR64, "xor64" },
{ BPF_INS_MOV64, "mov64" },
{ BPF_INS_ARSH64, "arsh64" },
{ BPF_INS_LE16, "le16" },
{ BPF_INS_LE32, "le32" },
{ BPF_INS_LE64, "le64" },
{ BPF_INS_BE16, "be16" },
{ BPF_INS_BE32, "be32" },
{ BPF_INS_BE64, "be64" },
{ BPF_INS_BSWAP16, "bswap16" },
{ BPF_INS_BSWAP32, "bswap32" },
{ BPF_INS_BSWAP64, "bswap64" },
{ BPF_INS_LDW, "ldw" },
{ BPF_INS_LDH, "ldh" },
{ BPF_INS_LDB, "ldb" },
{ BPF_INS_LDDW, "lddw" },
{ BPF_INS_LDXW, "ldxw" },
{ BPF_INS_LDXH, "ldxh" },
{ BPF_INS_LDXB, "ldxb" },
{ BPF_INS_LDXDW, "ldxdw" },
{ BPF_INS_STW, "stw" },
{ BPF_INS_STH, "sth" },
{ BPF_INS_STB, "stb" },
{ BPF_INS_STDW, "stdw" },
{ BPF_INS_STXW, "stxw" },
{ BPF_INS_STXH, "stxh" },
{ BPF_INS_STXB, "stxb" },
{ BPF_INS_STXDW, "stxdw" },
{ BPF_INS_XADDW, "xaddw" },
{ BPF_INS_XADDDW, "xadddw" },
{ BPF_INS_JMP, "jmp" },
{ BPF_INS_JEQ, "jeq" },
{ BPF_INS_JGT, "jgt" },
{ BPF_INS_JGE, "jge" },
{ BPF_INS_JSET, "jset" },
{ BPF_INS_JNE, "jne" },
{ BPF_INS_JSGT, "jsgt" },
{ BPF_INS_JSGE, "jsge" },
{ BPF_INS_CALL, "call" },
{ BPF_INS_CALLX, "callx" },
{ BPF_INS_EXIT, "exit" },
{ BPF_INS_JLT, "jlt" },
{ BPF_INS_JLE, "jle" },
{ BPF_INS_JSLT, "jslt" },
{ BPF_INS_JSLE, "jsle" },
{ BPF_INS_RET, "ret" },
{ BPF_INS_TAX, "tax" },
{ BPF_INS_TXA, "txa" },
};
#endif
const char *BPF_insn_name(csh handle, unsigned int id)
{
#ifndef CAPSTONE_DIET
/* We have some special cases because 'ld' in cBPF is equivalent to 'ldw'
* in eBPF, and we don't want to see 'ldw' appears in cBPF mode.
*/
if (!EBPF_MODE(handle)) {
switch (id) {
case BPF_INS_LD: return "ld";
case BPF_INS_LDX: return "ldx";
case BPF_INS_ST: return "st";
case BPF_INS_STX: return "stx";
}
}
return id2name(insn_name_maps, ARR_SIZE(insn_name_maps), id);
#else
return NULL;
#endif
}
const char *BPF_reg_name(csh handle, unsigned int reg)
{
#ifndef CAPSTONE_DIET
if (EBPF_MODE(handle)) {
if (reg < BPF_REG_R0 || reg > BPF_REG_R10)
return NULL;
static const char reg_names[11][4] = {
"r0", "r1", "r2", "r3", "r4",
"r5", "r6", "r7", "r8", "r9",
"r10"
};
return reg_names[reg - BPF_REG_R0];
}
/* cBPF mode */
if (reg == BPF_REG_A)
return "a";
else if (reg == BPF_REG_X)
return "x";
else
return NULL;
#else
return NULL;
#endif
}
static bpf_insn op2insn_ld(unsigned opcode)
{
#define CASE(c) case BPF_SIZE_##c: \
if (BPF_CLASS(opcode) == BPF_CLASS_LD) \
return BPF_INS_LD##c; \
else \
return BPF_INS_LDX##c;
switch (BPF_SIZE(opcode)) {
CASE(W);
CASE(H);
CASE(B);
CASE(DW);
}
#undef CASE
return BPF_INS_INVALID;
}
static bpf_insn op2insn_st(unsigned opcode)
{
/*
* - BPF_STX | BPF_XADD | BPF_{W,DW}
* - BPF_ST* | BPF_MEM | BPF_{W,H,B,DW}
*/
if (opcode == (BPF_CLASS_STX | BPF_MODE_XADD | BPF_SIZE_W))
return BPF_INS_XADDW;
if (opcode == (BPF_CLASS_STX | BPF_MODE_XADD | BPF_SIZE_DW))
return BPF_INS_XADDDW;
/* should be BPF_MEM */
#define CASE(c) case BPF_SIZE_##c: \
if (BPF_CLASS(opcode) == BPF_CLASS_ST) \
return BPF_INS_ST##c; \
else \
return BPF_INS_STX##c;
switch (BPF_SIZE(opcode)) {
CASE(W);
CASE(H);
CASE(B);
CASE(DW);
}
#undef CASE
return BPF_INS_INVALID;
}
static bpf_insn op2insn_alu(unsigned opcode)
{
/* Endian is a special case */
if (BPF_OP(opcode) == BPF_ALU_END) {
if (BPF_CLASS(opcode) == BPF_CLASS_ALU64) {
switch (opcode ^ BPF_CLASS_ALU64 ^ BPF_ALU_END ^ BPF_SRC_LITTLE) {
case (16 << 4):
return BPF_INS_BSWAP16;
case (32 << 4):
return BPF_INS_BSWAP32;
case (64 << 4):
return BPF_INS_BSWAP64;
default:
return BPF_INS_INVALID;
}
}
switch (opcode ^ BPF_CLASS_ALU ^ BPF_ALU_END) {
case BPF_SRC_LITTLE | (16 << 4):
return BPF_INS_LE16;
case BPF_SRC_LITTLE | (32 << 4):
return BPF_INS_LE32;
case BPF_SRC_LITTLE | (64 << 4):
return BPF_INS_LE64;
case BPF_SRC_BIG | (16 << 4):
return BPF_INS_BE16;
case BPF_SRC_BIG | (32 << 4):
return BPF_INS_BE32;
case BPF_SRC_BIG | (64 << 4):
return BPF_INS_BE64;
}
return BPF_INS_INVALID;
}
#define CASE(c) case BPF_ALU_##c: \
if (BPF_CLASS(opcode) == BPF_CLASS_ALU) \
return BPF_INS_##c; \
else \
return BPF_INS_##c##64;
switch (BPF_OP(opcode)) {
CASE(ADD);
CASE(SUB);
CASE(MUL);
CASE(DIV);
CASE(OR);
CASE(AND);
CASE(LSH);
CASE(RSH);
CASE(NEG);
CASE(MOD);
CASE(XOR);
CASE(MOV);
CASE(ARSH);
}
#undef CASE
return BPF_INS_INVALID;
}
static bpf_insn op2insn_jmp(unsigned opcode)
{
if (opcode == (BPF_CLASS_JMP | BPF_JUMP_CALL | BPF_SRC_X)) {
return BPF_INS_CALLX;
}
#define CASE(c) case BPF_JUMP_##c: return BPF_INS_##c
switch (BPF_OP(opcode)) {
case BPF_JUMP_JA:
return BPF_INS_JMP;
CASE(JEQ);
CASE(JGT);
CASE(JGE);
CASE(JSET);
CASE(JNE);
CASE(JSGT);
CASE(JSGE);
CASE(CALL);
CASE(EXIT);
CASE(JLT);
CASE(JLE);
CASE(JSLT);
CASE(JSLE);
}
#undef CASE
return BPF_INS_INVALID;
}
#ifndef CAPSTONE_DIET
static void update_regs_access(cs_struct *ud, cs_detail *detail,
bpf_insn insn_id, unsigned int opcode)
{
if (insn_id == BPF_INS_INVALID)
return;
#define PUSH_READ(r) do { \
detail->regs_read[detail->regs_read_count] = r; \
detail->regs_read_count++; \
} while (0)
#define PUSH_WRITE(r) do { \
detail->regs_write[detail->regs_write_count] = r; \
detail->regs_write_count++; \
} while (0)
/*
* In eBPF mode, only these instructions have implicit registers access:
* - legacy ld{w,h,b,dw} * // w: r0
* - exit // r: r0
*/
if (EBPF_MODE(ud)) {
switch (insn_id) {
default:
break;
case BPF_INS_LDW:
case BPF_INS_LDH:
case BPF_INS_LDB:
case BPF_INS_LDDW:
if (BPF_MODE(opcode) == BPF_MODE_ABS || BPF_MODE(opcode) == BPF_MODE_IND) {
PUSH_WRITE(BPF_REG_R0);
}
break;
case BPF_INS_EXIT:
PUSH_READ(BPF_REG_R0);
break;
}
return;
}
/* cBPF mode */
switch (BPF_CLASS(opcode)) {
default:
break;
case BPF_CLASS_LD:
PUSH_WRITE(BPF_REG_A);
break;
case BPF_CLASS_LDX:
PUSH_WRITE(BPF_REG_X);
break;
case BPF_CLASS_ST:
PUSH_READ(BPF_REG_A);
break;
case BPF_CLASS_STX:
PUSH_READ(BPF_REG_X);
break;
case BPF_CLASS_ALU:
PUSH_READ(BPF_REG_A);
PUSH_WRITE(BPF_REG_A);
break;
case BPF_CLASS_JMP:
if (insn_id != BPF_INS_JMP) // except the unconditional jump
PUSH_READ(BPF_REG_A);
break;
/* case BPF_CLASS_RET: */
case BPF_CLASS_MISC:
if (insn_id == BPF_INS_TAX) {
PUSH_READ(BPF_REG_A);
PUSH_WRITE(BPF_REG_X);
}
else {
PUSH_READ(BPF_REG_X);
PUSH_WRITE(BPF_REG_A);
}
break;
}
}
#endif
/*
* 1. Convert opcode(id) to BPF_INS_*
* 2. Set regs_read/regs_write/groups
*/
void BPF_get_insn_id(cs_struct *ud, cs_insn *insn, unsigned int opcode)
{
// No need to care the mode (cBPF or eBPF) since all checks has be done in
// BPF_getInstruction, we can simply map opcode to BPF_INS_*.
bpf_insn id = BPF_INS_INVALID;
#ifndef CAPSTONE_DIET
cs_detail *detail;
bpf_insn_group grp;
detail = insn->detail;
#define PUSH_GROUP(grp) do { \
if (detail) { \
detail->groups[detail->groups_count] = grp; \
detail->groups_count++; \
} \
} while(0)
#else
#define PUSH_GROUP(grp) do {} while(0)
#endif
switch (BPF_CLASS(opcode)) {
default: // will never happen
break;
case BPF_CLASS_LD:
case BPF_CLASS_LDX:
id = op2insn_ld(opcode);
PUSH_GROUP(BPF_GRP_LOAD);
break;
case BPF_CLASS_ST:
case BPF_CLASS_STX:
id = op2insn_st(opcode);
PUSH_GROUP(BPF_GRP_STORE);
break;
case BPF_CLASS_ALU:
id = op2insn_alu(opcode);
PUSH_GROUP(BPF_GRP_ALU);
break;
case BPF_CLASS_JMP:
id = op2insn_jmp(opcode);
#ifndef CAPSTONE_DIET
grp = BPF_GRP_JUMP;
if (id == BPF_INS_CALL || id == BPF_INS_CALLX)
grp = BPF_GRP_CALL;
else if (id == BPF_INS_EXIT)
grp = BPF_GRP_RETURN;
PUSH_GROUP(grp);
#endif
break;
case BPF_CLASS_RET:
id = BPF_INS_RET;
PUSH_GROUP(BPF_GRP_RETURN);
break;
// BPF_CLASS_MISC and BPF_CLASS_ALU64 have exactly same value
case BPF_CLASS_MISC:
/* case BPF_CLASS_ALU64: */
if (EBPF_MODE(ud)) {
// ALU64 in eBPF
id = op2insn_alu(opcode);
PUSH_GROUP(BPF_GRP_ALU);
}
else {
if (BPF_MISCOP(opcode) == BPF_MISCOP_TXA)
id = BPF_INS_TXA;
else
id = BPF_INS_TAX;
PUSH_GROUP(BPF_GRP_MISC);
}
break;
}
insn->id = id;
#undef PUSH_GROUP
#ifndef CAPSTONE_DIET
if (detail) {
update_regs_access(ud, detail, id, opcode);
}
#endif
}
static void sort_and_uniq(cs_regs arr, uint8_t n, uint8_t *new_n)
{
/* arr is always a tiny (usually n < 3) array,
* a simple O(n^2) sort is efficient enough. */
int i;
int j;
int iMin;
int tmp;
/* a modified selection sort for sorting and making unique */
for (j = 0; j < n; j++) {
/* arr[iMin] will be min(arr[j .. n-1]) */
iMin = j;
for (i = j + 1; i < n; i++) {
if (arr[i] < arr[iMin])
iMin = i;
}
if (j != 0 && arr[iMin] == arr[j - 1]) { // duplicate ele found
arr[iMin] = arr[n - 1];
--n;
}
else {
tmp = arr[iMin];
arr[iMin] = arr[j];
arr[j] = tmp;
}
}
*new_n = n;
}
void BPF_reg_access(const cs_insn *insn,
cs_regs regs_read, uint8_t *regs_read_count,
cs_regs regs_write, uint8_t *regs_write_count)
{
unsigned i;
uint8_t read_count, write_count;
const cs_bpf *bpf = &(insn->detail->bpf);
read_count = insn->detail->regs_read_count;
write_count = insn->detail->regs_write_count;
// implicit registers
memcpy(regs_read, insn->detail->regs_read, read_count * sizeof(insn->detail->regs_read[0]));
memcpy(regs_write, insn->detail->regs_write, write_count * sizeof(insn->detail->regs_write[0]));
for (i = 0; i < bpf->op_count; i++) {
const cs_bpf_op *op = &(bpf->operands[i]);
switch (op->type) {
default:
break;
case BPF_OP_REG:
if (op->access & CS_AC_READ) {
regs_read[read_count] = op->reg;
read_count++;
}
if (op->access & CS_AC_WRITE) {
regs_write[write_count] = op->reg;
write_count++;
}
break;
case BPF_OP_MEM:
if (op->mem.base != BPF_REG_INVALID) {
regs_read[read_count] = op->mem.base;
read_count++;
}
break;
}
}
sort_and_uniq(regs_read, read_count, regs_read_count);
sort_and_uniq(regs_write, write_count, regs_write_count);
}