#include "asmmon.h"
#define AM_ADDR (AM_ZM | AM_ZMX | AM_ZMY | \
AM_IND | AM_INDX | AM_INDY | \
AM_ABS | AM_ABX | AM_ABY | \
AM_AIND | AM_AINDX | AM_AINDY | \
AM_INDX2 | AM_ZM2)
static const uint64_t mem_size = 0x04000000;
token *tok_global;
uint8_t isexpr(uint8_t type, uint8_t dbg) {
switch (type) {
case EXPR_PLUS:
case EXPR_MINUS:
case EXPR_LOW:
case EXPR_HIGH:
case EXPR_OR:
case EXPR_LSHFT:
case EXPR_RSHFT:
return 1;
default:
return 0;
}
}
uint8_t get_rs(token *t, uint8_t dbg) {
if (t->id == TOK_RS) {
return t->type;
} else {
return 0xFF;
}
}
uint8_t get_of(token *t, uint8_t dbg) {
if (t->id == TOK_OF) {
return t->type;
} else {
return 0xFF;
}
}
uint8_t get_ind(uint8_t mne, uint8_t am, uint8_t dbg) {
uint8_t base_idx = 0;
uint8_t offset = 0;
switch (mne) {
case CMP: base_idx = CMP_IND; break;
case CPB: base_idx = CPB_IND; break;
case JMP: base_idx = JMP_IND; break;
case JSR: base_idx = JSR_IND; break;
case LDA: base_idx = LDA_IND; break;
case LDB: base_idx = LDB_IND; break;
case LDX: base_idx = LDX_IND; break;
case LDY: base_idx = LDY_IND; break;
case STA: base_idx = STA_IND; break;
case STB: base_idx = STB_IND; break;
case STX: base_idx = STX_IND; break;
case STY: base_idx = STY_IND; break;
}
switch (am) {
case IND : offset = 0; break;
case INDY: offset += 1; break;
case INDX: offset += 2; break;
}
return base_idx + offset;
}
uint8_t get_eind(uint8_t mne, uint8_t dbg) {
switch (mne) {
case DEC: return DEC_EIND;
case INC: return INC_EIND;
case STY: return STY_EIND;
case STA: return STA_EIND;
case STB: return STB_EIND;
case LDX: return LDX_EIND;
case STX: return STX_EIND;
case CPB: return CPB_EIND;
case CPX: return CPX_EIND;
case CPY: return CPY_EIND;
}
return 0xFF;
}
uint8_t get_ext_ortho(uint8_t mne, uint8_t dbg) {
switch (mne) {
case LEA: return OP_LEA;
case PEA: return OP_PEA;
case ADD: return OP_ADD;
case SUB: return OP_SUB;
case NOT: return OP_NOT;
case CLZ: return OP_CLZ;
case CLO: return OP_CLO;
case SWP: return OP_SWP;
case PCN: return OP_PCN;
}
return 0xFF;
}
static void write_value(uint64_t value, uint64_t address, uint8_t size) {
if (address < mem_size) {
size = (size > 7) ? 7 : size;
#if 1
if (size < 7) {
uint64_t mask = (-(uint64_t)1 >> ((7 - size) * 8));
*(uint64_t *)(addr+address) = (*(uint64_t *)(addr+address) & ~mask) | (value & mask);
} else {
*(uint64_t *)(addr+address) = value;
}
#else
#endif
}
}
uint64_t get_val(token *t, uint64_t addr, uint8_t size, uint8_t end_expr, uint8_t stop_comma, uint8_t dbg) {
uint64_t value = 0;
uint64_t tmp_val = 0;
uint8_t type = EXPR_NONE;
uint8_t isstart = 1;
int done = 0;
do {
if (t->id == TOK_EXPR) {
type = t->type;
t = t->next;
}
if (stop_comma && t->subtype == TOK_CSV) {
done = 1;
}
if (stop_comma && t->id == TOK_REG) {
break;
}
switch (t->id) {
case TOK_HEX:
case TOK_DEC:
case TOK_BIN:
case TOK_CHAR: tmp_val = t->qword; t = t->next; break;
case TOK_SYM:
case TOK_LABEL:
for (; t->sym && t->sym->isstruct && t->next && t->next->id == TOK_SYM; t = t->next);
tmp_val = (t->sym) ? t->sym->val : addr;
t = t->next;
break;
default: tmp_val = 0;
}
if (end_expr != 0xFF && type == end_expr) {
break;
}
switch (type) {
case EXPR_PLUS : (isstart) ? (value = tmp_val) : (value += tmp_val); break;
case EXPR_MINUS: (isstart) ? (value = -tmp_val) : (value -= tmp_val); break;
case EXPR_OR : value |= tmp_val; break;
case EXPR_LSHFT: value <<= tmp_val; break;
case EXPR_RSHFT: value >>= tmp_val; break;
case EXPR_LOW :
value = tmp_val;
switch (size) {
default:
case 2 : value &= 0xFFFFFFFF; break;
case 1 : value &= 0x0000FFFF; break;
case 0 : value &= 0x000000FF; break;
}
break;
case EXPR_HIGH :
value = tmp_val;
switch (size) {
default:
case 2 : value >>= 0x20; break;
case 1 : value >>= 0x10; break;
case 0 : value >>= 0x08; break;
}
break;
case EXPR_NONE : value = tmp_val; break;
}
isstart = 0;
if (dbg) {
printf("get_val(): Value: $%"PRIX64", Expression type: $%X, Expression Value: $%"PRIX64".\n", value, type, tmp_val);
}
} while (!done && t && t->id == TOK_EXPR && isexpr(t->type, dbg));
return value;
}
token *skip_expr(token *t, uint8_t end_expr, uint8_t stop_comma, uint8_t dbg) {
int done = 0;
int is_reg = 0;
do {
is_reg = (t && t->next && t->next->id == TOK_REG);
t = (t->id == TOK_EXPR && !is_reg) ? t->next : t;
if (is_reg || (stop_comma && (t->subtype == TOK_CSV))) {
done = 1;
}
switch (t->id) {
case TOK_HEX :
case TOK_DEC :
case TOK_BIN :
case TOK_CHAR :
case TOK_SYM :
case TOK_LABEL: t = t->next; break;
}
if (end_expr != 0xFF && t->id == TOK_EXPR && t->type == end_expr) {
break;
}
} while (!done && t && t->id == TOK_EXPR && isexpr(t->type, dbg));
return t;
}
uint8_t get_directivesize(uint8_t type, uint8_t dbg) {
switch (type) {
case DIR_QWORD: return 3;
case DIR_DWORD: return 2;
case DIR_WORD : return 1;
case DIR_BYTE : return 0;
}
return 0;
}
uint16_t handle_struct(line **ln, uint64_t address, uint16_t offset, uint8_t dbg) {
uint8_t is_struct = 0;
uint8_t done = 0;
uint8_t ismember = 0;
uint16_t size = 0;
uint16_t member_size = 0;
line *l = *ln;
symbol *strct = NULL;
token *tok = l->tok;
for (uint8_t found = 0; tok && !found; tok = tok->next) {
switch (tok->id) {
case TOK_DIR:
is_struct = (tok->type == DIR_STRUCT);
found = (tok->type == DIR_STRUCT || tok->type == DIR_UNION);
break;
case TOK_STRUCT: is_struct = 1;
case TOK_UNION : found = 1; break;
}
}
if (tok != NULL) {
strct = tok->sym;
}
if (l && l->next) {
l = l->next;
}
for (; l && !done; l = l->next) {
token *t = l->tok;
token *start = t;
symbol *member;
for (; t && !done; t = t->next) {
switch (t->id) {
case TOK_MEMBER: ismember = 1; member = t->sym; break;
case TOK_DIR :
ismember = (t->type == DIR_STRUCT || t->type == DIR_UNION) ? 1 : ismember;
done = ((is_struct && t->type == DIR_ENDSTRUCT) || (!is_struct && t->type == DIR_ENDUNION));
if (!done && ismember) {
switch (t->type) {
case DIR_BYTE : member_size = 1; break;
case DIR_WORD : member_size = 2; break;
case DIR_DWORD : member_size = 4; break;
case DIR_QWORD : member_size = 8; break;
case DIR_UNION :
case DIR_STRUCT: member_size = handle_struct(&l, address, offset, dbg); break;
case DIR_RES : member_size = get_val(t, address, 3, 0xFF, 0, dbg); t = skip_expr(t, 0xFF, 0, dbg); break;
}
if (member && t->type != DIR_UNION && t->type != DIR_STRUCT) {
member->val = offset;
}
if (is_struct) {
size += member_size;
offset += member_size;
} else if (size < member_size) {
size = member_size;
}
}
ismember = 0;
break;
}
}
if (done) {
break;
}
}
*ln = l;
if (strct != NULL) {
strct->val = size;
}
return size;
}
uint64_t handle_directive(token *t, bytecount *bc, uint8_t isasm, uint64_t address, uint8_t dbg) {
union reg val;
uint8_t c = 0;
uint8_t tmp = 0;
uint8_t type = t->type;
uint64_t tmpaddr = address;
t = t->next;
for (; t; t = t->next) {
tmp = 0;
switch (t->id) {
case TOK_HEX:
case TOK_DEC:
case TOK_BIN:
case TOK_CHAR:
case TOK_SYM:
case TOK_LABEL:
val.u64 = get_val(t, tmpaddr, get_directivesize(type, dbg), 0xFF, 0, dbg);
switch (type) {
case DIR_QWORD: tmp = 8; break;
case DIR_DWORD: tmp = 4; break;
case DIR_WORD : tmp = 2; break;
case DIR_BYTE : tmp = 1; break;
}
write_value(val.u64, tmpaddr, tmp-1);
tmpaddr += tmp;
bc->datasize += tmp;
if (t->next && t->next->id == TOK_EXPR && isexpr(t->next->type, dbg)) {
t = skip_expr(t, 0xFF, 0, dbg);
}
break;
case TOK_STRING:
if (type == DIR_BYTE) {
for (uint16_t k = 0; t->str[k] != '\0'; k++) {
switch (t->str[k]) {
case '\\':
switch (t->str[k+1]) {
case 'n' : c = '\n'; break;
case 'r' : c = '\r'; break;
case 't' : c = '\t'; break;
case '\"': c = '\"'; break;
case '\'': c = '\''; break;
case '\\': c = '\\'; break;
case '0' : c = '\0'; break;
}
k++;
break;
default: c = t->str[k]; break;
}
if (isasm) {
addr[tmpaddr] = c;
}
tmpaddr++;
bc->datasize++;
}
if (isasm) {
addr[tmpaddr] = '\0';
}
tmpaddr++;
bc->datasize++;
}
break;
}
if (t == NULL) {
break;
}
}
return tmpaddr;
}
static uint8_t write_inst(uint8_t prefix, uint8_t ext_prefix, uint8_t opcode, operand *op, uint64_t address, uint8_t *op_size, uint8_t isasm, uint8_t dbg) {
uint8_t inst_size = 0;
uint8_t op_ins_size[2];
union reg ins;
union reg op_ins[2];
memset(op_ins_size, 0, sizeof(op_ins_size));
memset(op_ins, 0, sizeof(op_ins_size));
if (prefix & 3) {
ins.u8[inst_size++] = prefix;
}
if ((ext_prefix & 0x0D) == 0x0D) {
ins.u8[inst_size++] = ext_prefix;
}
ins.u8[inst_size++] = opcode;
if (ext_prefix == 0x1D) {
for (int i = 0; i < 2 && op[i].type != 0xFF; i++) {
ins.u8[inst_size] |= (op[i].id << (!i*4));
}
inst_size++;
for (int i = 0; i < 2 && op[i].type != 0xFF; i++) {
int is_sib = 0;
if (op[i].type == 1) {
int reg = 0;
switch (op[i].id) {
case MEM_SIB : op_ins[i].u8[op_ins_size[i]++] = op[i].scale;
case MEM_ABSR :
case MEM_ARIND:
case MEM_AINDR:
case MEM_ZMR :
case MEM_ZRIND:
case MEM_ZINDR:
case MEM_RIND :
op_ins[i].u8[op_ins_size[i]] = op[i].rind[reg];
reg = (op[i].rind[1] != 0xFF);
op_ins[i].u8[op_ins_size[i]++] |= op[i].rind[reg] << 4;
break;
}
}
}
}
if (isasm) {
if (dbg) {
printf("$%04"PRIX64":\t", address);
for (int i = 0; i < inst_size; i++) {
printf("%02X", ins.u8[i]);
if (i < inst_size-1) {
putchar(' ');
}
}
}
write_value(ins.u64, address, inst_size-1);
}
for (int i = 0; i < 2 && op[i].type != 0xFF; i++) {
if (op[i].type == 1) {
if (ext_prefix == 0x1D && op_ins_size[i]) {
if (isasm) {
if (dbg) {
putchar(' ');
for (int j = 0; j < op_ins_size[i]; j++) {
printf("%02X", op_ins[i].u8[j]);
if (j < op_ins_size[i]-1) {
putchar(' ');
}
}
}
write_value(op_ins[i].u64, address+inst_size, op_ins_size[i]-1);
}
inst_size += op_ins_size[i];
}
if (op_size[i] && (ext_prefix != 0x1D || (ext_prefix == 0x1D && op[i].id != MEM_SIB && op[i].id != MEM_RIND))) {
if (isasm) {
if (dbg) {
uint8_t *tmp = (uint8_t *)&op[i].value;
putchar(' ');
for (int j = 0; j < op_size[i]; j++) {
printf("%02X", tmp[j]);
if (j < op_size[i]-1) {
putchar(' ');
}
}
}
write_value(op[i].value, address+inst_size, op_size[i]-1);
}
inst_size += op_size[i];
}
}
}
if (isasm && dbg) {
putchar('\n');
}
return inst_size;
}
token *get_operands(token *t, operand *op, uint64_t address, uint8_t rs, uint8_t dbg) {
uint8_t op_type;
uint8_t op_inst;
int i = 0;
int old_i = -1;
uint8_t expr_type = 0xFF;
uint8_t stop_comma = 0;
uint64_t value = 0;
uint8_t reg = 0;
uint8_t got_value = 0;
uint8_t is_sib = 0;
uint8_t brack_done = 0;
uint8_t is_comma = 0;
token *tmp = t;
for (; t && i < 2; t = t->next) {
if (t->subtype == TOK_IND) {
brack_done = 1;
}
switch (t->id) {
case TOK_HEX :
case TOK_DEC :
case TOK_BIN :
case TOK_SYM :
case TOK_CHAR :
case TOK_LABEL:
if (!got_value) {
expr_type = (expr_type == 0xFF && t->next && t->next->id == TOK_EXPR) ? t->next->type : expr_type;
switch (expr_type) {
default : stop_comma = 1; break;
case EXPR_MUL : stop_comma = 0; break;
}
is_sib = (!stop_comma && op[i].type && op[i].id == MEM_IND);
value = get_val(t, address, (rs != 0xFF) ? rs : 0, (!stop_comma) ? expr_type : 0xFF, stop_comma, dbg);
op[i].value = (!is_sib) ? value : op[i].value;
got_value = 1;
} else {
op[i].value = (!is_sib) ? value : op[i].value;
got_value = 0;
}
if ((op[i].type == 1 && op[i].id == MEM_RIND) || (!op[i].type)) {
op[i].is_ind = (op[i].type == 1 && op[i].id == MEM_RIND);
op[i].type = 1;
op[i].rind[0] = (op[i].rind[0] == 0xFF) ? op[i].id : op[i].rind[0];
op[i].id = MEM_ZMR;
}
is_comma += (t && (t->subtype == TOK_CSV || (t->next && t->next->subtype == TOK_CSV)));
is_comma = (is_comma >= 2) ? 0 : is_comma;
case TOK_MEM:
is_comma = (t->id == TOK_MEM) ? 0 : is_comma;
if (old_i != i) {
op[i].type = 1;
op[i].id = (t->id == TOK_MEM) ? t->type : 0xFF;
op[i].id = (is_sib) ? MEM_SIB : op[i].id;
op[i].scale = (is_sib) ? value : op[i].scale;
old_i = i;
} else {
if (!op[i].type && !is_sib) {
op[i].type = 1;
op[i].id = MEM_ZMR;
}
}
if (got_value && !is_comma) {
if (t && t->subtype != TOK_CSV) {
t = skip_expr(t, (!stop_comma) ? expr_type : 0xFF, stop_comma, dbg);
}
}
i += is_comma;
if (old_i != i) {
got_value = 0;
}
break;
case TOK_REG:
if (old_i != i) {
op[i].type = 0;
op[i].id = t->type;
old_i = i;
} else {
if (op[i].type == 1) {
if (op[i].id == MEM_IND) {
op[i].id = (!brack_done && got_value) ? MEM_ZRIND : MEM_RIND;
op[i].id = (brack_done && got_value) ? MEM_ZINDR : op[i].id;
brack_done = 0;
} else {
op[i].id = (got_value) ? MEM_ZMR : op[i].id;
}
op[i].rind[reg] = t->type;
reg++;
reg = (reg > 1) ? 0 : reg;
}
}
is_comma += (t && (t->subtype == TOK_CSV || (t->next && t->next->subtype == TOK_CSV)));
is_comma = (is_comma >= 2) ? 0 : is_comma;
i += is_comma;
break;
case TOK_EXPR:
expr_type = t->type;
switch (expr_type) {
default : stop_comma = 1; break;
case EXPR_MUL : stop_comma = 0; break;
}
if (!got_value) {
if (t->next && t->next->id != TOK_REG) {
value = get_val(t, address, (rs != 0xFF) ? rs : 0, (!stop_comma) ? expr_type : 0xFF, stop_comma, dbg);
got_value = 1;
}
} else {
got_value = 0;
}
break;
}
if (!t) {
break;
}
}
return t;
}
uint64_t handle_opcode(token *t, bytecount *bc, uint8_t isasm, uint64_t address, uint8_t dbg) {
union reg val;
uint8_t opsize;
uint8_t id;
uint8_t instr;
uint8_t opcode;
uint8_t ext_prefix = 0;
uint8_t type;
uint32_t am = 0;
uint8_t op_size[2];
uint8_t prefix = 0;
uint8_t rs = 0;
uint8_t of = 0;
uint8_t tmp_prefix = 0;
uint8_t inst_size = 0;
int is_ortho = 0;
val.u64 = 0;
instruction ins;
operand op[2];
memset(op, 0xFF, sizeof(op));
op_size[0] = 0;
op_size[1] = 0;
if (t) {
if (t->id == TOK_OPCODE || t->id == TOK_EXTOP || t->id == TOK_ORTHO) {
id = t->id;
instr = t->byte;
type = t->type;
} else {
return address;
}
opsize = 1;
opcode = 0;
if (t->next) {
rs = get_rs(t->next, dbg);
t = (rs != 0xFF) ? t->next : t;
if (t->next) {
is_ortho = (t->next->id == TOK_OS);
t = (is_ortho) ? t->next : t;
if (t->next) {
of = get_of(t->next, dbg);
t = (of != 0xFF) ? t->next : t;
}
}
}
get_operands(t, op, address, rs, dbg);
if (dbg) {
for (int i = 0; i < 2 && op[i].type != 0xFF; i++) {
printf("%i: op.type: %u, op.id: $%X, op.scale; $%X, op.rind[0]: $%X, op.rind[1]: $%X, op.value: $%"PRIX64"\n", i, op[i].type, op[i].id, op[i].scale, op[i].rind[0], op[i].rind[1], op[i].value);
}
}
uint8_t is_eind = (op[0].type == 1 && op[0].id == MEM_RIND && op[0].rind[0] == REG_E);
uint8_t is_mem = (op[0].type == 1 && op[0].id != MEM_IMM);
uint8_t is_idx = (is_mem && !op[1].type && (op[1].id == REG_X || op[1].id == REG_Y));
of = (is_mem && (op[0].id == MEM_RIND || op[0].id == MEM_ZMR) && op[0].rind[0] == REG_SP) ? 1 : of;
if (rs != 0xFF || of != 0xFF) {
tmp_prefix = (rs != 0xFF) ? (rs << 0) : tmp_prefix;
tmp_prefix |= (of != 0xFF) ? (of << 2) : tmp_prefix;
}
prefix = (tmp_prefix) ? ((tmp_prefix << 4) | 3) : 0;
uint8_t isincdec = (instr == INC || instr == DEC);
uint8_t isimplied = (op[0].type == 0xFF);
switch (id) {
case TOK_OPCODE: ins = inst[instr]; break;
case TOK_EXTOP : ins = ext_inst[instr]; break;
case TOK_ORTHO : ins = ortho_inst[instr]; break;
}
am = ins.am;
if (id == TOK_EXTOP || (id == TOK_OPCODE && is_eind)) {
ext_prefix = 0x0D;
}
if ((am & AM_IMPL) && isimplied) {
type = IMPL;
} else {
if (ins.am & AM_REL) {
type = REL;
}
}
opcode = ins.op;
uint64_t saveaddr = address;
uint64_t max_val = 0;
uint8_t i = 0;
uint8_t j = 1;
uint8_t type2 = 0xFF;
if (!is_ortho) {
is_ortho = (id == TOK_ORTHO);
}
int is_valid = 1;
if (!is_ortho && (is_mem || is_idx)) {
switch (op[0].id) {
case MEM_AINDR:
case MEM_ZINDR:
case MEM_ARIND:
case MEM_ZRIND:
case MEM_SIB : is_valid = 0; break;
case MEM_ZMR :
case MEM_ABSR :
case MEM_RIND :
default :
if (is_mem && !op[1].type && op[1].id != REG_X && op[1].id != REG_Y) {
is_valid = 0;
} else if (op[0].id == MEM_ZMR || op[0].id == MEM_ABSR || op[0].id == MEM_RIND) {
is_valid = (of != 0xFF || op[0].rind[0] == REG_SP);
} else if (is_mem && op[1].type == 1) {
is_valid = 0;
}
break;
}
} else if (is_ortho) {
is_valid = 0;
}
if (type == IMPL && (am & AM_IMPL)) {
if (id == TOK_OPCODE && instr == CPS) {
rs = 0;
}
opcode = opcode;
is_ortho = 0;
} else if (type == REL) {
rs = (rs != 0xFF) ? rs : 0;
op_size[0] = (1 << rs);
uint64_t max_sign = 0;
uint8_t offset = 1;
uint64_t tmp_val;
tmp_val = op[0].value;
offset += (prefix != 0);
tmp_val -= offset+op_size[0];
tmp_val -= address;
switch (rs) {
default: max_sign = (int8_t )(1 << 7); break;
case 1 : max_sign = (int16_t)(1 << 15); break;
case 2 : max_sign = (int32_t)(1 << 31); break;
case 3 : max_sign = (int64_t)((uint64_t)1 << 63); break;
}
if ((int64_t)tmp_val > ~(int64_t)max_sign || (int64_t)tmp_val < (int64_t)max_sign) {
offset += (!rs);
rs += (rs <= 3);
op_size[0] = (1 << rs);
tmp_val = op[i].value;
tmp_val -= offset+op_size[0];
tmp_val -= address;
prefix = ((rs << 4) | 3);
}
op[0].value = tmp_val;
is_ortho = 0;
} else if (id != TOK_ORTHO) {
if (!is_eind && is_valid && (is_mem || is_idx)) {
if (of != 0xFF) {
i = 8;
for (; i <= 64; i += 8, j++) {
max_val |= ((uint64_t)1 << (i-1));
if ((int64_t)op[0].value >= ~(int64_t)(max_val) || (int64_t)op[0].value <= (int64_t)(max_val)) {
opsize = j;
break;
}
}
} else {
for (; i <= 64; i += 8, j++) {
max_val |= (0xFF << i);
if (op[0].value <= max_val) {
opsize = j;
break;
}
}
}
if (is_idx) {
switch (op[1].id) {
case REG_X: type = (op[0].id == MEM_IND || (of == 1 && op[0].id == MEM_ZMR && op[0].is_ind)) ? INDX : ZMX; break;
case REG_Y: type = (op[0].id == MEM_IND || (of == 1 && op[0].id == MEM_ZMR && op[0].is_ind)) ? INDY : ZMY; break;
}
} else {
type = (op[0].id == MEM_IND || (of == 1 && op[0].id == MEM_ZMR && op[0].is_ind)) ? IND : 0xFF;
}
type2 = type;
if (type == 0xFF || (id == TOK_EXTOP && type2 != 0xFF)) {
switch (opsize-1) {
case 0: case 2: case 5: case 3: type = ZM ; break;
case 1: case 4: case 6: case 7: type = ABS; break;
}
}
switch (type2) {
case ZMX : type = (type == ABS) ? ABSX : type2; break;
case ZMY : type = (type == ABS) ? ABSY : type2; break;
case IND : type = (type == ABS) ? AIND : type2; break;
case INDX: type = (type == ABS) ? AINDX : type2; break;
case INDY: type = (type == ABS) ? AINDY : type2; break;
}
if (opsize) {
uint8_t is_abs = 0;
switch (type) {
case ABS :
case ABSX :
case ABSY :
case AIND :
case AINDX:
case AINDY: is_abs = 1; break;
}
if (!is_abs || (type2 != 0xFF && type == type2)) {
switch (opsize) {
case 2: opsize = 3; break;
case 5: opsize = 6; break;
}
}
prefix |= amp[opsize-1];
}
if (am & AM_ADDR) {
switch (type) {
case ZM:
if (am & AM_ZM) {
opcode += 0x04;
} else if (am & AM_ZM2) {
opcode += 0x20;
} else {
is_ortho = 1;
}
break;
case ZMX:
if (am & AM_ZMX) {
opcode += (id == TOK_OPCODE) ? 0x06 : 0x54;
} else {
is_ortho = 1;
}
break;
case ZMY:
if (am & AM_ZMY) {
opcode += 0x14;
} else {
is_ortho = 1;
}
break;
case INDX:
if (am & AM_INDX) {
opcode += (id == TOK_OPCODE) ? 0x16 : 0x94;
break;
} else {
is_ortho = 1;
}
case IND:
case INDY:
if ((id == TOK_OPCODE) && (am & (AM_IND|AM_INDY|AM_INDX2))) {
opcode = ind_ops[get_ind(instr, type, dbg)];
} else if (id = TOK_EXTOP) {
opcode += (type == IND) ? 0x44 : 0x84;
} else {
is_ortho = 1;
}
break;
case ABS:
if (am & AM_ABS) {
opcode += 0x10;
} else {
is_ortho = 1;
}
break;
case ABSX:
if (am & AM_ABX) {
opcode += 0x50;
} else {
is_ortho = 1;
}
break;
case ABSY:
if (am & AM_ABY) {
opcode += 0x00;
} else {
is_ortho = 1;
}
break;
case AIND:
if (am & AM_AIND) {
opcode += 0x40;
} else {
is_ortho = 1;
}
break;
case AINDX:
if (am & AM_AINDX) {
opcode += 0x90;
} else {
is_ortho = 1;
}
break;
case AINDY:
if (am & AM_AINDY) {
opcode += 0x80;
} else {
is_ortho = 1;
}
break;
}
op_size[0] = opsize;
}
} else if (op[1].type == 0xFF) {
if (!op[0].type) {
if (op[0].id == REG_B && (am & AM_BREG)) {
opcode += 0x14;
} else if (op[0].id == REG_A && (am & AM_IMPL)) {
opcode = opcode;
} else {
is_ortho = 1;
}
} else {
if (is_eind && (am & AM_EIND|AM_EIND2)) {
int eind_type = ((am & AM_EIND2) != 0);
switch (eind_type) {
case 0: opcode = (id == TOK_EXTOP) ? opcode+0x14 : opcode+0x10; break;
case 1: opcode = (id == TOK_EXTOP) ? opcode+0x01 : eind_base_ops[get_eind(instr, dbg)]; break;
}
} else if ((op[0].id == MEM_IMM) && (am & AM_IMM)) {
rs = (rs != 0xFF) ? rs : 0;
op_size[0] = (1 << rs);
} else {
is_ortho = 1;
}
}
} else {
is_ortho = 1;
}
}
if (is_ortho) {
ext_prefix = 0x1D;
if (id == TOK_EXTOP) {
opcode = ext_ortho_ops[get_ext_ortho(instr, dbg)];
}
for (int i = 0; i < 2 && op[i].type != 0xFF; i++) {
int i2 = 0;
int j = 0;
(op[i].type == 1) ? (opcode |= (1 << (3+!i))) : (opcode &= ~(1 << (3+!i)));
if (op[i].type == 1 && (am & AM_ORTHO|AM_ORTHO2)) {
switch (op[i].id) {
case MEM_RIND: break;
case MEM_IMM:
rs = (rs != 0xFF) ? rs : 0;
op_size[i] = (1 << rs);
break;
case MEM_ZRIND:
case MEM_ZINDR:
case MEM_ZMR:
case MEM_IND:
default:
if (of != 0xFF) {
max_val = 0;
for (i2 = 8, j = 1; i2 <= 64; i2 += 8, j++) {
max_val |= ((uint64_t)1 << (i2-1));
if ((int64_t)op[i].value >= ~(int64_t)(max_val) || (int64_t)op[i].value <= (int64_t)(max_val)) {
opsize = j;
break;
}
}
} else {
max_val = 0;
for (i2 = 0, j = 1; i2 <= 64; i2 += 8, j++) {
max_val |= (0xFF << i2);
if (op[i].value <= max_val) {
opsize = j;
break;
}
}
}
switch (opsize-1) {
case 0: case 2: case 5: case 3: type = ZM ; break;
case 1: case 4: case 6: case 7: type = ABS; break;
}
switch (op[i].id) {
case MEM_IND : op[i].id = (type == ABS) ? MEM_AIND : op[i].id; break;
case MEM_ZMR : op[i].id = (type == ABS) ? MEM_ABSR : op[i].id; break;
case MEM_ZRIND: op[i].id = (type == ABS) ? MEM_ARIND : op[i].id; break;
case MEM_ZINDR: op[i].id = (type == ABS) ? MEM_AINDR : op[i].id; break;
case 0xFF : op[i].id = (type == ABS) ? MEM_ABS : MEM_ZM; break;
}
if (opsize) {
uint8_t is_abs = (type == ABS);
if (!is_abs) {
switch (opsize) {
case 2: opsize = 3; break;
case 5: opsize = 6; break;
}
}
prefix |= amp[opsize-1];
}
op_size[i] = opsize;
if (isasm ) {
printf("op_size[%i]: %i, opsize: %u\n", i, op_size[i], opsize);
}
break;
}
}
}
}
inst_size = write_inst(prefix, ext_prefix, opcode, op, address, op_size, isasm, isasm);
address += inst_size;
bc->progsize += inst_size;
if (isasm ) {
printf("inst_size: $%X, bc->progsize: $%"PRIX64"\n", inst_size, bc->progsize);
}
}
return address;
}
uint64_t parse_tokens(token *t, line **l, bytecount *bc, uint8_t isasm, uint64_t address, uint8_t dbg) {
for (; t; t = t->next) {
switch (t->id) {
case TOK_DIR:
switch (t->type) {
case DIR_STRUCT:
case DIR_UNION : handle_struct(l, address, 0, dbg); break;
case DIR_RES: t = t->next; address += get_val(t, address, 3, 0xFF, 0, dbg); break;
case DIR_ORG: t = t->next; address = get_val(t, address, 3, 0xFF, 0, dbg); break;
case DIR_BYTE:
case DIR_WORD:
case DIR_DWORD:
case DIR_QWORD: address = handle_directive(t, bc, isasm, address, dbg); break;
}
break;
case TOK_ORTHO :
case TOK_EXTOP :
case TOK_OPCODE: address = handle_opcode(t, bc, isasm, address, dbg); break;
case TOK_COMMENT: break;
}
}
return address;
}
token *make_token(uint8_t id, uint8_t type, uint8_t space, uint8_t tab, uint64_t value, char *str, symbol *s) {
token *new_tok = malloc(sizeof(token));
(last_tok) ? (last_tok->next = new_tok) : (tokens = new_tok);
new_tok->id = id;
new_tok->type = type;
new_tok->subtype = 0xFF;
new_tok->tab = tab;
new_tok->space = space;
new_tok->subtab = 0;
new_tok->subspace = 0;
new_tok->digits = 0;
if (s) {
new_tok->sym = s;
} else if (str[0]) {
new_tok->str = str;
} else {
new_tok->qword = value;
}
new_tok->next = NULL;
last_tok = new_tok;
return new_tok;
}
void assemble(line *ln, bytecount *bc, uint8_t dbg) {
uint64_t address = 0;
line *l = ln;
for (int i = 0; i < 2; i++) {
l = ln;
for (; l; l = l->next) {
l->addr = address;
token *t = l->tok;
address = parse_tokens(t, &l, bc, 0, address, dbg);
}
l = ln;
for (; l; l = l->next) {
address = l->addr;
token *t = l->tok;
for (; t; t = t->next) {
if (t->id == TOK_LABEL && t->sym->val != address) {
t->sym->val = l->addr;
}
}
}
}
l = ln;
bc->progsize = 0;
bc->datasize = 0;
for (; l; l = l->next) {
address = parse_tokens(l->tok, &l, bc, 1, address, dbg);
}
}
static void find_dupsym() {
symbol *root = symbols;
symbol *s = symbols;
for (; s; s = s->next) {
root = symbols;
for (int i = 0; root; root = root->next) {
if (root == s) {
i++;
}
if (i > 1) {
printf("Found duplicate symbol, s->name: %s, root->name: %s\n", s->name, root->name);
i = 0;
}
}
}
}
static symbol *find_fixup(token *t) {
fixup* f = fixups;
for (; f && t != f->t; f = f->next);
return (f && t == f->t) ? f->s : NULL;
}
static void print_symval(symbol *s) {
if (s) {
if (s->down) {
print_symval(s->down);
}
if (s->name) {
printf("s->name: %s, s->val: $%"PRIX64"\n", s->name, s->val);
}
print_symval(s->next);
}
}
static void print_symtree(symbol *s, int depth) {
if (s) {
if (s->name != NULL) {
for (int i = depth; i; i--) {
printf("|%s", (i > 1) ? " " : "--");
}
printf("%s: $%"PRIX64"\n", s->name, s->val);
}
if (s->down != NULL) {
print_symtree(s->down, depth+1);
}
print_symtree(s->next, depth);
}
}
void fix_symtree(line *l) {
symbol *s = symbols;
symbol *cur_sym = NULL;
symbol *sym_struct = NULL;
symbols = NULL;
last_sym = NULL;
int islocal = 0;
int isanon = 0;
int is_struct = 0;
int is_structsym = 0;
for (; l; l = l->next) {
token *t = l->tok;
token *lt = NULL;
for (; t; t = t->next) {
int ismember = (t->id == TOK_MEMBER);
switch (t->id) {
case TOK_STRUCT:
case TOK_UNION : islocal = !(is_struct == 1 && lt && lt->id == TOK_DIR);
case TOK_SYM :
if (t->id == TOK_SYM && t != l->tok) {
break;
}
case TOK_MEMBER:
case TOK_LABEL :
if (symbols) {
(!islocal && s && !s->up) ? (last_sym = s) : (last_loc = s);
}
if (((t->type == 1 || ismember) && !islocal) || (islocal && ismember && is_structsym)) {
is_structsym = 0;
last_loc = NULL;
islocal = 1;
cur_sym = s;
s->down = t->sym;
s->down->up = s;
s = s->down;
if (s) {
s->next = NULL;
s->prev = NULL;
s->down = NULL;
}
locals = s;
} else if ((islocal || t->type == 0)) {
if (t->type == 0 && !is_struct && islocal && !ismember) {
islocal = 0;
if (s) {
s->up->down = locals;
s = s->up;
}
}
symbol *tmp = s;
s = t->sym;
if (s) {
s->prev = (tmp && tmp != s) ? tmp : NULL;
s->up = (s->prev) ? s->prev->up : s->up;
}
if (s && s->next) {
s->next = NULL;
}
}
if (!islocal) {
last_loc = NULL;
(last_sym) ? (last_sym->next = s) : (symbols = s);
cur_sym = s;
if (last_sym) {
last_sym->next->prev = last_sym;
last_sym->next->up = last_sym->up;
last_sym->next->down = NULL;
}
} else {
(last_loc) ? (last_loc->next = s) : (locals = s);
if (last_loc) {
last_loc->next->prev = last_loc;
last_loc->next->up = last_loc->up;
last_loc->next->down = NULL;
} else {
locals->prev = NULL;
locals->down = NULL;
}
}
break;
case TOK_DIR:
if (t->type == DIR_STRUCT || t->type == DIR_UNION) {
is_struct++;
is_structsym = (t->next && (t->next->id == TOK_STRUCT || t->next->id == TOK_UNION));
if ((!is_structsym) || (isanon && is_structsym)) {
isanon++;
}
}
break;
}
}
}
}
static inline void free_tokens(token *t) {
token *tok;
if (t != NULL) {
tok = t;
t = t->next;
free(tok);
free_tokens(t);
}
}
void free_lines(line *l) {
line *ln;
if (l != NULL) {
free_tokens(l->tok);
ln = l;
l = l->next;
free(ln);
free_lines(l);
}
}
static void free_symbols(symbol *s) {
symbol *sym;
sym = s;
if (sym != NULL) {
if (s && s->down) {
free_symbols(s->down);
}
if (sym->name != NULL) {
free(sym->name);
sym->name = NULL;
}
s = s->next;
free(sym);
sym = NULL;
free_symbols(s);
}
}
static inline void free_fixups(fixup *f) {
fixup *fix;
if (f != NULL) {
fix = f;
f = f->next;
free(fix);
fix = NULL;
free_fixups(f);
}
}
uint64_t get_tokmem(token *t) {
uint64_t i = 0;
for (; t; t = t->next, i++);
return i*sizeof(token);
}
void get_linemem(line *l) {
uint64_t i = 0;
uint64_t j = 0;
for (; l; j += get_tokmem(l->tok), l = l->next, i++);
printf("Bytes per line: %"PRIu64", Bytes per token: %"PRIu64", Total size of line table in bytes: %"PRIu64"\n", sizeof(line), sizeof(token), j+(i*sizeof(line)));
}
void cleanup() {
uint16_t i;
if (lines) {
free_lines(lines);
lines = NULL;
}
if (symbols) {
free_symbols(symbols);
symbols = NULL;
}
if (fixups) {
free_fixups(fixups);
fixups = NULL;
}
while (i < stridx || i < comidx) {
if (i < stridx && string[i]) {
free(string[i]);
string[i] = NULL;
}
if (i < comidx && comment[i]) {
free(comment[i]);
comment[i] = NULL;
}
i++;
}
}