Add registers data table

Change the validated primitive parse_register so that it uses the data
table instead
This commit is contained in:
omicron 2025-04-13 15:17:08 +02:00
parent ac14925a0a
commit d59559d327
3 changed files with 176 additions and 15 deletions

92
src/data/registers.c Normal file
View File

@ -0,0 +1,92 @@
#include "registers.h"
register_data_t *const registers[] = {
// Instruction pointer
&(register_data_t){"rip", REG_RIP, OPERAND_SIZE_64},
&(register_data_t){"eip", REG_RIP, OPERAND_SIZE_32},
&(register_data_t){"ip", REG_RIP, OPERAND_SIZE_16},
// 64-bit general purpose registers
&(register_data_t){"rax", REG_A, OPERAND_SIZE_64},
&(register_data_t){"rcx", REG_C, OPERAND_SIZE_64},
&(register_data_t){"rdx", REG_D, OPERAND_SIZE_64},
&(register_data_t){"rbx", REG_B, OPERAND_SIZE_64},
&(register_data_t){"rsp", REG_SP, OPERAND_SIZE_64},
&(register_data_t){"rbp", REG_BP, OPERAND_SIZE_64},
&(register_data_t){"rsi", REG_SI, OPERAND_SIZE_64},
&(register_data_t){"rdi", REG_DI, OPERAND_SIZE_64},
&(register_data_t){"r8", REG_8, OPERAND_SIZE_64},
&(register_data_t){"r9", REG_9, OPERAND_SIZE_64},
&(register_data_t){"r10", REG_10, OPERAND_SIZE_64},
&(register_data_t){"r11", REG_11, OPERAND_SIZE_64},
&(register_data_t){"r12", REG_12, OPERAND_SIZE_64},
&(register_data_t){"r13", REG_13, OPERAND_SIZE_64},
&(register_data_t){"r14", REG_14, OPERAND_SIZE_64},
&(register_data_t){"r15", REG_15, OPERAND_SIZE_64},
// 32-bit general purpose registers
&(register_data_t){"eax", REG_A, OPERAND_SIZE_32},
&(register_data_t){"ecx", REG_C, OPERAND_SIZE_32},
&(register_data_t){"edx", REG_D, OPERAND_SIZE_32},
&(register_data_t){"ebx", REG_B, OPERAND_SIZE_32},
&(register_data_t){"esp", REG_SP, OPERAND_SIZE_32},
&(register_data_t){"ebp", REG_BP, OPERAND_SIZE_32},
&(register_data_t){"esi", REG_SI, OPERAND_SIZE_32},
&(register_data_t){"edi", REG_DI, OPERAND_SIZE_32},
&(register_data_t){"r8d", REG_8, OPERAND_SIZE_32},
&(register_data_t){"r9d", REG_9, OPERAND_SIZE_32},
&(register_data_t){"r10d", REG_10, OPERAND_SIZE_32},
&(register_data_t){"r11d", REG_11, OPERAND_SIZE_32},
&(register_data_t){"r12d", REG_12, OPERAND_SIZE_32},
&(register_data_t){"r13d", REG_13, OPERAND_SIZE_32},
&(register_data_t){"r14d", REG_14, OPERAND_SIZE_32},
&(register_data_t){"r15d", REG_15, OPERAND_SIZE_32},
// 16-bit general purpose registers
&(register_data_t){"ax", REG_A, OPERAND_SIZE_16},
&(register_data_t){"cx", REG_C, OPERAND_SIZE_16},
&(register_data_t){"dx", REG_D, OPERAND_SIZE_16},
&(register_data_t){"bx", REG_B, OPERAND_SIZE_16},
&(register_data_t){"sp", REG_SP, OPERAND_SIZE_16},
&(register_data_t){"bp", REG_BP, OPERAND_SIZE_16},
&(register_data_t){"si", REG_SI, OPERAND_SIZE_16},
&(register_data_t){"di", REG_DI, OPERAND_SIZE_16},
&(register_data_t){"r8w", REG_8, OPERAND_SIZE_16},
&(register_data_t){"r9w", REG_9, OPERAND_SIZE_16},
&(register_data_t){"r10w", REG_10, OPERAND_SIZE_16},
&(register_data_t){"r11w", REG_11, OPERAND_SIZE_16},
&(register_data_t){"r12w", REG_12, OPERAND_SIZE_16},
&(register_data_t){"r13w", REG_13, OPERAND_SIZE_16},
&(register_data_t){"r14w", REG_14, OPERAND_SIZE_16},
&(register_data_t){"r15w", REG_15, OPERAND_SIZE_16},
// 8-bit general purpose registers (low byte)
&(register_data_t){"al", REG_A, OPERAND_SIZE_8 },
&(register_data_t){"cl", REG_C, OPERAND_SIZE_8 },
&(register_data_t){"dl", REG_D, OPERAND_SIZE_8 },
&(register_data_t){"bl", REG_B, OPERAND_SIZE_8 },
&(register_data_t){"spl", REG_SP, OPERAND_SIZE_8 },
&(register_data_t){"bpl", REG_BP, OPERAND_SIZE_8 },
&(register_data_t){"sil", REG_SI, OPERAND_SIZE_8 },
&(register_data_t){"dil", REG_DI, OPERAND_SIZE_8 },
&(register_data_t){"r8b", REG_8, OPERAND_SIZE_8 },
&(register_data_t){"r9b", REG_9, OPERAND_SIZE_8 },
&(register_data_t){"r10b", REG_10, OPERAND_SIZE_8 },
&(register_data_t){"r11b", REG_11, OPERAND_SIZE_8 },
&(register_data_t){"r12b", REG_12, OPERAND_SIZE_8 },
&(register_data_t){"r13b", REG_13, OPERAND_SIZE_8 },
&(register_data_t){"r14b", REG_14, OPERAND_SIZE_8 },
&(register_data_t){"r15b", REG_15, OPERAND_SIZE_8 },
// x87 floating point registers
&(register_data_t){"st0", REG_ST0, OPERAND_SIZE_80},
&(register_data_t){"st1", REG_ST1, OPERAND_SIZE_80},
&(register_data_t){"st2", REG_ST2, OPERAND_SIZE_80},
&(register_data_t){"st3", REG_ST3, OPERAND_SIZE_80},
&(register_data_t){"st4", REG_ST4, OPERAND_SIZE_80},
&(register_data_t){"st5", REG_ST5, OPERAND_SIZE_80},
&(register_data_t){"st6", REG_ST6, OPERAND_SIZE_80},
&(register_data_t){"st7", REG_ST7, OPERAND_SIZE_80},
nullptr,
};

82
src/data/registers.h Normal file
View File

@ -0,0 +1,82 @@
#ifndef INCLUDE_DATA_REGISTERS_H_
#define INCLUDE_DATA_REGISTERS_H_
typedef enum operand_size {
OPERAND_SIZE_INVALID = 0,
OPERAND_SIZE_8 = 1 << 0,
OPERAND_SIZE_16 = 1 << 1,
OPERAND_SIZE_32 = 1 << 2,
OPERAND_SIZE_64 = 1 << 3,
OPERAND_SIZE_80 = 1 << 4,
OPERAND_SIZE_128 = 1 << 5,
OPERAND_SIZE_256 = 1 << 6,
OPERAND_SIZE_512 = 1 << 7,
} operand_size_t;
static inline operand_size_t bits_to_operand_size(int bits) {
switch (bits) {
case 8:
return OPERAND_SIZE_8;
case 16:
return OPERAND_SIZE_16;
case 32:
return OPERAND_SIZE_32;
case 64:
return OPERAND_SIZE_64;
case 80:
return OPERAND_SIZE_80;
case 128:
return OPERAND_SIZE_128;
case 256:
return OPERAND_SIZE_256;
case 512:
return OPERAND_SIZE_512;
default:
return OPERAND_SIZE_INVALID;
}
}
typedef enum register_id {
// Special registers
REG_RIP = -1,
// General purpose registers
REG_A = 0x0000,
REG_C,
REG_D,
REG_B,
REG_SP,
REG_BP,
REG_SI,
REG_DI,
REG_8,
REG_9,
REG_10,
REG_11,
REG_12,
REG_13,
REG_14,
REG_15,
REG_ST0 = 0x1000,
REG_ST1,
REG_ST2,
REG_ST3,
REG_ST4,
REG_ST5,
REG_ST6,
REG_ST7,
} register_id_t;
typedef struct register_data {
const char *name;
register_id_t id;
operand_size_t size;
} register_data_t;
extern register_data_t *const registers[];
#endif // INCLUDE_DATA_REGISTERS_H_

View File

@ -1,5 +1,6 @@
#include "primitives.h" #include "primitives.h"
#include "../ast.h" #include "../ast.h"
#include "../data/registers.h"
#include <string.h> #include <string.h>
parse_result_t parse_identifier(tokenlist_entry_t *current) { parse_result_t parse_identifier(tokenlist_entry_t *current) {
@ -71,23 +72,9 @@ parse_result_t parse_label_reference(tokenlist_entry_t *current) {
nullptr); nullptr);
} }
const char *registers[] = {
// 64-bit registers
"rax", "rcx", "rdx", "rbx", "rsp", "rbp", "rsi", "rdi", "r8", "r9", "r10",
"r11", "r12", "r13", "r14", "r15",
// 32-bit registers
"eax", "ecx", "edx", "ebx", "esp", "ebp", "esi", "edi", "r8d", "r9d",
"r10d", "r11d", "r12d", "r13d", "r14d", "r15d",
// 16-bit registers
"ax", "cx", "dx", "bx", "sp", "bp", "si", "di", "r8w", "r9w", "r10w",
"r11w", "r12w", "r13w", "r14w", "r15w",
// 8-bit low registers
"al", "cl", "dl", "bl", "spl", "bpl", "sil", "dil", "r8b", "r9b", "r10b",
"r11b", "r12b", "r13b", "r14b", "r15b", nullptr};
bool is_register_token(lexer_token_t *token) { bool is_register_token(lexer_token_t *token) {
for (size_t i = 0; registers[i] != nullptr; ++i) for (size_t i = 0; registers[i] != nullptr; ++i)
if (strcmp(token->value, registers[i]) == 0) if (strcmp(token->value, registers[i]->name) == 0)
return true; return true;
return false; return false;
} }