Merge branch 'asm'

.gitignore

@@ -1,2 +1,4 @@
 *.o
+*.a
 m16vm
+m16as

Makefile

@@ -17,16 +17,29 @@
 #    Outputs the instructions executed in a human-readable format.
 #
 CC = gcc
-CFLAGS = -Iinclude -DMEM_SIZE=32 -DM16_DEBUG_MEM
+CFLAGS = -g -Ilib/include -DMEM_SIZE=32 -DM16_DEBUG_MEM
 LD = $(CC)
 
-VM = m16vm
+PROGRAMS = m16vm m16as
 
-$(VM) : src/vm.o src/cpu.o src/mm.o src/instr_decode.o src/syscall.o src/program.o
+all: $(PROGRAMS)
-	$(LD) $(LDFLAGS)-o $@ $^
+
+m16vm : src/vm.o src/cpu.o src/mm.o src/instr_decode.o src/syscall.o src/program.o
+	$(LD) $(LDFLAGS) -o $@ $^
+
+m16as : src/as/as.o src/as/parser.o src/as/lexer.o \
+	src/as/codegen.o src/as/error.o src/as/symtab.o \
+	src/as/ast.o lib/libm16.a
+	$(LD) $(LDFLAGS) -o $@ $^
+
+lib/libm16.a : lib/src/vector.o
+	$(AR) rcs $@ $^
 
 clean :
 	$(RM) src/*.o
+	$(RM) src/as/*.o
+	$(RM) lib/*.o
+	$(RM) lib/*.a
 
 distclean : clean
-	$(RM) $(VM)
+	$(RM) $(PROGRAMS)

asm/hello_world.as

@@ -0,0 +1,33 @@
+
+; -- Store "Hello World" string in memory.
+movl 	$0x1, 72 ; H
+sw	$0x0, $0x1, 0
+movl 	$0x1, 69 ; E
+sw	$0x0, $0x1, 1
+movl 	$0x1, 76 ; L
+sw	$0x0, $0x1, 2
+sw	$0x0, $0x1, 3
+movl 	$0x1, 79 ; O
+sw	$0x0, $0x1, 4
+movl 	$0x1, 32 ; Space
+sw	$0x0, $0x1, 5
+movl 	$0x1, 87 ; W
+sw	$0x0, $0x1, 6
+movl 	$0x1, 79 ; O
+sw	$0x0, $0x1, 7
+movl 	$0x1, 82 ; R
+sw	$0x0, $0x1, 8
+movl 	$0x1, 76 ; L
+sw	$0x0, $0x1, 9
+movl	$0x1, 68 ; D
+sw	$0x0, $0x1, 10
+
+; -- setup print loop.
+movl 	$0x1, 1		; Load 1 in R1 (used for increment the counter)
+_start:
+ld	$0xF, $0x0, 0		; Load memory address stored in R0 into R15
+int	$0xA, 2			; Print character
+add	$0x0, $0x0, $0x1	; Add 1 (R1) to counter (R0)
+beq	$0xF, $0x2, 1		; Branch to "noop" (skipping next instruction) if R15 = 0 (R2 holds 0)
+jmp     _start			; jump back to "_start" label
+noop

asm/mov_test.as

@@ -0,0 +1,15 @@
+
+; mov h/l test.
+; Using 2 instructions to store a 16-bit words is abit tricky to write code for
+
+; Storing 32767 = (2^15) - 1 (highest value in 2's complement 16-bit).
+; MSB (signed flag) = 0, rest 1.
+; H [0111 1111] L [1111 1111]
+movl	$0x0, -1
+movh    $0x0, 127
+
+; Storing -32768 = (2^15) (highest value in 2's complement 16-bit).
+; MSB (signed flag) = 1, rest 0.
+; H [1000 0000] L [0000 0000]
+movl	$0x1, 0
+movh    $0x1, -128

asm/test.as

@@ -0,0 +1,15 @@
+; NOTE: This is just to check the syntax.
+; No logic behind any of the instructions (will prob crash)
+movl 	$0, 3
+movl 	$5, 6500
+; Comment
+noop    ; Comment
+start0: add  	$3, $0, $1
+sw	$2, $3, $2
+ld	$5, $0, $3
+beq	$0, $1, 1
+beq	$0, $1, -25
+beq	$0, $1, -4000
+jr	$5, 0
+int 	$0,    1
+jmp	5

lib/include/instr.h

@@ -33,6 +33,7 @@
 #define OP_JMP  	7
 #define OP_JR   	8
 #define OP_INT 		15
+#define OP_NONE		0xFF
 
 /* Register type */
 struct instr_R {

lib/include/vector.h

@@ -0,0 +1,41 @@
+/* vector.h
+ *
+ *   Copyright (C) 2018   Henrik Hautakoski <henrik@fiktivkod.org>
+ *
+ *   This program 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 2 of the License, or
+ *   (at your option) any later version.
+ *
+ *   This program 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 this program; if not, write to the Free Software
+ *   Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
+ *   MA 02110-1301, USA.
+ */
+#ifndef VECTOR_H
+#define VECTOR_H
+
+#include <stddef.h>
+#include <stdint.h>
+
+struct vector  {
+	size_t blk_sz;
+	size_t alloc;
+	size_t size;
+	void  *base;
+};
+
+#define VECTOR_INIT(block_size) { block_size, 0, 0, NULL }
+
+void vector_init(struct vector* v, size_t block_size);
+
+void vector_destory(struct vector* v);
+
+void vector_append(struct vector* v, const void *ptr, size_t n);
+
+#endif /* VECTOR_H */

lib/src/vector.c

@@ -0,0 +1,54 @@
+/* vector.c
+ *
+ *   Copyright (C) 2018   Henrik Hautakoski <henrik@fiktivkod.org>
+ *
+ *   This program 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 2 of the License, or
+ *   (at your option) any later version.
+ *
+ *   This program 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 this program; if not, write to the Free Software
+ *   Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
+ *   MA 02110-1301, USA.
+ */
+#include <string.h>
+#include <stdlib.h>
+#include <vector.h>
+
+static void ensure_size(struct vector* v, size_t n) {
+
+	size_t old_alloc = v->alloc;
+	while(v->alloc < v->size + n)
+		v->alloc += v->blk_sz;
+
+	if (old_alloc != v->alloc)
+		v->base = realloc(v->base, v->alloc);
+}
+
+void vector_init(struct vector* v, size_t block_size) {
+
+	v->blk_sz = block_size > 0 ? block_size : 64;
+	v->alloc  = 0;
+	v->size   = 0;
+	v->base   = NULL;
+}
+
+void vector_destory(struct vector *v) {
+
+	if (v->base)
+		free(v->base);
+}
+
+void vector_append(struct vector *v, const void *ptr, size_t n) {
+
+	ensure_size(v, n);
+
+	memcpy(v->base + v->size, ptr, n);
+	v->size += n;
+}

src/as/as.c

@@ -0,0 +1,60 @@
+/* as.c
+ *
+ *   Copyright (C) 2012   Henrik Hautakoski <henrik@fiktivkod.org>
+ *
+ *   This program 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 2 of the License, or
+ *   (at your option) any later version.
+ *
+ *   This program 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 this program; if not, write to the Free Software
+ *   Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
+ *   MA 02110-1301, USA.
+ */
+#include <stdio.h>
+#include <stdlib.h>
+#include "parser.h"
+
+int usage(char *program) {
+
+	fprintf(stderr, "Usage: %s <inputfile> [ <outputfile ]\n", program);
+	return -1;
+}
+
+int main(int argc, char **argv) {
+
+	FILE *fd_in;
+	FILE *fd_out = stdout;
+
+	if (argc < 2)
+		return usage(argv[0]);
+
+	fd_in = fopen(argv[1], "r");
+	if (fd_in == NULL) {
+		perror("Could not open input file");
+		return -1;
+	}
+
+
+	// If we have a output file.
+	if (argc > 2) {
+		fd_out = fopen(argv[2], "w");
+		if (fd_out == NULL) {
+			perror("Could not open output file");
+			fclose(fd_in);
+			return -1;
+		}
+	}
+
+	parse(fd_in, fd_out);
+
+	fclose(fd_in);
+	fclose(fd_out);
+	return 0;
+}

src/as/ast.c

@@ -0,0 +1,95 @@
+
+#include <string.h>
+#include <stdlib.h>
+#include <vector.h>
+#include <stdio.h>
+#include "ast.h"
+
+void ast_init(struct ast* ast) {
+
+	vector_init(&ast->instr, 16 * sizeof(struct ast));
+	vector_init(&ast->labels, 16 * sizeof(char**));
+	ast->symbols = symtab_init();
+}
+
+void ast_free(struct ast* ast) {
+
+	int i;
+
+	// Free all label strings
+	for(int i = 0; i < ast->labels.size; i += sizeof(char**)) {
+		char *ptr = *((char**) ast->labels.base + i);
+		free(ptr);
+	}
+	vector_destory(&ast->labels);
+	vector_destory(&ast->instr);
+	symtab_free(ast->symbols);
+}
+
+void ast_instr(struct ast* ast, uint8_t opcode) {
+
+	struct ast_instr instr = { 0 };
+
+	instr.opcode = opcode;
+	vector_append(&ast->instr, &instr, sizeof(struct ast_instr));
+}
+
+void ast_instr_operand(struct ast* ast, enum ast_datatype type, void* value) {
+
+	struct ast_instr *instr;
+
+	if (ast->instr.size < 1)
+		return;
+
+	// Fetch latest instruction.
+	instr = ast->instr.base + (ast->instr.size - sizeof(struct ast_instr));
+
+	if (instr->n_operands < 3) {
+		struct ast_instr_operand *n = instr->operands + instr->n_operands;
+
+		n->type = type;
+		if (n->type == DATATYPE_STRING) {
+			n->s = strdup(value);
+			vector_append(&ast->labels, &n->s, sizeof(char**));
+		} else {
+			n->r = ((intptr_t) value) & 0xFF;
+		}
+
+		instr->n_operands++;
+	}
+}
+
+void ast_location(struct ast* ast, const char *label, uint16_t loc) {
+
+	symtab_set(ast->symbols, label, loc);
+}
+
+void ast_print(struct ast* ast) {
+
+	printf("{\n");
+
+	for(int i = 0; i < ast->instr.size; i += sizeof(struct ast_instr)) {
+		struct ast_instr *instr = ast->instr.base + i;
+
+		printf("\topcode: %u", instr->opcode);
+		for(int j = 0; j < instr->n_operands; j++) {
+			struct ast_instr_operand *op = instr->operands + j;
+			printf(", operand%i: ", j);
+			switch(op->type) {
+			case DATATYPE_REGISTER :
+				printf("%u (reg)", op->r);
+				break;
+			case DATATYPE_NUMBER :
+				printf("%i (num)", op->n);
+				break;
+			case DATATYPE_STRING :
+				printf("%s (string)", op->s);
+				break;
+			}
+		}
+		printf("\n");
+	}
+
+	printf("}\n");
+
+}

src/as/ast.h

@@ -0,0 +1,81 @@
+/* ast.h
+ *
+ *   Copyright (C) 2018   Henrik Hautakoski <henrik.hautakoski@gmail.com>
+ *
+ *   This program 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 2 of the License, or
+ *   (at your option) any later version.
+ *
+ *   This program 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 this program; if not, write to the Free Software
+ *   Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
+ *   MA 02110-1301, USA.
+ */
+#ifndef ASM_AST_H
+#define ASM_AST_H
+
+#include <vector.h>
+#include "symtab.h"
+
+// All datatypes in the language.
+enum ast_datatype {
+	DATATYPE_NONE = 0,   // Not an actual type.
+	DATATYPE_REGISTER,   // 4 bit unsigned.
+	DATATYPE_NUMBER,     // 8 bit signed.
+	DATATYPE_STRING
+};
+
+// Struct to represent a single operand to a instruction.
+struct ast_instr_operand {
+	enum ast_datatype type;
+	union {
+		uint8_t     r;
+		int8_t      n;
+		uint16_t    addr;
+		const char *s;
+	};
+};
+
+// Struct to represent a single instruction.
+struct ast_instr {
+	uint8_t opcode;
+	uint8_t n_operands;
+	struct ast_instr_operand operands[3];
+};
+
+// Structure to represent a program.
+struct ast {
+
+	// List of instructions.
+	struct vector instr;
+
+	// List for storing label pointers (so we can free)
+	struct vector labels;
+
+	// Symbol table with label - location mapping.
+	symtab_t *symbols;
+};
+
+void ast_init(struct ast* ast);
+
+void ast_free(struct ast* ast);
+
+// Add an instruction
+void ast_instr(struct ast* ast, uint8_t opcode);
+
+// Add an operand to the latest instruction.
+//void ast_instr_operand(struct ast* ast, const struct ast_instr_operand* op);
+void ast_instr_operand(struct ast* ast, enum ast_datatype type, void* value);
+
+void ast_location(struct ast* ast, const char *label, uint16_t loc);
+
+// For debugging :)
+void ast_print(struct ast* ast);
+
+#endif /* ASM_AST_H */

src/as/codegen.c

@@ -0,0 +1,56 @@
+/* codegen.c
+ *
+ *   Copyright (C) 2018   Henrik Hautakoski <henrik@fiktivkod.org>
+ *
+ *   This program 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 2 of the License, or
+ *   (at your option) any later version.
+ *
+ *   This program 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 this program; if not, write to the Free Software
+ *   Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
+ *   MA 02110-1301, USA.
+ */
+#include "codegen.h"
+
+void codegen_emit(struct ast_instr* instr, symtab_t* symbols, uint8_t* out) {
+
+	out[0] = instr->opcode << 4;
+
+	if (instr->n_operands < 1)
+		// No operands, return
+		return;
+
+	// J-Type
+	if (instr->operands[0].type == DATATYPE_STRING) {
+		uint16_t addr;
+
+		if (symtab_get(symbols, instr->operands[0].s, &addr) >= 0) {
+			out[0] |= (addr >> 8);
+			out[1]  =  addr;
+		}
+	}
+	// R/RI or I
+	else if (instr->operands[0].type == DATATYPE_REGISTER) {
+
+		out[0] |= instr->operands[0].r & 0xF;
+
+		// I
+		if (instr->operands[1].type == DATATYPE_NUMBER) {
+
+			out[1] = instr->operands[1].n;
+		}
+		// R/RI-Type
+		else if (instr->operands[1].type == DATATYPE_REGISTER &&
+			(instr->operands[2].type == DATATYPE_REGISTER || instr->operands[2].type == DATATYPE_NUMBER)) {
+
+			out[1] = (instr->operands[1].r << 4) | (instr->operands[2].r & 0xF);
+		}
+	}
+}

src/as/codegen.h

@@ -0,0 +1,27 @@
+/* codegen.h
+ *
+ *   Copyright (C) 2018   Henrik Hautakoski <henrik.hautakoski@gmail.com>
+ *
+ *   This program 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 2 of the License, or
+ *   (at your option) any later version.
+ *
+ *   This program 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 this program; if not, write to the Free Software
+ *   Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
+ *   MA 02110-1301, USA.
+ */
+#ifndef ASM_CODEGEN_H
+#define ASM_CODEGEN_H
+
+#include "ast.h"
+
+void codegen_emit(struct ast_instr* ast, symtab_t* symbols, uint8_t* out);
+
+#endif /* ASM_CODEGEN_H */

src/as/error.c

@@ -0,0 +1,46 @@
+/* error.c
+ *
+ *   Copyright (C) 2018   Henrik Hautakoski <henrik.hautakoski@gmail.com>
+ *
+ *   This program 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 2 of the License, or
+ *   (at your option) any later version.
+ *
+ *   This program 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 this program; if not, write to the Free Software
+ *   Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
+ *   MA 02110-1301, USA.
+ */
+#include <stdio.h>
+#include <stdarg.h>
+#include "error.h"
+
+void asm_warn(int lineno, const char *fmt, ...) {
+
+	va_list vl;
+
+	fprintf(stderr, "Line %i: Warning: ", lineno);
+	va_start(vl, fmt);
+	vfprintf(stderr, fmt, vl);
+	va_end(vl);
+	fprintf(stderr, "\n");
+}
+
+int asm_error(int lineno, const char *fmt, ...) {
+
+	va_list vl;
+
+	fprintf(stderr, "Line %i: Error: ", lineno);
+	va_start(vl, fmt);
+	vfprintf(stderr, fmt, vl);
+	va_end(vl);
+	fprintf(stderr, "\n");
+
+	return -1;
+}

src/as/error.h

@@ -0,0 +1,27 @@
+/* error.h
+ *
+ *   Copyright (C) 2018   Henrik Hautakoski <henrik.hautakoski@gmail.com>
+ *
+ *   This program 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 2 of the License, or
+ *   (at your option) any later version.
+ *
+ *   This program 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 this program; if not, write to the Free Software
+ *   Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
+ *   MA 02110-1301, USA.
+ */
+#ifndef ASM_ERROR_H
+#define ASM_ERROR_H
+
+void asm_warn(int lineno, const char *fmt, ...);
+
+int asm_error(int lineno, const char *fmt, ...);
+
+#endif /* ASM_ERROR_H */

src/as/lexer.c

@@ -0,0 +1,296 @@
+
+#include <stdio.h>
+#include <string.h>
+#include "error.h"
+#include "lexer.h"
+
+/**
+ * macros for the grammar.
+ */
+
+// Numbers is defined as [0-9]
+#define number(x) ((x) >= '0' && (x) <= '9')
+
+// The first digit can however also contain '-'
+#define first_number(x) (number(x) || (x) == '-' )
+
+// First character in strings can be [a-z][A-Z] or '_'
+#define first_string(x)			\
+	(  ((x) >= 'a' && (x) <= 'z')   \
+	|| ((x) >= 'A' && (x) <= 'Z')	\
+	||  (x) == '_'			)
+
+// All characters after can also include numbers or ':'
+#define string(x) \
+	(first_string(x) || number(x))
+
+#define space(x) ((x) == ' ' || (x) == '\t' || (x) == '\r')
+
+
+struct opcode_ent {
+	char *  name;
+	uint8_t code;
+};
+
+static const struct opcode_ent opcode_table[] = {
+	{ "noop", TOKEN_OPCODE_NOOP },
+	{ "add" , TOKEN_OPCODE_ADD  },
+	{ "movl", TOKEN_OPCODE_MOVL },
+	{ "movh", TOKEN_OPCODE_MOVH },
+	{ "ld"  , TOKEN_OPCODE_LD   },
+	{ "sw"  , TOKEN_OPCODE_SW   },
+	{ "beq" , TOKEN_OPCODE_BEQ  },
+	{ "jmp" , TOKEN_OPCODE_JMP  },
+	{ "jr"  , TOKEN_OPCODE_JR   },
+	{ "int" , TOKEN_OPCODE_INT  },
+	{ NULL  , 0                 },
+};
+
+/**
+ * Helper functions
+ */
+
+static int read_next(struct lexer *lex) {
+
+ 	int c, comment = 0;
+
+ 	while((c = fgetc(lex->fp)) != EOF) {
+
+ 		if (c == '\n')
+			break;
+
+ 		if (comment)
+ 			continue;
+
+ 		if (c == ';') {
+ 			comment = 1;
+ 		} else if (!space(c)) {
+ 			break;
+ 		}
+ 	}
+ 	return c;
+ }
+
+static int read_hex(FILE *fp, int *out) {
+
+	int c, val = 0;
+
+	while((c = fgetc(fp)) != EOF) {
+		char n = 0;
+		if (number(c)) {
+			n = c - '0';
+		}
+		else if (  (c >= 'a' && c <= 'f')
+			|| (c >= 'A' && c <= 'F')) {
+			n = (c % 0x20) + 9;
+		}
+		else {
+			ungetc(c, fp);
+			break;
+		}
+
+		val = (val * 16) + n;
+		if (val > 0xFF)
+			goto overflow;
+	}
+	*out = val;
+	return 0;
+
+overflow:
+	*out = 0xFF;
+	return -1;
+}
+
+static int read_dec(FILE *fp, int neg, int *out) {
+
+	int c, val = 0;
+
+	while((c = fgetc(fp)) != EOF) {
+		if (!number(c)) {
+			ungetc(c, fp);
+			break;
+		}
+		val = (val * 10) + (c - '0');
+
+		// Cool trick here.
+		// because the range is -128 (0x80) to +127 (0x7F)
+		// We can do 0x80 - 1 if it is NOT a negative number.
+		if (val > (0x80 - !neg))
+			goto overflow;
+	}
+
+	*out = neg ? -1 * val : val;
+	return 0;
+
+overflow:
+	*out = neg ? -1 * 0x80 : 0x7F;
+	return -1;
+}
+
+static int read_number(FILE *fp, int *out) {
+
+	int neg = 0, c = fgetc(fp);
+
+	// Check for '0x'.
+	if (c == '0') {
+		c = fgetc(fp);
+		if (c == 'x') {
+			// We have a hexadecimal number.
+			return read_hex(fp, out);
+		}
+		ungetc(c, fp);
+		ungetc('0', fp);
+	}
+	// While we are at it. check for a negative sign.
+	else if (c == '-') {
+		neg = 1;
+	}
+	// We got something else. put it back.
+	else {
+		ungetc(c, fp);
+	}
+
+	return read_dec(fp, neg, out);
+}
+
+static int parse_number(struct lexer *lex) {
+
+	int num;
+
+	if (read_number(lex->fp, &num) < 0)
+		asm_warn(lex->lineno, "Value truncated to %i", num);
+
+	lex->token.value.n = num;
+	return 0;
+}
+
+static int read_string(FILE *fp, char *buf, size_t len) {
+
+	int c, label_decl = 0, i = 0;
+
+	while((c = fgetc(fp)) != EOF && i < len) {
+
+		if (string(c)) {
+			buf[i++] = c;
+		} else {
+			if (c == ':') {
+				label_decl = 1;
+			} else {
+				ungetc(c, fp);
+			}
+			break;
+		}
+	}
+	buf[i] = '\0';
+
+	if (label_decl)
+		return TOKEN_LABEL_DECL;
+
+	for(i = 0; opcode_table[i].name; i++) {
+
+		if (!strcmp(opcode_table[i].name, buf))
+			return opcode_table[i].code;
+	}
+	return TOKEN_LABEL;
+}
+
+/**
+ * Exposed functions
+ */
+
+void lexer_init(struct lexer *lex, FILE *fp) {
+
+	lex->lineno = 1;
+	lex->fp = fp;
+	lex->token.type = TOKEN_EOI;
+}
+
+int lexer_get_next(struct lexer *lex) {
+
+	uint16_t num;
+	int ch = read_next(lex);
+
+	if (lex->token.type == TOKEN_EOL)
+		lex->lineno++;
+
+	switch(ch) {
+	case EOF : lex->token.type = TOKEN_EOI;
+		break;
+	case '\n' :
+		lex->token.type = TOKEN_EOL;
+		break;
+	case ',' : lex->token.type = TOKEN_ARG_SEP;
+		break;
+	case '$' :
+		lex->token.type = TOKEN_REG;
+		if (parse_number(lex) < 0)
+			return -1;
+		break;
+	default:
+		if (first_number(ch)) {
+			ungetc(ch, lex->fp);
+			lex->token.type = TOKEN_NUMBER;
+			if (parse_number(lex) < 0)
+				return -1;
+		} else if (first_string(ch)) {
+			char buf[32];
+			ungetc(ch, lex->fp);
+			lex->token.type = read_string(lex->fp, buf, sizeof(buf));
+			if (lex->token.type == TOKEN_LABEL_DECL || lex->token.type == TOKEN_LABEL)
+				strcpy(lex->token.value.s, buf);
+		} else {
+			return asm_error(lex->lineno, "Invalid character '%c'", ch);
+		}
+	}
+
+	lex->token.lineno = lex->lineno;
+	return 0;
+}
+
+void lexer_print_token(struct token *token) {
+
+	static int lineno = 0;
+
+	if (token->lineno != lineno) {
+		lineno = token->lineno;
+		printf("\n%i: ", lineno);
+	}
+
+	switch(token->type) {
+	case TOKEN_OPCODE_NOOP : printf(" [OP NOOP] ");
+		break;
+	case TOKEN_OPCODE_ADD : printf(" [OP ADD] ");
+		break;
+	case TOKEN_OPCODE_MOVL : printf(" [OP MOVL] ");
+		break;
+	case TOKEN_OPCODE_MOVH : printf(" [OP MOVH] ");
+		break;
+	case TOKEN_OPCODE_LD : printf(" [OP LD] ");
+		break;
+	case TOKEN_OPCODE_SW : printf(" [OP SW] ");
+		break;
+	case TOKEN_OPCODE_BEQ : printf(" [OP BEQ] ");
+		break;
+	case TOKEN_OPCODE_JMP : printf(" [OP JMP] ");
+		break;
+	case TOKEN_OPCODE_JR : printf(" [OP JR] ");
+		break;
+	case TOKEN_OPCODE_INT : printf(" [OP INT] ");
+		break;
+	case TOKEN_LABEL : printf(" [LABEL \"%s\"] ", token->value.s);
+		break;
+	case TOKEN_LABEL_DECL : printf(" [LABEL DECL \"%s\"] ", token->value.s);
+		break;
+	case TOKEN_REG : printf(" [REG %i] ", token->value.n);
+		break;
+	case TOKEN_ARG_SEP : printf(" [SEP] ");
+		break;
+	case TOKEN_NUMBER : printf(" [NUM %i] ", token->value.n);
+		break;
+	case TOKEN_EOI : printf(" [EOI] ");
+		break;
+	case TOKEN_EOL : printf(" [EOL] ");
+		break;
+	default: printf(" [U] ");
+	}
+}

src/as/lexer.h

@@ -0,0 +1,95 @@
+/* lexer.h
+ *
+ *   Copyright (C) 2018   Henrik Hautakoski <henrik@fiktivkod.org>
+ *
+ *   This program 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 2 of the License, or
+ *   (at your option) any later version.
+ *
+ *   This program 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 this program; if not, write to the Free Software
+ *   Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
+ *   MA 02110-1301, USA.
+ */
+#ifndef ASM_LEXER_H
+#define ASM_LEXER_H
+
+#include <stdlib.h>
+#include <stdint.h>
+
+/**
+ * All token types.
+ */
+enum token_type {
+	TOKEN_EOI = -1,
+	TOKEN_EOL,		// Newline
+	TOKEN_OPCODE_NOOP,
+	TOKEN_OPCODE_ADD,
+	TOKEN_OPCODE_MOVL,
+	TOKEN_OPCODE_MOVH,
+	TOKEN_OPCODE_LD,
+	TOKEN_OPCODE_SW,
+	TOKEN_OPCODE_BEQ,
+	TOKEN_OPCODE_JMP,
+	TOKEN_OPCODE_JR,
+	TOKEN_OPCODE_INT,
+	TOKEN_LABEL,
+	TOKEN_LABEL_DECL,
+	TOKEN_REG,
+	TOKEN_NUMBER,
+	TOKEN_ARG_SEP
+};
+
+/**
+ * Token structure.
+ *
+ * Holds information about a single token.
+ */
+struct token {
+	// Line number where the token was extracted from.
+	uint16_t 	lineno;
+
+	enum token_type	type;
+
+	/*
+	 * Token value, depending on type
+	 * this can be a string or unsigned short
+	 */
+	union {
+		int8_t n;
+		char   s[32];
+	} value;
+};
+
+/**
+ * Lexer state
+ */
+struct lexer {
+	uint16_t 	lineno;		// current line number
+	FILE *		fp;		// File being lexed.
+	struct token 	token;		// Current token
+};
+
+/**
+ * Initialize the lexer with a file pointer to the file
+ * that should be lexed.
+ */
+void lexer_init(struct lexer *lex, FILE *fp);
+
+/**
+ * Advance the lexer to the next token.
+ */
+int lexer_get_next(struct lexer *lex);
+
+/**
+ * For debugging, prints the token to standard output.
+ */
+void lexer_print_token(struct token *token);
+
+#endif /* ASM_LEXER_H */

src/as/parser.c

@@ -0,0 +1,263 @@
+/* parser.c
+ *
+ *   Copyright (C) 2012   Henrik Hautakoski <henrik@fiktivkod.org>
+ *
+ *   This program 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 2 of the License, or
+ *   (at your option) any later version.
+ *
+ *   This program 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 this program; if not, write to the Free Software
+ *   Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
+ *   MA 02110-1301, USA.
+ */
+#include <stdint.h>
+#include <stdio.h>
+#include <string.h>
+#include "error.h"
+#include "codegen.h"
+#include "ast.h"
+#include "lexer.h"
+#include "parser.h"
+
+/*
+ * Helper functions/macros for defining parser rules.
+ */
+
+enum number_size {
+	NUMBER_SIZE_U4,
+	NUMBER_SIZE_S4,
+	NUMBER_SIZE_S8,
+};
+
+// match the next token.
+// returns 0 if the token was of the correct type. -1 otherwise
+static int match_type(struct lexer* lex, enum token_type type) {
+
+	lexer_get_next(lex);
+	return lex->token.type == type ? 0 : -1;
+}
+
+static int validate_number(struct lexer* lex, enum number_size size) {
+
+	int8_t n = lex->token.value.n;
+
+	switch(size) {
+	case NUMBER_SIZE_U4 :
+		if (!(n >= 0x0 && n <= 0xF))
+			return asm_error(lex->lineno, "Value out of range %u", (uint8_t) n);
+		break;
+	case NUMBER_SIZE_S4 :
+		if (!(n >= -8 && n < 8))
+			return asm_error(lex->lineno, "Value out of range %i", n);
+		break;
+	case NUMBER_SIZE_S8 :
+		if (!(n >= -128 && n< 128))
+			return asm_error(lex->lineno, "Value out of range %i", n);
+		break;
+	}
+	return 0;
+}
+
+// Same as match_type() but also generates a operand in the AST.
+static int match_operand(struct lexer* lex, enum token_type type, struct ast *ast) {
+
+	if (match_type(lex, type) < 0)
+		return -1;
+
+	if (type == TOKEN_REG) {
+		ast_instr_operand(ast, DATATYPE_REGISTER, lex->token.value.n);
+	} else if (type == TOKEN_NUMBER) {
+		ast_instr_operand(ast, DATATYPE_NUMBER, lex->token.value.n);
+	} else {
+		ast_instr_operand(ast, DATATYPE_STRING, lex->token.value.s);
+	}
+
+	return 1;
+}
+
+/*
+ * Helper macros for matching tokens.
+ */
+#define match_reg(pos, ast) \
+	if (match_operand(lex, TOKEN_REG, ast) < 0 || validate_number(lex, NUMBER_SIZE_U4) < 0) \
+		return asm_error((lex)->lineno, "Expected register at argument %i", pos)
+
+#define match_imm(pos, size, ast)  \
+	if (match_operand(lex, TOKEN_NUMBER, ast) < 0 || validate_number(lex, size) < 0) \
+		return asm_error((lex)->lineno, "Expected number at argument %i", pos)
+
+#define match_arg(pos) \
+	if (match_type(lex, TOKEN_ARG_SEP) < 0) \
+		return asm_error((lex)->lineno, "Expected separator after argument %i", pos)
+
+#define match_end \
+	if (match_type(lex, TOKEN_EOL) < 0) \
+		return asm_error(lex->lineno, "Expected newline")
+
+/*
+ * Functions for matching complete instructions.
+ */
+
+// R-Type (rs : u4, r0 : u4, r1 : u4)
+static int match_typeR(struct lexer* lex, struct ast *ast) {
+
+	match_reg(1, ast); match_arg(1);
+	match_reg(2, ast); match_arg(2);
+	match_reg(3, ast);
+	match_end;
+
+	return 1;
+}
+
+// RI-Type (rs : u4, r0 : u4, offset : s4)
+static int match_typeRI(struct lexer* lex, struct ast *ast) {
+
+	match_reg(1, ast); match_arg(1);
+	match_reg(2, ast); match_arg(2);
+	match_imm(3, NUMBER_SIZE_S4, ast);
+	match_end;
+
+	return 1;
+}
+
+// I-Type (rs : u4, imm : s8)
+static int match_typeI(struct lexer* lex, struct ast *ast) {
+
+	match_reg(1, ast); match_arg(1);
+	match_imm(2, NUMBER_SIZE_S8, ast);
+	match_end;
+
+	return 1;
+}
+
+// J-Type (addr : string)
+static int match_typeJ(struct lexer* lex, struct ast *ast) {
+
+	if (match_operand(lex, TOKEN_LABEL, ast) < 0)
+		return asm_error(lex->lineno, "Expected label at argument 1");
+	match_end;
+	return 1;
+}
+
+static int match_label_decl(struct lexer* lex, struct ast *ast) {
+
+	uint16_t location = ast->instr.size / sizeof(struct ast_instr);
+
+	ast_location(ast, lex->token.value.s, location);
+
+	return 1;
+}
+
+#define opcode_guard(op) \
+	if (op_set == 0) { op_set = 1; ast_instr(ast, op); }
+
+/*
+ * Parse a single line.
+ */
+static int parse_line(struct lexer* lex, struct ast *ast) {
+
+	int op_set = 0;
+
+	if (lexer_get_next(lex) < 0)
+		return -1;
+
+	// Opcode should come first.
+	switch(lex->token.type) {
+	case TOKEN_EOI: return 0;
+	case TOKEN_EOL: break;
+	case TOKEN_OPCODE_NOOP : ast_instr(ast, OP_NOOP);
+		match_end;
+		break;
+	// Type-R
+	case TOKEN_OPCODE_ADD : opcode_guard(OP_ADD);
+		return match_typeR(lex, ast);
+	// Type-I
+	case TOKEN_OPCODE_MOVL : opcode_guard(OP_MOVL);
+	case TOKEN_OPCODE_MOVH : opcode_guard(OP_MOVH);
+	case TOKEN_OPCODE_JR   : opcode_guard(OP_JR);
+	case TOKEN_OPCODE_INT  : opcode_guard(OP_INT);
+		return match_typeI(lex, ast);
+	// Type-RI
+	case TOKEN_OPCODE_LD  : opcode_guard(OP_LW);
+	case TOKEN_OPCODE_SW  : opcode_guard(OP_SW);
+	case TOKEN_OPCODE_BEQ : opcode_guard(OP_BEQ);
+		return match_typeRI(lex, ast);
+	// Type-J
+	case TOKEN_OPCODE_JMP : opcode_guard(OP_JMP);
+		return match_typeJ(lex, ast);
+	case TOKEN_LABEL_DECL :
+		return match_label_decl(lex, ast);
+	default:
+		return asm_error(lex->lineno, "Opcode or label expected");
+	}
+
+	return 1;
+}
+
+// Check the semantics of the program's AST.
+// For now, we only need to check that all
+// referenced labels exist in the symbol table
+static int check_semantics(struct ast* ast) {
+
+	int i;
+
+	// TODO: Need to implement a iterator for vectors.
+	for(i = 0; i < ast->instr.size; i += sizeof(struct ast_instr)) {
+		struct ast_instr *instr = ast->instr.base + i;
+
+		// Only J-Type can have labels.
+		if (instr->opcode == OP_JMP
+			&& instr->operands[0].type == DATATYPE_STRING
+			&& symtab_get(ast->symbols, instr->operands[0].s, NULL) < 0)  {
+
+			return asm_error(0, "Label '%s' is not defined", instr->operands[0].s);
+		}
+	}
+
+	return 0;
+}
+
+/*
+ * Main parser function.
+ */
+int parse(FILE *source_fd, FILE *dest_fd) {
+
+	struct lexer lex;
+	struct ast ast;
+
+	ast_init(&ast);
+	lexer_init(&lex, source_fd);
+
+	// Parse and build AST.
+	for(;;) {
+		int rc = parse_line(&lex, &ast);
+		if (rc < 0)
+			goto done;
+		if (rc == 0)
+			break;
+	}
+
+	if (check_semantics(&ast) < 0)
+		goto done;
+
+	// Code generation
+	for(int i = 0; i < ast.instr.size; i += sizeof(struct ast_instr)) {
+		struct ast_instr *instr = ast.instr.base + i;
+		uint8_t code[2] = { 0 };
+
+		codegen_emit(instr, ast.symbols, &code);
+
+		fwrite(&code, sizeof(code), 1, dest_fd);
+	}
+
+	// Cleanup
+done:	ast_free(&ast);
+	return 0;
+}

src/as/parser.h

@@ -0,0 +1,33 @@
+/* parser.h
+ *
+ *   Copyright (C) 2012   Henrik Hautakoski <henrik@fiktivkod.org>
+ *
+ *   This program 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 2 of the License, or
+ *   (at your option) any later version.
+ *
+ *   This program 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 this program; if not, write to the Free Software
+ *   Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
+ *   MA 02110-1301, USA.
+ */
+#ifndef ASM_PARSER_H
+#define ASM_PARSER_H
+
+#include <instr.h>
+#include <stdio.h>
+
+struct parse_state {
+	struct instr *instr_ptr;
+	unsigned int num;
+};
+
+int parse(FILE *source_fd, FILE *dest_fd);
+
+#endif /* ASM_PARSER_H */

src/as/symtab.c

@@ -0,0 +1,121 @@
+/* symtab.c
+ *
+ *   Copyright (C) 2018   Henrik Hautakoski <henrik.hautakoski@gmail.com>
+ *
+ *   This program 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 2 of the License, or
+ *   (at your option) any later version.
+ *
+ *   This program 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 this program; if not, write to the Free Software
+ *   Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
+ *   MA 02110-1301, USA.
+ */
+#include <stdlib.h>
+#include <string.h>
+#include "symtab.h"
+
+// TODO: Right now, the symtab is implemented as a simple dynamic growing array.
+// This is fine for now, but for lookup speed. A hashtable is more efficient.
+
+// How much memory that should be allocated each time.
+// Note: this should be defined as number of `struct symbol` objects, not bytes.
+#define BLOCK_SIZE 32
+
+// Helper macro to calculate number of bytes the table need.
+#define __M_SIZE(t) \
+	((t)->size * sizeof(struct symbol))
+
+struct symbol {
+	// The label associated with the location.
+	char label[SYMTAB_LABEL_MAXLEN];
+
+	// Location (program adress)
+	uint16_t loc;
+};
+
+// Symbol table structure.
+struct symtab {
+	// Array of symbols.
+	struct symbol *data;
+
+	// Number of symbols in the array.
+	size_t num;
+
+	// Total number of symbols the array can hold.
+	size_t size;
+};
+
+symtab_t* symtab_init() {
+
+	struct symtab *table = malloc(sizeof(struct symtab));
+
+	table->num = 0;
+	table->size = BLOCK_SIZE;
+	table->data = malloc(__M_SIZE(table));
+
+	return table;
+}
+
+void symtab_free(symtab_t *table) {
+
+	if (!table)
+		return;
+
+	if (table->data)
+		free(table->data);
+
+	memset(table, 0, sizeof(struct symtab));
+	free(table);
+}
+
+int symtab_set(symtab_t *table, const char *label, uint16_t loc) {
+
+	// Check if it exist first.
+	int index = symtab_get(table, label, NULL);
+
+	// Entry did not exists. Insert
+	if (index < 0) {
+		struct symbol *sym;
+
+		// We have one more.
+		table->num += 1;
+
+		// Make sure we resize the memory.
+		if (table->num > table->size) {
+			table->size += BLOCK_SIZE;
+			table->data = realloc(table->data, __M_SIZE(table));
+		}
+
+		// insert the symbol at the end.
+		sym = table->data + (table->num - 1);
+		strncpy(sym->label, label, SYMTAB_LABEL_MAXLEN);
+		sym->loc = loc;
+
+		return 0;
+	}
+
+	// Could not insert. return as error.
+	return -1;
+}
+
+int symtab_get(symtab_t *table, const char *label, uint16_t *loc) {
+
+	// Linear search here for simplicity.
+	for(size_t i = 0; i < table->num; i++) {
+		struct symbol *sym = table->data + i;
+
+		if (!strncmp(sym->label, label, SYMTAB_LABEL_MAXLEN)) {
+
+			if (loc) *loc = sym->loc;
+			return i;
+		}
+	}
+	return -1;
+}

src/as/symtab.h

@@ -0,0 +1,37 @@
+/* symtab.h
+ *
+ *   Copyright (C) 2018   Henrik Hautakoski <henrik.hautakoski@gmail.com>
+ *
+ *   This program 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 2 of the License, or
+ *   (at your option) any later version.
+ *
+ *   This program 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 this program; if not, write to the Free Software
+ *   Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
+ *   MA 02110-1301, USA.
+ */
+#ifndef ASM_SYMTAB_H
+#define ASM_SYMTAB_H
+
+#include <stdint.h>
+
+#define SYMTAB_LABEL_MAXLEN 32
+
+typedef struct symtab symtab_t;
+
+symtab_t* symtab_init();
+
+void symtab_free(symtab_t *table);
+
+int symtab_set(symtab_t *table, const char *label, uint16_t loc);
+
+int symtab_get(symtab_t *table, const char *label, uint16_t *loc);
+
+#endif /* ASM_SYMTAB_H */