Add one dimensional arrays

Author: rui314

chibicc.h

@@ -127,17 +127,29 @@ typedef enum {
   TY_INT,
   TY_PTR,
   TY_FUNC,
+  TY_ARRAY,
 } TypeKind;
 
 struct Type {
   TypeKind kind;
-
-  // Pointer
+  int size;      // sizeof() value
+
+  // Pointer-to or array-of type. We intentionally use the same member
+  // to represent pointer/array duality in C.
+  //
+  // In many contexts in which a pointer is expected, we examine this
+  // member instead of "kind" member to determine whether a type is a
+  // pointer or not. That means in many contexts "array of T" is
+  // naturally handled as if it were "pointer to T", as required by
+  // the C spec.
   Type *base;
 
   // Declaration
   Token *name;
 
+  // Array
+  int array_len;
+
   // Function type
   Type *return_ty;
   Type *params;
@@ -150,6 +162,7 @@ bool is_integer(Type *ty);
 Type *copy_type(Type *ty);
 Type *pointer_to(Type *base);
 Type *func_type(Type *return_ty);
+Type *array_of(Type *base, int size);
 void add_type(Node *node);
 
 //

codegen.c

@@ -42,6 +42,27 @@ static void gen_addr(Node *node) {
   error_tok(node->tok, "not an lvalue");
 }
 
+// Load a value from where %rax is pointing to.
+static void load(Type *ty) {
+  if (ty->kind == TY_ARRAY) {
+    // If it is an array, do not attempt to load a value to the
+    // register because in general we can't load an entire array to a
+    // register. As a result, the result of an evaluation of an array
+    // becomes not the array itself but the address of the array.
+    // This is where "array is automatically converted to a pointer to
+    // the first element of the array in C" occurs.
+    return;
+  }
+
+  printf("  mov (%%rax), %%rax\n");
+}
+
+// Store %rax to an address that the stack top is pointing to.
+static void store(void) {
+  pop("%rdi");
+  printf("  mov %%rax, (%%rdi)\n");
+}
+
 // Generate code for a given node.
 static void gen_expr(Node *node) {
   switch (node->kind) {
@@ -54,11 +75,11 @@ static void gen_expr(Node *node) {
     return;
   case ND_VAR:
     gen_addr(node);
-    printf("  mov (%%rax), %%rax\n");
+    load(node->ty);
     return;
   case ND_DEREF:
     gen_expr(node->lhs);
-    printf("  mov (%%rax), %%rax\n");
+    load(node->ty);
     return;
   case ND_ADDR:
     gen_addr(node->lhs);
@@ -67,8 +88,7 @@ static void gen_expr(Node *node) {
     gen_addr(node->lhs);
     push();
     gen_expr(node->rhs);
-    pop("%rdi");
-    printf("  mov %%rax, (%%rdi)\n");
+    store();
     return;
   case ND_FUNCALL: {
     int nargs = 0;
@@ -181,7 +201,7 @@ static void assign_lvar_offsets(Function *prog) {
   for (Function *fn = prog; fn; fn = fn->next) {
     int offset = 0;
     for (Obj *var = fn->locals; var; var = var->next) {
-      offset += 8;
+      offset += var->ty->size;
       var->offset = -offset;
     }
     fn->stack_size = align_to(offset, 16);

parse.c

@@ -92,34 +92,49 @@ static char *get_ident(Token *tok) {
   return strndup(tok->loc, tok->len);
 }
 
+static int get_number(Token *tok) {
+  if (tok->kind != TK_NUM)
+    error_tok(tok, "expected a number");
+  return tok->val;
+}
+
 // typespec = "int"
 static Type *typespec(Token **rest, Token *tok) {
   *rest = skip(tok, "int");
   return ty_int;
 }
 
-// type-suffix = ("(" func-params? ")")?
-// func-params = param ("," param)*
+// func-params = (param ("," param)*)? ")"
 // param       = typespec declarator
-static Type *type_suffix(Token **rest, Token *tok, Type *ty) {
-  if (equal(tok, "(")) {
-    tok = tok->next;
+static Type *func_params(Token **rest, Token *tok, Type *ty) {
+  Type head = {};
+  Type *cur = &head;
+
+  while (!equal(tok, ")")) {
+    if (cur != &head)
+      tok = skip(tok, ",");
+    Type *basety = typespec(&tok, tok);
+    Type *ty = declarator(&tok, tok, basety);
+    cur = cur->next = copy_type(ty);
+  }
 
-    Type head = {};
-    Type *cur = &head;
+  ty = func_type(ty);
+  ty->params = head.next;
+  *rest = tok->next;
+  return ty;
+}
 
-    while (!equal(tok, ")")) {
-      if (cur != &head)
-        tok = skip(tok, ",");
-      Type *basety = typespec(&tok, tok);
-      Type *ty = declarator(&tok, tok, basety);
-      cur = cur->next = copy_type(ty);
-    }
+// type-suffix = "(" func-params
+//             | "[" num "]"
+//             | ε
+static Type *type_suffix(Token **rest, Token *tok, Type *ty) {
+  if (equal(tok, "("))
+    return func_params(rest, tok->next, ty);
 
-    ty = func_type(ty);
-    ty->params = head.next;
-    *rest = tok->next;
-    return ty;
+  if (equal(tok, "[")) {
+    int sz = get_number(tok->next);
+    *rest = skip(tok->next->next, "]");
+    return array_of(ty, sz);
   }
 
   *rest = tok;
@@ -353,7 +368,7 @@ static Node *new_add(Node *lhs, Node *rhs, Token *tok) {
   }
 
   // ptr + num
-  rhs = new_binary(ND_MUL, rhs, new_num(8, tok), tok);
+  rhs = new_binary(ND_MUL, rhs, new_num(lhs->ty->base->size, tok), tok);
   return new_binary(ND_ADD, lhs, rhs, tok);
 }
 
@@ -368,7 +383,7 @@ static Node *new_sub(Node *lhs, Node *rhs, Token *tok) {
 
   // ptr - num
   if (lhs->ty->base && is_integer(rhs->ty)) {
-    rhs = new_binary(ND_MUL, rhs, new_num(8, tok), tok);
+    rhs = new_binary(ND_MUL, rhs, new_num(lhs->ty->base->size, tok), tok);
     add_type(rhs);
     Node *node = new_binary(ND_SUB, lhs, rhs, tok);
     node->ty = lhs->ty;
@@ -379,7 +394,7 @@ static Node *new_sub(Node *lhs, Node *rhs, Token *tok) {
   if (lhs->ty->base && rhs->ty->base) {
     Node *node = new_binary(ND_SUB, lhs, rhs, tok);
     node->ty = ty_int;
-    return new_binary(ND_DIV, node, new_num(8, tok), tok);
+    return new_binary(ND_DIV, node, new_num(lhs->ty->base->size, tok), tok);
   }
 
   error_tok(tok, "invalid operands");

test.sh

@@ -112,4 +112,10 @@ assert 7 'int main() { return add2(3,4); } int add2(int x, int y) { return x+y;
 assert 1 'int main() { return sub2(4,3); } int sub2(int x, int y) { return x-y; }'
 assert 55 'int main() { return fib(9); } int fib(int x) { if (x<=1) return 1; return fib(x-1) + fib(x-2); }'
 
+assert 3 'int main() { int x[2]; int *y=&x; *y=3; return *x; }'
+
+assert 3 'int main() { int x[3]; *x=3; *(x+1)=4; *(x+2)=5; return *x; }'
+assert 4 'int main() { int x[3]; *x=3; *(x+1)=4; *(x+2)=5; return *(x+1); }'
+assert 5 'int main() { int x[3]; *x=3; *(x+1)=4; *(x+2)=5; return *(x+2); }'
+
 echo OK

type.c

@@ -1,6 +1,6 @@
 #include "chibicc.h"
 
-Type *ty_int = &(Type){TY_INT};
+Type *ty_int = &(Type){TY_INT, 8};
 
 bool is_integer(Type *ty) {
   return ty->kind == TY_INT;
@@ -15,6 +15,7 @@ Type *copy_type(Type *ty) {
 Type *pointer_to(Type *base) {
   Type *ty = calloc(1, sizeof(Type));
   ty->kind = TY_PTR;
+  ty->size = 8;
   ty->base = base;
   return ty;
 }
@@ -26,6 +27,15 @@ Type *func_type(Type *return_ty) {
   return ty;
 }
 
+Type *array_of(Type *base, int len) {
+  Type *ty = calloc(1, sizeof(Type));
+  ty->kind = TY_ARRAY;
+  ty->size = base->size * len;
+  ty->base = base;
+  ty->array_len = len;
+  return ty;
+}
+
 void add_type(Node *node) {
   if (!node || node->ty)
     return;
@@ -49,7 +59,11 @@ void add_type(Node *node) {
   case ND_MUL:
   case ND_DIV:
   case ND_NEG:
+    node->ty = node->lhs->ty;
+    return;
   case ND_ASSIGN:
+    if (node->lhs->ty->kind == TY_ARRAY)
+      error_tok(node->lhs->tok, "not an lvalue");
     node->ty = node->lhs->ty;
     return;
   case ND_EQ:
@@ -64,10 +78,13 @@ void add_type(Node *node) {
     node->ty = node->var->ty;
     return;
   case ND_ADDR:
-    node->ty = pointer_to(node->lhs->ty);
+    if (node->lhs->ty->kind == TY_ARRAY)
+      node->ty = pointer_to(node->lhs->ty->base);
+    else
+      node->ty = pointer_to(node->lhs->ty);
     return;
   case ND_DEREF:
-    if (node->lhs->ty->kind != TY_PTR)
+    if (!node->lhs->ty->base)
       error_tok(node->tok, "invalid pointer dereference");
     node->ty = node->lhs->ty->base;
     return;