fix: 5项立即修复 + 2项尽快修复

立即:
- lexer: token数组容量改为src_len+16 + idx越界防御
- symbol: 4个函数arena_alloc加NULL检查
- codegen: verify_err fallback用arena_strdup替代静态字符串
- codegen: cleanup_list从固定64改为arena动态扩容
- lexer: 标识符/字符串字面量65535字符上限

尽快:
- to_llvm_type: TYPE_STRUCT/TYPE_UNKNOWN/TYPE_ERROR显式case
- LLVMGetValueType2不存在(LLVM 22仍用旧名), 保留GlobalGetValueType

src/codegen/codegen.c

@@ -50,8 +50,9 @@ typedef struct {
     LLVMValueRef      strlen_fn;
     LLVMValueRef      memcpy_fn;
     // 自动内存管理: 追踪需要 free 的 str alloca
-    LLVMValueRef      cleanup_list[64];
+    LLVMValueRef*     cleanup_list;
     size_t            cleanup_count;
+    size_t            cleanup_cap;
 } CgCtx;
 
 // === 类型映射（需要 Context）===
@@ -61,6 +62,9 @@ static LLVMTypeRef to_llvm_type(CgCtx* ctx, TypeKind kind) {
         case TYPE_F64:  return LLVMDoubleTypeInContext(ctx->context);
         case TYPE_BOOL: return LLVMInt1TypeInContext(ctx->context);
         case TYPE_STR:  return LLVMPointerType(LLVMInt8TypeInContext(ctx->context), 0);
+        case TYPE_STRUCT:
+        case TYPE_UNKNOWN:
+        case TYPE_ERROR:
         default:        return LLVMVoidTypeInContext(ctx->context);
     }
 }
@@ -350,9 +354,16 @@ static LLVMValueRef codegen_expr(CgCtx* ctx, AstNode* node) {
 
 // === 自动内存管理: 作用域退出时释放 str 堆分配 ===
 static void cleanup_add(CgCtx* ctx, LLVMValueRef alloca) {
-    if (ctx->cleanup_count < 64) {
+    if (ctx->cleanup_count >= ctx->cleanup_cap) {
-        ctx->cleanup_list[ctx->cleanup_count++] = alloca;
+        size_t new_cap = ctx->cleanup_cap ? ctx->cleanup_cap * 2 : 16;
+        LLVMValueRef* new_list = arena_alloc(ctx->arena, new_cap * sizeof(LLVMValueRef));
+        if (!new_list) return;
+        if (ctx->cleanup_list)
+            memcpy(new_list, ctx->cleanup_list, ctx->cleanup_count * sizeof(LLVMValueRef));
+        ctx->cleanup_list = new_list;
+        ctx->cleanup_cap = new_cap;
     }
+    ctx->cleanup_list[ctx->cleanup_count++] = alloca;
 }
 
 // 释放从 mark 位置开始的所有 str 变量
@@ -645,7 +656,8 @@ LLVMModuleRef codegen_module(AstNode* ast, Arena* codegen_arena,
     // 验证模块（使用 ReturnStatus 以获取完整错误消息）
     char* verify_err = NULL;
     if (LLVMVerifyModule(ctx.module, LLVMReturnStatusAction, &verify_err)) {
-        *error_msg = verify_err ? verify_err : "模块验证失败（错误消息为 NULL）";
+        *error_msg = verify_err ? verify_err
+            : arena_strdup(ctx.arena, "LLVM 模块验证失败");
         LLVMDisposeBuilder(ctx.builder);
         *out_context = ctx.context;
         return NULL;

src/lexer/lexer.c

@@ -70,7 +70,10 @@ static TokenKind check_keyword(const Token* tok) {
 
 static Token lex_ident_or_keyword(Lexer* l) {
     int start = l->pos;
-    while (isalnum(peek(l)) || peek(l) == '_') advance(l);
+    while (isalnum(peek(l)) || peek(l) == '_') {
+        if (l->pos - start > 65535) break;  // 标识符长度上限
+        advance(l);
+    }
     Token t = make_token(l, TOK_IDENT, start, l->pos - start);
     t.kind = check_keyword(&t);
     return t;
@@ -79,8 +82,9 @@ static Token lex_ident_or_keyword(Lexer* l) {
 Token* lex(Arena* a, const char* source, const char* filename,
            size_t* count, ErrorInfo* error) {
     Lexer l = {.src = source, .filename = filename, .pos = 0, .line = 1, .col = 1};
-    // 预估容量：源码长度的 1/3
+    // 容量上限: 极端情况每个字符一个 token (如 "(){}+-"), src_len 足够
-    size_t cap = strlen(source) / 3 + 16;
+    size_t src_len = strlen(source);
+    size_t cap = src_len + 16;
     Token* tokens = arena_alloc(a, cap * sizeof(Token));
     if (!tokens) { *count = 0; return NULL; }
     size_t idx = 0;
@@ -88,6 +92,7 @@ Token* lex(Arena* a, const char* source, const char* filename,
     while (peek(&l) != '\0') {
         skip_whitespace(&l);
         if (peek(&l) == '\0') break;
+        if (idx >= cap) { *count = 0; return NULL; }  // 防御
 
         int line = l.line, col = l.col;
         char c = peek(&l);
@@ -96,7 +101,10 @@ Token* lex(Arena* a, const char* source, const char* filename,
         else if (c == '"') {
             advance(&l);  // 跳过开头的 "
             int start = l.pos;
-            while (peek(&l) != '"' && peek(&l) != '\0' && peek(&l) != '\n') advance(&l);
+            while (peek(&l) != '"' && peek(&l) != '\0' && peek(&l) != '\n') {
+                if (l.pos - start > 65535) break;  // 字符串长度上限
+                advance(&l);
+            }
             int len = l.pos - start;
             if (peek(&l) != '"') {
                 *error = (ErrorInfo){.message="未闭合的字符串", .filename=filename, .line=line, .col=col};

src/sema/symbol.c

@@ -4,6 +4,7 @@
 
 Scope* scope_new(void* alloc, Scope* parent) {
     Scope* s = (Scope*)arena_alloc_impl(alloc, sizeof(Scope));
+    if (!s) return NULL;
     s->head = NULL;
     s->parent = parent;
     return s;
@@ -26,6 +27,7 @@ Symbol* scope_insert(Scope* scope, void* alloc, const char* name,
         }
     }
     Symbol* sym = (Symbol*)arena_alloc_impl(alloc, sizeof(Symbol));
+    if (!sym) return NULL;
     sym->name = name; sym->kind = kind; sym->type = type;
     sym->is_mut = false; sym->return_type = TYPE_VOID;
     sym->param_types = NULL; sym->param_count = 0;
@@ -46,6 +48,7 @@ Symbol* scope_insert_function(Scope* scope, void* alloc, const char* name,
         }
     }
     Symbol* sym = (Symbol*)arena_alloc_impl(alloc, sizeof(Symbol));
+    if (!sym) return NULL;
     sym->name = name; sym->kind = SYM_FUNCTION; sym->type = TYPE_VOID;
     sym->return_type = ret; sym->param_types = pt; sym->param_count = pc;
     sym->struct_field_names = NULL;
@@ -66,6 +69,7 @@ Symbol* scope_insert_struct(Scope* scope, void* alloc, const char* name,
         }
     }
     Symbol* sym = (Symbol*)arena_alloc_impl(alloc, sizeof(Symbol));
+    if (!sym) return NULL;
     sym->name = name; sym->kind = SYM_STRUCT; sym->type = TYPE_STRUCT;
     sym->is_mut = false; sym->return_type = TYPE_VOID;
     sym->param_types = NULL; sym->param_count = 0;