Handle memory errors occurring in wasm scanners

* In WASM, use a custom, simple malloc implementation that lets us
  expicitly reset the heap with a new start location.
* When a WASM call traps or errors, propagate that as a parse failure.
* Reset the WASM heap after every parse.

Co-authored-by: Conrad <conrad@zed.dev>

cli/src/tests/helpers/allocations.rs

@@ -81,7 +81,6 @@ fn record_alloc(ptr: *mut c_void) {
 }
 
 fn record_dealloc(ptr: *mut c_void) {
-    assert!(!ptr.is_null(), "Zero pointer deallocation!");
     RECORDER.with(|recorder| {
         if recorder.enabled.load(SeqCst) {
             recorder

cli/src/tests/wasm_language_test.rs

@@ -29,7 +29,71 @@ fn test_wasm_stdlib_symbols() {
 }
 
 #[test]
-fn test_load_wasm_language() {
+fn test_load_wasm_ruby_language() {
+    allocations::record(|| {
+        let mut store = WasmStore::new(ENGINE.clone()).unwrap();
+        let mut parser = Parser::new();
+        let wasm = fs::read(WASM_DIR.join("tree-sitter-ruby.wasm")).unwrap();
+        let language = store.load_language("ruby", &wasm).unwrap();
+        parser.set_wasm_store(store).unwrap();
+        parser.set_language(&language).unwrap();
+        let tree = parser.parse("class A; end", None).unwrap();
+        assert_eq!(
+            tree.root_node().to_sexp(),
+            "(program (class name: (constant)))"
+        );
+    });
+}
+
+#[test]
+fn test_load_wasm_html_language() {
+    allocations::record(|| {
+        let mut store = WasmStore::new(ENGINE.clone()).unwrap();
+        let mut parser = Parser::new();
+        let wasm = fs::read(WASM_DIR.join("tree-sitter-html.wasm")).unwrap();
+        let language = store.load_language("html", &wasm).unwrap();
+        parser.set_wasm_store(store).unwrap();
+        parser.set_language(&language).unwrap();
+        let tree = parser
+            .parse("<div><span></span><p></p></div>", None)
+            .unwrap();
+        assert_eq!(
+            tree.root_node().to_sexp(),
+            "(document (element (start_tag (tag_name)) (element (start_tag (tag_name)) (end_tag (tag_name))) (element (start_tag (tag_name)) (end_tag (tag_name))) (end_tag (tag_name))))"
+        );
+    });
+}
+
+#[test]
+fn test_load_wasm_rust_language() {
+    allocations::record(|| {
+        let mut store = WasmStore::new(ENGINE.clone()).unwrap();
+        let mut parser = Parser::new();
+        let wasm = fs::read(WASM_DIR.join("tree-sitter-rust.wasm")).unwrap();
+        let language = store.load_language("rust", &wasm).unwrap();
+        parser.set_wasm_store(store).unwrap();
+        parser.set_language(&language).unwrap();
+        let tree = parser.parse("fn main() {}", None).unwrap();
+        assert_eq!(tree.root_node().to_sexp(), "(source_file (function_item name: (identifier) parameters: (parameters) body: (block)))");
+    });
+}
+
+#[test]
+fn test_load_wasm_javascript_language() {
+    allocations::record(|| {
+        let mut store = WasmStore::new(ENGINE.clone()).unwrap();
+        let mut parser = Parser::new();
+        let wasm = fs::read(WASM_DIR.join("tree-sitter-javascript.wasm")).unwrap();
+        let language = store.load_language("javascript", &wasm).unwrap();
+        parser.set_wasm_store(store).unwrap();
+        parser.set_language(&language).unwrap();
+        let tree = parser.parse("const a = b\nconst c = d", None).unwrap();
+        assert_eq!(tree.root_node().to_sexp(), "(program (lexical_declaration (variable_declarator name: (identifier) value: (identifier))) (lexical_declaration (variable_declarator name: (identifier) value: (identifier))))");
+    });
+}
+
+#[test]
+fn test_load_multiple_wasm_languages() {
     allocations::record(|| {
         let mut store = WasmStore::new(ENGINE.clone()).unwrap();
         let mut parser = Parser::new();
@@ -158,3 +222,28 @@ fn test_load_wasm_errors() {
         );
     });
 }
+
+#[test]
+fn test_wasm_oom() {
+    allocations::record(|| {
+        let mut store = WasmStore::new(ENGINE.clone()).unwrap();
+        let mut parser = Parser::new();
+        let wasm = fs::read(WASM_DIR.join("tree-sitter-html.wasm")).unwrap();
+        let language = store.load_language("html", &wasm).unwrap();
+        parser.set_wasm_store(store).unwrap();
+        parser.set_language(&language).unwrap();
+
+        let tag_name = "a-b".repeat(2 * 1024 * 1024);
+        let code = format!("<{tag_name}>hello world</{tag_name}>");
+        assert!(parser.parse(&code, None).is_none());
+
+        let tag_name = "a-b".repeat(20);
+        let code = format!("<{tag_name}>hello world</{tag_name}>");
+        parser.set_language(&language).unwrap();
+        let tree = parser.parse(&code, None).unwrap();
+        assert_eq!(
+            tree.root_node().to_sexp(),
+            "(document (element (start_tag (tag_name)) (text) (end_tag (tag_name))))"
+        );
+    });
+}

cli/src/wasm.rs

@@ -62,6 +62,7 @@ pub fn compile_language_to_wasm(
         "__cxa_atexit",
         "abort",
         "emscripten_notify_memory_growth",
+        "tree_sitter_debug_message",
         "proc_exit",
     ];