[DRAFT] add wasm-bindgen support

emcc.py

@@ -733,7 +733,7 @@ def compile_source_file(input_file):
       # Default to assuming the inputs are object files and pass them to the linker
       pass
 
-  return [f.value for f in linker_args]
+  return linker_args
 
 
 @ToolchainProfiler.profile()

src/runtime_common.js

@@ -98,7 +98,7 @@ var
   HEAPU64;
 #endif
 
-#if SUPPORT_BIG_ENDIAN
+#if SUPPORT_BIG_ENDIAN || WASM_BINDGEN
 /** @type {!DataView} */
 var HEAP_DATA_VIEW;
 #endif
@@ -145,8 +145,10 @@ function updateMemoryViews() {
   {{{ maybeExportHeap('HEAP64')  }}}HEAP64 = new BigInt64Array(b);
   {{{ maybeExportHeap('HEAPU64') }}}HEAPU64 = new BigUint64Array(b);
 #endif
+#if SUPPORT_BIG_ENDIAN || WASM_BINDGEN
+  {{{ maybeExportHeap('HEAP_DATA_VIEW') }}}HEAP_DATA_VIEW = new DataView(b);
+#endif
 #if SUPPORT_BIG_ENDIAN
-  {{{ maybeExportHeap('HEAP_DATA_VIEW') }}} HEAP_DATA_VIEW = new DataView(b);
   LE_HEAP_UPDATE();
 #endif
 }

src/settings.js

@@ -2164,6 +2164,10 @@ var LEGACY_RUNTIME = false;
 // [link]
 var SIGNATURE_CONVERSIONS = [];
 
+// Run wasm-bindgen and integrate the rust-exported symbols into the rest of Emscripten's JS output.
+// [link]
+var WASM_BINDGEN = 0;
+
 // Experimental support for wasm source phase imports.
 // This is only currently implemented in the pre-release/nightly version of node,
 // and not yet supported by browsers.

test/rust/basics/.cargo/config.toml

@@ -0,0 +1,2 @@
+[build]
+target = "wasm32-unknown-emscripten"

test/rust/bindgen_integration/.cargo/config.toml

@@ -0,0 +1,7 @@
+[build]
+target = "wasm32-unknown-emscripten"
+rustflags = [
+    "-Cllvm-args=-enable-emscripten-cxx-exceptions=0",
+    "-Cpanic=abort",
+    "-Crelocation-model=static",
+]

test/rust/bindgen_integration/Cargo.toml

@@ -0,0 +1,9 @@
+[package]
+name = "bindgen_integration"
+edition = "2021"
+
+[lib]
+crate-type = ["staticlib"]
+
+[dependencies]
+wasm-bindgen = { path = "../../../wasm-bindgen" }

test/rust/bindgen_integration/src/lib.rs

@@ -0,0 +1,6 @@
+use wasm_bindgen::prelude::*;
+
+#[wasm_bindgen]
+pub fn rs_add(a: i32, b: i32) -> i32 {
+    return a + b;
+}

test/test_other.py

@@ -15943,7 +15943,7 @@ def test_embool(self):
   @requires_rust
   def test_rust_integration_basics(self):
     copytree(test_file('rust/basics'), '.')
-    self.run_process(['cargo', 'build', '--target=wasm32-unknown-emscripten'])
+    self.run_process(['cargo', 'build'])
     lib = 'target/wasm32-unknown-emscripten/debug/libbasics.a'
     self.assertExists(lib)
 
@@ -15955,6 +15955,40 @@ def test_rust_integration_basics(self):
     }''')
     self.do_runf('main.cpp', 'Hello from rust!', emcc_args=[lib])
 
+  @requires_rust
+  def test_wasm_bindgen_integration(self):
+    copytree(test_file('rust/bindgen_integration'), '.')
+    self.run_process(['cargo', 'build'])
+    lib = 'target/wasm32-unknown-emscripten/debug/libbindgen_integration.a'
+    self.assertExists(lib)
+
+    create_file('main.cpp', '')
+    create_file('post.js', '''
+      Module.onRuntimeInitialized = () => {
+        out(Module.rs_add(17, 25));
+      };
+    ''')
+    exported_funcs = [
+        '___wbindgen_describe_rs_add',
+        '_rs_add',
+        '___externref_drop_slice',
+        '___externref_heap_live_count',
+        '___externref_table_alloc',
+        '___externref_table_dealloc',
+        '___wbindgen_exn_store',
+        '___wbindgen_free',
+        '___wbindgen_malloc',
+        '___wbindgen_realloc']
+    emcc_args = [
+        lib,
+        '-sWASM_BINDGEN',
+        '--post-js',
+        'post.js',
+        '-Wno-undefined',
+        '-sEXPORTED_FUNCTIONS=' + ','.join(exported_funcs),
+    ]
+    self.do_runf('main.cpp', '42', emcc_args=emcc_args)
+
   def test_relative_em_cache(self):
     with env_modify({'EM_CACHE': 'foo'}):
       err = self.expect_fail([EMCC, '-c', test_file('hello_world.c')])

tools/building.py

@@ -254,7 +254,32 @@ def lld_flags_for_executable(external_symbols):
   return cmd
 
 
-def link_lld(args, target, external_symbols=None):
+def get_wasm_bindgen_exported_symbols(input_files):
+  if not os.path.exists(LLVM_NM):
+    exit_with_error('llvm-nm not found in LLVM directory: %s', LLVM_NM)
+
+  nm_args = [
+      LLVM_NM,
+      '--defined-only',
+      '--extern-only',
+      '--format=just-symbols',
+      '--print-file-name',
+      '--quiet',
+  ]
+
+  result = run_process(nm_args + input_files, stdout=subprocess.PIPE)
+  symbols = []
+  for line in result.stdout.splitlines():
+    (path, symbol) = line.split()
+    # Skip mangled (non-C) symbols
+    if symbol.startswith('_Z') or symbol.startswith('_R') or symbol.startswith('anon.'):
+      continue
+    symbols.append(symbol)
+
+  return symbols
+
+
+def link_lld(args, target, external_symbols=None, linker_inputs=[]):
   if not os.path.exists(WASM_LD):
     exit_with_error('linker binary not found in LLVM directory: %s', WASM_LD)
   # runs lld to link things.
@@ -263,6 +288,10 @@ def link_lld(args, target, external_symbols=None):
   # grouping.
   args = [a for a in args if a not in ('--start-group', '--end-group')]
 
+  if settings.WASM_BINDGEN:
+    exported_symbols = get_wasm_bindgen_exported_symbols(linker_inputs)
+    args.extend(f'--export-if-defined={e}' for e in exported_symbols)
+
   # Emscripten currently expects linkable output (SIDE_MODULE/MAIN_MODULE) to
   # include all archive contents.
   if settings.LINKABLE:
@@ -1239,6 +1268,30 @@ def run_wasm_opt(infile, outfile=None, args=[], **kwargs):  # noqa
   return run_binaryen_command('wasm-opt', infile, outfile, args=args, **kwargs)
 
 
+def run_wasm_bindgen(infile, outfile=None, args=[], **kwargs):  # noqa
+  bindgen_out_dir = get_emscripten_temp_dir() + '/bindgen_out/'
+
+  cmd = config.WASM_BINDGEN + [
+    infile,
+    '--target',
+    'emscripten',
+    '--keep-lld-exports',
+    '--keep-debug',
+    '--out-dir',
+    bindgen_out_dir,
+  ]
+  check_call(cmd)
+
+  # Don't try to predict the .wasm filename that wasm-bindgen outputs. Instead
+  # just grab the .wasm file itself.
+  all_output_files = os.listdir(bindgen_out_dir)
+  new_wasm_file = list(filter(lambda x: x.endswith('.wasm'), all_output_files))[0]
+  if outfile == None:
+    outfile = infile
+
+  shutil.copyfile(bindgen_out_dir + new_wasm_file, outfile)
+
+
 intermediate_counter = 0
 
 

tools/config.py

@@ -28,6 +28,7 @@
 CACHE = None
 PORTS = None
 COMPILER_WRAPPER = None
+WASM_BINDGEN = None
 
 # Set by init()
 EM_CONFIG = None
@@ -61,6 +62,7 @@ def fix_js_engine(old, new):
 def normalize_config_settings():
   global CACHE, PORTS, LLVM_ADD_VERSION, CLANG_ADD_VERSION, CLOSURE_COMPILER
   global NODE_JS, NODE_JS_TEST, V8_ENGINE, JS_ENGINES, SPIDERMONKEY_ENGINE, WASM_ENGINES
+  global WASM_BINDGEN
 
   # EM_CONFIG stuff
   if not JS_ENGINES:
@@ -73,6 +75,7 @@ def normalize_config_settings():
   JS_ENGINES = [listify(engine) for engine in JS_ENGINES]
   WASM_ENGINES = [listify(engine) for engine in WASM_ENGINES]
   CLOSURE_COMPILER = listify(CLOSURE_COMPILER)
+  WASM_BINDGEN = listify(WASM_BINDGEN)
   if not CACHE:
     CACHE = path_from_root('cache')
   if not PORTS:
@@ -122,6 +125,7 @@ def parse_config_file():
     'CACHE',
     'PORTS',
     'COMPILER_WRAPPER',
+    'WASM_BINDGEN',
   )
 
   # Only propagate certain settings from the config file.

tools/emscripten.py

@@ -546,7 +546,8 @@ def finalize_wasm(infile, outfile, js_syms):
   if not need_name_section:
     strip_sections += ['name']
 
-  if strip_sections or not settings.GENERATE_DWARF:
+  # TODO(walkingeyerobot): make this work. it appears to not like the wasm file from wasm-bindgen
+  if not settings.WASM_BINDGEN and (strip_sections or not settings.GENERATE_DWARF):
     building.save_intermediate(outfile, 'strip.wasm')
     building.strip(infile, outfile, debug=not settings.GENERATE_DWARF,
                    sections=strip_sections)
@@ -643,7 +644,7 @@ def create_tsd_exported_runtime_methods(metadata):
   return utils.read_file(tsc_output_file)
 
 
-def create_tsd(metadata, embind_tsd):
+def create_tsd(metadata, embind_tsd, bindgen_tsd = None):
   out = '// TypeScript bindings for emscripten-generated code.  Automatically generated at compile time.\n'
   if settings.EXPORTED_RUNTIME_METHODS:
     out += create_tsd_exported_runtime_methods(metadata)
@@ -673,6 +674,12 @@ def create_tsd(metadata, embind_tsd):
   # Add in embind definitions.
   if embind_tsd:
     export_interfaces += ' & EmbindModule'
+  if settings.WASM_BINDGEN and bindgen_tsd:
+    for file_path in bindgen_tsd:
+      with open(file_path, 'r') as file:
+          for line in file:
+              out += f'{line}'
+    export_interfaces += ' & BindgenModule'
   out += f'export type MainModule = {export_interfaces};\n'
   if settings.MODULARIZE:
     return_type = 'MainModule'

tools/link.py

@@ -39,6 +39,7 @@
 from .shared import in_temp, safe_copy, do_replace
 from .shared import DEBUG, WINDOWS, DYLIB_EXTENSIONS
 from .shared import unsuffixed, unsuffixed_basename, get_file_suffix
+from .shared import get_emscripten_temp_dir
 from .settings import settings, default_setting, user_settings, JS_ONLY_SETTINGS, DEPRECATED_SETTINGS
 from .minimal_runtime_shell import generate_minimal_runtime_html
 
@@ -1900,7 +1901,7 @@ def phase_calculate_system_libraries(options):
 
 
 @ToolchainProfiler.profile_block('link')
-def phase_link(linker_args, wasm_target, js_syms):
+def phase_link(linker_args, linker_inputs, wasm_target, js_syms):
   logger.debug(f'linking: {linker_args}')
 
   # Make a final pass over settings.EXPORTED_FUNCTIONS to remove any
@@ -1922,11 +1923,12 @@ def phase_link(linker_args, wasm_target, js_syms):
     # TODO(sbc): Remove this double execution of wasm-ld if we ever find a way to
     # distinguish EMSCRIPTEN_KEEPALIVE exports from `--export-dynamic` exports.
     settings.LINKABLE = False
-    building.link_lld(linker_args, wasm_target, external_symbols=js_syms)
+    building.link_lld(linker_args, wasm_target, external_symbols=js_syms,
+                      linker_inputs=linker_inputs)
     settings.LINKABLE = True
     rtn = extract_metadata.extract_metadata(wasm_target)
 
-  building.link_lld(linker_args, wasm_target, external_symbols=js_syms)
+  building.link_lld(linker_args, wasm_target, external_symbols=js_syms, linker_inputs=linker_inputs)
   return rtn
 
 
@@ -1948,6 +1950,10 @@ def phase_post_link(options, in_wasm, wasm_target, target, js_syms, base_metadat
 
   settings.TARGET_JS_NAME = os.path.basename(js_target)
 
+  if settings.WASM_BINDGEN:
+    building.run_wasm_bindgen(in_wasm)
+    settings.JS_LIBRARIES += [get_emscripten_temp_dir() + '/bindgen_out/library_bindgen.js']
+
   metadata = phase_emscript(in_wasm, wasm_target, js_syms, base_metadata)
 
   if settings.EMBIND_AOT:
@@ -2062,7 +2068,10 @@ def phase_emit_tsd(options, wasm_target, js_target, js_syms, metadata):
   embind_tsd = ''
   if settings.EMBIND:
     embind_tsd = run_embind_gen(options, wasm_target, js_syms, {'EMBIND_AOT': False})
-  all_tsd = emscripten.create_tsd(metadata, embind_tsd)
+  bindgen_ts_files = glob.glob(get_emscripten_temp_dir() + "/bindgen_out/*.d.ts", recursive=False)
+  # This list comprehension then filters out any files that end with .wasm.d.ts.
+  bindgen_ts_files = [file for file in bindgen_ts_files if not file.endswith('.wasm.d.ts')]
+  all_tsd = emscripten.create_tsd(metadata, embind_tsd, bindgen_ts_files)
   out_file = os.path.join(os.path.dirname(js_target), filename)
   write_file(out_file, all_tsd)
 
@@ -3060,6 +3069,9 @@ def run(options, linker_args):
   if not linker_args:
     exit_with_error('no input files')
 
+  linker_inputs = [f.value for f in linker_args if f.is_file]
+  linker_args = [f.value for f in linker_args]
+
   if options.output_file and options.output_file.startswith('-'):
     exit_with_error(f'invalid output filename: `{options.output_file}`')
 
@@ -3109,7 +3121,7 @@ def add_js_deps(sym):
       settings.ASYNCIFY_IMPORTS_EXCEPT_JS_LIBS = settings.ASYNCIFY_IMPORTS[:]
       settings.ASYNCIFY_IMPORTS += ['*.' + x for x in js_info['asyncFuncs']]
 
-  base_metadata = phase_link(linker_args, wasm_target, js_syms)
+  base_metadata = phase_link(linker_args, linker_inputs, wasm_target, js_syms)
 
   # Special handling for when the user passed '-Wl,--version'.  In this case the linker
   # does not create the output file, but just prints its version and exits with 0.