[3.11] gh-95853: Add script to automate WASM build (GH-95828, GH-95985, GH-96045, GH-96389, GH-96744) (GH-96749)

Automate WASM build with a new Python script. The script provides
several build profiles with configure flags for Emscripten flavors
and WASI. The script can detect and use Emscripten SDK and WASI SDK from
default locations or env vars.

``configure`` now detects Node arguments and creates HOSTRUNNER
arguments for Node 16. It also sets correct arguments for
``wasm64-emscripten``.

Author: tiran

Lib/distutils/tests/test_sysconfig.py

@@ -49,6 +49,7 @@ def test_get_config_vars(self):
         self.assertIsInstance(cvars, dict)
         self.assertTrue(cvars)
 
+    @unittest.skipIf(is_wasi, "Incompatible with WASI mapdir and OOT builds")
     def test_srcdir(self):
         # See Issues #15322, #15364.
         srcdir = sysconfig.get_config_var('srcdir')

Lib/test/test_decimal.py

@@ -37,7 +37,7 @@
                           requires_legacy_unicode_capi, check_sanitizer)
 from test.support import (TestFailed,
                           run_with_locale, cpython_only,
-                          darwin_malloc_err_warning)
+                          darwin_malloc_err_warning, is_emscripten)
 from test.support.import_helper import import_fresh_module
 from test.support import threading_helper
 from test.support import warnings_helper
@@ -5623,6 +5623,7 @@ def __abs__(self):
     # Issue 41540:
     @unittest.skipIf(sys.platform.startswith("aix"),
                      "AIX: default ulimit: test is flaky because of extreme over-allocation")
+    @unittest.skipIf(is_emscripten, "Test is unstable on Emscripten")
     @unittest.skipIf(check_sanitizer(address=True, memory=True),
                      "ASAN/MSAN sanitizer defaults to crashing "
                      "instead of returning NULL for malloc failure.")

Lib/test/test_sysconfig.py

@@ -439,6 +439,7 @@ def test_platform_in_subprocess(self):
         self.assertEqual(status, 0)
         self.assertEqual(my_platform, test_platform)
 
+    @unittest.skipIf(is_wasi, "Incompatible with WASI mapdir and OOT builds")
     def test_srcdir(self):
         # See Issues #15322, #15364.
         srcdir = sysconfig.get_config_var('srcdir')

Lib/test/test_unicode_file_functions.py

@@ -6,6 +6,7 @@
 import warnings
 from unicodedata import normalize
 from test.support import os_helper
+from test import support
 
 
 filenames = [
@@ -123,6 +124,10 @@ def test_open(self):
     # NFKD in Python is useless, because darwin will normalize it later and so
     # open(), os.stat(), etc. don't raise any exception.
     @unittest.skipIf(sys.platform == 'darwin', 'irrelevant test on Mac OS X')
+    @unittest.skipIf(
+        support.is_emscripten or support.is_wasi,
+        "test fails on Emscripten/WASI when host platform is macOS."
+    )
     def test_normalize(self):
         files = set(self.files)
         others = set()

Lib/test/test_warnings/__init__.py

@@ -489,7 +489,14 @@ def test_warn_explicit_non_ascii_filename(self):
                 module=self.module) as w:
             self.module.resetwarnings()
             self.module.filterwarnings("always", category=UserWarning)
-            for filename in ("nonascii\xe9\u20ac", "surrogate\udc80"):
+            filenames = ["nonascii\xe9\u20ac"]
+            if not support.is_emscripten:
+                # JavaScript does not like surrogates.
+                # Invalid UTF-8 leading byte 0x80 encountered when
+                # deserializing a UTF-8 string in wasm memory to a JS
+                # string!
+                filenames.append("surrogate\udc80")
+            for filename in filenames:
                 try:
                     os.fsencode(filename)
                 except UnicodeEncodeError:

Makefile.pre.in

@@ -817,10 +817,11 @@ $(DLLLIBRARY) libpython$(LDVERSION).dll.a: $(LIBRARY_OBJS)
 # wasm assets directory is relative to current build dir, e.g. "./usr/local".
 # --preload-file turns a relative asset path into an absolute path.
 
+.PHONY: wasm_stdlib
+wasm_stdlib: $(WASM_STDLIB)
 $(WASM_STDLIB): $(srcdir)/Lib/*.py $(srcdir)/Lib/*/*.py \
 	    $(srcdir)/Tools/wasm/wasm_assets.py \
-	    Makefile pybuilddir.txt Modules/Setup.local \
-	    python.html python.worker.js
+	    Makefile pybuilddir.txt Modules/Setup.local
 	$(PYTHON_FOR_BUILD) $(srcdir)/Tools/wasm/wasm_assets.py \
 	    --buildroot . --prefix $(prefix)
 
@@ -1713,6 +1714,10 @@ buildbottest: all
 		fi
 		$(TESTRUNNER) -j 1 -u all -W --slowest --fail-env-changed --timeout=$(TESTTIMEOUT) $(TESTOPTS)
 
+# Like testall, but run Python tests with HOSTRUNNER directly.
+hostrunnertest: all
+	$(RUNSHARED) $(HOSTRUNNER) ./$(BUILDPYTHON) -m test -u all $(TESTOPTS)
+
 pythoninfo: all
 		$(RUNSHARED) $(HOSTRUNNER) ./$(BUILDPYTHON) -m test.pythoninfo
 

Misc/NEWS.d/next/Tools-Demos/2022-08-10-17-08-43.gh-issue-95853.HCjC2m.rst

@@ -0,0 +1,2 @@
+The new tool ``Tools/wasm/wasm_builder.py`` automates configure, compile, and
+test steps for building CPython on WebAssembly platforms.

Misc/NEWS.d/next/Tools-Demos/2022-08-29-17-25-13.gh-issue-95853.Ce17cT.rst

@@ -0,0 +1,2 @@
+The ``wasm_build.py`` script now pre-builds Emscripten ports, checks for
+broken EMSDK versions, and warns about pkg-config env vars.

Modules/pyexpat.c

@@ -775,7 +775,7 @@ readinst(char *buf, int buf_size, PyObject *meth)
     Py_ssize_t len;
     const char *ptr;
 
-    str = PyObject_CallFunction(meth, "n", buf_size);
+    str = PyObject_CallFunction(meth, "i", buf_size);
     if (str == NULL)
         goto error;
 

Python/sysmodule.c

@@ -2775,14 +2775,18 @@ EM_JS(char *, _Py_emscripten_runtime, (void), {
     if (typeof navigator == 'object') {
         info = navigator.userAgent;
     } else if (typeof process == 'object') {
-        info = "Node.js ".concat(process.version)
+        info = "Node.js ".concat(process.version);
     } else {
-        info = "UNKNOWN"
+        info = "UNKNOWN";
     }
     var len = lengthBytesUTF8(info) + 1;
     var res = _malloc(len);
-    stringToUTF8(info, res, len);
+    if (res) stringToUTF8(info, res, len);
+#if __wasm64__
+    return BigInt(res);
+#else
     return res;
+#endif
 });
 
 static PyObject *

Tools/wasm/README.md

@@ -1,11 +1,16 @@
 # Python WebAssembly (WASM) build
 
-**WARNING: WASM support is highly experimental! Lots of features are not working yet.**
+**WARNING: WASM support is work-in-progress! Lots of features are not working yet.**
 
 This directory contains configuration and helpers to facilitate cross
-compilation of CPython to WebAssembly (WASM). For now we support
-*wasm32-emscripten* builds for modern browser and for *Node.js*. WASI
-(*wasm32-wasi*) is work-in-progress
+compilation of CPython to WebAssembly (WASM). Python supports Emscripten
+(*wasm32-emscripten*) and WASI (*wasm32-wasi*) targets. Emscripten builds
+run in modern browsers and JavaScript runtimes like *Node.js*. WASI builds
+use WASM runtimes such as *wasmtime*.
+
+Users and developers are encouraged to use the script
+`Tools/wasm/wasm_build.py`. The tool automates the build process and provides
+assistance with installation of SDKs.
 
 ## wasm32-emscripten build
 
@@ -17,7 +22,7 @@ access the file system directly.
 
 Cross compiling to the wasm32-emscripten platform needs the
 [Emscripten](https://emscripten.org/) SDK and a build Python interpreter.
-Emscripten 3.1.8 or newer are recommended. All commands below are relative
+Emscripten 3.1.19 or newer are recommended. All commands below are relative
 to a repository checkout.
 
 Christian Heimes maintains a container image with Emscripten SDK, Python
@@ -35,7 +40,13 @@ docker run --rm -ti -v $(pwd):/python-wasm/cpython -w /python-wasm/cpython quay.
 
 ### Compile a build Python interpreter
 
-From within the container, run the following commands:
+From within the container, run the following command:
+
+```shell
+./Tools/wasm/wasm_build.py build
+```
+
+The command is roughly equivalent to:
 
 ```shell
 mkdir -p builddir/build
@@ -45,13 +56,13 @@ make -j$(nproc)
 popd
 ```
 
-### Fetch and build additional emscripten ports
+### Cross-compile to wasm32-emscripten for browser
 
 ```shell
-embuilder build zlib bzip2
+./Tools/wasm/wasm_build.py emscripten-browser
 ```
 
-### Cross compile to wasm32-emscripten for browser
+The command is roughly equivalent to:
 
 ```shell
 mkdir -p builddir/emscripten-browser
@@ -85,22 +96,29 @@ and header files with debug builds.
 ### Cross compile to wasm32-emscripten for node
 
 ```shell
-mkdir -p builddir/emscripten-node
-pushd builddir/emscripten-node
+./Tools/wasm/wasm_build.py emscripten-browser-dl
+```
+
+The command is roughly equivalent to:
+
+```shell
+mkdir -p builddir/emscripten-node-dl
+pushd builddir/emscripten-node-dl
 
 CONFIG_SITE=../../Tools/wasm/config.site-wasm32-emscripten \
   emconfigure ../../configure -C \
     --host=wasm32-unknown-emscripten \
     --build=$(../../config.guess) \
     --with-emscripten-target=node \
+    --enable-wasm-dynamic-linking \
     --with-build-python=$(pwd)/../build/python
 
 emmake make -j$(nproc)
 popd
 ```
 
 ```shell
-node --experimental-wasm-threads --experimental-wasm-bulk-memory --experimental-wasm-bigint builddir/emscripten-node/python.js
+node --experimental-wasm-threads --experimental-wasm-bulk-memory --experimental-wasm-bigint builddir/emscripten-node-dl/python.js
 ```
 
 (``--experimental-wasm-bigint`` is not needed with recent NodeJS versions)
@@ -199,6 +217,15 @@ Node builds use ``NODERAWFS``.
 - Node RawFS allows direct access to the host file system without need to
   perform ``FS.mount()`` call.
 
+## wasm64-emscripten
+
+- wasm64 requires recent NodeJS and ``--experimental-wasm-memory64``.
+- ``EM_JS`` functions must return ``BigInt()``.
+- ``Py_BuildValue()`` format strings must match size of types. Confusing 32
+  and 64 bits types leads to memory corruption, see
+  [gh-95876](https://github.com/python/cpython/issues/95876) and
+  [gh-95878](https://github.com/python/cpython/issues/95878).
+
 # Hosting Python WASM builds
 
 The simple REPL terminal uses SharedArrayBuffer. For security reasons
@@ -234,6 +261,12 @@ The script ``wasi-env`` sets necessary compiler and linker flags as well as
 ``pkg-config`` overrides. The script assumes that WASI-SDK is installed in
 ``/opt/wasi-sdk`` or ``$WASI_SDK_PATH``.
 
+```shell
+./Tools/wasm/wasm_build.py wasi
+```
+
+The command is roughly equivalent to:
+
 ```shell
 mkdir -p builddir/wasi
 pushd builddir/wasi
@@ -308,26 +341,46 @@ if os.name == "posix":
 ```python
 >>> import os, sys
 >>> os.uname()
-posix.uname_result(sysname='Emscripten', nodename='emscripten', release='1.0', version='#1', machine='wasm32')
+posix.uname_result(
+    sysname='Emscripten',
+    nodename='emscripten',
+    release='3.1.19',
+    version='#1',
+    machine='wasm32'
+)
 >>> os.name
 'posix'
 >>> sys.platform
 'emscripten'
 >>> sys._emscripten_info
 sys._emscripten_info(
-    emscripten_version=(3, 1, 8),
-    runtime='Mozilla/5.0 (X11; Fedora; Linux x86_64; rv:99.0) Gecko/20100101 Firefox/99.0',
+    emscripten_version=(3, 1, 10),
+    runtime='Mozilla/5.0 (X11; Linux x86_64; rv:104.0) Gecko/20100101 Firefox/104.0',
     pthreads=False,
     shared_memory=False
 )
+```
+
+```python
 >>> sys._emscripten_info
-sys._emscripten_info(emscripten_version=(3, 1, 8), runtime='Node.js v14.18.2', pthreads=True, shared_memory=True)
+sys._emscripten_info(
+    emscripten_version=(3, 1, 19),
+    runtime='Node.js v14.18.2',
+    pthreads=True,
+    shared_memory=True
+)
 ```
 
 ```python
 >>> import os, sys
 >>> os.uname()
-posix.uname_result(sysname='wasi', nodename='(none)', release='0.0.0', version='0.0.0', machine='wasm32')
+posix.uname_result(
+    sysname='wasi',
+    nodename='(none)',
+    release='0.0.0',
+    version='0.0.0',
+    machine='wasm32'
+)
 >>> os.name
 'posix'
 >>> sys.platform
@@ -418,7 +471,8 @@ embuilder build --pic zlib bzip2 MINIMAL_PIC
 
 **NOTE**: WASI-SDK's clang may show a warning on Fedora:
 ``/lib64/libtinfo.so.6: no version information available``,
-[RHBZ#1875587](https://bugzilla.redhat.com/show_bug.cgi?id=1875587).
+[RHBZ#1875587](https://bugzilla.redhat.com/show_bug.cgi?id=1875587). The
+warning can be ignored.
 
 ```shell
 export WASI_VERSION=16
@@ -443,6 +497,8 @@ ln -srf -t /usr/local/bin/ ~/.wasmtime/bin/wasmtime
 
 ### WASI debugging
 
-* ``wasmtime run -g`` generates debugging symbols for gdb and lldb.
+* ``wasmtime run -g`` generates debugging symbols for gdb and lldb. The
+  feature is currently broken, see
+  https://github.com/bytecodealliance/wasmtime/issues/4669 .
 * The environment variable ``RUST_LOG=wasi_common`` enables debug and
   trace logging.

Tools/wasm/wasi-env

@@ -72,4 +72,5 @@ export CFLAGS LDFLAGS
 export PKG_CONFIG_PATH PKG_CONFIG_LIBDIR PKG_CONFIG_SYSROOT_DIR
 export PATH
 
-exec "$@"
+# no exec, it makes arvg[0] path absolute.
+"$@"

Tools/wasm/wasm_assets.py

@@ -116,6 +116,14 @@
     "unittest/test/",
 )
 
+SYSCONFIG_NAMES = (
+    "_sysconfigdata__emscripten_wasm32-emscripten",
+    "_sysconfigdata__emscripten_wasm32-emscripten",
+    "_sysconfigdata__wasi_wasm32-wasi",
+    "_sysconfigdata__wasi_wasm64-wasi",
+)
+
+
 def get_builddir(args: argparse.Namespace) -> pathlib.Path:
     """Get builddir path from pybuilddir.txt
     """
@@ -128,7 +136,11 @@ def get_sysconfigdata(args: argparse.Namespace) -> pathlib.Path:
     """Get path to sysconfigdata relative to build root
     """
     data_name = sysconfig._get_sysconfigdata_name()
-    assert "emscripten_wasm32" in data_name
+    if not data_name.startswith(SYSCONFIG_NAMES):
+        raise ValueError(
+            f"Invalid sysconfig data name '{data_name}'.",
+            SYSCONFIG_NAMES
+        )
     filename = data_name + ".py"
     return args.builddir / filename