cpython/Tools/wasm
YAMAMOTO Takashi d8f87cdf94
gh-101538: Add experimental wasi-threads build (#101537)
Co-authored-by: Brett Cannon <brett@python.org>
Co-authored-by: Erlend E. Aasland <erlend.aasland@protonmail.com>
2023-06-22 11:26:10 -07:00
..
.editorconfig gh-84461: Improve WebAssembly in-browser demo (#91879) 2022-07-01 11:52:58 +02:00
README.md GH-102404, GH-100956: Document how to do a WASI build (GH-105251) 2023-06-02 15:15:41 -07:00
Setup.local.example gh-104773: PEP 594: Remove the audioop module (#104937) 2023-05-25 17:59:00 +02:00
build_wasi.sh GH-102404, GH-100956: Document how to do a WASI build (GH-105251) 2023-06-02 15:15:41 -07:00
config.site-wasm32-emscripten gh-98948: Remove obsolete readelf dependency (#98949) 2022-11-03 14:31:37 +01:00
config.site-wasm32-wasi gh-100340: Allows -Wno-int-conversion for wasm (#100341) 2023-01-18 19:43:49 +01:00
python.html gh-84461: Tool/wasm/python.html: Add subresource integrity hashes (#93953) 2022-07-01 11:54:30 +02:00
python.worker.js gh-84461: Improve WebAssembly in-browser demo (#91879) 2022-07-01 11:52:58 +02:00
wasi-env gh-95853: Add script to automate WASM build (GH-95828) 2022-08-13 21:56:08 +02:00
wasm_assets.py gh-104773: PEP 594: Remove the audioop module (#104937) 2023-05-25 17:59:00 +02:00
wasm_build.py gh-101538: Add experimental wasi-threads build (#101537) 2023-06-22 11:26:10 -07:00
wasm_webserver.py bpo-40280: WASM docs and smaller browser builds (GH-32412) 2022-04-10 09:29:51 +02:00

README.md

Python WebAssembly (WASM) build

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). 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, running tests, etc.

NOTE: If you are looking for information that is not directly related to building CPython for WebAssembly (or the resulting build), please see https://github.com/psf/webassembly for more information.

wasm32-emscripten

Build

For now the build system has two target flavors. The Emscripten/browser target (--with-emscripten-target=browser) is optimized for browsers. It comes with a reduced and preloaded stdlib without tests and threading support. The Emscripten/node target has threading enabled and can access the file system directly.

Cross compiling to the wasm32-emscripten platform needs the Emscripten SDK and a build Python interpreter. Emscripten 3.1.19 or newer are recommended. All commands below are relative to a repository checkout.

Toolchain

Container image

Christian Heimes maintains a container image with Emscripten SDK, Python build dependencies, WASI-SDK, wasmtime, and several additional tools.

From within your local CPython repo clone, run one of the following commands:

# Fedora, RHEL, CentOS
podman run --rm -ti -v $(pwd):/python-wasm/cpython:Z -w /python-wasm/cpython quay.io/tiran/cpythonbuild:emsdk3

# other
docker run --rm -ti -v $(pwd):/python-wasm/cpython -w /python-wasm/cpython quay.io/tiran/cpythonbuild:emsdk3
Manually
Install Emscripten SDK

NOTE: Follow the on-screen instructions how to add the SDK to PATH.

git clone https://github.com/emscripten-core/emsdk.git /opt/emsdk
/opt/emsdk/emsdk install latest
/opt/emsdk/emsdk activate latest
Optionally: enable ccache for EMSDK

The EM_COMPILER_WRAPPER must be set after the EMSDK environment is sourced. Otherwise the source script removes the environment variable.

. /opt/emsdk/emsdk_env.sh
EM_COMPILER_WRAPPER=ccache
Optionally: pre-build and cache static libraries

Emscripten SDK provides static builds of core libraries without PIC (position-independent code). Python builds with dlopen support require PIC. To populate the build cache, run:

. /opt/emsdk/emsdk_env.sh
embuilder build zlib bzip2 MINIMAL_PIC
embuilder build --pic zlib bzip2 MINIMAL_PIC

Compile a build Python interpreter

From within the container, run the following command:

./Tools/wasm/wasm_build.py build

The command is roughly equivalent to:

mkdir -p builddir/build
pushd builddir/build
../../configure -C
make -j$(nproc)
popd

Cross-compile to wasm32-emscripten for browser

./Tools/wasm/wasm_build.py emscripten-browser

The command is roughly equivalent to:

mkdir -p builddir/emscripten-browser
pushd builddir/emscripten-browser

CONFIG_SITE=../../Tools/wasm/config.site-wasm32-emscripten \
  emconfigure ../../configure -C \
    --host=wasm32-unknown-emscripten \
    --build=$(../../config.guess) \
    --with-emscripten-target=browser \
    --with-build-python=$(pwd)/../build/python

emmake make -j$(nproc)
popd

Serve python.html with a local webserver and open the file in a browser. Python comes with a minimal web server script that sets necessary HTTP headers like COOP, COEP, and mimetypes. Run the script outside the container and from the root of the CPython checkout.

./Tools/wasm/wasm_webserver.py

and open http://localhost:8000/builddir/emscripten-browser/python.html . This directory structure enables the C/C++ DevTools Support (DWARF) to load C and header files with debug builds.

Cross compile to wasm32-emscripten for node

./Tools/wasm/wasm_build.py emscripten-browser-dl

The command is roughly equivalent to:

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
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)

Limitations and issues

Emscripten before 3.1.8 has known bugs that can cause memory corruption and resource leaks. 3.1.8 contains several fixes for bugs in date and time functions.

Network stack

  • Python's socket module does not work with Emscripten's emulated POSIX sockets yet. Network modules like asyncio, urllib, selectors, etc. are not available.
  • Only AF_INET and AF_INET6 with SOCK_STREAM (TCP) or SOCK_DGRAM (UDP) are available. AF_UNIX is not supported.
  • socketpair does not work.
  • Blocking sockets are not available and non-blocking sockets don't work correctly, e.g. socket.accept crashes the runtime. gethostbyname does not resolve to a real IP address. IPv6 is not available.
  • The select module is limited. select.select() crashes the runtime due to lack of exectfd support.

processes, signals

  • Processes are not supported. System calls like fork, popen, and subprocess fail with ENOSYS or ENOSUP.
  • Signal support is limited. signal.alarm, itimer, sigaction are not available or do not work correctly. SIGTERM exits the runtime.
  • Keyboard interrupt (CTRL+C) handling is not implemented yet.
  • Resource-related functions like os.nice and most functions of the resource module are not available.

threading

  • Threading is disabled by default. The configure option --enable-wasm-pthreads adds compiler flag -pthread and linker flags -sUSE_PTHREADS -sPROXY_TO_PTHREAD.
  • pthread support requires WASM threads and SharedArrayBuffer (bulk memory). The Node.JS runtime keeps a pool of web workers around. Each web worker uses several file descriptors (eventfd, epoll, pipe).
  • It's not advised to enable threading when building for browsers or with dynamic linking support; there are performance and stability issues.

file system

  • Most user, group, and permission related function and modules are not supported or don't work as expected, e.g.pwd module, grp module, os.setgroups, os.chown, and so on. lchown and lchmod are not available.
  • umask is a no-op.
  • hard links (os.link) are not supported.
  • Offset and iovec I/O functions (e.g. os.pread, os.preadv) are not available.
  • os.mknod and os.mkfifo don't work and are disabled.
  • Large file support crashes the runtime and is disabled.
  • mmap module is unstable. flush (msync) can crash the runtime.

Misc

  • Heap memory and stack size are limited. Recursion or extensive memory consumption can crash Python.
  • Most stdlib modules with a dependency on external libraries are missing, e.g. ctypes, readline, ssl, and more.
  • Shared extension modules are not implemented yet. All extension modules are statically linked into the main binary. The experimental configure option --enable-wasm-dynamic-linking enables dynamic extensions supports. It's currently known to crash in combination with threading.
  • glibc extensions for date and time formatting are not available.
  • locales module is affected by musl libc issues, gh-90548.
  • Python's object allocator obmalloc is disabled by default.
  • ensurepip is not available.
  • Some ctypes features like c_longlong and c_longdouble may need NodeJS option --experimental-wasm-bigint.

In the browser

  • The interactive shell does not handle copy 'n paste and unicode support well.
  • The bundled stdlib is limited. Network-related modules, multiprocessing, dbm, tests and similar modules are not shipped. All other modules are bundled as pre-compiled pyc files.
  • In-memory file system (MEMFS) is not persistent and limited.
  • Test modules are disabled by default. Use --enable-test-modules build test modules like _testcapi.

wasm32-emscripten in node

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 and gh-95878.

Hosting Python WASM builds

The simple REPL terminal uses SharedArrayBuffer. For security reasons browsers only provide the feature in secure environents with cross-origin isolation. The webserver must send cross-origin headers and correct MIME types for the JavaScript and WASM files. Otherwise the terminal will fail to load with an error message like Browsers disable shared array buffer.

Apache HTTP .htaccess

Place a .htaccess file in the same directory as python.wasm.

# .htaccess
Header set Cross-Origin-Opener-Policy same-origin
Header set Cross-Origin-Embedder-Policy require-corp

AddType application/javascript js
AddType application/wasm wasm

<IfModule mod_deflate.c>
    AddOutputFilterByType DEFLATE text/html application/javascript application/wasm
</IfModule>

WASI (wasm32-wasi)

WASI builds require the WASI SDK 16.0+. See .devcontainer/Dockerfile for an example of how to download and install the WASI SDK.

Build

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.

There are two scripts you can use to do a WASI build from a source checkout. You can either use:

./Tools/wasm/wasm_build.py wasi build

or:

./Tools/wasm/build_wasi.sh

The commands are equivalent to the following steps:

  • Make sure Modules/Setup.local exists
  • Make sure the necessary build tools are installed:
    • WASI SDK (which ships with clang)
    • make
    • pkg-config (on Linux)
  • Create the build Python
    • mkdir -p builddir/build
    • pushd builddir/build
    • Get the build platform
      • Python: sysconfig.get_config_var("BUILD_GNU_TYPE")
      • Shell: ../../config.guess
    • ../../configure -C
    • make all
    • PYTHON_VERSION=`./python -c 'import sys; print(f"{sys.version_info.major}.{sys.version_info.minor}")'`
    • popd
  • Create the host/WASI Python
    • mkdir builddir/wasi
    • pushd builddir/wasi
    • ../../Tools/wasm/wasi-env ../../configure -C --host=wasm32-unknown-wasi --build=$(../../config.guess) --with-build-python=../build/python
      • CONFIG_SITE=../../Tools/wasm/config.site-wasm32-wasi
      • HOSTRUNNER="wasmtime run --mapdir /::$(dirname $(dirname $(pwd))) --env PYTHONPATH=/builddir/wasi/build/lib.wasi-wasm32-$PYTHON_VERSION $(pwd)/python.wasm --"
        • Maps the source checkout to / in the WASI runtime
        • Stdlib gets loaded from /Lib
        • Gets _sysconfigdata__wasi_wasm32-wasi.py on to sys.path via PYTHONPATH
      • Set by wasi-env
        • WASI_SDK_PATH
        • WASI_SYSROOT
        • CC
        • CPP
        • CXX
        • LDSHARED
        • AR
        • RANLIB
        • CFLAGS
        • LDFLAGS
        • PKG_CONFIG_PATH
        • PKG_CONFIG_LIBDIR
        • PKG_CONFIG_SYSROOT_DIR
        • PATH
    • make all

Running

If you followed the instructions above, you can run the interpreter via e.g., wasmtime from within the Tools/wasi directory (make sure to set/change $PYTHON_VERSION and do note the paths are relative to running inbuilddir/wasi for simplicity only):

wasmtime run --mapdir /::../.. --env PYTHONPATH=/builddir/wasi/build/lib.wasi-wasm32-$PYTHON_VERSION python.wasm -- <args>

There are also helpers provided by Tools/wasm/wasm_build.py as listed below. Also, if you used Tools/wasm/build_wasi.sh, a run_wasi.sh file will be created in builddir/wasi which will run the above command for you (it also uses absolute paths, so it can be executed from anywhere).

REPL

./Tools/wasm/wasm_build.py wasi repl

Tests

./Tools/wasm/wasm_build.py wasi test

Debugging

Detect WebAssembly builds

Python code

import os, sys

if sys.platform == "emscripten":
    # Python on Emscripten
if sys.platform == "wasi":
    # Python on WASI

if os.name == "posix":
    # WASM platforms identify as POSIX-like.
    # Windows does not provide os.uname().
    machine = os.uname().machine
    if machine.startswith("wasm"):
        # WebAssembly (wasm32, wasm64 in the future)
>>> import os, sys
>>> os.uname()
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, 10),
    runtime='Mozilla/5.0 (X11; Linux x86_64; rv:104.0) Gecko/20100101 Firefox/104.0',
    pthreads=False,
    shared_memory=False
)
>>> sys._emscripten_info
sys._emscripten_info(
    emscripten_version=(3, 1, 19),
    runtime='Node.js v14.18.2',
    pthreads=True,
    shared_memory=True
)
>>> import os, sys
>>> os.uname()
posix.uname_result(
    sysname='wasi',
    nodename='(none)',
    release='0.0.0',
    version='0.0.0',
    machine='wasm32'
)
>>> os.name
'posix'
>>> sys.platform
'wasi'

C code

Emscripten SDK and WASI SDK define several built-in macros. You can dump a full list of built-ins with emcc -dM -E - < /dev/null and /path/to/wasi-sdk/bin/clang -dM -E - < /dev/null.

  • WebAssembly __wasm__ (also __wasm)
  • wasm32 __wasm32__ (also __wasm32)
  • wasm64 __wasm64__
  • Emscripten __EMSCRIPTEN__ (also EMSCRIPTEN)
  • Emscripten version __EMSCRIPTEN_major__, __EMSCRIPTEN_minor__, __EMSCRIPTEN_tiny__
  • WASI __wasi__