gh-124621: Emscripten: Support pyrepl

Basic support for pyrepl in Emscripten. Limitations: * requires JSPI * no signal handling implemented As followup work, it would be nice to implement a webworker variant for when JSPI is not available and proper signal handling. Because it requires JSPI, it doesn't work in Safari. Firefox requires setting an experimental flag. All the Chromiums have full support since May. Until we make it work without JSPI, let's keep the original web_example around.
python · ambv · Jul 22, 2025 · Jul 21, 2025 · Jul 21, 2025 · Jul 21, 2025
commit 733846cfa64ce2e81508f568b995a3859ca924d4
@@ -1,6 +1,7 @@
 from __future__ import annotations
 
 import os
+import sys
 
 # types
 if False:
@@ -12,10 +13,22 @@
     trace_file = open(trace_filename, "a")
 
 
-def trace(line: str, *k: object, **kw: object) -> None:
-    if trace_file is None:
-        return
-    if k or kw:
-        line = line.format(*k, **kw)
-    trace_file.write(line + "\n")
-    trace_file.flush()
+
+if sys.platform == "emscripten":
+    from posix import _emscripten_log
+
+    def trace(line: str, *k: object, **kw: object) -> None:
+        if "PYREPL_TRACE" not in os.environ:
+            return
+        if k or kw:
+            line = line.format(*k, **kw)
+        _emscripten_log(line)
+
+else:
+    def trace(line: str, *k: object, **kw: object) -> None:
+        if trace_file is None:
+            return
+        if k or kw:
+            line = line.format(*k, **kw)
+        trace_file.write(line + "\n")
+        trace_file.flush()
@@ -804,7 +804,7 @@ build_wasm: check-clean-src $(BUILDPYTHON) platform sharedmods \
 		python-config checksharedmods
 
 .PHONY: build_emscripten
-build_emscripten: build_wasm web_example
+build_emscripten: build_wasm web_example web_example_pyrepl_jspi
 
 # Check that the source is clean when building out of source.
 .PHONY: check-clean-src
@@ -1095,26 +1095,28 @@ $(DLLLIBRARY) libpython$(LDVERSION).dll.a: $(LIBRARY_OBJS)
 
 # wasm32-emscripten browser web example
 
-WEBEX_DIR=$(srcdir)/Tools/wasm/emscripten/web_example/
+EMSCRIPTEN_DIR=$(srcdir)/Tools/wasm/emscripten
+WEBEX_DIR=$(EMSCRIPTEN_DIR)/web_example/
+
+ZIP_STDLIB=python$(VERSION)$(ABI_THREAD).zip
+$(ZIP_STDLIB): $(srcdir)/Lib/*.py $(srcdir)/Lib/*/*.py \
+	    $(EMSCRIPTEN_DIR)/wasm_assets.py \
+	    Makefile pybuilddir.txt Modules/Setup.local
+	$(PYTHON_FOR_BUILD) $(EMSCRIPTEN_DIR)/wasm_assets.py \
+	    --buildroot . --prefix $(prefix) -o $@
+
 web_example/index.html: $(WEBEX_DIR)/index.html
 	@mkdir -p web_example
 	@cp $< $@
 
-web_example/python.worker.mjs: $(WEBEX_DIR)/python.worker.mjs
+web_example/server.py: $(WEBEX_DIR)/server.py
 	@mkdir -p web_example
 	@cp $< $@
 
-web_example/server.py: $(WEBEX_DIR)/server.py
+web_example/$(ZIP_STDLIB): $(ZIP_STDLIB)
 	@mkdir -p web_example
 	@cp $< $@
 
-WEB_STDLIB=web_example/python$(VERSION)$(ABI_THREAD).zip
-$(WEB_STDLIB): $(srcdir)/Lib/*.py $(srcdir)/Lib/*/*.py \
-	    $(WEBEX_DIR)/wasm_assets.py \
-	    Makefile pybuilddir.txt Modules/Setup.local
-	$(PYTHON_FOR_BUILD) $(WEBEX_DIR)/wasm_assets.py \
-	    --buildroot . --prefix $(prefix) -o $@
-
 web_example/python.mjs web_example/python.wasm: $(BUILDPYTHON)
 	@if test $(HOST_GNU_TYPE) != 'wasm32-unknown-emscripten' ; then \
 		echo "Can only build web_example when target is Emscripten" ;\
@@ -1124,7 +1126,39 @@ web_example/python.mjs web_example/python.wasm: $(BUILDPYTHON)
 	cp python.wasm web_example/python.wasm
 
 .PHONY: web_example
-web_example: web_example/python.mjs web_example/python.worker.mjs web_example/index.html web_example/server.py $(WEB_STDLIB)
+web_example: web_example/python.mjs web_example/index.html web_example/src.mjs web_example/server.py web_example/$(ZIP_STDLIB)
+
+WEBEX2=web_example_pyrepl_jspi
+WEBEX2_DIR=$(EMSCRIPTEN_DIR)/$(WEBEX2)/
+
+$(WEBEX2)/python.mjs $(WEBEX2)/python.wasm: $(BUILDPYTHON)
+	@if test $(HOST_GNU_TYPE) != 'wasm32-unknown-emscripten' ; then \
+		echo "Can only build web_example when target is Emscripten" ;\
+		exit 1 ;\
+	fi
+	@mkdir -p $(WEBEX2)
+	@cp python.mjs $(WEBEX2)/python.mjs
+	@cp python.wasm $(WEBEX2)/python.wasm
+
+$(WEBEX2)/index.html: $(WEBEX2_DIR)/index.html
+	@mkdir -p $(WEBEX2)
+	@cp $< $@
+
+$(WEBEX2)/src.mjs: $(WEBEX2_DIR)/src.mjs
+	@mkdir -p $(WEBEX2)
+	@cp $< $@
+
+$(WEBEX2)/$(ZIP_STDLIB): $(ZIP_STDLIB)
+	@mkdir -p $(WEBEX2)
+	@cp $< $@
+
+$(WEBEX2)/src.mjs: $(WEBEX2_DIR)/src.mjs
+	@mkdir -p $(WEBEX2)
+	@cp $< $@
+
+.PHONY: web_example_pyrepl_jspi
+web_example_pyrepl_jspi: $(WEBEX2)/python.mjs $(WEBEX2)/index.html $(WEBEX2)/src.mjs $(WEBEX2)/$(ZIP_STDLIB)
+
 
 ############################################################################
 # Header files

diff --git a/Modules/clinic/posixmodule.c.h b/Modules/clinic/posixmodule.c.h
diff --git a/Modules/posixmodule.c b/Modules/posixmodule.c
@@ -16971,6 +16971,25 @@ os__emscripten_debugger_impl(PyObject *module)
     emscripten_debugger();
     Py_RETURN_NONE;
 }
+
+EM_JS(void, emscripten_log_impl_js, (const char* arg), {
+    console.warn(UTF8ToString(arg));
+});
+
+/*[clinic input]
+os._emscripten_log
+   arg: str
+
+Log something to the JS console. Emscripten only.
+[clinic start generated code]*/
+
+static PyObject *
+os__emscripten_log_impl(PyObject *module, const char *arg)
+/*[clinic end generated code: output=9749e5e293c42784 input=350aa1f70bc1e905]*/
+{
+    emscripten_log_impl_js(arg);
+    Py_RETURN_NONE;
+}
 #endif /* __EMSCRIPTEN__ */
 
 
@@ -17190,6 +17209,7 @@ static PyMethodDef posix_methods[] = {
     OS__IS_INPUTHOOK_INSTALLED_METHODDEF
     OS__CREATE_ENVIRON_METHODDEF
     OS__EMSCRIPTEN_DEBUGGER_METHODDEF
+    OS__EMSCRIPTEN_LOG_METHODDEF
     {NULL,              NULL}            /* Sentinel */
 };
 

diff --git a/Python/emscripten_syscalls.c b/Python/emscripten_syscalls.c
@@ -1,4 +1,5 @@
 #include "emscripten.h"
+#include "stdio.h"
 
 // If we're running in node, report the UID of the user in the native system as
 // the UID of the user. Since the nodefs will report the uid correctly, if we
@@ -40,6 +41,7 @@ int __syscall_umask(int mask) {
 
 #include <wasi/api.h>
 #include <errno.h>
+#include <fcntl.h>
 #undef errno
 
 // Variant of EM_JS that does C preprocessor substitution on the body
@@ -100,7 +102,7 @@ EM_JS_MACROS(void, _emscripten_promising_main_js, (void), {
         return;
     }
     const origResolveGlobalSymbol = resolveGlobalSymbol;
-    if (!Module.onExit && globalThis?.process?.exit) {
+    if (ENVIRONMENT_IS_NODE && !Module.onExit) {
         Module.onExit = (code) => process.exit(code);
     }
     // * wrap the main symbol with WebAssembly.promising,
@@ -115,7 +117,7 @@ EM_JS_MACROS(void, _emscripten_promising_main_js, (void), {
             orig.sym = (...args) => {
                 (async () => {
                     const ret = await main(...args);
-                    process?.exit?.(ret);
+                    Module.onExit?.(ret);
                 })();
                 _emscripten_exit_with_live_runtime();
             };
@@ -199,23 +201,101 @@ __wasi_errno_t __wasi_fd_read_orig(__wasi_fd_t fd, const __wasi_iovec_t *iovs,
 
 // Take a promise that resolves to __wasi_errno_t and suspend until it resolves,
 // get the output.
-EM_JS(__wasi_errno_t, __block_for_errno, (__externref_t p), {
+EM_JS(int, __block_for_int, (__externref_t p), {
     return p;
 }
 if (WebAssembly.Suspending) {
-    __block_for_errno = new WebAssembly.Suspending(__block_for_errno);
+    __block_for_int = new WebAssembly.Suspending(__block_for_int);
 }
 )
 
 // Replacement for fd_read syscall. Call __maybe_fd_read_async. If it returned
-// null, delegate back to __wasi_fd_read_orig. Otherwise, use __block_for_errno
+// null, delegate back to __wasi_fd_read_orig. Otherwise, use __block_for_int
 // to get the result.
 __wasi_errno_t __wasi_fd_read(__wasi_fd_t fd, const __wasi_iovec_t *iovs,
                               size_t iovs_len, __wasi_size_t *nread) {
   __externref_t p = __maybe_fd_read_async(fd, iovs, iovs_len, nread);
   if (__builtin_wasm_ref_is_null_extern(p)) {
     return __wasi_fd_read_orig(fd, iovs, iovs_len, nread);
   }
-  __wasi_errno_t res = __block_for_errno(p);
-  return res;
+  return __block_for_int(p);
+}
+
+#include <poll.h>
+#define POLLFD_FD 0
+#define POLLFD_EVENTS 4
+#define POLLFD_REVENTS 6
+#define POLLFD_SIZE 8
+_Static_assert(offsetof(struct pollfd, fd) == 0, "Unepxected pollfd struct layout");
+_Static_assert(offsetof(struct pollfd, events) == 4, "Unepxected pollfd struct layout");
+_Static_assert(offsetof(struct pollfd, revents) == 6, "Unepxected pollfd struct layout");
+_Static_assert(sizeof(struct pollfd) == 8, "Unepxected pollfd struct layout");
+
+EM_JS_MACROS(__externref_t, __maybe_poll_async, (intptr_t fds, int nfds, int timeout), {
+    if (!WebAssembly.promising) {
+        return null;
+    }
+    return (async function() {
+        try {
+            var nonzero = 0;
+            var promises = [];
+            for (var i = 0; i < nfds; i++) {
+                var pollfd = fds + POLLFD_SIZE * i;
+                var fd = HEAP32[(pollfd + POLLFD_FD)/4];
+                var events = HEAP16[(pollfd + POLLFD_EVENTS)/2];
+                var mask = POLLNVAL;
+                var stream = FS.getStream(fd);
+                if (stream) {
+                    mask = POLLIN | POLLOUT;
+                    if (stream.stream_ops.pollAsync) {
+                        promises.push(stream.stream_ops.pollAsync(stream, timeout).then((mask) => {
+                            mask &= events | POLLERR | POLLHUP;
+                            HEAP16[(pollfd + POLLFD_REVENTS)/2] = mask;
+                            if (mask) {
+                                nonzero ++;
+                            }
+                        }));
+                    } else if (stream.stream_ops.poll) {
+                        var mask = stream.stream_ops.poll(stream, timeout);
+                        mask &= events | POLLERR | POLLHUP;
+                        HEAP16[(pollfd + POLLFD_REVENTS)/2] = mask;
+                        if (mask) {
+                            nonzero ++;
+                        }
+                    }
+                }
+            }
+            await Promise.all(promises);
+            return nonzero;
+        } catch(e) {
+            if (e?.name !== "ErrnoError") throw e;
+            return -e.errno;
+        }
+    })();
+});
+
+// Bind original poll syscall to syscall_poll_orig().
+int syscall_poll_orig(intptr_t fds, int nfds, int timeout)
+    __attribute__((__import_module__("env"),
+                   __import_name__("__syscall_poll"), __warn_unused_result__));
+
+int __syscall_poll(intptr_t fds, int nfds, int timeout) {
+    __externref_t p = __maybe_poll_async(fds, nfds, timeout);
+    if (__builtin_wasm_ref_is_null_extern(p)) {
+        return syscall_poll_orig(fds, nfds, timeout);
+    }
+    return __block_for_int(p);
+}
+
+#include <sys/ioctl.h>
+
+int syscall_ioctl_orig(int fd, int request, void* varargs)
+    __attribute__((__import_module__("env"),
+                   __import_name__("__syscall_ioctl"), __warn_unused_result__));
+
+int __syscall_ioctl(int fd, int request, void* varargs) {
+    if (request == FIOCLEX || request == FIONCLEX) {
+        return 0;
+    }
+    return syscall_ioctl_orig(fd, request, varargs);
 }
diff --git a/Tools/wasm/config.site-wasm32-emscripten b/Tools/wasm/config.site-wasm32-emscripten
@@ -69,7 +69,6 @@ ac_cv_func_posix_fallocate=no
 
 # Syscalls that resulted in a segfault
 ac_cv_func_utimensat=no
-ac_cv_header_sys_ioctl_h=no
 
 # sockets are supported, but only AF_INET / AF_INET6 in non-blocking mode.
 # Disable AF_UNIX and AF_PACKET support, see socketmodule.h.