[GR-61894] Eagerly initialize native thread state.

graalpython/com.oracle.graal.python.cext/src/capi.c

@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018, 2025, Oracle and/or its affiliates. All rights reserved.
+ * Copyright (c) 2018, 2026, Oracle and/or its affiliates. All rights reserved.
  * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
  *
  * The Universal Permissive License (UPL), Version 1.0
@@ -272,9 +272,13 @@ PyObject* _Py_NotImplementedStructReference;
  */
 THREAD_LOCAL PyThreadState *tstate_current = NULL;
 
-static void initialize_globals() {
-    // store the thread state into a thread local variable
-    tstate_current = GraalPyPrivate_ThreadState_Get(&tstate_current);
+PyAPI_FUNC(PyThreadState **) GraalPyPrivate_InitThreadStateCurrent(PyThreadState *tstate) {
+    tstate_current = tstate;
+    return &tstate_current;
+}
+
+static void initialize_globals(PyThreadState *tstate) {
+    GraalPyPrivate_InitThreadStateCurrent(tstate);
     _Py_NoneStructReference = GraalPyPrivate_None();
     _Py_NotImplementedStructReference = GraalPyPrivate_NotImplemented();
     _Py_EllipsisObjectReference = GraalPyPrivate_Ellipsis();
@@ -667,7 +671,7 @@ Py_LOCAL_SYMBOL TruffleContext* TRUFFLE_CONTEXT;
  */
 Py_LOCAL_SYMBOL int8_t *_graalpy_finalizing = NULL;
 
-PyAPI_FUNC(void) initialize_graal_capi(TruffleEnv* env, void **builtin_closures, GCState *gc) {
+PyAPI_FUNC(PyThreadState **) initialize_graal_capi(TruffleEnv* env, void **builtin_closures, GCState *gc, PyThreadState *tstate) {
     clock_t t = clock();
 
     if (env) {
@@ -706,7 +710,7 @@ PyAPI_FUNC(void) initialize_graal_capi(TruffleEnv* env, void **builtin_closures,
 
     initialize_builtin_types_and_structs();
     // initialize global variables like '_Py_NoneStruct', etc.
-    initialize_globals();
+    initialize_globals(tstate);
     initialize_exceptions();
     initialize_hashes();
     initialize_bufferprocs();
@@ -717,6 +721,7 @@ PyAPI_FUNC(void) initialize_graal_capi(TruffleEnv* env, void **builtin_closures,
     Py_FileSystemDefaultEncoding = "utf-8"; // strdup(PyUnicode_AsUTF8(GraalPyPrivate_FileSystemDefaultEncoding()));
 
     GraalPyPrivate_Log(PY_TRUFFLE_LOG_FINE, "initialize_graal_capi: %fs", ((double) (clock() - t)) / CLOCKS_PER_SEC);
+    return &tstate_current;
 }
 
 /*

graalpython/com.oracle.graal.python.cext/src/pystate.c

@@ -1,4 +1,4 @@
-/* Copyright (c) 2024, 2025, Oracle and/or its affiliates.
+/* Copyright (c) 2024, 2026, Oracle and/or its affiliates.
  * Copyright (C) 1996-2024 Python Software Foundation
  *
  * Licensed under the PYTHON SOFTWARE FOUNDATION LICENSE VERSION 2
@@ -84,9 +84,16 @@ static inline PyThreadState *
 _get_thread_state() {
     PyThreadState *ts = tstate_current;
     if (UNLIKELY(ts == NULL)) {
-         ts = GraalPyPrivate_ThreadState_Get(&tstate_current);
-         tstate_current = ts;
+        /*
+         * Very unlikely fallback: this can happen if another thread initializes the C API while
+         * the current thread is attached to Python but blocked and therefore misses eager
+         * initialization of its native 'tstate_current' TLS slot.
+         */
+        ts = GraalPyPrivate_ThreadState_Get(&tstate_current);
+        assert(ts != NULL);
+        tstate_current = ts;
     }
+    assert(ts != NULL);
     return ts;
 }
 

graalpython/com.oracle.graal.python.test/src/tests/test_datetime.py

@@ -38,6 +38,11 @@
 # SOFTWARE.
 
 import datetime
+import os
+import subprocess
+import sys
+import textwrap
+import time
 import unittest
 
 class DateTest(unittest.TestCase):
@@ -542,12 +547,29 @@ def test_strptime(self):
         actual = datetime.datetime.strptime("+00:00 GMT", "%z %Z")
         self.assertEqual(actual.tzinfo.tzname(None), "GMT")
 
-        import time
         timezone_name = time.localtime().tm_zone
         self.assertIsNotNone(timezone_name)
         actual = datetime.datetime.strptime(f"+00:00 {timezone_name}", "%z %Z")
         self.assertEqual(actual.tzinfo.tzname(None), timezone_name)
 
+        if hasattr(time, "tzset") and sys.executable:
+            proc = subprocess.run(
+                [sys.executable, "-c", textwrap.dedent("""\
+                    import datetime
+                    import time
+
+                    time.tzset()
+                    timezone_name = time.localtime().tm_zone
+                    actual = datetime.datetime.strptime(f"+00:00 {timezone_name}", "%z %Z")
+                    assert actual.tzinfo.tzname(None) == timezone_name
+                """)],
+                env={**os.environ, "TZ": "Etc/GMT-1"},
+                stdout=subprocess.PIPE,
+                stderr=subprocess.PIPE,
+                text=True,
+            )
+            self.assertEqual(proc.returncode, 0, proc.stderr)
+
         # time zone name without utc offset is ignored
         actual = datetime.datetime.strptime("UTC", "%Z")
         self.assertIsNone(actual.tzinfo)

graalpython/com.oracle.graal.python.test/src/tests/test_socket.py

@@ -1,4 +1,4 @@
-# Copyright (c) 2020, 2025, Oracle and/or its affiliates. All rights reserved.
+# Copyright (c) 2020, 2026, Oracle and/or its affiliates. All rights reserved.
 # DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
 #
 # The Universal Permissive License (UPL), Version 1.0
@@ -59,6 +59,13 @@ def test_inet_aton_errs(self):
         self.assertRaises(OSError, lambda : socket.inet_aton('255.255.256.1'))
         self.assertRaises(TypeError, lambda : socket.inet_aton(255))
 
+
+class TestHostLookupErrors(unittest.TestCase):
+    def test_gethostbyname_ex_invalid_host_raises_gaierror(self):
+        with self.assertRaises(socket.gaierror):
+            socket.gethostbyname_ex("nonexistent.invalid")
+
+
 def test_get_name_info():
     import socket
     try :

graalpython/com.oracle.graal.python/src/com/oracle/graal/python/builtins/modules/ImpModuleBuiltins.java

@@ -284,7 +284,7 @@ private static int doExec(Node node, PythonContext context, PythonModule extensi
                 return 0;
             }
 
-            if (!context.hasCApiContext()) {
+            if (context.getCApiState() != PythonContext.CApiState.INITIALIZED) {
                 throw PRaiseNode.raiseStatic(node, PythonBuiltinClassType.SystemError, ErrorMessages.CAPI_NOT_YET_INITIALIZED);
             }
 

graalpython/com.oracle.graal.python/src/com/oracle/graal/python/builtins/modules/SocketModuleBuiltins.java

@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018, 2025, Oracle and/or its affiliates. All rights reserved.
+ * Copyright (c) 2018, 2026, Oracle and/or its affiliates. All rights reserved.
  * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
  *
  * The Universal Permissive License (UPL), Version 1.0
@@ -385,7 +385,7 @@ static Object get(VirtualFrame frame, Object nameObj,
                     addrInfoCursorLib.release(cursor);
                 }
             } catch (GetAddrInfoException e) {
-                throw constructAndRaiseNode.get(inliningTarget).executeWithArgsOnly(frame, SocketHError, new Object[]{e.getMessageAsTruffleString()});
+                throw constructAndRaiseNode.get(inliningTarget).executeWithArgsOnly(frame, SocketGAIError, new Object[]{e.getErrorCode(), e.getMessageAsTruffleString()});
             } catch (PosixException e) {
                 throw constructAndRaiseNode.get(inliningTarget).raiseOSErrorFromPosixException(frame, e);
             }

graalpython/com.oracle.graal.python/src/com/oracle/graal/python/builtins/modules/cext/PythonCextPyStateBuiltins.java

@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021, 2025, Oracle and/or its affiliates. All rights reserved.
+ * Copyright (c) 2021, 2026, Oracle and/or its affiliates. All rights reserved.
  * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
  *
  * The Universal Permissive License (UPL), Version 1.0
@@ -71,16 +71,13 @@
 import com.oracle.graal.python.runtime.GilNode;
 import com.oracle.graal.python.runtime.PythonContext;
 import com.oracle.graal.python.runtime.PythonContext.PythonThreadState;
-import com.oracle.graal.python.runtime.object.PFactory;
 import com.oracle.graal.python.util.OverflowException;
 import com.oracle.truffle.api.CompilerDirectives.TruffleBoundary;
 import com.oracle.truffle.api.ThreadLocalAction;
 import com.oracle.truffle.api.TruffleLogger;
 import com.oracle.truffle.api.dsl.Bind;
 import com.oracle.truffle.api.dsl.Cached;
 import com.oracle.truffle.api.dsl.Specialization;
-import com.oracle.truffle.api.interop.InteropLibrary;
-import com.oracle.truffle.api.library.CachedLibrary;
 import com.oracle.truffle.api.nodes.Node;
 import com.oracle.truffle.api.nodes.RootNode;
 
@@ -116,19 +113,39 @@ static Object restore(
         }
     }
 
-    @CApiBuiltin(ret = PyThreadState, args = {Pointer}, call = Ignored)
+    /**
+     * Very unlikely fallback for threads that were already attached when another thread initialized
+     * the C API, but were blocked at that time and therefore could not process the thread-local
+     * action that eagerly initializes their native 'tstate_current' TLS slot.
+     */
+    @CApiBuiltin(ret = PyThreadState, args = {Pointer}, acquireGil = false, call = Ignored)
     abstract static class GraalPyPrivate_ThreadState_Get extends CApiUnaryBuiltinNode {
+        private static final TruffleLogger LOGGER = CApiContext.getLogger(GraalPyPrivate_ThreadState_Get.class);
 
-        @Specialization(limit = "1")
-        static Object get(Object tstateCurrentPtr,
-                        @Bind Node inliningTarget,
-                        @Bind PythonContext context,
-                        @CachedLibrary("tstateCurrentPtr") InteropLibrary lib) {
-            PythonThreadState pythonThreadState = context.getThreadState(context.getLanguage(inliningTarget));
-            if (!lib.isNull(tstateCurrentPtr)) {
-                pythonThreadState.setNativeThreadLocalVarPointer(tstateCurrentPtr);
+        @Specialization
+        @TruffleBoundary
+        static Object get(Object tstateCurrentPtr) {
+            PythonContext context = PythonContext.get(null);
+            PythonThreadState threadState = context.getThreadState(context.getLanguage());
+
+            /*
+             * The C caller may have observed 'tstate_current == NULL' before entering this upcall.
+             * While entering this builtin, the same thread may process a queued thread-local action
+             * from C API initialization and initialize its native thread state eagerly. So the
+             * fallback decision made in C can be stale by the time we get here.
+             */
+            if (threadState.isNativeThreadStateInitialized()) {
+                LOGGER.fine(() -> String.format("Lazy initialization attempt of native thread state for thread %s aborted. Was initialized in the meantime.", Thread.currentThread()));
+                Object nativeThreadState = PThreadState.getNativeThreadState(threadState);
+                assert nativeThreadState != null;
+                return nativeThreadState;
             }
-            return PThreadState.getOrCreateNativeThreadState(pythonThreadState);
+
+            LOGGER.fine(() -> "Lazy (fallback) initialization of native thread state for thread " + Thread.currentThread());
+            assert PThreadState.getNativeThreadState(threadState) == null;
+            Object nativeThreadState = PThreadState.getOrCreateNativeThreadState(threadState);
+            threadState.setNativeThreadLocalVarPointer(tstateCurrentPtr);
+            return nativeThreadState;
         }
     }
 
@@ -151,12 +168,7 @@ static PDict get(
                         @Bind Node inliningTarget,
                         @Bind PythonContext context) {
             PythonThreadState threadState = context.getThreadState(context.getLanguage(inliningTarget));
-            PDict threadStateDict = threadState.getDict();
-            if (threadStateDict == null) {
-                threadStateDict = PFactory.createDict(context.getLanguage());
-                threadState.setDict(threadStateDict);
-            }
-            return threadStateDict;
+            return PThreadState.getOrCreateThreadStateDict(context, threadState);
         }
     }
 

graalpython/com.oracle.graal.python/src/com/oracle/graal/python/builtins/modules/datetime/DateTimeBuiltins.java

@@ -2310,22 +2310,14 @@ private static Object parse(String string, String format, PythonContext context,
                                 TimeZone timeZone = TimeModuleBuiltins.getGlobalTimeZone(context);
                                 String zoneName = timeZone.getDisplayName(false, TimeZone.SHORT);
                                 String zoneNameDaylightSaving = timeZone.getDisplayName(true, TimeZone.SHORT);
+                                String matchedZoneName = matchTimeZoneName(string, i, zoneName, zoneNameDaylightSaving, "UTC", "GMT");
 
-                                if (string.startsWith("UTC", i)) {
-                                    builder.setTimeZoneName("UTC");
-                                    i += 3;
-                                } else if (string.startsWith("GMT", i)) {
-                                    builder.setTimeZoneName("GMT");
-                                    i += 3;
-                                } else if (string.startsWith(zoneName, i)) {
-                                    builder.setTimeZoneName(zoneName);
-                                    i += zoneName.length();
-                                } else if (string.startsWith(zoneNameDaylightSaving, i)) {
-                                    builder.setTimeZoneName(zoneNameDaylightSaving);
-                                    i += zoneNameDaylightSaving.length();
-                                } else {
+                                if (matchedZoneName == null) {
                                     throw PRaiseNode.raiseStatic(inliningTarget, ValueError, ErrorMessages.TIME_DATA_S_DOES_NOT_MATCH_FORMAT_S, string, format);
                                 }
+
+                                builder.setTimeZoneName(matchedZoneName);
+                                i += matchedZoneName.length();
                             }
                             case 'j' -> {
                                 var pos = new ParsePosition(i);
@@ -2487,6 +2479,16 @@ private static Integer parseDigits(String source, int from, int digitsCount) {
             return result;
         }
 
+        private static String matchTimeZoneName(String string, int from, String... candidates) {
+            String matched = null;
+            for (String candidate : candidates) {
+                if (candidate != null && string.startsWith(candidate, from) && (matched == null || candidate.length() > matched.length())) {
+                    matched = candidate;
+                }
+            }
+            return matched;
+        }
+
         @TruffleBoundary
         private static Integer parseDigitsUpTo(String source, ParsePosition from, int maxDigitsCount) {
             int result = 0;