gh-142982: Specialize CALL_FUNCTION_EX (GH-143391)

Author: Fidget-Spinner

Include/internal/pycore_code.h

@@ -160,6 +160,12 @@ typedef struct {
 
 #define INLINE_CACHE_ENTRIES_CONTAINS_OP CACHE_ENTRIES(_PyContainsOpCache)
 
+typedef struct {
+    _Py_BackoffCounter counter;
+} _PyCallFunctionExCache;
+
+#define INLINE_CACHE_ENTRIES_CALL_FUNCTION_EX CACHE_ENTRIES(_PyCallFunctionExCache)
+
 /* "Locals plus" for a code object is the set of locals + cell vars +
  * free vars.  This relates to variable names as well as offsets into
  * the "fast locals" storage array of execution frames.  The compiler
@@ -326,6 +332,7 @@ PyAPI_FUNC(void) _Py_Specialize_Send(_PyStackRef receiver, _Py_CODEUNIT *instr);
 PyAPI_FUNC(void) _Py_Specialize_ToBool(_PyStackRef value, _Py_CODEUNIT *instr);
 PyAPI_FUNC(void) _Py_Specialize_ContainsOp(_PyStackRef value, _Py_CODEUNIT *instr);
 PyAPI_FUNC(void) _Py_GatherStats_GetIter(_PyStackRef iterable);
+PyAPI_FUNC(void) _Py_Specialize_CallFunctionEx(_PyStackRef func_st, _Py_CODEUNIT *instr);
 
 // Utility functions for reading/writing 32/64-bit values in the inline caches.
 // Great care should be taken to ensure that these functions remain correct and

Include/internal/pycore_interpframe.h

@@ -395,6 +395,10 @@ _PyEvalFramePushAndInit(PyThreadState *tstate, _PyStackRef func,
                         size_t argcount, PyObject *kwnames,
                         _PyInterpreterFrame *previous);
 
+PyAPI_FUNC(_PyInterpreterFrame *)
+_PyEvalFramePushAndInit_Ex(PyThreadState *tstate, _PyStackRef func,
+    PyObject *locals, Py_ssize_t nargs, PyObject *callargs, PyObject *kwargs, _PyInterpreterFrame *previous);
+
 #ifdef __cplusplus
 }
 #endif

Include/internal/pycore_magic_number.h

@@ -289,6 +289,7 @@ Known values:
     Python 3.15a2 3656 (Add TRACE_RECORD instruction, for platforms with switch based interpreter)
     Python 3.15a4 3657 (Add BINARY_OP_SUBSCR_USTR_INT)
     Python 3.15a4 3658 (Optimize bytecode for list/set called on genexp)
+    Python 3.15a4 3659 (Add CALL_FUNCTION_EX specialization)
 
 
     Python 3.16 will start with 3700
@@ -302,7 +303,7 @@ PC/launcher.c must also be updated.
 
 */
 
-#define PYC_MAGIC_NUMBER 3658
+#define PYC_MAGIC_NUMBER 3659
 /* This is equivalent to converting PYC_MAGIC_NUMBER to 2 bytes
    (little-endian) and then appending b'\r\n'. */
 #define PYC_MAGIC_NUMBER_TOKEN \