gh-105481: Generate the opcode lists in dis from data extracted from bytecodes.c (#106758)

.gitattributes

@@ -87,7 +87,6 @@ Programs/test_frozenmain.h                          generated
 Python/Python-ast.c                                 generated
 Python/executor_cases.c.h                           generated
 Python/generated_cases.c.h                          generated
-Include/internal/pycore_opcode_metadata.h           generated
 Python/opcode_targets.h                             generated
 Python/stdlib_module_names.h                        generated
 Tools/peg_generator/pegen/grammar_parser.py         generated

Doc/library/dis.rst

@@ -1803,15 +1803,12 @@ instructions:
    Sequence of bytecodes that access an attribute by name.
 
 
-.. data:: hasjrel
+.. data:: hasjump
 
-   Sequence of bytecodes that have a relative jump target.
+   Sequence of bytecodes that have a jump target. All jumps
-
+   are relative.
-
-.. data:: hasjabs
-
-   Sequence of bytecodes that have an absolute jump target.
 
+   .. versionadded:: 3.13
 
 .. data:: haslocal
 
@@ -1827,3 +1824,20 @@ instructions:
    Sequence of bytecodes that set an exception handler.
 
    .. versionadded:: 3.12
+
+
+.. data:: hasjrel
+
+   Sequence of bytecodes that have a relative jump target.
+
+   .. deprecated:: 3.13
+      All jumps are now relative. Use :data:`hasjump`.
+
+
+.. data:: hasjabs
+
+   Sequence of bytecodes that have an absolute jump target.
+
+   .. deprecated:: 3.13
+      All jumps are now relative. This list is empty.
+

Include/cpython/compile.h

@@ -73,3 +73,7 @@ PyAPI_FUNC(int) PyUnstable_OpcodeHasArg(int opcode);
 PyAPI_FUNC(int) PyUnstable_OpcodeHasConst(int opcode);
 PyAPI_FUNC(int) PyUnstable_OpcodeHasName(int opcode);
 PyAPI_FUNC(int) PyUnstable_OpcodeHasJump(int opcode);
+PyAPI_FUNC(int) PyUnstable_OpcodeHasFree(int opcode);
+PyAPI_FUNC(int) PyUnstable_OpcodeHasLocal(int opcode);
+PyAPI_FUNC(int) PyUnstable_OpcodeHasExc(int opcode);
+

Include/internal/pycore_opcode_metadata.h

@@ -955,10 +955,14 @@ enum InstructionFormat { INSTR_FMT_IB, INSTR_FMT_IBC, INSTR_FMT_IBC00, INSTR_FMT
 #define HAS_CONST_FLAG (2)
 #define HAS_NAME_FLAG (4)
 #define HAS_JUMP_FLAG (8)
+#define HAS_FREE_FLAG (16)
+#define HAS_LOCAL_FLAG (32)
 #define OPCODE_HAS_ARG(OP) (_PyOpcode_opcode_metadata[OP].flags & (HAS_ARG_FLAG))
 #define OPCODE_HAS_CONST(OP) (_PyOpcode_opcode_metadata[OP].flags & (HAS_CONST_FLAG))
 #define OPCODE_HAS_NAME(OP) (_PyOpcode_opcode_metadata[OP].flags & (HAS_NAME_FLAG))
 #define OPCODE_HAS_JUMP(OP) (_PyOpcode_opcode_metadata[OP].flags & (HAS_JUMP_FLAG))
+#define OPCODE_HAS_FREE(OP) (_PyOpcode_opcode_metadata[OP].flags & (HAS_FREE_FLAG))
+#define OPCODE_HAS_LOCAL(OP) (_PyOpcode_opcode_metadata[OP].flags & (HAS_LOCAL_FLAG))
 
 struct opcode_metadata {
     bool valid_entry;
@@ -995,16 +999,16 @@ const struct opcode_metadata _PyOpcode_opcode_metadata[OPCODE_METADATA_SIZE] = {
     [NOP] = { true, INSTR_FMT_IX, 0 },
     [RESUME] = { true, INSTR_FMT_IB, HAS_ARG_FLAG },
     [INSTRUMENTED_RESUME] = { true, INSTR_FMT_IB, HAS_ARG_FLAG },
-    [LOAD_CLOSURE] = { true, INSTR_FMT_IB, HAS_ARG_FLAG },
+    [LOAD_CLOSURE] = { true, INSTR_FMT_IB, HAS_ARG_FLAG | HAS_LOCAL_FLAG },
-    [LOAD_FAST_CHECK] = { true, INSTR_FMT_IB, HAS_ARG_FLAG },
+    [LOAD_FAST_CHECK] = { true, INSTR_FMT_IB, HAS_ARG_FLAG | HAS_LOCAL_FLAG },
-    [LOAD_FAST] = { true, INSTR_FMT_IB, HAS_ARG_FLAG },
+    [LOAD_FAST] = { true, INSTR_FMT_IB, HAS_ARG_FLAG | HAS_LOCAL_FLAG },
-    [LOAD_FAST_AND_CLEAR] = { true, INSTR_FMT_IB, HAS_ARG_FLAG },
+    [LOAD_FAST_AND_CLEAR] = { true, INSTR_FMT_IB, HAS_ARG_FLAG | HAS_LOCAL_FLAG },
-    [LOAD_FAST_LOAD_FAST] = { true, INSTR_FMT_IB, HAS_ARG_FLAG },
+    [LOAD_FAST_LOAD_FAST] = { true, INSTR_FMT_IB, HAS_ARG_FLAG | HAS_LOCAL_FLAG },
     [LOAD_CONST] = { true, INSTR_FMT_IB, HAS_ARG_FLAG | HAS_CONST_FLAG },
-    [STORE_FAST] = { true, INSTR_FMT_IB, HAS_ARG_FLAG },
+    [STORE_FAST] = { true, INSTR_FMT_IB, HAS_ARG_FLAG | HAS_LOCAL_FLAG },
-    [STORE_FAST_MAYBE_NULL] = { true, INSTR_FMT_IB, HAS_ARG_FLAG },
+    [STORE_FAST_MAYBE_NULL] = { true, INSTR_FMT_IB, HAS_ARG_FLAG | HAS_LOCAL_FLAG },
-    [STORE_FAST_LOAD_FAST] = { true, INSTR_FMT_IB, HAS_ARG_FLAG },
+    [STORE_FAST_LOAD_FAST] = { true, INSTR_FMT_IB, HAS_ARG_FLAG | HAS_LOCAL_FLAG },
-    [STORE_FAST_STORE_FAST] = { true, INSTR_FMT_IB, HAS_ARG_FLAG },
+    [STORE_FAST_STORE_FAST] = { true, INSTR_FMT_IB, HAS_ARG_FLAG | HAS_LOCAL_FLAG },
     [POP_TOP] = { true, INSTR_FMT_IX, 0 },
     [PUSH_NULL] = { true, INSTR_FMT_IX, 0 },
     [END_FOR] = { true, INSTR_FMT_IB, 0 },
@@ -1028,7 +1032,7 @@ const struct opcode_metadata _PyOpcode_opcode_metadata[OPCODE_METADATA_SIZE] = {
     [BINARY_OP_ADD_FLOAT] = { true, INSTR_FMT_IBC, 0 },
     [BINARY_OP_SUBTRACT_FLOAT] = { true, INSTR_FMT_IBC, 0 },
     [BINARY_OP_ADD_UNICODE] = { true, INSTR_FMT_IBC, 0 },
-    [BINARY_OP_INPLACE_ADD_UNICODE] = { true, INSTR_FMT_IB, 0 },
+    [BINARY_OP_INPLACE_ADD_UNICODE] = { true, INSTR_FMT_IB, HAS_LOCAL_FLAG },
     [BINARY_SUBSCR] = { true, INSTR_FMT_IXC, 0 },
     [BINARY_SLICE] = { true, INSTR_FMT_IX, 0 },
     [STORE_SLICE] = { true, INSTR_FMT_IX, 0 },
@@ -1080,12 +1084,12 @@ const struct opcode_metadata _PyOpcode_opcode_metadata[OPCODE_METADATA_SIZE] = {
     [LOAD_GLOBAL] = { true, INSTR_FMT_IBC000, HAS_ARG_FLAG | HAS_NAME_FLAG },
     [LOAD_GLOBAL_MODULE] = { true, INSTR_FMT_IBC000, HAS_ARG_FLAG },
     [LOAD_GLOBAL_BUILTIN] = { true, INSTR_FMT_IBC000, HAS_ARG_FLAG },
-    [DELETE_FAST] = { true, INSTR_FMT_IB, HAS_ARG_FLAG },
+    [DELETE_FAST] = { true, INSTR_FMT_IB, HAS_ARG_FLAG | HAS_LOCAL_FLAG },
-    [MAKE_CELL] = { true, INSTR_FMT_IB, HAS_ARG_FLAG },
+    [MAKE_CELL] = { true, INSTR_FMT_IB, HAS_ARG_FLAG | HAS_FREE_FLAG },
-    [DELETE_DEREF] = { true, INSTR_FMT_IB, HAS_ARG_FLAG },
+    [DELETE_DEREF] = { true, INSTR_FMT_IB, HAS_ARG_FLAG | HAS_FREE_FLAG },
-    [LOAD_FROM_DICT_OR_DEREF] = { true, INSTR_FMT_IB, HAS_ARG_FLAG },
+    [LOAD_FROM_DICT_OR_DEREF] = { true, INSTR_FMT_IB, HAS_ARG_FLAG | HAS_FREE_FLAG },
-    [LOAD_DEREF] = { true, INSTR_FMT_IB, HAS_ARG_FLAG },
+    [LOAD_DEREF] = { true, INSTR_FMT_IB, HAS_ARG_FLAG | HAS_FREE_FLAG },
-    [STORE_DEREF] = { true, INSTR_FMT_IB, HAS_ARG_FLAG },
+    [STORE_DEREF] = { true, INSTR_FMT_IB, HAS_ARG_FLAG | HAS_FREE_FLAG },
     [COPY_FREE_VARS] = { true, INSTR_FMT_IB, HAS_ARG_FLAG },
     [BUILD_STRING] = { true, INSTR_FMT_IB, HAS_ARG_FLAG },
     [BUILD_TUPLE] = { true, INSTR_FMT_IB, HAS_ARG_FLAG },

Lib/opcode.py

@@ -4,10 +4,12 @@ opcode module - potentially shared between dis and other modules which
 operate on bytecodes (e.g. peephole optimizers).
 """
 
-__all__ = ["cmp_op", "hasarg", "hasconst", "hasname", "hasjrel", "hasjabs",
-           "haslocal", "hascompare", "hasfree", "hasexc", "opname", "opmap",
-           "stack_effect", "HAVE_ARGUMENT", "EXTENDED_ARG"]
 
+# Note that __all__ is further extended below
+__all__ = ["cmp_op", "opname", "opmap", "stack_effect", "hascompare",
+           "HAVE_ARGUMENT", "EXTENDED_ARG"]
+
+import _opcode
 from _opcode import stack_effect
 
 import sys
@@ -17,55 +19,24 @@ if sys.version_info[:2] >= (3, 13):
 
 cmp_op = ('<', '<=', '==', '!=', '>', '>=')
 
-hasarg = []
-hasconst = []
-hasname = []
-hasjrel = []
-hasjabs = []
-haslocal = []
-hascompare = []
-hasfree = []
-hasexc = []
-
 
 ENABLE_SPECIALIZATION = True
 
 def is_pseudo(op):
     return op >= MIN_PSEUDO_OPCODE and op <= MAX_PSEUDO_OPCODE
 
-oplists = [hasarg, hasconst, hasname, hasjrel, hasjabs,
-           haslocal, hascompare, hasfree, hasexc]
-
 opmap = {}
 
-## pseudo opcodes (used in the compiler) mapped to the values
+# pseudo opcodes (used in the compiler) mapped to the values
-## they can become in the actual code.
+# they can become in the actual code.
 _pseudo_ops = {}
 
 def def_op(name, op):
     opmap[name] = op
 
-def name_op(name, op):
-    def_op(name, op)
-    hasname.append(op)
-
-def jrel_op(name, op):
-    def_op(name, op)
-    hasjrel.append(op)
-
-def jabs_op(name, op):
-    def_op(name, op)
-    hasjabs.append(op)
-
 def pseudo_op(name, op, real_ops):
     def_op(name, op)
     _pseudo_ops[name] = real_ops
-    # add the pseudo opcode to the lists its targets are in
-    for oplist in oplists:
-        res = [opmap[rop] in oplist for rop in real_ops]
-        if any(res):
-            assert all(res)
-            oplist.append(op)
 
 
 # Instruction opcodes for compiled code
@@ -137,74 +108,61 @@ def_op('POP_EXCEPT', 89)
 
 HAVE_ARGUMENT = 90             # real opcodes from here have an argument:
 
-name_op('STORE_NAME', 90)       # Index in name list
+def_op('STORE_NAME', 90)       # Index in name list
-name_op('DELETE_NAME', 91)      # ""
+def_op('DELETE_NAME', 91)      # ""
 def_op('UNPACK_SEQUENCE', 92)   # Number of tuple items
-jrel_op('FOR_ITER', 93)
+def_op('FOR_ITER', 93)
 def_op('UNPACK_EX', 94)
-name_op('STORE_ATTR', 95)       # Index in name list
+def_op('STORE_ATTR', 95)       # Index in name list
-name_op('DELETE_ATTR', 96)      # ""
+def_op('DELETE_ATTR', 96)      # ""
-name_op('STORE_GLOBAL', 97)     # ""
+def_op('STORE_GLOBAL', 97)     # ""
-name_op('DELETE_GLOBAL', 98)    # ""
+def_op('DELETE_GLOBAL', 98)    # ""
 def_op('SWAP', 99)
 def_op('LOAD_CONST', 100)       # Index in const list
-hasconst.append(100)
+def_op('LOAD_NAME', 101)       # Index in name list
-name_op('LOAD_NAME', 101)       # Index in name list
 def_op('BUILD_TUPLE', 102)      # Number of tuple items
 def_op('BUILD_LIST', 103)       # Number of list items
 def_op('BUILD_SET', 104)        # Number of set items
 def_op('BUILD_MAP', 105)        # Number of dict entries
-name_op('LOAD_ATTR', 106)       # Index in name list
+def_op('LOAD_ATTR', 106)       # Index in name list
 def_op('COMPARE_OP', 107)       # Comparison operator
-hascompare.append(107)
+def_op('IMPORT_NAME', 108)     # Index in name list
-name_op('IMPORT_NAME', 108)     # Index in name list
+def_op('IMPORT_FROM', 109)     # Index in name list
-name_op('IMPORT_FROM', 109)     # Index in name list
+def_op('JUMP_FORWARD', 110)    # Number of words to skip
-jrel_op('JUMP_FORWARD', 110)    # Number of words to skip
 
-jrel_op('POP_JUMP_IF_FALSE', 114)
+def_op('POP_JUMP_IF_FALSE', 114)
-jrel_op('POP_JUMP_IF_TRUE', 115)
+def_op('POP_JUMP_IF_TRUE', 115)
-name_op('LOAD_GLOBAL', 116)     # Index in name list
+def_op('LOAD_GLOBAL', 116)     # Index in name list
 def_op('IS_OP', 117)
 def_op('CONTAINS_OP', 118)
 def_op('RERAISE', 119)
 def_op('COPY', 120)
 def_op('RETURN_CONST', 121)
-hasconst.append(121)
 def_op('BINARY_OP', 122)
-jrel_op('SEND', 123)            # Number of words to skip
+def_op('SEND', 123)            # Number of words to skip
 def_op('LOAD_FAST', 124)        # Local variable number, no null check
-haslocal.append(124)
 def_op('STORE_FAST', 125)       # Local variable number
-haslocal.append(125)
 def_op('DELETE_FAST', 126)      # Local variable number
-haslocal.append(126)
 def_op('LOAD_FAST_CHECK', 127)  # Local variable number
-haslocal.append(127)
+def_op('POP_JUMP_IF_NOT_NONE', 128)
-jrel_op('POP_JUMP_IF_NOT_NONE', 128)
+def_op('POP_JUMP_IF_NONE', 129)
-jrel_op('POP_JUMP_IF_NONE', 129)
 def_op('RAISE_VARARGS', 130)    # Number of raise arguments (1, 2, or 3)
 def_op('GET_AWAITABLE', 131)
 def_op('BUILD_SLICE', 133)      # Number of items
-jrel_op('JUMP_BACKWARD_NO_INTERRUPT', 134) # Number of words to skip (backwards)
+def_op('JUMP_BACKWARD_NO_INTERRUPT', 134) # Number of words to skip (backwards)
 def_op('MAKE_CELL', 135)
-hasfree.append(135)
 def_op('LOAD_DEREF', 137)
-hasfree.append(137)
 def_op('STORE_DEREF', 138)
-hasfree.append(138)
 def_op('DELETE_DEREF', 139)
-hasfree.append(139)
+def_op('JUMP_BACKWARD', 140)    # Number of words to skip (backwards)
-jrel_op('JUMP_BACKWARD', 140)    # Number of words to skip (backwards)
+def_op('LOAD_SUPER_ATTR', 141)
-name_op('LOAD_SUPER_ATTR', 141)
 def_op('CALL_FUNCTION_EX', 142)  # Flags
 def_op('LOAD_FAST_AND_CLEAR', 143)  # Local variable number
-haslocal.append(143)
 def_op('EXTENDED_ARG', 144)
-EXTENDED_ARG = 144
+EXTENDED_ARG = opmap['EXTENDED_ARG']
 def_op('LIST_APPEND', 145)
 def_op('SET_ADD', 146)
 def_op('MAP_ADD', 147)
-hasfree.append(148)
 def_op('COPY_FREE_VARS', 149)
 def_op('YIELD_VALUE', 150)
 def_op('RESUME', 151)   # This must be kept in sync with deepfreeze.py
@@ -224,12 +182,10 @@ def_op('STORE_FAST_LOAD_FAST', 169)
 def_op('STORE_FAST_STORE_FAST', 170)
 def_op('CALL', 171)
 def_op('KW_NAMES', 172)
-hasconst.append(172)
 def_op('CALL_INTRINSIC_1', 173)
 def_op('CALL_INTRINSIC_2', 174)
-name_op('LOAD_FROM_DICT_OR_GLOBALS', 175)
+def_op('LOAD_FROM_DICT_OR_GLOBALS', 175)
 def_op('LOAD_FROM_DICT_OR_DEREF', 176)
-hasfree.append(176)
 def_op('SET_FUNCTION_ATTRIBUTE', 177)    # Attribute
 
 # Optimizer hook
@@ -258,16 +214,12 @@ def_op('INSTRUMENTED_INSTRUCTION', 253)
 def_op('INSTRUMENTED_LINE', 254)
 # 255 is reserved
 
-hasarg.extend([op for op in opmap.values() if op >= HAVE_ARGUMENT])
 
 MIN_PSEUDO_OPCODE = 256
 
 pseudo_op('SETUP_FINALLY', 256, ['NOP'])
-hasexc.append(256)
 pseudo_op('SETUP_CLEANUP', 257, ['NOP'])
-hasexc.append(257)
 pseudo_op('SETUP_WITH', 258, ['NOP'])
-hasexc.append(258)
 pseudo_op('POP_BLOCK', 259, ['NOP'])
 
 pseudo_op('JUMP', 260, ['JUMP_FORWARD', 'JUMP_BACKWARD'])
@@ -283,12 +235,29 @@ pseudo_op('LOAD_CLOSURE', 267, ['LOAD_FAST'])
 
 MAX_PSEUDO_OPCODE = MIN_PSEUDO_OPCODE + len(_pseudo_ops) - 1
 
-del def_op, name_op, jrel_op, jabs_op, pseudo_op
+del def_op, pseudo_op
 
 opname = ['<%r>' % (op,) for op in range(MAX_PSEUDO_OPCODE + 1)]
 for op, i in opmap.items():
     opname[i] = op
 
+# The build uses older versions of Python which do not have _opcode.has_* functions
+if sys.version_info[:2] >= (3, 13):
+    # These lists are documented as part of the dis module's API
+    hasarg = [op for op in opmap.values() if _opcode.has_arg(op)]
+    hasconst = [op for op in opmap.values() if _opcode.has_const(op)]
+    hasname = [op for op in opmap.values() if _opcode.has_name(op)]
+    hasjump = [op for op in opmap.values() if _opcode.has_jump(op)]
+    hasjrel = hasjump  # for backward compatibility
+    hasjabs = []
+    hasfree = [op for op in opmap.values() if _opcode.has_free(op)]
+    haslocal = [op for op in opmap.values() if _opcode.has_local(op)]
+    hasexc = [op for op in opmap.values() if _opcode.has_exc(op)]
+
+    __all__.extend(["hasarg", "hasconst", "hasname", "hasjump", "hasjrel",
+                    "hasjabs", "hasfree", "haslocal", "hasexc"])
+
+hascompare = [opmap["COMPARE_OP"]]
 
 _nb_ops = [
     ("NB_ADD", "+"),

Lib/test/test__opcode.py

@@ -7,16 +7,7 @@ _opcode = import_module("_opcode")
 from _opcode import stack_effect
 
 
-class OpcodeTests(unittest.TestCase):
+class OpListTests(unittest.TestCase):
-
-    def check_bool_function_result(self, func, ops, expected):
-        for op in ops:
-            if isinstance(op, str):
-                op = dis.opmap[op]
-            with self.subTest(opcode=op, func=func):
-                self.assertIsInstance(func(op), bool)
-                self.assertEqual(func(op), expected)
-
     def test_invalid_opcodes(self):
         invalid = [-100, -1, 255, 512, 513, 1000]
         self.check_bool_function_result(_opcode.is_valid, invalid, False)
@@ -24,6 +15,9 @@ class OpcodeTests(unittest.TestCase):
         self.check_bool_function_result(_opcode.has_const, invalid, False)
         self.check_bool_function_result(_opcode.has_name, invalid, False)
         self.check_bool_function_result(_opcode.has_jump, invalid, False)
+        self.check_bool_function_result(_opcode.has_free, invalid, False)
+        self.check_bool_function_result(_opcode.has_local, invalid, False)
+        self.check_bool_function_result(_opcode.has_exc, invalid, False)
 
     def test_is_valid(self):
         names = [
@@ -36,43 +30,24 @@ class OpcodeTests(unittest.TestCase):
         opcodes = [dis.opmap[opname] for opname in names]
         self.check_bool_function_result(_opcode.is_valid, opcodes, True)
 
-    def test_has_arg(self):
+    def test_oplists(self):
-        has_arg = ['SWAP', 'LOAD_FAST', 'INSTRUMENTED_POP_JUMP_IF_TRUE', 'JUMP']
+        def check_function(self, func, expected):
-        no_arg = ['SETUP_WITH', 'POP_TOP', 'NOP', 'CACHE']
+            for op in [-10, 520]:
-        self.check_bool_function_result(_opcode.has_arg, has_arg, True)
+                with self.subTest(opcode=op, func=func):
-        self.check_bool_function_result(_opcode.has_arg, no_arg, False)
+                    res = func(op)
+                    self.assertIsInstance(res, bool)
+                    self.assertEqual(res, op in expected)
 
-    def test_has_const(self):
+        check_function(self, _opcode.has_arg, dis.hasarg)
-        has_const = ['LOAD_CONST', 'RETURN_CONST', 'KW_NAMES']
+        check_function(self, _opcode.has_const, dis.hasconst)
-        no_const = ['SETUP_WITH', 'POP_TOP', 'NOP', 'CACHE']
+        check_function(self, _opcode.has_name, dis.hasname)
-        self.check_bool_function_result(_opcode.has_const, has_const, True)
+        check_function(self, _opcode.has_jump, dis.hasjump)
-        self.check_bool_function_result(_opcode.has_const, no_const, False)
+        check_function(self, _opcode.has_free, dis.hasfree)
+        check_function(self, _opcode.has_local, dis.haslocal)
+        check_function(self, _opcode.has_exc, dis.hasexc)
 
-    def test_has_name(self):
-        has_name = ['STORE_NAME', 'DELETE_ATTR', 'STORE_GLOBAL', 'IMPORT_FROM',
-                    'LOAD_FROM_DICT_OR_GLOBALS']
-        no_name = ['SETUP_WITH', 'POP_TOP', 'NOP', 'CACHE']
-        self.check_bool_function_result(_opcode.has_name, has_name, True)
-        self.check_bool_function_result(_opcode.has_name, no_name, False)
-
-    def test_has_jump(self):
-        has_jump = ['FOR_ITER', 'JUMP_FORWARD', 'JUMP', 'POP_JUMP_IF_TRUE', 'SEND']
-        no_jump = ['SETUP_WITH', 'POP_TOP', 'NOP', 'CACHE']
-        self.check_bool_function_result(_opcode.has_jump, has_jump, True)
-        self.check_bool_function_result(_opcode.has_jump, no_jump, False)
-
-    # the following test is part of the refactor, it will be removed soon
-    def test_against_legacy_bool_values(self):
-        # limiting to ops up to ENTER_EXECUTOR, because everything after that
-        # is not currently categorized correctly in opcode.py.
-        for op in range(0, opcode.opmap['ENTER_EXECUTOR']):
-            with self.subTest(op=op):
-                if opcode.opname[op] != f'<{op}>':
-                    self.assertEqual(op in dis.hasarg, _opcode.has_arg(op))
-                    self.assertEqual(op in dis.hasconst, _opcode.has_const(op))
-                    self.assertEqual(op in dis.hasname, _opcode.has_name(op))
-                    self.assertEqual(op in dis.hasjrel, _opcode.has_jump(op))
 
+class OpListTests(unittest.TestCase):
     def test_stack_effect(self):
         self.assertEqual(stack_effect(dis.opmap['POP_TOP']), -1)
         self.assertEqual(stack_effect(dis.opmap['BUILD_SLICE'], 0), -1)

Misc/NEWS.d/next/Library/2023-07-17-16-46-00.gh-issue-105481.fek_Nn.rst

@@ -0,0 +1 @@
+The various opcode lists in the :mod:`dis` module are now generated from bytecodes.c instead of explicitly constructed in opcode.py.

Modules/_opcode.c

@@ -147,6 +147,63 @@ _opcode_has_jump_impl(PyObject *module, int opcode)
 
 /*[clinic input]
 
+_opcode.has_free -> bool
+
+  opcode: int
+
+Return True if the opcode accesses a free variable, False otherwise.
+
+Note that 'free' in this context refers to names in the current scope
+that are referenced by inner scopes or names in outer scopes that are
+referenced from this scope. It does not include references to global
+or builtin scopes.
+[clinic start generated code]*/
+
+static int
+_opcode_has_free_impl(PyObject *module, int opcode)
+/*[clinic end generated code: output=d81ae4d79af0ee26 input=117dcd5c19c1139b]*/
+{
+    return PyUnstable_OpcodeIsValid(opcode) &&
+           PyUnstable_OpcodeHasFree(opcode);
+
+}
+
+/*[clinic input]
+
+_opcode.has_local -> bool
+
+  opcode: int
+
+Return True if the opcode accesses a local variable, False otherwise.
+[clinic start generated code]*/
+
+static int
+_opcode_has_local_impl(PyObject *module, int opcode)
+/*[clinic end generated code: output=da5a8616b7a5097b input=9a798ee24aaef49d]*/
+{
+    return PyUnstable_OpcodeIsValid(opcode) &&
+           PyUnstable_OpcodeHasLocal(opcode);
+}
+
+/*[clinic input]
+
+_opcode.has_exc -> bool
+
+  opcode: int
+
+Return True if the opcode sets an exception handler, False otherwise.
+[clinic start generated code]*/
+
+static int
+_opcode_has_exc_impl(PyObject *module, int opcode)
+/*[clinic end generated code: output=41b68dff0ec82a52 input=db0e4bdb9bf13fa5]*/
+{
+    return PyUnstable_OpcodeIsValid(opcode) &&
+           PyUnstable_OpcodeHasExc(opcode);
+}
+
+/*[clinic input]
+
 _opcode.get_specialization_stats
 
 Return the specialization stats
@@ -171,6 +228,9 @@ opcode_functions[] =  {
     _OPCODE_HAS_CONST_METHODDEF
     _OPCODE_HAS_NAME_METHODDEF
     _OPCODE_HAS_JUMP_METHODDEF
+    _OPCODE_HAS_FREE_METHODDEF
+    _OPCODE_HAS_LOCAL_METHODDEF
+    _OPCODE_HAS_EXC_METHODDEF
     _OPCODE_GET_SPECIALIZATION_STATS_METHODDEF
     {NULL, NULL, 0, NULL}
 };

Modules/clinic/_opcode.c.h

@@ -401,6 +401,200 @@ exit:
     return return_value;
 }
 
+PyDoc_STRVAR(_opcode_has_free__doc__,
+"has_free($module, /, opcode)\n"
+"--\n"
+"\n"
+"Return True if the opcode accesses a free variable, False otherwise.\n"
+"\n"
+"Note that \'free\' in this context refers to names in the current scope\n"
+"that are referenced by inner scopes or names in outer scopes that are\n"
+"referenced from this scope. It does not include references to global\n"
+"or builtin scopes.");
+
+#define _OPCODE_HAS_FREE_METHODDEF    \
+    {"has_free", _PyCFunction_CAST(_opcode_has_free), METH_FASTCALL|METH_KEYWORDS, _opcode_has_free__doc__},
+
+static int
+_opcode_has_free_impl(PyObject *module, int opcode);
+
+static PyObject *
+_opcode_has_free(PyObject *module, PyObject *const *args, Py_ssize_t nargs, PyObject *kwnames)
+{
+    PyObject *return_value = NULL;
+    #if defined(Py_BUILD_CORE) && !defined(Py_BUILD_CORE_MODULE)
+
+    #define NUM_KEYWORDS 1
+    static struct {
+        PyGC_Head _this_is_not_used;
+        PyObject_VAR_HEAD
+        PyObject *ob_item[NUM_KEYWORDS];
+    } _kwtuple = {
+        .ob_base = PyVarObject_HEAD_INIT(&PyTuple_Type, NUM_KEYWORDS)
+        .ob_item = { &_Py_ID(opcode), },
+    };
+    #undef NUM_KEYWORDS
+    #define KWTUPLE (&_kwtuple.ob_base.ob_base)
+
+    #else  // !Py_BUILD_CORE
+    #  define KWTUPLE NULL
+    #endif  // !Py_BUILD_CORE
+
+    static const char * const _keywords[] = {"opcode", NULL};
+    static _PyArg_Parser _parser = {
+        .keywords = _keywords,
+        .fname = "has_free",
+        .kwtuple = KWTUPLE,
+    };
+    #undef KWTUPLE
+    PyObject *argsbuf[1];
+    int opcode;
+    int _return_value;
+
+    args = _PyArg_UnpackKeywords(args, nargs, NULL, kwnames, &_parser, 1, 1, 0, argsbuf);
+    if (!args) {
+        goto exit;
+    }
+    opcode = _PyLong_AsInt(args[0]);
+    if (opcode == -1 && PyErr_Occurred()) {
+        goto exit;
+    }
+    _return_value = _opcode_has_free_impl(module, opcode);
+    if ((_return_value == -1) && PyErr_Occurred()) {
+        goto exit;
+    }
+    return_value = PyBool_FromLong((long)_return_value);
+
+exit:
+    return return_value;
+}
+
+PyDoc_STRVAR(_opcode_has_local__doc__,
+"has_local($module, /, opcode)\n"
+"--\n"
+"\n"
+"Return True if the opcode accesses a local variable, False otherwise.");
+
+#define _OPCODE_HAS_LOCAL_METHODDEF    \
+    {"has_local", _PyCFunction_CAST(_opcode_has_local), METH_FASTCALL|METH_KEYWORDS, _opcode_has_local__doc__},
+
+static int
+_opcode_has_local_impl(PyObject *module, int opcode);
+
+static PyObject *
+_opcode_has_local(PyObject *module, PyObject *const *args, Py_ssize_t nargs, PyObject *kwnames)
+{
+    PyObject *return_value = NULL;
+    #if defined(Py_BUILD_CORE) && !defined(Py_BUILD_CORE_MODULE)
+
+    #define NUM_KEYWORDS 1
+    static struct {
+        PyGC_Head _this_is_not_used;
+        PyObject_VAR_HEAD
+        PyObject *ob_item[NUM_KEYWORDS];
+    } _kwtuple = {
+        .ob_base = PyVarObject_HEAD_INIT(&PyTuple_Type, NUM_KEYWORDS)
+        .ob_item = { &_Py_ID(opcode), },
+    };
+    #undef NUM_KEYWORDS
+    #define KWTUPLE (&_kwtuple.ob_base.ob_base)
+
+    #else  // !Py_BUILD_CORE
+    #  define KWTUPLE NULL
+    #endif  // !Py_BUILD_CORE
+
+    static const char * const _keywords[] = {"opcode", NULL};
+    static _PyArg_Parser _parser = {
+        .keywords = _keywords,
+        .fname = "has_local",
+        .kwtuple = KWTUPLE,
+    };
+    #undef KWTUPLE
+    PyObject *argsbuf[1];
+    int opcode;
+    int _return_value;
+
+    args = _PyArg_UnpackKeywords(args, nargs, NULL, kwnames, &_parser, 1, 1, 0, argsbuf);
+    if (!args) {
+        goto exit;
+    }
+    opcode = _PyLong_AsInt(args[0]);
+    if (opcode == -1 && PyErr_Occurred()) {
+        goto exit;
+    }
+    _return_value = _opcode_has_local_impl(module, opcode);
+    if ((_return_value == -1) && PyErr_Occurred()) {
+        goto exit;
+    }
+    return_value = PyBool_FromLong((long)_return_value);
+
+exit:
+    return return_value;
+}
+
+PyDoc_STRVAR(_opcode_has_exc__doc__,
+"has_exc($module, /, opcode)\n"
+"--\n"
+"\n"
+"Return True if the opcode sets an exception handler, False otherwise.");
+
+#define _OPCODE_HAS_EXC_METHODDEF    \
+    {"has_exc", _PyCFunction_CAST(_opcode_has_exc), METH_FASTCALL|METH_KEYWORDS, _opcode_has_exc__doc__},
+
+static int
+_opcode_has_exc_impl(PyObject *module, int opcode);
+
+static PyObject *
+_opcode_has_exc(PyObject *module, PyObject *const *args, Py_ssize_t nargs, PyObject *kwnames)
+{
+    PyObject *return_value = NULL;
+    #if defined(Py_BUILD_CORE) && !defined(Py_BUILD_CORE_MODULE)
+
+    #define NUM_KEYWORDS 1
+    static struct {
+        PyGC_Head _this_is_not_used;
+        PyObject_VAR_HEAD
+        PyObject *ob_item[NUM_KEYWORDS];
+    } _kwtuple = {
+        .ob_base = PyVarObject_HEAD_INIT(&PyTuple_Type, NUM_KEYWORDS)
+        .ob_item = { &_Py_ID(opcode), },
+    };
+    #undef NUM_KEYWORDS
+    #define KWTUPLE (&_kwtuple.ob_base.ob_base)
+
+    #else  // !Py_BUILD_CORE
+    #  define KWTUPLE NULL
+    #endif  // !Py_BUILD_CORE
+
+    static const char * const _keywords[] = {"opcode", NULL};
+    static _PyArg_Parser _parser = {
+        .keywords = _keywords,
+        .fname = "has_exc",
+        .kwtuple = KWTUPLE,
+    };
+    #undef KWTUPLE
+    PyObject *argsbuf[1];
+    int opcode;
+    int _return_value;
+
+    args = _PyArg_UnpackKeywords(args, nargs, NULL, kwnames, &_parser, 1, 1, 0, argsbuf);
+    if (!args) {
+        goto exit;
+    }
+    opcode = _PyLong_AsInt(args[0]);
+    if (opcode == -1 && PyErr_Occurred()) {
+        goto exit;
+    }
+    _return_value = _opcode_has_exc_impl(module, opcode);
+    if ((_return_value == -1) && PyErr_Occurred()) {
+        goto exit;
+    }
+    return_value = PyBool_FromLong((long)_return_value);
+
+exit:
+    return return_value;
+}
+
 PyDoc_STRVAR(_opcode_get_specialization_stats__doc__,
 "get_specialization_stats($module, /)\n"
 "--\n"
@@ -418,4 +612,4 @@ _opcode_get_specialization_stats(PyObject *module, PyObject *Py_UNUSED(ignored))
 {
     return _opcode_get_specialization_stats_impl(module);
 }
-/*[clinic end generated code: output=ae2b2ef56d582180 input=a9049054013a1b77]*/
+/*[clinic end generated code: output=e507bf14fb2796f8 input=a9049054013a1b77]*/

Python/bytecodes.c

@@ -3604,7 +3604,7 @@ dummy_func(
             _PyBinaryOpCache *cache = (_PyBinaryOpCache *)next_instr;
             if (ADAPTIVE_COUNTER_IS_ZERO(cache->counter)) {
                 next_instr--;
-                _Py_Specialize_BinaryOp(lhs, rhs, next_instr, oparg, &GETLOCAL(0));
+                _Py_Specialize_BinaryOp(lhs, rhs, next_instr, oparg, LOCALS_ARRAY);
                 DISPATCH_SAME_OPARG();
             }
             STAT_INC(BINARY_OP, deferred);

Python/ceval_macros.h

@@ -234,6 +234,7 @@ GETITEM(PyObject *v, Py_ssize_t i) {
 
 /* Local variable macros */
 
+#define LOCALS_ARRAY    (frame->localsplus)
 #define GETLOCAL(i)     (frame->localsplus[i])
 
 /* The SETLOCAL() macro must not DECREF the local variable in-place and

Python/compile.c

@@ -896,6 +896,24 @@ PyUnstable_OpcodeHasJump(int opcode)
     return OPCODE_HAS_JUMP(opcode);
 }
 
+int
+PyUnstable_OpcodeHasFree(int opcode)
+{
+    return OPCODE_HAS_FREE(opcode);
+}
+
+int
+PyUnstable_OpcodeHasLocal(int opcode)
+{
+    return OPCODE_HAS_LOCAL(opcode);
+}
+
+int
+PyUnstable_OpcodeHasExc(int opcode)
+{
+    return IS_BLOCK_PUSH_OPCODE(opcode);
+}
+
 static int
 codegen_addop_noarg(instr_sequence *seq, int opcode, location loc)
 {

Python/executor_cases.c.h

@@ -2449,7 +2449,7 @@
             _PyBinaryOpCache *cache = (_PyBinaryOpCache *)next_instr;
             if (ADAPTIVE_COUNTER_IS_ZERO(cache->counter)) {
                 next_instr--;
-                _Py_Specialize_BinaryOp(lhs, rhs, next_instr, oparg, &GETLOCAL(0));
+                _Py_Specialize_BinaryOp(lhs, rhs, next_instr, oparg, LOCALS_ARRAY);
                 DISPATCH_SAME_OPARG();
             }
             STAT_INC(BINARY_OP, deferred);

Python/generated_cases.c.h

@@ -4451,7 +4451,7 @@
             _PyBinaryOpCache *cache = (_PyBinaryOpCache *)next_instr;
             if (ADAPTIVE_COUNTER_IS_ZERO(cache->counter)) {
                 next_instr--;
-                _Py_Specialize_BinaryOp(lhs, rhs, next_instr, oparg, &GETLOCAL(0));
+                _Py_Specialize_BinaryOp(lhs, rhs, next_instr, oparg, LOCALS_ARRAY);
                 DISPATCH_SAME_OPARG();
             }
             STAT_INC(BINARY_OP, deferred);

Tools/build/generate_opcode_h.py

@@ -84,13 +84,7 @@ def main(opcode_py,
     opcode = get_python_module_dict(opcode_py)
     opmap = opcode['opmap']
     opname = opcode['opname']
-    hasarg = opcode['hasarg']
-    hasconst = opcode['hasconst']
-    hasjrel = opcode['hasjrel']
-    hasjabs = opcode['hasjabs']
     is_pseudo = opcode['is_pseudo']
-    _pseudo_ops = opcode['_pseudo_ops']
-
 
     ENABLE_SPECIALIZATION = opcode["ENABLE_SPECIALIZATION"]
     MIN_PSEUDO_OPCODE = opcode["MIN_PSEUDO_OPCODE"]

Tools/cases_generator/generate_cases.py

@@ -261,6 +261,8 @@ class InstructionFlags:
     HAS_CONST_FLAG: bool
     HAS_NAME_FLAG: bool
     HAS_JUMP_FLAG: bool
+    HAS_FREE_FLAG: bool
+    HAS_LOCAL_FLAG: bool
 
     def __post_init__(self):
         self.bitmask = {
@@ -269,16 +271,25 @@ class InstructionFlags:
 
     @staticmethod
     def fromInstruction(instr: "AnyInstruction"):
+
+        has_free = (variable_used(instr, "PyCell_New") or
+                    variable_used(instr, "PyCell_GET") or
+                    variable_used(instr, "PyCell_SET"))
+
         return InstructionFlags(
             HAS_ARG_FLAG=variable_used(instr, "oparg"),
             HAS_CONST_FLAG=variable_used(instr, "FRAME_CO_CONSTS"),
             HAS_NAME_FLAG=variable_used(instr, "FRAME_CO_NAMES"),
             HAS_JUMP_FLAG=variable_used(instr, "JUMPBY"),
+            HAS_FREE_FLAG=has_free,
+            HAS_LOCAL_FLAG=(variable_used(instr, "GETLOCAL") or
+                            variable_used(instr, "SETLOCAL")) and
+                            not has_free,
         )
 
     @staticmethod
     def newEmpty():
-        return InstructionFlags(False, False, False, False)
+        return InstructionFlags(False, False, False, False, False, False)
 
     def add(self, other: "InstructionFlags") -> None:
         for name, value in dataclasses.asdict(other).items():