gh-135801: Add tests for filtering warnings by module (GH-140240)

.github/workflows/mypy.yml

@@ -16,6 +16,7 @@ on:
       - "Tools/build/check_extension_modules.py"
       - "Tools/build/check_warnings.py"
       - "Tools/build/compute-changes.py"
+      - "Tools/build/consts_getter.py"
       - "Tools/build/deepfreeze.py"
       - "Tools/build/generate-build-details.py"
       - "Tools/build/generate_sbom.py"

Doc/bugs.rst

@@ -19,6 +19,12 @@ If you find a bug in this documentation or would like to propose an improvement,
 please submit a bug report on the :ref:`issue tracker <using-the-tracker>`.  If you
 have a suggestion on how to fix it, include that as well.
 
+.. only:: translation
+
+   If the bug or suggested improvement concerns the translation of this
+   documentation, submit the report to the
+   `translation’s repository <TRANSLATION_REPO_>`_ instead.
+
 You can also open a discussion item on our
 `Documentation Discourse forum <https://discuss.python.org/c/documentation/26>`_.
 

Doc/c-api/long.rst

@@ -43,7 +43,7 @@ distinguished from a number.  Use :c:func:`PyErr_Occurred` to disambiguate.
    .. impl-detail::
 
       CPython keeps an array of integer objects for all integers
-      between ``-5`` and ``256``.  When you create an int in that range
+      between ``-5`` and ``1024``.  When you create an int in that range
       you actually just get back a reference to the existing object.
 
 

Doc/conf.py

@@ -444,6 +444,25 @@
 # https://github.com/sphinx-doc/sphinx/issues/12359
 epub_use_index = False
 
+# translation tag
+# ---------------
+
+language_code = None
+for arg in sys.argv:
+    if arg.startswith('language='):
+        language_code = arg.split('=', 1)[1]
+
+if language_code:
+    tags.add('translation')  # noqa: F821
+
+    rst_epilog += f"""\
+.. _TRANSLATION_REPO: https://github.com/python/python-docs-{language_code.replace("_", "-").lower()}
+"""  # noqa: F821
+else:
+    rst_epilog += """\
+.. _TRANSLATION_REPO: https://github.com/python
+"""
+
 # Options for the coverage checker
 # --------------------------------
 

Lib/test/test_warnings/__init__.py

@@ -241,6 +241,85 @@ def test_once(self):
                                     42)
             self.assertEqual(len(w), 0)
 
+    def test_filter_module(self):
+        MS_WINDOWS = (sys.platform == 'win32')
+        with self.module.catch_warnings(record=True) as w:
+            self.module.simplefilter('error')
+            self.module.filterwarnings('always', module=r'package\.module\z')
+            self.module.warn_explicit('msg', UserWarning, 'filename', 42,
+                                      module='package.module')
+            self.assertEqual(len(w), 1)
+            with self.assertRaises(UserWarning):
+                self.module.warn_explicit('msg', UserWarning, '/path/to/package/module', 42)
+            with self.assertRaises(UserWarning):
+                self.module.warn_explicit('msg', UserWarning, '/path/to/package/module.py', 42)
+
+        with self.module.catch_warnings(record=True) as w:
+            self.module.simplefilter('error')
+            self.module.filterwarnings('always', module='package')
+            self.module.warn_explicit('msg', UserWarning, 'filename', 42,
+                                      module='package.module')
+            self.assertEqual(len(w), 1)
+            with self.assertRaises(UserWarning):
+                self.module.warn_explicit('msg', UserWarning, 'filename', 42,
+                                          module='other.package.module')
+            with self.assertRaises(UserWarning):
+                self.module.warn_explicit('msg', UserWarning, '/path/to/otherpackage/module.py', 42)
+
+        with self.module.catch_warnings(record=True) as w:
+            self.module.simplefilter('error')
+            self.module.filterwarnings('always', module=r'/path/to/package/module\z')
+            self.module.warn_explicit('msg', UserWarning, '/path/to/package/module', 42)
+            self.assertEqual(len(w), 1)
+            self.module.warn_explicit('msg', UserWarning, '/path/to/package/module.py', 42)
+            self.assertEqual(len(w), 2)
+            with self.assertRaises(UserWarning):
+                self.module.warn_explicit('msg', UserWarning, '/PATH/TO/PACKAGE/MODULE', 42)
+            if MS_WINDOWS:
+                if self.module is py_warnings:
+                    self.module.warn_explicit('msg', UserWarning, r'/path/to/package/module.PY', 42)
+                    self.assertEqual(len(w), 3)
+                with self.assertRaises(UserWarning):
+                    self.module.warn_explicit('msg', UserWarning, r'/path/to/package/module/__init__.py', 42)
+                with self.assertRaises(UserWarning):
+                    self.module.warn_explicit('msg', UserWarning, r'/path/to/package/module.pyw', 42)
+                with self.assertRaises(UserWarning):
+                    self.module.warn_explicit('msg', UserWarning, r'\path\to\package\module', 42)
+
+        with self.module.catch_warnings(record=True) as w:
+            self.module.simplefilter('error')
+            self.module.filterwarnings('always', module=r'/path/to/package/__init__\z')
+            self.module.warn_explicit('msg', UserWarning, '/path/to/package/__init__.py', 42)
+            self.assertEqual(len(w), 1)
+            self.module.warn_explicit('msg', UserWarning, '/path/to/package/__init__', 42)
+            self.assertEqual(len(w), 2)
+
+        if MS_WINDOWS:
+            with self.module.catch_warnings(record=True) as w:
+                self.module.simplefilter('error')
+                self.module.filterwarnings('always', module=r'C:\\path\\to\\package\\module\z')
+                self.module.warn_explicit('msg', UserWarning, r'C:\path\to\package\module', 42)
+                self.assertEqual(len(w), 1)
+                self.module.warn_explicit('msg', UserWarning, r'C:\path\to\package\module.py', 42)
+                self.assertEqual(len(w), 2)
+                if self.module is py_warnings:
+                    self.module.warn_explicit('msg', UserWarning, r'C:\path\to\package\module.PY', 42)
+                    self.assertEqual(len(w), 3)
+                with self.assertRaises(UserWarning):
+                    self.module.warn_explicit('msg', UserWarning, r'C:\path\to\package\module.pyw', 42)
+                with self.assertRaises(UserWarning):
+                    self.module.warn_explicit('msg', UserWarning, r'C:\PATH\TO\PACKAGE\MODULE', 42)
+                with self.assertRaises(UserWarning):
+                    self.module.warn_explicit('msg', UserWarning, r'C:/path/to/package/module', 42)
+                with self.assertRaises(UserWarning):
+                    self.module.warn_explicit('msg', UserWarning, r'C:\path\to\package\module\__init__.py', 42)
+
+        with self.module.catch_warnings(record=True) as w:
+            self.module.simplefilter('error')
+            self.module.filterwarnings('always', module=r'<unknown>\z')
+            self.module.warn_explicit('msg', UserWarning, '', 42)
+            self.assertEqual(len(w), 1)
+
     def test_module_globals(self):
         with self.module.catch_warnings(record=True) as w:
             self.module.simplefilter("always", UserWarning)

Python/jit.c

@@ -167,11 +167,13 @@ set_bits(uint32_t *loc, uint8_t loc_start, uint64_t value, uint8_t value_start,
 
 // See https://developer.arm.com/documentation/ddi0602/2023-09/Base-Instructions
 // for instruction encodings:
-#define IS_AARCH64_ADD_OR_SUB(I) (((I) & 0x11C00000) == 0x11000000)
+#define IS_AARCH64_ADD_OR_SUB(I)  (((I) & 0x11C00000) == 0x11000000)
-#define IS_AARCH64_ADRP(I)       (((I) & 0x9F000000) == 0x90000000)
+#define IS_AARCH64_ADRP(I)        (((I) & 0x9F000000) == 0x90000000)
-#define IS_AARCH64_BRANCH(I)     (((I) & 0x7C000000) == 0x14000000)
+#define IS_AARCH64_BRANCH(I)      (((I) & 0x7C000000) == 0x14000000)
-#define IS_AARCH64_LDR_OR_STR(I) (((I) & 0x3B000000) == 0x39000000)
+#define IS_AARCH64_BRANCH_COND(I) (((I) & 0x7C000000) == 0x54000000)
-#define IS_AARCH64_MOV(I)        (((I) & 0x9F800000) == 0x92800000)
+#define IS_AARCH64_TEST_AND_BRANCH(I) (((I) & 0x7E000000) == 0x36000000)
+#define IS_AARCH64_LDR_OR_STR(I)  (((I) & 0x3B000000) == 0x39000000)
+#define IS_AARCH64_MOV(I)         (((I) & 0x9F800000) == 0x92800000)
 
 // LLD is a great reference for performing relocations... just keep in
 // mind that Tools/jit/build.py does filtering and preprocessing for us!
@@ -332,6 +334,21 @@ patch_aarch64_21rx(unsigned char *location, uint64_t value)
     patch_aarch64_21r(location, value);
 }
 
+// 21-bit relative branch.
+void
+patch_aarch64_19r(unsigned char *location, uint64_t value)
+{
+    uint32_t *loc32 = (uint32_t *)location;
+    assert(IS_AARCH64_BRANCH_COND(*loc32));
+    value -= (uintptr_t)location;
+    // Check that we're not out of range of 21 signed bits:
+    assert((int64_t)value >= -(1 << 20));
+    assert((int64_t)value < (1 << 20));
+    // Since instructions are 4-byte aligned, only use 19 bits:
+    assert(get_bits(value, 0, 2) == 0);
+    set_bits(loc32, 5, value, 2, 19);
+}
+
 // 28-bit relative branch.
 void
 patch_aarch64_26r(unsigned char *location, uint64_t value)

Tools/build/consts_getter.py

@@ -0,0 +1,20 @@
+from pathlib import Path
+
+ROOT = Path(__file__).resolve().parents[2]
+INTERNAL = ROOT / "Include" / "internal"
+
+def get_nsmallnegints_and_nsmallposints() -> tuple[int, int]:
+    nsmallposints = None
+    nsmallnegints = None
+    with open(INTERNAL / "pycore_runtime_structs.h") as infile:
+        for line in infile:
+            if line.startswith("#define _PY_NSMALLPOSINTS"):
+                nsmallposints = int(line.split()[-1])
+            elif line.startswith("#define _PY_NSMALLNEGINTS"):
+                nsmallnegints = int(line.split()[-1])
+                break
+        else:
+            raise NotImplementedError
+    assert nsmallposints
+    assert nsmallnegints
+    return nsmallnegints, nsmallposints

Tools/build/deepfreeze.py

@@ -17,6 +17,7 @@
 import time
 import types
 
+import consts_getter
 import umarshal
 
 TYPE_CHECKING = False
@@ -362,7 +363,9 @@ def _generate_int_for_bits(self, name: str, i: int, digit: int) -> None:
                 self.write(f".ob_digit = {{ {ds} }},")
 
     def generate_int(self, name: str, i: int) -> str:
-        if -5 <= i <= 256:
+        nsmallnegints, nsmallposints = consts_getter.get_nsmallnegints_and_nsmallposints()
+
+        if -nsmallnegints <= i <= nsmallposints:
             return f"(PyObject *)&_PyLong_SMALL_INTS[_PY_NSMALLNEGINTS + {i}]"
         if i >= 0:
             name = f"const_int_{i}"

Tools/build/generate_global_objects.py

@@ -3,6 +3,8 @@
 import os.path
 import re
 
+import consts_getter
+
 SCRIPT_NAME = 'Tools/build/generate_global_objects.py'
 __file__ = os.path.abspath(__file__)
 ROOT = os.path.dirname(os.path.dirname(os.path.dirname(__file__)))
@@ -274,20 +276,7 @@ def generate_global_strings(identifiers, strings):
 
 
 def generate_runtime_init(identifiers, strings):
-    # First get some info from the declarations.
+    nsmallnegints, nsmallposints = consts_getter.get_nsmallnegints_and_nsmallposints()
-    nsmallposints = None
-    nsmallnegints = None
-    with open(os.path.join(INTERNAL, 'pycore_runtime_structs.h')) as infile:
-        for line in infile:
-            if line.startswith('#define _PY_NSMALLPOSINTS'):
-                nsmallposints = int(line.split()[-1])
-            elif line.startswith('#define _PY_NSMALLNEGINTS'):
-                nsmallnegints = int(line.split()[-1])
-                break
-        else:
-            raise NotImplementedError
-    assert nsmallposints
-    assert nsmallnegints
 
     # Then target the runtime initializer.
     filename = os.path.join(INTERNAL, 'pycore_runtime_init_generated.h')

Tools/build/mypy.ini

@@ -6,6 +6,7 @@ files =
     Tools/build/check_extension_modules.py,
     Tools/build/check_warnings.py,
     Tools/build/compute-changes.py,
+    Tools/build/consts_getter.py,
     Tools/build/deepfreeze.py,
     Tools/build/generate-build-details.py,
     Tools/build/generate_sbom.py,

Tools/jit/_optimizers.py

@@ -39,6 +39,34 @@
 # Update with all of the inverted branches, too:
 _X86_BRANCHES |= {v: k for k, v in _X86_BRANCHES.items() if v}
 
+_AARCH64_COND_CODES = {
+    # https://developer.arm.com/documentation/dui0801/b/CJAJIHAD?lang=en
+    "eq": "ne",
+    "ne": "eq",
+    "lt": "ge",
+    "ge": "lt",
+    "gt": "le",
+    "le": "gt",
+    "vs": "vc",
+    "vc": "vs",
+    "mi": "pl",
+    "pl": "mi",
+    "cs": "cc",
+    "cc": "cs",
+    "hs": "lo",
+    "lo": "hs",
+    "hi": "ls",
+    "ls": "hi",
+}
+# Branches are either b.{cond} or bc.{cond}
+_AARCH64_BRANCHES = {
+    "b." + cond: ("b." + inverse if inverse else None)
+    for (cond, inverse) in _AARCH64_COND_CODES.items()
+} | {
+    "bc." + cond: ("bc." + inverse if inverse else None)
+    for (cond, inverse) in _AARCH64_COND_CODES.items()
+}
+
 
 @dataclasses.dataclass
 class _Block:
@@ -283,13 +311,28 @@ def run(self) -> None:
         self.path.write_text(self._body())
 
 
-class OptimizerAArch64(Optimizer):  # pylint: disable = too-few-public-methods
+# Mach-O does not support the 19 bit branch locations needed for branch reordering
-    """aarch64-apple-darwin/aarch64-pc-windows-msvc/aarch64-unknown-linux-gnu"""
+class OptimizerAArch64_MachO(Optimizer):  # pylint: disable = too-few-public-methods
+    """aarch64-apple-darwin"""
 
     # https://developer.arm.com/documentation/ddi0602/2025-03/Base-Instructions/B--Branch-
     _re_jump = re.compile(r"\s*b\s+(?P<target>[\w.]+)")
 
 
+class OptimizerAArch64(Optimizer):  # pylint: disable = too-few-public-methods
+    """aarch64-pc-windows-msvc/aarch64-unknown-linux-gnu"""
+
+    _branches = _AARCH64_BRANCHES
+    _re_branch = re.compile(
+        rf"\s*(?P<instruction>{'|'.join(_AARCH64_BRANCHES)})\s+(.+,\s+)*(?P<target>[\w.]+)"
+    )
+
+    # https://developer.arm.com/documentation/ddi0602/2025-03/Base-Instructions/B--Branch-
+    _re_jump = re.compile(r"\s*b\s+(?P<target>[\w.]+)")
+    # https://developer.arm.com/documentation/ddi0602/2025-09/Base-Instructions/RET--Return-from-subroutine-
+    _re_return = re.compile(r"\s*ret\b")
+
+
 class OptimizerX86(Optimizer):  # pylint: disable = too-few-public-methods
     """i686-pc-windows-msvc/x86_64-apple-darwin/x86_64-unknown-linux-gnu"""
 

Tools/jit/_schema.py

@@ -10,6 +10,7 @@
     "ARM64_RELOC_PAGEOFF12",
     "ARM64_RELOC_UNSIGNED",
     "IMAGE_REL_AMD64_REL32",
+    "IMAGE_REL_ARM64_BRANCH19",
     "IMAGE_REL_ARM64_BRANCH26",
     "IMAGE_REL_ARM64_PAGEBASE_REL21",
     "IMAGE_REL_ARM64_PAGEOFFSET_12A",
@@ -20,6 +21,7 @@
     "R_AARCH64_ADR_GOT_PAGE",
     "R_AARCH64_ADR_PREL_PG_HI21",
     "R_AARCH64_CALL26",
+    "R_AARCH64_CONDBR19",
     "R_AARCH64_JUMP26",
     "R_AARCH64_ADD_ABS_LO12_NC",
     "R_AARCH64_LD64_GOT_LO12_NC",

Tools/jit/_stencils.py

@@ -61,6 +61,7 @@ class HoleValue(enum.Enum):
     # x86_64-pc-windows-msvc:
     "IMAGE_REL_AMD64_REL32": "patch_x86_64_32rx",
     # aarch64-pc-windows-msvc:
+    "IMAGE_REL_ARM64_BRANCH19": "patch_aarch64_19r",
     "IMAGE_REL_ARM64_BRANCH26": "patch_aarch64_26r",
     "IMAGE_REL_ARM64_PAGEBASE_REL21": "patch_aarch64_21rx",
     "IMAGE_REL_ARM64_PAGEOFFSET_12A": "patch_aarch64_12",
@@ -74,6 +75,7 @@ class HoleValue(enum.Enum):
     "R_AARCH64_ADR_GOT_PAGE": "patch_aarch64_21rx",
     "R_AARCH64_ADR_PREL_PG_HI21": "patch_aarch64_21r",
     "R_AARCH64_CALL26": "patch_aarch64_26r",
+    "R_AARCH64_CONDBR19": "patch_aarch64_19r",
     "R_AARCH64_JUMP26": "patch_aarch64_26r",
     "R_AARCH64_LD64_GOT_LO12_NC": "patch_aarch64_12x",
     "R_AARCH64_MOVW_UABS_G0_NC": "patch_aarch64_16a",

Tools/jit/_targets.py

@@ -335,7 +335,8 @@ def _handle_relocation(
                 "Offset": offset,
                 "Symbol": s,
                 "Type": {
-                    "Name": "IMAGE_REL_ARM64_BRANCH26"
+                    "Name": "IMAGE_REL_ARM64_BRANCH19"
+                    | "IMAGE_REL_ARM64_BRANCH26"
                     | "IMAGE_REL_ARM64_PAGEBASE_REL21"
                     | "IMAGE_REL_ARM64_PAGEOFFSET_12A"
                     | "IMAGE_REL_ARM64_PAGEOFFSET_12L" as kind
@@ -564,7 +565,7 @@ def get_target(host: str) -> _COFF32 | _COFF64 | _ELF | _MachO:
     if re.fullmatch(r"aarch64-apple-darwin.*", host):
         host = "aarch64-apple-darwin"
         condition = "defined(__aarch64__) && defined(__APPLE__)"
-        optimizer = _optimizers.OptimizerAArch64
+        optimizer = _optimizers.OptimizerAArch64_MachO
         target = _MachO(host, condition, optimizer=optimizer)
     elif re.fullmatch(r"aarch64-pc-windows-msvc", host):
         host = "aarch64-pc-windows-msvc"