feat: add GNU_PROPERTY_X86_ISA_1_NEEDED detection (#535)

* chore: use a dataclass for ldd result

* feat: add GNU_PROPERTY_X86_ISA_1_NEEDED detection

ISA extensions usage is not defined by a PEP yet.

This first implementation fails to repair the wheel if the usage of x86-64-v[2-4] is required.

The check can be disabled with `--disable-isa-ext-check`.

The detection being related to a declaration when building, it will not detect the requirement for binaries where the declaration is missing.

All executables built on a manylinux_2_34  image will be detected as x86-64-v2.

Author: mayeut

scripts/create-arch-wheels.sh

@@ -0,0 +1,27 @@
+#!/bin/sh
+
+# This script is used to create wheels for unsupported architectures
+# in order to extend coverage and check errors with those.
+
+set -eux
+
+SCRIPT_DIR="$(CDPATH='' cd -- "$(dirname -- "$0")" && pwd -P)"
+INTEGRATION_TEST_DIR="${SCRIPT_DIR}/../tests/integration"
+mkdir -p "${INTEGRATION_TEST_DIR}/arch-wheels"
+
+# "mips64le" built with buildpack-deps:bookworm and renamed cp313-cp313
+# "386" "amd64" "arm/v5" "arm/v7" "arm64/v8"
+for ARCH in  "ppc64le" "riscv64" "s390x"; do
+  docker run --platform linux/${ARCH} -i --rm -v "${INTEGRATION_TEST_DIR}:/tests" debian:trixie-20250203 << "EOF"
+# for, "arm/v5" QEMU will report armv7l, running on aarch64 will report aarch64, force armv5l/armv7l
+case "$(dpkg --print-architecture)" in
+  armel) export _PYTHON_HOST_PLATFORM="linux-armv5l";;
+  armhf) export _PYTHON_HOST_PLATFORM="linux-armv7l";;
+  *) ;;
+esac
+DEBIAN_FRONTEND=noninteractive apt-get update
+DEBIAN_FRONTEND=noninteractive apt-get install -y --no-install-recommends gcc python3-pip python3-dev
+python3 -m pip wheel --no-deps -w /tests/arch-wheels /tests/testsimple
+EOF
+
+done

src/auditwheel/architecture.py

@@ -0,0 +1,71 @@
+from __future__ import annotations
+
+import functools
+import platform
+import struct
+import sys
+from enum import Enum
+
+
+class Architecture(Enum):
+    value: str
+
+    aarch64 = "aarch64"
+    armv7l = "armv7l"
+    i686 = "i686"
+    loongarch64 = "loongarch64"
+    ppc64 = "ppc64"
+    ppc64le = "ppc64le"
+    riscv64 = "riscv64"
+    s390x = "s390x"
+    x86_64 = "x86_64"
+    x86_64_v2 = "x86_64_v2"
+    x86_64_v3 = "x86_64_v3"
+    x86_64_v4 = "x86_64_v4"
+
+    def __str__(self):
+        return self.value
+
+    @property
+    def baseline(self):
+        if self.value.startswith("x86_64"):
+            return Architecture.x86_64
+        return self
+
+    @classmethod
+    @functools.lru_cache(None)
+    def _member_list(cls) -> list[Architecture]:
+        return list(cls)
+
+    def is_subset(self, other: Architecture) -> bool:
+        if self.baseline != other.baseline:
+            return False
+        member_list = Architecture._member_list()
+        return member_list.index(self) <= member_list.index(other)
+
+    def is_superset(self, other: Architecture) -> bool:
+        if self.baseline != other.baseline:
+            return False
+        return other.is_subset(self)
+
+    @staticmethod
+    def get_native_architecture(*, bits: int | None = None) -> Architecture:
+        machine = platform.machine()
+        if sys.platform.startswith("win"):
+            machine = {"AMD64": "x86_64", "ARM64": "aarch64", "x86": "i686"}.get(
+                machine, machine
+            )
+        elif sys.platform.startswith("darwin"):
+            machine = {"arm64": "aarch64"}.get(machine, machine)
+
+        if bits is None:
+            # c.f. https://github.com/pypa/packaging/pull/711
+            bits = 8 * struct.calcsize("P")
+
+        if machine in {"x86_64", "i686"}:
+            machine = {64: "x86_64", 32: "i686"}[bits]
+        elif machine in {"aarch64", "armv8l"}:
+            # use armv7l policy for 64-bit arm kernel in 32-bit mode (armv8l)
+            machine = {64: "aarch64", 32: "armv7l"}[bits]
+
+        return Architecture(machine)

src/auditwheel/json.py

@@ -0,0 +1,21 @@
+from __future__ import annotations
+
+import dataclasses
+import json
+from enum import Enum
+from typing import Any
+
+
+def _encode_value(value: Any) -> Any:
+    if dataclasses.is_dataclass(value) and not isinstance(value, type):
+        return dataclasses.asdict(value)
+    if isinstance(value, frozenset):
+        return sorted(value)
+    if isinstance(value, Enum):
+        return repr(value)
+    msg = f"object of type {value.__class__.__name__!r} can't be encoded to JSON"
+    raise TypeError(msg)
+
+
+def dumps(obj: Any) -> str:
+    return json.dumps(obj, indent=4, default=_encode_value)