gh-116622: Fix testPyObjectPrintOSError on Android

[mhsmith]

This test, which was added in #98749, was failing on Android as follows:

17:43:48.804  W  ======================================================================
17:43:48.804  W  FAIL: testPyObjectPrintOSError (test.test_capi.test_object.PrintTest.testPyObjectPrintOSError)
17:43:48.804  W  ----------------------------------------------------------------------
17:43:48.804  W  Traceback (most recent call last):
17:43:48.804  W    File "/data/user/0/org.python.testbed/files/python/lib/python3.13/test/test_capi/test_object.py", line 103, in testPyObjectPrintOSError
17:43:48.804  W      with self.assertRaises(OSError):
17:43:48.804  W          _testcapi.pyobject_print_os_error(output_filename)
17:43:48.804  W  AssertionError: OSError not raised
17:43:48.804  W
17:43:48.804  W  ----------------------------------------------------------------------

This was caused by a recently-fixed bug in OpenBSD's libc, on which Android's libc is based. Luckily there's a simple workaround.

This PR also addresses the comment in #121595 about it being inconvenient to build all architectures before running the testbed. The build script now allows architectures to be missing, as long as at least one is present.

• Issue: Add support for Android as a target platform #116622

[mhsmith]

CI failure seems unrelated:

• Thread sanitizer tests failing in GitHub actions #121946

[freakboy3742]

Annoying, but makes sense to me.

[miss-islington-app]

Thanks @mhsmith for the PR, and @freakboy3742 for merging it 🌮🎉.. I'm working now to backport this PR to: 3.13.
🐍🍒⛏🤖

[bedevere-app]

GH-122490 is a backport of this pull request to the 3.13 branch.