GeezerVison

test_getoutput fails on Windows whenever the console code page is not UTF-8: getstatusoutput() decodes the shell's localized error message as UTF-8 and raises.

File "Lib/subprocess.py", line 1382, in communicate
    stdout = self.stdout.read()
UnicodeDecodeError: 'utf-8' codec can't decode byte 0xd5 in position 4

UTF-8 is correct for Python's side of the pipe — the code page is the child's choice, not Python's — but this test only needs the exit status, so let the decoder replace the bytes it cannot decode. That keeps getstatusoutput() under test instead of switching to run(), as suggested in the issue.
Verified on zh-CN Windows 11 (CP936), current main:

# before
$ ./PCbuild/amd64/python_d.exe -m test test_subprocess -m test_getoutput
FAILED (errors=1)
# after
$ ./PCbuild/amd64/python_d.exe -m test test_subprocess -m test_getoutput
OK
$ ./PCbuild/amd64/python_d.exe -m test test_subprocess
Result: SUCCESS (run=371 skipped=216)

3.12 with -X utf8=1 fails the same way, so the test was encoding-dependent since UTF-8 mode existed; 3.15 only made it the default. Skip news (test only).
AI assistance: drafted with deepseek-v4.1-flash in the pi coding agent; commands above were run locally, text reviewed by the author.

• Issue: test_subprocess: test_getoutput fails on non-English Windows, as text mode decodes cmd.exe's OEM output as UTF-8 #160652

Fixes #160652