[Abhiroop3917]

Type of changes

• Bug fix
• New feature
• Documentation / docstrings
• Tests
• Other

AI?

• AI was used to generate this PR

Checklist

• I've run the latest black with default args on new code.
• I've updated CHANGELOG.md and CONTRIBUTORS.md where appropriate.
• I've added tests for new code.
• I accept that @pytest-dev/pytest may be pedantic in the code review.

Description

When a test mocks time.perf_counter or time.time (as freezegun, time-machine, or monkeypatch do), subtests report negative durations and a stop timestamp of 0.0 (1970-01-01) in JUnit XML output. Normal tests are unaffected by the same mocking.

Root cause: subtests.py imported the time module and called time.time() / time.perf_counter() directly. When a mock is installed inside the subtest body, the start-time is already captured from the real clock, but the stop-time reads the patched value (0.0), giving stop − start < 0.

The rest of pytest uses _pytest.timing which captures real function references at import time (before any test can patch them). This pattern was introduced specifically to prevent this class of bug (issue #185) and extended in #7764. subtests.py was added later and missed it.

Fix: Replace the four time.xxx() calls in subtests.py with timing.xxx().

Reproducer (before fix):

import time

def test_with_subtest(subtests, monkeypatch):
    with subtests.test(msg="frozen clock"):
        monkeypatch.setattr(time, "perf_counter", lambda: 0.0)
        monkeypatch.setattr(time, "time", lambda: 0.0)

<!-- JUnit XML output: -->
<testcase name="test_with_subtest" time="-1194.217" .../>

After fix: duration is always non-negative.

closes #15201

[RonnyPfannschmidt]

i already had a pr open for that since a while