The test now calls free_tool_id() so it can be run multiple times in
the same process. For example, the following command no longer fails:
python -m test test_monitoring -R 3:3
def test_gh108976(self):
sys.monitoring.use_tool_id(0, "test")
+ self.addCleanup(sys.monitoring.free_tool_id, 0)
sys.monitoring.set_events(0, 0)
sys.monitoring.register_callback(0, E.LINE, lambda *args: sys.monitoring.set_events(0, 0))
sys.monitoring.register_callback(0, E.INSTRUCTION, lambda *args: 0)