The same Test object may be used several times

[Totktonada]

See:

test-run/lib/worker.py

Lines 300 to 304 in 1f6d7ba

	def find_task(self, task_id):
	for cur_task in self.suite.tests:
	if cur_task.id == task_id:
	return cur_task
	raise ValueError('Cannot find test: %s' % str(task_id))

When we run something like ./test/test-run.py -j1 --force foo/bar.test.lua foo/bar.test.lua, the same Test object will be used, because it has the same ID ('foo/bar.test.lua', None) — (test_name, conf_name). Usually it does not lead to any problem, but there are corner cases.

Say, if we report a failure (last log lines and so on), we store the fact that the fail was reported in is_crash_reported in the Test instance. If both runs fails, the second will not show logs.

Not sure, whether it may lead to other problems.