scheduler: Add support for runtime parameter "platforms"

If parameter defined, we allow runtime to use only listed
platforms.
We might have limited set of platforms available in certain lab. Lab owners should have option to define such list.
As per suggestion: kernelci/kernelci-project#350 (comment)

Signed-off-by: Denys Fedoryshchenko <[email protected]>

src/scheduler.py

@@ -189,6 +189,18 @@ def _run(self, sub_id):
             event = self._api_helper.receive_event_data(sub_id)
             for job, runtime, platform, rules in self._sched.get_schedule(event):
                 input_node = self._api.node.get(event['id'])
+                if runtime.config.lab_type == 'lava':
+                    supported_platforms = runtime.platforms()
+                    if supported_platforms and platform not in supported_platforms:
+                        # TODO: debug level
+                        self.log.info(' '.join([
+                            input_node['id'],
+                            runtime.config.name,
+                            platform.name,
+                            job.name,
+                            f"Platform not supported in lab {runtime.config.name}",
+                        ]))
+                        continue
                 if self._api_helper.should_create_node(rules, input_node):
                     self._run_job(job, runtime, platform, input_node)