lava_rpc_connector: block_wait_for_jobs: Also handle OSError's

We had a case when TimeoutError was thrown during LAVA operation. Catch
base class, OSError, in case some other socker errors may occur either.

Signed-off-by: Paul Sokolovsky <paul.sokolovsky@linaro.org>
Change-Id: I32c40d031ff758f5e813de58a554663f097f2963
diff --git a/tfm_ci_pylib/lava_rpc_connector.py b/tfm_ci_pylib/lava_rpc_connector.py
index 1b88826..72864a8 100644
--- a/tfm_ci_pylib/lava_rpc_connector.py
+++ b/tfm_ci_pylib/lava_rpc_connector.py
@@ -248,11 +248,12 @@
                 # Check if the job is not running
                 try:
                     cur_status = self.get_job_info(job_id)
-                except xmlrpc.client.ProtocolError as e:
+                except (xmlrpc.client.ProtocolError, OSError) as e:
                     # There can be transient HTTP errors, e.g. "502 Proxy Error"
+                    # or socket timeout.
                     # Just continue with the next job, the faulted one will be
                     # re-checked on next iteration.
-                    _log.warning("block_wait_for_jobs: xmlrpc.client.ProtocolError %s occurred, ignore and continue", e.errmsg)
+                    _log.warning("block_wait_for_jobs: %r occurred, ignore and continue", e)
                     time.sleep(2)
                     continue
                 # If in queue or running wait