On 09:09 Wed 11 Mar , Ira Weiny wrote: > > OK! I miss-understood you. I thought you did not want to add another set of > timeout/retry values.
I didn't, but you was against me and thought that per port timeout/retries can be useful. > The below is just fine with me. Do you ever see > getting rid of the global default? I would leave it in order to simplify initial setup for diag tools - IOW don't want to enforce application to call mad_rpc_set_timeout() explicitly, the default should work just fine. Sasha _______________________________________________ general mailing list [email protected] http://lists.openfabrics.org/cgi-bin/mailman/listinfo/general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
