On Tue, Jul 30, 2019 at 4:03 PM Hsin-Yi Wang <[email protected]> wrote: > > On Wed, Jul 17, 2019 at 10:23 PM Saravana Kannan <[email protected]> wrote: > > > -free_required_tables: > > - _opp_table_free_required_tables(opp_table); > > -put_np: > > - of_node_put(np); > > + for (i = 0; i < src->required_opp_count; i++) { > > + if (src->required_opp_tables[i]) > > + continue; > > + > > + req_np = of_parse_required_opp(src_opp->np, i); > > + if (!req_np) > > + continue; > > + > > + req_table = _find_table_of_opp_np(req_np); > Not yet tested in v4, but in v3: > In _find_table_of_opp_np(), there's a lockdep check: > lockdep_assert_held(&opp_table_lock); > which would lead to lockdep warnings. > > Call trace: > _find_table_of_opp_np > _of_lazy_link_required_tables > dev_pm_opp_xlate_opp
Thanks for testing. I'll keep this in mind based on where the rest of the discussion goes. -Saravana
