On Tue, Nov 13, 2018 at 07:52:26AM -0500, Hal Rosenstock wrote: > On 11/13/2018 1:31 AM, Honggang LI wrote: > > From: Honggang Li <honli@xxxxxxxxxx> > > > > Signed-off-by: Honggang Li <honli@xxxxxxxxxx> > > --- > > opensm/osm_guid_info_rcv.c | 3 +-- > > 1 file changed, 1 insertion(+), 2 deletions(-) > > > > diff --git a/opensm/osm_guid_info_rcv.c b/opensm/osm_guid_info_rcv.c > > index f4d0592e..ea3c2c56 100644 > > --- a/opensm/osm_guid_info_rcv.c > > +++ b/opensm/osm_guid_info_rcv.c > > @@ -74,10 +74,9 @@ void osm_gi_rcv_process(IN void *context, IN void *data) > > ib_net64_t port_guid, node_guid; > > uint8_t block_num; > > > > - CL_ASSERT(sm); > > - > > OSM_LOG_ENTER(sm->p_log); > > > > + CL_ASSERT(sm); > > CL_ASSERT(p_madw); > > > > p_smp = osm_madw_get_smp_ptr(p_madw); > > > > The CL_ASSERT needs to come before the OSM_LOG_ENTER here. Otherwise, > OSM_LOG_ENTER can chase a null pointer in accessing sm->p_log. Yes, you are right. thanks
