> -----Original Message----- > From: Steve Wise [mailto:swise@xxxxxxxxxxxxxxxxxxxxx] > Sent: Friday, April 27, 2018 2:21 PM > To: Parav Pandit <parav@xxxxxxxxxxxx>; 'Saleem, Shiraz' > <shiraz.saleem@xxxxxxxxx>; 'Raju Rangoju' <rajur@xxxxxxxxxxx> > Cc: linux-rdma@xxxxxxxxxxxxxxx; 'Hefty, Sean' <sean.hefty@xxxxxxxxx> > Subject: RE: iwarp kernel mode applications are broken with commit f35faa4ba > > > Hi Shiraz, > > > > > -----Original Message----- > > > From: Saleem, Shiraz [mailto:shiraz.saleem@xxxxxxxxx] > > > Sent: Friday, April 27, 2018 1:26 PM > > > To: Parav Pandit <parav@xxxxxxxxxxxx>; Raju Rangoju > > <rajur@xxxxxxxxxxx> > > > Cc: linux-rdma@xxxxxxxxxxxxxxx; SWise OGC > > <swise@xxxxxxxxxxxxxxxxxxxxx>; > > > Hefty, Sean <sean.hefty@xxxxxxxxx> > > > Subject: RE: iwarp kernel mode applications are broken with commit > > f35faa4ba > > > > > > >Subject: RE: iwarp kernel mode applications are broken with commit > > > >f35faa4ba > > > > > > > >> > > > >> > diff --git a/drivers/infiniband/core/cma.c > > > >> > b/drivers/infiniband/core/cma.c index 8512f63..e119cff 100644 > > > >> > --- a/drivers/infiniband/core/cma.c > > > >> > +++ b/drivers/infiniband/core/cma.c > > > >> > @@ -863,7 +863,7 @@ void rdma_destroy_qp(struct rdma_cm_id *id) > > > >> > static int cma_modify_qp_rtr(struct rdma_id_private *id_priv, > > > >> > struct rdma_conn_param *conn_param) { > > > >> > - struct ib_qp_attr qp_attr; > > > >> > + struct ib_qp_attr qp_attr = {}; > > > >> > int qp_attr_mask, ret; > > > >> > union ib_gid sgid; > > > >> > > > > >-- > > > > > > Hi Parav - The patch resolves the issue. > > > But shouldn't we remove ib_query_gid() from cma_modify_qp_rtr() > > altogether? > > > What is its purpose in that function? > > I agree it is not needed. I looked at the commit dd5f03beb4f7 from > > Matan where sgid was used for smac address. > > But now the code has improved a lot over it and this query_gid should > > have been removed when smac thing was refactored, but it was not. > > For this rc cycle I think we just keep it and only make qp_attr {} > > which > is > > anyway good thing to do, just trying to avoid too many changes in rc now. > > But I am fine either ways. > > Let's just initialize qp_attr for this -rc. Please post a patch for Doug to merge. > And tag it for -stable if that is needed. ok Steve, we usually let Leon post it. I have already enqueued to Leon's internal tree, so mostly Sun/Mon it should be out to the ML. -- To unsubscribe from this list: send the line "unsubscribe linux-rdma" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
