Hi,
On Sun, Aug 11, 2024 at 10:41 AM Leon Romanovsky <leon@xxxxxxxxxx> wrote:
>
> On Fri, Aug 09, 2024 at 03:15:27PM +0200, Md Haris Iqbal wrote:
> > From: Jack Wang <jinpu.wang@xxxxxxxxx>
> >
> > In case of error happening during session stablishment, close_work is
> > running. A new RDMA CM event may arrive since we don't destroy cm_id
> > before destroying qp. To fix this, we first destroy cm_id after drain_qp,
> > so no new RDMA CM event will arrive afterwards.
> >
> > Fixes: 9cb837480424 ("RDMA/rtrs: server: main functionality")
> > Signed-off-by: Jack Wang <jinpu.wang@xxxxxxxxx>
> > Signed-off-by: Grzegorz Prajsner <grzegorz.prajsner@xxxxxxxxx>
> > Signed-off-by: Md Haris Iqbal <haris.iqbal@xxxxxxxxx>
> > ---
> > drivers/infiniband/ulp/rtrs/rtrs-srv.c | 2 +-
> > drivers/infiniband/ulp/rtrs/rtrs.c | 2 +-
> > 2 files changed, 2 insertions(+), 2 deletions(-)
> >
> > diff --git a/drivers/infiniband/ulp/rtrs/rtrs-srv.c b/drivers/infiniband/ulp/rtrs/rtrs-srv.c
> > index fb67b58a7f62..90ea25ad6720 100644
> > --- a/drivers/infiniband/ulp/rtrs/rtrs-srv.c
> > +++ b/drivers/infiniband/ulp/rtrs/rtrs-srv.c
> > @@ -1540,6 +1540,7 @@ static void rtrs_srv_close_work(struct work_struct *work)
> > con = to_srv_con(srv_path->s.con[i]);
> > rdma_disconnect(con->c.cm_id);
> > ib_drain_qp(con->c.qp);
> > + rdma_destroy_id(con->c.cm_id);
> > }
> >
> > /*
> > @@ -1564,7 +1565,6 @@ static void rtrs_srv_close_work(struct work_struct *work)
> > continue;
> > con = to_srv_con(srv_path->s.con[i]);
> > rtrs_cq_qp_destroy(&con->c);
> > - rdma_destroy_id(con->c.cm_id);
> > kfree(con);
> > }
> > rtrs_ib_dev_put(srv_path->s.dev);
> > diff --git a/drivers/infiniband/ulp/rtrs/rtrs.c b/drivers/infiniband/ulp/rtrs/rtrs.c
> > index 4e17d546d4cc..44167fd1c958 100644
> > --- a/drivers/infiniband/ulp/rtrs/rtrs.c
> > +++ b/drivers/infiniband/ulp/rtrs/rtrs.c
> > @@ -318,7 +318,7 @@ EXPORT_SYMBOL_GPL(rtrs_cq_qp_create);
> > void rtrs_cq_qp_destroy(struct rtrs_con *con)
> > {
> > if (con->qp) {
> > - rdma_destroy_qp(con->cm_id);
> > + ib_destroy_qp(con->qp);
>
> You created that QP with rdma_create_qp() and you should destroy it with rdma_destroy_qp().
We can't do it, as we move rdma_destroy_id before rtrs_cq_qp_destroy,
if we still call rdma_destroy_qp, which will lead to use after free as
cm_id could already be free-ed.
>
> Thanks
Thx!
>
> > con->qp = NULL;
> > }
> > destroy_cq(con);
> > --
> > 2.25.1
> >
