diff options
| author |   Alex Estrin <alex.estrin@intel.com> | 2017-10-09 12:38:33 -0700 |
|---|---|---|
| committer |   Doug Ledford <dledford@redhat.com> | 2017-10-18 10:13:00 -0400 |
| commit | f9586abfa333143d67e3362dfb2a19ae45d82441 (patch) | |
| tree | 6e6fe5990dff922feea6c4ba9360392d8e591329 /drivers/infiniband/sw/rdmavt | |
| parent | IB/hfi1: Set hdr_type when tx req is allocated (diff) | |
| download | linux-dev-f9586abfa333143d67e3362dfb2a19ae45d82441.tar.xz linux-dev-f9586abfa333143d67e3362dfb2a19ae45d82441.zip | |
IB/rdmavt: Don't wait for resources in QP reset
Per the IBTA spec, QP destroy shall fail if the QP is attached
to multicast groups, although the spec is silent on modify_qp
to reset state. It implies that ULP must deregister QP from
all mcast groups for destroy to succeed.
The faulty patch "IB/ipoib: Update broadcast object if PKey value
was changed in index 0" exposed two issues in rdmavt:
1. Rvt QP reset waits for qp references to go to zero.
This will hang if QP is attached to multicast groups.
2. The mcast group detach will fail for a QP in reset state
therefore preventing ULP from correcting the issue.
This patch moves the reference count wait to the the destroy QP
path and allows a QP mcast detach to work in the reset state.
Reviewed-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Alex Estrin <alex.estrin@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
Diffstat (limited to 'drivers/infiniband/sw/rdmavt')
| -rw-r--r-- | drivers/infiniband/sw/rdmavt/mcast.c | 2 | ||||
| -rw-r--r-- | drivers/infiniband/sw/rdmavt/qp.c | 2 |
2 files changed, 2 insertions, 2 deletions
diff --git a/drivers/infiniband/sw/rdmavt/mcast.c b/drivers/infiniband/sw/rdmavt/mcast.c index 1f12b69a0d07..b3a38c5e4cad 100644 --- a/drivers/infiniband/sw/rdmavt/mcast.c +++ b/drivers/infiniband/sw/rdmavt/mcast.c @@ -351,7 +351,7 @@ int rvt_detach_mcast(struct ib_qp *ibqp, union ib_gid *gid, u16 lid) int last = 0; int ret = 0; - if (ibqp->qp_num <= 1 || qp->state == IB_QPS_RESET) + if (ibqp->qp_num <= 1) return -EINVAL; spin_lock_irq(&ibp->lock); diff --git a/drivers/infiniband/sw/rdmavt/qp.c b/drivers/infiniband/sw/rdmavt/qp.c index 22df09ae809e..367c78618019 100644 --- a/drivers/infiniband/sw/rdmavt/qp.c +++ b/drivers/infiniband/sw/rdmavt/qp.c @@ -717,7 +717,6 @@ static void rvt_reset_qp(struct rvt_dev_info *rdi, struct rvt_qp *qp, /* take qp out the hash and wait for it to be unused */ rvt_remove_qp(rdi, qp); - wait_event(qp->wait, !atomic_read(&qp->refcount)); /* grab the lock b/c it was locked at call time */ spin_lock_irq(&qp->r_lock); @@ -1443,6 +1442,7 @@ int rvt_destroy_qp(struct ib_qp *ibqp) spin_unlock(&qp->s_hlock); spin_unlock_irq(&qp->r_lock); + wait_event(qp->wait, !atomic_read(&qp->refcount)); /* qpn is now available for use again */ rvt_free_qpn(&rdi->qp_dev->qpn_table, qp->ibqp.qp_num); |