IB/qib: Avoid returning EBUSY from MR deregister

A timing issue can occur where qib_mr_dereg can return -EBUSY if the
MR use count is not zero.

This can occur if the MR is de-registered while RDMA read response
packets are being progressed from the SDMA ring.  The suspicion is
that the peer sent an RDMA read request, which has already been copied
across to the peer.  The peer sees the completion of his request and
then communicates to the responder that the MR is not needed any
longer.  The responder tries to de-register the MR, catching some
responses remaining in the SDMA ring holding the MR use count.

The code now uses a get/put paradigm to track MR use counts and
coordinates with the MR de-registration process using a completion
when the count has reached zero.  A timeout on the delay is in place
to catch other EBUSY issues.

The reference count protocol is as follows:
- The return to the user counts as 1
- A reference from the lk_table or the qib_ibdev counts as 1.
- Transient I/O operations increase/decrease as necessary

A lot of code duplication has been folded into the new routines
init_qib_mregion() and deinit_qib_mregion().  Additionally, explicit
initialization of fields to zero is now handled by kzalloc().

Also, duplicated code 'while.*num_sge' that decrements reference
counts have been consolidated in qib_put_ss().

Reviewed-by: Ramkrishna Vepa <ramkrishna.vepa@intel.com>
Signed-off-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>

1 files changed, 5 insertions, 5 deletions

diff --git a/drivers/infiniband/hw/qib/qib_verbs.c b/drivers/infiniband/hw/qib/qib_verbs.c
index 7b6c3bffa9d9..76d7ce8a8c6e 100644
--- a/drivers/infiniband/hw/qib/qib_verbs.c
+++ b/drivers/infiniband/hw/qib/qib_verbs.c
@@ -183,7 +183,7 @@ void qib_copy_sge(struct qib_sge_state *ss, void *data, u32 length, int release)
                 sge->sge_length -= len;
                 if (sge->sge_length == 0) {
                         if (release)
-                                atomic_dec(&sge->mr->refcount);
+                                qib_put_mr(sge->mr);
                         if (--ss->num_sge)
                                 *sge = *ss->sg_list++;
                 } else if (sge->length == 0 && sge->mr->lkey) {
@@ -224,7 +224,7 @@ void qib_skip_sge(struct qib_sge_state *ss, u32 length, int release)
                 sge->sge_length -= len;
                 if (sge->sge_length == 0) {
                         if (release)
-                                atomic_dec(&sge->mr->refcount);
+                                qib_put_mr(sge->mr);
                         if (--ss->num_sge)
                                 *sge = *ss->sg_list++;
                 } else if (sge->length == 0 && sge->mr->lkey) {
@@ -435,7 +435,7 @@ bail_inval_free:
         while (j) {
                 struct qib_sge *sge = &wqe->sg_list[--j];
 
-                atomic_dec(&sge->mr->refcount);
+                qib_put_mr(sge->mr);
         }
 bail_inval:
         ret = -EINVAL;
@@ -978,7 +978,7 @@ void qib_put_txreq(struct qib_verbs_txreq *tx)
         if (atomic_dec_and_test(&qp->refcount))
                 wake_up(&qp->wait);
         if (tx->mr) {
-                atomic_dec(&tx->mr->refcount);
+                qib_put_mr(tx->mr);
                 tx->mr = NULL;
         }
         if (tx->txreq.flags & QIB_SDMA_TXREQ_F_FREEBUF) {
@@ -1336,7 +1336,7 @@ done:
         }
         qib_sendbuf_done(dd, pbufn);
         if (qp->s_rdma_mr) {
-                atomic_dec(&qp->s_rdma_mr->refcount);
+                qib_put_mr(qp->s_rdma_mr);
                 qp->s_rdma_mr = NULL;
         }
         if (qp->s_wqe) {