bnxt_re: Clean cq for kernel consumers only

Kernel space provider driver should clean the CQs belonging to kernel
space consumers only. The current implementation is doing reverse of it.

Fixing the same by avoiding the call to __clean_cq on a kernel qp during
destroy.

Fixes: c50866e2853a ("bnxt_re: fix the regression due to changes in alloc_pbl")
Signed-off-by: Devesh Sharma <devesh.sharma@broadcom.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>

diff --git a/drivers/infiniband/hw/bnxt_re/ib_verbs.c b/drivers/infiniband/hw/bnxt_re/ib_verbs.c
index 2409291..071b2fc 100644 (file)
--- a/drivers/infiniband/hw/bnxt_re/ib_verbs.c
+++ b/drivers/infiniband/hw/bnxt_re/ib_verbs.c
@@ -803,7 +803,7 @@ int bnxt_re_destroy_qp(struct ib_qp *ib_qp)
                return rc;
        }
 
-       if (!rdma_is_kernel_res(&qp->ib_qp.res)) {
+       if (rdma_is_kernel_res(&qp->ib_qp.res)) {
                flags = bnxt_re_lock_cqs(qp);
                bnxt_qplib_clean_qp(&qp->qplib_qp);
                bnxt_re_unlock_cqs(qp, flags);