bnxt_re: fix the regression due to changes in alloc_pbl

While adding the use of for_each_sg_dma_page iterator for Brodcom's rdma
driver, there was a regression added in the __alloc_pbl path. The change
left bnxt_re in DOA state in for-next branch.

Fixing the regression to avoid the host crash when a user space object is
created. Restricting the unconditional access to hwq.pg_arr when hwq is
initialized for user space objects.

Fixes: 161ebe24 ("RDMA/bnxt_re: Use for_each_sg_dma_page iterator on umem SGL")
Reported-by: NGal Pressman <galpress@amazon.com>
Signed-off-by: NSelvin Xavier <selvin.xavier@broadcom.com>
Signed-off-by: NDevesh Sharma <devesh.sharma@broadcom.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

drivers/infiniband/hw/bnxt_re/ib_verbs.c

@@ -793,8 +793,8 @@ int bnxt_re_destroy_qp(struct ib_qp *ib_qp)
 {
 	struct bnxt_re_qp *qp = container_of(ib_qp, struct bnxt_re_qp, ib_qp);
 	struct bnxt_re_dev *rdev = qp->rdev;
-	int rc;
 	unsigned int flags;
+	int rc;
 
 	bnxt_qplib_flush_cqn_wq(&qp->qplib_qp);
 	rc = bnxt_qplib_destroy_qp(&rdev->qplib_res, &qp->qplib_qp);
@@ -803,9 +803,12 @@ int bnxt_re_destroy_qp(struct ib_qp *ib_qp)
 		return rc;
 	}
 
-	flags = bnxt_re_lock_cqs(qp);
-	bnxt_qplib_clean_qp(&qp->qplib_qp);
-	bnxt_re_unlock_cqs(qp, flags);
+	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);
+	}
+
 	bnxt_qplib_free_qp_res(&rdev->qplib_res, &qp->qplib_qp);
 
 	if (ib_qp->qp_type == IB_QPT_GSI && rdev->qp1_sqp) {

drivers/infiniband/hw/bnxt_re/qplib_fp.c

@@ -862,18 +862,18 @@ int bnxt_qplib_create_qp1(struct bnxt_qplib_res *res, struct bnxt_qplib_qp *qp)
 int bnxt_qplib_create_qp(struct bnxt_qplib_res *res, struct bnxt_qplib_qp *qp)
 {
 	struct bnxt_qplib_rcfw *rcfw = res->rcfw;
-	struct sq_send *hw_sq_send_hdr, **hw_sq_send_ptr;
-	struct cmdq_create_qp req;
-	struct creq_create_qp_resp resp;
-	struct bnxt_qplib_pbl *pbl;
-	struct sq_psn_search **psn_search_ptr;
 	unsigned long int psn_search, poff = 0;
+	struct sq_psn_search **psn_search_ptr;
 	struct bnxt_qplib_q *sq = &qp->sq;
 	struct bnxt_qplib_q *rq = &qp->rq;
 	int i, rc, req_size, psn_sz = 0;
+	struct sq_send **hw_sq_send_ptr;
+	struct creq_create_qp_resp resp;
 	struct bnxt_qplib_hwq *xrrq;
 	u16 cmd_flags = 0, max_ssge;
-	u32 sw_prod, qp_flags = 0;
+	struct cmdq_create_qp req;
+	struct bnxt_qplib_pbl *pbl;
+	u32 qp_flags = 0;
 	u16 max_rsge;
 
 	RCFW_CMD_PREP(req, CREATE_QP, cmd_flags);
@@ -948,14 +948,6 @@ int bnxt_qplib_create_qp(struct bnxt_qplib_res *res, struct bnxt_qplib_qp *qp)
 				CMDQ_CREATE_QP_SQ_PG_SIZE_PG_1G :
 		 CMDQ_CREATE_QP_SQ_PG_SIZE_PG_4K);
 
-	/* initialize all SQ WQEs to LOCAL_INVALID (sq prep for hw fetch) */
-	hw_sq_send_ptr = (struct sq_send **)sq->hwq.pbl_ptr;
-	for (sw_prod = 0; sw_prod < sq->hwq.max_elements; sw_prod++) {
-		hw_sq_send_hdr = &hw_sq_send_ptr[get_sqe_pg(sw_prod)]
-						[get_sqe_idx(sw_prod)];
-		hw_sq_send_hdr->wqe_type = SQ_BASE_WQE_TYPE_LOCAL_INVALID;
-	}
-
 	if (qp->scq)
 		req.scq_cid = cpu_to_le32(qp->scq->id);
 

drivers/infiniband/hw/bnxt_re/qplib_res.c

@@ -119,11 +119,8 @@ static int __alloc_pbl(struct pci_dev *pdev, struct bnxt_qplib_pbl *pbl,
 		for_each_sg_dma_page (sghead, &sg_iter, pages, 0) {
 			pbl->pg_map_arr[i] = sg_page_iter_dma_address(&sg_iter);
 			pbl->pg_arr[i] = NULL;
-			if (!pbl->pg_arr[i])
-				goto fail;
-
-			i++;
 			pbl->pg_count++;
+			i++;
 		}
 	}