1 files changed, 77 insertions, 35 deletions

diff --git a/drivers/infiniband/sw/rxe/rxe_resp.c b/drivers/infiniband/sw/rxe/rxe_resp.c
index 68f6cd188d8e..711f73e0bbb1 100644
--- a/drivers/infiniband/sw/rxe/rxe_resp.c
+++ b/drivers/infiniband/sw/rxe/rxe_resp.c
@@ -49,18 +49,8 @@ static char *resp_state_name[] = {
 /* rxe_recv calls here to add a request packet to the input queue */
 void rxe_resp_queue_pkt(struct rxe_qp *qp, struct sk_buff *skb)
 {
-	int must_sched;
-	struct rxe_pkt_info *pkt = SKB_TO_PKT(skb);
-
 	skb_queue_tail(&qp->req_pkts, skb);
-
-	must_sched = (pkt->opcode == IB_OPCODE_RC_RDMA_READ_REQUEST) ||
-			(skb_queue_len(&qp->req_pkts) > 1);
-
-	if (must_sched)
-		rxe_sched_task(&qp->resp.task);
-	else
-		rxe_run_task(&qp->resp.task);
+	rxe_sched_task(&qp->recv_task);
 }
 
 static inline enum resp_states get_req(struct rxe_qp *qp,
@@ -351,9 +341,22 @@ static enum resp_states rxe_resp_check_length(struct rxe_qp *qp,
 	/*
 	 * See IBA C9-92
 	 * For UD QPs we only check if the packet will fit in the
-	 * receive buffer later. For rmda operations additional
+	 * receive buffer later. For RDMA operations additional
 	 * length checks are performed in check_rkey.
 	 */
+	if ((qp_type(qp) == IB_QPT_GSI) || (qp_type(qp) == IB_QPT_UD)) {
+		unsigned int payload = payload_size(pkt);
+		unsigned int recv_buffer_len = 0;
+		int i;
+
+		for (i = 0; i < qp->resp.wqe->dma.num_sge; i++)
+			recv_buffer_len += qp->resp.wqe->dma.sge[i].length;
+		if (payload + sizeof(union rdma_network_hdr) > recv_buffer_len) {
+			rxe_dbg_qp(qp, "The receive buffer is too small for this UD packet.\n");
+			return RESPST_ERR_LENGTH;
+		}
+	}
+
 	if (pkt->mask & RXE_PAYLOAD_MASK && ((qp_type(qp) == IB_QPT_RC) ||
 					     (qp_type(qp) == IB_QPT_UC))) {
 		unsigned int mtu = qp->mtu;
@@ -362,18 +365,18 @@ static enum resp_states rxe_resp_check_length(struct rxe_qp *qp,
 		if ((pkt->mask & RXE_START_MASK) &&
 		    (pkt->mask & RXE_END_MASK)) {
 			if (unlikely(payload > mtu)) {
-				rxe_dbg_qp(qp, "only packet too long");
+				rxe_dbg_qp(qp, "only packet too long\n");
 				return RESPST_ERR_LENGTH;
 			}
 		} else if ((pkt->mask & RXE_START_MASK) ||
 			   (pkt->mask & RXE_MIDDLE_MASK)) {
 			if (unlikely(payload != mtu)) {
-				rxe_dbg_qp(qp, "first or middle packet not mtu");
+				rxe_dbg_qp(qp, "first or middle packet not mtu\n");
 				return RESPST_ERR_LENGTH;
 			}
 		} else if (pkt->mask & RXE_END_MASK) {
 			if (unlikely((payload == 0) || (payload > mtu))) {
-				rxe_dbg_qp(qp, "last packet zero or too long");
+				rxe_dbg_qp(qp, "last packet zero or too long\n");
 				return RESPST_ERR_LENGTH;
 			}
 		}
@@ -382,12 +385,15 @@ static enum resp_states rxe_resp_check_length(struct rxe_qp *qp,
 	/* See IBA C9-94 */
 	if (pkt->mask & RXE_RETH_MASK) {
 		if (reth_len(pkt) > (1U << 31)) {
-			rxe_dbg_qp(qp, "dma length too long");
+			rxe_dbg_qp(qp, "dma length too long\n");
 			return RESPST_ERR_LENGTH;
 		}
 	}
 
-	return RESPST_CHK_RKEY;
+	if (pkt->mask & RXE_RDMA_OP_MASK)
+		return RESPST_CHK_RKEY;
+	else
+		return RESPST_EXECUTE;
 }
 
 /* if the reth length field is zero we can assume nothing
@@ -434,6 +440,10 @@ static enum resp_states check_rkey(struct rxe_qp *qp,
 	enum resp_states state;
 	int access = 0;
 
+	/* parse RETH or ATMETH header for first/only packets
+	 * for va, length, rkey, etc. or use current value for
+	 * middle/last packets.
+	 */
 	if (pkt->mask & (RXE_READ_OR_WRITE_MASK | RXE_ATOMIC_WRITE_MASK)) {
 		if (pkt->mask & RXE_RETH_MASK)
 			qp_resp_from_reth(qp, pkt);
@@ -454,7 +464,8 @@ static enum resp_states check_rkey(struct rxe_qp *qp,
 		qp_resp_from_atmeth(qp, pkt);
 		access = IB_ACCESS_REMOTE_ATOMIC;
 	} else {
-		return RESPST_EXECUTE;
+		/* shouldn't happen */
+		WARN_ON(1);
 	}
 
 	/* A zero-byte read or write op is not required to
@@ -489,8 +500,9 @@ static enum resp_states check_rkey(struct rxe_qp *qp,
 		if (mw->access & IB_ZERO_BASED)
 			qp->resp.offset = mw->addr;
 
-		rxe_put(mw);
 		rxe_get(mr);
+		rxe_put(mw);
+		mw = NULL;
 	} else {
 		mr = lookup_mr(qp->pd, access, rkey, RXE_LOOKUP_REMOTE);
 		if (!mr) {
@@ -690,10 +702,16 @@ static enum resp_states atomic_reply(struct rxe_qp *qp,
 	if (!res->replay) {
 		u64 iova = qp->resp.va + qp->resp.offset;
 
-		err = rxe_mr_do_atomic_op(mr, iova, pkt->opcode,
-					  atmeth_comp(pkt),
-					  atmeth_swap_add(pkt),
-					  &res->atomic.orig_val);
+		if (is_odp_mr(mr))
+			err = rxe_odp_atomic_op(mr, iova, pkt->opcode,
+						atmeth_comp(pkt),
+						atmeth_swap_add(pkt),
+						&res->atomic.orig_val);
+		else
+			err = rxe_mr_do_atomic_op(mr, iova, pkt->opcode,
+						  atmeth_comp(pkt),
+						  atmeth_swap_add(pkt),
+						  &res->atomic.orig_val);
 		if (err)
 			return err;
 
@@ -731,7 +749,16 @@ static enum resp_states atomic_write_reply(struct rxe_qp *qp,
 	value = *(u64 *)payload_addr(pkt);
 	iova = qp->resp.va + qp->resp.offset;
 
-	err = rxe_mr_do_atomic_write(mr, iova, value);
+	/* See IBA oA19-28 */
+	if (unlikely(mr->state != RXE_MR_STATE_VALID)) {
+		rxe_dbg_mr(mr, "mr not in valid state\n");
+		return RESPST_ERR_RKEY_VIOLATION;
+	}
+
+	if (is_odp_mr(mr))
+		err = rxe_odp_do_atomic_write(mr, iova, value);
+	else
+		err = rxe_mr_do_atomic_write(mr, iova, value);
 	if (err)
 		return err;
 
@@ -1047,6 +1074,7 @@ static enum resp_states do_complete(struct rxe_qp *qp,
 	struct ib_uverbs_wc *uwc = &cqe.uibwc;
 	struct rxe_recv_wqe *wqe = qp->resp.wqe;
 	struct rxe_dev *rxe = to_rdev(qp->ibqp.device);
+	unsigned long flags;
 
 	if (!wqe)
 		goto finish;
@@ -1123,7 +1151,7 @@ static enum resp_states do_complete(struct rxe_qp *qp,
 		}
 	} else {
 		if (wc->status != IB_WC_WR_FLUSH_ERR)
-			rxe_err_qp(qp, "non-flush error status = %d",
+			rxe_err_qp(qp, "non-flush error status = %d\n",
 				wc->status);
 	}
 
@@ -1137,12 +1165,12 @@ static enum resp_states do_complete(struct rxe_qp *qp,
 		return RESPST_ERR_CQ_OVERFLOW;
 
 finish:
-	spin_lock_bh(&qp->state_lock);
+	spin_lock_irqsave(&qp->state_lock, flags);
 	if (unlikely(qp_state(qp) == IB_QPS_ERR)) {
-		spin_unlock_bh(&qp->state_lock);
+		spin_unlock_irqrestore(&qp->state_lock, flags);
 		return RESPST_CHK_RESOURCE;
 	}
-	spin_unlock_bh(&qp->state_lock);
+	spin_unlock_irqrestore(&qp->state_lock, flags);
 
 	if (unlikely(!pkt))
 		return RESPST_DONE;
@@ -1432,7 +1460,7 @@ static int flush_recv_wqe(struct rxe_qp *qp, struct rxe_recv_wqe *wqe)
 
 	err = rxe_cq_post(qp->rcq, &cqe, 0);
 	if (err)
-		rxe_dbg_cq(qp->rcq, "post cq failed err = %d", err);
+		rxe_dbg_cq(qp->rcq, "post cq failed err = %d\n", err);
 
 	return err;
 }
@@ -1447,7 +1475,20 @@ static void flush_recv_queue(struct rxe_qp *qp, bool notify)
 	struct rxe_recv_wqe *wqe;
 	int err;
 
-	if (qp->srq)
+	if (qp->srq) {
+		if (notify && qp->ibqp.event_handler) {
+			struct ib_event ev;
+
+			ev.device = qp->ibqp.device;
+			ev.element.qp = &qp->ibqp;
+			ev.event = IB_EVENT_QP_LAST_WQE_REACHED;
+			qp->ibqp.event_handler(&ev, qp->ibqp.qp_context);
+		}
+		return;
+	}
+
+	/* recv queue not created. nothing to do. */
+	if (!qp->rq.queue)
 		return;
 
 	while ((wqe = queue_head(q, q->type))) {
@@ -1462,24 +1503,25 @@ static void flush_recv_queue(struct rxe_qp *qp, bool notify)
 	qp->resp.wqe = NULL;
 }
 
-int rxe_responder(struct rxe_qp *qp)
+int rxe_receiver(struct rxe_qp *qp)
 {
 	struct rxe_dev *rxe = to_rdev(qp->ibqp.device);
 	enum resp_states state;
 	struct rxe_pkt_info *pkt = NULL;
 	int ret;
+	unsigned long flags;
 
-	spin_lock_bh(&qp->state_lock);
+	spin_lock_irqsave(&qp->state_lock, flags);
 	if (!qp->valid || qp_state(qp) == IB_QPS_ERR ||
 			  qp_state(qp) == IB_QPS_RESET) {
 		bool notify = qp->valid && (qp_state(qp) == IB_QPS_ERR);
 
 		drain_req_pkts(qp);
 		flush_recv_queue(qp, notify);
-		spin_unlock_bh(&qp->state_lock);
+		spin_unlock_irqrestore(&qp->state_lock, flags);
 		goto exit;
 	}
-	spin_unlock_bh(&qp->state_lock);
+	spin_unlock_irqrestore(&qp->state_lock, flags);
 
 	qp->resp.aeth_syndrome = AETH_ACK_UNLIMITED;
 
@@ -1654,7 +1696,7 @@ int rxe_responder(struct rxe_qp *qp)
 	}
 
 	/* A non-zero return value will cause rxe_do_task to
-	 * exit its loop and end the tasklet. A zero return
+	 * exit its loop and end the work item. A zero return
 	 * will continue looping and return to rxe_responder
 	 */
 done: