Return-Path: Received: from mail-io0-f194.google.com ([209.85.223.194]:56553 "EHLO mail-io0-f194.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750781AbdJ0Otx (ORCPT ); Fri, 27 Oct 2017 10:49:53 -0400 Subject: [PATCH] svcrdma: Enqueue after setting XPT_CLOSE in completion handlers From: Chuck Lever To: bfields@fieldses.org Cc: linux-rdma@vger.kernel.org, linux-nfs@vger.kernel.org Date: Fri, 27 Oct 2017 10:49:51 -0400 Message-ID: <20171027144743.15444.3407.stgit@klimt.1015granger.net> MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Sender: linux-nfs-owner@vger.kernel.org List-ID: I noticed the server was sometimes not closing the connection after a flushed Send. For example, if the client responds with an RNR NAK to a Reply from the server, that client might be deadlocked, and thus wouldn't send any more traffic. Thus the server wouldn't have any opportunity to notice the XPT_CLOSE bit has been set. Enqueue the transport so that svcxprt notices the bit even if there is no more transport activity after a flushed completion, QP access error, or device removal event. Signed-off-by: Chuck Lever --- net/sunrpc/xprtrdma/svc_rdma_transport.c | 11 ++++++++--- 1 file changed, 8 insertions(+), 3 deletions(-) Hi Bruce- Please consider this patch for v4.15. Thanks! diff --git a/net/sunrpc/xprtrdma/svc_rdma_transport.c b/net/sunrpc/xprtrdma/svc_rdma_transport.c index 5caf8e7..46ec069 100644 --- a/net/sunrpc/xprtrdma/svc_rdma_transport.c +++ b/net/sunrpc/xprtrdma/svc_rdma_transport.c @@ -290,6 +290,7 @@ static void qp_event_handler(struct ib_event *event, void *context) ib_event_msg(event->event), event->event, event->element.qp); set_bit(XPT_CLOSE, &xprt->xpt_flags); + svc_xprt_enqueue(xprt); break; } } @@ -322,8 +323,7 @@ static void svc_rdma_wc_receive(struct ib_cq *cq, struct ib_wc *wc) set_bit(XPT_DATA, &xprt->sc_xprt.xpt_flags); if (test_bit(RDMAXPRT_CONN_PENDING, &xprt->sc_flags)) goto out; - svc_xprt_enqueue(&xprt->sc_xprt); - goto out; + goto out_enqueue; flushed: if (wc->status != IB_WC_WR_FLUSH_ERR) @@ -333,6 +333,8 @@ static void svc_rdma_wc_receive(struct ib_cq *cq, struct ib_wc *wc) set_bit(XPT_CLOSE, &xprt->sc_xprt.xpt_flags); svc_rdma_put_context(ctxt, 1); +out_enqueue: + svc_xprt_enqueue(&xprt->sc_xprt); out: svc_xprt_put(&xprt->sc_xprt); } @@ -358,6 +360,7 @@ void svc_rdma_wc_send(struct ib_cq *cq, struct ib_wc *wc) if (unlikely(wc->status != IB_WC_SUCCESS)) { set_bit(XPT_CLOSE, &xprt->sc_xprt.xpt_flags); + svc_xprt_enqueue(&xprt->sc_xprt); if (wc->status != IB_WC_WR_FLUSH_ERR) pr_err("svcrdma: Send: %s (%u/0x%x)\n", ib_wc_status_msg(wc->status), @@ -569,8 +572,10 @@ static int rdma_listen_handler(struct rdma_cm_id *cma_id, case RDMA_CM_EVENT_DEVICE_REMOVAL: dprintk("svcrdma: Device removal xprt=%p, cm_id=%p\n", xprt, cma_id); - if (xprt) + if (xprt) { set_bit(XPT_CLOSE, &xprt->sc_xprt.xpt_flags); + svc_xprt_enqueue(&xprt->sc_xprt); + } break; default: