Return-Path: Received: from fieldses.org ([173.255.197.46]:45992 "EHLO fieldses.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751992AbdKFPTp (ORCPT ); Mon, 6 Nov 2017 10:19:45 -0500 Date: Mon, 6 Nov 2017 10:19:44 -0500 From: Bruce Fields To: Chuck Lever Cc: linux-rdma , Linux NFS Mailing List , Devesh Sharma Subject: Re: [PATCH] svcrdma: Enqueue after setting XPT_CLOSE in completion handlers Message-ID: <20171106151944.GC599@fieldses.org> References: <20171027144743.15444.3407.stgit@klimt.1015granger.net> <28EB7B95-5EF6-41D0-8910-6BC7784CE28E@oracle.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii In-Reply-To: <28EB7B95-5EF6-41D0-8910-6BC7784CE28E@oracle.com> Sender: linux-nfs-owner@vger.kernel.org List-ID: On Fri, Nov 03, 2017 at 05:19:30PM -0400, Chuck Lever wrote: > Hi Bruce- > > > On Oct 28, 2017, at 3:43 AM, Devesh Sharma wrote: > > > > looks good > > > > Reveiwed-By: Devesh Sharma > > > > On Fri, Oct 27, 2017 at 8:19 PM, Chuck Lever wrote: > >> 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! > > I notice nfsd-next does not have this patch. Is there anything > else I need to do? Should I resend with Devesh's Reviewed-by > tag? Thanks for the reminder; applied for 4.15 with Devesh's Reviewed-by. --b.