Return-Path: Received: from p3plsmtpa07-02.prod.phx3.secureserver.net ([173.201.192.231]:56582 "EHLO p3plsmtpa07-02.prod.phx3.secureserver.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752943AbdIFBgn (ORCPT ); Tue, 5 Sep 2017 21:36:43 -0400 Subject: Re: [PATCH RFC 0/5] xprtrdma Send completion batching To: Chuck Lever , Jason Gunthorpe Cc: Sagi Grimberg , linux-rdma@vger.kernel.org, Linux NFS Mailing List References: <20170905164347.11106.27140.stgit@manet.1015granger.net> <20170905200608.GA4055@obsidianresearch.com> <0D960F54-0BC7-4A5A-8001-0391673892DF@oracle.com> From: Tom Talpey Message-ID: Date: Tue, 5 Sep 2017 21:28:52 -0400 MIME-Version: 1.0 In-Reply-To: <0D960F54-0BC7-4A5A-8001-0391673892DF@oracle.com> Content-Type: text/plain; charset=utf-8; format=flowed Sender: linux-nfs-owner@vger.kernel.org List-ID: On 9/5/2017 5:22 PM, Chuck Lever wrote: > I think it does bear mentioning that, given this diagnosis, it is > still safe to remove the ib_post_send counting mechanism in 5/5, > which has been in xprtrdma for as long as I can recall, and has 2007. :-) > been effective (with a few minor adjustments) at preventing SQ > overflow. > > I'm not able to test this change with every HCA the Linux kernel > currently supports, unfortunately. The best I can do is offer a > "proof of correctness" and hope that vendors will jump on this > and try it out. Someday, a proper set of verbs test cases? Alas. Tom.