From: "NeilBrown" Subject: Re: Link performance over NFS degraded in RHEL5. -- was : Read/Write NFS I/O performance degraded by FLUSH_STABLE page flushing Date: Tue, 16 Jun 2009 10:21:50 +1000 (EST) Message-ID: <99d4545537613ce76040d3655b78bdb7.squirrel@neil.brown.name> References: <1244049027.5603.5.camel@heimdal.trondhjem.org> <1244138698.5203.59.camel@heimdal.trondhjem.org> <4A2902E6.2080006@RedHat.com> <4A29144A.6030405@gmail.com> <4A291DE3.2070105@RedHat.com> <1244209956.5410.33.camel@heimdal.trondhjem.org> <4A29243F.8080008@RedHat.com> <20090605160544.GE10975@fieldses.org> <1244219715.5410.40.camel@heimdal.trondhjem.org> <20090615230801.GA18504@fieldses.org> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Cc: "Trond Myklebust" , "Steve Dickson" , "Tom Talpey" , "Linux NFS Mailing list" To: "J. Bruce Fields" Return-path: Received: from cantor2.suse.de ([195.135.220.15]:58592 "EHLO mx2.suse.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750993AbZFPAV6 (ORCPT ); Mon, 15 Jun 2009 20:21:58 -0400 In-Reply-To: <20090615230801.GA18504@fieldses.org> Sender: linux-nfs-owner@vger.kernel.org List-ID: On Tue, June 16, 2009 9:08 am, J. Bruce Fields wrote: > + if (host_err >= 0 && stable) > + wait_for_concurrent_writes(file, use_wgather, &host_err); > Surely you want this to be: if (host_err >= 0 && stable && use_wgather) host_err = wait_for_concurrent_writes(file); as - this is more readable - setting last_ino and last_dev is pointless when !use_wgather - we aren't interested in differentiation between non-negative values of host_err. NeilBrown