From mboxrd@z Thu Jan 1 00:00:00 1970 From: Eric Dumazet Subject: Re: [PATCH net] e1000e: Change wthresh to 1 to avoid possible Tx stalls. Date: Tue, 09 Oct 2012 08:07:43 +0200 Message-ID: <1349762863.21172.3848.camel@edumazet-glaptop> References: <20120606174355.823e9aa7.shimoda.hiroaki@gmail.com> <1339030752.2075.1.camel@jtkirshe-mobl> <1339043085.26966.77.camel@edumazet-glaptop> <1339044752.2075.14.camel@jtkirshe-mobl> Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 7bit Cc: netdev@vger.kernel.org, Jeff Kirsher To: Frank Reppin Return-path: Received: from mail-bk0-f46.google.com ([209.85.214.46]:35986 "EHLO mail-bk0-f46.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753030Ab2JIGHq (ORCPT ); Tue, 9 Oct 2012 02:07:46 -0400 Received: by mail-bk0-f46.google.com with SMTP id jk13so2393149bkc.19 for ; Mon, 08 Oct 2012 23:07:45 -0700 (PDT) In-Reply-To: Sender: netdev-owner@vger.kernel.org List-ID: On Tue, 2012-10-09 at 00:25 +0000, Frank Reppin wrote: > Jeff Kirsher intel.com> writes: > > > > On Thu, 2012-06-07 at 06:24 +0200, Eric Dumazet wrote: > > > On Wed, 2012-06-06 at 17:59 -0700, Jeff Kirsher wrote: > > > > > > > After further internal review, NACK. > > > > > > > > This patch will cause unacceptable performance issues with non-ESB2 > > > > parts. > > > > > > > > I am dropping this patch from my queue. > > > > > > > > > > I'd like you share your performance numbers before NACKing this patch. > > > > > > What is the alternative patch you guys have ? > > > > > > > Jesse did not share any performance numbers with me, I am sure he can > > give some background tomorrow when he is back online. > > > > I am working on an alternative patch now and should have something to > > share tomorrow. > Please allow me to ask if there's any progess here? > > I've tried 3.5.4 a couple of days ago on a SuperMicro X8SIE-LN4 (82574L) > and could still observe severe latency (up to 3000ms) spikes. > > Applying Hiroakis suggested patch did fix this for me as well. > [please note as well that I didn't had this issue in any 3.4.x kernel > before - so +1 for fixing the regression] Its a shame this bug is not yet fixed. I dont know what to say.