From mboxrd@z Thu Jan 1 00:00:00 1970 From: Yang Yingliang Subject: Re: [PATCH] net: less interrupt masking in NAPI Date: Thu, 4 Dec 2014 11:09:43 +0800 Message-ID: <547FD077.4060807@huawei.com> References: <1414937973.31792.37.camel@edumazet-glaptop2.roam.corp.google.com> <20141103.122538.387451917276174830.davem@davemloft.net> <547EBC66.4040301@huawei.com> <1417592482.5303.132.camel@edumazet-glaptop2.roam.corp.google.com> <547ED728.2010703@huawei.com> <1417607546.5303.133.camel@edumazet-glaptop2.roam.corp.google.com> <547FC2A1.60105@huawei.com> <1417661505.22424.1.camel@edumazet-glaptop2.roam.corp.google.com> Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 7bit Cc: David Miller , , To: Eric Dumazet Return-path: Received: from szxga01-in.huawei.com ([119.145.14.64]:10376 "EHLO szxga01-in.huawei.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751106AbaLDDJ4 (ORCPT ); Wed, 3 Dec 2014 22:09:56 -0500 In-Reply-To: <1417661505.22424.1.camel@edumazet-glaptop2.roam.corp.google.com> Sender: netdev-owner@vger.kernel.org List-ID: On 2014/12/4 10:51, Eric Dumazet wrote: > On Thu, 2014-12-04 at 10:10 +0800, Yang Yingliang wrote: >> On 2014/12/3 19:52, Eric Dumazet wrote: >>> On Wed, 2014-12-03 at 17:26 +0800, Yang Yingliang wrote: >>> >>>> Before this patch, when a large network flow arrives, some other processes >>>> response slowly or even don't response because the cpu is dealing with softirq. >>>> >>>> After this patch, under pressure, much more softirq is doing in ksoftirqd. The other >>>> processes be scheduled. >>>> >>>> My system has dual core. >>> >>> Which NIC driver are you using ? >>> >> The driver we developed ourself, it's not upstreamed. > > Is it a NAPI driver ? Yes. > > Quite frankly this patch is not a stable candidate, as it was a > performance improvement, with some regressions in a couple of drivers. > OK, thanks for your reply.