Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751832AbaLCJJY (ORCPT ); Wed, 3 Dec 2014 04:09:24 -0500 Received: from rtits2.realtek.com ([60.250.210.242]:37968 "EHLO rtits2.realtek.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750893AbaLCJJU convert rfc822-to-8bit (ORCPT ); Wed, 3 Dec 2014 04:09:20 -0500 Authenticated-By: X-SpamFilter-By: BOX Solutions SpamTrap 5.49 with qID sB399EWa016192, This message is accepted by code: ctloc85258 From: Hayes Wang To: Eric Dumazet CC: "netdev@vger.kernel.org" , nic_swsd , "linux-kernel@vger.kernel.org" , "linux-usb@vger.kernel.org" Subject: RE: [PATCH net-next] r8152: reduce memory copy for rx Thread-Topic: [PATCH net-next] r8152: reduce memory copy for rx Thread-Index: AQHQDrgS8Vmh5UfiJ0iekivC5Eu1N5x821QAgACRkiD//4FegIAAoPyA Date: Wed, 3 Dec 2014 09:09:12 +0000 Message-ID: <0835B3720019904CB8F7AA43166CEEB2ED364C@RTITMBSV03.realtek.com.tw> References: <1394712342-15778-105-Taiwan-albertk@realtek.com> <1417586858.5303.118.camel@edumazet-glaptop2.roam.corp.google.com> <0835B3720019904CB8F7AA43166CEEB2ED360A@RTITMBSV03.realtek.com.tw> <1417590925.5303.127.camel@edumazet-glaptop2.roam.corp.google.com> In-Reply-To: <1417590925.5303.127.camel@edumazet-glaptop2.roam.corp.google.com> Accept-Language: zh-TW, en-US Content-Language: zh-TW X-MS-Has-Attach: X-MS-TNEF-Correlator: x-originating-ip: [172.21.71.143] Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 8BIT MIME-Version: 1.0 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Eric Dumazet [mailto:eric.dumazet@gmail.com] > Sent: Wednesday, December 03, 2014 3:15 PM [...] > Have you tried using more concurrent RX flows, in a possibly lossy > environment (so that TCP is forced to queue packets in out of order > queue) ? I don't do the test. I would check it next time. > skb cloning prevents GRO and TCP coalescing from working. > > netfilter might also be forced to copy whole frame in case a mangle is > needed (eg with NAT ...) > > I would rather try to implement GRO, and/or using fragments instead of > pure linear skbs. > > (skb->head would be around 128 or 256 bytes, and you attach to skb the > frame as a page fragment) Thanks for your response. I would study the GRO first. Best Regards, Hayes -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/