[{"id":1762333,"web_url":"http://patchwork.ozlabs.org/comment/1762333/","msgid":"","list_archive_url":null,"date":"2017-09-03T21:17:24","subject":"Re: [LEDE-DEV] [PATCH] ar71xx: Add GRO support to ag71xx","submitter":{"id":70632,"url":"http://patchwork.ozlabs.org/api/people/70632/","name":"Michael Yartys via Lede-dev","email":"lede-dev@lists.infradead.org"},"content":"The sender domain has a DMARC Reject/Quarantine policy which disallows\nsending mailing list messages using the original \"From\" header.\n\nTo mitigate this problem, the original message has been wrapped\nautomatically by the mailing list software.\nOn Sunday, September 3, 2017 11:35:46 AM CEST Rosen Penev wrote:\n> On a TL-WN710N, this patch increases iperf performance from ~92.5 to ~93.5 mbps. Keep in mind the WN710N is a 100mbps device. I expect greater numbers from gigabit devices.\n> \n> Signed-off-by: Rosen Penev \n> ---\nI've done a quick test of the patch on my WD Range Extender.\n(It has a Atheros AR9344 rev 2 SoC @ CPU:560.000MHz, DDR:400.000MHz\nThe PHY is a AR8035, which supports 1 GBit/s Links)\n\nThe range extender (DUT) was running iperf3 server in both tests.\nAnother desktop PC was acting as the iperf3 client.\n\nwithout the patch:\n\nConnecting to host range-extender, port 5201\n[ 4] local 192.168.8.7 port 51518 connected to 192.168.8.204 port 5201\n[ ID] Interval Transfer Bandwidth Retr Cwnd\n[ 4] 0.00-1.00 sec 23.5 MBytes 197 Mbits/sec 0 105 KBytes\n[ 4] 1.00-2.00 sec 23.7 MBytes 199 Mbits/sec 0 105 KBytes\n[ 4] 2.00-3.00 sec 23.6 MBytes 198 Mbits/sec 0 105 KBytes\n[ 4] 3.00-4.00 sec 23.0 MBytes 193 Mbits/sec 0 105 KBytes\n[ 4] 4.00-5.00 sec 23.4 MBytes 197 Mbits/sec 0 105 KBytes\n[ 4] 5.00-6.00 sec 23.3 MBytes 195 Mbits/sec 0 105 KBytes\n[ 4] 6.00-7.00 sec 23.4 MBytes 196 Mbits/sec 0 105 KBytes\n[ 4] 7.00-8.00 sec 23.6 MBytes 198 Mbits/sec 0 105 KBytes\n[ 4] 8.00-9.00 sec 23.1 MBytes 194 Mbits/sec 0 105 KBytes\n[ 4] 9.00-10.00 sec 22.1 MBytes 185 Mbits/sec 0 105 KBytes\n- - - - - - - - - - - - - - - - - - - - - - - - -\n[ ID] Interval Transfer Bandwidth Retr\n[ 4] 0.00-10.00 sec 233 MBytes 195 Mbits/sec 0 sender\n[ 4] 0.00-10.00 sec 232 MBytes 195 Mbits/sec receiver\n\niperf Done.\n\nwith the patch (gro enabled - this is done by default):\n\nConnecting to host range-extender, port 5201\n[ 4] local 192.168.8.7 port 52004 connected to 192.168.8.204 port 5201\n[ ID] Interval Transfer Bandwidth Retr Cwnd\n[ 4] 0.00-1.00 sec 32.7 MBytes 274 Mbits/sec 0 106 KBytes \n[ 4] 1.00-2.00 sec 32.6 MBytes 274 Mbits/sec 0 106 KBytes \n[ 4] 2.00-3.00 sec 32.4 MBytes 272 Mbits/sec 0 106 KBytes \n[ 4] 3.00-4.00 sec 32.3 MBytes 271 Mbits/sec 0 106 KBytes \n[ 4] 4.00-5.00 sec 32.5 MBytes 273 Mbits/sec 0 106 KBytes \n[ 4] 5.00-6.00 sec 32.5 MBytes 273 Mbits/sec 0 106 KBytes \n[ 4] 6.00-7.00 sec 32.6 MBytes 273 Mbits/sec 0 106 KBytes \n[ 4] 7.00-8.00 sec 32.4 MBytes 272 Mbits/sec 0 106 KBytes \n[ 4] 8.00-9.00 sec 32.6 MBytes 273 Mbits/sec 0 106 KBytes \n[ 4] 9.00-10.00 sec 31.4 MBytes 264 Mbits/sec 0 106 KBytes \n- - - - - - - - - - - - - - - - - - - - - - - - -\n[ ID] Interval Transfer Bandwidth Retr\n[ 4] 0.00-10.00 sec 324 MBytes 272 Mbits/sec 0 sender\n[ 4] 0.00-10.00 sec 324 MBytes 272 Mbits/sec receiver\n\niperf Done.\n\n(range-extender) # ethtool -K eth0 gro off\n\nConnecting to host range-extender, port 5201\n[ 4] local 192.168.8.7 port 52120 connected to 192.168.8.204 port 5201\n[ ID] Interval Transfer Bandwidth Retr Cwnd\n[ 4] 0.00-1.00 sec 24.8 MBytes 208 Mbits/sec 0 105 KBytes \n[ 4] 1.00-2.00 sec 23.6 MBytes 198 Mbits/sec 0 105 KBytes \n[ 4] 2.00-3.00 sec 24.5 MBytes 206 Mbits/sec 0 105 KBytes \n[ 4] 3.00-4.00 sec 23.9 MBytes 201 Mbits/sec 0 105 KBytes \n[ 4] 4.00-5.00 sec 24.6 MBytes 207 Mbits/sec 0 105 KBytes \n[ 4] 5.00-6.00 sec 24.7 MBytes 207 Mbits/sec 0 105 KBytes \n[ 4] 6.00-7.00 sec 24.5 MBytes 206 Mbits/sec 0 105 KBytes \n[ 4] 7.00-8.00 sec 24.0 MBytes 201 Mbits/sec 0 105 KBytes \n[ 4] 8.00-9.00 sec 24.3 MBytes 204 Mbits/sec 0 105 KBytes \n[ 4] 9.00-10.00 sec 24.5 MBytes 206 Mbits/sec 0 105 KBytes \n- - - - - - - - - - - - - - - - - - - - - - - - -\n[ ID] Interval Transfer Bandwidth Retr\n[ 4] 0.00-10.00 sec 244 MBytes 204 Mbits/sec 0 sender\n[ 4] 0.00-10.00 sec 243 MBytes 204 Mbits/sec receiver\n\niperf Done.\n\nSo, the throughput went from 195 Mbits/sec to 272 Mbits/sec.\nThe gain would be (272 Mbps - 195 Mbps) / 195 Mbps = 0.3949 ~ 40%\n\nRegards,\nChristian","headers":{"Return-Path":"","X-Original-To":"incoming@patchwork.ozlabs.org","Delivered-To":"patchwork-incoming@bilbo.ozlabs.org","Authentication-Results":["ozlabs.org; spf=none (mailfrom)\n\tsmtp.mailfrom=lists.infradead.org (client-ip=65.50.211.133;\n\thelo=bombadil.infradead.org;\n\tenvelope-from=lede-dev-bounces+incoming=patchwork.ozlabs.org@lists.infradead.org;\n\treceiver=)","ozlabs.org; dkim=pass (2048-bit key;\n\tunprotected) header.d=lists.infradead.org\n\theader.i=@lists.infradead.org header.b=\"Hvp8lVm2\"; \n\tdkim-atps=neutral"],"Received":["from bombadil.infradead.org (bombadil.infradead.org\n\t[65.50.211.133])\n\t(using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256\n\tbits)) (No client certificate requested)\n\tby ozlabs.org (Postfix) with ESMTPS id 3xlm5n3Vz2z9sP3\n\tfor ;\n\tMon, 4 Sep 2017 07:17:33 +1000 (AEST)","from localhost ([127.0.0.1] helo=bombadil.infradead.org)\n\tby bombadil.infradead.org with esmtp (Exim 4.87 #1 (Red Hat Linux))\n\tid 1docGi-000683-Eu; Sun, 03 Sep 2017 21:17:24 +0000"],"DKIM-Signature":"v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed;\n\td=lists.infradead.org; s=bombadil.20170209;\n\th=Date:Sender:Content-Type:\n\tSubject:List-Help:Reply-To:List-Archive:List-Unsubscribe:List-Subscribe:Cc:\n\tFrom:List-Post:List-Id:Message-ID:MIME-Version:References:In-Reply-To:To:\n\tContent-Transfer-Encoding:Content-ID:Content-Description:Resent-Date:\n\tResent-From:Resent-Sender:Resent-To:Resent-Cc:Resent-Message-ID:List-Owner;\n\tbh=Xa+ocm94Rp4tiC0hqMmgUzIX11Sp18YZgKOFVzU3yJk=;\n\tb=Hvp8lVm2Viy7LUJaXovfIsM/x\n\t1IMVW4XUqWBeErf2h5xPQa5uDKgBdY2nDOqXYL2zjHqbQS7Cs4MvsVvnn6+Jd/TnCU0lnXYTAlN+O\n\tDvTTLlbg8Zb2J8BmTpVWl+mcZn1ycVrBpzUlliIc9m73E0en3jOJIA2TS43TqvusmTZAa/+Dd8eZw\n\tBJ6mtJniCKYoivtwWBzTHXunPXSr9PZkzA3a/9YYKLEBq0+hxAU2qiDDpZhXBHNecwq4swqMzXbfV\n\txjWGjV1RrEpbTKiQjedOWemEwEi3CwpbZL0KvD3LOhHZMGDHyw8u4cMot9jB/keKf0h0tCfPXUVNa\n\tojuAb96AQ==;","To":"lede-dev@lists.infradead.org","In-Reply-To":"<20170903183546.16749-1-rosenp@gmail.com>","References":"<20170903183546.16749-1-rosenp@gmail.com>","MIME-Version":"1.0","Message-ID":"","List-Id":"","List-Post":"","From":"Christian Lamparter via Lede-dev ","Precedence":"list","Cc":"Rosen Penev ","X-Mailman-Version":"2.1.21","X-BeenThere":"lede-dev@lists.infradead.org","List-Subscribe":",\n\t","List-Unsubscribe":",\n\t","List-Archive":"","Reply-To":"Christian Lamparter ","List-Help":"","Subject":"Re: [LEDE-DEV] [PATCH] ar71xx: Add GRO support to ag71xx","Content-Type":"multipart/mixed;\n\tboundary=\"===============2548893514386972219==\"","Sender":"\"Lede-dev\" ","Date":"Sun, 03 Sep 2017 21:17:24 +0000","Errors-To":"lede-dev-bounces+incoming=patchwork.ozlabs.org@lists.infradead.org"}},{"id":1762340,"web_url":"http://patchwork.ozlabs.org/comment/1762340/","msgid":"<1504478774.2387.8.camel@gmail.com>","list_archive_url":null,"date":"2017-09-03T22:46:14","subject":"Re: [LEDE-DEV] [PATCH] ar71xx: Add GRO support to ag71xx","submitter":{"id":70304,"url":"http://patchwork.ozlabs.org/api/people/70304/","name":"Rosen Penev","email":"rosenp@gmail.com"},"content":"That's...way better than I expected given how minimal my changes are.\n\nFor some context, the ag71xx driver is special in that it does not seem\nto do any hardware offloading to the NIC.\n\nAs far as I understand this change, GRO takes 1500 MTU packets and\npacks then into 64Kb blocks which the kernel then processes.\n\nI would be curious if anyone can do latency comparisons before this\nchange and after. I do know this driver to have lower latency than\nothers due to lack of offloads.\n\nI guess all that's left is to add GSO support to the driver. That seems\nlike a lot more work than a three line change though. ¯\\_(ツ)_/¯\n\nOn Sun, 2017-09-03 at 23:16 +0200, Christian Lamparter wrote:\n> On Sunday, September 3, 2017 11:35:46 AM CEST Rosen Penev wrote:\n> > On a TL-WN710N, this patch increases iperf performance from ~92.5\n> > to ~93.5 mbps. Keep in mind the WN710N is a 100mbps device. I\n> > expect greater numbers from gigabit devices.\n> > \n> > Signed-off-by: Rosen Penev \n> > ---\n> \n> I've done a quick test of the patch on my WD Range Extender.\n> (It has a Atheros AR9344 rev 2 SoC @ CPU:560.000MHz, DDR:400.000MHz\n> The PHY is a AR8035, which supports 1 GBit/s Links)\n> \n> The range extender (DUT) was running iperf3 server in both tests.\n> Another desktop PC was acting as the iperf3 client.\n> \n> without the patch:\n> \n> Connecting to host range-extender, port 5201\n> [ 4] local 192.168.8.7 port 51518 connected to 192.168.8.204 port\n> 5201\n> [ ID] Interval Transfer Bandwidth Retr Cwnd\n> [ 4] 0.00-1.00 sec 23.5 MBytes 197 Mbits/sec 0 105\n> KBytes\n> [ 4] 1.00-2.00 sec 23.7 MBytes 199 Mbits/sec 0 105\n> KBytes\n> [ 4] 2.00-3.00 sec 23.6 MBytes 198 Mbits/sec 0 105\n> KBytes\n> [ 4] 3.00-4.00 sec 23.0 MBytes 193 Mbits/sec 0 105\n> KBytes\n> [ 4] 4.00-5.00 sec 23.4 MBytes 197 Mbits/sec 0 105\n> KBytes\n> [ 4] 5.00-6.00 sec 23.3 MBytes 195 Mbits/sec 0 105\n> KBytes\n> [ 4] 6.00-7.00 sec 23.4 MBytes 196 Mbits/sec 0 105\n> KBytes\n> [ 4] 7.00-8.00 sec 23.6 MBytes 198 Mbits/sec 0 105\n> KBytes\n> [ 4] 8.00-9.00 sec 23.1 MBytes 194 Mbits/sec 0 105\n> KBytes\n> [ 4] 9.00-10.00 sec 22.1 MBytes 185 Mbits/sec 0 105\n> KBytes\n> - - - - - - - - - - - - - - - - - - - - - - - - -\n> [ ID] Interval Transfer Bandwidth Retr\n> [ 4] 0.00-10.00 sec 233 MBytes 195\n> Mbits/sec 0 sender\n> [ 4] 0.00-10.00 sec 232 MBytes 195\n> Mbits/sec receiver\n> \n> iperf Done.\n> \n> with the patch (gro enabled - this is done by default):\n> \n> Connecting to host range-extender, port 5201\n> [ 4] local 192.168.8.7 port 52004 connected to 192.168.8.204 port\n> 5201\n> [ ID] Interval Transfer Bandwidth Retr Cwnd\n> [ 4] 0.00-1.00 sec 32.7 MBytes 274 Mbits/sec 0 106\n> KBytes \n> [ 4] 1.00-2.00 sec 32.6 MBytes 274 Mbits/sec 0 106\n> KBytes \n> [ 4] 2.00-3.00 sec 32.4 MBytes 272 Mbits/sec 0 106\n> KBytes \n> [ 4] 3.00-4.00 sec 32.3 MBytes 271 Mbits/sec 0 106\n> KBytes \n> [ 4] 4.00-5.00 sec 32.5 MBytes 273 Mbits/sec 0 106\n> KBytes \n> [ 4] 5.00-6.00 sec 32.5 MBytes 273 Mbits/sec 0 106\n> KBytes \n> [ 4] 6.00-7.00 sec 32.6 MBytes 273 Mbits/sec 0 106\n> KBytes \n> [ 4] 7.00-8.00 sec 32.4 MBytes 272 Mbits/sec 0 106\n> KBytes \n> [ 4] 8.00-9.00 sec 32.6 MBytes 273 Mbits/sec 0 106\n> KBytes \n> [ 4] 9.00-10.00 sec 31.4 MBytes 264 Mbits/sec 0 106\n> KBytes \n> - - - - - - - - - - - - - - - - - - - - - - - - -\n> [ ID] Interval Transfer Bandwidth Retr\n> [ 4] 0.00-10.00 sec 324 MBytes 272\n> Mbits/sec 0 sender\n> [ 4] 0.00-10.00 sec 324 MBytes 272\n> Mbits/sec receiver\n> \n> iperf Done.\n> \n> (range-extender) # ethtool -K eth0 gro off\n> \n> Connecting to host range-extender, port 5201\n> [ 4] local 192.168.8.7 port 52120 connected to 192.168.8.204 port\n> 5201\n> [ ID] Interval Transfer Bandwidth Retr Cwnd\n> [ 4] 0.00-1.00 sec 24.8 MBytes 208 Mbits/sec 0 105\n> KBytes \n> [ 4] 1.00-2.00 sec 23.6 MBytes 198 Mbits/sec 0 105\n> KBytes \n> [ 4] 2.00-3.00 sec 24.5 MBytes 206 Mbits/sec 0 105\n> KBytes \n> [ 4] 3.00-4.00 sec 23.9 MBytes 201 Mbits/sec 0 105\n> KBytes \n> [ 4] 4.00-5.00 sec 24.6 MBytes 207 Mbits/sec 0 105\n> KBytes \n> [ 4] 5.00-6.00 sec 24.7 MBytes 207 Mbits/sec 0 105\n> KBytes \n> [ 4] 6.00-7.00 sec 24.5 MBytes 206 Mbits/sec 0 105\n> KBytes \n> [ 4] 7.00-8.00 sec 24.0 MBytes 201 Mbits/sec 0 105\n> KBytes \n> [ 4] 8.00-9.00 sec 24.3 MBytes 204 Mbits/sec 0 105\n> KBytes \n> [ 4] 9.00-10.00 sec 24.5 MBytes 206 Mbits/sec 0 105\n> KBytes \n> - - - - - - - - - - - - - - - - - - - - - - - - -\n> [ ID] Interval Transfer Bandwidth Retr\n> [ 4] 0.00-10.00 sec 244 MBytes 204\n> Mbits/sec 0 sender\n> [ 4] 0.00-10.00 sec 243 MBytes 204\n> Mbits/sec receiver\n> \n> iperf Done.\n> \n> So, the throughput went from 195 Mbits/sec to 272 Mbits/sec.\n> The gain would be (272 Mbps - 195 Mbps) / 195 Mbps = 0.3949 ~ 40%\n> \n> Regards,\n> Christian","headers":{"Return-Path":"","X-Original-To":"incoming@patchwork.ozlabs.org","Delivered-To":"patchwork-incoming@bilbo.ozlabs.org","Authentication-Results":["ozlabs.org; spf=none (mailfrom)\n\tsmtp.mailfrom=lists.infradead.org (client-ip=65.50.211.133;\n\thelo=bombadil.infradead.org;\n\tenvelope-from=lede-dev-bounces+incoming=patchwork.ozlabs.org@lists.infradead.org;\n\treceiver=)","ozlabs.org; dkim=pass (2048-bit key;\n\tunprotected) header.d=lists.infradead.org\n\theader.i=@lists.infradead.org header.b=\"VbIlLlMF\"; \n\tdkim=fail reason=\"signature verification failed\" (2048-bit key;\n\tunprotected) header.d=gmail.com header.i=@gmail.com\n\theader.b=\"RpDrVFoD\"; dkim-atps=neutral"],"Received":["from bombadil.infradead.org (bombadil.infradead.org\n\t[65.50.211.133])\n\t(using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256\n\tbits)) (No client certificate requested)\n\tby ozlabs.org (Postfix) with ESMTPS id 3xlp523xjqz9s06\n\tfor ;\n\tMon, 4 Sep 2017 08:47:02 +1000 (AEST)","from localhost ([127.0.0.1] helo=bombadil.infradead.org)\n\tby bombadil.infradead.org with esmtp (Exim 4.87 #1 (Red Hat Linux))\n\tid 1dodf7-00033k-0b; Sun, 03 Sep 2017 22:46:41 +0000","from mail-pg0-x231.google.com ([2607:f8b0:400e:c05::231])\n\tby bombadil.infradead.org with esmtps (Exim 4.87 #1 (Red Hat Linux))\n\tid 1dodf4-00030F-4n\n\tfor lede-dev@lists.infradead.org; Sun, 03 Sep 2017 22:46:39 +0000","by mail-pg0-x231.google.com with SMTP id b8so13863048pgn.5\n\tfor ;\n\tSun, 03 Sep 2017 15:46:17 -0700 (PDT)","from mangix.clevo ([2603:3024:2111:2b00::3daf])\n\tby smtp.gmail.com with ESMTPSA id\n\tb3sm7561165pgc.6.2017.09.03.15.46.15\n\t(version=TLS1_2 cipher=ECDHE-RSA-CHACHA20-POLY1305 bits=256/256);\n\tSun, 03 Sep 2017 15:46:16 -0700 (PDT)"],"DKIM-Signature":["v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed;\n\td=lists.infradead.org; s=bombadil.20170209; h=Sender:\n\tContent-Transfer-Encoding:Content-Type:List-Subscribe:List-Help:List-Post:\n\tList-Archive:List-Unsubscribe:List-Id:Subject:Mime-Version:References:\n\tIn-Reply-To:Date:To:From:Message-ID:Reply-To:Cc:Content-ID:\n\tContent-Description:Resent-Date:Resent-From:Resent-Sender:Resent-To:Resent-Cc\n\t:Resent-Message-ID:List-Owner;\n\tbh=tEnAC2Z8+IiF9R3iuYCM2MF4Pe3Ovh5xG5CDMRzhSgo=;\n\tb=VbIlLlMFbwzC7+8QjQYjaNtz1b\n\tANhNeLm2I/mcblezqGzfomuU0lXu7r30TM+G0Vu+qFomHPo7puzS/uylswZFOfz6RookoihnlrzSz\n\tloJCPHdNAnzSXpOlh9TJSOVf4MFZjy2EodQP0HW6Fdiez18r9t1ePo85+6gnRhNIrdQzIeq6+0fOG\n\tKRkP7gciatG7RJSQ180R57O4TaBTLLftGYsLyfQyO2LQ4aaEO0u0Ff0mN/+JL9lOUe/2tJcA+zQZJ\n\t2VNI2198d4Qmva/UaFQH4vAsdA617GBhvHjBv6qhigjrEhrLKEo+067JMTbEcrTHoKgnOQ08+qNxS\n\t4Q57QHlQ==;","v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025;\n\th=message-id:subject:from:to:date:in-reply-to:references:mime-version\n\t:content-transfer-encoding;\n\tbh=74kttzmTodVTyDyFBsoWQyw6Hcylh+ZZfrAMM2pGMpM=;\n\tb=RpDrVFoDQvGuugm/HwmF68Hg8fI+l90HTBbs0ZlzPKqoiMdYCj/1xdJIgHjPtoxTXn\n\tIarERLGW+FErsKj1z/Gjf9d97dfY/+WfsuZYdkpZrBTlM/OKdbW3wPhFc7BY4mKfbJSR\n\tRWAehNZONCCPxX/NyG08qU5eYIcHdVL6B//5TEcySq+1L0JYGNza8KG+RAwq0PicSAuM\n\tpCAWU+8TzOtY8W7nnahvHAdOBv4Jy63YKT6ZlgdVqfJmU0AsUNUgzppahQmwvJnbpeEI\n\tCeupmOYuSRiT7bPEEnxx1MxECT00DOW2K5/37Y+zdVYW4jA4NLZmhcGa2yFVOhqaf+S7\n\tOSaA=="],"X-Google-DKIM-Signature":"v=1; a=rsa-sha256; c=relaxed/relaxed;\n\td=1e100.net; s=20161025;\n\th=x-gm-message-state:message-id:subject:from:to:date:in-reply-to\n\t:references:mime-version:content-transfer-encoding;\n\tbh=74kttzmTodVTyDyFBsoWQyw6Hcylh+ZZfrAMM2pGMpM=;\n\tb=cfO8k37B1g/1BVK9/ajx7QkrJj3EI4r7AD2tP6WVBAC9AtylzjkdvKXCzVydEFOVkH\n\t2Pxpl5rDa8SpDW6j/8wi8gs56mvaXxo3iyPlpnTT0CHy2GSRnj/nsvUauxuB1Z+YP3xy\n\tueszG5L/0rkz68NeiblznzEeRJxm0IWzMIhqsaXlWcCkvW0zygjleR3N2Pu+nduYrUyg\n\tBdnHJaceiSLzXpHMNhMx9gnUx4Dv8MdsSoWm7e5Joej2k4BNs2XaeqdZ4QbpH8iKbFQv\n\t/JXCHiEvAriLhO+QG898BjPPwbYIxgXTQH5grqzA2lLwKuj6qw7gn/jH18h2IU/qnCVl\n\tkVGw==","X-Gm-Message-State":"AHPjjUha5FvclGuMVB9FjNeS9Mr31HxObVgz424ASJT6hLx+vy02uiTe\n\tV5W+tO9T7ATmN+egaHw=","X-Google-Smtp-Source":"ADKCNb4ypmfH2DtGUXkfqkOy1Tg115W+kU9ZwqAaX2xnGvoIZ0DWto3WMR0tioyCx8JCuQgHVMpWWg==","X-Received":"by 10.99.174.65 with SMTP id e1mr9876833pgp.149.1504478777180;\n\tSun, 03 Sep 2017 15:46:17 -0700 (PDT)","Message-ID":"<1504478774.2387.8.camel@gmail.com>","From":"rosenp@gmail.com","To":"Christian Lamparter , \n\tlede-dev@lists.infradead.org","Date":"Sun, 03 Sep 2017 15:46:14 -0700","In-Reply-To":"<1510043.ZdtWCe27Fq@debian64>","References":"<20170903183546.16749-1-rosenp@gmail.com>\n\t<1510043.ZdtWCe27Fq@debian64>","X-Mailer":"Evolution 3.24.5 (3.24.5-1.fc26) ","Mime-Version":"1.0","X-CRM114-Version":"20100106-BlameMichelson ( TRE 0.8.0 (BSD) ) MR-646709E3 ","X-CRM114-CacheID":"sfid-20170903_154638_245252_40C08925 ","X-CRM114-Status":"GOOD ( 17.19 )","X-Spam-Score":"-2.0 (--)","X-Spam-Report":"SpamAssassin version 3.4.1 on bombadil.infradead.org summary:\n\tContent analysis details: (-2.0 points)\n\tpts rule name description\n\t---- ----------------------\n\t--------------------------------------------------\n\t-0.0 RCVD_IN_DNSWL_NONE RBL: Sender listed at http://www.dnswl.org/,\n\tno\n\ttrust [2607:f8b0:400e:c05:0:0:0:231 listed in] [list.dnswl.org]\n\t-0.0 SPF_PASS SPF: sender matches SPF record\n\t0.0 FREEMAIL_FROM Sender email is commonly abused enduser mail\n\tprovider (rosenp[at]gmail.com)\n\t-1.9 BAYES_00 BODY: Bayes spam probability is 0 to 1%\n\t[score: 0.0000]\n\t-0.1 DKIM_VALID Message has at least one valid DKIM or DK signature\n\t0.1 DKIM_SIGNED Message has a DKIM or DK signature,\n\tnot necessarily valid\n\t-0.1 DKIM_VALID_AU Message has a valid DKIM or DK signature from\n\tauthor's domain","Subject":"Re: [LEDE-DEV] [PATCH] ar71xx: Add GRO support to ag71xx","X-BeenThere":"lede-dev@lists.infradead.org","X-Mailman-Version":"2.1.21","Precedence":"list","List-Id":"","List-Unsubscribe":",\n\t","List-Archive":"","List-Post":"","List-Help":"","List-Subscribe":",\n\t","Content-Type":"text/plain; charset=\"utf-8\"","Content-Transfer-Encoding":"base64","Sender":"\"Lede-dev\" ","Errors-To":"lede-dev-bounces+incoming=patchwork.ozlabs.org@lists.infradead.org"}},{"id":1762392,"web_url":"http://patchwork.ozlabs.org/comment/1762392/","msgid":"<937dc2ec-0951-85af-3b35-f42956e3d360@gmail.com>","list_archive_url":null,"date":"2017-09-04T02:10:02","subject":"Re: [LEDE-DEV] [PATCH] ar71xx: Add GRO support to ag71xx","submitter":{"id":2800,"url":"http://patchwork.ozlabs.org/api/people/2800/","name":"Florian Fainelli","email":"f.fainelli@gmail.com"},"content":"Le 09/03/17 à 15:46, rosenp@gmail.com a écrit :\n> That's...way better than I expected given how minimal my changes are.\n> \n> For some context, the ag71xx driver is special in that it does not seem\n> to do any hardware offloading to the NIC.\n> \n> As far as I understand this change, GRO takes 1500 MTU packets and\n> packs then into 64Kb blocks which the kernel then processes.\n> \n> I would be curious if anyone can do latency comparisons before this\n> change and after. I do know this driver to have lower latency than\n> others due to lack of offloads.\n> \n> I guess all that's left is to add GSO support to the driver. That seems\n> like a lot more work than a three line change though. ¯\\_(ツ)_/¯\n\nYou could look into adding software TSO:\n\nhttps://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/drivers/net/ethernet/marvell/mv643xx_eth.c?id=3ae8f4e0b98b640aadf410c21185ccb6b5b02351\n\nthis makes a huge difference with mv643xx_eth. There were a number of\nfixes on top of this initial commit but I would be curious to see what\nit gives you with ag71xx\n\n> \n> On Sun, 2017-09-03 at 23:16 +0200, Christian Lamparter wrote:\n>> On Sunday, September 3, 2017 11:35:46 AM CEST Rosen Penev wrote:\n>>> On a TL-WN710N, this patch increases iperf performance from ~92.5\n>>> to ~93.5 mbps. Keep in mind the WN710N is a 100mbps device. I\n>>> expect greater numbers from gigabit devices.\n>>>\n>>> Signed-off-by: Rosen Penev \n>>> ---\n>>\n>> I've done a quick test of the patch on my WD Range Extender.\n>> (It has a Atheros AR9344 rev 2 SoC @ CPU:560.000MHz, DDR:400.000MHz\n>> The PHY is a AR8035, which supports 1 GBit/s Links)\n>>\n>> The range extender (DUT) was running iperf3 server in both tests.\n>> Another desktop PC was acting as the iperf3 client.\n>>\n>> without the patch:\n>>\n>> Connecting to host range-extender, port 5201\n>> [ 4] local 192.168.8.7 port 51518 connected to 192.168.8.204 port\n>> 5201\n>> [ ID] Interval Transfer Bandwidth Retr Cwnd\n>> [ 4] 0.00-1.00 sec 23.5 MBytes 197 Mbits/sec 0 105\n>> KBytes\n>> [ 4] 1.00-2.00 sec 23.7 MBytes 199 Mbits/sec 0 105\n>> KBytes\n>> [ 4] 2.00-3.00 sec 23.6 MBytes 198 Mbits/sec 0 105\n>> KBytes\n>> [ 4] 3.00-4.00 sec 23.0 MBytes 193 Mbits/sec 0 105\n>> KBytes\n>> [ 4] 4.00-5.00 sec 23.4 MBytes 197 Mbits/sec 0 105\n>> KBytes\n>> [ 4] 5.00-6.00 sec 23.3 MBytes 195 Mbits/sec 0 105\n>> KBytes\n>> [ 4] 6.00-7.00 sec 23.4 MBytes 196 Mbits/sec 0 105\n>> KBytes\n>> [ 4] 7.00-8.00 sec 23.6 MBytes 198 Mbits/sec 0 105\n>> KBytes\n>> [ 4] 8.00-9.00 sec 23.1 MBytes 194 Mbits/sec 0 105\n>> KBytes\n>> [ 4] 9.00-10.00 sec 22.1 MBytes 185 Mbits/sec 0 105\n>> KBytes\n>> - - - - - - - - - - - - - - - - - - - - - - - - -\n>> [ ID] Interval Transfer Bandwidth Retr\n>> [ 4] 0.00-10.00 sec 233 MBytes 195\n>> Mbits/sec 0 sender\n>> [ 4] 0.00-10.00 sec 232 MBytes 195\n>> Mbits/sec receiver\n>>\n>> iperf Done.\n>>\n>> with the patch (gro enabled - this is done by default):\n>>\n>> Connecting to host range-extender, port 5201\n>> [ 4] local 192.168.8.7 port 52004 connected to 192.168.8.204 port\n>> 5201\n>> [ ID] Interval Transfer Bandwidth Retr Cwnd\n>> [ 4] 0.00-1.00 sec 32.7 MBytes 274 Mbits/sec 0 106\n>> KBytes \n>> [ 4] 1.00-2.00 sec 32.6 MBytes 274 Mbits/sec 0 106\n>> KBytes \n>> [ 4] 2.00-3.00 sec 32.4 MBytes 272 Mbits/sec 0 106\n>> KBytes \n>> [ 4] 3.00-4.00 sec 32.3 MBytes 271 Mbits/sec 0 106\n>> KBytes \n>> [ 4] 4.00-5.00 sec 32.5 MBytes 273 Mbits/sec 0 106\n>> KBytes \n>> [ 4] 5.00-6.00 sec 32.5 MBytes 273 Mbits/sec 0 106\n>> KBytes \n>> [ 4] 6.00-7.00 sec 32.6 MBytes 273 Mbits/sec 0 106\n>> KBytes \n>> [ 4] 7.00-8.00 sec 32.4 MBytes 272 Mbits/sec 0 106\n>> KBytes \n>> [ 4] 8.00-9.00 sec 32.6 MBytes 273 Mbits/sec 0 106\n>> KBytes \n>> [ 4] 9.00-10.00 sec 31.4 MBytes 264 Mbits/sec 0 106\n>> KBytes \n>> - - - - - - - - - - - - - - - - - - - - - - - - -\n>> [ ID] Interval Transfer Bandwidth Retr\n>> [ 4] 0.00-10.00 sec 324 MBytes 272\n>> Mbits/sec 0 sender\n>> [ 4] 0.00-10.00 sec 324 MBytes 272\n>> Mbits/sec receiver\n>>\n>> iperf Done.\n>>\n>> (range-extender) # ethtool -K eth0 gro off\n>>\n>> Connecting to host range-extender, port 5201\n>> [ 4] local 192.168.8.7 port 52120 connected to 192.168.8.204 port\n>> 5201\n>> [ ID] Interval Transfer Bandwidth Retr Cwnd\n>> [ 4] 0.00-1.00 sec 24.8 MBytes 208 Mbits/sec 0 105\n>> KBytes \n>> [ 4] 1.00-2.00 sec 23.6 MBytes 198 Mbits/sec 0 105\n>> KBytes \n>> [ 4] 2.00-3.00 sec 24.5 MBytes 206 Mbits/sec 0 105\n>> KBytes \n>> [ 4] 3.00-4.00 sec 23.9 MBytes 201 Mbits/sec 0 105\n>> KBytes \n>> [ 4] 4.00-5.00 sec 24.6 MBytes 207 Mbits/sec 0 105\n>> KBytes \n>> [ 4] 5.00-6.00 sec 24.7 MBytes 207 Mbits/sec 0 105\n>> KBytes \n>> [ 4] 6.00-7.00 sec 24.5 MBytes 206 Mbits/sec 0 105\n>> KBytes \n>> [ 4] 7.00-8.00 sec 24.0 MBytes 201 Mbits/sec 0 105\n>> KBytes \n>> [ 4] 8.00-9.00 sec 24.3 MBytes 204 Mbits/sec 0 105\n>> KBytes \n>> [ 4] 9.00-10.00 sec 24.5 MBytes 206 Mbits/sec 0 105\n>> KBytes \n>> - - - - - - - - - - - - - - - - - - - - - - - - -\n>> [ ID] Interval Transfer Bandwidth Retr\n>> [ 4] 0.00-10.00 sec 244 MBytes 204\n>> Mbits/sec 0 sender\n>> [ 4] 0.00-10.00 sec 243 MBytes 204\n>> Mbits/sec receiver\n>>\n>> iperf Done.\n>>\n>> So, the throughput went from 195 Mbits/sec to 272 Mbits/sec.\n>> The gain would be (272 Mbps - 195 Mbps) / 195 Mbps = 0.3949 ~ 40%\n>>\n>> Regards,\n>> Christian\n> \n> _______________________________________________\n> Lede-dev mailing list\n> Lede-dev@lists.infradead.org\n> http://lists.infradead.org/mailman/listinfo/lede-dev\n>","headers":{"Return-Path":"","X-Original-To":"incoming@patchwork.ozlabs.org","Delivered-To":"patchwork-incoming@bilbo.ozlabs.org","Authentication-Results":["ozlabs.org; spf=none (mailfrom)\n\tsmtp.mailfrom=lists.infradead.org (client-ip=65.50.211.133;\n\thelo=bombadil.infradead.org;\n\tenvelope-from=lede-dev-bounces+incoming=patchwork.ozlabs.org@lists.infradead.org;\n\treceiver=)","ozlabs.org; dkim=pass (2048-bit key;\n\tunprotected) header.d=lists.infradead.org\n\theader.i=@lists.infradead.org header.b=\"hglhkgsC\"; \n\tdkim=fail reason=\"signature verification failed\" (2048-bit key;\n\tunprotected) header.d=gmail.com header.i=@gmail.com\n\theader.b=\"AsdLeq1b\"; dkim-atps=neutral"],"Received":["from bombadil.infradead.org (bombadil.infradead.org\n\t[65.50.211.133])\n\t(using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256\n\tbits)) (No client certificate requested)\n\tby ozlabs.org (Postfix) with ESMTPS id 3xltcH53Xbz9sNr\n\tfor ;\n\tMon, 4 Sep 2017 12:10:55 +1000 (AEST)","from localhost ([127.0.0.1] helo=bombadil.infradead.org)\n\tby bombadil.infradead.org with esmtp (Exim 4.87 #1 (Red Hat Linux))\n\tid 1dogqO-0006GX-V1; Mon, 04 Sep 2017 02:10:32 +0000","from mail-oi0-x243.google.com ([2607:f8b0:4003:c06::243])\n\tby bombadil.infradead.org with esmtps (Exim 4.87 #1 (Red Hat Linux))\n\tid 1dogqK-0005NP-Ty\n\tfor lede-dev@lists.infradead.org; Mon, 04 Sep 2017 02:10:31 +0000","by mail-oi0-x243.google.com with SMTP id r203so4058226oih.5\n\tfor ;\n\tSun, 03 Sep 2017 19:10:06 -0700 (PDT)","from ?IPv6:2001:470:d:73f:89f7:5bc5:8791:e113?\n\t([2001:470:d:73f:89f7:5bc5:8791:e113])\n\tby smtp.googlemail.com with ESMTPSA id\n\te184sm8652610oih.35.2017.09.03.19.10.03\n\t(version=TLS1_2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128/128);\n\tSun, 03 Sep 2017 19:10:04 -0700 (PDT)"],"DKIM-Signature":["v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed;\n\td=lists.infradead.org; s=bombadil.20170209; h=Sender:\n\tContent-Transfer-Encoding:Content-Type:List-Subscribe:List-Help:List-Post:\n\tList-Archive:List-Unsubscribe:List-Id:Subject:In-Reply-To:MIME-Version:Date:\n\tMessage-ID:From:References:To:Reply-To:Cc:Content-ID:Content-Description:\n\tResent-Date:Resent-From:Resent-Sender:Resent-To:Resent-Cc:Resent-Message-ID:\n\tList-Owner; bh=yR1xKgfYipH7fYSyTaquUpoOLcdEfnByjvkGmBDB8Hk=;\n\tb=hglhkgsCnD4EuO\n\t3cCLxctnpmpJLWsCqZmJldDLZBldnVO3D3yztVWY6J6gQ/fbUBqIJ/1RlR+vi103BYcCYiB7x9ySr\n\tWhCNOckWyPtNWByMfTb83UsdKTVTL3xa2Lw4GnEtm0qSlnX/9Cw8m2oktPDp5VTNnCmD45JlxOxoR\n\tKlnpv6n2qZHl/n71aE19djS+ZHoJkFrFrZ9C7ks81YguiBHm+i7XHKRRWF5Qxwu0A5tHVUOPeJ0Yk\n\tUK3VwI5byPgR+8+DqEJFsQRwtEdWLpv0CH3bGgA9gFkToLkCkKJaOaBzvDcPKwkUG1DKgKNthoXuk\n\t+X1EbEGPgDfzhJgIsTlw==;","v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025;\n\th=subject:to:references:from:message-id:date:user-agent:mime-version\n\t:in-reply-to:content-language:content-transfer-encoding;\n\tbh=dBmdXQ/w3fVbDayGaUU9jKbMefKyldJ+t5qV/JA9VMA=;\n\tb=AsdLeq1bva7JAgUwgp5/34xRQ0w3/yCuh+q4tCxcRVq/5VqVObwSmbgmuuP4rh7FE5\n\tSgFsLmv0qAINSFlewwyvkpAgOucDV8zEYhyo7+GAVdShH94yi04oFFJZp1RbnNgWOPHZ\n\tOoIf/OtHZWXCySFKmn7UEcLR10SSbNtubOUKzNC0TyaQqt6d9Ttm6EYaMC87w+ubEPtc\n\t7dayY8/tgo3sZeifCRLSktmG8LkN6vZ76UuQ0P1/uDADMb/Gs+isU0xza1wWIy/U6KzO\n\tTfxEuPdu/X1QLMGbhQVDqRpJ04EUHauDUn1s06Gr+bgyQN8HIcyT9WpgqmJXwwSLW4th\n\tzs7w=="],"X-Google-DKIM-Signature":"v=1; a=rsa-sha256; c=relaxed/relaxed;\n\td=1e100.net; s=20161025;\n\th=x-gm-message-state:subject:to:references:from:message-id:date\n\t:user-agent:mime-version:in-reply-to:content-language\n\t:content-transfer-encoding;\n\tbh=dBmdXQ/w3fVbDayGaUU9jKbMefKyldJ+t5qV/JA9VMA=;\n\tb=EqvkcmvTySFEL5n79Juy7TrAy0XrqQxxdaeGkzMscBequqHt7kVuYrRkSy8LTSffl/\n\tx7RvrMJZBrRSx/MCYEVGFgbDAaP6R5Fk00LZgbX81RLbFRx+PcOQ7OwefFOSKm2mej4V\n\tSC7VPtHvecygxpVEfYLGNaWkQNChXNcapUnYNK1oXnmCDXVvgC2bnro/yo0cjDzhCmD0\n\topMELU9fAT33k5PLo61SqBenrTsyDympdiMET47mQeYa3acB4SEQ+tq8IgY9X0lCjWT2\n\toO8w8O6THcxaJjWzRWhCHaDOVaiwu2AdIvPzUJhFANeoJRMDiolsoEiDz/LgyShlTm+M\n\tbNhg==","X-Gm-Message-State":"AHPjjUhPRG4F5krNedWiUANk+jFtpKxYSUMTWv9JXmeD8zO003jf/vCD\n\tMc3+ZfOkGyOhitZwAM0=","X-Google-Smtp-Source":"ADKCNb7NJb2YDGz5BLLJbpLlyDh8gDJMpLDdPOweF+ALVosfFpGwXl1CYJLL68OBSKUBWe1GWZET/Q==","X-Received":"by 10.202.214.143 with SMTP id\n\tn137mr8939066oig.208.1504491005411; \n\tSun, 03 Sep 2017 19:10:05 -0700 (PDT)","To":"rosenp@gmail.com, Christian Lamparter ,\n\tlede-dev@lists.infradead.org","References":"<20170903183546.16749-1-rosenp@gmail.com>\n\t<1510043.ZdtWCe27Fq@debian64> <1504478774.2387.8.camel@gmail.com>","From":"Florian Fainelli ","Message-ID":"<937dc2ec-0951-85af-3b35-f42956e3d360@gmail.com>","Date":"Sun, 3 Sep 2017 19:10:02 -0700","User-Agent":"Mozilla/5.0 (X11; Linux x86_64; rv:52.0) Gecko/20100101\n\tThunderbird/52.2.1","MIME-Version":"1.0","In-Reply-To":"<1504478774.2387.8.camel@gmail.com>","Content-Language":"en-US","X-CRM114-Version":"20100106-BlameMichelson ( TRE 0.8.0 (BSD) ) MR-646709E3 ","X-CRM114-CacheID":"sfid-20170903_191029_054168_AC7423AB ","X-CRM114-Status":"GOOD ( 21.49 )","X-Spam-Score":"-1.0 (-)","X-Spam-Report":"SpamAssassin version 3.4.1 on bombadil.infradead.org summary:\n\tContent analysis details: (-1.0 points)\n\tpts rule name description\n\t---- ----------------------\n\t--------------------------------------------------\n\t-0.0 RCVD_IN_DNSWL_NONE RBL: Sender listed at http://www.dnswl.org/,\n\tno\n\ttrust [2607:f8b0:4003:c06:0:0:0:243 listed in] [list.dnswl.org]\n\t-0.0 SPF_PASS SPF: sender matches SPF record\n\t0.0 FREEMAIL_FROM Sender email is commonly abused enduser mail\n\tprovider (f.fainelli[at]gmail.com)\n\t-1.9 BAYES_00 BODY: Bayes spam probability is 0 to 1%\n\t[score: 0.0000]\n\t-0.1 DKIM_VALID Message has at least one valid DKIM or DK signature\n\t0.1 DKIM_SIGNED Message has a DKIM or DK signature,\n\tnot necessarily valid\n\t-0.1 DKIM_VALID_AU Message has a valid DKIM or DK signature from\n\tauthor's domain\n\t1.0 FREEMAIL_REPLY From and body contain different freemails","Subject":"Re: [LEDE-DEV] [PATCH] ar71xx: Add GRO support to ag71xx","X-BeenThere":"lede-dev@lists.infradead.org","X-Mailman-Version":"2.1.21","Precedence":"list","List-Id":"","List-Unsubscribe":",\n\t","List-Archive":"","List-Post":"","List-Help":"","List-Subscribe":",\n\t","Content-Type":"text/plain; charset=\"utf-8\"","Content-Transfer-Encoding":"base64","Sender":"\"Lede-dev\" ","Errors-To":"lede-dev-bounces+incoming=patchwork.ozlabs.org@lists.infradead.org"}},{"id":1762752,"web_url":"http://patchwork.ozlabs.org/comment/1762752/","msgid":"<1504541751.3761.4.camel@gmail.com>","list_archive_url":null,"date":"2017-09-04T16:15:51","subject":"Re: [LEDE-DEV] [PATCH] ar71xx: Add GRO support to ag71xx","submitter":{"id":70304,"url":"http://patchwork.ozlabs.org/api/people/70304/","name":"Rosen Penev","email":"rosenp@gmail.com"},"content":"After a bit more research, it turns out this change is really boring.\nEric Dumazet did this for many ethernet drivers early this year.\n\nAs far as TSO goes, no idea how to even go about that. As an\ninteresting side note, TSO support broke mv643xx_eth for a while.\n\nOn Sun, 2017-09-03 at 19:10 -0700, Florian Fainelli wrote:\n> Le 09/03/17 à 15:46, rosenp@gmail.com a écrit :\n> > That's...way better than I expected given how minimal my changes\n> > are.\n> > \n> > For some context, the ag71xx driver is special in that it does not\n> > seem\n> > to do any hardware offloading to the NIC.\n> > \n> > As far as I understand this change, GRO takes 1500 MTU packets and\n> > packs then into 64Kb blocks which the kernel then processes.\n> > \n> > I would be curious if anyone can do latency comparisons before this\n> > change and after. I do know this driver to have lower latency than\n> > others due to lack of offloads.\n> > \n> > I guess all that's left is to add GSO support to the driver. That\n> > seems\n> > like a lot more work than a three line change though. ¯\\_(ツ)_/¯\n> \n> You could look into adding software TSO:\n> \n> https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/co\n> mmit/drivers/net/ethernet/marvell/mv643xx_eth.c?id=3ae8f4e0b98b640aad\n> f410c21185ccb6b5b02351\n> \n> this makes a huge difference with mv643xx_eth. There were a number of\n> fixes on top of this initial commit but I would be curious to see\n> what\n> it gives you with ag71xx\n> \n> > \n> > On Sun, 2017-09-03 at 23:16 +0200, Christian Lamparter wrote:\n> > > On Sunday, September 3, 2017 11:35:46 AM CEST Rosen Penev wrote:\n> > > > On a TL-WN710N, this patch increases iperf performance from\n> > > > ~92.5\n> > > > to ~93.5 mbps. Keep in mind the WN710N is a 100mbps device. I\n> > > > expect greater numbers from gigabit devices.\n> > > > \n> > > > Signed-off-by: Rosen Penev \n> > > > ---\n> > > \n> > > I've done a quick test of the patch on my WD Range Extender.\n> > > (It has a Atheros AR9344 rev 2 SoC @ CPU:560.000MHz,\n> > > DDR:400.000MHz\n> > > The PHY is a AR8035, which supports 1 GBit/s Links)\n> > > \n> > > The range extender (DUT) was running iperf3 server in both tests.\n> > > Another desktop PC was acting as the iperf3 client.\n> > > \n> > > without the patch:\n> > > \n> > > Connecting to host range-extender, port 5201\n> > > [ 4] local 192.168.8.7 port 51518 connected to 192.168.8.204\n> > > port\n> > > 5201\n> > > [ ID] Interval Transfer Bandwidth Retr Cwnd\n> > > [ 4] 0.00-1.00 sec 23.5 MBytes 197 Mbits/sec 0 105\n> > > KBytes\n> > > [ 4] 1.00-2.00 sec 23.7 MBytes 199 Mbits/sec 0 105\n> > > KBytes\n> > > [ 4] 2.00-3.00 sec 23.6 MBytes 198 Mbits/sec 0 105\n> > > KBytes\n> > > [ 4] 3.00-4.00 sec 23.0 MBytes 193 Mbits/sec 0 105\n> > > KBytes\n> > > [ 4] 4.00-5.00 sec 23.4 MBytes 197 Mbits/sec 0 105\n> > > KBytes\n> > > [ 4] 5.00-6.00 sec 23.3 MBytes 195 Mbits/sec 0 105\n> > > KBytes\n> > > [ 4] 6.00-7.00 sec 23.4 MBytes 196 Mbits/sec 0 105\n> > > KBytes\n> > > [ 4] 7.00-8.00 sec 23.6 MBytes 198 Mbits/sec 0 105\n> > > KBytes\n> > > [ 4] 8.00-9.00 sec 23.1 MBytes 194 Mbits/sec 0 105\n> > > KBytes\n> > > [ 4] 9.00-10.00 sec 22.1 MBytes 185 Mbits/sec 0 105\n> > > KBytes\n> > > - - - - - - - - - - - - - - - - - - - - - - - - -\n> > > [ ID] Interval Transfer Bandwidth Retr\n> > > [ 4] 0.00-10.00 sec 233 MBytes 195\n> > > Mbits/sec 0 sender\n> > > [ 4] 0.00-10.00 sec 232 MBytes 195\n> > > Mbits/sec receiver\n> > > \n> > > iperf Done.\n> > > \n> > > with the patch (gro enabled - this is done by default):\n> > > \n> > > Connecting to host range-extender, port 5201\n> > > [ 4] local 192.168.8.7 port 52004 connected to 192.168.8.204\n> > > port\n> > > 5201\n> > > [ ID] Interval Transfer Bandwidth Retr Cwnd\n> > > [ 4] 0.00-1.00 sec 32.7 MBytes 274 Mbits/sec 0 106\n> > > KBytes \n> > > [ 4] 1.00-2.00 sec 32.6 MBytes 274 Mbits/sec 0 106\n> > > KBytes \n> > > [ 4] 2.00-3.00 sec 32.4 MBytes 272 Mbits/sec 0 106\n> > > KBytes \n> > > [ 4] 3.00-4.00 sec 32.3 MBytes 271 Mbits/sec 0 106\n> > > KBytes \n> > > [ 4] 4.00-5.00 sec 32.5 MBytes 273 Mbits/sec 0 106\n> > > KBytes \n> > > [ 4] 5.00-6.00 sec 32.5 MBytes 273 Mbits/sec 0 106\n> > > KBytes \n> > > [ 4] 6.00-7.00 sec 32.6 MBytes 273 Mbits/sec 0 106\n> > > KBytes \n> > > [ 4] 7.00-8.00 sec 32.4 MBytes 272 Mbits/sec 0 106\n> > > KBytes \n> > > [ 4] 8.00-9.00 sec 32.6 MBytes 273 Mbits/sec 0 106\n> > > KBytes \n> > > [ 4] 9.00-10.00 sec 31.4 MBytes 264 Mbits/sec 0 106\n> > > KBytes \n> > > - - - - - - - - - - - - - - - - - - - - - - - - -\n> > > [ ID] Interval Transfer Bandwidth Retr\n> > > [ 4] 0.00-10.00 sec 324 MBytes 272\n> > > Mbits/sec 0 sender\n> > > [ 4] 0.00-10.00 sec 324 MBytes 272\n> > > Mbits/sec receiver\n> > > \n> > > iperf Done.\n> > > \n> > > (range-extender) # ethtool -K eth0 gro off\n> > > \n> > > Connecting to host range-extender, port 5201\n> > > [ 4] local 192.168.8.7 port 52120 connected to 192.168.8.204\n> > > port\n> > > 5201\n> > > [ ID] Interval Transfer Bandwidth Retr Cwnd\n> > > [ 4] 0.00-1.00 sec 24.8 MBytes 208 Mbits/sec 0 105\n> > > KBytes \n> > > [ 4] 1.00-2.00 sec 23.6 MBytes 198 Mbits/sec 0 105\n> > > KBytes \n> > > [ 4] 2.00-3.00 sec 24.5 MBytes 206 Mbits/sec 0 105\n> > > KBytes \n> > > [ 4] 3.00-4.00 sec 23.9 MBytes 201 Mbits/sec 0 105\n> > > KBytes \n> > > [ 4] 4.00-5.00 sec 24.6 MBytes 207 Mbits/sec 0 105\n> > > KBytes \n> > > [ 4] 5.00-6.00 sec 24.7 MBytes 207 Mbits/sec 0 105\n> > > KBytes \n> > > [ 4] 6.00-7.00 sec 24.5 MBytes 206 Mbits/sec 0 105\n> > > KBytes \n> > > [ 4] 7.00-8.00 sec 24.0 MBytes 201 Mbits/sec 0 105\n> > > KBytes \n> > > [ 4] 8.00-9.00 sec 24.3 MBytes 204 Mbits/sec 0 105\n> > > KBytes \n> > > [ 4] 9.00-10.00 sec 24.5 MBytes 206 Mbits/sec 0 105\n> > > KBytes \n> > > - - - - - - - - - - - - - - - - - - - - - - - - -\n> > > [ ID] Interval Transfer Bandwidth Retr\n> > > [ 4] 0.00-10.00 sec 244 MBytes 204\n> > > Mbits/sec 0 sender\n> > > [ 4] 0.00-10.00 sec 243 MBytes 204\n> > > Mbits/sec receiver\n> > > \n> > > iperf Done.\n> > > \n> > > So, the throughput went from 195 Mbits/sec to 272 Mbits/sec.\n> > > The gain would be (272 Mbps - 195 Mbps) / 195 Mbps = 0.3949 ~ 40%\n> > > \n> > > Regards,\n> > > Christian\n> > \n> > _______________________________________________\n> > Lede-dev mailing list\n> > Lede-dev@lists.infradead.org\n> > http://lists.infradead.org/mailman/listinfo/lede-dev\n> > \n> \n>","headers":{"Return-Path":"","X-Original-To":"incoming@patchwork.ozlabs.org","Delivered-To":"patchwork-incoming@bilbo.ozlabs.org","Authentication-Results":["ozlabs.org; spf=none (mailfrom)\n\tsmtp.mailfrom=lists.infradead.org (client-ip=65.50.211.133;\n\thelo=bombadil.infradead.org;\n\tenvelope-from=lede-dev-bounces+incoming=patchwork.ozlabs.org@lists.infradead.org;\n\treceiver=)","ozlabs.org; dkim=pass (2048-bit key;\n\tunprotected) header.d=lists.infradead.org\n\theader.i=@lists.infradead.org header.b=\"JQ2+eflv\"; \n\tdkim=fail reason=\"signature verification failed\" (2048-bit key;\n\tunprotected) header.d=gmail.com header.i=@gmail.com\n\theader.b=\"ebEmlWOF\"; dkim-atps=neutral"],"Received":["from bombadil.infradead.org (bombadil.infradead.org\n\t[65.50.211.133])\n\t(using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256\n\tbits)) (No client certificate requested)\n\tby ozlabs.org (Postfix) with ESMTPS id 3xmFN80ySyz9t2Z\n\tfor ;\n\tTue, 5 Sep 2017 02:16:40 +1000 (AEST)","from localhost ([127.0.0.1] helo=bombadil.infradead.org)\n\tby bombadil.infradead.org with esmtp (Exim 4.87 #1 (Red Hat Linux))\n\tid 1dou2v-00040P-QO; Mon, 04 Sep 2017 16:16:21 +0000","from mail-pf0-x22f.google.com ([2607:f8b0:400e:c00::22f])\n\tby bombadil.infradead.org with esmtps (Exim 4.87 #1 (Red Hat Linux))\n\tid 1dou2q-0003hU-11\n\tfor lede-dev@lists.infradead.org; Mon, 04 Sep 2017 16:16:19 +0000","by mail-pf0-x22f.google.com with SMTP id g13so1995110pfm.2\n\tfor ;\n\tMon, 04 Sep 2017 09:15:55 -0700 (PDT)","from mangix.lan ([2001:470:1f05:c3e::8ce])\n\tby smtp.gmail.com with ESMTPSA id\n\tw15sm13720035pgo.63.2017.09.04.09.15.52\n\tfor \n\t(version=TLS1_2 cipher=ECDHE-RSA-CHACHA20-POLY1305 bits=256/256);\n\tMon, 04 Sep 2017 09:15:53 -0700 (PDT)"],"DKIM-Signature":["v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed;\n\td=lists.infradead.org; s=bombadil.20170209; h=Sender:\n\tContent-Transfer-Encoding:Content-Type:List-Subscribe:List-Help:List-Post:\n\tList-Archive:List-Unsubscribe:List-Id:Subject:Mime-Version:References:\n\tIn-Reply-To:Date:To:From:Message-ID:Reply-To:Cc:Content-ID:\n\tContent-Description:Resent-Date:Resent-From:Resent-Sender:Resent-To:Resent-Cc\n\t:Resent-Message-ID:List-Owner;\n\tbh=/Nh5ET7ZSsAPjV1AXmfysXr8+OzUnTfqcEFBUB6p2T8=;\n\tb=JQ2+eflvbtiNIh5cRX/SQXNbw2\n\tZhQHH0QaJTEQtBIclRl1AANuuH1195qfC1ynp/c7ko38G/fMw4QUndZBT5lHBGgZ0uuTgvgcc1KTS\n\tKIrK5tzFtpdsX0zcJxiQ6JTJi0q/rbdMFsLapUCsmanwwViOAz+elVV2jyOR9ub7IESFoSGto6H4S\n\tHO9140gPkwg9hzTLjApy68pnuhT0fnOahF7PlnVU+ksy1QjWzY9EY38kftusRVl73/TDklXYSWSsb\n\tWjKJ10uWLCZVYdh0oHgHSgxHFj7ol9Xn1EoAa7KsYg1jhC7qefKB0bwIfKYr6JmuSjz/ZAPDFsE22\n\t6SidziVg==;","v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025;\n\th=message-id:subject:from:to:date:in-reply-to:references:mime-version\n\t:content-transfer-encoding;\n\tbh=+CC282jKuD5rBF31hUyMAvDu3BpHHs+RQe9LxZOwWas=;\n\tb=ebEmlWOFXboGH0FLg/JmMC95RM8yZEJWLHRwH5A6vkrij5nwe9KackBfe1dQCPueAk\n\tPlEqwZI3hbWdHFxZzLCynrUP5+VEKSAjxZsc3JDIAjfQK5SmfJcyto1htaIuXO9WXrRN\n\thKjPRKEe3RTXVVEh5xesc6D9MbhGCRygHZqObYHfbQczJnY9/362Pa/RFiE0aWgdpzrz\n\tccvxcFjftlfAOZtocziBZBlKTMaZVpqnc3ucp3v+MDhN+0vkpfUuCuw/zG0wF0MXCVk9\n\ttBTO9g1I+tTbd1bGep2SHH+FaYR5uAaAgBFFNpJiREQ1JK5tmVuVGqzTEc2sC42yMEfc\n\thvsQ=="],"X-Google-DKIM-Signature":"v=1; a=rsa-sha256; c=relaxed/relaxed;\n\td=1e100.net; s=20161025;\n\th=x-gm-message-state:message-id:subject:from:to:date:in-reply-to\n\t:references:mime-version:content-transfer-encoding;\n\tbh=+CC282jKuD5rBF31hUyMAvDu3BpHHs+RQe9LxZOwWas=;\n\tb=JL8iJExi+JF1bUKm4LWl21sixlyNRvAerpUhBun/MZmy42pMh5pbdl0Q/qNtJx3t44\n\tsKy6gTFOZ8o3/nNhKY+aG+rIlW8HqBqQsxyLw03nbjjTtmPfm4XziLnRGI0j0YM005nJ\n\tgWzofE23sp4U83LE5m/jvoGKmAmypRhfMhd6POB8K93uI+6Rbkj5pJ7hyuasKMyjyUn2\n\tL8E/tfxKb7KSEcts/Io9iR8nebVuX4WvhAfviTygGptfj399xUGyPOwSCzgd5X3ho493\n\tvvQKkUQm5rfsJ0B7OzqtTMOIJgXPbr9X9rDxHhHiH20/X2q5cqB5QikGukxuATkplb3f\n\tGWfg==","X-Gm-Message-State":"AHPjjUi5bL0BadNpNv27oHuBIM/xtdF1FbsQtHbqtFjNJRMB2d+MXDKR\n\tA93m7wOkKyb4UPqlB2Q=","X-Google-Smtp-Source":"ADKCNb7LOHw/ljOzxr/41Z26vtQQaHJaqA59LxhnV97GjGxEvRWiWij6it+n64cEmgXYPyBE79DnpA==","X-Received":"by 10.84.248.1 with SMTP id p1mr1172254pll.307.1504541754196;\n\tMon, 04 Sep 2017 09:15:54 -0700 (PDT)","Message-ID":"<1504541751.3761.4.camel@gmail.com>","From":"rosenp@gmail.com","To":"lede-dev@lists.infradead.org","Date":"Mon, 04 Sep 2017 09:15:51 -0700","In-Reply-To":"<937dc2ec-0951-85af-3b35-f42956e3d360@gmail.com>","References":"<20170903183546.16749-1-rosenp@gmail.com>\n\t<1510043.ZdtWCe27Fq@debian64> <1504478774.2387.8.camel@gmail.com>\n\t<937dc2ec-0951-85af-3b35-f42956e3d360@gmail.com>","X-Mailer":"Evolution 3.24.5 (3.24.5-1.fc26) ","Mime-Version":"1.0","X-CRM114-Version":"20100106-BlameMichelson ( TRE 0.8.0 (BSD) ) MR-646709E3 ","X-CRM114-CacheID":"sfid-20170904_091616_184551_4E9CFBDD ","X-CRM114-Status":"GOOD ( 26.87 )","X-Spam-Score":"-2.0 (--)","X-Spam-Report":"SpamAssassin version 3.4.1 on bombadil.infradead.org summary:\n\tContent analysis details: (-2.0 points)\n\tpts rule name description\n\t---- ----------------------\n\t--------------------------------------------------\n\t-0.0 RCVD_IN_DNSWL_NONE RBL: Sender listed at http://www.dnswl.org/,\n\tno\n\ttrust [2607:f8b0:400e:c00:0:0:0:22f listed in] [list.dnswl.org]\n\t-0.0 SPF_PASS SPF: sender matches SPF record\n\t0.0 FREEMAIL_FROM Sender email is commonly abused enduser mail\n\tprovider (rosenp[at]gmail.com)\n\t-1.9 BAYES_00 BODY: Bayes spam probability is 0 to 1%\n\t[score: 0.0000]\n\t-0.1 DKIM_VALID Message has at least one valid DKIM or DK signature\n\t0.1 DKIM_SIGNED Message has a DKIM or DK signature,\n\tnot necessarily valid\n\t-0.1 DKIM_VALID_AU Message has a valid DKIM or DK signature from\n\tauthor's domain","Subject":"Re: [LEDE-DEV] [PATCH] ar71xx: Add GRO support to ag71xx","X-BeenThere":"lede-dev@lists.infradead.org","X-Mailman-Version":"2.1.21","Precedence":"list","List-Id":"","List-Unsubscribe":",\n\t","List-Archive":"","List-Post":"","List-Help":"","List-Subscribe":",\n\t","Content-Type":"text/plain; charset=\"utf-8\"","Content-Transfer-Encoding":"base64","Sender":"\"Lede-dev\" ","Errors-To":"lede-dev-bounces+incoming=patchwork.ozlabs.org@lists.infradead.org"}},{"id":1788437,"web_url":"http://patchwork.ozlabs.org/comment/1788437/","msgid":"<94797207-5159-5d01-22f4-8d752afac1e9@nbd.name>","list_archive_url":null,"date":"2017-10-17T14:02:42","subject":"Re: [LEDE-DEV] [PATCH] ar71xx: Add GRO support to ag71xx","submitter":{"id":8537,"url":"http://patchwork.ozlabs.org/api/people/8537/","name":"Felix Fietkau","email":"nbd@nbd.name"},"content":"On 2017-09-03 20:35, Rosen Penev wrote:\n> On a TL-WN710N, this patch increases iperf performance from ~92.5 to ~93.5 mbps.> Keep in mind the WN710N is a 100mbps device. I expect greater numbers\nfrom gigabit devices.\n> \n> Signed-off-by: Rosen Penev \nHi Rosen,\n\nSorry about that, but I will have to revert this change. It causes a\nserious regression in LAN->WAN routing performance on various devices.\nI did some digging and found out why:\nFor GRO to work properly, checksums of incoming packets have to be\nverified very early in the network stack. The Ethernet MAC does not\nsupport rx checksum offload, so this has to happen in software.\nDue to the very small cache size, this causes a significant increase in\nmemory bus traffic.\nIt might be possible in the future to avoid this by making use of the\nchecksum offload engine, but that's a separate component on the chip and\nnot present on every SoC (only the newer ones).\nIt also requires a significant rework of the Ethernet driver, which I\ndon't have any time for.\n\n- Felix","headers":{"Return-Path":"","X-Original-To":"incoming@patchwork.ozlabs.org","Delivered-To":"patchwork-incoming@bilbo.ozlabs.org","Authentication-Results":["ozlabs.org; spf=none (mailfrom)\n\tsmtp.mailfrom=lists.infradead.org (client-ip=65.50.211.133;\n\thelo=bombadil.infradead.org;\n\tenvelope-from=lede-dev-bounces+incoming=patchwork.ozlabs.org@lists.infradead.org;\n\treceiver=)","ozlabs.org; dkim=pass (2048-bit key;\n\tunprotected) header.d=lists.infradead.org\n\theader.i=@lists.infradead.org header.b=\"kGRWQcli\"; \n\tdkim=fail reason=\"signature verification failed\" (1024-bit key;\n\tunprotected) header.d=nbd.name header.i=@nbd.name header.b=\"qlWRSb90\";\n\tdkim-atps=neutral"],"Received":["from bombadil.infradead.org (bombadil.infradead.org\n\t[65.50.211.133])\n\t(using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256\n\tbits)) (No client certificate requested)\n\tby ozlabs.org (Postfix) with ESMTPS id 3yGcNZ3BkLz9sNc\n\tfor ;\n\tWed, 18 Oct 2017 01:03:24 +1100 (AEDT)","from localhost ([127.0.0.1] helo=bombadil.infradead.org)\n\tby bombadil.infradead.org with esmtp (Exim 4.87 #1 (Red Hat Linux))\n\tid 1e4SSc-0005RA-JS; Tue, 17 Oct 2017 14:03:10 +0000","from nbd.name ([2a01:4f8:131:30e2::2])\n\tby bombadil.infradead.org with esmtps (Exim 4.87 #1 (Red Hat Linux))\n\tid 1e4SSY-0005P1-RC\n\tfor lede-dev@lists.infradead.org; Tue, 17 Oct 2017 14:03:08 +0000"],"DKIM-Signature":["v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed;\n\td=lists.infradead.org; s=bombadil.20170209; h=Sender:\n\tContent-Transfer-Encoding:Content-Type:List-Subscribe:List-Help:List-Post:\n\tList-Archive:List-Unsubscribe:List-Id:Subject:In-Reply-To:MIME-Version:Date:\n\tMessage-ID:From:References:To:Reply-To:Cc:Content-ID:Content-Description:\n\tResent-Date:Resent-From:Resent-Sender:Resent-To:Resent-Cc:Resent-Message-ID:\n\tList-Owner; bh=4wKSoYYXPYgXRosH4EPJqYOSHF76/GS7M/shsaWZ5To=;\n\tb=kGRWQcliaK3IG0\n\tHqtoVwLodWWDDvkQhHAbbIg5RNBxamUQc6WAM5OhD0dz8/nYqoXgYKrtmE+kMUXF88jWO/FGhw6KE\n\tFwlmGSKQb15U5VvVSE/3G5mklG6BnMcUzREzGSUutC2KGcp1RlEHXcfH5lW6jw6x9AUVkhVembHZY\n\tOao2QkkoMG6mrOaj88coHjEMYHD0agAeY12LSqcRYSWdZywAkhDJ8IP5Nq8V3CI7NGcFQCh+HtAIH\n\t9xzHOsj2mX5XIowDx4C5SAGc3ggu9W5gIWfWzzwHJivmGGqce4vvTu0Z7EfTkadd6bhg2dmhx/JdK\n\tgDx5UZTiVgB+6E5MMMDg==;","v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=nbd.name; \n\ts=20160729;\n\th=Content-Transfer-Encoding:Content-Type:In-Reply-To:\n\tMIME-Version:Date:Message-ID:From:References:To:Subject:Sender:Reply-To:Cc:\n\tContent-ID:Content-Description:Resent-Date:Resent-From:Resent-Sender:\n\tResent-To:Resent-Cc:Resent-Message-ID:List-Id:List-Help:List-Unsubscribe:\n\tList-Subscribe:List-Post:List-Owner:List-Archive;\n\tbh=Mc/OtH3qj9gNiBNqPD1YEft08Lg12G+QLdbHLwZIAmA=;\n\tb=qlWRSb90BrcvnxGL3ckoe/Up/L\n\tsIen+VcaUNI8H0TBkUoBadZcQVGmKqqw+uJ5+unMP5CL4WHab0rZObnwDOLx1CoT9HrPr2XZvJvIj\n\tcuGWmZV2NFtiewjkNZVQ7nvOXFWtYe3ygvFfyoWCuyg56oGc6kK1xJAm/NxF7XjZRCYY=;"],"To":"Rosen Penev , lede-dev@lists.infradead.org","References":"<20170903183546.16749-1-rosenp@gmail.com>","From":"Felix Fietkau ","Message-ID":"<94797207-5159-5d01-22f4-8d752afac1e9@nbd.name>","Date":"Tue, 17 Oct 2017 16:02:42 +0200","User-Agent":"Mozilla/5.0 (Macintosh; Intel Mac OS X 10.12; rv:52.0)\n\tGecko/20100101 Thunderbird/52.4.0","MIME-Version":"1.0","In-Reply-To":"<20170903183546.16749-1-rosenp@gmail.com>","Content-Language":"en-US","X-CRM114-Version":"20100106-BlameMichelson ( TRE 0.8.0 (BSD) ) MR-646709E3 ","X-CRM114-CacheID":"sfid-20171017_070307_241922_364B206A ","X-CRM114-Status":"UNSURE ( 9.95 )","X-CRM114-Notice":"Please train this message.","X-Spam-Score":"-2.0 (--)","X-Spam-Report":"SpamAssassin version 3.4.1 on bombadil.infradead.org summary:\n\tContent analysis details: (-2.0 points)\n\tpts rule name description\n\t---- ----------------------\n\t--------------------------------------------------\n\t-0.0 RP_MATCHES_RCVD Envelope sender domain matches handover relay\n\tdomain\n\t-1.9 BAYES_00 BODY: Bayes spam probability is 0 to 1%\n\t[score: 0.0000]\n\t-0.1 DKIM_VALID Message has at least one valid DKIM or DK signature\n\t0.1 DKIM_SIGNED Message has a DKIM or DK signature,\n\tnot necessarily valid\n\t-0.1 DKIM_VALID_AU Message has a valid DKIM or DK signature from\n\tauthor's domain","Subject":"Re: [LEDE-DEV] [PATCH] ar71xx: Add GRO support to ag71xx","X-BeenThere":"lede-dev@lists.infradead.org","X-Mailman-Version":"2.1.21","Precedence":"list","List-Id":"","List-Unsubscribe":",\n\t","List-Archive":"","List-Post":"","List-Help":"","List-Subscribe":",\n\t","Content-Type":"text/plain; charset=\"us-ascii\"","Content-Transfer-Encoding":"7bit","Sender":"\"Lede-dev\" ","Errors-To":"lede-dev-bounces+incoming=patchwork.ozlabs.org@lists.infradead.org"}},{"id":1788550,"web_url":"http://patchwork.ozlabs.org/comment/1788550/","msgid":"<87po9lpzhf.fsf@nemesis.taht.net>","list_archive_url":null,"date":"2017-10-17T15:26:20","subject":"Re: [LEDE-DEV] [PATCH] ar71xx: Add GRO support to ag71xx","submitter":{"id":70796,"url":"http://patchwork.ozlabs.org/api/people/70796/","name":"Dave Taht","email":"dave@taht.net"},"content":"Felix Fietkau writes:\n\n> On 2017-09-03 20:35, Rosen Penev wrote:\n>> On a TL-WN710N, this patch increases iperf performance from ~92.5 to ~93.5\n>> mbps.> Keep in mind the WN710N is a 100mbps device. I expect greater numbers\n> from gigabit devices.\n>> \n>> Signed-off-by: Rosen Penev \n> Hi Rosen,\n>\n> Sorry about that, but I will have to revert this change. It causes a\n> serious regression in LAN->WAN routing performance on various devices.\n> I did some digging and found out why:\n> For GRO to work properly, checksums of incoming packets have to be\n> verified very early in the network stack. The Ethernet MAC does not\n> support rx checksum offload, so this has to happen in software.\n> Due to the very small cache size, this causes a significant increase in\n> memory bus traffic.\n\nAlso, for the record, if there is a need to manage the WAN side to lower\nspeeds (say, below 40Mbit/s) via sqm, GRO bulking up a microburst into a\nsuperpacket mandates sch_cake (rather than fq_codel) to peel it apart\nagain to hold latencies low there.\n\nThere are a lot of devices that do GRO that perhaps shouldn't. mvneta\nhas very agressive soft-GRO, in particular. Shipping out one 64k\nsuperpacket takes half a second at 1mbit.\n\n> It might be possible in the future to avoid this by making use of the\n> checksum offload engine, but that's a separate component on the chip and\n> not present on every SoC (only the newer ones).\n> It also requires a significant rework of the Ethernet driver, which I\n> don't have any time for.\n>\n> - Felix\n>\n> _______________________________________________\n> Lede-dev mailing list\n> Lede-dev@lists.infradead.org\n> http://lists.infradead.org/mailman/listinfo/lede-dev","headers":{"Return-Path":"","X-Original-To":"incoming@patchwork.ozlabs.org","Delivered-To":"patchwork-incoming@bilbo.ozlabs.org","Authentication-Results":["ozlabs.org; spf=none (mailfrom)\n\tsmtp.mailfrom=lists.infradead.org (client-ip=65.50.211.133;\n\thelo=bombadil.infradead.org;\n\tenvelope-from=lede-dev-bounces+incoming=patchwork.ozlabs.org@lists.infradead.org;\n\treceiver=)","ozlabs.org; dkim=pass (2048-bit key;\n\tunprotected) header.d=lists.infradead.org\n\theader.i=@lists.infradead.org header.b=\"eg3IfS+g\"; \n\tdkim-atps=neutral"],"Received":["from bombadil.infradead.org (bombadil.infradead.org\n\t[65.50.211.133])\n\t(using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256\n\tbits)) (No client certificate requested)\n\tby ozlabs.org (Postfix) with ESMTPS id 3yGfFT3kFWz9t2S\n\tfor ;\n\tWed, 18 Oct 2017 02:27:25 +1100 (AEDT)","from localhost ([127.0.0.1] helo=bombadil.infradead.org)\n\tby bombadil.infradead.org with esmtp (Exim 4.87 #1 (Red Hat Linux))\n\tid 1e4Tlm-0003nm-FO; Tue, 17 Oct 2017 15:27:02 +0000","from mail.taht.net ([2a01:7e00::f03c:91ff:feae:7028])\n\tby bombadil.infradead.org with esmtps (Exim 4.87 #1 (Red Hat Linux))\n\tid 1e4TlU-0003Wp-UT\n\tfor lede-dev@lists.infradead.org; Tue, 17 Oct 2017 15:26:47 +0000","from nemesis.taht.net (unknown\n\t[IPv6:2603:3024:1536:86f0:2e0:4cff:fec1:1206])\n\t(using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128\n\tbits)) (No client certificate requested)\n\tby mail.taht.net (Postfix) with ESMTPSA id 3A51C21425;\n\tTue, 17 Oct 2017 15:26:22 +0000 (UTC)"],"DKIM-Signature":"v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed;\n\td=lists.infradead.org; s=bombadil.20170209; h=Sender:\n\tContent-Transfer-Encoding:Content-Type:Cc:List-Subscribe:List-Help:List-Post:\n\tList-Archive:List-Unsubscribe:List-Id:Subject:MIME-Version:Message-ID:\n\tIn-Reply-To:Date:References:To:From:Reply-To:Content-ID:Content-Description:\n\tResent-Date:Resent-From:Resent-Sender:Resent-To:Resent-Cc:Resent-Message-ID:\n\tList-Owner; bh=Q8WndO16SoESNE4KgvPyLremH4ttluZJ1uzmC5DE+jU=;\n\tb=eg3IfS+gRL3b1r\n\t/h8st+k9x63PAqwOFUc5V3BZ5tWDqM7zRWFy8eSSt3FLZoQn+p/kSYoKjJhnJGZ2IIjYlAwiHI/Mt\n\tLb2bXtSv3HyHarUrPRQ28PbDnc+hhKgxf53q4b3h+2JGeVH3Iib4jcRLv8oQbfxZ0UWhI301F40wZ\n\t6w/73FgN3uNDR2zPaI8oDb6DRc3s78HD+t9rE138Rpsxvmr9OujAlJv1SnzYR1YdrRLwsx3B+l6s9\n\trImcKGLP9I979myQxl1Vsystqaj5+hGuIvFCrmIutzEHLC7IvLh8KaFB8iXk9AWpXINSEdhWuRgmY\n\tdFElQEECsTujUw8XG2kg==;","From":"Dave Taht ","To":"Felix Fietkau ","References":"<20170903183546.16749-1-rosenp@gmail.com>\n\t<94797207-5159-5d01-22f4-8d752afac1e9@nbd.name>","Date":"Tue, 17 Oct 2017 08:26:20 -0700","In-Reply-To":"<94797207-5159-5d01-22f4-8d752afac1e9@nbd.name> (Felix Fietkau's\n\tmessage of \"Tue, 17 Oct 2017 16:02:42 +0200\")","Message-ID":"<87po9lpzhf.fsf@nemesis.taht.net>","User-Agent":"Gnus/5.13 (Gnus v5.13) Emacs/24.5 (gnu/linux)","MIME-Version":"1.0","X-CRM114-Version":"20100106-BlameMichelson ( TRE 0.8.0 (BSD) ) MR-646709E3 ","X-CRM114-CacheID":"sfid-20171017_082645_138824_DC1E0EB0 ","X-CRM114-Status":"GOOD ( 16.54 )","X-Spam-Score":"-1.9 (-)","X-Spam-Report":"SpamAssassin version 3.4.1 on bombadil.infradead.org summary:\n\tContent analysis details: (-1.9 points)\n\tpts rule name description\n\t---- ----------------------\n\t--------------------------------------------------\n\t-0.0 SPF_PASS SPF: sender matches SPF record\n\t-0.0 RP_MATCHES_RCVD Envelope sender domain matches handover relay\n\tdomain\n\t-1.9 BAYES_00 BODY: Bayes spam probability is 0 to 1%\n\t[score: 0.0000]","Subject":"Re: [LEDE-DEV] [PATCH] ar71xx: Add GRO support to ag71xx","X-BeenThere":"lede-dev@lists.infradead.org","X-Mailman-Version":"2.1.21","Precedence":"list","List-Id":"","List-Unsubscribe":",\n\t","List-Archive":"","List-Post":"","List-Help":"","List-Subscribe":",\n\t","Cc":"lede-dev@lists.infradead.org, Rosen Penev ","Content-Type":"text/plain; charset=\"us-ascii\"","Content-Transfer-Encoding":"7bit","Sender":"\"Lede-dev\" ","Errors-To":"lede-dev-bounces+incoming=patchwork.ozlabs.org@lists.infradead.org"}}]