From mboxrd@z Thu Jan 1 00:00:00 1970 From: David Jander Subject: Re: BUG: fsl FEC ethernet tx checksum offloading doesn't work with RMII interface Date: Wed, 28 Oct 2015 12:19:18 +0100 Message-ID: <20151028121918.4884b42e@archvile> References: <20151028104854.025430ce@archvile> <20151028111414.GJ8644@n2100.arm.linux.org.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Cc: netdev@vger.kernel.org, linux-arm-kernel@lists.infradead.org To: Russell King - ARM Linux Return-path: Received: from protonic.xs4all.nl ([83.163.252.89]:27330 "EHLO protonic.xs4all.nl" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754999AbbJ1LTT (ORCPT ); Wed, 28 Oct 2015 07:19:19 -0400 In-Reply-To: <20151028111414.GJ8644@n2100.arm.linux.org.uk> Sender: netdev-owner@vger.kernel.org List-ID: On Wed, 28 Oct 2015 11:14:14 +0000 Russell King - ARM Linux wrote: > On Wed, Oct 28, 2015 at 10:48:54AM +0100, David Jander wrote: > > > > Hi all, > > > > I was unable to figure out who's maintaining > > drivers/net/ethernet/freescale/fec_main.c, so I hope someone can help out > > on this list... > > > > We have a board with a RMII phy connected to an i.MX6S. The hardware seems > > to be ok, since I can receive and transmit ethernet frames without drops or > > errors. However only simple things like ping and dhcp seemed to work. > > TCP/IP connections could not be made. When looking at both ends with > > tcpdump, I realized that all transmitted packages arrived at the other end > > with the TCP and IP header checksums zeroed-out. > > > > After issuing the following command, TCP/IP started working correctly: > > > > $ ethtool --offload eth0 tx off > > > > This works around the issue. For some reason, when the FEC is in RMII > > mode, it isn't filling in the checksums. > > > > On another board with an RGMII phy the same kernel works fine without the > > need to disable offloading. What can possibly relate this functionality to > > the choice of MAC interface? > > You don't mention which kernel version you're using. There has been a bug > here with older kernels... Sorry, I somehow assumed it was obvious I'd report against latest mainline... I'm on 4.3-rc7. Best regards, -- David Jander Protonic Holland.