From mboxrd@z Thu Jan 1 00:00:00 1970 From: Michal Schmidt Subject: Re: [PATCH 5/7] bnx2x: do not set TPA flags and features in bnx2x_init_bp Date: Tue, 30 Aug 2011 19:15:21 +0200 Message-ID: <4E5D1AA9.2010702@redhat.com> References: <1314714646-3642-1-git-send-email-mschmidt@redhat.com> <1314714646-3642-6-git-send-email-mschmidt@redhat.com> <201108301921.22838.vladz@broadcom.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Cc: "netdev@vger.kernel.org" , Dmitry Kravkov , Eilon Greenstein To: Vlad Zolotarov Return-path: Received: from mx1.redhat.com ([209.132.183.28]:32688 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754698Ab1H3RPY (ORCPT ); Tue, 30 Aug 2011 13:15:24 -0400 In-Reply-To: <201108301921.22838.vladz@broadcom.com> Sender: netdev-owner@vger.kernel.org List-ID: On 08/30/2011 06:21 PM, Vlad Zolotarov wrote: > NACK > > This patch will cause the bnx2x to initialize HW with LRO disabled on the > first ifup because our code considers the TPA_ENABLE_FLAG when desiding on > whether LRO is enabled or not. ethtool would still report the LRO on though! I see. I thought register_netdevice() would always call bnx2x_set_features(), but this is not the case. Patch 7/7 has a similar problem with RX_VLAN_STRIP_FLAG then. Michal