From owner-freebsd-net@FreeBSD.ORG Thu May 17 05:50:11 2007 Return-Path: X-Original-To: freebsd-net@freebsd.org Delivered-To: freebsd-net@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [69.147.83.52]) by hub.freebsd.org (Postfix) with ESMTP id D2A9B16A400 for ; Thu, 17 May 2007 05:50:11 +0000 (UTC) (envelope-from rea-fbsd@codelabs.ru) Received: from pobox.codelabs.ru (pobox.codelabs.ru [144.206.177.45]) by mx1.freebsd.org (Postfix) with ESMTP id 897B213C459 for ; Thu, 17 May 2007 05:50:11 +0000 (UTC) (envelope-from rea-fbsd@codelabs.ru) DomainKey-Signature: a=rsa-sha1; q=dns; c=simple; s=one; d=codelabs.ru; h=Received:Date:From:To:Cc:Message-ID:MIME-Version:Content-Type:Content-Disposition:Sender:X-Spam-Status:Subject; b=nZTf8oXo4pmrBPMk2FBMN30z63lYLBG3JdAD2PFHX/M8+Ozjuevjoyb5Igy8xKniCj6RFN6KBdI+H01ya3j7NTQ9z8QXxQ1Zzbb4iB9O5CWVPudNgR5/pgpsbQ5Wi7comP6tFQwx5gfTEYcydq9Y5YqIrXwboiefS+z2hRypHec=; Received: from void.codelabs.ru (void.codelabs.ru [144.206.177.25]) by pobox.codelabs.ru with esmtpsa (TLSv1:AES256-SHA:256) id 1HoYsJ-0001X4-VZ; Thu, 17 May 2007 09:50:08 +0400 Date: Thu, 17 May 2007 09:50:03 +0400 From: Eygene Ryabinkin To: jfv@freebsd.org Message-ID: <20070517055003.GB1325@void.codelabs.ru> MIME-Version: 1.0 Content-Type: text/plain; charset=koi8-r Content-Disposition: inline Sender: rea-fbsd@codelabs.ru X-Spam-Status: No, score=-4.4 required=4.0 tests=ALL_TRUSTED,BAYES_00 Cc: freebsd-net@freebsd.org Subject: Port of the new em(4) to RELENG_6 and some problems with 82540EM X-BeenThere: freebsd-net@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Networking and TCP/IP with FreeBSD List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 17 May 2007 05:50:11 -0000 Jack, good day. I happened to make a quick and dirty patch for your latest (for the 16.05.2007) em(4) driver from 7-CURRENT. I had seen that you mentioned TSO and 6.3, so I assume that you're going to merge the driver to the RELENG_6 someday, so maybe my effort may be helpful. To my problems. I have the '82540EM Gigabit Ethernet Controller' branded card (PWLA8390MT) and I am currently expiriencing troubles with the Gigabit switch (D-Link DGS-1008, the cheap one). Plugged into that switch I have - broken packets and SSH messages 'Disconnecting: Corrupted MAC on input.' when checksumming is turned on and gigabit speed, - abyssmal rates for the HTTP downloads (about 10Mbit/sec) when checksums are turned off and the speed is gigabit again, - having switched the media on the adapter to the 100BaseTX full-duplex, but still using the same switch the speed drops to 200Kbit/sec, - using another D-Link switch (sorry, have no other brand in hand) the above situation is completely reproducible. - working in the half-duplex mode (for 10BaseT/UTP and 100BaseTX) I have no problems to reach the 90% of the theoretical bandwidth of the link via HTTP either with checksums enabled or disabled. So, the full-duplex mode is not usable to me now. If someone can help in this situation, I am all ears. I thought that new driver may introduce some fixes, but it isn't: the situation is the same. I will continue my investigations, but just in case you're interested in my port, it is available from http://codelabs.ru/patches/fbsd/em-6.5.0-initial.patch.bz2 RELENG_6 and 7-CURRENT sources were CVSupped yesterday, 16 May 2007. TSO-related stuff is completely disabled for the RELENG_6. My limited testing of the driver revealed no problems yet, but if they will arise I will report them. -- Eygene