From owner-freebsd-net@FreeBSD.ORG Tue Jun 2 03:37:52 2009 Return-Path: Delivered-To: freebsd-net@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 8D7B0106566B for ; Tue, 2 Jun 2009 03:37:52 +0000 (UTC) (envelope-from cmb@pfsense.org) Received: from mail.pfsense.org (mail.pfsense.org [69.64.6.29]) by mx1.freebsd.org (Postfix) with ESMTP id 6099F8FC12 for ; Tue, 2 Jun 2009 03:37:52 +0000 (UTC) (envelope-from cmb@pfsense.org) Received: from localhost (localhost.localdomain [127.0.0.1]) by mail.pfsense.org (Postfix) with ESMTP id B849323E17 for ; Mon, 1 Jun 2009 22:19:56 -0500 (EST) X-Virus-Scanned: amavisd-new at mail.pfsense.org Received: from mail.pfsense.org ([127.0.0.1]) by localhost (localhost [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id MaErR6+1Zrjh for ; Mon, 1 Jun 2009 22:19:55 -0500 (EST) Received: from [10.0.64.15] (96-28-38-25.dhcp.insightbb.com [96.28.38.25]) by mail.pfsense.org (Postfix) with ESMTP id E372223DB8 for ; Mon, 1 Jun 2009 22:19:54 -0500 (EST) Message-ID: <4A249A57.1070900@pfsense.org> Date: Mon, 01 Jun 2009 23:19:51 -0400 From: Chris Buechler User-Agent: Thunderbird 2.0.0.21 (Windows/20090302) MIME-Version: 1.0 To: freebsd-net@freebsd.org Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Subject: ath(4) randomly changes MTU to 2290 after explicitly set to 1500 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: Tue, 02 Jun 2009 03:37:52 -0000 In FreeBSD 7.1 using this patch: http://people.freebsd.org/~sam/ath_hal-releng7.patch and 7.2 with stock ath(4) (the above does not cleanly apply to 7.2), there are numerous pfsense users seeing problems with ath when bridging. This did not occur in 7.0. Upon investigation of a few systems, though we explicitly configure the interface with MTU 1500 twice (once when bringing up the ath interface, and again when setting up the bridge interface), somehow it ends up switching MTU to 2290 when we never configure it as such. This breaks bridging to an Ethernet interface because if_bridge requires the MTU to be the same on both interfaces. None of the commands we're running to setup ath or the bridge will replicate it, no matter what I do to the interface it stays to 1500 unless it sits there for a while ("a while" might be hours or days). There isn't anything that runs in the background to touch interfaces, a completely idle untouched box will change its configuration. It is setup fine and works initially, but given time, it'll switch to 2290 and stop working. example bridge setup commands: /sbin/ifconfig bridge0 destroy /sbin/ifconfig bridge0 create /sbin/ifconfig ath0 mtu 1500 /sbin/ifconfig vr1 mtu 1500 /sbin/ifconfig ath0 up /sbin/ifconfig vr1 up /sbin/ifconfig bridge0 addm ath0 addm vr1 up example ath setup commands: /sbin/ifconfig ath0 down /sbin/ifconfig ath0 mode '11g' /sbin/ifconfig ath0 channel any /sbin/ifconfig ath0 -mediaopt turbo /sbin/ifconfig ath0 ssid 'cmb' /sbin/ifconfig ath0 -hidessid /sbin/ifconfig ath0 -mediaopt adhoc /sbin/ifconfig ath0 protmode 'off' /sbin/ifconfig ath0 -pureg /sbin/ifconfig ath0 apbridge /sbin/ifconfig ath0 -wme /sbin/ifconfig ath0 authmode open wepmode off /sbin/ifconfig ath0 txpower '99' /sbin/ifconfig ath0 mediaopt hostap /sbin/ifconfig ath0 mtu 1500 /sbin/ifconfig ath0 up /usr/sbin/hostapd -B /var/etc/hostapd_ath0.conf $ dmesg|grep ath ath0: mem 0xe00c0000-0xe00cffff irq 9 at device 12.0 on pci0 ath0: [ITHREAD] ath0: WARNING: using obsoleted if_watchdog interface ath0: Ethernet address: 00:0b:6b:84:3d:7c ath0: mac 5.9 phy 4.3 radio 3.6 ath0: promiscuous mode enabled another that can replicate it is: ath0: mem 0x80000000-0x8000ffff irq 12 at device 13.0 on pci0 $ sysctl -a hw.ath hw.ath.txbuf: 200 hw.ath.rxbuf: 40 hw.ath.regdomain: 0 hw.ath.countrycode: 0 hw.ath.xchanmode: 1 hw.ath.outdoor: 1 hw.ath.calibrate: 30 hw.ath.hal.swba_backoff: 0 hw.ath.hal.sw_brt: 10 hw.ath.hal.dma_brt: 2 I have access to several systems that can replicate this, please let me know if any further information would be helpful. thanks, Chris