From owner-freebsd-bugs@FreeBSD.ORG Wed Sep 16 02:29:52 2009 Return-Path: Delivered-To: freebsd-bugs@FreeBSD.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id BCA10106568F; Wed, 16 Sep 2009 02:29:52 +0000 (UTC) (envelope-from dokas@oitsec.umn.edu) Received: from mail.oitsec.umn.edu (mail.oitsec.umn.edu [128.101.238.120]) by mx1.freebsd.org (Postfix) with ESMTP id 919928FC15; Wed, 16 Sep 2009 02:29:52 +0000 (UTC) Received: from yog.oitsec.umn.edu (mrtoad.oitsec.umn.edu [128.101.238.97]) (Authenticated sender: dokas) by mail.oitsec.umn.edu (Postfix) with ESMTPSA id 36BC16D453; Tue, 15 Sep 2009 21:15:00 -0500 (CDT) Message-ID: <4AB04A25.501@oitsec.umn.edu> Date: Tue, 15 Sep 2009 21:15:01 -0500 From: Paul Dokas Organization: OIT Security & Assurance, University of Minnesota User-Agent: Thunderbird 2.0.0.23 (X11/20090823) MIME-Version: 1.0 To: linimon@FreeBSD.org References: <200909120053.n8C0rcbZ005357@freefall.freebsd.org> In-Reply-To: <200909120053.n8C0rcbZ005357@freefall.freebsd.org> X-Enigmail-Version: 0.95.6 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Cc: freebsd-net@FreeBSD.org, freebsd-bugs@FreeBSD.org Subject: Re: kern/138739: [wpi] wpi(4) does not work very well under 8.0-BETA4 X-BeenThere: freebsd-bugs@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list Reply-To: Paul Dokas List-Id: Bug reports List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 16 Sep 2009 02:29:52 -0000 linimon@FreeBSD.org wrote: > Old Synopsis: wpi(4) does not work very well under 8.0-BETA4 > New Synopsis: [wpi] wpi(4) does not work very well under 8.0-BETA4 > > Responsible-Changed-From-To: freebsd-bugs->freebsd-net > Responsible-Changed-By: linimon > Responsible-Changed-When: Sat Sep 12 00:53:04 UTC 2009 > Responsible-Changed-Why: > Over to maintainer(s). > > http://www.freebsd.org/cgi/query-pr.cgi?pr=138739 I was just looking for possible sol'ns to a very similar problem that I have with my FreeBSD 7.2 laptop, a Dell D820 with a 3945abg card: wpi0: mem 0xecfff000-0xecffffff irq 17 at device 0.0 on pci12 I've had what looks like exactly this same problem for ages. Back into the FreeBSD 6.X days when the wpi driver was first added. Also, I think that PR 127102 is probably related. I've noticed problems when I'm in an area with lots of SSIDs and APs. I get this particularly bad when at work where there are 5 SSIDs present on each AP and at least 4+ APs always within range. At home where I see only 3 SSIDs on 3 APs, I can associate with my Linksys "router" just fine, but I see the slow throughput problem. This might be because I'm only associated at 1Mbps: wpi0: flags=8843 metric 0 mtu 1500 inet 172.16.23.65 netmask 0xffffff80 broadcast 172.16.23.127 media: IEEE 802.11 Wireless Ethernet autoselect (DS/1Mbps) status: associated ssid SSID channel 6 (2437 Mhz 11g) bssid 00:21:19:ae:00:00 authmode WPA privacy ON deftxkey UNDEF TKIP 2:128-bit TKIP 3:128-bit txpower 50 bmiss 7 scanvalid 60 protmode CTS roaming MANUAL Anyway, at work today, I did a bunch of testing. I attempted to associate first with a WPA2-enterprise SSID and then with an open SSID. In both cases, my laptop went into a spiral of doom: associate, spew errors, disassociate, repeat: Sep 15 15:21:13 yog kernel: wpi0: link state changed to UP Sep 15 15:21:13 yog kernel: wpi0: fatal firmware error Sep 15 15:21:13 yog kernel: wpi0: timeout resetting Rx ring Sep 15 15:21:13 yog kernel: wpi0: link state changed to DOWN Sep 15 15:21:17 yog kernel: wpi0: link state changed to UP Sep 15 15:21:17 yog kernel: wpi0: fatal firmware error Sep 15 15:21:17 yog kernel: wpi0: timeout resetting Rx ring Sep 15 15:21:17 yog kernel: wpi0: link state changed to DOWN ... A network engineer pulled the logs from the network gear (Trapeze ABGN APs) and saw that my laptop was repeatedly roaming between all of the available APs. I strongly suspect that there's a bug lurking in wpi firmware related to roaming. A possible pointer for this problem might be here: http://www.openbsd.org/plus41.html "Fix firmware fatal errors on re-associations in wpi(4)." I'm willing to provide more data and test any possible fixes for 7.2, but I'm lacking the time and experience necessary to track this one down myself. Paul -- Paul Dokas dokas at oitsec.umn.edu ====================================================================== Don Juan Matus: "an enigma wrapped in mystery wrapped in a tortilla."