From owner-freebsd-current@FreeBSD.ORG Tue Sep 29 06:43:34 2009 Return-Path: Delivered-To: freebsd-current@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 9B5FF106566B for ; Tue, 29 Sep 2009 06:43:34 +0000 (UTC) (envelope-from ale@FreeBSD.org) Received: from lab.alexdupre.com (cl-506.trn-01.it.sixxs.net [IPv6:2001:1418:100:1f9::2]) by mx1.freebsd.org (Postfix) with ESMTP id CB4E38FC12 for ; Tue, 29 Sep 2009 06:43:33 +0000 (UTC) Received: (qmail 79816 invoked from network); 29 Sep 2009 06:43:32 -0000 Received: from athlon.alexdupre.com (192.168.178.2) by lab.alexdupre.com with SMTP; 29 Sep 2009 06:43:32 -0000 Message-ID: <4AC1ABE5.6090409@FreeBSD.org> Date: Tue, 29 Sep 2009 08:40:37 +0200 From: Alex Dupre User-Agent: Thunderbird 2.0.0.19 (X11/20090112) MIME-Version: 1.0 To: Sam Leffler References: <4AC0BB4A.40201@FreeBSD.org> <4AC17F33.80800@freebsd.org> In-Reply-To: <4AC17F33.80800@freebsd.org> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Cc: freebsd-current@freebsd.org Subject: Re: Atheros 9280 issue X-BeenThere: freebsd-current@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Discussions about the use of FreeBSD-current List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 29 Sep 2009 06:43:34 -0000 Sam Leffler wrote: >> Clients can connect to the AP, but obviously packets are dropped every >> time the card is reset. > > How do you know packets are dropped? Nothing should be lost. I'm pinging the access point from an associated wifi client: when this message appears one or two pings are lost. Previously, when such messages didn't appear, no pings were lost. Maybe the reply timeouts and is not completely lost, I'll do more tests asap. > "every few seconds" is not helpful. Actually the timing is highly variable, it may happen once a second, like once every 15 seconds or more. Usually (in the two hours I tried) every 2-5 seconds. > Look for something else in the > system that coincides w/ your msgs. It's possible that periodic > calibration done by the driver has a bad effect; you can turn it off by > setting a sysctl (look at sysctl -a and/or the driver code). hw.ath.shortcal = 0, right? I'll try, I haven't done additional tests before knowing what would be useful for debugging. -- Alex Dupre