From owner-freebsd-current@FreeBSD.ORG Thu Aug 31 15:58:23 2006 Return-Path: X-Original-To: current@freebsd.org Delivered-To: freebsd-current@FreeBSD.ORG Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 595AF16A4DA for ; Thu, 31 Aug 2006 15:58:23 +0000 (UTC) (envelope-from sam@errno.com) Received: from ebb.errno.com (ebb.errno.com [69.12.149.25]) by mx1.FreeBSD.org (Postfix) with ESMTP id 49FEF43D62 for ; Thu, 31 Aug 2006 15:58:20 +0000 (GMT) (envelope-from sam@errno.com) Received: from [10.0.0.248] (trouble.errno.com [10.0.0.248]) (authenticated bits=0) by ebb.errno.com (8.13.6/8.12.6) with ESMTP id k7VFwHhK024424 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO); Thu, 31 Aug 2006 08:58:19 -0700 (PDT) (envelope-from sam@errno.com) Message-ID: <44F70719.70501@errno.com> Date: Thu, 31 Aug 2006 08:58:17 -0700 From: Sam Leffler User-Agent: Thunderbird 1.5.0.4 (X11/20060724) MIME-Version: 1.0 To: dandee@volny.cz References: <005801c6cab0$d2c8aee0$6508280a@tocnet28.jspoj.czf> In-Reply-To: <005801c6cab0$d2c8aee0$6508280a@tocnet28.jspoj.czf> X-Enigmail-Version: 0.94.0.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit Cc: current@freebsd.org Subject: Re: athstats and flush stats 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: Thu, 31 Aug 2006 15:58:23 -0000 Daniel Dvořák wrote: > Hello Sam and all, > > could you tell me how to reset or flush the stats ? > > I use athstats on ahdemo links to watch OFDM timing errors as you have told > me last time, but it seems it is cumulative stats and iface down and up, it > does not help. > > How-to fulsh except for reboot the computer ? There is no way (currently) to clear the kernel stat counters. It'd be easy to add but I personally hate to lose data and would prefer that reporting stats relative to an event or time be done by user code (e.g. by recording the stat block and showing differences). Another option might be keeping 2 stat blocks in the kernel and adding an ioctl that snap's the state of the stats so user apps can report data relative to the snapshot. This issue is really kinda generic (i.e. relevant to all/many statistics kept in the kernel). Sam