From owner-freebsd-sparc64@FreeBSD.ORG Thu Sep 4 06:50:44 2003 Return-Path: Delivered-To: freebsd-sparc64@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 6433116A4BF for ; Thu, 4 Sep 2003 06:50:44 -0700 (PDT) Received: from geddar.km.ua (geddar.km.ua [62.149.0.130]) by mx1.FreeBSD.org (Postfix) with ESMTP id AF0D543FB1 for ; Thu, 4 Sep 2003 06:50:40 -0700 (PDT) (envelope-from maxim@geddar.km.ua) Received: from geddar.km.ua (localhost [127.0.0.1]) by geddar.km.ua (8.12.9/8.12.9) with ESMTP id h84DoaUu077584; Thu, 4 Sep 2003 16:50:36 +0300 (EEST) (envelope-from maxim@geddar.km.ua) Received: (from maxim@localhost) by geddar.km.ua (8.12.9/8.12.9/Submit) id h84DoaaN077583; Thu, 4 Sep 2003 16:50:36 +0300 (EEST) Date: Thu, 4 Sep 2003 16:50:36 +0300 From: Maxim Mazurok To: Thomas Moestl Message-ID: <20030904135036.GW8966@km.ua> References: <20030823180012.GA668@crow.dom2ip.de> <20030828100032.GI8966@km.ua> <20030828104212.GL8966@km.ua> <20030828155927.GA707@timesink.dyndns.org> <20030828174715.GR8966@km.ua> <20030903170000.GB20825@timesink.dyndns.org> <20030903194133.GN8966@km.ua> <20030903212129.GA21806@timesink.dyndns.org> <20030903225314.GS8966@km.ua> <20030904130426.GB5484@timesink.dyndns.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20030904130426.GB5484@timesink.dyndns.org> User-Agent: Mutt/1.5.4i cc: freebsd-sparc@freebsd.org Subject: Re: sio(4) driver X-BeenThere: freebsd-sparc64@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Porting FreeBSD to the Sparc List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 04 Sep 2003 13:50:44 -0000 On Thu, Sep 04, 2003 at 03:04:26PM +0200, Thomas Moestl wrote: >> >> >> >That's very strange. So you cannot see them in tcpdump if you do not >> >> >> >put the interface in promiscuous mode (by using the -p argument to >> >> >> >tcpdump)? Can you receive broadcast packets (e.g. a broadcast ping) at >> >> >> >all? >> >> >> >> >> >> so... experiment: >> >> >> [...] >> >> > >> >> >Hmmm. Can you please repeat the experiment and send a tcpdump, this >> >> >time in promiscuous mode (i.e. without the -p option), if possible >> >> >from both the e250 and the cisco? >> >> >> >> i not have e250 :( >> > >> >Oh, sorry, mixed that up. >> > >> >> i have Ultra AXi OEM motherboard..... >> >> tcpdump on cisco - it's problem. i can run packet debugger.... >> > >> >Hmmm, the tcpdump from the sparc64 box might do for now. >> > >> >> so, experiment two: >> >> >> >> root@fang:~#ifconfig rl3 >> >> rl3: flags=8843 mtu 1500 >> >> options=8 >> >> inet 193.201.116.249 netmask 0xfffffffc broadcast 193.201.116.251 >> >> ether 00:30:4f:21:bc:91 >> >> media: Ethernet 10baseT/UTP >> >> status: active >> >> root@fang:~#tcpdump -n -i rl3 >> > >> >Please include -e in the tcpdump flags; sorry for forgetting to >> >mention that. >> >> root@fang:~#tcpdump -e -n -i rl3 >> tcpdump: listening on rl3 >> 01:47:28.400964 0:0:c:47:8a:c1 ff:ff:ff:ff:ff:ff 0800 114: 193.201.116.250 > 255.255.255.255: icmp: echo request >> 01:47:29.796714 0:0:c:47:8a:c1 1:0:c:cc:cc:cc 0125 307: CDP v2, ttl=180s DevID 'ap-ec' Addr (1): IPv4 193.201.116.250 PortID 'Ethernet0' CAP 0x01[|cdp] >> 01:47:30.398474 0:0:c:47:8a:c1 ff:ff:ff:ff:ff:ff 0800 114: 193.201.116.250 > 255.255.255.255: icmp: echo request >> 01:47:32.398607 0:0:c:47:8a:c1 ff:ff:ff:ff:ff:ff 0800 114: 193.201.116.250 > 255.255.255.255: icmp: echo request >> 01:47:34.398588 0:0:c:47:8a:c1 ff:ff:ff:ff:ff:ff 0800 114: 193.201.116.250 > 255.255.255.255: icmp: echo request >> 01:47:36.401175 0:0:c:47:8a:c1 ff:ff:ff:ff:ff:ff 0800 114: 193.201.116.250 > 255.255.255.255: icmp: echo request > >OK, so arp is working fine; the real problem is that the station address >is not programmed correctly even with the patch that I sent you. The >issue is that it has do be programmed with 32-bit accesses - byte-wise >accesses, like my patch introduces, are seemingly not allowed. They >used to "work" before because we did not make the configuration >registers writable, so the byte writes would not change anything and >the address preloaded from the card ROM would be used. > >I've attached a patch that uses stream accesses to program the address >instead. I've borrowed myself a rl card, and it seems to work fine >now (in fact I'm writing this over it). my sources now from *default release=cvs tag=. date=2003.08.22.00.00.00 patch to my if_rl.c is rejected :( patch to my if_rlreg.h is applied. i need cvsup to latest date? -- Maxim Mazurok (MMP2-RIPE)