Date: Wed, 31 Mar 1999 11:29:23 +0200 From: Stefan Bethke <stefan@promo.de> To: "Alex Sel'kov" <as@shome.eu.org> Cc: freebsd-net@FreeBSD.ORG Subject: RE: 2 cards in one collision domain Message-ID: <327978.3131868563@d225.promo.de> In-Reply-To: <NCBBIKHEGKHONNFGOHCBCECACHAA.as@shome.eu.org>
next in thread | previous in thread | raw e-mail | index | archive | help
Alex Sel'kov <as@shome.eu.org> wrote: >> Did you add "options BRIDGE" to your kernel config file? >> >> I'm now somewhat confused. I *think* that you'd normally >> shouldn't see this >> message, unless you enabled bridging. >> > > Yes, I did. But net.link.ether.bridge: 0 is not enough to disable bridging > at all? No. In some parts of the kernel, some code is ommited or included specifically with options BRIDGE, for example, some ARP code in src/sys/netinet/if_ether.c. >> However, why would you want to enable bridging between two interfaces >> connected to the same net? > No, I don't. But this is only one of possible configurations. I must to > play with various scenario, with and without bridging. In this current > case I don't want any bridging functionality. > > Anyway, I can simply use a different kernel versions in different > scenario. That would seem resonable. >> Why have you connected both interfaces to the same net? If you want your >> host to be on two subnets on the same wire, simply configure both IPs >> (see /etc/rc.conf for details). > > Yes, I want have two connections to large shared media (large > bridged/switched net on logical link layer). I can use aliases, but this > is not exact what I need. because I can have different rules of game at > MAC layer and MAC address can make difference. > >> Hhm. Can you compile /sys/pci/if_de.c with TULIP_DEBUG enabled? And >> give us some more information on the make and model of the cards your >> using? >> > > de0: > > Vendor: Nexo (http://www.nexo.com.tw). Name - NEC-100D. Chip - INTEL > 21143-PD. > Transceiver - DM9101F. > > de1: > > Vendor: unknown. Name: unknown. Chip - Digital 21143-PC. Transceiver - > Kendin KS8761. > > dmesg: [ reordered for clarity ] > de0: <Digital 21143 Fast Ethernet> rev 0x41 int a irq 16 on pci0.19.0 > de0: 21143 [10-100Mb/s] pass 4.1 (invalid EESPROM checksum) > de0: address 00:00:1c:b0:d9:37 > de0: timeout: probing 100baseTX > de0: enabling 100baseTX port > de0: timeout: probing 100baseTX > de1: <Digital 21143 Fast Ethernet> rev 0x30 int a irq 16 on pci2.4.0 > de1: 21143 [10-100Mb/s] pass 3.0 > de1: address 00:c0:ca:11:78:ed > de1: timeout: probing 100baseTX > de1: timeout: probing 10baseT > de1: enabling 10baseT port > de1: timeout: probing 100baseTX > de1: timeout: probing 10baseT > de1: enabling 10baseT port It seems to me that the auto-sensing of your hub (a dual-speed hub?) and the cards don't get together well. The safest way in this case is to set the interface to the correct speed manually. As for de1, can you compile if_de.c with TULIP_DEBUG defined? This should give some more indications why the PHY isn't set up correctly for 100baseTX. I've just looked up the data sheets for both tranceivers, and the Kendin KS8761 (http://www.kendin.com/) doesn't support the MII control interface (the DM9101 does, http://www.davicom8.com/lan/dm9101.htm). In this case, the 2114x must specify the link rate to use via some other signal (the tranceiver needs two signals SEL10 and SEL100). I'll try to look into this. > arp: 192.168.10.254 is on de0 but got reply from 00:80:48:d9:76:77 on de1 > > I got last message because now kernel recompiled without bridging support. > But it looks > strange - it is not an error from my point of view... :) It's just a warning you can safely ignore. One could argue that it is pointless for the kernel to report stations moving from one interface to another, but usually, people don't connect two interfaces to the same net, either. Stefan -- Stefan Bethke Promo Datentechnik | Tel. +49-40-851744-18 + Systemberatung GmbH | Fax. +49-40-851744-44 Eduardstrasse 46-48 | e-mail: stefan@Promo.DE D-20257 Hamburg | http://www.Promo.DE/ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?327978.3131868563>