Date: Wed, 08 Dec 2004 04:11:45 +0200 From: Iasen Kostov <tbyte@OTEL.net> To: Tony Ackerman <tackerman@hub.freebsd.org> Cc: freebsd-net@freebsd.org Subject: Re: em(4) VLAN + PROMISC still doesn't work with latest CVS version Message-ID: <41B662E1.1040303@OTEL.net> In-Reply-To: <20041207232408.GA26544@hub.freebsd.org> References: <Pine.NEB.3.96L.1041128161659.94120E-100000@fledge.watson.org> <41AB0B98.6020600@OTEL.net> <41B5BC98.2080408@OTEL.net> <20041207232408.GA26544@hub.freebsd.org>
next in thread | previous in thread | raw e-mail | index | archive | help
Tony Ackerman wrote: >What is the purpose of putting em1 in promiscuous mode below? Is >the required or did you just notice the issue with this configuration? > >There was a change added some months ago in order to allow the >bridging of vlans. In order for vlan briding to work the interface >had to have vlan tagging/stripping disabled when promisc mode is >invoked (which is how bridge works). The side effect is that now >tcpdump which puts the interface in promisc mode by default will >in effect cripple the interface. > >However, if "tcpdump -p" does not put the interface in promisc mode >and it works just fine. > > > What about bridge over VLAN ? If you can make a bridge over vlans without putting interfaces in promisc mode I won't have some of the problems - but I don't think it is possible ... More, if you forget to put -p after tcpdump (or trafshow) in ssh session ... BOOM. At least until sshd times out and drops the session and kills tcpdump but of course this is rather annoying :). And some traffic accounters go into background so they won't die with the ssh session making you say large amount of not so nice words before calling the support asking them to reboot the machine :). >On Tue, Dec 07, 2004 at 04:22:16PM +0200, Iasen Kostov wrote: > > >>Iasen Kostov wrote: >> >> >> >>>Robert Watson wrote: >>> >>> >>> >>>>On Sat, 27 Nov 2004, Kevin Day wrote: >>>> >>>> >>>> >>>> >>>> >>>>>I recently upgraded to 5.3 on a system, and manually upgraded >>>>>src/sys/dev/em/* to the latest RELENG_5 versions. (1.44.2.4 of >>>>>if_em.c) >>>>> >>>>> >>>>I'm able to reproduce problems using the below configuration is 6.x >>>>also, >>>>and am investigating. Thanks for the report, hope to get back to you >>>>shortly with something concrete. >>>> >>>>Robert N M Watson FreeBSD Core Team, TrustedBSD Projects >>>>robert@fledge.watson.org Principal Research Scientist, McAfee >>>>Research >>>> >>>> >>>> >>>> >>>> >>>> >>>>>While the VLAN side of things works better than the stock 5.3 version, >>>>>there still is this problem: >>>>> >>>>>ifconfig vlan1 create >>>>>ifconfig vlan1 vlan 1 vlandev em1 link0 >>>>>ifconfig vlan2 create >>>>>ifconfig vlan2 vlan 2 vlandev em1 link0 >>>>>ifconfig vlan3 create >>>>>ifconfig vlan3 vlan 3 vlandev em1 link0 >>>>> >>>>>ifconfig vlan1 inet 192.aaa.bbb.129 netmask 255.255.255.0 >>>>>ifconfig vlan2 inet 64.ccc.ddd.61 netmask 255.255.255.192 >>>>>ifconfig vlan3 inet 64.eee.fff.61 netmask 255.255.255.192 >>>>> >>>>>ifconfig em1 up >>>>>ifconfig em1 promisc >>>>> >>>>>If I do this, vlan1 and vlan3 work fine. Vlan2 can receive packets, >>>>>but anything sent out vlan2 doesn't seem to be heard by any foreign >>>>>hosts. Setting "ifconfig em1 -promisc" makes all vlans work properly. >>>>> >>>>>This is better than the stock 5.3 version of em(4) where none of the >>>>>vlans worked, but something still isn't right. >>>>> >>>>>Is this a known problem still or am I just doing something wrong? >>>>> >>>>> >>>>> >>>>> >>>>> >>>>> >>>>> >>>Saddly I can just confirm that :( >>> >>> regards >>> >>>_______________________________________________ >>>freebsd-net@freebsd.org mailing list >>>http://lists.freebsd.org/mailman/listinfo/freebsd-net >>>To unsubscribe, send any mail to "freebsd-net-unsubscribe@freebsd.org" >>> >>> >>> >> Is there an update on this case or I should find a way to disable >>all hw "things" in the driver ?:) (because things are getting hot here :). >> >> regards >> >> > > >
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?41B662E1.1040303>