NetBSD Problem Report #47313

From  Wed Dec 12 14:55:05 2012
Return-Path: <>
Received: from ( [])
	by (Postfix) with ESMTP id 7278F63E33E
	for <>; Wed, 12 Dec 2012 14:55:05 +0000 (UTC)
Message-Id: <>
Date: Wed, 12 Dec 2012 15:55:02 +0100
Subject: xen/bridge multicast reception problem
X-Send-Pr-Version: 3.95

>Number:         47313
>Category:       port-xen
>Synopsis:       dom0 doesn't react on domU's multicast packets
>Confidential:   no
>Severity:       serious
>Priority:       medium
>Responsible:    port-xen-maintainer
>State:          closed
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Wed Dec 12 15:00:00 +0000 2012
>Closed-Date:    Fri Jan 21 08:29:34 +0000 2022
>Last-Modified:  Fri Jan 21 08:29:34 +0000 2022
>Release:        NetBSD 6.0_RC2
computer science department, university of Bonn, Germany

System: NetBSD jaguar-eta 6.0_RC2 NetBSD 6.0_RC2 (XEN3PAE_DOMU) #8: Thu Sep 27 17:08:35 CEST 2012 i386
Architecture: i386
Machine: i386

	The original problem was that I can't ssh to a dom0 from a
hosted domU. Further examination showed that multicast packets (both
IPv4 and IPv6) from the domU can be tcpdump'ed on the dom0, but the
dom0 doesn't react on them. Bridging them to the cable works, though;
I see pongs and can connect to other machines on the LAN.

This has been tested with both xenkernel41/xentools41 and
xenkernel33/xentools33, on different machines with different
ethernet hardware. (bge vs. bnx)

All virtual and real machines run 6.0-postbeta1 or later, dom0 are amd64,
domU are i386-PAE.

	From a DomU:
	ssh -6 name-of-dom0 (from dom-U)
	ping6 ff02::1%xennet0



Responsible-Changed-From-To: kern-bug-people->port-xen-maintainer
Responsible-Changed-When: Fri, 17 Jul 2020 11:01:01 +0000
This seems to be xen support issue.

State-Changed-From-To: open->feedback
State-Changed-When: Fri, 17 Jul 2020 11:01:01 +0000
There was recently some multicast fix for xennet(4) which might have fixed
this. Can you check with up-to-date -current?

Subject: Re: port-xen/47313 xen/bridge multicast reception problem
Date: Thu, 20 Jan 2022 18:35:54 +0100


 > There was recently some multicast fix for xennet(4) which might have fixed
 > this. Can you check with up-to-date -current?

 I can confirm that the problem is fixed, at least with

 - xenkernel411 (and xentools411) on the Xen host
 - netbsd9.99.92 kernel/modules on the Xen host, wm driver
 - netbsd 9.2 userland on the Xen host 
 - netbsd-7.1 netbsd/i386 on the VM

 (The machine where I originally saw the problem is also still alive,
 but I seem to have botched the upgrade, or there's a different kernel
 or Xen problem affecting that hardware, so no test right now for the
 original configuration.)

 Has there been a pullup to netbsd-9? I could test that, too.


State-Changed-From-To: feedback->closed
State-Changed-When: Fri, 21 Jan 2022 08:29:34 +0000
There were some changes around checksums, those are most likely to fix
it. Perhaps it can be determined by bisection. For now no netbsd-9 pullup.


NetBSD Home
NetBSD PR Database Search

(Contact us) $NetBSD: query-full-pr,v 1.46 2020/01/03 16:35:01 leot Exp $
$NetBSD:,v 1.9 2014/08/02 14:16:04 spz Exp $
Copyright © 1994-2020 The NetBSD Foundation, Inc. ALL RIGHTS RESERVED.