From owner-freebsd-xen@FreeBSD.ORG Mon Aug 1 03:45:17 2011 Return-Path: Delivered-To: freebsd-xen@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 4F1CA106566C for ; Mon, 1 Aug 2011 03:45:17 +0000 (UTC) (envelope-from armondbc@gmail.com) Received: from mail-qw0-f54.google.com (mail-qw0-f54.google.com [209.85.216.54]) by mx1.freebsd.org (Postfix) with ESMTP id 162768FC0C for ; Mon, 1 Aug 2011 03:45:16 +0000 (UTC) Received: by qwc9 with SMTP id 9so3626258qwc.13 for ; Sun, 31 Jul 2011 20:45:16 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=mime-version:date:message-id:subject:from:to:content-type; bh=8jyLjZNK72qA8/ZLXeIo9NNu7P9fyO4Ssn71t8sMsP4=; b=vTv9bKhhuntjRqMc95czSc2DXt/A2XgeTBmM1r7EzkEvnkGaQB1f92nOIoLu4Kwshw w7rOlZhU52Ev6AcU/Q5R1j8BmDWQZ51iI2qJF2KB5Y4XX0tMNttQYw1vp7Uw9tlGMcxG xtJhxYxdyKbWYo4YyftXHjwYX8P3VV+XPqlHQ= MIME-Version: 1.0 Received: by 10.229.14.158 with SMTP id g30mr2384240qca.34.1312168932325; Sun, 31 Jul 2011 20:22:12 -0700 (PDT) Received: by 10.229.187.18 with HTTP; Sun, 31 Jul 2011 20:22:12 -0700 (PDT) Date: Sun, 31 Jul 2011 22:22:12 -0500 Message-ID: From: "Ben C." To: freebsd-xen@freebsd.org Content-Type: text/plain; charset=ISO-8859-1 Subject: Networking issues with FreeBSD HVM + SMP X-BeenThere: freebsd-xen@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Discussion of the freebsd port to xen - implementation and usage List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 01 Aug 2011 03:45:17 -0000 Hello all, Thanks for taking time to read this post. It's a fairly long winded tale of confusion, but here goes: * I originally installed FreeBSD 8.2-RELEASE fine working under 100% HVM. Using the GENERIC kernel with xen's qemu * I trashed the box (kind of willingly, it's been a while since I've had the pleasure of FreeBSD and think of it as my first pancake) * Reinstalled. Networking failed. I had no idea why. Several weeks past, I finally got networking back working after about 20-30 install attempts. * Setup the box properly. * Did some changes throughout the dom0/domu's - I had originally had the vcpu's pinned to physical cpu's and all this mojo. I found that my load averages on the dom0 were significantly lower when the vcpu's were bound to 'any physical' -- but that's beside the point. The point is, FreeBSD changed vcpu's from 1 to 2 - and networking stopped. * I honestly thought it surely was something else. I triple-checked and confirmed, when vcpu=1, networking worked as expected. When vcpu was >1, networking failed with a vauge .. em0 (or re0, see below) .. watchdog timeout spewing all over. If you take a look at my configuration below, I originally had model=e1000 in the vif, which I *thought* was what made it work originally. I was apprently wrong. vcpu pinning doesn't matter. Essentially, if it has more than 1 (v)cpu's, networking fails. The dom0 is running NetBSD -CURRENT [current there is just like current here], on an Intel i7. I could try the -current branch or using the xenhvm kernel/drivers. I'm fairly sure the latter may help the situation, but honestly I'm just a bit bothered by the strangeness of this "bug". Does anyone else have any suggestions on what I should try next? Thanks for your time, Ben name ="a5freebsd" memory = 2048 kernel = "/usr/pkg/lib/xen/boot/hvmloader" builder='hvm' device_model = '/usr/pkg/libexec/qemu-dm' disk = [ 'phy:/dev/mapper/rxenpool-a5freebsd_root,ioemu:hda,w', 'file:/home/xen/iso/FreeBSD-8.2-RELEASE-amd64-disc1.iso,ioemu:hdc:cdrom,r' ] boot = 'cd' vcpus=2 #cpus=['2'] #vif=[ 'type=ioemu,bridge=bridge0,model=e1000,mac=00:16:3e:4f:bb:78' ] vif=[ 'type=ioemu,bridge=bridge0,mac=00:16:3e:4f:bb:78' ] vnc = 1 vncdisplay = 2 vnclisten = "1.2.3.4" vncpasswd = "removed" on_reboot = 'restart' on_crash = 'restart' usbdevice = 'tablet'