This is an archived copy of the Xen.org mailing list, which we have preserved to ensure that existing links to archives are not broken. The live archive, which contains the latest emails, can be found at http://lists.xen.org/
Home Products Support Community News


Re: [Xen-devel] [PATCH 2/2] tools/hotplug, Use udev rules instead of qem

To: Stefano Stabellini <Stefano.Stabellini@xxxxxxxxxxxxx>
Subject: Re: [Xen-devel] [PATCH 2/2] tools/hotplug, Use udev rules instead of qemu script to setup the bridge.
From: Anthony PERARD <anthony.perard@xxxxxxxxxx>
Date: Fri, 06 Aug 2010 16:49:12 +0100
Cc: "xen-devel@xxxxxxxxxxxxxxxxxxx" <xen-devel@xxxxxxxxxxxxxxxxxxx>
Delivery-date: Fri, 06 Aug 2010 08:50:22 -0700
Envelope-to: www-data@xxxxxxxxxxxxxxxxxxx
In-reply-to: <alpine.DEB.2.00.1008041835230.19809@kaball-desktop>
List-help: <mailto:xen-devel-request@lists.xensource.com?subject=help>
List-id: Xen developer discussion <xen-devel.lists.xensource.com>
List-post: <mailto:xen-devel@lists.xensource.com>
List-subscribe: <http://lists.xensource.com/mailman/listinfo/xen-devel>, <mailto:xen-devel-request@lists.xensource.com?subject=subscribe>
List-unsubscribe: <http://lists.xensource.com/mailman/listinfo/xen-devel>, <mailto:xen-devel-request@lists.xensource.com?subject=unsubscribe>
References: <1280938265-12107-1-git-send-email-anthony.perard@xxxxxxxxxx> <1280938265-12107-3-git-send-email-anthony.perard@xxxxxxxxxx> <alpine.DEB.2.00.1008041835230.19809@kaball-desktop>
Sender: xen-devel-bounces@xxxxxxxxxxxxxxxxxxx
User-agent: Mozilla-Thunderbird (X11/20100328)
Stefano Stabellini wrote:
On Wed, 4 Aug 2010, anthony.perard@xxxxxxxxxx wrote:
From: Anthony PERARD <anthony.perard@xxxxxxxxxx>

This patch adds a second argument to vif-bridge script. It can be "vif"
or "tap". "vif" give the default behavior and "tap" just add the
interface to the found bridge when the action is "add".

Signed-off-by: Anthony PERARD <anthony.perard@xxxxxxxxxx>
 tools/hotplug/Linux/vif-bridge        |   42 +++++++++++++++++++++++++++-----
 tools/hotplug/Linux/vif-common.sh     |   12 ++++++---
 tools/hotplug/Linux/xen-backend.rules |    5 ++-
 tools/libxl/libxl.c                   |    4 +-
 4 files changed, 48 insertions(+), 15 deletions(-)

diff --git a/tools/hotplug/Linux/vif-bridge b/tools/hotplug/Linux/vif-bridge
index d35144e..63223f3 100644
--- a/tools/hotplug/Linux/vif-bridge
+++ b/tools/hotplug/Linux/vif-bridge
@@ -29,6 +29,13 @@
 # rules for its ip addresses (if any).
+# Older versions of Xen do not pass in the type as an argument
+if [ $# -lt 2 ]; then
+    type_if=vif
+    type_if=$2
 dir=$(dirname "$0")
 . "$dir/vif-common.sh"
@@ -79,22 +86,43 @@ then
     fatal "Could not find bridge device $bridge"
+case "$type_if" in
+    vif)
+        dev=$vif
+        ;;
+    tap)
+        dev=$INTERFACE
+        ;;
 case "$command" in
-       setup_bridge_port "$vif"
-       add_to_bridge "$bridge" "$vif"
+        if [ "$type_if" = vif ]; then
+            setup_bridge_port "$vif"
+            add_to_bridge "$bridge" "$vif"
+        fi
-        do_without_error brctl delif "$bridge" "$vif"
-        do_without_error ifconfig "$vif" down
+        if [ "$type_if" = vif ]; then
+            do_without_error brctl delif "$bridge" "$vif"
+            do_without_error ifconfig "$vif" down
+        fi
+        ;;
+    add)
+        if [ "$type_if" = tap ]; then
+            add_to_bridge "$bridge" "$dev"
+        fi

Can we limit the amount of "if" introduced by this patch somehow?
If we use a different command (add instead of online), why do we need
the "if" under the online and offline cases?
If add is only used with tap devices, why do we need an if under add?

The command come directly from the $ACTION in udev, and for some reason, this event is different in both cases. Here, I just check if we really want to setup a vif or a tap. This is too much? I can check the argument before that, in vif-common.sh, like is already done for a vif.

+if [ "$type_if" = vif ]; then
+    handle_iptable

why is handle_iptable only called with vifs?

Because iptable wasn't handled before with the qemu script and it doesn't seem to be necessary to handle iptable for a tap device.

-log debug "Successful vif-bridge $command for $vif, bridge $bridge."
-if [ "$command" == "online" ]
+log debug "Successful vif-bridge $command for $dev, bridge $bridge."
+if [ "$type_if" = vif -a "$command" = "online" ]
diff --git a/tools/hotplug/Linux/vif-common.sh 
index 44dd342..4cf00c0 100644
--- a/tools/hotplug/Linux/vif-common.sh
+++ b/tools/hotplug/Linux/vif-common.sh
@@ -33,7 +33,9 @@ fi
case "$command" in
     add | remove)
-        exit 0
+        if [ "$type_if" != tap ]; then
+            exit 0
+        fi
@@ -48,9 +50,11 @@ evalVariables "$@"
 ip=$(xenstore_read_default "$XENBUS_PATH/ip" "$ip")
-# Check presence of compulsory args.
+if [ "$type_if" != tap ]; then
+    # Check presence of compulsory args.
+    vif="${vif:?}"

vifs get the parameter from xenstore, how do tap devices get the

The interface name is get from $INTERFACE, and the bridge is found automatically by the script. But there must be a way to take the bridge from xenstore, by using the name of the device (tap32.0) to have the path.

Anthony PERARD

Xen-devel mailing list

<Prev in Thread] Current Thread [Next in Thread>