[Qemu-devel] [RFC] net: Peer with existing NIC in netdev_add

[Stefan Hajnoczi]

Allow netdev_del followed by netdev_add to re-peer a NIC and its netdev:

  (qemu) info network
  virtio-net-pci.0: type=nic,model=virtio-net-pci,macaddr=52:54:00:12:34:56
   \ netdev0: type=user,net=10.0.2.0,restrict=off

  (qemu) netdev_del netdev0

  (qemu) netdev_add socket,id=netdev0,listen=:1234

  (qemu) info network
  virtio-net-pci.0: type=nic,model=virtio-net-pci,macaddr=52:54:00:12:34:56
   \ netdev0: type=socket,

This makes it possible to switch netdev while the guest is running.  It
is not necessary to reset the NIC.

Note that the NIC's link goes down in netdev_del and back up again in
netdev_add.  Therefore the guest becomes aware that the network has
changed, although this depends on the emulated NIC model providing link
status change interrupts.

Signed-off-by: Stefan Hajnoczi <address@hidden>
---
This patch addresses the netdev hotplug issue that has been discussed
previously on the mailing list:

  https://www.redhat.com/archives/libvir-list/2012-October/msg00629.html

Initially I thought we need to add a new monitor command.  After looking at the
code it seems that simply allowing netdev_del followed by netdev_add (as
suggested by Laine) is the simplest way to solve the problem.  No new monitor
commands are necessary.

The current code is not safe with virtio-net + tap, where the virtio-net device
reports offload features from the tap device to the guest.  If you try to
switch to a -netdev user or -netdev socket device, those offload capabilities
are not present and network communication will fail.

Laine: Please try this out and see if it works for your use case.

 net.c | 32 ++++++++++++++++++++++++++++++++
 1 file changed, 32 insertions(+)

diff --git a/net.c b/net.c
index ae4bc0d..d7c60b0 100644
--- a/net.c
+++ b/net.c
@@ -619,6 +619,19 @@ static int (* const 
net_client_init_fun[NET_CLIENT_OPTIONS_KIND_MAX])(
         [NET_CLIENT_OPTIONS_KIND_HUBPORT]   = net_init_hubport,
 };
 
+static NICState *net_find_nic_for_peername(const char *name)
+{
+    NetClientState *nc;
+    QTAILQ_FOREACH(nc, &net_clients, next) {
+        if (nc->info->type != NET_CLIENT_OPTIONS_KIND_NIC) {
+            continue;
+        }
+        if (nc->peer && strcmp(nc->peer->name, name) == 0) {
+            return (NICState *)nc;
+        }
+    }
+    return NULL;
+}
 
 static int net_client_init1(const void *object, int is_netdev, Error **errp)
 {
@@ -678,6 +691,25 @@ static int net_client_init1(const void *object, int 
is_netdev, Error **errp)
                       NetClientOptionsKind_lookup[opts->kind]);
             return -1;
         }
+
+        /* Peer with existing nic if netdev name matches */
+        if (is_netdev && opts->kind != NET_CLIENT_OPTIONS_KIND_NIC) {
+            NICState *nic = net_find_nic_for_peername(name);
+            if (nic && nic->peer_deleted) {
+                NetClientState *netdev = qemu_find_netdev(name);
+
+                /* TODO vnet_hdr compatibility check */
+
+                qemu_free_net_client(nic->nc.peer);
+                nic->nc.peer = netdev;
+                netdev->peer = &nic->nc;
+                nic->peer_deleted = false;
+                nic->nc.link_down = false;
+                if (nic->nc.info->link_status_changed) {
+                    nic->nc.info->link_status_changed(&nic->nc);
+                }
+            }
+        }
     }
     return 0;
 }
-- 
1.7.11.7