Skip to contentSkip to conversation

Emulate status feature in vhost-vdpa net

PATCH In review Eugenio Pérez 2 participants 5 messages Oct 28, 2022 – Oct 29, 2022 Networking

lore
Share & tools
lore
#0 lore raw

v2 cover Emulate status feature in vhost-vdpa net

The net config space is already copied from the device so it can me modified by qemu. In particular, this is already done to fix cases where the NIC does not expose the right fields.

It's trivial to emulate _F_STATE with qemu if not supported by the device, sice a valid approach is to always show the link as up. If the feature is already supported by the device, no config space modification is needed.

This is a pre requisite to use other features like _F_GUEST_ANNOUNCE, since _F_STATUS is needed for the guest to access the status.

These patches are sent on top of [1] series, so trivial conflicts could arise if it is applied directly on master. Future versions can be not based on it is more convenient.

v2:
* Move feature handling to vhost_net, instead of force depending on the backend.

Eugenio Pérez (3):
virtio_net: Modify virtio_net_get_config to early return
virtio_net: Handle _F_STATUS emulation in virtio_net_get_config
vhost_vdpa: move VIRTIO_NET_F_STATUS handling to vhost_net

 hw/net/vhost_net.c  | 13 +++++++++++++
 hw/net/virtio-net.c | 39 ++++++++++++++++++++++++---------------
 net/vhost-vdpa.c    |  1 -
 3 files changed, 37 insertions(+), 16 deletions(-)
Signature · 1 line
2.31.1
#1 lore raw

PATCH v2 1/3 virtio_net: Modify virtio_net_get_config to early return

Next patches introduce more code on vhost-vdpa branch, with already have too much indentation.

---
 hw/net/virtio-net.c | 28 +++++++++++++++-------------
 1 file changed, 15 insertions(+), 13 deletions(-)
hw/net/virtio-net.c +15 -13
index e9f696b4cf..56ff219196 100644
@@ -158,20 +158,22 @@ static void virtio_net_get_config(VirtIODevice *vdev, uint8_t *config)
     if (nc->peer && nc->peer->info->type == NET_CLIENT_DRIVER_VHOST_VDPA) {
         ret = vhost_net_get_config(get_vhost_net(nc->peer), (uint8_t *)&netcfg,
                                    n->config_size);
-        if (ret != -1) {
-            /*
-             * Some NIC/kernel combinations present 0 as the mac address.  As
-             * that is not a legal address, try to proceed with the
-             * address from the QEMU command line in the hope that the
-             * address has been configured correctly elsewhere - just not
-             * reported by the device.
-             */
-            if (memcmp(&netcfg.mac, &zero, sizeof(zero)) == 0) {
-                info_report("Zero hardware mac address detected. Ignoring.");
-                memcpy(netcfg.mac, n->mac, ETH_ALEN);
-            }
-            memcpy(config, &netcfg, n->config_size);
+        if (ret == -1) {
+            return;
         }
+
+        /*
+         * Some NIC/kernel combinations present 0 as the mac address.  As that
+         * is not a legal address, try to proceed with the address from the
+         * QEMU command line in the hope that the address has been configured
+         * correctly elsewhere - just not reported by the device.
+         */
+        if (memcmp(&netcfg.mac, &zero, sizeof(zero)) == 0) {
+            info_report("Zero hardware mac address detected. Ignoring.");
+            memcpy(netcfg.mac, n->mac, ETH_ALEN);
+        }
+
+        memcpy(config, &netcfg, n->config_size);
     }
 }
 
Signature · 1 line
2.31.1
#2 lore raw

PATCH v2 2/3 virtio_net: Handle _F_STATUS emulation in virtio_net_get_config

In reply to Eugenio Pérez

At this moment this code path is not reached, but vdpa devices can offer VIRTIO_NET_F_STATUS unconditionally. While the guest must assume that link is always up by the standard, qemu will set the status bit to 1 always in this case.

This makes little use by itself, but VIRTIO_NET_F_STATUS is needed for the guest to read status bit VIRTIO_NET_F_GUEST_ANNOUNCE, used by feature VIRTIO_NET_F_GUEST_ANNOUNCE. So qemu must emulate status feature in case it needs to emulate the guest announce feature.

---
 hw/net/virtio-net.c | 11 +++++++++--
 1 file changed, 9 insertions(+), 2 deletions(-)
hw/net/virtio-net.c +9 -2
index 56ff219196..6d4d75615b 100644
@@ -156,8 +156,9 @@ static void virtio_net_get_config(VirtIODevice *vdev, uint8_t *config)
      * disconnect/reconnect a VDPA peer.
      */
     if (nc->peer && nc->peer->info->type == NET_CLIENT_DRIVER_VHOST_VDPA) {
-        ret = vhost_net_get_config(get_vhost_net(nc->peer), (uint8_t *)&netcfg,
-                                   n->config_size);
+        struct vhost_net *net = get_vhost_net(nc->peer);
+
+        ret = vhost_net_get_config(net, (uint8_t *)&netcfg, n->config_size);
         if (ret == -1) {
             return;
         }
@@ -173,6 +174,12 @@ static void virtio_net_get_config(VirtIODevice *vdev, uint8_t *config)
             memcpy(netcfg.mac, n->mac, ETH_ALEN);
         }
 
+        if (vdev->guest_features & BIT_ULL(VIRTIO_NET_F_STATUS) &&
+            !(net->dev.features & BIT_ULL(VIRTIO_NET_F_STATUS))) {
+            /* Emulating link up in qemu */
+            netcfg.status |= virtio_tswap16(vdev, VIRTIO_NET_S_LINK_UP);
+        }
+
         memcpy(config, &netcfg, n->config_size);
     }
 }
Signature · 1 line
2.31.1
#3 lore raw

PATCH v2 3/3 vhost_vdpa: move VIRTIO_NET_F_STATUS handling to vhost_net

In reply to Eugenio Pérez

Since it is emulated on all vhost backends it makes sense to move it.

Although this feature can be emulated by qemu it benefits from information from the device. Ack it as long as the guest ack it.

---
 hw/net/vhost_net.c | 13 +++++++++++++
 net/vhost-vdpa.c   |  1 -
 2 files changed, 13 insertions(+), 1 deletion(-)
hw/net/vhost_net.c +13 -0
index d28f8b974b..b533744211 100644
@@ -127,8 +127,21 @@ int vhost_net_set_config(struct vhost_net *net, const uint8_t *data,
 
 void vhost_net_ack_features(struct vhost_net *net, uint64_t features)
 {
+    static const int status_feature_bit[] = {
+        VIRTIO_NET_F_STATUS,
+        VHOST_INVALID_FEATURE_BIT,
+    };
+
     net->dev.acked_features = net->dev.backend_features;
     vhost_ack_features(&net->dev, vhost_net_get_feature_bits(net), features);
+    if (net->dev.features & BIT_ULL(VIRTIO_NET_F_STATUS)) {
+        /*
+         * If device support _F_STATUS qemu should ack it so it reports link
+         * status changes. If not supported qemu emulates it reporting an
+         * always up link.
+         */
+        vhost_ack_features(&net->dev, status_feature_bit, features);
+    }
 }
 
 uint64_t vhost_net_get_max_queues(VHostNetState *net)
net/vhost-vdpa.c +0 -1
index 6d64000202..854b27186c 100644
@@ -72,7 +72,6 @@ const int vdpa_feature_bits[] = {
     VIRTIO_NET_F_RSS,
     VIRTIO_NET_F_HASH_REPORT,
     VIRTIO_NET_F_GUEST_ANNOUNCE,
-    VIRTIO_NET_F_STATUS,
     VHOST_INVALID_FEATURE_BIT
 };
 
Signature · 1 line
2.31.1
Reviewed-by on v2 3/3
#4 lore raw

Re: [PATCH v2 3/3] vhost_vdpa: move VIRTIO_NET_F_STATUS handling to vhost_net

Quoted 6 lines · Eugenio Pérez

Since it is emulated on all vhost backends it makes sense to move it.

Although this feature can be emulated by qemu it benefits from information from the device. Ack it as long as the guest ack it.

So I assume Jason's tree too?

Quoted 45 lines
---
 hw/net/vhost_net.c | 13 +++++++++++++
 net/vhost-vdpa.c   |  1 -
 2 files changed, 13 insertions(+), 1 deletion(-)
hw/net/vhost_net.c +13 -0
index d28f8b974b..b533744211 100644
@@ -127,8 +127,21 @@ int vhost_net_set_config(struct vhost_net *net, const uint8_t *data,
 
 void vhost_net_ack_features(struct vhost_net *net, uint64_t features)
 {
+    static const int status_feature_bit[] = {
+        VIRTIO_NET_F_STATUS,
+        VHOST_INVALID_FEATURE_BIT,
+    };
+
     net->dev.acked_features = net->dev.backend_features;
     vhost_ack_features(&net->dev, vhost_net_get_feature_bits(net), features);
+    if (net->dev.features & BIT_ULL(VIRTIO_NET_F_STATUS)) {
+        /*
+         * If device support _F_STATUS qemu should ack it so it reports link
+         * status changes. If not supported qemu emulates it reporting an
+         * always up link.
+         */
+        vhost_ack_features(&net->dev, status_feature_bit, features);
+    }
 }
 
 uint64_t vhost_net_get_max_queues(VHostNetState *net)
net/vhost-vdpa.c +0 -1
index 6d64000202..854b27186c 100644
@@ -72,7 +72,6 @@ const int vdpa_feature_bits[] = {
     VIRTIO_NET_F_RSS,
     VIRTIO_NET_F_HASH_REPORT,
     VIRTIO_NET_F_GUEST_ANNOUNCE,
-    VIRTIO_NET_F_STATUS,
     VHOST_INVALID_FEATURE_BIT
 };
 
Signature · 1 line
2.31.1