On Fri, Jun 16, 2017 at 06:48:38PM +0800, Wei Wang wrote:
This patch enables the virtio-net tx queue size to be configurable
between 256 (the default queue size) and 1024 by the user when the
vhost-user backend is used.
Currently, the maximum tx queue size for other backends is 512 due
to the following limitations:
- QEMU backend: the QEMU backend implementation in some cases may
send 1024+1 iovs to writev.
- Vhost_net backend: there are possibilities that the guest sends
a vring_desc of memory which corsses a MemoryRegion thereby
generating more than 1024 iovs in total after translattion from
guest-physical address in the backend.
Signed-off-by: Wei Wang <address@hidden>
---
hw/net/virtio-net.c | 46
++++++++++++++++++++++++++++++++++--------
include/hw/virtio/virtio-net.h | 1 +
2 files changed, 39 insertions(+), 8 deletions(-)
diff --git a/hw/net/virtio-net.c b/hw/net/virtio-net.c
index 7d091c9..e1a08fd 100644
--- a/hw/net/virtio-net.c
+++ b/hw/net/virtio-net.c
@@ -33,8 +33,11 @@
/* previously fixed value */
#define VIRTIO_NET_RX_QUEUE_DEFAULT_SIZE 256
+#define VIRTIO_NET_TX_QUEUE_DEFAULT_SIZE 256
+
/* for now, only allow larger queues; with virtio-1, guest can downsize */
#define VIRTIO_NET_RX_QUEUE_MIN_SIZE VIRTIO_NET_RX_QUEUE_DEFAULT_SIZE
+#define VIRTIO_NET_TX_QUEUE_MIN_SIZE VIRTIO_NET_TX_QUEUE_DEFAULT_SIZE
/*
* Calculate the number of bytes up to and including the given 'field' of
@@ -1491,18 +1494,33 @@ static void virtio_net_tx_bh(void *opaque)
static void virtio_net_add_queue(VirtIONet *n, int index)
{
VirtIODevice *vdev = VIRTIO_DEVICE(n);
+ NetClientState *nc = qemu_get_queue(n->nic);
n->vqs[index].rx_vq = virtio_add_queue(vdev, n->net_conf.rx_queue_size,
virtio_net_handle_rx);
+
+ /*
+ * Currently, backends other than vhost-user don't support 1024 queue
+ * size.
+ */
+ if (n->net_conf.tx_queue_size == VIRTQUEUE_MAX_SIZE &&
+ nc->peer->info->type != NET_CLIENT_DRIVER_VHOST_USER) {
+ fprintf(stderr, "warning: %s: queue size %d not supported\n",
+ __func__, n->net_conf.tx_queue_size);
+ n->net_conf.tx_queue_size = VIRTIO_NET_TX_QUEUE_DEFAULT_SIZE;
+ }
+
Also, I suspect we can get here with no peer, and above will crash.
It seems ugly to do this on each virtio_net_add_queue.
How about moving this to realize?