1 files changed, 401 insertions, 46 deletions

diff --git a/drivers/vdpa/vdpa_sim/vdpa_sim_net.c b/drivers/vdpa/vdpa_sim/vdpa_sim_net.c
index d5324f6fd8c7..6caf09a1907b 100644
--- a/drivers/vdpa/vdpa_sim/vdpa_sim_net.c
+++ b/drivers/vdpa/vdpa_sim/vdpa_sim_net.c
@@ -11,10 +11,10 @@
 #include <linux/module.h>
 #include <linux/device.h>
 #include <linux/kernel.h>
-#include <linux/sched.h>
 #include <linux/etherdevice.h>
 #include <linux/vringh.h>
 #include <linux/vdpa.h>
+#include <net/netlink.h>
 #include <uapi/linux/virtio_net.h>
 #include <uapi/linux/vdpa.h>
 
@@ -26,81 +26,385 @@
 #define DRV_LICENSE  "GPL v2"
 
 #define VDPASIM_NET_FEATURES	(VDPASIM_FEATURES | \
-				 (1ULL << VIRTIO_NET_F_MAC))
+				 (1ULL << VIRTIO_NET_F_MAC) | \
+				 (1ULL << VIRTIO_NET_F_STATUS) | \
+				 (1ULL << VIRTIO_NET_F_MTU) | \
+				 (1ULL << VIRTIO_NET_F_CTRL_VQ) | \
+				 (1ULL << VIRTIO_NET_F_CTRL_MAC_ADDR))
+
+/* 3 virtqueues, 2 address spaces, 2 virtqueue groups */
+#define VDPASIM_NET_VQ_NUM	3
+#define VDPASIM_NET_AS_NUM	2
+#define VDPASIM_NET_GROUP_NUM	2
+
+struct vdpasim_dataq_stats {
+	struct u64_stats_sync syncp;
+	u64 pkts;
+	u64 bytes;
+	u64 drops;
+	u64 errors;
+	u64 overruns;
+};
+
+struct vdpasim_cq_stats {
+	struct u64_stats_sync syncp;
+	u64 requests;
+	u64 successes;
+	u64 errors;
+};
+
+struct vdpasim_net{
+	struct vdpasim vdpasim;
+	struct vdpasim_dataq_stats tx_stats;
+	struct vdpasim_dataq_stats rx_stats;
+	struct vdpasim_cq_stats cq_stats;
+	void *buffer;
+};
+
+static struct vdpasim_net *sim_to_net(struct vdpasim *vdpasim)
+{
+	return container_of(vdpasim, struct vdpasim_net, vdpasim);
+}
+
+static void vdpasim_net_complete(struct vdpasim_virtqueue *vq, size_t len)
+{
+	/* Make sure data is wrote before advancing index */
+	smp_wmb();
+
+	vringh_complete_iotlb(&vq->vring, vq->head, len);
+
+	/* Make sure used is visible before rasing the interrupt. */
+	smp_wmb();
+
+	local_bh_disable();
+	if (vringh_need_notify_iotlb(&vq->vring) > 0)
+		vringh_notify(&vq->vring);
+	local_bh_enable();
+}
+
+static bool receive_filter(struct vdpasim *vdpasim, size_t len)
+{
+	bool modern = vdpasim->features & (1ULL << VIRTIO_F_VERSION_1);
+	size_t hdr_len = modern ? sizeof(struct virtio_net_hdr_v1) :
+				  sizeof(struct virtio_net_hdr);
+	struct virtio_net_config *vio_config = vdpasim->config;
+	struct vdpasim_net *net = sim_to_net(vdpasim);
+
+	if (len < ETH_ALEN + hdr_len)
+		return false;
+
+	if (is_broadcast_ether_addr(net->buffer + hdr_len) ||
+	    is_multicast_ether_addr(net->buffer + hdr_len))
+		return true;
+	if (!strncmp(net->buffer + hdr_len, vio_config->mac, ETH_ALEN))
+		return true;
+
+	return false;
+}
+
+static virtio_net_ctrl_ack vdpasim_handle_ctrl_mac(struct vdpasim *vdpasim,
+						   u8 cmd)
+{
+	struct virtio_net_config *vio_config = vdpasim->config;
+	struct vdpasim_virtqueue *cvq = &vdpasim->vqs[2];
+	virtio_net_ctrl_ack status = VIRTIO_NET_ERR;
+	size_t read;
+
+	switch (cmd) {
+	case VIRTIO_NET_CTRL_MAC_ADDR_SET:
+		read = vringh_iov_pull_iotlb(&cvq->vring, &cvq->in_iov,
+					     vio_config->mac, ETH_ALEN);
+		if (read == ETH_ALEN)
+			status = VIRTIO_NET_OK;
+		break;
+	default:
+		break;
+	}
 
-#define VDPASIM_NET_VQ_NUM	2
+	return status;
+}
 
-static void vdpasim_net_work(struct work_struct *work)
+static void vdpasim_handle_cvq(struct vdpasim *vdpasim)
+{
+	struct vdpasim_virtqueue *cvq = &vdpasim->vqs[2];
+	struct vdpasim_net *net = sim_to_net(vdpasim);
+	virtio_net_ctrl_ack status = VIRTIO_NET_ERR;
+	struct virtio_net_ctrl_hdr ctrl;
+	size_t read, write;
+	u64 requests = 0, errors = 0, successes = 0;
+	int err;
+
+	if (!(vdpasim->features & (1ULL << VIRTIO_NET_F_CTRL_VQ)))
+		return;
+
+	if (!cvq->ready)
+		return;
+
+	while (true) {
+		err = vringh_getdesc_iotlb(&cvq->vring, &cvq->in_iov,
+					   &cvq->out_iov,
+					   &cvq->head, GFP_ATOMIC);
+		if (err <= 0)
+			break;
+
+		++requests;
+		read = vringh_iov_pull_iotlb(&cvq->vring, &cvq->in_iov, &ctrl,
+					     sizeof(ctrl));
+		if (read != sizeof(ctrl)) {
+			++errors;
+			break;
+		}
+
+		switch (ctrl.class) {
+		case VIRTIO_NET_CTRL_MAC:
+			status = vdpasim_handle_ctrl_mac(vdpasim, ctrl.cmd);
+			break;
+		default:
+			break;
+		}
+
+		if (status == VIRTIO_NET_OK)
+			++successes;
+		else
+			++errors;
+
+		/* Make sure data is wrote before advancing index */
+		smp_wmb();
+
+		write = vringh_iov_push_iotlb(&cvq->vring, &cvq->out_iov,
+					      &status, sizeof(status));
+		vringh_complete_iotlb(&cvq->vring, cvq->head, write);
+		vringh_kiov_cleanup(&cvq->in_iov);
+		vringh_kiov_cleanup(&cvq->out_iov);
+
+		/* Make sure used is visible before rasing the interrupt. */
+		smp_wmb();
+
+		local_bh_disable();
+		if (cvq->cb)
+			cvq->cb(cvq->private);
+		local_bh_enable();
+	}
+
+	u64_stats_update_begin(&net->cq_stats.syncp);
+	net->cq_stats.requests += requests;
+	net->cq_stats.errors += errors;
+	net->cq_stats.successes += successes;
+	u64_stats_update_end(&net->cq_stats.syncp);
+}
+
+static void vdpasim_net_work(struct vdpasim *vdpasim)
 {
-	struct vdpasim *vdpasim = container_of(work, struct vdpasim, work);
 	struct vdpasim_virtqueue *txq = &vdpasim->vqs[1];
 	struct vdpasim_virtqueue *rxq = &vdpasim->vqs[0];
+	struct vdpasim_net *net = sim_to_net(vdpasim);
 	ssize_t read, write;
-	size_t total_write;
-	int pkts = 0;
+	u64 tx_pkts = 0, rx_pkts = 0, tx_bytes = 0, rx_bytes = 0;
+	u64 rx_drops = 0, rx_overruns = 0, rx_errors = 0, tx_errors = 0;
 	int err;
 
-	spin_lock(&vdpasim->lock);
+	mutex_lock(&vdpasim->mutex);
+
+	if (!vdpasim->running)
+		goto out;
 
 	if (!(vdpasim->status & VIRTIO_CONFIG_S_DRIVER_OK))
 		goto out;
 
+	vdpasim_handle_cvq(vdpasim);
+
 	if (!txq->ready || !rxq->ready)
 		goto out;
 
 	while (true) {
-		total_write = 0;
 		err = vringh_getdesc_iotlb(&txq->vring, &txq->out_iov, NULL,
 					   &txq->head, GFP_ATOMIC);
-		if (err <= 0)
-			break;
-
-		err = vringh_getdesc_iotlb(&rxq->vring, NULL, &rxq->in_iov,
-					   &rxq->head, GFP_ATOMIC);
 		if (err <= 0) {
-			vringh_complete_iotlb(&txq->vring, txq->head, 0);
+			if (err)
+				++tx_errors;
 			break;
 		}
 
-		while (true) {
-			read = vringh_iov_pull_iotlb(&txq->vring, &txq->out_iov,
-						     vdpasim->buffer,
-						     PAGE_SIZE);
-			if (read <= 0)
-				break;
+		++tx_pkts;
+		read = vringh_iov_pull_iotlb(&txq->vring, &txq->out_iov,
+					     net->buffer, PAGE_SIZE);
 
-			write = vringh_iov_push_iotlb(&rxq->vring, &rxq->in_iov,
-						      vdpasim->buffer, read);
-			if (write <= 0)
-				break;
+		tx_bytes += read;
 
-			total_write += write;
+		if (!receive_filter(vdpasim, read)) {
+			++rx_drops;
+			vdpasim_net_complete(txq, 0);
+			continue;
 		}
 
-		/* Make sure data is wrote before advancing index */
-		smp_wmb();
+		err = vringh_getdesc_iotlb(&rxq->vring, NULL, &rxq->in_iov,
+					   &rxq->head, GFP_ATOMIC);
+		if (err <= 0) {
+			++rx_overruns;
+			vdpasim_net_complete(txq, 0);
+			break;
+		}
 
-		vringh_complete_iotlb(&txq->vring, txq->head, 0);
-		vringh_complete_iotlb(&rxq->vring, rxq->head, total_write);
+		write = vringh_iov_push_iotlb(&rxq->vring, &rxq->in_iov,
+					      net->buffer, read);
+		if (write <= 0) {
+			++rx_errors;
+			break;
+		}
 
-		/* Make sure used is visible before rasing the interrupt. */
-		smp_wmb();
+		++rx_pkts;
+		rx_bytes += write;
 
-		local_bh_disable();
-		if (vringh_need_notify_iotlb(&txq->vring) > 0)
-			vringh_notify(&txq->vring);
-		if (vringh_need_notify_iotlb(&rxq->vring) > 0)
-			vringh_notify(&rxq->vring);
-		local_bh_enable();
+		vdpasim_net_complete(txq, 0);
+		vdpasim_net_complete(rxq, write);
 
-		if (++pkts > 4) {
-			schedule_work(&vdpasim->work);
+		if (tx_pkts > 4) {
+			vdpasim_schedule_work(vdpasim);
 			goto out;
 		}
 	}
 
 out:
-	spin_unlock(&vdpasim->lock);
+	mutex_unlock(&vdpasim->mutex);
+
+	u64_stats_update_begin(&net->tx_stats.syncp);
+	net->tx_stats.pkts += tx_pkts;
+	net->tx_stats.bytes += tx_bytes;
+	net->tx_stats.errors += tx_errors;
+	u64_stats_update_end(&net->tx_stats.syncp);
+
+	u64_stats_update_begin(&net->rx_stats.syncp);
+	net->rx_stats.pkts += rx_pkts;
+	net->rx_stats.bytes += rx_bytes;
+	net->rx_stats.drops += rx_drops;
+	net->rx_stats.errors += rx_errors;
+	net->rx_stats.overruns += rx_overruns;
+	u64_stats_update_end(&net->rx_stats.syncp);
+}
+
+static int vdpasim_net_get_stats(struct vdpasim *vdpasim, u16 idx,
+				 struct sk_buff *msg,
+				 struct netlink_ext_ack *extack)
+{
+	struct vdpasim_net *net = sim_to_net(vdpasim);
+	u64 rx_pkts, rx_bytes, rx_errors, rx_overruns, rx_drops;
+	u64 tx_pkts, tx_bytes, tx_errors, tx_drops;
+	u64 cq_requests, cq_successes, cq_errors;
+	unsigned int start;
+	int err = -EMSGSIZE;
+
+	switch(idx) {
+	case 0:
+		do {
+			start = u64_stats_fetch_begin(&net->rx_stats.syncp);
+			rx_pkts = net->rx_stats.pkts;
+			rx_bytes = net->rx_stats.bytes;
+			rx_errors = net->rx_stats.errors;
+			rx_overruns = net->rx_stats.overruns;
+			rx_drops = net->rx_stats.drops;
+		} while (u64_stats_fetch_retry(&net->rx_stats.syncp, start));
+
+		if (nla_put_string(msg, VDPA_ATTR_DEV_VENDOR_ATTR_NAME,
+					"rx packets"))
+			break;
+		if (nla_put_u64_64bit(msg, VDPA_ATTR_DEV_VENDOR_ATTR_VALUE,
+				      rx_pkts, VDPA_ATTR_PAD))
+			break;
+		if (nla_put_string(msg, VDPA_ATTR_DEV_VENDOR_ATTR_NAME,
+				  "rx bytes"))
+			break;
+		if (nla_put_u64_64bit(msg, VDPA_ATTR_DEV_VENDOR_ATTR_VALUE,
+				      rx_bytes, VDPA_ATTR_PAD))
+			break;
+		if (nla_put_string(msg, VDPA_ATTR_DEV_VENDOR_ATTR_NAME,
+				  "rx errors"))
+			break;
+		if (nla_put_u64_64bit(msg, VDPA_ATTR_DEV_VENDOR_ATTR_VALUE,
+				      rx_errors, VDPA_ATTR_PAD))
+			break;
+		if (nla_put_string(msg, VDPA_ATTR_DEV_VENDOR_ATTR_NAME,
+				  "rx overruns"))
+			break;
+		if (nla_put_u64_64bit(msg, VDPA_ATTR_DEV_VENDOR_ATTR_VALUE,
+				      rx_overruns, VDPA_ATTR_PAD))
+			break;
+		if (nla_put_string(msg, VDPA_ATTR_DEV_VENDOR_ATTR_NAME,
+				  "rx drops"))
+			break;
+		if (nla_put_u64_64bit(msg, VDPA_ATTR_DEV_VENDOR_ATTR_VALUE,
+				      rx_drops, VDPA_ATTR_PAD))
+			break;
+		err = 0;
+		break;
+	case 1:
+		do {
+			start = u64_stats_fetch_begin(&net->tx_stats.syncp);
+			tx_pkts = net->tx_stats.pkts;
+			tx_bytes = net->tx_stats.bytes;
+			tx_errors = net->tx_stats.errors;
+			tx_drops = net->tx_stats.drops;
+		} while (u64_stats_fetch_retry(&net->tx_stats.syncp, start));
+
+		if (nla_put_string(msg, VDPA_ATTR_DEV_VENDOR_ATTR_NAME,
+				  "tx packets"))
+			break;
+		if (nla_put_u64_64bit(msg, VDPA_ATTR_DEV_VENDOR_ATTR_VALUE,
+				      tx_pkts, VDPA_ATTR_PAD))
+			break;
+		if (nla_put_string(msg, VDPA_ATTR_DEV_VENDOR_ATTR_NAME,
+				  "tx bytes"))
+			break;
+		if (nla_put_u64_64bit(msg, VDPA_ATTR_DEV_VENDOR_ATTR_VALUE,
+				      tx_bytes, VDPA_ATTR_PAD))
+			break;
+		if (nla_put_string(msg, VDPA_ATTR_DEV_VENDOR_ATTR_NAME,
+				  "tx errors"))
+			break;
+		if (nla_put_u64_64bit(msg, VDPA_ATTR_DEV_VENDOR_ATTR_VALUE,
+				      tx_errors, VDPA_ATTR_PAD))
+			break;
+		if (nla_put_string(msg, VDPA_ATTR_DEV_VENDOR_ATTR_NAME,
+				  "tx drops"))
+			break;
+		if (nla_put_u64_64bit(msg, VDPA_ATTR_DEV_VENDOR_ATTR_VALUE,
+				      tx_drops, VDPA_ATTR_PAD))
+			break;
+		err = 0;
+		break;
+	case 2:
+		do {
+			start = u64_stats_fetch_begin(&net->cq_stats.syncp);
+			cq_requests = net->cq_stats.requests;
+			cq_successes = net->cq_stats.successes;
+			cq_errors = net->cq_stats.errors;
+		} while (u64_stats_fetch_retry(&net->cq_stats.syncp, start));
+
+		if (nla_put_string(msg, VDPA_ATTR_DEV_VENDOR_ATTR_NAME,
+				  "cvq requests"))
+			break;
+		if (nla_put_u64_64bit(msg, VDPA_ATTR_DEV_VENDOR_ATTR_VALUE,
+				      cq_requests, VDPA_ATTR_PAD))
+			break;
+		if (nla_put_string(msg, VDPA_ATTR_DEV_VENDOR_ATTR_NAME,
+				  "cvq successes"))
+			break;
+		if (nla_put_u64_64bit(msg, VDPA_ATTR_DEV_VENDOR_ATTR_VALUE,
+				      cq_successes, VDPA_ATTR_PAD))
+			break;
+		if (nla_put_string(msg, VDPA_ATTR_DEV_VENDOR_ATTR_NAME,
+				  "cvq errors"))
+			break;
+		if (nla_put_u64_64bit(msg, VDPA_ATTR_DEV_VENDOR_ATTR_VALUE,
+				      cq_errors, VDPA_ATTR_PAD))
+			break;
+		err = 0;
+		break;
+	default:
+		err = -EINVAL;
+		break;
+	}
+
+	return err;
 }
 
 static void vdpasim_net_get_config(struct vdpasim *vdpasim, void *config)
@@ -110,6 +414,24 @@ static void vdpasim_net_get_config(struct vdpasim *vdpasim, void *config)
 	net_config->status = cpu_to_vdpasim16(vdpasim, VIRTIO_NET_S_LINK_UP);
 }
 
+static int vdpasim_net_set_attr(struct vdpa_mgmt_dev *mdev, struct vdpa_device *dev,
+				const struct vdpa_dev_set_config *config)
+{
+	struct vdpasim *vdpasim = container_of(dev, struct vdpasim, vdpa);
+	struct virtio_net_config *vio_config = vdpasim->config;
+
+	mutex_lock(&vdpasim->mutex);
+
+	if (config->mask & (1 << VDPA_ATTR_DEV_NET_CFG_MACADDR)) {
+		ether_addr_copy(vio_config->mac, config->net.mac);
+		mutex_unlock(&vdpasim->mutex);
+		return 0;
+	}
+
+	mutex_unlock(&vdpasim->mutex);
+	return -EOPNOTSUPP;
+}
+
 static void vdpasim_net_setup_config(struct vdpasim *vdpasim,
 				     const struct vdpa_dev_set_config *config)
 {
@@ -124,6 +446,13 @@ static void vdpasim_net_setup_config(struct vdpasim *vdpasim,
 		vio_config->mtu = cpu_to_vdpasim16(vdpasim, 1500);
 }
 
+static void vdpasim_net_free(struct vdpasim *vdpasim)
+{
+	struct vdpasim_net *net = sim_to_net(vdpasim);
+
+	kvfree(net->buffer);
+}
+
 static void vdpasim_net_mgmtdev_release(struct device *dev)
 {
 }
@@ -137,6 +466,7 @@ static int vdpasim_net_dev_add(struct vdpa_mgmt_dev *mdev, const char *name,
 			       const struct vdpa_dev_set_config *config)
 {
 	struct vdpasim_dev_attr dev_attr = {};
+	struct vdpasim_net *net;
 	struct vdpasim *simdev;
 	int ret;
 
@@ -145,17 +475,38 @@ static int vdpasim_net_dev_add(struct vdpa_mgmt_dev *mdev, const char *name,
 	dev_attr.id = VIRTIO_ID_NET;
 	dev_attr.supported_features = VDPASIM_NET_FEATURES;
 	dev_attr.nvqs = VDPASIM_NET_VQ_NUM;
+	dev_attr.ngroups = VDPASIM_NET_GROUP_NUM;
+	dev_attr.nas = VDPASIM_NET_AS_NUM;
+	dev_attr.alloc_size = sizeof(struct vdpasim_net);
 	dev_attr.config_size = sizeof(struct virtio_net_config);
 	dev_attr.get_config = vdpasim_net_get_config;
 	dev_attr.work_fn = vdpasim_net_work;
-	dev_attr.buffer_size = PAGE_SIZE;
+	dev_attr.get_stats = vdpasim_net_get_stats;
+	dev_attr.free = vdpasim_net_free;
 
-	simdev = vdpasim_create(&dev_attr);
+	simdev = vdpasim_create(&dev_attr, config);
 	if (IS_ERR(simdev))
 		return PTR_ERR(simdev);
 
 	vdpasim_net_setup_config(simdev, config);
 
+	net = sim_to_net(simdev);
+
+	u64_stats_init(&net->tx_stats.syncp);
+	u64_stats_init(&net->rx_stats.syncp);
+	u64_stats_init(&net->cq_stats.syncp);
+
+	net->buffer = kvmalloc(PAGE_SIZE, GFP_KERNEL);
+	if (!net->buffer) {
+		ret = -ENOMEM;
+		goto reg_err;
+	}
+
+	/*
+	 * Initialization must be completed before this call, since it can
+	 * connect the device to the vDPA bus, so requests can arrive after
+	 * this call.
+	 */
 	ret = _vdpa_register_device(&simdev->vdpa, VDPASIM_NET_VQ_NUM);
 	if (ret)
 		goto reg_err;
@@ -177,7 +528,8 @@ static void vdpasim_net_dev_del(struct vdpa_mgmt_dev *mdev,
 
 static const struct vdpa_mgmtdev_ops vdpasim_net_mgmtdev_ops = {
 	.dev_add = vdpasim_net_dev_add,
-	.dev_del = vdpasim_net_dev_del
+	.dev_del = vdpasim_net_dev_del,
+	.dev_set_attr = vdpasim_net_set_attr
 };
 
 static struct virtio_device_id id_table[] = {
@@ -190,7 +542,8 @@ static struct vdpa_mgmt_dev mgmt_dev = {
 	.id_table = id_table,
 	.ops = &vdpasim_net_mgmtdev_ops,
 	.config_attr_mask = (1 << VDPA_ATTR_DEV_NET_CFG_MACADDR |
-			     1 << VDPA_ATTR_DEV_NET_CFG_MTU),
+			     1 << VDPA_ATTR_DEV_NET_CFG_MTU |
+		             1 << VDPA_ATTR_DEV_FEATURES),
 	.max_supported_vqs = VDPASIM_NET_VQ_NUM,
 	.supported_features = VDPASIM_NET_FEATURES,
 };
@@ -200,8 +553,10 @@ static int __init vdpasim_net_init(void)
 	int ret;
 
 	ret = device_register(&vdpasim_net_mgmtdev);
-	if (ret)
+	if (ret) {
+		put_device(&vdpasim_net_mgmtdev);
 		return ret;
+	}
 
 	ret = vdpa_mgmtdev_register(&mgmt_dev);
 	if (ret)