Resguardo inicial

linux-7.2.3/drivers/vdpa/vdpa_sim/Makefile

@@ -0,0 +1,4 @@
+# SPDX-License-Identifier: GPL-2.0
+obj-$(CONFIG_VDPA_SIM) += vdpa_sim.o
+obj-$(CONFIG_VDPA_SIM_NET) += vdpa_sim_net.o
+obj-$(CONFIG_VDPA_SIM_BLOCK) += vdpa_sim_blk.o

linux-7.2.3/drivers/vdpa/vdpa_sim/vdpa_sim.c

@@ -0,0 +1,850 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * VDPA device simulator core.
+ *
+ * Copyright (c) 2020, Red Hat Inc. All rights reserved.
+ *     Author: Jason Wang <jasowang@redhat.com>
+ *
+ */
+
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/device.h>
+#include <linux/kernel.h>
+#include <linux/kthread.h>
+#include <linux/slab.h>
+#include <linux/dma-map-ops.h>
+#include <linux/vringh.h>
+#include <linux/vdpa.h>
+#include <linux/vhost_iotlb.h>
+#include <uapi/linux/vdpa.h>
+#include <uapi/linux/vhost_types.h>
+
+#include "vdpa_sim.h"
+
+#define DRV_VERSION  "0.1"
+#define DRV_AUTHOR   "Jason Wang <jasowang@redhat.com>"
+#define DRV_DESC     "vDPA Device Simulator core"
+#define DRV_LICENSE  "GPL v2"
+
+static int batch_mapping = 1;
+module_param(batch_mapping, int, 0444);
+MODULE_PARM_DESC(batch_mapping, "Batched mapping 1 -Enable; 0 - Disable");
+
+static int max_iotlb_entries = 2048;
+module_param(max_iotlb_entries, int, 0444);
+MODULE_PARM_DESC(max_iotlb_entries,
+		 "Maximum number of iotlb entries for each address space. (default: 2048)");
+
+static bool use_va = true;
+module_param(use_va, bool, 0444);
+MODULE_PARM_DESC(use_va, "Enable/disable the device's ability to use VA");
+
+#define VDPASIM_QUEUE_ALIGN PAGE_SIZE
+#define VDPASIM_QUEUE_MAX 256
+#define VDPASIM_VENDOR_ID 0
+
+struct vdpasim_mm_work {
+	struct kthread_work work;
+	struct vdpasim *vdpasim;
+	struct mm_struct *mm_to_bind;
+	int ret;
+};
+
+static void vdpasim_mm_work_fn(struct kthread_work *work)
+{
+	struct vdpasim_mm_work *mm_work =
+		container_of(work, struct vdpasim_mm_work, work);
+	struct vdpasim *vdpasim = mm_work->vdpasim;
+
+	mm_work->ret = 0;
+
+	//TODO: should we attach the cgroup of the mm owner?
+	vdpasim->mm_bound = mm_work->mm_to_bind;
+}
+
+static void vdpasim_worker_change_mm_sync(struct vdpasim *vdpasim,
+					  struct vdpasim_mm_work *mm_work)
+{
+	struct kthread_work *work = &mm_work->work;
+
+	kthread_init_work(work, vdpasim_mm_work_fn);
+	kthread_queue_work(vdpasim->worker, work);
+
+	kthread_flush_work(work);
+}
+
+static struct vdpasim *vdpa_to_sim(struct vdpa_device *vdpa)
+{
+	return container_of(vdpa, struct vdpasim, vdpa);
+}
+
+static void vdpasim_vq_notify(struct vringh *vring)
+{
+	struct vdpasim_virtqueue *vq =
+		container_of(vring, struct vdpasim_virtqueue, vring);
+
+	if (!vq->cb)
+		return;
+
+	vq->cb(vq->private);
+}
+
+static void vdpasim_queue_ready(struct vdpasim *vdpasim, unsigned int idx)
+{
+	struct vdpasim_virtqueue *vq = &vdpasim->vqs[idx];
+	uint16_t last_avail_idx = vq->vring.last_avail_idx;
+	struct vring_desc *desc = (struct vring_desc *)
+				  (uintptr_t)vq->desc_addr;
+	struct vring_avail *avail = (struct vring_avail *)
+				    (uintptr_t)vq->driver_addr;
+	struct vring_used *used = (struct vring_used *)
+				  (uintptr_t)vq->device_addr;
+
+	if (use_va && vdpasim->mm_bound) {
+		vringh_init_iotlb_va(&vq->vring, vdpasim->features, vq->num,
+				     true, desc, avail, used);
+	} else {
+		vringh_init_iotlb(&vq->vring, vdpasim->features, vq->num,
+				  true, desc, avail, used);
+	}
+
+	vq->vring.last_avail_idx = last_avail_idx;
+
+	/*
+	 * Since vdpa_sim does not support receive inflight descriptors as a
+	 * destination of a migration, let's set both avail_idx and used_idx
+	 * the same at vq start.  This is how vhost-user works in a
+	 * VHOST_SET_VRING_BASE call.
+	 *
+	 * Although the simple fix is to set last_used_idx at
+	 * vdpasim_set_vq_state, it would be reset at vdpasim_queue_ready.
+	 */
+	vq->vring.last_used_idx = last_avail_idx;
+	vq->vring.notify = vdpasim_vq_notify;
+}
+
+static void vdpasim_vq_reset(struct vdpasim *vdpasim,
+			     struct vdpasim_virtqueue *vq)
+{
+	vq->ready = false;
+	vq->desc_addr = 0;
+	vq->driver_addr = 0;
+	vq->device_addr = 0;
+	vq->cb = NULL;
+	vq->private = NULL;
+	vringh_init_iotlb(&vq->vring, vdpasim->dev_attr.supported_features,
+			  VDPASIM_QUEUE_MAX, false, NULL, NULL, NULL);
+
+	vq->vring.notify = NULL;
+}
+
+static void vdpasim_do_reset(struct vdpasim *vdpasim, u32 flags)
+{
+	int i;
+
+	spin_lock(&vdpasim->iommu_lock);
+
+	for (i = 0; i < vdpasim->dev_attr.nvqs; i++) {
+		vdpasim_vq_reset(vdpasim, &vdpasim->vqs[i]);
+		vringh_set_iotlb(&vdpasim->vqs[i].vring, &vdpasim->iommu[0],
+				 &vdpasim->iommu_lock);
+	}
+
+	if (flags & VDPA_RESET_F_CLEAN_MAP) {
+		for (i = 0; i < vdpasim->dev_attr.nas; i++) {
+			vhost_iotlb_reset(&vdpasim->iommu[i]);
+			vhost_iotlb_add_range(&vdpasim->iommu[i], 0, ULONG_MAX,
+					      0, VHOST_MAP_RW);
+			vdpasim->iommu_pt[i] = true;
+		}
+	}
+
+	vdpasim->running = false;
+	spin_unlock(&vdpasim->iommu_lock);
+
+	vdpasim->features = 0;
+	vdpasim->status = 0;
+	++vdpasim->generation;
+}
+
+static const struct vdpa_config_ops vdpasim_config_ops;
+static const struct vdpa_config_ops vdpasim_batch_config_ops;
+
+static void vdpasim_work_fn(struct kthread_work *work)
+{
+	struct vdpasim *vdpasim = container_of(work, struct vdpasim, work);
+	struct mm_struct *mm = vdpasim->mm_bound;
+
+	if (use_va && mm) {
+		if (!mmget_not_zero(mm))
+			return;
+		kthread_use_mm(mm);
+	}
+
+	vdpasim->dev_attr.work_fn(vdpasim);
+
+	if (use_va && mm) {
+		kthread_unuse_mm(mm);
+		mmput(mm);
+	}
+}
+
+struct vdpasim *vdpasim_create(struct vdpasim_dev_attr *dev_attr,
+			       const struct vdpa_dev_set_config *config)
+{
+	const struct vdpa_config_ops *ops;
+	struct vdpa_device *vdpa;
+	struct vdpasim *vdpasim;
+	struct device *dev;
+	int i, ret = -ENOMEM;
+
+	if (!dev_attr->alloc_size)
+		return ERR_PTR(-EINVAL);
+	if (max_iotlb_entries < 2)
+		return ERR_PTR(-EINVAL);
+
+	if (config->mask & BIT_ULL(VDPA_ATTR_DEV_FEATURES)) {
+		if (config->device_features &
+		    ~dev_attr->supported_features)
+			return ERR_PTR(-EINVAL);
+		dev_attr->supported_features =
+			config->device_features;
+	}
+
+	if (batch_mapping)
+		ops = &vdpasim_batch_config_ops;
+	else
+		ops = &vdpasim_config_ops;
+
+	vdpa = __vdpa_alloc_device(NULL, ops, NULL,
+				   dev_attr->ngroups, dev_attr->nas,
+				   dev_attr->alloc_size,
+				   dev_attr->name, use_va);
+	if (IS_ERR(vdpa)) {
+		ret = PTR_ERR(vdpa);
+		goto err_alloc;
+	}
+
+	vdpasim = vdpa_to_sim(vdpa);
+	vdpasim->dev_attr = *dev_attr;
+	dev = &vdpasim->vdpa.dev;
+
+	kthread_init_work(&vdpasim->work, vdpasim_work_fn);
+	vdpasim->worker = kthread_run_worker(0, "vDPA sim worker: %s",
+						dev_attr->name);
+	if (IS_ERR(vdpasim->worker))
+		goto err_iommu;
+
+	mutex_init(&vdpasim->mutex);
+	spin_lock_init(&vdpasim->iommu_lock);
+
+	dev->dma_mask = &dev->coherent_dma_mask;
+	if (dma_set_mask_and_coherent(dev, DMA_BIT_MASK(64)))
+		goto err_iommu;
+	vdpasim->vdpa.mdev = dev_attr->mgmt_dev;
+
+	vdpasim->config = kzalloc(dev_attr->config_size, GFP_KERNEL);
+	if (!vdpasim->config)
+		goto err_iommu;
+
+	vdpasim->vqs = kzalloc_objs(struct vdpasim_virtqueue, dev_attr->nvqs);
+	if (!vdpasim->vqs)
+		goto err_iommu;
+
+	vdpasim->iommu = kmalloc_objs(*vdpasim->iommu, vdpasim->dev_attr.nas);
+	if (!vdpasim->iommu)
+		goto err_iommu;
+
+	vdpasim->iommu_pt = kmalloc_objs(*vdpasim->iommu_pt,
+					 vdpasim->dev_attr.nas);
+	if (!vdpasim->iommu_pt)
+		goto err_iommu;
+
+	for (i = 0; i < vdpasim->dev_attr.nas; i++) {
+		vhost_iotlb_init(&vdpasim->iommu[i], max_iotlb_entries, 0);
+		ret = vhost_iotlb_add_range(&vdpasim->iommu[i], 0, ULONG_MAX,
+					    0, VHOST_MAP_RW);
+		if (ret)
+			goto err_iommu;
+		vdpasim->iommu_pt[i] = true;
+	}
+
+	for (i = 0; i < dev_attr->nvqs; i++)
+		vringh_set_iotlb(&vdpasim->vqs[i].vring, &vdpasim->iommu[0],
+				 &vdpasim->iommu_lock);
+
+	vdpasim->vdpa.vmap.dma_dev = dev;
+
+	return vdpasim;
+
+err_iommu:
+	put_device(dev);
+err_alloc:
+	return ERR_PTR(ret);
+}
+EXPORT_SYMBOL_GPL(vdpasim_create);
+
+void vdpasim_schedule_work(struct vdpasim *vdpasim)
+{
+	kthread_queue_work(vdpasim->worker, &vdpasim->work);
+}
+EXPORT_SYMBOL_GPL(vdpasim_schedule_work);
+
+static int vdpasim_set_vq_address(struct vdpa_device *vdpa, u16 idx,
+				  u64 desc_area, u64 driver_area,
+				  u64 device_area)
+{
+	struct vdpasim *vdpasim = vdpa_to_sim(vdpa);
+	struct vdpasim_virtqueue *vq = &vdpasim->vqs[idx];
+
+	vq->desc_addr = desc_area;
+	vq->driver_addr = driver_area;
+	vq->device_addr = device_area;
+
+	return 0;
+}
+
+static void vdpasim_set_vq_num(struct vdpa_device *vdpa, u16 idx, u32 num)
+{
+	struct vdpasim *vdpasim = vdpa_to_sim(vdpa);
+	struct vdpasim_virtqueue *vq = &vdpasim->vqs[idx];
+
+	vq->num = num;
+}
+
+static u16 vdpasim_get_vq_size(struct vdpa_device *vdpa, u16 idx)
+{
+	struct vdpasim *vdpasim = vdpa_to_sim(vdpa);
+	struct vdpasim_virtqueue *vq = &vdpasim->vqs[idx];
+
+	if (vdpasim->status & VIRTIO_CONFIG_S_DRIVER_OK)
+		return vq->num;
+	else
+		return VDPASIM_QUEUE_MAX;
+}
+
+static void vdpasim_kick_vq(struct vdpa_device *vdpa, u16 idx)
+{
+	struct vdpasim *vdpasim = vdpa_to_sim(vdpa);
+	struct vdpasim_virtqueue *vq = &vdpasim->vqs[idx];
+
+	if (!vdpasim->running &&
+	    (vdpasim->status & VIRTIO_CONFIG_S_DRIVER_OK)) {
+		vdpasim->pending_kick = true;
+		return;
+	}
+
+	if (vq->ready)
+		vdpasim_schedule_work(vdpasim);
+}
+
+static void vdpasim_set_vq_cb(struct vdpa_device *vdpa, u16 idx,
+			      struct vdpa_callback *cb)
+{
+	struct vdpasim *vdpasim = vdpa_to_sim(vdpa);
+	struct vdpasim_virtqueue *vq = &vdpasim->vqs[idx];
+
+	vq->cb = cb->callback;
+	vq->private = cb->private;
+}
+
+static void vdpasim_set_vq_ready(struct vdpa_device *vdpa, u16 idx, bool ready)
+{
+	struct vdpasim *vdpasim = vdpa_to_sim(vdpa);
+	struct vdpasim_virtqueue *vq = &vdpasim->vqs[idx];
+	bool old_ready;
+
+	mutex_lock(&vdpasim->mutex);
+	old_ready = vq->ready;
+	vq->ready = ready;
+	if (vq->ready && !old_ready) {
+		vdpasim_queue_ready(vdpasim, idx);
+	}
+	mutex_unlock(&vdpasim->mutex);
+}
+
+static bool vdpasim_get_vq_ready(struct vdpa_device *vdpa, u16 idx)
+{
+	struct vdpasim *vdpasim = vdpa_to_sim(vdpa);
+	struct vdpasim_virtqueue *vq = &vdpasim->vqs[idx];
+
+	return vq->ready;
+}
+
+static int vdpasim_set_vq_state(struct vdpa_device *vdpa, u16 idx,
+				const struct vdpa_vq_state *state)
+{
+	struct vdpasim *vdpasim = vdpa_to_sim(vdpa);
+	struct vdpasim_virtqueue *vq = &vdpasim->vqs[idx];
+	struct vringh *vrh = &vq->vring;
+
+	mutex_lock(&vdpasim->mutex);
+	vrh->last_avail_idx = state->split.avail_index;
+	mutex_unlock(&vdpasim->mutex);
+
+	return 0;
+}
+
+static int vdpasim_get_vq_state(struct vdpa_device *vdpa, u16 idx,
+				struct vdpa_vq_state *state)
+{
+	struct vdpasim *vdpasim = vdpa_to_sim(vdpa);
+	struct vdpasim_virtqueue *vq = &vdpasim->vqs[idx];
+	struct vringh *vrh = &vq->vring;
+
+	state->split.avail_index = vrh->last_avail_idx;
+	return 0;
+}
+
+static int vdpasim_get_vq_stats(struct vdpa_device *vdpa, u16 idx,
+				struct sk_buff *msg,
+				struct netlink_ext_ack *extack)
+{
+	struct vdpasim *vdpasim = vdpa_to_sim(vdpa);
+
+	if (vdpasim->dev_attr.get_stats)
+		return vdpasim->dev_attr.get_stats(vdpasim, idx,
+						   msg, extack);
+	return -EOPNOTSUPP;
+}
+
+static u32 vdpasim_get_vq_align(struct vdpa_device *vdpa)
+{
+	return VDPASIM_QUEUE_ALIGN;
+}
+
+static u32 vdpasim_get_vq_group(struct vdpa_device *vdpa, u16 idx)
+{
+	/* RX and TX belongs to group 0, CVQ belongs to group 1 */
+	if (idx == 2)
+		return 1;
+	else
+		return 0;
+}
+
+static u64 vdpasim_get_device_features(struct vdpa_device *vdpa)
+{
+	struct vdpasim *vdpasim = vdpa_to_sim(vdpa);
+
+	return vdpasim->dev_attr.supported_features;
+}
+
+static u64 vdpasim_get_backend_features(const struct vdpa_device *vdpa)
+{
+	return BIT_ULL(VHOST_BACKEND_F_ENABLE_AFTER_DRIVER_OK);
+}
+
+static int vdpasim_set_driver_features(struct vdpa_device *vdpa, u64 features)
+{
+	struct vdpasim *vdpasim = vdpa_to_sim(vdpa);
+
+	/* DMA mapping must be done by driver */
+	if (!(features & (1ULL << VIRTIO_F_ACCESS_PLATFORM)))
+		return -EINVAL;
+
+	vdpasim->features = features & vdpasim->dev_attr.supported_features;
+
+	return 0;
+}
+
+static u64 vdpasim_get_driver_features(struct vdpa_device *vdpa)
+{
+	struct vdpasim *vdpasim = vdpa_to_sim(vdpa);
+
+	return vdpasim->features;
+}
+
+static void vdpasim_set_config_cb(struct vdpa_device *vdpa,
+				  struct vdpa_callback *cb)
+{
+	/* We don't support config interrupt */
+}
+
+static u16 vdpasim_get_vq_num_max(struct vdpa_device *vdpa)
+{
+	return VDPASIM_QUEUE_MAX;
+}
+
+static u32 vdpasim_get_device_id(struct vdpa_device *vdpa)
+{
+	struct vdpasim *vdpasim = vdpa_to_sim(vdpa);
+
+	return vdpasim->dev_attr.id;
+}
+
+static u32 vdpasim_get_vendor_id(struct vdpa_device *vdpa)
+{
+	return VDPASIM_VENDOR_ID;
+}
+
+static u8 vdpasim_get_status(struct vdpa_device *vdpa)
+{
+	struct vdpasim *vdpasim = vdpa_to_sim(vdpa);
+	u8 status;
+
+	mutex_lock(&vdpasim->mutex);
+	status = vdpasim->status;
+	mutex_unlock(&vdpasim->mutex);
+
+	return status;
+}
+
+static void vdpasim_set_status(struct vdpa_device *vdpa, u8 status)
+{
+	struct vdpasim *vdpasim = vdpa_to_sim(vdpa);
+
+	mutex_lock(&vdpasim->mutex);
+	vdpasim->status = status;
+	vdpasim->running = (status & VIRTIO_CONFIG_S_DRIVER_OK) != 0;
+	mutex_unlock(&vdpasim->mutex);
+}
+
+static int vdpasim_compat_reset(struct vdpa_device *vdpa, u32 flags)
+{
+	struct vdpasim *vdpasim = vdpa_to_sim(vdpa);
+
+	mutex_lock(&vdpasim->mutex);
+	vdpasim->status = 0;
+	vdpasim_do_reset(vdpasim, flags);
+	mutex_unlock(&vdpasim->mutex);
+
+	return 0;
+}
+
+static int vdpasim_reset(struct vdpa_device *vdpa)
+{
+	return vdpasim_compat_reset(vdpa, 0);
+}
+
+static int vdpasim_suspend(struct vdpa_device *vdpa)
+{
+	struct vdpasim *vdpasim = vdpa_to_sim(vdpa);
+
+	mutex_lock(&vdpasim->mutex);
+	vdpasim->running = false;
+	mutex_unlock(&vdpasim->mutex);
+
+	return 0;
+}
+
+static int vdpasim_resume(struct vdpa_device *vdpa)
+{
+	struct vdpasim *vdpasim = vdpa_to_sim(vdpa);
+	int i;
+
+	mutex_lock(&vdpasim->mutex);
+	vdpasim->running = true;
+
+	if (vdpasim->pending_kick) {
+		/* Process pending descriptors */
+		for (i = 0; i < vdpasim->dev_attr.nvqs; ++i)
+			vdpasim_kick_vq(vdpa, i);
+
+		vdpasim->pending_kick = false;
+	}
+
+	mutex_unlock(&vdpasim->mutex);
+
+	return 0;
+}
+
+static size_t vdpasim_get_config_size(struct vdpa_device *vdpa)
+{
+	struct vdpasim *vdpasim = vdpa_to_sim(vdpa);
+
+	return vdpasim->dev_attr.config_size;
+}
+
+static void vdpasim_get_config(struct vdpa_device *vdpa, unsigned int offset,
+			     void *buf, unsigned int len)
+{
+	struct vdpasim *vdpasim = vdpa_to_sim(vdpa);
+
+	if (offset + len > vdpasim->dev_attr.config_size)
+		return;
+
+	if (vdpasim->dev_attr.get_config)
+		vdpasim->dev_attr.get_config(vdpasim, vdpasim->config);
+
+	memcpy(buf, vdpasim->config + offset, len);
+}
+
+static void vdpasim_set_config(struct vdpa_device *vdpa, unsigned int offset,
+			     const void *buf, unsigned int len)
+{
+	struct vdpasim *vdpasim = vdpa_to_sim(vdpa);
+
+	if (offset + len > vdpasim->dev_attr.config_size)
+		return;
+
+	memcpy(vdpasim->config + offset, buf, len);
+
+	if (vdpasim->dev_attr.set_config)
+		vdpasim->dev_attr.set_config(vdpasim, vdpasim->config);
+}
+
+static u32 vdpasim_get_generation(struct vdpa_device *vdpa)
+{
+	struct vdpasim *vdpasim = vdpa_to_sim(vdpa);
+
+	return vdpasim->generation;
+}
+
+static struct vdpa_iova_range vdpasim_get_iova_range(struct vdpa_device *vdpa)
+{
+	struct vdpa_iova_range range = {
+		.first = 0ULL,
+		.last = ULLONG_MAX,
+	};
+
+	return range;
+}
+
+static int vdpasim_set_group_asid(struct vdpa_device *vdpa, unsigned int group,
+				  unsigned int asid)
+{
+	struct vdpasim *vdpasim = vdpa_to_sim(vdpa);
+	struct vhost_iotlb *iommu;
+	int i;
+
+	iommu = &vdpasim->iommu[asid];
+
+	mutex_lock(&vdpasim->mutex);
+
+	for (i = 0; i < vdpasim->dev_attr.nvqs; i++)
+		if (vdpasim_get_vq_group(vdpa, i) == group)
+			vringh_set_iotlb(&vdpasim->vqs[i].vring, iommu,
+					 &vdpasim->iommu_lock);
+
+	mutex_unlock(&vdpasim->mutex);
+
+	return 0;
+}
+
+static int vdpasim_set_map(struct vdpa_device *vdpa, unsigned int asid,
+			   struct vhost_iotlb *iotlb)
+{
+	struct vdpasim *vdpasim = vdpa_to_sim(vdpa);
+	struct vhost_iotlb_map *map;
+	struct vhost_iotlb *iommu;
+	u64 start = 0ULL, last = 0ULL - 1;
+	int ret;
+
+	if (asid >= vdpasim->dev_attr.nas)
+		return -EINVAL;
+
+	spin_lock(&vdpasim->iommu_lock);
+
+	iommu = &vdpasim->iommu[asid];
+	vhost_iotlb_reset(iommu);
+	vdpasim->iommu_pt[asid] = false;
+
+	for (map = vhost_iotlb_itree_first(iotlb, start, last); map;
+	     map = vhost_iotlb_itree_next(map, start, last)) {
+		ret = vhost_iotlb_add_range(iommu, map->start,
+					    map->last, map->addr, map->perm);
+		if (ret)
+			goto err;
+	}
+	spin_unlock(&vdpasim->iommu_lock);
+	return 0;
+
+err:
+	vhost_iotlb_reset(iommu);
+	spin_unlock(&vdpasim->iommu_lock);
+	return ret;
+}
+
+static int vdpasim_reset_map(struct vdpa_device *vdpa, unsigned int asid)
+{
+	struct vdpasim *vdpasim = vdpa_to_sim(vdpa);
+
+	if (asid >= vdpasim->dev_attr.nas)
+		return -EINVAL;
+
+	spin_lock(&vdpasim->iommu_lock);
+	if (vdpasim->iommu_pt[asid])
+		goto out;
+	vhost_iotlb_reset(&vdpasim->iommu[asid]);
+	vhost_iotlb_add_range(&vdpasim->iommu[asid], 0, ULONG_MAX,
+			      0, VHOST_MAP_RW);
+	vdpasim->iommu_pt[asid] = true;
+out:
+	spin_unlock(&vdpasim->iommu_lock);
+	return 0;
+}
+
+static int vdpasim_bind_mm(struct vdpa_device *vdpa, struct mm_struct *mm)
+{
+	struct vdpasim *vdpasim = vdpa_to_sim(vdpa);
+	struct vdpasim_mm_work mm_work;
+
+	mm_work.vdpasim = vdpasim;
+	mm_work.mm_to_bind = mm;
+
+	vdpasim_worker_change_mm_sync(vdpasim, &mm_work);
+
+	return mm_work.ret;
+}
+
+static void vdpasim_unbind_mm(struct vdpa_device *vdpa)
+{
+	struct vdpasim *vdpasim = vdpa_to_sim(vdpa);
+	struct vdpasim_mm_work mm_work;
+
+	mm_work.vdpasim = vdpasim;
+	mm_work.mm_to_bind = NULL;
+
+	vdpasim_worker_change_mm_sync(vdpasim, &mm_work);
+}
+
+static int vdpasim_dma_map(struct vdpa_device *vdpa, unsigned int asid,
+			   u64 iova, u64 size,
+			   u64 pa, u32 perm, void *opaque)
+{
+	struct vdpasim *vdpasim = vdpa_to_sim(vdpa);
+	int ret;
+
+	if (asid >= vdpasim->dev_attr.nas)
+		return -EINVAL;
+
+	spin_lock(&vdpasim->iommu_lock);
+	if (vdpasim->iommu_pt[asid]) {
+		vhost_iotlb_reset(&vdpasim->iommu[asid]);
+		vdpasim->iommu_pt[asid] = false;
+	}
+	ret = vhost_iotlb_add_range_ctx(&vdpasim->iommu[asid], iova,
+					iova + size - 1, pa, perm, opaque);
+	spin_unlock(&vdpasim->iommu_lock);
+
+	return ret;
+}
+
+static int vdpasim_dma_unmap(struct vdpa_device *vdpa, unsigned int asid,
+			     u64 iova, u64 size)
+{
+	struct vdpasim *vdpasim = vdpa_to_sim(vdpa);
+
+	if (asid >= vdpasim->dev_attr.nas)
+		return -EINVAL;
+
+	if (vdpasim->iommu_pt[asid]) {
+		vhost_iotlb_reset(&vdpasim->iommu[asid]);
+		vdpasim->iommu_pt[asid] = false;
+	}
+
+	spin_lock(&vdpasim->iommu_lock);
+	vhost_iotlb_del_range(&vdpasim->iommu[asid], iova, iova + size - 1);
+	spin_unlock(&vdpasim->iommu_lock);
+
+	return 0;
+}
+
+static void vdpasim_free(struct vdpa_device *vdpa)
+{
+	struct vdpasim *vdpasim = vdpa_to_sim(vdpa);
+	int i;
+
+	kthread_cancel_work_sync(&vdpasim->work);
+	kthread_destroy_worker(vdpasim->worker);
+
+	for (i = 0; i < vdpasim->dev_attr.nvqs; i++) {
+		vringh_kiov_cleanup(&vdpasim->vqs[i].out_iov);
+		vringh_kiov_cleanup(&vdpasim->vqs[i].in_iov);
+	}
+
+	vdpasim->dev_attr.free(vdpasim);
+
+	for (i = 0; i < vdpasim->dev_attr.nas; i++)
+		vhost_iotlb_reset(&vdpasim->iommu[i]);
+	kfree(vdpasim->iommu);
+	kfree(vdpasim->iommu_pt);
+	kfree(vdpasim->vqs);
+	kfree(vdpasim->config);
+}
+
+static const struct vdpa_config_ops vdpasim_config_ops = {
+	.set_vq_address         = vdpasim_set_vq_address,
+	.set_vq_num             = vdpasim_set_vq_num,
+	.kick_vq                = vdpasim_kick_vq,
+	.set_vq_cb              = vdpasim_set_vq_cb,
+	.set_vq_ready           = vdpasim_set_vq_ready,
+	.get_vq_ready           = vdpasim_get_vq_ready,
+	.set_vq_state           = vdpasim_set_vq_state,
+	.get_vendor_vq_stats    = vdpasim_get_vq_stats,
+	.get_vq_state           = vdpasim_get_vq_state,
+	.get_vq_align           = vdpasim_get_vq_align,
+	.get_vq_group           = vdpasim_get_vq_group,
+	.get_device_features    = vdpasim_get_device_features,
+	.get_backend_features   = vdpasim_get_backend_features,
+	.set_driver_features    = vdpasim_set_driver_features,
+	.get_driver_features    = vdpasim_get_driver_features,
+	.set_config_cb          = vdpasim_set_config_cb,
+	.get_vq_num_max         = vdpasim_get_vq_num_max,
+	.get_vq_size		= vdpasim_get_vq_size,
+	.get_device_id          = vdpasim_get_device_id,
+	.get_vendor_id          = vdpasim_get_vendor_id,
+	.get_status             = vdpasim_get_status,
+	.set_status             = vdpasim_set_status,
+	.reset			= vdpasim_reset,
+	.compat_reset		= vdpasim_compat_reset,
+	.suspend		= vdpasim_suspend,
+	.resume			= vdpasim_resume,
+	.get_config_size        = vdpasim_get_config_size,
+	.get_config             = vdpasim_get_config,
+	.set_config             = vdpasim_set_config,
+	.get_generation         = vdpasim_get_generation,
+	.get_iova_range         = vdpasim_get_iova_range,
+	.set_group_asid         = vdpasim_set_group_asid,
+	.dma_map                = vdpasim_dma_map,
+	.dma_unmap              = vdpasim_dma_unmap,
+	.reset_map              = vdpasim_reset_map,
+	.bind_mm		= vdpasim_bind_mm,
+	.unbind_mm		= vdpasim_unbind_mm,
+	.free                   = vdpasim_free,
+};
+
+static const struct vdpa_config_ops vdpasim_batch_config_ops = {
+	.set_vq_address         = vdpasim_set_vq_address,
+	.set_vq_num             = vdpasim_set_vq_num,
+	.kick_vq                = vdpasim_kick_vq,
+	.set_vq_cb              = vdpasim_set_vq_cb,
+	.set_vq_ready           = vdpasim_set_vq_ready,
+	.get_vq_ready           = vdpasim_get_vq_ready,
+	.set_vq_state           = vdpasim_set_vq_state,
+	.get_vendor_vq_stats    = vdpasim_get_vq_stats,
+	.get_vq_state           = vdpasim_get_vq_state,
+	.get_vq_align           = vdpasim_get_vq_align,
+	.get_vq_group           = vdpasim_get_vq_group,
+	.get_device_features    = vdpasim_get_device_features,
+	.get_backend_features   = vdpasim_get_backend_features,
+	.set_driver_features    = vdpasim_set_driver_features,
+	.get_driver_features    = vdpasim_get_driver_features,
+	.set_config_cb          = vdpasim_set_config_cb,
+	.get_vq_num_max         = vdpasim_get_vq_num_max,
+	.get_device_id          = vdpasim_get_device_id,
+	.get_vendor_id          = vdpasim_get_vendor_id,
+	.get_status             = vdpasim_get_status,
+	.set_status             = vdpasim_set_status,
+	.reset			= vdpasim_reset,
+	.compat_reset		= vdpasim_compat_reset,
+	.suspend		= vdpasim_suspend,
+	.resume			= vdpasim_resume,
+	.get_config_size        = vdpasim_get_config_size,
+	.get_config             = vdpasim_get_config,
+	.set_config             = vdpasim_set_config,
+	.get_generation         = vdpasim_get_generation,
+	.get_iova_range         = vdpasim_get_iova_range,
+	.set_group_asid         = vdpasim_set_group_asid,
+	.set_map                = vdpasim_set_map,
+	.reset_map              = vdpasim_reset_map,
+	.bind_mm		= vdpasim_bind_mm,
+	.unbind_mm		= vdpasim_unbind_mm,
+	.free                   = vdpasim_free,
+};
+
+MODULE_VERSION(DRV_VERSION);
+MODULE_LICENSE(DRV_LICENSE);
+MODULE_AUTHOR(DRV_AUTHOR);
+MODULE_DESCRIPTION(DRV_DESC);

linux-7.2.3/drivers/vdpa/vdpa_sim/vdpa_sim.h

@@ -0,0 +1,121 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * Copyright (c) 2020, Red Hat Inc. All rights reserved.
+ */
+
+#ifndef _VDPA_SIM_H
+#define _VDPA_SIM_H
+
+#include <linux/iova.h>
+#include <linux/vringh.h>
+#include <linux/vdpa.h>
+#include <linux/virtio_byteorder.h>
+#include <linux/vhost_iotlb.h>
+#include <uapi/linux/virtio_config.h>
+
+#define VDPASIM_FEATURES	((1ULL << VIRTIO_F_ANY_LAYOUT) | \
+				 (1ULL << VIRTIO_F_VERSION_1)  | \
+				 (1ULL << VIRTIO_F_ACCESS_PLATFORM))
+
+struct vdpasim;
+
+struct vdpasim_virtqueue {
+	struct vringh vring;
+	struct vringh_kiov in_iov;
+	struct vringh_kiov out_iov;
+	unsigned short head;
+	bool ready;
+	u64 desc_addr;
+	u64 device_addr;
+	u64 driver_addr;
+	u32 num;
+	void *private;
+	irqreturn_t (*cb)(void *data);
+};
+
+struct vdpasim_dev_attr {
+	struct vdpa_mgmt_dev *mgmt_dev;
+	const char *name;
+	u64 supported_features;
+	size_t alloc_size;
+	size_t config_size;
+	int nvqs;
+	u32 id;
+	u32 ngroups;
+	u32 nas;
+
+	void (*work_fn)(struct vdpasim *vdpasim);
+	void (*get_config)(struct vdpasim *vdpasim, void *config);
+	void (*set_config)(struct vdpasim *vdpasim, const void *config);
+	int (*get_stats)(struct vdpasim *vdpasim, u16 idx,
+			 struct sk_buff *msg,
+			 struct netlink_ext_ack *extack);
+	void (*free)(struct vdpasim *vdpasim);
+};
+
+/* State of each vdpasim device */
+struct vdpasim {
+	struct vdpa_device vdpa;
+	struct vdpasim_virtqueue *vqs;
+	struct kthread_worker *worker;
+	struct kthread_work work;
+	struct mm_struct *mm_bound;
+	struct vdpasim_dev_attr dev_attr;
+	/* mutex to synchronize virtqueue state */
+	struct mutex mutex;
+	/* virtio config according to device type */
+	void *config;
+	struct vhost_iotlb *iommu;
+	bool *iommu_pt;
+	u32 status;
+	u32 generation;
+	u64 features;
+	u32 groups;
+	bool running;
+	bool pending_kick;
+	/* spinlock to synchronize iommu table */
+	spinlock_t iommu_lock;
+};
+
+struct vdpasim *vdpasim_create(struct vdpasim_dev_attr *attr,
+			       const struct vdpa_dev_set_config *config);
+void vdpasim_schedule_work(struct vdpasim *vdpasim);
+
+/* TODO: cross-endian support */
+static inline bool vdpasim_is_little_endian(struct vdpasim *vdpasim)
+{
+	return virtio_legacy_is_little_endian() ||
+		(vdpasim->features & (1ULL << VIRTIO_F_VERSION_1));
+}
+
+static inline u16 vdpasim16_to_cpu(struct vdpasim *vdpasim, __virtio16 val)
+{
+	return __virtio16_to_cpu(vdpasim_is_little_endian(vdpasim), val);
+}
+
+static inline __virtio16 cpu_to_vdpasim16(struct vdpasim *vdpasim, u16 val)
+{
+	return __cpu_to_virtio16(vdpasim_is_little_endian(vdpasim), val);
+}
+
+static inline u32 vdpasim32_to_cpu(struct vdpasim *vdpasim, __virtio32 val)
+{
+	return __virtio32_to_cpu(vdpasim_is_little_endian(vdpasim), val);
+}
+
+static inline __virtio32 cpu_to_vdpasim32(struct vdpasim *vdpasim, u32 val)
+{
+	return __cpu_to_virtio32(vdpasim_is_little_endian(vdpasim), val);
+}
+
+static inline u64 vdpasim64_to_cpu(struct vdpasim *vdpasim, __virtio64 val)
+{
+	return __virtio64_to_cpu(vdpasim_is_little_endian(vdpasim), val);
+}
+
+static inline __virtio64 cpu_to_vdpasim64(struct vdpasim *vdpasim, u64 val)
+{
+	return __cpu_to_virtio64(vdpasim_is_little_endian(vdpasim), val);
+}
+
+#endif

linux-7.2.3/drivers/vdpa/vdpa_sim/vdpa_sim_blk.c

@@ -0,0 +1,519 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * VDPA simulator for block device.
+ *
+ * Copyright (c) 2020, NVIDIA CORPORATION. All rights reserved.
+ * Copyright (c) 2021, Red Hat Inc. All rights reserved.
+ *
+ */
+
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/device.h>
+#include <linux/kernel.h>
+#include <linux/blkdev.h>
+#include <linux/vringh.h>
+#include <linux/vdpa.h>
+#include <uapi/linux/virtio_blk.h>
+
+#include "vdpa_sim.h"
+
+#define DRV_VERSION  "0.1"
+#define DRV_AUTHOR   "Max Gurtovoy <mgurtovoy@nvidia.com>"
+#define DRV_DESC     "vDPA Device Simulator for block device"
+#define DRV_LICENSE  "GPL v2"
+
+#define VDPASIM_BLK_FEATURES	(VDPASIM_FEATURES | \
+				 (1ULL << VIRTIO_BLK_F_FLUSH)    | \
+				 (1ULL << VIRTIO_BLK_F_SIZE_MAX) | \
+				 (1ULL << VIRTIO_BLK_F_SEG_MAX)  | \
+				 (1ULL << VIRTIO_BLK_F_BLK_SIZE) | \
+				 (1ULL << VIRTIO_BLK_F_TOPOLOGY) | \
+				 (1ULL << VIRTIO_BLK_F_MQ)       | \
+				 (1ULL << VIRTIO_BLK_F_DISCARD)  | \
+				 (1ULL << VIRTIO_BLK_F_WRITE_ZEROES))
+
+#define VDPASIM_BLK_CAPACITY	0x40000
+#define VDPASIM_BLK_SIZE_MAX	0x1000
+#define VDPASIM_BLK_SEG_MAX	32
+#define VDPASIM_BLK_DWZ_MAX_SECTORS UINT_MAX
+
+/* 1 virtqueue, 1 address space, 1 virtqueue group */
+#define VDPASIM_BLK_VQ_NUM	1
+#define VDPASIM_BLK_AS_NUM	1
+#define VDPASIM_BLK_GROUP_NUM	1
+
+struct vdpasim_blk {
+	struct vdpasim vdpasim;
+	void *buffer;
+	bool shared_backend;
+};
+
+static struct vdpasim_blk *sim_to_blk(struct vdpasim *vdpasim)
+{
+	return container_of(vdpasim, struct vdpasim_blk, vdpasim);
+}
+
+static char vdpasim_blk_id[VIRTIO_BLK_ID_BYTES] = "vdpa_blk_sim";
+
+static bool shared_backend;
+module_param(shared_backend, bool, 0444);
+MODULE_PARM_DESC(shared_backend, "Enable the shared backend between virtio-blk devices");
+
+static void *shared_buffer;
+/* mutex to synchronize shared_buffer access */
+static DEFINE_MUTEX(shared_buffer_mutex);
+
+static void vdpasim_blk_buffer_lock(struct vdpasim_blk *blk)
+{
+	if (blk->shared_backend)
+		mutex_lock(&shared_buffer_mutex);
+}
+
+static void vdpasim_blk_buffer_unlock(struct vdpasim_blk *blk)
+{
+	if (blk->shared_backend)
+		mutex_unlock(&shared_buffer_mutex);
+}
+
+static bool vdpasim_blk_check_range(struct vdpasim *vdpasim, u64 start_sector,
+				    u64 num_sectors, u64 max_sectors)
+{
+	if (start_sector > VDPASIM_BLK_CAPACITY) {
+		dev_dbg(&vdpasim->vdpa.dev,
+			"starting sector exceeds the capacity - start: 0x%llx capacity: 0x%x\n",
+			start_sector, VDPASIM_BLK_CAPACITY);
+	}
+
+	if (num_sectors > max_sectors) {
+		dev_dbg(&vdpasim->vdpa.dev,
+			"number of sectors exceeds the max allowed in a request - num: 0x%llx max: 0x%llx\n",
+			num_sectors, max_sectors);
+		return false;
+	}
+
+	if (num_sectors > VDPASIM_BLK_CAPACITY - start_sector) {
+		dev_dbg(&vdpasim->vdpa.dev,
+			"request exceeds the capacity - start: 0x%llx num: 0x%llx capacity: 0x%x\n",
+			start_sector, num_sectors, VDPASIM_BLK_CAPACITY);
+		return false;
+	}
+
+	return true;
+}
+
+/* Returns 'true' if the request is handled (with or without an I/O error)
+ * and the status is correctly written in the last byte of the 'in iov',
+ * 'false' otherwise.
+ */
+static bool vdpasim_blk_handle_req(struct vdpasim *vdpasim,
+				   struct vdpasim_virtqueue *vq)
+{
+	struct vdpasim_blk *blk = sim_to_blk(vdpasim);
+	size_t pushed = 0, to_pull, to_push;
+	struct virtio_blk_outhdr hdr;
+	bool handled = false;
+	ssize_t bytes;
+	loff_t offset;
+	u64 sector;
+	u8 status;
+	u32 type;
+	int ret;
+
+	ret = vringh_getdesc_iotlb(&vq->vring, &vq->out_iov, &vq->in_iov,
+				   &vq->head, GFP_ATOMIC);
+	if (ret != 1)
+		return false;
+
+	if (vq->out_iov.used < 1 || vq->in_iov.used < 1) {
+		dev_dbg(&vdpasim->vdpa.dev, "missing headers - out_iov: %u in_iov %u\n",
+			vq->out_iov.used, vq->in_iov.used);
+		goto err;
+	}
+
+	if (vq->in_iov.iov[vq->in_iov.used - 1].iov_len < 1) {
+		dev_dbg(&vdpasim->vdpa.dev, "request in header too short\n");
+		goto err;
+	}
+
+	/* The last byte is the status and we checked if the last iov has
+	 * enough room for it.
+	 */
+	to_push = vringh_kiov_length(&vq->in_iov) - 1;
+
+	to_pull = vringh_kiov_length(&vq->out_iov);
+
+	bytes = vringh_iov_pull_iotlb(&vq->vring, &vq->out_iov, &hdr,
+				      sizeof(hdr));
+	if (bytes != sizeof(hdr)) {
+		dev_dbg(&vdpasim->vdpa.dev, "request out header too short\n");
+		goto err;
+	}
+
+	to_pull -= bytes;
+
+	type = vdpasim32_to_cpu(vdpasim, hdr.type);
+	sector = vdpasim64_to_cpu(vdpasim, hdr.sector);
+	offset = sector << SECTOR_SHIFT;
+	status = VIRTIO_BLK_S_OK;
+
+	if (type != VIRTIO_BLK_T_IN && type != VIRTIO_BLK_T_OUT &&
+	    sector != 0) {
+		dev_dbg(&vdpasim->vdpa.dev,
+			"sector must be 0 for %u request - sector: 0x%llx\n",
+			type, sector);
+		status = VIRTIO_BLK_S_IOERR;
+		goto err_status;
+	}
+
+	switch (type) {
+	case VIRTIO_BLK_T_IN:
+		if (!vdpasim_blk_check_range(vdpasim, sector,
+					     to_push >> SECTOR_SHIFT,
+					     VDPASIM_BLK_SIZE_MAX * VDPASIM_BLK_SEG_MAX)) {
+			status = VIRTIO_BLK_S_IOERR;
+			break;
+		}
+
+		vdpasim_blk_buffer_lock(blk);
+		bytes = vringh_iov_push_iotlb(&vq->vring, &vq->in_iov,
+					      blk->buffer + offset, to_push);
+		vdpasim_blk_buffer_unlock(blk);
+		if (bytes < 0) {
+			dev_dbg(&vdpasim->vdpa.dev,
+				"vringh_iov_push_iotlb() error: %zd offset: 0x%llx len: 0x%zx\n",
+				bytes, offset, to_push);
+			status = VIRTIO_BLK_S_IOERR;
+			break;
+		}
+
+		pushed += bytes;
+		break;
+
+	case VIRTIO_BLK_T_OUT:
+		if (!vdpasim_blk_check_range(vdpasim, sector,
+					     to_pull >> SECTOR_SHIFT,
+					     VDPASIM_BLK_SIZE_MAX * VDPASIM_BLK_SEG_MAX)) {
+			status = VIRTIO_BLK_S_IOERR;
+			break;
+		}
+
+		vdpasim_blk_buffer_lock(blk);
+		bytes = vringh_iov_pull_iotlb(&vq->vring, &vq->out_iov,
+					      blk->buffer + offset, to_pull);
+		vdpasim_blk_buffer_unlock(blk);
+		if (bytes < 0) {
+			dev_dbg(&vdpasim->vdpa.dev,
+				"vringh_iov_pull_iotlb() error: %zd offset: 0x%llx len: 0x%zx\n",
+				bytes, offset, to_pull);
+			status = VIRTIO_BLK_S_IOERR;
+			break;
+		}
+		break;
+
+	case VIRTIO_BLK_T_GET_ID:
+		bytes = vringh_iov_push_iotlb(&vq->vring, &vq->in_iov,
+					      vdpasim_blk_id,
+					      VIRTIO_BLK_ID_BYTES);
+		if (bytes < 0) {
+			dev_dbg(&vdpasim->vdpa.dev,
+				"vringh_iov_push_iotlb() error: %zd\n", bytes);
+			status = VIRTIO_BLK_S_IOERR;
+			break;
+		}
+
+		pushed += bytes;
+		break;
+
+	case VIRTIO_BLK_T_FLUSH:
+		/* nothing to do */
+		break;
+
+	case VIRTIO_BLK_T_DISCARD:
+	case VIRTIO_BLK_T_WRITE_ZEROES: {
+		struct virtio_blk_discard_write_zeroes range;
+		u32 num_sectors, flags;
+
+		if (to_pull != sizeof(range)) {
+			dev_dbg(&vdpasim->vdpa.dev,
+				"discard/write_zeroes header len: 0x%zx [expected: 0x%zx]\n",
+				to_pull, sizeof(range));
+			status = VIRTIO_BLK_S_IOERR;
+			break;
+		}
+
+		bytes = vringh_iov_pull_iotlb(&vq->vring, &vq->out_iov, &range,
+					      to_pull);
+		if (bytes < 0) {
+			dev_dbg(&vdpasim->vdpa.dev,
+				"vringh_iov_pull_iotlb() error: %zd offset: 0x%llx len: 0x%zx\n",
+				bytes, offset, to_pull);
+			status = VIRTIO_BLK_S_IOERR;
+			break;
+		}
+
+		sector = le64_to_cpu(range.sector);
+		offset = sector << SECTOR_SHIFT;
+		num_sectors = le32_to_cpu(range.num_sectors);
+		flags = le32_to_cpu(range.flags);
+
+		if (type == VIRTIO_BLK_T_DISCARD && flags != 0) {
+			dev_dbg(&vdpasim->vdpa.dev,
+				"discard unexpected flags set - flags: 0x%x\n",
+				flags);
+			status = VIRTIO_BLK_S_UNSUPP;
+			break;
+		}
+
+		if (type == VIRTIO_BLK_T_WRITE_ZEROES &&
+		    flags & ~VIRTIO_BLK_WRITE_ZEROES_FLAG_UNMAP) {
+			dev_dbg(&vdpasim->vdpa.dev,
+				"write_zeroes unexpected flags set - flags: 0x%x\n",
+				flags);
+			status = VIRTIO_BLK_S_UNSUPP;
+			break;
+		}
+
+		if (!vdpasim_blk_check_range(vdpasim, sector, num_sectors,
+					     VDPASIM_BLK_DWZ_MAX_SECTORS)) {
+			status = VIRTIO_BLK_S_IOERR;
+			break;
+		}
+
+		if (type == VIRTIO_BLK_T_WRITE_ZEROES) {
+			vdpasim_blk_buffer_lock(blk);
+			memset(blk->buffer + offset, 0,
+			       num_sectors << SECTOR_SHIFT);
+			vdpasim_blk_buffer_unlock(blk);
+		}
+
+		break;
+	}
+	default:
+		dev_dbg(&vdpasim->vdpa.dev,
+			"Unsupported request type %d\n", type);
+		status = VIRTIO_BLK_S_IOERR;
+		break;
+	}
+
+err_status:
+	/* If some operations fail, we need to skip the remaining bytes
+	 * to put the status in the last byte
+	 */
+	if (to_push - pushed > 0)
+		vringh_kiov_advance(&vq->in_iov, to_push - pushed);
+
+	/* Last byte is the status */
+	bytes = vringh_iov_push_iotlb(&vq->vring, &vq->in_iov, &status, 1);
+	if (bytes != 1)
+		goto err;
+
+	pushed += bytes;
+
+	/* Make sure data is wrote before advancing index */
+	smp_wmb();
+
+	handled = true;
+
+err:
+	vringh_complete_iotlb(&vq->vring, vq->head, pushed);
+
+	return handled;
+}
+
+static void vdpasim_blk_work(struct vdpasim *vdpasim)
+{
+	bool reschedule = false;
+	int i;
+
+	mutex_lock(&vdpasim->mutex);
+
+	if (!(vdpasim->status & VIRTIO_CONFIG_S_DRIVER_OK))
+		goto out;
+
+	if (!vdpasim->running)
+		goto out;
+
+	for (i = 0; i < VDPASIM_BLK_VQ_NUM; i++) {
+		struct vdpasim_virtqueue *vq = &vdpasim->vqs[i];
+		int reqs = 0;
+
+		if (!vq->ready)
+			continue;
+
+		while (vdpasim_blk_handle_req(vdpasim, vq)) {
+			/* 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();
+
+			if (++reqs > 4) {
+				reschedule = true;
+				break;
+			}
+		}
+	}
+out:
+	mutex_unlock(&vdpasim->mutex);
+
+	if (reschedule)
+		vdpasim_schedule_work(vdpasim);
+}
+
+static void vdpasim_blk_get_config(struct vdpasim *vdpasim, void *config)
+{
+	struct virtio_blk_config *blk_config = config;
+
+	memset(config, 0, sizeof(struct virtio_blk_config));
+
+	blk_config->capacity = cpu_to_vdpasim64(vdpasim, VDPASIM_BLK_CAPACITY);
+	blk_config->size_max = cpu_to_vdpasim32(vdpasim, VDPASIM_BLK_SIZE_MAX);
+	blk_config->seg_max = cpu_to_vdpasim32(vdpasim, VDPASIM_BLK_SEG_MAX);
+	blk_config->num_queues = cpu_to_vdpasim16(vdpasim, VDPASIM_BLK_VQ_NUM);
+	blk_config->min_io_size = cpu_to_vdpasim16(vdpasim, 1);
+	blk_config->opt_io_size = cpu_to_vdpasim32(vdpasim, 1);
+	blk_config->blk_size = cpu_to_vdpasim32(vdpasim, SECTOR_SIZE);
+	/* VIRTIO_BLK_F_DISCARD */
+	blk_config->discard_sector_alignment =
+		cpu_to_vdpasim32(vdpasim, SECTOR_SIZE);
+	blk_config->max_discard_sectors =
+		cpu_to_vdpasim32(vdpasim, VDPASIM_BLK_DWZ_MAX_SECTORS);
+	blk_config->max_discard_seg = cpu_to_vdpasim32(vdpasim, 1);
+	/* VIRTIO_BLK_F_WRITE_ZEROES */
+	blk_config->max_write_zeroes_sectors =
+		cpu_to_vdpasim32(vdpasim, VDPASIM_BLK_DWZ_MAX_SECTORS);
+	blk_config->max_write_zeroes_seg = cpu_to_vdpasim32(vdpasim, 1);
+
+}
+
+static void vdpasim_blk_free(struct vdpasim *vdpasim)
+{
+	struct vdpasim_blk *blk = sim_to_blk(vdpasim);
+
+	if (!blk->shared_backend)
+		kvfree(blk->buffer);
+}
+
+static struct device *vdpasim_blk_mgmtdev;
+
+static int vdpasim_blk_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_blk *blk;
+	struct vdpasim *simdev;
+	int ret;
+
+	dev_attr.mgmt_dev = mdev;
+	dev_attr.name = name;
+	dev_attr.id = VIRTIO_ID_BLOCK;
+	dev_attr.supported_features = VDPASIM_BLK_FEATURES;
+	dev_attr.nvqs = VDPASIM_BLK_VQ_NUM;
+	dev_attr.ngroups = VDPASIM_BLK_GROUP_NUM;
+	dev_attr.nas = VDPASIM_BLK_AS_NUM;
+	dev_attr.alloc_size = sizeof(struct vdpasim_blk);
+	dev_attr.config_size = sizeof(struct virtio_blk_config);
+	dev_attr.get_config = vdpasim_blk_get_config;
+	dev_attr.work_fn = vdpasim_blk_work;
+	dev_attr.free = vdpasim_blk_free;
+
+	simdev = vdpasim_create(&dev_attr, config);
+	if (IS_ERR(simdev))
+		return PTR_ERR(simdev);
+
+	blk = sim_to_blk(simdev);
+	blk->shared_backend = shared_backend;
+
+	if (blk->shared_backend) {
+		blk->buffer = shared_buffer;
+	} else {
+		blk->buffer = kvzalloc(VDPASIM_BLK_CAPACITY << SECTOR_SHIFT,
+				       GFP_KERNEL);
+		if (!blk->buffer) {
+			ret = -ENOMEM;
+			goto put_dev;
+		}
+	}
+
+	ret = _vdpa_register_device(&simdev->vdpa, VDPASIM_BLK_VQ_NUM);
+	if (ret)
+		goto put_dev;
+
+	return 0;
+
+put_dev:
+	put_device(&simdev->vdpa.dev);
+	return ret;
+}
+
+static void vdpasim_blk_dev_del(struct vdpa_mgmt_dev *mdev,
+				struct vdpa_device *dev)
+{
+	struct vdpasim *simdev = container_of(dev, struct vdpasim, vdpa);
+
+	_vdpa_unregister_device(&simdev->vdpa);
+}
+
+static const struct vdpa_mgmtdev_ops vdpasim_blk_mgmtdev_ops = {
+	.dev_add = vdpasim_blk_dev_add,
+	.dev_del = vdpasim_blk_dev_del
+};
+
+static struct virtio_device_id id_table[] = {
+	{ VIRTIO_ID_BLOCK, VIRTIO_DEV_ANY_ID },
+	{ 0 },
+};
+
+static struct vdpa_mgmt_dev mgmt_dev = {
+	.id_table = id_table,
+	.ops = &vdpasim_blk_mgmtdev_ops,
+};
+
+static int __init vdpasim_blk_init(void)
+{
+	int ret;
+
+	vdpasim_blk_mgmtdev = root_device_register("vdpasim_blk");
+	if (IS_ERR(vdpasim_blk_mgmtdev))
+		return PTR_ERR(vdpasim_blk_mgmtdev);
+
+	mgmt_dev.device = vdpasim_blk_mgmtdev;
+	ret = vdpa_mgmtdev_register(&mgmt_dev);
+	if (ret)
+		goto parent_err;
+
+	if (shared_backend) {
+		shared_buffer = kvzalloc(VDPASIM_BLK_CAPACITY << SECTOR_SHIFT,
+					 GFP_KERNEL);
+		if (!shared_buffer) {
+			ret = -ENOMEM;
+			goto mgmt_dev_err;
+		}
+	}
+
+	return 0;
+mgmt_dev_err:
+	vdpa_mgmtdev_unregister(&mgmt_dev);
+parent_err:
+	root_device_unregister(vdpasim_blk_mgmtdev);
+
+	return ret;
+}
+
+static void __exit vdpasim_blk_exit(void)
+{
+	kvfree(shared_buffer);
+	vdpa_mgmtdev_unregister(&mgmt_dev);
+	root_device_unregister(vdpasim_blk_mgmtdev);
+}
+
+module_init(vdpasim_blk_init)
+module_exit(vdpasim_blk_exit)
+
+MODULE_VERSION(DRV_VERSION);
+MODULE_LICENSE(DRV_LICENSE);
+MODULE_AUTHOR(DRV_AUTHOR);
+MODULE_DESCRIPTION(DRV_DESC);

linux-7.2.3/drivers/vdpa/vdpa_sim/vdpa_sim_net.c

@@ -0,0 +1,574 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * VDPA simulator for networking device.
+ *
+ * Copyright (c) 2020, Red Hat Inc. All rights reserved.
+ *     Author: Jason Wang <jasowang@redhat.com>
+ *
+ */
+
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/device.h>
+#include <linux/kernel.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>
+
+#include "vdpa_sim.h"
+
+#define DRV_VERSION  "0.1"
+#define DRV_AUTHOR   "Jason Wang <jasowang@redhat.com>"
+#define DRV_DESC     "vDPA Device Simulator for networking device"
+#define DRV_LICENSE  "GPL v2"
+
+#define VDPASIM_NET_FEATURES	(VDPASIM_FEATURES | \
+				 (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;
+	}
+
+	return status;
+}
+
+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_virtqueue *txq = &vdpasim->vqs[1];
+	struct vdpasim_virtqueue *rxq = &vdpasim->vqs[0];
+	struct vdpasim_net *net = sim_to_net(vdpasim);
+	ssize_t read, write;
+	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;
+
+	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) {
+		err = vringh_getdesc_iotlb(&txq->vring, &txq->out_iov, NULL,
+					   &txq->head, GFP_ATOMIC);
+		if (err <= 0) {
+			if (err)
+				++tx_errors;
+			break;
+		}
+
+		++tx_pkts;
+		read = vringh_iov_pull_iotlb(&txq->vring, &txq->out_iov,
+					     net->buffer, PAGE_SIZE);
+
+		tx_bytes += read;
+
+		if (!receive_filter(vdpasim, read)) {
+			++rx_drops;
+			vdpasim_net_complete(txq, 0);
+			continue;
+		}
+
+		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;
+		}
+
+		write = vringh_iov_push_iotlb(&rxq->vring, &rxq->in_iov,
+					      net->buffer, read);
+		if (write <= 0) {
+			++rx_errors;
+			break;
+		}
+
+		++rx_pkts;
+		rx_bytes += write;
+
+		vdpasim_net_complete(txq, 0);
+		vdpasim_net_complete(rxq, write);
+
+		if (tx_pkts > 4) {
+			vdpasim_schedule_work(vdpasim);
+			goto out;
+		}
+	}
+
+out:
+	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)
+{
+	struct virtio_net_config *net_config = 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)
+{
+	struct virtio_net_config *vio_config = vdpasim->config;
+
+	if (config->mask & (1 << VDPA_ATTR_DEV_NET_CFG_MACADDR))
+		memcpy(vio_config->mac, config->net.mac, ETH_ALEN);
+	if (config->mask & (1 << VDPA_ATTR_DEV_NET_CFG_MTU))
+		vio_config->mtu = cpu_to_vdpasim16(vdpasim, config->net.mtu);
+	else
+		/* Setup default MTU to be 1500 */
+		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 struct device *vdpasim_net_mgmtdev;
+
+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;
+
+	dev_attr.mgmt_dev = mdev;
+	dev_attr.name = 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.get_stats = vdpasim_net_get_stats;
+	dev_attr.free = vdpasim_net_free;
+
+	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;
+
+	return 0;
+
+reg_err:
+	put_device(&simdev->vdpa.dev);
+	return ret;
+}
+
+static void vdpasim_net_dev_del(struct vdpa_mgmt_dev *mdev,
+				struct vdpa_device *dev)
+{
+	struct vdpasim *simdev = container_of(dev, struct vdpasim, vdpa);
+
+	_vdpa_unregister_device(&simdev->vdpa);
+}
+
+static const struct vdpa_mgmtdev_ops vdpasim_net_mgmtdev_ops = {
+	.dev_add = vdpasim_net_dev_add,
+	.dev_del = vdpasim_net_dev_del,
+	.dev_set_attr = vdpasim_net_set_attr
+};
+
+static struct virtio_device_id id_table[] = {
+	{ VIRTIO_ID_NET, VIRTIO_DEV_ANY_ID },
+	{ 0 },
+};
+
+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_FEATURES),
+	.max_supported_vqs = VDPASIM_NET_VQ_NUM,
+	.supported_features = VDPASIM_NET_FEATURES,
+};
+
+static int __init vdpasim_net_init(void)
+{
+	int ret;
+
+	vdpasim_net_mgmtdev = root_device_register("vdpasim_net");
+	if (IS_ERR(vdpasim_net_mgmtdev))
+		return PTR_ERR(vdpasim_net_mgmtdev);
+
+	mgmt_dev.device = vdpasim_net_mgmtdev;
+	ret = vdpa_mgmtdev_register(&mgmt_dev);
+	if (ret)
+		goto parent_err;
+	return 0;
+
+parent_err:
+	root_device_unregister(vdpasim_net_mgmtdev);
+	return ret;
+}
+
+static void __exit vdpasim_net_exit(void)
+{
+	vdpa_mgmtdev_unregister(&mgmt_dev);
+	root_device_unregister(vdpasim_net_mgmtdev);
+}
+
+module_init(vdpasim_net_init);
+module_exit(vdpasim_net_exit);
+
+MODULE_VERSION(DRV_VERSION);
+MODULE_LICENSE(DRV_LICENSE);
+MODULE_AUTHOR(DRV_AUTHOR);
+MODULE_DESCRIPTION(DRV_DESC);