Merge tag 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mst/vhost

Pull virtio updates from Michael Tsirkin:

 - vdpa/mlx5: support for resumable vqs

 - virtio_scsi: mq_poll support

 - 3virtio_pmem: support SHMEM_REGION

 - virtio_balloon: stay awake while adjusting balloon

 - virtio: support for no-reset virtio PCI PM

 - Fixes, cleanups

* tag 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mst/vhost:
  vdpa/mlx5: Add mkey leak detection
  vdpa/mlx5: Introduce reference counting to mrs
  vdpa/mlx5: Use vq suspend/resume during .set_map
  vdpa/mlx5: Mark vq state for modification in hw vq
  vdpa/mlx5: Mark vq addrs for modification in hw vq
  vdpa/mlx5: Introduce per vq and device resume
  vdpa/mlx5: Allow modifying multiple vq fields in one modify command
  vdpa/mlx5: Expose resumable vq capability
  vdpa: Block vq property changes in DRIVER_OK
  vdpa: Track device suspended state
  scsi: virtio_scsi: Add mq_poll support
  virtio_pmem: support feature SHMEM_REGION
  virtio_balloon: stay awake while adjusting balloon
  vdpa: Remove usage of the deprecated ida_simple_xx() API
  virtio: Add support for no-reset virtio PCI PM
  virtio_net: fix missing dma unmap for resize
  vhost-vdpa: account iommu allocations
  vdpa: Fix an error handling path in eni_vdpa_probe()

drivers/nvdimm/virtio_pmem.c

@@ -29,12 +29,27 @@ static int init_vq(struct virtio_pmem *vpmem)
 	return 0;
 };
 
+static int virtio_pmem_validate(struct virtio_device *vdev)
+{
+	struct virtio_shm_region shm_reg;
+
+	if (virtio_has_feature(vdev, VIRTIO_PMEM_F_SHMEM_REGION) &&
+		!virtio_get_shm_region(vdev, &shm_reg, (u8)VIRTIO_PMEM_SHMEM_REGION_ID)
+	) {
+		dev_notice(&vdev->dev, "failed to get shared memory region %d\n",
+				VIRTIO_PMEM_SHMEM_REGION_ID);
+		__virtio_clear_bit(vdev, VIRTIO_PMEM_F_SHMEM_REGION);
+	}
+	return 0;
+}
+
 static int virtio_pmem_probe(struct virtio_device *vdev)
 {
 	struct nd_region_desc ndr_desc = {};
 	struct nd_region *nd_region;
 	struct virtio_pmem *vpmem;
 	struct resource res;
+	struct virtio_shm_region shm_reg;
 	int err = 0;
 
 	if (!vdev->config->get) {
@@ -57,10 +72,16 @@ static int virtio_pmem_probe(struct virtio_device *vdev)
 		goto out_err;
 	}
 
-	virtio_cread_le(vpmem->vdev, struct virtio_pmem_config,
-			start, &vpmem->start);
-	virtio_cread_le(vpmem->vdev, struct virtio_pmem_config,
-			size, &vpmem->size);
+	if (virtio_has_feature(vdev, VIRTIO_PMEM_F_SHMEM_REGION)) {
+		virtio_get_shm_region(vdev, &shm_reg, (u8)VIRTIO_PMEM_SHMEM_REGION_ID);
+		vpmem->start = shm_reg.addr;
+		vpmem->size = shm_reg.len;
+	} else {
+		virtio_cread_le(vpmem->vdev, struct virtio_pmem_config,
+				start, &vpmem->start);
+		virtio_cread_le(vpmem->vdev, struct virtio_pmem_config,
+				size, &vpmem->size);
+	}
 
 	res.start = vpmem->start;
 	res.end   = vpmem->start + vpmem->size - 1;
@@ -122,10 +143,17 @@ static void virtio_pmem_remove(struct virtio_device *vdev)
 	virtio_reset_device(vdev);
 }
 
+static unsigned int features[] = {
+	VIRTIO_PMEM_F_SHMEM_REGION,
+};
+
 static struct virtio_driver virtio_pmem_driver = {
+	.feature_table		= features,
+	.feature_table_size	= ARRAY_SIZE(features),
 	.driver.name		= KBUILD_MODNAME,
 	.driver.owner		= THIS_MODULE,
 	.id_table		= id_table,
+	.validate		= virtio_pmem_validate,
 	.probe			= virtio_pmem_probe,
 	.remove			= virtio_pmem_remove,
 };

drivers/scsi/virtio_scsi.c

@@ -37,6 +37,11 @@
 #define VIRTIO_SCSI_EVENT_LEN 8
 #define VIRTIO_SCSI_VQ_BASE 2
 
+static unsigned int virtscsi_poll_queues;
+module_param(virtscsi_poll_queues, uint, 0644);
+MODULE_PARM_DESC(virtscsi_poll_queues,
+		 "The number of dedicated virtqueues for polling I/O");
+
 /* Command queue element */
 struct virtio_scsi_cmd {
 	struct scsi_cmnd *sc;
@@ -76,6 +81,7 @@ struct virtio_scsi {
 	struct virtio_scsi_event_node event_list[VIRTIO_SCSI_EVENT_LEN];
 
 	u32 num_queues;
+	int io_queues[HCTX_MAX_TYPES];
 
 	struct hlist_node node;
 
@@ -722,9 +728,49 @@ static int virtscsi_abort(struct scsi_cmnd *sc)
 static void virtscsi_map_queues(struct Scsi_Host *shost)
 {
 	struct virtio_scsi *vscsi = shost_priv(shost);
-	struct blk_mq_queue_map *qmap = &shost->tag_set.map[HCTX_TYPE_DEFAULT];
+	int i, qoff;
+
+	for (i = 0, qoff = 0; i < shost->nr_maps; i++) {
+		struct blk_mq_queue_map *map = &shost->tag_set.map[i];
+
+		map->nr_queues = vscsi->io_queues[i];
+		map->queue_offset = qoff;
+		qoff += map->nr_queues;
+
+		if (map->nr_queues == 0)
+			continue;
+
+		/*
+		 * Regular queues have interrupts and hence CPU affinity is
+		 * defined by the core virtio code, but polling queues have
+		 * no interrupts so we let the block layer assign CPU affinity.
+		 */
+		if (i == HCTX_TYPE_POLL)
+			blk_mq_map_queues(map);
+		else
+			blk_mq_virtio_map_queues(map, vscsi->vdev, 2);
+	}
+}
+
+static int virtscsi_mq_poll(struct Scsi_Host *shost, unsigned int queue_num)
+{
+	struct virtio_scsi *vscsi = shost_priv(shost);
+	struct virtio_scsi_vq *virtscsi_vq = &vscsi->req_vqs[queue_num];
+	unsigned long flags;
+	unsigned int len;
+	int found = 0;
+	void *buf;
+
+	spin_lock_irqsave(&virtscsi_vq->vq_lock, flags);
+
+	while ((buf = virtqueue_get_buf(virtscsi_vq->vq, &len)) != NULL) {
+		virtscsi_complete_cmd(vscsi, buf);
+		found++;
+	}
+
+	spin_unlock_irqrestore(&virtscsi_vq->vq_lock, flags);
 
-	blk_mq_virtio_map_queues(qmap, vscsi->vdev, 2);
+	return found;
 }
 
 static void virtscsi_commit_rqs(struct Scsi_Host *shost, u16 hwq)
@@ -751,6 +797,7 @@ static const struct scsi_host_template virtscsi_host_template = {
 	.this_id = -1,
 	.cmd_size = sizeof(struct virtio_scsi_cmd),
 	.queuecommand = virtscsi_queuecommand,
+	.mq_poll = virtscsi_mq_poll,
 	.commit_rqs = virtscsi_commit_rqs,
 	.change_queue_depth = virtscsi_change_queue_depth,
 	.eh_abort_handler = virtscsi_abort,
@@ -795,13 +842,14 @@ static int virtscsi_init(struct virtio_device *vdev,
 {
 	int err;
 	u32 i;
-	u32 num_vqs;
+	u32 num_vqs, num_poll_vqs, num_req_vqs;
 	vq_callback_t **callbacks;
 	const char **names;
 	struct virtqueue **vqs;
 	struct irq_affinity desc = { .pre_vectors = 2 };
 
-	num_vqs = vscsi->num_queues + VIRTIO_SCSI_VQ_BASE;
+	num_req_vqs = vscsi->num_queues;
+	num_vqs = num_req_vqs + VIRTIO_SCSI_VQ_BASE;
 	vqs = kmalloc_array(num_vqs, sizeof(struct virtqueue *), GFP_KERNEL);
 	callbacks = kmalloc_array(num_vqs, sizeof(vq_callback_t *),
 				  GFP_KERNEL);
@@ -812,15 +860,31 @@ static int virtscsi_init(struct virtio_device *vdev,
 		goto out;
 	}
 
+	num_poll_vqs = min_t(unsigned int, virtscsi_poll_queues,
+			     num_req_vqs - 1);
+	vscsi->io_queues[HCTX_TYPE_DEFAULT] = num_req_vqs - num_poll_vqs;
+	vscsi->io_queues[HCTX_TYPE_READ] = 0;
+	vscsi->io_queues[HCTX_TYPE_POLL] = num_poll_vqs;
+
+	dev_info(&vdev->dev, "%d/%d/%d default/read/poll queues\n",
+		 vscsi->io_queues[HCTX_TYPE_DEFAULT],
+		 vscsi->io_queues[HCTX_TYPE_READ],
+		 vscsi->io_queues[HCTX_TYPE_POLL]);
+
 	callbacks[0] = virtscsi_ctrl_done;
 	callbacks[1] = virtscsi_event_done;
 	names[0] = "control";
 	names[1] = "event";
-	for (i = VIRTIO_SCSI_VQ_BASE; i < num_vqs; i++) {
+	for (i = VIRTIO_SCSI_VQ_BASE; i < num_vqs - num_poll_vqs; i++) {
 		callbacks[i] = virtscsi_req_done;
 		names[i] = "request";
 	}
 
+	for (; i < num_vqs; i++) {
+		callbacks[i] = NULL;
+		names[i] = "request_poll";
+	}
+
 	/* Discover virtqueues and write information to configuration.  */
 	err = virtio_find_vqs(vdev, num_vqs, vqs, callbacks, names, &desc);
 	if (err)
@@ -874,6 +938,7 @@ static int virtscsi_probe(struct virtio_device *vdev)
 
 	sg_elems = virtscsi_config_get(vdev, seg_max) ?: 1;
 	shost->sg_tablesize = sg_elems;
+	shost->nr_maps = 1;
 	vscsi = shost_priv(shost);
 	vscsi->vdev = vdev;
 	vscsi->num_queues = num_queues;
@@ -883,6 +948,9 @@ static int virtscsi_probe(struct virtio_device *vdev)
 	if (err)
 		goto virtscsi_init_failed;
 
+	if (vscsi->io_queues[HCTX_TYPE_POLL])
+		shost->nr_maps = HCTX_TYPE_POLL + 1;
+
 	shost->can_queue = virtqueue_get_vring_size(vscsi->req_vqs[0].vq);
 
 	cmd_per_lun = virtscsi_config_get(vdev, cmd_per_lun) ?: 1;

drivers/vdpa/alibaba/eni_vdpa.c

@@ -497,7 +497,7 @@ static int eni_vdpa_probe(struct pci_dev *pdev, const struct pci_device_id *id)
 	if (!eni_vdpa->vring) {
 		ret = -ENOMEM;
 		ENI_ERR(pdev, "failed to allocate virtqueues\n");
-		goto err;
+		goto err_remove_vp_legacy;
 	}
 
 	for (i = 0; i < eni_vdpa->queues; i++) {
@@ -509,11 +509,13 @@ static int eni_vdpa_probe(struct pci_dev *pdev, const struct pci_device_id *id)
 	ret = vdpa_register_device(&eni_vdpa->vdpa, eni_vdpa->queues);
 	if (ret) {
 		ENI_ERR(pdev, "failed to register to vdpa bus\n");
-		goto err;
+		goto err_remove_vp_legacy;
 	}
 
 	return 0;
 
+err_remove_vp_legacy:
+	vp_legacy_remove(&eni_vdpa->ldev);
 err:
 	put_device(&eni_vdpa->vdpa.dev);
 	return ret;

drivers/vdpa/mlx5/core/mlx5_vdpa.h

@@ -35,6 +35,9 @@ struct mlx5_vdpa_mr {
 	struct vhost_iotlb *iotlb;
 
 	bool user_mr;
+
+	refcount_t refcount;
+	struct list_head mr_list;
 };
 
 struct mlx5_vdpa_resources {
@@ -93,6 +96,7 @@ struct mlx5_vdpa_dev {
 	u32 generation;
 
 	struct mlx5_vdpa_mr *mr[MLX5_VDPA_NUM_AS];
+	struct list_head mr_list_head;
 	/* serialize mr access */
 	struct mutex mr_mtx;
 	struct mlx5_control_vq cvq;
@@ -118,8 +122,10 @@ int mlx5_vdpa_destroy_mkey(struct mlx5_vdpa_dev *mvdev, u32 mkey);
 struct mlx5_vdpa_mr *mlx5_vdpa_create_mr(struct mlx5_vdpa_dev *mvdev,
 					 struct vhost_iotlb *iotlb);
 void mlx5_vdpa_destroy_mr_resources(struct mlx5_vdpa_dev *mvdev);
-void mlx5_vdpa_destroy_mr(struct mlx5_vdpa_dev *mvdev,
-			  struct mlx5_vdpa_mr *mr);
+void mlx5_vdpa_get_mr(struct mlx5_vdpa_dev *mvdev,
+		      struct mlx5_vdpa_mr *mr);
+void mlx5_vdpa_put_mr(struct mlx5_vdpa_dev *mvdev,
+		      struct mlx5_vdpa_mr *mr);
 void mlx5_vdpa_update_mr(struct mlx5_vdpa_dev *mvdev,
 			 struct mlx5_vdpa_mr *mr,
 			 unsigned int asid);

drivers/vdpa/mlx5/core/mr.c

@@ -498,32 +498,54 @@ static void destroy_user_mr(struct mlx5_vdpa_dev *mvdev, struct mlx5_vdpa_mr *mr
 
 static void _mlx5_vdpa_destroy_mr(struct mlx5_vdpa_dev *mvdev, struct mlx5_vdpa_mr *mr)
 {
+	if (WARN_ON(!mr))
+		return;
+
 	if (mr->user_mr)
 		destroy_user_mr(mvdev, mr);
 	else
 		destroy_dma_mr(mvdev, mr);
 
 	vhost_iotlb_free(mr->iotlb);
+
+	list_del(&mr->mr_list);
+
+	kfree(mr);
 }
 
-void mlx5_vdpa_destroy_mr(struct mlx5_vdpa_dev *mvdev,
-			  struct mlx5_vdpa_mr *mr)
+static void _mlx5_vdpa_put_mr(struct mlx5_vdpa_dev *mvdev,
+			      struct mlx5_vdpa_mr *mr)
 {
 	if (!mr)
 		return;
 
+	if (refcount_dec_and_test(&mr->refcount))
+		_mlx5_vdpa_destroy_mr(mvdev, mr);
+}
+
+void mlx5_vdpa_put_mr(struct mlx5_vdpa_dev *mvdev,
+		      struct mlx5_vdpa_mr *mr)
+{
 	mutex_lock(&mvdev->mr_mtx);
+	_mlx5_vdpa_put_mr(mvdev, mr);
+	mutex_unlock(&mvdev->mr_mtx);
+}
 
-	_mlx5_vdpa_destroy_mr(mvdev, mr);
+static void _mlx5_vdpa_get_mr(struct mlx5_vdpa_dev *mvdev,
+			      struct mlx5_vdpa_mr *mr)
+{
+	if (!mr)
+		return;
 
-	for (int i = 0; i < MLX5_VDPA_NUM_AS; i++) {
-		if (mvdev->mr[i] == mr)
-			mvdev->mr[i] = NULL;
-	}
+	refcount_inc(&mr->refcount);
+}
 
+void mlx5_vdpa_get_mr(struct mlx5_vdpa_dev *mvdev,
+		      struct mlx5_vdpa_mr *mr)
+{
+	mutex_lock(&mvdev->mr_mtx);
+	_mlx5_vdpa_get_mr(mvdev, mr);
 	mutex_unlock(&mvdev->mr_mtx);
-
-	kfree(mr);
 }
 
 void mlx5_vdpa_update_mr(struct mlx5_vdpa_dev *mvdev,
@@ -534,10 +556,23 @@ void mlx5_vdpa_update_mr(struct mlx5_vdpa_dev *mvdev,
 
 	mutex_lock(&mvdev->mr_mtx);
 
+	_mlx5_vdpa_put_mr(mvdev, old_mr);
 	mvdev->mr[asid] = new_mr;
-	if (old_mr) {
-		_mlx5_vdpa_destroy_mr(mvdev, old_mr);
-		kfree(old_mr);
+
+	mutex_unlock(&mvdev->mr_mtx);
+}
+
+static void mlx5_vdpa_show_mr_leaks(struct mlx5_vdpa_dev *mvdev)
+{
+	struct mlx5_vdpa_mr *mr;
+
+	mutex_lock(&mvdev->mr_mtx);
+
+	list_for_each_entry(mr, &mvdev->mr_list_head, mr_list) {
+
+		mlx5_vdpa_warn(mvdev, "mkey still alive after resource delete: "
+				      "mr: %p, mkey: 0x%x, refcount: %u\n",
+				       mr, mr->mkey, refcount_read(&mr->refcount));
 	}
 
 	mutex_unlock(&mvdev->mr_mtx);
@@ -547,9 +582,11 @@ void mlx5_vdpa_update_mr(struct mlx5_vdpa_dev *mvdev,
 void mlx5_vdpa_destroy_mr_resources(struct mlx5_vdpa_dev *mvdev)
 {
 	for (int i = 0; i < MLX5_VDPA_NUM_AS; i++)
-		mlx5_vdpa_destroy_mr(mvdev, mvdev->mr[i]);
+		mlx5_vdpa_update_mr(mvdev, NULL, i);
 
 	prune_iotlb(mvdev->cvq.iotlb);
+
+	mlx5_vdpa_show_mr_leaks(mvdev);
 }
 
 static int _mlx5_vdpa_create_mr(struct mlx5_vdpa_dev *mvdev,
@@ -576,6 +613,8 @@ static int _mlx5_vdpa_create_mr(struct mlx5_vdpa_dev *mvdev,
 	if (err)
 		goto err_iotlb;
 
+	list_add_tail(&mr->mr_list, &mvdev->mr_list_head);
+
 	return 0;
 
 err_iotlb:
@@ -607,6 +646,8 @@ struct mlx5_vdpa_mr *mlx5_vdpa_create_mr(struct mlx5_vdpa_dev *mvdev,
 	if (err)
 		goto out_err;
 
+	refcount_set(&mr->refcount, 1);
+
 	return mr;
 
 out_err:
@@ -651,7 +692,7 @@ int mlx5_vdpa_reset_mr(struct mlx5_vdpa_dev *mvdev, unsigned int asid)
 	if (asid >= MLX5_VDPA_NUM_AS)
 		return -EINVAL;
 
-	mlx5_vdpa_destroy_mr(mvdev, mvdev->mr[asid]);
+	mlx5_vdpa_update_mr(mvdev, NULL, asid);
 
 	if (asid == 0 && MLX5_CAP_GEN(mvdev->mdev, umem_uid_0)) {
 		if (mlx5_vdpa_create_dma_mr(mvdev))

drivers/vdpa/vdpa.c

@@ -131,7 +131,7 @@ static void vdpa_release_dev(struct device *d)
 	if (ops->free)
 		ops->free(vdev);
 
-	ida_simple_remove(&vdpa_index_ida, vdev->index);
+	ida_free(&vdpa_index_ida, vdev->index);
 	kfree(vdev->driver_override);
 	kfree(vdev);
 }
@@ -205,7 +205,7 @@ struct vdpa_device *__vdpa_alloc_device(struct device *parent,
 	return vdev;
 
 err_name:
-	ida_simple_remove(&vdpa_index_ida, vdev->index);
+	ida_free(&vdpa_index_ida, vdev->index);
 err_ida:
 	kfree(vdev);
 err:

drivers/vhost/vdpa.c

@@ -59,6 +59,7 @@ struct vhost_vdpa {
 	int in_batch;
 	struct vdpa_iova_range range;
 	u32 batch_asid;
+	bool suspended;
 };
 
 static DEFINE_IDA(vhost_vdpa_ida);
@@ -232,6 +233,8 @@ static int _compat_vdpa_reset(struct vhost_vdpa *v)
 	struct vdpa_device *vdpa = v->vdpa;
 	u32 flags = 0;
 
+	v->suspended = false;
+
 	if (v->vdev.vqs) {
 		flags |= !vhost_backend_has_feature(v->vdev.vqs[0],
 						    VHOST_BACKEND_F_IOTLB_PERSIST) ?
@@ -590,11 +593,16 @@ static long vhost_vdpa_suspend(struct vhost_vdpa *v)
 {
 	struct vdpa_device *vdpa = v->vdpa;
 	const struct vdpa_config_ops *ops = vdpa->config;
+	int ret;
 
 	if (!ops->suspend)
 		return -EOPNOTSUPP;
 
-	return ops->suspend(vdpa);
+	ret = ops->suspend(vdpa);
+	if (!ret)
+		v->suspended = true;
+
+	return ret;
 }
 
 /* After a successful return of this ioctl the device resumes processing
@@ -605,11 +613,16 @@ static long vhost_vdpa_resume(struct vhost_vdpa *v)
 {
 	struct vdpa_device *vdpa = v->vdpa;
 	const struct vdpa_config_ops *ops = vdpa->config;
+	int ret;
 
 	if (!ops->resume)
 		return -EOPNOTSUPP;
 
-	return ops->resume(vdpa);
+	ret = ops->resume(vdpa);
+	if (!ret)
+		v->suspended = false;
+
+	return ret;
 }
 
 static long vhost_vdpa_vring_ioctl(struct vhost_vdpa *v, unsigned int cmd,
@@ -690,6 +703,9 @@ static long vhost_vdpa_vring_ioctl(struct vhost_vdpa *v, unsigned int cmd,
 
 	switch (cmd) {
 	case VHOST_SET_VRING_ADDR:
+		if ((ops->get_status(vdpa) & VIRTIO_CONFIG_S_DRIVER_OK) && !v->suspended)
+			return -EINVAL;
+
 		if (ops->set_vq_address(vdpa, idx,
 					(u64)(uintptr_t)vq->desc,
 					(u64)(uintptr_t)vq->avail,
@@ -698,6 +714,9 @@ static long vhost_vdpa_vring_ioctl(struct vhost_vdpa *v, unsigned int cmd,
 		break;
 
 	case VHOST_SET_VRING_BASE:
+		if ((ops->get_status(vdpa) & VIRTIO_CONFIG_S_DRIVER_OK) && !v->suspended)
+			return -EINVAL;
+
 		if (vhost_has_feature(vq, VIRTIO_F_RING_PACKED)) {
 			vq_state.packed.last_avail_idx = vq->last_avail_idx & 0x7fff;
 			vq_state.packed.last_avail_counter = !!(vq->last_avail_idx & 0x8000);
@@ -968,7 +987,8 @@ static int vhost_vdpa_map(struct vhost_vdpa *v, struct vhost_iotlb *iotlb,
 			r = ops->set_map(vdpa, asid, iotlb);
 	} else {
 		r = iommu_map(v->domain, iova, pa, size,
-			      perm_to_iommu_flags(perm), GFP_KERNEL);
+			      perm_to_iommu_flags(perm),
+			      GFP_KERNEL_ACCOUNT);
 	}
 	if (r) {
 		vhost_iotlb_del_range(iotlb, iova, iova + size - 1);

drivers/virtio/virtio_balloon.c

@@ -119,6 +119,11 @@ struct virtio_balloon {
 	/* Free page reporting device */
 	struct virtqueue *reporting_vq;
 	struct page_reporting_dev_info pr_dev_info;
+
+	/* State for keeping the wakeup_source active while adjusting the balloon */
+	spinlock_t adjustment_lock;
+	bool adjustment_signal_pending;
+	bool adjustment_in_progress;
 };
 
 static const struct virtio_device_id id_table[] = {
@@ -437,6 +442,31 @@ static void virtio_balloon_queue_free_page_work(struct virtio_balloon *vb)
 	queue_work(vb->balloon_wq, &vb->report_free_page_work);
 }
 
+static void start_update_balloon_size(struct virtio_balloon *vb)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&vb->adjustment_lock, flags);
+	vb->adjustment_signal_pending = true;
+	if (!vb->adjustment_in_progress) {
+		vb->adjustment_in_progress = true;
+		pm_stay_awake(vb->vdev->dev.parent);
+	}
+	spin_unlock_irqrestore(&vb->adjustment_lock, flags);
+
+	queue_work(system_freezable_wq, &vb->update_balloon_size_work);
+}
+
+static void end_update_balloon_size(struct virtio_balloon *vb)
+{
+	spin_lock_irq(&vb->adjustment_lock);
+	if (!vb->adjustment_signal_pending && vb->adjustment_in_progress) {
+		vb->adjustment_in_progress = false;
+		pm_relax(vb->vdev->dev.parent);
+	}
+	spin_unlock_irq(&vb->adjustment_lock);
+}
+
 static void virtballoon_changed(struct virtio_device *vdev)
 {
 	struct virtio_balloon *vb = vdev->priv;
@@ -444,8 +474,7 @@ static void virtballoon_changed(struct virtio_device *vdev)
 
 	spin_lock_irqsave(&vb->stop_update_lock, flags);
 	if (!vb->stop_update) {
-		queue_work(system_freezable_wq,
-			   &vb->update_balloon_size_work);
+		start_update_balloon_size(vb);
 		virtio_balloon_queue_free_page_work(vb);
 	}
 	spin_unlock_irqrestore(&vb->stop_update_lock, flags);
@@ -476,19 +505,25 @@ static void update_balloon_size_func(struct work_struct *work)
 
 	vb = container_of(work, struct virtio_balloon,
 			  update_balloon_size_work);
-	diff = towards_target(vb);
 
-	if (!diff)
-		return;
+	spin_lock_irq(&vb->adjustment_lock);
+	vb->adjustment_signal_pending = false;
+	spin_unlock_irq(&vb->adjustment_lock);
 
-	if (diff > 0)
-		diff -= fill_balloon(vb, diff);
-	else
-		diff += leak_balloon(vb, -diff);
-	update_balloon_size(vb);
+	diff = towards_target(vb);
+
+	if (diff) {
+		if (diff > 0)
+			diff -= fill_balloon(vb, diff);
+		else
+			diff += leak_balloon(vb, -diff);
+		update_balloon_size(vb);
+	}
 
 	if (diff)
 		queue_work(system_freezable_wq, work);
+	else
+		end_update_balloon_size(vb);
 }
 
 static int init_vqs(struct virtio_balloon *vb)
@@ -992,6 +1027,8 @@ static int virtballoon_probe(struct virtio_device *vdev)
 			goto out_unregister_oom;
 	}
 
+	spin_lock_init(&vb->adjustment_lock);
+
 	virtio_device_ready(vdev);
 
 	if (towards_target(vb))

drivers/virtio/virtio_pci_common.c

@@ -495,8 +495,40 @@ static int virtio_pci_restore(struct device *dev)
 	return virtio_device_restore(&vp_dev->vdev);
 }
 
+static bool vp_supports_pm_no_reset(struct device *dev)
+{
+	struct pci_dev *pci_dev = to_pci_dev(dev);
+	u16 pmcsr;
+
+	if (!pci_dev->pm_cap)
+		return false;
+
+	pci_read_config_word(pci_dev, pci_dev->pm_cap + PCI_PM_CTRL, &pmcsr);
+	if (PCI_POSSIBLE_ERROR(pmcsr)) {
+		dev_err(dev, "Unable to query pmcsr");
+		return false;
+	}
+
+	return pmcsr & PCI_PM_CTRL_NO_SOFT_RESET;
+}
+
+static int virtio_pci_suspend(struct device *dev)
+{
+	return vp_supports_pm_no_reset(dev) ? 0 : virtio_pci_freeze(dev);
+}
+
+static int virtio_pci_resume(struct device *dev)
+{
+	return vp_supports_pm_no_reset(dev) ? 0 : virtio_pci_restore(dev);
+}
+
 static const struct dev_pm_ops virtio_pci_pm_ops = {
-	SET_SYSTEM_SLEEP_PM_OPS(virtio_pci_freeze, virtio_pci_restore)
+	.suspend = virtio_pci_suspend,
+	.resume = virtio_pci_resume,
+	.freeze = virtio_pci_freeze,
+	.thaw = virtio_pci_restore,
+	.poweroff = virtio_pci_freeze,
+	.restore = virtio_pci_restore,
 };
 #endif
 

include/linux/mlx5/mlx5_ifc.h

@@ -1241,7 +1241,8 @@ struct mlx5_ifc_virtio_emulation_cap_bits {
 
 	u8	   reserved_at_c0[0x13];
 	u8         desc_group_mkey_supported[0x1];
-	u8         reserved_at_d4[0xc];
+	u8         freeze_to_rdy_supported[0x1];
+	u8         reserved_at_d5[0xb];
 
 	u8         reserved_at_e0[0x20];
 

include/linux/mlx5/mlx5_ifc_vdpa.h

@@ -145,6 +145,10 @@ enum {
 	MLX5_VIRTQ_MODIFY_MASK_STATE                    = (u64)1 << 0,
 	MLX5_VIRTQ_MODIFY_MASK_DIRTY_BITMAP_PARAMS      = (u64)1 << 3,
 	MLX5_VIRTQ_MODIFY_MASK_DIRTY_BITMAP_DUMP_ENABLE = (u64)1 << 4,
+	MLX5_VIRTQ_MODIFY_MASK_VIRTIO_Q_ADDRS           = (u64)1 << 6,
+	MLX5_VIRTQ_MODIFY_MASK_VIRTIO_Q_AVAIL_IDX       = (u64)1 << 7,
+	MLX5_VIRTQ_MODIFY_MASK_VIRTIO_Q_USED_IDX        = (u64)1 << 8,
+	MLX5_VIRTQ_MODIFY_MASK_VIRTIO_Q_MKEY            = (u64)1 << 11,
 	MLX5_VIRTQ_MODIFY_MASK_DESC_GROUP_MKEY          = (u64)1 << 14,
 };
 

include/uapi/linux/virtio_pmem.h

@@ -14,6 +14,13 @@
 #include <linux/virtio_ids.h>
 #include <linux/virtio_config.h>
 
+/* Feature bits */
+/* guest physical address range will be indicated as shared memory region 0 */
+#define VIRTIO_PMEM_F_SHMEM_REGION 0
+
+/* shmid of the shared memory region corresponding to the pmem */
+#define VIRTIO_PMEM_SHMEM_REGION_ID 0
+
 struct virtio_pmem_config {
 	__le64 start;
 	__le64 size;