virtio: features, fixes, cleanups

Several new features here:
 
 	virtio-balloon supports new stats
 
 	vdpa supports setting mac address
 
 	vdpa/mlx5 suspend/resume as well as MKEY ops are now faster
 
 	virtio_fs supports new sysfs entries for queue info
 
 	virtio/vsock performance has been improved
 
 Fixes, cleanups all over the place.
 
 Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
 -----BEGIN PGP SIGNATURE-----
 
 iQFDBAABCAAtFiEEXQn9CHHI+FuUyooNKB8NuNKNVGkFAmbz7ykPHG1zdEByZWRo
 YXQuY29tAAoJECgfDbjSjVRpkk8H/A3vMRYXBzne9anezZLvADKS/CpX7v0DFEVj
 VfSMWXvYdUariYDyyb7pZsvK5QR22pE0pIaW6Kcgv9fNwq27M/H6g6NJk5ny8a7d
 216AQs1J28pXPPY+q03fhf3SzE3yHP8aeD9lyiO9QJYfs9vjtoyZeBGt3a4IUSX4
 ZeNBAx8xWTBcEDIIcZLdY1DNDTbZ4+qQ12Ln9IKq7D4xkE6l7Xh+HGdgTWTnDZ8P
 qEUUOmJTFKTQdOiVuU4NN3wzgHKWHdwKg0uWXo7ereYr3kYe3q//jCcLMv88a1x0
 XP7NRBQg/rsErwTMdLz6ffyqXJs6lGGqNXzRfZKEwAvmnh/+zs4=
 =gNBq
 -----END PGP SIGNATURE-----

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

Pull virtio updates from Michael Tsirkin:
 "Several new features here:

   - virtio-balloon supports new stats

   - vdpa supports setting mac address

   - vdpa/mlx5 suspend/resume as well as MKEY ops are now faster

   - virtio_fs supports new sysfs entries for queue info

   - virtio/vsock performance has been improved

  And fixes, cleanups all over the place"

* tag 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mst/vhost: (34 commits)
  vsock/virtio: avoid queuing packets when intermediate queue is empty
  vsock/virtio: refactor virtio_transport_send_pkt_work
  fw_cfg: Constify struct kobj_type
  vdpa/mlx5: Postpone MR deletion
  vdpa/mlx5: Introduce init/destroy for MR resources
  vdpa/mlx5: Rename mr_mtx -> lock
  vdpa/mlx5: Extract mr members in own resource struct
  vdpa/mlx5: Rename function
  vdpa/mlx5: Delete direct MKEYs in parallel
  vdpa/mlx5: Create direct MKEYs in parallel
  MAINTAINERS: add virtio-vsock driver in the VIRTIO CORE section
  virtio_fs: add sysfs entries for queue information
  virtio_fs: introduce virtio_fs_put_locked helper
  vdpa: Remove unused declarations
  vdpa/mlx5: Parallelize VQ suspend/resume for CVQ MQ command
  vdpa/mlx5: Small improvement for change_num_qps()
  vdpa/mlx5: Keep notifiers during suspend but ignore
  vdpa/mlx5: Parallelize device resume
  vdpa/mlx5: Parallelize device suspend
  vdpa/mlx5: Use async API for vq modify commands
  ...

MAINTAINERS

@@ -24464,6 +24464,7 @@ F:	include/linux/vdpa.h
 F:	include/linux/virtio*.h
 F:	include/linux/vringh.h
 F:	include/uapi/linux/virtio_*.h
+F:	net/vmw_vsock/virtio*
 F:	tools/virtio/
 F:	tools/testing/selftests/drivers/net/virtio_net/
 

drivers/firmware/qemu_fw_cfg.c

@@ -452,7 +452,7 @@ static void fw_cfg_sysfs_release_entry(struct kobject *kobj)
 }
 
 /* kobj_type: ties together all properties required to register an entry */
-static struct kobj_type fw_cfg_sysfs_entry_ktype = {
+static const struct kobj_type fw_cfg_sysfs_entry_ktype = {
 	.default_groups = fw_cfg_sysfs_entry_groups,
 	.sysfs_ops = &fw_cfg_sysfs_attr_ops,
 	.release = fw_cfg_sysfs_release_entry,

drivers/net/ethernet/mellanox/mlx5/core/cmd.c

@@ -1887,10 +1887,12 @@ static int cmd_exec(struct mlx5_core_dev *dev, void *in, int in_size, void *out,
 
 	throttle_op = mlx5_cmd_is_throttle_opcode(opcode);
 	if (throttle_op) {
-		/* atomic context may not sleep */
+		if (callback) {
-		if (callback)
+			if (down_trylock(&dev->cmd.vars.throttle_sem))
-			return -EINVAL;
+				return -EBUSY;
-		down(&dev->cmd.vars.throttle_sem);
+		} else {
+			down(&dev->cmd.vars.throttle_sem);
+		}
 	}
 
 	pages_queue = is_manage_pages(in);
@@ -2096,10 +2098,19 @@ static void mlx5_cmd_exec_cb_handler(int status, void *_work)
 {
 	struct mlx5_async_work *work = _work;
 	struct mlx5_async_ctx *ctx;
+	struct mlx5_core_dev *dev;
+	u16 opcode;
 
 	ctx = work->ctx;
-	status = cmd_status_err(ctx->dev, status, work->opcode, work->op_mod, work->out);
+	dev = ctx->dev;
+	opcode = work->opcode;
+	status = cmd_status_err(dev, status, work->opcode, work->op_mod, work->out);
 	work->user_callback(status, work);
+	/* Can't access "work" from this point on. It could have been freed in
+	 * the callback.
+	 */
+	if (mlx5_cmd_is_throttle_opcode(opcode))
+		up(&dev->cmd.vars.throttle_sem);
 	if (atomic_dec_and_test(&ctx->num_inflight))
 		complete(&ctx->inflight_done);
 }

drivers/nvdimm/nd_virtio.c

@@ -44,6 +44,15 @@ static int virtio_pmem_flush(struct nd_region *nd_region)
 	unsigned long flags;
 	int err, err1;
 
+	/*
+	 * Don't bother to submit the request to the device if the device is
+	 * not activated.
+	 */
+	if (vdev->config->get_status(vdev) & VIRTIO_CONFIG_S_NEEDS_RESET) {
+		dev_info(&vdev->dev, "virtio pmem device needs a reset\n");
+		return -EIO;
+	}
+
 	might_sleep();
 	req_data = kmalloc(sizeof(*req_data), GFP_KERNEL);
 	if (!req_data)

drivers/vdpa/ifcvf/ifcvf_base.h

@@ -112,15 +112,12 @@ void ifcvf_write_dev_config(struct ifcvf_hw *hw, u64 offset,
 			    const void *src, int length);
 u8 ifcvf_get_status(struct ifcvf_hw *hw);
 void ifcvf_set_status(struct ifcvf_hw *hw, u8 status);
-void io_write64_twopart(u64 val, u32 *lo, u32 *hi);
 void ifcvf_reset(struct ifcvf_hw *hw);
 u64 ifcvf_get_dev_features(struct ifcvf_hw *hw);
 u64 ifcvf_get_hw_features(struct ifcvf_hw *hw);
 int ifcvf_verify_min_features(struct ifcvf_hw *hw, u64 features);
 u16 ifcvf_get_vq_state(struct ifcvf_hw *hw, u16 qid);
 int ifcvf_set_vq_state(struct ifcvf_hw *hw, u16 qid, u16 num);
-struct ifcvf_adapter *vf_to_adapter(struct ifcvf_hw *hw);
-int ifcvf_probed_virtio_net(struct ifcvf_hw *hw);
 u32 ifcvf_get_config_size(struct ifcvf_hw *hw);
 u16 ifcvf_set_vq_vector(struct ifcvf_hw *hw, u16 qid, int vector);
 u16 ifcvf_set_config_vector(struct ifcvf_hw *hw, int vector);

drivers/vdpa/mlx5/core/mlx5_vdpa.h

@@ -83,10 +83,28 @@ enum {
 	MLX5_VDPA_NUM_AS = 2
 };
 
+struct mlx5_vdpa_mr_resources {
+	struct mlx5_vdpa_mr *mr[MLX5_VDPA_NUM_AS];
+	unsigned int group2asid[MLX5_VDPA_NUMVQ_GROUPS];
+
+	/* Pre-deletion mr list */
+	struct list_head mr_list_head;
+
+	/* Deferred mr list */
+	struct list_head mr_gc_list_head;
+	struct workqueue_struct *wq_gc;
+	struct delayed_work gc_dwork_ent;
+
+	struct mutex lock;
+
+	atomic_t shutdown;
+};
+
 struct mlx5_vdpa_dev {
 	struct vdpa_device vdev;
 	struct mlx5_core_dev *mdev;
 	struct mlx5_vdpa_resources res;
+	struct mlx5_vdpa_mr_resources mres;
 
 	u64 mlx_features;
 	u64 actual_features;
@@ -95,14 +113,23 @@ struct mlx5_vdpa_dev {
 	u16 max_idx;
 	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;
 	struct workqueue_struct *wq;
-	unsigned int group2asid[MLX5_VDPA_NUMVQ_GROUPS];
 	bool suspended;
+
+	struct mlx5_async_ctx async_ctx;
+};
+
+struct mlx5_vdpa_async_cmd {
+	int err;
+	struct mlx5_async_work cb_work;
+	struct completion cmd_done;
+
+	void *in;
+	size_t inlen;
+
+	void *out;
+	size_t outlen;
 };
 
 int mlx5_vdpa_create_tis(struct mlx5_vdpa_dev *mvdev, void *in, u32 *tisn);
@@ -121,7 +148,9 @@ int mlx5_vdpa_create_mkey(struct mlx5_vdpa_dev *mvdev, u32 *mkey, u32 *in,
 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);
+int mlx5_vdpa_init_mr_resources(struct mlx5_vdpa_dev *mvdev);
 void mlx5_vdpa_destroy_mr_resources(struct mlx5_vdpa_dev *mvdev);
+void mlx5_vdpa_clean_mrs(struct mlx5_vdpa_dev *mvdev);
 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,
@@ -134,6 +163,14 @@ int mlx5_vdpa_update_cvq_iotlb(struct mlx5_vdpa_dev *mvdev,
 				unsigned int asid);
 int mlx5_vdpa_create_dma_mr(struct mlx5_vdpa_dev *mvdev);
 int mlx5_vdpa_reset_mr(struct mlx5_vdpa_dev *mvdev, unsigned int asid);
+int mlx5_vdpa_exec_async_cmds(struct mlx5_vdpa_dev *mvdev,
+			      struct mlx5_vdpa_async_cmd *cmds,
+			      int num_cmds);
+
+#define mlx5_vdpa_err(__dev, format, ...)                                                          \
+	dev_err((__dev)->mdev->device, "%s:%d:(pid %d) error: " format, __func__, __LINE__,        \
+		 current->pid, ##__VA_ARGS__)
+
 
 #define mlx5_vdpa_warn(__dev, format, ...)                                                         \
 	dev_warn((__dev)->mdev->device, "%s:%d:(pid %d) warning: " format, __func__, __LINE__,     \

drivers/vdpa/mlx5/core/mr.c

@@ -49,17 +49,23 @@ static void populate_mtts(struct mlx5_vdpa_direct_mr *mr, __be64 *mtt)
 	}
 }
 
-static int create_direct_mr(struct mlx5_vdpa_dev *mvdev, struct mlx5_vdpa_direct_mr *mr)
+struct mlx5_create_mkey_mem {
-{
+	u8 out[MLX5_ST_SZ_BYTES(create_mkey_out)];
-	int inlen;
+	u8 in[MLX5_ST_SZ_BYTES(create_mkey_in)];
-	void *mkc;
+	__be64 mtt[];
-	void *in;
+};
-	int err;
 
-	inlen = MLX5_ST_SZ_BYTES(create_mkey_in) + roundup(MLX5_ST_SZ_BYTES(mtt) * mr->nsg, 16);
+struct mlx5_destroy_mkey_mem {
-	in = kvzalloc(inlen, GFP_KERNEL);
+	u8 out[MLX5_ST_SZ_BYTES(destroy_mkey_out)];
-	if (!in)
+	u8 in[MLX5_ST_SZ_BYTES(destroy_mkey_in)];
-		return -ENOMEM;
+};
+
+static void fill_create_direct_mr(struct mlx5_vdpa_dev *mvdev,
+				  struct mlx5_vdpa_direct_mr *mr,
+				  struct mlx5_create_mkey_mem *mem)
+{
+	void *in = &mem->in;
+	void *mkc;
 
 	MLX5_SET(create_mkey_in, in, uid, mvdev->res.uid);
 	mkc = MLX5_ADDR_OF(create_mkey_in, in, memory_key_mkey_entry);
@@ -76,18 +82,36 @@ static int create_direct_mr(struct mlx5_vdpa_dev *mvdev, struct mlx5_vdpa_direct
 	MLX5_SET(create_mkey_in, in, translations_octword_actual_size,
 		 get_octo_len(mr->end - mr->start, mr->log_size));
 	populate_mtts(mr, MLX5_ADDR_OF(create_mkey_in, in, klm_pas_mtt));
-	err = mlx5_vdpa_create_mkey(mvdev, &mr->mr, in, inlen);
-	kvfree(in);
-	if (err) {
-		mlx5_vdpa_warn(mvdev, "Failed to create direct MR\n");
-		return err;
-	}
 
-	return 0;
+	MLX5_SET(create_mkey_in, in, opcode, MLX5_CMD_OP_CREATE_MKEY);
+	MLX5_SET(create_mkey_in, in, uid, mvdev->res.uid);
+}
+
+static void create_direct_mr_end(struct mlx5_vdpa_dev *mvdev,
+				 struct mlx5_vdpa_direct_mr *mr,
+				 struct mlx5_create_mkey_mem *mem)
+{
+	u32 mkey_index = MLX5_GET(create_mkey_out, mem->out, mkey_index);
+
+	mr->mr = mlx5_idx_to_mkey(mkey_index);
+}
+
+static void fill_destroy_direct_mr(struct mlx5_vdpa_dev *mvdev,
+				   struct mlx5_vdpa_direct_mr *mr,
+				   struct mlx5_destroy_mkey_mem *mem)
+{
+	void *in = &mem->in;
+
+	MLX5_SET(destroy_mkey_in, in, uid, mvdev->res.uid);
+	MLX5_SET(destroy_mkey_in, in, opcode, MLX5_CMD_OP_DESTROY_MKEY);
+	MLX5_SET(destroy_mkey_in, in, mkey_index, mlx5_mkey_to_idx(mr->mr));
 }
 
 static void destroy_direct_mr(struct mlx5_vdpa_dev *mvdev, struct mlx5_vdpa_direct_mr *mr)
 {
+	if (!mr->mr)
+		return;
+
 	mlx5_vdpa_destroy_mkey(mvdev, mr->mr);
 }
 
@@ -179,6 +203,123 @@ static int klm_byte_size(int nklms)
 	return 16 * ALIGN(nklms, 4);
 }
 
+#define MLX5_VDPA_MTT_ALIGN 16
+
+static int create_direct_keys(struct mlx5_vdpa_dev *mvdev, struct mlx5_vdpa_mr *mr)
+{
+	struct mlx5_vdpa_async_cmd *cmds;
+	struct mlx5_vdpa_direct_mr *dmr;
+	int err = 0;
+	int i = 0;
+
+	cmds = kvcalloc(mr->num_directs, sizeof(*cmds), GFP_KERNEL);
+	if (!cmds)
+		return -ENOMEM;
+
+	list_for_each_entry(dmr, &mr->head, list) {
+		struct mlx5_create_mkey_mem *cmd_mem;
+		int mttlen, mttcount;
+
+		mttlen = roundup(MLX5_ST_SZ_BYTES(mtt) * dmr->nsg, MLX5_VDPA_MTT_ALIGN);
+		mttcount = mttlen / sizeof(cmd_mem->mtt[0]);
+		cmd_mem = kvcalloc(1, struct_size(cmd_mem, mtt, mttcount), GFP_KERNEL);
+		if (!cmd_mem) {
+			err = -ENOMEM;
+			goto done;
+		}
+
+		cmds[i].out = cmd_mem->out;
+		cmds[i].outlen = sizeof(cmd_mem->out);
+		cmds[i].in = cmd_mem->in;
+		cmds[i].inlen = struct_size(cmd_mem, mtt, mttcount);
+
+		fill_create_direct_mr(mvdev, dmr, cmd_mem);
+
+		i++;
+	}
+
+	err = mlx5_vdpa_exec_async_cmds(mvdev, cmds, mr->num_directs);
+	if (err) {
+
+		mlx5_vdpa_err(mvdev, "error issuing MTT mkey creation for direct mrs: %d\n", err);
+		goto done;
+	}
+
+	i = 0;
+	list_for_each_entry(dmr, &mr->head, list) {
+		struct mlx5_vdpa_async_cmd *cmd = &cmds[i++];
+		struct mlx5_create_mkey_mem *cmd_mem;
+
+		cmd_mem = container_of(cmd->out, struct mlx5_create_mkey_mem, out);
+
+		if (!cmd->err) {
+			create_direct_mr_end(mvdev, dmr, cmd_mem);
+		} else {
+			err = err ? err : cmd->err;
+			mlx5_vdpa_err(mvdev, "error creating MTT mkey [0x%llx, 0x%llx]: %d\n",
+				dmr->start, dmr->end, cmd->err);
+		}
+	}
+
+done:
+	for (i = i-1; i >= 0; i--) {
+		struct mlx5_create_mkey_mem *cmd_mem;
+
+		cmd_mem = container_of(cmds[i].out, struct mlx5_create_mkey_mem, out);
+		kvfree(cmd_mem);
+	}
+
+	kvfree(cmds);
+	return err;
+}
+
+DEFINE_FREE(free_cmds, struct mlx5_vdpa_async_cmd *, kvfree(_T))
+DEFINE_FREE(free_cmd_mem, struct mlx5_destroy_mkey_mem *, kvfree(_T))
+
+static int destroy_direct_keys(struct mlx5_vdpa_dev *mvdev, struct mlx5_vdpa_mr *mr)
+{
+	struct mlx5_destroy_mkey_mem *cmd_mem __free(free_cmd_mem) = NULL;
+	struct mlx5_vdpa_async_cmd *cmds __free(free_cmds) = NULL;
+	struct mlx5_vdpa_direct_mr *dmr;
+	int err = 0;
+	int i = 0;
+
+	cmds = kvcalloc(mr->num_directs, sizeof(*cmds), GFP_KERNEL);
+	cmd_mem = kvcalloc(mr->num_directs, sizeof(*cmd_mem), GFP_KERNEL);
+	if (!cmds || !cmd_mem)
+		return -ENOMEM;
+
+	list_for_each_entry(dmr, &mr->head, list) {
+		cmds[i].out = cmd_mem[i].out;
+		cmds[i].outlen = sizeof(cmd_mem[i].out);
+		cmds[i].in = cmd_mem[i].in;
+		cmds[i].inlen = sizeof(cmd_mem[i].in);
+		fill_destroy_direct_mr(mvdev, dmr, &cmd_mem[i]);
+		i++;
+	}
+
+	err = mlx5_vdpa_exec_async_cmds(mvdev, cmds, mr->num_directs);
+	if (err) {
+
+		mlx5_vdpa_err(mvdev, "error issuing MTT mkey deletion for direct mrs: %d\n", err);
+		return err;
+	}
+
+	i = 0;
+	list_for_each_entry(dmr, &mr->head, list) {
+		struct mlx5_vdpa_async_cmd *cmd = &cmds[i++];
+
+		dmr->mr = 0;
+		if (cmd->err) {
+			err = err ? err : cmd->err;
+			mlx5_vdpa_err(mvdev, "error deleting MTT mkey [0x%llx, 0x%llx]: %d\n",
+				dmr->start, dmr->end, cmd->err);
+		}
+	}
+
+	return err;
+}
+
 static int create_indirect_key(struct mlx5_vdpa_dev *mvdev, struct mlx5_vdpa_mr *mr)
 {
 	int inlen;
@@ -279,14 +420,8 @@ done:
 		goto err_map;
 	}
 
-	err = create_direct_mr(mvdev, mr);
-	if (err)
-		goto err_direct;
-
 	return 0;
 
-err_direct:
-	dma_unmap_sg_attrs(dma, mr->sg_head.sgl, mr->nsg, DMA_BIDIRECTIONAL, 0);
 err_map:
 	sg_free_table(&mr->sg_head);
 	return err;
@@ -401,6 +536,10 @@ static int create_user_mr(struct mlx5_vdpa_dev *mvdev,
 	if (err)
 		goto err_chain;
 
+	err = create_direct_keys(mvdev, mr);
+	if (err)
+		goto err_chain;
+
 	/* Create the memory key that defines the guests's address space. This
 	 * memory key refers to the direct keys that contain the MTT
 	 * translations
@@ -489,6 +628,7 @@ static void destroy_user_mr(struct mlx5_vdpa_dev *mvdev, struct mlx5_vdpa_mr *mr
 	struct mlx5_vdpa_direct_mr *n;
 
 	destroy_indirect_key(mvdev, mr);
+	destroy_direct_keys(mvdev, mr);
 	list_for_each_entry_safe_reverse(dmr, n, &mr->head, list) {
 		list_del_init(&dmr->list);
 		unmap_direct_mr(mvdev, dmr);
@@ -513,22 +653,58 @@ static void _mlx5_vdpa_destroy_mr(struct mlx5_vdpa_dev *mvdev, struct mlx5_vdpa_
 	kfree(mr);
 }
 
+/* There can be multiple .set_map() operations in quick succession.
+ * This large delay is a simple way to prevent the MR cleanup from blocking
+ * .set_map() MR creation in this scenario.
+ */
+#define MLX5_VDPA_MR_GC_TRIGGER_MS 2000
+
+static void mlx5_vdpa_mr_gc_handler(struct work_struct *work)
+{
+	struct mlx5_vdpa_mr_resources *mres;
+	struct mlx5_vdpa_mr *mr, *tmp;
+	struct mlx5_vdpa_dev *mvdev;
+
+	mres = container_of(work, struct mlx5_vdpa_mr_resources, gc_dwork_ent.work);
+
+	if (atomic_read(&mres->shutdown)) {
+		mutex_lock(&mres->lock);
+	} else if (!mutex_trylock(&mres->lock)) {
+		queue_delayed_work(mres->wq_gc, &mres->gc_dwork_ent,
+				   msecs_to_jiffies(MLX5_VDPA_MR_GC_TRIGGER_MS));
+		return;
+	}
+
+	mvdev = container_of(mres, struct mlx5_vdpa_dev, mres);
+
+	list_for_each_entry_safe(mr, tmp, &mres->mr_gc_list_head, mr_list) {
+		_mlx5_vdpa_destroy_mr(mvdev, mr);
+	}
+
+	mutex_unlock(&mres->lock);
+}
+
 static void _mlx5_vdpa_put_mr(struct mlx5_vdpa_dev *mvdev,
 			      struct mlx5_vdpa_mr *mr)
 {
+	struct mlx5_vdpa_mr_resources *mres = &mvdev->mres;
+
 	if (!mr)
 		return;
 
-	if (refcount_dec_and_test(&mr->refcount))
+	if (refcount_dec_and_test(&mr->refcount)) {
-		_mlx5_vdpa_destroy_mr(mvdev, mr);
+		list_move_tail(&mr->mr_list, &mres->mr_gc_list_head);
+		queue_delayed_work(mres->wq_gc, &mres->gc_dwork_ent,
+				   msecs_to_jiffies(MLX5_VDPA_MR_GC_TRIGGER_MS));
+	}
 }
 
 void mlx5_vdpa_put_mr(struct mlx5_vdpa_dev *mvdev,
 		      struct mlx5_vdpa_mr *mr)
 {
-	mutex_lock(&mvdev->mr_mtx);
+	mutex_lock(&mvdev->mres.lock);
 	_mlx5_vdpa_put_mr(mvdev, mr);
-	mutex_unlock(&mvdev->mr_mtx);
+	mutex_unlock(&mvdev->mres.lock);
 }
 
 static void _mlx5_vdpa_get_mr(struct mlx5_vdpa_dev *mvdev,
@@ -543,44 +719,47 @@ static void _mlx5_vdpa_get_mr(struct mlx5_vdpa_dev *mvdev,
 void mlx5_vdpa_get_mr(struct mlx5_vdpa_dev *mvdev,
 		      struct mlx5_vdpa_mr *mr)
 {
-	mutex_lock(&mvdev->mr_mtx);
+	mutex_lock(&mvdev->mres.lock);
 	_mlx5_vdpa_get_mr(mvdev, mr);
-	mutex_unlock(&mvdev->mr_mtx);
+	mutex_unlock(&mvdev->mres.lock);
 }
 
 void mlx5_vdpa_update_mr(struct mlx5_vdpa_dev *mvdev,
 			 struct mlx5_vdpa_mr *new_mr,
 			 unsigned int asid)
 {
-	struct mlx5_vdpa_mr *old_mr = mvdev->mr[asid];
+	struct mlx5_vdpa_mr *old_mr = mvdev->mres.mr[asid];
 
-	mutex_lock(&mvdev->mr_mtx);
+	mutex_lock(&mvdev->mres.lock);
 
 	_mlx5_vdpa_put_mr(mvdev, old_mr);
-	mvdev->mr[asid] = new_mr;
+	mvdev->mres.mr[asid] = new_mr;
 
-	mutex_unlock(&mvdev->mr_mtx);
+	mutex_unlock(&mvdev->mres.lock);
 }
 
 static void mlx5_vdpa_show_mr_leaks(struct mlx5_vdpa_dev *mvdev)
 {
 	struct mlx5_vdpa_mr *mr;
 
-	mutex_lock(&mvdev->mr_mtx);
+	mutex_lock(&mvdev->mres.lock);
 
-	list_for_each_entry(mr, &mvdev->mr_list_head, mr_list) {
+	list_for_each_entry(mr, &mvdev->mres.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);
+	mutex_unlock(&mvdev->mres.lock);
 
 }
 
-void mlx5_vdpa_destroy_mr_resources(struct mlx5_vdpa_dev *mvdev)
+void mlx5_vdpa_clean_mrs(struct mlx5_vdpa_dev *mvdev)
 {
+	if (!mvdev->res.valid)
+		return;
+
 	for (int i = 0; i < MLX5_VDPA_NUM_AS; i++)
 		mlx5_vdpa_update_mr(mvdev, NULL, i);
 
@@ -613,7 +792,7 @@ 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);
+	list_add_tail(&mr->mr_list, &mvdev->mres.mr_list_head);
 
 	return 0;
 
@@ -639,9 +818,9 @@ struct mlx5_vdpa_mr *mlx5_vdpa_create_mr(struct mlx5_vdpa_dev *mvdev,
 	if (!mr)
 		return ERR_PTR(-ENOMEM);
 
-	mutex_lock(&mvdev->mr_mtx);
+	mutex_lock(&mvdev->mres.lock);
 	err = _mlx5_vdpa_create_mr(mvdev, mr, iotlb);
-	mutex_unlock(&mvdev->mr_mtx);
+	mutex_unlock(&mvdev->mres.lock);
 
 	if (err)
 		goto out_err;
@@ -661,7 +840,7 @@ int mlx5_vdpa_update_cvq_iotlb(struct mlx5_vdpa_dev *mvdev,
 {
 	int err;
 
-	if (mvdev->group2asid[MLX5_VDPA_CVQ_GROUP] != asid)
+	if (mvdev->mres.group2asid[MLX5_VDPA_CVQ_GROUP] != asid)
 		return 0;
 
 	spin_lock(&mvdev->cvq.iommu_lock);
@@ -703,3 +882,33 @@ int mlx5_vdpa_reset_mr(struct mlx5_vdpa_dev *mvdev, unsigned int asid)
 
 	return 0;
 }
+
+int mlx5_vdpa_init_mr_resources(struct mlx5_vdpa_dev *mvdev)
+{
+	struct mlx5_vdpa_mr_resources *mres = &mvdev->mres;
+
+	mres->wq_gc = create_singlethread_workqueue("mlx5_vdpa_mr_gc");
+	if (!mres->wq_gc)
+		return -ENOMEM;
+
+	INIT_DELAYED_WORK(&mres->gc_dwork_ent, mlx5_vdpa_mr_gc_handler);
+
+	mutex_init(&mres->lock);
+
+	INIT_LIST_HEAD(&mres->mr_list_head);
+	INIT_LIST_HEAD(&mres->mr_gc_list_head);
+
+	return 0;
+}
+
+void mlx5_vdpa_destroy_mr_resources(struct mlx5_vdpa_dev *mvdev)
+{
+	struct mlx5_vdpa_mr_resources *mres = &mvdev->mres;
+
+	atomic_set(&mres->shutdown, 1);
+
+	flush_delayed_work(&mres->gc_dwork_ent);
+	destroy_workqueue(mres->wq_gc);
+	mres->wq_gc = NULL;
+	mutex_destroy(&mres->lock);
+}

drivers/vdpa/mlx5/core/resources.c

@@ -256,7 +256,6 @@ int mlx5_vdpa_alloc_resources(struct mlx5_vdpa_dev *mvdev)
 		mlx5_vdpa_warn(mvdev, "resources already allocated\n");
 		return -EINVAL;
 	}
-	mutex_init(&mvdev->mr_mtx);
 	res->uar = mlx5_get_uars_page(mdev);
 	if (IS_ERR(res->uar)) {
 		err = PTR_ERR(res->uar);
@@ -301,7 +300,6 @@ err_pd:
 err_uctx:
 	mlx5_put_uars_page(mdev, res->uar);
 err_uars:
-	mutex_destroy(&mvdev->mr_mtx);
 	return err;
 }
 
@@ -318,6 +316,78 @@ void mlx5_vdpa_free_resources(struct mlx5_vdpa_dev *mvdev)
 	dealloc_pd(mvdev, res->pdn, res->uid);
 	destroy_uctx(mvdev, res->uid);
 	mlx5_put_uars_page(mvdev->mdev, res->uar);
-	mutex_destroy(&mvdev->mr_mtx);
 	res->valid = false;
 }
+
+static void virtqueue_cmd_callback(int status, struct mlx5_async_work *context)
+{
+	struct mlx5_vdpa_async_cmd *cmd =
+		container_of(context, struct mlx5_vdpa_async_cmd, cb_work);
+
+	cmd->err = mlx5_cmd_check(context->ctx->dev, status, cmd->in, cmd->out);
+	complete(&cmd->cmd_done);
+}
+
+static int issue_async_cmd(struct mlx5_vdpa_dev *mvdev,
+			   struct mlx5_vdpa_async_cmd *cmds,
+			   int issued,
+			   int *completed)
+
+{
+	struct mlx5_vdpa_async_cmd *cmd = &cmds[issued];
+	int err;
+
+retry:
+	err = mlx5_cmd_exec_cb(&mvdev->async_ctx,
+			       cmd->in, cmd->inlen,
+			       cmd->out, cmd->outlen,
+			       virtqueue_cmd_callback,
+			       &cmd->cb_work);
+	if (err == -EBUSY) {
+		if (*completed < issued) {
+			/* Throttled by own commands: wait for oldest completion. */
+			wait_for_completion(&cmds[*completed].cmd_done);
+			(*completed)++;
+
+			goto retry;
+		} else {
+			/* Throttled by external commands: switch to sync api. */
+			err = mlx5_cmd_exec(mvdev->mdev,
+					    cmd->in, cmd->inlen,
+					    cmd->out, cmd->outlen);
+			if (!err)
+				(*completed)++;
+		}
+	}
+
+	return err;
+}
+
+int mlx5_vdpa_exec_async_cmds(struct mlx5_vdpa_dev *mvdev,
+			      struct mlx5_vdpa_async_cmd *cmds,
+			      int num_cmds)
+{
+	int completed = 0;
+	int issued = 0;
+	int err = 0;
+
+	for (int i = 0; i < num_cmds; i++)
+		init_completion(&cmds[i].cmd_done);
+
+	while (issued < num_cmds) {
+
+		err = issue_async_cmd(mvdev, cmds, issued, &completed);
+		if (err) {
+			mlx5_vdpa_err(mvdev, "error issuing command %d of %d: %d\n",
+				      issued, num_cmds, err);
+			break;
+		}
+
+		issued++;
+	}
+
+	while (completed < issued)
+		wait_for_completion(&cmds[completed++].cmd_done);
+
+	return err;
+}

drivers/vdpa/mlx5/net/mlx5_vnet.c

@@ -941,11 +941,11 @@ static int create_virtqueue(struct mlx5_vdpa_net *ndev,
 		MLX5_SET64(virtio_q, vq_ctx, used_addr, mvq->device_addr);
 		MLX5_SET64(virtio_q, vq_ctx, available_addr, mvq->driver_addr);
 
-		vq_mr = mvdev->mr[mvdev->group2asid[MLX5_VDPA_DATAVQ_GROUP]];
+		vq_mr = mvdev->mres.mr[mvdev->mres.group2asid[MLX5_VDPA_DATAVQ_GROUP]];
 		if (vq_mr)
 			MLX5_SET(virtio_q, vq_ctx, virtio_q_mkey, vq_mr->mkey);
 
-		vq_desc_mr = mvdev->mr[mvdev->group2asid[MLX5_VDPA_DATAVQ_DESC_GROUP]];
+		vq_desc_mr = mvdev->mres.mr[mvdev->mres.group2asid[MLX5_VDPA_DATAVQ_DESC_GROUP]];
 		if (vq_desc_mr &&
 		    MLX5_CAP_DEV_VDPA_EMULATION(mvdev->mdev, desc_group_mkey_supported))
 			MLX5_SET(virtio_q, vq_ctx, desc_group_mkey, vq_desc_mr->mkey);
@@ -953,11 +953,11 @@ static int create_virtqueue(struct mlx5_vdpa_net *ndev,
 		/* If there is no mr update, make sure that the existing ones are set
 		 * modify to ready.
 		 */
-		vq_mr = mvdev->mr[mvdev->group2asid[MLX5_VDPA_DATAVQ_GROUP]];
+		vq_mr = mvdev->mres.mr[mvdev->mres.group2asid[MLX5_VDPA_DATAVQ_GROUP]];
 		if (vq_mr)
 			mvq->modified_fields |= MLX5_VIRTQ_MODIFY_MASK_VIRTIO_Q_MKEY;
 
-		vq_desc_mr = mvdev->mr[mvdev->group2asid[MLX5_VDPA_DATAVQ_DESC_GROUP]];
+		vq_desc_mr = mvdev->mres.mr[mvdev->mres.group2asid[MLX5_VDPA_DATAVQ_DESC_GROUP]];
 		if (vq_desc_mr)
 			mvq->modified_fields |= MLX5_VIRTQ_MODIFY_MASK_DESC_GROUP_MKEY;
 	}
@@ -1184,40 +1184,92 @@ struct mlx5_virtq_attr {
 	u16 used_index;
 };
 
-static int query_virtqueue(struct mlx5_vdpa_net *ndev, struct mlx5_vdpa_virtqueue *mvq,
+struct mlx5_virtqueue_query_mem {
-			   struct mlx5_virtq_attr *attr)
+	u8 in[MLX5_ST_SZ_BYTES(query_virtio_net_q_in)];
+	u8 out[MLX5_ST_SZ_BYTES(query_virtio_net_q_out)];
+};
+
+struct mlx5_virtqueue_modify_mem {
+	u8 in[MLX5_ST_SZ_BYTES(modify_virtio_net_q_in)];
+	u8 out[MLX5_ST_SZ_BYTES(modify_virtio_net_q_out)];
+};
+
+static void fill_query_virtqueue_cmd(struct mlx5_vdpa_net *ndev,
+				     struct mlx5_vdpa_virtqueue *mvq,
+				     struct mlx5_virtqueue_query_mem *cmd)
 {
-	int outlen = MLX5_ST_SZ_BYTES(query_virtio_net_q_out);
+	void *cmd_hdr = MLX5_ADDR_OF(query_virtio_net_q_in, cmd->in, general_obj_in_cmd_hdr);
-	u32 in[MLX5_ST_SZ_DW(query_virtio_net_q_in)] = {};
-	void *out;
-	void *obj_context;
-	void *cmd_hdr;
-	int err;
-
-	out = kzalloc(outlen, GFP_KERNEL);
-	if (!out)
-		return -ENOMEM;
-
-	cmd_hdr = MLX5_ADDR_OF(query_virtio_net_q_in, in, general_obj_in_cmd_hdr);
 
 	MLX5_SET(general_obj_in_cmd_hdr, cmd_hdr, opcode, MLX5_CMD_OP_QUERY_GENERAL_OBJECT);
 	MLX5_SET(general_obj_in_cmd_hdr, cmd_hdr, obj_type, MLX5_OBJ_TYPE_VIRTIO_NET_Q);
 	MLX5_SET(general_obj_in_cmd_hdr, cmd_hdr, obj_id, mvq->virtq_id);
 	MLX5_SET(general_obj_in_cmd_hdr, cmd_hdr, uid, ndev->mvdev.res.uid);
-	err = mlx5_cmd_exec(ndev->mvdev.mdev, in, sizeof(in), out, outlen);
+}
-	if (err)
+
-		goto err_cmd;
+static void query_virtqueue_end(struct mlx5_vdpa_net *ndev,
+				struct mlx5_virtqueue_query_mem *cmd,
+				struct mlx5_virtq_attr *attr)
+{
+	void *obj_context = MLX5_ADDR_OF(query_virtio_net_q_out, cmd->out, obj_context);
 
-	obj_context = MLX5_ADDR_OF(query_virtio_net_q_out, out, obj_context);
 	memset(attr, 0, sizeof(*attr));
 	attr->state = MLX5_GET(virtio_net_q_object, obj_context, state);
 	attr->available_index = MLX5_GET(virtio_net_q_object, obj_context, hw_available_index);
 	attr->used_index = MLX5_GET(virtio_net_q_object, obj_context, hw_used_index);
-	kfree(out);
+}
-	return 0;
 
-err_cmd:
+static int query_virtqueues(struct mlx5_vdpa_net *ndev,
-	kfree(out);
+			    int start_vq,
+			    int num_vqs,
+			    struct mlx5_virtq_attr *attrs)
+{
+	struct mlx5_vdpa_dev *mvdev = &ndev->mvdev;
+	struct mlx5_virtqueue_query_mem *cmd_mem;
+	struct mlx5_vdpa_async_cmd *cmds;
+	int err = 0;
+
+	WARN(start_vq + num_vqs > mvdev->max_vqs, "query vq range invalid [%d, %d), max_vqs: %u\n",
+	     start_vq, start_vq + num_vqs, mvdev->max_vqs);
+
+	cmds = kvcalloc(num_vqs, sizeof(*cmds), GFP_KERNEL);
+	cmd_mem = kvcalloc(num_vqs, sizeof(*cmd_mem), GFP_KERNEL);
+	if (!cmds || !cmd_mem) {
+		err = -ENOMEM;
+		goto done;
+	}
+
+	for (int i = 0; i < num_vqs; i++) {
+		cmds[i].in = &cmd_mem[i].in;
+		cmds[i].inlen = sizeof(cmd_mem[i].in);
+		cmds[i].out = &cmd_mem[i].out;
+		cmds[i].outlen = sizeof(cmd_mem[i].out);
+		fill_query_virtqueue_cmd(ndev, &ndev->vqs[start_vq + i], &cmd_mem[i]);
+	}
+
+	err = mlx5_vdpa_exec_async_cmds(&ndev->mvdev, cmds, num_vqs);
+	if (err) {
+		mlx5_vdpa_err(mvdev, "error issuing query cmd for vq range [%d, %d): %d\n",
+			      start_vq, start_vq + num_vqs, err);
+		goto done;
+	}
+
+	for (int i = 0; i < num_vqs; i++) {
+		struct mlx5_vdpa_async_cmd *cmd = &cmds[i];
+		int vq_idx = start_vq + i;
+
+		if (cmd->err) {
+			mlx5_vdpa_err(mvdev, "query vq %d failed, err: %d\n", vq_idx, err);
+			if (!err)
+				err = cmd->err;
+			continue;
+		}
+
+		query_virtqueue_end(ndev, &cmd_mem[i], &attrs[i]);
+	}
+
+done:
+	kvfree(cmd_mem);
+	kvfree(cmds);
 	return err;
 }
 
@@ -1251,51 +1303,30 @@ static bool modifiable_virtqueue_fields(struct mlx5_vdpa_virtqueue *mvq)
 	return true;
 }
 
-static int modify_virtqueue(struct mlx5_vdpa_net *ndev,
+static void fill_modify_virtqueue_cmd(struct mlx5_vdpa_net *ndev,
-			    struct mlx5_vdpa_virtqueue *mvq,
+				      struct mlx5_vdpa_virtqueue *mvq,
-			    int state)
+				      int state,
+				      struct mlx5_virtqueue_modify_mem *cmd)
 {
-	int inlen = MLX5_ST_SZ_BYTES(modify_virtio_net_q_in);
-	u32 out[MLX5_ST_SZ_DW(modify_virtio_net_q_out)] = {};
 	struct mlx5_vdpa_dev *mvdev = &ndev->mvdev;
 	struct mlx5_vdpa_mr *desc_mr = NULL;
 	struct mlx5_vdpa_mr *vq_mr = NULL;
-	bool state_change = false;
 	void *obj_context;
 	void *cmd_hdr;
 	void *vq_ctx;
-	void *in;
-	int err;
 
-	if (mvq->fw_state == MLX5_VIRTIO_NET_Q_OBJECT_NONE)
+	cmd_hdr = MLX5_ADDR_OF(modify_virtio_net_q_in, cmd->in, general_obj_in_cmd_hdr);
-		return 0;
-
-	if (!modifiable_virtqueue_fields(mvq))
-		return -EINVAL;
-
-	in = kzalloc(inlen, GFP_KERNEL);
-	if (!in)
-		return -ENOMEM;
-
-	cmd_hdr = MLX5_ADDR_OF(modify_virtio_net_q_in, in, general_obj_in_cmd_hdr);
 
 	MLX5_SET(general_obj_in_cmd_hdr, cmd_hdr, opcode, MLX5_CMD_OP_MODIFY_GENERAL_OBJECT);
 	MLX5_SET(general_obj_in_cmd_hdr, cmd_hdr, obj_type, MLX5_OBJ_TYPE_VIRTIO_NET_Q);
 	MLX5_SET(general_obj_in_cmd_hdr, cmd_hdr, obj_id, mvq->virtq_id);
 	MLX5_SET(general_obj_in_cmd_hdr, cmd_hdr, uid, ndev->mvdev.res.uid);
 
-	obj_context = MLX5_ADDR_OF(modify_virtio_net_q_in, in, obj_context);
+	obj_context = MLX5_ADDR_OF(modify_virtio_net_q_in, cmd->in, obj_context);
 	vq_ctx = MLX5_ADDR_OF(virtio_net_q_object, obj_context, virtio_q_context);
 
-	if (mvq->modified_fields & MLX5_VIRTQ_MODIFY_MASK_STATE) {
+	if (mvq->modified_fields & MLX5_VIRTQ_MODIFY_MASK_STATE)
-		if (!is_valid_state_change(mvq->fw_state, state, is_resumable(ndev))) {
-			err = -EINVAL;
-			goto done;
-		}
-
 		MLX5_SET(virtio_net_q_object, obj_context, state, state);
-		state_change = true;
-	}
 
 	if (mvq->modified_fields & MLX5_VIRTQ_MODIFY_MASK_VIRTIO_Q_ADDRS) {
 		MLX5_SET64(virtio_q, vq_ctx, desc_addr, mvq->desc_addr);
@@ -1323,7 +1354,7 @@ static int modify_virtqueue(struct mlx5_vdpa_net *ndev,
 	}
 
 	if (mvq->modified_fields & MLX5_VIRTQ_MODIFY_MASK_VIRTIO_Q_MKEY) {
-		vq_mr = mvdev->mr[mvdev->group2asid[MLX5_VDPA_DATAVQ_GROUP]];
+		vq_mr = mvdev->mres.mr[mvdev->mres.group2asid[MLX5_VDPA_DATAVQ_GROUP]];
 
 		if (vq_mr)
 			MLX5_SET(virtio_q, vq_ctx, virtio_q_mkey, vq_mr->mkey);
@@ -1332,7 +1363,7 @@ static int modify_virtqueue(struct mlx5_vdpa_net *ndev,
 	}
 
 	if (mvq->modified_fields & MLX5_VIRTQ_MODIFY_MASK_DESC_GROUP_MKEY) {
-		desc_mr = mvdev->mr[mvdev->group2asid[MLX5_VDPA_DATAVQ_DESC_GROUP]];
+		desc_mr = mvdev->mres.mr[mvdev->mres.group2asid[MLX5_VDPA_DATAVQ_DESC_GROUP]];
 
 		if (desc_mr && MLX5_CAP_DEV_VDPA_EMULATION(mvdev->mdev, desc_group_mkey_supported))
 			MLX5_SET(virtio_q, vq_ctx, desc_group_mkey, desc_mr->mkey);
@@ -1341,38 +1372,36 @@ static int modify_virtqueue(struct mlx5_vdpa_net *ndev,
 	}
 
 	MLX5_SET64(virtio_net_q_object, obj_context, modify_field_select, mvq->modified_fields);
-	err = mlx5_cmd_exec(ndev->mvdev.mdev, in, inlen, out, sizeof(out));
+}
-	if (err)
-		goto done;
 
-	if (state_change)
+static void modify_virtqueue_end(struct mlx5_vdpa_net *ndev,
-		mvq->fw_state = state;
+				 struct mlx5_vdpa_virtqueue *mvq,
+				 int state)
+{
+	struct mlx5_vdpa_dev *mvdev = &ndev->mvdev;
 
 	if (mvq->modified_fields & MLX5_VIRTQ_MODIFY_MASK_VIRTIO_Q_MKEY) {
+		unsigned int asid = mvdev->mres.group2asid[MLX5_VDPA_DATAVQ_GROUP];
+		struct mlx5_vdpa_mr *vq_mr = mvdev->mres.mr[asid];
+
 		mlx5_vdpa_put_mr(mvdev, mvq->vq_mr);
 		mlx5_vdpa_get_mr(mvdev, vq_mr);
 		mvq->vq_mr = vq_mr;
 	}
 
 	if (mvq->modified_fields & MLX5_VIRTQ_MODIFY_MASK_DESC_GROUP_MKEY) {
+		unsigned int asid = mvdev->mres.group2asid[MLX5_VDPA_DATAVQ_DESC_GROUP];
+		struct mlx5_vdpa_mr *desc_mr = mvdev->mres.mr[asid];
+
 		mlx5_vdpa_put_mr(mvdev, mvq->desc_mr);
 		mlx5_vdpa_get_mr(mvdev, desc_mr);
 		mvq->desc_mr = desc_mr;
 	}
 
+	if (mvq->modified_fields & MLX5_VIRTQ_MODIFY_MASK_STATE)
+		mvq->fw_state = state;
+
 	mvq->modified_fields = 0;
-
-done:
-	kfree(in);
-	return err;
-}
-
-static int modify_virtqueue_state(struct mlx5_vdpa_net *ndev,
-				  struct mlx5_vdpa_virtqueue *mvq,
-				  unsigned int state)
-{
-	mvq->modified_fields |= MLX5_VIRTQ_MODIFY_MASK_STATE;
-	return modify_virtqueue(ndev, mvq, state);
 }
 
 static int counter_set_alloc(struct mlx5_vdpa_net *ndev, struct mlx5_vdpa_virtqueue *mvq)
@@ -1525,53 +1554,136 @@ err_fwqp:
 	return err;
 }
 
-static int suspend_vq(struct mlx5_vdpa_net *ndev, struct mlx5_vdpa_virtqueue *mvq)
+static int modify_virtqueues(struct mlx5_vdpa_net *ndev, int start_vq, int num_vqs, int state)
 {
-	struct mlx5_virtq_attr attr;
+	struct mlx5_vdpa_dev *mvdev = &ndev->mvdev;
+	struct mlx5_virtqueue_modify_mem *cmd_mem;
+	struct mlx5_vdpa_async_cmd *cmds;
+	int err = 0;
+
+	WARN(start_vq + num_vqs > mvdev->max_vqs, "modify vq range invalid [%d, %d), max_vqs: %u\n",
+	     start_vq, start_vq + num_vqs, mvdev->max_vqs);
+
+	cmds = kvcalloc(num_vqs, sizeof(*cmds), GFP_KERNEL);
+	cmd_mem = kvcalloc(num_vqs, sizeof(*cmd_mem), GFP_KERNEL);
+	if (!cmds || !cmd_mem) {
+		err = -ENOMEM;
+		goto done;
+	}
+
+	for (int i = 0; i < num_vqs; i++) {
+		struct mlx5_vdpa_async_cmd *cmd = &cmds[i];
+		struct mlx5_vdpa_virtqueue *mvq;
+		int vq_idx = start_vq + i;
+
+		mvq = &ndev->vqs[vq_idx];
+
+		if (!modifiable_virtqueue_fields(mvq)) {
+			err = -EINVAL;
+			goto done;
+		}
+
+		if (mvq->fw_state != state) {
+			if (!is_valid_state_change(mvq->fw_state, state, is_resumable(ndev))) {
+				err = -EINVAL;
+				goto done;
+			}
+
+			mvq->modified_fields |= MLX5_VIRTQ_MODIFY_MASK_STATE;
+		}
+
+		cmd->in = &cmd_mem[i].in;
+		cmd->inlen = sizeof(cmd_mem[i].in);
+		cmd->out = &cmd_mem[i].out;
+		cmd->outlen = sizeof(cmd_mem[i].out);
+		fill_modify_virtqueue_cmd(ndev, mvq, state, &cmd_mem[i]);
+	}
+
+	err = mlx5_vdpa_exec_async_cmds(&ndev->mvdev, cmds, num_vqs);
+	if (err) {
+		mlx5_vdpa_err(mvdev, "error issuing modify cmd for vq range [%d, %d)\n",
+			      start_vq, start_vq + num_vqs);
+		goto done;
+	}
+
+	for (int i = 0; i < num_vqs; i++) {
+		struct mlx5_vdpa_async_cmd *cmd = &cmds[i];
+		struct mlx5_vdpa_virtqueue *mvq;
+		int vq_idx = start_vq + i;
+
+		mvq = &ndev->vqs[vq_idx];
+
+		if (cmd->err) {
+			mlx5_vdpa_err(mvdev, "modify vq %d failed, state: %d -> %d, err: %d\n",
+				      vq_idx, mvq->fw_state, state, err);
+			if (!err)
+				err = cmd->err;
+			continue;
+		}
+
+		modify_virtqueue_end(ndev, mvq, state);
+	}
+
+done:
+	kvfree(cmd_mem);
+	kvfree(cmds);
+	return err;
+}
+
+static int suspend_vqs(struct mlx5_vdpa_net *ndev, int start_vq, int num_vqs)
+{
+	struct mlx5_vdpa_virtqueue *mvq;
+	struct mlx5_virtq_attr *attrs;
+	int vq_idx, i;
 	int err;
 
+	if (start_vq >= ndev->cur_num_vqs)
+		return -EINVAL;
+
+	mvq = &ndev->vqs[start_vq];
 	if (!mvq->initialized)
 		return 0;
 
 	if (mvq->fw_state != MLX5_VIRTIO_NET_Q_OBJECT_STATE_RDY)
 		return 0;
 
-	err = modify_virtqueue_state(ndev, mvq, MLX5_VIRTIO_NET_Q_OBJECT_STATE_SUSPEND);
+	err = modify_virtqueues(ndev, start_vq, num_vqs, MLX5_VIRTIO_NET_Q_OBJECT_STATE_SUSPEND);
-	if (err) {
+	if (err)
-		mlx5_vdpa_warn(&ndev->mvdev, "modify to suspend failed, err: %d\n", err);
 		return err;
+
+	attrs = kcalloc(num_vqs, sizeof(struct mlx5_virtq_attr), GFP_KERNEL);
+	if (!attrs)
+		return -ENOMEM;
+
+	err = query_virtqueues(ndev, start_vq, num_vqs, attrs);
+	if (err)
+		goto done;
+
+	for (i = 0, vq_idx = start_vq; i < num_vqs; i++, vq_idx++) {
+		mvq = &ndev->vqs[vq_idx];
+		mvq->avail_idx = attrs[i].available_index;
+		mvq->used_idx = attrs[i].used_index;
 	}
 
-	err = query_virtqueue(ndev, mvq, &attr);
+done:
-	if (err) {
+	kfree(attrs);
-		mlx5_vdpa_warn(&ndev->mvdev, "failed to query virtqueue, err: %d\n", err);
-		return err;
-	}
-
-	mvq->avail_idx = attr.available_index;
-	mvq->used_idx = attr.used_index;
-
-	return 0;
-}
-
-static int suspend_vqs(struct mlx5_vdpa_net *ndev)
-{
-	int err = 0;
-	int i;
-
-	for (i = 0; i < ndev->cur_num_vqs; i++) {
-		int local_err = suspend_vq(ndev, &ndev->vqs[i]);
-
-		err = local_err ? local_err : err;
-	}
-
 	return err;
 }
 
-static int resume_vq(struct mlx5_vdpa_net *ndev, struct mlx5_vdpa_virtqueue *mvq)
+static int suspend_vq(struct mlx5_vdpa_net *ndev, struct mlx5_vdpa_virtqueue *mvq)
 {
+	return suspend_vqs(ndev, mvq->index, 1);
+}
+
+static int resume_vqs(struct mlx5_vdpa_net *ndev, int start_vq, int num_vqs)
+{
+	struct mlx5_vdpa_virtqueue *mvq;
 	int err;
 
+	if (start_vq >= ndev->mvdev.max_vqs)
+		return -EINVAL;
+
+	mvq = &ndev->vqs[start_vq];
 	if (!mvq->initialized)
 		return 0;
 
@@ -1583,13 +1695,9 @@ static int resume_vq(struct mlx5_vdpa_net *ndev, struct mlx5_vdpa_virtqueue *mvq
 		/* Due to a FW quirk we need to modify the VQ fields first then change state.
 		 * This should be fixed soon. After that, a single command can be used.
 		 */
-		err = modify_virtqueue(ndev, mvq, 0);
+		err = modify_virtqueues(ndev, start_vq, num_vqs, mvq->fw_state);
-		if (err) {
+		if (err)
-			mlx5_vdpa_warn(&ndev->mvdev,
-				"modify vq properties failed for vq %u, err: %d\n",
-				mvq->index, err);
 			return err;
-		}
 		break;
 	case MLX5_VIRTIO_NET_Q_OBJECT_STATE_SUSPEND:
 		if (!is_resumable(ndev)) {
@@ -1600,30 +1708,17 @@ static int resume_vq(struct mlx5_vdpa_net *ndev, struct mlx5_vdpa_virtqueue *mvq
 	case MLX5_VIRTIO_NET_Q_OBJECT_STATE_RDY:
 		return 0;
 	default:
-		mlx5_vdpa_warn(&ndev->mvdev, "resume vq %u called from bad state %d\n",
+		mlx5_vdpa_err(&ndev->mvdev, "resume vq %u called from bad state %d\n",
 			       mvq->index, mvq->fw_state);
 		return -EINVAL;
 	}
 
-	err = modify_virtqueue_state(ndev, mvq, MLX5_VIRTIO_NET_Q_OBJECT_STATE_RDY);
+	return modify_virtqueues(ndev, start_vq, num_vqs, MLX5_VIRTIO_NET_Q_OBJECT_STATE_RDY);
-	if (err)
-		mlx5_vdpa_warn(&ndev->mvdev, "modify to resume failed for vq %u, err: %d\n",
-			       mvq->index, err);
-
-	return err;
 }
 
-static int resume_vqs(struct mlx5_vdpa_net *ndev)
+static int resume_vq(struct mlx5_vdpa_net *ndev, struct mlx5_vdpa_virtqueue *mvq)
 {
-	int err = 0;
+	return resume_vqs(ndev, mvq->index, 1);
-
-	for (int i = 0; i < ndev->cur_num_vqs; i++) {
-		int local_err = resume_vq(ndev, &ndev->vqs[i]);
-
-		err = local_err ? local_err : err;
-	}
-
-	return err;
 }
 
 static void teardown_vq(struct mlx5_vdpa_net *ndev, struct mlx5_vdpa_virtqueue *mvq)
@@ -2002,13 +2097,13 @@ static int setup_steering(struct mlx5_vdpa_net *ndev)
 
 	ns = mlx5_get_flow_namespace(ndev->mvdev.mdev, MLX5_FLOW_NAMESPACE_BYPASS);
 	if (!ns) {
-		mlx5_vdpa_warn(&ndev->mvdev, "failed to get flow namespace\n");
+		mlx5_vdpa_err(&ndev->mvdev, "failed to get flow namespace\n");
 		return -EOPNOTSUPP;
 	}
 
 	ndev->rxft = mlx5_create_auto_grouped_flow_table(ns, &ft_attr);
 	if (IS_ERR(ndev->rxft)) {
-		mlx5_vdpa_warn(&ndev->mvdev, "failed to create flow table\n");
+		mlx5_vdpa_err(&ndev->mvdev, "failed to create flow table\n");
 		return PTR_ERR(ndev->rxft);
 	}
 	mlx5_vdpa_add_rx_flow_table(ndev);
@@ -2124,45 +2219,48 @@ static virtio_net_ctrl_ack handle_ctrl_mac(struct mlx5_vdpa_dev *mvdev, u8 cmd)
 static int change_num_qps(struct mlx5_vdpa_dev *mvdev, int newqps)
 {
 	struct mlx5_vdpa_net *ndev = to_mlx5_vdpa_ndev(mvdev);
-	int cur_qps = ndev->cur_num_vqs / 2;
+	int cur_vqs = ndev->cur_num_vqs;
+	int new_vqs = newqps * 2;
 	int err;
 	int i;
 
-	if (cur_qps > newqps) {
+	if (cur_vqs > new_vqs) {
-		err = modify_rqt(ndev, 2 * newqps);
+		err = modify_rqt(ndev, new_vqs);
 		if (err)
 			return err;
 
-		for (i = ndev->cur_num_vqs - 1; i >= 2 * newqps; i--) {
+		if (is_resumable(ndev)) {
-			struct mlx5_vdpa_virtqueue *mvq = &ndev->vqs[i];
+			suspend_vqs(ndev, new_vqs, cur_vqs - new_vqs);
-
+		} else {
-			if (is_resumable(ndev))
+			for (i = new_vqs; i < cur_vqs; i++)
-				suspend_vq(ndev, mvq);
+				teardown_vq(ndev, &ndev->vqs[i]);
-			else
-				teardown_vq(ndev, mvq);
 		}
 
-		ndev->cur_num_vqs = 2 * newqps;
+		ndev->cur_num_vqs = new_vqs;
 	} else {
-		ndev->cur_num_vqs = 2 * newqps;
+		ndev->cur_num_vqs = new_vqs;
-		for (i = cur_qps * 2; i < 2 * newqps; i++) {
-			struct mlx5_vdpa_virtqueue *mvq = &ndev->vqs[i];
 
-			err = mvq->initialized ? resume_vq(ndev, mvq) : setup_vq(ndev, mvq, true);
+		for (i = cur_vqs; i < new_vqs; i++) {
+			err = setup_vq(ndev, &ndev->vqs[i], false);
 			if (err)
 				goto clean_added;
 		}
-		err = modify_rqt(ndev, 2 * newqps);
+
+		err = resume_vqs(ndev, cur_vqs, new_vqs - cur_vqs);
+		if (err)
+			goto clean_added;
+
+		err = modify_rqt(ndev, new_vqs);
 		if (err)
 			goto clean_added;
 	}
 	return 0;
 
 clean_added:
-	for (--i; i >= 2 * cur_qps; --i)
+	for (--i; i >= cur_vqs; --i)
 		teardown_vq(ndev, &ndev->vqs[i]);
 
-	ndev->cur_num_vqs = 2 * cur_qps;
+	ndev->cur_num_vqs = cur_vqs;
 
 	return err;
 }
@@ -2528,9 +2626,9 @@ static int mlx5_vdpa_get_vq_state(struct vdpa_device *vdev, u16 idx, struct vdpa
 		return 0;
 	}
 
-	err = query_virtqueue(ndev, mvq, &attr);
+	err = query_virtqueues(ndev, mvq->index, 1, &attr);
 	if (err) {
-		mlx5_vdpa_warn(mvdev, "failed to query virtqueue\n");
+		mlx5_vdpa_err(mvdev, "failed to query virtqueue\n");
 		return err;
 	}
 	state->split.avail_index = attr.used_index;
@@ -2755,6 +2853,9 @@ static int event_handler(struct notifier_block *nb, unsigned long event, void *p
 	struct mlx5_eqe *eqe = param;
 	int ret = NOTIFY_DONE;
 
+	if (ndev->mvdev.suspended)
+		return NOTIFY_DONE;
+
 	if (event == MLX5_EVENT_TYPE_PORT_CHANGE) {
 		switch (eqe->sub_type) {
 		case MLX5_PORT_CHANGE_SUBTYPE_DOWN:
@@ -2879,7 +2980,7 @@ static int save_channel_info(struct mlx5_vdpa_net *ndev, struct mlx5_vdpa_virtqu
 	int err;
 
 	if (mvq->initialized) {
-		err = query_virtqueue(ndev, mvq, &attr);
+		err = query_virtqueues(ndev, mvq->index, 1, &attr);
 		if (err)
 			return err;
 	}
@@ -2948,7 +3049,7 @@ static int mlx5_vdpa_change_map(struct mlx5_vdpa_dev *mvdev,
 	bool teardown = !is_resumable(ndev);
 	int err;
 
-	suspend_vqs(ndev);
+	suspend_vqs(ndev, 0, ndev->cur_num_vqs);
 	if (teardown) {
 		err = save_channels_info(ndev);
 		if (err)
@@ -2973,7 +3074,7 @@ static int mlx5_vdpa_change_map(struct mlx5_vdpa_dev *mvdev,
 			return err;
 	}
 
-	resume_vqs(ndev);
+	resume_vqs(ndev, 0, ndev->cur_num_vqs);
 
 	return 0;
 }
@@ -3097,7 +3198,7 @@ static void mlx5_vdpa_set_status(struct vdpa_device *vdev, u8 status)
 				teardown_vq_resources(ndev);
 
 			if (ndev->setup) {
-				err = resume_vqs(ndev);
+				err = resume_vqs(ndev, 0, ndev->cur_num_vqs);
 				if (err) {
 					mlx5_vdpa_warn(mvdev, "failed to resume VQs\n");
 					goto err_driver;
@@ -3122,7 +3223,7 @@ static void mlx5_vdpa_set_status(struct vdpa_device *vdev, u8 status)
 err_driver:
 	unregister_link_notifier(ndev);
 err_setup:
-	mlx5_vdpa_destroy_mr_resources(&ndev->mvdev);
+	mlx5_vdpa_clean_mrs(&ndev->mvdev);
 	ndev->mvdev.status |= VIRTIO_CONFIG_S_FAILED;
 err_clear:
 	up_write(&ndev->reslock);
@@ -3134,7 +3235,7 @@ static void init_group_to_asid_map(struct mlx5_vdpa_dev *mvdev)
 
 	/* default mapping all groups are mapped to asid 0 */
 	for (i = 0; i < MLX5_VDPA_NUMVQ_GROUPS; i++)
-		mvdev->group2asid[i] = 0;
+		mvdev->mres.group2asid[i] = 0;
 }
 
 static bool needs_vqs_reset(const struct mlx5_vdpa_dev *mvdev)
@@ -3174,7 +3275,7 @@ static int mlx5_vdpa_compat_reset(struct vdpa_device *vdev, u32 flags)
 	}
 
 	if (flags & VDPA_RESET_F_CLEAN_MAP)
-		mlx5_vdpa_destroy_mr_resources(&ndev->mvdev);
+		mlx5_vdpa_clean_mrs(&ndev->mvdev);
 	ndev->mvdev.status = 0;
 	ndev->mvdev.suspended = false;
 	ndev->cur_num_vqs = MLX5V_DEFAULT_VQ_COUNT;
@@ -3189,7 +3290,7 @@ static int mlx5_vdpa_compat_reset(struct vdpa_device *vdev, u32 flags)
 	if ((flags & VDPA_RESET_F_CLEAN_MAP) &&
 	    MLX5_CAP_GEN(mvdev->mdev, umem_uid_0)) {
 		if (mlx5_vdpa_create_dma_mr(mvdev))
-			mlx5_vdpa_warn(mvdev, "create MR failed\n");
+			mlx5_vdpa_err(mvdev, "create MR failed\n");
 	}
 	if (vq_reset)
 		setup_vq_resources(ndev, false);
@@ -3244,7 +3345,7 @@ static int set_map_data(struct mlx5_vdpa_dev *mvdev, struct vhost_iotlb *iotlb,
 		new_mr = mlx5_vdpa_create_mr(mvdev, iotlb);
 		if (IS_ERR(new_mr)) {
 			err = PTR_ERR(new_mr);
-			mlx5_vdpa_warn(mvdev, "create map failed(%d)\n", err);
+			mlx5_vdpa_err(mvdev, "create map failed(%d)\n", err);
 			return err;
 		}
 	} else {
@@ -3252,12 +3353,12 @@ static int set_map_data(struct mlx5_vdpa_dev *mvdev, struct vhost_iotlb *iotlb,
 		new_mr = NULL;
 	}
 
-	if (!mvdev->mr[asid]) {
+	if (!mvdev->mres.mr[asid]) {
 		mlx5_vdpa_update_mr(mvdev, new_mr, asid);
 	} else {
 		err = mlx5_vdpa_change_map(mvdev, new_mr, asid);
 		if (err) {
-			mlx5_vdpa_warn(mvdev, "change map failed(%d)\n", err);
+			mlx5_vdpa_err(mvdev, "change map failed(%d)\n", err);
 			goto out_err;
 		}
 	}
@@ -3332,7 +3433,10 @@ static void mlx5_vdpa_free(struct vdpa_device *vdev)
 	ndev = to_mlx5_vdpa_ndev(mvdev);
 
 	free_fixed_resources(ndev);
-	mlx5_vdpa_destroy_mr_resources(mvdev);
+	mlx5_vdpa_clean_mrs(mvdev);
+	mlx5_vdpa_destroy_mr_resources(&ndev->mvdev);
+	mlx5_cmd_cleanup_async_ctx(&mvdev->async_ctx);
+
 	if (!is_zero_ether_addr(ndev->config.mac)) {
 		pfmdev = pci_get_drvdata(pci_physfn(mvdev->mdev->pdev));
 		mlx5_mpfs_del_mac(pfmdev, ndev->config.mac);
@@ -3500,8 +3604,7 @@ static int mlx5_vdpa_suspend(struct vdpa_device *vdev)
 	mlx5_vdpa_info(mvdev, "suspending device\n");
 
 	down_write(&ndev->reslock);
-	unregister_link_notifier(ndev);
+	err = suspend_vqs(ndev, 0, ndev->cur_num_vqs);
-	err = suspend_vqs(ndev);
 	mlx5_vdpa_cvq_suspend(mvdev);
 	mvdev->suspended = true;
 	up_write(&ndev->reslock);
@@ -3521,8 +3624,8 @@ static int mlx5_vdpa_resume(struct vdpa_device *vdev)
 
 	down_write(&ndev->reslock);
 	mvdev->suspended = false;
-	err = resume_vqs(ndev);
+	err = resume_vqs(ndev, 0, ndev->cur_num_vqs);
-	register_link_notifier(ndev);
+	queue_link_work(ndev);
 	up_write(&ndev->reslock);
 
 	return err;
@@ -3537,12 +3640,12 @@ static int mlx5_set_group_asid(struct vdpa_device *vdev, u32 group,
 	if (group >= MLX5_VDPA_NUMVQ_GROUPS)
 		return -EINVAL;
 
-	mvdev->group2asid[group] = asid;
+	mvdev->mres.group2asid[group] = asid;
 
-	mutex_lock(&mvdev->mr_mtx);
+	mutex_lock(&mvdev->mres.lock);
-	if (group == MLX5_VDPA_CVQ_GROUP && mvdev->mr[asid])
+	if (group == MLX5_VDPA_CVQ_GROUP && mvdev->mres.mr[asid])
-		err = mlx5_vdpa_update_cvq_iotlb(mvdev, mvdev->mr[asid]->iotlb, asid);
+		err = mlx5_vdpa_update_cvq_iotlb(mvdev, mvdev->mres.mr[asid]->iotlb, asid);
-	mutex_unlock(&mvdev->mr_mtx);
+	mutex_unlock(&mvdev->mres.lock);
 
 	return err;
 }
@@ -3854,18 +3957,22 @@ static int mlx5_vdpa_dev_add(struct vdpa_mgmt_dev *v_mdev, const char *name,
 		ndev->rqt_size = 1;
 	}
 
+	mlx5_cmd_init_async_ctx(mdev, &mvdev->async_ctx);
+
 	ndev->mvdev.mlx_features = device_features;
 	mvdev->vdev.dma_dev = &mdev->pdev->dev;
 	err = mlx5_vdpa_alloc_resources(&ndev->mvdev);
 	if (err)
 		goto err_mpfs;
 
-	INIT_LIST_HEAD(&mvdev->mr_list_head);
+	err = mlx5_vdpa_init_mr_resources(mvdev);
+	if (err)
+		goto err_res;
 
 	if (MLX5_CAP_GEN(mvdev->mdev, umem_uid_0)) {
 		err = mlx5_vdpa_create_dma_mr(mvdev);
 		if (err)
-			goto err_res;
+			goto err_mr_res;
 	}
 
 	err = alloc_fixed_resources(ndev);
@@ -3906,6 +4013,8 @@ err_reg:
 err_res2:
 	free_fixed_resources(ndev);
 err_mr:
+	mlx5_vdpa_clean_mrs(mvdev);
+err_mr_res:
 	mlx5_vdpa_destroy_mr_resources(mvdev);
 err_res:
 	mlx5_vdpa_free_resources(&ndev->mvdev);
@@ -3937,9 +4046,37 @@ static void mlx5_vdpa_dev_del(struct vdpa_mgmt_dev *v_mdev, struct vdpa_device *
 	mgtdev->ndev = NULL;
 }
 
+static int mlx5_vdpa_set_attr(struct vdpa_mgmt_dev *v_mdev, struct vdpa_device *dev,
+			      const struct vdpa_dev_set_config *add_config)
+{
+	struct virtio_net_config *config;
+	struct mlx5_core_dev *pfmdev;
+	struct mlx5_vdpa_dev *mvdev;
+	struct mlx5_vdpa_net *ndev;
+	struct mlx5_core_dev *mdev;
+	int err = -EOPNOTSUPP;
+
+	mvdev = to_mvdev(dev);
+	ndev = to_mlx5_vdpa_ndev(mvdev);
+	mdev = mvdev->mdev;
+	config = &ndev->config;
+
+	down_write(&ndev->reslock);
+	if (add_config->mask & (1 << VDPA_ATTR_DEV_NET_CFG_MACADDR)) {
+		pfmdev = pci_get_drvdata(pci_physfn(mdev->pdev));
+		err = mlx5_mpfs_add_mac(pfmdev, config->mac);
+		if (!err)
+			ether_addr_copy(config->mac, add_config->net.mac);
+	}
+
+	up_write(&ndev->reslock);
+	return err;
+}
+
 static const struct vdpa_mgmtdev_ops mdev_ops = {
 	.dev_add = mlx5_vdpa_dev_add,
 	.dev_del = mlx5_vdpa_dev_del,
+	.dev_set_attr = mlx5_vdpa_set_attr,
 };
 
 static struct virtio_device_id id_table[] = {

drivers/vdpa/pds/cmds.h

@@ -14,5 +14,4 @@ int pds_vdpa_cmd_init_vq(struct pds_vdpa_device *pdsv, u16 qid, u16 invert_idx,
 			 struct pds_vdpa_vq_info *vq_info);
 int pds_vdpa_cmd_reset_vq(struct pds_vdpa_device *pdsv, u16 qid, u16 invert_idx,
 			  struct pds_vdpa_vq_info *vq_info);
-int pds_vdpa_cmd_set_features(struct pds_vdpa_device *pdsv, u64 features);
 #endif /* _VDPA_CMDS_H_ */

drivers/vdpa/vdpa.c

@@ -1361,6 +1361,80 @@ dev_err:
 	return err;
 }
 
+static int vdpa_dev_net_device_attr_set(struct vdpa_device *vdev,
+					struct genl_info *info)
+{
+	struct vdpa_dev_set_config set_config = {};
+	struct vdpa_mgmt_dev *mdev = vdev->mdev;
+	struct nlattr **nl_attrs = info->attrs;
+	const u8 *macaddr;
+	int err = -EOPNOTSUPP;
+
+	down_write(&vdev->cf_lock);
+	if (nl_attrs[VDPA_ATTR_DEV_NET_CFG_MACADDR]) {
+		set_config.mask |= BIT_ULL(VDPA_ATTR_DEV_NET_CFG_MACADDR);
+		macaddr = nla_data(nl_attrs[VDPA_ATTR_DEV_NET_CFG_MACADDR]);
+
+		if (is_valid_ether_addr(macaddr)) {
+			ether_addr_copy(set_config.net.mac, macaddr);
+			if (mdev->ops->dev_set_attr) {
+				err = mdev->ops->dev_set_attr(mdev, vdev,
+							      &set_config);
+			} else {
+				NL_SET_ERR_MSG_FMT_MOD(info->extack,
+						       "Operation not supported by the device.");
+			}
+		} else {
+			NL_SET_ERR_MSG_FMT_MOD(info->extack,
+					       "Invalid MAC address");
+		}
+	}
+	up_write(&vdev->cf_lock);
+	return err;
+}
+
+static int vdpa_nl_cmd_dev_attr_set_doit(struct sk_buff *skb,
+					 struct genl_info *info)
+{
+	struct vdpa_device *vdev;
+	struct device *dev;
+	const char *name;
+	u64 classes;
+	int err = 0;
+
+	if (!info->attrs[VDPA_ATTR_DEV_NAME])
+		return -EINVAL;
+
+	name = nla_data(info->attrs[VDPA_ATTR_DEV_NAME]);
+
+	down_write(&vdpa_dev_lock);
+	dev = bus_find_device(&vdpa_bus, NULL, name, vdpa_name_match);
+	if (!dev) {
+		NL_SET_ERR_MSG_MOD(info->extack, "device not found");
+		err = -ENODEV;
+		goto dev_err;
+	}
+	vdev = container_of(dev, struct vdpa_device, dev);
+	if (!vdev->mdev) {
+		NL_SET_ERR_MSG_MOD(info->extack, "unmanaged vdpa device");
+		err = -EINVAL;
+		goto mdev_err;
+	}
+	classes = vdpa_mgmtdev_get_classes(vdev->mdev, NULL);
+	if (classes & BIT_ULL(VIRTIO_ID_NET)) {
+		err = vdpa_dev_net_device_attr_set(vdev, info);
+	} else {
+		NL_SET_ERR_MSG_FMT_MOD(info->extack, "%s device not supported",
+				       name);
+	}
+
+mdev_err:
+	put_device(dev);
+dev_err:
+	up_write(&vdpa_dev_lock);
+	return err;
+}
+
 static int vdpa_dev_config_dump(struct device *dev, void *data)
 {
 	struct vdpa_device *vdev = container_of(dev, struct vdpa_device, dev);
@@ -1497,6 +1571,11 @@ static const struct genl_ops vdpa_nl_ops[] = {
 		.doit = vdpa_nl_cmd_dev_stats_get_doit,
 		.flags = GENL_ADMIN_PERM,
 	},
+	{
+		.cmd = VDPA_CMD_DEV_ATTR_SET,
+		.doit = vdpa_nl_cmd_dev_attr_set_doit,
+		.flags = GENL_ADMIN_PERM,
+	},
 };
 
 static struct genl_family vdpa_nl_family __ro_after_init = {

drivers/vdpa/vdpa_sim/vdpa_sim_net.c

@@ -414,6 +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)
 {
@@ -510,7 +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[] = {

drivers/vhost/vdpa.c

@@ -209,11 +209,9 @@ static void vhost_vdpa_setup_vq_irq(struct vhost_vdpa *v, u16 qid)
 	if (irq < 0)
 		return;
 
-	irq_bypass_unregister_producer(&vq->call_ctx.producer);
 	if (!vq->call_ctx.ctx)
 		return;
 
-	vq->call_ctx.producer.token = vq->call_ctx.ctx;
 	vq->call_ctx.producer.irq = irq;
 	ret = irq_bypass_register_producer(&vq->call_ctx.producer);
 	if (unlikely(ret))
@@ -709,6 +707,14 @@ static long vhost_vdpa_vring_ioctl(struct vhost_vdpa *v, unsigned int cmd,
 			vq->last_avail_idx = vq_state.split.avail_index;
 		}
 		break;
+	case VHOST_SET_VRING_CALL:
+		if (vq->call_ctx.ctx) {
+			if (ops->get_status(vdpa) &
+			    VIRTIO_CONFIG_S_DRIVER_OK)
+				vhost_vdpa_unsetup_vq_irq(v, idx);
+			vq->call_ctx.producer.token = NULL;
+		}
+		break;
 	}
 
 	r = vhost_vring_ioctl(&v->vdev, cmd, argp);
@@ -747,13 +753,16 @@ static long vhost_vdpa_vring_ioctl(struct vhost_vdpa *v, unsigned int cmd,
 			cb.callback = vhost_vdpa_virtqueue_cb;
 			cb.private = vq;
 			cb.trigger = vq->call_ctx.ctx;
+			vq->call_ctx.producer.token = vq->call_ctx.ctx;
+			if (ops->get_status(vdpa) &
+			    VIRTIO_CONFIG_S_DRIVER_OK)
+				vhost_vdpa_setup_vq_irq(v, idx);
 		} else {
 			cb.callback = NULL;
 			cb.private = NULL;
 			cb.trigger = NULL;
 		}
 		ops->set_vq_cb(vdpa, idx, &cb);
-		vhost_vdpa_setup_vq_irq(v, idx);
 		break;
 
 	case VHOST_SET_VRING_NUM:
@@ -1419,6 +1428,7 @@ static int vhost_vdpa_open(struct inode *inode, struct file *filep)
 	for (i = 0; i < nvqs; i++) {
 		vqs[i] = &v->vqs[i];
 		vqs[i]->handle_kick = handle_vq_kick;
+		vqs[i]->call_ctx.ctx = NULL;
 	}
 	vhost_dev_init(dev, vqs, nvqs, 0, 0, 0, false,
 		       vhost_vdpa_process_iotlb_msg);

drivers/virtio/virtio_balloon.c

@@ -355,6 +355,8 @@ static inline unsigned int update_balloon_vm_stats(struct virtio_balloon *vb)
 {
 	unsigned long events[NR_VM_EVENT_ITEMS];
 	unsigned int idx = 0;
+	unsigned int zid;
+	unsigned long stall = 0;
 
 	all_vm_events(events);
 	update_stat(vb, idx++, VIRTIO_BALLOON_S_SWAP_IN,
@@ -363,6 +365,22 @@ static inline unsigned int update_balloon_vm_stats(struct virtio_balloon *vb)
 		    pages_to_bytes(events[PSWPOUT]));
 	update_stat(vb, idx++, VIRTIO_BALLOON_S_MAJFLT, events[PGMAJFAULT]);
 	update_stat(vb, idx++, VIRTIO_BALLOON_S_MINFLT, events[PGFAULT]);
+	update_stat(vb, idx++, VIRTIO_BALLOON_S_OOM_KILL, events[OOM_KILL]);
+
+	/* sum all the stall events */
+	for (zid = 0; zid < MAX_NR_ZONES; zid++)
+		stall += events[ALLOCSTALL_NORMAL - ZONE_NORMAL + zid];
+
+	update_stat(vb, idx++, VIRTIO_BALLOON_S_ALLOC_STALL, stall);
+
+	update_stat(vb, idx++, VIRTIO_BALLOON_S_ASYNC_SCAN,
+		    pages_to_bytes(events[PGSCAN_KSWAPD]));
+	update_stat(vb, idx++, VIRTIO_BALLOON_S_DIRECT_SCAN,
+		    pages_to_bytes(events[PGSCAN_DIRECT]));
+	update_stat(vb, idx++, VIRTIO_BALLOON_S_ASYNC_RECLAIM,
+		    pages_to_bytes(events[PGSTEAL_KSWAPD]));
+	update_stat(vb, idx++, VIRTIO_BALLOON_S_DIRECT_RECLAIM,
+		    pages_to_bytes(events[PGSTEAL_DIRECT]));
 
 #ifdef CONFIG_HUGETLB_PAGE
 	update_stat(vb, idx++, VIRTIO_BALLOON_S_HTLB_PGALLOC,

fs/fuse/virtio_fs.c

@@ -56,12 +56,14 @@ struct virtio_fs_vq {
 	bool connected;
 	long in_flight;
 	struct completion in_flight_zero; /* No inflight requests */
+	struct kobject *kobj;
 	char name[VQ_NAME_LEN];
 } ____cacheline_aligned_in_smp;
 
 /* A virtio-fs device instance */
 struct virtio_fs {
 	struct kobject kobj;
+	struct kobject *mqs_kobj;
 	struct list_head list;    /* on virtio_fs_instances */
 	char *tag;
 	struct virtio_fs_vq *vqs;
@@ -200,19 +202,94 @@ static const struct kobj_type virtio_fs_ktype = {
 	.default_groups = virtio_fs_groups,
 };
 
+static struct virtio_fs_vq *virtio_fs_kobj_to_vq(struct virtio_fs *fs,
+		struct kobject *kobj)
+{
+	int i;
+
+	for (i = 0; i < fs->nvqs; i++) {
+		if (kobj == fs->vqs[i].kobj)
+			return &fs->vqs[i];
+	}
+	return NULL;
+}
+
+static ssize_t name_show(struct kobject *kobj,
+		struct kobj_attribute *attr, char *buf)
+{
+	struct virtio_fs *fs = container_of(kobj->parent->parent, struct virtio_fs, kobj);
+	struct virtio_fs_vq *fsvq = virtio_fs_kobj_to_vq(fs, kobj);
+
+	if (!fsvq)
+		return -EINVAL;
+	return sysfs_emit(buf, "%s\n", fsvq->name);
+}
+
+static struct kobj_attribute virtio_fs_vq_name_attr = __ATTR_RO(name);
+
+static ssize_t cpu_list_show(struct kobject *kobj,
+		struct kobj_attribute *attr, char *buf)
+{
+	struct virtio_fs *fs = container_of(kobj->parent->parent, struct virtio_fs, kobj);
+	struct virtio_fs_vq *fsvq = virtio_fs_kobj_to_vq(fs, kobj);
+	unsigned int cpu, qid;
+	const size_t size = PAGE_SIZE - 1;
+	bool first = true;
+	int ret = 0, pos = 0;
+
+	if (!fsvq)
+		return -EINVAL;
+
+	qid = fsvq->vq->index;
+	for (cpu = 0; cpu < nr_cpu_ids; cpu++) {
+		if (qid < VQ_REQUEST || (fs->mq_map[cpu] == qid - VQ_REQUEST)) {
+			if (first)
+				ret = snprintf(buf + pos, size - pos, "%u", cpu);
+			else
+				ret = snprintf(buf + pos, size - pos, ", %u", cpu);
+
+			if (ret >= size - pos)
+				break;
+			first = false;
+			pos += ret;
+		}
+	}
+	ret = snprintf(buf + pos, size + 1 - pos, "\n");
+	return pos + ret;
+}
+
+static struct kobj_attribute virtio_fs_vq_cpu_list_attr = __ATTR_RO(cpu_list);
+
+static struct attribute *virtio_fs_vq_attrs[] = {
+	&virtio_fs_vq_name_attr.attr,
+	&virtio_fs_vq_cpu_list_attr.attr,
+	NULL
+};
+
+static struct attribute_group virtio_fs_vq_attr_group = {
+	.attrs = virtio_fs_vq_attrs,
+};
+
 /* Make sure virtiofs_mutex is held */
+static void virtio_fs_put_locked(struct virtio_fs *fs)
+{
+	lockdep_assert_held(&virtio_fs_mutex);
+
+	kobject_put(&fs->kobj);
+}
+
 static void virtio_fs_put(struct virtio_fs *fs)
 {
-	kobject_put(&fs->kobj);
+	mutex_lock(&virtio_fs_mutex);
+	virtio_fs_put_locked(fs);
+	mutex_unlock(&virtio_fs_mutex);
 }
 
 static void virtio_fs_fiq_release(struct fuse_iqueue *fiq)
 {
 	struct virtio_fs *vfs = fiq->priv;
 
-	mutex_lock(&virtio_fs_mutex);
 	virtio_fs_put(vfs);
-	mutex_unlock(&virtio_fs_mutex);
 }
 
 static void virtio_fs_drain_queue(struct virtio_fs_vq *fsvq)
@@ -273,6 +350,50 @@ static void virtio_fs_start_all_queues(struct virtio_fs *fs)
 	}
 }
 
+static void virtio_fs_delete_queues_sysfs(struct virtio_fs *fs)
+{
+	struct virtio_fs_vq *fsvq;
+	int i;
+
+	for (i = 0; i < fs->nvqs; i++) {
+		fsvq = &fs->vqs[i];
+		kobject_put(fsvq->kobj);
+	}
+}
+
+static int virtio_fs_add_queues_sysfs(struct virtio_fs *fs)
+{
+	struct virtio_fs_vq *fsvq;
+	char buff[12];
+	int i, j, ret;
+
+	for (i = 0; i < fs->nvqs; i++) {
+		fsvq = &fs->vqs[i];
+
+		sprintf(buff, "%d", i);
+		fsvq->kobj = kobject_create_and_add(buff, fs->mqs_kobj);
+		if (!fs->mqs_kobj) {
+			ret = -ENOMEM;
+			goto out_del;
+		}
+
+		ret = sysfs_create_group(fsvq->kobj, &virtio_fs_vq_attr_group);
+		if (ret) {
+			kobject_put(fsvq->kobj);
+			goto out_del;
+		}
+	}
+
+	return 0;
+
+out_del:
+	for (j = 0; j < i; j++) {
+		fsvq = &fs->vqs[j];
+		kobject_put(fsvq->kobj);
+	}
+	return ret;
+}
+
 /* Add a new instance to the list or return -EEXIST if tag name exists*/
 static int virtio_fs_add_instance(struct virtio_device *vdev,
 				  struct virtio_fs *fs)
@@ -296,17 +417,22 @@ static int virtio_fs_add_instance(struct virtio_device *vdev,
 	 */
 	fs->kobj.kset = virtio_fs_kset;
 	ret = kobject_add(&fs->kobj, NULL, "%d", vdev->index);
-	if (ret < 0) {
+	if (ret < 0)
-		mutex_unlock(&virtio_fs_mutex);
+		goto out_unlock;
-		return ret;
+
+	fs->mqs_kobj = kobject_create_and_add("mqs", &fs->kobj);
+	if (!fs->mqs_kobj) {
+		ret = -ENOMEM;
+		goto out_del;
 	}
 
 	ret = sysfs_create_link(&fs->kobj, &vdev->dev.kobj, "device");
-	if (ret < 0) {
+	if (ret < 0)
-		kobject_del(&fs->kobj);
+		goto out_put;
-		mutex_unlock(&virtio_fs_mutex);
+
-		return ret;
+	ret = virtio_fs_add_queues_sysfs(fs);
-	}
+	if (ret)
+		goto out_remove;
 
 	list_add_tail(&fs->list, &virtio_fs_instances);
 
@@ -315,6 +441,16 @@ static int virtio_fs_add_instance(struct virtio_device *vdev,
 	kobject_uevent(&fs->kobj, KOBJ_ADD);
 
 	return 0;
+
+out_remove:
+	sysfs_remove_link(&fs->kobj, "device");
+out_put:
+	kobject_put(fs->mqs_kobj);
+out_del:
+	kobject_del(&fs->kobj);
+out_unlock:
+	mutex_unlock(&virtio_fs_mutex);
+	return ret;
 }
 
 /* Return the virtio_fs with a given tag, or NULL */
@@ -1043,7 +1179,9 @@ static void virtio_fs_remove(struct virtio_device *vdev)
 	mutex_lock(&virtio_fs_mutex);
 	/* This device is going away. No one should get new reference */
 	list_del_init(&fs->list);
+	virtio_fs_delete_queues_sysfs(fs);
 	sysfs_remove_link(&fs->kobj, "device");
+	kobject_put(fs->mqs_kobj);
 	kobject_del(&fs->kobj);
 	virtio_fs_stop_all_queues(fs);
 	virtio_fs_drain_all_queues_locked(fs);
@@ -1052,7 +1190,7 @@ static void virtio_fs_remove(struct virtio_device *vdev)
 
 	vdev->priv = NULL;
 	/* Put device reference on virtio_fs object */
-	virtio_fs_put(fs);
+	virtio_fs_put_locked(fs);
 	mutex_unlock(&virtio_fs_mutex);
 }
 
@@ -1581,9 +1719,7 @@ static int virtio_fs_get_tree(struct fs_context *fsc)
 
 out_err:
 	kfree(fc);
-	mutex_lock(&virtio_fs_mutex);
 	virtio_fs_put(fs);
-	mutex_unlock(&virtio_fs_mutex);
 	return err;
 }
 

include/linux/vdpa.h

@@ -582,11 +582,20 @@ void vdpa_set_status(struct vdpa_device *vdev, u8 status);
  *	     @dev: vdpa device to remove
  *	     Driver need to remove the specified device by calling
  *	     _vdpa_unregister_device().
+ * @dev_set_attr: change a vdpa device's attr after it was create
+ *	     @mdev: parent device to use for device
+ *	     @dev: vdpa device structure
+ *	     @config:Attributes to be set for the device.
+ *	     The driver needs to check the mask of the structure and then set
+ *	     the related information to the vdpa device. The driver must return 0
+ *	     if set successfully.
  */
 struct vdpa_mgmtdev_ops {
 	int (*dev_add)(struct vdpa_mgmt_dev *mdev, const char *name,
 		       const struct vdpa_dev_set_config *config);
 	void (*dev_del)(struct vdpa_mgmt_dev *mdev, struct vdpa_device *dev);
+	int (*dev_set_attr)(struct vdpa_mgmt_dev *mdev, struct vdpa_device *dev,
+			    const struct vdpa_dev_set_config *config);
 };
 
 /**

include/uapi/linux/vdpa.h

@@ -19,6 +19,7 @@ enum vdpa_command {
 	VDPA_CMD_DEV_GET,		/* can dump */
 	VDPA_CMD_DEV_CONFIG_GET,	/* can dump */
 	VDPA_CMD_DEV_VSTATS_GET,
+	VDPA_CMD_DEV_ATTR_SET,
 };
 
 enum vdpa_attr {

include/uapi/linux/virtio_balloon.h

@@ -71,7 +71,13 @@ struct virtio_balloon_config {
 #define VIRTIO_BALLOON_S_CACHES   7   /* Disk caches */
 #define VIRTIO_BALLOON_S_HTLB_PGALLOC  8  /* Hugetlb page allocations */
 #define VIRTIO_BALLOON_S_HTLB_PGFAIL   9  /* Hugetlb page allocation failures */
-#define VIRTIO_BALLOON_S_NR       10
+#define VIRTIO_BALLOON_S_OOM_KILL      10 /* OOM killer invocations */
+#define VIRTIO_BALLOON_S_ALLOC_STALL   11 /* Stall count of memory allocatoin */
+#define VIRTIO_BALLOON_S_ASYNC_SCAN    12 /* Amount of memory scanned asynchronously */
+#define VIRTIO_BALLOON_S_DIRECT_SCAN   13 /* Amount of memory scanned directly */
+#define VIRTIO_BALLOON_S_ASYNC_RECLAIM 14 /* Amount of memory reclaimed asynchronously */
+#define VIRTIO_BALLOON_S_DIRECT_RECLAIM 15 /* Amount of memory reclaimed directly */
+#define VIRTIO_BALLOON_S_NR       16
 
 #define VIRTIO_BALLOON_S_NAMES_WITH_PREFIX(VIRTIO_BALLOON_S_NAMES_prefix) { \
 	VIRTIO_BALLOON_S_NAMES_prefix "swap-in", \
@@ -83,7 +89,13 @@ struct virtio_balloon_config {
 	VIRTIO_BALLOON_S_NAMES_prefix "available-memory", \
 	VIRTIO_BALLOON_S_NAMES_prefix "disk-caches", \
 	VIRTIO_BALLOON_S_NAMES_prefix "hugetlb-allocations", \
-	VIRTIO_BALLOON_S_NAMES_prefix "hugetlb-failures" \
+	VIRTIO_BALLOON_S_NAMES_prefix "hugetlb-failures", \
+	VIRTIO_BALLOON_S_NAMES_prefix "oom-kills", \
+	VIRTIO_BALLOON_S_NAMES_prefix "alloc-stalls", \
+	VIRTIO_BALLOON_S_NAMES_prefix "async-scans", \
+	VIRTIO_BALLOON_S_NAMES_prefix "direct-scans", \
+	VIRTIO_BALLOON_S_NAMES_prefix "async-reclaims", \
+	VIRTIO_BALLOON_S_NAMES_prefix "direct-reclaims" \
 }
 
 #define VIRTIO_BALLOON_S_NAMES VIRTIO_BALLOON_S_NAMES_WITH_PREFIX("")

net/vmw_vsock/virtio_transport.c

@@ -94,6 +94,63 @@ out_rcu:
 	return ret;
 }
 
+/* Caller need to hold vsock->tx_lock on vq */
+static int virtio_transport_send_skb(struct sk_buff *skb, struct virtqueue *vq,
+				     struct virtio_vsock *vsock)
+{
+	int ret, in_sg = 0, out_sg = 0;
+	struct scatterlist **sgs;
+
+	sgs = vsock->out_sgs;
+	sg_init_one(sgs[out_sg], virtio_vsock_hdr(skb),
+		    sizeof(*virtio_vsock_hdr(skb)));
+	out_sg++;
+
+	if (!skb_is_nonlinear(skb)) {
+		if (skb->len > 0) {
+			sg_init_one(sgs[out_sg], skb->data, skb->len);
+			out_sg++;
+		}
+	} else {
+		struct skb_shared_info *si;
+		int i;
+
+		/* If skb is nonlinear, then its buffer must contain
+		 * only header and nothing more. Data is stored in
+		 * the fragged part.
+		 */
+		WARN_ON_ONCE(skb_headroom(skb) != sizeof(*virtio_vsock_hdr(skb)));
+
+		si = skb_shinfo(skb);
+
+		for (i = 0; i < si->nr_frags; i++) {
+			skb_frag_t *skb_frag = &si->frags[i];
+			void *va;
+
+			/* We will use 'page_to_virt()' for the userspace page
+			 * here, because virtio or dma-mapping layers will call
+			 * 'virt_to_phys()' later to fill the buffer descriptor.
+			 * We don't touch memory at "virtual" address of this page.
+			 */
+			va = page_to_virt(skb_frag_page(skb_frag));
+			sg_init_one(sgs[out_sg],
+				    va + skb_frag_off(skb_frag),
+				    skb_frag_size(skb_frag));
+			out_sg++;
+		}
+	}
+
+	ret = virtqueue_add_sgs(vq, sgs, out_sg, in_sg, skb, GFP_KERNEL);
+	/* Usually this means that there is no more space available in
+	 * the vq
+	 */
+	if (ret < 0)
+		return ret;
+
+	virtio_transport_deliver_tap_pkt(skb);
+	return 0;
+}
+
 static void
 virtio_transport_send_pkt_work(struct work_struct *work)
 {
@@ -111,66 +168,22 @@ virtio_transport_send_pkt_work(struct work_struct *work)
 	vq = vsock->vqs[VSOCK_VQ_TX];
 
 	for (;;) {
-		int ret, in_sg = 0, out_sg = 0;
-		struct scatterlist **sgs;
 		struct sk_buff *skb;
 		bool reply;
+		int ret;
 
 		skb = virtio_vsock_skb_dequeue(&vsock->send_pkt_queue);
 		if (!skb)
 			break;
 
 		reply = virtio_vsock_skb_reply(skb);
-		sgs = vsock->out_sgs;
-		sg_init_one(sgs[out_sg], virtio_vsock_hdr(skb),
-			    sizeof(*virtio_vsock_hdr(skb)));
-		out_sg++;
 
-		if (!skb_is_nonlinear(skb)) {
+		ret = virtio_transport_send_skb(skb, vq, vsock);
-			if (skb->len > 0) {
-				sg_init_one(sgs[out_sg], skb->data, skb->len);
-				out_sg++;
-			}
-		} else {
-			struct skb_shared_info *si;
-			int i;
-
-			/* If skb is nonlinear, then its buffer must contain
-			 * only header and nothing more. Data is stored in
-			 * the fragged part.
-			 */
-			WARN_ON_ONCE(skb_headroom(skb) != sizeof(*virtio_vsock_hdr(skb)));
-
-			si = skb_shinfo(skb);
-
-			for (i = 0; i < si->nr_frags; i++) {
-				skb_frag_t *skb_frag = &si->frags[i];
-				void *va;
-
-				/* We will use 'page_to_virt()' for the userspace page
-				 * here, because virtio or dma-mapping layers will call
-				 * 'virt_to_phys()' later to fill the buffer descriptor.
-				 * We don't touch memory at "virtual" address of this page.
-				 */
-				va = page_to_virt(skb_frag_page(skb_frag));
-				sg_init_one(sgs[out_sg],
-					    va + skb_frag_off(skb_frag),
-					    skb_frag_size(skb_frag));
-				out_sg++;
-			}
-		}
-
-		ret = virtqueue_add_sgs(vq, sgs, out_sg, in_sg, skb, GFP_KERNEL);
-		/* Usually this means that there is no more space available in
-		 * the vq
-		 */
 		if (ret < 0) {
 			virtio_vsock_skb_queue_head(&vsock->send_pkt_queue, skb);
 			break;
 		}
 
-		virtio_transport_deliver_tap_pkt(skb);
-
 		if (reply) {
 			struct virtqueue *rx_vq = vsock->vqs[VSOCK_VQ_RX];
 			int val;
@@ -195,6 +208,28 @@ out:
 		queue_work(virtio_vsock_workqueue, &vsock->rx_work);
 }
 
+/* Caller need to hold RCU for vsock.
+ * Returns 0 if the packet is successfully put on the vq.
+ */
+static int virtio_transport_send_skb_fast_path(struct virtio_vsock *vsock, struct sk_buff *skb)
+{
+	struct virtqueue *vq = vsock->vqs[VSOCK_VQ_TX];
+	int ret;
+
+	/* Inside RCU, can't sleep! */
+	ret = mutex_trylock(&vsock->tx_lock);
+	if (unlikely(ret == 0))
+		return -EBUSY;
+
+	ret = virtio_transport_send_skb(skb, vq, vsock);
+	if (ret == 0)
+		virtqueue_kick(vq);
+
+	mutex_unlock(&vsock->tx_lock);
+
+	return ret;
+}
+
 static int
 virtio_transport_send_pkt(struct sk_buff *skb)
 {
@@ -218,11 +253,20 @@ virtio_transport_send_pkt(struct sk_buff *skb)
 		goto out_rcu;
 	}
 
-	if (virtio_vsock_skb_reply(skb))
+	/* If send_pkt_queue is empty, we can safely bypass this queue
-		atomic_inc(&vsock->queued_replies);
+	 * because packet order is maintained and (try) to put the packet
+	 * on the virtqueue using virtio_transport_send_skb_fast_path.
+	 * If this fails we simply put the packet on the intermediate
+	 * queue and schedule the worker.
+	 */
+	if (!skb_queue_empty_lockless(&vsock->send_pkt_queue) ||
+	    virtio_transport_send_skb_fast_path(vsock, skb)) {
+		if (virtio_vsock_skb_reply(skb))
+			atomic_inc(&vsock->queued_replies);
 
-	virtio_vsock_skb_queue_tail(&vsock->send_pkt_queue, skb);
+		virtio_vsock_skb_queue_tail(&vsock->send_pkt_queue, skb);
-	queue_work(virtio_vsock_workqueue, &vsock->send_pkt_work);
+		queue_work(virtio_vsock_workqueue, &vsock->send_pkt_work);
+	}
 
 out_rcu:
 	rcu_read_unlock();

tools/virtio/ringtest/main.c

@@ -276,7 +276,7 @@ static void help(void)
 	fprintf(stderr, "Usage: <test> [--help]"
 		" [--host-affinity H]"
 		" [--guest-affinity G]"
-		" [--ring-size R (default: %d)]"
+		" [--ring-size R (default: %u)]"
 		" [--run-cycles C (default: %d)]"
 		" [--batch b]"
 		" [--outstanding o]"