Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::3:3 as permitted sender) client-ip=2620:137:e000::3:3;
Received-SPF: Pass (protection.outlook.com: domain of nvidia.com designates
 216.228.117.161 as permitted sender) receiver=protection.outlook.com;
 client-ip=216.228.117.161; helo=mail.nvidia.com; pr=C
From:   Dragos Tatulea <dtatulea@nvidia.com>
To:     "Michael S . Tsirkin" <mst@redhat.com>,
        Jason Wang <jasowang@redhat.com>,
        Eugenio Perez Martin <eperezma@redhat.com>,
        Si-Wei Liu <si-wei.liu@oracle.com>,
        Saeed Mahameed <saeedm@nvidia.com>,
        Leon Romanovsky <leon@kernel.org>,
        <virtualization@lists.linux-foundation.org>
CC:     Dragos Tatulea <dtatulea@nvidia.com>, <kvm@vger.kernel.org>,
        <linux-kernel@vger.kernel.org>, Parav Pandit <parav@nvidia.com>,
        Xuan Zhuo <xuanzhuo@linux.alibaba.com>
Subject: [PATCH vhost v4 13/16] vdpa/mlx5: Introduce mr for vq descriptor
Date:   Wed, 18 Oct 2023 20:14:52 +0300
Message-ID: <20231018171456.1624030-15-dtatulea@nvidia.com>
In-Reply-To: <20231018171456.1624030-2-dtatulea@nvidia.com>
References: <20231018171456.1624030-2-dtatulea@nvidia.com>
MIME-Version: 1.0
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: 8bit
X-MS-Exchange-CrossTenant-OriginalArrivalTime: 18 Oct 2023 17:16:25.7881
 (UTC)
X-MS-Exchange-CrossTenant-Network-Message-Id: cd83ee9a-e141-43fa-95b0-08dbcffdf5bb
X-MS-Exchange-CrossTenant-Id: 43083d15-7273-40c1-b7db-39efd9ccc17a
X-MS-Exchange-CrossTenant-OriginalAttributedTenantConnectingIp: TenantId=43083d15-7273-40c1-b7db-39efd9ccc17a;Ip=[216.228.117.161];Helo=[mail.nvidia.com]
X-MS-Exchange-CrossTenant-AuthSource: CY4PEPF0000FCBF.namprd03.prod.outlook.com
X-MS-Exchange-CrossTenant-AuthAs: Anonymous
X-MS-Exchange-CrossTenant-FromEntityHeader: HybridOnPrem
X-MS-Exchange-Transport-CrossTenantHeadersStamped: SA0PR12MB4527
X-Spam-Status: No, score=-0.9 required=5.0 tests=DKIMWL_WL_HIGH,DKIM_SIGNED,
	DKIM_VALID,DKIM_VALID_AU,HEADER_FROM_DIFFERENT_DOMAINS,
	MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS autolearn=unavailable
	autolearn_force=no version=3.4.6
X-Spam-Checker-Version: SpamAssassin 3.4.6 (2021-04-09) on lipwig.vger.email
Precedence: bulk
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org
X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.6.4 (lipwig.vger.email [0.0.0.0]); Wed, 18 Oct 2023 10:18:06 -0700 (PDT)

Introduce the vq descriptor group and mr per ASID. Until now
.set_map on ASID 1 was only updating the cvq iotlb. From now on it also
creates a mkey for it. The current patch doesn't use it but follow-up
patches will add hardware support for mapping the vq descriptors.

Acked-by: Jason Wang <jasowang@redhat.com>
Acked-by: Eugenio Pérez <eperezma@redhat.com>
Signed-off-by: Dragos Tatulea <dtatulea@nvidia.com>
---
 drivers/vdpa/mlx5/core/mlx5_vdpa.h |  5 +++--
 drivers/vdpa/mlx5/core/mr.c        | 14 +++++++++-----
 drivers/vdpa/mlx5/net/mlx5_vnet.c  | 20 +++++++++++++-------
 3 files changed, 25 insertions(+), 14 deletions(-)

diff --git a/drivers/vdpa/mlx5/core/mlx5_vdpa.h b/drivers/vdpa/mlx5/core/mlx5_vdpa.h
index bbe4335106bd..ae09296f4270 100644
--- a/drivers/vdpa/mlx5/core/mlx5_vdpa.h
+++ b/drivers/vdpa/mlx5/core/mlx5_vdpa.h
@@ -70,11 +70,12 @@ struct mlx5_vdpa_wq_ent {
 enum {
 	MLX5_VDPA_DATAVQ_GROUP,
 	MLX5_VDPA_CVQ_GROUP,
+	MLX5_VDPA_DATAVQ_DESC_GROUP,
 	MLX5_VDPA_NUMVQ_GROUPS
 };
 
 enum {
-	MLX5_VDPA_NUM_AS = MLX5_VDPA_NUMVQ_GROUPS
+	MLX5_VDPA_NUM_AS = 2
 };
 
 struct mlx5_vdpa_dev {
@@ -89,7 +90,7 @@ struct mlx5_vdpa_dev {
 	u16 max_idx;
 	u32 generation;
 
-	struct mlx5_vdpa_mr *mr;
+	struct mlx5_vdpa_mr *mr[MLX5_VDPA_NUM_AS];
 	/* serialize mr access */
 	struct mutex mr_mtx;
 	struct mlx5_control_vq cvq;
diff --git a/drivers/vdpa/mlx5/core/mr.c b/drivers/vdpa/mlx5/core/mr.c
index 00eff5a07152..3dee6d9bed6b 100644
--- a/drivers/vdpa/mlx5/core/mr.c
+++ b/drivers/vdpa/mlx5/core/mr.c
@@ -511,8 +511,10 @@ void mlx5_vdpa_destroy_mr(struct mlx5_vdpa_dev *mvdev,
 
 	_mlx5_vdpa_destroy_mr(mvdev, mr);
 
-	if (mvdev->mr == mr)
-		mvdev->mr = NULL;
+	for (int i = 0; i < MLX5_VDPA_NUM_AS; i++) {
+		if (mvdev->mr[i] == mr)
+			mvdev->mr[i] = NULL;
+	}
 
 	mutex_unlock(&mvdev->mr_mtx);
 
@@ -523,11 +525,11 @@ 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;
+	struct mlx5_vdpa_mr *old_mr = mvdev->mr[asid];
 
 	mutex_lock(&mvdev->mr_mtx);
 
-	mvdev->mr = new_mr;
+	mvdev->mr[asid] = new_mr;
 	if (old_mr) {
 		_mlx5_vdpa_destroy_mr(mvdev, old_mr);
 		kfree(old_mr);
@@ -539,7 +541,9 @@ void mlx5_vdpa_update_mr(struct mlx5_vdpa_dev *mvdev,
 
 void mlx5_vdpa_destroy_mr_resources(struct mlx5_vdpa_dev *mvdev)
 {
-	mlx5_vdpa_destroy_mr(mvdev, mvdev->mr);
+	for (int i = 0; i < MLX5_VDPA_NUM_AS; i++)
+		mlx5_vdpa_destroy_mr(mvdev, mvdev->mr[i]);
+
 	prune_iotlb(mvdev);
 }
 
diff --git a/drivers/vdpa/mlx5/net/mlx5_vnet.c b/drivers/vdpa/mlx5/net/mlx5_vnet.c
index 7b878995b6aa..ea76c0b4b78e 100644
--- a/drivers/vdpa/mlx5/net/mlx5_vnet.c
+++ b/drivers/vdpa/mlx5/net/mlx5_vnet.c
@@ -821,6 +821,8 @@ static int create_virtqueue(struct mlx5_vdpa_net *ndev, struct mlx5_vdpa_virtque
 {
 	int inlen = MLX5_ST_SZ_BYTES(create_virtio_net_q_in);
 	u32 out[MLX5_ST_SZ_DW(create_virtio_net_q_out)] = {};
+	struct mlx5_vdpa_dev *mvdev = &ndev->mvdev;
+	struct mlx5_vdpa_mr *vq_mr;
 	void *obj_context;
 	u16 mlx_features;
 	void *cmd_hdr;
@@ -873,7 +875,9 @@ static int create_virtqueue(struct mlx5_vdpa_net *ndev, struct mlx5_vdpa_virtque
 	MLX5_SET64(virtio_q, vq_ctx, desc_addr, mvq->desc_addr);
 	MLX5_SET64(virtio_q, vq_ctx, used_addr, mvq->device_addr);
 	MLX5_SET64(virtio_q, vq_ctx, available_addr, mvq->driver_addr);
-	MLX5_SET(virtio_q, vq_ctx, virtio_q_mkey, ndev->mvdev.mr->mkey);
+	vq_mr = mvdev->mr[mvdev->group2asid[MLX5_VDPA_DATAVQ_GROUP]];
+	if (vq_mr)
+		MLX5_SET(virtio_q, vq_ctx, virtio_q_mkey, vq_mr->mkey);
 	MLX5_SET(virtio_q, vq_ctx, umem_1_id, mvq->umem1.id);
 	MLX5_SET(virtio_q, vq_ctx, umem_1_size, mvq->umem1.size);
 	MLX5_SET(virtio_q, vq_ctx, umem_2_id, mvq->umem2.id);
@@ -2633,7 +2637,8 @@ static void restore_channels_info(struct mlx5_vdpa_net *ndev)
 }
 
 static int mlx5_vdpa_change_map(struct mlx5_vdpa_dev *mvdev,
-				struct mlx5_vdpa_mr *new_mr, unsigned int asid)
+				struct mlx5_vdpa_mr *new_mr,
+				unsigned int asid)
 {
 	struct mlx5_vdpa_net *ndev = to_mlx5_vdpa_ndev(mvdev);
 	int err;
@@ -2652,8 +2657,10 @@ static int mlx5_vdpa_change_map(struct mlx5_vdpa_dev *mvdev,
 
 	restore_channels_info(ndev);
 	err = setup_driver(mvdev);
+	if (err)
+		return err;
 
-	return err;
+	return 0;
 }
 
 /* reslock must be held for this function */
@@ -2869,8 +2876,8 @@ static int set_map_data(struct mlx5_vdpa_dev *mvdev, struct vhost_iotlb *iotlb,
 	struct mlx5_vdpa_mr *new_mr;
 	int err;
 
-	if (mvdev->group2asid[MLX5_VDPA_DATAVQ_GROUP] != asid)
-		goto end;
+	if (asid >= MLX5_VDPA_NUM_AS)
+		return -EINVAL;
 
 	if (vhost_iotlb_itree_first(iotlb, 0, U64_MAX)) {
 		new_mr = mlx5_vdpa_create_mr(mvdev, iotlb);
@@ -2884,7 +2891,7 @@ static int set_map_data(struct mlx5_vdpa_dev *mvdev, struct vhost_iotlb *iotlb,
 		new_mr = NULL;
 	}
 
-	if (!mvdev->mr) {
+	if (!mvdev->mr[asid]) {
 		mlx5_vdpa_update_mr(mvdev, new_mr, asid);
 	} else {
 		err = mlx5_vdpa_change_map(mvdev, new_mr, asid);
@@ -2894,7 +2901,6 @@ static int set_map_data(struct mlx5_vdpa_dev *mvdev, struct vhost_iotlb *iotlb,
 		}
 	}
 
-end:
 	return mlx5_vdpa_update_cvq_iotlb(mvdev, iotlb, asid);
 
 out_err:
-- 
2.41.0