net: hns3: fix race conditions between reset and module loading & unloading

When loading or unloading module, it should wait for the reset task
done before it un-initializes the client, otherwise the reset task
may cause a NULL pointer reference.

Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: Peng Li <lipeng321@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
index 0772ca176d93eb454969f3fcc736b3d60d298623..53dbef9ba77184b5fed63976f765d34fb8a9ad43 100644 (file)
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
@@ -8282,13 +8282,21 @@ static int hclge_init_nic_client_instance(struct hnae3_ae_dev *ae_dev,
 {
        struct hnae3_client *client = vport->nic.client;
        struct hclge_dev *hdev = ae_dev->priv;
+       int rst_cnt;
        int ret;
 
+       rst_cnt = hdev->rst_stats.reset_cnt;
        ret = client->ops->init_instance(&vport->nic);
        if (ret)
                return ret;
 
        set_bit(HCLGE_STATE_NIC_REGISTERED, &hdev->state);
+       if (test_bit(HCLGE_STATE_RST_HANDLING, &hdev->state) ||
+           rst_cnt != hdev->rst_stats.reset_cnt) {
+               ret = -EBUSY;
+               goto init_nic_err;
+       }
+
        hnae3_set_client_init_flag(client, ae_dev, 1);
 
        /* Enable nic hw error interrupts */
@@ -8301,6 +8309,15 @@ static int hclge_init_nic_client_instance(struct hnae3_ae_dev *ae_dev,
                hclge_info_show(hdev);
 
        return ret;
+
+init_nic_err:
+       clear_bit(HCLGE_STATE_NIC_REGISTERED, &hdev->state);
+       while (test_bit(HCLGE_STATE_RST_HANDLING, &hdev->state))
+               msleep(HCLGE_WAIT_RESET_DONE);
+
+       client->ops->uninit_instance(&vport->nic, 0);
+
+       return ret;
 }
 
 static int hclge_init_roce_client_instance(struct hnae3_ae_dev *ae_dev,
@@ -8308,6 +8325,7 @@ static int hclge_init_roce_client_instance(struct hnae3_ae_dev *ae_dev,
 {
        struct hnae3_client *client = vport->roce.client;
        struct hclge_dev *hdev = ae_dev->priv;
+       int rst_cnt;
        int ret;
 
        if (!hnae3_dev_roce_supported(hdev) || !hdev->roce_client ||
@@ -8319,14 +8337,30 @@ static int hclge_init_roce_client_instance(struct hnae3_ae_dev *ae_dev,
        if (ret)
                return ret;
 
+       rst_cnt = hdev->rst_stats.reset_cnt;
        ret = client->ops->init_instance(&vport->roce);
        if (ret)
                return ret;
 
        set_bit(HCLGE_STATE_ROCE_REGISTERED, &hdev->state);
+       if (test_bit(HCLGE_STATE_RST_HANDLING, &hdev->state) ||
+           rst_cnt != hdev->rst_stats.reset_cnt) {
+               ret = -EBUSY;
+               goto init_roce_err;
+       }
+
        hnae3_set_client_init_flag(client, ae_dev, 1);
 
        return 0;
+
+init_roce_err:
+       clear_bit(HCLGE_STATE_ROCE_REGISTERED, &hdev->state);
+       while (test_bit(HCLGE_STATE_RST_HANDLING, &hdev->state))
+               msleep(HCLGE_WAIT_RESET_DONE);
+
+       hdev->roce_client->ops->uninit_instance(&vport->roce, 0);
+
+       return ret;
 }
 
 static int hclge_init_client_instance(struct hnae3_client *client,
@@ -8398,6 +8432,9 @@ static void hclge_uninit_client_instance(struct hnae3_client *client,
                vport = &hdev->vport[i];
                if (hdev->roce_client) {
                        clear_bit(HCLGE_STATE_ROCE_REGISTERED, &hdev->state);
+                       while (test_bit(HCLGE_STATE_RST_HANDLING, &hdev->state))
+                               msleep(HCLGE_WAIT_RESET_DONE);
+
                        hdev->roce_client->ops->uninit_instance(&vport->roce,
                                                                0);
                        hdev->roce_client = NULL;
@@ -8407,6 +8444,9 @@ static void hclge_uninit_client_instance(struct hnae3_client *client,
                        return;
                if (hdev->nic_client && client->ops->uninit_instance) {
                        clear_bit(HCLGE_STATE_NIC_REGISTERED, &hdev->state);
+                       while (test_bit(HCLGE_STATE_RST_HANDLING, &hdev->state))
+                               msleep(HCLGE_WAIT_RESET_DONE);
+
                        client->ops->uninit_instance(&vport->nic, 0);
                        hdev->nic_client = NULL;
                        vport->nic.client = NULL;

diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h
index 8c4e47a4c266ecb0cdb0847bad63c55ef5078f10..6a12285f4c763edf94f931ccb020cf26081c2dae 100644 (file)
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h
@@ -700,6 +700,7 @@ struct hclge_mac_tnl_stats {
 };
 
 #define HCLGE_RESET_INTERVAL   (10 * HZ)
+#define HCLGE_WAIT_RESET_DONE  100
 
 #pragma pack(1)
 struct hclge_vf_vlan_cfg {