net: hns3: Fix for PF mailbox receving unknown message

Before the firmware updates the crq's tail pointer, if the PF driver
reads the data in the crq, the data may be incomplete at this time,
which will lead to the driver read an unknown message.

This patch fixes it by checking if crq is not empty before reading the
message.

Fixes: c1a81619d73a ("net: hns3: Add mailbox interrupt handling to PF driver")
Signed-off-by: Xi Wang <wangxi11@huawei.com>
Signed-off-by: Peng Li <lipeng321@huawei.com>
Signed-off-by: Salil Mehta <salil.mehta@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_mbx.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_mbx.c
index b6ae26b..31f3d9a 100644 (file)
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_mbx.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_mbx.c
@@ -382,6 +382,13 @@ static void hclge_reset_vf(struct hclge_vport *vport,
        hclge_func_reset_cmd(hdev, mbx_req->mbx_src_vfid);
 }
 
+static bool hclge_cmd_crq_empty(struct hclge_hw *hw)
+{
+       u32 tail = hclge_read_dev(hw, HCLGE_NIC_CRQ_TAIL_REG);
+
+       return tail == hw->cmq.crq.next_to_use;
+}
+
 void hclge_mbx_handler(struct hclge_dev *hdev)
 {
        struct hclge_cmq_ring *crq = &hdev->hw.cmq.crq;
@@ -390,12 +397,23 @@ void hclge_mbx_handler(struct hclge_dev *hdev)
        struct hclge_desc *desc;
        int ret, flag;
 
-       flag = le16_to_cpu(crq->desc[crq->next_to_use].flag);
        /* handle all the mailbox requests in the queue */
-       while (hnae_get_bit(flag, HCLGE_CMDQ_RX_OUTVLD_B)) {
+       while (!hclge_cmd_crq_empty(&hdev->hw)) {
                desc = &crq->desc[crq->next_to_use];
                req = (struct hclge_mbx_vf_to_pf_cmd *)desc->data;
 
+               flag = le16_to_cpu(crq->desc[crq->next_to_use].flag);
+               if (unlikely(!hnae_get_bit(flag, HCLGE_CMDQ_RX_OUTVLD_B))) {
+                       dev_warn(&hdev->pdev->dev,
+                                "dropped invalid mailbox message, code = %d\n",
+                                req->msg[0]);
+
+                       /* dropping/not processing this invalid message */
+                       crq->desc[crq->next_to_use].flag = 0;
+                       hclge_mbx_ring_ptr_move_crq(crq);
+                       continue;
+               }
+
                vport = &hdev->vport[req->mbx_src_vfid];
 
                switch (req->msg[0]) {
@@ -470,7 +488,6 @@ void hclge_mbx_handler(struct hclge_dev *hdev)
                }
                crq->desc[crq->next_to_use].flag = 0;
                hclge_mbx_ring_ptr_move_crq(crq);
-               flag = le16_to_cpu(crq->desc[crq->next_to_use].flag);
        }
 
        /* Write back CMDQ_RQ header pointer, M7 need this pointer */