be2net: stop issuing FW cmds if any cmd times out

BugLink: http://bugs.launchpad.net/bugs/921793

A FW cmd timeout (with a sufficiently large timeout value in the
order of tens of seconds) indicates an unresponsive FW. In this state
issuing further cmds and waiting for a completion will only stall the process.

Signed-off-by: Sathya Perla <sathya.perla@emulex.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Tim Gardner <tim.gardner@canonical.com>

diff --git a/drivers/net/ethernet/emulex/benet/be.h b/drivers/net/ethernet/emulex/benet/be.h
index 4163980..34f162d 100644 (file)
--- a/drivers/net/ethernet/emulex/benet/be.h
+++ b/drivers/net/ethernet/emulex/benet/be.h
@@ -350,6 +350,8 @@ struct be_adapter {
        u32 beacon_state;       /* for set_phys_id */
 
        bool eeh_err;
+       bool ue_detected;
+       bool fw_timeout;
        u32 port_num;
        bool promiscuous;
        bool wol;
@@ -357,7 +359,6 @@ struct be_adapter {
        u32 function_caps;
        u32 rx_fc;              /* Rx flow control */
        u32 tx_fc;              /* Tx flow control */
-       bool ue_detected;
        bool stats_cmd_sent;
        int link_speed;
        u8 port_type;
@@ -522,6 +523,11 @@ static inline bool be_multi_rxq(const struct be_adapter *adapter)
        return adapter->num_rx_qs > 1;
 }
 
+static inline bool be_error(struct be_adapter *adapter)
+{
+       return adapter->eeh_err || adapter->ue_detected || adapter->fw_timeout;
+}
+
 extern void be_cq_notify(struct be_adapter *adapter, u16 qid, bool arm,
                u16 num_popped);
 extern void be_link_status_update(struct be_adapter *adapter, u32 link_status);

diff --git a/drivers/net/ethernet/emulex/benet/be_cmds.c b/drivers/net/ethernet/emulex/benet/be_cmds.c
index 94cd77c..ad3eef0 100644 (file)
--- a/drivers/net/ethernet/emulex/benet/be_cmds.c
+++ b/drivers/net/ethernet/emulex/benet/be_cmds.c
@@ -31,7 +31,7 @@ static void be_mcc_notify(struct be_adapter *adapter)
        struct be_queue_info *mccq = &adapter->mcc_obj.q;
        u32 val = 0;
 
-       if (adapter->eeh_err)
+       if (be_error(adapter))
                return;
 
        val |= mccq->id & DB_MCCQ_RING_ID_MASK;
@@ -263,10 +263,10 @@ static int be_mcc_wait_compl(struct be_adapter *adapter)
        int i, num, status = 0;
        struct be_mcc_obj *mcc_obj = &adapter->mcc_obj;
 
-       if (adapter->eeh_err)
-               return -EIO;
-
        for (i = 0; i < mcc_timeout; i++) {
+               if (be_error(adapter))
+                       return -EIO;
+
                num = be_process_mcc(adapter, &status);
                if (num)
                        be_cq_notify(adapter, mcc_obj->cq.id,
@@ -277,7 +277,8 @@ static int be_mcc_wait_compl(struct be_adapter *adapter)
                udelay(100);
        }
        if (i == mcc_timeout) {
-               dev_err(&adapter->pdev->dev, "mccq poll timed out\n");
+               dev_err(&adapter->pdev->dev, "FW not responding\n");
+               adapter->fw_timeout = true;
                return -1;
        }
        return status;
@@ -295,10 +296,10 @@ static int be_mbox_db_ready_wait(struct be_adapter *adapter, void __iomem *db)
        int msecs = 0;
        u32 ready;
 
-       if (adapter->eeh_err)
-               return -EIO;
-
        do {
+               if (be_error(adapter))
+                       return -EIO;
+
                ready = ioread32(db);
                if (ready == 0xffffffff)
                        return -1;
@@ -308,7 +309,8 @@ static int be_mbox_db_ready_wait(struct be_adapter *adapter, void __iomem *db)
                        break;
 
                if (msecs > 4000) {
-                       dev_err(&adapter->pdev->dev, "mbox poll timed out\n");
+                       dev_err(&adapter->pdev->dev, "FW not responding\n");
+                       adapter->fw_timeout = true;
                        be_detect_dump_ue(adapter);
                        return -1;
                }
@@ -546,9 +548,6 @@ int be_cmd_fw_clean(struct be_adapter *adapter)
        u8 *wrb;
        int status;
 
-       if (adapter->eeh_err)
-               return -EIO;
-
        if (mutex_lock_interruptible(&adapter->mbox_lock))
                return -1;
 
@@ -1012,9 +1011,6 @@ int be_cmd_q_destroy(struct be_adapter *adapter, struct be_queue_info *q,
        u8 subsys = 0, opcode = 0;
        int status;
 
-       if (adapter->eeh_err)
-               return -EIO;
-
        if (mutex_lock_interruptible(&adapter->mbox_lock))
                return -1;
 
@@ -1136,9 +1132,6 @@ int be_cmd_if_destroy(struct be_adapter *adapter, int interface_id, u32 domain)
        struct be_cmd_req_if_destroy *req;
        int status;
 
-       if (adapter->eeh_err)
-               return -EIO;
-
        if (interface_id == -1)
                return 0;
 

diff --git a/drivers/net/ethernet/emulex/benet/be_main.c b/drivers/net/ethernet/emulex/benet/be_main.c
index 0e97b6d..ce20d64 100644 (file)
--- a/drivers/net/ethernet/emulex/benet/be_main.c
+++ b/drivers/net/ethernet/emulex/benet/be_main.c
@@ -3569,6 +3569,8 @@ static pci_ers_result_t be_eeh_reset(struct pci_dev *pdev)
 
        dev_info(&adapter->pdev->dev, "EEH reset\n");
        adapter->eeh_err = false;
+       adapter->ue_detected = false;
+       adapter->fw_timeout = false;
 
        status = pci_enable_device(pdev);
        if (status)