net/bnxt: add separate mutex for FW health check

def_cp_lock was added to sync race between dev_configure and
int_handler. It should not be used to synchronize scheduling of FW
health check between dev_start and async event handler as well,
use a separate mutex for the same.

Fixes: a73b8e939f10 ("net/bnxt: fix race between start and interrupt handler")
Cc: stable@dpdk.org
Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
Reviewed-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>

diff --git a/drivers/net/bnxt/bnxt.h b/drivers/net/bnxt/bnxt.h
index f0b0800..bb26599 100644 (file)
--- a/drivers/net/bnxt/bnxt.h
+++ b/drivers/net/bnxt/bnxt.h
@@ -712,6 +712,7 @@ struct bnxt {
        rte_iova_t                      hwrm_short_cmd_req_dma_addr;
        rte_spinlock_t                  hwrm_lock;
        pthread_mutex_t                 def_cp_lock;
+       pthread_mutex_t                 health_check_lock;
        uint16_t                        max_req_len;
        uint16_t                        max_resp_len;
        uint16_t                        hwrm_max_ext_req_len;

diff --git a/drivers/net/bnxt/bnxt_ethdev.c b/drivers/net/bnxt/bnxt_ethdev.c
index 7c27e24..05e9a6a 100644 (file)
--- a/drivers/net/bnxt/bnxt_ethdev.c
+++ b/drivers/net/bnxt/bnxt_ethdev.c
@@ -1252,9 +1252,7 @@ static int bnxt_dev_start_op(struct rte_eth_dev *eth_dev)
        eth_dev->rx_pkt_burst = bnxt_receive_function(eth_dev);
        eth_dev->tx_pkt_burst = bnxt_transmit_function(eth_dev);
 
-       pthread_mutex_lock(&bp->def_cp_lock);
        bnxt_schedule_fw_health_check(bp);
-       pthread_mutex_unlock(&bp->def_cp_lock);
 
        return 0;
 
@@ -4675,17 +4673,22 @@ void bnxt_schedule_fw_health_check(struct bnxt *bp)
 {
        uint32_t polling_freq;
 
+       pthread_mutex_lock(&bp->health_check_lock);
+
        if (!bnxt_is_recovery_enabled(bp))
-               return;
+               goto done;
 
        if (bp->flags & BNXT_FLAG_FW_HEALTH_CHECK_SCHEDULED)
-               return;
+               goto done;
 
        polling_freq = bp->recovery_info->driver_polling_freq;
 
        rte_eal_alarm_set(US_PER_MS * polling_freq,
                          bnxt_check_fw_health, (void *)bp);
        bp->flags |= BNXT_FLAG_FW_HEALTH_CHECK_SCHEDULED;
+
+done:
+       pthread_mutex_unlock(&bp->health_check_lock);
 }
 
 static void bnxt_cancel_fw_health_check(struct bnxt *bp)
@@ -5473,6 +5476,10 @@ bnxt_init_locks(struct bnxt *bp)
        err = pthread_mutex_init(&bp->def_cp_lock, NULL);
        if (err)
                PMD_DRV_LOG(ERR, "Unable to initialize def_cp_lock\n");
+
+       err = pthread_mutex_init(&bp->health_check_lock, NULL);
+       if (err)
+               PMD_DRV_LOG(ERR, "Unable to initialize health_check_lock\n");
        return err;
 }
 
@@ -5884,6 +5891,7 @@ bnxt_uninit_locks(struct bnxt *bp)
 {
        pthread_mutex_destroy(&bp->flow_lock);
        pthread_mutex_destroy(&bp->def_cp_lock);
+       pthread_mutex_destroy(&bp->health_check_lock);
        if (bp->rep_info) {
                pthread_mutex_destroy(&bp->rep_info->vfr_lock);
                pthread_mutex_destroy(&bp->rep_info->vfr_start_lock);