[dpdk-dev] [PATCH v2 11/26] net/ena: add checking for admin queue state

[Michal Krawczyk <mk@semihalf.com>]

The admin queue can stop responding or became inactive due to unexpected
behaviour of the device. In that case, the whole device should be
restarted.

Signed-off-by: Michal Krawczyk <mk@semihalf.com>
---
 drivers/net/ena/ena_ethdev.c | 37 +++++++++++++++++++++++++++++--------
 drivers/net/ena/ena_ethdev.h |  2 ++
 2 files changed, 31 insertions(+), 8 deletions(-)

diff --git a/drivers/net/ena/ena_ethdev.c b/drivers/net/ena/ena_ethdev.c
index 98f1b6c0a..d8447215a 100644
--- a/drivers/net/ena/ena_ethdev.c
+++ b/drivers/net/ena/ena_ethdev.c
@@ -527,6 +527,8 @@ ena_dev_reset(struct rte_eth_dev *dev)
 	for (i = 0; i < nb_queues; ++i)
 		ena_tx_queue_setup(eth_dev, i, adapter->tx_ring_size, 0, NULL);
 
+	adapter->trigger_reset = false;
+
 	return 0;
 }
 
@@ -1400,21 +1402,40 @@ static void ena_interrupt_handler_rte(void *cb_arg)
 		ena_com_aenq_intr_handler(ena_dev, adapter);
 }
 
+static void check_for_missing_keep_alive(struct ena_adapter *adapter)
+{
+	if (adapter->keep_alive_timeout == ENA_HW_HINTS_NO_TIMEOUT)
+		return;
+
+	if (unlikely((rte_get_timer_cycles() - adapter->timestamp_wd) >=
+	    adapter->keep_alive_timeout)) {
+		RTE_LOG(ERR, PMD, "Keep alive timeout\n");
+		adapter->reset_reason = ENA_REGS_RESET_KEEP_ALIVE_TO;
+		adapter->trigger_reset = true;
+	}
+}
+
+/* Check if admin queue is enabled */
+static void check_for_admin_com_state(struct ena_adapter *adapter)
+{
+	if (unlikely(!ena_com_get_admin_running_state(&adapter->ena_dev))) {
+		RTE_LOG(ERR, PMD, "ENA admin queue is not in running state!\n");
+		adapter->reset_reason = ENA_REGS_RESET_ADMIN_TO;
+		adapter->trigger_reset = true;
+	}
+}
+
 static void ena_timer_wd_callback(__rte_unused struct rte_timer *timer,
 				  void *arg)
 {
 	struct ena_adapter *adapter = (struct ena_adapter *)arg;
 	struct rte_eth_dev *dev = adapter->rte_dev;
 
-	if (adapter->keep_alive_timeout == ENA_HW_HINTS_NO_TIMEOUT)
-		return;
+	check_for_missing_keep_alive(adapter);
+	check_for_admin_com_state(adapter);
 
-	/* Within reasonable timing range no memory barriers are needed */
-	if ((rte_get_timer_cycles() - adapter->timestamp_wd) >=
-	    adapter->keep_alive_timeout) {
-		RTE_LOG(ERR, PMD, "The ENA device is not responding - "
-			"performing device reset...");
-		adapter->reset_reason = ENA_REGS_RESET_KEEP_ALIVE_TO;
+	if (unlikely(adapter->trigger_reset)) {
+		RTE_LOG(ERR, PMD, "Trigger reset is on\n");
 		_rte_eth_dev_callback_process(dev, RTE_ETH_EVENT_INTR_RESET,
 			NULL);
 	}
diff --git a/drivers/net/ena/ena_ethdev.h b/drivers/net/ena/ena_ethdev.h
index b44cca23e..1f6a7062f 100644
--- a/drivers/net/ena/ena_ethdev.h
+++ b/drivers/net/ena/ena_ethdev.h
@@ -194,6 +194,8 @@ struct ena_adapter {
 	struct rte_timer timer_wd;
 	uint64_t timestamp_wd;
 	uint64_t keep_alive_timeout;
+
+	bool trigger_reset;
 };
 
 #endif /* _ENA_ETHDEV_H_ */
-- 
2.14.1