]> git.proxmox.com Git - mirror_ubuntu-zesty-kernel.git/blobdiff - drivers/net/ethernet/amazon/ena/ena_netdev.c
net: ena: allow the driver to work with small number of msix vectors
[mirror_ubuntu-zesty-kernel.git] / drivers / net / ethernet / amazon / ena / ena_netdev.c
index d6076cb2434b10f71079a6722819cb780222b099..5c05db40936883a5f9c0a91d4b230730a17813d2 100644 (file)
@@ -87,6 +87,7 @@ static void ena_tx_timeout(struct net_device *dev)
        if (test_and_set_bit(ENA_FLAG_TRIGGER_RESET, &adapter->flags))
                return;
 
+       adapter->reset_reason = ENA_REGS_RESET_OS_NETDEV_WD;
        u64_stats_update_begin(&adapter->syncp);
        adapter->dev_stats.tx_timeout++;
        u64_stats_update_end(&adapter->syncp);
@@ -303,6 +304,24 @@ static void ena_free_all_io_tx_resources(struct ena_adapter *adapter)
                ena_free_tx_resources(adapter, i);
 }
 
+static inline int validate_rx_req_id(struct ena_ring *rx_ring, u16 req_id)
+{
+       if (likely(req_id < rx_ring->ring_size))
+               return 0;
+
+       netif_err(rx_ring->adapter, rx_err, rx_ring->netdev,
+                 "Invalid rx req_id: %hu\n", req_id);
+
+       u64_stats_update_begin(&rx_ring->syncp);
+       rx_ring->rx_stats.bad_req_id++;
+       u64_stats_update_end(&rx_ring->syncp);
+
+       /* Trigger device reset */
+       rx_ring->adapter->reset_reason = ENA_REGS_RESET_INV_RX_REQ_ID;
+       set_bit(ENA_FLAG_TRIGGER_RESET, &rx_ring->adapter->flags);
+       return -EFAULT;
+}
+
 /* ena_setup_rx_resources - allocate I/O Rx resources (Descriptors)
  * @adapter: network interface device structure
  * @qid: queue index
@@ -314,7 +333,7 @@ static int ena_setup_rx_resources(struct ena_adapter *adapter,
 {
        struct ena_ring *rx_ring = &adapter->rx_ring[qid];
        struct ena_irq *ena_irq = &adapter->irq_tbl[ENA_IO_IRQ_IDX(qid)];
-       int size, node;
+       int size, node, i;
 
        if (rx_ring->rx_buffer_info) {
                netif_err(adapter, ifup, adapter->netdev,
@@ -335,6 +354,20 @@ static int ena_setup_rx_resources(struct ena_adapter *adapter,
                        return -ENOMEM;
        }
 
+       size = sizeof(u16) * rx_ring->ring_size;
+       rx_ring->free_rx_ids = vzalloc_node(size, node);
+       if (!rx_ring->free_rx_ids) {
+               rx_ring->free_rx_ids = vzalloc(size);
+               if (!rx_ring->free_rx_ids) {
+                       vfree(rx_ring->rx_buffer_info);
+                       return -ENOMEM;
+               }
+       }
+
+       /* Req id ring for receiving RX pkts out of order */
+       for (i = 0; i < rx_ring->ring_size; i++)
+               rx_ring->free_rx_ids[i] = i;
+
        /* Reset rx statistics */
        memset(&rx_ring->rx_stats, 0x0, sizeof(rx_ring->rx_stats));
 
@@ -358,6 +391,9 @@ static void ena_free_rx_resources(struct ena_adapter *adapter,
 
        vfree(rx_ring->rx_buffer_info);
        rx_ring->rx_buffer_info = NULL;
+
+       vfree(rx_ring->free_rx_ids);
+       rx_ring->free_rx_ids = NULL;
 }
 
 /* ena_setup_all_rx_resources - allocate I/O Rx queues resources for all queues
@@ -463,15 +499,22 @@ static void ena_free_rx_page(struct ena_ring *rx_ring,
 
 static int ena_refill_rx_bufs(struct ena_ring *rx_ring, u32 num)
 {
-       u16 next_to_use;
+       u16 next_to_use, req_id;
        u32 i;
        int rc;
 
        next_to_use = rx_ring->next_to_use;
 
        for (i = 0; i < num; i++) {
-               struct ena_rx_buffer *rx_info =
-                       &rx_ring->rx_buffer_info[next_to_use];
+               struct ena_rx_buffer *rx_info;
+
+               req_id = rx_ring->free_rx_ids[next_to_use];
+               rc = validate_rx_req_id(rx_ring, req_id);
+               if (unlikely(rc < 0))
+                       break;
+
+               rx_info = &rx_ring->rx_buffer_info[req_id];
+
 
                rc = ena_alloc_rx_page(rx_ring, rx_info,
                                       __GFP_COLD | GFP_ATOMIC | __GFP_COMP);
@@ -483,7 +526,7 @@ static int ena_refill_rx_bufs(struct ena_ring *rx_ring, u32 num)
                }
                rc = ena_com_add_single_rx_desc(rx_ring->ena_com_io_sq,
                                                &rx_info->ena_buf,
-                                               next_to_use);
+                                               req_id);
                if (unlikely(rc)) {
                        netif_warn(rx_ring->adapter, rx_status, rx_ring->netdev,
                                   "failed to add buffer for rx queue %d\n",
@@ -670,6 +713,7 @@ static int validate_tx_req_id(struct ena_ring *tx_ring, u16 req_id)
        u64_stats_update_end(&tx_ring->syncp);
 
        /* Trigger device reset */
+       tx_ring->adapter->reset_reason = ENA_REGS_RESET_INV_TX_REQ_ID;
        set_bit(ENA_FLAG_TRIGGER_RESET, &tx_ring->adapter->flags);
        return -EFAULT;
 }
@@ -787,13 +831,14 @@ static struct sk_buff *ena_rx_skb(struct ena_ring *rx_ring,
                                  u16 *next_to_clean)
 {
        struct sk_buff *skb;
-       struct ena_rx_buffer *rx_info =
-               &rx_ring->rx_buffer_info[*next_to_clean];
-       u32 len;
-       u32 buf = 0;
+       struct ena_rx_buffer *rx_info;
+       u16 len, req_id, buf = 0;
        void *va;
 
-       len = ena_bufs[0].len;
+       len = ena_bufs[buf].len;
+       req_id = ena_bufs[buf].req_id;
+       rx_info = &rx_ring->rx_buffer_info[req_id];
+
        if (unlikely(!rx_info->page)) {
                netif_err(rx_ring->adapter, rx_err, rx_ring->netdev,
                          "Page is NULL\n");
@@ -865,13 +910,18 @@ static struct sk_buff *ena_rx_skb(struct ena_ring *rx_ring,
                          skb->len, skb->data_len);
 
                rx_info->page = NULL;
+
+               rx_ring->free_rx_ids[*next_to_clean] = req_id;
                *next_to_clean =
                        ENA_RX_RING_IDX_NEXT(*next_to_clean,
                                             rx_ring->ring_size);
                if (likely(--descs == 0))
                        break;
-               rx_info = &rx_ring->rx_buffer_info[*next_to_clean];
-               len = ena_bufs[++buf].len;
+
+               buf++;
+               len = ena_bufs[buf].len;
+               req_id = ena_bufs[buf].req_id;
+               rx_info = &rx_ring->rx_buffer_info[req_id];
        } while (1);
 
        return skb;
@@ -972,6 +1022,7 @@ static int ena_clean_rx_irq(struct ena_ring *rx_ring, struct napi_struct *napi,
        int rc = 0;
        int total_len = 0;
        int rx_copybreak_pkt = 0;
+       int i;
 
        netif_dbg(rx_ring->adapter, rx_status, rx_ring->netdev,
                  "%s qid %d\n", __func__, rx_ring->qid);
@@ -1001,9 +1052,13 @@ static int ena_clean_rx_irq(struct ena_ring *rx_ring, struct napi_struct *napi,
 
                /* exit if we failed to retrieve a buffer */
                if (unlikely(!skb)) {
-                       next_to_clean = ENA_RX_RING_IDX_ADD(next_to_clean,
-                                                           ena_rx_ctx.descs,
-                                                           rx_ring->ring_size);
+                       for (i = 0; i < ena_rx_ctx.descs; i++) {
+                               rx_ring->free_tx_ids[next_to_clean] =
+                                       rx_ring->ena_bufs[i].req_id;
+                               next_to_clean =
+                                       ENA_RX_RING_IDX_NEXT(next_to_clean,
+                                                            rx_ring->ring_size);
+                       }
                        break;
                }
 
@@ -1055,6 +1110,7 @@ error:
        u64_stats_update_end(&rx_ring->syncp);
 
        /* Too many desc from the device. Trigger reset */
+       adapter->reset_reason = ENA_REGS_RESET_TOO_MANY_RX_DESCS;
        set_bit(ENA_FLAG_TRIGGER_RESET, &adapter->flags);
 
        return 0;
@@ -1213,9 +1269,20 @@ static irqreturn_t ena_intr_msix_io(int irq, void *data)
        return IRQ_HANDLED;
 }
 
+/* Reserve a single MSI-X vector for management (admin + aenq).
+ * plus reserve one vector for each potential io queue.
+ * the number of potential io queues is the minimum of what the device
+ * supports and the number of vCPUs.
+ */
 static int ena_enable_msix(struct ena_adapter *adapter, int num_queues)
 {
-       int msix_vecs, rc;
+       int msix_vecs, irq_cnt;
+
+       if (test_bit(ENA_FLAG_MSIX_ENABLED, &adapter->flags)) {
+               netif_err(adapter, probe, adapter->netdev,
+                         "Error, MSI-X is already enabled\n");
+               return -EPERM;
+       }
 
        /* Reserved the max msix vectors we might need */
        msix_vecs = ENA_MAX_MSIX_VEC(num_queues);
@@ -1223,25 +1290,28 @@ static int ena_enable_msix(struct ena_adapter *adapter, int num_queues)
        netif_dbg(adapter, probe, adapter->netdev,
                  "trying to enable MSI-X, vectors %d\n", msix_vecs);
 
-       rc = pci_alloc_irq_vectors(adapter->pdev, msix_vecs, msix_vecs,
-                       PCI_IRQ_MSIX);
-       if (rc < 0) {
+       irq_cnt = pci_alloc_irq_vectors(adapter->pdev, ENA_MIN_MSIX_VEC,
+                                       msix_vecs, PCI_IRQ_MSIX);
+
+       if (irq_cnt < 0) {
                netif_err(adapter, probe, adapter->netdev,
-                         "Failed to enable MSI-X, vectors %d rc %d\n",
-                         msix_vecs, rc);
+                         "Failed to enable MSI-X. irq_cnt %d\n", irq_cnt);
                return -ENOSPC;
        }
 
-       netif_dbg(adapter, probe, adapter->netdev, "enable MSI-X, vectors %d\n",
-                 msix_vecs);
-
-       if (msix_vecs >= 1) {
-               if (ena_init_rx_cpu_rmap(adapter))
-                       netif_warn(adapter, probe, adapter->netdev,
-                                  "Failed to map IRQs to CPUs\n");
+       if (irq_cnt != msix_vecs) {
+               netif_notice(adapter, probe, adapter->netdev,
+                            "enable only %d MSI-X (out of %d), reduce the number of queues\n",
+                            irq_cnt, msix_vecs);
+               adapter->num_queues = irq_cnt - ENA_ADMIN_MSIX_VEC;
        }
 
-       adapter->msix_vecs = msix_vecs;
+       if (ena_init_rx_cpu_rmap(adapter))
+               netif_warn(adapter, probe, adapter->netdev,
+                          "Failed to map IRQs to CPUs\n");
+
+       adapter->msix_vecs = irq_cnt;
+       set_bit(ENA_FLAG_MSIX_ENABLED, &adapter->flags);
 
        return 0;
 }
@@ -1318,6 +1388,12 @@ static int ena_request_io_irq(struct ena_adapter *adapter)
        struct ena_irq *irq;
        int rc = 0, i, k;
 
+       if (!test_bit(ENA_FLAG_MSIX_ENABLED, &adapter->flags)) {
+               netif_err(adapter, ifup, adapter->netdev,
+                         "Failed to request I/O IRQ: MSI-X is not enabled\n");
+               return -EINVAL;
+       }
+
        for (i = ENA_IO_IRQ_FIRST_IDX; i < adapter->msix_vecs; i++) {
                irq = &adapter->irq_tbl[i];
                rc = request_irq(irq->vector, irq->handler, flags, irq->name,
@@ -1376,6 +1452,12 @@ static void ena_free_io_irq(struct ena_adapter *adapter)
        }
 }
 
+static void ena_disable_msix(struct ena_adapter *adapter)
+{
+       if (test_and_clear_bit(ENA_FLAG_MSIX_ENABLED, &adapter->flags))
+               pci_free_irq_vectors(adapter->pdev);
+}
+
 static void ena_disable_io_intr_sync(struct ena_adapter *adapter)
 {
        int i;
@@ -1446,7 +1528,7 @@ static int ena_rss_configure(struct ena_adapter *adapter)
        /* In case the RSS table wasn't initialized by probe */
        if (!ena_dev->rss.tbl_log_size) {
                rc = ena_rss_init_default(adapter);
-               if (rc && (rc != -EPERM)) {
+               if (rc && (rc != -EOPNOTSUPP)) {
                        netif_err(adapter, ifup, adapter->netdev,
                                  "Failed to init RSS rc: %d\n", rc);
                        return rc;
@@ -1455,17 +1537,17 @@ static int ena_rss_configure(struct ena_adapter *adapter)
 
        /* Set indirect table */
        rc = ena_com_indirect_table_set(ena_dev);
-       if (unlikely(rc && rc != -EPERM))
+       if (unlikely(rc && rc != -EOPNOTSUPP))
                return rc;
 
        /* Configure hash function (if supported) */
        rc = ena_com_set_hash_function(ena_dev);
-       if (unlikely(rc && (rc != -EPERM)))
+       if (unlikely(rc && (rc != -EOPNOTSUPP)))
                return rc;
 
        /* Configure hash inputs (if supported) */
        rc = ena_com_set_hash_ctrl(ena_dev);
-       if (unlikely(rc && (rc != -EPERM)))
+       if (unlikely(rc && (rc != -EOPNOTSUPP)))
                return rc;
 
        return 0;
@@ -1720,7 +1802,7 @@ static void ena_down(struct ena_adapter *adapter)
        if (test_bit(ENA_FLAG_TRIGGER_RESET, &adapter->flags)) {
                int rc;
 
-               rc = ena_com_dev_reset(adapter->ena_dev);
+               rc = ena_com_dev_reset(adapter->ena_dev, adapter->reset_reason);
                if (rc)
                        dev_err(&adapter->pdev->dev, "Device reset failed\n");
        }
@@ -2144,7 +2226,7 @@ static void ena_config_host_info(struct ena_com_dev *ena_dev)
 
        rc = ena_com_set_host_attributes(ena_dev);
        if (rc) {
-               if (rc == -EPERM)
+               if (rc == -EOPNOTSUPP)
                        pr_warn("Cannot set host attributes\n");
                else
                        pr_err("Cannot set host attributes\n");
@@ -2181,7 +2263,7 @@ static void ena_config_debug_area(struct ena_adapter *adapter)
 
        rc = ena_com_set_host_attributes(adapter->ena_dev);
        if (rc) {
-               if (rc == -EPERM)
+               if (rc == -EOPNOTSUPP)
                        netif_warn(adapter, drv, adapter->netdev,
                                   "Cannot set host attributes\n");
                else
@@ -2355,7 +2437,7 @@ static int ena_device_init(struct ena_com_dev *ena_dev, struct pci_dev *pdev,
        readless_supported = !(pdev->revision & ENA_MMIO_DISABLE_REG_READ);
        ena_com_set_mmio_read_mode(ena_dev, readless_supported);
 
-       rc = ena_com_dev_reset(ena_dev);
+       rc = ena_com_dev_reset(ena_dev, ENA_REGS_RESET_NORMAL);
        if (rc) {
                dev_err(dev, "Can not reset device\n");
                goto err_mmio_read_less;
@@ -2466,7 +2548,8 @@ static int ena_enable_msix_and_set_admin_interrupts(struct ena_adapter *adapter,
        return 0;
 
 err_disable_msix:
-       pci_free_irq_vectors(adapter->pdev);
+       ena_disable_msix(adapter);
+
        return rc;
 }
 
@@ -2504,7 +2587,7 @@ static void ena_fw_reset_device(struct work_struct *work)
 
        ena_free_mgmnt_irq(adapter);
 
-       pci_free_irq_vectors(adapter->pdev);
+       ena_disable_msix(adapter);
 
        ena_com_abort_admin_commands(ena_dev);
 
@@ -2514,6 +2597,7 @@ static void ena_fw_reset_device(struct work_struct *work)
 
        ena_com_mmio_reg_read_request_destroy(ena_dev);
 
+       adapter->reset_reason = ENA_REGS_RESET_NORMAL;
        clear_bit(ENA_FLAG_TRIGGER_RESET, &adapter->flags);
 
        /* Finish with the destroy part. Start the init part */
@@ -2555,7 +2639,7 @@ static void ena_fw_reset_device(struct work_struct *work)
        return;
 err_disable_msix:
        ena_free_mgmnt_irq(adapter);
-       pci_free_irq_vectors(adapter->pdev);
+       ena_disable_msix(adapter);
 err_device_destroy:
        ena_com_admin_destroy(ena_dev);
 err:
@@ -2579,7 +2663,7 @@ static int check_missing_comp_in_queue(struct ena_adapter *adapter,
                tx_buf = &tx_ring->tx_buffer_info[i];
                last_jiffies = tx_buf->last_jiffies;
                if (unlikely(last_jiffies &&
-                            time_is_before_jiffies(last_jiffies + TX_TIMEOUT))) {
+                            time_is_before_jiffies(last_jiffies + adapter->missing_tx_completion_to))) {
                        if (!tx_buf->print_once)
                                netif_notice(adapter, tx_err, adapter->netdev,
                                             "Found a Tx that wasn't completed on time, qid %d, index %d.\n",
@@ -2588,10 +2672,13 @@ static int check_missing_comp_in_queue(struct ena_adapter *adapter,
                        tx_buf->print_once = 1;
                        missed_tx++;
 
-                       if (unlikely(missed_tx > MAX_NUM_OF_TIMEOUTED_PACKETS)) {
+                       if (unlikely(missed_tx > adapter->missing_tx_completion_threshold)) {
                                netif_err(adapter, tx_err, adapter->netdev,
                                          "The number of lost tx completions is above the threshold (%d > %d). Reset the device\n",
-                                         missed_tx, MAX_NUM_OF_TIMEOUTED_PACKETS);
+                                         missed_tx,
+                                         adapter->missing_tx_completion_threshold);
+                               adapter->reset_reason =
+                                       ENA_REGS_RESET_MISS_TX_CMPL;
                                set_bit(ENA_FLAG_TRIGGER_RESET, &adapter->flags);
                                return -EIO;
                        }
@@ -2615,6 +2702,9 @@ static void check_for_missing_tx_completions(struct ena_adapter *adapter)
        if (test_bit(ENA_FLAG_TRIGGER_RESET, &adapter->flags))
                return;
 
+       if (adapter->missing_tx_completion_to == ENA_HW_HINTS_NO_TIMEOUT)
+               return;
+
        budget = ENA_MONITORED_TX_QUEUES;
 
        for (i = adapter->last_monitored_tx_qid; i < adapter->num_queues; i++) {
@@ -2692,14 +2782,18 @@ static void check_for_missing_keep_alive(struct ena_adapter *adapter)
        if (!adapter->wd_state)
                return;
 
-       keep_alive_expired = round_jiffies(adapter->last_keep_alive_jiffies
-                                          + ENA_DEVICE_KALIVE_TIMEOUT);
+       if (adapter->keep_alive_timeout == ENA_HW_HINTS_NO_TIMEOUT)
+               return;
+
+       keep_alive_expired = round_jiffies(adapter->last_keep_alive_jiffies +
+                                          adapter->keep_alive_timeout);
        if (unlikely(time_is_before_jiffies(keep_alive_expired))) {
                netif_err(adapter, drv, adapter->netdev,
                          "Keep alive watchdog timeout.\n");
                u64_stats_update_begin(&adapter->syncp);
                adapter->dev_stats.wd_expired++;
                u64_stats_update_end(&adapter->syncp);
+               adapter->reset_reason = ENA_REGS_RESET_KEEP_ALIVE_TO;
                set_bit(ENA_FLAG_TRIGGER_RESET, &adapter->flags);
        }
 }
@@ -2712,10 +2806,49 @@ static void check_for_admin_com_state(struct ena_adapter *adapter)
                u64_stats_update_begin(&adapter->syncp);
                adapter->dev_stats.admin_q_pause++;
                u64_stats_update_end(&adapter->syncp);
+               adapter->reset_reason = ENA_REGS_RESET_ADMIN_TO;
                set_bit(ENA_FLAG_TRIGGER_RESET, &adapter->flags);
        }
 }
 
+static void ena_update_hints(struct ena_adapter *adapter,
+                            struct ena_admin_ena_hw_hints *hints)
+{
+       struct net_device *netdev = adapter->netdev;
+
+       if (hints->admin_completion_tx_timeout)
+               adapter->ena_dev->admin_queue.completion_timeout =
+                       hints->admin_completion_tx_timeout * 1000;
+
+       if (hints->mmio_read_timeout)
+               /* convert to usec */
+               adapter->ena_dev->mmio_read.reg_read_to =
+                       hints->mmio_read_timeout * 1000;
+
+       if (hints->missed_tx_completion_count_threshold_to_reset)
+               adapter->missing_tx_completion_threshold =
+                       hints->missed_tx_completion_count_threshold_to_reset;
+
+       if (hints->missing_tx_completion_timeout) {
+               if (hints->missing_tx_completion_timeout == ENA_HW_HINTS_NO_TIMEOUT)
+                       adapter->missing_tx_completion_to = ENA_HW_HINTS_NO_TIMEOUT;
+               else
+                       adapter->missing_tx_completion_to =
+                               msecs_to_jiffies(hints->missing_tx_completion_timeout);
+       }
+
+       if (hints->netdev_wd_timeout)
+               netdev->watchdog_timeo = msecs_to_jiffies(hints->netdev_wd_timeout);
+
+       if (hints->driver_watchdog_timeout) {
+               if (hints->driver_watchdog_timeout == ENA_HW_HINTS_NO_TIMEOUT)
+                       adapter->keep_alive_timeout = ENA_HW_HINTS_NO_TIMEOUT;
+               else
+                       adapter->keep_alive_timeout =
+                               msecs_to_jiffies(hints->driver_watchdog_timeout);
+       }
+}
+
 static void ena_update_host_info(struct ena_admin_host_info *host_info,
                                 struct net_device *netdev)
 {
@@ -2888,7 +3021,7 @@ static int ena_rss_init_default(struct ena_adapter *adapter)
                val = ethtool_rxfh_indir_default(i, adapter->num_queues);
                rc = ena_com_indirect_table_fill_entry(ena_dev, i,
                                                       ENA_IO_RXQ_IDX(val));
-               if (unlikely(rc && (rc != -EPERM))) {
+               if (unlikely(rc && (rc != -EOPNOTSUPP))) {
                        dev_err(dev, "Cannot fill indirect table\n");
                        goto err_fill_indir;
                }
@@ -2896,13 +3029,13 @@ static int ena_rss_init_default(struct ena_adapter *adapter)
 
        rc = ena_com_fill_hash_function(ena_dev, ENA_ADMIN_CRC32, NULL,
                                        ENA_HASH_KEY_SIZE, 0xFFFFFFFF);
-       if (unlikely(rc && (rc != -EPERM))) {
+       if (unlikely(rc && (rc != -EOPNOTSUPP))) {
                dev_err(dev, "Cannot fill hash function\n");
                goto err_fill_indir;
        }
 
        rc = ena_com_set_default_hash_ctrl(ena_dev);
-       if (unlikely(rc && (rc != -EPERM))) {
+       if (unlikely(rc && (rc != -EOPNOTSUPP))) {
                dev_err(dev, "Cannot fill hash control\n");
                goto err_fill_indir;
        }
@@ -3078,6 +3211,7 @@ static int ena_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
        ena_set_conf_feat_params(adapter, &get_feat_ctx);
 
        adapter->msg_enable = netif_msg_init(debug, DEFAULT_MSG_ENABLE);
+       adapter->reset_reason = ENA_REGS_RESET_NORMAL;
 
        adapter->tx_ring_size = queue_size;
        adapter->rx_ring_size = queue_size;
@@ -3116,7 +3250,7 @@ static int ena_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
                goto err_worker_destroy;
        }
        rc = ena_rss_init_default(adapter);
-       if (rc && (rc != -EPERM)) {
+       if (rc && (rc != -EOPNOTSUPP)) {
                dev_err(&pdev->dev, "Cannot init RSS rc: %d\n", rc);
                goto err_free_msix;
        }
@@ -3138,6 +3272,11 @@ static int ena_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
        INIT_WORK(&adapter->reset_task, ena_fw_reset_device);
 
        adapter->last_keep_alive_jiffies = jiffies;
+       adapter->keep_alive_timeout = ENA_DEVICE_KALIVE_TIMEOUT;
+       adapter->missing_tx_completion_to = TX_TIMEOUT;
+       adapter->missing_tx_completion_threshold = MAX_NUM_OF_TIMEOUTED_PACKETS;
+
+       ena_update_hints(adapter, &get_feat_ctx.hw_hints);
 
        setup_timer(&adapter->timer_service, ena_timer_service,
                    (unsigned long)adapter);
@@ -3157,9 +3296,9 @@ err_rss:
        ena_com_delete_debug_area(ena_dev);
        ena_com_rss_destroy(ena_dev);
 err_free_msix:
-       ena_com_dev_reset(ena_dev);
+       ena_com_dev_reset(ena_dev, ENA_REGS_RESET_INIT_ERR);
        ena_free_mgmnt_irq(adapter);
-       pci_free_irq_vectors(adapter->pdev);
+       ena_disable_msix(adapter);
 err_worker_destroy:
        ena_com_destroy_interrupt_moderation(ena_dev);
        del_timer(&adapter->timer_service);
@@ -3240,11 +3379,11 @@ static void ena_remove(struct pci_dev *pdev)
 
        /* Reset the device only if the device is running. */
        if (test_bit(ENA_FLAG_DEVICE_RUNNING, &adapter->flags))
-               ena_com_dev_reset(ena_dev);
+               ena_com_dev_reset(ena_dev, adapter->reset_reason);
 
        ena_free_mgmnt_irq(adapter);
 
-       pci_free_irq_vectors(adapter->pdev);
+       ena_disable_msix(adapter);
 
        free_netdev(netdev);
 
@@ -3339,6 +3478,7 @@ static void ena_notification(void *adapter_data,
                             struct ena_admin_aenq_entry *aenq_e)
 {
        struct ena_adapter *adapter = (struct ena_adapter *)adapter_data;
+       struct ena_admin_ena_hw_hints *hints;
 
        WARN(aenq_e->aenq_common_desc.group != ENA_ADMIN_NOTIFICATION,
             "Invalid group(%x) expected %x\n",
@@ -3356,6 +3496,11 @@ static void ena_notification(void *adapter_data,
        case ENA_ADMIN_RESUME:
                queue_work(ena_wq, &adapter->resume_io_task);
                break;
+       case ENA_ADMIN_UPDATE_HINTS:
+               hints = (struct ena_admin_ena_hw_hints *)
+                       (&aenq_e->inline_data_w4);
+               ena_update_hints(adapter, hints);
+               break;
        default:
                netif_err(adapter, drv, adapter->netdev,
                          "Invalid aenq notification link state %d\n",