1 files changed, 135 insertions, 44 deletions

diff --git a/drivers/net/ethernet/intel/igc/igc_main.c b/drivers/net/ethernet/intel/igc/igc_main.c
index 4d975d620a8e..b5bcabab7a1d 100644
--- a/drivers/net/ethernet/intel/igc/igc_main.c
+++ b/drivers/net/ethernet/intel/igc/igc_main.c
@@ -2712,8 +2712,7 @@ static struct sk_buff *igc_construct_skb_zc(struct igc_ring *ring,
 
 	net_prefetch(xdp->data_meta);
 
-	skb = __napi_alloc_skb(&ring->q_vector->napi, totalsize,
-			       GFP_ATOMIC | __GFP_NOWARN);
+	skb = napi_alloc_skb(&ring->q_vector->napi, totalsize);
 	if (unlikely(!skb))
 		return NULL;
 
@@ -2874,6 +2873,89 @@ static void igc_update_tx_stats(struct igc_q_vector *q_vector,
 	q_vector->tx.total_packets += packets;
 }
 
+static void igc_xsk_request_timestamp(void *_priv)
+{
+	struct igc_metadata_request *meta_req = _priv;
+	struct igc_ring *tx_ring = meta_req->tx_ring;
+	struct igc_tx_timestamp_request *tstamp;
+	u32 tx_flags = IGC_TX_FLAGS_TSTAMP;
+	struct igc_adapter *adapter;
+	unsigned long lock_flags;
+	bool found = false;
+	int i;
+
+	if (test_bit(IGC_RING_FLAG_TX_HWTSTAMP, &tx_ring->flags)) {
+		adapter = netdev_priv(tx_ring->netdev);
+
+		spin_lock_irqsave(&adapter->ptp_tx_lock, lock_flags);
+
+		/* Search for available tstamp regs */
+		for (i = 0; i < IGC_MAX_TX_TSTAMP_REGS; i++) {
+			tstamp = &adapter->tx_tstamp[i];
+
+			/* tstamp->skb and tstamp->xsk_tx_buffer are in union.
+			 * When tstamp->skb is equal to NULL,
+			 * tstamp->xsk_tx_buffer is equal to NULL as well.
+			 * This condition means that the particular tstamp reg
+			 * is not occupied by other packet.
+			 */
+			if (!tstamp->skb) {
+				found = true;
+				break;
+			}
+		}
+
+		/* Return if no available tstamp regs */
+		if (!found) {
+			adapter->tx_hwtstamp_skipped++;
+			spin_unlock_irqrestore(&adapter->ptp_tx_lock,
+					       lock_flags);
+			return;
+		}
+
+		tstamp->start = jiffies;
+		tstamp->xsk_queue_index = tx_ring->queue_index;
+		tstamp->xsk_tx_buffer = meta_req->tx_buffer;
+		tstamp->buffer_type = IGC_TX_BUFFER_TYPE_XSK;
+
+		/* Hold the transmit completion until timestamp is ready */
+		meta_req->tx_buffer->xsk_pending_ts = true;
+
+		/* Keep the pointer to tx_timestamp, which is located in XDP
+		 * metadata area. It is the location to store the value of
+		 * tx hardware timestamp.
+		 */
+		xsk_tx_metadata_to_compl(meta_req->meta, &tstamp->xsk_meta);
+
+		/* Set timestamp bit based on the _TSTAMP(_X) bit. */
+		tx_flags |= tstamp->flags;
+		meta_req->cmd_type |= IGC_SET_FLAG(tx_flags,
+						   IGC_TX_FLAGS_TSTAMP,
+						   (IGC_ADVTXD_MAC_TSTAMP));
+		meta_req->cmd_type |= IGC_SET_FLAG(tx_flags,
+						   IGC_TX_FLAGS_TSTAMP_1,
+						   (IGC_ADVTXD_TSTAMP_REG_1));
+		meta_req->cmd_type |= IGC_SET_FLAG(tx_flags,
+						   IGC_TX_FLAGS_TSTAMP_2,
+						   (IGC_ADVTXD_TSTAMP_REG_2));
+		meta_req->cmd_type |= IGC_SET_FLAG(tx_flags,
+						   IGC_TX_FLAGS_TSTAMP_3,
+						   (IGC_ADVTXD_TSTAMP_REG_3));
+
+		spin_unlock_irqrestore(&adapter->ptp_tx_lock, lock_flags);
+	}
+}
+
+static u64 igc_xsk_fill_timestamp(void *_priv)
+{
+	return *(u64 *)_priv;
+}
+
+const struct xsk_tx_metadata_ops igc_xsk_tx_metadata_ops = {
+	.tmo_request_timestamp		= igc_xsk_request_timestamp,
+	.tmo_fill_timestamp		= igc_xsk_fill_timestamp,
+};
+
 static void igc_xdp_xmit_zc(struct igc_ring *ring)
 {
 	struct xsk_buff_pool *pool = ring->xsk_pool;
@@ -2895,24 +2977,34 @@ static void igc_xdp_xmit_zc(struct igc_ring *ring)
 	budget = igc_desc_unused(ring);
 
 	while (xsk_tx_peek_desc(pool, &xdp_desc) && budget--) {
-		u32 cmd_type, olinfo_status;
+		struct igc_metadata_request meta_req;
+		struct xsk_tx_metadata *meta = NULL;
 		struct igc_tx_buffer *bi;
+		u32 olinfo_status;
 		dma_addr_t dma;
 
-		cmd_type = IGC_ADVTXD_DTYP_DATA | IGC_ADVTXD_DCMD_DEXT |
-			   IGC_ADVTXD_DCMD_IFCS | IGC_TXD_DCMD |
-			   xdp_desc.len;
+		meta_req.cmd_type = IGC_ADVTXD_DTYP_DATA |
+				    IGC_ADVTXD_DCMD_DEXT |
+				    IGC_ADVTXD_DCMD_IFCS |
+				    IGC_TXD_DCMD | xdp_desc.len;
 		olinfo_status = xdp_desc.len << IGC_ADVTXD_PAYLEN_SHIFT;
 
 		dma = xsk_buff_raw_get_dma(pool, xdp_desc.addr);
+		meta = xsk_buff_get_metadata(pool, xdp_desc.addr);
 		xsk_buff_raw_dma_sync_for_device(pool, dma, xdp_desc.len);
+		bi = &ring->tx_buffer_info[ntu];
+
+		meta_req.tx_ring = ring;
+		meta_req.tx_buffer = bi;
+		meta_req.meta = meta;
+		xsk_tx_metadata_request(meta, &igc_xsk_tx_metadata_ops,
+					&meta_req);
 
 		tx_desc = IGC_TX_DESC(ring, ntu);
-		tx_desc->read.cmd_type_len = cpu_to_le32(cmd_type);
+		tx_desc->read.cmd_type_len = cpu_to_le32(meta_req.cmd_type);
 		tx_desc->read.olinfo_status = cpu_to_le32(olinfo_status);
 		tx_desc->read.buffer_addr = cpu_to_le64(dma);
 
-		bi = &ring->tx_buffer_info[ntu];
 		bi->type = IGC_TX_BUFFER_TYPE_XSK;
 		bi->protocol = 0;
 		bi->bytecount = xdp_desc.len;
@@ -2975,6 +3067,13 @@ static bool igc_clean_tx_irq(struct igc_q_vector *q_vector, int napi_budget)
 		if (!(eop_desc->wb.status & cpu_to_le32(IGC_TXD_STAT_DD)))
 			break;
 
+		/* Hold the completions while there's a pending tx hardware
+		 * timestamp request from XDP Tx metadata.
+		 */
+		if (tx_buffer->type == IGC_TX_BUFFER_TYPE_XSK &&
+		    tx_buffer->xsk_pending_ts)
+			break;
+
 		/* clear next_to_watch to prevent false hangs */
 		tx_buffer->next_to_watch = NULL;
 
@@ -5176,7 +5275,7 @@ static int igc_change_mtu(struct net_device *netdev, int new_mtu)
 		igc_down(adapter);
 
 	netdev_dbg(netdev, "changing MTU from %d to %d\n", netdev->mtu, new_mtu);
-	netdev->mtu = new_mtu;
+	WRITE_ONCE(netdev->mtu, new_mtu);
 
 	if (netif_running(netdev))
 		igc_up(adapter);
@@ -5930,15 +6029,6 @@ static int __igc_open(struct net_device *netdev, bool resuming)
 	if (err)
 		goto err_req_irq;
 
-	/* Notify the stack of the actual queue counts. */
-	err = netif_set_real_num_tx_queues(netdev, adapter->num_tx_queues);
-	if (err)
-		goto err_set_queues;
-
-	err = netif_set_real_num_rx_queues(netdev, adapter->num_rx_queues);
-	if (err)
-		goto err_set_queues;
-
 	clear_bit(__IGC_DOWN, &adapter->state);
 
 	for (i = 0; i < adapter->num_q_vectors; i++)
@@ -5959,8 +6049,6 @@ static int __igc_open(struct net_device *netdev, bool resuming)
 
 	return IGC_SUCCESS;
 
-err_set_queues:
-	igc_free_irq(adapter);
 err_req_irq:
 	igc_release_hw_control(adapter);
 	igc_power_down_phy_copper_base(&adapter->hw);
@@ -5977,6 +6065,17 @@ err_setup_tx:
 
 int igc_open(struct net_device *netdev)
 {
+	struct igc_adapter *adapter = netdev_priv(netdev);
+	int err;
+
+	/* Notify the stack of the actual queue counts. */
+	err = netif_set_real_num_queues(netdev, adapter->num_tx_queues,
+					adapter->num_rx_queues);
+	if (err) {
+		netdev_err(netdev, "error setting real queue count\n");
+		return err;
+	}
+
 	return __igc_open(netdev, false);
 }
 
@@ -6803,6 +6902,7 @@ static int igc_probe(struct pci_dev *pdev,
 
 	netdev->netdev_ops = &igc_netdev_ops;
 	netdev->xdp_metadata_ops = &igc_xdp_metadata_ops;
+	netdev->xsk_tx_metadata_ops = &igc_xsk_tx_metadata_ops;
 	igc_ethtool_set_ops(netdev);
 	netdev->watchdog_timeo = 5 * HZ;
 
@@ -6928,8 +7028,6 @@ static int igc_probe(struct pci_dev *pdev,
 	device_set_wakeup_enable(&adapter->pdev->dev,
 				 adapter->flags & IGC_FLAG_WOL_SUPPORTED);
 
-	igc_ptp_init(adapter);
-
 	igc_tsn_clear_schedule(adapter);
 
 	/* reset the hardware with the new settings */
@@ -6951,6 +7049,9 @@ static int igc_probe(struct pci_dev *pdev,
 	/* Check if Media Autosense is enabled */
 	adapter->ei = *ei;
 
+	/* do hw tstamp init after resetting */
+	igc_ptp_init(adapter);
+
 	/* print pcie link status and MAC address */
 	pcie_print_link_status(pdev);
 	netdev_info(netdev, "MAC: %pM\n", netdev->dev_addr);
@@ -7108,8 +7209,7 @@ static int __igc_shutdown(struct pci_dev *pdev, bool *enable_wake,
 	return 0;
 }
 
-#ifdef CONFIG_PM
-static int __maybe_unused igc_runtime_suspend(struct device *dev)
+static int igc_runtime_suspend(struct device *dev)
 {
 	return __igc_shutdown(to_pci_dev(dev), NULL, 1);
 }
@@ -7144,7 +7244,7 @@ static void igc_deliver_wake_packet(struct net_device *netdev)
 	netif_rx(skb);
 }
 
-static int __maybe_unused igc_resume(struct device *dev)
+static int igc_resume(struct device *dev)
 {
 	struct pci_dev *pdev = to_pci_dev(dev);
 	struct net_device *netdev = pci_get_drvdata(pdev);
@@ -7186,23 +7286,21 @@ static int __maybe_unused igc_resume(struct device *dev)
 
 	wr32(IGC_WUS, ~0);
 
-	rtnl_lock();
-	if (!err && netif_running(netdev))
+	if (netif_running(netdev)) {
 		err = __igc_open(netdev, true);
-
-	if (!err)
-		netif_device_attach(netdev);
-	rtnl_unlock();
+		if (!err)
+			netif_device_attach(netdev);
+	}
 
 	return err;
 }
 
-static int __maybe_unused igc_runtime_resume(struct device *dev)
+static int igc_runtime_resume(struct device *dev)
 {
 	return igc_resume(dev);
 }
 
-static int __maybe_unused igc_suspend(struct device *dev)
+static int igc_suspend(struct device *dev)
 {
 	return __igc_shutdown(to_pci_dev(dev), NULL, 0);
 }
@@ -7217,7 +7315,6 @@ static int __maybe_unused igc_runtime_idle(struct device *dev)
 
 	return -EBUSY;
 }
-#endif /* CONFIG_PM */
 
 static void igc_shutdown(struct pci_dev *pdev)
 {
@@ -7332,22 +7429,16 @@ static const struct pci_error_handlers igc_err_handler = {
 	.resume = igc_io_resume,
 };
 
-#ifdef CONFIG_PM
-static const struct dev_pm_ops igc_pm_ops = {
-	SET_SYSTEM_SLEEP_PM_OPS(igc_suspend, igc_resume)
-	SET_RUNTIME_PM_OPS(igc_runtime_suspend, igc_runtime_resume,
-			   igc_runtime_idle)
-};
-#endif
+static _DEFINE_DEV_PM_OPS(igc_pm_ops, igc_suspend, igc_resume,
+			  igc_runtime_suspend, igc_runtime_resume,
+			  igc_runtime_idle);
 
 static struct pci_driver igc_driver = {
 	.name     = igc_driver_name,
 	.id_table = igc_pci_tbl,
 	.probe    = igc_probe,
 	.remove   = igc_remove,
-#ifdef CONFIG_PM
-	.driver.pm = &igc_pm_ops,
-#endif
+	.driver.pm = pm_ptr(&igc_pm_ops),
 	.shutdown = igc_shutdown,
 	.err_handler = &igc_err_handler,
 };