bnxt_en: New Broadcom ethernet driver.

Broadcom ethernet driver for the new family of NetXtreme-C/E
ethernet devices.

v5:
  - Removed empty blank lines at end of files (noted by David Miller).
  - Moved busy poll helper functions to bnxt.h to at least make the
    .c file look less cluttered with #ifdef (noted by Stephen Hemminger).

v4:
  - Broke up 2 long message strings with "\n" (suggested by John Linville)
  - Constify an array of strings (suggested by Stephen Hemminger)
  - Improve bnxt_vf_pciid() (suggested by Stephen Hemminger)
  - Use PCI_VDEVICE() to populate pci_device_id table for more compact
    source.

v3:
  - Fixed 2 more sparse warnings.
  - Removed some unused structures in .h files.

v2:
  - Fixed all kbuild test robot reported warnings.
  - Fixed many of the checkpatch.pl errors and warnings.
  - Fixed the Kconfig description (noted by Dmitry Kravkov).

Acked-by: Eddie Wai <eddie.wai@broadcom.com>
Acked-by: Jeffrey Huang <huangjw@broadcom.com>
Signed-off-by: Prashant Sreedharan <prashant@broadcom.com>
Signed-off-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

12 files changed, 13034 insertions, 0 deletions

diff --git a/drivers/net/ethernet/broadcom/Kconfig b/drivers/net/ethernet/broadcom/Kconfig
index e930aa9a3cfb..67a7d520d9f5 100644
--- a/drivers/net/ethernet/broadcom/Kconfig
+++ b/drivers/net/ethernet/broadcom/Kconfig
@@ -170,4 +170,23 @@ config SYSTEMPORT
 	  Broadcom BCM7xxx Set Top Box family chipset using an internal
 	  Ethernet switch.
 
+config BNXT
+	tristate "Broadcom NetXtreme-C/E support"
+	depends on PCI
+	select FW_LOADER
+	select LIBCRC32C
+	---help---
+	  This driver supports Broadcom NetXtreme-C/E 10/25/40/50 gigabit
+	  Ethernet cards.  To compile this driver as a module, choose M here:
+	  the module will be called bnxt_en.  This is recommended.
+
+config BNXT_SRIOV
+	bool "Broadcom NetXtreme-C/E SR-IOV support"
+	depends on BNXT && PCI_IOV
+	default y
+	---help---
+	  This configuration parameter enables Single Root Input Output
+	  Virtualization support in the NetXtreme-C/E products. This
+	  allows for virtual function acceleration in virtual environments.
+
 endif # NET_VENDOR_BROADCOM
diff --git a/drivers/net/ethernet/broadcom/Makefile b/drivers/net/ethernet/broadcom/Makefile
index e2a958a657e0..00584d78b3e0 100644
--- a/drivers/net/ethernet/broadcom/Makefile
+++ b/drivers/net/ethernet/broadcom/Makefile
@@ -12,3 +12,4 @@ obj-$(CONFIG_SB1250_MAC) += sb1250-mac.o
 obj-$(CONFIG_TIGON3) += tg3.o
 obj-$(CONFIG_BGMAC) += bgmac.o
 obj-$(CONFIG_SYSTEMPORT) += bcmsysport.o
+obj-$(CONFIG_BNXT) += bnxt/
diff --git a/drivers/net/ethernet/broadcom/bnxt/Makefile b/drivers/net/ethernet/broadcom/bnxt/Makefile
new file mode 100644
index 000000000000..97e78e217928
--- /dev/null
+++ b/drivers/net/ethernet/broadcom/bnxt/Makefile
@@ -0,0 +1,3 @@
+obj-$(CONFIG_BNXT) += bnxt_en.o
+
+bnxt_en-y := bnxt.o bnxt_sriov.o bnxt_ethtool.o
diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
new file mode 100644
index 000000000000..94b11a59e703
--- /dev/null
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
@@ -0,0 +1,5746 @@
+/* Broadcom NetXtreme-C/E network driver.
+ *
+ * Copyright (c) 2014-2015 Broadcom Corporation
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation.
+ */
+
+#include <linux/module.h>
+
+#include <linux/stringify.h>
+#include <linux/kernel.h>
+#include <linux/timer.h>
+#include <linux/errno.h>
+#include <linux/ioport.h>
+#include <linux/slab.h>
+#include <linux/vmalloc.h>
+#include <linux/interrupt.h>
+#include <linux/pci.h>
+#include <linux/netdevice.h>
+#include <linux/etherdevice.h>
+#include <linux/skbuff.h>
+#include <linux/dma-mapping.h>
+#include <linux/bitops.h>
+#include <linux/io.h>
+#include <linux/irq.h>
+#include <linux/delay.h>
+#include <asm/byteorder.h>
+#include <asm/page.h>
+#include <linux/time.h>
+#include <linux/mii.h>
+#include <linux/if.h>
+#include <linux/if_vlan.h>
+#include <net/ip.h>
+#include <net/tcp.h>
+#include <net/udp.h>
+#include <net/checksum.h>
+#include <net/ip6_checksum.h>
+#if defined(CONFIG_VXLAN) || defined(CONFIG_VXLAN_MODULE)
+#include <net/vxlan.h>
+#endif
+#ifdef CONFIG_NET_RX_BUSY_POLL
+#include <net/busy_poll.h>
+#endif
+#include <linux/workqueue.h>
+#include <linux/prefetch.h>
+#include <linux/cache.h>
+#include <linux/log2.h>
+#include <linux/aer.h>
+#include <linux/bitmap.h>
+#include <linux/cpu_rmap.h>
+
+#include "bnxt_hsi.h"
+#include "bnxt.h"
+#include "bnxt_sriov.h"
+#include "bnxt_ethtool.h"
+
+#define BNXT_TX_TIMEOUT		(5 * HZ)
+
+static const char version[] =
+	"Broadcom NetXtreme-C/E driver " DRV_MODULE_NAME " v" DRV_MODULE_VERSION "\n";
+
+MODULE_LICENSE("GPL");
+MODULE_DESCRIPTION("Broadcom BCM573xx network driver");
+MODULE_VERSION(DRV_MODULE_VERSION);
+
+#define BNXT_RX_OFFSET (NET_SKB_PAD + NET_IP_ALIGN)
+#define BNXT_RX_DMA_OFFSET NET_SKB_PAD
+#define BNXT_RX_COPY_THRESH 256
+
+#define BNXT_TX_PUSH_THRESH 92
+
+enum board_idx {
+	BCM57302,
+	BCM57304,
+	BCM57404,
+	BCM57406,
+	BCM57304_VF,
+	BCM57404_VF,
+};
+
+/* indexed by enum above */
+static const struct {
+	char *name;
+} board_info[] = {
+	{ "Broadcom BCM57302 NetXtreme-C Single-port 10Gb/25Gb/40Gb/50Gb Ethernet" },
+	{ "Broadcom BCM57304 NetXtreme-C Dual-port 10Gb/25Gb/40Gb/50Gb Ethernet" },
+	{ "Broadcom BCM57404 NetXtreme-E Dual-port 10Gb/25Gb Ethernet" },
+	{ "Broadcom BCM57406 NetXtreme-E Dual-port 10Gb Ethernet" },
+	{ "Broadcom BCM57304 NetXtreme-C Ethernet Virtual Function" },
+	{ "Broadcom BCM57404 NetXtreme-E Ethernet Virtual Function" },
+};
+
+static const struct pci_device_id bnxt_pci_tbl[] = {
+	{ PCI_VDEVICE(BROADCOM, 0x16c9), .driver_data = BCM57302 },
+	{ PCI_VDEVICE(BROADCOM, 0x16ca), .driver_data = BCM57304 },
+	{ PCI_VDEVICE(BROADCOM, 0x16d1), .driver_data = BCM57404 },
+	{ PCI_VDEVICE(BROADCOM, 0x16d2), .driver_data = BCM57406 },
+#ifdef CONFIG_BNXT_SRIOV
+	{ PCI_VDEVICE(BROADCOM, 0x16cb), .driver_data = BCM57304_VF },
+	{ PCI_VDEVICE(BROADCOM, 0x16d3), .driver_data = BCM57404_VF },
+#endif
+	{ 0 }
+};
+
+MODULE_DEVICE_TABLE(pci, bnxt_pci_tbl);
+
+static const u16 bnxt_vf_req_snif[] = {
+	HWRM_FUNC_CFG,
+	HWRM_PORT_PHY_QCFG,
+	HWRM_CFA_L2_FILTER_ALLOC,
+};
+
+static bool bnxt_vf_pciid(enum board_idx idx)
+{
+	return (idx == BCM57304_VF || idx == BCM57404_VF);
+}
+
+#define DB_CP_REARM_FLAGS	(DB_KEY_CP | DB_IDX_VALID)
+#define DB_CP_FLAGS		(DB_KEY_CP | DB_IDX_VALID | DB_IRQ_DIS)
+#define DB_CP_IRQ_DIS_FLAGS	(DB_KEY_CP | DB_IRQ_DIS)
+
+#define BNXT_CP_DB_REARM(db, raw_cons)					\
+		writel(DB_CP_REARM_FLAGS | RING_CMP(raw_cons), db)
+
+#define BNXT_CP_DB(db, raw_cons)					\
+		writel(DB_CP_FLAGS | RING_CMP(raw_cons), db)
+
+#define BNXT_CP_DB_IRQ_DIS(db)						\
+		writel(DB_CP_IRQ_DIS_FLAGS, db)
+
+static inline u32 bnxt_tx_avail(struct bnxt *bp, struct bnxt_tx_ring_info *txr)
+{
+	/* Tell compiler to fetch tx indices from memory. */
+	barrier();
+
+	return bp->tx_ring_size -
+		((txr->tx_prod - txr->tx_cons) & bp->tx_ring_mask);
+}
+
+static const u16 bnxt_lhint_arr[] = {
+	TX_BD_FLAGS_LHINT_512_AND_SMALLER,
+	TX_BD_FLAGS_LHINT_512_TO_1023,
+	TX_BD_FLAGS_LHINT_1024_TO_2047,
+	TX_BD_FLAGS_LHINT_1024_TO_2047,
+	TX_BD_FLAGS_LHINT_2048_AND_LARGER,
+	TX_BD_FLAGS_LHINT_2048_AND_LARGER,
+	TX_BD_FLAGS_LHINT_2048_AND_LARGER,
+	TX_BD_FLAGS_LHINT_2048_AND_LARGER,
+	TX_BD_FLAGS_LHINT_2048_AND_LARGER,
+	TX_BD_FLAGS_LHINT_2048_AND_LARGER,
+	TX_BD_FLAGS_LHINT_2048_AND_LARGER,
+	TX_BD_FLAGS_LHINT_2048_AND_LARGER,
+	TX_BD_FLAGS_LHINT_2048_AND_LARGER,
+	TX_BD_FLAGS_LHINT_2048_AND_LARGER,
+	TX_BD_FLAGS_LHINT_2048_AND_LARGER,
+	TX_BD_FLAGS_LHINT_2048_AND_LARGER,
+	TX_BD_FLAGS_LHINT_2048_AND_LARGER,
+	TX_BD_FLAGS_LHINT_2048_AND_LARGER,
+	TX_BD_FLAGS_LHINT_2048_AND_LARGER,
+};
+
+static netdev_tx_t bnxt_start_xmit(struct sk_buff *skb, struct net_device *dev)
+{
+	struct bnxt *bp = netdev_priv(dev);
+	struct tx_bd *txbd;
+	struct tx_bd_ext *txbd1;
+	struct netdev_queue *txq;
+	int i;
+	dma_addr_t mapping;
+	unsigned int length, pad = 0;
+	u32 len, free_size, vlan_tag_flags, cfa_action, flags;
+	u16 prod, last_frag;
+	struct pci_dev *pdev = bp->pdev;
+	struct bnxt_napi *bnapi;
+	struct bnxt_tx_ring_info *txr;
+	struct bnxt_sw_tx_bd *tx_buf;
+
+	i = skb_get_queue_mapping(skb);
+	if (unlikely(i >= bp->tx_nr_rings)) {
+		dev_kfree_skb_any(skb);
+		return NETDEV_TX_OK;
+	}
+
+	bnapi = bp->bnapi[i];
+	txr = &bnapi->tx_ring;
+	txq = netdev_get_tx_queue(dev, i);
+	prod = txr->tx_prod;
+
+	free_size = bnxt_tx_avail(bp, txr);
+	if (unlikely(free_size < skb_shinfo(skb)->nr_frags + 2)) {
+		netif_tx_stop_queue(txq);
+		return NETDEV_TX_BUSY;
+	}
+
+	length = skb->len;
+	len = skb_headlen(skb);
+	last_frag = skb_shinfo(skb)->nr_frags;
+
+	txbd = &txr->tx_desc_ring[TX_RING(prod)][TX_IDX(prod)];
+
+	txbd->tx_bd_opaque = prod;
+
+	tx_buf = &txr->tx_buf_ring[prod];
+	tx_buf->skb = skb;
+	tx_buf->nr_frags = last_frag;
+
+	vlan_tag_flags = 0;
+	cfa_action = 0;
+	if (skb_vlan_tag_present(skb)) {
+		vlan_tag_flags = TX_BD_CFA_META_KEY_VLAN |
+				 skb_vlan_tag_get(skb);
+		/* Currently supports 8021Q, 8021AD vlan offloads
+		 * QINQ1, QINQ2, QINQ3 vlan headers are deprecated
+		 */
+		if (skb->vlan_proto == htons(ETH_P_8021Q))
+			vlan_tag_flags |= 1 << TX_BD_CFA_META_TPID_SHIFT;
+	}
+
+	if (free_size == bp->tx_ring_size && length <= bp->tx_push_thresh) {
+		struct tx_push_bd *push = txr->tx_push;
+		struct tx_bd *tx_push = &push->txbd1;
+		struct tx_bd_ext *tx_push1 = &push->txbd2;
+		void *pdata = tx_push1 + 1;
+		int j;
+
+		/* Set COAL_NOW to be ready quickly for the next push */
+		tx_push->tx_bd_len_flags_type =
+			cpu_to_le32((length << TX_BD_LEN_SHIFT) |
+					TX_BD_TYPE_LONG_TX_BD |
+					TX_BD_FLAGS_LHINT_512_AND_SMALLER |
+					TX_BD_FLAGS_COAL_NOW |
+					TX_BD_FLAGS_PACKET_END |
+					(2 << TX_BD_FLAGS_BD_CNT_SHIFT));
+
+		if (skb->ip_summed == CHECKSUM_PARTIAL)
+			tx_push1->tx_bd_hsize_lflags =
+					cpu_to_le32(TX_BD_FLAGS_TCP_UDP_CHKSUM);
+		else
+			tx_push1->tx_bd_hsize_lflags = 0;
+
+		tx_push1->tx_bd_cfa_meta = cpu_to_le32(vlan_tag_flags);
+		tx_push1->tx_bd_cfa_action = cpu_to_le32(cfa_action);
+
+		skb_copy_from_linear_data(skb, pdata, len);
+		pdata += len;
+		for (j = 0; j < last_frag; j++) {
+			skb_frag_t *frag = &skb_shinfo(skb)->frags[j];
+			void *fptr;
+
+			fptr = skb_frag_address_safe(frag);
+			if (!fptr)
+				goto normal_tx;
+
+			memcpy(pdata, fptr, skb_frag_size(frag));
+			pdata += skb_frag_size(frag);
+		}
+
+		memcpy(txbd, tx_push, sizeof(*txbd));
+		prod = NEXT_TX(prod);
+		txbd = &txr->tx_desc_ring[TX_RING(prod)][TX_IDX(prod)];
+		memcpy(txbd, tx_push1, sizeof(*txbd));
+		prod = NEXT_TX(prod);
+		push->doorbell =
+			cpu_to_le32(DB_KEY_TX_PUSH | DB_LONG_TX_PUSH | prod);
+		txr->tx_prod = prod;
+
+		netdev_tx_sent_queue(txq, skb->len);
+
+		__iowrite64_copy(txr->tx_doorbell, push,
+				 (length + sizeof(*push) + 8) / 8);
+
+		tx_buf->is_push = 1;
+
+		goto tx_done;
+	}
+
+normal_tx:
+	if (length < BNXT_MIN_PKT_SIZE) {
+		pad = BNXT_MIN_PKT_SIZE - length;
+		if (skb_pad(skb, pad)) {
+			/* SKB already freed. */
+			tx_buf->skb = NULL;
+			return NETDEV_TX_OK;
+		}
+		length = BNXT_MIN_PKT_SIZE;
+	}
+
+	mapping = dma_map_single(&pdev->dev, skb->data, len, DMA_TO_DEVICE);
+
+	if (unlikely(dma_mapping_error(&pdev->dev, mapping))) {
+		dev_kfree_skb_any(skb);
+		tx_buf->skb = NULL;
+		return NETDEV_TX_OK;
+	}
+
+	dma_unmap_addr_set(tx_buf, mapping, mapping);
+	flags = (len << TX_BD_LEN_SHIFT) | TX_BD_TYPE_LONG_TX_BD |
+		((last_frag + 2) << TX_BD_FLAGS_BD_CNT_SHIFT);
+
+	txbd->tx_bd_haddr = cpu_to_le64(mapping);
+
+	prod = NEXT_TX(prod);
+	txbd1 = (struct tx_bd_ext *)
+		&txr->tx_desc_ring[TX_RING(prod)][TX_IDX(prod)];
+
+	txbd1->tx_bd_hsize_lflags = 0;
+	if (skb_is_gso(skb)) {
+		u32 hdr_len;
+
+		if (skb->encapsulation)
+			hdr_len = skb_inner_network_offset(skb) +
+				skb_inner_network_header_len(skb) +
+				inner_tcp_hdrlen(skb);
+		else
+			hdr_len = skb_transport_offset(skb) +
+				tcp_hdrlen(skb);
+
+		txbd1->tx_bd_hsize_lflags = cpu_to_le32(TX_BD_FLAGS_LSO |
+					TX_BD_FLAGS_T_IPID |
+					(hdr_len << (TX_BD_HSIZE_SHIFT - 1)));
+		length = skb_shinfo(skb)->gso_size;
+		txbd1->tx_bd_mss = cpu_to_le32(length);
+		length += hdr_len;
+	} else if (skb->ip_summed == CHECKSUM_PARTIAL) {
+		txbd1->tx_bd_hsize_lflags =
+			cpu_to_le32(TX_BD_FLAGS_TCP_UDP_CHKSUM);
+		txbd1->tx_bd_mss = 0;
+	}
+
+	length >>= 9;
+	flags |= bnxt_lhint_arr[length];
+	txbd->tx_bd_len_flags_type = cpu_to_le32(flags);
+
+	txbd1->tx_bd_cfa_meta = cpu_to_le32(vlan_tag_flags);
+	txbd1->tx_bd_cfa_action = cpu_to_le32(cfa_action);
+	for (i = 0; i < last_frag; i++) {
+		skb_frag_t *frag = &skb_shinfo(skb)->frags[i];
+
+		prod = NEXT_TX(prod);
+		txbd = &txr->tx_desc_ring[TX_RING(prod)][TX_IDX(prod)];
+
+		len = skb_frag_size(frag);
+		mapping = skb_frag_dma_map(&pdev->dev, frag, 0, len,
+					   DMA_TO_DEVICE);
+
+		if (unlikely(dma_mapping_error(&pdev->dev, mapping)))
+			goto tx_dma_error;
+
+		tx_buf = &txr->tx_buf_ring[prod];
+		dma_unmap_addr_set(tx_buf, mapping, mapping);
+
+		txbd->tx_bd_haddr = cpu_to_le64(mapping);
+
+		flags = len << TX_BD_LEN_SHIFT;
+		txbd->tx_bd_len_flags_type = cpu_to_le32(flags);
+	}
+
+	flags &= ~TX_BD_LEN;
+	txbd->tx_bd_len_flags_type =
+		cpu_to_le32(((len + pad) << TX_BD_LEN_SHIFT) | flags |
+			    TX_BD_FLAGS_PACKET_END);
+
+	netdev_tx_sent_queue(txq, skb->len);
+
+	/* Sync BD data before updating doorbell */
+	wmb();
+
+	prod = NEXT_TX(prod);
+	txr->tx_prod = prod;
+
+	writel(DB_KEY_TX | prod, txr->tx_doorbell);
+	writel(DB_KEY_TX | prod, txr->tx_doorbell);
+
+tx_done:
+
+	mmiowb();
+
+	if (unlikely(bnxt_tx_avail(bp, txr) <= MAX_SKB_FRAGS + 1)) {
+		netif_tx_stop_queue(txq);
+
+		/* netif_tx_stop_queue() must be done before checking
+		 * tx index in bnxt_tx_avail() below, because in
+		 * bnxt_tx_int(), we update tx index before checking for
+		 * netif_tx_queue_stopped().
+		 */
+		smp_mb();
+		if (bnxt_tx_avail(bp, txr) > bp->tx_wake_thresh)
+			netif_tx_wake_queue(txq);
+	}
+	return NETDEV_TX_OK;
+
+tx_dma_error:
+	last_frag = i;
+
+	/* start back at beginning and unmap skb */
+	prod = txr->tx_prod;
+	tx_buf = &txr->tx_buf_ring[prod];
+	tx_buf->skb = NULL;
+	dma_unmap_single(&pdev->dev, dma_unmap_addr(tx_buf, mapping),
+			 skb_headlen(skb), PCI_DMA_TODEVICE);
+	prod = NEXT_TX(prod);
+
+	/* unmap remaining mapped pages */
+	for (i = 0; i < last_frag; i++) {
+		prod = NEXT_TX(prod);
+		tx_buf = &txr->tx_buf_ring[prod];
+		dma_unmap_page(&pdev->dev, dma_unmap_addr(tx_buf, mapping),
+			       skb_frag_size(&skb_shinfo(skb)->frags[i]),
+			       PCI_DMA_TODEVICE);
+	}
+
+	dev_kfree_skb_any(skb);
+	return NETDEV_TX_OK;
+}
+
+static void bnxt_tx_int(struct bnxt *bp, struct bnxt_napi *bnapi, int nr_pkts)
+{
+	struct bnxt_tx_ring_info *txr = &bnapi->tx_ring;
+	int index = bnapi->index;
+	struct netdev_queue *txq = netdev_get_tx_queue(bp->dev, index);
+	u16 cons = txr->tx_cons;
+	struct pci_dev *pdev = bp->pdev;
+	int i;
+	unsigned int tx_bytes = 0;
+
+	for (i = 0; i < nr_pkts; i++) {
+		struct bnxt_sw_tx_bd *tx_buf;
+		struct sk_buff *skb;
+		int j, last;
+
+		tx_buf = &txr->tx_buf_ring[cons];
+		cons = NEXT_TX(cons);
+		skb = tx_buf->skb;
+		tx_buf->skb = NULL;
+
+		if (tx_buf->is_push) {
+			tx_buf->is_push = 0;
+			goto next_tx_int;
+		}
+
+		dma_unmap_single(&pdev->dev, dma_unmap_addr(tx_buf, mapping),
+				 skb_headlen(skb), PCI_DMA_TODEVICE);
+		last = tx_buf->nr_frags;
+
+		for (j = 0; j < last; j++) {
+			cons = NEXT_TX(cons);
+			tx_buf = &txr->tx_buf_ring[cons];
+			dma_unmap_page(
+				&pdev->dev,
+				dma_unmap_addr(tx_buf, mapping),
+				skb_frag_size(&skb_shinfo(skb)->frags[j]),
+				PCI_DMA_TODEVICE);
+		}
+
+next_tx_int:
+		cons = NEXT_TX(cons);
+
+		tx_bytes += skb->len;
+		dev_kfree_skb_any(skb);
+	}
+
+	netdev_tx_completed_queue(txq, nr_pkts, tx_bytes);
+	txr->tx_cons = cons;
+
+	/* Need to make the tx_cons update visible to bnxt_start_xmit()
+	 * before checking for netif_tx_queue_stopped().  Without the
+	 * memory barrier, there is a small possibility that bnxt_start_xmit()
+	 * will miss it and cause the queue to be stopped forever.
+	 */
+	smp_mb();
+
+	if (unlikely(netif_tx_queue_stopped(txq)) &&
+	    (bnxt_tx_avail(bp, txr) > bp->tx_wake_thresh)) {
+		__netif_tx_lock(txq, smp_processor_id());
+		if (netif_tx_queue_stopped(txq) &&
+		    bnxt_tx_avail(bp, txr) > bp->tx_wake_thresh &&
+		    txr->dev_state != BNXT_DEV_STATE_CLOSING)
+			netif_tx_wake_queue(txq);
+		__netif_tx_unlock(txq);
+	}
+}
+
+static inline u8 *__bnxt_alloc_rx_data(struct bnxt *bp, dma_addr_t *mapping,
+				       gfp_t gfp)
+{
+	u8 *data;
+	struct pci_dev *pdev = bp->pdev;
+
+	data = kmalloc(bp->rx_buf_size, gfp);
+	if (!data)
+		return NULL;
+
+	*mapping = dma_map_single(&pdev->dev, data + BNXT_RX_DMA_OFFSET,
+				  bp->rx_buf_use_size, PCI_DMA_FROMDEVICE);
+
+	if (dma_mapping_error(&pdev->dev, *mapping)) {
+		kfree(data);
+		data = NULL;
+	}
+	return data;
+}
+
+static inline int bnxt_alloc_rx_data(struct bnxt *bp,
+				     struct bnxt_rx_ring_info *rxr,
+				     u16 prod, gfp_t gfp)
+{
+	struct rx_bd *rxbd = &rxr->rx_desc_ring[RX_RING(prod)][RX_IDX(prod)];
+	struct bnxt_sw_rx_bd *rx_buf = &rxr->rx_buf_ring[prod];
+	u8 *data;
+	dma_addr_t mapping;
+
+	data = __bnxt_alloc_rx_data(bp, &mapping, gfp);
+	if (!data)
+		return -ENOMEM;
+
+	rx_buf->data = data;
+	dma_unmap_addr_set(rx_buf, mapping, mapping);
+
+	rxbd->rx_bd_haddr = cpu_to_le64(mapping);
+
+	return 0;
+}
+
+static void bnxt_reuse_rx_data(struct bnxt_rx_ring_info *rxr, u16 cons,
+			       u8 *data)
+{
+	u16 prod = rxr->rx_prod;
+	struct bnxt_sw_rx_bd *cons_rx_buf, *prod_rx_buf;
+	struct rx_bd *cons_bd, *prod_bd;
+
+	prod_rx_buf = &rxr->rx_buf_ring[prod];
+	cons_rx_buf = &rxr->rx_buf_ring[cons];
+
+	prod_rx_buf->data = data;
+
+	dma_unmap_addr_set(prod_rx_buf, mapping,
+			   dma_unmap_addr(cons_rx_buf, mapping));
+
+	prod_bd = &rxr->rx_desc_ring[RX_RING(prod)][RX_IDX(prod)];
+	cons_bd = &rxr->rx_desc_ring[RX_RING(cons)][RX_IDX(cons)];
+
+	prod_bd->rx_bd_haddr = cons_bd->rx_bd_haddr;
+}
+
+static inline u16 bnxt_find_next_agg_idx(struct bnxt_rx_ring_info *rxr, u16 idx)
+{
+	u16 next, max = rxr->rx_agg_bmap_size;
+
+	next = find_next_zero_bit(rxr->rx_agg_bmap, max, idx);
+	if (next >= max)
+		next = find_first_zero_bit(rxr->rx_agg_bmap, max);
+	return next;
+}
+
+static inline int bnxt_alloc_rx_page(struct bnxt *bp,
+				     struct bnxt_rx_ring_info *rxr,
+				     u16 prod, gfp_t gfp)
+{
+	struct rx_bd *rxbd =
+		&rxr->rx_agg_desc_ring[RX_RING(prod)][RX_IDX(prod)];
+	struct bnxt_sw_rx_agg_bd *rx_agg_buf;
+	struct pci_dev *pdev = bp->pdev;
+	struct page *page;
+	dma_addr_t mapping;
+	u16 sw_prod = rxr->rx_sw_agg_prod;
+
+	page = alloc_page(gfp);
+	if (!page)
+		return -ENOMEM;
+
+	mapping = dma_map_page(&pdev->dev, page, 0, PAGE_SIZE,
+			       PCI_DMA_FROMDEVICE);
+	if (dma_mapping_error(&pdev->dev, mapping)) {
+		__free_page(page);
+		return -EIO;
+	}
+
+	if (unlikely(test_bit(sw_prod, rxr->rx_agg_bmap)))
+		sw_prod = bnxt_find_next_agg_idx(rxr, sw_prod);
+
+	__set_bit(sw_prod, rxr->rx_agg_bmap);
+	rx_agg_buf = &rxr->rx_agg_ring[sw_prod];
+	rxr->rx_sw_agg_prod = NEXT_RX_AGG(sw_prod);
+
+	rx_agg_buf->page = page;
+	rx_agg_buf->mapping = mapping;
+	rxbd->rx_bd_haddr = cpu_to_le64(mapping);
+	rxbd->rx_bd_opaque = sw_prod;
+	return 0;
+}
+
+static void bnxt_reuse_rx_agg_bufs(struct bnxt_napi *bnapi, u16 cp_cons,
+				   u32 agg_bufs)
+{
+	struct bnxt *bp = bnapi->bp;
+	struct bnxt_cp_ring_info *cpr = &bnapi->cp_ring;
+	struct bnxt_rx_ring_info *rxr = &bnapi->rx_ring;
+	u16 prod = rxr->rx_agg_prod;
+	u16 sw_prod = rxr->rx_sw_agg_prod;
+	u32 i;
+
+	for (i = 0; i < agg_bufs; i++) {
+		u16 cons;
+		struct rx_agg_cmp *agg;
+		struct bnxt_sw_rx_agg_bd *cons_rx_buf, *prod_rx_buf;
+		struct rx_bd *prod_bd;
+		struct page *page;
+
+		agg = (struct rx_agg_cmp *)
+			&cpr->cp_desc_ring[CP_RING(cp_cons)][CP_IDX(cp_cons)];
+		cons = agg->rx_agg_cmp_opaque;
+		__clear_bit(cons, rxr->rx_agg_bmap);
+
+		if (unlikely(test_bit(sw_prod, rxr->rx_agg_bmap)))
+			sw_prod = bnxt_find_next_agg_idx(rxr, sw_prod);
+
+		__set_bit(sw_prod, rxr->rx_agg_bmap);
+		prod_rx_buf = &rxr->rx_agg_ring[sw_prod];
+		cons_rx_buf = &rxr->rx_agg_ring[cons];
+
+		/* It is possible for sw_prod to be equal to cons, so
+		 * set cons_rx_buf->page to NULL first.
+		 */
+		page = cons_rx_buf->page;
+		cons_rx_buf->page = NULL;
+		prod_rx_buf->page = page;
+
+		prod_rx_buf->mapping = cons_rx_buf->mapping;
+
+		prod_bd = &rxr->rx_agg_desc_ring[RX_RING(prod)][RX_IDX(prod)];
+
+		prod_bd->rx_bd_haddr = cpu_to_le64(cons_rx_buf->mapping);
+		prod_bd->rx_bd_opaque = sw_prod;
+
+		prod = NEXT_RX_AGG(prod);
+		sw_prod = NEXT_RX_AGG(sw_prod);
+		cp_cons = NEXT_CMP(cp_cons);
+	}
+	rxr->rx_agg_prod = prod;
+	rxr->rx_sw_agg_prod = sw_prod;
+}
+
+static struct sk_buff *bnxt_rx_skb(struct bnxt *bp,
+				   struct bnxt_rx_ring_info *rxr, u16 cons,
+				   u16 prod, u8 *data, dma_addr_t dma_addr,
+				   unsigned int len)
+{
+	int err;
+	struct sk_buff *skb;
+
+	err = bnxt_alloc_rx_data(bp, rxr, prod, GFP_ATOMIC);
+	if (unlikely(err)) {
+		bnxt_reuse_rx_data(rxr, cons, data);
+		return NULL;
+	}
+
+	skb = build_skb(data, 0);
+	dma_unmap_single(&bp->pdev->dev, dma_addr, bp->rx_buf_use_size,
+			 PCI_DMA_FROMDEVICE);
+	if (!skb) {
+		kfree(data);
+		return NULL;
+	}
+
+	skb_reserve(skb, BNXT_RX_OFFSET);
+	skb_put(skb, len);
+	return skb;
+}
+
+static struct sk_buff *bnxt_rx_pages(struct bnxt *bp, struct bnxt_napi *bnapi,
+				     struct sk_buff *skb, u16 cp_cons,
+				     u32 agg_bufs)
+{
+	struct pci_dev *pdev = bp->pdev;
+	struct bnxt_cp_ring_info *cpr = &bnapi->cp_ring;
+	struct bnxt_rx_ring_info *rxr = &bnapi->rx_ring;
+	u16 prod = rxr->rx_agg_prod;
+	u32 i;
+
+	for (i = 0; i < agg_bufs; i++) {
+		u16 cons, frag_len;
+		struct rx_agg_cmp *agg;
+		struct bnxt_sw_rx_agg_bd *cons_rx_buf;
+		struct page *page;
+		dma_addr_t mapping;
+
+		agg = (struct rx_agg_cmp *)
+			&cpr->cp_desc_ring[CP_RING(cp_cons)][CP_IDX(cp_cons)];
+		cons = agg->rx_agg_cmp_opaque;
+		frag_len = (le32_to_cpu(agg->rx_agg_cmp_len_flags_type) &
+			    RX_AGG_CMP_LEN) >> RX_AGG_CMP_LEN_SHIFT;
+
+		cons_rx_buf = &rxr->rx_agg_ring[cons];
+		skb_fill_page_desc(skb, i, cons_rx_buf->page, 0, frag_len);
+		__clear_bit(cons, rxr->rx_agg_bmap);
+
+		/* It is possible for bnxt_alloc_rx_page() to allocate
+		 * a sw_prod index that equals the cons index, so we
+		 * need to clear the cons entry now.
+		 */
+		mapping = dma_unmap_addr(cons_rx_buf, mapping);
+		page = cons_rx_buf->page;
+		cons_rx_buf->page = NULL;
+
+		if (bnxt_alloc_rx_page(bp, rxr, prod, GFP_ATOMIC) != 0) {
+			struct skb_shared_info *shinfo;
+			unsigned int nr_frags;
+
+			shinfo = skb_shinfo(skb);
+			nr_frags = --shinfo->nr_frags;
+			__skb_frag_set_page(&shinfo->frags[nr_frags], NULL);
+
+			dev_kfree_skb(skb);
+
+			cons_rx_buf->page = page;
+
+			/* Update prod since possibly some pages have been
+			 * allocated already.
+			 */
+			rxr->rx_agg_prod = prod;
+			bnxt_reuse_rx_agg_bufs(bnapi, cp_cons, agg_bufs - i);
+			return NULL;
+		}
+
+		dma_unmap_page(&pdev->dev, mapping, PAGE_SIZE,
+			       PCI_DMA_FROMDEVICE);
+
+		skb->data_len += frag_len;
+		skb->len += frag_len;
+		skb->truesize += PAGE_SIZE;
+
+		prod = NEXT_RX_AGG(prod);
+		cp_cons = NEXT_CMP(cp_cons);
+	}
+	rxr->rx_agg_prod = prod;
+	return skb;
+}
+
+static int bnxt_agg_bufs_valid(struct bnxt *bp, struct bnxt_cp_ring_info *cpr,
+			       u8 agg_bufs, u32 *raw_cons)
+{
+	u16 last;
+	struct rx_agg_cmp *agg;
+
+	*raw_cons = ADV_RAW_CMP(*raw_cons, agg_bufs);
+	last = RING_CMP(*raw_cons);
+	agg = (struct rx_agg_cmp *)
+		&cpr->cp_desc_ring[CP_RING(last)][CP_IDX(last)];
+	return RX_AGG_CMP_VALID(agg, *raw_cons);
+}
+
+static inline struct sk_buff *bnxt_copy_skb(struct bnxt_napi *bnapi, u8 *data,
+					    unsigned int len,
+					    dma_addr_t mapping)
+{
+	struct bnxt *bp = bnapi->bp;
+	struct pci_dev *pdev = bp->pdev;
+	struct sk_buff *skb;
+
+	skb = napi_alloc_skb(&bnapi->napi, len);
+	if (!skb)
+		return NULL;
+
+	dma_sync_single_for_cpu(&pdev->dev, mapping,
+				bp->rx_copy_thresh, PCI_DMA_FROMDEVICE);
+
+	memcpy(skb->data - BNXT_RX_OFFSET, data, len + BNXT_RX_OFFSET);
+
+	dma_sync_single_for_device(&pdev->dev, mapping,
+				   bp->rx_copy_thresh,
+				   PCI_DMA_FROMDEVICE);
+
+	skb_put(skb, len);
+	return skb;
+}
+
+static void bnxt_tpa_start(struct bnxt *bp, struct bnxt_rx_ring_info *rxr,
+			   struct rx_tpa_start_cmp *tpa_start,
+			   struct rx_tpa_start_cmp_ext *tpa_start1)
+{
+	u8 agg_id = TPA_START_AGG_ID(tpa_start);
+	u16 cons, prod;
+	struct bnxt_tpa_info *tpa_info;
+	struct bnxt_sw_rx_bd *cons_rx_buf, *prod_rx_buf;
+	struct rx_bd *prod_bd;
+	dma_addr_t mapping;
+
+	cons = tpa_start->rx_tpa_start_cmp_opaque;
+	prod = rxr->rx_prod;
+	cons_rx_buf = &rxr->rx_buf_ring[cons];
+	prod_rx_buf = &rxr->rx_buf_ring[prod];
+	tpa_info = &rxr->rx_tpa[agg_id];
+
+	prod_rx_buf->data = tpa_info->data;
+
+	mapping = tpa_info->mapping;
+	dma_unmap_addr_set(prod_rx_buf, mapping, mapping);
+
+	prod_bd = &rxr->rx_desc_ring[RX_RING(prod)][RX_IDX(prod)];
+
+	prod_bd->rx_bd_haddr = cpu_to_le64(mapping);
+
+	tpa_info->data = cons_rx_buf->data;
+	cons_rx_buf->data = NULL;
+	tpa_info->mapping = dma_unmap_addr(cons_rx_buf, mapping);
+
+	tpa_info->len =
+		le32_to_cpu(tpa_start->rx_tpa_start_cmp_len_flags_type) >>
+				RX_TPA_START_CMP_LEN_SHIFT;
+	if (likely(TPA_START_HASH_VALID(tpa_start))) {
+		u32 hash_type = TPA_START_HASH_TYPE(tpa_start);
+
+		tpa_info->hash_type = PKT_HASH_TYPE_L4;
+		tpa_info->gso_type = SKB_GSO_TCPV4;
+		/* RSS profiles 1 and 3 with extract code 0 for inner 4-tuple */
+		if (hash_type == 3)
+			tpa_info->gso_type = SKB_GSO_TCPV6;
+		tpa_info->rss_hash =
+			le32_to_cpu(tpa_start->rx_tpa_start_cmp_rss_hash);
+	} else {
+		tpa_info->hash_type = PKT_HASH_TYPE_NONE;
+		tpa_info->gso_type = 0;
+		if (netif_msg_rx_err(bp))
+			netdev_warn(bp->dev, "TPA packet without valid hash\n");
+	}
+	tpa_info->flags2 = le32_to_cpu(tpa_start1->rx_tpa_start_cmp_flags2);
+	tpa_info->metadata = le32_to_cpu(tpa_start1->rx_tpa_start_cmp_metadata);
+
+	rxr->rx_prod = NEXT_RX(prod);
+	cons = NEXT_RX(cons);
+	cons_rx_buf = &rxr->rx_buf_ring[cons];
+
+	bnxt_reuse_rx_data(rxr, cons, cons_rx_buf->data);
+	rxr->rx_prod = NEXT_RX(rxr->rx_prod);
+	cons_rx_buf->data = NULL;
+}
+
+static void bnxt_abort_tpa(struct bnxt *bp, struct bnxt_napi *bnapi,
+			   u16 cp_cons, u32 agg_bufs)
+{
+	if (agg_bufs)
+		bnxt_reuse_rx_agg_bufs(bnapi, cp_cons, agg_bufs);
+}
+
+#define BNXT_IPV4_HDR_SIZE	(sizeof(struct iphdr) + sizeof(struct tcphdr))
+#define BNXT_IPV6_HDR_SIZE	(sizeof(struct ipv6hdr) + sizeof(struct tcphdr))
+
+static inline struct sk_buff *bnxt_gro_skb(struct bnxt_tpa_info *tpa_info,
+					   struct rx_tpa_end_cmp *tpa_end,
+					   struct rx_tpa_end_cmp_ext *tpa_end1,
+					   struct sk_buff *skb)
+{
+	struct tcphdr *th;
+	int payload_off, tcp_opt_len = 0;
+	int len, nw_off;
+
+	NAPI_GRO_CB(skb)->count = TPA_END_TPA_SEGS(tpa_end);
+	skb_shinfo(skb)->gso_size =
+		le32_to_cpu(tpa_end1->rx_tpa_end_cmp_seg_len);
+	skb_shinfo(skb)->gso_type = tpa_info->gso_type;
+	payload_off = (le32_to_cpu(tpa_end->rx_tpa_end_cmp_misc_v1) &
+		       RX_TPA_END_CMP_PAYLOAD_OFFSET) >>
+		      RX_TPA_END_CMP_PAYLOAD_OFFSET_SHIFT;
+	if (TPA_END_GRO_TS(tpa_end))
+		tcp_opt_len = 12;
+
+#ifdef CONFIG_INET
+	if (tpa_info->gso_type == SKB_GSO_TCPV4) {
+		struct iphdr *iph;
+
+		nw_off = payload_off - BNXT_IPV4_HDR_SIZE - tcp_opt_len -
+			 ETH_HLEN;
+		skb_set_network_header(skb, nw_off);
+		iph = ip_hdr(skb);
+		skb_set_transport_header(skb, nw_off + sizeof(struct iphdr));
+		len = skb->len - skb_transport_offset(skb);
+		th = tcp_hdr(skb);
+		th->check = ~tcp_v4_check(len, iph->saddr, iph->daddr, 0);
+	} else if (tpa_info->gso_type == SKB_GSO_TCPV6) {
+		struct ipv6hdr *iph;
+
+		nw_off = payload_off - BNXT_IPV6_HDR_SIZE - tcp_opt_len -
+			 ETH_HLEN;
+		skb_set_network_header(skb, nw_off);
+		iph = ipv6_hdr(skb);
+		skb_set_transport_header(skb, nw_off + sizeof(struct ipv6hdr));
+		len = skb->len - skb_transport_offset(skb);
+		th = tcp_hdr(skb);
+		th->check = ~tcp_v6_check(len, &iph->saddr, &iph->daddr, 0);
+	} else {
+		dev_kfree_skb_any(skb);
+		return NULL;
+	}
+	tcp_gro_complete(skb);
+
+	if (nw_off) { /* tunnel */
+		struct udphdr *uh = NULL;
+
+		if (skb->protocol == htons(ETH_P_IP)) {
+			struct iphdr *iph = (struct iphdr *)skb->data;
+
+			if (iph->protocol == IPPROTO_UDP)
+				uh = (struct udphdr *)(iph + 1);
+		} else {
+			struct ipv6hdr *iph = (struct ipv6hdr *)skb->data;
+
+			if (iph->nexthdr == IPPROTO_UDP)
+				uh = (struct udphdr *)(iph + 1);
+		}
+		if (uh) {
+			if (uh->check)
+				skb_shinfo(skb)->gso_type |=
+					SKB_GSO_UDP_TUNNEL_CSUM;
+			else
+				skb_shinfo(skb)->gso_type |= SKB_GSO_UDP_TUNNEL;
+		}
+	}
+#endif
+	return skb;
+}
+
+static inline struct sk_buff *bnxt_tpa_end(struct bnxt *bp,
+					   struct bnxt_napi *bnapi,
+					   u32 *raw_cons,
+					   struct rx_tpa_end_cmp *tpa_end,
+					   struct rx_tpa_end_cmp_ext *tpa_end1,
+					   bool *agg_event)
+{
+	struct bnxt_cp_ring_info *cpr = &bnapi->cp_ring;
+	struct bnxt_rx_ring_info *rxr = &bnapi->rx_ring;
+	u8 agg_id = TPA_END_AGG_ID(tpa_end);
+	u8 *data, agg_bufs;
+	u16 cp_cons = RING_CMP(*raw_cons);
+	unsigned int len;
+	struct bnxt_tpa_info *tpa_info;
+	dma_addr_t mapping;
+	struct sk_buff *skb;
+
+	tpa_info = &rxr->rx_tpa[agg_id];
+	data = tpa_info->data;
+	prefetch(data);
+	len = tpa_info->len;
+	mapping = tpa_info->mapping;
+
+	agg_bufs = (le32_to_cpu(tpa_end->rx_tpa_end_cmp_misc_v1) &
+		    RX_TPA_END_CMP_AGG_BUFS) >> RX_TPA_END_CMP_AGG_BUFS_SHIFT;
+
+	if (agg_bufs) {
+		if (!bnxt_agg_bufs_valid(bp, cpr, agg_bufs, raw_cons))
+			return ERR_PTR(-EBUSY);
+
+		*agg_event = true;
+		cp_cons = NEXT_CMP(cp_cons);
+	}
+
+	if (unlikely(agg_bufs > MAX_SKB_FRAGS)) {
+		bnxt_abort_tpa(bp, bnapi, cp_cons, agg_bufs);
+		netdev_warn(bp->dev, "TPA frags %d exceeded MAX_SKB_FRAGS %d\n",
+			    agg_bufs, (int)MAX_SKB_FRAGS);
+		return NULL;
+	}
+
+	if (len <= bp->rx_copy_thresh) {
+		skb = bnxt_copy_skb(bnapi, data, len, mapping);
+		if (!skb) {
+			bnxt_abort_tpa(bp, bnapi, cp_cons, agg_bufs);
+			return NULL;
+		}
+	} else {
+		u8 *new_data;
+		dma_addr_t new_mapping;
+
+		new_data = __bnxt_alloc_rx_data(bp, &new_mapping, GFP_ATOMIC)