use FW 7.2.16

The patch integrates FW 7.2.16 HSI and implements driver
part of GRO flow.

FW 7.2.16 adds the ability to aggregate packets for GRO
(and not just LRO) and also fixes some bugs.

1. Added new aggregation mode: GRO. In this mode packets are aggregated
   such that the original packets can be reconstructed by the OS.
2. 57712 HW bug workaround - initialized all CAM TM registers to 0x32.
3. Adding the FCoE statistics structures to the BNX2X HSI.
4. Wrong configuration of TX HW input buffer size may cause theoretical
   performance effect. Performed configuration fix.
5. FCOE - Arrival of packets beyond task IO size can lead to crash.
   Fix firmware data-in flow.
6. iSCSI - In rare cases of on-chip termination the graceful termination
   timer hangs, and the termination doesn't complete. Firmware fix to MSL
   timer tolerance.
7. iSCSI - Chip hangs when target sends FIN out-of-order or with isles
   open at the initiator side. Firmware implementation corrected to drop
   FIN received out-of-order or with isles still open.
8. iSCSI - Chip hangs when in case of retransmission not aligned to 4-bytes
   from the beginning of iSCSI PDU. Firmware implementation corrected
   to support arbitrary aligned retransmissions.
9. iSCSI - Arrival of target-initiated NOP-IN during intense ISCSI traffic
   might lead to crash. Firmware fix to relevant flow.

Signed-off-by: Dmitry Kravkov <dmitry@broadcom.com>
Signed-off-by: Eilon Greenstein <eilong@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

drivers/net/ethernet/broadcom/bnx2x/bnx2x.h

@@ -445,6 +445,8 @@ struct bnx2x_agg_info {
 	u16			vlan_tag;
 	u16			len_on_bd;
 	u32			rxhash;
+	u16			gro_size;
+	u16			full_page;
 };
 
 #define Q_STATS_OFFSET32(stat_name) \
@@ -473,6 +475,11 @@ struct bnx2x_fp_txdata {
 	int			txq_index;
 };
 
+enum bnx2x_tpa_mode_t {
+	TPA_MODE_LRO,
+	TPA_MODE_GRO
+};
+
 struct bnx2x_fastpath {
 	struct bnx2x		*bp; /* parent */
 
@@ -489,6 +496,8 @@ struct bnx2x_fastpath {
 
 	dma_addr_t		status_blk_mapping;
 
+	enum bnx2x_tpa_mode_t	mode;
+
 	u8			max_cos; /* actual number of active tx coses */
 	struct bnx2x_fp_txdata	txdata[BNX2X_MULTI_TX_COS];
 
@@ -1199,6 +1208,8 @@ struct bnx2x {
 #define ETH_MIN_PACKET_SIZE		60
 #define ETH_MAX_PACKET_SIZE		1500
 #define ETH_MAX_JUMBO_PACKET_SIZE	9600
+/* TCP with Timestamp Option (32) + IPv6 (40) */
+#define ETH_MAX_TPA_HEADER_SIZE		72
 
 	/* Max supported alignment is 256 (8 shift) */
 #define BNX2X_RX_ALIGN_SHIFT		min(8, L1_CACHE_SHIFT)
@@ -1269,6 +1280,7 @@ struct bnx2x {
 #define NO_MCP_FLAG			(1 << 9)
 
 #define BP_NOMCP(bp)			(bp->flags & NO_MCP_FLAG)
+#define GRO_ENABLE_FLAG			(1 << 10)
 #define MF_FUNC_DIS			(1 << 11)
 #define OWN_CNIC_IRQ			(1 << 12)
 #define NO_ISCSI_OOO_FLAG		(1 << 13)

drivers/net/ethernet/broadcom/bnx2x/bnx2x_cmn.c

@@ -209,13 +209,11 @@ static inline void bnx2x_update_last_max_sge(struct bnx2x_fastpath *fp,
 		fp->last_max_sge = idx;
 }
 
-static void bnx2x_update_sge_prod(struct bnx2x_fastpath *fp,
-				  struct eth_fast_path_rx_cqe *fp_cqe)
+static inline void bnx2x_update_sge_prod(struct bnx2x_fastpath *fp,
+					 u16 sge_len,
+					 struct eth_end_agg_rx_cqe *cqe)
 {
 	struct bnx2x *bp = fp->bp;
-	u16 sge_len = SGE_PAGE_ALIGN(le16_to_cpu(fp_cqe->pkt_len) -
-				     le16_to_cpu(fp_cqe->len_on_bd)) >>
-		      SGE_PAGE_SHIFT;
 	u16 last_max, last_elem, first_elem;
 	u16 delta = 0;
 	u16 i;
@@ -226,15 +224,15 @@ static void bnx2x_update_sge_prod(struct bnx2x_fastpath *fp,
 	/* First mark all used pages */
 	for (i = 0; i < sge_len; i++)
 		BIT_VEC64_CLEAR_BIT(fp->sge_mask,
-			RX_SGE(le16_to_cpu(fp_cqe->sgl_or_raw_data.sgl[i])));
+			RX_SGE(le16_to_cpu(cqe->sgl_or_raw_data.sgl[i])));
 
 	DP(NETIF_MSG_RX_STATUS, "fp_cqe->sgl[%d] = %d\n",
-	   sge_len - 1, le16_to_cpu(fp_cqe->sgl_or_raw_data.sgl[sge_len - 1]));
+	   sge_len - 1, le16_to_cpu(cqe->sgl_or_raw_data.sgl[sge_len - 1]));
 
 	/* Here we assume that the last SGE index is the biggest */
 	prefetch((void *)(fp->sge_mask));
 	bnx2x_update_last_max_sge(fp,
-		le16_to_cpu(fp_cqe->sgl_or_raw_data.sgl[sge_len - 1]));
+		le16_to_cpu(cqe->sgl_or_raw_data.sgl[sge_len - 1]));
 
 	last_max = RX_SGE(fp->last_max_sge);
 	last_elem = last_max >> BIT_VEC64_ELEM_SHIFT;
@@ -328,6 +326,12 @@ static void bnx2x_tpa_start(struct bnx2x_fastpath *fp, u16 queue,
 	tpa_info->len_on_bd = le16_to_cpu(cqe->len_on_bd);
 	tpa_info->placement_offset = cqe->placement_offset;
 	tpa_info->rxhash = bnx2x_get_rxhash(bp, cqe);
+	if (fp->mode == TPA_MODE_GRO) {
+		u16 gro_size = le16_to_cpu(cqe->pkt_len_or_gro_seg_len);
+		tpa_info->full_page =
+			SGE_PAGE_SIZE * PAGES_PER_SGE / gro_size * gro_size;
+		tpa_info->gro_size = gro_size;
+	}
 
 #ifdef BNX2X_STOP_ON_ERROR
 	fp->tpa_queue_used |= (1 << queue);
@@ -384,25 +388,40 @@ static inline u16 bnx2x_set_lro_mss(struct bnx2x *bp, u16 parsing_flags,
 }
 
 static int bnx2x_fill_frag_skb(struct bnx2x *bp, struct bnx2x_fastpath *fp,
-			       u16 queue, struct sk_buff *skb,
+			       struct bnx2x_agg_info *tpa_info,
+			       u16 pages,
+			       struct sk_buff *skb,
 			       struct eth_end_agg_rx_cqe *cqe,
 			       u16 cqe_idx)
 {
 	struct sw_rx_page *rx_pg, old_rx_pg;
-	u32 i, frag_len, frag_size, pages;
-	int err;
-	int j;
-	struct bnx2x_agg_info *tpa_info = &fp->tpa_info[queue];
+	u32 i, frag_len, frag_size;
+	int err, j, frag_id = 0;
 	u16 len_on_bd = tpa_info->len_on_bd;
+	u16 full_page = 0, gro_size = 0;
 
 	frag_size = le16_to_cpu(cqe->pkt_len) - len_on_bd;
-	pages = SGE_PAGE_ALIGN(frag_size) >> SGE_PAGE_SHIFT;
+
+	if (fp->mode == TPA_MODE_GRO) {
+		gro_size = tpa_info->gro_size;
+		full_page = tpa_info->full_page;
+	}
 
 	/* This is needed in order to enable forwarding support */
-	if (frag_size)
+	if (frag_size) {
 		skb_shinfo(skb)->gso_size = bnx2x_set_lro_mss(bp,
 					tpa_info->parsing_flags, len_on_bd);
 
+		/* set for GRO */
+		if (fp->mode == TPA_MODE_GRO)
+			skb_shinfo(skb)->gso_type =
+			    (GET_FLAG(tpa_info->parsing_flags,
+				      PARSING_FLAGS_OVER_ETHERNET_PROTOCOL) ==
+						PRS_FLAG_OVERETH_IPV6) ?
+				SKB_GSO_TCPV6 : SKB_GSO_TCPV4;
+	}
+
+
 #ifdef BNX2X_STOP_ON_ERROR
 	if (pages > min_t(u32, 8, MAX_SKB_FRAGS)*SGE_PAGE_SIZE*PAGES_PER_SGE) {
 		BNX2X_ERR("SGL length is too long: %d. CQE index is %d\n",
@@ -419,7 +438,12 @@ static int bnx2x_fill_frag_skb(struct bnx2x *bp, struct bnx2x_fastpath *fp,
 
 		/* FW gives the indices of the SGE as if the ring is an array
 		   (meaning that "next" element will consume 2 indices) */
-		frag_len = min(frag_size, (u32)(SGE_PAGE_SIZE*PAGES_PER_SGE));
+		if (fp->mode == TPA_MODE_GRO)
+			frag_len = min_t(u32, frag_size, (u32)full_page);
+		else /* LRO */
+			frag_len = min_t(u32, frag_size,
+					 (u32)(SGE_PAGE_SIZE * PAGES_PER_SGE));
+
 		rx_pg = &fp->rx_page_ring[sge_idx];
 		old_rx_pg = *rx_pg;
 
@@ -435,9 +459,21 @@ static int bnx2x_fill_frag_skb(struct bnx2x *bp, struct bnx2x_fastpath *fp,
 		dma_unmap_page(&bp->pdev->dev,
 			       dma_unmap_addr(&old_rx_pg, mapping),
 			       SGE_PAGE_SIZE*PAGES_PER_SGE, DMA_FROM_DEVICE);
-
 		/* Add one frag and update the appropriate fields in the skb */
-		skb_fill_page_desc(skb, j, old_rx_pg.page, 0, frag_len);
+		if (fp->mode == TPA_MODE_LRO)
+			skb_fill_page_desc(skb, j, old_rx_pg.page, 0, frag_len);
+		else { /* GRO */
+			int rem;
+			int offset = 0;
+			for (rem = frag_len; rem > 0; rem -= gro_size) {
+				int len = rem > gro_size ? gro_size : rem;
+				skb_fill_page_desc(skb, frag_id++,
+						   old_rx_pg.page, offset, len);
+				if (offset)
+					get_page(old_rx_pg.page);
+				offset += len;
+			}
+		}
 
 		skb->data_len += frag_len;
 		skb->truesize += SGE_PAGE_SIZE * PAGES_PER_SGE;
@@ -449,18 +485,17 @@ static int bnx2x_fill_frag_skb(struct bnx2x *bp, struct bnx2x_fastpath *fp,
 	return 0;
 }
 
-static void bnx2x_tpa_stop(struct bnx2x *bp, struct bnx2x_fastpath *fp,
-			   u16 queue, struct eth_end_agg_rx_cqe *cqe,
-			   u16 cqe_idx)
+static inline void bnx2x_tpa_stop(struct bnx2x *bp, struct bnx2x_fastpath *fp,
+				  struct bnx2x_agg_info *tpa_info,
+				  u16 pages,
+				  struct eth_end_agg_rx_cqe *cqe,
+				  u16 cqe_idx)
 {
-	struct bnx2x_agg_info *tpa_info = &fp->tpa_info[queue];
 	struct sw_rx_bd *rx_buf = &tpa_info->first_buf;
-	u32 pad = tpa_info->placement_offset;
+	u8 pad = tpa_info->placement_offset;
 	u16 len = tpa_info->len_on_bd;
 	struct sk_buff *skb = NULL;
-	u8 *data = rx_buf->data;
-	/* alloc new skb */
-	u8 *new_data;
+	u8 *new_data, *data = rx_buf->data;
 	u8 old_tpa_state = tpa_info->tpa_state;
 
 	tpa_info->tpa_state = BNX2X_TPA_STOP;
@@ -500,7 +535,8 @@ static void bnx2x_tpa_stop(struct bnx2x *bp, struct bnx2x_fastpath *fp,
 		skb->protocol = eth_type_trans(skb, bp->dev);
 		skb->ip_summed = CHECKSUM_UNNECESSARY;
 
-		if (!bnx2x_fill_frag_skb(bp, fp, queue, skb, cqe, cqe_idx)) {
+		if (!bnx2x_fill_frag_skb(bp, fp, tpa_info, pages,
+					 skb, cqe, cqe_idx)) {
 			if (tpa_info->parsing_flags & PARSING_FLAGS_VLAN)
 				__vlan_hwaccel_put_tag(skb, tpa_info->vlan_tag);
 			napi_gro_receive(&fp->napi, skb);
@@ -565,7 +601,7 @@ int bnx2x_rx_int(struct bnx2x_fastpath *fp, int budget)
 		struct eth_fast_path_rx_cqe *cqe_fp;
 		u8 cqe_fp_flags;
 		enum eth_rx_cqe_type cqe_fp_type;
-		u16 len, pad;
+		u16 len, pad, queue;
 		u8 *data;
 
 #ifdef BNX2X_STOP_ON_ERROR
@@ -586,17 +622,21 @@ int bnx2x_rx_int(struct bnx2x_fastpath *fp, int budget)
 		   "  queue %x  vlan %x  len %u\n", CQE_TYPE(cqe_fp_flags),
 		   cqe_fp_flags, cqe_fp->status_flags,
 		   le32_to_cpu(cqe_fp->rss_hash_result),
-		   le16_to_cpu(cqe_fp->vlan_tag), le16_to_cpu(cqe_fp->pkt_len));
+		   le16_to_cpu(cqe_fp->vlan_tag),
+		   le16_to_cpu(cqe_fp->pkt_len_or_gro_seg_len));
 
 		/* is this a slowpath msg? */
 		if (unlikely(CQE_TYPE_SLOW(cqe_fp_type))) {
 			bnx2x_sp_event(fp, cqe);
 			goto next_cqe;
 		}
+
 		rx_buf = &fp->rx_buf_ring[bd_cons];
 		data = rx_buf->data;
 
 		if (!CQE_TYPE_FAST(cqe_fp_type)) {
+			struct bnx2x_agg_info *tpa_info;
+			u16 frag_size, pages;
 #ifdef BNX2X_STOP_ON_ERROR
 			/* sanity check */
 			if (fp->disable_tpa &&
@@ -616,28 +656,38 @@ int bnx2x_rx_int(struct bnx2x_fastpath *fp, int budget)
 				bnx2x_tpa_start(fp, queue,
 						bd_cons, bd_prod,
 						cqe_fp);
+
 				goto next_rx;
-			} else {
-				u16 queue =
-					cqe->end_agg_cqe.queue_index;
-				DP(NETIF_MSG_RX_STATUS,
-				   "calling tpa_stop on queue %d\n",
-				   queue);
 
-				bnx2x_tpa_stop(bp, fp, queue,
-					       &cqe->end_agg_cqe,
-					       comp_ring_cons);
+			}
+			queue = cqe->end_agg_cqe.queue_index;
+			tpa_info = &fp->tpa_info[queue];
+			DP(NETIF_MSG_RX_STATUS,
+			   "calling tpa_stop on queue %d\n",
+			   queue);
+
+			frag_size = le16_to_cpu(cqe->end_agg_cqe.pkt_len) -
+				    tpa_info->len_on_bd;
+
+			if (fp->mode == TPA_MODE_GRO)
+				pages = (frag_size + tpa_info->full_page - 1) /
+					 tpa_info->full_page;
+			else
+				pages = SGE_PAGE_ALIGN(frag_size) >>
+					SGE_PAGE_SHIFT;
+
+			bnx2x_tpa_stop(bp, fp, tpa_info, pages,
+				       &cqe->end_agg_cqe, comp_ring_cons);
 #ifdef BNX2X_STOP_ON_ERROR
-				if (bp->panic)
-					return 0;
+			if (bp->panic)
+				return 0;
 #endif
 
-				bnx2x_update_sge_prod(fp, cqe_fp);
-				goto next_cqe;
-			}
+			bnx2x_update_sge_prod(fp, pages, &cqe->end_agg_cqe);
+			goto next_cqe;
 		}
 		/* non TPA */
-		len = le16_to_cpu(cqe_fp->pkt_len);
+		len = le16_to_cpu(cqe_fp->pkt_len_or_gro_seg_len);
 		pad = cqe_fp->placement_offset;
 		dma_sync_single_for_cpu(&bp->pdev->dev,
 					dma_unmap_addr(rx_buf, mapping),
@@ -3440,13 +3490,15 @@ int bnx2x_change_mtu(struct net_device *dev, int new_mtu)
 }
 
 netdev_features_t bnx2x_fix_features(struct net_device *dev,
-	netdev_features_t features)
+				     netdev_features_t features)
 {
 	struct bnx2x *bp = netdev_priv(dev);
 
 	/* TPA requires Rx CSUM offloading */
-	if (!(features & NETIF_F_RXCSUM) || bp->disable_tpa)
+	if (!(features & NETIF_F_RXCSUM) || bp->disable_tpa) {
 		features &= ~NETIF_F_LRO;
+		features &= ~NETIF_F_GRO;
+	}
 
 	return features;
 }
@@ -3462,6 +3514,11 @@ int bnx2x_set_features(struct net_device *dev, netdev_features_t features)
 	else
 		flags &= ~TPA_ENABLE_FLAG;
 
+	if (features & NETIF_F_GRO)
+		flags |= GRO_ENABLE_FLAG;
+	else
+		flags &= ~GRO_ENABLE_FLAG;
+
 	if (features & NETIF_F_LOOPBACK) {
 		if (bp->link_params.loopback_mode != LOOPBACK_BMAC) {
 			bp->link_params.loopback_mode = LOOPBACK_BMAC;

drivers/net/ethernet/broadcom/bnx2x/bnx2x_cmn.h

@@ -534,8 +534,9 @@ int bnx2x_change_mtu(struct net_device *dev, int new_mtu);
  */
 int bnx2x_fcoe_get_wwn(struct net_device *dev, u64 *wwn, int type);
 #endif
+
 netdev_features_t bnx2x_fix_features(struct net_device *dev,
-	netdev_features_t features);
+				     netdev_features_t features);
 int bnx2x_set_features(struct net_device *dev, netdev_features_t features);
 
 /**
@@ -1491,6 +1492,18 @@ static inline u16 bnx2x_extract_max_cfg(struct bnx2x *bp, u32 mf_cfg)
 	return max_cfg;
 }
 
+/* checks if HW supports GRO for given MTU */
+static inline bool bnx2x_mtu_allows_gro(int mtu)
+{
+	/* gro frags per page */
+	int fpp = SGE_PAGE_SIZE / (mtu - ETH_MAX_TPA_HEADER_SIZE);
+
+	/*
+	 * 1. number of frags should not grow above MAX_SKB_FRAGS
+	 * 2. frag must fit the page
+	 */
+	return mtu <= SGE_PAGE_SIZE && (U_ETH_SGL_SIZE * fpp) <= MAX_SKB_FRAGS;
+}
 /**
  * bnx2x_bz_fp - zero content of the fastpath structure.
  *
@@ -1556,7 +1569,14 @@ static inline void bnx2x_bz_fp(struct bnx2x *bp, int index)
 	 * set the tpa flag for each queue. The tpa flag determines the queue
 	 * minimal size so it must be set prior to queue memory allocation
 	 */
-	fp->disable_tpa = (bp->flags & TPA_ENABLE_FLAG) == 0;
+	fp->disable_tpa = !(bp->flags & TPA_ENABLE_FLAG ||
+				  (bp->flags & GRO_ENABLE_FLAG &&
+				   bnx2x_mtu_allows_gro(bp->dev->mtu)));
+	if (bp->flags & TPA_ENABLE_FLAG)
+		fp->mode = TPA_MODE_LRO;
+	else if (bp->flags & GRO_ENABLE_FLAG)
+		fp->mode = TPA_MODE_GRO;
+
 #ifdef BCM_CNIC
 	/* We don't want TPA on an FCoE L2 ring */
 	if (IS_FCOE_FP(fp))

drivers/net/ethernet/broadcom/bnx2x/bnx2x_ethtool.c

@@ -1931,7 +1931,7 @@ static int bnx2x_run_loopback(struct bnx2x *bp, int loopback_mode)
 	if (!CQE_TYPE_FAST(cqe_fp_type) || (cqe_fp_flags & ETH_RX_ERROR_FALGS))
 		goto test_loopback_rx_exit;
 
-	len = le16_to_cpu(cqe->fast_path_cqe.pkt_len);
+	len = le16_to_cpu(cqe->fast_path_cqe.pkt_len_or_gro_seg_len);
 	if (len != pkt_size)
 		goto test_loopback_rx_exit;
 

drivers/net/ethernet/broadcom/bnx2x/bnx2x_hsi.h

@@ -34,9 +34,10 @@ struct license_key {
 };
 
 
-#define PORT_0              0
-#define PORT_1              1
-#define PORT_MAX            2
+#define PORT_0			0
+#define PORT_1			1
+#define PORT_MAX		2
+#define NVM_PATH_MAX		2
 
 /****************************************************************************
  * Shared HW configuration                                                  *
@@ -618,12 +619,6 @@ struct port_hw_cfg {		    /* port 0: 0x12c  port 1: 0x2bc */
 	#define PORT_HW_CFG_ENABLE_CMS_DISABLED                      0x00000000
 	#define PORT_HW_CFG_ENABLE_CMS_ENABLED                       0x00200000
 
-	/*  Enable RJ45 magjack pair swapping on 10GBase-T PHY, 84833 only */
-	#define PORT_HW_CFG_RJ45_PR_SWP_MASK                0x00400000
-	#define PORT_HW_CFG_RJ45_PR_SWP_SHIFT			     22
-	#define PORT_HW_CFG_RJ45_PR_SWP_DISABLED		     0x00000000
-	#define PORT_HW_CFG_RJ45_PR_SWP_ENABLED			     0x00400000
-
 	/*  Determine the Serdes electrical interface   */
 	#define PORT_HW_CFG_NET_SERDES_IF_MASK              0x0F000000
 	#define PORT_HW_CFG_NET_SERDES_IF_SHIFT                      24
@@ -898,11 +893,6 @@ struct port_feat_cfg {		    /* port 0: 0x454  port 1: 0x4c8 */
 		#define PORT_FEAT_CFG_DCBX_DISABLED                  0x00000000
 		#define PORT_FEAT_CFG_DCBX_ENABLED                   0x00000100
 
-	#define PORT_FEAT_CFG_AUTOGREEN_MASK                0x00000200
-	#define PORT_FEAT_CFG_AUTOGREEN_SHIFT                        9
-	#define PORT_FEAT_CFG_AUTOGREEN_DISABLED                     0x00000000
-	#define PORT_FEAT_CFG_AUTOGREEN_ENABLED                      0x00000200
-
 	#define PORT_FEATURE_EN_SIZE_MASK                   0x0f000000
 	#define PORT_FEATURE_EN_SIZE_SHIFT                           24
 	#define PORT_FEATURE_WOL_ENABLED                             0x01000000
@@ -1139,9 +1129,6 @@ struct shm_dev_info {				/* size */
 
 #define FW_ACK_NUM_OF_POLL  (FW_ACK_TIME_OUT_MS/FW_ACK_POLL_TIME_MS)
 
-/* LED Blink rate that will achieve ~15.9Hz */
-#define LED_BLINK_RATE_VAL      480
-
 /****************************************************************************
  * Driver <-> FW Mailbox                                                    *
  ****************************************************************************/
@@ -1407,7 +1394,7 @@ struct port_mf_cfg {
 	#define PORT_MF_CFG_E1HOV_TAG_SHIFT             0
 	#define PORT_MF_CFG_E1HOV_TAG_DEFAULT         PORT_MF_CFG_E1HOV_TAG_MASK
 
-	u32 reserved[3];
+	u32 reserved[1];
 
 };
 
@@ -1493,7 +1480,8 @@ struct func_ext_cfg {
 struct mf_cfg {
 
 	struct shared_mf_cfg    shared_mf_config;       /* 0x4 */
-	struct port_mf_cfg  port_mf_config[PORT_MAX];   /* 0x10 * 2 = 0x20 */
+							/* 0x8*2*2=0x20 */
+	struct port_mf_cfg  port_mf_config[NVM_PATH_MAX][PORT_MAX];
 	/* for all chips, there are 8 mf functions */
 	struct func_mf_cfg  func_mf_config[E1H_FUNC_MAX]; /* 0x18 * 8 = 0xc0 */
 	/*
@@ -2002,6 +1990,7 @@ struct shmem2_region {
 #define DRV_INFO_CONTROL_VER_SHIFT         0
 #define DRV_INFO_CONTROL_OP_CODE_MASK      0x0000ff00
 #define DRV_INFO_CONTROL_OP_CODE_SHIFT     8
+	u32 ibft_host_addr; /* initialized by option ROM */
 };
 
 
@@ -2700,8 +2689,8 @@ union drv_info_to_mcp {
 	struct iscsi_stats_info	iscsi_stat;
 };
 #define BCM_5710_FW_MAJOR_VERSION			7
-#define BCM_5710_FW_MINOR_VERSION			0
-#define BCM_5710_FW_REVISION_VERSION		29
+#define BCM_5710_FW_MINOR_VERSION			2
+#define BCM_5710_FW_REVISION_VERSION		16
 #define BCM_5710_FW_ENGINEERING_VERSION		0
 #define BCM_5710_FW_COMPILE_FLAGS			1
 
@@ -3308,8 +3297,10 @@ struct client_init_rx_data {
 #define CLIENT_INIT_RX_DATA_TPA_EN_IPV4_SHIFT 0
 #define CLIENT_INIT_RX_DATA_TPA_EN_IPV6 (0x1<<1)
 #define CLIENT_INIT_RX_DATA_TPA_EN_IPV6_SHIFT 1
-#define CLIENT_INIT_RX_DATA_RESERVED5 (0x3F<<2)
-#define CLIENT_INIT_RX_DATA_RESERVED5_SHIFT 2
+#define CLIENT_INIT_RX_DATA_TPA_MODE (0x1<<2)
+#define CLIENT_INIT_RX_DATA_TPA_MODE_SHIFT 2
+#define CLIENT_INIT_RX_DATA_RESERVED5 (0x1F<<3)
+#define CLIENT_INIT_RX_DATA_RESERVED5_SHIFT 3
 	u8 vmqueue_mode_en_flg;
 	u8 extra_data_over_sgl_en_flg;
 	u8 cache_line_alignment_log_size;
@@ -3324,7 +3315,7 @@ struct client_init_rx_data {
 	u8 outer_vlan_removal_enable_flg;
 	u8 status_block_id;
 	u8 rx_sb_index_number;
-	u8 reserved0;
+	u8 dont_verify_rings_pause_thr_flg;
 	u8 max_tpa_queues;
 	u8 silent_vlan_removal_flg;
 	__le16 max_bytes_on_bd;
@@ -3657,7 +3648,7 @@ struct eth_fast_path_rx_cqe {
 	u8 placement_offset;
 	__le32 rss_hash_result;
 	__le16 vlan_tag;
-	__le16 pkt_len;
+	__le16 pkt_len_or_gro_seg_len;
 	__le16 len_on_bd;
 	struct parsing_flags pars_flags;
 	union eth_sgl_or_raw_data sgl_or_raw_data;
@@ -4214,6 +4205,15 @@ enum set_mac_action_type {
 };
 
 
+/*
+ * Ethernet TPA Modes
+ */
+enum tpa_mode {
+	TPA_LRO,
+	TPA_GRO,
+	MAX_TPA_MODE};
+
+
 /*
  * tpa update ramrod data
  */
@@ -4224,7 +4224,8 @@ struct tpa_update_ramrod_data {
 	u8 max_tpa_queues;
 	u8 max_sges_for_packet;
 	u8 complete_on_both_clients;
-	__le16 reserved1;
+	u8 dont_verify_rings_pause_thr_flg;
+	u8 tpa_mode;
 	__le16 sge_buff_size;
 	__le16 max_agg_size;
 	__le32 sge_page_base_lo;
@@ -4447,13 +4448,13 @@ enum common_spqe_cmd_id {
 	RAMROD_CMD_ID_COMMON_UNUSED,
 	RAMROD_CMD_ID_COMMON_FUNCTION_START,
 	RAMROD_CMD_ID_COMMON_FUNCTION_STOP,
+	RAMROD_CMD_ID_COMMON_FUNCTION_UPDATE,
 	RAMROD_CMD_ID_COMMON_CFC_DEL,
 	RAMROD_CMD_ID_COMMON_CFC_DEL_WB,
 	RAMROD_CMD_ID_COMMON_STAT_QUERY,
 	RAMROD_CMD_ID_COMMON_STOP_TRAFFIC,
 	RAMROD_CMD_ID_COMMON_START_TRAFFIC,
 	RAMROD_CMD_ID_COMMON_RESERVED1,
-	RAMROD_CMD_ID_COMMON_RESERVED2,
 	MAX_COMMON_SPQE_CMD_ID
 };
 
@@ -4733,8 +4734,8 @@ enum event_ring_opcode {
 	EVENT_RING_OPCODE_MALICIOUS_VF,
 	EVENT_RING_OPCODE_FORWARD_SETUP,
 	EVENT_RING_OPCODE_RSS_UPDATE_RULES,
+	EVENT_RING_OPCODE_FUNCTION_UPDATE,
 	EVENT_RING_OPCODE_RESERVED1,
-	EVENT_RING_OPCODE_RESERVED2,
 	EVENT_RING_OPCODE_SET_MAC,
 	EVENT_RING_OPCODE_CLASSIFICATION_RULES,
 	EVENT_RING_OPCODE_FILTERS_RULES,

drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c

@@ -2702,6 +2702,8 @@ static inline unsigned long bnx2x_get_q_flags(struct bnx2x *bp,
 	if (!fp->disable_tpa) {
 		__set_bit(BNX2X_Q_FLG_TPA, &flags);
 		__set_bit(BNX2X_Q_FLG_TPA_IPV6, &flags);
+		if (fp->mode == TPA_MODE_GRO)
+			__set_bit(BNX2X_Q_FLG_TPA_GRO, &flags);
 	}
 
 	if (leading) {
@@ -10167,10 +10169,10 @@ static int __devinit bnx2x_init_bp(struct bnx2x *bp)
 
 	/* Set TPA flags */
 	if (bp->disable_tpa) {
-		bp->flags &= ~TPA_ENABLE_FLAG;
+		bp->flags &= ~(TPA_ENABLE_FLAG | GRO_ENABLE_FLAG);
 		bp->dev->features &= ~NETIF_F_LRO;
 	} else {
-		bp->flags |= TPA_ENABLE_FLAG;
+		bp->flags |= (TPA_ENABLE_FLAG | GRO_ENABLE_FLAG);
 		bp->dev->features |= NETIF_F_LRO;
 	}
 
@@ -10716,8 +10718,9 @@ static int __devinit bnx2x_init_dev(struct pci_dev *pdev,
 	dev->priv_flags |= IFF_UNICAST_FLT;
 
 	dev->hw_features = NETIF_F_SG | NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM |
-		NETIF_F_TSO | NETIF_F_TSO_ECN | NETIF_F_TSO6 | NETIF_F_LRO |
-		NETIF_F_RXCSUM | NETIF_F_RXHASH | NETIF_F_HW_VLAN_TX;
+		NETIF_F_TSO | NETIF_F_TSO_ECN | NETIF_F_TSO6 |
+		NETIF_F_RXCSUM | NETIF_F_LRO | NETIF_F_GRO |
+		NETIF_F_RXHASH | NETIF_F_HW_VLAN_TX;
 
 	dev->vlan_features = NETIF_F_SG | NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM |
 		NETIF_F_TSO | NETIF_F_TSO_ECN | NETIF_F_TSO6 | NETIF_F_HIGHDMA;

drivers/net/ethernet/broadcom/bnx2x/bnx2x_sp.c

@@ -4431,9 +4431,10 @@ static void bnx2x_q_fill_init_rx_data(struct bnx2x_queue_sp_obj *o,
 				struct client_init_rx_data *rx_data,
 				unsigned long *flags)
 {
-		/* Rx data */
 	rx_data->tpa_en = test_bit(BNX2X_Q_FLG_TPA, flags) *
 				CLIENT_INIT_RX_DATA_TPA_EN_IPV4;
+	rx_data->tpa_en |= test_bit(BNX2X_Q_FLG_TPA_GRO, flags) *
+				CLIENT_INIT_RX_DATA_TPA_MODE;
 	rx_data->vmqueue_mode_en_flg = 0;
 
 	rx_data->cache_line_alignment_log_size =

drivers/net/ethernet/broadcom/bnx2x/bnx2x_sp.h

@@ -774,6 +774,7 @@ enum bnx2x_queue_cmd {
 enum {
 	BNX2X_Q_FLG_TPA,
 	BNX2X_Q_FLG_TPA_IPV6,
+	BNX2X_Q_FLG_TPA_GRO,
 	BNX2X_Q_FLG_STATS,
 	BNX2X_Q_FLG_ZERO_STATS,
 	BNX2X_Q_FLG_ACTIVE,