bnx2x: Using the new FW

The new FW improves the packets per second rate. It required a lot of change in
the FW which implies many changes in the driver to support it. It is now also
possible for the driver to use a separate MSI-X vector for Rx and Tx - this also
add some to the complicity of this change.

All things said - after this patch, practically all performance matrixes show
improvement.
Though Vladislav Zolotarov is not signed on this patch, he did most of the job
and deserves credit for that.
Signed-off-by: NEilon Greenstein <eilong@broadcom.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

drivers/net/bnx2x.h

@@ -142,6 +142,9 @@ struct sw_rx_bd {
 struct sw_tx_bd {
 	struct sk_buff	*skb;
 	u16		first_bd;
+	u8		flags;
+/* Set on the first BD descriptor when there is a split BD */
+#define BNX2X_TSO_SPLIT_BD		(1<<0)
 };
 
 struct sw_rx_page {
@@ -149,6 +152,11 @@ struct sw_rx_page {
 	DECLARE_PCI_UNMAP_ADDR(mapping)
 };
 
+union db_prod {
+	struct doorbell_set_prod data;
+	u32		raw;
+};
+
 
 /* MC hsi */
 #define BCM_PAGE_SHIFT			12
@@ -234,15 +242,14 @@ struct bnx2x_fastpath {
 
 	struct napi_struct	napi;
 
+	u8			is_rx_queue;
+
 	struct host_status_block *status_blk;
 	dma_addr_t		status_blk_mapping;
 
-	struct eth_tx_db_data	*hw_tx_prods;
-	dma_addr_t		tx_prods_mapping;
-
 	struct sw_tx_bd		*tx_buf_ring;
 
-	struct eth_tx_bd	*tx_desc_ring;
+	union eth_tx_bd_types	*tx_desc_ring;
 	dma_addr_t		tx_desc_mapping;
 
 	struct sw_rx_bd		*rx_buf_ring;	/* BDs mappings ring */
@@ -272,6 +279,8 @@ struct bnx2x_fastpath {
 	u8			cl_id;	/* eth client id */
 	u8			sb_id;	/* status block number in HW */
 
+	union db_prod		tx_db;
+
 	u16			tx_pkt_prod;
 	u16			tx_pkt_cons;
 	u16			tx_bd_prod;
@@ -309,21 +318,24 @@ struct bnx2x_fastpath {
 	struct xstorm_per_client_stats old_xclient;
 	struct bnx2x_eth_q_stats eth_q_stats;
 
-	char			name[IFNAMSIZ];
+	/* The size is calculated using the following:
+	     sizeof name field from netdev structure +
+	     4 ('-Xx-' string) +
+	     4 (for the digits and to make it DWORD aligned) */
+#define FP_NAME_SIZE		(sizeof(((struct net_device *)0)->name) + 8)
+	char			name[FP_NAME_SIZE];
 	struct bnx2x		*bp; /* parent */
 };
 
 #define bnx2x_fp(bp, nr, var)		(bp->fp[nr].var)
 
-#define BNX2X_HAS_WORK(fp)	(bnx2x_has_rx_work(fp) || bnx2x_has_tx_work(fp))
-
 
 /* MC hsi */
 #define MAX_FETCH_BD			13	/* HW max BDs per packet */
 #define RX_COPY_THRESH			92
 
 #define NUM_TX_RINGS			16
-#define TX_DESC_CNT		(BCM_PAGE_SIZE / sizeof(struct eth_tx_bd))
+#define TX_DESC_CNT		(BCM_PAGE_SIZE / sizeof(union eth_tx_bd_types))
 #define MAX_TX_DESC_CNT			(TX_DESC_CNT - 1)
 #define NUM_TX_BD			(TX_DESC_CNT * NUM_TX_RINGS)
 #define MAX_TX_BD			(NUM_TX_BD - 1)
@@ -395,7 +407,7 @@ struct bnx2x_fastpath {
 #define DPM_TRIGER_TYPE			0x40
 #define DOORBELL(bp, cid, val) \
 	do { \
-		writel((u32)val, (bp)->doorbells + (BCM_PAGE_SIZE * cid) + \
+		writel((u32)(val), bp->doorbells + (BCM_PAGE_SIZE * (cid)) + \
 		       DPM_TRIGER_TYPE); \
 	} while (0)
 
@@ -902,8 +914,6 @@ struct bnx2x {
 	u16			rx_quick_cons_trip;
 	u16			rx_ticks_int;
 	u16			rx_ticks;
-/* Maximal coalescing timeout in us */
-#define BNX2X_MAX_COALESCE_TOUT		(0xf0*12)
 
 	u32			lin_cnt;
 
@@ -985,19 +995,20 @@ struct bnx2x {
 };
 
 
-#define BNX2X_MAX_QUEUES(bp)	(IS_E1HMF(bp) ? (MAX_CONTEXT / E1HVN_MAX) : \
-						 MAX_CONTEXT)
-#define BNX2X_NUM_QUEUES(bp)	max(bp->num_rx_queues, bp->num_tx_queues)
-#define is_multi(bp)		(BNX2X_NUM_QUEUES(bp) > 1)
+#define BNX2X_MAX_QUEUES(bp)	(IS_E1HMF(bp) ? (MAX_CONTEXT/(2 * E1HVN_MAX)) \
+					      : (MAX_CONTEXT/2))
+#define BNX2X_NUM_QUEUES(bp)	(bp->num_rx_queues + bp->num_tx_queues)
+#define is_multi(bp)		(BNX2X_NUM_QUEUES(bp) > 2)
 
 #define for_each_rx_queue(bp, var) \
 			for (var = 0; var < bp->num_rx_queues; var++)
 #define for_each_tx_queue(bp, var) \
-			for (var = 0; var < bp->num_tx_queues; var++)
+			for (var = bp->num_rx_queues; \
+			     var < BNX2X_NUM_QUEUES(bp); var++)
 #define for_each_queue(bp, var) \
 			for (var = 0; var < BNX2X_NUM_QUEUES(bp); var++)
 #define for_each_nondefault_queue(bp, var) \
-			for (var = 1; var < BNX2X_NUM_QUEUES(bp); var++)
+			for (var = 1; var < bp->num_rx_queues; var++)
 
 
 void bnx2x_read_dmae(struct bnx2x *bp, u32 src_addr, u32 len32);

drivers/net/bnx2x_reg.h

@@ -370,7 +370,6 @@
 #define CFC_REG_NUM_LCIDS_LEAVING				 0x104018
 /* [RW 8] The event id for aggregated interrupt 0 */
 #define CSDM_REG_AGG_INT_EVENT_0				 0xc2038
-#define CSDM_REG_AGG_INT_EVENT_1				 0xc203c
 #define CSDM_REG_AGG_INT_EVENT_10				 0xc2060
 #define CSDM_REG_AGG_INT_EVENT_11				 0xc2064
 #define CSDM_REG_AGG_INT_EVENT_12				 0xc2068
@@ -378,37 +377,27 @@
 #define CSDM_REG_AGG_INT_EVENT_14				 0xc2070
 #define CSDM_REG_AGG_INT_EVENT_15				 0xc2074
 #define CSDM_REG_AGG_INT_EVENT_16				 0xc2078
-#define CSDM_REG_AGG_INT_EVENT_17				 0xc207c
-#define CSDM_REG_AGG_INT_EVENT_18				 0xc2080
-#define CSDM_REG_AGG_INT_EVENT_19				 0xc2084
 #define CSDM_REG_AGG_INT_EVENT_2				 0xc2040
-#define CSDM_REG_AGG_INT_EVENT_20				 0xc2088
-#define CSDM_REG_AGG_INT_EVENT_21				 0xc208c
-#define CSDM_REG_AGG_INT_EVENT_22				 0xc2090
-#define CSDM_REG_AGG_INT_EVENT_23				 0xc2094
-#define CSDM_REG_AGG_INT_EVENT_24				 0xc2098
-#define CSDM_REG_AGG_INT_EVENT_25				 0xc209c
-#define CSDM_REG_AGG_INT_EVENT_26				 0xc20a0
-#define CSDM_REG_AGG_INT_EVENT_27				 0xc20a4
-#define CSDM_REG_AGG_INT_EVENT_28				 0xc20a8
-#define CSDM_REG_AGG_INT_EVENT_29				 0xc20ac
 #define CSDM_REG_AGG_INT_EVENT_3				 0xc2044
-#define CSDM_REG_AGG_INT_EVENT_30				 0xc20b0
-#define CSDM_REG_AGG_INT_EVENT_31				 0xc20b4
 #define CSDM_REG_AGG_INT_EVENT_4				 0xc2048
-/* [RW 1] The T bit for aggregated interrupt 0 */
-#define CSDM_REG_AGG_INT_T_0					 0xc20b8
-#define CSDM_REG_AGG_INT_T_1					 0xc20bc
-#define CSDM_REG_AGG_INT_T_10					 0xc20e0
-#define CSDM_REG_AGG_INT_T_11					 0xc20e4
-#define CSDM_REG_AGG_INT_T_12					 0xc20e8
-#define CSDM_REG_AGG_INT_T_13					 0xc20ec
-#define CSDM_REG_AGG_INT_T_14					 0xc20f0
-#define CSDM_REG_AGG_INT_T_15					 0xc20f4
-#define CSDM_REG_AGG_INT_T_16					 0xc20f8
-#define CSDM_REG_AGG_INT_T_17					 0xc20fc
-#define CSDM_REG_AGG_INT_T_18					 0xc2100
-#define CSDM_REG_AGG_INT_T_19					 0xc2104
+#define CSDM_REG_AGG_INT_EVENT_5				 0xc204c
+#define CSDM_REG_AGG_INT_EVENT_6				 0xc2050
+#define CSDM_REG_AGG_INT_EVENT_7				 0xc2054
+#define CSDM_REG_AGG_INT_EVENT_8				 0xc2058
+#define CSDM_REG_AGG_INT_EVENT_9				 0xc205c
+/* [RW 1] For each aggregated interrupt index whether the mode is normal (0)
+   or auto-mask-mode (1) */
+#define CSDM_REG_AGG_INT_MODE_10				 0xc21e0
+#define CSDM_REG_AGG_INT_MODE_11				 0xc21e4
+#define CSDM_REG_AGG_INT_MODE_12				 0xc21e8
+#define CSDM_REG_AGG_INT_MODE_13				 0xc21ec
+#define CSDM_REG_AGG_INT_MODE_14				 0xc21f0
+#define CSDM_REG_AGG_INT_MODE_15				 0xc21f4
+#define CSDM_REG_AGG_INT_MODE_16				 0xc21f8
+#define CSDM_REG_AGG_INT_MODE_6 				 0xc21d0
+#define CSDM_REG_AGG_INT_MODE_7 				 0xc21d4
+#define CSDM_REG_AGG_INT_MODE_8 				 0xc21d8
+#define CSDM_REG_AGG_INT_MODE_9 				 0xc21dc
 /* [RW 13] The start address in the internal RAM for the cfc_rsp lcid */
 #define CSDM_REG_CFC_RSP_START_ADDR				 0xc2008
 /* [RW 16] The maximum value of the competion counter #0 */
@@ -1421,6 +1410,8 @@
 /* [RW 1] e1hmf for WOL. If clr WOL signal o the PXP will be send on bit 0
    only. */
 #define MISC_REG_E1HMF_MODE					 0xa5f8
+/* [RW 32] Debug only: spare RW register reset by core reset */
+#define MISC_REG_GENERIC_CR_0					 0xa460
 /* [RW 32] GPIO. [31-28] FLOAT port 0; [27-24] FLOAT port 0; When any of
    these bits is written as a '1'; the corresponding SPIO bit will turn off
    it's drivers and become an input. This is the reset state of all GPIO
@@ -1729,6 +1720,7 @@
 /* [RW 3] for port0 enable for llfc ppp and pause. b0 - brb1 enable; b1-
    tsdm enable; b2- usdm enable */
 #define NIG_REG_LLFC_EGRESS_SRC_ENABLE_0			 0x16070
+#define NIG_REG_LLFC_EGRESS_SRC_ENABLE_1			 0x16074
 /* [RW 1] SAFC enable for port0. This register may get 1 only when
    ~ppp_enable.ppp_enable = 0 and pause_enable.pause_enable =0 for the same
    port */
@@ -2079,6 +2071,7 @@
 #define PXP2_REG_PGL_ADDR_94_F0 				 0x120540
 #define PXP2_REG_PGL_CONTROL0					 0x120490
 #define PXP2_REG_PGL_CONTROL1					 0x120514
+#define PXP2_REG_PGL_DEBUG					 0x120520
 /* [RW 32] third dword data of expansion rom request. this register is
    special. reading from it provides a vector outstanding read requests. if
    a bit is zero it means that a read request on the corresponding tag did
@@ -2238,6 +2231,9 @@
 /* [RW 8] The maximum number of blocks in Tetris Buffer that can be
    allocated for vq22 */
 #define PXP2_REG_RD_MAX_BLKS_VQ22				 0x1203d0
+/* [RW 8] The maximum number of blocks in Tetris Buffer that can be
+   allocated for vq25 */
+#define PXP2_REG_RD_MAX_BLKS_VQ25				 0x1203dc
 /* [RW 8] The maximum number of blocks in Tetris Buffer that can be
    allocated for vq6 */
 #define PXP2_REG_RD_MAX_BLKS_VQ6				 0x120390
@@ -3835,6 +3831,7 @@
 #define TM_REG_LIN0_PHY_ADDR					 0x164270
 /* [RW 1] Linear0 physical address valid. */
 #define TM_REG_LIN0_PHY_ADDR_VALID				 0x164248
+#define TM_REG_LIN0_SCAN_ON					 0x1640d0
 /* [RW 24] Linear0 array scan timeout. */
 #define TM_REG_LIN0_SCAN_TIME					 0x16403c
 /* [RW 32] Linear1 logic address. */
@@ -4363,6 +4360,7 @@
 #define USDM_REG_AGG_INT_EVENT_31				 0xc40b4
 #define USDM_REG_AGG_INT_EVENT_4				 0xc4048
 #define USDM_REG_AGG_INT_EVENT_5				 0xc404c
+#define USDM_REG_AGG_INT_EVENT_6				 0xc4050
 /* [RW 1] For each aggregated interrupt index whether the mode is normal (0)
    or auto-mask-mode (1) */
 #define USDM_REG_AGG_INT_MODE_0 				 0xc41b8
@@ -4379,6 +4377,10 @@
 #define USDM_REG_AGG_INT_MODE_19				 0xc4204
 #define USDM_REG_AGG_INT_MODE_4 				 0xc41c8
 #define USDM_REG_AGG_INT_MODE_5 				 0xc41cc
+#define USDM_REG_AGG_INT_MODE_6 				 0xc41d0
+/* [RW 1] The T bit for aggregated interrupt 5 */
+#define USDM_REG_AGG_INT_T_5					 0xc40cc
+#define USDM_REG_AGG_INT_T_6					 0xc40d0
 /* [RW 13] The start address in the internal RAM for the cfc_rsp lcid */
 #define USDM_REG_CFC_RSP_START_ADDR				 0xc4008
 /* [RW 16] The maximum value of the competion counter #0 */

firmware/Makefile

@@ -32,7 +32,7 @@ fw-shipped-$(CONFIG_ADAPTEC_STARFIRE) += adaptec/starfire_rx.bin \
 					 adaptec/starfire_tx.bin
 fw-shipped-$(CONFIG_ATARI_DSP56K) += dsp56k/bootstrap.bin
 fw-shipped-$(CONFIG_ATM_AMBASSADOR) += atmsar11.fw
-fw-shipped-$(CONFIG_BNX2X) += bnx2x-e1-4.8.53.0.fw bnx2x-e1h-4.8.53.0.fw
+fw-shipped-$(CONFIG_BNX2X) += bnx2x-e1-5.0.21.0.fw bnx2x-e1h-5.0.21.0.fw
 fw-shipped-$(CONFIG_BNX2) += bnx2/bnx2-mips-09-4.6.17.fw \
 			     bnx2/bnx2-rv2p-09-4.6.15.fw \
 			     bnx2/bnx2-mips-06-4.6.16.fw \