Merge git://github.com/davem330/net

* git://github.com/davem330/net: (62 commits)
  ipv6: don't use inetpeer to store metrics for routes.
  can: ti_hecc: include linux/io.h
  IRDA: Fix global type conflicts in net/irda/irsysctl.c v2
  net: Handle different key sizes between address families in flow cache
  net: Align AF-specific flowi structs to long
  ipv4: Fix fib_info->fib_metrics leak
  caif: fix a potential NULL dereference
  sctp: deal with multiple COOKIE_ECHO chunks
  ibmveth: Fix checksum offload failure handling
  ibmveth: Checksum offload is always disabled
  ibmveth: Fix issue with DMA mapping failure
  ibmveth: Fix DMA unmap error
  pch_gbe: support ML7831 IOH
  pch_gbe: added the process of FIFO over run error
  pch_gbe: fixed the issue which receives an unnecessary packet.
  sfc: Use 64-bit writes for TX push where possible
  Revert "sfc: Use write-combining to reduce TX latency" and follow-ups
  bnx2x: Fix ethtool advertisement
  bnx2x: Fix 578xx link LED
  bnx2x: Fix XMAC loopback test
  ...

Documentation/networking/dmfe.txt

@@ -1,3 +1,5 @@
+Note: This driver doesn't have a maintainer.
+
 Davicom DM9102(A)/DM9132/DM9801 fast ethernet driver for Linux.
 
 This program is free software; you can redistribute it and/or
@@ -55,7 +57,6 @@ Test and make sure PCI latency is now correct for all cases.
 Authors:
 
 Sten Wang <[email protected] >   : Original Author
-Tobias Ringstrom <[email protected]> : Current Maintainer
 
 Contributors:
 

MAINTAINERS

@@ -1278,7 +1278,6 @@ F:	drivers/input/misc/ati_remote2.c
 ATLX ETHERNET DRIVERS
 M:	Jay Cliburn <[email protected]>
 M:	Chris Snook <[email protected]>
-M:	Jie Yang <[email protected]>
 L:	[email protected]
 W:	http://sourceforge.net/projects/atl1
 W:	http://atl1.sourceforge.net
@@ -1574,7 +1573,6 @@ F:	drivers/scsi/bfa/
 
 BROCADE BNA 10 GIGABIT ETHERNET DRIVER
 M:	Rasesh Mody <[email protected]>
-M:	Debashis Dutt <[email protected]>
 L:	[email protected]
 S:	Supported
 F:	drivers/net/bna/
@@ -1758,7 +1756,6 @@ F:	Documentation/zh_CN/
 
 CISCO VIC ETHERNET NIC DRIVER
 M:	Christian Benvenuti <[email protected]>
-M:	Vasanthy Kolluri <[email protected]>
 M:	Roopa Prabhu <[email protected]>
 M:	David Wang <[email protected]>
 S:	Supported
@@ -4415,7 +4412,8 @@ L:	[email protected]
 L:	[email protected]
 W:	http://www.netfilter.org/
 W:	http://www.iptables.org/
-T:	git git://git.kernel.org/pub/scm/linux/kernel/git/kaber/nf-2.6.git
+T:	git git://git.kernel.org/pub/scm/linux/kernel/git/netfilter/nf-2.6.git
+T:	git git://git.kernel.org/pub/scm/linux/kernel/git/netfilter/nf-next-2.6.git
 S:	Supported
 F:	include/linux/netfilter*
 F:	include/linux/netfilter/

drivers/net/Kconfig

@@ -2535,7 +2535,7 @@ config S6GMAC
 source "drivers/net/stmmac/Kconfig"
 
 config PCH_GBE
-	tristate "Intel EG20T PCH / OKI SEMICONDUCTOR ML7223 IOH GbE"
+	tristate "Intel EG20T PCH/OKI SEMICONDUCTOR IOH(ML7223/ML7831) GbE"
 	depends on PCI
 	select MII
 	---help---
@@ -2548,10 +2548,11 @@ config PCH_GBE
 	  This driver enables Gigabit Ethernet function.
 
 	  This driver also can be used for OKI SEMICONDUCTOR IOH(Input/
-	  Output Hub), ML7223.
-	  ML7223 IOH is for MP(Media Phone) use.
-	  ML7223 is companion chip for Intel Atom E6xx series.
-	  ML7223 is completely compatible for Intel EG20T PCH.
+	  Output Hub), ML7223/ML7831.
+	  ML7223 IOH is for MP(Media Phone) use. ML7831 IOH is for general
+	  purpose use.
+	  ML7223/ML7831 is companion chip for Intel Atom E6xx series.
+	  ML7223/ML7831 is completely compatible for Intel EG20T PCH.
 
 config FTGMAC100
 	tristate "Faraday FTGMAC100 Gigabit Ethernet support"

drivers/net/bnx2x/bnx2x.h

@@ -315,6 +315,14 @@ union db_prod {
 	u32		raw;
 };
 
+/* dropless fc FW/HW related params */
+#define BRB_SIZE(bp)		(CHIP_IS_E3(bp) ? 1024 : 512)
+#define MAX_AGG_QS(bp)		(CHIP_IS_E1(bp) ? \
+					ETH_MAX_AGGREGATION_QUEUES_E1 :\
+					ETH_MAX_AGGREGATION_QUEUES_E1H_E2)
+#define FW_DROP_LEVEL(bp)	(3 + MAX_SPQ_PENDING + MAX_AGG_QS(bp))
+#define FW_PREFETCH_CNT		16
+#define DROPLESS_FC_HEADROOM	100
 
 /* MC hsi */
 #define BCM_PAGE_SHIFT		12
@@ -331,15 +339,35 @@ union db_prod {
 /* SGE ring related macros */
 #define NUM_RX_SGE_PAGES	2
 #define RX_SGE_CNT		(BCM_PAGE_SIZE / sizeof(struct eth_rx_sge))
-#define MAX_RX_SGE_CNT		(RX_SGE_CNT - 2)
+#define NEXT_PAGE_SGE_DESC_CNT	2
+#define MAX_RX_SGE_CNT		(RX_SGE_CNT - NEXT_PAGE_SGE_DESC_CNT)
 /* RX_SGE_CNT is promised to be a power of 2 */
 #define RX_SGE_MASK		(RX_SGE_CNT - 1)
 #define NUM_RX_SGE		(RX_SGE_CNT * NUM_RX_SGE_PAGES)
 #define MAX_RX_SGE		(NUM_RX_SGE - 1)
 #define NEXT_SGE_IDX(x)		((((x) & RX_SGE_MASK) == \
-				  (MAX_RX_SGE_CNT - 1)) ? (x) + 3 : (x) + 1)
+				  (MAX_RX_SGE_CNT - 1)) ? \
+					(x) + 1 + NEXT_PAGE_SGE_DESC_CNT : \
+					(x) + 1)
 #define RX_SGE(x)		((x) & MAX_RX_SGE)
 
+/*
+ * Number of required  SGEs is the sum of two:
+ * 1. Number of possible opened aggregations (next packet for
+ *    these aggregations will probably consume SGE immidiatelly)
+ * 2. Rest of BRB blocks divided by 2 (block will consume new SGE only
+ *    after placement on BD for new TPA aggregation)
+ *
+ * Takes into account NEXT_PAGE_SGE_DESC_CNT "next" elements on each page
+ */
+#define NUM_SGE_REQ		(MAX_AGG_QS(bp) + \
+					(BRB_SIZE(bp) - MAX_AGG_QS(bp)) / 2)
+#define NUM_SGE_PG_REQ		((NUM_SGE_REQ + MAX_RX_SGE_CNT - 1) / \
+						MAX_RX_SGE_CNT)
+#define SGE_TH_LO(bp)		(NUM_SGE_REQ + \
+				 NUM_SGE_PG_REQ * NEXT_PAGE_SGE_DESC_CNT)
+#define SGE_TH_HI(bp)		(SGE_TH_LO(bp) + DROPLESS_FC_HEADROOM)
+
 /* Manipulate a bit vector defined as an array of u64 */
 
 /* Number of bits in one sge_mask array element */
@@ -551,24 +579,43 @@ struct bnx2x_fastpath {
 
 #define NUM_TX_RINGS		16
 #define TX_DESC_CNT		(BCM_PAGE_SIZE / sizeof(union eth_tx_bd_types))
-#define MAX_TX_DESC_CNT		(TX_DESC_CNT - 1)
+#define NEXT_PAGE_TX_DESC_CNT	1
+#define MAX_TX_DESC_CNT		(TX_DESC_CNT - NEXT_PAGE_TX_DESC_CNT)
 #define NUM_TX_BD		(TX_DESC_CNT * NUM_TX_RINGS)
 #define MAX_TX_BD		(NUM_TX_BD - 1)
 #define MAX_TX_AVAIL		(MAX_TX_DESC_CNT * NUM_TX_RINGS - 2)
 #define NEXT_TX_IDX(x)		((((x) & MAX_TX_DESC_CNT) == \
-				  (MAX_TX_DESC_CNT - 1)) ? (x) + 2 : (x) + 1)
+				  (MAX_TX_DESC_CNT - 1)) ? \
+					(x) + 1 + NEXT_PAGE_TX_DESC_CNT : \
+					(x) + 1)
 #define TX_BD(x)		((x) & MAX_TX_BD)
 #define TX_BD_POFF(x)		((x) & MAX_TX_DESC_CNT)
 
 /* The RX BD ring is special, each bd is 8 bytes but the last one is 16 */
 #define NUM_RX_RINGS		8
 #define RX_DESC_CNT		(BCM_PAGE_SIZE / sizeof(struct eth_rx_bd))
-#define MAX_RX_DESC_CNT		(RX_DESC_CNT - 2)
+#define NEXT_PAGE_RX_DESC_CNT	2
+#define MAX_RX_DESC_CNT		(RX_DESC_CNT - NEXT_PAGE_RX_DESC_CNT)
 #define RX_DESC_MASK		(RX_DESC_CNT - 1)
 #define NUM_RX_BD		(RX_DESC_CNT * NUM_RX_RINGS)
 #define MAX_RX_BD		(NUM_RX_BD - 1)
 #define MAX_RX_AVAIL		(MAX_RX_DESC_CNT * NUM_RX_RINGS - 2)
-#define MIN_RX_AVAIL		128
+
+/* dropless fc calculations for BDs
+ *
+ * Number of BDs should as number of buffers in BRB:
+ * Low threshold takes into account NEXT_PAGE_RX_DESC_CNT
+ * "next" elements on each page
+ */
+#define NUM_BD_REQ		BRB_SIZE(bp)
+#define NUM_BD_PG_REQ		((NUM_BD_REQ + MAX_RX_DESC_CNT - 1) / \
+					      MAX_RX_DESC_CNT)
+#define BD_TH_LO(bp)		(NUM_BD_REQ + \
+				 NUM_BD_PG_REQ * NEXT_PAGE_RX_DESC_CNT + \
+				 FW_DROP_LEVEL(bp))
+#define BD_TH_HI(bp)		(BD_TH_LO(bp) + DROPLESS_FC_HEADROOM)
+
+#define MIN_RX_AVAIL		((bp)->dropless_fc ? BD_TH_HI(bp) + 128 : 128)
 
 #define MIN_RX_SIZE_TPA_HW	(CHIP_IS_E1(bp) ? \
 					ETH_MIN_RX_CQES_WITH_TPA_E1 : \
@@ -579,7 +626,9 @@ struct bnx2x_fastpath {
 								MIN_RX_AVAIL))
 
 #define NEXT_RX_IDX(x)		((((x) & RX_DESC_MASK) == \
-				  (MAX_RX_DESC_CNT - 1)) ? (x) + 3 : (x) + 1)
+				  (MAX_RX_DESC_CNT - 1)) ? \
+					(x) + 1 + NEXT_PAGE_RX_DESC_CNT : \
+					(x) + 1)
 #define RX_BD(x)		((x) & MAX_RX_BD)
 
 /*
@@ -589,14 +638,31 @@ struct bnx2x_fastpath {
 #define CQE_BD_REL	(sizeof(union eth_rx_cqe) / sizeof(struct eth_rx_bd))
 #define NUM_RCQ_RINGS		(NUM_RX_RINGS * CQE_BD_REL)
 #define RCQ_DESC_CNT		(BCM_PAGE_SIZE / sizeof(union eth_rx_cqe))
-#define MAX_RCQ_DESC_CNT	(RCQ_DESC_CNT - 1)
+#define NEXT_PAGE_RCQ_DESC_CNT	1
+#define MAX_RCQ_DESC_CNT	(RCQ_DESC_CNT - NEXT_PAGE_RCQ_DESC_CNT)
 #define NUM_RCQ_BD		(RCQ_DESC_CNT * NUM_RCQ_RINGS)
 #define MAX_RCQ_BD		(NUM_RCQ_BD - 1)
 #define MAX_RCQ_AVAIL		(MAX_RCQ_DESC_CNT * NUM_RCQ_RINGS - 2)
 #define NEXT_RCQ_IDX(x)		((((x) & MAX_RCQ_DESC_CNT) == \
-				  (MAX_RCQ_DESC_CNT - 1)) ? (x) + 2 : (x) + 1)
+				  (MAX_RCQ_DESC_CNT - 1)) ? \
+					(x) + 1 + NEXT_PAGE_RCQ_DESC_CNT : \
+					(x) + 1)
 #define RCQ_BD(x)		((x) & MAX_RCQ_BD)
 
+/* dropless fc calculations for RCQs
+ *
+ * Number of RCQs should be as number of buffers in BRB:
+ * Low threshold takes into account NEXT_PAGE_RCQ_DESC_CNT
+ * "next" elements on each page
+ */
+#define NUM_RCQ_REQ		BRB_SIZE(bp)
+#define NUM_RCQ_PG_REQ		((NUM_BD_REQ + MAX_RCQ_DESC_CNT - 1) / \
+					      MAX_RCQ_DESC_CNT)
+#define RCQ_TH_LO(bp)		(NUM_RCQ_REQ + \
+				 NUM_RCQ_PG_REQ * NEXT_PAGE_RCQ_DESC_CNT + \
+				 FW_DROP_LEVEL(bp))
+#define RCQ_TH_HI(bp)		(RCQ_TH_LO(bp) + DROPLESS_FC_HEADROOM)
+
 
 /* This is needed for determining of last_max */
 #define SUB_S16(a, b)		(s16)((s16)(a) - (s16)(b))
@@ -685,24 +751,17 @@ struct bnx2x_fastpath {
 #define FP_CSB_FUNC_OFF	\
 			offsetof(struct cstorm_status_block_c, func)
 
-#define HC_INDEX_TOE_RX_CQ_CONS		0 /* Formerly Ustorm TOE CQ index */
-					  /* (HC_INDEX_U_TOE_RX_CQ_CONS)  */
-#define HC_INDEX_ETH_RX_CQ_CONS		1 /* Formerly Ustorm ETH CQ index */
-					  /* (HC_INDEX_U_ETH_RX_CQ_CONS)  */
-#define HC_INDEX_ETH_RX_BD_CONS		2 /* Formerly Ustorm ETH BD index */
-					  /* (HC_INDEX_U_ETH_RX_BD_CONS)  */
-
-#define HC_INDEX_TOE_TX_CQ_CONS		4 /* Formerly Cstorm TOE CQ index   */
-					  /* (HC_INDEX_C_TOE_TX_CQ_CONS)    */
-#define HC_INDEX_ETH_TX_CQ_CONS_COS0	5 /* Formerly Cstorm ETH CQ index   */
-					  /* (HC_INDEX_C_ETH_TX_CQ_CONS)    */
-#define HC_INDEX_ETH_TX_CQ_CONS_COS1	6 /* Formerly Cstorm ETH CQ index   */
-					  /* (HC_INDEX_C_ETH_TX_CQ_CONS)    */
-#define HC_INDEX_ETH_TX_CQ_CONS_COS2	7 /* Formerly Cstorm ETH CQ index   */
-					  /* (HC_INDEX_C_ETH_TX_CQ_CONS)    */
+#define HC_INDEX_ETH_RX_CQ_CONS		1
 
-#define HC_INDEX_ETH_FIRST_TX_CQ_CONS	HC_INDEX_ETH_TX_CQ_CONS_COS0
+#define HC_INDEX_OOO_TX_CQ_CONS		4
 
+#define HC_INDEX_ETH_TX_CQ_CONS_COS0	5
+
+#define HC_INDEX_ETH_TX_CQ_CONS_COS1	6
+
+#define HC_INDEX_ETH_TX_CQ_CONS_COS2	7
+
+#define HC_INDEX_ETH_FIRST_TX_CQ_CONS	HC_INDEX_ETH_TX_CQ_CONS_COS0
 
 #define BNX2X_RX_SB_INDEX \
 	(&fp->sb_index_values[HC_INDEX_ETH_RX_CQ_CONS])
@@ -1100,11 +1159,12 @@ struct bnx2x {
 #define BP_PORT(bp)			(bp->pfid & 1)
 #define BP_FUNC(bp)			(bp->pfid)
 #define BP_ABS_FUNC(bp)			(bp->pf_num)
-#define BP_E1HVN(bp)			(bp->pfid >> 1)
-#define BP_VN(bp)			(BP_E1HVN(bp)) /*remove when approved*/
-#define BP_L_ID(bp)			(BP_E1HVN(bp) << 2)
-#define BP_FW_MB_IDX(bp)		(BP_PORT(bp) +\
-	  BP_VN(bp) * ((CHIP_IS_E1x(bp) || (CHIP_MODE_IS_4_PORT(bp))) ? 2  : 1))
+#define BP_VN(bp)			((bp)->pfid >> 1)
+#define BP_MAX_VN_NUM(bp)		(CHIP_MODE_IS_4_PORT(bp) ? 2 : 4)
+#define BP_L_ID(bp)			(BP_VN(bp) << 2)
+#define BP_FW_MB_IDX_VN(bp, vn)		(BP_PORT(bp) +\
+	  (vn) * ((CHIP_IS_E1x(bp) || (CHIP_MODE_IS_4_PORT(bp))) ? 2  : 1))
+#define BP_FW_MB_IDX(bp)		BP_FW_MB_IDX_VN(bp, BP_VN(bp))
 
 	struct net_device	*dev;
 	struct pci_dev		*pdev;
@@ -1767,7 +1827,7 @@ static inline u32 reg_poll(struct bnx2x *bp, u32 reg, u32 expected, int ms,
 
 #define MAX_DMAE_C_PER_PORT		8
 #define INIT_DMAE_C(bp)			(BP_PORT(bp) * MAX_DMAE_C_PER_PORT + \
-					 BP_E1HVN(bp))
+					 BP_VN(bp))
 #define PMF_DMAE_C(bp)			(BP_PORT(bp) * MAX_DMAE_C_PER_PORT + \
 					 E1HVN_MAX)
 
@@ -1793,7 +1853,7 @@ static inline u32 reg_poll(struct bnx2x *bp, u32 reg, u32 expected, int ms,
 
 /* must be used on a CID before placing it on a HW ring */
 #define HW_CID(bp, x)			((BP_PORT(bp) << 23) | \
-					 (BP_E1HVN(bp) << BNX2X_SWCID_SHIFT) | \
+					 (BP_VN(bp) << BNX2X_SWCID_SHIFT) | \
 					 (x))
 
 #define SP_DESC_CNT		(BCM_PAGE_SIZE / sizeof(struct eth_spe))

drivers/net/bnx2x/bnx2x_cmn.c

@@ -987,8 +987,6 @@ void __bnx2x_link_report(struct bnx2x *bp)
 void bnx2x_init_rx_rings(struct bnx2x *bp)
 {
 	int func = BP_FUNC(bp);
-	int max_agg_queues = CHIP_IS_E1(bp) ? ETH_MAX_AGGREGATION_QUEUES_E1 :
-					      ETH_MAX_AGGREGATION_QUEUES_E1H_E2;
 	u16 ring_prod;
 	int i, j;
 
@@ -1001,7 +999,7 @@ void bnx2x_init_rx_rings(struct bnx2x *bp)
 
 		if (!fp->disable_tpa) {
 			/* Fill the per-aggregtion pool */
-			for (i = 0; i < max_agg_queues; i++) {
+			for (i = 0; i < MAX_AGG_QS(bp); i++) {
 				struct bnx2x_agg_info *tpa_info =
 					&fp->tpa_info[i];
 				struct sw_rx_bd *first_buf =
@@ -1041,7 +1039,7 @@ void bnx2x_init_rx_rings(struct bnx2x *bp)
 					bnx2x_free_rx_sge_range(bp, fp,
 								ring_prod);
 					bnx2x_free_tpa_pool(bp, fp,
-							    max_agg_queues);
+							    MAX_AGG_QS(bp));
 					fp->disable_tpa = 1;
 					ring_prod = 0;
 					break;
@@ -1137,9 +1135,7 @@ static void bnx2x_free_rx_skbs(struct bnx2x *bp)
 		bnx2x_free_rx_bds(fp);
 
 		if (!fp->disable_tpa)
-			bnx2x_free_tpa_pool(bp, fp, CHIP_IS_E1(bp) ?
-					    ETH_MAX_AGGREGATION_QUEUES_E1 :
-					    ETH_MAX_AGGREGATION_QUEUES_E1H_E2);
+			bnx2x_free_tpa_pool(bp, fp, MAX_AGG_QS(bp));
 	}
 }
 
@@ -3095,15 +3091,20 @@ static int bnx2x_alloc_fp_mem_at(struct bnx2x *bp, int index)
 	struct bnx2x_fastpath *fp = &bp->fp[index];
 	int ring_size = 0;
 	u8 cos;
+	int rx_ring_size = 0;
 
 	/* if rx_ring_size specified - use it */
-	int rx_ring_size = bp->rx_ring_size ? bp->rx_ring_size :
-			   MAX_RX_AVAIL/BNX2X_NUM_RX_QUEUES(bp);
+	if (!bp->rx_ring_size) {
 
-	/* allocate at least number of buffers required by FW */
-	rx_ring_size = max_t(int, bp->disable_tpa ? MIN_RX_SIZE_NONTPA :
-						    MIN_RX_SIZE_TPA,
-				  rx_ring_size);
+		rx_ring_size = MAX_RX_AVAIL/BNX2X_NUM_RX_QUEUES(bp);
+
+		/* allocate at least number of buffers required by FW */
+		rx_ring_size = max_t(int, bp->disable_tpa ? MIN_RX_SIZE_NONTPA :
+				     MIN_RX_SIZE_TPA, rx_ring_size);
+
+		bp->rx_ring_size = rx_ring_size;
+	} else
+		rx_ring_size = bp->rx_ring_size;
 
 	/* Common */
 	sb = &bnx2x_fp(bp, index, status_blk);