Merge master.kernel.org:/pub/scm/linux/kernel/git/torvalds/linux-2.6

Conflicts:

	drivers/usb/input/Makefile
	drivers/usb/input/gtco.c

16 files changed, 806 insertions, 282 deletions

diff --git a/drivers/net/cxgb3/adapter.h b/drivers/net/cxgb3/adapter.h
index 5c97a64451ce..80c3d8f268a7 100644
--- a/drivers/net/cxgb3/adapter.h
+++ b/drivers/net/cxgb3/adapter.h
@@ -74,6 +74,11 @@ enum {				/* adapter flags */
 struct rx_desc;
 struct rx_sw_desc;
 
+struct sge_fl_page {
+	struct skb_frag_struct frag;
+	unsigned char *va;
+};
+
 struct sge_fl {			/* SGE per free-buffer list state */
 	unsigned int buf_size;	/* size of each Rx buffer */
 	unsigned int credits;	/* # of available Rx buffers */
@@ -81,11 +86,13 @@ struct sge_fl {			/* SGE per free-buffer list state */
 	unsigned int cidx;	/* consumer index */
 	unsigned int pidx;	/* producer index */
 	unsigned int gen;	/* free list generation */
+	unsigned int cntxt_id;	/* SGE context id for the free list */
+	struct sge_fl_page page;
 	struct rx_desc *desc;	/* address of HW Rx descriptor ring */
 	struct rx_sw_desc *sdesc;	/* address of SW Rx descriptor ring */
 	dma_addr_t phys_addr;	/* physical address of HW ring start */
-	unsigned int cntxt_id;	/* SGE context id for the free list */
 	unsigned long empty;	/* # of times queue ran out of buffers */
+	unsigned long alloc_failed; /* # of times buffer allocation failed */
 };
 
 /*
@@ -121,6 +128,8 @@ struct sge_rspq {		/* state for an SGE response queue */
 	unsigned long empty;	/* # of times queue ran out of credits */
 	unsigned long nomem;	/* # of responses deferred due to no mem */
 	unsigned long unhandled_irqs;	/* # of spurious intrs */
+	unsigned long starved;
+	unsigned long restarted;
 };
 
 struct tx_desc;
diff --git a/drivers/net/cxgb3/common.h b/drivers/net/cxgb3/common.h
index e23deeb7d06d..8d1379633698 100644
--- a/drivers/net/cxgb3/common.h
+++ b/drivers/net/cxgb3/common.h
@@ -112,8 +112,7 @@ enum {
 };
 
 enum {
-	SUPPORTED_OFFLOAD = 1 << 24,
-	SUPPORTED_IRQ = 1 << 25
+	SUPPORTED_IRQ      = 1 << 24
 };
 
 enum {				/* adapter interrupt-maintained statistics */
@@ -260,6 +259,10 @@ struct mac_stats {
 	unsigned long serdes_signal_loss;
 	unsigned long xaui_pcs_ctc_err;
 	unsigned long xaui_pcs_align_change;
+
+	unsigned long num_toggled; /* # times toggled TxEn due to stuck TX */
+	unsigned long num_resets;  /* # times reset due to stuck TX */
+
 };
 
 struct tp_mib_stats {
@@ -354,6 +357,9 @@ enum {
 	MC5_MODE_72_BIT = 2
 };
 
+/* MC5 min active region size */
+enum { MC5_MIN_TIDS = 16 };
+
 struct vpd_params {
 	unsigned int cclk;
 	unsigned int mclk;
@@ -398,6 +404,13 @@ struct adapter_params {
 	unsigned int stats_update_period;	/* MAC stats accumulation period */
 	unsigned int linkpoll_period;	/* link poll period in 0.1s */
 	unsigned int rev;	/* chip revision */
+	unsigned int offload;
+};
+
+enum {					    /* chip revisions */
+	T3_REV_A  = 0,
+	T3_REV_B  = 2,
+	T3_REV_B2 = 3,
 };
 
 struct trace_params {
@@ -465,6 +478,13 @@ struct cmac {
 	struct adapter *adapter;
 	unsigned int offset;
 	unsigned int nucast;	/* # of address filters for unicast MACs */
+	unsigned int tx_tcnt;
+	unsigned int tx_xcnt;
+	u64 tx_mcnt;
+	unsigned int rx_xcnt;
+	u64 rx_mcnt;
+	unsigned int toggle_cnt;
+	unsigned int txen;
 	struct mac_stats stats;
 };
 
@@ -588,7 +608,7 @@ static inline int is_10G(const struct adapter *adap)
 
 static inline int is_offload(const struct adapter *adap)
 {
-	return adapter_info(adap)->caps & SUPPORTED_OFFLOAD;
+	return adap->params.offload;
 }
 
 static inline unsigned int core_ticks_per_usec(const struct adapter *adap)
@@ -666,6 +686,7 @@ int t3_mac_set_address(struct cmac *mac, unsigned int idx, u8 addr[6]);
 int t3_mac_set_num_ucast(struct cmac *mac, int n);
 const struct mac_stats *t3_mac_update_stats(struct cmac *mac);
 int t3_mac_set_speed_duplex_fc(struct cmac *mac, int speed, int duplex, int fc);
+int t3b2_mac_watchdog_task(struct cmac *mac);
 
 void t3_mc5_prep(struct adapter *adapter, struct mc5 *mc5, int mode);
 int t3_mc5_init(struct mc5 *mc5, unsigned int nservers, unsigned int nfilters,
diff --git a/drivers/net/cxgb3/cxgb3_defs.h b/drivers/net/cxgb3/cxgb3_defs.h
index 16e004990c59..483a594210a7 100644
--- a/drivers/net/cxgb3/cxgb3_defs.h
+++ b/drivers/net/cxgb3/cxgb3_defs.h
@@ -1,6 +1,5 @@
 /*
  * Copyright (c) 2006-2007 Chelsio, Inc. All rights reserved.
- * Copyright (c) 2006-2007 Open Grid Computing, Inc. All rights reserved.
  *
  * This software is available to you under a choice of one of two
  * licenses.  You may choose to be licensed under the terms of the GNU
@@ -68,7 +67,10 @@ static inline union listen_entry *stid2entry(const struct tid_info *t,
 static inline struct t3c_tid_entry *lookup_tid(const struct tid_info *t,
 					       unsigned int tid)
 {
-	return tid < t->ntids ? &(t->tid_tab[tid]) : NULL;
+	struct t3c_tid_entry *t3c_tid = tid < t->ntids ?
+	    &(t->tid_tab[tid]) : NULL;
+
+	return (t3c_tid && t3c_tid->client) ? t3c_tid : NULL;
 }
 
 /*
diff --git a/drivers/net/cxgb3/cxgb3_ioctl.h b/drivers/net/cxgb3/cxgb3_ioctl.h
index a94281861a66..0a82fcddf2d8 100644
--- a/drivers/net/cxgb3/cxgb3_ioctl.h
+++ b/drivers/net/cxgb3/cxgb3_ioctl.h
@@ -36,28 +36,17 @@
  * Ioctl commands specific to this driver.
  */
 enum {
-	CHELSIO_SETREG = 1024,
-	CHELSIO_GETREG,
-	CHELSIO_SETTPI,
-	CHELSIO_GETTPI,
-	CHELSIO_GETMTUTAB,
-	CHELSIO_SETMTUTAB,
-	CHELSIO_GETMTU,
-	CHELSIO_SET_PM,
-	CHELSIO_GET_PM,
-	CHELSIO_GET_TCAM,
-	CHELSIO_SET_TCAM,
-	CHELSIO_GET_TCB,
-	CHELSIO_GET_MEM,
-	CHELSIO_LOAD_FW,
-	CHELSIO_GET_PROTO,
-	CHELSIO_SET_PROTO,
-	CHELSIO_SET_TRACE_FILTER,
-	CHELSIO_SET_QSET_PARAMS,
-	CHELSIO_GET_QSET_PARAMS,
-	CHELSIO_SET_QSET_NUM,
-	CHELSIO_GET_QSET_NUM,
-	CHELSIO_SET_PKTSCHED,
+	CHELSIO_GETMTUTAB 		= 1029,
+	CHELSIO_SETMTUTAB 		= 1030,
+	CHELSIO_SET_PM 			= 1032,
+	CHELSIO_GET_PM			= 1033,
+	CHELSIO_GET_MEM			= 1038,
+	CHELSIO_LOAD_FW			= 1041,
+	CHELSIO_SET_TRACE_FILTER	= 1044,
+	CHELSIO_SET_QSET_PARAMS		= 1045,
+	CHELSIO_GET_QSET_PARAMS		= 1046,
+	CHELSIO_SET_QSET_NUM		= 1047,
+	CHELSIO_GET_QSET_NUM		= 1048,
 };
 
 struct ch_reg {
diff --git a/drivers/net/cxgb3/cxgb3_main.c b/drivers/net/cxgb3/cxgb3_main.c
index 43583ed655ab..67b4b219d927 100644
--- a/drivers/net/cxgb3/cxgb3_main.c
+++ b/drivers/net/cxgb3/cxgb3_main.c
@@ -42,6 +42,7 @@
 #include <linux/workqueue.h>
 #include <linux/proc_fs.h>
 #include <linux/rtnetlink.h>
+#include <linux/firmware.h>
 #include <asm/uaccess.h>
 
 #include "common.h"
@@ -184,16 +185,24 @@ void t3_os_link_changed(struct adapter *adapter, int port_id, int link_stat,
 			int speed, int duplex, int pause)
 {
 	struct net_device *dev = adapter->port[port_id];
+	struct port_info *pi = netdev_priv(dev);
+	struct cmac *mac = &pi->mac;
 
 	/* Skip changes from disabled ports. */
 	if (!netif_running(dev))
 		return;
 
 	if (link_stat != netif_carrier_ok(dev)) {
-		if (link_stat)
+		if (link_stat) {
+			t3_mac_enable(mac, MAC_DIRECTION_RX);
 			netif_carrier_on(dev);
-		else
+		} else {
 			netif_carrier_off(dev);
+			pi->phy.ops->power_down(&pi->phy, 1);
+			t3_mac_disable(mac, MAC_DIRECTION_RX);
+			t3_link_start(&pi->phy, mac, &pi->link_config);
+		}
+
 		link_report(dev);
 	}
 }
@@ -406,7 +415,7 @@ static void quiesce_rx(struct adapter *adap)
 static int setup_sge_qsets(struct adapter *adap)
 {
 	int i, j, err, irq_idx = 0, qset_idx = 0, dummy_dev_idx = 0;
-	unsigned int ntxq = is_offload(adap) ? SGE_TXQ_PER_SET : 1;
+	unsigned int ntxq = SGE_TXQ_PER_SET;
 
 	if (adap->params.rev > 0 && !(adap->flags & USING_MSI))
 		irq_idx = -1;
@@ -434,27 +443,25 @@ static int setup_sge_qsets(struct adapter *adap)
 
 static ssize_t attr_show(struct device *d, struct device_attribute *attr,
 			 char *buf,
-			 ssize_t(*format) (struct adapter *, char *))
+			 ssize_t(*format) (struct net_device *, char *))
 {
 	ssize_t len;
-	struct adapter *adap = to_net_dev(d)->priv;
 
 	/* Synchronize with ioctls that may shut down the device */
 	rtnl_lock();
-	len = (*format) (adap, buf);
+	len = (*format) (to_net_dev(d), buf);
 	rtnl_unlock();
 	return len;
 }
 
 static ssize_t attr_store(struct device *d, struct device_attribute *attr,
 			  const char *buf, size_t len,
-			  ssize_t(*set) (struct adapter *, unsigned int),
+			  ssize_t(*set) (struct net_device *, unsigned int),
 			  unsigned int min_val, unsigned int max_val)
 {
 	char *endp;
 	ssize_t ret;
 	unsigned int val;
-	struct adapter *adap = to_net_dev(d)->priv;
 
 	if (!capable(CAP_NET_ADMIN))
 		return -EPERM;
@@ -464,7 +471,7 @@ static ssize_t attr_store(struct device *d, struct device_attribute *attr,
 		return -EINVAL;
 
 	rtnl_lock();
-	ret = (*set) (adap, val);
+	ret = (*set) (to_net_dev(d), val);
 	if (!ret)
 		ret = len;
 	rtnl_unlock();
@@ -472,8 +479,9 @@ static ssize_t attr_store(struct device *d, struct device_attribute *attr,
 }
 
 #define CXGB3_SHOW(name, val_expr) \
-static ssize_t format_##name(struct adapter *adap, char *buf) \
+static ssize_t format_##name(struct net_device *dev, char *buf) \
 { \
+	struct adapter *adap = dev->priv; \
 	return sprintf(buf, "%u\n", val_expr); \
 } \
 static ssize_t show_##name(struct device *d, struct device_attribute *attr, \
@@ -482,13 +490,17 @@ static ssize_t show_##name(struct device *d, struct device_attribute *attr, \
 	return attr_show(d, attr, buf, format_##name); \
 }
 
-static ssize_t set_nfilters(struct adapter *adap, unsigned int val)
+static ssize_t set_nfilters(struct net_device *dev, unsigned int val)
 {
+	struct adapter *adap = dev->priv;
+	int min_tids = is_offload(adap) ? MC5_MIN_TIDS : 0;
+
 	if (adap->flags & FULL_INIT_DONE)
 		return -EBUSY;
 	if (val && adap->params.rev == 0)
 		return -EINVAL;
-	if (val > t3_mc5_size(&adap->mc5) - adap->params.mc5.nservers)
+	if (val > t3_mc5_size(&adap->mc5) - adap->params.mc5.nservers -
+	    min_tids)
 		return -EINVAL;
 	adap->params.mc5.nfilters = val;
 	return 0;
@@ -500,11 +512,14 @@ static ssize_t store_nfilters(struct device *d, struct device_attribute *attr,
 	return attr_store(d, attr, buf, len, set_nfilters, 0, ~0);
 }
 
-static ssize_t set_nservers(struct adapter *adap, unsigned int val)
+static ssize_t set_nservers(struct net_device *dev, unsigned int val)
 {
+	struct adapter *adap = dev->priv;
+
 	if (adap->flags & FULL_INIT_DONE)
 		return -EBUSY;
-	if (val > t3_mc5_size(&adap->mc5) - adap->params.mc5.nfilters)
+	if (val > t3_mc5_size(&adap->mc5) - adap->params.mc5.nfilters -
+	    MC5_MIN_TIDS)
 		return -EINVAL;
 	adap->params.mc5.nservers = val;
 	return 0;
@@ -704,6 +719,28 @@ static void bind_qsets(struct adapter *adap)
 	}
 }
 
+#define FW_FNAME "t3fw-%d.%d.%d.bin"
+
+static int upgrade_fw(struct adapter *adap)
+{
+	int ret;
+	char buf[64];
+	const struct firmware *fw;
+	struct device *dev = &adap->pdev->dev;
+
+	snprintf(buf, sizeof(buf), FW_FNAME, FW_VERSION_MAJOR,
+		 FW_VERSION_MINOR, FW_VERSION_MICRO);
+	ret = request_firmware(&fw, buf, dev);
+	if (ret < 0) {
+		dev_err(dev, "could not upgrade firmware: unable to load %s\n",
+			buf);
+		return ret;
+	}
+	ret = t3_load_fw(adap, fw->data, fw->size);
+	release_firmware(fw);
+	return ret;
+}
+
 /**
  *	cxgb_up - enable the adapter
  *	@adapter: adapter being enabled
@@ -720,6 +757,8 @@ static int cxgb_up(struct adapter *adap)
 
 	if (!(adap->flags & FULL_INIT_DONE)) {
 		err = t3_check_fw_version(adap);
+		if (err == -EINVAL)
+			err = upgrade_fw(adap);
 		if (err)
 			goto out;
 
@@ -731,6 +770,8 @@ static int cxgb_up(struct adapter *adap)
 		if (err)
 			goto out;
 
+		t3_write_reg(adap, A_ULPRX_TDDP_PSZ, V_HPZ0(PAGE_SHIFT - 12));
+		
 		err = setup_sge_qsets(adap);
 		if (err)
 			goto out;
@@ -891,7 +932,7 @@ static int cxgb_open(struct net_device *dev)
 		return err;
 
 	set_bit(pi->port_id, &adapter->open_device_map);
-	if (!ofld_disable) {
+	if (is_offload(adapter) && !ofld_disable) {
 		err = offload_open(dev);
 		if (err)
 			printk(KERN_WARNING
@@ -1028,7 +1069,11 @@ static char stats_strings[][ETH_GSTRING_LEN] = {
 	"VLANinsertions     ",
 	"TxCsumOffload      ",
 	"RxCsumGood         ",
-	"RxDrops            "
+	"RxDrops            ",
+
+	"CheckTXEnToggled   ",
+	"CheckResets        ",
+
 };
 
 static int get_stats_count(struct net_device *dev)
@@ -1142,6 +1187,9 @@ static void get_stats(struct net_device *dev, struct ethtool_stats *stats,
 	*data++ = collect_sge_port_stats(adapter, pi, SGE_PSTAT_TX_CSUM);
 	*data++ = collect_sge_port_stats(adapter, pi, SGE_PSTAT_RX_CSUM_GOOD);
 	*data++ = s->rx_cong_drops;
+
+	*data++ = s->num_toggled;
+	*data++ = s->num_resets;
 }
 
 static inline void reg_block_dump(struct adapter *ap, void *buf,
@@ -1359,23 +1407,27 @@ static int set_rx_csum(struct net_device *dev, u32 data)
 
 static void get_sge_param(struct net_device *dev, struct ethtool_ringparam *e)
 {
-	struct adapter *adapter = dev->priv;
+	const struct adapter *adapter = dev->priv;
+	const struct port_info *pi = netdev_priv(dev);
+	const struct qset_params *q = &adapter->params.sge.qset[pi->first_qset];
 
 	e->rx_max_pending = MAX_RX_BUFFERS;
 	e->rx_mini_max_pending = 0;
 	e->rx_jumbo_max_pending = MAX_RX_JUMBO_BUFFERS;
 	e->tx_max_pending = MAX_TXQ_ENTRIES;
 
-	e->rx_pending = adapter->params.sge.qset[0].fl_size;
-	e->rx_mini_pending = adapter->params.sge.qset[0].rspq_size;
-	e->rx_jumbo_pending = adapter->params.sge.qset[0].jumbo_size;
-	e->tx_pending = adapter->params.sge.qset[0].txq_size[0];
+	e->rx_pending = q->fl_size;
+	e->rx_mini_pending = q->rspq_size;
+	e->rx_jumbo_pending = q->jumbo_size;
+	e->tx_pending = q->txq_size[0];
 }
 
 static int set_sge_param(struct net_device *dev, struct ethtool_ringparam *e)
 {
 	int i;
+	struct qset_params *q;
 	struct adapter *adapter = dev->priv;
+	const struct port_info *pi = netdev_priv(dev);
 
 	if (e->rx_pending > MAX_RX_BUFFERS ||
 	    e->rx_jumbo_pending > MAX_RX_JUMBO_BUFFERS ||
@@ -1390,9 +1442,8 @@ static int set_sge_param(struct net_device *dev, struct ethtool_ringparam *e)
 	if (adapter->flags & FULL_INIT_DONE)
 		return -EBUSY;
 
-	for (i = 0; i < SGE_QSETS; ++i) {
-		struct qset_params *q = &adapter->params.sge.qset[i];
-
+	q = &adapter->params.sge.qset[pi->first_qset];
+	for (i = 0; i < pi->nqsets; ++i, ++q) {
 		q->rspq_size = e->rx_mini_pending;
 		q->fl_size = e->rx_pending;
 		q->jumbo_size = e->rx_jumbo_pending;
@@ -1549,32 +1600,6 @@ static int cxgb_extension_ioctl(struct net_device *dev, void __user *useraddr)
 		return -EFAULT;
 
 	switch (cmd) {
-	case CHELSIO_SETREG:{
-		struct ch_reg edata;
-
-		if (!capable(CAP_NET_ADMIN))
-			return -EPERM;
-		if (copy_from_user(&edata, useraddr, sizeof(edata)))
-			return -EFAULT;
-		if ((edata.addr & 3) != 0
-			|| edata.addr >= adapter->mmio_len)
-			return -EINVAL;
-		writel(edata.val, adapter->regs + edata.addr);
-		break;
-	}
-	case CHELSIO_GETREG:{
-		struct ch_reg edata;
-
-		if (copy_from_user(&edata, useraddr, sizeof(edata)))
-			return -EFAULT;
-		if ((edata.addr & 3) != 0
-			|| edata.addr >= adapter->mmio_len)
-			return -EINVAL;
-		edata.val = readl(adapter->regs + edata.addr);
-		if (copy_to_user(useraddr, &edata, sizeof(edata)))
-			return -EFAULT;
-		break;
-	}
 	case CHELSIO_SET_QSET_PARAMS:{
 		int i;
 		struct qset_params *q;
@@ -1838,10 +1863,10 @@ static int cxgb_extension_ioctl(struct net_device *dev, void __user *useraddr)
 			return -EINVAL;
 
 		/*
-			* Version scheme:
-			* bits 0..9: chip version
-			* bits 10..15: chip revision
-			*/
+		 * Version scheme:
+		 * bits 0..9: chip version
+		 * bits 10..15: chip revision
+		 */
 		t.version = 3 | (adapter->params.rev << 10);
 		if (copy_to_user(useraddr, &t, sizeof(t)))
 			return -EFAULT;
@@ -1890,20 +1915,6 @@ static int cxgb_extension_ioctl(struct net_device *dev, void __user *useraddr)
 						t.trace_rx);
 		break;
 	}
-	case CHELSIO_SET_PKTSCHED:{
-		struct ch_pktsched_params p;
-
-		if (!capable(CAP_NET_ADMIN))
-				return -EPERM;
-		if (!adapter->open_device_map)
-				return -EAGAIN;	/* uP and SGE must be running */
-		if (copy_from_user(&p, useraddr, sizeof(p)))
-				return -EFAULT;
-		send_pktsched_cmd(adapter, p.sched, p.idx, p.min, p.max,
-				  p.binding);
-		break;
-			
-	}
 	default:
 		return -EOPNOTSUPP;
 	}
@@ -2104,6 +2115,42 @@ static void check_link_status(struct adapter *adapter)
 	}
 }
 
+static void check_t3b2_mac(struct adapter *adapter)
+{
+	int i;
+
+	if (!rtnl_trylock())	/* synchronize with ifdown */
+		return;
+
+	for_each_port(adapter, i) {
+		struct net_device *dev = adapter->port[i];
+		struct port_info *p = netdev_priv(dev);
+		int status;
+
+		if (!netif_running(dev))
+			continue;
+
+		status = 0;
+		if (netif_running(dev) && netif_carrier_ok(dev))
+			status = t3b2_mac_watchdog_task(&p->mac);
+		if (status == 1)
+			p->mac.stats.num_toggled++;
+		else if (status == 2) {
+			struct cmac *mac = &p->mac;
+
+			t3_mac_set_mtu(mac, dev->mtu);
+			t3_mac_set_address(mac, 0, dev->dev_addr);
+			cxgb_set_rxmode(dev);
+			t3_link_start(&p->phy, mac, &p->link_config);
+			t3_mac_enable(mac, MAC_DIRECTION_RX | MAC_DIRECTION_TX);
+			t3_port_intr_enable(adapter, p->port_id);
+			p->mac.stats.num_resets++;
+		}
+	}
+	rtnl_unlock();
+}
+
+
 static void t3_adap_check_task(struct work_struct *work)
 {
 	struct adapter *adapter = container_of(work, struct adapter,
@@ -2124,6 +2171,9 @@ static void t3_adap_check_task(struct work_struct *work)
 		adapter->check_task_cnt = 0;
 	}
 
+	if (p->rev == T3_REV_B2)
+		check_t3b2_mac(adapter);
+
 	/* Schedule the next check update if any port is active. */
 	spin_lock(&adapter->work_lock);
 	if (adapter->open_device_map & PORT_MASK)
@@ -2232,9 +2282,9 @@ static void __devinit print_port_info(struct adapter *adap,
 
 		if (!test_bit(i, &adap->registered_device_map))
 			continue;
-		printk(KERN_INFO "%s: %s %s RNIC (rev %d) %s%s\n",
+		printk(KERN_INFO "%s: %s %s %sNIC (rev %d) %s%s\n",
 		       dev->name, ai->desc, pi->port_type->desc,
-		       adap->params.rev, buf,
+		       is_offload(adap) ? "R" : "", adap->params.rev, buf,
 		       (adap->flags & USING_MSIX) ? " MSI-X" :
 		       (adap->flags & USING_MSI) ? " MSI" : "");
 		if (adap->name == dev->name && adap->params.vpd.mclk)
diff --git a/drivers/net/cxgb3/cxgb3_offload.c b/drivers/net/cxgb3/cxgb3_offload.c
index c6b726643185..ebcf35e4cf5b 100644
--- a/drivers/net/cxgb3/cxgb3_offload.c
+++ b/drivers/net/cxgb3/cxgb3_offload.c
@@ -1,6 +1,5 @@
 /*
  * Copyright (c) 2006-2007 Chelsio, Inc. All rights reserved.
- * Copyright (c) 2006-2007 Open Grid Computing, Inc. All rights reserved.
  *
  * This software is available to you under a choice of one of two
  * licenses.  You may choose to be licensed under the terms of the GNU
@@ -161,14 +160,16 @@ static struct net_device *get_iff_from_mac(struct adapter *adapter,
 	int i;
 
 	for_each_port(adapter, i) {
-		const struct vlan_group *grp;
+		struct vlan_group *grp;
 		struct net_device *dev = adapter->port[i];
 		const struct port_info *p = netdev_priv(dev);
 
 		if (!memcmp(dev->dev_addr, mac, ETH_ALEN)) {
 			if (vlan && vlan != VLAN_VID_MASK) {
 				grp = p->vlan_grp;
-				dev = grp ? grp->vlan_devices[vlan] : NULL;
+				dev = NULL;
+				if (grp)
+					dev = vlan_group_get_device(grp, vlan);
 			} else
 				while (dev->master)
 					dev = dev->master;
@@ -507,6 +508,7 @@ void cxgb3_queue_tid_release(struct t3cdev *tdev, unsigned int tid)
 
 	spin_lock_bh(&td->tid_release_lock);
 	p->ctx = (void *)td->tid_release_list;
+	p->client = NULL;
 	td->tid_release_list = p;
 	if (!p->ctx)
 		schedule_work(&td->tid_release_task);
@@ -552,7 +554,9 @@ int cxgb3_alloc_atid(struct t3cdev *tdev, struct cxgb3_client *client,
 	struct tid_info *t = &(T3C_DATA(tdev))->tid_maps;
 
 	spin_lock_bh(&t->atid_lock);
-	if (t->afree) {
+	if (t->afree &&
+	    t->atids_in_use + atomic_read(&t->tids_in_use) + MC5_MIN_TIDS <=
+	    t->ntids) {
 		union active_open_entry *p = t->afree;
 
 		atid = (p - t->atid_tab) + t->atid_base;
@@ -620,7 +624,8 @@ static int do_act_open_rpl(struct t3cdev *dev, struct sk_buff *skb)
 	struct t3c_tid_entry *t3c_tid;
 
 	t3c_tid = lookup_atid(&(T3C_DATA(dev))->tid_maps, atid);
-	if (t3c_tid->ctx && t3c_tid->client && t3c_tid->client->handlers &&
+	if (t3c_tid && t3c_tid->ctx && t3c_tid->client &&
+	    t3c_tid->client->handlers &&
 	    t3c_tid->client->handlers[CPL_ACT_OPEN_RPL]) {
 		return t3c_tid->client->handlers[CPL_ACT_OPEN_RPL] (dev, skb,
 								    t3c_tid->
@@ -639,7 +644,7 @@ static int do_stid_rpl(struct t3cdev *dev, struct sk_buff *skb)
 	struct t3c_tid_entry *t3c_tid;
 
 	t3c_tid = lookup_stid(&(T3C_DATA(dev))->tid_maps, stid);
-	if (t3c_tid->ctx && t3c_tid->client->handlers &&
+	if (t3c_tid && t3c_tid->ctx && t3c_tid->client->handlers &&
 	    t3c_tid->client->handlers[p->opcode]) {
 		return t3c_tid->client->handlers[p->opcode] (dev, skb,
 							     t3c_tid->ctx);
@@ -657,7 +662,7 @@ static int do_hwtid_rpl(struct t3cdev *dev, struct sk_buff *skb)
 	struct t3c_tid_entry *t3c_tid;
 
 	t3c_tid = lookup_tid(&(T3C_DATA(dev))->tid_maps, hwtid);
-	if (t3c_tid->ctx && t3c_tid->client->handlers &&
+	if (t3c_tid && t3c_tid->ctx && t3c_tid->client->handlers &&
 	    t3c_tid->client->handlers[p->opcode]) {
 		return t3c_tid->client->handlers[p->opcode]
 		    (dev, skb, t3c_tid->ctx);
@@ -686,6 +691,28 @@ static int do_cr(struct t3cdev *dev, struct sk_buff *skb)
 	}
 }
 
+/*
+ * Returns an sk_buff for a reply CPL message of size len.  If the input
+ * sk_buff has no other users it is trimmed and reused, otherwise a new buffer
+ * is allocated.  The input skb must be of size at least len.  Note that this
+ * operation does not destroy the original skb data even if it decides to reuse
+ * the buffer.
+ */
+static struct sk_buff *cxgb3_get_cpl_reply_skb(struct sk_buff *skb, size_t len,
+					       int gfp)
+{
+	if (likely(!skb_cloned(skb))) {
+		BUG_ON(skb->len < len);
+		__skb_trim(skb, len);
+		skb_get(skb);
+	} else {
+		skb = alloc_skb(len, gfp);
+		if (skb)
+			__skb_put(skb, len);
+	}
+	return skb;
+}
+
 static int do_abort_req_rss(struct t3cdev *dev, struct sk_buff *skb)
 {
 	union opcode_tid *p = cplhdr(skb);
@@ -693,30 +720,39 @@ static int do_abort_req_rss(struct t3cdev *dev, struct sk_buff *skb)
 	struct t3c_tid_entry *t3c_tid;
 
 	t3c_tid = lookup_tid(&(T3C_DATA(dev))->tid_maps, hwtid);
-	if (t3c_tid->ctx && t3c_tid->client->handlers &&
+	if (t3c_tid && t3c_tid->ctx && t3c_tid->client->handlers &&
 	    t3c_tid->client->handlers[p->opcode]) {
 		return t3c_tid->client->handlers[p->opcode]
 		    (dev, skb, t3c_tid->ctx);
 	} else {
 		struct cpl_abort_req_rss *req = cplhdr(skb);
 		struct cpl_abort_rpl *rpl;
+		struct sk_buff *reply_skb;
+		unsigned int tid = GET_TID(req);
+		u8 cmd = req->status;
+
+		if (req->status == CPL_ERR_RTX_NEG_ADVICE ||
+		    req->status == CPL_ERR_PERSIST_NEG_ADVICE)
+			goto out;
 
-		struct sk_buff *skb =
-		    alloc_skb(sizeof(struct cpl_abort_rpl), GFP_ATOMIC);
-		if (!skb) {
+		reply_skb = cxgb3_get_cpl_reply_skb(skb,
+						    sizeof(struct
+							   cpl_abort_rpl),
+						    GFP_ATOMIC);
+
+		if (!reply_skb) {
 			printk("do_abort_req_rss: couldn't get skb!\n");
 			goto out;
 		}
-		skb->priority = CPL_PRIORITY_DATA;
-		__skb_put(skb, sizeof(struct cpl_abort_rpl));
-		rpl = cplhdr(skb);
+		reply_skb->priority = CPL_PRIORITY_DATA;
+		__skb_put(reply_skb, sizeof(struct cpl_abort_rpl));
+		rpl = cplhdr(reply_skb);
 		rpl->wr.wr_hi =
 		    htonl(V_WR_OP(FW_WROPCODE_OFLD_HOST_ABORT_CON_RPL));
-		rpl->wr.wr_lo = htonl(V_WR_TID(GET_TID(req)));
-		OPCODE_TID(rpl) =
-		    htonl(MK_OPCODE_TID(CPL_ABORT_RPL, GET_TID(req)));
-		rpl->cmd = req->status;
-		cxgb3_ofld_send(dev, skb);
+		rpl->wr.wr_lo = htonl(V_WR_TID(tid));
+		OPCODE_TID(rpl) = htonl(MK_OPCODE_TID(CPL_ABORT_RPL, tid));
+		rpl->cmd = cmd;
+		cxgb3_ofld_send(dev, reply_skb);
 out:
 		return CPL_RET_BUF_DONE;
 	}
@@ -729,7 +765,7 @@ static int do_act_establish(struct t3cdev *dev, struct sk_buff *skb)
 	struct t3c_tid_entry *t3c_tid;
 
 	t3c_tid = lookup_atid(&(T3C_DATA(dev))->tid_maps, atid);
-	if (t3c_tid->ctx && t3c_tid->client->handlers &&
+	if (t3c_tid && t3c_tid->ctx && t3c_tid->client->handlers &&
 	    t3c_tid->client->handlers[CPL_ACT_ESTABLISH]) {
 		return t3c_tid->client->handlers[CPL_ACT_ESTABLISH]
 		    (dev, skb, t3c_tid->ctx);
@@ -740,17 +776,6 @@ static int do_act_establish(struct t3cdev *dev, struct sk_buff *skb)
 	}
 }
 
-static int do_set_tcb_rpl(struct t3cdev *dev, struct sk_buff *skb)
-{
-	struct cpl_set_tcb_rpl *rpl = cplhdr(skb);
-
-	if (rpl->status != CPL_ERR_NONE)
-		printk(KERN_ERR
-		       "Unexpected SET_TCB_RPL status %u for tid %u\n",
-		       rpl->status, GET_TID(rpl));
-	return CPL_RET_BUF_DONE;
-}
-
 static int do_trace(struct t3cdev *dev, struct sk_buff *skb)
 {
 	struct cpl_trace_pkt *p = cplhdr(skb);
@@ -758,7 +783,7 @@ static int do_trace(struct t3cdev *dev, struct sk_buff *skb)
 	skb->protocol = htons(0xffff);
 	skb->dev = dev->lldev;
 	skb_pull(skb, sizeof(*p));
-	skb->mac.raw = skb->data;
+	skb_reset_mac_header(skb);
 	netif_receive_skb(skb);
 	return 0;
 }
@@ -770,7 +795,7 @@ static int do_term(struct t3cdev *dev, struct sk_buff *skb)
 	struct t3c_tid_entry *t3c_tid;
 
 	t3c_tid = lookup_tid(&(T3C_DATA(dev))->tid_maps, hwtid);
-	if (t3c_tid->ctx && t3c_tid->client->handlers &&
+	if (t3c_tid && t3c_tid->ctx && t3c_tid->client->handlers &&
 	    t3c_tid->client->handlers[opcode]) {
 		return t3c_tid->client->handlers[opcode] (dev, skb,
 							  t3c_tid->ctx);
@@ -969,7 +994,7 @@ void cxgb_redirect(struct dst_entry *old, struct dst_entry *new)
 	for (tid = 0; tid < ti->ntids; tid++) {
 		te = lookup_tid(ti, tid);
 		BUG_ON(!te);
-		if (te->ctx && te->client && te->client->redirect) {
+		if (te && te->ctx && te->client && te->client->redirect) {
 			update_tcb = te->client->redirect(te->ctx, old, new, e);
 			if (update_tcb) {
 				l2t_hold(L2DATA(tdev), e);
@@ -1212,7 +1237,8 @@ void __init cxgb3_offload_init(void)
 	t3_register_cpl_handler(CPL_CLOSE_CON_RPL, do_hwtid_rpl);
 	t3_register_cpl_handler(CPL_ABORT_REQ_RSS, do_abort_req_rss);
 	t3_register_cpl_handler(CPL_ACT_ESTABLISH, do_act_establish);
-	t3_register_cpl_handler(CPL_SET_TCB_RPL, do_set_tcb_rpl);
+	t3_register_cpl_handler(CPL_SET_TCB_RPL, do_hwtid_rpl);
+	t3_register_cpl_handler(CPL_GET_TCB_RPL, do_hwtid_rpl);
 	t3_register_cpl_handler(CPL_RDMA_TERMINATE, do_term);
 	t3_register_cpl_handler(CPL_RDMA_EC_STATUS, do_hwtid_rpl);
 	t3_register_cpl_handler(CPL_TRACE_PKT, do_trace);
diff --git a/drivers/net/cxgb3/cxgb3_offload.h b/drivers/net/cxgb3/cxgb3_offload.h
index 0e6beb69ba17..f15446a32efc 100644
--- a/drivers/net/cxgb3/cxgb3_offload.h
+++ b/drivers/net/cxgb3/cxgb3_offload.h
@@ -1,6 +1,5 @@
 /*
  * Copyright (c) 2006-2007 Chelsio, Inc. All rights reserved.
- * Copyright (c) 2006-2007 Open Grid Computing, Inc. All rights reserved.
  *
  * This software is available to you under a choice of one of two
  * licenses.  You may choose to be licensed under the terms of the GNU
diff --git a/drivers/net/cxgb3/l2t.c b/drivers/net/cxgb3/l2t.c
index 3c0cb8557058..d660af74606e 100644
--- a/drivers/net/cxgb3/l2t.c
+++ b/drivers/net/cxgb3/l2t.c
@@ -1,6 +1,5 @@
 /*
  * Copyright (c) 2003-2007 Chelsio, Inc. All rights reserved.
- * Copyright (c) 2006-2007 Open Grid Computing, Inc. All rights reserved.
  *
  * This software is available to you under a choice of one of two
  * licenses.  You may choose to be licensed under the terms of the GNU
diff --git a/drivers/net/cxgb3/l2t.h b/drivers/net/cxgb3/l2t.h
index ba5d2cbd7241..d79001336cfd 100644
--- a/drivers/net/cxgb3/l2t.h
+++ b/drivers/net/cxgb3/l2t.h
@@ -1,6 +1,5 @@
 /*
  * Copyright (c) 2003-2007 Chelsio, Inc. All rights reserved.
- * Copyright (c) 2006-2007 Open Grid Computing, Inc. All rights reserved.
  *
  * This software is available to you under a choice of one of two
  * licenses.  You may choose to be licensed under the terms of the GNU
diff --git a/drivers/net/cxgb3/mc5.c b/drivers/net/cxgb3/mc5.c
index 644d62ea86a6..84c1ffa8e2d3 100644
--- a/drivers/net/cxgb3/mc5.c
+++ b/drivers/net/cxgb3/mc5.c
@@ -328,6 +328,9 @@ int t3_mc5_init(struct mc5 *mc5, unsigned int nservers, unsigned int nfilters,
 	unsigned int tcam_size = mc5->tcam_size;
 	struct adapter *adap = mc5->adapter;
 
+	if (!tcam_size)
+		return 0;
+
 	if (nroutes > MAX_ROUTES || nroutes + nservers + nfilters > tcam_size)
 		return -EINVAL;
 
diff --git a/drivers/net/cxgb3/regs.h b/drivers/net/cxgb3/regs.h
index b56c5f52bcdc..e5a553410e24 100644
--- a/drivers/net/cxgb3/regs.h
+++ b/drivers/net/cxgb3/regs.h
@@ -1206,6 +1206,14 @@
 
 #define A_TP_RX_TRC_KEY0 0x120
 
+#define A_TP_TX_DROP_CNT_CH0 0x12d
+
+#define S_TXDROPCNTCH0RCVD    0
+#define M_TXDROPCNTCH0RCVD    0xffff
+#define V_TXDROPCNTCH0RCVD(x) ((x) << S_TXDROPCNTCH0RCVD)
+#define G_TXDROPCNTCH0RCVD(x) (((x) >> S_TXDROPCNTCH0RCVD) & \
+			       M_TXDROPCNTCH0RCVD)
+
 #define A_ULPRX_CTL 0x500
 
 #define S_ROUND_ROBIN    4
@@ -1226,9 +1234,15 @@
 
 #define A_ULPRX_ISCSI_TAGMASK 0x514
 
+#define S_HPZ0    0
+#define M_HPZ0    0xf
+#define V_HPZ0(x) ((x) << S_HPZ0)
+#define G_HPZ0(x) (((x) >> S_HPZ0) & M_HPZ0)
+
 #define A_ULPRX_TDDP_LLIMIT 0x51c
 
 #define A_ULPRX_TDDP_ULIMIT 0x520
+#define A_ULPRX_TDDP_PSZ 0x528
 
 #define A_ULPRX_STAG_LLIMIT 0x52c
 
@@ -1834,6 +1848,8 @@
 #define V_TXPAUSEEN(x) ((x) << S_TXPAUSEEN)
 #define F_TXPAUSEEN    V_TXPAUSEEN(1U)
 
+#define A_XGM_TX_PAUSE_QUANTA 0x808
+
 #define A_XGM_RX_CTRL 0x80c
 
 #define S_RXEN    0
@@ -1920,11 +1936,20 @@
 
 #define A_XGM_TXFIFO_CFG 0x888
 
+#define S_TXIPG    13
+#define M_TXIPG    0xff
+#define V_TXIPG(x) ((x) << S_TXIPG)
+#define G_TXIPG(x) (((x) >> S_TXIPG) & M_TXIPG)
+
 #define S_TXFIFOTHRESH    4
 #define M_TXFIFOTHRESH    0x1ff
 
 #define V_TXFIFOTHRESH(x) ((x) << S_TXFIFOTHRESH)
 
+#define S_ENDROPPKT    21
+#define V_ENDROPPKT(x) ((x) << S_ENDROPPKT)
+#define F_ENDROPPKT    V_ENDROPPKT(1U)
+
 #define A_XGM_SERDES_CTRL 0x890
 #define A_XGM_SERDES_CTRL0 0x8e0
 
@@ -2190,6 +2215,13 @@
 
 #define A_XGM_RX_MAX_PKT_SIZE_ERR_CNT 0x9a4
 
+#define A_XGM_TX_SPI4_SOP_EOP_CNT 0x9a8
+
+#define S_TXSPI4SOPCNT    16
+#define M_TXSPI4SOPCNT    0xffff
+#define V_TXSPI4SOPCNT(x) ((x) << S_TXSPI4SOPCNT)
+#define G_TXSPI4SOPCNT(x) (((x) >> S_TXSPI4SOPCNT) & M_TXSPI4SOPCNT)
+
 #define A_XGM_RX_SPI4_SOP_EOP_CNT 0x9ac
 
 #define XGMAC0_1_BASE_ADDR 0xa00
diff --git a/drivers/net/cxgb3/sge.c b/drivers/net/cxgb3/sge.c
index 3f2cf8a07c61..3666586a4831 100644
--- a/drivers/net/cxgb3/sge.c
+++ b/drivers/net/cxgb3/sge.c
@@ -45,9 +45,25 @@
 #define USE_GTS 0
 
 #define SGE_RX_SM_BUF_SIZE 1536
+
+/*
+ * If USE_RX_PAGE is defined, the small freelist populated with (partial)
+ * pages instead of skbs. Pages are carved up into RX_PAGE_SIZE chunks (must
+ * be a multiple of the host page size).
+ */
+#define USE_RX_PAGE
+#define RX_PAGE_SIZE 2048
+
+/*
+ * skb freelist packets are copied into a new skb (and the freelist one is 
+ * reused) if their len is <= 
+ */
 #define SGE_RX_COPY_THRES  256
 
-# define SGE_RX_DROP_THRES 16
+/*
+ * Minimum number of freelist entries before we start dropping TUNNEL frames.
+ */
+#define SGE_RX_DROP_THRES 16
 
 /*
  * Period of the Tx buffer reclaim timer.  This timer does not need to run
@@ -85,7 +101,10 @@ struct tx_sw_desc {		/* SW state per Tx descriptor */
 };
 
 struct rx_sw_desc {		/* SW state per Rx descriptor */
-	struct sk_buff *skb;
+	union {
+		struct sk_buff *skb;
+		struct sge_fl_page page;
+	} t;
 	 DECLARE_PCI_UNMAP_ADDR(dma_addr);
 };
 
@@ -105,6 +124,15 @@ struct unmap_info {		/* packet unmapping info, overlays skb->cb */
 };
 
 /*
+ * Holds unmapping information for Tx packets that need deferred unmapping.
+ * This structure lives at skb->head and must be allocated by callers.
+ */
+struct deferred_unmap_info {
+	struct pci_dev *pdev;
+	dma_addr_t addr[MAX_SKB_FRAGS + 1];
+};
+
+/*
  * Maps a number of flits to the number of Tx descriptors that can hold them.
  * The formula is
  *
@@ -252,10 +280,13 @@ static void free_tx_desc(struct adapter *adapter, struct sge_txq *q,
 	struct pci_dev *pdev = adapter->pdev;
 	unsigned int cidx = q->cidx;
 
+	const int need_unmap = need_skb_unmap() &&
+			       q->cntxt_id >= FW_TUNNEL_SGEEC_START;
+
 	d = &q->sdesc[cidx];
 	while (n--) {
 		if (d->skb) {	/* an SGL is present */
-			if (need_skb_unmap())
+			if (need_unmap)
 				unmap_skb(d->skb, q, cidx, pdev);
 			if (d->skb->priority == cidx)
 				kfree_skb(d->skb);
@@ -320,16 +351,27 @@ static void free_rx_bufs(struct pci_dev *pdev, struct sge_fl *q)
 
 		pci_unmap_single(pdev, pci_unmap_addr(d, dma_addr),
 				 q->buf_size, PCI_DMA_FROMDEVICE);
-		kfree_skb(d->skb);
-		d->skb = NULL;
+
+		if (q->buf_size != RX_PAGE_SIZE) {
+			kfree_skb(d->t.skb);
+			d->t.skb = NULL;
+		} else {
+			if (d->t.page.frag.page)
+				put_page(d->t.page.frag.page);
+			d->t.page.frag.page = NULL;
+		}
 		if (++cidx == q->size)
 			cidx = 0;
 	}
+
+	if (q->page.frag.page)
+		put_page(q->page.frag.page);
+	q->page.frag.page = NULL;
 }
 
 /**
  *	add_one_rx_buf - add a packet buffer to a free-buffer list
- *	@skb: the buffer to add
+ *	@va: va of the buffer to add
  *	@len: the buffer length
  *	@d: the HW Rx descriptor to write
  *	@sd: the SW Rx descriptor to write
@@ -339,14 +381,13 @@ static void free_rx_bufs(struct pci_dev *pdev, struct sge_fl *q)
  *	Add a buffer of the given length to the supplied HW and SW Rx
  *	descriptors.
  */
-static inline void add_one_rx_buf(struct sk_buff *skb, unsigned int len,
+static inline void add_one_rx_buf(unsigned char *va, unsigned int len,
 				  struct rx_desc *d, struct rx_sw_desc *sd,
 				  unsigned int gen, struct pci_dev *pdev)
 {
 	dma_addr_t mapping;
 
-	sd->skb = skb;
-	mapping = pci_map_single(pdev, skb->data, len, PCI_DMA_FROMDEVICE);
+	mapping = pci_map_single(pdev, va, len, PCI_DMA_FROMDEVICE);
 	pci_unmap_addr_set(sd, dma_addr, mapping);
 
 	d->addr_lo = cpu_to_be32(mapping);
@@ -371,14 +412,47 @@ static void refill_fl(struct adapter *adap, struct sge_fl *q, int n, gfp_t gfp)
 {
 	struct rx_sw_desc *sd = &q->sdesc[q->pidx];
 	struct rx_desc *d = &q->desc[q->pidx];
+	struct sge_fl_page *p = &q->page;
 
 	while (n--) {
-		struct sk_buff *skb = alloc_skb(q->buf_size, gfp);
+		unsigned char *va;
 
-		if (!skb)
-			break;
+		if (unlikely(q->buf_size != RX_PAGE_SIZE)) {
+			struct sk_buff *skb = alloc_skb(q->buf_size, gfp);
+
+			if (!skb) {
+				q->alloc_failed++;
+				break;
+			}
+			va = skb->data;
+			sd->t.skb = skb;
+		} else {
+			if (!p->frag.page) {
+				p->frag.page = alloc_pages(gfp, 0);
+				if (unlikely(!p->frag.page)) {
+					q->alloc_failed++;
+					break;
+				} else {
+					p->frag.size = RX_PAGE_SIZE;
+					p->frag.page_offset = 0;
+					p->va = page_address(p->frag.page);
+				}
+			}
+
+			memcpy(&sd->t, p, sizeof(*p));
+			va = p->va;
+
+			p->frag.page_offset += RX_PAGE_SIZE;
+			BUG_ON(p->frag.page_offset > PAGE_SIZE);
+			p->va += RX_PAGE_SIZE;
+			if (p->frag.page_offset == PAGE_SIZE)
+				p->frag.page = NULL;
+			else
+				get_page(p->frag.page);
+		}
+
+		add_one_rx_buf(va, q->buf_size, d, sd, q->gen, adap->pdev);
 
-		add_one_rx_buf(skb, q->buf_size, d, sd, q->gen, adap->pdev);
 		d++;
 		sd++;
 		if (++q->pidx == q->size) {
@@ -413,7 +487,7 @@ static void recycle_rx_buf(struct adapter *adap, struct sge_fl *q,
 	struct rx_desc *from = &q->desc[idx];
 	struct rx_desc *to = &q->desc[q->pidx];
 
-	q->sdesc[q->pidx] = q->sdesc[idx];
+	memcpy(&q->sdesc[q->pidx], &q->sdesc[idx], sizeof(struct rx_sw_desc));
 	to->addr_lo = from->addr_lo;	/* already big endian */
 	to->addr_hi = from->addr_hi;	/* likewise */
 	wmb();
@@ -446,7 +520,7 @@ static void recycle_rx_buf(struct adapter *adap, struct sge_fl *q,
  *	of the SW ring.
  */
 static void *alloc_ring(struct pci_dev *pdev, size_t nelem, size_t elem_size,
-			size_t sw_size, dma_addr_t *phys, void *metadata)
+			size_t sw_size, dma_addr_t * phys, void *metadata)
 {
 	size_t len = nelem * elem_size;
 	void *s = NULL;
@@ -576,61 +650,6 @@ static inline unsigned int flits_to_desc(unsigned int n)
 }
 
 /**
- *	get_packet - return the next ingress packet buffer from a free list
- *	@adap: the adapter that received the packet
- *	@fl: the SGE free list holding the packet
- *	@len: the packet length including any SGE padding
- *	@drop_thres: # of remaining buffers before we start dropping packets
- *
- *	Get the next packet from a free list and complete setup of the
- *	sk_buff.  If the packet is small we make a copy and recycle the
- *	original buffer, otherwise we use the original buffer itself.  If a
- *	positive drop threshold is supplied packets are dropped and their
- *	buffers recycled if (a) the number of remaining buffers is under the
- *	threshold and the packet is too big to copy, or (b) the packet should
- *	be copied but there is no memory for the copy.
- */
-static struct sk_buff *get_packet(struct adapter *adap, struct sge_fl *fl,
-				  unsigned int len, unsigned int drop_thres)
-{
-	struct sk_buff *skb = NULL;
-	struct rx_sw_desc *sd = &fl->sdesc[fl->cidx];
-
-	prefetch(sd->skb->data);
-
-	if (len <= SGE_RX_COPY_THRES) {
-		skb = alloc_skb(len, GFP_ATOMIC);
-		if (likely(skb != NULL)) {
-			__skb_put(skb, len);
-			pci_dma_sync_single_for_cpu(adap->pdev,
-						    pci_unmap_addr(sd,
-								   dma_addr),
-						    len, PCI_DMA_FROMDEVICE);
-			memcpy(skb->data, sd->skb->data, len);
-			pci_dma_sync_single_for_device(adap->pdev,
-						       pci_unmap_addr(sd,
-								      dma_addr),
-						       len, PCI_DMA_FROMDEVICE);
-		} else if (!drop_thres)
-			goto use_orig_buf;
-	      recycle:
-		recycle_rx_buf(adap, fl, fl->cidx);
-		return skb;
-	}
-
-	if (unlikely(fl->credits < drop_thres))
-		goto recycle;
-
-      use_orig_buf:
-	pci_unmap_single(adap->pdev, pci_unmap_addr(sd, dma_addr),
-			 fl->buf_size, PCI_DMA_FROMDEVICE);
-	skb = sd->skb;
-	skb_put(skb, len);
-	__refill_fl(adap, fl);
-	return skb;
-}
-
-/**
  *	get_imm_packet - return the next ingress packet buffer from a response
  *	@resp: the response descriptor containing the packet data
  *
@@ -642,7 +661,7 @@ static inline struct sk_buff *get_imm_packet(const struct rsp_desc *resp)
 
 	if (skb) {
 		__skb_put(skb, IMMED_PKT_SIZE);
-		memcpy(skb->data, resp->imm_data, IMMED_PKT_SIZE);
+		skb_copy_to_linear_data(skb, resp->imm_data, IMMED_PKT_SIZE);
 	}
 	return skb;
 }
@@ -878,11 +897,11 @@ static void write_tx_pkt_wr(struct adapter *adap, struct sk_buff *skb,
 		d->flit[2] = 0;
 		cntrl |= V_TXPKT_OPCODE(CPL_TX_PKT_LSO);
 		hdr->cntrl = htonl(cntrl);
-		eth_type = skb->nh.raw - skb->data == ETH_HLEN ?
+		eth_type = skb_network_offset(skb) == ETH_HLEN ?
 		    CPL_ETH_II : CPL_ETH_II_VLAN;
 		tso_info |= V_LSO_ETH_TYPE(eth_type) |
-		    V_LSO_IPHDR_WORDS(skb->nh.iph->ihl) |
-		    V_LSO_TCPHDR_WORDS(skb->h.th->doff);
+		    V_LSO_IPHDR_WORDS(ip_hdr(skb)->ihl) |
+		    V_LSO_TCPHDR_WORDS(tcp_hdr(skb)->doff);
 		hdr->lso_info = htonl(tso_info);
 		flits = 3;
 	} else {
@@ -894,7 +913,8 @@ static void write_tx_pkt_wr(struct adapter *adap, struct sk_buff *skb,
 		if (skb->len <= WR_LEN - sizeof(*cpl)) {
 			q->sdesc[pidx].skb = NULL;
 			if (!skb->data_len)
-				memcpy(&d->flit[2], skb->data, skb->len);
+				skb_copy_from_linear_data(skb, &d->flit[2],
+							  skb->len);
 			else
 				skb_copy_bits(skb, 0, &d->flit[2], skb->len);
 
@@ -1227,6 +1247,50 @@ int t3_mgmt_tx(struct adapter *adap, struct sk_buff *skb)
 }
 
 /**
+ *	deferred_unmap_destructor - unmap a packet when it is freed
+ *	@skb: the packet
+ *
+ *	This is the packet destructor used for Tx packets that need to remain
+ *	mapped until they are freed rather than until their Tx descriptors are
+ *	freed.
+ */
+static void deferred_unmap_destructor(struct sk_buff *skb)
+{
+	int i;
+	const dma_addr_t *p;
+	const struct skb_shared_info *si;
+	const struct deferred_unmap_info *dui;
+	const struct unmap_info *ui = (struct unmap_info *)skb->cb;
+
+	dui = (struct deferred_unmap_info *)skb->head;
+	p = dui->addr;
+
+	if (ui->len)
+		pci_unmap_single(dui->pdev, *p++, ui->len, PCI_DMA_TODEVICE);
+
+	si = skb_shinfo(skb);
+	for (i = 0; i < si->nr_frags; i++)
+		pci_unmap_page(dui->pdev, *p++, si->frags[i].size,
+			       PCI_DMA_TODEVICE);
+}
+
+static void setup_deferred_unmapping(struct sk_buff *skb, struct pci_dev *pdev,
+				     const struct sg_ent *sgl, int sgl_flits)
+{
+	dma_addr_t *p;
+	struct deferred_unmap_info *dui;
+
+	dui = (struct deferred_unmap_info *)skb->head;
+	dui->pdev = pdev;
+	for (p = dui->addr; sgl_flits >= 3; sgl++, sgl_flits -= 3) {
+		*p++ = be64_to_cpu(sgl->addr[0]);
+		*p++ = be64_to_cpu(sgl->addr[1]);
+	}
+	if (sgl_flits)
+		*p = be64_to_cpu(sgl->addr[0]);
+}
+
+/**
  *	write_ofld_wr - write an offload work request
  *	@adap: the adapter
  *	@skb: the packet to send
@@ -1256,14 +1320,20 @@ static void write_ofld_wr(struct adapter *adap, struct sk_buff *skb,
 	/* Only TX_DATA builds SGLs */
 
 	from = (struct work_request_hdr *)skb->data;
-	memcpy(&d->flit[1], &from[1], skb->h.raw - skb->data - sizeof(*from));
+	memcpy(&d->flit[1], &from[1],
+	       skb_transport_offset(skb) - sizeof(*from));
 
-	flits = (skb->h.raw - skb->data) / 8;
+	flits = skb_transport_offset(skb) / 8;
 	sgp = ndesc == 1 ? (struct sg_ent *)&d->flit[flits] : sgl;
-	sgl_flits = make_sgl(skb, sgp, skb->h.raw, skb->tail - skb->h.raw,
+	sgl_flits = make_sgl(skb, sgp, skb_transport_header(skb),
+			     skb->tail - skb->transport_header,
 			     adap->pdev);
-	if (need_skb_unmap())
-		((struct unmap_info *)skb->cb)->len = skb->tail - skb->h.raw;
+	if (need_skb_unmap()) {
+		setup_deferred_unmapping(skb, adap->pdev, sgp, sgl_flits);
+		skb->destructor = deferred_unmap_destructor;
+		((struct unmap_info *)skb->cb)->len = (skb->tail -
+						       skb->transport_header);
+	}
 
 	write_wr_hdr_sgl(ndesc, skb, d, pidx, q, sgl, flits, sgl_flits,
 			 gen, from->wr_hi, from->wr_lo);
@@ -1283,8 +1353,8 @@ static inline unsigned int calc_tx_descs_ofld(const struct sk_buff *skb)
 	if (skb->len <= WR_LEN && cnt == 0)
 		return 1;	/* packet fits as immediate data */
 
-	flits = (skb->h.raw - skb->data) / 8;	/* headers */
-	if (skb->tail != skb->h.raw)
+	flits = skb_transport_offset(skb) / 8;	/* headers */
+	if (skb->tail != skb->transport_header)
 		cnt++;
 	return flits_to_desc(flits + sgl_len(cnt));
 }
@@ -1554,7 +1624,9 @@ static inline int rx_offload(struct t3cdev *tdev, struct sge_rspq *rq,
 			     unsigned int gather_idx)
 {
 	rq->offload_pkts++;
-	skb->mac.raw = skb->nh.raw = skb->h.raw = skb->data;
+	skb_reset_mac_header(skb);
+	skb_reset_network_header(skb);
+	skb_reset_transport_header(skb);
 
 	if (rq->polling) {
 		rx_gather[gather_idx++] = skb;
@@ -1617,11 +1689,9 @@ static void rx_eth(struct adapter *adap, struct sge_rspq *rq,
 	struct cpl_rx_pkt *p = (struct cpl_rx_pkt *)(skb->data + pad);
 	struct port_info *pi;
 
-	rq->eth_pkts++;
 	skb_pull(skb, sizeof(*p) + pad);
-	skb->dev = adap->port[p->iff];
 	skb->dev->last_rx = jiffies;
-	skb->protocol = eth_type_trans(skb, skb->dev);
+	skb->protocol = eth_type_trans(skb, adap->port[p->iff]);
 	pi = netdev_priv(skb->dev);
 	if (pi->rx_csum_offload && p->csum_valid && p->csum == 0xffff &&
 	    !p->fragment) {
@@ -1645,6 +1715,85 @@ static void rx_eth(struct adapter *adap, struct sge_rspq *rq,
 		netif_rx(skb);
 }
 
+#define SKB_DATA_SIZE 128
+
+static void skb_data_init(struct sk_buff *skb, struct sge_fl_page *p,
+			  unsigned int len)
+{
+	skb->len = len;
+	if (len <= SKB_DATA_SIZE) {
+		skb_copy_to_linear_data(skb, p->va, len);
+		skb->tail += len;
+		put_page(p->frag.page);
+	} else {
+		skb_copy_to_linear_data(skb, p->va, SKB_DATA_SIZE);
+		skb_shinfo(skb)->frags[0].page = p->frag.page;
+		skb_shinfo(skb)->frags[0].page_offset =
+		    p->frag.page_offset + SKB_DATA_SIZE;
+		skb_shinfo(skb)->frags[0].size = len - SKB_DATA_SIZE;
+		skb_shinfo(skb)->nr_frags = 1;
+		skb->data_len = len - SKB_DATA_SIZE;
+		skb->tail += SKB_DATA_SIZE;
+		skb->truesize += skb->data_len;
+	}
+}
+
+/**
+*      get_packet - return the next ingress packet buffer from a free list
+*      @adap: the adapter that received the packet
+*      @fl: the SGE free list holding the packet
+*      @len: the packet length including any SGE padding
+*      @drop_thres: # of remaining buffers before we start dropping packets
+*
+*      Get the next packet from a free list and complete setup of the
+*      sk_buff.  If the packet is small we make a copy and recycle the
+*      original buffer, otherwise we use the original buffer itself.  If a
+*      positive drop threshold is supplied packets are dropped and their
+*      buffers recycled if (a) the number of remaining buffers is under the
+*      threshold and the packet is too big to copy, or (b) the packet should
+*      be copied but there is no memory for the copy.
+*/
+static struct sk_buff *get_packet(struct adapter *adap, struct sge_fl *fl,
+				  unsigned int len, unsigned int drop_thres)
+{
+	struct sk_buff *skb = NULL;
+	struct rx_sw_desc *sd = &fl->sdesc[fl->cidx];
+
+	prefetch(sd->t.skb->data);
+
+	if (len <= SGE_RX_COPY_THRES) {
+		skb = alloc_skb(len, GFP_ATOMIC);
+		if (likely(skb != NULL)) {
+			struct rx_desc *d = &fl->desc[fl->cidx];
+			dma_addr_t mapping =
+			    (dma_addr_t)((u64) be32_to_cpu(d->addr_hi) << 32 |
+					 be32_to_cpu(d->addr_lo));
+
+			__skb_put(skb, len);
+			pci_dma_sync_single_for_cpu(adap->pdev, mapping, len,
+						    PCI_DMA_FROMDEVICE);
+			skb_copy_from_linear_data(sd->t.skb, skb->data, len);
+			pci_dma_sync_single_for_device(adap->pdev, mapping, len,
+						       PCI_DMA_FROMDEVICE);
+		} else if (!drop_thres)
+			goto use_orig_buf;
+recycle:
+		recycle_rx_buf(adap, fl, fl->cidx);
+		return skb;
+	}
+
+	if (unlikely(fl->credits < drop_thres))
+		goto recycle;
+
+use_orig_buf:
+	pci_unmap_single(adap->pdev, pci_unmap_addr(sd, dma_addr),
+			 fl->buf_size, PCI_DMA_FROMDEVICE);
+	skb = sd->t.skb;
+	skb_put(skb, len);
+	__refill_fl(adap, fl);
+	return skb;
+}
+
 /**
  *	handle_rsp_cntrl_info - handles control information in a response
  *	@qs: the queue set corresponding to the response
@@ -1767,7 +1916,7 @@ static int process_responses(struct adapter *adap, struct sge_qset *qs,
 	q->next_holdoff = q->holdoff_tmr;
 
 	while (likely(budget_left && is_new_response(r, q))) {
-		int eth, ethpad = 0;
+		int eth, ethpad = 2;
 		struct sk_buff *skb = NULL;
 		u32 len, flags = ntohl(r->flags);
 		u32 rss_hi = *(const u32 *)r, rss_lo = r->rss_hdr.rss_hash_val;
@@ -1794,18 +1943,56 @@ static int process_responses(struct adapter *adap, struct sge_qset *qs,
 				break;
 			}
 			q->imm_data++;
+			ethpad = 0;
 		} else if ((len = ntohl(r->len_cq)) != 0) {
-			struct sge_fl *fl;
+			struct sge_fl *fl =
+			    (len & F_RSPD_FLQ) ? &qs->fl[1] : &qs->fl[0];
+
+			if (fl->buf_size == RX_PAGE_SIZE) {
+				struct rx_sw_desc *sd = &fl->sdesc[fl->cidx];
+				struct sge_fl_page *p = &sd->t.page;
+
+				prefetch(p->va);
+				prefetch(p->va + L1_CACHE_BYTES);
+
+				__refill_fl(adap, fl);
+
+				pci_unmap_single(adap->pdev,
+						 pci_unmap_addr(sd, dma_addr),
+						 fl->buf_size,
+						 PCI_DMA_FROMDEVICE);
+
+				if (eth) {
+					if (unlikely(fl->credits <
+						     SGE_RX_DROP_THRES))
+						goto eth_recycle;
+
+					skb = alloc_skb(SKB_DATA_SIZE,
+							GFP_ATOMIC);
+					if (unlikely(!skb)) {
+eth_recycle:
+						q->rx_drops++;
+						recycle_rx_buf(adap, fl,
+							       fl->cidx);
+						goto eth_done;
+					}
+				} else {
+					skb = alloc_skb(SKB_DATA_SIZE,
+							GFP_ATOMIC);
+					if (unlikely(!skb))
+						goto no_mem;
+				}
+
+				skb_data_init(skb, p, G_RSPD_LEN(len));
+eth_done:
+				fl->credits--;
+				q->eth_pkts++;
+			} else {
+				fl->credits--;
+				skb = get_packet(adap, fl, G_RSPD_LEN(len),
+						 eth ? SGE_RX_DROP_THRES : 0);
+			}
 
-			fl = (len & F_RSPD_FLQ) ? &qs->fl[1] : &qs->fl[0];
-			fl->credits--;
-			skb = get_packet(adap, fl, G_RSPD_LEN(len),
-					 eth ? SGE_RX_DROP_THRES : 0);
-			if (!skb)
-				q->rx_drops++;
-			else if (r->rss_hdr.opcode == CPL_TRACE_PKT)
-				__skb_pull(skb, 2);
-			ethpad = 2;
 			if (++fl->cidx == fl->size)
 				fl->cidx = 0;
 		} else
@@ -1829,18 +2016,23 @@ static int process_responses(struct adapter *adap, struct sge_qset *qs,
 			q->credits = 0;
 		}
 
-		if (likely(skb != NULL)) {
+		if (skb) {
+			/* Preserve the RSS info in csum & priority */
+			skb->csum = rss_hi;
+			skb->priority = rss_lo;
+
 			if (eth)
 				rx_eth(adap, q, skb, ethpad);
 			else {
-				/* Preserve the RSS info in csum & priority */
-				skb->csum = rss_hi;
-				skb->priority = rss_lo;
-				ngathered = rx_offload(&adap->tdev, q, skb,
-						       offload_skbs, ngathered);
+				if (unlikely(r->rss_hdr.opcode ==
+					     CPL_TRACE_PKT))
+					__skb_pull(skb, ethpad);
+
+				ngathered = rx_offload(&adap->tdev, q,
+						       skb, offload_skbs,
+						       ngathered);
 			}
 		}
-
 		--budget_left;
 	}
 
@@ -2320,10 +2512,23 @@ static void sge_timer_cb(unsigned long data)
 	    &adap->sge.qs[0].rspq.lock;
 	if (spin_trylock_irq(lock)) {
 		if (!napi_is_scheduled(qs->netdev)) {
+			u32 status = t3_read_reg(adap, A_SG_RSPQ_FL_STATUS);
+
 			if (qs->fl[0].credits < qs->fl[0].size)
 				__refill_fl(adap, &qs->fl[0]);
 			if (qs->fl[1].credits < qs->fl[1].size)
 				__refill_fl(adap, &qs->fl[1]);
+
+			if (status & (1 << qs->rspq.cntxt_id)) {
+				qs->rspq.starved++;
+				if (qs->rspq.credits) {
+					refill_rspq(adap, &qs->rspq, 1);
+					qs->rspq.credits--;
+					qs->rspq.restarted++;
+					t3_write_reg(adap, A_SG_RSPQ_FL_STATUS,
+						     1 << qs->rspq.cntxt_id);
+				}
+			}
 		}
 		spin_unlock_irq(lock);
 	}
@@ -2431,14 +2636,22 @@ int t3_sge_alloc_qset(struct adapter *adapter, unsigned int id, int nports,
 	q->txq[TXQ_ETH].stop_thres = nports *
 	    flits_to_desc(sgl_len(MAX_SKB_FRAGS + 1) + 3);
 
-	if (ntxq == 1) {
+	if (!is_offload(adapter)) {
+#ifdef USE_RX_PAGE
+		q->fl[0].buf_size = RX_PAGE_SIZE;
+#else
 		q->fl[0].buf_size = SGE_RX_SM_BUF_SIZE + 2 +
 		    sizeof(struct cpl_rx_pkt);
+#endif
 		q->fl[1].buf_size = MAX_FRAME_SIZE + 2 +
 		    sizeof(struct cpl_rx_pkt);
 	} else {
+#ifdef USE_RX_PAGE
+		q->fl[0].buf_size = RX_PAGE_SIZE;
+#else
 		q->fl[0].buf_size = SGE_RX_SM_BUF_SIZE +
 		    sizeof(struct cpl_rx_data);
+#endif
 		q->fl[1].buf_size = (16 * 1024) -
 		    SKB_DATA_ALIGN(sizeof(struct skb_shared_info));
 	}
@@ -2632,7 +2845,7 @@ void __devinit t3_sge_prep(struct adapter *adap, struct sge_params *p)
 		q->polling = adap->params.rev > 0;
 		q->coalesce_usecs = 5;
 		q->rspq_size = 1024;
-		q->fl_size = 4096;
+		q->fl_size = 1024;
 		q->jumbo_size = 512;
 		q->txq_size[TXQ_ETH] = 1024;
 		q->txq_size[TXQ_OFLD] = 1024;
diff --git a/drivers/net/cxgb3/t3_hw.c b/drivers/net/cxgb3/t3_hw.c
index 365a7f5b1f94..fb485d0a43d8 100644
--- a/drivers/net/cxgb3/t3_hw.c
+++ b/drivers/net/cxgb3/t3_hw.c
@@ -438,23 +438,23 @@ static const struct adapter_info t3_adap_info[] = {
 	{2, 0, 0, 0,
 	 F_GPIO2_OEN | F_GPIO4_OEN |
 	 F_GPIO2_OUT_VAL | F_GPIO4_OUT_VAL, F_GPIO3 | F_GPIO5,
-	 SUPPORTED_OFFLOAD,
+	 0,
 	 &mi1_mdio_ops, "Chelsio PE9000"},
 	{2, 0, 0, 0,
 	 F_GPIO2_OEN | F_GPIO4_OEN |
 	 F_GPIO2_OUT_VAL | F_GPIO4_OUT_VAL, F_GPIO3 | F_GPIO5,
-	 SUPPORTED_OFFLOAD,
+	 0,
 	 &mi1_mdio_ops, "Chelsio T302"},
 	{1, 0, 0, 0,
 	 F_GPIO1_OEN | F_GPIO6_OEN | F_GPIO7_OEN | F_GPIO10_OEN |
 	 F_GPIO1_OUT_VAL | F_GPIO6_OUT_VAL | F_GPIO10_OUT_VAL, 0,
-	 SUPPORTED_10000baseT_Full | SUPPORTED_AUI | SUPPORTED_OFFLOAD,
+	 SUPPORTED_10000baseT_Full | SUPPORTED_AUI,
 	 &mi1_mdio_ext_ops, "Chelsio T310"},
 	{2, 0, 0, 0,
 	 F_GPIO1_OEN | F_GPIO2_OEN | F_GPIO4_OEN | F_GPIO5_OEN | F_GPIO6_OEN |
 	 F_GPIO7_OEN | F_GPIO10_OEN | F_GPIO11_OEN | F_GPIO1_OUT_VAL |
 	 F_GPIO5_OUT_VAL | F_GPIO6_OUT_VAL | F_GPIO10_OUT_VAL, 0,
-	 SUPPORTED_10000baseT_Full | SUPPORTED_AUI | SUPPORTED_OFFLOAD,
+	 SUPPORTED_10000baseT_Full | SUPPORTED_AUI,
 	 &mi1_mdio_ext_ops, "Chelsio T320"},
 };
 
@@ -681,7 +681,8 @@ enum {
 	SF_ERASE_SECTOR = 0xd8,	/* erase sector */
 
 	FW_FLASH_BOOT_ADDR = 0x70000,	/* start address of FW in flash */
-	FW_VERS_ADDR = 0x77ffc	/* flash address holding FW version */
+	FW_VERS_ADDR = 0x77ffc,    /* flash address holding FW version */
+	FW_MIN_SIZE = 8            /* at least version and csum */
 };
 
 /**
@@ -884,11 +885,13 @@ int t3_check_fw_version(struct adapter *adapter)
 	major = G_FW_VERSION_MAJOR(vers);
 	minor = G_FW_VERSION_MINOR(vers);
 
-	if (type == FW_VERSION_T3 && major == 3 && minor == 1)
+	if (type == FW_VERSION_T3 && major == FW_VERSION_MAJOR &&
+	    minor == FW_VERSION_MINOR)
 		return 0;
 
 	CH_ERR(adapter, "found wrong FW version(%u.%u), "
-	       "driver needs version 3.1\n", major, minor);
+	       "driver needs version %u.%u\n", major, minor,
+	       FW_VERSION_MAJOR, FW_VERSION_MINOR);
 	return -EINVAL;
 }
 
@@ -933,7 +936,7 @@ int t3_load_fw(struct adapter *adapter, const u8 *fw_data, unsigned int size)
 	const u32 *p = (const u32 *)fw_data;
 	int ret, addr, fw_sector = FW_FLASH_BOOT_ADDR >> 16;
 
-	if (size & 3)
+	if ((size & 3) || size < FW_MIN_SIZE)
 		return -EINVAL;
 	if (size > FW_VERS_ADDR + 8 - FW_FLASH_BOOT_ADDR)
 		return -EFBIG;
@@ -1520,19 +1523,25 @@ static int mac_intr_handler(struct adapter *adap, unsigned int idx)
  */
 int t3_phy_intr_handler(struct adapter *adapter)
 {
-	static const int intr_gpio_bits[] = { 8, 0x20 };
-
+	u32 mask, gpi = adapter_info(adapter)->gpio_intr;
 	u32 i, cause = t3_read_reg(adapter, A_T3DBG_INT_CAUSE);
 
 	for_each_port(adapter, i) {
-		if (cause & intr_gpio_bits[i]) {
-			struct cphy *phy = &adap2pinfo(adapter, i)->phy;
-			int phy_cause = phy->ops->intr_handler(phy);
+		struct port_info *p = adap2pinfo(adapter, i);
+
+		mask = gpi - (gpi & (gpi - 1));
+		gpi -= mask;
+
+		if (!(p->port_type->caps & SUPPORTED_IRQ))
+			continue;
+
+		if (cause & mask) {
+			int phy_cause = p->phy.ops->intr_handler(&p->phy);
 
 			if (phy_cause & cphy_cause_link_change)
 				t3_link_changed(adapter, i);
 			if (phy_cause & cphy_cause_fifo_error)
-				phy->fifo_errors++;
+				p->phy.fifo_errors++;
 		}
 	}
 
@@ -2897,6 +2906,9 @@ static int mc7_init(struct mc7 *mc7, unsigned int mc7_clock, int mem_type)
 	struct adapter *adapter = mc7->adapter;
 	const struct mc7_timing_params *p = &mc7_timings[mem_type];
 
+	if (!mc7->size)
+		return 0;
+
 	val = t3_read_reg(adapter, mc7->offset + A_MC7_CFG);
 	slow = val & F_SLOW;
 	width = G_WIDTH(val);
@@ -3097,8 +3109,10 @@ int t3_init_hw(struct adapter *adapter, u32 fw_params)
 	do {			/* wait for uP to initialize */
 		msleep(20);
 	} while (t3_read_reg(adapter, A_CIM_HOST_ACC_DATA) && --attempts);
-	if (!attempts)
+	if (!attempts) {
+		CH_ERR(adapter, "uP initialization timed out\n");
 		goto out_err;
+	}
 
 	err = 0;
 out_err:
@@ -3198,7 +3212,7 @@ static void __devinit mc7_prep(struct adapter *adapter, struct mc7 *mc7,
 	mc7->name = name;
 	mc7->offset = base_addr - MC7_PMRX_BASE_ADDR;
 	cfg = t3_read_reg(adapter, mc7->offset + A_MC7_CFG);
-	mc7->size = mc7_calc_size(cfg);
+	mc7->size = mc7->size = G_DEN(cfg) == M_DEN ? 0 : mc7_calc_size(cfg);
 	mc7->width = G_WIDTH(cfg);
 }
 
@@ -3225,6 +3239,7 @@ void early_hw_init(struct adapter *adapter, const struct adapter_info *ai)
 		     V_I2C_CLKDIV(adapter->params.vpd.cclk / 80 - 1));
 	t3_write_reg(adapter, A_T3DBG_GPIO_EN,
 		     ai->gpio_out | F_GPIO0_OEN | F_GPIO0_OUT_VAL);
+	t3_write_reg(adapter, A_MC5_DB_SERVER_INDEX, 0);
 
 	if (adapter->params.rev == 0 || !uses_xaui(adapter))
 		val |= F_ENRGMII;
@@ -3241,15 +3256,17 @@ void early_hw_init(struct adapter *adapter, const struct adapter_info *ai)
 }
 
 /*
- * Reset the adapter.  PCIe cards lose their config space during reset, PCI-X
+ * Reset the adapter. 
+ * Older PCIe cards lose their config space during reset, PCI-X
  * ones don't.
  */
 int t3_reset_adapter(struct adapter *adapter)
 {
-	int i;
+	int i, save_and_restore_pcie = 
+	    adapter->params.rev < T3_REV_B2 && is_pcie(adapter);
 	uint16_t devid = 0;
 
-	if (is_pcie(adapter))
+	if (save_and_restore_pcie)
 		pci_save_state(adapter->pdev);
 	t3_write_reg(adapter, A_PL_RST, F_CRSTWRM | F_CRSTWRMMODE);
 
@@ -3267,7 +3284,7 @@ int t3_reset_adapter(struct adapter *adapter)
 	if (devid != 0x1425)
 		return -1;
 
-	if (is_pcie(adapter))
+	if (save_and_restore_pcie)
 		pci_restore_state(adapter->pdev);
 	return 0;
 }
@@ -3321,7 +3338,13 @@ int __devinit t3_prep_adapter(struct adapter *adapter,
 		p->tx_num_pgs = pm_num_pages(p->chan_tx_size, p->tx_pg_size);
 		p->ntimer_qs = p->cm_size >= (128 << 20) ||
 		    adapter->params.rev > 0 ? 12 : 6;
+	}
+
+	adapter->params.offload = t3_mc7_size(&adapter->pmrx) &&
+				  t3_mc7_size(&adapter->pmtx) &&
+				  t3_mc7_size(&adapter->cm);
 
+	if (is_offload(adapter)) {
 		adapter->params.mc5.nservers = DEFAULT_NSERVERS;
 		adapter->params.mc5.nfilters = adapter->params.rev > 0 ?
 		    DEFAULT_NFILTERS : 0;
diff --git a/drivers/net/cxgb3/t3cdev.h b/drivers/net/cxgb3/t3cdev.h
index 9af3bcd64b3b..fa4099bc0416 100644
--- a/drivers/net/cxgb3/t3cdev.h
+++ b/drivers/net/cxgb3/t3cdev.h
@@ -1,6 +1,5 @@
 /*
  * Copyright (C) 2006-2007 Chelsio Communications.  All rights reserved.
- * Copyright (C) 2006-2007 Open Grid Computing, Inc.  All rights reserved.
  *
  * This software is available to you under a choice of one of two
  * licenses.  You may choose to be licensed under the terms of the GNU
diff --git a/drivers/net/cxgb3/version.h b/drivers/net/cxgb3/version.h
index 2b67dd523cc1..042e27e291cd 100644
--- a/drivers/net/cxgb3/version.h
+++ b/drivers/net/cxgb3/version.h
@@ -35,5 +35,10 @@
 #define DRV_DESC "Chelsio T3 Network Driver"
 #define DRV_NAME "cxgb3"
 /* Driver version */
-#define DRV_VERSION "1.0"
+#define DRV_VERSION "1.0-ko"
+
+/* Firmware version */
+#define FW_VERSION_MAJOR 3
+#define FW_VERSION_MINOR 3
+#define FW_VERSION_MICRO 0
 #endif				/* __CHELSIO_VERSION_H */
diff --git a/drivers/net/cxgb3/xgmac.c b/drivers/net/cxgb3/xgmac.c
index 907a272ae32d..a506792f9575 100644
--- a/drivers/net/cxgb3/xgmac.c
+++ b/drivers/net/cxgb3/xgmac.c
@@ -124,9 +124,6 @@ int t3_mac_reset(struct cmac *mac)
 			xaui_serdes_reset(mac);
 	}
 
-	if (adap->params.rev > 0)
-		t3_write_reg(adap, A_XGM_PAUSE_TIMER + oft, 0xf000);
-
 	val = F_MAC_RESET_;
 	if (is_10G(adap))
 		val |= F_PCS_RESET_;
@@ -145,6 +142,58 @@ int t3_mac_reset(struct cmac *mac)
 	return 0;
 }
 
+int t3b2_mac_reset(struct cmac *mac)
+{
+	struct adapter *adap = mac->adapter;
+	unsigned int oft = mac->offset;
+	u32 val;
+
+	if (!macidx(mac)) 
+		t3_set_reg_field(adap, A_MPS_CFG, F_PORT0ACTIVE, 0);
+	else
+		t3_set_reg_field(adap, A_MPS_CFG, F_PORT1ACTIVE, 0);
+
+	t3_write_reg(adap, A_XGM_RESET_CTRL + oft, F_MAC_RESET_);
+	t3_read_reg(adap, A_XGM_RESET_CTRL + oft);    /* flush */
+
+	msleep(10);
+
+	/* Check for xgm Rx fifo empty */
+	if (t3_wait_op_done(adap, A_XGM_RX_MAX_PKT_SIZE_ERR_CNT + oft,
+			    0x80000000, 1, 5, 2)) {
+		CH_ERR(adap, "MAC %d Rx fifo drain failed\n",
+		       macidx(mac));
+		return -1;
+	}
+
+	t3_write_reg(adap, A_XGM_RESET_CTRL + oft, 0);
+	t3_read_reg(adap, A_XGM_RESET_CTRL + oft);    /* flush */
+
+	val = F_MAC_RESET_;
+	if (is_10G(adap))
+		val |= F_PCS_RESET_;
+	else if (uses_xaui(adap))
+		val |= F_PCS_RESET_ | F_XG2G_RESET_;
+	else
+		val |= F_RGMII_RESET_ | F_XG2G_RESET_;
+	t3_write_reg(adap, A_XGM_RESET_CTRL + oft, val);
+	t3_read_reg(adap, A_XGM_RESET_CTRL + oft);  /* flush */
+	if ((val & F_PCS_RESET_) && adap->params.rev) {
+		msleep(1);
+		t3b_pcs_reset(mac);
+	}
+	t3_write_reg(adap, A_XGM_RX_CFG + oft, 
+		     F_DISPAUSEFRAMES | F_EN1536BFRAMES |
+		     F_RMFCS | F_ENJUMBO | F_ENHASHMCAST);
+
+	if (!macidx(mac)) 
+		t3_set_reg_field(adap, A_MPS_CFG, 0, F_PORT0ACTIVE);
+	else
+		t3_set_reg_field(adap, A_MPS_CFG, 0, F_PORT1ACTIVE);
+
+	return 0;
+}
+
 /*
  * Set the exact match register 'idx' to recognize the given Ethernet address.
  */
@@ -251,9 +300,11 @@ int t3_mac_set_mtu(struct cmac *mac, unsigned int mtu)
 	 * Adjust the PAUSE frame watermarks.  We always set the LWM, and the
 	 * HWM only if flow-control is enabled.
 	 */
-	hwm = max(MAC_RXFIFO_SIZE - 3 * mtu, MAC_RXFIFO_SIZE / 2U);
-	hwm = min(hwm, 3 * MAC_RXFIFO_SIZE / 4 + 1024);
-	lwm = hwm - 1024;
+	hwm = max_t(unsigned int, MAC_RXFIFO_SIZE - 3 * mtu, 
+		    MAC_RXFIFO_SIZE * 38 / 100);
+	hwm = min(hwm, MAC_RXFIFO_SIZE - 8192);
+	lwm = min(3 * (int)mtu, MAC_RXFIFO_SIZE / 4);
+
 	v = t3_read_reg(adap, A_XGM_RXFIFO_CFG + mac->offset);
 	v &= ~V_RXFIFOPAUSELWM(M_RXFIFOPAUSELWM);
 	v |= V_RXFIFOPAUSELWM(lwm / 8);
@@ -270,7 +321,15 @@ int t3_mac_set_mtu(struct cmac *mac, unsigned int mtu)
 	thres = mtu > thres ? (mtu - thres + 7) / 8 : 0;
 	thres = max(thres, 8U);	/* need at least 8 */
 	t3_set_reg_field(adap, A_XGM_TXFIFO_CFG + mac->offset,
-			 V_TXFIFOTHRESH(M_TXFIFOTHRESH), V_TXFIFOTHRESH(thres));
+			 V_TXFIFOTHRESH(M_TXFIFOTHRESH) | V_TXIPG(M_TXIPG),
+			 V_TXFIFOTHRESH(thres) | V_TXIPG(1));
+
+	if (adap->params.rev > 0)
+		t3_write_reg(adap, A_XGM_PAUSE_TIMER + mac->offset,
+			     (hwm - lwm) * 4 / 8);
+	t3_write_reg(adap, A_XGM_TX_PAUSE_QUANTA + mac->offset,
+		     MAC_RXFIFO_SIZE * 4 * 8 / 512);
+
 	return 0;
 }
 
@@ -298,12 +357,6 @@ int t3_mac_set_speed_duplex_fc(struct cmac *mac, int speed, int duplex, int fc)
 				 V_PORTSPEED(M_PORTSPEED), val);
 	}
 
-	val = t3_read_reg(adap, A_XGM_RXFIFO_CFG + oft);
-	val &= ~V_RXFIFOPAUSEHWM(M_RXFIFOPAUSEHWM);
-	if (fc & PAUSE_TX)
-		val |= V_RXFIFOPAUSEHWM(G_RXFIFOPAUSELWM(val) + 128);	/* +1KB */
-	t3_write_reg(adap, A_XGM_RXFIFO_CFG + oft, val);
-
 	t3_set_reg_field(adap, A_XGM_TX_CFG + oft, F_TXPAUSEEN,
 			 (fc & PAUSE_RX) ? F_TXPAUSEEN : 0);
 	return 0;
@@ -314,13 +367,28 @@ int t3_mac_enable(struct cmac *mac, int which)
 	int idx = macidx(mac);
 	struct adapter *adap = mac->adapter;
 	unsigned int oft = mac->offset;
-
+	struct mac_stats *s = &mac->stats;
+	
 	if (which & MAC_DIRECTION_TX) {
 		t3_write_reg(adap, A_XGM_TX_CTRL + oft, F_TXEN);
 		t3_write_reg(adap, A_TP_PIO_ADDR, A_TP_TX_DROP_CFG_CH0 + idx);
-		t3_write_reg(adap, A_TP_PIO_DATA, 0xbf000001);
+		t3_write_reg(adap, A_TP_PIO_DATA, 0xc0ede401);
 		t3_write_reg(adap, A_TP_PIO_ADDR, A_TP_TX_DROP_MODE);
 		t3_set_reg_field(adap, A_TP_PIO_DATA, 1 << idx, 1 << idx);
+
+		t3_write_reg(adap, A_TP_PIO_ADDR, A_TP_TX_DROP_CNT_CH0 + idx);
+		mac->tx_mcnt = s->tx_frames;
+		mac->tx_tcnt = (G_TXDROPCNTCH0RCVD(t3_read_reg(adap,
+							A_TP_PIO_DATA)));
+		mac->tx_xcnt = (G_TXSPI4SOPCNT(t3_read_reg(adap,
+						A_XGM_TX_SPI4_SOP_EOP_CNT +
+						oft)));
+		mac->rx_mcnt = s->rx_frames;
+		mac->rx_xcnt = (G_TXSPI4SOPCNT(t3_read_reg(adap,
+						A_XGM_RX_SPI4_SOP_EOP_CNT +
+						oft)));
+		mac->txen = F_TXEN;
+		mac->toggle_cnt = 0;
 	}
 	if (which & MAC_DIRECTION_RX)
 		t3_write_reg(adap, A_XGM_RX_CTRL + oft, F_RXEN);
@@ -331,19 +399,102 @@ int t3_mac_disable(struct cmac *mac, int which)
 {
 	int idx = macidx(mac);
 	struct adapter *adap = mac->adapter;
+	int val;
 
 	if (which & MAC_DIRECTION_TX) {
 		t3_write_reg(adap, A_XGM_TX_CTRL + mac->offset, 0);
 		t3_write_reg(adap, A_TP_PIO_ADDR, A_TP_TX_DROP_CFG_CH0 + idx);
 		t3_write_reg(adap, A_TP_PIO_DATA, 0xc000001f);
 		t3_write_reg(adap, A_TP_PIO_ADDR, A_TP_TX_DROP_MODE);
-		t3_set_reg_field(adap, A_TP_PIO_DATA, 1 << idx, 0);
+		t3_set_reg_field(adap, A_TP_PIO_DATA, 1 << idx, 1 << idx);
+		mac->txen = 0;
 	}
-	if (which & MAC_DIRECTION_RX)
+	if (which & MAC_DIRECTION_RX) {
+		t3_set_reg_field(mac->adapter, A_XGM_RESET_CTRL + mac->offset,
+				 F_PCS_RESET_, 0);
+		msleep(100);
 		t3_write_reg(adap, A_XGM_RX_CTRL + mac->offset, 0);
+		val = F_MAC_RESET_;
+		if (is_10G(adap))
+			val |= F_PCS_RESET_;
+		else if (uses_xaui(adap))
+			val |= F_PCS_RESET_ | F_XG2G_RESET_;
+		else
+			val |= F_RGMII_RESET_ | F_XG2G_RESET_;
+		t3_write_reg(mac->adapter, A_XGM_RESET_CTRL + mac->offset, val);
+	}
 	return 0;
 }
 
+int t3b2_mac_watchdog_task(struct cmac *mac)
+{
+	struct adapter *adap = mac->adapter;
+	struct mac_stats *s = &mac->stats;
+	unsigned int tx_tcnt, tx_xcnt;
+	unsigned int tx_mcnt = s->tx_frames;
+	unsigned int rx_mcnt = s->rx_frames;
+	unsigned int rx_xcnt;
+	int status;
+
+	if (tx_mcnt == mac->tx_mcnt) {
+		tx_xcnt = (G_TXSPI4SOPCNT(t3_read_reg(adap,
+						A_XGM_TX_SPI4_SOP_EOP_CNT +
+					       	mac->offset)));
+		if (tx_xcnt == 0) {
+			t3_write_reg(adap, A_TP_PIO_ADDR,
+				     A_TP_TX_DROP_CNT_CH0 + macidx(mac));
+			tx_tcnt = (G_TXDROPCNTCH0RCVD(t3_read_reg(adap,
+						      A_TP_PIO_DATA)));
+		} else {
+			mac->toggle_cnt = 0;
+			return 0;
+		}
+	} else {
+		mac->toggle_cnt = 0;
+		return 0;
+	}
+
+	if (((tx_tcnt != mac->tx_tcnt) &&
+	     (tx_xcnt == 0) && (mac->tx_xcnt == 0)) ||
+	    ((mac->tx_mcnt == tx_mcnt) &&
+	     (tx_xcnt != 0) && (mac->tx_xcnt != 0))) {
+		if (mac->toggle_cnt > 4)
+			status = 2;
+		else 
+			status = 1;
+	} else {
+		mac->toggle_cnt = 0;
+		return 0;
+	}
+
+	if (rx_mcnt != mac->rx_mcnt)
+		rx_xcnt = (G_TXSPI4SOPCNT(t3_read_reg(adap,
+						A_XGM_RX_SPI4_SOP_EOP_CNT +
+						mac->offset)));
+	else 
+		return 0;
+
+	if (mac->rx_mcnt != s->rx_frames && rx_xcnt == 0 && mac->rx_xcnt == 0) 
+		status = 2;
+	
+	mac->tx_tcnt = tx_tcnt;
+	mac->tx_xcnt = tx_xcnt;
+	mac->tx_mcnt = s->tx_frames;
+	mac->rx_xcnt = rx_xcnt;
+	mac->rx_mcnt = s->rx_frames;
+	if (status == 1) {
+		t3_write_reg(adap, A_XGM_TX_CTRL + mac->offset, 0);
+		t3_read_reg(adap, A_XGM_TX_CTRL + mac->offset);  /* flush */
+		t3_write_reg(adap, A_XGM_TX_CTRL + mac->offset, mac->txen);
+		t3_read_reg(adap, A_XGM_TX_CTRL + mac->offset);  /* flush */
+		mac->toggle_cnt++;
+	} else if (status == 2) {
+		t3b2_mac_reset(mac);
+		mac->toggle_cnt = 0;
+	}
+	return status;
+}
+
 /*
  * This function is called periodically to accumulate the current values of the
  * RMON counters into the port statistics.  Since the packet counters are only
@@ -373,7 +524,11 @@ const struct mac_stats *t3_mac_update_stats(struct cmac *mac)
 	RMON_UPDATE(mac, rx_symbol_errs, RX_SYM_CODE_ERR_FRAMES);
 
 	RMON_UPDATE(mac, rx_too_long, RX_OVERSIZE_FRAMES);
-	mac->stats.rx_too_long += RMON_READ(mac, A_XGM_RX_MAX_PKT_SIZE_ERR_CNT);
+
+	v = RMON_READ(mac, A_XGM_RX_MAX_PKT_SIZE_ERR_CNT);
+	if (mac->adapter->params.rev == T3_REV_B2)
+		v &= 0x7fffffff;
+	mac->stats.rx_too_long += v;
 
 	RMON_UPDATE(mac, rx_frames_64, RX_64B_FRAMES);
 	RMON_UPDATE(mac, rx_frames_65_127, RX_65_127B_FRAMES);