Please pull from 'davem-next' branch of
master.kernel.org:/pub/scm/linux/kernel/git/jgarzik/netdev-2.6.git davem-next

to receive the following updates:

 Documentation/powerpc/booting-without-of.txt |   21 +
 drivers/net/Kconfig                          |    4 +-
 drivers/net/forcedeth.c                      |   61 +-
 drivers/net/fs_enet/fs_enet-main.c           |   31 +-
 drivers/net/myri10ge/myri10ge.c              | 1076 ++++++++++++++++++++------
 drivers/net/ns83820.c                        |    9 +-
 drivers/net/phy/Kconfig                      |    6 +
 drivers/net/phy/Makefile                     |    1 +
 drivers/net/phy/mdio-ofgpio.c                |  205 +++++
 drivers/net/ps3_gelic_net.c                  |   10 +-
 drivers/net/ps3_gelic_net.h                  |    2 +-
 drivers/net/ps3_gelic_wireless.c             |   40 +-
 drivers/net/ps3_gelic_wireless.h             |    4 +-
 drivers/net/s2io.c                           |    2 +-
 drivers/net/sfc/Kconfig                      |    2 +
 drivers/net/sfc/Makefile                     |    2 +-
 drivers/net/sfc/boards.c                     |    2 +-
 drivers/net/sfc/boards.h                     |    3 +-
 drivers/net/sfc/efx.c                        |    2 +
 drivers/net/sfc/falcon.c                     |   74 ++-
 drivers/net/sfc/i2c-direct.c                 |  381 ---------
 drivers/net/sfc/i2c-direct.h                 |   91 ---
 drivers/net/sfc/net_driver.h                 |   11 +-
 drivers/net/sfc/sfe4001.c                    |  126 ++--
 drivers/net/sky2.c                           |  173 ++++-
 drivers/net/sky2.h                           |   23 +-
 drivers/net/tlan.c                           |  490 ++++++------
 drivers/net/tlan.h                           |   26 +-
 drivers/net/tsi108_eth.c                     |    6 +-
 29 files changed, 1749 insertions(+), 1135 deletions(-)
 create mode 100644 drivers/net/phy/mdio-ofgpio.c
 delete mode 100644 drivers/net/sfc/i2c-direct.c
 delete mode 100644 drivers/net/sfc/i2c-direct.h

Ben Hutchings (2):
      sfc: Use kernel I2C system and i2c-algo-bit driver
      sfc: Reduce I2C udelay to 5 resulting in a clock frequency of 100 kHz

Brice Goglin (3):
      myri10ge: add routines for multislices
      myri10ge: add multislices support
      myri10ge: add Direct Cache Access support

Daniel Walker (3):
      ps3: gelic: scan_lock semaphore to mutex
      ps3: gelic: assoc_stat_lock semaphore to mutex
      ps3: gelic: updown_lock semaphore to mutex

Laurent Pinchart (2):
      net: OpenFirmware GPIO based MDIO bitbang driver
      fs_enet: MDIO on GPIO support

Marcin Slusarz (1):
      net/s2io: set_rxd_buffer_pointer returns -ENOMEM, not ENOMEM

Randy Dunlap (1):
      cxgb3: fix build error when INET=n

Stephen Hemminger (12):
      tlan: get rid of padding buffer
      tlan: use netdevice stats
      tlan: remove unused devName field
      tlan: 64bit conversion
      tlan: manage rx allocation failure better
      tlan: proper shared IRQ support
      tlan: wrap source lines
      sky2: split phy power into two functions
      sky2: put PHY in sleep when down
      sky2: pci power savings
      ts108: use netdev_alloc_skb
      ns8320: use netdev_alloc_skb

Tobias Diedrich (3):
      [netdrvr] forcedeth: setup wake-on-lan before shutting down
      [netdrvr] forcedeth: save/restore device configuration space
      [netdrvr] forcedeth: reorder suspend/resume code

diff --git a/Documentation/powerpc/booting-without-of.txt b/Documentation/powerpc/booting-without-of.txt
index 1d2a772..46a9dba 100644
--- a/Documentation/powerpc/booting-without-of.txt
+++ b/Documentation/powerpc/booting-without-of.txt
@@ -58,6 +58,7 @@ Table of Contents
       o) Xilinx IP cores
       p) Freescale Synchronous Serial Interface
 	  q) USB EHCI controllers
+      r) MDIO on GPIOs
 
   VII - Marvell Discovery mv64[345]6x System Controller chips
     1) The /system-controller node
@@ -2870,6 +2871,26 @@ platforms are moved over to use the flattened-device-tree model.
 		reg = <0xe8000000 32>;
 	};
 
+   r) MDIO on GPIOs
+
+   Currently defined compatibles:
+   - virtual,gpio-mdio
+
+   MDC and MDIO lines connected to GPIO controllers are listed in the
+   gpios property as described in section VIII.1 in the following order:
+
+   MDC, MDIO.
+
+   Example:
+
+	mdio {
+		compatible = "virtual,mdio-gpio";
+		#address-cells = <1>;
+		#size-cells = <0>;
+		gpios = <&qe_pio_a 11
+			 &qe_pio_c 6>;
+	};
+
 VII - Marvell Discovery mv64[345]6x System Controller chips
 ===========================================================
 
diff --git a/drivers/net/Kconfig b/drivers/net/Kconfig
index 8178a4d..d415d81 100644
--- a/drivers/net/Kconfig
+++ b/drivers/net/Kconfig
@@ -1670,7 +1670,7 @@ config SUNDANCE_MMIO
 
 config TLAN
 	tristate "TI ThunderLAN support"
-	depends on NET_PCI && (PCI || EISA) && !64BIT
+	depends on NET_PCI && (PCI || EISA)
 	---help---
 	  If you have a PCI Ethernet network card based on the ThunderLAN chip
 	  which is supported by this driver, say Y and read the
@@ -2408,7 +2408,7 @@ config CHELSIO_T1_NAPI
 
 config CHELSIO_T3
 	tristate "Chelsio Communications T3 10Gb Ethernet support"
-	depends on PCI
+	depends on PCI && INET
 	select FW_LOADER
 	select INET_LRO
 	help
diff --git a/drivers/net/forcedeth.c b/drivers/net/forcedeth.c
index 9eca97f..e4d6978 100644
--- a/drivers/net/forcedeth.c
+++ b/drivers/net/forcedeth.c
@@ -426,6 +426,7 @@ union ring_type {
 #define NV_PCI_REGSZ_VER1      	0x270
 #define NV_PCI_REGSZ_VER2      	0x2d4
 #define NV_PCI_REGSZ_VER3      	0x604
+#define NV_PCI_REGSZ_MAX       	0x604
 
 /* various timeout delays: all in usec */
 #define NV_TXRX_RESET_DELAY	4
@@ -784,6 +785,9 @@ struct fe_priv {
 
 	/* flow control */
 	u32 pause_flags;
+
+	/* power saved state */
+	u32 saved_config_space[NV_PCI_REGSZ_MAX/4];
 };
 
 /*
@@ -5785,50 +5789,66 @@ static int nv_suspend(struct pci_dev *pdev, pm_message_t state)
 {
 	struct net_device *dev = pci_get_drvdata(pdev);
 	struct fe_priv *np = netdev_priv(dev);
+	u8 __iomem *base = get_hwbase(dev);
+	int i;
 
-	if (!netif_running(dev))
-		goto out;
-
+	if (netif_running(dev)) {
+		// Gross.
+		nv_close(dev);
+	}
 	netif_device_detach(dev);
 
-	// Gross.
-	nv_close(dev);
+	/* save non-pci configuration space */
+	for (i = 0;i <= np->register_size/sizeof(u32); i++)
+		np->saved_config_space[i] = readl(base + i*sizeof(u32));
 
 	pci_save_state(pdev);
 	pci_enable_wake(pdev, pci_choose_state(pdev, state), np->wolenabled);
+	pci_disable_device(pdev);
 	pci_set_power_state(pdev, pci_choose_state(pdev, state));
-out:
 	return 0;
 }
 
 static int nv_resume(struct pci_dev *pdev)
 {
 	struct net_device *dev = pci_get_drvdata(pdev);
+	struct fe_priv *np = netdev_priv(dev);
 	u8 __iomem *base = get_hwbase(dev);
-	int rc = 0;
-	u32 txreg;
-
-	if (!netif_running(dev))
-		goto out;
-
-	netif_device_attach(dev);
+	int i, rc = 0;
 
 	pci_set_power_state(pdev, PCI_D0);
 	pci_restore_state(pdev);
+	/* ack any pending wake events, disable PME */
 	pci_enable_wake(pdev, PCI_D0, 0);
 
-	/* restore mac address reverse flag */
-	txreg = readl(base + NvRegTransmitPoll);
-	txreg |= NVREG_TRANSMITPOLL_MAC_ADDR_REV;
-	writel(txreg, base + NvRegTransmitPoll);
+	/* restore non-pci configuration space */
+	for (i = 0;i <= np->register_size/sizeof(u32); i++)
+		writel(np->saved_config_space[i], base+i*sizeof(u32));
 
-	rc = nv_open(dev);
-	nv_set_multicast(dev);
-out:
+	netif_device_attach(dev);
+	if (netif_running(dev)) {
+		rc = nv_open(dev);
+		nv_set_multicast(dev);
+	}
 	return rc;
 }
+
+static void nv_shutdown(struct pci_dev *pdev)
+{
+	struct net_device *dev = pci_get_drvdata(pdev);
+	struct fe_priv *np = netdev_priv(dev);
+
+	if (netif_running(dev))
+		nv_close(dev);
+
+	pci_enable_wake(pdev, PCI_D3hot, np->wolenabled);
+	pci_enable_wake(pdev, PCI_D3cold, np->wolenabled);
+	pci_disable_device(pdev);
+	pci_set_power_state(pdev, PCI_D3hot);
+}
 #else
 #define nv_suspend NULL
+#define nv_shutdown NULL
 #define nv_resume NULL
 #endif /* CONFIG_PM */
 
@@ -5999,6 +6019,7 @@ static struct pci_driver driver = {
 	.remove		= __devexit_p(nv_remove),
 	.suspend	= nv_suspend,
 	.resume		= nv_resume,
+	.shutdown	= nv_shutdown,
 };
 
 static int __init init_nic(void)
diff --git a/drivers/net/fs_enet/fs_enet-main.c b/drivers/net/fs_enet/fs_enet-main.c
index a5baaf5..fb7c477 100644
--- a/drivers/net/fs_enet/fs_enet-main.c
+++ b/drivers/net/fs_enet/fs_enet-main.c
@@ -43,6 +43,7 @@
 #include <asm/uaccess.h>
 
 #ifdef CONFIG_PPC_CPM_NEW_BINDING
+#include <linux/of_gpio.h>
 #include <asm/of_platform.h>
 #endif
 
@@ -1172,8 +1173,7 @@ static int __devinit find_phy(struct device_node *np,
                               struct fs_platform_info *fpi)
 {
 	struct device_node *phynode, *mdionode;
-	struct resource res;
-	int ret = 0, len;
+	int ret = 0, len, bus_id;
 	const u32 *data;
 
 	data  = of_get_property(np, "fixed-link", NULL);
@@ -1190,19 +1190,28 @@ static int __devinit find_phy(struct device_node *np,
 	if (!phynode)
 		return -EINVAL;
 
-	mdionode = of_get_parent(phynode);
-	if (!mdionode)
+	data = of_get_property(phynode, "reg", &len);
+	if (!data || len != 4) {
+		ret = -EINVAL;
 		goto out_put_phy;
+	}
 
-	ret = of_address_to_resource(mdionode, 0, &res);
-	if (ret)
-		goto out_put_mdio;
+	mdionode = of_get_parent(phynode);
+	if (!mdionode) {
+		ret = -EINVAL;
+		goto out_put_phy;
+	}
 
-	data = of_get_property(phynode, "reg", &len);
-	if (!data || len != 4)
-		goto out_put_mdio;
+	bus_id = of_get_gpio(mdionode, 0);
+	if (bus_id < 0) {
+		struct resource res;
+		ret = of_address_to_resource(mdionode, 0, &res);
+		if (ret)
+			goto out_put_mdio;
+		bus_id = res.start;
+	}
 
-	snprintf(fpi->bus_id, 16, "%x:%02x", res.start, *data);
+	snprintf(fpi->bus_id, 16, "%x:%02x", bus_id, *data);
 
 out_put_mdio:
 	of_node_put(mdionode);
diff --git a/drivers/net/myri10ge/myri10ge.c b/drivers/net/myri10ge/myri10ge.c
index 36be6ef..93007d3 100644
--- a/drivers/net/myri10ge/myri10ge.c
+++ b/drivers/net/myri10ge/myri10ge.c
@@ -49,6 +49,7 @@
 #include <linux/if_ether.h>
 #include <linux/if_vlan.h>
 #include <linux/inet_lro.h>
+#include <linux/dca.h>
 #include <linux/ip.h>
 #include <linux/inet.h>
 #include <linux/in.h>
@@ -185,11 +186,18 @@ struct myri10ge_slice_state {
 	dma_addr_t fw_stats_bus;
 	int watchdog_tx_done;
 	int watchdog_tx_req;
+#ifdef CONFIG_DCA
+	int cached_dca_tag;
+	int cpu;
+	__be32 __iomem *dca_tag;
+#endif
+	char irq_desc[32];
 };
 
 struct myri10ge_priv {
-	struct myri10ge_slice_state ss;
+	struct myri10ge_slice_state *ss;
 	int tx_boundary;	/* boundary transmits cannot cross */
+	int num_slices;
 	int running;		/* running?             */
 	int csum_flag;		/* rx_csums?            */
 	int small_bytes;
@@ -208,6 +216,11 @@ struct myri10ge_priv {
 	dma_addr_t cmd_bus;
 	struct pci_dev *pdev;
 	int msi_enabled;
+	int msix_enabled;
+	struct msix_entry *msix_vectors;
+#ifdef CONFIG_DCA
+	int dca_enabled;
+#endif
 	u32 link_state;
 	unsigned int rdma_tags_available;
 	int intr_coal_delay;
@@ -244,6 +257,8 @@ struct myri10ge_priv {
 
 static char *myri10ge_fw_unaligned = "myri10ge_ethp_z8e.dat";
 static char *myri10ge_fw_aligned = "myri10ge_eth_z8e.dat";
+static char *myri10ge_fw_rss_unaligned = "myri10ge_rss_ethp_z8e.dat";
+static char *myri10ge_fw_rss_aligned = "myri10ge_rss_eth_z8e.dat";
 
 static char *myri10ge_fw_name = NULL;
 module_param(myri10ge_fw_name, charp, S_IRUGO | S_IWUSR);
@@ -321,6 +336,18 @@ static int myri10ge_wcfifo = 0;
 module_param(myri10ge_wcfifo, int, S_IRUGO);
 MODULE_PARM_DESC(myri10ge_wcfifo, "Enable WC Fifo when WC is enabled");
 
+static int myri10ge_max_slices = 1;
+module_param(myri10ge_max_slices, int, S_IRUGO);
+MODULE_PARM_DESC(myri10ge_max_slices, "Max tx/rx queues");
+
+static int myri10ge_rss_hash = MXGEFW_RSS_HASH_TYPE_SRC_PORT;
+module_param(myri10ge_rss_hash, int, S_IRUGO);
+MODULE_PARM_DESC(myri10ge_rss_hash, "Type of RSS hashing to do");
+
+static int myri10ge_dca = 1;
+module_param(myri10ge_dca, int, S_IRUGO);
+MODULE_PARM_DESC(myri10ge_dca, "Enable DCA if possible");
+
 #define MYRI10GE_FW_OFFSET 1024*1024
 #define MYRI10GE_HIGHPART_TO_U32(X) \
 (sizeof (X) == 8) ? ((u32)((u64)(X) >> 32)) : (0)
@@ -657,7 +684,7 @@ static int myri10ge_get_firmware_capabilities(struct myri10ge_priv *mgp)
 	return 0;
 }
 
-static int myri10ge_load_firmware(struct myri10ge_priv *mgp)
+static int myri10ge_load_firmware(struct myri10ge_priv *mgp, int adopt)
 {
 	char __iomem *submit;
 	__be32 buf[16] __attribute__ ((__aligned__(8)));
@@ -667,6 +694,8 @@ static int myri10ge_load_firmware(struct myri10ge_priv *mgp)
 	size = 0;
 	status = myri10ge_load_hotplug_firmware(mgp, &size);
 	if (status) {
+		if (!adopt)
+			return status;
 		dev_warn(&mgp->pdev->dev, "hotplug firmware loading failed\n");
 
 		/* Do not attempt to adopt firmware if there
@@ -859,8 +888,12 @@ abort:
 static int myri10ge_reset(struct myri10ge_priv *mgp)
 {
 	struct myri10ge_cmd cmd;
-	int status;
+	struct myri10ge_slice_state *ss;
+	int i, status;
 	size_t bytes;
+#ifdef CONFIG_DCA
+	unsigned long dca_tag_off;
+#endif
 
 	/* try to send a reset command to the card to see if it
 	 * is alive */
@@ -872,20 +905,74 @@ static int myri10ge_reset(struct myri10ge_priv *mgp)
 	}
 
 	(void)myri10ge_dma_test(mgp, MXGEFW_DMA_TEST);
+	/*
+	 * Use non-ndis mcp_slot (eg, 4 bytes total,
+	 * no toeplitz hash value returned.  Older firmware will
+	 * not understand this command, but will use the correct
+	 * sized mcp_slot, so we ignore error returns
+	 */
+	cmd.data0 = MXGEFW_RSS_MCP_SLOT_TYPE_MIN;
+	(void)myri10ge_send_cmd(mgp, MXGEFW_CMD_SET_RSS_MCP_SLOT_TYPE, &cmd, 0);
 
 	/* Now exchange information about interrupts  */
 
-	bytes = mgp->max_intr_slots * sizeof(*mgp->ss.rx_done.entry);
-	memset(mgp->ss.rx_done.entry, 0, bytes);
+	bytes = mgp->max_intr_slots * sizeof(*mgp->ss[0].rx_done.entry);
 	cmd.data0 = (u32) bytes;
 	status = myri10ge_send_cmd(mgp, MXGEFW_CMD_SET_INTRQ_SIZE, &cmd, 0);
-	cmd.data0 = MYRI10GE_LOWPART_TO_U32(mgp->ss.rx_done.bus);
-	cmd.data1 = MYRI10GE_HIGHPART_TO_U32(mgp->ss.rx_done.bus);
-	status |= myri10ge_send_cmd(mgp, MXGEFW_CMD_SET_INTRQ_DMA, &cmd, 0);
+
+	/*
+	 * Even though we already know how many slices are supported
+	 * via myri10ge_probe_slices() MXGEFW_CMD_GET_MAX_RSS_QUEUES
+	 * has magic side effects, and must be called after a reset.
+	 * It must be called prior to calling any RSS related cmds,
+	 * including assigning an interrupt queue for anything but
+	 * slice 0.  It must also be called *after*
+	 * MXGEFW_CMD_SET_INTRQ_SIZE, since the intrq size is used by
+	 * the firmware to compute offsets.
+	 */
+
+	if (mgp->num_slices > 1) {
+
+		/* ask the maximum number of slices it supports */
+		status = myri10ge_send_cmd(mgp, MXGEFW_CMD_GET_MAX_RSS_QUEUES,
+					   &cmd, 0);
+		if (status != 0) {
+			dev_err(&mgp->pdev->dev,
+				"failed to get number of slices\n");
+		}
+
+		/*
+		 * MXGEFW_CMD_ENABLE_RSS_QUEUES must be called prior
+		 * to setting up the interrupt queue DMA
+		 */
+
+		cmd.data0 = mgp->num_slices;
+		cmd.data1 = 1;	/* use MSI-X */
+		status = myri10ge_send_cmd(mgp, MXGEFW_CMD_ENABLE_RSS_QUEUES,
+					   &cmd, 0);
+		if (status != 0) {
+			dev_err(&mgp->pdev->dev,
+				"failed to set number of slices\n");
+
+			return status;
+		}
+	}
+	for (i = 0; i < mgp->num_slices; i++) {
+		ss = &mgp->ss[i];
+		cmd.data0 = MYRI10GE_LOWPART_TO_U32(ss->rx_done.bus);
+		cmd.data1 = MYRI10GE_HIGHPART_TO_U32(ss->rx_done.bus);
+		cmd.data2 = i;
+		status |= myri10ge_send_cmd(mgp, MXGEFW_CMD_SET_INTRQ_DMA,
+					    &cmd, 0);
+	};
 
 	status |=
 	    myri10ge_send_cmd(mgp, MXGEFW_CMD_GET_IRQ_ACK_OFFSET, &cmd, 0);
-	mgp->ss.irq_claim = (__iomem __be32 *) (mgp->sram + cmd.data0);
+	for (i = 0; i < mgp->num_slices; i++) {
+		ss = &mgp->ss[i];
+		ss->irq_claim =
+		    (__iomem __be32 *) (mgp->sram + cmd.data0 + 8 * i);
+	}
 	status |= myri10ge_send_cmd(mgp, MXGEFW_CMD_GET_IRQ_DEASSERT_OFFSET,
 				    &cmd, 0);
 	mgp->irq_deassert = (__iomem __be32 *) (mgp->sram + cmd.data0);
@@ -899,24 +986,116 @@ static int myri10ge_reset(struct myri10ge_priv *mgp)
 	}
 	put_be32(htonl(mgp->intr_coal_delay), mgp->intr_coal_delay_ptr);
 
-	memset(mgp->ss.rx_done.entry, 0, bytes);
+#ifdef CONFIG_DCA
+	status = myri10ge_send_cmd(mgp, MXGEFW_CMD_GET_DCA_OFFSET, &cmd, 0);
+	dca_tag_off = cmd.data0;
+	for (i = 0; i < mgp->num_slices; i++) {
+		ss = &mgp->ss[i];
+		if (status == 0) {
+			ss->dca_tag = (__iomem __be32 *)
+			    (mgp->sram + dca_tag_off + 4 * i);
+		} else {
+			ss->dca_tag = NULL;
+		}
+	}
+#endif				/* CONFIG_DCA */
 
 	/* reset mcp/driver shared state back to 0 */
-	mgp->ss.tx.req = 0;
-	mgp->ss.tx.done = 0;
-	mgp->ss.tx.pkt_start = 0;
-	mgp->ss.tx.pkt_done = 0;
-	mgp->ss.rx_big.cnt = 0;
-	mgp->ss.rx_small.cnt = 0;
-	mgp->ss.rx_done.idx = 0;
-	mgp->ss.rx_done.cnt = 0;
+
 	mgp->link_changes = 0;
+	for (i = 0; i < mgp->num_slices; i++) {
+		ss = &mgp->ss[i];
+
+		memset(ss->rx_done.entry, 0, bytes);
+		ss->tx.req = 0;
+		ss->tx.done = 0;
+		ss->tx.pkt_start = 0;
+		ss->tx.pkt_done = 0;
+		ss->rx_big.cnt = 0;
+		ss->rx_small.cnt = 0;
+		ss->rx_done.idx = 0;
+		ss->rx_done.cnt = 0;
+		ss->tx.wake_queue = 0;
+		ss->tx.stop_queue = 0;
+	}
+
 	status = myri10ge_update_mac_address(mgp, mgp->dev->dev_addr);
 	myri10ge_change_pause(mgp, mgp->pause);
 	myri10ge_set_multicast_list(mgp->dev);
 	return status;
 }
 
+#ifdef CONFIG_DCA
+static void
+myri10ge_write_dca(struct myri10ge_slice_state *ss, int cpu, int tag)
+{
+	ss->cpu = cpu;
+	ss->cached_dca_tag = tag;
+	put_be32(htonl(tag), ss->dca_tag);
+}
+
+static inline void myri10ge_update_dca(struct myri10ge_slice_state *ss)
+{
+	int cpu = get_cpu();
+	int tag;
+
+	if (cpu != ss->cpu) {
+		tag = dca_get_tag(cpu);
+		if (ss->cached_dca_tag != tag)
+			myri10ge_write_dca(ss, cpu, tag);
+	}
+	put_cpu();
+}
+
+static void myri10ge_setup_dca(struct myri10ge_priv *mgp)
+{
+	int err, i;
+	struct pci_dev *pdev = mgp->pdev;
+
+	if (mgp->ss[0].dca_tag == NULL || mgp->dca_enabled)
+		return;
+	if (!myri10ge_dca) {
+		dev_err(&pdev->dev, "dca disabled by administrator\n");
+		return;
+	}
+	err = dca_add_requester(&pdev->dev);
+	if (err) {
+		dev_err(&pdev->dev,
+			"dca_add_requester() failed, err=%d\n", err);
+		return;
+	}
+	mgp->dca_enabled = 1;
+	for (i = 0; i < mgp->num_slices; i++)
+		myri10ge_write_dca(&mgp->ss[i], -1, 0);
+}
+
+static void myri10ge_teardown_dca(struct myri10ge_priv *mgp)
+{
+	struct pci_dev *pdev = mgp->pdev;
+	int err;
+
+	if (!mgp->dca_enabled)
+		return;
+	mgp->dca_enabled = 0;
+	err = dca_remove_requester(&pdev->dev);
+}
+
+static int myri10ge_notify_dca_device(struct device *dev, void *data)
+{
+	struct myri10ge_priv *mgp;
+	unsigned long event;
+
+	mgp = dev_get_drvdata(dev);
+	event = *(unsigned long *)data;
+
+	if (event == DCA_PROVIDER_ADD)
+		myri10ge_setup_dca(mgp);
+	else if (event == DCA_PROVIDER_REMOVE)
+		myri10ge_teardown_dca(mgp);
+	return 0;
+}
+#endif				/* CONFIG_DCA */
+
 static inline void
 myri10ge_submit_8rx(struct mcp_kreq_ether_recv __iomem * dst,
 		    struct mcp_kreq_ether_recv *src)
@@ -1095,9 +1274,10 @@ myri10ge_rx_done(struct myri10ge_slice_state *ss, struct myri10ge_rx_buf *rx,
 		rx_frags[0].size -= MXGEFW_PAD;
 		len -= MXGEFW_PAD;
 		lro_receive_frags(&ss->rx_done.lro_mgr, rx_frags,
-				  len, len,
 				  /* opaque, will come back in get_frag_header */
+				  len, len,
 				  (void *)(__force unsigned long)csum, csum);
+
 		return 1;
 	}
 
@@ -1236,7 +1416,7 @@ myri10ge_clean_rx_done(struct myri10ge_slice_state *ss, int budget)
 
 static inline void myri10ge_check_statblock(struct myri10ge_priv *mgp)
 {
-	struct mcp_irq_data *stats = mgp->ss.fw_stats;
+	struct mcp_irq_data *stats = mgp->ss[0].fw_stats;
 
 	if (unlikely(stats->stats_updated)) {
 		unsigned link_up = ntohl(stats->link_up);
@@ -1283,6 +1463,11 @@ static int myri10ge_poll(struct napi_struct *napi, int budget)
 	struct net_device *netdev = ss->mgp->dev;
 	int work_done;
 
+#ifdef CONFIG_DCA
+	if (ss->mgp->dca_enabled)
+		myri10ge_update_dca(ss);
+#endif
+
 	/* process as many rx events as NAPI will allow */
 	work_done = myri10ge_clean_rx_done(ss, budget);
 
@@ -1302,6 +1487,13 @@ static irqreturn_t myri10ge_intr(int irq, void *arg)
 	u32 send_done_count;
 	int i;
 
+	/* an interrupt on a non-zero slice is implicitly valid
+	 * since MSI-X irqs are not shared */
+	if (ss != mgp->ss) {
+		netif_rx_schedule(ss->dev, &ss->napi);
+		return (IRQ_HANDLED);
+	}
+
 	/* make sure it is our IRQ, and that the DMA has finished */
 	if (unlikely(!stats->valid))
 		return (IRQ_NONE);
@@ -1311,7 +1503,7 @@ static irqreturn_t myri10ge_intr(int irq, void *arg)
 	if (stats->valid & 1)
 		netif_rx_schedule(ss->dev, &ss->napi);
 
-	if (!mgp->msi_enabled) {
+	if (!mgp->msi_enabled && !mgp->msix_enabled) {
 		put_be32(0, mgp->irq_deassert);
 		if (!myri10ge_deassert_wait)
 			stats->valid = 0;
@@ -1446,10 +1638,10 @@ myri10ge_get_ringparam(struct net_device *netdev,
 {
 	struct myri10ge_priv *mgp = netdev_priv(netdev);
 
-	ring->rx_mini_max_pending = mgp->ss.rx_small.mask + 1;
-	ring->rx_max_pending = mgp->ss.rx_big.mask + 1;
+	ring->rx_mini_max_pending = mgp->ss[0].rx_small.mask + 1;
+	ring->rx_max_pending = mgp->ss[0].rx_big.mask + 1;
 	ring->rx_jumbo_max_pending = 0;
-	ring->tx_max_pending = mgp->ss.rx_small.mask + 1;
+	ring->tx_max_pending = mgp->ss[0].rx_small.mask + 1;
 	ring->rx_mini_pending = ring->rx_mini_max_pending;
 	ring->rx_pending = ring->rx_max_pending;
 	ring->rx_jumbo_pending = ring->rx_jumbo_max_pending;
@@ -1497,9 +1689,12 @@ static const char myri10ge_gstrings_main_stats[][ETH_GSTRING_LEN] = {
 	"tx_aborted_errors", "tx_carrier_errors", "tx_fifo_errors",
 	"tx_heartbeat_errors", "tx_window_errors",
 	/* device-specific stats */
-	"tx_boundary", "WC", "irq", "MSI",
+	"tx_boundary", "WC", "irq", "MSI", "MSIX",
 	"read_dma_bw_MBs", "write_dma_bw_MBs", "read_write_dma_bw_MBs",
 	"serial_number", "watchdog_resets",
+#ifdef CONFIG_DCA
+	"dca_capable", "dca_enabled",
+#endif
 	"link_changes", "link_up", "dropped_link_overflow",
 	"dropped_link_error_or_filtered",
 	"dropped_pause", "dropped_bad_phy", "dropped_bad_crc32",
@@ -1524,23 +1719,31 @@ static const char myri10ge_gstrings_slice_stats[][ETH_GSTRING_LEN] = {
 static void
 myri10ge_get_strings(struct net_device *netdev, u32 stringset, u8 * data)
 {
+	struct myri10ge_priv *mgp = netdev_priv(netdev);
+	int i;
+
 	switch (stringset) {
 	case ETH_SS_STATS:
 		memcpy(data, *myri10ge_gstrings_main_stats,
 		       sizeof(myri10ge_gstrings_main_stats));
 		data += sizeof(myri10ge_gstrings_main_stats);
-		memcpy(data, *myri10ge_gstrings_slice_stats,
-		       sizeof(myri10ge_gstrings_slice_stats));
-		data += sizeof(myri10ge_gstrings_slice_stats);
+		for (i = 0; i < mgp->num_slices; i++) {
+			memcpy(data, *myri10ge_gstrings_slice_stats,
+			       sizeof(myri10ge_gstrings_slice_stats));
+			data += sizeof(myri10ge_gstrings_slice_stats);
+		}
 		break;
 	}
 }
 
 static int myri10ge_get_sset_count(struct net_device *netdev, int sset)
 {
+	struct myri10ge_priv *mgp = netdev_priv(netdev);
+
 	switch (sset) {
 	case ETH_SS_STATS:
-		return MYRI10GE_MAIN_STATS_LEN + MYRI10GE_SLICE_STATS_LEN;
+		return MYRI10GE_MAIN_STATS_LEN +
+		    mgp->num_slices * MYRI10GE_SLICE_STATS_LEN;
 	default:
 		return -EOPNOTSUPP;
 	}
@@ -1552,6 +1755,7 @@ myri10ge_get_ethtool_stats(struct net_device *netdev,
 {
 	struct myri10ge_priv *mgp = netdev_priv(netdev);
 	struct myri10ge_slice_state *ss;
+	int slice;
 	int i;
 
 	for (i = 0; i < MYRI10GE_NET_STATS_LEN; i++)
@@ -1561,15 +1765,20 @@ myri10ge_get_ethtool_stats(struct net_device *netdev,
 	data[i++] = (unsigned int)mgp->wc_enabled;
 	data[i++] = (unsigned int)mgp->pdev->irq;
 	data[i++] = (unsigned int)mgp->msi_enabled;
+	data[i++] = (unsigned int)mgp->msix_enabled;
 	data[i++] = (unsigned int)mgp->read_dma;
 	data[i++] = (unsigned int)mgp->write_dma;
 	data[i++] = (unsigned int)mgp->read_write_dma;
 	data[i++] = (unsigned int)mgp->serial_number;
 	data[i++] = (unsigned int)mgp->watchdog_resets;
+#ifdef CONFIG_DCA
+	data[i++] = (unsigned int)(mgp->ss[0].dca_tag != NULL);
+	data[i++] = (unsigned int)(mgp->dca_enabled);
+#endif
 	data[i++] = (unsigned int)mgp->link_changes;
 
 	/* firmware stats are useful only in the first slice */
-	ss = &mgp->ss;
+	ss = &mgp->ss[0];
 	data[i++] = (unsigned int)ntohl(ss->fw_stats->link_up);
 	data[i++] = (unsigned int)ntohl(ss->fw_stats->dropped_link_overflow);
 	data[i++] =
@@ -1585,24 +1794,27 @@ myri10ge_get_ethtool_stats(struct net_device *netdev,
 	data[i++] = (unsigned int)ntohl(ss->fw_stats->dropped_no_small_buffer);
 	data[i++] = (unsigned int)ntohl(ss->fw_stats->dropped_no_big_buffer);
 
-	data[i++] = 0;
-	data[i++] = (unsigned int)ss->tx.pkt_start;
-	data[i++] = (unsigned int)ss->tx.pkt_done;
-	data[i++] = (unsigned int)ss->tx.req;
-	data[i++] = (unsigned int)ss->tx.done;
-	data[i++] = (unsigned int)ss->rx_small.cnt;
-	data[i++] = (unsigned int)ss->rx_big.cnt;
-	data[i++] = (unsigned int)ss->tx.wake_queue;
-	data[i++] = (unsigned int)ss->tx.stop_queue;
-	data[i++] = (unsigned int)ss->tx.linearized;
-	data[i++] = ss->rx_done.lro_mgr.stats.aggregated;
-	data[i++] = ss->rx_done.lro_mgr.stats.flushed;
-	if (ss->rx_done.lro_mgr.stats.flushed)
-		data[i++] = ss->rx_done.lro_mgr.stats.aggregated /
-		    ss->rx_done.lro_mgr.stats.flushed;
-	else
-		data[i++] = 0;
-	data[i++] = ss->rx_done.lro_mgr.stats.no_desc;
+	for (slice = 0; slice < mgp->num_slices; slice++) {
+		ss = &mgp->ss[slice];
+		data[i++] = slice;
+		data[i++] = (unsigned int)ss->tx.pkt_start;
+		data[i++] = (unsigned int)ss->tx.pkt_done;
+		data[i++] = (unsigned int)ss->tx.req;
+		data[i++] = (unsigned int)ss->tx.done;
+		data[i++] = (unsigned int)ss->rx_small.cnt;
+		data[i++] = (unsigned int)ss->rx_big.cnt;
+		data[i++] = (unsigned int)ss->tx.wake_queue;
+		data[i++] = (unsigned int)ss->tx.stop_queue;
+		data[i++] = (unsigned int)ss->tx.linearized;
+		data[i++] = ss->rx_done.lro_mgr.stats.aggregated;
+		data[i++] = ss->rx_done.lro_mgr.stats.flushed;
+		if (ss->rx_done.lro_mgr.stats.flushed)
+			data[i++] = ss->rx_done.lro_mgr.stats.aggregated /
+			    ss->rx_done.lro_mgr.stats.flushed;
+		else
+			data[i++] = 0;
+		data[i++] = ss->rx_done.lro_mgr.stats.no_desc;
+	}
 }
 
 static void myri10ge_set_msglevel(struct net_device *netdev, u32 value)
@@ -1645,12 +1857,15 @@ static int myri10ge_allocate_rings(struct myri10ge_slice_state *ss)
 	struct net_device *dev = mgp->dev;
 	int tx_ring_size, rx_ring_size;
 	int tx_ring_entries, rx_ring_entries;
-	int i, status;
+	int i, slice, status;
 	size_t bytes;
 
 	/* get ring sizes */
+	slice = ss - mgp->ss;
+	cmd.data0 = slice;
 	status = myri10ge_send_cmd(mgp, MXGEFW_CMD_GET_SEND_RING_SIZE, &cmd, 0);
 	tx_ring_size = cmd.data0;
+	cmd.data0 = slice;
 	status |= myri10ge_send_cmd(mgp, MXGEFW_CMD_GET_RX_RING_SIZE, &cmd, 0);
 	if (status != 0)
 		return status;
@@ -1715,15 +1930,17 @@ static int myri10ge_allocate_rings(struct myri10ge_slice_state *ss)
 				mgp->small_bytes + MXGEFW_PAD, 0);
 
 	if (ss->rx_small.fill_cnt < ss->rx_small.mask + 1) {
-		printk(KERN_ERR "myri10ge: %s: alloced only %d small bufs\n",
-		       dev->name, ss->rx_small.fill_cnt);
+		printk(KERN_ERR
+		       "myri10ge: %s:slice-%d: alloced only %d small bufs\n",
+		       dev->name, slice, ss->rx_small.fill_cnt);
 		goto abort_with_rx_small_ring;
 	}
 
 	myri10ge_alloc_rx_pages(mgp, &ss->rx_big, mgp->big_bytes, 0);
 	if (ss->rx_big.fill_cnt < ss->rx_big.mask + 1) {
-		printk(KERN_ERR "myri10ge: %s: alloced only %d big bufs\n",
-		       dev->name, ss->rx_big.fill_cnt);
+		printk(KERN_ERR
+		       "myri10ge: %s:slice-%d: alloced only %d big bufs\n",
+		       dev->name, slice, ss->rx_big.fill_cnt);
 		goto abort_with_rx_big_ring;
 	}
 
@@ -1775,6 +1992,10 @@ static void myri10ge_free_rings(struct myri10ge_slice_state *ss)
 	struct myri10ge_tx_buf *tx;
 	int i, len, idx;
 
+	/* If not allocated, skip it */
+	if (ss->tx.req_list == NULL)
+		return;
+
 	for (i = ss->rx_big.cnt; i < ss->rx_big.fill_cnt; i++) {
 		idx = i & ss->rx_big.mask;
 		if (i == ss->rx_big.fill_cnt - 1)
@@ -1837,25 +2058,67 @@ static void myri10ge_free_rings(struct myri10ge_slice_state *ss)
 static int myri10ge_request_irq(struct myri10ge_priv *mgp)
 {
 	struct pci_dev *pdev = mgp->pdev;
+	struct myri10ge_slice_state *ss;
+	struct net_device *netdev = mgp->dev;
+	int i;
 	int status;
 
+	mgp->msi_enabled = 0;
+	mgp->msix_enabled = 0;
+	status = 0;
 	if (myri10ge_msi) {
-		status = pci_enable_msi(pdev);
-		if (status != 0)
-			dev_err(&pdev->dev,
-				"Error %d setting up MSI; falling back to xPIC\n",
-				status);
-		else
-			mgp->msi_enabled = 1;
-	} else {
-		mgp->msi_enabled = 0;
+		if (mgp->num_slices > 1) {
+			status =
+			    pci_enable_msix(pdev, mgp->msix_vectors,
+					    mgp->num_slices);
+			if (status == 0) {
+				mgp->msix_enabled = 1;
+			} else {
+				dev_err(&pdev->dev,
+					"Error %d setting up MSI-X\n", status);
+				return status;
+			}
+		}
+		if (mgp->msix_enabled == 0) {
+			status = pci_enable_msi(pdev);
+			if (status != 0) {
+				dev_err(&pdev->dev,
+					"Error %d setting up MSI; falling back to xPIC\n",
+					status);
+			} else {
+				mgp->msi_enabled = 1;
+			}
+		}
 	}
-	status = request_irq(pdev->irq, myri10ge_intr, IRQF_SHARED,
-			     mgp->dev->name, mgp);
-	if (status != 0) {
-		dev_err(&pdev->dev, "failed to allocate IRQ\n");
-		if (mgp->msi_enabled)
-			pci_disable_msi(pdev);
+	if (mgp->msix_enabled) {
+		for (i = 0; i < mgp->num_slices; i++) {
+			ss = &mgp->ss[i];
+			snprintf(ss->irq_desc, sizeof(ss->irq_desc),
+				 "%s:slice-%d", netdev->name, i);
+			status = request_irq(mgp->msix_vectors[i].vector,
+					     myri10ge_intr, 0, ss->irq_desc,
+					     ss);
+			if (status != 0) {
+				dev_err(&pdev->dev,
+					"slice %d failed to allocate IRQ\n", i);
+				i--;
+				while (i >= 0) {
+					free_irq(mgp->msix_vectors[i].vector,
+						 &mgp->ss[i]);
+					i--;
+				}
+				pci_disable_msix(pdev);
+				return status;
+			}
+		}
+	} else {
+		status = request_irq(pdev->irq, myri10ge_intr, IRQF_SHARED,
+				     mgp->dev->name, &mgp->ss[0]);
+		if (status != 0) {
+			dev_err(&pdev->dev, "failed to allocate IRQ\n");
+			if (mgp->msi_enabled)
+				pci_disable_msi(pdev);
+		}
 	}
 	return status;
 }
@@ -1863,10 +2126,18 @@ static int myri10ge_request_irq(struct myri10ge_priv *mgp)
 static void myri10ge_free_irq(struct myri10ge_priv *mgp)
 {
 	struct pci_dev *pdev = mgp->pdev;
+	int i;
 
-	free_irq(pdev->irq, mgp);
+	if (mgp->msix_enabled) {
+		for (i = 0; i < mgp->num_slices; i++)
+			free_irq(mgp->msix_vectors[i].vector, &mgp->ss[i]);
+	} else {
+		free_irq(pdev->irq, &mgp->ss[0]);
+	}
 	if (mgp->msi_enabled)
 		pci_disable_msi(pdev);
+	if (mgp->msix_enabled)
+		pci_disable_msix(pdev);
 }
 
 static int
@@ -1928,12 +2199,82 @@ myri10ge_get_frag_header(struct skb_frag_struct *frag, void **mac_hdr,
 	return 0;
 }
 
+static int myri10ge_get_txrx(struct myri10ge_priv *mgp, int slice)
+{
+	struct myri10ge_cmd cmd;
+	struct myri10ge_slice_state *ss;
+	int status;
+
+	ss = &mgp->ss[slice];
+	cmd.data0 = 0;		/* single slice for now */
+	status = myri10ge_send_cmd(mgp, MXGEFW_CMD_GET_SEND_OFFSET, &cmd, 0);
+	ss->tx.lanai = (struct mcp_kreq_ether_send __iomem *)
+	    (mgp->sram + cmd.data0);
+
+	cmd.data0 = slice;
+	status |= myri10ge_send_cmd(mgp, MXGEFW_CMD_GET_SMALL_RX_OFFSET,
+				    &cmd, 0);
+	ss->rx_small.lanai = (struct mcp_kreq_ether_recv __iomem *)
+	    (mgp->sram + cmd.data0);
+
+	cmd.data0 = slice;
+	status |= myri10ge_send_cmd(mgp, MXGEFW_CMD_GET_BIG_RX_OFFSET, &cmd, 0);
+	ss->rx_big.lanai = (struct mcp_kreq_ether_recv __iomem *)
+	    (mgp->sram + cmd.data0);
+
+	if (myri10ge_wcfifo && mgp->wc_enabled) {
+		ss->tx.wc_fifo = (u8 __iomem *)
+		    mgp->sram + MXGEFW_ETH_SEND_4 + 64 * slice;
+		ss->rx_small.wc_fifo = (u8 __iomem *)
+		    mgp->sram + MXGEFW_ETH_RECV_SMALL + 64 * slice;
+		ss->rx_big.wc_fifo = (u8 __iomem *)
+		    mgp->sram + MXGEFW_ETH_RECV_BIG + 64 * slice;
+	} else {
+		ss->tx.wc_fifo = NULL;
+		ss->rx_small.wc_fifo = NULL;
+		ss->rx_big.wc_fifo = NULL;
+	}
+	return status;
+
+}
+
+static int myri10ge_set_stats(struct myri10ge_priv *mgp, int slice)
+{
+	struct myri10ge_cmd cmd;
+	struct myri10ge_slice_state *ss;
+	int status;
+
+	ss = &mgp->ss[slice];
+	cmd.data0 = MYRI10GE_LOWPART_TO_U32(ss->fw_stats_bus);
+	cmd.data1 = MYRI10GE_HIGHPART_TO_U32(ss->fw_stats_bus);
+	cmd.data2 = sizeof(struct mcp_irq_data);
+	status = myri10ge_send_cmd(mgp, MXGEFW_CMD_SET_STATS_DMA_V2, &cmd, 0);
+	if (status == -ENOSYS) {
+		dma_addr_t bus = ss->fw_stats_bus;
+		if (slice != 0)
+			return -EINVAL;
+		bus += offsetof(struct mcp_irq_data, send_done_count);
+		cmd.data0 = MYRI10GE_LOWPART_TO_U32(bus);
+		cmd.data1 = MYRI10GE_HIGHPART_TO_U32(bus);
+		status = myri10ge_send_cmd(mgp,
+					   MXGEFW_CMD_SET_STATS_DMA_OBSOLETE,
+					   &cmd, 0);
+		/* Firmware cannot support multicast without STATS_DMA_V2 */
+		mgp->fw_multicast_support = 0;
+	} else {
+		mgp->fw_multicast_support = 1;
+	}
+	return 0;
+}
+
 static int myri10ge_open(struct net_device *dev)
 {
+	struct myri10ge_slice_state *ss;
 	struct myri10ge_priv *mgp = netdev_priv(dev);
 	struct myri10ge_cmd cmd;
+	int i, status, big_pow2, slice;
+	u8 *itable;
 	struct net_lro_mgr *lro_mgr;
-	int status, big_pow2;
 
 	if (mgp->running != MYRI10GE_ETH_STOPPED)
 		return -EBUSY;
@@ -1945,6 +2286,48 @@ static int myri10ge_open(struct net_device *dev)
 		goto abort_with_nothing;
 	}
 
+	if (mgp->num_slices > 1) {
+		cmd.data0 = mgp->num_slices;
+		cmd.data1 = 1;	/* use MSI-X */
+		status = myri10ge_send_cmd(mgp, MXGEFW_CMD_ENABLE_RSS_QUEUES,
+					   &cmd, 0);
+		if (status != 0) {
+			printk(KERN_ERR
+			       "myri10ge: %s: failed to set number of slices\n",
+			       dev->name);
+			goto abort_with_nothing;
+		}
+		/* setup the indirection table */
+		cmd.data0 = mgp->num_slices;
+		status = myri10ge_send_cmd(mgp, MXGEFW_CMD_SET_RSS_TABLE_SIZE,
+					   &cmd, 0);
+
+		status |= myri10ge_send_cmd(mgp,
+					    MXGEFW_CMD_GET_RSS_TABLE_OFFSET,
+					    &cmd, 0);
+		if (status != 0) {
+			printk(KERN_ERR
+			       "myri10ge: %s: failed to setup rss tables\n",
+			       dev->name);
+		}
+
+		/* just enable an identity mapping */
+		itable = mgp->sram + cmd.data0;
+		for (i = 0; i < mgp->num_slices; i++)
+			__raw_writeb(i, &itable[i]);
+
+		cmd.data0 = 1;
+		cmd.data1 = myri10ge_rss_hash;
+		status = myri10ge_send_cmd(mgp, MXGEFW_CMD_SET_RSS_ENABLE,
+					   &cmd, 0);
+		if (status != 0) {
+			printk(KERN_ERR
+			       "myri10ge: %s: failed to enable slices\n",
+			       dev->name);
+			goto abort_with_nothing;
+		}
+	}
+
 	status = myri10ge_request_irq(mgp);
 	if (status != 0)
 		goto abort_with_nothing;
@@ -1968,41 +2351,6 @@ static int myri10ge_open(struct net_device *dev)
 	if (myri10ge_small_bytes > 0)
 		mgp->small_bytes = myri10ge_small_bytes;
 
-	/* get the lanai pointers to the send and receive rings */
-
-	status |= myri10ge_send_cmd(mgp, MXGEFW_CMD_GET_SEND_OFFSET, &cmd, 0);
-	mgp->ss.tx.lanai =
-	    (struct mcp_kreq_ether_send __iomem *)(mgp->sram + cmd.data0);
-
-	status |=
-	    myri10ge_send_cmd(mgp, MXGEFW_CMD_GET_SMALL_RX_OFFSET, &cmd, 0);
-	mgp->ss.rx_small.lanai =
-	    (struct mcp_kreq_ether_recv __iomem *)(mgp->sram + cmd.data0);
-
-	status |= myri10ge_send_cmd(mgp, MXGEFW_CMD_GET_BIG_RX_OFFSET, &cmd, 0);
-	mgp->ss.rx_big.lanai =
-	    (struct mcp_kreq_ether_recv __iomem *)(mgp->sram + cmd.data0);
-
-	if (status != 0) {
-		printk(KERN_ERR
-		       "myri10ge: %s: failed to get ring sizes or locations\n",
-		       dev->name);
-		mgp->running = MYRI10GE_ETH_STOPPED;
-		goto abort_with_irq;
-	}
-
-	if (myri10ge_wcfifo && mgp->wc_enabled) {
-		mgp->ss.tx.wc_fifo = (u8 __iomem *) mgp->sram + MXGEFW_ETH_SEND_4;
-		mgp->ss.rx_small.wc_fifo =
-		    (u8 __iomem *) mgp->sram + MXGEFW_ETH_RECV_SMALL;
-		mgp->ss.rx_big.wc_fifo =
-		    (u8 __iomem *) mgp->sram + MXGEFW_ETH_RECV_BIG;
-	} else {
-		mgp->ss.tx.wc_fifo = NULL;
-		mgp->ss.rx_small.wc_fifo = NULL;
-		mgp->ss.rx_big.wc_fifo = NULL;
-	}
-
 	/* Firmware needs the big buff size as a power of 2.  Lie and
 	 * tell him the buffer is larger, because we only use 1
 	 * buffer/pkt, and the mtu will prevent overruns.
@@ -2017,9 +2365,44 @@ static int myri10ge_open(struct net_device *dev)
 		mgp->big_bytes = big_pow2;
 	}
 
-	status = myri10ge_allocate_rings(&mgp->ss);
-	if (status != 0)
-		goto abort_with_irq;
+	/* setup the per-slice data structures */
+	for (slice = 0; slice < mgp->num_slices; slice++) {
+		ss = &mgp->ss[slice];
+
+		status = myri10ge_get_txrx(mgp, slice);
+		if (status != 0) {
+			printk(KERN_ERR
+			       "myri10ge: %s: failed to get ring sizes or locations\n",
+			       dev->name);
+			goto abort_with_rings;
+		}
+		status = myri10ge_allocate_rings(ss);
+		if (status != 0)
+			goto abort_with_rings;
+		if (slice == 0)
+			status = myri10ge_set_stats(mgp, slice);
+		if (status) {
+			printk(KERN_ERR
+			       "myri10ge: %s: Couldn't set stats DMA\n",
+			       dev->name);
+			goto abort_with_rings;
+		}
+
+		lro_mgr = &ss->rx_done.lro_mgr;
+		lro_mgr->dev = dev;
+		lro_mgr->features = LRO_F_NAPI;
+		lro_mgr->ip_summed = CHECKSUM_COMPLETE;
+		lro_mgr->ip_summed_aggr = CHECKSUM_UNNECESSARY;
+		lro_mgr->max_desc = MYRI10GE_MAX_LRO_DESCRIPTORS;
+		lro_mgr->lro_arr = ss->rx_done.lro_desc;
+		lro_mgr->get_frag_header = myri10ge_get_frag_header;
+		lro_mgr->max_aggr = myri10ge_lro_max_pkts;
+		if (lro_mgr->max_aggr > MAX_SKB_FRAGS)
+			lro_mgr->max_aggr = MAX_SKB_FRAGS;
+
+		/* must happen prior to any irq */
+		napi_enable(&(ss)->napi);
+	}
 
 	/* now give firmware buffers sizes, and MTU */
 	cmd.data0 = dev->mtu + ETH_HLEN + VLAN_HLEN;
@@ -2036,25 +2419,15 @@ static int myri10ge_open(struct net_device *dev)
 		goto abort_with_rings;
 	}
 
-	cmd.data0 = MYRI10GE_LOWPART_TO_U32(mgp->ss.fw_stats_bus);
-	cmd.data1 = MYRI10GE_HIGHPART_TO_U32(mgp->ss.fw_stats_bus);
-	cmd.data2 = sizeof(struct mcp_irq_data);
-	status = myri10ge_send_cmd(mgp, MXGEFW_CMD_SET_STATS_DMA_V2, &cmd, 0);
-	if (status == -ENOSYS) {
-		dma_addr_t bus = mgp->ss.fw_stats_bus;
-		bus += offsetof(struct mcp_irq_data, send_done_count);
-		cmd.data0 = MYRI10GE_LOWPART_TO_U32(bus);
-		cmd.data1 = MYRI10GE_HIGHPART_TO_U32(bus);
-		status = myri10ge_send_cmd(mgp,
-					   MXGEFW_CMD_SET_STATS_DMA_OBSOLETE,
-					   &cmd, 0);
-		/* Firmware cannot support multicast without STATS_DMA_V2 */
-		mgp->fw_multicast_support = 0;
-	} else {
-		mgp->fw_multicast_support = 1;
-	}
-	if (status) {
-		printk(KERN_ERR "myri10ge: %s: Couldn't set stats DMA\n",
+	/*
+	 * Set Linux style TSO mode; this is needed only on newer
+	 *  firmware versions.  Older versions default to Linux
+	 *  style TSO
+	 */
+	cmd.data0 = 0;
+	status = myri10ge_send_cmd(mgp, MXGEFW_CMD_SET_TSO_MODE, &cmd, 0);
+	if (status && status != -ENOSYS) {
+		printk(KERN_ERR "myri10ge: %s: Couldn't set TSO mode\n",
 		       dev->name);
 		goto abort_with_rings;
 	}
@@ -2062,21 +2435,6 @@ static int myri10ge_open(struct net_device *dev)
 	mgp->link_state = ~0U;
 	mgp->rdma_tags_available = 15;
 
-	lro_mgr = &mgp->ss.rx_done.lro_mgr;
-	lro_mgr->dev = dev;
-	lro_mgr->features = LRO_F_NAPI;
-	lro_mgr->ip_summed = CHECKSUM_COMPLETE;
-	lro_mgr->ip_summed_aggr = CHECKSUM_UNNECESSARY;
-	lro_mgr->max_desc = MYRI10GE_MAX_LRO_DESCRIPTORS;
-	lro_mgr->lro_arr = mgp->ss.rx_done.lro_desc;
-	lro_mgr->get_frag_header = myri10ge_get_frag_header;
-	lro_mgr->max_aggr = myri10ge_lro_max_pkts;
-	lro_mgr->frag_align_pad = 2;
-	if (lro_mgr->max_aggr > MAX_SKB_FRAGS)
-		lro_mgr->max_aggr = MAX_SKB_FRAGS;
-
-	napi_enable(&mgp->ss.napi);	/* must happen prior to any irq */
-
 	status = myri10ge_send_cmd(mgp, MXGEFW_CMD_ETHERNET_UP, &cmd, 0);
 	if (status) {
 		printk(KERN_ERR "myri10ge: %s: Couldn't bring up link\n",
@@ -2084,8 +2442,6 @@ static int myri10ge_open(struct net_device *dev)
 		goto abort_with_rings;
 	}
 
-	mgp->ss.tx.wake_queue = 0;
-	mgp->ss.tx.stop_queue = 0;
 	mgp->running = MYRI10GE_ETH_RUNNING;
 	mgp->watchdog_timer.expires = jiffies + myri10ge_watchdog_timeout * HZ;
 	add_timer(&mgp->watchdog_timer);
@@ -2093,9 +2449,9 @@ static int myri10ge_open(struct net_device *dev)
 	return 0;
 
 abort_with_rings:
-	myri10ge_free_rings(&mgp->ss);
+	for (i = 0; i < mgp->num_slices; i++)
+		myri10ge_free_rings(&mgp->ss[i]);
 
-abort_with_irq:
 	myri10ge_free_irq(mgp);
 
 abort_with_nothing:
@@ -2108,16 +2464,19 @@ static int myri10ge_close(struct net_device *dev)
 	struct myri10ge_priv *mgp = netdev_priv(dev);
 	struct myri10ge_cmd cmd;
 	int status, old_down_cnt;
+	int i;
 
 	if (mgp->running != MYRI10GE_ETH_RUNNING)
 		return 0;
 
-	if (mgp->ss.tx.req_bytes == NULL)
+	if (mgp->ss[0].tx.req_bytes == NULL)
 		return 0;
 
 	del_timer_sync(&mgp->watchdog_timer);
 	mgp->running = MYRI10GE_ETH_STOPPING;
-	napi_disable(&mgp->ss.napi);
+	for (i = 0; i < mgp->num_slices; i++) {
+		napi_disable(&mgp->ss[i].napi);
+	}
 	netif_carrier_off(dev);
 	netif_stop_queue(dev);
 	old_down_cnt = mgp->down_cnt;
@@ -2133,7 +2492,8 @@ static int myri10ge_close(struct net_device *dev)
 
 	netif_tx_disable(dev);
 	myri10ge_free_irq(mgp);
-	myri10ge_free_rings(&mgp->ss);
+	for (i = 0; i < mgp->num_slices; i++)
+		myri10ge_free_rings(&mgp->ss[i]);
 
 	mgp->running = MYRI10GE_ETH_STOPPED;
 	return 0;
@@ -2254,7 +2614,7 @@ static int myri10ge_xmit(struct sk_buff *skb, struct net_device *dev)
 	u8 flags, odd_flag;
 
 	/* always transmit through slot 0 */
-	ss = &mgp->ss;
+	ss = mgp->ss;
 	tx = &ss->tx;
 again:
 	req = tx->req_list;
@@ -2559,7 +2919,21 @@ drop:
 static struct net_device_stats *myri10ge_get_stats(struct net_device *dev)
 {
 	struct myri10ge_priv *mgp = netdev_priv(dev);
-	return &mgp->stats;
+	struct myri10ge_slice_netstats *slice_stats;
+	struct net_device_stats *stats = &mgp->stats;
+	int i;
+
+	memset(stats, 0, sizeof(*stats));
+	for (i = 0; i < mgp->num_slices; i++) {
+		slice_stats = &mgp->ss[i].stats;
+		stats->rx_packets += slice_stats->rx_packets;
+		stats->tx_packets += slice_stats->tx_packets;
+		stats->rx_bytes += slice_stats->rx_bytes;
+		stats->tx_bytes += slice_stats->tx_bytes;
+		stats->rx_dropped += slice_stats->rx_dropped;
+		stats->tx_dropped += slice_stats->tx_dropped;
+	}
+	return stats;
 }
 
 static void myri10ge_set_multicast_list(struct net_device *dev)
@@ -2770,10 +3144,10 @@ static void myri10ge_enable_ecrc(struct myri10ge_priv *mgp)
  *
  * If the driver can neither enable ECRC nor verify that it has
  * already been enabled, then it must use a firmware image which works
- * around unaligned completion packets (myri10ge_ethp_z8e.dat), and it
+ * around unaligned completion packets (myri10ge_rss_ethp_z8e.dat), and it
  * should also ensure that it never gives the device a Read-DMA which is
  * larger than 2KB by setting the tx_boundary to 2KB.  If ECRC is
- * enabled, then the driver should use the aligned (myri10ge_eth_z8e.dat)
+ * enabled, then the driver should use the aligned (myri10ge_rss_eth_z8e.dat)
  * firmware image, and set tx_boundary to 4KB.
  */
 
@@ -2802,7 +3176,7 @@ static void myri10ge_firmware_probe(struct myri10ge_priv *mgp)
 	 * completions) in order to see if it works on this host.
 	 */
 	mgp->fw_name = myri10ge_fw_aligned;
-	status = myri10ge_load_firmware(mgp);
+	status = myri10ge_load_firmware(mgp, 1);
 	if (status != 0) {
 		goto abort;
 	}
@@ -2983,6 +3357,7 @@ static void myri10ge_watchdog(struct work_struct *work)
 	struct myri10ge_tx_buf *tx;
 	u32 reboot;
 	int status;
+	int i;
 	u16 cmd, vendor;
 
 	mgp->watchdog_resets++;
@@ -3030,20 +3405,26 @@ static void myri10ge_watchdog(struct work_struct *work)
 
 		printk(KERN_ERR "myri10ge: %s: device timeout, resetting\n",
 		       mgp->dev->name);
-		tx = &mgp->ss.tx;
-		printk(KERN_INFO "myri10ge: %s: %d %d %d %d %d\n",
-		       mgp->dev->name, tx->req, tx->done,
-		       tx->pkt_start, tx->pkt_done,
-		       (int)ntohl(mgp->ss.fw_stats->send_done_count));
-		msleep(2000);
-		printk(KERN_INFO "myri10ge: %s: %d %d %d %d %d\n",
-		       mgp->dev->name, tx->req, tx->done,
-		       tx->pkt_start, tx->pkt_done,
-		       (int)ntohl(mgp->ss.fw_stats->send_done_count));
+		for (i = 0; i < mgp->num_slices; i++) {
+			tx = &mgp->ss[i].tx;
+			printk(KERN_INFO
+			       "myri10ge: %s: (%d): %d %d %d %d %d\n",
+			       mgp->dev->name, i, tx->req, tx->done,
+			       tx->pkt_start, tx->pkt_done,
+			       (int)ntohl(mgp->ss[i].fw_stats->
+					  send_done_count));
+			msleep(2000);
+			printk(KERN_INFO
+			       "myri10ge: %s: (%d): %d %d %d %d %d\n",
+			       mgp->dev->name, i, tx->req, tx->done,
+			       tx->pkt_start, tx->pkt_done,
+			       (int)ntohl(mgp->ss[i].fw_stats->
+					  send_done_count));
+		}
 	}
 	rtnl_lock();
 	myri10ge_close(mgp->dev);
-	status = myri10ge_load_firmware(mgp);
+	status = myri10ge_load_firmware(mgp, 1);
 	if (status != 0)
 		printk(KERN_ERR "myri10ge: %s: failed to load firmware\n",
 		       mgp->dev->name);
@@ -3063,47 +3444,241 @@ static void myri10ge_watchdog_timer(unsigned long arg)
 {
 	struct myri10ge_priv *mgp;
 	struct myri10ge_slice_state *ss;
+	int i, reset_needed;
 	u32 rx_pause_cnt;
 
 	mgp = (struct myri10ge_priv *)arg;
 
-	rx_pause_cnt = ntohl(mgp->ss.fw_stats->dropped_pause);
+	rx_pause_cnt = ntohl(mgp->ss[0].fw_stats->dropped_pause);
+	for (i = 0, reset_needed = 0;
+	     i < mgp->num_slices && reset_needed == 0; ++i) {
+
+		ss = &mgp->ss[i];
+		if (ss->rx_small.watchdog_needed) {
+			myri10ge_alloc_rx_pages(mgp, &ss->rx_small,
+						mgp->small_bytes + MXGEFW_PAD,
+						1);
+			if (ss->rx_small.fill_cnt - ss->rx_small.cnt >=
+			    myri10ge_fill_thresh)
+				ss->rx_small.watchdog_needed = 0;
+		}
+		if (ss->rx_big.watchdog_needed) {
+			myri10ge_alloc_rx_pages(mgp, &ss->rx_big,
+						mgp->big_bytes, 1);
+			if (ss->rx_big.fill_cnt - ss->rx_big.cnt >=
+			    myri10ge_fill_thresh)
+				ss->rx_big.watchdog_needed = 0;
+		}
 
-	ss = &mgp->ss;
-	if (ss->rx_small.watchdog_needed) {
-		myri10ge_alloc_rx_pages(mgp, &ss->rx_small,
-					mgp->small_bytes + MXGEFW_PAD, 1);
-		if (ss->rx_small.fill_cnt - ss->rx_small.cnt >=
-		    myri10ge_fill_thresh)
-			ss->rx_small.watchdog_needed = 0;
-	}
-	if (ss->rx_big.watchdog_needed) {
-		myri10ge_alloc_rx_pages(mgp, &ss->rx_big, mgp->big_bytes, 1);
-		if (ss->rx_big.fill_cnt - ss->rx_big.cnt >=
-		    myri10ge_fill_thresh)
-			ss->rx_big.watchdog_needed = 0;
-	}
-
-	if (ss->tx.req != ss->tx.done &&
-	    ss->tx.done == ss->watchdog_tx_done &&
-	    ss->watchdog_tx_req != ss->watchdog_tx_done) {
-		/* nic seems like it might be stuck.. */
-		if (rx_pause_cnt != mgp->watchdog_pause) {
-			if (net_ratelimit())
-				printk(KERN_WARNING "myri10ge %s:"
-				       "TX paused, check link partner\n",
-				       mgp->dev->name);
-		} else {
-			schedule_work(&mgp->watchdog_work);
-			return;
+		if (ss->tx.req != ss->tx.done &&
+		    ss->tx.done == ss->watchdog_tx_done &&
+		    ss->watchdog_tx_req != ss->watchdog_tx_done) {
+			/* nic seems like it might be stuck.. */
+			if (rx_pause_cnt != mgp->watchdog_pause) {
+				if (net_ratelimit())
+					printk(KERN_WARNING "myri10ge %s:"
+					       "TX paused, check link partner\n",
+					       mgp->dev->name);
+			} else {
+				reset_needed = 1;
+			}
 		}
+		ss->watchdog_tx_done = ss->tx.done;
+		ss->watchdog_tx_req = ss->tx.req;
 	}
-	/* rearm timer */
-	mod_timer(&mgp->watchdog_timer,
-		  jiffies + myri10ge_watchdog_timeout * HZ);
-	ss->watchdog_tx_done = ss->tx.done;
-	ss->watchdog_tx_req = ss->tx.req;
 	mgp->watchdog_pause = rx_pause_cnt;
+
+	if (reset_needed) {
+		schedule_work(&mgp->watchdog_work);
+	} else {
+		/* rearm timer */
+		mod_timer(&mgp->watchdog_timer,
+			  jiffies + myri10ge_watchdog_timeout * HZ);
+	}
+}
+
+static void myri10ge_free_slices(struct myri10ge_priv *mgp)
+{
+	struct myri10ge_slice_state *ss;
+	struct pci_dev *pdev = mgp->pdev;
+	size_t bytes;
+	int i;
+
+	if (mgp->ss == NULL)
+		return;
+
+	for (i = 0; i < mgp->num_slices; i++) {
+		ss = &mgp->ss[i];
+		if (ss->rx_done.entry != NULL) {
+			bytes = mgp->max_intr_slots *
+			    sizeof(*ss->rx_done.entry);
+			dma_free_coherent(&pdev->dev, bytes,
+					  ss->rx_done.entry, ss->rx_done.bus);
+			ss->rx_done.entry = NULL;
+		}
+		if (ss->fw_stats != NULL) {
+			bytes = sizeof(*ss->fw_stats);
+			dma_free_coherent(&pdev->dev, bytes,
+					  ss->fw_stats, ss->fw_stats_bus);
+			ss->fw_stats = NULL;
+		}
+	}
+	kfree(mgp->ss);
+	mgp->ss = NULL;
+}
+
+static int myri10ge_alloc_slices(struct myri10ge_priv *mgp)
+{
+	struct myri10ge_slice_state *ss;
+	struct pci_dev *pdev = mgp->pdev;
+	size_t bytes;
+	int i;
+
+	bytes = sizeof(*mgp->ss) * mgp->num_slices;
+	mgp->ss = kzalloc(bytes, GFP_KERNEL);
+	if (mgp->ss == NULL) {
+		return -ENOMEM;
+	}
+
+	for (i = 0; i < mgp->num_slices; i++) {
+		ss = &mgp->ss[i];
+		bytes = mgp->max_intr_slots * sizeof(*ss->rx_done.entry);
+		ss->rx_done.entry = dma_alloc_coherent(&pdev->dev, bytes,
+						       &ss->rx_done.bus,
+						       GFP_KERNEL);
+		if (ss->rx_done.entry == NULL)
+			goto abort;
+		memset(ss->rx_done.entry, 0, bytes);
+		bytes = sizeof(*ss->fw_stats);
+		ss->fw_stats = dma_alloc_coherent(&pdev->dev, bytes,
+						  &ss->fw_stats_bus,
+						  GFP_KERNEL);
+		if (ss->fw_stats == NULL)
+			goto abort;
+		ss->mgp = mgp;
+		ss->dev = mgp->dev;
+		netif_napi_add(ss->dev, &ss->napi, myri10ge_poll,
+			       myri10ge_napi_weight);
+	}
+	return 0;
+abort:
+	myri10ge_free_slices(mgp);
+	return -ENOMEM;
+}
+
+/*
+ * This function determines the number of slices supported.
+ * The number slices is the minumum of the number of CPUS,
+ * the number of MSI-X irqs supported, the number of slices
+ * supported by the firmware
+ */
+static void myri10ge_probe_slices(struct myri10ge_priv *mgp)
+{
+	struct myri10ge_cmd cmd;
+	struct pci_dev *pdev = mgp->pdev;
+	char *old_fw;
+	int i, status, ncpus, msix_cap;
+
+	mgp->num_slices = 1;
+	msix_cap = pci_find_capability(pdev, PCI_CAP_ID_MSIX);
+	ncpus = num_online_cpus();
+
+	if (myri10ge_max_slices == 1 || msix_cap == 0 ||
+	    (myri10ge_max_slices == -1 && ncpus < 2))
+		return;
+
+	/* try to load the slice aware rss firmware */
+	old_fw = mgp->fw_name;
+	if (old_fw == myri10ge_fw_aligned)
+		mgp->fw_name = myri10ge_fw_rss_aligned;
+	else
+		mgp->fw_name = myri10ge_fw_rss_unaligned;
+	status = myri10ge_load_firmware(mgp, 0);
+	if (status != 0) {
+		dev_info(&pdev->dev, "Rss firmware not found\n");
+		return;
+	}
+
+	/* hit the board with a reset to ensure it is alive */
+	memset(&cmd, 0, sizeof(cmd));
+	status = myri10ge_send_cmd(mgp, MXGEFW_CMD_RESET, &cmd, 0);
+	if (status != 0) {
+		dev_err(&mgp->pdev->dev, "failed reset\n");
+		goto abort_with_fw;
+		return;
+	}
+
+	mgp->max_intr_slots = cmd.data0 / sizeof(struct mcp_slot);
+
+	/* tell it the size of the interrupt queues */
+	cmd.data0 = mgp->max_intr_slots * sizeof(struct mcp_slot);
+	status = myri10ge_send_cmd(mgp, MXGEFW_CMD_SET_INTRQ_SIZE, &cmd, 0);
+	if (status != 0) {
+		dev_err(&mgp->pdev->dev, "failed MXGEFW_CMD_SET_INTRQ_SIZE\n");
+		goto abort_with_fw;
+	}
+
+	/* ask the maximum number of slices it supports */
+	status = myri10ge_send_cmd(mgp, MXGEFW_CMD_GET_MAX_RSS_QUEUES, &cmd, 0);
+	if (status != 0)
+		goto abort_with_fw;
+	else
+		mgp->num_slices = cmd.data0;
+
+	/* Only allow multiple slices if MSI-X is usable */
+	if (!myri10ge_msi) {
+		goto abort_with_fw;
+	}
+
+	/* if the admin did not specify a limit to how many
+	 * slices we should use, cap it automatically to the
+	 * number of CPUs currently online */
+	if (myri10ge_max_slices == -1)
+		myri10ge_max_slices = ncpus;
+
+	if (mgp->num_slices > myri10ge_max_slices)
+		mgp->num_slices = myri10ge_max_slices;
+
+	/* Now try to allocate as many MSI-X vectors as we have
+	 * slices. We give up on MSI-X if we can only get a single
+	 * vector. */
+
+	mgp->msix_vectors = kzalloc(mgp->num_slices *
+				    sizeof(*mgp->msix_vectors), GFP_KERNEL);
+	if (mgp->msix_vectors == NULL)
+		goto disable_msix;
+	for (i = 0; i < mgp->num_slices; i++) {
+		mgp->msix_vectors[i].entry = i;
+	}
+
+	while (mgp->num_slices > 1) {
+		/* make sure it is a power of two */
+		while (!is_power_of_2(mgp->num_slices))
+			mgp->num_slices--;
+		if (mgp->num_slices == 1)
+			goto disable_msix;
+		status = pci_enable_msix(pdev, mgp->msix_vectors,
+					 mgp->num_slices);
+		if (status == 0) {
+			pci_disable_msix(pdev);
+			return;
+		}
+		if (status > 0)
+			mgp->num_slices = status;
+		else
+			goto disable_msix;
+	}
+
+disable_msix:
+	if (mgp->msix_vectors != NULL) {
+		kfree(mgp->msix_vectors);
+		mgp->msix_vectors = NULL;
+	}
+
+abort_with_fw:
+	mgp->num_slices = 1;
+	mgp->fw_name = old_fw;
+	myri10ge_load_firmware(mgp, 0);
 }
 
 static int myri10ge_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
@@ -3111,7 +3686,6 @@ static int myri10ge_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 	struct net_device *netdev;
 	struct myri10ge_priv *mgp;
 	struct device *dev = &pdev->dev;
-	size_t bytes;
 	int i;
 	int status = -ENXIO;
 	int dac_enabled;
@@ -3126,7 +3700,6 @@ static int myri10ge_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 
 	mgp = netdev_priv(netdev);
 	mgp->dev = netdev;
-	netif_napi_add(netdev, &mgp->ss.napi, myri10ge_poll, myri10ge_napi_weight);
 	mgp->pdev = pdev;
 	mgp->csum_flag = MXGEFW_FLAGS_CKSUM;
 	mgp->pause = myri10ge_flow_control;
@@ -3172,11 +3745,6 @@ static int myri10ge_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 	if (mgp->cmd == NULL)
 		goto abort_with_netdev;
 
-	mgp->ss.fw_stats = dma_alloc_coherent(&pdev->dev, sizeof(*mgp->ss.fw_stats),
-					   &mgp->ss.fw_stats_bus, GFP_KERNEL);
-	if (mgp->ss.fw_stats == NULL)
-		goto abort_with_cmd;
-
 	mgp->board_span = pci_resource_len(pdev, 0);
 	mgp->iomem_base = pci_resource_start(pdev, 0);
 	mgp->mtrr = -1;
@@ -3213,28 +3781,28 @@ static int myri10ge_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 	for (i = 0; i < ETH_ALEN; i++)
 		netdev->dev_addr[i] = mgp->mac_addr[i];
 
-	/* allocate rx done ring */
-	bytes = mgp->max_intr_slots * sizeof(*mgp->ss.rx_done.entry);
-	mgp->ss.rx_done.entry = dma_alloc_coherent(&pdev->dev, bytes,
-						&mgp->ss.rx_done.bus, GFP_KERNEL);
-	if (mgp->ss.rx_done.entry == NULL)
-		goto abort_with_ioremap;
-	memset(mgp->ss.rx_done.entry, 0, bytes);
-
 	myri10ge_select_firmware(mgp);
 
-	status = myri10ge_load_firmware(mgp);
+	status = myri10ge_load_firmware(mgp, 1);
 	if (status != 0) {
 		dev_err(&pdev->dev, "failed to load firmware\n");
-		goto abort_with_rx_done;
+		goto abort_with_ioremap;
+	}
+	myri10ge_probe_slices(mgp);
+	status = myri10ge_alloc_slices(mgp);
+	if (status != 0) {
+		dev_err(&pdev->dev, "failed to alloc slice state\n");
+		goto abort_with_firmware;
 	}
 
 	status = myri10ge_reset(mgp);
 	if (status != 0) {
 		dev_err(&pdev->dev, "failed reset\n");
-		goto abort_with_firmware;
+		goto abort_with_slices;
 	}
-
+#ifdef CONFIG_DCA
+	myri10ge_setup_dca(mgp);
+#endif
 	pci_set_drvdata(pdev, mgp);
 	if ((myri10ge_initial_mtu + ETH_HLEN) > MYRI10GE_MAX_ETHER_MTU)
 		myri10ge_initial_mtu = MYRI10GE_MAX_ETHER_MTU - ETH_HLEN;
@@ -3277,24 +3845,27 @@ static int myri10ge_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 		dev_err(&pdev->dev, "register_netdev failed: %d\n", status);
 		goto abort_with_state;
 	}
-	dev_info(dev, "%s IRQ %d, tx bndry %d, fw %s, WC %s\n",
-		 (mgp->msi_enabled ? "MSI" : "xPIC"),
-		 netdev->irq, mgp->tx_boundary, mgp->fw_name,
-		 (mgp->wc_enabled ? "Enabled" : "Disabled"));
+	if (mgp->msix_enabled)
+		dev_info(dev, "%d MSI-X IRQs, tx bndry %d, fw %s, WC %s\n",
+			 mgp->num_slices, mgp->tx_boundary, mgp->fw_name,
+			 (mgp->wc_enabled ? "Enabled" : "Disabled"));
+	else
+		dev_info(dev, "%s IRQ %d, tx bndry %d, fw %s, WC %s\n",
+			 mgp->msi_enabled ? "MSI" : "xPIC",
+			 netdev->irq, mgp->tx_boundary, mgp->fw_name,
+			 (mgp->wc_enabled ? "Enabled" : "Disabled"));
 
 	return 0;
 
 abort_with_state:
 	pci_restore_state(pdev);
 
+abort_with_slices:
+	myri10ge_free_slices(mgp);
+
 abort_with_firmware:
 	myri10ge_dummy_rdma(mgp, 0);
 
-abort_with_rx_done:
-	bytes = mgp->max_intr_slots * sizeof(*mgp->ss.rx_done.entry);
-	dma_free_coherent(&pdev->dev, bytes,
-			  mgp->ss.rx_done.entry, mgp->ss.rx_done.bus);
-
 abort_with_ioremap:
 	iounmap(mgp->sram);
 
@@ -3303,10 +3874,6 @@ abort_with_wc:
 	if (mgp->mtrr >= 0)
 		mtrr_del(mgp->mtrr, mgp->iomem_base, mgp->board_span);
 #endif
-	dma_free_coherent(&pdev->dev, sizeof(*mgp->ss.fw_stats),
-			  mgp->ss.fw_stats, mgp->ss.fw_stats_bus);
-
-abort_with_cmd:
 	dma_free_coherent(&pdev->dev, sizeof(*mgp->cmd),
 			  mgp->cmd, mgp->cmd_bus);
 
@@ -3327,7 +3894,6 @@ static void myri10ge_remove(struct pci_dev *pdev)
 {
 	struct myri10ge_priv *mgp;
 	struct net_device *netdev;
-	size_t bytes;
 
 	mgp = pci_get_drvdata(pdev);
 	if (mgp == NULL)
@@ -3337,24 +3903,23 @@ static void myri10ge_remove(struct pci_dev *pdev)
 	netdev = mgp->dev;
 	unregister_netdev(netdev);
 
+#ifdef CONFIG_DCA
+	myri10ge_teardown_dca(mgp);
+#endif
 	myri10ge_dummy_rdma(mgp, 0);
 
 	/* avoid a memory leak */
 	pci_restore_state(pdev);
 
-	bytes = mgp->max_intr_slots * sizeof(*mgp->ss.rx_done.entry);
-	dma_free_coherent(&pdev->dev, bytes,
-			  mgp->ss.rx_done.entry, mgp->ss.rx_done.bus);
-
 	iounmap(mgp->sram);
 
 #ifdef CONFIG_MTRR
 	if (mgp->mtrr >= 0)
 		mtrr_del(mgp->mtrr, mgp->iomem_base, mgp->board_span);
 #endif
-	dma_free_coherent(&pdev->dev, sizeof(*mgp->ss.fw_stats),
-			  mgp->ss.fw_stats, mgp->ss.fw_stats_bus);
-
+	myri10ge_free_slices(mgp);
+	if (mgp->msix_vectors != NULL)
+		kfree(mgp->msix_vectors);
 	dma_free_coherent(&pdev->dev, sizeof(*mgp->cmd),
 			  mgp->cmd, mgp->cmd_bus);
 
@@ -3383,10 +3948,42 @@ static struct pci_driver myri10ge_driver = {
 #endif
 };
 
+#ifdef CONFIG_DCA
+static int
+myri10ge_notify_dca(struct notifier_block *nb, unsigned long event, void *p)
+{
+	int err = driver_for_each_device(&myri10ge_driver.driver,
+					 NULL, &event,
+					 myri10ge_notify_dca_device);
+
+	if (err)
+		return NOTIFY_BAD;
+	return NOTIFY_DONE;
+}
+
+static struct notifier_block myri10ge_dca_notifier = {
+	.notifier_call = myri10ge_notify_dca,
+	.next = NULL,
+	.priority = 0,
+};
+#endif				/* CONFIG_DCA */
+
 static __init int myri10ge_init_module(void)
 {
 	printk(KERN_INFO "%s: Version %s\n", myri10ge_driver.name,
 	       MYRI10GE_VERSION_STR);
+
+	if (myri10ge_rss_hash > MXGEFW_RSS_HASH_TYPE_SRC_PORT ||
+	    myri10ge_rss_hash < MXGEFW_RSS_HASH_TYPE_IPV4) {
+		printk(KERN_ERR
+		       "%s: Illegal rssh hash type %d, defaulting to source port\n",
+		       myri10ge_driver.name, myri10ge_rss_hash);
+		myri10ge_rss_hash = MXGEFW_RSS_HASH_TYPE_SRC_PORT;
+	}
+#ifdef CONFIG_DCA
+	dca_register_notify(&myri10ge_dca_notifier);
+#endif
+
 	return pci_register_driver(&myri10ge_driver);
 }
 
@@ -3394,6 +3991,9 @@ module_init(myri10ge_init_module);
 
 static __exit void myri10ge_cleanup_module(void)
 {
+#ifdef CONFIG_DCA
+	dca_unregister_notify(&myri10ge_dca_notifier);
+#endif
 	pci_unregister_driver(&myri10ge_driver);
 }
 
diff --git a/drivers/net/ns83820.c b/drivers/net/ns83820.c
index b42c05f..ff44961 100644
--- a/drivers/net/ns83820.c
+++ b/drivers/net/ns83820.c
@@ -585,16 +585,13 @@ static inline int rx_refill(struct net_device *ndev, gfp_t gfp)
 	for (i=0; i<NR_RX_DESC; i++) {
 		struct sk_buff *skb;
 		long res;
+
 		/* extra 16 bytes for alignment */
-		skb = __dev_alloc_skb(REAL_RX_BUF_SIZE+16, gfp);
+		skb = __netdev_alloc_skb(ndev, REAL_RX_BUF_SIZE+16, gfp);
 		if (unlikely(!skb))
 			break;
 
-		res = (long)skb->data & 0xf;
-		res = 0x10 - res;
-		res &= 0xf;
-		skb_reserve(skb, res);
-
+		skb_reserve(skb, skb->data - PTR_ALIGN(skb->data, 16));
 		if (gfp != GFP_ATOMIC)
 			spin_lock_irqsave(&dev->rx_info.lock, flags);
 		res = ns83820_add_rx_skb(dev, skb);
diff --git a/drivers/net/phy/Kconfig b/drivers/net/phy/Kconfig
index 284217c..d55932a 100644
--- a/drivers/net/phy/Kconfig
+++ b/drivers/net/phy/Kconfig
@@ -84,4 +84,10 @@ config MDIO_BITBANG
 
 	  If in doubt, say N.
 
+config MDIO_OF_GPIO
+	tristate "Support for GPIO lib-based bitbanged MDIO buses"
+	depends on MDIO_BITBANG && OF_GPIO
+	---help---
+	  Supports GPIO lib-based MDIO busses.
+
 endif # PHYLIB
diff --git a/drivers/net/phy/Makefile b/drivers/net/phy/Makefile
index 5997d6e..eee329f 100644
--- a/drivers/net/phy/Makefile
+++ b/drivers/net/phy/Makefile
@@ -15,3 +15,4 @@ obj-$(CONFIG_ICPLUS_PHY)	+= icplus.o
 obj-$(CONFIG_REALTEK_PHY)	+= realtek.o
 obj-$(CONFIG_FIXED_PHY)		+= fixed.o
 obj-$(CONFIG_MDIO_BITBANG)	+= mdio-bitbang.o
+obj-$(CONFIG_MDIO_OF_GPIO)	+= mdio-ofgpio.o
diff --git a/drivers/net/phy/mdio-ofgpio.c b/drivers/net/phy/mdio-ofgpio.c
new file mode 100644
index 0000000..7edfc0c
--- /dev/null
+++ b/drivers/net/phy/mdio-ofgpio.c
@@ -0,0 +1,205 @@
+/*
+ * OpenFirmware GPIO based MDIO bitbang driver.
+ *
+ * Copyright (c) 2008 CSE Semaphore Belgium.
+ *  by Laurent Pinchart <laurentp@cse-semaphore.com>
+ *
+ * Based on earlier work by
+ *
+ * Copyright (c) 2003 Intracom S.A.
+ *  by Pantelis Antoniou <panto@intracom.gr>
+ *
+ * 2005 (c) MontaVista Software, Inc.
+ * Vitaly Bordug <vbordug@ru.mvista.com>
+ *
+ * This file is licensed under the terms of the GNU General Public License
+ * version 2. This program is licensed "as is" without any warranty of any
+ * kind, whether express or implied.
+ */
+
+#include <linux/module.h>
+#include <linux/slab.h>
+#include <linux/init.h>
+#include <linux/interrupt.h>
+#include <linux/mdio-bitbang.h>
+#include <linux/of_gpio.h>
+#include <linux/of_platform.h>
+
+struct mdio_gpio_info {
+	struct mdiobb_ctrl ctrl;
+	int mdc, mdio;
+};
+
+static void mdio_dir(struct mdiobb_ctrl *ctrl, int dir)
+{
+	struct mdio_gpio_info *bitbang =
+		container_of(ctrl, struct mdio_gpio_info, ctrl);
+
+	if (dir)
+		gpio_direction_output(bitbang->mdio, 1);
+	else
+		gpio_direction_input(bitbang->mdio);
+}
+
+static int mdio_read(struct mdiobb_ctrl *ctrl)
+{
+	struct mdio_gpio_info *bitbang =
+		container_of(ctrl, struct mdio_gpio_info, ctrl);
+
+	return gpio_get_value(bitbang->mdio);
+}
+
+static void mdio(struct mdiobb_ctrl *ctrl, int what)
+{
+	struct mdio_gpio_info *bitbang =
+		container_of(ctrl, struct mdio_gpio_info, ctrl);
+
+	gpio_set_value(bitbang->mdio, what);
+}
+
+static void mdc(struct mdiobb_ctrl *ctrl, int what)
+{
+	struct mdio_gpio_info *bitbang =
+		container_of(ctrl, struct mdio_gpio_info, ctrl);
+
+	gpio_set_value(bitbang->mdc, what);
+}
+
+static struct mdiobb_ops mdio_gpio_ops = {
+	.owner = THIS_MODULE,
+	.set_mdc = mdc,
+	.set_mdio_dir = mdio_dir,
+	.set_mdio_data = mdio,
+	.get_mdio_data = mdio_read,
+};
+
+static int __devinit mdio_ofgpio_bitbang_init(struct mii_bus *bus,
+                                         struct device_node *np)
+{
+	struct mdio_gpio_info *bitbang = bus->priv;
+
+	bitbang->mdc = of_get_gpio(np, 0);
+	bitbang->mdio = of_get_gpio(np, 1);
+
+	if (bitbang->mdc < 0 || bitbang->mdio < 0)
+		return -ENODEV;
+
+	snprintf(bus->id, MII_BUS_ID_SIZE, "%x", bitbang->mdc);
+	return 0;
+}
+
+static void __devinit add_phy(struct mii_bus *bus, struct device_node *np)
+{
+	const u32 *data;
+	int len, id, irq;
+
+	data = of_get_property(np, "reg", &len);
+	if (!data || len != 4)
+		return;
+
+	id = *data;
+	bus->phy_mask &= ~(1 << id);
+
+	irq = of_irq_to_resource(np, 0, NULL);
+	if (irq != NO_IRQ)
+		bus->irq[id] = irq;
+}
+
+static int __devinit mdio_ofgpio_probe(struct of_device *ofdev,
+                                        const struct of_device_id *match)
+{
+	struct device_node *np = NULL;
+	struct mii_b