lpc32xx: Added ethernet driver
authorstigge@antcom.de <stigge@antcom.de>
Thu, 8 Mar 2012 11:49:17 +0000 (11:49 +0000)
committerDavid S. Miller <davem@davemloft.net>
Tue, 13 Mar 2012 00:08:34 +0000 (17:08 -0700)
This patch adds an ethernet driver for the LPC32xx ARM SoC.

Signed-off-by: Roland Stigge <stigge@antcom.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
drivers/net/ethernet/Kconfig
drivers/net/ethernet/Makefile
drivers/net/ethernet/nxp/Kconfig [new file with mode: 0644]
drivers/net/ethernet/nxp/Makefile [new file with mode: 0644]
drivers/net/ethernet/nxp/lpc_eth.c [new file with mode: 0644]

index 3474a61d470501adb915b6dc2e0cef487672a971..c63a64cb608546b0a13fcd48842998786261dca9 100644 (file)
@@ -126,6 +126,7 @@ config NET_NETX
 
 source "drivers/net/ethernet/nuvoton/Kconfig"
 source "drivers/net/ethernet/nvidia/Kconfig"
+source "drivers/net/ethernet/nxp/Kconfig"
 source "drivers/net/ethernet/octeon/Kconfig"
 source "drivers/net/ethernet/oki-semi/Kconfig"
 
index 08d5f03888777b1a977d0c07c7bc8248971a4ae3..9676a5109d94a42d7ed8ee35a9a6ddbe776ecac7 100644 (file)
@@ -47,6 +47,7 @@ obj-$(CONFIG_NET_VENDOR_NATSEMI) += natsemi/
 obj-$(CONFIG_NET_NETX) += netx-eth.o
 obj-$(CONFIG_NET_VENDOR_NUVOTON) += nuvoton/
 obj-$(CONFIG_NET_VENDOR_NVIDIA) += nvidia/
+obj-$(CONFIG_LPC_ENET) += nxp/
 obj-$(CONFIG_OCTEON_MGMT_ETHERNET) += octeon/
 obj-$(CONFIG_NET_VENDOR_OKI) += oki-semi/
 obj-$(CONFIG_ETHOC) += ethoc.o
diff --git a/drivers/net/ethernet/nxp/Kconfig b/drivers/net/ethernet/nxp/Kconfig
new file mode 100644 (file)
index 0000000..0d9baf9
--- /dev/null
@@ -0,0 +1,8 @@
+config LPC_ENET
+        tristate "NXP ethernet MAC on LPC devices"
+        depends on ARCH_LPC32XX
+        select PHYLIB
+        help
+         Say Y or M here if you want to use the NXP ethernet MAC included on
+         some NXP LPC devices. You can safely enable this option for LPC32xx
+         SoC. Also available as a module.
diff --git a/drivers/net/ethernet/nxp/Makefile b/drivers/net/ethernet/nxp/Makefile
new file mode 100644 (file)
index 0000000..a128114
--- /dev/null
@@ -0,0 +1 @@
+obj-$(CONFIG_LPC_ENET) += lpc_eth.o
diff --git a/drivers/net/ethernet/nxp/lpc_eth.c b/drivers/net/ethernet/nxp/lpc_eth.c
new file mode 100644 (file)
index 0000000..6944424
--- /dev/null
@@ -0,0 +1,1604 @@
+/*
+ * drivers/net/ethernet/nxp/lpc_eth.c
+ *
+ * Author: Kevin Wells <kevin.wells@nxp.com>
+ *
+ * Copyright (C) 2010 NXP Semiconductors
+ * Copyright (C) 2012 Roland Stigge <stigge@antcom.de>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
+
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/kernel.h>
+#include <linux/sched.h>
+#include <linux/slab.h>
+#include <linux/delay.h>
+#include <linux/interrupt.h>
+#include <linux/errno.h>
+#include <linux/ioport.h>
+#include <linux/crc32.h>
+#include <linux/platform_device.h>
+#include <linux/spinlock.h>
+#include <linux/ethtool.h>
+#include <linux/mii.h>
+#include <linux/clk.h>
+#include <linux/workqueue.h>
+#include <linux/netdevice.h>
+#include <linux/etherdevice.h>
+#include <linux/skbuff.h>
+#include <linux/phy.h>
+#include <linux/dma-mapping.h>
+#include <linux/of_net.h>
+#include <linux/types.h>
+
+#include <linux/delay.h>
+#include <linux/io.h>
+#include <mach/board.h>
+#include <mach/platform.h>
+#include <mach/hardware.h>
+
+#define MODNAME "lpc-eth"
+#define DRV_VERSION "1.00"
+#define PHYDEF_ADDR 0x00
+
+#define ENET_MAXF_SIZE 1536
+#define ENET_RX_DESC 48
+#define ENET_TX_DESC 16
+
+#define NAPI_WEIGHT 16
+
+/*
+ * Ethernet MAC controller Register offsets
+ */
+#define LPC_ENET_MAC1(x)                       (x + 0x000)
+#define LPC_ENET_MAC2(x)                       (x + 0x004)
+#define LPC_ENET_IPGT(x)                       (x + 0x008)
+#define LPC_ENET_IPGR(x)                       (x + 0x00C)
+#define LPC_ENET_CLRT(x)                       (x + 0x010)
+#define LPC_ENET_MAXF(x)                       (x + 0x014)
+#define LPC_ENET_SUPP(x)                       (x + 0x018)
+#define LPC_ENET_TEST(x)                       (x + 0x01C)
+#define LPC_ENET_MCFG(x)                       (x + 0x020)
+#define LPC_ENET_MCMD(x)                       (x + 0x024)
+#define LPC_ENET_MADR(x)                       (x + 0x028)
+#define LPC_ENET_MWTD(x)                       (x + 0x02C)
+#define LPC_ENET_MRDD(x)                       (x + 0x030)
+#define LPC_ENET_MIND(x)                       (x + 0x034)
+#define LPC_ENET_SA0(x)                                (x + 0x040)
+#define LPC_ENET_SA1(x)                                (x + 0x044)
+#define LPC_ENET_SA2(x)                                (x + 0x048)
+#define LPC_ENET_COMMAND(x)                    (x + 0x100)
+#define LPC_ENET_STATUS(x)                     (x + 0x104)
+#define LPC_ENET_RXDESCRIPTOR(x)               (x + 0x108)
+#define LPC_ENET_RXSTATUS(x)                   (x + 0x10C)
+#define LPC_ENET_RXDESCRIPTORNUMBER(x)         (x + 0x110)
+#define LPC_ENET_RXPRODUCEINDEX(x)             (x + 0x114)
+#define LPC_ENET_RXCONSUMEINDEX(x)             (x + 0x118)
+#define LPC_ENET_TXDESCRIPTOR(x)               (x + 0x11C)
+#define LPC_ENET_TXSTATUS(x)                   (x + 0x120)
+#define LPC_ENET_TXDESCRIPTORNUMBER(x)         (x + 0x124)
+#define LPC_ENET_TXPRODUCEINDEX(x)             (x + 0x128)
+#define LPC_ENET_TXCONSUMEINDEX(x)             (x + 0x12C)
+#define LPC_ENET_TSV0(x)                       (x + 0x158)
+#define LPC_ENET_TSV1(x)                       (x + 0x15C)
+#define LPC_ENET_RSV(x)                                (x + 0x160)
+#define LPC_ENET_FLOWCONTROLCOUNTER(x)         (x + 0x170)
+#define LPC_ENET_FLOWCONTROLSTATUS(x)          (x + 0x174)
+#define LPC_ENET_RXFILTER_CTRL(x)              (x + 0x200)
+#define LPC_ENET_RXFILTERWOLSTATUS(x)          (x + 0x204)
+#define LPC_ENET_RXFILTERWOLCLEAR(x)           (x + 0x208)
+#define LPC_ENET_HASHFILTERL(x)                        (x + 0x210)
+#define LPC_ENET_HASHFILTERH(x)                        (x + 0x214)
+#define LPC_ENET_INTSTATUS(x)                  (x + 0xFE0)
+#define LPC_ENET_INTENABLE(x)                  (x + 0xFE4)
+#define LPC_ENET_INTCLEAR(x)                   (x + 0xFE8)
+#define LPC_ENET_INTSET(x)                     (x + 0xFEC)
+#define LPC_ENET_POWERDOWN(x)                  (x + 0xFF4)
+
+/*
+ * mac1 register definitions
+ */
+#define LPC_MAC1_RECV_ENABLE                   (1 << 0)
+#define LPC_MAC1_PASS_ALL_RX_FRAMES            (1 << 1)
+#define LPC_MAC1_RX_FLOW_CONTROL               (1 << 2)
+#define LPC_MAC1_TX_FLOW_CONTROL               (1 << 3)
+#define LPC_MAC1_LOOPBACK                      (1 << 4)
+#define LPC_MAC1_RESET_TX                      (1 << 8)
+#define LPC_MAC1_RESET_MCS_TX                  (1 << 9)
+#define LPC_MAC1_RESET_RX                      (1 << 10)
+#define LPC_MAC1_RESET_MCS_RX                  (1 << 11)
+#define LPC_MAC1_SIMULATION_RESET              (1 << 14)
+#define LPC_MAC1_SOFT_RESET                    (1 << 15)
+
+/*
+ * mac2 register definitions
+ */
+#define LPC_MAC2_FULL_DUPLEX                   (1 << 0)
+#define LPC_MAC2_FRAME_LENGTH_CHECKING         (1 << 1)
+#define LPC_MAC2_HUGH_LENGTH_CHECKING          (1 << 2)
+#define LPC_MAC2_DELAYED_CRC                   (1 << 3)
+#define LPC_MAC2_CRC_ENABLE                    (1 << 4)
+#define LPC_MAC2_PAD_CRC_ENABLE                        (1 << 5)
+#define LPC_MAC2_VLAN_PAD_ENABLE               (1 << 6)
+#define LPC_MAC2_AUTO_DETECT_PAD_ENABLE                (1 << 7)
+#define LPC_MAC2_PURE_PREAMBLE_ENFORCEMENT     (1 << 8)
+#define LPC_MAC2_LONG_PREAMBLE_ENFORCEMENT     (1 << 9)
+#define LPC_MAC2_NO_BACKOFF                    (1 << 12)
+#define LPC_MAC2_BACK_PRESSURE                 (1 << 13)
+#define LPC_MAC2_EXCESS_DEFER                  (1 << 14)
+
+/*
+ * ipgt register definitions
+ */
+#define LPC_IPGT_LOAD(n)                       ((n) & 0x7F)
+
+/*
+ * ipgr register definitions
+ */
+#define LPC_IPGR_LOAD_PART2(n)                 ((n) & 0x7F)
+#define LPC_IPGR_LOAD_PART1(n)                 (((n) & 0x7F) << 8)
+
+/*
+ * clrt register definitions
+ */
+#define LPC_CLRT_LOAD_RETRY_MAX(n)             ((n) & 0xF)
+#define LPC_CLRT_LOAD_COLLISION_WINDOW(n)      (((n) & 0x3F) << 8)
+
+/*
+ * maxf register definitions
+ */
+#define LPC_MAXF_LOAD_MAX_FRAME_LEN(n)         ((n) & 0xFFFF)
+
+/*
+ * supp register definitions
+ */
+#define LPC_SUPP_SPEED                         (1 << 8)
+#define LPC_SUPP_RESET_RMII                    (1 << 11)
+
+/*
+ * test register definitions
+ */
+#define LPC_TEST_SHORTCUT_PAUSE_QUANTA         (1 << 0)
+#define LPC_TEST_PAUSE                         (1 << 1)
+#define LPC_TEST_BACKPRESSURE                  (1 << 2)
+
+/*
+ * mcfg register definitions
+ */
+#define LPC_MCFG_SCAN_INCREMENT                        (1 << 0)
+#define LPC_MCFG_SUPPRESS_PREAMBLE             (1 << 1)
+#define LPC_MCFG_CLOCK_SELECT(n)               (((n) & 0x7) << 2)
+#define LPC_MCFG_CLOCK_HOST_DIV_4              0
+#define LPC_MCFG_CLOCK_HOST_DIV_6              2
+#define LPC_MCFG_CLOCK_HOST_DIV_8              3
+#define LPC_MCFG_CLOCK_HOST_DIV_10             4
+#define LPC_MCFG_CLOCK_HOST_DIV_14             5
+#define LPC_MCFG_CLOCK_HOST_DIV_20             6
+#define LPC_MCFG_CLOCK_HOST_DIV_28             7
+#define LPC_MCFG_RESET_MII_MGMT                        (1 << 15)
+
+/*
+ * mcmd register definitions
+ */
+#define LPC_MCMD_READ                          (1 << 0)
+#define LPC_MCMD_SCAN                          (1 << 1)
+
+/*
+ * madr register definitions
+ */
+#define LPC_MADR_REGISTER_ADDRESS(n)           ((n) & 0x1F)
+#define LPC_MADR_PHY_0ADDRESS(n)               (((n) & 0x1F) << 8)
+
+/*
+ * mwtd register definitions
+ */
+#define LPC_MWDT_WRITE(n)                      ((n) & 0xFFFF)
+
+/*
+ * mrdd register definitions
+ */
+#define LPC_MRDD_READ_MASK                     0xFFFF
+
+/*
+ * mind register definitions
+ */
+#define LPC_MIND_BUSY                          (1 << 0)
+#define LPC_MIND_SCANNING                      (1 << 1)
+#define LPC_MIND_NOT_VALID                     (1 << 2)
+#define LPC_MIND_MII_LINK_FAIL                 (1 << 3)
+
+/*
+ * command register definitions
+ */
+#define LPC_COMMAND_RXENABLE                   (1 << 0)
+#define LPC_COMMAND_TXENABLE                   (1 << 1)
+#define LPC_COMMAND_REG_RESET                  (1 << 3)
+#define LPC_COMMAND_TXRESET                    (1 << 4)
+#define LPC_COMMAND_RXRESET                    (1 << 5)
+#define LPC_COMMAND_PASSRUNTFRAME              (1 << 6)
+#define LPC_COMMAND_PASSRXFILTER               (1 << 7)
+#define LPC_COMMAND_TXFLOWCONTROL              (1 << 8)
+#define LPC_COMMAND_RMII                       (1 << 9)
+#define LPC_COMMAND_FULLDUPLEX                 (1 << 10)
+
+/*
+ * status register definitions
+ */
+#define LPC_STATUS_RXACTIVE                    (1 << 0)
+#define LPC_STATUS_TXACTIVE                    (1 << 1)
+
+/*
+ * tsv0 register definitions
+ */
+#define LPC_TSV0_CRC_ERROR                     (1 << 0)
+#define LPC_TSV0_LENGTH_CHECK_ERROR            (1 << 1)
+#define LPC_TSV0_LENGTH_OUT_OF_RANGE           (1 << 2)
+#define LPC_TSV0_DONE                          (1 << 3)
+#define LPC_TSV0_MULTICAST                     (1 << 4)
+#define LPC_TSV0_BROADCAST                     (1 << 5)
+#define LPC_TSV0_PACKET_DEFER                  (1 << 6)
+#define LPC_TSV0_ESCESSIVE_DEFER               (1 << 7)
+#define LPC_TSV0_ESCESSIVE_COLLISION           (1 << 8)
+#define LPC_TSV0_LATE_COLLISION                        (1 << 9)
+#define LPC_TSV0_GIANT                         (1 << 10)
+#define LPC_TSV0_UNDERRUN                      (1 << 11)
+#define LPC_TSV0_TOTAL_BYTES(n)                        (((n) >> 12) & 0xFFFF)
+#define LPC_TSV0_CONTROL_FRAME                 (1 << 28)
+#define LPC_TSV0_PAUSE                         (1 << 29)
+#define LPC_TSV0_BACKPRESSURE                  (1 << 30)
+#define LPC_TSV0_VLAN                          (1 << 31)
+
+/*
+ * tsv1 register definitions
+ */
+#define LPC_TSV1_TRANSMIT_BYTE_COUNT(n)                ((n) & 0xFFFF)
+#define LPC_TSV1_COLLISION_COUNT(n)            (((n) >> 16) & 0xF)
+
+/*
+ * rsv register definitions
+ */
+#define LPC_RSV_RECEIVED_BYTE_COUNT(n)         ((n) & 0xFFFF)
+#define LPC_RSV_RXDV_EVENT_IGNORED             (1 << 16)
+#define LPC_RSV_RXDV_EVENT_PREVIOUSLY_SEEN     (1 << 17)
+#define LPC_RSV_CARRIER_EVNT_PREVIOUS_SEEN     (1 << 18)
+#define LPC_RSV_RECEIVE_CODE_VIOLATION         (1 << 19)
+#define LPC_RSV_CRC_ERROR                      (1 << 20)
+#define LPC_RSV_LENGTH_CHECK_ERROR             (1 << 21)
+#define LPC_RSV_LENGTH_OUT_OF_RANGE            (1 << 22)
+#define LPC_RSV_RECEIVE_OK                     (1 << 23)
+#define LPC_RSV_MULTICAST                      (1 << 24)
+#define LPC_RSV_BROADCAST                      (1 << 25)
+#define LPC_RSV_DRIBBLE_NIBBLE                 (1 << 26)
+#define LPC_RSV_CONTROL_FRAME                  (1 << 27)
+#define LPC_RSV_PAUSE                          (1 << 28)
+#define LPC_RSV_UNSUPPORTED_OPCODE             (1 << 29)
+#define LPC_RSV_VLAN                           (1 << 30)
+
+/*
+ * flowcontrolcounter register definitions
+ */
+#define LPC_FCCR_MIRRORCOUNTER(n)              ((n) & 0xFFFF)
+#define LPC_FCCR_PAUSETIMER(n)                 (((n) >> 16) & 0xFFFF)
+
+/*
+ * flowcontrolstatus register definitions
+ */
+#define LPC_FCCR_MIRRORCOUNTERCURRENT(n)       ((n) & 0xFFFF)
+
+/*
+ * rxfliterctrl, rxfilterwolstatus, and rxfilterwolclear shared
+ * register definitions
+ */
+#define LPC_RXFLTRW_ACCEPTUNICAST              (1 << 0)
+#define LPC_RXFLTRW_ACCEPTUBROADCAST           (1 << 1)
+#define LPC_RXFLTRW_ACCEPTUMULTICAST           (1 << 2)
+#define LPC_RXFLTRW_ACCEPTUNICASTHASH          (1 << 3)
+#define LPC_RXFLTRW_ACCEPTUMULTICASTHASH       (1 << 4)
+#define LPC_RXFLTRW_ACCEPTPERFECT              (1 << 5)
+
+/*
+ * rxfliterctrl register definitions
+ */
+#define LPC_RXFLTRWSTS_MAGICPACKETENWOL                (1 << 12)
+#define LPC_RXFLTRWSTS_RXFILTERENWOL           (1 << 13)
+
+/*
+ * rxfilterwolstatus/rxfilterwolclear register definitions
+ */
+#define LPC_RXFLTRWSTS_RXFILTERWOL             (1 << 7)
+#define LPC_RXFLTRWSTS_MAGICPACKETWOL          (1 << 8)
+
+/*
+ * intstatus, intenable, intclear, and Intset shared register
+ * definitions
+ */
+#define LPC_MACINT_RXOVERRUNINTEN              (1 << 0)
+#define LPC_MACINT_RXERRORONINT                        (1 << 1)
+#define LPC_MACINT_RXFINISHEDINTEN             (1 << 2)
+#define LPC_MACINT_RXDONEINTEN                 (1 << 3)
+#define LPC_MACINT_TXUNDERRUNINTEN             (1 << 4)
+#define LPC_MACINT_TXERRORINTEN                        (1 << 5)
+#define LPC_MACINT_TXFINISHEDINTEN             (1 << 6)
+#define LPC_MACINT_TXDONEINTEN                 (1 << 7)
+#define LPC_MACINT_SOFTINTEN                   (1 << 12)
+#define LPC_MACINT_WAKEUPINTEN                 (1 << 13)
+
+/*
+ * powerdown register definitions
+ */
+#define LPC_POWERDOWN_MACAHB                   (1 << 31)
+
+/* Upon the upcoming introduction of device tree usage in LPC32xx,
+ * lpc_phy_interface_mode() and use_iram_for_net() will be extended with a
+ * device parameter for access to device tree information at runtime, instead
+ * of defining the values at compile time
+ */
+static inline phy_interface_t lpc_phy_interface_mode(void)
+{
+#ifdef CONFIG_ARCH_LPC32XX_MII_SUPPORT
+       return PHY_INTERFACE_MODE_MII;
+#else
+       return PHY_INTERFACE_MODE_RMII;
+#endif
+}
+
+static inline int use_iram_for_net(void)
+{
+#ifdef CONFIG_ARCH_LPC32XX_IRAM_FOR_NET
+       return 1;
+#else
+       return 0;
+#endif
+}
+
+/* Receive Status information word */
+#define RXSTATUS_SIZE                  0x000007FF
+#define RXSTATUS_CONTROL               (1 << 18)
+#define RXSTATUS_VLAN                  (1 << 19)
+#define RXSTATUS_FILTER                        (1 << 20)
+#define RXSTATUS_MULTICAST             (1 << 21)
+#define RXSTATUS_BROADCAST             (1 << 22)
+#define RXSTATUS_CRC                   (1 << 23)
+#define RXSTATUS_SYMBOL                        (1 << 24)
+#define RXSTATUS_LENGTH                        (1 << 25)
+#define RXSTATUS_RANGE                 (1 << 26)
+#define RXSTATUS_ALIGN                 (1 << 27)
+#define RXSTATUS_OVERRUN               (1 << 28)
+#define RXSTATUS_NODESC                        (1 << 29)
+#define RXSTATUS_LAST                  (1 << 30)
+#define RXSTATUS_ERROR                 (1 << 31)
+
+#define RXSTATUS_STATUS_ERROR \
+       (RXSTATUS_NODESC | RXSTATUS_OVERRUN | RXSTATUS_ALIGN | \
+        RXSTATUS_RANGE | RXSTATUS_LENGTH | RXSTATUS_SYMBOL | RXSTATUS_CRC)
+
+/* Receive Descriptor control word */
+#define RXDESC_CONTROL_SIZE            0x000007FF
+#define RXDESC_CONTROL_INT             (1 << 31)
+
+/* Transmit Status information word */
+#define TXSTATUS_COLLISIONS_GET(x)     (((x) >> 21) & 0xF)
+#define TXSTATUS_DEFER                 (1 << 25)
+#define TXSTATUS_EXCESSDEFER           (1 << 26)
+#define TXSTATUS_EXCESSCOLL            (1 << 27)
+#define TXSTATUS_LATECOLL              (1 << 28)
+#define TXSTATUS_UNDERRUN              (1 << 29)
+#define TXSTATUS_NODESC                        (1 << 30)
+#define TXSTATUS_ERROR                 (1 << 31)
+
+/* Transmit Descriptor control word */
+#define TXDESC_CONTROL_SIZE            0x000007FF
+#define TXDESC_CONTROL_OVERRIDE                (1 << 26)
+#define TXDESC_CONTROL_HUGE            (1 << 27)
+#define TXDESC_CONTROL_PAD             (1 << 28)
+#define TXDESC_CONTROL_CRC             (1 << 29)
+#define TXDESC_CONTROL_LAST            (1 << 30)
+#define TXDESC_CONTROL_INT             (1 << 31)
+
+static int lpc_eth_hard_start_xmit(struct sk_buff *skb,
+                                  struct net_device *ndev);
+
+/*
+ * Structure of a TX/RX descriptors and RX status
+ */
+struct txrx_desc_t {
+       __le32 packet;
+       __le32 control;
+};
+struct rx_status_t {
+       __le32 statusinfo;
+       __le32 statushashcrc;
+};
+
+/*
+ * Device driver data structure
+ */
+struct netdata_local {
+       struct platform_device  *pdev;
+       struct net_device       *ndev;
+       spinlock_t              lock;
+       void __iomem            *net_base;
+       u32                     msg_enable;
+       struct sk_buff          *skb[ENET_TX_DESC];
+       unsigned int            last_tx_idx;
+       unsigned int            num_used_tx_buffs;
+       struct mii_bus          *mii_bus;
+       struct phy_device       *phy_dev;
+       struct clk              *clk;
+       dma_addr_t              dma_buff_base_p;
+       void                    *dma_buff_base_v;
+       size_t                  dma_buff_size;
+       struct txrx_desc_t      *tx_desc_v;
+       u32                     *tx_stat_v;
+       void                    *tx_buff_v;
+       struct txrx_desc_t      *rx_desc_v;
+       struct rx_status_t      *rx_stat_v;
+       void                    *rx_buff_v;
+       int                     link;
+       int                     speed;
+       int                     duplex;
+       struct napi_struct      napi;
+};
+
+/*
+ * MAC support functions
+ */
+static void __lpc_set_mac(struct netdata_local *pldat, u8 *mac)
+{
+       u32 tmp;
+
+       /* Set station address */
+       tmp = mac[0] | ((u32)mac[1] << 8);
+       writel(tmp, LPC_ENET_SA2(pldat->net_base));
+       tmp = mac[2] | ((u32)mac[3] << 8);
+       writel(tmp, LPC_ENET_SA1(pldat->net_base));
+       tmp = mac[4] | ((u32)mac[5] << 8);
+       writel(tmp, LPC_ENET_SA0(pldat->net_base));
+
+       netdev_dbg(pldat->ndev, "Ethernet MAC address %pM\n", mac);
+}
+
+static void __lpc_get_mac(struct netdata_local *pldat, u8 *mac)
+{
+       u32 tmp;
+
+       /* Get station address */
+       tmp = readl(LPC_ENET_SA2(pldat->net_base));
+       mac[0] = tmp & 0xFF;
+       mac[1] = tmp >> 8;
+       tmp = readl(LPC_ENET_SA1(pldat->net_base));
+       mac[2] = tmp & 0xFF;
+       mac[3] = tmp >> 8;
+       tmp = readl(LPC_ENET_SA0(pldat->net_base));
+       mac[4] = tmp & 0xFF;
+       mac[5] = tmp >> 8;
+}
+
+static void __lpc_eth_clock_enable(struct netdata_local *pldat,
+                                  bool enable)
+{
+       if (enable)
+               clk_enable(pldat->clk);
+       else
+               clk_disable(pldat->clk);
+}
+
+static void __lpc_params_setup(struct netdata_local *pldat)
+{
+       u32 tmp;
+
+       if (pldat->duplex == DUPLEX_FULL) {
+               tmp = readl(LPC_ENET_MAC2(pldat->net_base));
+               tmp |= LPC_MAC2_FULL_DUPLEX;
+               writel(tmp, LPC_ENET_MAC2(pldat->net_base));
+               tmp = readl(LPC_ENET_COMMAND(pldat->net_base));
+               tmp |= LPC_COMMAND_FULLDUPLEX;
+               writel(tmp, LPC_ENET_COMMAND(pldat->net_base));
+               writel(LPC_IPGT_LOAD(0x15), LPC_ENET_IPGT(pldat->net_base));
+       } else {
+               tmp = readl(LPC_ENET_MAC2(pldat->net_base));
+               tmp &= ~LPC_MAC2_FULL_DUPLEX;
+               writel(tmp, LPC_ENET_MAC2(pldat->net_base));
+               tmp = readl(LPC_ENET_COMMAND(pldat->net_base));
+               tmp &= ~LPC_COMMAND_FULLDUPLEX;
+               writel(tmp, LPC_ENET_COMMAND(pldat->net_base));
+               writel(LPC_IPGT_LOAD(0x12), LPC_ENET_IPGT(pldat->net_base));
+       }
+
+       if (pldat->speed == SPEED_100)
+               writel(LPC_SUPP_SPEED, LPC_ENET_SUPP(pldat->net_base));
+       else
+               writel(0, LPC_ENET_SUPP(pldat->net_base));
+}
+
+static void __lpc_eth_reset(struct netdata_local *pldat)
+{
+       /* Reset all MAC logic */
+       writel((LPC_MAC1_RESET_TX | LPC_MAC1_RESET_MCS_TX | LPC_MAC1_RESET_RX |
+               LPC_MAC1_RESET_MCS_RX | LPC_MAC1_SIMULATION_RESET |
+               LPC_MAC1_SOFT_RESET), LPC_ENET_MAC1(pldat->net_base));
+       writel((LPC_COMMAND_REG_RESET | LPC_COMMAND_TXRESET |
+               LPC_COMMAND_RXRESET), LPC_ENET_COMMAND(pldat->net_base));
+}
+
+static int __lpc_mii_mngt_reset(struct netdata_local *pldat)
+{
+       /* Reset MII management hardware */
+       writel(LPC_MCFG_RESET_MII_MGMT, LPC_ENET_MCFG(pldat->net_base));
+
+       /* Setup MII clock to slowest rate with a /28 divider */
+       writel(LPC_MCFG_CLOCK_SELECT(LPC_MCFG_CLOCK_HOST_DIV_28),
+              LPC_ENET_MCFG(pldat->net_base));
+
+       return 0;
+}
+
+static inline phys_addr_t __va_to_pa(void *addr, struct netdata_local *pldat)
+{
+       phys_addr_t phaddr;
+
+       phaddr = addr - pldat->dma_buff_base_v;
+       phaddr += pldat->dma_buff_base_p;
+
+       return phaddr;
+}
+
+static void lpc_eth_enable_int(void __iomem *regbase)
+{
+       writel((LPC_MACINT_RXDONEINTEN | LPC_MACINT_TXDONEINTEN),
+              LPC_ENET_INTENABLE(regbase));
+}
+
+static void lpc_eth_disable_int(void __iomem *regbase)
+{
+       writel(0, LPC_ENET_INTENABLE(regbase));
+}
+
+/* Setup TX/RX descriptors */
+static void __lpc_txrx_desc_setup(struct netdata_local *pldat)
+{
+       u32 *ptxstat;
+       void *tbuff;
+       int i;
+       struct txrx_desc_t *ptxrxdesc;
+       struct rx_status_t *prxstat;
+
+       tbuff = PTR_ALIGN(pldat->dma_buff_base_v, 16);
+
+       /* Setup TX descriptors, status, and buffers */
+       pldat->tx_desc_v = tbuff;
+       tbuff += sizeof(struct txrx_desc_t) * ENET_TX_DESC;
+
+       pldat->tx_stat_v = tbuff;
+       tbuff += sizeof(u32) * ENET_TX_DESC;
+
+       tbuff = PTR_ALIGN(tbuff, 16);
+       pldat->tx_buff_v = tbuff;
+       tbuff += ENET_MAXF_SIZE * ENET_TX_DESC;
+
+       /* Setup RX descriptors, status, and buffers */
+       pldat->rx_desc_v = tbuff;
+       tbuff += sizeof(struct txrx_desc_t) * ENET_RX_DESC;
+
+       tbuff = PTR_ALIGN(tbuff, 16);
+       pldat->rx_stat_v = tbuff;
+       tbuff += sizeof(struct rx_status_t) * ENET_RX_DESC;
+
+       tbuff = PTR_ALIGN(tbuff, 16);
+       pldat->rx_buff_v = tbuff;
+       tbuff += ENET_MAXF_SIZE * ENET_RX_DESC;
+
+       /* Map the TX descriptors to the TX buffers in hardware */
+       for (i = 0; i < ENET_TX_DESC; i++) {
+               ptxstat = &pldat->tx_stat_v[i];
+               ptxrxdesc = &pldat->tx_desc_v[i];
+
+               ptxrxdesc->packet = __va_to_pa(
+                               pldat->tx_buff_v + i * ENET_MAXF_SIZE, pldat);
+               ptxrxdesc->control = 0;
+               *ptxstat = 0;
+       }
+
+       /* Map the RX descriptors to the RX buffers in hardware */
+       for (i = 0; i < ENET_RX_DESC; i++) {
+               prxstat = &pldat->rx_stat_v[i];
+               ptxrxdesc = &pldat->rx_desc_v[i];
+
+               ptxrxdesc->packet = __va_to_pa(
+                               pldat->rx_buff_v + i * ENET_MAXF_SIZE, pldat);
+               ptxrxdesc->control = RXDESC_CONTROL_INT | (ENET_MAXF_SIZE - 1);
+               prxstat->statusinfo = 0;
+               prxstat->statushashcrc = 0;
+       }
+
+       /* Setup base addresses in hardware to point to buffers and
+        * descriptors
+        */
+       writel((ENET_TX_DESC - 1),
+              LPC_ENET_TXDESCRIPTORNUMBER(pldat->net_base));
+       writel(__va_to_pa(pldat->tx_desc_v, pldat),
+              LPC_ENET_TXDESCRIPTOR(pldat->net_base));
+       writel(__va_to_pa(pldat->tx_stat_v, pldat),
+              LPC_ENET_TXSTATUS(pldat->net_base));
+       writel((ENET_RX_DESC - 1),
+              LPC_ENET_RXDESCRIPTORNUMBER(pldat->net_base));
+       writel(__va_to_pa(pldat->rx_desc_v, pldat),
+              LPC_ENET_RXDESCRIPTOR(pldat->net_base));
+       writel(__va_to_pa(pldat->rx_stat_v, pldat),
+              LPC_ENET_RXSTATUS(pldat->net_base));
+}
+
+static void __lpc_eth_init(struct netdata_local *pldat)
+{
+       u32 tmp;
+
+       /* Disable controller and reset */
+       tmp = readl(LPC_ENET_COMMAND(pldat->net_base));
+       tmp &= ~LPC_COMMAND_RXENABLE | LPC_COMMAND_TXENABLE;
+       writel(tmp, LPC_ENET_COMMAND(pldat->net_base));
+       tmp = readl(LPC_ENET_MAC1(pldat->net_base));
+       tmp &= ~LPC_MAC1_RECV_ENABLE;
+       writel(tmp, LPC_ENET_MAC1(pldat->net_base));
+
+       /* Initial MAC setup */
+       writel(LPC_MAC1_PASS_ALL_RX_FRAMES, LPC_ENET_MAC1(pldat->net_base));
+       writel((LPC_MAC2_PAD_CRC_ENABLE | LPC_MAC2_CRC_ENABLE),
+              LPC_ENET_MAC2(pldat->net_base));
+       writel(ENET_MAXF_SIZE, LPC_ENET_MAXF(pldat->net_base));
+
+       /* Collision window, gap */
+       writel((LPC_CLRT_LOAD_RETRY_MAX(0xF) |
+               LPC_CLRT_LOAD_COLLISION_WINDOW(0x37)),
+              LPC_ENET_CLRT(pldat->net_base));
+       writel(LPC_IPGR_LOAD_PART2(0x12), LPC_ENET_IPGR(pldat->net_base));
+
+       if (lpc_phy_interface_mode() == PHY_INTERFACE_MODE_MII)
+               writel(LPC_COMMAND_PASSRUNTFRAME,
+                      LPC_ENET_COMMAND(pldat->net_base));
+       else {
+               writel((LPC_COMMAND_PASSRUNTFRAME | LPC_COMMAND_RMII),
+                      LPC_ENET_COMMAND(pldat->net_base));
+               writel(LPC_SUPP_RESET_RMII, LPC_ENET_SUPP(pldat->net_base));
+       }
+
+       __lpc_params_setup(pldat);
+
+       /* Setup TX and RX descriptors */
+       __lpc_txrx_desc_setup(pldat);
+
+       /* Setup packet filtering */
+       writel((LPC_RXFLTRW_ACCEPTUBROADCAST | LPC_RXFLTRW_ACCEPTPERFECT),
+              LPC_ENET_RXFILTER_CTRL(pldat->net_base));
+
+       /* Get the next TX buffer output index */
+       pldat->num_used_tx_buffs = 0;
+       pldat->last_tx_idx =
+               readl(LPC_ENET_TXCONSUMEINDEX(pldat->net_base));
+
+       /* Clear and enable interrupts */
+       writel(0xFFFF, LPC_ENET_INTCLEAR(pldat->net_base));
+       smp_wmb();
+       lpc_eth_enable_int(pldat->net_base);
+
+       /* Enable controller */
+       tmp = readl(LPC_ENET_COMMAND(pldat->net_base));
+       tmp |= LPC_COMMAND_RXENABLE | LPC_COMMAND_TXENABLE;
+       writel(tmp, LPC_ENET_COMMAND(pldat->net_base));
+       tmp = readl(LPC_ENET_MAC1(pldat->net_base));
+       tmp |= LPC_MAC1_RECV_ENABLE;
+       writel(tmp, LPC_ENET_MAC1(pldat->net_base));
+}
+
+static void __lpc_eth_shutdown(struct netdata_local *pldat)
+{
+       /* Reset ethernet and power down PHY */
+       __lpc_eth_reset(pldat);
+       writel(0, LPC_ENET_MAC1(pldat->net_base));
+       writel(0, LPC_ENET_MAC2(pldat->net_base));
+}
+
+/*
+ * MAC<--->PHY support functions
+ */
+static int lpc_mdio_read(struct mii_bus *bus, int phy_id, int phyreg)
+{
+       struct netdata_local *pldat = bus->priv;
+       unsigned long timeout = jiffies + msecs_to_jiffies(100);
+       int lps;
+
+       writel(((phy_id << 8) | phyreg), LPC_ENET_MADR(pldat->net_base));
+       writel(LPC_MCMD_READ, LPC_ENET_MCMD(pldat->net_base));
+
+       /* Wait for unbusy status */
+       while (readl(LPC_ENET_MIND(pldat->net_base)) & LPC_MIND_BUSY) {
+               if (time_after(jiffies, timeout))
+                       return -EIO;
+               cpu_relax();
+       }
+
+       lps = readl(LPC_ENET_MRDD(pldat->net_base));
+       writel(0, LPC_ENET_MCMD(pldat->net_base));
+
+       return lps;
+}
+
+static int lpc_mdio_write(struct mii_bus *bus, int phy_id, int phyreg,
+                       u16 phydata)
+{
+       struct netdata_local *pldat = bus->priv;
+       unsigned long timeout = jiffies + msecs_to_jiffies(100);
+
+       writel(((phy_id << 8) | phyreg), LPC_ENET_MADR(pldat->net_base));
+       writel(phydata, LPC_ENET_MWTD(pldat->net_base));
+
+       /* Wait for completion */
+       while (readl(LPC_ENET_MIND(pldat->net_base)) & LPC_MIND_BUSY) {
+               if (time_after(jiffies, timeout))
+                       return -EIO;
+               cpu_relax();
+       }
+
+       return 0;
+}
+
+static int lpc_mdio_reset(struct mii_bus *bus)
+{
+       return __lpc_mii_mngt_reset((struct netdata_local *)bus->priv);
+}
+
+static void lpc_handle_link_change(struct net_device *ndev)
+{
+       struct netdata_local *pldat = netdev_priv(ndev);
+       struct phy_device *phydev = pldat->phy_dev;
+       unsigned long flags;
+
+       bool status_change = false;
+
+       spin_lock_irqsave(&pldat->lock, flags);
+
+       if (phydev->link) {
+               if ((pldat->speed != phydev->speed) ||
+                   (pldat->duplex != phydev->duplex)) {
+                       pldat->speed = phydev->speed;
+                       pldat->duplex = phydev->duplex;
+                       status_change = true;
+               }
+       }
+
+       if (phydev->link != pldat->link) {
+               if (!phydev->link) {
+                       pldat->speed = 0;
+                       pldat->duplex = -1;
+               }
+               pldat->link = phydev->link;
+
+               status_change = true;
+       }
+
+       spin_unlock_irqrestore(&pldat->lock, flags);
+
+       if (status_change)
+               __lpc_params_setup(pldat);
+}
+
+static int lpc_mii_probe(struct net_device *ndev)
+{
+       struct netdata_local *pldat = netdev_priv(ndev);
+       struct phy_device *phydev = phy_find_first(pldat->mii_bus);
+
+       if (!phydev) {
+               netdev_err(ndev, "no PHY found\n");
+               return -ENODEV;
+       }
+
+       /* Attach to the PHY */
+       if (lpc_phy_interface_mode() == PHY_INTERFACE_MODE_MII)
+               netdev_info(ndev, "using MII interface\n");
+       else
+               netdev_info(ndev, "using RMII interface\n");
+       phydev = phy_connect(ndev, dev_name(&phydev->dev),
+               &lpc_handle_link_change, 0, lpc_phy_interface_mode());
+
+       if (IS_ERR(phydev)) {
+               netdev_err(ndev, "Could not attach to PHY\n");
+               return PTR_ERR(phydev);
+       }
+
+       /* mask with MAC supported features */
+       phydev->supported &= PHY_BASIC_FEATURES;
+
+       phydev->advertising = phydev->supported;
+
+       pldat->link = 0;
+       pldat->speed = 0;
+       pldat->duplex = -1;
+       pldat->phy_dev = phydev;
+
+       netdev_info(ndev,
+               "attached PHY driver [%s] (mii_bus:phy_addr=%s, irq=%d)\n",
+               phydev->drv->name, dev_name(&phydev->dev), phydev->irq);
+       return 0;
+}
+
+static int lpc_mii_init(struct netdata_local *pldat)
+{
+       int err = -ENXIO, i;
+
+       pldat->mii_bus = mdiobus_alloc();
+       if (!pldat->mii_bus) {
+               err = -ENOMEM;
+               goto err_out;
+       }
+
+       /* Setup MII mode */
+       if (lpc_phy_interface_mode() == PHY_INTERFACE_MODE_MII)
+               writel(LPC_COMMAND_PASSRUNTFRAME,
+                      LPC_ENET_COMMAND(pldat->net_base));
+       else {
+               writel((LPC_COMMAND_PASSRUNTFRAME | LPC_COMMAND_RMII),
+                      LPC_ENET_COMMAND(pldat->net_base));
+               writel(LPC_SUPP_RESET_RMII, LPC_ENET_SUPP(pldat->net_base));
+       }
+
+       pldat->mii_bus->name = "lpc_mii_bus";
+       pldat->mii_bus->read = &lpc_mdio_read;
+       pldat->mii_bus->write = &lpc_mdio_write;
+       pldat->mii_bus->reset = &lpc_mdio_reset;
+       snprintf(pldat->mii_bus->id, MII_BUS_ID_SIZE, "%s-%x",
+                pldat->pdev->name, pldat->pdev->id);
+       pldat->mii_bus->priv = pldat;
+       pldat->mii_bus->parent = &pldat->pdev->dev;
+
+       pldat->mii_bus->irq = kmalloc(sizeof(int) * PHY_MAX_ADDR, GFP_KERNEL);
+       if (!pldat->mii_bus->irq) {
+               err = -ENOMEM;
+               goto err_out_1;
+       }
+
+       for (i = 0; i < PHY_MAX_ADDR; i++)
+               pldat->mii_bus->irq[i] = PHY_POLL;
+
+       platform_set_drvdata(pldat->pdev, pldat->mii_bus);
+
+       if (mdiobus_register(pldat->mii_bus))
+               goto err_out_free_mdio_irq;
+
+       if (lpc_mii_probe(pldat->ndev) != 0)
+               goto err_out_unregister_bus;
+
+       return 0;
+
+err_out_unregister_bus:
+       mdiobus_unregister(pldat->mii_bus);
+err_out_free_mdio_irq:
+       kfree(pldat->mii_bus->irq);
+err_out_1:
+       mdiobus_free(pldat->mii_bus);
+err_out:
+       return err;
+}
+
+static void __lpc_handle_xmit(struct net_device *ndev)
+{
+       struct netdata_local *pldat = netdev_priv(ndev);
+       struct sk_buff *skb;
+       u32 txcidx, *ptxstat, txstat;
+
+       txcidx = readl(LPC_ENET_TXCONSUMEINDEX(pldat->net_base));
+       while (pldat->last_tx_idx != txcidx) {
+               skb = pldat->skb[pldat->last_tx_idx];
+
+               /* A buffer is available, get buffer status */
+               ptxstat = &pldat->tx_stat_v[pldat->last_tx_idx];
+               txstat = *ptxstat;
+
+               /* Next buffer and decrement used buffer counter */
+               pldat->num_used_tx_buffs--;
+               pldat->last_tx_idx++;
+               if (pldat->last_tx_idx >= ENET_TX_DESC)
+                       pldat->last_tx_idx = 0;
+
+               /* Update collision counter */
+               ndev->stats.collisions += TXSTATUS_COLLISIONS_GET(txstat);
+
+               /* Any errors occurred? */
+               if (txstat & TXSTATUS_ERROR) {
+                       if (txstat & TXSTATUS_UNDERRUN) {
+                               /* FIFO underrun */
+                               ndev->stats.tx_fifo_errors++;
+                       }
+                       if (txstat & TXSTATUS_LATECOLL) {
+                               /* Late collision */
+                               ndev->stats.tx_aborted_errors++;
+                       }
+                       if (txstat & TXSTATUS_EXCESSCOLL) {
+                               /* Excessive collision */
+                               ndev->stats.tx_aborted_errors++;
+                       }
+                       if (txstat & TXSTATUS_EXCESSDEFER) {
+                               /* Defer limit */
+                               ndev->stats.tx_aborted_errors++;
+                       }
+                       ndev->stats.tx_errors++;
+               } else {
+                       /* Update stats */
+                       ndev->stats.tx_packets++;
+                       ndev->stats.tx_bytes += skb->len;
+
+                       /* Free buffer */
+                       dev_kfree_skb_irq(skb);
+               }
+
+               txcidx = readl(LPC_ENET_TXCONSUMEINDEX(pldat->net_base));
+       }
+
+       if (netif_queue_stopped(ndev))
+               netif_wake_queue(ndev);
+}
+
+static int __lpc_handle_recv(struct net_device *ndev, int budget)
+{
+       struct netdata_local *pldat = netdev_priv(ndev);
+       struct sk_buff *skb;
+       u32 rxconsidx, len, ethst;
+       struct rx_status_t *prxstat;
+       u8 *prdbuf;
+       int rx_done = 0;
+
+       /* Get the current RX buffer indexes */
+       rxconsidx = readl(LPC_ENET_RXCONSUMEINDEX(pldat->net_base));
+       while (rx_done < budget && rxconsidx !=
+                       readl(LPC_ENET_RXPRODUCEINDEX(pldat->net_base))) {
+               /* Get pointer to receive status */
+               prxstat = &pldat->rx_stat_v[rxconsidx];
+               len = (prxstat->statusinfo & RXSTATUS_SIZE) + 1;
+
+               /* Status error? */
+               ethst = prxstat->statusinfo;
+               if ((ethst & (RXSTATUS_ERROR | RXSTATUS_STATUS_ERROR)) ==
+                   (RXSTATUS_ERROR | RXSTATUS_RANGE))
+                       ethst &= ~RXSTATUS_ERROR;
+
+               if (ethst & RXSTATUS_ERROR) {
+                       int si = prxstat->statusinfo;
+                       /* Check statuses */
+                       if (si & RXSTATUS_OVERRUN) {
+                               /* Overrun error */
+                               ndev->stats.rx_fifo_errors++;
+                       } else if (si & RXSTATUS_CRC) {
+                               /* CRC error */
+                               ndev->stats.rx_crc_errors++;
+                       } else if (si & RXSTATUS_LENGTH) {
+                               /* Length error */
+                               ndev->stats.rx_length_errors++;
+                       } else if (si & RXSTATUS_ERROR) {
+                               /* Other error */
+                               ndev->stats.rx_length_errors++;
+                       }
+                       ndev->stats.rx_errors++;
+               } else {
+                       /* Packet is good */
+                       skb = dev_alloc_skb(len + 8);
+                       if (!skb)
+                               ndev->stats.rx_dropped++;
+                       else {
+                               prdbuf = skb_put(skb, len);
+
+                               /* Copy packet from buffer */
+                               memcpy(prdbuf, pldat->rx_buff_v +
+                                       rxconsidx * ENET_MAXF_SIZE, len);
+
+                               /* Pass to upper layer */
+                               skb->protocol = eth_type_trans(skb, ndev);
+                               netif_receive_skb(skb);
+                               ndev->stats.rx_packets++;
+                               ndev->stats.rx_bytes += len;
+                       }
+               }
+
+               /* Increment consume index */
+               rxconsidx = rxconsidx + 1;
+               if (rxconsidx >= ENET_RX_DESC)
+                       rxconsidx = 0;
+               writel(rxconsidx,
+                      LPC_ENET_RXCONSUMEINDEX(pldat->net_base));
+               rx_done++;
+       }
+
+       return rx_done;
+}
+
+static int lpc_eth_poll(struct napi_struct *napi, int budget)
+{
+       struct netdata_local *pldat = container_of(napi,
+                       struct netdata_local, napi);
+       struct net_device *ndev = pldat->ndev;
+       int rx_done = 0;
+       struct netdev_queue *txq = netdev_get_tx_queue(ndev, 0);
+
+       __netif_tx_lock(txq, smp_processor_id());
+       __lpc_handle_xmit(ndev);
+       __netif_tx_unlock(txq);
+       rx_done = __lpc_handle_recv(ndev, budget);
+
+       if (rx_done < budget) {
+               napi_complete(napi);
+               lpc_eth_enable_int(pldat->net_base);
+       }
+
+       return rx_done;
+}
+
+static irqreturn_t __lpc_eth_interrupt(int irq, void *dev_id)
+{
+       struct net_device *ndev = dev_id;
+       struct netdata_local *pldat = netdev_priv(ndev);
+       u32 tmp;
+
+       spin_lock(&pldat->lock);
+
+       tmp = readl(LPC_ENET_INTSTATUS(pldat->net_base));
+       /* Clear interrupts */
+       writel(tmp, LPC_ENET_INTCLEAR(pldat->net_base));
+
+       lpc_eth_disable_int(pldat->net_base);
+       if (likely(napi_schedule_prep(&pldat->napi)))
+               __napi_schedule(&pldat->napi);
+
+       spin_unlock(&pldat->lock);
+
+       return IRQ_HANDLED;
+}
+
+static int lpc_eth_close(struct net_device *ndev)
+{
+       unsigned long flags;
+       struct netdata_local *pldat = netdev_priv(ndev);
+
+       if (netif_msg_ifdown(pldat))
+               dev_dbg(&pldat->pdev->dev, "shutting down %s\n", ndev->name);
+
+       napi_disable(&pldat->napi);
+       netif_stop_queue(ndev);
+
+       if (pldat->phy_dev)
+               phy_stop(pldat->phy_dev);
+
+       spin_lock_irqsave(&pldat->lock, flags);
+       __lpc_eth_reset(pldat);
+       netif_carrier_off(ndev);
+       writel(0, LPC_ENET_MAC1(pldat->net_base));
+       writel(0, LPC_ENET_MAC2(pldat->net_base));
+       spin_unlock_irqrestore(&pldat->lock, flags);
+
+       __lpc_eth_clock_enable(pldat, false);
+
+       return 0;
+}
+
+static int lpc_eth_hard_start_xmit(struct sk_buff *skb, struct net_device *ndev)
+{
+       struct netdata_local *pldat = netdev_priv(ndev);
+       u32 len, txidx;
+       u32 *ptxstat;
+       struct txrx_desc_t *ptxrxdesc;
+
+       len = skb->len;
+
+       spin_lock_irq(&pldat->lock);
+
+       if (pldat->num_used_tx_buffs >= (ENET_TX_DESC - 1)) {
+               /* This function should never be called when there are no
+                  buffers */
+               netif_stop_queue(ndev);
+               spin_unlock_irq(&pldat->lock);
+               WARN(1, "BUG! TX request when no free TX buffers!\n");
+               return NETDEV_TX_BUSY;
+       }
+
+       /* Get the next TX descriptor index */
+       txidx = readl(LPC_ENET_TXPRODUCEINDEX(pldat->net_base));
+
+       /* Setup control for the transfer */
+       ptxstat = &pldat->tx_stat_v[txidx];
+       *ptxstat = 0;
+       ptxrxdesc = &pldat->tx_desc_v[txidx];
+       ptxrxdesc->control =
+               (len - 1) | TXDESC_CONTROL_LAST | TXDESC_CONTROL_INT;
+
+       /* Copy data to the DMA buffer */
+       memcpy(pldat->tx_buff_v + txidx * ENET_MAXF_SIZE, skb->data, len);
+
+       /* Save the buffer and increment the buffer counter */
+       pldat->skb[txidx] = skb;
+       pldat->num_used_tx_buffs++;
+
+       /* Start transmit */
+       txidx++;
+       if (txidx >= ENET_TX_DESC)
+               txidx = 0;
+       writel(txidx, LPC_ENET_TXPRODUCEINDEX(pldat->net_base));
+
+       /* Stop queue if no more TX buffers */
+       if (pldat->num_used_tx_buffs >= (ENET_TX_DESC - 1))
+               netif_stop_queue(ndev);
+
+       spin_unlock_irq(&pldat->lock);
+
+       return NETDEV_TX_OK;
+}
+
+static int lpc_set_mac_address(struct net_device *ndev, void *p)
+{
+       struct sockaddr *addr = p;
+       struct netdata_local *pldat = netdev_priv(ndev);
+       unsigned long flags;
+
+       if (!is_valid_ether_addr(addr->sa_data))
+               return -EADDRNOTAVAIL;
+       memcpy(ndev->dev_addr, addr->sa_data, ETH_ALEN);
+
+       spin_lock_irqsave(&pldat->lock, flags);
+
+       /* Set station address */
+       __lpc_set_mac(pldat, ndev->dev_addr);
+
+       spin_unlock_irqrestore(&pldat->lock, flags);
+
+       return 0;
+}
+
+static void lpc_eth_set_multicast_list(struct net_device *ndev)
+{
+       struct netdata_local *pldat = netdev_priv(ndev);
+       struct netdev_hw_addr_list *mcptr = &ndev->mc;
+       struct netdev_hw_addr *ha;
+       u32 tmp32, hash_val, hashlo, hashhi;
+       unsigned long flags;
+
+       spin_lock_irqsave(&pldat->lock, flags);
+
+       /* Set station address */
+       __lpc_set_mac(pldat, ndev->dev_addr);
+
+       tmp32 =  LPC_RXFLTRW_ACCEPTUBROADCAST | LPC_RXFLTRW_ACCEPTPERFECT;
+
+       if (ndev->flags & IFF_PROMISC)
+               tmp32 |= LPC_RXFLTRW_ACCEPTUNICAST |
+                       LPC_RXFLTRW_ACCEPTUMULTICAST;
+       if (ndev->flags & IFF_ALLMULTI)
+               tmp32 |= LPC_RXFLTRW_ACCEPTUMULTICAST;
+
+       if (netdev_hw_addr_list_count(mcptr))
+               tmp32 |= LPC_RXFLTRW_ACCEPTUMULTICASTHASH;
+
+       writel(tmp32, LPC_ENET_RXFILTER_CTRL(pldat->net_base));
+
+
+       /* Set initial hash table */
+       hashlo = 0x0;
+       hashhi = 0x0;
+
+       /* 64 bits : multicast address in hash table */
+       netdev_hw_addr_list_for_each(ha, mcptr) {
+               hash_val = (ether_crc(6, ha->addr) >> 23) & 0x3F;
+
+               if (hash_val >= 32)
+                       hashhi |= 1 << (hash_val - 32);
+               else
+                       hashlo |= 1 << hash_val;
+       }
+
+       writel(hashlo, LPC_ENET_HASHFILTERL(pldat->net_base));
+       writel(hashhi, LPC_ENET_HASHFILTERH(pldat->net_base));
+
+       spin_unlock_irqrestore(&pldat->lock, flags);
+}
+
+static int lpc_eth_ioctl(struct net_device *ndev, struct ifreq *req, int cmd)
+{
+       struct netdata_local *pldat = netdev_priv(ndev);
+       struct phy_device *phydev = pldat->phy_dev;
+
+       if (!netif_running(ndev))
+               return -EINVAL;
+
+       if (!phydev)
+               return -ENODEV;
+
+       return phy_mii_ioctl(phydev, req, cmd);
+}
+
+static int lpc_eth_open(struct net_device *ndev)
+{
+       struct netdata_local *pldat = netdev_priv(ndev);
+
+       if (netif_msg_ifup(pldat))
+               dev_dbg(&pldat->pdev->dev, "enabling %s\n", ndev->name);
+
+       if (!is_valid_ether_addr(ndev->dev_addr))
+               return -EADDRNOTAVAIL;
+
+       __lpc_eth_clock_enable(pldat, true);
+
+       /* Reset and initialize */
+       __lpc_eth_reset(pldat);
+       __lpc_eth_init(pldat);
+
+       /* schedule a link state check */
+       phy_start(pldat->phy_dev);
+       netif_start_queue(ndev);
+       napi_enable(&pldat->napi);
+
+       return 0;
+}
+
+/*
+ * Ethtool ops
+ */
+static void lpc_eth_ethtool_getdrvinfo(struct net_device *ndev,
+       struct ethtool_drvinfo *info)
+{
+       strcpy(info->driver, MODNAME);
+       strcpy(info->version, DRV_VERSION);
+       strcpy(info->bus_info, dev_name(ndev->dev.parent));
+}
+
+static u32 lpc_eth_ethtool_getmsglevel(struct net_device *ndev)
+{
+       struct netdata_local *pldat = netdev_priv(ndev);
+
+       return pldat->msg_enable;
+}
+
+static void lpc_eth_ethtool_setmsglevel(struct net_device *ndev, u32 level)
+{
+       struct netdata_local *pldat = netdev_priv(ndev);
+
+       pldat->msg_enable = level;
+}
+
+static int lpc_eth_ethtool_getsettings(struct net_device *ndev,
+       struct ethtool_cmd *cmd)
+{
+       struct netdata_local *pldat = netdev_priv(ndev);
+       struct phy_device *phydev = pldat->phy_dev;
+
+       if (!phydev)
+               return -EOPNOTSUPP;
+
+       return phy_ethtool_gset(phydev, cmd);
+}
+
+static int lpc_eth_ethtool_setsettings(struct net_device *ndev,
+       struct ethtool_cmd *cmd)
+{
+       struct netdata_local *pldat = netdev_priv(ndev);
+       struct phy_device *phydev = pldat->phy_dev;
+
+       if (!phydev)
+               return -EOPNOTSUPP;
+
+       return phy_ethtool_sset(phydev, cmd);
+}
+
+static const struct ethtool_ops lpc_eth_ethtool_ops = {
+       .get_drvinfo    = lpc_eth_ethtool_getdrvinfo,
+       .get_settings   = lpc_eth_ethtool_getsettings,
+       .set_settings   = lpc_eth_ethtool_setsettings,
+       .get_msglevel   = lpc_eth_ethtool_getmsglevel,
+       .set_msglevel   = lpc_eth_ethtool_setmsglevel,
+       .get_link       = ethtool_op_get_link,
+};
+
+static const struct net_device_ops lpc_netdev_ops = {
+       .ndo_open               = lpc_eth_open,
+       .ndo_stop               = lpc_eth_close,
+       .ndo_start_xmit         = lpc_eth_hard_start_xmit,
+       .ndo_set_rx_mode        = lpc_eth_set_multicast_list,
+       .ndo_do_ioctl           = lpc_eth_ioctl,
+       .ndo_set_mac_address    = lpc_set_mac_address,
+};
+
+static int lpc_eth_drv_probe(struct platform_device *pdev)
+{
+       struct resource *res;
+       struct resource *dma_res;
+       struct net_device *ndev;
+       struct netdata_local *pldat;
+       struct phy_device *phydev;
+       dma_addr_t dma_handle;
+       int irq, ret;
+
+       /* Get platform resources */
+       res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+       dma_res = platform_get_resource(pdev, IORESOURCE_MEM, 1);
+       irq = platform_get_irq(pdev, 0);
+       if ((!res) || (!dma_res) || (irq < 0) || (irq >= NR_IRQS)) {
+               dev_err(&pdev->dev, "error getting resources.\n");
+               ret = -ENXIO;
+               goto err_exit;
+       }
+
+       /* Allocate net driver data structure */
+       ndev = alloc_etherdev(sizeof(struct netdata_local));
+       if (!ndev) {
+               dev_err(&pdev->dev, "could not allocate device.\n");
+               ret = -ENOMEM;
+               goto err_exit;
+       }
+
+       SET_NETDEV_DEV(ndev, &pdev->dev);
+
+       pldat = netdev_priv(ndev);
+       pldat->pdev = pdev;
+       pldat->ndev = ndev;
+
+       spin_lock_init(&pldat->lock);
+
+       /* Save resources */
+       ndev->irq = irq;
+
+       /* Get clock for the device */
+       pldat->clk = clk_get(&pdev->dev, NULL);
+       if (IS_ERR(pldat->clk)) {
+               dev_err(&pdev->dev, "error getting clock.\n");
+               ret = PTR_ERR(pldat->clk);
+               goto err_out_free_dev;
+       }
+
+       /* Enable network clock */
+       __lpc_eth_clock_enable(pldat, true);
+
+       /* Map IO space */
+       pldat->net_base = ioremap(res->start, res->end - res->start + 1);
+       if (!pldat->net_base) {
+               dev_err(&pdev->dev, "failed to map registers\n");
+               ret = -ENOMEM;
+               goto err_out_disable_clocks;
+       }
+       ret = request_irq(ndev->irq, __lpc_eth_interrupt, 0,
+                         ndev->name, ndev);
+       if (ret) {
+               dev_err(&pdev->dev, "error requesting interrupt.\n");
+               goto err_out_iounmap;
+       }
+
+       /* Fill in the fields of the device structure with ethernet values. */
+       ether_setup(ndev);
+
+       /* Setup driver functions */
+       ndev->netdev_ops = &lpc_netdev_ops;
+       ndev->ethtool_ops = &lpc_eth_ethtool_ops;
+       ndev->watchdog_timeo = msecs_to_jiffies(2500);
+
+       /* Get size of DMA buffers/descriptors region */
+       pldat->dma_buff_size = (ENET_TX_DESC + ENET_RX_DESC) * (ENET_MAXF_SIZE +
+               sizeof(struct txrx_desc_t) + sizeof(struct rx_status_t));
+       pldat->dma_buff_base_v = 0;
+
+       if (use_iram_for_net()) {
+               dma_handle = dma_res->start;
+               if (pldat->dma_buff_size <= lpc32xx_return_iram_size())
+                       pldat->dma_buff_base_v =
+                               io_p2v(dma_res->start);
+               else
+                       netdev_err(ndev,
+                               "IRAM not big enough for net buffers, using SDRAM instead.\n");
+       }
+
+       if (pldat->dma_buff_base_v == 0) {
+               pldat->dma_buff_size = PAGE_ALIGN(pldat->dma_buff_size);
+
+               /* Allocate a chunk of memory for the DMA ethernet buffers
+                  and descriptors */
+               pldat->dma_buff_base_v =
+                       dma_alloc_coherent(&pldat->pdev->dev,
+                                          pldat->dma_buff_size, &dma_handle,
+                                          GFP_KERNEL);
+
+               if (pldat->dma_buff_base_v == NULL) {
+                       dev_err(&pdev->dev, "error getting DMA region.\n");
+                       ret = -ENOMEM;
+                       goto err_out_free_irq;
+               }
+       }
+       pldat->dma_buff_base_p = dma_handle;
+
+       netdev_dbg(ndev, "IO address start     :0x%08x\n",
+                       res->start);
+       netdev_dbg(ndev, "IO address size      :%d\n",
+                       res->end - res->start + 1);
+       netdev_err(ndev, "IO address (mapped)  :0x%p\n",
+                       pldat->net_base);
+       netdev_dbg(ndev, "IRQ number           :%d\n", ndev->irq);
+       netdev_dbg(ndev, "DMA buffer size      :%d\n", pldat->dma_buff_size);
+       netdev_dbg(ndev, "DMA buffer P address :0x%08x\n",
+                       pldat->dma_buff_base_p);
+       netdev_dbg(ndev, "DMA buffer V address :0x%p\n",
+                       pldat->dma_buff_base_v);
+
+       /* Get MAC address from current HW setting (POR state is all zeros) */
+       __lpc_get_mac(pldat, ndev->dev_addr);
+
+#ifdef CONFIG_OF_NET
+       if (!is_valid_ether_addr(ndev->dev_addr)) {
+               const char *macaddr = of_get_mac_address(pdev->dev.of_node);
+               if (macaddr)
+                       memcpy(ndev->dev_addr, macaddr, ETH_ALEN);
+       }
+#endif
+       if (!is_valid_ether_addr(ndev->dev_addr))
+               dev_hw_addr_random(ndev, ndev->dev_addr);
+
+       /* Reset the ethernet controller */
+       __lpc_eth_reset(pldat);
+
+       /* then shut everything down to save power */
+       __lpc_eth_shutdown(pldat);
+
+       /* Set default parameters */
+       pldat->msg_enable = NETIF_MSG_LINK;
+
+       /* Force an MII interface reset and clock setup */
+       __lpc_mii_mngt_reset(pldat);
+
+       /* Force default PHY interface setup in chip, this will probably be
+          changed by the PHY driver */
+       pldat->link = 0;
+       pldat->speed = 100;
+       pldat->duplex = DUPLEX_FULL;
+       __lpc_params_setup(pldat);
+
+       netif_napi_add(ndev, &pldat->napi, lpc_eth_poll, NAPI_WEIGHT);
+
+       ret = register_netdev(ndev);
+       if (ret) {
+               dev_err(&pdev->dev, "Cannot register net device, aborting.\n");
+               goto err_out_dma_unmap;
+       }
+       platform_set_drvdata(pdev, ndev);
+
+       if (lpc_mii_init(pldat) != 0)
+               goto err_out_unregister_netdev;
+
+       netdev_info(ndev, "LPC mac at 0x%08x irq %d\n",
+              res->start, ndev->irq);
+
+       phydev = pldat->phy_dev;
+
+       device_init_wakeup(&pdev->dev, 1);
+       device_set_wakeup_enable(&pdev->dev, 0);
+
+       return 0;
+
+err_out_unregister_netdev:
+       platform_set_drvdata(pdev, NULL);
+       unregister_netdev(ndev);
+err_out_dma_unmap:
+       if (!use_iram_for_net() ||
+           pldat->dma_buff_size > lpc32xx_return_iram_size())
+               dma_free_coherent(&pldat->pdev->dev, pldat->dma_buff_size,
+                                 pldat->dma_buff_base_v,
+                                 pldat->dma_buff_base_p);
+err_out_free_irq:
+       free_irq(ndev->irq, ndev);
+err_out_iounmap:
+       iounmap(pldat->net_base);
+err_out_disable_clocks:
+       clk_disable(pldat->clk);
+       clk_put(pldat->clk);
+err_out_free_dev:
+       free_netdev(ndev);
+err_exit:
+       pr_err("%s: not found (%d).\n", MODNAME, ret);
+       return ret;
+}
+
+static int lpc_eth_drv_remove(struct platform_device *pdev)
+{
+       struct net_device *ndev = platform_get_drvdata(pdev);
+       struct netdata_local *pldat = netdev_priv(ndev);
+
+       unregister_netdev(ndev);
+       platform_set_drvdata(pdev, NULL);
+
+       if (!use_iram_for_net() ||
+           pldat->dma_buff_size > lpc32xx_return_iram_size())
+               dma_free_coherent(&pldat->pdev->dev, pldat->dma_buff_size,
+                                 pldat->dma_buff_base_v,
+                                 pldat->dma_buff_base_p);
+       free_irq(ndev->irq, ndev);
+       iounmap(pldat->net_base);
+       mdiobus_free(pldat->mii_bus);
+       clk_disable(pldat->clk);
+       clk_put(pldat->clk);
+       free_netdev(ndev);
+
+       return 0;
+}
+
+#ifdef CONFIG_PM
+static int lpc_eth_drv_suspend(struct platform_device *pdev,
+       pm_message_t state)
+{
+       struct net_device *ndev = platform_get_drvdata(pdev);
+       struct netdata_local *pldat = netdev_priv(ndev);
+
+       if (device_may_wakeup(&pdev->dev))
+               enable_irq_wake(ndev->irq);
+
+       if (ndev) {
+               if (netif_running(ndev)) {
+                       netif_device_detach(ndev);
+                       __lpc_eth_shutdown(pldat);
+                       clk_disable(pldat->clk);
+
+                       /*
+                        * Reset again now clock is disable to be sure
+                        * EMC_MDC is down
+                        */
+                       __lpc_eth_reset(pldat);
+               }
+       }
+
+       return 0;
+}
+
+static int lpc_eth_drv_resume(struct platform_device *pdev)
+{
+       struct net_device *ndev = platform_get_drvdata(pdev);
+       struct netdata_local *pldat;
+
+       if (device_may_wakeup(&pdev->dev))
+               disable_irq_wake(ndev->irq);
+
+       if (ndev) {
+               if (netif_running(ndev)) {
+                       pldat = netdev_priv(ndev);
+
+                       /* Enable interface clock */
+                       clk_enable(pldat->clk);
+
+                       /* Reset and initialize */
+                       __lpc_eth_reset(pldat);
+                       __lpc_eth_init(pldat);
+
+                       netif_device_attach(ndev);
+               }
+       }
+
+       return 0;
+}
+#endif
+
+static struct platform_driver lpc_eth_driver = {
+       .probe          = lpc_eth_drv_probe,
+       .remove         = __devexit_p(lpc_eth_drv_remove),
+#ifdef CONFIG_PM
+       .suspend        = lpc_eth_drv_suspend,
+       .resume         = lpc_eth_drv_resume,
+#endif
+       .driver         = {
+               .name   = MODNAME,
+       },
+};
+
+module_platform_driver(lpc_eth_driver);
+
+MODULE_AUTHOR("Kevin Wells <kevin.wells@nxp.com>");
+MODULE_AUTHOR("Roland Stigge <stigge@antcom.de>");
+MODULE_DESCRIPTION("LPC Ethernet Driver");
+MODULE_LICENSE("GPL");