From 4796417417a62e2ae83d92cb92e1ecf9ec67b5f5 Mon Sep 17 00:00:00 2001 From: Ilya Yanok Date: Wed, 11 Mar 2009 23:26:02 -0700 Subject: dnet: Dave DNET ethernet controller driver (updated) Driver for Dave DNET ethernet controller found on Dave/DENX QongEVB-LITE FPGA. Heavily based on Dave sources, I've just adopted it to current kernel version and done some code cleanup. Signed-off-by: Ilya Yanok Signed-off-by: David S. Miller --- drivers/net/dnet.c | 994 +++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 994 insertions(+) create mode 100644 drivers/net/dnet.c (limited to 'drivers/net/dnet.c') diff --git a/drivers/net/dnet.c b/drivers/net/dnet.c new file mode 100644 index 00000000000..92c3bd3a472 --- /dev/null +++ b/drivers/net/dnet.c @@ -0,0 +1,994 @@ +/* + * Dave DNET Ethernet Controller driver + * + * Copyright (C) 2008 Dave S.r.l. + * Copyright (C) 2009 Ilya Yanok, Emcraft Systems Ltd, + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include "dnet.h" + +#undef DEBUG + +/* function for reading internal MAC register */ +u16 dnet_readw_mac(struct dnet *bp, u16 reg) +{ + u16 data_read; + + /* issue a read */ + dnet_writel(bp, reg, MACREG_ADDR); + + /* since a read/write op to the MAC is very slow, + * we must wait before reading the data */ + ndelay(500); + + /* read data read from the MAC register */ + data_read = dnet_readl(bp, MACREG_DATA); + + /* all done */ + return data_read; +} + +/* function for writing internal MAC register */ +void dnet_writew_mac(struct dnet *bp, u16 reg, u16 val) +{ + /* load data to write */ + dnet_writel(bp, val, MACREG_DATA); + + /* issue a write */ + dnet_writel(bp, reg | DNET_INTERNAL_WRITE, MACREG_ADDR); + + /* since a read/write op to the MAC is very slow, + * we must wait before exiting */ + ndelay(500); +} + +static void __dnet_set_hwaddr(struct dnet *bp) +{ + u16 tmp; + + tmp = cpu_to_be16(*((u16 *) bp->dev->dev_addr)); + dnet_writew_mac(bp, DNET_INTERNAL_MAC_ADDR_0_REG, tmp); + tmp = cpu_to_be16(*((u16 *) (bp->dev->dev_addr + 2))); + dnet_writew_mac(bp, DNET_INTERNAL_MAC_ADDR_1_REG, tmp); + tmp = cpu_to_be16(*((u16 *) (bp->dev->dev_addr + 4))); + dnet_writew_mac(bp, DNET_INTERNAL_MAC_ADDR_2_REG, tmp); +} + +static void __devinit dnet_get_hwaddr(struct dnet *bp) +{ + u16 tmp; + u8 addr[6]; + + /* + * from MAC docs: + * "Note that the MAC address is stored in the registers in Hexadecimal + * form. For example, to set the MAC Address to: AC-DE-48-00-00-80 + * would require writing 0xAC (octet 0) to address 0x0B (high byte of + * Mac_addr[15:0]), 0xDE (octet 1) to address 0x0A (Low byte of + * Mac_addr[15:0]), 0x48 (octet 2) to address 0x0D (high byte of + * Mac_addr[15:0]), 0x00 (octet 3) to address 0x0C (Low byte of + * Mac_addr[15:0]), 0x00 (octet 4) to address 0x0F (high byte of + * Mac_addr[15:0]), and 0x80 (octet 5) to address * 0x0E (Low byte of + * Mac_addr[15:0]). + */ + tmp = dnet_readw_mac(bp, DNET_INTERNAL_MAC_ADDR_0_REG); + *((u16 *) addr) = be16_to_cpu(tmp); + tmp = dnet_readw_mac(bp, DNET_INTERNAL_MAC_ADDR_1_REG); + *((u16 *) (addr + 2)) = be16_to_cpu(tmp); + tmp = dnet_readw_mac(bp, DNET_INTERNAL_MAC_ADDR_2_REG); + *((u16 *) (addr + 4)) = be16_to_cpu(tmp); + + if (is_valid_ether_addr(addr)) + memcpy(bp->dev->dev_addr, addr, sizeof(addr)); +} + +static int dnet_mdio_read(struct mii_bus *bus, int mii_id, int regnum) +{ + struct dnet *bp = bus->priv; + u16 value; + + while (!(dnet_readw_mac(bp, DNET_INTERNAL_GMII_MNG_CTL_REG) + & DNET_INTERNAL_GMII_MNG_CMD_FIN)) + cpu_relax(); + + /* only 5 bits allowed for phy-addr and reg_offset */ + mii_id &= 0x1f; + regnum &= 0x1f; + + /* prepare reg_value for a read */ + value = (mii_id << 8); + value |= regnum; + + /* write control word */ + dnet_writew_mac(bp, DNET_INTERNAL_GMII_MNG_CTL_REG, value); + + /* wait for end of transfer */ + while (!(dnet_readw_mac(bp, DNET_INTERNAL_GMII_MNG_CTL_REG) + & DNET_INTERNAL_GMII_MNG_CMD_FIN)) + cpu_relax(); + + value = dnet_readw_mac(bp, DNET_INTERNAL_GMII_MNG_DAT_REG); + + pr_debug("mdio_read %02x:%02x <- %04x\n", mii_id, regnum, value); + + return value; +} + +static int dnet_mdio_write(struct mii_bus *bus, int mii_id, int regnum, + u16 value) +{ + struct dnet *bp = bus->priv; + u16 tmp; + + pr_debug("mdio_write %02x:%02x <- %04x\n", mii_id, regnum, value); + + while (!(dnet_readw_mac(bp, DNET_INTERNAL_GMII_MNG_CTL_REG) + & DNET_INTERNAL_GMII_MNG_CMD_FIN)) + cpu_relax(); + + /* prepare for a write operation */ + tmp = (1 << 13); + + /* only 5 bits allowed for phy-addr and reg_offset */ + mii_id &= 0x1f; + regnum &= 0x1f; + + /* only 16 bits on data */ + value &= 0xffff; + + /* prepare reg_value for a write */ + tmp |= (mii_id << 8); + tmp |= regnum; + + /* write data to write first */ + dnet_writew_mac(bp, DNET_INTERNAL_GMII_MNG_DAT_REG, value); + + /* write control word */ + dnet_writew_mac(bp, DNET_INTERNAL_GMII_MNG_CTL_REG, tmp); + + while (!(dnet_readw_mac(bp, DNET_INTERNAL_GMII_MNG_CTL_REG) + & DNET_INTERNAL_GMII_MNG_CMD_FIN)) + cpu_relax(); + + return 0; +} + +static int dnet_mdio_reset(struct mii_bus *bus) +{ + return 0; +} + +static void dnet_handle_link_change(struct net_device *dev) +{ + struct dnet *bp = netdev_priv(dev); + struct phy_device *phydev = bp->phy_dev; + unsigned long flags; + u32 mode_reg, ctl_reg; + + int status_change = 0; + + spin_lock_irqsave(&bp->lock, flags); + + mode_reg = dnet_readw_mac(bp, DNET_INTERNAL_MODE_REG); + ctl_reg = dnet_readw_mac(bp, DNET_INTERNAL_RXTX_CONTROL_REG); + + if (phydev->link) { + if (bp->duplex != phydev->duplex) { + if (phydev->duplex) + ctl_reg &= + ~(DNET_INTERNAL_RXTX_CONTROL_ENABLEHALFDUP); + else + ctl_reg |= + DNET_INTERNAL_RXTX_CONTROL_ENABLEHALFDUP; + + bp->duplex = phydev->duplex; + status_change = 1; + } + + if (bp->speed != phydev->speed) { + status_change = 1; + switch (phydev->speed) { + case 1000: + mode_reg |= DNET_INTERNAL_MODE_GBITEN; + break; + case 100: + case 10: + mode_reg &= ~DNET_INTERNAL_MODE_GBITEN; + break; + default: + printk(KERN_WARNING + "%s: Ack! Speed (%d) is not " + "10/100/1000!\n", dev->name, + phydev->speed); + break; + } + bp->speed = phydev->speed; + } + } + + if (phydev->link != bp->link) { + if (phydev->link) { + mode_reg |= + (DNET_INTERNAL_MODE_RXEN | DNET_INTERNAL_MODE_TXEN); + } else { + mode_reg &= + ~(DNET_INTERNAL_MODE_RXEN | + DNET_INTERNAL_MODE_TXEN); + bp->speed = 0; + bp->duplex = -1; + } + bp->link = phydev->link; + + status_change = 1; + } + + if (status_change) { + dnet_writew_mac(bp, DNET_INTERNAL_RXTX_CONTROL_REG, ctl_reg); + dnet_writew_mac(bp, DNET_INTERNAL_MODE_REG, mode_reg); + } + + spin_unlock_irqrestore(&bp->lock, flags); + + if (status_change) { + if (phydev->link) + printk(KERN_INFO "%s: link up (%d/%s)\n", + dev->name, phydev->speed, + DUPLEX_FULL == phydev->duplex ? "Full" : "Half"); + else + printk(KERN_INFO "%s: link down\n", dev->name); + } +} + +static int dnet_mii_probe(struct net_device *dev) +{ + struct dnet *bp = netdev_priv(dev); + struct phy_device *phydev = NULL; + int phy_addr; + + /* find the first phy */ + for (phy_addr = 0; phy_addr < PHY_MAX_ADDR; phy_addr++) { + if (bp->mii_bus->phy_map[phy_addr]) { + phydev = bp->mii_bus->phy_map[phy_addr]; + break; + } + } + + if (!phydev) { + printk(KERN_ERR "%s: no PHY found\n", dev->name); + return -ENODEV; + } + + /* TODO : add pin_irq */ + + /* attach the mac to the phy */ + if (bp->capabilities & DNET_HAS_RMII) { + phydev = phy_connect(dev, phydev->dev.bus_id, + &dnet_handle_link_change, 0, + PHY_INTERFACE_MODE_RMII); + } else { + phydev = phy_connect(dev, phydev->dev.bus_id, + &dnet_handle_link_change, 0, + PHY_INTERFACE_MODE_MII); + } + + if (IS_ERR(phydev)) { + printk(KERN_ERR "%s: Could not attach to PHY\n", dev->name); + return PTR_ERR(phydev); + } + + /* mask with MAC supported features */ + if (bp->capabilities & DNET_HAS_GIGABIT) + phydev->supported &= PHY_GBIT_FEATURES; + else + phydev->supported &= PHY_BASIC_FEATURES; + + phydev->supported |= SUPPORTED_Asym_Pause | SUPPORTED_Pause; + + phydev->advertising = phydev->supported; + + bp->link = 0; + bp->speed = 0; + bp->duplex = -1; + bp->phy_dev = phydev; + + return 0; +} + +static int dnet_mii_init(struct dnet *bp) +{ + int err, i; + + bp->mii_bus = mdiobus_alloc(); + if (bp->mii_bus == NULL) + return -ENOMEM; + + bp->mii_bus->name = "dnet_mii_bus"; + bp->mii_bus->read = &dnet_mdio_read; + bp->mii_bus->write = &dnet_mdio_write; + bp->mii_bus->reset = &dnet_mdio_reset; + + snprintf(bp->mii_bus->id, MII_BUS_ID_SIZE, "%x", 0); + + bp->mii_bus->priv = bp; + + bp->mii_bus->irq = kmalloc(sizeof(int) * PHY_MAX_ADDR, GFP_KERNEL); + if (!bp->mii_bus->irq) { + err = -ENOMEM; + goto err_out; + } + + for (i = 0; i < PHY_MAX_ADDR; i++) + bp->mii_bus->irq[i] = PHY_POLL; + + platform_set_drvdata(bp->dev, bp->mii_bus); + + if (mdiobus_register(bp->mii_bus)) { + err = -ENXIO; + goto err_out_free_mdio_irq; + } + + if (dnet_mii_probe(bp->dev) != 0) { + err = -ENXIO; + goto err_out_unregister_bus; + } + + return 0; + +err_out_unregister_bus: + mdiobus_unregister(bp->mii_bus); +err_out_free_mdio_irq: + kfree(bp->mii_bus->irq); +err_out: + mdiobus_free(bp->mii_bus); + return err; +} + +/* For Neptune board: LINK1000 as Link LED and TX as activity LED */ +int dnet_phy_marvell_fixup(struct phy_device *phydev) +{ + return phy_write(phydev, 0x18, 0x4148); +} + +static void dnet_update_stats(struct dnet *bp) +{ + u32 __iomem *reg = bp->regs + DNET_RX_PKT_IGNR_CNT; + u32 *p = &bp->hw_stats.rx_pkt_ignr; + u32 *end = &bp->hw_stats.rx_byte + 1; + + WARN_ON((unsigned long)(end - p - 1) != + (DNET_RX_BYTE_CNT - DNET_RX_PKT_IGNR_CNT) / 4); + + for (; p < end; p++, reg++) + *p += readl(reg); + + reg = bp->regs + DNET_TX_UNICAST_CNT; + p = &bp->hw_stats.tx_unicast; + end = &bp->hw_stats.tx_byte + 1; + + WARN_ON((unsigned long)(end - p - 1) != + (DNET_TX_BYTE_CNT - DNET_TX_UNICAST_CNT) / 4); + + for (; p < end; p++, reg++) + *p += readl(reg); +} + +static int dnet_poll(struct napi_struct *napi, int budget) +{ + struct dnet *bp = container_of(napi, struct dnet, napi); + struct net_device *dev = bp->dev; + int npackets = 0; + unsigned int pkt_len; + struct sk_buff *skb; + unsigned int *data_ptr; + u32 int_enable; + u32 cmd_word; + int i; + + while (npackets < budget) { + /* + * break out of while loop if there are no more + * packets waiting + */ + if (!(dnet_readl(bp, RX_FIFO_WCNT) >> 16)) { + netif_rx_complete(napi); + int_enable = dnet_readl(bp, INTR_ENB); + int_enable |= DNET_INTR_SRC_RX_CMDFIFOAF; + dnet_writel(bp, int_enable, INTR_ENB); + return 0; + } + + cmd_word = dnet_readl(bp, RX_LEN_FIFO); + pkt_len = cmd_word & 0xFFFF; + + if (cmd_word & 0xDF180000) + printk(KERN_ERR "%s packet receive error %x\n", + __func__, cmd_word); + + skb = dev_alloc_skb(pkt_len + 5); + if (skb != NULL) { + /* Align IP on 16 byte boundaries */ + skb_reserve(skb, 2); + /* + * 'skb_put()' points to the start of sk_buff + * data area. + */ + data_ptr = (unsigned int *)skb_put(skb, pkt_len); + for (i = 0; i < (pkt_len + 3) >> 2; i++) + *data_ptr++ = dnet_readl(bp, RX_DATA_FIFO); + skb->protocol = eth_type_trans(skb, dev); + netif_receive_skb(skb); + npackets++; + } else + printk(KERN_NOTICE + "%s: No memory to allocate a sk_buff of " + "size %u.\n", dev->name, pkt_len); + } + + budget -= npackets; + + if (npackets < budget) { + /* We processed all packets available. Tell NAPI it can + * stop polling then re-enable rx interrupts */ + netif_rx_complete(napi); + int_enable = dnet_readl(bp, INTR_ENB); + int_enable |= DNET_INTR_SRC_RX_CMDFIFOAF; + dnet_writel(bp, int_enable, INTR_ENB); + return 0; + } + + /* There are still packets waiting */ + return 1; +} + +static irqreturn_t dnet_interrupt(int irq, void *dev_id) +{ + struct net_device *dev = dev_id; + struct dnet *bp = netdev_priv(dev); + u32 int_src, int_enable, int_current; + unsigned long flags; + unsigned int handled = 0; + + spin_lock_irqsave(&bp->lock, flags); + + /* read and clear the DNET irq (clear on read) */ + int_src = dnet_readl(bp, INTR_SRC); + int_enable = dnet_readl(bp, INTR_ENB); + int_current = int_src & int_enable; + + /* restart the queue if we had stopped it for TX fifo almost full */ + if (int_current & DNET_INTR_SRC_TX_FIFOAE) { + int_enable = dnet_readl(bp, INTR_ENB); + int_enable &= ~DNET_INTR_ENB_TX_FIFOAE; + dnet_writel(bp, int_enable, INTR_ENB); + netif_wake_queue(dev); + handled = 1; + } + + /* RX FIFO error checking */ + if (int_current & + (DNET_INTR_SRC_RX_CMDFIFOFF | DNET_INTR_SRC_RX_DATAFIFOFF)) { + printk(KERN_ERR "%s: RX fifo error %x, irq %x\n", __func__, + dnet_readl(bp, RX_STATUS), int_current); + /* we can only flush the RX FIFOs */ + dnet_writel(bp, DNET_SYS_CTL_RXFIFOFLUSH, SYS_CTL); + ndelay(500); + dnet_writel(bp, 0, SYS_CTL); + handled = 1; + } + + /* TX FIFO error checking */ + if (int_current & + (DNET_INTR_SRC_TX_FIFOFULL | DNET_INTR_SRC_TX_DISCFRM)) { + printk(KERN_ERR "%s: TX fifo error %x, irq %x\n", __func__, + dnet_readl(bp, TX_STATUS), int_current); + /* we can only flush the TX FIFOs */ + dnet_writel(bp, DNET_SYS_CTL_TXFIFOFLUSH, SYS_CTL); + ndelay(500); + dnet_writel(bp, 0, SYS_CTL); + handled = 1; + } + + if (int_current & DNET_INTR_SRC_RX_CMDFIFOAF) { + if (netif_rx_schedule_prep(&bp->napi)) { + /* + * There's no point taking any more interrupts + * until we have processed the buffers + */ + /* Disable Rx interrupts and schedule NAPI poll */ + int_enable = dnet_readl(bp, INTR_ENB); + int_enable &= ~DNET_INTR_SRC_RX_CMDFIFOAF; + dnet_writel(bp, int_enable, INTR_ENB); + __netif_rx_schedule(&bp->napi); + } + handled = 1; + } + + if (!handled) + pr_debug("%s: irq %x remains\n", __func__, int_current); + + spin_unlock_irqrestore(&bp->lock, flags); + + return IRQ_RETVAL(handled); +} + +#ifdef DEBUG +static inline void dnet_print_skb(struct sk_buff *skb) +{ + int k; + printk(KERN_DEBUG PFX "data:"); + for (k = 0; k < skb->len; k++) + printk(" %02x", (unsigned int)skb->data[k]); + printk("\n"); +} +#else +#define dnet_print_skb(skb) do {} while (0) +#endif + +static int dnet_start_xmit(struct sk_buff *skb, struct net_device *dev) +{ + + struct dnet *bp = netdev_priv(dev); + u32 tx_status, irq_enable; + unsigned int len, i, tx_cmd, wrsz; + unsigned long flags; + unsigned int *bufp; + + tx_status = dnet_readl(bp, TX_STATUS); + + pr_debug("start_xmit: len %u head %p data %p tail %p end %p\n", + skb->len, skb->head, skb->data, skb->tail, skb->end); + dnet_print_skb(skb); + + /* frame size (words) */ + len = (skb->len + 3) >> 2; + + spin_lock_irqsave(&bp->lock, flags); + + tx_status = dnet_readl(bp, TX_STATUS); + + bufp = (unsigned int *)(((u32) skb->data) & 0xFFFFFFFC); + wrsz = (u32) skb->len + 3; + wrsz += ((u32) skb->data) & 0x3; + wrsz >>= 2; + tx_cmd = ((((unsigned int)(skb->data)) & 0x03) << 16) | (u32) skb->len; + + /* check if there is enough room for the current frame */ + if (wrsz < (DNET_FIFO_SIZE - dnet_readl(bp, TX_FIFO_WCNT))) { + for (i = 0; i < wrsz; i++) + dnet_writel(bp, *bufp++, TX_DATA_FIFO); + + /* + * inform MAC that a packet's written and ready to be + * shipped out + */ + dnet_writel(bp, tx_cmd, TX_LEN_FIFO); + } + + if (dnet_readl(bp, TX_FIFO_WCNT) > DNET_FIFO_TX_DATA_AF_TH) { + netif_stop_queue(dev); + tx_status = dnet_readl(bp, INTR_SRC); + irq_enable = dnet_readl(bp, INTR_ENB); + irq_enable |= DNET_INTR_ENB_TX_FIFOAE; + dnet_writel(bp, irq_enable, INTR_ENB); + } + + /* free the buffer */ + dev_kfree_skb(skb); + + spin_unlock_irqrestore(&bp->lock, flags); + + dev->trans_start = jiffies; + + return 0; +} + +static void dnet_reset_hw(struct dnet *bp) +{ + /* put ts_mac in IDLE state i.e. disable rx/tx */ + dnet_writew_mac(bp, DNET_INTERNAL_MODE_REG, DNET_INTERNAL_MODE_FCEN); + + /* + * RX FIFO almost full threshold: only cmd FIFO almost full is + * implemented for RX side + */ + dnet_writel(bp, DNET_FIFO_RX_CMD_AF_TH, RX_FIFO_TH); + /* + * TX FIFO almost empty threshold: only data FIFO almost empty + * is implemented for TX side + */ + dnet_writel(bp, DNET_FIFO_TX_DATA_AE_TH, TX_FIFO_TH); + + /* flush rx/tx fifos */ + dnet_writel(bp, DNET_SYS_CTL_RXFIFOFLUSH | DNET_SYS_CTL_TXFIFOFLUSH, + SYS_CTL); + msleep(1); + dnet_writel(bp, 0, SYS_CTL); +} + +static void dnet_init_hw(struct dnet *bp) +{ + u32 config; + + dnet_reset_hw(bp); + __dnet_set_hwaddr(bp); + + config = dnet_readw_mac(bp, DNET_INTERNAL_RXTX_CONTROL_REG); + + if (bp->dev->flags & IFF_PROMISC) + /* Copy All Frames */ + config |= DNET_INTERNAL_RXTX_CONTROL_ENPROMISC; + if (!(bp->dev->flags & IFF_BROADCAST)) + /* No BroadCast */ + config |= DNET_INTERNAL_RXTX_CONTROL_RXMULTICAST; + + config |= DNET_INTERNAL_RXTX_CONTROL_RXPAUSE | + DNET_INTERNAL_RXTX_CONTROL_RXBROADCAST | + DNET_INTERNAL_RXTX_CONTROL_DROPCONTROL | + DNET_INTERNAL_RXTX_CONTROL_DISCFXFCS; + + dnet_writew_mac(bp, DNET_INTERNAL_RXTX_CONTROL_REG, config); + + /* clear irq before enabling them */ + config = dnet_readl(bp, INTR_SRC); + + /* enable RX/TX interrupt, recv packet ready interrupt */ + dnet_writel(bp, DNET_INTR_ENB_GLOBAL_ENABLE | DNET_INTR_ENB_RX_SUMMARY | + DNET_INTR_ENB_TX_SUMMARY | DNET_INTR_ENB_RX_FIFOERR | + DNET_INTR_ENB_RX_ERROR | DNET_INTR_ENB_RX_FIFOFULL | + DNET_INTR_ENB_TX_FIFOFULL | DNET_INTR_ENB_TX_DISCFRM | + DNET_INTR_ENB_RX_PKTRDY, INTR_ENB); +} + +static int dnet_open(struct net_device *dev) +{ + struct dnet *bp = netdev_priv(dev); + + /* if the phy is not yet register, retry later */ + if (!bp->phy_dev) + return -EAGAIN; + + if (!is_valid_ether_addr(dev->dev_addr)) + return -EADDRNOTAVAIL; + + napi_enable(&bp->napi); + dnet_init_hw(bp); + + phy_start_aneg(bp->phy_dev); + + /* schedule a link state check */ + phy_start(bp->phy_dev); + + netif_start_queue(dev); + + return 0; +} + +static int dnet_close(struct net_device *dev) +{ + struct dnet *bp = netdev_priv(dev); + + netif_stop_queue(dev); + napi_disable(&bp->napi); + + if (bp->phy_dev) + phy_stop(bp->phy_dev); + + dnet_reset_hw(bp); + netif_carrier_off(dev); + + return 0; +} + +static inline void dnet_print_pretty_hwstats(struct dnet_stats *hwstat) +{ + pr_debug("%s\n", __func__); + pr_debug("----------------------------- RX statistics " + "-------------------------------\n"); + pr_debug("RX_PKT_IGNR_CNT %-8x\n", hwstat->rx_pkt_ignr); + pr_debug("RX_LEN_CHK_ERR_CNT %-8x\n", hwstat->rx_len_chk_err); + pr_debug("RX_LNG_FRM_CNT %-8x\n", hwstat->rx_lng_frm); + pr_debug("RX_SHRT_FRM_CNT %-8x\n", hwstat->rx_shrt_frm); + pr_debug("RX_IPG_VIOL_CNT %-8x\n", hwstat->rx_ipg_viol); + pr_debug("RX_CRC_ERR_CNT %-8x\n", hwstat->rx_crc_err); + pr_debug("RX_OK_PKT_CNT %-8x\n", hwstat->rx_ok_pkt); + pr_debug("RX_CTL_FRM_CNT %-8x\n", hwstat->rx_ctl_frm); + pr_debug("RX_PAUSE_FRM_CNT %-8x\n", hwstat->rx_pause_frm); + pr_debug("RX_MULTICAST_CNT %-8x\n", hwstat->rx_multicast); + pr_debug("RX_BROADCAST_CNT %-8x\n", hwstat->rx_broadcast); + pr_debug("RX_VLAN_TAG_CNT %-8x\n", hwstat->rx_vlan_tag); + pr_debug("RX_PRE_SHRINK_CNT %-8x\n", hwstat->rx_pre_shrink); + pr_debug("RX_DRIB_NIB_CNT %-8x\n", hwstat->rx_drib_nib); + pr_debug("RX_UNSUP_OPCD_CNT %-8x\n", hwstat->rx_unsup_opcd); + pr_debug("RX_BYTE_CNT %-8x\n", hwstat->rx_byte); + pr_debug("----------------------------- TX statistics " + "-------------------------------\n"); + pr_debug("TX_UNICAST_CNT %-8x\n", hwstat->tx_unicast); + pr_debug("TX_PAUSE_FRM_CNT %-8x\n", hwstat->tx_pause_frm); + pr_debug("TX_MULTICAST_CNT %-8x\n", hwstat->tx_multicast); + pr_debug("TX_BRDCAST_CNT %-8x\n", hwstat->tx_brdcast); + pr_debug("TX_VLAN_TAG_CNT %-8x\n", hwstat->tx_vlan_tag); + pr_debug("TX_BAD_FCS_CNT %-8x\n", hwstat->tx_bad_fcs); + pr_debug("TX_JUMBO_CNT %-8x\n", hwstat->tx_jumbo); + pr_debug("TX_BYTE_CNT %-8x\n", hwstat->tx_byte); +} + +static struct net_device_stats *dnet_get_stats(struct net_device *dev) +{ + + struct dnet *bp = netdev_priv(dev); + struct net_device_stats *nstat = &dev->stats; + struct dnet_stats *hwstat = &bp->hw_stats; + + /* read stats from hardware */ + dnet_update_stats(bp); + + /* Convert HW stats into netdevice stats */ + nstat->rx_errors = (hwstat->rx_len_chk_err + + hwstat->rx_lng_frm + hwstat->rx_shrt_frm + + /* ignore IGP violation error + hwstat->rx_ipg_viol + */ + hwstat->rx_crc_err + + hwstat->rx_pre_shrink + + hwstat->rx_drib_nib + hwstat->rx_unsup_opcd); + nstat->tx_errors = hwstat->tx_bad_fcs; + nstat->rx_length_errors = (hwstat->rx_len_chk_err + + hwstat->rx_lng_frm + + hwstat->rx_shrt_frm + hwstat->rx_pre_shrink); + nstat->rx_crc_errors = hwstat->rx_crc_err; + nstat->rx_frame_errors = hwstat->rx_pre_shrink + hwstat->rx_drib_nib; + nstat->rx_packets = hwstat->rx_ok_pkt; + nstat->tx_packets = (hwstat->tx_unicast + + hwstat->tx_multicast + hwstat->tx_brdcast); + nstat->rx_bytes = hwstat->rx_byte; + nstat->tx_bytes = hwstat->tx_byte; + nstat->multicast = hwstat->rx_multicast; + nstat->rx_missed_errors = hwstat->rx_pkt_ignr; + + dnet_print_pretty_hwstats(hwstat); + + return nstat; +} + +static int dnet_get_settings(struct net_device *dev, struct ethtool_cmd *cmd) +{ + struct dnet *bp = netdev_priv(dev); + struct phy_device *phydev = bp->phy_dev; + + if (!phydev) + return -ENODEV; + + return phy_ethtool_gset(phydev, cmd); +} + +static int dnet_set_settings(struct net_device *dev, struct ethtool_cmd *cmd) +{ + struct dnet *bp = netdev_priv(dev); + struct phy_device *phydev = bp->phy_dev; + + if (!phydev) + return -ENODEV; + + return phy_ethtool_sset(phydev, cmd); +} + +static int dnet_ioctl(struct net_device *dev, struct ifreq *rq, int cmd) +{ + struct dnet *bp = netdev_priv(dev); + struct phy_device *phydev = bp->phy_dev; + + if (!netif_running(dev)) + return -EINVAL; + + if (!phydev) + return -ENODEV; + + return phy_mii_ioctl(phydev, if_mii(rq), cmd); +} + +static void dnet_get_drvinfo(struct net_device *dev, + struct ethtool_drvinfo *info) +{ + strcpy(info->driver, DRV_NAME); + strcpy(info->version, DRV_VERSION); + strcpy(info->bus_info, "0"); +} + +static const struct ethtool_ops dnet_ethtool_ops = { + .get_settings = dnet_get_settings, + .set_settings = dnet_set_settings, + .get_drvinfo = dnet_get_drvinfo, + .get_link = ethtool_op_get_link, +}; + +static const struct net_device_ops dnet_netdev_ops = { + .ndo_open = dnet_open, + .ndo_stop = dnet_close, + .ndo_get_stats = dnet_get_stats, + .ndo_start_xmit = dnet_start_xmit, + .ndo_do_ioctl = dnet_ioctl, + .ndo_set_mac_address = eth_mac_addr, + .ndo_validate_addr = eth_validate_addr, + .ndo_change_mtu = eth_change_mtu, +}; + +static int __devinit dnet_probe(struct platform_device *pdev) +{ + struct resource *res; + struct net_device *dev; + struct dnet *bp; + struct phy_device *phydev; + int err = -ENXIO; + unsigned int mem_base, mem_size, irq; + + res = platform_get_resource(pdev, IORESOURCE_MEM, 0); + if (!res) { + dev_err(&pdev->dev, "no mmio resource defined\n"); + goto err_out; + } + mem_base = res->start; + mem_size = resource_size(res); + irq = platform_get_irq(pdev, 0); + + if (!request_mem_region(mem_base, mem_size, DRV_NAME)) { + dev_err(&pdev->dev, "no memory region available\n"); + err = -EBUSY; + goto err_out; + } + + err = -ENOMEM; + dev = alloc_etherdev(sizeof(*bp)); + if (!dev) { + dev_err(&pdev->dev, "etherdev alloc failed, aborting.\n"); + goto err_out; + } + + /* TODO: Actually, we have some interesting features... */ + dev->features |= 0; + + bp = netdev_priv(dev); + bp->dev = dev; + + SET_NETDEV_DEV(dev, &pdev->dev); + + spin_lock_init(&bp->lock); + + bp->regs = ioremap(mem_base, mem_size); + if (!bp->regs) { + dev_err(&pdev->dev, "failed to map registers, aborting.\n"); + err = -ENOMEM; + goto err_out_free_dev; + } + + dev->irq = irq; + err = request_irq(dev->irq, dnet_interrupt, 0, DRV_NAME, dev); + if (err) { + dev_err(&pdev->dev, "Unable to request IRQ %d (error %d)\n", + irq, err); + goto err_out_iounmap; + } + + dev->netdev_ops = &dnet_netdev_ops; + netif_napi_add(dev, &bp->napi, dnet_poll, 64); + dev->ethtool_ops = &dnet_ethtool_ops; + + dev->base_addr = (unsigned long)bp->regs; + + bp->capabilities = dnet_readl(bp, VERCAPS) & DNET_CAPS_MASK; + + dnet_get_hwaddr(bp); + + if (!is_valid_ether_addr(dev->dev_addr)) { + /* choose a random ethernet address */ + random_ether_addr(dev->dev_addr); + __dnet_set_hwaddr(bp); + } + + err = register_netdev(dev); + if (err) { + dev_err(&pdev->dev, "Cannot register net device, aborting.\n"); + goto err_out_free_irq; + } + + /* register the PHY board fixup (for Marvell 88E1111) */ + err = phy_register_fixup_for_uid(0x01410cc0, 0xfffffff0, + dnet_phy_marvell_fixup); + /* we can live without it, so just issue a warning */ + if (err) + dev_warn(&pdev->dev, "Cannot register PHY board fixup.\n"); + + if (dnet_mii_init(bp) != 0) + goto err_out_unregister_netdev; + + dev_info(&pdev->dev, "Dave DNET at 0x%p (0x%08x) irq %d %pM\n", + bp->regs, mem_base, dev->irq, dev->dev_addr); + dev_info(&pdev->dev, "has %smdio, %sirq, %sgigabit, %sdma \n", + (bp->capabilities & DNET_HAS_MDIO) ? "" : "no ", + (bp->capabilities & DNET_HAS_IRQ) ? "" : "no ", + (bp->capabilities & DNET_HAS_GIGABIT) ? "" : "no ", + (bp->capabilities & DNET_HAS_DMA) ? "" : "no "); + phydev = bp->phy_dev; + dev_info(&pdev->dev, "attached PHY driver [%s] " + "(mii_bus:phy_addr=%s, irq=%d)\n", + phydev->drv->name, phydev->dev.bus_id, phydev->irq); + + return 0; + +err_out_unregister_netdev: + unregister_netdev(dev); +err_out_free_irq: + free_irq(dev->irq, dev); +err_out_iounmap: + iounmap(bp->regs); +err_out_free_dev: + free_netdev(dev); +err_out: + return err; +} + +static int __devexit dnet_remove(struct platform_device *pdev) +{ + + struct net_device *dev; + struct dnet *bp; + + dev = platform_get_drvdata(pdev); + + if (dev) { + bp = netdev_priv(dev); + if (bp->phy_dev) + phy_disconnect(bp->phy_dev); + mdiobus_unregister(bp->mii_bus); + kfree(bp->mii_bus->irq); + mdiobus_free(bp->mii_bus); + unregister_netdev(dev); + free_irq(dev->irq, dev); + iounmap(bp->regs); + free_netdev(dev); + } + + return 0; +} + +static struct platform_driver dnet_driver = { + .probe = dnet_probe, + .remove = __devexit_p(dnet_remove), + .driver = { + .name = "dnet", + }, +}; + +static int __init dnet_init(void) +{ + return platform_driver_register(&dnet_driver); +} + +static void __exit dnet_exit(void) +{ + platform_driver_unregister(&dnet_driver); +} + +module_init(dnet_init); +module_exit(dnet_exit); + +MODULE_LICENSE("GPL"); +MODULE_DESCRIPTION("Dave DNET Ethernet driver"); +MODULE_AUTHOR("Ilya Yanok , " + "Matteo Vit "); -- cgit v1.2.3-70-g09d2 From 2c5849ea38fdad477d72dcf1c8c4842db4b33aae Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Wed, 11 Mar 2009 23:28:57 -0700 Subject: dnet: Fix warnings on 64-bit. Signed-off-by: David S. Miller --- drivers/net/dnet.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) (limited to 'drivers/net/dnet.c') diff --git a/drivers/net/dnet.c b/drivers/net/dnet.c index 92c3bd3a472..4b96974fe76 100644 --- a/drivers/net/dnet.c +++ b/drivers/net/dnet.c @@ -553,8 +553,8 @@ static int dnet_start_xmit(struct sk_buff *skb, struct net_device *dev) tx_status = dnet_readl(bp, TX_STATUS); - pr_debug("start_xmit: len %u head %p data %p tail %p end %p\n", - skb->len, skb->head, skb->data, skb->tail, skb->end); + pr_debug("start_xmit: len %u head %p data %p\n", + skb->len, skb->head, skb->data); dnet_print_skb(skb); /* frame size (words) */ @@ -564,11 +564,11 @@ static int dnet_start_xmit(struct sk_buff *skb, struct net_device *dev) tx_status = dnet_readl(bp, TX_STATUS); - bufp = (unsigned int *)(((u32) skb->data) & 0xFFFFFFFC); + bufp = (unsigned int *)(((unsigned long) skb->data) & ~0x3UL); wrsz = (u32) skb->len + 3; - wrsz += ((u32) skb->data) & 0x3; + wrsz += ((unsigned long) skb->data) & 0x3; wrsz >>= 2; - tx_cmd = ((((unsigned int)(skb->data)) & 0x03) << 16) | (u32) skb->len; + tx_cmd = ((((unsigned long)(skb->data)) & 0x03) << 16) | (u32) skb->len; /* check if there is enough room for the current frame */ if (wrsz < (DNET_FIFO_SIZE - dnet_readl(bp, TX_FIFO_WCNT))) { -- cgit v1.2.3-70-g09d2 From 9fae6c3f648e38f023b99b5f5a5280907b2e796e Mon Sep 17 00:00:00 2001 From: Ilya Yanok Date: Fri, 13 Mar 2009 09:51:46 -0700 Subject: dnet: replace obsolete *netif_rx_* functions with *napi_* *netif_rx_* functions is obsolete and removed in newer kernels so we need to use corresponding *napi_* functions instead. Signed-off-by: Ilya Yanok Signed-off-by: David S. Miller --- drivers/net/dnet.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'drivers/net/dnet.c') diff --git a/drivers/net/dnet.c b/drivers/net/dnet.c index 4b96974fe76..5c347f70cb6 100644 --- a/drivers/net/dnet.c +++ b/drivers/net/dnet.c @@ -408,7 +408,7 @@ static int dnet_poll(struct napi_struct *napi, int budget) * packets waiting */ if (!(dnet_readl(bp, RX_FIFO_WCNT) >> 16)) { - netif_rx_complete(napi); + napi_complete(napi); int_enable = dnet_readl(bp, INTR_ENB); int_enable |= DNET_INTR_SRC_RX_CMDFIFOAF; dnet_writel(bp, int_enable, INTR_ENB); @@ -447,7 +447,7 @@ static int dnet_poll(struct napi_struct *napi, int budget) if (npackets < budget) { /* We processed all packets available. Tell NAPI it can * stop polling then re-enable rx interrupts */ - netif_rx_complete(napi); + napi_complete(napi); int_enable = dnet_readl(bp, INTR_ENB); int_enable |= DNET_INTR_SRC_RX_CMDFIFOAF; dnet_writel(bp, int_enable, INTR_ENB); @@ -507,7 +507,7 @@ static irqreturn_t dnet_interrupt(int irq, void *dev_id) } if (int_current & DNET_INTR_SRC_RX_CMDFIFOAF) { - if (netif_rx_schedule_prep(&bp->napi)) { + if (napi_schedule_prep(&bp->napi)) { /* * There's no point taking any more interrupts * until we have processed the buffers @@ -516,7 +516,7 @@ static irqreturn_t dnet_interrupt(int irq, void *dev_id) int_enable = dnet_readl(bp, INTR_ENB); int_enable &= ~DNET_INTR_SRC_RX_CMDFIFOAF; dnet_writel(bp, int_enable, INTR_ENB); - __netif_rx_schedule(&bp->napi); + __napi_schedule(&bp->napi); } handled = 1; } -- cgit v1.2.3-70-g09d2 From 6580f57d485f70851218813fa053d971915f61fb Mon Sep 17 00:00:00 2001 From: Stephen Rothwell Date: Sun, 22 Mar 2009 21:22:48 -0700 Subject: net: update dnet.c for bus_id removal Signed-off-by: Stephen Rothwell Signed-off-by: David S. Miller --- drivers/net/dnet.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'drivers/net/dnet.c') diff --git a/drivers/net/dnet.c b/drivers/net/dnet.c index 5c347f70cb6..1b4063222a8 100644 --- a/drivers/net/dnet.c +++ b/drivers/net/dnet.c @@ -280,11 +280,11 @@ static int dnet_mii_probe(struct net_device *dev) /* attach the mac to the phy */ if (bp->capabilities & DNET_HAS_RMII) { - phydev = phy_connect(dev, phydev->dev.bus_id, + phydev = phy_connect(dev, dev_name(&phydev->dev), &dnet_handle_link_change, 0, PHY_INTERFACE_MODE_RMII); } else { - phydev = phy_connect(dev, phydev->dev.bus_id, + phydev = phy_connect(dev, dev_name(&phydev->dev), &dnet_handle_link_change, 0, PHY_INTERFACE_MODE_MII); } @@ -927,7 +927,7 @@ static int __devinit dnet_probe(struct platform_device *pdev) phydev = bp->phy_dev; dev_info(&pdev->dev, "attached PHY driver [%s] " "(mii_bus:phy_addr=%s, irq=%d)\n", - phydev->drv->name, phydev->dev.bus_id, phydev->irq); + phydev->drv->name, dev_name(&phydev->dev), phydev->irq); return 0; -- cgit v1.2.3-70-g09d2