summaryrefslogtreecommitdiffstats
path: root/drivers/net/mv643xx_eth.c
diff options
context:
space:
mode:
Diffstat (limited to 'drivers/net/mv643xx_eth.c')
-rw-r--r--drivers/net/mv643xx_eth.c1558
1 files changed, 573 insertions, 985 deletions
diff --git a/drivers/net/mv643xx_eth.c b/drivers/net/mv643xx_eth.c
index c0998ef938e0..9f2661355a4a 100644
--- a/drivers/net/mv643xx_eth.c
+++ b/drivers/net/mv643xx_eth.c
@@ -10,7 +10,7 @@
*
* Copyright (C) 2003 Ralf Baechle <ralf@linux-mips.org>
*
- * Copyright (C) 2004-2005 MontaVista Software, Inc.
+ * Copyright (C) 2004-2006 MontaVista Software, Inc.
* Dale Farnsworth <dale@farnsworth.org>
*
* Copyright (C) 2004 Steven J. Hill <sjhill1@rockwellcollins.com>
@@ -37,8 +37,6 @@
#include <linux/tcp.h>
#include <linux/udp.h>
#include <linux/etherdevice.h>
-#include <linux/in.h>
-#include <linux/ip.h>
#include <linux/bitops.h>
#include <linux/delay.h>
@@ -52,39 +50,16 @@
#include <asm/delay.h>
#include "mv643xx_eth.h"
-/*
- * The first part is the high level driver of the gigE ethernet ports.
- */
-
-/* Constants */
-#define VLAN_HLEN 4
-#define FCS_LEN 4
-#define DMA_ALIGN 8 /* hw requires 8-byte alignment */
-#define HW_IP_ALIGN 2 /* hw aligns IP header */
-#define WRAP HW_IP_ALIGN + ETH_HLEN + VLAN_HLEN + FCS_LEN
-#define RX_SKB_SIZE ((dev->mtu + WRAP + 7) & ~0x7)
-
-#define INT_UNMASK_ALL 0x0007ffff
-#define INT_UNMASK_ALL_EXT 0x0011ffff
-#define INT_MASK_ALL 0x00000000
-#define INT_MASK_ALL_EXT 0x00000000
-#define INT_CAUSE_CHECK_BITS INT_CAUSE_UNMASK_ALL
-#define INT_CAUSE_CHECK_BITS_EXT INT_CAUSE_UNMASK_ALL_EXT
-
-#ifdef MV643XX_CHECKSUM_OFFLOAD_TX
-#define MAX_DESCS_PER_SKB (MAX_SKB_FRAGS + 1)
-#else
-#define MAX_DESCS_PER_SKB 1
-#endif
-
-#define PHY_WAIT_ITERATIONS 1000 /* 1000 iterations * 10uS = 10mS max */
-#define PHY_WAIT_MICRO_SECONDS 10
-
/* Static function declarations */
-static int eth_port_link_is_up(unsigned int eth_port_num);
static void eth_port_uc_addr_get(struct net_device *dev,
unsigned char *MacAddr);
static void eth_port_set_multicast_list(struct net_device *);
+static void mv643xx_eth_port_enable_tx(unsigned int port_num,
+ unsigned int queues);
+static void mv643xx_eth_port_enable_rx(unsigned int port_num,
+ unsigned int queues);
+static unsigned int mv643xx_eth_port_disable_tx(unsigned int port_num);
+static unsigned int mv643xx_eth_port_disable_rx(unsigned int port_num);
static int mv643xx_eth_open(struct net_device *);
static int mv643xx_eth_stop(struct net_device *);
static int mv643xx_eth_change_mtu(struct net_device *, int);
@@ -93,8 +68,12 @@ static void eth_port_init_mac_tables(unsigned int eth_port_num);
#ifdef MV643XX_NAPI
static int mv643xx_poll(struct net_device *dev, int *budget);
#endif
+static int ethernet_phy_get(unsigned int eth_port_num);
static void ethernet_phy_set(unsigned int eth_port_num, int phy_addr);
static int ethernet_phy_detect(unsigned int eth_port_num);
+static int mv643xx_mdio_read(struct net_device *dev, int phy_id, int location);
+static void mv643xx_mdio_write(struct net_device *dev, int phy_id, int location, int val);
+static int mv643xx_eth_do_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd);
static struct ethtool_ops mv643xx_ethtool_ops;
static char mv643xx_driver_name[] = "mv643xx_eth";
@@ -153,67 +132,53 @@ static int mv643xx_eth_change_mtu(struct net_device *dev, int new_mtu)
}
/*
- * mv643xx_eth_rx_task
+ * mv643xx_eth_rx_refill_descs
*
* Fills / refills RX queue on a certain gigabit ethernet port
*
* Input : pointer to ethernet interface network device structure
* Output : N/A
*/
-static void mv643xx_eth_rx_task(void *data)
+static void mv643xx_eth_rx_refill_descs(struct net_device *dev)
{
- struct net_device *dev = (struct net_device *)data;
struct mv643xx_private *mp = netdev_priv(dev);
struct pkt_info pkt_info;
struct sk_buff *skb;
int unaligned;
- if (test_and_set_bit(0, &mp->rx_task_busy))
- panic("%s: Error in test_set_bit / clear_bit", dev->name);
-
- while (mp->rx_ring_skbs < (mp->rx_ring_size - 5)) {
- skb = dev_alloc_skb(RX_SKB_SIZE + DMA_ALIGN);
+ while (mp->rx_desc_count < mp->rx_ring_size) {
+ skb = dev_alloc_skb(ETH_RX_SKB_SIZE + ETH_DMA_ALIGN);
if (!skb)
break;
- mp->rx_ring_skbs++;
- unaligned = (u32)skb->data & (DMA_ALIGN - 1);
+ mp->rx_desc_count++;
+ unaligned = (u32)skb->data & (ETH_DMA_ALIGN - 1);
if (unaligned)
- skb_reserve(skb, DMA_ALIGN - unaligned);
+ skb_reserve(skb, ETH_DMA_ALIGN - unaligned);
pkt_info.cmd_sts = ETH_RX_ENABLE_INTERRUPT;
- pkt_info.byte_cnt = RX_SKB_SIZE;
- pkt_info.buf_ptr = dma_map_single(NULL, skb->data, RX_SKB_SIZE,
- DMA_FROM_DEVICE);
+ pkt_info.byte_cnt = ETH_RX_SKB_SIZE;
+ pkt_info.buf_ptr = dma_map_single(NULL, skb->data,
+ ETH_RX_SKB_SIZE, DMA_FROM_DEVICE);
pkt_info.return_info = skb;
if (eth_rx_return_buff(mp, &pkt_info) != ETH_OK) {
printk(KERN_ERR
"%s: Error allocating RX Ring\n", dev->name);
break;
}
- skb_reserve(skb, HW_IP_ALIGN);
+ skb_reserve(skb, ETH_HW_IP_ALIGN);
}
- clear_bit(0, &mp->rx_task_busy);
/*
* If RX ring is empty of SKB, set a timer to try allocating
- * again in a later time .
+ * again at a later time.
*/
- if ((mp->rx_ring_skbs == 0) && (mp->rx_timer_flag == 0)) {
+ if (mp->rx_desc_count == 0) {
printk(KERN_INFO "%s: Rx ring is empty\n", dev->name);
- /* After 100mSec */
- mp->timeout.expires = jiffies + (HZ / 10);
+ mp->timeout.expires = jiffies + (HZ / 10); /* 100 mSec */
add_timer(&mp->timeout);
- mp->rx_timer_flag = 1;
- }
-#ifdef MV643XX_RX_QUEUE_FILL_ON_TASK
- else {
- /* Return interrupts */
- mv_write(MV643XX_ETH_INTERRUPT_MASK_REG(mp->port_num),
- INT_UNMASK_ALL);
}
-#endif
}
/*
- * mv643xx_eth_rx_task_timer_wrapper
+ * mv643xx_eth_rx_refill_descs_timer_wrapper
*
* Timer routine to wake up RX queue filling task. This function is
* used only in case the RX queue is empty, and all alloc_skb has
@@ -222,13 +187,9 @@ static void mv643xx_eth_rx_task(void *data)
* Input : pointer to ethernet interface network device structure
* Output : N/A
*/
-static void mv643xx_eth_rx_task_timer_wrapper(unsigned long data)
+static inline void mv643xx_eth_rx_refill_descs_timer_wrapper(unsigned long data)
{
- struct net_device *dev = (struct net_device *)data;
- struct mv643xx_private *mp = netdev_priv(dev);
-
- mp->rx_timer_flag = 0;
- mv643xx_eth_rx_task((void *)data);
+ mv643xx_eth_rx_refill_descs((struct net_device *)data);
}
/*
@@ -245,8 +206,7 @@ static void mv643xx_eth_update_mac_address(struct net_device *dev)
unsigned int port_num = mp->port_num;
eth_port_init_mac_tables(port_num);
- memcpy(mp->port_mac_addr, dev->dev_addr, 6);
- eth_port_uc_addr_set(port_num, mp->port_mac_addr);
+ eth_port_uc_addr_set(port_num, dev->dev_addr);
}
/*
@@ -260,13 +220,14 @@ static void mv643xx_eth_update_mac_address(struct net_device *dev)
static void mv643xx_eth_set_rx_mode(struct net_device *dev)
{
struct mv643xx_private *mp = netdev_priv(dev);
+ u32 config_reg;
+ config_reg = mv_read(MV643XX_ETH_PORT_CONFIG_REG(mp->port_num));
if (dev->flags & IFF_PROMISC)
- mp->port_config |= (u32) MV643XX_ETH_UNICAST_PROMISCUOUS_MODE;
+ config_reg |= (u32) MV643XX_ETH_UNICAST_PROMISCUOUS_MODE;
else
- mp->port_config &= ~(u32) MV643XX_ETH_UNICAST_PROMISCUOUS_MODE;
-
- mv_write(MV643XX_ETH_PORT_CONFIG_REG(mp->port_num), mp->port_config);
+ config_reg &= ~(u32) MV643XX_ETH_UNICAST_PROMISCUOUS_MODE;
+ mv_write(MV643XX_ETH_PORT_CONFIG_REG(mp->port_num), config_reg);
eth_port_set_multicast_list(dev);
}
@@ -322,53 +283,82 @@ static void mv643xx_eth_tx_timeout_task(struct net_device *dev)
netif_device_detach(dev);
eth_port_reset(mp->port_num);
- eth_port_start(mp);
+ eth_port_start(dev);
netif_device_attach(dev);
}
-/*
- * mv643xx_eth_free_tx_queue
- *
- * Input : dev - a pointer to the required interface
+/**
+ * mv643xx_eth_free_tx_descs - Free the tx desc data for completed descriptors
*
- * Output : 0 if was able to release skb , nonzero otherwise
+ * If force is non-zero, frees uncompleted descriptors as well
*/
-static int mv643xx_eth_free_tx_queue(struct net_device *dev,
- unsigned int eth_int_cause_ext)
+int mv643xx_eth_free_tx_descs(struct net_device *dev, int force)
{
struct mv643xx_private *mp = netdev_priv(dev);
- struct net_device_stats *stats = &mp->stats;
- struct pkt_info pkt_info;
- int released = 1;
+ struct eth_tx_desc *desc;
+ u32 cmd_sts;
+ struct sk_buff *skb;
+ unsigned long flags;
+ int tx_index;
+ dma_addr_t addr;
+ int count;
+ int released = 0;
+
+ while (mp->tx_desc_count > 0) {
+ spin_lock_irqsave(&mp->lock, flags);
+ tx_index = mp->tx_used_desc_q;
+ desc = &mp->p_tx_desc_area[tx_index];
+ cmd_sts = desc->cmd_sts;
+
+ if (!force && (cmd_sts & ETH_BUFFER_OWNED_BY_DMA)) {
+ spin_unlock_irqrestore(&mp->lock, flags);
+ return released;
+ }
- if (!(eth_int_cause_ext & (BIT0 | BIT8)))
- return released;
+ mp->tx_used_desc_q = (tx_index + 1) % mp->tx_ring_size;
+ mp->tx_desc_count--;
- /* Check only queue 0 */
- while (eth_tx_return_desc(mp, &pkt_info) == ETH_OK) {
- if (pkt_info.cmd_sts & BIT0) {
+ addr = desc->buf_ptr;
+ count = desc->byte_cnt;
+ skb = mp->tx_skb[tx_index];
+ if (skb)
+ mp->tx_skb[tx_index] = NULL;
+
+ spin_unlock_irqrestore(&mp->lock, flags);
+
+ if (cmd_sts & ETH_ERROR_SUMMARY) {
printk("%s: Error in TX\n", dev->name);
- stats->tx_errors++;
+ mp->stats.tx_errors++;
}
- if (pkt_info.cmd_sts & ETH_TX_FIRST_DESC)
- dma_unmap_single(NULL, pkt_info.buf_ptr,
- pkt_info.byte_cnt,
- DMA_TO_DEVICE);
+ if (cmd_sts & ETH_TX_FIRST_DESC)
+ dma_unmap_single(NULL, addr, count, DMA_TO_DEVICE);
else
- dma_unmap_page(NULL, pkt_info.buf_ptr,
- pkt_info.byte_cnt,
- DMA_TO_DEVICE);
+ dma_unmap_page(NULL, addr, count, DMA_TO_DEVICE);
- if (pkt_info.return_info) {
- dev_kfree_skb_irq(pkt_info.return_info);
- released = 0;
- }
+ if (skb)
+ dev_kfree_skb_irq(skb);
+
+ released = 1;
}
return released;
}
+static void mv643xx_eth_free_completed_tx_descs(struct net_device *dev)
+{
+ struct mv643xx_private *mp = netdev_priv(dev);
+
+ if (mv643xx_eth_free_tx_descs(dev, 0) &&
+ mp->tx_ring_size - mp->tx_desc_count >= MAX_DESCS_PER_SKB)
+ netif_wake_queue(dev);
+}
+
+static void mv643xx_eth_free_all_tx_descs(struct net_device *dev)
+{
+ mv643xx_eth_free_tx_descs(dev, 1);
+}
+
/*
* mv643xx_eth_receive
*
@@ -380,11 +370,7 @@ static int mv643xx_eth_free_tx_queue(struct net_device *dev,
*
* Output : number of served packets
*/
-#ifdef MV643XX_NAPI
static int mv643xx_eth_receive_queue(struct net_device *dev, int budget)
-#else
-static int mv643xx_eth_receive_queue(struct net_device *dev)
-#endif
{
struct mv643xx_private *mp = netdev_priv(dev);
struct net_device_stats *stats = &mp->stats;
@@ -392,15 +378,14 @@ static int mv643xx_eth_receive_queue(struct net_device *dev)
struct sk_buff *skb;
struct pkt_info pkt_info;
-#ifdef MV643XX_NAPI
while (budget-- > 0 && eth_port_receive(mp, &pkt_info) == ETH_OK) {
-#else
- while (eth_port_receive(mp, &pkt_info) == ETH_OK) {
-#endif
- mp->rx_ring_skbs--;
+ mp->rx_desc_count--;
received_packets++;
- /* Update statistics. Note byte count includes 4 byte CRC count */
+ /*
+ * Update statistics.
+ * Note byte count includes 4 byte CRC count
+ */
stats->rx_packets++;
stats->rx_bytes += pkt_info.byte_cnt;
skb = pkt_info.return_info;
@@ -448,10 +433,61 @@ static int mv643xx_eth_receive_queue(struct net_device *dev)
}
dev->last_rx = jiffies;
}
+ mv643xx_eth_rx_refill_descs(dev); /* Fill RX ring with skb's */
return received_packets;
}
+/* Set the mv643xx port configuration register for the speed/duplex mode. */
+static void mv643xx_eth_update_pscr(struct net_device *dev,
+ struct ethtool_cmd *ecmd)
+{
+ struct mv643xx_private *mp = netdev_priv(dev);
+ int port_num = mp->port_num;
+ u32 o_pscr, n_pscr;
+ unsigned int queues;
+
+ o_pscr = mv_read(MV643XX_ETH_PORT_SERIAL_CONTROL_REG(port_num));
+ n_pscr = o_pscr;
+
+ /* clear speed, duplex and rx buffer size fields */
+ n_pscr &= ~(MV643XX_ETH_SET_MII_SPEED_TO_100 |
+ MV643XX_ETH_SET_GMII_SPEED_TO_1000 |
+ MV643XX_ETH_SET_FULL_DUPLEX_MODE |
+ MV643XX_ETH_MAX_RX_PACKET_MASK);
+
+ if (ecmd->duplex == DUPLEX_FULL)
+ n_pscr |= MV643XX_ETH_SET_FULL_DUPLEX_MODE;
+
+ if (ecmd->speed == SPEED_1000)
+ n_pscr |= MV643XX_ETH_SET_GMII_SPEED_TO_1000 |
+ MV643XX_ETH_MAX_RX_PACKET_9700BYTE;
+ else {
+ if (ecmd->speed == SPEED_100)
+ n_pscr |= MV643XX_ETH_SET_MII_SPEED_TO_100;
+ n_pscr |= MV643XX_ETH_MAX_RX_PACKET_1522BYTE;
+ }
+
+ if (n_pscr != o_pscr) {
+ if ((o_pscr & MV643XX_ETH_SERIAL_PORT_ENABLE) == 0)
+ mv_write(MV643XX_ETH_PORT_SERIAL_CONTROL_REG(port_num),
+ n_pscr);
+ else {
+ queues = mv643xx_eth_port_disable_tx(port_num);
+
+ o_pscr &= ~MV643XX_ETH_SERIAL_PORT_ENABLE;
+ mv_write(MV643XX_ETH_PORT_SERIAL_CONTROL_REG(port_num),
+ o_pscr);
+ mv_write(MV643XX_ETH_PORT_SERIAL_CONTROL_REG(port_num),
+ n_pscr);
+ mv_write(MV643XX_ETH_PORT_SERIAL_CONTROL_REG(port_num),
+ n_pscr);
+ if (queues)
+ mv643xx_eth_port_enable_tx(port_num, queues);
+ }
+ }
+}
+
/*
* mv643xx_eth_int_handler
*
@@ -473,78 +509,52 @@ static irqreturn_t mv643xx_eth_int_handler(int irq, void *dev_id,
/* Read interrupt cause registers */
eth_int_cause = mv_read(MV643XX_ETH_INTERRUPT_CAUSE_REG(port_num)) &
- INT_UNMASK_ALL;
-
- if (eth_int_cause & BIT1)
+ ETH_INT_UNMASK_ALL;
+ if (eth_int_cause & ETH_INT_CAUSE_EXT) {
eth_int_cause_ext = mv_read(
MV643XX_ETH_INTERRUPT_CAUSE_EXTEND_REG(port_num)) &
- INT_UNMASK_ALL_EXT;
+ ETH_INT_UNMASK_ALL_EXT;
+ mv_write(MV643XX_ETH_INTERRUPT_CAUSE_EXTEND_REG(port_num),
+ ~eth_int_cause_ext);
+ }
-#ifdef MV643XX_NAPI
- if (!(eth_int_cause & 0x0007fffd)) {
- /* Dont ack the Rx interrupt */
-#endif
- /*
- * Clear specific ethernet port intrerrupt registers by
- * acknowleding relevant bits.
- */
- mv_write(MV643XX_ETH_INTERRUPT_CAUSE_REG(port_num),
- ~eth_int_cause);
- if (eth_int_cause_ext != 0x0)
- mv_write(MV643XX_ETH_INTERRUPT_CAUSE_EXTEND_REG
- (port_num), ~eth_int_cause_ext);
-
- /* UDP change : We may need this */
- if ((eth_int_cause_ext & 0x0000ffff) &&
- (mv643xx_eth_free_tx_queue(dev, eth_int_cause_ext) == 0) &&
- (mp->tx_ring_size > mp->tx_ring_skbs + MAX_DESCS_PER_SKB))
- netif_wake_queue(dev);
-#ifdef MV643XX_NAPI
- } else {
- if (netif_rx_schedule_prep(dev)) {
- /* Mask all the interrupts */
- mv_write(MV643XX_ETH_INTERRUPT_MASK_REG(port_num),
- INT_MASK_ALL);
- /* wait for previous write to complete */
- mv_read(MV643XX_ETH_INTERRUPT_MASK_REG(port_num));
- __netif_rx_schedule(dev);
+ /* PHY status changed */
+ if (eth_int_cause_ext & ETH_INT_CAUSE_PHY) {
+ struct ethtool_cmd cmd;
+
+ if (mii_link_ok(&mp->mii)) {
+ mii_ethtool_gset(&mp->mii, &cmd);
+ mv643xx_eth_update_pscr(dev, &cmd);
+ mv643xx_eth_port_enable_tx(port_num,
+ ETH_TX_QUEUES_ENABLED);
+ if (!netif_carrier_ok(dev)) {
+ netif_carrier_on(dev);
+ if (mp->tx_ring_size - mp->tx_desc_count >=
+ MAX_DESCS_PER_SKB)
+ netif_wake_queue(dev);
+ }
+ } else if (netif_carrier_ok(dev)) {
+ netif_stop_queue(dev);
+ netif_carrier_off(dev);
}
-#else
- if (eth_int_cause & (BIT2 | BIT11))
- mv643xx_eth_receive_queue(dev, 0);
+ }
- /*
- * After forwarded received packets to upper layer, add a task
- * in an interrupts enabled context that refills the RX ring
- * with skb's.
- */
-#ifdef MV643XX_RX_QUEUE_FILL_ON_TASK
- /* Mask all interrupts on ethernet port */
+#ifdef MV643XX_NAPI
+ if (eth_int_cause & ETH_INT_CAUSE_RX) {
+ /* schedule the NAPI poll routine to maintain port */
mv_write(MV643XX_ETH_INTERRUPT_MASK_REG(port_num),
- INT_MASK_ALL);
- /* wait for previous write to take effect */
+ ETH_INT_MASK_ALL);
+ /* wait for previous write to complete */
mv_read(MV643XX_ETH_INTERRUPT_MASK_REG(port_num));
- queue_task(&mp->rx_task, &tq_immediate);
- mark_bh(IMMEDIATE_BH);
+ netif_rx_schedule(dev);
+ }
#else
- mp->rx_task.func(dev);
+ if (eth_int_cause & ETH_INT_CAUSE_RX)
+ mv643xx_eth_receive_queue(dev, INT_MAX);
+ if (eth_int_cause_ext & ETH_INT_CAUSE_TX)
+ mv643xx_eth_free_completed_tx_descs(dev);
#endif
-#endif
- }
- /* PHY status changed */
- if (eth_int_cause_ext & (BIT16 | BIT20)) {
- if (eth_port_link_is_up(port_num)) {
- netif_carrier_on(dev);
- netif_wake_queue(dev);
- /* Start TX queue */
- mv_write(MV643XX_ETH_TRANSMIT_QUEUE_COMMAND_REG
- (port_num), 1);
- } else {
- netif_carrier_off(dev);
- netif_stop_queue(dev);
- }
- }
/*
* If no real interrupt occured, exit.
@@ -670,9 +680,6 @@ static void ether_init_rx_desc_ring(struct mv643xx_private *mp)
mp->rx_used_desc_q = 0;
mp->rx_desc_area_size = rx_desc_num * sizeof(struct eth_rx_desc);
-
- /* Add the queue to the list of RX queues of this port */
- mp->port_rx_queue_command |= 1;
}
/*
@@ -712,14 +719,36 @@ static void ether_init_tx_desc_ring(struct mv643xx_private *mp)
mp->tx_curr_desc_q = 0;
mp->tx_used_desc_q = 0;
-#ifdef MV643XX_CHECKSUM_OFFLOAD_TX
- mp->tx_first_desc_q = 0;
-#endif
mp->tx_desc_area_size = tx_desc_num * sizeof(struct eth_tx_desc);
+}
- /* Add the queue to the list of Tx queues of this port */
- mp->port_tx_queue_command |= 1;
+static int mv643xx_set_settings(struct net_device *dev, struct ethtool_cmd *cmd)
+{
+ struct mv643xx_private *mp = netdev_priv(dev);
+ int err;
+
+ spin_lock_irq(&mp->lock);
+ err = mii_ethtool_sset(&mp->mii, cmd);
+ spin_unlock_irq(&mp->lock);
+
+ return err;
+}
+
+static int mv643xx_get_settings(struct net_device *dev, struct ethtool_cmd *cmd)
+{
+ struct mv643xx_private *mp = netdev_priv(dev);
+ int err;
+
+ spin_lock_irq(&mp->lock);
+ err = mii_ethtool_gset(&mp->mii, cmd);
+ spin_unlock_irq(&mp->lock);
+
+ /* The PHY may support 1000baseT_Half, but the mv643xx does not */
+ cmd->supported &= ~SUPPORTED_1000baseT_Half;
+ cmd->advertising &= ~ADVERTISED_1000baseT_Half;
+
+ return err;
}
/*
@@ -750,23 +779,12 @@ static int mv643xx_eth_open(struct net_device *dev)
return -EAGAIN;
}
- /* Stop RX Queues */
- mv_write(MV643XX_ETH_RECEIVE_QUEUE_COMMAND_REG(port_num), 0x0000ff00);
-
- /* Set the MAC Address */
- memcpy(mp->port_mac_addr, dev->dev_addr, 6);
-
eth_port_init(mp);
- INIT_WORK(&mp->rx_task, (void (*)(void *))mv643xx_eth_rx_task, dev);
-
memset(&mp->timeout, 0, sizeof(struct timer_list));
- mp->timeout.function = mv643xx_eth_rx_task_timer_wrapper;
+ mp->timeout.function = mv643xx_eth_rx_refill_descs_timer_wrapper;
mp->timeout.data = (unsigned long)dev;
- mp->rx_task_busy = 0;
- mp->rx_timer_flag = 0;
-
/* Allocate RX and TX skb rings */
mp->rx_skb = kmalloc(sizeof(*mp->rx_skb) * mp->rx_ring_size,
GFP_KERNEL);
@@ -784,7 +802,7 @@ static int mv643xx_eth_open(struct net_device *dev)
}
/* Allocate TX ring */
- mp->tx_ring_skbs = 0;
+ mp->tx_desc_count = 0;
size = mp->tx_ring_size * sizeof(struct eth_tx_desc);
mp->tx_desc_area_size = size;
@@ -809,7 +827,7 @@ static int mv643xx_eth_open(struct net_device *dev)
ether_init_tx_desc_ring(mp);
/* Allocate RX ring */
- mp->rx_ring_skbs = 0;
+ mp->rx_desc_count = 0;
size = mp->rx_ring_size * sizeof(struct eth_rx_desc);
mp->rx_desc_area_size = size;
@@ -839,9 +857,13 @@ static int mv643xx_eth_open(struct net_device *dev)
ether_init_rx_desc_ring(mp);
- mv643xx_eth_rx_task(dev); /* Fill RX ring with skb's */
+ mv643xx_eth_rx_refill_descs(dev); /* Fill RX ring with skb's */
+
+ /* Clear any pending ethernet port interrupts */
+ mv_write(MV643XX_ETH_INTERRUPT_CAUSE_REG(port_num), 0);
+ mv_write(MV643XX_ETH_INTERRUPT_CAUSE_EXTEND_REG(port_num), 0);
- eth_port_start(mp);
+ eth_port_start(dev);
/* Interrupt Coalescing */
@@ -853,16 +875,13 @@ static int mv643xx_eth_open(struct net_device *dev)
mp->tx_int_coal =
eth_port_set_tx_coal(port_num, 133000000, MV643XX_TX_COAL);
- /* Clear any pending ethernet port interrupts */
- mv_write(MV643XX_ETH_INTERRUPT_CAUSE_REG(port_num), 0);
- mv_write(MV643XX_ETH_INTERRUPT_CAUSE_EXTEND_REG(port_num), 0);
-
/* Unmask phy and link status changes interrupts */
mv_write(MV643XX_ETH_INTERRUPT_EXTEND_MASK_REG(port_num),
- INT_UNMASK_ALL_EXT);
+ ETH_INT_UNMASK_ALL_EXT);
/* Unmask RX buffer and TX end interrupt */
- mv_write(MV643XX_ETH_INTERRUPT_MASK_REG(port_num), INT_UNMASK_ALL);
+ mv_write(MV643XX_ETH_INTERRUPT_MASK_REG(port_num), ETH_INT_UNMASK_ALL);
+
return 0;
out_free_tx_skb:
@@ -878,25 +897,14 @@ out_free_irq:
static void mv643xx_eth_free_tx_rings(struct net_device *dev)
{
struct mv643xx_private *mp = netdev_priv(dev);
- unsigned int port_num = mp->port_num;
- unsigned int curr;
- struct sk_buff *skb;
/* Stop Tx Queues */
- mv_write(MV643XX_ETH_TRANSMIT_QUEUE_COMMAND_REG(port_num), 0x0000ff00);
-
- /* Free outstanding skb's on TX rings */
- for (curr = 0; mp->tx_ring_skbs && curr < mp->tx_ring_size; curr++) {
- skb = mp->tx_skb[curr];
- if (skb) {
- mp->tx_ring_skbs -= skb_shinfo(skb)->nr_frags;
- dev_kfree_skb(skb);
- mp->tx_ring_skbs--;
- }
- }
- if (mp->tx_ring_skbs)
- printk("%s: Error on Tx descriptor free - could not free %d"
- " descriptors\n", dev->name, mp->tx_ring_skbs);
+ mv643xx_eth_port_disable_tx(mp->port_num);
+
+ /* Free outstanding skb's on TX ring */
+ mv643xx_eth_free_all_tx_descs(dev);
+
+ BUG_ON(mp->tx_used_desc_q != mp->tx_curr_desc_q);
/* Free TX ring */
if (mp->tx_sram_size)
@@ -913,21 +921,21 @@ static void mv643xx_eth_free_rx_rings(struct net_device *dev)
int curr;
/* Stop RX Queues */
- mv_write(MV643XX_ETH_RECEIVE_QUEUE_COMMAND_REG(port_num), 0x0000ff00);
+ mv643xx_eth_port_disable_rx(port_num);
/* Free preallocated skb's on RX rings */
- for (curr = 0; mp->rx_ring_skbs && curr < mp->rx_ring_size; curr++) {
+ for (curr = 0; mp->rx_desc_count && curr < mp->rx_ring_size; curr++) {
if (mp->rx_skb[curr]) {
dev_kfree_skb(mp->rx_skb[curr]);
- mp->rx_ring_skbs--;
+ mp->rx_desc_count--;
}
}
- if (mp->rx_ring_skbs)
+ if (mp->rx_desc_count)
printk(KERN_ERR
"%s: Error in freeing Rx Ring. %d skb's still"
" stuck in RX Ring - ignoring them\n", dev->name,
- mp->rx_ring_skbs);
+ mp->rx_desc_count);
/* Free RX ring */
if (mp->rx_sram_size)
iounmap(mp->p_rx_desc_area);
@@ -952,7 +960,7 @@ static int mv643xx_eth_stop(struct net_device *dev)
unsigned int port_num = mp->port_num;
/* Mask all interrupts on ethernet port */
- mv_write(MV643XX_ETH_INTERRUPT_MASK_REG(port_num), INT_MASK_ALL);
+ mv_write(MV643XX_ETH_INTERRUPT_MASK_REG(port_num), ETH_INT_MASK_ALL);
/* wait for previous write to complete */
mv_read(MV643XX_ETH_INTERRUPT_MASK_REG(port_num));
@@ -977,30 +985,6 @@ static int mv643xx_eth_stop(struct net_device *dev)
}
#ifdef MV643XX_NAPI
-static void mv643xx_tx(struct net_device *dev)
-{
- struct mv643xx_private *mp = netdev_priv(dev);
- struct pkt_info pkt_info;
-
- while (eth_tx_return_desc(mp, &pkt_info) == ETH_OK) {
- if (pkt_info.cmd_sts & ETH_TX_FIRST_DESC)
- dma_unmap_single(NULL, pkt_info.buf_ptr,
- pkt_info.byte_cnt,
- DMA_TO_DEVICE);
- else
- dma_unmap_page(NULL, pkt_info.buf_ptr,
- pkt_info.byte_cnt,
- DMA_TO_DEVICE);
-
- if (pkt_info.return_info)
- dev_kfree_skb_irq(pkt_info.return_info);
- }
-
- if (netif_queue_stopped(dev) &&
- mp->tx_ring_size > mp->tx_ring_skbs + MAX_DESCS_PER_SKB)
- netif_wake_queue(dev);
-}
-
/*
* mv643xx_poll
*
@@ -1014,7 +998,7 @@ static int mv643xx_poll(struct net_device *dev, int *budget)
#ifdef MV643XX_TX_FAST_REFILL
if (++mp->tx_clean_threshold > 5) {
- mv643xx_tx(dev);
+ mv643xx_eth_free_completed_tx_descs(dev);
mp->tx_clean_threshold = 0;
}
#endif
@@ -1025,7 +1009,6 @@ static int mv643xx_poll(struct net_device *dev, int *budget)
if (orig_budget > dev->quota)
orig_budget = dev->quota;
work_done = mv643xx_eth_receive_queue(dev, orig_budget);
- mp->rx_task.func(dev);
*budget -= work_done;
dev->quota -= work_done;
if (work_done >= orig_budget)
@@ -1037,14 +1020,17 @@ static int mv643xx_poll(struct net_device *dev, int *budget)
mv_write(MV643XX_ETH_INTERRUPT_CAUSE_REG(port_num), 0);
mv_write(MV643XX_ETH_INTERRUPT_CAUSE_EXTEND_REG(port_num), 0);
mv_write(MV643XX_ETH_INTERRUPT_MASK_REG(port_num),
- INT_UNMASK_ALL);
+ ETH_INT_UNMASK_ALL);
}
return done ? 0 : 1;
}
#endif
-/* Hardware can't handle unaligned fragments smaller than 9 bytes.
+/**
+ * has_tiny_unaligned_frags - check if skb has any small, unaligned fragments
+ *
+ * Hardware can't handle unaligned fragments smaller than 9 bytes.
* This helper function detects that case.
*/
@@ -1061,223 +1047,166 @@ static inline unsigned int has_tiny_unaligned_frags(struct sk_buff *skb)
return 0;
}
+/**
+ * eth_alloc_tx_desc_index - return the index of the next available tx desc
+ */
+static int eth_alloc_tx_desc_index(struct mv643xx_private *mp)
+{
+ int tx_desc_curr;
-/*
- * mv643xx_eth_start_xmit
- *
- * This function is queues a packet in the Tx descriptor for
- * required port.
- *
- * Input : skb - a pointer to socket buffer
- * dev - a pointer to the required port
+ BUG_ON(mp->tx_desc_count >= mp->tx_ring_size);
+
+ tx_desc_curr = mp->tx_curr_desc_q;
+ mp->tx_curr_desc_q = (tx_desc_curr + 1) % mp->tx_ring_size;
+
+ BUG_ON(mp->tx_curr_desc_q == mp->tx_used_desc_q);
+
+ return tx_desc_curr;
+}
+
+/**
+ * eth_tx_fill_frag_descs - fill tx hw descriptors for an skb's fragments.
*
- * Output : zero upon success
+ * Ensure the data for each fragment to be transmitted is mapped properly,
+ * then fill in descriptors in the tx hw queue.
*/
-static int mv643xx_eth_start_xmit(struct sk_buff *skb, struct net_device *dev)
+static void eth_tx_fill_frag_descs(struct mv643xx_private *mp,
+ struct sk_buff *skb)
{
- struct mv643xx_private *mp = netdev_priv(dev);
- struct net_device_stats *stats = &mp->stats;
- ETH_FUNC_RET_STATUS status;
- unsigned long flags;
- struct pkt_info pkt_info;
+ int frag;
+ int tx_index;
+ struct eth_tx_desc *desc;
- if (netif_queue_stopped(dev)) {
- printk(KERN_ERR
- "%s: Tried sending packet when interface is stopped\n",
- dev->name);
- return 1;
+ for (frag = 0; frag < skb_shinfo(skb)->nr_frags; frag++) {
+ skb_frag_t *this_frag = &skb_shinfo(skb)->frags[frag];
+
+ tx_index = eth_alloc_tx_desc_index(mp);
+ desc = &mp->p_tx_desc_area[tx_index];
+
+ desc->cmd_sts = ETH_BUFFER_OWNED_BY_DMA;
+ /* Last Frag enables interrupt and frees the skb */
+ if (frag == (skb_shinfo(skb)->nr_frags - 1)) {
+ desc->cmd_sts |= ETH_ZERO_PADDING |
+ ETH_TX_LAST_DESC |
+ ETH_TX_ENABLE_INTERRUPT;
+ mp->tx_skb[tx_index] = skb;
+ } else
+ mp->tx_skb[tx_index] = 0;
+
+ desc = &mp->p_tx_desc_area[tx_index];
+ desc->l4i_chk = 0;
+ desc->byte_cnt = this_frag->size;
+ desc->buf_ptr = dma_map_page(NULL, this_frag->page,
+ this_frag->page_offset,
+ this_frag->size,
+ DMA_TO_DEVICE);
}
+}
- /* This is a hard error, log it. */
- if ((mp->tx_ring_size - mp->tx_ring_skbs) <=
- (skb_shinfo(skb)->nr_frags + 1)) {
- netif_stop_queue(dev);
- printk(KERN_ERR
- "%s: Bug in mv643xx_eth - Trying to transmit when"
- " queue full !\n", dev->name);
- return 1;
- }
+/**
+ * eth_tx_submit_descs_for_skb - submit data from an skb to the tx hw
+ *
+ * Ensure the data for an skb to be transmitted is mapped properly,
+ * then fill in descriptors in the tx hw queue and start the hardware.
+ */
+static void eth_tx_submit_descs_for_skb(struct mv643xx_private *mp,
+ struct sk_buff *skb)
+{
+ int tx_index;
+ struct eth_tx_desc *desc;
+ u32 cmd_sts;
+ int length;
+ int nr_frags = skb_shinfo(skb)->nr_frags;
- /* Paranoid check - this shouldn't happen */
- if (skb == NULL) {
- stats->tx_dropped++;
- printk(KERN_ERR "mv64320_eth paranoid check failed\n");
- return 1;
- }
+ cmd_sts = ETH_TX_FIRST_DESC | ETH_GEN_CRC | ETH_BUFFER_OWNED_BY_DMA;
-#ifdef MV643XX_CHECKSUM_OFFLOAD_TX
- if (has_tiny_unaligned_frags(skb)) {
- if ((skb_linearize(skb, GFP_ATOMIC) != 0)) {
- stats->tx_dropped++;
- printk(KERN_DEBUG "%s: failed to linearize tiny "
- "unaligned fragment\n", dev->name);
- return 1;
- }
- }
+ tx_index = eth_alloc_tx_desc_index(mp);
+ desc = &mp->p_tx_desc_area[tx_index];
- spin_lock_irqsave(&mp->lock, flags);
+ if (nr_frags) {
+ eth_tx_fill_frag_descs(mp, skb);
- if (!skb_shinfo(skb)->nr_frags) {
- if (skb->ip_summed != CHECKSUM_HW) {
- /* Errata BTS #50, IHL must be 5 if no HW checksum */
- pkt_info.cmd_sts = ETH_TX_ENABLE_INTERRUPT |
- ETH_TX_FIRST_DESC |
- ETH_TX_LAST_DESC |
- 5 << ETH_TX_IHL_SHIFT;
- pkt_info.l4i_chk = 0;
- } else {
- pkt_info.cmd_sts = ETH_TX_ENABLE_INTERRUPT |
- ETH_TX_FIRST_DESC |
- ETH_TX_LAST_DESC |
- ETH_GEN_TCP_UDP_CHECKSUM |
- ETH_GEN_IP_V_4_CHECKSUM |
- skb->nh.iph->ihl << ETH_TX_IHL_SHIFT;
- /* CPU already calculated pseudo header checksum. */
- if ((skb->protocol == ETH_P_IP) &&
- (skb->nh.iph->protocol == IPPROTO_UDP) ) {
- pkt_info.cmd_sts |= ETH_UDP_FRAME;
- pkt_info.l4i_chk = skb->h.uh->check;
- } else if ((skb->protocol == ETH_P_IP) &&
- (skb->nh.iph->protocol == IPPROTO_TCP))
- pkt_info.l4i_chk = skb->h.th->check;
- else {
- printk(KERN_ERR
- "%s: chksum proto != IPv4 TCP or UDP\n",
- dev->name);
- spin_unlock_irqrestore(&mp->lock, flags);
- return 1;
- }
- }
- pkt_info.byte_cnt = skb->len;
- pkt_info.buf_ptr = dma_map_single(NULL, skb->data, skb->len,
- DMA_TO_DEVICE);
- pkt_info.return_info = skb;
- status = eth_port_send(mp, &pkt_info);
- if ((status == ETH_ERROR) || (status == ETH_QUEUE_FULL))
- printk(KERN_ERR "%s: Error on transmitting packet\n",
- dev->name);
- stats->tx_bytes += pkt_info.byte_cnt;
+ length = skb_headlen(skb);
+ mp->tx_skb[tx_index] = 0;
} else {
- unsigned int frag;
+ cmd_sts |= ETH_ZERO_PADDING |
+ ETH_TX_LAST_DESC |
+ ETH_TX_ENABLE_INTERRUPT;
+ length = skb->len;
+ mp->tx_skb[tx_index] = skb;
+ }
- /* first frag which is skb header */
- pkt_info.byte_cnt = skb_headlen(skb);
- pkt_info.buf_ptr = dma_map_single(NULL, skb->data,
- skb_headlen(skb),
- DMA_TO_DEVICE);
- pkt_info.l4i_chk = 0;
- pkt_info.return_info = 0;
-
- if (skb->ip_summed != CHECKSUM_HW)
- /* Errata BTS #50, IHL must be 5 if no HW checksum */
- pkt_info.cmd_sts = ETH_TX_FIRST_DESC |
- 5 << ETH_TX_IHL_SHIFT;
- else {
- pkt_info.cmd_sts = ETH_TX_FIRST_DESC |
- ETH_GEN_TCP_UDP_CHECKSUM |
- ETH_GEN_IP_V_4_CHECKSUM |
- skb->nh.iph->ihl << ETH_TX_IHL_SHIFT;
- /* CPU already calculated pseudo header checksum. */
- if ((skb->protocol == ETH_P_IP) &&
- (skb->nh.iph->protocol == IPPROTO_UDP)) {
- pkt_info.cmd_sts |= ETH_UDP_FRAME;
- pkt_info.l4i_chk = skb->h.uh->check;
- } else if ((skb->protocol == ETH_P_IP) &&
- (skb->nh.iph->protocol == IPPROTO_TCP))
- pkt_info.l4i_chk = skb->h.th->check;
- else {
- printk(KERN_ERR
- "%s: chksum proto != IPv4 TCP or UDP\n",
- dev->name);
- spin_unlock_irqrestore(&mp->lock, flags);
- return 1;
- }
- }
+ desc->byte_cnt = length;
+ desc->buf_ptr = dma_map_single(NULL, skb->data, length, DMA_TO_DEVICE);
- status = eth_port_send(mp, &pkt_info);
- if (status != ETH_OK) {
- if ((status == ETH_ERROR))
- printk(KERN_ERR
- "%s: Error on transmitting packet\n",
- dev->name);
- if (status == ETH_QUEUE_FULL)
- printk("Error on Queue Full \n");
- if (status == ETH_QUEUE_LAST_RESOURCE)
- printk("Tx resource error \n");
+ if (skb->ip_summed == CHECKSUM_HW) {
+ BUG_ON(skb->protocol != ETH_P_IP);
+
+ cmd_sts |= ETH_GEN_TCP_UDP_CHECKSUM |
+ ETH_GEN_IP_V_4_CHECKSUM |
+ skb->nh.iph->ihl << ETH_TX_IHL_SHIFT;
+
+ switch (skb->nh.iph->protocol) {
+ case IPPROTO_UDP:
+ cmd_sts |= ETH_UDP_FRAME;
+ desc->l4i_chk = skb->h.uh->check;
+ break;
+ case IPPROTO_TCP:
+ desc->l4i_chk = skb->h.th->check;
+ break;
+ default:
+ BUG();
}
- stats->tx_bytes += pkt_info.byte_cnt;
-
- /* Check for the remaining frags */
- for (frag = 0; frag < skb_shinfo(skb)->nr_frags; frag++) {
- skb_frag_t *this_frag = &skb_shinfo(skb)->frags[frag];
- pkt_info.l4i_chk = 0x0000;
- pkt_info.cmd_sts = 0x00000000;
-
- /* Last Frag enables interrupt and frees the skb */
- if (frag == (skb_shinfo(skb)->nr_frags - 1)) {
- pkt_info.cmd_sts |= ETH_TX_ENABLE_INTERRUPT |
- ETH_TX_LAST_DESC;
- pkt_info.return_info = skb;
- } else {
- pkt_info.return_info = 0;
- }
- pkt_info.l4i_chk = 0;
- pkt_info.byte_cnt = this_frag->size;
+ } else {
+ /* Errata BTS #50, IHL must be 5 if no HW checksum */
+ cmd_sts |= 5 << ETH_TX_IHL_SHIFT;
+ desc->l4i_chk = 0;
+ }
+
+ /* ensure all other descriptors are written before first cmd_sts */
+ wmb();
+ desc->cmd_sts = cmd_sts;
- pkt_info.buf_ptr = dma_map_page(NULL, this_frag->page,
- this_frag->page_offset,
- this_frag->size,
- DMA_TO_DEVICE);
+ /* ensure all descriptors are written before poking hardware */
+ wmb();
+ mv643xx_eth_port_enable_tx(mp->port_num, ETH_TX_QUEUES_ENABLED);
- status = eth_port_send(mp, &pkt_info);
+ mp->tx_desc_count += nr_frags + 1;
+}
- if (status != ETH_OK) {
- if ((status == ETH_ERROR))
- printk(KERN_ERR "%s: Error on "
- "transmitting packet\n",
- dev->name);
+/**
+ * mv643xx_eth_start_xmit - queue an skb to the hardware for transmission
+ *
+ */
+static int mv643xx_eth_start_xmit(struct sk_buff *skb, struct net_device *dev)
+{
+ struct mv643xx_private *mp = netdev_priv(dev);
+ struct net_device_stats *stats = &mp->stats;
+ unsigned long flags;
- if (status == ETH_QUEUE_LAST_RESOURCE)
- printk("Tx resource error \n");
+ BUG_ON(netif_queue_stopped(dev));
+ BUG_ON(skb == NULL);
+ BUG_ON(mp->tx_ring_size - mp->tx_desc_count < MAX_DESCS_PER_SKB);
- if (status == ETH_QUEUE_FULL)
- printk("Queue is full \n");
- }
- stats->tx_bytes += pkt_info.byte_cnt;
+ if (has_tiny_unaligned_frags(skb)) {
+ if ((skb_linearize(skb, GFP_ATOMIC) != 0)) {
+ stats->tx_dropped++;
+ printk(KERN_DEBUG "%s: failed to linearize tiny "
+ "unaligned fragment\n", dev->name);
+ return 1;
}
}
-#else
- spin_lock_irqsave(&mp->lock, flags);
- pkt_info.cmd_sts = ETH_TX_ENABLE_INTERRUPT | ETH_TX_FIRST_DESC |
- ETH_TX_LAST_DESC;
- pkt_info.l4i_chk = 0;
- pkt_info.byte_cnt = skb->len;
- pkt_info.buf_ptr = dma_map_single(NULL, skb->data, skb->len,
- DMA_TO_DEVICE);
- pkt_info.return_info = skb;
- status = eth_port_send(mp, &pkt_info);
- if ((status == ETH_ERROR) || (status == ETH_QUEUE_FULL))
- printk(KERN_ERR "%s: Error on transmitting packet\n",
- dev->name);
- stats->tx_bytes += pkt_info.byte_cnt;
-#endif
-
- /* Check if TX queue can handle another skb. If not, then
- * signal higher layers to stop requesting TX
- */
- if (mp->tx_ring_size <= (mp->tx_ring_skbs + MAX_DESCS_PER_SKB))
- /*
- * Stop getting skb's from upper layers.
- * Getting skb's from upper layers will be enabled again after
- * packets are released.
- */
- netif_stop_queue(dev);
+ spin_lock_irqsave(&mp->lock, flags);
- /* Update statistics and start of transmittion time */
+ eth_tx_submit_descs_for_skb(mp, skb);
+ stats->tx_bytes = skb->len;
stats->tx_packets++;
dev->trans_start = jiffies;
+ if (mp->tx_ring_size - mp->tx_desc_count < MAX_DESCS_PER_SKB)
+ netif_stop_queue(dev);
+
spin_unlock_irqrestore(&mp->lock, flags);
return 0; /* success */
@@ -1306,16 +1235,45 @@ static void mv643xx_netpoll(struct net_device *netdev)
struct mv643xx_private *mp = netdev_priv(netdev);
int port_num = mp->port_num;
- mv_write(MV643XX_ETH_INTERRUPT_MASK_REG(port_num), INT_MASK_ALL);
+ mv_write(MV643XX_ETH_INTERRUPT_MASK_REG(port_num), ETH_INT_MASK_ALL);
/* wait for previous write to complete */
mv_read(MV643XX_ETH_INTERRUPT_MASK_REG(port_num));
mv643xx_eth_int_handler(netdev->irq, netdev, NULL);
- mv_write(MV643XX_ETH_INTERRUPT_MASK_REG(port_num), INT_UNMASK_ALL);
+ mv_write(MV643XX_ETH_INTERRUPT_MASK_REG(port_num), ETH_INT_UNMASK_ALL);
}
#endif
+static void mv643xx_init_ethtool_cmd(struct net_device *dev, int phy_address,
+ int speed, int duplex,
+ struct ethtool_cmd *cmd)
+{
+ struct mv643xx_private *mp = netdev_priv(dev);
+
+ memset(cmd, 0, sizeof(*cmd));
+
+ cmd->port = PORT_MII;
+ cmd->transceiver = XCVR_INTERNAL;
+ cmd->phy_address = phy_address;
+
+ if (speed == 0) {
+ cmd->autoneg = AUTONEG_ENABLE;
+ /* mii lib checks, but doesn't use speed on AUTONEG_ENABLE */
+ cmd->speed = SPEED_100;
+ cmd->advertising = ADVERTISED_10baseT_Half |
+ ADVERTISED_10baseT_Full |
+ ADVERTISED_100baseT_Half |
+ ADVERTISED_100baseT_Full;
+ if (mp->mii.supports_gmii)
+ cmd->advertising |= ADVERTISED_1000baseT_Full;
+ } else {
+ cmd->autoneg = AUTONEG_DISABLE;
+ cmd->speed = speed;
+ cmd->duplex = duplex;
+ }
+}
+
/*/
* mv643xx_eth_probe
*
@@ -1336,6 +1294,9 @@ static int mv643xx_eth_probe(struct platform_device *pdev)
u8 *p;
struct resource *res;
int err;
+ struct ethtool_cmd cmd;
+ int duplex = DUPLEX_HALF;
+ int speed = 0; /* default to auto-negotiation */
dev = alloc_etherdev(sizeof(struct mv643xx_private));
if (!dev)
@@ -1373,6 +1334,7 @@ static int mv643xx_eth_probe(struct platform_device *pdev)
dev->tx_queue_len = mp->tx_ring_size;
dev->base_addr = 0;
dev->change_mtu = mv643xx_eth_change_mtu;
+ dev->do_ioctl = mv643xx_eth_do_ioctl;
SET_ETHTOOL_OPS(dev, &mv643xx_ethtool_ops);
#ifdef MV643XX_CHECKSUM_OFFLOAD_TX
@@ -1393,33 +1355,17 @@ static int mv643xx_eth_probe(struct platform_device *pdev)
/* set default config values */
eth_port_uc_addr_get(dev, dev->dev_addr);
- mp->port_config = MV643XX_ETH_PORT_CONFIG_DEFAULT_VALUE;
- mp->port_config_extend = MV643XX_ETH_PORT_CONFIG_EXTEND_DEFAULT_VALUE;
- mp->port_sdma_config = MV643XX_ETH_PORT_SDMA_CONFIG_DEFAULT_VALUE;
- mp->port_serial_control = MV643XX_ETH_PORT_SERIAL_CONTROL_DEFAULT_VALUE;
mp->rx_ring_size = MV643XX_ETH_PORT_DEFAULT_RECEIVE_QUEUE_SIZE;
mp->tx_ring_size = MV643XX_ETH_PORT_DEFAULT_TRANSMIT_QUEUE_SIZE;
pd = pdev->dev.platform_data;
if (pd) {
- if (pd->mac_addr != NULL)
+ if (pd->mac_addr)
memcpy(dev->dev_addr, pd->mac_addr, 6);
if (pd->phy_addr || pd->force_phy_addr)
ethernet_phy_set(port_num, pd->phy_addr);
- if (pd->port_config || pd->force_port_config)
- mp->port_config = pd->port_config;
-
- if (pd->port_config_extend || pd->force_port_config_extend)
- mp->port_config_extend = pd->port_config_extend;
-
- if (pd->port_sdma_config || pd->force_port_sdma_config)
- mp->port_sdma_config = pd->port_sdma_config;
-
- if (pd->port_serial_control || pd->force_port_serial_control)
- mp->port_serial_control = pd->port_serial_control;
-
if (pd->rx_queue_size)
mp->rx_ring_size = pd->rx_queue_size;
@@ -1435,16 +1381,33 @@ static int mv643xx_eth_probe(struct platform_device *pdev)
mp->rx_sram_size = pd->rx_sram_size;
mp->rx_sram_addr = pd->rx_sram_addr;
}
+
+ duplex = pd->duplex;
+ speed = pd->speed;
}
+ /* Hook up MII support for ethtool */
+ mp->mii.dev = dev;
+ mp->mii.mdio_read = mv643xx_mdio_read;
+ mp->mii.mdio_write = mv643xx_mdio_write;
+ mp->mii.phy_id = ethernet_phy_get(port_num);
+ mp->mii.phy_id_mask = 0x3f;
+ mp->mii.reg_num_mask = 0x1f;
+
err = ethernet_phy_detect(port_num);
if (err) {
pr_debug("MV643xx ethernet port %d: "
"No PHY detected at addr %d\n",
port_num, ethernet_phy_get(port_num));
- return err;
+ goto out;
}
+ ethernet_phy_reset(port_num);
+ mp->mii.supports_gmii = mii_check_gmii_support(&mp->mii);
+ mv643xx_init_ethtool_cmd(dev, mp->mii.phy_id, speed, duplex, &cmd);
+ mv643xx_eth_update_pscr(dev, &cmd);
+ mv643xx_set_settings(dev, &cmd);
+
err = register_netdev(dev);
if (err)
goto out;
@@ -1689,26 +1652,9 @@ MODULE_DESCRIPTION("Ethernet driver for Marvell MV643XX");
* to the Rx descriptor ring to enable the reuse of this source.
* Return Rx resource is done using the eth_rx_return_buff API.
*
- * Transmit operation:
- * The eth_port_send API supports Scatter-Gather which enables to
- * send a packet spanned over multiple buffers. This means that
- * for each packet info structure given by the user and put into
- * the Tx descriptors ring, will be transmitted only if the 'LAST'
- * bit will be set in the packet info command status field. This
- * API also consider restriction regarding buffer alignments and
- * sizes.
- * The user must return a Tx resource after ensuring the buffer
- * has been transmitted to enable the Tx ring indexes to update.
- *
- * BOARD LAYOUT
- * This device is on-board. No jumper diagram is necessary.
- *
- * EXTERNAL INTERFACE
- *
* Prior to calling the initialization routine eth_port_init() the user
* must set the following fields under mv643xx_private struct:
* port_num User Ethernet port number.
- * port_mac_addr[6] User defined port MAC address.
* port_config User port configuration value.
* port_config_extend User port config extend value.
* port_sdma_config User port SDMA config value.
@@ -1725,20 +1671,12 @@ MODULE_DESCRIPTION("Ethernet driver for Marvell MV643XX");
* return_info Tx/Rx user resource return information.
*/
-/* defines */
-/* SDMA command macros */
-#define ETH_ENABLE_TX_QUEUE(eth_port) \
- mv_write(MV643XX_ETH_TRANSMIT_QUEUE_COMMAND_REG(eth_port), 1)
-
-/* locals */
-
/* PHY routines */
static int ethernet_phy_get(unsigned int eth_port_num);
static void ethernet_phy_set(unsigned int eth_port_num, int phy_addr);
/* Ethernet Port routines */
-static int eth_port_uc_addr(unsigned int eth_port_num, unsigned char uc_nibble,
- int option);
+static void eth_port_set_filter_table_entry(int table, unsigned char entry);
/*
* eth_port_init - Initialize the Ethernet port driver
@@ -1766,17 +1704,11 @@ static int eth_port_uc_addr(unsigned int eth_port_num, unsigned char uc_nibble,
*/
static void eth_port_init(struct mv643xx_private *mp)
{
- mp->port_rx_queue_command = 0;
- mp->port_tx_queue_command = 0;
-
mp->rx_resource_err = 0;
- mp->tx_resource_err = 0;
eth_port_reset(mp->port_num);
eth_port_init_mac_tables(mp->port_num);
-
- ethernet_phy_reset(mp->port_num);
}
/*
@@ -1798,7 +1730,7 @@ static void eth_port_init(struct mv643xx_private *mp)
* and ether_init_rx_desc_ring for Rx queues).
*
* INPUT:
- * struct mv643xx_private *mp Ethernet port control struct
+ * dev - a pointer to the required interface
*
* OUTPUT:
* Ethernet port is ready to receive and transmit.
@@ -1806,10 +1738,13 @@ static void eth_port_init(struct mv643xx_private *mp)
* RETURN:
* None.
*/
-static void eth_port_start(struct mv643xx_private *mp)
+static void eth_port_start(struct net_device *dev)
{
+ struct mv643xx_private *mp = netdev_priv(dev);
unsigned int port_num = mp->port_num;
int tx_curr_desc, rx_curr_desc;
+ u32 pscr;
+ struct ethtool_cmd ethtool_cmd;
/* Assignment of Tx CTRP of given queue */
tx_curr_desc = mp->tx_curr_desc_q;
@@ -1822,37 +1757,45 @@ static void eth_port_start(struct mv643xx_private *mp)
(u32)((struct eth_rx_desc *)mp->rx_desc_dma + rx_curr_desc));
/* Add the assigned Ethernet address to the port's address table */
- eth_port_uc_addr_set(port_num, mp->port_mac_addr);
+ eth_port_uc_addr_set(port_num, dev->dev_addr);
/* Assign port configuration and command. */
- mv_write(MV643XX_ETH_PORT_CONFIG_REG(port_num), mp->port_config);
+ mv_write(MV643XX_ETH_PORT_CONFIG_REG(port_num),
+ MV643XX_ETH_PORT_CONFIG_DEFAULT_VALUE);
mv_write(MV643XX_ETH_PORT_CONFIG_EXTEND_REG(port_num),
- mp->port_config_extend);
+ MV643XX_ETH_PORT_CONFIG_EXTEND_DEFAULT_VALUE);
+ pscr = mv_read(MV643XX_ETH_PORT_SERIAL_CONTROL_REG(port_num));
- /* Increase the Rx side buffer size if supporting GigE */
- if (mp->port_serial_control & MV643XX_ETH_SET_GMII_SPEED_TO_1000)
- mv_write(MV643XX_ETH_PORT_SERIAL_CONTROL_REG(port_num),
- (mp->port_serial_control & 0xfff1ffff) | (0x5 << 17));
- else
- mv_write(MV643XX_ETH_PORT_SERIAL_CONTROL_REG(port_num),
- mp->port_serial_control);
+ pscr &= ~(MV643XX_ETH_SERIAL_PORT_ENABLE | MV643XX_ETH_FORCE_LINK_PASS);
+ mv_write(MV643XX_ETH_PORT_SERIAL_CONTROL_REG(port_num), pscr);
+
+ pscr |= MV643XX_ETH_DISABLE_AUTO_NEG_FOR_FLOW_CTRL |
+ MV643XX_ETH_DISABLE_AUTO_NEG_SPEED_GMII |
+ MV643XX_ETH_DISABLE_AUTO_NEG_FOR_DUPLX |
+ MV643XX_ETH_DO_NOT_FORCE_LINK_FAIL |
+ MV643XX_ETH_SERIAL_PORT_CONTROL_RESERVED;
- mv_write(MV643XX_ETH_PORT_SERIAL_CONTROL_REG(port_num),
- mv_read(MV643XX_ETH_PORT_SERIAL_CONTROL_REG(port_num)) |
- MV643XX_ETH_SERIAL_PORT_ENABLE);
+ mv_write(MV643XX_ETH_PORT_SERIAL_CONTROL_REG(port_num), pscr);
+
+ pscr |= MV643XX_ETH_SERIAL_PORT_ENABLE;
+ mv_write(MV643XX_ETH_PORT_SERIAL_CONTROL_REG(port_num), pscr);
/* Assign port SDMA configuration */
mv_write(MV643XX_ETH_SDMA_CONFIG_REG(port_num),
- mp->port_sdma_config);
+ MV643XX_ETH_PORT_SDMA_CONFIG_DEFAULT_VALUE);
/* Enable port Rx. */
- mv_write(MV643XX_ETH_RECEIVE_QUEUE_COMMAND_REG(port_num),
- mp->port_rx_queue_command);
+ mv643xx_eth_port_enable_rx(port_num, ETH_RX_QUEUES_ENABLED);
/* Disable port bandwidth limits by clearing MTU register */
mv_write(MV643XX_ETH_MAXIMUM_TRANSMIT_UNIT(port_num), 0);
+
+ /* save phy settings across reset */
+ mv643xx_get_settings(dev, &ethtool_cmd);
+ ethernet_phy_reset(mp->port_num);
+ mv643xx_set_settings(dev, &ethtool_cmd);
}
/*
@@ -1866,8 +1809,9 @@ static void eth_port_start(struct mv643xx_private *mp)
* char * p_addr Address to be set
*
* OUTPUT:
- * Set MAC address low and high registers. also calls eth_port_uc_addr()
- * To set the unicast table with the proper information.
+ * Set MAC address low and high registers. also calls
+ * eth_port_set_filter_table_entry() to set the unicast
+ * table with the proper information.
*
* RETURN:
* N/A.
@@ -1878,6 +1822,7 @@ static void eth_port_uc_addr_set(unsigned int eth_port_num,
{
unsigned int mac_h;
unsigned int mac_l;
+ int table;
mac_l = (p_addr[4] << 8) | (p_addr[5]);
mac_h = (p_addr[0] << 24) | (p_addr[1] << 16) | (p_addr[2] << 8) |
@@ -1887,9 +1832,8 @@ static void eth_port_uc_addr_set(unsigned int eth_port_num,
mv_write(MV643XX_ETH_MAC_ADDR_HIGH(eth_port_num), mac_h);
/* Accept frames of this address */
- eth_port_uc_addr(eth_port_num, p_addr[5], ACCEPT_MAC_ADDR);
-
- return;
+ table = MV643XX_ETH_DA_FILTER_UNICAST_TABLE_BASE(eth_port_num);
+ eth_port_set_filter_table_entry(table, p_addr[5] & 0x0f);
}
/*
@@ -1928,72 +1872,6 @@ static void eth_port_uc_addr_get(struct net_device *dev, unsigned char *p_addr)
}
/*
- * eth_port_uc_addr - This function Set the port unicast address table
- *
- * DESCRIPTION:
- * This function locates the proper entry in the Unicast table for the
- * specified MAC nibble and sets its properties according to function
- * parameters.
- *
- * INPUT:
- * unsigned int eth_port_num Port number.
- * unsigned char uc_nibble Unicast MAC Address last nibble.
- * int option 0 = Add, 1 = remove address.
- *
- * OUTPUT:
- * This function add/removes MAC addresses from the port unicast address
- * table.
- *
- * RETURN:
- * true is output succeeded.
- * false if option parameter is invalid.
- *
- */
-static int eth_port_uc_addr(unsigned int eth_port_num, unsigned char uc_nibble,
- int option)
-{
- unsigned int unicast_reg;
- unsigned int tbl_offset;
- unsigned int reg_offset;
-
- /* Locate the Unicast table entry */
- uc_nibble = (0xf & uc_nibble);
- tbl_offset = (uc_nibble / 4) * 4; /* Register offset from unicast table base */
- reg_offset = uc_nibble % 4; /* Entry offset within the above register */
-
- switch (option) {
- case REJECT_MAC_ADDR:
- /* Clear accepts frame bit at given unicast DA table entry */
- unicast_reg = mv_read((MV643XX_ETH_DA_FILTER_UNICAST_TABLE_BASE
- (eth_port_num) + tbl_offset));
-
- unicast_reg &= (0x0E << (8 * reg_offset));
-
- mv_write((MV643XX_ETH_DA_FILTER_UNICAST_TABLE_BASE
- (eth_port_num) + tbl_offset), unicast_reg);
- break;
-
- case ACCEPT_MAC_ADDR:
- /* Set accepts frame bit at unicast DA filter table entry */
- unicast_reg =
- mv_read((MV643XX_ETH_DA_FILTER_UNICAST_TABLE_BASE
- (eth_port_num) + tbl_offset));
-
- unicast_reg |= (0x01 << (8 * reg_offset));
-
- mv_write((MV643XX_ETH_DA_FILTER_UNICAST_TABLE_BASE
- (eth_port_num) + tbl_offset), unicast_reg);
-
- break;
-
- default:
- return 0;
- }
-
- return 1;
-}
-
-/*
* The entries in each table are indexed by a hash of a packet's MAC
* address. One bit in each entry determines whether the packet is
* accepted. There are 4 entries (each 8 bits wide) in each register
@@ -2205,8 +2083,8 @@ static void eth_port_init_mac_tables(unsigned int eth_port_num)
/* Clear DA filter unicast table (Ex_dFUT) */
for (table_index = 0; table_index <= 0xC; table_index += 4)
- mv_write((MV643XX_ETH_DA_FILTER_UNICAST_TABLE_BASE
- (eth_port_num) + table_index), 0);
+ mv_write(MV643XX_ETH_DA_FILTER_UNICAST_TABLE_BASE
+ (eth_port_num) + table_index, 0);
for (table_index = 0; table_index <= 0xFC; table_index += 4) {
/* Clear DA filter special multicast table (Ex_dFSMT) */
@@ -2389,6 +2267,73 @@ static void ethernet_phy_reset(unsigned int eth_port_num)
eth_port_read_smi_reg(eth_port_num, 0, &phy_reg_data);
phy_reg_data |= 0x8000; /* Set bit 15 to reset the PHY */
eth_port_write_smi_reg(eth_port_num, 0, phy_reg_data);
+
+ /* wait for PHY to come out of reset */
+ do {
+ udelay(1);
+ eth_port_read_smi_reg(eth_port_num, 0, &phy_reg_data);
+ } while (phy_reg_data & 0x8000);
+}
+
+static void mv643xx_eth_port_enable_tx(unsigned int port_num,
+ unsigned int queues)
+{
+ mv_write(MV643XX_ETH_TRANSMIT_QUEUE_COMMAND_REG(port_num), queues);
+}
+
+static void mv643xx_eth_port_enable_rx(unsigned int port_num,
+ unsigned int queues)
+{
+ mv_write(MV643XX_ETH_RECEIVE_QUEUE_COMMAND_REG(port_num), queues);
+}
+
+static unsigned int mv643xx_eth_port_disable_tx(unsigned int port_num)
+{
+ u32 queues;
+
+ /* Stop Tx port activity. Check port Tx activity. */
+ queues = mv_read(MV643XX_ETH_TRANSMIT_QUEUE_COMMAND_REG(port_num))
+ & 0xFF;
+ if (queues) {
+ /* Issue stop command for active queues only */
+ mv_write(MV643XX_ETH_TRANSMIT_QUEUE_COMMAND_REG(port_num),
+ (queues << 8));
+
+ /* Wait for all Tx activity to terminate. */
+ /* Check port cause register that all Tx queues are stopped */
+ while (mv_read(MV643XX_ETH_TRANSMIT_QUEUE_COMMAND_REG(port_num))
+ & 0xFF)
+ udelay(PHY_WAIT_MICRO_SECONDS);
+
+ /* Wait for Tx FIFO to empty */
+ while (mv_read(MV643XX_ETH_PORT_STATUS_REG(port_num)) &
+ ETH_PORT_TX_FIFO_EMPTY)
+ udelay(PHY_WAIT_MICRO_SECONDS);
+ }
+
+ return queues;
+}
+
+static unsigned int mv643xx_eth_port_disable_rx(unsigned int port_num)
+{
+ u32 queues;
+
+ /* Stop Rx port activity. Check port Rx activity. */
+ queues = mv_read(MV643XX_ETH_RECEIVE_QUEUE_COMMAND_REG(port_num))
+ & 0xFF;
+ if (queues) {
+ /* Issue stop command for active queues only */
+ mv_write(MV643XX_ETH_RECEIVE_QUEUE_COMMAND_REG(port_num),
+ (queues << 8));
+
+ /* Wait for all Rx activity to terminate. */
+ /* Check port cause register that all Rx queues are stopped */
+ while (mv_read(MV643XX_ETH_RECEIVE_QUEUE_COMMAND_REG(port_num))
+ & 0xFF)
+ udelay(PHY_WAIT_MICRO_SECONDS);
+ }
+
+ return queues;
}
/*
@@ -2413,70 +2358,21 @@ static void eth_port_reset(unsigned int port_num)
{
unsigned int reg_data;
- /* Stop Tx port activity. Check port Tx activity. */
- reg_data = mv_read(MV643XX_ETH_TRANSMIT_QUEUE_COMMAND_REG(port_num));
-
- if (reg_data & 0xFF) {
- /* Issue stop command for active channels only */
- mv_write(MV643XX_ETH_TRANSMIT_QUEUE_COMMAND_REG(port_num),
- (reg_data << 8));
-
- /* Wait for all Tx activity to terminate. */
- /* Check port cause register that all Tx queues are stopped */
- while (mv_read(MV643XX_ETH_TRANSMIT_QUEUE_COMMAND_REG(port_num))
- & 0xFF)
- udelay(10);
- }
-
- /* Stop Rx port activity. Check port Rx activity. */
- reg_data = mv_read(MV643XX_ETH_RECEIVE_QUEUE_COMMAND_REG(port_num));
-
- if (reg_data & 0xFF) {
- /* Issue stop command for active channels only */
- mv_write(MV643XX_ETH_RECEIVE_QUEUE_COMMAND_REG(port_num),
- (reg_data << 8));
-
- /* Wait for all Rx activity to terminate. */
- /* Check port cause register that all Rx queues are stopped */
- while (mv_read(MV643XX_ETH_RECEIVE_QUEUE_COMMAND_REG(port_num))
- & 0xFF)
- udelay(10);
- }
+ mv643xx_eth_port_disable_tx(port_num);
+ mv643xx_eth_port_disable_rx(port_num);
/* Clear all MIB counters */
eth_clear_mib_counters(port_num);
/* Reset the Enable bit in the Configuration Register */
reg_data = mv_read(MV643XX_ETH_PORT_SERIAL_CONTROL_REG(port_num));
- reg_data &= ~MV643XX_ETH_SERIAL_PORT_ENABLE;
+ reg_data &= ~(MV643XX_ETH_SERIAL_PORT_ENABLE |
+ MV643XX_ETH_DO_NOT_FORCE_LINK_FAIL |
+ MV643XX_ETH_FORCE_LINK_PASS);
mv_write(MV643XX_ETH_PORT_SERIAL_CONTROL_REG(port_num), reg_data);
}
-static int eth_port_autoneg_supported(unsigned int eth_port_num)
-{
- unsigned int phy_reg_data0;
-
- eth_port_read_smi_reg(eth_port_num, 0, &phy_reg_data0);
-
- return phy_reg_data0 & 0x1000;
-}
-
-static int eth_port_link_is_up(unsigned int eth_port_num)
-{
- unsigned int phy_reg_data1;
-
- eth_port_read_smi_reg(eth_port_num, 1, &phy_reg_data1);
-
- if (eth_port_autoneg_supported(eth_port_num)) {
- if (phy_reg_data1 & 0x20) /* auto-neg complete */
- return 1;
- } else if (phy_reg_data1 & 0x4) /* link up */
- return 1;
-
- return 0;
-}
-
/*
* eth_port_read_smi_reg - Read PHY registers
*
@@ -2582,250 +2478,21 @@ out:
}
/*
- * eth_port_send - Send an Ethernet packet
- *
- * DESCRIPTION:
- * This routine send a given packet described by p_pktinfo parameter. It
- * supports transmitting of a packet spaned over multiple buffers. The
- * routine updates 'curr' and 'first' indexes according to the packet
- * segment passed to the routine. In case the packet segment is first,
- * the 'first' index is update. In any case, the 'curr' index is updated.
- * If the routine get into Tx resource error it assigns 'curr' index as
- * 'first'. This way the function can abort Tx process of multiple
- * descriptors per packet.
- *
- * INPUT:
- * struct mv643xx_private *mp Ethernet Port Control srtuct.
- * struct pkt_info *p_pkt_info User packet buffer.
- *
- * OUTPUT:
- * Tx ring 'curr' and 'first' indexes are updated.
- *
- * RETURN:
- * ETH_QUEUE_FULL in case of Tx resource error.
- * ETH_ERROR in case the routine can not access Tx desc ring.
- * ETH_QUEUE_LAST_RESOURCE if the routine uses the last Tx resource.
- * ETH_OK otherwise.
- *
- */
-#ifdef MV643XX_CHECKSUM_OFFLOAD_TX
-/*
- * Modified to include the first descriptor pointer in case of SG
+ * Wrappers for MII support library.
*/
-static ETH_FUNC_RET_STATUS eth_port_send(struct mv643xx_private *mp,
- struct pkt_info *p_pkt_info)
-{
- int tx_desc_curr, tx_desc_used, tx_first_desc, tx_next_desc;
- struct eth_tx_desc *current_descriptor;
- struct eth_tx_desc *first_descriptor;
- u32 command;
-
- /* Do not process Tx ring in case of Tx ring resource error */
- if (mp->tx_resource_err)
- return ETH_QUEUE_FULL;
-
- /*
- * The hardware requires that each buffer that is <= 8 bytes
- * in length must be aligned on an 8 byte boundary.
- */
- if (p_pkt_info->byte_cnt <= 8 && p_pkt_info->buf_ptr & 0x7) {
- printk(KERN_ERR
- "mv643xx_eth port %d: packet size <= 8 problem\n",
- mp->port_num);
- return ETH_ERROR;
- }
-
- mp->tx_ring_skbs++;
- BUG_ON(mp->tx_ring_skbs > mp->tx_ring_size);
-
- /* Get the Tx Desc ring indexes */
- tx_desc_curr = mp->tx_curr_desc_q;
- tx_desc_used = mp->tx_used_desc_q;
-
- current_descriptor = &mp->p_tx_desc_area[tx_desc_curr];
-
- tx_next_desc = (tx_desc_curr + 1) % mp->tx_ring_size;
-
- current_descriptor->buf_ptr = p_pkt_info->buf_ptr;
- current_descriptor->byte_cnt = p_pkt_info->byte_cnt;
- current_descriptor->l4i_chk = p_pkt_info->l4i_chk;
- mp->tx_skb[tx_desc_curr] = p_pkt_info->return_info;
-
- command = p_pkt_info->cmd_sts | ETH_ZERO_PADDING | ETH_GEN_CRC |
- ETH_BUFFER_OWNED_BY_DMA;
- if (command & ETH_TX_FIRST_DESC) {
- tx_first_desc = tx_desc_curr;
- mp->tx_first_desc_q = tx_first_desc;
- first_descriptor = current_descriptor;
- mp->tx_first_command = command;
- } else {
- tx_first_desc = mp->tx_first_desc_q;
- first_descriptor = &mp->p_tx_desc_area[tx_first_desc];
- BUG_ON(first_descriptor == NULL);
- current_descriptor->cmd_sts = command;
- }
-
- if (command & ETH_TX_LAST_DESC) {
- wmb();
- first_descriptor->cmd_sts = mp->tx_first_command;
-
- wmb();
- ETH_ENABLE_TX_QUEUE(mp->port_num);
-
- /*
- * Finish Tx packet. Update first desc in case of Tx resource
- * error */
- tx_first_desc = tx_next_desc;
- mp->tx_first_desc_q = tx_first_desc;
- }
-
- /* Check for ring index overlap in the Tx desc ring */
- if (tx_next_desc == tx_desc_used) {
- mp->tx_resource_err = 1;
- mp->tx_curr_desc_q = tx_first_desc;
-
- return ETH_QUEUE_LAST_RESOURCE;
- }
-
- mp->tx_curr_desc_q = tx_next_desc;
-
- return ETH_OK;
-}
-#else
-static ETH_FUNC_RET_STATUS eth_port_send(struct mv643xx_private *mp,
- struct pkt_info *p_pkt_info)
+static int mv643xx_mdio_read(struct net_device *dev, int phy_id, int location)
{
- int tx_desc_curr;
- int tx_desc_used;
- struct eth_tx_desc *current_descriptor;
- unsigned int command_status;
-
- /* Do not process Tx ring in case of Tx ring resource error */
- if (mp->tx_resource_err)
- return ETH_QUEUE_FULL;
-
- mp->tx_ring_skbs++;
- BUG_ON(mp->tx_ring_skbs > mp->tx_ring_size);
-
- /* Get the Tx Desc ring indexes */
- tx_desc_curr = mp->tx_curr_desc_q;
- tx_desc_used = mp->tx_used_desc_q;
- current_descriptor = &mp->p_tx_desc_area[tx_desc_curr];
-
- command_status = p_pkt_info->cmd_sts | ETH_ZERO_PADDING | ETH_GEN_CRC;
- current_descriptor->buf_ptr = p_pkt_info->buf_ptr;
- current_descriptor->byte_cnt = p_pkt_info->byte_cnt;
- mp->tx_skb[tx_desc_curr] = p_pkt_info->return_info;
-
- /* Set last desc with DMA ownership and interrupt enable. */
- wmb();
- current_descriptor->cmd_sts = command_status |
- ETH_BUFFER_OWNED_BY_DMA | ETH_TX_ENABLE_INTERRUPT;
-
- wmb();
- ETH_ENABLE_TX_QUEUE(mp->port_num);
-
- /* Finish Tx packet. Update first desc in case of Tx resource error */
- tx_desc_curr = (tx_desc_curr + 1) % mp->tx_ring_size;
-
- /* Update the current descriptor */
- mp->tx_curr_desc_q = tx_desc_curr;
-
- /* Check for ring index overlap in the Tx desc ring */
- if (tx_desc_curr == tx_desc_used) {
- mp->tx_resource_err = 1;
- return ETH_QUEUE_LAST_RESOURCE;
- }
+ int val;
+ struct mv643xx_private *mp = netdev_priv(dev);
- return ETH_OK;
+ eth_port_read_smi_reg(mp->port_num, location, &val);
+ return val;
}
-#endif
-/*
- * eth_tx_return_desc - Free all used Tx descriptors
- *
- * DESCRIPTION:
- * This routine returns the transmitted packet information to the caller.
- * It uses the 'first' index to support Tx desc return in case a transmit
- * of a packet spanned over multiple buffer still in process.
- * In case the Tx queue was in "resource error" condition, where there are
- * no available Tx resources, the function resets the resource error flag.
- *
- * INPUT:
- * struct mv643xx_private *mp Ethernet Port Control srtuct.
- * struct pkt_info *p_pkt_info User packet buffer.
- *
- * OUTPUT:
- * Tx ring 'first' and 'used' indexes are updated.
- *
- * RETURN:
- * ETH_OK on success
- * ETH_ERROR otherwise.
- *
- */
-static ETH_FUNC_RET_STATUS eth_tx_return_desc(struct mv643xx_private *mp,
- struct pkt_info *p_pkt_info)
+static void mv643xx_mdio_write(struct net_device *dev, int phy_id, int location, int val)
{
- int tx_desc_used;
- int tx_busy_desc;
- struct eth_tx_desc *p_tx_desc_used;
- unsigned int command_status;
- unsigned long flags;
- int err = ETH_OK;
-
- spin_lock_irqsave(&mp->lock, flags);
-
-#ifdef MV643XX_CHECKSUM_OFFLOAD_TX
- tx_busy_desc = mp->tx_first_desc_q;
-#else
- tx_busy_desc = mp->tx_curr_desc_q;
-#endif
-
- /* Get the Tx Desc ring indexes */
- tx_desc_used = mp->tx_used_desc_q;
-
- p_tx_desc_used = &mp->p_tx_desc_area[tx_desc_used];
-
- /* Sanity check */
- if (p_tx_desc_used == NULL) {
- err = ETH_ERROR;
- goto out;
- }
-
- /* Stop release. About to overlap the current available Tx descriptor */
- if (tx_desc_used == tx_busy_desc && !mp->tx_resource_err) {
- err = ETH_ERROR;
- goto out;
- }
-
- command_status = p_tx_desc_used->cmd_sts;
-
- /* Still transmitting... */
- if (command_status & (ETH_BUFFER_OWNED_BY_DMA)) {
- err = ETH_ERROR;
- goto out;
- }
-
- /* Pass the packet information to the caller */
- p_pkt_info->cmd_sts = command_status;
- p_pkt_info->return_info = mp->tx_skb[tx_desc_used];
- p_pkt_info->buf_ptr = p_tx_desc_used->buf_ptr;
- p_pkt_info->byte_cnt = p_tx_desc_used->byte_cnt;
- mp->tx_skb[tx_desc_used] = NULL;
-
- /* Update the next descriptor to release. */
- mp->tx_used_desc_q = (tx_desc_used + 1) % mp->tx_ring_size;
-
- /* Any Tx return cancels the Tx resource error status */
- mp->tx_resource_err = 0;
-
- BUG_ON(mp->tx_ring_skbs == 0);
- mp->tx_ring_skbs--;
-
-out:
- spin_unlock_irqrestore(&mp->lock, flags);
-
- return err;
+ struct mv643xx_private *mp = netdev_priv(dev);
+ eth_port_write_smi_reg(mp->port_num, location, val);
}
/*
@@ -3017,111 +2684,6 @@ static const struct mv643xx_stats mv643xx_gstrings_stats[] = {
#define MV643XX_STATS_LEN \
sizeof(mv643xx_gstrings_stats) / sizeof(struct mv643xx_stats)
-static int
-mv643xx_get_settings(struct net_device *netdev, struct ethtool_cmd *ecmd)
-{
- struct mv643xx_private *mp = netdev->priv;
- int port_num = mp->port_num;
- int autoneg = eth_port_autoneg_supported(port_num);
- int mode_10_bit;
- int auto_duplex;
- int half_duplex = 0;
- int full_duplex = 0;
- int auto_speed;
- int speed_10 = 0;
- int speed_100 = 0;
- int speed_1000 = 0;
-
- u32 pcs = mv_read(MV643XX_ETH_PORT_SERIAL_CONTROL_REG(port_num));
- u32 psr = mv_read(MV643XX_ETH_PORT_STATUS_REG(port_num));
-
- mode_10_bit = psr & MV643XX_ETH_PORT_STATUS_MODE_10_BIT;
-
- if (mode_10_bit) {
- ecmd->supported = SUPPORTED_10baseT_Half;
- } else {
- ecmd->supported = (SUPPORTED_10baseT_Half |
- SUPPORTED_10baseT_Full |
- SUPPORTED_100baseT_Half |
- SUPPORTED_100baseT_Full |
- SUPPORTED_1000baseT_Full |
- (autoneg ? SUPPORTED_Autoneg : 0) |
- SUPPORTED_TP);
-
- auto_duplex = !(pcs & MV643XX_ETH_DISABLE_AUTO_NEG_FOR_DUPLX);
- auto_speed = !(pcs & MV643XX_ETH_DISABLE_AUTO_NEG_SPEED_GMII);
-
- ecmd->advertising = ADVERTISED_TP;
-
- if (autoneg) {
- ecmd->advertising |= ADVERTISED_Autoneg;
-
- if (auto_duplex) {
- half_duplex = 1;
- full_duplex = 1;
- } else {
- if (pcs & MV643XX_ETH_SET_FULL_DUPLEX_MODE)
- full_duplex = 1;
- else
- half_duplex = 1;
- }
-
- if (auto_speed) {
- speed_10 = 1;
- speed_100 = 1;
- speed_1000 = 1;
- } else {
- if (pcs & MV643XX_ETH_SET_GMII_SPEED_TO_1000)
- speed_1000 = 1;
- else if (pcs & MV643XX_ETH_SET_MII_SPEED_TO_100)
- speed_100 = 1;
- else
- speed_10 = 1;
- }
-
- if (speed_10 & half_duplex)
- ecmd->advertising |= ADVERTISED_10baseT_Half;
- if (speed_10 & full_duplex)
- ecmd->advertising |= ADVERTISED_10baseT_Full;
- if (speed_100 & half_duplex)
- ecmd->advertising |= ADVERTISED_100baseT_Half;
- if (speed_100 & full_duplex)
- ecmd->advertising |= ADVERTISED_100baseT_Full;
- if (speed_1000)
- ecmd->advertising |= ADVERTISED_1000baseT_Full;
- }
- }
-
- ecmd->port = PORT_TP;
- ecmd->phy_address = ethernet_phy_get(port_num);
-
- ecmd->transceiver = XCVR_EXTERNAL;
-
- if (netif_carrier_ok(netdev)) {
- if (mode_10_bit)
- ecmd->speed = SPEED_10;
- else {
- if (psr & MV643XX_ETH_PORT_STATUS_GMII_1000)
- ecmd->speed = SPEED_1000;
- else if (psr & MV643XX_ETH_PORT_STATUS_MII_100)
- ecmd->speed = SPEED_100;
- else
- ecmd->speed = SPEED_10;
- }
-
- if (psr & MV643XX_ETH_PORT_STATUS_FULL_DUPLEX)
- ecmd->duplex = DUPLEX_FULL;
- else
- ecmd->duplex = DUPLEX_HALF;
- } else {
- ecmd->speed = -1;
- ecmd->duplex = -1;
- }
-
- ecmd->autoneg = autoneg ? AUTONEG_ENABLE : AUTONEG_DISABLE;
- return 0;
-}
-
static void mv643xx_get_drvinfo(struct net_device *netdev,
struct ethtool_drvinfo *drvinfo)
{
@@ -3168,15 +2730,41 @@ static void mv643xx_get_strings(struct net_device *netdev, uint32_t stringset,
}
}
+static u32 mv643xx_eth_get_link(struct net_device *dev)
+{
+ struct mv643xx_private *mp = netdev_priv(dev);
+
+ return mii_link_ok(&mp->mii);
+}
+
+static int mv643xx_eth_nway_restart(struct net_device *dev)
+{
+ struct mv643xx_private *mp = netdev_priv(dev);
+
+ return mii_nway_restart(&mp->mii);
+}
+
+static int mv643xx_eth_do_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd)
+{
+ struct mv643xx_private *mp = netdev_priv(dev);
+
+ return generic_mii_ioctl(&mp->mii, if_mii(ifr), cmd, NULL);
+}
+
static struct ethtool_ops mv643xx_ethtool_ops = {
.get_settings = mv643xx_get_settings,
+ .set_settings = mv643xx_set_settings,
.get_drvinfo = mv643xx_get_drvinfo,
- .get_link = ethtool_op_get_link,
+ .get_link = mv643xx_eth_get_link,
.get_sg = ethtool_op_get_sg,
.set_sg = ethtool_op_set_sg,
.get_strings = mv643xx_get_strings,
.get_stats_count = mv643xx_get_stats_count,
.get_ethtool_stats = mv643xx_get_ethtool_stats,
+ .get_strings = mv643xx_get_strings,
+ .get_stats_count = mv643xx_get_stats_count,
+ .get_ethtool_stats = mv643xx_get_ethtool_stats,
+ .nway_reset = mv643xx_eth_nway_restart,
};
/************* End ethtool support *************************/