diff options
author | Luka Perkov <luka@openwrt.org> | 2012-11-11 20:03:56 +0000 |
---|---|---|
committer | Luka Perkov <luka@openwrt.org> | 2012-11-11 20:03:56 +0000 |
commit | e38adfc912d8232651c06202cc4f170359482169 (patch) | |
tree | 4213b27971de2cdd758b007d806be1552f29b67a /target/linux | |
parent | 306894b88fde3858773b8fd430eb6a88ed3c6c43 (diff) | |
download | upstream-e38adfc912d8232651c06202cc4f170359482169.tar.gz upstream-e38adfc912d8232651c06202cc4f170359482169.tar.bz2 upstream-e38adfc912d8232651c06202cc4f170359482169.zip |
cns3xxx: convert ethernet patches to files
SVN-Revision: 34161
Diffstat (limited to 'target/linux')
10 files changed, 1400 insertions, 2090 deletions
diff --git a/target/linux/cns3xxx/files/drivers/net/ethernet/cavium/Kconfig b/target/linux/cns3xxx/files/drivers/net/ethernet/cavium/Kconfig new file mode 100644 index 0000000000..79eebe2863 --- /dev/null +++ b/target/linux/cns3xxx/files/drivers/net/ethernet/cavium/Kconfig @@ -0,0 +1,24 @@ +config NET_VENDOR_CAVIUM + bool "Cavium devices" + default y + depends on ARCH_CNS3XXX + ---help--- + If you have a network (Ethernet) chipset belonging to this class, + say Y. + + Note that the answer to this question does not directly affect + the kernel: saying N will just case the configurator to skip all + the questions regarding AMD chipsets. If you say Y, you will be asked + for your specific chipset/driver in the following questions. + +if NET_VENDOR_CAVIUM + +config CNS3XXX_ETH + tristate "Cavium CNS3xxx Ethernet support" + depends on ARCH_CNS3XXX + select PHYLIB + help + Say Y here if you want to use built-in Ethernet ports + on CNS3XXX processor. + +endif diff --git a/target/linux/cns3xxx/files/drivers/net/ethernet/cavium/Makefile b/target/linux/cns3xxx/files/drivers/net/ethernet/cavium/Makefile new file mode 100644 index 0000000000..badd2404a6 --- /dev/null +++ b/target/linux/cns3xxx/files/drivers/net/ethernet/cavium/Makefile @@ -0,0 +1,5 @@ +# +# Makefile for the Cavium ethernet device drivers. +# + +obj-$(CONFIG_CNS3XXX_ETH) += cns3xxx_eth.o diff --git a/target/linux/cns3xxx/files/drivers/net/ethernet/cavium/cns3xxx_eth.c b/target/linux/cns3xxx/files/drivers/net/ethernet/cavium/cns3xxx_eth.c new file mode 100644 index 0000000000..a9edefe4bb --- /dev/null +++ b/target/linux/cns3xxx/files/drivers/net/ethernet/cavium/cns3xxx_eth.c @@ -0,0 +1,1349 @@ +/* + * Cavium CNS3xxx Gigabit driver for Linux + * + * Copyright 2011 Gateworks Corporation + * Chris Lang <clang@gateworks.com> + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of version 2 of the GNU General Public License + * as published by the Free Software Foundation. + * + */ + +#include <linux/delay.h> +#include <linux/module.h> +#include <linux/dma-mapping.h> +#include <linux/dmapool.h> +#include <linux/etherdevice.h> +#include <linux/interrupt.h> +#include <linux/io.h> +#include <linux/kernel.h> +#include <linux/phy.h> +#include <linux/platform_device.h> +#include <linux/skbuff.h> +#include <mach/irqs.h> +#include <mach/platform.h> + +#define DRV_NAME "cns3xxx_eth" + +#define RX_DESCS 128 +#define TX_DESCS 128 +#define TX_DESC_RESERVE 20 + +#define RX_POOL_ALLOC_SIZE (sizeof(struct rx_desc) * RX_DESCS) +#define TX_POOL_ALLOC_SIZE (sizeof(struct tx_desc) * TX_DESCS) +#define REGS_SIZE 336 + +#define RX_BUFFER_ALIGN 64 +#define RX_BUFFER_ALIGN_MASK (~(RX_BUFFER_ALIGN - 1)) + +#define SKB_HEAD_ALIGN (((PAGE_SIZE - NET_SKB_PAD) % RX_BUFFER_ALIGN) + NET_SKB_PAD + NET_IP_ALIGN) +#define RX_SEGMENT_ALLOC_SIZE 4096 +#define RX_SEGMENT_BUFSIZE (SKB_WITH_OVERHEAD(RX_SEGMENT_ALLOC_SIZE)) +#define RX_SEGMENT_MRU (((RX_SEGMENT_BUFSIZE - SKB_HEAD_ALIGN) & RX_BUFFER_ALIGN_MASK) - NET_IP_ALIGN) +#define MAX_MTU 9500 + +#define NAPI_WEIGHT 64 + +/* MDIO Defines */ +#define MDIO_CMD_COMPLETE 0x00008000 +#define MDIO_WRITE_COMMAND 0x00002000 +#define MDIO_READ_COMMAND 0x00004000 +#define MDIO_REG_OFFSET 8 +#define MDIO_VALUE_OFFSET 16 + +/* Descritor Defines */ +#define END_OF_RING 0x40000000 +#define FIRST_SEGMENT 0x20000000 +#define LAST_SEGMENT 0x10000000 +#define FORCE_ROUTE 0x04000000 +#define IP_CHECKSUM 0x00040000 +#define UDP_CHECKSUM 0x00020000 +#define TCP_CHECKSUM 0x00010000 + +/* Port Config Defines */ +#define PORT_BP_ENABLE 0x00020000 +#define PORT_DISABLE 0x00040000 +#define PORT_LEARN_DIS 0x00080000 +#define PORT_BLOCK_STATE 0x00100000 +#define PORT_BLOCK_MODE 0x00200000 + +#define PROMISC_OFFSET 29 + +/* Global Config Defines */ +#define UNKNOWN_VLAN_TO_CPU 0x02000000 +#define ACCEPT_CRC_PACKET 0x00200000 +#define CRC_STRIPPING 0x00100000 + +/* VLAN Config Defines */ +#define NIC_MODE 0x00008000 +#define VLAN_UNAWARE 0x00000001 + +/* DMA AUTO Poll Defines */ +#define TS_POLL_EN 0x00000020 +#define TS_SUSPEND 0x00000010 +#define FS_POLL_EN 0x00000002 +#define FS_SUSPEND 0x00000001 + +/* DMA Ring Control Defines */ +#define QUEUE_THRESHOLD 0x000000f0 +#define CLR_FS_STATE 0x80000000 + +/* Interrupt Status Defines */ +#define MAC0_STATUS_CHANGE 0x00004000 +#define MAC1_STATUS_CHANGE 0x00008000 +#define MAC2_STATUS_CHANGE 0x00010000 +#define MAC0_RX_ERROR 0x00100000 +#define MAC1_RX_ERROR 0x00200000 +#define MAC2_RX_ERROR 0x00400000 + +struct tx_desc +{ + u32 sdp; /* segment data pointer */ + + union { + struct { + u32 sdl:16; /* segment data length */ + u32 tco:1; + u32 uco:1; + u32 ico:1; + u32 rsv_1:3; /* reserve */ + u32 pri:3; + u32 fp:1; /* force priority */ + u32 fr:1; + u32 interrupt:1; + u32 lsd:1; + u32 fsd:1; + u32 eor:1; + u32 cown:1; + }; + u32 config0; + }; + + union { + struct { + u32 ctv:1; + u32 stv:1; + u32 sid:4; + u32 inss:1; + u32 dels:1; + u32 rsv_2:9; + u32 pmap:5; + u32 mark:3; + u32 ewan:1; + u32 fewan:1; + u32 rsv_3:5; + }; + u32 config1; + }; + + union { + struct { + u32 c_vid:12; + u32 c_cfs:1; + u32 c_pri:3; + u32 s_vid:12; + u32 s_dei:1; + u32 s_pri:3; + }; + u32 config2; + }; + + u8 alignment[16]; /* for 32 byte */ +}; + +struct rx_desc +{ + u32 sdp; /* segment data pointer */ + + union { + struct { + u32 sdl:16; /* segment data length */ + u32 l4f:1; + u32 ipf:1; + u32 prot:4; + u32 hr:6; + u32 lsd:1; + u32 fsd:1; + u32 eor:1; + u32 cown:1; + }; + u32 config0; + }; + + union { + struct { + u32 ctv:1; + u32 stv:1; + u32 unv:1; + u32 iwan:1; + u32 exdv:1; + u32 e_wan:1; + u32 rsv_1:2; + u32 sp:3; + u32 crc_err:1; + u32 un_eth:1; + u32 tc:2; + u32 rsv_2:1; + u32 ip_offset:5; + u32 rsv_3:11; + }; + u32 config1; + }; + + union { + struct { + u32 c_vid:12; + u32 c_cfs:1; + u32 c_pri:3; + u32 s_vid:12; + u32 s_dei:1; + u32 s_pri:3; + }; + u32 config2; + }; + + u8 alignment[16]; /* for 32 byte alignment */ +}; + + +struct switch_regs { + u32 phy_control; + u32 phy_auto_addr; + u32 mac_glob_cfg; + u32 mac_cfg[4]; + u32 mac_pri_ctrl[5], __res; + u32 etype[2]; + u32 udp_range[4]; + u32 prio_etype_udp; + u32 prio_ipdscp[8]; + u32 tc_ctrl; + u32 rate_ctrl; + u32 fc_glob_thrs; + u32 fc_port_thrs; + u32 mc_fc_glob_thrs; + u32 dc_glob_thrs; + u32 arl_vlan_cmd; + u32 arl_ctrl[3]; + u32 vlan_cfg; + u32 pvid[2]; + u32 vlan_ctrl[3]; + u32 session_id[8]; + u32 intr_stat; + u32 intr_mask; + u32 sram_test; + u32 mem_queue; + u32 farl_ctrl; + u32 fc_input_thrs, __res1[2]; + u32 clk_skew_ctrl; + u32 mac_glob_cfg_ext, __res2[2]; + u32 dma_ring_ctrl; + u32 dma_auto_poll_cfg; + u32 delay_intr_cfg, __res3; + u32 ts_dma_ctrl0; + u32 ts_desc_ptr0; + u32 ts_desc_base_addr0, __res4; + u32 fs_dma_ctrl0; + u32 fs_desc_ptr0; + u32 fs_desc_base_addr0, __res5; + u32 ts_dma_ctrl1; + u32 ts_desc_ptr1; + u32 ts_desc_base_addr1, __res6; + u32 fs_dma_ctrl1; + u32 fs_desc_ptr1; + u32 fs_desc_base_addr1; + u32 __res7[109]; + u32 mac_counter0[13]; +}; + +struct _tx_ring { + struct tx_desc *desc; + dma_addr_t phys_addr; + struct tx_desc *cur_addr; + struct sk_buff *buff_tab[TX_DESCS]; + unsigned int phys_tab[TX_DESCS]; + u32 free_index; + u32 count_index; + u32 cur_index; + int num_used; + int num_count; + bool stopped; +}; + +struct _rx_ring { + struct rx_desc *desc; + dma_addr_t phys_addr; + struct rx_desc *cur_addr; + void *buff_tab[RX_DESCS]; + unsigned int phys_tab[RX_DESCS]; + u32 cur_index; + u32 alloc_index; + int alloc_count; +}; + +struct sw { + struct resource *mem_res; + struct switch_regs __iomem *regs; + struct napi_struct napi; + struct cns3xxx_plat_info *plat; + struct _tx_ring *tx_ring; + struct _rx_ring *rx_ring; + struct sk_buff *frag_first; + struct sk_buff *frag_last; +}; + +struct port { + struct net_device *netdev; + struct phy_device *phydev; + struct sw *sw; + int id; /* logical port ID */ + int speed, duplex; +}; + +static spinlock_t mdio_lock; +static DEFINE_SPINLOCK(tx_lock); +static struct switch_regs __iomem *mdio_regs; /* mdio command and status only */ +struct mii_bus *mdio_bus; +static int ports_open; +static struct port *switch_port_tab[4]; +static struct dma_pool *rx_dma_pool; +static struct dma_pool *tx_dma_pool; +struct net_device *napi_dev; + +static int cns3xxx_mdio_cmd(struct mii_bus *bus, int phy_id, int location, + int write, u16 cmd) +{ + int cycles = 0; + u32 temp = 0; + + temp = __raw_readl(&mdio_regs->phy_control); + temp |= MDIO_CMD_COMPLETE; + __raw_writel(temp, &mdio_regs->phy_control); + udelay(10); + + if (write) { + temp = (cmd << MDIO_VALUE_OFFSET); + temp |= MDIO_WRITE_COMMAND; + } else { + temp = MDIO_READ_COMMAND; + } + temp |= ((location & 0x1f) << MDIO_REG_OFFSET); + temp |= (phy_id & 0x1f); + + __raw_writel(temp, &mdio_regs->phy_control); + + while (((__raw_readl(&mdio_regs->phy_control) & MDIO_CMD_COMPLETE) == 0) + && cycles < 5000) { + udelay(1); + cycles++; + } + + if (cycles == 5000) { + printk(KERN_ERR "%s #%i: MII transaction failed\n", bus->name, + phy_id); + return -1; + } + + temp = __raw_readl(&mdio_regs->phy_control); + temp |= MDIO_CMD_COMPLETE; + __raw_writel(temp, &mdio_regs->phy_control); + + if (write) + return 0; + + return ((temp >> MDIO_VALUE_OFFSET) & 0xFFFF); +} + +static int cns3xxx_mdio_read(struct mii_bus *bus, int phy_id, int location) +{ + unsigned long flags; + int ret; + + spin_lock_irqsave(&mdio_lock, flags); + ret = cns3xxx_mdio_cmd(bus, phy_id, location, 0, 0); + spin_unlock_irqrestore(&mdio_lock, flags); + return ret; +} + +static int cns3xxx_mdio_write(struct mii_bus *bus, int phy_id, int location, + u16 val) +{ + unsigned long flags; + int ret; + + spin_lock_irqsave(&mdio_lock, flags); + ret = cns3xxx_mdio_cmd(bus, phy_id, location, 1, val); + spin_unlock_irqrestore(&mdio_lock, flags); + return ret; +} + +static int cns3xxx_mdio_register(void) +{ + int err; + + if (!(mdio_bus = mdiobus_alloc())) + return -ENOMEM; + + mdio_regs = (struct switch_regs __iomem *)CNS3XXX_SWITCH_BASE_VIRT; + + spin_lock_init(&mdio_lock); + mdio_bus->name = "CNS3xxx MII Bus"; + mdio_bus->read = &cns3xxx_mdio_read; + mdio_bus->write = &cns3xxx_mdio_write; + strcpy(mdio_bus->id, "0"); + + if ((err = mdiobus_register(mdio_bus))) + mdiobus_free(mdio_bus); + return err; +} + +static void cns3xxx_mdio_remove(void) +{ + mdiobus_unregister(mdio_bus); + mdiobus_free(mdio_bus); +} + +static void enable_tx_dma(struct sw *sw) +{ + __raw_writel(0x1, &sw->regs->ts_dma_ctrl0); +} + +static void enable_rx_dma(struct sw *sw) +{ + __raw_writel(0x1, &sw->regs->fs_dma_ctrl0); +} + +static void cns3xxx_adjust_link(struct net_device *dev) +{ + struct port *port = netdev_priv(dev); + struct phy_device *phydev = port->phydev; + + if (!phydev->link) { + if (port->speed) { + port->speed = 0; + printk(KERN_INFO "%s: link down\n", dev->name); + } + return; + } + + if (port->speed == phydev->speed && port->duplex == phydev->duplex) + return; + + port->speed = phydev->speed; + port->duplex = phydev->duplex; + + printk(KERN_INFO "%s: link up, speed %u Mb/s, %s duplex\n", + dev->name, port->speed, port->duplex ? "full" : "half"); +} + +irqreturn_t eth_rx_irq(int irq, void *pdev) +{ + struct net_device *dev = pdev; + struct sw *sw = netdev_priv(dev); + if (likely(napi_schedule_prep(&sw->napi))) { + disable_irq_nosync(IRQ_CNS3XXX_SW_R0RXC); + __napi_schedule(&sw->napi); + } + return (IRQ_HANDLED); +} + +irqreturn_t eth_stat_irq(int irq, void *pdev) +{ + struct net_device *dev = pdev; + struct sw *sw = netdev_priv(dev); + u32 cfg; + u32 stat = __raw_readl(&sw->regs->intr_stat); + __raw_writel(0xffffffff, &sw->regs->intr_stat); + + if (stat & MAC2_RX_ERROR) + switch_port_tab[3]->netdev->stats.rx_dropped++; + if (stat & MAC1_RX_ERROR) + switch_port_tab[1]->netdev->stats.rx_dropped++; + if (stat & MAC0_RX_ERROR) + switch_port_tab[0]->netdev->stats.rx_dropped++; + + if (stat & MAC0_STATUS_CHANGE) { + cfg = __raw_readl(&sw->regs->mac_cfg[0]); + switch_port_tab[0]->phydev->link = (cfg & 0x1); + switch_port_tab[0]->phydev->duplex = ((cfg >> 4) & 0x1); + if (((cfg >> 2) & 0x3) == 2) + switch_port_tab[0]->phydev->speed = 1000; + else if (((cfg >> 2) & 0x3) == 1) + switch_port_tab[0]->phydev->speed = 100; + else + switch_port_tab[0]->phydev->speed = 10; + cns3xxx_adjust_link(switch_port_tab[0]->netdev); + } + + if (stat & MAC1_STATUS_CHANGE) { + cfg = __raw_readl(&sw->regs->mac_cfg[1]); + switch_port_tab[1]->phydev->link = (cfg & 0x1); + switch_port_tab[1]->phydev->duplex = ((cfg >> 4) & 0x1); + if (((cfg >> 2) & 0x3) == 2) + switch_port_tab[1]->phydev->speed = 1000; + else if (((cfg >> 2) & 0x3) == 1) + switch_port_tab[1]->phydev->speed = 100; + else + switch_port_tab[1]->phydev->speed = 10; + cns3xxx_adjust_link(switch_port_tab[1]->netdev); + } + + if (stat & MAC2_STATUS_CHANGE) { + cfg = __raw_readl(&sw->regs->mac_cfg[3]); + switch_port_tab[3]->phydev->link = (cfg & 0x1); + switch_port_tab[3]->phydev->duplex = ((cfg >> 4) & 0x1); + if (((cfg >> 2) & 0x3) == 2) + switch_port_tab[3]->phydev->speed = 1000; + else if (((cfg >> 2) & 0x3) == 1) + switch_port_tab[3]->phydev->speed = 100; + else + switch_port_tab[3]->phydev->speed = 10; + cns3xxx_adjust_link(switch_port_tab[3]->netdev); + } + + return (IRQ_HANDLED); +} + + +static void cns3xxx_alloc_rx_buf(struct sw *sw, int received) +{ + struct _rx_ring *rx_ring = sw->rx_ring; + unsigned int i = rx_ring->alloc_index; + struct rx_desc *desc = &(rx_ring)->desc[i]; + void *buf; + unsigned int phys; + + for (received += rx_ring->alloc_count; received > 0; received--) { + buf = kzalloc(RX_SEGMENT_ALLOC_SIZE, GFP_ATOMIC); + if (!buf) + goto out; + + phys = dma_map_single(NULL, buf + SKB_HEAD_ALIGN, + RX_SEGMENT_MRU, DMA_FROM_DEVICE); + if (dma_mapping_error(NULL, phys)) { + kfree(buf); + goto out; + } + + desc->sdl = RX_SEGMENT_MRU; + desc->sdp = phys; + + /* put the new buffer on RX-free queue */ + rx_ring->buff_tab[i] = buf; + rx_ring->phys_tab[i] = phys; + if (i == RX_DESCS - 1) { + i = 0; + desc->config0 = END_OF_RING | FIRST_SEGMENT | + LAST_SEGMENT | RX_SEGMENT_MRU; + desc = &(rx_ring)->desc[i]; + } else { + desc->config0 = FIRST_SEGMENT | LAST_SEGMENT | + RX_SEGMENT_MRU; + i++; + desc++; + } + } +out: + rx_ring->alloc_count = received; + rx_ring->alloc_index = i; +} + +static void eth_check_num_used(struct _tx_ring *tx_ring) +{ + bool stop = false; + int i; + + if (tx_ring->num_used >= TX_DESCS - TX_DESC_RESERVE) + stop = true; + + if (tx_ring->stopped == stop) + return; + + tx_ring->stopped = stop; + for (i = 0; i < 4; i++) { + struct port *port = switch_port_tab[i]; + struct net_device *dev; + + if (!port) + continue; + + dev = port->netdev; + if (stop) + netif_stop_queue(dev); + else + netif_wake_queue(dev); + } +} + +static void eth_complete_tx(struct sw *sw) +{ + struct _tx_ring *tx_ring = sw->tx_ring; + struct tx_desc *desc; + int i; + int index; + int num_used = tx_ring->num_used; + struct sk_buff *skb; + + index = tx_ring->free_index; + desc = &(tx_ring)->desc[index]; + for (i = 0; i < num_used; i++) { + if (desc->cown) { + skb = tx_ring->buff_tab[index]; + tx_ring->buff_tab[index] = 0; + if (skb) + dev_kfree_skb_any(skb); + dma_unmap_single(NULL, tx_ring->phys_tab[index], + desc->sdl, DMA_TO_DEVICE); + if (++index == TX_DESCS) { + index = 0; + desc = &(tx_ring)->desc[index]; + } else { + desc++; + } + } else { + break; + } + } + tx_ring->free_index = index; + tx_ring->num_used -= i; + eth_check_num_used(tx_ring); +} + +static int eth_poll(struct napi_struct *napi, int budget) +{ + struct sw *sw = container_of(napi, struct sw, napi); + struct _rx_ring *rx_ring = sw->rx_ring; + int received = 0; + unsigned int length; + unsigned int i = rx_ring->cur_index; + struct rx_desc *desc = &(rx_ring)->desc[i]; + + while (desc->cown) { + struct sk_buff *skb; + int reserve = SKB_HEAD_ALIGN; + + if (received >= budget) + break; + + /* process received frame */ + dma_unmap_single(NULL, rx_ring->phys_tab[i], + RX_SEGMENT_MRU, DMA_FROM_DEVICE); + + skb = build_skb(rx_ring->buff_tab[i]); + if (!skb) + break; + + skb->dev = switch_port_tab[desc->sp]->netdev; + + length = desc->sdl; + if (desc->fsd && !desc->lsd) + length = RX_SEGMENT_MRU; + + if (!desc->fsd) { + reserve -= NET_IP_ALIGN; + if (!desc->lsd) + length += NET_IP_ALIGN; + } + + skb_reserve(skb, reserve); + skb_put(skb, length); + + if (!sw->frag_first) + sw->frag_first = skb; + else { + if (sw->frag_first == sw->frag_last) + skb_frag_add_head(sw->frag_first, skb); + else + sw->frag_last->next = skb; + sw->frag_first->len += skb->len; + sw->frag_first->data_len += skb->len; + sw->frag_first->truesize += skb->truesize; + } + sw->frag_last = skb; + + if (desc->lsd) { + struct net_device *dev; + + skb = sw->frag_first; + dev = skb->dev; + skb->protocol = eth_type_trans(skb, dev); + + dev->stats.rx_packets++; + dev->stats.rx_bytes += skb->len; + + /* RX Hardware checksum offload */ + skb->ip_summed = CHECKSUM_NONE; + switch (desc->prot) { + case 1: + case 2: + case 5: + case 6: + case 13: + case 14: + if (!desc->l4f) { + skb->ip_summed = CHECKSUM_UNNECESSARY; + napi_gro_receive(napi, skb); + break; + } + /* fall through */ + default: + netif_receive_skb(skb); + break; + } + + sw->frag_first = NULL; + sw->frag_last = NULL; + } + + received++; + if (++i == RX_DESCS) { + i = 0; + desc = &(rx_ring)->desc[i]; + } else { + desc++; + } + } + + cns3xxx_alloc_rx_buf(sw, received); + + rx_ring->cur_index = i; + + if (received != budget) { + napi_complete(napi); + enable_irq(IRQ_CNS3XXX_SW_R0RXC); + } + + enable_rx_dma(sw); + + spin_lock_bh(&tx_lock); + eth_complete_tx(sw); + spin_unlock_bh(&tx_lock); + + return received; +} + +static void eth_set_desc(struct _tx_ring *tx_ring, int index, int index_last, + void *data, int len, u32 config0, u32 pmap) +{ + struct tx_desc *tx_desc = &(tx_ring)->desc[index]; + unsigned int phys; + + phys = dma_map_single(NULL, data, len, DMA_TO_DEVICE); + tx_desc->sdp = phys; + tx_desc->pmap = pmap; + tx_ring->phys_tab[index] = phys; + + config0 |= len; + if (index == TX_DESCS - 1) + config0 |= END_OF_RING; + if (index == index_last) + config0 |= LAST_SEGMENT; + + mb(); + tx_desc->config0 = config0; +} + +static int eth_xmit(struct sk_buff *skb, struct net_device *dev) +{ + struct port *port = netdev_priv(dev); + struct sw *sw = port->sw; + struct _tx_ring *tx_ring = sw->tx_ring; + struct sk_buff *skb1; + char pmap = (1 << port->id); + int nr_frags = skb_shinfo(skb)->nr_frags; + int nr_desc = nr_frags; + int index0, index, index_last; + int len0; + unsigned int i; + u32 config0; + + if (pmap == 8) + pmap = (1 << 4); + + skb_walk_frags(skb, skb1) + nr_desc++; + + spin_lock_bh(&tx_lock); + + eth_complete_tx(sw); + if ((tx_ring->num_used + nr_desc + 1) >= TX_DESCS) { + spin_unlock_bh(&tx_lock); + return NETDEV_TX_BUSY; + } + + index = index0 = tx_ring->cur_index; + index_last = (index0 + nr_desc) % TX_DESCS; + tx_ring->cur_index = (index_last + 1) % TX_DESCS; + + spin_unlock_bh(&tx_lock); + + config0 = FORCE_ROUTE; + if (skb->ip_summed == CHECKSUM_PARTIAL) + config0 |= UDP_CHECKSUM | TCP_CHECKSUM; + + len0 = skb->len; + + /* fragments */ + for (i = 0; i < nr_frags; i++) { + struct skb_frag_struct *frag; + void *addr; + + index = (index + 1) % TX_DESCS; + + frag = &skb_shinfo(skb)->frags[i]; + addr = page_address(skb_frag_page(frag)) + frag->page_offset; + + eth_set_desc(tx_ring, index, index_last, addr, frag->size, + config0, pmap); + } + + if (nr_frags) + len0 = skb->len - skb->data_len; + + skb_walk_frags(skb, skb1) { + index = (index + 1) % TX_DESCS; + len0 -= skb1->len; + + eth_set_desc(tx_ring, index, index_last, skb1->data, skb1->len, + config0, pmap); + } + + tx_ring->buff_tab[index0] = skb; + eth_set_desc(tx_ring, index0, index_last, skb->data, len0, + config0 | FIRST_SEGMENT, pmap); + + mb(); + + spin_lock(&tx_lock); + tx_ring->num_used += nr_desc + 1; + spin_unlock(&tx_lock); + + dev->stats.tx_packets++; + dev->stats.tx_bytes += skb->len; + + enable_tx_dma(sw); + + return NETDEV_TX_OK; +} + +static int eth_ioctl(struct net_device *dev, struct ifreq *req, int cmd) +{ + struct port *port = netdev_priv(dev); + + if (!netif_running(dev)) + return -EINVAL; + return phy_mii_ioctl(port->phydev, req, cmd); +} + +/* ethtool support */ + +static void cns3xxx_get_drvinfo(struct net_device *dev, + struct ethtool_drvinfo *info) +{ + strcpy(info->driver, DRV_NAME); + strcpy(info->bus_info, "internal"); +} + +static int cns3xxx_get_settings(struct net_device *dev, struct ethtool_cmd *cmd) +{ + struct port *port = netdev_priv(dev); + return phy_ethtool_gset(port->phydev, cmd); +} + +static int cns3xxx_set_settings(struct net_device *dev, struct ethtool_cmd *cmd) +{ + struct port *port = netdev_priv(dev); + return phy_ethtool_sset(port->phydev, cmd); +} + +static int cns3xxx_nway_reset(struct net_device *dev) +{ + struct port *port = netdev_priv(dev); + return phy_start_aneg(port->phydev); +} + +static struct ethtool_ops cns3xxx_ethtool_ops = { + .get_drvinfo = cns3xxx_get_drvinfo, + .get_settings = cns3xxx_get_settings, + .set_settings = cns3xxx_set_settings, + .nway_reset = cns3xxx_nway_reset, + .get_link = ethtool_op_get_link, +}; + + +static int init_rings(struct sw *sw) +{ + int i; + struct _rx_ring *rx_ring = sw->rx_ring; + struct _tx_ring *tx_ring = sw->tx_ring; + + __raw_writel(0, &sw->regs->fs_dma_ctrl0); + __raw_writel(TS_SUSPEND | FS_SUSPEND, &sw->regs->dma_auto_poll_cfg); + __raw_writel(QUEUE_THRESHOLD, &sw->regs->dma_ring_ctrl); + __raw_writel(CLR_FS_STATE | QUEUE_THRESHOLD, &sw->regs->dma_ring_ctrl); + + __raw_writel(QUEUE_THRESHOLD, &sw->regs->dma_ring_ctrl); + + if (!(rx_dma_pool = dma_pool_create(DRV_NAME, NULL, + RX_POOL_ALLOC_SIZE, 32, 0))) + return -ENOMEM; + + if (!(rx_ring->desc = dma_pool_alloc(rx_dma_pool, GFP_KERNEL, + &rx_ring->phys_addr))) + return -ENOMEM; + memset(rx_ring->desc, 0, RX_POOL_ALLOC_SIZE); + + /* Setup RX buffers */ + for (i = 0; i < RX_DESCS; i++) { + struct rx_desc *desc = &(rx_ring)->desc[i]; + void *buf; + + buf = kzalloc(RX_SEGMENT_ALLOC_SIZE, GFP_KERNEL); + if (!buf) + return -ENOMEM; + + desc->sdl = RX_SEGMENT_MRU; + if (i == (RX_DESCS - 1)) + desc->eor = 1; + desc->fsd = 1; + desc->lsd = 1; + + desc->sdp = dma_map_single(NULL, buf + SKB_HEAD_ALIGN, + RX_SEGMENT_MRU, DMA_FROM_DEVICE); + if (dma_mapping_error(NULL, desc->sdp)) + return -EIO; + + rx_ring->buff_tab[i] = buf; + rx_ring->phys_tab[i] = desc->sdp; + desc->cown = 0; + } + __raw_writel(rx_ring->phys_addr, &sw->regs->fs_desc_ptr0); + __raw_writel(rx_ring->phys_addr, &sw->regs->fs_desc_base_addr0); + + if (!(tx_dma_pool = dma_pool_create(DRV_NAME, NULL, + TX_POOL_ALLOC_SIZE, 32, 0))) + return -ENOMEM; + + if (!(tx_ring->desc = dma_pool_alloc(tx_dma_pool, GFP_KERNEL, + &tx_ring->phys_addr))) + return -ENOMEM; + memset(tx_ring->desc, 0, TX_POOL_ALLOC_SIZE); + + /* Setup TX buffers */ + for (i = 0; i < TX_DESCS; i++) { + struct tx_desc *desc = &(tx_ring)->desc[i]; + tx_ring->buff_tab[i] = 0; + + if (i == (TX_DESCS - 1)) + desc->eor = 1; + desc->cown = 1; + } + __raw_writel(tx_ring->phys_addr, &sw->regs->ts_desc_ptr0); + __raw_writel(tx_ring->phys_addr, &sw->regs->ts_desc_base_addr0); + + return 0; +} + +static void destroy_rings(struct sw *sw) +{ + int i; + if (sw->rx_ring->desc) { + for (i = 0; i < RX_DESCS; i++) { + struct _rx_ring *rx_ring = sw->rx_ring; + struct rx_desc *desc = &(rx_ring)->desc[i]; + struct sk_buff *skb = sw->rx_ring->buff_tab[i]; + + if (!skb) + continue; + + dma_unmap_single(NULL, desc->sdp, RX_SEGMENT_MRU, + DMA_FROM_DEVICE); + dev_kfree_skb(skb); + } + dma_pool_free(rx_dma_pool, sw->rx_ring->desc, sw->rx_ring->phys_addr); + dma_pool_destroy(rx_dma_pool); + rx_dma_pool = 0; + sw->rx_ring->desc = 0; + } + if (sw->tx_ring->desc) { + for (i = 0; i < TX_DESCS; i++) { + struct _tx_ring *tx_ring = sw->tx_ring; + struct tx_desc *desc = &(tx_ring)->desc[i]; + struct sk_buff *skb = sw->tx_ring->buff_tab[i]; + if (skb) { + dma_unmap_single(NULL, desc->sdp, + skb->len, DMA_TO_DEVICE); + dev_kfree_skb(skb); + } + } + dma_pool_free(tx_dma_pool, sw->tx_ring->desc, sw->tx_ring->phys_addr); + dma_pool_destroy(tx_dma_pool); + tx_dma_pool = 0; + sw->tx_ring->desc = 0; + } +} + +static int eth_open(struct net_device *dev) +{ + struct port *port = netdev_priv(dev); + struct sw *sw = port->sw; + u32 temp; + + port->speed = 0; /* force "link up" message */ + phy_start(port->phydev); + + netif_start_queue(dev); + + if (!ports_open) { + request_irq(IRQ_CNS3XXX_SW_R0RXC, eth_rx_irq, IRQF_SHARED, "gig_switch", napi_dev); + request_irq(IRQ_CNS3XXX_SW_STATUS, eth_stat_irq, IRQF_SHARED, "gig_stat", napi_dev); + napi_enable(&sw->napi); + netif_start_queue(napi_dev); + + __raw_writel(~(MAC0_STATUS_CHANGE | MAC1_STATUS_CHANGE | MAC2_STATUS_CHANGE | + MAC0_RX_ERROR | MAC1_RX_ERROR | MAC2_RX_ERROR), &sw->regs->intr_mask); + + temp = __raw_readl(&sw->regs->mac_cfg[2]); + temp &= ~(PORT_DISABLE); + __raw_writel(temp, &sw->regs->mac_cfg[2]); + + temp = __raw_readl(&sw->regs->dma_auto_poll_cfg); + temp &= ~(TS_SUSPEND | FS_SUSPEND); + __raw_writel(temp, &sw->regs->dma_auto_poll_cfg); + + enable_rx_dma(sw); + } + temp = __raw_readl(&sw->regs->mac_cfg[port->id]); + temp &= ~(PORT_DISABLE); + __raw_writel(temp, &sw->regs->mac_cfg[port->id]); + + ports_open++; + netif_carrier_on(dev); + + return 0; +} + +static int eth_close(struct net_device *dev) +{ + struct port *port = netdev_priv(dev); + struct sw *sw = port->sw; + u32 temp; + + ports_open--; + + temp = __raw_readl(&sw->regs->mac_cfg[port->id]); + temp |= (PORT_DISABLE); + __raw_writel(temp, &sw->regs->mac_cfg[port->id]); + + netif_stop_queue(dev); + + phy_stop(port->phydev); + + if (!ports_open) { + disable_irq(IRQ_CNS3XXX_SW_R0RXC); + free_irq(IRQ_CNS3XXX_SW_R0RXC, napi_dev); + disable_irq(IRQ_CNS3XXX_SW_STATUS); + free_irq(IRQ_CNS3XXX_SW_STATUS, napi_dev); + napi_disable(&sw->napi); + netif_stop_queue(napi_dev); + temp = __raw_readl(&sw->regs->mac_cfg[2]); + temp |= (PORT_DISABLE); + __raw_writel(temp, &sw->regs->mac_cfg[2]); + + __raw_writel(TS_SUSPEND | FS_SUSPEND, + &sw->regs->dma_auto_poll_cfg); + } + + netif_carrier_off(dev); + return 0; +} + +static void eth_rx_mode(struct net_device *dev) +{ + struct port *port = netdev_priv(dev); + struct sw *sw = port->sw; + u32 temp; + + temp = __raw_readl(&sw->regs->mac_glob_cfg); + + if (dev->flags & IFF_PROMISC) { + if (port->id == 3) + temp |= ((1 << 2) << PROMISC_OFFSET); + else + temp |= ((1 << port->id) << PROMISC_OFFSET); + } else { + if (port->id == 3) + temp &= ~((1 << 2) << PROMISC_OFFSET); + else + temp &= ~((1 << port->id) << PROMISC_OFFSET); + } + __raw_writel(temp, &sw->regs->mac_glob_cfg); +} + +static int eth_set_mac(struct net_device *netdev, void *p) +{ + struct port *port = netdev_priv(netdev); + struct sw *sw = port->sw; + struct sockaddr *addr = p; + u32 cycles = 0; + + if (!is_valid_ether_addr(addr->sa_data)) + return -EADDRNOTAVAIL; + + /* Invalidate old ARL Entry */ + if (port->id == 3) + __raw_writel((port->id << 16) | (0x4 << 9), &sw->regs->arl_ctrl[0]); + else + __raw_writel(((port->id + 1) << 16) | (0x4 << 9), &sw->regs->arl_ctrl[0]); + __raw_writel( ((netdev->dev_addr[0] << 24) | (netdev->dev_addr[1] << 16) | + (netdev->dev_addr[2] << 8) | (netdev->dev_addr[3])), + &sw->regs->arl_ctrl[1]); + + __raw_writel( ((netdev->dev_addr[4] << 24) | (netdev->dev_addr[5] << 16) | + (1 << 1)), + &sw->regs->arl_ctrl[2]); + __raw_writel((1 << 19), &sw->regs->arl_vlan_cmd); + + while (((__raw_readl(&sw->regs->arl_vlan_cmd) & (1 << 21)) == 0) + && cycles < 5000) { + udelay(1); + cycles++; + } + + cycles = 0; + memcpy(netdev->dev_addr, addr->sa_data, netdev->addr_len); + + if (port->id == 3) + __raw_writel((port->id << 16) | (0x4 << 9), &sw->regs->arl_ctrl[0]); + else + __raw_writel(((port->id + 1) << 16) | (0x4 << 9), &sw->regs->arl_ctrl[0]); + __raw_writel( ((addr->sa_data[0] << 24) | (addr->sa_data[1] << 16) | + (addr->sa_data[2] << 8) | (addr->sa_data[3])), + &sw->regs->arl_ctrl[1]); + + __raw_writel( ((addr->sa_data[4] << 24) | (addr->sa_data[5] << 16) | + (7 << 4) | (1 << 1)), &sw->regs->arl_ctrl[2]); + __raw_writel((1 << 19), &sw->regs->arl_vlan_cmd); + + while (((__raw_readl(&sw->regs->arl_vlan_cmd) & (1 << 21)) == 0) + && cycles < 5000) { + udelay(1); + cycles++; + } + return 0; +} + +static int cns3xxx_change_mtu(struct net_device *dev, int new_mtu) +{ + if (new_mtu > MAX_MTU) + return -EINVAL; + + dev->mtu = new_mtu; + return 0; +} + +static const struct net_device_ops cns3xxx_netdev_ops = { + .ndo_open = eth_open, + .ndo_stop = eth_close, + .ndo_start_xmit = eth_xmit, + .ndo_set_rx_mode = eth_rx_mode, + .ndo_do_ioctl = eth_ioctl, + .ndo_change_mtu = cns3xxx_change_mtu, + .ndo_set_mac_address = eth_set_mac, + .ndo_validate_addr = eth_validate_addr, +}; + +static int __devinit eth_init_one(struct platform_device *pdev) +{ + int i; + struct port *port; + struct sw *sw; + struct net_device *dev; + struct cns3xxx_plat_info *plat = pdev->dev.platform_data; + u32 regs_phys; + char phy_id[MII_BUS_ID_SIZE + 3]; + int err; + u32 temp; + + if (!(napi_dev = alloc_etherdev(sizeof(struct sw)))) + return -ENOMEM; + strcpy(napi_dev->name, "switch%d"); + napi_dev->features = NETIF_F_IP_CSUM | NETIF_F_SG | NETIF_F_FRAGLIST; + + SET_NETDEV_DEV(napi_dev, &pdev->dev); + sw = netdev_priv(napi_dev); + memset(sw, 0, sizeof(struct sw)); + sw->regs = (struct switch_regs __iomem *)CNS3XXX_SWITCH_BASE_VIRT; + regs_phys = CNS3XXX_SWITCH_BASE; + sw->mem_res = request_mem_region(regs_phys, REGS_SIZE, napi_dev->name); + if (!sw->mem_res) { + err = -EBUSY; + goto err_free; + } + + temp = __raw_readl(&sw->regs->phy_auto_addr); + temp |= (3 << 30); /* maximum frame length: 9600 bytes */ + __raw_writel(temp, &sw->regs->phy_auto_addr); + + for (i = 0; i < 4; i++) { + temp = __raw_readl(&sw->regs->mac_cfg[i]); + temp |= (PORT_DISABLE); + __raw_writel(temp, &sw->regs->mac_cfg[i]); + } + + temp = PORT_DISABLE; + __raw_writel(temp, &sw->regs->mac_cfg[2]); + + temp = __raw_readl(&sw->regs->vlan_cfg); + temp |= NIC_MODE | VLAN_UNAWARE; + __raw_writel(temp, &sw->regs->vlan_cfg); + + __raw_writel(UNKNOWN_VLAN_TO_CPU | + CRC_STRIPPING, &sw->regs->mac_glob_cfg); + + if (!(sw->rx_ring = kmalloc(sizeof(struct _rx_ring), GFP_KERNEL))) { + err = -ENOMEM; + goto err_free; + } + memset(sw->rx_ring, 0, sizeof(struct _rx_ring)); + + if (!(sw->tx_ring = kmalloc(sizeof(struct _tx_ring), GFP_KERNEL))) { + err = -ENOMEM; + goto err_free_rx; + } + memset(sw->tx_ring, 0, sizeof(struct _tx_ring)); + + if ((err = init_rings(sw)) != 0) { + destroy_rings(sw); + err = -ENOMEM; + goto err_free_rings; + } + platform_set_drvdata(pdev, napi_dev); + + netif_napi_add(napi_dev, &sw->napi, eth_poll, NAPI_WEIGHT); + + for (i = 0; i < 3; i++) { + if (!(plat->ports & (1 << i))) { + continue; + } + + if (!(dev = alloc_etherdev(sizeof(struct port)))) { + goto free_ports; + } + + port = netdev_priv(dev); + port->netdev = dev; + if (i == 2) + port->id = 3; + else + port->id = i; + port->sw = sw; + + temp = __raw_readl(&sw->regs->mac_cfg[port->id]); + temp |= (PORT_DISABLE | PORT_BLOCK_STATE | PORT_LEARN_DIS); + __raw_writel(temp, &sw->regs->mac_cfg[port->id]); + + dev->netdev_ops = &cns3xxx_netdev_ops; + dev->ethtool_ops = &cns3xxx_ethtool_ops; + dev->tx_queue_len = 1000; + dev->features = NETIF_F_IP_CSUM | NETIF_F_SG | NETIF_F_FRAGLIST; + + switch_port_tab[port->id] = port; + memcpy(dev->dev_addr, &plat->hwaddr[i], ETH_ALEN); + + snprintf(phy_id, MII_BUS_ID_SIZE + 3, PHY_ID_FMT, "0", plat->phy[i]); + port->phydev = phy_connect(dev, phy_id, &cns3xxx_adjust_link, 0, + PHY_INTERFACE_MODE_RGMII); + if ((err = IS_ERR(port->phydev))) { + switch_port_tab[port->id] = 0; + free_netdev(dev); + goto free_ports; + } + + port->phydev->irq = PHY_IGNORE_INTERRUPT; + + if ((err = register_netdev(dev))) { + phy_disconnect(port->phydev); + switch_port_tab[port->id] = 0; + free_netdev(dev); + goto free_ports; + } + + printk(KERN_INFO "%s: RGMII PHY %i on cns3xxx Switch\n", dev->name, plat->phy[i]); + netif_carrier_off(dev); + dev = 0; + } + + return 0; + +free_ports: + err = -ENOMEM; + for (--i; i >= 0; i--) { + if (switch_port_tab[i]) { + port = switch_port_tab[i]; + dev = port->netdev; + unregister_netdev(dev); + phy_disconnect(port->phydev); + switch_port_tab[i] = 0; + free_netdev(dev); + } + } +err_free_rings: + kfree(sw->tx_ring); +err_free_rx: + kfree(sw->rx_ring); +err_free: + free_netdev(napi_dev); + return err; +} + +static int __devexit eth_remove_one(struct platform_device *pdev) +{ + struct net_device *dev = platform_get_drvdata(pdev); + struct sw *sw = netdev_priv(dev); + int i; + destroy_rings(sw); + + for (i = 3; i >= 0; i--) { + if (switch_port_tab[i]) { + struct port *port = switch_port_tab[i]; + struct net_device *dev = port->netdev; + unregister_netdev(dev); + phy_disconnect(port->phydev); + switch_port_tab[i] = 0; + free_netdev(dev); + } + } + + release_resource(sw->mem_res); + free_netdev(napi_dev); + return 0; +} + +static struct platform_driver cns3xxx_eth_driver = { + .driver.name = DRV_NAME, + .probe = eth_init_one, + .remove = eth_remove_one, +}; + +static int __init eth_init_module(void) +{ + int err; + if ((err = cns3xxx_mdio_register())) + return err; + return platform_driver_register(&cns3xxx_eth_driver); +} + +static void __exit eth_cleanup_module(void) +{ + platform_driver_unregister(&cns3xxx_eth_driver); + cns3xxx_mdio_remove(); +} + +module_init(eth_init_module); +module_exit(eth_cleanup_module); + +MODULE_AUTHOR("Chris Lang"); +MODULE_DESCRIPTION("Cavium CNS3xxx Ethernet driver"); +MODULE_LICENSE("GPL v2"); +MODULE_ALIAS("platform:cns3xxx_eth"); diff --git a/target/linux/cns3xxx/patches-3.3/051-cns3xxx_gigabit.patch b/target/linux/cns3xxx/patches-3.3/051-cns3xxx_gigabit.patch deleted file mode 100644 index 439403ac0e..0000000000 --- a/target/linux/cns3xxx/patches-3.3/051-cns3xxx_gigabit.patch +++ /dev/null @@ -1,1356 +0,0 @@ ---- /dev/null -+++ b/drivers/net/ethernet/cavium/cns3xxx_eth.c -@@ -0,0 +1,1298 @@ -+/* -+ * Cavium CNS3xxx Gigabit driver for Linux -+ * -+ * Copyright 2011 Gateworks Corporation -+ * Chris Lang <clang@gateworks.com> -+ * -+ * This program is free software; you can redistribute it and/or modify it -+ * under the terms of version 2 of the GNU General Public License -+ * as published by the Free Software Foundation. -+ * -+ */ -+ -+#include <linux/delay.h> -+#include <linux/module.h> -+#include <linux/dma-mapping.h> -+#include <linux/dmapool.h> -+#include <linux/etherdevice.h> -+#include <linux/interrupt.h> -+#include <linux/io.h> -+#include <linux/kernel.h> -+#include <linux/phy.h> -+#include <linux/platform_device.h> -+#include <linux/skbuff.h> -+#include <mach/irqs.h> -+#include <mach/platform.h> -+ -+#define DRV_NAME "cns3xxx_eth" -+ -+#define RX_DESCS 512 -+#define TX_DESCS 512 -+#define SKB_DMA_REALIGN ((PAGE_SIZE - NET_SKB_PAD) % SMP_CACHE_BYTES) -+ -+#define RX_POOL_ALLOC_SIZE (sizeof(struct rx_desc) * RX_DESCS) -+#define TX_POOL_ALLOC_SIZE (sizeof(struct tx_desc) * TX_DESCS) -+#define REGS_SIZE 336 -+#define MAX_MRU (1536 + SKB_DMA_REALIGN) -+#define CNS3XXX_MAX_MTU (1536) -+ -+#define NAPI_WEIGHT 64 -+ -+/* MDIO Defines */ -+#define MDIO_CMD_COMPLETE 0x00008000 -+#define MDIO_WRITE_COMMAND 0x00002000 -+#define MDIO_READ_COMMAND 0x00004000 -+#define MDIO_REG_OFFSET 8 -+#define MDIO_VALUE_OFFSET 16 -+ -+/* Descritor Defines */ -+#define END_OF_RING 0x40000000 -+#define FIRST_SEGMENT 0x20000000 -+#define LAST_SEGMENT 0x10000000 -+#define FORCE_ROUTE 0x04000000 -+#define IP_CHECKSUM 0x00040000 -+#define UDP_CHECKSUM 0x00020000 -+#define TCP_CHECKSUM 0x00010000 -+ -+/* Port Config Defines */ -+#define PORT_BP_ENABLE 0x00020000 -+#define PORT_DISABLE 0x00040000 -+#define PORT_LEARN_DIS 0x00080000 -+#define PORT_BLOCK_STATE 0x00100000 -+#define PORT_BLOCK_MODE 0x00200000 -+ -+#define PROMISC_OFFSET 29 -+ -+/* Global Config Defines */ -+#define UNKNOWN_VLAN_TO_CPU 0x02000000 -+#define ACCEPT_CRC_PACKET 0x00200000 -+#define CRC_STRIPPING 0x00100000 -+ -+/* VLAN Config Defines */ -+#define NIC_MODE 0x00008000 -+#define VLAN_UNAWARE 0x00000001 -+ -+/* DMA AUTO Poll Defines */ -+#define TS_POLL_EN 0x00000020 -+#define TS_SUSPEND 0x00000010 -+#define FS_POLL_EN 0x00000002 -+#define FS_SUSPEND 0x00000001 -+ -+/* DMA Ring Control Defines */ -+#define QUEUE_THRESHOLD 0x000000f0 -+#define CLR_FS_STATE 0x80000000 -+ -+/* Interrupt Status Defines */ -+#define MAC0_STATUS_CHANGE 0x00004000 -+#define MAC1_STATUS_CHANGE 0x00008000 -+#define MAC2_STATUS_CHANGE 0x00010000 -+#define MAC0_RX_ERROR 0x00100000 -+#define MAC1_RX_ERROR 0x00200000 -+#define MAC2_RX_ERROR 0x00400000 -+ -+struct tx_desc -+{ -+ u32 sdp; /* segment data pointer */ -+ -+ union { -+ struct { -+ u32 sdl:16; /* segment data length */ -+ u32 tco:1; -+ u32 uco:1; -+ u32 ico:1; -+ u32 rsv_1:3; /* reserve */ -+ u32 pri:3; -+ u32 fp:1; /* force priority */ -+ u32 fr:1; -+ u32 interrupt:1; -+ u32 lsd:1; -+ u32 fsd:1; -+ u32 eor:1; -+ u32 cown:1; -+ }; -+ u32 config0; -+ }; -+ -+ union { -+ struct { -+ u32 ctv:1; -+ u32 stv:1; -+ u32 sid:4; -+ u32 inss:1; -+ u32 dels:1; -+ u32 rsv_2:9; -+ u32 pmap:5; -+ u32 mark:3; -+ u32 ewan:1; -+ u32 fewan:1; -+ u32 rsv_3:5; -+ }; -+ u32 config1; -+ }; -+ -+ union { -+ struct { -+ u32 c_vid:12; -+ u32 c_cfs:1; -+ u32 c_pri:3; -+ u32 s_vid:12; -+ u32 s_dei:1; -+ u32 s_pri:3; -+ }; -+ u32 config2; -+ }; -+ -+ u8 alignment[16]; /* for 32 byte */ -+}; -+ -+struct rx_desc -+{ -+ u32 sdp; /* segment data pointer */ -+ -+ union { -+ struct { -+ u32 sdl:16; /* segment data length */ -+ u32 l4f:1; -+ u32 ipf:1; -+ u32 prot:4; -+ u32 hr:6; -+ u32 lsd:1; -+ u32 fsd:1; -+ u32 eor:1; -+ u32 cown:1; -+ }; -+ u32 config0; -+ }; -+ -+ union { -+ struct { -+ u32 ctv:1; -+ u32 stv:1; -+ u32 unv:1; -+ u32 iwan:1; -+ u32 exdv:1; -+ u32 e_wan:1; -+ u32 rsv_1:2; -+ u32 sp:3; -+ u32 crc_err:1; -+ u32 un_eth:1; -+ u32 tc:2; -+ u32 rsv_2:1; -+ u32 ip_offset:5; -+ u32 rsv_3:11; -+ }; -+ u32 config1; -+ }; -+ -+ union { -+ struct { -+ u32 c_vid:12; -+ u32 c_cfs:1; -+ u32 c_pri:3; -+ u32 s_vid:12; -+ u32 s_dei:1; -+ u32 s_pri:3; -+ }; -+ u32 config2; -+ }; -+ -+ u8 alignment[16]; /* for 32 byte alignment */ -+}; -+ -+ -+struct switch_regs { -+ u32 phy_control; -+ u32 phy_auto_addr; -+ u32 mac_glob_cfg; -+ u32 mac_cfg[4]; -+ u32 mac_pri_ctrl[5], __res; -+ u32 etype[2]; -+ u32 udp_range[4]; -+ u32 prio_etype_udp; -+ u32 prio_ipdscp[8]; -+ u32 tc_ctrl; -+ u32 rate_ctrl; -+ u32 fc_glob_thrs; -+ u32 fc_port_thrs; -+ u32 mc_fc_glob_thrs; -+ u32 dc_glob_thrs; -+ u32 arl_vlan_cmd; -+ u32 arl_ctrl[3]; -+ u32 vlan_cfg; -+ u32 pvid[2]; -+ u32 vlan_ctrl[3]; -+ u32 session_id[8]; -+ u32 intr_stat; -+ u32 intr_mask; -+ u32 sram_test; -+ u32 mem_queue; -+ u32 farl_ctrl; -+ u32 fc_input_thrs, __res1[2]; -+ u32 clk_skew_ctrl; -+ u32 mac_glob_cfg_ext, __res2[2]; -+ u32 dma_ring_ctrl; -+ u32 dma_auto_poll_cfg; -+ u32 delay_intr_cfg, __res3; -+ u32 ts_dma_ctrl0; -+ u32 ts_desc_ptr0; -+ u32 ts_desc_base_addr0, __res4; -+ u32 fs_dma_ctrl0; -+ u32 fs_desc_ptr0; -+ u32 fs_desc_base_addr0, __res5; -+ u32 ts_dma_ctrl1; -+ u32 ts_desc_ptr1; -+ u32 ts_desc_base_addr1, __res6; -+ u32 fs_dma_ctrl1; -+ u32 fs_desc_ptr1; -+ u32 fs_desc_base_addr1; -+ u32 __res7[109]; -+ u32 mac_counter0[13]; -+}; -+ -+struct _tx_ring { -+ struct tx_desc *desc; -+ dma_addr_t phys_addr; -+ struct tx_desc *cur_addr; -+ struct sk_buff *buff_tab[TX_DESCS]; -+ unsigned int phys_tab[TX_DESCS]; -+ u32 free_index; -+ u32 count_index; -+ u32 cur_index; -+ int num_used; -+ int num_count; -+}; -+ -+struct _rx_ring { -+ struct rx_desc *desc; -+ dma_addr_t phys_addr; -+ struct rx_desc *cur_addr; -+ struct sk_buff *buff_tab[RX_DESCS]; -+ unsigned int phys_tab[RX_DESCS]; -+ u32 cur_index; -+ u32 alloc_index; -+ int alloc_count; -+}; -+ -+struct sw { -+ struct resource *mem_res; -+ struct switch_regs __iomem *regs; -+ struct napi_struct napi; -+ struct cns3xxx_plat_info *plat; -+ struct _tx_ring *tx_ring; -+ struct _rx_ring *rx_ring; -+}; -+ -+struct port { -+ struct net_device *netdev; -+ struct phy_device *phydev; -+ struct sw *sw; -+ int id; /* logical port ID */ -+ int speed, duplex; -+}; -+ -+static spinlock_t mdio_lock; -+static DEFINE_SPINLOCK(tx_lock); -+static struct switch_regs __iomem *mdio_regs; /* mdio command and status only */ -+struct mii_bus *mdio_bus; -+static int ports_open; -+static struct port *switch_port_tab[4]; -+static struct dma_pool *rx_dma_pool; -+static struct dma_pool *tx_dma_pool; -+struct net_device *napi_dev; -+ -+static int cns3xxx_mdio_cmd(struct mii_bus *bus, int phy_id, int location, -+ int write, u16 cmd) -+{ -+ int cycles = 0; -+ u32 temp = 0; -+ -+ temp = __raw_readl(&mdio_regs->phy_control); -+ temp |= MDIO_CMD_COMPLETE; -+ __raw_writel(temp, &mdio_regs->phy_control); -+ udelay(10); -+ -+ if (write) { -+ temp = (cmd << MDIO_VALUE_OFFSET); -+ temp |= MDIO_WRITE_COMMAND; -+ } else { -+ temp = MDIO_READ_COMMAND; -+ } -+ temp |= ((location & 0x1f) << MDIO_REG_OFFSET); -+ temp |= (phy_id & 0x1f); -+ -+ __raw_writel(temp, &mdio_regs->phy_control); -+ -+ while (((__raw_readl(&mdio_regs->phy_control) & MDIO_CMD_COMPLETE) == 0) -+ && cycles < 5000) { -+ udelay(1); -+ cycles++; -+ } -+ -+ if (cycles == 5000) { -+ printk(KERN_ERR "%s #%i: MII transaction failed\n", bus->name, -+ phy_id); -+ return -1; -+ } -+ -+ temp = __raw_readl(&mdio_regs->phy_control); -+ temp |= MDIO_CMD_COMPLETE; -+ __raw_writel(temp, &mdio_regs->phy_control); -+ -+ if (write) -+ return 0; -+ -+ return ((temp >> MDIO_VALUE_OFFSET) & 0xFFFF); -+} -+ -+static int cns3xxx_mdio_read(struct mii_bus *bus, int phy_id, int location) -+{ -+ unsigned long flags; -+ int ret; -+ -+ spin_lock_irqsave(&mdio_lock, flags); -+ ret = cns3xxx_mdio_cmd(bus, phy_id, location, 0, 0); -+ spin_unlock_irqrestore(&mdio_lock, flags); -+ return ret; -+} -+ -+static int cns3xxx_mdio_write(struct mii_bus *bus, int phy_id, int location, -+ u16 val) -+{ -+ unsigned long flags; -+ int ret; -+ -+ spin_lock_irqsave(&mdio_lock, flags); -+ ret = cns3xxx_mdio_cmd(bus, phy_id, location, 1, val); -+ spin_unlock_irqrestore(&mdio_lock, flags); -+ return ret; -+} -+ -+static int cns3xxx_mdio_register(void) -+{ -+ int err; -+ -+ if (!(mdio_bus = mdiobus_alloc())) -+ return -ENOMEM; -+ -+ mdio_regs = (struct switch_regs __iomem *)CNS3XXX_SWITCH_BASE_VIRT; -+ -+ spin_lock_init(&mdio_lock); -+ mdio_bus->name = "CNS3xxx MII Bus"; -+ mdio_bus->read = &cns3xxx_mdio_read; -+ mdio_bus->write = &cns3xxx_mdio_write; -+ strcpy(mdio_bus->id, "0"); -+ -+ if ((err = mdiobus_register(mdio_bus))) -+ mdiobus_free(mdio_bus); -+ return err; -+} -+ -+static void cns3xxx_mdio_remove(void) -+{ -+ mdiobus_unregister(mdio_bus); -+ mdiobus_free(mdio_bus); -+} -+ -+static void enable_tx_dma(struct sw *sw) -+{ -+ __raw_writel(0x1, &sw->regs->ts_dma_ctrl0); -+} -+ -+static void enable_rx_dma(struct sw *sw) -+{ -+ __raw_writel(0x1, &sw->regs->fs_dma_ctrl0); -+} -+ -+static void cns3xxx_adjust_link(struct net_device *dev) -+{ -+ struct port *port = netdev_priv(dev); -+ struct phy_device *phydev = port->phydev; -+ -+ if (!phydev->link) { -+ if (port->speed) { -+ port->speed = 0; -+ printk(KERN_INFO "%s: link down\n", dev->name); -+ } -+ return; -+ } -+ -+ if (port->speed == phydev->speed && port->duplex == phydev->duplex) -+ return; -+ -+ port->speed = phydev->speed; -+ port->duplex = phydev->duplex; -+ -+ printk(KERN_INFO "%s: link up, speed %u Mb/s, %s duplex\n", -+ dev->name, port->speed, port->duplex ? "full" : "half"); -+} -+ -+irqreturn_t eth_rx_irq(int irq, void *pdev) -+{ -+ struct net_device *dev = pdev; -+ struct sw *sw = netdev_priv(dev); -+ if (likely(napi_schedule_prep(&sw->napi))) { -+ disable_irq_nosync(IRQ_CNS3XXX_SW_R0RXC); -+ __napi_schedule(&sw->napi); -+ } -+ return (IRQ_HANDLED); -+} -+ -+irqreturn_t eth_stat_irq(int irq, void *pdev) -+{ -+ struct net_device *dev = pdev; -+ struct sw *sw = netdev_priv(dev); -+ u32 cfg; -+ u32 stat = __raw_readl(&sw->regs->intr_stat); -+ __raw_writel(0xffffffff, &sw->regs->intr_stat); -+ -+ if (stat & MAC2_RX_ERROR) -+ switch_port_tab[3]->netdev->stats.rx_dropped++; -+ if (stat & MAC1_RX_ERROR) -+ switch_port_tab[1]->netdev->stats.rx_dropped++; -+ if (stat & MAC0_RX_ERROR) -+ switch_port_tab[0]->netdev->stats.rx_dropped++; -+ -+ if (stat & MAC0_STATUS_CHANGE) { -+ cfg = __raw_readl(&sw->regs->mac_cfg[0]); -+ switch_port_tab[0]->phydev->link = (cfg & 0x1); -+ switch_port_tab[0]->phydev->duplex = ((cfg >> 4) & 0x1); -+ if (((cfg >> 2) & 0x3) == 2) -+ switch_port_tab[0]->phydev->speed = 1000; -+ else if (((cfg >> 2) & 0x3) == 1) -+ switch_port_tab[0]->phydev->speed = 100; -+ else -+ switch_port_tab[0]->phydev->speed = 10; -+ cns3xxx_adjust_link(switch_port_tab[0]->netdev); -+ } -+ -+ if (stat & MAC1_STATUS_CHANGE) { -+ cfg = __raw_readl(&sw->regs->mac_cfg[1]); -+ switch_port_tab[1]->phydev->link = (cfg & 0x1); -+ switch_port_tab[1]->phydev->duplex = ((cfg >> 4) & 0x1); -+ if (((cfg >> 2) & 0x3) == 2) -+ switch_port_tab[1]->phydev->speed = 1000; -+ else if (((cfg >> 2) & 0x3) == 1) -+ switch_port_tab[1]->phydev->speed = 100; -+ else -+ switch_port_tab[1]->phydev->speed = 10; -+ cns3xxx_adjust_link(switch_port_tab[1]->netdev); -+ } -+ -+ if (stat & MAC2_STATUS_CHANGE) { -+ cfg = __raw_readl(&sw->regs->mac_cfg[3]); -+ switch_port_tab[3]->phydev->link = (cfg & 0x1); -+ switch_port_tab[3]->phydev->duplex = ((cfg >> 4) & 0x1); -+ if (((cfg >> 2) & 0x3) == 2) -+ switch_port_tab[3]->phydev->speed = 1000; -+ else if (((cfg >> 2) & 0x3) == 1) -+ switch_port_tab[3]->phydev->speed = 100; -+ else -+ switch_port_tab[3]->phydev->speed = 10; -+ cns3xxx_adjust_link(switch_port_tab[3]->netdev); -+ } -+ -+ return (IRQ_HANDLED); -+} -+ -+ -+static void cns3xxx_alloc_rx_buf(struct sw *sw, int received) -+{ -+ struct _rx_ring *rx_ring = sw->rx_ring; -+ unsigned int i = rx_ring->alloc_index; -+ struct rx_desc *desc = &(rx_ring)->desc[i]; -+ struct sk_buff *skb; -+ unsigned int phys; -+ -+ for (received += rx_ring->alloc_count; received > 0; received--) { -+ if ((skb = dev_alloc_skb(MAX_MRU))) { -+ if (SKB_DMA_REALIGN) -+ skb_reserve(skb, SKB_DMA_REALIGN); -+ skb_reserve(skb, NET_IP_ALIGN); -+ phys = dma_map_single(NULL, skb->data, -+ CNS3XXX_MAX_MTU, DMA_FROM_DEVICE); -+ if (dma_mapping_error(NULL, phys)) { -+ dev_kfree_skb(skb); -+ /* Failed to map, better luck next time */ -+ goto out;; -+ } -+ desc->sdp = phys; -+ } else { -+ /* Failed to allocate skb, try again next time */ -+ goto out; -+ } -+ -+ /* put the new buffer on RX-free queue */ -+ rx_ring->buff_tab[i] = skb; -+ rx_ring->phys_tab[i] = phys; -+ if (i == RX_DESCS - 1) { -+ i = 0; -+ desc->config0 = END_OF_RING | FIRST_SEGMENT | -+ LAST_SEGMENT | CNS3XXX_MAX_MTU; -+ desc = &(rx_ring)->desc[i]; -+ } else { -+ desc->config0 = FIRST_SEGMENT | LAST_SEGMENT | CNS3XXX_MAX_MTU; -+ i++; -+ desc++; -+ } -+ } -+out: -+ rx_ring->alloc_count = received; -+ rx_ring->alloc_index = i; -+} -+ -+static void clear_tx_desc(struct sw *sw) -+{ -+ struct _tx_ring *tx_ring = sw->tx_ring; -+ struct tx_desc *desc; -+ int i; -+ int index; -+ int num_used = tx_ring->num_used; -+ struct sk_buff *skb; -+ -+ if (num_used < (TX_DESCS >> 1)) -+ return; -+ -+ index = tx_ring->free_index; -+ desc = &(tx_ring)->desc[index]; -+ for (i = 0; i < num_used; i++) { -+ if (desc->cown) { -+ skb = tx_ring->buff_tab[index]; -+ tx_ring->buff_tab[index] = 0; -+ if (skb) -+ dev_kfree_skb_any(skb); -+ dma_unmap_single(NULL, tx_ring->phys_tab[index], -+ desc->sdl, DMA_TO_DEVICE); -+ if (++index == TX_DESCS) { -+ index = 0; -+ desc = &(tx_ring)->desc[index]; -+ } else { -+ desc++; -+ } -+ } else { -+ break; -+ } -+ } -+ tx_ring->free_index = index; -+ tx_ring->num_used -= i; -+} -+ -+static int eth_poll(struct napi_struct *napi, int budget) -+{ -+ struct sw *sw = container_of(napi, struct sw, napi); -+ struct net_device *dev; -+ struct _rx_ring *rx_ring = sw->rx_ring; -+ int received = 0; -+ unsigned int length; -+ unsigned int i = rx_ring->cur_index; -+ struct rx_desc *desc = &(rx_ring)->desc[i]; -+ -+ while (desc->cown) { -+ struct sk_buff *skb; -+ -+ if (received >= budget) -+ break; -+ -+ skb = rx_ring->buff_tab[i]; -+ -+ dev = switch_port_tab[desc->sp]->netdev; -+ -+ length = desc->sdl; -+ /* process received frame */ -+ dma_unmap_single(&dev->dev, rx_ring->phys_tab[i], -+ length, DMA_FROM_DEVICE); -+ -+ skb_put(skb, length); -+ -+ skb->dev = dev; -+ skb->protocol = eth_type_trans(skb, dev); -+ -+ dev->stats.rx_packets++; -+ dev->stats.rx_bytes += length; -+ -+ /* RX Hardware checksum offload */ -+ switch (desc->prot) { -+ case 1: -+ case 2: -+ case 5: -+ case 6: -+ case 13: -+ case 14: -+ if (desc->l4f) -+ skb->ip_summed = CHECKSUM_NONE; -+ else -+ skb->ip_summed = CHECKSUM_UNNECESSARY; -+ break; -+ default: -+ skb->ip_summed = CHECKSUM_NONE; -+ break; -+ } -+ -+ napi_gro_receive(napi, skb); -+ -+ received++; -+ -+ if (++i == RX_DESCS) { -+ i = 0; -+ desc = &(rx_ring)->desc[i]; -+ } else { -+ desc++; -+ } -+ } -+ -+ cns3xxx_alloc_rx_buf(sw, received); -+ -+ rx_ring->cur_index = i; -+ -+ if (received != budget) { -+ napi_complete(napi); -+ enable_irq(IRQ_CNS3XXX_SW_R0RXC); -+ } -+ -+ enable_rx_dma(sw); -+ -+ return received; -+} -+ -+static int eth_xmit(struct sk_buff *skb, struct net_device *dev) -+{ -+ struct port *port = netdev_priv(dev); -+ struct sw *sw = port->sw; -+ struct _tx_ring *tx_ring = sw->tx_ring; -+ struct tx_desc *tx_desc; -+ int index; -+ int len; -+ char pmap = (1 << port->id); -+ unsigned int phys; -+ int nr_frags = skb_shinfo(skb)->nr_frags; -+ struct skb_frag_struct *frag; -+ unsigned int i; -+ -+ if (pmap == 8) -+ pmap = (1 << 4); -+ -+ if (skb->len > CNS3XXX_MAX_MTU) { -+ dev_kfree_skb(skb); -+ dev->stats.tx_errors++; -+ return NETDEV_TX_OK; -+ } -+ -+ spin_lock(&tx_lock); -+ -+ if ((tx_ring->num_used + nr_frags) >= TX_DESCS) { -+ clear_tx_desc(sw); -+ if ((tx_ring->num_used + nr_frags) >= TX_DESCS) { -+ spin_unlock(&tx_lock); -+ return NETDEV_TX_BUSY; -+ } -+ } -+ -+ index = tx_ring->cur_index; -+ tx_ring->cur_index = ((tx_ring->cur_index + nr_frags + 1) % TX_DESCS); -+ -+ spin_unlock(&tx_lock); -+ -+ if (!nr_frags) { -+ tx_desc = &(tx_ring)->desc[index]; -+ -+ len = skb->len; -+ -+ phys = dma_map_single(NULL, skb->data, len, -+ DMA_TO_DEVICE); -+ -+ tx_desc->sdp = phys; -+ tx_desc->pmap = pmap; -+ tx_ring->phys_tab[index] = phys; -+ -+ tx_ring->buff_tab[index] = skb; -+ if (index == TX_DESCS - 1) { -+ tx_desc->config0 = END_OF_RING | FIRST_SEGMENT | LAST_SEGMENT | -+ FORCE_ROUTE | IP_CHECKSUM | UDP_CHECKSUM | -+ TCP_CHECKSUM | len; -+ } else { -+ tx_desc->config0 = FIRST_SEGMENT | LAST_SEGMENT | -+ FORCE_ROUTE | IP_CHECKSUM | UDP_CHECKSUM | -+ TCP_CHECKSUM | len; -+ } -+ } else { -+ unsigned int config; -+ -+ index = ((index + nr_frags) % TX_DESCS); -+ tx_desc = &(tx_ring)->desc[index]; -+ -+ /* fragments */ -+ for (i = nr_frags; i > 0; i--) { -+ void *addr; -+ -+ frag = &skb_shinfo(skb)->frags[i-1]; -+ len = frag->size; -+ -+ addr = page_address(skb_frag_page(frag)) + -+ frag->page_offset; -+ phys = dma_map_single(NULL, addr, len, DMA_TO_DEVICE); -+ -+ tx_desc->sdp = phys; -+ -+ tx_desc->pmap = pmap; -+ tx_ring->phys_tab[index] = phys; -+ -+ config = FORCE_ROUTE | IP_CHECKSUM | UDP_CHECKSUM | -+ TCP_CHECKSUM | len; -+ if (i == nr_frags) { -+ config |= LAST_SEGMENT; -+ tx_ring->buff_tab[index] = skb; -+ } -+ if (index == TX_DESCS - 1) -+ config |= END_OF_RING; -+ tx_desc->config0 = config; -+ -+ if (index == 0) { -+ index = TX_DESCS - 1; -+ tx_desc = &(tx_ring)->desc[index]; -+ } else { -+ index--; -+ tx_desc--; -+ } -+ } -+ -+ /* header */ -+ len = skb->len - skb->data_len; -+ -+ phys = dma_map_single(NULL, skb->data, len, -+ DMA_TO_DEVICE); -+ -+ tx_desc->sdp = phys; -+ tx_desc->pmap = pmap; -+ tx_ring->phys_tab[index] = phys; -+ -+ if (index == TX_DESCS - 1) { -+ tx_desc->config0 = END_OF_RING | FIRST_SEGMENT | -+ FORCE_ROUTE | IP_CHECKSUM | UDP_CHECKSUM | -+ TCP_CHECKSUM | len; -+ } else { -+ tx_desc->config0 = FIRST_SEGMENT | -+ FORCE_ROUTE | IP_CHECKSUM | UDP_CHECKSUM | -+ TCP_CHECKSUM | len; -+ } -+ } -+ -+ mb(); -+ -+ spin_lock(&tx_lock); -+ tx_ring->num_used += nr_frags + 1; -+ spin_unlock(&tx_lock); -+ -+ dev->stats.tx_packets++; -+ dev->stats.tx_bytes += skb->len; -+ -+ enable_tx_dma(sw); -+ -+ return NETDEV_TX_OK; -+} -+ -+static int eth_ioctl(struct net_device *dev, struct ifreq *req, int cmd) -+{ -+ struct port *port = netdev_priv(dev); -+ -+ if (!netif_running(dev)) -+ return -EINVAL; -+ return phy_mii_ioctl(port->phydev, req, cmd); -+} -+ -+/* ethtool support */ -+ -+static void cns3xxx_get_drvinfo(struct net_device *dev, -+ struct ethtool_drvinfo *info) -+{ -+ strcpy(info->driver, DRV_NAME); -+ strcpy(info->bus_info, "internal"); -+} -+ -+static int cns3xxx_get_settings(struct net_device *dev, struct ethtool_cmd *cmd) -+{ -+ struct port *port = netdev_priv(dev); -+ return phy_ethtool_gset(port->phydev, cmd); -+} -+ -+static int cns3xxx_set_settings(struct net_device *dev, struct ethtool_cmd *cmd) -+{ -+ struct port *port = netdev_priv(dev); -+ return phy_ethtool_sset(port->phydev, cmd); -+} -+ -+static int cns3xxx_nway_reset(struct net_device *dev) -+{ -+ struct port *port = netdev_priv(dev); -+ return phy_start_aneg(port->phydev); -+} -+ -+static struct ethtool_ops cns3xxx_ethtool_ops = { -+ .get_drvinfo = cns3xxx_get_drvinfo, -+ .get_settings = cns3xxx_get_settings, -+ .set_settings = cns3xxx_set_settings, -+ .nway_reset = cns3xxx_nway_reset, -+ .get_link = ethtool_op_get_link, -+}; -+ -+ -+static int init_rings(struct sw *sw) -+{ -+ int i; -+ struct _rx_ring *rx_ring = sw->rx_ring; -+ struct _tx_ring *tx_ring = sw->tx_ring; -+ -+ __raw_writel(0, &sw->regs->fs_dma_ctrl0); -+ __raw_writel(TS_SUSPEND | FS_SUSPEND, &sw->regs->dma_auto_poll_cfg); -+ __raw_writel(QUEUE_THRESHOLD, &sw->regs->dma_ring_ctrl); -+ __raw_writel(CLR_FS_STATE | QUEUE_THRESHOLD, &sw->regs->dma_ring_ctrl); -+ -+ __raw_writel(QUEUE_THRESHOLD, &sw->regs->dma_ring_ctrl); -+ -+ if (!(rx_dma_pool = dma_pool_create(DRV_NAME, NULL, -+ RX_POOL_ALLOC_SIZE, 32, 0))) -+ return -ENOMEM; -+ -+ if (!(rx_ring->desc = dma_pool_alloc(rx_dma_pool, GFP_KERNEL, -+ &rx_ring->phys_addr))) -+ return -ENOMEM; -+ memset(rx_ring->desc, 0, RX_POOL_ALLOC_SIZE); -+ -+ /* Setup RX buffers */ -+ for (i = 0; i < RX_DESCS; i++) { -+ struct rx_desc *desc = &(rx_ring)->desc[i]; -+ struct sk_buff *skb; -+ if (!(skb = dev_alloc_skb(MAX_MRU))) -+ return -ENOMEM; -+ if (SKB_DMA_REALIGN) -+ skb_reserve(skb, SKB_DMA_REALIGN); -+ skb_reserve(skb, NET_IP_ALIGN); -+ desc->sdl = CNS3XXX_MAX_MTU; -+ if (i == (RX_DESCS - 1)) -+ desc->eor = 1; -+ desc->fsd = 1; -+ desc->lsd = 1; -+ -+ desc->sdp = dma_map_single(NULL, skb->data, -+ CNS3XXX_MAX_MTU, DMA_FROM_DEVICE); -+ if (dma_mapping_error(NULL, desc->sdp)) { -+ return -EIO; -+ } -+ rx_ring->buff_tab[i] = skb; -+ rx_ring->phys_tab[i] = desc->sdp; -+ desc->cown = 0; -+ } -+ __raw_writel(rx_ring->phys_addr, &sw->regs->fs_desc_ptr0); -+ __raw_writel(rx_ring->phys_addr, &sw->regs->fs_desc_base_addr0); -+ -+ if (!(tx_dma_pool = dma_pool_create(DRV_NAME, NULL, -+ TX_POOL_ALLOC_SIZE, 32, 0))) -+ return -ENOMEM; -+ -+ if (!(tx_ring->desc = dma_pool_alloc(tx_dma_pool, GFP_KERNEL, -+ &tx_ring->phys_addr))) -+ return -ENOMEM; -+ memset(tx_ring->desc, 0, TX_POOL_ALLOC_SIZE); -+ -+ /* Setup TX buffers */ -+ for (i = 0; i < TX_DESCS; i++) { -+ struct tx_desc *desc = &(tx_ring)->desc[i]; -+ tx_ring->buff_tab[i] = 0; -+ -+ if (i == (TX_DESCS - 1)) -+ desc->eor = 1; -+ desc->cown = 1; -+ } -+ __raw_writel(tx_ring->phys_addr, &sw->regs->ts_desc_ptr0); -+ __raw_writel(tx_ring->phys_addr, &sw->regs->ts_desc_base_addr0); -+ -+ return 0; -+} -+ -+static void destroy_rings(struct sw *sw) -+{ -+ int i; -+ if (sw->rx_ring->desc) { -+ for (i = 0; i < RX_DESCS; i++) { -+ struct _rx_ring *rx_ring = sw->rx_ring; -+ struct rx_desc *desc = &(rx_ring)->desc[i]; -+ struct sk_buff *skb = sw->rx_ring->buff_tab[i]; -+ if (skb) { -+ dma_unmap_single(NULL, -+ desc->sdp, -+ CNS3XXX_MAX_MTU, DMA_FROM_DEVICE); -+ dev_kfree_skb(skb); -+ } -+ } -+ dma_pool_free(rx_dma_pool, sw->rx_ring->desc, sw->rx_ring->phys_addr); -+ dma_pool_destroy(rx_dma_pool); -+ rx_dma_pool = 0; -+ sw->rx_ring->desc = 0; -+ } -+ if (sw->tx_ring->desc) { -+ for (i = 0; i < TX_DESCS; i++) { -+ struct _tx_ring *tx_ring = sw->tx_ring; -+ struct tx_desc *desc = &(tx_ring)->desc[i]; -+ struct sk_buff *skb = sw->tx_ring->buff_tab[i]; -+ if (skb) { -+ dma_unmap_single(NULL, desc->sdp, -+ skb->len, DMA_TO_DEVICE); -+ dev_kfree_skb(skb); -+ } -+ } -+ dma_pool_free(tx_dma_pool, sw->tx_ring->desc, sw->tx_ring->phys_addr); -+ dma_pool_destroy(tx_dma_pool); -+ tx_dma_pool = 0; -+ sw->tx_ring->desc = 0; -+ } -+} -+ -+static int eth_open(struct net_device *dev) -+{ -+ struct port *port = netdev_priv(dev); -+ struct sw *sw = port->sw; -+ u32 temp; -+ -+ port->speed = 0; /* force "link up" message */ -+ phy_start(port->phydev); -+ -+ netif_start_queue(dev); -+ -+ if (!ports_open) { -+ request_irq(IRQ_CNS3XXX_SW_R0RXC, eth_rx_irq, IRQF_SHARED, "gig_switch", napi_dev); -+ request_irq(IRQ_CNS3XXX_SW_STATUS, eth_stat_irq, IRQF_SHARED, "gig_stat", napi_dev); -+ napi_enable(&sw->napi); -+ netif_start_queue(napi_dev); -+ -+ __raw_writel(~(MAC0_STATUS_CHANGE | MAC1_STATUS_CHANGE | MAC2_STATUS_CHANGE | -+ MAC0_RX_ERROR | MAC1_RX_ERROR | MAC2_RX_ERROR), &sw->regs->intr_mask); -+ -+ temp = __raw_readl(&sw->regs->mac_cfg[2]); -+ temp &= ~(PORT_DISABLE); -+ __raw_writel(temp, &sw->regs->mac_cfg[2]); -+ -+ temp = __raw_readl(&sw->regs->dma_auto_poll_cfg); -+ temp &= ~(TS_SUSPEND | FS_SUSPEND); -+ __raw_writel(temp, &sw->regs->dma_auto_poll_cfg); -+ -+ enable_rx_dma(sw); -+ } -+ temp = __raw_readl(&sw->regs->mac_cfg[port->id]); -+ temp &= ~(PORT_DISABLE); -+ __raw_writel(temp, &sw->regs->mac_cfg[port->id]); -+ -+ ports_open++; -+ netif_carrier_on(dev); -+ -+ return 0; -+} -+ -+static int eth_close(struct net_device *dev) -+{ -+ struct port *port = netdev_priv(dev); -+ struct sw *sw = port->sw; -+ u32 temp; -+ -+ ports_open--; -+ -+ temp = __raw_readl(&sw->regs->mac_cfg[port->id]); -+ temp |= (PORT_DISABLE); -+ __raw_writel(temp, &sw->regs->mac_cfg[port->id]); -+ -+ netif_stop_queue(dev); -+ -+ phy_stop(port->phydev); -+ -+ if (!ports_open) { -+ disable_irq(IRQ_CNS3XXX_SW_R0RXC); -+ free_irq(IRQ_CNS3XXX_SW_R0RXC, napi_dev); -+ disable_irq(IRQ_CNS3XXX_SW_STATUS); -+ free_irq(IRQ_CNS3XXX_SW_STATUS, napi_dev); -+ napi_disable(&sw->napi); -+ netif_stop_queue(napi_dev); -+ temp = __raw_readl(&sw->regs->mac_cfg[2]); -+ temp |= (PORT_DISABLE); -+ __raw_writel(temp, &sw->regs->mac_cfg[2]); -+ -+ __raw_writel(TS_SUSPEND | FS_SUSPEND, -+ &sw->regs->dma_auto_poll_cfg); -+ } -+ -+ netif_carrier_off(dev); -+ return 0; -+} -+ -+static void eth_rx_mode(struct net_device *dev) -+{ -+ struct port *port = netdev_priv(dev); -+ struct sw *sw = port->sw; -+ u32 temp; -+ -+ temp = __raw_readl(&sw->regs->mac_glob_cfg); -+ -+ if (dev->flags & IFF_PROMISC) { -+ if (port->id == 3) -+ temp |= ((1 << 2) << PROMISC_OFFSET); -+ else -+ temp |= ((1 << port->id) << PROMISC_OFFSET); -+ } else { -+ if (port->id == 3) -+ temp &= ~((1 << 2) << PROMISC_OFFSET); -+ else -+ temp &= ~((1 << port->id) << PROMISC_OFFSET); -+ } -+ __raw_writel(temp, &sw->regs->mac_glob_cfg); -+} -+ -+static int eth_set_mac(struct net_device *netdev, void *p) -+{ -+ struct port *port = netdev_priv(netdev); -+ struct sw *sw = port->sw; -+ struct sockaddr *addr = p; -+ u32 cycles = 0; -+ -+ if (!is_valid_ether_addr(addr->sa_data)) -+ return -EADDRNOTAVAIL; -+ -+ /* Invalidate old ARL Entry */ -+ if (port->id == 3) -+ __raw_writel((port->id << 16) | (0x4 << 9), &sw->regs->arl_ctrl[0]); -+ else -+ __raw_writel(((port->id + 1) << 16) | (0x4 << 9), &sw->regs->arl_ctrl[0]); -+ __raw_writel( ((netdev->dev_addr[0] << 24) | (netdev->dev_addr[1] << 16) | -+ (netdev->dev_addr[2] << 8) | (netdev->dev_addr[3])), -+ &sw->regs->arl_ctrl[1]); -+ -+ __raw_writel( ((netdev->dev_addr[4] << 24) | (netdev->dev_addr[5] << 16) | -+ (1 << 1)), -+ &sw->regs->arl_ctrl[2]); -+ __raw_writel((1 << 19), &sw->regs->arl_vlan_cmd); -+ -+ while (((__raw_readl(&sw->regs->arl_vlan_cmd) & (1 << 21)) == 0) -+ && cycles < 5000) { -+ udelay(1); -+ cycles++; -+ } -+ -+ cycles = 0; -+ memcpy(netdev->dev_addr, addr->sa_data, netdev->addr_len); -+ -+ if (port->id == 3) -+ __raw_writel((port->id << 16) | (0x4 << 9), &sw->regs->arl_ctrl[0]); -+ else -+ __raw_writel(((port->id + 1) << 16) | (0x4 << 9), &sw->regs->arl_ctrl[0]); -+ __raw_writel( ((addr->sa_data[0] << 24) | (addr->sa_data[1] << 16) | -+ (addr->sa_data[2] << 8) | (addr->sa_data[3])), -+ &sw->regs->arl_ctrl[1]); -+ -+ __raw_writel( ((addr->sa_data[4] << 24) | (addr->sa_data[5] << 16) | -+ (7 << 4) | (1 << 1)), &sw->regs->arl_ctrl[2]); -+ __raw_writel((1 << 19), &sw->regs->arl_vlan_cmd); -+ -+ while (((__raw_readl(&sw->regs->arl_vlan_cmd) & (1 << 21)) == 0) -+ && cycles < 5000) { -+ udelay(1); -+ cycles++; -+ } -+ return 0; -+} -+ -+static const struct net_device_ops cns3xxx_netdev_ops = { -+ .ndo_open = eth_open, -+ .ndo_stop = eth_close, -+ .ndo_start_xmit = eth_xmit, -+ .ndo_set_rx_mode = eth_rx_mode, -+ .ndo_do_ioctl = eth_ioctl, -+ .ndo_change_mtu = eth_change_mtu, -+ .ndo_set_mac_address = eth_set_mac, -+ .ndo_validate_addr = eth_validate_addr, -+}; -+ -+static int __devinit eth_init_one(struct platform_device *pdev) -+{ -+ int i; -+ struct port *port; -+ struct sw *sw; -+ struct net_device *dev; -+ struct cns3xxx_plat_info *plat = pdev->dev.platform_data; -+ u32 regs_phys; -+ char phy_id[MII_BUS_ID_SIZE + 3]; -+ int err; -+ u32 temp; -+ -+ if (!(napi_dev = alloc_etherdev(sizeof(struct sw)))) -+ return -ENOMEM; -+ strcpy(napi_dev->name, "switch%d"); -+ napi_dev->features = NETIF_F_IP_CSUM | NETIF_F_SG; -+ -+ SET_NETDEV_DEV(napi_dev, &pdev->dev); -+ sw = netdev_priv(napi_dev); -+ memset(sw, 0, sizeof(struct sw)); -+ sw->regs = (struct switch_regs __iomem *)CNS3XXX_SWITCH_BASE_VIRT; -+ regs_phys = CNS3XXX_SWITCH_BASE; -+ sw->mem_res = request_mem_region(regs_phys, REGS_SIZE, napi_dev->name); -+ if (!sw->mem_res) { -+ err = -EBUSY; -+ goto err_free; -+ } -+ -+ for (i = 0; i < 4; i++) { -+ temp = __raw_readl(&sw->regs->mac_cfg[i]); -+ temp |= (PORT_DISABLE); -+ __raw_writel(temp, &sw->regs->mac_cfg[i]); -+ } -+ -+ temp = PORT_DISABLE; -+ __raw_writel(temp, &sw->regs->mac_cfg[2]); -+ -+ temp = __raw_readl(&sw->regs->vlan_cfg); -+ temp |= NIC_MODE | VLAN_UNAWARE; -+ __raw_writel(temp, &sw->regs->vlan_cfg); -+ -+ __raw_writel(UNKNOWN_VLAN_TO_CPU | -+ CRC_STRIPPING, &sw->regs->mac_glob_cfg); -+ -+ if (!(sw->rx_ring = kmalloc(sizeof(struct _rx_ring), GFP_KERNEL))) { -+ err = -ENOMEM; -+ goto err_free; -+ } -+ memset(sw->rx_ring, 0, sizeof(struct _rx_ring)); -+ -+ if (!(sw->tx_ring = kmalloc(sizeof(struct _tx_ring), GFP_KERNEL))) { -+ err = -ENOMEM; -+ goto err_free_rx; -+ } -+ memset(sw->tx_ring, 0, sizeof(struct _tx_ring)); -+ -+ if ((err = init_rings(sw)) != 0) { -+ destroy_rings(sw); -+ err = -ENOMEM; -+ goto err_free_rings; -+ } -+ platform_set_drvdata(pdev, napi_dev); -+ -+ netif_napi_add(napi_dev, &sw->napi, eth_poll, NAPI_WEIGHT); -+ -+ for (i = 0; i < 3; i++) { -+ if (!(plat->ports & (1 << i))) { -+ continue; -+ } -+ -+ if (!(dev = alloc_etherdev(sizeof(struct port)))) { -+ goto free_ports; -+ } -+ -+ port = netdev_priv(dev); -+ port->netdev = dev; -+ if (i == 2) -+ port->id = 3; -+ else -+ port->id = i; -+ port->sw = sw; -+ -+ temp = __raw_readl(&sw->regs->mac_cfg[port->id]); -+ temp |= (PORT_DISABLE | PORT_BLOCK_STATE | PORT_LEARN_DIS); -+ __raw_writel(temp, &sw->regs->mac_cfg[port->id]); -+ -+ dev->netdev_ops = &cns3xxx_netdev_ops; -+ dev->ethtool_ops = &cns3xxx_ethtool_ops; -+ dev->tx_queue_len = 1000; -+ dev->features = NETIF_F_IP_CSUM | NETIF_F_SG; -+ -+ switch_port_tab[port->id] = port; -+ memcpy(dev->dev_addr, &plat->hwaddr[i], ETH_ALEN); -+ -+ snprintf(phy_id, MII_BUS_ID_SIZE + 3, PHY_ID_FMT, "0", plat->phy[i]); -+ port->phydev = phy_connect(dev, phy_id, &cns3xxx_adjust_link, 0, -+ PHY_INTERFACE_MODE_RGMII); -+ if ((err = IS_ERR(port->phydev))) { -+ switch_port_tab[port->id] = 0; -+ free_netdev(dev); -+ goto free_ports; -+ } -+ -+ port->phydev->irq = PHY_IGNORE_INTERRUPT; -+ -+ if ((err = register_netdev(dev))) { -+ phy_disconnect(port->phydev); -+ switch_port_tab[port->id] = 0; -+ free_netdev(dev); -+ goto free_ports; -+ } -+ -+ printk(KERN_INFO "%s: RGMII PHY %i on cns3xxx Switch\n", dev->name, plat->phy[i]); -+ netif_carrier_off(dev); -+ dev = 0; -+ } -+ -+ return 0; -+ -+free_ports: -+ err = -ENOMEM; -+ for (--i; i >= 0; i--) { -+ if (switch_port_tab[i]) { -+ port = switch_port_tab[i]; -+ dev = port->netdev; -+ unregister_netdev(dev); -+ phy_disconnect(port->phydev); -+ switch_port_tab[i] = 0; -+ free_netdev(dev); -+ } -+ } -+err_free_rings: -+ kfree(sw->tx_ring); -+err_free_rx: -+ kfree(sw->rx_ring); -+err_free: -+ free_netdev(napi_dev); -+ return err; -+} -+ -+static int __devexit eth_remove_one(struct platform_device *pdev) -+{ -+ struct net_device *dev = platform_get_drvdata(pdev); -+ struct sw *sw = netdev_priv(dev); -+ int i; -+ destroy_rings(sw); -+ -+ for (i = 3; i >= 0; i--) { -+ if (switch_port_tab[i]) { -+ struct port *port = switch_port_tab[i]; -+ struct net_device *dev = port->netdev; -+ unregister_netdev(dev); -+ phy_disconnect(port->phydev); -+ switch_port_tab[i] = 0; -+ free_netdev(dev); -+ } -+ } -+ -+ release_resource(sw->mem_res); -+ free_netdev(napi_dev); -+ return 0; -+} -+ -+static struct platform_driver cns3xxx_eth_driver = { -+ .driver.name = DRV_NAME, -+ .probe = eth_init_one, -+ .remove = eth_remove_one, -+}; -+ -+static int __init eth_init_module(void) -+{ -+ int err; -+ if ((err = cns3xxx_mdio_register())) -+ return err; -+ return platform_driver_register(&cns3xxx_eth_driver); -+} -+ -+static void __exit eth_cleanup_module(void) -+{ -+ platform_driver_unregister(&cns3xxx_eth_driver); -+ cns3xxx_mdio_remove(); -+} -+ -+module_init(eth_init_module); -+module_exit(eth_cleanup_module); -+ -+MODULE_AUTHOR("Chris Lang"); -+MODULE_DESCRIPTION("Cavium CNS3xxx Ethernet driver"); -+MODULE_LICENSE("GPL v2"); -+MODULE_ALIAS("platform:cns3xxx_eth"); ---- a/drivers/net/ethernet/Kconfig -+++ b/drivers/net/ethernet/Kconfig -@@ -32,6 +32,7 @@ source "drivers/net/ethernet/calxeda/Kco - source "drivers/net/ethernet/chelsio/Kconfig" - source "drivers/net/ethernet/cirrus/Kconfig" - source "drivers/net/ethernet/cisco/Kconfig" -+source "drivers/net/ethernet/cavium/Kconfig" - source "drivers/net/ethernet/davicom/Kconfig" - - config DNET ---- a/drivers/net/ethernet/Makefile -+++ b/drivers/net/ethernet/Makefile -@@ -15,6 +15,7 @@ obj-$(CONFIG_NET_BFIN) += adi/ - obj-$(CONFIG_NET_VENDOR_BROADCOM) += broadcom/ - obj-$(CONFIG_NET_VENDOR_BROCADE) += brocade/ - obj-$(CONFIG_NET_CALXEDA_XGMAC) += calxeda/ -+obj-$(CONFIG_NET_VENDOR_CAVIUM) += cavium/ - obj-$(CONFIG_NET_VENDOR_CHELSIO) += chelsio/ - obj-$(CONFIG_NET_VENDOR_CIRRUS) += cirrus/ - obj-$(CONFIG_NET_VENDOR_CISCO) += cisco/ ---- /dev/null -+++ b/drivers/net/ethernet/cavium/Kconfig -@@ -0,0 +1,24 @@ -+config NET_VENDOR_CAVIUM -+ bool "Cavium devices" -+ default y -+ depends on ARCH_CNS3XXX -+ ---help--- -+ If you have a network (Ethernet) chipset belonging to this class, -+ say Y. -+ -+ Note that the answer to this question does not directly affect -+ the kernel: saying N will just case the configurator to skip all -+ the questions regarding AMD chipsets. If you say Y, you will be asked -+ for your specific chipset/driver in the following questions. -+ -+if NET_VENDOR_CAVIUM -+ -+config CNS3XXX_ETH -+ tristate "Cavium CNS3xxx Ethernet support" -+ depends on ARCH_CNS3XXX -+ select PHYLIB -+ help -+ Say Y here if you want to use built-in Ethernet ports -+ on CNS3XXX processor. -+ -+endif ---- /dev/null -+++ b/drivers/net/ethernet/cavium/Makefile -@@ -0,0 +1,5 @@ -+# -+# Makefile for the Cavium ethernet device drivers. -+# -+ -+obj-$(CONFIG_CNS3XXX_ETH) += cns3xxx_eth.o diff --git a/target/linux/cns3xxx/patches-3.3/400-ethernet_fix_tx_csum_offload.patch b/target/linux/cns3xxx/patches-3.3/400-ethernet_fix_tx_csum_offload.patch deleted file mode 100644 index e5dc838b9c..0000000000 --- a/target/linux/cns3xxx/patches-3.3/400-ethernet_fix_tx_csum_offload.patch +++ /dev/null @@ -1,89 +0,0 @@ ---- a/drivers/net/ethernet/cavium/cns3xxx_eth.c -+++ b/drivers/net/ethernet/cavium/cns3xxx_eth.c -@@ -666,6 +666,7 @@ static int eth_xmit(struct sk_buff *skb, - int nr_frags = skb_shinfo(skb)->nr_frags; - struct skb_frag_struct *frag; - unsigned int i; -+ u32 config0 = 0; - - if (pmap == 8) - pmap = (1 << 4); -@@ -691,6 +692,10 @@ static int eth_xmit(struct sk_buff *skb, - - spin_unlock(&tx_lock); - -+ config0 = FORCE_ROUTE; -+ if (skb->ip_summed == CHECKSUM_PARTIAL) -+ config0 |= UDP_CHECKSUM | TCP_CHECKSUM; -+ - if (!nr_frags) { - tx_desc = &(tx_ring)->desc[index]; - -@@ -704,23 +709,14 @@ static int eth_xmit(struct sk_buff *skb, - tx_ring->phys_tab[index] = phys; - - tx_ring->buff_tab[index] = skb; -- if (index == TX_DESCS - 1) { -- tx_desc->config0 = END_OF_RING | FIRST_SEGMENT | LAST_SEGMENT | -- FORCE_ROUTE | IP_CHECKSUM | UDP_CHECKSUM | -- TCP_CHECKSUM | len; -- } else { -- tx_desc->config0 = FIRST_SEGMENT | LAST_SEGMENT | -- FORCE_ROUTE | IP_CHECKSUM | UDP_CHECKSUM | -- TCP_CHECKSUM | len; -- } -+ config0 |= FIRST_SEGMENT | LAST_SEGMENT; - } else { -- unsigned int config; -- - index = ((index + nr_frags) % TX_DESCS); - tx_desc = &(tx_ring)->desc[index]; - - /* fragments */ - for (i = nr_frags; i > 0; i--) { -+ u32 config; - void *addr; - - frag = &skb_shinfo(skb)->frags[i-1]; -@@ -735,8 +731,7 @@ static int eth_xmit(struct sk_buff *skb, - tx_desc->pmap = pmap; - tx_ring->phys_tab[index] = phys; - -- config = FORCE_ROUTE | IP_CHECKSUM | UDP_CHECKSUM | -- TCP_CHECKSUM | len; -+ config = config0 | len; - if (i == nr_frags) { - config |= LAST_SEGMENT; - tx_ring->buff_tab[index] = skb; -@@ -757,24 +752,19 @@ static int eth_xmit(struct sk_buff *skb, - /* header */ - len = skb->len - skb->data_len; - -- phys = dma_map_single(NULL, skb->data, len, -- DMA_TO_DEVICE); -+ phys = dma_map_single(NULL, skb->data, len, DMA_TO_DEVICE); - - tx_desc->sdp = phys; - tx_desc->pmap = pmap; - tx_ring->phys_tab[index] = phys; -- -- if (index == TX_DESCS - 1) { -- tx_desc->config0 = END_OF_RING | FIRST_SEGMENT | -- FORCE_ROUTE | IP_CHECKSUM | UDP_CHECKSUM | -- TCP_CHECKSUM | len; -- } else { -- tx_desc->config0 = FIRST_SEGMENT | -- FORCE_ROUTE | IP_CHECKSUM | UDP_CHECKSUM | -- TCP_CHECKSUM | len; -- } -+ config0 |= FIRST_SEGMENT; - } - -+ if (index == TX_DESCS - 1) -+ config0 |= END_OF_RING; -+ -+ tx_desc->config0 = config0 | len; -+ - mb(); - - spin_lock(&tx_lock); diff --git a/target/linux/cns3xxx/patches-3.3/410-ethernet_fix_jumbo_frame.patch b/target/linux/cns3xxx/patches-3.3/410-ethernet_fix_jumbo_frame.patch deleted file mode 100644 index 3365523c7c..0000000000 --- a/target/linux/cns3xxx/patches-3.3/410-ethernet_fix_jumbo_frame.patch +++ /dev/null @@ -1,506 +0,0 @@ ---- a/drivers/net/ethernet/cavium/cns3xxx_eth.c -+++ b/drivers/net/ethernet/cavium/cns3xxx_eth.c -@@ -26,15 +26,21 @@ - - #define DRV_NAME "cns3xxx_eth" - --#define RX_DESCS 512 --#define TX_DESCS 512 --#define SKB_DMA_REALIGN ((PAGE_SIZE - NET_SKB_PAD) % SMP_CACHE_BYTES) -+#define RX_DESCS 128 -+#define TX_DESCS 128 - - #define RX_POOL_ALLOC_SIZE (sizeof(struct rx_desc) * RX_DESCS) - #define TX_POOL_ALLOC_SIZE (sizeof(struct tx_desc) * TX_DESCS) - #define REGS_SIZE 336 --#define MAX_MRU (1536 + SKB_DMA_REALIGN) --#define CNS3XXX_MAX_MTU (1536) -+ -+#define RX_BUFFER_ALIGN 64 -+#define RX_BUFFER_ALIGN_MASK (~(RX_BUFFER_ALIGN - 1)) -+ -+#define SKB_HEAD_ALIGN (((PAGE_SIZE - NET_SKB_PAD) % RX_BUFFER_ALIGN) + NET_SKB_PAD + NET_IP_ALIGN) -+#define RX_SEGMENT_ALLOC_SIZE 4096 -+#define RX_SEGMENT_BUFSIZE (SKB_WITH_OVERHEAD(RX_SEGMENT_ALLOC_SIZE)) -+#define RX_SEGMENT_MRU (((RX_SEGMENT_BUFSIZE - SKB_HEAD_ALIGN) & RX_BUFFER_ALIGN_MASK) - NET_IP_ALIGN) -+#define MAX_MTU 9500 - - #define NAPI_WEIGHT 64 - -@@ -266,7 +272,7 @@ struct _rx_ring { - struct rx_desc *desc; - dma_addr_t phys_addr; - struct rx_desc *cur_addr; -- struct sk_buff *buff_tab[RX_DESCS]; -+ void *buff_tab[RX_DESCS]; - unsigned int phys_tab[RX_DESCS]; - u32 cur_index; - u32 alloc_index; -@@ -280,6 +286,8 @@ struct sw { - struct cns3xxx_plat_info *plat; - struct _tx_ring *tx_ring; - struct _rx_ring *rx_ring; -+ struct sk_buff *frag_first; -+ struct sk_buff *frag_last; - }; - - struct port { -@@ -500,37 +508,35 @@ static void cns3xxx_alloc_rx_buf(struct - struct _rx_ring *rx_ring = sw->rx_ring; - unsigned int i = rx_ring->alloc_index; - struct rx_desc *desc = &(rx_ring)->desc[i]; -- struct sk_buff *skb; -+ void *buf; - unsigned int phys; - - for (received += rx_ring->alloc_count; received > 0; received--) { -- if ((skb = dev_alloc_skb(MAX_MRU))) { -- if (SKB_DMA_REALIGN) -- skb_reserve(skb, SKB_DMA_REALIGN); -- skb_reserve(skb, NET_IP_ALIGN); -- phys = dma_map_single(NULL, skb->data, -- CNS3XXX_MAX_MTU, DMA_FROM_DEVICE); -- if (dma_mapping_error(NULL, phys)) { -- dev_kfree_skb(skb); -- /* Failed to map, better luck next time */ -- goto out;; -- } -- desc->sdp = phys; -- } else { -- /* Failed to allocate skb, try again next time */ -+ buf = kzalloc(RX_SEGMENT_ALLOC_SIZE, GFP_ATOMIC); -+ if (!buf) -+ goto out; -+ -+ phys = dma_map_single(NULL, buf + SKB_HEAD_ALIGN, -+ RX_SEGMENT_MRU, DMA_FROM_DEVICE); -+ if (dma_mapping_error(NULL, phys)) { -+ kfree(buf); - goto out; - } - -+ desc->sdl = RX_SEGMENT_MRU; -+ desc->sdp = phys; -+ - /* put the new buffer on RX-free queue */ -- rx_ring->buff_tab[i] = skb; -+ rx_ring->buff_tab[i] = buf; - rx_ring->phys_tab[i] = phys; - if (i == RX_DESCS - 1) { - i = 0; - desc->config0 = END_OF_RING | FIRST_SEGMENT | -- LAST_SEGMENT | CNS3XXX_MAX_MTU; -+ LAST_SEGMENT | RX_SEGMENT_MRU; - desc = &(rx_ring)->desc[i]; - } else { -- desc->config0 = FIRST_SEGMENT | LAST_SEGMENT | CNS3XXX_MAX_MTU; -+ desc->config0 = FIRST_SEGMENT | LAST_SEGMENT | -+ RX_SEGMENT_MRU; - i++; - desc++; - } -@@ -579,7 +585,6 @@ static void clear_tx_desc(struct sw *sw) - static int eth_poll(struct napi_struct *napi, int budget) - { - struct sw *sw = container_of(napi, struct sw, napi); -- struct net_device *dev; - struct _rx_ring *rx_ring = sw->rx_ring; - int received = 0; - unsigned int length; -@@ -588,49 +593,82 @@ static int eth_poll(struct napi_struct * - - while (desc->cown) { - struct sk_buff *skb; -+ int reserve = SKB_HEAD_ALIGN; - - if (received >= budget) - break; - -- skb = rx_ring->buff_tab[i]; -+ /* process received frame */ -+ dma_unmap_single(NULL, rx_ring->phys_tab[i], -+ RX_SEGMENT_MRU, DMA_FROM_DEVICE); -+ -+ skb = build_skb(rx_ring->buff_tab[i]); -+ if (!skb) -+ break; - -- dev = switch_port_tab[desc->sp]->netdev; -+ skb->dev = switch_port_tab[desc->sp]->netdev; - - length = desc->sdl; -- /* process received frame */ -- dma_unmap_single(&dev->dev, rx_ring->phys_tab[i], -- length, DMA_FROM_DEVICE); -+ if (desc->fsd && !desc->lsd) -+ length = RX_SEGMENT_MRU; -+ -+ if (!desc->fsd) { -+ reserve -= NET_IP_ALIGN; -+ if (!desc->lsd) -+ length += NET_IP_ALIGN; -+ } - -+ skb_reserve(skb, reserve); - skb_put(skb, length); - -- skb->dev = dev; -- skb->protocol = eth_type_trans(skb, dev); -+ if (!sw->frag_first) -+ sw->frag_first = skb; -+ else { -+ if (sw->frag_first == sw->frag_last) -+ skb_frag_add_head(sw->frag_first, skb); -+ else -+ sw->frag_last->next = skb; -+ sw->frag_first->len += skb->len; -+ sw->frag_first->data_len += skb->len; -+ sw->frag_first->truesize += skb->truesize; -+ } -+ sw->frag_last = skb; - -- dev->stats.rx_packets++; -- dev->stats.rx_bytes += length; -+ if (desc->lsd) { -+ struct net_device *dev; -+ -+ skb = sw->frag_first; -+ dev = skb->dev; -+ skb->protocol = eth_type_trans(skb, dev); -+ -+ dev->stats.rx_packets++; -+ dev->stats.rx_bytes += skb->len; -+ -+ /* RX Hardware checksum offload */ -+ skb->ip_summed = CHECKSUM_NONE; -+ switch (desc->prot) { -+ case 1: -+ case 2: -+ case 5: -+ case 6: -+ case 13: -+ case 14: -+ if (desc->l4f) -+ break; - -- /* RX Hardware checksum offload */ -- switch (desc->prot) { -- case 1: -- case 2: -- case 5: -- case 6: -- case 13: -- case 14: -- if (desc->l4f) -- skb->ip_summed = CHECKSUM_NONE; -- else - skb->ip_summed = CHECKSUM_UNNECESSARY; -- break; -- default: -- skb->ip_summed = CHECKSUM_NONE; -- break; -- } -+ break; -+ default: -+ break; -+ } - -- napi_gro_receive(napi, skb); -+ napi_gro_receive(napi, skb); - -- received++; -+ sw->frag_first = NULL; -+ sw->frag_last = NULL; -+ } - -+ received++; - if (++i == RX_DESCS) { - i = 0; - desc = &(rx_ring)->desc[i]; -@@ -653,42 +691,60 @@ static int eth_poll(struct napi_struct * - return received; - } - -+static void eth_set_desc(struct _tx_ring *tx_ring, int index, int index_last, -+ void *data, int len, u32 config0, u32 pmap) -+{ -+ struct tx_desc *tx_desc = &(tx_ring)->desc[index]; -+ unsigned int phys; -+ -+ phys = dma_map_single(NULL, data, len, DMA_TO_DEVICE); -+ tx_desc->sdp = phys; -+ tx_desc->pmap = pmap; -+ tx_ring->phys_tab[index] = phys; -+ -+ config0 |= len; -+ if (index == TX_DESCS - 1) -+ config0 |= END_OF_RING; -+ if (index == index_last) -+ config0 |= LAST_SEGMENT; -+ -+ mb(); -+ tx_desc->config0 = config0; -+} -+ - static int eth_xmit(struct sk_buff *skb, struct net_device *dev) - { - struct port *port = netdev_priv(dev); - struct sw *sw = port->sw; - struct _tx_ring *tx_ring = sw->tx_ring; -- struct tx_desc *tx_desc; -- int index; -- int len; -+ struct sk_buff *skb1; - char pmap = (1 << port->id); -- unsigned int phys; - int nr_frags = skb_shinfo(skb)->nr_frags; -- struct skb_frag_struct *frag; -+ int nr_desc = nr_frags; -+ int index0, index, index_last; -+ int len0; - unsigned int i; -- u32 config0 = 0; -+ u32 config0; - - if (pmap == 8) - pmap = (1 << 4); - -- if (skb->len > CNS3XXX_MAX_MTU) { -- dev_kfree_skb(skb); -- dev->stats.tx_errors++; -- return NETDEV_TX_OK; -- } -+ skb_walk_frags(skb, skb1) -+ nr_desc++; - - spin_lock(&tx_lock); - -- if ((tx_ring->num_used + nr_frags) >= TX_DESCS) { -+ if ((tx_ring->num_used + nr_desc + 1) >= TX_DESCS) { - clear_tx_desc(sw); -- if ((tx_ring->num_used + nr_frags) >= TX_DESCS) { -+ if ((tx_ring->num_used + nr_desc + 1) >= TX_DESCS) { - spin_unlock(&tx_lock); - return NETDEV_TX_BUSY; - } - } - -- index = tx_ring->cur_index; -- tx_ring->cur_index = ((tx_ring->cur_index + nr_frags + 1) % TX_DESCS); -+ index = index0 = tx_ring->cur_index; -+ index_last = (index0 + nr_desc) % TX_DESCS; -+ tx_ring->cur_index = (index_last + 1) % TX_DESCS; - - spin_unlock(&tx_lock); - -@@ -696,79 +752,41 @@ static int eth_xmit(struct sk_buff *skb, - if (skb->ip_summed == CHECKSUM_PARTIAL) - config0 |= UDP_CHECKSUM | TCP_CHECKSUM; - -- if (!nr_frags) { -- tx_desc = &(tx_ring)->desc[index]; -- -- len = skb->len; -+ len0 = skb->len; - -- phys = dma_map_single(NULL, skb->data, len, -- DMA_TO_DEVICE); -+ /* fragments */ -+ for (i = 0; i < nr_frags; i++) { -+ struct skb_frag_struct *frag; -+ void *addr; - -- tx_desc->sdp = phys; -- tx_desc->pmap = pmap; -- tx_ring->phys_tab[index] = phys; -+ index = (index + 1) % TX_DESCS; - -- tx_ring->buff_tab[index] = skb; -- config0 |= FIRST_SEGMENT | LAST_SEGMENT; -- } else { -- index = ((index + nr_frags) % TX_DESCS); -- tx_desc = &(tx_ring)->desc[index]; -+ frag = &skb_shinfo(skb)->frags[i]; -+ addr = page_address(skb_frag_page(frag)) + frag->page_offset; - -- /* fragments */ -- for (i = nr_frags; i > 0; i--) { -- u32 config; -- void *addr; -- -- frag = &skb_shinfo(skb)->frags[i-1]; -- len = frag->size; -- -- addr = page_address(skb_frag_page(frag)) + -- frag->page_offset; -- phys = dma_map_single(NULL, addr, len, DMA_TO_DEVICE); -- -- tx_desc->sdp = phys; -- -- tx_desc->pmap = pmap; -- tx_ring->phys_tab[index] = phys; -- -- config = config0 | len; -- if (i == nr_frags) { -- config |= LAST_SEGMENT; -- tx_ring->buff_tab[index] = skb; -- } -- if (index == TX_DESCS - 1) -- config |= END_OF_RING; -- tx_desc->config0 = config; -- -- if (index == 0) { -- index = TX_DESCS - 1; -- tx_desc = &(tx_ring)->desc[index]; -- } else { -- index--; -- tx_desc--; -- } -- } -+ eth_set_desc(tx_ring, index, index_last, addr, frag->size, -+ config0, pmap); -+ } - -- /* header */ -- len = skb->len - skb->data_len; -+ if (nr_frags) -+ len0 = skb->len - skb->data_len; - -- phys = dma_map_single(NULL, skb->data, len, DMA_TO_DEVICE); -+ skb_walk_frags(skb, skb1) { -+ index = (index + 1) % TX_DESCS; -+ len0 -= skb1->len; - -- tx_desc->sdp = phys; -- tx_desc->pmap = pmap; -- tx_ring->phys_tab[index] = phys; -- config0 |= FIRST_SEGMENT; -+ eth_set_desc(tx_ring, index, index_last, skb1->data, skb1->len, -+ config0, pmap); - } - -- if (index == TX_DESCS - 1) -- config0 |= END_OF_RING; -- -- tx_desc->config0 = config0 | len; -+ tx_ring->buff_tab[index0] = skb; -+ eth_set_desc(tx_ring, index0, index_last, skb->data, len0, -+ config0 | FIRST_SEGMENT, pmap); - - mb(); - - spin_lock(&tx_lock); -- tx_ring->num_used += nr_frags + 1; -+ tx_ring->num_used += nr_desc + 1; - spin_unlock(&tx_lock); - - dev->stats.tx_packets++; -@@ -849,24 +867,24 @@ static int init_rings(struct sw *sw) - /* Setup RX buffers */ - for (i = 0; i < RX_DESCS; i++) { - struct rx_desc *desc = &(rx_ring)->desc[i]; -- struct sk_buff *skb; -- if (!(skb = dev_alloc_skb(MAX_MRU))) -+ void *buf; -+ -+ buf = kzalloc(RX_SEGMENT_ALLOC_SIZE, GFP_KERNEL); -+ if (!buf) - return -ENOMEM; -- if (SKB_DMA_REALIGN) -- skb_reserve(skb, SKB_DMA_REALIGN); -- skb_reserve(skb, NET_IP_ALIGN); -- desc->sdl = CNS3XXX_MAX_MTU; -+ -+ desc->sdl = RX_SEGMENT_MRU; - if (i == (RX_DESCS - 1)) - desc->eor = 1; - desc->fsd = 1; - desc->lsd = 1; - -- desc->sdp = dma_map_single(NULL, skb->data, -- CNS3XXX_MAX_MTU, DMA_FROM_DEVICE); -- if (dma_mapping_error(NULL, desc->sdp)) { -+ desc->sdp = dma_map_single(NULL, buf + SKB_HEAD_ALIGN, -+ RX_SEGMENT_MRU, DMA_FROM_DEVICE); -+ if (dma_mapping_error(NULL, desc->sdp)) - return -EIO; -- } -- rx_ring->buff_tab[i] = skb; -+ -+ rx_ring->buff_tab[i] = buf; - rx_ring->phys_tab[i] = desc->sdp; - desc->cown = 0; - } -@@ -905,12 +923,13 @@ static void destroy_rings(struct sw *sw) - struct _rx_ring *rx_ring = sw->rx_ring; - struct rx_desc *desc = &(rx_ring)->desc[i]; - struct sk_buff *skb = sw->rx_ring->buff_tab[i]; -- if (skb) { -- dma_unmap_single(NULL, -- desc->sdp, -- CNS3XXX_MAX_MTU, DMA_FROM_DEVICE); -- dev_kfree_skb(skb); -- } -+ -+ if (!skb) -+ continue; -+ -+ dma_unmap_single(NULL, desc->sdp, RX_SEGMENT_MRU, -+ DMA_FROM_DEVICE); -+ dev_kfree_skb(skb); - } - dma_pool_free(rx_dma_pool, sw->rx_ring->desc, sw->rx_ring->phys_addr); - dma_pool_destroy(rx_dma_pool); -@@ -1085,13 +1104,22 @@ static int eth_set_mac(struct net_device - return 0; - } - -+static int cns3xxx_change_mtu(struct net_device *dev, int new_mtu) -+{ -+ if (new_mtu > MAX_MTU) -+ return -EINVAL; -+ -+ dev->mtu = new_mtu; -+ return 0; -+} -+ - static const struct net_device_ops cns3xxx_netdev_ops = { - .ndo_open = eth_open, - .ndo_stop = eth_close, - .ndo_start_xmit = eth_xmit, - .ndo_set_rx_mode = eth_rx_mode, - .ndo_do_ioctl = eth_ioctl, -- .ndo_change_mtu = eth_change_mtu, -+ .ndo_change_mtu = cns3xxx_change_mtu, - .ndo_set_mac_address = eth_set_mac, - .ndo_validate_addr = eth_validate_addr, - }; -@@ -1111,7 +1139,7 @@ static int __devinit eth_init_one(struct - if (!(napi_dev = alloc_etherdev(sizeof(struct sw)))) - return -ENOMEM; - strcpy(napi_dev->name, "switch%d"); -- napi_dev->features = NETIF_F_IP_CSUM | NETIF_F_SG; -+ napi_dev->features = NETIF_F_IP_CSUM | NETIF_F_SG | NETIF_F_FRAGLIST; - - SET_NETDEV_DEV(napi_dev, &pdev->dev); - sw = netdev_priv(napi_dev); -@@ -1124,6 +1152,10 @@ static int __devinit eth_init_one(struct - goto err_free; - } - -+ temp = __raw_readl(&sw->regs->phy_auto_addr); -+ temp |= (3 << 30); /* maximum frame length: 9600 bytes */ -+ __raw_writel(temp, &sw->regs->phy_auto_addr); -+ - for (i = 0; i < 4; i++) { - temp = __raw_readl(&sw->regs->mac_cfg[i]); - temp |= (PORT_DISABLE); -@@ -1185,7 +1217,7 @@ static int __devinit eth_init_one(struct - dev->netdev_ops = &cns3xxx_netdev_ops; - dev->ethtool_ops = &cns3xxx_ethtool_ops; - dev->tx_queue_len = 1000; -- dev->features = NETIF_F_IP_CSUM | NETIF_F_SG; -+ dev->features = NETIF_F_IP_CSUM | NETIF_F_SG | NETIF_F_FRAGLIST; - - switch_port_tab[port->id] = port; - memcpy(dev->dev_addr, &plat->hwaddr[i], ETH_ALEN); diff --git a/target/linux/cns3xxx/patches-3.3/420-ethernet_optimize_rx_offload.patch b/target/linux/cns3xxx/patches-3.3/420-ethernet_optimize_rx_offload.patch deleted file mode 100644 index 67929ef895..0000000000 --- a/target/linux/cns3xxx/patches-3.3/420-ethernet_optimize_rx_offload.patch +++ /dev/null @@ -1,27 +0,0 @@ ---- a/drivers/net/ethernet/cavium/cns3xxx_eth.c -+++ b/drivers/net/ethernet/cavium/cns3xxx_eth.c -@@ -653,17 +653,17 @@ static int eth_poll(struct napi_struct * - case 6: - case 13: - case 14: -- if (desc->l4f) -+ if (!desc->l4f) { -+ skb->ip_summed = CHECKSUM_UNNECESSARY; -+ napi_gro_receive(napi, skb); - break; -- -- skb->ip_summed = CHECKSUM_UNNECESSARY; -- break; -+ } -+ /* fall through */ - default: -- break; -+ netif_receive_skb(skb); -+ break; - } - -- napi_gro_receive(napi, skb); -- - sw->frag_first = NULL; - sw->frag_last = NULL; - } diff --git a/target/linux/cns3xxx/patches-3.3/430-ethernet_fix_tx_completion.patch b/target/linux/cns3xxx/patches-3.3/430-ethernet_fix_tx_completion.patch deleted file mode 100644 index 9a16c99f8f..0000000000 --- a/target/linux/cns3xxx/patches-3.3/430-ethernet_fix_tx_completion.patch +++ /dev/null @@ -1,110 +0,0 @@ ---- a/drivers/net/ethernet/cavium/cns3xxx_eth.c -+++ b/drivers/net/ethernet/cavium/cns3xxx_eth.c -@@ -28,6 +28,7 @@ - - #define RX_DESCS 128 - #define TX_DESCS 128 -+#define TX_DESC_RESERVE 20 - - #define RX_POOL_ALLOC_SIZE (sizeof(struct rx_desc) * RX_DESCS) - #define TX_POOL_ALLOC_SIZE (sizeof(struct tx_desc) * TX_DESCS) -@@ -266,6 +267,7 @@ struct _tx_ring { - u32 cur_index; - int num_used; - int num_count; -+ bool stopped; - }; - - struct _rx_ring { -@@ -546,7 +548,34 @@ out: - rx_ring->alloc_index = i; - } - --static void clear_tx_desc(struct sw *sw) -+static void eth_check_num_used(struct _tx_ring *tx_ring) -+{ -+ bool stop = false; -+ int i; -+ -+ if (tx_ring->num_used >= TX_DESCS - TX_DESC_RESERVE) -+ stop = true; -+ -+ if (tx_ring->stopped == stop) -+ return; -+ -+ tx_ring->stopped = stop; -+ for (i = 0; i < 4; i++) { -+ struct port *port = switch_port_tab[i]; -+ struct net_device *dev; -+ -+ if (!port) -+ continue; -+ -+ dev = port->netdev; -+ if (stop) -+ netif_stop_queue(dev); -+ else -+ netif_wake_queue(dev); -+ } -+} -+ -+static void eth_complete_tx(struct sw *sw) - { - struct _tx_ring *tx_ring = sw->tx_ring; - struct tx_desc *desc; -@@ -555,9 +584,6 @@ static void clear_tx_desc(struct sw *sw) - int num_used = tx_ring->num_used; - struct sk_buff *skb; - -- if (num_used < (TX_DESCS >> 1)) -- return; -- - index = tx_ring->free_index; - desc = &(tx_ring)->desc[index]; - for (i = 0; i < num_used; i++) { -@@ -580,6 +606,7 @@ static void clear_tx_desc(struct sw *sw) - } - tx_ring->free_index = index; - tx_ring->num_used -= i; -+ eth_check_num_used(tx_ring); - } - - static int eth_poll(struct napi_struct *napi, int budget) -@@ -688,6 +715,10 @@ static int eth_poll(struct napi_struct * - - enable_rx_dma(sw); - -+ spin_lock_bh(&tx_lock); -+ eth_complete_tx(sw); -+ spin_unlock_bh(&tx_lock); -+ - return received; - } - -@@ -732,21 +763,19 @@ static int eth_xmit(struct sk_buff *skb, - skb_walk_frags(skb, skb1) - nr_desc++; - -- spin_lock(&tx_lock); -+ spin_lock_bh(&tx_lock); - -+ eth_complete_tx(sw); - if ((tx_ring->num_used + nr_desc + 1) >= TX_DESCS) { -- clear_tx_desc(sw); -- if ((tx_ring->num_used + nr_desc + 1) >= TX_DESCS) { -- spin_unlock(&tx_lock); -- return NETDEV_TX_BUSY; -- } -+ spin_unlock_bh(&tx_lock); -+ return NETDEV_TX_BUSY; - } - - index = index0 = tx_ring->cur_index; - index_last = (index0 + nr_desc) % TX_DESCS; - tx_ring->cur_index = (index_last + 1) % TX_DESCS; - -- spin_unlock(&tx_lock); -+ spin_unlock_bh(&tx_lock); - - config0 = FORCE_ROUTE; - if (skb->ip_summed == CHECKSUM_PARTIAL) diff --git a/target/linux/cns3xxx/patches-3.3/460-cns3xxx_fiq_support.patch b/target/linux/cns3xxx/patches-3.3/460-cns3xxx_fiq_support.patch index 083cd1d7ca..4e5cb46925 100644 --- a/target/linux/cns3xxx/patches-3.3/460-cns3xxx_fiq_support.patch +++ b/target/linux/cns3xxx/patches-3.3/460-cns3xxx_fiq_support.patch @@ -1,8 +1,8 @@ --- a/arch/arm/Kconfig +++ b/arch/arm/Kconfig -@@ -375,6 +375,7 @@ config ARCH_CNS3XXX +@@ -374,6 +374,7 @@ config ARCH_CNS3XXX + select MIGHT_HAVE_PCI select PCI_DOMAINS if PCI - select HAVE_ARM_TWD select HAVE_SMP + select FIQ help diff --git a/target/linux/cns3xxx/patches-3.3/600-cns3xxx_ethernet.patch b/target/linux/cns3xxx/patches-3.3/600-cns3xxx_ethernet.patch new file mode 100644 index 0000000000..84548a322c --- /dev/null +++ b/target/linux/cns3xxx/patches-3.3/600-cns3xxx_ethernet.patch @@ -0,0 +1,20 @@ +--- a/drivers/net/ethernet/Kconfig ++++ b/drivers/net/ethernet/Kconfig +@@ -32,6 +32,7 @@ source "drivers/net/ethernet/calxeda/Kco + source "drivers/net/ethernet/chelsio/Kconfig" + source "drivers/net/ethernet/cirrus/Kconfig" + source "drivers/net/ethernet/cisco/Kconfig" ++source "drivers/net/ethernet/cavium/Kconfig" + source "drivers/net/ethernet/davicom/Kconfig" + + config DNET +--- a/drivers/net/ethernet/Makefile ++++ b/drivers/net/ethernet/Makefile +@@ -15,6 +15,7 @@ obj-$(CONFIG_NET_BFIN) += adi/ + obj-$(CONFIG_NET_VENDOR_BROADCOM) += broadcom/ + obj-$(CONFIG_NET_VENDOR_BROCADE) += brocade/ + obj-$(CONFIG_NET_CALXEDA_XGMAC) += calxeda/ ++obj-$(CONFIG_NET_VENDOR_CAVIUM) += cavium/ + obj-$(CONFIG_NET_VENDOR_CHELSIO) += chelsio/ + obj-$(CONFIG_NET_VENDOR_CIRRUS) += cirrus/ + obj-$(CONFIG_NET_VENDOR_CISCO) += cisco/ |