/* * Driver for Marvell Discovery (MV643XX) and Marvell Orion ethernet ports * Copyright (C) 2002 Matthew Dharm * * Based on the 64360 driver from: * Copyright (C) 2002 Rabeeh Khoury * Rabeeh Khoury * * Copyright (C) 2003 PMC-Sierra, Inc., * written by Manish Lachwani * * Copyright (C) 2003 Ralf Baechle * * Copyright (C) 2004-2006 MontaVista Software, Inc. * Dale Farnsworth * * Copyright (C) 2004 Steven J. Hill * * * Copyright (C) 2007-2008 Marvell Semiconductor * Lennert Buytenhek * * This program is free software; you can redistribute it and/or * modify it under the terms of the GNU General Public License * as published by the Free Software Foundation; either version 2 * of the License, or (at your option) any later version. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #define MV643XX_CHECKSUM_OFFLOAD_TX #define MV643XX_NAPI #define MV643XX_TX_FAST_REFILL #undef MV643XX_COAL #define MV643XX_TX_COAL 100 #ifdef MV643XX_COAL #define MV643XX_RX_COAL 100 #endif #ifdef MV643XX_CHECKSUM_OFFLOAD_TX #define MAX_DESCS_PER_SKB (MAX_SKB_FRAGS + 1) #else #define MAX_DESCS_PER_SKB 1 #endif #define ETH_VLAN_HLEN 4 #define ETH_FCS_LEN 4 #define ETH_HW_IP_ALIGN 2 /* hw aligns IP header */ #define ETH_WRAPPER_LEN (ETH_HW_IP_ALIGN + ETH_HLEN + \ ETH_VLAN_HLEN + ETH_FCS_LEN) #define ETH_RX_SKB_SIZE (dev->mtu + ETH_WRAPPER_LEN + \ dma_get_cache_alignment()) /* * Registers shared between all ports. */ #define PHY_ADDR_REG 0x0000 #define SMI_REG 0x0004 /* * Per-port registers. */ #define PORT_CONFIG_REG(p) (0x0400 + ((p) << 10)) #define PORT_CONFIG_EXTEND_REG(p) (0x0404 + ((p) << 10)) #define MAC_ADDR_LOW(p) (0x0414 + ((p) << 10)) #define MAC_ADDR_HIGH(p) (0x0418 + ((p) << 10)) #define SDMA_CONFIG_REG(p) (0x041c + ((p) << 10)) #define PORT_SERIAL_CONTROL_REG(p) (0x043c + ((p) << 10)) #define PORT_STATUS_REG(p) (0x0444 + ((p) << 10)) #define TRANSMIT_QUEUE_COMMAND_REG(p) (0x0448 + ((p) << 10)) #define MAXIMUM_TRANSMIT_UNIT(p) (0x0458 + ((p) << 10)) #define INTERRUPT_CAUSE_REG(p) (0x0460 + ((p) << 10)) #define INTERRUPT_CAUSE_EXTEND_REG(p) (0x0464 + ((p) << 10)) #define INTERRUPT_MASK_REG(p) (0x0468 + ((p) << 10)) #define INTERRUPT_EXTEND_MASK_REG(p) (0x046c + ((p) << 10)) #define TX_FIFO_URGENT_THRESHOLD_REG(p) (0x0474 + ((p) << 10)) #define RX_CURRENT_QUEUE_DESC_PTR_0(p) (0x060c + ((p) << 10)) #define RECEIVE_QUEUE_COMMAND_REG(p) (0x0680 + ((p) << 10)) #define TX_CURRENT_QUEUE_DESC_PTR_0(p) (0x06c0 + ((p) << 10)) #define MIB_COUNTERS_BASE(p) (0x1000 + ((p) << 7)) #define DA_FILTER_SPECIAL_MULTICAST_TABLE_BASE(p) (0x1400 + ((p) << 10)) #define DA_FILTER_OTHER_MULTICAST_TABLE_BASE(p) (0x1500 + ((p) << 10)) #define DA_FILTER_UNICAST_TABLE_BASE(p) (0x1600 + ((p) << 10)) /* These macros describe Ethernet Port configuration reg (Px_cR) bits */ #define UNICAST_NORMAL_MODE (0 << 0) #define UNICAST_PROMISCUOUS_MODE (1 << 0) #define DEFAULT_RX_QUEUE(queue) ((queue) << 1) #define DEFAULT_RX_ARP_QUEUE(queue) ((queue) << 4) #define RECEIVE_BC_IF_NOT_IP_OR_ARP (0 << 7) #define REJECT_BC_IF_NOT_IP_OR_ARP (1 << 7) #define RECEIVE_BC_IF_IP (0 << 8) #define REJECT_BC_IF_IP (1 << 8) #define RECEIVE_BC_IF_ARP (0 << 9) #define REJECT_BC_IF_ARP (1 << 9) #define TX_AM_NO_UPDATE_ERROR_SUMMARY (1 << 12) #define CAPTURE_TCP_FRAMES_DIS (0 << 14) #define CAPTURE_TCP_FRAMES_EN (1 << 14) #define CAPTURE_UDP_FRAMES_DIS (0 << 15) #define CAPTURE_UDP_FRAMES_EN (1 << 15) #define DEFAULT_RX_TCP_QUEUE(queue) ((queue) << 16) #define DEFAULT_RX_UDP_QUEUE(queue) ((queue) << 19) #define DEFAULT_RX_BPDU_QUEUE(queue) ((queue) << 22) #define PORT_CONFIG_DEFAULT_VALUE \ UNICAST_NORMAL_MODE | \ DEFAULT_RX_QUEUE(0) | \ DEFAULT_RX_ARP_QUEUE(0) | \ RECEIVE_BC_IF_NOT_IP_OR_ARP | \ RECEIVE_BC_IF_IP | \ RECEIVE_BC_IF_ARP | \ CAPTURE_TCP_FRAMES_DIS | \ CAPTURE_UDP_FRAMES_DIS | \ DEFAULT_RX_TCP_QUEUE(0) | \ DEFAULT_RX_UDP_QUEUE(0) | \ DEFAULT_RX_BPDU_QUEUE(0) /* These macros describe Ethernet Port configuration extend reg (Px_cXR) bits*/ #define CLASSIFY_EN (1 << 0) #define SPAN_BPDU_PACKETS_AS_NORMAL (0 << 1) #define SPAN_BPDU_PACKETS_TO_RX_QUEUE_7 (1 << 1) #define PARTITION_DISABLE (0 << 2) #define PARTITION_ENABLE (1 << 2) #define PORT_CONFIG_EXTEND_DEFAULT_VALUE \ SPAN_BPDU_PACKETS_AS_NORMAL | \ PARTITION_DISABLE /* These macros describe Ethernet Port Sdma configuration reg (SDCR) bits */ #define RIFB (1 << 0) #define RX_BURST_SIZE_1_64BIT (0 << 1) #define RX_BURST_SIZE_2_64BIT (1 << 1) #define RX_BURST_SIZE_4_64BIT (2 << 1) #define RX_BURST_SIZE_8_64BIT (3 << 1) #define RX_BURST_SIZE_16_64BIT (4 << 1) #define BLM_RX_NO_SWAP (1 << 4) #define BLM_RX_BYTE_SWAP (0 << 4) #define BLM_TX_NO_SWAP (1 << 5) #define BLM_TX_BYTE_SWAP (0 << 5) #define DESCRIPTORS_BYTE_SWAP (1 << 6) #define DESCRIPTORS_NO_SWAP (0 << 6) #define IPG_INT_RX(value) (((value) & 0x3fff) << 8) #define TX_BURST_SIZE_1_64BIT (0 << 22) #define TX_BURST_SIZE_2_64BIT (1 << 22) #define TX_BURST_SIZE_4_64BIT (2 << 22) #define TX_BURST_SIZE_8_64BIT (3 << 22) #define TX_BURST_SIZE_16_64BIT (4 << 22) #if defined(__BIG_ENDIAN) #define PORT_SDMA_CONFIG_DEFAULT_VALUE \ RX_BURST_SIZE_4_64BIT | \ IPG_INT_RX(0) | \ TX_BURST_SIZE_4_64BIT #elif defined(__LITTLE_ENDIAN) #define PORT_SDMA_CONFIG_DEFAULT_VALUE \ RX_BURST_SIZE_4_64BIT | \ BLM_RX_NO_SWAP | \ BLM_TX_NO_SWAP | \ IPG_INT_RX(0) | \ TX_BURST_SIZE_4_64BIT #else #error One of __BIG_ENDIAN or __LITTLE_ENDIAN must be defined #endif /* These macros describe Ethernet Port serial control reg (PSCR) bits */ #define SERIAL_PORT_DISABLE (0 << 0) #define SERIAL_PORT_ENABLE (1 << 0) #define DO_NOT_FORCE_LINK_PASS (0 << 1) #define FORCE_LINK_PASS (1 << 1) #define ENABLE_AUTO_NEG_FOR_DUPLX (0 << 2) #define DISABLE_AUTO_NEG_FOR_DUPLX (1 << 2) #define ENABLE_AUTO_NEG_FOR_FLOW_CTRL (0 << 3) #define DISABLE_AUTO_NEG_FOR_FLOW_CTRL (1 << 3) #define ADV_NO_FLOW_CTRL (0 << 4) #define ADV_SYMMETRIC_FLOW_CTRL (1 << 4) #define FORCE_FC_MODE_NO_PAUSE_DIS_TX (0 << 5) #define FORCE_FC_MODE_TX_PAUSE_DIS (1 << 5) #define FORCE_BP_MODE_NO_JAM (0 << 7) #define FORCE_BP_MODE_JAM_TX (1 << 7) #define FORCE_BP_MODE_JAM_TX_ON_RX_ERR (2 << 7) #define SERIAL_PORT_CONTROL_RESERVED (1 << 9) #define FORCE_LINK_FAIL (0 << 10) #define DO_NOT_FORCE_LINK_FAIL (1 << 10) #define RETRANSMIT_16_ATTEMPTS (0 << 11) #define RETRANSMIT_FOREVER (1 << 11) #define ENABLE_AUTO_NEG_SPEED_GMII (0 << 13) #define DISABLE_AUTO_NEG_SPEED_GMII (1 << 13) #define DTE_ADV_0 (0 << 14) #define DTE_ADV_1 (1 << 14) #define DISABLE_AUTO_NEG_BYPASS (0 << 15) #define ENABLE_AUTO_NEG_BYPASS (1 << 15) #define AUTO_NEG_NO_CHANGE (0 << 16) #define RESTART_AUTO_NEG (1 << 16) #define MAX_RX_PACKET_1518BYTE (0 << 17) #define MAX_RX_PACKET_1522BYTE (1 << 17) #define MAX_RX_PACKET_1552BYTE (2 << 17) #define MAX_RX_PACKET_9022BYTE (3 << 17) #define MAX_RX_PACKET_9192BYTE (4 << 17) #define MAX_RX_PACKET_9700BYTE (5 << 17) #define MAX_RX_PACKET_MASK (7 << 17) #define CLR_EXT_LOOPBACK (0 << 20) #define SET_EXT_LOOPBACK (1 << 20) #define SET_HALF_DUPLEX_MODE (0 << 21) #define SET_FULL_DUPLEX_MODE (1 << 21) #define DISABLE_FLOW_CTRL_TX_RX_IN_FULL_DUPLEX (0 << 22) #define ENABLE_FLOW_CTRL_TX_RX_IN_FULL_DUPLEX (1 << 22) #define SET_GMII_SPEED_TO_10_100 (0 << 23) #define SET_GMII_SPEED_TO_1000 (1 << 23) #define SET_MII_SPEED_TO_10 (0 << 24) #define SET_MII_SPEED_TO_100 (1 << 24) #define PORT_SERIAL_CONTROL_DEFAULT_VALUE \ DO_NOT_FORCE_LINK_PASS | \ ENABLE_AUTO_NEG_FOR_DUPLX | \ DISABLE_AUTO_NEG_FOR_FLOW_CTRL | \ ADV_SYMMETRIC_FLOW_CTRL | \ FORCE_FC_MODE_NO_PAUSE_DIS_TX | \ FORCE_BP_MODE_NO_JAM | \ (1 << 9) /* reserved */ | \ DO_NOT_FORCE_LINK_FAIL | \ RETRANSMIT_16_ATTEMPTS | \ ENABLE_AUTO_NEG_SPEED_GMII | \ DTE_ADV_0 | \ DISABLE_AUTO_NEG_BYPASS | \ AUTO_NEG_NO_CHANGE | \ MAX_RX_PACKET_9700BYTE | \ CLR_EXT_LOOPBACK | \ SET_FULL_DUPLEX_MODE | \ ENABLE_FLOW_CTRL_TX_RX_IN_FULL_DUPLEX /* These macros describe Ethernet Serial Status reg (PSR) bits */ #define PORT_STATUS_MODE_10_BIT (1 << 0) #define PORT_STATUS_LINK_UP (1 << 1) #define PORT_STATUS_FULL_DUPLEX (1 << 2) #define PORT_STATUS_FLOW_CONTROL (1 << 3) #define PORT_STATUS_GMII_1000 (1 << 4) #define PORT_STATUS_MII_100 (1 << 5) /* PSR bit 6 is undocumented */ #define PORT_STATUS_TX_IN_PROGRESS (1 << 7) #define PORT_STATUS_AUTONEG_BYPASSED (1 << 8) #define PORT_STATUS_PARTITION (1 << 9) #define PORT_STATUS_TX_FIFO_EMPTY (1 << 10) /* PSR bits 11-31 are reserved */ #define PORT_DEFAULT_TRANSMIT_QUEUE_SIZE 800 #define PORT_DEFAULT_RECEIVE_QUEUE_SIZE 400 #define DESC_SIZE 64 #define ETH_RX_QUEUES_ENABLED (1 << 0) /* use only Q0 for receive */ #define ETH_TX_QUEUES_ENABLED (1 << 0) /* use only Q0 for transmit */ #define ETH_INT_CAUSE_RX_DONE (ETH_RX_QUEUES_ENABLED << 2) #define ETH_INT_CAUSE_RX_ERROR (ETH_RX_QUEUES_ENABLED << 9) #define ETH_INT_CAUSE_RX (ETH_INT_CAUSE_RX_DONE | ETH_INT_CAUSE_RX_ERROR) #define ETH_INT_CAUSE_EXT 0x00000002 #define ETH_INT_UNMASK_ALL (ETH_INT_CAUSE_RX | ETH_INT_CAUSE_EXT) #define ETH_INT_CAUSE_TX_DONE (ETH_TX_QUEUES_ENABLED << 0) #define ETH_INT_CAUSE_TX_ERROR (ETH_TX_QUEUES_ENABLED << 8) #define ETH_INT_CAUSE_TX (ETH_INT_CAUSE_TX_DONE | ETH_INT_CAUSE_TX_ERROR) #define ETH_INT_CAUSE_PHY 0x00010000 #define ETH_INT_CAUSE_STATE 0x00100000 #define ETH_INT_UNMASK_ALL_EXT (ETH_INT_CAUSE_TX | ETH_INT_CAUSE_PHY | \ ETH_INT_CAUSE_STATE) #define ETH_INT_MASK_ALL 0x00000000 #define ETH_INT_MASK_ALL_EXT 0x00000000 #define PHY_WAIT_ITERATIONS 1000 /* 1000 iterations * 10uS = 10mS max */ #define PHY_WAIT_MICRO_SECONDS 10 /* Buffer offset from buffer pointer */ #define RX_BUF_OFFSET 0x2 /* Gigabit Ethernet Unit Global Registers */ /* MIB Counters register definitions */ #define ETH_MIB_GOOD_OCTETS_RECEIVED_LOW 0x0 #define ETH_MIB_GOOD_OCTETS_RECEIVED_HIGH 0x4 #define ETH_MIB_BAD_OCTETS_RECEIVED 0x8 #define ETH_MIB_INTERNAL_MAC_TRANSMIT_ERR 0xc #define ETH_MIB_GOOD_FRAMES_RECEIVED 0x10 #define ETH_MIB_BAD_FRAMES_RECEIVED 0x14 #define ETH_MIB_BROADCAST_FRAMES_RECEIVED 0x18 #define ETH_MIB_MULTICAST_FRAMES_RECEIVED 0x1c #define ETH_MIB_FRAMES_64_OCTETS 0x20 #define ETH_MIB_FRAMES_65_TO_127_OCTETS 0x24 #define ETH_MIB_FRAMES_128_TO_255_OCTETS 0x28 #define ETH_MIB_FRAMES_256_TO_511_OCTETS 0x2c #define ETH_MIB_FRAMES_512_TO_1023_OCTETS 0x30 #define ETH_MIB_FRAMES_1024_TO_MAX_OCTETS 0x34 #define ETH_MIB_GOOD_OCTETS_SENT_LOW 0x38 #define ETH_MIB_GOOD_OCTETS_SENT_HIGH 0x3c #define ETH_MIB_GOOD_FRAMES_SENT 0x40 #define ETH_MIB_EXCESSIVE_COLLISION 0x44 #define ETH_MIB_MULTICAST_FRAMES_SENT 0x48 #define ETH_MIB_BROADCAST_FRAMES_SENT 0x4c #define ETH_MIB_UNREC_MAC_CONTROL_RECEIVED 0x50 #define ETH_MIB_FC_SENT 0x54 #define ETH_MIB_GOOD_FC_RECEIVED 0x58 #define ETH_MIB_BAD_FC_RECEIVED 0x5c #define ETH_MIB_UNDERSIZE_RECEIVED 0x60 #define ETH_MIB_FRAGMENTS_RECEIVED 0x64 #define ETH_MIB_OVERSIZE_RECEIVED 0x68 #define ETH_MIB_JABBER_RECEIVED 0x6c #define ETH_MIB_MAC_RECEIVE_ERROR 0x70 #define ETH_MIB_BAD_CRC_EVENT 0x74 #define ETH_MIB_COLLISION 0x78 #define ETH_MIB_LATE_COLLISION 0x7c /* Port serial status reg (PSR) */ #define ETH_INTERFACE_PCM 0x00000001 #define ETH_LINK_IS_UP 0x00000002 #define ETH_PORT_AT_FULL_DUPLEX 0x00000004 #define ETH_RX_FLOW_CTRL_ENABLED 0x00000008 #define ETH_GMII_SPEED_1000 0x00000010 #define ETH_MII_SPEED_100 0x00000020 #define ETH_TX_IN_PROGRESS 0x00000080 #define ETH_BYPASS_ACTIVE 0x00000100 #define ETH_PORT_AT_PARTITION_STATE 0x00000200 #define ETH_PORT_TX_FIFO_EMPTY 0x00000400 /* SMI reg */ #define ETH_SMI_BUSY 0x10000000 /* 0 - Write, 1 - Read */ #define ETH_SMI_READ_VALID 0x08000000 /* 0 - Write, 1 - Read */ #define ETH_SMI_OPCODE_WRITE 0 /* Completion of Read */ #define ETH_SMI_OPCODE_READ 0x04000000 /* Operation is in progress */ /* Interrupt Cause Register Bit Definitions */ /* SDMA command status fields macros */ /* Tx & Rx descriptors status */ #define ETH_ERROR_SUMMARY 0x00000001 /* Tx & Rx descriptors command */ #define ETH_BUFFER_OWNED_BY_DMA 0x80000000 /* Tx descriptors status */ #define ETH_LC_ERROR 0 #define ETH_UR_ERROR 0x00000002 #define ETH_RL_ERROR 0x00000004 #define ETH_LLC_SNAP_FORMAT 0x00000200 /* Rx descriptors status */ #define ETH_OVERRUN_ERROR 0x00000002 #define ETH_MAX_FRAME_LENGTH_ERROR 0x00000004 #define ETH_RESOURCE_ERROR 0x00000006 #define ETH_VLAN_TAGGED 0x00080000 #define ETH_BPDU_FRAME 0x00100000 #define ETH_UDP_FRAME_OVER_IP_V_4 0x00200000 #define ETH_OTHER_FRAME_TYPE 0x00400000 #define ETH_LAYER_2_IS_ETH_V_2 0x00800000 #define ETH_FRAME_TYPE_IP_V_4 0x01000000 #define ETH_FRAME_HEADER_OK 0x02000000 #define ETH_RX_LAST_DESC 0x04000000 #define ETH_RX_FIRST_DESC 0x08000000 #define ETH_UNKNOWN_DESTINATION_ADDR 0x10000000 #define ETH_RX_ENABLE_INTERRUPT 0x20000000 #define ETH_LAYER_4_CHECKSUM_OK 0x40000000 /* Rx descriptors byte count */ #define ETH_FRAME_FRAGMENTED 0x00000004 /* Tx descriptors command */ #define ETH_LAYER_4_CHECKSUM_FIRST_DESC 0x00000400 #define ETH_FRAME_SET_TO_VLAN 0x00008000 #define ETH_UDP_FRAME 0x00010000 #define ETH_GEN_TCP_UDP_CHECKSUM 0x00020000 #define ETH_GEN_IP_V_4_CHECKSUM 0x00040000 #define ETH_ZERO_PADDING 0x00080000 #define ETH_TX_LAST_DESC 0x00100000 #define ETH_TX_FIRST_DESC 0x00200000 #define ETH_GEN_CRC 0x00400000 #define ETH_TX_ENABLE_INTERRUPT 0x00800000 #define ETH_AUTO_MODE 0x40000000 #define ETH_TX_IHL_SHIFT 11 /* typedefs */ typedef enum _eth_func_ret_status { ETH_OK, /* Returned as expected. */ ETH_ERROR, /* Fundamental error. */ ETH_RETRY, /* Could not process request. Try later.*/ ETH_END_OF_JOB, /* Ring has nothing to process. */ ETH_QUEUE_FULL, /* Ring resource error. */ ETH_QUEUE_LAST_RESOURCE /* Ring resources about to exhaust. */ } ETH_FUNC_RET_STATUS; /* These are for big-endian machines. Little endian needs different * definitions. */ #if defined(__BIG_ENDIAN) struct eth_rx_desc { u16 byte_cnt; /* Descriptor buffer byte count */ u16 buf_size; /* Buffer size */ u32 cmd_sts; /* Descriptor command status */ u32 next_desc_ptr; /* Next descriptor pointer */ u32 buf_ptr; /* Descriptor buffer pointer */ }; struct eth_tx_desc { u16 byte_cnt; /* buffer byte count */ u16 l4i_chk; /* CPU provided TCP checksum */ u32 cmd_sts; /* Command/status field */ u32 next_desc_ptr; /* Pointer to next descriptor */ u32 buf_ptr; /* pointer to buffer for this descriptor*/ }; #elif defined(__LITTLE_ENDIAN) struct eth_rx_desc { u32 cmd_sts; /* Descriptor command status */ u16 buf_size; /* Buffer size */ u16 byte_cnt; /* Descriptor buffer byte count */ u32 buf_ptr; /* Descriptor buffer pointer */ u32 next_desc_ptr; /* Next descriptor pointer */ }; struct eth_tx_desc { u32 cmd_sts; /* Command/status field */ u16 l4i_chk; /* CPU provided TCP checksum */ u16 byte_cnt; /* buffer byte count */ u32 buf_ptr; /* pointer to buffer for this descriptor*/ u32 next_desc_ptr; /* Pointer to next descriptor */ }; #else #error One of __BIG_ENDIAN or __LITTLE_ENDIAN must be defined #endif /* Unified struct for Rx and Tx operations. The user is not required to */ /* be familier with neither Tx nor Rx descriptors. */ struct pkt_info { unsigned short byte_cnt; /* Descriptor buffer byte count */ unsigned short l4i_chk; /* Tx CPU provided TCP Checksum */ unsigned int cmd_sts; /* Descriptor command status */ dma_addr_t buf_ptr; /* Descriptor buffer pointer */ struct sk_buff *return_info; /* User resource return information */ }; /* Ethernet port specific information */ struct mv643xx_mib_counters { u64 good_octets_received; u32 bad_octets_received; u32 internal_mac_transmit_err; u32 good_frames_received; u32 bad_frames_received; u32 broadcast_frames_received; u32 multicast_frames_received; u32 frames_64_octets; u32 frames_65_to_127_octets; u32 frames_128_to_255_octets; u32 frames_256_to_511_octets; u32 frames_512_to_1023_octets; u32 frames_1024_to_max_octets; u64 good_octets_sent; u32 good_frames_sent; u32 excessive_collision; u32 multicast_frames_sent; u32 broadcast_frames_sent; u32 unrec_mac_control_received; u32 fc_sent; u32 good_fc_received; u32 bad_fc_received; u32 undersize_received; u32 fragments_received; u32 oversize_received; u32 jabber_received; u32 mac_receive_error; u32 bad_crc_event; u32 collision; u32 late_collision; }; struct mv643xx_shared_private { void __iomem *eth_base; /* used to protect SMI_REG, which is shared across ports */ spinlock_t phy_lock; }; struct mv643xx_private { struct mv643xx_shared_private *shared; int port_num; /* User Ethernet port number */ u32 rx_sram_addr; /* Base address of rx sram area */ u32 rx_sram_size; /* Size of rx sram area */ u32 tx_sram_addr; /* Base address of tx sram area */ u32 tx_sram_size; /* Size of tx sram area */ int rx_resource_err; /* Rx ring resource error flag */ /* Tx/Rx rings managment indexes fields. For driver use */ /* Next available and first returning Rx resource */ int rx_curr_desc_q, rx_used_desc_q; /* Next available and first returning Tx resource */ int tx_curr_desc_q, tx_used_desc_q; #ifdef MV643XX_TX_FAST_REFILL u32 tx_clean_threshold; #endif struct eth_rx_desc *p_rx_desc_area; dma_addr_t rx_desc_dma; int rx_desc_area_size; struct sk_buff **rx_skb; struct eth_tx_desc *p_tx_desc_area; dma_addr_t tx_desc_dma; int tx_desc_area_size; struct sk_buff **tx_skb; struct work_struct tx_timeout_task; struct net_device *dev; struct napi_struct napi; struct net_device_stats stats; struct mv643xx_mib_counters mib_counters; spinlock_t lock; /* Size of Tx Ring per queue */ int tx_ring_size; /* Number of tx descriptors in use */ int tx_desc_count; /* Size of Rx Ring per queue */ int rx_ring_size; /* Number of rx descriptors in use */ int rx_desc_count; /* * Used in case RX Ring is empty, which can be caused when * system does not have resources (skb's) */ struct timer_list timeout; u32 rx_int_coal; u32 tx_int_coal; struct mii_if_info mii; }; /* Static function declarations */ static void eth_port_init(struct mv643xx_private *mp); static void eth_port_reset(struct mv643xx_private *mp); static void eth_port_start(struct net_device *dev); static void ethernet_phy_reset(struct mv643xx_private *mp); static void eth_port_write_smi_reg(struct mv643xx_private *mp, unsigned int phy_reg, unsigned int value); static void eth_port_read_smi_reg(struct mv643xx_private *mp, unsigned int phy_reg, unsigned int *value); static void eth_clear_mib_counters(struct mv643xx_private *mp); static ETH_FUNC_RET_STATUS eth_port_receive(struct mv643xx_private *mp, struct pkt_info *p_pkt_info); static ETH_FUNC_RET_STATUS eth_rx_return_buff(struct mv643xx_private *mp, struct pkt_info *p_pkt_info); static void eth_port_uc_addr_get(struct mv643xx_private *mp, unsigned char *p_addr); static void eth_port_uc_addr_set(struct mv643xx_private *mp, unsigned char *p_addr); static void eth_port_set_multicast_list(struct net_device *); static void mv643xx_eth_port_enable_tx(struct mv643xx_private *mp, unsigned int queues); static void mv643xx_eth_port_enable_rx(struct mv643xx_private *mp, unsigned int queues); static unsigned int mv643xx_eth_port_disable_tx(struct mv643xx_private *mp); static unsigned int mv643xx_eth_port_disable_rx(struct mv643xx_private *mp); static int mv643xx_eth_open(struct net_device *); static int mv643xx_eth_stop(struct net_device *); static void eth_port_init_mac_tables(struct mv643xx_private *mp); #ifdef MV643XX_NAPI static int mv643xx_poll(struct napi_struct *napi, int budget); #endif static int ethernet_phy_get(struct mv643xx_private *mp); static void ethernet_phy_set(struct mv643xx_private *mp, int phy_addr); static int ethernet_phy_detect(struct mv643xx_private *mp); 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 const struct ethtool_ops mv643xx_ethtool_ops; static char mv643xx_driver_name[] = "mv643xx_eth"; static char mv643xx_driver_version[] = "1.0"; static inline u32 rdl(struct mv643xx_private *mp, int offset) { return readl(mp->shared->eth_base + offset); } static inline void wrl(struct mv643xx_private *mp, int offset, u32 data) { writel(data, mp->shared->eth_base + offset); } /* * Changes MTU (maximum transfer unit) of the gigabit ethenret port * * Input : pointer to ethernet interface network device structure * new mtu size * Output : 0 upon success, -EINVAL upon failure */ static int mv643xx_eth_change_mtu(struct net_device *dev, int new_mtu) { if ((new_mtu > 9500) || (new_mtu < 64)) return -EINVAL; dev->mtu = new_mtu; if (!netif_running(dev)) return 0; /* * Stop and then re-open the interface. This will allocate RX * skbs of the new MTU. * There is a possible danger that the open will not succeed, * due to memory being full, which might fail the open function. */ mv643xx_eth_stop(dev); if (mv643xx_eth_open(dev)) { printk(KERN_ERR "%s: Fatal error on opening device\n", dev->name); } return 0; } /* * 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_refill_descs(struct net_device *dev) { struct mv643xx_private *mp = netdev_priv(dev); struct pkt_info pkt_info; struct sk_buff *skb; int unaligned; while (mp->rx_desc_count < mp->rx_ring_size) { skb = dev_alloc_skb(ETH_RX_SKB_SIZE + dma_get_cache_alignment()); if (!skb) break; mp->rx_desc_count++; unaligned = (u32)skb->data & (dma_get_cache_alignment() - 1); if (unaligned) skb_reserve(skb, dma_get_cache_alignment() - unaligned); pkt_info.cmd_sts = ETH_RX_ENABLE_INTERRUPT; 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, ETH_HW_IP_ALIGN); } /* * If RX ring is empty of SKB, set a timer to try allocating * again at a later time. */ if (mp->rx_desc_count == 0) { printk(KERN_INFO "%s: Rx ring is empty\n", dev->name); mp->timeout.expires = jiffies + (HZ / 10); /* 100 mSec */ add_timer(&mp->timeout); } } /* * 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 * failed (due to out of memory event). * * Input : pointer to ethernet interface network device structure * Output : N/A */ static inline void mv643xx_eth_rx_refill_descs_timer_wrapper(unsigned long data) { mv643xx_eth_rx_refill_descs((struct net_device *)data); } /* * mv643xx_eth_update_mac_address * * Update the MAC address of the port in the address table * * Input : pointer to ethernet interface network device structure * Output : N/A */ static void mv643xx_eth_update_mac_address(struct net_device *dev) { struct mv643xx_private *mp = netdev_priv(dev); eth_port_init_mac_tables(mp); eth_port_uc_addr_set(mp, dev->dev_addr); } /* * mv643xx_eth_set_rx_mode * * Change from promiscuos to regular rx mode * * Input : pointer to ethernet interface network device structure * Output : N/A */ static void mv643xx_eth_set_rx_mode(struct net_device *dev) { struct mv643xx_private *mp = netdev_priv(dev); u32 config_reg; config_reg = rdl(mp, PORT_CONFIG_REG(mp->port_num)); if (dev->flags & IFF_PROMISC) config_reg |= (u32) UNICAST_PROMISCUOUS_MODE; else config_reg &= ~(u32) UNICAST_PROMISCUOUS_MODE; wrl(mp, PORT_CONFIG_REG(mp->port_num), config_reg); eth_port_set_multicast_list(dev); } /* * mv643xx_eth_set_mac_address * * Change the interface's mac address. * No special hardware thing should be done because interface is always * put in promiscuous mode. * * Input : pointer to ethernet interface network device structure and * a pointer to the designated entry to be added to the cache. * Output : zero upon success, negative upon failure */ static int mv643xx_eth_set_mac_address(struct net_device *dev, void *addr) { int i; for (i = 0; i < 6; i++) /* +2 is for the offset of the HW addr type */ dev->dev_addr[i] = ((unsigned char *)addr)[i + 2]; mv643xx_eth_update_mac_address(dev); return 0; } /* * mv643xx_eth_tx_timeout * * Called upon a timeout on transmitting a packet * * Input : pointer to ethernet interface network device structure. * Output : N/A */ static void mv643xx_eth_tx_timeout(struct net_device *dev) { struct mv643xx_private *mp = netdev_priv(dev); printk(KERN_INFO "%s: TX timeout ", dev->name); /* Do the reset outside of interrupt context */ schedule_work(&mp->tx_timeout_task); } /* * mv643xx_eth_tx_timeout_task * * Actual routine to reset the adapter when a timeout on Tx has occurred */ static void mv643xx_eth_tx_timeout_task(struct work_struct *ugly) { struct mv643xx_private *mp = container_of(ugly, struct mv643xx_private, tx_timeout_task); struct net_device *dev = mp->dev; if (!netif_running(dev)) return; netif_stop_queue(dev); eth_port_reset(mp); eth_port_start(dev); if (mp->tx_ring_size - mp->tx_desc_count >= MAX_DESCS_PER_SKB) netif_wake_queue(dev); } /** * mv643xx_eth_free_tx_descs - Free the tx desc data for completed descriptors * * If force is non-zero, frees uncompleted descriptors as well */ static int mv643xx_eth_free_tx_descs(struct net_device *dev, int force) { struct mv643xx_private *mp = netdev_priv(dev); 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_desc_count might have changed before acquiring the lock */ if (mp->tx_desc_count <= 0) { spin_unlock_irqrestore(&mp->lock, flags); return released; } 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; } mp->tx_used_desc_q = (tx_index + 1) % mp->tx_ring_size; mp->tx_desc_count--; addr = desc->buf_ptr; count = desc->byte_cnt; skb = mp->tx_skb[tx_index]; if (skb) mp->tx_skb[tx_index] = NULL; if (cmd_sts & ETH_ERROR_SUMMARY) { printk("%s: Error in TX\n", dev->name); dev->stats.tx_errors++; } spin_unlock_irqrestore(&mp->lock, flags); if (cmd_sts & ETH_TX_FIRST_DESC) dma_unmap_single(NULL, addr, count, DMA_TO_DEVICE); else dma_unmap_page(NULL, addr, count, DMA_TO_DEVICE); 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 * * This function is forward packets that are received from the port's * queues toward kernel core or FastRoute them to another interface. * * Input : dev - a pointer to the required interface * max - maximum number to receive (0 means unlimted) * * Output : number of served packets */ static int mv643xx_eth_receive_queue(struct net_device *dev, int budget) { struct mv643xx_private *mp = netdev_priv(dev); struct net_device_stats *stats = &dev->stats; unsigned int received_packets = 0; struct sk_buff *skb; struct pkt_info pkt_info; while (budget-- > 0 && eth_port_receive(mp, &pkt_info) == ETH_OK) { dma_unmap_single(NULL, pkt_info.buf_ptr, ETH_RX_SKB_SIZE, DMA_FROM_DEVICE); mp->rx_desc_count--; received_packets++; /* * 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; /* * In case received a packet without first / last bits on OR * the error summary bit is on, the packets needs to be dropeed. */ if (((pkt_info.cmd_sts & (ETH_RX_FIRST_DESC | ETH_RX_LAST_DESC)) != (ETH_RX_FIRST_DESC | ETH_RX_LAST_DESC)) || (pkt_info.cmd_sts & ETH_ERROR_SUMMARY)) { stats->rx_dropped++; if ((pkt_info.cmd_sts & (ETH_RX_FIRST_DESC | ETH_RX_LAST_DESC)) != (ETH_RX_FIRST_DESC | ETH_RX_LAST_DESC)) { if (net_ratelimit()) printk(KERN_ERR "%s: Received packet spread " "on multiple descriptors\n", dev->name); } if (pkt_info.cmd_sts & ETH_ERROR_SUMMARY) stats->rx_errors++; dev_kfree_skb_irq(skb); } else { /* * The -4 is for the CRC in the trailer of the * received packet */ skb_put(skb, pkt_info.byte_cnt - 4); if (pkt_info.cmd_sts & ETH_LAYER_4_CHECKSUM_OK) { skb->ip_summed = CHECKSUM_UNNECESSARY; skb->csum = htons( (pkt_info.cmd_sts & 0x0007fff8) >> 3); } skb->protocol = eth_type_trans(skb, dev); #ifdef MV643XX_NAPI netif_receive_skb(skb); #else netif_rx(skb); #endif } 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 = rdl(mp, PORT_SERIAL_CONTROL_REG(port_num)); n_pscr = o_pscr; /* clear speed, duplex and rx buffer size fields */ n_pscr &= ~(SET_MII_SPEED_TO_100 | SET_GMII_SPEED_TO_1000 | SET_FULL_DUPLEX_MODE | MAX_RX_PACKET_MASK); if (ecmd->duplex == DUPLEX_FULL) n_pscr |= SET_FULL_DUPLEX_MODE; if (ecmd->speed == SPEED_1000) n_pscr |= SET_GMII_SPEED_TO_1000 | MAX_RX_PACKET_9700BYTE; else { if (ecmd->speed == SPEED_100) n_pscr |= SET_MII_SPEED_TO_100; n_pscr |= MAX_RX_PACKET_1522BYTE; } if (n_pscr != o_pscr) { if ((o_pscr & SERIAL_PORT_ENABLE) == 0) wrl(mp, PORT_SERIAL_CONTROL_REG(port_num), n_pscr); else { queues = mv643xx_eth_port_disable_tx(mp); o_pscr &= ~SERIAL_PORT_ENABLE; wrl(mp, PORT_SERIAL_CONTROL_REG(port_num), o_pscr); wrl(mp, PORT_SERIAL_CONTROL_REG(port_num), n_pscr); wrl(mp, PORT_SERIAL_CONTROL_REG(port_num), n_pscr); if (queues) mv643xx_eth_port_enable_tx(mp, queues); } } } /* * mv643xx_eth_int_handler * * Main interrupt handler for the gigbit ethernet ports * * Input : irq - irq number (not used) * dev_id - a pointer to the required interface's data structure * regs - not used * Output : N/A */ static irqreturn_t mv643xx_eth_int_handler(int irq, void *dev_id) { struct net_device *dev = (struct net_device *)dev_id; struct mv643xx_private *mp = netdev_priv(dev); u32 eth_int_cause, eth_int_cause_ext = 0; unsigned int port_num = mp->port_num; /* Read interrupt cause registers */ eth_int_cause = rdl(mp, INTERRUPT_CAUSE_REG(port_num)) & ETH_INT_UNMASK_ALL; if (eth_int_cause & ETH_INT_CAUSE_EXT) { eth_int_cause_ext = rdl(mp, INTERRUPT_CAUSE_EXTEND_REG(port_num)) & ETH_INT_UNMASK_ALL_EXT; wrl(mp, INTERRUPT_CAUSE_EXTEND_REG(port_num), ~eth_int_cause_ext); } /* PHY status changed */ if (eth_int_cause_ext & (ETH_INT_CAUSE_PHY | ETH_INT_CAUSE_STATE)) { 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(mp, 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); } } #ifdef MV643XX_NAPI if (eth_int_cause & ETH_INT_CAUSE_RX) { /* schedule the NAPI poll routine to maintain port */ wrl(mp, INTERRUPT_MASK_REG(port_num), ETH_INT_MASK_ALL); /* wait for previous write to complete */ rdl(mp, INTERRUPT_MASK_REG(port_num)); netif_rx_schedule(dev, &mp->napi); } #else if (eth_int_cause & ETH_INT_CAUSE_RX) mv643xx_eth_receive_queue(dev, INT_MAX); #endif if (eth_int_cause_ext & ETH_INT_CAUSE_TX) mv643xx_eth_free_completed_tx_descs(dev); /* * If no real interrupt occured, exit. * This can happen when using gigE interrupt coalescing mechanism. */ if ((eth_int_cause == 0x0) && (eth_int_cause_ext == 0x0)) return IRQ_NONE; return IRQ_HANDLED; } #ifdef MV643XX_COAL /* * eth_port_set_rx_coal - Sets coalescing interrupt mechanism on RX path * * DESCRIPTION: * This routine sets the RX coalescing interrupt mechanism parameter. * This parameter is a timeout counter, that counts in 64 t_clk * chunks ; that when timeout event occurs a maskable interrupt * occurs. * The parameter is calculated using the tClk of the MV-643xx chip * , and the required delay of the interrupt in usec. * * INPUT: * struct mv643xx_private *mp Ethernet port * unsigned int t_clk t_clk of the MV-643xx chip in HZ units * unsigned int delay Delay in usec * * OUTPUT: * Interrupt coalescing mechanism value is set in MV-643xx chip. * * RETURN: * The interrupt coalescing value set in the gigE port. * */ static unsigned int eth_port_set_rx_coal(struct mv643xx_private *mp, unsigned int t_clk, unsigned int delay) { unsigned int port_num = mp->port_num; unsigned int coal = ((t_clk / 1000000) * delay) / 64; /* Set RX Coalescing mechanism */ wrl(mp, SDMA_CONFIG_REG(port_num), ((coal & 0x3fff) << 8) | (rdl(mp, SDMA_CONFIG_REG(port_num)) & 0xffc000ff)); return coal; } #endif /* * eth_port_set_tx_coal - Sets coalescing interrupt mechanism on TX path * * DESCRIPTION: * This routine sets the TX coalescing interrupt mechanism parameter. * This parameter is a timeout counter, that counts in 64 t_clk * chunks ; that when timeout event occurs a maskable interrupt * occurs. * The parameter is calculated using the t_cLK frequency of the * MV-643xx chip and the required delay in the interrupt in uSec * * INPUT: * struct mv643xx_private *mp Ethernet port * unsigned int t_clk t_clk of the MV-643xx chip in HZ units * unsigned int delay Delay in uSeconds * * OUTPUT: * Interrupt coalescing mechanism value is set in MV-643xx chip. * * RETURN: * The interrupt coalescing value set in the gigE port. * */ static unsigned int eth_port_set_tx_coal(struct mv643xx_private *mp, unsigned int t_clk, unsigned int delay) { unsigned int coal = ((t_clk / 1000000) * delay) / 64; /* Set TX Coalescing mechanism */ wrl(mp, TX_FIFO_URGENT_THRESHOLD_REG(mp->port_num), coal << 4); return coal; } /* * ether_init_rx_desc_ring - Curve a Rx chain desc list and buffer in memory. * * DESCRIPTION: * This function prepares a Rx chained list of descriptors and packet * buffers in a form of a ring. The routine must be called after port * initialization routine and before port start routine. * The Ethernet SDMA engine uses CPU bus addresses to access the various * devices in the system (i.e. DRAM). This function uses the ethernet * struct 'virtual to physical' routine (set by the user) to set the ring * with physical addresses. * * INPUT: * struct mv643xx_private *mp Ethernet Port Control srtuct. * * OUTPUT: * The routine updates the Ethernet port control struct with information * regarding the Rx descriptors and buffers. * * RETURN: * None. */ static void ether_init_rx_desc_ring(struct mv643xx_private *mp) { volatile struct eth_rx_desc *p_rx_desc; int rx_desc_num = mp->rx_ring_size; int i; /* initialize the next_desc_ptr links in the Rx descriptors ring */ p_rx_desc = (struct eth_rx_desc *)mp->p_rx_desc_area; for (i = 0; i < rx_desc_num; i++) { p_rx_desc[i].next_desc_ptr = mp->rx_desc_dma + ((i + 1) % rx_desc_num) * sizeof(struct eth_rx_desc); } /* Save Rx desc pointer to driver struct. */ mp->rx_curr_desc_q = 0; mp->rx_used_desc_q = 0; mp->rx_desc_area_size = rx_desc_num * sizeof(struct eth_rx_desc); } /* * ether_init_tx_desc_ring - Curve a Tx chain desc list and buffer in memory. * * DESCRIPTION: * This function prepares a Tx chained list of descriptors and packet * buffers in a form of a ring. The routine must be called after port * initialization routine and before port start routine. * The Ethernet SDMA engine uses CPU bus addresses to access the various * devices in the system (i.e. DRAM). This function uses the ethernet * struct 'virtual to physical' routine (set by the user) to set the ring * with physical addresses. * * INPUT: * struct mv643xx_private *mp Ethernet Port Control srtuct. * * OUTPUT: * The routine updates the Ethernet port control struct with information * regarding the Tx descriptors and buffers. * * RETURN: * None. */ static void ether_init_tx_desc_ring(struct mv643xx_private *mp) { int tx_desc_num = mp->tx_ring_size; struct eth_tx_desc *p_tx_desc; int i; /* Initialize the next_desc_ptr links in the Tx descriptors ring */ p_tx_desc = (struct eth_tx_desc *)mp->p_tx_desc_area; for (i = 0; i < tx_desc_num; i++) { p_tx_desc[i].next_desc_ptr = mp->tx_desc_dma + ((i + 1) % tx_desc_num) * sizeof(struct eth_tx_desc); } mp->tx_curr_desc_q = 0; mp->tx_used_desc_q = 0; mp->tx_desc_area_size = tx_desc_num * sizeof(struct eth_tx_desc); } 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; } /* * mv643xx_eth_open * * This function is called when openning the network device. The function * should initialize all the hardware, initialize cyclic Rx/Tx * descriptors chain and buffers and allocate an IRQ to the network * device. * * Input : a pointer to the network device structure * * Output : zero of success , nonzero if fails. */ static int mv643xx_eth_open(struct net_device *dev) { struct mv643xx_private *mp = netdev_priv(dev); unsigned int port_num = mp->port_num; unsigned int size; int err; /* Clear any pending ethernet port interrupts */ wrl(mp, INTERRUPT_CAUSE_REG(port_num), 0); wrl(mp, INTERRUPT_CAUSE_EXTEND_REG(port_num), 0); /* wait for previous write to complete */ rdl(mp, INTERRUPT_CAUSE_EXTEND_REG(port_num)); err = request_irq(dev->irq, mv643xx_eth_int_handler, IRQF_SHARED | IRQF_SAMPLE_RANDOM, dev->name, dev); if (err) { printk(KERN_ERR "%s: Can not assign IRQ\n", dev->name); return -EAGAIN; } eth_port_init(mp); memset(&mp->timeout, 0, sizeof(struct timer_list)); mp->timeout.function = mv643xx_eth_rx_refill_descs_timer_wrapper; mp->timeout.data = (unsigned long)dev; /* Allocate RX and TX skb rings */ mp->rx_skb = kmalloc(sizeof(*mp->rx_skb) * mp->rx_ring_size, GFP_KERNEL); if (!mp->rx_skb) { printk(KERN_ERR "%s: Cannot allocate Rx skb ring\n", dev->name); err = -ENOMEM; goto out_free_irq; } mp->tx_skb = kmalloc(sizeof(*mp->tx_skb) * mp->tx_ring_size, GFP_KERNEL); if (!mp->tx_skb) { printk(KERN_ERR "%s: Cannot allocate Tx skb ring\n", dev->name); err = -ENOMEM; goto out_free_rx_skb; } /* Allocate TX ring */ mp->tx_desc_count = 0; size = mp->tx_ring_size * sizeof(struct eth_tx_desc); mp->tx_desc_area_size = size; if (mp->tx_sram_size) { mp->p_tx_desc_area = ioremap(mp->tx_sram_addr, mp->tx_sram_size); mp->tx_desc_dma = mp->tx_sram_addr; } else mp->p_tx_desc_area = dma_alloc_coherent(NULL, size, &mp->tx_desc_dma, GFP_KERNEL); if (!mp->p_tx_desc_area) { printk(KERN_ERR "%s: Cannot allocate Tx Ring (size %d bytes)\n", dev->name, size); err = -ENOMEM; goto out_free_tx_skb; } BUG_ON((u32) mp->p_tx_desc_area & 0xf); /* check 16-byte alignment */ memset((void *)mp->p_tx_desc_area, 0, mp->tx_desc_area_size); ether_init_tx_desc_ring(mp); /* Allocate RX ring */ mp->rx_desc_count = 0; size = mp->rx_ring_size * sizeof(struct eth_rx_desc); mp->rx_desc_area_size = size; if (mp->rx_sram_size) { mp->p_rx_desc_area = ioremap(mp->rx_sram_addr, mp->rx_sram_size); mp->rx_desc_dma = mp->rx_sram_addr; } else mp->p_rx_desc_area = dma_alloc_coherent(NULL, size, &mp->rx_desc_dma, GFP_KERNEL); if (!mp->p_rx_desc_area) { printk(KERN_ERR "%s: Cannot allocate Rx ring (size %d bytes)\n", dev->name, size); printk(KERN_ERR "%s: Freeing previously allocated TX queues...", dev->name); if (mp->rx_sram_size) iounmap(mp->p_tx_desc_area); else dma_free_coherent(NULL, mp->tx_desc_area_size, mp->p_tx_desc_area, mp->tx_desc_dma); err = -ENOMEM; goto out_free_tx_skb; } memset((void *)mp->p_rx_desc_area, 0, size); ether_init_rx_desc_ring(mp); mv643xx_eth_rx_refill_descs(dev); /* Fill RX ring with skb's */ #ifdef MV643XX_NAPI napi_enable(&mp->napi); #endif eth_port_start(dev); /* Interrupt Coalescing */ #ifdef MV643XX_COAL mp->rx_int_coal = eth_port_set_rx_coal(mp, 133000000, MV643XX_RX_COAL); #endif mp->tx_int_coal = eth_port_set_tx_coal(mp, 133000000, MV643XX_TX_COAL); /* Unmask phy and link status changes interrupts */ wrl(mp, INTERRUPT_EXTEND_MASK_REG(port_num), ETH_INT_UNMASK_ALL_EXT); /* Unmask RX buffer and TX end interrupt */ wrl(mp, INTERRUPT_MASK_REG(port_num), ETH_INT_UNMASK_ALL); return 0; out_free_tx_skb: kfree(mp->tx_skb); out_free_rx_skb: kfree(mp->rx_skb); out_free_irq: free_irq(dev->irq, dev); return err; } static void mv643xx_eth_free_tx_rings(struct net_device *dev) { struct mv643xx_private *mp = netdev_priv(dev); /* Stop Tx Queues */ mv643xx_eth_port_disable_tx(mp); /* 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) iounmap(mp->p_tx_desc_area); else dma_free_coherent(NULL, mp->tx_desc_area_size, mp->p_tx_desc_area, mp->tx_desc_dma); } static void mv643xx_eth_free_rx_rings(struct net_device *dev) { struct mv643xx_private *mp = netdev_priv(dev); int curr; /* Stop RX Queues */ mv643xx_eth_port_disable_rx(mp); /* Free preallocated skb's on RX rings */ 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_desc_count--; } } 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_desc_count); /* Free RX ring */ if (mp->rx_sram_size) iounmap(mp->p_rx_desc_area); else dma_free_coherent(NULL, mp->rx_desc_area_size, mp->p_rx_desc_area, mp->rx_desc_dma); } /* * mv643xx_eth_stop * * This function is used when closing the network device. * It updates the hardware, * release all memory that holds buffers and descriptors and release the IRQ. * Input : a pointer to the device structure * Output : zero if success , nonzero if fails */ static int mv643xx_eth_stop(struct net_device *dev) { struct mv643xx_private *mp = netdev_priv(dev); unsigned int port_num = mp->port_num; /* Mask all interrupts on ethernet port */ wrl(mp, INTERRUPT_MASK_REG(port_num), ETH_INT_MASK_ALL); /* wait for previous write to complete */ rdl(mp, INTERRUPT_MASK_REG(port_num)); #ifdef MV643XX_NAPI napi_disable(&mp->napi); #endif netif_carrier_off(dev); netif_stop_queue(dev); eth_port_reset(mp); mv643xx_eth_free_tx_rings(dev); mv643xx_eth_free_rx_rings(dev); free_irq(dev->irq, dev); return 0; } #ifdef MV643XX_NAPI /* * mv643xx_poll * * This function is used in case of NAPI */ static int mv643xx_poll(struct napi_struct *napi, int budget) { struct mv643xx_private *mp = container_of(napi, struct mv643xx_private, napi); struct net_device *dev = mp->dev; unsigned int port_num = mp->port_num; int work_done; #ifdef MV643XX_TX_FAST_REFILL if (++mp->tx_clean_threshold > 5) { mv643xx_eth_free_completed_tx_descs(dev); mp->tx_clean_threshold = 0; } #endif work_done = 0; if ((rdl(mp, RX_CURRENT_QUEUE_DESC_PTR_0(port_num))) != (u32) mp->rx_used_desc_q) work_done = mv643xx_eth_receive_queue(dev, budget); if (work_done < budget) { netif_rx_complete(dev, napi); wrl(mp, INTERRUPT_CAUSE_REG(port_num), 0); wrl(mp, INTERRUPT_CAUSE_EXTEND_REG(port_num), 0); wrl(mp, INTERRUPT_MASK_REG(port_num), ETH_INT_UNMASK_ALL); } return work_done; } #endif /** * 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. */ static inline unsigned int has_tiny_unaligned_frags(struct sk_buff *skb) { unsigned int frag; skb_frag_t *fragp; for (frag = 0; frag < skb_shinfo(skb)->nr_frags; frag++) { fragp = &skb_shinfo(skb)->frags[frag]; if (fragp->size <= 8 && fragp->page_offset & 0x7) return 1; } 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; 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. * * Ensure the data for each fragment to be transmitted is mapped properly, * then fill in descriptors in the tx hw queue. */ static void eth_tx_fill_frag_descs(struct mv643xx_private *mp, struct sk_buff *skb) { int frag; int tx_index; struct eth_tx_desc *desc; 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] = NULL; 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); } } static inline __be16 sum16_as_be(__sum16 sum) { return (__force __be16)sum; } /** * 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; cmd_sts = ETH_TX_FIRST_DESC | ETH_GEN_CRC | ETH_BUFFER_OWNED_BY_DMA; tx_index = eth_alloc_tx_desc_index(mp); desc = &mp->p_tx_desc_area[tx_index]; if (nr_frags) { eth_tx_fill_frag_descs(mp, skb); length = skb_headlen(skb); mp->tx_skb[tx_index] = NULL; } else { cmd_sts |= ETH_ZERO_PADDING | ETH_TX_LAST_DESC | ETH_TX_ENABLE_INTERRUPT; length = skb->len; mp->tx_skb[tx_index] = skb; } desc->byte_cnt = length; desc->buf_ptr = dma_map_single(NULL, skb->data, length, DMA_TO_DEVICE); if (skb->ip_summed == CHECKSUM_PARTIAL) { BUG_ON(skb->protocol != htons(ETH_P_IP)); cmd_sts |= ETH_GEN_TCP_UDP_CHECKSUM | ETH_GEN_IP_V_4_CHECKSUM | ip_hdr(skb)->ihl << ETH_TX_IHL_SHIFT; switch (ip_hdr(skb)->protocol) { case IPPROTO_UDP: cmd_sts |= ETH_UDP_FRAME; desc->l4i_chk = ntohs(sum16_as_be(udp_hdr(skb)->check)); break; case IPPROTO_TCP: desc->l4i_chk = ntohs(sum16_as_be(tcp_hdr(skb)->check)); break; default: BUG(); } } 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; /* ensure all descriptors are written before poking hardware */ wmb(); mv643xx_eth_port_enable_tx(mp, ETH_TX_QUEUES_ENABLED); mp->tx_desc_count += nr_frags + 1; } /** * 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 = &dev->stats; unsigned long flags; BUG_ON(netif_queue_stopped(dev)); if (has_tiny_unaligned_frags(skb) && __skb_linearize(skb)) { stats->tx_dropped++; printk(KERN_DEBUG "%s: failed to linearize tiny " "unaligned fragment\n", dev->name); return NETDEV_TX_BUSY; } spin_lock_irqsave(&mp->lock, flags); if (mp->tx_ring_size - mp->tx_desc_count < MAX_DESCS_PER_SKB) { printk(KERN_ERR "%s: transmit with queue full\n", dev->name); netif_stop_queue(dev); spin_unlock_irqrestore(&mp->lock, flags); return NETDEV_TX_BUSY; } 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 NETDEV_TX_OK; } #ifdef CONFIG_NET_POLL_CONTROLLER static void mv643xx_netpoll(struct net_device *netdev) { struct mv643xx_private *mp = netdev_priv(netdev); int port_num = mp->port_num; wrl(mp, INTERRUPT_MASK_REG(port_num), ETH_INT_MASK_ALL); /* wait for previous write to complete */ rdl(mp, INTERRUPT_MASK_REG(port_num)); mv643xx_eth_int_handler(netdev->irq, netdev); wrl(mp, 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 * * First function called after registering the network device. * It's purpose is to initialize the device as an ethernet device, * fill the ethernet device structure with pointers * to functions, * and set the MAC address of the interface * * Input : struct device * * Output : -ENOMEM if failed , 0 if success */ static int mv643xx_eth_probe(struct platform_device *pdev) { struct mv643xx_eth_platform_data *pd; int port_num; struct mv643xx_private *mp; struct net_device *dev; u8 *p; struct resource *res; int err; struct ethtool_cmd cmd; int duplex = DUPLEX_HALF; int speed = 0; /* default to auto-negotiation */ DECLARE_MAC_BUF(mac); pd = pdev->dev.platform_data; if (pd == NULL) { printk(KERN_ERR "No mv643xx_eth_platform_data\n"); return -ENODEV; } if (pd->shared == NULL) { printk(KERN_ERR "No mv643xx_eth_platform_data->shared\n"); return -ENODEV; } dev = alloc_etherdev(sizeof(struct mv643xx_private)); if (!dev) return -ENOMEM; platform_set_drvdata(pdev, dev); mp = netdev_priv(dev); mp->dev = dev; #ifdef MV643XX_NAPI netif_napi_add(dev, &mp->napi, mv643xx_poll, 64); #endif res = platform_get_resource(pdev, IORESOURCE_IRQ, 0); BUG_ON(!res); dev->irq = res->start; dev->open = mv643xx_eth_open; dev->stop = mv643xx_eth_stop; dev->hard_start_xmit = mv643xx_eth_start_xmit; dev->set_mac_address = mv643xx_eth_set_mac_address; dev->set_multicast_list = mv643xx_eth_set_rx_mode; /* No need to Tx Timeout */ dev->tx_timeout = mv643xx_eth_tx_timeout; #ifdef CONFIG_NET_POLL_CONTROLLER dev->poll_controller = mv643xx_netpoll; #endif dev->watchdog_timeo = 2 * HZ; 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 #ifdef MAX_SKB_FRAGS /* * Zero copy can only work if we use Discovery II memory. Else, we will * have to map the buffers to ISA memory which is only 16 MB */ dev->features = NETIF_F_SG | NETIF_F_IP_CSUM; #endif #endif /* Configure the timeout task */ INIT_WORK(&mp->tx_timeout_task, mv643xx_eth_tx_timeout_task); spin_lock_init(&mp->lock); mp->shared = platform_get_drvdata(pd->shared); port_num = mp->port_num = pd->port_number; /* set default config values */ eth_port_uc_addr_get(mp, dev->dev_addr); mp->rx_ring_size = PORT_DEFAULT_RECEIVE_QUEUE_SIZE; mp->tx_ring_size = PORT_DEFAULT_TRANSMIT_QUEUE_SIZE; if (is_valid_ether_addr(pd->mac_addr)) memcpy(dev->dev_addr, pd->mac_addr, 6); if (pd->phy_addr || pd->force_phy_addr) ethernet_phy_set(mp, pd->phy_addr); if (pd->rx_queue_size) mp->rx_ring_size = pd->rx_queue_size; if (pd->tx_queue_size) mp->tx_ring_size = pd->tx_queue_size; if (pd->tx_sram_size) { mp->tx_sram_size = pd->tx_sram_size; mp->tx_sram_addr = pd->tx_sram_addr; } if (pd->rx_sram_size) { 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(mp); mp->mii.phy_id_mask = 0x3f; mp->mii.reg_num_mask = 0x1f; err = ethernet_phy_detect(mp); if (err) { pr_debug("%s: No PHY detected at addr %d\n", dev->name, ethernet_phy_get(mp)); goto out; } ethernet_phy_reset(mp); 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); SET_NETDEV_DEV(dev, &pdev->dev); err = register_netdev(dev); if (err) goto out; p = dev->dev_addr; printk(KERN_NOTICE "%s: port %d with MAC address %s\n", dev->name, port_num, print_mac(mac, p)); if (dev->features & NETIF_F_SG) printk(KERN_NOTICE "%s: Scatter Gather Enabled\n", dev->name); if (dev->features & NETIF_F_IP_CSUM) printk(KERN_NOTICE "%s: TX TCP/IP Checksumming Supported\n", dev->name); #ifdef MV643XX_CHECKSUM_OFFLOAD_TX printk(KERN_NOTICE "%s: RX TCP/UDP Checksum Offload ON \n", dev->name); #endif #ifdef MV643XX_COAL printk(KERN_NOTICE "%s: TX and RX Interrupt Coalescing ON \n", dev->name); #endif #ifdef MV643XX_NAPI printk(KERN_NOTICE "%s: RX NAPI Enabled \n", dev->name); #endif if (mp->tx_sram_size > 0) printk(KERN_NOTICE "%s: Using SRAM\n", dev->name); return 0; out: free_netdev(dev); return err; } static int mv643xx_eth_remove(struct platform_device *pdev) { struct net_device *dev = platform_get_drvdata(pdev); unregister_netdev(dev); flush_scheduled_work(); free_netdev(dev); platform_set_drvdata(pdev, NULL); return 0; } static int mv643xx_eth_shared_probe(struct platform_device *pdev) { static int mv643xx_version_printed = 0; struct mv643xx_shared_private *msp; struct resource *res; int ret; if (!mv643xx_version_printed++) printk(KERN_NOTICE "MV-643xx 10/100/1000 Ethernet Driver\n"); ret = -EINVAL; res = platform_get_resource(pdev, IORESOURCE_MEM, 0); if (res == NULL) goto out; ret = -ENOMEM; msp = kmalloc(sizeof(*msp), GFP_KERNEL); if (msp == NULL) goto out; memset(msp, 0, sizeof(*msp)); msp->eth_base = ioremap(res->start, res->end - res->start + 1); if (msp->eth_base == NULL) goto out_free; spin_lock_init(&msp->phy_lock); platform_set_drvdata(pdev, msp); return 0; out_free: kfree(msp); out: return ret; } static int mv643xx_eth_shared_remove(struct platform_device *pdev) { struct mv643xx_shared_private *msp = platform_get_drvdata(pdev); iounmap(msp->eth_base); kfree(msp); return 0; } static void mv643xx_eth_shutdown(struct platform_device *pdev) { struct net_device *dev = platform_get_drvdata(pdev); struct mv643xx_private *mp = netdev_priv(dev); unsigned int port_num = mp->port_num; /* Mask all interrupts on ethernet port */ wrl(mp, INTERRUPT_MASK_REG(port_num), 0); rdl(mp, INTERRUPT_MASK_REG(port_num)); eth_port_reset(mp); } static struct platform_driver mv643xx_eth_driver = { .probe = mv643xx_eth_probe, .remove = mv643xx_eth_remove, .shutdown = mv643xx_eth_shutdown, .driver = { .name = MV643XX_ETH_NAME, .owner = THIS_MODULE, }, }; static struct platform_driver mv643xx_eth_shared_driver = { .probe = mv643xx_eth_shared_probe, .remove = mv643xx_eth_shared_remove, .driver = { .name = MV643XX_ETH_SHARED_NAME, .owner = THIS_MODULE, }, }; /* * mv643xx_init_module * * Registers the network drivers into the Linux kernel * * Input : N/A * * Output : N/A */ static int __init mv643xx_init_module(void) { int rc; rc = platform_driver_register(&mv643xx_eth_shared_driver); if (!rc) { rc = platform_driver_register(&mv643xx_eth_driver); if (rc) platform_driver_unregister(&mv643xx_eth_shared_driver); } return rc; } /* * mv643xx_cleanup_module * * Registers the network drivers into the Linux kernel * * Input : N/A * * Output : N/A */ static void __exit mv643xx_cleanup_module(void) { platform_driver_unregister(&mv643xx_eth_driver); platform_driver_unregister(&mv643xx_eth_shared_driver); } module_init(mv643xx_init_module); module_exit(mv643xx_cleanup_module); MODULE_LICENSE("GPL"); MODULE_AUTHOR( "Rabeeh Khoury, Assaf Hoffman, Matthew Dharm, Manish Lachwani" " and Dale Farnsworth"); MODULE_DESCRIPTION("Ethernet driver for Marvell MV643XX"); MODULE_ALIAS("platform:" MV643XX_ETH_NAME); MODULE_ALIAS("platform:" MV643XX_ETH_SHARED_NAME); /* * The second part is the low level driver of the gigE ethernet ports. */ /* * Marvell's Gigabit Ethernet controller low level driver * * DESCRIPTION: * This file introduce low level API to Marvell's Gigabit Ethernet * controller. This Gigabit Ethernet Controller driver API controls * 1) Operations (i.e. port init, start, reset etc'). * 2) Data flow (i.e. port send, receive etc'). * Each Gigabit Ethernet port is controlled via * struct mv643xx_private. * This struct includes user configuration information as well as * driver internal data needed for its operations. * * Supported Features: * - This low level driver is OS independent. Allocating memory for * the descriptor rings and buffers are not within the scope of * this driver. * - The user is free from Rx/Tx queue managing. * - This low level driver introduce functionality API that enable * the to operate Marvell's Gigabit Ethernet Controller in a * convenient way. * - Simple Gigabit Ethernet port operation API. * - Simple Gigabit Ethernet port data flow API. * - Data flow and operation API support per queue functionality. * - Support cached descriptors for better performance. * - Enable access to all four DRAM banks and internal SRAM memory * spaces. * - PHY access and control API. * - Port control register configuration API. * - Full control over Unicast and Multicast MAC configurations. * * Operation flow: * * Initialization phase * This phase complete the initialization of the the * mv643xx_private struct. * User information regarding port configuration has to be set * prior to calling the port initialization routine. * * In this phase any port Tx/Rx activity is halted, MIB counters * are cleared, PHY address is set according to user parameter and * access to DRAM and internal SRAM memory spaces. * * Driver ring initialization * Allocating memory for the descriptor rings and buffers is not * within the scope of this driver. Thus, the user is required to * allocate memory for the descriptors ring and buffers. Those * memory parameters are used by the Rx and Tx ring initialization * routines in order to curve the descriptor linked list in a form * of a ring. * Note: Pay special attention to alignment issues when using * cached descriptors/buffers. In this phase the driver store * information in the mv643xx_private struct regarding each queue * ring. * * Driver start * This phase prepares the Ethernet port for Rx and Tx activity. * It uses the information stored in the mv643xx_private struct to * initialize the various port registers. * * Data flow: * All packet references to/from the driver are done using * struct pkt_info. * This struct is a unified struct used with Rx and Tx operations. * This way the user is not required to be familiar with neither * Tx nor Rx descriptors structures. * The driver's descriptors rings are management by indexes. * Those indexes controls the ring resources and used to indicate * a SW resource error: * 'current' * This index points to the current available resource for use. For * example in Rx process this index will point to the descriptor * that will be passed to the user upon calling the receive * routine. In Tx process, this index will point to the descriptor * that will be assigned with the user packet info and transmitted. * 'used' * This index points to the descriptor that need to restore its * resources. For example in Rx process, using the Rx buffer return * API will attach the buffer returned in packet info to the * descriptor pointed by 'used'. In Tx process, using the Tx * descriptor return will merely return the user packet info with * the command status of the transmitted buffer pointed by the * 'used' index. Nevertheless, it is essential to use this routine * to update the 'used' index. * 'first' * This index supports Tx Scatter-Gather. It points to the first * descriptor of a packet assembled of multiple buffers. For * example when in middle of Such packet we have a Tx resource * error the 'curr' index get the value of 'first' to indicate * that the ring returned to its state before trying to transmit * this packet. * * Receive operation: * The eth_port_receive API set the packet information struct, * passed by the caller, with received information from the * 'current' SDMA descriptor. * It is the user responsibility to return this resource back * to the Rx descriptor ring to enable the reuse of this source. * Return Rx resource is done using the eth_rx_return_buff API. * * 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_config User port configuration value. * port_config_extend User port config extend value. * port_sdma_config User port SDMA config value. * port_serial_control User port serial control value. * * This driver data flow is done using the struct pkt_info which * is a unified struct for Rx and Tx operations: * * byte_cnt Tx/Rx descriptor buffer byte count. * l4i_chk CPU provided TCP Checksum. For Tx operation * only. * cmd_sts Tx/Rx descriptor command status. * buf_ptr Tx/Rx descriptor buffer pointer. * return_info Tx/Rx user resource return information. */ /* Ethernet Port routines */ static void eth_port_set_filter_table_entry(struct mv643xx_private *mp, int table, unsigned char entry); /* * eth_port_init - Initialize the Ethernet port driver * * DESCRIPTION: * This function prepares the ethernet port to start its activity: * 1) Completes the ethernet port driver struct initialization toward port * start routine. * 2) Resets the device to a quiescent state in case of warm reboot. * 3) Enable SDMA access to all four DRAM banks as well as internal SRAM. * 4) Clean MAC tables. The reset status of those tables is unknown. * 5) Set PHY address. * Note: Call this routine prior to eth_port_start routine and after * setting user values in the user fields of Ethernet port control * struct. * * INPUT: * struct mv643xx_private *mp Ethernet port control struct * * OUTPUT: * See description. * * RETURN: * None. */ static void eth_port_init(struct mv643xx_private *mp) { mp->rx_resource_err = 0; eth_port_reset(mp); eth_port_init_mac_tables(mp); } /* * eth_port_start - Start the Ethernet port activity. * * DESCRIPTION: * This routine prepares the Ethernet port for Rx and Tx activity: * 1. Initialize Tx and Rx Current Descriptor Pointer for each queue that * has been initialized a descriptor's ring (using * ether_init_tx_desc_ring for Tx and ether_init_rx_desc_ring for Rx) * 2. Initialize and enable the Ethernet configuration port by writing to * the port's configuration and command registers. * 3. Initialize and enable the SDMA by writing to the SDMA's * configuration and command registers. After completing these steps, * the ethernet port SDMA can starts to perform Rx and Tx activities. * * Note: Each Rx and Tx queue descriptor's list must be initialized prior * to calling this function (use ether_init_tx_desc_ring for Tx queues * and ether_init_rx_desc_ring for Rx queues). * * INPUT: * dev - a pointer to the required interface * * OUTPUT: * Ethernet port is ready to receive and transmit. * * RETURN: * None. */ 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; wrl(mp, TX_CURRENT_QUEUE_DESC_PTR_0(port_num), (u32)((struct eth_tx_desc *)mp->tx_desc_dma + tx_curr_desc)); /* Assignment of Rx CRDP of given queue */ rx_curr_desc = mp->rx_curr_desc_q; wrl(mp, RX_CURRENT_QUEUE_DESC_PTR_0(port_num), (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(mp, dev->dev_addr); /* Assign port configuration and command. */ wrl(mp, PORT_CONFIG_REG(port_num), PORT_CONFIG_DEFAULT_VALUE); wrl(mp, PORT_CONFIG_EXTEND_REG(port_num), PORT_CONFIG_EXTEND_DEFAULT_VALUE); pscr = rdl(mp, PORT_SERIAL_CONTROL_REG(port_num)); pscr &= ~(SERIAL_PORT_ENABLE | FORCE_LINK_PASS); wrl(mp, PORT_SERIAL_CONTROL_REG(port_num), pscr); pscr |= DISABLE_AUTO_NEG_FOR_FLOW_CTRL | DISABLE_AUTO_NEG_SPEED_GMII | DISABLE_AUTO_NEG_FOR_DUPLX | DO_NOT_FORCE_LINK_FAIL | SERIAL_PORT_CONTROL_RESERVED; wrl(mp, PORT_SERIAL_CONTROL_REG(port_num), pscr); pscr |= SERIAL_PORT_ENABLE; wrl(mp, PORT_SERIAL_CONTROL_REG(port_num), pscr); /* Assign port SDMA configuration */ wrl(mp, SDMA_CONFIG_REG(port_num), PORT_SDMA_CONFIG_DEFAULT_VALUE); /* Enable port Rx. */ mv643xx_eth_port_enable_rx(mp, ETH_RX_QUEUES_ENABLED); /* Disable port bandwidth limits by clearing MTU register */ wrl(mp, MAXIMUM_TRANSMIT_UNIT(port_num), 0); /* save phy settings across reset */ mv643xx_get_settings(dev, ðtool_cmd); ethernet_phy_reset(mp); mv643xx_set_settings(dev, ðtool_cmd); } /* * eth_port_uc_addr_set - Write a MAC address into the port's hw registers */ static void eth_port_uc_addr_set(struct mv643xx_private *mp, unsigned char *p_addr) { unsigned int port_num = mp->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) | (p_addr[3] << 0); wrl(mp, MAC_ADDR_LOW(port_num), mac_l); wrl(mp, MAC_ADDR_HIGH(port_num), mac_h); /* Accept frames with this address */ table = DA_FILTER_UNICAST_TABLE_BASE(port_num); eth_port_set_filter_table_entry(mp, table, p_addr[5] & 0x0f); } /* * eth_port_uc_addr_get - Read the MAC address from the port's hw registers */ static void eth_port_uc_addr_get(struct mv643xx_private *mp, unsigned char *p_addr) { unsigned int port_num = mp->port_num; unsigned int mac_h; unsigned int mac_l; mac_h = rdl(mp, MAC_ADDR_HIGH(port_num)); mac_l = rdl(mp, MAC_ADDR_LOW(port_num)); p_addr[0] = (mac_h >> 24) & 0xff; p_addr[1] = (mac_h >> 16) & 0xff; p_addr[2] = (mac_h >> 8) & 0xff; p_addr[3] = mac_h & 0xff; p_addr[4] = (mac_l >> 8) & 0xff; p_addr[5] = mac_l & 0xff; } /* * 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 * of the table. The bits in each entry are defined as follows: * 0 Accept=1, Drop=0 * 3-1 Queue (ETH_Q0=0) * 7-4 Reserved = 0; */ static void eth_port_set_filter_table_entry(struct mv643xx_private *mp, int table, unsigned char entry) { unsigned int table_reg; unsigned int tbl_offset; unsigned int reg_offset; tbl_offset = (entry / 4) * 4; /* Register offset of DA table entry */ reg_offset = entry % 4; /* Entry offset within the register */ /* Set "accepts frame bit" at specified table entry */ table_reg = rdl(mp, table + tbl_offset); table_reg |= 0x01 << (8 * reg_offset); wrl(mp, table + tbl_offset, table_reg); } /* * eth_port_mc_addr - Multicast address settings. * * The MV device supports multicast using two tables: * 1) Special Multicast Table for MAC addresses of the form * 0x01-00-5E-00-00-XX (where XX is between 0x00 and 0x_FF). * The MAC DA[7:0] bits are used as a pointer to the Special Multicast * Table entries in the DA-Filter table. * 2) Other Multicast Table for multicast of another type. A CRC-8bit * is used as an index to the Other Multicast Table entries in the * DA-Filter table. This function calculates the CRC-8bit value. * In either case, eth_port_set_filter_table_entry() is then called * to set to set the actual table entry. */ static void eth_port_mc_addr(struct mv643xx_private *mp, unsigned char *p_addr) { unsigned int port_num = mp->port_num; unsigned int mac_h; unsigned int mac_l; unsigned char crc_result = 0; int table; int mac_array[48]; int crc[8]; int i; if ((p_addr[0] == 0x01) && (p_addr[1] == 0x00) && (p_addr[2] == 0x5E) && (p_addr[3] == 0x00) && (p_addr[4] == 0x00)) { table = DA_FILTER_SPECIAL_MULTICAST_TABLE_BASE(port_num); eth_port_set_filter_table_entry(mp, table, p_addr[5]); return; } /* Calculate CRC-8 out of the given address */ mac_h = (p_addr[0] << 8) | (p_addr[1]); mac_l = (p_addr[2] << 24) | (p_addr[3] << 16) | (p_addr[4] << 8) | (p_addr[5] << 0); for (i = 0; i < 32; i++) mac_array[i] = (mac_l >> i) & 0x1; for (i = 32; i < 48; i++) mac_array[i] = (mac_h >> (i - 32)) & 0x1; crc[0] = mac_array[45] ^ mac_array[43] ^ mac_array[40] ^ mac_array[39] ^ mac_array[35] ^ mac_array[34] ^ mac_array[31] ^ mac_array[30] ^ mac_array[28] ^ mac_array[23] ^ mac_array[21] ^ mac_array[19] ^ mac_array[18] ^ mac_array[16] ^ mac_array[14] ^ mac_array[12] ^ mac_array[8] ^ mac_array[7] ^ mac_array[6] ^ mac_array[0]; crc[1] = mac_array[46] ^ mac_array[45] ^ mac_array[44] ^ mac_array[43] ^ mac_array[41] ^ mac_array[39] ^ mac_array[36] ^ mac_array[34] ^ mac_array[32] ^ mac_array[30] ^ mac_array[29] ^ mac_array[28] ^ mac_array[24] ^ mac_array[23] ^ mac_array[22] ^ mac_array[21] ^ mac_array[20] ^ mac_array[18] ^ mac_array[17] ^ mac_array[16] ^ mac_array[15] ^ mac_array[14] ^ mac_array[13] ^ mac_array[12] ^ mac_array[9] ^ mac_array[6] ^ mac_array[1] ^ mac_array[0]; crc[2] = mac_array[47] ^ mac_array[46] ^ mac_array[44] ^ mac_array[43] ^ mac_array[42] ^ mac_array[39] ^ mac_array[37] ^ mac_array[34] ^ mac_array[33] ^ mac_array[29] ^ mac_array[28] ^ mac_array[25] ^ mac_array[24] ^ mac_array[22] ^ mac_array[17] ^ mac_array[15] ^ mac_array[13] ^ mac_array[12] ^ mac_array[10] ^ mac_array[8] ^ mac_array[6] ^ mac_array[2] ^ mac_array[1] ^ mac_array[0]; crc[3] = mac_array[47] ^ mac_array[45] ^ mac_array[44] ^ mac_array[43] ^ mac_array[40] ^ mac_array[38] ^ mac_array[35] ^ mac_array[34] ^ mac_array[30] ^ mac_array[29] ^ mac_array[26] ^ mac_array[25] ^ mac_array[23] ^ mac_array[18] ^ mac_array[16] ^ mac_array[14] ^ mac_array[13] ^ mac_array[11] ^ mac_array[9] ^ mac_array[7] ^ mac_array[3] ^ mac_array[2] ^ mac_array[1]; crc[4] = mac_array[46] ^ mac_array[45] ^ mac_array[44] ^ mac_array[41] ^ mac_array[39] ^ mac_array[36] ^ mac_array[35] ^ mac_array[31] ^ mac_array[30] ^ mac_array[27] ^ mac_array[26] ^ mac_array[24] ^ mac_array[19] ^ mac_array[17] ^ mac_array[15] ^ mac_array[14] ^ mac_array[12] ^ mac_array[10] ^ mac_array[8] ^ mac_array[4] ^ mac_array[3] ^ mac_array[2]; crc[5] = mac_array[47] ^ mac_array[46] ^ mac_array[45] ^ mac_array[42] ^ mac_array[40] ^ mac_array[37] ^ mac_array[36] ^ mac_array[32] ^ mac_array[31] ^ mac_array[28] ^ mac_array[27] ^ mac_array[25] ^ mac_array[20] ^ mac_array[18] ^ mac_array[16] ^ mac_array[15] ^ mac_array[13] ^ mac_array[11] ^ mac_array[9] ^ mac_array[5] ^ mac_array[4] ^ mac_array[3]; crc[6] = mac_array[47] ^ mac_array[46] ^ mac_array[43] ^ mac_array[41] ^ mac_array[38] ^ mac_array[37] ^ mac_array[33] ^ mac_array[32] ^ mac_array[29] ^ mac_array[28] ^ mac_array[26] ^ mac_array[21] ^ mac_array[19] ^ mac_array[17] ^ mac_array[16] ^ mac_array[14] ^ mac_array[12] ^ mac_array[10] ^ mac_array[6] ^ mac_array[5] ^ mac_array[4]; crc[7] = mac_array[47] ^ mac_array[44] ^ mac_array[42] ^ mac_array[39] ^ mac_array[38] ^ mac_array[34] ^ mac_array[33] ^ mac_array[30] ^ mac_array[29] ^ mac_array[27] ^ mac_array[22] ^ mac_array[20] ^ mac_array[18] ^ mac_array[17] ^ mac_array[15] ^ mac_array[13] ^ mac_array[11] ^ mac_array[7] ^ mac_array[6] ^ mac_array[5]; for (i = 0; i < 8; i++) crc_result = crc_result | (crc[i] << i); table = DA_FILTER_OTHER_MULTICAST_TABLE_BASE(port_num); eth_port_set_filter_table_entry(mp, table, crc_result); } /* * Set the entire multicast list based on dev->mc_list. */ static void eth_port_set_multicast_list(struct net_device *dev) { struct dev_mc_list *mc_list; int i; int table_index; struct mv643xx_private *mp = netdev_priv(dev); unsigned int eth_port_num = mp->port_num; /* If the device is in promiscuous mode or in all multicast mode, * we will fully populate both multicast tables with accept. * This is guaranteed to yield a match on all multicast addresses... */ if ((dev->flags & IFF_PROMISC) || (dev->flags & IFF_ALLMULTI)) { for (table_index = 0; table_index <= 0xFC; table_index += 4) { /* Set all entries in DA filter special multicast * table (Ex_dFSMT) * Set for ETH_Q0 for now * Bits * 0 Accept=1, Drop=0 * 3-1 Queue ETH_Q0=0 * 7-4 Reserved = 0; */ wrl(mp, DA_FILTER_SPECIAL_MULTICAST_TABLE_BASE(eth_port_num) + table_index, 0x01010101); /* Set all entries in DA filter other multicast * table (Ex_dFOMT) * Set for ETH_Q0 for now * Bits * 0 Accept=1, Drop=0 * 3-1 Queue ETH_Q0=0 * 7-4 Reserved = 0; */ wrl(mp, DA_FILTER_OTHER_MULTICAST_TABLE_BASE(eth_port_num) + table_index, 0x01010101); } return; } /* We will clear out multicast tables every time we get the list. * Then add the entire new list... */ for (table_index = 0; table_index <= 0xFC; table_index += 4) { /* Clear DA filter special multicast table (Ex_dFSMT) */ wrl(mp, DA_FILTER_SPECIAL_MULTICAST_TABLE_BASE (eth_port_num) + table_index, 0); /* Clear DA filter other multicast table (Ex_dFOMT) */ wrl(mp, DA_FILTER_OTHER_MULTICAST_TABLE_BASE (eth_port_num) + table_index, 0); } /* Get pointer to net_device multicast list and add each one... */ for (i = 0, mc_list = dev->mc_list; (i < 256) && (mc_list != NULL) && (i < dev->mc_count); i++, mc_list = mc_list->next) if (mc_list->dmi_addrlen == 6) eth_port_mc_addr(mp, mc_list->dmi_addr); } /* * eth_port_init_mac_tables - Clear all entrance in the UC, SMC and OMC tables * * DESCRIPTION: * Go through all the DA filter tables (Unicast, Special Multicast & * Other Multicast) and set each entry to 0. * * INPUT: * struct mv643xx_private *mp Ethernet Port. * * OUTPUT: * Multicast and Unicast packets are rejected. * * RETURN: * None. */ static void eth_port_init_mac_tables(struct mv643xx_private *mp) { unsigned int port_num = mp->port_num; int table_index; /* Clear DA filter unicast table (Ex_dFUT) */ for (table_index = 0; table_index <= 0xC; table_index += 4) wrl(mp, DA_FILTER_UNICAST_TABLE_BASE(port_num) + table_index, 0); for (table_index = 0; table_index <= 0xFC; table_index += 4) { /* Clear DA filter special multicast table (Ex_dFSMT) */ wrl(mp, DA_FILTER_SPECIAL_MULTICAST_TABLE_BASE(port_num) + table_index, 0); /* Clear DA filter other multicast table (Ex_dFOMT) */ wrl(mp, DA_FILTER_OTHER_MULTICAST_TABLE_BASE(port_num) + table_index, 0); } } /* * eth_clear_mib_counters - Clear all MIB counters * * DESCRIPTION: * This function clears all MIB counters of a specific ethernet port. * A read from the MIB counter will reset the counter. * * INPUT: * struct mv643xx_private *mp Ethernet Port. * * OUTPUT: * After reading all MIB counters, the counters resets. * * RETURN: * MIB counter value. * */ static void eth_clear_mib_counters(struct mv643xx_private *mp) { unsigned int port_num = mp->port_num; int i; /* Perform dummy reads from MIB counters */ for (i = ETH_MIB_GOOD_OCTETS_RECEIVED_LOW; i < ETH_MIB_LATE_COLLISION; i += 4) rdl(mp, MIB_COUNTERS_BASE(port_num) + i); } static inline u32 read_mib(struct mv643xx_private *mp, int offset) { return rdl(mp, MIB_COUNTERS_BASE(mp->port_num) + offset); } static void eth_update_mib_counters(struct mv643xx_private *mp) { struct mv643xx_mib_counters *p = &mp->mib_counters; int offset; p->good_octets_received += read_mib(mp, ETH_MIB_GOOD_OCTETS_RECEIVED_LOW); p->good_octets_received += (u64)read_mib(mp, ETH_MIB_GOOD_OCTETS_RECEIVED_HIGH) << 32; for (offset = ETH_MIB_BAD_OCTETS_RECEIVED; offset <= ETH_MIB_FRAMES_1024_TO_MAX_OCTETS; offset += 4) *(u32 *)((char *)p + offset) += read_mib(mp, offset); p->good_octets_sent += read_mib(mp, ETH_MIB_GOOD_OCTETS_SENT_LOW); p->good_octets_sent += (u64)read_mib(mp, ETH_MIB_GOOD_OCTETS_SENT_HIGH) << 32; for (offset = ETH_MIB_GOOD_FRAMES_SENT; offset <= ETH_MIB_LATE_COLLISION; offset += 4) *(u32 *)((char *)p + offset) += read_mib(mp, offset); } /* * ethernet_phy_detect - Detect whether a phy is present * * DESCRIPTION: * This function tests whether there is a PHY present on * the specified port. * * INPUT: * struct mv643xx_private *mp Ethernet Port. * * OUTPUT: * None * * RETURN: * 0 on success * -ENODEV on failure * */ static int ethernet_phy_detect(struct mv643xx_private *mp) { unsigned int phy_reg_data0; int auto_neg; eth_port_read_smi_reg(mp, 0, &phy_reg_data0); auto_neg = phy_reg_data0 & 0x1000; phy_reg_data0 ^= 0x1000; /* invert auto_neg */ eth_port_write_smi_reg(mp, 0, phy_reg_data0); eth_port_read_smi_reg(mp, 0, &phy_reg_data0); if ((phy_reg_data0 & 0x1000) == auto_neg) return -ENODEV; /* change didn't take */ phy_reg_data0 ^= 0x1000; eth_port_write_smi_reg(mp, 0, phy_reg_data0); return 0; } /* * ethernet_phy_get - Get the ethernet port PHY address. * * DESCRIPTION: * This routine returns the given ethernet port PHY address. * * INPUT: * struct mv643xx_private *mp Ethernet Port. * * OUTPUT: * None. * * RETURN: * PHY address. * */ static int ethernet_phy_get(struct mv643xx_private *mp) { unsigned int reg_data; reg_data = rdl(mp, PHY_ADDR_REG); return ((reg_data >> (5 * mp->port_num)) & 0x1f); } /* * ethernet_phy_set - Set the ethernet port PHY address. * * DESCRIPTION: * This routine sets the given ethernet port PHY address. * * INPUT: * struct mv643xx_private *mp Ethernet Port. * int phy_addr PHY address. * * OUTPUT: * None. * * RETURN: * None. * */ static void ethernet_phy_set(struct mv643xx_private *mp, int phy_addr) { u32 reg_data; int addr_shift = 5 * mp->port_num; reg_data = rdl(mp, PHY_ADDR_REG); reg_data &= ~(0x1f << addr_shift); reg_data |= (phy_addr & 0x1f) << addr_shift; wrl(mp, PHY_ADDR_REG, reg_data); } /* * ethernet_phy_reset - Reset Ethernet port PHY. * * DESCRIPTION: * This routine utilizes the SMI interface to reset the ethernet port PHY. * * INPUT: * struct mv643xx_private *mp Ethernet Port. * * OUTPUT: * The PHY is reset. * * RETURN: * None. * */ static void ethernet_phy_reset(struct mv643xx_private *mp) { unsigned int phy_reg_data; /* Reset the PHY */ eth_port_read_smi_reg(mp, 0, &phy_reg_data); phy_reg_data |= 0x8000; /* Set bit 15 to reset the PHY */ eth_port_write_smi_reg(mp, 0, phy_reg_data); /* wait for PHY to come out of reset */ do { udelay(1); eth_port_read_smi_reg(mp, 0, &phy_reg_data); } while (phy_reg_data & 0x8000); } static void mv643xx_eth_port_enable_tx(struct mv643xx_private *mp, unsigned int queues) { wrl(mp, TRANSMIT_QUEUE_COMMAND_REG(mp->port_num), queues); } static void mv643xx_eth_port_enable_rx(struct mv643xx_private *mp, unsigned int queues) { wrl(mp, RECEIVE_QUEUE_COMMAND_REG(mp->port_num), queues); } static unsigned int mv643xx_eth_port_disable_tx(struct mv643xx_private *mp) { unsigned int port_num = mp->port_num; u32 queues; /* Stop Tx port activity. Check port Tx activity. */ queues = rdl(mp, TRANSMIT_QUEUE_COMMAND_REG(port_num)) & 0xFF; if (queues) { /* Issue stop command for active queues only */ wrl(mp, 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 (rdl(mp, TRANSMIT_QUEUE_COMMAND_REG(port_num)) & 0xFF) udelay(PHY_WAIT_MICRO_SECONDS); /* Wait for Tx FIFO to empty */ while (rdl(mp, 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(struct mv643xx_private *mp) { unsigned int port_num = mp->port_num; u32 queues; /* Stop Rx port activity. Check port Rx activity. */ queues = rdl(mp, RECEIVE_QUEUE_COMMAND_REG(port_num)) & 0xFF; if (queues) { /* Issue stop command for active queues only */ wrl(mp, 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 (rdl(mp, RECEIVE_QUEUE_COMMAND_REG(port_num)) & 0xFF) udelay(PHY_WAIT_MICRO_SECONDS); } return queues; } /* * eth_port_reset - Reset Ethernet port * * DESCRIPTION: * This routine resets the chip by aborting any SDMA engine activity and * clearing the MIB counters. The Receiver and the Transmit unit are in * idle state after this command is performed and the port is disabled. * * INPUT: * struct mv643xx_private *mp Ethernet Port. * * OUTPUT: * Channel activity is halted. * * RETURN: * None. * */ static void eth_port_reset(struct mv643xx_private *mp) { unsigned int port_num = mp->port_num; unsigned int reg_data; mv643xx_eth_port_disable_tx(mp); mv643xx_eth_port_disable_rx(mp); /* Clear all MIB counters */ eth_clear_mib_counters(mp); /* Reset the Enable bit in the Configuration Register */ reg_data = rdl(mp, PORT_SERIAL_CONTROL_REG(port_num)); reg_data &= ~(SERIAL_PORT_ENABLE | DO_NOT_FORCE_LINK_FAIL | FORCE_LINK_PASS); wrl(mp, PORT_SERIAL_CONTROL_REG(port_num), reg_data); } /* * eth_port_read_smi_reg - Read PHY registers * * DESCRIPTION: * This routine utilize the SMI interface to interact with the PHY in * order to perform PHY register read. * * INPUT: * struct mv643xx_private *mp Ethernet Port. * unsigned int phy_reg PHY register address offset. * unsigned int *value Register value buffer. * * OUTPUT: * Write the value of a specified PHY register into given buffer. * * RETURN: * false if the PHY is busy or read data is not in valid state. * true otherwise. * */ static void eth_port_read_smi_reg(struct mv643xx_private *mp, unsigned int phy_reg, unsigned int *value) { int phy_addr = ethernet_phy_get(mp); unsigned long flags; int i; /* the SMI register is a shared resource */ spin_lock_irqsave(&mp->shared->phy_lock, flags); /* wait for the SMI register to become available */ for (i = 0; rdl(mp, SMI_REG) & ETH_SMI_BUSY; i++) { if (i == PHY_WAIT_ITERATIONS) { printk("%s: PHY busy timeout\n", mp->dev->name); goto out; } udelay(PHY_WAIT_MICRO_SECONDS); } wrl(mp, SMI_REG, (phy_addr << 16) | (phy_reg << 21) | ETH_SMI_OPCODE_READ); /* now wait for the data to be valid */ for (i = 0; !(rdl(mp, SMI_REG) & ETH_SMI_READ_VALID); i++) { if (i == PHY_WAIT_ITERATIONS) { printk("%s: PHY read timeout\n", mp->dev->name); goto out; } udelay(PHY_WAIT_MICRO_SECONDS); } *value = rdl(mp, SMI_REG) & 0xffff; out: spin_unlock_irqrestore(&mp->shared->phy_lock, flags); } /* * eth_port_write_smi_reg - Write to PHY registers * * DESCRIPTION: * This routine utilize the SMI interface to interact with the PHY in * order to perform writes to PHY registers. * * INPUT: * struct mv643xx_private *mp Ethernet Port. * unsigned int phy_reg PHY register address offset. * unsigned int value Register value. * * OUTPUT: * Write the given value to the specified PHY register. * * RETURN: * false if the PHY is busy. * true otherwise. * */ static void eth_port_write_smi_reg(struct mv643xx_private *mp, unsigned int phy_reg, unsigned int value) { int phy_addr; int i; unsigned long flags; phy_addr = ethernet_phy_get(mp); /* the SMI register is a shared resource */ spin_lock_irqsave(&mp->shared->phy_lock, flags); /* wait for the SMI register to become available */ for (i = 0; rdl(mp, SMI_REG) & ETH_SMI_BUSY; i++) { if (i == PHY_WAIT_ITERATIONS) { printk("%s: PHY busy timeout\n", mp->dev->name); goto out; } udelay(PHY_WAIT_MICRO_SECONDS); } wrl(mp, SMI_REG, (phy_addr << 16) | (phy_reg << 21) | ETH_SMI_OPCODE_WRITE | (value & 0xffff)); out: spin_unlock_irqrestore(&mp->shared->phy_lock, flags); } /* * Wrappers for MII support library. */ static int mv643xx_mdio_read(struct net_device *dev, int phy_id, int location) { struct mv643xx_private *mp = netdev_priv(dev); int val; eth_port_read_smi_reg(mp, location, &val); return val; } static void mv643xx_mdio_write(struct net_device *dev, int phy_id, int location, int val) { struct mv643xx_private *mp = netdev_priv(dev); eth_port_write_smi_reg(mp, location, val); } /* * eth_port_receive - Get received information from Rx ring. * * DESCRIPTION: * This routine returns the received data to the caller. There is no * data copying during routine operation. All information is returned * using pointer to packet information struct passed from the caller. * If the routine exhausts Rx ring resources then the resource error flag * is set. * * INPUT: * struct mv643xx_private *mp Ethernet Port Control srtuct. * struct pkt_info *p_pkt_info User packet buffer. * * OUTPUT: * Rx ring current and used indexes are updated. * * RETURN: * ETH_ERROR in case the routine can not access Rx desc ring. * ETH_QUEUE_FULL if Rx ring resources are exhausted. * ETH_END_OF_JOB if there is no received data. * ETH_OK otherwise. */ static ETH_FUNC_RET_STATUS eth_port_receive(struct mv643xx_private *mp, struct pkt_info *p_pkt_info) { int rx_next_curr_desc, rx_curr_desc, rx_used_desc; volatile struct eth_rx_desc *p_rx_desc; unsigned int command_status; unsigned long flags; /* Do not process Rx ring in case of Rx ring resource error */ if (mp->rx_resource_err) return ETH_QUEUE_FULL; spin_lock_irqsave(&mp->lock, flags); /* Get the Rx Desc ring 'curr and 'used' indexes */ rx_curr_desc = mp->rx_curr_desc_q; rx_used_desc = mp->rx_used_desc_q; p_rx_desc = &mp->p_rx_desc_area[rx_curr_desc]; /* The following parameters are used to save readings from memory */ command_status = p_rx_desc->cmd_sts; rmb(); /* Nothing to receive... */ if (command_status & (ETH_BUFFER_OWNED_BY_DMA)) { spin_unlock_irqrestore(&mp->lock, flags); return ETH_END_OF_JOB; } p_pkt_info->byte_cnt = (p_rx_desc->byte_cnt) - RX_BUF_OFFSET; p_pkt_info->cmd_sts = command_status; p_pkt_info->buf_ptr = (p_rx_desc->buf_ptr) + RX_BUF_OFFSET; p_pkt_info->return_info = mp->rx_skb[rx_curr_desc]; p_pkt_info->l4i_chk = p_rx_desc->buf_size; /* * Clean the return info field to indicate that the * packet has been moved to the upper layers */ mp->rx_skb[rx_curr_desc] = NULL; /* Update current index in data structure */ rx_next_curr_desc = (rx_curr_desc + 1) % mp->rx_ring_size; mp->rx_curr_desc_q = rx_next_curr_desc; /* Rx descriptors exhausted. Set the Rx ring resource error flag */ if (rx_next_curr_desc == rx_used_desc) mp->rx_resource_err = 1; spin_unlock_irqrestore(&mp->lock, flags); return ETH_OK; } /* * eth_rx_return_buff - Returns a Rx buffer back to the Rx ring. * * DESCRIPTION: * This routine returns a Rx buffer back to the Rx ring. It retrieves the * next 'used' descriptor and attached the returned buffer to it. * In case the Rx ring was in "resource error" condition, where there are * no available Rx resources, the function resets the resource error flag. * * INPUT: * struct mv643xx_private *mp Ethernet Port Control srtuct. * struct pkt_info *p_pkt_info Information on returned buffer. * * OUTPUT: * New available Rx resource in Rx descriptor ring. * * RETURN: * ETH_ERROR in case the routine can not access Rx desc ring. * ETH_OK otherwise. */ static ETH_FUNC_RET_STATUS eth_rx_return_buff(struct mv643xx_private *mp, struct pkt_info *p_pkt_info) { int used_rx_desc; /* Where to return Rx resource */ volatile struct eth_rx_desc *p_used_rx_desc; unsigned long flags; spin_lock_irqsave(&mp->lock, flags); /* Get 'used' Rx descriptor */ used_rx_desc = mp->rx_used_desc_q; p_used_rx_desc = &mp->p_rx_desc_area[used_rx_desc]; p_used_rx_desc->buf_ptr = p_pkt_info->buf_ptr; p_used_rx_desc->buf_size = p_pkt_info->byte_cnt; mp->rx_skb[used_rx_desc] = p_pkt_info->return_info; /* Flush the write pipe */ /* Return the descriptor to DMA ownership */ wmb(); p_used_rx_desc->cmd_sts = ETH_BUFFER_OWNED_BY_DMA | ETH_RX_ENABLE_INTERRUPT; wmb(); /* Move the used descriptor pointer to the next descriptor */ mp->rx_used_desc_q = (used_rx_desc + 1) % mp->rx_ring_size; /* Any Rx return cancels the Rx resource error status */ mp->rx_resource_err = 0; spin_unlock_irqrestore(&mp->lock, flags); return ETH_OK; } /************* Begin ethtool support *************************/ struct mv643xx_stats { char stat_string[ETH_GSTRING_LEN]; int sizeof_stat; int stat_offset; }; #define MV643XX_STAT(m) FIELD_SIZEOF(struct mv643xx_private, m), \ offsetof(struct mv643xx_private, m) static const struct mv643xx_stats mv643xx_gstrings_stats[] = { { "rx_packets", MV643XX_STAT(stats.rx_packets) }, { "tx_packets", MV643XX_STAT(stats.tx_packets) }, { "rx_bytes", MV643XX_STAT(stats.rx_bytes) }, { "tx_bytes", MV643XX_STAT(stats.tx_bytes) }, { "rx_errors", MV643XX_STAT(stats.rx_errors) }, { "tx_errors", MV643XX_STAT(stats.tx_errors) }, { "rx_dropped", MV643XX_STAT(stats.rx_dropped) }, { "tx_dropped", MV643XX_STAT(stats.tx_dropped) }, { "good_octets_received", MV643XX_STAT(mib_counters.good_octets_received) }, { "bad_octets_received", MV643XX_STAT(mib_counters.bad_octets_received) }, { "internal_mac_transmit_err", MV643XX_STAT(mib_counters.internal_mac_transmit_err) }, { "good_frames_received", MV643XX_STAT(mib_counters.good_frames_received) }, { "bad_frames_received", MV643XX_STAT(mib_counters.bad_frames_received) }, { "broadcast_frames_received", MV643XX_STAT(mib_counters.broadcast_frames_received) }, { "multicast_frames_received", MV643XX_STAT(mib_counters.multicast_frames_received) }, { "frames_64_octets", MV643XX_STAT(mib_counters.frames_64_octets) }, { "frames_65_to_127_octets", MV643XX_STAT(mib_counters.frames_65_to_127_octets) }, { "frames_128_to_255_octets", MV643XX_STAT(mib_counters.frames_128_to_255_octets) }, { "frames_256_to_511_octets", MV643XX_STAT(mib_counters.frames_256_to_511_octets) }, { "frames_512_to_1023_octets", MV643XX_STAT(mib_counters.frames_512_to_1023_octets) }, { "frames_1024_to_max_octets", MV643XX_STAT(mib_counters.frames_1024_to_max_octets) }, { "good_octets_sent", MV643XX_STAT(mib_counters.good_octets_sent) }, { "good_frames_sent", MV643XX_STAT(mib_counters.good_frames_sent) }, { "excessive_collision", MV643XX_STAT(mib_counters.excessive_collision) }, { "multicast_frames_sent", MV643XX_STAT(mib_counters.multicast_frames_sent) }, { "broadcast_frames_sent", MV643XX_STAT(mib_counters.broadcast_frames_sent) }, { "unrec_mac_control_received", MV643XX_STAT(mib_counters.unrec_mac_control_received) }, { "fc_sent", MV643XX_STAT(mib_counters.fc_sent) }, { "good_fc_received", MV643XX_STAT(mib_counters.good_fc_received) }, { "bad_fc_received", MV643XX_STAT(mib_counters.bad_fc_received) }, { "undersize_received", MV643XX_STAT(mib_counters.undersize_received) }, { "fragments_received", MV643XX_STAT(mib_counters.fragments_received) }, { "oversize_received", MV643XX_STAT(mib_counters.oversize_received) }, { "jabber_received", MV643XX_STAT(mib_counters.jabber_received) }, { "mac_receive_error", MV643XX_STAT(mib_counters.mac_receive_error) }, { "bad_crc_event", MV643XX_STAT(mib_counters.bad_crc_event) }, { "collision", MV643XX_STAT(mib_counters.collision) }, { "late_collision", MV643XX_STAT(mib_counters.late_collision) }, }; #define MV643XX_STATS_LEN ARRAY_SIZE(mv643xx_gstrings_stats) static void mv643xx_get_drvinfo(struct net_device *netdev, struct ethtool_drvinfo *drvinfo) { strncpy(drvinfo->driver, mv643xx_driver_name, 32); strncpy(drvinfo->version, mv643xx_driver_version, 32); strncpy(drvinfo->fw_version, "N/A", 32); strncpy(drvinfo->bus_info, "mv643xx", 32); drvinfo->n_stats = MV643XX_STATS_LEN; } static int mv643xx_get_sset_count(struct net_device *netdev, int sset) { switch (sset) { case ETH_SS_STATS: return MV643XX_STATS_LEN; default: return -EOPNOTSUPP; } } static void mv643xx_get_ethtool_stats(struct net_device *netdev, struct ethtool_stats *stats, uint64_t *data) { struct mv643xx_private *mp = netdev->priv; int i; eth_update_mib_counters(mp); for (i = 0; i < MV643XX_STATS_LEN; i++) { char *p = (char *)mp+mv643xx_gstrings_stats[i].stat_offset; data[i] = (mv643xx_gstrings_stats[i].sizeof_stat == sizeof(uint64_t)) ? *(uint64_t *)p : *(uint32_t *)p; } } static void mv643xx_get_strings(struct net_device *netdev, uint32_t stringset, uint8_t *data) { int i; switch(stringset) { case ETH_SS_STATS: for (i=0; i < MV643XX_STATS_LEN; i++) { memcpy(data + i * ETH_GSTRING_LEN, mv643xx_gstrings_stats[i].stat_string, ETH_GSTRING_LEN); } break; } } 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 const struct ethtool_ops mv643xx_ethtool_ops = { .get_settings = mv643xx_get_settings, .set_settings = mv643xx_set_settings, .get_drvinfo = mv643xx_get_drvinfo, .get_link = mv643xx_eth_get_link, .set_sg = ethtool_op_set_sg, .get_sset_count = mv643xx_get_sset_count, .get_ethtool_stats = mv643xx_get_ethtool_stats, .get_strings = mv643xx_get_strings, .nway_reset = mv643xx_eth_nway_restart, }; /************* End ethtool support *************************/