blob: 9b530d7509a42ab41372fcb8fb0d86e0cef832ce [file] [log] [blame]
// SPDX-License-Identifier: GPL-2.0-or-later
/*
* forcedeth: Ethernet driver for NVIDIA nForce media access controllers.
*
* Note: This driver is a cleanroom reimplementation based on reverse
* engineered documentation written by Carl-Daniel Hailfinger
* and Andrew de Quincey.
*
* NVIDIA, nForce and other NVIDIA marks are trademarks or registered
* trademarks of NVIDIA Corporation in the United States and other
* countries.
*
* Copyright (C) 2003,4,5 Manfred Spraul
* Copyright (C) 2004 Andrew de Quincey (wol support)
* Copyright (C) 2004 Carl-Daniel Hailfinger (invalid MAC handling, insane
* IRQ rate fixes, bigendian fixes, cleanups, verification)
* Copyright (c) 2004,2005,2006,2007,2008,2009 NVIDIA Corporation
*
* Known bugs:
* We suspect that on some hardware no TX done interrupts are generated.
* This means recovery from netif_stop_queue only happens if the hw timer
* interrupt fires (100 times/second, configurable with NVREG_POLL_DEFAULT)
* and the timer is active in the IRQMask, or if a rx packet arrives by chance.
* If your hardware reliably generates tx done interrupts, then you can remove
* DEV_NEED_TIMERIRQ from the driver_data flags.
* DEV_NEED_TIMERIRQ will not harm you on sane hardware, only generating a few
* superfluous timer interrupts from the nic.
*/
#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
#define FORCEDETH_VERSION "0.64"
#define DRV_NAME "forcedeth"
#include <linux/module.h>
#include <linux/types.h>
#include <linux/pci.h>
#include <linux/interrupt.h>
#include <linux/netdevice.h>
#include <linux/etherdevice.h>
#include <linux/delay.h>
#include <linux/sched.h>
#include <linux/spinlock.h>
#include <linux/ethtool.h>
#include <linux/timer.h>
#include <linux/skbuff.h>
#include <linux/mii.h>
#include <linux/random.h>
#include <linux/if_vlan.h>
#include <linux/dma-mapping.h>
#include <linux/slab.h>
#include <linux/uaccess.h>
#include <linux/prefetch.h>
#include <linux/u64_stats_sync.h>
#include <linux/io.h>
#include <asm/irq.h>
#define TX_WORK_PER_LOOP 64
#define RX_WORK_PER_LOOP 64
/*
* Hardware access:
*/
#define DEV_NEED_TIMERIRQ 0x0000001 /* set the timer irq flag in the irq mask */
#define DEV_NEED_LINKTIMER 0x0000002 /* poll link settings. Relies on the timer irq */
#define DEV_HAS_LARGEDESC 0x0000004 /* device supports jumbo frames and needs packet format 2 */
#define DEV_HAS_HIGH_DMA 0x0000008 /* device supports 64bit dma */
#define DEV_HAS_CHECKSUM 0x0000010 /* device supports tx and rx checksum offloads */
#define DEV_HAS_VLAN 0x0000020 /* device supports vlan tagging and striping */
#define DEV_HAS_MSI 0x0000040 /* device supports MSI */
#define DEV_HAS_MSI_X 0x0000080 /* device supports MSI-X */
#define DEV_HAS_POWER_CNTRL 0x0000100 /* device supports power savings */
#define DEV_HAS_STATISTICS_V1 0x0000200 /* device supports hw statistics version 1 */
#define DEV_HAS_STATISTICS_V2 0x0000400 /* device supports hw statistics version 2 */
#define DEV_HAS_STATISTICS_V3 0x0000800 /* device supports hw statistics version 3 */
#define DEV_HAS_STATISTICS_V12 0x0000600 /* device supports hw statistics version 1 and 2 */
#define DEV_HAS_STATISTICS_V123 0x0000e00 /* device supports hw statistics version 1, 2, and 3 */
#define DEV_HAS_TEST_EXTENDED 0x0001000 /* device supports extended diagnostic test */
#define DEV_HAS_MGMT_UNIT 0x0002000 /* device supports management unit */
#define DEV_HAS_CORRECT_MACADDR 0x0004000 /* device supports correct mac address order */
#define DEV_HAS_COLLISION_FIX 0x0008000 /* device supports tx collision fix */
#define DEV_HAS_PAUSEFRAME_TX_V1 0x0010000 /* device supports tx pause frames version 1 */
#define DEV_HAS_PAUSEFRAME_TX_V2 0x0020000 /* device supports tx pause frames version 2 */
#define DEV_HAS_PAUSEFRAME_TX_V3 0x0040000 /* device supports tx pause frames version 3 */
#define DEV_NEED_TX_LIMIT 0x0080000 /* device needs to limit tx */
#define DEV_NEED_TX_LIMIT2 0x0180000 /* device needs to limit tx, expect for some revs */
#define DEV_HAS_GEAR_MODE 0x0200000 /* device supports gear mode */
#define DEV_NEED_PHY_INIT_FIX 0x0400000 /* device needs specific phy workaround */
#define DEV_NEED_LOW_POWER_FIX 0x0800000 /* device needs special power up workaround */
#define DEV_NEED_MSI_FIX 0x1000000 /* device needs msi workaround */
enum {
NvRegIrqStatus = 0x000,
#define NVREG_IRQSTAT_MIIEVENT 0x040
#define NVREG_IRQSTAT_MASK 0x83ff
NvRegIrqMask = 0x004,
#define NVREG_IRQ_RX_ERROR 0x0001
#define NVREG_IRQ_RX 0x0002
#define NVREG_IRQ_RX_NOBUF 0x0004
#define NVREG_IRQ_TX_ERR 0x0008
#define NVREG_IRQ_TX_OK 0x0010
#define NVREG_IRQ_TIMER 0x0020
#define NVREG_IRQ_LINK 0x0040
#define NVREG_IRQ_RX_FORCED 0x0080
#define NVREG_IRQ_TX_FORCED 0x0100
#define NVREG_IRQ_RECOVER_ERROR 0x8200
#define NVREG_IRQMASK_THROUGHPUT 0x00df
#define NVREG_IRQMASK_CPU 0x0060
#define NVREG_IRQ_TX_ALL (NVREG_IRQ_TX_ERR|NVREG_IRQ_TX_OK|NVREG_IRQ_TX_FORCED)
#define NVREG_IRQ_RX_ALL (NVREG_IRQ_RX_ERROR|NVREG_IRQ_RX|NVREG_IRQ_RX_NOBUF|NVREG_IRQ_RX_FORCED)
#define NVREG_IRQ_OTHER (NVREG_IRQ_TIMER|NVREG_IRQ_LINK|NVREG_IRQ_RECOVER_ERROR)
NvRegUnknownSetupReg6 = 0x008,
#define NVREG_UNKSETUP6_VAL 3
/*
* NVREG_POLL_DEFAULT is the interval length of the timer source on the nic
* NVREG_POLL_DEFAULT=97 would result in an interval length of 1 ms
*/
NvRegPollingInterval = 0x00c,
#define NVREG_POLL_DEFAULT_THROUGHPUT 65535 /* backup tx cleanup if loop max reached */
#define NVREG_POLL_DEFAULT_CPU 13
NvRegMSIMap0 = 0x020,
NvRegMSIMap1 = 0x024,
NvRegMSIIrqMask = 0x030,
#define NVREG_MSI_VECTOR_0_ENABLED 0x01
NvRegMisc1 = 0x080,
#define NVREG_MISC1_PAUSE_TX 0x01
#define NVREG_MISC1_HD 0x02
#define NVREG_MISC1_FORCE 0x3b0f3c
NvRegMacReset = 0x34,
#define NVREG_MAC_RESET_ASSERT 0x0F3
NvRegTransmitterControl = 0x084,
#define NVREG_XMITCTL_START 0x01
#define NVREG_XMITCTL_MGMT_ST 0x40000000
#define NVREG_XMITCTL_SYNC_MASK 0x000f0000
#define NVREG_XMITCTL_SYNC_NOT_READY 0x0
#define NVREG_XMITCTL_SYNC_PHY_INIT 0x00040000
#define NVREG_XMITCTL_MGMT_SEMA_MASK 0x00000f00
#define NVREG_XMITCTL_MGMT_SEMA_FREE 0x0
#define NVREG_XMITCTL_HOST_SEMA_MASK 0x0000f000
#define NVREG_XMITCTL_HOST_SEMA_ACQ 0x0000f000
#define NVREG_XMITCTL_HOST_LOADED 0x00004000
#define NVREG_XMITCTL_TX_PATH_EN 0x01000000
#define NVREG_XMITCTL_DATA_START 0x00100000
#define NVREG_XMITCTL_DATA_READY 0x00010000
#define NVREG_XMITCTL_DATA_ERROR 0x00020000
NvRegTransmitterStatus = 0x088,
#define NVREG_XMITSTAT_BUSY 0x01
NvRegPacketFilterFlags = 0x8c,
#define NVREG_PFF_PAUSE_RX 0x08
#define NVREG_PFF_ALWAYS 0x7F0000
#define NVREG_PFF_PROMISC 0x80
#define NVREG_PFF_MYADDR 0x20
#define NVREG_PFF_LOOPBACK 0x10
NvRegOffloadConfig = 0x90,
#define NVREG_OFFLOAD_HOMEPHY 0x601
#define NVREG_OFFLOAD_NORMAL RX_NIC_BUFSIZE
NvRegReceiverControl = 0x094,
#define NVREG_RCVCTL_START 0x01
#define NVREG_RCVCTL_RX_PATH_EN 0x01000000
NvRegReceiverStatus = 0x98,
#define NVREG_RCVSTAT_BUSY 0x01
NvRegSlotTime = 0x9c,
#define NVREG_SLOTTIME_LEGBF_ENABLED 0x80000000
#define NVREG_SLOTTIME_10_100_FULL 0x00007f00
#define NVREG_SLOTTIME_1000_FULL 0x0003ff00
#define NVREG_SLOTTIME_HALF 0x0000ff00
#define NVREG_SLOTTIME_DEFAULT 0x00007f00
#define NVREG_SLOTTIME_MASK 0x000000ff
NvRegTxDeferral = 0xA0,
#define NVREG_TX_DEFERRAL_DEFAULT 0x15050f
#define NVREG_TX_DEFERRAL_RGMII_10_100 0x16070f
#define NVREG_TX_DEFERRAL_RGMII_1000 0x14050f
#define NVREG_TX_DEFERRAL_RGMII_STRETCH_10 0x16190f
#define NVREG_TX_DEFERRAL_RGMII_STRETCH_100 0x16300f
#define NVREG_TX_DEFERRAL_MII_STRETCH 0x152000
NvRegRxDeferral = 0xA4,
#define NVREG_RX_DEFERRAL_DEFAULT 0x16
NvRegMacAddrA = 0xA8,
NvRegMacAddrB = 0xAC,
NvRegMulticastAddrA = 0xB0,
#define NVREG_MCASTADDRA_FORCE 0x01
NvRegMulticastAddrB = 0xB4,
NvRegMulticastMaskA = 0xB8,
#define NVREG_MCASTMASKA_NONE 0xffffffff
NvRegMulticastMaskB = 0xBC,
#define NVREG_MCASTMASKB_NONE 0xffff
NvRegPhyInterface = 0xC0,
#define PHY_RGMII 0x10000000
NvRegBackOffControl = 0xC4,
#define NVREG_BKOFFCTRL_DEFAULT 0x70000000
#define NVREG_BKOFFCTRL_SEED_MASK 0x000003ff
#define NVREG_BKOFFCTRL_SELECT 24
#define NVREG_BKOFFCTRL_GEAR 12
NvRegTxRingPhysAddr = 0x100,
NvRegRxRingPhysAddr = 0x104,
NvRegRingSizes = 0x108,
#define NVREG_RINGSZ_TXSHIFT 0
#define NVREG_RINGSZ_RXSHIFT 16
NvRegTransmitPoll = 0x10c,
#define NVREG_TRANSMITPOLL_MAC_ADDR_REV 0x00008000
NvRegLinkSpeed = 0x110,
#define NVREG_LINKSPEED_FORCE 0x10000
#define NVREG_LINKSPEED_10 1000
#define NVREG_LINKSPEED_100 100
#define NVREG_LINKSPEED_1000 50
#define NVREG_LINKSPEED_MASK (0xFFF)
NvRegUnknownSetupReg5 = 0x130,
#define NVREG_UNKSETUP5_BIT31 (1<<31)
NvRegTxWatermark = 0x13c,
#define NVREG_TX_WM_DESC1_DEFAULT 0x0200010
#define NVREG_TX_WM_DESC2_3_DEFAULT 0x1e08000
#define NVREG_TX_WM_DESC2_3_1000 0xfe08000
NvRegTxRxControl = 0x144,
#define NVREG_TXRXCTL_KICK 0x0001
#define NVREG_TXRXCTL_BIT1 0x0002
#define NVREG_TXRXCTL_BIT2 0x0004
#define NVREG_TXRXCTL_IDLE 0x0008
#define NVREG_TXRXCTL_RESET 0x0010
#define NVREG_TXRXCTL_RXCHECK 0x0400
#define NVREG_TXRXCTL_DESC_1 0
#define NVREG_TXRXCTL_DESC_2 0x002100
#define NVREG_TXRXCTL_DESC_3 0xc02200
#define NVREG_TXRXCTL_VLANSTRIP 0x00040
#define NVREG_TXRXCTL_VLANINS 0x00080
NvRegTxRingPhysAddrHigh = 0x148,
NvRegRxRingPhysAddrHigh = 0x14C,
NvRegTxPauseFrame = 0x170,
#define NVREG_TX_PAUSEFRAME_DISABLE 0x0fff0080
#define NVREG_TX_PAUSEFRAME_ENABLE_V1 0x01800010
#define NVREG_TX_PAUSEFRAME_ENABLE_V2 0x056003f0
#define NVREG_TX_PAUSEFRAME_ENABLE_V3 0x09f00880
NvRegTxPauseFrameLimit = 0x174,
#define NVREG_TX_PAUSEFRAMELIMIT_ENABLE 0x00010000
NvRegMIIStatus = 0x180,
#define NVREG_MIISTAT_ERROR 0x0001
#define NVREG_MIISTAT_LINKCHANGE 0x0008
#define NVREG_MIISTAT_MASK_RW 0x0007
#define NVREG_MIISTAT_MASK_ALL 0x000f
NvRegMIIMask = 0x184,
#define NVREG_MII_LINKCHANGE 0x0008
NvRegAdapterControl = 0x188,
#define NVREG_ADAPTCTL_START 0x02
#define NVREG_ADAPTCTL_LINKUP 0x04
#define NVREG_ADAPTCTL_PHYVALID 0x40000
#define NVREG_ADAPTCTL_RUNNING 0x100000
#define NVREG_ADAPTCTL_PHYSHIFT 24
NvRegMIISpeed = 0x18c,
#define NVREG_MIISPEED_BIT8 (1<<8)
#define NVREG_MIIDELAY 5
NvRegMIIControl = 0x190,
#define NVREG_MIICTL_INUSE 0x08000
#define NVREG_MIICTL_WRITE 0x00400
#define NVREG_MIICTL_ADDRSHIFT 5
NvRegMIIData = 0x194,
NvRegTxUnicast = 0x1a0,
NvRegTxMulticast = 0x1a4,
NvRegTxBroadcast = 0x1a8,
NvRegWakeUpFlags = 0x200,
#define NVREG_WAKEUPFLAGS_VAL 0x7770
#define NVREG_WAKEUPFLAGS_BUSYSHIFT 24
#define NVREG_WAKEUPFLAGS_ENABLESHIFT 16
#define NVREG_WAKEUPFLAGS_D3SHIFT 12
#define NVREG_WAKEUPFLAGS_D2SHIFT 8
#define NVREG_WAKEUPFLAGS_D1SHIFT 4
#define NVREG_WAKEUPFLAGS_D0SHIFT 0
#define NVREG_WAKEUPFLAGS_ACCEPT_MAGPAT 0x01
#define NVREG_WAKEUPFLAGS_ACCEPT_WAKEUPPAT 0x02
#define NVREG_WAKEUPFLAGS_ACCEPT_LINKCHANGE 0x04
#define NVREG_WAKEUPFLAGS_ENABLE 0x1111
NvRegMgmtUnitGetVersion = 0x204,
#define NVREG_MGMTUNITGETVERSION 0x01
NvRegMgmtUnitVersion = 0x208,
#define NVREG_MGMTUNITVERSION 0x08
NvRegPowerCap = 0x268,
#define NVREG_POWERCAP_D3SUPP (1<<30)
#define NVREG_POWERCAP_D2SUPP (1<<26)
#define NVREG_POWERCAP_D1SUPP (1<<25)
NvRegPowerState = 0x26c,
#define NVREG_POWERSTATE_POWEREDUP 0x8000
#define NVREG_POWERSTATE_VALID 0x0100
#define NVREG_POWERSTATE_MASK 0x0003
#define NVREG_POWERSTATE_D0 0x0000
#define NVREG_POWERSTATE_D1 0x0001
#define NVREG_POWERSTATE_D2 0x0002
#define NVREG_POWERSTATE_D3 0x0003
NvRegMgmtUnitControl = 0x278,
#define NVREG_MGMTUNITCONTROL_INUSE 0x20000
NvRegTxCnt = 0x280,
NvRegTxZeroReXmt = 0x284,
NvRegTxOneReXmt = 0x288,
NvRegTxManyReXmt = 0x28c,
NvRegTxLateCol = 0x290,
NvRegTxUnderflow = 0x294,
NvRegTxLossCarrier = 0x298,
NvRegTxExcessDef = 0x29c,
NvRegTxRetryErr = 0x2a0,
NvRegRxFrameErr = 0x2a4,
NvRegRxExtraByte = 0x2a8,
NvRegRxLateCol = 0x2ac,
NvRegRxRunt = 0x2b0,
NvRegRxFrameTooLong = 0x2b4,
NvRegRxOverflow = 0x2b8,
NvRegRxFCSErr = 0x2bc,
NvRegRxFrameAlignErr = 0x2c0,
NvRegRxLenErr = 0x2c4,
NvRegRxUnicast = 0x2c8,
NvRegRxMulticast = 0x2cc,
NvRegRxBroadcast = 0x2d0,
NvRegTxDef = 0x2d4,
NvRegTxFrame = 0x2d8,
NvRegRxCnt = 0x2dc,
NvRegTxPause = 0x2e0,
NvRegRxPause = 0x2e4,
NvRegRxDropFrame = 0x2e8,
NvRegVlanControl = 0x300,
#define NVREG_VLANCONTROL_ENABLE 0x2000
NvRegMSIXMap0 = 0x3e0,
NvRegMSIXMap1 = 0x3e4,
NvRegMSIXIrqStatus = 0x3f0,
NvRegPowerState2 = 0x600,
#define NVREG_POWERSTATE2_POWERUP_MASK 0x0F15
#define NVREG_POWERSTATE2_POWERUP_REV_A3 0x0001
#define NVREG_POWERSTATE2_PHY_RESET 0x0004
#define NVREG_POWERSTATE2_GATE_CLOCKS 0x0F00
};
/* Big endian: should work, but is untested */
struct ring_desc {
__le32 buf;
__le32 flaglen;
};
struct ring_desc_ex {
__le32 bufhigh;
__le32 buflow;
__le32 txvlan;
__le32 flaglen;
};
union ring_type {
struct ring_desc *orig;
struct ring_desc_ex *ex;
};
#define FLAG_MASK_V1 0xffff0000
#define FLAG_MASK_V2 0xffffc000
#define LEN_MASK_V1 (0xffffffff ^ FLAG_MASK_V1)
#define LEN_MASK_V2 (0xffffffff ^ FLAG_MASK_V2)
#define NV_TX_LASTPACKET (1<<16)
#define NV_TX_RETRYERROR (1<<19)
#define NV_TX_RETRYCOUNT_MASK (0xF<<20)
#define NV_TX_FORCED_INTERRUPT (1<<24)
#define NV_TX_DEFERRED (1<<26)
#define NV_TX_CARRIERLOST (1<<27)
#define NV_TX_LATECOLLISION (1<<28)
#define NV_TX_UNDERFLOW (1<<29)
#define NV_TX_ERROR (1<<30)
#define NV_TX_VALID (1<<31)
#define NV_TX2_LASTPACKET (1<<29)
#define NV_TX2_RETRYERROR (1<<18)
#define NV_TX2_RETRYCOUNT_MASK (0xF<<19)
#define NV_TX2_FORCED_INTERRUPT (1<<30)
#define NV_TX2_DEFERRED (1<<25)
#define NV_TX2_CARRIERLOST (1<<26)
#define NV_TX2_LATECOLLISION (1<<27)
#define NV_TX2_UNDERFLOW (1<<28)
/* error and valid are the same for both */
#define NV_TX2_ERROR (1<<30)
#define NV_TX2_VALID (1<<31)
#define NV_TX2_TSO (1<<28)
#define NV_TX2_TSO_SHIFT 14
#define NV_TX2_TSO_MAX_SHIFT 14
#define NV_TX2_TSO_MAX_SIZE (1<<NV_TX2_TSO_MAX_SHIFT)
#define NV_TX2_CHECKSUM_L3 (1<<27)
#define NV_TX2_CHECKSUM_L4 (1<<26)
#define NV_TX3_VLAN_TAG_PRESENT (1<<18)
#define NV_RX_DESCRIPTORVALID (1<<16)
#define NV_RX_MISSEDFRAME (1<<17)
#define NV_RX_SUBTRACT1 (1<<18)
#define NV_RX_ERROR1 (1<<23)
#define NV_RX_ERROR2 (1<<24)
#define NV_RX_ERROR3 (1<<25)
#define NV_RX_ERROR4 (1<<26)
#define NV_RX_CRCERR (1<<27)
#define NV_RX_OVERFLOW (1<<28)
#define NV_RX_FRAMINGERR (1<<29)
#define NV_RX_ERROR (1<<30)
#define NV_RX_AVAIL (1<<31)
#define NV_RX_ERROR_MASK (NV_RX_ERROR1|NV_RX_ERROR2|NV_RX_ERROR3|NV_RX_ERROR4|NV_RX_CRCERR|NV_RX_OVERFLOW|NV_RX_FRAMINGERR)
#define NV_RX2_CHECKSUMMASK (0x1C000000)
#define NV_RX2_CHECKSUM_IP (0x10000000)
#define NV_RX2_CHECKSUM_IP_TCP (0x14000000)
#define NV_RX2_CHECKSUM_IP_UDP (0x18000000)
#define NV_RX2_DESCRIPTORVALID (1<<29)
#define NV_RX2_SUBTRACT1 (1<<25)
#define NV_RX2_ERROR1 (1<<18)
#define NV_RX2_ERROR2 (1<<19)
#define NV_RX2_ERROR3 (1<<20)
#define NV_RX2_ERROR4 (1<<21)
#define NV_RX2_CRCERR (1<<22)
#define NV_RX2_OVERFLOW (1<<23)
#define NV_RX2_FRAMINGERR (1<<24)
/* error and avail are the same for both */
#define NV_RX2_ERROR (1<<30)
#define NV_RX2_AVAIL (1<<31)
#define NV_RX2_ERROR_MASK (NV_RX2_ERROR1|NV_RX2_ERROR2|NV_RX2_ERROR3|NV_RX2_ERROR4|NV_RX2_CRCERR|NV_RX2_OVERFLOW|NV_RX2_FRAMINGERR)
#define NV_RX3_VLAN_TAG_PRESENT (1<<16)
#define NV_RX3_VLAN_TAG_MASK (0x0000FFFF)
/* Miscellaneous hardware related defines: */
#define NV_PCI_REGSZ_VER1 0x270
#define NV_PCI_REGSZ_VER2 0x2d4
#define NV_PCI_REGSZ_VER3 0x604
#define NV_PCI_REGSZ_MAX 0x604
/* various timeout delays: all in usec */
#define NV_TXRX_RESET_DELAY 4
#define NV_TXSTOP_DELAY1 10
#define NV_TXSTOP_DELAY1MAX 500000
#define NV_TXSTOP_DELAY2 100
#define NV_RXSTOP_DELAY1 10
#define NV_RXSTOP_DELAY1MAX 500000
#define NV_RXSTOP_DELAY2 100
#define NV_SETUP5_DELAY 5
#define NV_SETUP5_DELAYMAX 50000
#define NV_POWERUP_DELAY 5
#define NV_POWERUP_DELAYMAX 5000
#define NV_MIIBUSY_DELAY 50
#define NV_MIIPHY_DELAY 10
#define NV_MIIPHY_DELAYMAX 10000
#define NV_MAC_RESET_DELAY 64
#define NV_WAKEUPPATTERNS 5
#define NV_WAKEUPMASKENTRIES 4
/* General driver defaults */
#define NV_WATCHDOG_TIMEO (5*HZ)
#define RX_RING_DEFAULT 512
#define TX_RING_DEFAULT 256
#define RX_RING_MIN 128
#define TX_RING_MIN 64
#define RING_MAX_DESC_VER_1 1024
#define RING_MAX_DESC_VER_2_3 16384
/* rx/tx mac addr + type + vlan + align + slack*/
#define NV_RX_HEADERS (64)
/* even more slack. */
#define NV_RX_ALLOC_PAD (64)
/* maximum mtu size */
#define NV_PKTLIMIT_1 ETH_DATA_LEN /* hard limit not known */
#define NV_PKTLIMIT_2 9100 /* Actual limit according to NVidia: 9202 */
#define OOM_REFILL (1+HZ/20)
#define POLL_WAIT (1+HZ/100)
#define LINK_TIMEOUT (3*HZ)
#define STATS_INTERVAL (10*HZ)
/*
* desc_ver values:
* The nic supports three different descriptor types:
* - DESC_VER_1: Original
* - DESC_VER_2: support for jumbo frames.
* - DESC_VER_3: 64-bit format.
*/
#define DESC_VER_1 1
#define DESC_VER_2 2
#define DESC_VER_3 3
/* PHY defines */
#define PHY_OUI_MARVELL 0x5043
#define PHY_OUI_CICADA 0x03f1
#define PHY_OUI_VITESSE 0x01c1
#define PHY_OUI_REALTEK 0x0732
#define PHY_OUI_REALTEK2 0x0020
#define PHYID1_OUI_MASK 0x03ff
#define PHYID1_OUI_SHFT 6
#define PHYID2_OUI_MASK 0xfc00
#define PHYID2_OUI_SHFT 10
#define PHYID2_MODEL_MASK 0x03f0
#define PHY_MODEL_REALTEK_8211 0x0110
#define PHY_REV_MASK 0x0001
#define PHY_REV_REALTEK_8211B 0x0000
#define PHY_REV_REALTEK_8211C 0x0001
#define PHY_MODEL_REALTEK_8201 0x0200
#define PHY_MODEL_MARVELL_E3016 0x0220
#define PHY_MARVELL_E3016_INITMASK 0x0300
#define PHY_CICADA_INIT1 0x0f000
#define PHY_CICADA_INIT2 0x0e00
#define PHY_CICADA_INIT3 0x01000
#define PHY_CICADA_INIT4 0x0200
#define PHY_CICADA_INIT5 0x0004
#define PHY_CICADA_INIT6 0x02000
#define PHY_VITESSE_INIT_REG1 0x1f
#define PHY_VITESSE_INIT_REG2 0x10
#define PHY_VITESSE_INIT_REG3 0x11
#define PHY_VITESSE_INIT_REG4 0x12
#define PHY_VITESSE_INIT_MSK1 0xc
#define PHY_VITESSE_INIT_MSK2 0x0180
#define PHY_VITESSE_INIT1 0x52b5
#define PHY_VITESSE_INIT2 0xaf8a
#define PHY_VITESSE_INIT3 0x8
#define PHY_VITESSE_INIT4 0x8f8a
#define PHY_VITESSE_INIT5 0xaf86
#define PHY_VITESSE_INIT6 0x8f86
#define PHY_VITESSE_INIT7 0xaf82
#define PHY_VITESSE_INIT8 0x0100
#define PHY_VITESSE_INIT9 0x8f82
#define PHY_VITESSE_INIT10 0x0
#define PHY_REALTEK_INIT_REG1 0x1f
#define PHY_REALTEK_INIT_REG2 0x19
#define PHY_REALTEK_INIT_REG3 0x13
#define PHY_REALTEK_INIT_REG4 0x14
#define PHY_REALTEK_INIT_REG5 0x18
#define PHY_REALTEK_INIT_REG6 0x11
#define PHY_REALTEK_INIT_REG7 0x01
#define PHY_REALTEK_INIT1 0x0000
#define PHY_REALTEK_INIT2 0x8e00
#define PHY_REALTEK_INIT3 0x0001
#define PHY_REALTEK_INIT4 0xad17
#define PHY_REALTEK_INIT5 0xfb54
#define PHY_REALTEK_INIT6 0xf5c7
#define PHY_REALTEK_INIT7 0x1000
#define PHY_REALTEK_INIT8 0x0003
#define PHY_REALTEK_INIT9 0x0008
#define PHY_REALTEK_INIT10 0x0005
#define PHY_REALTEK_INIT11 0x0200
#define PHY_REALTEK_INIT_MSK1 0x0003
#define PHY_GIGABIT 0x0100
#define PHY_TIMEOUT 0x1
#define PHY_ERROR 0x2
#define PHY_100 0x1
#define PHY_1000 0x2
#define PHY_HALF 0x100
#define NV_PAUSEFRAME_RX_CAPABLE 0x0001
#define NV_PAUSEFRAME_TX_CAPABLE 0x0002
#define NV_PAUSEFRAME_RX_ENABLE 0x0004
#define NV_PAUSEFRAME_TX_ENABLE 0x0008
#define NV_PAUSEFRAME_RX_REQ 0x0010
#define NV_PAUSEFRAME_TX_REQ 0x0020
#define NV_PAUSEFRAME_AUTONEG 0x0040
/* MSI/MSI-X defines */
#define NV_MSI_X_MAX_VECTORS 8
#define NV_MSI_X_VECTORS_MASK 0x000f
#define NV_MSI_CAPABLE 0x0010
#define NV_MSI_X_CAPABLE 0x0020
#define NV_MSI_ENABLED 0x0040
#define NV_MSI_X_ENABLED 0x0080
#define NV_MSI_X_VECTOR_ALL 0x0
#define NV_MSI_X_VECTOR_RX 0x0
#define NV_MSI_X_VECTOR_TX 0x1
#define NV_MSI_X_VECTOR_OTHER 0x2
#define NV_MSI_PRIV_OFFSET 0x68
#define NV_MSI_PRIV_VALUE 0xffffffff
#define NV_RESTART_TX 0x1
#define NV_RESTART_RX 0x2
#define NV_TX_LIMIT_COUNT 16
#define NV_DYNAMIC_THRESHOLD 4
#define NV_DYNAMIC_MAX_QUIET_COUNT 2048
/* statistics */
struct nv_ethtool_str {
char name[ETH_GSTRING_LEN];
};
static const struct nv_ethtool_str nv_estats_str[] = {
{ "tx_bytes" }, /* includes Ethernet FCS CRC */
{ "tx_zero_rexmt" },
{ "tx_one_rexmt" },
{ "tx_many_rexmt" },
{ "tx_late_collision" },
{ "tx_fifo_errors" },
{ "tx_carrier_errors" },
{ "tx_excess_deferral" },
{ "tx_retry_error" },
{ "rx_frame_error" },
{ "rx_extra_byte" },
{ "rx_late_collision" },
{ "rx_runt" },
{ "rx_frame_too_long" },
{ "rx_over_errors" },
{ "rx_crc_errors" },
{ "rx_frame_align_error" },
{ "rx_length_error" },
{ "rx_unicast" },
{ "rx_multicast" },
{ "rx_broadcast" },
{ "rx_packets" },
{ "rx_errors_total" },
{ "tx_errors_total" },
/* version 2 stats */
{ "tx_deferral" },
{ "tx_packets" },
{ "rx_bytes" }, /* includes Ethernet FCS CRC */
{ "tx_pause" },
{ "rx_pause" },
{ "rx_drop_frame" },
/* version 3 stats */
{ "tx_unicast" },
{ "tx_multicast" },
{ "tx_broadcast" }
};
struct nv_ethtool_stats {
u64 tx_bytes; /* should be ifconfig->tx_bytes + 4*tx_packets */
u64 tx_zero_rexmt;
u64 tx_one_rexmt;
u64 tx_many_rexmt;
u64 tx_late_collision;
u64 tx_fifo_errors;
u64 tx_carrier_errors;
u64 tx_excess_deferral;
u64 tx_retry_error;
u64 rx_frame_error;
u64 rx_extra_byte;
u64 rx_late_collision;
u64 rx_runt;
u64 rx_frame_too_long;
u64 rx_over_errors;
u64 rx_crc_errors;
u64 rx_frame_align_error;
u64 rx_length_error;
u64 rx_unicast;
u64 rx_multicast;
u64 rx_broadcast;
u64 rx_packets; /* should be ifconfig->rx_packets */
u64 rx_errors_total;
u64 tx_errors_total;
/* version 2 stats */
u64 tx_deferral;
u64 tx_packets; /* should be ifconfig->tx_packets */
u64 rx_bytes; /* should be ifconfig->rx_bytes + 4*rx_packets */
u64 tx_pause;
u64 rx_pause;
u64 rx_drop_frame;
/* version 3 stats */
u64 tx_unicast;
u64 tx_multicast;
u64 tx_broadcast;
};
#define NV_DEV_STATISTICS_V3_COUNT (sizeof(struct nv_ethtool_stats)/sizeof(u64))
#define NV_DEV_STATISTICS_V2_COUNT (NV_DEV_STATISTICS_V3_COUNT - 3)
#define NV_DEV_STATISTICS_V1_COUNT (NV_DEV_STATISTICS_V2_COUNT - 6)
/* diagnostics */
#define NV_TEST_COUNT_BASE 3
#define NV_TEST_COUNT_EXTENDED 4
static const struct nv_ethtool_str nv_etests_str[] = {
{ "link (online/offline)" },
{ "register (offline) " },
{ "interrupt (offline) " },
{ "loopback (offline) " }
};
struct register_test {
__u32 reg;
__u32 mask;
};
static const struct register_test nv_registers_test[] = {
{ NvRegUnknownSetupReg6, 0x01 },
{ NvRegMisc1, 0x03c },
{ NvRegOffloadConfig, 0x03ff },
{ NvRegMulticastAddrA, 0xffffffff },
{ NvRegTxWatermark, 0x0ff },
{ NvRegWakeUpFlags, 0x07777 },
{ 0, 0 }
};
struct nv_skb_map {
struct sk_buff *skb;
dma_addr_t dma;
unsigned int dma_len:31;
unsigned int dma_single:1;
struct ring_desc_ex *first_tx_desc;
struct nv_skb_map *next_tx_ctx;
};
struct nv_txrx_stats {
u64 stat_rx_packets;
u64 stat_rx_bytes; /* not always available in HW */
u64 stat_rx_missed_errors;
u64 stat_rx_dropped;
u64 stat_tx_packets; /* not always available in HW */
u64 stat_tx_bytes;
u64 stat_tx_dropped;
};
#define nv_txrx_stats_inc(member) \
__this_cpu_inc(np->txrx_stats->member)
#define nv_txrx_stats_add(member, count) \
__this_cpu_add(np->txrx_stats->member, (count))
/*
* SMP locking:
* All hardware access under netdev_priv(dev)->lock, except the performance
* critical parts:
* - rx is (pseudo-) lockless: it relies on the single-threading provided
* by the arch code for interrupts.
* - tx setup is lockless: it relies on netif_tx_lock. Actual submission
* needs netdev_priv(dev)->lock :-(
* - set_multicast_list: preparation lockless, relies on netif_tx_lock.
*
* Hardware stats updates are protected by hwstats_lock:
* - updated by nv_do_stats_poll (timer). This is meant to avoid
* integer wraparound in the NIC stats registers, at low frequency
* (0.1 Hz)
* - updated by nv_get_ethtool_stats + nv_get_stats64
*
* Software stats are accessed only through 64b synchronization points
* and are not subject to other synchronization techniques (single
* update thread on the TX or RX paths).
*/
/* in dev: base, irq */
struct fe_priv {
spinlock_t lock;
struct net_device *dev;
struct napi_struct napi;
/* hardware stats are updated in syscall and timer */
spinlock_t hwstats_lock;
struct nv_ethtool_stats estats;
int in_shutdown;
u32 linkspeed;
int duplex;
int autoneg;
int fixed_mode;
int phyaddr;
int wolenabled;
unsigned int phy_oui;
unsigned int phy_model;
unsigned int phy_rev;
u16 gigabit;
int intr_test;
int recover_error;
int quiet_count;
/* General data: RO fields */
dma_addr_t ring_addr;
struct pci_dev *pci_dev;
u32 orig_mac[2];
u32 events;
u32 irqmask;
u32 desc_ver;
u32 txrxctl_bits;
u32 vlanctl_bits;
u32 driver_data;
u32 device_id;
u32 register_size;
u32 mac_in_use;
int mgmt_version;
int mgmt_sema;
void __iomem *base;
/* rx specific fields.
* Locking: Within irq hander or disable_irq+spin_lock(&np->lock);
*/
union ring_type get_rx, put_rx, last_rx;
struct nv_skb_map *get_rx_ctx, *put_rx_ctx;
struct nv_skb_map *last_rx_ctx;
struct nv_skb_map *rx_skb;
union ring_type rx_ring;
unsigned int rx_buf_sz;
unsigned int pkt_limit;
struct timer_list oom_kick;
struct timer_list nic_poll;
struct timer_list stats_poll;
u32 nic_poll_irq;
int rx_ring_size;
/* RX software stats */
struct u64_stats_sync swstats_rx_syncp;
struct nv_txrx_stats __percpu *txrx_stats;
/* media detection workaround.
* Locking: Within irq hander or disable_irq+spin_lock(&np->lock);
*/
int need_linktimer;
unsigned long link_timeout;
/*
* tx specific fields.
*/
union ring_type get_tx, put_tx, last_tx;
struct nv_skb_map *get_tx_ctx, *put_tx_ctx;
struct nv_skb_map *last_tx_ctx;
struct nv_skb_map *tx_skb;
union ring_type tx_ring;
u32 tx_flags;
int tx_ring_size;
int tx_limit;
u32 tx_pkts_in_progress;
struct nv_skb_map *tx_change_owner;
struct nv_skb_map *tx_end_flip;
int tx_stop;
/* TX software stats */
struct u64_stats_sync swstats_tx_syncp;
/* msi/msi-x fields */
u32 msi_flags;
struct msix_entry msi_x_entry[NV_MSI_X_MAX_VECTORS];
/* flow control */
u32 pause_flags;
/* power saved state */
u32 saved_config_space[NV_PCI_REGSZ_MAX/4];
/* for different msi-x irq type */
char name_rx[IFNAMSIZ + 3]; /* -rx */
char name_tx[IFNAMSIZ + 3]; /* -tx */
char name_other[IFNAMSIZ + 6]; /* -other */
};
/*
* Maximum number of loops until we assume that a bit in the irq mask
* is stuck. Overridable with module param.
*/
static int max_interrupt_work = 4;
/*
* Optimization can be either throuput mode or cpu mode
*
* Throughput Mode: Every tx and rx packet will generate an interrupt.
* CPU Mode: Interrupts are controlled by a timer.
*/
enum {
NV_OPTIMIZATION_MODE_THROUGHPUT,
NV_OPTIMIZATION_MODE_CPU,
NV_OPTIMIZATION_MODE_DYNAMIC
};
static int optimization_mode = NV_OPTIMIZATION_MODE_DYNAMIC;
/*
* Poll interval for timer irq
*
* This interval determines how frequent an interrupt is generated.
* The is value is determined by [(time_in_micro_secs * 100) / (2^10)]
* Min = 0, and Max = 65535
*/
static int poll_interval = -1;
/*
* MSI interrupts
*/
enum {
NV_MSI_INT_DISABLED,
NV_MSI_INT_ENABLED
};
static int msi = NV_MSI_INT_ENABLED;
/*
* MSIX interrupts
*/
enum {
NV_MSIX_INT_DISABLED,
NV_MSIX_INT_ENABLED
};
static int msix = NV_MSIX_INT_ENABLED;
/*
* DMA 64bit
*/
enum {
NV_DMA_64BIT_DISABLED,
NV_DMA_64BIT_ENABLED
};
static int dma_64bit = NV_DMA_64BIT_ENABLED;
/*
* Debug output control for tx_timeout
*/
static bool debug_tx_timeout = false;
/*
* Crossover Detection
* Realtek 8201 phy + some OEM boards do not work properly.
*/
enum {
NV_CROSSOVER_DETECTION_DISABLED,
NV_CROSSOVER_DETECTION_ENABLED
};
static int phy_cross = NV_CROSSOVER_DETECTION_DISABLED;
/*
* Power down phy when interface is down (persists through reboot;
* older Linux and other OSes may not power it up again)
*/
static int phy_power_down;
static inline struct fe_priv *get_nvpriv(struct net_device *dev)
{
return netdev_priv(dev);
}
static inline u8 __iomem *get_hwbase(struct net_device *dev)
{
return ((struct fe_priv *)netdev_priv(dev))->base;
}
static inline void pci_push(u8 __iomem *base)
{
/* force out pending posted writes */
readl(base);
}
static inline u32 nv_descr_getlength(struct ring_desc *prd, u32 v)
{
return le32_to_cpu(prd->flaglen)
& ((v == DESC_VER_1) ? LEN_MASK_V1 : LEN_MASK_V2);
}
static inline u32 nv_descr_getlength_ex(struct ring_desc_ex *prd, u32 v)
{
return le32_to_cpu(prd->flaglen) & LEN_MASK_V2;
}
static bool nv_optimized(struct fe_priv *np)
{
if (np->desc_ver == DESC_VER_1 || np->desc_ver == DESC_VER_2)
return false;
return true;
}
static int reg_delay(struct net_device *dev, int offset, u32 mask, u32 target,
int delay, int delaymax)
{
u8 __iomem *base = get_hwbase(dev);
pci_push(base);
do {
udelay(delay);
delaymax -= delay;
if (delaymax < 0)
return 1;
} while ((readl(base + offset) & mask) != target);
return 0;
}
#define NV_SETUP_RX_RING 0x01
#define NV_SETUP_TX_RING 0x02
static inline u32 dma_low(dma_addr_t addr)
{
return addr;
}
static inline u32 dma_high(dma_addr_t addr)
{
return addr>>31>>1; /* 0 if 32bit, shift down by 32 if 64bit */
}
static void setup_hw_rings(struct net_device *dev, int rxtx_flags)
{
struct fe_priv *np = get_nvpriv(dev);
u8 __iomem *base = get_hwbase(dev);
if (!nv_optimized(np)) {
if (rxtx_flags & NV_SETUP_RX_RING)
writel(dma_low(np->ring_addr), base + NvRegRxRingPhysAddr);
if (rxtx_flags & NV_SETUP_TX_RING)
writel(dma_low(np->ring_addr + np->rx_ring_size*sizeof(struct ring_desc)), base + NvRegTxRingPhysAddr);
} else {
if (rxtx_flags & NV_SETUP_RX_RING) {
writel(dma_low(np->ring_addr), base + NvRegRxRingPhysAddr);
writel(dma_high(np->ring_addr), base + NvRegRxRingPhysAddrHigh);
}
if (rxtx_flags & NV_SETUP_TX_RING) {
writel(dma_low(np->ring_addr + np->rx_ring_size*sizeof(struct ring_desc_ex)), base + NvRegTxRingPhysAddr);
writel(dma_high(np->ring_addr + np->rx_ring_size*sizeof(struct ring_desc_ex)), base + NvRegTxRingPhysAddrHigh);
}
}
}
static void free_rings(struct net_device *dev)
{
struct fe_priv *np = get_nvpriv(dev);
if (!nv_optimized(np)) {
if (np->rx_ring.orig)
dma_free_coherent(&np->pci_dev->dev,
sizeof(struct ring_desc) *
(np->rx_ring_size +
np->tx_ring_size),
np->rx_ring.orig, np->ring_addr);
} else {
if (np->rx_ring.ex)
dma_free_coherent(&np->pci_dev->dev,
sizeof(struct ring_desc_ex) *
(np->rx_ring_size +
np->tx_ring_size),
np->rx_ring.ex, np->ring_addr);
}
kfree(np->rx_skb);
kfree(np->tx_skb);
}
static int using_multi_irqs(struct net_device *dev)
{
struct fe_priv *np = get_nvpriv(dev);
if (!(np->msi_flags & NV_MSI_X_ENABLED) ||
((np->msi_flags & NV_MSI_X_VECTORS_MASK) == 0x1))
return 0;
else
return 1;
}
static void nv_txrx_gate(struct net_device *dev, bool gate)
{
struct fe_priv *np = get_nvpriv(dev);
u8 __iomem *base = get_hwbase(dev);
u32 powerstate;
if (!np->mac_in_use &&
(np->driver_data & DEV_HAS_POWER_CNTRL)) {
powerstate = readl(base + NvRegPowerState2);
if (gate)
powerstate |= NVREG_POWERSTATE2_GATE_CLOCKS;
else
powerstate &= ~NVREG_POWERSTATE2_GATE_CLOCKS;
writel(powerstate, base + NvRegPowerState2);
}
}
static void nv_enable_irq(struct net_device *dev)
{
struct fe_priv *np = get_nvpriv(dev);
if (!using_multi_irqs(dev)) {
if (np->msi_flags & NV_MSI_X_ENABLED)
enable_irq(np->msi_x_entry[NV_MSI_X_VECTOR_ALL].vector);
else
enable_irq(np->pci_dev->irq);
} else {
enable_irq(np->msi_x_entry[NV_MSI_X_VECTOR_RX].vector);
enable_irq(np->msi_x_entry[NV_MSI_X_VECTOR_TX].vector);
enable_irq(np->msi_x_entry[NV_MSI_X_VECTOR_OTHER].vector);
}
}
static void nv_disable_irq(struct net_device *dev)
{
struct fe_priv *np = get_nvpriv(dev);
if (!using_multi_irqs(dev)) {
if (np->msi_flags & NV_MSI_X_ENABLED)
disable_irq(np->msi_x_entry[NV_MSI_X_VECTOR_ALL].vector);
else
disable_irq(np->pci_dev->irq);
} else {
disable_irq(np->msi_x_entry[NV_MSI_X_VECTOR_RX].vector);
disable_irq(np->msi_x_entry[NV_MSI_X_VECTOR_TX].vector);
disable_irq(np->msi_x_entry[NV_MSI_X_VECTOR_OTHER].vector);
}
}
/* In MSIX mode, a write to irqmask behaves as XOR */
static void nv_enable_hw_interrupts(struct net_device *dev, u32 mask)
{
u8 __iomem *base = get_hwbase(dev);
writel(mask, base + NvRegIrqMask);
}
static void nv_disable_hw_interrupts(struct net_device *dev, u32 mask)
{
struct fe_priv *np = get_nvpriv(dev);
u8 __iomem *base = get_hwbase(dev);
if (np->msi_flags & NV_MSI_X_ENABLED) {
writel(mask, base + NvRegIrqMask);
} else {
if (np->msi_flags & NV_MSI_ENABLED)
writel(0, base + NvRegMSIIrqMask);
writel(0, base + NvRegIrqMask);
}
}
static void nv_napi_enable(struct net_device *dev)
{
struct fe_priv *np = get_nvpriv(dev);
napi_enable(&np->napi);
}
static void nv_napi_disable(struct net_device *dev)
{
struct fe_priv *np = get_nvpriv(dev);
napi_disable(&np->napi);
}
#define MII_READ (-1)
/* mii_rw: read/write a register on the PHY.
*
* Caller must guarantee serialization
*/
static int mii_rw(struct net_device *dev, int addr, int miireg, int value)
{
u8 __iomem *base = get_hwbase(dev);
u32 reg;
int retval;
writel(NVREG_MIISTAT_MASK_RW, base + NvRegMIIStatus);
reg = readl(base + NvRegMIIControl);
if (reg & NVREG_MIICTL_INUSE) {
writel(NVREG_MIICTL_INUSE, base + NvRegMIIControl);
udelay(NV_MIIBUSY_DELAY);
}
reg = (addr << NVREG_MIICTL_ADDRSHIFT) | miireg;
if (value != MII_READ) {
writel(value, base + NvRegMIIData);
reg |= NVREG_MIICTL_WRITE;
}
writel(reg, base + NvRegMIIControl);
if (reg_delay(dev, NvRegMIIControl, NVREG_MIICTL_INUSE, 0,
NV_MIIPHY_DELAY, NV_MIIPHY_DELAYMAX)) {
retval = -1;
} else if (value != MII_READ) {
/* it was a write operation - fewer failures are detectable */
retval = 0;
} else if (readl(base + NvRegMIIStatus) & NVREG_MIISTAT_ERROR) {
retval = -1;
} else {
retval = readl(base + NvRegMIIData);
}
return retval;
}
static int phy_reset(struct net_device *dev, u32 bmcr_setup)
{
struct fe_priv *np = netdev_priv(dev);
u32 miicontrol;
unsigned int tries = 0;
miicontrol = BMCR_RESET | bmcr_setup;
if (mii_rw(dev, np->phyaddr, MII_BMCR, miicontrol))
return -1;
/* wait for 500ms */
msleep(500);
/* must wait till reset is deasserted */
while (miicontrol & BMCR_RESET) {
usleep_range(10000, 20000);
miicontrol = mii_rw(dev, np->phyaddr, MII_BMCR, MII_READ);
/* FIXME: 100 tries seem excessive */
if (tries++ > 100)
return -1;
}
return 0;
}
static int init_realtek_8211b(struct net_device *dev, struct fe_priv *np)
{
static const struct {
int reg;
int init;
} ri[] = {
{ PHY_REALTEK_INIT_REG1, PHY_REALTEK_INIT1 },
{ PHY_REALTEK_INIT_REG2, PHY_REALTEK_INIT2 },
{ PHY_REALTEK_INIT_REG1, PHY_REALTEK_INIT3 },
{ PHY_REALTEK_INIT_REG3, PHY_REALTEK_INIT4 },
{ PHY_REALTEK_INIT_REG4, PHY_REALTEK_INIT5 },
{ PHY_REALTEK_INIT_REG5, PHY_REALTEK_INIT6 },
{ PHY_REALTEK_INIT_REG1, PHY_REALTEK_INIT1 },
};
int i;
for (i = 0; i < ARRAY_SIZE(ri); i++) {
if (mii_rw(dev, np->phyaddr, ri[i].reg, ri[i].init))
return PHY_ERROR;
}
return 0;
}
static int init_realtek_8211c(struct net_device *dev, struct fe_priv *np)
{
u32 reg;
u8 __iomem *base = get_hwbase(dev);
u32 powerstate = readl(base + NvRegPowerState2);
/* need to perform hw phy reset */
powerstate |= NVREG_POWERSTATE2_PHY_RESET;
writel(powerstate, base + NvRegPowerState2);
msleep(25);
powerstate &= ~NVREG_POWERSTATE2_PHY_RESET;
writel(powerstate, base + NvRegPowerState2);
msleep(25);
reg = mii_rw(dev, np->phyaddr, PHY_REALTEK_INIT_REG6, MII_READ);
reg |= PHY_REALTEK_INIT9;
if (mii_rw(dev, np->phyaddr, PHY_REALTEK_INIT_REG6, reg))
return PHY_ERROR;
if (mii_rw(dev, np->phyaddr,
PHY_REALTEK_INIT_REG1, PHY_REALTEK_INIT10))
return PHY_ERROR;
reg = mii_rw(dev, np->phyaddr, PHY_REALTEK_INIT_REG7, MII_READ);
if (!(reg & PHY_REALTEK_INIT11)) {
reg |= PHY_REALTEK_INIT11;
if (mii_rw(dev, np->phyaddr, PHY_REALTEK_INIT_REG7, reg))
return PHY_ERROR;
}
if (mii_rw(dev, np->phyaddr,
PHY_REALTEK_INIT_REG1, PHY_REALTEK_INIT1))
return PHY_ERROR;
return 0;
}
static int init_realtek_8201(struct net_device *dev, struct fe_priv *np)
{
u32 phy_reserved;
if (np->driver_data & DEV_NEED_PHY_INIT_FIX) {
phy_reserved = mii_rw(dev, np->phyaddr,
PHY_REALTEK_INIT_REG6, MII_READ);
phy_reserved |= PHY_REALTEK_INIT7;
if (mii_rw(dev, np->phyaddr,
PHY_REALTEK_INIT_REG6, phy_reserved))
return PHY_ERROR;
}
return 0;
}
static int init_realtek_8201_cross(struct net_device *dev, struct fe_priv *np)
{
u32 phy_reserved;
if (phy_cross == NV_CROSSOVER_DETECTION_DISABLED) {
if (mii_rw(dev, np->phyaddr,
PHY_REALTEK_INIT_REG1, PHY_REALTEK_INIT3))
return PHY_ERROR;
phy_reserved = mii_rw(dev, np->phyaddr,
PHY_REALTEK_INIT_REG2, MII_READ);
phy_reserved &= ~PHY_REALTEK_INIT_MSK1;
phy_reserved |= PHY_REALTEK_INIT3;
if (mii_rw(dev, np->phyaddr,
PHY_REALTEK_INIT_REG2, phy_reserved))
return PHY_ERROR;
if (mii_rw(dev, np->phyaddr,
PHY_REALTEK_INIT_REG1, PHY_REALTEK_INIT1))
return PHY_ERROR;
}
return 0;
}
static int init_cicada(struct net_device *dev, struct fe_priv *np,
u32 phyinterface)
{
u32 phy_reserved;
if (phyinterface & PHY_RGMII) {
phy_reserved = mii_rw(dev, np->phyaddr, MII_RESV1, MII_READ);
phy_reserved &= ~(PHY_CICADA_INIT1 | PHY_CICADA_INIT2);
phy_reserved |= (PHY_CICADA_INIT3 | PHY_CICADA_INIT4);
if (mii_rw(dev, np->phyaddr, MII_RESV1, phy_reserved))
return PHY_ERROR;
phy_reserved = mii_rw(dev, np->phyaddr, MII_NCONFIG, MII_READ);
phy_reserved |= PHY_CICADA_INIT5;
if (mii_rw(dev, np->phyaddr, MII_NCONFIG, phy_reserved))
return PHY_ERROR;
}
phy_reserved = mii_rw(dev, np->phyaddr, MII_SREVISION, MII_READ);
phy_reserved |= PHY_CICADA_INIT6;
if (mii_rw(dev, np->phyaddr, MII_SREVISION, phy_reserved))
return PHY_ERROR;
return 0;
}
static int init_vitesse(struct net_device *dev, struct fe_priv *np)
{
u32 phy_reserved;
if (mii_rw(dev, np->phyaddr,
PHY_VITESSE_INIT_REG1, PHY_VITESSE_INIT1))
return PHY_ERROR;
if (mii_rw(dev, np->phyaddr,
PHY_VITESSE_INIT_REG2, PHY_VITESSE_INIT2))
return PHY_ERROR;
phy_reserved = mii_rw(dev, np->phyaddr,
PHY_VITESSE_INIT_REG4, MII_READ);
if (mii_rw(dev, np->phyaddr, PHY_VITESSE_INIT_REG4, phy_reserved))
return PHY_ERROR;
phy_reserved = mii_rw(dev, np->phyaddr,
PHY_VITESSE_INIT_REG3, MII_READ);
phy_reserved &= ~PHY_VITESSE_INIT_MSK1;
phy_reserved |= PHY_VITESSE_INIT3;
if (mii_rw(dev, np->phyaddr, PHY_VITESSE_INIT_REG3, phy_reserved))
return PHY_ERROR;
if (mii_rw(dev, np->phyaddr,
PHY_VITESSE_INIT_REG2, PHY_VITESSE_INIT4))
return PHY_ERROR;
if (mii_rw(dev, np->phyaddr,
PHY_VITESSE_INIT_REG2, PHY_VITESSE_INIT5))
return PHY_ERROR;
phy_reserved = mii_rw(dev, np->phyaddr,
PHY_VITESSE_INIT_REG4, MII_READ);
phy_reserved &= ~PHY_VITESSE_INIT_MSK1;
phy_reserved |= PHY_VITESSE_INIT3;
if (mii_rw(dev, np->phyaddr, PHY_VITESSE_INIT_REG4, phy_reserved))
return PHY_ERROR;
phy_reserved = mii_rw(dev, np->phyaddr,
PHY_VITESSE_INIT_REG3, MII_READ);
if (mii_rw(dev, np->phyaddr, PHY_VITESSE_INIT_REG3, phy_reserved))
return PHY_ERROR;
if (mii_rw(dev, np->phyaddr,
PHY_VITESSE_INIT_REG2, PHY_VITESSE_INIT6))
return PHY_ERROR;
if (mii_rw(dev, np->phyaddr,
PHY_VITESSE_INIT_REG2, PHY_VITESSE_INIT7))
return PHY_ERROR;
phy_reserved = mii_rw(dev, np->phyaddr,
PHY_VITESSE_INIT_REG4, MII_READ);
if (mii_rw(dev, np->phyaddr, PHY_VITESSE_INIT_REG4, phy_reserved))
return PHY_ERROR;
phy_reserved = mii_rw(dev, np->phyaddr,
PHY_VITESSE_INIT_REG3, MII_READ);
phy_reserved &= ~PHY_VITESSE_INIT_MSK2;
phy_reserved |= PHY_VITESSE_INIT8;
if (mii_rw(dev, np->phyaddr, PHY_VITESSE_INIT_REG3, phy_reserved))
return PHY_ERROR;
if (mii_rw(dev, np->phyaddr,
PHY_VITESSE_INIT_REG2, PHY_VITESSE_INIT9))
return PHY_ERROR;
if (mii_rw(dev, np->phyaddr,
PHY_VITESSE_INIT_REG1, PHY_VITESSE_INIT10))
return PHY_ERROR;
return 0;
}
static int phy_init(struct net_device *dev)
{
struct fe_priv *np = get_nvpriv(dev);
u8 __iomem *base = get_hwbase(dev);
u32 phyinterface;
u32 mii_status, mii_control, mii_control_1000, reg;
/* phy errata for E3016 phy */
if (np->phy_model == PHY_MODEL_MARVELL_E3016) {
reg = mii_rw(dev, np->phyaddr, MII_NCONFIG, MII_READ);
reg &= ~PHY_MARVELL_E3016_INITMASK;
if (mii_rw(dev, np->phyaddr, MII_NCONFIG, reg)) {
netdev_info(dev, "%s: phy write to errata reg failed\n",
pci_name(np->pci_dev));
return PHY_ERROR;
}
}
if (np->phy_oui == PHY_OUI_REALTEK) {
if (np->phy_model == PHY_MODEL_REALTEK_8211 &&
np->phy_rev == PHY_REV_REALTEK_8211B) {
if (init_realtek_8211b(dev, np)) {
netdev_info(dev, "%s: phy init failed\n",
pci_name(np->pci_dev));
return PHY_ERROR;
}
} else if (np->phy_model == PHY_MODEL_REALTEK_8211 &&
np->phy_rev == PHY_REV_REALTEK_8211C) {
if (init_realtek_8211c(dev, np)) {
netdev_info(dev, "%s: phy init failed\n",
pci_name(np->pci_dev));
return PHY_ERROR;
}
} else if (np->phy_model == PHY_MODEL_REALTEK_8201) {
if (init_realtek_8201(dev, np)) {
netdev_info(dev, "%s: phy init failed\n",
pci_name(np->pci_dev));
return PHY_ERROR;
}
}
}
/* set advertise register */
reg = mii_rw(dev, np->phyaddr, MII_ADVERTISE, MII_READ);
reg |= (ADVERTISE_10HALF | ADVERTISE_10FULL |
ADVERTISE_100HALF | ADVERTISE_100FULL |
ADVERTISE_PAUSE_ASYM | ADVERTISE_PAUSE_CAP);
if (mii_rw(dev, np->phyaddr, MII_ADVERTISE, reg)) {
netdev_info(dev, "%s: phy write to advertise failed\n",
pci_name(np->pci_dev));
return PHY_ERROR;
}
/* get phy interface type */
phyinterface = readl(base + NvRegPhyInterface);
/* see if gigabit phy */
mii_status = mii_rw(dev, np->phyaddr, MII_BMSR, MII_READ);
if (mii_status & PHY_GIGABIT) {
np->gigabit = PHY_GIGABIT;
mii_control_1000 = mii_rw(dev, np->phyaddr,
MII_CTRL1000, MII_READ);
mii_control_1000 &= ~ADVERTISE_1000HALF;
if (phyinterface & PHY_RGMII)
mii_control_1000 |= ADVERTISE_1000FULL;
else
mii_control_1000 &= ~ADVERTISE_1000FULL;
if (mii_rw(dev, np->phyaddr, MII_CTRL1000, mii_control_1000)) {
netdev_info(dev, "%s: phy init failed\n",
pci_name(np->pci_dev));
return PHY_ERROR;
}
} else
np->gigabit = 0;
mii_control = mii_rw(dev, np->phyaddr, MII_BMCR, MII_READ);
mii_control |= BMCR_ANENABLE;
if (np->phy_oui == PHY_OUI_REALTEK &&
np->phy_model == PHY_MODEL_REALTEK_8211 &&
np->phy_rev == PHY_REV_REALTEK_8211C) {
/* start autoneg since we already performed hw reset above */
mii_control |= BMCR_ANRESTART;
if (mii_rw(dev, np->phyaddr, MII_BMCR, mii_control)) {
netdev_info(dev, "%s: phy init failed\n",
pci_name(np->pci_dev));
return PHY_ERROR;
}
} else {
/* reset the phy
* (certain phys need bmcr to be setup with reset)
*/
if (phy_reset(dev, mii_control)) {
netdev_info(dev, "%s: phy reset failed\n",
pci_name(np->pci_dev));
return PHY_ERROR;
}
}
/* phy vendor specific configuration */
if (np->phy_oui == PHY_OUI_CICADA) {
if (init_cicada(dev, np, phyinterface)) {
netdev_info(dev, "%s: phy init failed\n",
pci_name(np->pci_dev));
return PHY_ERROR;
}
} else if (np->phy_oui == PHY_OUI_VITESSE) {
if (init_vitesse(dev, np)) {
netdev_info(dev, "%s: phy init failed\n",
pci_name(np->pci_dev));
return PHY_ERROR;
}
} else if (np->phy_oui == PHY_OUI_REALTEK) {
if (np->phy_model == PHY_MODEL_REALTEK_8211 &&
np->phy_rev == PHY_REV_REALTEK_8211B) {
/* reset could have cleared these out, set them back */
if (init_realtek_8211b(dev, np)) {
netdev_info(dev, "%s: phy init failed\n",
pci_name(np->pci_dev));
return PHY_ERROR;
}
} else if (np->phy_model == PHY_MODEL_REALTEK_8201) {
if (init_realtek_8201(dev, np) ||
init_realtek_8201_cross(dev, np)) {
netdev_info(dev, "%s: phy init failed\n",
pci_name(np->pci_dev));
return PHY_ERROR;
}
}
}
/* some phys clear out pause advertisement on reset, set it back */
mii_rw(dev, np->phyaddr, MII_ADVERTISE, reg);
/* restart auto negotiation, power down phy */
mii_control = mii_rw(dev, np->phyaddr, MII_BMCR, MII_READ);
mii_control |= (BMCR_ANRESTART | BMCR_ANENABLE);
if (phy_power_down)
mii_control |= BMCR_PDOWN;
if (mii_rw(dev, np->phyaddr, MII_BMCR, mii_control))
return PHY_ERROR;
return 0;
}
static void nv_start_rx(struct net_device *dev)
{
struct fe_priv *np = netdev_priv(dev);
u8 __iomem *base = get_hwbase(dev);
u32 rx_ctrl = readl(base + NvRegReceiverControl);
/* Already running? Stop it. */
if ((readl(base + NvRegReceiverControl) & NVREG_RCVCTL_START) && !np->mac_in_use) {
rx_ctrl &= ~NVREG_RCVCTL_START;
writel(rx_ctrl, base + NvRegReceiverControl);
pci_push(base);
}
writel(np->linkspeed, base + NvRegLinkSpeed);
pci_push(base);
rx_ctrl |= NVREG_RCVCTL_START;
if (np->mac_in_use)
rx_ctrl &= ~NVREG_RCVCTL_RX_PATH_EN;
writel(rx_ctrl, base + NvRegReceiverControl);
pci_push(base);
}
static void nv_stop_rx(struct net_device *dev)
{
struct fe_priv *np = netdev_priv(dev);
u8 __iomem *base = get_hwbase(dev);
u32 rx_ctrl = readl(base + NvRegReceiverControl);
if (!np->mac_in_use)
rx_ctrl &= ~NVREG_RCVCTL_START;
else
rx_ctrl |= NVREG_RCVCTL_RX_PATH_EN;
writel(rx_ctrl, base + NvRegReceiverControl);
if (reg_delay(dev, NvRegReceiverStatus, NVREG_RCVSTAT_BUSY, 0,
NV_RXSTOP_DELAY1, NV_RXSTOP_DELAY1MAX))
netdev_info(dev, "%s: ReceiverStatus remained busy\n",
__func__);
udelay(NV_RXSTOP_DELAY2);
if (!np->mac_in_use)
writel(0, base + NvRegLinkSpeed);
}
static void nv_start_tx(struct net_device *dev)
{
struct fe_priv *np = netdev_priv(dev);
u8 __iomem *base = get_hwbase(dev);
u32 tx_ctrl = readl(base + NvRegTransmitterControl);
tx_ctrl |= NVREG_XMITCTL_START;
if (np->mac_in_use)
tx_ctrl &= ~NVREG_XMITCTL_TX_PATH_EN;
writel(tx_ctrl, base + NvRegTransmitterControl);
pci_push(base);
}
static void nv_stop_tx(struct net_device *dev)
{
struct fe_priv *np = netdev_priv(dev);
u8 __iomem *base = get_hwbase(dev);
u32 tx_ctrl = readl(base + NvRegTransmitterControl);
if (!np->mac_in_use)
tx_ctrl &= ~NVREG_XMITCTL_START;
else
tx_ctrl |= NVREG_XMITCTL_TX_PATH_EN;
writel(tx_ctrl, base + NvRegTransmitterControl);
if (reg_delay(dev, NvRegTransmitterStatus, NVREG_XMITSTAT_BUSY, 0,
NV_TXSTOP_DELAY1, NV_TXSTOP_DELAY1MAX))
netdev_info(dev, "%s: TransmitterStatus remained busy\n",
__func__);
udelay(NV_TXSTOP_DELAY2);
if (!np->mac_in_use)
writel(readl(base + NvRegTransmitPoll) & NVREG_TRANSMITPOLL_MAC_ADDR_REV,
base + NvRegTransmitPoll);
}
static void nv_start_rxtx(struct net_device *dev)
{
nv_start_rx(dev);
nv_start_tx(dev);
}
static void nv_stop_rxtx(struct net_device *dev)
{
nv_stop_rx(dev);
nv_stop_tx(dev);
}
static void nv_txrx_reset(struct net_device *dev)
{
struct fe_priv *np = netdev_priv(dev);
u8 __iomem *base = get_hwbase(dev);
writel(NVREG_TXRXCTL_BIT2 | NVREG_TXRXCTL_RESET | np->txrxctl_bits, base + NvRegTxRxControl);
pci_push(base);
udelay(NV_TXRX_RESET_DELAY);
writel(NVREG_TXRXCTL_BIT2 | np->txrxctl_bits, base + NvRegTxRxControl);
pci_push(base);
}
static void nv_mac_reset(struct net_device *dev)
{
struct fe_priv *np = netdev_priv(dev);
u8 __iomem *base = get_hwbase(dev);
u32 temp1, temp2, temp3;
writel(NVREG_TXRXCTL_BIT2 | NVREG_TXRXCTL_RESET | np->txrxctl_bits, base + NvRegTxRxControl);
pci_push(base);
/* save registers since they will be cleared on reset */
temp1 = readl(base + NvRegMacAddrA);
temp2 = readl(base + NvRegMacAddrB);
temp3 = readl(base + NvRegTransmitPoll);
writel(NVREG_MAC_RESET_ASSERT, base + NvRegMacReset);
pci_push(base);
udelay(NV_MAC_RESET_DELAY);
writel(0, base + NvRegMacReset);
pci_push(base);
udelay(NV_MAC_RESET_DELAY);
/* restore saved registers */
writel(temp1, base + NvRegMacAddrA);
writel(temp2, base + NvRegMacAddrB);
writel(temp3, base + NvRegTransmitPoll);
writel(NVREG_TXRXCTL_BIT2 | np->txrxctl_bits, base + NvRegTxRxControl);
pci_push(base);
}
/* Caller must appropriately lock netdev_priv(dev)->hwstats_lock */
static void nv_update_stats(struct net_device *dev)
{
struct fe_priv *np = netdev_priv(dev);
u8 __iomem *base = get_hwbase(dev);
lockdep_assert_held(&np->hwstats_lock);
/* query hardware */
np->estats.tx_bytes += readl(base + NvRegTxCnt);
np->estats.tx_zero_rexmt += readl(base + NvRegTxZeroReXmt);
np->estats.tx_one_rexmt += readl(base + NvRegTxOneReXmt);
np->estats.tx_many_rexmt += readl(base + NvRegTxManyReXmt);
np->estats.tx_late_collision += readl(base + NvRegTxLateCol);
np->estats.tx_fifo_errors += readl(base + NvRegTxUnderflow);
np->estats.tx_carrier_errors += readl(base + NvRegTxLossCarrier);
np->estats.tx_excess_deferral += readl(base + NvRegTxExcessDef);
np->estats.tx_retry_error += readl(base + NvRegTxRetryErr);
np->estats.rx_frame_error += readl(base + NvRegRxFrameErr);
np->estats.rx_extra_byte += readl(base + NvRegRxExtraByte);
np->estats.rx_late_collision += readl(base + NvRegRxLateCol);
np->estats.rx_runt += readl(base + NvRegRxRunt);
np->estats.rx_frame_too_long += readl(base + NvRegRxFrameTooLong);
np->estats.rx_over_errors += readl(base + NvRegRxOverflow);
np->estats.rx_crc_errors += readl(base + NvRegRxFCSErr);
np->estats.rx_frame_align_error += readl(base + NvRegRxFrameAlignErr);
np->estats.rx_length_error += readl(base + NvRegRxLenErr);
np->estats.rx_unicast += readl(base + NvRegRxUnicast);
np->estats.rx_multicast += readl(base + NvRegRxMulticast);
np->estats.rx_broadcast += readl(base + NvRegRxBroadcast);
np->estats.rx_packets =
np->estats.rx_unicast +
np->estats.rx_multicast +
np->estats.rx_broadcast;
np->estats.rx_errors_total =
np->estats.rx_crc_errors +
np->estats.rx_over_errors +
np->estats.rx_frame_error +
(np->estats.rx_frame_align_error - np->estats.rx_extra_byte) +
np->estats.rx_late_collision +
np->estats.rx_runt +
np->estats.rx_frame_too_long;
np->estats.tx_errors_total =
np->estats.tx_late_collision +
np->estats.tx_fifo_errors +
np->estats.tx_carrier_errors +
np->estats.tx_excess_deferral +
np->estats.tx_retry_error;
if (np->driver_data & DEV_HAS_STATISTICS_V2) {
np->estats.tx_deferral += readl(base + NvRegTxDef);
np->estats.tx_packets += readl(base + NvRegTxFrame);
np->estats.rx_bytes += readl(base + NvRegRxCnt);
np->estats.tx_pause += readl(base + NvRegTxPause);
np->estats.rx_pause += readl(base + NvRegRxPause);
np->estats.rx_drop_frame += readl(base + NvRegRxDropFrame);
np->estats.rx_errors_total += np->estats.rx_drop_frame;
}
if (np->driver_data & DEV_HAS_STATISTICS_V3) {
np->estats.tx_unicast += readl(base + NvRegTxUnicast);
np->estats.tx_multicast += readl(base + NvRegTxMulticast);
np->estats.tx_broadcast += readl(base + NvRegTxBroadcast);
}
}
static void nv_get_stats(int cpu, struct fe_priv *np,
struct rtnl_link_stats64 *storage)
{
struct nv_txrx_stats *src = per_cpu_ptr(np->txrx_stats, cpu);
unsigned int syncp_start;
u64 rx_packets, rx_bytes, rx_dropped, rx_missed_errors;
u64 tx_packets, tx_bytes, tx_dropped;
do {
syncp_start = u64_stats_fetch_begin_irq(&np->swstats_rx_syncp);
rx_packets = src->stat_rx_packets;
rx_bytes = src->stat_rx_bytes;
rx_dropped = src->stat_rx_dropped;
rx_missed_errors = src->stat_rx_missed_errors;
} while (u64_stats_fetch_retry_irq(&np->swstats_rx_syncp, syncp_start));
storage->rx_packets += rx_packets;
storage->rx_bytes += rx_bytes;
storage->rx_dropped += rx_dropped;
storage->rx_missed_errors += rx_missed_errors;
do {
syncp_start = u64_stats_fetch_begin_irq(&np->swstats_tx_syncp);
tx_packets = src->stat_tx_packets;
tx_bytes = src->stat_tx_bytes;
tx_dropped = src->stat_tx_dropped;
} while (u64_stats_fetch_retry_irq(&np->swstats_tx_syncp, syncp_start));
storage->tx_packets += tx_packets;
storage->tx_bytes += tx_bytes;
storage->tx_dropped += tx_dropped;
}
/*
* nv_get_stats64: dev->ndo_get_stats64 function
* Get latest stats value from the nic.
* Called with read_lock(&dev_base_lock) held for read -
* only synchronized against unregister_netdevice.
*/
static void
nv_get_stats64(struct net_device *dev, struct rtnl_link_stats64 *storage)
__acquires(&netdev_priv(dev)->hwstats_lock)
__releases(&netdev_priv(dev)->hwstats_lock)
{
struct fe_priv *np = netdev_priv(dev);
int cpu;
/*
* Note: because HW stats are not always available and for
* consistency reasons, the following ifconfig stats are
* managed by software: rx_bytes, tx_bytes, rx_packets and
* tx_packets. The related hardware stats reported by ethtool
* should be equivalent to these ifconfig stats, with 4
* additional bytes per packet (Ethernet FCS CRC), except for
* tx_packets when TSO kicks in.
*/
/* software stats */
for_each_online_cpu(cpu)
nv_get_stats(cpu, np, storage);
/* If the nic supports hw counters then retrieve latest values */
if (np->driver_data & DEV_HAS_STATISTICS_V123) {
spin_lock_bh(&np->hwstats_lock);
nv_update_stats(dev);
/* generic stats */
storage->rx_errors = np->estats.rx_errors_total;
storage->tx_errors = np->estats.tx_errors_total;
/* meaningful only when NIC supports stats v3 */
storage->multicast = np->estats.rx_multicast;
/* detailed rx_errors */
storage->rx_length_errors = np->estats.rx_length_error;
storage->rx_over_errors = np->estats.rx_over_errors;
storage->rx_crc_errors = np->estats.rx_crc_errors;
storage->rx_frame_errors = np->estats.rx_frame_align_error;
storage->rx_fifo_errors = np->estats.rx_drop_frame;
/* detailed tx_errors */
storage->tx_carrier_errors = np->estats.tx_carrier_errors;
storage->tx_fifo_errors = np->estats.tx_fifo_errors;
spin_unlock_bh(&np->hwstats_lock);
}
}
/*
* nv_alloc_rx: fill rx ring entries.
* Return 1 if the allocations for the skbs failed and the
* rx engine is without Available descriptors
*/
static int nv_alloc_rx(struct net_device *dev)
{
struct fe_priv *np = netdev_priv(dev);
struct ring_desc *less_rx;
less_rx = np->get_rx.orig;
if (less_rx-- == np->rx_ring.orig)
less_rx = np->last_rx.orig;
while (np->put_rx.orig != less_rx) {
struct sk_buff *skb = netdev_alloc_skb(dev, np->rx_buf_sz + NV_RX_ALLOC_PAD);
if (likely(skb)) {
np->put_rx_ctx->skb = skb;
np->put_rx_ctx->dma = dma_map_single(&np->pci_dev->dev,
skb->data,
skb_tailroom(skb),
DMA_FROM_DEVICE);
if (unlikely(dma_mapping_error(&np->pci_dev->dev,
np->put_rx_ctx->dma))) {
kfree_skb(skb);
goto packet_dropped;
}
np->put_rx_ctx->dma_len = skb_tailroom(skb);
np->put_rx.orig->buf = cpu_to_le32(np->put_rx_ctx->dma);
wmb();
np->put_rx.orig->flaglen = cpu_to_le32(np->rx_buf_sz | NV_RX_AVAIL);
if (unlikely(np->put_rx.orig++ == np->last_rx.orig))
np->put_rx.orig = np->rx_ring.orig;
if (unlikely(np->put_rx_ctx++ == np->last_rx_ctx))
np->put_rx_ctx = np->rx_skb;
} else {
packet_dropped:
u64_stats_update_begin(&np->swstats_rx_syncp);
nv_txrx_stats_inc(stat_rx_dropped);
u64_stats_update_end(&np->swstats_rx_syncp);
return 1;
}
}
return 0;
}
static int nv_alloc_rx_optimized(struct net_device *dev)
{
struct fe_priv *np = netdev_priv(dev);
struct ring_desc_ex *less_rx;
less_rx = np->get_rx.ex;
if (less_rx-- == np->rx_ring.ex)
less_rx = np->last_rx.ex;
while (np->put_rx.ex != less_rx) {
struct sk_buff *skb = netdev_alloc_skb(dev, np->rx_buf_sz + NV_RX_ALLOC_PAD);
if (likely(skb)) {
np->put_rx_ctx->skb = skb;
np->put_rx_ctx->dma = dma_map_single(&np->pci_dev->dev,
skb->data,
skb_tailroom(skb),
DMA_FROM_DEVICE);
if (unlikely(dma_mapping_error(&np->pci_dev->dev,
np->put_rx_ctx->dma))) {
kfree_skb(skb);
goto packet_dropped;
}
np->put_rx_ctx->dma_len = skb_tailroom(skb);
np->put_rx.ex->bufhigh = cpu_to_le32(dma_high(np->put_rx_ctx->dma));
np->put_rx.ex->buflow = cpu_to_le32(dma_low(np->put_rx_ctx->dma));
wmb();
np->put_rx.ex->flaglen = cpu_to_le32(np->rx_buf_sz | NV_RX2_AVAIL);
if (unlikely(np->put_rx.ex++ == np->last_rx.ex))
np->put_rx.ex = np->rx_ring.ex;
if (unlikely(np->put_rx_ctx++ == np->last_rx_ctx))
np->put_rx_ctx = np->rx_skb;
} else {
packet_dropped:
u64_stats_update_begin(&np->swstats_rx_syncp);
nv_txrx_stats_inc(stat_rx_dropped);
u64_stats_update_end(&np->swstats_rx_syncp);
return 1;
}
}
return 0;
}
/* If rx bufs are exhausted called after 50ms to attempt to refresh */
static void nv_do_rx_refill(struct timer_list *t)
{
struct fe_priv *np = from_timer(np, t, oom_kick);
/* Just reschedule NAPI rx processing */
napi_schedule(&np->napi);
}
static void nv_init_rx(struct net_device *dev)
{
struct fe_priv *np = netdev_priv(dev);
int i;
np->get_rx = np->rx_ring;
np->put_rx = np->rx_ring;
if (!nv_optimized(np))
np->last_rx.orig = &np->rx_ring.orig[np->rx_ring_size-1];
else
np->last_rx.ex = &np->rx_ring.ex[np->rx_ring_size-1];
np->get_rx_ctx = np->rx_skb;
np->put_rx_ctx = np->rx_skb;
np->last_rx_ctx = &np->rx_skb[np->rx_ring_size-1];
for (i = 0; i < np->rx_ring_size; i++) {
if (!nv_optimized(np)) {
np->rx_ring.orig[i].flaglen = 0;
np->rx_ring.orig[i].buf = 0;
} else {
np->rx_ring.ex[i].flaglen = 0;
np->rx_ring.ex[i].txvlan = 0;
np->rx_ring.ex[i].bufhigh = 0;
np->rx_ring.ex[i].buflow = 0;
}
np->rx_skb[i].skb = NULL;
np->rx_skb[i].dma = 0;
}
}
static void nv_init_tx(struct net_device *dev)
{
struct fe_priv *np = netdev_priv(dev);
int i;
np->get_tx = np->tx_ring;
np->put_tx = np->tx_ring;
if (!nv_optimized(np))
np->last_tx.orig = &np->tx_ring.orig[np->tx_ring_size-1];
else
np->last_tx.ex = &np->tx_ring.ex[np->tx_ring_size-1];
np->get_tx_ctx = np->tx_skb;
np->put_tx_ctx = np->tx_skb;
np->last_tx_ctx = &np->tx_skb[np->tx_ring_size-1];
netdev_reset_queue(np->dev);
np->tx_pkts_in_progress = 0;
np->tx_change_owner = NULL;
np->tx_end_flip = NULL;
np->tx_stop = 0;
for (i = 0; i < np->tx_ring_size; i++) {
if (!nv_optimized(np)) {
np->tx_ring.orig[i].flaglen = 0;
np->tx_ring.orig[i].buf = 0;
} else {
np->tx_ring.ex[i].flaglen = 0;
np->tx_ring.ex[i].txvlan = 0;
np->tx_ring.ex[i].bufhigh = 0;
np->tx_ring.ex[i].buflow = 0;
}
np->tx_skb[i].skb = NULL;
np->tx_skb[i].dma = 0;
np->tx_skb[i].dma_len = 0;
np->tx_skb[i].dma_single = 0;
np->tx_skb[i].first_tx_desc = NULL;
np->tx_skb[i].next_tx_ctx = NULL;
}
}
static int nv_init_ring(struct net_device *dev)
{
struct fe_priv *np = netdev_priv(dev);
nv_init_tx(dev);
nv_init_rx(dev);
if (!nv_optimized(np))
return nv_alloc_rx(dev);
else
return nv_alloc_rx_optimized(dev);
}
static void nv_unmap_txskb(struct fe_priv *np, struct nv_skb_map *tx_skb)
{
if (tx_skb->dma) {
if (tx_skb->dma_single)
dma_unmap_single(&np->pci_dev->dev, tx_skb->dma,
tx_skb->dma_len,
DMA_TO_DEVICE);
else
dma_unmap_page(&np->pci_dev->dev, tx_skb->dma,
tx_skb->dma_len,
DMA_TO_DEVICE);
tx_skb->dma = 0;
}
}
static int nv_release_txskb(struct fe_priv *np, struct nv_skb_map *tx_skb)
{
nv_unmap_txskb(np, tx_skb);
if (tx_skb->skb) {
dev_kfree_skb_any(tx_skb->skb);
tx_skb->skb = NULL;
return 1;
}
return 0;
}
static void nv_drain_tx(struct net_device *dev)
{
struct fe_priv *np = netdev_priv(dev);
unsigned int i;
for (i = 0; i < np->tx_ring_size; i++) {
if (!nv_optimized(np)) {
np->tx_ring.orig[i].flaglen = 0;
np->tx_ring.orig[i].buf = 0;
} else {
np->tx_ring.ex[i].flaglen = 0;
np->tx_ring.ex[i].txvlan = 0;
np->tx_ring.ex[i].bufhigh = 0;
np->tx_ring.ex[i].buflow = 0;
}
if (nv_release_txskb(np, &np->tx_skb[i])) {
u64_stats_update_begin(&np->swstats_tx_syncp);
nv_txrx_stats_inc(stat_tx_dropped);
u64_stats_update_end(&np->swstats_tx_syncp);
}
np->tx_skb[i].dma = 0;
np->tx_skb[i].dma_len = 0;
np->tx_skb[i].dma_single = 0;
np->tx_skb[i].first_tx_desc = NULL;
np->tx_skb[i].next_tx_ctx = NULL;
}
np->tx_pkts_in_progress = 0;
np->tx_change_owner = NULL;
np->tx_end_flip = NULL;
}
static void nv_drain_rx(struct net_device *dev)
{
struct fe_priv *np = netdev_priv(dev);
int i;
for (i = 0; i < np->rx_ring_size; i++) {
if (!nv_optimized(np)) {
np->rx_ring.orig[i].flaglen = 0;
np->rx_ring.orig[i].buf = 0;
} else {
np->rx_ring.ex[i].flaglen = 0;
np->rx_ring.ex[i].txvlan = 0;
np->rx_ring.ex[i].bufhigh = 0;
np->rx_ring.ex[i].buflow = 0;
}
wmb();
if (np->rx_skb[i].skb) {
dma_unmap_single(&np->pci_dev->dev, np->rx_skb[i].dma,
(skb_end_pointer(np->rx_skb[i].skb) -
np->rx_skb[i].skb->data),
DMA_FROM_DEVICE);
dev_kfree_skb(np->rx_skb[i].skb);
np->rx_skb[i].skb = NULL;
}
}
}
static void nv_drain_rxtx(struct net_device *dev)
{
nv_drain_tx(dev);
nv_drain_rx(dev);
}
static inline u32 nv_get_empty_tx_slots(struct fe_priv *np)
{
return (u32)(np->tx_ring_size - ((np->tx_ring_size + (np->put_tx_ctx - np->get_tx_ctx)) % np->tx_ring_size));
}
static void nv_legacybackoff_reseed(struct net_device *dev)
{
u8 __iomem *base = get_hwbase(dev);
u32 reg;
u32 low;
int tx_status = 0;
reg = readl(base + NvRegSlotTime) & ~NVREG_SLOTTIME_MASK;
get_random_bytes(&low, sizeof(low));
reg |= low & NVREG_SLOTTIME_MASK;
/* Need to stop tx before change takes effect.
* Caller has already gained np->lock.
*/
tx_status = readl(base + NvRegTransmitterControl) & NVREG_XMITCTL_START;
if (tx_status)
nv_stop_tx(dev);
nv_stop_rx(dev);
writel(reg, base + NvRegSlotTime);
if (tx_status)
nv_start_tx(dev);
nv_start_rx(dev);
}
/* Gear Backoff Seeds */
#define BACKOFF_SEEDSET_ROWS 8
#define BACKOFF_SEEDSET_LFSRS 15
/* Known Good seed sets */
static const u32 main_seedset[BACKOFF_SEEDSET_ROWS][BACKOFF_SEEDSET_LFSRS] = {
{145, 155, 165, 175, 185, 196, 235, 245, 255, 265, 275, 285, 660, 690, 874},
{245, 255, 265, 575, 385, 298, 335, 345, 355, 366, 375, 385, 761, 790, 974},
{145, 155, 165, 175, 185, 196, 235, 245, 255, 265, 275, 285, 660, 690, 874},
{245, 255, 265, 575, 385, 298, 335, 345, 355, 366, 375, 386, 761, 790, 974},
{266, 265, 276, 585, 397, 208, 345, 355, 365, 376, 385, 396, 771, 700, 984},
{266, 265, 276, 586, 397, 208, 346, 355, 365, 376, 285, 396, 771, 700, 984},
{366, 365, 376, 686, 497, 308, 447, 455, 466, 476, 485, 496, 871, 800, 84},
{466, 465, 476, 786, 597, 408, 547, 555, 566, 576, 585, 597, 971, 900, 184} };
static const u32 gear_seedset[BACKOFF_SEEDSET_ROWS][BACKOFF_SEEDSET_LFSRS] = {
{251, 262, 273, 324, 319, 508, 375, 364, 341, 371, 398, 193, 375, 30, 295},
{351, 375, 373, 469, 551, 639, 477, 464, 441, 472, 498, 293, 476, 130, 395},
{351, 375, 373, 469, 551, 639, 477, 464, 441, 472, 498, 293, 476, 130, 397},
{251, 262, 273, 324, 319, 508, 375, 364, 341, 371, 398, 193, 375, 30, 295},
{251, 262, 273, 324, 319, 508, 375, 364, 341, 371, 398, 193, 375, 30, 295},
{351, 375, 373, 469, 551, 639, 477, 464, 441, 472, 498, 293, 476, 130, 395},
{351, 375, 373, 469, 551, 639, 477, 464, 441, 472, 498, 293, 476, 130, 395},
{351, 375, 373, 469, 551, 639, 477, 464, 441, 472, 498, 293, 476, 130, 395} };
static void nv_gear_backoff_reseed(struct net_device *dev)
{
u8 __iomem *base = get_hwbase(dev);
u32 miniseed1, miniseed2, miniseed2_reversed, miniseed3, miniseed3_reversed;
u32 temp, seedset, combinedSeed;
int i;
/* Setup seed for free running LFSR */
/* We are going to read the time stamp counter 3 times
and swizzle bits around to increase randomness */
get_random_bytes(&miniseed1, sizeof(miniseed1));
miniseed1 &= 0x0fff;
if (miniseed1 == 0)
miniseed1 = 0xabc;
get_random_bytes(&miniseed2, sizeof(miniseed2));
miniseed2 &= 0x0fff;
if (miniseed2 == 0)
miniseed2 = 0xabc;
miniseed2_reversed =
((miniseed2 & 0xF00) >> 8) |
(miniseed2 & 0x0F0) |
((miniseed2 & 0x00F) << 8);
get_random_bytes(&miniseed3, sizeof(miniseed3));
miniseed3 &= 0x0fff;
if (miniseed3 == 0)
miniseed3 = 0xabc;
miniseed3_reversed =
((miniseed3 & 0xF00) >> 8) |
(miniseed3 & 0x0F0) |
((miniseed3 & 0x00F) << 8);
combinedSeed = ((miniseed1 ^ miniseed2_reversed) << 12) |
(miniseed2 ^ miniseed3_reversed);
/* Seeds can not be zero */
if ((combinedSeed & NVREG_BKOFFCTRL_SEED_MASK) == 0)
combinedSeed |= 0x08;
if ((combinedSeed & (NVREG_BKOFFCTRL_SEED_MASK << NVREG_BKOFFCTRL_GEAR)) == 0)
combinedSeed |= 0x8000;
/* No need to disable tx here */
temp = NVREG_BKOFFCTRL_DEFAULT | (0 << NVREG_BKOFFCTRL_SELECT);
temp |= combinedSeed & NVREG_BKOFFCTRL_SEED_MASK;
temp |= combinedSeed >> NVREG_BKOFFCTRL_GEAR;
writel(temp, base + NvRegBackOffControl);
/* Setup seeds for all gear LFSRs. */
get_random_bytes(&seedset, sizeof(seedset));
seedset = seedset % BACKOFF_SEEDSET_ROWS;
for (i = 1; i <= BACKOFF_SEEDSET_LFSRS; i++) {
temp = NVREG_BKOFFCTRL_DEFAULT | (i << NVREG_BKOFFCTRL_SELECT);
temp |= main_seedset[seedset][i-1] & 0x3ff;
temp |= ((gear_seedset[seedset][i-1] & 0x3ff) << NVREG_BKOFFCTRL_GEAR);
writel(temp, base + NvRegBackOffControl);
}
}
/*
* nv_start_xmit: dev->hard_start_xmit function
* Called with netif_tx_lock held.
*/
static netdev_tx_t nv_start_xmit(struct sk_buff *skb, struct net_device *dev)
{
struct fe_priv *np = netdev_priv(dev);
u32 tx_flags = 0;
u32 tx_flags_extra = (np->desc_ver == DESC_VER_1 ? NV_TX_LASTPACKET : NV_TX2_LASTPACKET);
unsigned int fragments = skb_shinfo(skb)->nr_frags;
unsigned int i;
u32 offset = 0;
u32 bcnt;
u32 size = skb_headlen(skb);
u32 entries = (size >> NV_TX2_TSO_MAX_SHIFT) + ((size & (NV_TX2_TSO_MAX_SIZE-1)) ? 1 : 0);
u32 empty_slots;
struct ring_desc *put_tx;
struct ring_desc *start_tx;
struct ring_desc *prev_tx;
struct nv_skb_map *prev_tx_ctx;
struct nv_skb_map *tmp_tx_ctx = NULL, *start_tx_ctx = NULL;
unsigned long flags;
netdev_tx_t ret = NETDEV_TX_OK;
/* add fragments to entries count */
for (i = 0; i < fragments; i++) {
u32 frag_size = skb_frag_size(&skb_shinfo(skb)->frags[i]);
entries += (frag_size >> NV_TX2_TSO_MAX_SHIFT) +
((frag_size & (NV_TX2_TSO_MAX_SIZE-1)) ? 1 : 0);
}
spin_lock_irqsave(&np->lock, flags);
empty_slots = nv_get_empty_tx_slots(np);
if (unlikely(empty_slots <= entries)) {
netif_stop_queue(dev);
np->tx_stop = 1;
spin_unlock_irqrestore(&np->lock, flags);
/* When normal packets and/or xmit_more packets fill up
* tx_desc, it is necessary to trigger NIC tx reg.
*/
ret = NETDEV_TX_BUSY;
goto txkick;
}
spin_unlock_irqrestore(&np->lock, flags);
start_tx = put_tx = np->put_tx.orig;
/* setup the header buffer */
do {
bcnt = (size > NV_TX2_TSO_MAX_SIZE) ? NV_TX2_TSO_MAX_SIZE : size;
np->put_tx_ctx->dma = dma_map_single(&np->pci_dev->dev,
skb->data + offset, bcnt,
DMA_TO_DEVICE);
if (unlikely(dma_mapping_error(&np->pci_dev->dev,
np->put_tx_ctx->dma))) {
/* on DMA mapping error - drop the packet */
dev_kfree_skb_any(skb);
u64_stats_update_begin(&np->swstats_tx_syncp);
nv_txrx_stats_inc(stat_tx_dropped);
u64_stats_update_end(&np->swstats_tx_syncp);
ret = NETDEV_TX_OK;
goto dma_error;
}
np->put_tx_ctx->dma_len = bcnt;
np->put_tx_ctx->dma_single = 1;
put_tx->buf = cpu_to_le32(np->put_tx_ctx->dma);
put_tx->flaglen = cpu_to_le32((bcnt-1) | tx_flags);
tx_flags = np->tx_flags;
offset += bcnt;
size -= bcnt;
if (unlikely(put_tx++ == np->last_tx.orig))
put_tx = np->tx_ring.orig;
if (unlikely(np->put_tx_ctx++ == np->last_tx_ctx))
np->put_tx_ctx = np->tx_skb;
} while (size);
/* setup the fragments */
for (i = 0; i < fragments; i++) {
const skb_frag_t *frag = &skb_shinfo(skb)->frags[i];
u32 frag_size = skb_frag_size(frag);
offset = 0;
do {
if (!start_tx_ctx)
start_tx_ctx = tmp_tx_ctx = np->put_tx_ctx;
bcnt = (frag_size > NV_TX2_TSO_MAX_SIZE) ? NV_TX2_TSO_MAX_SIZE : frag_size;
np->put_tx_ctx->dma = skb_frag_dma_map(
&np->pci_dev->dev,
frag, offset,
bcnt,
DMA_TO_DEVICE);
if (unlikely(dma_mapping_error(&np->pci_dev->dev,
np->put_tx_ctx->dma))) {
/* Unwind the mapped fragments */
do {
nv_unmap_txskb(np, start_tx_ctx);
if (unlikely(tmp_tx_ctx++ == np->last_tx_ctx))
tmp_tx_ctx = np->tx_skb;
} while (tmp_tx_ctx != np->put_tx_ctx);
dev_kfree_skb_any(skb);
np->put_tx_ctx = start_tx_ctx;
u64_stats_update_begin(&np->swstats_tx_syncp);
nv_txrx_stats_inc(stat_tx_dropped);
u64_stats_update_end(&np->swstats_tx_syncp);
ret = NETDEV_TX_OK;
goto dma_error;
}
np->put_tx_ctx->dma_len = bcnt;
np->put_tx_ctx->dma_single = 0;
put_tx->buf = cpu_to_le32(np->put_tx_ctx->dma);
put_tx->flaglen = cpu_to_le32((bcnt-1) | tx_flags);
offset += bcnt;
frag_size -= bcnt;
if (unlikely(put_tx++ == np->last_tx.orig))
put_tx = np->tx_ring.orig;
if (unlikely(np->put_tx_ctx++ == np->last_tx_ctx))
np->put_tx_ctx = np->tx_skb;
} while (frag_size);
}
if (unlikely(put_tx == np->tx_ring.orig))
prev_tx = np->last_tx.orig;
else
prev_tx = put_tx - 1;
if (unlikely(np->put_tx_ctx == np->tx_skb))
prev_tx_ctx = np->last_tx_ctx;
else
prev_tx_ctx = np->put_tx_ctx - 1;
/* set last fragment flag */
prev_tx->flaglen |= cpu_to_le32(tx_flags_extra);
/* save skb in this slot's context area */
prev_tx_ctx->skb = skb;
if (skb_is_gso(skb))
tx_flags_extra = NV_TX2_TSO | (skb_shinfo(skb)->gso_size << NV_TX2_TSO_SHIFT);
else
tx_flags_extra = skb->ip_summed == CHECKSUM_PARTIAL ?
NV_TX2_CHECKSUM_L3 | NV_TX2_CHECKSUM_L4 : 0;
spin_lock_irqsave(&np->lock, flags);
/* set tx flags */
start_tx->flaglen |= cpu_to_le32(tx_flags | tx_flags_extra);
netdev_sent_queue(np->dev, skb->len);
skb_tx_timestamp(skb);
np->put_tx.orig = put_tx;
spin_unlock_irqrestore(&np->lock, flags);
txkick:
if (netif_queue_stopped(dev) || !netdev_xmit_more()) {
u32 txrxctl_kick;
dma_error:
txrxctl_kick = NVREG_TXRXCTL_KICK | np->txrxctl_bits;
writel(txrxctl_kick, get_hwbase(dev) + NvRegTxRxControl);
}
return ret;
}
static netdev_tx_t nv_start_xmit_optimized(struct sk_buff *skb,
struct net_device *dev)
{
struct fe_priv *np = netdev_priv(dev);
u32 tx_flags = 0;
u32 tx_flags_extra;
unsigned int fragments = skb_shinfo(skb)->nr_frags;
unsigned int i;
u32 offset = 0;
u32 bcnt;
u32 size = skb_headlen(skb);
u32 entries = (size >> NV_TX2_TSO_MAX_SHIFT) + ((size & (NV_TX2_TSO_MAX_SIZE-1)) ? 1 : 0);
u32 empty_slots;
struct ring_desc_ex *put_tx;
struct ring_desc_ex *start_tx;
struct ring_desc_ex *prev_tx;
struct nv_skb_map *prev_tx_ctx;
struct nv_skb_map *start_tx_ctx = NULL;
struct nv_skb_map *tmp_tx_ctx = NULL;
unsigned long flags;
netdev_tx_t ret = NETDEV_TX_OK;
/* add fragments to entries count */
for (i = 0; i < fragments; i++) {
u32 frag_size = skb_frag_size(&skb_shinfo(skb)->frags[i]);
entries += (frag_size >> NV_TX2_TSO_MAX_SHIFT) +
((frag_size & (NV_TX2_TSO_MAX_SIZE-1)) ? 1 : 0);
}
spin_lock_irqsave(&np->lock, flags);
empty_slots = nv_get_empty_tx_slots(np);
if (unlikely(empty_slots <= entries)) {
netif_stop_queue(dev);
np->tx_stop = 1;
spin_unlock_irqrestore(&np->lock, flags);
/* When normal packets and/or xmit_more packets fill up
* tx_desc, it is necessary to trigger NIC tx reg.
*/
ret = NETDEV_TX_BUSY;
goto txkick;
}
spin_unlock_irqrestore(&np->lock, flags);
start_tx = put_tx = np->put_tx.ex;
start_tx_ctx = np->put_tx_ctx;
/* setup the header buffer */
do {
bcnt = (size > NV_TX2_TSO_MAX_SIZE) ? NV_TX2_TSO_MAX_SIZE : size;
np->put_tx_ctx->dma = dma_map_single(&np->pci_dev->dev,
skb->data + offset, bcnt,
DMA_TO_DEVICE);
if (unlikely(dma_mapping_error(&np->pci_dev->dev,
np->put_tx_ctx->dma))) {
/* on DMA mapping error - drop the packet */
dev_kfree_skb_any(skb);
u64_stats_update_begin(&np->swstats_tx_syncp);
nv_txrx_stats_inc(stat_tx_dropped);
u64_stats_update_end(&np->swstats_tx_syncp);
ret = NETDEV_TX_OK;
goto dma_error;
}
np->put_tx_ctx->dma_len = bcnt;
np->put_tx_ctx->dma_single = 1;
put_tx->bufhigh = cpu_to_le32(dma_high(np->put_tx_ctx->dma));
put_tx->buflow = cpu_to_le32(dma_low(np->put_tx_ctx->dma));
put_tx->flaglen = cpu_to_le32((bcnt-1) | tx_flags);
tx_flags = NV_TX2_VALID;
offset += bcnt;
size -= bcnt;
if (unlikely(put_tx++ == np->last_tx.ex))
put_tx = np->tx_ring.ex;
if (unlikely(np->put_tx_ctx++ == np->last_tx_ctx))
np->put_tx_ctx = np->tx_skb;
} while (size);
/* setup the fragments */
for (i = 0; i < fragments; i++) {
skb_frag_t *frag = &skb_shinfo(skb)->frags[i];
u32 frag_size = skb_frag_size(frag);
offset = 0;
do {
bcnt = (frag_size > NV_TX2_TSO_MAX_SIZE) ? NV_TX2_TSO_MAX_SIZE : frag_size;
if (!start_tx_ctx)
start_tx_ctx = tmp_tx_ctx = np->put_tx_ctx;
np->put_tx_ctx->dma = skb_frag_dma_map(
&np->pci_dev->dev,
frag, offset,
bcnt,
DMA_TO_DEVICE);
if (unlikely(dma_mapping_error(&np->pci_dev->dev,
np->put_tx_ctx->dma))) {
/* Unwind the mapped fragments */
do {
nv_unmap_txskb(np, start_tx_ctx);
if (unlikely(tmp_tx_ctx++ == np->last_tx_ctx))
tmp_tx_ctx = np->tx_skb;
} while (tmp_tx_ctx != np->put_tx_ctx);
dev_kfree_skb_any(skb);
np->put_tx_ctx = start_tx_ctx;
u64_stats_update_begin(&np->swstats_tx_syncp);
nv_txrx_stats_inc(stat_tx_dropped);
u64_stats_update_end(&np->swstats_tx_syncp);
ret = NETDEV_TX_OK;
goto dma_error;
}
np->put_tx_ctx->dma_len = bcnt;
np->put_tx_ctx->dma_single = 0;
put_tx->bufhigh = cpu_to_le32(dma_high(np->put_tx_ctx->dma));
put_tx->buflow = cpu_to_le32(dma_low(np->put_tx_ctx->dma));
put_tx->flaglen = cpu_to_le32((bcnt-1) | tx_flags);
offset += bcnt;
frag_size -= bcnt;
if (unlikely(put_tx++ == np->last_tx.ex))
put_tx = np->tx_ring.ex;
if (unlikely(np->put_tx_ctx++ == np->last_tx_ctx))
np->put_tx_ctx = np->tx_skb;
} while (frag_size);
}
if (unlikely(put_tx == np->tx_ring.ex))
prev_tx = np->last_tx.ex;
else
prev_tx = put_tx - 1;
if (unlikely(np->put_tx_ctx == np->tx_skb))
prev_tx_ctx = np->last_tx_ctx;
else
prev_tx_ctx = np->put_tx_ctx - 1;
/* set last fragment flag */
prev_tx->flaglen |= cpu_to_le32(NV_TX2_LASTPACKET);
/* save skb in this slot's context area */
prev_tx_ctx->skb = skb;
if (skb_is_gso(skb))
tx_flags_extra = NV_TX2_TSO | (skb_shinfo(skb)->gso_size << NV_TX2_TSO_SHIFT);
else
tx_flags_extra = skb->ip_summed == CHECKSUM_PARTIAL ?
NV_TX2_CHECKSUM_L3 | NV_TX2_CHECKSUM_L4 : 0;
/* vlan tag */
if (skb_vlan_tag_present(skb))
start_tx->txvlan = cpu_to_le32(NV_TX3_VLAN_TAG_PRESENT |
skb_vlan_tag_get(skb));
else
start_tx->txvlan = 0;
spin_lock_irqsave(&np->lock, flags);
if (np->tx_limit) {
/* Limit the number of outstanding tx. Setup all fragments, but
* do not set the VALID bit on the first descriptor. Save a pointer
* to that descriptor and also for next skb_map element.
*/
if (np->tx_pkts_in_progress == NV_TX_LIMIT_COUNT) {
if (!np->tx_change_owner)
np->tx_change_owner = start_tx_ctx;
/* remove VALID bit */
tx_flags &= ~NV_TX2_VALID;
start_tx_ctx->first_tx_desc = start_tx;
start_tx_ctx->next_tx_ctx = np->put_tx_ctx;
np->tx_end_flip = np->put_tx_ctx;
} else {
np->tx_pkts_in_progress++;
}
}
/* set tx flags */
start_tx->flaglen |= cpu_to_le32(tx_flags | tx_flags_extra);
netdev_sent_queue(np->dev, skb->len);
skb_tx_timestamp(skb);
np->put_tx.ex = put_tx;
spin_unlock_irqrestore(&np->lock, flags);
txkick:
if (netif_queue_stopped(dev) || !netdev_xmit_more()) {
u32 txrxctl_kick;
dma_error:
txrxctl_kick = NVREG_TXRXCTL_KICK | np->txrxctl_bits;
writel(txrxctl_kick, get_hwbase(dev) + NvRegTxRxControl);
}
return ret;
}
static inline void nv_tx_flip_ownership(struct net_device *dev)
{
struct fe_priv *np = netdev_priv(dev);
np->tx_pkts_in_progress--;
if (np->tx_change_owner) {
np->tx_change_owner->first_tx_desc->flaglen |=
cpu_to_le32(NV_TX2_VALID);
np->tx_pkts_in_progress++;
np->tx_change_owner = np->tx_change_owner->next_tx_ctx;
if (np->tx_change_owner == np->tx_end_flip)
np->tx_change_owner = NULL;
writel(NVREG_TXRXCTL_KICK|np->txrxctl_bits, get_hwbase(dev) + NvRegTxRxControl);
}
}
/*
* nv_tx_done: check for completed packets, release the skbs.
*
* Caller must own np->lock.
*/
static int nv_tx_done(struct net_device *dev, int limit)
{
struct fe_priv *np = netdev_priv(dev);
u32 flags;
int tx_work = 0;
struct ring_desc *orig_get_tx = np->get_tx.orig;
unsigned int bytes_compl = 0;
while ((np->get_tx.orig != np->put_tx.orig) &&
!((flags = le32_to_cpu(np->get_tx.orig->flaglen)) & NV_TX_VALID) &&
(tx_work < limit)) {
nv_unmap_txskb(np, np->get_tx_ctx);
if (np->desc_ver == DESC_VER_1) {
if (flags & NV_TX_LASTPACKET) {
if (unlikely(flags & NV_TX_ERROR)) {
if ((flags & NV_TX_RETRYERROR)
&& !(flags & NV_TX_RETRYCOUNT_MASK))
nv_legacybackoff_reseed(dev);
} else {
unsigned int len;
u64_stats_update_begin(&np->swstats_tx_syncp);
nv_txrx_stats_inc(stat_tx_packets);
len = np->get_tx_ctx->skb->len;
nv_txrx_stats_add(stat_tx_bytes, len);
u64_stats_update_end(&np->swstats_tx_syncp);
}
bytes_compl += np->get_tx_ctx->skb->len;
dev_kfree_skb_any(np->get_tx_ctx->skb);
np->get_tx_ctx->skb = NULL;
tx_work++;
}
} else {
if (flags & NV_TX2_LASTPACKET) {
if (unlikely(flags & NV_TX2_ERROR)) {
if ((flags & NV_TX2_RETRYERROR)
&& !(flags & NV_TX2_RETRYCOUNT_MASK))
nv_legacybackoff_reseed(dev);
} else {
unsigned int len;
u64_stats_update_begin(&np->swstats_tx_syncp);
nv_txrx_stats_inc(stat_tx_packets);
len = np->get_tx_ctx->skb->len;
nv_txrx_stats_add(stat_tx_bytes, len);
u64_stats_update_end(&np->swstats_tx_syncp);
}
bytes_compl += np->get_tx_ctx->skb->len;
dev_kfree_skb_any(np->get_tx_ctx->skb);
np->get_tx_ctx->skb = NULL;
tx_work++;
}
}
if (unlikely(np->get_tx.orig++ == np->last_tx.orig))
np->get_tx.orig = np->tx_ring.orig;
if (unlikely(np->get_tx_ctx++ == np->last_tx_ctx))
np->get_tx_ctx = np->tx_skb;
}
netdev_completed_queue(np->dev, tx_work, bytes_compl);
if (unlikely((np->tx_stop == 1) && (np->get_tx.orig != orig_get_tx))) {
np->tx_stop = 0;
netif_wake_queue(dev);
}
return tx_work;
}
static int nv_tx_done_optimized(struct net_device *dev, int limit)
{
struct fe_priv *np = netdev_priv(dev);
u32 flags;
int tx_work = 0;
struct ring_desc_ex *orig_get_tx = np->get_tx.ex;
unsigned long bytes_cleaned = 0;
while ((np->get_tx.ex != np->put_tx.ex) &&
!((flags = le32_to_cpu(np->get_tx.ex->flaglen)) & NV_TX2_VALID) &&
(tx_work < limit)) {
nv_unmap_txskb(np, np->get_tx_ctx);
if (flags & NV_TX2_LASTPACKET) {
if (unlikely(flags & NV_TX2_ERROR)) {
if ((flags & NV_TX2_RETRYERROR)
&& !(flags & NV_TX2_RETRYCOUNT_MASK)) {
if (np->driver_data & DEV_HAS_GEAR_MODE)
nv_gear_backoff_reseed(dev);
else
nv_legacybackoff_reseed(dev);
}
} else {
unsigned int len;
u64_stats_update_begin(&np->swstats_tx_syncp);
nv_txrx_stats_inc(stat_tx_packets);
len = np->get_tx_ctx->skb->len;
nv_txrx_stats_add(stat_tx_bytes, len);
u64_stats_update_end(&np->swstats_tx_syncp);
}
bytes_cleaned += np->get_tx_ctx->skb->len;
dev_kfree_skb_any(np->get_tx_ctx->skb);
np->get_tx_ctx->skb = NULL;
tx_work++;
if (np->tx_limit)
nv_tx_flip_ownership(dev);
}
if (unlikely(np->get_tx.ex++ == np->last_tx.ex))
np->get_tx.ex = np->tx_ring.ex;
if (unlikely(np->get_tx_ctx++ == np->last_tx_ctx))
np->get_tx_ctx = np->tx_skb;
}
netdev_completed_queue(np->dev, tx_work, bytes_cleaned);
if (unlikely((np->tx_stop == 1) && (np->get_tx.ex != orig_get_tx))) {
np->tx_stop = 0;
netif_wake_queue(dev);
}
return tx_work;
}
/*
* nv_tx_timeout: dev->tx_timeout function
* Called with netif_tx_lock held.
*/
static void nv_tx_timeout(struct net_device *dev, unsigned int txqueue)
{
struct fe_priv *np = netdev_priv(dev);
u8 __iomem *base = get_hwbase(dev);
u32 status;
union ring_type put_tx;
int saved_tx_limit;
if (np->msi_flags & NV_MSI_X_ENABLED)
status = readl(base + NvRegMSIXIrqStatus) & NVREG_IRQSTAT_MASK;
else
status = readl(base + NvRegIrqStatus) & NVREG_IRQSTAT_MASK;
netdev_warn(dev, "Got tx_timeout. irq status: %08x\n", status);
if (unlikely(debug_tx_timeout)) {
int i;
netdev_info(dev, "Ring at %lx\n", (unsigned long)np->ring_addr);
netdev_info(dev, "Dumping tx registers\n");
for (i = 0; i <= np->register_size; i += 32) {
netdev_info(dev,
"%3x: %08x %08x %08x %08x "
"%08x %08x %08x %08x\n",
i,
readl(base + i + 0), readl(base + i + 4),
readl(base + i + 8), readl(base + i + 12),
readl(base + i + 16), readl(base + i + 20),
readl(base + i + 24), readl(base + i + 28));
}
netdev_info(dev, "Dumping tx ring\n");
for (i = 0; i < np->tx_ring_size; i += 4) {
if (!nv_optimized(np)) {
netdev_info(dev,
"%03x: %08x %08x // %08x %08x "
"// %08x %08x // %08x %08x\n",
i,
le32_to_cpu(np->tx_ring.orig[i].buf),
le32_to_cpu(np->tx_ring.orig[i].flaglen),
le32_to_cpu(np->tx_ring.orig[i+1].buf),
le32_to_cpu(np->tx_ring.orig[i+1].flaglen),
le32_to_cpu(np->tx_ring.orig[i+2].buf),
le32_to_cpu(np->tx_ring.orig[i+2].flaglen),
le32_to_cpu(np->tx_ring.orig[i+3].buf),
le32_to_cpu(np->tx_ring.orig[i+3].flaglen));
} else {
netdev_info(dev,
"%03x: %08x %08x %08x "
"// %08x %08x %08x "
"// %08x %08x %08x "
"// %08x %08x %08x\n",
i,
le32_to_cpu(np->tx_ring.ex[i].bufhigh),
le32_to_cpu(np->tx_ring.ex[i].buflow),
le32_to_cpu(np->tx_ring.ex[i].flaglen),
le32_to_cpu(np->tx_ring.ex[i+1].bufhigh),
le32_to_cpu(np->tx_ring.ex[i+1].buflow),
le32_to_cpu(np->tx_ring.ex[i+1].flaglen),
le32_to_cpu(np->tx_ring.ex[i+2].bufhigh),
le32_to_cpu(np->tx_ring.ex[i+2].buflow),
le32_to_cpu(np->tx_ring.ex[i+2].flaglen),
le32_to_cpu(np->tx_ring.ex[i+3].bufhigh),
le32_to_cpu(np->tx_ring.ex[i+3].buflow),
le32_to_cpu(np->tx_ring.ex[i+3].flaglen));
}
}
}
spin_lock_irq(&np->lock);
/* 1) stop tx engine */
nv_stop_tx(dev);
/* 2) complete any outstanding tx and do not give HW any limited tx pkts */
saved_tx_limit = np->tx_limit;
np->tx_limit = 0; /* prevent giving HW any limited pkts */
np->tx_stop = 0; /* prevent waking tx queue */
if (!nv_optimized(np))
nv_tx_done(dev, np->tx_ring_size);
else
nv_tx_done_optimized(dev, np->tx_ring_size);
/* save current HW position */
if (np->tx_change_owner)
put_tx.ex = np->tx_change_owner->first_tx_desc;
else
put_tx = np->put_tx;
/* 3) clear all tx state */
nv_drain_tx(dev);
nv_init_tx(dev);
/* 4) restore state to current HW position */
np->get_tx = np->put_tx = put_tx;
np->tx_limit = saved_tx_limit;
/* 5) restart tx engine */
nv_start_tx(dev);
netif_wake_queue(dev);
spin_unlock_irq(&np->lock);
}
/*
* Called when the nic notices a mismatch between the actual data len on the
* wire and the len indicated in the 802 header
*/
static int nv_getlen(struct net_device *dev, void *packet, int datalen)
{
int hdrlen; /* length of the 802 header */
int protolen; /* length as stored in the proto field */
/* 1) calculate len according to header */
if (((struct vlan_ethhdr *)packet)->h_vlan_proto == htons(ETH_P_8021Q)) {
protolen = ntohs(((struct vlan_ethhdr *)packet)->h_vlan_encapsulated_proto);
hdrlen = VLAN_HLEN;
} else {
protolen = ntohs(((struct ethhdr *)packet)->h_proto);
hdrlen = ETH_HLEN;
}
if (protolen > ETH_DATA_LEN)
return datalen; /* Value in proto field not a len, no checks possible */
protolen += hdrlen;
/* consistency checks: */
if (datalen > ETH_ZLEN) {
if (datalen >= protolen) {
/* more data on wire than in 802 header, trim of
* additional data.
*/
return protolen;
} else {
/* less data on wire than mentioned in header.
* Discard the packet.
*/
return -1;
}
} else {
/* short packet. Accept only if 802 values are also short */
if (protolen > ETH_ZLEN) {
return -1;
}
return datalen;
}
}
static void rx_missing_handler(u32 flags, struct fe_priv *np)
{
if (flags & NV_RX_MISSEDFRAME) {
u64_stats_update_begin(&np->swstats_rx_syncp);
nv_txrx_stats_inc(stat_rx_missed_errors);
u64_stats_update_end(&np->swstats_rx_syncp);
}
}
static int nv_rx_process(struct net_device *dev, int limit)
{
struct fe_priv *np = netdev_priv(dev);
u32 flags;
int rx_work = 0;
struct sk_buff *skb;
int len;
while ((np->get_rx.orig != np->put_rx.orig) &&
!((flags = le32_to_cpu(np->get_rx.orig->flaglen)) & NV_RX_AVAIL) &&
(rx_work < limit)) {
/*
* the packet is for us - immediately tear down the pci mapping.
* TODO: check if a prefetch of the first cacheline improves
* the performance.
*/
dma_unmap_single(&np->pci_dev->dev, np->get_rx_ctx->dma,
np->get_rx_ctx->dma_len,
DMA_FROM_DEVICE);
skb = np->get_rx_ctx->skb;
np->get_rx_ctx->skb = NULL;
/* look at what we actually got: */
if (np->desc_ver == DESC_VER_1) {
if (likely(flags & NV_RX_DESCRIPTORVALID)) {
len = flags & LEN_MASK_V1;
if (unlikely(flags & NV_RX_ERROR)) {
if ((flags & NV_RX_ERROR_MASK) == NV_RX_ERROR4) {
len = nv_getlen(dev, skb->data, len);
if (len < 0) {
dev_kfree_skb(skb);
goto next_pkt;
}
}
/* framing errors are soft errors */
else if ((flags & NV_RX_ERROR_MASK) == NV_RX_FRAMINGERR) {
if (flags & NV_RX_SUBTRACT1)
len--;
}
/* the rest a