igbvf: add new driver to support 82576 virtual functions

This adds an igbvf driver to handle virtual functions provided by the
igb driver when SR-IOV has been enabled.  A virtual function is a
lightweight pci-e function that supports a single queue and shares
resources with the 82576 physical function contained within the igb
driver.

To spawn virtual functions from the igb driver all that is needed is to
enable CONFIG_PCI_IOV and have an 82576 Ethernet adapter on a system that
supports SR-IOV in the BIOS.  The virtual functions will appear after the
interface is loaded.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

drivers/net/Kconfig

@@ -2058,6 +2058,27 @@ config IGB_DCA
 	  driver.  DCA is a method for warming the CPU cache before data
 	  is used, with the intent of lessening the impact of cache misses.
 
+config IGBVF
+       tristate "Intel(R) 82576 Virtual Function Ethernet support"
+       depends on PCI
+       ---help---
+         This driver supports Intel(R) 82576 virtual functions.  For more
+         information on how to identify your adapter, go to the Adapter &
+         Driver ID Guide at:
+
+         <http://support.intel.com/support/network/adapter/pro100/21397.htm>
+
+         For general information and support, go to the Intel support
+         website at:
+
+         <http://support.intel.com>
+
+         More specific information on configuring the driver is in
+         <file:Documentation/networking/e1000.txt>.
+
+         To compile this driver as a module, choose M here. The module
+         will be called igbvf.
+
 source "drivers/net/ixp2000/Kconfig"
 
 config MYRI_SBUS

drivers/net/Makefile

@@ -6,6 +6,7 @@ obj-$(CONFIG_E1000) += e1000/
 obj-$(CONFIG_E1000E) += e1000e/
 obj-$(CONFIG_IBM_NEW_EMAC) += ibm_newemac/
 obj-$(CONFIG_IGB) += igb/
+obj-$(CONFIG_IGBVF) += igbvf/
 obj-$(CONFIG_IXGBE) += ixgbe/
 obj-$(CONFIG_IXGB) += ixgb/
 obj-$(CONFIG_IP1000) += ipg.o

drivers/net/igbvf/Makefile

+################################################################################
+#
+# Intel(R) 82576 Virtual Function Linux driver
+# Copyright(c) 2009 Intel Corporation.
+#
+# This program is free software; you can redistribute it and/or modify it
+# under the terms and conditions of the GNU General Public License,
+# version 2, as published by the Free Software Foundation.
+#
+# This program is distributed in the hope 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.,
+# 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA.
+#
+# The full GNU General Public License is included in this distribution in
+# the file called "COPYING".
+#
+# Contact Information:
+# e1000-devel Mailing List <e1000-devel@lists.sourceforge.net>
+# Intel Corporation, 5200 N.E. Elam Young Parkway, Hillsboro, OR 97124-6497
+#
+################################################################################
+
+#
+# Makefile for the Intel(R) 82576 VF ethernet driver
+#
+
+obj-$(CONFIG_IGBVF) += igbvf.o
+
+igbvf-objs := vf.o \
+              mbx.o \
+              ethtool.o \
+              netdev.o
+

drivers/net/igbvf/defines.h

+/*******************************************************************************
+
+  Intel(R) 82576 Virtual Function Linux driver
+  Copyright(c) 1999 - 2009 Intel Corporation.
+
+  This program is free software; you can redistribute it and/or modify it
+  under the terms and conditions of the GNU General Public License,
+  version 2, as published by the Free Software Foundation.
+
+  This program is distributed in the hope 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.,
+  51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA.
+
+  The full GNU General Public License is included in this distribution in
+  the file called "COPYING".
+
+  Contact Information:
+  e1000-devel Mailing List <e1000-devel@lists.sourceforge.net>
+  Intel Corporation, 5200 N.E. Elam Young Parkway, Hillsboro, OR 97124-6497
+
+*******************************************************************************/
+
+#ifndef _E1000_DEFINES_H_
+#define _E1000_DEFINES_H_
+
+/* Number of Transmit and Receive Descriptors must be a multiple of 8 */
+#define REQ_TX_DESCRIPTOR_MULTIPLE  8
+#define REQ_RX_DESCRIPTOR_MULTIPLE  8
+
+/* IVAR valid bit */
+#define E1000_IVAR_VALID        0x80
+
+/* Receive Descriptor bit definitions */
+#define E1000_RXD_STAT_DD       0x01    /* Descriptor Done */
+#define E1000_RXD_STAT_EOP      0x02    /* End of Packet */
+#define E1000_RXD_STAT_IXSM     0x04    /* Ignore checksum */
+#define E1000_RXD_STAT_VP       0x08    /* IEEE VLAN Packet */
+#define E1000_RXD_STAT_UDPCS    0x10    /* UDP xsum calculated */
+#define E1000_RXD_STAT_TCPCS    0x20    /* TCP xsum calculated */
+#define E1000_RXD_STAT_IPCS     0x40    /* IP xsum calculated */
+#define E1000_RXD_ERR_SE        0x02    /* Symbol Error */
+#define E1000_RXD_SPC_VLAN_MASK 0x0FFF  /* VLAN ID is in lower 12 bits */
+
+#define E1000_RXDEXT_STATERR_CE    0x01000000
+#define E1000_RXDEXT_STATERR_SE    0x02000000
+#define E1000_RXDEXT_STATERR_SEQ   0x04000000
+#define E1000_RXDEXT_STATERR_CXE   0x10000000
+#define E1000_RXDEXT_STATERR_TCPE  0x20000000
+#define E1000_RXDEXT_STATERR_IPE   0x40000000
+#define E1000_RXDEXT_STATERR_RXE   0x80000000
+
+
+/* Same mask, but for extended and packet split descriptors */
+#define E1000_RXDEXT_ERR_FRAME_ERR_MASK ( \
+    E1000_RXDEXT_STATERR_CE  |            \
+    E1000_RXDEXT_STATERR_SE  |            \
+    E1000_RXDEXT_STATERR_SEQ |            \
+    E1000_RXDEXT_STATERR_CXE |            \
+    E1000_RXDEXT_STATERR_RXE)
+
+/* Device Control */
+#define E1000_CTRL_RST      0x04000000  /* Global reset */
+
+/* Device Status */
+#define E1000_STATUS_FD         0x00000001      /* Full duplex.0=half,1=full */
+#define E1000_STATUS_LU         0x00000002      /* Link up.0=no,1=link */
+#define E1000_STATUS_TXOFF      0x00000010      /* transmission paused */
+#define E1000_STATUS_SPEED_10   0x00000000      /* Speed 10Mb/s */
+#define E1000_STATUS_SPEED_100  0x00000040      /* Speed 100Mb/s */
+#define E1000_STATUS_SPEED_1000 0x00000080      /* Speed 1000Mb/s */
+
+#define SPEED_10    10
+#define SPEED_100   100
+#define SPEED_1000  1000
+#define HALF_DUPLEX 1
+#define FULL_DUPLEX 2
+
+/* Transmit Descriptor bit definitions */
+#define E1000_TXD_POPTS_IXSM 0x01       /* Insert IP checksum */
+#define E1000_TXD_POPTS_TXSM 0x02       /* Insert TCP/UDP checksum */
+#define E1000_TXD_CMD_DEXT   0x20000000 /* Descriptor extension (0 = legacy) */
+#define E1000_TXD_STAT_DD    0x00000001 /* Descriptor Done */
+
+#define MAX_JUMBO_FRAME_SIZE    0x3F00
+
+/* 802.1q VLAN Packet Size */
+#define VLAN_TAG_SIZE              4    /* 802.3ac tag (not DMA'd) */
+
+/* Error Codes */
+#define E1000_SUCCESS      0
+#define E1000_ERR_CONFIG   3
+#define E1000_ERR_MAC_INIT 5
+#define E1000_ERR_MBX      15
+
+#ifndef ETH_ADDR_LEN
+#define ETH_ADDR_LEN                 6
+#endif
+
+/* SRRCTL bit definitions */
+#define E1000_SRRCTL_BSIZEPKT_SHIFT                     10 /* Shift _right_ */
+#define E1000_SRRCTL_BSIZEHDRSIZE_MASK                  0x00000F00
+#define E1000_SRRCTL_BSIZEHDRSIZE_SHIFT                 2  /* Shift _left_ */
+#define E1000_SRRCTL_DESCTYPE_ADV_ONEBUF                0x02000000
+#define E1000_SRRCTL_DESCTYPE_HDR_SPLIT_ALWAYS          0x0A000000
+#define E1000_SRRCTL_DESCTYPE_MASK                      0x0E000000
+#define E1000_SRRCTL_DROP_EN                            0x80000000
+
+#define E1000_SRRCTL_BSIZEPKT_MASK      0x0000007F
+#define E1000_SRRCTL_BSIZEHDR_MASK      0x00003F00
+
+/* Additional Descriptor Control definitions */
+#define E1000_TXDCTL_QUEUE_ENABLE  0x02000000 /* Enable specific Tx Queue */
+#define E1000_RXDCTL_QUEUE_ENABLE  0x02000000 /* Enable specific Rx Queue */
+
+/* Direct Cache Access (DCA) definitions */
+#define E1000_DCA_TXCTRL_TX_WB_RO_EN (1 << 11) /* Tx Desc writeback RO bit */
+
+#define E1000_VF_INIT_TIMEOUT 200 /* Number of retries to clear RSTI */
+
+#endif /* _E1000_DEFINES_H_ */

drivers/net/igbvf/igbvf.h

+/*******************************************************************************
+
+  Intel(R) 82576 Virtual Function Linux driver
+  Copyright(c) 2009 Intel Corporation.
+
+  This program is free software; you can redistribute it and/or modify it
+  under the terms and conditions of the GNU General Public License,
+  version 2, as published by the Free Software Foundation.
+
+  This program is distributed in the hope 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.,
+  51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA.
+
+  The full GNU General Public License is included in this distribution in
+  the file called "COPYING".
+
+  Contact Information:
+  e1000-devel Mailing List <e1000-devel@lists.sourceforge.net>
+  Intel Corporation, 5200 N.E. Elam Young Parkway, Hillsboro, OR 97124-6497
+
+*******************************************************************************/
+
+/* Linux PRO/1000 Ethernet Driver main header file */
+
+#ifndef _IGBVF_H_
+#define _IGBVF_H_
+
+#include <linux/types.h>
+#include <linux/timer.h>
+#include <linux/io.h>
+#include <linux/netdevice.h>
+
+
+#include "vf.h"
+
+/* Forward declarations */
+struct igbvf_info;
+struct igbvf_adapter;
+
+/* Interrupt defines */
+#define IGBVF_START_ITR                 648 /* ~6000 ints/sec */
+
+/* Interrupt modes, as used by the IntMode paramter */
+#define IGBVF_INT_MODE_LEGACY           0
+#define IGBVF_INT_MODE_MSI              1
+#define IGBVF_INT_MODE_MSIX             2
+
+/* Tx/Rx descriptor defines */
+#define IGBVF_DEFAULT_TXD               256
+#define IGBVF_MAX_TXD                   4096
+#define IGBVF_MIN_TXD                   80
+
+#define IGBVF_DEFAULT_RXD               256
+#define IGBVF_MAX_RXD                   4096
+#define IGBVF_MIN_RXD                   80
+
+#define IGBVF_MIN_ITR_USECS             10 /* 100000 irq/sec */
+#define IGBVF_MAX_ITR_USECS             10000 /* 100    irq/sec */
+
+/* RX descriptor control thresholds.
+ * PTHRESH - MAC will consider prefetch if it has fewer than this number of
+ *           descriptors available in its onboard memory.
+ *           Setting this to 0 disables RX descriptor prefetch.
+ * HTHRESH - MAC will only prefetch if there are at least this many descriptors
+ *           available in host memory.
+ *           If PTHRESH is 0, this should also be 0.
+ * WTHRESH - RX descriptor writeback threshold - MAC will delay writing back
+ *           descriptors until either it has this many to write back, or the
+ *           ITR timer expires.
+ */
+#define IGBVF_RX_PTHRESH                16
+#define IGBVF_RX_HTHRESH                8
+#define IGBVF_RX_WTHRESH                1
+
+/* this is the size past which hardware will drop packets when setting LPE=0 */
+#define MAXIMUM_ETHERNET_VLAN_SIZE      1522
+
+#define IGBVF_FC_PAUSE_TIME             0x0680 /* 858 usec */
+
+/* How many Tx Descriptors do we need to call netif_wake_queue ? */
+#define IGBVF_TX_QUEUE_WAKE             32
+/* How many Rx Buffers do we bundle into one write to the hardware ? */
+#define IGBVF_RX_BUFFER_WRITE           16 /* Must be power of 2 */
+
+#define AUTO_ALL_MODES                  0
+#define IGBVF_EEPROM_APME               0x0400
+
+#define IGBVF_MNG_VLAN_NONE             (-1)
+
+/* Number of packet split data buffers (not including the header buffer) */
+#define PS_PAGE_BUFFERS                 (MAX_PS_BUFFERS - 1)
+
+enum igbvf_boards {
+	board_vf,
+};
+
+struct igbvf_queue_stats {
+	u64 packets;
+	u64 bytes;
+};
+
+/*
+ * wrappers around a pointer to a socket buffer,
+ * so a DMA handle can be stored along with the buffer
+ */
+struct igbvf_buffer {
+	dma_addr_t dma;
+	struct sk_buff *skb;
+	union {
+		/* Tx */
+		struct {
+			unsigned long time_stamp;
+			u16 length;
+			u16 next_to_watch;
+		};
+		/* Rx */
+		struct {
+			struct page *page;
+			u64 page_dma;
+			unsigned int page_offset;
+		};
+	};
+	struct page *page;
+};
+
+union igbvf_desc {
+	union e1000_adv_rx_desc rx_desc;
+	union e1000_adv_tx_desc tx_desc;
+	struct e1000_adv_tx_context_desc tx_context_desc;
+};
+
+struct igbvf_ring {
+	struct igbvf_adapter *adapter;  /* backlink */
+	union igbvf_desc *desc;         /* pointer to ring memory  */
+	dma_addr_t dma;                 /* phys address of ring    */
+	unsigned int size;              /* length of ring in bytes */
+	unsigned int count;             /* number of desc. in ring */
+
+	u16 next_to_use;
+	u16 next_to_clean;
+
+	u16 head;
+	u16 tail;
+
+	/* array of buffer information structs */
+	struct igbvf_buffer *buffer_info;
+	struct napi_struct napi;
+
+	char name[IFNAMSIZ + 5];
+	u32 eims_value;
+	u32 itr_val;
+	u16 itr_register;
+	int set_itr;
+
+	struct sk_buff *rx_skb_top;
+
+	struct igbvf_queue_stats stats;
+};
+
+/* board specific private data structure */
+struct igbvf_adapter {
+	struct timer_list watchdog_timer;
+	struct timer_list blink_timer;
+
+	struct work_struct reset_task;
+	struct work_struct watchdog_task;
+
+	const struct igbvf_info *ei;
+
+	struct vlan_group *vlgrp;
+	u32 bd_number;
+	u32 rx_buffer_len;
+	u32 polling_interval;
+	u16 mng_vlan_id;
+	u16 link_speed;
+	u16 link_duplex;
+
+	spinlock_t tx_queue_lock; /* prevent concurrent tail updates */
+
+	/* track device up/down/testing state */
+	unsigned long state;
+
+	/* Interrupt Throttle Rate */
+	u32 itr;
+	u32 itr_setting;
+	u16 tx_itr;
+	u16 rx_itr;
+
+	/*
+	 * Tx
+	 */
+	struct igbvf_ring *tx_ring /* One per active queue */
+	____cacheline_aligned_in_smp;
+
+	unsigned long tx_queue_len;
+	unsigned int restart_queue;
+	u32 txd_cmd;
+
+	bool detect_tx_hung;
+	u8 tx_timeout_factor;
+
+	u32 tx_int_delay;
+	u32 tx_abs_int_delay;
+
+	unsigned int total_tx_bytes;
+	unsigned int total_tx_packets;
+	unsigned int total_rx_bytes;
+	unsigned int total_rx_packets;
+
+	/* Tx stats */
+	u32 tx_timeout_count;
+	u32 tx_fifo_head;
+	u32 tx_head_addr;
+	u32 tx_fifo_size;
+	u32 tx_dma_failed;
+
+	/*
+	 * Rx
+	 */
+	struct igbvf_ring *rx_ring;
+
+	u32 rx_int_delay;
+	u32 rx_abs_int_delay;
+
+	/* Rx stats */
+	u64 hw_csum_err;
+	u64 hw_csum_good;
+	u64 rx_hdr_split;
+	u32 alloc_rx_buff_failed;
+	u32 rx_dma_failed;
+
+	unsigned int rx_ps_hdr_size;
+	u32 max_frame_size;
+	u32 min_frame_size;
+
+	/* OS defined structs */
+	struct net_device *netdev;
+	struct pci_dev *pdev;
+	struct net_device_stats net_stats;
+	spinlock_t stats_lock;      /* prevent concurrent stats updates */
+
+	/* structs defined in e1000_hw.h */
+	struct e1000_hw hw;
+
+	/* The VF counters don't clear on read so we have to get a base
+	 * count on driver start up and always subtract that base on
+	 * on the first update, thus the flag..
+	 */
+	struct e1000_vf_stats stats;
+	u64 zero_base;
+
+	struct igbvf_ring test_tx_ring;
+	struct igbvf_ring test_rx_ring;
+	u32 test_icr;
+
+	u32 msg_enable;
+	struct msix_entry *msix_entries;
+	int int_mode;
+	u32 eims_enable_mask;
+	u32 eims_other;
+	u32 int_counter0;
+	u32 int_counter1;
+
+	u32 eeprom_wol;
+	u32 wol;
+	u32 pba;
+
+	bool fc_autoneg;
+
+	unsigned long led_status;
+
+	unsigned int flags;
+};
+
+struct igbvf_info {
+	enum e1000_mac_type     mac;
+	unsigned int            flags;
+	u32                     pba;
+	void                    (*init_ops)(struct e1000_hw *);
+	s32                     (*get_variants)(struct igbvf_adapter *);
+};
+
+/* hardware capability, feature, and workaround flags */
+#define FLAG_HAS_HW_VLAN_FILTER           (1 << 0)
+#define FLAG_HAS_JUMBO_FRAMES             (1 << 1)
+#define FLAG_MSI_ENABLED                  (1 << 2)
+#define FLAG_RX_CSUM_ENABLED              (1 << 3)
+#define FLAG_TSO_FORCE                    (1 << 4)
+
+#define IGBVF_RX_DESC_ADV(R, i)     \
+	(&((((R).desc))[i].rx_desc))
+#define IGBVF_TX_DESC_ADV(R, i)     \
+	(&((((R).desc))[i].tx_desc))
+#define IGBVF_TX_CTXTDESC_ADV(R, i) \
+	(&((((R).desc))[i].tx_context_desc))
+
+enum igbvf_state_t {
+	__IGBVF_TESTING,
+	__IGBVF_RESETTING,
+	__IGBVF_DOWN
+};
+
+enum latency_range {
+	lowest_latency = 0,
+	low_latency = 1,
+	bulk_latency = 2,
+	latency_invalid = 255
+};
+
+extern char igbvf_driver_name[];
+extern const char igbvf_driver_version[];
+
+extern void igbvf_check_options(struct igbvf_adapter *);
+extern void igbvf_set_ethtool_ops(struct net_device *);
+
+extern int igbvf_up(struct igbvf_adapter *);
+extern void igbvf_down(struct igbvf_adapter *);
+extern void igbvf_reinit_locked(struct igbvf_adapter *);
+extern void igbvf_reset(struct igbvf_adapter *);
+extern int igbvf_setup_rx_resources(struct igbvf_adapter *, struct igbvf_ring *);
+extern int igbvf_setup_tx_resources(struct igbvf_adapter *, struct igbvf_ring *);
+extern void igbvf_free_rx_resources(struct igbvf_ring *);
+extern void igbvf_free_tx_resources(struct igbvf_ring *);
+extern void igbvf_update_stats(struct igbvf_adapter *);
+extern void igbvf_set_interrupt_capability(struct igbvf_adapter *);
+extern void igbvf_reset_interrupt_capability(struct igbvf_adapter *);
+
+extern unsigned int copybreak;
+
+#endif /* _IGBVF_H_ */

drivers/net/igbvf/mbx.c

+/*******************************************************************************
+
+  Intel(R) 82576 Virtual Function Linux driver
+  Copyright(c) 2009 Intel Corporation.
+
+  This program is free software; you can redistribute it and/or modify it
+  under the terms and conditions of the GNU General Public License,
+  version 2, as published by the Free Software Foundation.
+
+  This program is distributed in the hope 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.,
+  51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA.
+
+  The full GNU General Public License is included in this distribution in
+  the file called "COPYING".
+
+  Contact Information:
+  e1000-devel Mailing List <e1000-devel@lists.sourceforge.net>
+  Intel Corporation, 5200 N.E. Elam Young Parkway, Hillsboro, OR 97124-6497
+
+*******************************************************************************/
+
+#include "mbx.h"
+
+/**
+ *  e1000_poll_for_msg - Wait for message notification
+ *  @hw: pointer to the HW structure
+ *
+ *  returns SUCCESS if it successfully received a message notification
+ **/
+static s32 e1000_poll_for_msg(struct e1000_hw *hw)
+{
+	struct e1000_mbx_info *mbx = &hw->mbx;
+	int countdown = mbx->timeout;
+
+	if (!mbx->ops.check_for_msg)
+		goto out;
+
+	while (countdown && mbx->ops.check_for_msg(hw)) {
+		countdown--;
+		udelay(mbx->usec_delay);
+	}
+
+	/* if we failed, all future posted messages fail until reset */
+	if (!countdown)
+		mbx->timeout = 0;
+out:
+	return countdown ? E1000_SUCCESS : -E1000_ERR_MBX;
+}
+
+/**
+ *  e1000_poll_for_ack - Wait for message acknowledgement
+ *  @hw: pointer to the HW structure
+ *
+ *  returns SUCCESS if it successfully received a message acknowledgement
+ **/
+static s32 e1000_poll_for_ack(struct e1000_hw *hw)
+{
+	struct e1000_mbx_info *mbx = &hw->mbx;
+	int countdown = mbx->timeout;
+
+	if (!mbx->ops.check_for_ack)
+		goto out;
+
+	while (countdown && mbx->ops.check_for_ack(hw)) {
+		countdown--;
+		udelay(mbx->usec_delay);
+	}
+
+	/* if we failed, all future posted messages fail until reset */
+	if (!countdown)
+		mbx->timeout = 0;
+out:
+	return countdown ? E1000_SUCCESS : -E1000_ERR_MBX;
+}
+
+/**
+ *  e1000_read_posted_mbx - Wait for message notification and receive message
+ *  @hw: pointer to the HW structure
+ *  @msg: The message buffer
+ *  @size: Length of buffer
+ *
+ *  returns SUCCESS if it successfully received a message notification and
+ *  copied it into the receive buffer.
+ **/
+static s32 e1000_read_posted_mbx(struct e1000_hw *hw, u32 *msg, u16 size)
+{
+	struct e1000_mbx_info *mbx = &hw->mbx;
+	s32 ret_val = -E1000_ERR_MBX;
+
+	if (!mbx->ops.read)
+		goto out;
+
+	ret_val = e1000_poll_for_msg(hw);
+
+	/* if ack received read message, otherwise we timed out */
+	if (!ret_val)
+		ret_val = mbx->ops.read(hw, msg, size);
+out:
+	return ret_val;
+}
+
+/**
+ *  e1000_write_posted_mbx - Write a message to the mailbox, wait for ack
+ *  @hw: pointer to the HW structure
+ *  @msg: The message buffer
+ *  @size: Length of buffer
+ *
+ *  returns SUCCESS if it successfully copied message into the buffer and
+ *  received an ack to that message within delay * timeout period
+ **/
+static s32 e1000_write_posted_mbx(struct e1000_hw *hw, u32 *msg, u16 size)
+{
+	struct e1000_mbx_info *mbx = &hw->mbx;
+	s32 ret_val = -E1000_ERR_MBX;
+
+	/* exit if we either can't write or there isn't a defined timeout */
+	if (!mbx->ops.write || !mbx->timeout)
+		goto out;
+
+	/* send msg*/
+	ret_val = mbx->ops.write(hw, msg, size);
+
+	/* if msg sent wait until we receive an ack */
+	if (!ret_val)
+		ret_val = e1000_poll_for_ack(hw);
+out:
+	return ret_val;
+}
+
+/**
+ *  e1000_read_v2p_mailbox - read v2p mailbox
+ *  @hw: pointer to the HW structure
+ *
+ *  This function is used to read the v2p mailbox without losing the read to
+ *  clear status bits.
+ **/
+static u32 e1000_read_v2p_mailbox(struct e1000_hw *hw)
+{
+	u32 v2p_mailbox = er32(V2PMAILBOX(0));
+
+	v2p_mailbox |= hw->dev_spec.vf.v2p_mailbox;
+	hw->dev_spec.vf.v2p_mailbox |= v2p_mailbox & E1000_V2PMAILBOX_R2C_BITS;
+
+	return v2p_mailbox;
+}
+
+/**
+ *  e1000_check_for_bit_vf - Determine if a status bit was set
+ *  @hw: pointer to the HW structure
+ *  @mask: bitmask for bits to be tested and cleared
+ *
+ *  This function is used to check for the read to clear bits within
+ *  the V2P mailbox.
+ **/
+static s32 e1000_check_for_bit_vf(struct e1000_hw *hw, u32 mask)
+{
+	u32 v2p_mailbox = e1000_read_v2p_mailbox(hw);
+	s32 ret_val = -E1000_ERR_MBX;
+
+	if (v2p_mailbox & mask)
+		ret_val = E1000_SUCCESS;
+
+	hw->dev_spec.vf.v2p_mailbox &= ~mask;
+
+	return ret_val;
+}
+
+/**
+ *  e1000_check_for_msg_vf - checks to see if the PF has sent mail
+ *  @hw: pointer to the HW structure
+ *
+ *  returns SUCCESS if the PF has set the Status bit or else ERR_MBX
+ **/
+static s32 e1000_check_for_msg_vf(struct e1000_hw *hw)
+{
+	s32 ret_val = -E1000_ERR_MBX;
+
+	if (!e1000_check_for_bit_vf(hw, E1000_V2PMAILBOX_PFSTS)) {
+		ret_val = E1000_SUCCESS;
+		hw->mbx.stats.reqs++;
+	}
+
+	return ret_val;
+}
+
+/**
+ *  e1000_check_for_ack_vf - checks to see if the PF has ACK'd
+ *  @hw: pointer to the HW structure
+ *
+ *  returns SUCCESS if the PF has set the ACK bit or else ERR_MBX
+ **/
+static s32 e1000_check_for_ack_vf(struct e1000_hw *hw)
+{
+	s32 ret_val = -E1000_ERR_MBX;
+
+	if (!e1000_check_for_bit_vf(hw, E1000_V2PMAILBOX_PFACK)) {
+		ret_val = E1000_SUCCESS;
+		hw->mbx.stats.acks++;
+	}
+
+	return ret_val;
+}
+
+/**
+ *  e1000_check_for_rst_vf - checks to see if the PF has reset
+ *  @hw: pointer to the HW structure
+ *
+ *  returns true if the PF has set the reset done bit or else false
+ **/
+static s32 e1000_check_for_rst_vf(struct e1000_hw *hw)
+{
+	s32 ret_val = -E1000_ERR_MBX;
+
+	if (!e1000_check_for_bit_vf(hw, (E1000_V2PMAILBOX_RSTD |
+	                                 E1000_V2PMAILBOX_RSTI))) {
+		ret_val = E1000_SUCCESS;
+		hw->mbx.stats.rsts++;
+	}
+
+	return ret_val;
+}
+
+/**
+ *  e1000_obtain_mbx_lock_vf - obtain mailbox lock
+ *  @hw: pointer to the HW structure
+ *
+ *  return SUCCESS if we obtained the mailbox lock
+ **/
+static s32 e1000_obtain_mbx_lock_vf(struct e1000_hw *hw)
+{
+	s32 ret_val = -E1000_ERR_MBX;
+
+	/* Take ownership of the buffer */
+	ew32(V2PMAILBOX(0), E1000_V2PMAILBOX_VFU);
+
+	/* reserve mailbox for vf use */
+	if (e1000_read_v2p_mailbox(hw) & E1000_V2PMAILBOX_VFU)
+		ret_val = E1000_SUCCESS;
+
+	return ret_val;
+}
+
+/**
+ *  e1000_write_mbx_vf - Write a message to the mailbox
+ *  @hw: pointer to the HW structure
+ *  @msg: The message buffer
+ *  @size: Length of buffer
+ *
+ *  returns SUCCESS if it successfully copied message into the buffer
+ **/
+static s32 e1000_write_mbx_vf(struct e1000_hw *hw, u32 *msg, u16 size)
+{
+	s32 err;
+	u16 i;
+
+	/* lock the mailbox to prevent pf/vf race condition */
+	err = e1000_obtain_mbx_lock_vf(hw);
+	if (err)
+		goto out_no_write;
+
+	/* flush any ack or msg as we are going to overwrite mailbox */
+	e1000_check_for_ack_vf(hw);
+	e1000_check_for_msg_vf(hw);
+
+	/* copy the caller specified message to the mailbox memory buffer */
+	for (i = 0; i < size; i++)
+		array_ew32(VMBMEM(0), i, msg[i]);
+
+	/* update stats */
+	hw->mbx.stats.msgs_tx++;
+
+	/* Drop VFU and interrupt the PF to tell it a message has been sent */
+	ew32(V2PMAILBOX(0), E1000_V2PMAILBOX_REQ);
+
+out_no_write:
+	return err;
+}
+
+/**
+ *  e1000_read_mbx_vf - Reads a message from the inbox intended for vf
+ *  @hw: pointer to the HW structure
+ *  @msg: The message buffer
+ *  @size: Length of buffer
+ *
+ *  returns SUCCESS if it successfuly read message from buffer
+ **/
+static s32 e1000_read_mbx_vf(struct e1000_hw *hw, u32 *msg, u16 size)
+{
+	s32 err;
+	u16 i;
+
+	/* lock the mailbox to prevent pf/vf race condition */
+	err = e1000_obtain_mbx_lock_vf(hw);
+	if (err)
+		goto out_no_read;
+
+	/* copy the message from the mailbox memory buffer */
+	for (i = 0; i < size; i++)
+		msg[i] = array_er32(VMBMEM(0), i);
+
+	/* Acknowledge receipt and release mailbox, then we're done */
+	ew32(V2PMAILBOX(0), E1000_V2PMAILBOX_ACK);
+
+	/* update stats */
+	hw->mbx.stats.msgs_rx++;
+
+out_no_read:
+	return err;
+}
+
+/**
+ *  e1000_init_mbx_params_vf - set initial values for vf mailbox
+ *  @hw: pointer to the HW structure
+ *
+ *  Initializes the hw->mbx struct to correct values for vf mailbox
+ */
+s32 e1000_init_mbx_params_vf(struct e1000_hw *hw)
+{
+	struct e1000_mbx_info *mbx = &hw->mbx;
+
+	/* start mailbox as timed out and let the reset_hw call set the timeout
+	 * value to being communications */
+	mbx->timeout = 0;
+	mbx->usec_delay = E1000_VF_MBX_INIT_DELAY;
+
+	mbx->size = E1000_VFMAILBOX_SIZE;
+
+	mbx->ops.read = e1000_read_mbx_vf;
+	mbx->ops.write = e1000_write_mbx_vf;
+	mbx->ops.read_posted = e1000_read_posted_mbx;
+	mbx->ops.write_posted = e1000_write_posted_mbx;
+	mbx->ops.check_for_msg = e1000_check_for_msg_vf;
+	mbx->ops.check_for_ack = e1000_check_for_ack_vf;
+	mbx->ops.check_for_rst = e1000_check_for_rst_vf;
+
+	mbx->stats.msgs_tx = 0;
+	mbx->stats.msgs_rx = 0;
+	mbx->stats.reqs = 0;
+	mbx->stats.acks = 0;
+	mbx->stats.rsts = 0;
+
+	return E1000_SUCCESS;
+}
+

drivers/net/igbvf/mbx.h

+/*******************************************************************************
+
+  Intel(R) 82576 Virtual Function Linux driver
+  Copyright(c) 1999 - 2009 Intel Corporation.
+
+  This program is free software; you can redistribute it and/or modify it
+  under the terms and conditions of the GNU General Public License,
+  version 2, as published by the Free Software Foundation.
+
+  This program is distributed in the hope 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.,
+  51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA.
+
+  The full GNU General Public License is included in this distribution in
+  the file called "COPYING".
+
+  Contact Information:
+  e1000-devel Mailing List <e1000-devel@lists.sourceforge.net>
+  Intel Corporation, 5200 N.E. Elam Young Parkway, Hillsboro, OR 97124-6497
+
+*******************************************************************************/
+
+#ifndef _E1000_MBX_H_
+#define _E1000_MBX_H_
+
+#include "vf.h"
+
+#define E1000_V2PMAILBOX_REQ   0x00000001 /* Request for PF Ready bit */
+#define E1000_V2PMAILBOX_ACK   0x00000002 /* Ack PF message received */
+#define E1000_V2PMAILBOX_VFU   0x00000004 /* VF owns the mailbox buffer */
+#define E1000_V2PMAILBOX_PFU   0x00000008 /* PF owns the mailbox buffer */
+#define E1000_V2PMAILBOX_PFSTS 0x00000010 /* PF wrote a message in the MB */
+#define E1000_V2PMAILBOX_PFACK 0x00000020 /* PF ack the previous VF msg */
+#define E1000_V2PMAILBOX_RSTI  0x00000040 /* PF has reset indication */
+#define E1000_V2PMAILBOX_RSTD  0x00000080 /* PF has indicated reset done */
+#define E1000_V2PMAILBOX_R2C_BITS 0x000000B0 /* All read to clear bits */
+
+#define E1000_VFMAILBOX_SIZE   16 /* 16 32 bit words - 64 bytes */
+
+/* If it's a E1000_VF_* msg then it originates in the VF and is sent to the
+ * PF.  The reverse is true if it is E1000_PF_*.
+ * Message ACK's are the value or'd with 0xF0000000
+ */
+#define E1000_VT_MSGTYPE_ACK      0x80000000  /* Messages below or'd with
+                                               * this are the ACK */
+#define E1000_VT_MSGTYPE_NACK     0x40000000  /* Messages below or'd with
+                                               * this are the NACK */
+#define E1000_VT_MSGTYPE_CTS      0x20000000  /* Indicates that VF is still
+                                                 clear to send requests */
+
+/* We have a total wait time of 1s for vf mailbox posted messages */
+#define E1000_VF_MBX_INIT_TIMEOUT 2000 /* retry count for mailbox timeout */
+#define E1000_VF_MBX_INIT_DELAY   500  /* usec delay between retries */
+
+#define E1000_VT_MSGINFO_SHIFT    16
+/* bits 23:16 are used for exra info for certain messages */
+#define E1000_VT_MSGINFO_MASK     (0xFF << E1000_VT_MSGINFO_SHIFT)
+
+#define E1000_VF_RESET            0x01 /* VF requests reset */
+#define E1000_VF_SET_MAC_ADDR     0x02 /* VF requests PF to set MAC addr */
+#define E1000_VF_SET_MULTICAST    0x03 /* VF requests PF to set MC addr */
+#define E1000_VF_SET_VLAN         0x04 /* VF requests PF to set VLAN */
+#define E1000_VF_SET_LPE          0x05 /* VF requests PF to set VMOLR.LPE */
+
+#define E1000_PF_CONTROL_MSG      0x0100 /* PF control message */
+
+void e1000_init_mbx_ops_generic(struct e1000_hw *hw);
+s32 e1000_init_mbx_params_vf(struct e1000_hw *);
+
+#endif /* _E1000_MBX_H_ */

drivers/net/igbvf/regs.h

+/*******************************************************************************
+
+  Intel(R) 82576 Virtual Function Linux driver
+  Copyright(c) 2009 Intel Corporation.
+
+  This program is free software; you can redistribute it and/or modify it
+  under the terms and conditions of the GNU General Public License,
+  version 2, as published by the Free Software Foundation.
+
+  This program is distributed in the hope 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.,
+  51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA.
+
+  The full GNU General Public License is included in this distribution in
+  the file called "COPYING".
+
+  Contact Information:
+  e1000-devel Mailing List <e1000-devel@lists.sourceforge.net>
+  Intel Corporation, 5200 N.E. Elam Young Parkway, Hillsboro, OR 97124-6497
+
+*******************************************************************************/
+
+#ifndef _E1000_REGS_H_
+#define _E1000_REGS_H_
+
+#define E1000_CTRL      0x00000 /* Device Control - RW */
+#define E1000_STATUS    0x00008 /* Device Status - RO */
+#define E1000_ITR       0x000C4 /* Interrupt Throttling Rate - RW */
+#define E1000_EICR      0x01580 /* Ext. Interrupt Cause Read - R/clr */
+#define E1000_EITR(_n)  (0x01680 + (0x4 * (_n)))
+#define E1000_EICS      0x01520 /* Ext. Interrupt Cause Set - W0 */
+#define E1000_EIMS      0x01524 /* Ext. Interrupt Mask Set/Read - RW */
+#define E1000_EIMC      0x01528 /* Ext. Interrupt Mask Clear - WO */
+#define E1000_EIAC      0x0152C /* Ext. Interrupt Auto Clear - RW */
+#define E1000_EIAM      0x01530 /* Ext. Interrupt Ack Auto Clear Mask - RW */
+#define E1000_IVAR0     0x01700 /* Interrupt Vector Allocation (array) - RW */
+#define E1000_IVAR_MISC 0x01740 /* IVAR for "other" causes - RW */
+/*
+ * Convenience macros
+ *
+ * Note: "_n" is the queue number of the register to be written to.
+ *
+ * Example usage:
+ * E1000_RDBAL_REG(current_rx_queue)
+ */
+#define E1000_RDBAL(_n)      ((_n) < 4 ? (0x02800 + ((_n) * 0x100)) : \
+                                         (0x0C000 + ((_n) * 0x40)))
+#define E1000_RDBAH(_n)      ((_n) < 4 ? (0x02804 + ((_n) * 0x100)) : \
+                                         (0x0C004 + ((_n) * 0x40)))
+#define E1000_RDLEN(_n)      ((_n) < 4 ? (0x02808 + ((_n) * 0x100)) : \
+                                         (0x0C008 + ((_n) * 0x40)))
+#define E1000_SRRCTL(_n)     ((_n) < 4 ? (0x0280C + ((_n) * 0x100)) : \
+                                         (0x0C00C + ((_n) * 0x40)))
+#define E1000_RDH(_n)        ((_n) < 4 ? (0x02810 + ((_n) * 0x100)) : \
+                                         (0x0C010 + ((_n) * 0x40)))
+#define E1000_RDT(_n)        ((_n) < 4 ? (0x02818 + ((_n) * 0x100)) : \
+                                         (0x0C018 + ((_n) * 0x40)))
+#define E1000_RXDCTL(_n)     ((_n) < 4 ? (0x02828 + ((_n) * 0x100)) : \
+                                         (0x0C028 + ((_n) * 0x40)))
+#define E1000_TDBAL(_n)      ((_n) < 4 ? (0x03800 + ((_n) * 0x100)) : \
+                                         (0x0E000 + ((_n) * 0x40)))
+#define E1000_TDBAH(_n)      ((_n) < 4 ? (0x03804 + ((_n) * 0x100)) : \
+                                         (0x0E004 + ((_n) * 0x40)))
+#define E1000_TDLEN(_n)      ((_n) < 4 ? (0x03808 + ((_n) * 0x100)) : \
+                                         (0x0E008 + ((_n) * 0x40)))
+#define E1000_TDH(_n)        ((_n) < 4 ? (0x03810 + ((_n) * 0x100)) : \
+                                         (0x0E010 + ((_n) * 0x40)))
+#define E1000_TDT(_n)        ((_n) < 4 ? (0x03818 + ((_n) * 0x100)) : \
+                                         (0x0E018 + ((_n) * 0x40)))
+#define E1000_TXDCTL(_n)     ((_n) < 4 ? (0x03828 + ((_n) * 0x100)) : \
+                                         (0x0E028 + ((_n) * 0x40)))
+#define E1000_DCA_TXCTRL(_n) (0x03814 + (_n << 8))
+#define E1000_DCA_RXCTRL(_n) (0x02814 + (_n << 8))
+#define E1000_RAL(_i)  (((_i) <= 15) ? (0x05400 + ((_i) * 8)) : \
+                                       (0x054E0 + ((_i - 16) * 8)))
+#define E1000_RAH(_i)  (((_i) <= 15) ? (0x05404 + ((_i) * 8)) : \
+                                       (0x054E4 + ((_i - 16) * 8)))
+
+/* Statistics registers */
+#define E1000_VFGPRC    0x00F10
+#define E1000_VFGORC    0x00F18
+#define E1000_VFMPRC    0x00F3C
+#define E1000_VFGPTC    0x00F14
+#define E1000_VFGOTC    0x00F34
+#define E1000_VFGOTLBC  0x00F50
+#define E1000_VFGPTLBC  0x00F44
+#define E1000_VFGORLBC  0x00F48
+#define E1000_VFGPRLBC  0x00F40
+
+/* These act per VF so an array friendly macro is used */
+#define E1000_V2PMAILBOX(_n)   (0x00C40 + (4 * (_n)))
+#define E1000_VMBMEM(_n)       (0x00800 + (64 * (_n)))
+
+/* Define macros for handling registers */
+#define er32(reg) readl(hw->hw_addr + E1000_##reg)
+#define ew32(reg, val) writel((val), hw->hw_addr +  E1000_##reg)
+#define array_er32(reg, offset) \
+	readl(hw->hw_addr + E1000_##reg + (offset << 2))
+#define array_ew32(reg, offset, val) \
+	writel((val), hw->hw_addr +  E1000_##reg + (offset << 2))
+#define e1e_flush() er32(STATUS)
+
+#endif