Merge branch 'topic/core-cleanup' into for-linus
[safe/jmp/linux-2.6] / drivers / net / sfc / falcon.c
index ff15b9d..08278e7 100644 (file)
@@ -1,7 +1,7 @@
 /****************************************************************************
  * Driver for Solarflare Solarstorm network controllers and boards
  * Copyright 2005-2006 Fen Systems Ltd.
- * Copyright 2006-2008 Solarflare Communications Inc.
+ * Copyright 2006-2009 Solarflare Communications Inc.
  *
  * This program is free software; you can redistribute it and/or modify it
  * under the terms of the GNU General Public License version 2 as published
 #include <linux/seq_file.h>
 #include <linux/i2c.h>
 #include <linux/mii.h>
+#include <linux/slab.h>
 #include "net_driver.h"
 #include "bitfield.h"
 #include "efx.h"
 #include "mac.h"
 #include "spi.h"
-#include "falcon.h"
+#include "nic.h"
 #include "regs.h"
 #include "io.h"
 #include "mdio_10g.h"
 #include "phy.h"
 #include "workarounds.h"
 
-/* Falcon hardware control.
- * Falcon is the internal codename for the SFC4000 controller that is
- * present in SFE400X evaluation boards
- */
-
-/**************************************************************************
- *
- * Configurable values
- *
- **************************************************************************
- */
-
-static int disable_dma_stats;
-
-/* This is set to 16 for a good reason.  In summary, if larger than
- * 16, the descriptor cache holds more than a default socket
- * buffer's worth of packets (for UDP we can only have at most one
- * socket buffer's worth outstanding).  This combined with the fact
- * that we only get 1 TX event per descriptor cache means the NIC
- * goes idle.
- */
-#define TX_DC_ENTRIES 16
-#define TX_DC_ENTRIES_ORDER 0
-#define TX_DC_BASE 0x130000
-
-#define RX_DC_ENTRIES 64
-#define RX_DC_ENTRIES_ORDER 2
-#define RX_DC_BASE 0x100000
+/* Hardware control for SFC4000 (aka Falcon). */
 
 static const unsigned int
 /* "Large" EEPROM device: Atmel AT25640 or similar
@@ -70,90 +44,6 @@ default_flash_type = ((17 << SPI_DEV_TYPE_SIZE_LBN)
                      | (15 << SPI_DEV_TYPE_ERASE_SIZE_LBN)
                      | (8 << SPI_DEV_TYPE_BLOCK_SIZE_LBN));
 
-/* RX FIFO XOFF watermark
- *
- * When the amount of the RX FIFO increases used increases past this
- * watermark send XOFF. Only used if RX flow control is enabled (ethtool -A)
- * This also has an effect on RX/TX arbitration
- */
-static int rx_xoff_thresh_bytes = -1;
-module_param(rx_xoff_thresh_bytes, int, 0644);
-MODULE_PARM_DESC(rx_xoff_thresh_bytes, "RX fifo XOFF threshold");
-
-/* RX FIFO XON watermark
- *
- * When the amount of the RX FIFO used decreases below this
- * watermark send XON. Only used if TX flow control is enabled (ethtool -A)
- * This also has an effect on RX/TX arbitration
- */
-static int rx_xon_thresh_bytes = -1;
-module_param(rx_xon_thresh_bytes, int, 0644);
-MODULE_PARM_DESC(rx_xon_thresh_bytes, "RX fifo XON threshold");
-
-/* If FALCON_MAX_INT_ERRORS internal errors occur within
- * FALCON_INT_ERROR_EXPIRE seconds, we consider the NIC broken and
- * disable it.
- */
-#define FALCON_INT_ERROR_EXPIRE 3600
-#define FALCON_MAX_INT_ERRORS 5
-
-/* We poll for events every FLUSH_INTERVAL ms, and check FLUSH_POLL_COUNT times
- */
-#define FALCON_FLUSH_INTERVAL 10
-#define FALCON_FLUSH_POLL_COUNT 100
-
-/**************************************************************************
- *
- * Falcon constants
- *
- **************************************************************************
- */
-
-/* Size and alignment of special buffers (4KB) */
-#define FALCON_BUF_SIZE 4096
-
-/* Dummy SRAM size code */
-#define SRM_NB_BSZ_ONCHIP_ONLY (-1)
-
-#define FALCON_IS_DUAL_FUNC(efx)               \
-       (falcon_rev(efx) < FALCON_REV_B0)
-
-/**************************************************************************
- *
- * Falcon hardware access
- *
- **************************************************************************/
-
-static inline void falcon_write_buf_tbl(struct efx_nic *efx, efx_qword_t *value,
-                                       unsigned int index)
-{
-       efx_sram_writeq(efx, efx->membase + efx->type->buf_tbl_base,
-                       value, index);
-}
-
-/* Read the current event from the event queue */
-static inline efx_qword_t *falcon_event(struct efx_channel *channel,
-                                       unsigned int index)
-{
-       return (((efx_qword_t *) (channel->eventq.addr)) + index);
-}
-
-/* See if an event is present
- *
- * We check both the high and low dword of the event for all ones.  We
- * wrote all ones when we cleared the event, and no valid event can
- * have all ones in either its high or low dwords.  This approach is
- * robust against reordering.
- *
- * Note that using a single 64-bit comparison is incorrect; even
- * though the CPU read will be atomic, the DMA write may not be.
- */
-static inline int falcon_event_present(efx_qword_t *event)
-{
-       return (!(EFX_DWORD_IS_ALL_ONES(event->dword[0]) |
-                 EFX_DWORD_IS_ALL_ONES(event->dword[1])));
-}
-
 /**************************************************************************
  *
  * I2C bus - this is a bit-bashing interface using GPIO pins
@@ -210,843 +100,7 @@ static struct i2c_algo_bit_data falcon_i2c_bit_operations = {
        .timeout        = DIV_ROUND_UP(HZ, 20),
 };
 
-/**************************************************************************
- *
- * Falcon special buffer handling
- * Special buffers are used for event queues and the TX and RX
- * descriptor rings.
- *
- *************************************************************************/
-
-/*
- * Initialise a Falcon special buffer
- *
- * This will define a buffer (previously allocated via
- * falcon_alloc_special_buffer()) in Falcon's buffer table, allowing
- * it to be used for event queues, descriptor rings etc.
- */
-static void
-falcon_init_special_buffer(struct efx_nic *efx,
-                          struct efx_special_buffer *buffer)
-{
-       efx_qword_t buf_desc;
-       int index;
-       dma_addr_t dma_addr;
-       int i;
-
-       EFX_BUG_ON_PARANOID(!buffer->addr);
-
-       /* Write buffer descriptors to NIC */
-       for (i = 0; i < buffer->entries; i++) {
-               index = buffer->index + i;
-               dma_addr = buffer->dma_addr + (i * 4096);
-               EFX_LOG(efx, "mapping special buffer %d at %llx\n",
-                       index, (unsigned long long)dma_addr);
-               EFX_POPULATE_QWORD_3(buf_desc,
-                                    FRF_AZ_BUF_ADR_REGION, 0,
-                                    FRF_AZ_BUF_ADR_FBUF, dma_addr >> 12,
-                                    FRF_AZ_BUF_OWNER_ID_FBUF, 0);
-               falcon_write_buf_tbl(efx, &buf_desc, index);
-       }
-}
-
-/* Unmaps a buffer from Falcon and clears the buffer table entries */
-static void
-falcon_fini_special_buffer(struct efx_nic *efx,
-                          struct efx_special_buffer *buffer)
-{
-       efx_oword_t buf_tbl_upd;
-       unsigned int start = buffer->index;
-       unsigned int end = (buffer->index + buffer->entries - 1);
-
-       if (!buffer->entries)
-               return;
-
-       EFX_LOG(efx, "unmapping special buffers %d-%d\n",
-               buffer->index, buffer->index + buffer->entries - 1);
-
-       EFX_POPULATE_OWORD_4(buf_tbl_upd,
-                            FRF_AZ_BUF_UPD_CMD, 0,
-                            FRF_AZ_BUF_CLR_CMD, 1,
-                            FRF_AZ_BUF_CLR_END_ID, end,
-                            FRF_AZ_BUF_CLR_START_ID, start);
-       efx_writeo(efx, &buf_tbl_upd, FR_AZ_BUF_TBL_UPD);
-}
-
-/*
- * Allocate a new Falcon special buffer
- *
- * This allocates memory for a new buffer, clears it and allocates a
- * new buffer ID range.  It does not write into Falcon's buffer table.
- *
- * This call will allocate 4KB buffers, since Falcon can't use 8KB
- * buffers for event queues and descriptor rings.
- */
-static int falcon_alloc_special_buffer(struct efx_nic *efx,
-                                      struct efx_special_buffer *buffer,
-                                      unsigned int len)
-{
-       len = ALIGN(len, FALCON_BUF_SIZE);
-
-       buffer->addr = pci_alloc_consistent(efx->pci_dev, len,
-                                           &buffer->dma_addr);
-       if (!buffer->addr)
-               return -ENOMEM;
-       buffer->len = len;
-       buffer->entries = len / FALCON_BUF_SIZE;
-       BUG_ON(buffer->dma_addr & (FALCON_BUF_SIZE - 1));
-
-       /* All zeros is a potentially valid event so memset to 0xff */
-       memset(buffer->addr, 0xff, len);
-
-       /* Select new buffer ID */
-       buffer->index = efx->next_buffer_table;
-       efx->next_buffer_table += buffer->entries;
-
-       EFX_LOG(efx, "allocating special buffers %d-%d at %llx+%x "
-               "(virt %p phys %llx)\n", buffer->index,
-               buffer->index + buffer->entries - 1,
-               (u64)buffer->dma_addr, len,
-               buffer->addr, (u64)virt_to_phys(buffer->addr));
-
-       return 0;
-}
-
-static void falcon_free_special_buffer(struct efx_nic *efx,
-                                      struct efx_special_buffer *buffer)
-{
-       if (!buffer->addr)
-               return;
-
-       EFX_LOG(efx, "deallocating special buffers %d-%d at %llx+%x "
-               "(virt %p phys %llx)\n", buffer->index,
-               buffer->index + buffer->entries - 1,
-               (u64)buffer->dma_addr, buffer->len,
-               buffer->addr, (u64)virt_to_phys(buffer->addr));
-
-       pci_free_consistent(efx->pci_dev, buffer->len, buffer->addr,
-                           buffer->dma_addr);
-       buffer->addr = NULL;
-       buffer->entries = 0;
-}
-
-/**************************************************************************
- *
- * Falcon generic buffer handling
- * These buffers are used for interrupt status and MAC stats
- *
- **************************************************************************/
-
-static int falcon_alloc_buffer(struct efx_nic *efx,
-                              struct efx_buffer *buffer, unsigned int len)
-{
-       buffer->addr = pci_alloc_consistent(efx->pci_dev, len,
-                                           &buffer->dma_addr);
-       if (!buffer->addr)
-               return -ENOMEM;
-       buffer->len = len;
-       memset(buffer->addr, 0, len);
-       return 0;
-}
-
-static void falcon_free_buffer(struct efx_nic *efx, struct efx_buffer *buffer)
-{
-       if (buffer->addr) {
-               pci_free_consistent(efx->pci_dev, buffer->len,
-                                   buffer->addr, buffer->dma_addr);
-               buffer->addr = NULL;
-       }
-}
-
-/**************************************************************************
- *
- * Falcon TX path
- *
- **************************************************************************/
-
-/* Returns a pointer to the specified transmit descriptor in the TX
- * descriptor queue belonging to the specified channel.
- */
-static inline efx_qword_t *falcon_tx_desc(struct efx_tx_queue *tx_queue,
-                                              unsigned int index)
-{
-       return (((efx_qword_t *) (tx_queue->txd.addr)) + index);
-}
-
-/* This writes to the TX_DESC_WPTR; write pointer for TX descriptor ring */
-static inline void falcon_notify_tx_desc(struct efx_tx_queue *tx_queue)
-{
-       unsigned write_ptr;
-       efx_dword_t reg;
-
-       write_ptr = tx_queue->write_count & EFX_TXQ_MASK;
-       EFX_POPULATE_DWORD_1(reg, FRF_AZ_TX_DESC_WPTR_DWORD, write_ptr);
-       efx_writed_page(tx_queue->efx, &reg,
-                       FR_AZ_TX_DESC_UPD_DWORD_P0, tx_queue->queue);
-}
-
-
-/* For each entry inserted into the software descriptor ring, create a
- * descriptor in the hardware TX descriptor ring (in host memory), and
- * write a doorbell.
- */
-void falcon_push_buffers(struct efx_tx_queue *tx_queue)
-{
-
-       struct efx_tx_buffer *buffer;
-       efx_qword_t *txd;
-       unsigned write_ptr;
-
-       BUG_ON(tx_queue->write_count == tx_queue->insert_count);
-
-       do {
-               write_ptr = tx_queue->write_count & EFX_TXQ_MASK;
-               buffer = &tx_queue->buffer[write_ptr];
-               txd = falcon_tx_desc(tx_queue, write_ptr);
-               ++tx_queue->write_count;
-
-               /* Create TX descriptor ring entry */
-               EFX_POPULATE_QWORD_4(*txd,
-                                    FSF_AZ_TX_KER_CONT, buffer->continuation,
-                                    FSF_AZ_TX_KER_BYTE_COUNT, buffer->len,
-                                    FSF_AZ_TX_KER_BUF_REGION, 0,
-                                    FSF_AZ_TX_KER_BUF_ADDR, buffer->dma_addr);
-       } while (tx_queue->write_count != tx_queue->insert_count);
-
-       wmb(); /* Ensure descriptors are written before they are fetched */
-       falcon_notify_tx_desc(tx_queue);
-}
-
-/* Allocate hardware resources for a TX queue */
-int falcon_probe_tx(struct efx_tx_queue *tx_queue)
-{
-       struct efx_nic *efx = tx_queue->efx;
-       BUILD_BUG_ON(EFX_TXQ_SIZE < 512 || EFX_TXQ_SIZE > 4096 ||
-                    EFX_TXQ_SIZE & EFX_TXQ_MASK);
-       return falcon_alloc_special_buffer(efx, &tx_queue->txd,
-                                          EFX_TXQ_SIZE * sizeof(efx_qword_t));
-}
-
-void falcon_init_tx(struct efx_tx_queue *tx_queue)
-{
-       efx_oword_t tx_desc_ptr;
-       struct efx_nic *efx = tx_queue->efx;
-
-       tx_queue->flushed = false;
-
-       /* Pin TX descriptor ring */
-       falcon_init_special_buffer(efx, &tx_queue->txd);
-
-       /* Push TX descriptor ring to card */
-       EFX_POPULATE_OWORD_10(tx_desc_ptr,
-                             FRF_AZ_TX_DESCQ_EN, 1,
-                             FRF_AZ_TX_ISCSI_DDIG_EN, 0,
-                             FRF_AZ_TX_ISCSI_HDIG_EN, 0,
-                             FRF_AZ_TX_DESCQ_BUF_BASE_ID, tx_queue->txd.index,
-                             FRF_AZ_TX_DESCQ_EVQ_ID,
-                             tx_queue->channel->channel,
-                             FRF_AZ_TX_DESCQ_OWNER_ID, 0,
-                             FRF_AZ_TX_DESCQ_LABEL, tx_queue->queue,
-                             FRF_AZ_TX_DESCQ_SIZE,
-                             __ffs(tx_queue->txd.entries),
-                             FRF_AZ_TX_DESCQ_TYPE, 0,
-                             FRF_BZ_TX_NON_IP_DROP_DIS, 1);
-
-       if (falcon_rev(efx) >= FALCON_REV_B0) {
-               int csum = tx_queue->queue == EFX_TX_QUEUE_OFFLOAD_CSUM;
-               EFX_SET_OWORD_FIELD(tx_desc_ptr, FRF_BZ_TX_IP_CHKSM_DIS, !csum);
-               EFX_SET_OWORD_FIELD(tx_desc_ptr, FRF_BZ_TX_TCP_CHKSM_DIS,
-                                   !csum);
-       }
-
-       efx_writeo_table(efx, &tx_desc_ptr, efx->type->txd_ptr_tbl_base,
-                        tx_queue->queue);
-
-       if (falcon_rev(efx) < FALCON_REV_B0) {
-               efx_oword_t reg;
-
-               /* Only 128 bits in this register */
-               BUILD_BUG_ON(EFX_TX_QUEUE_COUNT >= 128);
-
-               efx_reado(efx, &reg, FR_AA_TX_CHKSM_CFG);
-               if (tx_queue->queue == EFX_TX_QUEUE_OFFLOAD_CSUM)
-                       clear_bit_le(tx_queue->queue, (void *)&reg);
-               else
-                       set_bit_le(tx_queue->queue, (void *)&reg);
-               efx_writeo(efx, &reg, FR_AA_TX_CHKSM_CFG);
-       }
-}
-
-static void falcon_flush_tx_queue(struct efx_tx_queue *tx_queue)
-{
-       struct efx_nic *efx = tx_queue->efx;
-       efx_oword_t tx_flush_descq;
-
-       /* Post a flush command */
-       EFX_POPULATE_OWORD_2(tx_flush_descq,
-                            FRF_AZ_TX_FLUSH_DESCQ_CMD, 1,
-                            FRF_AZ_TX_FLUSH_DESCQ, tx_queue->queue);
-       efx_writeo(efx, &tx_flush_descq, FR_AZ_TX_FLUSH_DESCQ);
-}
-
-void falcon_fini_tx(struct efx_tx_queue *tx_queue)
-{
-       struct efx_nic *efx = tx_queue->efx;
-       efx_oword_t tx_desc_ptr;
-
-       /* The queue should have been flushed */
-       WARN_ON(!tx_queue->flushed);
-
-       /* Remove TX descriptor ring from card */
-       EFX_ZERO_OWORD(tx_desc_ptr);
-       efx_writeo_table(efx, &tx_desc_ptr, efx->type->txd_ptr_tbl_base,
-                        tx_queue->queue);
-
-       /* Unpin TX descriptor ring */
-       falcon_fini_special_buffer(efx, &tx_queue->txd);
-}
-
-/* Free buffers backing TX queue */
-void falcon_remove_tx(struct efx_tx_queue *tx_queue)
-{
-       falcon_free_special_buffer(tx_queue->efx, &tx_queue->txd);
-}
-
-/**************************************************************************
- *
- * Falcon RX path
- *
- **************************************************************************/
-
-/* Returns a pointer to the specified descriptor in the RX descriptor queue */
-static inline efx_qword_t *falcon_rx_desc(struct efx_rx_queue *rx_queue,
-                                              unsigned int index)
-{
-       return (((efx_qword_t *) (rx_queue->rxd.addr)) + index);
-}
-
-/* This creates an entry in the RX descriptor queue */
-static inline void falcon_build_rx_desc(struct efx_rx_queue *rx_queue,
-                                       unsigned index)
-{
-       struct efx_rx_buffer *rx_buf;
-       efx_qword_t *rxd;
-
-       rxd = falcon_rx_desc(rx_queue, index);
-       rx_buf = efx_rx_buffer(rx_queue, index);
-       EFX_POPULATE_QWORD_3(*rxd,
-                            FSF_AZ_RX_KER_BUF_SIZE,
-                            rx_buf->len -
-                            rx_queue->efx->type->rx_buffer_padding,
-                            FSF_AZ_RX_KER_BUF_REGION, 0,
-                            FSF_AZ_RX_KER_BUF_ADDR, rx_buf->dma_addr);
-}
-
-/* This writes to the RX_DESC_WPTR register for the specified receive
- * descriptor ring.
- */
-void falcon_notify_rx_desc(struct efx_rx_queue *rx_queue)
-{
-       efx_dword_t reg;
-       unsigned write_ptr;
-
-       while (rx_queue->notified_count != rx_queue->added_count) {
-               falcon_build_rx_desc(rx_queue,
-                                    rx_queue->notified_count &
-                                    EFX_RXQ_MASK);
-               ++rx_queue->notified_count;
-       }
-
-       wmb();
-       write_ptr = rx_queue->added_count & EFX_RXQ_MASK;
-       EFX_POPULATE_DWORD_1(reg, FRF_AZ_RX_DESC_WPTR_DWORD, write_ptr);
-       efx_writed_page(rx_queue->efx, &reg,
-                       FR_AZ_RX_DESC_UPD_DWORD_P0, rx_queue->queue);
-}
-
-int falcon_probe_rx(struct efx_rx_queue *rx_queue)
-{
-       struct efx_nic *efx = rx_queue->efx;
-       BUILD_BUG_ON(EFX_RXQ_SIZE < 512 || EFX_RXQ_SIZE > 4096 ||
-                    EFX_RXQ_SIZE & EFX_RXQ_MASK);
-       return falcon_alloc_special_buffer(efx, &rx_queue->rxd,
-                                          EFX_RXQ_SIZE * sizeof(efx_qword_t));
-}
-
-void falcon_init_rx(struct efx_rx_queue *rx_queue)
-{
-       efx_oword_t rx_desc_ptr;
-       struct efx_nic *efx = rx_queue->efx;
-       bool is_b0 = falcon_rev(efx) >= FALCON_REV_B0;
-       bool iscsi_digest_en = is_b0;
-
-       EFX_LOG(efx, "RX queue %d ring in special buffers %d-%d\n",
-               rx_queue->queue, rx_queue->rxd.index,
-               rx_queue->rxd.index + rx_queue->rxd.entries - 1);
-
-       rx_queue->flushed = false;
-
-       /* Pin RX descriptor ring */
-       falcon_init_special_buffer(efx, &rx_queue->rxd);
-
-       /* Push RX descriptor ring to card */
-       EFX_POPULATE_OWORD_10(rx_desc_ptr,
-                             FRF_AZ_RX_ISCSI_DDIG_EN, iscsi_digest_en,
-                             FRF_AZ_RX_ISCSI_HDIG_EN, iscsi_digest_en,
-                             FRF_AZ_RX_DESCQ_BUF_BASE_ID, rx_queue->rxd.index,
-                             FRF_AZ_RX_DESCQ_EVQ_ID,
-                             rx_queue->channel->channel,
-                             FRF_AZ_RX_DESCQ_OWNER_ID, 0,
-                             FRF_AZ_RX_DESCQ_LABEL, rx_queue->queue,
-                             FRF_AZ_RX_DESCQ_SIZE,
-                             __ffs(rx_queue->rxd.entries),
-                             FRF_AZ_RX_DESCQ_TYPE, 0 /* kernel queue */ ,
-                             /* For >=B0 this is scatter so disable */
-                             FRF_AZ_RX_DESCQ_JUMBO, !is_b0,
-                             FRF_AZ_RX_DESCQ_EN, 1);
-       efx_writeo_table(efx, &rx_desc_ptr, efx->type->rxd_ptr_tbl_base,
-                        rx_queue->queue);
-}
-
-static void falcon_flush_rx_queue(struct efx_rx_queue *rx_queue)
-{
-       struct efx_nic *efx = rx_queue->efx;
-       efx_oword_t rx_flush_descq;
-
-       /* Post a flush command */
-       EFX_POPULATE_OWORD_2(rx_flush_descq,
-                            FRF_AZ_RX_FLUSH_DESCQ_CMD, 1,
-                            FRF_AZ_RX_FLUSH_DESCQ, rx_queue->queue);
-       efx_writeo(efx, &rx_flush_descq, FR_AZ_RX_FLUSH_DESCQ);
-}
-
-void falcon_fini_rx(struct efx_rx_queue *rx_queue)
-{
-       efx_oword_t rx_desc_ptr;
-       struct efx_nic *efx = rx_queue->efx;
-
-       /* The queue should already have been flushed */
-       WARN_ON(!rx_queue->flushed);
-
-       /* Remove RX descriptor ring from card */
-       EFX_ZERO_OWORD(rx_desc_ptr);
-       efx_writeo_table(efx, &rx_desc_ptr, efx->type->rxd_ptr_tbl_base,
-                        rx_queue->queue);
-
-       /* Unpin RX descriptor ring */
-       falcon_fini_special_buffer(efx, &rx_queue->rxd);
-}
-
-/* Free buffers backing RX queue */
-void falcon_remove_rx(struct efx_rx_queue *rx_queue)
-{
-       falcon_free_special_buffer(rx_queue->efx, &rx_queue->rxd);
-}
-
-/**************************************************************************
- *
- * Falcon event queue processing
- * Event queues are processed by per-channel tasklets.
- *
- **************************************************************************/
-
-/* Update a channel's event queue's read pointer (RPTR) register
- *
- * This writes the EVQ_RPTR_REG register for the specified channel's
- * event queue.
- *
- * Note that EVQ_RPTR_REG contains the index of the "last read" event,
- * whereas channel->eventq_read_ptr contains the index of the "next to
- * read" event.
- */
-void falcon_eventq_read_ack(struct efx_channel *channel)
-{
-       efx_dword_t reg;
-       struct efx_nic *efx = channel->efx;
-
-       EFX_POPULATE_DWORD_1(reg, FRF_AZ_EVQ_RPTR, channel->eventq_read_ptr);
-       efx_writed_table(efx, &reg, efx->type->evq_rptr_tbl_base,
-                           channel->channel);
-}
-
-/* Use HW to insert a SW defined event */
-void falcon_generate_event(struct efx_channel *channel, efx_qword_t *event)
-{
-       efx_oword_t drv_ev_reg;
-
-       BUILD_BUG_ON(FRF_AZ_DRV_EV_DATA_LBN != 0 ||
-                    FRF_AZ_DRV_EV_DATA_WIDTH != 64);
-       drv_ev_reg.u32[0] = event->u32[0];
-       drv_ev_reg.u32[1] = event->u32[1];
-       drv_ev_reg.u32[2] = 0;
-       drv_ev_reg.u32[3] = 0;
-       EFX_SET_OWORD_FIELD(drv_ev_reg, FRF_AZ_DRV_EV_QID, channel->channel);
-       efx_writeo(channel->efx, &drv_ev_reg, FR_AZ_DRV_EV);
-}
-
-/* Handle a transmit completion event
- *
- * Falcon batches TX completion events; the message we receive is of
- * the form "complete all TX events up to this index".
- */
-static void falcon_handle_tx_event(struct efx_channel *channel,
-                                  efx_qword_t *event)
-{
-       unsigned int tx_ev_desc_ptr;
-       unsigned int tx_ev_q_label;
-       struct efx_tx_queue *tx_queue;
-       struct efx_nic *efx = channel->efx;
-
-       if (likely(EFX_QWORD_FIELD(*event, FSF_AZ_TX_EV_COMP))) {
-               /* Transmit completion */
-               tx_ev_desc_ptr = EFX_QWORD_FIELD(*event, FSF_AZ_TX_EV_DESC_PTR);
-               tx_ev_q_label = EFX_QWORD_FIELD(*event, FSF_AZ_TX_EV_Q_LABEL);
-               tx_queue = &efx->tx_queue[tx_ev_q_label];
-               channel->irq_mod_score +=
-                       (tx_ev_desc_ptr - tx_queue->read_count) &
-                       EFX_TXQ_MASK;
-               efx_xmit_done(tx_queue, tx_ev_desc_ptr);
-       } else if (EFX_QWORD_FIELD(*event, FSF_AZ_TX_EV_WQ_FF_FULL)) {
-               /* Rewrite the FIFO write pointer */
-               tx_ev_q_label = EFX_QWORD_FIELD(*event, FSF_AZ_TX_EV_Q_LABEL);
-               tx_queue = &efx->tx_queue[tx_ev_q_label];
-
-               if (efx_dev_registered(efx))
-                       netif_tx_lock(efx->net_dev);
-               falcon_notify_tx_desc(tx_queue);
-               if (efx_dev_registered(efx))
-                       netif_tx_unlock(efx->net_dev);
-       } else if (EFX_QWORD_FIELD(*event, FSF_AZ_TX_EV_PKT_ERR) &&
-                  EFX_WORKAROUND_10727(efx)) {
-               efx_schedule_reset(efx, RESET_TYPE_TX_DESC_FETCH);
-       } else {
-               EFX_ERR(efx, "channel %d unexpected TX event "
-                       EFX_QWORD_FMT"\n", channel->channel,
-                       EFX_QWORD_VAL(*event));
-       }
-}
-
-/* Detect errors included in the rx_evt_pkt_ok bit. */
-static void falcon_handle_rx_not_ok(struct efx_rx_queue *rx_queue,
-                                   const efx_qword_t *event,
-                                   bool *rx_ev_pkt_ok,
-                                   bool *discard)
-{
-       struct efx_nic *efx = rx_queue->efx;
-       bool rx_ev_buf_owner_id_err, rx_ev_ip_hdr_chksum_err;
-       bool rx_ev_tcp_udp_chksum_err, rx_ev_eth_crc_err;
-       bool rx_ev_frm_trunc, rx_ev_drib_nib, rx_ev_tobe_disc;
-       bool rx_ev_other_err, rx_ev_pause_frm;
-       bool rx_ev_ip_frag_err, rx_ev_hdr_type, rx_ev_mcast_pkt;
-       unsigned rx_ev_pkt_type;
-
-       rx_ev_hdr_type = EFX_QWORD_FIELD(*event, FSF_AZ_RX_EV_HDR_TYPE);
-       rx_ev_mcast_pkt = EFX_QWORD_FIELD(*event, FSF_AZ_RX_EV_MCAST_PKT);
-       rx_ev_tobe_disc = EFX_QWORD_FIELD(*event, FSF_AZ_RX_EV_TOBE_DISC);
-       rx_ev_pkt_type = EFX_QWORD_FIELD(*event, FSF_AZ_RX_EV_PKT_TYPE);
-       rx_ev_buf_owner_id_err = EFX_QWORD_FIELD(*event,
-                                                FSF_AZ_RX_EV_BUF_OWNER_ID_ERR);
-       rx_ev_ip_frag_err = EFX_QWORD_FIELD(*event, FSF_AZ_RX_EV_IP_FRAG_ERR);
-       rx_ev_ip_hdr_chksum_err = EFX_QWORD_FIELD(*event,
-                                                 FSF_AZ_RX_EV_IP_HDR_CHKSUM_ERR);
-       rx_ev_tcp_udp_chksum_err = EFX_QWORD_FIELD(*event,
-                                                  FSF_AZ_RX_EV_TCP_UDP_CHKSUM_ERR);
-       rx_ev_eth_crc_err = EFX_QWORD_FIELD(*event, FSF_AZ_RX_EV_ETH_CRC_ERR);
-       rx_ev_frm_trunc = EFX_QWORD_FIELD(*event, FSF_AZ_RX_EV_FRM_TRUNC);
-       rx_ev_drib_nib = ((falcon_rev(efx) >= FALCON_REV_B0) ?
-                         0 : EFX_QWORD_FIELD(*event, FSF_AA_RX_EV_DRIB_NIB));
-       rx_ev_pause_frm = EFX_QWORD_FIELD(*event, FSF_AZ_RX_EV_PAUSE_FRM_ERR);
-
-       /* Every error apart from tobe_disc and pause_frm */
-       rx_ev_other_err = (rx_ev_drib_nib | rx_ev_tcp_udp_chksum_err |
-                          rx_ev_buf_owner_id_err | rx_ev_eth_crc_err |
-                          rx_ev_frm_trunc | rx_ev_ip_hdr_chksum_err);
-
-       /* Count errors that are not in MAC stats.  Ignore expected
-        * checksum errors during self-test. */
-       if (rx_ev_frm_trunc)
-               ++rx_queue->channel->n_rx_frm_trunc;
-       else if (rx_ev_tobe_disc)
-               ++rx_queue->channel->n_rx_tobe_disc;
-       else if (!efx->loopback_selftest) {
-               if (rx_ev_ip_hdr_chksum_err)
-                       ++rx_queue->channel->n_rx_ip_hdr_chksum_err;
-               else if (rx_ev_tcp_udp_chksum_err)
-                       ++rx_queue->channel->n_rx_tcp_udp_chksum_err;
-       }
-       if (rx_ev_ip_frag_err)
-               ++rx_queue->channel->n_rx_ip_frag_err;
-
-       /* The frame must be discarded if any of these are true. */
-       *discard = (rx_ev_eth_crc_err | rx_ev_frm_trunc | rx_ev_drib_nib |
-                   rx_ev_tobe_disc | rx_ev_pause_frm);
-
-       /* TOBE_DISC is expected on unicast mismatches; don't print out an
-        * error message.  FRM_TRUNC indicates RXDP dropped the packet due
-        * to a FIFO overflow.
-        */
-#ifdef EFX_ENABLE_DEBUG
-       if (rx_ev_other_err) {
-               EFX_INFO_RL(efx, " RX queue %d unexpected RX event "
-                           EFX_QWORD_FMT "%s%s%s%s%s%s%s%s\n",
-                           rx_queue->queue, EFX_QWORD_VAL(*event),
-                           rx_ev_buf_owner_id_err ? " [OWNER_ID_ERR]" : "",
-                           rx_ev_ip_hdr_chksum_err ?
-                           " [IP_HDR_CHKSUM_ERR]" : "",
-                           rx_ev_tcp_udp_chksum_err ?
-                           " [TCP_UDP_CHKSUM_ERR]" : "",
-                           rx_ev_eth_crc_err ? " [ETH_CRC_ERR]" : "",
-                           rx_ev_frm_trunc ? " [FRM_TRUNC]" : "",
-                           rx_ev_drib_nib ? " [DRIB_NIB]" : "",
-                           rx_ev_tobe_disc ? " [TOBE_DISC]" : "",
-                           rx_ev_pause_frm ? " [PAUSE]" : "");
-       }
-#endif
-}
-
-/* Handle receive events that are not in-order. */
-static void falcon_handle_rx_bad_index(struct efx_rx_queue *rx_queue,
-                                      unsigned index)
-{
-       struct efx_nic *efx = rx_queue->efx;
-       unsigned expected, dropped;
-
-       expected = rx_queue->removed_count & EFX_RXQ_MASK;
-       dropped = (index - expected) & EFX_RXQ_MASK;
-       EFX_INFO(efx, "dropped %d events (index=%d expected=%d)\n",
-               dropped, index, expected);
-
-       efx_schedule_reset(efx, EFX_WORKAROUND_5676(efx) ?
-                          RESET_TYPE_RX_RECOVERY : RESET_TYPE_DISABLE);
-}
-
-/* Handle a packet received event
- *
- * Falcon silicon gives a "discard" flag if it's a unicast packet with the
- * wrong destination address
- * Also "is multicast" and "matches multicast filter" flags can be used to
- * discard non-matching multicast packets.
- */
-static void falcon_handle_rx_event(struct efx_channel *channel,
-                                  const efx_qword_t *event)
-{
-       unsigned int rx_ev_desc_ptr, rx_ev_byte_cnt;
-       unsigned int rx_ev_hdr_type, rx_ev_mcast_pkt;
-       unsigned expected_ptr;
-       bool rx_ev_pkt_ok, discard = false, checksummed;
-       struct efx_rx_queue *rx_queue;
-       struct efx_nic *efx = channel->efx;
-
-       /* Basic packet information */
-       rx_ev_byte_cnt = EFX_QWORD_FIELD(*event, FSF_AZ_RX_EV_BYTE_CNT);
-       rx_ev_pkt_ok = EFX_QWORD_FIELD(*event, FSF_AZ_RX_EV_PKT_OK);
-       rx_ev_hdr_type = EFX_QWORD_FIELD(*event, FSF_AZ_RX_EV_HDR_TYPE);
-       WARN_ON(EFX_QWORD_FIELD(*event, FSF_AZ_RX_EV_JUMBO_CONT));
-       WARN_ON(EFX_QWORD_FIELD(*event, FSF_AZ_RX_EV_SOP) != 1);
-       WARN_ON(EFX_QWORD_FIELD(*event, FSF_AZ_RX_EV_Q_LABEL) !=
-               channel->channel);
-
-       rx_queue = &efx->rx_queue[channel->channel];
-
-       rx_ev_desc_ptr = EFX_QWORD_FIELD(*event, FSF_AZ_RX_EV_DESC_PTR);
-       expected_ptr = rx_queue->removed_count & EFX_RXQ_MASK;
-       if (unlikely(rx_ev_desc_ptr != expected_ptr))
-               falcon_handle_rx_bad_index(rx_queue, rx_ev_desc_ptr);
-
-       if (likely(rx_ev_pkt_ok)) {
-               /* If packet is marked as OK and packet type is TCP/IPv4 or
-                * UDP/IPv4, then we can rely on the hardware checksum.
-                */
-               checksummed =
-                       efx->rx_checksum_enabled &&
-                       (rx_ev_hdr_type == FSE_AB_RX_EV_HDR_TYPE_IPV4_TCP ||
-                        rx_ev_hdr_type == FSE_AB_RX_EV_HDR_TYPE_IPV4_UDP);
-       } else {
-               falcon_handle_rx_not_ok(rx_queue, event, &rx_ev_pkt_ok,
-                                       &discard);
-               checksummed = false;
-       }
-
-       /* Detect multicast packets that didn't match the filter */
-       rx_ev_mcast_pkt = EFX_QWORD_FIELD(*event, FSF_AZ_RX_EV_MCAST_PKT);
-       if (rx_ev_mcast_pkt) {
-               unsigned int rx_ev_mcast_hash_match =
-                       EFX_QWORD_FIELD(*event, FSF_AZ_RX_EV_MCAST_HASH_MATCH);
-
-               if (unlikely(!rx_ev_mcast_hash_match))
-                       discard = true;
-       }
-
-       channel->irq_mod_score += 2;
-
-       /* Handle received packet */
-       efx_rx_packet(rx_queue, rx_ev_desc_ptr, rx_ev_byte_cnt,
-                     checksummed, discard);
-}
-
-/* Global events are basically PHY events */
-static void falcon_handle_global_event(struct efx_channel *channel,
-                                      efx_qword_t *event)
-{
-       struct efx_nic *efx = channel->efx;
-       bool handled = false;
-
-       if (EFX_QWORD_FIELD(*event, FSF_AB_GLB_EV_G_PHY0_INTR) ||
-           EFX_QWORD_FIELD(*event, FSF_AB_GLB_EV_XG_PHY0_INTR) ||
-           EFX_QWORD_FIELD(*event, FSF_AB_GLB_EV_XFP_PHY0_INTR)) {
-               efx->phy_op->clear_interrupt(efx);
-               queue_work(efx->workqueue, &efx->phy_work);
-               handled = true;
-       }
-
-       if ((falcon_rev(efx) >= FALCON_REV_B0) &&
-           EFX_QWORD_FIELD(*event, FSF_BB_GLB_EV_XG_MGT_INTR)) {
-               queue_work(efx->workqueue, &efx->mac_work);
-               handled = true;
-       }
-
-       if (falcon_rev(efx) <= FALCON_REV_A1 ?
-           EFX_QWORD_FIELD(*event, FSF_AA_GLB_EV_RX_RECOVERY) :
-           EFX_QWORD_FIELD(*event, FSF_BB_GLB_EV_RX_RECOVERY)) {
-               EFX_ERR(efx, "channel %d seen global RX_RESET "
-                       "event. Resetting.\n", channel->channel);
-
-               atomic_inc(&efx->rx_reset);
-               efx_schedule_reset(efx, EFX_WORKAROUND_6555(efx) ?
-                                  RESET_TYPE_RX_RECOVERY : RESET_TYPE_DISABLE);
-               handled = true;
-       }
-
-       if (!handled)
-               EFX_ERR(efx, "channel %d unknown global event "
-                       EFX_QWORD_FMT "\n", channel->channel,
-                       EFX_QWORD_VAL(*event));
-}
-
-static void falcon_handle_driver_event(struct efx_channel *channel,
-                                      efx_qword_t *event)
-{
-       struct efx_nic *efx = channel->efx;
-       unsigned int ev_sub_code;
-       unsigned int ev_sub_data;
-
-       ev_sub_code = EFX_QWORD_FIELD(*event, FSF_AZ_DRIVER_EV_SUBCODE);
-       ev_sub_data = EFX_QWORD_FIELD(*event, FSF_AZ_DRIVER_EV_SUBDATA);
-
-       switch (ev_sub_code) {
-       case FSE_AZ_TX_DESCQ_FLS_DONE_EV:
-               EFX_TRACE(efx, "channel %d TXQ %d flushed\n",
-                         channel->channel, ev_sub_data);
-               break;
-       case FSE_AZ_RX_DESCQ_FLS_DONE_EV:
-               EFX_TRACE(efx, "channel %d RXQ %d flushed\n",
-                         channel->channel, ev_sub_data);
-               break;
-       case FSE_AZ_EVQ_INIT_DONE_EV:
-               EFX_LOG(efx, "channel %d EVQ %d initialised\n",
-                       channel->channel, ev_sub_data);
-               break;
-       case FSE_AZ_SRM_UPD_DONE_EV:
-               EFX_TRACE(efx, "channel %d SRAM update done\n",
-                         channel->channel);
-               break;
-       case FSE_AZ_WAKE_UP_EV:
-               EFX_TRACE(efx, "channel %d RXQ %d wakeup event\n",
-                         channel->channel, ev_sub_data);
-               break;
-       case FSE_AZ_TIMER_EV:
-               EFX_TRACE(efx, "channel %d RX queue %d timer expired\n",
-                         channel->channel, ev_sub_data);
-               break;
-       case FSE_AA_RX_RECOVER_EV:
-               EFX_ERR(efx, "channel %d seen DRIVER RX_RESET event. "
-                       "Resetting.\n", channel->channel);
-               atomic_inc(&efx->rx_reset);
-               efx_schedule_reset(efx,
-                                  EFX_WORKAROUND_6555(efx) ?
-                                  RESET_TYPE_RX_RECOVERY :
-                                  RESET_TYPE_DISABLE);
-               break;
-       case FSE_BZ_RX_DSC_ERROR_EV:
-               EFX_ERR(efx, "RX DMA Q %d reports descriptor fetch error."
-                       " RX Q %d is disabled.\n", ev_sub_data, ev_sub_data);
-               efx_schedule_reset(efx, RESET_TYPE_RX_DESC_FETCH);
-               break;
-       case FSE_BZ_TX_DSC_ERROR_EV:
-               EFX_ERR(efx, "TX DMA Q %d reports descriptor fetch error."
-                       " TX Q %d is disabled.\n", ev_sub_data, ev_sub_data);
-               efx_schedule_reset(efx, RESET_TYPE_TX_DESC_FETCH);
-               break;
-       default:
-               EFX_TRACE(efx, "channel %d unknown driver event code %d "
-                         "data %04x\n", channel->channel, ev_sub_code,
-                         ev_sub_data);
-               break;
-       }
-}
-
-int falcon_process_eventq(struct efx_channel *channel, int rx_quota)
-{
-       unsigned int read_ptr;
-       efx_qword_t event, *p_event;
-       int ev_code;
-       int rx_packets = 0;
-
-       read_ptr = channel->eventq_read_ptr;
-
-       do {
-               p_event = falcon_event(channel, read_ptr);
-               event = *p_event;
-
-               if (!falcon_event_present(&event))
-                       /* End of events */
-                       break;
-
-               EFX_TRACE(channel->efx, "channel %d event is "EFX_QWORD_FMT"\n",
-                         channel->channel, EFX_QWORD_VAL(event));
-
-               /* Clear this event by marking it all ones */
-               EFX_SET_QWORD(*p_event);
-
-               ev_code = EFX_QWORD_FIELD(event, FSF_AZ_EV_CODE);
-
-               switch (ev_code) {
-               case FSE_AZ_EV_CODE_RX_EV:
-                       falcon_handle_rx_event(channel, &event);
-                       ++rx_packets;
-                       break;
-               case FSE_AZ_EV_CODE_TX_EV:
-                       falcon_handle_tx_event(channel, &event);
-                       break;
-               case FSE_AZ_EV_CODE_DRV_GEN_EV:
-                       channel->eventq_magic = EFX_QWORD_FIELD(
-                               event, FSF_AZ_DRV_GEN_EV_MAGIC);
-                       EFX_LOG(channel->efx, "channel %d received generated "
-                               "event "EFX_QWORD_FMT"\n", channel->channel,
-                               EFX_QWORD_VAL(event));
-                       break;
-               case FSE_AZ_EV_CODE_GLOBAL_EV:
-                       falcon_handle_global_event(channel, &event);
-                       break;
-               case FSE_AZ_EV_CODE_DRIVER_EV:
-                       falcon_handle_driver_event(channel, &event);
-                       break;
-               default:
-                       EFX_ERR(channel->efx, "channel %d unknown event type %d"
-                               " (data " EFX_QWORD_FMT ")\n", channel->channel,
-                               ev_code, EFX_QWORD_VAL(event));
-               }
-
-               /* Increment read pointer */
-               read_ptr = (read_ptr + 1) & EFX_EVQ_MASK;
-
-       } while (rx_packets < rx_quota);
-
-       channel->eventq_read_ptr = read_ptr;
-       return rx_packets;
-}
-
-void falcon_set_int_moderation(struct efx_channel *channel)
+static void falcon_push_irq_moderation(struct efx_channel *channel)
 {
        efx_dword_t timer_cmd;
        struct efx_nic *efx = channel->efx;
@@ -1067,265 +121,18 @@ void falcon_set_int_moderation(struct efx_channel *channel)
        BUILD_BUG_ON(FR_AA_TIMER_COMMAND_KER != FR_BZ_TIMER_COMMAND_P0);
        efx_writed_page_locked(efx, &timer_cmd, FR_BZ_TIMER_COMMAND_P0,
                               channel->channel);
-
-}
-
-/* Allocate buffer table entries for event queue */
-int falcon_probe_eventq(struct efx_channel *channel)
-{
-       struct efx_nic *efx = channel->efx;
-       BUILD_BUG_ON(EFX_EVQ_SIZE < 512 || EFX_EVQ_SIZE > 32768 ||
-                    EFX_EVQ_SIZE & EFX_EVQ_MASK);
-       return falcon_alloc_special_buffer(efx, &channel->eventq,
-                                          EFX_EVQ_SIZE * sizeof(efx_qword_t));
-}
-
-void falcon_init_eventq(struct efx_channel *channel)
-{
-       efx_oword_t evq_ptr;
-       struct efx_nic *efx = channel->efx;
-
-       EFX_LOG(efx, "channel %d event queue in special buffers %d-%d\n",
-               channel->channel, channel->eventq.index,
-               channel->eventq.index + channel->eventq.entries - 1);
-
-       /* Pin event queue buffer */
-       falcon_init_special_buffer(efx, &channel->eventq);
-
-       /* Fill event queue with all ones (i.e. empty events) */
-       memset(channel->eventq.addr, 0xff, channel->eventq.len);
-
-       /* Push event queue to card */
-       EFX_POPULATE_OWORD_3(evq_ptr,
-                            FRF_AZ_EVQ_EN, 1,
-                            FRF_AZ_EVQ_SIZE, __ffs(channel->eventq.entries),
-                            FRF_AZ_EVQ_BUF_BASE_ID, channel->eventq.index);
-       efx_writeo_table(efx, &evq_ptr, efx->type->evq_ptr_tbl_base,
-                        channel->channel);
-
-       falcon_set_int_moderation(channel);
-}
-
-void falcon_fini_eventq(struct efx_channel *channel)
-{
-       efx_oword_t eventq_ptr;
-       struct efx_nic *efx = channel->efx;
-
-       /* Remove event queue from card */
-       EFX_ZERO_OWORD(eventq_ptr);
-       efx_writeo_table(efx, &eventq_ptr, efx->type->evq_ptr_tbl_base,
-                        channel->channel);
-
-       /* Unpin event queue */
-       falcon_fini_special_buffer(efx, &channel->eventq);
-}
-
-/* Free buffers backing event queue */
-void falcon_remove_eventq(struct efx_channel *channel)
-{
-       falcon_free_special_buffer(channel->efx, &channel->eventq);
-}
-
-
-/* Generates a test event on the event queue.  A subsequent call to
- * process_eventq() should pick up the event and place the value of
- * "magic" into channel->eventq_magic;
- */
-void falcon_generate_test_event(struct efx_channel *channel, unsigned int magic)
-{
-       efx_qword_t test_event;
-
-       EFX_POPULATE_QWORD_2(test_event, FSF_AZ_EV_CODE,
-                            FSE_AZ_EV_CODE_DRV_GEN_EV,
-                            FSF_AZ_DRV_GEN_EV_MAGIC, magic);
-       falcon_generate_event(channel, &test_event);
-}
-
-void falcon_sim_phy_event(struct efx_nic *efx)
-{
-       efx_qword_t phy_event;
-
-       EFX_POPULATE_QWORD_1(phy_event, FSF_AZ_EV_CODE,
-                            FSE_AZ_EV_CODE_GLOBAL_EV);
-       if (EFX_IS10G(efx))
-               EFX_SET_QWORD_FIELD(phy_event, FSF_AB_GLB_EV_XG_PHY0_INTR, 1);
-       else
-               EFX_SET_QWORD_FIELD(phy_event, FSF_AB_GLB_EV_G_PHY0_INTR, 1);
-
-       falcon_generate_event(&efx->channel[0], &phy_event);
-}
-
-/**************************************************************************
- *
- * Flush handling
- *
- **************************************************************************/
-
-
-static void falcon_poll_flush_events(struct efx_nic *efx)
-{
-       struct efx_channel *channel = &efx->channel[0];
-       struct efx_tx_queue *tx_queue;
-       struct efx_rx_queue *rx_queue;
-       unsigned int read_ptr = channel->eventq_read_ptr;
-       unsigned int end_ptr = (read_ptr - 1) & EFX_EVQ_MASK;
-
-       do {
-               efx_qword_t *event = falcon_event(channel, read_ptr);
-               int ev_code, ev_sub_code, ev_queue;
-               bool ev_failed;
-
-               if (!falcon_event_present(event))
-                       break;
-
-               ev_code = EFX_QWORD_FIELD(*event, FSF_AZ_EV_CODE);
-               ev_sub_code = EFX_QWORD_FIELD(*event,
-                                             FSF_AZ_DRIVER_EV_SUBCODE);
-               if (ev_code == FSE_AZ_EV_CODE_DRIVER_EV &&
-                   ev_sub_code == FSE_AZ_TX_DESCQ_FLS_DONE_EV) {
-                       ev_queue = EFX_QWORD_FIELD(*event,
-                                                  FSF_AZ_DRIVER_EV_SUBDATA);
-                       if (ev_queue < EFX_TX_QUEUE_COUNT) {
-                               tx_queue = efx->tx_queue + ev_queue;
-                               tx_queue->flushed = true;
-                       }
-               } else if (ev_code == FSE_AZ_EV_CODE_DRIVER_EV &&
-                          ev_sub_code == FSE_AZ_RX_DESCQ_FLS_DONE_EV) {
-                       ev_queue = EFX_QWORD_FIELD(
-                               *event, FSF_AZ_DRIVER_EV_RX_DESCQ_ID);
-                       ev_failed = EFX_QWORD_FIELD(
-                               *event, FSF_AZ_DRIVER_EV_RX_FLUSH_FAIL);
-                       if (ev_queue < efx->n_rx_queues) {
-                               rx_queue = efx->rx_queue + ev_queue;
-
-                               /* retry the rx flush */
-                               if (ev_failed)
-                                       falcon_flush_rx_queue(rx_queue);
-                               else
-                                       rx_queue->flushed = true;
-                       }
-               }
-
-               read_ptr = (read_ptr + 1) & EFX_EVQ_MASK;
-       } while (read_ptr != end_ptr);
-}
-
-/* Handle tx and rx flushes at the same time, since they run in
- * parallel in the hardware and there's no reason for us to
- * serialise them */
-int falcon_flush_queues(struct efx_nic *efx)
-{
-       struct efx_rx_queue *rx_queue;
-       struct efx_tx_queue *tx_queue;
-       int i;
-       bool outstanding;
-
-       /* Issue flush requests */
-       efx_for_each_tx_queue(tx_queue, efx) {
-               tx_queue->flushed = false;
-               falcon_flush_tx_queue(tx_queue);
-       }
-       efx_for_each_rx_queue(rx_queue, efx) {
-               rx_queue->flushed = false;
-               falcon_flush_rx_queue(rx_queue);
-       }
-
-       /* Poll the evq looking for flush completions. Since we're not pushing
-        * any more rx or tx descriptors at this point, we're in no danger of
-        * overflowing the evq whilst we wait */
-       for (i = 0; i < FALCON_FLUSH_POLL_COUNT; ++i) {
-               msleep(FALCON_FLUSH_INTERVAL);
-               falcon_poll_flush_events(efx);
-
-               /* Check if every queue has been succesfully flushed */
-               outstanding = false;
-               efx_for_each_tx_queue(tx_queue, efx)
-                       outstanding |= !tx_queue->flushed;
-               efx_for_each_rx_queue(rx_queue, efx)
-                       outstanding |= !rx_queue->flushed;
-               if (!outstanding)
-                       return 0;
-       }
-
-       /* Mark the queues as all flushed. We're going to return failure
-        * leading to a reset, or fake up success anyway. "flushed" now
-        * indicates that we tried to flush. */
-       efx_for_each_tx_queue(tx_queue, efx) {
-               if (!tx_queue->flushed)
-                       EFX_ERR(efx, "tx queue %d flush command timed out\n",
-                               tx_queue->queue);
-               tx_queue->flushed = true;
-       }
-       efx_for_each_rx_queue(rx_queue, efx) {
-               if (!rx_queue->flushed)
-                       EFX_ERR(efx, "rx queue %d flush command timed out\n",
-                               rx_queue->queue);
-               rx_queue->flushed = true;
-       }
-
-       if (EFX_WORKAROUND_7803(efx))
-               return 0;
-
-       return -ETIMEDOUT;
 }
 
-/**************************************************************************
- *
- * Falcon hardware interrupts
- * The hardware interrupt handler does very little work; all the event
- * queue processing is carried out by per-channel tasklets.
- *
- **************************************************************************/
-
-/* Enable/disable/generate Falcon interrupts */
-static inline void falcon_interrupts(struct efx_nic *efx, int enabled,
-                                    int force)
-{
-       efx_oword_t int_en_reg_ker;
-
-       EFX_POPULATE_OWORD_2(int_en_reg_ker,
-                            FRF_AZ_KER_INT_KER, force,
-                            FRF_AZ_DRV_INT_EN_KER, enabled);
-       efx_writeo(efx, &int_en_reg_ker, FR_AZ_INT_EN_KER);
-}
-
-void falcon_enable_interrupts(struct efx_nic *efx)
-{
-       efx_oword_t int_adr_reg_ker;
-       struct efx_channel *channel;
-
-       EFX_ZERO_OWORD(*((efx_oword_t *) efx->irq_status.addr));
-       wmb(); /* Ensure interrupt vector is clear before interrupts enabled */
-
-       /* Program address */
-       EFX_POPULATE_OWORD_2(int_adr_reg_ker,
-                            FRF_AZ_NORM_INT_VEC_DIS_KER,
-                            EFX_INT_MODE_USE_MSI(efx),
-                            FRF_AZ_INT_ADR_KER, efx->irq_status.dma_addr);
-       efx_writeo(efx, &int_adr_reg_ker, FR_AZ_INT_ADR_KER);
-
-       /* Enable interrupts */
-       falcon_interrupts(efx, 1, 0);
-
-       /* Force processing of all the channels to get the EVQ RPTRs up to
-          date */
-       efx_for_each_channel(channel, efx)
-               efx_schedule_channel(channel);
-}
+static void falcon_deconfigure_mac_wrapper(struct efx_nic *efx);
 
-void falcon_disable_interrupts(struct efx_nic *efx)
+static void falcon_prepare_flush(struct efx_nic *efx)
 {
-       /* Disable interrupts */
-       falcon_interrupts(efx, 0, 0);
-}
+       falcon_deconfigure_mac_wrapper(efx);
 
-/* Generate a Falcon test interrupt
- * Interrupt must already have been enabled, otherwise nasty things
- * may happen.
- */
-void falcon_generate_interrupt(struct efx_nic *efx)
-{
-       falcon_interrupts(efx, 1, 1);
+       /* Wait for the tx and rx fifo's to get to the next packet boundary
+        * (~1ms without back-pressure), then to drain the remainder of the
+        * fifo's at data path speeds (negligible), with a healthy margin. */
+       msleep(10);
 }
 
 /* Acknowledge a legacy interrupt from Falcon
@@ -1338,7 +145,7 @@ void falcon_generate_interrupt(struct efx_nic *efx)
  *
  * NB most hardware supports MSI interrupts
  */
-static inline void falcon_irq_ack_a1(struct efx_nic *efx)
+inline void falcon_irq_ack_a1(struct efx_nic *efx)
 {
        efx_dword_t reg;
 
@@ -1347,104 +154,8 @@ static inline void falcon_irq_ack_a1(struct efx_nic *efx)
        efx_readd(efx, &reg, FR_AA_WORK_AROUND_BROKEN_PCI_READS);
 }
 
-/* Process a fatal interrupt
- * Disable bus mastering ASAP and schedule a reset
- */
-static irqreturn_t falcon_fatal_interrupt(struct efx_nic *efx)
-{
-       struct falcon_nic_data *nic_data = efx->nic_data;
-       efx_oword_t *int_ker = efx->irq_status.addr;
-       efx_oword_t fatal_intr;
-       int error, mem_perr;
-
-       efx_reado(efx, &fatal_intr, FR_AZ_FATAL_INTR_KER);
-       error = EFX_OWORD_FIELD(fatal_intr, FRF_AZ_FATAL_INTR);
-
-       EFX_ERR(efx, "SYSTEM ERROR " EFX_OWORD_FMT " status "
-               EFX_OWORD_FMT ": %s\n", EFX_OWORD_VAL(*int_ker),
-               EFX_OWORD_VAL(fatal_intr),
-               error ? "disabling bus mastering" : "no recognised error");
-       if (error == 0)
-               goto out;
-
-       /* If this is a memory parity error dump which blocks are offending */
-       mem_perr = EFX_OWORD_FIELD(fatal_intr, FRF_AZ_MEM_PERR_INT_KER);
-       if (mem_perr) {
-               efx_oword_t reg;
-               efx_reado(efx, &reg, FR_AZ_MEM_STAT);
-               EFX_ERR(efx, "SYSTEM ERROR: memory parity error "
-                       EFX_OWORD_FMT "\n", EFX_OWORD_VAL(reg));
-       }
-
-       /* Disable both devices */
-       pci_clear_master(efx->pci_dev);
-       if (FALCON_IS_DUAL_FUNC(efx))
-               pci_clear_master(nic_data->pci_dev2);
-       falcon_disable_interrupts(efx);
-
-       /* Count errors and reset or disable the NIC accordingly */
-       if (efx->int_error_count == 0 ||
-           time_after(jiffies, efx->int_error_expire)) {
-               efx->int_error_count = 0;
-               efx->int_error_expire =
-                       jiffies + FALCON_INT_ERROR_EXPIRE * HZ;
-       }
-       if (++efx->int_error_count < FALCON_MAX_INT_ERRORS) {
-               EFX_ERR(efx, "SYSTEM ERROR - reset scheduled\n");
-               efx_schedule_reset(efx, RESET_TYPE_INT_ERROR);
-       } else {
-               EFX_ERR(efx, "SYSTEM ERROR - max number of errors seen."
-                       "NIC will be disabled\n");
-               efx_schedule_reset(efx, RESET_TYPE_DISABLE);
-       }
-out:
-       return IRQ_HANDLED;
-}
-
-/* Handle a legacy interrupt from Falcon
- * Acknowledges the interrupt and schedule event queue processing.
- */
-static irqreturn_t falcon_legacy_interrupt_b0(int irq, void *dev_id)
-{
-       struct efx_nic *efx = dev_id;
-       efx_oword_t *int_ker = efx->irq_status.addr;
-       irqreturn_t result = IRQ_NONE;
-       struct efx_channel *channel;
-       efx_dword_t reg;
-       u32 queues;
-       int syserr;
-
-       /* Read the ISR which also ACKs the interrupts */
-       efx_readd(efx, &reg, FR_BZ_INT_ISR0);
-       queues = EFX_EXTRACT_DWORD(reg, 0, 31);
-
-       /* Check to see if we have a serious error condition */
-       syserr = EFX_OWORD_FIELD(*int_ker, FSF_AZ_NET_IVEC_FATAL_INT);
-       if (unlikely(syserr))
-               return falcon_fatal_interrupt(efx);
-
-       /* Schedule processing of any interrupting queues */
-       efx_for_each_channel(channel, efx) {
-               if ((queues & 1) ||
-                   falcon_event_present(
-                           falcon_event(channel, channel->eventq_read_ptr))) {
-                       efx_schedule_channel(channel);
-                       result = IRQ_HANDLED;
-               }
-               queues >>= 1;
-       }
-
-       if (result == IRQ_HANDLED) {
-               efx->last_irq_cpu = raw_smp_processor_id();
-               EFX_TRACE(efx, "IRQ %d on CPU %d status " EFX_DWORD_FMT "\n",
-                         irq, raw_smp_processor_id(), EFX_DWORD_VAL(reg));
-       }
-
-       return result;
-}
-
 
-static irqreturn_t falcon_legacy_interrupt_a1(int irq, void *dev_id)
+irqreturn_t falcon_legacy_interrupt_a1(int irq, void *dev_id)
 {
        struct efx_nic *efx = dev_id;
        efx_oword_t *int_ker = efx->irq_status.addr;
@@ -1467,13 +178,13 @@ static irqreturn_t falcon_legacy_interrupt_a1(int irq, void *dev_id)
        /* Check to see if we have a serious error condition */
        syserr = EFX_OWORD_FIELD(*int_ker, FSF_AZ_NET_IVEC_FATAL_INT);
        if (unlikely(syserr))
-               return falcon_fatal_interrupt(efx);
+               return efx_nic_fatal_interrupt(efx);
 
        /* Determine interrupting queues, clear interrupt status
         * register and acknowledge the device interrupt.
         */
-       BUILD_BUG_ON(INT_EVQS_WIDTH > EFX_MAX_CHANNELS);
-       queues = EFX_OWORD_FIELD(*int_ker, INT_EVQS);
+       BUILD_BUG_ON(FSF_AZ_NET_IVEC_INT_Q_WIDTH > EFX_MAX_CHANNELS);
+       queues = EFX_OWORD_FIELD(*int_ker, FSF_AZ_NET_IVEC_INT_Q);
        EFX_ZERO_OWORD(*int_ker);
        wmb(); /* Ensure the vector is cleared before interrupt ack */
        falcon_irq_ack_a1(efx);
@@ -1489,126 +200,6 @@ static irqreturn_t falcon_legacy_interrupt_a1(int irq, void *dev_id)
 
        return IRQ_HANDLED;
 }
-
-/* Handle an MSI interrupt from Falcon
- *
- * Handle an MSI hardware interrupt.  This routine schedules event
- * queue processing.  No interrupt acknowledgement cycle is necessary.
- * Also, we never need to check that the interrupt is for us, since
- * MSI interrupts cannot be shared.
- */
-static irqreturn_t falcon_msi_interrupt(int irq, void *dev_id)
-{
-       struct efx_channel *channel = dev_id;
-       struct efx_nic *efx = channel->efx;
-       efx_oword_t *int_ker = efx->irq_status.addr;
-       int syserr;
-
-       efx->last_irq_cpu = raw_smp_processor_id();
-       EFX_TRACE(efx, "IRQ %d on CPU %d status " EFX_OWORD_FMT "\n",
-                 irq, raw_smp_processor_id(), EFX_OWORD_VAL(*int_ker));
-
-       /* Check to see if we have a serious error condition */
-       syserr = EFX_OWORD_FIELD(*int_ker, FATAL_INT);
-       if (unlikely(syserr))
-               return falcon_fatal_interrupt(efx);
-
-       /* Schedule processing of the channel */
-       efx_schedule_channel(channel);
-
-       return IRQ_HANDLED;
-}
-
-
-/* Setup RSS indirection table.
- * This maps from the hash value of the packet to RXQ
- */
-static void falcon_setup_rss_indir_table(struct efx_nic *efx)
-{
-       int i = 0;
-       unsigned long offset;
-       efx_dword_t dword;
-
-       if (falcon_rev(efx) < FALCON_REV_B0)
-               return;
-
-       for (offset = FR_BZ_RX_INDIRECTION_TBL;
-            offset < FR_BZ_RX_INDIRECTION_TBL + 0x800;
-            offset += 0x10) {
-               EFX_POPULATE_DWORD_1(dword, FRF_BZ_IT_QUEUE,
-                                    i % efx->n_rx_queues);
-               efx_writed(efx, &dword, offset);
-               i++;
-       }
-}
-
-/* Hook interrupt handler(s)
- * Try MSI and then legacy interrupts.
- */
-int falcon_init_interrupt(struct efx_nic *efx)
-{
-       struct efx_channel *channel;
-       int rc;
-
-       if (!EFX_INT_MODE_USE_MSI(efx)) {
-               irq_handler_t handler;
-               if (falcon_rev(efx) >= FALCON_REV_B0)
-                       handler = falcon_legacy_interrupt_b0;
-               else
-                       handler = falcon_legacy_interrupt_a1;
-
-               rc = request_irq(efx->legacy_irq, handler, IRQF_SHARED,
-                                efx->name, efx);
-               if (rc) {
-                       EFX_ERR(efx, "failed to hook legacy IRQ %d\n",
-                               efx->pci_dev->irq);
-                       goto fail1;
-               }
-               return 0;
-       }
-
-       /* Hook MSI or MSI-X interrupt */
-       efx_for_each_channel(channel, efx) {
-               rc = request_irq(channel->irq, falcon_msi_interrupt,
-                                IRQF_PROBE_SHARED, /* Not shared */
-                                channel->name, channel);
-               if (rc) {
-                       EFX_ERR(efx, "failed to hook IRQ %d\n", channel->irq);
-                       goto fail2;
-               }
-       }
-
-       return 0;
-
- fail2:
-       efx_for_each_channel(channel, efx)
-               free_irq(channel->irq, channel);
- fail1:
-       return rc;
-}
-
-void falcon_fini_interrupt(struct efx_nic *efx)
-{
-       struct efx_channel *channel;
-       efx_oword_t reg;
-
-       /* Disable MSI/MSI-X interrupts */
-       efx_for_each_channel(channel, efx) {
-               if (channel->irq)
-                       free_irq(channel->irq, channel);
-       }
-
-       /* ACK legacy interrupt */
-       if (falcon_rev(efx) >= FALCON_REV_B0)
-               efx_reado(efx, &reg, FR_BZ_INT_ISR0);
-       else
-               falcon_irq_ack_a1(efx);
-
-       /* Disable legacy interrupt */
-       if (efx->legacy_irq)
-               free_irq(efx->legacy_irq, efx);
-}
-
 /**************************************************************************
  *
  * EEPROM/flash
@@ -1652,11 +243,10 @@ static int falcon_spi_wait(struct efx_nic *efx)
        }
 }
 
-int falcon_spi_cmd(const struct efx_spi_device *spi,
+int falcon_spi_cmd(struct efx_nic *efx, const struct efx_spi_device *spi,
                   unsigned int command, int address,
                   const void *in, void *out, size_t len)
 {
-       struct efx_nic *efx = spi->efx;
        bool addressed = (address >= 0);
        bool reading = (out != NULL);
        efx_oword_t reg;
@@ -1725,15 +315,15 @@ efx_spi_munge_command(const struct efx_spi_device *spi,
 }
 
 /* Wait up to 10 ms for buffered write completion */
-int falcon_spi_wait_write(const struct efx_spi_device *spi)
+int
+falcon_spi_wait_write(struct efx_nic *efx, const struct efx_spi_device *spi)
 {
-       struct efx_nic *efx = spi->efx;
        unsigned long timeout = jiffies + 1 + DIV_ROUND_UP(HZ, 100);
        u8 status;
        int rc;
 
        for (;;) {
-               rc = falcon_spi_cmd(spi, SPI_RDSR, -1, NULL,
+               rc = falcon_spi_cmd(efx, spi, SPI_RDSR, -1, NULL,
                                    &status, sizeof(status));
                if (rc)
                        return rc;
@@ -1749,8 +339,8 @@ int falcon_spi_wait_write(const struct efx_spi_device *spi)
        }
 }
 
-int falcon_spi_read(const struct efx_spi_device *spi, loff_t start,
-                   size_t len, size_t *retlen, u8 *buffer)
+int falcon_spi_read(struct efx_nic *efx, const struct efx_spi_device *spi,
+                   loff_t start, size_t len, size_t *retlen, u8 *buffer)
 {
        size_t block_len, pos = 0;
        unsigned int command;
@@ -1760,7 +350,7 @@ int falcon_spi_read(const struct efx_spi_device *spi, loff_t start,
                block_len = min(len - pos, FALCON_SPI_MAX_LEN);
 
                command = efx_spi_munge_command(spi, SPI_READ, start + pos);
-               rc = falcon_spi_cmd(spi, command, start + pos, NULL,
+               rc = falcon_spi_cmd(efx, spi, command, start + pos, NULL,
                                    buffer + pos, block_len);
                if (rc)
                        break;
@@ -1779,8 +369,9 @@ int falcon_spi_read(const struct efx_spi_device *spi, loff_t start,
        return rc;
 }
 
-int falcon_spi_write(const struct efx_spi_device *spi, loff_t start,
-                    size_t len, size_t *retlen, const u8 *buffer)
+int
+falcon_spi_write(struct efx_nic *efx, const struct efx_spi_device *spi,
+                loff_t start, size_t len, size_t *retlen, const u8 *buffer)
 {
        u8 verify_buffer[FALCON_SPI_MAX_LEN];
        size_t block_len, pos = 0;
@@ -1788,24 +379,24 @@ int falcon_spi_write(const struct efx_spi_device *spi, loff_t start,
        int rc = 0;
 
        while (pos < len) {
-               rc = falcon_spi_cmd(spi, SPI_WREN, -1, NULL, NULL, 0);
+               rc = falcon_spi_cmd(efx, spi, SPI_WREN, -1, NULL, NULL, 0);
                if (rc)
                        break;
 
                block_len = min(len - pos,
                                falcon_spi_write_limit(spi, start + pos));
                command = efx_spi_munge_command(spi, SPI_WRITE, start + pos);
-               rc = falcon_spi_cmd(spi, command, start + pos,
+               rc = falcon_spi_cmd(efx, spi, command, start + pos,
                                    buffer + pos, NULL, block_len);
                if (rc)
                        break;
 
-               rc = falcon_spi_wait_write(spi);
+               rc = falcon_spi_wait_write(efx, spi);
                if (rc)
                        break;
 
                command = efx_spi_munge_command(spi, SPI_READ, start + pos);
-               rc = falcon_spi_cmd(spi, command, start + pos,
+               rc = falcon_spi_cmd(efx, spi, command, start + pos,
                                    NULL, verify_buffer, block_len);
                if (memcmp(verify_buffer, buffer + pos, block_len)) {
                        rc = -EIO;
@@ -1834,12 +425,23 @@ int falcon_spi_write(const struct efx_spi_device *spi, loff_t start,
  **************************************************************************
  */
 
-static int falcon_reset_macs(struct efx_nic *efx)
+static void falcon_push_multicast_hash(struct efx_nic *efx)
 {
-       efx_oword_t reg;
+       union efx_multicast_hash *mc_hash = &efx->multicast_hash;
+
+       WARN_ON(!mutex_is_locked(&efx->mac_lock));
+
+       efx_writeo(efx, &mc_hash->oword[0], FR_AB_MAC_MC_HASH_REG0);
+       efx_writeo(efx, &mc_hash->oword[1], FR_AB_MAC_MC_HASH_REG1);
+}
+
+static void falcon_reset_macs(struct efx_nic *efx)
+{
+       struct falcon_nic_data *nic_data = efx->nic_data;
+       efx_oword_t reg, mac_ctrl;
        int count;
 
-       if (falcon_rev(efx) < FALCON_REV_B0) {
+       if (efx_nic_rev(efx) < EFX_REV_FALCON_B0) {
                /* It's not safe to use GLB_CTL_REG to reset the
                 * macs, so instead use the internal MAC resets
                 */
@@ -1851,7 +453,7 @@ static int falcon_reset_macs(struct efx_nic *efx)
                        EFX_POPULATE_OWORD_1(reg, FRF_AB_GM_SW_RST, 0);
                        efx_writeo(efx, &reg, FR_AB_GM_CFG1);
                        udelay(1000);
-                       return 0;
+                       return;
                } else {
                        EFX_POPULATE_OWORD_1(reg, FRF_AB_XM_CORE_RST, 1);
                        efx_writeo(efx, &reg, FR_AB_XM_GLB_CFG);
@@ -1860,22 +462,20 @@ static int falcon_reset_macs(struct efx_nic *efx)
                                efx_reado(efx, &reg, FR_AB_XM_GLB_CFG);
                                if (EFX_OWORD_FIELD(reg, FRF_AB_XM_CORE_RST) ==
                                    0)
-                                       return 0;
+                                       return;
                                udelay(10);
                        }
 
                        EFX_ERR(efx, "timed out waiting for XMAC core reset\n");
-                       return -ETIMEDOUT;
                }
        }
 
-       /* MAC stats will fail whilst the TX fifo is draining. Serialise
-        * the drain sequence with the statistics fetch */
-       efx_stats_disable(efx);
+       /* Mac stats will fail whist the TX fifo is draining */
+       WARN_ON(nic_data->stats_disable_count == 0);
 
-       efx_reado(efx, &reg, FR_AB_MAC_CTRL);
-       EFX_SET_OWORD_FIELD(reg, FRF_BB_TXFIFO_DRAIN_EN, 1);
-       efx_writeo(efx, &reg, FR_AB_MAC_CTRL);
+       efx_reado(efx, &mac_ctrl, FR_AB_MAC_CTRL);
+       EFX_SET_OWORD_FIELD(mac_ctrl, FRF_BB_TXFIFO_DRAIN_EN, 1);
+       efx_writeo(efx, &mac_ctrl, FR_AB_MAC_CTRL);
 
        efx_reado(efx, &reg, FR_AB_GLB_CTL);
        EFX_SET_OWORD_FIELD(reg, FRF_AB_RST_XGTX, 1);
@@ -1901,21 +501,16 @@ static int falcon_reset_macs(struct efx_nic *efx)
                udelay(10);
        }
 
-       efx_stats_enable(efx);
-
-       /* If we've reset the EM block and the link is up, then
-        * we'll have to kick the XAUI link so the PHY can recover */
-       if (efx->link_up && EFX_IS10G(efx) && EFX_WORKAROUND_5147(efx))
-               falcon_reset_xaui(efx);
-
-       return 0;
+       /* Ensure the correct MAC is selected before statistics
+        * are re-enabled by the caller */
+       efx_writeo(efx, &mac_ctrl, FR_AB_MAC_CTRL);
 }
 
 void falcon_drain_tx_fifo(struct efx_nic *efx)
 {
        efx_oword_t reg;
 
-       if ((falcon_rev(efx) < FALCON_REV_B0) ||
+       if ((efx_nic_rev(efx) < EFX_REV_FALCON_B0) ||
            (efx->loopback_mode != LOOPBACK_NONE))
                return;
 
@@ -1927,11 +522,11 @@ void falcon_drain_tx_fifo(struct efx_nic *efx)
        falcon_reset_macs(efx);
 }
 
-void falcon_deconfigure_mac_wrapper(struct efx_nic *efx)
+static void falcon_deconfigure_mac_wrapper(struct efx_nic *efx)
 {
        efx_oword_t reg;
 
-       if (falcon_rev(efx) < FALCON_REV_B0)
+       if (efx_nic_rev(efx) < EFX_REV_FALCON_B0)
                return;
 
        /* Isolate the MAC -> RX */
@@ -1939,17 +534,17 @@ void falcon_deconfigure_mac_wrapper(struct efx_nic *efx)
        EFX_SET_OWORD_FIELD(reg, FRF_BZ_RX_INGR_EN, 0);
        efx_writeo(efx, &reg, FR_AZ_RX_CFG);
 
-       if (!efx->link_up)
-               falcon_drain_tx_fifo(efx);
+       /* Isolate TX -> MAC */
+       falcon_drain_tx_fifo(efx);
 }
 
 void falcon_reconfigure_mac_wrapper(struct efx_nic *efx)
 {
+       struct efx_link_state *link_state = &efx->link_state;
        efx_oword_t reg;
        int link_speed;
-       bool tx_fc;
 
-       switch (efx->link_speed) {
+       switch (link_state->speed) {
        case 10000: link_speed = 3; break;
        case 1000:  link_speed = 2; break;
        case 100:   link_speed = 1; break;
@@ -1967,48 +562,39 @@ void falcon_reconfigure_mac_wrapper(struct efx_nic *efx)
                             FRF_AB_MAC_SPEED, link_speed);
        /* On B0, MAC backpressure can be disabled and packets get
         * discarded. */
-       if (falcon_rev(efx) >= FALCON_REV_B0) {
+       if (efx_nic_rev(efx) >= EFX_REV_FALCON_B0) {
                EFX_SET_OWORD_FIELD(reg, FRF_BB_TXFIFO_DRAIN_EN,
-                                   !efx->link_up);
+                                   !link_state->up);
        }
 
        efx_writeo(efx, &reg, FR_AB_MAC_CTRL);
 
        /* Restore the multicast hash registers. */
-       falcon_set_multicast_hash(efx);
+       falcon_push_multicast_hash(efx);
 
-       /* Transmission of pause frames when RX crosses the threshold is
-        * covered by RX_XOFF_MAC_EN and XM_TX_CFG_REG:XM_FCNTL.
-        * Action on receipt of pause frames is controller by XM_DIS_FCNTL */
-       tx_fc = !!(efx->link_fc & EFX_FC_TX);
        efx_reado(efx, &reg, FR_AZ_RX_CFG);
-       EFX_SET_OWORD_FIELD(reg, FRF_AZ_RX_XOFF_MAC_EN, tx_fc);
-
+       /* Enable XOFF signal from RX FIFO (we enabled it during NIC
+        * initialisation but it may read back as 0) */
+       EFX_SET_OWORD_FIELD(reg, FRF_AZ_RX_XOFF_MAC_EN, 1);
        /* Unisolate the MAC -> RX */
-       if (falcon_rev(efx) >= FALCON_REV_B0)
+       if (efx_nic_rev(efx) >= EFX_REV_FALCON_B0)
                EFX_SET_OWORD_FIELD(reg, FRF_BZ_RX_INGR_EN, 1);
        efx_writeo(efx, &reg, FR_AZ_RX_CFG);
 }
 
-int falcon_dma_stats(struct efx_nic *efx, unsigned int done_offset)
+static void falcon_stats_request(struct efx_nic *efx)
 {
+       struct falcon_nic_data *nic_data = efx->nic_data;
        efx_oword_t reg;
-       u32 *dma_done;
-       int i;
 
-       if (disable_dma_stats)
-               return 0;
+       WARN_ON(nic_data->stats_pending);
+       WARN_ON(nic_data->stats_disable_count);
 
-       /* Statistics fetch will fail if the MAC is in TX drain */
-       if (falcon_rev(efx) >= FALCON_REV_B0) {
-               efx_oword_t temp;
-               efx_reado(efx, &temp, FR_AB_MAC_CTRL);
-               if (EFX_OWORD_FIELD(temp, FRF_BB_TXFIFO_DRAIN_EN))
-                       return 0;
-       }
+       if (nic_data->stats_dma_done == NULL)
+               return; /* no mac selected */
 
-       dma_done = (efx->stats_buffer.addr + done_offset);
-       *dma_done = FALCON_STATS_NOT_DONE;
+       *nic_data->stats_dma_done = FALCON_STATS_NOT_DONE;
+       nic_data->stats_pending = true;
        wmb(); /* ensure done flag is clear */
 
        /* Initiate DMA transfer of stats */
@@ -2018,17 +604,90 @@ int falcon_dma_stats(struct efx_nic *efx, unsigned int done_offset)
                             efx->stats_buffer.dma_addr);
        efx_writeo(efx, &reg, FR_AB_MAC_STAT_DMA);
 
-       /* Wait for transfer to complete */
-       for (i = 0; i < 400; i++) {
-               if (*(volatile u32 *)dma_done == FALCON_STATS_DONE) {
-                       rmb(); /* Ensure the stats are valid. */
-                       return 0;
-               }
-               udelay(10);
+       mod_timer(&nic_data->stats_timer, round_jiffies_up(jiffies + HZ / 2));
+}
+
+static void falcon_stats_complete(struct efx_nic *efx)
+{
+       struct falcon_nic_data *nic_data = efx->nic_data;
+
+       if (!nic_data->stats_pending)
+               return;
+
+       nic_data->stats_pending = 0;
+       if (*nic_data->stats_dma_done == FALCON_STATS_DONE) {
+               rmb(); /* read the done flag before the stats */
+               efx->mac_op->update_stats(efx);
+       } else {
+               EFX_ERR(efx, "timed out waiting for statistics\n");
        }
+}
 
-       EFX_ERR(efx, "timed out waiting for statistics\n");
-       return -ETIMEDOUT;
+static void falcon_stats_timer_func(unsigned long context)
+{
+       struct efx_nic *efx = (struct efx_nic *)context;
+       struct falcon_nic_data *nic_data = efx->nic_data;
+
+       spin_lock(&efx->stats_lock);
+
+       falcon_stats_complete(efx);
+       if (nic_data->stats_disable_count == 0)
+               falcon_stats_request(efx);
+
+       spin_unlock(&efx->stats_lock);
+}
+
+static void falcon_switch_mac(struct efx_nic *efx);
+
+static bool falcon_loopback_link_poll(struct efx_nic *efx)
+{
+       struct efx_link_state old_state = efx->link_state;
+
+       WARN_ON(!mutex_is_locked(&efx->mac_lock));
+       WARN_ON(!LOOPBACK_INTERNAL(efx));
+
+       efx->link_state.fd = true;
+       efx->link_state.fc = efx->wanted_fc;
+       efx->link_state.up = true;
+
+       if (efx->loopback_mode == LOOPBACK_GMAC)
+               efx->link_state.speed = 1000;
+       else
+               efx->link_state.speed = 10000;
+
+       return !efx_link_state_equal(&efx->link_state, &old_state);
+}
+
+static int falcon_reconfigure_port(struct efx_nic *efx)
+{
+       int rc;
+
+       WARN_ON(efx_nic_rev(efx) > EFX_REV_FALCON_B0);
+
+       /* Poll the PHY link state *before* reconfiguring it. This means we
+        * will pick up the correct speed (in loopback) to select the correct
+        * MAC.
+        */
+       if (LOOPBACK_INTERNAL(efx))
+               falcon_loopback_link_poll(efx);
+       else
+               efx->phy_op->poll(efx);
+
+       falcon_stop_nic_stats(efx);
+       falcon_deconfigure_mac_wrapper(efx);
+
+       falcon_switch_mac(efx);
+
+       efx->phy_op->reconfigure(efx);
+       rc = efx->mac_op->reconfigure(efx);
+       BUG_ON(rc);
+
+       falcon_start_nic_stats(efx);
+
+       /* Synchronise efx->link_state with the kernel */
+       efx_link_status_changed(efx);
+
+       return 0;
 }
 
 /**************************************************************************
@@ -2041,18 +700,18 @@ int falcon_dma_stats(struct efx_nic *efx, unsigned int done_offset)
 /* Wait for GMII access to complete */
 static int falcon_gmii_wait(struct efx_nic *efx)
 {
-       efx_dword_t md_stat;
+       efx_oword_t md_stat;
        int count;
 
        /* wait upto 50ms - taken max from datasheet */
        for (count = 0; count < 5000; count++) {
-               efx_readd(efx, &md_stat, FR_AB_MD_STAT);
-               if (EFX_DWORD_FIELD(md_stat, FRF_AB_MD_BSY) == 0) {
-                       if (EFX_DWORD_FIELD(md_stat, FRF_AB_MD_LNFL) != 0 ||
-                           EFX_DWORD_FIELD(md_stat, FRF_AB_MD_BSERR) != 0) {
+               efx_reado(efx, &md_stat, FR_AB_MD_STAT);
+               if (EFX_OWORD_FIELD(md_stat, FRF_AB_MD_BSY) == 0) {
+                       if (EFX_OWORD_FIELD(md_stat, FRF_AB_MD_LNFL) != 0 ||
+                           EFX_OWORD_FIELD(md_stat, FRF_AB_MD_BSERR) != 0) {
                                EFX_ERR(efx, "error from GMII access "
-                                       EFX_DWORD_FMT"\n",
-                                       EFX_DWORD_VAL(md_stat));
+                                       EFX_OWORD_FMT"\n",
+                                       EFX_OWORD_VAL(md_stat));
                                return -EIO;
                        }
                        return 0;
@@ -2074,7 +733,7 @@ static int falcon_mdio_write(struct net_device *net_dev,
        EFX_REGDUMP(efx, "writing MDIO %d register %d.%d with 0x%04x\n",
                    prtad, devad, addr, value);
 
-       spin_lock_bh(&efx->phy_lock);
+       mutex_lock(&efx->mdio_lock);
 
        /* Check MDIO not currently being accessed */
        rc = falcon_gmii_wait(efx);
@@ -2109,8 +768,8 @@ static int falcon_mdio_write(struct net_device *net_dev,
                udelay(10);
        }
 
- out:
-       spin_unlock_bh(&efx->phy_lock);
+out:
+       mutex_unlock(&efx->mdio_lock);
        return rc;
 }
 
@@ -2122,7 +781,7 @@ static int falcon_mdio_read(struct net_device *net_dev,
        efx_oword_t reg;
        int rc;
 
-       spin_lock_bh(&efx->phy_lock);
+       mutex_lock(&efx->mdio_lock);
 
        /* Check MDIO not currently being accessed */
        rc = falcon_gmii_wait(efx);
@@ -2158,39 +817,20 @@ static int falcon_mdio_read(struct net_device *net_dev,
                        prtad, devad, addr, rc);
        }
 
- out:
-       spin_unlock_bh(&efx->phy_lock);
+out:
+       mutex_unlock(&efx->mdio_lock);
        return rc;
 }
 
-int falcon_switch_mac(struct efx_nic *efx)
+static void falcon_clock_mac(struct efx_nic *efx)
 {
-       struct efx_mac_operations *old_mac_op = efx->mac_op;
-       efx_oword_t nic_stat;
        unsigned strap_val;
-       int rc = 0;
-
-       /* Don't try to fetch MAC stats while we're switching MACs */
-       efx_stats_disable(efx);
-
-       /* Internal loopbacks override the phy speed setting */
-       if (efx->loopback_mode == LOOPBACK_GMAC) {
-               efx->link_speed = 1000;
-               efx->link_fd = true;
-       } else if (LOOPBACK_INTERNAL(efx)) {
-               efx->link_speed = 10000;
-               efx->link_fd = true;
-       }
-
-       WARN_ON(!mutex_is_locked(&efx->mac_lock));
-       efx->mac_op = (EFX_IS10G(efx) ?
-                      &falcon_xmac_operations : &falcon_gmac_operations);
+       efx_oword_t nic_stat;
 
-       /* Always push the NIC_STAT_REG setting even if the mac hasn't
-        * changed, because this function is run post online reset */
+       /* Configure the NIC generated MAC clock correctly */
        efx_reado(efx, &nic_stat, FR_AB_NIC_STAT);
        strap_val = EFX_IS10G(efx) ? 5 : 3;
-       if (falcon_rev(efx) >= FALCON_REV_B0) {
+       if (efx_nic_rev(efx) >= EFX_REV_FALCON_B0) {
                EFX_SET_OWORD_FIELD(nic_stat, FRF_BB_EE_STRAP_EN, 1);
                EFX_SET_OWORD_FIELD(nic_stat, FRF_BB_EE_STRAP, strap_val);
                efx_writeo(efx, &nic_stat, FR_AB_NIC_STAT);
@@ -2200,22 +840,39 @@ int falcon_switch_mac(struct efx_nic *efx)
                BUG_ON(EFX_OWORD_FIELD(nic_stat, FRF_AB_STRAP_PINS) !=
                       strap_val);
        }
+}
+
+static void falcon_switch_mac(struct efx_nic *efx)
+{
+       struct efx_mac_operations *old_mac_op = efx->mac_op;
+       struct falcon_nic_data *nic_data = efx->nic_data;
+       unsigned int stats_done_offset;
+
+       WARN_ON(!mutex_is_locked(&efx->mac_lock));
+       WARN_ON(nic_data->stats_disable_count == 0);
+
+       efx->mac_op = (EFX_IS10G(efx) ?
+                      &falcon_xmac_operations : &falcon_gmac_operations);
+
+       if (EFX_IS10G(efx))
+               stats_done_offset = XgDmaDone_offset;
+       else
+               stats_done_offset = GDmaDone_offset;
+       nic_data->stats_dma_done = efx->stats_buffer.addr + stats_done_offset;
 
        if (old_mac_op == efx->mac_op)
-               goto out;
+               return;
+
+       falcon_clock_mac(efx);
 
        EFX_LOG(efx, "selected %cMAC\n", EFX_IS10G(efx) ? 'X' : 'G');
        /* Not all macs support a mac-level link state */
-       efx->mac_up = true;
-
-       rc = falcon_reset_macs(efx);
-out:
-       efx_stats_enable(efx);
-       return rc;
+       efx->xmac_poll_required = false;
+       falcon_reset_macs(efx);
 }
 
 /* This call is responsible for hooking in the MAC and PHY operations */
-int falcon_probe_port(struct efx_nic *efx)
+static int falcon_probe_port(struct efx_nic *efx)
 {
        int rc;
 
@@ -2237,29 +894,28 @@ int falcon_probe_port(struct efx_nic *efx)
                return -ENODEV;
        }
 
-       if (efx->phy_op->macs & EFX_XMAC)
-               efx->loopback_modes |= ((1 << LOOPBACK_XGMII) |
-                                       (1 << LOOPBACK_XGXS) |
-                                       (1 << LOOPBACK_XAUI));
-       if (efx->phy_op->macs & EFX_GMAC)
-               efx->loopback_modes |= (1 << LOOPBACK_GMAC);
-       efx->loopback_modes |= efx->phy_op->loopbacks;
-
-       /* Set up MDIO structure for PHY */
-       efx->mdio.mmds = efx->phy_op->mmds;
-       efx->mdio.mode_support = MDIO_SUPPORTS_C45 | MDIO_EMULATE_C22;
+       /* Fill out MDIO structure and loopback modes */
        efx->mdio.mdio_read = falcon_mdio_read;
        efx->mdio.mdio_write = falcon_mdio_write;
+       rc = efx->phy_op->probe(efx);
+       if (rc != 0)
+               return rc;
+
+       /* Initial assumption */
+       efx->link_state.speed = 10000;
+       efx->link_state.fd = true;
 
        /* Hardware flow ctrl. FalconA RX FIFO too small for pause generation */
-       if (falcon_rev(efx) >= FALCON_REV_B0)
+       if (efx_nic_rev(efx) >= EFX_REV_FALCON_B0)
                efx->wanted_fc = EFX_FC_RX | EFX_FC_TX;
        else
                efx->wanted_fc = EFX_FC_RX;
+       if (efx->mdio.mmds & MDIO_DEVS_AN)
+               efx->wanted_fc |= EFX_FC_AUTO;
 
        /* Allocate buffer for stats */
-       rc = falcon_alloc_buffer(efx, &efx->stats_buffer,
-                                FALCON_MAC_STATS_SIZE);
+       rc = efx_nic_alloc_buffer(efx, &efx->stats_buffer,
+                                 FALCON_MAC_STATS_SIZE);
        if (rc)
                return rc;
        EFX_LOG(efx, "stats buffer at %llx (virt %p phys %llx)\n",
@@ -2270,40 +926,20 @@ int falcon_probe_port(struct efx_nic *efx)
        return 0;
 }
 
-void falcon_remove_port(struct efx_nic *efx)
-{
-       falcon_free_buffer(efx, &efx->stats_buffer);
-}
-
-/**************************************************************************
- *
- * Multicast filtering
- *
- **************************************************************************
- */
-
-void falcon_set_multicast_hash(struct efx_nic *efx)
+static void falcon_remove_port(struct efx_nic *efx)
 {
-       union efx_multicast_hash *mc_hash = &efx->multicast_hash;
-
-       /* Broadcast packets go through the multicast hash filter.
-        * ether_crc_le() of the broadcast address is 0xbe2612ff
-        * so we always add bit 0xff to the mask.
-        */
-       set_bit_le(0xff, mc_hash->byte);
-
-       efx_writeo(efx, &mc_hash->oword[0], FR_AB_MAC_MC_HASH_REG0);
-       efx_writeo(efx, &mc_hash->oword[1], FR_AB_MAC_MC_HASH_REG1);
+       efx->phy_op->remove(efx);
+       efx_nic_free_buffer(efx, &efx->stats_buffer);
 }
 
-
 /**************************************************************************
  *
  * Falcon test code
  *
  **************************************************************************/
 
-int falcon_read_nvram(struct efx_nic *efx, struct falcon_nvconfig *nvconfig_out)
+static int
+falcon_read_nvram(struct efx_nic *efx, struct falcon_nvconfig *nvconfig_out)
 {
        struct falcon_nvconfig *nvconfig;
        struct efx_spi_device *spi;
@@ -2322,7 +958,7 @@ int falcon_read_nvram(struct efx_nic *efx, struct falcon_nvconfig *nvconfig_out)
        nvconfig = region + FALCON_NVCONFIG_OFFSET;
 
        mutex_lock(&efx->spi_lock);
-       rc = falcon_spi_read(spi, 0, FALCON_NVCONFIG_END, NULL, region);
+       rc = falcon_spi_read(efx, spi, 0, FALCON_NVCONFIG_END, NULL, region);
        mutex_unlock(&efx->spi_lock);
        if (rc) {
                EFX_ERR(efx, "Failed to read %s\n",
@@ -2366,13 +1002,14 @@ int falcon_read_nvram(struct efx_nic *efx, struct falcon_nvconfig *nvconfig_out)
        return rc;
 }
 
-/* Registers tested in the falcon register test */
-static struct {
-       unsigned address;
-       efx_oword_t mask;
-} efx_test_registers[] = {
+static int falcon_test_nvram(struct efx_nic *efx)
+{
+       return falcon_read_nvram(efx, NULL);
+}
+
+static const struct efx_nic_register_test falcon_b0_register_tests[] = {
        { FR_AZ_ADR_REGION,
-         EFX_OWORD32(0x0001FFFF, 0x0001FFFF, 0x0001FFFF, 0x0001FFFF) },
+         EFX_OWORD32(0x0003FFFF, 0x0003FFFF, 0x0003FFFF, 0x0003FFFF) },
        { FR_AZ_RX_CFG,
          EFX_OWORD32(0xFFFFFFFE, 0x00017FFF, 0x00000000, 0x00000000) },
        { FR_AZ_TX_CFG,
@@ -2409,64 +1046,10 @@ static struct {
          EFX_OWORD32(0x0003FF0F, 0x00000000, 0x00000000, 0x00000000) },
 };
 
-static bool efx_masked_compare_oword(const efx_oword_t *a, const efx_oword_t *b,
-                                    const efx_oword_t *mask)
-{
-       return ((a->u64[0] ^ b->u64[0]) & mask->u64[0]) ||
-               ((a->u64[1] ^ b->u64[1]) & mask->u64[1]);
-}
-
-int falcon_test_registers(struct efx_nic *efx)
+static int falcon_b0_test_registers(struct efx_nic *efx)
 {
-       unsigned address = 0, i, j;
-       efx_oword_t mask, imask, original, reg, buf;
-
-       /* Falcon should be in loopback to isolate the XMAC from the PHY */
-       WARN_ON(!LOOPBACK_INTERNAL(efx));
-
-       for (i = 0; i < ARRAY_SIZE(efx_test_registers); ++i) {
-               address = efx_test_registers[i].address;
-               mask = imask = efx_test_registers[i].mask;
-               EFX_INVERT_OWORD(imask);
-
-               efx_reado(efx, &original, address);
-
-               /* bit sweep on and off */
-               for (j = 0; j < 128; j++) {
-                       if (!EFX_EXTRACT_OWORD32(mask, j, j))
-                               continue;
-
-                       /* Test this testable bit can be set in isolation */
-                       EFX_AND_OWORD(reg, original, mask);
-                       EFX_SET_OWORD32(reg, j, j, 1);
-
-                       efx_writeo(efx, &reg, address);
-                       efx_reado(efx, &buf, address);
-
-                       if (efx_masked_compare_oword(&reg, &buf, &mask))
-                               goto fail;
-
-                       /* Test this testable bit can be cleared in isolation */
-                       EFX_OR_OWORD(reg, original, mask);
-                       EFX_SET_OWORD32(reg, j, j, 0);
-
-                       efx_writeo(efx, &reg, address);
-                       efx_reado(efx, &buf, address);
-
-                       if (efx_masked_compare_oword(&reg, &buf, &mask))
-                               goto fail;
-               }
-
-               efx_writeo(efx, &original, address);
-       }
-
-       return 0;
-
-fail:
-       EFX_ERR(efx, "wrote "EFX_OWORD_FMT" read "EFX_OWORD_FMT
-               " at address 0x%x mask "EFX_OWORD_FMT"\n", EFX_OWORD_VAL(reg),
-               EFX_OWORD_VAL(buf), address, EFX_OWORD_VAL(mask));
-       return -EIO;
+       return efx_nic_test_registers(efx, falcon_b0_register_tests,
+                                     ARRAY_SIZE(falcon_b0_register_tests));
 }
 
 /**************************************************************************
@@ -2478,13 +1061,13 @@ fail:
 
 /* Resets NIC to known state.  This routine must be called in process
  * context and is allowed to sleep. */
-int falcon_reset_hw(struct efx_nic *efx, enum reset_type method)
+static int falcon_reset_hw(struct efx_nic *efx, enum reset_type method)
 {
        struct falcon_nic_data *nic_data = efx->nic_data;
        efx_oword_t glb_ctl_reg_ker;
        int rc;
 
-       EFX_LOG(efx, "performing hardware reset (%d)\n", method);
+       EFX_LOG(efx, "performing %s hardware reset\n", RESET_TYPE(method));
 
        /* Initiate device reset */
        if (method == RESET_TYPE_WORLD) {
@@ -2494,7 +1077,7 @@ int falcon_reset_hw(struct efx_nic *efx, enum reset_type method)
                                "function prior to hardware reset\n");
                        goto fail1;
                }
-               if (FALCON_IS_DUAL_FUNC(efx)) {
+               if (efx_nic_is_dual_func(efx)) {
                        rc = pci_save_state(nic_data->pci_dev2);
                        if (rc) {
                                EFX_ERR(efx, "failed to backup PCI state of "
@@ -2529,7 +1112,7 @@ int falcon_reset_hw(struct efx_nic *efx, enum reset_type method)
 
        /* Restore PCI configuration if needed */
        if (method == RESET_TYPE_WORLD) {
-               if (FALCON_IS_DUAL_FUNC(efx)) {
+               if (efx_nic_is_dual_func(efx)) {
                        rc = pci_restore_state(nic_data->pci_dev2);
                        if (rc) {
                                EFX_ERR(efx, "failed to restore PCI config for "
@@ -2567,6 +1150,44 @@ fail5:
        return rc;
 }
 
+static void falcon_monitor(struct efx_nic *efx)
+{
+       bool link_changed;
+       int rc;
+
+       BUG_ON(!mutex_is_locked(&efx->mac_lock));
+
+       rc = falcon_board(efx)->type->monitor(efx);
+       if (rc) {
+               EFX_ERR(efx, "Board sensor %s; shutting down PHY\n",
+                       (rc == -ERANGE) ? "reported fault" : "failed");
+               efx->phy_mode |= PHY_MODE_LOW_POWER;
+               rc = __efx_reconfigure_port(efx);
+               WARN_ON(rc);
+       }
+
+       if (LOOPBACK_INTERNAL(efx))
+               link_changed = falcon_loopback_link_poll(efx);
+       else
+               link_changed = efx->phy_op->poll(efx);
+
+       if (link_changed) {
+               falcon_stop_nic_stats(efx);
+               falcon_deconfigure_mac_wrapper(efx);
+
+               falcon_switch_mac(efx);
+               rc = efx->mac_op->reconfigure(efx);
+               BUG_ON(rc);
+
+               falcon_start_nic_stats(efx);
+
+               efx_link_status_changed(efx);
+       }
+
+       if (EFX_IS10G(efx))
+               falcon_poll_xmac(efx);
+}
+
 /* Zeroes out the SRAM contents.  This routine must be called in
  * process context and is allowed to sleep.
  */
@@ -2633,8 +1254,6 @@ static int falcon_spi_device_init(struct efx_nic *efx,
                spi_device->block_size =
                        1 << SPI_DEV_TYPE_FIELD(device_type,
                                                SPI_DEV_TYPE_BLOCK_SIZE);
-
-               spi_device->efx = efx;
        } else {
                spi_device = NULL;
        }
@@ -2644,7 +1263,6 @@ static int falcon_spi_device_init(struct efx_nic *efx,
        return 0;
 }
 
-
 static void falcon_remove_spi_devices(struct efx_nic *efx)
 {
        kfree(efx->spi_eeprom);
@@ -2702,7 +1320,9 @@ static int falcon_probe_nvconfig(struct efx_nic *efx)
 
        EFX_LOG(efx, "PHY is %d phy_id %d\n", efx->phy_type, efx->mdio.prtad);
 
-       falcon_probe_board(efx, board_rev);
+       rc = falcon_probe_board(efx, board_rev);
+       if (rc)
+               goto fail2;
 
        kfree(nvconfig);
        return 0;
@@ -2714,49 +1334,6 @@ static int falcon_probe_nvconfig(struct efx_nic *efx)
        return rc;
 }
 
-/* Probe the NIC variant (revision, ASIC vs FPGA, function count, port
- * count, port speed).  Set workaround and feature flags accordingly.
- */
-static int falcon_probe_nic_variant(struct efx_nic *efx)
-{
-       efx_oword_t altera_build;
-       efx_oword_t nic_stat;
-
-       efx_reado(efx, &altera_build, FR_AZ_ALTERA_BUILD);
-       if (EFX_OWORD_FIELD(altera_build, FRF_AZ_ALTERA_BUILD_VER)) {
-               EFX_ERR(efx, "Falcon FPGA not supported\n");
-               return -ENODEV;
-       }
-
-       efx_reado(efx, &nic_stat, FR_AB_NIC_STAT);
-
-       switch (falcon_rev(efx)) {
-       case FALCON_REV_A0:
-       case 0xff:
-               EFX_ERR(efx, "Falcon rev A0 not supported\n");
-               return -ENODEV;
-
-       case FALCON_REV_A1:
-               if (EFX_OWORD_FIELD(nic_stat, FRF_AA_STRAP_PCIE) == 0) {
-                       EFX_ERR(efx, "Falcon rev A1 PCI-X not supported\n");
-                       return -ENODEV;
-               }
-               break;
-
-       case FALCON_REV_B0:
-               break;
-
-       default:
-               EFX_ERR(efx, "Unknown Falcon rev %d\n", falcon_rev(efx));
-               return -ENODEV;
-       }
-
-       /* Initial assumed speed */
-       efx->link_speed = EFX_OWORD_FIELD(nic_stat, FRF_AB_STRAP_10G) ? 10000 : 1000;
-
-       return 0;
-}
-
 /* Probe all SPI devices on the NIC */
 static void falcon_probe_spi_devices(struct efx_nic *efx)
 {
@@ -2796,7 +1373,7 @@ static void falcon_probe_spi_devices(struct efx_nic *efx)
                                       large_eeprom_type);
 }
 
-int falcon_probe_nic(struct efx_nic *efx)
+static int falcon_probe_nic(struct efx_nic *efx)
 {
        struct falcon_nic_data *nic_data;
        struct falcon_board *board;
@@ -2808,15 +1385,33 @@ int falcon_probe_nic(struct efx_nic *efx)
                return -ENOMEM;
        efx->nic_data = nic_data;
 
-       /* Determine number of ports etc. */
-       rc = falcon_probe_nic_variant(efx);
-       if (rc)
+       rc = -ENODEV;
+
+       if (efx_nic_fpga_ver(efx) != 0) {
+               EFX_ERR(efx, "Falcon FPGA not supported\n");
                goto fail1;
+       }
 
-       /* Probe secondary function if expected */
-       if (FALCON_IS_DUAL_FUNC(efx)) {
-               struct pci_dev *dev = pci_dev_get(efx->pci_dev);
+       if (efx_nic_rev(efx) <= EFX_REV_FALCON_A1) {
+               efx_oword_t nic_stat;
+               struct pci_dev *dev;
+               u8 pci_rev = efx->pci_dev->revision;
+
+               if ((pci_rev == 0xff) || (pci_rev == 0)) {
+                       EFX_ERR(efx, "Falcon rev A0 not supported\n");
+                       goto fail1;
+               }
+               efx_reado(efx, &nic_stat, FR_AB_NIC_STAT);
+               if (EFX_OWORD_FIELD(nic_stat, FRF_AB_STRAP_10G) == 0) {
+                       EFX_ERR(efx, "Falcon rev A1 1G not supported\n");
+                       goto fail1;
+               }
+               if (EFX_OWORD_FIELD(nic_stat, FRF_AA_STRAP_PCIE) == 0) {
+                       EFX_ERR(efx, "Falcon rev A1 PCI-X not supported\n");
+                       goto fail1;
+               }
 
+               dev = pci_dev_get(efx->pci_dev);
                while ((dev = pci_get_device(EFX_VENDID_SFC, FALCON_A_S_DEVID,
                                             dev))) {
                        if (dev->bus == efx->pci_dev->bus &&
@@ -2840,7 +1435,7 @@ int falcon_probe_nic(struct efx_nic *efx)
        }
 
        /* Allocate memory for INT_KER */
-       rc = falcon_alloc_buffer(efx, &efx->irq_status, sizeof(efx_oword_t));
+       rc = efx_nic_alloc_buffer(efx, &efx->irq_status, sizeof(efx_oword_t));
        if (rc)
                goto fail4;
        BUG_ON(efx->irq_status.dma_addr & 0x0f);
@@ -2869,12 +1464,16 @@ int falcon_probe_nic(struct efx_nic *efx)
        if (rc)
                goto fail5;
 
-       rc = falcon_board(efx)->init(efx);
+       rc = falcon_board(efx)->type->init(efx);
        if (rc) {
                EFX_ERR(efx, "failed to initialise board\n");
                goto fail6;
        }
 
+       nic_data->stats_disable_count = 1;
+       setup_timer(&nic_data->stats_timer, &falcon_stats_timer_func,
+                   (unsigned long)efx);
+
        return 0;
 
  fail6:
@@ -2882,7 +1481,7 @@ int falcon_probe_nic(struct efx_nic *efx)
        memset(&board->i2c_adap, 0, sizeof(board->i2c_adap));
  fail5:
        falcon_remove_spi_devices(efx);
-       falcon_free_buffer(efx, &efx->irq_status);
+       efx_nic_free_buffer(efx, &efx->irq_status);
  fail4:
  fail3:
        if (nic_data->pci_dev2) {
@@ -2905,12 +1504,12 @@ static void falcon_init_rx_cfg(struct efx_nic *efx)
        const unsigned ctrl_xon_thr = 20;
        const unsigned ctrl_xoff_thr = 25;
        /* RX data FIFO thresholds (256-byte units; size varies) */
-       int data_xon_thr = rx_xon_thresh_bytes >> 8;
-       int data_xoff_thr = rx_xoff_thresh_bytes >> 8;
+       int data_xon_thr = efx_nic_rx_xon_thresh >> 8;
+       int data_xoff_thr = efx_nic_rx_xoff_thresh >> 8;
        efx_oword_t reg;
 
        efx_reado(efx, &reg, FR_AZ_RX_CFG);
-       if (falcon_rev(efx) <= FALCON_REV_A1) {
+       if (efx_nic_rev(efx) <= EFX_REV_FALCON_A1) {
                /* Data FIFO size is 5.5K */
                if (data_xon_thr < 0)
                        data_xon_thr = 512 >> 8;
@@ -2938,6 +1537,9 @@ static void falcon_init_rx_cfg(struct efx_nic *efx)
                EFX_SET_OWORD_FIELD(reg, FRF_BZ_RX_XOFF_TX_TH, ctrl_xoff_thr);
                EFX_SET_OWORD_FIELD(reg, FRF_BZ_RX_INGR_EN, 1);
        }
+       /* Always enable XOFF signal from RX FIFO.  We enable
+        * or disable transmission of pause frames at the MAC. */
+       EFX_SET_OWORD_FIELD(reg, FRF_AZ_RX_XOFF_MAC_EN, 1);
        efx_writeo(efx, &reg, FR_AZ_RX_CFG);
 }
 
@@ -2945,7 +1547,7 @@ static void falcon_init_rx_cfg(struct efx_nic *efx)
  * defining the descriptor cache sizes and number of RSS channels.
  * It does not set up any buffers, descriptor rings or event queues.
  */
-int falcon_init_nic(struct efx_nic *efx)
+static int falcon_init_nic(struct efx_nic *efx)
 {
        efx_oword_t temp;
        int rc;
@@ -2956,36 +1558,19 @@ int falcon_init_nic(struct efx_nic *efx)
        efx_writeo(efx, &temp, FR_AB_NIC_STAT);
 
        /* Set the source of the GMAC clock */
-       if (falcon_rev(efx) == FALCON_REV_B0) {
+       if (efx_nic_rev(efx) == EFX_REV_FALCON_B0) {
                efx_reado(efx, &temp, FR_AB_GPIO_CTL);
                EFX_SET_OWORD_FIELD(temp, FRF_AB_USE_NIC_CLK, true);
                efx_writeo(efx, &temp, FR_AB_GPIO_CTL);
        }
 
+       /* Select the correct MAC */
+       falcon_clock_mac(efx);
+
        rc = falcon_reset_sram(efx);
        if (rc)
                return rc;
 
-       /* Set positions of descriptor caches in SRAM. */
-       EFX_POPULATE_OWORD_1(temp, FRF_AZ_SRM_TX_DC_BASE_ADR, TX_DC_BASE / 8);
-       efx_writeo(efx, &temp, FR_AZ_SRM_TX_DC_CFG);
-       EFX_POPULATE_OWORD_1(temp, FRF_AZ_SRM_RX_DC_BASE_ADR, RX_DC_BASE / 8);
-       efx_writeo(efx, &temp, FR_AZ_SRM_RX_DC_CFG);
-
-       /* Set TX descriptor cache size. */
-       BUILD_BUG_ON(TX_DC_ENTRIES != (16 << TX_DC_ENTRIES_ORDER));
-       EFX_POPULATE_OWORD_1(temp, FRF_AZ_TX_DC_SIZE, TX_DC_ENTRIES_ORDER);
-       efx_writeo(efx, &temp, FR_AZ_TX_DC_CFG);
-
-       /* Set RX descriptor cache size.  Set low watermark to size-8, as
-        * this allows most efficient prefetching.
-        */
-       BUILD_BUG_ON(RX_DC_ENTRIES != (16 << RX_DC_ENTRIES_ORDER));
-       EFX_POPULATE_OWORD_1(temp, FRF_AZ_RX_DC_SIZE, RX_DC_ENTRIES_ORDER);
-       efx_writeo(efx, &temp, FR_AZ_RX_DC_CFG);
-       EFX_POPULATE_OWORD_1(temp, FRF_AZ_RX_DC_PF_LWM, RX_DC_ENTRIES - 8);
-       efx_writeo(efx, &temp, FR_AZ_RX_DC_PF_WM);
-
        /* Clear the parity enables on the TX data fifos as
         * they produce false parity errors because of timing issues
         */
@@ -2995,19 +1580,6 @@ int falcon_init_nic(struct efx_nic *efx)
                efx_writeo(efx, &temp, FR_AZ_CSR_SPARE);
        }
 
-       /* Enable all the genuinely fatal interrupts.  (They are still
-        * masked by the overall interrupt mask, controlled by
-        * falcon_interrupts()).
-        *
-        * Note: All other fatal interrupts are enabled
-        */
-       EFX_POPULATE_OWORD_3(temp,
-                            FRF_AZ_ILL_ADR_INT_KER_EN, 1,
-                            FRF_AZ_RBUF_OWN_INT_KER_EN, 1,
-                            FRF_AZ_TBUF_OWN_INT_KER_EN, 1);
-       EFX_INVERT_OWORD(temp);
-       efx_writeo(efx, &temp, FR_AZ_FATAL_INTR_KER);
-
        if (EFX_WORKAROUND_7244(efx)) {
                efx_reado(efx, &temp, FR_BZ_RX_FILTER_CTL);
                EFX_SET_OWORD_FIELD(temp, FRF_BZ_UDP_FULL_SRCH_LIMIT, 8);
@@ -3017,8 +1589,6 @@ int falcon_init_nic(struct efx_nic *efx)
                efx_writeo(efx, &temp, FR_BZ_RX_FILTER_CTL);
        }
 
-       falcon_setup_rss_indir_table(efx);
-
        /* XXX This is documented only for Falcon A0/A1 */
        /* Setup RX.  Wait for descriptor is broken and must
         * be disabled.  RXDP recovery shouldn't be needed, but is.
@@ -3030,24 +1600,6 @@ int falcon_init_nic(struct efx_nic *efx)
                EFX_SET_OWORD_FIELD(temp, FRF_AA_RX_ISCSI_DIS, 1);
        efx_writeo(efx, &temp, FR_AA_RX_SELF_RST);
 
-       /* Disable the ugly timer-based TX DMA backoff and allow TX DMA to be
-        * controlled by the RX FIFO fill level. Set arbitration to one pkt/Q.
-        */
-       efx_reado(efx, &temp, FR_AZ_TX_RESERVED);
-       EFX_SET_OWORD_FIELD(temp, FRF_AZ_TX_RX_SPACER, 0xfe);
-       EFX_SET_OWORD_FIELD(temp, FRF_AZ_TX_RX_SPACER_EN, 1);
-       EFX_SET_OWORD_FIELD(temp, FRF_AZ_TX_ONE_PKT_PER_Q, 1);
-       EFX_SET_OWORD_FIELD(temp, FRF_AZ_TX_PUSH_EN, 0);
-       EFX_SET_OWORD_FIELD(temp, FRF_AZ_TX_DIS_NON_IP_EV, 1);
-       /* Enable SW_EV to inherit in char driver - assume harmless here */
-       EFX_SET_OWORD_FIELD(temp, FRF_AZ_TX_SOFT_EVT_EN, 1);
-       /* Prefetch threshold 2 => fetch when descriptor cache half empty */
-       EFX_SET_OWORD_FIELD(temp, FRF_AZ_TX_PREF_THRESHOLD, 2);
-       /* Squash TX of packets of 16 bytes or less */
-       if (falcon_rev(efx) >= FALCON_REV_B0 && EFX_WORKAROUND_9141(efx))
-               EFX_SET_OWORD_FIELD(temp, FRF_BZ_TX_FLUSH_MIN_LEN_EN, 1);
-       efx_writeo(efx, &temp, FR_AZ_TX_RESERVED);
-
        /* Do not enable TX_NO_EOP_DISC_EN, since it limits packets to 16
         * descriptors (which is bad).
         */
@@ -3058,21 +1610,23 @@ int falcon_init_nic(struct efx_nic *efx)
        falcon_init_rx_cfg(efx);
 
        /* Set destination of both TX and RX Flush events */
-       if (falcon_rev(efx) >= FALCON_REV_B0) {
+       if (efx_nic_rev(efx) >= EFX_REV_FALCON_B0) {
                EFX_POPULATE_OWORD_1(temp, FRF_BZ_FLS_EVQ_ID, 0);
                efx_writeo(efx, &temp, FR_BZ_DP_CTRL);
        }
 
+       efx_nic_init_common(efx);
+
        return 0;
 }
 
-void falcon_remove_nic(struct efx_nic *efx)
+static void falcon_remove_nic(struct efx_nic *efx)
 {
        struct falcon_nic_data *nic_data = efx->nic_data;
        struct falcon_board *board = falcon_board(efx);
        int rc;
 
-       falcon_board(efx)->fini(efx);
+       board->type->fini(efx);
 
        /* Remove I2C adapter and clear it in preparation for a retry */
        rc = i2c_del_adapter(&board->i2c_adap);
@@ -3080,7 +1634,7 @@ void falcon_remove_nic(struct efx_nic *efx)
        memset(&board->i2c_adap, 0, sizeof(board->i2c_adap));
 
        falcon_remove_spi_devices(efx);
-       falcon_free_buffer(efx, &efx->irq_status);
+       efx_nic_free_buffer(efx, &efx->irq_status);
 
        falcon_reset_hw(efx, RESET_TYPE_ALL);
 
@@ -3095,23 +1649,119 @@ void falcon_remove_nic(struct efx_nic *efx)
        efx->nic_data = NULL;
 }
 
-void falcon_update_nic_stats(struct efx_nic *efx)
+static void falcon_update_nic_stats(struct efx_nic *efx)
 {
+       struct falcon_nic_data *nic_data = efx->nic_data;
        efx_oword_t cnt;
 
+       if (nic_data->stats_disable_count)
+               return;
+
        efx_reado(efx, &cnt, FR_AZ_RX_NODESC_DROP);
        efx->n_rx_nodesc_drop_cnt +=
                EFX_OWORD_FIELD(cnt, FRF_AB_RX_NODESC_DROP_CNT);
+
+       if (nic_data->stats_pending &&
+           *nic_data->stats_dma_done == FALCON_STATS_DONE) {
+               nic_data->stats_pending = false;
+               rmb(); /* read the done flag before the stats */
+               efx->mac_op->update_stats(efx);
+       }
+}
+
+void falcon_start_nic_stats(struct efx_nic *efx)
+{
+       struct falcon_nic_data *nic_data = efx->nic_data;
+
+       spin_lock_bh(&efx->stats_lock);
+       if (--nic_data->stats_disable_count == 0)
+               falcon_stats_request(efx);
+       spin_unlock_bh(&efx->stats_lock);
+}
+
+void falcon_stop_nic_stats(struct efx_nic *efx)
+{
+       struct falcon_nic_data *nic_data = efx->nic_data;
+       int i;
+
+       might_sleep();
+
+       spin_lock_bh(&efx->stats_lock);
+       ++nic_data->stats_disable_count;
+       spin_unlock_bh(&efx->stats_lock);
+
+       del_timer_sync(&nic_data->stats_timer);
+
+       /* Wait enough time for the most recent transfer to
+        * complete. */
+       for (i = 0; i < 4 && nic_data->stats_pending; i++) {
+               if (*nic_data->stats_dma_done == FALCON_STATS_DONE)
+                       break;
+               msleep(1);
+       }
+
+       spin_lock_bh(&efx->stats_lock);
+       falcon_stats_complete(efx);
+       spin_unlock_bh(&efx->stats_lock);
+}
+
+static void falcon_set_id_led(struct efx_nic *efx, enum efx_led_mode mode)
+{
+       falcon_board(efx)->type->set_id_led(efx, mode);
+}
+
+/**************************************************************************
+ *
+ * Wake on LAN
+ *
+ **************************************************************************
+ */
+
+static void falcon_get_wol(struct efx_nic *efx, struct ethtool_wolinfo *wol)
+{
+       wol->supported = 0;
+       wol->wolopts = 0;
+       memset(&wol->sopass, 0, sizeof(wol->sopass));
+}
+
+static int falcon_set_wol(struct efx_nic *efx, u32 type)
+{
+       if (type != 0)
+               return -EINVAL;
+       return 0;
 }
 
 /**************************************************************************
  *
- * Revision-dependent attributes used by efx.c
+ * Revision-dependent attributes used by efx.c and nic.c
  *
  **************************************************************************
  */
 
-struct efx_nic_type falcon_a_nic_type = {
+struct efx_nic_type falcon_a1_nic_type = {
+       .probe = falcon_probe_nic,
+       .remove = falcon_remove_nic,
+       .init = falcon_init_nic,
+       .fini = efx_port_dummy_op_void,
+       .monitor = falcon_monitor,
+       .reset = falcon_reset_hw,
+       .probe_port = falcon_probe_port,
+       .remove_port = falcon_remove_port,
+       .prepare_flush = falcon_prepare_flush,
+       .update_stats = falcon_update_nic_stats,
+       .start_stats = falcon_start_nic_stats,
+       .stop_stats = falcon_stop_nic_stats,
+       .set_id_led = falcon_set_id_led,
+       .push_irq_moderation = falcon_push_irq_moderation,
+       .push_multicast_hash = falcon_push_multicast_hash,
+       .reconfigure_port = falcon_reconfigure_port,
+       .get_wol = falcon_get_wol,
+       .set_wol = falcon_set_wol,
+       .resume_wol = efx_port_dummy_op_void,
+       .test_nvram = falcon_test_nvram,
+       .default_mac_ops = &falcon_xmac_operations,
+
+       .revision = EFX_REV_FALCON_A1,
        .mem_map_size = 0x20000,
        .txd_ptr_tbl_base = FR_AA_TX_DESC_PTR_TBL_KER,
        .rxd_ptr_tbl_base = FR_AA_RX_DESC_PTR_TBL_KER,
@@ -3122,9 +1772,37 @@ struct efx_nic_type falcon_a_nic_type = {
        .rx_buffer_padding = 0x24,
        .max_interrupt_mode = EFX_INT_MODE_MSI,
        .phys_addr_channels = 4,
+       .tx_dc_base = 0x130000,
+       .rx_dc_base = 0x100000,
+       .offload_features = NETIF_F_IP_CSUM,
+       .reset_world_flags = ETH_RESET_IRQ,
 };
 
-struct efx_nic_type falcon_b_nic_type = {
+struct efx_nic_type falcon_b0_nic_type = {
+       .probe = falcon_probe_nic,
+       .remove = falcon_remove_nic,
+       .init = falcon_init_nic,
+       .fini = efx_port_dummy_op_void,
+       .monitor = falcon_monitor,
+       .reset = falcon_reset_hw,
+       .probe_port = falcon_probe_port,
+       .remove_port = falcon_remove_port,
+       .prepare_flush = falcon_prepare_flush,
+       .update_stats = falcon_update_nic_stats,
+       .start_stats = falcon_start_nic_stats,
+       .stop_stats = falcon_stop_nic_stats,
+       .set_id_led = falcon_set_id_led,
+       .push_irq_moderation = falcon_push_irq_moderation,
+       .push_multicast_hash = falcon_push_multicast_hash,
+       .reconfigure_port = falcon_reconfigure_port,
+       .get_wol = falcon_get_wol,
+       .set_wol = falcon_set_wol,
+       .resume_wol = efx_port_dummy_op_void,
+       .test_registers = falcon_b0_test_registers,
+       .test_nvram = falcon_test_nvram,
+       .default_mac_ops = &falcon_xmac_operations,
+
+       .revision = EFX_REV_FALCON_B0,
        /* Map everything up to and including the RSS indirection
         * table.  Don't map MSI-X table, MSI-X PBA since Linux
         * requires that they not be mapped.  */
@@ -3142,5 +1820,9 @@ struct efx_nic_type falcon_b_nic_type = {
        .phys_addr_channels = 32, /* Hardware limit is 64, but the legacy
                                   * interrupt handler only supports 32
                                   * channels */
+       .tx_dc_base = 0x130000,
+       .rx_dc_base = 0x100000,
+       .offload_features = NETIF_F_IP_CSUM,
+       .reset_world_flags = ETH_RESET_IRQ,
 };