linux/drivers/net/ethernet/sfc/nic_common.h
Edward Cree 51b35a454e sfc: skeleton EF100 PF driver
No TX or RX path, no MCDI, not even an ifup/down handler.
Besides stubs, the bulk of the patch deals with reading the Xilinx
 extended PCIe capability, which tells us where to find our BAR.

Though in the same module, EF100 has its own struct pci_driver,
 which is named sfc_ef100.

A small number of additional nic_type methods are added; those in the
 TX (tx_enqueue) and RX (rx_packet) paths are called through indirect
 call wrappers to minimise the performance impact.

Signed-off-by: Edward Cree <ecree@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2020-07-27 12:26:55 -07:00

287 lines
9.0 KiB
C

/* SPDX-License-Identifier: GPL-2.0-only */
/****************************************************************************
* Driver for Solarflare network controllers and boards
* Copyright 2005-2006 Fen Systems Ltd.
* Copyright 2006-2013 Solarflare Communications Inc.
* Copyright 2019-2020 Xilinx Inc.
*/
#ifndef EFX_NIC_COMMON_H
#define EFX_NIC_COMMON_H
#include "net_driver.h"
#include "efx_common.h"
#include "mcdi.h"
#include "ptp.h"
enum {
/* Revisions 0-2 were Falcon A0, A1 and B0 respectively.
* They are not supported by this driver but these revision numbers
* form part of the ethtool API for register dumping.
*/
EFX_REV_SIENA_A0 = 3,
EFX_REV_HUNT_A0 = 4,
EFX_REV_EF100 = 5,
};
static inline int efx_nic_rev(struct efx_nic *efx)
{
return efx->type->revision;
}
/* Read the current event from the event queue */
static inline efx_qword_t *efx_event(struct efx_channel *channel,
unsigned int index)
{
return ((efx_qword_t *) (channel->eventq.buf.addr)) +
(index & channel->eventq_mask);
}
/* 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 efx_event_present(efx_qword_t *event)
{
return !(EFX_DWORD_IS_ALL_ONES(event->dword[0]) |
EFX_DWORD_IS_ALL_ONES(event->dword[1]));
}
/* Returns a pointer to the specified transmit descriptor in the TX
* descriptor queue belonging to the specified channel.
*/
static inline efx_qword_t *
efx_tx_desc(struct efx_tx_queue *tx_queue, unsigned int index)
{
return ((efx_qword_t *) (tx_queue->txd.buf.addr)) + index;
}
/* Report whether this TX queue would be empty for the given write_count.
* May return false negative.
*/
static inline bool __efx_nic_tx_is_empty(struct efx_tx_queue *tx_queue,
unsigned int write_count)
{
unsigned int empty_read_count = READ_ONCE(tx_queue->empty_read_count);
if (empty_read_count == 0)
return false;
return ((empty_read_count ^ write_count) & ~EFX_EMPTY_COUNT_VALID) == 0;
}
/* Report whether the NIC considers this TX queue empty, using
* packet_write_count (the write count recorded for the last completable
* doorbell push). May return false negative. EF10 only, which is OK
* because only EF10 supports PIO.
*/
static inline bool efx_nic_tx_is_empty(struct efx_tx_queue *tx_queue)
{
EFX_WARN_ON_ONCE_PARANOID(!tx_queue->efx->type->option_descriptors);
return __efx_nic_tx_is_empty(tx_queue, tx_queue->packet_write_count);
}
/* Get partner of a TX queue, seen as part of the same net core queue */
/* XXX is this a thing on EF100? */
static inline struct efx_tx_queue *efx_tx_queue_partner(struct efx_tx_queue *tx_queue)
{
if (tx_queue->label & EFX_TXQ_TYPE_OFFLOAD)
return tx_queue - EFX_TXQ_TYPE_OFFLOAD;
else
return tx_queue + EFX_TXQ_TYPE_OFFLOAD;
}
/* Decide whether we can use TX PIO, ie. write packet data directly into
* a buffer on the device. This can reduce latency at the expense of
* throughput, so we only do this if both hardware and software TX rings
* are empty. This also ensures that only one packet at a time can be
* using the PIO buffer.
*/
static inline bool efx_nic_may_tx_pio(struct efx_tx_queue *tx_queue)
{
struct efx_tx_queue *partner = efx_tx_queue_partner(tx_queue);
return tx_queue->piobuf && efx_nic_tx_is_empty(tx_queue) &&
efx_nic_tx_is_empty(partner);
}
int efx_enqueue_skb_tso(struct efx_tx_queue *tx_queue, struct sk_buff *skb,
bool *data_mapped);
/* Decide whether to push a TX descriptor to the NIC vs merely writing
* the doorbell. This can reduce latency when we are adding a single
* descriptor to an empty queue, but is otherwise pointless. Further,
* Falcon and Siena have hardware bugs (SF bug 33851) that may be
* triggered if we don't check this.
* We use the write_count used for the last doorbell push, to get the
* NIC's view of the tx queue.
*/
static inline bool efx_nic_may_push_tx_desc(struct efx_tx_queue *tx_queue,
unsigned int write_count)
{
bool was_empty = __efx_nic_tx_is_empty(tx_queue, write_count);
tx_queue->empty_read_count = 0;
return was_empty && tx_queue->write_count - write_count == 1;
}
/* Returns a pointer to the specified descriptor in the RX descriptor queue */
static inline efx_qword_t *
efx_rx_desc(struct efx_rx_queue *rx_queue, unsigned int index)
{
return ((efx_qword_t *) (rx_queue->rxd.buf.addr)) + index;
}
/* Alignment of PCIe DMA boundaries (4KB) */
#define EFX_PAGE_SIZE 4096
/* Size and alignment of buffer table entries (same) */
#define EFX_BUF_SIZE EFX_PAGE_SIZE
/* NIC-generic software stats */
enum {
GENERIC_STAT_rx_noskb_drops,
GENERIC_STAT_rx_nodesc_trunc,
GENERIC_STAT_COUNT
};
#define EFX_GENERIC_SW_STAT(ext_name) \
[GENERIC_STAT_ ## ext_name] = { #ext_name, 0, 0 }
/* TX data path */
static inline int efx_nic_probe_tx(struct efx_tx_queue *tx_queue)
{
return tx_queue->efx->type->tx_probe(tx_queue);
}
static inline void efx_nic_init_tx(struct efx_tx_queue *tx_queue)
{
tx_queue->efx->type->tx_init(tx_queue);
}
static inline void efx_nic_remove_tx(struct efx_tx_queue *tx_queue)
{
if (tx_queue->efx->type->tx_remove)
tx_queue->efx->type->tx_remove(tx_queue);
}
static inline void efx_nic_push_buffers(struct efx_tx_queue *tx_queue)
{
tx_queue->efx->type->tx_write(tx_queue);
}
/* RX data path */
static inline int efx_nic_probe_rx(struct efx_rx_queue *rx_queue)
{
return rx_queue->efx->type->rx_probe(rx_queue);
}
static inline void efx_nic_init_rx(struct efx_rx_queue *rx_queue)
{
rx_queue->efx->type->rx_init(rx_queue);
}
static inline void efx_nic_remove_rx(struct efx_rx_queue *rx_queue)
{
rx_queue->efx->type->rx_remove(rx_queue);
}
static inline void efx_nic_notify_rx_desc(struct efx_rx_queue *rx_queue)
{
rx_queue->efx->type->rx_write(rx_queue);
}
static inline void efx_nic_generate_fill_event(struct efx_rx_queue *rx_queue)
{
rx_queue->efx->type->rx_defer_refill(rx_queue);
}
/* Event data path */
static inline int efx_nic_probe_eventq(struct efx_channel *channel)
{
return channel->efx->type->ev_probe(channel);
}
static inline int efx_nic_init_eventq(struct efx_channel *channel)
{
return channel->efx->type->ev_init(channel);
}
static inline void efx_nic_fini_eventq(struct efx_channel *channel)
{
channel->efx->type->ev_fini(channel);
}
static inline void efx_nic_remove_eventq(struct efx_channel *channel)
{
channel->efx->type->ev_remove(channel);
}
static inline int
efx_nic_process_eventq(struct efx_channel *channel, int quota)
{
return channel->efx->type->ev_process(channel, quota);
}
static inline void efx_nic_eventq_read_ack(struct efx_channel *channel)
{
channel->efx->type->ev_read_ack(channel);
}
void efx_nic_event_test_start(struct efx_channel *channel);
bool efx_nic_event_present(struct efx_channel *channel);
static inline void efx_sensor_event(struct efx_nic *efx, efx_qword_t *ev)
{
if (efx->type->sensor_event)
efx->type->sensor_event(efx, ev);
}
/* Some statistics are computed as A - B where A and B each increase
* linearly with some hardware counter(s) and the counters are read
* asynchronously. If the counters contributing to B are always read
* after those contributing to A, the computed value may be lower than
* the true value by some variable amount, and may decrease between
* subsequent computations.
*
* We should never allow statistics to decrease or to exceed the true
* value. Since the computed value will never be greater than the
* true value, we can achieve this by only storing the computed value
* when it increases.
*/
static inline void efx_update_diff_stat(u64 *stat, u64 diff)
{
if ((s64)(diff - *stat) > 0)
*stat = diff;
}
/* Interrupts */
int efx_nic_init_interrupt(struct efx_nic *efx);
int efx_nic_irq_test_start(struct efx_nic *efx);
void efx_nic_fini_interrupt(struct efx_nic *efx);
static inline int efx_nic_event_test_irq_cpu(struct efx_channel *channel)
{
return READ_ONCE(channel->event_test_cpu);
}
static inline int efx_nic_irq_test_irq_cpu(struct efx_nic *efx)
{
return READ_ONCE(efx->last_irq_cpu);
}
/* Global Resources */
int efx_nic_alloc_buffer(struct efx_nic *efx, struct efx_buffer *buffer,
unsigned int len, gfp_t gfp_flags);
void efx_nic_free_buffer(struct efx_nic *efx, struct efx_buffer *buffer);
size_t efx_nic_get_regs_len(struct efx_nic *efx);
void efx_nic_get_regs(struct efx_nic *efx, void *buf);
#define EFX_MC_STATS_GENERATION_INVALID ((__force __le64)(-1))
size_t efx_nic_describe_stats(const struct efx_hw_stat_desc *desc, size_t count,
const unsigned long *mask, u8 *names);
int efx_nic_copy_stats(struct efx_nic *efx, __le64 *dest);
void efx_nic_update_stats(const struct efx_hw_stat_desc *desc, size_t count,
const unsigned long *mask, u64 *stats,
const void *dma_buf, bool accumulate);
void efx_nic_fix_nodesc_drop_stat(struct efx_nic *efx, u64 *stat);
#define EFX_MAX_FLUSH_TIME 5000
#endif /* EFX_NIC_COMMON_H */