diff options
Diffstat (limited to 'drivers/net/ethernet/sfc/efx.c')
| -rw-r--r-- | drivers/net/ethernet/sfc/efx.c | 2039 |
1 files changed, 1296 insertions, 743 deletions
diff --git a/drivers/net/ethernet/sfc/efx.c b/drivers/net/ethernet/sfc/efx.c index e43702f33b6..1e274045970 100644 --- a/drivers/net/ethernet/sfc/efx.c +++ b/drivers/net/ethernet/sfc/efx.c @@ -1,7 +1,7 @@ /**************************************************************************** - * Driver for Solarflare Solarstorm network controllers and boards + * Driver for Solarflare network controllers and boards * Copyright 2005-2006 Fen Systems Ltd. - * Copyright 2005-2011 Solarflare Communications Inc. + * Copyright 2005-2013 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 @@ -17,14 +17,15 @@ #include <linux/ip.h> #include <linux/tcp.h> #include <linux/in.h> -#include <linux/crc32.h> #include <linux/ethtool.h> #include <linux/topology.h> #include <linux/gfp.h> -#include <linux/cpu_rmap.h> +#include <linux/aer.h> +#include <linux/interrupt.h> #include "net_driver.h" #include "efx.h" #include "nic.h" +#include "selftest.h" #include "mcdi.h" #include "workarounds.h" @@ -38,15 +39,15 @@ /* Loopback mode names (see LOOPBACK_MODE()) */ const unsigned int efx_loopback_mode_max = LOOPBACK_MAX; -const char *efx_loopback_mode_names[] = { +const char *const efx_loopback_mode_names[] = { [LOOPBACK_NONE] = "NONE", [LOOPBACK_DATA] = "DATAPATH", [LOOPBACK_GMAC] = "GMAC", [LOOPBACK_XGMII] = "XGMII", [LOOPBACK_XGXS] = "XGXS", - [LOOPBACK_XAUI] = "XAUI", - [LOOPBACK_GMII] = "GMII", - [LOOPBACK_SGMII] = "SGMII", + [LOOPBACK_XAUI] = "XAUI", + [LOOPBACK_GMII] = "GMII", + [LOOPBACK_SGMII] = "SGMII", [LOOPBACK_XGBR] = "XGBR", [LOOPBACK_XFI] = "XFI", [LOOPBACK_XAUI_FAR] = "XAUI_FAR", @@ -55,42 +56,49 @@ const char *efx_loopback_mode_names[] = { [LOOPBACK_XFI_FAR] = "XFI_FAR", [LOOPBACK_GPHY] = "GPHY", [LOOPBACK_PHYXS] = "PHYXS", - [LOOPBACK_PCS] = "PCS", - [LOOPBACK_PMAPMD] = "PMA/PMD", + [LOOPBACK_PCS] = "PCS", + [LOOPBACK_PMAPMD] = "PMA/PMD", [LOOPBACK_XPORT] = "XPORT", [LOOPBACK_XGMII_WS] = "XGMII_WS", - [LOOPBACK_XAUI_WS] = "XAUI_WS", + [LOOPBACK_XAUI_WS] = "XAUI_WS", [LOOPBACK_XAUI_WS_FAR] = "XAUI_WS_FAR", [LOOPBACK_XAUI_WS_NEAR] = "XAUI_WS_NEAR", - [LOOPBACK_GMII_WS] = "GMII_WS", + [LOOPBACK_GMII_WS] = "GMII_WS", [LOOPBACK_XFI_WS] = "XFI_WS", [LOOPBACK_XFI_WS_FAR] = "XFI_WS_FAR", - [LOOPBACK_PHYXS_WS] = "PHYXS_WS", + [LOOPBACK_PHYXS_WS] = "PHYXS_WS", }; const unsigned int efx_reset_type_max = RESET_TYPE_MAX; -const char *efx_reset_type_names[] = { - [RESET_TYPE_INVISIBLE] = "INVISIBLE", - [RESET_TYPE_ALL] = "ALL", - [RESET_TYPE_WORLD] = "WORLD", - [RESET_TYPE_DISABLE] = "DISABLE", - [RESET_TYPE_TX_WATCHDOG] = "TX_WATCHDOG", - [RESET_TYPE_INT_ERROR] = "INT_ERROR", - [RESET_TYPE_RX_RECOVERY] = "RX_RECOVERY", - [RESET_TYPE_RX_DESC_FETCH] = "RX_DESC_FETCH", - [RESET_TYPE_TX_DESC_FETCH] = "TX_DESC_FETCH", - [RESET_TYPE_TX_SKIP] = "TX_SKIP", - [RESET_TYPE_MC_FAILURE] = "MC_FAILURE", +const char *const efx_reset_type_names[] = { + [RESET_TYPE_INVISIBLE] = "INVISIBLE", + [RESET_TYPE_ALL] = "ALL", + [RESET_TYPE_RECOVER_OR_ALL] = "RECOVER_OR_ALL", + [RESET_TYPE_WORLD] = "WORLD", + [RESET_TYPE_RECOVER_OR_DISABLE] = "RECOVER_OR_DISABLE", + [RESET_TYPE_MC_BIST] = "MC_BIST", + [RESET_TYPE_DISABLE] = "DISABLE", + [RESET_TYPE_TX_WATCHDOG] = "TX_WATCHDOG", + [RESET_TYPE_INT_ERROR] = "INT_ERROR", + [RESET_TYPE_RX_RECOVERY] = "RX_RECOVERY", + [RESET_TYPE_DMA_ERROR] = "DMA_ERROR", + [RESET_TYPE_TX_SKIP] = "TX_SKIP", + [RESET_TYPE_MC_FAILURE] = "MC_FAILURE", + [RESET_TYPE_MCDI_TIMEOUT] = "MCDI_TIMEOUT (FLR)", }; -#define EFX_MAX_MTU (9 * 1024) - /* Reset workqueue. If any NIC has a hardware failure then a reset will be * queued onto this work queue. This is not a per-nic work queue, because * efx_reset_work() acquires the rtnl lock, so resets are naturally serialised. */ static struct workqueue_struct *reset_workqueue; +/* How often and how many times to poll for a reset while waiting for a + * BIST that another function started to complete. + */ +#define BIST_WAIT_DELAY_MS 100 +#define BIST_WAIT_DELAY_COUNT 100 + /************************************************************************** * * Configurable values @@ -105,8 +113,8 @@ static struct workqueue_struct *reset_workqueue; * * This is only used in MSI-X interrupt mode */ -static unsigned int separate_tx_channels; -module_param(separate_tx_channels, uint, 0444); +static bool separate_tx_channels; +module_param(separate_tx_channels, bool, 0444); MODULE_PARM_DESC(separate_tx_channels, "Use separate channels for TX and RX"); @@ -116,21 +124,15 @@ MODULE_PARM_DESC(separate_tx_channels, static int napi_weight = 64; /* This is the time (in jiffies) between invocations of the hardware - * monitor. On Falcon-based NICs, this will: + * monitor. + * On Falcon-based NICs, this will: * - Check the on-board hardware monitor; * - Poll the link state and reconfigure the hardware as necessary. + * On Siena-based NICs for power systems with EEH support, this will give EEH a + * chance to start. */ static unsigned int efx_monitor_interval = 1 * HZ; -/* This controls whether or not the driver will initialise devices - * with invalid MAC addresses stored in the EEPROM or flash. If true, - * such devices will be initialised with a random locally-generated - * MAC address. This allows for loading the sfc_mtd driver to - * reprogram the flash, even if the flash contents (including the MAC - * address) have previously been erased. - */ -static unsigned int allow_bad_hwaddr; - /* Initial interrupt moderation settings. They can be modified after * module load with ethtool. * @@ -162,22 +164,22 @@ static unsigned int interrupt_mode; * interrupt handling. * * Cards without MSI-X will only target one CPU via legacy or MSI interrupt. - * The default (0) means to assign an interrupt to each package (level II cache) + * The default (0) means to assign an interrupt to each core. */ static unsigned int rss_cpus; module_param(rss_cpus, uint, 0444); MODULE_PARM_DESC(rss_cpus, "Number of CPUs to use for Receive-Side Scaling"); -static int phy_flash_cfg; -module_param(phy_flash_cfg, int, 0644); +static bool phy_flash_cfg; +module_param(phy_flash_cfg, bool, 0644); MODULE_PARM_DESC(phy_flash_cfg, "Set PHYs into reflash mode initially"); -static unsigned irq_adapt_low_thresh = 10000; +static unsigned irq_adapt_low_thresh = 8000; module_param(irq_adapt_low_thresh, uint, 0644); MODULE_PARM_DESC(irq_adapt_low_thresh, "Threshold score for reducing IRQ moderation"); -static unsigned irq_adapt_high_thresh = 20000; +static unsigned irq_adapt_high_thresh = 16000; module_param(irq_adapt_high_thresh, uint, 0644); MODULE_PARM_DESC(irq_adapt_high_thresh, "Threshold score for increasing IRQ moderation"); @@ -195,9 +197,13 @@ MODULE_PARM_DESC(debug, "Bitmapped debugging message enable value"); * *************************************************************************/ +static int efx_soft_enable_interrupts(struct efx_nic *efx); +static void efx_soft_disable_interrupts(struct efx_nic *efx); +static void efx_remove_channel(struct efx_channel *channel); static void efx_remove_channels(struct efx_nic *efx); +static const struct efx_channel_type efx_default_channel_type; static void efx_remove_port(struct efx_nic *efx); -static void efx_init_napi(struct efx_nic *efx); +static void efx_init_napi_channel(struct efx_channel *channel); static void efx_fini_napi(struct efx_nic *efx); static void efx_fini_napi_channel(struct efx_channel *channel); static void efx_fini_struct(struct efx_nic *efx); @@ -206,11 +212,22 @@ static void efx_stop_all(struct efx_nic *efx); #define EFX_ASSERT_RESET_SERIALISED(efx) \ do { \ - if ((efx->state == STATE_RUNNING) || \ + if ((efx->state == STATE_READY) || \ + (efx->state == STATE_RECOVERY) || \ (efx->state == STATE_DISABLED)) \ ASSERT_RTNL(); \ } while (0) +static int efx_check_disabled(struct efx_nic *efx) +{ + if (efx->state == STATE_DISABLED || efx->state == STATE_RECOVERY) { + netif_err(efx, drv, efx->net_dev, + "device is disabled due to earlier errors\n"); + return -EIO; + } + return 0; +} + /************************************************************************** * * Event queue processing @@ -226,47 +243,23 @@ static void efx_stop_all(struct efx_nic *efx); */ static int efx_process_channel(struct efx_channel *channel, int budget) { - struct efx_nic *efx = channel->efx; int spent; - if (unlikely(efx->reset_pending || !channel->enabled)) + if (unlikely(!channel->enabled)) return 0; spent = efx_nic_process_eventq(channel, budget); - if (spent == 0) - return 0; + if (spent && efx_channel_has_rx_queue(channel)) { + struct efx_rx_queue *rx_queue = + efx_channel_get_rx_queue(channel); - /* Deliver last RX packet. */ - if (channel->rx_pkt) { - __efx_rx_packet(channel, channel->rx_pkt, - channel->rx_pkt_csummed); - channel->rx_pkt = NULL; + efx_rx_flush_packet(channel); + efx_fast_push_rx_descriptors(rx_queue, true); } - efx_rx_strategy(channel); - - efx_fast_push_rx_descriptors(efx_channel_get_rx_queue(channel)); - return spent; } -/* Mark channel as finished processing - * - * Note that since we will not receive further interrupts for this - * channel before we finish processing and call the eventq_read_ack() - * method, there is no need to use the interrupt hold-off timers. - */ -static inline void efx_channel_processed(struct efx_channel *channel) -{ - /* The interrupt handler for this channel may set work_pending - * as soon as we acknowledge the events we've seen. Make sure - * it's cleared before then. */ - channel->work_pending = false; - smp_wmb(); - - efx_nic_eventq_read_ack(channel); -} - /* NAPI poll handler * * NAPI guarantees serialisation of polls of the same device, which @@ -286,7 +279,7 @@ static int efx_poll(struct napi_struct *napi, int budget) spent = efx_process_channel(channel, budget); if (spent < budget) { - if (channel->channel < efx->n_rx_channels && + if (efx_channel_has_rx_queue(channel) && efx->irq_rx_adaptive && unlikely(++channel->irq_count == 1000)) { if (unlikely(channel->irq_mod_score < @@ -311,58 +304,16 @@ static int efx_poll(struct napi_struct *napi, int budget) /* There is no race here; although napi_disable() will * only wait for napi_complete(), this isn't a problem - * since efx_channel_processed() will have no effect if + * since efx_nic_eventq_read_ack() will have no effect if * interrupts have already been disabled. */ napi_complete(napi); - efx_channel_processed(channel); + efx_nic_eventq_read_ack(channel); } return spent; } -/* Process the eventq of the specified channel immediately on this CPU - * - * Disable hardware generated interrupts, wait for any existing - * processing to finish, then directly poll (and ack ) the eventq. - * Finally reenable NAPI and interrupts. - * - * This is for use only during a loopback self-test. It must not - * deliver any packets up the stack as this can result in deadlock. - */ -void efx_process_channel_now(struct efx_channel *channel) -{ - struct efx_nic *efx = channel->efx; - - BUG_ON(channel->channel >= efx->n_channels); - BUG_ON(!channel->enabled); - BUG_ON(!efx->loopback_selftest); - - /* Disable interrupts and wait for ISRs to complete */ - efx_nic_disable_interrupts(efx); - if (efx->legacy_irq) { - synchronize_irq(efx->legacy_irq); - efx->legacy_irq_enabled = false; - } - if (channel->irq) - synchronize_irq(channel->irq); - - /* Wait for any NAPI processing to complete */ - napi_disable(&channel->napi_str); - - /* Poll the channel */ - efx_process_channel(channel, channel->eventq_mask + 1); - - /* Ack the eventq. This may cause an interrupt to be generated - * when they are reenabled */ - efx_channel_processed(channel); - - napi_enable(&channel->napi_str); - if (efx->legacy_irq) - efx->legacy_irq_enabled = true; - efx_nic_enable_interrupts(efx); -} - /* Create event queue * Event queue memory allocations are done only once. If the channel * is reset, the memory buffer will be reused; this guards against @@ -373,7 +324,7 @@ static int efx_probe_eventq(struct efx_channel *channel) struct efx_nic *efx = channel->efx; unsigned long entries; - netif_dbg(channel->efx, probe, channel->efx->net_dev, + netif_dbg(efx, probe, efx->net_dev, "chan %d create event queue\n", channel->channel); /* Build an event queue with room for one event per tx and rx buffer, @@ -386,22 +337,59 @@ static int efx_probe_eventq(struct efx_channel *channel) } /* Prepare channel's event queue */ -static void efx_init_eventq(struct efx_channel *channel) +static int efx_init_eventq(struct efx_channel *channel) { - netif_dbg(channel->efx, drv, channel->efx->net_dev, + struct efx_nic *efx = channel->efx; + int rc; + + EFX_WARN_ON_PARANOID(channel->eventq_init); + + netif_dbg(efx, drv, efx->net_dev, "chan %d init event queue\n", channel->channel); - channel->eventq_read_ptr = 0; + rc = efx_nic_init_eventq(channel); + if (rc == 0) { + efx->type->push_irq_moderation(channel); + channel->eventq_read_ptr = 0; + channel->eventq_init = true; + } + return rc; +} + +/* Enable event queue processing and NAPI */ +static void efx_start_eventq(struct efx_channel *channel) +{ + netif_dbg(channel->efx, ifup, channel->efx->net_dev, + "chan %d start event queue\n", channel->channel); + + /* Make sure the NAPI handler sees the enabled flag set */ + channel->enabled = true; + smp_wmb(); + + napi_enable(&channel->napi_str); + efx_nic_eventq_read_ack(channel); +} + +/* Disable event queue processing and NAPI */ +static void efx_stop_eventq(struct efx_channel *channel) +{ + if (!channel->enabled) + return; - efx_nic_init_eventq(channel); + napi_disable(&channel->napi_str); + channel->enabled = false; } static void efx_fini_eventq(struct efx_channel *channel) { + if (!channel->eventq_init) + return; + netif_dbg(channel->efx, drv, channel->efx->net_dev, "chan %d fini event queue\n", channel->channel); efx_nic_fini_eventq(channel); + channel->eventq_init = false; } static void efx_remove_eventq(struct efx_channel *channel) @@ -418,8 +406,7 @@ static void efx_remove_eventq(struct efx_channel *channel) * *************************************************************************/ -/* Allocate and initialise a channel structure, optionally copying - * parameters (but not resources) from an old channel structure. */ +/* Allocate and initialise a channel structure. */ static struct efx_channel * efx_alloc_channel(struct efx_nic *efx, int i, struct efx_channel *old_channel) { @@ -428,45 +415,60 @@ efx_alloc_channel(struct efx_nic *efx, int i, struct efx_channel *old_channel) struct efx_tx_queue *tx_queue; int j; - if (old_channel) { - channel = kmalloc(sizeof(*channel), GFP_KERNEL); - if (!channel) - return NULL; + channel = kzalloc(sizeof(*channel), GFP_KERNEL); + if (!channel) + return NULL; + + channel->efx = efx; + channel->channel = i; + channel->type = &efx_default_channel_type; + + for (j = 0; j < EFX_TXQ_TYPES; j++) { + tx_queue = &channel->tx_queue[j]; + tx_queue->efx = efx; + tx_queue->queue = i * EFX_TXQ_TYPES + j; + tx_queue->channel = channel; + } + + rx_queue = &channel->rx_queue; + rx_queue->efx = efx; + setup_timer(&rx_queue->slow_fill, efx_rx_slow_fill, + (unsigned long)rx_queue); - *channel = *old_channel; + return channel; +} - channel->napi_dev = NULL; - memset(&channel->eventq, 0, sizeof(channel->eventq)); +/* Allocate and initialise a channel structure, copying parameters + * (but not resources) from an old channel structure. + */ +static struct efx_channel * +efx_copy_channel(const struct efx_channel *old_channel) +{ + struct efx_channel *channel; + struct efx_rx_queue *rx_queue; + struct efx_tx_queue *tx_queue; + int j; - rx_queue = &channel->rx_queue; - rx_queue->buffer = NULL; - memset(&rx_queue->rxd, 0, sizeof(rx_queue->rxd)); + channel = kmalloc(sizeof(*channel), GFP_KERNEL); + if (!channel) + return NULL; - for (j = 0; j < EFX_TXQ_TYPES; j++) { - tx_queue = &channel->tx_queue[j]; - if (tx_queue->channel) - tx_queue->channel = channel; - tx_queue->buffer = NULL; - memset(&tx_queue->txd, 0, sizeof(tx_queue->txd)); - } - } else { - channel = kzalloc(sizeof(*channel), GFP_KERNEL); - if (!channel) - return NULL; + *channel = *old_channel; - channel->efx = efx; - channel->channel = i; + channel->napi_dev = NULL; + memset(&channel->eventq, 0, sizeof(channel->eventq)); - for (j = 0; j < EFX_TXQ_TYPES; j++) { - tx_queue = &channel->tx_queue[j]; - tx_queue->efx = efx; - tx_queue->queue = i * EFX_TXQ_TYPES + j; + for (j = 0; j < EFX_TXQ_TYPES; j++) { + tx_queue = &channel->tx_queue[j]; + if (tx_queue->channel) tx_queue->channel = channel; - } + tx_queue->buffer = NULL; + memset(&tx_queue->txd, 0, sizeof(tx_queue->txd)); } rx_queue = &channel->rx_queue; - rx_queue->efx = efx; + rx_queue->buffer = NULL; + memset(&rx_queue->rxd, 0, sizeof(rx_queue->rxd)); setup_timer(&rx_queue->slow_fill, efx_rx_slow_fill, (unsigned long)rx_queue); @@ -482,57 +484,60 @@ static int efx_probe_channel(struct efx_channel *channel) netif_dbg(channel->efx, probe, channel->efx->net_dev, "creating channel %d\n", channel->channel); + rc = channel->type->pre_probe(channel); + if (rc) + goto fail; + rc = efx_probe_eventq(channel); if (rc) - goto fail1; + goto fail; efx_for_each_channel_tx_queue(tx_queue, channel) { rc = efx_probe_tx_queue(tx_queue); if (rc) - goto fail2; + goto fail; } efx_for_each_channel_rx_queue(rx_queue, channel) { rc = efx_probe_rx_queue(rx_queue); if (rc) - goto fail3; + goto fail; } - channel->n_rx_frm_trunc = 0; - return 0; - fail3: - efx_for_each_channel_rx_queue(rx_queue, channel) - efx_remove_rx_queue(rx_queue); - fail2: - efx_for_each_channel_tx_queue(tx_queue, channel) - efx_remove_tx_queue(tx_queue); - fail1: +fail: + efx_remove_channel(channel); return rc; } +static void +efx_get_channel_name(struct efx_channel *channel, char *buf, size_t len) +{ + struct efx_nic *efx = channel->efx; + const char *type; + int number; + + number = channel->channel; + if (efx->tx_channel_offset == 0) { + type = ""; + } else if (channel->channel < efx->tx_channel_offset) { + type = "-rx"; + } else { + type = "-tx"; + number -= efx->tx_channel_offset; + } + snprintf(buf, len, "%s%s-%d", efx->name, type, number); +} static void efx_set_channel_names(struct efx_nic *efx) { struct efx_channel *channel; - const char *type = ""; - int number; - efx_for_each_channel(channel, efx) { - number = channel->channel; - if (efx->n_channels > efx->n_rx_channels) { - if (channel->channel < efx->n_rx_channels) { - type = "-rx"; - } else { - type = "-tx"; - number -= efx->n_rx_channels; - } - } - snprintf(efx->channel_name[channel->channel], - sizeof(efx->channel_name[0]), - "%s%s-%d", efx->name, type, number); - } + efx_for_each_channel(channel, efx) + channel->type->get_name(channel, + efx->msi_context[channel->channel].name, + sizeof(efx->msi_context[0].name)); } static int efx_probe_channels(struct efx_nic *efx) @@ -543,7 +548,12 @@ static int efx_probe_channels(struct efx_nic *efx) /* Restart special buffer allocation */ efx->next_buffer_table = 0; - efx_for_each_channel(channel, efx) { + /* Probe channels in reverse, so that any 'extra' channels + * use the start of the buffer table. This allows the traffic + * channels to be resized without moving them or wasting the + * entries before them. + */ + efx_for_each_channel_rev(channel, efx) { rc = efx_probe_channel(channel); if (rc) { netif_err(efx, probe, efx->net_dev, @@ -565,87 +575,91 @@ fail: * to propagate configuration changes (mtu, checksum offload), or * to clear hardware error conditions */ -static void efx_init_channels(struct efx_nic *efx) +static void efx_start_datapath(struct efx_nic *efx) { + bool old_rx_scatter = efx->rx_scatter; struct efx_tx_queue *tx_queue; struct efx_rx_queue *rx_queue; struct efx_channel *channel; + size_t rx_buf_len; /* Calculate the rx buffer allocation parameters required to * support the current MTU, including padding for header * alignment and overruns. */ - efx->rx_buffer_len = (max(EFX_PAGE_IP_ALIGN, NET_IP_ALIGN) + - EFX_MAX_FRAME_LEN(efx->net_dev->mtu) + - efx->type->rx_buffer_hash_size + - efx->type->rx_buffer_padding); - efx->rx_buffer_order = get_order(efx->rx_buffer_len + - sizeof(struct efx_rx_page_state)); + efx->rx_dma_len = (efx->rx_prefix_size + + EFX_MAX_FRAME_LEN(efx->net_dev->mtu) + + efx->type->rx_buffer_padding); + rx_buf_len = (sizeof(struct efx_rx_page_state) + + efx->rx_ip_align + efx->rx_dma_len); + if (rx_buf_len <= PAGE_SIZE) { + efx->rx_scatter = efx->type->always_rx_scatter; + efx->rx_buffer_order = 0; + } else if (efx->type->can_rx_scatter) { + BUILD_BUG_ON(EFX_RX_USR_BUF_SIZE % L1_CACHE_BYTES); + BUILD_BUG_ON(sizeof(struct efx_rx_page_state) + + 2 * ALIGN(NET_IP_ALIGN + EFX_RX_USR_BUF_SIZE, + EFX_RX_BUF_ALIGNMENT) > + PAGE_SIZE); + efx->rx_scatter = true; + efx->rx_dma_len = EFX_RX_USR_BUF_SIZE; + efx->rx_buffer_order = 0; + } else { + efx->rx_scatter = false; + efx->rx_buffer_order = get_order(rx_buf_len); + } + + efx_rx_config_page_split(efx); + if (efx->rx_buffer_order) + netif_dbg(efx, drv, efx->net_dev, + "RX buf len=%u; page order=%u batch=%u\n", + efx->rx_dma_len, efx->rx_buffer_order, + efx->rx_pages_per_batch); + else + netif_dbg(efx, drv, efx->net_dev, + "RX buf len=%u step=%u bpp=%u; page batch=%u\n", + efx->rx_dma_len, efx->rx_page_buf_step, + efx->rx_bufs_per_page, efx->rx_pages_per_batch); + + /* RX filters may also have scatter-enabled flags */ + if (efx->rx_scatter != old_rx_scatter) + efx->type->filter_update_rx_scatter(efx); + + /* We must keep at least one descriptor in a TX ring empty. + * We could avoid this when the queue size does not exactly + * match the hardware ring size, but it's not that important. + * Therefore we stop the queue when one more skb might fill + * the ring completely. We wake it when half way back to + * empty. + */ + efx->txq_stop_thresh = efx->txq_entries - efx_tx_max_skb_descs(efx); + efx->txq_wake_thresh = efx->txq_stop_thresh / 2; /* Initialise the channels */ efx_for_each_channel(channel, efx) { - netif_dbg(channel->efx, drv, channel->efx->net_dev, - "init chan %d\n", channel->channel); - - efx_init_eventq(channel); - - efx_for_each_channel_tx_queue(tx_queue, channel) + efx_for_each_channel_tx_queue(tx_queue, channel) { efx_init_tx_queue(tx_queue); + atomic_inc(&efx->active_queues); + } - /* The rx buffer allocation strategy is MTU dependent */ - efx_rx_strategy(channel); - - efx_for_each_channel_rx_queue(rx_queue, channel) + efx_for_each_channel_rx_queue(rx_queue, channel) { efx_init_rx_queue(rx_queue); + atomic_inc(&efx->active_queues); + efx_stop_eventq(channel); + efx_fast_push_rx_descriptors(rx_queue, false); + efx_start_eventq(channel); + } - WARN_ON(channel->rx_pkt != NULL); - efx_rx_strategy(channel); + WARN_ON(channel->rx_pkt_n_frags); } -} -/* This enables event queue processing and packet transmission. - * - * Note that this function is not allowed to fail, since that would - * introduce too much complexity into the suspend/resume path. - */ -static void efx_start_channel(struct efx_channel *channel) -{ - struct efx_rx_queue *rx_queue; - - netif_dbg(channel->efx, ifup, channel->efx->net_dev, - "starting chan %d\n", channel->channel); + efx_ptp_start_datapath(efx); - /* The interrupt handler for this channel may set work_pending - * as soon as we enable it. Make sure it's cleared before - * then. Similarly, make sure it sees the enabled flag set. */ - channel->work_pending = false; - channel->enabled = true; - smp_wmb(); - - /* Fill the queues before enabling NAPI */ - efx_for_each_channel_rx_queue(rx_queue, channel) - efx_fast_push_rx_descriptors(rx_queue); - - napi_enable(&channel->napi_str); -} - -/* This disables event queue processing and packet transmission. - * This function does not guarantee that all queue processing - * (e.g. RX refill) is complete. - */ -static void efx_stop_channel(struct efx_channel *channel) -{ - if (!channel->enabled) - return; - - netif_dbg(channel->efx, ifdown, channel->efx->net_dev, - "stop chan %d\n", channel->channel); - - channel->enabled = false; - napi_disable(&channel->napi_str); + if (netif_device_present(efx->net_dev)) + netif_tx_wake_all_queues(efx->net_dev); } -static void efx_fini_channels(struct efx_nic *efx) +static void efx_stop_datapath(struct efx_nic *efx) { struct efx_channel *channel; struct efx_tx_queue *tx_queue; @@ -655,12 +669,34 @@ static void efx_fini_channels(struct efx_nic *efx) EFX_ASSERT_RESET_SERIALISED(efx); BUG_ON(efx->port_enabled); - rc = efx_nic_flush_queues(efx); + efx_ptp_stop_datapath(efx); + + /* Stop RX refill */ + efx_for_each_channel(channel, efx) { + efx_for_each_channel_rx_queue(rx_queue, channel) + rx_queue->refill_enabled = false; + } + + efx_for_each_channel(channel, efx) { + /* RX packet processing is pipelined, so wait for the + * NAPI handler to complete. At least event queue 0 + * might be kept active by non-data events, so don't + * use napi_synchronize() but actually disable NAPI + * temporarily. + */ + if (efx_channel_has_rx_queue(channel)) { + efx_stop_eventq(channel); + efx_start_eventq(channel); + } + } + + rc = efx->type->fini_dmaq(efx); if (rc && EFX_WORKAROUND_7803(efx)) { /* Schedule a reset to recover from the flush failure. The * descriptor caches reference memory we're about to free, * but falcon_reconfigure_mac_wrapper() won't reconnect - * the MACs because of the pending reset. */ + * the MACs because of the pending reset. + */ netif_err(efx, drv, efx->net_dev, "Resetting to recover from flush failure\n"); efx_schedule_reset(efx, RESET_TYPE_ALL); @@ -672,14 +708,10 @@ static void efx_fini_channels(struct efx_nic *efx) } efx_for_each_channel(channel, efx) { - netif_dbg(channel->efx, drv, channel->efx->net_dev, - "shut down chan %d\n", channel->channel); - efx_for_each_channel_rx_queue(rx_queue, channel) efx_fini_rx_queue(rx_queue); efx_for_each_possible_channel_tx_queue(tx_queue, channel) efx_fini_tx_queue(tx_queue); - efx_fini_eventq(channel); } } @@ -696,6 +728,7 @@ static void efx_remove_channel(struct efx_channel *channel) efx_for_each_possible_channel_tx_queue(tx_queue, channel) efx_remove_tx_queue(tx_queue); efx_remove_eventq(channel); + channel->type->post_remove(channel); } static void efx_remove_channels(struct efx_nic *efx) @@ -711,16 +744,45 @@ efx_realloc_channels(struct efx_nic *efx, u32 rxq_entries, u32 txq_entries) { struct efx_channel *other_channel[EFX_MAX_CHANNELS], *channel; u32 old_rxq_entries, old_txq_entries; - unsigned i; - int rc; + unsigned i, next_buffer_table = 0; + int rc, rc2; + + rc = efx_check_disabled(efx); + if (rc) + return rc; + /* Not all channels should be reallocated. We must avoid + * reallocating their buffer table entries. + */ + efx_for_each_channel(channel, efx) { + struct efx_rx_queue *rx_queue; + struct efx_tx_queue *tx_queue; + + if (channel->type->copy) + continue; + next_buffer_table = max(next_buffer_table, + channel->eventq.index + + channel->eventq.entries); + efx_for_each_channel_rx_queue(rx_queue, channel) + next_buffer_table = max(next_buffer_table, + rx_queue->rxd.index + + rx_queue->rxd.entries); + efx_for_each_channel_tx_queue(tx_queue, channel) + next_buffer_table = max(next_buffer_table, + tx_queue->txd.index + + tx_queue->txd.entries); + } + + efx_device_detach_sync(efx); efx_stop_all(efx); - efx_fini_channels(efx); + efx_soft_disable_interrupts(efx); - /* Clone channels */ + /* Clone channels (where possible) */ memset(other_channel, 0, sizeof(other_channel)); for (i = 0; i < efx->n_channels; i++) { - channel = efx_alloc_channel(efx, i, efx->channel[i]); + channel = efx->channel[i]; + if (channel->type->copy) + channel = channel->type->copy(channel); if (!channel) { rc = -ENOMEM; goto out; @@ -739,24 +801,40 @@ efx_realloc_channels(struct efx_nic *efx, u32 rxq_entries, u32 txq_entries) other_channel[i] = channel; } - rc = efx_probe_channels(efx); - if (rc) - goto rollback; - - efx_init_napi(efx); + /* Restart buffer table allocation */ + efx->next_buffer_table = next_buffer_table; - /* Destroy old channels */ for (i = 0; i < efx->n_channels; i++) { - efx_fini_napi_channel(other_channel[i]); - efx_remove_channel(other_channel[i]); + channel = efx->channel[i]; + if (!channel->type->copy) + continue; + rc = efx_probe_channel(channel); + if (rc) + goto rollback; + efx_init_napi_channel(efx->channel[i]); } + out: - /* Free unused channel structures */ - for (i = 0; i < efx->n_channels; i++) - kfree(other_channel[i]); + /* Destroy unused channel structures */ + for (i = 0; i < efx->n_channels; i++) { + channel = other_channel[i]; + if (channel && channel->type->copy) { + efx_fini_napi_channel(channel); + efx_remove_channel(channel); + kfree(channel); + } + } - efx_init_channels(efx); - efx_start_all(efx); + rc2 = efx_soft_enable_interrupts(efx); + if (rc2) { + rc = rc ? rc : rc2; + netif_err(efx, drv, efx->net_dev, + "unable to restart interrupts on channel reallocation\n"); + efx_schedule_reset(efx, RESET_TYPE_DISABLE); + } else { + efx_start_all(efx); + netif_device_attach(efx->net_dev); + } return rc; rollback: @@ -776,6 +854,23 @@ void efx_schedule_slow_fill(struct efx_rx_queue *rx_queue) mod_timer(&rx_queue->slow_fill, jiffies + msecs_to_jiffies(100)); } +static const struct efx_channel_type efx_default_channel_type = { + .pre_probe = efx_channel_dummy_op_int, + .post_remove = efx_channel_dummy_op_void, + .get_name = efx_get_channel_name, + .copy = efx_copy_channel, + .keep_eventq = false, +}; + +int efx_channel_dummy_op_int(struct efx_channel *channel) +{ + return 0; +} + +void efx_channel_dummy_op_void(struct efx_channel *channel) +{ +} + /************************************************************************** * * Port handling @@ -807,16 +902,13 @@ void efx_link_status_changed(struct efx_nic *efx) } /* Status message for kernel log */ - if (link_state->up) { + if (link_state->up) netif_info(efx, link, efx->net_dev, - "link up at %uMbps %s-duplex (MTU %d)%s\n", + "link up at %uMbps %s-duplex (MTU %d)\n", link_state->speed, link_state->fd ? "full" : "half", - efx->net_dev->mtu, - (efx->promiscuous ? " [PROMISC]" : "")); - } else { + efx->net_dev->mtu); + else netif_info(efx, link, efx->net_dev, "link down\n"); - } - } void efx_link_set_advertising(struct efx_nic *efx, u32 advertising) @@ -863,12 +955,6 @@ int __efx_reconfigure_port(struct efx_nic *efx) WARN_ON(!mutex_is_locked(&efx->mac_lock)); - /* Serialise the promiscuous flag with efx_set_multicast_list. */ - if (efx_dev_registered(efx)) { - netif_addr_lock_bh(efx->net_dev); - netif_addr_unlock_bh(efx->net_dev); - } - /* Disable PHY transmit in mac level loopbacks */ phy_mode = efx->phy_mode; if (LOOPBACK_INTERNAL(efx)) @@ -907,16 +993,13 @@ static void efx_mac_work(struct work_struct *data) struct efx_nic *efx = container_of(data, struct efx_nic, mac_work); mutex_lock(&efx->mac_lock); - if (efx->port_enabled) { - efx->type->push_multicast_hash(efx); - efx->mac_op->reconfigure(efx); - } + if (efx->port_enabled) + efx->type->reconfigure_mac(efx); mutex_unlock(&efx->mac_lock); } static int efx_probe_port(struct efx_nic *efx) { - unsigned char *perm_addr; int rc; netif_dbg(efx, probe, efx->net_dev, "create port\n"); @@ -929,28 +1012,10 @@ static int efx_probe_port(struct efx_nic *efx) if (rc) return rc; - /* Sanity check MAC address */ - perm_addr = efx->net_dev->perm_addr; - if (is_valid_ether_addr(perm_addr)) { - memcpy(efx->net_dev->dev_addr, perm_addr, ETH_ALEN); - } else { - netif_err(efx, probe, efx->net_dev, "invalid MAC address %pM\n", - perm_addr); - if (!allow_bad_hwaddr) { - rc = -EINVAL; - goto err; - } - random_ether_addr(efx->net_dev->dev_addr); - netif_info(efx, probe, efx->net_dev, - "using locally-generated MAC %pM\n", - efx->net_dev->dev_addr); - } + /* Initialise MAC address to permanent address */ + ether_addr_copy(efx->net_dev->dev_addr, efx->net_dev->perm_addr); return 0; - - err: - efx->type->remove_port(efx); - return rc; } static int efx_init_port(struct efx_nic *efx) @@ -969,7 +1034,7 @@ static int efx_init_port(struct efx_nic *efx) /* Reconfigure the MAC before creating dma queues (required for * Falcon/A1 where RX_INGR_EN/TX_DRAIN_EN isn't supported) */ - efx->mac_op->reconfigure(efx); + efx->type->reconfigure_mac(efx); /* Ensure the PHY advertises the correct flow control settings */ rc = efx->phy_op->reconfigure(efx); @@ -994,28 +1059,34 @@ static void efx_start_port(struct efx_nic *efx) mutex_lock(&efx->mac_lock); efx->port_enabled = true; - /* efx_mac_work() might have been scheduled after efx_stop_port(), - * and then cancelled by efx_flush_all() */ - efx->type->push_multicast_hash(efx); - efx->mac_op->reconfigure(efx); + /* Ensure MAC ingress/egress is enabled */ + efx->type->reconfigure_mac(efx); mutex_unlock(&efx->mac_lock); } -/* Prevent efx_mac_work() and efx_monitor() from working */ +/* Cancel work for MAC reconfiguration, periodic hardware monitoring + * and the async self-test, wait for them to finish and prevent them + * being scheduled again. This doesn't cover online resets, which + * should only be cancelled when removing the device. + */ static void efx_stop_port(struct efx_nic *efx) { netif_dbg(efx, ifdown, efx->net_dev, "stop port\n"); + EFX_ASSERT_RESET_SERIALISED(efx); + mutex_lock(&efx->mac_lock); efx->port_enabled = false; mutex_unlock(&efx->mac_lock); /* Serialise against efx_set_multicast_list() */ - if (efx_dev_registered(efx)) { - netif_addr_lock_bh(efx->net_dev); - netif_addr_unlock_bh(efx->net_dev); - } + netif_addr_lock_bh(efx->net_dev); + netif_addr_unlock_bh(efx->net_dev); + + cancel_delayed_work_sync(&efx->monitor_work); + efx_selftest_async_cancel(efx); + cancel_work_sync(&efx->mac_work); } static void efx_fini_port(struct efx_nic *efx) @@ -1045,11 +1116,83 @@ static void efx_remove_port(struct efx_nic *efx) * **************************************************************************/ +static LIST_HEAD(efx_primary_list); +static LIST_HEAD(efx_unassociated_list); + +static bool efx_same_controller(struct efx_nic *left, struct efx_nic *right) +{ + return left->type == right->type && + left->vpd_sn && right->vpd_sn && + !strcmp(left->vpd_sn, right->vpd_sn); +} + +static void efx_associate(struct efx_nic *efx) +{ + struct efx_nic *other, *next; + + if (efx->primary == efx) { + /* Adding primary function; look for secondaries */ + + netif_dbg(efx, probe, efx->net_dev, "adding to primary list\n"); + list_add_tail(&efx->node, &efx_primary_list); + + list_for_each_entry_safe(other, next, &efx_unassociated_list, + node) { + if (efx_same_controller(efx, other)) { + list_del(&other->node); + netif_dbg(other, probe, other->net_dev, + "moving to secondary list of %s %s\n", + pci_name(efx->pci_dev), + efx->net_dev->name); + list_add_tail(&other->node, + &efx->secondary_list); + other->primary = efx; + } + } + } else { + /* Adding secondary function; look for primary */ + + list_for_each_entry(other, &efx_primary_list, node) { + if (efx_same_controller(efx, other)) { + netif_dbg(efx, probe, efx->net_dev, + "adding to secondary list of %s %s\n", + pci_name(other->pci_dev), + other->net_dev->name); + list_add_tail(&efx->node, + &other->secondary_list); + efx->primary = other; + return; + } + } + + netif_dbg(efx, probe, efx->net_dev, + "adding to unassociated list\n"); + list_add_tail(&efx->node, &efx_unassociated_list); + } +} + +static void efx_dissociate(struct efx_nic *efx) +{ + struct efx_nic *other, *next; + + list_del(&efx->node); + efx->primary = NULL; + + list_for_each_entry_safe(other, next, &efx->secondary_list, node) { + list_del(&other->node); + netif_dbg(other, probe, other->net_dev, + "moving to unassociated list\n"); + list_add_tail(&other->node, &efx_unassociated_list); + other->primary = NULL; + } +} + /* This configures the PCI device to enable I/O and DMA. */ static int efx_init_io(struct efx_nic *efx) { struct pci_dev *pci_dev = efx->pci_dev; dma_addr_t dma_mask = efx->type->max_dma_mask; + unsigned int mem_map_size = efx->type->mem_map_size(efx); int rc; netif_dbg(efx, probe, efx->net_dev, "initialising I/O\n"); @@ -1069,9 +1212,11 @@ static int efx_init_io(struct efx_nic *efx) * masks event though they reject 46 bit masks. */ while (dma_mask > 0x7fffffffUL) { - if (pci_dma_supported(pci_dev, dma_mask) && - ((rc = pci_set_dma_mask(pci_dev, dma_mask)) == 0)) - break; + if (dma_supported(&pci_dev->dev, dma_mask)) { + rc = dma_set_mask_and_coherent(&pci_dev->dev, dma_mask); + if (rc == 0) + break; + } dma_mask >>= 1; } if (rc) { @@ -1081,16 +1226,6 @@ static int efx_init_io(struct efx_nic *efx) } netif_dbg(efx, probe, efx->net_dev, "using DMA mask %llx\n", (unsigned long long) dma_mask); - rc = pci_set_consistent_dma_mask(pci_dev, dma_mask); - if (rc) { - /* pci_set_consistent_dma_mask() is not *allowed* to - * fail with a mask that pci_set_dma_mask() accepted, - * but just in case... - */ - netif_err(efx, probe, efx->net_dev, - "failed to set consistent DMA mask\n"); - goto fail2; - } efx->membase_phys = pci_resource_start(efx->pci_dev, EFX_MEM_BAR); rc = pci_request_region(pci_dev, EFX_MEM_BAR, "sfc"); @@ -1100,20 +1235,18 @@ static int efx_init_io(struct efx_nic *efx) rc = -EIO; goto fail3; } - efx->membase = ioremap_nocache(efx->membase_phys, - efx->type->mem_map_size); + efx->membase = ioremap_nocache(efx->membase_phys, mem_map_size); if (!efx->membase) { netif_err(efx, probe, efx->net_dev, "could not map memory BAR at %llx+%x\n", - (unsigned long long)efx->membase_phys, - efx->type->mem_map_size); + (unsigned long long)efx->membase_phys, mem_map_size); rc = -ENOMEM; goto fail4; } netif_dbg(efx, probe, efx->net_dev, "memory BAR at %llx+%x (virtual %p)\n", - (unsigned long long)efx->membase_phys, - efx->type->mem_map_size, efx->membase); + (unsigned long long)efx->membase_phys, mem_map_size, + efx->membase); return 0; @@ -1144,116 +1277,106 @@ static void efx_fini_io(struct efx_nic *efx) pci_disable_device(efx->pci_dev); } -/* Get number of channels wanted. Each channel will have its own IRQ, - * 1 RX queue and/or 2 TX queues. */ -static int efx_wanted_channels(void) +static unsigned int efx_wanted_parallelism(struct efx_nic *efx) { - cpumask_var_t core_mask; - int count; + cpumask_var_t thread_mask; + unsigned int count; int cpu; - if (rss_cpus) - return rss_cpus; + if (rss_cpus) { + count = rss_cpus; + } else { + if (unlikely(!zalloc_cpumask_var(&thread_mask, GFP_KERNEL))) { + netif_warn(efx, probe, efx->net_dev, + "RSS disabled due to allocation failure\n"); + return 1; + } - if (unlikely(!zalloc_cpumask_var(&core_mask, GFP_KERNEL))) { - printk(KERN_WARNING - "sfc: RSS disabled due to allocation failure\n"); - return 1; + count = 0; + for_each_online_cpu(cpu) { + if (!cpumask_test_cpu(cpu, thread_mask)) { + ++count; + cpumask_or(thread_mask, thread_mask, + topology_thread_cpumask(cpu)); + } + } + + free_cpumask_var(thread_mask); } - count = 0; - for_each_online_cpu(cpu) { - if (!cpumask_test_cpu(cpu, core_mask)) { - ++count; - cpumask_or(core_mask, core_mask, - topology_core_cpumask(cpu)); - } + /* If RSS is requested for the PF *and* VFs then we can't write RSS + * table entries that are inaccessible to VFs + */ + if (efx_sriov_wanted(efx) && efx_vf_size(efx) > 1 && + count > efx_vf_size(efx)) { + netif_warn(efx, probe, efx->net_dev, + "Reducing number of RSS channels from %u to %u for " + "VF support. Increase vf-msix-limit to use more " + "channels on the PF.\n", + count, efx_vf_size(efx)); + count = efx_vf_size(efx); } - free_cpumask_var(core_mask); return count; } -static int -efx_init_rx_cpu_rmap(struct efx_nic *efx, struct msix_entry *xentries) -{ -#ifdef CONFIG_RFS_ACCEL - int i, rc; - - efx->net_dev->rx_cpu_rmap = alloc_irq_cpu_rmap(efx->n_rx_channels); - if (!efx->net_dev->rx_cpu_rmap) - return -ENOMEM; - for (i = 0; i < efx->n_rx_channels; i++) { - rc = irq_cpu_rmap_add(efx->net_dev->rx_cpu_rmap, - xentries[i].vector); - if (rc) { - free_irq_cpu_rmap(efx->net_dev->rx_cpu_rmap); - efx->net_dev->rx_cpu_rmap = NULL; - return rc; - } - } -#endif - return 0; -} - /* Probe the number and type of interrupts we are able to obtain, and * the resulting numbers of channels and RX queues. */ static int efx_probe_interrupts(struct efx_nic *efx) { - int max_channels = - min_t(int, efx->type->phys_addr_channels, EFX_MAX_CHANNELS); - int rc, i; + unsigned int extra_channels = 0; + unsigned int i, j; + int rc; + + for (i = 0; i < EFX_MAX_EXTRA_CHANNELS; i++) + if (efx->extra_channel_type[i]) + ++extra_channels; if (efx->interrupt_mode == EFX_INT_MODE_MSIX) { struct msix_entry xentries[EFX_MAX_CHANNELS]; - int n_channels; + unsigned int n_channels; - n_channels = efx_wanted_channels(); + n_channels = efx_wanted_parallelism(efx); if (separate_tx_channels) n_channels *= 2; - n_channels = min(n_channels, max_channels); + n_channels += extra_channels; + n_channels = min(n_channels, efx->max_channels); for (i = 0; i < n_channels; i++) xentries[i].entry = i; - rc = pci_enable_msix(efx->pci_dev, xentries, n_channels); - if (rc > 0) { + rc = pci_enable_msix_range(efx->pci_dev, + xentries, 1, n_channels); + if (rc < 0) { + /* Fall back to single channel MSI */ + efx->interrupt_mode = EFX_INT_MODE_MSI; + netif_err(efx, drv, efx->net_dev, + "could not enable MSI-X\n"); + } else if (rc < n_channels) { netif_err(efx, drv, efx->net_dev, "WARNING: Insufficient MSI-X vectors" - " available (%d < %d).\n", rc, n_channels); + " available (%d < %u).\n", rc, n_channels); netif_err(efx, drv, efx->net_dev, "WARNING: Performance may be reduced.\n"); - EFX_BUG_ON_PARANOID(rc >= n_channels); n_channels = rc; - rc = pci_enable_msix(efx->pci_dev, xentries, - n_channels); } - if (rc == 0) { + if (rc > 0) { efx->n_channels = n_channels; + if (n_channels > extra_channels) + n_channels -= extra_channels; if (separate_tx_channels) { - efx->n_tx_channels = - max(efx->n_channels / 2, 1U); - efx->n_rx_channels = - max(efx->n_channels - - efx->n_tx_channels, 1U); + efx->n_tx_channels = max(n_channels / 2, 1U); + efx->n_rx_channels = max(n_channels - + efx->n_tx_channels, + 1U); } else { - efx->n_tx_channels = efx->n_channels; - efx->n_rx_channels = efx->n_channels; - } - rc = efx_init_rx_cpu_rmap(efx, xentries); - if (rc) { - pci_disable_msix(efx->pci_dev); - return rc; + efx->n_tx_channels = n_channels; + efx->n_rx_channels = n_channels; } - for (i = 0; i < n_channels; i++) + for (i = 0; i < efx->n_channels; i++) efx_get_channel(efx, i)->irq = xentries[i].vector; - } else { - /* Fall back to single channel MSI */ - efx->interrupt_mode = EFX_INT_MODE_MSI; - netif_err(efx, drv, efx->net_dev, - "could not enable MSI-X\n"); } } @@ -1280,7 +1403,145 @@ static int efx_probe_interrupts(struct efx_nic *efx) efx->legacy_irq = efx->pci_dev->irq; } + /* Assign extra channels if possible */ + j = efx->n_channels; + for (i = 0; i < EFX_MAX_EXTRA_CHANNELS; i++) { + if (!efx->extra_channel_type[i]) + continue; + if (efx->interrupt_mode != EFX_INT_MODE_MSIX || + efx->n_channels <= extra_channels) { + efx->extra_channel_type[i]->handle_no_channel(efx); + } else { + --j; + efx_get_channel(efx, j)->type = + efx->extra_channel_type[i]; + } + } + + /* RSS might be usable on VFs even if it is disabled on the PF */ + efx->rss_spread = ((efx->n_rx_channels > 1 || !efx_sriov_wanted(efx)) ? + efx->n_rx_channels : efx_vf_size(efx)); + + return 0; +} + +static int efx_soft_enable_interrupts(struct efx_nic *efx) +{ + struct efx_channel *channel, *end_channel; + int rc; + + BUG_ON(efx->state == STATE_DISABLED); + + efx->irq_soft_enabled = true; + smp_wmb(); + + efx_for_each_channel(channel, efx) { + if (!channel->type->keep_eventq) { + rc = efx_init_eventq(channel); + if (rc) + goto fail; + } + efx_start_eventq(channel); + } + + efx_mcdi_mode_event(efx); + return 0; +fail: + end_channel = channel; + efx_for_each_channel(channel, efx) { + if (channel == end_channel) + break; + efx_stop_eventq(channel); + if (!channel->type->keep_eventq) + efx_fini_eventq(channel); + } + + return rc; +} + +static void efx_soft_disable_interrupts(struct efx_nic *efx) +{ + struct efx_channel *channel; + + if (efx->state == STATE_DISABLED) + return; + + efx_mcdi_mode_poll(efx); + + efx->irq_soft_enabled = false; + smp_wmb(); + + if (efx->legacy_irq) + synchronize_irq(efx->legacy_irq); + + efx_for_each_channel(channel, efx) { + if (channel->irq) + synchronize_irq(channel->irq); + + efx_stop_eventq(channel); + if (!channel->type->keep_eventq) + efx_fini_eventq(channel); + } + + /* Flush the asynchronous MCDI request queue */ + efx_mcdi_flush_async(efx); +} + +static int efx_enable_interrupts(struct efx_nic *efx) +{ + struct efx_channel *channel, *end_channel; + int rc; + + BUG_ON(efx->state == STATE_DISABLED); + + if (efx->eeh_disabled_legacy_irq) { + enable_irq(efx->legacy_irq); + efx->eeh_disabled_legacy_irq = false; + } + + efx->type->irq_enable_master(efx); + + efx_for_each_channel(channel, efx) { + if (channel->type->keep_eventq) { + rc = efx_init_eventq(channel); + if (rc) + goto fail; + } + } + + rc = efx_soft_enable_interrupts(efx); + if (rc) + goto fail; + + return 0; + +fail: + end_channel = channel; + efx_for_each_channel(channel, efx) { + if (channel == end_channel) + break; + if (channel->type->keep_eventq) + efx_fini_eventq(channel); + } + + efx->type->irq_disable_non_ev(efx); + + return rc; +} + +static void efx_disable_interrupts(struct efx_nic *efx) +{ + struct efx_channel *channel; + + efx_soft_disable_interrupts(efx); + + efx_for_each_channel(channel, efx) { + if (channel->type->keep_eventq) + efx_fini_eventq(channel); + } + + efx->type->irq_disable_non_ev(efx); } static void efx_remove_interrupts(struct efx_nic *efx) @@ -1305,10 +1566,16 @@ static void efx_set_channels(struct efx_nic *efx) efx->tx_channel_offset = separate_tx_channels ? efx->n_channels - efx->n_tx_channels : 0; - /* We need to adjust the TX queue numbers if we have separate + /* We need to mark which channels really have RX and TX + * queues, and adjust the TX queue numbers if we have separate * RX-only and TX-only channels. */ efx_for_each_channel(channel, efx) { + if (channel->channel < efx->n_rx_channels) + channel->rx_queue.core_index = channel->channel; + else + channel->rx_queue.core_index = -1; + efx_for_each_channel_tx_queue(tx_queue, channel) tx_queue->queue -= (efx->tx_channel_offset * EFX_TXQ_TYPES); @@ -1331,15 +1598,20 @@ static int efx_probe_nic(struct efx_nic *efx) * in MSI-X interrupts. */ rc = efx_probe_interrupts(efx); if (rc) - goto fail; + goto fail1; + + efx_set_channels(efx); + + rc = efx->type->dimension_resources(efx); + if (rc) + goto fail2; if (efx->n_channels > 1) get_random_bytes(&efx->rx_hash_key, sizeof(efx->rx_hash_key)); for (i = 0; i < ARRAY_SIZE(efx->rx_indir_table); i++) efx->rx_indir_table[i] = - ethtool_rxfh_indir_default(i, efx->n_rx_channels); + ethtool_rxfh_indir_default(i, efx->rss_spread); - efx_set_channels(efx); netif_set_real_num_tx_queues(efx->net_dev, efx->n_tx_channels); netif_set_real_num_rx_queues(efx->net_dev, efx->n_rx_channels); @@ -1349,7 +1621,9 @@ static int efx_probe_nic(struct efx_nic *efx) return 0; -fail: +fail2: + efx_remove_interrupts(efx); +fail1: efx->type->remove(efx); return rc; } @@ -1362,6 +1636,44 @@ static void efx_remove_nic(struct efx_nic *efx) efx->type->remove(efx); } +static int efx_probe_filters(struct efx_nic *efx) +{ + int rc; + + spin_lock_init(&efx->filter_lock); + + rc = efx->type->filter_table_probe(efx); + if (rc) + return rc; + +#ifdef CONFIG_RFS_ACCEL + if (efx->type->offload_features & NETIF_F_NTUPLE) { + efx->rps_flow_id = kcalloc(efx->type->max_rx_ip_filters, + sizeof(*efx->rps_flow_id), + GFP_KERNEL); + if (!efx->rps_flow_id) { + efx->type->filter_table_remove(efx); + return -ENOMEM; + } + } +#endif + + return 0; +} + +static void efx_remove_filters(struct efx_nic *efx) +{ +#ifdef CONFIG_RFS_ACCEL + kfree(efx->rps_flow_id); +#endif + efx->type->filter_table_remove(efx); +} + +static void efx_restore_filters(struct efx_nic *efx) +{ + efx->type->filter_table_restore(efx); +} + /************************************************************************** * * NIC startup/shutdown @@ -1384,22 +1696,28 @@ static int efx_probe_all(struct efx_nic *efx) goto fail2; } - efx->rxq_entries = efx->txq_entries = EFX_DEFAULT_DMAQ_SIZE; - rc = efx_probe_channels(efx); - if (rc) + BUILD_BUG_ON(EFX_DEFAULT_DMAQ_SIZE < EFX_RXQ_MIN_ENT); + if (WARN_ON(EFX_DEFAULT_DMAQ_SIZE < EFX_TXQ_MIN_ENT(efx))) { + rc = -EINVAL; goto fail3; + } + efx->rxq_entries = efx->txq_entries = EFX_DEFAULT_DMAQ_SIZE; rc = efx_probe_filters(efx); if (rc) { netif_err(efx, probe, efx->net_dev, "failed to create filter tables\n"); - goto fail4; + goto fail3; } + rc = efx_probe_channels(efx); + if (rc) + goto fail4; + return 0; fail4: - efx_remove_channels(efx); + efx_remove_filters(efx); fail3: efx_remove_port(efx); fail2: @@ -1408,56 +1726,36 @@ static int efx_probe_all(struct efx_nic *efx) return rc; } -/* Called after previous invocation(s) of efx_stop_all, restarts the - * port, kernel transmit queue, NAPI processing and hardware interrupts, - * and ensures that the port is scheduled to be reconfigured. - * This function is safe to call multiple times when the NIC is in any - * state. */ +/* If the interface is supposed to be running but is not, start + * the hardware and software data path, regular activity for the port + * (MAC statistics, link polling, etc.) and schedule the port to be + * reconfigured. Interrupts must already be enabled. This function + * is safe to call multiple times, so long as the NIC is not disabled. + * Requires the RTNL lock. + */ static void efx_start_all(struct efx_nic *efx) { - struct efx_channel *channel; - EFX_ASSERT_RESET_SERIALISED(efx); + BUG_ON(efx->state == STATE_DISABLED); /* Check that it is appropriate to restart the interface. All * of these flags are safe to read under just the rtnl lock */ - if (efx->port_enabled) - return; - if ((efx->state != STATE_RUNNING) && (efx->state != STATE_INIT)) - return; - if (efx_dev_registered(efx) && !netif_running(efx->net_dev)) + if (efx->port_enabled || !netif_running(efx->net_dev) || + efx->reset_pending) return; - /* Mark the port as enabled so port reconfigurations can start, then - * restart the transmit interface early so the watchdog timer stops */ efx_start_port(efx); + efx_start_datapath(efx); - if (efx_dev_registered(efx) && netif_device_present(efx->net_dev)) - netif_tx_wake_all_queues(efx->net_dev); - - efx_for_each_channel(channel, efx) - efx_start_channel(channel); - - if (efx->legacy_irq) - efx->legacy_irq_enabled = true; - efx_nic_enable_interrupts(efx); - - /* Switch to event based MCDI completions after enabling interrupts. - * If a reset has been scheduled, then we need to stay in polled mode. - * Rather than serialising efx_mcdi_mode_event() [which sleeps] and - * reset_pending [modified from an atomic context], we instead guarantee - * that efx_mcdi_mode_poll() isn't reverted erroneously */ - efx_mcdi_mode_event(efx); - if (efx->reset_pending) - efx_mcdi_mode_poll(efx); - - /* Start the hardware monitor if there is one. Otherwise (we're link - * event driven), we have to poll the PHY because after an event queue - * flush, we could have a missed a link state change */ - if (efx->type->monitor != NULL) { + /* Start the hardware monitor if there is one */ + if (efx->type->monitor != NULL) queue_delayed_work(efx->workqueue, &efx->monitor_work, efx_monitor_interval); - } else { + + /* If link state detection is normally event-driven, we have + * to poll now because we could have missed a change + */ + if (efx_nic_rev(efx) >= EFX_REV_SIENA_A0) { mutex_lock(&efx->mac_lock); if (efx->phy_op->poll(efx)) efx_link_status_changed(efx); @@ -1465,75 +1763,50 @@ static void efx_start_all(struct efx_nic *efx) } efx->type->start_stats(efx); + efx->type->pull_stats(efx); + spin_lock_bh(&efx->stats_lock); + efx->type->update_stats(efx, NULL, NULL); + spin_unlock_bh(&efx->stats_lock); } -/* Flush all delayed work. Should only be called when no more delayed work - * will be scheduled. This doesn't flush pending online resets (efx_reset), - * since we're holding the rtnl_lock at this point. */ -static void efx_flush_all(struct efx_nic *efx) -{ - /* Make sure the hardware monitor is stopped */ - cancel_delayed_work_sync(&efx->monitor_work); - /* Stop scheduled port reconfigurations */ - cancel_work_sync(&efx->mac_work); -} - -/* Quiesce hardware and software without bringing the link down. - * Safe to call multiple times, when the nic and interface is in any - * state. The caller is guaranteed to subsequently be in a position - * to modify any hardware and software state they see fit without - * taking locks. */ +/* Quiesce the hardware and software data path, and regular activity + * for the port without bringing the link down. Safe to call multiple + * times with the NIC in almost any state, but interrupts should be + * enabled. Requires the RTNL lock. + */ static void efx_stop_all(struct efx_nic *efx) { - struct efx_channel *channel; - EFX_ASSERT_RESET_SERIALISED(efx); /* port_enabled can be read safely under the rtnl lock */ if (!efx->port_enabled) return; + /* update stats before we go down so we can accurately count + * rx_nodesc_drops + */ + efx->type->pull_stats(efx); + spin_lock_bh(&efx->stats_lock); + efx->type->update_stats(efx, NULL, NULL); + spin_unlock_bh(&efx->stats_lock); efx->type->stop_stats(efx); - - /* Switch to MCDI polling on Siena before disabling interrupts */ - efx_mcdi_mode_poll(efx); - - /* Disable interrupts and wait for ISR to complete */ - efx_nic_disable_interrupts(efx); - if (efx->legacy_irq) { - synchronize_irq(efx->legacy_irq); - efx->legacy_irq_enabled = false; - } - efx_for_each_channel(channel, efx) { - if (channel->irq) - synchronize_irq(channel->irq); - } - - /* Stop all NAPI processing and synchronous rx refills */ - efx_for_each_channel(channel, efx) - efx_stop_channel(channel); - - /* Stop all asynchronous port reconfigurations. Since all - * event processing has already been stopped, there is no - * window to loose phy events */ efx_stop_port(efx); - /* Flush efx_mac_work(), refill_workqueue, monitor_work */ - efx_flush_all(efx); + /* Stop the kernel transmit interface. This is only valid if + * the device is stopped or detached; otherwise the watchdog + * may fire immediately. + */ + WARN_ON(netif_running(efx->net_dev) && + netif_device_present(efx->net_dev)); + netif_tx_disable(efx->net_dev); - /* Stop the kernel transmit interface late, so the watchdog - * timer isn't ticking over the flush */ - if (efx_dev_registered(efx)) { - netif_tx_stop_all_queues(efx->net_dev); - netif_tx_lock_bh(efx->net_dev); - netif_tx_unlock_bh(efx->net_dev); - } + efx_stop_datapath(efx); } static void efx_remove_all(struct efx_nic *efx) { - efx_remove_filters(efx); efx_remove_channels(efx); + efx_remove_filters(efx); efx_remove_port(efx); efx_remove_nic(efx); } @@ -1544,13 +1817,13 @@ static void efx_remove_all(struct efx_nic *efx) * **************************************************************************/ -static unsigned int irq_mod_ticks(unsigned int usecs, unsigned int resolution) +static unsigned int irq_mod_ticks(unsigned int usecs, unsigned int quantum_ns) { if (usecs == 0) return 0; - if (usecs < resolution) + if (usecs * 1000 < quantum_ns) return 1; /* never round down to 0 */ - return usecs / resolution; + return usecs * 1000 / quantum_ns; } /* Set interrupt moderation parameters */ @@ -1559,14 +1832,20 @@ int efx_init_irq_moderation(struct efx_nic *efx, unsigned int tx_usecs, bool rx_may_override_tx) { struct efx_channel *channel; - unsigned tx_ticks = irq_mod_ticks(tx_usecs, EFX_IRQ_MOD_RESOLUTION); - unsigned rx_ticks = irq_mod_ticks(rx_usecs, EFX_IRQ_MOD_RESOLUTION); + unsigned int irq_mod_max = DIV_ROUND_UP(efx->type->timer_period_max * + efx->timer_quantum_ns, + 1000); + unsigned int tx_ticks; + unsigned int rx_ticks; EFX_ASSERT_RESET_SERIALISED(efx); - if (tx_ticks > EFX_IRQ_MOD_MAX || rx_ticks > EFX_IRQ_MOD_MAX) + if (tx_usecs > irq_mod_max || rx_usecs > irq_mod_max) return -EINVAL; + tx_ticks = irq_mod_ticks(tx_usecs, efx->timer_quantum_ns); + rx_ticks = irq_mod_ticks(rx_usecs, efx->timer_quantum_ns); + if (tx_ticks != rx_ticks && efx->tx_channel_offset == 0 && !rx_may_override_tx) { netif_err(efx, drv, efx->net_dev, "Channels are shared. " @@ -1589,8 +1868,14 @@ int efx_init_irq_moderation(struct efx_nic *efx, unsigned int tx_usecs, void efx_get_irq_moderation(struct efx_nic *efx, unsigned int *tx_usecs, unsigned int *rx_usecs, bool *rx_adaptive) { + /* We must round up when converting ticks to microseconds + * because we round down when converting the other way. + */ + *rx_adaptive = efx->irq_rx_adaptive; - *rx_usecs = efx->irq_rx_moderation * EFX_IRQ_MOD_RESOLUTION; + *rx_usecs = DIV_ROUND_UP(efx->irq_rx_moderation * + efx->timer_quantum_ns, + 1000); /* If channels are shared between RX and TX, so is IRQ * moderation. Otherwise, IRQ moderation is the same for all @@ -1599,9 +1884,10 @@ void efx_get_irq_moderation(struct efx_nic *efx, unsigned int *tx_usecs, if (efx->tx_channel_offset == 0) *tx_usecs = *rx_usecs; else - *tx_usecs = + *tx_usecs = DIV_ROUND_UP( efx->channel[efx->tx_channel_offset]->irq_moderation * - EFX_IRQ_MOD_RESOLUTION; + efx->timer_quantum_ns, + 1000); } /************************************************************************** @@ -1648,7 +1934,10 @@ static int efx_ioctl(struct net_device *net_dev, struct ifreq *ifr, int cmd) struct efx_nic *efx = netdev_priv(net_dev); struct mii_ioctl_data *data = if_mii(ifr); - EFX_ASSERT_RESET_SERIALISED(efx); + if (cmd == SIOCSHWTSTAMP) + return efx_ptp_set_ts_config(efx, ifr); + if (cmd == SIOCGHWTSTAMP) + return efx_ptp_get_ts_config(efx, ifr); /* Convert phy_id from older PRTAD/DEVAD format */ if ((cmd == SIOCGMIIREG || cmd == SIOCSMIIREG) && @@ -1664,15 +1953,21 @@ static int efx_ioctl(struct net_device *net_dev, struct ifreq *ifr, int cmd) * **************************************************************************/ +static void efx_init_napi_channel(struct efx_channel *channel) +{ + struct efx_nic *efx = channel->efx; + + channel->napi_dev = efx->net_dev; + netif_napi_add(channel->napi_dev, &channel->napi_str, + efx_poll, napi_weight); +} + static void efx_init_napi(struct efx_nic *efx) { struct efx_channel *channel; - efx_for_each_channel(channel, efx) { - channel->napi_dev = efx->net_dev; - netif_napi_add(channel->napi_dev, &channel->napi_str, - efx_poll, napi_weight); - } + efx_for_each_channel(channel, efx) + efx_init_napi_channel(channel); } static void efx_fini_napi_channel(struct efx_channel *channel) @@ -1723,13 +2018,14 @@ static void efx_netpoll(struct net_device *net_dev) static int efx_net_open(struct net_device *net_dev) { struct efx_nic *efx = netdev_priv(net_dev); - EFX_ASSERT_RESET_SERIALISED(efx); + int rc; netif_dbg(efx, ifup, efx->net_dev, "opening device on CPU %d\n", raw_smp_processor_id()); - if (efx->state == STATE_DISABLED) - return -EIO; + rc = efx_check_disabled(efx); + if (rc) + return rc; if (efx->phy_mode & PHY_MODE_SPECIAL) return -EBUSY; if (efx_mcdi_poll_reboot(efx) && efx_reset(efx, RESET_TYPE_ALL)) @@ -1740,6 +2036,7 @@ static int efx_net_open(struct net_device *net_dev) efx_link_status_changed(efx); efx_start_all(efx); + efx_selftest_async_start(efx); return 0; } @@ -1754,48 +2051,22 @@ static int efx_net_stop(struct net_device *net_dev) netif_dbg(efx, ifdown, efx->net_dev, "closing on CPU %d\n", raw_smp_processor_id()); - if (efx->state != STATE_DISABLED) { - /* Stop the device and flush all the channels */ - efx_stop_all(efx); - efx_fini_channels(efx); - efx_init_channels(efx); - } + /* Stop the device and flush all the channels */ + efx_stop_all(efx); return 0; } /* Context: process, dev_base_lock or RTNL held, non-blocking. */ -static struct rtnl_link_stats64 *efx_net_stats(struct net_device *net_dev, struct rtnl_link_stats64 *stats) +static struct rtnl_link_stats64 *efx_net_stats(struct net_device *net_dev, + struct rtnl_link_stats64 *stats) { struct efx_nic *efx = netdev_priv(net_dev); - struct efx_mac_stats *mac_stats = &efx->mac_stats; spin_lock_bh(&efx->stats_lock); - efx->type->update_stats(efx); + efx->type->update_stats(efx, NULL, stats); spin_unlock_bh(&efx->stats_lock); - stats->rx_packets = mac_stats->rx_packets; - stats->tx_packets = mac_stats->tx_packets; - stats->rx_bytes = mac_stats->rx_bytes; - stats->tx_bytes = mac_stats->tx_bytes; - stats->rx_dropped = efx->n_rx_nodesc_drop_cnt; - stats->multicast = mac_stats->rx_multicast; - stats->collisions = mac_stats->tx_collision; - stats->rx_length_errors = (mac_stats->rx_gtjumbo + - mac_stats->rx_length_error); - stats->rx_crc_errors = mac_stats->rx_bad; - stats->rx_frame_errors = mac_stats->rx_align_error; - stats->rx_fifo_errors = mac_stats->rx_overflow; - stats->rx_missed_errors = mac_stats->rx_missed; - stats->tx_window_errors = mac_stats->tx_late_collision; - - stats->rx_errors = (stats->rx_length_errors + - stats->rx_crc_errors + - stats->rx_frame_errors + - mac_stats->rx_symbol_error); - stats->tx_errors = (stats->tx_window_errors + - mac_stats->tx_bad); - return stats; } @@ -1816,85 +2087,57 @@ static void efx_watchdog(struct net_device *net_dev) static int efx_change_mtu(struct net_device *net_dev, int new_mtu) { struct efx_nic *efx = netdev_priv(net_dev); - int rc = 0; - - EFX_ASSERT_RESET_SERIALISED(efx); + int rc; + rc = efx_check_disabled(efx); + if (rc) + return rc; if (new_mtu > EFX_MAX_MTU) return -EINVAL; - efx_stop_all(efx); - netif_dbg(efx, drv, efx->net_dev, "changing MTU to %d\n", new_mtu); - efx_fini_channels(efx); + efx_device_detach_sync(efx); + efx_stop_all(efx); mutex_lock(&efx->mac_lock); - /* Reconfigure the MAC before enabling the dma queues so that - * the RX buffers don't overflow */ net_dev->mtu = new_mtu; - efx->mac_op->reconfigure(efx); + efx->type->reconfigure_mac(efx); mutex_unlock(&efx->mac_lock); - efx_init_channels(efx); - efx_start_all(efx); - return rc; + netif_device_attach(efx->net_dev); + return 0; } static int efx_set_mac_address(struct net_device *net_dev, void *data) { struct efx_nic *efx = netdev_priv(net_dev); struct sockaddr *addr = data; - char *new_addr = addr->sa_data; - - EFX_ASSERT_RESET_SERIALISED(efx); + u8 *new_addr = addr->sa_data; if (!is_valid_ether_addr(new_addr)) { netif_err(efx, drv, efx->net_dev, "invalid ethernet MAC address requested: %pM\n", new_addr); - return -EINVAL; + return -EADDRNOTAVAIL; } - memcpy(net_dev->dev_addr, new_addr, net_dev->addr_len); + ether_addr_copy(net_dev->dev_addr, new_addr); + efx_sriov_mac_address_changed(efx); /* Reconfigure the MAC */ mutex_lock(&efx->mac_lock); - efx->mac_op->reconfigure(efx); + efx->type->reconfigure_mac(efx); mutex_unlock(&efx->mac_lock); return 0; } /* Context: netif_addr_lock held, BHs disabled. */ -static void efx_set_multicast_list(struct net_device *net_dev) +static void efx_set_rx_mode(struct net_device *net_dev) { struct efx_nic *efx = netdev_priv(net_dev); - struct netdev_hw_addr *ha; - union efx_multicast_hash *mc_hash = &efx->multicast_hash; - u32 crc; - int bit; - - efx->promiscuous = !!(net_dev->flags & IFF_PROMISC); - - /* Build multicast hash table */ - if (efx->promiscuous || (net_dev->flags & IFF_ALLMULTI)) { - memset(mc_hash, 0xff, sizeof(*mc_hash)); - } else { - memset(mc_hash, 0x00, sizeof(*mc_hash)); - netdev_for_each_mc_addr(ha, net_dev) { - crc = ether_crc_le(ETH_ALEN, ha->addr); - bit = crc & (EFX_MCAST_HASH_ENTRIES - 1); - set_bit_le(bit, mc_hash->byte); - } - - /* 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); - } if (efx->port_enabled) queue_work(efx->workqueue, &efx->mac_work); @@ -1907,12 +2150,12 @@ static int efx_set_features(struct net_device *net_dev, netdev_features_t data) /* If disabling RX n-tuple filtering, clear existing filters */ if (net_dev->features & ~data & NETIF_F_NTUPLE) - efx_filter_clear_rx(efx, EFX_FILTER_PRI_MANUAL); + return efx->type->filter_clear_rx(efx, EFX_FILTER_PRI_MANUAL); return 0; } -static const struct net_device_ops efx_netdev_ops = { +static const struct net_device_ops efx_farch_netdev_ops = { .ndo_open = efx_net_open, .ndo_stop = efx_net_stop, .ndo_get_stats64 = efx_net_stats, @@ -1922,8 +2165,14 @@ static const struct net_device_ops efx_netdev_ops = { .ndo_do_ioctl = efx_ioctl, .ndo_change_mtu = efx_change_mtu, .ndo_set_mac_address = efx_set_mac_address, - .ndo_set_rx_mode = efx_set_multicast_list, + .ndo_set_rx_mode = efx_set_rx_mode, .ndo_set_features = efx_set_features, +#ifdef CONFIG_SFC_SRIOV + .ndo_set_vf_mac = efx_sriov_set_vf_mac, + .ndo_set_vf_vlan = efx_sriov_set_vf_vlan, + .ndo_set_vf_spoofchk = efx_sriov_set_vf_spoofchk, + .ndo_get_vf_config = efx_sriov_get_vf_config, +#endif #ifdef CONFIG_NET_POLL_CONTROLLER .ndo_poll_controller = efx_netpoll, #endif @@ -1933,6 +2182,26 @@ static const struct net_device_ops efx_netdev_ops = { #endif }; +static const struct net_device_ops efx_ef10_netdev_ops = { + .ndo_open = efx_net_open, + .ndo_stop = efx_net_stop, + .ndo_get_stats64 = efx_net_stats, + .ndo_tx_timeout = efx_watchdog, + .ndo_start_xmit = efx_hard_start_xmit, + .ndo_validate_addr = eth_validate_addr, + .ndo_do_ioctl = efx_ioctl, + .ndo_change_mtu = efx_change_mtu, + .ndo_set_mac_address = efx_set_mac_address, + .ndo_set_rx_mode = efx_set_rx_mode, + .ndo_set_features = efx_set_features, +#ifdef CONFIG_NET_POLL_CONTROLLER + .ndo_poll_controller = efx_netpoll, +#endif +#ifdef CONFIG_RFS_ACCEL + .ndo_rx_flow_steer = efx_filter_rfs, +#endif +}; + static void efx_update_name(struct efx_nic *efx) { strcpy(efx->name, efx->net_dev->name); @@ -1943,9 +2212,10 @@ static void efx_update_name(struct efx_nic *efx) static int efx_netdev_event(struct notifier_block *this, unsigned long event, void *ptr) { - struct net_device *net_dev = ptr; + struct net_device *net_dev = netdev_notifier_info_to_dev(ptr); - if (net_dev->netdev_ops == &efx_netdev_ops && + if ((net_dev->netdev_ops == &efx_farch_netdev_ops || + net_dev->netdev_ops == &efx_ef10_netdev_ops) && event == NETDEV_CHANGENAME) efx_update_name(netdev_priv(net_dev)); @@ -1962,7 +2232,7 @@ show_phy_type(struct device *dev, struct device_attribute *attr, char *buf) struct efx_nic *efx = pci_get_drvdata(to_pci_dev(dev)); return sprintf(buf, "%d\n", efx->phy_type); } -static DEVICE_ATTR(phy_type, 0644, show_phy_type, NULL); +static DEVICE_ATTR(phy_type, 0444, show_phy_type, NULL); static int efx_register_netdev(struct efx_nic *efx) { @@ -1972,20 +2242,38 @@ static int efx_register_netdev(struct efx_nic *efx) net_dev->watchdog_timeo = 5 * HZ; net_dev->irq = efx->pci_dev->irq; - net_dev->netdev_ops = &efx_netdev_ops; - SET_ETHTOOL_OPS(net_dev, &efx_ethtool_ops); - - /* Clear MAC statistics */ - efx->mac_op->update_stats(efx); - memset(&efx->mac_stats, 0, sizeof(efx->mac_stats)); + if (efx_nic_rev(efx) >= EFX_REV_HUNT_A0) { + net_dev->netdev_ops = &efx_ef10_netdev_ops; + net_dev->priv_flags |= IFF_UNICAST_FLT; + } else { + net_dev->netdev_ops = &efx_farch_netdev_ops; + } + net_dev->ethtool_ops = &efx_ethtool_ops; + net_dev->gso_max_segs = EFX_TSO_MAX_SEGS; rtnl_lock(); + /* Enable resets to be scheduled and check whether any were + * already requested. If so, the NIC is probably hosed so we + * abort. + */ + efx->state = STATE_READY; + smp_mb(); /* ensure we change state before checking reset_pending */ + if (efx->reset_pending) { + netif_err(efx, probe, efx->net_dev, + "aborting probe due to scheduled reset\n"); + rc = -EIO; + goto fail_locked; + } + rc = dev_alloc_name(net_dev, net_dev->name); if (rc < 0) goto fail_locked; efx_update_name(efx); + /* Always start with carrier off; PHY events will detect the link */ + netif_carrier_off(net_dev); + rc = register_netdevice(net_dev); if (rc) goto fail_locked; @@ -1996,8 +2284,7 @@ static int efx_register_netdev(struct efx_nic *efx) efx_init_tx_queue_core_txq(tx_queue); } - /* Always start with carrier off; PHY events will detect the link */ - netif_carrier_off(efx->net_dev); + efx_associate(efx); rtnl_unlock(); @@ -2010,39 +2297,31 @@ static int efx_register_netdev(struct efx_nic *efx) return 0; +fail_registered: + rtnl_lock(); + efx_dissociate(efx); + unregister_netdevice(net_dev); fail_locked: + efx->state = STATE_UNINIT; rtnl_unlock(); netif_err(efx, drv, efx->net_dev, "could not register net dev\n"); return rc; - -fail_registered: - unregister_netdev(net_dev); - return rc; } static void efx_unregister_netdev(struct efx_nic *efx) { - struct efx_channel *channel; - struct efx_tx_queue *tx_queue; - if (!efx->net_dev) return; BUG_ON(netdev_priv(efx->net_dev) != efx); - /* Free up any skbs still remaining. This has to happen before - * we try to unregister the netdev as running their destructors - * may be needed to get the device ref. count to 0. */ - efx_for_each_channel(channel, efx) { - efx_for_each_channel_tx_queue(tx_queue, channel) - efx_release_tx_buffers(tx_queue); - } + strlcpy(efx->name, pci_name(efx->pci_dev), sizeof(efx->name)); + device_remove_file(&efx->pci_dev->dev, &dev_attr_phy_type); - if (efx_dev_registered(efx)) { - strlcpy(efx->name, pci_name(efx->pci_dev), sizeof(efx->name)); - device_remove_file(&efx->pci_dev->dev, &dev_attr_phy_type); - unregister_netdev(efx->net_dev); - } + rtnl_lock(); + unregister_netdevice(efx->net_dev); + efx->state = STATE_UNINIT; + rtnl_unlock(); } /************************************************************************** @@ -2057,10 +2336,13 @@ void efx_reset_down(struct efx_nic *efx, enum reset_type method) { EFX_ASSERT_RESET_SERIALISED(efx); + if (method == RESET_TYPE_MCDI_TIMEOUT) + efx->type->prepare_flr(efx); + efx_stop_all(efx); - mutex_lock(&efx->mac_lock); + efx_disable_interrupts(efx); - efx_fini_channels(efx); + mutex_lock(&efx->mac_lock); if (efx->port_initialized && method != RESET_TYPE_INVISIBLE) efx->phy_op->fini(efx); efx->type->fini(efx); @@ -2077,6 +2359,10 @@ int efx_reset_up(struct efx_nic *efx, enum reset_type method, bool ok) EFX_ASSERT_RESET_SERIALISED(efx); + if (method == RESET_TYPE_MCDI_TIMEOUT) + efx->type->finish_flr(efx); + + /* Ensure that SRAM is initialised even if we're disabling the device */ rc = efx->type->init(efx); if (rc) { netif_err(efx, drv, efx->net_dev, "failed to initialise NIC\n"); @@ -2095,10 +2381,11 @@ int efx_reset_up(struct efx_nic *efx, enum reset_type method, bool ok) "could not restore PHY settings\n"); } - efx->mac_op->reconfigure(efx); - - efx_init_channels(efx); + rc = efx_enable_interrupts(efx); + if (rc) + goto fail; efx_restore_filters(efx); + efx_sriov_reset(efx); mutex_unlock(&efx->mac_lock); @@ -2127,7 +2414,7 @@ int efx_reset(struct efx_nic *efx, enum reset_type method) netif_info(efx, drv, efx->net_dev, "resetting (%s)\n", RESET_TYPE(method)); - netif_device_detach(efx->net_dev); + efx_device_detach_sync(efx); efx_reset_down(efx, method); rc = efx->type->reset(efx, method); @@ -2139,7 +2426,10 @@ int efx_reset(struct efx_nic *efx, enum reset_type method) /* Clear flags for the scopes we covered. We assume the NIC and * driver are now quiescent so that there is no race here. */ - efx->reset_pending &= -(1 << (method + 1)); + if (method < RESET_TYPE_MAX_METHOD) + efx->reset_pending &= -(1 << (method + 1)); + else /* it doesn't fit into the well-ordered scope hierarchy */ + __clear_bit(method, &efx->reset_pending); /* Reinitialise bus-mastering, which may have been turned off before * the reset was scheduled. This is still appropriate, even in the @@ -2149,7 +2439,9 @@ int efx_reset(struct efx_nic *efx, enum reset_type method) out: /* Leave device stopped if necessary */ - disabled = rc || method == RESET_TYPE_DISABLE; + disabled = rc || + method == RESET_TYPE_DISABLE || + method == RESET_TYPE_RECOVER_OR_DISABLE; rc2 = efx_reset_up(efx, method, !disabled); if (rc2) { disabled = true; @@ -2168,27 +2460,82 @@ out: return rc; } +/* Try recovery mechanisms. + * For now only EEH is supported. + * Returns 0 if the recovery mechanisms are unsuccessful. + * Returns a non-zero value otherwise. + */ +int efx_try_recovery(struct efx_nic *efx) +{ +#ifdef CONFIG_EEH + /* A PCI error can occur and not be seen by EEH because nothing + * happens on the PCI bus. In this case the driver may fail and + * schedule a 'recover or reset', leading to this recovery handler. + * Manually call the eeh failure check function. + */ + struct eeh_dev *eehdev = + of_node_to_eeh_dev(pci_device_to_OF_node(efx->pci_dev)); + + if (eeh_dev_check_failure(eehdev)) { + /* The EEH mechanisms will handle the error and reset the + * device if necessary. + */ + return 1; + } +#endif + return 0; +} + +static void efx_wait_for_bist_end(struct efx_nic *efx) +{ + int i; + + for (i = 0; i < BIST_WAIT_DELAY_COUNT; ++i) { + if (efx_mcdi_poll_reboot(efx)) + goto out; + msleep(BIST_WAIT_DELAY_MS); + } + + netif_err(efx, drv, efx->net_dev, "Warning: No MC reboot after BIST mode\n"); +out: + /* Either way unset the BIST flag. If we found no reboot we probably + * won't recover, but we should try. + */ + efx->mc_bist_for_other_fn = false; +} + /* The worker thread exists so that code that cannot sleep can * schedule a reset for later. */ static void efx_reset_work(struct work_struct *data) { struct efx_nic *efx = container_of(data, struct efx_nic, reset_work); - unsigned long pending = ACCESS_ONCE(efx->reset_pending); + unsigned long pending; + enum reset_type method; - if (!pending) + pending = ACCESS_ONCE(efx->reset_pending); + method = fls(pending) - 1; + + if (method == RESET_TYPE_MC_BIST) + efx_wait_for_bist_end(efx); + + if ((method == RESET_TYPE_RECOVER_OR_DISABLE || + method == RESET_TYPE_RECOVER_OR_ALL) && + efx_try_recovery(efx)) return; - /* If we're not RUNNING then don't reset. Leave the reset_pending - * flags set so that efx_pci_probe_main will be retried */ - if (efx->state != STATE_RUNNING) { - netif_info(efx, drv, efx->net_dev, - "scheduled reset quenched. NIC not RUNNING\n"); + if (!pending) return; - } rtnl_lock(); - (void)efx_reset(efx, fls(pending) - 1); + + /* We checked the state in efx_schedule_reset() but it may + * have changed by now. Now that we have the RTNL lock, + * it cannot change again. + */ + if (efx->state == STATE_READY) + (void)efx_reset(efx, method); + rtnl_unlock(); } @@ -2196,11 +2543,22 @@ void efx_schedule_reset(struct efx_nic *efx, enum reset_type type) { enum reset_type method; + if (efx->state == STATE_RECOVERY) { + netif_dbg(efx, drv, efx->net_dev, + "recovering: skip scheduling %s reset\n", + RESET_TYPE(type)); + return; + } + switch (type) { case RESET_TYPE_INVISIBLE: case RESET_TYPE_ALL: + case RESET_TYPE_RECOVER_OR_ALL: case RESET_TYPE_WORLD: case RESET_TYPE_DISABLE: + case RESET_TYPE_RECOVER_OR_DISABLE: + case RESET_TYPE_MC_BIST: + case RESET_TYPE_MCDI_TIMEOUT: method = type; netif_dbg(efx, drv, efx->net_dev, "scheduling %s reset\n", RESET_TYPE(method)); @@ -2214,6 +2572,13 @@ void efx_schedule_reset(struct efx_nic *efx, enum reset_type type) } set_bit(method, &efx->reset_pending); + smp_mb(); /* ensure we change reset_pending before checking state */ + + /* If we're not READY then just leave the flags set as the cue + * to abort probing or reschedule the reset later. + */ + if (ACCESS_ONCE(efx->state) != STATE_READY) + return; /* efx_process_channel() will no longer read events once a * reset is scheduled. So switch back to poll'd MCDI completions. */ @@ -2240,6 +2605,8 @@ static DEFINE_PCI_DEVICE_TABLE(efx_pci_table) = { .driver_data = (unsigned long) &siena_a0_nic_type}, {PCI_DEVICE(PCI_VENDOR_ID_SOLARFLARE, 0x0813), /* SFL9021 */ .driver_data = (unsigned long) &siena_a0_nic_type}, + {PCI_DEVICE(PCI_VENDOR_ID_SOLARFLARE, 0x0903), /* SFC9120 PF */ + .driver_data = (unsigned long) &efx_hunt_a0_nic_type}, {0} /* end of list */ }; @@ -2279,42 +2646,49 @@ static const struct efx_phy_operations efx_dummy_phy_operations = { /* This zeroes out and then fills in the invariants in a struct * efx_nic (including all sub-structures). */ -static int efx_init_struct(struct efx_nic *efx, const struct efx_nic_type *type, +static int efx_init_struct(struct efx_nic *efx, struct pci_dev *pci_dev, struct net_device *net_dev) { int i; /* Initialise common structures */ - memset(efx, 0, sizeof(*efx)); + INIT_LIST_HEAD(&efx->node); + INIT_LIST_HEAD(&efx->secondary_list); spin_lock_init(&efx->biu_lock); #ifdef CONFIG_SFC_MTD INIT_LIST_HEAD(&efx->mtd_list); #endif INIT_WORK(&efx->reset_work, efx_reset_work); INIT_DELAYED_WORK(&efx->monitor_work, efx_monitor); + INIT_DELAYED_WORK(&efx->selftest_work, efx_selftest_async_work); efx->pci_dev = pci_dev; efx->msg_enable = debug; - efx->state = STATE_INIT; + efx->state = STATE_UNINIT; strlcpy(efx->name, pci_name(pci_dev), sizeof(efx->name)); efx->net_dev = net_dev; + efx->rx_prefix_size = efx->type->rx_prefix_size; + efx->rx_ip_align = + NET_IP_ALIGN ? (efx->rx_prefix_size + NET_IP_ALIGN) % 4 : 0; + efx->rx_packet_hash_offset = + efx->type->rx_hash_offset - efx->type->rx_prefix_size; + efx->rx_packet_ts_offset = + efx->type->rx_ts_offset - efx->type->rx_prefix_size; spin_lock_init(&efx->stats_lock); mutex_init(&efx->mac_lock); - efx->mac_op = type->default_mac_ops; efx->phy_op = &efx_dummy_phy_operations; efx->mdio.dev = net_dev; INIT_WORK(&efx->mac_work, efx_mac_work); + init_waitqueue_head(&efx->flush_wq); for (i = 0; i < EFX_MAX_CHANNELS; i++) { efx->channel[i] = efx_alloc_channel(efx, i, NULL); if (!efx->channel[i]) goto fail; + efx->msi_context[i].efx = efx; + efx->msi_context[i].index = i; } - efx->type = type; - - EFX_BUG_ON_PARANOID(efx->type->phys_addr_channels > EFX_MAX_CHANNELS); - /* Higher numbered interrupt modes are less capable! */ efx->interrupt_mode = max(efx->type->max_interrupt_mode, interrupt_mode); @@ -2340,6 +2714,8 @@ static void efx_fini_struct(struct efx_nic *efx) for (i = 0; i < EFX_MAX_CHANNELS; i++) kfree(efx->channel[i]); + kfree(efx->vpd_sn); + if (efx->workqueue) { destroy_workqueue(efx->workqueue); efx->workqueue = NULL; @@ -2357,12 +2733,14 @@ static void efx_fini_struct(struct efx_nic *efx) */ static void efx_pci_remove_main(struct efx_nic *efx) { -#ifdef CONFIG_RFS_ACCEL - free_irq_cpu_rmap(efx->net_dev->rx_cpu_rmap); - efx->net_dev->rx_cpu_rmap = NULL; -#endif + /* Flush reset_work. It can no longer be scheduled since we + * are not READY. + */ + BUG_ON(efx->state == STATE_READY); + cancel_work_sync(&efx->reset_work); + + efx_disable_interrupts(efx); efx_nic_fini_interrupt(efx); - efx_fini_channels(efx); efx_fini_port(efx); efx->type->fini(efx); efx_fini_napi(efx); @@ -2382,32 +2760,100 @@ static void efx_pci_remove(struct pci_dev *pci_dev) /* Mark the NIC as fini, then stop the interface */ rtnl_lock(); - efx->state = STATE_FINI; + efx_dissociate(efx); dev_close(efx->net_dev); - - /* Allow any queued efx_resets() to complete */ + efx_disable_interrupts(efx); rtnl_unlock(); + efx_sriov_fini(efx); efx_unregister_netdev(efx); efx_mtd_remove(efx); - /* Wait for any scheduled resets to complete. No more will be - * scheduled from this point because efx_stop_all() has been - * called, we are no longer registered with driverlink, and - * the net_device's have been removed. */ - cancel_work_sync(&efx->reset_work); - efx_pci_remove_main(efx); efx_fini_io(efx); netif_dbg(efx, drv, efx->net_dev, "shutdown successful\n"); - pci_set_drvdata(pci_dev, NULL); efx_fini_struct(efx); free_netdev(efx->net_dev); + + pci_disable_pcie_error_reporting(pci_dev); }; +/* NIC VPD information + * Called during probe to display the part number of the + * installed NIC. VPD is potentially very large but this should + * always appear within the first 512 bytes. + */ +#define SFC_VPD_LEN 512 +static void efx_probe_vpd_strings(struct efx_nic *efx) +{ + struct pci_dev *dev = efx->pci_dev; + char vpd_data[SFC_VPD_LEN]; + ssize_t vpd_size; + int ro_start, ro_size, i, j; + + /* Get the vpd data from the device */ + vpd_size = pci_read_vpd(dev, 0, sizeof(vpd_data), vpd_data); + if (vpd_size <= 0) { + netif_err(efx, drv, efx->net_dev, "Unable to read VPD\n"); + return; + } + + /* Get the Read only section */ + ro_start = pci_vpd_find_tag(vpd_data, 0, vpd_size, PCI_VPD_LRDT_RO_DATA); + if (ro_start < 0) { + netif_err(efx, drv, efx->net_dev, "VPD Read-only not found\n"); + return; + } + + ro_size = pci_vpd_lrdt_size(&vpd_data[ro_start]); + j = ro_size; + i = ro_start + PCI_VPD_LRDT_TAG_SIZE; + if (i + j > vpd_size) + j = vpd_size - i; + + /* Get the Part number */ + i = pci_vpd_find_info_keyword(vpd_data, i, j, "PN"); + if (i < 0) { + netif_err(efx, drv, efx->net_dev, "Part number not found\n"); + return; + } + + j = pci_vpd_info_field_size(&vpd_data[i]); + i += PCI_VPD_INFO_FLD_HDR_SIZE; + if (i + j > vpd_size) { + netif_err(efx, drv, efx->net_dev, "Incomplete part number\n"); + return; + } + + netif_info(efx, drv, efx->net_dev, + "Part Number : %.*s\n", j, &vpd_data[i]); + + i = ro_start + PCI_VPD_LRDT_TAG_SIZE; + j = ro_size; + i = pci_vpd_find_info_keyword(vpd_data, i, j, "SN"); + if (i < 0) { + netif_err(efx, drv, efx->net_dev, "Serial number not found\n"); + return; + } + + j = pci_vpd_info_field_size(&vpd_data[i]); + i += PCI_VPD_INFO_FLD_HDR_SIZE; + if (i + j > vpd_size) { + netif_err(efx, drv, efx->net_dev, "Incomplete serial number\n"); + return; + } + + efx->vpd_sn = kmalloc(j + 1, GFP_KERNEL); + if (!efx->vpd_sn) + return; + + snprintf(efx->vpd_sn, j + 1, "%s", &vpd_data[i]); +} + + /* Main body of NIC initialisation * This is called at module load (or hotplug insertion, theoretically). */ @@ -2436,16 +2882,18 @@ static int efx_pci_probe_main(struct efx_nic *efx) goto fail4; } - efx_init_channels(efx); - rc = efx_nic_init_interrupt(efx); if (rc) goto fail5; + rc = efx_enable_interrupts(efx); + if (rc) + goto fail6; return 0; + fail6: + efx_nic_fini_interrupt(efx); fail5: - efx_fini_channels(efx); efx_fini_port(efx); fail4: efx->type->fini(efx); @@ -2459,29 +2907,30 @@ static int efx_pci_probe_main(struct efx_nic *efx) /* NIC initialisation * * This is called at module load (or hotplug insertion, - * theoretically). It sets up PCI mappings, tests and resets the NIC, + * theoretically). It sets up PCI mappings, resets the NIC, * sets up and registers the network devices with the kernel and hooks * the interrupt service routine. It does not prepare the device for * transmission; this is left to the first time one of the network * interfaces is brought up (i.e. efx_net_open). */ -static int __devinit efx_pci_probe(struct pci_dev *pci_dev, - const struct pci_device_id *entry) +static int efx_pci_probe(struct pci_dev *pci_dev, + const struct pci_device_id *entry) { - const struct efx_nic_type *type = (const struct efx_nic_type *) entry->driver_data; struct net_device *net_dev; struct efx_nic *efx; - int i, rc; + int rc; /* Allocate and initialise a struct net_device and struct efx_nic */ net_dev = alloc_etherdev_mqs(sizeof(*efx), EFX_MAX_CORE_TX_QUEUES, EFX_MAX_RX_QUEUES); if (!net_dev) return -ENOMEM; - net_dev->features |= (type->offload_features | NETIF_F_SG | + efx = netdev_priv(net_dev); + efx->type = (const struct efx_nic_type *) entry->driver_data; + net_dev->features |= (efx->type->offload_features | NETIF_F_SG | NETIF_F_HIGHDMA | NETIF_F_TSO | NETIF_F_RXCSUM); - if (type->offload_features & NETIF_F_V6_CSUM) + if (efx->type->offload_features & NETIF_F_V6_CSUM) net_dev->features |= NETIF_F_TSO6; /* Mask for features that also apply to VLAN devices */ net_dev->vlan_features |= (NETIF_F_ALL_CSUM | NETIF_F_SG | @@ -2489,75 +2938,54 @@ static int __devinit efx_pci_probe(struct pci_dev *pci_dev, NETIF_F_RXCSUM); /* All offloads can be toggled */ net_dev->hw_features = net_dev->features & ~NETIF_F_HIGHDMA; - efx = netdev_priv(net_dev); pci_set_drvdata(pci_dev, efx); SET_NETDEV_DEV(net_dev, &pci_dev->dev); - rc = efx_init_struct(efx, type, pci_dev, net_dev); + rc = efx_init_struct(efx, pci_dev, net_dev); if (rc) goto fail1; netif_info(efx, probe, efx->net_dev, "Solarflare NIC detected\n"); + efx_probe_vpd_strings(efx); + /* Set up basic I/O (BAR mappings etc) */ rc = efx_init_io(efx); if (rc) goto fail2; - /* No serialisation is required with the reset path because - * we're in STATE_INIT. */ - for (i = 0; i < 5; i++) { - rc = efx_pci_probe_main(efx); - - /* Serialise against efx_reset(). No more resets will be - * scheduled since efx_stop_all() has been called, and we - * have not and never have been registered with either - * the rtnetlink or driverlink layers. */ - cancel_work_sync(&efx->reset_work); - - if (rc == 0) { - if (efx->reset_pending) { - /* If there was a scheduled reset during - * probe, the NIC is probably hosed anyway */ - efx_pci_remove_main(efx); - rc = -EIO; - } else { - break; - } - } - - /* Retry if a recoverably reset event has been scheduled */ - if (efx->reset_pending & - ~(1 << RESET_TYPE_INVISIBLE | 1 << RESET_TYPE_ALL) || - !efx->reset_pending) - goto fail3; - - efx->reset_pending = 0; - } + rc = efx_pci_probe_main(efx); + if (rc) + goto fail3; - if (rc) { - netif_err(efx, probe, efx->net_dev, "Could not reset NIC\n"); + rc = efx_register_netdev(efx); + if (rc) goto fail4; - } - /* Switch to the running state before we expose the device to the OS, - * so that dev_open()|efx_start_all() will actually start the device */ - efx->state = STATE_RUNNING; - - rc = efx_register_netdev(efx); + rc = efx_sriov_init(efx); if (rc) - goto fail5; + netif_err(efx, probe, efx->net_dev, + "SR-IOV can't be enabled rc %d\n", rc); netif_dbg(efx, probe, efx->net_dev, "initialisation successful\n"); + /* Try to create MTDs, but allow this to fail */ rtnl_lock(); - efx_mtd_probe(efx); /* allowed to fail */ + rc = efx_mtd_probe(efx); rtnl_unlock(); + if (rc) + netif_warn(efx, probe, efx->net_dev, + "failed to create MTDs (%d)\n", rc); + + rc = pci_enable_pcie_error_reporting(pci_dev); + if (rc && rc != -EINVAL) + netif_warn(efx, probe, efx->net_dev, + "pci_enable_pcie_error_reporting failed (%d)\n", rc); + return 0; - fail5: - efx_pci_remove_main(efx); fail4: + efx_pci_remove_main(efx); fail3: efx_fini_io(efx); fail2: @@ -2573,40 +3001,58 @@ static int efx_pm_freeze(struct device *dev) { struct efx_nic *efx = pci_get_drvdata(to_pci_dev(dev)); - efx->state = STATE_FINI; + rtnl_lock(); + + if (efx->state != STATE_DISABLED) { + efx->state = STATE_UNINIT; - netif_device_detach(efx->net_dev); + efx_device_detach_sync(efx); - efx_stop_all(efx); - efx_fini_channels(efx); + efx_stop_all(efx); + efx_disable_interrupts(efx); + } + + rtnl_unlock(); return 0; } static int efx_pm_thaw(struct device *dev) { + int rc; struct efx_nic *efx = pci_get_drvdata(to_pci_dev(dev)); - efx->state = STATE_INIT; + rtnl_lock(); + + if (efx->state != STATE_DISABLED) { + rc = efx_enable_interrupts(efx); + if (rc) + goto fail; - efx_init_channels(efx); + mutex_lock(&efx->mac_lock); + efx->phy_op->reconfigure(efx); + mutex_unlock(&efx->mac_lock); - mutex_lock(&efx->mac_lock); - efx->phy_op->reconfigure(efx); - mutex_unlock(&efx->mac_lock); + efx_start_all(efx); - efx_start_all(efx); + netif_device_attach(efx->net_dev); - netif_device_attach(efx->net_dev); + efx->state = STATE_READY; - efx->state = STATE_RUNNING; + efx->type->resume_wol(efx); + } - efx->type->resume_wol(efx); + rtnl_unlock(); /* Reschedule any quenched resets scheduled during efx_pm_freeze() */ queue_work(reset_workqueue, &efx->reset_work); return 0; + +fail: + rtnl_unlock(); + + return rc; } static int efx_pm_poweroff(struct device *dev) @@ -2643,8 +3089,8 @@ static int efx_pm_resume(struct device *dev) rc = efx->type->init(efx); if (rc) return rc; - efx_pm_thaw(dev); - return 0; + rc = efx_pm_thaw(dev); + return rc; } static int efx_pm_suspend(struct device *dev) @@ -2658,7 +3104,7 @@ static int efx_pm_suspend(struct device *dev) return rc; } -static struct dev_pm_ops efx_pm_ops = { +static const struct dev_pm_ops efx_pm_ops = { .suspend = efx_pm_suspend, .resume = efx_pm_resume, .freeze = efx_pm_freeze, @@ -2667,12 +3113,112 @@ static struct dev_pm_ops efx_pm_ops = { .restore = efx_pm_resume, }; +/* A PCI error affecting this device was detected. + * At this point MMIO and DMA may be disabled. + * Stop the software path and request a slot reset. + */ +static pci_ers_result_t efx_io_error_detected(struct pci_dev *pdev, + enum pci_channel_state state) +{ + pci_ers_result_t status = PCI_ERS_RESULT_RECOVERED; + struct efx_nic *efx = pci_get_drvdata(pdev); + + if (state == pci_channel_io_perm_failure) + return PCI_ERS_RESULT_DISCONNECT; + + rtnl_lock(); + + if (efx->state != STATE_DISABLED) { + efx->state = STATE_RECOVERY; + efx->reset_pending = 0; + + efx_device_detach_sync(efx); + + efx_stop_all(efx); + efx_disable_interrupts(efx); + + status = PCI_ERS_RESULT_NEED_RESET; + } else { + /* If the interface is disabled we don't want to do anything + * with it. + */ + status = PCI_ERS_RESULT_RECOVERED; + } + + rtnl_unlock(); + + pci_disable_device(pdev); + + return status; +} + +/* Fake a successfull reset, which will be performed later in efx_io_resume. */ +static pci_ers_result_t efx_io_slot_reset(struct pci_dev *pdev) +{ + struct efx_nic *efx = pci_get_drvdata(pdev); + pci_ers_result_t status = PCI_ERS_RESULT_RECOVERED; + int rc; + + if (pci_enable_device(pdev)) { + netif_err(efx, hw, efx->net_dev, + "Cannot re-enable PCI device after reset.\n"); + status = PCI_ERS_RESULT_DISCONNECT; + } + + rc = pci_cleanup_aer_uncorrect_error_status(pdev); + if (rc) { + netif_err(efx, hw, efx->net_dev, + "pci_cleanup_aer_uncorrect_error_status failed (%d)\n", rc); + /* Non-fatal error. Continue. */ + } + + return status; +} + +/* Perform the actual reset and resume I/O operations. */ +static void efx_io_resume(struct pci_dev *pdev) +{ + struct efx_nic *efx = pci_get_drvdata(pdev); + int rc; + + rtnl_lock(); + + if (efx->state == STATE_DISABLED) + goto out; + + rc = efx_reset(efx, RESET_TYPE_ALL); + if (rc) { + netif_err(efx, hw, efx->net_dev, + "efx_reset failed after PCI error (%d)\n", rc); + } else { + efx->state = STATE_READY; + netif_dbg(efx, hw, efx->net_dev, + "Done resetting and resuming IO after PCI error.\n"); + } + +out: + rtnl_unlock(); +} + +/* For simplicity and reliability, we always require a slot reset and try to + * reset the hardware when a pci error affecting the device is detected. + * We leave both the link_reset and mmio_enabled callback unimplemented: + * with our request for slot reset the mmio_enabled callback will never be + * called, and the link_reset callback is not used by AER or EEH mechanisms. + */ +static struct pci_error_handlers efx_err_handlers = { + .error_detected = efx_io_error_detected, + .slot_reset = efx_io_slot_reset, + .resume = efx_io_resume, +}; + static struct pci_driver efx_pci_driver = { .name = KBUILD_MODNAME, .id_table = efx_pci_table, .probe = efx_pci_probe, .remove = efx_pci_remove, .driver.pm = &efx_pm_ops, + .err_handler = &efx_err_handlers, }; /************************************************************************** @@ -2695,6 +3241,10 @@ static int __init efx_init_module(void) if (rc) goto err_notifier; + rc = efx_init_sriov(); + if (rc) + goto err_sriov; + reset_workqueue = create_singlethread_workqueue("sfc_reset"); if (!reset_workqueue) { rc = -ENOMEM; @@ -2710,6 +3260,8 @@ static int __init efx_init_module(void) err_pci: destroy_workqueue(reset_workqueue); err_reset: + efx_fini_sriov(); + err_sriov: unregister_netdevice_notifier(&efx_netdev_notifier); err_notifier: return rc; @@ -2721,6 +3273,7 @@ static void __exit efx_exit_module(void) pci_unregister_driver(&efx_pci_driver); destroy_workqueue(reset_workqueue); + efx_fini_sriov(); unregister_netdevice_notifier(&efx_netdev_notifier); } @@ -2730,6 +3283,6 @@ module_exit(efx_exit_module); MODULE_AUTHOR("Solarflare Communications and " "Michael Brown <mbrown@fensystems.co.uk>"); -MODULE_DESCRIPTION("Solarflare Communications network driver"); +MODULE_DESCRIPTION("Solarflare network driver"); MODULE_LICENSE("GPL"); MODULE_DEVICE_TABLE(pci, efx_pci_table); |