diff options
Diffstat (limited to 'drivers/net')
244 files changed, 14685 insertions, 8383 deletions
diff --git a/drivers/net/Kconfig b/drivers/net/Kconfig index 01e2657e4c26..dee79588d2b1 100644 --- a/drivers/net/Kconfig +++ b/drivers/net/Kconfig @@ -549,6 +549,8 @@ source "drivers/net/hyperv/Kconfig" config NETDEVSIM tristate "Simulated networking device" depends on DEBUG_FS + depends on INET + depends on IPV6 || IPV6=n select NET_DEVLINK help This driver is a developer testing tool and software model that can diff --git a/drivers/net/can/m_can/tcan4x5x.c b/drivers/net/can/m_can/tcan4x5x.c index 4e1789ea2bc3..eacd428e07e9 100644 --- a/drivers/net/can/m_can/tcan4x5x.c +++ b/drivers/net/can/m_can/tcan4x5x.c @@ -102,6 +102,7 @@ #define TCAN4X5X_MODE_NORMAL BIT(7) #define TCAN4X5X_DISABLE_WAKE_MSK (BIT(31) | BIT(30)) +#define TCAN4X5X_DISABLE_INH_MSK BIT(9) #define TCAN4X5X_SW_RESET BIT(2) @@ -166,6 +167,28 @@ static void tcan4x5x_check_wake(struct tcan4x5x_priv *priv) } } +static int tcan4x5x_reset(struct tcan4x5x_priv *priv) +{ + int ret = 0; + + if (priv->reset_gpio) { + gpiod_set_value(priv->reset_gpio, 1); + + /* tpulse_width minimum 30us */ + usleep_range(30, 100); + gpiod_set_value(priv->reset_gpio, 0); + } else { + ret = regmap_write(priv->regmap, TCAN4X5X_CONFIG, + TCAN4X5X_SW_RESET); + if (ret) + return ret; + } + + usleep_range(700, 1000); + + return ret; +} + static int regmap_spi_gather_write(void *context, const void *reg, size_t reg_len, const void *val, size_t val_len) @@ -348,14 +371,23 @@ static int tcan4x5x_disable_wake(struct m_can_classdev *cdev) TCAN4X5X_DISABLE_WAKE_MSK, 0x00); } +static int tcan4x5x_disable_state(struct m_can_classdev *cdev) +{ + struct tcan4x5x_priv *tcan4x5x = cdev->device_data; + + return regmap_update_bits(tcan4x5x->regmap, TCAN4X5X_CONFIG, + TCAN4X5X_DISABLE_INH_MSK, 0x01); +} + static int tcan4x5x_parse_config(struct m_can_classdev *cdev) { struct tcan4x5x_priv *tcan4x5x = cdev->device_data; + int ret; tcan4x5x->device_wake_gpio = devm_gpiod_get(cdev->dev, "device-wake", GPIOD_OUT_HIGH); if (IS_ERR(tcan4x5x->device_wake_gpio)) { - if (PTR_ERR(tcan4x5x->power) == -EPROBE_DEFER) + if (PTR_ERR(tcan4x5x->device_wake_gpio) == -EPROBE_DEFER) return -EPROBE_DEFER; tcan4x5x_disable_wake(cdev); @@ -366,18 +398,17 @@ static int tcan4x5x_parse_config(struct m_can_classdev *cdev) if (IS_ERR(tcan4x5x->reset_gpio)) tcan4x5x->reset_gpio = NULL; - usleep_range(700, 1000); + ret = tcan4x5x_reset(tcan4x5x); + if (ret) + return ret; tcan4x5x->device_state_gpio = devm_gpiod_get_optional(cdev->dev, "device-state", GPIOD_IN); - if (IS_ERR(tcan4x5x->device_state_gpio)) + if (IS_ERR(tcan4x5x->device_state_gpio)) { tcan4x5x->device_state_gpio = NULL; - - tcan4x5x->power = devm_regulator_get_optional(cdev->dev, - "vsup"); - if (PTR_ERR(tcan4x5x->power) == -EPROBE_DEFER) - return -EPROBE_DEFER; + tcan4x5x_disable_state(cdev); + } return 0; } @@ -412,6 +443,12 @@ static int tcan4x5x_can_probe(struct spi_device *spi) if (!priv) return -ENOMEM; + priv->power = devm_regulator_get_optional(&spi->dev, "vsup"); + if (PTR_ERR(priv->power) == -EPROBE_DEFER) + return -EPROBE_DEFER; + else + priv->power = NULL; + mcan_class->device_data = priv; m_can_class_get_clocks(mcan_class); @@ -451,11 +488,17 @@ static int tcan4x5x_can_probe(struct spi_device *spi) priv->regmap = devm_regmap_init(&spi->dev, &tcan4x5x_bus, &spi->dev, &tcan4x5x_regmap); - ret = tcan4x5x_parse_config(mcan_class); + ret = tcan4x5x_power_enable(priv->power, 1); if (ret) goto out_clk; - tcan4x5x_power_enable(priv->power, 1); + ret = tcan4x5x_parse_config(mcan_class); + if (ret) + goto out_power; + + ret = tcan4x5x_init(mcan_class); + if (ret) + goto out_power; ret = m_can_class_register(mcan_class); if (ret) diff --git a/drivers/net/can/mscan/mscan.c b/drivers/net/can/mscan/mscan.c index 8caf7af0dee2..99101d7027a8 100644 --- a/drivers/net/can/mscan/mscan.c +++ b/drivers/net/can/mscan/mscan.c @@ -381,13 +381,12 @@ static int mscan_rx_poll(struct napi_struct *napi, int quota) struct net_device *dev = napi->dev; struct mscan_regs __iomem *regs = priv->reg_base; struct net_device_stats *stats = &dev->stats; - int npackets = 0; - int ret = 1; + int work_done = 0; struct sk_buff *skb; struct can_frame *frame; u8 canrflg; - while (npackets < quota) { + while (work_done < quota) { canrflg = in_8(®s->canrflg); if (!(canrflg & (MSCAN_RXF | MSCAN_ERR_IF))) break; @@ -408,18 +407,18 @@ static int mscan_rx_poll(struct napi_struct *napi, int quota) stats->rx_packets++; stats->rx_bytes += frame->can_dlc; - npackets++; + work_done++; netif_receive_skb(skb); } - if (!(in_8(®s->canrflg) & (MSCAN_RXF | MSCAN_ERR_IF))) { - napi_complete(&priv->napi); - clear_bit(F_RX_PROGRESS, &priv->flags); - if (priv->can.state < CAN_STATE_BUS_OFF) - out_8(®s->canrier, priv->shadow_canrier); - ret = 0; + if (work_done < quota) { + if (likely(napi_complete_done(&priv->napi, work_done))) { + clear_bit(F_RX_PROGRESS, &priv->flags); + if (priv->can.state < CAN_STATE_BUS_OFF) + out_8(®s->canrier, priv->shadow_canrier); + } } - return ret; + return work_done; } static irqreturn_t mscan_isr(int irq, void *dev_id) diff --git a/drivers/net/can/usb/gs_usb.c b/drivers/net/can/usb/gs_usb.c index 2f74f6704c12..a4b4b742c80c 100644 --- a/drivers/net/can/usb/gs_usb.c +++ b/drivers/net/can/usb/gs_usb.c @@ -918,7 +918,7 @@ static int gs_usb_probe(struct usb_interface *intf, GS_USB_BREQ_HOST_FORMAT, USB_DIR_OUT | USB_TYPE_VENDOR | USB_RECIP_INTERFACE, 1, - intf->altsetting[0].desc.bInterfaceNumber, + intf->cur_altsetting->desc.bInterfaceNumber, hconf, sizeof(*hconf), 1000); @@ -941,7 +941,7 @@ static int gs_usb_probe(struct usb_interface *intf, GS_USB_BREQ_DEVICE_CONFIG, USB_DIR_IN | USB_TYPE_VENDOR | USB_RECIP_INTERFACE, 1, - intf->altsetting[0].desc.bInterfaceNumber, + intf->cur_altsetting->desc.bInterfaceNumber, dconf, sizeof(*dconf), 1000); diff --git a/drivers/net/can/usb/kvaser_usb/kvaser_usb_hydra.c b/drivers/net/can/usb/kvaser_usb/kvaser_usb_hydra.c index 5fc0be564274..7ab87a758754 100644 --- a/drivers/net/can/usb/kvaser_usb/kvaser_usb_hydra.c +++ b/drivers/net/can/usb/kvaser_usb/kvaser_usb_hydra.c @@ -1590,7 +1590,7 @@ static int kvaser_usb_hydra_setup_endpoints(struct kvaser_usb *dev) struct usb_endpoint_descriptor *ep; int i; - iface_desc = &dev->intf->altsetting[0]; + iface_desc = dev->intf->cur_altsetting; for (i = 0; i < iface_desc->desc.bNumEndpoints; ++i) { ep = &iface_desc->endpoint[i].desc; diff --git a/drivers/net/can/usb/kvaser_usb/kvaser_usb_leaf.c b/drivers/net/can/usb/kvaser_usb/kvaser_usb_leaf.c index ae4c37e1bb75..1b9957f12459 100644 --- a/drivers/net/can/usb/kvaser_usb/kvaser_usb_leaf.c +++ b/drivers/net/can/usb/kvaser_usb/kvaser_usb_leaf.c @@ -1310,7 +1310,7 @@ static int kvaser_usb_leaf_setup_endpoints(struct kvaser_usb *dev) struct usb_endpoint_descriptor *endpoint; int i; - iface_desc = &dev->intf->altsetting[0]; + iface_desc = dev->intf->cur_altsetting; for (i = 0; i < iface_desc->desc.bNumEndpoints; ++i) { endpoint = &iface_desc->endpoint[i].desc; diff --git a/drivers/net/dsa/b53/b53_common.c b/drivers/net/dsa/b53/b53_common.c index edacacfc9365..060497512159 100644 --- a/drivers/net/dsa/b53/b53_common.c +++ b/drivers/net/dsa/b53/b53_common.c @@ -371,8 +371,6 @@ static void b53_enable_vlan(struct b53_device *dev, bool enable, b53_read8(dev, B53_VLAN_PAGE, B53_VLAN_CTRL5, &vc5); } - mgmt &= ~SM_SW_FWD_MODE; - if (enable) { vc0 |= VC0_VLAN_EN | VC0_VID_CHK_EN | VC0_VID_HASH_VID; vc1 |= VC1_RX_MCST_UNTAG_EN | VC1_RX_MCST_FWD_EN; @@ -573,9 +571,8 @@ EXPORT_SYMBOL(b53_disable_port); void b53_brcm_hdr_setup(struct dsa_switch *ds, int port) { - bool tag_en = !(ds->ops->get_tag_protocol(ds, port) == - DSA_TAG_PROTO_NONE); struct b53_device *dev = ds->priv; + bool tag_en = !(dev->tag_protocol == DSA_TAG_PROTO_NONE); u8 hdr_ctl, val; u16 reg; @@ -595,6 +592,22 @@ void b53_brcm_hdr_setup(struct dsa_switch *ds, int port) break; } + /* Enable management mode if tagging is requested */ + b53_read8(dev, B53_CTRL_PAGE, B53_SWITCH_MODE, &hdr_ctl); + if (tag_en) + hdr_ctl |= SM_SW_FWD_MODE; + else + hdr_ctl &= ~SM_SW_FWD_MODE; + b53_write8(dev, B53_CTRL_PAGE, B53_SWITCH_MODE, hdr_ctl); + + /* Configure the appropriate IMP port */ + b53_read8(dev, B53_MGMT_PAGE, B53_GLOBAL_CONFIG, &hdr_ctl); + if (port == 8) + hdr_ctl |= GC_FRM_MGMT_PORT_MII; + else if (port == 5) + hdr_ctl |= GC_FRM_MGMT_PORT_M; + b53_write8(dev, B53_MGMT_PAGE, B53_GLOBAL_CONFIG, hdr_ctl); + /* Enable Broadcom tags for IMP port */ b53_read8(dev, B53_MGMT_PAGE, B53_BRCM_HDR, &hdr_ctl); if (tag_en) @@ -1866,36 +1879,57 @@ static bool b53_possible_cpu_port(struct dsa_switch *ds, int port) return false; } -static bool b53_can_enable_brcm_tags(struct dsa_switch *ds, int port) +static bool b53_can_enable_brcm_tags(struct dsa_switch *ds, int port, + enum dsa_tag_protocol tag_protocol) { bool ret = b53_possible_cpu_port(ds, port); - if (!ret) + if (!ret) { dev_warn(ds->dev, "Port %d is not Broadcom tag capable\n", port); + return ret; + } + + switch (tag_protocol) { + case DSA_TAG_PROTO_BRCM: + case DSA_TAG_PROTO_BRCM_PREPEND: + dev_warn(ds->dev, + "Port %d is stacked to Broadcom tag switch\n", port); + ret = false; + break; + default: + ret = true; + break; + } + return ret; } -enum dsa_tag_protocol b53_get_tag_protocol(struct dsa_switch *ds, int port) +enum dsa_tag_protocol b53_get_tag_protocol(struct dsa_switch *ds, int port, + enum dsa_tag_protocol mprot) { struct b53_device *dev = ds->priv; /* Older models (5325, 5365) support a different tag format that we do - * not support in net/dsa/tag_brcm.c yet. 539x and 531x5 require managed - * mode to be turned on which means we need to specifically manage ARL - * misses on multicast addresses (TBD). + * not support in net/dsa/tag_brcm.c yet. */ - if (is5325(dev) || is5365(dev) || is539x(dev) || is531x5(dev) || - !b53_can_enable_brcm_tags(ds, port)) - return DSA_TAG_PROTO_NONE; + if (is5325(dev) || is5365(dev) || + !b53_can_enable_brcm_tags(ds, port, mprot)) { + dev->tag_protocol = DSA_TAG_PROTO_NONE; + goto out; + } /* Broadcom BCM58xx chips have a flow accelerator on Port 8 * which requires us to use the prepended Broadcom tag type */ - if (dev->chip_id == BCM58XX_DEVICE_ID && port == B53_CPU_PORT) - return DSA_TAG_PROTO_BRCM_PREPEND; + if (dev->chip_id == BCM58XX_DEVICE_ID && port == B53_CPU_PORT) { + dev->tag_protocol = DSA_TAG_PROTO_BRCM_PREPEND; + goto out; + } - return DSA_TAG_PROTO_BRCM; + dev->tag_protocol = DSA_TAG_PROTO_BRCM; +out: + return dev->tag_protocol; } EXPORT_SYMBOL(b53_get_tag_protocol); diff --git a/drivers/net/dsa/b53/b53_priv.h b/drivers/net/dsa/b53/b53_priv.h index 1877acf05081..3c30f3a7eb29 100644 --- a/drivers/net/dsa/b53/b53_priv.h +++ b/drivers/net/dsa/b53/b53_priv.h @@ -118,6 +118,7 @@ struct b53_device { u8 jumbo_size_reg; int reset_gpio; u8 num_arl_entries; + enum dsa_tag_protocol tag_protocol; /* used ports mask */ u16 enabled_ports; @@ -359,7 +360,8 @@ int b53_mdb_del(struct dsa_switch *ds, int port, const struct switchdev_obj_port_mdb *mdb); int b53_mirror_add(struct dsa_switch *ds, int port, struct dsa_mall_mirror_tc_entry *mirror, bool ingress); -enum dsa_tag_protocol b53_get_tag_protocol(struct dsa_switch *ds, int port); +enum dsa_tag_protocol b53_get_tag_protocol(struct dsa_switch *ds, int port, + enum dsa_tag_protocol mprot); void b53_mirror_del(struct dsa_switch *ds, int port, struct dsa_mall_mirror_tc_entry *mirror); int b53_enable_port(struct dsa_switch *ds, int port, struct phy_device *phy); diff --git a/drivers/net/dsa/bcm_sf2.c b/drivers/net/dsa/bcm_sf2.c index e43040c9f9ee..3e8635311d0d 100644 --- a/drivers/net/dsa/bcm_sf2.c +++ b/drivers/net/dsa/bcm_sf2.c @@ -68,7 +68,7 @@ static void bcm_sf2_imp_setup(struct dsa_switch *ds, int port) /* Force link status for IMP port */ reg = core_readl(priv, offset); - reg |= (MII_SW_OR | LINK_STS); + reg |= (MII_SW_OR | LINK_STS | GMII_SPEED_UP_2G); core_writel(priv, reg, offset); /* Enable Broadcast, Multicast, Unicast forwarding to IMP port */ diff --git a/drivers/net/dsa/dsa_loop.c b/drivers/net/dsa/dsa_loop.c index c8d7ef27fd72..fdcb70b9f0e4 100644 --- a/drivers/net/dsa/dsa_loop.c +++ b/drivers/net/dsa/dsa_loop.c @@ -61,7 +61,8 @@ struct dsa_loop_priv { static struct phy_device *phydevs[PHY_MAX_ADDR]; static enum dsa_tag_protocol dsa_loop_get_protocol(struct dsa_switch *ds, - int port) + int port, + enum dsa_tag_protocol mp) { dev_dbg(ds->dev, "%s: port: %d\n", __func__, port); diff --git a/drivers/net/dsa/lan9303-core.c b/drivers/net/dsa/lan9303-core.c index e3c333a8f45d..cc17a44dd3a8 100644 --- a/drivers/net/dsa/lan9303-core.c +++ b/drivers/net/dsa/lan9303-core.c @@ -883,7 +883,8 @@ static int lan9303_check_device(struct lan9303 *chip) /* ---------------------------- DSA -----------------------------------*/ static enum dsa_tag_protocol lan9303_get_tag_protocol(struct dsa_switch *ds, - int port) + int port, + enum dsa_tag_protocol mp) { return DSA_TAG_PROTO_LAN9303; } diff --git a/drivers/net/dsa/lantiq_gswip.c b/drivers/net/dsa/lantiq_gswip.c index 955324968b74..0369c22fe3e1 100644 --- a/drivers/net/dsa/lantiq_gswip.c +++ b/drivers/net/dsa/lantiq_gswip.c @@ -841,7 +841,8 @@ static int gswip_setup(struct dsa_switch *ds) } static enum dsa_tag_protocol gswip_get_tag_protocol(struct dsa_switch *ds, - int port) + int port, + enum dsa_tag_protocol mp) { return DSA_TAG_PROTO_GSWIP; } diff --git a/drivers/net/dsa/microchip/ksz8795.c b/drivers/net/dsa/microchip/ksz8795.c index 24a5e99f7fd5..47d65b77caf7 100644 --- a/drivers/net/dsa/microchip/ksz8795.c +++ b/drivers/net/dsa/microchip/ksz8795.c @@ -645,7 +645,8 @@ static void ksz8795_w_phy(struct ksz_device *dev, u16 phy, u16 reg, u16 val) } static enum dsa_tag_protocol ksz8795_get_tag_protocol(struct dsa_switch *ds, - int port) + int port, + enum dsa_tag_protocol mp) { return DSA_TAG_PROTO_KSZ8795; } diff --git a/drivers/net/dsa/microchip/ksz9477.c b/drivers/net/dsa/microchip/ksz9477.c index 50ffc63d6231..9a51b8a4de5d 100644 --- a/drivers/net/dsa/microchip/ksz9477.c +++ b/drivers/net/dsa/microchip/ksz9477.c @@ -295,7 +295,8 @@ static void ksz9477_port_init_cnt(struct ksz_device *dev, int port) } static enum dsa_tag_protocol ksz9477_get_tag_protocol(struct dsa_switch *ds, - int port) + int port, + enum dsa_tag_protocol mp) { enum dsa_tag_protocol proto = DSA_TAG_PROTO_KSZ9477; struct ksz_device *dev = ds->priv; diff --git a/drivers/net/dsa/mt7530.c b/drivers/net/dsa/mt7530.c index ed1ec10ec62b..022466ca1c19 100644 --- a/drivers/net/dsa/mt7530.c +++ b/drivers/net/dsa/mt7530.c @@ -1223,7 +1223,8 @@ mt7530_port_vlan_del(struct dsa_switch *ds, int port, } static enum dsa_tag_protocol -mtk_get_tag_protocol(struct dsa_switch *ds, int port) +mtk_get_tag_protocol(struct dsa_switch *ds, int port, + enum dsa_tag_protocol mp) { struct mt7530_priv *priv = ds->priv; diff --git a/drivers/net/dsa/mv88e6060.c b/drivers/net/dsa/mv88e6060.c index a5a37f47b320..24b8219fd607 100644 --- a/drivers/net/dsa/mv88e6060.c +++ b/drivers/net/dsa/mv88e6060.c @@ -43,7 +43,8 @@ static const char *mv88e6060_get_name(struct mii_bus *bus, int sw_addr) } static enum dsa_tag_protocol mv88e6060_get_tag_protocol(struct dsa_switch *ds, - int port) + int port, + enum dsa_tag_protocol m) { return DSA_TAG_PROTO_TRAILER; } diff --git a/drivers/net/dsa/mv88e6xxx/chip.c b/drivers/net/dsa/mv88e6xxx/chip.c index 99816ca9e5e4..8c9289549688 100644 --- a/drivers/net/dsa/mv88e6xxx/chip.c +++ b/drivers/net/dsa/mv88e6xxx/chip.c @@ -3845,6 +3845,9 @@ static const struct mv88e6xxx_ops mv88e6190_ops = { .serdes_irq_mapping = mv88e6390_serdes_irq_mapping, .serdes_irq_enable = mv88e6390_serdes_irq_enable, .serdes_irq_status = mv88e6390_serdes_irq_status, + .serdes_get_strings = mv88e6390_serdes_get_strings, + .serdes_get_stats = mv88e6390_serdes_get_stats, + .phylink_validate = mv88e6390_phylink_validate, .gpio_ops = &mv88e6352_gpio_ops, .phylink_validate = mv88e6390_phylink_validate, }; @@ -3896,6 +3899,9 @@ static const struct mv88e6xxx_ops mv88e6190x_ops = { .serdes_irq_mapping = mv88e6390_serdes_irq_mapping, .serdes_irq_enable = mv88e6390_serdes_irq_enable, .serdes_irq_status = mv88e6390_serdes_irq_status, + .serdes_get_strings = mv88e6390_serdes_get_strings, + .serdes_get_stats = mv88e6390_serdes_get_stats, + .phylink_validate = mv88e6390_phylink_validate, .gpio_ops = &mv88e6352_gpio_ops, .phylink_validate = mv88e6390x_phylink_validate, }; @@ -3946,6 +3952,9 @@ static const struct mv88e6xxx_ops mv88e6191_ops = { .serdes_irq_mapping = mv88e6390_serdes_irq_mapping, .serdes_irq_enable = mv88e6390_serdes_irq_enable, .serdes_irq_status = mv88e6390_serdes_irq_status, + .serdes_get_strings = mv88e6390_serdes_get_strings, + .serdes_get_stats = mv88e6390_serdes_get_stats, + .phylink_validate = mv88e6390_phylink_validate, .avb_ops = &mv88e6390_avb_ops, .ptp_ops = &mv88e6352_ptp_ops, .phylink_validate = mv88e6390_phylink_validate, @@ -4092,6 +4101,9 @@ static const struct mv88e6xxx_ops mv88e6290_ops = { .serdes_irq_mapping = mv88e6390_serdes_irq_mapping, .serdes_irq_enable = mv88e6390_serdes_irq_enable, .serdes_irq_status = mv88e6390_serdes_irq_status, + .serdes_get_strings = mv88e6390_serdes_get_strings, + .serdes_get_stats = mv88e6390_serdes_get_stats, + .phylink_validate = mv88e6390_phylink_validate, .gpio_ops = &mv88e6352_gpio_ops, .avb_ops = &mv88e6390_avb_ops, .ptp_ops = &mv88e6352_ptp_ops, @@ -4486,6 +4498,9 @@ static const struct mv88e6xxx_ops mv88e6390x_ops = { .serdes_irq_mapping = mv88e6390_serdes_irq_mapping, .serdes_irq_enable = mv88e6390_serdes_irq_enable, .serdes_irq_status = mv88e6390_serdes_irq_status, + .serdes_get_sset_count = mv88e6390_serdes_get_sset_count, + .serdes_get_strings = mv88e6390_serdes_get_strings, + .serdes_get_stats = mv88e6390_serdes_get_stats, .gpio_ops = &mv88e6352_gpio_ops, .avb_ops = &mv88e6390_avb_ops, .ptp_ops = &mv88e6352_ptp_ops, @@ -5217,7 +5232,8 @@ static struct mv88e6xxx_chip *mv88e6xxx_alloc_chip(struct device *dev) } static enum dsa_tag_protocol mv88e6xxx_get_tag_protocol(struct dsa_switch *ds, - int port) + int port, + enum dsa_tag_protocol m) { struct mv88e6xxx_chip *chip = ds->priv; diff --git a/drivers/net/dsa/mv88e6xxx/global1.c b/drivers/net/dsa/mv88e6xxx/global1.c index 120a65d3e3ef..b016cc205f81 100644 --- a/drivers/net/dsa/mv88e6xxx/global1.c +++ b/drivers/net/dsa/mv88e6xxx/global1.c @@ -360,6 +360,11 @@ int mv88e6390_g1_set_cpu_port(struct mv88e6xxx_chip *chip, int port) { u16 ptr = MV88E6390_G1_MONITOR_MGMT_CTL_PTR_CPU_DEST; + /* Use the default high priority for management frames sent to + * the CPU. + */ + port |= MV88E6390_G1_MONITOR_MGMT_CTL_PTR_CPU_DEST_MGMTPRI; + return mv88e6390_g1_monitor_write(chip, ptr, port); } diff --git a/drivers/net/dsa/mv88e6xxx/global1.h b/drivers/net/dsa/mv88e6xxx/global1.h index bc5a6b2bb1e4..5324c6f4ae90 100644 --- a/drivers/net/dsa/mv88e6xxx/global1.h +++ b/drivers/net/dsa/mv88e6xxx/global1.h @@ -211,6 +211,7 @@ #define MV88E6390_G1_MONITOR_MGMT_CTL_PTR_INGRESS_DEST 0x2000 #define MV88E6390_G1_MONITOR_MGMT_CTL_PTR_EGRESS_DEST 0x2100 #define MV88E6390_G1_MONITOR_MGMT_CTL_PTR_CPU_DEST 0x3000 +#define MV88E6390_G1_MONITOR_MGMT_CTL_PTR_CPU_DEST_MGMTPRI 0x00e0 #define MV88E6390_G1_MONITOR_MGMT_CTL_DATA_MASK 0x00ff /* Offset 0x1C: Global Control 2 */ diff --git a/drivers/net/dsa/mv88e6xxx/port.c b/drivers/net/dsa/mv88e6xxx/port.c index 7fe256c5739d..0b43c650e100 100644 --- a/drivers/net/dsa/mv88e6xxx/port.c +++ b/drivers/net/dsa/mv88e6xxx/port.c @@ -393,7 +393,7 @@ phy_interface_t mv88e6390x_port_max_speed_mode(int port) } static int mv88e6xxx_port_set_cmode(struct mv88e6xxx_chip *chip, int port, - phy_interface_t mode) + phy_interface_t mode, bool force) { u8 lane; u16 cmode; @@ -427,8 +427,8 @@ static int mv88e6xxx_port_set_cmode(struct mv88e6xxx_chip *chip, int port, cmode = 0; } - /* cmode doesn't change, nothing to do for us */ - if (cmode == chip->ports[port].cmode) + /* cmode doesn't change, nothing to do for us unless forced */ + if (cmode == chip->ports[port].cmode && !force) return 0; lane = mv88e6xxx_serdes_get_lane(chip, port); @@ -484,7 +484,7 @@ int mv88e6390x_port_set_cmode(struct mv88e6xxx_chip *chip, int port, if (port != 9 && port != 10) return -EOPNOTSUPP; - return mv88e6xxx_port_set_cmode(chip, port, mode); + return mv88e6xxx_port_set_cmode(chip, port, mode, false); } int mv88e6390_port_set_cmode(struct mv88e6xxx_chip *chip, int port, @@ -504,7 +504,7 @@ int mv88e6390_port_set_cmode(struct mv88e6xxx_chip *chip, int port, break; } - return mv88e6xxx_port_set_cmode(chip, port, mode); + return mv88e6xxx_port_set_cmode(chip, port, mode, false); } static int mv88e6341_port_set_cmode_writable(struct mv88e6xxx_chip *chip, @@ -555,7 +555,7 @@ int mv88e6341_port_set_cmode(struct mv88e6xxx_chip *chip, int port, if (err) return err; - return mv88e6xxx_port_set_cmode(chip, port, mode); + return mv88e6xxx_port_set_cmode(chip, port, mode, true); } int mv88e6185_port_get_cmode(struct mv88e6xxx_chip *chip, int port, u8 *cmode) diff --git a/drivers/net/dsa/ocelot/felix.c b/drivers/net/dsa/ocelot/felix.c index f072dd75cea2..3257962c147e 100644 --- a/drivers/net/dsa/ocelot/felix.c +++ b/drivers/net/dsa/ocelot/felix.c @@ -16,7 +16,8 @@ #include "felix.h" static enum dsa_tag_protocol felix_get_tag_protocol(struct dsa_switch *ds, - int port) + int port, + enum dsa_tag_protocol mp) { return DSA_TAG_PROTO_OCELOT; } @@ -171,11 +172,10 @@ static void felix_phylink_validate(struct dsa_switch *ds, int port, phylink_set(mask, Autoneg); phylink_set(mask, Pause); phylink_set(mask, Asym_Pause); - if (state->interface != PHY_INTERFACE_MODE_2500BASEX) { - phylink_set(mask, 10baseT_Full); - phylink_set(mask, 100baseT_Full); - phylink_set(mask, 1000baseT_Full); - } + phylink_set(mask, 10baseT_Full); + phylink_set(mask, 100baseT_Full); + phylink_set(mask, 1000baseT_Full); + /* The internal ports that run at 2.5G are overclocked GMII */ if (state->interface == PHY_INTERFACE_MODE_GMII || state->interface == PHY_INTERFACE_MODE_2500BASEX || @@ -221,8 +221,12 @@ static void felix_phylink_mac_config(struct dsa_switch *ds, int port, * specification in incoming pause frames. */ mac_fc_cfg = SYS_MAC_FC_CFG_FC_LINK_SPEED(state->speed); - if (state->pause & MLO_PAUSE_RX) - mac_fc_cfg |= SYS_MAC_FC_CFG_RX_FC_ENA; + + /* handle Rx pause in all cases, with 2500base-X this is used for rate + * adaptation. + */ + mac_fc_cfg |= SYS_MAC_FC_CFG_RX_FC_ENA; + if (state->pause & MLO_PAUSE_TX) mac_fc_cfg |= SYS_MAC_FC_CFG_TX_FC_ENA | SYS_MAC_FC_CFG_PAUSE_VAL_CFG(0xffff) | @@ -322,7 +326,7 @@ static int felix_parse_ports_node(struct felix *felix, struct device *dev = felix->ocelot.dev; struct device_node *child; - for_each_child_of_node(ports_node, child) { + for_each_available_child_of_node(ports_node, child) { phy_interface_t phy_mode; u32 port; int err; diff --git a/drivers/net/dsa/ocelot/felix_vsc9959.c b/drivers/net/dsa/ocelot/felix_vsc9959.c index 03482616faa7..2c812b481778 100644 --- a/drivers/net/dsa/ocelot/felix_vsc9959.c +++ b/drivers/net/dsa/ocelot/felix_vsc9959.c @@ -645,6 +645,27 @@ static void vsc9959_pcs_init_sgmii(struct phy_device *pcs, const struct phylink_link_state *state) { if (link_an_mode == MLO_AN_INBAND) { + int bmsr, bmcr; + + /* Some PHYs like VSC8234 don't like it when AN restarts on + * their system side and they restart line side AN too, going + * into an endless link up/down loop. Don't restart PCS AN if + * link is up already. + * We do check that AN is enabled just in case this is the 1st + * call, PCS detects a carrier but AN is disabled from power on + * or by boot loader. + */ + bmcr = phy_read(pcs, MII_BMCR); + if (bmcr < 0) + return; + + bmsr = phy_read(pcs, MII_BMSR); + if (bmsr < 0) + return; + + if ((bmcr & BMCR_ANENABLE) && (bmsr & BMSR_LSTATUS)) + return; + /* SGMII spec requires tx_config_Reg[15:0] to be exactly 0x4001 * for the MAC PCS in order to acknowledge the AN. */ @@ -892,7 +913,6 @@ static void vsc9959_pcs_link_state_usxgmii(struct phy_device *pcs, break; } - pcs->link = USXGMII_LPA_LNKS(lpa); if (USXGMII_LPA_DUPLEX(lpa)) pcs->duplex = DUPLEX_FULL; else diff --git a/drivers/net/dsa/qca/ar9331.c b/drivers/net/dsa/qca/ar9331.c index da3bece75e21..de25f99e995a 100644 --- a/drivers/net/dsa/qca/ar9331.c +++ b/drivers/net/dsa/qca/ar9331.c @@ -347,7 +347,8 @@ static void ar9331_sw_port_disable(struct dsa_switch *ds, int port) } static enum dsa_tag_protocol ar9331_sw_get_tag_protocol(struct dsa_switch *ds, - int port) + int port, + enum dsa_tag_protocol m) { return DSA_TAG_PROTO_AR9331; } diff --git a/drivers/net/dsa/qca8k.c b/drivers/net/dsa/qca8k.c index e548289df31e..9f4205b4439b 100644 --- a/drivers/net/dsa/qca8k.c +++ b/drivers/net/dsa/qca8k.c @@ -1017,7 +1017,8 @@ qca8k_port_fdb_dump(struct dsa_switch *ds, int port, } static enum dsa_tag_protocol -qca8k_get_tag_protocol(struct dsa_switch *ds, int port) +qca8k_get_tag_protocol(struct dsa_switch *ds, int port, + enum dsa_tag_protocol mp) { return DSA_TAG_PROTO_QCA; } diff --git a/drivers/net/dsa/rtl8366rb.c b/drivers/net/dsa/rtl8366rb.c index f5cc8b0a7c74..fd1977590cb4 100644 --- a/drivers/net/dsa/rtl8366rb.c +++ b/drivers/net/dsa/rtl8366rb.c @@ -964,7 +964,8 @@ static int rtl8366rb_setup(struct dsa_switch *ds) } static enum dsa_tag_protocol rtl8366_get_tag_protocol(struct dsa_switch *ds, - int port) + int port, + enum dsa_tag_protocol mp) { /* For now, the RTL switches are handled without any custom tags. * diff --git a/drivers/net/dsa/sja1105/sja1105_main.c b/drivers/net/dsa/sja1105/sja1105_main.c index 61795833c8f5..03ba6d25f7fe 100644 --- a/drivers/net/dsa/sja1105/sja1105_main.c +++ b/drivers/net/dsa/sja1105/sja1105_main.c @@ -574,7 +574,7 @@ static int sja1105_parse_ports_node(struct sja1105_private *priv, struct device *dev = &priv->spidev->dev; struct device_node *child; - for_each_child_of_node(ports_node, child) { + for_each_available_child_of_node(ports_node, child) { struct device_node *phy_node; phy_interface_t phy_mode; u32 index; @@ -1534,7 +1534,8 @@ static int sja1105_setup_8021q_tagging(struct dsa_switch *ds, bool enabled) } static enum dsa_tag_protocol -sja1105_get_tag_protocol(struct dsa_switch *ds, int port) +sja1105_get_tag_protocol(struct dsa_switch *ds, int port, + enum dsa_tag_protocol mp) { return DSA_TAG_PROTO_SJA1105; } diff --git a/drivers/net/dsa/vitesse-vsc73xx-core.c b/drivers/net/dsa/vitesse-vsc73xx-core.c index 69fc0110ce04..6e21a2a5cf01 100644 --- a/drivers/net/dsa/vitesse-vsc73xx-core.c +++ b/drivers/net/dsa/vitesse-vsc73xx-core.c @@ -542,7 +542,8 @@ static int vsc73xx_phy_write(struct dsa_switch *ds, int phy, int regnum, } static enum dsa_tag_protocol vsc73xx_get_tag_protocol(struct dsa_switch *ds, - int port) + int port, + enum dsa_tag_protocol mp) { /* The switch internally uses a 8 byte header with length, * source port, tag, LPA and priority. This is supposedly diff --git a/drivers/net/ethernet/3com/3c59x.c b/drivers/net/ethernet/3com/3c59x.c index fc046797c0ea..a2b7f7ab8170 100644 --- a/drivers/net/ethernet/3com/3c59x.c +++ b/drivers/net/ethernet/3com/3c59x.c @@ -1548,7 +1548,7 @@ vortex_up(struct net_device *dev) struct vortex_private *vp = netdev_priv(dev); void __iomem *ioaddr = vp->ioaddr; unsigned int config; - int i, mii_reg1, mii_reg5, err = 0; + int i, mii_reg5, err = 0; if (VORTEX_PCI(vp)) { pci_set_power_state(VORTEX_PCI(vp), PCI_D0); /* Go active */ @@ -1605,7 +1605,7 @@ vortex_up(struct net_device *dev) window_write32(vp, config, 3, Wn3_Config); if (dev->if_port == XCVR_MII || dev->if_port == XCVR_NWAY) { - mii_reg1 = mdio_read(dev, vp->phys[0], MII_BMSR); + mdio_read(dev, vp->phys[0], MII_BMSR); mii_reg5 = mdio_read(dev, vp->phys[0], MII_LPA); vp->partner_flow_ctrl = ((mii_reg5 & 0x0400) != 0); vp->mii.full_duplex = vp->full_duplex; diff --git a/drivers/net/ethernet/agere/et131x.c b/drivers/net/ethernet/agere/et131x.c index 3c51d8c502ed..cb6a761d5c11 100644 --- a/drivers/net/ethernet/agere/et131x.c +++ b/drivers/net/ethernet/agere/et131x.c @@ -3651,15 +3651,6 @@ static int et131x_close(struct net_device *netdev) return del_timer_sync(&adapter->error_timer); } -static int et131x_ioctl(struct net_device *netdev, struct ifreq *reqbuf, - int cmd) -{ - if (!netdev->phydev) - return -EINVAL; - - return phy_mii_ioctl(netdev->phydev, reqbuf, cmd); -} - /* et131x_set_packet_filter - Configures the Rx Packet filtering */ static int et131x_set_packet_filter(struct et131x_adapter *adapter) { @@ -3899,7 +3890,7 @@ static const struct net_device_ops et131x_netdev_ops = { .ndo_set_mac_address = eth_mac_addr, .ndo_validate_addr = eth_validate_addr, .ndo_get_stats = et131x_stats, - .ndo_do_ioctl = et131x_ioctl, + .ndo_do_ioctl = phy_do_ioctl, }; static int et131x_pci_setup(struct pci_dev *pdev, diff --git a/drivers/net/ethernet/amd/a2065.c b/drivers/net/ethernet/amd/a2065.c index a3faf4feb204..2f808dbc8b0e 100644 --- a/drivers/net/ethernet/amd/a2065.c +++ b/drivers/net/ethernet/amd/a2065.c @@ -118,10 +118,6 @@ struct lance_private { int auto_select; /* cable-selection by carrier */ unsigned short busmaster_regval; -#ifdef CONFIG_SUNLANCE - struct Linux_SBus_DMA *ledma; /* if set this points to ledma and arch=4m */ - int burst_sizes; /* ledma SBus burst sizes */ -#endif struct timer_list multicast_timer; struct net_device *dev; }; @@ -551,11 +547,10 @@ static netdev_tx_t lance_start_xmit(struct sk_buff *skb, if (!lance_tx_buffs_avail(lp)) goto out_free; -#ifdef DEBUG /* dump the packet */ - print_hex_dump(KERN_DEBUG, "skb->data: ", DUMP_PREFIX_NONE, - 16, 1, skb->data, 64, true); -#endif + print_hex_dump_debug("skb->data: ", DUMP_PREFIX_NONE, 16, 1, skb->data, + 64, true); + entry = lp->tx_new & lp->tx_ring_mod_mask; ib->btx_ring[entry].length = (-skblen) | 0xf000; ib->btx_ring[entry].misc = 0; diff --git a/drivers/net/ethernet/amd/declance.c b/drivers/net/ethernet/amd/declance.c index 6592a2db9efb..7282ce55ffb8 100644 --- a/drivers/net/ethernet/amd/declance.c +++ b/drivers/net/ethernet/amd/declance.c @@ -608,7 +608,7 @@ static int lance_rx(struct net_device *dev) len = (*rds_ptr(rd, mblength, lp->type) & 0xfff) - 4; skb = netdev_alloc_skb(dev, len + 2); - if (skb == 0) { + if (!skb) { dev->stats.rx_dropped++; *rds_ptr(rd, mblength, lp->type) = 0; *rds_ptr(rd, rmd1, lp->type) = diff --git a/drivers/net/ethernet/amd/xgbe/xgbe-phy-v2.c b/drivers/net/ethernet/amd/xgbe/xgbe-phy-v2.c index 128cd648ba99..46c3c1ca38d6 100644 --- a/drivers/net/ethernet/amd/xgbe/xgbe-phy-v2.c +++ b/drivers/net/ethernet/amd/xgbe/xgbe-phy-v2.c @@ -1227,7 +1227,7 @@ static bool xgbe_phy_sfp_verify_eeprom(u8 cc_in, u8 *buf, unsigned int len) for (cc = 0; len; buf++, len--) cc += *buf; - return (cc == cc_in) ? true : false; + return cc == cc_in; } static int xgbe_phy_sfp_read_eeprom(struct xgbe_prv_data *pdata) diff --git a/drivers/net/ethernet/aquantia/atlantic/aq_nic.c b/drivers/net/ethernet/aquantia/atlantic/aq_nic.c index a17a4da7bc15..c85e3e29012c 100644 --- a/drivers/net/ethernet/aquantia/atlantic/aq_nic.c +++ b/drivers/net/ethernet/aquantia/atlantic/aq_nic.c @@ -403,6 +403,8 @@ int aq_nic_start(struct aq_nic_s *self) if (err < 0) goto err_exit; + aq_nic_set_loopback(self); + err = self->aq_hw_ops->hw_start(self->aq_hw); if (err < 0) goto err_exit; @@ -413,8 +415,6 @@ int aq_nic_start(struct aq_nic_s *self) INIT_WORK(&self->service_task, aq_nic_service_task); - aq_nic_set_loopback(self); - timer_setup(&self->service_timer, aq_nic_service_timer_cb, 0); aq_nic_service_timer_cb(&self->service_timer); diff --git a/drivers/net/ethernet/aquantia/atlantic/hw_atl/hw_atl_b0.c b/drivers/net/ethernet/aquantia/atlantic/hw_atl/hw_atl_b0.c index 58e891af6e09..ec041f78d063 100644 --- a/drivers/net/ethernet/aquantia/atlantic/hw_atl/hw_atl_b0.c +++ b/drivers/net/ethernet/aquantia/atlantic/hw_atl/hw_atl_b0.c @@ -1525,9 +1525,6 @@ const struct aq_hw_ops hw_atl_ops_b0 = { .rx_extract_ts = hw_atl_b0_rx_extract_ts, .extract_hwts = hw_atl_b0_extract_hwts, .hw_set_offload = hw_atl_b0_hw_offload_set, - .hw_get_hw_stats = hw_atl_utils_get_hw_stats, - .hw_get_fw_version = hw_atl_utils_get_fw_version, - .hw_set_offload = hw_atl_b0_hw_offload_set, .hw_set_loopback = hw_atl_b0_set_loopback, .hw_set_fc = hw_atl_b0_set_fc, }; diff --git a/drivers/net/ethernet/aquantia/atlantic/hw_atl/hw_atl_utils.c b/drivers/net/ethernet/aquantia/atlantic/hw_atl/hw_atl_utils.c index 8910b62e67ed..f547baa6c954 100644 --- a/drivers/net/ethernet/aquantia/atlantic/hw_atl/hw_atl_utils.c +++ b/drivers/net/ethernet/aquantia/atlantic/hw_atl/hw_atl_utils.c @@ -667,9 +667,7 @@ int hw_atl_utils_mpi_get_link_status(struct aq_hw_s *self) u32 speed; mpi_state = hw_atl_utils_mpi_get_state(self); - speed = mpi_state & (FW2X_RATE_100M | FW2X_RATE_1G | - FW2X_RATE_2G5 | FW2X_RATE_5G | - FW2X_RATE_10G); + speed = mpi_state >> HW_ATL_MPI_SPEED_SHIFT; if (!speed) { link_status->mbps = 0U; diff --git a/drivers/net/ethernet/atheros/ag71xx.c b/drivers/net/ethernet/atheros/ag71xx.c index 88e4e1500a8e..5ce2df482d8c 100644 --- a/drivers/net/ethernet/atheros/ag71xx.c +++ b/drivers/net/ethernet/atheros/ag71xx.c @@ -1394,14 +1394,6 @@ err_drop: return NETDEV_TX_OK; } -static int ag71xx_do_ioctl(struct net_device *ndev, struct ifreq *ifr, int cmd) -{ - if (!ndev->phydev) - return -EINVAL; - - return phy_mii_ioctl(ndev->phydev, ifr, cmd); -} - static void ag71xx_oom_timer_handler(struct timer_list *t) { struct ag71xx *ag = from_timer(ag, t, oom_timer); @@ -1618,7 +1610,7 @@ static const struct net_device_ops ag71xx_netdev_ops = { .ndo_open = ag71xx_open, .ndo_stop = ag71xx_stop, .ndo_start_xmit = ag71xx_hard_start_xmit, - .ndo_do_ioctl = ag71xx_do_ioctl, + .ndo_do_ioctl = phy_do_ioctl, .ndo_tx_timeout = ag71xx_tx_timeout, .ndo_change_mtu = ag71xx_change_mtu, .ndo_set_mac_address = eth_mac_addr, diff --git a/drivers/net/ethernet/aurora/nb8800.c b/drivers/net/ethernet/aurora/nb8800.c index 30b455013bf3..bc273e0db7ff 100644 --- a/drivers/net/ethernet/aurora/nb8800.c +++ b/drivers/net/ethernet/aurora/nb8800.c @@ -1005,18 +1005,13 @@ static int nb8800_stop(struct net_device *dev) return 0; } -static int nb8800_ioctl(struct net_device *dev, struct ifreq *rq, int cmd) -{ - return phy_mii_ioctl(dev->phydev, rq, cmd); -} - static const struct net_device_ops nb8800_netdev_ops = { .ndo_open = nb8800_open, .ndo_stop = nb8800_stop, .ndo_start_xmit = nb8800_xmit, .ndo_set_mac_address = nb8800_set_mac_address, .ndo_set_rx_mode = nb8800_set_rx_mode, - .ndo_do_ioctl = nb8800_ioctl, + .ndo_do_ioctl = phy_do_ioctl, .ndo_validate_addr = eth_validate_addr, }; diff --git a/drivers/net/ethernet/broadcom/b44.c b/drivers/net/ethernet/broadcom/b44.c index 5b3464c3e8d1..a780b7215021 100644 --- a/drivers/net/ethernet/broadcom/b44.c +++ b/drivers/net/ethernet/broadcom/b44.c @@ -1516,8 +1516,10 @@ static int b44_magic_pattern(u8 *macaddr, u8 *ppattern, u8 *pmask, int offset) int ethaddr_bytes = ETH_ALEN; memset(ppattern + offset, 0xff, magicsync); - for (j = 0; j < magicsync; j++) - set_bit(len++, (unsigned long *) pmask); + for (j = 0; j < magicsync; j++) { + pmask[len >> 3] |= BIT(len & 7); + len++; + } for (j = 0; j < B44_MAX_PATTERNS; j++) { if ((B44_PATTERN_SIZE - len) >= ETH_ALEN) @@ -1529,7 +1531,8 @@ static int b44_magic_pattern(u8 *macaddr, u8 *ppattern, u8 *pmask, int offset) for (k = 0; k< ethaddr_bytes; k++) { ppattern[offset + magicsync + (j * ETH_ALEN) + k] = macaddr[k]; - set_bit(len++, (unsigned long *) pmask); + pmask[len >> 3] |= BIT(len & 7); + len++; } } return len - 1; diff --git a/drivers/net/ethernet/broadcom/bcmsysport.c b/drivers/net/ethernet/broadcom/bcmsysport.c index 1907e47fd0af..f07ac0e0af59 100644 --- a/drivers/net/ethernet/broadcom/bcmsysport.c +++ b/drivers/net/ethernet/broadcom/bcmsysport.c @@ -2323,7 +2323,7 @@ static int bcm_sysport_map_queues(struct notifier_block *nb, ring->switch_queue = qp; ring->switch_port = port; ring->inspect = true; - priv->ring_map[q + port * num_tx_queues] = ring; + priv->ring_map[qp + port * num_tx_queues] = ring; qp++; } @@ -2338,7 +2338,7 @@ static int bcm_sysport_unmap_queues(struct notifier_block *nb, struct net_device *slave_dev; unsigned int num_tx_queues; struct net_device *dev; - unsigned int q, port; + unsigned int q, qp, port; priv = container_of(nb, struct bcm_sysport_priv, dsa_notifier); if (priv->netdev != info->master) @@ -2364,7 +2364,8 @@ static int bcm_sysport_unmap_queues(struct notifier_block *nb, continue; ring->inspect = false; - priv->ring_map[q + port * num_tx_queues] = NULL; + qp = ring->switch_queue; + priv->ring_map[qp + port * num_tx_queues] = NULL; } return 0; diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c b/drivers/net/ethernet/broadcom/bnxt/bnxt.c index 7b0fe19d9ecb..198c69dceeef 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c @@ -944,6 +944,7 @@ static struct sk_buff *bnxt_rx_page_skb(struct bnxt *bp, dma_addr -= bp->rx_dma_offset; dma_unmap_page_attrs(&bp->pdev->dev, dma_addr, PAGE_SIZE, bp->rx_dir, DMA_ATTR_WEAK_ORDERING); + page_pool_release_page(rxr->page_pool, page); if (unlikely(!payload)) payload = eth_get_headlen(bp->dev, data_ptr, len); @@ -10822,6 +10823,7 @@ static void bnxt_fw_reset_task(struct work_struct *work) smp_mb__before_atomic(); clear_bit(BNXT_STATE_IN_FW_RESET, &bp->state); bnxt_ulp_start(bp, rc); + bnxt_dl_health_recovery_done(bp); bnxt_dl_health_status_update(bp, true); rtnl_unlock(); break; @@ -11065,11 +11067,23 @@ static bool bnxt_fltr_match(struct bnxt_ntuple_filter *f1, struct flow_keys *keys1 = &f1->fkeys; struct flow_keys *keys2 = &f2->fkeys; - if (keys1->addrs.v4addrs.src == keys2->addrs.v4addrs.src && - keys1->addrs.v4addrs.dst == keys2->addrs.v4addrs.dst && - keys1->ports.ports == keys2->ports.ports && - keys1->basic.ip_proto == keys2->basic.ip_proto && - keys1->basic.n_proto == keys2->basic.n_proto && + if (keys1->basic.n_proto != keys2->basic.n_proto || + keys1->basic.ip_proto != keys2->basic.ip_proto) + return false; + + if (keys1->basic.n_proto == htons(ETH_P_IP)) { + if (keys1->addrs.v4addrs.src != keys2->addrs.v4addrs.src || + keys1->addrs.v4addrs.dst != keys2->addrs.v4addrs.dst) + return false; + } else { + if (memcmp(&keys1->addrs.v6addrs.src, &keys2->addrs.v6addrs.src, + sizeof(keys1->addrs.v6addrs.src)) || + memcmp(&keys1->addrs.v6addrs.dst, &keys2->addrs.v6addrs.dst, + sizeof(keys1->addrs.v6addrs.dst))) + return false; + } + + if (keys1->ports.ports == keys2->ports.ports && keys1->control.flags == keys2->control.flags && ether_addr_equal(f1->src_mac_addr, f2->src_mac_addr) && ether_addr_equal(f1->dst_mac_addr, f2->dst_mac_addr)) @@ -11361,7 +11375,7 @@ int bnxt_get_port_parent_id(struct net_device *dev, return -EOPNOTSUPP; /* The PF and it's VF-reps only support the switchdev framework */ - if (!BNXT_PF(bp)) + if (!BNXT_PF(bp) || !(bp->flags & BNXT_FLAG_DSN_VALID)) return -EOPNOTSUPP; ppid->id_len = sizeof(bp->switch_id); @@ -11734,6 +11748,7 @@ static int bnxt_pcie_dsn_get(struct bnxt *bp, u8 dsn[]) put_unaligned_le32(dw, &dsn[0]); pci_read_config_dword(pdev, pos + 4, &dw); put_unaligned_le32(dw, &dsn[4]); + bp->flags |= BNXT_FLAG_DSN_VALID; return 0; } @@ -11845,9 +11860,7 @@ static int bnxt_init_one(struct pci_dev *pdev, const struct pci_device_id *ent) if (BNXT_PF(bp)) { /* Read the adapter's DSN to use as the eswitch switch_id */ - rc = bnxt_pcie_dsn_get(bp, bp->switch_id); - if (rc) - goto init_err_pci_clean; + bnxt_pcie_dsn_get(bp, bp->switch_id); } /* MTU range: 60 - FW defined max */ diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.h b/drivers/net/ethernet/broadcom/bnxt/bnxt.h index 505af5cfb1bd..f14335433a64 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt.h +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.h @@ -1532,6 +1532,7 @@ struct bnxt { #define BNXT_FLAG_NO_AGG_RINGS 0x20000 #define BNXT_FLAG_RX_PAGE_MODE 0x40000 #define BNXT_FLAG_MULTI_HOST 0x100000 + #define BNXT_FLAG_DSN_VALID 0x200000 #define BNXT_FLAG_DOUBLE_DB 0x400000 #define BNXT_FLAG_CHIP_NITRO_A0 0x1000000 #define BNXT_FLAG_DIM 0x2000000 @@ -1936,9 +1937,6 @@ static inline bool bnxt_cfa_hwrm_message(u16 req_type) case HWRM_CFA_ENCAP_RECORD_FREE: case HWRM_CFA_DECAP_FILTER_ALLOC: case HWRM_CFA_DECAP_FILTER_FREE: - case HWRM_CFA_NTUPLE_FILTER_ALLOC: - case HWRM_CFA_NTUPLE_FILTER_FREE: - case HWRM_CFA_NTUPLE_FILTER_CFG: case HWRM_CFA_EM_FLOW_ALLOC: case HWRM_CFA_EM_FLOW_FREE: case HWRM_CFA_EM_FLOW_CFG: diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_devlink.c b/drivers/net/ethernet/broadcom/bnxt/bnxt_devlink.c index 3eedd4477218..0c3d224637b9 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt_devlink.c +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_devlink.c @@ -89,7 +89,7 @@ static int bnxt_fw_reset_recover(struct devlink_health_reporter *reporter, return -EOPNOTSUPP; bnxt_fw_reset(bp); - return 0; + return -EINPROGRESS; } static const @@ -116,7 +116,7 @@ static int bnxt_fw_fatal_recover(struct devlink_health_reporter *reporter, else if (event == BNXT_FW_EXCEPTION_SP_EVENT) bnxt_fw_exception(bp); - return 0; + return -EINPROGRESS; } static const @@ -262,6 +262,16 @@ void bnxt_dl_health_status_update(struct bnxt *bp, bool healthy) health->fatal = false; } +void bnxt_dl_health_recovery_done(struct bnxt *bp) +{ + struct bnxt_fw_health *hlth = bp->fw_health; + + if (hlth->fatal) + devlink_health_reporter_recovery_done(hlth->fw_fatal_reporter); + else + devlink_health_reporter_recovery_done(hlth->fw_reset_reporter); +} + static const struct devlink_ops bnxt_dl_ops = { #ifdef CONFIG_BNXT_SRIOV .eswitch_mode_set = bnxt_dl_eswitch_mode_set, diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_devlink.h b/drivers/net/ethernet/broadcom/bnxt/bnxt_devlink.h index 6db6c3dac472..08aaa4441c78 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt_devlink.h +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_devlink.h @@ -58,6 +58,7 @@ struct bnxt_dl_nvm_param { void bnxt_devlink_health_report(struct bnxt *bp, unsigned long event); void bnxt_dl_health_status_update(struct bnxt *bp, bool healthy); +void bnxt_dl_health_recovery_done(struct bnxt *bp); void bnxt_dl_fw_reporters_create(struct bnxt *bp); void bnxt_dl_fw_reporters_destroy(struct bnxt *bp, bool all); int bnxt_dl_register(struct bnxt *bp); diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_vfr.c b/drivers/net/ethernet/broadcom/bnxt/bnxt_vfr.c index f9bf7d7250ab..b010b34cdaf8 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt_vfr.c +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_vfr.c @@ -398,6 +398,9 @@ static int bnxt_vf_reps_create(struct bnxt *bp) struct net_device *dev; int rc, i; + if (!(bp->flags & BNXT_FLAG_DSN_VALID)) + return -ENODEV; + bp->vf_reps = kcalloc(num_vfs, sizeof(vf_rep), GFP_KERNEL); if (!bp->vf_reps) return -ENOMEM; diff --git a/drivers/net/ethernet/broadcom/tg3.c b/drivers/net/ethernet/broadcom/tg3.c index 460b4992914a..88466255bf66 100644 --- a/drivers/net/ethernet/broadcom/tg3.c +++ b/drivers/net/ethernet/broadcom/tg3.c @@ -7874,8 +7874,8 @@ static netdev_tx_t tg3_start_xmit(struct sk_buff *, struct net_device *); static int tg3_tso_bug(struct tg3 *tp, struct tg3_napi *tnapi, struct netdev_queue *txq, struct sk_buff *skb) { - struct sk_buff *segs, *nskb; u32 frag_cnt_est = skb_shinfo(skb)->gso_segs * 3; + struct sk_buff *segs, *seg, *next; /* Estimate the number of fragments in the worst case */ if (unlikely(tg3_tx_avail(tnapi) <= frag_cnt_est)) { @@ -7898,12 +7898,10 @@ static int tg3_tso_bug(struct tg3 *tp, struct tg3_napi *tnapi, if (IS_ERR(segs) || !segs) goto tg3_tso_bug_end; - do { - nskb = segs; - segs = segs->next; - nskb->next = NULL; - tg3_start_xmit(nskb, tp->dev); - } while (segs); + skb_list_walk_safe(segs, seg, next) { + skb_mark_not_on_list(seg); + tg3_start_xmit(seg, tp->dev); + } tg3_tso_bug_end: dev_consume_skb_any(skb); diff --git a/drivers/net/ethernet/cadence/macb.h b/drivers/net/ethernet/cadence/macb.h index 19fe4f4867c7..dbf7070fcdba 100644 --- a/drivers/net/ethernet/cadence/macb.h +++ b/drivers/net/ethernet/cadence/macb.h @@ -630,10 +630,17 @@ #define GEM_CLK_DIV96 5 /* Constants for MAN register */ -#define MACB_MAN_SOF 1 -#define MACB_MAN_WRITE 1 -#define MACB_MAN_READ 2 -#define MACB_MAN_CODE 2 +#define MACB_MAN_C22_SOF 1 +#define MACB_MAN_C22_WRITE 1 +#define MACB_MAN_C22_READ 2 +#define MACB_MAN_C22_CODE 2 + +#define MACB_MAN_C45_SOF 0 +#define MACB_MAN_C45_ADDR 0 +#define MACB_MAN_C45_WRITE 1 +#define MACB_MAN_C45_POST_READ_INCR 2 +#define MACB_MAN_C45_READ 3 +#define MACB_MAN_C45_CODE 2 /* Capability mask bits */ #define MACB_CAPS_ISR_CLEAR_ON_WRITE 0x00000001 diff --git a/drivers/net/ethernet/cadence/macb_main.c b/drivers/net/ethernet/cadence/macb_main.c index c5ee363ca5dc..7a2fe63d1136 100644 --- a/drivers/net/ethernet/cadence/macb_main.c +++ b/drivers/net/ethernet/cadence/macb_main.c @@ -337,11 +337,30 @@ static int macb_mdio_read(struct mii_bus *bus, int mii_id, int regnum) if (status < 0) goto mdio_read_exit; - macb_writel(bp, MAN, (MACB_BF(SOF, MACB_MAN_SOF) - | MACB_BF(RW, MACB_MAN_READ) - | MACB_BF(PHYA, mii_id) - | MACB_BF(REGA, regnum) - | MACB_BF(CODE, MACB_MAN_CODE))); + if (regnum & MII_ADDR_C45) { + macb_writel(bp, MAN, (MACB_BF(SOF, MACB_MAN_C45_SOF) + | MACB_BF(RW, MACB_MAN_C45_ADDR) + | MACB_BF(PHYA, mii_id) + | MACB_BF(REGA, (regnum >> 16) & 0x1F) + | MACB_BF(DATA, regnum & 0xFFFF) + | MACB_BF(CODE, MACB_MAN_C45_CODE))); + + status = macb_mdio_wait_for_idle(bp); + if (status < 0) + goto mdio_read_exit; + + macb_writel(bp, MAN, (MACB_BF(SOF, MACB_MAN_C45_SOF) + | MACB_BF(RW, MACB_MAN_C45_READ) + | MACB_BF(PHYA, mii_id) + | MACB_BF(REGA, (regnum >> 16) & 0x1F) + | MACB_BF(CODE, MACB_MAN_C45_CODE))); + } else { + macb_writel(bp, MAN, (MACB_BF(SOF, MACB_MAN_C22_SOF) + | MACB_BF(RW, MACB_MAN_C22_READ) + | MACB_BF(PHYA, mii_id) + | MACB_BF(REGA, regnum) + | MACB_BF(CODE, MACB_MAN_C22_CODE))); + } status = macb_mdio_wait_for_idle(bp); if (status < 0) @@ -370,12 +389,32 @@ static int macb_mdio_write(struct mii_bus *bus, int mii_id, int regnum, if (status < 0) goto mdio_write_exit; - macb_writel(bp, MAN, (MACB_BF(SOF, MACB_MAN_SOF) - | MACB_BF(RW, MACB_MAN_WRITE) - | MACB_BF(PHYA, mii_id) - | MACB_BF(REGA, regnum) - | MACB_BF(CODE, MACB_MAN_CODE) - | MACB_BF(DATA, value))); + if (regnum & MII_ADDR_C45) { + macb_writel(bp, MAN, (MACB_BF(SOF, MACB_MAN_C45_SOF) + | MACB_BF(RW, MACB_MAN_C45_ADDR) + | MACB_BF(PHYA, mii_id) + | MACB_BF(REGA, (regnum >> 16) & 0x1F) + | MACB_BF(DATA, regnum & 0xFFFF) + | MACB_BF(CODE, MACB_MAN_C45_CODE))); + + status = macb_mdio_wait_for_idle(bp); + if (status < 0) + goto mdio_write_exit; + + macb_writel(bp, MAN, (MACB_BF(SOF, MACB_MAN_C45_SOF) + | MACB_BF(RW, MACB_MAN_C45_WRITE) + | MACB_BF(PHYA, mii_id) + | MACB_BF(REGA, (regnum >> 16) & 0x1F) + | MACB_BF(CODE, MACB_MAN_C45_CODE) + | MACB_BF(DATA, value))); + } else { + macb_writel(bp, MAN, (MACB_BF(SOF, MACB_MAN_C22_SOF) + | MACB_BF(RW, MACB_MAN_C22_WRITE) + | MACB_BF(PHYA, mii_id) + | MACB_BF(REGA, regnum) + | MACB_BF(CODE, MACB_MAN_C22_CODE) + | MACB_BF(DATA, value))); + } status = macb_mdio_wait_for_idle(bp); if (status < 0) @@ -611,21 +650,24 @@ static const struct phylink_mac_ops macb_phylink_ops = { .mac_link_up = macb_mac_link_up, }; +static bool macb_phy_handle_exists(struct device_node *dn) +{ + dn = of_parse_phandle(dn, "phy-handle", 0); + of_node_put(dn); + return dn != NULL; +} + static int macb_phylink_connect(struct macb *bp) { + struct device_node *dn = bp->pdev->dev.of_node; struct net_device *dev = bp->dev; struct phy_device *phydev; int ret; - if (bp->pdev->dev.of_node && - of_parse_phandle(bp->pdev->dev.of_node, "phy-handle", 0)) { - ret = phylink_of_phy_connect(bp->phylink, bp->pdev->dev.of_node, - 0); - if (ret) { - netdev_err(dev, "Could not attach PHY (%d)\n", ret); - return ret; - } - } else { + if (dn) + ret = phylink_of_phy_connect(bp->phylink, dn, 0); + + if (!dn || (ret && !macb_phy_handle_exists(dn))) { phydev = phy_find_first(bp->mii_bus); if (!phydev) { netdev_err(dev, "no PHY found\n"); @@ -634,10 +676,11 @@ static int macb_phylink_connect(struct macb *bp) /* attach the mac to the phy */ ret = phylink_connect_phy(bp->phylink, phydev); - if (ret) { - netdev_err(dev, "Could not attach to PHY (%d)\n", ret); - return ret; - } + } + + if (ret) { + netdev_err(dev, "Could not attach PHY (%d)\n", ret); + return ret; } phylink_start(bp->phylink); @@ -4088,7 +4131,7 @@ static int fu540_c000_clk_init(struct platform_device *pdev, struct clk **pclk, mgmt->rate = 0; mgmt->hw.init = &init; - *tx_clk = clk_register(NULL, &mgmt->hw); + *tx_clk = devm_clk_register(&pdev->dev, &mgmt->hw); if (IS_ERR(*tx_clk)) return PTR_ERR(*tx_clk); @@ -4416,7 +4459,6 @@ err_out_free_netdev: err_disable_clocks: clk_disable_unprepare(tx_clk); - clk_unregister(tx_clk); clk_disable_unprepare(hclk); clk_disable_unprepare(pclk); clk_disable_unprepare(rx_clk); @@ -4446,7 +4488,6 @@ static int macb_remove(struct platform_device *pdev) pm_runtime_dont_use_autosuspend(&pdev->dev); if (!pm_runtime_suspended(&pdev->dev)) { clk_disable_unprepare(bp->tx_clk); - clk_unregister(bp->tx_clk); clk_disable_unprepare(bp->hclk); clk_disable_unprepare(bp->pclk); clk_disable_unprepare(bp->rx_clk); diff --git a/drivers/net/ethernet/cavium/octeon/octeon_mgmt.c b/drivers/net/ethernet/cavium/octeon/octeon_mgmt.c index cdd7e5da4a74..e9575887a4f8 100644 --- a/drivers/net/ethernet/cavium/octeon/octeon_mgmt.c +++ b/drivers/net/ethernet/cavium/octeon/octeon_mgmt.c @@ -790,9 +790,7 @@ static int octeon_mgmt_ioctl(struct net_device *netdev, case SIOCSHWTSTAMP: return octeon_mgmt_ioctl_hwtstamp(netdev, rq, cmd); default: - if (netdev->phydev) - return phy_mii_ioctl(netdev->phydev, rq, cmd); - return -EINVAL; + return phy_do_ioctl(netdev, rq, cmd); } } diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c index 1930e39f195e..649842a8aa28 100644 --- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c +++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c @@ -3163,9 +3163,9 @@ static int cxgb_set_tx_maxrate(struct net_device *dev, int index, u32 rate) { struct port_info *pi = netdev_priv(dev); struct adapter *adap = pi->adapter; + struct ch_sched_queue qe = { 0 }; + struct ch_sched_params p = { 0 }; struct sched_class *e; - struct ch_sched_params p; - struct ch_sched_queue qe; u32 req_rate; int err = 0; @@ -3182,6 +3182,15 @@ static int cxgb_set_tx_maxrate(struct net_device *dev, int index, u32 rate) return -EINVAL; } + qe.queue = index; + e = cxgb4_sched_queue_lookup(dev, &qe); + if (e && e->info.u.params.level != SCHED_CLASS_LEVEL_CL_RL) { + dev_err(adap->pdev_dev, + "Queue %u already bound to class %u of type: %u\n", + index, e->idx, e->info.u.params.level); + return -EBUSY; + } + /* Convert from Mbps to Kbps */ req_rate = rate * 1000; @@ -3211,7 +3220,6 @@ static int cxgb_set_tx_maxrate(struct net_device *dev, int index, u32 rate) return 0; /* Fetch any available unused or matching scheduling class */ - memset(&p, 0, sizeof(p)); p.type = SCHED_CLASS_TYPE_PACKET; p.u.params.level = SCHED_CLASS_LEVEL_CL_RL; p.u.params.mode = SCHED_CLASS_MODE_CLASS; diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_matchall.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_matchall.c index 24c3c2dc7171..1b7681a4eb32 100644 --- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_matchall.c +++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_matchall.c @@ -15,6 +15,8 @@ static int cxgb4_matchall_egress_validate(struct net_device *dev, struct flow_action *actions = &cls->rule->action; struct port_info *pi = netdev2pinfo(dev); struct flow_action_entry *entry; + struct ch_sched_queue qe; + struct sched_class *e; u64 max_link_rate; u32 i, speed; int ret; @@ -60,9 +62,61 @@ static int cxgb4_matchall_egress_validate(struct net_device *dev, } } + for (i = 0; i < pi->nqsets; i++) { + memset(&qe, 0, sizeof(qe)); + qe.queue = i; + + e = cxgb4_sched_queue_lookup(dev, &qe); + if (e && e->info.u.params.level != SCHED_CLASS_LEVEL_CH_RL) { + NL_SET_ERR_MSG_MOD(extack, + "Some queues are already bound to different class"); + return -EBUSY; + } + } + return 0; } +static int cxgb4_matchall_tc_bind_queues(struct net_device *dev, u32 tc) +{ + struct port_info *pi = netdev2pinfo(dev); + struct ch_sched_queue qe; + int ret; + u32 i; + + for (i = 0; i < pi->nqsets; i++) { + qe.queue = i; + qe.class = tc; + ret = cxgb4_sched_class_bind(dev, &qe, SCHED_QUEUE); + if (ret) + goto out_free; + } + + return 0; + +out_free: + while (i--) { + qe.queue = i; + qe.class = SCHED_CLS_NONE; + cxgb4_sched_class_unbind(dev, &qe, SCHED_QUEUE); + } + + return ret; +} + +static void cxgb4_matchall_tc_unbind_queues(struct net_device *dev) +{ + struct port_info *pi = netdev2pinfo(dev); + struct ch_sched_queue qe; + u32 i; + + for (i = 0; i < pi->nqsets; i++) { + qe.queue = i; + qe.class = SCHED_CLS_NONE; + cxgb4_sched_class_unbind(dev, &qe, SCHED_QUEUE); + } +} + static int cxgb4_matchall_alloc_tc(struct net_device *dev, struct tc_cls_matchall_offload *cls) { @@ -83,6 +137,7 @@ static int cxgb4_matchall_alloc_tc(struct net_device *dev, struct adapter *adap = netdev2adap(dev); struct flow_action_entry *entry; struct sched_class *e; + int ret; u32 i; tc_port_matchall = &adap->tc_matchall->port_matchall[pi->port_id]; @@ -101,10 +156,21 @@ static int cxgb4_matchall_alloc_tc(struct net_device *dev, return -ENOMEM; } + ret = cxgb4_matchall_tc_bind_queues(dev, e->idx); + if (ret) { + NL_SET_ERR_MSG_MOD(extack, + "Could not bind queues to traffic class"); + goto out_free; + } + tc_port_matchall->egress.hwtc = e->idx; tc_port_matchall->egress.cookie = cls->cookie; tc_port_matchall->egress.state = CXGB4_MATCHALL_STATE_ENABLED; return 0; + +out_free: + cxgb4_sched_class_free(dev, e->idx); + return ret; } static void cxgb4_matchall_free_tc(struct net_device *dev) @@ -114,6 +180,7 @@ static void cxgb4_matchall_free_tc(struct net_device *dev) struct adapter *adap = netdev2adap(dev); tc_port_matchall = &adap->tc_matchall->port_matchall[pi->port_id]; + cxgb4_matchall_tc_unbind_queues(dev); cxgb4_sched_class_free(dev, tc_port_matchall->egress.hwtc); tc_port_matchall->egress.hwtc = SCHED_CLS_NONE; diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_mqprio.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_mqprio.c index 8971dddcdb7a..ec3eb45ee3b4 100644 --- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_mqprio.c +++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_mqprio.c @@ -12,8 +12,9 @@ static int cxgb4_mqprio_validate(struct net_device *dev, struct port_info *pi = netdev2pinfo(dev); struct adapter *adap = netdev2adap(dev); u32 speed, qcount = 0, qoffset = 0; + u32 start_a, start_b, end_a, end_b; int ret; - u8 i; + u8 i, j; if (!mqprio->qopt.num_tc) return 0; @@ -47,6 +48,31 @@ static int cxgb4_mqprio_validate(struct net_device *dev, qoffset = max_t(u16, mqprio->qopt.offset[i], qoffset); qcount += mqprio->qopt.count[i]; + start_a = mqprio->qopt.offset[i]; + end_a = start_a + mqprio->qopt.count[i] - 1; + for (j = i + 1; j < mqprio->qopt.num_tc; j++) { + start_b = mqprio->qopt.offset[j]; + end_b = start_b + mqprio->qopt.count[j] - 1; + + /* If queue count is 0, then the traffic + * belonging to this class will not use + * ETHOFLD queues. So, no need to validate + * further. + */ + if (!mqprio->qopt.count[i]) + break; + + if (!mqprio->qopt.count[j]) + continue; + + if (max_t(u32, start_a, start_b) <= + min_t(u32, end_a, end_b)) { + netdev_err(dev, + "Queues can't overlap across tc\n"); + return -EINVAL; + } + } + /* Convert byte per second to bits per second */ min_rate += (mqprio->min_rate[i] * 8); max_rate += (mqprio->max_rate[i] * 8); diff --git a/drivers/net/ethernet/chelsio/cxgb4/sched.c b/drivers/net/ethernet/chelsio/cxgb4/sched.c index 3e61bd5d0c29..cebe1412d960 100644 --- a/drivers/net/ethernet/chelsio/cxgb4/sched.c +++ b/drivers/net/ethernet/chelsio/cxgb4/sched.c @@ -165,6 +165,22 @@ static void *t4_sched_entry_lookup(struct port_info *pi, return found; } +struct sched_class *cxgb4_sched_queue_lookup(struct net_device *dev, + struct ch_sched_queue *p) +{ + struct port_info *pi = netdev2pinfo(dev); + struct sched_queue_entry *qe = NULL; + struct adapter *adap = pi->adapter; + struct sge_eth_txq *txq; + + if (p->queue < 0 || p->queue >= pi->nqsets) + return NULL; + + txq = &adap->sge.ethtxq[pi->first_qset + p->queue]; + qe = t4_sched_entry_lookup(pi, SCHED_QUEUE, txq->q.cntxt_id); + return qe ? &pi->sched_tbl->tab[qe->param.class] : NULL; +} + static int t4_sched_queue_unbind(struct port_info *pi, struct ch_sched_queue *p) { struct sched_queue_entry *qe = NULL; diff --git a/drivers/net/ethernet/chelsio/cxgb4/sched.h b/drivers/net/ethernet/chelsio/cxgb4/sched.h index e92ff68bdd0a..5cc74a5a1774 100644 --- a/drivers/net/ethernet/chelsio/cxgb4/sched.h +++ b/drivers/net/ethernet/chelsio/cxgb4/sched.h @@ -103,6 +103,8 @@ static inline bool valid_class_id(struct net_device *dev, u8 class_id) return true; } +struct sched_class *cxgb4_sched_queue_lookup(struct net_device *dev, + struct ch_sched_queue *p); int cxgb4_sched_class_bind(struct net_device *dev, void *arg, enum sched_bind_type type); int cxgb4_sched_class_unbind(struct net_device *dev, void *arg, diff --git a/drivers/net/ethernet/faraday/ftgmac100.c b/drivers/net/ethernet/faraday/ftgmac100.c index 48b3b72fe02e..4572797f00d7 100644 --- a/drivers/net/ethernet/faraday/ftgmac100.c +++ b/drivers/net/ethernet/faraday/ftgmac100.c @@ -1536,15 +1536,6 @@ static int ftgmac100_stop(struct net_device *netdev) return 0; } -/* optional */ -static int ftgmac100_do_ioctl(struct net_device *netdev, struct ifreq *ifr, int cmd) -{ - if (!netdev->phydev) - return -ENXIO; - - return phy_mii_ioctl(netdev->phydev, ifr, cmd); -} - static void ftgmac100_tx_timeout(struct net_device *netdev, unsigned int txqueue) { struct ftgmac100 *priv = netdev_priv(netdev); @@ -1597,7 +1588,7 @@ static const struct net_device_ops ftgmac100_netdev_ops = { .ndo_start_xmit = ftgmac100_hard_start_xmit, .ndo_set_mac_address = ftgmac100_set_mac_addr, .ndo_validate_addr = eth_validate_addr, - .ndo_do_ioctl = ftgmac100_do_ioctl, + .ndo_do_ioctl = phy_do_ioctl, .ndo_tx_timeout = ftgmac100_tx_timeout, .ndo_set_rx_mode = ftgmac100_set_rx_mode, .ndo_set_features = ftgmac100_set_features, diff --git a/drivers/net/ethernet/freescale/Makefile b/drivers/net/ethernet/freescale/Makefile index 6a93293d31e0..67c436400352 100644 --- a/drivers/net/ethernet/freescale/Makefile +++ b/drivers/net/ethernet/freescale/Makefile @@ -25,4 +25,5 @@ obj-$(CONFIG_FSL_DPAA_ETH) += dpaa/ obj-$(CONFIG_FSL_DPAA2_ETH) += dpaa2/ obj-$(CONFIG_FSL_ENETC) += enetc/ +obj-$(CONFIG_FSL_ENETC_MDIO) += enetc/ obj-$(CONFIG_FSL_ENETC_VF) += enetc/ diff --git a/drivers/net/ethernet/freescale/enetc/enetc_qos.c b/drivers/net/ethernet/freescale/enetc/enetc_qos.c index 00382b7c5bd8..0c6bf3a55a9a 100644 --- a/drivers/net/ethernet/freescale/enetc/enetc_qos.c +++ b/drivers/net/ethernet/freescale/enetc/enetc_qos.c @@ -36,7 +36,6 @@ void enetc_sched_speed_set(struct net_device *ndev) case SPEED_10: default: pspeed = ENETC_PMR_PSPEED_10M; - netdev_err(ndev, "Qbv PSPEED set speed link down.\n"); } priv->speed = speed; diff --git a/drivers/net/ethernet/freescale/fec_main.c b/drivers/net/ethernet/freescale/fec_main.c index 798fed37be46..4432a59904c7 100644 --- a/drivers/net/ethernet/freescale/fec_main.c +++ b/drivers/net/ethernet/freescale/fec_main.c @@ -2199,8 +2199,14 @@ static void fec_enet_get_regs(struct net_device *ndev, { struct fec_enet_private *fep = netdev_priv(ndev); u32 __iomem *theregs = (u32 __iomem *)fep->hwp; + struct device *dev = &fep->pdev->dev; u32 *buf = (u32 *)regbuf; u32 i, off; + int ret; + + ret = pm_runtime_get_sync(dev); + if (ret < 0) + return; regs->version = fec_enet_register_version; @@ -2216,6 +2222,9 @@ static void fec_enet_get_regs(struct net_device *ndev, off >>= 2; buf[off] = readl(&theregs[off]); } + + pm_runtime_mark_last_busy(dev); + pm_runtime_put_autosuspend(dev); } static int fec_enet_get_ts_info(struct net_device *ndev, diff --git a/drivers/net/ethernet/freescale/fec_mpc52xx.c b/drivers/net/ethernet/freescale/fec_mpc52xx.c index de5278485062..7a3f066e611d 100644 --- a/drivers/net/ethernet/freescale/fec_mpc52xx.c +++ b/drivers/net/ethernet/freescale/fec_mpc52xx.c @@ -785,16 +785,6 @@ static const struct ethtool_ops mpc52xx_fec_ethtool_ops = { }; -static int mpc52xx_fec_ioctl(struct net_device *dev, struct ifreq *rq, int cmd) -{ - struct phy_device *phydev = dev->phydev; - - if (!phydev) - return -ENOTSUPP; - - return phy_mii_ioctl(phydev, rq, cmd); -} - static const struct net_device_ops mpc52xx_fec_netdev_ops = { .ndo_open = mpc52xx_fec_open, .ndo_stop = mpc52xx_fec_close, @@ -802,7 +792,7 @@ static const struct net_device_ops mpc52xx_fec_netdev_ops = { .ndo_set_rx_mode = mpc52xx_fec_set_multicast_list, .ndo_set_mac_address = mpc52xx_fec_set_mac_address, .ndo_validate_addr = eth_validate_addr, - .ndo_do_ioctl = mpc52xx_fec_ioctl, + .ndo_do_ioctl = phy_do_ioctl, .ndo_tx_timeout = mpc52xx_fec_tx_timeout, .ndo_get_stats = mpc52xx_fec_get_stats, #ifdef CONFIG_NET_POLL_CONTROLLER diff --git a/drivers/net/ethernet/google/gve/gve_rx.c b/drivers/net/ethernet/google/gve/gve_rx.c index edec61dfc868..9f52e72ff641 100644 --- a/drivers/net/ethernet/google/gve/gve_rx.c +++ b/drivers/net/ethernet/google/gve/gve_rx.c @@ -418,8 +418,6 @@ bool gve_clean_rx_done(struct gve_rx_ring *rx, int budget, rx->cnt = cnt; rx->fill_cnt += work_done; - /* restock desc ring slots */ - dma_wmb(); /* Ensure descs are visible before ringing doorbell */ gve_rx_write_doorbell(priv, rx); return gve_rx_work_pending(rx); } diff --git a/drivers/net/ethernet/google/gve/gve_tx.c b/drivers/net/ethernet/google/gve/gve_tx.c index f4889431f9b7..d0244feb0301 100644 --- a/drivers/net/ethernet/google/gve/gve_tx.c +++ b/drivers/net/ethernet/google/gve/gve_tx.c @@ -487,10 +487,6 @@ netdev_tx_t gve_tx(struct sk_buff *skb, struct net_device *dev) * may have added descriptors without ringing the doorbell. */ - /* Ensure tx descs from a prior gve_tx are visible before - * ringing doorbell. - */ - dma_wmb(); gve_tx_put_doorbell(priv, tx->q_resources, tx->req); return NETDEV_TX_BUSY; } @@ -505,8 +501,6 @@ netdev_tx_t gve_tx(struct sk_buff *skb, struct net_device *dev) if (!netif_xmit_stopped(tx->netdev_txq) && netdev_xmit_more()) return NETDEV_TX_OK; - /* Ensure tx descs are visible before ringing doorbell */ - dma_wmb(); gve_tx_put_doorbell(priv, tx->q_resources, tx->req); return NETDEV_TX_OK; } diff --git a/drivers/net/ethernet/hisilicon/hns/hns_enet.c b/drivers/net/ethernet/hisilicon/hns/hns_enet.c index e45553ec114a..2fbb476584bc 100644 --- a/drivers/net/ethernet/hisilicon/hns/hns_enet.c +++ b/drivers/net/ethernet/hisilicon/hns/hns_enet.c @@ -565,7 +565,6 @@ static int hns_nic_poll_rx_skb(struct hns_nic_ring_data *ring_data, skb = *out_skb = napi_alloc_skb(&ring_data->napi, HNS_RX_HEAD_SIZE); if (unlikely(!skb)) { - netdev_err(ndev, "alloc rx skb fail\n"); ring->stats.sw_err_cnt++; return -ENOMEM; } @@ -1056,7 +1055,6 @@ static int hns_nic_common_poll(struct napi_struct *napi, int budget) container_of(napi, struct hns_nic_ring_data, napi); struct hnae_ring *ring = ring_data->ring; -try_again: clean_complete += ring_data->poll_one( ring_data, budget - clean_complete, ring_data->ex_process); @@ -1066,7 +1064,7 @@ try_again: napi_complete(napi); ring->q->handle->dev->ops->toggle_ring_irq(ring, 0); } else { - goto try_again; + return budget; } } diff --git a/drivers/net/ethernet/hisilicon/hns3/hnae3.h b/drivers/net/ethernet/hisilicon/hns3/hnae3.h index 6b131ab36c7e..a3e4081b84ba 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hnae3.h +++ b/drivers/net/ethernet/hisilicon/hns3/hnae3.h @@ -167,11 +167,6 @@ enum hnae3_reset_type { HNAE3_MAX_RESET, }; -enum hnae3_flr_state { - HNAE3_FLR_DOWN, - HNAE3_FLR_DONE, -}; - enum hnae3_port_base_vlan_state { HNAE3_PORT_BASE_VLAN_DISABLE, HNAE3_PORT_BASE_VLAN_ENABLE, diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_debugfs.c b/drivers/net/ethernet/hisilicon/hns3/hns3_debugfs.c index 6b328a259efc..1d4ffc5f408a 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3_debugfs.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3_debugfs.c @@ -176,7 +176,7 @@ static int hns3_dbg_bd_info(struct hnae3_handle *h, const char *cmd_buf) return -EINVAL; } - ring = &priv->ring[q_num]; + ring = &priv->ring[q_num]; value = readl_relaxed(ring->tqp->io_base + HNS3_RING_TX_RING_TAIL_REG); tx_index = (cnt == 1) ? value : tx_index; @@ -209,10 +209,10 @@ static int hns3_dbg_bd_info(struct hnae3_handle *h, const char *cmd_buf) le16_to_cpu(tx_desc->tx.bdtp_fe_sc_vld_ra_ri)); dev_info(dev, "(TX)mss: %u\n", le16_to_cpu(tx_desc->tx.mss)); - ring = &priv->ring[q_num + h->kinfo.num_tqps]; + ring = &priv->ring[q_num + h->kinfo.num_tqps]; value = readl_relaxed(ring->tqp->io_base + HNS3_RING_RX_RING_TAIL_REG); rx_index = (cnt == 1) ? value : tx_index; - rx_desc = &ring->desc[rx_index]; + rx_desc = &ring->desc[rx_index]; addr = le64_to_cpu(rx_desc->addr); dev_info(dev, "RX Queue Num: %u, BD Index: %u\n", q_num, rx_index); @@ -297,8 +297,8 @@ static ssize_t hns3_dbg_cmd_read(struct file *filp, char __user *buffer, if (!buf) return -ENOMEM; - len = snprintf(buf, HNS3_DBG_READ_LEN, "%s\n", - "Please echo help to cmd to get help information"); + len = scnprintf(buf, HNS3_DBG_READ_LEN, "%s\n", + "Please echo help to cmd to get help information"); uncopy_bytes = copy_to_user(buffer, buf, len); kfree(buf); diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c index e240d99f7ca8..acb796cc10d0 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c @@ -60,6 +60,8 @@ MODULE_PARM_DESC(debug, " Network interface message level setting"); #define HNS3_INNER_VLAN_TAG 1 #define HNS3_OUTER_VLAN_TAG 2 +#define HNS3_MIN_TX_LEN 33U + /* hns3_pci_tbl - PCI Device ID Table * * Last entry must be all 0s @@ -1409,6 +1411,10 @@ netdev_tx_t hns3_nic_net_xmit(struct sk_buff *skb, struct net_device *netdev) int bd_num = 0; int ret; + /* Hardware can only handle short frames above 32 bytes */ + if (skb_put_padto(skb, HNS3_MIN_TX_LEN)) + return NETDEV_TX_OK; + /* Prefetch the data used later */ prefetch(skb->data); @@ -2088,10 +2094,8 @@ static int hns3_probe(struct pci_dev *pdev, const struct pci_device_id *ent) int ret; ae_dev = devm_kzalloc(&pdev->dev, sizeof(*ae_dev), GFP_KERNEL); - if (!ae_dev) { - ret = -ENOMEM; - return ret; - } + if (!ae_dev) + return -ENOMEM; ae_dev->pdev = pdev; ae_dev->flag = ent->driver_data; @@ -2534,8 +2538,8 @@ void hns3_clean_tx_ring(struct hns3_enet_ring *ring) rmb(); /* Make sure head is ready before touch any data */ if (unlikely(!is_valid_clean_head(ring, head))) { - netdev_err(netdev, "wrong head (%d, %d-%d)\n", head, - ring->next_to_use, ring->next_to_clean); + hns3_rl_err(netdev, "wrong head (%d, %d-%d)\n", head, + ring->next_to_use, ring->next_to_clean); u64_stats_update_begin(&ring->syncp); ring->stats.io_err_cnt++; @@ -2621,6 +2625,12 @@ static void hns3_nic_alloc_rx_buffers(struct hns3_enet_ring *ring, writel_relaxed(i, ring->tqp->io_base + HNS3_RING_RX_RING_HEAD_REG); } +static bool hns3_page_is_reusable(struct page *page) +{ + return page_to_nid(page) == numa_mem_id() && + !page_is_pfmemalloc(page); +} + static void hns3_nic_reuse_page(struct sk_buff *skb, int i, struct hns3_enet_ring *ring, int pull_len, struct hns3_desc_cb *desc_cb) @@ -2635,7 +2645,7 @@ static void hns3_nic_reuse_page(struct sk_buff *skb, int i, /* Avoid re-using remote pages, or the stack is still using the page * when page_offset rollback to zero, flag default unreuse */ - if (unlikely(page_to_nid(desc_cb->priv) != numa_mem_id()) || + if (unlikely(!hns3_page_is_reusable(desc_cb->priv)) || (!desc_cb->page_offset && page_count(desc_cb->priv) > 1)) return; @@ -2854,7 +2864,7 @@ static int hns3_alloc_skb(struct hns3_enet_ring *ring, unsigned int length, memcpy(__skb_put(skb, length), va, ALIGN(length, sizeof(long))); /* We can reuse buffer as-is, just make sure it is local */ - if (likely(page_to_nid(desc_cb->priv) == numa_mem_id())) + if (likely(hns3_page_is_reusable(desc_cb->priv))) desc_cb->reuse_flag = 1; else /* This page cannot be reused so discard it */ put_page(desc_cb->priv); @@ -4704,7 +4714,7 @@ static int __init hns3_init_module(void) pr_info("%s: %s\n", hns3_driver_name, hns3_copyright); client.type = HNAE3_CLIENT_KNIC; - snprintf(client.name, HNAE3_CLIENT_NAME_LENGTH - 1, "%s", + snprintf(client.name, HNAE3_CLIENT_NAME_LENGTH, "%s", hns3_driver_name); client.ops = &client_ops; diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c b/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c index 6e0212b79438..c03856e63320 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c @@ -423,9 +423,8 @@ static void *hns3_update_strings(u8 *data, const struct hns3_stats *stats, data[ETH_GSTRING_LEN - 1] = '\0'; /* first, prepend the prefix string */ - n1 = snprintf(data, MAX_PREFIX_SIZE, "%s%d_", - prefix, i); - n1 = min_t(uint, n1, MAX_PREFIX_SIZE - 1); + n1 = scnprintf(data, MAX_PREFIX_SIZE, "%s%d_", + prefix, i); size_left = (ETH_GSTRING_LEN - 1) - n1; /* now, concatenate the stats string to it */ diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_debugfs.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_debugfs.c index f3d4cbd28913..67fad80035d3 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_debugfs.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_debugfs.c @@ -73,8 +73,6 @@ static struct hclge_dbg_reg_type_info hclge_dbg_reg_info[] = { static int hclge_dbg_get_dfx_bd_num(struct hclge_dev *hdev, int offset) { -#define HCLGE_GET_DFX_REG_TYPE_CNT 4 - struct hclge_desc desc[HCLGE_GET_DFX_REG_TYPE_CNT]; int entries_per_desc; int index; diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_err.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_err.c index f8127d72d645..c85b72dc44d2 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_err.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_err.c @@ -1898,10 +1898,8 @@ static int hclge_handle_all_hw_msix_error(struct hclge_dev *hdev, bd_num = max_t(u32, mpf_bd_num, pf_bd_num); desc = kcalloc(bd_num, sizeof(struct hclge_desc), GFP_KERNEL); - if (!desc) { - ret = -ENOMEM; - goto out; - } + if (!desc) + return -ENOMEM; ret = hclge_handle_mpf_msix_error(hdev, desc, mpf_bd_num, reset_requests); diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c index 8bddda7f9092..ec5f6eeb639b 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c @@ -862,9 +862,7 @@ static int hclge_query_function_status(struct hclge_dev *hdev) usleep_range(1000, 2000); } while (timeout++ < HCLGE_QUERY_MAX_CNT); - ret = hclge_parse_func_status(hdev, req); - - return ret; + return hclge_parse_func_status(hdev, req); } static int hclge_query_pf_resource(struct hclge_dev *hdev) @@ -882,12 +880,12 @@ static int hclge_query_pf_resource(struct hclge_dev *hdev) } req = (struct hclge_pf_res_cmd *)desc.data; - hdev->num_tqps = __le16_to_cpu(req->tqp_num); - hdev->pkt_buf_size = __le16_to_cpu(req->buf_size) << HCLGE_BUF_UNIT_S; + hdev->num_tqps = le16_to_cpu(req->tqp_num); + hdev->pkt_buf_size = le16_to_cpu(req->buf_size) << HCLGE_BUF_UNIT_S; if (req->tx_buf_size) hdev->tx_buf_size = - __le16_to_cpu(req->tx_buf_size) << HCLGE_BUF_UNIT_S; + le16_to_cpu(req->tx_buf_size) << HCLGE_BUF_UNIT_S; else hdev->tx_buf_size = HCLGE_DEFAULT_TX_BUF; @@ -895,7 +893,7 @@ static int hclge_query_pf_resource(struct hclge_dev *hdev) if (req->dv_buf_size) hdev->dv_buf_size = - __le16_to_cpu(req->dv_buf_size) << HCLGE_BUF_UNIT_S; + le16_to_cpu(req->dv_buf_size) << HCLGE_BUF_UNIT_S; else hdev->dv_buf_size = HCLGE_DEFAULT_DV; @@ -903,10 +901,10 @@ static int hclge_query_pf_resource(struct hclge_dev *hdev) if (hnae3_dev_roce_supported(hdev)) { hdev->roce_base_msix_offset = - hnae3_get_field(__le16_to_cpu(req->msixcap_localid_ba_rocee), + hnae3_get_field(le16_to_cpu(req->msixcap_localid_ba_rocee), HCLGE_MSIX_OFT_ROCEE_M, HCLGE_MSIX_OFT_ROCEE_S); hdev->num_roce_msi = - hnae3_get_field(__le16_to_cpu(req->pf_intr_vector_number), + hnae3_get_field(le16_to_cpu(req->pf_intr_vector_number), HCLGE_PF_VEC_NUM_M, HCLGE_PF_VEC_NUM_S); /* nic's msix numbers is always equals to the roce's. */ @@ -919,7 +917,7 @@ static int hclge_query_pf_resource(struct hclge_dev *hdev) hdev->roce_base_msix_offset; } else { hdev->num_msi = - hnae3_get_field(__le16_to_cpu(req->pf_intr_vector_number), + hnae3_get_field(le16_to_cpu(req->pf_intr_vector_number), HCLGE_PF_VEC_NUM_M, HCLGE_PF_VEC_NUM_S); hdev->num_nic_msi = hdev->num_msi; @@ -1333,11 +1331,7 @@ static int hclge_get_cap(struct hclge_dev *hdev) } /* get pf resource */ - ret = hclge_query_pf_resource(hdev); - if (ret) - dev_err(&hdev->pdev->dev, "query pf resource error %d.\n", ret); - - return ret; + return hclge_query_pf_resource(hdev); } static void hclge_init_kdump_kernel_config(struct hclge_dev *hdev) @@ -2621,30 +2615,21 @@ static int hclge_mac_init(struct hclge_dev *hdev) hdev->hw.mac.duplex = HCLGE_MAC_FULL; ret = hclge_cfg_mac_speed_dup_hw(hdev, hdev->hw.mac.speed, hdev->hw.mac.duplex); - if (ret) { - dev_err(&hdev->pdev->dev, - "Config mac speed dup fail ret=%d\n", ret); + if (ret) return ret; - } if (hdev->hw.mac.support_autoneg) { ret = hclge_set_autoneg_en(hdev, hdev->hw.mac.autoneg); - if (ret) { - dev_err(&hdev->pdev->dev, - "Config mac autoneg fail ret=%d\n", ret); + if (ret) return ret; - } } mac->link = 0; if (mac->user_fec_mode & BIT(HNAE3_FEC_USER_DEF)) { ret = hclge_set_fec_hw(hdev, mac->user_fec_mode); - if (ret) { - dev_err(&hdev->pdev->dev, - "Fec mode init fail, ret = %d\n", ret); + if (ret) return ret; - } } ret = hclge_set_mac_mtu(hdev, hdev->mps); @@ -2916,7 +2901,7 @@ static int hclge_get_status(struct hnae3_handle *handle) static struct hclge_vport *hclge_get_vf_vport(struct hclge_dev *hdev, int vf) { - if (pci_num_vf(hdev->pdev) == 0) { + if (!pci_num_vf(hdev->pdev)) { dev_err(&hdev->pdev->dev, "SRIOV is disabled, can not get vport(%d) info.\n", vf); return NULL; @@ -3260,7 +3245,8 @@ static int hclge_notify_roce_client(struct hclge_dev *hdev, static int hclge_reset_wait(struct hclge_dev *hdev) { #define HCLGE_RESET_WATI_MS 100 -#define HCLGE_RESET_WAIT_CNT 200 +#define HCLGE_RESET_WAIT_CNT 350 + u32 val, reg, reg_bit; u32 cnt = 0; @@ -3277,8 +3263,6 @@ static int hclge_reset_wait(struct hclge_dev *hdev) reg = HCLGE_FUN_RST_ING; reg_bit = HCLGE_FUN_RST_ING_B; break; - case HNAE3_FLR_RESET: - break; default: dev_err(&hdev->pdev->dev, "Wait for unsupported reset type: %d\n", @@ -3286,20 +3270,6 @@ static int hclge_reset_wait(struct hclge_dev *hdev) return -EINVAL; } - if (hdev->reset_type == HNAE3_FLR_RESET) { - while (!test_bit(HNAE3_FLR_DONE, &hdev->flr_state) && - cnt++ < HCLGE_RESET_WAIT_CNT) - msleep(HCLGE_RESET_WATI_MS); - - if (!test_bit(HNAE3_FLR_DONE, &hdev->flr_state)) { - dev_err(&hdev->pdev->dev, - "flr wait timeout: %u\n", cnt); - return -EBUSY; - } - - return 0; - } - val = hclge_read_dev(&hdev->hw, reg); while (hnae3_get_bit(val, reg_bit) && cnt < HCLGE_RESET_WAIT_CNT) { msleep(HCLGE_RESET_WATI_MS); @@ -3377,7 +3347,7 @@ static void hclge_mailbox_service_task(struct hclge_dev *hdev) clear_bit(HCLGE_STATE_MBX_HANDLING, &hdev->state); } -static int hclge_func_reset_sync_vf(struct hclge_dev *hdev) +static void hclge_func_reset_sync_vf(struct hclge_dev *hdev) { struct hclge_pf_rst_sync_cmd *req; struct hclge_desc desc; @@ -3397,20 +3367,19 @@ static int hclge_func_reset_sync_vf(struct hclge_dev *hdev) */ if (ret == -EOPNOTSUPP) { msleep(HCLGE_RESET_SYNC_TIME); - return 0; + return; } else if (ret) { - dev_err(&hdev->pdev->dev, "sync with VF fail %d!\n", - ret); - return ret; + dev_warn(&hdev->pdev->dev, "sync with VF fail %d!\n", + ret); + return; } else if (req->all_vf_ready) { - return 0; + return; } msleep(HCLGE_PF_RESET_SYNC_TIME); hclge_cmd_reuse_desc(&desc, true); } while (cnt++ < HCLGE_PF_RESET_SYNC_CNT); - dev_err(&hdev->pdev->dev, "sync with VF timeout!\n"); - return -ETIME; + dev_warn(&hdev->pdev->dev, "sync with VF timeout!\n"); } void hclge_report_hw_error(struct hclge_dev *hdev, @@ -3490,12 +3459,6 @@ static void hclge_do_reset(struct hclge_dev *hdev) set_bit(HNAE3_FUNC_RESET, &hdev->reset_pending); hclge_reset_task_schedule(hdev); break; - case HNAE3_FLR_RESET: - dev_info(&pdev->dev, "FLR requested\n"); - /* schedule again to check later */ - set_bit(HNAE3_FLR_RESET, &hdev->reset_pending); - hclge_reset_task_schedule(hdev); - break; default: dev_warn(&pdev->dev, "Unsupported reset type: %d\n", hdev->reset_type); @@ -3584,23 +3547,6 @@ static void hclge_clear_reset_cause(struct hclge_dev *hdev) hclge_enable_vector(&hdev->misc_vector, true); } -static int hclge_reset_prepare_down(struct hclge_dev *hdev) -{ - int ret = 0; - - switch (hdev->reset_type) { - case HNAE3_FUNC_RESET: - /* fall through */ - case HNAE3_FLR_RESET: - ret = hclge_set_all_vf_rst(hdev, true); - break; - default: - break; - } - - return ret; -} - static void hclge_reset_handshake(struct hclge_dev *hdev, bool enable) { u32 reg_val; @@ -3614,6 +3560,19 @@ static void hclge_reset_handshake(struct hclge_dev *hdev, bool enable) hclge_write_dev(&hdev->hw, HCLGE_NIC_CSQ_DEPTH_REG, reg_val); } +static int hclge_func_reset_notify_vf(struct hclge_dev *hdev) +{ + int ret; + + ret = hclge_set_all_vf_rst(hdev, true); + if (ret) + return ret; + + hclge_func_reset_sync_vf(hdev); + + return 0; +} + static int hclge_reset_prepare_wait(struct hclge_dev *hdev) { u32 reg_val; @@ -3621,10 +3580,7 @@ static int hclge_reset_prepare_wait(struct hclge_dev *hdev) switch (hdev->reset_type) { case HNAE3_FUNC_RESET: - /* to confirm whether all running VF is ready - * before request PF reset - */ - ret = hclge_func_reset_sync_vf(hdev); + ret = hclge_func_reset_notify_vf(hdev); if (ret) return ret; @@ -3644,16 +3600,9 @@ static int hclge_reset_prepare_wait(struct hclge_dev *hdev) hdev->rst_stats.pf_rst_cnt++; break; case HNAE3_FLR_RESET: - /* to confirm whether all running VF is ready - * before request PF reset - */ - ret = hclge_func_reset_sync_vf(hdev); + ret = hclge_func_reset_notify_vf(hdev); if (ret) return ret; - - set_bit(HCLGE_STATE_CMD_DISABLE, &hdev->state); - set_bit(HNAE3_FLR_DOWN, &hdev->flr_state); - hdev->rst_stats.flr_rst_cnt++; break; case HNAE3_IMP_RESET: hclge_handle_imp_error(hdev); @@ -3782,10 +3731,9 @@ static int hclge_reset_stack(struct hclge_dev *hdev) return hclge_notify_client(hdev, HNAE3_RESTORE_CLIENT); } -static void hclge_reset(struct hclge_dev *hdev) +static int hclge_reset_prepare(struct hclge_dev *hdev) { struct hnae3_ae_dev *ae_dev = pci_get_drvdata(hdev->pdev); - enum hnae3_reset_type reset_level; int ret; /* Initialize ae_dev reset status as well, in case enet layer wants to @@ -3796,45 +3744,41 @@ static void hclge_reset(struct hclge_dev *hdev) /* perform reset of the stack & ae device for a client */ ret = hclge_notify_roce_client(hdev, HNAE3_DOWN_CLIENT); if (ret) - goto err_reset; - - ret = hclge_reset_prepare_down(hdev); - if (ret) - goto err_reset; + return ret; rtnl_lock(); ret = hclge_notify_client(hdev, HNAE3_DOWN_CLIENT); - if (ret) - goto err_reset_lock; - rtnl_unlock(); - - ret = hclge_reset_prepare_wait(hdev); if (ret) - goto err_reset; + return ret; - if (hclge_reset_wait(hdev)) - goto err_reset; + return hclge_reset_prepare_wait(hdev); +} + +static int hclge_reset_rebuild(struct hclge_dev *hdev) +{ + struct hnae3_ae_dev *ae_dev = pci_get_drvdata(hdev->pdev); + enum hnae3_reset_type reset_level; + int ret; hdev->rst_stats.hw_reset_done_cnt++; ret = hclge_notify_roce_client(hdev, HNAE3_UNINIT_CLIENT); if (ret) - goto err_reset; + return ret; rtnl_lock(); - ret = hclge_reset_stack(hdev); + rtnl_unlock(); if (ret) - goto err_reset_lock; + return ret; hclge_clear_reset_cause(hdev); ret = hclge_reset_prepare_up(hdev); if (ret) - goto err_reset_lock; + return ret; - rtnl_unlock(); ret = hclge_notify_roce_client(hdev, HNAE3_INIT_CLIENT); /* ignore RoCE notify error if it fails HCLGE_RESET_MAX_FAIL_CNT - 1 @@ -3842,19 +3786,17 @@ static void hclge_reset(struct hclge_dev *hdev) */ if (ret && hdev->rst_stats.reset_fail_cnt < HCLGE_RESET_MAX_FAIL_CNT - 1) - goto err_reset; + return ret; rtnl_lock(); - ret = hclge_notify_client(hdev, HNAE3_UP_CLIENT); - if (ret) - goto err_reset_lock; - rtnl_unlock(); + if (ret) + return ret; ret = hclge_notify_roce_client(hdev, HNAE3_UP_CLIENT); if (ret) - goto err_reset; + return ret; hdev->last_reset_time = jiffies; hdev->rst_stats.reset_fail_cnt = 0; @@ -3871,10 +3813,22 @@ static void hclge_reset(struct hclge_dev *hdev) if (reset_level != HNAE3_NONE_RESET) set_bit(reset_level, &hdev->reset_request); + return 0; +} + +static void hclge_reset(struct hclge_dev *hdev) +{ + if (hclge_reset_prepare(hdev)) + goto err_reset; + + if (hclge_reset_wait(hdev)) + goto err_reset; + + if (hclge_reset_rebuild(hdev)) + goto err_reset; + return; -err_reset_lock: - rtnl_unlock(); err_reset: if (hclge_reset_err_handle(hdev)) hclge_reset_task_schedule(hdev); @@ -3980,12 +3934,13 @@ static void hclge_reset_service_task(struct hclge_dev *hdev) if (!test_and_clear_bit(HCLGE_STATE_RST_SERVICE_SCHED, &hdev->state)) return; - if (test_and_set_bit(HCLGE_STATE_RST_HANDLING, &hdev->state)) - return; + down(&hdev->reset_sem); + set_bit(HCLGE_STATE_RST_HANDLING, &hdev->state); hclge_reset_subtask(hdev); clear_bit(HCLGE_STATE_RST_HANDLING, &hdev->state); + up(&hdev->reset_sem); } static void hclge_update_vport_alive(struct hclge_dev *hdev) @@ -4131,7 +4086,7 @@ static int hclge_put_vector(struct hnae3_handle *handle, int vector) vector_id = hclge_get_vector_index(hdev, vector); if (vector_id < 0) { dev_err(&hdev->pdev->dev, - "Get vector index fail. vector_id =%d\n", vector_id); + "Get vector index fail. vector = %d\n", vector); return vector_id; } @@ -6614,7 +6569,7 @@ static int hclge_set_serdes_loopback(struct hclge_dev *hdev, bool en, hclge_cfg_mac_mode(hdev, en); - ret = hclge_mac_phy_link_status_wait(hdev, en, FALSE); + ret = hclge_mac_phy_link_status_wait(hdev, en, false); if (ret) dev_err(&hdev->pdev->dev, "serdes loopback config mac mode timeout\n"); @@ -6672,7 +6627,7 @@ static int hclge_set_phy_loopback(struct hclge_dev *hdev, bool en) hclge_cfg_mac_mode(hdev, en); - ret = hclge_mac_phy_link_status_wait(hdev, en, TRUE); + ret = hclge_mac_phy_link_status_wait(hdev, en, true); if (ret) dev_err(&hdev->pdev->dev, "phy loopback config mac mode timeout\n"); @@ -9332,30 +9287,53 @@ static void hclge_state_uninit(struct hclge_dev *hdev) static void hclge_flr_prepare(struct hnae3_ae_dev *ae_dev) { -#define HCLGE_FLR_WAIT_MS 100 -#define HCLGE_FLR_WAIT_CNT 50 - struct hclge_dev *hdev = ae_dev->priv; - int cnt = 0; +#define HCLGE_FLR_RETRY_WAIT_MS 500 +#define HCLGE_FLR_RETRY_CNT 5 - clear_bit(HNAE3_FLR_DOWN, &hdev->flr_state); - clear_bit(HNAE3_FLR_DONE, &hdev->flr_state); - set_bit(HNAE3_FLR_RESET, &hdev->default_reset_request); - hclge_reset_event(hdev->pdev, NULL); + struct hclge_dev *hdev = ae_dev->priv; + int retry_cnt = 0; + int ret; - while (!test_bit(HNAE3_FLR_DOWN, &hdev->flr_state) && - cnt++ < HCLGE_FLR_WAIT_CNT) - msleep(HCLGE_FLR_WAIT_MS); +retry: + down(&hdev->reset_sem); + set_bit(HCLGE_STATE_RST_HANDLING, &hdev->state); + hdev->reset_type = HNAE3_FLR_RESET; + ret = hclge_reset_prepare(hdev); + if (ret) { + dev_err(&hdev->pdev->dev, "fail to prepare FLR, ret=%d\n", + ret); + if (hdev->reset_pending || + retry_cnt++ < HCLGE_FLR_RETRY_CNT) { + dev_err(&hdev->pdev->dev, + "reset_pending:0x%lx, retry_cnt:%d\n", + hdev->reset_pending, retry_cnt); + clear_bit(HCLGE_STATE_RST_HANDLING, &hdev->state); + up(&hdev->reset_sem); + msleep(HCLGE_FLR_RETRY_WAIT_MS); + goto retry; + } + } - if (!test_bit(HNAE3_FLR_DOWN, &hdev->flr_state)) - dev_err(&hdev->pdev->dev, - "flr wait down timeout: %d\n", cnt); + /* disable misc vector before FLR done */ + hclge_enable_vector(&hdev->misc_vector, false); + set_bit(HCLGE_STATE_CMD_DISABLE, &hdev->state); + hdev->rst_stats.flr_rst_cnt++; } static void hclge_flr_done(struct hnae3_ae_dev *ae_dev) { struct hclge_dev *hdev = ae_dev->priv; + int ret; + + hclge_enable_vector(&hdev->misc_vector, true); - set_bit(HNAE3_FLR_DONE, &hdev->flr_state); + ret = hclge_reset_rebuild(hdev); + if (ret) + dev_err(&hdev->pdev->dev, "fail to rebuild, ret=%d\n", ret); + + hdev->reset_type = HNAE3_NONE_RESET; + clear_bit(HCLGE_STATE_RST_HANDLING, &hdev->state); + up(&hdev->reset_sem); } static void hclge_clear_resetting_state(struct hclge_dev *hdev) @@ -9398,19 +9376,16 @@ static int hclge_init_ae_dev(struct hnae3_ae_dev *ae_dev) mutex_init(&hdev->vport_lock); spin_lock_init(&hdev->fd_rule_lock); + sema_init(&hdev->reset_sem, 1); ret = hclge_pci_init(hdev); - if (ret) { - dev_err(&pdev->dev, "PCI init failed\n"); + if (ret) goto out; - } /* Firmware command queue initialize */ ret = hclge_cmd_queue_init(hdev); - if (ret) { - dev_err(&pdev->dev, "Cmd queue init failed, ret = %d.\n", ret); + if (ret) goto err_pci_uninit; - } /* Firmware command initialize */ ret = hclge_cmd_init(hdev); @@ -9418,11 +9393,8 @@ static int hclge_init_ae_dev(struct hnae3_ae_dev *ae_dev) goto err_cmd_uninit; ret = hclge_get_cap(hdev); - if (ret) { - dev_err(&pdev->dev, "get hw capability error, ret = %d.\n", - ret); + if (ret) goto err_cmd_uninit; - } ret = hclge_configure(hdev); if (ret) { @@ -9437,12 +9409,8 @@ static int hclge_init_ae_dev(struct hnae3_ae_dev *ae_dev) } ret = hclge_misc_irq_init(hdev); - if (ret) { - dev_err(&pdev->dev, - "Misc IRQ(vector0) init error, ret = %d.\n", - ret); + if (ret) goto err_msi_uninit; - } ret = hclge_alloc_tqps(hdev); if (ret) { @@ -9451,31 +9419,22 @@ static int hclge_init_ae_dev(struct hnae3_ae_dev *ae_dev) } ret = hclge_alloc_vport(hdev); - if (ret) { - dev_err(&pdev->dev, "Allocate vport error, ret = %d.\n", ret); + if (ret) goto err_msi_irq_uninit; - } ret = hclge_map_tqp(hdev); - if (ret) { - dev_err(&pdev->dev, "Map tqp error, ret = %d.\n", ret); + if (ret) goto err_msi_irq_uninit; - } if (hdev->hw.mac.media_type == HNAE3_MEDIA_TYPE_COPPER) { ret = hclge_mac_mdio_config(hdev); - if (ret) { - dev_err(&hdev->pdev->dev, - "mdio config fail ret=%d\n", ret); + if (ret) goto err_msi_irq_uninit; - } } ret = hclge_init_umv_space(hdev); - if (ret) { - dev_err(&pdev->dev, "umv space init error, ret=%d.\n", ret); + if (ret) goto err_mdiobus_unreg; - } ret = hclge_mac_init(hdev); if (ret) { @@ -10210,10 +10169,8 @@ static int hclge_get_dfx_reg_bd_num(struct hclge_dev *hdev, int *bd_num_list, u32 type_num) { -#define HCLGE_DFX_REG_BD_NUM 4 - u32 entries_per_desc, desc_index, index, offset, i; - struct hclge_desc desc[HCLGE_DFX_REG_BD_NUM]; + struct hclge_desc desc[HCLGE_GET_DFX_REG_TYPE_CNT]; int ret; ret = hclge_query_bd_num_cmd_send(hdev, desc); @@ -10326,10 +10283,8 @@ static int hclge_get_dfx_reg(struct hclge_dev *hdev, void *data) buf_len = sizeof(*desc_src) * bd_num_max; desc_src = kzalloc(buf_len, GFP_KERNEL); - if (!desc_src) { - dev_err(&hdev->pdev->dev, "%s kzalloc failed\n", __func__); + if (!desc_src) return -ENOMEM; - } for (i = 0; i < dfx_reg_type_num; i++) { bd_num = bd_num_list[i]; diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h index 4e5cfda16eb7..f78cbb4cc85e 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h +++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h @@ -139,6 +139,8 @@ #define HCLGE_PHY_MDIX_STATUS_B 6 #define HCLGE_PHY_SPEED_DUP_RESOLVE_B 11 +#define HCLGE_GET_DFX_REG_TYPE_CNT 4 + /* Factor used to calculate offset and bitmap of VF num */ #define HCLGE_VF_NUM_PER_CMD 64 @@ -720,6 +722,7 @@ struct hclge_dev { unsigned long reset_request; /* reset has been requested */ unsigned long reset_pending; /* client rst is pending to be served */ struct hclge_rst_stats rst_stats; + struct semaphore reset_sem; /* protect reset process */ u32 fw_version; u16 num_vmdq_vport; /* Num vmdq vport this PF has set up */ u16 num_tqps; /* Num task queue pairs of this PF */ diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c index cfa797ef9906..d6597206e692 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c @@ -1316,14 +1316,13 @@ static int hclgevf_set_vlan_filter(struct hnae3_handle *handle, msg_data[0] = is_kill; memcpy(&msg_data[1], &vlan_id, sizeof(vlan_id)); memcpy(&msg_data[3], &proto, sizeof(proto)); - ret = hclgevf_send_mbx_msg(hdev, HCLGE_MBX_SET_VLAN, - HCLGE_MBX_VLAN_FILTER, msg_data, - HCLGEVF_VLAN_MBX_MSG_LEN, true, NULL, 0); - /* when remove hw vlan filter failed, record the vlan id, * and try to remove it from hw later, to be consistence * with stack. */ + ret = hclgevf_send_mbx_msg(hdev, HCLGE_MBX_SET_VLAN, + HCLGE_MBX_VLAN_FILTER, msg_data, + HCLGEVF_VLAN_MBX_MSG_LEN, true, NULL, 0); if (is_kill && ret) set_bit(vlan_id, hdev->vlan_del_fail_bmap); @@ -1411,32 +1410,6 @@ static int hclgevf_notify_client(struct hclgevf_dev *hdev, return ret; } -static void hclgevf_flr_done(struct hnae3_ae_dev *ae_dev) -{ - struct hclgevf_dev *hdev = ae_dev->priv; - - set_bit(HNAE3_FLR_DONE, &hdev->flr_state); -} - -static int hclgevf_flr_poll_timeout(struct hclgevf_dev *hdev, - unsigned long delay_us, - unsigned long wait_cnt) -{ - unsigned long cnt = 0; - - while (!test_bit(HNAE3_FLR_DONE, &hdev->flr_state) && - cnt++ < wait_cnt) - usleep_range(delay_us, delay_us * 2); - - if (!test_bit(HNAE3_FLR_DONE, &hdev->flr_state)) { - dev_err(&hdev->pdev->dev, - "flr wait timeout\n"); - return -ETIMEDOUT; - } - - return 0; -} - static int hclgevf_reset_wait(struct hclgevf_dev *hdev) { #define HCLGEVF_RESET_WAIT_US 20000 @@ -1447,11 +1420,7 @@ static int hclgevf_reset_wait(struct hclgevf_dev *hdev) u32 val; int ret; - if (hdev->reset_type == HNAE3_FLR_RESET) - return hclgevf_flr_poll_timeout(hdev, - HCLGEVF_RESET_WAIT_US, - HCLGEVF_RESET_WAIT_CNT); - else if (hdev->reset_type == HNAE3_VF_RESET) + if (hdev->reset_type == HNAE3_VF_RESET) ret = readl_poll_timeout(hdev->hw.io_base + HCLGEVF_VF_RST_ING, val, !(val & HCLGEVF_VF_RST_ING_BIT), @@ -1523,7 +1492,8 @@ static int hclgevf_reset_stack(struct hclgevf_dev *hdev) /* clear handshake status with IMP */ hclgevf_reset_handshake(hdev, false); - return 0; + /* bring up the nic to enable TX/RX again */ + return hclgevf_notify_client(hdev, HNAE3_UP_CLIENT); } static int hclgevf_reset_prepare_wait(struct hclgevf_dev *hdev) @@ -1532,18 +1502,10 @@ static int hclgevf_reset_prepare_wait(struct hclgevf_dev *hdev) int ret = 0; - switch (hdev->reset_type) { - case HNAE3_VF_FUNC_RESET: + if (hdev->reset_type == HNAE3_VF_FUNC_RESET) { ret = hclgevf_send_mbx_msg(hdev, HCLGE_MBX_RESET, 0, NULL, 0, true, NULL, sizeof(u8)); hdev->rst_stats.vf_func_rst_cnt++; - break; - case HNAE3_FLR_RESET: - set_bit(HNAE3_FLR_DOWN, &hdev->flr_state); - hdev->rst_stats.flr_rst_cnt++; - break; - default: - break; } set_bit(HCLGEVF_STATE_CMD_DISABLE, &hdev->state); @@ -1603,7 +1565,7 @@ static void hclgevf_reset_err_handle(struct hclgevf_dev *hdev) } } -static int hclgevf_reset(struct hclgevf_dev *hdev) +static int hclgevf_reset_prepare(struct hclgevf_dev *hdev) { struct hnae3_ae_dev *ae_dev = pci_get_drvdata(hdev->pdev); int ret; @@ -1613,62 +1575,64 @@ static int hclgevf_reset(struct hclgevf_dev *hdev) */ ae_dev->reset_type = hdev->reset_type; hdev->rst_stats.rst_cnt++; - rtnl_lock(); + rtnl_lock(); /* bring down the nic to stop any ongoing TX/RX */ ret = hclgevf_notify_client(hdev, HNAE3_DOWN_CLIENT); - if (ret) - goto err_reset_lock; - rtnl_unlock(); - - ret = hclgevf_reset_prepare_wait(hdev); if (ret) - goto err_reset; + return ret; - /* check if VF could successfully fetch the hardware reset completion - * status from the hardware - */ - ret = hclgevf_reset_wait(hdev); - if (ret) { - /* can't do much in this situation, will disable VF */ - dev_err(&hdev->pdev->dev, - "VF failed(=%d) to fetch H/W reset completion status\n", - ret); - goto err_reset; - } + return hclgevf_reset_prepare_wait(hdev); +} + +static int hclgevf_reset_rebuild(struct hclgevf_dev *hdev) +{ + struct hnae3_ae_dev *ae_dev = pci_get_drvdata(hdev->pdev); + int ret; hdev->rst_stats.hw_rst_done_cnt++; rtnl_lock(); - /* now, re-initialize the nic client and ae device */ ret = hclgevf_reset_stack(hdev); + rtnl_unlock(); if (ret) { dev_err(&hdev->pdev->dev, "failed to reset VF stack\n"); - goto err_reset_lock; + return ret; } - /* bring up the nic to enable TX/RX again */ - ret = hclgevf_notify_client(hdev, HNAE3_UP_CLIENT); - if (ret) - goto err_reset_lock; - - rtnl_unlock(); - hdev->last_reset_time = jiffies; ae_dev->reset_type = HNAE3_NONE_RESET; hdev->rst_stats.rst_done_cnt++; hdev->rst_stats.rst_fail_cnt = 0; clear_bit(HCLGEVF_STATE_RST_FAIL, &hdev->state); - return ret; -err_reset_lock: - rtnl_unlock(); + return 0; +} + +static void hclgevf_reset(struct hclgevf_dev *hdev) +{ + if (hclgevf_reset_prepare(hdev)) + goto err_reset; + + /* check if VF could successfully fetch the hardware reset completion + * status from the hardware + */ + if (hclgevf_reset_wait(hdev)) { + /* can't do much in this situation, will disable VF */ + dev_err(&hdev->pdev->dev, + "failed to fetch H/W reset completion status\n"); + goto err_reset; + } + + if (hclgevf_reset_rebuild(hdev)) + goto err_reset; + + return; + err_reset: hclgevf_reset_err_handle(hdev); - - return ret; } static enum hnae3_reset_type hclgevf_get_reset_level(struct hclgevf_dev *hdev, @@ -1731,25 +1695,60 @@ static void hclgevf_set_def_reset_request(struct hnae3_ae_dev *ae_dev, set_bit(rst_type, &hdev->default_reset_request); } +static void hclgevf_enable_vector(struct hclgevf_misc_vector *vector, bool en) +{ + writel(en ? 1 : 0, vector->addr); +} + static void hclgevf_flr_prepare(struct hnae3_ae_dev *ae_dev) { -#define HCLGEVF_FLR_WAIT_MS 100 -#define HCLGEVF_FLR_WAIT_CNT 50 +#define HCLGEVF_FLR_RETRY_WAIT_MS 500 +#define HCLGEVF_FLR_RETRY_CNT 5 + struct hclgevf_dev *hdev = ae_dev->priv; - int cnt = 0; + int retry_cnt = 0; + int ret; + +retry: + down(&hdev->reset_sem); + set_bit(HCLGEVF_STATE_RST_HANDLING, &hdev->state); + hdev->reset_type = HNAE3_FLR_RESET; + ret = hclgevf_reset_prepare(hdev); + if (ret) { + dev_err(&hdev->pdev->dev, "fail to prepare FLR, ret=%d\n", + ret); + if (hdev->reset_pending || + retry_cnt++ < HCLGEVF_FLR_RETRY_CNT) { + dev_err(&hdev->pdev->dev, + "reset_pending:0x%lx, retry_cnt:%d\n", + hdev->reset_pending, retry_cnt); + clear_bit(HCLGEVF_STATE_RST_HANDLING, &hdev->state); + up(&hdev->reset_sem); + msleep(HCLGEVF_FLR_RETRY_WAIT_MS); + goto retry; + } + } - clear_bit(HNAE3_FLR_DOWN, &hdev->flr_state); - clear_bit(HNAE3_FLR_DONE, &hdev->flr_state); - set_bit(HNAE3_FLR_RESET, &hdev->default_reset_request); - hclgevf_reset_event(hdev->pdev, NULL); + /* disable misc vector before FLR done */ + hclgevf_enable_vector(&hdev->misc_vector, false); + hdev->rst_stats.flr_rst_cnt++; +} - while (!test_bit(HNAE3_FLR_DOWN, &hdev->flr_state) && - cnt++ < HCLGEVF_FLR_WAIT_CNT) - msleep(HCLGEVF_FLR_WAIT_MS); +static void hclgevf_flr_done(struct hnae3_ae_dev *ae_dev) +{ + struct hclgevf_dev *hdev = ae_dev->priv; + int ret; - if (!test_bit(HNAE3_FLR_DOWN, &hdev->flr_state)) - dev_err(&hdev->pdev->dev, - "flr wait down timeout: %d\n", cnt); + hclgevf_enable_vector(&hdev->misc_vector, true); + + ret = hclgevf_reset_rebuild(hdev); + if (ret) + dev_warn(&hdev->pdev->dev, "fail to rebuild, ret=%d\n", + ret); + + hdev->reset_type = HNAE3_NONE_RESET; + clear_bit(HCLGEVF_STATE_RST_HANDLING, &hdev->state); + up(&hdev->reset_sem); } static u32 hclgevf_get_fw_version(struct hnae3_handle *handle) @@ -1802,13 +1801,11 @@ static void hclgevf_reset_service_task(struct hclgevf_dev *hdev) { #define HCLGEVF_MAX_RESET_ATTEMPTS_CNT 3 - int ret; - if (!test_and_clear_bit(HCLGEVF_STATE_RST_SERVICE_SCHED, &hdev->state)) return; - if (test_and_set_bit(HCLGEVF_STATE_RST_HANDLING, &hdev->state)) - return; + down(&hdev->reset_sem); + set_bit(HCLGEVF_STATE_RST_HANDLING, &hdev->state); if (test_and_clear_bit(HCLGEVF_RESET_PENDING, &hdev->reset_state)) { @@ -1822,12 +1819,8 @@ static void hclgevf_reset_service_task(struct hclgevf_dev *hdev) hdev->last_reset_time = jiffies; while ((hdev->reset_type = hclgevf_get_reset_level(hdev, &hdev->reset_pending)) - != HNAE3_NONE_RESET) { - ret = hclgevf_reset(hdev); - if (ret) - dev_err(&hdev->pdev->dev, - "VF stack reset failed %d.\n", ret); - } + != HNAE3_NONE_RESET) + hclgevf_reset(hdev); } else if (test_and_clear_bit(HCLGEVF_RESET_REQUESTED, &hdev->reset_state)) { /* we could be here when either of below happens: @@ -1868,7 +1861,9 @@ static void hclgevf_reset_service_task(struct hclgevf_dev *hdev) hclgevf_reset_task_schedule(hdev); } + hdev->reset_type = HNAE3_NONE_RESET; clear_bit(HCLGEVF_STATE_RST_HANDLING, &hdev->state); + up(&hdev->reset_sem); } static void hclgevf_mailbox_service_task(struct hclgevf_dev *hdev) @@ -2012,11 +2007,6 @@ static enum hclgevf_evt_cause hclgevf_check_evt_cause(struct hclgevf_dev *hdev, return HCLGEVF_VECTOR0_EVENT_OTHER; } -static void hclgevf_enable_vector(struct hclgevf_misc_vector *vector, bool en) -{ - writel(en ? 1 : 0, vector->addr); -} - static irqreturn_t hclgevf_misc_irq_handle(int irq, void *data) { enum hclgevf_evt_cause event_cause; @@ -2291,6 +2281,7 @@ static void hclgevf_state_init(struct hclgevf_dev *hdev) INIT_DELAYED_WORK(&hdev->service_task, hclgevf_service_task); mutex_init(&hdev->mbx_resp.mbx_mutex); + sema_init(&hdev->reset_sem, 1); /* bring the device down */ set_bit(HCLGEVF_STATE_DOWN, &hdev->state); @@ -2606,11 +2597,11 @@ static int hclgevf_query_vf_resource(struct hclgevf_dev *hdev) if (hnae3_dev_roce_supported(hdev)) { hdev->roce_base_msix_offset = - hnae3_get_field(__le16_to_cpu(req->msixcap_localid_ba_rocee), + hnae3_get_field(le16_to_cpu(req->msixcap_localid_ba_rocee), HCLGEVF_MSIX_OFT_ROCEE_M, HCLGEVF_MSIX_OFT_ROCEE_S); hdev->num_roce_msix = - hnae3_get_field(__le16_to_cpu(req->vf_intr_vector_number), + hnae3_get_field(le16_to_cpu(req->vf_intr_vector_number), HCLGEVF_VEC_NUM_M, HCLGEVF_VEC_NUM_S); /* nic's msix numbers is always equals to the roce's. */ @@ -2623,7 +2614,7 @@ static int hclgevf_query_vf_resource(struct hclgevf_dev *hdev) hdev->roce_base_msix_offset; } else { hdev->num_msi = - hnae3_get_field(__le16_to_cpu(req->vf_intr_vector_number), + hnae3_get_field(le16_to_cpu(req->vf_intr_vector_number), HCLGEVF_VEC_NUM_M, HCLGEVF_VEC_NUM_S); hdev->num_nic_msix = hdev->num_msi; @@ -2720,16 +2711,12 @@ static int hclgevf_init_hdev(struct hclgevf_dev *hdev) int ret; ret = hclgevf_pci_init(hdev); - if (ret) { - dev_err(&pdev->dev, "PCI initialization failed\n"); + if (ret) return ret; - } ret = hclgevf_cmd_queue_init(hdev); - if (ret) { - dev_err(&pdev->dev, "Cmd queue init failed: %d\n", ret); + if (ret) goto err_cmd_queue_init; - } ret = hclgevf_cmd_init(hdev); if (ret) @@ -2737,11 +2724,8 @@ static int hclgevf_init_hdev(struct hclgevf_dev *hdev) /* Get vf resource */ ret = hclgevf_query_vf_resource(hdev); - if (ret) { - dev_err(&hdev->pdev->dev, - "Query vf status error, ret = %d.\n", ret); + if (ret) goto err_cmd_init; - } ret = hclgevf_init_msi(hdev); if (ret) { @@ -2751,13 +2735,11 @@ static int hclgevf_init_hdev(struct hclgevf_dev *hdev) hclgevf_state_init(hdev); hdev->reset_level = HNAE3_VF_FUNC_RESET; + hdev->reset_type = HNAE3_NONE_RESET; ret = hclgevf_misc_irq_init(hdev); - if (ret) { - dev_err(&pdev->dev, "failed(%d) to init Misc IRQ(vector0)\n", - ret); + if (ret) goto err_misc_irq_init; - } set_bit(HCLGEVF_STATE_IRQ_INITED, &hdev->state); @@ -2774,10 +2756,8 @@ static int hclgevf_init_hdev(struct hclgevf_dev *hdev) } ret = hclgevf_set_handle_info(hdev); - if (ret) { - dev_err(&pdev->dev, "failed(%d) to set handle info\n", ret); + if (ret) goto err_config; - } ret = hclgevf_config_gro(hdev, true); if (ret) diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.h b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.h index 2cbc7dfe2f91..fee8d97f323c 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.h +++ b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.h @@ -253,6 +253,7 @@ struct hclgevf_dev { unsigned long reset_state; /* requested, pending */ struct hclgevf_rst_stats rst_stats; u32 reset_attempts; + struct semaphore reset_sem; /* protect reset process */ u32 fw_version; u16 num_tqps; /* num task queue pairs of this PF */ diff --git a/drivers/net/ethernet/intel/e1000e/e1000.h b/drivers/net/ethernet/intel/e1000e/e1000.h index 6c51b1bad8c4..37a2314d3e6b 100644 --- a/drivers/net/ethernet/intel/e1000e/e1000.h +++ b/drivers/net/ethernet/intel/e1000e/e1000.h @@ -185,13 +185,12 @@ struct e1000_phy_regs { /* board specific private data structure */ struct e1000_adapter { + struct timer_list watchdog_timer; struct timer_list phy_info_timer; struct timer_list blink_timer; struct work_struct reset_task; - struct delayed_work watchdog_task; - - struct workqueue_struct *e1000_workqueue; + struct work_struct watchdog_task; const struct e1000_info *ei; diff --git a/drivers/net/ethernet/intel/e1000e/netdev.c b/drivers/net/ethernet/intel/e1000e/netdev.c index 8797913b2702..db4ea58bac82 100644 --- a/drivers/net/ethernet/intel/e1000e/netdev.c +++ b/drivers/net/ethernet/intel/e1000e/netdev.c @@ -1780,8 +1780,7 @@ static irqreturn_t e1000_intr_msi(int __always_unused irq, void *data) } /* guard against interrupt when we're going down */ if (!test_bit(__E1000_DOWN, &adapter->state)) - mod_delayed_work(adapter->e1000_workqueue, - &adapter->watchdog_task, HZ); + mod_timer(&adapter->watchdog_timer, jiffies + 1); } /* Reset on uncorrectable ECC error */ @@ -1861,8 +1860,7 @@ static irqreturn_t e1000_intr(int __always_unused irq, void *data) } /* guard against interrupt when we're going down */ if (!test_bit(__E1000_DOWN, &adapter->state)) - mod_delayed_work(adapter->e1000_workqueue, - &adapter->watchdog_task, HZ); + mod_timer(&adapter->watchdog_timer, jiffies + 1); } /* Reset on uncorrectable ECC error */ @@ -1907,8 +1905,7 @@ static irqreturn_t e1000_msix_other(int __always_unused irq, void *data) hw->mac.get_link_status = true; /* guard against interrupt when we're going down */ if (!test_bit(__E1000_DOWN, &adapter->state)) - mod_delayed_work(adapter->e1000_workqueue, - &adapter->watchdog_task, HZ); + mod_timer(&adapter->watchdog_timer, jiffies + 1); } if (!test_bit(__E1000_DOWN, &adapter->state)) @@ -4284,6 +4281,7 @@ void e1000e_down(struct e1000_adapter *adapter, bool reset) napi_synchronize(&adapter->napi); + del_timer_sync(&adapter->watchdog_timer); del_timer_sync(&adapter->phy_info_timer); spin_lock(&adapter->stats64_lock); @@ -5156,11 +5154,25 @@ static void e1000e_check_82574_phy_workaround(struct e1000_adapter *adapter) } } +/** + * e1000_watchdog - Timer Call-back + * @data: pointer to adapter cast into an unsigned long + **/ +static void e1000_watchdog(struct timer_list *t) +{ + struct e1000_adapter *adapter = from_timer(adapter, t, watchdog_timer); + + /* Do the rest outside of interrupt context */ + schedule_work(&adapter->watchdog_task); + + /* TODO: make this use queue_delayed_work() */ +} + static void e1000_watchdog_task(struct work_struct *work) { struct e1000_adapter *adapter = container_of(work, struct e1000_adapter, - watchdog_task.work); + watchdog_task); struct net_device *netdev = adapter->netdev; struct e1000_mac_info *mac = &adapter->hw.mac; struct e1000_phy_info *phy = &adapter->hw.phy; @@ -5408,9 +5420,8 @@ link_up: /* Reset the timer */ if (!test_bit(__E1000_DOWN, &adapter->state)) - queue_delayed_work(adapter->e1000_workqueue, - &adapter->watchdog_task, - round_jiffies(2 * HZ)); + mod_timer(&adapter->watchdog_timer, + round_jiffies(jiffies + 2 * HZ)); } #define E1000_TX_FLAGS_CSUM 0x00000001 @@ -7450,21 +7461,11 @@ static int e1000_probe(struct pci_dev *pdev, const struct pci_device_id *ent) goto err_eeprom; } - adapter->e1000_workqueue = alloc_workqueue("%s", WQ_MEM_RECLAIM, 0, - e1000e_driver_name); - - if (!adapter->e1000_workqueue) { - err = -ENOMEM; - goto err_workqueue; - } - - INIT_DELAYED_WORK(&adapter->watchdog_task, e1000_watchdog_task); - queue_delayed_work(adapter->e1000_workqueue, &adapter->watchdog_task, - 0); - + timer_setup(&adapter->watchdog_timer, e1000_watchdog, 0); timer_setup(&adapter->phy_info_timer, e1000_update_phy_info, 0); INIT_WORK(&adapter->reset_task, e1000_reset_task); + INIT_WORK(&adapter->watchdog_task, e1000_watchdog_task); INIT_WORK(&adapter->downshift_task, e1000e_downshift_workaround); INIT_WORK(&adapter->update_phy_task, e1000e_update_phy_task); INIT_WORK(&adapter->print_hang_task, e1000_print_hw_hang); @@ -7558,9 +7559,6 @@ static int e1000_probe(struct pci_dev *pdev, const struct pci_device_id *ent) return 0; err_register: - flush_workqueue(adapter->e1000_workqueue); - destroy_workqueue(adapter->e1000_workqueue); -err_workqueue: if (!(adapter->flags & FLAG_HAS_AMT)) e1000e_release_hw_control(adapter); err_eeprom: @@ -7605,17 +7603,15 @@ static void e1000_remove(struct pci_dev *pdev) * from being rescheduled. */ set_bit(__E1000_DOWN, &adapter->state); + del_timer_sync(&adapter->watchdog_timer); del_timer_sync(&adapter->phy_info_timer); cancel_work_sync(&adapter->reset_task); + cancel_work_sync(&adapter->watchdog_task); cancel_work_sync(&adapter->downshift_task); cancel_work_sync(&adapter->update_phy_task); cancel_work_sync(&adapter->print_hang_task); - cancel_delayed_work(&adapter->watchdog_task); - flush_workqueue(adapter->e1000_workqueue); - destroy_workqueue(adapter->e1000_workqueue); - if (adapter->flags & FLAG_HAS_HW_TIMESTAMP) { cancel_work_sync(&adapter->tx_hwtstamp_work); if (adapter->tx_hwtstamp_skb) { diff --git a/drivers/net/ethernet/intel/fm10k/fm10k_netdev.c b/drivers/net/ethernet/intel/fm10k/fm10k_netdev.c index ba2566e2123d..0637ccadee79 100644 --- a/drivers/net/ethernet/intel/fm10k/fm10k_netdev.c +++ b/drivers/net/ethernet/intel/fm10k/fm10k_netdev.c @@ -696,21 +696,24 @@ static netdev_tx_t fm10k_xmit_frame(struct sk_buff *skb, struct net_device *dev) /** * fm10k_tx_timeout - Respond to a Tx Hang * @netdev: network interface device structure + * @txqueue: the index of the Tx queue that timed out **/ static void fm10k_tx_timeout(struct net_device *netdev, unsigned int txqueue) { struct fm10k_intfc *interface = netdev_priv(netdev); + struct fm10k_ring *tx_ring; bool real_tx_hang = false; - int i; - -#define TX_TIMEO_LIMIT 16000 - for (i = 0; i < interface->num_tx_queues; i++) { - struct fm10k_ring *tx_ring = interface->tx_ring[i]; - if (check_for_tx_hang(tx_ring) && fm10k_check_tx_hang(tx_ring)) - real_tx_hang = true; + if (txqueue >= interface->num_tx_queues) { + WARN(1, "invalid Tx queue index %d", txqueue); + return; } + tx_ring = interface->tx_ring[txqueue]; + if (check_for_tx_hang(tx_ring) && fm10k_check_tx_hang(tx_ring)) + real_tx_hang = true; + +#define TX_TIMEO_LIMIT 16000 if (real_tx_hang) { fm10k_tx_timeout_reset(interface); } else { diff --git a/drivers/net/ethernet/intel/i40e/i40e_adminq.c b/drivers/net/ethernet/intel/i40e/i40e_adminq.c index 9f0a4e92a231..37514a75f928 100644 --- a/drivers/net/ethernet/intel/i40e/i40e_adminq.c +++ b/drivers/net/ethernet/intel/i40e/i40e_adminq.c @@ -536,6 +536,11 @@ static void i40e_set_hw_flags(struct i40e_hw *hw) (aq->api_maj_ver == 1 && aq->api_min_ver >= I40E_MINOR_VER_FW_LLDP_STOPPABLE_X722)) hw->flags |= I40E_HW_FLAG_FW_LLDP_STOPPABLE; + + if (aq->api_maj_ver > 1 || + (aq->api_maj_ver == 1 && + aq->api_min_ver >= I40E_MINOR_VER_GET_LINK_INFO_X722)) + hw->flags |= I40E_HW_FLAG_AQ_PHY_ACCESS_CAPABLE; /* fall through */ default: break; diff --git a/drivers/net/ethernet/intel/i40e/i40e_main.c b/drivers/net/ethernet/intel/i40e/i40e_main.c index 33912cf964eb..8c3e753bfb9d 100644 --- a/drivers/net/ethernet/intel/i40e/i40e_main.c +++ b/drivers/net/ethernet/intel/i40e/i40e_main.c @@ -307,37 +307,18 @@ static void i40e_tx_timeout(struct net_device *netdev, unsigned int txqueue) struct i40e_vsi *vsi = np->vsi; struct i40e_pf *pf = vsi->back; struct i40e_ring *tx_ring = NULL; - unsigned int i, hung_queue = 0; + unsigned int i; u32 head, val; pf->tx_timeout_count++; - /* find the stopped queue the same way the stack does */ - for (i = 0; i < netdev->num_tx_queues; i++) { - struct netdev_queue *q; - unsigned long trans_start; - - q = netdev_get_tx_queue(netdev, i); - trans_start = q->trans_start; - if (netif_xmit_stopped(q) && - time_after(jiffies, - (trans_start + netdev->watchdog_timeo))) { - hung_queue = i; - break; - } - } - - if (i == netdev->num_tx_queues) { - netdev_info(netdev, "tx_timeout: no netdev hung queue found\n"); - } else { - /* now that we have an index, find the tx_ring struct */ - for (i = 0; i < vsi->num_queue_pairs; i++) { - if (vsi->tx_rings[i] && vsi->tx_rings[i]->desc) { - if (hung_queue == - vsi->tx_rings[i]->queue_index) { - tx_ring = vsi->tx_rings[i]; - break; - } + /* with txqueue index, find the tx_ring struct */ + for (i = 0; i < vsi->num_queue_pairs; i++) { + if (vsi->tx_rings[i] && vsi->tx_rings[i]->desc) { + if (txqueue == + vsi->tx_rings[i]->queue_index) { + tx_ring = vsi->tx_rings[i]; + break; } } } @@ -363,14 +344,14 @@ static void i40e_tx_timeout(struct net_device *netdev, unsigned int txqueue) val = rd32(&pf->hw, I40E_PFINT_DYN_CTL0); netdev_info(netdev, "tx_timeout: VSI_seid: %d, Q %d, NTC: 0x%x, HWB: 0x%x, NTU: 0x%x, TAIL: 0x%x, INT: 0x%x\n", - vsi->seid, hung_queue, tx_ring->next_to_clean, + vsi->seid, txqueue, tx_ring->next_to_clean, head, tx_ring->next_to_use, readl(tx_ring->tail), val); } pf->tx_timeout_last_recovery = jiffies; - netdev_info(netdev, "tx_timeout recovery level %d, hung_queue %d\n", - pf->tx_timeout_recovery_level, hung_queue); + netdev_info(netdev, "tx_timeout recovery level %d, txqueue %d\n", + pf->tx_timeout_recovery_level, txqueue); switch (pf->tx_timeout_recovery_level) { case 1: diff --git a/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c b/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c index 6a3f0fc56c3b..69523ac85639 100644 --- a/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c +++ b/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c @@ -2322,6 +2322,22 @@ static int i40e_ctrl_vf_rx_rings(struct i40e_vsi *vsi, unsigned long q_map, } /** + * i40e_vc_validate_vqs_bitmaps - validate Rx/Tx queue bitmaps from VIRTHCHNL + * @vqs: virtchnl_queue_select structure containing bitmaps to validate + * + * Returns true if validation was successful, else false. + */ +static bool i40e_vc_validate_vqs_bitmaps(struct virtchnl_queue_select *vqs) +{ + if ((!vqs->rx_queues && !vqs->tx_queues) || + vqs->rx_queues >= BIT(I40E_MAX_VF_QUEUES) || + vqs->tx_queues >= BIT(I40E_MAX_VF_QUEUES)) + return false; + + return true; +} + +/** * i40e_vc_enable_queues_msg * @vf: pointer to the VF info * @msg: pointer to the msg buffer @@ -2346,7 +2362,7 @@ static int i40e_vc_enable_queues_msg(struct i40e_vf *vf, u8 *msg) goto error_param; } - if ((0 == vqs->rx_queues) && (0 == vqs->tx_queues)) { + if (i40e_vc_validate_vqs_bitmaps(vqs)) { aq_ret = I40E_ERR_PARAM; goto error_param; } @@ -2408,9 +2424,7 @@ static int i40e_vc_disable_queues_msg(struct i40e_vf *vf, u8 *msg) goto error_param; } - if ((vqs->rx_queues == 0 && vqs->tx_queues == 0) || - vqs->rx_queues > I40E_MAX_VF_QUEUES || - vqs->tx_queues > I40E_MAX_VF_QUEUES) { + if (i40e_vc_validate_vqs_bitmaps(vqs)) { aq_ret = I40E_ERR_PARAM; goto error_param; } diff --git a/drivers/net/ethernet/intel/iavf/iavf.h b/drivers/net/ethernet/intel/iavf/iavf.h index 29de3ae96ef2..bd1b1ed323f4 100644 --- a/drivers/net/ethernet/intel/iavf/iavf.h +++ b/drivers/net/ethernet/intel/iavf/iavf.h @@ -415,4 +415,6 @@ void iavf_enable_channels(struct iavf_adapter *adapter); void iavf_disable_channels(struct iavf_adapter *adapter); void iavf_add_cloud_filter(struct iavf_adapter *adapter); void iavf_del_cloud_filter(struct iavf_adapter *adapter); +struct iavf_mac_filter *iavf_add_filter(struct iavf_adapter *adapter, + const u8 *macaddr); #endif /* _IAVF_H_ */ diff --git a/drivers/net/ethernet/intel/iavf/iavf_main.c b/drivers/net/ethernet/intel/iavf/iavf_main.c index 0a8824871618..62fe56ddcb6e 100644 --- a/drivers/net/ethernet/intel/iavf/iavf_main.c +++ b/drivers/net/ethernet/intel/iavf/iavf_main.c @@ -743,9 +743,8 @@ iavf_mac_filter *iavf_find_filter(struct iavf_adapter *adapter, * * Returns ptr to the filter object or NULL when no memory available. **/ -static struct -iavf_mac_filter *iavf_add_filter(struct iavf_adapter *adapter, - const u8 *macaddr) +struct iavf_mac_filter *iavf_add_filter(struct iavf_adapter *adapter, + const u8 *macaddr) { struct iavf_mac_filter *f; @@ -2065,9 +2064,9 @@ static void iavf_reset_task(struct work_struct *work) struct virtchnl_vf_resource *vfres = adapter->vf_res; struct net_device *netdev = adapter->netdev; struct iavf_hw *hw = &adapter->hw; + struct iavf_mac_filter *f, *ftmp; struct iavf_vlan_filter *vlf; struct iavf_cloud_filter *cf; - struct iavf_mac_filter *f; u32 reg_val; int i = 0, err; bool running; @@ -2181,6 +2180,16 @@ continue_reset: spin_lock_bh(&adapter->mac_vlan_list_lock); + /* Delete filter for the current MAC address, it could have + * been changed by the PF via administratively set MAC. + * Will be re-added via VIRTCHNL_OP_GET_VF_RESOURCES. + */ + list_for_each_entry_safe(f, ftmp, &adapter->mac_filter_list, list) { + if (ether_addr_equal(f->macaddr, adapter->hw.mac.addr)) { + list_del(&f->list); + kfree(f); + } + } /* re-add all MAC filters */ list_for_each_entry(f, &adapter->mac_filter_list, list) { f->add = true; diff --git a/drivers/net/ethernet/intel/iavf/iavf_virtchnl.c b/drivers/net/ethernet/intel/iavf/iavf_virtchnl.c index c46770eba320..1ab9cb339acb 100644 --- a/drivers/net/ethernet/intel/iavf/iavf_virtchnl.c +++ b/drivers/net/ethernet/intel/iavf/iavf_virtchnl.c @@ -1359,6 +1359,9 @@ void iavf_virtchnl_completion(struct iavf_adapter *adapter, ether_addr_copy(netdev->perm_addr, adapter->hw.mac.addr); } + spin_lock_bh(&adapter->mac_vlan_list_lock); + iavf_add_filter(adapter, adapter->hw.mac.addr); + spin_unlock_bh(&adapter->mac_vlan_list_lock); iavf_process_config(adapter); } break; diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c index bf539483e25e..eb9d00608e9a 100644 --- a/drivers/net/ethernet/intel/ice/ice_main.c +++ b/drivers/net/ethernet/intel/ice/ice_main.c @@ -5086,36 +5086,17 @@ static void ice_tx_timeout(struct net_device *netdev, unsigned int txqueue) struct ice_ring *tx_ring = NULL; struct ice_vsi *vsi = np->vsi; struct ice_pf *pf = vsi->back; - int hung_queue = -1; u32 i; pf->tx_timeout_count++; - /* find the stopped queue the same way dev_watchdog() does */ - for (i = 0; i < netdev->num_tx_queues; i++) { - unsigned long trans_start; - struct netdev_queue *q; - - q = netdev_get_tx_queue(netdev, i); - trans_start = q->trans_start; - if (netif_xmit_stopped(q) && - time_after(jiffies, - trans_start + netdev->watchdog_timeo)) { - hung_queue = i; - break; - } - } - - if (i == netdev->num_tx_queues) - netdev_info(netdev, "tx_timeout: no netdev hung queue found\n"); - else - /* now that we have an index, find the tx_ring struct */ - for (i = 0; i < vsi->num_txq; i++) - if (vsi->tx_rings[i] && vsi->tx_rings[i]->desc) - if (hung_queue == vsi->tx_rings[i]->q_index) { - tx_ring = vsi->tx_rings[i]; - break; - } + /* now that we have an index, find the tx_ring struct */ + for (i = 0; i < vsi->num_txq; i++) + if (vsi->tx_rings[i] && vsi->tx_rings[i]->desc) + if (txqueue == vsi->tx_rings[i]->q_index) { + tx_ring = vsi->tx_rings[i]; + break; + } /* Reset recovery level if enough time has elapsed after last timeout. * Also ensure no new reset action happens before next timeout period. @@ -5130,19 +5111,19 @@ static void ice_tx_timeout(struct net_device *netdev, unsigned int txqueue) struct ice_hw *hw = &pf->hw; u32 head, val = 0; - head = (rd32(hw, QTX_COMM_HEAD(vsi->txq_map[hung_queue])) & + head = (rd32(hw, QTX_COMM_HEAD(vsi->txq_map[txqueue])) & QTX_COMM_HEAD_HEAD_M) >> QTX_COMM_HEAD_HEAD_S; /* Read interrupt register */ val = rd32(hw, GLINT_DYN_CTL(tx_ring->q_vector->reg_idx)); netdev_info(netdev, "tx_timeout: VSI_num: %d, Q %d, NTC: 0x%x, HW_HEAD: 0x%x, NTU: 0x%x, INT: 0x%x\n", - vsi->vsi_num, hung_queue, tx_ring->next_to_clean, + vsi->vsi_num, txqueue, tx_ring->next_to_clean, head, tx_ring->next_to_use, val); } pf->tx_timeout_last_recovery = jiffies; - netdev_info(netdev, "tx_timeout recovery level %d, hung_queue %d\n", - pf->tx_timeout_recovery_level, hung_queue); + netdev_info(netdev, "tx_timeout recovery level %d, txqueue %d\n", + pf->tx_timeout_recovery_level, txqueue); switch (pf->tx_timeout_recovery_level) { case 1: diff --git a/drivers/net/ethernet/intel/ice/ice_xsk.c b/drivers/net/ethernet/intel/ice/ice_xsk.c index 0e401f116d54..149dca0012ba 100644 --- a/drivers/net/ethernet/intel/ice/ice_xsk.c +++ b/drivers/net/ethernet/intel/ice/ice_xsk.c @@ -1020,8 +1020,8 @@ bool ice_clean_tx_irq_zc(struct ice_ring *xdp_ring, int budget) s16 ntc = xdp_ring->next_to_clean; struct ice_tx_desc *tx_desc; struct ice_tx_buf *tx_buf; - bool xmit_done = true; u32 xsk_frames = 0; + bool xmit_done; tx_desc = ICE_TX_DESC(xdp_ring, ntc); tx_buf = &xdp_ring->tx_buf[ntc]; diff --git a/drivers/net/ethernet/intel/igb/e1000_82575.c b/drivers/net/ethernet/intel/igb/e1000_82575.c index 8a6ef3514129..438b42ce2cd9 100644 --- a/drivers/net/ethernet/intel/igb/e1000_82575.c +++ b/drivers/net/ethernet/intel/igb/e1000_82575.c @@ -530,7 +530,7 @@ static s32 igb_set_sfp_media_type_82575(struct e1000_hw *hw) dev_spec->module_plugged = true; if (eth_flags->e1000_base_lx || eth_flags->e1000_base_sx) { hw->phy.media_type = e1000_media_type_internal_serdes; - } else if (eth_flags->e100_base_fx) { + } else if (eth_flags->e100_base_fx || eth_flags->e100_base_lx) { dev_spec->sgmii_active = true; hw->phy.media_type = e1000_media_type_internal_serdes; } else if (eth_flags->e1000_base_t) { @@ -657,14 +657,10 @@ static s32 igb_get_invariants_82575(struct e1000_hw *hw) break; } - /* do not change link mode for 100BaseFX */ - if (dev_spec->eth_flags.e100_base_fx) - break; - /* change current link mode setting */ ctrl_ext &= ~E1000_CTRL_EXT_LINK_MODE_MASK; - if (hw->phy.media_type == e1000_media_type_copper) + if (dev_spec->sgmii_active) ctrl_ext |= E1000_CTRL_EXT_LINK_MODE_SGMII; else ctrl_ext |= E1000_CTRL_EXT_LINK_MODE_PCIE_SERDES; diff --git a/drivers/net/ethernet/intel/igb/igb_ethtool.c b/drivers/net/ethernet/intel/igb/igb_ethtool.c index 43c438365389..f96ffa83efbe 100644 --- a/drivers/net/ethernet/intel/igb/igb_ethtool.c +++ b/drivers/net/ethernet/intel/igb/igb_ethtool.c @@ -181,7 +181,7 @@ static int igb_get_link_ksettings(struct net_device *netdev, advertising &= ~ADVERTISED_1000baseKX_Full; } } - if (eth_flags->e100_base_fx) { + if (eth_flags->e100_base_fx || eth_flags->e100_base_lx) { supported |= SUPPORTED_100baseT_Full; advertising |= ADVERTISED_100baseT_Full; } diff --git a/drivers/net/ethernet/intel/igc/igc_base.c b/drivers/net/ethernet/intel/igc/igc_base.c index db289bcce21d..5a506440560a 100644 --- a/drivers/net/ethernet/intel/igc/igc_base.c +++ b/drivers/net/ethernet/intel/igc/igc_base.c @@ -212,6 +212,7 @@ static s32 igc_get_invariants_base(struct igc_hw *hw) case IGC_DEV_ID_I225_I: case IGC_DEV_ID_I220_V: case IGC_DEV_ID_I225_K: + case IGC_DEV_ID_I225_BLANK_NVM: mac->type = igc_i225; break; default: diff --git a/drivers/net/ethernet/intel/igc/igc_defines.h b/drivers/net/ethernet/intel/igc/igc_defines.h index 2121fc34e300..58efa7a02c68 100644 --- a/drivers/net/ethernet/intel/igc/igc_defines.h +++ b/drivers/net/ethernet/intel/igc/igc_defines.h @@ -282,6 +282,10 @@ #define IGC_TXD_STAT_TC 0x00000004 /* Tx Underrun */ #define IGC_TXD_EXTCMD_TSTAMP 0x00000010 /* IEEE1588 Timestamp packet */ +/* IPSec Encrypt Enable */ +#define IGC_ADVTXD_L4LEN_SHIFT 8 /* Adv ctxt L4LEN shift */ +#define IGC_ADVTXD_MSS_SHIFT 16 /* Adv ctxt MSS shift */ + /* Transmit Control */ #define IGC_TCTL_EN 0x00000002 /* enable Tx */ #define IGC_TCTL_PSP 0x00000008 /* pad short packets */ @@ -460,6 +464,7 @@ /* PHY Status Register */ #define MII_SR_LINK_STATUS 0x0004 /* Link Status 1 = link */ #define MII_SR_AUTONEG_COMPLETE 0x0020 /* Auto Neg Complete */ +#define IGC_PHY_RST_COMP 0x0100 /* Internal PHY reset completion */ /* PHY 1000 MII Register/Bit Definitions */ /* PHY Registers defined by IEEE */ diff --git a/drivers/net/ethernet/intel/igc/igc_hw.h b/drivers/net/ethernet/intel/igc/igc_hw.h index 20f710645746..90ac0e0144d8 100644 --- a/drivers/net/ethernet/intel/igc/igc_hw.h +++ b/drivers/net/ethernet/intel/igc/igc_hw.h @@ -21,8 +21,7 @@ #define IGC_DEV_ID_I225_I 0x15F8 #define IGC_DEV_ID_I220_V 0x15F7 #define IGC_DEV_ID_I225_K 0x3100 - -#define IGC_FUNC_0 0 +#define IGC_DEV_ID_I225_BLANK_NVM 0x15FD /* Function pointers for the MAC. */ struct igc_mac_operations { diff --git a/drivers/net/ethernet/intel/igc/igc_main.c b/drivers/net/ethernet/intel/igc/igc_main.c index c359f3d9fb25..d9d5425fe8d9 100644 --- a/drivers/net/ethernet/intel/igc/igc_main.c +++ b/drivers/net/ethernet/intel/igc/igc_main.c @@ -45,6 +45,7 @@ static const struct pci_device_id igc_pci_tbl[] = { { PCI_VDEVICE(INTEL, IGC_DEV_ID_I225_I), board_base }, { PCI_VDEVICE(INTEL, IGC_DEV_ID_I220_V), board_base }, { PCI_VDEVICE(INTEL, IGC_DEV_ID_I225_K), board_base }, + { PCI_VDEVICE(INTEL, IGC_DEV_ID_I225_BLANK_NVM), board_base }, /* required last entry */ {0, } }; @@ -880,7 +881,7 @@ static void igc_tx_ctxtdesc(struct igc_ring *tx_ring, /* set bits to identify this as an advanced context descriptor */ type_tucmd |= IGC_TXD_CMD_DEXT | IGC_ADVTXD_DTYP_CTXT; - /* For 82575, context index must be unique per ring. */ + /* For i225, context index must be unique per ring. */ if (test_bit(IGC_RING_FLAG_TX_CTX_IDX, &tx_ring->flags)) mss_l4len_idx |= tx_ring->reg_idx << 4; @@ -999,6 +1000,10 @@ static u32 igc_tx_cmd_type(struct sk_buff *skb, u32 tx_flags) IGC_ADVTXD_DCMD_DEXT | IGC_ADVTXD_DCMD_IFCS; + /* set segmentation bits for TSO */ + cmd_type |= IGC_SET_FLAG(tx_flags, IGC_TX_FLAGS_TSO, + (IGC_ADVTXD_DCMD_TSE)); + /* set timestamp bit if present */ cmd_type |= IGC_SET_FLAG(tx_flags, IGC_TX_FLAGS_TSTAMP, (IGC_ADVTXD_MAC_TSTAMP)); @@ -1170,6 +1175,100 @@ dma_error: return -1; } +static int igc_tso(struct igc_ring *tx_ring, + struct igc_tx_buffer *first, + u8 *hdr_len) +{ + u32 vlan_macip_lens, type_tucmd, mss_l4len_idx; + struct sk_buff *skb = first->skb; + union { + struct iphdr *v4; + struct ipv6hdr *v6; + unsigned char *hdr; + } ip; + union { + struct tcphdr *tcp; + struct udphdr *udp; + unsigned char *hdr; + } l4; + u32 paylen, l4_offset; + int err; + + if (skb->ip_summed != CHECKSUM_PARTIAL) + return 0; + + if (!skb_is_gso(skb)) + return 0; + + err = skb_cow_head(skb, 0); + if (err < 0) + return err; + + ip.hdr = skb_network_header(skb); + l4.hdr = skb_checksum_start(skb); + + /* ADV DTYP TUCMD MKRLOC/ISCSIHEDLEN */ + type_tucmd = IGC_ADVTXD_TUCMD_L4T_TCP; + + /* initialize outer IP header fields */ + if (ip.v4->version == 4) { + unsigned char *csum_start = skb_checksum_start(skb); + unsigned char *trans_start = ip.hdr + (ip.v4->ihl * 4); + + /* IP header will have to cancel out any data that + * is not a part of the outer IP header + */ + ip.v4->check = csum_fold(csum_partial(trans_start, + csum_start - trans_start, + 0)); + type_tucmd |= IGC_ADVTXD_TUCMD_IPV4; + + ip.v4->tot_len = 0; + first->tx_flags |= IGC_TX_FLAGS_TSO | + IGC_TX_FLAGS_CSUM | + IGC_TX_FLAGS_IPV4; + } else { + ip.v6->payload_len = 0; + first->tx_flags |= IGC_TX_FLAGS_TSO | + IGC_TX_FLAGS_CSUM; + } + + /* determine offset of inner transport header */ + l4_offset = l4.hdr - skb->data; + + /* remove payload length from inner checksum */ + paylen = skb->len - l4_offset; + if (type_tucmd & IGC_ADVTXD_TUCMD_L4T_TCP) { + /* compute length of segmentation header */ + *hdr_len = (l4.tcp->doff * 4) + l4_offset; + csum_replace_by_diff(&l4.tcp->check, + (__force __wsum)htonl(paylen)); + } else { + /* compute length of segmentation header */ + *hdr_len = sizeof(*l4.udp) + l4_offset; + csum_replace_by_diff(&l4.udp->check, + (__force __wsum)htonl(paylen)); + } + + /* update gso size and bytecount with header size */ + first->gso_segs = skb_shinfo(skb)->gso_segs; + first->bytecount += (first->gso_segs - 1) * *hdr_len; + + /* MSS L4LEN IDX */ + mss_l4len_idx = (*hdr_len - l4_offset) << IGC_ADVTXD_L4LEN_SHIFT; + mss_l4len_idx |= skb_shinfo(skb)->gso_size << IGC_ADVTXD_MSS_SHIFT; + + /* VLAN MACLEN IPLEN */ + vlan_macip_lens = l4.hdr - ip.hdr; + vlan_macip_lens |= (ip.hdr - skb->data) << IGC_ADVTXD_MACLEN_SHIFT; + vlan_macip_lens |= first->tx_flags & IGC_TX_FLAGS_VLAN_MASK; + + igc_tx_ctxtdesc(tx_ring, first, vlan_macip_lens, + type_tucmd, mss_l4len_idx); + + return 1; +} + static netdev_tx_t igc_xmit_frame_ring(struct sk_buff *skb, struct igc_ring *tx_ring) { @@ -1179,6 +1278,7 @@ static netdev_tx_t igc_xmit_frame_ring(struct sk_buff *skb, u32 tx_flags = 0; unsigned short f; u8 hdr_len = 0; + int tso = 0; /* need: 1 descriptor per page * PAGE_SIZE/IGC_MAX_DATA_PER_TXD, * + 1 desc for skb_headlen/IGC_MAX_DATA_PER_TXD, @@ -1225,11 +1325,21 @@ static netdev_tx_t igc_xmit_frame_ring(struct sk_buff *skb, first->tx_flags = tx_flags; first->protocol = protocol; - igc_tx_csum(tx_ring, first); + tso = igc_tso(tx_ring, first, &hdr_len); + if (tso < 0) + goto out_drop; + else if (!tso) + igc_tx_csum(tx_ring, first); igc_tx_map(tx_ring, first, hdr_len); return NETDEV_TX_OK; + +out_drop: + dev_kfree_skb_any(first->skb); + first->skb = NULL; + + return NETDEV_TX_OK; } static inline struct igc_ring *igc_tx_queue_mapping(struct igc_adapter *adapter, @@ -4588,6 +4698,8 @@ static int igc_probe(struct pci_dev *pdev, /* Add supported features to the features list*/ netdev->features |= NETIF_F_SG; + netdev->features |= NETIF_F_TSO; + netdev->features |= NETIF_F_TSO6; netdev->features |= NETIF_F_RXCSUM; netdev->features |= NETIF_F_HW_CSUM; netdev->features |= NETIF_F_SCTP_CRC; diff --git a/drivers/net/ethernet/intel/igc/igc_phy.c b/drivers/net/ethernet/intel/igc/igc_phy.c index f4b05af0dd2f..8e1799508edc 100644 --- a/drivers/net/ethernet/intel/igc/igc_phy.c +++ b/drivers/net/ethernet/intel/igc/igc_phy.c @@ -173,6 +173,7 @@ s32 igc_check_downshift(struct igc_hw *hw) s32 igc_phy_hw_reset(struct igc_hw *hw) { struct igc_phy_info *phy = &hw->phy; + u32 phpm = 0, timeout = 10000; s32 ret_val; u32 ctrl; @@ -186,6 +187,8 @@ s32 igc_phy_hw_reset(struct igc_hw *hw) if (ret_val) goto out; + phpm = rd32(IGC_I225_PHPM); + ctrl = rd32(IGC_CTRL); wr32(IGC_CTRL, ctrl | IGC_CTRL_PHY_RST); wrfl(); @@ -195,7 +198,18 @@ s32 igc_phy_hw_reset(struct igc_hw *hw) wr32(IGC_CTRL, ctrl); wrfl(); - usleep_range(1500, 2000); + /* SW should guarantee 100us for the completion of the PHY reset */ + usleep_range(100, 150); + do { + phpm = rd32(IGC_I225_PHPM); + timeout--; + udelay(1); + } while (!(phpm & IGC_PHY_RST_COMP) && timeout); + + if (!timeout) + hw_dbg("Timeout is expired after a phy reset\n"); + + usleep_range(100, 150); phy->ops.release(hw); diff --git a/drivers/net/ethernet/intel/igc/igc_regs.h b/drivers/net/ethernet/intel/igc/igc_regs.h index c82111051898..c9029b549b90 100644 --- a/drivers/net/ethernet/intel/igc/igc_regs.h +++ b/drivers/net/ethernet/intel/igc/igc_regs.h @@ -12,6 +12,7 @@ #define IGC_MDIC 0x00020 /* MDI Control - RW */ #define IGC_MDICNFG 0x00E04 /* MDC/MDIO Configuration - RW */ #define IGC_CONNSW 0x00034 /* Copper/Fiber switch control - RW */ +#define IGC_I225_PHPM 0x00E14 /* I225 PHY Power Management */ /* Internal Packet Buffer Size Registers */ #define IGC_RXPBS 0x02404 /* Rx Packet Buffer Size - RW */ diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c b/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c index 4c13cca656b2..718931d951bc 100644 --- a/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c +++ b/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c @@ -5239,7 +5239,7 @@ static void ixgbe_fdir_filter_restore(struct ixgbe_adapter *adapter) struct ixgbe_hw *hw = &adapter->hw; struct hlist_node *node2; struct ixgbe_fdir_filter *filter; - u64 action; + u8 queue; spin_lock(&adapter->fdir_perfect_lock); @@ -5248,17 +5248,34 @@ static void ixgbe_fdir_filter_restore(struct ixgbe_adapter *adapter) hlist_for_each_entry_safe(filter, node2, &adapter->fdir_filter_list, fdir_node) { - action = filter->action; - if (action != IXGBE_FDIR_DROP_QUEUE && action != 0) - action = - (action >> ETHTOOL_RX_FLOW_SPEC_RING_VF_OFF) - 1; + if (filter->action == IXGBE_FDIR_DROP_QUEUE) { + queue = IXGBE_FDIR_DROP_QUEUE; + } else { + u32 ring = ethtool_get_flow_spec_ring(filter->action); + u8 vf = ethtool_get_flow_spec_ring_vf(filter->action); + + if (!vf && (ring >= adapter->num_rx_queues)) { + e_err(drv, "FDIR restore failed without VF, ring: %u\n", + ring); + continue; + } else if (vf && + ((vf > adapter->num_vfs) || + ring >= adapter->num_rx_queues_per_pool)) { + e_err(drv, "FDIR restore failed with VF, vf: %hhu, ring: %u\n", + vf, ring); + continue; + } + + /* Map the ring onto the absolute queue index */ + if (!vf) + queue = adapter->rx_ring[ring]->reg_idx; + else + queue = ((vf - 1) * + adapter->num_rx_queues_per_pool) + ring; + } ixgbe_fdir_write_perfect_filter_82599(hw, - &filter->filter, - filter->sw_idx, - (action == IXGBE_FDIR_DROP_QUEUE) ? - IXGBE_FDIR_DROP_QUEUE : - adapter->rx_ring[action]->reg_idx); + &filter->filter, filter->sw_idx, queue); } spin_unlock(&adapter->fdir_perfect_lock); diff --git a/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c b/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c index fa286694ac2c..4622c4ea2e46 100644 --- a/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c +++ b/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c @@ -2081,11 +2081,6 @@ static int ixgbevf_write_uc_addr_list(struct net_device *netdev) struct ixgbe_hw *hw = &adapter->hw; int count = 0; - if ((netdev_uc_count(netdev)) > 10) { - pr_err("Too many unicast filters - No Space\n"); - return -ENOSPC; - } - if (!netdev_uc_empty(netdev)) { struct netdev_hw_addr *ha; diff --git a/drivers/net/ethernet/lantiq_etop.c b/drivers/net/ethernet/lantiq_etop.c index 028e3e6222e9..578c31697cc0 100644 --- a/drivers/net/ethernet/lantiq_etop.c +++ b/drivers/net/ethernet/lantiq_etop.c @@ -510,13 +510,6 @@ ltq_etop_change_mtu(struct net_device *dev, int new_mtu) } static int -ltq_etop_ioctl(struct net_device *dev, struct ifreq *rq, int cmd) -{ - /* TODO: mii-toll reports "No MII transceiver present!." ?!*/ - return phy_mii_ioctl(dev->phydev, rq, cmd); -} - -static int ltq_etop_set_mac_address(struct net_device *dev, void *p) { int ret = eth_mac_addr(dev, p); @@ -616,7 +609,7 @@ static const struct net_device_ops ltq_eth_netdev_ops = { .ndo_stop = ltq_etop_stop, .ndo_start_xmit = ltq_etop_tx, .ndo_change_mtu = ltq_etop_change_mtu, - .ndo_do_ioctl = ltq_etop_ioctl, + .ndo_do_ioctl = phy_do_ioctl, .ndo_set_mac_address = ltq_etop_set_mac_address, .ndo_validate_addr = eth_validate_addr, .ndo_set_rx_mode = ltq_etop_set_multicast_list, diff --git a/drivers/net/ethernet/marvell/mvneta.c b/drivers/net/ethernet/marvell/mvneta.c index 71a872d46bc4..0449d4b28ade 100644 --- a/drivers/net/ethernet/marvell/mvneta.c +++ b/drivers/net/ethernet/marvell/mvneta.c @@ -2081,7 +2081,11 @@ static int mvneta_run_xdp(struct mvneta_port *pp, struct mvneta_rx_queue *rxq, struct bpf_prog *prog, struct xdp_buff *xdp) { - u32 ret, act = bpf_prog_run_xdp(prog, xdp); + unsigned int len; + u32 ret, act; + + len = xdp->data_end - xdp->data_hard_start - pp->rx_offset_correction; + act = bpf_prog_run_xdp(prog, xdp); switch (act) { case XDP_PASS: @@ -2094,9 +2098,8 @@ mvneta_run_xdp(struct mvneta_port *pp, struct mvneta_rx_queue *rxq, if (err) { ret = MVNETA_XDP_DROPPED; __page_pool_put_page(rxq->page_pool, - virt_to_head_page(xdp->data), - xdp->data_end - xdp->data_hard_start, - true); + virt_to_head_page(xdp->data), + len, true); } else { ret = MVNETA_XDP_REDIR; } @@ -2106,9 +2109,8 @@ mvneta_run_xdp(struct mvneta_port *pp, struct mvneta_rx_queue *rxq, ret = mvneta_xdp_xmit_back(pp, xdp); if (ret != MVNETA_XDP_TX) __page_pool_put_page(rxq->page_pool, - virt_to_head_page(xdp->data), - xdp->data_end - xdp->data_hard_start, - true); + virt_to_head_page(xdp->data), + len, true); break; default: bpf_warn_invalid_xdp_action(act); @@ -2119,8 +2121,7 @@ mvneta_run_xdp(struct mvneta_port *pp, struct mvneta_rx_queue *rxq, case XDP_DROP: __page_pool_put_page(rxq->page_pool, virt_to_head_page(xdp->data), - xdp->data_end - xdp->data_hard_start, - true); + len, true); ret = MVNETA_XDP_DROPPED; break; } @@ -3071,7 +3072,7 @@ static int mvneta_create_page_pool(struct mvneta_port *pp, .order = 0, .flags = PP_FLAG_DMA_MAP | PP_FLAG_DMA_SYNC_DEV, .pool_size = size, - .nid = cpu_to_node(0), + .nid = NUMA_NO_NODE, .dev = pp->dev->dev.parent, .dma_dir = xdp_prog ? DMA_BIDIRECTIONAL : DMA_FROM_DEVICE, .offset = pp->rx_offset_correction, diff --git a/drivers/net/ethernet/marvell/pxa168_eth.c b/drivers/net/ethernet/marvell/pxa168_eth.c index 1a6877902dd6..7a0d785b826c 100644 --- a/drivers/net/ethernet/marvell/pxa168_eth.c +++ b/drivers/net/ethernet/marvell/pxa168_eth.c @@ -1344,15 +1344,6 @@ static int pxa168_smi_write(struct mii_bus *bus, int phy_addr, int regnum, return 0; } -static int pxa168_eth_do_ioctl(struct net_device *dev, struct ifreq *ifr, - int cmd) -{ - if (dev->phydev) - return phy_mii_ioctl(dev->phydev, ifr, cmd); - - return -EOPNOTSUPP; -} - #ifdef CONFIG_NET_POLL_CONTROLLER static void pxa168_eth_netpoll(struct net_device *dev) { @@ -1387,7 +1378,7 @@ static const struct net_device_ops pxa168_eth_netdev_ops = { .ndo_set_rx_mode = pxa168_eth_set_rx_mode, .ndo_set_mac_address = pxa168_eth_set_mac_address, .ndo_validate_addr = eth_validate_addr, - .ndo_do_ioctl = pxa168_eth_do_ioctl, + .ndo_do_ioctl = phy_do_ioctl, .ndo_change_mtu = pxa168_eth_change_mtu, .ndo_tx_timeout = pxa168_eth_tx_timeout, #ifdef CONFIG_NET_POLL_CONTROLLER diff --git a/drivers/net/ethernet/mellanox/mlx4/crdump.c b/drivers/net/ethernet/mellanox/mlx4/crdump.c index eaf08f7ad128..64ed725aec28 100644 --- a/drivers/net/ethernet/mellanox/mlx4/crdump.c +++ b/drivers/net/ethernet/mellanox/mlx4/crdump.c @@ -182,7 +182,7 @@ int mlx4_crdump_collect(struct mlx4_dev *dev) crdump_enable_crspace_access(dev, cr_space); /* Get the available snapshot ID for the dumps */ - id = devlink_region_shapshot_id_get(devlink); + id = devlink_region_snapshot_id_get(devlink); /* Try to capture dumps */ mlx4_crdump_collect_crspace(dev, cr_space, id); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/Makefile b/drivers/net/ethernet/mellanox/mlx5/core/Makefile index a6f390fdb971..d3e06cec8317 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/Makefile +++ b/drivers/net/ethernet/mellanox/mlx5/core/Makefile @@ -42,7 +42,7 @@ mlx5_core-$(CONFIG_PCI_HYPERV_INTERFACE) += en/hv_vhca_stats.o # Core extra # mlx5_core-$(CONFIG_MLX5_ESWITCH) += eswitch.o eswitch_offloads.o eswitch_offloads_termtbl.o \ - ecpf.o rdma.o + ecpf.o rdma.o eswitch_offloads_chains.o mlx5_core-$(CONFIG_MLX5_MPFS) += lib/mpfs.o mlx5_core-$(CONFIG_VXLAN) += lib/vxlan.o mlx5_core-$(CONFIG_PTP_1588_CLOCK) += lib/clock.o diff --git a/drivers/net/ethernet/mellanox/mlx5/core/alloc.c b/drivers/net/ethernet/mellanox/mlx5/core/alloc.c index 549f962cd86e..42198e64a7f4 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/alloc.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/alloc.c @@ -71,8 +71,8 @@ static void *mlx5_dma_zalloc_coherent_node(struct mlx5_core_dev *dev, return cpu_handle; } -int mlx5_buf_alloc_node(struct mlx5_core_dev *dev, int size, - struct mlx5_frag_buf *buf, int node) +static int mlx5_buf_alloc_node(struct mlx5_core_dev *dev, int size, + struct mlx5_frag_buf *buf, int node) { dma_addr_t t; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en.h b/drivers/net/ethernet/mellanox/mlx5/core/en.h index 9c8427698238..fc80b59db9a8 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en.h @@ -135,7 +135,7 @@ struct page_pool; #define MLX5E_LOG_INDIR_RQT_SIZE 0x7 #define MLX5E_INDIR_RQT_SIZE BIT(MLX5E_LOG_INDIR_RQT_SIZE) #define MLX5E_MIN_NUM_CHANNELS 0x1 -#define MLX5E_MAX_NUM_CHANNELS (MLX5E_INDIR_RQT_SIZE >> 1) +#define MLX5E_MAX_NUM_CHANNELS MLX5E_INDIR_RQT_SIZE #define MLX5E_MAX_NUM_SQS (MLX5E_MAX_NUM_CHANNELS * MLX5E_MAX_NUM_TC) #define MLX5E_TX_CQ_POLL_BUDGET 128 #define MLX5E_TX_XSK_POLL_BUDGET 64 @@ -1175,11 +1175,11 @@ int mlx5e_attach_netdev(struct mlx5e_priv *priv); void mlx5e_detach_netdev(struct mlx5e_priv *priv); void mlx5e_destroy_netdev(struct mlx5e_priv *priv); void mlx5e_set_netdev_mtu_boundaries(struct mlx5e_priv *priv); -void mlx5e_build_nic_params(struct mlx5_core_dev *mdev, +void mlx5e_build_nic_params(struct mlx5e_priv *priv, struct mlx5e_xsk *xsk, struct mlx5e_rss_params *rss_params, struct mlx5e_params *params, - u16 max_channels, u16 mtu); + u16 mtu); void mlx5e_build_rq_params(struct mlx5_core_dev *mdev, struct mlx5e_params *params); void mlx5e_build_rss_params(struct mlx5e_rss_params *rss_params, diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/fs.h b/drivers/net/ethernet/mellanox/mlx5/core/en/fs.h index 68d593074f6c..d48292ccda29 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/fs.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/fs.h @@ -122,6 +122,22 @@ enum { #endif }; +#define MLX5E_TTC_NUM_GROUPS 3 +#define MLX5E_TTC_GROUP1_SIZE (BIT(3) + MLX5E_NUM_TUNNEL_TT) +#define MLX5E_TTC_GROUP2_SIZE BIT(1) +#define MLX5E_TTC_GROUP3_SIZE BIT(0) +#define MLX5E_TTC_TABLE_SIZE (MLX5E_TTC_GROUP1_SIZE +\ + MLX5E_TTC_GROUP2_SIZE +\ + MLX5E_TTC_GROUP3_SIZE) + +#define MLX5E_INNER_TTC_NUM_GROUPS 3 +#define MLX5E_INNER_TTC_GROUP1_SIZE BIT(3) +#define MLX5E_INNER_TTC_GROUP2_SIZE BIT(1) +#define MLX5E_INNER_TTC_GROUP3_SIZE BIT(0) +#define MLX5E_INNER_TTC_TABLE_SIZE (MLX5E_INNER_TTC_GROUP1_SIZE +\ + MLX5E_INNER_TTC_GROUP2_SIZE +\ + MLX5E_INNER_TTC_GROUP3_SIZE) + #ifdef CONFIG_MLX5_EN_RXNFC struct mlx5e_ethtool_table { diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/health.c b/drivers/net/ethernet/mellanox/mlx5/core/en/health.c index 1d6b58860da6..3a975641f902 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/health.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/health.c @@ -197,9 +197,10 @@ int mlx5e_health_report(struct mlx5e_priv *priv, struct devlink_health_reporter *reporter, char *err_str, struct mlx5e_err_ctx *err_ctx) { - if (!reporter) { - netdev_err(priv->netdev, err_str); + netdev_err(priv->netdev, err_str); + + if (!reporter) return err_ctx->recover(&err_ctx->ctx); - } + return devlink_health_report(reporter, err_str, err_ctx); } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c b/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c index 15b7f0f1427c..73d3dc07331f 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c @@ -904,22 +904,6 @@ del_rules: return err; } -#define MLX5E_TTC_NUM_GROUPS 3 -#define MLX5E_TTC_GROUP1_SIZE (BIT(3) + MLX5E_NUM_TUNNEL_TT) -#define MLX5E_TTC_GROUP2_SIZE BIT(1) -#define MLX5E_TTC_GROUP3_SIZE BIT(0) -#define MLX5E_TTC_TABLE_SIZE (MLX5E_TTC_GROUP1_SIZE +\ - MLX5E_TTC_GROUP2_SIZE +\ - MLX5E_TTC_GROUP3_SIZE) - -#define MLX5E_INNER_TTC_NUM_GROUPS 3 -#define MLX5E_INNER_TTC_GROUP1_SIZE BIT(3) -#define MLX5E_INNER_TTC_GROUP2_SIZE BIT(1) -#define MLX5E_INNER_TTC_GROUP3_SIZE BIT(0) -#define MLX5E_INNER_TTC_TABLE_SIZE (MLX5E_INNER_TTC_GROUP1_SIZE +\ - MLX5E_INNER_TTC_GROUP2_SIZE +\ - MLX5E_INNER_TTC_GROUP3_SIZE) - static int mlx5e_create_ttc_table_groups(struct mlx5e_ttc_table *ttc, bool use_ipv) { diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c b/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c index acd946f2ddbe..3bc2ac3d53fc 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c @@ -58,6 +58,7 @@ static struct mlx5e_ethtool_table *get_flow_table(struct mlx5e_priv *priv, struct ethtool_rx_flow_spec *fs, int num_tuples) { + struct mlx5_flow_table_attr ft_attr = {}; struct mlx5e_ethtool_table *eth_ft; struct mlx5_flow_namespace *ns; struct mlx5_flow_table *ft; @@ -102,9 +103,11 @@ static struct mlx5e_ethtool_table *get_flow_table(struct mlx5e_priv *priv, table_size = min_t(u32, BIT(MLX5_CAP_FLOWTABLE(priv->mdev, flow_table_properties_nic_receive.log_max_ft_size)), MLX5E_ETHTOOL_NUM_ENTRIES); - ft = mlx5_create_auto_grouped_flow_table(ns, prio, - table_size, - MLX5E_ETHTOOL_NUM_GROUPS, 0, 0); + + ft_attr.prio = prio; + ft_attr.max_fte = table_size; + ft_attr.autogroup.max_num_groups = MLX5E_ETHTOOL_NUM_GROUPS; + ft = mlx5_create_auto_grouped_flow_table(ns, &ft_attr); if (IS_ERR(ft)) return (void *)ft; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c index 319b39f25592..78737fd42616 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c @@ -4739,17 +4739,19 @@ void mlx5e_build_rss_params(struct mlx5e_rss_params *rss_params, tirc_default_config[tt].rx_hash_fields; } -void mlx5e_build_nic_params(struct mlx5_core_dev *mdev, +void mlx5e_build_nic_params(struct mlx5e_priv *priv, struct mlx5e_xsk *xsk, struct mlx5e_rss_params *rss_params, struct mlx5e_params *params, - u16 max_channels, u16 mtu) + u16 mtu) { + struct mlx5_core_dev *mdev = priv->mdev; u8 rx_cq_period_mode; params->sw_mtu = mtu; params->hard_mtu = MLX5E_ETH_HARD_MTU; - params->num_channels = max_channels; + params->num_channels = min_t(unsigned int, MLX5E_MAX_NUM_CHANNELS / 2, + priv->max_nch); params->num_tc = 1; /* SQ */ @@ -4986,8 +4988,8 @@ static int mlx5e_nic_init(struct mlx5_core_dev *mdev, if (err) return err; - mlx5e_build_nic_params(mdev, &priv->xsk, rss, &priv->channels.params, - priv->max_nch, netdev->mtu); + mlx5e_build_nic_params(priv, &priv->xsk, rss, &priv->channels.params, + netdev->mtu); mlx5e_timestamp_init(priv); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c index f175cb24bb67..446eb4d6c983 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c @@ -41,6 +41,7 @@ #include <net/ipv6_stubs.h> #include "eswitch.h" +#include "eswitch_offloads_chains.h" #include "en.h" #include "en_rep.h" #include "en_tc.h" @@ -1247,8 +1248,7 @@ static int mlx5e_rep_setup_tc_cb(enum tc_setup_type type, void *type_data, static int mlx5e_rep_setup_ft_cb(enum tc_setup_type type, void *type_data, void *cb_priv) { - struct flow_cls_offload *f = type_data; - struct flow_cls_offload cls_flower; + struct flow_cls_offload tmp, *f = type_data; struct mlx5e_priv *priv = cb_priv; struct mlx5_eswitch *esw; unsigned long flags; @@ -1261,16 +1261,30 @@ static int mlx5e_rep_setup_ft_cb(enum tc_setup_type type, void *type_data, switch (type) { case TC_SETUP_CLSFLOWER: - if (!mlx5_eswitch_prios_supported(esw) || f->common.chain_index) + memcpy(&tmp, f, sizeof(*f)); + + if (!mlx5_esw_chains_prios_supported(esw) || + tmp.common.chain_index) return -EOPNOTSUPP; /* Re-use tc offload path by moving the ft flow to the * reserved ft chain. + * + * FT offload can use prio range [0, INT_MAX], so we normalize + * it to range [1, mlx5_esw_chains_get_prio_range(esw)] + * as with tc, where prio 0 isn't supported. + * + * We only support chain 0 of FT offload. */ - memcpy(&cls_flower, f, sizeof(*f)); - cls_flower.common.chain_index = FDB_FT_CHAIN; - err = mlx5e_rep_setup_tc_cls_flower(priv, &cls_flower, flags); - memcpy(&f->stats, &cls_flower.stats, sizeof(f->stats)); + if (tmp.common.prio >= mlx5_esw_chains_get_prio_range(esw)) + return -EOPNOTSUPP; + if (tmp.common.chain_index != 0) + return -EOPNOTSUPP; + + tmp.common.chain_index = mlx5_esw_chains_get_ft_chain(esw); + tmp.common.prio++; + err = mlx5e_rep_setup_tc_cls_flower(priv, &tmp, flags); + memcpy(&f->stats, &tmp.stats, sizeof(f->stats)); return err; default: return -EOPNOTSUPP; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c index 9f09253f9f46..4291db78efc9 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c @@ -297,6 +297,9 @@ static void mlx5e_grp_sw_update_stats(struct mlx5e_priv *priv) s->tx_tls_drop_bypass_req += sq_stats->tls_drop_bypass_req; #endif s->tx_cqes += sq_stats->cqes; + + /* https://gcc.gnu.org/bugzilla/show_bug.cgi?id=92657 */ + barrier(); } } } @@ -1130,6 +1133,7 @@ static void mlx5e_grp_per_port_buffer_congest_update_stats(struct mlx5e_priv *pr static const struct counter_desc pport_per_prio_traffic_stats_desc[] = { { "rx_prio%d_bytes", PPORT_PER_PRIO_OFF(rx_octets) }, { "rx_prio%d_packets", PPORT_PER_PRIO_OFF(rx_frames) }, + { "rx_prio%d_discards", PPORT_PER_PRIO_OFF(rx_discards) }, { "tx_prio%d_bytes", PPORT_PER_PRIO_OFF(tx_octets) }, { "tx_prio%d_packets", PPORT_PER_PRIO_OFF(tx_frames) }, }; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c index 9b32a9c0f497..26f559b453dc 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c @@ -51,6 +51,7 @@ #include "en_rep.h" #include "en_tc.h" #include "eswitch.h" +#include "eswitch_offloads_chains.h" #include "fs_core.h" #include "en/port.h" #include "en/tc_tun.h" @@ -592,7 +593,7 @@ static void mlx5e_hairpin_set_ttc_params(struct mlx5e_hairpin *hp, for (tt = 0; tt < MLX5E_NUM_INDIR_TIRS; tt++) ttc_params->indir_tirn[tt] = hp->indir_tirn[tt]; - ft_attr->max_fte = MLX5E_NUM_TT; + ft_attr->max_fte = MLX5E_TTC_TABLE_SIZE; ft_attr->level = MLX5E_TC_TTC_FT_LEVEL; ft_attr->prio = MLX5E_TC_PRIO; } @@ -960,7 +961,8 @@ mlx5e_tc_add_nic_flow(struct mlx5e_priv *priv, mutex_lock(&priv->fs.tc.t_lock); if (IS_ERR_OR_NULL(priv->fs.tc.t)) { - int tc_grp_size, tc_tbl_size; + struct mlx5_flow_table_attr ft_attr = {}; + int tc_grp_size, tc_tbl_size, tc_num_grps; u32 max_flow_counter; max_flow_counter = (MLX5_CAP_GEN(dev, max_flow_counter_31_16) << 16) | @@ -970,13 +972,15 @@ mlx5e_tc_add_nic_flow(struct mlx5e_priv *priv, tc_tbl_size = min_t(int, tc_grp_size * MLX5E_TC_TABLE_NUM_GROUPS, BIT(MLX5_CAP_FLOWTABLE_NIC_RX(dev, log_max_ft_size))); + tc_num_grps = MLX5E_TC_TABLE_NUM_GROUPS; + ft_attr.prio = MLX5E_TC_PRIO; + ft_attr.max_fte = tc_tbl_size; + ft_attr.level = MLX5E_TC_FT_LEVEL; + ft_attr.autogroup.max_num_groups = tc_num_grps; priv->fs.tc.t = mlx5_create_auto_grouped_flow_table(priv->fs.ns, - MLX5E_TC_PRIO, - tc_tbl_size, - MLX5E_TC_TABLE_NUM_GROUPS, - MLX5E_TC_FT_LEVEL, 0); + &ft_attr); if (IS_ERR(priv->fs.tc.t)) { mutex_unlock(&priv->fs.tc.t_lock); NL_SET_ERR_MSG_MOD(extack, @@ -1080,7 +1084,7 @@ mlx5e_tc_offload_to_slow_path(struct mlx5_eswitch *esw, memcpy(slow_attr, flow->esw_attr, sizeof(*slow_attr)); slow_attr->action = MLX5_FLOW_CONTEXT_ACTION_FWD_DEST; slow_attr->split_count = 0; - slow_attr->dest_chain = FDB_TC_SLOW_PATH_CHAIN; + slow_attr->flags |= MLX5_ESW_ATTR_FLAG_SLOW_PATH; rule = mlx5e_tc_offload_fdb_rules(esw, flow, spec, slow_attr); if (!IS_ERR(rule)) @@ -1097,7 +1101,7 @@ mlx5e_tc_unoffload_from_slow_path(struct mlx5_eswitch *esw, memcpy(slow_attr, flow->esw_attr, sizeof(*slow_attr)); slow_attr->action = MLX5_FLOW_CONTEXT_ACTION_FWD_DEST; slow_attr->split_count = 0; - slow_attr->dest_chain = FDB_TC_SLOW_PATH_CHAIN; + slow_attr->flags |= MLX5_ESW_ATTR_FLAG_SLOW_PATH; mlx5e_tc_unoffload_fdb_rules(esw, flow, slow_attr); flow_flag_clear(flow, SLOW); } @@ -1157,19 +1161,18 @@ mlx5e_tc_add_fdb_flow(struct mlx5e_priv *priv, struct netlink_ext_ack *extack) { struct mlx5_eswitch *esw = priv->mdev->priv.eswitch; - u32 max_chain = mlx5_eswitch_get_chain_range(esw); struct mlx5_esw_flow_attr *attr = flow->esw_attr; struct mlx5e_tc_flow_parse_attr *parse_attr = attr->parse_attr; - u16 max_prio = mlx5_eswitch_get_prio_range(esw); struct net_device *out_dev, *encap_dev = NULL; struct mlx5_fc *counter = NULL; struct mlx5e_rep_priv *rpriv; struct mlx5e_priv *out_priv; bool encap_valid = true; + u32 max_prio, max_chain; int err = 0; int out_index; - if (!mlx5_eswitch_prios_supported(esw) && attr->prio != 1) { + if (!mlx5_esw_chains_prios_supported(esw) && attr->prio != 1) { NL_SET_ERR_MSG(extack, "E-switch priorities unsupported, upgrade FW"); return -EOPNOTSUPP; } @@ -1179,11 +1182,13 @@ mlx5e_tc_add_fdb_flow(struct mlx5e_priv *priv, * FDB_FT_CHAIN which is outside tc range. * See mlx5e_rep_setup_ft_cb(). */ + max_chain = mlx5_esw_chains_get_chain_range(esw); if (!mlx5e_is_ft_flow(flow) && attr->chain > max_chain) { NL_SET_ERR_MSG(extack, "Requested chain is out of supported range"); return -EOPNOTSUPP; } + max_prio = mlx5_esw_chains_get_prio_range(esw); if (attr->prio > max_prio) { NL_SET_ERR_MSG(extack, "Requested priority is out of supported range"); return -EOPNOTSUPP; @@ -2842,6 +2847,10 @@ static int parse_tc_nic_actions(struct mlx5e_priv *priv, flow_action_for_each(i, act, flow_action) { switch (act->id) { + case FLOW_ACTION_ACCEPT: + action |= MLX5_FLOW_CONTEXT_ACTION_FWD_DEST | + MLX5_FLOW_CONTEXT_ACTION_COUNT; + break; case FLOW_ACTION_DROP: action |= MLX5_FLOW_CONTEXT_ACTION_DROP; if (MLX5_CAP_FLOWTABLE(priv->mdev, @@ -2999,6 +3008,25 @@ static struct ip_tunnel_info *dup_tun_info(const struct ip_tunnel_info *tun_info return kmemdup(tun_info, tun_size, GFP_KERNEL); } +static bool is_duplicated_encap_entry(struct mlx5e_priv *priv, + struct mlx5e_tc_flow *flow, + int out_index, + struct mlx5e_encap_entry *e, + struct netlink_ext_ack *extack) +{ + int i; + + for (i = 0; i < out_index; i++) { + if (flow->encaps[i].e != e) + continue; + NL_SET_ERR_MSG_MOD(extack, "can't duplicate encap action"); + netdev_err(priv->netdev, "can't duplicate encap action\n"); + return true; + } + + return false; +} + static int mlx5e_attach_encap(struct mlx5e_priv *priv, struct mlx5e_tc_flow *flow, struct net_device *mirred_dev, @@ -3034,6 +3062,12 @@ static int mlx5e_attach_encap(struct mlx5e_priv *priv, /* must verify if encap is valid or not */ if (e) { + /* Check that entry was not already attached to this flow */ + if (is_duplicated_encap_entry(priv, flow, out_index, e, extack)) { + err = -EOPNOTSUPP; + goto out_err; + } + mutex_unlock(&esw->offloads.encap_tbl_lock); wait_for_completion(&e->res_ready); @@ -3220,6 +3254,26 @@ bool mlx5e_is_valid_eswitch_fwd_dev(struct mlx5e_priv *priv, same_hw_devs(priv, netdev_priv(out_dev)); } +static bool is_duplicated_output_device(struct net_device *dev, + struct net_device *out_dev, + int *ifindexes, int if_count, + struct netlink_ext_ack *extack) +{ + int i; + + for (i = 0; i < if_count; i++) { + if (ifindexes[i] == out_dev->ifindex) { + NL_SET_ERR_MSG_MOD(extack, + "can't duplicate output to same device"); + netdev_err(dev, "can't duplicate output to same device: %s\n", + out_dev->name); + return true; + } + } + + return false; +} + static int parse_tc_fdb_actions(struct mlx5e_priv *priv, struct flow_action *flow_action, struct mlx5e_tc_flow *flow, @@ -3231,11 +3285,12 @@ static int parse_tc_fdb_actions(struct mlx5e_priv *priv, struct mlx5e_tc_flow_parse_attr *parse_attr = attr->parse_attr; struct mlx5e_rep_priv *rpriv = priv->ppriv; const struct ip_tunnel_info *info = NULL; + int ifindexes[MLX5_MAX_FLOW_FWD_VPORTS]; bool ft_flow = mlx5e_is_ft_flow(flow); const struct flow_action_entry *act; + int err, i, if_count = 0; bool encap = false; u32 action = 0; - int err, i; if (!flow_action_has_entries(flow_action)) return -EINVAL; @@ -3312,6 +3367,16 @@ static int parse_tc_fdb_actions(struct mlx5e_priv *priv, struct net_device *uplink_dev = mlx5_eswitch_uplink_get_proto_dev(esw, REP_ETH); struct net_device *uplink_upper; + if (is_duplicated_output_device(priv->netdev, + out_dev, + ifindexes, + if_count, + extack)) + return -EOPNOTSUPP; + + ifindexes[if_count] = out_dev->ifindex; + if_count++; + rcu_read_lock(); uplink_upper = netdev_master_upper_dev_get_rcu(uplink_dev); @@ -3406,7 +3471,7 @@ static int parse_tc_fdb_actions(struct mlx5e_priv *priv, break; case FLOW_ACTION_GOTO: { u32 dest_chain = act->chain_index; - u32 max_chain = mlx5_eswitch_get_chain_range(esw); + u32 max_chain = mlx5_esw_chains_get_chain_range(esw); if (ft_flow) { NL_SET_ERR_MSG_MOD(extack, "Goto action is not supported"); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eq.c b/drivers/net/ethernet/mellanox/mlx5/core/eq.c index 580c71cb9dfa..cccea3a8eddd 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eq.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/eq.c @@ -156,7 +156,8 @@ static int mlx5_eq_comp_int(struct notifier_block *nb, cq->comp(cq, eqe); mlx5_cq_put(cq); } else { - mlx5_core_warn(eq->dev, "Completion event for bogus CQ 0x%x\n", cqn); + dev_dbg_ratelimited(eq->dev->device, + "Completion event for bogus CQ 0x%x\n", cqn); } ++eq->cons_index; @@ -563,6 +564,39 @@ static void gather_async_events_mask(struct mlx5_core_dev *dev, u64 mask[4]) gather_user_async_events(dev, mask); } +static int +setup_async_eq(struct mlx5_core_dev *dev, struct mlx5_eq_async *eq, + struct mlx5_eq_param *param, const char *name) +{ + int err; + + eq->irq_nb.notifier_call = mlx5_eq_async_int; + + err = create_async_eq(dev, &eq->core, param); + if (err) { + mlx5_core_warn(dev, "failed to create %s EQ %d\n", name, err); + return err; + } + err = mlx5_eq_enable(dev, &eq->core, &eq->irq_nb); + if (err) { + mlx5_core_warn(dev, "failed to enable %s EQ %d\n", name, err); + destroy_async_eq(dev, &eq->core); + } + return err; +} + +static void cleanup_async_eq(struct mlx5_core_dev *dev, + struct mlx5_eq_async *eq, const char *name) +{ + int err; + + mlx5_eq_disable(dev, &eq->core, &eq->irq_nb); + err = destroy_async_eq(dev, &eq->core); + if (err) + mlx5_core_err(dev, "failed to destroy %s eq, err(%d)\n", + name, err); +} + static int create_async_eqs(struct mlx5_core_dev *dev) { struct mlx5_eq_table *table = dev->priv.eq_table; @@ -572,77 +606,45 @@ static int create_async_eqs(struct mlx5_core_dev *dev) MLX5_NB_INIT(&table->cq_err_nb, cq_err_event_notifier, CQ_ERROR); mlx5_eq_notifier_register(dev, &table->cq_err_nb); - table->cmd_eq.irq_nb.notifier_call = mlx5_eq_async_int; param = (struct mlx5_eq_param) { .irq_index = 0, .nent = MLX5_NUM_CMD_EQE, + .mask[0] = 1ull << MLX5_EVENT_TYPE_CMD, }; - - param.mask[0] = 1ull << MLX5_EVENT_TYPE_CMD; - err = create_async_eq(dev, &table->cmd_eq.core, ¶m); - if (err) { - mlx5_core_warn(dev, "failed to create cmd EQ %d\n", err); - goto err0; - } - err = mlx5_eq_enable(dev, &table->cmd_eq.core, &table->cmd_eq.irq_nb); - if (err) { - mlx5_core_warn(dev, "failed to enable cmd EQ %d\n", err); + err = setup_async_eq(dev, &table->cmd_eq, ¶m, "cmd"); + if (err) goto err1; - } + mlx5_cmd_use_events(dev); - table->async_eq.irq_nb.notifier_call = mlx5_eq_async_int; param = (struct mlx5_eq_param) { .irq_index = 0, .nent = MLX5_NUM_ASYNC_EQE, }; gather_async_events_mask(dev, param.mask); - err = create_async_eq(dev, &table->async_eq.core, ¶m); - if (err) { - mlx5_core_warn(dev, "failed to create async EQ %d\n", err); + err = setup_async_eq(dev, &table->async_eq, ¶m, "async"); + if (err) goto err2; - } - err = mlx5_eq_enable(dev, &table->async_eq.core, - &table->async_eq.irq_nb); - if (err) { - mlx5_core_warn(dev, "failed to enable async EQ %d\n", err); - goto err3; - } - table->pages_eq.irq_nb.notifier_call = mlx5_eq_async_int; param = (struct mlx5_eq_param) { .irq_index = 0, .nent = /* TODO: sriov max_vf + */ 1, + .mask[0] = 1ull << MLX5_EVENT_TYPE_PAGE_REQUEST, }; - param.mask[0] = 1ull << MLX5_EVENT_TYPE_PAGE_REQUEST; - err = create_async_eq(dev, &table->pages_eq.core, ¶m); - if (err) { - mlx5_core_warn(dev, "failed to create pages EQ %d\n", err); - goto err4; - } - err = mlx5_eq_enable(dev, &table->pages_eq.core, - &table->pages_eq.irq_nb); - if (err) { - mlx5_core_warn(dev, "failed to enable pages EQ %d\n", err); - goto err5; - } + err = setup_async_eq(dev, &table->pages_eq, ¶m, "pages"); + if (err) + goto err3; - return err; + return 0; -err5: - destroy_async_eq(dev, &table->pages_eq.core); -err4: - mlx5_eq_disable(dev, &table->async_eq.core, &table->async_eq.irq_nb); err3: - destroy_async_eq(dev, &table->async_eq.core); + cleanup_async_eq(dev, &table->async_eq, "async"); err2: mlx5_cmd_use_polling(dev); - mlx5_eq_disable(dev, &table->cmd_eq.core, &table->cmd_eq.irq_nb); + cleanup_async_eq(dev, &table->cmd_eq, "cmd"); err1: - destroy_async_eq(dev, &table->cmd_eq.core); -err0: mlx5_eq_notifier_unregister(dev, &table->cq_err_nb); return err; } @@ -650,28 +652,11 @@ err0: static void destroy_async_eqs(struct mlx5_core_dev *dev) { struct mlx5_eq_table *table = dev->priv.eq_table; - int err; - - mlx5_eq_disable(dev, &table->pages_eq.core, &table->pages_eq.irq_nb); - err = destroy_async_eq(dev, &table->pages_eq.core); - if (err) - mlx5_core_err(dev, "failed to destroy pages eq, err(%d)\n", - err); - - mlx5_eq_disable(dev, &table->async_eq.core, &table->async_eq.irq_nb); - err = destroy_async_eq(dev, &table->async_eq.core); - if (err) - mlx5_core_err(dev, "failed to destroy async eq, err(%d)\n", - err); + cleanup_async_eq(dev, &table->pages_eq, "pages"); + cleanup_async_eq(dev, &table->async_eq, "async"); mlx5_cmd_use_polling(dev); - - mlx5_eq_disable(dev, &table->cmd_eq.core, &table->cmd_eq.irq_nb); - err = destroy_async_eq(dev, &table->cmd_eq.core); - if (err) - mlx5_core_err(dev, "failed to destroy command eq, err(%d)\n", - err); - + cleanup_async_eq(dev, &table->cmd_eq, "cmd"); mlx5_eq_notifier_unregister(dev, &table->cq_err_nb); } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c index 2c965ad0d744..05b13a1e829c 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c @@ -277,6 +277,7 @@ enum { static int esw_create_legacy_vepa_table(struct mlx5_eswitch *esw) { + struct mlx5_flow_table_attr ft_attr = {}; struct mlx5_core_dev *dev = esw->dev; struct mlx5_flow_namespace *root_ns; struct mlx5_flow_table *fdb; @@ -289,8 +290,10 @@ static int esw_create_legacy_vepa_table(struct mlx5_eswitch *esw) } /* num FTE 2, num FG 2 */ - fdb = mlx5_create_auto_grouped_flow_table(root_ns, LEGACY_VEPA_PRIO, - 2, 2, 0, 0); + ft_attr.prio = LEGACY_VEPA_PRIO; + ft_attr.max_fte = 2; + ft_attr.autogroup.max_num_groups = 2; + fdb = mlx5_create_auto_grouped_flow_table(root_ns, &ft_attr); if (IS_ERR(fdb)) { err = PTR_ERR(fdb); esw_warn(dev, "Failed to create VEPA FDB err %d\n", err); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h index ffcff3ba3701..4472710ccc9c 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h @@ -157,7 +157,7 @@ enum offloads_fdb_flags { ESW_FDB_CHAINS_AND_PRIOS_SUPPORTED = BIT(0), }; -extern const unsigned int ESW_POOLS[4]; +struct mlx5_esw_chains_priv; struct mlx5_eswitch_fdb { union { @@ -182,14 +182,7 @@ struct mlx5_eswitch_fdb { struct mlx5_flow_handle *miss_rule_multi; int vlan_push_pop_refcount; - struct { - struct mlx5_flow_table *fdb; - u32 num_rules; - } fdb_prio[FDB_NUM_CHAINS][FDB_TC_MAX_PRIO + 1][FDB_TC_LEVELS_PER_PRIO]; - /* Protects fdb_prio table */ - struct mutex fdb_prio_lock; - - int fdb_left[ARRAY_SIZE(ESW_POOLS)]; + struct mlx5_esw_chains_priv *esw_chains_priv; } offloads; }; u32 flags; @@ -355,15 +348,6 @@ mlx5_eswitch_del_fwd_rule(struct mlx5_eswitch *esw, struct mlx5_flow_handle *rule, struct mlx5_esw_flow_attr *attr); -bool -mlx5_eswitch_prios_supported(struct mlx5_eswitch *esw); - -u16 -mlx5_eswitch_get_prio_range(struct mlx5_eswitch *esw); - -u32 -mlx5_eswitch_get_chain_range(struct mlx5_eswitch *esw); - struct mlx5_flow_handle * mlx5_eswitch_create_vport_rx_rule(struct mlx5_eswitch *esw, u16 vport, struct mlx5_flow_destination *dest); @@ -388,6 +372,11 @@ enum { MLX5_ESW_DEST_ENCAP_VALID = BIT(1), }; +enum { + MLX5_ESW_ATTR_FLAG_VLAN_HANDLED = BIT(0), + MLX5_ESW_ATTR_FLAG_SLOW_PATH = BIT(1), +}; + struct mlx5_esw_flow_attr { struct mlx5_eswitch_rep *in_rep; struct mlx5_core_dev *in_mdev; @@ -401,7 +390,6 @@ struct mlx5_esw_flow_attr { u16 vlan_vid[MLX5_FS_VLAN_DEPTH]; u8 vlan_prio[MLX5_FS_VLAN_DEPTH]; u8 total_vlan; - bool vlan_handled; struct { u32 flags; struct mlx5_eswitch_rep *rep; @@ -416,6 +404,7 @@ struct mlx5_esw_flow_attr { u32 chain; u16 prio; u32 dest_chain; + u32 flags; struct mlx5e_tc_flow_parse_attr *parse_attr; }; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c index 243a5440867e..a6d0b62ef234 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c @@ -37,6 +37,7 @@ #include <linux/mlx5/fs.h> #include "mlx5_core.h" #include "eswitch.h" +#include "eswitch_offloads_chains.h" #include "rdma.h" #include "en.h" #include "fs_core.h" @@ -47,10 +48,6 @@ * one for multicast. */ #define MLX5_ESW_MISS_FLOWS (2) - -#define fdb_prio_table(esw, chain, prio, level) \ - (esw)->fdb_table.offloads.fdb_prio[(chain)][(prio)][(level)] - #define UPLINK_REP_INDEX 0 static struct mlx5_eswitch_rep *mlx5_eswitch_get_rep(struct mlx5_eswitch *esw, @@ -62,32 +59,6 @@ static struct mlx5_eswitch_rep *mlx5_eswitch_get_rep(struct mlx5_eswitch *esw, return &esw->offloads.vport_reps[idx]; } -static struct mlx5_flow_table * -esw_get_prio_table(struct mlx5_eswitch *esw, u32 chain, u16 prio, int level); -static void -esw_put_prio_table(struct mlx5_eswitch *esw, u32 chain, u16 prio, int level); - -bool mlx5_eswitch_prios_supported(struct mlx5_eswitch *esw) -{ - return (!!(esw->fdb_table.flags & ESW_FDB_CHAINS_AND_PRIOS_SUPPORTED)); -} - -u32 mlx5_eswitch_get_chain_range(struct mlx5_eswitch *esw) -{ - if (esw->fdb_table.flags & ESW_FDB_CHAINS_AND_PRIOS_SUPPORTED) - return FDB_TC_MAX_CHAIN; - - return 0; -} - -u16 mlx5_eswitch_get_prio_range(struct mlx5_eswitch *esw) -{ - if (esw->fdb_table.flags & ESW_FDB_CHAINS_AND_PRIOS_SUPPORTED) - return FDB_TC_MAX_PRIO; - - return 1; -} - static bool esw_check_ingress_prio_tag_enabled(const struct mlx5_eswitch *esw, const struct mlx5_vport *vport) @@ -175,10 +146,17 @@ mlx5_eswitch_add_offloaded_rule(struct mlx5_eswitch *esw, } if (flow_act.action & MLX5_FLOW_CONTEXT_ACTION_FWD_DEST) { - if (attr->dest_chain) { - struct mlx5_flow_table *ft; + struct mlx5_flow_table *ft; - ft = esw_get_prio_table(esw, attr->dest_chain, 1, 0); + if (attr->flags & MLX5_ESW_ATTR_FLAG_SLOW_PATH) { + flow_act.flags |= FLOW_ACT_IGNORE_FLOW_LEVEL; + dest[i].type = MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE; + dest[i].ft = mlx5_esw_chains_get_tc_end_ft(esw); + i++; + } else if (attr->dest_chain) { + flow_act.flags |= FLOW_ACT_IGNORE_FLOW_LEVEL; + ft = mlx5_esw_chains_get_table(esw, attr->dest_chain, + 1, 0); if (IS_ERR(ft)) { rule = ERR_CAST(ft); goto err_create_goto_table; @@ -223,7 +201,8 @@ mlx5_eswitch_add_offloaded_rule(struct mlx5_eswitch *esw, if (flow_act.action & MLX5_FLOW_CONTEXT_ACTION_MOD_HDR) flow_act.modify_hdr = attr->modify_hdr; - fdb = esw_get_prio_table(esw, attr->chain, attr->prio, !!split); + fdb = mlx5_esw_chains_get_table(esw, attr->chain, attr->prio, + !!split); if (IS_ERR(fdb)) { rule = ERR_CAST(fdb); goto err_esw_get; @@ -242,10 +221,10 @@ mlx5_eswitch_add_offloaded_rule(struct mlx5_eswitch *esw, return rule; err_add_rule: - esw_put_prio_table(esw, attr->chain, attr->prio, !!split); + mlx5_esw_chains_put_table(esw, attr->chain, attr->prio, !!split); err_esw_get: - if (attr->dest_chain) - esw_put_prio_table(esw, attr->dest_chain, 1, 0); + if (!(attr->flags & MLX5_ESW_ATTR_FLAG_SLOW_PATH) && attr->dest_chain) + mlx5_esw_chains_put_table(esw, attr->dest_chain, 1, 0); err_create_goto_table: return rule; } @@ -262,13 +241,13 @@ mlx5_eswitch_add_fwd_rule(struct mlx5_eswitch *esw, struct mlx5_flow_handle *rule; int i; - fast_fdb = esw_get_prio_table(esw, attr->chain, attr->prio, 0); + fast_fdb = mlx5_esw_chains_get_table(esw, attr->chain, attr->prio, 0); if (IS_ERR(fast_fdb)) { rule = ERR_CAST(fast_fdb); goto err_get_fast; } - fwd_fdb = esw_get_prio_table(esw, attr->chain, attr->prio, 1); + fwd_fdb = mlx5_esw_chains_get_table(esw, attr->chain, attr->prio, 1); if (IS_ERR(fwd_fdb)) { rule = ERR_CAST(fwd_fdb); goto err_get_fwd; @@ -296,6 +275,7 @@ mlx5_eswitch_add_fwd_rule(struct mlx5_eswitch *esw, if (attr->outer_match_level != MLX5_MATCH_NONE) spec->match_criteria_enable |= MLX5_MATCH_OUTER_HEADERS; + flow_act.flags |= FLOW_ACT_IGNORE_FLOW_LEVEL; rule = mlx5_add_flow_rules(fast_fdb, spec, &flow_act, dest, i); if (IS_ERR(rule)) @@ -305,9 +285,9 @@ mlx5_eswitch_add_fwd_rule(struct mlx5_eswitch *esw, return rule; add_err: - esw_put_prio_table(esw, attr->chain, attr->prio, 1); + mlx5_esw_chains_put_table(esw, attr->chain, attr->prio, 1); err_get_fwd: - esw_put_prio_table(esw, attr->chain, attr->prio, 0); + mlx5_esw_chains_put_table(esw, attr->chain, attr->prio, 0); err_get_fast: return rule; } @@ -332,12 +312,13 @@ __mlx5_eswitch_del_rule(struct mlx5_eswitch *esw, atomic64_dec(&esw->offloads.num_flows); if (fwd_rule) { - esw_put_prio_table(esw, attr->chain, attr->prio, 1); - esw_put_prio_table(esw, attr->chain, attr->prio, 0); + mlx5_esw_chains_put_table(esw, attr->chain, attr->prio, 1); + mlx5_esw_chains_put_table(esw, attr->chain, attr->prio, 0); } else { - esw_put_prio_table(esw, attr->chain, attr->prio, !!split); + mlx5_esw_chains_put_table(esw, attr->chain, attr->prio, + !!split); if (attr->dest_chain) - esw_put_prio_table(esw, attr->dest_chain, 1, 0); + mlx5_esw_chains_put_table(esw, attr->dest_chain, 1, 0); } } @@ -451,7 +432,7 @@ int mlx5_eswitch_add_vlan_action(struct mlx5_eswitch *esw, if (err) goto unlock; - attr->vlan_handled = false; + attr->flags &= ~MLX5_ESW_ATTR_FLAG_VLAN_HANDLED; vport = esw_vlan_action_get_vport(attr, push, pop); @@ -459,7 +440,7 @@ int mlx5_eswitch_add_vlan_action(struct mlx5_eswitch *esw, /* tracks VF --> wire rules without vlan push action */ if (attr->dests[0].rep->vport == MLX5_VPORT_UPLINK) { vport->vlan_refcount++; - attr->vlan_handled = true; + attr->flags |= MLX5_ESW_ATTR_FLAG_VLAN_HANDLED; } goto unlock; @@ -490,7 +471,7 @@ skip_set_push: } out: if (!err) - attr->vlan_handled = true; + attr->flags |= MLX5_ESW_ATTR_FLAG_VLAN_HANDLED; unlock: mutex_unlock(&esw->state_lock); return err; @@ -508,7 +489,7 @@ int mlx5_eswitch_del_vlan_action(struct mlx5_eswitch *esw, if (mlx5_eswitch_vlan_actions_supported(esw->dev, 1)) return 0; - if (!attr->vlan_handled) + if (!(attr->flags & MLX5_ESW_ATTR_FLAG_VLAN_HANDLED)) return 0; push = !!(attr->action & MLX5_FLOW_CONTEXT_ACTION_VLAN_PUSH); @@ -582,8 +563,8 @@ mlx5_eswitch_add_send_to_vport_rule(struct mlx5_eswitch *esw, u16 vport, dest.vport.num = vport; flow_act.action = MLX5_FLOW_CONTEXT_ACTION_FWD_DEST; - flow_rule = mlx5_add_flow_rules(esw->fdb_table.offloads.slow_fdb, spec, - &flow_act, &dest, 1); + flow_rule = mlx5_add_flow_rules(esw->fdb_table.offloads.slow_fdb, + spec, &flow_act, &dest, 1); if (IS_ERR(flow_rule)) esw_warn(esw->dev, "FDB: Failed to add send to vport rule err %ld\n", PTR_ERR(flow_rule)); out: @@ -824,8 +805,8 @@ static int esw_add_fdb_miss_rule(struct mlx5_eswitch *esw) dest.vport.num = esw->manager_vport; flow_act.action = MLX5_FLOW_CONTEXT_ACTION_FWD_DEST; - flow_rule = mlx5_add_flow_rules(esw->fdb_table.offloads.slow_fdb, spec, - &flow_act, &dest, 1); + flow_rule = mlx5_add_flow_rules(esw->fdb_table.offloads.slow_fdb, + spec, &flow_act, &dest, 1); if (IS_ERR(flow_rule)) { err = PTR_ERR(flow_rule); esw_warn(esw->dev, "FDB: Failed to add unicast miss flow rule err %d\n", err); @@ -839,8 +820,8 @@ static int esw_add_fdb_miss_rule(struct mlx5_eswitch *esw) dmac_v = MLX5_ADDR_OF(fte_match_param, headers_v, outer_headers.dmac_47_16); dmac_v[0] = 0x01; - flow_rule = mlx5_add_flow_rules(esw->fdb_table.offloads.slow_fdb, spec, - &flow_act, &dest, 1); + flow_rule = mlx5_add_flow_rules(esw->fdb_table.offloads.slow_fdb, + spec, &flow_act, &dest, 1); if (IS_ERR(flow_rule)) { err = PTR_ERR(flow_rule); esw_warn(esw->dev, "FDB: Failed to add multicast miss flow rule err %d\n", err); @@ -855,174 +836,6 @@ out: return err; } -#define ESW_OFFLOADS_NUM_GROUPS 4 - -/* Firmware currently has 4 pool of 4 sizes that it supports (ESW_POOLS), - * and a virtual memory region of 16M (ESW_SIZE), this region is duplicated - * for each flow table pool. We can allocate up to 16M of each pool, - * and we keep track of how much we used via put/get_sz_to_pool. - * Firmware doesn't report any of this for now. - * ESW_POOL is expected to be sorted from large to small - */ -#define ESW_SIZE (16 * 1024 * 1024) -const unsigned int ESW_POOLS[4] = { 4 * 1024 * 1024, 1 * 1024 * 1024, - 64 * 1024, 4 * 1024 }; - -static int -get_sz_from_pool(struct mlx5_eswitch *esw) -{ - int sz = 0, i; - - for (i = 0; i < ARRAY_SIZE(ESW_POOLS); i++) { - if (esw->fdb_table.offloads.fdb_left[i]) { - --esw->fdb_table.offloads.fdb_left[i]; - sz = ESW_POOLS[i]; - break; - } - } - - return sz; -} - -static void -put_sz_to_pool(struct mlx5_eswitch *esw, int sz) -{ - int i; - - for (i = 0; i < ARRAY_SIZE(ESW_POOLS); i++) { - if (sz >= ESW_POOLS[i]) { - ++esw->fdb_table.offloads.fdb_left[i]; - break; - } - } -} - -static struct mlx5_flow_table * -create_next_size_table(struct mlx5_eswitch *esw, - struct mlx5_flow_namespace *ns, - u16 table_prio, - int level, - u32 flags) -{ - struct mlx5_flow_table *fdb; - int sz; - - sz = get_sz_from_pool(esw); - if (!sz) - return ERR_PTR(-ENOSPC); - - fdb = mlx5_create_auto_grouped_flow_table(ns, - table_prio, - sz, - ESW_OFFLOADS_NUM_GROUPS, - level, - flags); - if (IS_ERR(fdb)) { - esw_warn(esw->dev, "Failed to create FDB Table err %d (table prio: %d, level: %d, size: %d)\n", - (int)PTR_ERR(fdb), table_prio, level, sz); - put_sz_to_pool(esw, sz); - } - - return fdb; -} - -static struct mlx5_flow_table * -esw_get_prio_table(struct mlx5_eswitch *esw, u32 chain, u16 prio, int level) -{ - struct mlx5_core_dev *dev = esw->dev; - struct mlx5_flow_table *fdb = NULL; - struct mlx5_flow_namespace *ns; - int table_prio, l = 0; - u32 flags = 0; - - if (chain == FDB_TC_SLOW_PATH_CHAIN) - return esw->fdb_table.offloads.slow_fdb; - - mutex_lock(&esw->fdb_table.offloads.fdb_prio_lock); - - fdb = fdb_prio_table(esw, chain, prio, level).fdb; - if (fdb) { - /* take ref on earlier levels as well */ - while (level >= 0) - fdb_prio_table(esw, chain, prio, level--).num_rules++; - mutex_unlock(&esw->fdb_table.offloads.fdb_prio_lock); - return fdb; - } - - ns = mlx5_get_fdb_sub_ns(dev, chain); - if (!ns) { - esw_warn(dev, "Failed to get FDB sub namespace\n"); - mutex_unlock(&esw->fdb_table.offloads.fdb_prio_lock); - return ERR_PTR(-EOPNOTSUPP); - } - - if (esw->offloads.encap != DEVLINK_ESWITCH_ENCAP_MODE_NONE) - flags |= (MLX5_FLOW_TABLE_TUNNEL_EN_REFORMAT | - MLX5_FLOW_TABLE_TUNNEL_EN_DECAP); - - table_prio = prio - 1; - - /* create earlier levels for correct fs_core lookup when - * connecting tables - */ - for (l = 0; l <= level; l++) { - if (fdb_prio_table(esw, chain, prio, l).fdb) { - fdb_prio_table(esw, chain, prio, l).num_rules++; - continue; - } - - fdb = create_next_size_table(esw, ns, table_prio, l, flags); - if (IS_ERR(fdb)) { - l--; - goto err_create_fdb; - } - - fdb_prio_table(esw, chain, prio, l).fdb = fdb; - fdb_prio_table(esw, chain, prio, l).num_rules = 1; - } - - mutex_unlock(&esw->fdb_table.offloads.fdb_prio_lock); - return fdb; - -err_create_fdb: - mutex_unlock(&esw->fdb_table.offloads.fdb_prio_lock); - if (l >= 0) - esw_put_prio_table(esw, chain, prio, l); - - return fdb; -} - -static void -esw_put_prio_table(struct mlx5_eswitch *esw, u32 chain, u16 prio, int level) -{ - int l; - - if (chain == FDB_TC_SLOW_PATH_CHAIN) - return; - - mutex_lock(&esw->fdb_table.offloads.fdb_prio_lock); - - for (l = level; l >= 0; l--) { - if (--(fdb_prio_table(esw, chain, prio, l).num_rules) > 0) - continue; - - put_sz_to_pool(esw, fdb_prio_table(esw, chain, prio, l).fdb->max_fte); - mlx5_destroy_flow_table(fdb_prio_table(esw, chain, prio, l).fdb); - fdb_prio_table(esw, chain, prio, l).fdb = NULL; - } - - mutex_unlock(&esw->fdb_table.offloads.fdb_prio_lock); -} - -static void esw_destroy_offloads_fast_fdb_tables(struct mlx5_eswitch *esw) -{ - /* If lazy creation isn't supported, deref the fast path tables */ - if (!(esw->fdb_table.flags & ESW_FDB_CHAINS_AND_PRIOS_SUPPORTED)) { - esw_put_prio_table(esw, 0, 1, 1); - esw_put_prio_table(esw, 0, 1, 0); - } -} - #define MAX_PF_SQ 256 #define MAX_SQ_NVPORTS 32 @@ -1055,16 +868,16 @@ static int esw_create_offloads_fdb_tables(struct mlx5_eswitch *esw, int nvports) int inlen = MLX5_ST_SZ_BYTES(create_flow_group_in); struct mlx5_flow_table_attr ft_attr = {}; struct mlx5_core_dev *dev = esw->dev; - u32 *flow_group_in, max_flow_counter; struct mlx5_flow_namespace *root_ns; struct mlx5_flow_table *fdb = NULL; - int table_size, ix, err = 0, i; + u32 flags = 0, *flow_group_in; + int table_size, ix, err = 0; struct mlx5_flow_group *g; - u32 flags = 0, fdb_max; void *match_criteria; u8 *dmac; esw_debug(esw->dev, "Create offloads FDB Tables\n"); + flow_group_in = kvzalloc(inlen, GFP_KERNEL); if (!flow_group_in) return -ENOMEM; @@ -1083,19 +896,6 @@ static int esw_create_offloads_fdb_tables(struct mlx5_eswitch *esw, int nvports) goto ns_err; } - max_flow_counter = (MLX5_CAP_GEN(dev, max_flow_counter_31_16) << 16) | - MLX5_CAP_GEN(dev, max_flow_counter_15_0); - fdb_max = 1 << MLX5_CAP_ESW_FLOWTABLE_FDB(dev, log_max_ft_size); - - esw_debug(dev, "Create offloads FDB table, min (max esw size(2^%d), max counters(%d), groups(%d), max flow table size(%d))\n", - MLX5_CAP_ESW_FLOWTABLE_FDB(dev, log_max_ft_size), - max_flow_counter, ESW_OFFLOADS_NUM_GROUPS, - fdb_max); - - for (i = 0; i < ARRAY_SIZE(ESW_POOLS); i++) - esw->fdb_table.offloads.fdb_left[i] = - ESW_POOLS[i] <= fdb_max ? ESW_SIZE / ESW_POOLS[i] : 0; - table_size = nvports * MAX_SQ_NVPORTS + MAX_PF_SQ + MLX5_ESW_MISS_FLOWS + esw->total_vports; @@ -1118,16 +918,10 @@ static int esw_create_offloads_fdb_tables(struct mlx5_eswitch *esw, int nvports) } esw->fdb_table.offloads.slow_fdb = fdb; - /* If lazy creation isn't supported, open the fast path tables now */ - if (!MLX5_CAP_ESW_FLOWTABLE(esw->dev, multi_fdb_encap) && - esw->offloads.encap != DEVLINK_ESWITCH_ENCAP_MODE_NONE) { - esw->fdb_table.flags &= ~ESW_FDB_CHAINS_AND_PRIOS_SUPPORTED; - esw_warn(dev, "Lazy creation of flow tables isn't supported, ignoring priorities\n"); - esw_get_prio_table(esw, 0, 1, 0); - esw_get_prio_table(esw, 0, 1, 1); - } else { - esw_debug(dev, "Lazy creation of flow tables supported, deferring table opening\n"); - esw->fdb_table.flags |= ESW_FDB_CHAINS_AND_PRIOS_SUPPORTED; + err = mlx5_esw_chains_create(esw); + if (err) { + esw_warn(dev, "Failed to create fdb chains err(%d)\n", err); + goto fdb_chains_err; } /* create send-to-vport group */ @@ -1218,7 +1012,8 @@ miss_err: peer_miss_err: mlx5_destroy_flow_group(esw->fdb_table.offloads.send_to_vport_grp); send_vport_err: - esw_destroy_offloads_fast_fdb_tables(esw); + mlx5_esw_chains_destroy(esw); +fdb_chains_err: mlx5_destroy_flow_table(esw->fdb_table.offloads.slow_fdb); slow_fdb_err: /* Holds true only as long as DMFS is the default */ @@ -1240,8 +1035,8 @@ static void esw_destroy_offloads_fdb_tables(struct mlx5_eswitch *esw) mlx5_destroy_flow_group(esw->fdb_table.offloads.peer_miss_grp); mlx5_destroy_flow_group(esw->fdb_table.offloads.miss_grp); + mlx5_esw_chains_destroy(esw); mlx5_destroy_flow_table(esw->fdb_table.offloads.slow_fdb); - esw_destroy_offloads_fast_fdb_tables(esw); /* Holds true only as long as DMFS is the default */ mlx5_flow_namespace_set_mode(esw->fdb_table.offloads.ns, MLX5_FLOW_STEERING_MODE_DMFS); @@ -2111,7 +1906,6 @@ static int esw_offloads_steering_init(struct mlx5_eswitch *esw) total_vports = num_vfs + MLX5_SPECIAL_VPORTS(esw->dev); memset(&esw->fdb_table.offloads, 0, sizeof(struct offloads_fdb)); - mutex_init(&esw->fdb_table.offloads.fdb_prio_lock); err = esw_create_uplink_offloads_acl_tables(esw); if (err) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads_chains.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads_chains.c new file mode 100644 index 000000000000..3a60eb5360bd --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads_chains.c @@ -0,0 +1,758 @@ +// SPDX-License-Identifier: GPL-2.0 OR Linux-OpenIB +// Copyright (c) 2020 Mellanox Technologies. + +#include <linux/mlx5/driver.h> +#include <linux/mlx5/mlx5_ifc.h> +#include <linux/mlx5/fs.h> + +#include "eswitch_offloads_chains.h" +#include "mlx5_core.h" +#include "fs_core.h" +#include "eswitch.h" +#include "en.h" + +#define esw_chains_priv(esw) ((esw)->fdb_table.offloads.esw_chains_priv) +#define esw_chains_lock(esw) (esw_chains_priv(esw)->lock) +#define esw_chains_ht(esw) (esw_chains_priv(esw)->chains_ht) +#define esw_prios_ht(esw) (esw_chains_priv(esw)->prios_ht) +#define fdb_pool_left(esw) (esw_chains_priv(esw)->fdb_left) +#define tc_slow_fdb(esw) ((esw)->fdb_table.offloads.slow_fdb) +#define tc_end_fdb(esw) (esw_chains_priv(esw)->tc_end_fdb) +#define fdb_ignore_flow_level_supported(esw) \ + (MLX5_CAP_ESW_FLOWTABLE_FDB((esw)->dev, ignore_flow_level)) + +#define ESW_OFFLOADS_NUM_GROUPS 4 + +/* Firmware currently has 4 pool of 4 sizes that it supports (ESW_POOLS), + * and a virtual memory region of 16M (ESW_SIZE), this region is duplicated + * for each flow table pool. We can allocate up to 16M of each pool, + * and we keep track of how much we used via get_next_avail_sz_from_pool. + * Firmware doesn't report any of this for now. + * ESW_POOL is expected to be sorted from large to small and match firmware + * pools. + */ +#define ESW_SIZE (16 * 1024 * 1024) +const unsigned int ESW_POOLS[] = { 4 * 1024 * 1024, + 1 * 1024 * 1024, + 64 * 1024, + 4 * 1024, }; + +struct mlx5_esw_chains_priv { + struct rhashtable chains_ht; + struct rhashtable prios_ht; + /* Protects above chains_ht and prios_ht */ + struct mutex lock; + + struct mlx5_flow_table *tc_end_fdb; + + int fdb_left[ARRAY_SIZE(ESW_POOLS)]; +}; + +struct fdb_chain { + struct rhash_head node; + + u32 chain; + + int ref; + + struct mlx5_eswitch *esw; + struct list_head prios_list; +}; + +struct fdb_prio_key { + u32 chain; + u32 prio; + u32 level; +}; + +struct fdb_prio { + struct rhash_head node; + struct list_head list; + + struct fdb_prio_key key; + + int ref; + + struct fdb_chain *fdb_chain; + struct mlx5_flow_table *fdb; + struct mlx5_flow_table *next_fdb; + struct mlx5_flow_group *miss_group; + struct mlx5_flow_handle *miss_rule; +}; + +static const struct rhashtable_params chain_params = { + .head_offset = offsetof(struct fdb_chain, node), + .key_offset = offsetof(struct fdb_chain, chain), + .key_len = sizeof_field(struct fdb_chain, chain), + .automatic_shrinking = true, +}; + +static const struct rhashtable_params prio_params = { + .head_offset = offsetof(struct fdb_prio, node), + .key_offset = offsetof(struct fdb_prio, key), + .key_len = sizeof_field(struct fdb_prio, key), + .automatic_shrinking = true, +}; + +bool mlx5_esw_chains_prios_supported(struct mlx5_eswitch *esw) +{ + return esw->fdb_table.flags & ESW_FDB_CHAINS_AND_PRIOS_SUPPORTED; +} + +u32 mlx5_esw_chains_get_chain_range(struct mlx5_eswitch *esw) +{ + if (!mlx5_esw_chains_prios_supported(esw)) + return 1; + + if (fdb_ignore_flow_level_supported(esw)) + return UINT_MAX - 1; + + return FDB_TC_MAX_CHAIN; +} + +u32 mlx5_esw_chains_get_ft_chain(struct mlx5_eswitch *esw) +{ + return mlx5_esw_chains_get_chain_range(esw) + 1; +} + +u32 mlx5_esw_chains_get_prio_range(struct mlx5_eswitch *esw) +{ + if (!mlx5_esw_chains_prios_supported(esw)) + return 1; + + if (fdb_ignore_flow_level_supported(esw)) + return UINT_MAX; + + return FDB_TC_MAX_PRIO; +} + +static unsigned int mlx5_esw_chains_get_level_range(struct mlx5_eswitch *esw) +{ + if (fdb_ignore_flow_level_supported(esw)) + return UINT_MAX; + + return FDB_TC_LEVELS_PER_PRIO; +} + +#define POOL_NEXT_SIZE 0 +static int +mlx5_esw_chains_get_avail_sz_from_pool(struct mlx5_eswitch *esw, + int desired_size) +{ + int i, found_i = -1; + + for (i = ARRAY_SIZE(ESW_POOLS) - 1; i >= 0; i--) { + if (fdb_pool_left(esw)[i] && ESW_POOLS[i] > desired_size) { + found_i = i; + if (desired_size != POOL_NEXT_SIZE) + break; + } + } + + if (found_i != -1) { + --fdb_pool_left(esw)[found_i]; + return ESW_POOLS[found_i]; + } + + return 0; +} + +static void +mlx5_esw_chains_put_sz_to_pool(struct mlx5_eswitch *esw, int sz) +{ + int i; + + for (i = ARRAY_SIZE(ESW_POOLS) - 1; i >= 0; i--) { + if (sz == ESW_POOLS[i]) { + ++fdb_pool_left(esw)[i]; + return; + } + } + + WARN_ONCE(1, "Couldn't find size %d in fdb size pool", sz); +} + +static void +mlx5_esw_chains_init_sz_pool(struct mlx5_eswitch *esw) +{ + u32 fdb_max; + int i; + + fdb_max = 1 << MLX5_CAP_ESW_FLOWTABLE_FDB(esw->dev, log_max_ft_size); + + for (i = ARRAY_SIZE(ESW_POOLS) - 1; i >= 0; i--) + fdb_pool_left(esw)[i] = + ESW_POOLS[i] <= fdb_max ? ESW_SIZE / ESW_POOLS[i] : 0; +} + +static struct mlx5_flow_table * +mlx5_esw_chains_create_fdb_table(struct mlx5_eswitch *esw, + u32 chain, u32 prio, u32 level) +{ + struct mlx5_flow_table_attr ft_attr = {}; + struct mlx5_flow_namespace *ns; + struct mlx5_flow_table *fdb; + int sz; + + if (esw->offloads.encap != DEVLINK_ESWITCH_ENCAP_MODE_NONE) + ft_attr.flags |= (MLX5_FLOW_TABLE_TUNNEL_EN_REFORMAT | + MLX5_FLOW_TABLE_TUNNEL_EN_DECAP); + + sz = mlx5_esw_chains_get_avail_sz_from_pool(esw, POOL_NEXT_SIZE); + if (!sz) + return ERR_PTR(-ENOSPC); + ft_attr.max_fte = sz; + + /* We use tc_slow_fdb(esw) as the table's next_ft till + * ignore_flow_level is allowed on FT creation and not just for FTEs. + * Instead caller should add an explicit miss rule if needed. + */ + ft_attr.next_ft = tc_slow_fdb(esw); + + /* The root table(chain 0, prio 1, level 0) is required to be + * connected to the previous prio (FDB_BYPASS_PATH if exists). + * We always create it, as a managed table, in order to align with + * fs_core logic. + */ + if (!fdb_ignore_flow_level_supported(esw) || + (chain == 0 && prio == 1 && level == 0)) { + ft_attr.level = level; + ft_attr.prio = prio - 1; + ns = mlx5_get_fdb_sub_ns(esw->dev, chain); + } else { + ft_attr.flags |= MLX5_FLOW_TABLE_UNMANAGED; + ft_attr.prio = FDB_TC_OFFLOAD; + /* Firmware doesn't allow us to create another level 0 table, + * so we create all unmanaged tables as level 1. + * + * To connect them, we use explicit miss rules with + * ignore_flow_level. Caller is responsible to create + * these rules (if needed). + */ + ft_attr.level = 1; + ns = mlx5_get_flow_namespace(esw->dev, MLX5_FLOW_NAMESPACE_FDB); + } + + ft_attr.autogroup.num_reserved_entries = 2; + ft_attr.autogroup.max_num_groups = ESW_OFFLOADS_NUM_GROUPS; + fdb = mlx5_create_auto_grouped_flow_table(ns, &ft_attr); + if (IS_ERR(fdb)) { + esw_warn(esw->dev, + "Failed to create FDB table err %d (chain: %d, prio: %d, level: %d, size: %d)\n", + (int)PTR_ERR(fdb), chain, prio, level, sz); + mlx5_esw_chains_put_sz_to_pool(esw, sz); + return fdb; + } + + return fdb; +} + +static void +mlx5_esw_chains_destroy_fdb_table(struct mlx5_eswitch *esw, + struct mlx5_flow_table *fdb) +{ + mlx5_esw_chains_put_sz_to_pool(esw, fdb->max_fte); + mlx5_destroy_flow_table(fdb); +} + +static struct fdb_chain * +mlx5_esw_chains_create_fdb_chain(struct mlx5_eswitch *esw, u32 chain) +{ + struct fdb_chain *fdb_chain = NULL; + int err; + + fdb_chain = kvzalloc(sizeof(*fdb_chain), GFP_KERNEL); + if (!fdb_chain) + return ERR_PTR(-ENOMEM); + + fdb_chain->esw = esw; + fdb_chain->chain = chain; + INIT_LIST_HEAD(&fdb_chain->prios_list); + + err = rhashtable_insert_fast(&esw_chains_ht(esw), &fdb_chain->node, + chain_params); + if (err) + goto err_insert; + + return fdb_chain; + +err_insert: + kvfree(fdb_chain); + return ERR_PTR(err); +} + +static void +mlx5_esw_chains_destroy_fdb_chain(struct fdb_chain *fdb_chain) +{ + struct mlx5_eswitch *esw = fdb_chain->esw; + + rhashtable_remove_fast(&esw_chains_ht(esw), &fdb_chain->node, + chain_params); + kvfree(fdb_chain); +} + +static struct fdb_chain * +mlx5_esw_chains_get_fdb_chain(struct mlx5_eswitch *esw, u32 chain) +{ + struct fdb_chain *fdb_chain; + + fdb_chain = rhashtable_lookup_fast(&esw_chains_ht(esw), &chain, + chain_params); + if (!fdb_chain) { + fdb_chain = mlx5_esw_chains_create_fdb_chain(esw, chain); + if (IS_ERR(fdb_chain)) + return fdb_chain; + } + + fdb_chain->ref++; + + return fdb_chain; +} + +static struct mlx5_flow_handle * +mlx5_esw_chains_add_miss_rule(struct mlx5_flow_table *fdb, + struct mlx5_flow_table *next_fdb) +{ + static const struct mlx5_flow_spec spec = {}; + struct mlx5_flow_destination dest = {}; + struct mlx5_flow_act act = {}; + + act.flags = FLOW_ACT_IGNORE_FLOW_LEVEL | FLOW_ACT_NO_APPEND; + act.action = MLX5_FLOW_CONTEXT_ACTION_FWD_DEST; + dest.type = MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE; + dest.ft = next_fdb; + + return mlx5_add_flow_rules(fdb, &spec, &act, &dest, 1); +} + +static int +mlx5_esw_chains_update_prio_prevs(struct fdb_prio *fdb_prio, + struct mlx5_flow_table *next_fdb) +{ + struct mlx5_flow_handle *miss_rules[FDB_TC_LEVELS_PER_PRIO + 1] = {}; + struct fdb_chain *fdb_chain = fdb_prio->fdb_chain; + struct fdb_prio *pos; + int n = 0, err; + + if (fdb_prio->key.level) + return 0; + + /* Iterate in reverse order until reaching the level 0 rule of + * the previous priority, adding all the miss rules first, so we can + * revert them if any of them fails. + */ + pos = fdb_prio; + list_for_each_entry_continue_reverse(pos, + &fdb_chain->prios_list, + list) { + miss_rules[n] = mlx5_esw_chains_add_miss_rule(pos->fdb, + next_fdb); + if (IS_ERR(miss_rules[n])) { + err = PTR_ERR(miss_rules[n]); + goto err_prev_rule; + } + + n++; + if (!pos->key.level) + break; + } + + /* Success, delete old miss rules, and update the pointers. */ + n = 0; + pos = fdb_prio; + list_for_each_entry_continue_reverse(pos, + &fdb_chain->prios_list, + list) { + mlx5_del_flow_rules(pos->miss_rule); + + pos->miss_rule = miss_rules[n]; + pos->next_fdb = next_fdb; + + n++; + if (!pos->key.level) + break; + } + + return 0; + +err_prev_rule: + while (--n >= 0) + mlx5_del_flow_rules(miss_rules[n]); + + return err; +} + +static void +mlx5_esw_chains_put_fdb_chain(struct fdb_chain *fdb_chain) +{ + if (--fdb_chain->ref == 0) + mlx5_esw_chains_destroy_fdb_chain(fdb_chain); +} + +static struct fdb_prio * +mlx5_esw_chains_create_fdb_prio(struct mlx5_eswitch *esw, + u32 chain, u32 prio, u32 level) +{ + int inlen = MLX5_ST_SZ_BYTES(create_flow_group_in); + struct mlx5_flow_handle *miss_rule = NULL; + struct mlx5_flow_group *miss_group; + struct fdb_prio *fdb_prio = NULL; + struct mlx5_flow_table *next_fdb; + struct fdb_chain *fdb_chain; + struct mlx5_flow_table *fdb; + struct list_head *pos; + u32 *flow_group_in; + int err; + + fdb_chain = mlx5_esw_chains_get_fdb_chain(esw, chain); + if (IS_ERR(fdb_chain)) + return ERR_CAST(fdb_chain); + + fdb_prio = kvzalloc(sizeof(*fdb_prio), GFP_KERNEL); + flow_group_in = kvzalloc(inlen, GFP_KERNEL); + if (!fdb_prio || !flow_group_in) { + err = -ENOMEM; + goto err_alloc; + } + + /* Chain's prio list is sorted by prio and level. + * And all levels of some prio point to the next prio's level 0. + * Example list (prio, level): + * (3,0)->(3,1)->(5,0)->(5,1)->(6,1)->(7,0) + * In hardware, we will we have the following pointers: + * (3,0) -> (5,0) -> (7,0) -> Slow path + * (3,1) -> (5,0) + * (5,1) -> (7,0) + * (6,1) -> (7,0) + */ + + /* Default miss for each chain: */ + next_fdb = (chain == mlx5_esw_chains_get_ft_chain(esw)) ? + tc_slow_fdb(esw) : + tc_end_fdb(esw); + list_for_each(pos, &fdb_chain->prios_list) { + struct fdb_prio *p = list_entry(pos, struct fdb_prio, list); + + /* exit on first pos that is larger */ + if (prio < p->key.prio || (prio == p->key.prio && + level < p->key.level)) { + /* Get next level 0 table */ + next_fdb = p->key.level == 0 ? p->fdb : p->next_fdb; + break; + } + } + + fdb = mlx5_esw_chains_create_fdb_table(esw, chain, prio, level); + if (IS_ERR(fdb)) { + err = PTR_ERR(fdb); + goto err_create; + } + + MLX5_SET(create_flow_group_in, flow_group_in, start_flow_index, + fdb->max_fte - 2); + MLX5_SET(create_flow_group_in, flow_group_in, end_flow_index, + fdb->max_fte - 1); + miss_group = mlx5_create_flow_group(fdb, flow_group_in); + if (IS_ERR(miss_group)) { + err = PTR_ERR(miss_group); + goto err_group; + } + + /* Add miss rule to next_fdb */ + miss_rule = mlx5_esw_chains_add_miss_rule(fdb, next_fdb); + if (IS_ERR(miss_rule)) { + err = PTR_ERR(miss_rule); + goto err_miss_rule; + } + + fdb_prio->miss_group = miss_group; + fdb_prio->miss_rule = miss_rule; + fdb_prio->next_fdb = next_fdb; + fdb_prio->fdb_chain = fdb_chain; + fdb_prio->key.chain = chain; + fdb_prio->key.prio = prio; + fdb_prio->key.level = level; + fdb_prio->fdb = fdb; + + err = rhashtable_insert_fast(&esw_prios_ht(esw), &fdb_prio->node, + prio_params); + if (err) + goto err_insert; + + list_add(&fdb_prio->list, pos->prev); + + /* Table is ready, connect it */ + err = mlx5_esw_chains_update_prio_prevs(fdb_prio, fdb); + if (err) + goto err_update; + + kvfree(flow_group_in); + return fdb_prio; + +err_update: + list_del(&fdb_prio->list); + rhashtable_remove_fast(&esw_prios_ht(esw), &fdb_prio->node, + prio_params); +err_insert: + mlx5_del_flow_rules(miss_rule); +err_miss_rule: + mlx5_destroy_flow_group(miss_group); +err_group: + mlx5_esw_chains_destroy_fdb_table(esw, fdb); +err_create: +err_alloc: + kvfree(fdb_prio); + kvfree(flow_group_in); + mlx5_esw_chains_put_fdb_chain(fdb_chain); + return ERR_PTR(err); +} + +static void +mlx5_esw_chains_destroy_fdb_prio(struct mlx5_eswitch *esw, + struct fdb_prio *fdb_prio) +{ + struct fdb_chain *fdb_chain = fdb_prio->fdb_chain; + + WARN_ON(mlx5_esw_chains_update_prio_prevs(fdb_prio, + fdb_prio->next_fdb)); + + list_del(&fdb_prio->list); + rhashtable_remove_fast(&esw_prios_ht(esw), &fdb_prio->node, + prio_params); + mlx5_del_flow_rules(fdb_prio->miss_rule); + mlx5_destroy_flow_group(fdb_prio->miss_group); + mlx5_esw_chains_destroy_fdb_table(esw, fdb_prio->fdb); + mlx5_esw_chains_put_fdb_chain(fdb_chain); + kvfree(fdb_prio); +} + +struct mlx5_flow_table * +mlx5_esw_chains_get_table(struct mlx5_eswitch *esw, u32 chain, u32 prio, + u32 level) +{ + struct mlx5_flow_table *prev_fts; + struct fdb_prio *fdb_prio; + struct fdb_prio_key key; + int l = 0; + + if ((chain > mlx5_esw_chains_get_chain_range(esw) && + chain != mlx5_esw_chains_get_ft_chain(esw)) || + prio > mlx5_esw_chains_get_prio_range(esw) || + level > mlx5_esw_chains_get_level_range(esw)) + return ERR_PTR(-EOPNOTSUPP); + + /* create earlier levels for correct fs_core lookup when + * connecting tables. + */ + for (l = 0; l < level; l++) { + prev_fts = mlx5_esw_chains_get_table(esw, chain, prio, l); + if (IS_ERR(prev_fts)) { + fdb_prio = ERR_CAST(prev_fts); + goto err_get_prevs; + } + } + + key.chain = chain; + key.prio = prio; + key.level = level; + + mutex_lock(&esw_chains_lock(esw)); + fdb_prio = rhashtable_lookup_fast(&esw_prios_ht(esw), &key, + prio_params); + if (!fdb_prio) { + fdb_prio = mlx5_esw_chains_create_fdb_prio(esw, chain, + prio, level); + if (IS_ERR(fdb_prio)) + goto err_create_prio; + } + + ++fdb_prio->ref; + mutex_unlock(&esw_chains_lock(esw)); + + return fdb_prio->fdb; + +err_create_prio: + mutex_unlock(&esw_chains_lock(esw)); +err_get_prevs: + while (--l >= 0) + mlx5_esw_chains_put_table(esw, chain, prio, l); + return ERR_CAST(fdb_prio); +} + +void +mlx5_esw_chains_put_table(struct mlx5_eswitch *esw, u32 chain, u32 prio, + u32 level) +{ + struct fdb_prio *fdb_prio; + struct fdb_prio_key key; + + key.chain = chain; + key.prio = prio; + key.level = level; + + mutex_lock(&esw_chains_lock(esw)); + fdb_prio = rhashtable_lookup_fast(&esw_prios_ht(esw), &key, + prio_params); + if (!fdb_prio) + goto err_get_prio; + + if (--fdb_prio->ref == 0) + mlx5_esw_chains_destroy_fdb_prio(esw, fdb_prio); + mutex_unlock(&esw_chains_lock(esw)); + + while (level-- > 0) + mlx5_esw_chains_put_table(esw, chain, prio, level); + + return; + +err_get_prio: + mutex_unlock(&esw_chains_lock(esw)); + WARN_ONCE(1, + "Couldn't find table: (chain: %d prio: %d level: %d)", + chain, prio, level); +} + +struct mlx5_flow_table * +mlx5_esw_chains_get_tc_end_ft(struct mlx5_eswitch *esw) +{ + return tc_end_fdb(esw); +} + +static int +mlx5_esw_chains_init(struct mlx5_eswitch *esw) +{ + struct mlx5_esw_chains_priv *chains_priv; + struct mlx5_core_dev *dev = esw->dev; + u32 max_flow_counter, fdb_max; + int err; + + chains_priv = kzalloc(sizeof(*chains_priv), GFP_KERNEL); + if (!chains_priv) + return -ENOMEM; + esw_chains_priv(esw) = chains_priv; + + max_flow_counter = (MLX5_CAP_GEN(dev, max_flow_counter_31_16) << 16) | + MLX5_CAP_GEN(dev, max_flow_counter_15_0); + fdb_max = 1 << MLX5_CAP_ESW_FLOWTABLE_FDB(dev, log_max_ft_size); + + esw_debug(dev, + "Init esw offloads chains, max counters(%d), groups(%d), max flow table size(%d)\n", + max_flow_counter, ESW_OFFLOADS_NUM_GROUPS, fdb_max); + + mlx5_esw_chains_init_sz_pool(esw); + + if (!MLX5_CAP_ESW_FLOWTABLE(esw->dev, multi_fdb_encap) && + esw->offloads.encap != DEVLINK_ESWITCH_ENCAP_MODE_NONE) { + esw->fdb_table.flags &= ~ESW_FDB_CHAINS_AND_PRIOS_SUPPORTED; + esw_warn(dev, "Tc chains and priorities offload aren't supported, update firmware if needed\n"); + } else { + esw->fdb_table.flags |= ESW_FDB_CHAINS_AND_PRIOS_SUPPORTED; + esw_info(dev, "Supported tc offload range - chains: %u, prios: %u\n", + mlx5_esw_chains_get_chain_range(esw), + mlx5_esw_chains_get_prio_range(esw)); + } + + err = rhashtable_init(&esw_chains_ht(esw), &chain_params); + if (err) + goto init_chains_ht_err; + + err = rhashtable_init(&esw_prios_ht(esw), &prio_params); + if (err) + goto init_prios_ht_err; + + mutex_init(&esw_chains_lock(esw)); + + return 0; + +init_prios_ht_err: + rhashtable_destroy(&esw_chains_ht(esw)); +init_chains_ht_err: + kfree(chains_priv); + return err; +} + +static void +mlx5_esw_chains_cleanup(struct mlx5_eswitch *esw) +{ + mutex_destroy(&esw_chains_lock(esw)); + rhashtable_destroy(&esw_prios_ht(esw)); + rhashtable_destroy(&esw_chains_ht(esw)); + + kfree(esw_chains_priv(esw)); +} + +static int +mlx5_esw_chains_open(struct mlx5_eswitch *esw) +{ + struct mlx5_flow_table *ft; + int err; + + /* Create tc_end_fdb(esw) which is the always created ft chain */ + ft = mlx5_esw_chains_get_table(esw, mlx5_esw_chains_get_ft_chain(esw), + 1, 0); + if (IS_ERR(ft)) + return PTR_ERR(ft); + + tc_end_fdb(esw) = ft; + + /* Always open the root for fast path */ + ft = mlx5_esw_chains_get_table(esw, 0, 1, 0); + if (IS_ERR(ft)) { + err = PTR_ERR(ft); + goto level_0_err; + } + + /* Open level 1 for split rules now if prios isn't supported */ + if (!mlx5_esw_chains_prios_supported(esw)) { + ft = mlx5_esw_chains_get_table(esw, 0, 1, 1); + + if (IS_ERR(ft)) { + err = PTR_ERR(ft); + goto level_1_err; + } + } + + return 0; + +level_1_err: + mlx5_esw_chains_put_table(esw, 0, 1, 0); +level_0_err: + mlx5_esw_chains_put_table(esw, mlx5_esw_chains_get_ft_chain(esw), 1, 0); + return err; +} + +static void +mlx5_esw_chains_close(struct mlx5_eswitch *esw) +{ + if (!mlx5_esw_chains_prios_supported(esw)) + mlx5_esw_chains_put_table(esw, 0, 1, 1); + mlx5_esw_chains_put_table(esw, 0, 1, 0); + mlx5_esw_chains_put_table(esw, mlx5_esw_chains_get_ft_chain(esw), 1, 0); +} + +int +mlx5_esw_chains_create(struct mlx5_eswitch *esw) +{ + int err; + + err = mlx5_esw_chains_init(esw); + if (err) + return err; + + err = mlx5_esw_chains_open(esw); + if (err) + goto err_open; + + return 0; + +err_open: + mlx5_esw_chains_cleanup(esw); + return err; +} + +void +mlx5_esw_chains_destroy(struct mlx5_eswitch *esw) +{ + mlx5_esw_chains_close(esw); + mlx5_esw_chains_cleanup(esw); +} diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads_chains.h b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads_chains.h new file mode 100644 index 000000000000..2e13097fe348 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads_chains.h @@ -0,0 +1,30 @@ +/* SPDX-License-Identifier: GPL-2.0 OR Linux-OpenIB */ +/* Copyright (c) 2020 Mellanox Technologies. */ + +#ifndef __ML5_ESW_CHAINS_H__ +#define __ML5_ESW_CHAINS_H__ + +bool +mlx5_esw_chains_prios_supported(struct mlx5_eswitch *esw); +u32 +mlx5_esw_chains_get_prio_range(struct mlx5_eswitch *esw); +u32 +mlx5_esw_chains_get_chain_range(struct mlx5_eswitch *esw); +u32 +mlx5_esw_chains_get_ft_chain(struct mlx5_eswitch *esw); + +struct mlx5_flow_table * +mlx5_esw_chains_get_table(struct mlx5_eswitch *esw, u32 chain, u32 prio, + u32 level); +void +mlx5_esw_chains_put_table(struct mlx5_eswitch *esw, u32 chain, u32 prio, + u32 level); + +struct mlx5_flow_table * +mlx5_esw_chains_get_tc_end_ft(struct mlx5_eswitch *esw); + +int mlx5_esw_chains_create(struct mlx5_eswitch *esw); +void mlx5_esw_chains_destroy(struct mlx5_eswitch *esw); + +#endif /* __ML5_ESW_CHAINS_H__ */ + diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads_termtbl.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads_termtbl.c index 366bda1bb1c3..dc08ed9339ab 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads_termtbl.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads_termtbl.c @@ -50,8 +50,8 @@ mlx5_eswitch_termtbl_create(struct mlx5_core_dev *dev, struct mlx5_flow_act *flow_act) { static const struct mlx5_flow_spec spec = {}; + struct mlx5_flow_table_attr ft_attr = {}; struct mlx5_flow_namespace *root_ns; - int prio, flags; int err; root_ns = mlx5_get_flow_namespace(dev, MLX5_FLOW_NAMESPACE_FDB); @@ -63,10 +63,11 @@ mlx5_eswitch_termtbl_create(struct mlx5_core_dev *dev, /* As this is the terminating action then the termination table is the * same prio as the slow path */ - prio = FDB_SLOW_PATH; - flags = MLX5_FLOW_TABLE_TERMINATION; - tt->termtbl = mlx5_create_auto_grouped_flow_table(root_ns, prio, 1, 1, - 0, flags); + ft_attr.flags = MLX5_FLOW_TABLE_TERMINATION; + ft_attr.prio = FDB_SLOW_PATH; + ft_attr.max_fte = 1; + ft_attr.autogroup.max_num_groups = 1; + tt->termtbl = mlx5_create_auto_grouped_flow_table(root_ns, &ft_attr); if (IS_ERR(tt->termtbl)) { esw_warn(dev, "Failed to create termination table\n"); return -EOPNOTSUPP; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c index 3c816e81f8d9..b25465d9e030 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c @@ -432,6 +432,9 @@ static int mlx5_cmd_set_fte(struct mlx5_core_dev *dev, MLX5_SET(set_fte_in, in, table_type, ft->type); MLX5_SET(set_fte_in, in, table_id, ft->id); MLX5_SET(set_fte_in, in, flow_index, fte->index); + MLX5_SET(set_fte_in, in, ignore_flow_level, + !!(fte->action.flags & FLOW_ACT_IGNORE_FLOW_LEVEL)); + if (ft->vport) { MLX5_SET(set_fte_in, in, vport_number, ft->vport); MLX5_SET(set_fte_in, in, other_vport, 1); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c index 9a48c4310887..c7a16ae05fa8 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c @@ -531,16 +531,9 @@ static void del_hw_fte(struct fs_node *node) } } -static void del_sw_fte_rcu(struct rcu_head *head) -{ - struct fs_fte *fte = container_of(head, struct fs_fte, rcu); - struct mlx5_flow_steering *steering = get_steering(&fte->node); - - kmem_cache_free(steering->ftes_cache, fte); -} - static void del_sw_fte(struct fs_node *node) { + struct mlx5_flow_steering *steering = get_steering(node); struct mlx5_flow_group *fg; struct fs_fte *fte; int err; @@ -553,8 +546,7 @@ static void del_sw_fte(struct fs_node *node) rhash_fte); WARN_ON(err); ida_simple_remove(&fg->fte_allocator, fte->index - fg->start_index); - - call_rcu(&fte->rcu, del_sw_fte_rcu); + kmem_cache_free(steering->ftes_cache, fte); } static void del_hw_flow_group(struct fs_node *node) @@ -587,7 +579,9 @@ static void del_sw_flow_group(struct fs_node *node) rhashtable_destroy(&fg->ftes_hash); ida_destroy(&fg->fte_allocator); - if (ft->autogroup.active && fg->max_ftes == ft->autogroup.group_size) + if (ft->autogroup.active && + fg->max_ftes == ft->autogroup.group_size && + fg->start_index < ft->autogroup.max_fte) ft->autogroup.num_groups--; err = rhltable_remove(&ft->fgs_hash, &fg->hash, @@ -1014,7 +1008,8 @@ static struct mlx5_flow_table *__mlx5_create_flow_table(struct mlx5_flow_namespa u16 vport) { struct mlx5_flow_root_namespace *root = find_root(&ns->node); - struct mlx5_flow_table *next_ft = NULL; + bool unmanaged = ft_attr->flags & MLX5_FLOW_TABLE_UNMANAGED; + struct mlx5_flow_table *next_ft; struct fs_prio *fs_prio = NULL; struct mlx5_flow_table *ft; int log_table_sz; @@ -1031,14 +1026,21 @@ static struct mlx5_flow_table *__mlx5_create_flow_table(struct mlx5_flow_namespa err = -EINVAL; goto unlock_root; } - if (ft_attr->level >= fs_prio->num_levels) { - err = -ENOSPC; - goto unlock_root; + if (!unmanaged) { + /* The level is related to the + * priority level range. + */ + if (ft_attr->level >= fs_prio->num_levels) { + err = -ENOSPC; + goto unlock_root; + } + + ft_attr->level += fs_prio->start_level; } + /* The level is related to the * priority level range. */ - ft_attr->level += fs_prio->start_level; ft = alloc_flow_table(ft_attr->level, vport, ft_attr->max_fte ? roundup_pow_of_two(ft_attr->max_fte) : 0, @@ -1051,19 +1053,27 @@ static struct mlx5_flow_table *__mlx5_create_flow_table(struct mlx5_flow_namespa tree_init_node(&ft->node, del_hw_flow_table, del_sw_flow_table); log_table_sz = ft->max_fte ? ilog2(ft->max_fte) : 0; - next_ft = find_next_chained_ft(fs_prio); + next_ft = unmanaged ? ft_attr->next_ft : + find_next_chained_ft(fs_prio); ft->def_miss_action = ns->def_miss_action; err = root->cmds->create_flow_table(root, ft, log_table_sz, next_ft); if (err) goto free_ft; - err = connect_flow_table(root->dev, ft, fs_prio); - if (err) - goto destroy_ft; + if (!unmanaged) { + err = connect_flow_table(root->dev, ft, fs_prio); + if (err) + goto destroy_ft; + } + ft->node.active = true; down_write_ref_node(&fs_prio->node, false); - tree_add_node(&ft->node, &fs_prio->node); - list_add_flow_table(ft, fs_prio); + if (!unmanaged) { + tree_add_node(&ft->node, &fs_prio->node); + list_add_flow_table(ft, fs_prio); + } else { + ft->node.root = fs_prio->node.root; + } fs_prio->num_ft++; up_write_ref_node(&fs_prio->node, false); mutex_unlock(&root->chain_lock); @@ -1111,31 +1121,27 @@ EXPORT_SYMBOL(mlx5_create_lag_demux_flow_table); struct mlx5_flow_table* mlx5_create_auto_grouped_flow_table(struct mlx5_flow_namespace *ns, - int prio, - int num_flow_table_entries, - int max_num_groups, - u32 level, - u32 flags) + struct mlx5_flow_table_attr *ft_attr) { - struct mlx5_flow_table_attr ft_attr = {}; + int num_reserved_entries = ft_attr->autogroup.num_reserved_entries; + int autogroups_max_fte = ft_attr->max_fte - num_reserved_entries; + int max_num_groups = ft_attr->autogroup.max_num_groups; struct mlx5_flow_table *ft; - if (max_num_groups > num_flow_table_entries) + if (max_num_groups > autogroups_max_fte) + return ERR_PTR(-EINVAL); + if (num_reserved_entries > ft_attr->max_fte) return ERR_PTR(-EINVAL); - ft_attr.max_fte = num_flow_table_entries; - ft_attr.prio = prio; - ft_attr.level = level; - ft_attr.flags = flags; - - ft = mlx5_create_flow_table(ns, &ft_attr); + ft = mlx5_create_flow_table(ns, ft_attr); if (IS_ERR(ft)) return ft; ft->autogroup.active = true; ft->autogroup.required_groups = max_num_groups; + ft->autogroup.max_fte = autogroups_max_fte; /* We save place for flow groups in addition to max types */ - ft->autogroup.group_size = ft->max_fte / (max_num_groups + 1); + ft->autogroup.group_size = autogroups_max_fte / (max_num_groups + 1); return ft; } @@ -1157,7 +1163,7 @@ struct mlx5_flow_group *mlx5_create_flow_group(struct mlx5_flow_table *ft, struct mlx5_flow_group *fg; int err; - if (ft->autogroup.active) + if (ft->autogroup.active && start_index < ft->autogroup.max_fte) return ERR_PTR(-EPERM); down_write_ref_node(&ft->node, false); @@ -1330,9 +1336,10 @@ static struct mlx5_flow_group *alloc_auto_flow_group(struct mlx5_flow_table *ft const struct mlx5_flow_spec *spec) { struct list_head *prev = &ft->node.children; - struct mlx5_flow_group *fg; + u32 max_fte = ft->autogroup.max_fte; unsigned int candidate_index = 0; unsigned int group_size = 0; + struct mlx5_flow_group *fg; if (!ft->autogroup.active) return ERR_PTR(-ENOENT); @@ -1340,7 +1347,7 @@ static struct mlx5_flow_group *alloc_auto_flow_group(struct mlx5_flow_table *ft if (ft->autogroup.num_groups < ft->autogroup.required_groups) group_size = ft->autogroup.group_size; - /* ft->max_fte == ft->autogroup.max_types */ + /* max_fte == ft->autogroup.max_types */ if (group_size == 0) group_size = 1; @@ -1353,7 +1360,7 @@ static struct mlx5_flow_group *alloc_auto_flow_group(struct mlx5_flow_table *ft prev = &fg->node.list; } - if (candidate_index + group_size > ft->max_fte) + if (candidate_index + group_size > max_fte) return ERR_PTR(-ENOSPC); fg = alloc_insert_flow_group(ft, @@ -1537,18 +1544,30 @@ static bool counter_is_valid(u32 action) } static bool dest_is_valid(struct mlx5_flow_destination *dest, - u32 action, + struct mlx5_flow_act *flow_act, struct mlx5_flow_table *ft) { + bool ignore_level = flow_act->flags & FLOW_ACT_IGNORE_FLOW_LEVEL; + u32 action = flow_act->action; + if (dest && (dest->type == MLX5_FLOW_DESTINATION_TYPE_COUNTER)) return counter_is_valid(action); if (!(action & MLX5_FLOW_CONTEXT_ACTION_FWD_DEST)) return true; + if (ignore_level) { + if (ft->type != FS_FT_FDB) + return false; + + if (dest->type == MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE && + dest->ft->type != FS_FT_FDB) + return false; + } + if (!dest || ((dest->type == MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE) && - (dest->ft->level <= ft->level))) + (dest->ft->level <= ft->level && !ignore_level))) return false; return true; } @@ -1633,47 +1652,22 @@ static u64 matched_fgs_get_version(struct list_head *match_head) } static struct fs_fte * -lookup_fte_for_write_locked(struct mlx5_flow_group *g, const u32 *match_value) -{ - struct fs_fte *fte_tmp; - - nested_down_write_ref_node(&g->node, FS_LOCK_PARENT); - - fte_tmp = rhashtable_lookup_fast(&g->ftes_hash, match_value, rhash_fte); - if (!fte_tmp || !tree_get_node(&fte_tmp->node)) { - fte_tmp = NULL; - goto out; - } - - if (!fte_tmp->node.active) { - tree_put_node(&fte_tmp->node, false); - fte_tmp = NULL; - goto out; - } - nested_down_write_ref_node(&fte_tmp->node, FS_LOCK_CHILD); - -out: - up_write_ref_node(&g->node, false); - return fte_tmp; -} - -static struct fs_fte * -lookup_fte_for_read_locked(struct mlx5_flow_group *g, const u32 *match_value) +lookup_fte_locked(struct mlx5_flow_group *g, + const u32 *match_value, + bool take_write) { struct fs_fte *fte_tmp; - if (!tree_get_node(&g->node)) - return NULL; - - rcu_read_lock(); - fte_tmp = rhashtable_lookup(&g->ftes_hash, match_value, rhash_fte); + if (take_write) + nested_down_write_ref_node(&g->node, FS_LOCK_PARENT); + else + nested_down_read_ref_node(&g->node, FS_LOCK_PARENT); + fte_tmp = rhashtable_lookup_fast(&g->ftes_hash, match_value, + rhash_fte); if (!fte_tmp || !tree_get_node(&fte_tmp->node)) { - rcu_read_unlock(); fte_tmp = NULL; goto out; } - rcu_read_unlock(); - if (!fte_tmp->node.active) { tree_put_node(&fte_tmp->node, false); fte_tmp = NULL; @@ -1681,19 +1675,12 @@ lookup_fte_for_read_locked(struct mlx5_flow_group *g, const u32 *match_value) } nested_down_write_ref_node(&fte_tmp->node, FS_LOCK_CHILD); - out: - tree_put_node(&g->node, false); - return fte_tmp; -} - -static struct fs_fte * -lookup_fte_locked(struct mlx5_flow_group *g, const u32 *match_value, bool write) -{ - if (write) - return lookup_fte_for_write_locked(g, match_value); + if (take_write) + up_write_ref_node(&g->node, false); else - return lookup_fte_for_read_locked(g, match_value); + up_read_ref_node(&g->node); + return fte_tmp; } static struct mlx5_flow_handle * @@ -1810,7 +1797,7 @@ _mlx5_add_flow_rules(struct mlx5_flow_table *ft, return ERR_PTR(-EINVAL); for (i = 0; i < dest_num; i++) { - if (!dest_is_valid(&dest[i], flow_act->action, ft)) + if (!dest_is_valid(&dest[i], flow_act, ft)) return ERR_PTR(-EINVAL); } nested_down_read_ref_node(&ft->node, FS_LOCK_GRANDPARENT); @@ -2073,7 +2060,8 @@ int mlx5_destroy_flow_table(struct mlx5_flow_table *ft) int err = 0; mutex_lock(&root->chain_lock); - err = disconnect_flow_table(ft); + if (!(ft->flags & MLX5_FLOW_TABLE_UNMANAGED)) + err = disconnect_flow_table(ft); if (err) { mutex_unlock(&root->chain_lock); return err; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.h b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.h index e8cd997f413e..be5f5e32c1e8 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.h @@ -164,6 +164,7 @@ struct mlx5_flow_table { unsigned int required_groups; unsigned int group_size; unsigned int num_groups; + unsigned int max_fte; } autogroup; /* Protect fwd_rules */ struct mutex lock; @@ -203,7 +204,6 @@ struct fs_fte { enum fs_fte_status status; struct mlx5_fc *counter; struct rhash_head hash; - struct rcu_head rcu; int modify_mask; }; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fw.c b/drivers/net/ethernet/mellanox/mlx5/core/fw.c index a19790dee7b2..d89ff1d09119 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/fw.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/fw.c @@ -131,11 +131,11 @@ static int mlx5_get_pcam_reg(struct mlx5_core_dev *dev) MLX5_PCAM_REGS_5000_TO_507F); } -static int mlx5_get_mcam_reg(struct mlx5_core_dev *dev) +static int mlx5_get_mcam_access_reg_group(struct mlx5_core_dev *dev, + enum mlx5_mcam_reg_groups group) { - return mlx5_query_mcam_reg(dev, dev->caps.mcam, - MLX5_MCAM_FEATURE_ENHANCED_FEATURES, - MLX5_MCAM_REGS_FIRST_128); + return mlx5_query_mcam_reg(dev, dev->caps.mcam[group], + MLX5_MCAM_FEATURE_ENHANCED_FEATURES, group); } static int mlx5_get_qcam_reg(struct mlx5_core_dev *dev) @@ -221,8 +221,11 @@ int mlx5_query_hca_caps(struct mlx5_core_dev *dev) if (MLX5_CAP_GEN(dev, pcam_reg)) mlx5_get_pcam_reg(dev); - if (MLX5_CAP_GEN(dev, mcam_reg)) - mlx5_get_mcam_reg(dev); + if (MLX5_CAP_GEN(dev, mcam_reg)) { + mlx5_get_mcam_access_reg_group(dev, MLX5_MCAM_REGS_FIRST_128); + mlx5_get_mcam_access_reg_group(dev, MLX5_MCAM_REGS_0x9080_0x90FF); + mlx5_get_mcam_access_reg_group(dev, MLX5_MCAM_REGS_0x9100_0x917F); + } if (MLX5_CAP_GEN(dev, qcam_reg)) mlx5_get_qcam_reg(dev); @@ -245,6 +248,13 @@ int mlx5_query_hca_caps(struct mlx5_core_dev *dev) return err; } + if (MLX5_CAP_GEN_64(dev, general_obj_types) & + MLX5_GENERAL_OBJ_TYPES_CAP_VIRTIO_NET_Q) { + err = mlx5_core_get_caps(dev, MLX5_CAP_VDPA_EMULATION); + if (err) + return err; + } + return 0; } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib.c b/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib.c index 3ed8ab2d703d..7c87f523e370 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib.c @@ -87,8 +87,8 @@ int mlx5i_init(struct mlx5_core_dev *mdev, mlx5e_set_netdev_mtu_boundaries(priv); netdev->mtu = netdev->max_mtu; - mlx5e_build_nic_params(mdev, NULL, &priv->rss_params, &priv->channels.params, - priv->max_nch, netdev->mtu); + mlx5e_build_nic_params(priv, NULL, &priv->rss_params, &priv->channels.params, + netdev->mtu); mlx5i_build_nic_params(mdev, &priv->channels.params); mlx5e_timestamp_init(priv); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/main.c b/drivers/net/ethernet/mellanox/mlx5/core/main.c index 173e2c12e1c7..cf7b8da0f010 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/main.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/main.c @@ -1193,6 +1193,12 @@ int mlx5_load_one(struct mlx5_core_dev *dev, bool boot) if (err) goto err_load; + if (boot) { + err = mlx5_devlink_register(priv_to_devlink(dev), dev->device); + if (err) + goto err_devlink_reg; + } + if (mlx5_device_registered(dev)) { mlx5_attach_device(dev); } else { @@ -1210,6 +1216,9 @@ out: return err; err_reg_dev: + if (boot) + mlx5_devlink_unregister(priv_to_devlink(dev)); +err_devlink_reg: mlx5_unload(dev); err_load: if (boot) @@ -1347,10 +1356,6 @@ static int init_one(struct pci_dev *pdev, const struct pci_device_id *id) request_module_nowait(MLX5_IB_MOD); - err = mlx5_devlink_register(devlink, &pdev->dev); - if (err) - goto clean_load; - err = mlx5_crdump_enable(dev); if (err) dev_err(&pdev->dev, "mlx5_crdump_enable failed with error code %d\n", err); @@ -1358,9 +1363,6 @@ static int init_one(struct pci_dev *pdev, const struct pci_device_id *id) pci_save_state(pdev); return 0; -clean_load: - mlx5_unload_one(dev, true); - err_load_one: mlx5_pci_close(dev); pci_init_err: diff --git a/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_action.c b/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_action.c index 004c56c2fc0c..9359eed10889 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_action.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_action.c @@ -690,9 +690,9 @@ int mlx5dr_actions_build_ste_arr(struct mlx5dr_matcher *matcher, /* get the relevant addresses */ if (!action->dest_tbl.fw_tbl.rx_icm_addr) { - ret = mlx5dr_cmd_query_flow_table(action->dest_tbl.fw_tbl.mdev, - action->dest_tbl.fw_tbl.ft->type, - action->dest_tbl.fw_tbl.ft->id, + ret = mlx5dr_cmd_query_flow_table(dmn->mdev, + action->dest_tbl.fw_tbl.type, + action->dest_tbl.fw_tbl.id, &output); if (!ret) { action->dest_tbl.fw_tbl.tx_icm_addr = @@ -982,8 +982,106 @@ dec_ref: } struct mlx5dr_action * -mlx5dr_create_action_dest_flow_fw_table(struct mlx5_flow_table *ft, - struct mlx5_core_dev *mdev) +mlx5dr_action_create_mult_dest_tbl(struct mlx5dr_domain *dmn, + struct mlx5dr_action_dest *dests, + u32 num_of_dests) +{ + struct mlx5dr_cmd_flow_destination_hw_info *hw_dests; + struct mlx5dr_action **ref_actions; + struct mlx5dr_action *action; + bool reformat_req = false; + u32 num_of_ref = 0; + int ret; + int i; + + if (dmn->type != MLX5DR_DOMAIN_TYPE_FDB) { + mlx5dr_err(dmn, "Multiple destination support is for FDB only\n"); + return NULL; + } + + hw_dests = kzalloc(sizeof(*hw_dests) * num_of_dests, GFP_KERNEL); + if (!hw_dests) + return NULL; + + ref_actions = kzalloc(sizeof(*ref_actions) * num_of_dests * 2, GFP_KERNEL); + if (!ref_actions) + goto free_hw_dests; + + for (i = 0; i < num_of_dests; i++) { + struct mlx5dr_action *reformat_action = dests[i].reformat; + struct mlx5dr_action *dest_action = dests[i].dest; + + ref_actions[num_of_ref++] = dest_action; + + switch (dest_action->action_type) { + case DR_ACTION_TYP_VPORT: + hw_dests[i].vport.flags = MLX5_FLOW_DEST_VPORT_VHCA_ID; + hw_dests[i].type = MLX5_FLOW_DESTINATION_TYPE_VPORT; + hw_dests[i].vport.num = dest_action->vport.caps->num; + hw_dests[i].vport.vhca_id = dest_action->vport.caps->vhca_gvmi; + if (reformat_action) { + reformat_req = true; + hw_dests[i].vport.reformat_id = + reformat_action->reformat.reformat_id; + ref_actions[num_of_ref++] = reformat_action; + hw_dests[i].vport.flags |= MLX5_FLOW_DEST_VPORT_REFORMAT_ID; + } + break; + + case DR_ACTION_TYP_FT: + hw_dests[i].type = MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE; + if (dest_action->dest_tbl.is_fw_tbl) + hw_dests[i].ft_id = dest_action->dest_tbl.fw_tbl.id; + else + hw_dests[i].ft_id = dest_action->dest_tbl.tbl->table_id; + break; + + default: + mlx5dr_dbg(dmn, "Invalid multiple destinations action\n"); + goto free_ref_actions; + } + } + + action = dr_action_create_generic(DR_ACTION_TYP_FT); + if (!action) + goto free_ref_actions; + + ret = mlx5dr_fw_create_md_tbl(dmn, + hw_dests, + num_of_dests, + reformat_req, + &action->dest_tbl.fw_tbl.id, + &action->dest_tbl.fw_tbl.group_id); + if (ret) + goto free_action; + + refcount_inc(&dmn->refcount); + + for (i = 0; i < num_of_ref; i++) + refcount_inc(&ref_actions[i]->refcount); + + action->dest_tbl.is_fw_tbl = true; + action->dest_tbl.fw_tbl.dmn = dmn; + action->dest_tbl.fw_tbl.type = FS_FT_FDB; + action->dest_tbl.fw_tbl.ref_actions = ref_actions; + action->dest_tbl.fw_tbl.num_of_ref_actions = num_of_ref; + + kfree(hw_dests); + + return action; + +free_action: + kfree(action); +free_ref_actions: + kfree(ref_actions); +free_hw_dests: + kfree(hw_dests); + return NULL; +} + +struct mlx5dr_action * +mlx5dr_action_create_dest_flow_fw_table(struct mlx5dr_domain *dmn, + struct mlx5_flow_table *ft) { struct mlx5dr_action *action; @@ -992,8 +1090,11 @@ mlx5dr_create_action_dest_flow_fw_table(struct mlx5_flow_table *ft, return NULL; action->dest_tbl.is_fw_tbl = 1; - action->dest_tbl.fw_tbl.ft = ft; - action->dest_tbl.fw_tbl.mdev = mdev; + action->dest_tbl.fw_tbl.type = ft->type; + action->dest_tbl.fw_tbl.id = ft->id; + action->dest_tbl.fw_tbl.dmn = dmn; + + refcount_inc(&dmn->refcount); return action; } @@ -1559,8 +1660,26 @@ int mlx5dr_action_destroy(struct mlx5dr_action *action) switch (action->action_type) { case DR_ACTION_TYP_FT: - if (!action->dest_tbl.is_fw_tbl) + if (action->dest_tbl.is_fw_tbl) + refcount_dec(&action->dest_tbl.fw_tbl.dmn->refcount); + else refcount_dec(&action->dest_tbl.tbl->refcount); + + if (action->dest_tbl.is_fw_tbl && + action->dest_tbl.fw_tbl.num_of_ref_actions) { + struct mlx5dr_action **ref_actions; + int i; + + ref_actions = action->dest_tbl.fw_tbl.ref_actions; + for (i = 0; i < action->dest_tbl.fw_tbl.num_of_ref_actions; i++) + refcount_dec(&ref_actions[i]->refcount); + + kfree(ref_actions); + + mlx5dr_fw_destroy_md_tbl(action->dest_tbl.fw_tbl.dmn, + action->dest_tbl.fw_tbl.id, + action->dest_tbl.fw_tbl.group_id); + } break; case DR_ACTION_TYP_TNL_L2_TO_L2: refcount_dec(&action->reformat.dmn->refcount); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_cmd.c b/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_cmd.c index 41662c4e2664..461b39376daf 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_cmd.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_cmd.c @@ -320,12 +320,7 @@ int mlx5dr_cmd_destroy_flow_group(struct mlx5_core_dev *mdev, } int mlx5dr_cmd_create_flow_table(struct mlx5_core_dev *mdev, - u32 table_type, - u64 icm_addr_rx, - u64 icm_addr_tx, - u8 level, - bool sw_owner, - bool term_tbl, + struct mlx5dr_cmd_create_flow_table_attr *attr, u64 *fdb_rx_icm_addr, u32 *table_id) { @@ -335,37 +330,43 @@ int mlx5dr_cmd_create_flow_table(struct mlx5_core_dev *mdev, int err; MLX5_SET(create_flow_table_in, in, opcode, MLX5_CMD_OP_CREATE_FLOW_TABLE); - MLX5_SET(create_flow_table_in, in, table_type, table_type); + MLX5_SET(create_flow_table_in, in, table_type, attr->table_type); ft_mdev = MLX5_ADDR_OF(create_flow_table_in, in, flow_table_context); - MLX5_SET(flow_table_context, ft_mdev, termination_table, term_tbl); - MLX5_SET(flow_table_context, ft_mdev, sw_owner, sw_owner); - MLX5_SET(flow_table_context, ft_mdev, level, level); + MLX5_SET(flow_table_context, ft_mdev, termination_table, attr->term_tbl); + MLX5_SET(flow_table_context, ft_mdev, sw_owner, attr->sw_owner); + MLX5_SET(flow_table_context, ft_mdev, level, attr->level); - if (sw_owner) { + if (attr->sw_owner) { /* icm_addr_0 used for FDB RX / NIC TX / NIC_RX * icm_addr_1 used for FDB TX */ - if (table_type == MLX5_FLOW_TABLE_TYPE_NIC_RX) { + if (attr->table_type == MLX5_FLOW_TABLE_TYPE_NIC_RX) { MLX5_SET64(flow_table_context, ft_mdev, - sw_owner_icm_root_0, icm_addr_rx); - } else if (table_type == MLX5_FLOW_TABLE_TYPE_NIC_TX) { + sw_owner_icm_root_0, attr->icm_addr_rx); + } else if (attr->table_type == MLX5_FLOW_TABLE_TYPE_NIC_TX) { MLX5_SET64(flow_table_context, ft_mdev, - sw_owner_icm_root_0, icm_addr_tx); - } else if (table_type == MLX5_FLOW_TABLE_TYPE_FDB) { + sw_owner_icm_root_0, attr->icm_addr_tx); + } else if (attr->table_type == MLX5_FLOW_TABLE_TYPE_FDB) { MLX5_SET64(flow_table_context, ft_mdev, - sw_owner_icm_root_0, icm_addr_rx); + sw_owner_icm_root_0, attr->icm_addr_rx); MLX5_SET64(flow_table_context, ft_mdev, - sw_owner_icm_root_1, icm_addr_tx); + sw_owner_icm_root_1, attr->icm_addr_tx); } } + MLX5_SET(create_flow_table_in, in, flow_table_context.decap_en, + attr->decap_en); + MLX5_SET(create_flow_table_in, in, flow_table_context.reformat_en, + attr->reformat_en); + err = mlx5_cmd_exec(mdev, in, sizeof(in), out, sizeof(out)); if (err) return err; *table_id = MLX5_GET(create_flow_table_out, out, table_id); - if (!sw_owner && table_type == MLX5_FLOW_TABLE_TYPE_FDB) + if (!attr->sw_owner && attr->table_type == MLX5_FLOW_TABLE_TYPE_FDB && + fdb_rx_icm_addr) *fdb_rx_icm_addr = (u64)MLX5_GET(create_flow_table_out, out, icm_address_31_0) | (u64)MLX5_GET(create_flow_table_out, out, icm_address_39_32) << 32 | @@ -478,3 +479,208 @@ int mlx5dr_cmd_query_gid(struct mlx5_core_dev *mdev, u8 vhca_port_num, return 0; } + +static int mlx5dr_cmd_set_extended_dest(struct mlx5_core_dev *dev, + struct mlx5dr_cmd_fte_info *fte, + bool *extended_dest) +{ + int fw_log_max_fdb_encap_uplink = MLX5_CAP_ESW(dev, log_max_fdb_encap_uplink); + int num_fwd_destinations = 0; + int num_encap = 0; + int i; + + *extended_dest = false; + if (!(fte->action.action & MLX5_FLOW_CONTEXT_ACTION_FWD_DEST)) + return 0; + for (i = 0; i < fte->dests_size; i++) { + if (fte->dest_arr[i].type == MLX5_FLOW_DESTINATION_TYPE_COUNTER) + continue; + if (fte->dest_arr[i].type == MLX5_FLOW_DESTINATION_TYPE_VPORT && + fte->dest_arr[i].vport.flags & MLX5_FLOW_DEST_VPORT_REFORMAT_ID) + num_encap++; + num_fwd_destinations++; + } + + if (num_fwd_destinations > 1 && num_encap > 0) + *extended_dest = true; + + if (*extended_dest && !fw_log_max_fdb_encap_uplink) { + mlx5_core_warn(dev, "FW does not support extended destination"); + return -EOPNOTSUPP; + } + if (num_encap > (1 << fw_log_max_fdb_encap_uplink)) { + mlx5_core_warn(dev, "FW does not support more than %d encaps", + 1 << fw_log_max_fdb_encap_uplink); + return -EOPNOTSUPP; + } + + return 0; +} + +int mlx5dr_cmd_set_fte(struct mlx5_core_dev *dev, + int opmod, int modify_mask, + struct mlx5dr_cmd_ft_info *ft, + u32 group_id, + struct mlx5dr_cmd_fte_info *fte) +{ + u32 out[MLX5_ST_SZ_DW(set_fte_out)] = {}; + void *in_flow_context, *vlan; + bool extended_dest = false; + void *in_match_value; + unsigned int inlen; + int dst_cnt_size; + void *in_dests; + u32 *in; + int err; + int i; + + if (mlx5dr_cmd_set_extended_dest(dev, fte, &extended_dest)) + return -EOPNOTSUPP; + + if (!extended_dest) + dst_cnt_size = MLX5_ST_SZ_BYTES(dest_format_struct); + else + dst_cnt_size = MLX5_ST_SZ_BYTES(extended_dest_format); + + inlen = MLX5_ST_SZ_BYTES(set_fte_in) + fte->dests_size * dst_cnt_size; + in = kvzalloc(inlen, GFP_KERNEL); + if (!in) + return -ENOMEM; + + MLX5_SET(set_fte_in, in, opcode, MLX5_CMD_OP_SET_FLOW_TABLE_ENTRY); + MLX5_SET(set_fte_in, in, op_mod, opmod); + MLX5_SET(set_fte_in, in, modify_enable_mask, modify_mask); + MLX5_SET(set_fte_in, in, table_type, ft->type); + MLX5_SET(set_fte_in, in, table_id, ft->id); + MLX5_SET(set_fte_in, in, flow_index, fte->index); + if (ft->vport) { + MLX5_SET(set_fte_in, in, vport_number, ft->vport); + MLX5_SET(set_fte_in, in, other_vport, 1); + } + + in_flow_context = MLX5_ADDR_OF(set_fte_in, in, flow_context); + MLX5_SET(flow_context, in_flow_context, group_id, group_id); + + MLX5_SET(flow_context, in_flow_context, flow_tag, + fte->flow_context.flow_tag); + MLX5_SET(flow_context, in_flow_context, flow_source, + fte->flow_context.flow_source); + + MLX5_SET(flow_context, in_flow_context, extended_destination, + extended_dest); + if (extended_dest) { + u32 action; + + action = fte->action.action & + ~MLX5_FLOW_CONTEXT_ACTION_PACKET_REFORMAT; + MLX5_SET(flow_context, in_flow_context, action, action); + } else { + MLX5_SET(flow_context, in_flow_context, action, + fte->action.action); + if (fte->action.pkt_reformat) + MLX5_SET(flow_context, in_flow_context, packet_reformat_id, + fte->action.pkt_reformat->id); + } + if (fte->action.modify_hdr) + MLX5_SET(flow_context, in_flow_context, modify_header_id, + fte->action.modify_hdr->id); + + vlan = MLX5_ADDR_OF(flow_context, in_flow_context, push_vlan); + + MLX5_SET(vlan, vlan, ethtype, fte->action.vlan[0].ethtype); + MLX5_SET(vlan, vlan, vid, fte->action.vlan[0].vid); + MLX5_SET(vlan, vlan, prio, fte->action.vlan[0].prio); + + vlan = MLX5_ADDR_OF(flow_context, in_flow_context, push_vlan_2); + + MLX5_SET(vlan, vlan, ethtype, fte->action.vlan[1].ethtype); + MLX5_SET(vlan, vlan, vid, fte->action.vlan[1].vid); + MLX5_SET(vlan, vlan, prio, fte->action.vlan[1].prio); + + in_match_value = MLX5_ADDR_OF(flow_context, in_flow_context, + match_value); + memcpy(in_match_value, fte->val, sizeof(u32) * MLX5_ST_SZ_DW_MATCH_PARAM); + + in_dests = MLX5_ADDR_OF(flow_context, in_flow_context, destination); + if (fte->action.action & MLX5_FLOW_CONTEXT_ACTION_FWD_DEST) { + int list_size = 0; + + for (i = 0; i < fte->dests_size; i++) { + unsigned int id, type = fte->dest_arr[i].type; + + if (type == MLX5_FLOW_DESTINATION_TYPE_COUNTER) + continue; + + switch (type) { + case MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE_NUM: + id = fte->dest_arr[i].ft_num; + type = MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE; + break; + case MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE: + id = fte->dest_arr[i].ft_id; + break; + case MLX5_FLOW_DESTINATION_TYPE_VPORT: + id = fte->dest_arr[i].vport.num; + MLX5_SET(dest_format_struct, in_dests, + destination_eswitch_owner_vhca_id_valid, + !!(fte->dest_arr[i].vport.flags & + MLX5_FLOW_DEST_VPORT_VHCA_ID)); + MLX5_SET(dest_format_struct, in_dests, + destination_eswitch_owner_vhca_id, + fte->dest_arr[i].vport.vhca_id); + if (extended_dest && (fte->dest_arr[i].vport.flags & + MLX5_FLOW_DEST_VPORT_REFORMAT_ID)) { + MLX5_SET(dest_format_struct, in_dests, + packet_reformat, + !!(fte->dest_arr[i].vport.flags & + MLX5_FLOW_DEST_VPORT_REFORMAT_ID)); + MLX5_SET(extended_dest_format, in_dests, + packet_reformat_id, + fte->dest_arr[i].vport.reformat_id); + } + break; + default: + id = fte->dest_arr[i].tir_num; + } + + MLX5_SET(dest_format_struct, in_dests, destination_type, + type); + MLX5_SET(dest_format_struct, in_dests, destination_id, id); + in_dests += dst_cnt_size; + list_size++; + } + + MLX5_SET(flow_context, in_flow_context, destination_list_size, + list_size); + } + + if (fte->action.action & MLX5_FLOW_CONTEXT_ACTION_COUNT) { + int max_list_size = BIT(MLX5_CAP_FLOWTABLE_TYPE(dev, + log_max_flow_counter, + ft->type)); + int list_size = 0; + + for (i = 0; i < fte->dests_size; i++) { + if (fte->dest_arr[i].type != + MLX5_FLOW_DESTINATION_TYPE_COUNTER) + continue; + + MLX5_SET(flow_counter_list, in_dests, flow_counter_id, + fte->dest_arr[i].counter_id); + in_dests += dst_cnt_size; + list_size++; + } + if (list_size > max_list_size) { + err = -EINVAL; + goto err_out; + } + + MLX5_SET(flow_context, in_flow_context, flow_counter_list_size, + list_size); + } + + err = mlx5_cmd_exec(dev, in, inlen, out, sizeof(out)); +err_out: + kvfree(in); + return err; +} diff --git a/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_fw.c b/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_fw.c index 60ef6e6171e3..1fbcd012bb85 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_fw.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_fw.c @@ -7,6 +7,7 @@ struct mlx5dr_fw_recalc_cs_ft * mlx5dr_fw_create_recalc_cs_ft(struct mlx5dr_domain *dmn, u32 vport_num) { + struct mlx5dr_cmd_create_flow_table_attr ft_attr = {}; struct mlx5dr_fw_recalc_cs_ft *recalc_cs_ft; u32 table_id, group_id, modify_hdr_id; u64 rx_icm_addr, modify_ttl_action; @@ -16,9 +17,14 @@ mlx5dr_fw_create_recalc_cs_ft(struct mlx5dr_domain *dmn, u32 vport_num) if (!recalc_cs_ft) return NULL; - ret = mlx5dr_cmd_create_flow_table(dmn->mdev, MLX5_FLOW_TABLE_TYPE_FDB, - 0, 0, dmn->info.caps.max_ft_level - 1, - false, true, &rx_icm_addr, &table_id); + ft_attr.table_type = MLX5_FLOW_TABLE_TYPE_FDB; + ft_attr.level = dmn->info.caps.max_ft_level - 1; + ft_attr.term_tbl = true; + + ret = mlx5dr_cmd_create_flow_table(dmn->mdev, + &ft_attr, + &rx_icm_addr, + &table_id); if (ret) { mlx5dr_err(dmn, "Failed creating TTL W/A FW flow table %d\n", ret); goto free_ttl_tbl; @@ -91,3 +97,70 @@ void mlx5dr_fw_destroy_recalc_cs_ft(struct mlx5dr_domain *dmn, kfree(recalc_cs_ft); } + +int mlx5dr_fw_create_md_tbl(struct mlx5dr_domain *dmn, + struct mlx5dr_cmd_flow_destination_hw_info *dest, + int num_dest, + bool reformat_req, + u32 *tbl_id, + u32 *group_id) +{ + struct mlx5dr_cmd_create_flow_table_attr ft_attr = {}; + struct mlx5dr_cmd_fte_info fte_info = {}; + u32 val[MLX5_ST_SZ_DW_MATCH_PARAM] = {}; + struct mlx5dr_cmd_ft_info ft_info = {}; + int ret; + + ft_attr.table_type = MLX5_FLOW_TABLE_TYPE_FDB; + ft_attr.level = dmn->info.caps.max_ft_level - 2; + ft_attr.reformat_en = reformat_req; + ft_attr.decap_en = reformat_req; + + ret = mlx5dr_cmd_create_flow_table(dmn->mdev, &ft_attr, NULL, tbl_id); + if (ret) { + mlx5dr_err(dmn, "Failed creating multi dest FW flow table %d\n", ret); + return ret; + } + + ret = mlx5dr_cmd_create_empty_flow_group(dmn->mdev, + MLX5_FLOW_TABLE_TYPE_FDB, + *tbl_id, group_id); + if (ret) { + mlx5dr_err(dmn, "Failed creating multi dest FW flow group %d\n", ret); + goto free_flow_table; + } + + ft_info.id = *tbl_id; + ft_info.type = FS_FT_FDB; + fte_info.action.action = MLX5_FLOW_CONTEXT_ACTION_FWD_DEST; + fte_info.dests_size = num_dest; + fte_info.val = val; + fte_info.dest_arr = dest; + + ret = mlx5dr_cmd_set_fte(dmn->mdev, 0, 0, &ft_info, *group_id, &fte_info); + if (ret) { + mlx5dr_err(dmn, "Failed setting fte into table %d\n", ret); + goto free_flow_group; + } + + return 0; + +free_flow_group: + mlx5dr_cmd_destroy_flow_group(dmn->mdev, MLX5_FLOW_TABLE_TYPE_FDB, + *tbl_id, *group_id); +free_flow_table: + mlx5dr_cmd_destroy_flow_table(dmn->mdev, *tbl_id, + MLX5_FLOW_TABLE_TYPE_FDB); + return ret; +} + +void mlx5dr_fw_destroy_md_tbl(struct mlx5dr_domain *dmn, + u32 tbl_id, u32 group_id) +{ + mlx5dr_cmd_del_flow_table_entry(dmn->mdev, FS_FT_FDB, tbl_id); + mlx5dr_cmd_destroy_flow_group(dmn->mdev, + MLX5_FLOW_TABLE_TYPE_FDB, + tbl_id, group_id); + mlx5dr_cmd_destroy_flow_table(dmn->mdev, tbl_id, + MLX5_FLOW_TABLE_TYPE_FDB); +} diff --git a/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_rule.c b/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_rule.c index 32e94d2ee5e4..e4cff7abb348 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_rule.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_rule.c @@ -209,7 +209,7 @@ static void dr_rule_rehash_copy_ste_ctrl(struct mlx5dr_matcher *matcher, /* We need to copy the refcount since this ste * may have been traversed several times */ - refcount_set(&new_ste->refcount, refcount_read(&cur_ste->refcount)); + new_ste->refcount = cur_ste->refcount; /* Link old STEs rule_mem list to the new ste */ mlx5dr_rule_update_rule_member(cur_ste, new_ste); @@ -638,6 +638,9 @@ static int dr_rule_add_member(struct mlx5dr_rule_rx_tx *nic_rule, if (!rule_mem) return -ENOMEM; + INIT_LIST_HEAD(&rule_mem->list); + INIT_LIST_HEAD(&rule_mem->use_ste_list); + rule_mem->ste = ste; list_add_tail(&rule_mem->list, &nic_rule->rule_members_list); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_ste.c b/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_ste.c index a5a266983dd3..c6c7d1defbd7 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_ste.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_ste.c @@ -348,7 +348,7 @@ static void dr_ste_replace(struct mlx5dr_ste *dst, struct mlx5dr_ste *src) if (dst->next_htbl) dst->next_htbl->pointing_ste = dst; - refcount_set(&dst->refcount, refcount_read(&src->refcount)); + dst->refcount = src->refcount; INIT_LIST_HEAD(&dst->rule_list); list_splice_tail_init(&src->rule_list, &dst->rule_list); @@ -565,7 +565,7 @@ bool mlx5dr_ste_is_not_valid_entry(u8 *p_hw_ste) bool mlx5dr_ste_not_used_ste(struct mlx5dr_ste *ste) { - return !refcount_read(&ste->refcount); + return !ste->refcount; } /* Init one ste as a pattern for ste data array */ @@ -689,14 +689,14 @@ struct mlx5dr_ste_htbl *mlx5dr_ste_htbl_alloc(struct mlx5dr_icm_pool *pool, htbl->ste_arr = chunk->ste_arr; htbl->hw_ste_arr = chunk->hw_ste_arr; htbl->miss_list = chunk->miss_list; - refcount_set(&htbl->refcount, 0); + htbl->refcount = 0; for (i = 0; i < chunk->num_of_entries; i++) { struct mlx5dr_ste *ste = &htbl->ste_arr[i]; ste->hw_ste = htbl->hw_ste_arr + i * DR_STE_SIZE_REDUCED; ste->htbl = htbl; - refcount_set(&ste->refcount, 0); + ste->refcount = 0; INIT_LIST_HEAD(&ste->miss_list_node); INIT_LIST_HEAD(&htbl->miss_list[i]); INIT_LIST_HEAD(&ste->rule_list); @@ -713,7 +713,7 @@ out_free_htbl: int mlx5dr_ste_htbl_free(struct mlx5dr_ste_htbl *htbl) { - if (refcount_read(&htbl->refcount)) + if (htbl->refcount) return -EBUSY; mlx5dr_icm_free_chunk(htbl->chunk); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_table.c b/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_table.c index e178d8d3dbc9..14ce2d7dbb66 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_table.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_table.c @@ -211,6 +211,9 @@ static int dr_table_destroy_sw_owned_tbl(struct mlx5dr_table *tbl) static int dr_table_create_sw_owned_tbl(struct mlx5dr_table *tbl) { + bool en_encap = !!(tbl->flags & MLX5_FLOW_TABLE_TUNNEL_EN_REFORMAT); + bool en_decap = !!(tbl->flags & MLX5_FLOW_TABLE_TUNNEL_EN_DECAP); + struct mlx5dr_cmd_create_flow_table_attr ft_attr = {}; u64 icm_addr_rx = 0; u64 icm_addr_tx = 0; int ret; @@ -221,18 +224,21 @@ static int dr_table_create_sw_owned_tbl(struct mlx5dr_table *tbl) if (tbl->tx.s_anchor) icm_addr_tx = tbl->tx.s_anchor->chunk->icm_addr; - ret = mlx5dr_cmd_create_flow_table(tbl->dmn->mdev, - tbl->table_type, - icm_addr_rx, - icm_addr_tx, - tbl->dmn->info.caps.max_ft_level - 1, - true, false, NULL, - &tbl->table_id); + ft_attr.table_type = tbl->table_type; + ft_attr.icm_addr_rx = icm_addr_rx; + ft_attr.icm_addr_tx = icm_addr_tx; + ft_attr.level = tbl->dmn->info.caps.max_ft_level - 1; + ft_attr.sw_owner = true; + ft_attr.decap_en = en_decap; + ft_attr.reformat_en = en_encap; + + ret = mlx5dr_cmd_create_flow_table(tbl->dmn->mdev, &ft_attr, + NULL, &tbl->table_id); return ret; } -struct mlx5dr_table *mlx5dr_table_create(struct mlx5dr_domain *dmn, u32 level) +struct mlx5dr_table *mlx5dr_table_create(struct mlx5dr_domain *dmn, u32 level, u32 flags) { struct mlx5dr_table *tbl; int ret; @@ -245,6 +251,7 @@ struct mlx5dr_table *mlx5dr_table_create(struct mlx5dr_domain *dmn, u32 level) tbl->dmn = dmn; tbl->level = level; + tbl->flags = flags; refcount_set(&tbl->refcount, 1); ret = dr_table_init(tbl); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_types.h b/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_types.h index 290fe61c33d0..dffe35145d19 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_types.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_types.h @@ -123,7 +123,7 @@ struct mlx5dr_matcher_rx_tx; struct mlx5dr_ste { u8 *hw_ste; /* refcount: indicates the num of rules that using this ste */ - refcount_t refcount; + u32 refcount; /* attached to the miss_list head at each htbl entry */ struct list_head miss_list_node; @@ -155,7 +155,7 @@ struct mlx5dr_ste_htbl_ctrl { struct mlx5dr_ste_htbl { u8 lu_type; u16 byte_mask; - refcount_t refcount; + u32 refcount; struct mlx5dr_icm_chunk *chunk; struct mlx5dr_ste *ste_arr; u8 *hw_ste_arr; @@ -206,13 +206,14 @@ int mlx5dr_ste_htbl_free(struct mlx5dr_ste_htbl *htbl); static inline void mlx5dr_htbl_put(struct mlx5dr_ste_htbl *htbl) { - if (refcount_dec_and_test(&htbl->refcount)) + htbl->refcount--; + if (!htbl->refcount) mlx5dr_ste_htbl_free(htbl); } static inline void mlx5dr_htbl_get(struct mlx5dr_ste_htbl *htbl) { - refcount_inc(&htbl->refcount); + htbl->refcount++; } /* STE utils */ @@ -254,14 +255,15 @@ static inline void mlx5dr_ste_put(struct mlx5dr_ste *ste, struct mlx5dr_matcher *matcher, struct mlx5dr_matcher_rx_tx *nic_matcher) { - if (refcount_dec_and_test(&ste->refcount)) + ste->refcount--; + if (!ste->refcount) mlx5dr_ste_free(ste, matcher, nic_matcher); } /* initial as 0, increased only when ste appears in a new rule */ static inline void mlx5dr_ste_get(struct mlx5dr_ste *ste) { - refcount_inc(&ste->refcount); + ste->refcount++; } void mlx5dr_ste_set_hit_addr_by_next_htbl(u8 *hw_ste, @@ -679,6 +681,7 @@ struct mlx5dr_table { u32 level; u32 table_type; u32 table_id; + u32 flags; struct list_head matcher_list; struct mlx5dr_action *miss_action; refcount_t refcount; @@ -742,10 +745,14 @@ struct mlx5dr_action { union { struct mlx5dr_table *tbl; struct { - struct mlx5_flow_table *ft; + struct mlx5dr_domain *dmn; + u32 id; + u32 group_id; + enum fs_flow_table_type type; u64 rx_icm_addr; u64 tx_icm_addr; - struct mlx5_core_dev *mdev; + struct mlx5dr_action **ref_actions; + u32 num_of_ref_actions; } fw_tbl; }; } dest_tbl; @@ -867,6 +874,17 @@ struct mlx5dr_cmd_query_flow_table_details { u64 sw_owner_icm_root_0; }; +struct mlx5dr_cmd_create_flow_table_attr { + u32 table_type; + u64 icm_addr_rx; + u64 icm_addr_tx; + u8 level; + bool sw_owner; + bool term_tbl; + bool decap_en; + bool reformat_en; +}; + /* internal API functions */ int mlx5dr_cmd_query_device(struct mlx5_core_dev *mdev, struct mlx5dr_cmd_caps *caps); @@ -904,12 +922,7 @@ int mlx5dr_cmd_destroy_flow_group(struct mlx5_core_dev *mdev, u32 table_id, u32 group_id); int mlx5dr_cmd_create_flow_table(struct mlx5_core_dev *mdev, - u32 table_type, - u64 icm_addr_rx, - u64 icm_addr_tx, - u8 level, - bool sw_owner, - bool term_tbl, + struct mlx5dr_cmd_create_flow_table_attr *attr, u64 *fdb_rx_icm_addr, u32 *table_id); int mlx5dr_cmd_destroy_flow_table(struct mlx5_core_dev *mdev, @@ -1051,6 +1064,43 @@ int mlx5dr_send_postsend_formatted_htbl(struct mlx5dr_domain *dmn, int mlx5dr_send_postsend_action(struct mlx5dr_domain *dmn, struct mlx5dr_action *action); +struct mlx5dr_cmd_ft_info { + u32 id; + u16 vport; + enum fs_flow_table_type type; +}; + +struct mlx5dr_cmd_flow_destination_hw_info { + enum mlx5_flow_destination_type type; + union { + u32 tir_num; + u32 ft_num; + u32 ft_id; + u32 counter_id; + struct { + u16 num; + u16 vhca_id; + u32 reformat_id; + u8 flags; + } vport; + }; +}; + +struct mlx5dr_cmd_fte_info { + u32 dests_size; + u32 index; + struct mlx5_flow_context flow_context; + u32 *val; + struct mlx5_flow_act action; + struct mlx5dr_cmd_flow_destination_hw_info *dest_arr; +}; + +int mlx5dr_cmd_set_fte(struct mlx5_core_dev *dev, + int opmod, int modify_mask, + struct mlx5dr_cmd_ft_info *ft, + u32 group_id, + struct mlx5dr_cmd_fte_info *fte); + struct mlx5dr_fw_recalc_cs_ft { u64 rx_icm_addr; u32 table_id; @@ -1065,4 +1115,12 @@ void mlx5dr_fw_destroy_recalc_cs_ft(struct mlx5dr_domain *dmn, int mlx5dr_domain_cache_get_recalc_cs_ft_addr(struct mlx5dr_domain *dmn, u32 vport_num, u64 *rx_icm_addr); +int mlx5dr_fw_create_md_tbl(struct mlx5dr_domain *dmn, + struct mlx5dr_cmd_flow_destination_hw_info *dest, + int num_dest, + bool reformat_req, + u32 *tbl_id, + u32 *group_id); +void mlx5dr_fw_destroy_md_tbl(struct mlx5dr_domain *dmn, u32 tbl_id, + u32 group_id); #endif /* _DR_TYPES_H_ */ diff --git a/drivers/net/ethernet/mellanox/mlx5/core/steering/fs_dr.c b/drivers/net/ethernet/mellanox/mlx5/core/steering/fs_dr.c index 3d587d0bdbbe..b43275cde8bf 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/steering/fs_dr.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/steering/fs_dr.c @@ -74,7 +74,7 @@ static int mlx5_cmd_dr_create_flow_table(struct mlx5_flow_root_namespace *ns, next_ft); tbl = mlx5dr_table_create(ns->fs_dr_domain.dr_domain, - ft->level); + ft->level, ft->flags); if (!tbl) { mlx5_core_err(ns->dev, "Failed creating dr flow_table\n"); return -EINVAL; @@ -184,13 +184,13 @@ static struct mlx5dr_action *create_vport_action(struct mlx5dr_domain *domain, dest_attr->vport.vhca_id); } -static struct mlx5dr_action *create_ft_action(struct mlx5_core_dev *dev, +static struct mlx5dr_action *create_ft_action(struct mlx5dr_domain *domain, struct mlx5_flow_rule *dst) { struct mlx5_flow_table *dest_ft = dst->dest_attr.ft; if (mlx5_dr_is_fw_table(dest_ft->flags)) - return mlx5dr_create_action_dest_flow_fw_table(dest_ft, dev); + return mlx5dr_action_create_dest_flow_fw_table(domain, dest_ft); return mlx5dr_action_create_dest_table(dest_ft->fs_dr_table.dr_table); } @@ -206,6 +206,12 @@ static struct mlx5dr_action *create_action_push_vlan(struct mlx5dr_domain *domai return mlx5dr_action_create_push_vlan(domain, htonl(vlan_hdr)); } +static bool contain_vport_reformat_action(struct mlx5_flow_rule *dst) +{ + return dst->dest_attr.type == MLX5_FLOW_DESTINATION_TYPE_VPORT && + dst->dest_attr.vport.flags & MLX5_FLOW_DEST_VPORT_REFORMAT_ID; +} + #define MLX5_FLOW_CONTEXT_ACTION_MAX 20 static int mlx5_cmd_dr_create_fte(struct mlx5_flow_root_namespace *ns, struct mlx5_flow_table *ft, @@ -213,7 +219,7 @@ static int mlx5_cmd_dr_create_fte(struct mlx5_flow_root_namespace *ns, struct fs_fte *fte) { struct mlx5dr_domain *domain = ns->fs_dr_domain.dr_domain; - struct mlx5dr_action *term_action = NULL; + struct mlx5dr_action_dest *term_actions; struct mlx5dr_match_parameters params; struct mlx5_core_dev *dev = ns->dev; struct mlx5dr_action **fs_dr_actions; @@ -223,6 +229,7 @@ static int mlx5_cmd_dr_create_fte(struct mlx5_flow_root_namespace *ns, struct mlx5dr_rule *rule; struct mlx5_flow_rule *dst; int fs_dr_num_actions = 0; + int num_term_actions = 0; int num_actions = 0; size_t match_sz; int err = 0; @@ -233,18 +240,38 @@ static int mlx5_cmd_dr_create_fte(struct mlx5_flow_root_namespace *ns, actions = kcalloc(MLX5_FLOW_CONTEXT_ACTION_MAX, sizeof(*actions), GFP_KERNEL); - if (!actions) - return -ENOMEM; + if (!actions) { + err = -ENOMEM; + goto out_err; + } fs_dr_actions = kcalloc(MLX5_FLOW_CONTEXT_ACTION_MAX, sizeof(*fs_dr_actions), GFP_KERNEL); if (!fs_dr_actions) { - kfree(actions); - return -ENOMEM; + err = -ENOMEM; + goto free_actions_alloc; + } + + term_actions = kcalloc(MLX5_FLOW_CONTEXT_ACTION_MAX, + sizeof(*term_actions), GFP_KERNEL); + if (!term_actions) { + err = -ENOMEM; + goto free_fs_dr_actions_alloc; } match_sz = sizeof(fte->val); + /* Drop reformat action bit if destination vport set with reformat */ + if (fte->action.action & MLX5_FLOW_CONTEXT_ACTION_FWD_DEST) { + list_for_each_entry(dst, &fte->node.children, node.list) { + if (!contain_vport_reformat_action(dst)) + continue; + + fte->action.action &= ~MLX5_FLOW_CONTEXT_ACTION_PACKET_REFORMAT; + break; + } + } + /* The order of the actions are must to be keep, only the following * order is supported by SW steering: * TX: push vlan -> modify header -> encap @@ -335,7 +362,7 @@ static int mlx5_cmd_dr_create_fte(struct mlx5_flow_root_namespace *ns, goto free_actions; } fs_dr_actions[fs_dr_num_actions++] = tmp_action; - term_action = tmp_action; + term_actions[num_term_actions++].dest = tmp_action; } if (fte->flow_context.flow_tag) { @@ -354,7 +381,8 @@ static int mlx5_cmd_dr_create_fte(struct mlx5_flow_root_namespace *ns, enum mlx5_flow_destination_type type = dst->dest_attr.type; u32 id; - if (num_actions == MLX5_FLOW_CONTEXT_ACTION_MAX) { + if (num_actions == MLX5_FLOW_CONTEXT_ACTION_MAX || + num_term_actions >= MLX5_FLOW_CONTEXT_ACTION_MAX) { err = -ENOSPC; goto free_actions; } @@ -373,13 +401,13 @@ static int mlx5_cmd_dr_create_fte(struct mlx5_flow_root_namespace *ns, actions[num_actions++] = tmp_action; break; case MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE: - tmp_action = create_ft_action(dev, dst); + tmp_action = create_ft_action(domain, dst); if (!tmp_action) { err = -ENOMEM; goto free_actions; } fs_dr_actions[fs_dr_num_actions++] = tmp_action; - term_action = tmp_action; + term_actions[num_term_actions++].dest = tmp_action; break; case MLX5_FLOW_DESTINATION_TYPE_VPORT: tmp_action = create_vport_action(domain, dst); @@ -388,7 +416,14 @@ static int mlx5_cmd_dr_create_fte(struct mlx5_flow_root_namespace *ns, goto free_actions; } fs_dr_actions[fs_dr_num_actions++] = tmp_action; - term_action = tmp_action; + term_actions[num_term_actions].dest = tmp_action; + + if (dst->dest_attr.vport.flags & + MLX5_FLOW_DEST_VPORT_REFORMAT_ID) + term_actions[num_term_actions].reformat = + dst->dest_attr.vport.pkt_reformat->action.dr_action; + + num_term_actions++; break; default: err = -EOPNOTSUPP; @@ -399,9 +434,22 @@ static int mlx5_cmd_dr_create_fte(struct mlx5_flow_root_namespace *ns, params.match_sz = match_sz; params.match_buf = (u64 *)fte->val; - - if (term_action) - actions[num_actions++] = term_action; + if (num_term_actions == 1) { + if (term_actions->reformat) + actions[num_actions++] = term_actions->reformat; + + actions[num_actions++] = term_actions->dest; + } else if (num_term_actions > 1) { + tmp_action = mlx5dr_action_create_mult_dest_tbl(domain, + term_actions, + num_term_actions); + if (!tmp_action) { + err = -EOPNOTSUPP; + goto free_actions; + } + fs_dr_actions[fs_dr_num_actions++] = tmp_action; + actions[num_actions++] = tmp_action; + } rule = mlx5dr_rule_create(group->fs_dr_matcher.dr_matcher, ¶ms, @@ -412,7 +460,9 @@ static int mlx5_cmd_dr_create_fte(struct mlx5_flow_root_namespace *ns, goto free_actions; } + kfree(term_actions); kfree(actions); + fte->fs_dr_rule.dr_rule = rule; fte->fs_dr_rule.num_actions = fs_dr_num_actions; fte->fs_dr_rule.dr_actions = fs_dr_actions; @@ -420,13 +470,18 @@ static int mlx5_cmd_dr_create_fte(struct mlx5_flow_root_namespace *ns, return 0; free_actions: - for (i = 0; i < fs_dr_num_actions; i++) + /* Free in reverse order to handle action dependencies */ + for (i = fs_dr_num_actions - 1; i >= 0; i--) if (!IS_ERR_OR_NULL(fs_dr_actions[i])) mlx5dr_action_destroy(fs_dr_actions[i]); - mlx5_core_err(dev, "Failed to create dr rule err(%d)\n", err); - kfree(actions); + kfree(term_actions); +free_fs_dr_actions_alloc: kfree(fs_dr_actions); +free_actions_alloc: + kfree(actions); +out_err: + mlx5_core_err(dev, "Failed to create dr rule err(%d)\n", err); return err; } @@ -533,7 +588,8 @@ static int mlx5_cmd_dr_delete_fte(struct mlx5_flow_root_namespace *ns, if (err) return err; - for (i = 0; i < rule->num_actions; i++) + /* Free in reverse order to handle action dependencies */ + for (i = rule->num_actions - 1; i >= 0; i--) if (!IS_ERR_OR_NULL(rule->dr_actions[i])) mlx5dr_action_destroy(rule->dr_actions[i]); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/steering/mlx5dr.h b/drivers/net/ethernet/mellanox/mlx5/core/steering/mlx5dr.h index adda9cbfba45..e1edc9c247b7 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/steering/mlx5dr.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/steering/mlx5dr.h @@ -33,6 +33,11 @@ struct mlx5dr_match_parameters { u64 *match_buf; /* Device spec format */ }; +struct mlx5dr_action_dest { + struct mlx5dr_action *dest; + struct mlx5dr_action *reformat; +}; + #ifdef CONFIG_MLX5_SW_STEERING struct mlx5dr_domain * @@ -46,7 +51,7 @@ void mlx5dr_domain_set_peer(struct mlx5dr_domain *dmn, struct mlx5dr_domain *peer_dmn); struct mlx5dr_table * -mlx5dr_table_create(struct mlx5dr_domain *domain, u32 level); +mlx5dr_table_create(struct mlx5dr_domain *domain, u32 level, u32 flags); int mlx5dr_table_destroy(struct mlx5dr_table *table); @@ -75,14 +80,19 @@ struct mlx5dr_action * mlx5dr_action_create_dest_table(struct mlx5dr_table *table); struct mlx5dr_action * -mlx5dr_create_action_dest_flow_fw_table(struct mlx5_flow_table *ft, - struct mlx5_core_dev *mdev); +mlx5dr_action_create_dest_flow_fw_table(struct mlx5dr_domain *domain, + struct mlx5_flow_table *ft); struct mlx5dr_action * mlx5dr_action_create_dest_vport(struct mlx5dr_domain *domain, u32 vport, u8 vhca_id_valid, u16 vhca_id); +struct mlx5dr_action * +mlx5dr_action_create_mult_dest_tbl(struct mlx5dr_domain *dmn, + struct mlx5dr_action_dest *dests, + u32 num_of_dests); + struct mlx5dr_action *mlx5dr_action_create_drop(void); struct mlx5dr_action *mlx5dr_action_create_tag(u32 tag_value); @@ -131,7 +141,7 @@ mlx5dr_domain_set_peer(struct mlx5dr_domain *dmn, struct mlx5dr_domain *peer_dmn) { } static inline struct mlx5dr_table * -mlx5dr_table_create(struct mlx5dr_domain *domain, u32 level) { return NULL; } +mlx5dr_table_create(struct mlx5dr_domain *domain, u32 level, u32 flags) { return NULL; } static inline int mlx5dr_table_destroy(struct mlx5dr_table *table) { return 0; } @@ -165,8 +175,8 @@ static inline struct mlx5dr_action * mlx5dr_action_create_dest_table(struct mlx5dr_table *table) { return NULL; } static inline struct mlx5dr_action * -mlx5dr_create_action_dest_flow_fw_table(struct mlx5_flow_table *ft, - struct mlx5_core_dev *mdev) { return NULL; } +mlx5dr_action_create_dest_flow_fw_table(struct mlx5dr_domain *domain, + struct mlx5_flow_table *ft) { return NULL; } static inline struct mlx5dr_action * mlx5dr_action_create_dest_vport(struct mlx5dr_domain *domain, @@ -174,6 +184,11 @@ mlx5dr_action_create_dest_vport(struct mlx5dr_domain *domain, u16 vhca_id) { return NULL; } static inline struct mlx5dr_action * +mlx5dr_action_create_mult_dest_tbl(struct mlx5dr_domain *dmn, + struct mlx5dr_action_dest *dests, + u32 num_of_dests) { return NULL; } + +static inline struct mlx5dr_action * mlx5dr_action_create_drop(void) { return NULL; } static inline struct mlx5dr_action * diff --git a/drivers/net/ethernet/mellanox/mlxsw/reg.h b/drivers/net/ethernet/mellanox/mlxsw/reg.h index fd59280cf979..0b80e75e87c3 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/reg.h +++ b/drivers/net/ethernet/mellanox/mlxsw/reg.h @@ -5513,6 +5513,7 @@ enum mlxsw_reg_htgt_discard_trap_group { MLXSW_REG_HTGT_DISCARD_TRAP_GROUP_BASE = MLXSW_REG_HTGT_TRAP_GROUP_MAX, MLXSW_REG_HTGT_TRAP_GROUP_SP_L2_DISCARDS, MLXSW_REG_HTGT_TRAP_GROUP_SP_L3_DISCARDS, + MLXSW_REG_HTGT_TRAP_GROUP_SP_TUNNEL_DISCARDS, }; /* reg_htgt_trap_group @@ -10140,6 +10141,92 @@ static inline void mlxsw_reg_tigcr_pack(char *payload, bool ttlc, u8 ttl_uc) mlxsw_reg_tigcr_ttl_uc_set(payload, ttl_uc); } +/* TIEEM - Tunneling IPinIP Encapsulation ECN Mapping Register + * ----------------------------------------------------------- + * The TIEEM register maps ECN of the IP header at the ingress to the + * encapsulation to the ECN of the underlay network. + */ +#define MLXSW_REG_TIEEM_ID 0xA812 +#define MLXSW_REG_TIEEM_LEN 0x0C + +MLXSW_REG_DEFINE(tieem, MLXSW_REG_TIEEM_ID, MLXSW_REG_TIEEM_LEN); + +/* reg_tieem_overlay_ecn + * ECN of the IP header in the overlay network. + * Access: Index + */ +MLXSW_ITEM32(reg, tieem, overlay_ecn, 0x04, 24, 2); + +/* reg_tineem_underlay_ecn + * ECN of the IP header in the underlay network. + * Access: RW + */ +MLXSW_ITEM32(reg, tieem, underlay_ecn, 0x04, 16, 2); + +static inline void mlxsw_reg_tieem_pack(char *payload, u8 overlay_ecn, + u8 underlay_ecn) +{ + MLXSW_REG_ZERO(tieem, payload); + mlxsw_reg_tieem_overlay_ecn_set(payload, overlay_ecn); + mlxsw_reg_tieem_underlay_ecn_set(payload, underlay_ecn); +} + +/* TIDEM - Tunneling IPinIP Decapsulation ECN Mapping Register + * ----------------------------------------------------------- + * The TIDEM register configures the actions that are done in the + * decapsulation. + */ +#define MLXSW_REG_TIDEM_ID 0xA813 +#define MLXSW_REG_TIDEM_LEN 0x0C + +MLXSW_REG_DEFINE(tidem, MLXSW_REG_TIDEM_ID, MLXSW_REG_TIDEM_LEN); + +/* reg_tidem_underlay_ecn + * ECN field of the IP header in the underlay network. + * Access: Index + */ +MLXSW_ITEM32(reg, tidem, underlay_ecn, 0x04, 24, 2); + +/* reg_tidem_overlay_ecn + * ECN field of the IP header in the overlay network. + * Access: Index + */ +MLXSW_ITEM32(reg, tidem, overlay_ecn, 0x04, 16, 2); + +/* reg_tidem_eip_ecn + * Egress IP ECN. ECN field of the IP header of the packet which goes out + * from the decapsulation. + * Access: RW + */ +MLXSW_ITEM32(reg, tidem, eip_ecn, 0x04, 8, 2); + +/* reg_tidem_trap_en + * Trap enable: + * 0 - No trap due to decap ECN + * 1 - Trap enable with trap_id + * Access: RW + */ +MLXSW_ITEM32(reg, tidem, trap_en, 0x08, 28, 4); + +/* reg_tidem_trap_id + * Trap ID. Either DECAP_ECN0 or DECAP_ECN1. + * Reserved when trap_en is '0'. + * Access: RW + */ +MLXSW_ITEM32(reg, tidem, trap_id, 0x08, 0, 9); + +static inline void mlxsw_reg_tidem_pack(char *payload, u8 underlay_ecn, + u8 overlay_ecn, u8 eip_ecn, + bool trap_en, u16 trap_id) +{ + MLXSW_REG_ZERO(tidem, payload); + mlxsw_reg_tidem_underlay_ecn_set(payload, underlay_ecn); + mlxsw_reg_tidem_overlay_ecn_set(payload, overlay_ecn); + mlxsw_reg_tidem_eip_ecn_set(payload, eip_ecn); + mlxsw_reg_tidem_trap_en_set(payload, trap_en); + mlxsw_reg_tidem_trap_id_set(payload, trap_id); +} + /* SBPR - Shared Buffer Pools Register * ----------------------------------- * The SBPR configures and retrieves the shared buffer pools and configuration. @@ -10684,6 +10771,8 @@ static const struct mlxsw_reg_info *mlxsw_reg_infos[] = { MLXSW_REG(tndem), MLXSW_REG(tnpc), MLXSW_REG(tigcr), + MLXSW_REG(tieem), + MLXSW_REG(tidem), MLXSW_REG(sbpr), MLXSW_REG(sbcm), MLXSW_REG(sbpm), diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum.c index 431c3765b545..9eb3ac7669f7 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum.c @@ -45,11 +45,9 @@ #include "spectrum_ptp.h" #include "../mlxfw/mlxfw.h" -#define MLXSW_SP_FWREV_MINOR_TO_BRANCH(minor) ((minor) / 100) - #define MLXSW_SP1_FWREV_MAJOR 13 #define MLXSW_SP1_FWREV_MINOR 2000 -#define MLXSW_SP1_FWREV_SUBMINOR 2308 +#define MLXSW_SP1_FWREV_SUBMINOR 2714 #define MLXSW_SP1_FWREV_CAN_RESET_MINOR 1702 static const struct mlxsw_fw_rev mlxsw_sp1_fw_rev = { @@ -66,7 +64,7 @@ static const struct mlxsw_fw_rev mlxsw_sp1_fw_rev = { #define MLXSW_SP2_FWREV_MAJOR 29 #define MLXSW_SP2_FWREV_MINOR 2000 -#define MLXSW_SP2_FWREV_SUBMINOR 2308 +#define MLXSW_SP2_FWREV_SUBMINOR 2714 static const struct mlxsw_fw_rev mlxsw_sp2_fw_rev = { .major = MLXSW_SP2_FWREV_MAJOR, @@ -197,6 +195,10 @@ struct mlxsw_sp_ptp_ops { u64 *data, int data_index); }; +struct mlxsw_sp_span_ops { + u32 (*buffsize_get)(int mtu, u32 speed); +}; + static int mlxsw_sp_component_query(struct mlxfw_dev *mlxfw_dev, u16 component_index, u32 *p_max_size, u8 *p_align_bits, u16 *p_max_write_size) @@ -423,13 +425,12 @@ static int mlxsw_sp_fw_rev_validate(struct mlxsw_sp *mlxsw_sp) rev->major, req_rev->major); return -EINVAL; } - if (MLXSW_SP_FWREV_MINOR_TO_BRANCH(rev->minor) == - MLXSW_SP_FWREV_MINOR_TO_BRANCH(req_rev->minor) && - mlxsw_core_fw_rev_minor_subminor_validate(rev, req_rev)) + if (mlxsw_core_fw_rev_minor_subminor_validate(rev, req_rev)) return 0; - dev_info(mlxsw_sp->bus_info->dev, "The firmware version %d.%d.%d is incompatible with the driver\n", - rev->major, rev->minor, rev->subminor); + dev_err(mlxsw_sp->bus_info->dev, "The firmware version %d.%d.%d is incompatible with the driver (required >= %d.%d.%d)\n", + rev->major, rev->minor, rev->subminor, req_rev->major, + req_rev->minor, req_rev->subminor); dev_info(mlxsw_sp->bus_info->dev, "Flashing firmware using file %s\n", fw_filename); @@ -860,23 +861,17 @@ static netdev_tx_t mlxsw_sp_port_xmit(struct sk_buff *skb, u64 len; int err; + if (skb_cow_head(skb, MLXSW_TXHDR_LEN)) { + this_cpu_inc(mlxsw_sp_port->pcpu_stats->tx_dropped); + dev_kfree_skb_any(skb); + return NETDEV_TX_OK; + } + memset(skb->cb, 0, sizeof(struct mlxsw_skb_cb)); if (mlxsw_core_skb_transmit_busy(mlxsw_sp->core, &tx_info)) return NETDEV_TX_BUSY; - if (unlikely(skb_headroom(skb) < MLXSW_TXHDR_LEN)) { - struct sk_buff *skb_orig = skb; - - skb = skb_realloc_headroom(skb, MLXSW_TXHDR_LEN); - if (!skb) { - this_cpu_inc(mlxsw_sp_port->pcpu_stats->tx_dropped); - dev_kfree_skb_any(skb_orig); - return NETDEV_TX_OK; - } - dev_consume_skb_any(skb_orig); - } - if (eth_skb_pad(skb)) { this_cpu_inc(mlxsw_sp_port->pcpu_stats->tx_dropped); return NETDEV_TX_OK; @@ -1215,6 +1210,9 @@ static void update_stats_cache(struct work_struct *work) periodic_hw_stats.update_dw.work); if (!netif_carrier_ok(mlxsw_sp_port->dev)) + /* Note: mlxsw_sp_port_down_wipe_counters() clears the cache as + * necessary when port goes down. + */ goto out; mlxsw_sp_port_get_hw_stats(mlxsw_sp_port->dev, @@ -3541,6 +3539,27 @@ mlxsw_sp_port_speed_by_width_set(struct mlxsw_sp_port *mlxsw_sp_port) return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ptys), ptys_pl); } +int mlxsw_sp_port_speed_get(struct mlxsw_sp_port *mlxsw_sp_port, u32 *speed) +{ + const struct mlxsw_sp_port_type_speed_ops *port_type_speed_ops; + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + char ptys_pl[MLXSW_REG_PTYS_LEN]; + u32 eth_proto_oper; + int err; + + port_type_speed_ops = mlxsw_sp->port_type_speed_ops; + port_type_speed_ops->reg_ptys_eth_pack(mlxsw_sp, ptys_pl, + mlxsw_sp_port->local_port, 0, + false); + err = mlxsw_reg_query(mlxsw_sp->core, MLXSW_REG(ptys), ptys_pl); + if (err) + return err; + port_type_speed_ops->reg_ptys_eth_unpack(mlxsw_sp, ptys_pl, NULL, NULL, + ð_proto_oper); + *speed = port_type_speed_ops->from_ptys_speed(mlxsw_sp, eth_proto_oper); + return 0; +} + int mlxsw_sp_port_ets_set(struct mlxsw_sp_port *mlxsw_sp_port, enum mlxsw_reg_qeec_hr hr, u8 index, u8 next_index, bool dwrr, u8 dwrr_weight) @@ -3889,6 +3908,8 @@ static int mlxsw_sp_port_create(struct mlxsw_sp *mlxsw_sp, u8 local_port, INIT_DELAYED_WORK(&mlxsw_sp_port->ptp.shaper_dw, mlxsw_sp->ptp_ops->shaper_work); + INIT_DELAYED_WORK(&mlxsw_sp_port->span.speed_update_dw, + mlxsw_sp_span_speed_update_work); mlxsw_sp->ports[local_port] = mlxsw_sp_port; err = register_netdev(dev); @@ -3945,6 +3966,7 @@ static void mlxsw_sp_port_remove(struct mlxsw_sp *mlxsw_sp, u8 local_port) struct mlxsw_sp_port *mlxsw_sp_port = mlxsw_sp->ports[local_port]; cancel_delayed_work_sync(&mlxsw_sp_port->periodic_hw_stats.update_dw); + cancel_delayed_work_sync(&mlxsw_sp_port->span.speed_update_dw); cancel_delayed_work_sync(&mlxsw_sp_port->ptp.shaper_dw); mlxsw_sp_port_ptp_clear(mlxsw_sp_port); mlxsw_core_port_clear(mlxsw_sp->core, local_port, mlxsw_sp); @@ -4325,6 +4347,15 @@ static int mlxsw_sp_port_unsplit(struct mlxsw_core *mlxsw_core, u8 local_port, return 0; } +static void +mlxsw_sp_port_down_wipe_counters(struct mlxsw_sp_port *mlxsw_sp_port) +{ + int i; + + for (i = 0; i < TC_MAX_QUEUE; i++) + mlxsw_sp_port->periodic_hw_stats.xstats.backlog[i] = 0; +} + static void mlxsw_sp_pude_event_func(const struct mlxsw_reg_info *reg, char *pude_pl, void *priv) { @@ -4343,9 +4374,11 @@ static void mlxsw_sp_pude_event_func(const struct mlxsw_reg_info *reg, netdev_info(mlxsw_sp_port->dev, "link up\n"); netif_carrier_on(mlxsw_sp_port->dev); mlxsw_core_schedule_dw(&mlxsw_sp_port->ptp.shaper_dw, 0); + mlxsw_core_schedule_dw(&mlxsw_sp_port->span.speed_update_dw, 0); } else { netdev_info(mlxsw_sp_port->dev, "link down\n"); netif_carrier_off(mlxsw_sp_port->dev); + mlxsw_sp_port_down_wipe_counters(mlxsw_sp_port); } } @@ -4541,8 +4574,6 @@ static const struct mlxsw_listener mlxsw_sp_listener[] = { false), MLXSW_SP_RXL_MARK(ROUTER_ALERT_IPV4, TRAP_TO_CPU, ROUTER_EXP, false), MLXSW_SP_RXL_MARK(ROUTER_ALERT_IPV6, TRAP_TO_CPU, ROUTER_EXP, false), - MLXSW_SP_RXL_MARK(IPIP_DECAP_ERROR, TRAP_TO_CPU, ROUTER_EXP, false), - MLXSW_SP_RXL_MARK(DECAP_ECN0, TRAP_TO_CPU, ROUTER_EXP, false), MLXSW_SP_RXL_MARK(IPV4_VRRP, TRAP_TO_CPU, VRRP, false), MLXSW_SP_RXL_MARK(IPV6_VRRP, TRAP_TO_CPU, VRRP, false), MLXSW_SP_RXL_NO_MARK(DISCARD_ING_ROUTER_SIP_CLASS_E, FORWARD, @@ -4891,6 +4922,33 @@ static const struct mlxsw_sp_ptp_ops mlxsw_sp2_ptp_ops = { .get_stats = mlxsw_sp2_get_stats, }; +static u32 mlxsw_sp1_span_buffsize_get(int mtu, u32 speed) +{ + return mtu * 5 / 2; +} + +static const struct mlxsw_sp_span_ops mlxsw_sp1_span_ops = { + .buffsize_get = mlxsw_sp1_span_buffsize_get, +}; + +#define MLXSW_SP2_SPAN_EG_MIRROR_BUFFER_FACTOR 38 + +static u32 mlxsw_sp2_span_buffsize_get(int mtu, u32 speed) +{ + return 3 * mtu + MLXSW_SP2_SPAN_EG_MIRROR_BUFFER_FACTOR * speed / 1000; +} + +static const struct mlxsw_sp_span_ops mlxsw_sp2_span_ops = { + .buffsize_get = mlxsw_sp2_span_buffsize_get, +}; + +u32 mlxsw_sp_span_buffsize_get(struct mlxsw_sp *mlxsw_sp, int mtu, u32 speed) +{ + u32 buffsize = mlxsw_sp->span_ops->buffsize_get(speed, mtu); + + return mlxsw_sp_bytes_cells(mlxsw_sp, buffsize) + 1; +} + static int mlxsw_sp_netdevice_event(struct notifier_block *unused, unsigned long event, void *ptr); @@ -5112,6 +5170,7 @@ static int mlxsw_sp1_init(struct mlxsw_core *mlxsw_core, mlxsw_sp->sb_vals = &mlxsw_sp1_sb_vals; mlxsw_sp->port_type_speed_ops = &mlxsw_sp1_port_type_speed_ops; mlxsw_sp->ptp_ops = &mlxsw_sp1_ptp_ops; + mlxsw_sp->span_ops = &mlxsw_sp1_span_ops; mlxsw_sp->listeners = mlxsw_sp1_listener; mlxsw_sp->listeners_count = ARRAY_SIZE(mlxsw_sp1_listener); @@ -5137,6 +5196,29 @@ static int mlxsw_sp2_init(struct mlxsw_core *mlxsw_core, mlxsw_sp->sb_vals = &mlxsw_sp2_sb_vals; mlxsw_sp->port_type_speed_ops = &mlxsw_sp2_port_type_speed_ops; mlxsw_sp->ptp_ops = &mlxsw_sp2_ptp_ops; + mlxsw_sp->span_ops = &mlxsw_sp2_span_ops; + + return mlxsw_sp_init(mlxsw_core, mlxsw_bus_info, extack); +} + +static int mlxsw_sp3_init(struct mlxsw_core *mlxsw_core, + const struct mlxsw_bus_info *mlxsw_bus_info, + struct netlink_ext_ack *extack) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_core_driver_priv(mlxsw_core); + + mlxsw_sp->kvdl_ops = &mlxsw_sp2_kvdl_ops; + mlxsw_sp->afa_ops = &mlxsw_sp2_act_afa_ops; + mlxsw_sp->afk_ops = &mlxsw_sp2_afk_ops; + mlxsw_sp->mr_tcam_ops = &mlxsw_sp2_mr_tcam_ops; + mlxsw_sp->acl_tcam_ops = &mlxsw_sp2_acl_tcam_ops; + mlxsw_sp->nve_ops_arr = mlxsw_sp2_nve_ops_arr; + mlxsw_sp->mac_mask = mlxsw_sp2_mac_mask; + mlxsw_sp->rif_ops_arr = mlxsw_sp2_rif_ops_arr; + mlxsw_sp->sb_vals = &mlxsw_sp2_sb_vals; + mlxsw_sp->port_type_speed_ops = &mlxsw_sp2_port_type_speed_ops; + mlxsw_sp->ptp_ops = &mlxsw_sp2_ptp_ops; + mlxsw_sp->span_ops = &mlxsw_sp2_span_ops; return mlxsw_sp_init(mlxsw_core, mlxsw_bus_info, extack); } @@ -5643,7 +5725,7 @@ static struct mlxsw_driver mlxsw_sp2_driver = { static struct mlxsw_driver mlxsw_sp3_driver = { .kind = mlxsw_sp3_driver_name, .priv_size = sizeof(struct mlxsw_sp), - .init = mlxsw_sp2_init, + .init = mlxsw_sp3_init, .fini = mlxsw_sp_fini, .basic_trap_groups_set = mlxsw_sp_basic_trap_groups_set, .port_split = mlxsw_sp_port_split, diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum.h b/drivers/net/ethernet/mellanox/mlxsw/spectrum.h index 948ef4720d40..5f3b74360dc8 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum.h +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum.h @@ -140,6 +140,7 @@ struct mlxsw_sp_sb_vals; struct mlxsw_sp_port_type_speed_ops; struct mlxsw_sp_ptp_state; struct mlxsw_sp_ptp_ops; +struct mlxsw_sp_span_ops; struct mlxsw_sp_port_mapping { u8 module; @@ -185,6 +186,7 @@ struct mlxsw_sp { const struct mlxsw_sp_sb_vals *sb_vals; const struct mlxsw_sp_port_type_speed_ops *port_type_speed_ops; const struct mlxsw_sp_ptp_ops *ptp_ops; + const struct mlxsw_sp_span_ops *span_ops; const struct mlxsw_listener *listeners; size_t listeners_count; }; @@ -292,6 +294,9 @@ struct mlxsw_sp_port { struct mlxsw_sp_ptp_port_stats stats; } ptp; u8 split_base_local_port; + struct { + struct delayed_work speed_update_dw; + } span; }; struct mlxsw_sp_port_type_speed_ops { @@ -471,6 +476,7 @@ extern struct notifier_block mlxsw_sp_switchdev_notifier; /* spectrum.c */ void mlxsw_sp_rx_listener_no_mark_func(struct sk_buff *skb, u8 local_port, void *priv); +int mlxsw_sp_port_speed_get(struct mlxsw_sp_port *mlxsw_sp_port, u32 *speed); int mlxsw_sp_port_ets_set(struct mlxsw_sp_port *mlxsw_sp_port, enum mlxsw_reg_qeec_hr hr, u8 index, u8 next_index, bool dwrr, u8 dwrr_weight); @@ -501,6 +507,7 @@ int mlxsw_sp_flow_counter_alloc(struct mlxsw_sp *mlxsw_sp, unsigned int *p_counter_index); void mlxsw_sp_flow_counter_free(struct mlxsw_sp *mlxsw_sp, unsigned int counter_index); +u32 mlxsw_sp_span_buffsize_get(struct mlxsw_sp *mlxsw_sp, int mtu, u32 speed); bool mlxsw_sp_port_dev_check(const struct net_device *dev); struct mlxsw_sp *mlxsw_sp_lower_get(struct net_device *dev); struct mlxsw_sp_port *mlxsw_sp_port_dev_lower_find(struct net_device *dev); diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.c index 6400cd644b7a..a8525992528f 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.c @@ -3,8 +3,10 @@ #include <net/ip_tunnels.h> #include <net/ip6_tunnel.h> +#include <net/inet_ecn.h> #include "spectrum_ipip.h" +#include "reg.h" struct ip_tunnel_parm mlxsw_sp_ipip_netdev_parms4(const struct net_device *ol_dev) @@ -338,3 +340,61 @@ static const struct mlxsw_sp_ipip_ops mlxsw_sp_ipip_gre4_ops = { const struct mlxsw_sp_ipip_ops *mlxsw_sp_ipip_ops_arr[] = { [MLXSW_SP_IPIP_TYPE_GRE4] = &mlxsw_sp_ipip_gre4_ops, }; + +static int mlxsw_sp_ipip_ecn_encap_init_one(struct mlxsw_sp *mlxsw_sp, + u8 inner_ecn, u8 outer_ecn) +{ + char tieem_pl[MLXSW_REG_TIEEM_LEN]; + + mlxsw_reg_tieem_pack(tieem_pl, inner_ecn, outer_ecn); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(tieem), tieem_pl); +} + +int mlxsw_sp_ipip_ecn_encap_init(struct mlxsw_sp *mlxsw_sp) +{ + int i; + + /* Iterate over inner ECN values */ + for (i = INET_ECN_NOT_ECT; i <= INET_ECN_CE; i++) { + u8 outer_ecn = INET_ECN_encapsulate(0, i); + int err; + + err = mlxsw_sp_ipip_ecn_encap_init_one(mlxsw_sp, i, outer_ecn); + if (err) + return err; + } + + return 0; +} + +static int mlxsw_sp_ipip_ecn_decap_init_one(struct mlxsw_sp *mlxsw_sp, + u8 inner_ecn, u8 outer_ecn) +{ + char tidem_pl[MLXSW_REG_TIDEM_LEN]; + bool trap_en, set_ce = false; + u8 new_inner_ecn; + + trap_en = __INET_ECN_decapsulate(outer_ecn, inner_ecn, &set_ce); + new_inner_ecn = set_ce ? INET_ECN_CE : inner_ecn; + + mlxsw_reg_tidem_pack(tidem_pl, outer_ecn, inner_ecn, new_inner_ecn, + trap_en, trap_en ? MLXSW_TRAP_ID_DECAP_ECN0 : 0); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(tidem), tidem_pl); +} + +int mlxsw_sp_ipip_ecn_decap_init(struct mlxsw_sp *mlxsw_sp) +{ + int i, j, err; + + /* Iterate over inner ECN values */ + for (i = INET_ECN_NOT_ECT; i <= INET_ECN_CE; i++) { + /* Iterate over outer ECN values */ + for (j = INET_ECN_NOT_ECT; j <= INET_ECN_CE; j++) { + err = mlxsw_sp_ipip_ecn_decap_init_one(mlxsw_sp, i, j); + if (err) + return err; + } + } + + return 0; +} diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_ptp.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ptp.c index 4aaaa4937b1a..34f7c3501b08 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_ptp.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ptp.c @@ -1016,27 +1016,17 @@ mlxsw_sp1_ptp_port_shaper_set(struct mlxsw_sp_port *mlxsw_sp_port, bool enable) static int mlxsw_sp1_ptp_port_shaper_check(struct mlxsw_sp_port *mlxsw_sp_port) { - const struct mlxsw_sp_port_type_speed_ops *port_type_speed_ops; - struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; - char ptys_pl[MLXSW_REG_PTYS_LEN]; - u32 eth_proto_oper, speed; bool ptps = false; int err, i; + u32 speed; if (!mlxsw_sp1_ptp_hwtstamp_enabled(mlxsw_sp_port)) return mlxsw_sp1_ptp_port_shaper_set(mlxsw_sp_port, false); - port_type_speed_ops = mlxsw_sp->port_type_speed_ops; - port_type_speed_ops->reg_ptys_eth_pack(mlxsw_sp, ptys_pl, - mlxsw_sp_port->local_port, 0, - false); - err = mlxsw_reg_query(mlxsw_sp->core, MLXSW_REG(ptys), ptys_pl); + err = mlxsw_sp_port_speed_get(mlxsw_sp_port, &speed); if (err) return err; - port_type_speed_ops->reg_ptys_eth_unpack(mlxsw_sp, ptys_pl, NULL, NULL, - ð_proto_oper); - speed = port_type_speed_ops->from_ptys_speed(mlxsw_sp, eth_proto_oper); for (i = 0; i < MLXSW_SP1_PTP_SHAPER_PARAMS_LEN; i++) { if (mlxsw_sp1_ptp_shaper_params[i].ethtool_speed == speed) { ptps = true; diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_qdisc.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_qdisc.c index 81a2c087f534..d57c9b15f45e 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_qdisc.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_qdisc.c @@ -196,6 +196,20 @@ mlxsw_sp_qdisc_get_xstats(struct mlxsw_sp_port *mlxsw_sp_port, return -EOPNOTSUPP; } +static u64 +mlxsw_sp_xstats_backlog(struct mlxsw_sp_port_xstats *xstats, int tclass_num) +{ + return xstats->backlog[tclass_num] + + xstats->backlog[tclass_num + 8]; +} + +static u64 +mlxsw_sp_xstats_tail_drop(struct mlxsw_sp_port_xstats *xstats, int tclass_num) +{ + return xstats->tail_drop[tclass_num] + + xstats->tail_drop[tclass_num + 8]; +} + static void mlxsw_sp_qdisc_bstats_per_priority_get(struct mlxsw_sp_port_xstats *xstats, u8 prio_bitmap, u64 *tx_packets, @@ -270,7 +284,7 @@ mlxsw_sp_setup_tc_qdisc_red_clean_stats(struct mlxsw_sp_port *mlxsw_sp_port, &stats_base->tx_bytes); red_base->prob_mark = xstats->ecn; red_base->prob_drop = xstats->wred_drop[tclass_num]; - red_base->pdrop = xstats->tail_drop[tclass_num]; + red_base->pdrop = mlxsw_sp_xstats_tail_drop(xstats, tclass_num); stats_base->overlimits = red_base->prob_drop + red_base->prob_mark; stats_base->drops = red_base->prob_drop + red_base->pdrop; @@ -371,7 +385,8 @@ mlxsw_sp_qdisc_get_red_xstats(struct mlxsw_sp_port *mlxsw_sp_port, early_drops = xstats->wred_drop[tclass_num] - xstats_base->prob_drop; marks = xstats->ecn - xstats_base->prob_mark; - pdrops = xstats->tail_drop[tclass_num] - xstats_base->pdrop; + pdrops = mlxsw_sp_xstats_tail_drop(xstats, tclass_num) - + xstats_base->pdrop; res->pdrop += pdrops; res->prob_drop += early_drops; @@ -404,9 +419,10 @@ mlxsw_sp_qdisc_get_red_stats(struct mlxsw_sp_port *mlxsw_sp_port, overlimits = xstats->wred_drop[tclass_num] + xstats->ecn - stats_base->overlimits; - drops = xstats->wred_drop[tclass_num] + xstats->tail_drop[tclass_num] - + drops = xstats->wred_drop[tclass_num] + + mlxsw_sp_xstats_tail_drop(xstats, tclass_num) - stats_base->drops; - backlog = xstats->backlog[tclass_num]; + backlog = mlxsw_sp_xstats_backlog(xstats, tclass_num); _bstats_update(stats_ptr->bstats, tx_bytes, tx_packets); stats_ptr->qstats->overlimits += overlimits; @@ -626,9 +642,9 @@ mlxsw_sp_qdisc_get_prio_stats(struct mlxsw_sp_port *mlxsw_sp_port, tx_packets = stats->tx_packets - stats_base->tx_packets; for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++) { - drops += xstats->tail_drop[i]; + drops += mlxsw_sp_xstats_tail_drop(xstats, i); drops += xstats->wred_drop[i]; - backlog += xstats->backlog[i]; + backlog += mlxsw_sp_xstats_backlog(xstats, i); } drops = drops - stats_base->drops; @@ -664,7 +680,7 @@ mlxsw_sp_setup_tc_qdisc_prio_clean_stats(struct mlxsw_sp_port *mlxsw_sp_port, stats_base->drops = 0; for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++) { - stats_base->drops += xstats->tail_drop[i]; + stats_base->drops += mlxsw_sp_xstats_tail_drop(xstats, i); stats_base->drops += xstats->wred_drop[i]; } @@ -767,6 +783,13 @@ __mlxsw_sp_qdisc_ets_graft(struct mlxsw_sp_port *mlxsw_sp_port, mlxsw_sp_port->tclass_qdiscs[tclass_num].handle == child_handle) return 0; + if (!child_handle) { + /* This is an invisible FIFO replacing the original Qdisc. + * Ignore it--the original Qdisc's destroy will follow. + */ + return 0; + } + /* See if the grafted qdisc is already offloaded on any tclass. If so, * unoffload it. */ diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c index 7f70aa799064..ce707723f8cf 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c @@ -3235,20 +3235,6 @@ mlxsw_sp_nexthop_fib_entries_update(struct mlxsw_sp *mlxsw_sp, return 0; } -static void -mlxsw_sp_fib_entry_offload_refresh(struct mlxsw_sp_fib_entry *fib_entry, - enum mlxsw_reg_ralue_op op, int err); - -static void -mlxsw_sp_nexthop_fib_entries_refresh(struct mlxsw_sp_nexthop_group *nh_grp) -{ - enum mlxsw_reg_ralue_op op = MLXSW_REG_RALUE_OP_WRITE_WRITE; - struct mlxsw_sp_fib_entry *fib_entry; - - list_for_each_entry(fib_entry, &nh_grp->fib_list, nexthop_group_node) - mlxsw_sp_fib_entry_offload_refresh(fib_entry, op, 0); -} - static void mlxsw_sp_adj_grp_size_round_up(u16 *p_adj_grp_size) { /* Valid sizes for an adjacency group are: @@ -3352,6 +3338,73 @@ mlxsw_sp_nexthop_group_rebalance(struct mlxsw_sp_nexthop_group *nh_grp) } } +static struct mlxsw_sp_nexthop * +mlxsw_sp_rt6_nexthop(struct mlxsw_sp_nexthop_group *nh_grp, + const struct mlxsw_sp_rt6 *mlxsw_sp_rt6); + +static void +mlxsw_sp_nexthop4_group_offload_refresh(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop_group *nh_grp) +{ + int i; + + for (i = 0; i < nh_grp->count; i++) { + struct mlxsw_sp_nexthop *nh = &nh_grp->nexthops[i]; + + if (nh->offloaded) + nh->key.fib_nh->fib_nh_flags |= RTNH_F_OFFLOAD; + else + nh->key.fib_nh->fib_nh_flags &= ~RTNH_F_OFFLOAD; + } +} + +static void +__mlxsw_sp_nexthop6_group_offload_refresh(struct mlxsw_sp_nexthop_group *nh_grp, + struct mlxsw_sp_fib6_entry *fib6_entry) +{ + struct mlxsw_sp_rt6 *mlxsw_sp_rt6; + + list_for_each_entry(mlxsw_sp_rt6, &fib6_entry->rt6_list, list) { + struct fib6_nh *fib6_nh = mlxsw_sp_rt6->rt->fib6_nh; + struct mlxsw_sp_nexthop *nh; + + nh = mlxsw_sp_rt6_nexthop(nh_grp, mlxsw_sp_rt6); + if (nh && nh->offloaded) + fib6_nh->fib_nh_flags |= RTNH_F_OFFLOAD; + else + fib6_nh->fib_nh_flags &= ~RTNH_F_OFFLOAD; + } +} + +static void +mlxsw_sp_nexthop6_group_offload_refresh(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop_group *nh_grp) +{ + struct mlxsw_sp_fib6_entry *fib6_entry; + + /* Unfortunately, in IPv6 the route and the nexthop are described by + * the same struct, so we need to iterate over all the routes using the + * nexthop group and set / clear the offload indication for them. + */ + list_for_each_entry(fib6_entry, &nh_grp->fib_list, + common.nexthop_group_node) + __mlxsw_sp_nexthop6_group_offload_refresh(nh_grp, fib6_entry); +} + +static void +mlxsw_sp_nexthop_group_offload_refresh(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop_group *nh_grp) +{ + switch (mlxsw_sp_nexthop_group_type(nh_grp)) { + case AF_INET: + mlxsw_sp_nexthop4_group_offload_refresh(mlxsw_sp, nh_grp); + break; + case AF_INET6: + mlxsw_sp_nexthop6_group_offload_refresh(mlxsw_sp, nh_grp); + break; + } +} + static void mlxsw_sp_nexthop_group_refresh(struct mlxsw_sp *mlxsw_sp, struct mlxsw_sp_nexthop_group *nh_grp) @@ -3425,6 +3478,8 @@ mlxsw_sp_nexthop_group_refresh(struct mlxsw_sp *mlxsw_sp, goto set_trap; } + mlxsw_sp_nexthop_group_offload_refresh(mlxsw_sp, nh_grp); + if (!old_adj_index_valid) { /* The trap was set for fib entries, so we have to call * fib entry update to unset it and use adjacency index. @@ -3446,9 +3501,6 @@ mlxsw_sp_nexthop_group_refresh(struct mlxsw_sp *mlxsw_sp, goto set_trap; } - /* Offload state within the group changed, so update the flags. */ - mlxsw_sp_nexthop_fib_entries_refresh(nh_grp); - return; set_trap: @@ -3461,6 +3513,7 @@ set_trap: err = mlxsw_sp_nexthop_fib_entries_update(mlxsw_sp, nh_grp); if (err) dev_warn(mlxsw_sp->bus_info->dev, "Failed to set traps for fib entries.\n"); + mlxsw_sp_nexthop_group_offload_refresh(mlxsw_sp, nh_grp); if (old_adj_index_valid) mlxsw_sp_kvdl_free(mlxsw_sp, MLXSW_SP_KVDL_ENTRY_TYPE_ADJ, nh_grp->ecmp_size, nh_grp->adj_index); @@ -4043,131 +4096,128 @@ mlxsw_sp_rt6_nexthop(struct mlxsw_sp_nexthop_group *nh_grp, } static void -mlxsw_sp_fib4_entry_offload_set(struct mlxsw_sp_fib_entry *fib_entry) +mlxsw_sp_fib4_entry_hw_flags_set(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry) { - struct mlxsw_sp_nexthop_group *nh_grp = fib_entry->nh_group; - int i; - - if (fib_entry->type == MLXSW_SP_FIB_ENTRY_TYPE_LOCAL || - fib_entry->type == MLXSW_SP_FIB_ENTRY_TYPE_BLACKHOLE || - fib_entry->type == MLXSW_SP_FIB_ENTRY_TYPE_IPIP_DECAP || - fib_entry->type == MLXSW_SP_FIB_ENTRY_TYPE_NVE_DECAP) { - nh_grp->nexthops->key.fib_nh->fib_nh_flags |= RTNH_F_OFFLOAD; - return; - } - - for (i = 0; i < nh_grp->count; i++) { - struct mlxsw_sp_nexthop *nh = &nh_grp->nexthops[i]; + struct fib_info *fi = mlxsw_sp_nexthop4_group_fi(fib_entry->nh_group); + u32 *p_dst = (u32 *) fib_entry->fib_node->key.addr; + int dst_len = fib_entry->fib_node->key.prefix_len; + struct mlxsw_sp_fib4_entry *fib4_entry; + struct fib_rt_info fri; + bool should_offload; - if (nh->offloaded) - nh->key.fib_nh->fib_nh_flags |= RTNH_F_OFFLOAD; - else - nh->key.fib_nh->fib_nh_flags &= ~RTNH_F_OFFLOAD; - } + should_offload = mlxsw_sp_fib_entry_should_offload(fib_entry); + fib4_entry = container_of(fib_entry, struct mlxsw_sp_fib4_entry, + common); + fri.fi = fi; + fri.tb_id = fib4_entry->tb_id; + fri.dst = cpu_to_be32(*p_dst); + fri.dst_len = dst_len; + fri.tos = fib4_entry->tos; + fri.type = fib4_entry->type; + fri.offload = should_offload; + fri.trap = !should_offload; + fib_alias_hw_flags_set(mlxsw_sp_net(mlxsw_sp), &fri); } static void -mlxsw_sp_fib4_entry_offload_unset(struct mlxsw_sp_fib_entry *fib_entry) +mlxsw_sp_fib4_entry_hw_flags_clear(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry) { - struct mlxsw_sp_nexthop_group *nh_grp = fib_entry->nh_group; - int i; - - if (!list_is_singular(&nh_grp->fib_list)) - return; - - for (i = 0; i < nh_grp->count; i++) { - struct mlxsw_sp_nexthop *nh = &nh_grp->nexthops[i]; + struct fib_info *fi = mlxsw_sp_nexthop4_group_fi(fib_entry->nh_group); + u32 *p_dst = (u32 *) fib_entry->fib_node->key.addr; + int dst_len = fib_entry->fib_node->key.prefix_len; + struct mlxsw_sp_fib4_entry *fib4_entry; + struct fib_rt_info fri; - nh->key.fib_nh->fib_nh_flags &= ~RTNH_F_OFFLOAD; - } + fib4_entry = container_of(fib_entry, struct mlxsw_sp_fib4_entry, + common); + fri.fi = fi; + fri.tb_id = fib4_entry->tb_id; + fri.dst = cpu_to_be32(*p_dst); + fri.dst_len = dst_len; + fri.tos = fib4_entry->tos; + fri.type = fib4_entry->type; + fri.offload = false; + fri.trap = false; + fib_alias_hw_flags_set(mlxsw_sp_net(mlxsw_sp), &fri); } static void -mlxsw_sp_fib6_entry_offload_set(struct mlxsw_sp_fib_entry *fib_entry) +mlxsw_sp_fib6_entry_hw_flags_set(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry) { struct mlxsw_sp_fib6_entry *fib6_entry; struct mlxsw_sp_rt6 *mlxsw_sp_rt6; + bool should_offload; + should_offload = mlxsw_sp_fib_entry_should_offload(fib_entry); + + /* In IPv6 a multipath route is represented using multiple routes, so + * we need to set the flags on all of them. + */ fib6_entry = container_of(fib_entry, struct mlxsw_sp_fib6_entry, common); - - if (fib_entry->type == MLXSW_SP_FIB_ENTRY_TYPE_LOCAL || - fib_entry->type == MLXSW_SP_FIB_ENTRY_TYPE_BLACKHOLE) { - list_first_entry(&fib6_entry->rt6_list, struct mlxsw_sp_rt6, - list)->rt->fib6_nh->fib_nh_flags |= RTNH_F_OFFLOAD; - return; - } - - list_for_each_entry(mlxsw_sp_rt6, &fib6_entry->rt6_list, list) { - struct mlxsw_sp_nexthop_group *nh_grp = fib_entry->nh_group; - struct fib6_nh *fib6_nh = mlxsw_sp_rt6->rt->fib6_nh; - struct mlxsw_sp_nexthop *nh; - - nh = mlxsw_sp_rt6_nexthop(nh_grp, mlxsw_sp_rt6); - if (nh && nh->offloaded) - fib6_nh->fib_nh_flags |= RTNH_F_OFFLOAD; - else - fib6_nh->fib_nh_flags &= ~RTNH_F_OFFLOAD; - } + list_for_each_entry(mlxsw_sp_rt6, &fib6_entry->rt6_list, list) + fib6_info_hw_flags_set(mlxsw_sp_rt6->rt, should_offload, + !should_offload); } static void -mlxsw_sp_fib6_entry_offload_unset(struct mlxsw_sp_fib_entry *fib_entry) +mlxsw_sp_fib6_entry_hw_flags_clear(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry) { struct mlxsw_sp_fib6_entry *fib6_entry; struct mlxsw_sp_rt6 *mlxsw_sp_rt6; fib6_entry = container_of(fib_entry, struct mlxsw_sp_fib6_entry, common); - list_for_each_entry(mlxsw_sp_rt6, &fib6_entry->rt6_list, list) { - struct fib6_info *rt = mlxsw_sp_rt6->rt; - - rt->fib6_nh->fib_nh_flags &= ~RTNH_F_OFFLOAD; - } + list_for_each_entry(mlxsw_sp_rt6, &fib6_entry->rt6_list, list) + fib6_info_hw_flags_set(mlxsw_sp_rt6->rt, false, false); } -static void mlxsw_sp_fib_entry_offload_set(struct mlxsw_sp_fib_entry *fib_entry) +static void +mlxsw_sp_fib_entry_hw_flags_set(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry) { switch (fib_entry->fib_node->fib->proto) { case MLXSW_SP_L3_PROTO_IPV4: - mlxsw_sp_fib4_entry_offload_set(fib_entry); + mlxsw_sp_fib4_entry_hw_flags_set(mlxsw_sp, fib_entry); break; case MLXSW_SP_L3_PROTO_IPV6: - mlxsw_sp_fib6_entry_offload_set(fib_entry); + mlxsw_sp_fib6_entry_hw_flags_set(mlxsw_sp, fib_entry); break; } } static void -mlxsw_sp_fib_entry_offload_unset(struct mlxsw_sp_fib_entry *fib_entry) +mlxsw_sp_fib_entry_hw_flags_clear(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry) { switch (fib_entry->fib_node->fib->proto) { case MLXSW_SP_L3_PROTO_IPV4: - mlxsw_sp_fib4_entry_offload_unset(fib_entry); + mlxsw_sp_fib4_entry_hw_flags_clear(mlxsw_sp, fib_entry); break; case MLXSW_SP_L3_PROTO_IPV6: - mlxsw_sp_fib6_entry_offload_unset(fib_entry); + mlxsw_sp_fib6_entry_hw_flags_clear(mlxsw_sp, fib_entry); break; } } static void -mlxsw_sp_fib_entry_offload_refresh(struct mlxsw_sp_fib_entry *fib_entry, - enum mlxsw_reg_ralue_op op, int err) +mlxsw_sp_fib_entry_hw_flags_refresh(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry, + enum mlxsw_reg_ralue_op op) { switch (op) { - case MLXSW_REG_RALUE_OP_WRITE_DELETE: - return mlxsw_sp_fib_entry_offload_unset(fib_entry); case MLXSW_REG_RALUE_OP_WRITE_WRITE: - if (err) - return; - if (mlxsw_sp_fib_entry_should_offload(fib_entry)) - mlxsw_sp_fib_entry_offload_set(fib_entry); - else - mlxsw_sp_fib_entry_offload_unset(fib_entry); - return; + mlxsw_sp_fib_entry_hw_flags_set(mlxsw_sp, fib_entry); + break; + case MLXSW_REG_RALUE_OP_WRITE_DELETE: + mlxsw_sp_fib_entry_hw_flags_clear(mlxsw_sp, fib_entry); + break; default: - return; + break; } } @@ -4394,7 +4444,10 @@ static int mlxsw_sp_fib_entry_op(struct mlxsw_sp *mlxsw_sp, { int err = __mlxsw_sp_fib_entry_op(mlxsw_sp, fib_entry, op); - mlxsw_sp_fib_entry_offload_refresh(fib_entry, op, err); + if (err) + return err; + + mlxsw_sp_fib_entry_hw_flags_refresh(mlxsw_sp, fib_entry, op); return err; } @@ -4830,7 +4883,7 @@ mlxsw_sp_router_fib4_replace(struct mlxsw_sp *mlxsw_sp, if (!replaced) return 0; - mlxsw_sp_fib_entry_offload_unset(replaced); + mlxsw_sp_fib_entry_hw_flags_clear(mlxsw_sp, replaced); fib4_replaced = container_of(replaced, struct mlxsw_sp_fib4_entry, common); mlxsw_sp_fib4_entry_destroy(mlxsw_sp, fib4_replaced); @@ -5113,6 +5166,11 @@ static int mlxsw_sp_nexthop6_group_get(struct mlxsw_sp *mlxsw_sp, &nh_grp->fib_list); fib6_entry->common.nh_group = nh_grp; + /* The route and the nexthop are described by the same struct, so we + * need to the update the nexthop offload indication for the new route. + */ + __mlxsw_sp_nexthop6_group_offload_refresh(nh_grp, fib6_entry); + return 0; } @@ -5393,7 +5451,7 @@ static int mlxsw_sp_router_fib6_replace(struct mlxsw_sp *mlxsw_sp, if (!replaced) return 0; - mlxsw_sp_fib_entry_offload_unset(replaced); + mlxsw_sp_fib_entry_hw_flags_clear(mlxsw_sp, replaced); fib6_replaced = container_of(replaced, struct mlxsw_sp_fib6_entry, common); mlxsw_sp_fib6_entry_destroy(mlxsw_sp, fib6_replaced); @@ -7733,8 +7791,18 @@ mlxsw_sp_ipip_config_tigcr(struct mlxsw_sp *mlxsw_sp) static int mlxsw_sp_ipips_init(struct mlxsw_sp *mlxsw_sp) { + int err; + mlxsw_sp->router->ipip_ops_arr = mlxsw_sp_ipip_ops_arr; INIT_LIST_HEAD(&mlxsw_sp->router->ipip_list); + + err = mlxsw_sp_ipip_ecn_encap_init(mlxsw_sp); + if (err) + return err; + err = mlxsw_sp_ipip_ecn_decap_init(mlxsw_sp); + if (err) + return err; + return mlxsw_sp_ipip_config_tigcr(mlxsw_sp); } diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.h b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.h index cc1de91e8217..c9b94f435cdd 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.h +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.h @@ -104,4 +104,7 @@ static inline bool mlxsw_sp_l3addr_eq(const union mlxsw_sp_l3addr *addr1, return !memcmp(addr1, addr2, sizeof(*addr1)); } +int mlxsw_sp_ipip_ecn_encap_init(struct mlxsw_sp *mlxsw_sp); +int mlxsw_sp_ipip_ecn_decap_init(struct mlxsw_sp *mlxsw_sp); + #endif /* _MLXSW_ROUTER_H_*/ diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_span.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_span.c index 200d324e6d99..0cdd7954a085 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_span.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_span.c @@ -748,33 +748,50 @@ static bool mlxsw_sp_span_is_egress_mirror(struct mlxsw_sp_port *port) return false; } -static int mlxsw_sp_span_mtu_to_buffsize(const struct mlxsw_sp *mlxsw_sp, - int mtu) +static int +mlxsw_sp_span_port_buffsize_update(struct mlxsw_sp_port *mlxsw_sp_port, u16 mtu) { - return mlxsw_sp_bytes_cells(mlxsw_sp, mtu * 5 / 2) + 1; + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + char sbib_pl[MLXSW_REG_SBIB_LEN]; + u32 buffsize; + u32 speed; + int err; + + err = mlxsw_sp_port_speed_get(mlxsw_sp_port, &speed); + if (err) + return err; + if (speed == SPEED_UNKNOWN) + speed = 0; + + buffsize = mlxsw_sp_span_buffsize_get(mlxsw_sp, speed, mtu); + mlxsw_reg_sbib_pack(sbib_pl, mlxsw_sp_port->local_port, buffsize); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(sbib), sbib_pl); } int mlxsw_sp_span_port_mtu_update(struct mlxsw_sp_port *port, u16 mtu) { - struct mlxsw_sp *mlxsw_sp = port->mlxsw_sp; - char sbib_pl[MLXSW_REG_SBIB_LEN]; - int err; - /* If port is egress mirrored, the shared buffer size should be * updated according to the mtu value */ - if (mlxsw_sp_span_is_egress_mirror(port)) { - u32 buffsize = mlxsw_sp_span_mtu_to_buffsize(mlxsw_sp, mtu); + if (mlxsw_sp_span_is_egress_mirror(port)) + return mlxsw_sp_span_port_buffsize_update(port, mtu); + return 0; +} - mlxsw_reg_sbib_pack(sbib_pl, port->local_port, buffsize); - err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(sbib), sbib_pl); - if (err) { - netdev_err(port->dev, "Could not update shared buffer for mirroring\n"); - return err; - } - } +void mlxsw_sp_span_speed_update_work(struct work_struct *work) +{ + struct delayed_work *dwork = to_delayed_work(work); + struct mlxsw_sp_port *mlxsw_sp_port; - return 0; + mlxsw_sp_port = container_of(dwork, struct mlxsw_sp_port, + span.speed_update_dw); + + /* If port is egress mirrored, the shared buffer size should be + * updated according to the speed value. + */ + if (mlxsw_sp_span_is_egress_mirror(mlxsw_sp_port)) + mlxsw_sp_span_port_buffsize_update(mlxsw_sp_port, + mlxsw_sp_port->dev->mtu); } static struct mlxsw_sp_span_inspected_port * @@ -836,15 +853,9 @@ mlxsw_sp_span_inspected_port_add(struct mlxsw_sp_port *port, /* if it is an egress SPAN, bind a shared buffer to it */ if (type == MLXSW_SP_SPAN_EGRESS) { - u32 buffsize = mlxsw_sp_span_mtu_to_buffsize(mlxsw_sp, - port->dev->mtu); - - mlxsw_reg_sbib_pack(sbib_pl, port->local_port, buffsize); - err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(sbib), sbib_pl); - if (err) { - netdev_err(port->dev, "Could not create shared buffer for mirroring\n"); + err = mlxsw_sp_span_port_buffsize_update(port, port->dev->mtu); + if (err) return err; - } } if (bind) { diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_span.h b/drivers/net/ethernet/mellanox/mlxsw/spectrum_span.h index 5e04252f2a11..59724335525f 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_span.h +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_span.h @@ -74,5 +74,6 @@ void mlxsw_sp_span_entry_invalidate(struct mlxsw_sp *mlxsw_sp, struct mlxsw_sp_span_entry *span_entry); int mlxsw_sp_span_port_mtu_update(struct mlxsw_sp_port *port, u16 mtu); +void mlxsw_sp_span_speed_update_work(struct work_struct *work); #endif diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_trap.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_trap.c index e0d7c49ffae0..60205aa3f6a5 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_trap.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_trap.c @@ -9,6 +9,20 @@ #include "reg.h" #include "spectrum.h" +/* All driver-specific traps must be documented in + * Documentation/networking/devlink/mlxsw.rst + */ +enum { + DEVLINK_MLXSW_TRAP_ID_BASE = DEVLINK_TRAP_GENERIC_ID_MAX, + DEVLINK_MLXSW_TRAP_ID_IRIF_DISABLED, + DEVLINK_MLXSW_TRAP_ID_ERIF_DISABLED, +}; + +#define DEVLINK_MLXSW_TRAP_NAME_IRIF_DISABLED \ + "irif_disabled" +#define DEVLINK_MLXSW_TRAP_NAME_ERIF_DISABLED \ + "erif_disabled" + #define MLXSW_SP_TRAP_METADATA DEVLINK_TRAP_METADATA_TYPE_F_IN_PORT static void mlxsw_sp_rx_drop_listener(struct sk_buff *skb, u8 local_port, @@ -21,6 +35,12 @@ static void mlxsw_sp_rx_exception_listener(struct sk_buff *skb, u8 local_port, DEVLINK_TRAP_GROUP_GENERIC(_group_id), \ MLXSW_SP_TRAP_METADATA) +#define MLXSW_SP_TRAP_DRIVER_DROP(_id, _group_id) \ + DEVLINK_TRAP_DRIVER(DROP, DROP, DEVLINK_MLXSW_TRAP_ID_##_id, \ + DEVLINK_MLXSW_TRAP_NAME_##_id, \ + DEVLINK_TRAP_GROUP_GENERIC(_group_id), \ + MLXSW_SP_TRAP_METADATA) + #define MLXSW_SP_TRAP_EXCEPTION(_id, _group_id) \ DEVLINK_TRAP_GENERIC(EXCEPTION, TRAP, _id, \ DEVLINK_TRAP_GROUP_GENERIC(_group_id), \ @@ -58,6 +78,11 @@ static struct devlink_trap mlxsw_sp_traps_arr[] = { MLXSW_SP_TRAP_EXCEPTION(UNRESOLVED_NEIGH, L3_DROPS), MLXSW_SP_TRAP_EXCEPTION(IPV4_LPM_UNICAST_MISS, L3_DROPS), MLXSW_SP_TRAP_EXCEPTION(IPV6_LPM_UNICAST_MISS, L3_DROPS), + MLXSW_SP_TRAP_DRIVER_DROP(IRIF_DISABLED, L3_DROPS), + MLXSW_SP_TRAP_DRIVER_DROP(ERIF_DISABLED, L3_DROPS), + MLXSW_SP_TRAP_DROP(NON_ROUTABLE, L3_DROPS), + MLXSW_SP_TRAP_EXCEPTION(DECAP_ERROR, TUNNEL_DROPS), + MLXSW_SP_TRAP_DROP(OVERLAY_SMAC_MC, TUNNEL_DROPS), }; static struct mlxsw_listener mlxsw_sp_listeners_arr[] = { @@ -90,6 +115,15 @@ static struct mlxsw_listener mlxsw_sp_listeners_arr[] = { TRAP_EXCEPTION_TO_CPU), MLXSW_SP_RXL_EXCEPTION(DISCARD_ROUTER_LPM6, ROUTER_EXP, TRAP_EXCEPTION_TO_CPU), + MLXSW_SP_RXL_DISCARD(ROUTER_IRIF_EN, L3_DISCARDS), + MLXSW_SP_RXL_DISCARD(ROUTER_ERIF_EN, L3_DISCARDS), + MLXSW_SP_RXL_DISCARD(NON_ROUTABLE, L3_DISCARDS), + MLXSW_SP_RXL_EXCEPTION(DECAP_ECN0, ROUTER_EXP, TRAP_EXCEPTION_TO_CPU), + MLXSW_SP_RXL_EXCEPTION(IPIP_DECAP_ERROR, ROUTER_EXP, + TRAP_EXCEPTION_TO_CPU), + MLXSW_SP_RXL_EXCEPTION(DISCARD_DEC_PKT, TUNNEL_DISCARDS, + TRAP_EXCEPTION_TO_CPU), + MLXSW_SP_RXL_DISCARD(OVERLAY_SMAC_MC, TUNNEL_DISCARDS), }; /* Mapping between hardware trap and devlink trap. Multiple hardware traps can @@ -123,6 +157,13 @@ static u16 mlxsw_sp_listener_devlink_map[] = { DEVLINK_TRAP_GENERIC_ID_UNRESOLVED_NEIGH, DEVLINK_TRAP_GENERIC_ID_IPV4_LPM_UNICAST_MISS, DEVLINK_TRAP_GENERIC_ID_IPV6_LPM_UNICAST_MISS, + DEVLINK_MLXSW_TRAP_ID_IRIF_DISABLED, + DEVLINK_MLXSW_TRAP_ID_ERIF_DISABLED, + DEVLINK_TRAP_GENERIC_ID_NON_ROUTABLE, + DEVLINK_TRAP_GENERIC_ID_DECAP_ERROR, + DEVLINK_TRAP_GENERIC_ID_DECAP_ERROR, + DEVLINK_TRAP_GENERIC_ID_DECAP_ERROR, + DEVLINK_TRAP_GENERIC_ID_OVERLAY_SMAC_MC, }; static int mlxsw_sp_rx_listener(struct mlxsw_sp *mlxsw_sp, struct sk_buff *skb, @@ -304,8 +345,9 @@ mlxsw_sp_trap_group_policer_init(struct mlxsw_sp *mlxsw_sp, u32 rate; switch (group->id) { - case DEVLINK_TRAP_GROUP_GENERIC_ID_L3_DROPS:/* fall through */ - case DEVLINK_TRAP_GROUP_GENERIC_ID_L2_DROPS: + case DEVLINK_TRAP_GROUP_GENERIC_ID_L2_DROPS: /* fall through */ + case DEVLINK_TRAP_GROUP_GENERIC_ID_L3_DROPS: /* fall through */ + case DEVLINK_TRAP_GROUP_GENERIC_ID_TUNNEL_DROPS: policer_id = MLXSW_SP_DISCARD_POLICER_ID; ir_units = MLXSW_REG_QPCR_IR_UNITS_M; is_bytes = false; @@ -342,6 +384,12 @@ __mlxsw_sp_trap_group_init(struct mlxsw_sp *mlxsw_sp, priority = 0; tc = 1; break; + case DEVLINK_TRAP_GROUP_GENERIC_ID_TUNNEL_DROPS: + group_id = MLXSW_REG_HTGT_TRAP_GROUP_SP_TUNNEL_DISCARDS; + policer_id = MLXSW_SP_DISCARD_POLICER_ID; + priority = 0; + tc = 1; + break; default: return -EINVAL; } diff --git a/drivers/net/ethernet/mellanox/mlxsw/switchx2.c b/drivers/net/ethernet/mellanox/mlxsw/switchx2.c index de6cb22f68b1..f0e98ec8f1ee 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/switchx2.c +++ b/drivers/net/ethernet/mellanox/mlxsw/switchx2.c @@ -299,22 +299,17 @@ static netdev_tx_t mlxsw_sx_port_xmit(struct sk_buff *skb, u64 len; int err; + if (skb_cow_head(skb, MLXSW_TXHDR_LEN)) { + this_cpu_inc(mlxsw_sx_port->pcpu_stats->tx_dropped); + dev_kfree_skb_any(skb); + return NETDEV_TX_OK; + } + memset(skb->cb, 0, sizeof(struct mlxsw_skb_cb)); if (mlxsw_core_skb_transmit_busy(mlxsw_sx->core, &tx_info)) return NETDEV_TX_BUSY; - if (unlikely(skb_headroom(skb) < MLXSW_TXHDR_LEN)) { - struct sk_buff *skb_orig = skb; - - skb = skb_realloc_headroom(skb, MLXSW_TXHDR_LEN); - if (!skb) { - this_cpu_inc(mlxsw_sx_port->pcpu_stats->tx_dropped); - dev_kfree_skb_any(skb_orig); - return NETDEV_TX_OK; - } - dev_consume_skb_any(skb_orig); - } mlxsw_sx_txhdr_construct(skb, &tx_info); /* TX header is consumed by HW on the way so we shouldn't count its * bytes as being sent. diff --git a/drivers/net/ethernet/mellanox/mlxsw/trap.h b/drivers/net/ethernet/mellanox/mlxsw/trap.h index 3d2331be05d8..12e1fa998d42 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/trap.h +++ b/drivers/net/ethernet/mellanox/mlxsw/trap.h @@ -67,6 +67,7 @@ enum { MLXSW_TRAP_ID_NVE_ENCAP_ARP = 0xBD, MLXSW_TRAP_ID_ROUTER_ALERT_IPV4 = 0xD6, MLXSW_TRAP_ID_ROUTER_ALERT_IPV6 = 0xD7, + MLXSW_TRAP_ID_DISCARD_NON_ROUTABLE = 0x11A, MLXSW_TRAP_ID_DISCARD_ROUTER2 = 0x130, MLXSW_TRAP_ID_DISCARD_ROUTER3 = 0x131, MLXSW_TRAP_ID_DISCARD_ING_PACKET_SMAC_MC = 0x140, @@ -88,8 +89,12 @@ enum { MLXSW_TRAP_ID_DISCARD_ING_ROUTER_IPV4_SIP_BC = 0x16A, MLXSW_TRAP_ID_DISCARD_ING_ROUTER_IPV4_DIP_LOCAL_NET = 0x16B, MLXSW_TRAP_ID_DISCARD_ING_ROUTER_DIP_LINK_LOCAL = 0x16C, + MLXSW_TRAP_ID_DISCARD_ROUTER_IRIF_EN = 0x178, + MLXSW_TRAP_ID_DISCARD_ROUTER_ERIF_EN = 0x179, MLXSW_TRAP_ID_DISCARD_ROUTER_LPM4 = 0x17B, MLXSW_TRAP_ID_DISCARD_ROUTER_LPM6 = 0x17C, + MLXSW_TRAP_ID_DISCARD_DEC_PKT = 0x188, + MLXSW_TRAP_ID_DISCARD_OVERLAY_SMAC_MC = 0x190, MLXSW_TRAP_ID_DISCARD_IPV6_MC_DIP_RESERVED_SCOPE = 0x1B0, MLXSW_TRAP_ID_DISCARD_IPV6_MC_DIP_INTERFACE_LOCAL_SCOPE = 0x1B1, MLXSW_TRAP_ID_ACL0 = 0x1C0, diff --git a/drivers/net/ethernet/myricom/myri10ge/myri10ge.c b/drivers/net/ethernet/myricom/myri10ge/myri10ge.c index c979f38a2e0c..2ee0d0be113a 100644 --- a/drivers/net/ethernet/myricom/myri10ge/myri10ge.c +++ b/drivers/net/ethernet/myricom/myri10ge/myri10ge.c @@ -2892,7 +2892,7 @@ drop: static netdev_tx_t myri10ge_sw_tso(struct sk_buff *skb, struct net_device *dev) { - struct sk_buff *segs, *curr; + struct sk_buff *segs, *curr, *next; struct myri10ge_priv *mgp = netdev_priv(dev); struct myri10ge_slice_state *ss; netdev_tx_t status; @@ -2901,10 +2901,8 @@ static netdev_tx_t myri10ge_sw_tso(struct sk_buff *skb, if (IS_ERR(segs)) goto drop; - while (segs) { - curr = segs; - segs = segs->next; - curr->next = NULL; + skb_list_walk_safe(segs, curr, next) { + skb_mark_not_on_list(curr); status = myri10ge_xmit(curr, dev); if (status != 0) { dev_kfree_skb_any(curr); diff --git a/drivers/net/ethernet/netronome/Kconfig b/drivers/net/ethernet/netronome/Kconfig index bac5be4d4f43..a3f68a718813 100644 --- a/drivers/net/ethernet/netronome/Kconfig +++ b/drivers/net/ethernet/netronome/Kconfig @@ -31,6 +31,7 @@ config NFP_APP_FLOWER bool "NFP4000/NFP6000 TC Flower offload support" depends on NFP depends on NET_SWITCHDEV + depends on IPV6!=m || NFP=m default y ---help--- Enable driver support for TC Flower offload on NFP4000 and NFP6000. diff --git a/drivers/net/ethernet/rdc/r6040.c b/drivers/net/ethernet/rdc/r6040.c index c23cb61bbd30..f5ecc410ff85 100644 --- a/drivers/net/ethernet/rdc/r6040.c +++ b/drivers/net/ethernet/rdc/r6040.c @@ -498,14 +498,6 @@ static int r6040_close(struct net_device *dev) return 0; } -static int r6040_ioctl(struct net_device *dev, struct ifreq *rq, int cmd) -{ - if (!dev->phydev) - return -EINVAL; - - return phy_mii_ioctl(dev->phydev, rq, cmd); -} - static int r6040_rx(struct net_device *dev, int limit) { struct r6040_private *priv = netdev_priv(dev); @@ -957,7 +949,7 @@ static const struct net_device_ops r6040_netdev_ops = { .ndo_set_rx_mode = r6040_multicast_list, .ndo_validate_addr = eth_validate_addr, .ndo_set_mac_address = eth_mac_addr, - .ndo_do_ioctl = r6040_ioctl, + .ndo_do_ioctl = phy_do_ioctl, .ndo_tx_timeout = r6040_tx_timeout, #ifdef CONFIG_NET_POLL_CONTROLLER .ndo_poll_controller = r6040_poll_controller, diff --git a/drivers/net/ethernet/realtek/Makefile b/drivers/net/ethernet/realtek/Makefile index d5304bad2372..2e1d78b106b0 100644 --- a/drivers/net/ethernet/realtek/Makefile +++ b/drivers/net/ethernet/realtek/Makefile @@ -6,5 +6,5 @@ obj-$(CONFIG_8139CP) += 8139cp.o obj-$(CONFIG_8139TOO) += 8139too.o obj-$(CONFIG_ATP) += atp.o -r8169-objs += r8169_main.o r8169_firmware.o +r8169-objs += r8169_main.o r8169_firmware.o r8169_phy_config.o obj-$(CONFIG_R8169) += r8169.o diff --git a/drivers/net/ethernet/realtek/r8169.h b/drivers/net/ethernet/realtek/r8169.h new file mode 100644 index 000000000000..22a6a057b11e --- /dev/null +++ b/drivers/net/ethernet/realtek/r8169.h @@ -0,0 +1,78 @@ +/* SPDX-License-Identifier: GPL-2.0-only */ +/* r8169.h: RealTek 8169/8168/8101 ethernet driver. + * + * Copyright (c) 2002 ShuChen <shuchen@realtek.com.tw> + * Copyright (c) 2003 - 2007 Francois Romieu <romieu@fr.zoreil.com> + * Copyright (c) a lot of people too. Please respect their work. + * + * See MAINTAINERS file for support contact information. + */ + +#include <linux/types.h> +#include <linux/phy.h> + +enum mac_version { + /* support for ancient RTL_GIGA_MAC_VER_01 has been removed */ + RTL_GIGA_MAC_VER_02, + RTL_GIGA_MAC_VER_03, + RTL_GIGA_MAC_VER_04, + RTL_GIGA_MAC_VER_05, + RTL_GIGA_MAC_VER_06, + RTL_GIGA_MAC_VER_07, + RTL_GIGA_MAC_VER_08, + RTL_GIGA_MAC_VER_09, + RTL_GIGA_MAC_VER_10, + RTL_GIGA_MAC_VER_11, + RTL_GIGA_MAC_VER_12, + RTL_GIGA_MAC_VER_13, + RTL_GIGA_MAC_VER_14, + RTL_GIGA_MAC_VER_15, + RTL_GIGA_MAC_VER_16, + RTL_GIGA_MAC_VER_17, + RTL_GIGA_MAC_VER_18, + RTL_GIGA_MAC_VER_19, + RTL_GIGA_MAC_VER_20, + RTL_GIGA_MAC_VER_21, + RTL_GIGA_MAC_VER_22, + RTL_GIGA_MAC_VER_23, + RTL_GIGA_MAC_VER_24, + RTL_GIGA_MAC_VER_25, + RTL_GIGA_MAC_VER_26, + RTL_GIGA_MAC_VER_27, + RTL_GIGA_MAC_VER_28, + RTL_GIGA_MAC_VER_29, + RTL_GIGA_MAC_VER_30, + RTL_GIGA_MAC_VER_31, + RTL_GIGA_MAC_VER_32, + RTL_GIGA_MAC_VER_33, + RTL_GIGA_MAC_VER_34, + RTL_GIGA_MAC_VER_35, + RTL_GIGA_MAC_VER_36, + RTL_GIGA_MAC_VER_37, + RTL_GIGA_MAC_VER_38, + RTL_GIGA_MAC_VER_39, + RTL_GIGA_MAC_VER_40, + RTL_GIGA_MAC_VER_41, + RTL_GIGA_MAC_VER_42, + RTL_GIGA_MAC_VER_43, + RTL_GIGA_MAC_VER_44, + RTL_GIGA_MAC_VER_45, + RTL_GIGA_MAC_VER_46, + RTL_GIGA_MAC_VER_47, + RTL_GIGA_MAC_VER_48, + RTL_GIGA_MAC_VER_49, + RTL_GIGA_MAC_VER_50, + RTL_GIGA_MAC_VER_51, + RTL_GIGA_MAC_VER_52, + RTL_GIGA_MAC_VER_60, + RTL_GIGA_MAC_VER_61, + RTL_GIGA_MAC_NONE +}; + +struct rtl8169_private; + +void r8169_apply_firmware(struct rtl8169_private *tp); +u16 rtl8168h_2_get_adc_bias_ioffset(struct rtl8169_private *tp); +u8 rtl8168d_efuse_read(struct rtl8169_private *tp, int reg_addr); +void r8169_hw_phy_config(struct rtl8169_private *tp, struct phy_device *phydev, + enum mac_version ver); diff --git a/drivers/net/ethernet/realtek/r8169_main.c b/drivers/net/ethernet/realtek/r8169_main.c index 0161d839fa87..7a5fe11378aa 100644 --- a/drivers/net/ethernet/realtek/r8169_main.c +++ b/drivers/net/ethernet/realtek/r8169_main.c @@ -31,6 +31,7 @@ #include <linux/ipv6.h> #include <net/ip6_checksum.h> +#include "r8169.h" #include "r8169_firmware.h" #define MODULENAME "r8169" @@ -84,64 +85,6 @@ #define RTL_R16(tp, reg) readw(tp->mmio_addr + (reg)) #define RTL_R32(tp, reg) readl(tp->mmio_addr + (reg)) -enum mac_version { - /* support for ancient RTL_GIGA_MAC_VER_01 has been removed */ - RTL_GIGA_MAC_VER_02, - RTL_GIGA_MAC_VER_03, - RTL_GIGA_MAC_VER_04, - RTL_GIGA_MAC_VER_05, - RTL_GIGA_MAC_VER_06, - RTL_GIGA_MAC_VER_07, - RTL_GIGA_MAC_VER_08, - RTL_GIGA_MAC_VER_09, - RTL_GIGA_MAC_VER_10, - RTL_GIGA_MAC_VER_11, - RTL_GIGA_MAC_VER_12, - RTL_GIGA_MAC_VER_13, - RTL_GIGA_MAC_VER_14, - RTL_GIGA_MAC_VER_15, - RTL_GIGA_MAC_VER_16, - RTL_GIGA_MAC_VER_17, - RTL_GIGA_MAC_VER_18, - RTL_GIGA_MAC_VER_19, - RTL_GIGA_MAC_VER_20, - RTL_GIGA_MAC_VER_21, - RTL_GIGA_MAC_VER_22, - RTL_GIGA_MAC_VER_23, - RTL_GIGA_MAC_VER_24, - RTL_GIGA_MAC_VER_25, - RTL_GIGA_MAC_VER_26, - RTL_GIGA_MAC_VER_27, - RTL_GIGA_MAC_VER_28, - RTL_GIGA_MAC_VER_29, - RTL_GIGA_MAC_VER_30, - RTL_GIGA_MAC_VER_31, - RTL_GIGA_MAC_VER_32, - RTL_GIGA_MAC_VER_33, - RTL_GIGA_MAC_VER_34, - RTL_GIGA_MAC_VER_35, - RTL_GIGA_MAC_VER_36, - RTL_GIGA_MAC_VER_37, - RTL_GIGA_MAC_VER_38, - RTL_GIGA_MAC_VER_39, - RTL_GIGA_MAC_VER_40, - RTL_GIGA_MAC_VER_41, - RTL_GIGA_MAC_VER_42, - RTL_GIGA_MAC_VER_43, - RTL_GIGA_MAC_VER_44, - RTL_GIGA_MAC_VER_45, - RTL_GIGA_MAC_VER_46, - RTL_GIGA_MAC_VER_47, - RTL_GIGA_MAC_VER_48, - RTL_GIGA_MAC_VER_49, - RTL_GIGA_MAC_VER_50, - RTL_GIGA_MAC_VER_51, - RTL_GIGA_MAC_VER_52, - RTL_GIGA_MAC_VER_60, - RTL_GIGA_MAC_VER_61, - RTL_GIGA_MAC_NONE -}; - #define JUMBO_1K ETH_DATA_LEN #define JUMBO_4K (4*1024 - ETH_HLEN - 2) #define JUMBO_6K (6*1024 - ETH_HLEN - 2) @@ -492,6 +435,7 @@ enum rtl_register_content { /* CPlusCmd p.31 */ EnableBist = (1 << 15), // 8168 8101 Mac_dbgo_oe = (1 << 14), // 8168 8101 + EnAnaPLL = (1 << 14), // 8169 Normal_mode = (1 << 13), // unused Force_half_dup = (1 << 12), // 8168 8101 Force_rxflow_en = (1 << 11), // 8168 8101 @@ -1078,52 +1022,6 @@ static int rtl_readphy(struct rtl8169_private *tp, int location) } } -static void rtl_patchphy(struct rtl8169_private *tp, int reg_addr, int value) -{ - rtl_writephy(tp, reg_addr, rtl_readphy(tp, reg_addr) | value); -} - -static void rtl_w0w1_phy(struct rtl8169_private *tp, int reg_addr, int p, int m) -{ - int val; - - val = rtl_readphy(tp, reg_addr); - rtl_writephy(tp, reg_addr, (val & ~m) | p); -} - -static void r8168d_modify_extpage(struct phy_device *phydev, int extpage, - int reg, u16 mask, u16 val) -{ - int oldpage = phy_select_page(phydev, 0x0007); - - __phy_write(phydev, 0x1e, extpage); - __phy_modify(phydev, reg, mask, val); - - phy_restore_page(phydev, oldpage, 0); -} - -static void r8168d_phy_param(struct phy_device *phydev, u16 parm, - u16 mask, u16 val) -{ - int oldpage = phy_select_page(phydev, 0x0005); - - __phy_write(phydev, 0x05, parm); - __phy_modify(phydev, 0x06, mask, val); - - phy_restore_page(phydev, oldpage, 0); -} - -static void r8168g_phy_param(struct phy_device *phydev, u16 parm, - u16 mask, u16 val) -{ - int oldpage = phy_select_page(phydev, 0x0a43); - - __phy_write(phydev, 0x13, parm); - __phy_modify(phydev, 0x14, mask, val); - - phy_restore_page(phydev, oldpage, 0); -} - DECLARE_RTL_COND(rtl_ephyar_cond) { return RTL_R32(tp, EPHYAR) & EPHYAR_FLAG; @@ -1372,7 +1270,7 @@ DECLARE_RTL_COND(rtl_efusear_cond) return RTL_R32(tp, EFUSEAR) & EFUSEAR_FLAG; } -static u8 rtl8168d_efuse_read(struct rtl8169_private *tp, int reg_addr) +u8 rtl8168d_efuse_read(struct rtl8169_private *tp, int reg_addr) { RTL_W32(tp, EFUSEAR, (reg_addr & EFUSEAR_REG_MASK) << EFUSEAR_REG_SHIFT); @@ -2268,22 +2166,6 @@ static void rtl8169_get_mac_version(struct rtl8169_private *tp) } } -struct phy_reg { - u16 reg; - u16 val; -}; - -static void __rtl_writephy_batch(struct rtl8169_private *tp, - const struct phy_reg *regs, int len) -{ - while (len-- > 0) { - rtl_writephy(tp, regs->reg, regs->val); - regs++; - } -} - -#define rtl_writephy_batch(tp, a) __rtl_writephy_batch(tp, a, ARRAY_SIZE(a)) - static void rtl_release_firmware(struct rtl8169_private *tp) { if (tp->rtl_fw) { @@ -2293,7 +2175,7 @@ static void rtl_release_firmware(struct rtl8169_private *tp) } } -static void rtl_apply_firmware(struct rtl8169_private *tp) +void r8169_apply_firmware(struct rtl8169_private *tp) { /* TODO: release firmware if rtl_fw_write_firmware signals failure. */ if (tp->rtl_fw) @@ -2315,594 +2197,6 @@ static void rtl8125_config_eee_mac(struct rtl8169_private *tp) r8168_mac_ocp_modify(tp, 0xeb62, 0, BIT(2) | BIT(1)); } -static void rtl8168f_config_eee_phy(struct rtl8169_private *tp) -{ - struct phy_device *phydev = tp->phydev; - - r8168d_modify_extpage(phydev, 0x0020, 0x15, 0, BIT(8)); - r8168d_phy_param(phydev, 0x8b85, 0, BIT(13)); -} - -static void rtl8168g_config_eee_phy(struct rtl8169_private *tp) -{ - phy_modify_paged(tp->phydev, 0x0a43, 0x11, 0, BIT(4)); -} - -static void rtl8168h_config_eee_phy(struct rtl8169_private *tp) -{ - struct phy_device *phydev = tp->phydev; - - rtl8168g_config_eee_phy(tp); - - phy_modify_paged(phydev, 0xa4a, 0x11, 0x0000, 0x0200); - phy_modify_paged(phydev, 0xa42, 0x14, 0x0000, 0x0080); -} - -static void rtl8125_config_eee_phy(struct rtl8169_private *tp) -{ - struct phy_device *phydev = tp->phydev; - - rtl8168h_config_eee_phy(tp); - - phy_modify_paged(phydev, 0xa6d, 0x12, 0x0001, 0x0000); - phy_modify_paged(phydev, 0xa6d, 0x14, 0x0010, 0x0000); -} - -static void rtl8169s_hw_phy_config(struct rtl8169_private *tp) -{ - static const struct phy_reg phy_reg_init[] = { - { 0x1f, 0x0001 }, - { 0x06, 0x006e }, - { 0x08, 0x0708 }, - { 0x15, 0x4000 }, - { 0x18, 0x65c7 }, - - { 0x1f, 0x0001 }, - { 0x03, 0x00a1 }, - { 0x02, 0x0008 }, - { 0x01, 0x0120 }, - { 0x00, 0x1000 }, - { 0x04, 0x0800 }, - { 0x04, 0x0000 }, - - { 0x03, 0xff41 }, - { 0x02, 0xdf60 }, - { 0x01, 0x0140 }, - { 0x00, 0x0077 }, - { 0x04, 0x7800 }, - { 0x04, 0x7000 }, - - { 0x03, 0x802f }, - { 0x02, 0x4f02 }, - { 0x01, 0x0409 }, - { 0x00, 0xf0f9 }, - { 0x04, 0x9800 }, - { 0x04, 0x9000 }, - - { 0x03, 0xdf01 }, - { 0x02, 0xdf20 }, - { 0x01, 0xff95 }, - { 0x00, 0xba00 }, - { 0x04, 0xa800 }, - { 0x04, 0xa000 }, - - { 0x03, 0xff41 }, - { 0x02, 0xdf20 }, - { 0x01, 0x0140 }, - { 0x00, 0x00bb }, - { 0x04, 0xb800 }, - { 0x04, 0xb000 }, - - { 0x03, 0xdf41 }, - { 0x02, 0xdc60 }, - { 0x01, 0x6340 }, - { 0x00, 0x007d }, - { 0x04, 0xd800 }, - { 0x04, 0xd000 }, - - { 0x03, 0xdf01 }, - { 0x02, 0xdf20 }, - { 0x01, 0x100a }, - { 0x00, 0xa0ff }, - { 0x04, 0xf800 }, - { 0x04, 0xf000 }, - - { 0x1f, 0x0000 }, - { 0x0b, 0x0000 }, - { 0x00, 0x9200 } - }; - - rtl_writephy_batch(tp, phy_reg_init); -} - -static void rtl8169sb_hw_phy_config(struct rtl8169_private *tp) -{ - phy_write_paged(tp->phydev, 0x0002, 0x01, 0x90d0); -} - -static void rtl8169scd_hw_phy_config_quirk(struct rtl8169_private *tp) -{ - struct pci_dev *pdev = tp->pci_dev; - - if ((pdev->subsystem_vendor != PCI_VENDOR_ID_GIGABYTE) || - (pdev->subsystem_device != 0xe000)) - return; - - phy_write_paged(tp->phydev, 0x0001, 0x10, 0xf01b); -} - -static void rtl8169scd_hw_phy_config(struct rtl8169_private *tp) -{ - static const struct phy_reg phy_reg_init[] = { - { 0x1f, 0x0001 }, - { 0x04, 0x0000 }, - { 0x03, 0x00a1 }, - { 0x02, 0x0008 }, - { 0x01, 0x0120 }, - { 0x00, 0x1000 }, - { 0x04, 0x0800 }, - { 0x04, 0x9000 }, - { 0x03, 0x802f }, - { 0x02, 0x4f02 }, - { 0x01, 0x0409 }, - { 0x00, 0xf099 }, - { 0x04, 0x9800 }, - { 0x04, 0xa000 }, - { 0x03, 0xdf01 }, - { 0x02, 0xdf20 }, - { 0x01, 0xff95 }, - { 0x00, 0xba00 }, - { 0x04, 0xa800 }, - { 0x04, 0xf000 }, - { 0x03, 0xdf01 }, - { 0x02, 0xdf20 }, - { 0x01, 0x101a }, - { 0x00, 0xa0ff }, - { 0x04, 0xf800 }, - { 0x04, 0x0000 }, - { 0x1f, 0x0000 }, - - { 0x1f, 0x0001 }, - { 0x10, 0xf41b }, - { 0x14, 0xfb54 }, - { 0x18, 0xf5c7 }, - { 0x1f, 0x0000 }, - - { 0x1f, 0x0001 }, - { 0x17, 0x0cc0 }, - { 0x1f, 0x0000 } - }; - - rtl_writephy_batch(tp, phy_reg_init); - - rtl8169scd_hw_phy_config_quirk(tp); -} - -static void rtl8169sce_hw_phy_config(struct rtl8169_private *tp) -{ - static const struct phy_reg phy_reg_init[] = { - { 0x1f, 0x0001 }, - { 0x04, 0x0000 }, - { 0x03, 0x00a1 }, - { 0x02, 0x0008 }, - { 0x01, 0x0120 }, - { 0x00, 0x1000 }, - { 0x04, 0x0800 }, - { 0x04, 0x9000 }, - { 0x03, 0x802f }, - { 0x02, 0x4f02 }, - { 0x01, 0x0409 }, - { 0x00, 0xf099 }, - { 0x04, 0x9800 }, - { 0x04, 0xa000 }, - { 0x03, 0xdf01 }, - { 0x02, 0xdf20 }, - { 0x01, 0xff95 }, - { 0x00, 0xba00 }, - { 0x04, 0xa800 }, - { 0x04, 0xf000 }, - { 0x03, 0xdf01 }, - { 0x02, 0xdf20 }, - { 0x01, 0x101a }, - { 0x00, 0xa0ff }, - { 0x04, 0xf800 }, - { 0x04, 0x0000 }, - { 0x1f, 0x0000 }, - - { 0x1f, 0x0001 }, - { 0x0b, 0x8480 }, - { 0x1f, 0x0000 }, - - { 0x1f, 0x0001 }, - { 0x18, 0x67c7 }, - { 0x04, 0x2000 }, - { 0x03, 0x002f }, - { 0x02, 0x4360 }, - { 0x01, 0x0109 }, - { 0x00, 0x3022 }, - { 0x04, 0x2800 }, - { 0x1f, 0x0000 }, - - { 0x1f, 0x0001 }, - { 0x17, 0x0cc0 }, - { 0x1f, 0x0000 } - }; - - rtl_writephy_batch(tp, phy_reg_init); -} - -static void rtl8168bb_hw_phy_config(struct rtl8169_private *tp) -{ - rtl_writephy(tp, 0x1f, 0x0001); - rtl_patchphy(tp, 0x16, 1 << 0); - rtl_writephy(tp, 0x10, 0xf41b); - rtl_writephy(tp, 0x1f, 0x0000); -} - -static void rtl8168bef_hw_phy_config(struct rtl8169_private *tp) -{ - phy_write_paged(tp->phydev, 0x0001, 0x10, 0xf41b); -} - -static void rtl8168cp_1_hw_phy_config(struct rtl8169_private *tp) -{ - phy_write(tp->phydev, 0x1d, 0x0f00); - phy_write_paged(tp->phydev, 0x0002, 0x0c, 0x1ec8); -} - -static void rtl8168cp_2_hw_phy_config(struct rtl8169_private *tp) -{ - phy_set_bits(tp->phydev, 0x14, BIT(5)); - phy_set_bits(tp->phydev, 0x0d, BIT(5)); - phy_write_paged(tp->phydev, 0x0001, 0x1d, 0x3d98); -} - -static void rtl8168c_1_hw_phy_config(struct rtl8169_private *tp) -{ - static const struct phy_reg phy_reg_init[] = { - { 0x1f, 0x0001 }, - { 0x12, 0x2300 }, - { 0x1f, 0x0002 }, - { 0x00, 0x88d4 }, - { 0x01, 0x82b1 }, - { 0x03, 0x7002 }, - { 0x08, 0x9e30 }, - { 0x09, 0x01f0 }, - { 0x0a, 0x5500 }, - { 0x0c, 0x00c8 }, - { 0x1f, 0x0003 }, - { 0x12, 0xc096 }, - { 0x16, 0x000a }, - { 0x1f, 0x0000 }, - { 0x1f, 0x0000 }, - { 0x09, 0x2000 }, - { 0x09, 0x0000 } - }; - - rtl_writephy_batch(tp, phy_reg_init); - - rtl_patchphy(tp, 0x14, 1 << 5); - rtl_patchphy(tp, 0x0d, 1 << 5); - rtl_writephy(tp, 0x1f, 0x0000); -} - -static void rtl8168c_2_hw_phy_config(struct rtl8169_private *tp) -{ - static const struct phy_reg phy_reg_init[] = { - { 0x1f, 0x0001 }, - { 0x12, 0x2300 }, - { 0x03, 0x802f }, - { 0x02, 0x4f02 }, - { 0x01, 0x0409 }, - { 0x00, 0xf099 }, - { 0x04, 0x9800 }, - { 0x04, 0x9000 }, - { 0x1d, 0x3d98 }, - { 0x1f, 0x0002 }, - { 0x0c, 0x7eb8 }, - { 0x06, 0x0761 }, - { 0x1f, 0x0003 }, - { 0x16, 0x0f0a }, - { 0x1f, 0x0000 } - }; - - rtl_writephy_batch(tp, phy_reg_init); - - rtl_patchphy(tp, 0x16, 1 << 0); - rtl_patchphy(tp, 0x14, 1 << 5); - rtl_patchphy(tp, 0x0d, 1 << 5); - rtl_writephy(tp, 0x1f, 0x0000); -} - -static void rtl8168c_3_hw_phy_config(struct rtl8169_private *tp) -{ - static const struct phy_reg phy_reg_init[] = { - { 0x1f, 0x0001 }, - { 0x12, 0x2300 }, - { 0x1d, 0x3d98 }, - { 0x1f, 0x0002 }, - { 0x0c, 0x7eb8 }, - { 0x06, 0x5461 }, - { 0x1f, 0x0003 }, - { 0x16, 0x0f0a }, - { 0x1f, 0x0000 } - }; - - rtl_writephy_batch(tp, phy_reg_init); - - rtl_patchphy(tp, 0x16, 1 << 0); - rtl_patchphy(tp, 0x14, 1 << 5); - rtl_patchphy(tp, 0x0d, 1 << 5); - rtl_writephy(tp, 0x1f, 0x0000); -} - -static const struct phy_reg rtl8168d_1_phy_reg_init_0[] = { - /* Channel Estimation */ - { 0x1f, 0x0001 }, - { 0x06, 0x4064 }, - { 0x07, 0x2863 }, - { 0x08, 0x059c }, - { 0x09, 0x26b4 }, - { 0x0a, 0x6a19 }, - { 0x0b, 0xdcc8 }, - { 0x10, 0xf06d }, - { 0x14, 0x7f68 }, - { 0x18, 0x7fd9 }, - { 0x1c, 0xf0ff }, - { 0x1d, 0x3d9c }, - { 0x1f, 0x0003 }, - { 0x12, 0xf49f }, - { 0x13, 0x070b }, - { 0x1a, 0x05ad }, - { 0x14, 0x94c0 }, - - /* - * Tx Error Issue - * Enhance line driver power - */ - { 0x1f, 0x0002 }, - { 0x06, 0x5561 }, - { 0x1f, 0x0005 }, - { 0x05, 0x8332 }, - { 0x06, 0x5561 }, - - /* - * Can not link to 1Gbps with bad cable - * Decrease SNR threshold form 21.07dB to 19.04dB - */ - { 0x1f, 0x0001 }, - { 0x17, 0x0cc0 }, - - { 0x1f, 0x0000 }, - { 0x0d, 0xf880 } -}; - -static const struct phy_reg rtl8168d_1_phy_reg_init_1[] = { - { 0x1f, 0x0002 }, - { 0x05, 0x669a }, - { 0x1f, 0x0005 }, - { 0x05, 0x8330 }, - { 0x06, 0x669a }, - { 0x1f, 0x0002 } -}; - -static void rtl8168d_apply_firmware_cond(struct rtl8169_private *tp, u16 val) -{ - u16 reg_val; - - rtl_writephy(tp, 0x1f, 0x0005); - rtl_writephy(tp, 0x05, 0x001b); - reg_val = rtl_readphy(tp, 0x06); - rtl_writephy(tp, 0x1f, 0x0000); - - if (reg_val != val) - netif_warn(tp, hw, tp->dev, "chipset not ready for firmware\n"); - else - rtl_apply_firmware(tp); -} - -static void rtl8168d_1_hw_phy_config(struct rtl8169_private *tp) -{ - rtl_writephy_batch(tp, rtl8168d_1_phy_reg_init_0); - - /* - * Rx Error Issue - * Fine Tune Switching regulator parameter - */ - rtl_writephy(tp, 0x1f, 0x0002); - rtl_w0w1_phy(tp, 0x0b, 0x0010, 0x00ef); - rtl_w0w1_phy(tp, 0x0c, 0xa200, 0x5d00); - - if (rtl8168d_efuse_read(tp, 0x01) == 0xb1) { - int val; - - rtl_writephy_batch(tp, rtl8168d_1_phy_reg_init_1); - - val = rtl_readphy(tp, 0x0d); - - if ((val & 0x00ff) != 0x006c) { - static const u32 set[] = { - 0x0065, 0x0066, 0x0067, 0x0068, - 0x0069, 0x006a, 0x006b, 0x006c - }; - int i; - - rtl_writephy(tp, 0x1f, 0x0002); - - val &= 0xff00; - for (i = 0; i < ARRAY_SIZE(set); i++) - rtl_writephy(tp, 0x0d, val | set[i]); - } - } else { - phy_write_paged(tp->phydev, 0x0002, 0x05, 0x6662); - r8168d_phy_param(tp->phydev, 0x8330, 0xffff, 0x6662); - } - - /* RSET couple improve */ - rtl_writephy(tp, 0x1f, 0x0002); - rtl_patchphy(tp, 0x0d, 0x0300); - rtl_patchphy(tp, 0x0f, 0x0010); - - /* Fine tune PLL performance */ - rtl_writephy(tp, 0x1f, 0x0002); - rtl_w0w1_phy(tp, 0x02, 0x0100, 0x0600); - rtl_w0w1_phy(tp, 0x03, 0x0000, 0xe000); - rtl_writephy(tp, 0x1f, 0x0000); - - rtl8168d_apply_firmware_cond(tp, 0xbf00); -} - -static void rtl8168d_2_hw_phy_config(struct rtl8169_private *tp) -{ - rtl_writephy_batch(tp, rtl8168d_1_phy_reg_init_0); - - if (rtl8168d_efuse_read(tp, 0x01) == 0xb1) { - int val; - - rtl_writephy_batch(tp, rtl8168d_1_phy_reg_init_1); - - val = rtl_readphy(tp, 0x0d); - if ((val & 0x00ff) != 0x006c) { - static const u32 set[] = { - 0x0065, 0x0066, 0x0067, 0x0068, - 0x0069, 0x006a, 0x006b, 0x006c - }; - int i; - - rtl_writephy(tp, 0x1f, 0x0002); - - val &= 0xff00; - for (i = 0; i < ARRAY_SIZE(set); i++) - rtl_writephy(tp, 0x0d, val | set[i]); - } - } else { - phy_write_paged(tp->phydev, 0x0002, 0x05, 0x2642); - r8168d_phy_param(tp->phydev, 0x8330, 0xffff, 0x2642); - } - - /* Fine tune PLL performance */ - rtl_writephy(tp, 0x1f, 0x0002); - rtl_w0w1_phy(tp, 0x02, 0x0100, 0x0600); - rtl_w0w1_phy(tp, 0x03, 0x0000, 0xe000); - - /* Switching regulator Slew rate */ - rtl_writephy(tp, 0x1f, 0x0002); - rtl_patchphy(tp, 0x0f, 0x0017); - rtl_writephy(tp, 0x1f, 0x0000); - - rtl8168d_apply_firmware_cond(tp, 0xb300); -} - -static void rtl8168d_3_hw_phy_config(struct rtl8169_private *tp) -{ - static const struct phy_reg phy_reg_init[] = { - { 0x1f, 0x0002 }, - { 0x10, 0x0008 }, - { 0x0d, 0x006c }, - - { 0x1f, 0x0000 }, - { 0x0d, 0xf880 }, - - { 0x1f, 0x0001 }, - { 0x17, 0x0cc0 }, - - { 0x1f, 0x0001 }, - { 0x0b, 0xa4d8 }, - { 0x09, 0x281c }, - { 0x07, 0x2883 }, - { 0x0a, 0x6b35 }, - { 0x1d, 0x3da4 }, - { 0x1c, 0xeffd }, - { 0x14, 0x7f52 }, - { 0x18, 0x7fc6 }, - { 0x08, 0x0601 }, - { 0x06, 0x4063 }, - { 0x10, 0xf074 }, - { 0x1f, 0x0003 }, - { 0x13, 0x0789 }, - { 0x12, 0xf4bd }, - { 0x1a, 0x04fd }, - { 0x14, 0x84b0 }, - { 0x1f, 0x0000 }, - { 0x00, 0x9200 }, - - { 0x1f, 0x0005 }, - { 0x01, 0x0340 }, - { 0x1f, 0x0001 }, - { 0x04, 0x4000 }, - { 0x03, 0x1d21 }, - { 0x02, 0x0c32 }, - { 0x01, 0x0200 }, - { 0x00, 0x5554 }, - { 0x04, 0x4800 }, - { 0x04, 0x4000 }, - { 0x04, 0xf000 }, - { 0x03, 0xdf01 }, - { 0x02, 0xdf20 }, - { 0x01, 0x101a }, - { 0x00, 0xa0ff }, - { 0x04, 0xf800 }, - { 0x04, 0xf000 }, - { 0x1f, 0x0000 }, - }; - - rtl_writephy_batch(tp, phy_reg_init); - - r8168d_modify_extpage(tp->phydev, 0x0023, 0x16, 0xffff, 0x0000); -} - -static void rtl8168d_4_hw_phy_config(struct rtl8169_private *tp) -{ - phy_write_paged(tp->phydev, 0x0001, 0x17, 0x0cc0); - r8168d_modify_extpage(tp->phydev, 0x002d, 0x18, 0xffff, 0x0040); - phy_set_bits(tp->phydev, 0x0d, BIT(5)); -} - -static void rtl8168e_1_hw_phy_config(struct rtl8169_private *tp) -{ - static const struct phy_reg phy_reg_init[] = { - /* Channel estimation fine tune */ - { 0x1f, 0x0001 }, - { 0x0b, 0x6c20 }, - { 0x07, 0x2872 }, - { 0x1c, 0xefff }, - { 0x1f, 0x0003 }, - { 0x14, 0x6420 }, - { 0x1f, 0x0000 }, - }; - struct phy_device *phydev = tp->phydev; - - rtl_apply_firmware(tp); - - /* Enable Delay cap */ - r8168d_phy_param(phydev, 0x8b80, 0xffff, 0xc896); - - rtl_writephy_batch(tp, phy_reg_init); - - /* Update PFM & 10M TX idle timer */ - r8168d_modify_extpage(phydev, 0x002f, 0x15, 0xffff, 0x1919); - - r8168d_modify_extpage(phydev, 0x00ac, 0x18, 0xffff, 0x0006); - - /* DCO enable for 10M IDLE Power */ - r8168d_modify_extpage(phydev, 0x0023, 0x17, 0x0000, 0x0006); - - /* For impedance matching */ - phy_modify_paged(phydev, 0x0002, 0x08, 0x7f00, 0x8000); - - /* PHY auto speed down */ - r8168d_modify_extpage(phydev, 0x002d, 0x18, 0x0000, 0x0050); - phy_set_bits(phydev, 0x14, BIT(15)); - - r8168d_phy_param(phydev, 0x8b86, 0x0000, 0x0001); - r8168d_phy_param(phydev, 0x8b85, 0x2000, 0x0000); - - r8168d_modify_extpage(phydev, 0x0020, 0x15, 0x1100, 0x0000); - phy_write_paged(phydev, 0x0006, 0x00, 0x5a00); - - phy_write_mmd(phydev, MDIO_MMD_AN, MDIO_AN_EEE_ADV, 0x0000); -} - static void rtl_rar_exgmac_set(struct rtl8169_private *tp, u8 *addr) { const u16 w[] = { @@ -2917,286 +2211,7 @@ static void rtl_rar_exgmac_set(struct rtl8169_private *tp, u8 *addr) rtl_eri_write(tp, 0xf4, ERIAR_MASK_1111, w[1] | (w[2] << 16)); } -static void rtl8168e_2_hw_phy_config(struct rtl8169_private *tp) -{ - struct phy_device *phydev = tp->phydev; - - rtl_apply_firmware(tp); - - /* Enable Delay cap */ - r8168d_modify_extpage(phydev, 0x00ac, 0x18, 0xffff, 0x0006); - - /* Channel estimation fine tune */ - phy_write_paged(phydev, 0x0003, 0x09, 0xa20f); - - /* Green Setting */ - r8168d_phy_param(phydev, 0x8b5b, 0xffff, 0x9222); - r8168d_phy_param(phydev, 0x8b6d, 0xffff, 0x8000); - r8168d_phy_param(phydev, 0x8b76, 0xffff, 0x8000); - - /* For 4-corner performance improve */ - rtl_writephy(tp, 0x1f, 0x0005); - rtl_writephy(tp, 0x05, 0x8b80); - rtl_w0w1_phy(tp, 0x17, 0x0006, 0x0000); - rtl_writephy(tp, 0x1f, 0x0000); - - /* PHY auto speed down */ - r8168d_modify_extpage(phydev, 0x002d, 0x18, 0x0000, 0x0010); - phy_set_bits(phydev, 0x14, BIT(15)); - - /* improve 10M EEE waveform */ - r8168d_phy_param(phydev, 0x8b86, 0x0000, 0x0001); - - /* Improve 2-pair detection performance */ - r8168d_phy_param(phydev, 0x8b85, 0x0000, 0x4000); - - rtl8168f_config_eee_phy(tp); - - /* Green feature */ - rtl_writephy(tp, 0x1f, 0x0003); - rtl_w0w1_phy(tp, 0x19, 0x0001, 0x0000); - rtl_w0w1_phy(tp, 0x10, 0x0400, 0x0000); - rtl_writephy(tp, 0x1f, 0x0000); - rtl_writephy(tp, 0x1f, 0x0005); - rtl_w0w1_phy(tp, 0x01, 0x0100, 0x0000); - rtl_writephy(tp, 0x1f, 0x0000); -} - -static void rtl8168f_hw_phy_config(struct rtl8169_private *tp) -{ - struct phy_device *phydev = tp->phydev; - - /* For 4-corner performance improve */ - r8168d_phy_param(phydev, 0x8b80, 0x0000, 0x0006); - - /* PHY auto speed down */ - r8168d_modify_extpage(phydev, 0x002d, 0x18, 0x0000, 0x0010); - phy_set_bits(phydev, 0x14, BIT(15)); - - /* Improve 10M EEE waveform */ - r8168d_phy_param(phydev, 0x8b86, 0x0000, 0x0001); - - rtl8168f_config_eee_phy(tp); -} - -static void rtl8168f_1_hw_phy_config(struct rtl8169_private *tp) -{ - struct phy_device *phydev = tp->phydev; - - rtl_apply_firmware(tp); - - /* Channel estimation fine tune */ - phy_write_paged(phydev, 0x0003, 0x09, 0xa20f); - - /* Modify green table for giga & fnet */ - r8168d_phy_param(phydev, 0x8b55, 0xffff, 0x0000); - r8168d_phy_param(phydev, 0x8b5e, 0xffff, 0x0000); - r8168d_phy_param(phydev, 0x8b67, 0xffff, 0x0000); - r8168d_phy_param(phydev, 0x8b70, 0xffff, 0x0000); - r8168d_modify_extpage(phydev, 0x0078, 0x17, 0xffff, 0x0000); - r8168d_modify_extpage(phydev, 0x0078, 0x19, 0xffff, 0x00fb); - - /* Modify green table for 10M */ - r8168d_phy_param(phydev, 0x8b79, 0xffff, 0xaa00); - - /* Disable hiimpedance detection (RTCT) */ - phy_write_paged(phydev, 0x0003, 0x01, 0x328a); - - rtl8168f_hw_phy_config(tp); - - /* Improve 2-pair detection performance */ - r8168d_phy_param(phydev, 0x8b85, 0x0000, 0x4000); -} - -static void rtl8168f_2_hw_phy_config(struct rtl8169_private *tp) -{ - rtl_apply_firmware(tp); - - rtl8168f_hw_phy_config(tp); -} - -static void rtl8411_hw_phy_config(struct rtl8169_private *tp) -{ - struct phy_device *phydev = tp->phydev; - - rtl_apply_firmware(tp); - - rtl8168f_hw_phy_config(tp); - - /* Improve 2-pair detection performance */ - r8168d_phy_param(phydev, 0x8b85, 0x0000, 0x4000); - - /* Channel estimation fine tune */ - phy_write_paged(phydev, 0x0003, 0x09, 0xa20f); - - /* Modify green table for giga & fnet */ - r8168d_phy_param(phydev, 0x8b55, 0xffff, 0x0000); - r8168d_phy_param(phydev, 0x8b5e, 0xffff, 0x0000); - r8168d_phy_param(phydev, 0x8b67, 0xffff, 0x0000); - r8168d_phy_param(phydev, 0x8b70, 0xffff, 0x0000); - r8168d_modify_extpage(phydev, 0x0078, 0x17, 0xffff, 0x0000); - r8168d_modify_extpage(phydev, 0x0078, 0x19, 0xffff, 0x00aa); - - /* Modify green table for 10M */ - r8168d_phy_param(phydev, 0x8b79, 0xffff, 0xaa00); - - /* Disable hiimpedance detection (RTCT) */ - phy_write_paged(phydev, 0x0003, 0x01, 0x328a); - - /* Modify green table for giga */ - r8168d_phy_param(phydev, 0x8b54, 0x0800, 0x0000); - r8168d_phy_param(phydev, 0x8b5d, 0x0800, 0x0000); - r8168d_phy_param(phydev, 0x8a7c, 0x0100, 0x0000); - r8168d_phy_param(phydev, 0x8a7f, 0x0000, 0x0100); - r8168d_phy_param(phydev, 0x8a82, 0x0100, 0x0000); - r8168d_phy_param(phydev, 0x8a85, 0x0100, 0x0000); - r8168d_phy_param(phydev, 0x8a88, 0x0100, 0x0000); - - /* uc same-seed solution */ - r8168d_phy_param(phydev, 0x8b85, 0x0000, 0x8000); - - /* Green feature */ - rtl_writephy(tp, 0x1f, 0x0003); - rtl_w0w1_phy(tp, 0x19, 0x0000, 0x0001); - rtl_w0w1_phy(tp, 0x10, 0x0000, 0x0400); - rtl_writephy(tp, 0x1f, 0x0000); -} - -static void rtl8168g_disable_aldps(struct rtl8169_private *tp) -{ - phy_modify_paged(tp->phydev, 0x0a43, 0x10, BIT(2), 0); -} - -static void rtl8168g_phy_adjust_10m_aldps(struct rtl8169_private *tp) -{ - struct phy_device *phydev = tp->phydev; - - phy_modify_paged(phydev, 0x0bcc, 0x14, BIT(8), 0); - phy_modify_paged(phydev, 0x0a44, 0x11, 0, BIT(7) | BIT(6)); - r8168g_phy_param(phydev, 0x8084, 0x6000, 0x0000); - phy_modify_paged(phydev, 0x0a43, 0x10, 0x0000, 0x1003); -} - -static void rtl8168g_1_hw_phy_config(struct rtl8169_private *tp) -{ - int ret; - - rtl_apply_firmware(tp); - - ret = phy_read_paged(tp->phydev, 0x0a46, 0x10); - if (ret & BIT(8)) - phy_modify_paged(tp->phydev, 0x0bcc, 0x12, BIT(15), 0); - else - phy_modify_paged(tp->phydev, 0x0bcc, 0x12, 0, BIT(15)); - - ret = phy_read_paged(tp->phydev, 0x0a46, 0x13); - if (ret & BIT(8)) - phy_modify_paged(tp->phydev, 0x0c41, 0x15, 0, BIT(1)); - else - phy_modify_paged(tp->phydev, 0x0c41, 0x15, BIT(1), 0); - - /* Enable PHY auto speed down */ - phy_modify_paged(tp->phydev, 0x0a44, 0x11, 0, BIT(3) | BIT(2)); - - rtl8168g_phy_adjust_10m_aldps(tp); - - /* EEE auto-fallback function */ - phy_modify_paged(tp->phydev, 0x0a4b, 0x11, 0, BIT(2)); - - /* Enable UC LPF tune function */ - r8168g_phy_param(tp->phydev, 0x8012, 0x0000, 0x8000); - - phy_modify_paged(tp->phydev, 0x0c42, 0x11, BIT(13), BIT(14)); - - /* Improve SWR Efficiency */ - rtl_writephy(tp, 0x1f, 0x0bcd); - rtl_writephy(tp, 0x14, 0x5065); - rtl_writephy(tp, 0x14, 0xd065); - rtl_writephy(tp, 0x1f, 0x0bc8); - rtl_writephy(tp, 0x11, 0x5655); - rtl_writephy(tp, 0x1f, 0x0bcd); - rtl_writephy(tp, 0x14, 0x1065); - rtl_writephy(tp, 0x14, 0x9065); - rtl_writephy(tp, 0x14, 0x1065); - rtl_writephy(tp, 0x1f, 0x0000); - - rtl8168g_disable_aldps(tp); - rtl8168g_config_eee_phy(tp); -} - -static void rtl8168g_2_hw_phy_config(struct rtl8169_private *tp) -{ - rtl_apply_firmware(tp); - rtl8168g_config_eee_phy(tp); -} - -static void rtl8168h_1_hw_phy_config(struct rtl8169_private *tp) -{ - struct phy_device *phydev = tp->phydev; - u16 dout_tapbin; - u32 data; - - rtl_apply_firmware(tp); - - /* CHN EST parameters adjust - giga master */ - r8168g_phy_param(phydev, 0x809b, 0xf800, 0x8000); - r8168g_phy_param(phydev, 0x80a2, 0xff00, 0x8000); - r8168g_phy_param(phydev, 0x80a4, 0xff00, 0x8500); - r8168g_phy_param(phydev, 0x809c, 0xff00, 0xbd00); - - /* CHN EST parameters adjust - giga slave */ - r8168g_phy_param(phydev, 0x80ad, 0xf800, 0x7000); - r8168g_phy_param(phydev, 0x80b4, 0xff00, 0x5000); - r8168g_phy_param(phydev, 0x80ac, 0xff00, 0x4000); - - /* CHN EST parameters adjust - fnet */ - r8168g_phy_param(phydev, 0x808e, 0xff00, 0x1200); - r8168g_phy_param(phydev, 0x8090, 0xff00, 0xe500); - r8168g_phy_param(phydev, 0x8092, 0xff00, 0x9f00); - - /* enable R-tune & PGA-retune function */ - dout_tapbin = 0; - data = phy_read_paged(phydev, 0x0a46, 0x13); - data &= 3; - data <<= 2; - dout_tapbin |= data; - data = phy_read_paged(phydev, 0x0a46, 0x12); - data &= 0xc000; - data >>= 14; - dout_tapbin |= data; - dout_tapbin = ~(dout_tapbin^0x08); - dout_tapbin <<= 12; - dout_tapbin &= 0xf000; - - r8168g_phy_param(phydev, 0x827a, 0xf000, dout_tapbin); - r8168g_phy_param(phydev, 0x827b, 0xf000, dout_tapbin); - r8168g_phy_param(phydev, 0x827c, 0xf000, dout_tapbin); - r8168g_phy_param(phydev, 0x827d, 0xf000, dout_tapbin); - r8168g_phy_param(phydev, 0x0811, 0x0000, 0x0800); - phy_modify_paged(phydev, 0x0a42, 0x16, 0x0000, 0x0002); - - /* enable GPHY 10M */ - phy_modify_paged(tp->phydev, 0x0a44, 0x11, 0, BIT(11)); - - /* SAR ADC performance */ - phy_modify_paged(tp->phydev, 0x0bca, 0x17, BIT(12) | BIT(13), BIT(14)); - - r8168g_phy_param(phydev, 0x803f, 0x3000, 0x0000); - r8168g_phy_param(phydev, 0x8047, 0x3000, 0x0000); - r8168g_phy_param(phydev, 0x804f, 0x3000, 0x0000); - r8168g_phy_param(phydev, 0x8057, 0x3000, 0x0000); - r8168g_phy_param(phydev, 0x805f, 0x3000, 0x0000); - r8168g_phy_param(phydev, 0x8067, 0x3000, 0x0000); - r8168g_phy_param(phydev, 0x806f, 0x3000, 0x0000); - - /* disable phy pfm mode */ - phy_modify_paged(tp->phydev, 0x0a44, 0x11, BIT(7), 0); - - rtl8168g_disable_aldps(tp); - rtl8168h_config_eee_phy(tp); -} - -static u16 rtl8168h_2_get_adc_bias_ioffset(struct rtl8169_private *tp) +u16 rtl8168h_2_get_adc_bias_ioffset(struct rtl8169_private *tp) { u16 data1, data2, ioffset; @@ -3212,410 +2227,28 @@ static u16 rtl8168h_2_get_adc_bias_ioffset(struct rtl8169_private *tp) return ioffset; } -static void rtl8168h_2_hw_phy_config(struct rtl8169_private *tp) -{ - struct phy_device *phydev = tp->phydev; - u16 ioffset, rlen; - u32 data; - - rtl_apply_firmware(tp); - - /* CHIN EST parameter update */ - r8168g_phy_param(phydev, 0x808a, 0x003f, 0x000a); - - /* enable R-tune & PGA-retune function */ - r8168g_phy_param(phydev, 0x0811, 0x0000, 0x0800); - phy_modify_paged(phydev, 0x0a42, 0x16, 0x0000, 0x0002); - - /* enable GPHY 10M */ - phy_modify_paged(tp->phydev, 0x0a44, 0x11, 0, BIT(11)); - - ioffset = rtl8168h_2_get_adc_bias_ioffset(tp); - if (ioffset != 0xffff) - phy_write_paged(phydev, 0x0bcf, 0x16, ioffset); - - /* Modify rlen (TX LPF corner frequency) level */ - data = phy_read_paged(phydev, 0x0bcd, 0x16); - data &= 0x000f; - rlen = 0; - if (data > 3) - rlen = data - 3; - data = rlen | (rlen<<4) | (rlen<<8) | (rlen<<12); - phy_write_paged(phydev, 0x0bcd, 0x17, data); - - /* disable phy pfm mode */ - phy_modify_paged(phydev, 0x0a44, 0x11, BIT(7), 0); - - rtl8168g_disable_aldps(tp); - rtl8168g_config_eee_phy(tp); -} - -static void rtl8168ep_1_hw_phy_config(struct rtl8169_private *tp) -{ - struct phy_device *phydev = tp->phydev; - - /* Enable PHY auto speed down */ - phy_modify_paged(phydev, 0x0a44, 0x11, 0, BIT(3) | BIT(2)); - - rtl8168g_phy_adjust_10m_aldps(tp); - - /* Enable EEE auto-fallback function */ - phy_modify_paged(phydev, 0x0a4b, 0x11, 0, BIT(2)); - - /* Enable UC LPF tune function */ - r8168g_phy_param(phydev, 0x8012, 0x0000, 0x8000); - - /* set rg_sel_sdm_rate */ - phy_modify_paged(phydev, 0x0c42, 0x11, BIT(13), BIT(14)); - - rtl8168g_disable_aldps(tp); - rtl8168g_config_eee_phy(tp); -} - -static void rtl8168ep_2_hw_phy_config(struct rtl8169_private *tp) -{ - struct phy_device *phydev = tp->phydev; - - rtl8168g_phy_adjust_10m_aldps(tp); - - /* Enable UC LPF tune function */ - r8168g_phy_param(phydev, 0x8012, 0x0000, 0x8000); - - /* Set rg_sel_sdm_rate */ - phy_modify_paged(tp->phydev, 0x0c42, 0x11, BIT(13), BIT(14)); - - /* Channel estimation parameters */ - r8168g_phy_param(phydev, 0x80f3, 0xff00, 0x8b00); - r8168g_phy_param(phydev, 0x80f0, 0xff00, 0x3a00); - r8168g_phy_param(phydev, 0x80ef, 0xff00, 0x0500); - r8168g_phy_param(phydev, 0x80f6, 0xff00, 0x6e00); - r8168g_phy_param(phydev, 0x80ec, 0xff00, 0x6800); - r8168g_phy_param(phydev, 0x80ed, 0xff00, 0x7c00); - r8168g_phy_param(phydev, 0x80f2, 0xff00, 0xf400); - r8168g_phy_param(phydev, 0x80f4, 0xff00, 0x8500); - r8168g_phy_param(phydev, 0x8110, 0xff00, 0xa800); - r8168g_phy_param(phydev, 0x810f, 0xff00, 0x1d00); - r8168g_phy_param(phydev, 0x8111, 0xff00, 0xf500); - r8168g_phy_param(phydev, 0x8113, 0xff00, 0x6100); - r8168g_phy_param(phydev, 0x8115, 0xff00, 0x9200); - r8168g_phy_param(phydev, 0x810e, 0xff00, 0x0400); - r8168g_phy_param(phydev, 0x810c, 0xff00, 0x7c00); - r8168g_phy_param(phydev, 0x810b, 0xff00, 0x5a00); - r8168g_phy_param(phydev, 0x80d1, 0xff00, 0xff00); - r8168g_phy_param(phydev, 0x80cd, 0xff00, 0x9e00); - r8168g_phy_param(phydev, 0x80d3, 0xff00, 0x0e00); - r8168g_phy_param(phydev, 0x80d5, 0xff00, 0xca00); - r8168g_phy_param(phydev, 0x80d7, 0xff00, 0x8400); - - /* Force PWM-mode */ - rtl_writephy(tp, 0x1f, 0x0bcd); - rtl_writephy(tp, 0x14, 0x5065); - rtl_writephy(tp, 0x14, 0xd065); - rtl_writephy(tp, 0x1f, 0x0bc8); - rtl_writephy(tp, 0x12, 0x00ed); - rtl_writephy(tp, 0x1f, 0x0bcd); - rtl_writephy(tp, 0x14, 0x1065); - rtl_writephy(tp, 0x14, 0x9065); - rtl_writephy(tp, 0x14, 0x1065); - rtl_writephy(tp, 0x1f, 0x0000); - - rtl8168g_disable_aldps(tp); - rtl8168g_config_eee_phy(tp); -} - -static void rtl8117_hw_phy_config(struct rtl8169_private *tp) -{ - struct phy_device *phydev = tp->phydev; - - /* CHN EST parameters adjust - fnet */ - r8168g_phy_param(phydev, 0x808e, 0xff00, 0x4800); - r8168g_phy_param(phydev, 0x8090, 0xff00, 0xcc00); - r8168g_phy_param(phydev, 0x8092, 0xff00, 0xb000); - - r8168g_phy_param(phydev, 0x8088, 0xff00, 0x6000); - r8168g_phy_param(phydev, 0x808b, 0x3f00, 0x0b00); - r8168g_phy_param(phydev, 0x808d, 0x1f00, 0x0600); - r8168g_phy_param(phydev, 0x808c, 0xff00, 0xb000); - r8168g_phy_param(phydev, 0x80a0, 0xff00, 0x2800); - r8168g_phy_param(phydev, 0x80a2, 0xff00, 0x5000); - r8168g_phy_param(phydev, 0x809b, 0xf800, 0xb000); - r8168g_phy_param(phydev, 0x809a, 0xff00, 0x4b00); - r8168g_phy_param(phydev, 0x809d, 0x3f00, 0x0800); - r8168g_phy_param(phydev, 0x80a1, 0xff00, 0x7000); - r8168g_phy_param(phydev, 0x809f, 0x1f00, 0x0300); - r8168g_phy_param(phydev, 0x809e, 0xff00, 0x8800); - r8168g_phy_param(phydev, 0x80b2, 0xff00, 0x2200); - r8168g_phy_param(phydev, 0x80ad, 0xf800, 0x9800); - r8168g_phy_param(phydev, 0x80af, 0x3f00, 0x0800); - r8168g_phy_param(phydev, 0x80b3, 0xff00, 0x6f00); - r8168g_phy_param(phydev, 0x80b1, 0x1f00, 0x0300); - r8168g_phy_param(phydev, 0x80b0, 0xff00, 0x9300); - - r8168g_phy_param(phydev, 0x8011, 0x0000, 0x0800); - - /* enable GPHY 10M */ - phy_modify_paged(tp->phydev, 0x0a44, 0x11, 0, BIT(11)); - - r8168g_phy_param(phydev, 0x8016, 0x0000, 0x0400); - - rtl8168g_disable_aldps(tp); - rtl8168h_config_eee_phy(tp); -} - -static void rtl8102e_hw_phy_config(struct rtl8169_private *tp) -{ - static const struct phy_reg phy_reg_init[] = { - { 0x1f, 0x0003 }, - { 0x08, 0x441d }, - { 0x01, 0x9100 }, - { 0x1f, 0x0000 } - }; - - rtl_writephy(tp, 0x1f, 0x0000); - rtl_patchphy(tp, 0x11, 1 << 12); - rtl_patchphy(tp, 0x19, 1 << 13); - rtl_patchphy(tp, 0x10, 1 << 15); - - rtl_writephy_batch(tp, phy_reg_init); -} - -static void rtl8105e_hw_phy_config(struct rtl8169_private *tp) -{ - /* Disable ALDPS before ram code */ - phy_write(tp->phydev, 0x18, 0x0310); - msleep(100); - - rtl_apply_firmware(tp); - - phy_write_paged(tp->phydev, 0x0005, 0x1a, 0x0000); - phy_write_paged(tp->phydev, 0x0004, 0x1c, 0x0000); - phy_write_paged(tp->phydev, 0x0001, 0x15, 0x7701); -} - -static void rtl8402_hw_phy_config(struct rtl8169_private *tp) -{ - /* Disable ALDPS before setting firmware */ - phy_write(tp->phydev, 0x18, 0x0310); - msleep(20); - - rtl_apply_firmware(tp); - - /* EEE setting */ - rtl_eri_write(tp, 0x1b0, ERIAR_MASK_0011, 0x0000); - rtl_writephy(tp, 0x1f, 0x0004); - rtl_writephy(tp, 0x10, 0x401f); - rtl_writephy(tp, 0x19, 0x7030); - rtl_writephy(tp, 0x1f, 0x0000); -} - -static void rtl8106e_hw_phy_config(struct rtl8169_private *tp) -{ - static const struct phy_reg phy_reg_init[] = { - { 0x1f, 0x0004 }, - { 0x10, 0xc07f }, - { 0x19, 0x7030 }, - { 0x1f, 0x0000 } - }; - - /* Disable ALDPS before ram code */ - phy_write(tp->phydev, 0x18, 0x0310); - msleep(100); - - rtl_apply_firmware(tp); - - rtl_eri_write(tp, 0x1b0, ERIAR_MASK_0011, 0x0000); - rtl_writephy_batch(tp, phy_reg_init); - - rtl_eri_write(tp, 0x1d0, ERIAR_MASK_0011, 0x0000); -} - -static void rtl8125_1_hw_phy_config(struct rtl8169_private *tp) -{ - struct phy_device *phydev = tp->phydev; - - phy_modify_paged(phydev, 0xad4, 0x10, 0x03ff, 0x0084); - phy_modify_paged(phydev, 0xad4, 0x17, 0x0000, 0x0010); - phy_modify_paged(phydev, 0xad1, 0x13, 0x03ff, 0x0006); - phy_modify_paged(phydev, 0xad3, 0x11, 0x003f, 0x0006); - phy_modify_paged(phydev, 0xac0, 0x14, 0x0000, 0x1100); - phy_modify_paged(phydev, 0xac8, 0x15, 0xf000, 0x7000); - phy_modify_paged(phydev, 0xad1, 0x14, 0x0000, 0x0400); - phy_modify_paged(phydev, 0xad1, 0x15, 0x0000, 0x03ff); - phy_modify_paged(phydev, 0xad1, 0x16, 0x0000, 0x03ff); - - r8168g_phy_param(phydev, 0x80ea, 0xff00, 0xc400); - r8168g_phy_param(phydev, 0x80eb, 0x0700, 0x0300); - r8168g_phy_param(phydev, 0x80f8, 0xff00, 0x1c00); - r8168g_phy_param(phydev, 0x80f1, 0xff00, 0x3000); - r8168g_phy_param(phydev, 0x80fe, 0xff00, 0xa500); - r8168g_phy_param(phydev, 0x8102, 0xff00, 0x5000); - r8168g_phy_param(phydev, 0x8105, 0xff00, 0x3300); - r8168g_phy_param(phydev, 0x8100, 0xff00, 0x7000); - r8168g_phy_param(phydev, 0x8104, 0xff00, 0xf000); - r8168g_phy_param(phydev, 0x8106, 0xff00, 0x6500); - r8168g_phy_param(phydev, 0x80dc, 0xff00, 0xed00); - r8168g_phy_param(phydev, 0x80df, 0x0000, 0x0100); - r8168g_phy_param(phydev, 0x80e1, 0x0100, 0x0000); - - phy_modify_paged(phydev, 0xbf0, 0x13, 0x003f, 0x0038); - r8168g_phy_param(phydev, 0x819f, 0xffff, 0xd0b6); - - phy_write_paged(phydev, 0xbc3, 0x12, 0x5555); - phy_modify_paged(phydev, 0xbf0, 0x15, 0x0e00, 0x0a00); - phy_modify_paged(phydev, 0xa5c, 0x10, 0x0400, 0x0000); - phy_modify_paged(phydev, 0xa44, 0x11, 0x0000, 0x0800); - - rtl8125_config_eee_phy(tp); -} - -static void rtl8125_2_hw_phy_config(struct rtl8169_private *tp) -{ - struct phy_device *phydev = tp->phydev; - int i; - - phy_modify_paged(phydev, 0xad4, 0x17, 0x0000, 0x0010); - phy_modify_paged(phydev, 0xad1, 0x13, 0x03ff, 0x03ff); - phy_modify_paged(phydev, 0xad3, 0x11, 0x003f, 0x0006); - phy_modify_paged(phydev, 0xac0, 0x14, 0x1100, 0x0000); - phy_modify_paged(phydev, 0xacc, 0x10, 0x0003, 0x0002); - phy_modify_paged(phydev, 0xad4, 0x10, 0x00e7, 0x0044); - phy_modify_paged(phydev, 0xac1, 0x12, 0x0080, 0x0000); - phy_modify_paged(phydev, 0xac8, 0x10, 0x0300, 0x0000); - phy_modify_paged(phydev, 0xac5, 0x17, 0x0007, 0x0002); - phy_write_paged(phydev, 0xad4, 0x16, 0x00a8); - phy_write_paged(phydev, 0xac5, 0x16, 0x01ff); - phy_modify_paged(phydev, 0xac8, 0x15, 0x00f0, 0x0030); - - phy_write(phydev, 0x1f, 0x0b87); - phy_write(phydev, 0x16, 0x80a2); - phy_write(phydev, 0x17, 0x0153); - phy_write(phydev, 0x16, 0x809c); - phy_write(phydev, 0x17, 0x0153); - phy_write(phydev, 0x1f, 0x0000); - - phy_write(phydev, 0x1f, 0x0a43); - phy_write(phydev, 0x13, 0x81B3); - phy_write(phydev, 0x14, 0x0043); - phy_write(phydev, 0x14, 0x00A7); - phy_write(phydev, 0x14, 0x00D6); - phy_write(phydev, 0x14, 0x00EC); - phy_write(phydev, 0x14, 0x00F6); - phy_write(phydev, 0x14, 0x00FB); - phy_write(phydev, 0x14, 0x00FD); - phy_write(phydev, 0x14, 0x00FF); - phy_write(phydev, 0x14, 0x00BB); - phy_write(phydev, 0x14, 0x0058); - phy_write(phydev, 0x14, 0x0029); - phy_write(phydev, 0x14, 0x0013); - phy_write(phydev, 0x14, 0x0009); - phy_write(phydev, 0x14, 0x0004); - phy_write(phydev, 0x14, 0x0002); - for (i = 0; i < 25; i++) - phy_write(phydev, 0x14, 0x0000); - phy_write(phydev, 0x1f, 0x0000); - - r8168g_phy_param(phydev, 0x8257, 0xffff, 0x020F); - r8168g_phy_param(phydev, 0x80ea, 0xffff, 0x7843); - - rtl_apply_firmware(tp); - - phy_modify_paged(phydev, 0xd06, 0x14, 0x0000, 0x2000); - - r8168g_phy_param(phydev, 0x81a2, 0x0000, 0x0100); - - phy_modify_paged(phydev, 0xb54, 0x16, 0xff00, 0xdb00); - phy_modify_paged(phydev, 0xa45, 0x12, 0x0001, 0x0000); - phy_modify_paged(phydev, 0xa5d, 0x12, 0x0000, 0x0020); - phy_modify_paged(phydev, 0xad4, 0x17, 0x0010, 0x0000); - phy_modify_paged(phydev, 0xa86, 0x15, 0x0001, 0x0000); - phy_modify_paged(phydev, 0xa44, 0x11, 0x0000, 0x0800); - - rtl8125_config_eee_phy(tp); -} - -static void rtl_hw_phy_config(struct net_device *dev) -{ - static const rtl_generic_fct phy_configs[] = { - /* PCI devices. */ - [RTL_GIGA_MAC_VER_02] = rtl8169s_hw_phy_config, - [RTL_GIGA_MAC_VER_03] = rtl8169s_hw_phy_config, - [RTL_GIGA_MAC_VER_04] = rtl8169sb_hw_phy_config, - [RTL_GIGA_MAC_VER_05] = rtl8169scd_hw_phy_config, - [RTL_GIGA_MAC_VER_06] = rtl8169sce_hw_phy_config, - /* PCI-E devices. */ - [RTL_GIGA_MAC_VER_07] = rtl8102e_hw_phy_config, - [RTL_GIGA_MAC_VER_08] = rtl8102e_hw_phy_config, - [RTL_GIGA_MAC_VER_09] = rtl8102e_hw_phy_config, - [RTL_GIGA_MAC_VER_10] = NULL, - [RTL_GIGA_MAC_VER_11] = rtl8168bb_hw_phy_config, - [RTL_GIGA_MAC_VER_12] = rtl8168bef_hw_phy_config, - [RTL_GIGA_MAC_VER_13] = NULL, - [RTL_GIGA_MAC_VER_14] = NULL, - [RTL_GIGA_MAC_VER_15] = NULL, - [RTL_GIGA_MAC_VER_16] = NULL, - [RTL_GIGA_MAC_VER_17] = rtl8168bef_hw_phy_config, - [RTL_GIGA_MAC_VER_18] = rtl8168cp_1_hw_phy_config, - [RTL_GIGA_MAC_VER_19] = rtl8168c_1_hw_phy_config, - [RTL_GIGA_MAC_VER_20] = rtl8168c_2_hw_phy_config, - [RTL_GIGA_MAC_VER_21] = rtl8168c_3_hw_phy_config, - [RTL_GIGA_MAC_VER_22] = rtl8168c_3_hw_phy_config, - [RTL_GIGA_MAC_VER_23] = rtl8168cp_2_hw_phy_config, - [RTL_GIGA_MAC_VER_24] = rtl8168cp_2_hw_phy_config, - [RTL_GIGA_MAC_VER_25] = rtl8168d_1_hw_phy_config, - [RTL_GIGA_MAC_VER_26] = rtl8168d_2_hw_phy_config, - [RTL_GIGA_MAC_VER_27] = rtl8168d_3_hw_phy_config, - [RTL_GIGA_MAC_VER_28] = rtl8168d_4_hw_phy_config, - [RTL_GIGA_MAC_VER_29] = rtl8105e_hw_phy_config, - [RTL_GIGA_MAC_VER_30] = rtl8105e_hw_phy_config, - [RTL_GIGA_MAC_VER_31] = NULL, - [RTL_GIGA_MAC_VER_32] = rtl8168e_1_hw_phy_config, - [RTL_GIGA_MAC_VER_33] = rtl8168e_1_hw_phy_config, - [RTL_GIGA_MAC_VER_34] = rtl8168e_2_hw_phy_config, - [RTL_GIGA_MAC_VER_35] = rtl8168f_1_hw_phy_config, - [RTL_GIGA_MAC_VER_36] = rtl8168f_2_hw_phy_config, - [RTL_GIGA_MAC_VER_37] = rtl8402_hw_phy_config, - [RTL_GIGA_MAC_VER_38] = rtl8411_hw_phy_config, - [RTL_GIGA_MAC_VER_39] = rtl8106e_hw_phy_config, - [RTL_GIGA_MAC_VER_40] = rtl8168g_1_hw_phy_config, - [RTL_GIGA_MAC_VER_41] = NULL, - [RTL_GIGA_MAC_VER_42] = rtl8168g_2_hw_phy_config, - [RTL_GIGA_MAC_VER_43] = rtl8168g_2_hw_phy_config, - [RTL_GIGA_MAC_VER_44] = rtl8168g_2_hw_phy_config, - [RTL_GIGA_MAC_VER_45] = rtl8168h_1_hw_phy_config, - [RTL_GIGA_MAC_VER_46] = rtl8168h_2_hw_phy_config, - [RTL_GIGA_MAC_VER_47] = rtl8168h_1_hw_phy_config, - [RTL_GIGA_MAC_VER_48] = rtl8168h_2_hw_phy_config, - [RTL_GIGA_MAC_VER_49] = rtl8168ep_1_hw_phy_config, - [RTL_GIGA_MAC_VER_50] = rtl8168ep_2_hw_phy_config, - [RTL_GIGA_MAC_VER_51] = rtl8168ep_2_hw_phy_config, - [RTL_GIGA_MAC_VER_52] = rtl8117_hw_phy_config, - [RTL_GIGA_MAC_VER_60] = rtl8125_1_hw_phy_config, - [RTL_GIGA_MAC_VER_61] = rtl8125_2_hw_phy_config, - }; - struct rtl8169_private *tp = netdev_priv(dev); - - if (phy_configs[tp->mac_version]) - phy_configs[tp->mac_version](tp); -} - static void rtl_schedule_task(struct rtl8169_private *tp, enum rtl_flag flag) { if (!test_and_set_bit(flag, tp->wk.flags)) schedule_work(&tp->wk.work); } -static void rtl8169_init_phy(struct net_device *dev, struct rtl8169_private *tp) +static void rtl8169_init_phy(struct rtl8169_private *tp) { - rtl_hw_phy_config(dev); + r8169_hw_phy_config(tp, tp->phydev, tp->mac_version); if (tp->mac_version <= RTL_GIGA_MAC_VER_06) { pci_write_config_byte(tp->pci_dev, PCI_LATENCY_TIMER, 0x40); pci_write_config_byte(tp->pci_dev, PCI_CACHE_LINE_SIZE, 0x08); - netif_dbg(tp, drv, dev, - "Set MAC Reg C+CR Offset 0x82h = 0x01h\n"); + /* set undocumented MAC Reg C+CR Offset 0x82h */ RTL_W8(tp, 0x82, 0x01); } + if (tp->mac_version == RTL_GIGA_MAC_VER_05 && + tp->pci_dev->subsystem_vendor == PCI_VENDOR_ID_GIGABYTE && + tp->pci_dev->subsystem_device == 0xe000) + phy_write_paged(tp->phydev, 0x0001, 0x10, 0xf01b); + /* We may have called phy_speed_down before */ phy_speed_up(tp->phydev); @@ -3665,16 +2298,6 @@ static int rtl_set_mac_address(struct net_device *dev, void *p) return 0; } -static int rtl8169_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd) -{ - struct rtl8169_private *tp = netdev_priv(dev); - - if (!netif_running(dev)) - return -ENODEV; - - return phy_mii_ioctl(tp->phydev, ifr, cmd); -} - static void rtl_wol_suspend_quirk(struct rtl8169_private *tp) { switch (tp->mac_version) { @@ -4700,9 +3323,7 @@ static void rtl_hw_start_8168h_1(struct rtl8169_private *tp) rtl_pcie_state_l2l3_disable(tp); - rtl_writephy(tp, 0x1f, 0x0c42); - rg_saw_cnt = (rtl_readphy(tp, 0x13) & 0x3fff); - rtl_writephy(tp, 0x1f, 0x0000); + rg_saw_cnt = phy_read_paged(tp->phydev, 0x0c42, 0x13) & 0x3fff; if (rg_saw_cnt > 0) { u16 sw_cnt_1ms_ini; @@ -4877,7 +3498,7 @@ static void rtl_hw_start_8117(struct rtl8169_private *tp) r8168_mac_ocp_write(tp, 0xc09e, 0x0000); /* firmware is for MAC only */ - rtl_apply_firmware(tp); + r8169_apply_firmware(tp); rtl_hw_aspm_clkreq_enable(tp, true); } @@ -4981,6 +3602,9 @@ static void rtl_hw_start_8402(struct rtl8169_private *tp) rtl_eri_write(tp, 0xb8, ERIAR_MASK_0011, 0x0000); rtl_w0w1_eri(tp, 0x0d4, ERIAR_MASK_0011, 0x0e00, 0xff00); + /* disable EEE */ + rtl_eri_write(tp, 0x1b0, ERIAR_MASK_0011, 0x0000); + rtl_pcie_state_l2l3_disable(tp); } @@ -4995,6 +3619,11 @@ static void rtl_hw_start_8106(struct rtl8169_private *tp) RTL_W8(tp, MCU, RTL_R8(tp, MCU) | EN_NDP | EN_OOB_RESET); RTL_W8(tp, DLLPR, RTL_R8(tp, DLLPR) & ~PFM_EN); + rtl_eri_write(tp, 0x1d0, ERIAR_MASK_0011, 0x0000); + + /* disable EEE */ + rtl_eri_write(tp, 0x1b0, ERIAR_MASK_0011, 0x0000); + rtl_pcie_state_l2l3_disable(tp); rtl_hw_aspm_clkreq_enable(tp, true); } @@ -5212,11 +3841,8 @@ static void rtl_hw_start_8169(struct rtl8169_private *tp) tp->cp_cmd |= PCIMulRW; if (tp->mac_version == RTL_GIGA_MAC_VER_02 || - tp->mac_version == RTL_GIGA_MAC_VER_03) { - netif_dbg(tp, drv, tp->dev, - "Set MAC Reg C+CR Offset 0xe0. Bit 3 and Bit 14 MUST be 1\n"); - tp->cp_cmd |= (1 << 14); - } + tp->mac_version == RTL_GIGA_MAC_VER_03) + tp->cp_cmd |= EnAnaPLL; RTL_W16(tp, CPlusCmd, tp->cp_cmd); @@ -6230,7 +4856,7 @@ static int rtl_open(struct net_device *dev) napi_enable(&tp->napi); - rtl8169_init_phy(dev, tp); + rtl8169_init_phy(tp); rtl_pll_power_up(tp); @@ -6361,7 +4987,7 @@ static void __rtl8169_resume(struct net_device *dev) netif_device_attach(dev); rtl_pll_power_up(tp); - rtl8169_init_phy(dev, tp); + rtl8169_init_phy(tp); phy_start(tp->phydev); @@ -6532,7 +5158,7 @@ static const struct net_device_ops rtl_netdev_ops = { .ndo_fix_features = rtl8169_fix_features, .ndo_set_features = rtl8169_set_features, .ndo_set_mac_address = rtl_set_mac_address, - .ndo_do_ioctl = rtl8169_ioctl, + .ndo_do_ioctl = phy_do_ioctl_running, .ndo_set_rx_mode = rtl_set_rx_mode, #ifdef CONFIG_NET_POLL_CONTROLLER .ndo_poll_controller = rtl8169_netpoll, diff --git a/drivers/net/ethernet/realtek/r8169_phy_config.c b/drivers/net/ethernet/realtek/r8169_phy_config.c new file mode 100644 index 000000000000..e367e77c773b --- /dev/null +++ b/drivers/net/ethernet/realtek/r8169_phy_config.c @@ -0,0 +1,1307 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * r8169_phy_config.c: RealTek 8169/8168/8101 ethernet driver. + * + * Copyright (c) 2002 ShuChen <shuchen@realtek.com.tw> + * Copyright (c) 2003 - 2007 Francois Romieu <romieu@fr.zoreil.com> + * Copyright (c) a lot of people too. Please respect their work. + * + * See MAINTAINERS file for support contact information. + */ + +#include <linux/delay.h> +#include <linux/phy.h> + +#include "r8169.h" + +typedef void (*rtl_phy_cfg_fct)(struct rtl8169_private *tp, + struct phy_device *phydev); + +static void r8168d_modify_extpage(struct phy_device *phydev, int extpage, + int reg, u16 mask, u16 val) +{ + int oldpage = phy_select_page(phydev, 0x0007); + + __phy_write(phydev, 0x1e, extpage); + __phy_modify(phydev, reg, mask, val); + + phy_restore_page(phydev, oldpage, 0); +} + +static void r8168d_phy_param(struct phy_device *phydev, u16 parm, + u16 mask, u16 val) +{ + int oldpage = phy_select_page(phydev, 0x0005); + + __phy_write(phydev, 0x05, parm); + __phy_modify(phydev, 0x06, mask, val); + + phy_restore_page(phydev, oldpage, 0); +} + +static void r8168g_phy_param(struct phy_device *phydev, u16 parm, + u16 mask, u16 val) +{ + int oldpage = phy_select_page(phydev, 0x0a43); + + __phy_write(phydev, 0x13, parm); + __phy_modify(phydev, 0x14, mask, val); + + phy_restore_page(phydev, oldpage, 0); +} + +struct phy_reg { + u16 reg; + u16 val; +}; + +static void __rtl_writephy_batch(struct phy_device *phydev, + const struct phy_reg *regs, int len) +{ + phy_lock_mdio_bus(phydev); + + while (len-- > 0) { + __phy_write(phydev, regs->reg, regs->val); + regs++; + } + + phy_unlock_mdio_bus(phydev); +} + +#define rtl_writephy_batch(p, a) __rtl_writephy_batch(p, a, ARRAY_SIZE(a)) + +static void rtl8168f_config_eee_phy(struct phy_device *phydev) +{ + r8168d_modify_extpage(phydev, 0x0020, 0x15, 0, BIT(8)); + r8168d_phy_param(phydev, 0x8b85, 0, BIT(13)); +} + +static void rtl8168g_config_eee_phy(struct phy_device *phydev) +{ + phy_modify_paged(phydev, 0x0a43, 0x11, 0, BIT(4)); +} + +static void rtl8168h_config_eee_phy(struct phy_device *phydev) +{ + rtl8168g_config_eee_phy(phydev); + + phy_modify_paged(phydev, 0xa4a, 0x11, 0x0000, 0x0200); + phy_modify_paged(phydev, 0xa42, 0x14, 0x0000, 0x0080); +} + +static void rtl8125_config_eee_phy(struct phy_device *phydev) +{ + rtl8168h_config_eee_phy(phydev); + + phy_modify_paged(phydev, 0xa6d, 0x12, 0x0001, 0x0000); + phy_modify_paged(phydev, 0xa6d, 0x14, 0x0010, 0x0000); +} + +static void rtl8169s_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + static const struct phy_reg phy_reg_init[] = { + { 0x1f, 0x0001 }, + { 0x06, 0x006e }, + { 0x08, 0x0708 }, + { 0x15, 0x4000 }, + { 0x18, 0x65c7 }, + + { 0x1f, 0x0001 }, + { 0x03, 0x00a1 }, + { 0x02, 0x0008 }, + { 0x01, 0x0120 }, + { 0x00, 0x1000 }, + { 0x04, 0x0800 }, + { 0x04, 0x0000 }, + + { 0x03, 0xff41 }, + { 0x02, 0xdf60 }, + { 0x01, 0x0140 }, + { 0x00, 0x0077 }, + { 0x04, 0x7800 }, + { 0x04, 0x7000 }, + + { 0x03, 0x802f }, + { 0x02, 0x4f02 }, + { 0x01, 0x0409 }, + { 0x00, 0xf0f9 }, + { 0x04, 0x9800 }, + { 0x04, 0x9000 }, + + { 0x03, 0xdf01 }, + { 0x02, 0xdf20 }, + { 0x01, 0xff95 }, + { 0x00, 0xba00 }, + { 0x04, 0xa800 }, + { 0x04, 0xa000 }, + + { 0x03, 0xff41 }, + { 0x02, 0xdf20 }, + { 0x01, 0x0140 }, + { 0x00, 0x00bb }, + { 0x04, 0xb800 }, + { 0x04, 0xb000 }, + + { 0x03, 0xdf41 }, + { 0x02, 0xdc60 }, + { 0x01, 0x6340 }, + { 0x00, 0x007d }, + { 0x04, 0xd800 }, + { 0x04, 0xd000 }, + + { 0x03, 0xdf01 }, + { 0x02, 0xdf20 }, + { 0x01, 0x100a }, + { 0x00, 0xa0ff }, + { 0x04, 0xf800 }, + { 0x04, 0xf000 }, + + { 0x1f, 0x0000 }, + { 0x0b, 0x0000 }, + { 0x00, 0x9200 } + }; + + rtl_writephy_batch(phydev, phy_reg_init); +} + +static void rtl8169sb_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + phy_write_paged(phydev, 0x0002, 0x01, 0x90d0); +} + +static void rtl8169scd_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + static const struct phy_reg phy_reg_init[] = { + { 0x1f, 0x0001 }, + { 0x04, 0x0000 }, + { 0x03, 0x00a1 }, + { 0x02, 0x0008 }, + { 0x01, 0x0120 }, + { 0x00, 0x1000 }, + { 0x04, 0x0800 }, + { 0x04, 0x9000 }, + { 0x03, 0x802f }, + { 0x02, 0x4f02 }, + { 0x01, 0x0409 }, + { 0x00, 0xf099 }, + { 0x04, 0x9800 }, + { 0x04, 0xa000 }, + { 0x03, 0xdf01 }, + { 0x02, 0xdf20 }, + { 0x01, 0xff95 }, + { 0x00, 0xba00 }, + { 0x04, 0xa800 }, + { 0x04, 0xf000 }, + { 0x03, 0xdf01 }, + { 0x02, 0xdf20 }, + { 0x01, 0x101a }, + { 0x00, 0xa0ff }, + { 0x04, 0xf800 }, + { 0x04, 0x0000 }, + { 0x1f, 0x0000 }, + + { 0x1f, 0x0001 }, + { 0x10, 0xf41b }, + { 0x14, 0xfb54 }, + { 0x18, 0xf5c7 }, + { 0x1f, 0x0000 }, + + { 0x1f, 0x0001 }, + { 0x17, 0x0cc0 }, + { 0x1f, 0x0000 } + }; + + rtl_writephy_batch(phydev, phy_reg_init); +} + +static void rtl8169sce_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + static const struct phy_reg phy_reg_init[] = { + { 0x1f, 0x0001 }, + { 0x04, 0x0000 }, + { 0x03, 0x00a1 }, + { 0x02, 0x0008 }, + { 0x01, 0x0120 }, + { 0x00, 0x1000 }, + { 0x04, 0x0800 }, + { 0x04, 0x9000 }, + { 0x03, 0x802f }, + { 0x02, 0x4f02 }, + { 0x01, 0x0409 }, + { 0x00, 0xf099 }, + { 0x04, 0x9800 }, + { 0x04, 0xa000 }, + { 0x03, 0xdf01 }, + { 0x02, 0xdf20 }, + { 0x01, 0xff95 }, + { 0x00, 0xba00 }, + { 0x04, 0xa800 }, + { 0x04, 0xf000 }, + { 0x03, 0xdf01 }, + { 0x02, 0xdf20 }, + { 0x01, 0x101a }, + { 0x00, 0xa0ff }, + { 0x04, 0xf800 }, + { 0x04, 0x0000 }, + { 0x1f, 0x0000 }, + + { 0x1f, 0x0001 }, + { 0x0b, 0x8480 }, + { 0x1f, 0x0000 }, + + { 0x1f, 0x0001 }, + { 0x18, 0x67c7 }, + { 0x04, 0x2000 }, + { 0x03, 0x002f }, + { 0x02, 0x4360 }, + { 0x01, 0x0109 }, + { 0x00, 0x3022 }, + { 0x04, 0x2800 }, + { 0x1f, 0x0000 }, + + { 0x1f, 0x0001 }, + { 0x17, 0x0cc0 }, + { 0x1f, 0x0000 } + }; + + rtl_writephy_batch(phydev, phy_reg_init); +} + +static void rtl8168bb_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + phy_write(phydev, 0x1f, 0x0001); + phy_set_bits(phydev, 0x16, BIT(0)); + phy_write(phydev, 0x10, 0xf41b); + phy_write(phydev, 0x1f, 0x0000); +} + +static void rtl8168bef_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + phy_write_paged(phydev, 0x0001, 0x10, 0xf41b); +} + +static void rtl8168cp_1_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + phy_write(phydev, 0x1d, 0x0f00); + phy_write_paged(phydev, 0x0002, 0x0c, 0x1ec8); +} + +static void rtl8168cp_2_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + phy_set_bits(phydev, 0x14, BIT(5)); + phy_set_bits(phydev, 0x0d, BIT(5)); + phy_write_paged(phydev, 0x0001, 0x1d, 0x3d98); +} + +static void rtl8168c_1_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + static const struct phy_reg phy_reg_init[] = { + { 0x1f, 0x0001 }, + { 0x12, 0x2300 }, + { 0x1f, 0x0002 }, + { 0x00, 0x88d4 }, + { 0x01, 0x82b1 }, + { 0x03, 0x7002 }, + { 0x08, 0x9e30 }, + { 0x09, 0x01f0 }, + { 0x0a, 0x5500 }, + { 0x0c, 0x00c8 }, + { 0x1f, 0x0003 }, + { 0x12, 0xc096 }, + { 0x16, 0x000a }, + { 0x1f, 0x0000 }, + { 0x1f, 0x0000 }, + { 0x09, 0x2000 }, + { 0x09, 0x0000 } + }; + + rtl_writephy_batch(phydev, phy_reg_init); + + phy_set_bits(phydev, 0x14, BIT(5)); + phy_set_bits(phydev, 0x0d, BIT(5)); +} + +static void rtl8168c_2_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + static const struct phy_reg phy_reg_init[] = { + { 0x1f, 0x0001 }, + { 0x12, 0x2300 }, + { 0x03, 0x802f }, + { 0x02, 0x4f02 }, + { 0x01, 0x0409 }, + { 0x00, 0xf099 }, + { 0x04, 0x9800 }, + { 0x04, 0x9000 }, + { 0x1d, 0x3d98 }, + { 0x1f, 0x0002 }, + { 0x0c, 0x7eb8 }, + { 0x06, 0x0761 }, + { 0x1f, 0x0003 }, + { 0x16, 0x0f0a }, + { 0x1f, 0x0000 } + }; + + rtl_writephy_batch(phydev, phy_reg_init); + + phy_set_bits(phydev, 0x16, BIT(0)); + phy_set_bits(phydev, 0x14, BIT(5)); + phy_set_bits(phydev, 0x0d, BIT(5)); +} + +static void rtl8168c_3_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + static const struct phy_reg phy_reg_init[] = { + { 0x1f, 0x0001 }, + { 0x12, 0x2300 }, + { 0x1d, 0x3d98 }, + { 0x1f, 0x0002 }, + { 0x0c, 0x7eb8 }, + { 0x06, 0x5461 }, + { 0x1f, 0x0003 }, + { 0x16, 0x0f0a }, + { 0x1f, 0x0000 } + }; + + rtl_writephy_batch(phydev, phy_reg_init); + + phy_set_bits(phydev, 0x16, BIT(0)); + phy_set_bits(phydev, 0x14, BIT(5)); + phy_set_bits(phydev, 0x0d, BIT(5)); +} + +static const struct phy_reg rtl8168d_1_phy_reg_init_0[] = { + /* Channel Estimation */ + { 0x1f, 0x0001 }, + { 0x06, 0x4064 }, + { 0x07, 0x2863 }, + { 0x08, 0x059c }, + { 0x09, 0x26b4 }, + { 0x0a, 0x6a19 }, + { 0x0b, 0xdcc8 }, + { 0x10, 0xf06d }, + { 0x14, 0x7f68 }, + { 0x18, 0x7fd9 }, + { 0x1c, 0xf0ff }, + { 0x1d, 0x3d9c }, + { 0x1f, 0x0003 }, + { 0x12, 0xf49f }, + { 0x13, 0x070b }, + { 0x1a, 0x05ad }, + { 0x14, 0x94c0 }, + + /* + * Tx Error Issue + * Enhance line driver power + */ + { 0x1f, 0x0002 }, + { 0x06, 0x5561 }, + { 0x1f, 0x0005 }, + { 0x05, 0x8332 }, + { 0x06, 0x5561 }, + + /* + * Can not link to 1Gbps with bad cable + * Decrease SNR threshold form 21.07dB to 19.04dB + */ + { 0x1f, 0x0001 }, + { 0x17, 0x0cc0 }, + + { 0x1f, 0x0000 }, + { 0x0d, 0xf880 } +}; + +static const struct phy_reg rtl8168d_1_phy_reg_init_1[] = { + { 0x1f, 0x0002 }, + { 0x05, 0x669a }, + { 0x1f, 0x0005 }, + { 0x05, 0x8330 }, + { 0x06, 0x669a }, + { 0x1f, 0x0002 } +}; + +static void rtl8168d_apply_firmware_cond(struct rtl8169_private *tp, + struct phy_device *phydev, + u16 val) +{ + u16 reg_val; + + phy_write(phydev, 0x1f, 0x0005); + phy_write(phydev, 0x05, 0x001b); + reg_val = phy_read(phydev, 0x06); + phy_write(phydev, 0x1f, 0x0000); + + if (reg_val != val) + phydev_warn(phydev, "chipset not ready for firmware\n"); + else + r8169_apply_firmware(tp); +} + +static void rtl8168d_1_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + rtl_writephy_batch(phydev, rtl8168d_1_phy_reg_init_0); + + /* + * Rx Error Issue + * Fine Tune Switching regulator parameter + */ + phy_write(phydev, 0x1f, 0x0002); + phy_modify(phydev, 0x0b, 0x00ef, 0x0010); + phy_modify(phydev, 0x0c, 0x5d00, 0xa200); + + if (rtl8168d_efuse_read(tp, 0x01) == 0xb1) { + int val; + + rtl_writephy_batch(phydev, rtl8168d_1_phy_reg_init_1); + + val = phy_read(phydev, 0x0d); + + if ((val & 0x00ff) != 0x006c) { + static const u32 set[] = { + 0x0065, 0x0066, 0x0067, 0x0068, + 0x0069, 0x006a, 0x006b, 0x006c + }; + int i; + + phy_write(phydev, 0x1f, 0x0002); + + val &= 0xff00; + for (i = 0; i < ARRAY_SIZE(set); i++) + phy_write(phydev, 0x0d, val | set[i]); + } + } else { + phy_write_paged(phydev, 0x0002, 0x05, 0x6662); + r8168d_phy_param(phydev, 0x8330, 0xffff, 0x6662); + } + + /* RSET couple improve */ + phy_write(phydev, 0x1f, 0x0002); + phy_set_bits(phydev, 0x0d, 0x0300); + phy_set_bits(phydev, 0x0f, 0x0010); + + /* Fine tune PLL performance */ + phy_write(phydev, 0x1f, 0x0002); + phy_modify(phydev, 0x02, 0x0600, 0x0100); + phy_clear_bits(phydev, 0x03, 0xe000); + phy_write(phydev, 0x1f, 0x0000); + + rtl8168d_apply_firmware_cond(tp, phydev, 0xbf00); +} + +static void rtl8168d_2_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + rtl_writephy_batch(phydev, rtl8168d_1_phy_reg_init_0); + + if (rtl8168d_efuse_read(tp, 0x01) == 0xb1) { + int val; + + rtl_writephy_batch(phydev, rtl8168d_1_phy_reg_init_1); + + val = phy_read(phydev, 0x0d); + if ((val & 0x00ff) != 0x006c) { + static const u32 set[] = { + 0x0065, 0x0066, 0x0067, 0x0068, + 0x0069, 0x006a, 0x006b, 0x006c + }; + int i; + + phy_write(phydev, 0x1f, 0x0002); + + val &= 0xff00; + for (i = 0; i < ARRAY_SIZE(set); i++) + phy_write(phydev, 0x0d, val | set[i]); + } + } else { + phy_write_paged(phydev, 0x0002, 0x05, 0x2642); + r8168d_phy_param(phydev, 0x8330, 0xffff, 0x2642); + } + + /* Fine tune PLL performance */ + phy_write(phydev, 0x1f, 0x0002); + phy_modify(phydev, 0x02, 0x0600, 0x0100); + phy_clear_bits(phydev, 0x03, 0xe000); + phy_write(phydev, 0x1f, 0x0000); + + /* Switching regulator Slew rate */ + phy_modify_paged(phydev, 0x0002, 0x0f, 0x0000, 0x0017); + + rtl8168d_apply_firmware_cond(tp, phydev, 0xb300); +} + +static void rtl8168d_3_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + static const struct phy_reg phy_reg_init[] = { + { 0x1f, 0x0002 }, + { 0x10, 0x0008 }, + { 0x0d, 0x006c }, + + { 0x1f, 0x0000 }, + { 0x0d, 0xf880 }, + + { 0x1f, 0x0001 }, + { 0x17, 0x0cc0 }, + + { 0x1f, 0x0001 }, + { 0x0b, 0xa4d8 }, + { 0x09, 0x281c }, + { 0x07, 0x2883 }, + { 0x0a, 0x6b35 }, + { 0x1d, 0x3da4 }, + { 0x1c, 0xeffd }, + { 0x14, 0x7f52 }, + { 0x18, 0x7fc6 }, + { 0x08, 0x0601 }, + { 0x06, 0x4063 }, + { 0x10, 0xf074 }, + { 0x1f, 0x0003 }, + { 0x13, 0x0789 }, + { 0x12, 0xf4bd }, + { 0x1a, 0x04fd }, + { 0x14, 0x84b0 }, + { 0x1f, 0x0000 }, + { 0x00, 0x9200 }, + + { 0x1f, 0x0005 }, + { 0x01, 0x0340 }, + { 0x1f, 0x0001 }, + { 0x04, 0x4000 }, + { 0x03, 0x1d21 }, + { 0x02, 0x0c32 }, + { 0x01, 0x0200 }, + { 0x00, 0x5554 }, + { 0x04, 0x4800 }, + { 0x04, 0x4000 }, + { 0x04, 0xf000 }, + { 0x03, 0xdf01 }, + { 0x02, 0xdf20 }, + { 0x01, 0x101a }, + { 0x00, 0xa0ff }, + { 0x04, 0xf800 }, + { 0x04, 0xf000 }, + { 0x1f, 0x0000 }, + }; + + rtl_writephy_batch(phydev, phy_reg_init); + r8168d_modify_extpage(phydev, 0x0023, 0x16, 0xffff, 0x0000); +} + +static void rtl8168d_4_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + phy_write_paged(phydev, 0x0001, 0x17, 0x0cc0); + r8168d_modify_extpage(phydev, 0x002d, 0x18, 0xffff, 0x0040); + phy_set_bits(phydev, 0x0d, BIT(5)); +} + +static void rtl8168e_1_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + static const struct phy_reg phy_reg_init[] = { + /* Channel estimation fine tune */ + { 0x1f, 0x0001 }, + { 0x0b, 0x6c20 }, + { 0x07, 0x2872 }, + { 0x1c, 0xefff }, + { 0x1f, 0x0003 }, + { 0x14, 0x6420 }, + { 0x1f, 0x0000 }, + }; + + r8169_apply_firmware(tp); + + /* Enable Delay cap */ + r8168d_phy_param(phydev, 0x8b80, 0xffff, 0xc896); + + rtl_writephy_batch(phydev, phy_reg_init); + + /* Update PFM & 10M TX idle timer */ + r8168d_modify_extpage(phydev, 0x002f, 0x15, 0xffff, 0x1919); + + r8168d_modify_extpage(phydev, 0x00ac, 0x18, 0xffff, 0x0006); + + /* DCO enable for 10M IDLE Power */ + r8168d_modify_extpage(phydev, 0x0023, 0x17, 0x0000, 0x0006); + + /* For impedance matching */ + phy_modify_paged(phydev, 0x0002, 0x08, 0x7f00, 0x8000); + + /* PHY auto speed down */ + r8168d_modify_extpage(phydev, 0x002d, 0x18, 0x0000, 0x0050); + phy_set_bits(phydev, 0x14, BIT(15)); + + r8168d_phy_param(phydev, 0x8b86, 0x0000, 0x0001); + r8168d_phy_param(phydev, 0x8b85, 0x2000, 0x0000); + + r8168d_modify_extpage(phydev, 0x0020, 0x15, 0x1100, 0x0000); + phy_write_paged(phydev, 0x0006, 0x00, 0x5a00); + + phy_write_mmd(phydev, MDIO_MMD_AN, MDIO_AN_EEE_ADV, 0x0000); +} + +static void rtl8168e_2_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + r8169_apply_firmware(tp); + + /* Enable Delay cap */ + r8168d_modify_extpage(phydev, 0x00ac, 0x18, 0xffff, 0x0006); + + /* Channel estimation fine tune */ + phy_write_paged(phydev, 0x0003, 0x09, 0xa20f); + + /* Green Setting */ + r8168d_phy_param(phydev, 0x8b5b, 0xffff, 0x9222); + r8168d_phy_param(phydev, 0x8b6d, 0xffff, 0x8000); + r8168d_phy_param(phydev, 0x8b76, 0xffff, 0x8000); + + /* For 4-corner performance improve */ + phy_write(phydev, 0x1f, 0x0005); + phy_write(phydev, 0x05, 0x8b80); + phy_set_bits(phydev, 0x17, 0x0006); + phy_write(phydev, 0x1f, 0x0000); + + /* PHY auto speed down */ + r8168d_modify_extpage(phydev, 0x002d, 0x18, 0x0000, 0x0010); + phy_set_bits(phydev, 0x14, BIT(15)); + + /* improve 10M EEE waveform */ + r8168d_phy_param(phydev, 0x8b86, 0x0000, 0x0001); + + /* Improve 2-pair detection performance */ + r8168d_phy_param(phydev, 0x8b85, 0x0000, 0x4000); + + rtl8168f_config_eee_phy(phydev); + + /* Green feature */ + phy_write(phydev, 0x1f, 0x0003); + phy_set_bits(phydev, 0x19, BIT(0)); + phy_set_bits(phydev, 0x10, BIT(10)); + phy_write(phydev, 0x1f, 0x0000); + phy_modify_paged(phydev, 0x0005, 0x01, 0, BIT(8)); +} + +static void rtl8168f_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + /* For 4-corner performance improve */ + r8168d_phy_param(phydev, 0x8b80, 0x0000, 0x0006); + + /* PHY auto speed down */ + r8168d_modify_extpage(phydev, 0x002d, 0x18, 0x0000, 0x0010); + phy_set_bits(phydev, 0x14, BIT(15)); + + /* Improve 10M EEE waveform */ + r8168d_phy_param(phydev, 0x8b86, 0x0000, 0x0001); + + rtl8168f_config_eee_phy(phydev); +} + +static void rtl8168f_1_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + r8169_apply_firmware(tp); + + /* Channel estimation fine tune */ + phy_write_paged(phydev, 0x0003, 0x09, 0xa20f); + + /* Modify green table for giga & fnet */ + r8168d_phy_param(phydev, 0x8b55, 0xffff, 0x0000); + r8168d_phy_param(phydev, 0x8b5e, 0xffff, 0x0000); + r8168d_phy_param(phydev, 0x8b67, 0xffff, 0x0000); + r8168d_phy_param(phydev, 0x8b70, 0xffff, 0x0000); + r8168d_modify_extpage(phydev, 0x0078, 0x17, 0xffff, 0x0000); + r8168d_modify_extpage(phydev, 0x0078, 0x19, 0xffff, 0x00fb); + + /* Modify green table for 10M */ + r8168d_phy_param(phydev, 0x8b79, 0xffff, 0xaa00); + + /* Disable hiimpedance detection (RTCT) */ + phy_write_paged(phydev, 0x0003, 0x01, 0x328a); + + rtl8168f_hw_phy_config(tp, phydev); + + /* Improve 2-pair detection performance */ + r8168d_phy_param(phydev, 0x8b85, 0x0000, 0x4000); +} + +static void rtl8168f_2_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + r8169_apply_firmware(tp); + + rtl8168f_hw_phy_config(tp, phydev); +} + +static void rtl8411_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + r8169_apply_firmware(tp); + + rtl8168f_hw_phy_config(tp, phydev); + + /* Improve 2-pair detection performance */ + r8168d_phy_param(phydev, 0x8b85, 0x0000, 0x4000); + + /* Channel estimation fine tune */ + phy_write_paged(phydev, 0x0003, 0x09, 0xa20f); + + /* Modify green table for giga & fnet */ + r8168d_phy_param(phydev, 0x8b55, 0xffff, 0x0000); + r8168d_phy_param(phydev, 0x8b5e, 0xffff, 0x0000); + r8168d_phy_param(phydev, 0x8b67, 0xffff, 0x0000); + r8168d_phy_param(phydev, 0x8b70, 0xffff, 0x0000); + r8168d_modify_extpage(phydev, 0x0078, 0x17, 0xffff, 0x0000); + r8168d_modify_extpage(phydev, 0x0078, 0x19, 0xffff, 0x00aa); + + /* Modify green table for 10M */ + r8168d_phy_param(phydev, 0x8b79, 0xffff, 0xaa00); + + /* Disable hiimpedance detection (RTCT) */ + phy_write_paged(phydev, 0x0003, 0x01, 0x328a); + + /* Modify green table for giga */ + r8168d_phy_param(phydev, 0x8b54, 0x0800, 0x0000); + r8168d_phy_param(phydev, 0x8b5d, 0x0800, 0x0000); + r8168d_phy_param(phydev, 0x8a7c, 0x0100, 0x0000); + r8168d_phy_param(phydev, 0x8a7f, 0x0000, 0x0100); + r8168d_phy_param(phydev, 0x8a82, 0x0100, 0x0000); + r8168d_phy_param(phydev, 0x8a85, 0x0100, 0x0000); + r8168d_phy_param(phydev, 0x8a88, 0x0100, 0x0000); + + /* uc same-seed solution */ + r8168d_phy_param(phydev, 0x8b85, 0x0000, 0x8000); + + /* Green feature */ + phy_write(phydev, 0x1f, 0x0003); + phy_clear_bits(phydev, 0x19, BIT(0)); + phy_clear_bits(phydev, 0x10, BIT(10)); + phy_write(phydev, 0x1f, 0x0000); +} + +static void rtl8168g_disable_aldps(struct phy_device *phydev) +{ + phy_modify_paged(phydev, 0x0a43, 0x10, BIT(2), 0); +} + +static void rtl8168g_phy_adjust_10m_aldps(struct phy_device *phydev) +{ + phy_modify_paged(phydev, 0x0bcc, 0x14, BIT(8), 0); + phy_modify_paged(phydev, 0x0a44, 0x11, 0, BIT(7) | BIT(6)); + r8168g_phy_param(phydev, 0x8084, 0x6000, 0x0000); + phy_modify_paged(phydev, 0x0a43, 0x10, 0x0000, 0x1003); +} + +static void rtl8168g_1_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + int ret; + + r8169_apply_firmware(tp); + + ret = phy_read_paged(phydev, 0x0a46, 0x10); + if (ret & BIT(8)) + phy_modify_paged(phydev, 0x0bcc, 0x12, BIT(15), 0); + else + phy_modify_paged(phydev, 0x0bcc, 0x12, 0, BIT(15)); + + ret = phy_read_paged(phydev, 0x0a46, 0x13); + if (ret & BIT(8)) + phy_modify_paged(phydev, 0x0c41, 0x15, 0, BIT(1)); + else + phy_modify_paged(phydev, 0x0c41, 0x15, BIT(1), 0); + + /* Enable PHY auto speed down */ + phy_modify_paged(phydev, 0x0a44, 0x11, 0, BIT(3) | BIT(2)); + + rtl8168g_phy_adjust_10m_aldps(phydev); + + /* EEE auto-fallback function */ + phy_modify_paged(phydev, 0x0a4b, 0x11, 0, BIT(2)); + + /* Enable UC LPF tune function */ + r8168g_phy_param(phydev, 0x8012, 0x0000, 0x8000); + + phy_modify_paged(phydev, 0x0c42, 0x11, BIT(13), BIT(14)); + + /* Improve SWR Efficiency */ + phy_write(phydev, 0x1f, 0x0bcd); + phy_write(phydev, 0x14, 0x5065); + phy_write(phydev, 0x14, 0xd065); + phy_write(phydev, 0x1f, 0x0bc8); + phy_write(phydev, 0x11, 0x5655); + phy_write(phydev, 0x1f, 0x0bcd); + phy_write(phydev, 0x14, 0x1065); + phy_write(phydev, 0x14, 0x9065); + phy_write(phydev, 0x14, 0x1065); + phy_write(phydev, 0x1f, 0x0000); + + rtl8168g_disable_aldps(phydev); + rtl8168g_config_eee_phy(phydev); +} + +static void rtl8168g_2_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + r8169_apply_firmware(tp); + rtl8168g_config_eee_phy(phydev); +} + +static void rtl8168h_1_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + u16 dout_tapbin; + u32 data; + + r8169_apply_firmware(tp); + + /* CHN EST parameters adjust - giga master */ + r8168g_phy_param(phydev, 0x809b, 0xf800, 0x8000); + r8168g_phy_param(phydev, 0x80a2, 0xff00, 0x8000); + r8168g_phy_param(phydev, 0x80a4, 0xff00, 0x8500); + r8168g_phy_param(phydev, 0x809c, 0xff00, 0xbd00); + + /* CHN EST parameters adjust - giga slave */ + r8168g_phy_param(phydev, 0x80ad, 0xf800, 0x7000); + r8168g_phy_param(phydev, 0x80b4, 0xff00, 0x5000); + r8168g_phy_param(phydev, 0x80ac, 0xff00, 0x4000); + + /* CHN EST parameters adjust - fnet */ + r8168g_phy_param(phydev, 0x808e, 0xff00, 0x1200); + r8168g_phy_param(phydev, 0x8090, 0xff00, 0xe500); + r8168g_phy_param(phydev, 0x8092, 0xff00, 0x9f00); + + /* enable R-tune & PGA-retune function */ + dout_tapbin = 0; + data = phy_read_paged(phydev, 0x0a46, 0x13); + data &= 3; + data <<= 2; + dout_tapbin |= data; + data = phy_read_paged(phydev, 0x0a46, 0x12); + data &= 0xc000; + data >>= 14; + dout_tapbin |= data; + dout_tapbin = ~(dout_tapbin ^ 0x08); + dout_tapbin <<= 12; + dout_tapbin &= 0xf000; + + r8168g_phy_param(phydev, 0x827a, 0xf000, dout_tapbin); + r8168g_phy_param(phydev, 0x827b, 0xf000, dout_tapbin); + r8168g_phy_param(phydev, 0x827c, 0xf000, dout_tapbin); + r8168g_phy_param(phydev, 0x827d, 0xf000, dout_tapbin); + r8168g_phy_param(phydev, 0x0811, 0x0000, 0x0800); + phy_modify_paged(phydev, 0x0a42, 0x16, 0x0000, 0x0002); + + /* enable GPHY 10M */ + phy_modify_paged(phydev, 0x0a44, 0x11, 0, BIT(11)); + + /* SAR ADC performance */ + phy_modify_paged(phydev, 0x0bca, 0x17, BIT(12) | BIT(13), BIT(14)); + + r8168g_phy_param(phydev, 0x803f, 0x3000, 0x0000); + r8168g_phy_param(phydev, 0x8047, 0x3000, 0x0000); + r8168g_phy_param(phydev, 0x804f, 0x3000, 0x0000); + r8168g_phy_param(phydev, 0x8057, 0x3000, 0x0000); + r8168g_phy_param(phydev, 0x805f, 0x3000, 0x0000); + r8168g_phy_param(phydev, 0x8067, 0x3000, 0x0000); + r8168g_phy_param(phydev, 0x806f, 0x3000, 0x0000); + + /* disable phy pfm mode */ + phy_modify_paged(phydev, 0x0a44, 0x11, BIT(7), 0); + + rtl8168g_disable_aldps(phydev); + rtl8168h_config_eee_phy(phydev); +} + +static void rtl8168h_2_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + u16 ioffset, rlen; + u32 data; + + r8169_apply_firmware(tp); + + /* CHIN EST parameter update */ + r8168g_phy_param(phydev, 0x808a, 0x003f, 0x000a); + + /* enable R-tune & PGA-retune function */ + r8168g_phy_param(phydev, 0x0811, 0x0000, 0x0800); + phy_modify_paged(phydev, 0x0a42, 0x16, 0x0000, 0x0002); + + /* enable GPHY 10M */ + phy_modify_paged(phydev, 0x0a44, 0x11, 0, BIT(11)); + + ioffset = rtl8168h_2_get_adc_bias_ioffset(tp); + if (ioffset != 0xffff) + phy_write_paged(phydev, 0x0bcf, 0x16, ioffset); + + /* Modify rlen (TX LPF corner frequency) level */ + data = phy_read_paged(phydev, 0x0bcd, 0x16); + data &= 0x000f; + rlen = 0; + if (data > 3) + rlen = data - 3; + data = rlen | (rlen << 4) | (rlen << 8) | (rlen << 12); + phy_write_paged(phydev, 0x0bcd, 0x17, data); + + /* disable phy pfm mode */ + phy_modify_paged(phydev, 0x0a44, 0x11, BIT(7), 0); + + rtl8168g_disable_aldps(phydev); + rtl8168g_config_eee_phy(phydev); +} + +static void rtl8168ep_1_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + /* Enable PHY auto speed down */ + phy_modify_paged(phydev, 0x0a44, 0x11, 0, BIT(3) | BIT(2)); + + rtl8168g_phy_adjust_10m_aldps(phydev); + + /* Enable EEE auto-fallback function */ + phy_modify_paged(phydev, 0x0a4b, 0x11, 0, BIT(2)); + + /* Enable UC LPF tune function */ + r8168g_phy_param(phydev, 0x8012, 0x0000, 0x8000); + + /* set rg_sel_sdm_rate */ + phy_modify_paged(phydev, 0x0c42, 0x11, BIT(13), BIT(14)); + + rtl8168g_disable_aldps(phydev); + rtl8168g_config_eee_phy(phydev); +} + +static void rtl8168ep_2_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + rtl8168g_phy_adjust_10m_aldps(phydev); + + /* Enable UC LPF tune function */ + r8168g_phy_param(phydev, 0x8012, 0x0000, 0x8000); + + /* Set rg_sel_sdm_rate */ + phy_modify_paged(phydev, 0x0c42, 0x11, BIT(13), BIT(14)); + + /* Channel estimation parameters */ + r8168g_phy_param(phydev, 0x80f3, 0xff00, 0x8b00); + r8168g_phy_param(phydev, 0x80f0, 0xff00, 0x3a00); + r8168g_phy_param(phydev, 0x80ef, 0xff00, 0x0500); + r8168g_phy_param(phydev, 0x80f6, 0xff00, 0x6e00); + r8168g_phy_param(phydev, 0x80ec, 0xff00, 0x6800); + r8168g_phy_param(phydev, 0x80ed, 0xff00, 0x7c00); + r8168g_phy_param(phydev, 0x80f2, 0xff00, 0xf400); + r8168g_phy_param(phydev, 0x80f4, 0xff00, 0x8500); + r8168g_phy_param(phydev, 0x8110, 0xff00, 0xa800); + r8168g_phy_param(phydev, 0x810f, 0xff00, 0x1d00); + r8168g_phy_param(phydev, 0x8111, 0xff00, 0xf500); + r8168g_phy_param(phydev, 0x8113, 0xff00, 0x6100); + r8168g_phy_param(phydev, 0x8115, 0xff00, 0x9200); + r8168g_phy_param(phydev, 0x810e, 0xff00, 0x0400); + r8168g_phy_param(phydev, 0x810c, 0xff00, 0x7c00); + r8168g_phy_param(phydev, 0x810b, 0xff00, 0x5a00); + r8168g_phy_param(phydev, 0x80d1, 0xff00, 0xff00); + r8168g_phy_param(phydev, 0x80cd, 0xff00, 0x9e00); + r8168g_phy_param(phydev, 0x80d3, 0xff00, 0x0e00); + r8168g_phy_param(phydev, 0x80d5, 0xff00, 0xca00); + r8168g_phy_param(phydev, 0x80d7, 0xff00, 0x8400); + + /* Force PWM-mode */ + phy_write(phydev, 0x1f, 0x0bcd); + phy_write(phydev, 0x14, 0x5065); + phy_write(phydev, 0x14, 0xd065); + phy_write(phydev, 0x1f, 0x0bc8); + phy_write(phydev, 0x12, 0x00ed); + phy_write(phydev, 0x1f, 0x0bcd); + phy_write(phydev, 0x14, 0x1065); + phy_write(phydev, 0x14, 0x9065); + phy_write(phydev, 0x14, 0x1065); + phy_write(phydev, 0x1f, 0x0000); + + rtl8168g_disable_aldps(phydev); + rtl8168g_config_eee_phy(phydev); +} + +static void rtl8117_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + /* CHN EST parameters adjust - fnet */ + r8168g_phy_param(phydev, 0x808e, 0xff00, 0x4800); + r8168g_phy_param(phydev, 0x8090, 0xff00, 0xcc00); + r8168g_phy_param(phydev, 0x8092, 0xff00, 0xb000); + + r8168g_phy_param(phydev, 0x8088, 0xff00, 0x6000); + r8168g_phy_param(phydev, 0x808b, 0x3f00, 0x0b00); + r8168g_phy_param(phydev, 0x808d, 0x1f00, 0x0600); + r8168g_phy_param(phydev, 0x808c, 0xff00, 0xb000); + r8168g_phy_param(phydev, 0x80a0, 0xff00, 0x2800); + r8168g_phy_param(phydev, 0x80a2, 0xff00, 0x5000); + r8168g_phy_param(phydev, 0x809b, 0xf800, 0xb000); + r8168g_phy_param(phydev, 0x809a, 0xff00, 0x4b00); + r8168g_phy_param(phydev, 0x809d, 0x3f00, 0x0800); + r8168g_phy_param(phydev, 0x80a1, 0xff00, 0x7000); + r8168g_phy_param(phydev, 0x809f, 0x1f00, 0x0300); + r8168g_phy_param(phydev, 0x809e, 0xff00, 0x8800); + r8168g_phy_param(phydev, 0x80b2, 0xff00, 0x2200); + r8168g_phy_param(phydev, 0x80ad, 0xf800, 0x9800); + r8168g_phy_param(phydev, 0x80af, 0x3f00, 0x0800); + r8168g_phy_param(phydev, 0x80b3, 0xff00, 0x6f00); + r8168g_phy_param(phydev, 0x80b1, 0x1f00, 0x0300); + r8168g_phy_param(phydev, 0x80b0, 0xff00, 0x9300); + + r8168g_phy_param(phydev, 0x8011, 0x0000, 0x0800); + + /* enable GPHY 10M */ + phy_modify_paged(phydev, 0x0a44, 0x11, 0, BIT(11)); + + r8168g_phy_param(phydev, 0x8016, 0x0000, 0x0400); + + rtl8168g_disable_aldps(phydev); + rtl8168h_config_eee_phy(phydev); +} + +static void rtl8102e_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + static const struct phy_reg phy_reg_init[] = { + { 0x1f, 0x0003 }, + { 0x08, 0x441d }, + { 0x01, 0x9100 }, + { 0x1f, 0x0000 } + }; + + phy_set_bits(phydev, 0x11, BIT(12)); + phy_set_bits(phydev, 0x19, BIT(13)); + phy_set_bits(phydev, 0x10, BIT(15)); + + rtl_writephy_batch(phydev, phy_reg_init); +} + +static void rtl8105e_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + /* Disable ALDPS before ram code */ + phy_write(phydev, 0x18, 0x0310); + msleep(100); + + r8169_apply_firmware(tp); + + phy_write_paged(phydev, 0x0005, 0x1a, 0x0000); + phy_write_paged(phydev, 0x0004, 0x1c, 0x0000); + phy_write_paged(phydev, 0x0001, 0x15, 0x7701); +} + +static void rtl8402_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + /* Disable ALDPS before setting firmware */ + phy_write(phydev, 0x18, 0x0310); + msleep(20); + + r8169_apply_firmware(tp); + + /* EEE setting */ + phy_write(phydev, 0x1f, 0x0004); + phy_write(phydev, 0x10, 0x401f); + phy_write(phydev, 0x19, 0x7030); + phy_write(phydev, 0x1f, 0x0000); +} + +static void rtl8106e_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + static const struct phy_reg phy_reg_init[] = { + { 0x1f, 0x0004 }, + { 0x10, 0xc07f }, + { 0x19, 0x7030 }, + { 0x1f, 0x0000 } + }; + + /* Disable ALDPS before ram code */ + phy_write(phydev, 0x18, 0x0310); + msleep(100); + + r8169_apply_firmware(tp); + + rtl_writephy_batch(phydev, phy_reg_init); +} + +static void rtl8125_1_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + phy_modify_paged(phydev, 0xad4, 0x10, 0x03ff, 0x0084); + phy_modify_paged(phydev, 0xad4, 0x17, 0x0000, 0x0010); + phy_modify_paged(phydev, 0xad1, 0x13, 0x03ff, 0x0006); + phy_modify_paged(phydev, 0xad3, 0x11, 0x003f, 0x0006); + phy_modify_paged(phydev, 0xac0, 0x14, 0x0000, 0x1100); + phy_modify_paged(phydev, 0xac8, 0x15, 0xf000, 0x7000); + phy_modify_paged(phydev, 0xad1, 0x14, 0x0000, 0x0400); + phy_modify_paged(phydev, 0xad1, 0x15, 0x0000, 0x03ff); + phy_modify_paged(phydev, 0xad1, 0x16, 0x0000, 0x03ff); + + r8168g_phy_param(phydev, 0x80ea, 0xff00, 0xc400); + r8168g_phy_param(phydev, 0x80eb, 0x0700, 0x0300); + r8168g_phy_param(phydev, 0x80f8, 0xff00, 0x1c00); + r8168g_phy_param(phydev, 0x80f1, 0xff00, 0x3000); + r8168g_phy_param(phydev, 0x80fe, 0xff00, 0xa500); + r8168g_phy_param(phydev, 0x8102, 0xff00, 0x5000); + r8168g_phy_param(phydev, 0x8105, 0xff00, 0x3300); + r8168g_phy_param(phydev, 0x8100, 0xff00, 0x7000); + r8168g_phy_param(phydev, 0x8104, 0xff00, 0xf000); + r8168g_phy_param(phydev, 0x8106, 0xff00, 0x6500); + r8168g_phy_param(phydev, 0x80dc, 0xff00, 0xed00); + r8168g_phy_param(phydev, 0x80df, 0x0000, 0x0100); + r8168g_phy_param(phydev, 0x80e1, 0x0100, 0x0000); + + phy_modify_paged(phydev, 0xbf0, 0x13, 0x003f, 0x0038); + r8168g_phy_param(phydev, 0x819f, 0xffff, 0xd0b6); + + phy_write_paged(phydev, 0xbc3, 0x12, 0x5555); + phy_modify_paged(phydev, 0xbf0, 0x15, 0x0e00, 0x0a00); + phy_modify_paged(phydev, 0xa5c, 0x10, 0x0400, 0x0000); + phy_modify_paged(phydev, 0xa44, 0x11, 0x0000, 0x0800); + + rtl8125_config_eee_phy(phydev); +} + +static void rtl8125_2_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + int i; + + phy_modify_paged(phydev, 0xad4, 0x17, 0x0000, 0x0010); + phy_modify_paged(phydev, 0xad1, 0x13, 0x03ff, 0x03ff); + phy_modify_paged(phydev, 0xad3, 0x11, 0x003f, 0x0006); + phy_modify_paged(phydev, 0xac0, 0x14, 0x1100, 0x0000); + phy_modify_paged(phydev, 0xacc, 0x10, 0x0003, 0x0002); + phy_modify_paged(phydev, 0xad4, 0x10, 0x00e7, 0x0044); + phy_modify_paged(phydev, 0xac1, 0x12, 0x0080, 0x0000); + phy_modify_paged(phydev, 0xac8, 0x10, 0x0300, 0x0000); + phy_modify_paged(phydev, 0xac5, 0x17, 0x0007, 0x0002); + phy_write_paged(phydev, 0xad4, 0x16, 0x00a8); + phy_write_paged(phydev, 0xac5, 0x16, 0x01ff); + phy_modify_paged(phydev, 0xac8, 0x15, 0x00f0, 0x0030); + + phy_write(phydev, 0x1f, 0x0b87); + phy_write(phydev, 0x16, 0x80a2); + phy_write(phydev, 0x17, 0x0153); + phy_write(phydev, 0x16, 0x809c); + phy_write(phydev, 0x17, 0x0153); + phy_write(phydev, 0x1f, 0x0000); + + phy_write(phydev, 0x1f, 0x0a43); + phy_write(phydev, 0x13, 0x81B3); + phy_write(phydev, 0x14, 0x0043); + phy_write(phydev, 0x14, 0x00A7); + phy_write(phydev, 0x14, 0x00D6); + phy_write(phydev, 0x14, 0x00EC); + phy_write(phydev, 0x14, 0x00F6); + phy_write(phydev, 0x14, 0x00FB); + phy_write(phydev, 0x14, 0x00FD); + phy_write(phydev, 0x14, 0x00FF); + phy_write(phydev, 0x14, 0x00BB); + phy_write(phydev, 0x14, 0x0058); + phy_write(phydev, 0x14, 0x0029); + phy_write(phydev, 0x14, 0x0013); + phy_write(phydev, 0x14, 0x0009); + phy_write(phydev, 0x14, 0x0004); + phy_write(phydev, 0x14, 0x0002); + for (i = 0; i < 25; i++) + phy_write(phydev, 0x14, 0x0000); + phy_write(phydev, 0x1f, 0x0000); + + r8168g_phy_param(phydev, 0x8257, 0xffff, 0x020F); + r8168g_phy_param(phydev, 0x80ea, 0xffff, 0x7843); + + r8169_apply_firmware(tp); + + phy_modify_paged(phydev, 0xd06, 0x14, 0x0000, 0x2000); + + r8168g_phy_param(phydev, 0x81a2, 0x0000, 0x0100); + + phy_modify_paged(phydev, 0xb54, 0x16, 0xff00, 0xdb00); + phy_modify_paged(phydev, 0xa45, 0x12, 0x0001, 0x0000); + phy_modify_paged(phydev, 0xa5d, 0x12, 0x0000, 0x0020); + phy_modify_paged(phydev, 0xad4, 0x17, 0x0010, 0x0000); + phy_modify_paged(phydev, 0xa86, 0x15, 0x0001, 0x0000); + phy_modify_paged(phydev, 0xa44, 0x11, 0x0000, 0x0800); + + rtl8125_config_eee_phy(phydev); +} + +void r8169_hw_phy_config(struct rtl8169_private *tp, struct phy_device *phydev, + enum mac_version ver) +{ + static const rtl_phy_cfg_fct phy_configs[] = { + /* PCI devices. */ + [RTL_GIGA_MAC_VER_02] = rtl8169s_hw_phy_config, + [RTL_GIGA_MAC_VER_03] = rtl8169s_hw_phy_config, + [RTL_GIGA_MAC_VER_04] = rtl8169sb_hw_phy_config, + [RTL_GIGA_MAC_VER_05] = rtl8169scd_hw_phy_config, + [RTL_GIGA_MAC_VER_06] = rtl8169sce_hw_phy_config, + /* PCI-E devices. */ + [RTL_GIGA_MAC_VER_07] = rtl8102e_hw_phy_config, + [RTL_GIGA_MAC_VER_08] = rtl8102e_hw_phy_config, + [RTL_GIGA_MAC_VER_09] = rtl8102e_hw_phy_config, + [RTL_GIGA_MAC_VER_10] = NULL, + [RTL_GIGA_MAC_VER_11] = rtl8168bb_hw_phy_config, + [RTL_GIGA_MAC_VER_12] = rtl8168bef_hw_phy_config, + [RTL_GIGA_MAC_VER_13] = NULL, + [RTL_GIGA_MAC_VER_14] = NULL, + [RTL_GIGA_MAC_VER_15] = NULL, + [RTL_GIGA_MAC_VER_16] = NULL, + [RTL_GIGA_MAC_VER_17] = rtl8168bef_hw_phy_config, + [RTL_GIGA_MAC_VER_18] = rtl8168cp_1_hw_phy_config, + [RTL_GIGA_MAC_VER_19] = rtl8168c_1_hw_phy_config, + [RTL_GIGA_MAC_VER_20] = rtl8168c_2_hw_phy_config, + [RTL_GIGA_MAC_VER_21] = rtl8168c_3_hw_phy_config, + [RTL_GIGA_MAC_VER_22] = rtl8168c_3_hw_phy_config, + [RTL_GIGA_MAC_VER_23] = rtl8168cp_2_hw_phy_config, + [RTL_GIGA_MAC_VER_24] = rtl8168cp_2_hw_phy_config, + [RTL_GIGA_MAC_VER_25] = rtl8168d_1_hw_phy_config, + [RTL_GIGA_MAC_VER_26] = rtl8168d_2_hw_phy_config, + [RTL_GIGA_MAC_VER_27] = rtl8168d_3_hw_phy_config, + [RTL_GIGA_MAC_VER_28] = rtl8168d_4_hw_phy_config, + [RTL_GIGA_MAC_VER_29] = rtl8105e_hw_phy_config, + [RTL_GIGA_MAC_VER_30] = rtl8105e_hw_phy_config, + [RTL_GIGA_MAC_VER_31] = NULL, + [RTL_GIGA_MAC_VER_32] = rtl8168e_1_hw_phy_config, + [RTL_GIGA_MAC_VER_33] = rtl8168e_1_hw_phy_config, + [RTL_GIGA_MAC_VER_34] = rtl8168e_2_hw_phy_config, + [RTL_GIGA_MAC_VER_35] = rtl8168f_1_hw_phy_config, + [RTL_GIGA_MAC_VER_36] = rtl8168f_2_hw_phy_config, + [RTL_GIGA_MAC_VER_37] = rtl8402_hw_phy_config, + [RTL_GIGA_MAC_VER_38] = rtl8411_hw_phy_config, + [RTL_GIGA_MAC_VER_39] = rtl8106e_hw_phy_config, + [RTL_GIGA_MAC_VER_40] = rtl8168g_1_hw_phy_config, + [RTL_GIGA_MAC_VER_41] = NULL, + [RTL_GIGA_MAC_VER_42] = rtl8168g_2_hw_phy_config, + [RTL_GIGA_MAC_VER_43] = rtl8168g_2_hw_phy_config, + [RTL_GIGA_MAC_VER_44] = rtl8168g_2_hw_phy_config, + [RTL_GIGA_MAC_VER_45] = rtl8168h_1_hw_phy_config, + [RTL_GIGA_MAC_VER_46] = rtl8168h_2_hw_phy_config, + [RTL_GIGA_MAC_VER_47] = rtl8168h_1_hw_phy_config, + [RTL_GIGA_MAC_VER_48] = rtl8168h_2_hw_phy_config, + [RTL_GIGA_MAC_VER_49] = rtl8168ep_1_hw_phy_config, + [RTL_GIGA_MAC_VER_50] = rtl8168ep_2_hw_phy_config, + [RTL_GIGA_MAC_VER_51] = rtl8168ep_2_hw_phy_config, + [RTL_GIGA_MAC_VER_52] = rtl8117_hw_phy_config, + [RTL_GIGA_MAC_VER_60] = rtl8125_1_hw_phy_config, + [RTL_GIGA_MAC_VER_61] = rtl8125_2_hw_phy_config, + }; + + if (phy_configs[ver]) + phy_configs[ver](tp, phydev); +} diff --git a/drivers/net/ethernet/renesas/sh_eth.c b/drivers/net/ethernet/renesas/sh_eth.c index cdd8ab2eb910..c922d7a553f2 100644 --- a/drivers/net/ethernet/renesas/sh_eth.c +++ b/drivers/net/ethernet/renesas/sh_eth.c @@ -2204,24 +2204,28 @@ static size_t __sh_eth_get_regs(struct net_device *ndev, u32 *buf) if (cd->tsu) { add_tsu_reg(ARSTR); add_tsu_reg(TSU_CTRST); - add_tsu_reg(TSU_FWEN0); - add_tsu_reg(TSU_FWEN1); - add_tsu_reg(TSU_FCM); - add_tsu_reg(TSU_BSYSL0); - add_tsu_reg(TSU_BSYSL1); - add_tsu_reg(TSU_PRISL0); - add_tsu_reg(TSU_PRISL1); - add_tsu_reg(TSU_FWSL0); - add_tsu_reg(TSU_FWSL1); + if (cd->dual_port) { + add_tsu_reg(TSU_FWEN0); + add_tsu_reg(TSU_FWEN1); + add_tsu_reg(TSU_FCM); + add_tsu_reg(TSU_BSYSL0); + add_tsu_reg(TSU_BSYSL1); + add_tsu_reg(TSU_PRISL0); + add_tsu_reg(TSU_PRISL1); + add_tsu_reg(TSU_FWSL0); + add_tsu_reg(TSU_FWSL1); + } add_tsu_reg(TSU_FWSLC); - add_tsu_reg(TSU_QTAGM0); - add_tsu_reg(TSU_QTAGM1); - add_tsu_reg(TSU_FWSR); - add_tsu_reg(TSU_FWINMK); - add_tsu_reg(TSU_ADQT0); - add_tsu_reg(TSU_ADQT1); - add_tsu_reg(TSU_VTAG0); - add_tsu_reg(TSU_VTAG1); + if (cd->dual_port) { + add_tsu_reg(TSU_QTAGM0); + add_tsu_reg(TSU_QTAGM1); + add_tsu_reg(TSU_FWSR); + add_tsu_reg(TSU_FWINMK); + add_tsu_reg(TSU_ADQT0); + add_tsu_reg(TSU_ADQT1); + add_tsu_reg(TSU_VTAG0); + add_tsu_reg(TSU_VTAG1); + } add_tsu_reg(TSU_ADSBSY); add_tsu_reg(TSU_TEN); add_tsu_reg(TSU_POST1); diff --git a/drivers/net/ethernet/samsung/sxgbe/sxgbe_main.c b/drivers/net/ethernet/samsung/sxgbe/sxgbe_main.c index 7d3a1c0df09c..c705743d69f7 100644 --- a/drivers/net/ethernet/samsung/sxgbe/sxgbe_main.c +++ b/drivers/net/ethernet/samsung/sxgbe/sxgbe_main.c @@ -1939,9 +1939,7 @@ static int sxgbe_ioctl(struct net_device *dev, struct ifreq *rq, int cmd) case SIOCGMIIPHY: case SIOCGMIIREG: case SIOCSMIIREG: - if (!dev->phydev) - return -EINVAL; - ret = phy_mii_ioctl(dev->phydev, rq, cmd); + ret = phy_do_ioctl(dev, rq, cmd); break; default: break; diff --git a/drivers/net/ethernet/sfc/Makefile b/drivers/net/ethernet/sfc/Makefile index c5c297e78d06..890fd65caa2d 100644 --- a/drivers/net/ethernet/sfc/Makefile +++ b/drivers/net/ethernet/sfc/Makefile @@ -1,7 +1,10 @@ # SPDX-License-Identifier: GPL-2.0 -sfc-y += efx.o nic.o farch.o siena.o ef10.o tx.o rx.o \ - selftest.o ethtool.o ptp.o tx_tso.o \ - mcdi.o mcdi_port.o mcdi_mon.o +sfc-y += efx.o efx_common.o efx_channels.o nic.o \ + farch.o siena.o ef10.o \ + tx.o tx_common.o tx_tso.o rx.o rx_common.o \ + selftest.o ethtool.o ethtool_common.o ptp.o \ + mcdi.o mcdi_port.o mcdi_port_common.o \ + mcdi_functions.o mcdi_mon.o sfc-$(CONFIG_SFC_MTD) += mtd.o sfc-$(CONFIG_SFC_SRIOV) += sriov.o siena_sriov.o ef10_sriov.o diff --git a/drivers/net/ethernet/sfc/ef10.c b/drivers/net/ethernet/sfc/ef10.c index 4d9bbccc6f89..4997f61de3d6 100644 --- a/drivers/net/ethernet/sfc/ef10.c +++ b/drivers/net/ethernet/sfc/ef10.c @@ -5,10 +5,13 @@ */ #include "net_driver.h" +#include "rx_common.h" #include "ef10_regs.h" #include "io.h" #include "mcdi.h" #include "mcdi_pcol.h" +#include "mcdi_port_common.h" +#include "mcdi_functions.h" #include "nic.h" #include "workarounds.h" #include "selftest.h" @@ -185,24 +188,6 @@ static bool efx_ef10_is_vf(struct efx_nic *efx) return efx->type->is_vf; } -static int efx_ef10_get_pf_index(struct efx_nic *efx) -{ - MCDI_DECLARE_BUF(outbuf, MC_CMD_GET_FUNCTION_INFO_OUT_LEN); - struct efx_ef10_nic_data *nic_data = efx->nic_data; - size_t outlen; - int rc; - - rc = efx_mcdi_rpc(efx, MC_CMD_GET_FUNCTION_INFO, NULL, 0, outbuf, - sizeof(outbuf), &outlen); - if (rc) - return rc; - if (outlen < sizeof(outbuf)) - return -EIO; - - nic_data->pf_index = MCDI_DWORD(outbuf, GET_FUNCTION_INFO_OUT_PF); - return 0; -} - #ifdef CONFIG_SFC_SRIOV static int efx_ef10_get_vf_index(struct efx_nic *efx) { @@ -273,24 +258,9 @@ static int efx_ef10_init_datapath_caps(struct efx_nic *efx) u8 vi_window_mode = MCDI_BYTE(outbuf, GET_CAPABILITIES_V3_OUT_VI_WINDOW_MODE); - switch (vi_window_mode) { - case MC_CMD_GET_CAPABILITIES_V3_OUT_VI_WINDOW_MODE_8K: - efx->vi_stride = 8192; - break; - case MC_CMD_GET_CAPABILITIES_V3_OUT_VI_WINDOW_MODE_16K: - efx->vi_stride = 16384; - break; - case MC_CMD_GET_CAPABILITIES_V3_OUT_VI_WINDOW_MODE_64K: - efx->vi_stride = 65536; - break; - default: - netif_err(efx, probe, efx->net_dev, - "Unrecognised VI window mode %d\n", - vi_window_mode); - return -EIO; - } - netif_dbg(efx, probe, efx->net_dev, "vi_stride = %u\n", - efx->vi_stride); + rc = efx_mcdi_window_mode_to_stride(efx, vi_window_mode); + if (rc) + return rc; } else { /* keep default VI stride */ netif_dbg(efx, probe, efx->net_dev, @@ -689,7 +659,7 @@ static int efx_ef10_probe(struct efx_nic *efx) } nic_data->warm_boot_count = rc; - efx->rss_context.context_id = EFX_EF10_RSS_CONTEXT_INVALID; + efx->rss_context.context_id = EFX_MCDI_RSS_CONTEXT_INVALID; nic_data->vport_id = EVB_PORT_ID_ASSIGNED; @@ -725,7 +695,7 @@ static int efx_ef10_probe(struct efx_nic *efx) if (rc) goto fail4; - rc = efx_ef10_get_pf_index(efx); + rc = efx_get_pf_index(efx, &nic_data->pf_index); if (rc) goto fail5; @@ -831,22 +801,6 @@ fail1: return rc; } -static int efx_ef10_free_vis(struct efx_nic *efx) -{ - MCDI_DECLARE_BUF_ERR(outbuf); - size_t outlen; - int rc = efx_mcdi_rpc_quiet(efx, MC_CMD_FREE_VIS, NULL, 0, - outbuf, sizeof(outbuf), &outlen); - - /* -EALREADY means nothing to free, so ignore */ - if (rc == -EALREADY) - rc = 0; - if (rc) - efx_mcdi_display_error(efx, MC_CMD_FREE_VIS, 0, outbuf, outlen, - rc); - return rc; -} - #ifdef EFX_USE_PIO static void efx_ef10_free_piobufs(struct efx_nic *efx) @@ -1089,7 +1043,7 @@ static void efx_ef10_remove(struct efx_nic *efx) if (nic_data->wc_membase) iounmap(nic_data->wc_membase); - rc = efx_ef10_free_vis(efx); + rc = efx_mcdi_free_vis(efx); WARN_ON(rc != 0); if (!nic_data->must_restore_piobufs) @@ -1260,28 +1214,10 @@ static int efx_ef10_probe_vf(struct efx_nic *efx __attribute__ ((unused))) static int efx_ef10_alloc_vis(struct efx_nic *efx, unsigned int min_vis, unsigned int max_vis) { - MCDI_DECLARE_BUF(inbuf, MC_CMD_ALLOC_VIS_IN_LEN); - MCDI_DECLARE_BUF(outbuf, MC_CMD_ALLOC_VIS_OUT_LEN); struct efx_ef10_nic_data *nic_data = efx->nic_data; - size_t outlen; - int rc; - - MCDI_SET_DWORD(inbuf, ALLOC_VIS_IN_MIN_VI_COUNT, min_vis); - MCDI_SET_DWORD(inbuf, ALLOC_VIS_IN_MAX_VI_COUNT, max_vis); - rc = efx_mcdi_rpc(efx, MC_CMD_ALLOC_VIS, inbuf, sizeof(inbuf), - outbuf, sizeof(outbuf), &outlen); - if (rc != 0) - return rc; - - if (outlen < MC_CMD_ALLOC_VIS_OUT_LEN) - return -EIO; - netif_dbg(efx, drv, efx->net_dev, "base VI is A0x%03x\n", - MCDI_DWORD(outbuf, ALLOC_VIS_OUT_VI_BASE)); - - nic_data->vi_base = MCDI_DWORD(outbuf, ALLOC_VIS_OUT_VI_BASE); - nic_data->n_allocated_vis = MCDI_DWORD(outbuf, ALLOC_VIS_OUT_VI_COUNT); - return 0; + return efx_mcdi_alloc_vis(efx, min_vis, max_vis, &nic_data->vi_base, + &nic_data->n_allocated_vis); } /* Note that the failure path of this function does not free @@ -1363,7 +1299,7 @@ static int efx_ef10_dimension_resources(struct efx_nic *efx) } /* In case the last attached driver failed to free VIs, do it now */ - rc = efx_ef10_free_vis(efx); + rc = efx_mcdi_free_vis(efx); if (rc != 0) return rc; @@ -1384,7 +1320,7 @@ static int efx_ef10_dimension_resources(struct efx_nic *efx) efx->max_tx_channels = nic_data->n_allocated_vis / EFX_TXQ_TYPES; - efx_ef10_free_vis(efx); + efx_mcdi_free_vis(efx); return -EAGAIN; } @@ -1503,7 +1439,7 @@ static void efx_ef10_reset_mc_allocations(struct efx_nic *efx) nic_data->must_restore_filters = true; nic_data->must_restore_piobufs = true; efx_ef10_forget_old_piobufs(efx); - efx->rss_context.context_id = EFX_EF10_RSS_CONTEXT_INVALID; + efx->rss_context.context_id = EFX_MCDI_RSS_CONTEXT_INVALID; /* Driver-created vswitches and vports must be re-created */ nic_data->must_probe_vswitching = true; @@ -2408,20 +2344,15 @@ static u32 efx_ef10_tso_versions(struct efx_nic *efx) static void efx_ef10_tx_init(struct efx_tx_queue *tx_queue) { - MCDI_DECLARE_BUF(inbuf, MC_CMD_INIT_TXQ_IN_LEN(EFX_MAX_DMAQ_SIZE * 8 / - EFX_BUF_SIZE)); bool csum_offload = tx_queue->queue & EFX_TXQ_TYPE_OFFLOAD; - size_t entries = tx_queue->txd.buf.len / EFX_BUF_SIZE; struct efx_channel *channel = tx_queue->channel; struct efx_nic *efx = tx_queue->efx; - struct efx_ef10_nic_data *nic_data = efx->nic_data; + struct efx_ef10_nic_data *nic_data; bool tso_v2 = false; - size_t inlen; - dma_addr_t dma_addr; efx_qword_t *txd; int rc; - int i; - BUILD_BUG_ON(MC_CMD_INIT_TXQ_OUT_LEN != 0); + + nic_data = efx->nic_data; /* Only attempt to enable TX timestamping if we have the license for it, * otherwise TXQ init will fail @@ -2448,51 +2379,9 @@ static void efx_ef10_tx_init(struct efx_tx_queue *tx_queue) channel->channel); } - MCDI_SET_DWORD(inbuf, INIT_TXQ_IN_SIZE, tx_queue->ptr_mask + 1); - MCDI_SET_DWORD(inbuf, INIT_TXQ_IN_TARGET_EVQ, channel->channel); - MCDI_SET_DWORD(inbuf, INIT_TXQ_IN_LABEL, tx_queue->queue); - MCDI_SET_DWORD(inbuf, INIT_TXQ_IN_INSTANCE, tx_queue->queue); - MCDI_SET_DWORD(inbuf, INIT_TXQ_IN_OWNER_ID, 0); - MCDI_SET_DWORD(inbuf, INIT_TXQ_IN_PORT_ID, nic_data->vport_id); - - dma_addr = tx_queue->txd.buf.dma_addr; - - netif_dbg(efx, hw, efx->net_dev, "pushing TXQ %d. %zu entries (%llx)\n", - tx_queue->queue, entries, (u64)dma_addr); - - for (i = 0; i < entries; ++i) { - MCDI_SET_ARRAY_QWORD(inbuf, INIT_TXQ_IN_DMA_ADDR, i, dma_addr); - dma_addr += EFX_BUF_SIZE; - } - - inlen = MC_CMD_INIT_TXQ_IN_LEN(entries); - - do { - MCDI_POPULATE_DWORD_4(inbuf, INIT_TXQ_IN_FLAGS, - /* This flag was removed from mcdi_pcol.h for - * the non-_EXT version of INIT_TXQ. However, - * firmware still honours it. - */ - INIT_TXQ_EXT_IN_FLAG_TSOV2_EN, tso_v2, - INIT_TXQ_IN_FLAG_IP_CSUM_DIS, !csum_offload, - INIT_TXQ_IN_FLAG_TCP_CSUM_DIS, !csum_offload, - INIT_TXQ_EXT_IN_FLAG_TIMESTAMP, - tx_queue->timestamping); - - rc = efx_mcdi_rpc_quiet(efx, MC_CMD_INIT_TXQ, inbuf, inlen, - NULL, 0, NULL); - if (rc == -ENOSPC && tso_v2) { - /* Retry without TSOv2 if we're short on contexts. */ - tso_v2 = false; - netif_warn(efx, probe, efx->net_dev, - "TSOv2 context not available to segment in hardware. TCP performance may be reduced.\n"); - } else if (rc) { - efx_mcdi_display_error(efx, MC_CMD_INIT_TXQ, - MC_CMD_INIT_TXQ_EXT_IN_LEN, - NULL, 0, rc); - goto fail; - } - } while (rc); + rc = efx_mcdi_tx_init(tx_queue, tso_v2); + if (rc) + goto fail; /* A previous user of this TX queue might have set us up the * bomb by writing a descriptor to the TX push collector but @@ -2530,35 +2419,6 @@ fail: tx_queue->queue); } -static void efx_ef10_tx_fini(struct efx_tx_queue *tx_queue) -{ - MCDI_DECLARE_BUF(inbuf, MC_CMD_FINI_TXQ_IN_LEN); - MCDI_DECLARE_BUF_ERR(outbuf); - struct efx_nic *efx = tx_queue->efx; - size_t outlen; - int rc; - - MCDI_SET_DWORD(inbuf, FINI_TXQ_IN_INSTANCE, - tx_queue->queue); - - rc = efx_mcdi_rpc_quiet(efx, MC_CMD_FINI_TXQ, inbuf, sizeof(inbuf), - outbuf, sizeof(outbuf), &outlen); - - if (rc && rc != -EALREADY) - goto fail; - - return; - -fail: - efx_mcdi_display_error(efx, MC_CMD_FINI_TXQ, MC_CMD_FINI_TXQ_IN_LEN, - outbuf, outlen, rc); -} - -static void efx_ef10_tx_remove(struct efx_tx_queue *tx_queue) -{ - efx_nic_free_buffer(tx_queue->efx, &tx_queue->txd.buf); -} - /* This writes to the TX_DESC_WPTR; write pointer for TX descriptor ring */ static inline void efx_ef10_notify_tx_desc(struct efx_tx_queue *tx_queue) { @@ -2737,7 +2597,7 @@ static int efx_ef10_alloc_rss_context(struct efx_nic *efx, bool exclusive, EFX_EF10_MAX_SHARED_RSS_CONTEXT_SIZE); if (!exclusive && rss_spread == 1) { - ctx->context_id = EFX_EF10_RSS_CONTEXT_INVALID; + ctx->context_id = EFX_MCDI_RSS_CONTEXT_INVALID; if (context_size) *context_size = 1; return 0; @@ -2824,11 +2684,11 @@ static void efx_ef10_rx_free_indir_table(struct efx_nic *efx) { int rc; - if (efx->rss_context.context_id != EFX_EF10_RSS_CONTEXT_INVALID) { + if (efx->rss_context.context_id != EFX_MCDI_RSS_CONTEXT_INVALID) { rc = efx_ef10_free_rss_context(efx, efx->rss_context.context_id); WARN_ON(rc != 0); } - efx->rss_context.context_id = EFX_EF10_RSS_CONTEXT_INVALID; + efx->rss_context.context_id = EFX_MCDI_RSS_CONTEXT_INVALID; } static int efx_ef10_rx_push_shared_rss_config(struct efx_nic *efx, @@ -2854,7 +2714,7 @@ static int efx_ef10_rx_push_exclusive_rss_config(struct efx_nic *efx, struct efx_ef10_nic_data *nic_data = efx->nic_data; int rc; - if (efx->rss_context.context_id == EFX_EF10_RSS_CONTEXT_INVALID || + if (efx->rss_context.context_id == EFX_MCDI_RSS_CONTEXT_INVALID || !nic_data->rx_rss_context_exclusive) { rc = efx_ef10_alloc_rss_context(efx, true, &efx->rss_context, NULL); @@ -2870,7 +2730,7 @@ static int efx_ef10_rx_push_exclusive_rss_config(struct efx_nic *efx, goto fail2; if (efx->rss_context.context_id != old_rx_rss_context && - old_rx_rss_context != EFX_EF10_RSS_CONTEXT_INVALID) + old_rx_rss_context != EFX_MCDI_RSS_CONTEXT_INVALID) WARN_ON(efx_ef10_free_rss_context(efx, old_rx_rss_context) != 0); nic_data->rx_rss_context_exclusive = true; if (rx_indir_table != efx->rss_context.rx_indir_table) @@ -2901,7 +2761,7 @@ static int efx_ef10_rx_push_rss_context_config(struct efx_nic *efx, WARN_ON(!mutex_is_locked(&efx->rss_lock)); - if (ctx->context_id == EFX_EF10_RSS_CONTEXT_INVALID) { + if (ctx->context_id == EFX_MCDI_RSS_CONTEXT_INVALID) { rc = efx_ef10_alloc_rss_context(efx, true, ctx, NULL); if (rc) return rc; @@ -2936,7 +2796,7 @@ static int efx_ef10_rx_pull_rss_context_config(struct efx_nic *efx, BUILD_BUG_ON(MC_CMD_RSS_CONTEXT_GET_TABLE_IN_LEN != MC_CMD_RSS_CONTEXT_GET_KEY_IN_LEN); - if (ctx->context_id == EFX_EF10_RSS_CONTEXT_INVALID) + if (ctx->context_id == EFX_MCDI_RSS_CONTEXT_INVALID) return -ENOENT; MCDI_SET_DWORD(inbuf, RSS_CONTEXT_GET_TABLE_IN_RSS_CONTEXT_ID, @@ -2997,7 +2857,7 @@ static void efx_ef10_rx_restore_rss_contexts(struct efx_nic *efx) list_for_each_entry(ctx, &efx->rss_context.list, list) { /* previous NIC RSS context is gone */ - ctx->context_id = EFX_EF10_RSS_CONTEXT_INVALID; + ctx->context_id = EFX_MCDI_RSS_CONTEXT_INVALID; /* so try to allocate a new one */ rc = efx_ef10_rx_push_rss_context_config(efx, ctx, ctx->rx_indir_table, @@ -3068,96 +2928,11 @@ static int efx_ef10_vf_rx_push_rss_config(struct efx_nic *efx, bool user, { if (user) return -EOPNOTSUPP; - if (efx->rss_context.context_id != EFX_EF10_RSS_CONTEXT_INVALID) + if (efx->rss_context.context_id != EFX_MCDI_RSS_CONTEXT_INVALID) return 0; return efx_ef10_rx_push_shared_rss_config(efx, NULL); } -static int efx_ef10_rx_probe(struct efx_rx_queue *rx_queue) -{ - return efx_nic_alloc_buffer(rx_queue->efx, &rx_queue->rxd.buf, - (rx_queue->ptr_mask + 1) * - sizeof(efx_qword_t), - GFP_KERNEL); -} - -static void efx_ef10_rx_init(struct efx_rx_queue *rx_queue) -{ - MCDI_DECLARE_BUF(inbuf, - MC_CMD_INIT_RXQ_IN_LEN(EFX_MAX_DMAQ_SIZE * 8 / - EFX_BUF_SIZE)); - struct efx_channel *channel = efx_rx_queue_channel(rx_queue); - size_t entries = rx_queue->rxd.buf.len / EFX_BUF_SIZE; - struct efx_nic *efx = rx_queue->efx; - struct efx_ef10_nic_data *nic_data = efx->nic_data; - size_t inlen; - dma_addr_t dma_addr; - int rc; - int i; - BUILD_BUG_ON(MC_CMD_INIT_RXQ_OUT_LEN != 0); - - rx_queue->scatter_n = 0; - rx_queue->scatter_len = 0; - - MCDI_SET_DWORD(inbuf, INIT_RXQ_IN_SIZE, rx_queue->ptr_mask + 1); - MCDI_SET_DWORD(inbuf, INIT_RXQ_IN_TARGET_EVQ, channel->channel); - MCDI_SET_DWORD(inbuf, INIT_RXQ_IN_LABEL, efx_rx_queue_index(rx_queue)); - MCDI_SET_DWORD(inbuf, INIT_RXQ_IN_INSTANCE, - efx_rx_queue_index(rx_queue)); - MCDI_POPULATE_DWORD_2(inbuf, INIT_RXQ_IN_FLAGS, - INIT_RXQ_IN_FLAG_PREFIX, 1, - INIT_RXQ_IN_FLAG_TIMESTAMP, 1); - MCDI_SET_DWORD(inbuf, INIT_RXQ_IN_OWNER_ID, 0); - MCDI_SET_DWORD(inbuf, INIT_RXQ_IN_PORT_ID, nic_data->vport_id); - - dma_addr = rx_queue->rxd.buf.dma_addr; - - netif_dbg(efx, hw, efx->net_dev, "pushing RXQ %d. %zu entries (%llx)\n", - efx_rx_queue_index(rx_queue), entries, (u64)dma_addr); - - for (i = 0; i < entries; ++i) { - MCDI_SET_ARRAY_QWORD(inbuf, INIT_RXQ_IN_DMA_ADDR, i, dma_addr); - dma_addr += EFX_BUF_SIZE; - } - - inlen = MC_CMD_INIT_RXQ_IN_LEN(entries); - - rc = efx_mcdi_rpc(efx, MC_CMD_INIT_RXQ, inbuf, inlen, - NULL, 0, NULL); - if (rc) - netdev_WARN(efx->net_dev, "failed to initialise RXQ %d\n", - efx_rx_queue_index(rx_queue)); -} - -static void efx_ef10_rx_fini(struct efx_rx_queue *rx_queue) -{ - MCDI_DECLARE_BUF(inbuf, MC_CMD_FINI_RXQ_IN_LEN); - MCDI_DECLARE_BUF_ERR(outbuf); - struct efx_nic *efx = rx_queue->efx; - size_t outlen; - int rc; - - MCDI_SET_DWORD(inbuf, FINI_RXQ_IN_INSTANCE, - efx_rx_queue_index(rx_queue)); - - rc = efx_mcdi_rpc_quiet(efx, MC_CMD_FINI_RXQ, inbuf, sizeof(inbuf), - outbuf, sizeof(outbuf), &outlen); - - if (rc && rc != -EALREADY) - goto fail; - - return; - -fail: - efx_mcdi_display_error(efx, MC_CMD_FINI_RXQ, MC_CMD_FINI_RXQ_IN_LEN, - outbuf, outlen, rc); -} - -static void efx_ef10_rx_remove(struct efx_rx_queue *rx_queue) -{ - efx_nic_free_buffer(rx_queue->efx, &rx_queue->rxd.buf); -} - /* This creates an entry in the RX descriptor queue */ static inline void efx_ef10_build_rx_desc(struct efx_rx_queue *rx_queue, unsigned int index) @@ -3229,106 +3004,20 @@ efx_ef10_rx_defer_refill_complete(struct efx_nic *efx, unsigned long cookie, /* nothing to do */ } -static int efx_ef10_ev_probe(struct efx_channel *channel) -{ - return efx_nic_alloc_buffer(channel->efx, &channel->eventq.buf, - (channel->eventq_mask + 1) * - sizeof(efx_qword_t), - GFP_KERNEL); -} - -static void efx_ef10_ev_fini(struct efx_channel *channel) -{ - MCDI_DECLARE_BUF(inbuf, MC_CMD_FINI_EVQ_IN_LEN); - MCDI_DECLARE_BUF_ERR(outbuf); - struct efx_nic *efx = channel->efx; - size_t outlen; - int rc; - - MCDI_SET_DWORD(inbuf, FINI_EVQ_IN_INSTANCE, channel->channel); - - rc = efx_mcdi_rpc_quiet(efx, MC_CMD_FINI_EVQ, inbuf, sizeof(inbuf), - outbuf, sizeof(outbuf), &outlen); - - if (rc && rc != -EALREADY) - goto fail; - - return; - -fail: - efx_mcdi_display_error(efx, MC_CMD_FINI_EVQ, MC_CMD_FINI_EVQ_IN_LEN, - outbuf, outlen, rc); -} - static int efx_ef10_ev_init(struct efx_channel *channel) { - MCDI_DECLARE_BUF(inbuf, - MC_CMD_INIT_EVQ_V2_IN_LEN(EFX_MAX_EVQ_SIZE * 8 / - EFX_BUF_SIZE)); - MCDI_DECLARE_BUF(outbuf, MC_CMD_INIT_EVQ_V2_OUT_LEN); - size_t entries = channel->eventq.buf.len / EFX_BUF_SIZE; struct efx_nic *efx = channel->efx; struct efx_ef10_nic_data *nic_data; - size_t inlen, outlen; unsigned int enabled, implemented; - dma_addr_t dma_addr; + bool use_v2, cut_thru; int rc; - int i; nic_data = efx->nic_data; - - /* Fill event queue with all ones (i.e. empty events) */ - memset(channel->eventq.buf.addr, 0xff, channel->eventq.buf.len); - - MCDI_SET_DWORD(inbuf, INIT_EVQ_IN_SIZE, channel->eventq_mask + 1); - MCDI_SET_DWORD(inbuf, INIT_EVQ_IN_INSTANCE, channel->channel); - /* INIT_EVQ expects index in vector table, not absolute */ - MCDI_SET_DWORD(inbuf, INIT_EVQ_IN_IRQ_NUM, channel->channel); - MCDI_SET_DWORD(inbuf, INIT_EVQ_IN_TMR_MODE, - MC_CMD_INIT_EVQ_IN_TMR_MODE_DIS); - MCDI_SET_DWORD(inbuf, INIT_EVQ_IN_TMR_LOAD, 0); - MCDI_SET_DWORD(inbuf, INIT_EVQ_IN_TMR_RELOAD, 0); - MCDI_SET_DWORD(inbuf, INIT_EVQ_IN_COUNT_MODE, - MC_CMD_INIT_EVQ_IN_COUNT_MODE_DIS); - MCDI_SET_DWORD(inbuf, INIT_EVQ_IN_COUNT_THRSHLD, 0); - - if (nic_data->datapath_caps2 & - 1 << MC_CMD_GET_CAPABILITIES_V2_OUT_INIT_EVQ_V2_LBN) { - /* Use the new generic approach to specifying event queue - * configuration, requesting lower latency or higher throughput. - * The options that actually get used appear in the output. - */ - MCDI_POPULATE_DWORD_2(inbuf, INIT_EVQ_V2_IN_FLAGS, - INIT_EVQ_V2_IN_FLAG_INTERRUPTING, 1, - INIT_EVQ_V2_IN_FLAG_TYPE, - MC_CMD_INIT_EVQ_V2_IN_FLAG_TYPE_AUTO); - } else { - bool cut_thru = !(nic_data->datapath_caps & - 1 << MC_CMD_GET_CAPABILITIES_OUT_RX_BATCHING_LBN); - - MCDI_POPULATE_DWORD_4(inbuf, INIT_EVQ_IN_FLAGS, - INIT_EVQ_IN_FLAG_INTERRUPTING, 1, - INIT_EVQ_IN_FLAG_RX_MERGE, 1, - INIT_EVQ_IN_FLAG_TX_MERGE, 1, - INIT_EVQ_IN_FLAG_CUT_THRU, cut_thru); - } - - dma_addr = channel->eventq.buf.dma_addr; - for (i = 0; i < entries; ++i) { - MCDI_SET_ARRAY_QWORD(inbuf, INIT_EVQ_IN_DMA_ADDR, i, dma_addr); - dma_addr += EFX_BUF_SIZE; - } - - inlen = MC_CMD_INIT_EVQ_IN_LEN(entries); - - rc = efx_mcdi_rpc(efx, MC_CMD_INIT_EVQ, inbuf, inlen, - outbuf, sizeof(outbuf), &outlen); - - if (outlen >= MC_CMD_INIT_EVQ_V2_OUT_LEN) - netif_dbg(efx, drv, efx->net_dev, - "Channel %d using event queue flags %08x\n", - channel->channel, - MCDI_DWORD(outbuf, INIT_EVQ_V2_OUT_FLAGS)); + use_v2 = nic_data->datapath_caps2 & + 1 << MC_CMD_GET_CAPABILITIES_V2_OUT_INIT_EVQ_V2_LBN; + cut_thru = !(nic_data->datapath_caps & + 1 << MC_CMD_GET_CAPABILITIES_OUT_RX_BATCHING_LBN); + rc = efx_mcdi_ev_init(channel, cut_thru, use_v2); /* IRQ return is ignored */ if (channel->channel || rc) @@ -3386,15 +3075,10 @@ static int efx_ef10_ev_init(struct efx_channel *channel) return 0; fail: - efx_ef10_ev_fini(channel); + efx_mcdi_ev_fini(channel); return rc; } -static void efx_ef10_ev_remove(struct efx_channel *channel) -{ - efx_nic_free_buffer(channel->efx, &channel->eventq.buf); -} - static void efx_ef10_handle_rx_wrong_queue(struct efx_rx_queue *rx_queue, unsigned int rx_queue_label) { @@ -3976,9 +3660,9 @@ static int efx_ef10_fini_dmaq(struct efx_nic *efx) if (efx->state != STATE_RECOVERY) { efx_for_each_channel(channel, efx) { efx_for_each_channel_rx_queue(rx_queue, channel) - efx_ef10_rx_fini(rx_queue); + efx_mcdi_rx_fini(rx_queue); efx_for_each_channel_tx_queue(tx_queue, channel) - efx_ef10_tx_fini(tx_queue); + efx_mcdi_tx_fini(tx_queue); } wait_event_timeout(efx->flush_wq, @@ -4165,7 +3849,7 @@ static void efx_ef10_filter_push_prep(struct efx_nic *efx, */ if (WARN_ON_ONCE(!ctx)) flags &= ~EFX_FILTER_FLAG_RX_RSS; - else if (WARN_ON_ONCE(ctx->context_id == EFX_EF10_RSS_CONTEXT_INVALID)) + else if (WARN_ON_ONCE(ctx->context_id == EFX_MCDI_RSS_CONTEXT_INVALID)) flags &= ~EFX_FILTER_FLAG_RX_RSS; } @@ -4344,7 +4028,7 @@ static s32 efx_ef10_filter_insert_locked(struct efx_nic *efx, rc = -ENOENT; goto out_unlock; } - if (ctx->context_id == EFX_EF10_RSS_CONTEXT_INVALID) { + if (ctx->context_id == EFX_MCDI_RSS_CONTEXT_INVALID) { rc = -EOPNOTSUPP; goto out_unlock; } @@ -5085,7 +4769,7 @@ static void efx_ef10_filter_table_restore(struct efx_nic *efx) invalid_filters++; goto not_restored; } - if (ctx->context_id == EFX_EF10_RSS_CONTEXT_INVALID) { + if (ctx->context_id == EFX_MCDI_RSS_CONTEXT_INVALID) { netif_warn(efx, drv, efx->net_dev, "Warning: unable to restore a filter with RSS context %u as it was not created.\n", spec->rss_context); @@ -6650,20 +6334,20 @@ const struct efx_nic_type efx_hunt_a0_vf_nic_type = { .irq_handle_legacy = efx_ef10_legacy_interrupt, .tx_probe = efx_ef10_tx_probe, .tx_init = efx_ef10_tx_init, - .tx_remove = efx_ef10_tx_remove, + .tx_remove = efx_mcdi_tx_remove, .tx_write = efx_ef10_tx_write, .tx_limit_len = efx_ef10_tx_limit_len, .rx_push_rss_config = efx_ef10_vf_rx_push_rss_config, .rx_pull_rss_config = efx_ef10_rx_pull_rss_config, - .rx_probe = efx_ef10_rx_probe, - .rx_init = efx_ef10_rx_init, - .rx_remove = efx_ef10_rx_remove, + .rx_probe = efx_mcdi_rx_probe, + .rx_init = efx_mcdi_rx_init, + .rx_remove = efx_mcdi_rx_remove, .rx_write = efx_ef10_rx_write, .rx_defer_refill = efx_ef10_rx_defer_refill, - .ev_probe = efx_ef10_ev_probe, + .ev_probe = efx_mcdi_ev_probe, .ev_init = efx_ef10_ev_init, - .ev_fini = efx_ef10_ev_fini, - .ev_remove = efx_ef10_ev_remove, + .ev_fini = efx_mcdi_ev_fini, + .ev_remove = efx_mcdi_ev_remove, .ev_process = efx_ef10_ev_process, .ev_read_ack = efx_ef10_ev_read_ack, .ev_test_generate = efx_ef10_ev_test_generate, @@ -6759,7 +6443,7 @@ const struct efx_nic_type efx_hunt_a0_nic_type = { .irq_handle_legacy = efx_ef10_legacy_interrupt, .tx_probe = efx_ef10_tx_probe, .tx_init = efx_ef10_tx_init, - .tx_remove = efx_ef10_tx_remove, + .tx_remove = efx_mcdi_tx_remove, .tx_write = efx_ef10_tx_write, .tx_limit_len = efx_ef10_tx_limit_len, .rx_push_rss_config = efx_ef10_pf_rx_push_rss_config, @@ -6767,15 +6451,15 @@ const struct efx_nic_type efx_hunt_a0_nic_type = { .rx_push_rss_context_config = efx_ef10_rx_push_rss_context_config, .rx_pull_rss_context_config = efx_ef10_rx_pull_rss_context_config, .rx_restore_rss_contexts = efx_ef10_rx_restore_rss_contexts, - .rx_probe = efx_ef10_rx_probe, - .rx_init = efx_ef10_rx_init, - .rx_remove = efx_ef10_rx_remove, + .rx_probe = efx_mcdi_rx_probe, + .rx_init = efx_mcdi_rx_init, + .rx_remove = efx_mcdi_rx_remove, .rx_write = efx_ef10_rx_write, .rx_defer_refill = efx_ef10_rx_defer_refill, - .ev_probe = efx_ef10_ev_probe, + .ev_probe = efx_mcdi_ev_probe, .ev_init = efx_ef10_ev_init, - .ev_fini = efx_ef10_ev_fini, - .ev_remove = efx_ef10_ev_remove, + .ev_fini = efx_mcdi_ev_fini, + .ev_remove = efx_mcdi_ev_remove, .ev_process = efx_ef10_ev_process, .ev_read_ack = efx_ef10_ev_read_ack, .ev_test_generate = efx_ef10_ev_test_generate, diff --git a/drivers/net/ethernet/sfc/efx.c b/drivers/net/ethernet/sfc/efx.c index 033907e6fdb0..4481f21a1f43 100644 --- a/drivers/net/ethernet/sfc/efx.c +++ b/drivers/net/ethernet/sfc/efx.c @@ -23,6 +23,10 @@ #include <net/gre.h> #include <net/udp_tunnel.h> #include "efx.h" +#include "efx_common.h" +#include "efx_channels.h" +#include "rx_common.h" +#include "tx_common.h" #include "nic.h" #include "io.h" #include "selftest.h" @@ -39,56 +43,6 @@ ************************************************************************** */ -/* Loopback mode names (see LOOPBACK_MODE()) */ -const unsigned int efx_loopback_mode_max = LOOPBACK_MAX; -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_XGBR] = "XGBR", - [LOOPBACK_XFI] = "XFI", - [LOOPBACK_XAUI_FAR] = "XAUI_FAR", - [LOOPBACK_GMII_FAR] = "GMII_FAR", - [LOOPBACK_SGMII_FAR] = "SGMII_FAR", - [LOOPBACK_XFI_FAR] = "XFI_FAR", - [LOOPBACK_GPHY] = "GPHY", - [LOOPBACK_PHYXS] = "PHYXS", - [LOOPBACK_PCS] = "PCS", - [LOOPBACK_PMAPMD] = "PMA/PMD", - [LOOPBACK_XPORT] = "XPORT", - [LOOPBACK_XGMII_WS] = "XGMII_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_XFI_WS] = "XFI_WS", - [LOOPBACK_XFI_WS_FAR] = "XFI_WS_FAR", - [LOOPBACK_PHYXS_WS] = "PHYXS_WS", -}; - -const unsigned int efx_reset_type_max = RESET_TYPE_MAX; -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_DATAPATH] = "DATAPATH", - [RESET_TYPE_MC_BIST] = "MC_BIST", - [RESET_TYPE_DISABLE] = "DISABLE", - [RESET_TYPE_TX_WATCHDOG] = "TX_WATCHDOG", - [RESET_TYPE_INT_ERROR] = "INT_ERROR", - [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)", -}; - /* UDP tunnel type names */ static const char *const efx_udp_tunnel_type_names[] = { [TUNNEL_ENCAP_UDP_PORT_ENTRY_VXLAN] = "vxlan", @@ -104,18 +58,6 @@ void efx_get_udp_tunnel_type_name(u16 type, char *buf, size_t buflen) snprintf(buf, buflen, "type %d", type); } -/* 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 @@ -135,21 +77,6 @@ module_param(efx_separate_tx_channels, bool, 0444); MODULE_PARM_DESC(efx_separate_tx_channels, "Use separate channels for TX and RX"); -/* This is the weight assigned to each of the (per-channel) virtual - * NAPI devices. - */ -static int napi_weight = 64; - -/* This is the time (in jiffies) between invocations of the hardware - * 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; - /* Initial interrupt moderation settings. They can be modified after * module load with ethtool. * @@ -169,38 +96,10 @@ static unsigned int rx_irq_mod_usec = 60; */ static unsigned int tx_irq_mod_usec = 150; -/* This is the first interrupt mode to try out of: - * 0 => MSI-X - * 1 => MSI - * 2 => legacy - */ -static unsigned int interrupt_mode; - -/* This is the requested number of CPUs to use for Receive-Side Scaling (RSS), - * i.e. the number of CPUs among which we may distribute simultaneous - * 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 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 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 = 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 = 16000; -module_param(irq_adapt_high_thresh, uint, 0644); -MODULE_PARM_DESC(irq_adapt_high_thresh, - "Threshold score for increasing IRQ moderation"); - static unsigned debug = (NETIF_MSG_DRV | NETIF_MSG_PROBE | NETIF_MSG_LINK | NETIF_MSG_IFDOWN | NETIF_MSG_IFUP | NETIF_MSG_RX_ERR | @@ -214,18 +113,8 @@ 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_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); -static void efx_start_all(struct efx_nic *efx); -static void efx_stop_all(struct efx_nic *efx); static int efx_xdp_setup_prog(struct efx_nic *efx, struct bpf_prog *prog); static int efx_xdp(struct net_device *dev, struct netdev_bpf *xdp); static int efx_xdp_xmit(struct net_device *dev, int n, struct xdp_frame **xdpfs, @@ -239,776 +128,12 @@ static int efx_xdp_xmit(struct net_device *dev, int n, struct xdp_frame **xdpfs, 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 - * - *************************************************************************/ - -/* Process channel's event queue - * - * This function is responsible for processing the event queue of a - * single channel. The caller must guarantee that this function will - * never be concurrently called more than once on the same channel, - * though different channels may be being processed concurrently. - */ -static int efx_process_channel(struct efx_channel *channel, int budget) -{ - struct efx_tx_queue *tx_queue; - struct list_head rx_list; - int spent; - - if (unlikely(!channel->enabled)) - return 0; - - /* Prepare the batch receive list */ - EFX_WARN_ON_PARANOID(channel->rx_list != NULL); - INIT_LIST_HEAD(&rx_list); - channel->rx_list = &rx_list; - - efx_for_each_channel_tx_queue(tx_queue, channel) { - tx_queue->pkts_compl = 0; - tx_queue->bytes_compl = 0; - } - - spent = efx_nic_process_eventq(channel, budget); - if (spent && efx_channel_has_rx_queue(channel)) { - struct efx_rx_queue *rx_queue = - efx_channel_get_rx_queue(channel); - - efx_rx_flush_packet(channel); - efx_fast_push_rx_descriptors(rx_queue, true); - } - - /* Update BQL */ - efx_for_each_channel_tx_queue(tx_queue, channel) { - if (tx_queue->bytes_compl) { - netdev_tx_completed_queue(tx_queue->core_txq, - tx_queue->pkts_compl, tx_queue->bytes_compl); - } - } - - /* Receive any packets we queued up */ - netif_receive_skb_list(channel->rx_list); - channel->rx_list = NULL; - - return spent; -} - -/* NAPI poll handler - * - * NAPI guarantees serialisation of polls of the same device, which - * provides the guarantee required by efx_process_channel(). - */ -static void efx_update_irq_mod(struct efx_nic *efx, struct efx_channel *channel) -{ - int step = efx->irq_mod_step_us; - - if (channel->irq_mod_score < irq_adapt_low_thresh) { - if (channel->irq_moderation_us > step) { - channel->irq_moderation_us -= step; - efx->type->push_irq_moderation(channel); - } - } else if (channel->irq_mod_score > irq_adapt_high_thresh) { - if (channel->irq_moderation_us < - efx->irq_rx_moderation_us) { - channel->irq_moderation_us += step; - efx->type->push_irq_moderation(channel); - } - } - - channel->irq_count = 0; - channel->irq_mod_score = 0; -} - -static int efx_poll(struct napi_struct *napi, int budget) -{ - struct efx_channel *channel = - container_of(napi, struct efx_channel, napi_str); - struct efx_nic *efx = channel->efx; - int spent; - - netif_vdbg(efx, intr, efx->net_dev, - "channel %d NAPI poll executing on CPU %d\n", - channel->channel, raw_smp_processor_id()); - - spent = efx_process_channel(channel, budget); - - xdp_do_flush_map(); - - if (spent < budget) { - if (efx_channel_has_rx_queue(channel) && - efx->irq_rx_adaptive && - unlikely(++channel->irq_count == 1000)) { - efx_update_irq_mod(efx, channel); - } - -#ifdef CONFIG_RFS_ACCEL - /* Perhaps expire some ARFS filters */ - mod_delayed_work(system_wq, &channel->filter_work, 0); -#endif - - /* There is no race here; although napi_disable() will - * only wait for napi_complete(), this isn't a problem - * since efx_nic_eventq_read_ack() will have no effect if - * interrupts have already been disabled. - */ - if (napi_complete_done(napi, spent)) - efx_nic_eventq_read_ack(channel); - } - - return spent; -} - -/* 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 - * errors during channel reset and also simplifies interrupt handling. - */ -static int efx_probe_eventq(struct efx_channel *channel) -{ - struct efx_nic *efx = channel->efx; - unsigned long entries; - - 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, - * plus some extra for link state events and MCDI completions. */ - entries = roundup_pow_of_two(efx->rxq_entries + efx->txq_entries + 128); - EFX_WARN_ON_PARANOID(entries > EFX_MAX_EVQ_SIZE); - channel->eventq_mask = max(entries, EFX_MIN_EVQ_SIZE) - 1; - - return efx_nic_probe_eventq(channel); -} - -/* Prepare channel's event queue */ -static int efx_init_eventq(struct efx_channel *channel) -{ - 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); - - 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 */ -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 */ -void efx_stop_eventq(struct efx_channel *channel) -{ - if (!channel->enabled) - return; - - 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) -{ - netif_dbg(channel->efx, drv, channel->efx->net_dev, - "chan %d remove event queue\n", channel->channel); - - efx_nic_remove_eventq(channel); -} - -/************************************************************************** - * - * Channel handling - * - *************************************************************************/ - -/* Allocate and initialise a channel structure. */ -static struct efx_channel * -efx_alloc_channel(struct efx_nic *efx, int i, struct efx_channel *old_channel) -{ - struct efx_channel *channel; - struct efx_rx_queue *rx_queue; - struct efx_tx_queue *tx_queue; - int j; - - 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; - } - -#ifdef CONFIG_RFS_ACCEL - INIT_DELAYED_WORK(&channel->filter_work, efx_filter_rfs_expire); -#endif - - rx_queue = &channel->rx_queue; - rx_queue->efx = efx; - timer_setup(&rx_queue->slow_fill, efx_rx_slow_fill, 0); - - return channel; -} - -/* 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; - - channel = kmalloc(sizeof(*channel), GFP_KERNEL); - if (!channel) - return NULL; - - *channel = *old_channel; - - channel->napi_dev = NULL; - INIT_HLIST_NODE(&channel->napi_str.napi_hash_node); - channel->napi_str.napi_id = 0; - channel->napi_str.state = 0; - memset(&channel->eventq, 0, sizeof(channel->eventq)); - - 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->buffer = NULL; - memset(&rx_queue->rxd, 0, sizeof(rx_queue->rxd)); - timer_setup(&rx_queue->slow_fill, efx_rx_slow_fill, 0); -#ifdef CONFIG_RFS_ACCEL - INIT_DELAYED_WORK(&channel->filter_work, efx_filter_rfs_expire); -#endif - - return channel; -} - -static int efx_probe_channel(struct efx_channel *channel) -{ - struct efx_tx_queue *tx_queue; - struct efx_rx_queue *rx_queue; - int rc; - - 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 fail; - - efx_for_each_channel_tx_queue(tx_queue, channel) { - rc = efx_probe_tx_queue(tx_queue); - if (rc) - goto fail; - } - - efx_for_each_channel_rx_queue(rx_queue, channel) { - rc = efx_probe_rx_queue(rx_queue); - if (rc) - goto fail; - } - - channel->rx_list = NULL; - - return 0; - -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 (number >= efx->xdp_channel_offset && - !WARN_ON_ONCE(!efx->n_xdp_channels)) { - type = "-xdp"; - number -= efx->xdp_channel_offset; - } else if (efx->tx_channel_offset == 0) { - type = ""; - } else if (number < 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; - - 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) -{ - struct efx_channel *channel; - int rc; - - /* Restart special buffer allocation */ - efx->next_buffer_table = 0; - - /* 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, - "failed to create channel %d\n", - channel->channel); - goto fail; - } - } - efx_set_channel_names(efx); - - return 0; - -fail: - efx_remove_channels(efx); - return rc; -} - -/* Channels are shutdown and reinitialised whilst the NIC is running - * to propagate configuration changes (mtu, checksum offload), or - * to clear hardware error conditions - */ -static void efx_start_datapath(struct efx_nic *efx) -{ - netdev_features_t old_features = efx->net_dev->features; - 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_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) + XDP_PACKET_HEADROOM + - 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); - - /* Restore previously fixed features in hw_features and remove - * features which are fixed now - */ - efx->net_dev->hw_features |= efx->net_dev->features; - efx->net_dev->hw_features &= ~efx->fixed_features; - efx->net_dev->features |= efx->fixed_features; - if (efx->net_dev->features != old_features) - netdev_features_change(efx->net_dev); - - /* 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) { - efx_for_each_channel_tx_queue(tx_queue, channel) { - efx_init_tx_queue(tx_queue); - atomic_inc(&efx->active_queues); - } - - 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_n_frags); - } - - efx_ptp_start_datapath(efx); - - if (netif_device_present(efx->net_dev)) - netif_tx_wake_all_queues(efx->net_dev); -} - -static void efx_stop_datapath(struct efx_nic *efx) -{ - struct efx_channel *channel; - struct efx_tx_queue *tx_queue; - struct efx_rx_queue *rx_queue; - int rc; - - EFX_ASSERT_RESET_SERIALISED(efx); - BUG_ON(efx->port_enabled); - - 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) { - netif_err(efx, drv, efx->net_dev, "failed to flush queues\n"); - } else { - netif_dbg(efx, drv, efx->net_dev, - "successfully flushed all queues\n"); - } - - efx_for_each_channel(channel, efx) { - 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->xdp_rxq_info_failed = false; -} - -static void efx_remove_channel(struct efx_channel *channel) -{ - struct efx_tx_queue *tx_queue; - struct efx_rx_queue *rx_queue; - - netif_dbg(channel->efx, drv, channel->efx->net_dev, - "destroy chan %d\n", channel->channel); - - efx_for_each_channel_rx_queue(rx_queue, channel) - efx_remove_rx_queue(rx_queue); - 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) -{ - struct efx_channel *channel; - - efx_for_each_channel(channel, efx) - efx_remove_channel(channel); - - kfree(efx->xdp_tx_queues); -} - -int -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, 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_soft_disable_interrupts(efx); - - /* Clone channels (where possible) */ - memset(other_channel, 0, sizeof(other_channel)); - for (i = 0; i < efx->n_channels; i++) { - channel = efx->channel[i]; - if (channel->type->copy) - channel = channel->type->copy(channel); - if (!channel) { - rc = -ENOMEM; - goto out; - } - other_channel[i] = channel; - } - - /* Swap entry counts and channel pointers */ - old_rxq_entries = efx->rxq_entries; - old_txq_entries = efx->txq_entries; - efx->rxq_entries = rxq_entries; - efx->txq_entries = txq_entries; - for (i = 0; i < efx->n_channels; i++) { - channel = efx->channel[i]; - efx->channel[i] = other_channel[i]; - other_channel[i] = channel; - } - - /* Restart buffer table allocation */ - efx->next_buffer_table = next_buffer_table; - - for (i = 0; i < efx->n_channels; 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: - /* 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); - } - } - - 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); - efx_device_attach_if_not_resetting(efx); - } - return rc; - -rollback: - /* Swap back */ - efx->rxq_entries = old_rxq_entries; - efx->txq_entries = old_txq_entries; - for (i = 0; i < efx->n_channels; i++) { - channel = efx->channel[i]; - efx->channel[i] = other_channel[i]; - other_channel[i] = channel; - } - goto out; -} - -void efx_schedule_slow_fill(struct efx_rx_queue *rx_queue) -{ - mod_timer(&rx_queue->slow_fill, jiffies + msecs_to_jiffies(10)); -} - -static bool efx_default_channel_want_txqs(struct efx_channel *channel) -{ - return channel->channel - channel->efx->tx_channel_offset < - channel->efx->n_tx_channels; -} - -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, - .want_txqs = efx_default_channel_want_txqs, - .keep_eventq = false, - .want_pio = true, -}; - -int efx_channel_dummy_op_int(struct efx_channel *channel) -{ - return 0; -} - -void efx_channel_dummy_op_void(struct efx_channel *channel) -{ -} - /************************************************************************** * * Port handling * **************************************************************************/ -/* This ensures that the kernel is kept informed (via - * netif_carrier_on/off) of the link status, and also maintains the - * link status's stop on the port's TX queue. - */ -void efx_link_status_changed(struct efx_nic *efx) -{ - struct efx_link_state *link_state = &efx->link_state; - - /* SFC Bug 5356: A net_dev notifier is registered, so we must ensure - * that no events are triggered between unregister_netdev() and the - * driver unloading. A more general condition is that NETDEV_CHANGE - * can only be generated between NETDEV_UP and NETDEV_DOWN */ - if (!netif_running(efx->net_dev)) - return; - - if (link_state->up != netif_carrier_ok(efx->net_dev)) { - efx->n_link_state_changes++; - - if (link_state->up) - netif_carrier_on(efx->net_dev); - else - netif_carrier_off(efx->net_dev); - } - - /* Status message for kernel log */ - if (link_state->up) - netif_info(efx, link, efx->net_dev, - "link up at %uMbps %s-duplex (MTU %d)\n", - link_state->speed, link_state->fd ? "full" : "half", - efx->net_dev->mtu); - else - netif_info(efx, link, efx->net_dev, "link down\n"); -} - -void efx_link_set_advertising(struct efx_nic *efx, - const unsigned long *advertising) -{ - memcpy(efx->link_advertising, advertising, - sizeof(__ETHTOOL_DECLARE_LINK_MODE_MASK())); - - efx->link_advertising[0] |= ADVERTISED_Autoneg; - if (advertising[0] & ADVERTISED_Pause) - efx->wanted_fc |= (EFX_FC_TX | EFX_FC_RX); - else - efx->wanted_fc &= ~(EFX_FC_TX | EFX_FC_RX); - if (advertising[0] & ADVERTISED_Asym_Pause) - efx->wanted_fc ^= EFX_FC_TX; -} - /* Equivalent to efx_link_set_advertising with all-zeroes, except does not * force the Autoneg bit on. */ @@ -1035,73 +160,6 @@ void efx_link_set_wanted_fc(struct efx_nic *efx, u8 wanted_fc) static void efx_fini_port(struct efx_nic *efx); -/* We assume that efx->type->reconfigure_mac will always try to sync RX - * filters and therefore needs to read-lock the filter table against freeing - */ -void efx_mac_reconfigure(struct efx_nic *efx) -{ - down_read(&efx->filter_sem); - efx->type->reconfigure_mac(efx); - up_read(&efx->filter_sem); -} - -/* Push loopback/power/transmit disable settings to the PHY, and reconfigure - * the MAC appropriately. All other PHY configuration changes are pushed - * through phy_op->set_settings(), and pushed asynchronously to the MAC - * through efx_monitor(). - * - * Callers must hold the mac_lock - */ -int __efx_reconfigure_port(struct efx_nic *efx) -{ - enum efx_phy_mode phy_mode; - int rc; - - WARN_ON(!mutex_is_locked(&efx->mac_lock)); - - /* Disable PHY transmit in mac level loopbacks */ - phy_mode = efx->phy_mode; - if (LOOPBACK_INTERNAL(efx)) - efx->phy_mode |= PHY_MODE_TX_DISABLED; - else - efx->phy_mode &= ~PHY_MODE_TX_DISABLED; - - rc = efx->type->reconfigure_port(efx); - - if (rc) - efx->phy_mode = phy_mode; - - return rc; -} - -/* Reinitialise the MAC to pick up new PHY settings, even if the port is - * disabled. */ -int efx_reconfigure_port(struct efx_nic *efx) -{ - int rc; - - EFX_ASSERT_RESET_SERIALISED(efx); - - mutex_lock(&efx->mac_lock); - rc = __efx_reconfigure_port(efx); - mutex_unlock(&efx->mac_lock); - - return rc; -} - -/* Asynchronous work item for changing MAC promiscuity and multicast - * hash. Avoid a drain/rx_ingress enable by reconfiguring the current - * MAC directly. */ -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_mac_reconfigure(efx); - mutex_unlock(&efx->mac_lock); -} - static int efx_probe_port(struct efx_nic *efx) { int rc; @@ -1155,44 +213,6 @@ fail1: return rc; } -static void efx_start_port(struct efx_nic *efx) -{ - netif_dbg(efx, ifup, efx->net_dev, "start port\n"); - BUG_ON(efx->port_enabled); - - mutex_lock(&efx->mac_lock); - efx->port_enabled = true; - - /* Ensure MAC ingress/egress is enabled */ - efx_mac_reconfigure(efx); - - mutex_unlock(&efx->mac_lock); -} - -/* 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() */ - 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) { netif_dbg(efx, drv, efx->net_dev, "shut down port\n"); @@ -1291,582 +311,6 @@ static void efx_dissociate(struct efx_nic *efx) } } -/* 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, bar; - - netif_dbg(efx, probe, efx->net_dev, "initialising I/O\n"); - - bar = efx->type->mem_bar(efx); - - rc = pci_enable_device(pci_dev); - if (rc) { - netif_err(efx, probe, efx->net_dev, - "failed to enable PCI device\n"); - goto fail1; - } - - pci_set_master(pci_dev); - - /* Set the PCI DMA mask. Try all possibilities from our genuine mask - * down to 32 bits, because some architectures will allow 40 bit - * masks event though they reject 46 bit masks. - */ - while (dma_mask > 0x7fffffffUL) { - rc = dma_set_mask_and_coherent(&pci_dev->dev, dma_mask); - if (rc == 0) - break; - dma_mask >>= 1; - } - if (rc) { - netif_err(efx, probe, efx->net_dev, - "could not find a suitable DMA mask\n"); - goto fail2; - } - netif_dbg(efx, probe, efx->net_dev, - "using DMA mask %llx\n", (unsigned long long) dma_mask); - - efx->membase_phys = pci_resource_start(efx->pci_dev, bar); - rc = pci_request_region(pci_dev, bar, "sfc"); - if (rc) { - netif_err(efx, probe, efx->net_dev, - "request for memory BAR failed\n"); - rc = -EIO; - goto fail3; - } - 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, 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, mem_map_size, - efx->membase); - - return 0; - - fail4: - pci_release_region(efx->pci_dev, bar); - fail3: - efx->membase_phys = 0; - fail2: - pci_disable_device(efx->pci_dev); - fail1: - return rc; -} - -static void efx_fini_io(struct efx_nic *efx) -{ - int bar; - - netif_dbg(efx, drv, efx->net_dev, "shutting down I/O\n"); - - if (efx->membase) { - iounmap(efx->membase); - efx->membase = NULL; - } - - if (efx->membase_phys) { - bar = efx->type->mem_bar(efx); - pci_release_region(efx->pci_dev, bar); - efx->membase_phys = 0; - } - - /* Don't disable bus-mastering if VFs are assigned */ - if (!pci_vfs_assigned(efx->pci_dev)) - pci_disable_device(efx->pci_dev); -} - -void efx_set_default_rx_indir_table(struct efx_nic *efx, - struct efx_rss_context *ctx) -{ - size_t i; - - for (i = 0; i < ARRAY_SIZE(ctx->rx_indir_table); i++) - ctx->rx_indir_table[i] = - ethtool_rxfh_indir_default(i, efx->rss_spread); -} - -static unsigned int efx_wanted_parallelism(struct efx_nic *efx) -{ - cpumask_var_t thread_mask; - unsigned int count; - int cpu; - - 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; - } - - count = 0; - for_each_online_cpu(cpu) { - if (!cpumask_test_cpu(cpu, thread_mask)) { - ++count; - cpumask_or(thread_mask, thread_mask, - topology_sibling_cpumask(cpu)); - } - } - - free_cpumask_var(thread_mask); - } - - if (count > EFX_MAX_RX_QUEUES) { - netif_cond_dbg(efx, probe, efx->net_dev, !rss_cpus, warn, - "Reducing number of rx queues from %u to %u.\n", - count, EFX_MAX_RX_QUEUES); - count = EFX_MAX_RX_QUEUES; - } - - /* If RSS is requested for the PF *and* VFs then we can't write RSS - * table entries that are inaccessible to VFs - */ -#ifdef CONFIG_SFC_SRIOV - if (efx->type->sriov_wanted) { - if (efx->type->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); - } - } -#endif - - return count; -} - -static int efx_allocate_msix_channels(struct efx_nic *efx, - unsigned int max_channels, - unsigned int extra_channels, - unsigned int parallelism) -{ - unsigned int n_channels = parallelism; - int vec_count; - int n_xdp_tx; - int n_xdp_ev; - - if (efx_separate_tx_channels) - n_channels *= 2; - n_channels += extra_channels; - - /* To allow XDP transmit to happen from arbitrary NAPI contexts - * we allocate a TX queue per CPU. We share event queues across - * multiple tx queues, assuming tx and ev queues are both - * maximum size. - */ - - n_xdp_tx = num_possible_cpus(); - n_xdp_ev = DIV_ROUND_UP(n_xdp_tx, EFX_TXQ_TYPES); - - vec_count = pci_msix_vec_count(efx->pci_dev); - if (vec_count < 0) - return vec_count; - - max_channels = min_t(unsigned int, vec_count, max_channels); - - /* Check resources. - * We need a channel per event queue, plus a VI per tx queue. - * This may be more pessimistic than it needs to be. - */ - if (n_channels + n_xdp_ev > max_channels) { - netif_err(efx, drv, efx->net_dev, - "Insufficient resources for %d XDP event queues (%d other channels, max %d)\n", - n_xdp_ev, n_channels, max_channels); - efx->n_xdp_channels = 0; - efx->xdp_tx_per_channel = 0; - efx->xdp_tx_queue_count = 0; - } else { - efx->n_xdp_channels = n_xdp_ev; - efx->xdp_tx_per_channel = EFX_TXQ_TYPES; - efx->xdp_tx_queue_count = n_xdp_tx; - n_channels += n_xdp_ev; - netif_dbg(efx, drv, efx->net_dev, - "Allocating %d TX and %d event queues for XDP\n", - n_xdp_tx, n_xdp_ev); - } - - if (vec_count < n_channels) { - netif_err(efx, drv, efx->net_dev, - "WARNING: Insufficient MSI-X vectors available (%d < %u).\n", - vec_count, n_channels); - netif_err(efx, drv, efx->net_dev, - "WARNING: Performance may be reduced.\n"); - n_channels = vec_count; - } - - n_channels = min(n_channels, max_channels); - - efx->n_channels = n_channels; - - /* Ignore XDP tx channels when creating rx channels. */ - n_channels -= efx->n_xdp_channels; - - if (efx_separate_tx_channels) { - efx->n_tx_channels = - min(max(n_channels / 2, 1U), - efx->max_tx_channels); - efx->tx_channel_offset = - n_channels - efx->n_tx_channels; - efx->n_rx_channels = - max(n_channels - - efx->n_tx_channels, 1U); - } else { - efx->n_tx_channels = min(n_channels, efx->max_tx_channels); - efx->tx_channel_offset = 0; - efx->n_rx_channels = n_channels; - } - - efx->n_rx_channels = min(efx->n_rx_channels, parallelism); - efx->n_tx_channels = min(efx->n_tx_channels, parallelism); - - efx->xdp_channel_offset = n_channels; - - netif_dbg(efx, drv, efx->net_dev, - "Allocating %u RX channels\n", - efx->n_rx_channels); - - return efx->n_channels; -} - -/* 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) -{ - unsigned int extra_channels = 0; - unsigned int rss_spread; - 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) { - unsigned int parallelism = efx_wanted_parallelism(efx); - struct msix_entry xentries[EFX_MAX_CHANNELS]; - unsigned int n_channels; - - rc = efx_allocate_msix_channels(efx, efx->max_channels, - extra_channels, parallelism); - if (rc >= 0) { - n_channels = rc; - for (i = 0; i < n_channels; i++) - xentries[i].entry = i; - rc = pci_enable_msix_range(efx->pci_dev, xentries, 1, - n_channels); - } - if (rc < 0) { - /* Fall back to single channel MSI */ - netif_err(efx, drv, efx->net_dev, - "could not enable MSI-X\n"); - if (efx->type->min_interrupt_mode >= EFX_INT_MODE_MSI) - efx->interrupt_mode = EFX_INT_MODE_MSI; - else - return rc; - } else if (rc < n_channels) { - netif_err(efx, drv, efx->net_dev, - "WARNING: Insufficient MSI-X vectors" - " available (%d < %u).\n", rc, n_channels); - netif_err(efx, drv, efx->net_dev, - "WARNING: Performance may be reduced.\n"); - n_channels = rc; - } - - if (rc > 0) { - for (i = 0; i < efx->n_channels; i++) - efx_get_channel(efx, i)->irq = - xentries[i].vector; - } - } - - /* Try single interrupt MSI */ - if (efx->interrupt_mode == EFX_INT_MODE_MSI) { - efx->n_channels = 1; - efx->n_rx_channels = 1; - efx->n_tx_channels = 1; - efx->n_xdp_channels = 0; - efx->xdp_channel_offset = efx->n_channels; - rc = pci_enable_msi(efx->pci_dev); - if (rc == 0) { - efx_get_channel(efx, 0)->irq = efx->pci_dev->irq; - } else { - netif_err(efx, drv, efx->net_dev, - "could not enable MSI\n"); - if (efx->type->min_interrupt_mode >= EFX_INT_MODE_LEGACY) - efx->interrupt_mode = EFX_INT_MODE_LEGACY; - else - return rc; - } - } - - /* Assume legacy interrupts */ - if (efx->interrupt_mode == EFX_INT_MODE_LEGACY) { - efx->n_channels = 1 + (efx_separate_tx_channels ? 1 : 0); - efx->n_rx_channels = 1; - efx->n_tx_channels = 1; - efx->n_xdp_channels = 0; - efx->xdp_channel_offset = efx->n_channels; - efx->legacy_irq = efx->pci_dev->irq; - } - - /* Assign extra channels if possible, before XDP channels */ - efx->n_extra_tx_channels = 0; - j = efx->xdp_channel_offset; - for (i = 0; i < EFX_MAX_EXTRA_CHANNELS; i++) { - if (!efx->extra_channel_type[i]) - continue; - if (j <= efx->tx_channel_offset + efx->n_tx_channels) { - efx->extra_channel_type[i]->handle_no_channel(efx); - } else { - --j; - efx_get_channel(efx, j)->type = - efx->extra_channel_type[i]; - if (efx_channel_has_tx_queues(efx_get_channel(efx, j))) - efx->n_extra_tx_channels++; - } - } - - rss_spread = efx->n_rx_channels; - /* RSS might be usable on VFs even if it is disabled on the PF */ -#ifdef CONFIG_SFC_SRIOV - if (efx->type->sriov_wanted) { - efx->rss_spread = ((rss_spread > 1 || - !efx->type->sriov_wanted(efx)) ? - rss_spread : efx_vf_size(efx)); - return 0; - } -#endif - efx->rss_spread = rss_spread; - - return 0; -} - -#if defined(CONFIG_SMP) -static void efx_set_interrupt_affinity(struct efx_nic *efx) -{ - struct efx_channel *channel; - unsigned int cpu; - - efx_for_each_channel(channel, efx) { - cpu = cpumask_local_spread(channel->channel, - pcibus_to_node(efx->pci_dev->bus)); - irq_set_affinity_hint(channel->irq, cpumask_of(cpu)); - } -} - -static void efx_clear_interrupt_affinity(struct efx_nic *efx) -{ - struct efx_channel *channel; - - efx_for_each_channel(channel, efx) - irq_set_affinity_hint(channel->irq, NULL); -} -#else -static void -efx_set_interrupt_affinity(struct efx_nic *efx __attribute__ ((unused))) -{ -} - -static void -efx_clear_interrupt_affinity(struct efx_nic *efx __attribute__ ((unused))) -{ -} -#endif /* CONFIG_SMP */ - -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) -{ - struct efx_channel *channel; - - /* Remove MSI/MSI-X interrupts */ - efx_for_each_channel(channel, efx) - channel->irq = 0; - pci_disable_msi(efx->pci_dev); - pci_disable_msix(efx->pci_dev); - - /* Remove legacy interrupt */ - efx->legacy_irq = 0; -} - -static int efx_set_channels(struct efx_nic *efx) -{ - struct efx_channel *channel; - struct efx_tx_queue *tx_queue; - int xdp_queue_number; - - efx->tx_channel_offset = - efx_separate_tx_channels ? - efx->n_channels - efx->n_tx_channels : 0; - - if (efx->xdp_tx_queue_count) { - EFX_WARN_ON_PARANOID(efx->xdp_tx_queues); - - /* Allocate array for XDP TX queue lookup. */ - efx->xdp_tx_queues = kcalloc(efx->xdp_tx_queue_count, - sizeof(*efx->xdp_tx_queues), - GFP_KERNEL); - if (!efx->xdp_tx_queues) - return -ENOMEM; - } - - /* 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. - */ - xdp_queue_number = 0; - 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); - - if (efx_channel_is_xdp_tx(channel) && - xdp_queue_number < efx->xdp_tx_queue_count) { - efx->xdp_tx_queues[xdp_queue_number] = tx_queue; - xdp_queue_number++; - } - } - } - return 0; -} - static int efx_probe_nic(struct efx_nic *efx) { int rc; @@ -1939,70 +383,6 @@ static void efx_remove_nic(struct efx_nic *efx) efx->type->remove(efx); } -static int efx_probe_filters(struct efx_nic *efx) -{ - int rc; - - init_rwsem(&efx->filter_sem); - mutex_lock(&efx->mac_lock); - down_write(&efx->filter_sem); - rc = efx->type->filter_table_probe(efx); - if (rc) - goto out_unlock; - -#ifdef CONFIG_RFS_ACCEL - if (efx->type->offload_features & NETIF_F_NTUPLE) { - struct efx_channel *channel; - int i, success = 1; - - efx_for_each_channel(channel, efx) { - channel->rps_flow_id = - kcalloc(efx->type->max_rx_ip_filters, - sizeof(*channel->rps_flow_id), - GFP_KERNEL); - if (!channel->rps_flow_id) - success = 0; - else - for (i = 0; - i < efx->type->max_rx_ip_filters; - ++i) - channel->rps_flow_id[i] = - RPS_FLOW_ID_INVALID; - channel->rfs_expire_index = 0; - channel->rfs_filter_count = 0; - } - - if (!success) { - efx_for_each_channel(channel, efx) - kfree(channel->rps_flow_id); - efx->type->filter_table_remove(efx); - rc = -ENOMEM; - goto out_unlock; - } - } -#endif -out_unlock: - up_write(&efx->filter_sem); - mutex_unlock(&efx->mac_lock); - return rc; -} - -static void efx_remove_filters(struct efx_nic *efx) -{ -#ifdef CONFIG_RFS_ACCEL - struct efx_channel *channel; - - efx_for_each_channel(channel, efx) { - cancel_delayed_work_sync(&channel->filter_work); - kfree(channel->rps_flow_id); - } -#endif - down_write(&efx->filter_sem); - efx->type->filter_table_remove(efx); - up_write(&efx->filter_sem); -} - - /************************************************************************** * * NIC startup/shutdown @@ -2067,81 +447,6 @@ static int efx_probe_all(struct efx_nic *efx) return rc; } -/* 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) -{ - 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 || !netif_running(efx->net_dev) || - efx->reset_pending) - return; - - efx_start_port(efx); - efx_start_datapath(efx); - - /* Start the hardware monitor if there is one */ - if (efx->type->monitor != NULL) - queue_delayed_work(efx->workqueue, &efx->monitor_work, - efx_monitor_interval); - - /* Link state detection is normally event-driven; we have - * to poll now because we could have missed a change - */ - mutex_lock(&efx->mac_lock); - if (efx->phy_op->poll(efx)) - efx_link_status_changed(efx); - mutex_unlock(&efx->mac_lock); - - 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); -} - -/* 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) -{ - 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); - efx_stop_port(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); - - efx_stop_datapath(efx); -} - static void efx_remove_all(struct efx_nic *efx) { rtnl_lock(); @@ -2237,36 +542,6 @@ void efx_get_irq_moderation(struct efx_nic *efx, unsigned int *tx_usecs, /************************************************************************** * - * Hardware monitor - * - **************************************************************************/ - -/* Run periodically off the general workqueue */ -static void efx_monitor(struct work_struct *data) -{ - struct efx_nic *efx = container_of(data, struct efx_nic, - monitor_work.work); - - netif_vdbg(efx, timer, efx->net_dev, - "hardware monitor executing on CPU %d\n", - raw_smp_processor_id()); - BUG_ON(efx->type->monitor == NULL); - - /* If the mac_lock is already held then it is likely a port - * reconfiguration is already in place, which will likely do - * most of the work of monitor() anyway. */ - if (mutex_trylock(&efx->mac_lock)) { - if (efx->port_enabled) - efx->type->monitor(efx); - mutex_unlock(&efx->mac_lock); - } - - queue_delayed_work(efx->workqueue, &efx->monitor_work, - efx_monitor_interval); -} - -/************************************************************************** - * * ioctls * *************************************************************************/ @@ -2294,45 +569,6 @@ static int efx_ioctl(struct net_device *net_dev, struct ifreq *ifr, int cmd) /************************************************************************** * - * NAPI interface - * - **************************************************************************/ - -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) - efx_init_napi_channel(channel); -} - -static void efx_fini_napi_channel(struct efx_channel *channel) -{ - if (channel->napi_dev) - netif_napi_del(&channel->napi_str); - - channel->napi_dev = NULL; -} - -static void efx_fini_napi(struct efx_nic *efx) -{ - struct efx_channel *channel; - - efx_for_each_channel(channel, efx) - efx_fini_napi_channel(channel); -} - -/************************************************************************** - * * Kernel net device interface * *************************************************************************/ @@ -2382,17 +618,6 @@ int efx_net_stop(struct net_device *net_dev) return 0; } -/* Context: process, dev_base_lock or RTNL held, non-blocking. */ -static void efx_net_stats(struct net_device *net_dev, - struct rtnl_link_stats64 *stats) -{ - struct efx_nic *efx = netdev_priv(net_dev); - - spin_lock_bh(&efx->stats_lock); - efx->type->update_stats(efx, NULL, stats); - spin_unlock_bh(&efx->stats_lock); -} - /* Context: netif_tx_lock held, BHs disabled. */ static void efx_watchdog(struct net_device *net_dev, unsigned int txqueue) { @@ -2405,51 +630,6 @@ static void efx_watchdog(struct net_device *net_dev, unsigned int txqueue) efx_schedule_reset(efx, RESET_TYPE_TX_WATCHDOG); } -static unsigned int efx_xdp_max_mtu(struct efx_nic *efx) -{ - /* The maximum MTU that we can fit in a single page, allowing for - * framing, overhead and XDP headroom. - */ - int overhead = EFX_MAX_FRAME_LEN(0) + sizeof(struct efx_rx_page_state) + - efx->rx_prefix_size + efx->type->rx_buffer_padding + - efx->rx_ip_align + XDP_PACKET_HEADROOM; - - return PAGE_SIZE - overhead; -} - -/* Context: process, rtnl_lock() held. */ -static int efx_change_mtu(struct net_device *net_dev, int new_mtu) -{ - struct efx_nic *efx = netdev_priv(net_dev); - int rc; - - rc = efx_check_disabled(efx); - if (rc) - return rc; - - if (rtnl_dereference(efx->xdp_prog) && - new_mtu > efx_xdp_max_mtu(efx)) { - netif_err(efx, drv, efx->net_dev, - "Requested MTU of %d too big for XDP (max: %d)\n", - new_mtu, efx_xdp_max_mtu(efx)); - return -EINVAL; - } - - netif_dbg(efx, drv, efx->net_dev, "changing MTU to %d\n", new_mtu); - - efx_device_detach_sync(efx); - efx_stop_all(efx); - - mutex_lock(&efx->mac_lock); - net_dev->mtu = new_mtu; - efx_mac_reconfigure(efx); - mutex_unlock(&efx->mac_lock); - - efx_start_all(efx); - efx_device_attach_if_not_resetting(efx); - return 0; -} - static int efx_set_mac_address(struct net_device *net_dev, void *data) { struct efx_nic *efx = netdev_priv(net_dev); @@ -2726,28 +906,6 @@ show_phy_type(struct device *dev, struct device_attribute *attr, char *buf) } static DEVICE_ATTR(phy_type, 0444, show_phy_type, NULL); -#ifdef CONFIG_SFC_MCDI_LOGGING -static ssize_t show_mcdi_log(struct device *dev, struct device_attribute *attr, - char *buf) -{ - struct efx_nic *efx = dev_get_drvdata(dev); - struct efx_mcdi_iface *mcdi = efx_mcdi(efx); - - return scnprintf(buf, PAGE_SIZE, "%d\n", mcdi->logging_enabled); -} -static ssize_t set_mcdi_log(struct device *dev, struct device_attribute *attr, - const char *buf, size_t count) -{ - struct efx_nic *efx = dev_get_drvdata(dev); - struct efx_mcdi_iface *mcdi = efx_mcdi(efx); - bool enable = count > 0 && *buf != '0'; - - mcdi->logging_enabled = enable; - return count; -} -static DEVICE_ATTR(mcdi_logging, 0644, show_mcdi_log, set_mcdi_log); -#endif - static int efx_register_netdev(struct efx_nic *efx) { struct net_device *net_dev = efx->net_dev; @@ -2807,21 +965,11 @@ static int efx_register_netdev(struct efx_nic *efx) "failed to init net dev attributes\n"); goto fail_registered; } -#ifdef CONFIG_SFC_MCDI_LOGGING - rc = device_create_file(&efx->pci_dev->dev, &dev_attr_mcdi_logging); - if (rc) { - netif_err(efx, drv, efx->net_dev, - "failed to init net dev attributes\n"); - goto fail_attr_mcdi_logging; - } -#endif + + efx_init_mcdi_logging(efx); return 0; -#ifdef CONFIG_SFC_MCDI_LOGGING -fail_attr_mcdi_logging: - device_remove_file(&efx->pci_dev->dev, &dev_attr_phy_type); -#endif fail_registered: rtnl_lock(); efx_dissociate(efx); @@ -2842,9 +990,7 @@ static void efx_unregister_netdev(struct efx_nic *efx) if (efx_dev_registered(efx)) { strlcpy(efx->name, pci_name(efx->pci_dev), sizeof(efx->name)); -#ifdef CONFIG_SFC_MCDI_LOGGING - device_remove_file(&efx->pci_dev->dev, &dev_attr_mcdi_logging); -#endif + efx_fini_mcdi_logging(efx); device_remove_file(&efx->pci_dev->dev, &dev_attr_phy_type); unregister_netdev(efx->net_dev); } @@ -2852,292 +998,6 @@ static void efx_unregister_netdev(struct efx_nic *efx) /************************************************************************** * - * Device reset and suspend - * - **************************************************************************/ - -/* Tears down the entire software state and most of the hardware state - * before reset. */ -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); - efx_disable_interrupts(efx); - - mutex_lock(&efx->mac_lock); - down_write(&efx->filter_sem); - mutex_lock(&efx->rss_lock); - if (efx->port_initialized && method != RESET_TYPE_INVISIBLE && - method != RESET_TYPE_DATAPATH) - efx->phy_op->fini(efx); - efx->type->fini(efx); -} - -/* This function will always ensure that the locks acquired in - * efx_reset_down() are released. A failure return code indicates - * that we were unable to reinitialise the hardware, and the - * driver should be disabled. If ok is false, then the rx and tx - * engines are not restarted, pending a RESET_DISABLE. */ -int efx_reset_up(struct efx_nic *efx, enum reset_type method, bool ok) -{ - int rc; - - 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"); - goto fail; - } - - if (!ok) - goto fail; - - if (efx->port_initialized && method != RESET_TYPE_INVISIBLE && - method != RESET_TYPE_DATAPATH) { - rc = efx->phy_op->init(efx); - if (rc) - goto fail; - rc = efx->phy_op->reconfigure(efx); - if (rc && rc != -EPERM) - netif_err(efx, drv, efx->net_dev, - "could not restore PHY settings\n"); - } - - rc = efx_enable_interrupts(efx); - if (rc) - goto fail; - -#ifdef CONFIG_SFC_SRIOV - rc = efx->type->vswitching_restore(efx); - if (rc) /* not fatal; the PF will still work fine */ - netif_warn(efx, probe, efx->net_dev, - "failed to restore vswitching rc=%d;" - " VFs may not function\n", rc); -#endif - - if (efx->type->rx_restore_rss_contexts) - efx->type->rx_restore_rss_contexts(efx); - mutex_unlock(&efx->rss_lock); - efx->type->filter_table_restore(efx); - up_write(&efx->filter_sem); - if (efx->type->sriov_reset) - efx->type->sriov_reset(efx); - - mutex_unlock(&efx->mac_lock); - - efx_start_all(efx); - - if (efx->type->udp_tnl_push_ports) - efx->type->udp_tnl_push_ports(efx); - - return 0; - -fail: - efx->port_initialized = false; - - mutex_unlock(&efx->rss_lock); - up_write(&efx->filter_sem); - mutex_unlock(&efx->mac_lock); - - return rc; -} - -/* Reset the NIC using the specified method. Note that the reset may - * fail, in which case the card will be left in an unusable state. - * - * Caller must hold the rtnl_lock. - */ -int efx_reset(struct efx_nic *efx, enum reset_type method) -{ - int rc, rc2; - bool disabled; - - netif_info(efx, drv, efx->net_dev, "resetting (%s)\n", - RESET_TYPE(method)); - - efx_device_detach_sync(efx); - efx_reset_down(efx, method); - - rc = efx->type->reset(efx, method); - if (rc) { - netif_err(efx, drv, efx->net_dev, "failed to reset hardware\n"); - goto out; - } - - /* Clear flags for the scopes we covered. We assume the NIC and - * driver are now quiescent so that there is no race here. - */ - 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 - * RESET_TYPE_DISABLE since this driver generally assumes the hardware - * can respond to requests. */ - pci_set_master(efx->pci_dev); - -out: - /* Leave device stopped if necessary */ - disabled = rc || - method == RESET_TYPE_DISABLE || - method == RESET_TYPE_RECOVER_OR_DISABLE; - rc2 = efx_reset_up(efx, method, !disabled); - if (rc2) { - disabled = true; - if (!rc) - rc = rc2; - } - - if (disabled) { - dev_close(efx->net_dev); - netif_err(efx, drv, efx->net_dev, "has been disabled\n"); - efx->state = STATE_DISABLED; - } else { - netif_dbg(efx, drv, efx->net_dev, "reset complete\n"); - efx_device_attach_if_not_resetting(efx); - } - 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 = pci_dev_to_eeh_dev(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; - enum reset_type method; - - pending = READ_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 (!pending) - return; - - rtnl_lock(); - - /* 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(); -} - -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_DATAPATH: - 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)); - break; - default: - method = efx->type->map_reset_reason(type); - netif_dbg(efx, drv, efx->net_dev, - "scheduling %s reset for %s\n", - RESET_TYPE(method), RESET_TYPE(type)); - break; - } - - 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 (READ_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. */ - efx_mcdi_mode_poll(efx); - - queue_work(reset_workqueue, &efx->reset_work); -} - -/************************************************************************** - * * List of NICs we support * **************************************************************************/ @@ -3169,139 +1029,10 @@ static const struct pci_device_id efx_pci_table[] = { /************************************************************************** * - * Dummy PHY/MAC operations - * - * Can be used for some unimplemented operations - * Needed so all function pointers are valid and do not have to be tested - * before use - * - **************************************************************************/ -int efx_port_dummy_op_int(struct efx_nic *efx) -{ - return 0; -} -void efx_port_dummy_op_void(struct efx_nic *efx) {} - -static bool efx_port_dummy_op_poll(struct efx_nic *efx) -{ - return false; -} - -static const struct efx_phy_operations efx_dummy_phy_operations = { - .init = efx_port_dummy_op_int, - .reconfigure = efx_port_dummy_op_int, - .poll = efx_port_dummy_op_poll, - .fini = efx_port_dummy_op_void, -}; - -/************************************************************************** - * * Data housekeeping * **************************************************************************/ -/* 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, - struct pci_dev *pci_dev, struct net_device *net_dev) -{ - int rc = -ENOMEM, i; - - /* Initialise common structures */ - 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_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; - INIT_LIST_HEAD(&efx->rss_context.list); - mutex_init(&efx->rss_lock); - spin_lock_init(&efx->stats_lock); - efx->vi_stride = EFX_DEFAULT_VI_STRIDE; - efx->num_mac_stats = MC_CMD_MAC_NSTATS; - BUILD_BUG_ON(MC_CMD_MAC_NSTATS - 1 != MC_CMD_MAC_GENERATION_END); - mutex_init(&efx->mac_lock); -#ifdef CONFIG_RFS_ACCEL - mutex_init(&efx->rps_mutex); - spin_lock_init(&efx->rps_hash_lock); - /* Failure to allocate is not fatal, but may degrade ARFS performance */ - efx->rps_hash_table = kcalloc(EFX_ARFS_HASH_TABLE_SIZE, - sizeof(*efx->rps_hash_table), GFP_KERNEL); -#endif - 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; - } - - /* Higher numbered interrupt modes are less capable! */ - if (WARN_ON_ONCE(efx->type->max_interrupt_mode > - efx->type->min_interrupt_mode)) { - rc = -EIO; - goto fail; - } - efx->interrupt_mode = max(efx->type->max_interrupt_mode, - interrupt_mode); - efx->interrupt_mode = min(efx->type->min_interrupt_mode, - interrupt_mode); - - /* Would be good to use the net_dev name, but we're too early */ - snprintf(efx->workqueue_name, sizeof(efx->workqueue_name), "sfc%s", - pci_name(pci_dev)); - efx->workqueue = create_singlethread_workqueue(efx->workqueue_name); - if (!efx->workqueue) - goto fail; - - return 0; - -fail: - efx_fini_struct(efx); - return rc; -} - -static void efx_fini_struct(struct efx_nic *efx) -{ - int i; - -#ifdef CONFIG_RFS_ACCEL - kfree(efx->rps_hash_table); -#endif - - 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; - } -} - void efx_update_sw_stats(struct efx_nic *efx, u64 *stats) { u64 n_rx_nodesc_trunc = 0; @@ -3313,197 +1044,6 @@ void efx_update_sw_stats(struct efx_nic *efx, u64 *stats) stats[GENERIC_STAT_rx_noskb_drops] = atomic_read(&efx->n_rx_noskb_drops); } -bool efx_filter_spec_equal(const struct efx_filter_spec *left, - const struct efx_filter_spec *right) -{ - if ((left->match_flags ^ right->match_flags) | - ((left->flags ^ right->flags) & - (EFX_FILTER_FLAG_RX | EFX_FILTER_FLAG_TX))) - return false; - - return memcmp(&left->outer_vid, &right->outer_vid, - sizeof(struct efx_filter_spec) - - offsetof(struct efx_filter_spec, outer_vid)) == 0; -} - -u32 efx_filter_spec_hash(const struct efx_filter_spec *spec) -{ - BUILD_BUG_ON(offsetof(struct efx_filter_spec, outer_vid) & 3); - return jhash2((const u32 *)&spec->outer_vid, - (sizeof(struct efx_filter_spec) - - offsetof(struct efx_filter_spec, outer_vid)) / 4, - 0); -} - -#ifdef CONFIG_RFS_ACCEL -bool efx_rps_check_rule(struct efx_arfs_rule *rule, unsigned int filter_idx, - bool *force) -{ - if (rule->filter_id == EFX_ARFS_FILTER_ID_PENDING) { - /* ARFS is currently updating this entry, leave it */ - return false; - } - if (rule->filter_id == EFX_ARFS_FILTER_ID_ERROR) { - /* ARFS tried and failed to update this, so it's probably out - * of date. Remove the filter and the ARFS rule entry. - */ - rule->filter_id = EFX_ARFS_FILTER_ID_REMOVING; - *force = true; - return true; - } else if (WARN_ON(rule->filter_id != filter_idx)) { /* can't happen */ - /* ARFS has moved on, so old filter is not needed. Since we did - * not mark the rule with EFX_ARFS_FILTER_ID_REMOVING, it will - * not be removed by efx_rps_hash_del() subsequently. - */ - *force = true; - return true; - } - /* Remove it iff ARFS wants to. */ - return true; -} - -static -struct hlist_head *efx_rps_hash_bucket(struct efx_nic *efx, - const struct efx_filter_spec *spec) -{ - u32 hash = efx_filter_spec_hash(spec); - - lockdep_assert_held(&efx->rps_hash_lock); - if (!efx->rps_hash_table) - return NULL; - return &efx->rps_hash_table[hash % EFX_ARFS_HASH_TABLE_SIZE]; -} - -struct efx_arfs_rule *efx_rps_hash_find(struct efx_nic *efx, - const struct efx_filter_spec *spec) -{ - struct efx_arfs_rule *rule; - struct hlist_head *head; - struct hlist_node *node; - - head = efx_rps_hash_bucket(efx, spec); - if (!head) - return NULL; - hlist_for_each(node, head) { - rule = container_of(node, struct efx_arfs_rule, node); - if (efx_filter_spec_equal(spec, &rule->spec)) - return rule; - } - return NULL; -} - -struct efx_arfs_rule *efx_rps_hash_add(struct efx_nic *efx, - const struct efx_filter_spec *spec, - bool *new) -{ - struct efx_arfs_rule *rule; - struct hlist_head *head; - struct hlist_node *node; - - head = efx_rps_hash_bucket(efx, spec); - if (!head) - return NULL; - hlist_for_each(node, head) { - rule = container_of(node, struct efx_arfs_rule, node); - if (efx_filter_spec_equal(spec, &rule->spec)) { - *new = false; - return rule; - } - } - rule = kmalloc(sizeof(*rule), GFP_ATOMIC); - *new = true; - if (rule) { - memcpy(&rule->spec, spec, sizeof(rule->spec)); - hlist_add_head(&rule->node, head); - } - return rule; -} - -void efx_rps_hash_del(struct efx_nic *efx, const struct efx_filter_spec *spec) -{ - struct efx_arfs_rule *rule; - struct hlist_head *head; - struct hlist_node *node; - - head = efx_rps_hash_bucket(efx, spec); - if (WARN_ON(!head)) - return; - hlist_for_each(node, head) { - rule = container_of(node, struct efx_arfs_rule, node); - if (efx_filter_spec_equal(spec, &rule->spec)) { - /* Someone already reused the entry. We know that if - * this check doesn't fire (i.e. filter_id == REMOVING) - * then the REMOVING mark was put there by our caller, - * because caller is holding a lock on filter table and - * only holders of that lock set REMOVING. - */ - if (rule->filter_id != EFX_ARFS_FILTER_ID_REMOVING) - return; - hlist_del(node); - kfree(rule); - return; - } - } - /* We didn't find it. */ - WARN_ON(1); -} -#endif - -/* RSS contexts. We're using linked lists and crappy O(n) algorithms, because - * (a) this is an infrequent control-plane operation and (b) n is small (max 64) - */ -struct efx_rss_context *efx_alloc_rss_context_entry(struct efx_nic *efx) -{ - struct list_head *head = &efx->rss_context.list; - struct efx_rss_context *ctx, *new; - u32 id = 1; /* Don't use zero, that refers to the master RSS context */ - - WARN_ON(!mutex_is_locked(&efx->rss_lock)); - - /* Search for first gap in the numbering */ - list_for_each_entry(ctx, head, list) { - if (ctx->user_id != id) - break; - id++; - /* Check for wrap. If this happens, we have nearly 2^32 - * allocated RSS contexts, which seems unlikely. - */ - if (WARN_ON_ONCE(!id)) - return NULL; - } - - /* Create the new entry */ - new = kmalloc(sizeof(struct efx_rss_context), GFP_KERNEL); - if (!new) - return NULL; - new->context_id = EFX_EF10_RSS_CONTEXT_INVALID; - new->rx_hash_udp_4tuple = false; - - /* Insert the new entry into the gap */ - new->user_id = id; - list_add_tail(&new->list, &ctx->list); - return new; -} - -struct efx_rss_context *efx_find_rss_context_entry(struct efx_nic *efx, u32 id) -{ - struct list_head *head = &efx->rss_context.list; - struct efx_rss_context *ctx; - - WARN_ON(!mutex_is_locked(&efx->rss_lock)); - - list_for_each_entry(ctx, head, list) - if (ctx->user_id == id) - return ctx; - return NULL; -} - -void efx_free_rss_context_entry(struct efx_rss_context *ctx) -{ - list_del(&ctx->list); - kfree(ctx); -} - /************************************************************************** * * PCI interface @@ -3519,7 +1059,7 @@ static void efx_pci_remove_main(struct efx_nic *efx) * are not READY. */ BUG_ON(efx->state == STATE_READY); - cancel_work_sync(&efx->reset_work); + efx_flush_reset_workqueue(efx); efx_disable_interrupts(efx); efx_clear_interrupt_affinity(efx); @@ -3559,7 +1099,7 @@ static void efx_pci_remove(struct pci_dev *pci_dev) efx_pci_remove_main(efx); - efx_fini_io(efx); + efx_fini_io(efx, efx->type->mem_bar(efx)); netif_dbg(efx, drv, efx->net_dev, "shutdown successful\n"); efx_fini_struct(efx); @@ -3782,7 +1322,8 @@ static int efx_pci_probe(struct pci_dev *pci_dev, efx_probe_vpd_strings(efx); /* Set up basic I/O (BAR mappings etc) */ - rc = efx_init_io(efx); + rc = efx_init_io(efx, efx->type->mem_bar(efx), efx->type->max_dma_mask, + efx->type->mem_map_size(efx)); if (rc) goto fail2; @@ -3826,7 +1367,7 @@ static int efx_pci_probe(struct pci_dev *pci_dev, return 0; fail3: - efx_fini_io(efx); + efx_fini_io(efx, efx->type->mem_bar(efx)); fail2: efx_fini_struct(efx); fail1: @@ -3904,7 +1445,7 @@ static int efx_pm_thaw(struct device *dev) rtnl_unlock(); /* Reschedule any quenched resets scheduled during efx_pm_freeze() */ - queue_work(reset_workqueue, &efx->reset_work); + efx_queue_reset_work(efx); return 0; @@ -4083,10 +1624,6 @@ static struct pci_driver efx_pci_driver = { * *************************************************************************/ -module_param(interrupt_mode, uint, 0444); -MODULE_PARM_DESC(interrupt_mode, - "Interrupt mode (0=>MSIX 1=>MSI 2=>legacy)"); - static int __init efx_init_module(void) { int rc; @@ -4103,11 +1640,9 @@ static int __init efx_init_module(void) goto err_sriov; #endif - reset_workqueue = create_singlethread_workqueue("sfc_reset"); - if (!reset_workqueue) { - rc = -ENOMEM; + rc = efx_create_reset_workqueue(); + if (rc) goto err_reset; - } rc = pci_register_driver(&efx_pci_driver); if (rc < 0) @@ -4116,7 +1651,7 @@ static int __init efx_init_module(void) return 0; err_pci: - destroy_workqueue(reset_workqueue); + efx_destroy_reset_workqueue(); err_reset: #ifdef CONFIG_SFC_SRIOV efx_fini_sriov(); @@ -4132,7 +1667,7 @@ static void __exit efx_exit_module(void) printk(KERN_INFO "Solarflare NET driver unloading\n"); pci_unregister_driver(&efx_pci_driver); - destroy_workqueue(reset_workqueue); + efx_destroy_reset_workqueue(); #ifdef CONFIG_SFC_SRIOV efx_fini_sriov(); #endif diff --git a/drivers/net/ethernet/sfc/efx.h b/drivers/net/ethernet/sfc/efx.h index 2dd8d5002315..f1bdb04efbe4 100644 --- a/drivers/net/ethernet/sfc/efx.h +++ b/drivers/net/ethernet/sfc/efx.h @@ -15,31 +15,17 @@ int efx_net_open(struct net_device *net_dev); int efx_net_stop(struct net_device *net_dev); /* TX */ -int efx_probe_tx_queue(struct efx_tx_queue *tx_queue); -void efx_remove_tx_queue(struct efx_tx_queue *tx_queue); -void efx_init_tx_queue(struct efx_tx_queue *tx_queue); void efx_init_tx_queue_core_txq(struct efx_tx_queue *tx_queue); -void efx_fini_tx_queue(struct efx_tx_queue *tx_queue); netdev_tx_t efx_hard_start_xmit(struct sk_buff *skb, struct net_device *net_dev); netdev_tx_t efx_enqueue_skb(struct efx_tx_queue *tx_queue, struct sk_buff *skb); void efx_xmit_done(struct efx_tx_queue *tx_queue, unsigned int index); int efx_setup_tc(struct net_device *net_dev, enum tc_setup_type type, void *type_data); -unsigned int efx_tx_max_skb_descs(struct efx_nic *efx); extern unsigned int efx_piobuf_size; extern bool efx_separate_tx_channels; /* RX */ -void efx_set_default_rx_indir_table(struct efx_nic *efx, - struct efx_rss_context *ctx); -void efx_rx_config_page_split(struct efx_nic *efx); -int efx_probe_rx_queue(struct efx_rx_queue *rx_queue); -void efx_remove_rx_queue(struct efx_rx_queue *rx_queue); -void efx_init_rx_queue(struct efx_rx_queue *rx_queue); -void efx_fini_rx_queue(struct efx_rx_queue *rx_queue); -void efx_fast_push_rx_descriptors(struct efx_rx_queue *rx_queue, bool atomic); -void efx_rx_slow_fill(struct timer_list *t); void __efx_rx_packet(struct efx_channel *channel); void efx_rx_packet(struct efx_rx_queue *rx_queue, unsigned int index, unsigned int n_frags, unsigned int len, u16 flags); @@ -48,7 +34,6 @@ static inline void efx_rx_flush_packet(struct efx_channel *channel) if (channel->rx_pkt_n_frags) __efx_rx_packet(channel); } -void efx_schedule_slow_fill(struct efx_rx_queue *rx_queue); #define EFX_MAX_DMAQ_SIZE 4096UL #define EFX_DEFAULT_DMAQ_SIZE 1024UL @@ -80,8 +65,6 @@ static inline bool efx_rss_enabled(struct efx_nic *efx) /* Filters */ -void efx_mac_reconfigure(struct efx_nic *efx); - /** * efx_filter_insert_filter - add or replace a filter * @efx: NIC in which to insert the filter @@ -186,58 +169,17 @@ static inline void efx_filter_rfs_expire(struct work_struct *data) static inline void efx_filter_rfs_expire(struct work_struct *data) {} #define efx_filter_rfs_enabled() 0 #endif -bool efx_filter_is_mc_recipient(const struct efx_filter_spec *spec); - -bool efx_filter_spec_equal(const struct efx_filter_spec *left, - const struct efx_filter_spec *right); -u32 efx_filter_spec_hash(const struct efx_filter_spec *spec); - -#ifdef CONFIG_RFS_ACCEL -bool efx_rps_check_rule(struct efx_arfs_rule *rule, unsigned int filter_idx, - bool *force); - -struct efx_arfs_rule *efx_rps_hash_find(struct efx_nic *efx, - const struct efx_filter_spec *spec); - -/* @new is written to indicate if entry was newly added (true) or if an old - * entry was found and returned (false). - */ -struct efx_arfs_rule *efx_rps_hash_add(struct efx_nic *efx, - const struct efx_filter_spec *spec, - bool *new); - -void efx_rps_hash_del(struct efx_nic *efx, const struct efx_filter_spec *spec); -#endif /* RSS contexts */ -struct efx_rss_context *efx_alloc_rss_context_entry(struct efx_nic *efx); -struct efx_rss_context *efx_find_rss_context_entry(struct efx_nic *efx, u32 id); -void efx_free_rss_context_entry(struct efx_rss_context *ctx); static inline bool efx_rss_active(struct efx_rss_context *ctx) { - return ctx->context_id != EFX_EF10_RSS_CONTEXT_INVALID; + return ctx->context_id != EFX_MCDI_RSS_CONTEXT_INVALID; } -/* Channels */ -int efx_channel_dummy_op_int(struct efx_channel *channel); -void efx_channel_dummy_op_void(struct efx_channel *channel); -int efx_realloc_channels(struct efx_nic *efx, u32 rxq_entries, u32 txq_entries); - -/* Ports */ -int efx_reconfigure_port(struct efx_nic *efx); -int __efx_reconfigure_port(struct efx_nic *efx); - /* Ethtool support */ extern const struct ethtool_ops efx_ethtool_ops; -/* Reset handling */ -int efx_reset(struct efx_nic *efx, enum reset_type method); -void efx_reset_down(struct efx_nic *efx, enum reset_type method); -int efx_reset_up(struct efx_nic *efx, enum reset_type method, bool ok); -int efx_try_recovery(struct efx_nic *efx); - /* Global */ -void efx_schedule_reset(struct efx_nic *efx, enum reset_type type); unsigned int efx_usecs_to_ticks(struct efx_nic *efx, unsigned int usecs); unsigned int efx_ticks_to_usecs(struct efx_nic *efx, unsigned int ticks); int efx_init_irq_moderation(struct efx_nic *efx, unsigned int tx_usecs, @@ -245,8 +187,6 @@ int efx_init_irq_moderation(struct efx_nic *efx, unsigned int tx_usecs, bool rx_may_override_tx); void efx_get_irq_moderation(struct efx_nic *efx, unsigned int *tx_usecs, unsigned int *rx_usecs, bool *rx_adaptive); -void efx_stop_eventq(struct efx_channel *channel); -void efx_start_eventq(struct efx_channel *channel); /* Dummy PHY ops for PHY drivers */ int efx_port_dummy_op_int(struct efx_nic *efx); @@ -293,9 +233,6 @@ static inline void efx_schedule_channel_irq(struct efx_channel *channel) efx_schedule_channel(channel); } -void efx_link_status_changed(struct efx_nic *efx); -void efx_link_set_advertising(struct efx_nic *efx, - const unsigned long *advertising); void efx_link_clear_advertising(struct efx_nic *efx); void efx_link_set_wanted_fc(struct efx_nic *efx, u8); diff --git a/drivers/net/ethernet/sfc/efx_channels.c b/drivers/net/ethernet/sfc/efx_channels.c new file mode 100644 index 000000000000..aeb5e8aa2f2a --- /dev/null +++ b/drivers/net/ethernet/sfc/efx_channels.c @@ -0,0 +1,1234 @@ +// SPDX-License-Identifier: GPL-2.0-only +/**************************************************************************** + * Driver for Solarflare network controllers and boards + * Copyright 2018 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 + * by the Free Software Foundation, incorporated herein by reference. + */ + +#include "net_driver.h" +#include <linux/module.h> +#include "efx_channels.h" +#include "efx.h" +#include "efx_common.h" +#include "tx_common.h" +#include "rx_common.h" +#include "nic.h" +#include "sriov.h" + +/* This is the first interrupt mode to try out of: + * 0 => MSI-X + * 1 => MSI + * 2 => legacy + */ +static unsigned int interrupt_mode; +module_param(interrupt_mode, uint, 0444); +MODULE_PARM_DESC(interrupt_mode, + "Interrupt mode (0=>MSIX 1=>MSI 2=>legacy)"); + +/* This is the requested number of CPUs to use for Receive-Side Scaling (RSS), + * i.e. the number of CPUs among which we may distribute simultaneous + * 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 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 unsigned int 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 int 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"); + +/* This is the weight assigned to each of the (per-channel) virtual + * NAPI devices. + */ +static int napi_weight = 64; + +/*************** + * Housekeeping + ***************/ + +int efx_channel_dummy_op_int(struct efx_channel *channel) +{ + return 0; +} + +void efx_channel_dummy_op_void(struct efx_channel *channel) +{ +} + +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, + .want_txqs = efx_default_channel_want_txqs, + .keep_eventq = false, + .want_pio = true, +}; + +/************* + * INTERRUPTS + *************/ + +static unsigned int efx_wanted_parallelism(struct efx_nic *efx) +{ + cpumask_var_t thread_mask; + unsigned int count; + int cpu; + + 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; + } + + count = 0; + for_each_online_cpu(cpu) { + if (!cpumask_test_cpu(cpu, thread_mask)) { + ++count; + cpumask_or(thread_mask, thread_mask, + topology_sibling_cpumask(cpu)); + } + } + + free_cpumask_var(thread_mask); + } + + if (count > EFX_MAX_RX_QUEUES) { + netif_cond_dbg(efx, probe, efx->net_dev, !rss_cpus, warn, + "Reducing number of rx queues from %u to %u.\n", + count, EFX_MAX_RX_QUEUES); + count = EFX_MAX_RX_QUEUES; + } + + /* If RSS is requested for the PF *and* VFs then we can't write RSS + * table entries that are inaccessible to VFs + */ +#ifdef CONFIG_SFC_SRIOV + if (efx->type->sriov_wanted) { + if (efx->type->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); + } + } +#endif + + return count; +} + +static int efx_allocate_msix_channels(struct efx_nic *efx, + unsigned int max_channels, + unsigned int extra_channels, + unsigned int parallelism) +{ + unsigned int n_channels = parallelism; + int vec_count; + int n_xdp_tx; + int n_xdp_ev; + + if (efx_separate_tx_channels) + n_channels *= 2; + n_channels += extra_channels; + + /* To allow XDP transmit to happen from arbitrary NAPI contexts + * we allocate a TX queue per CPU. We share event queues across + * multiple tx queues, assuming tx and ev queues are both + * maximum size. + */ + + n_xdp_tx = num_possible_cpus(); + n_xdp_ev = DIV_ROUND_UP(n_xdp_tx, EFX_TXQ_TYPES); + + vec_count = pci_msix_vec_count(efx->pci_dev); + if (vec_count < 0) + return vec_count; + + max_channels = min_t(unsigned int, vec_count, max_channels); + + /* Check resources. + * We need a channel per event queue, plus a VI per tx queue. + * This may be more pessimistic than it needs to be. + */ + if (n_channels + n_xdp_ev > max_channels) { + netif_err(efx, drv, efx->net_dev, + "Insufficient resources for %d XDP event queues (%d other channels, max %d)\n", + n_xdp_ev, n_channels, max_channels); + efx->n_xdp_channels = 0; + efx->xdp_tx_per_channel = 0; + efx->xdp_tx_queue_count = 0; + } else { + efx->n_xdp_channels = n_xdp_ev; + efx->xdp_tx_per_channel = EFX_TXQ_TYPES; + efx->xdp_tx_queue_count = n_xdp_tx; + n_channels += n_xdp_ev; + netif_dbg(efx, drv, efx->net_dev, + "Allocating %d TX and %d event queues for XDP\n", + n_xdp_tx, n_xdp_ev); + } + + if (vec_count < n_channels) { + netif_err(efx, drv, efx->net_dev, + "WARNING: Insufficient MSI-X vectors available (%d < %u).\n", + vec_count, n_channels); + netif_err(efx, drv, efx->net_dev, + "WARNING: Performance may be reduced.\n"); + n_channels = vec_count; + } + + n_channels = min(n_channels, max_channels); + + efx->n_channels = n_channels; + + /* Ignore XDP tx channels when creating rx channels. */ + n_channels -= efx->n_xdp_channels; + + if (efx_separate_tx_channels) { + efx->n_tx_channels = + min(max(n_channels / 2, 1U), + efx->max_tx_channels); + efx->tx_channel_offset = + n_channels - efx->n_tx_channels; + efx->n_rx_channels = + max(n_channels - + efx->n_tx_channels, 1U); + } else { + efx->n_tx_channels = min(n_channels, efx->max_tx_channels); + efx->tx_channel_offset = 0; + efx->n_rx_channels = n_channels; + } + + efx->n_rx_channels = min(efx->n_rx_channels, parallelism); + efx->n_tx_channels = min(efx->n_tx_channels, parallelism); + + efx->xdp_channel_offset = n_channels; + + netif_dbg(efx, drv, efx->net_dev, + "Allocating %u RX channels\n", + efx->n_rx_channels); + + return efx->n_channels; +} + +/* Probe the number and type of interrupts we are able to obtain, and + * the resulting numbers of channels and RX queues. + */ +int efx_probe_interrupts(struct efx_nic *efx) +{ + unsigned int extra_channels = 0; + unsigned int rss_spread; + 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) { + unsigned int parallelism = efx_wanted_parallelism(efx); + struct msix_entry xentries[EFX_MAX_CHANNELS]; + unsigned int n_channels; + + rc = efx_allocate_msix_channels(efx, efx->max_channels, + extra_channels, parallelism); + if (rc >= 0) { + n_channels = rc; + for (i = 0; i < n_channels; i++) + xentries[i].entry = i; + rc = pci_enable_msix_range(efx->pci_dev, xentries, 1, + n_channels); + } + if (rc < 0) { + /* Fall back to single channel MSI */ + netif_err(efx, drv, efx->net_dev, + "could not enable MSI-X\n"); + if (efx->type->min_interrupt_mode >= EFX_INT_MODE_MSI) + efx->interrupt_mode = EFX_INT_MODE_MSI; + else + return rc; + } else if (rc < n_channels) { + netif_err(efx, drv, efx->net_dev, + "WARNING: Insufficient MSI-X vectors" + " available (%d < %u).\n", rc, n_channels); + netif_err(efx, drv, efx->net_dev, + "WARNING: Performance may be reduced.\n"); + n_channels = rc; + } + + if (rc > 0) { + for (i = 0; i < efx->n_channels; i++) + efx_get_channel(efx, i)->irq = + xentries[i].vector; + } + } + + /* Try single interrupt MSI */ + if (efx->interrupt_mode == EFX_INT_MODE_MSI) { + efx->n_channels = 1; + efx->n_rx_channels = 1; + efx->n_tx_channels = 1; + efx->n_xdp_channels = 0; + efx->xdp_channel_offset = efx->n_channels; + rc = pci_enable_msi(efx->pci_dev); + if (rc == 0) { + efx_get_channel(efx, 0)->irq = efx->pci_dev->irq; + } else { + netif_err(efx, drv, efx->net_dev, + "could not enable MSI\n"); + if (efx->type->min_interrupt_mode >= EFX_INT_MODE_LEGACY) + efx->interrupt_mode = EFX_INT_MODE_LEGACY; + else + return rc; + } + } + + /* Assume legacy interrupts */ + if (efx->interrupt_mode == EFX_INT_MODE_LEGACY) { + efx->n_channels = 1 + (efx_separate_tx_channels ? 1 : 0); + efx->n_rx_channels = 1; + efx->n_tx_channels = 1; + efx->n_xdp_channels = 0; + efx->xdp_channel_offset = efx->n_channels; + efx->legacy_irq = efx->pci_dev->irq; + } + + /* Assign extra channels if possible, before XDP channels */ + efx->n_extra_tx_channels = 0; + j = efx->xdp_channel_offset; + for (i = 0; i < EFX_MAX_EXTRA_CHANNELS; i++) { + if (!efx->extra_channel_type[i]) + continue; + if (j <= efx->tx_channel_offset + efx->n_tx_channels) { + efx->extra_channel_type[i]->handle_no_channel(efx); + } else { + --j; + efx_get_channel(efx, j)->type = + efx->extra_channel_type[i]; + if (efx_channel_has_tx_queues(efx_get_channel(efx, j))) + efx->n_extra_tx_channels++; + } + } + + rss_spread = efx->n_rx_channels; + /* RSS might be usable on VFs even if it is disabled on the PF */ +#ifdef CONFIG_SFC_SRIOV + if (efx->type->sriov_wanted) { + efx->rss_spread = ((rss_spread > 1 || + !efx->type->sriov_wanted(efx)) ? + rss_spread : efx_vf_size(efx)); + return 0; + } +#endif + efx->rss_spread = rss_spread; + + return 0; +} + +#if defined(CONFIG_SMP) +void efx_set_interrupt_affinity(struct efx_nic *efx) +{ + struct efx_channel *channel; + unsigned int cpu; + + efx_for_each_channel(channel, efx) { + cpu = cpumask_local_spread(channel->channel, + pcibus_to_node(efx->pci_dev->bus)); + irq_set_affinity_hint(channel->irq, cpumask_of(cpu)); + } +} + +void efx_clear_interrupt_affinity(struct efx_nic *efx) +{ + struct efx_channel *channel; + + efx_for_each_channel(channel, efx) + irq_set_affinity_hint(channel->irq, NULL); +} +#else +void +efx_set_interrupt_affinity(struct efx_nic *efx __attribute__ ((unused))) +{ +} + +void +efx_clear_interrupt_affinity(struct efx_nic *efx __attribute__ ((unused))) +{ +} +#endif /* CONFIG_SMP */ + +void efx_remove_interrupts(struct efx_nic *efx) +{ + struct efx_channel *channel; + + /* Remove MSI/MSI-X interrupts */ + efx_for_each_channel(channel, efx) + channel->irq = 0; + pci_disable_msi(efx->pci_dev); + pci_disable_msix(efx->pci_dev); + + /* Remove legacy interrupt */ + efx->legacy_irq = 0; +} + +/*************** + * EVENT QUEUES + ***************/ + +/* 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 + * errors during channel reset and also simplifies interrupt handling. + */ +int efx_probe_eventq(struct efx_channel *channel) +{ + struct efx_nic *efx = channel->efx; + unsigned long entries; + + 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, + * plus some extra for link state events and MCDI completions. + */ + entries = roundup_pow_of_two(efx->rxq_entries + efx->txq_entries + 128); + EFX_WARN_ON_PARANOID(entries > EFX_MAX_EVQ_SIZE); + channel->eventq_mask = max(entries, EFX_MIN_EVQ_SIZE) - 1; + + return efx_nic_probe_eventq(channel); +} + +/* Prepare channel's event queue */ +int efx_init_eventq(struct efx_channel *channel) +{ + 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); + + 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 */ +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 */ +void efx_stop_eventq(struct efx_channel *channel) +{ + if (!channel->enabled) + return; + + napi_disable(&channel->napi_str); + channel->enabled = false; +} + +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; +} + +void efx_remove_eventq(struct efx_channel *channel) +{ + netif_dbg(channel->efx, drv, channel->efx->net_dev, + "chan %d remove event queue\n", channel->channel); + + efx_nic_remove_eventq(channel); +} + +/************************************************************************** + * + * Channel handling + * + *************************************************************************/ + +/* Allocate and initialise a channel structure. */ +struct efx_channel * +efx_alloc_channel(struct efx_nic *efx, int i, struct efx_channel *old_channel) +{ + struct efx_rx_queue *rx_queue; + struct efx_tx_queue *tx_queue; + struct efx_channel *channel; + int j; + + 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; + } + +#ifdef CONFIG_RFS_ACCEL + INIT_DELAYED_WORK(&channel->filter_work, efx_filter_rfs_expire); +#endif + + rx_queue = &channel->rx_queue; + rx_queue->efx = efx; + timer_setup(&rx_queue->slow_fill, efx_rx_slow_fill, 0); + + return channel; +} + +int efx_init_channels(struct efx_nic *efx) +{ + unsigned int i; + + for (i = 0; i < EFX_MAX_CHANNELS; i++) { + efx->channel[i] = efx_alloc_channel(efx, i, NULL); + if (!efx->channel[i]) + return -ENOMEM; + efx->msi_context[i].efx = efx; + efx->msi_context[i].index = i; + } + + /* Higher numbered interrupt modes are less capable! */ + if (WARN_ON_ONCE(efx->type->max_interrupt_mode > + efx->type->min_interrupt_mode)) { + return -EIO; + } + efx->interrupt_mode = max(efx->type->max_interrupt_mode, + interrupt_mode); + efx->interrupt_mode = min(efx->type->min_interrupt_mode, + interrupt_mode); + + return 0; +} + +void efx_fini_channels(struct efx_nic *efx) +{ + unsigned int i; + + for (i = 0; i < EFX_MAX_CHANNELS; i++) + if (efx->channel[i]) { + kfree(efx->channel[i]); + efx->channel[i] = NULL; + } +} + +/* Allocate and initialise a channel structure, copying parameters + * (but not resources) from an old channel structure. + */ +struct efx_channel *efx_copy_channel(const struct efx_channel *old_channel) +{ + struct efx_rx_queue *rx_queue; + struct efx_tx_queue *tx_queue; + struct efx_channel *channel; + int j; + + channel = kmalloc(sizeof(*channel), GFP_KERNEL); + if (!channel) + return NULL; + + *channel = *old_channel; + + channel->napi_dev = NULL; + INIT_HLIST_NODE(&channel->napi_str.napi_hash_node); + channel->napi_str.napi_id = 0; + channel->napi_str.state = 0; + memset(&channel->eventq, 0, sizeof(channel->eventq)); + + 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->buffer = NULL; + memset(&rx_queue->rxd, 0, sizeof(rx_queue->rxd)); + timer_setup(&rx_queue->slow_fill, efx_rx_slow_fill, 0); +#ifdef CONFIG_RFS_ACCEL + INIT_DELAYED_WORK(&channel->filter_work, efx_filter_rfs_expire); +#endif + + return channel; +} + +static int efx_probe_channel(struct efx_channel *channel) +{ + struct efx_tx_queue *tx_queue; + struct efx_rx_queue *rx_queue; + int rc; + + 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 fail; + + efx_for_each_channel_tx_queue(tx_queue, channel) { + rc = efx_probe_tx_queue(tx_queue); + if (rc) + goto fail; + } + + efx_for_each_channel_rx_queue(rx_queue, channel) { + rc = efx_probe_rx_queue(rx_queue); + if (rc) + goto fail; + } + + channel->rx_list = NULL; + + return 0; + +fail: + efx_remove_channel(channel); + return rc; +} + +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 (number >= efx->xdp_channel_offset && + !WARN_ON_ONCE(!efx->n_xdp_channels)) { + type = "-xdp"; + number -= efx->xdp_channel_offset; + } else if (efx->tx_channel_offset == 0) { + type = ""; + } else if (number < efx->tx_channel_offset) { + type = "-rx"; + } else { + type = "-tx"; + number -= efx->tx_channel_offset; + } + snprintf(buf, len, "%s%s-%d", efx->name, type, number); +} + +void efx_set_channel_names(struct efx_nic *efx) +{ + struct efx_channel *channel; + + efx_for_each_channel(channel, efx) + channel->type->get_name(channel, + efx->msi_context[channel->channel].name, + sizeof(efx->msi_context[0].name)); +} + +int efx_probe_channels(struct efx_nic *efx) +{ + struct efx_channel *channel; + int rc; + + /* Restart special buffer allocation */ + efx->next_buffer_table = 0; + + /* 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, + "failed to create channel %d\n", + channel->channel); + goto fail; + } + } + efx_set_channel_names(efx); + + return 0; + +fail: + efx_remove_channels(efx); + return rc; +} + +void efx_remove_channel(struct efx_channel *channel) +{ + struct efx_tx_queue *tx_queue; + struct efx_rx_queue *rx_queue; + + netif_dbg(channel->efx, drv, channel->efx->net_dev, + "destroy chan %d\n", channel->channel); + + efx_for_each_channel_rx_queue(rx_queue, channel) + efx_remove_rx_queue(rx_queue); + 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); +} + +void efx_remove_channels(struct efx_nic *efx) +{ + struct efx_channel *channel; + + efx_for_each_channel(channel, efx) + efx_remove_channel(channel); + + kfree(efx->xdp_tx_queues); +} + +int efx_realloc_channels(struct efx_nic *efx, u32 rxq_entries, u32 txq_entries) +{ + struct efx_channel *other_channel[EFX_MAX_CHANNELS], *channel; + unsigned int i, next_buffer_table = 0; + u32 old_rxq_entries, old_txq_entries; + 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_soft_disable_interrupts(efx); + + /* Clone channels (where possible) */ + memset(other_channel, 0, sizeof(other_channel)); + for (i = 0; i < efx->n_channels; i++) { + channel = efx->channel[i]; + if (channel->type->copy) + channel = channel->type->copy(channel); + if (!channel) { + rc = -ENOMEM; + goto out; + } + other_channel[i] = channel; + } + + /* Swap entry counts and channel pointers */ + old_rxq_entries = efx->rxq_entries; + old_txq_entries = efx->txq_entries; + efx->rxq_entries = rxq_entries; + efx->txq_entries = txq_entries; + for (i = 0; i < efx->n_channels; i++) { + channel = efx->channel[i]; + efx->channel[i] = other_channel[i]; + other_channel[i] = channel; + } + + /* Restart buffer table allocation */ + efx->next_buffer_table = next_buffer_table; + + for (i = 0; i < efx->n_channels; 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: + /* 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); + } + } + + 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); + efx_device_attach_if_not_resetting(efx); + } + return rc; + +rollback: + /* Swap back */ + efx->rxq_entries = old_rxq_entries; + efx->txq_entries = old_txq_entries; + for (i = 0; i < efx->n_channels; i++) { + channel = efx->channel[i]; + efx->channel[i] = other_channel[i]; + other_channel[i] = channel; + } + goto out; +} + +int efx_set_channels(struct efx_nic *efx) +{ + struct efx_channel *channel; + struct efx_tx_queue *tx_queue; + int xdp_queue_number; + + efx->tx_channel_offset = + efx_separate_tx_channels ? + efx->n_channels - efx->n_tx_channels : 0; + + if (efx->xdp_tx_queue_count) { + EFX_WARN_ON_PARANOID(efx->xdp_tx_queues); + + /* Allocate array for XDP TX queue lookup. */ + efx->xdp_tx_queues = kcalloc(efx->xdp_tx_queue_count, + sizeof(*efx->xdp_tx_queues), + GFP_KERNEL); + if (!efx->xdp_tx_queues) + return -ENOMEM; + } + + /* 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. + */ + xdp_queue_number = 0; + 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); + + if (efx_channel_is_xdp_tx(channel) && + xdp_queue_number < efx->xdp_tx_queue_count) { + efx->xdp_tx_queues[xdp_queue_number] = tx_queue; + xdp_queue_number++; + } + } + } + return 0; +} + +bool efx_default_channel_want_txqs(struct efx_channel *channel) +{ + return channel->channel - channel->efx->tx_channel_offset < + channel->efx->n_tx_channels; +} + +/************* + * START/STOP + *************/ + +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; +} + +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); +} + +int efx_enable_interrupts(struct efx_nic *efx) +{ + struct efx_channel *channel, *end_channel; + int rc; + + /* TODO: Is this really a bug? */ + 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; +} + +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); +} + +void efx_start_channels(struct efx_nic *efx) +{ + struct efx_tx_queue *tx_queue; + struct efx_rx_queue *rx_queue; + struct efx_channel *channel; + + efx_for_each_channel(channel, efx) { + efx_for_each_channel_tx_queue(tx_queue, channel) { + efx_init_tx_queue(tx_queue); + atomic_inc(&efx->active_queues); + } + + 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_n_frags); + } +} + +void efx_stop_channels(struct efx_nic *efx) +{ + struct efx_tx_queue *tx_queue; + struct efx_rx_queue *rx_queue; + struct efx_channel *channel; + int rc = 0; + + /* 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); + } + } + + if (efx->type->fini_dmaq) + rc = efx->type->fini_dmaq(efx); + + if (rc) { + netif_err(efx, drv, efx->net_dev, "failed to flush queues\n"); + } else { + netif_dbg(efx, drv, efx->net_dev, + "successfully flushed all queues\n"); + } + + efx_for_each_channel(channel, efx) { + 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); + } +} + +/************************************************************************** + * + * NAPI interface + * + *************************************************************************/ + +/* Process channel's event queue + * + * This function is responsible for processing the event queue of a + * single channel. The caller must guarantee that this function will + * never be concurrently called more than once on the same channel, + * though different channels may be being processed concurrently. + */ +static int efx_process_channel(struct efx_channel *channel, int budget) +{ + struct efx_tx_queue *tx_queue; + struct list_head rx_list; + int spent; + + if (unlikely(!channel->enabled)) + return 0; + + /* Prepare the batch receive list */ + EFX_WARN_ON_PARANOID(channel->rx_list != NULL); + INIT_LIST_HEAD(&rx_list); + channel->rx_list = &rx_list; + + efx_for_each_channel_tx_queue(tx_queue, channel) { + tx_queue->pkts_compl = 0; + tx_queue->bytes_compl = 0; + } + + spent = efx_nic_process_eventq(channel, budget); + if (spent && efx_channel_has_rx_queue(channel)) { + struct efx_rx_queue *rx_queue = + efx_channel_get_rx_queue(channel); + + efx_rx_flush_packet(channel); + efx_fast_push_rx_descriptors(rx_queue, true); + } + + /* Update BQL */ + efx_for_each_channel_tx_queue(tx_queue, channel) { + if (tx_queue->bytes_compl) { + netdev_tx_completed_queue(tx_queue->core_txq, + tx_queue->pkts_compl, + tx_queue->bytes_compl); + } + } + + /* Receive any packets we queued up */ + netif_receive_skb_list(channel->rx_list); + channel->rx_list = NULL; + + return spent; +} + +static void efx_update_irq_mod(struct efx_nic *efx, struct efx_channel *channel) +{ + int step = efx->irq_mod_step_us; + + if (channel->irq_mod_score < irq_adapt_low_thresh) { + if (channel->irq_moderation_us > step) { + channel->irq_moderation_us -= step; + efx->type->push_irq_moderation(channel); + } + } else if (channel->irq_mod_score > irq_adapt_high_thresh) { + if (channel->irq_moderation_us < + efx->irq_rx_moderation_us) { + channel->irq_moderation_us += step; + efx->type->push_irq_moderation(channel); + } + } + + channel->irq_count = 0; + channel->irq_mod_score = 0; +} + +/* NAPI poll handler + * + * NAPI guarantees serialisation of polls of the same device, which + * provides the guarantee required by efx_process_channel(). + */ +static int efx_poll(struct napi_struct *napi, int budget) +{ + struct efx_channel *channel = + container_of(napi, struct efx_channel, napi_str); + struct efx_nic *efx = channel->efx; + int spent; + + netif_vdbg(efx, intr, efx->net_dev, + "channel %d NAPI poll executing on CPU %d\n", + channel->channel, raw_smp_processor_id()); + + spent = efx_process_channel(channel, budget); + + xdp_do_flush_map(); + + if (spent < budget) { + if (efx_channel_has_rx_queue(channel) && + efx->irq_rx_adaptive && + unlikely(++channel->irq_count == 1000)) { + efx_update_irq_mod(efx, channel); + } + +#ifdef CONFIG_RFS_ACCEL + /* Perhaps expire some ARFS filters */ + mod_delayed_work(system_wq, &channel->filter_work, 0); +#endif + + /* There is no race here; although napi_disable() will + * only wait for napi_complete(), this isn't a problem + * since efx_nic_eventq_read_ack() will have no effect if + * interrupts have already been disabled. + */ + if (napi_complete_done(napi, spent)) + efx_nic_eventq_read_ack(channel); + } + + return spent; +} + +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); +} + +void efx_init_napi(struct efx_nic *efx) +{ + struct efx_channel *channel; + + efx_for_each_channel(channel, efx) + efx_init_napi_channel(channel); +} + +void efx_fini_napi_channel(struct efx_channel *channel) +{ + if (channel->napi_dev) + netif_napi_del(&channel->napi_str); + + channel->napi_dev = NULL; +} + +void efx_fini_napi(struct efx_nic *efx) +{ + struct efx_channel *channel; + + efx_for_each_channel(channel, efx) + efx_fini_napi_channel(channel); +} diff --git a/drivers/net/ethernet/sfc/efx_channels.h b/drivers/net/ethernet/sfc/efx_channels.h new file mode 100644 index 000000000000..8d7b8c4142d7 --- /dev/null +++ b/drivers/net/ethernet/sfc/efx_channels.h @@ -0,0 +1,55 @@ +/* SPDX-License-Identifier: GPL-2.0-only */ +/**************************************************************************** + * Driver for Solarflare network controllers and boards + * Copyright 2018 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 + * by the Free Software Foundation, incorporated herein by reference. + */ + +#ifndef EFX_CHANNELS_H +#define EFX_CHANNELS_H + +int efx_probe_interrupts(struct efx_nic *efx); +void efx_remove_interrupts(struct efx_nic *efx); +int efx_soft_enable_interrupts(struct efx_nic *efx); +void efx_soft_disable_interrupts(struct efx_nic *efx); +int efx_enable_interrupts(struct efx_nic *efx); +void efx_disable_interrupts(struct efx_nic *efx); + +void efx_set_interrupt_affinity(struct efx_nic *efx); +void efx_clear_interrupt_affinity(struct efx_nic *efx); + +int efx_probe_eventq(struct efx_channel *channel); +int efx_init_eventq(struct efx_channel *channel); +void efx_start_eventq(struct efx_channel *channel); +void efx_stop_eventq(struct efx_channel *channel); +void efx_fini_eventq(struct efx_channel *channel); +void efx_remove_eventq(struct efx_channel *channel); + +struct efx_channel * +efx_alloc_channel(struct efx_nic *efx, int i, struct efx_channel *old_channel); +int efx_realloc_channels(struct efx_nic *efx, u32 rxq_entries, u32 txq_entries); +void efx_get_channel_name(struct efx_channel *channel, char *buf, size_t len); +void efx_set_channel_names(struct efx_nic *efx); +int efx_init_channels(struct efx_nic *efx); +int efx_probe_channels(struct efx_nic *efx); +int efx_set_channels(struct efx_nic *efx); +bool efx_default_channel_want_txqs(struct efx_channel *channel); +void efx_remove_channel(struct efx_channel *channel); +void efx_remove_channels(struct efx_nic *efx); +void efx_fini_channels(struct efx_nic *efx); +struct efx_channel *efx_copy_channel(const struct efx_channel *old_channel); +void efx_start_channels(struct efx_nic *efx); +void efx_stop_channels(struct efx_nic *efx); + +void efx_init_napi_channel(struct efx_channel *channel); +void efx_init_napi(struct efx_nic *efx); +void efx_fini_napi_channel(struct efx_channel *channel); +void efx_fini_napi(struct efx_nic *efx); + +int efx_channel_dummy_op_int(struct efx_channel *channel); +void efx_channel_dummy_op_void(struct efx_channel *channel); + +#endif diff --git a/drivers/net/ethernet/sfc/efx_common.c b/drivers/net/ethernet/sfc/efx_common.c new file mode 100644 index 000000000000..ab0ce62f81c1 --- /dev/null +++ b/drivers/net/ethernet/sfc/efx_common.c @@ -0,0 +1,1102 @@ +// SPDX-License-Identifier: GPL-2.0-only +/**************************************************************************** + * Driver for Solarflare network controllers and boards + * Copyright 2018 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 + * by the Free Software Foundation, incorporated herein by reference. + */ + +#include "net_driver.h" +#include <linux/module.h> +#include <linux/netdevice.h> +#include "efx_common.h" +#include "efx_channels.h" +#include "efx.h" +#include "mcdi.h" +#include "selftest.h" +#include "rx_common.h" +#include "tx_common.h" +#include "nic.h" +#include "io.h" +#include "mcdi_pcol.h" + +static unsigned int debug = (NETIF_MSG_DRV | NETIF_MSG_PROBE | + NETIF_MSG_LINK | NETIF_MSG_IFDOWN | + NETIF_MSG_IFUP | NETIF_MSG_RX_ERR | + NETIF_MSG_TX_ERR | NETIF_MSG_HW); +module_param(debug, uint, 0); +MODULE_PARM_DESC(debug, "Bitmapped debugging message enable value"); + +/* This is the time (in jiffies) between invocations of the hardware + * 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; + +/* 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 + +/* Default stats update time */ +#define STATS_PERIOD_MS_DEFAULT 1000 + +const unsigned int efx_reset_type_max = RESET_TYPE_MAX; +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_DATAPATH] = "DATAPATH", + [RESET_TYPE_MC_BIST] = "MC_BIST", + [RESET_TYPE_DISABLE] = "DISABLE", + [RESET_TYPE_TX_WATCHDOG] = "TX_WATCHDOG", + [RESET_TYPE_INT_ERROR] = "INT_ERROR", + [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 RESET_TYPE(type) \ + STRING_TABLE_LOOKUP(type, efx_reset_type) + +/* Loopback mode names (see LOOPBACK_MODE()) */ +const unsigned int efx_loopback_mode_max = LOOPBACK_MAX; +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_XGBR] = "XGBR", + [LOOPBACK_XFI] = "XFI", + [LOOPBACK_XAUI_FAR] = "XAUI_FAR", + [LOOPBACK_GMII_FAR] = "GMII_FAR", + [LOOPBACK_SGMII_FAR] = "SGMII_FAR", + [LOOPBACK_XFI_FAR] = "XFI_FAR", + [LOOPBACK_GPHY] = "GPHY", + [LOOPBACK_PHYXS] = "PHYXS", + [LOOPBACK_PCS] = "PCS", + [LOOPBACK_PMAPMD] = "PMA/PMD", + [LOOPBACK_XPORT] = "XPORT", + [LOOPBACK_XGMII_WS] = "XGMII_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_XFI_WS] = "XFI_WS", + [LOOPBACK_XFI_WS_FAR] = "XFI_WS_FAR", + [LOOPBACK_PHYXS_WS] = "PHYXS_WS", +}; + +/* 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; + +int efx_create_reset_workqueue(void) +{ + reset_workqueue = create_singlethread_workqueue("sfc_reset"); + if (!reset_workqueue) { + printk(KERN_ERR "Failed to create reset workqueue\n"); + return -ENOMEM; + } + + return 0; +} + +void efx_queue_reset_work(struct efx_nic *efx) +{ + queue_work(reset_workqueue, &efx->reset_work); +} + +void efx_flush_reset_workqueue(struct efx_nic *efx) +{ + cancel_work_sync(&efx->reset_work); +} + +void efx_destroy_reset_workqueue(void) +{ + if (reset_workqueue) { + destroy_workqueue(reset_workqueue); + reset_workqueue = NULL; + } +} + +/* We assume that efx->type->reconfigure_mac will always try to sync RX + * filters and therefore needs to read-lock the filter table against freeing + */ +void efx_mac_reconfigure(struct efx_nic *efx) +{ + if (efx->type->reconfigure_mac) { + down_read(&efx->filter_sem); + efx->type->reconfigure_mac(efx); + up_read(&efx->filter_sem); + } +} + +/* Asynchronous work item for changing MAC promiscuity and multicast + * hash. Avoid a drain/rx_ingress enable by reconfiguring the current + * MAC directly. + */ +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_mac_reconfigure(efx); + mutex_unlock(&efx->mac_lock); +} + +/* This ensures that the kernel is kept informed (via + * netif_carrier_on/off) of the link status, and also maintains the + * link status's stop on the port's TX queue. + */ +void efx_link_status_changed(struct efx_nic *efx) +{ + struct efx_link_state *link_state = &efx->link_state; + + /* SFC Bug 5356: A net_dev notifier is registered, so we must ensure + * that no events are triggered between unregister_netdev() and the + * driver unloading. A more general condition is that NETDEV_CHANGE + * can only be generated between NETDEV_UP and NETDEV_DOWN + */ + if (!netif_running(efx->net_dev)) + return; + + if (link_state->up != netif_carrier_ok(efx->net_dev)) { + efx->n_link_state_changes++; + + if (link_state->up) + netif_carrier_on(efx->net_dev); + else + netif_carrier_off(efx->net_dev); + } + + /* Status message for kernel log */ + if (link_state->up) + netif_info(efx, link, efx->net_dev, + "link up at %uMbps %s-duplex (MTU %d)\n", + link_state->speed, link_state->fd ? "full" : "half", + efx->net_dev->mtu); + else + netif_info(efx, link, efx->net_dev, "link down\n"); +} + +unsigned int efx_xdp_max_mtu(struct efx_nic *efx) +{ + /* The maximum MTU that we can fit in a single page, allowing for + * framing, overhead and XDP headroom. + */ + int overhead = EFX_MAX_FRAME_LEN(0) + sizeof(struct efx_rx_page_state) + + efx->rx_prefix_size + efx->type->rx_buffer_padding + + efx->rx_ip_align + XDP_PACKET_HEADROOM; + + return PAGE_SIZE - overhead; +} + +/* Context: process, rtnl_lock() held. */ +int efx_change_mtu(struct net_device *net_dev, int new_mtu) +{ + struct efx_nic *efx = netdev_priv(net_dev); + int rc; + + rc = efx_check_disabled(efx); + if (rc) + return rc; + + if (rtnl_dereference(efx->xdp_prog) && + new_mtu > efx_xdp_max_mtu(efx)) { + netif_err(efx, drv, efx->net_dev, + "Requested MTU of %d too big for XDP (max: %d)\n", + new_mtu, efx_xdp_max_mtu(efx)); + return -EINVAL; + } + + netif_dbg(efx, drv, efx->net_dev, "changing MTU to %d\n", new_mtu); + + efx_device_detach_sync(efx); + efx_stop_all(efx); + + mutex_lock(&efx->mac_lock); + net_dev->mtu = new_mtu; + efx_mac_reconfigure(efx); + mutex_unlock(&efx->mac_lock); + + efx_start_all(efx); + efx_device_attach_if_not_resetting(efx); + return 0; +} + +/************************************************************************** + * + * Hardware monitor + * + **************************************************************************/ + +/* Run periodically off the general workqueue */ +static void efx_monitor(struct work_struct *data) +{ + struct efx_nic *efx = container_of(data, struct efx_nic, + monitor_work.work); + + netif_vdbg(efx, timer, efx->net_dev, + "hardware monitor executing on CPU %d\n", + raw_smp_processor_id()); + BUG_ON(efx->type->monitor == NULL); + + /* If the mac_lock is already held then it is likely a port + * reconfiguration is already in place, which will likely do + * most of the work of monitor() anyway. + */ + if (mutex_trylock(&efx->mac_lock)) { + if (efx->port_enabled && efx->type->monitor) + efx->type->monitor(efx); + mutex_unlock(&efx->mac_lock); + } + + efx_start_monitor(efx); +} + +void efx_start_monitor(struct efx_nic *efx) +{ + if (efx->type->monitor) + queue_delayed_work(efx->workqueue, &efx->monitor_work, + efx_monitor_interval); +} + +/************************************************************************** + * + * Event queue processing + * + *************************************************************************/ + +/* Channels are shutdown and reinitialised whilst the NIC is running + * to propagate configuration changes (mtu, checksum offload), or + * to clear hardware error conditions + */ +static void efx_start_datapath(struct efx_nic *efx) +{ + netdev_features_t old_features = efx->net_dev->features; + bool old_rx_scatter = efx->rx_scatter; + 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_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) + XDP_PACKET_HEADROOM + + 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); + + /* Restore previously fixed features in hw_features and remove + * features which are fixed now + */ + efx->net_dev->hw_features |= efx->net_dev->features; + efx->net_dev->hw_features &= ~efx->fixed_features; + efx->net_dev->features |= efx->fixed_features; + if (efx->net_dev->features != old_features) + netdev_features_change(efx->net_dev); + + /* RX filters may also have scatter-enabled flags */ + if ((efx->rx_scatter != old_rx_scatter) && + efx->type->filter_update_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_start_channels(efx); + + efx_ptp_start_datapath(efx); + + if (netif_device_present(efx->net_dev)) + netif_tx_wake_all_queues(efx->net_dev); +} + +static void efx_stop_datapath(struct efx_nic *efx) +{ + EFX_ASSERT_RESET_SERIALISED(efx); + BUG_ON(efx->port_enabled); + + efx_ptp_stop_datapath(efx); + + efx_stop_channels(efx); +} + +/************************************************************************** + * + * Port handling + * + **************************************************************************/ + +static void efx_start_port(struct efx_nic *efx) +{ + netif_dbg(efx, ifup, efx->net_dev, "start port\n"); + BUG_ON(efx->port_enabled); + + mutex_lock(&efx->mac_lock); + efx->port_enabled = true; + + /* Ensure MAC ingress/egress is enabled */ + efx_mac_reconfigure(efx); + + mutex_unlock(&efx->mac_lock); +} + +/* 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() */ + 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); +} + +/* 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. + */ +void efx_start_all(struct efx_nic *efx) +{ + 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 || !netif_running(efx->net_dev) || + efx->reset_pending) + return; + + efx_start_port(efx); + efx_start_datapath(efx); + + /* Start the hardware monitor if there is one */ + efx_start_monitor(efx); + + /* Link state detection is normally event-driven; we have + * to poll now because we could have missed a change + */ + mutex_lock(&efx->mac_lock); + if (efx->phy_op->poll(efx)) + efx_link_status_changed(efx); + mutex_unlock(&efx->mac_lock); + + if (efx->type->start_stats) { + 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); + } +} + +/* 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. + */ +void efx_stop_all(struct efx_nic *efx) +{ + EFX_ASSERT_RESET_SERIALISED(efx); + + /* port_enabled can be read safely under the rtnl lock */ + if (!efx->port_enabled) + return; + + if (efx->type->update_stats) { + /* 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); + } + + efx_stop_port(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); + + efx_stop_datapath(efx); +} + +/* Context: process, dev_base_lock or RTNL held, non-blocking. */ +void efx_net_stats(struct net_device *net_dev, struct rtnl_link_stats64 *stats) +{ + struct efx_nic *efx = netdev_priv(net_dev); + + spin_lock_bh(&efx->stats_lock); + efx->type->update_stats(efx, NULL, stats); + spin_unlock_bh(&efx->stats_lock); +} + +/* Push loopback/power/transmit disable settings to the PHY, and reconfigure + * the MAC appropriately. All other PHY configuration changes are pushed + * through phy_op->set_settings(), and pushed asynchronously to the MAC + * through efx_monitor(). + * + * Callers must hold the mac_lock + */ +int __efx_reconfigure_port(struct efx_nic *efx) +{ + enum efx_phy_mode phy_mode; + int rc = 0; + + WARN_ON(!mutex_is_locked(&efx->mac_lock)); + + /* Disable PHY transmit in mac level loopbacks */ + phy_mode = efx->phy_mode; + if (LOOPBACK_INTERNAL(efx)) + efx->phy_mode |= PHY_MODE_TX_DISABLED; + else + efx->phy_mode &= ~PHY_MODE_TX_DISABLED; + + if (efx->type->reconfigure_port) + rc = efx->type->reconfigure_port(efx); + + if (rc) + efx->phy_mode = phy_mode; + + return rc; +} + +/* Reinitialise the MAC to pick up new PHY settings, even if the port is + * disabled. + */ +int efx_reconfigure_port(struct efx_nic *efx) +{ + int rc; + + EFX_ASSERT_RESET_SERIALISED(efx); + + mutex_lock(&efx->mac_lock); + rc = __efx_reconfigure_port(efx); + mutex_unlock(&efx->mac_lock); + + return rc; +} + +/************************************************************************** + * + * Device reset and suspend + * + **************************************************************************/ + +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; +} + +/* 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 = pci_dev_to_eeh_dev(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; +} + +/* Tears down the entire software state and most of the hardware state + * before reset. + */ +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); + efx_disable_interrupts(efx); + + mutex_lock(&efx->mac_lock); + down_write(&efx->filter_sem); + mutex_lock(&efx->rss_lock); + if (efx->port_initialized && method != RESET_TYPE_INVISIBLE && + method != RESET_TYPE_DATAPATH) + efx->phy_op->fini(efx); + efx->type->fini(efx); +} + +/* This function will always ensure that the locks acquired in + * efx_reset_down() are released. A failure return code indicates + * that we were unable to reinitialise the hardware, and the + * driver should be disabled. If ok is false, then the rx and tx + * engines are not restarted, pending a RESET_DISABLE. + */ +int efx_reset_up(struct efx_nic *efx, enum reset_type method, bool ok) +{ + int rc; + + 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"); + goto fail; + } + + if (!ok) + goto fail; + + if (efx->port_initialized && method != RESET_TYPE_INVISIBLE && + method != RESET_TYPE_DATAPATH) { + rc = efx->phy_op->init(efx); + if (rc) + goto fail; + rc = efx->phy_op->reconfigure(efx); + if (rc && rc != -EPERM) + netif_err(efx, drv, efx->net_dev, + "could not restore PHY settings\n"); + } + + rc = efx_enable_interrupts(efx); + if (rc) + goto fail; + +#ifdef CONFIG_SFC_SRIOV + rc = efx->type->vswitching_restore(efx); + if (rc) /* not fatal; the PF will still work fine */ + netif_warn(efx, probe, efx->net_dev, + "failed to restore vswitching rc=%d;" + " VFs may not function\n", rc); +#endif + + if (efx->type->rx_restore_rss_contexts) + efx->type->rx_restore_rss_contexts(efx); + mutex_unlock(&efx->rss_lock); + efx->type->filter_table_restore(efx); + up_write(&efx->filter_sem); + if (efx->type->sriov_reset) + efx->type->sriov_reset(efx); + + mutex_unlock(&efx->mac_lock); + + efx_start_all(efx); + + if (efx->type->udp_tnl_push_ports) + efx->type->udp_tnl_push_ports(efx); + + return 0; + +fail: + efx->port_initialized = false; + + mutex_unlock(&efx->rss_lock); + up_write(&efx->filter_sem); + mutex_unlock(&efx->mac_lock); + + return rc; +} + +/* Reset the NIC using the specified method. Note that the reset may + * fail, in which case the card will be left in an unusable state. + * + * Caller must hold the rtnl_lock. + */ +int efx_reset(struct efx_nic *efx, enum reset_type method) +{ + bool disabled; + int rc, rc2; + + netif_info(efx, drv, efx->net_dev, "resetting (%s)\n", + RESET_TYPE(method)); + + efx_device_detach_sync(efx); + efx_reset_down(efx, method); + + rc = efx->type->reset(efx, method); + if (rc) { + netif_err(efx, drv, efx->net_dev, "failed to reset hardware\n"); + goto out; + } + + /* Clear flags for the scopes we covered. We assume the NIC and + * driver are now quiescent so that there is no race here. + */ + 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 + * RESET_TYPE_DISABLE since this driver generally assumes the hardware + * can respond to requests. + */ + pci_set_master(efx->pci_dev); + +out: + /* Leave device stopped if necessary */ + disabled = rc || + method == RESET_TYPE_DISABLE || + method == RESET_TYPE_RECOVER_OR_DISABLE; + rc2 = efx_reset_up(efx, method, !disabled); + if (rc2) { + disabled = true; + if (!rc) + rc = rc2; + } + + if (disabled) { + dev_close(efx->net_dev); + netif_err(efx, drv, efx->net_dev, "has been disabled\n"); + efx->state = STATE_DISABLED; + } else { + netif_dbg(efx, drv, efx->net_dev, "reset complete\n"); + efx_device_attach_if_not_resetting(efx); + } + return rc; +} + +/* 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; + enum reset_type method; + + pending = READ_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 (!pending) + return; + + rtnl_lock(); + + /* 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(); +} + +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_DATAPATH: + 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)); + break; + default: + method = efx->type->map_reset_reason(type); + netif_dbg(efx, drv, efx->net_dev, + "scheduling %s reset for %s\n", + RESET_TYPE(method), RESET_TYPE(type)); + break; + } + + 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 (READ_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. + */ + efx_mcdi_mode_poll(efx); + + efx_queue_reset_work(efx); +} + +/************************************************************************** + * + * Dummy PHY/MAC operations + * + * Can be used for some unimplemented operations + * Needed so all function pointers are valid and do not have to be tested + * before use + * + **************************************************************************/ +int efx_port_dummy_op_int(struct efx_nic *efx) +{ + return 0; +} +void efx_port_dummy_op_void(struct efx_nic *efx) {} + +static bool efx_port_dummy_op_poll(struct efx_nic *efx) +{ + return false; +} + +static const struct efx_phy_operations efx_dummy_phy_operations = { + .init = efx_port_dummy_op_int, + .reconfigure = efx_port_dummy_op_int, + .poll = efx_port_dummy_op_poll, + .fini = efx_port_dummy_op_void, +}; + +/************************************************************************** + * + * Data housekeeping + * + **************************************************************************/ + +/* This zeroes out and then fills in the invariants in a struct + * efx_nic (including all sub-structures). + */ +int efx_init_struct(struct efx_nic *efx, + struct pci_dev *pci_dev, struct net_device *net_dev) +{ + int rc = -ENOMEM; + + /* Initialise common structures */ + 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); + efx_selftest_async_init(efx); + efx->pci_dev = pci_dev; + efx->msg_enable = debug; + 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; + INIT_LIST_HEAD(&efx->rss_context.list); + mutex_init(&efx->rss_lock); + spin_lock_init(&efx->stats_lock); + efx->vi_stride = EFX_DEFAULT_VI_STRIDE; + efx->num_mac_stats = MC_CMD_MAC_NSTATS; + BUILD_BUG_ON(MC_CMD_MAC_NSTATS - 1 != MC_CMD_MAC_GENERATION_END); + mutex_init(&efx->mac_lock); +#ifdef CONFIG_RFS_ACCEL + mutex_init(&efx->rps_mutex); + spin_lock_init(&efx->rps_hash_lock); + /* Failure to allocate is not fatal, but may degrade ARFS performance */ + efx->rps_hash_table = kcalloc(EFX_ARFS_HASH_TABLE_SIZE, + sizeof(*efx->rps_hash_table), GFP_KERNEL); +#endif + 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); + + rc = efx_init_channels(efx); + if (rc) + goto fail; + + /* Would be good to use the net_dev name, but we're too early */ + snprintf(efx->workqueue_name, sizeof(efx->workqueue_name), "sfc%s", + pci_name(pci_dev)); + efx->workqueue = create_singlethread_workqueue(efx->workqueue_name); + if (!efx->workqueue) { + rc = -ENOMEM; + goto fail; + } + + return 0; + +fail: + efx_fini_struct(efx); + return rc; +} + +void efx_fini_struct(struct efx_nic *efx) +{ +#ifdef CONFIG_RFS_ACCEL + kfree(efx->rps_hash_table); +#endif + + efx_fini_channels(efx); + + kfree(efx->vpd_sn); + + if (efx->workqueue) { + destroy_workqueue(efx->workqueue); + efx->workqueue = NULL; + } +} + +/* This configures the PCI device to enable I/O and DMA. */ +int efx_init_io(struct efx_nic *efx, int bar, dma_addr_t dma_mask, + unsigned int mem_map_size) +{ + struct pci_dev *pci_dev = efx->pci_dev; + int rc; + + netif_dbg(efx, probe, efx->net_dev, "initialising I/O\n"); + + rc = pci_enable_device(pci_dev); + if (rc) { + netif_err(efx, probe, efx->net_dev, + "failed to enable PCI device\n"); + goto fail1; + } + + pci_set_master(pci_dev); + + /* Set the PCI DMA mask. Try all possibilities from our + * genuine mask down to 32 bits, because some architectures + * (e.g. x86_64 with iommu_sac_force set) will allow 40 bit + * masks event though they reject 46 bit masks. + */ + while (dma_mask > 0x7fffffffUL) { + rc = dma_set_mask_and_coherent(&pci_dev->dev, dma_mask); + if (rc == 0) + break; + dma_mask >>= 1; + } + if (rc) { + netif_err(efx, probe, efx->net_dev, + "could not find a suitable DMA mask\n"); + goto fail2; + } + netif_dbg(efx, probe, efx->net_dev, + "using DMA mask %llx\n", (unsigned long long)dma_mask); + + efx->membase_phys = pci_resource_start(efx->pci_dev, bar); + if (!efx->membase_phys) { + netif_err(efx, probe, efx->net_dev, + "ERROR: No BAR%d mapping from the BIOS. " + "Try pci=realloc on the kernel command line\n", bar); + rc = -ENODEV; + goto fail3; + } + + rc = pci_request_region(pci_dev, bar, "sfc"); + if (rc) { + netif_err(efx, probe, efx->net_dev, + "request for memory BAR failed\n"); + rc = -EIO; + goto fail3; + } + + 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, 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, mem_map_size, + efx->membase); + + return 0; + +fail4: + pci_release_region(efx->pci_dev, bar); +fail3: + efx->membase_phys = 0; +fail2: + pci_disable_device(efx->pci_dev); +fail1: + return rc; +} + +void efx_fini_io(struct efx_nic *efx, int bar) +{ + netif_dbg(efx, drv, efx->net_dev, "shutting down I/O\n"); + + if (efx->membase) { + iounmap(efx->membase); + efx->membase = NULL; + } + + if (efx->membase_phys) { + pci_release_region(efx->pci_dev, bar); + efx->membase_phys = 0; + } + + /* Don't disable bus-mastering if VFs are assigned */ + if (!pci_vfs_assigned(efx->pci_dev)) + pci_disable_device(efx->pci_dev); +} + +#ifdef CONFIG_SFC_MCDI_LOGGING +static ssize_t show_mcdi_log(struct device *dev, struct device_attribute *attr, + char *buf) +{ + struct efx_nic *efx = dev_get_drvdata(dev); + struct efx_mcdi_iface *mcdi = efx_mcdi(efx); + + return scnprintf(buf, PAGE_SIZE, "%d\n", mcdi->logging_enabled); +} + +static ssize_t set_mcdi_log(struct device *dev, struct device_attribute *attr, + const char *buf, size_t count) +{ + struct efx_nic *efx = dev_get_drvdata(dev); + struct efx_mcdi_iface *mcdi = efx_mcdi(efx); + bool enable = count > 0 && *buf != '0'; + + mcdi->logging_enabled = enable; + return count; +} + +static DEVICE_ATTR(mcdi_logging, 0644, show_mcdi_log, set_mcdi_log); + +void efx_init_mcdi_logging(struct efx_nic *efx) +{ + int rc = device_create_file(&efx->pci_dev->dev, &dev_attr_mcdi_logging); + + if (rc) { + netif_warn(efx, drv, efx->net_dev, + "failed to init net dev attributes\n"); + } +} + +void efx_fini_mcdi_logging(struct efx_nic *efx) +{ + device_remove_file(&efx->pci_dev->dev, &dev_attr_mcdi_logging); +} +#endif diff --git a/drivers/net/ethernet/sfc/efx_common.h b/drivers/net/ethernet/sfc/efx_common.h new file mode 100644 index 000000000000..fa2fc681e7f9 --- /dev/null +++ b/drivers/net/ethernet/sfc/efx_common.h @@ -0,0 +1,73 @@ +/* SPDX-License-Identifier: GPL-2.0-only */ +/**************************************************************************** + * Driver for Solarflare network controllers and boards + * Copyright 2018 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 + * by the Free Software Foundation, incorporated herein by reference. + */ + +#ifndef EFX_COMMON_H +#define EFX_COMMON_H + +int efx_init_io(struct efx_nic *efx, int bar, dma_addr_t dma_mask, + unsigned int mem_map_size); +void efx_fini_io(struct efx_nic *efx, int bar); +int efx_init_struct(struct efx_nic *efx, struct pci_dev *pci_dev, + struct net_device *net_dev); +void efx_fini_struct(struct efx_nic *efx); + +void efx_start_all(struct efx_nic *efx); +void efx_stop_all(struct efx_nic *efx); + +void efx_net_stats(struct net_device *net_dev, struct rtnl_link_stats64 *stats); + +int efx_create_reset_workqueue(void); +void efx_queue_reset_work(struct efx_nic *efx); +void efx_flush_reset_workqueue(struct efx_nic *efx); +void efx_destroy_reset_workqueue(void); + +void efx_start_monitor(struct efx_nic *efx); + +int __efx_reconfigure_port(struct efx_nic *efx); +int efx_reconfigure_port(struct efx_nic *efx); + +#define EFX_ASSERT_RESET_SERIALISED(efx) \ + do { \ + if ((efx->state == STATE_READY) || \ + (efx->state == STATE_RECOVERY) || \ + (efx->state == STATE_DISABLED)) \ + ASSERT_RTNL(); \ + } while (0) + +int efx_try_recovery(struct efx_nic *efx); +void efx_reset_down(struct efx_nic *efx, enum reset_type method); +int efx_reset_up(struct efx_nic *efx, enum reset_type method, bool ok); +int efx_reset(struct efx_nic *efx, enum reset_type method); +void efx_schedule_reset(struct efx_nic *efx, enum reset_type type); + +static inline 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; +} + +#ifdef CONFIG_SFC_MCDI_LOGGING +void efx_init_mcdi_logging(struct efx_nic *efx); +void efx_fini_mcdi_logging(struct efx_nic *efx); +#else +static inline void efx_init_mcdi_logging(struct efx_nic *efx) {} +static inline void efx_fini_mcdi_logging(struct efx_nic *efx) {} +#endif + +void efx_mac_reconfigure(struct efx_nic *efx); +void efx_link_status_changed(struct efx_nic *efx); +unsigned int efx_xdp_max_mtu(struct efx_nic *efx); +int efx_change_mtu(struct net_device *net_dev, int new_mtu); + +#endif diff --git a/drivers/net/ethernet/sfc/ethtool.c b/drivers/net/ethernet/sfc/ethtool.c index b31032da4bcb..993b5769525b 100644 --- a/drivers/net/ethernet/sfc/ethtool.c +++ b/drivers/net/ethernet/sfc/ethtool.c @@ -13,92 +13,13 @@ #include "workarounds.h" #include "selftest.h" #include "efx.h" +#include "efx_channels.h" +#include "rx_common.h" +#include "tx_common.h" +#include "ethtool_common.h" #include "filter.h" #include "nic.h" -struct efx_sw_stat_desc { - const char *name; - enum { - EFX_ETHTOOL_STAT_SOURCE_nic, - EFX_ETHTOOL_STAT_SOURCE_channel, - EFX_ETHTOOL_STAT_SOURCE_tx_queue - } source; - unsigned offset; - u64(*get_stat) (void *field); /* Reader function */ -}; - -/* Initialiser for a struct efx_sw_stat_desc with type-checking */ -#define EFX_ETHTOOL_STAT(stat_name, source_name, field, field_type, \ - get_stat_function) { \ - .name = #stat_name, \ - .source = EFX_ETHTOOL_STAT_SOURCE_##source_name, \ - .offset = ((((field_type *) 0) == \ - &((struct efx_##source_name *)0)->field) ? \ - offsetof(struct efx_##source_name, field) : \ - offsetof(struct efx_##source_name, field)), \ - .get_stat = get_stat_function, \ -} - -static u64 efx_get_uint_stat(void *field) -{ - return *(unsigned int *)field; -} - -static u64 efx_get_atomic_stat(void *field) -{ - return atomic_read((atomic_t *) field); -} - -#define EFX_ETHTOOL_ATOMIC_NIC_ERROR_STAT(field) \ - EFX_ETHTOOL_STAT(field, nic, field, \ - atomic_t, efx_get_atomic_stat) - -#define EFX_ETHTOOL_UINT_CHANNEL_STAT(field) \ - EFX_ETHTOOL_STAT(field, channel, n_##field, \ - unsigned int, efx_get_uint_stat) -#define EFX_ETHTOOL_UINT_CHANNEL_STAT_NO_N(field) \ - EFX_ETHTOOL_STAT(field, channel, field, \ - unsigned int, efx_get_uint_stat) - -#define EFX_ETHTOOL_UINT_TXQ_STAT(field) \ - EFX_ETHTOOL_STAT(tx_##field, tx_queue, field, \ - unsigned int, efx_get_uint_stat) - -static const struct efx_sw_stat_desc efx_sw_stat_desc[] = { - EFX_ETHTOOL_UINT_TXQ_STAT(merge_events), - EFX_ETHTOOL_UINT_TXQ_STAT(tso_bursts), - EFX_ETHTOOL_UINT_TXQ_STAT(tso_long_headers), - EFX_ETHTOOL_UINT_TXQ_STAT(tso_packets), - EFX_ETHTOOL_UINT_TXQ_STAT(tso_fallbacks), - EFX_ETHTOOL_UINT_TXQ_STAT(pushes), - EFX_ETHTOOL_UINT_TXQ_STAT(pio_packets), - EFX_ETHTOOL_UINT_TXQ_STAT(cb_packets), - EFX_ETHTOOL_ATOMIC_NIC_ERROR_STAT(rx_reset), - EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_tobe_disc), - EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_ip_hdr_chksum_err), - EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_tcp_udp_chksum_err), - EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_inner_ip_hdr_chksum_err), - EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_inner_tcp_udp_chksum_err), - EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_outer_ip_hdr_chksum_err), - EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_outer_tcp_udp_chksum_err), - EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_eth_crc_err), - EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_mcast_mismatch), - EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_frm_trunc), - EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_merge_events), - EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_merge_packets), - EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_xdp_drops), - EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_xdp_bad_drops), - EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_xdp_tx), - EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_xdp_redirect), -#ifdef CONFIG_RFS_ACCEL - EFX_ETHTOOL_UINT_CHANNEL_STAT_NO_N(rfs_filter_count), - EFX_ETHTOOL_UINT_CHANNEL_STAT(rfs_succeeded), - EFX_ETHTOOL_UINT_CHANNEL_STAT(rfs_failed), -#endif -}; - -#define EFX_ETHTOOL_SW_STAT_COUNT ARRAY_SIZE(efx_sw_stat_desc) - #define EFX_ETHTOOL_EEPROM_MAGIC 0xEFAB /************************************************************************** @@ -185,18 +106,6 @@ efx_ethtool_set_link_ksettings(struct net_device *net_dev, return rc; } -static void efx_ethtool_get_drvinfo(struct net_device *net_dev, - struct ethtool_drvinfo *info) -{ - struct efx_nic *efx = netdev_priv(net_dev); - - strlcpy(info->driver, KBUILD_MODNAME, sizeof(info->driver)); - strlcpy(info->version, EFX_DRIVER_VERSION, sizeof(info->version)); - efx_mcdi_print_fwver(efx, info->fw_version, - sizeof(info->fw_version)); - strlcpy(info->bus_info, pci_name(efx->pci_dev), sizeof(info->bus_info)); -} - static int efx_ethtool_get_regs_len(struct net_device *net_dev) { return efx_nic_get_regs_len(netdev_priv(net_dev)); @@ -211,341 +120,6 @@ static void efx_ethtool_get_regs(struct net_device *net_dev, efx_nic_get_regs(efx, buf); } -static u32 efx_ethtool_get_msglevel(struct net_device *net_dev) -{ - struct efx_nic *efx = netdev_priv(net_dev); - return efx->msg_enable; -} - -static void efx_ethtool_set_msglevel(struct net_device *net_dev, u32 msg_enable) -{ - struct efx_nic *efx = netdev_priv(net_dev); - efx->msg_enable = msg_enable; -} - -/** - * efx_fill_test - fill in an individual self-test entry - * @test_index: Index of the test - * @strings: Ethtool strings, or %NULL - * @data: Ethtool test results, or %NULL - * @test: Pointer to test result (used only if data != %NULL) - * @unit_format: Unit name format (e.g. "chan\%d") - * @unit_id: Unit id (e.g. 0 for "chan0") - * @test_format: Test name format (e.g. "loopback.\%s.tx.sent") - * @test_id: Test id (e.g. "PHYXS" for "loopback.PHYXS.tx_sent") - * - * Fill in an individual self-test entry. - */ -static void efx_fill_test(unsigned int test_index, u8 *strings, u64 *data, - int *test, const char *unit_format, int unit_id, - const char *test_format, const char *test_id) -{ - char unit_str[ETH_GSTRING_LEN], test_str[ETH_GSTRING_LEN]; - - /* Fill data value, if applicable */ - if (data) - data[test_index] = *test; - - /* Fill string, if applicable */ - if (strings) { - if (strchr(unit_format, '%')) - snprintf(unit_str, sizeof(unit_str), - unit_format, unit_id); - else - strcpy(unit_str, unit_format); - snprintf(test_str, sizeof(test_str), test_format, test_id); - snprintf(strings + test_index * ETH_GSTRING_LEN, - ETH_GSTRING_LEN, - "%-6s %-24s", unit_str, test_str); - } -} - -#define EFX_CHANNEL_NAME(_channel) "chan%d", _channel->channel -#define EFX_TX_QUEUE_NAME(_tx_queue) "txq%d", _tx_queue->queue -#define EFX_RX_QUEUE_NAME(_rx_queue) "rxq%d", _rx_queue->queue -#define EFX_LOOPBACK_NAME(_mode, _counter) \ - "loopback.%s." _counter, STRING_TABLE_LOOKUP(_mode, efx_loopback_mode) - -/** - * efx_fill_loopback_test - fill in a block of loopback self-test entries - * @efx: Efx NIC - * @lb_tests: Efx loopback self-test results structure - * @mode: Loopback test mode - * @test_index: Starting index of the test - * @strings: Ethtool strings, or %NULL - * @data: Ethtool test results, or %NULL - * - * Fill in a block of loopback self-test entries. Return new test - * index. - */ -static int efx_fill_loopback_test(struct efx_nic *efx, - struct efx_loopback_self_tests *lb_tests, - enum efx_loopback_mode mode, - unsigned int test_index, - u8 *strings, u64 *data) -{ - struct efx_channel *channel = - efx_get_channel(efx, efx->tx_channel_offset); - struct efx_tx_queue *tx_queue; - - efx_for_each_channel_tx_queue(tx_queue, channel) { - efx_fill_test(test_index++, strings, data, - &lb_tests->tx_sent[tx_queue->queue], - EFX_TX_QUEUE_NAME(tx_queue), - EFX_LOOPBACK_NAME(mode, "tx_sent")); - efx_fill_test(test_index++, strings, data, - &lb_tests->tx_done[tx_queue->queue], - EFX_TX_QUEUE_NAME(tx_queue), - EFX_LOOPBACK_NAME(mode, "tx_done")); - } - efx_fill_test(test_index++, strings, data, - &lb_tests->rx_good, - "rx", 0, - EFX_LOOPBACK_NAME(mode, "rx_good")); - efx_fill_test(test_index++, strings, data, - &lb_tests->rx_bad, - "rx", 0, - EFX_LOOPBACK_NAME(mode, "rx_bad")); - - return test_index; -} - -/** - * efx_ethtool_fill_self_tests - get self-test details - * @efx: Efx NIC - * @tests: Efx self-test results structure, or %NULL - * @strings: Ethtool strings, or %NULL - * @data: Ethtool test results, or %NULL - * - * Get self-test number of strings, strings, and/or test results. - * Return number of strings (== number of test results). - * - * The reason for merging these three functions is to make sure that - * they can never be inconsistent. - */ -static int efx_ethtool_fill_self_tests(struct efx_nic *efx, - struct efx_self_tests *tests, - u8 *strings, u64 *data) -{ - struct efx_channel *channel; - unsigned int n = 0, i; - enum efx_loopback_mode mode; - - efx_fill_test(n++, strings, data, &tests->phy_alive, - "phy", 0, "alive", NULL); - efx_fill_test(n++, strings, data, &tests->nvram, - "core", 0, "nvram", NULL); - efx_fill_test(n++, strings, data, &tests->interrupt, - "core", 0, "interrupt", NULL); - - /* Event queues */ - efx_for_each_channel(channel, efx) { - efx_fill_test(n++, strings, data, - &tests->eventq_dma[channel->channel], - EFX_CHANNEL_NAME(channel), - "eventq.dma", NULL); - efx_fill_test(n++, strings, data, - &tests->eventq_int[channel->channel], - EFX_CHANNEL_NAME(channel), - "eventq.int", NULL); - } - - efx_fill_test(n++, strings, data, &tests->memory, - "core", 0, "memory", NULL); - efx_fill_test(n++, strings, data, &tests->registers, - "core", 0, "registers", NULL); - - if (efx->phy_op->run_tests != NULL) { - EFX_WARN_ON_PARANOID(efx->phy_op->test_name == NULL); - - for (i = 0; true; ++i) { - const char *name; - - EFX_WARN_ON_PARANOID(i >= EFX_MAX_PHY_TESTS); - name = efx->phy_op->test_name(efx, i); - if (name == NULL) - break; - - efx_fill_test(n++, strings, data, &tests->phy_ext[i], - "phy", 0, name, NULL); - } - } - - /* Loopback tests */ - for (mode = LOOPBACK_NONE; mode <= LOOPBACK_TEST_MAX; mode++) { - if (!(efx->loopback_modes & (1 << mode))) - continue; - n = efx_fill_loopback_test(efx, - &tests->loopback[mode], mode, n, - strings, data); - } - - return n; -} - -static size_t efx_describe_per_queue_stats(struct efx_nic *efx, u8 *strings) -{ - size_t n_stats = 0; - struct efx_channel *channel; - - efx_for_each_channel(channel, efx) { - if (efx_channel_has_tx_queues(channel)) { - n_stats++; - if (strings != NULL) { - snprintf(strings, ETH_GSTRING_LEN, - "tx-%u.tx_packets", - channel->tx_queue[0].queue / - EFX_TXQ_TYPES); - - strings += ETH_GSTRING_LEN; - } - } - } - efx_for_each_channel(channel, efx) { - if (efx_channel_has_rx_queue(channel)) { - n_stats++; - if (strings != NULL) { - snprintf(strings, ETH_GSTRING_LEN, - "rx-%d.rx_packets", channel->channel); - strings += ETH_GSTRING_LEN; - } - } - } - if (efx->xdp_tx_queue_count && efx->xdp_tx_queues) { - unsigned short xdp; - - for (xdp = 0; xdp < efx->xdp_tx_queue_count; xdp++) { - n_stats++; - if (strings) { - snprintf(strings, ETH_GSTRING_LEN, - "tx-xdp-cpu-%hu.tx_packets", xdp); - strings += ETH_GSTRING_LEN; - } - } - } - - return n_stats; -} - -static int efx_ethtool_get_sset_count(struct net_device *net_dev, - int string_set) -{ - struct efx_nic *efx = netdev_priv(net_dev); - - switch (string_set) { - case ETH_SS_STATS: - return efx->type->describe_stats(efx, NULL) + - EFX_ETHTOOL_SW_STAT_COUNT + - efx_describe_per_queue_stats(efx, NULL) + - efx_ptp_describe_stats(efx, NULL); - case ETH_SS_TEST: - return efx_ethtool_fill_self_tests(efx, NULL, NULL, NULL); - default: - return -EINVAL; - } -} - -static void efx_ethtool_get_strings(struct net_device *net_dev, - u32 string_set, u8 *strings) -{ - struct efx_nic *efx = netdev_priv(net_dev); - int i; - - switch (string_set) { - case ETH_SS_STATS: - strings += (efx->type->describe_stats(efx, strings) * - ETH_GSTRING_LEN); - for (i = 0; i < EFX_ETHTOOL_SW_STAT_COUNT; i++) - strlcpy(strings + i * ETH_GSTRING_LEN, - efx_sw_stat_desc[i].name, ETH_GSTRING_LEN); - strings += EFX_ETHTOOL_SW_STAT_COUNT * ETH_GSTRING_LEN; - strings += (efx_describe_per_queue_stats(efx, strings) * - ETH_GSTRING_LEN); - efx_ptp_describe_stats(efx, strings); - break; - case ETH_SS_TEST: - efx_ethtool_fill_self_tests(efx, NULL, strings, NULL); - break; - default: - /* No other string sets */ - break; - } -} - -static void efx_ethtool_get_stats(struct net_device *net_dev, - struct ethtool_stats *stats, - u64 *data) -{ - struct efx_nic *efx = netdev_priv(net_dev); - const struct efx_sw_stat_desc *stat; - struct efx_channel *channel; - struct efx_tx_queue *tx_queue; - struct efx_rx_queue *rx_queue; - int i; - - spin_lock_bh(&efx->stats_lock); - - /* Get NIC statistics */ - data += efx->type->update_stats(efx, data, NULL); - - /* Get software statistics */ - for (i = 0; i < EFX_ETHTOOL_SW_STAT_COUNT; i++) { - stat = &efx_sw_stat_desc[i]; - switch (stat->source) { - case EFX_ETHTOOL_STAT_SOURCE_nic: - data[i] = stat->get_stat((void *)efx + stat->offset); - break; - case EFX_ETHTOOL_STAT_SOURCE_channel: - data[i] = 0; - efx_for_each_channel(channel, efx) - data[i] += stat->get_stat((void *)channel + - stat->offset); - break; - case EFX_ETHTOOL_STAT_SOURCE_tx_queue: - data[i] = 0; - efx_for_each_channel(channel, efx) { - efx_for_each_channel_tx_queue(tx_queue, channel) - data[i] += - stat->get_stat((void *)tx_queue - + stat->offset); - } - break; - } - } - data += EFX_ETHTOOL_SW_STAT_COUNT; - - spin_unlock_bh(&efx->stats_lock); - - efx_for_each_channel(channel, efx) { - if (efx_channel_has_tx_queues(channel)) { - *data = 0; - efx_for_each_channel_tx_queue(tx_queue, channel) { - *data += tx_queue->tx_packets; - } - data++; - } - } - efx_for_each_channel(channel, efx) { - if (efx_channel_has_rx_queue(channel)) { - *data = 0; - efx_for_each_channel_rx_queue(rx_queue, channel) { - *data += rx_queue->rx_packets; - } - data++; - } - } - if (efx->xdp_tx_queue_count && efx->xdp_tx_queues) { - int xdp; - - for (xdp = 0; xdp < efx->xdp_tx_queue_count; xdp++) { - data[0] = efx->xdp_tx_queues[xdp]->tx_packets; - data++; - } - } - - efx_ptp_update_stats(efx, data); -} - static void efx_ethtool_self_test(struct net_device *net_dev, struct ethtool_test *test, u64 *data) { @@ -787,16 +361,6 @@ out: return rc; } -static void efx_ethtool_get_pauseparam(struct net_device *net_dev, - struct ethtool_pauseparam *pause) -{ - struct efx_nic *efx = netdev_priv(net_dev); - - pause->rx_pause = !!(efx->wanted_fc & EFX_FC_RX); - pause->tx_pause = !!(efx->wanted_fc & EFX_FC_TX); - pause->autoneg = !!(efx->wanted_fc & EFX_FC_AUTO); -} - static void efx_ethtool_get_wol(struct net_device *net_dev, struct ethtool_wolinfo *wol) { @@ -1456,7 +1020,7 @@ static int efx_ethtool_set_rxfh_context(struct net_device *net_dev, rc = -ENOMEM; goto out_unlock; } - ctx->context_id = EFX_EF10_RSS_CONTEXT_INVALID; + ctx->context_id = EFX_MCDI_RSS_CONTEXT_INVALID; /* Initialise indir table and key to defaults */ efx_set_default_rx_indir_table(efx, ctx); netdev_rss_key_fill(ctx->rx_hash_key, sizeof(ctx->rx_hash_key)); diff --git a/drivers/net/ethernet/sfc/ethtool_common.c b/drivers/net/ethernet/sfc/ethtool_common.c new file mode 100644 index 000000000000..b8d281ab6c7a --- /dev/null +++ b/drivers/net/ethernet/sfc/ethtool_common.c @@ -0,0 +1,457 @@ +// SPDX-License-Identifier: GPL-2.0-only +/**************************************************************************** + * Driver for Solarflare network controllers and boards + * Copyright 2019 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 + * by the Free Software Foundation, incorporated herein by reference. + */ +#include <linux/module.h> +#include <linux/netdevice.h> +#include "net_driver.h" +#include "mcdi.h" +#include "nic.h" +#include "selftest.h" +#include "ethtool_common.h" + +struct efx_sw_stat_desc { + const char *name; + enum { + EFX_ETHTOOL_STAT_SOURCE_nic, + EFX_ETHTOOL_STAT_SOURCE_channel, + EFX_ETHTOOL_STAT_SOURCE_tx_queue + } source; + unsigned int offset; + u64 (*get_stat)(void *field); /* Reader function */ +}; + +/* Initialiser for a struct efx_sw_stat_desc with type-checking */ +#define EFX_ETHTOOL_STAT(stat_name, source_name, field, field_type, \ + get_stat_function) { \ + .name = #stat_name, \ + .source = EFX_ETHTOOL_STAT_SOURCE_##source_name, \ + .offset = ((((field_type *) 0) == \ + &((struct efx_##source_name *)0)->field) ? \ + offsetof(struct efx_##source_name, field) : \ + offsetof(struct efx_##source_name, field)), \ + .get_stat = get_stat_function, \ +} + +static u64 efx_get_uint_stat(void *field) +{ + return *(unsigned int *)field; +} + +static u64 efx_get_atomic_stat(void *field) +{ + return atomic_read((atomic_t *) field); +} + +#define EFX_ETHTOOL_ATOMIC_NIC_ERROR_STAT(field) \ + EFX_ETHTOOL_STAT(field, nic, field, \ + atomic_t, efx_get_atomic_stat) + +#define EFX_ETHTOOL_UINT_CHANNEL_STAT(field) \ + EFX_ETHTOOL_STAT(field, channel, n_##field, \ + unsigned int, efx_get_uint_stat) +#define EFX_ETHTOOL_UINT_CHANNEL_STAT_NO_N(field) \ + EFX_ETHTOOL_STAT(field, channel, field, \ + unsigned int, efx_get_uint_stat) + +#define EFX_ETHTOOL_UINT_TXQ_STAT(field) \ + EFX_ETHTOOL_STAT(tx_##field, tx_queue, field, \ + unsigned int, efx_get_uint_stat) + +static const struct efx_sw_stat_desc efx_sw_stat_desc[] = { + EFX_ETHTOOL_UINT_TXQ_STAT(merge_events), + EFX_ETHTOOL_UINT_TXQ_STAT(tso_bursts), + EFX_ETHTOOL_UINT_TXQ_STAT(tso_long_headers), + EFX_ETHTOOL_UINT_TXQ_STAT(tso_packets), + EFX_ETHTOOL_UINT_TXQ_STAT(tso_fallbacks), + EFX_ETHTOOL_UINT_TXQ_STAT(pushes), + EFX_ETHTOOL_UINT_TXQ_STAT(pio_packets), + EFX_ETHTOOL_UINT_TXQ_STAT(cb_packets), + EFX_ETHTOOL_ATOMIC_NIC_ERROR_STAT(rx_reset), + EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_tobe_disc), + EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_ip_hdr_chksum_err), + EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_tcp_udp_chksum_err), + EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_inner_ip_hdr_chksum_err), + EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_inner_tcp_udp_chksum_err), + EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_outer_ip_hdr_chksum_err), + EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_outer_tcp_udp_chksum_err), + EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_eth_crc_err), + EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_mcast_mismatch), + EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_frm_trunc), + EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_merge_events), + EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_merge_packets), + EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_xdp_drops), + EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_xdp_bad_drops), + EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_xdp_tx), + EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_xdp_redirect), +#ifdef CONFIG_RFS_ACCEL + EFX_ETHTOOL_UINT_CHANNEL_STAT_NO_N(rfs_filter_count), + EFX_ETHTOOL_UINT_CHANNEL_STAT(rfs_succeeded), + EFX_ETHTOOL_UINT_CHANNEL_STAT(rfs_failed), +#endif +}; + +#define EFX_ETHTOOL_SW_STAT_COUNT ARRAY_SIZE(efx_sw_stat_desc) + +void efx_ethtool_get_drvinfo(struct net_device *net_dev, + struct ethtool_drvinfo *info) +{ + struct efx_nic *efx = netdev_priv(net_dev); + + strlcpy(info->driver, KBUILD_MODNAME, sizeof(info->driver)); + strlcpy(info->version, EFX_DRIVER_VERSION, sizeof(info->version)); + efx_mcdi_print_fwver(efx, info->fw_version, + sizeof(info->fw_version)); + strlcpy(info->bus_info, pci_name(efx->pci_dev), sizeof(info->bus_info)); +} + +u32 efx_ethtool_get_msglevel(struct net_device *net_dev) +{ + struct efx_nic *efx = netdev_priv(net_dev); + + return efx->msg_enable; +} + +void efx_ethtool_set_msglevel(struct net_device *net_dev, u32 msg_enable) +{ + struct efx_nic *efx = netdev_priv(net_dev); + + efx->msg_enable = msg_enable; +} + +void efx_ethtool_get_pauseparam(struct net_device *net_dev, + struct ethtool_pauseparam *pause) +{ + struct efx_nic *efx = netdev_priv(net_dev); + + pause->rx_pause = !!(efx->wanted_fc & EFX_FC_RX); + pause->tx_pause = !!(efx->wanted_fc & EFX_FC_TX); + pause->autoneg = !!(efx->wanted_fc & EFX_FC_AUTO); +} + +/** + * efx_fill_test - fill in an individual self-test entry + * @test_index: Index of the test + * @strings: Ethtool strings, or %NULL + * @data: Ethtool test results, or %NULL + * @test: Pointer to test result (used only if data != %NULL) + * @unit_format: Unit name format (e.g. "chan\%d") + * @unit_id: Unit id (e.g. 0 for "chan0") + * @test_format: Test name format (e.g. "loopback.\%s.tx.sent") + * @test_id: Test id (e.g. "PHYXS" for "loopback.PHYXS.tx_sent") + * + * Fill in an individual self-test entry. + */ +static void efx_fill_test(unsigned int test_index, u8 *strings, u64 *data, + int *test, const char *unit_format, int unit_id, + const char *test_format, const char *test_id) +{ + char unit_str[ETH_GSTRING_LEN], test_str[ETH_GSTRING_LEN]; + + /* Fill data value, if applicable */ + if (data) + data[test_index] = *test; + + /* Fill string, if applicable */ + if (strings) { + if (strchr(unit_format, '%')) + snprintf(unit_str, sizeof(unit_str), + unit_format, unit_id); + else + strcpy(unit_str, unit_format); + snprintf(test_str, sizeof(test_str), test_format, test_id); + snprintf(strings + test_index * ETH_GSTRING_LEN, + ETH_GSTRING_LEN, + "%-6s %-24s", unit_str, test_str); + } +} + +#define EFX_CHANNEL_NAME(_channel) "chan%d", _channel->channel +#define EFX_TX_QUEUE_NAME(_tx_queue) "txq%d", _tx_queue->queue +#define EFX_RX_QUEUE_NAME(_rx_queue) "rxq%d", _rx_queue->queue +#define EFX_LOOPBACK_NAME(_mode, _counter) \ + "loopback.%s." _counter, STRING_TABLE_LOOKUP(_mode, efx_loopback_mode) + +/** + * efx_fill_loopback_test - fill in a block of loopback self-test entries + * @efx: Efx NIC + * @lb_tests: Efx loopback self-test results structure + * @mode: Loopback test mode + * @test_index: Starting index of the test + * @strings: Ethtool strings, or %NULL + * @data: Ethtool test results, or %NULL + * + * Fill in a block of loopback self-test entries. Return new test + * index. + */ +static int efx_fill_loopback_test(struct efx_nic *efx, + struct efx_loopback_self_tests *lb_tests, + enum efx_loopback_mode mode, + unsigned int test_index, + u8 *strings, u64 *data) +{ + struct efx_channel *channel = + efx_get_channel(efx, efx->tx_channel_offset); + struct efx_tx_queue *tx_queue; + + efx_for_each_channel_tx_queue(tx_queue, channel) { + efx_fill_test(test_index++, strings, data, + &lb_tests->tx_sent[tx_queue->queue], + EFX_TX_QUEUE_NAME(tx_queue), + EFX_LOOPBACK_NAME(mode, "tx_sent")); + efx_fill_test(test_index++, strings, data, + &lb_tests->tx_done[tx_queue->queue], + EFX_TX_QUEUE_NAME(tx_queue), + EFX_LOOPBACK_NAME(mode, "tx_done")); + } + efx_fill_test(test_index++, strings, data, + &lb_tests->rx_good, + "rx", 0, + EFX_LOOPBACK_NAME(mode, "rx_good")); + efx_fill_test(test_index++, strings, data, + &lb_tests->rx_bad, + "rx", 0, + EFX_LOOPBACK_NAME(mode, "rx_bad")); + + return test_index; +} + +/** + * efx_ethtool_fill_self_tests - get self-test details + * @efx: Efx NIC + * @tests: Efx self-test results structure, or %NULL + * @strings: Ethtool strings, or %NULL + * @data: Ethtool test results, or %NULL + * + * Get self-test number of strings, strings, and/or test results. + * Return number of strings (== number of test results). + * + * The reason for merging these three functions is to make sure that + * they can never be inconsistent. + */ +int efx_ethtool_fill_self_tests(struct efx_nic *efx, + struct efx_self_tests *tests, + u8 *strings, u64 *data) +{ + struct efx_channel *channel; + unsigned int n = 0, i; + enum efx_loopback_mode mode; + + efx_fill_test(n++, strings, data, &tests->phy_alive, + "phy", 0, "alive", NULL); + efx_fill_test(n++, strings, data, &tests->nvram, + "core", 0, "nvram", NULL); + efx_fill_test(n++, strings, data, &tests->interrupt, + "core", 0, "interrupt", NULL); + + /* Event queues */ + efx_for_each_channel(channel, efx) { + efx_fill_test(n++, strings, data, + &tests->eventq_dma[channel->channel], + EFX_CHANNEL_NAME(channel), + "eventq.dma", NULL); + efx_fill_test(n++, strings, data, + &tests->eventq_int[channel->channel], + EFX_CHANNEL_NAME(channel), + "eventq.int", NULL); + } + + efx_fill_test(n++, strings, data, &tests->memory, + "core", 0, "memory", NULL); + efx_fill_test(n++, strings, data, &tests->registers, + "core", 0, "registers", NULL); + + if (efx->phy_op->run_tests != NULL) { + EFX_WARN_ON_PARANOID(efx->phy_op->test_name == NULL); + + for (i = 0; true; ++i) { + const char *name; + + EFX_WARN_ON_PARANOID(i >= EFX_MAX_PHY_TESTS); + name = efx->phy_op->test_name(efx, i); + if (name == NULL) + break; + + efx_fill_test(n++, strings, data, &tests->phy_ext[i], + "phy", 0, name, NULL); + } + } + + /* Loopback tests */ + for (mode = LOOPBACK_NONE; mode <= LOOPBACK_TEST_MAX; mode++) { + if (!(efx->loopback_modes & (1 << mode))) + continue; + n = efx_fill_loopback_test(efx, + &tests->loopback[mode], mode, n, + strings, data); + } + + return n; +} + +static size_t efx_describe_per_queue_stats(struct efx_nic *efx, u8 *strings) +{ + size_t n_stats = 0; + struct efx_channel *channel; + + efx_for_each_channel(channel, efx) { + if (efx_channel_has_tx_queues(channel)) { + n_stats++; + if (strings != NULL) { + snprintf(strings, ETH_GSTRING_LEN, + "tx-%u.tx_packets", + channel->tx_queue[0].queue / + EFX_TXQ_TYPES); + + strings += ETH_GSTRING_LEN; + } + } + } + efx_for_each_channel(channel, efx) { + if (efx_channel_has_rx_queue(channel)) { + n_stats++; + if (strings != NULL) { + snprintf(strings, ETH_GSTRING_LEN, + "rx-%d.rx_packets", channel->channel); + strings += ETH_GSTRING_LEN; + } + } + } + if (efx->xdp_tx_queue_count && efx->xdp_tx_queues) { + unsigned short xdp; + + for (xdp = 0; xdp < efx->xdp_tx_queue_count; xdp++) { + n_stats++; + if (strings) { + snprintf(strings, ETH_GSTRING_LEN, + "tx-xdp-cpu-%hu.tx_packets", xdp); + strings += ETH_GSTRING_LEN; + } + } + } + + return n_stats; +} + +int efx_ethtool_get_sset_count(struct net_device *net_dev, int string_set) +{ + struct efx_nic *efx = netdev_priv(net_dev); + + switch (string_set) { + case ETH_SS_STATS: + return efx->type->describe_stats(efx, NULL) + + EFX_ETHTOOL_SW_STAT_COUNT + + efx_describe_per_queue_stats(efx, NULL) + + efx_ptp_describe_stats(efx, NULL); + case ETH_SS_TEST: + return efx_ethtool_fill_self_tests(efx, NULL, NULL, NULL); + default: + return -EINVAL; + } +} + +void efx_ethtool_get_strings(struct net_device *net_dev, + u32 string_set, u8 *strings) +{ + struct efx_nic *efx = netdev_priv(net_dev); + int i; + + switch (string_set) { + case ETH_SS_STATS: + strings += (efx->type->describe_stats(efx, strings) * + ETH_GSTRING_LEN); + for (i = 0; i < EFX_ETHTOOL_SW_STAT_COUNT; i++) + strlcpy(strings + i * ETH_GSTRING_LEN, + efx_sw_stat_desc[i].name, ETH_GSTRING_LEN); + strings += EFX_ETHTOOL_SW_STAT_COUNT * ETH_GSTRING_LEN; + strings += (efx_describe_per_queue_stats(efx, strings) * + ETH_GSTRING_LEN); + efx_ptp_describe_stats(efx, strings); + break; + case ETH_SS_TEST: + efx_ethtool_fill_self_tests(efx, NULL, strings, NULL); + break; + default: + /* No other string sets */ + break; + } +} + +void efx_ethtool_get_stats(struct net_device *net_dev, + struct ethtool_stats *stats, + u64 *data) +{ + struct efx_nic *efx = netdev_priv(net_dev); + const struct efx_sw_stat_desc *stat; + struct efx_channel *channel; + struct efx_tx_queue *tx_queue; + struct efx_rx_queue *rx_queue; + int i; + + spin_lock_bh(&efx->stats_lock); + + /* Get NIC statistics */ + data += efx->type->update_stats(efx, data, NULL); + + /* Get software statistics */ + for (i = 0; i < EFX_ETHTOOL_SW_STAT_COUNT; i++) { + stat = &efx_sw_stat_desc[i]; + switch (stat->source) { + case EFX_ETHTOOL_STAT_SOURCE_nic: + data[i] = stat->get_stat((void *)efx + stat->offset); + break; + case EFX_ETHTOOL_STAT_SOURCE_channel: + data[i] = 0; + efx_for_each_channel(channel, efx) + data[i] += stat->get_stat((void *)channel + + stat->offset); + break; + case EFX_ETHTOOL_STAT_SOURCE_tx_queue: + data[i] = 0; + efx_for_each_channel(channel, efx) { + efx_for_each_channel_tx_queue(tx_queue, channel) + data[i] += + stat->get_stat((void *)tx_queue + + stat->offset); + } + break; + } + } + data += EFX_ETHTOOL_SW_STAT_COUNT; + + spin_unlock_bh(&efx->stats_lock); + + efx_for_each_channel(channel, efx) { + if (efx_channel_has_tx_queues(channel)) { + *data = 0; + efx_for_each_channel_tx_queue(tx_queue, channel) { + *data += tx_queue->tx_packets; + } + data++; + } + } + efx_for_each_channel(channel, efx) { + if (efx_channel_has_rx_queue(channel)) { + *data = 0; + efx_for_each_channel_rx_queue(rx_queue, channel) { + *data += rx_queue->rx_packets; + } + data++; + } + } + if (efx->xdp_tx_queue_count && efx->xdp_tx_queues) { + int xdp; + + for (xdp = 0; xdp < efx->xdp_tx_queue_count; xdp++) { + data[0] = efx->xdp_tx_queues[xdp]->tx_packets; + data++; + } + } + + efx_ptp_update_stats(efx, data); +} diff --git a/drivers/net/ethernet/sfc/ethtool_common.h b/drivers/net/ethernet/sfc/ethtool_common.h new file mode 100644 index 000000000000..fa624313f330 --- /dev/null +++ b/drivers/net/ethernet/sfc/ethtool_common.h @@ -0,0 +1,30 @@ +/* SPDX-License-Identifier: GPL-2.0-only */ +/**************************************************************************** + * Driver for Solarflare network controllers and boards + * Copyright 2019 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 + * by the Free Software Foundation, incorporated herein by reference. + */ + +#ifndef EFX_ETHTOOL_COMMON_H +#define EFX_ETHTOOL_COMMON_H + +void efx_ethtool_get_drvinfo(struct net_device *net_dev, + struct ethtool_drvinfo *info); +u32 efx_ethtool_get_msglevel(struct net_device *net_dev); +void efx_ethtool_set_msglevel(struct net_device *net_dev, u32 msg_enable); +void efx_ethtool_get_pauseparam(struct net_device *net_dev, + struct ethtool_pauseparam *pause); +int efx_ethtool_fill_self_tests(struct efx_nic *efx, + struct efx_self_tests *tests, + u8 *strings, u64 *data); +int efx_ethtool_get_sset_count(struct net_device *net_dev, int string_set); +void efx_ethtool_get_strings(struct net_device *net_dev, u32 string_set, + u8 *strings); +void efx_ethtool_get_stats(struct net_device *net_dev, + struct ethtool_stats *stats __attribute__ ((unused)), + u64 *data); + +#endif diff --git a/drivers/net/ethernet/sfc/farch.c b/drivers/net/ethernet/sfc/farch.c index eedd32e2bfcb..dbbb898adddb 100644 --- a/drivers/net/ethernet/sfc/farch.c +++ b/drivers/net/ethernet/sfc/farch.c @@ -15,6 +15,7 @@ #include "net_driver.h" #include "bitfield.h" #include "efx.h" +#include "rx_common.h" #include "nic.h" #include "farch_regs.h" #include "sriov.h" diff --git a/drivers/net/ethernet/sfc/mcdi.h b/drivers/net/ethernet/sfc/mcdi.h index 9081f84a2604..54a45010b576 100644 --- a/drivers/net/ethernet/sfc/mcdi.h +++ b/drivers/net/ethernet/sfc/mcdi.h @@ -346,11 +346,8 @@ int efx_mcdi_flush_rxqs(struct efx_nic *efx); int efx_mcdi_port_probe(struct efx_nic *efx); void efx_mcdi_port_remove(struct efx_nic *efx); int efx_mcdi_port_reconfigure(struct efx_nic *efx); -int efx_mcdi_port_get_number(struct efx_nic *efx); u32 efx_mcdi_phy_get_caps(struct efx_nic *efx); void efx_mcdi_process_link_change(struct efx_nic *efx, efx_qword_t *ev); -int efx_mcdi_set_mac(struct efx_nic *efx); -#define EFX_MC_STATS_GENERATION_INVALID ((__force __le64)(-1)) void efx_mcdi_mac_start_stats(struct efx_nic *efx); void efx_mcdi_mac_stop_stats(struct efx_nic *efx); void efx_mcdi_mac_pull_stats(struct efx_nic *efx); diff --git a/drivers/net/ethernet/sfc/mcdi_functions.c b/drivers/net/ethernet/sfc/mcdi_functions.c new file mode 100644 index 000000000000..dcfe78b0fa5a --- /dev/null +++ b/drivers/net/ethernet/sfc/mcdi_functions.c @@ -0,0 +1,386 @@ +// SPDX-License-Identifier: GPL-2.0-only +/**************************************************************************** + * Driver for Solarflare network controllers and boards + * Copyright 2019 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 + * by the Free Software Foundation, incorporated herein by reference. + */ + +#include "net_driver.h" +#include "efx.h" +#include "nic.h" +#include "mcdi_functions.h" +#include "mcdi.h" +#include "mcdi_pcol.h" + +int efx_mcdi_free_vis(struct efx_nic *efx) +{ + MCDI_DECLARE_BUF_ERR(outbuf); + size_t outlen; + int rc = efx_mcdi_rpc_quiet(efx, MC_CMD_FREE_VIS, NULL, 0, + outbuf, sizeof(outbuf), &outlen); + + /* -EALREADY means nothing to free, so ignore */ + if (rc == -EALREADY) + rc = 0; + if (rc) + efx_mcdi_display_error(efx, MC_CMD_FREE_VIS, 0, outbuf, outlen, + rc); + return rc; +} + +int efx_mcdi_alloc_vis(struct efx_nic *efx, unsigned int min_vis, + unsigned int max_vis, unsigned int *vi_base, + unsigned int *allocated_vis) +{ + MCDI_DECLARE_BUF(outbuf, MC_CMD_ALLOC_VIS_OUT_LEN); + MCDI_DECLARE_BUF(inbuf, MC_CMD_ALLOC_VIS_IN_LEN); + size_t outlen; + int rc; + + MCDI_SET_DWORD(inbuf, ALLOC_VIS_IN_MIN_VI_COUNT, min_vis); + MCDI_SET_DWORD(inbuf, ALLOC_VIS_IN_MAX_VI_COUNT, max_vis); + rc = efx_mcdi_rpc(efx, MC_CMD_ALLOC_VIS, inbuf, sizeof(inbuf), + outbuf, sizeof(outbuf), &outlen); + if (rc != 0) + return rc; + + if (outlen < MC_CMD_ALLOC_VIS_OUT_LEN) + return -EIO; + + netif_dbg(efx, drv, efx->net_dev, "base VI is A0x%03x\n", + MCDI_DWORD(outbuf, ALLOC_VIS_OUT_VI_BASE)); + + if (vi_base) + *vi_base = MCDI_DWORD(outbuf, ALLOC_VIS_OUT_VI_BASE); + if (allocated_vis) + *allocated_vis = MCDI_DWORD(outbuf, ALLOC_VIS_OUT_VI_COUNT); + return 0; +} + +int efx_mcdi_ev_probe(struct efx_channel *channel) +{ + return efx_nic_alloc_buffer(channel->efx, &channel->eventq.buf, + (channel->eventq_mask + 1) * + sizeof(efx_qword_t), + GFP_KERNEL); +} + +int efx_mcdi_ev_init(struct efx_channel *channel, bool v1_cut_thru, bool v2) +{ + MCDI_DECLARE_BUF(inbuf, + MC_CMD_INIT_EVQ_V2_IN_LEN(EFX_MAX_EVQ_SIZE * 8 / + EFX_BUF_SIZE)); + MCDI_DECLARE_BUF(outbuf, MC_CMD_INIT_EVQ_V2_OUT_LEN); + size_t entries = channel->eventq.buf.len / EFX_BUF_SIZE; + struct efx_nic *efx = channel->efx; + size_t inlen, outlen; + dma_addr_t dma_addr; + int rc, i; + + /* Fill event queue with all ones (i.e. empty events) */ + memset(channel->eventq.buf.addr, 0xff, channel->eventq.buf.len); + + MCDI_SET_DWORD(inbuf, INIT_EVQ_IN_SIZE, channel->eventq_mask + 1); + MCDI_SET_DWORD(inbuf, INIT_EVQ_IN_INSTANCE, channel->channel); + /* INIT_EVQ expects index in vector table, not absolute */ + MCDI_SET_DWORD(inbuf, INIT_EVQ_IN_IRQ_NUM, channel->channel); + MCDI_SET_DWORD(inbuf, INIT_EVQ_IN_TMR_MODE, + MC_CMD_INIT_EVQ_IN_TMR_MODE_DIS); + MCDI_SET_DWORD(inbuf, INIT_EVQ_IN_TMR_LOAD, 0); + MCDI_SET_DWORD(inbuf, INIT_EVQ_IN_TMR_RELOAD, 0); + MCDI_SET_DWORD(inbuf, INIT_EVQ_IN_COUNT_MODE, + MC_CMD_INIT_EVQ_IN_COUNT_MODE_DIS); + MCDI_SET_DWORD(inbuf, INIT_EVQ_IN_COUNT_THRSHLD, 0); + + if (v2) { + /* Use the new generic approach to specifying event queue + * configuration, requesting lower latency or higher throughput. + * The options that actually get used appear in the output. + */ + MCDI_POPULATE_DWORD_2(inbuf, INIT_EVQ_V2_IN_FLAGS, + INIT_EVQ_V2_IN_FLAG_INTERRUPTING, 1, + INIT_EVQ_V2_IN_FLAG_TYPE, + MC_CMD_INIT_EVQ_V2_IN_FLAG_TYPE_AUTO); + } else { + MCDI_POPULATE_DWORD_4(inbuf, INIT_EVQ_IN_FLAGS, + INIT_EVQ_IN_FLAG_INTERRUPTING, 1, + INIT_EVQ_IN_FLAG_RX_MERGE, 1, + INIT_EVQ_IN_FLAG_TX_MERGE, 1, + INIT_EVQ_IN_FLAG_CUT_THRU, v1_cut_thru); + } + + dma_addr = channel->eventq.buf.dma_addr; + for (i = 0; i < entries; ++i) { + MCDI_SET_ARRAY_QWORD(inbuf, INIT_EVQ_IN_DMA_ADDR, i, dma_addr); + dma_addr += EFX_BUF_SIZE; + } + + inlen = MC_CMD_INIT_EVQ_IN_LEN(entries); + + rc = efx_mcdi_rpc(efx, MC_CMD_INIT_EVQ, inbuf, inlen, + outbuf, sizeof(outbuf), &outlen); + + if (outlen >= MC_CMD_INIT_EVQ_V2_OUT_LEN) + netif_dbg(efx, drv, efx->net_dev, + "Channel %d using event queue flags %08x\n", + channel->channel, + MCDI_DWORD(outbuf, INIT_EVQ_V2_OUT_FLAGS)); + + return rc; +} + +void efx_mcdi_ev_remove(struct efx_channel *channel) +{ + efx_nic_free_buffer(channel->efx, &channel->eventq.buf); +} + +void efx_mcdi_ev_fini(struct efx_channel *channel) +{ + MCDI_DECLARE_BUF(inbuf, MC_CMD_FINI_EVQ_IN_LEN); + MCDI_DECLARE_BUF_ERR(outbuf); + struct efx_nic *efx = channel->efx; + size_t outlen; + int rc; + + MCDI_SET_DWORD(inbuf, FINI_EVQ_IN_INSTANCE, channel->channel); + + rc = efx_mcdi_rpc_quiet(efx, MC_CMD_FINI_EVQ, inbuf, sizeof(inbuf), + outbuf, sizeof(outbuf), &outlen); + + if (rc && rc != -EALREADY) + goto fail; + + return; + +fail: + efx_mcdi_display_error(efx, MC_CMD_FINI_EVQ, MC_CMD_FINI_EVQ_IN_LEN, + outbuf, outlen, rc); +} + +int efx_mcdi_tx_init(struct efx_tx_queue *tx_queue, bool tso_v2) +{ + MCDI_DECLARE_BUF(inbuf, MC_CMD_INIT_TXQ_IN_LEN(EFX_MAX_DMAQ_SIZE * 8 / + EFX_BUF_SIZE)); + bool csum_offload = tx_queue->queue & EFX_TXQ_TYPE_OFFLOAD; + size_t entries = tx_queue->txd.buf.len / EFX_BUF_SIZE; + struct efx_channel *channel = tx_queue->channel; + struct efx_nic *efx = tx_queue->efx; + struct efx_ef10_nic_data *nic_data; + dma_addr_t dma_addr; + size_t inlen; + int rc, i; + + BUILD_BUG_ON(MC_CMD_INIT_TXQ_OUT_LEN != 0); + + nic_data = efx->nic_data; + + MCDI_SET_DWORD(inbuf, INIT_TXQ_IN_SIZE, tx_queue->ptr_mask + 1); + MCDI_SET_DWORD(inbuf, INIT_TXQ_IN_TARGET_EVQ, channel->channel); + MCDI_SET_DWORD(inbuf, INIT_TXQ_IN_LABEL, tx_queue->queue); + MCDI_SET_DWORD(inbuf, INIT_TXQ_IN_INSTANCE, tx_queue->queue); + MCDI_SET_DWORD(inbuf, INIT_TXQ_IN_OWNER_ID, 0); + MCDI_SET_DWORD(inbuf, INIT_TXQ_IN_PORT_ID, nic_data->vport_id); + + dma_addr = tx_queue->txd.buf.dma_addr; + + netif_dbg(efx, hw, efx->net_dev, "pushing TXQ %d. %zu entries (%llx)\n", + tx_queue->queue, entries, (u64)dma_addr); + + for (i = 0; i < entries; ++i) { + MCDI_SET_ARRAY_QWORD(inbuf, INIT_TXQ_IN_DMA_ADDR, i, dma_addr); + dma_addr += EFX_BUF_SIZE; + } + + inlen = MC_CMD_INIT_TXQ_IN_LEN(entries); + + do { + MCDI_POPULATE_DWORD_4(inbuf, INIT_TXQ_IN_FLAGS, + /* This flag was removed from mcdi_pcol.h for + * the non-_EXT version of INIT_TXQ. However, + * firmware still honours it. + */ + INIT_TXQ_EXT_IN_FLAG_TSOV2_EN, tso_v2, + INIT_TXQ_IN_FLAG_IP_CSUM_DIS, !csum_offload, + INIT_TXQ_IN_FLAG_TCP_CSUM_DIS, !csum_offload, + INIT_TXQ_EXT_IN_FLAG_TIMESTAMP, + tx_queue->timestamping); + + rc = efx_mcdi_rpc_quiet(efx, MC_CMD_INIT_TXQ, inbuf, inlen, + NULL, 0, NULL); + if (rc == -ENOSPC && tso_v2) { + /* Retry without TSOv2 if we're short on contexts. */ + tso_v2 = false; + netif_warn(efx, probe, efx->net_dev, + "TSOv2 context not available to segment in " + "hardware. TCP performance may be reduced.\n" + ); + } else if (rc) { + efx_mcdi_display_error(efx, MC_CMD_INIT_TXQ, + MC_CMD_INIT_TXQ_EXT_IN_LEN, + NULL, 0, rc); + goto fail; + } + } while (rc); + + return 0; + +fail: + return rc; +} + +void efx_mcdi_tx_remove(struct efx_tx_queue *tx_queue) +{ + efx_nic_free_buffer(tx_queue->efx, &tx_queue->txd.buf); +} + +void efx_mcdi_tx_fini(struct efx_tx_queue *tx_queue) +{ + MCDI_DECLARE_BUF(inbuf, MC_CMD_FINI_TXQ_IN_LEN); + MCDI_DECLARE_BUF_ERR(outbuf); + struct efx_nic *efx = tx_queue->efx; + size_t outlen; + int rc; + + MCDI_SET_DWORD(inbuf, FINI_TXQ_IN_INSTANCE, + tx_queue->queue); + + rc = efx_mcdi_rpc_quiet(efx, MC_CMD_FINI_TXQ, inbuf, sizeof(inbuf), + outbuf, sizeof(outbuf), &outlen); + + if (rc && rc != -EALREADY) + goto fail; + + return; + +fail: + efx_mcdi_display_error(efx, MC_CMD_FINI_TXQ, MC_CMD_FINI_TXQ_IN_LEN, + outbuf, outlen, rc); +} + +int efx_mcdi_rx_probe(struct efx_rx_queue *rx_queue) +{ + return efx_nic_alloc_buffer(rx_queue->efx, &rx_queue->rxd.buf, + (rx_queue->ptr_mask + 1) * + sizeof(efx_qword_t), + GFP_KERNEL); +} + +void efx_mcdi_rx_init(struct efx_rx_queue *rx_queue) +{ + MCDI_DECLARE_BUF(inbuf, + MC_CMD_INIT_RXQ_IN_LEN(EFX_MAX_DMAQ_SIZE * 8 / + EFX_BUF_SIZE)); + struct efx_channel *channel = efx_rx_queue_channel(rx_queue); + size_t entries = rx_queue->rxd.buf.len / EFX_BUF_SIZE; + struct efx_nic *efx = rx_queue->efx; + struct efx_ef10_nic_data *nic_data = efx->nic_data; + dma_addr_t dma_addr; + size_t inlen; + int rc; + int i; + BUILD_BUG_ON(MC_CMD_INIT_RXQ_OUT_LEN != 0); + + rx_queue->scatter_n = 0; + rx_queue->scatter_len = 0; + + MCDI_SET_DWORD(inbuf, INIT_RXQ_IN_SIZE, rx_queue->ptr_mask + 1); + MCDI_SET_DWORD(inbuf, INIT_RXQ_IN_TARGET_EVQ, channel->channel); + MCDI_SET_DWORD(inbuf, INIT_RXQ_IN_LABEL, efx_rx_queue_index(rx_queue)); + MCDI_SET_DWORD(inbuf, INIT_RXQ_IN_INSTANCE, + efx_rx_queue_index(rx_queue)); + MCDI_POPULATE_DWORD_2(inbuf, INIT_RXQ_IN_FLAGS, + INIT_RXQ_IN_FLAG_PREFIX, 1, + INIT_RXQ_IN_FLAG_TIMESTAMP, 1); + MCDI_SET_DWORD(inbuf, INIT_RXQ_IN_OWNER_ID, 0); + MCDI_SET_DWORD(inbuf, INIT_RXQ_IN_PORT_ID, nic_data->vport_id); + + dma_addr = rx_queue->rxd.buf.dma_addr; + + netif_dbg(efx, hw, efx->net_dev, "pushing RXQ %d. %zu entries (%llx)\n", + efx_rx_queue_index(rx_queue), entries, (u64)dma_addr); + + for (i = 0; i < entries; ++i) { + MCDI_SET_ARRAY_QWORD(inbuf, INIT_RXQ_IN_DMA_ADDR, i, dma_addr); + dma_addr += EFX_BUF_SIZE; + } + + inlen = MC_CMD_INIT_RXQ_IN_LEN(entries); + + rc = efx_mcdi_rpc(efx, MC_CMD_INIT_RXQ, inbuf, inlen, + NULL, 0, NULL); + if (rc) + netdev_WARN(efx->net_dev, "failed to initialise RXQ %d\n", + efx_rx_queue_index(rx_queue)); +} + +void efx_mcdi_rx_remove(struct efx_rx_queue *rx_queue) +{ + efx_nic_free_buffer(rx_queue->efx, &rx_queue->rxd.buf); +} + +void efx_mcdi_rx_fini(struct efx_rx_queue *rx_queue) +{ + MCDI_DECLARE_BUF(inbuf, MC_CMD_FINI_RXQ_IN_LEN); + MCDI_DECLARE_BUF_ERR(outbuf); + struct efx_nic *efx = rx_queue->efx; + size_t outlen; + int rc; + + MCDI_SET_DWORD(inbuf, FINI_RXQ_IN_INSTANCE, + efx_rx_queue_index(rx_queue)); + + rc = efx_mcdi_rpc_quiet(efx, MC_CMD_FINI_RXQ, inbuf, sizeof(inbuf), + outbuf, sizeof(outbuf), &outlen); + + if (rc && rc != -EALREADY) + goto fail; + + return; + +fail: + efx_mcdi_display_error(efx, MC_CMD_FINI_RXQ, MC_CMD_FINI_RXQ_IN_LEN, + outbuf, outlen, rc); +} + +int efx_mcdi_window_mode_to_stride(struct efx_nic *efx, u8 vi_window_mode) +{ + switch (vi_window_mode) { + case MC_CMD_GET_CAPABILITIES_V3_OUT_VI_WINDOW_MODE_8K: + efx->vi_stride = 8192; + break; + case MC_CMD_GET_CAPABILITIES_V3_OUT_VI_WINDOW_MODE_16K: + efx->vi_stride = 16384; + break; + case MC_CMD_GET_CAPABILITIES_V3_OUT_VI_WINDOW_MODE_64K: + efx->vi_stride = 65536; + break; + default: + netif_err(efx, probe, efx->net_dev, + "Unrecognised VI window mode %d\n", + vi_window_mode); + return -EIO; + } + netif_dbg(efx, probe, efx->net_dev, "vi_stride = %u\n", + efx->vi_stride); + return 0; +} + +int efx_get_pf_index(struct efx_nic *efx, unsigned int *pf_index) +{ + MCDI_DECLARE_BUF(outbuf, MC_CMD_GET_FUNCTION_INFO_OUT_LEN); + size_t outlen; + int rc; + + rc = efx_mcdi_rpc(efx, MC_CMD_GET_FUNCTION_INFO, NULL, 0, outbuf, + sizeof(outbuf), &outlen); + if (rc) + return rc; + if (outlen < sizeof(outbuf)) + return -EIO; + + *pf_index = MCDI_DWORD(outbuf, GET_FUNCTION_INFO_OUT_PF); + return 0; +} diff --git a/drivers/net/ethernet/sfc/mcdi_functions.h b/drivers/net/ethernet/sfc/mcdi_functions.h new file mode 100644 index 000000000000..ca4a5ac1a66b --- /dev/null +++ b/drivers/net/ethernet/sfc/mcdi_functions.h @@ -0,0 +1,32 @@ +/* SPDX-License-Identifier: GPL-2.0-only */ +/**************************************************************************** + * Driver for Solarflare network controllers and boards + * Copyright 2018 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 + * by the Free Software Foundation, incorporated herein by reference. + */ +#ifndef EFX_MCDI_FUNCTIONS_H +#define EFX_MCDI_FUNCTIONS_H + +int efx_mcdi_alloc_vis(struct efx_nic *efx, unsigned int min_vis, + unsigned int max_vis, unsigned int *vi_base, + unsigned int *allocated_vis); +int efx_mcdi_free_vis(struct efx_nic *efx); + +int efx_mcdi_ev_probe(struct efx_channel *channel); +int efx_mcdi_ev_init(struct efx_channel *channel, bool v1_cut_thru, bool v2); +void efx_mcdi_ev_remove(struct efx_channel *channel); +void efx_mcdi_ev_fini(struct efx_channel *channel); +int efx_mcdi_tx_init(struct efx_tx_queue *tx_queue, bool tso_v2); +void efx_mcdi_tx_remove(struct efx_tx_queue *tx_queue); +void efx_mcdi_tx_fini(struct efx_tx_queue *tx_queue); +int efx_mcdi_rx_probe(struct efx_rx_queue *rx_queue); +void efx_mcdi_rx_init(struct efx_rx_queue *rx_queue); +void efx_mcdi_rx_remove(struct efx_rx_queue *rx_queue); +void efx_mcdi_rx_fini(struct efx_rx_queue *rx_queue); +int efx_mcdi_window_mode_to_stride(struct efx_nic *efx, u8 vi_window_mode); +int efx_get_pf_index(struct efx_nic *efx, unsigned int *pf_index); + +#endif diff --git a/drivers/net/ethernet/sfc/mcdi_port.c b/drivers/net/ethernet/sfc/mcdi_port.c index fb7cde4980ed..ab5227b13ae6 100644 --- a/drivers/net/ethernet/sfc/mcdi_port.c +++ b/drivers/net/ethernet/sfc/mcdi_port.c @@ -14,106 +14,7 @@ #include "mcdi_pcol.h" #include "nic.h" #include "selftest.h" - -struct efx_mcdi_phy_data { - u32 flags; - u32 type; - u32 supported_cap; - u32 channel; - u32 port; - u32 stats_mask; - u8 name[20]; - u32 media; - u32 mmd_mask; - u8 revision[20]; - u32 forced_cap; -}; - -static int -efx_mcdi_get_phy_cfg(struct efx_nic *efx, struct efx_mcdi_phy_data *cfg) -{ - MCDI_DECLARE_BUF(outbuf, MC_CMD_GET_PHY_CFG_OUT_LEN); - size_t outlen; - int rc; - - BUILD_BUG_ON(MC_CMD_GET_PHY_CFG_IN_LEN != 0); - BUILD_BUG_ON(MC_CMD_GET_PHY_CFG_OUT_NAME_LEN != sizeof(cfg->name)); - - rc = efx_mcdi_rpc(efx, MC_CMD_GET_PHY_CFG, NULL, 0, - outbuf, sizeof(outbuf), &outlen); - if (rc) - goto fail; - - if (outlen < MC_CMD_GET_PHY_CFG_OUT_LEN) { - rc = -EIO; - goto fail; - } - - cfg->flags = MCDI_DWORD(outbuf, GET_PHY_CFG_OUT_FLAGS); - cfg->type = MCDI_DWORD(outbuf, GET_PHY_CFG_OUT_TYPE); - cfg->supported_cap = - MCDI_DWORD(outbuf, GET_PHY_CFG_OUT_SUPPORTED_CAP); - cfg->channel = MCDI_DWORD(outbuf, GET_PHY_CFG_OUT_CHANNEL); - cfg->port = MCDI_DWORD(outbuf, GET_PHY_CFG_OUT_PRT); - cfg->stats_mask = MCDI_DWORD(outbuf, GET_PHY_CFG_OUT_STATS_MASK); - memcpy(cfg->name, MCDI_PTR(outbuf, GET_PHY_CFG_OUT_NAME), - sizeof(cfg->name)); - cfg->media = MCDI_DWORD(outbuf, GET_PHY_CFG_OUT_MEDIA_TYPE); - cfg->mmd_mask = MCDI_DWORD(outbuf, GET_PHY_CFG_OUT_MMD_MASK); - memcpy(cfg->revision, MCDI_PTR(outbuf, GET_PHY_CFG_OUT_REVISION), - sizeof(cfg->revision)); - - return 0; - -fail: - netif_err(efx, hw, efx->net_dev, "%s: failed rc=%d\n", __func__, rc); - return rc; -} - -static int efx_mcdi_set_link(struct efx_nic *efx, u32 capabilities, - u32 flags, u32 loopback_mode, - u32 loopback_speed) -{ - MCDI_DECLARE_BUF(inbuf, MC_CMD_SET_LINK_IN_LEN); - int rc; - - BUILD_BUG_ON(MC_CMD_SET_LINK_OUT_LEN != 0); - - MCDI_SET_DWORD(inbuf, SET_LINK_IN_CAP, capabilities); - MCDI_SET_DWORD(inbuf, SET_LINK_IN_FLAGS, flags); - MCDI_SET_DWORD(inbuf, SET_LINK_IN_LOOPBACK_MODE, loopback_mode); - MCDI_SET_DWORD(inbuf, SET_LINK_IN_LOOPBACK_SPEED, loopback_speed); - - rc = efx_mcdi_rpc(efx, MC_CMD_SET_LINK, inbuf, sizeof(inbuf), - NULL, 0, NULL); - return rc; -} - -static int efx_mcdi_loopback_modes(struct efx_nic *efx, u64 *loopback_modes) -{ - MCDI_DECLARE_BUF(outbuf, MC_CMD_GET_LOOPBACK_MODES_OUT_LEN); - size_t outlen; - int rc; - - rc = efx_mcdi_rpc(efx, MC_CMD_GET_LOOPBACK_MODES, NULL, 0, - outbuf, sizeof(outbuf), &outlen); - if (rc) - goto fail; - - if (outlen < (MC_CMD_GET_LOOPBACK_MODES_OUT_SUGGESTED_OFST + - MC_CMD_GET_LOOPBACK_MODES_OUT_SUGGESTED_LEN)) { - rc = -EIO; - goto fail; - } - - *loopback_modes = MCDI_QWORD(outbuf, GET_LOOPBACK_MODES_OUT_SUGGESTED); - - return 0; - -fail: - netif_err(efx, hw, efx->net_dev, "%s: failed rc=%d\n", __func__, rc); - return rc; -} +#include "mcdi_port_common.h" static int efx_mcdi_mdio_read(struct net_device *net_dev, int prtad, int devad, u16 addr) @@ -168,246 +69,6 @@ static int efx_mcdi_mdio_write(struct net_device *net_dev, return 0; } -static void mcdi_to_ethtool_linkset(u32 media, u32 cap, unsigned long *linkset) -{ - #define SET_BIT(name) __set_bit(ETHTOOL_LINK_MODE_ ## name ## _BIT, \ - linkset) - - bitmap_zero(linkset, __ETHTOOL_LINK_MODE_MASK_NBITS); - switch (media) { - case MC_CMD_MEDIA_KX4: - SET_BIT(Backplane); - if (cap & (1 << MC_CMD_PHY_CAP_1000FDX_LBN)) - SET_BIT(1000baseKX_Full); - if (cap & (1 << MC_CMD_PHY_CAP_10000FDX_LBN)) - SET_BIT(10000baseKX4_Full); - if (cap & (1 << MC_CMD_PHY_CAP_40000FDX_LBN)) - SET_BIT(40000baseKR4_Full); - break; - - case MC_CMD_MEDIA_XFP: - case MC_CMD_MEDIA_SFP_PLUS: - case MC_CMD_MEDIA_QSFP_PLUS: - SET_BIT(FIBRE); - if (cap & (1 << MC_CMD_PHY_CAP_1000FDX_LBN)) - SET_BIT(1000baseT_Full); - if (cap & (1 << MC_CMD_PHY_CAP_10000FDX_LBN)) - SET_BIT(10000baseT_Full); - if (cap & (1 << MC_CMD_PHY_CAP_40000FDX_LBN)) - SET_BIT(40000baseCR4_Full); - if (cap & (1 << MC_CMD_PHY_CAP_100000FDX_LBN)) - SET_BIT(100000baseCR4_Full); - if (cap & (1 << MC_CMD_PHY_CAP_25000FDX_LBN)) - SET_BIT(25000baseCR_Full); - if (cap & (1 << MC_CMD_PHY_CAP_50000FDX_LBN)) - SET_BIT(50000baseCR2_Full); - break; - - case MC_CMD_MEDIA_BASE_T: - SET_BIT(TP); - if (cap & (1 << MC_CMD_PHY_CAP_10HDX_LBN)) - SET_BIT(10baseT_Half); - if (cap & (1 << MC_CMD_PHY_CAP_10FDX_LBN)) - SET_BIT(10baseT_Full); - if (cap & (1 << MC_CMD_PHY_CAP_100HDX_LBN)) - SET_BIT(100baseT_Half); - if (cap & (1 << MC_CMD_PHY_CAP_100FDX_LBN)) - SET_BIT(100baseT_Full); - if (cap & (1 << MC_CMD_PHY_CAP_1000HDX_LBN)) - SET_BIT(1000baseT_Half); - if (cap & (1 << MC_CMD_PHY_CAP_1000FDX_LBN)) - SET_BIT(1000baseT_Full); - if (cap & (1 << MC_CMD_PHY_CAP_10000FDX_LBN)) - SET_BIT(10000baseT_Full); - break; - } - - if (cap & (1 << MC_CMD_PHY_CAP_PAUSE_LBN)) - SET_BIT(Pause); - if (cap & (1 << MC_CMD_PHY_CAP_ASYM_LBN)) - SET_BIT(Asym_Pause); - if (cap & (1 << MC_CMD_PHY_CAP_AN_LBN)) - SET_BIT(Autoneg); - - #undef SET_BIT -} - -static u32 ethtool_linkset_to_mcdi_cap(const unsigned long *linkset) -{ - u32 result = 0; - - #define TEST_BIT(name) test_bit(ETHTOOL_LINK_MODE_ ## name ## _BIT, \ - linkset) - - if (TEST_BIT(10baseT_Half)) - result |= (1 << MC_CMD_PHY_CAP_10HDX_LBN); - if (TEST_BIT(10baseT_Full)) - result |= (1 << MC_CMD_PHY_CAP_10FDX_LBN); - if (TEST_BIT(100baseT_Half)) - result |= (1 << MC_CMD_PHY_CAP_100HDX_LBN); - if (TEST_BIT(100baseT_Full)) - result |= (1 << MC_CMD_PHY_CAP_100FDX_LBN); - if (TEST_BIT(1000baseT_Half)) - result |= (1 << MC_CMD_PHY_CAP_1000HDX_LBN); - if (TEST_BIT(1000baseT_Full) || TEST_BIT(1000baseKX_Full)) - result |= (1 << MC_CMD_PHY_CAP_1000FDX_LBN); - if (TEST_BIT(10000baseT_Full) || TEST_BIT(10000baseKX4_Full)) - result |= (1 << MC_CMD_PHY_CAP_10000FDX_LBN); - if (TEST_BIT(40000baseCR4_Full) || TEST_BIT(40000baseKR4_Full)) - result |= (1 << MC_CMD_PHY_CAP_40000FDX_LBN); - if (TEST_BIT(100000baseCR4_Full)) - result |= (1 << MC_CMD_PHY_CAP_100000FDX_LBN); - if (TEST_BIT(25000baseCR_Full)) - result |= (1 << MC_CMD_PHY_CAP_25000FDX_LBN); - if (TEST_BIT(50000baseCR2_Full)) - result |= (1 << MC_CMD_PHY_CAP_50000FDX_LBN); - if (TEST_BIT(Pause)) - result |= (1 << MC_CMD_PHY_CAP_PAUSE_LBN); - if (TEST_BIT(Asym_Pause)) - result |= (1 << MC_CMD_PHY_CAP_ASYM_LBN); - if (TEST_BIT(Autoneg)) - result |= (1 << MC_CMD_PHY_CAP_AN_LBN); - - #undef TEST_BIT - - return result; -} - -static u32 efx_get_mcdi_phy_flags(struct efx_nic *efx) -{ - struct efx_mcdi_phy_data *phy_cfg = efx->phy_data; - enum efx_phy_mode mode, supported; - u32 flags; - - /* TODO: Advertise the capabilities supported by this PHY */ - supported = 0; - if (phy_cfg->flags & (1 << MC_CMD_GET_PHY_CFG_OUT_TXDIS_LBN)) - supported |= PHY_MODE_TX_DISABLED; - if (phy_cfg->flags & (1 << MC_CMD_GET_PHY_CFG_OUT_LOWPOWER_LBN)) - supported |= PHY_MODE_LOW_POWER; - if (phy_cfg->flags & (1 << MC_CMD_GET_PHY_CFG_OUT_POWEROFF_LBN)) - supported |= PHY_MODE_OFF; - - mode = efx->phy_mode & supported; - - flags = 0; - if (mode & PHY_MODE_TX_DISABLED) - flags |= (1 << MC_CMD_SET_LINK_IN_TXDIS_LBN); - if (mode & PHY_MODE_LOW_POWER) - flags |= (1 << MC_CMD_SET_LINK_IN_LOWPOWER_LBN); - if (mode & PHY_MODE_OFF) - flags |= (1 << MC_CMD_SET_LINK_IN_POWEROFF_LBN); - - return flags; -} - -static u8 mcdi_to_ethtool_media(u32 media) -{ - switch (media) { - case MC_CMD_MEDIA_XAUI: - case MC_CMD_MEDIA_CX4: - case MC_CMD_MEDIA_KX4: - return PORT_OTHER; - - case MC_CMD_MEDIA_XFP: - case MC_CMD_MEDIA_SFP_PLUS: - case MC_CMD_MEDIA_QSFP_PLUS: - return PORT_FIBRE; - - case MC_CMD_MEDIA_BASE_T: - return PORT_TP; - - default: - return PORT_OTHER; - } -} - -static void efx_mcdi_phy_decode_link(struct efx_nic *efx, - struct efx_link_state *link_state, - u32 speed, u32 flags, u32 fcntl) -{ - switch (fcntl) { - case MC_CMD_FCNTL_AUTO: - WARN_ON(1); /* This is not a link mode */ - link_state->fc = EFX_FC_AUTO | EFX_FC_TX | EFX_FC_RX; - break; - case MC_CMD_FCNTL_BIDIR: - link_state->fc = EFX_FC_TX | EFX_FC_RX; - break; - case MC_CMD_FCNTL_RESPOND: - link_state->fc = EFX_FC_RX; - break; - default: - WARN_ON(1); - /* Fall through */ - case MC_CMD_FCNTL_OFF: - link_state->fc = 0; - break; - } - - link_state->up = !!(flags & (1 << MC_CMD_GET_LINK_OUT_LINK_UP_LBN)); - link_state->fd = !!(flags & (1 << MC_CMD_GET_LINK_OUT_FULL_DUPLEX_LBN)); - link_state->speed = speed; -} - -/* The semantics of the ethtool FEC mode bitmask are not well defined, - * particularly the meaning of combinations of bits. Which means we get to - * define our own semantics, as follows: - * OFF overrides any other bits, and means "disable all FEC" (with the - * exception of 25G KR4/CR4, where it is not possible to reject it if AN - * partner requests it). - * AUTO on its own means use cable requirements and link partner autoneg with - * fw-default preferences for the cable type. - * AUTO and either RS or BASER means use the specified FEC type if cable and - * link partner support it, otherwise autoneg/fw-default. - * RS or BASER alone means use the specified FEC type if cable and link partner - * support it and either requests it, otherwise no FEC. - * Both RS and BASER (whether AUTO or not) means use FEC if cable and link - * partner support it, preferring RS to BASER. - */ -static u32 ethtool_fec_caps_to_mcdi(u32 ethtool_cap) -{ - u32 ret = 0; - - if (ethtool_cap & ETHTOOL_FEC_OFF) - return 0; - - if (ethtool_cap & ETHTOOL_FEC_AUTO) - ret |= (1 << MC_CMD_PHY_CAP_BASER_FEC_LBN) | - (1 << MC_CMD_PHY_CAP_25G_BASER_FEC_LBN) | - (1 << MC_CMD_PHY_CAP_RS_FEC_LBN); - if (ethtool_cap & ETHTOOL_FEC_RS) - ret |= (1 << MC_CMD_PHY_CAP_RS_FEC_LBN) | - (1 << MC_CMD_PHY_CAP_RS_FEC_REQUESTED_LBN); - if (ethtool_cap & ETHTOOL_FEC_BASER) - ret |= (1 << MC_CMD_PHY_CAP_BASER_FEC_LBN) | - (1 << MC_CMD_PHY_CAP_25G_BASER_FEC_LBN) | - (1 << MC_CMD_PHY_CAP_BASER_FEC_REQUESTED_LBN) | - (1 << MC_CMD_PHY_CAP_25G_BASER_FEC_REQUESTED_LBN); - return ret; -} - -/* Invert ethtool_fec_caps_to_mcdi. There are two combinations that function - * can never produce, (baser xor rs) and neither req; the implementation below - * maps both of those to AUTO. This should never matter, and it's not clear - * what a better mapping would be anyway. - */ -static u32 mcdi_fec_caps_to_ethtool(u32 caps, bool is_25g) -{ - bool rs = caps & (1 << MC_CMD_PHY_CAP_RS_FEC_LBN), - rs_req = caps & (1 << MC_CMD_PHY_CAP_RS_FEC_REQUESTED_LBN), - baser = is_25g ? caps & (1 << MC_CMD_PHY_CAP_25G_BASER_FEC_LBN) - : caps & (1 << MC_CMD_PHY_CAP_BASER_FEC_LBN), - baser_req = is_25g ? caps & (1 << MC_CMD_PHY_CAP_25G_BASER_FEC_REQUESTED_LBN) - : caps & (1 << MC_CMD_PHY_CAP_BASER_FEC_REQUESTED_LBN); - - if (!baser && !rs) - return ETHTOOL_FEC_OFF; - return (rs_req ? ETHTOOL_FEC_RS : 0) | - (baser_req ? ETHTOOL_FEC_BASER : 0) | - (baser == baser_req && rs == rs_req ? 0 : ETHTOOL_FEC_AUTO); -} - static int efx_mcdi_phy_probe(struct efx_nic *efx) { struct efx_mcdi_phy_data *phy_data; @@ -527,58 +188,6 @@ int efx_mcdi_port_reconfigure(struct efx_nic *efx) efx->loopback_mode, 0); } -/* Verify that the forced flow control settings (!EFX_FC_AUTO) are - * supported by the link partner. Warn the user if this isn't the case - */ -static void efx_mcdi_phy_check_fcntl(struct efx_nic *efx, u32 lpa) -{ - struct efx_mcdi_phy_data *phy_cfg = efx->phy_data; - u32 rmtadv; - - /* The link partner capabilities are only relevant if the - * link supports flow control autonegotiation */ - if (~phy_cfg->supported_cap & (1 << MC_CMD_PHY_CAP_AN_LBN)) - return; - - /* If flow control autoneg is supported and enabled, then fine */ - if (efx->wanted_fc & EFX_FC_AUTO) - return; - - rmtadv = 0; - if (lpa & (1 << MC_CMD_PHY_CAP_PAUSE_LBN)) - rmtadv |= ADVERTISED_Pause; - if (lpa & (1 << MC_CMD_PHY_CAP_ASYM_LBN)) - rmtadv |= ADVERTISED_Asym_Pause; - - if ((efx->wanted_fc & EFX_FC_TX) && rmtadv == ADVERTISED_Asym_Pause) - netif_err(efx, link, efx->net_dev, - "warning: link partner doesn't support pause frames"); -} - -static bool efx_mcdi_phy_poll(struct efx_nic *efx) -{ - struct efx_link_state old_state = efx->link_state; - MCDI_DECLARE_BUF(outbuf, MC_CMD_GET_LINK_OUT_LEN); - int rc; - - WARN_ON(!mutex_is_locked(&efx->mac_lock)); - - BUILD_BUG_ON(MC_CMD_GET_LINK_IN_LEN != 0); - - rc = efx_mcdi_rpc(efx, MC_CMD_GET_LINK, NULL, 0, - outbuf, sizeof(outbuf), NULL); - if (rc) - efx->link_state.up = false; - else - efx_mcdi_phy_decode_link( - efx, &efx->link_state, - MCDI_DWORD(outbuf, GET_LINK_OUT_LINK_SPEED), - MCDI_DWORD(outbuf, GET_LINK_OUT_FLAGS), - MCDI_DWORD(outbuf, GET_LINK_OUT_FCNTL)); - - return !efx_link_state_equal(&efx->link_state, &old_state); -} - static void efx_mcdi_phy_remove(struct efx_nic *efx) { struct efx_mcdi_phy_data *phy_data = efx->phy_data; @@ -666,58 +275,6 @@ efx_mcdi_phy_set_link_ksettings(struct efx_nic *efx, return 0; } -static int efx_mcdi_phy_get_fecparam(struct efx_nic *efx, - struct ethtool_fecparam *fec) -{ - MCDI_DECLARE_BUF(outbuf, MC_CMD_GET_LINK_OUT_V2_LEN); - u32 caps, active, speed; /* MCDI format */ - bool is_25g = false; - size_t outlen; - int rc; - - BUILD_BUG_ON(MC_CMD_GET_LINK_IN_LEN != 0); - rc = efx_mcdi_rpc(efx, MC_CMD_GET_LINK, NULL, 0, - outbuf, sizeof(outbuf), &outlen); - if (rc) - return rc; - if (outlen < MC_CMD_GET_LINK_OUT_V2_LEN) - return -EOPNOTSUPP; - - /* behaviour for 25G/50G links depends on 25G BASER bit */ - speed = MCDI_DWORD(outbuf, GET_LINK_OUT_V2_LINK_SPEED); - is_25g = speed == 25000 || speed == 50000; - - caps = MCDI_DWORD(outbuf, GET_LINK_OUT_V2_CAP); - fec->fec = mcdi_fec_caps_to_ethtool(caps, is_25g); - /* BASER is never supported on 100G */ - if (speed == 100000) - fec->fec &= ~ETHTOOL_FEC_BASER; - - active = MCDI_DWORD(outbuf, GET_LINK_OUT_V2_FEC_TYPE); - switch (active) { - case MC_CMD_FEC_NONE: - fec->active_fec = ETHTOOL_FEC_OFF; - break; - case MC_CMD_FEC_BASER: - fec->active_fec = ETHTOOL_FEC_BASER; - break; - case MC_CMD_FEC_RS: - fec->active_fec = ETHTOOL_FEC_RS; - break; - default: - netif_warn(efx, hw, efx->net_dev, - "Firmware reports unrecognised FEC_TYPE %u\n", - active); - /* We don't know what firmware has picked. AUTO is as good a - * "can't happen" value as any other. - */ - fec->active_fec = ETHTOOL_FEC_AUTO; - break; - } - - return 0; -} - static int efx_mcdi_phy_set_fecparam(struct efx_nic *efx, const struct ethtool_fecparam *fec) { @@ -745,27 +302,6 @@ static int efx_mcdi_phy_set_fecparam(struct efx_nic *efx, return 0; } -static int efx_mcdi_phy_test_alive(struct efx_nic *efx) -{ - MCDI_DECLARE_BUF(outbuf, MC_CMD_GET_PHY_STATE_OUT_LEN); - size_t outlen; - int rc; - - BUILD_BUG_ON(MC_CMD_GET_PHY_STATE_IN_LEN != 0); - - rc = efx_mcdi_rpc(efx, MC_CMD_GET_PHY_STATE, NULL, 0, - outbuf, sizeof(outbuf), &outlen); - if (rc) - return rc; - - if (outlen < MC_CMD_GET_PHY_STATE_OUT_LEN) - return -EIO; - if (MCDI_DWORD(outbuf, GET_PHY_STATE_OUT_STATE) != MC_CMD_PHY_STATE_OK) - return -EINVAL; - - return 0; -} - static const char *const mcdi_sft9001_cable_diag_names[] = { "cable.pairA.length", "cable.pairB.length", @@ -1139,84 +675,6 @@ u32 efx_mcdi_phy_get_caps(struct efx_nic *efx) return phy_data->supported_cap; } -static unsigned int efx_mcdi_event_link_speed[] = { - [MCDI_EVENT_LINKCHANGE_SPEED_100M] = 100, - [MCDI_EVENT_LINKCHANGE_SPEED_1G] = 1000, - [MCDI_EVENT_LINKCHANGE_SPEED_10G] = 10000, - [MCDI_EVENT_LINKCHANGE_SPEED_40G] = 40000, - [MCDI_EVENT_LINKCHANGE_SPEED_25G] = 25000, - [MCDI_EVENT_LINKCHANGE_SPEED_50G] = 50000, - [MCDI_EVENT_LINKCHANGE_SPEED_100G] = 100000, -}; - -void efx_mcdi_process_link_change(struct efx_nic *efx, efx_qword_t *ev) -{ - u32 flags, fcntl, speed, lpa; - - speed = EFX_QWORD_FIELD(*ev, MCDI_EVENT_LINKCHANGE_SPEED); - EFX_WARN_ON_PARANOID(speed >= ARRAY_SIZE(efx_mcdi_event_link_speed)); - speed = efx_mcdi_event_link_speed[speed]; - - flags = EFX_QWORD_FIELD(*ev, MCDI_EVENT_LINKCHANGE_LINK_FLAGS); - fcntl = EFX_QWORD_FIELD(*ev, MCDI_EVENT_LINKCHANGE_FCNTL); - lpa = EFX_QWORD_FIELD(*ev, MCDI_EVENT_LINKCHANGE_LP_CAP); - - /* efx->link_state is only modified by efx_mcdi_phy_get_link(), - * which is only run after flushing the event queues. Therefore, it - * is safe to modify the link state outside of the mac_lock here. - */ - efx_mcdi_phy_decode_link(efx, &efx->link_state, speed, flags, fcntl); - - efx_mcdi_phy_check_fcntl(efx, lpa); - - efx_link_status_changed(efx); -} - -int efx_mcdi_set_mac(struct efx_nic *efx) -{ - u32 fcntl; - MCDI_DECLARE_BUF(cmdbytes, MC_CMD_SET_MAC_IN_LEN); - - BUILD_BUG_ON(MC_CMD_SET_MAC_OUT_LEN != 0); - - /* This has no effect on EF10 */ - ether_addr_copy(MCDI_PTR(cmdbytes, SET_MAC_IN_ADDR), - efx->net_dev->dev_addr); - - MCDI_SET_DWORD(cmdbytes, SET_MAC_IN_MTU, - EFX_MAX_FRAME_LEN(efx->net_dev->mtu)); - MCDI_SET_DWORD(cmdbytes, SET_MAC_IN_DRAIN, 0); - - /* Set simple MAC filter for Siena */ - MCDI_POPULATE_DWORD_1(cmdbytes, SET_MAC_IN_REJECT, - SET_MAC_IN_REJECT_UNCST, efx->unicast_filter); - - MCDI_POPULATE_DWORD_1(cmdbytes, SET_MAC_IN_FLAGS, - SET_MAC_IN_FLAG_INCLUDE_FCS, - !!(efx->net_dev->features & NETIF_F_RXFCS)); - - switch (efx->wanted_fc) { - case EFX_FC_RX | EFX_FC_TX: - fcntl = MC_CMD_FCNTL_BIDIR; - break; - case EFX_FC_RX: - fcntl = MC_CMD_FCNTL_RESPOND; - break; - default: - fcntl = MC_CMD_FCNTL_OFF; - break; - } - if (efx->wanted_fc & EFX_FC_AUTO) - fcntl = MC_CMD_FCNTL_AUTO; - if (efx->fc_disable) - fcntl = MC_CMD_FCNTL_OFF; - - MCDI_SET_DWORD(cmdbytes, SET_MAC_IN_FCNTL, fcntl); - - return efx_mcdi_rpc(efx, MC_CMD_SET_MAC, cmdbytes, sizeof(cmdbytes), - NULL, 0, NULL); -} - bool efx_mcdi_mac_check_fault(struct efx_nic *efx) { MCDI_DECLARE_BUF(outbuf, MC_CMD_GET_LINK_OUT_LEN); @@ -1348,17 +806,3 @@ void efx_mcdi_port_remove(struct efx_nic *efx) efx->phy_op->remove(efx); efx_nic_free_buffer(efx, &efx->stats_buffer); } - -/* Get physical port number (EF10 only; on Siena it is same as PF number) */ -int efx_mcdi_port_get_number(struct efx_nic *efx) -{ - MCDI_DECLARE_BUF(outbuf, MC_CMD_GET_PORT_ASSIGNMENT_OUT_LEN); - int rc; - - rc = efx_mcdi_rpc(efx, MC_CMD_GET_PORT_ASSIGNMENT, NULL, 0, - outbuf, sizeof(outbuf), NULL); - if (rc) - return rc; - - return MCDI_DWORD(outbuf, GET_PORT_ASSIGNMENT_OUT_PORT); -} diff --git a/drivers/net/ethernet/sfc/mcdi_port_common.c b/drivers/net/ethernet/sfc/mcdi_port_common.c new file mode 100644 index 000000000000..a6a072ba46d3 --- /dev/null +++ b/drivers/net/ethernet/sfc/mcdi_port_common.c @@ -0,0 +1,568 @@ +// SPDX-License-Identifier: GPL-2.0-only +/**************************************************************************** + * Driver for Solarflare network controllers and boards + * Copyright 2018 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 + * by the Free Software Foundation, incorporated herein by reference. + */ + +#include "mcdi_port_common.h" +#include "efx_common.h" + +int efx_mcdi_get_phy_cfg(struct efx_nic *efx, struct efx_mcdi_phy_data *cfg) +{ + MCDI_DECLARE_BUF(outbuf, MC_CMD_GET_PHY_CFG_OUT_LEN); + size_t outlen; + int rc; + + BUILD_BUG_ON(MC_CMD_GET_PHY_CFG_IN_LEN != 0); + BUILD_BUG_ON(MC_CMD_GET_PHY_CFG_OUT_NAME_LEN != sizeof(cfg->name)); + + rc = efx_mcdi_rpc(efx, MC_CMD_GET_PHY_CFG, NULL, 0, + outbuf, sizeof(outbuf), &outlen); + if (rc) + goto fail; + + if (outlen < MC_CMD_GET_PHY_CFG_OUT_LEN) { + rc = -EIO; + goto fail; + } + + cfg->flags = MCDI_DWORD(outbuf, GET_PHY_CFG_OUT_FLAGS); + cfg->type = MCDI_DWORD(outbuf, GET_PHY_CFG_OUT_TYPE); + cfg->supported_cap = + MCDI_DWORD(outbuf, GET_PHY_CFG_OUT_SUPPORTED_CAP); + cfg->channel = MCDI_DWORD(outbuf, GET_PHY_CFG_OUT_CHANNEL); + cfg->port = MCDI_DWORD(outbuf, GET_PHY_CFG_OUT_PRT); + cfg->stats_mask = MCDI_DWORD(outbuf, GET_PHY_CFG_OUT_STATS_MASK); + memcpy(cfg->name, MCDI_PTR(outbuf, GET_PHY_CFG_OUT_NAME), + sizeof(cfg->name)); + cfg->media = MCDI_DWORD(outbuf, GET_PHY_CFG_OUT_MEDIA_TYPE); + cfg->mmd_mask = MCDI_DWORD(outbuf, GET_PHY_CFG_OUT_MMD_MASK); + memcpy(cfg->revision, MCDI_PTR(outbuf, GET_PHY_CFG_OUT_REVISION), + sizeof(cfg->revision)); + + return 0; + +fail: + netif_err(efx, hw, efx->net_dev, "%s: failed rc=%d\n", __func__, rc); + return rc; +} + +void efx_link_set_advertising(struct efx_nic *efx, + const unsigned long *advertising) +{ + memcpy(efx->link_advertising, advertising, + sizeof(__ETHTOOL_DECLARE_LINK_MODE_MASK())); + + efx->link_advertising[0] |= ADVERTISED_Autoneg; + if (advertising[0] & ADVERTISED_Pause) + efx->wanted_fc |= (EFX_FC_TX | EFX_FC_RX); + else + efx->wanted_fc &= ~(EFX_FC_TX | EFX_FC_RX); + if (advertising[0] & ADVERTISED_Asym_Pause) + efx->wanted_fc ^= EFX_FC_TX; +} + +int efx_mcdi_set_link(struct efx_nic *efx, u32 capabilities, + u32 flags, u32 loopback_mode, u32 loopback_speed) +{ + MCDI_DECLARE_BUF(inbuf, MC_CMD_SET_LINK_IN_LEN); + int rc; + + BUILD_BUG_ON(MC_CMD_SET_LINK_OUT_LEN != 0); + + MCDI_SET_DWORD(inbuf, SET_LINK_IN_CAP, capabilities); + MCDI_SET_DWORD(inbuf, SET_LINK_IN_FLAGS, flags); + MCDI_SET_DWORD(inbuf, SET_LINK_IN_LOOPBACK_MODE, loopback_mode); + MCDI_SET_DWORD(inbuf, SET_LINK_IN_LOOPBACK_SPEED, loopback_speed); + + rc = efx_mcdi_rpc(efx, MC_CMD_SET_LINK, inbuf, sizeof(inbuf), + NULL, 0, NULL); + return rc; +} + +int efx_mcdi_loopback_modes(struct efx_nic *efx, u64 *loopback_modes) +{ + MCDI_DECLARE_BUF(outbuf, MC_CMD_GET_LOOPBACK_MODES_OUT_LEN); + size_t outlen; + int rc; + + rc = efx_mcdi_rpc(efx, MC_CMD_GET_LOOPBACK_MODES, NULL, 0, + outbuf, sizeof(outbuf), &outlen); + if (rc) + goto fail; + + if (outlen < (MC_CMD_GET_LOOPBACK_MODES_OUT_SUGGESTED_OFST + + MC_CMD_GET_LOOPBACK_MODES_OUT_SUGGESTED_LEN)) { + rc = -EIO; + goto fail; + } + + *loopback_modes = MCDI_QWORD(outbuf, GET_LOOPBACK_MODES_OUT_SUGGESTED); + + return 0; + +fail: + netif_err(efx, hw, efx->net_dev, "%s: failed rc=%d\n", __func__, rc); + return rc; +} + +void mcdi_to_ethtool_linkset(u32 media, u32 cap, unsigned long *linkset) +{ + #define SET_BIT(name) __set_bit(ETHTOOL_LINK_MODE_ ## name ## _BIT, \ + linkset) + + bitmap_zero(linkset, __ETHTOOL_LINK_MODE_MASK_NBITS); + switch (media) { + case MC_CMD_MEDIA_KX4: + SET_BIT(Backplane); + if (cap & (1 << MC_CMD_PHY_CAP_1000FDX_LBN)) + SET_BIT(1000baseKX_Full); + if (cap & (1 << MC_CMD_PHY_CAP_10000FDX_LBN)) + SET_BIT(10000baseKX4_Full); + if (cap & (1 << MC_CMD_PHY_CAP_40000FDX_LBN)) + SET_BIT(40000baseKR4_Full); + break; + + case MC_CMD_MEDIA_XFP: + case MC_CMD_MEDIA_SFP_PLUS: + case MC_CMD_MEDIA_QSFP_PLUS: + SET_BIT(FIBRE); + if (cap & (1 << MC_CMD_PHY_CAP_1000FDX_LBN)) + SET_BIT(1000baseT_Full); + if (cap & (1 << MC_CMD_PHY_CAP_10000FDX_LBN)) + SET_BIT(10000baseT_Full); + if (cap & (1 << MC_CMD_PHY_CAP_40000FDX_LBN)) + SET_BIT(40000baseCR4_Full); + if (cap & (1 << MC_CMD_PHY_CAP_100000FDX_LBN)) + SET_BIT(100000baseCR4_Full); + if (cap & (1 << MC_CMD_PHY_CAP_25000FDX_LBN)) + SET_BIT(25000baseCR_Full); + if (cap & (1 << MC_CMD_PHY_CAP_50000FDX_LBN)) + SET_BIT(50000baseCR2_Full); + break; + + case MC_CMD_MEDIA_BASE_T: + SET_BIT(TP); + if (cap & (1 << MC_CMD_PHY_CAP_10HDX_LBN)) + SET_BIT(10baseT_Half); + if (cap & (1 << MC_CMD_PHY_CAP_10FDX_LBN)) + SET_BIT(10baseT_Full); + if (cap & (1 << MC_CMD_PHY_CAP_100HDX_LBN)) + SET_BIT(100baseT_Half); + if (cap & (1 << MC_CMD_PHY_CAP_100FDX_LBN)) + SET_BIT(100baseT_Full); + if (cap & (1 << MC_CMD_PHY_CAP_1000HDX_LBN)) + SET_BIT(1000baseT_Half); + if (cap & (1 << MC_CMD_PHY_CAP_1000FDX_LBN)) + SET_BIT(1000baseT_Full); + if (cap & (1 << MC_CMD_PHY_CAP_10000FDX_LBN)) + SET_BIT(10000baseT_Full); + break; + } + + if (cap & (1 << MC_CMD_PHY_CAP_PAUSE_LBN)) + SET_BIT(Pause); + if (cap & (1 << MC_CMD_PHY_CAP_ASYM_LBN)) + SET_BIT(Asym_Pause); + if (cap & (1 << MC_CMD_PHY_CAP_AN_LBN)) + SET_BIT(Autoneg); + + #undef SET_BIT +} + +u32 ethtool_linkset_to_mcdi_cap(const unsigned long *linkset) +{ + u32 result = 0; + + #define TEST_BIT(name) test_bit(ETHTOOL_LINK_MODE_ ## name ## _BIT, \ + linkset) + + if (TEST_BIT(10baseT_Half)) + result |= (1 << MC_CMD_PHY_CAP_10HDX_LBN); + if (TEST_BIT(10baseT_Full)) + result |= (1 << MC_CMD_PHY_CAP_10FDX_LBN); + if (TEST_BIT(100baseT_Half)) + result |= (1 << MC_CMD_PHY_CAP_100HDX_LBN); + if (TEST_BIT(100baseT_Full)) + result |= (1 << MC_CMD_PHY_CAP_100FDX_LBN); + if (TEST_BIT(1000baseT_Half)) + result |= (1 << MC_CMD_PHY_CAP_1000HDX_LBN); + if (TEST_BIT(1000baseT_Full) || TEST_BIT(1000baseKX_Full)) + result |= (1 << MC_CMD_PHY_CAP_1000FDX_LBN); + if (TEST_BIT(10000baseT_Full) || TEST_BIT(10000baseKX4_Full)) + result |= (1 << MC_CMD_PHY_CAP_10000FDX_LBN); + if (TEST_BIT(40000baseCR4_Full) || TEST_BIT(40000baseKR4_Full)) + result |= (1 << MC_CMD_PHY_CAP_40000FDX_LBN); + if (TEST_BIT(100000baseCR4_Full)) + result |= (1 << MC_CMD_PHY_CAP_100000FDX_LBN); + if (TEST_BIT(25000baseCR_Full)) + result |= (1 << MC_CMD_PHY_CAP_25000FDX_LBN); + if (TEST_BIT(50000baseCR2_Full)) + result |= (1 << MC_CMD_PHY_CAP_50000FDX_LBN); + if (TEST_BIT(Pause)) + result |= (1 << MC_CMD_PHY_CAP_PAUSE_LBN); + if (TEST_BIT(Asym_Pause)) + result |= (1 << MC_CMD_PHY_CAP_ASYM_LBN); + if (TEST_BIT(Autoneg)) + result |= (1 << MC_CMD_PHY_CAP_AN_LBN); + + #undef TEST_BIT + + return result; +} + +u32 efx_get_mcdi_phy_flags(struct efx_nic *efx) +{ + struct efx_mcdi_phy_data *phy_cfg = efx->phy_data; + enum efx_phy_mode mode, supported; + u32 flags; + + /* TODO: Advertise the capabilities supported by this PHY */ + supported = 0; + if (phy_cfg->flags & (1 << MC_CMD_GET_PHY_CFG_OUT_TXDIS_LBN)) + supported |= PHY_MODE_TX_DISABLED; + if (phy_cfg->flags & (1 << MC_CMD_GET_PHY_CFG_OUT_LOWPOWER_LBN)) + supported |= PHY_MODE_LOW_POWER; + if (phy_cfg->flags & (1 << MC_CMD_GET_PHY_CFG_OUT_POWEROFF_LBN)) + supported |= PHY_MODE_OFF; + + mode = efx->phy_mode & supported; + + flags = 0; + if (mode & PHY_MODE_TX_DISABLED) + flags |= (1 << MC_CMD_SET_LINK_IN_TXDIS_LBN); + if (mode & PHY_MODE_LOW_POWER) + flags |= (1 << MC_CMD_SET_LINK_IN_LOWPOWER_LBN); + if (mode & PHY_MODE_OFF) + flags |= (1 << MC_CMD_SET_LINK_IN_POWEROFF_LBN); + + return flags; +} + +u8 mcdi_to_ethtool_media(u32 media) +{ + switch (media) { + case MC_CMD_MEDIA_XAUI: + case MC_CMD_MEDIA_CX4: + case MC_CMD_MEDIA_KX4: + return PORT_OTHER; + + case MC_CMD_MEDIA_XFP: + case MC_CMD_MEDIA_SFP_PLUS: + case MC_CMD_MEDIA_QSFP_PLUS: + return PORT_FIBRE; + + case MC_CMD_MEDIA_BASE_T: + return PORT_TP; + + default: + return PORT_OTHER; + } +} + +void efx_mcdi_phy_decode_link(struct efx_nic *efx, + struct efx_link_state *link_state, + u32 speed, u32 flags, u32 fcntl) +{ + switch (fcntl) { + case MC_CMD_FCNTL_AUTO: + WARN_ON(1); /* This is not a link mode */ + link_state->fc = EFX_FC_AUTO | EFX_FC_TX | EFX_FC_RX; + break; + case MC_CMD_FCNTL_BIDIR: + link_state->fc = EFX_FC_TX | EFX_FC_RX; + break; + case MC_CMD_FCNTL_RESPOND: + link_state->fc = EFX_FC_RX; + break; + default: + WARN_ON(1); + /* Fall through */ + case MC_CMD_FCNTL_OFF: + link_state->fc = 0; + break; + } + + link_state->up = !!(flags & (1 << MC_CMD_GET_LINK_OUT_LINK_UP_LBN)); + link_state->fd = !!(flags & (1 << MC_CMD_GET_LINK_OUT_FULL_DUPLEX_LBN)); + link_state->speed = speed; +} + +/* The semantics of the ethtool FEC mode bitmask are not well defined, + * particularly the meaning of combinations of bits. Which means we get to + * define our own semantics, as follows: + * OFF overrides any other bits, and means "disable all FEC" (with the + * exception of 25G KR4/CR4, where it is not possible to reject it if AN + * partner requests it). + * AUTO on its own means use cable requirements and link partner autoneg with + * fw-default preferences for the cable type. + * AUTO and either RS or BASER means use the specified FEC type if cable and + * link partner support it, otherwise autoneg/fw-default. + * RS or BASER alone means use the specified FEC type if cable and link partner + * support it and either requests it, otherwise no FEC. + * Both RS and BASER (whether AUTO or not) means use FEC if cable and link + * partner support it, preferring RS to BASER. + */ +u32 ethtool_fec_caps_to_mcdi(u32 ethtool_cap) +{ + u32 ret = 0; + + if (ethtool_cap & ETHTOOL_FEC_OFF) + return 0; + + if (ethtool_cap & ETHTOOL_FEC_AUTO) + ret |= (1 << MC_CMD_PHY_CAP_BASER_FEC_LBN) | + (1 << MC_CMD_PHY_CAP_25G_BASER_FEC_LBN) | + (1 << MC_CMD_PHY_CAP_RS_FEC_LBN); + if (ethtool_cap & ETHTOOL_FEC_RS) + ret |= (1 << MC_CMD_PHY_CAP_RS_FEC_LBN) | + (1 << MC_CMD_PHY_CAP_RS_FEC_REQUESTED_LBN); + if (ethtool_cap & ETHTOOL_FEC_BASER) + ret |= (1 << MC_CMD_PHY_CAP_BASER_FEC_LBN) | + (1 << MC_CMD_PHY_CAP_25G_BASER_FEC_LBN) | + (1 << MC_CMD_PHY_CAP_BASER_FEC_REQUESTED_LBN) | + (1 << MC_CMD_PHY_CAP_25G_BASER_FEC_REQUESTED_LBN); + return ret; +} + +/* Invert ethtool_fec_caps_to_mcdi. There are two combinations that function + * can never produce, (baser xor rs) and neither req; the implementation below + * maps both of those to AUTO. This should never matter, and it's not clear + * what a better mapping would be anyway. + */ +u32 mcdi_fec_caps_to_ethtool(u32 caps, bool is_25g) +{ + bool rs = caps & (1 << MC_CMD_PHY_CAP_RS_FEC_LBN), + rs_req = caps & (1 << MC_CMD_PHY_CAP_RS_FEC_REQUESTED_LBN), + baser = is_25g ? caps & (1 << MC_CMD_PHY_CAP_25G_BASER_FEC_LBN) + : caps & (1 << MC_CMD_PHY_CAP_BASER_FEC_LBN), + baser_req = is_25g ? caps & (1 << MC_CMD_PHY_CAP_25G_BASER_FEC_REQUESTED_LBN) + : caps & (1 << MC_CMD_PHY_CAP_BASER_FEC_REQUESTED_LBN); + + if (!baser && !rs) + return ETHTOOL_FEC_OFF; + return (rs_req ? ETHTOOL_FEC_RS : 0) | + (baser_req ? ETHTOOL_FEC_BASER : 0) | + (baser == baser_req && rs == rs_req ? 0 : ETHTOOL_FEC_AUTO); +} + +/* Verify that the forced flow control settings (!EFX_FC_AUTO) are + * supported by the link partner. Warn the user if this isn't the case + */ +void efx_mcdi_phy_check_fcntl(struct efx_nic *efx, u32 lpa) +{ + struct efx_mcdi_phy_data *phy_cfg = efx->phy_data; + u32 rmtadv; + + /* The link partner capabilities are only relevant if the + * link supports flow control autonegotiation + */ + if (~phy_cfg->supported_cap & (1 << MC_CMD_PHY_CAP_AN_LBN)) + return; + + /* If flow control autoneg is supported and enabled, then fine */ + if (efx->wanted_fc & EFX_FC_AUTO) + return; + + rmtadv = 0; + if (lpa & (1 << MC_CMD_PHY_CAP_PAUSE_LBN)) + rmtadv |= ADVERTISED_Pause; + if (lpa & (1 << MC_CMD_PHY_CAP_ASYM_LBN)) + rmtadv |= ADVERTISED_Asym_Pause; + + if ((efx->wanted_fc & EFX_FC_TX) && rmtadv == ADVERTISED_Asym_Pause) + netif_err(efx, link, efx->net_dev, + "warning: link partner doesn't support pause frames"); +} + +bool efx_mcdi_phy_poll(struct efx_nic *efx) +{ + struct efx_link_state old_state = efx->link_state; + MCDI_DECLARE_BUF(outbuf, MC_CMD_GET_LINK_OUT_LEN); + int rc; + + WARN_ON(!mutex_is_locked(&efx->mac_lock)); + + BUILD_BUG_ON(MC_CMD_GET_LINK_IN_LEN != 0); + + rc = efx_mcdi_rpc(efx, MC_CMD_GET_LINK, NULL, 0, + outbuf, sizeof(outbuf), NULL); + if (rc) + efx->link_state.up = false; + else + efx_mcdi_phy_decode_link( + efx, &efx->link_state, + MCDI_DWORD(outbuf, GET_LINK_OUT_LINK_SPEED), + MCDI_DWORD(outbuf, GET_LINK_OUT_FLAGS), + MCDI_DWORD(outbuf, GET_LINK_OUT_FCNTL)); + + return !efx_link_state_equal(&efx->link_state, &old_state); +} + +int efx_mcdi_phy_get_fecparam(struct efx_nic *efx, struct ethtool_fecparam *fec) +{ + MCDI_DECLARE_BUF(outbuf, MC_CMD_GET_LINK_OUT_V2_LEN); + u32 caps, active, speed; /* MCDI format */ + bool is_25g = false; + size_t outlen; + int rc; + + BUILD_BUG_ON(MC_CMD_GET_LINK_IN_LEN != 0); + rc = efx_mcdi_rpc(efx, MC_CMD_GET_LINK, NULL, 0, + outbuf, sizeof(outbuf), &outlen); + if (rc) + return rc; + if (outlen < MC_CMD_GET_LINK_OUT_V2_LEN) + return -EOPNOTSUPP; + + /* behaviour for 25G/50G links depends on 25G BASER bit */ + speed = MCDI_DWORD(outbuf, GET_LINK_OUT_V2_LINK_SPEED); + is_25g = speed == 25000 || speed == 50000; + + caps = MCDI_DWORD(outbuf, GET_LINK_OUT_V2_CAP); + fec->fec = mcdi_fec_caps_to_ethtool(caps, is_25g); + /* BASER is never supported on 100G */ + if (speed == 100000) + fec->fec &= ~ETHTOOL_FEC_BASER; + + active = MCDI_DWORD(outbuf, GET_LINK_OUT_V2_FEC_TYPE); + switch (active) { + case MC_CMD_FEC_NONE: + fec->active_fec = ETHTOOL_FEC_OFF; + break; + case MC_CMD_FEC_BASER: + fec->active_fec = ETHTOOL_FEC_BASER; + break; + case MC_CMD_FEC_RS: + fec->active_fec = ETHTOOL_FEC_RS; + break; + default: + netif_warn(efx, hw, efx->net_dev, + "Firmware reports unrecognised FEC_TYPE %u\n", + active); + /* We don't know what firmware has picked. AUTO is as good a + * "can't happen" value as any other. + */ + fec->active_fec = ETHTOOL_FEC_AUTO; + break; + } + + return 0; +} + +int efx_mcdi_phy_test_alive(struct efx_nic *efx) +{ + MCDI_DECLARE_BUF(outbuf, MC_CMD_GET_PHY_STATE_OUT_LEN); + size_t outlen; + int rc; + + BUILD_BUG_ON(MC_CMD_GET_PHY_STATE_IN_LEN != 0); + + rc = efx_mcdi_rpc(efx, MC_CMD_GET_PHY_STATE, NULL, 0, + outbuf, sizeof(outbuf), &outlen); + if (rc) + return rc; + + if (outlen < MC_CMD_GET_PHY_STATE_OUT_LEN) + return -EIO; + if (MCDI_DWORD(outbuf, GET_PHY_STATE_OUT_STATE) != MC_CMD_PHY_STATE_OK) + return -EINVAL; + + return 0; +} + +int efx_mcdi_set_mac(struct efx_nic *efx) +{ + u32 fcntl; + MCDI_DECLARE_BUF(cmdbytes, MC_CMD_SET_MAC_IN_LEN); + + BUILD_BUG_ON(MC_CMD_SET_MAC_OUT_LEN != 0); + + /* This has no effect on EF10 */ + ether_addr_copy(MCDI_PTR(cmdbytes, SET_MAC_IN_ADDR), + efx->net_dev->dev_addr); + + MCDI_SET_DWORD(cmdbytes, SET_MAC_IN_MTU, + EFX_MAX_FRAME_LEN(efx->net_dev->mtu)); + MCDI_SET_DWORD(cmdbytes, SET_MAC_IN_DRAIN, 0); + + /* Set simple MAC filter for Siena */ + MCDI_POPULATE_DWORD_1(cmdbytes, SET_MAC_IN_REJECT, + SET_MAC_IN_REJECT_UNCST, efx->unicast_filter); + + MCDI_POPULATE_DWORD_1(cmdbytes, SET_MAC_IN_FLAGS, + SET_MAC_IN_FLAG_INCLUDE_FCS, + !!(efx->net_dev->features & NETIF_F_RXFCS)); + + switch (efx->wanted_fc) { + case EFX_FC_RX | EFX_FC_TX: + fcntl = MC_CMD_FCNTL_BIDIR; + break; + case EFX_FC_RX: + fcntl = MC_CMD_FCNTL_RESPOND; + break; + default: + fcntl = MC_CMD_FCNTL_OFF; + break; + } + if (efx->wanted_fc & EFX_FC_AUTO) + fcntl = MC_CMD_FCNTL_AUTO; + if (efx->fc_disable) + fcntl = MC_CMD_FCNTL_OFF; + + MCDI_SET_DWORD(cmdbytes, SET_MAC_IN_FCNTL, fcntl); + + return efx_mcdi_rpc(efx, MC_CMD_SET_MAC, cmdbytes, sizeof(cmdbytes), + NULL, 0, NULL); +} + +/* Get physical port number (EF10 only; on Siena it is same as PF number) */ +int efx_mcdi_port_get_number(struct efx_nic *efx) +{ + MCDI_DECLARE_BUF(outbuf, MC_CMD_GET_PORT_ASSIGNMENT_OUT_LEN); + int rc; + + rc = efx_mcdi_rpc(efx, MC_CMD_GET_PORT_ASSIGNMENT, NULL, 0, + outbuf, sizeof(outbuf), NULL); + if (rc) + return rc; + + return MCDI_DWORD(outbuf, GET_PORT_ASSIGNMENT_OUT_PORT); +} + +static unsigned int efx_mcdi_event_link_speed[] = { + [MCDI_EVENT_LINKCHANGE_SPEED_100M] = 100, + [MCDI_EVENT_LINKCHANGE_SPEED_1G] = 1000, + [MCDI_EVENT_LINKCHANGE_SPEED_10G] = 10000, + [MCDI_EVENT_LINKCHANGE_SPEED_40G] = 40000, + [MCDI_EVENT_LINKCHANGE_SPEED_25G] = 25000, + [MCDI_EVENT_LINKCHANGE_SPEED_50G] = 50000, + [MCDI_EVENT_LINKCHANGE_SPEED_100G] = 100000, +}; + +void efx_mcdi_process_link_change(struct efx_nic *efx, efx_qword_t *ev) +{ + u32 flags, fcntl, speed, lpa; + + speed = EFX_QWORD_FIELD(*ev, MCDI_EVENT_LINKCHANGE_SPEED); + EFX_WARN_ON_PARANOID(speed >= ARRAY_SIZE(efx_mcdi_event_link_speed)); + speed = efx_mcdi_event_link_speed[speed]; + + flags = EFX_QWORD_FIELD(*ev, MCDI_EVENT_LINKCHANGE_LINK_FLAGS); + fcntl = EFX_QWORD_FIELD(*ev, MCDI_EVENT_LINKCHANGE_FCNTL); + lpa = EFX_QWORD_FIELD(*ev, MCDI_EVENT_LINKCHANGE_LP_CAP); + + /* efx->link_state is only modified by efx_mcdi_phy_get_link(), + * which is only run after flushing the event queues. Therefore, it + * is safe to modify the link state outside of the mac_lock here. + */ + efx_mcdi_phy_decode_link(efx, &efx->link_state, speed, flags, fcntl); + + efx_mcdi_phy_check_fcntl(efx, lpa); + + efx_link_status_changed(efx); +} diff --git a/drivers/net/ethernet/sfc/mcdi_port_common.h b/drivers/net/ethernet/sfc/mcdi_port_common.h new file mode 100644 index 000000000000..b16f11265269 --- /dev/null +++ b/drivers/net/ethernet/sfc/mcdi_port_common.h @@ -0,0 +1,57 @@ +/* SPDX-License-Identifier: GPL-2.0-only */ +/**************************************************************************** + * Driver for Solarflare network controllers and boards + * Copyright 2018 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 + * by the Free Software Foundation, incorporated herein by reference. + */ +#ifndef EFX_MCDI_PORT_COMMON_H +#define EFX_MCDI_PORT_COMMON_H + +#include "net_driver.h" +#include "mcdi.h" +#include "mcdi_pcol.h" + +struct efx_mcdi_phy_data { + u32 flags; + u32 type; + u32 supported_cap; + u32 channel; + u32 port; + u32 stats_mask; + u8 name[20]; + u32 media; + u32 mmd_mask; + u8 revision[20]; + u32 forced_cap; +}; + +#define EFX_MC_STATS_GENERATION_INVALID ((__force __le64)(-1)) + +int efx_mcdi_get_phy_cfg(struct efx_nic *efx, struct efx_mcdi_phy_data *cfg); +void efx_link_set_advertising(struct efx_nic *efx, + const unsigned long *advertising); +int efx_mcdi_set_link(struct efx_nic *efx, u32 capabilities, + u32 flags, u32 loopback_mode, u32 loopback_speed); +int efx_mcdi_loopback_modes(struct efx_nic *efx, u64 *loopback_modes); +void mcdi_to_ethtool_linkset(u32 media, u32 cap, unsigned long *linkset); +u32 ethtool_linkset_to_mcdi_cap(const unsigned long *linkset); +u32 efx_get_mcdi_phy_flags(struct efx_nic *efx); +u8 mcdi_to_ethtool_media(u32 media); +void efx_mcdi_phy_decode_link(struct efx_nic *efx, + struct efx_link_state *link_state, + u32 speed, u32 flags, u32 fcntl); +u32 ethtool_fec_caps_to_mcdi(u32 ethtool_cap); +u32 mcdi_fec_caps_to_ethtool(u32 caps, bool is_25g); +void efx_mcdi_phy_check_fcntl(struct efx_nic *efx, u32 lpa); +bool efx_mcdi_phy_poll(struct efx_nic *efx); +int efx_mcdi_phy_get_fecparam(struct efx_nic *efx, + struct ethtool_fecparam *fec); +int efx_mcdi_phy_test_alive(struct efx_nic *efx); +int efx_mcdi_set_mac(struct efx_nic *efx); +int efx_mcdi_port_get_number(struct efx_nic *efx); +void efx_mcdi_process_link_change(struct efx_nic *efx, efx_qword_t *ev); + +#endif diff --git a/drivers/net/ethernet/sfc/net_driver.h b/drivers/net/ethernet/sfc/net_driver.h index 709172a6995e..9f9886f222c8 100644 --- a/drivers/net/ethernet/sfc/net_driver.h +++ b/drivers/net/ethernet/sfc/net_driver.h @@ -138,6 +138,8 @@ struct efx_special_buffer { * freed when descriptor completes * @xdpf: When @flags & %EFX_TX_BUF_XDP, the XDP frame information; its @data * member is the associated buffer to drop a page reference on. + * @option: When @flags & %EFX_TX_BUF_OPTION, an EF10-specific option + * descriptor. * @dma_addr: DMA address of the fragment. * @flags: Flags for allocation and DMA mapping type * @len: Length of this fragment. @@ -152,7 +154,7 @@ struct efx_tx_buffer { struct xdp_frame *xdpf; }; union { - efx_qword_t option; + efx_qword_t option; /* EF10 */ dma_addr_t dma_addr; }; unsigned short flags; @@ -742,13 +744,13 @@ union efx_multicast_hash { struct vfdi_status; /* The reserved RSS context value */ -#define EFX_EF10_RSS_CONTEXT_INVALID 0xffffffff +#define EFX_MCDI_RSS_CONTEXT_INVALID 0xffffffff /** * struct efx_rss_context - A user-defined RSS context for filtering * @list: node of linked list on which this struct is stored * @context_id: the RSS_CONTEXT_ID returned by MC firmware, or - * %EFX_EF10_RSS_CONTEXT_INVALID if this context is not present on the NIC. - * For Siena, 0 if RSS is active, else %EFX_EF10_RSS_CONTEXT_INVALID. + * %EFX_MCDI_RSS_CONTEXT_INVALID if this context is not present on the NIC. + * For Siena, 0 if RSS is active, else %EFX_MCDI_RSS_CONTEXT_INVALID. * @user_id: the rss_context ID exposed to userspace over ethtool. * @rx_hash_udp_4tuple: UDP 4-tuple hashing enabled * @rx_hash_key: Toeplitz hash key for this RSS context @@ -1610,6 +1612,15 @@ static inline struct efx_rx_buffer *efx_rx_buffer(struct efx_rx_queue *rx_queue, return &rx_queue->buffer[index]; } +static inline struct efx_rx_buffer * +efx_rx_buf_next(struct efx_rx_queue *rx_queue, struct efx_rx_buffer *rx_buf) +{ + if (unlikely(rx_buf == efx_rx_buffer(rx_queue, rx_queue->ptr_mask))) + return efx_rx_buffer(rx_queue, 0); + else + return rx_buf + 1; +} + /** * EFX_MAX_FRAME_LEN - calculate maximum frame length * diff --git a/drivers/net/ethernet/sfc/nic.h b/drivers/net/ethernet/sfc/nic.h index bf0bdb22cc64..6670fda8f35a 100644 --- a/drivers/net/ethernet/sfc/nic.h +++ b/drivers/net/ethernet/sfc/nic.h @@ -11,6 +11,7 @@ #include <linux/net_tstamp.h> #include "net_driver.h" #include "efx.h" +#include "efx_common.h" #include "mcdi.h" enum { @@ -505,6 +506,9 @@ static inline void efx_nic_push_buffers(struct efx_tx_queue *tx_queue) tx_queue->efx->type->tx_write(tx_queue); } +int efx_enqueue_skb_tso(struct efx_tx_queue *tx_queue, struct sk_buff *skb, + bool *data_mapped); + /* RX data path */ static inline int efx_nic_probe_rx(struct efx_rx_queue *rx_queue) { @@ -553,6 +557,7 @@ 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); /* Falcon/Siena queue operations */ @@ -670,6 +675,7 @@ struct efx_farch_register_test { unsigned address; efx_oword_t mask; }; + int efx_farch_test_registers(struct efx_nic *efx, const struct efx_farch_register_test *regs, size_t n_regs); diff --git a/drivers/net/ethernet/sfc/rx.c b/drivers/net/ethernet/sfc/rx.c index c29bf862a94c..a2042f16babc 100644 --- a/drivers/net/ethernet/sfc/rx.c +++ b/drivers/net/ethernet/sfc/rx.c @@ -21,6 +21,7 @@ #include <linux/bpf_trace.h> #include "net_driver.h" #include "efx.h" +#include "rx_common.h" #include "filter.h" #include "nic.h" #include "selftest.h" @@ -32,60 +33,13 @@ /* Maximum rx prefix used by any architecture. */ #define EFX_MAX_RX_PREFIX_SIZE 16 -/* Number of RX buffers to recycle pages for. When creating the RX page recycle - * ring, this number is divided by the number of buffers per page to calculate - * the number of pages to store in the RX page recycle ring. - */ -#define EFX_RECYCLE_RING_SIZE_IOMMU 4096 -#define EFX_RECYCLE_RING_SIZE_NOIOMMU (2 * EFX_RX_PREFERRED_BATCH) - /* Size of buffer allocated for skb header area. */ #define EFX_SKB_HEADERS 128u -/* This is the percentage fill level below which new RX descriptors - * will be added to the RX descriptor ring. - */ -static unsigned int rx_refill_threshold; - /* Each packet can consume up to ceil(max_frame_len / buffer_size) buffers */ #define EFX_RX_MAX_FRAGS DIV_ROUND_UP(EFX_MAX_FRAME_LEN(EFX_MAX_MTU), \ EFX_RX_USR_BUF_SIZE) -/* - * RX maximum head room required. - * - * This must be at least 1 to prevent overflow, plus one packet-worth - * to allow pipelined receives. - */ -#define EFX_RXD_HEAD_ROOM (1 + EFX_RX_MAX_FRAGS) - -static inline u8 *efx_rx_buf_va(struct efx_rx_buffer *buf) -{ - return page_address(buf->page) + buf->page_offset; -} - -static inline u32 efx_rx_buf_hash(struct efx_nic *efx, const u8 *eh) -{ -#if defined(CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS) - return __le32_to_cpup((const __le32 *)(eh + efx->rx_packet_hash_offset)); -#else - const u8 *data = eh + efx->rx_packet_hash_offset; - return (u32)data[0] | - (u32)data[1] << 8 | - (u32)data[2] << 16 | - (u32)data[3] << 24; -#endif -} - -static inline struct efx_rx_buffer * -efx_rx_buf_next(struct efx_rx_queue *rx_queue, struct efx_rx_buffer *rx_buf) -{ - if (unlikely(rx_buf == efx_rx_buffer(rx_queue, rx_queue->ptr_mask))) - return efx_rx_buffer(rx_queue, 0); - else - return rx_buf + 1; -} - static inline void efx_sync_rx_buffer(struct efx_nic *efx, struct efx_rx_buffer *rx_buf, unsigned int len) @@ -94,301 +48,6 @@ static inline void efx_sync_rx_buffer(struct efx_nic *efx, DMA_FROM_DEVICE); } -void efx_rx_config_page_split(struct efx_nic *efx) -{ - efx->rx_page_buf_step = ALIGN(efx->rx_dma_len + efx->rx_ip_align + - XDP_PACKET_HEADROOM, - EFX_RX_BUF_ALIGNMENT); - efx->rx_bufs_per_page = efx->rx_buffer_order ? 1 : - ((PAGE_SIZE - sizeof(struct efx_rx_page_state)) / - efx->rx_page_buf_step); - efx->rx_buffer_truesize = (PAGE_SIZE << efx->rx_buffer_order) / - efx->rx_bufs_per_page; - efx->rx_pages_per_batch = DIV_ROUND_UP(EFX_RX_PREFERRED_BATCH, - efx->rx_bufs_per_page); -} - -/* Check the RX page recycle ring for a page that can be reused. */ -static struct page *efx_reuse_page(struct efx_rx_queue *rx_queue) -{ - struct efx_nic *efx = rx_queue->efx; - struct page *page; - struct efx_rx_page_state *state; - unsigned index; - - index = rx_queue->page_remove & rx_queue->page_ptr_mask; - page = rx_queue->page_ring[index]; - if (page == NULL) - return NULL; - - rx_queue->page_ring[index] = NULL; - /* page_remove cannot exceed page_add. */ - if (rx_queue->page_remove != rx_queue->page_add) - ++rx_queue->page_remove; - - /* If page_count is 1 then we hold the only reference to this page. */ - if (page_count(page) == 1) { - ++rx_queue->page_recycle_count; - return page; - } else { - state = page_address(page); - dma_unmap_page(&efx->pci_dev->dev, state->dma_addr, - PAGE_SIZE << efx->rx_buffer_order, - DMA_FROM_DEVICE); - put_page(page); - ++rx_queue->page_recycle_failed; - } - - return NULL; -} - -/** - * efx_init_rx_buffers - create EFX_RX_BATCH page-based RX buffers - * - * @rx_queue: Efx RX queue - * - * This allocates a batch of pages, maps them for DMA, and populates - * struct efx_rx_buffers for each one. Return a negative error code or - * 0 on success. If a single page can be used for multiple buffers, - * then the page will either be inserted fully, or not at all. - */ -static int efx_init_rx_buffers(struct efx_rx_queue *rx_queue, bool atomic) -{ - struct efx_nic *efx = rx_queue->efx; - struct efx_rx_buffer *rx_buf; - struct page *page; - unsigned int page_offset; - struct efx_rx_page_state *state; - dma_addr_t dma_addr; - unsigned index, count; - - count = 0; - do { - page = efx_reuse_page(rx_queue); - if (page == NULL) { - page = alloc_pages(__GFP_COMP | - (atomic ? GFP_ATOMIC : GFP_KERNEL), - efx->rx_buffer_order); - if (unlikely(page == NULL)) - return -ENOMEM; - dma_addr = - dma_map_page(&efx->pci_dev->dev, page, 0, - PAGE_SIZE << efx->rx_buffer_order, - DMA_FROM_DEVICE); - if (unlikely(dma_mapping_error(&efx->pci_dev->dev, - dma_addr))) { - __free_pages(page, efx->rx_buffer_order); - return -EIO; - } - state = page_address(page); - state->dma_addr = dma_addr; - } else { - state = page_address(page); - dma_addr = state->dma_addr; - } - - dma_addr += sizeof(struct efx_rx_page_state); - page_offset = sizeof(struct efx_rx_page_state); - - do { - index = rx_queue->added_count & rx_queue->ptr_mask; - rx_buf = efx_rx_buffer(rx_queue, index); - rx_buf->dma_addr = dma_addr + efx->rx_ip_align + - XDP_PACKET_HEADROOM; - rx_buf->page = page; - rx_buf->page_offset = page_offset + efx->rx_ip_align + - XDP_PACKET_HEADROOM; - rx_buf->len = efx->rx_dma_len; - rx_buf->flags = 0; - ++rx_queue->added_count; - get_page(page); - dma_addr += efx->rx_page_buf_step; - page_offset += efx->rx_page_buf_step; - } while (page_offset + efx->rx_page_buf_step <= PAGE_SIZE); - - rx_buf->flags = EFX_RX_BUF_LAST_IN_PAGE; - } while (++count < efx->rx_pages_per_batch); - - return 0; -} - -/* Unmap a DMA-mapped page. This function is only called for the final RX - * buffer in a page. - */ -static void efx_unmap_rx_buffer(struct efx_nic *efx, - struct efx_rx_buffer *rx_buf) -{ - struct page *page = rx_buf->page; - - if (page) { - struct efx_rx_page_state *state = page_address(page); - dma_unmap_page(&efx->pci_dev->dev, - state->dma_addr, - PAGE_SIZE << efx->rx_buffer_order, - DMA_FROM_DEVICE); - } -} - -static void efx_free_rx_buffers(struct efx_rx_queue *rx_queue, - struct efx_rx_buffer *rx_buf, - unsigned int num_bufs) -{ - do { - if (rx_buf->page) { - put_page(rx_buf->page); - rx_buf->page = NULL; - } - rx_buf = efx_rx_buf_next(rx_queue, rx_buf); - } while (--num_bufs); -} - -/* Attempt to recycle the page if there is an RX recycle ring; the page can - * only be added if this is the final RX buffer, to prevent pages being used in - * the descriptor ring and appearing in the recycle ring simultaneously. - */ -static void efx_recycle_rx_page(struct efx_channel *channel, - struct efx_rx_buffer *rx_buf) -{ - struct page *page = rx_buf->page; - struct efx_rx_queue *rx_queue = efx_channel_get_rx_queue(channel); - struct efx_nic *efx = rx_queue->efx; - unsigned index; - - /* Only recycle the page after processing the final buffer. */ - if (!(rx_buf->flags & EFX_RX_BUF_LAST_IN_PAGE)) - return; - - index = rx_queue->page_add & rx_queue->page_ptr_mask; - if (rx_queue->page_ring[index] == NULL) { - unsigned read_index = rx_queue->page_remove & - rx_queue->page_ptr_mask; - - /* The next slot in the recycle ring is available, but - * increment page_remove if the read pointer currently - * points here. - */ - if (read_index == index) - ++rx_queue->page_remove; - rx_queue->page_ring[index] = page; - ++rx_queue->page_add; - return; - } - ++rx_queue->page_recycle_full; - efx_unmap_rx_buffer(efx, rx_buf); - put_page(rx_buf->page); -} - -static void efx_fini_rx_buffer(struct efx_rx_queue *rx_queue, - struct efx_rx_buffer *rx_buf) -{ - /* Release the page reference we hold for the buffer. */ - if (rx_buf->page) - put_page(rx_buf->page); - - /* If this is the last buffer in a page, unmap and free it. */ - if (rx_buf->flags & EFX_RX_BUF_LAST_IN_PAGE) { - efx_unmap_rx_buffer(rx_queue->efx, rx_buf); - efx_free_rx_buffers(rx_queue, rx_buf, 1); - } - rx_buf->page = NULL; -} - -/* Recycle the pages that are used by buffers that have just been received. */ -static void efx_recycle_rx_pages(struct efx_channel *channel, - struct efx_rx_buffer *rx_buf, - unsigned int n_frags) -{ - struct efx_rx_queue *rx_queue = efx_channel_get_rx_queue(channel); - - do { - efx_recycle_rx_page(channel, rx_buf); - rx_buf = efx_rx_buf_next(rx_queue, rx_buf); - } while (--n_frags); -} - -static void efx_discard_rx_packet(struct efx_channel *channel, - struct efx_rx_buffer *rx_buf, - unsigned int n_frags) -{ - struct efx_rx_queue *rx_queue = efx_channel_get_rx_queue(channel); - - efx_recycle_rx_pages(channel, rx_buf, n_frags); - - efx_free_rx_buffers(rx_queue, rx_buf, n_frags); -} - -/** - * efx_fast_push_rx_descriptors - push new RX descriptors quickly - * @rx_queue: RX descriptor queue - * - * This will aim to fill the RX descriptor queue up to - * @rx_queue->@max_fill. If there is insufficient atomic - * memory to do so, a slow fill will be scheduled. - * - * The caller must provide serialisation (none is used here). In practise, - * this means this function must run from the NAPI handler, or be called - * when NAPI is disabled. - */ -void efx_fast_push_rx_descriptors(struct efx_rx_queue *rx_queue, bool atomic) -{ - struct efx_nic *efx = rx_queue->efx; - unsigned int fill_level, batch_size; - int space, rc = 0; - - if (!rx_queue->refill_enabled) - return; - - /* Calculate current fill level, and exit if we don't need to fill */ - fill_level = (rx_queue->added_count - rx_queue->removed_count); - EFX_WARN_ON_ONCE_PARANOID(fill_level > rx_queue->efx->rxq_entries); - if (fill_level >= rx_queue->fast_fill_trigger) - goto out; - - /* Record minimum fill level */ - if (unlikely(fill_level < rx_queue->min_fill)) { - if (fill_level) - rx_queue->min_fill = fill_level; - } - - batch_size = efx->rx_pages_per_batch * efx->rx_bufs_per_page; - space = rx_queue->max_fill - fill_level; - EFX_WARN_ON_ONCE_PARANOID(space < batch_size); - - netif_vdbg(rx_queue->efx, rx_status, rx_queue->efx->net_dev, - "RX queue %d fast-filling descriptor ring from" - " level %d to level %d\n", - efx_rx_queue_index(rx_queue), fill_level, - rx_queue->max_fill); - - - do { - rc = efx_init_rx_buffers(rx_queue, atomic); - if (unlikely(rc)) { - /* Ensure that we don't leave the rx queue empty */ - efx_schedule_slow_fill(rx_queue); - goto out; - } - } while ((space -= batch_size) >= batch_size); - - netif_vdbg(rx_queue->efx, rx_status, rx_queue->efx->net_dev, - "RX queue %d fast-filled descriptor ring " - "to level %d\n", efx_rx_queue_index(rx_queue), - rx_queue->added_count - rx_queue->removed_count); - - out: - if (rx_queue->notified_count != rx_queue->added_count) - efx_nic_notify_rx_desc(rx_queue); -} - -void efx_rx_slow_fill(struct timer_list *t) -{ - struct efx_rx_queue *rx_queue = from_timer(rx_queue, t, slow_fill); - - /* Post an event to cause NAPI to run and refill the queue */ - efx_nic_generate_fill_event(rx_queue); - ++rx_queue->slow_fill_count; -} - static void efx_rx_packet__check_len(struct efx_rx_queue *rx_queue, struct efx_rx_buffer *rx_buf, int len) @@ -412,53 +71,6 @@ static void efx_rx_packet__check_len(struct efx_rx_queue *rx_queue, efx_rx_queue_channel(rx_queue)->n_rx_overlength++; } -/* Pass a received packet up through GRO. GRO can handle pages - * regardless of checksum state and skbs with a good checksum. - */ -static void -efx_rx_packet_gro(struct efx_channel *channel, struct efx_rx_buffer *rx_buf, - unsigned int n_frags, u8 *eh) -{ - struct napi_struct *napi = &channel->napi_str; - struct efx_nic *efx = channel->efx; - struct sk_buff *skb; - - skb = napi_get_frags(napi); - if (unlikely(!skb)) { - struct efx_rx_queue *rx_queue; - - rx_queue = efx_channel_get_rx_queue(channel); - efx_free_rx_buffers(rx_queue, rx_buf, n_frags); - return; - } - - if (efx->net_dev->features & NETIF_F_RXHASH) - skb_set_hash(skb, efx_rx_buf_hash(efx, eh), - PKT_HASH_TYPE_L3); - skb->ip_summed = ((rx_buf->flags & EFX_RX_PKT_CSUMMED) ? - CHECKSUM_UNNECESSARY : CHECKSUM_NONE); - skb->csum_level = !!(rx_buf->flags & EFX_RX_PKT_CSUM_LEVEL); - - for (;;) { - skb_fill_page_desc(skb, skb_shinfo(skb)->nr_frags, - rx_buf->page, rx_buf->page_offset, - rx_buf->len); - rx_buf->page = NULL; - skb->len += rx_buf->len; - if (skb_shinfo(skb)->nr_frags == n_frags) - break; - - rx_buf = efx_rx_buf_next(&channel->rx_queue, rx_buf); - } - - skb->data_len = skb->len; - skb->truesize += n_frags * efx->rx_buffer_truesize; - - skb_record_rx_queue(skb, channel->rx_queue.core_index); - - napi_gro_frags(napi); -} - /* Allocate and construct an SKB around page fragments */ static struct sk_buff *efx_rx_mk_skb(struct efx_channel *channel, struct efx_rx_buffer *rx_buf, @@ -805,174 +417,6 @@ out: channel->rx_pkt_n_frags = 0; } -int efx_probe_rx_queue(struct efx_rx_queue *rx_queue) -{ - struct efx_nic *efx = rx_queue->efx; - unsigned int entries; - int rc; - - /* Create the smallest power-of-two aligned ring */ - entries = max(roundup_pow_of_two(efx->rxq_entries), EFX_MIN_DMAQ_SIZE); - EFX_WARN_ON_PARANOID(entries > EFX_MAX_DMAQ_SIZE); - rx_queue->ptr_mask = entries - 1; - - netif_dbg(efx, probe, efx->net_dev, - "creating RX queue %d size %#x mask %#x\n", - efx_rx_queue_index(rx_queue), efx->rxq_entries, - rx_queue->ptr_mask); - - /* Allocate RX buffers */ - rx_queue->buffer = kcalloc(entries, sizeof(*rx_queue->buffer), - GFP_KERNEL); - if (!rx_queue->buffer) - return -ENOMEM; - - rc = efx_nic_probe_rx(rx_queue); - if (rc) { - kfree(rx_queue->buffer); - rx_queue->buffer = NULL; - } - - return rc; -} - -static void efx_init_rx_recycle_ring(struct efx_nic *efx, - struct efx_rx_queue *rx_queue) -{ - unsigned int bufs_in_recycle_ring, page_ring_size; - - /* Set the RX recycle ring size */ -#ifdef CONFIG_PPC64 - bufs_in_recycle_ring = EFX_RECYCLE_RING_SIZE_IOMMU; -#else - if (iommu_present(&pci_bus_type)) - bufs_in_recycle_ring = EFX_RECYCLE_RING_SIZE_IOMMU; - else - bufs_in_recycle_ring = EFX_RECYCLE_RING_SIZE_NOIOMMU; -#endif /* CONFIG_PPC64 */ - - page_ring_size = roundup_pow_of_two(bufs_in_recycle_ring / - efx->rx_bufs_per_page); - rx_queue->page_ring = kcalloc(page_ring_size, - sizeof(*rx_queue->page_ring), GFP_KERNEL); - rx_queue->page_ptr_mask = page_ring_size - 1; -} - -void efx_init_rx_queue(struct efx_rx_queue *rx_queue) -{ - struct efx_nic *efx = rx_queue->efx; - unsigned int max_fill, trigger, max_trigger; - int rc = 0; - - netif_dbg(rx_queue->efx, drv, rx_queue->efx->net_dev, - "initialising RX queue %d\n", efx_rx_queue_index(rx_queue)); - - /* Initialise ptr fields */ - rx_queue->added_count = 0; - rx_queue->notified_count = 0; - rx_queue->removed_count = 0; - rx_queue->min_fill = -1U; - efx_init_rx_recycle_ring(efx, rx_queue); - - rx_queue->page_remove = 0; - rx_queue->page_add = rx_queue->page_ptr_mask + 1; - rx_queue->page_recycle_count = 0; - rx_queue->page_recycle_failed = 0; - rx_queue->page_recycle_full = 0; - - /* Initialise limit fields */ - max_fill = efx->rxq_entries - EFX_RXD_HEAD_ROOM; - max_trigger = - max_fill - efx->rx_pages_per_batch * efx->rx_bufs_per_page; - if (rx_refill_threshold != 0) { - trigger = max_fill * min(rx_refill_threshold, 100U) / 100U; - if (trigger > max_trigger) - trigger = max_trigger; - } else { - trigger = max_trigger; - } - - rx_queue->max_fill = max_fill; - rx_queue->fast_fill_trigger = trigger; - rx_queue->refill_enabled = true; - - /* Initialise XDP queue information */ - rc = xdp_rxq_info_reg(&rx_queue->xdp_rxq_info, efx->net_dev, - rx_queue->core_index); - - if (rc) { - netif_err(efx, rx_err, efx->net_dev, - "Failure to initialise XDP queue information rc=%d\n", - rc); - efx->xdp_rxq_info_failed = true; - } else { - rx_queue->xdp_rxq_info_valid = true; - } - - /* Set up RX descriptor ring */ - efx_nic_init_rx(rx_queue); -} - -void efx_fini_rx_queue(struct efx_rx_queue *rx_queue) -{ - int i; - struct efx_nic *efx = rx_queue->efx; - struct efx_rx_buffer *rx_buf; - - netif_dbg(rx_queue->efx, drv, rx_queue->efx->net_dev, - "shutting down RX queue %d\n", efx_rx_queue_index(rx_queue)); - - del_timer_sync(&rx_queue->slow_fill); - - /* Release RX buffers from the current read ptr to the write ptr */ - if (rx_queue->buffer) { - for (i = rx_queue->removed_count; i < rx_queue->added_count; - i++) { - unsigned index = i & rx_queue->ptr_mask; - rx_buf = efx_rx_buffer(rx_queue, index); - efx_fini_rx_buffer(rx_queue, rx_buf); - } - } - - /* Unmap and release the pages in the recycle ring. Remove the ring. */ - for (i = 0; i <= rx_queue->page_ptr_mask; i++) { - struct page *page = rx_queue->page_ring[i]; - struct efx_rx_page_state *state; - - if (page == NULL) - continue; - - state = page_address(page); - dma_unmap_page(&efx->pci_dev->dev, state->dma_addr, - PAGE_SIZE << efx->rx_buffer_order, - DMA_FROM_DEVICE); - put_page(page); - } - kfree(rx_queue->page_ring); - rx_queue->page_ring = NULL; - - if (rx_queue->xdp_rxq_info_valid) - xdp_rxq_info_unreg(&rx_queue->xdp_rxq_info); - - rx_queue->xdp_rxq_info_valid = false; -} - -void efx_remove_rx_queue(struct efx_rx_queue *rx_queue) -{ - netif_dbg(rx_queue->efx, drv, rx_queue->efx->net_dev, - "destroying RX queue %d\n", efx_rx_queue_index(rx_queue)); - - efx_nic_remove_rx(rx_queue); - - kfree(rx_queue->buffer); - rx_queue->buffer = NULL; -} - - -module_param(rx_refill_threshold, uint, 0444); -MODULE_PARM_DESC(rx_refill_threshold, - "RX descriptor ring refill threshold (%)"); - #ifdef CONFIG_RFS_ACCEL static void efx_filter_rfs_work(struct work_struct *data) @@ -1206,37 +650,3 @@ bool __efx_filter_rfs_expire(struct efx_channel *channel, unsigned int quota) } #endif /* CONFIG_RFS_ACCEL */ - -/** - * efx_filter_is_mc_recipient - test whether spec is a multicast recipient - * @spec: Specification to test - * - * Return: %true if the specification is a non-drop RX filter that - * matches a local MAC address I/G bit value of 1 or matches a local - * IPv4 or IPv6 address value in the respective multicast address - * range. Otherwise %false. - */ -bool efx_filter_is_mc_recipient(const struct efx_filter_spec *spec) -{ - if (!(spec->flags & EFX_FILTER_FLAG_RX) || - spec->dmaq_id == EFX_FILTER_RX_DMAQ_ID_DROP) - return false; - - if (spec->match_flags & - (EFX_FILTER_MATCH_LOC_MAC | EFX_FILTER_MATCH_LOC_MAC_IG) && - is_multicast_ether_addr(spec->loc_mac)) - return true; - - if ((spec->match_flags & - (EFX_FILTER_MATCH_ETHER_TYPE | EFX_FILTER_MATCH_LOC_HOST)) == - (EFX_FILTER_MATCH_ETHER_TYPE | EFX_FILTER_MATCH_LOC_HOST)) { - if (spec->ether_type == htons(ETH_P_IP) && - ipv4_is_multicast(spec->loc_host[0])) - return true; - if (spec->ether_type == htons(ETH_P_IPV6) && - ((const u8 *)spec->loc_host)[0] == 0xff) - return true; - } - - return false; -} diff --git a/drivers/net/ethernet/sfc/rx_common.c b/drivers/net/ethernet/sfc/rx_common.c new file mode 100644 index 000000000000..ee8beb87bdc1 --- /dev/null +++ b/drivers/net/ethernet/sfc/rx_common.c @@ -0,0 +1,851 @@ +// SPDX-License-Identifier: GPL-2.0-only +/**************************************************************************** + * Driver for Solarflare network controllers and boards + * Copyright 2018 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 + * by the Free Software Foundation, incorporated herein by reference. + */ + +#include "net_driver.h" +#include <linux/module.h> +#include <linux/iommu.h> +#include "efx.h" +#include "nic.h" +#include "rx_common.h" + +/* This is the percentage fill level below which new RX descriptors + * will be added to the RX descriptor ring. + */ +static unsigned int rx_refill_threshold; +module_param(rx_refill_threshold, uint, 0444); +MODULE_PARM_DESC(rx_refill_threshold, + "RX descriptor ring refill threshold (%)"); + +/* Number of RX buffers to recycle pages for. When creating the RX page recycle + * ring, this number is divided by the number of buffers per page to calculate + * the number of pages to store in the RX page recycle ring. + */ +#define EFX_RECYCLE_RING_SIZE_IOMMU 4096 +#define EFX_RECYCLE_RING_SIZE_NOIOMMU (2 * EFX_RX_PREFERRED_BATCH) + +/* RX maximum head room required. + * + * This must be at least 1 to prevent overflow, plus one packet-worth + * to allow pipelined receives. + */ +#define EFX_RXD_HEAD_ROOM (1 + EFX_RX_MAX_FRAGS) + +/* Check the RX page recycle ring for a page that can be reused. */ +static struct page *efx_reuse_page(struct efx_rx_queue *rx_queue) +{ + struct efx_nic *efx = rx_queue->efx; + struct efx_rx_page_state *state; + unsigned int index; + struct page *page; + + index = rx_queue->page_remove & rx_queue->page_ptr_mask; + page = rx_queue->page_ring[index]; + if (page == NULL) + return NULL; + + rx_queue->page_ring[index] = NULL; + /* page_remove cannot exceed page_add. */ + if (rx_queue->page_remove != rx_queue->page_add) + ++rx_queue->page_remove; + + /* If page_count is 1 then we hold the only reference to this page. */ + if (page_count(page) == 1) { + ++rx_queue->page_recycle_count; + return page; + } else { + state = page_address(page); + dma_unmap_page(&efx->pci_dev->dev, state->dma_addr, + PAGE_SIZE << efx->rx_buffer_order, + DMA_FROM_DEVICE); + put_page(page); + ++rx_queue->page_recycle_failed; + } + + return NULL; +} + +/* Attempt to recycle the page if there is an RX recycle ring; the page can + * only be added if this is the final RX buffer, to prevent pages being used in + * the descriptor ring and appearing in the recycle ring simultaneously. + */ +static void efx_recycle_rx_page(struct efx_channel *channel, + struct efx_rx_buffer *rx_buf) +{ + struct efx_rx_queue *rx_queue = efx_channel_get_rx_queue(channel); + struct efx_nic *efx = rx_queue->efx; + struct page *page = rx_buf->page; + unsigned int index; + + /* Only recycle the page after processing the final buffer. */ + if (!(rx_buf->flags & EFX_RX_BUF_LAST_IN_PAGE)) + return; + + index = rx_queue->page_add & rx_queue->page_ptr_mask; + if (rx_queue->page_ring[index] == NULL) { + unsigned int read_index = rx_queue->page_remove & + rx_queue->page_ptr_mask; + + /* The next slot in the recycle ring is available, but + * increment page_remove if the read pointer currently + * points here. + */ + if (read_index == index) + ++rx_queue->page_remove; + rx_queue->page_ring[index] = page; + ++rx_queue->page_add; + return; + } + ++rx_queue->page_recycle_full; + efx_unmap_rx_buffer(efx, rx_buf); + put_page(rx_buf->page); +} + +/* Recycle the pages that are used by buffers that have just been received. */ +void efx_recycle_rx_pages(struct efx_channel *channel, + struct efx_rx_buffer *rx_buf, + unsigned int n_frags) +{ + struct efx_rx_queue *rx_queue = efx_channel_get_rx_queue(channel); + + do { + efx_recycle_rx_page(channel, rx_buf); + rx_buf = efx_rx_buf_next(rx_queue, rx_buf); + } while (--n_frags); +} + +void efx_discard_rx_packet(struct efx_channel *channel, + struct efx_rx_buffer *rx_buf, + unsigned int n_frags) +{ + struct efx_rx_queue *rx_queue = efx_channel_get_rx_queue(channel); + + efx_recycle_rx_pages(channel, rx_buf, n_frags); + + efx_free_rx_buffers(rx_queue, rx_buf, n_frags); +} + +static void efx_init_rx_recycle_ring(struct efx_rx_queue *rx_queue) +{ + unsigned int bufs_in_recycle_ring, page_ring_size; + struct efx_nic *efx = rx_queue->efx; + + /* Set the RX recycle ring size */ +#ifdef CONFIG_PPC64 + bufs_in_recycle_ring = EFX_RECYCLE_RING_SIZE_IOMMU; +#else + if (iommu_present(&pci_bus_type)) + bufs_in_recycle_ring = EFX_RECYCLE_RING_SIZE_IOMMU; + else + bufs_in_recycle_ring = EFX_RECYCLE_RING_SIZE_NOIOMMU; +#endif /* CONFIG_PPC64 */ + + page_ring_size = roundup_pow_of_two(bufs_in_recycle_ring / + efx->rx_bufs_per_page); + rx_queue->page_ring = kcalloc(page_ring_size, + sizeof(*rx_queue->page_ring), GFP_KERNEL); + rx_queue->page_ptr_mask = page_ring_size - 1; +} + +static void efx_fini_rx_recycle_ring(struct efx_rx_queue *rx_queue) +{ + struct efx_nic *efx = rx_queue->efx; + int i; + + /* Unmap and release the pages in the recycle ring. Remove the ring. */ + for (i = 0; i <= rx_queue->page_ptr_mask; i++) { + struct page *page = rx_queue->page_ring[i]; + struct efx_rx_page_state *state; + + if (page == NULL) + continue; + + state = page_address(page); + dma_unmap_page(&efx->pci_dev->dev, state->dma_addr, + PAGE_SIZE << efx->rx_buffer_order, + DMA_FROM_DEVICE); + put_page(page); + } + kfree(rx_queue->page_ring); + rx_queue->page_ring = NULL; +} + +static void efx_fini_rx_buffer(struct efx_rx_queue *rx_queue, + struct efx_rx_buffer *rx_buf) +{ + /* Release the page reference we hold for the buffer. */ + if (rx_buf->page) + put_page(rx_buf->page); + + /* If this is the last buffer in a page, unmap and free it. */ + if (rx_buf->flags & EFX_RX_BUF_LAST_IN_PAGE) { + efx_unmap_rx_buffer(rx_queue->efx, rx_buf); + efx_free_rx_buffers(rx_queue, rx_buf, 1); + } + rx_buf->page = NULL; +} + +int efx_probe_rx_queue(struct efx_rx_queue *rx_queue) +{ + struct efx_nic *efx = rx_queue->efx; + unsigned int entries; + int rc; + + /* Create the smallest power-of-two aligned ring */ + entries = max(roundup_pow_of_two(efx->rxq_entries), EFX_MIN_DMAQ_SIZE); + EFX_WARN_ON_PARANOID(entries > EFX_MAX_DMAQ_SIZE); + rx_queue->ptr_mask = entries - 1; + + netif_dbg(efx, probe, efx->net_dev, + "creating RX queue %d size %#x mask %#x\n", + efx_rx_queue_index(rx_queue), efx->rxq_entries, + rx_queue->ptr_mask); + + /* Allocate RX buffers */ + rx_queue->buffer = kcalloc(entries, sizeof(*rx_queue->buffer), + GFP_KERNEL); + if (!rx_queue->buffer) + return -ENOMEM; + + rc = efx_nic_probe_rx(rx_queue); + if (rc) { + kfree(rx_queue->buffer); + rx_queue->buffer = NULL; + } + + return rc; +} + +void efx_init_rx_queue(struct efx_rx_queue *rx_queue) +{ + unsigned int max_fill, trigger, max_trigger; + struct efx_nic *efx = rx_queue->efx; + int rc = 0; + + netif_dbg(rx_queue->efx, drv, rx_queue->efx->net_dev, + "initialising RX queue %d\n", efx_rx_queue_index(rx_queue)); + + /* Initialise ptr fields */ + rx_queue->added_count = 0; + rx_queue->notified_count = 0; + rx_queue->removed_count = 0; + rx_queue->min_fill = -1U; + efx_init_rx_recycle_ring(rx_queue); + + rx_queue->page_remove = 0; + rx_queue->page_add = rx_queue->page_ptr_mask + 1; + rx_queue->page_recycle_count = 0; + rx_queue->page_recycle_failed = 0; + rx_queue->page_recycle_full = 0; + + /* Initialise limit fields */ + max_fill = efx->rxq_entries - EFX_RXD_HEAD_ROOM; + max_trigger = + max_fill - efx->rx_pages_per_batch * efx->rx_bufs_per_page; + if (rx_refill_threshold != 0) { + trigger = max_fill * min(rx_refill_threshold, 100U) / 100U; + if (trigger > max_trigger) + trigger = max_trigger; + } else { + trigger = max_trigger; + } + + rx_queue->max_fill = max_fill; + rx_queue->fast_fill_trigger = trigger; + rx_queue->refill_enabled = true; + + /* Initialise XDP queue information */ + rc = xdp_rxq_info_reg(&rx_queue->xdp_rxq_info, efx->net_dev, + rx_queue->core_index); + + if (rc) { + netif_err(efx, rx_err, efx->net_dev, + "Failure to initialise XDP queue information rc=%d\n", + rc); + efx->xdp_rxq_info_failed = true; + } else { + rx_queue->xdp_rxq_info_valid = true; + } + + /* Set up RX descriptor ring */ + efx_nic_init_rx(rx_queue); +} + +void efx_fini_rx_queue(struct efx_rx_queue *rx_queue) +{ + struct efx_rx_buffer *rx_buf; + int i; + + netif_dbg(rx_queue->efx, drv, rx_queue->efx->net_dev, + "shutting down RX queue %d\n", efx_rx_queue_index(rx_queue)); + + del_timer_sync(&rx_queue->slow_fill); + + /* Release RX buffers from the current read ptr to the write ptr */ + if (rx_queue->buffer) { + for (i = rx_queue->removed_count; i < rx_queue->added_count; + i++) { + unsigned int index = i & rx_queue->ptr_mask; + + rx_buf = efx_rx_buffer(rx_queue, index); + efx_fini_rx_buffer(rx_queue, rx_buf); + } + } + + efx_fini_rx_recycle_ring(rx_queue); + + if (rx_queue->xdp_rxq_info_valid) + xdp_rxq_info_unreg(&rx_queue->xdp_rxq_info); + + rx_queue->xdp_rxq_info_valid = false; +} + +void efx_remove_rx_queue(struct efx_rx_queue *rx_queue) +{ + netif_dbg(rx_queue->efx, drv, rx_queue->efx->net_dev, + "destroying RX queue %d\n", efx_rx_queue_index(rx_queue)); + + efx_nic_remove_rx(rx_queue); + + kfree(rx_queue->buffer); + rx_queue->buffer = NULL; +} + +/* Unmap a DMA-mapped page. This function is only called for the final RX + * buffer in a page. + */ +void efx_unmap_rx_buffer(struct efx_nic *efx, + struct efx_rx_buffer *rx_buf) +{ + struct page *page = rx_buf->page; + + if (page) { + struct efx_rx_page_state *state = page_address(page); + + dma_unmap_page(&efx->pci_dev->dev, + state->dma_addr, + PAGE_SIZE << efx->rx_buffer_order, + DMA_FROM_DEVICE); + } +} + +void efx_free_rx_buffers(struct efx_rx_queue *rx_queue, + struct efx_rx_buffer *rx_buf, + unsigned int num_bufs) +{ + do { + if (rx_buf->page) { + put_page(rx_buf->page); + rx_buf->page = NULL; + } + rx_buf = efx_rx_buf_next(rx_queue, rx_buf); + } while (--num_bufs); +} + +void efx_rx_slow_fill(struct timer_list *t) +{ + struct efx_rx_queue *rx_queue = from_timer(rx_queue, t, slow_fill); + + /* Post an event to cause NAPI to run and refill the queue */ + efx_nic_generate_fill_event(rx_queue); + ++rx_queue->slow_fill_count; +} + +void efx_schedule_slow_fill(struct efx_rx_queue *rx_queue) +{ + mod_timer(&rx_queue->slow_fill, jiffies + msecs_to_jiffies(10)); +} + +/* efx_init_rx_buffers - create EFX_RX_BATCH page-based RX buffers + * + * @rx_queue: Efx RX queue + * + * This allocates a batch of pages, maps them for DMA, and populates + * struct efx_rx_buffers for each one. Return a negative error code or + * 0 on success. If a single page can be used for multiple buffers, + * then the page will either be inserted fully, or not at all. + */ +static int efx_init_rx_buffers(struct efx_rx_queue *rx_queue, bool atomic) +{ + unsigned int page_offset, index, count; + struct efx_nic *efx = rx_queue->efx; + struct efx_rx_page_state *state; + struct efx_rx_buffer *rx_buf; + dma_addr_t dma_addr; + struct page *page; + + count = 0; + do { + page = efx_reuse_page(rx_queue); + if (page == NULL) { + page = alloc_pages(__GFP_COMP | + (atomic ? GFP_ATOMIC : GFP_KERNEL), + efx->rx_buffer_order); + if (unlikely(page == NULL)) + return -ENOMEM; + dma_addr = + dma_map_page(&efx->pci_dev->dev, page, 0, + PAGE_SIZE << efx->rx_buffer_order, + DMA_FROM_DEVICE); + if (unlikely(dma_mapping_error(&efx->pci_dev->dev, + dma_addr))) { + __free_pages(page, efx->rx_buffer_order); + return -EIO; + } + state = page_address(page); + state->dma_addr = dma_addr; + } else { + state = page_address(page); + dma_addr = state->dma_addr; + } + + dma_addr += sizeof(struct efx_rx_page_state); + page_offset = sizeof(struct efx_rx_page_state); + + do { + index = rx_queue->added_count & rx_queue->ptr_mask; + rx_buf = efx_rx_buffer(rx_queue, index); + rx_buf->dma_addr = dma_addr + efx->rx_ip_align + + XDP_PACKET_HEADROOM; + rx_buf->page = page; + rx_buf->page_offset = page_offset + efx->rx_ip_align + + XDP_PACKET_HEADROOM; + rx_buf->len = efx->rx_dma_len; + rx_buf->flags = 0; + ++rx_queue->added_count; + get_page(page); + dma_addr += efx->rx_page_buf_step; + page_offset += efx->rx_page_buf_step; + } while (page_offset + efx->rx_page_buf_step <= PAGE_SIZE); + + rx_buf->flags = EFX_RX_BUF_LAST_IN_PAGE; + } while (++count < efx->rx_pages_per_batch); + + return 0; +} + +void efx_rx_config_page_split(struct efx_nic *efx) +{ + efx->rx_page_buf_step = ALIGN(efx->rx_dma_len + efx->rx_ip_align + + XDP_PACKET_HEADROOM, + EFX_RX_BUF_ALIGNMENT); + efx->rx_bufs_per_page = efx->rx_buffer_order ? 1 : + ((PAGE_SIZE - sizeof(struct efx_rx_page_state)) / + efx->rx_page_buf_step); + efx->rx_buffer_truesize = (PAGE_SIZE << efx->rx_buffer_order) / + efx->rx_bufs_per_page; + efx->rx_pages_per_batch = DIV_ROUND_UP(EFX_RX_PREFERRED_BATCH, + efx->rx_bufs_per_page); +} + +/* efx_fast_push_rx_descriptors - push new RX descriptors quickly + * @rx_queue: RX descriptor queue + * + * This will aim to fill the RX descriptor queue up to + * @rx_queue->@max_fill. If there is insufficient atomic + * memory to do so, a slow fill will be scheduled. + * + * The caller must provide serialisation (none is used here). In practise, + * this means this function must run from the NAPI handler, or be called + * when NAPI is disabled. + */ +void efx_fast_push_rx_descriptors(struct efx_rx_queue *rx_queue, bool atomic) +{ + struct efx_nic *efx = rx_queue->efx; + unsigned int fill_level, batch_size; + int space, rc = 0; + + if (!rx_queue->refill_enabled) + return; + + /* Calculate current fill level, and exit if we don't need to fill */ + fill_level = (rx_queue->added_count - rx_queue->removed_count); + EFX_WARN_ON_ONCE_PARANOID(fill_level > rx_queue->efx->rxq_entries); + if (fill_level >= rx_queue->fast_fill_trigger) + goto out; + + /* Record minimum fill level */ + if (unlikely(fill_level < rx_queue->min_fill)) { + if (fill_level) + rx_queue->min_fill = fill_level; + } + + batch_size = efx->rx_pages_per_batch * efx->rx_bufs_per_page; + space = rx_queue->max_fill - fill_level; + EFX_WARN_ON_ONCE_PARANOID(space < batch_size); + + netif_vdbg(rx_queue->efx, rx_status, rx_queue->efx->net_dev, + "RX queue %d fast-filling descriptor ring from" + " level %d to level %d\n", + efx_rx_queue_index(rx_queue), fill_level, + rx_queue->max_fill); + + do { + rc = efx_init_rx_buffers(rx_queue, atomic); + if (unlikely(rc)) { + /* Ensure that we don't leave the rx queue empty */ + efx_schedule_slow_fill(rx_queue); + goto out; + } + } while ((space -= batch_size) >= batch_size); + + netif_vdbg(rx_queue->efx, rx_status, rx_queue->efx->net_dev, + "RX queue %d fast-filled descriptor ring " + "to level %d\n", efx_rx_queue_index(rx_queue), + rx_queue->added_count - rx_queue->removed_count); + + out: + if (rx_queue->notified_count != rx_queue->added_count) + efx_nic_notify_rx_desc(rx_queue); +} + +/* Pass a received packet up through GRO. GRO can handle pages + * regardless of checksum state and skbs with a good checksum. + */ +void +efx_rx_packet_gro(struct efx_channel *channel, struct efx_rx_buffer *rx_buf, + unsigned int n_frags, u8 *eh) +{ + struct napi_struct *napi = &channel->napi_str; + struct efx_nic *efx = channel->efx; + struct sk_buff *skb; + + skb = napi_get_frags(napi); + if (unlikely(!skb)) { + struct efx_rx_queue *rx_queue; + + rx_queue = efx_channel_get_rx_queue(channel); + efx_free_rx_buffers(rx_queue, rx_buf, n_frags); + return; + } + + if (efx->net_dev->features & NETIF_F_RXHASH) + skb_set_hash(skb, efx_rx_buf_hash(efx, eh), + PKT_HASH_TYPE_L3); + skb->ip_summed = ((rx_buf->flags & EFX_RX_PKT_CSUMMED) ? + CHECKSUM_UNNECESSARY : CHECKSUM_NONE); + skb->csum_level = !!(rx_buf->flags & EFX_RX_PKT_CSUM_LEVEL); + + for (;;) { + skb_fill_page_desc(skb, skb_shinfo(skb)->nr_frags, + rx_buf->page, rx_buf->page_offset, + rx_buf->len); + rx_buf->page = NULL; + skb->len += rx_buf->len; + if (skb_shinfo(skb)->nr_frags == n_frags) + break; + + rx_buf = efx_rx_buf_next(&channel->rx_queue, rx_buf); + } + + skb->data_len = skb->len; + skb->truesize += n_frags * efx->rx_buffer_truesize; + + skb_record_rx_queue(skb, channel->rx_queue.core_index); + + napi_gro_frags(napi); +} + +/* RSS contexts. We're using linked lists and crappy O(n) algorithms, because + * (a) this is an infrequent control-plane operation and (b) n is small (max 64) + */ +struct efx_rss_context *efx_alloc_rss_context_entry(struct efx_nic *efx) +{ + struct list_head *head = &efx->rss_context.list; + struct efx_rss_context *ctx, *new; + u32 id = 1; /* Don't use zero, that refers to the master RSS context */ + + WARN_ON(!mutex_is_locked(&efx->rss_lock)); + + /* Search for first gap in the numbering */ + list_for_each_entry(ctx, head, list) { + if (ctx->user_id != id) + break; + id++; + /* Check for wrap. If this happens, we have nearly 2^32 + * allocated RSS contexts, which seems unlikely. + */ + if (WARN_ON_ONCE(!id)) + return NULL; + } + + /* Create the new entry */ + new = kmalloc(sizeof(*new), GFP_KERNEL); + if (!new) + return NULL; + new->context_id = EFX_MCDI_RSS_CONTEXT_INVALID; + new->rx_hash_udp_4tuple = false; + + /* Insert the new entry into the gap */ + new->user_id = id; + list_add_tail(&new->list, &ctx->list); + return new; +} + +struct efx_rss_context *efx_find_rss_context_entry(struct efx_nic *efx, u32 id) +{ + struct list_head *head = &efx->rss_context.list; + struct efx_rss_context *ctx; + + WARN_ON(!mutex_is_locked(&efx->rss_lock)); + + list_for_each_entry(ctx, head, list) + if (ctx->user_id == id) + return ctx; + return NULL; +} + +void efx_free_rss_context_entry(struct efx_rss_context *ctx) +{ + list_del(&ctx->list); + kfree(ctx); +} + +void efx_set_default_rx_indir_table(struct efx_nic *efx, + struct efx_rss_context *ctx) +{ + size_t i; + + for (i = 0; i < ARRAY_SIZE(ctx->rx_indir_table); i++) + ctx->rx_indir_table[i] = + ethtool_rxfh_indir_default(i, efx->rss_spread); +} + +/** + * efx_filter_is_mc_recipient - test whether spec is a multicast recipient + * @spec: Specification to test + * + * Return: %true if the specification is a non-drop RX filter that + * matches a local MAC address I/G bit value of 1 or matches a local + * IPv4 or IPv6 address value in the respective multicast address + * range. Otherwise %false. + */ +bool efx_filter_is_mc_recipient(const struct efx_filter_spec *spec) +{ + if (!(spec->flags & EFX_FILTER_FLAG_RX) || + spec->dmaq_id == EFX_FILTER_RX_DMAQ_ID_DROP) + return false; + + if (spec->match_flags & + (EFX_FILTER_MATCH_LOC_MAC | EFX_FILTER_MATCH_LOC_MAC_IG) && + is_multicast_ether_addr(spec->loc_mac)) + return true; + + if ((spec->match_flags & + (EFX_FILTER_MATCH_ETHER_TYPE | EFX_FILTER_MATCH_LOC_HOST)) == + (EFX_FILTER_MATCH_ETHER_TYPE | EFX_FILTER_MATCH_LOC_HOST)) { + if (spec->ether_type == htons(ETH_P_IP) && + ipv4_is_multicast(spec->loc_host[0])) + return true; + if (spec->ether_type == htons(ETH_P_IPV6) && + ((const u8 *)spec->loc_host)[0] == 0xff) + return true; + } + + return false; +} + +bool efx_filter_spec_equal(const struct efx_filter_spec *left, + const struct efx_filter_spec *right) +{ + if ((left->match_flags ^ right->match_flags) | + ((left->flags ^ right->flags) & + (EFX_FILTER_FLAG_RX | EFX_FILTER_FLAG_TX))) + return false; + + return memcmp(&left->outer_vid, &right->outer_vid, + sizeof(struct efx_filter_spec) - + offsetof(struct efx_filter_spec, outer_vid)) == 0; +} + +u32 efx_filter_spec_hash(const struct efx_filter_spec *spec) +{ + BUILD_BUG_ON(offsetof(struct efx_filter_spec, outer_vid) & 3); + return jhash2((const u32 *)&spec->outer_vid, + (sizeof(struct efx_filter_spec) - + offsetof(struct efx_filter_spec, outer_vid)) / 4, + 0); +} + +#ifdef CONFIG_RFS_ACCEL +bool efx_rps_check_rule(struct efx_arfs_rule *rule, unsigned int filter_idx, + bool *force) +{ + if (rule->filter_id == EFX_ARFS_FILTER_ID_PENDING) { + /* ARFS is currently updating this entry, leave it */ + return false; + } + if (rule->filter_id == EFX_ARFS_FILTER_ID_ERROR) { + /* ARFS tried and failed to update this, so it's probably out + * of date. Remove the filter and the ARFS rule entry. + */ + rule->filter_id = EFX_ARFS_FILTER_ID_REMOVING; + *force = true; + return true; + } else if (WARN_ON(rule->filter_id != filter_idx)) { /* can't happen */ + /* ARFS has moved on, so old filter is not needed. Since we did + * not mark the rule with EFX_ARFS_FILTER_ID_REMOVING, it will + * not be removed by efx_rps_hash_del() subsequently. + */ + *force = true; + return true; + } + /* Remove it iff ARFS wants to. */ + return true; +} + +static +struct hlist_head *efx_rps_hash_bucket(struct efx_nic *efx, + const struct efx_filter_spec *spec) +{ + u32 hash = efx_filter_spec_hash(spec); + + lockdep_assert_held(&efx->rps_hash_lock); + if (!efx->rps_hash_table) + return NULL; + return &efx->rps_hash_table[hash % EFX_ARFS_HASH_TABLE_SIZE]; +} + +struct efx_arfs_rule *efx_rps_hash_find(struct efx_nic *efx, + const struct efx_filter_spec *spec) +{ + struct efx_arfs_rule *rule; + struct hlist_head *head; + struct hlist_node *node; + + head = efx_rps_hash_bucket(efx, spec); + if (!head) + return NULL; + hlist_for_each(node, head) { + rule = container_of(node, struct efx_arfs_rule, node); + if (efx_filter_spec_equal(spec, &rule->spec)) + return rule; + } + return NULL; +} + +struct efx_arfs_rule *efx_rps_hash_add(struct efx_nic *efx, + const struct efx_filter_spec *spec, + bool *new) +{ + struct efx_arfs_rule *rule; + struct hlist_head *head; + struct hlist_node *node; + + head = efx_rps_hash_bucket(efx, spec); + if (!head) + return NULL; + hlist_for_each(node, head) { + rule = container_of(node, struct efx_arfs_rule, node); + if (efx_filter_spec_equal(spec, &rule->spec)) { + *new = false; + return rule; + } + } + rule = kmalloc(sizeof(*rule), GFP_ATOMIC); + *new = true; + if (rule) { + memcpy(&rule->spec, spec, sizeof(rule->spec)); + hlist_add_head(&rule->node, head); + } + return rule; +} + +void efx_rps_hash_del(struct efx_nic *efx, const struct efx_filter_spec *spec) +{ + struct efx_arfs_rule *rule; + struct hlist_head *head; + struct hlist_node *node; + + head = efx_rps_hash_bucket(efx, spec); + if (WARN_ON(!head)) + return; + hlist_for_each(node, head) { + rule = container_of(node, struct efx_arfs_rule, node); + if (efx_filter_spec_equal(spec, &rule->spec)) { + /* Someone already reused the entry. We know that if + * this check doesn't fire (i.e. filter_id == REMOVING) + * then the REMOVING mark was put there by our caller, + * because caller is holding a lock on filter table and + * only holders of that lock set REMOVING. + */ + if (rule->filter_id != EFX_ARFS_FILTER_ID_REMOVING) + return; + hlist_del(node); + kfree(rule); + return; + } + } + /* We didn't find it. */ + WARN_ON(1); +} +#endif + +int efx_probe_filters(struct efx_nic *efx) +{ + int rc; + + init_rwsem(&efx->filter_sem); + mutex_lock(&efx->mac_lock); + down_write(&efx->filter_sem); + rc = efx->type->filter_table_probe(efx); + if (rc) + goto out_unlock; + +#ifdef CONFIG_RFS_ACCEL + if (efx->type->offload_features & NETIF_F_NTUPLE) { + struct efx_channel *channel; + int i, success = 1; + + efx_for_each_channel(channel, efx) { + channel->rps_flow_id = + kcalloc(efx->type->max_rx_ip_filters, + sizeof(*channel->rps_flow_id), + GFP_KERNEL); + if (!channel->rps_flow_id) + success = 0; + else + for (i = 0; + i < efx->type->max_rx_ip_filters; + ++i) + channel->rps_flow_id[i] = + RPS_FLOW_ID_INVALID; + channel->rfs_expire_index = 0; + channel->rfs_filter_count = 0; + } + + if (!success) { + efx_for_each_channel(channel, efx) + kfree(channel->rps_flow_id); + efx->type->filter_table_remove(efx); + rc = -ENOMEM; + goto out_unlock; + } + } +#endif +out_unlock: + up_write(&efx->filter_sem); + mutex_unlock(&efx->mac_lock); + return rc; +} + +void efx_remove_filters(struct efx_nic *efx) +{ +#ifdef CONFIG_RFS_ACCEL + struct efx_channel *channel; + + efx_for_each_channel(channel, efx) { + cancel_delayed_work_sync(&channel->filter_work); + kfree(channel->rps_flow_id); + } +#endif + down_write(&efx->filter_sem); + efx->type->filter_table_remove(efx); + up_write(&efx->filter_sem); +} diff --git a/drivers/net/ethernet/sfc/rx_common.h b/drivers/net/ethernet/sfc/rx_common.h new file mode 100644 index 000000000000..c41f12a89477 --- /dev/null +++ b/drivers/net/ethernet/sfc/rx_common.h @@ -0,0 +1,97 @@ +/* SPDX-License-Identifier: GPL-2.0-only */ +/**************************************************************************** + * Driver for Solarflare network controllers and boards + * Copyright 2018 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 + * by the Free Software Foundation, incorporated herein by reference. + */ + +#ifndef EFX_RX_COMMON_H +#define EFX_RX_COMMON_H + +/* Preferred number of descriptors to fill at once */ +#define EFX_RX_PREFERRED_BATCH 8U + +/* Each packet can consume up to ceil(max_frame_len / buffer_size) buffers */ +#define EFX_RX_MAX_FRAGS DIV_ROUND_UP(EFX_MAX_FRAME_LEN(EFX_MAX_MTU), \ + EFX_RX_USR_BUF_SIZE) + +static inline u8 *efx_rx_buf_va(struct efx_rx_buffer *buf) +{ + return page_address(buf->page) + buf->page_offset; +} + +static inline u32 efx_rx_buf_hash(struct efx_nic *efx, const u8 *eh) +{ +#if defined(CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS) + return __le32_to_cpup((const __le32 *)(eh + efx->rx_packet_hash_offset)); +#else + const u8 *data = eh + efx->rx_packet_hash_offset; + + return (u32)data[0] | + (u32)data[1] << 8 | + (u32)data[2] << 16 | + (u32)data[3] << 24; +#endif +} + +void efx_rx_slow_fill(struct timer_list *t); + +void efx_recycle_rx_pages(struct efx_channel *channel, + struct efx_rx_buffer *rx_buf, + unsigned int n_frags); +void efx_discard_rx_packet(struct efx_channel *channel, + struct efx_rx_buffer *rx_buf, + unsigned int n_frags); + +int efx_probe_rx_queue(struct efx_rx_queue *rx_queue); +void efx_init_rx_queue(struct efx_rx_queue *rx_queue); +void efx_fini_rx_queue(struct efx_rx_queue *rx_queue); +void efx_remove_rx_queue(struct efx_rx_queue *rx_queue); +void efx_destroy_rx_queue(struct efx_rx_queue *rx_queue); + +void efx_init_rx_buffer(struct efx_rx_queue *rx_queue, + struct page *page, + unsigned int page_offset, + u16 flags); +void efx_unmap_rx_buffer(struct efx_nic *efx, struct efx_rx_buffer *rx_buf); +void efx_free_rx_buffers(struct efx_rx_queue *rx_queue, + struct efx_rx_buffer *rx_buf, + unsigned int num_bufs); + +void efx_schedule_slow_fill(struct efx_rx_queue *rx_queue); +void efx_rx_config_page_split(struct efx_nic *efx); +void efx_fast_push_rx_descriptors(struct efx_rx_queue *rx_queue, bool atomic); + +void +efx_rx_packet_gro(struct efx_channel *channel, struct efx_rx_buffer *rx_buf, + unsigned int n_frags, u8 *eh); + +struct efx_rss_context *efx_alloc_rss_context_entry(struct efx_nic *efx); +struct efx_rss_context *efx_find_rss_context_entry(struct efx_nic *efx, u32 id); +void efx_free_rss_context_entry(struct efx_rss_context *ctx); +void efx_set_default_rx_indir_table(struct efx_nic *efx, + struct efx_rss_context *ctx); + +bool efx_filter_is_mc_recipient(const struct efx_filter_spec *spec); +bool efx_filter_spec_equal(const struct efx_filter_spec *left, + const struct efx_filter_spec *right); +u32 efx_filter_spec_hash(const struct efx_filter_spec *spec); + +#ifdef CONFIG_RFS_ACCEL +bool efx_rps_check_rule(struct efx_arfs_rule *rule, unsigned int filter_idx, + bool *force); +struct efx_arfs_rule *efx_rps_hash_find(struct efx_nic *efx, + const struct efx_filter_spec *spec); +struct efx_arfs_rule *efx_rps_hash_add(struct efx_nic *efx, + const struct efx_filter_spec *spec, + bool *new); +void efx_rps_hash_del(struct efx_nic *efx, const struct efx_filter_spec *spec); +#endif + +int efx_probe_filters(struct efx_nic *efx); +void efx_remove_filters(struct efx_nic *efx); + +#endif diff --git a/drivers/net/ethernet/sfc/selftest.c b/drivers/net/ethernet/sfc/selftest.c index 8474cf8ea7d3..1ae369022d7d 100644 --- a/drivers/net/ethernet/sfc/selftest.c +++ b/drivers/net/ethernet/sfc/selftest.c @@ -18,6 +18,8 @@ #include <linux/slab.h> #include "net_driver.h" #include "efx.h" +#include "efx_common.h" +#include "efx_channels.h" #include "nic.h" #include "selftest.h" #include "workarounds.h" @@ -783,7 +785,7 @@ void efx_selftest_async_cancel(struct efx_nic *efx) cancel_delayed_work_sync(&efx->selftest_work); } -void efx_selftest_async_work(struct work_struct *data) +static void efx_selftest_async_work(struct work_struct *data) { struct efx_nic *efx = container_of(data, struct efx_nic, selftest_work.work); @@ -802,3 +804,8 @@ void efx_selftest_async_work(struct work_struct *data) channel->channel, cpu); } } + +void efx_selftest_async_init(struct efx_nic *efx) +{ + INIT_DELAYED_WORK(&efx->selftest_work, efx_selftest_async_work); +} diff --git a/drivers/net/ethernet/sfc/selftest.h b/drivers/net/ethernet/sfc/selftest.h index a3553816d92c..ca88ebb4f6b1 100644 --- a/drivers/net/ethernet/sfc/selftest.h +++ b/drivers/net/ethernet/sfc/selftest.h @@ -45,8 +45,8 @@ void efx_loopback_rx_packet(struct efx_nic *efx, const char *buf_ptr, int pkt_len); int efx_selftest(struct efx_nic *efx, struct efx_self_tests *tests, unsigned flags); +void efx_selftest_async_init(struct efx_nic *efx); void efx_selftest_async_start(struct efx_nic *efx); void efx_selftest_async_cancel(struct efx_nic *efx); -void efx_selftest_async_work(struct work_struct *data); #endif /* EFX_SELFTEST_H */ diff --git a/drivers/net/ethernet/sfc/siena.c b/drivers/net/ethernet/sfc/siena.c index 81499244a4b4..baa464161626 100644 --- a/drivers/net/ethernet/sfc/siena.c +++ b/drivers/net/ethernet/sfc/siena.c @@ -14,12 +14,14 @@ #include "net_driver.h" #include "bitfield.h" #include "efx.h" +#include "efx_common.h" #include "nic.h" #include "farch_regs.h" #include "io.h" #include "workarounds.h" #include "mcdi.h" #include "mcdi_pcol.h" +#include "mcdi_port_common.h" #include "selftest.h" #include "siena_sriov.h" diff --git a/drivers/net/ethernet/sfc/siena_sriov.c b/drivers/net/ethernet/sfc/siena_sriov.c index dfbdf05dcf79..83dcfcae3d4b 100644 --- a/drivers/net/ethernet/sfc/siena_sriov.c +++ b/drivers/net/ethernet/sfc/siena_sriov.c @@ -7,6 +7,7 @@ #include <linux/module.h> #include "net_driver.h" #include "efx.h" +#include "efx_channels.h" #include "nic.h" #include "io.h" #include "mcdi.h" diff --git a/drivers/net/ethernet/sfc/tx.c b/drivers/net/ethernet/sfc/tx.c index 00c1c4402451..04d7f41d7ed9 100644 --- a/drivers/net/ethernet/sfc/tx.c +++ b/drivers/net/ethernet/sfc/tx.c @@ -20,6 +20,7 @@ #include "io.h" #include "nic.h" #include "tx.h" +#include "tx_common.h" #include "workarounds.h" #include "ef10_regs.h" @@ -56,72 +57,6 @@ u8 *efx_tx_get_copy_buffer_limited(struct efx_tx_queue *tx_queue, return efx_tx_get_copy_buffer(tx_queue, buffer); } -static void efx_dequeue_buffer(struct efx_tx_queue *tx_queue, - struct efx_tx_buffer *buffer, - unsigned int *pkts_compl, - unsigned int *bytes_compl) -{ - if (buffer->unmap_len) { - struct device *dma_dev = &tx_queue->efx->pci_dev->dev; - dma_addr_t unmap_addr = buffer->dma_addr - buffer->dma_offset; - if (buffer->flags & EFX_TX_BUF_MAP_SINGLE) - dma_unmap_single(dma_dev, unmap_addr, buffer->unmap_len, - DMA_TO_DEVICE); - else - dma_unmap_page(dma_dev, unmap_addr, buffer->unmap_len, - DMA_TO_DEVICE); - buffer->unmap_len = 0; - } - - if (buffer->flags & EFX_TX_BUF_SKB) { - struct sk_buff *skb = (struct sk_buff *)buffer->skb; - - EFX_WARN_ON_PARANOID(!pkts_compl || !bytes_compl); - (*pkts_compl)++; - (*bytes_compl) += skb->len; - if (tx_queue->timestamping && - (tx_queue->completed_timestamp_major || - tx_queue->completed_timestamp_minor)) { - struct skb_shared_hwtstamps hwtstamp; - - hwtstamp.hwtstamp = - efx_ptp_nic_to_kernel_time(tx_queue); - skb_tstamp_tx(skb, &hwtstamp); - - tx_queue->completed_timestamp_major = 0; - tx_queue->completed_timestamp_minor = 0; - } - dev_consume_skb_any((struct sk_buff *)buffer->skb); - netif_vdbg(tx_queue->efx, tx_done, tx_queue->efx->net_dev, - "TX queue %d transmission id %x complete\n", - tx_queue->queue, tx_queue->read_count); - } else if (buffer->flags & EFX_TX_BUF_XDP) { - xdp_return_frame_rx_napi(buffer->xdpf); - } - - buffer->len = 0; - buffer->flags = 0; -} - -unsigned int efx_tx_max_skb_descs(struct efx_nic *efx) -{ - /* Header and payload descriptor for each output segment, plus - * one for every input fragment boundary within a segment - */ - unsigned int max_descs = EFX_TSO_MAX_SEGS * 2 + MAX_SKB_FRAGS; - - /* Possibly one more per segment for option descriptors */ - if (efx_nic_rev(efx) >= EFX_REV_HUNT_A0) - max_descs += EFX_TSO_MAX_SEGS; - - /* Possibly more for PCIe page boundaries within input fragments */ - if (PAGE_SIZE > EFX_PAGE_SIZE) - max_descs += max_t(unsigned int, MAX_SKB_FRAGS, - DIV_ROUND_UP(GSO_MAX_SIZE, EFX_PAGE_SIZE)); - - return max_descs; -} - static void efx_tx_maybe_stop_queue(struct efx_tx_queue *txq1) { /* We need to consider both queues that the net core sees as one */ @@ -333,125 +268,6 @@ static int efx_enqueue_skb_pio(struct efx_tx_queue *tx_queue, } #endif /* EFX_USE_PIO */ -static struct efx_tx_buffer *efx_tx_map_chunk(struct efx_tx_queue *tx_queue, - dma_addr_t dma_addr, - size_t len) -{ - const struct efx_nic_type *nic_type = tx_queue->efx->type; - struct efx_tx_buffer *buffer; - unsigned int dma_len; - - /* Map the fragment taking account of NIC-dependent DMA limits. */ - do { - buffer = efx_tx_queue_get_insert_buffer(tx_queue); - dma_len = nic_type->tx_limit_len(tx_queue, dma_addr, len); - - buffer->len = dma_len; - buffer->dma_addr = dma_addr; - buffer->flags = EFX_TX_BUF_CONT; - len -= dma_len; - dma_addr += dma_len; - ++tx_queue->insert_count; - } while (len); - - return buffer; -} - -/* Map all data from an SKB for DMA and create descriptors on the queue. - */ -static int efx_tx_map_data(struct efx_tx_queue *tx_queue, struct sk_buff *skb, - unsigned int segment_count) -{ - struct efx_nic *efx = tx_queue->efx; - struct device *dma_dev = &efx->pci_dev->dev; - unsigned int frag_index, nr_frags; - dma_addr_t dma_addr, unmap_addr; - unsigned short dma_flags; - size_t len, unmap_len; - - nr_frags = skb_shinfo(skb)->nr_frags; - frag_index = 0; - - /* Map header data. */ - len = skb_headlen(skb); - dma_addr = dma_map_single(dma_dev, skb->data, len, DMA_TO_DEVICE); - dma_flags = EFX_TX_BUF_MAP_SINGLE; - unmap_len = len; - unmap_addr = dma_addr; - - if (unlikely(dma_mapping_error(dma_dev, dma_addr))) - return -EIO; - - if (segment_count) { - /* For TSO we need to put the header in to a separate - * descriptor. Map this separately if necessary. - */ - size_t header_len = skb_transport_header(skb) - skb->data + - (tcp_hdr(skb)->doff << 2u); - - if (header_len != len) { - tx_queue->tso_long_headers++; - efx_tx_map_chunk(tx_queue, dma_addr, header_len); - len -= header_len; - dma_addr += header_len; - } - } - - /* Add descriptors for each fragment. */ - do { - struct efx_tx_buffer *buffer; - skb_frag_t *fragment; - - buffer = efx_tx_map_chunk(tx_queue, dma_addr, len); - - /* The final descriptor for a fragment is responsible for - * unmapping the whole fragment. - */ - buffer->flags = EFX_TX_BUF_CONT | dma_flags; - buffer->unmap_len = unmap_len; - buffer->dma_offset = buffer->dma_addr - unmap_addr; - - if (frag_index >= nr_frags) { - /* Store SKB details with the final buffer for - * the completion. - */ - buffer->skb = skb; - buffer->flags = EFX_TX_BUF_SKB | dma_flags; - return 0; - } - - /* Move on to the next fragment. */ - fragment = &skb_shinfo(skb)->frags[frag_index++]; - len = skb_frag_size(fragment); - dma_addr = skb_frag_dma_map(dma_dev, fragment, - 0, len, DMA_TO_DEVICE); - dma_flags = 0; - unmap_len = len; - unmap_addr = dma_addr; - - if (unlikely(dma_mapping_error(dma_dev, dma_addr))) - return -EIO; - } while (1); -} - -/* Remove buffers put into a tx_queue for the current packet. - * None of the buffers must have an skb attached. - */ -static void efx_enqueue_unwind(struct efx_tx_queue *tx_queue, - unsigned int insert_count) -{ - struct efx_tx_buffer *buffer; - unsigned int bytes_compl = 0; - unsigned int pkts_compl = 0; - - /* Work backwards until we hit the original insert pointer value */ - while (tx_queue->insert_count != insert_count) { - --tx_queue->insert_count; - buffer = __efx_tx_queue_get_insert_buffer(tx_queue); - efx_dequeue_buffer(tx_queue, buffer, &pkts_compl, &bytes_compl); - } -} - /* * Fallback to software TSO. * @@ -473,12 +289,9 @@ static int efx_tx_tso_fallback(struct efx_tx_queue *tx_queue, dev_consume_skb_any(skb); skb = segments; - while (skb) { - next = skb->next; - skb->next = NULL; - + skb_list_walk_safe(skb, skb, next) { + skb_mark_not_on_list(skb); efx_enqueue_skb(tx_queue, skb); - skb = next; } return 0; @@ -687,41 +500,6 @@ int efx_xdp_tx_buffers(struct efx_nic *efx, int n, struct xdp_frame **xdpfs, return i; } -/* Remove packets from the TX queue - * - * This removes packets from the TX queue, up to and including the - * specified index. - */ -static void efx_dequeue_buffers(struct efx_tx_queue *tx_queue, - unsigned int index, - unsigned int *pkts_compl, - unsigned int *bytes_compl) -{ - struct efx_nic *efx = tx_queue->efx; - unsigned int stop_index, read_ptr; - - stop_index = (index + 1) & tx_queue->ptr_mask; - read_ptr = tx_queue->read_count & tx_queue->ptr_mask; - - while (read_ptr != stop_index) { - struct efx_tx_buffer *buffer = &tx_queue->buffer[read_ptr]; - - if (!(buffer->flags & EFX_TX_BUF_OPTION) && - unlikely(buffer->len == 0)) { - netif_err(efx, tx_err, efx->net_dev, - "TX queue %d spurious TX completion id %x\n", - tx_queue->queue, read_ptr); - efx_schedule_reset(efx, RESET_TYPE_TX_SKIP); - return; - } - - efx_dequeue_buffer(tx_queue, buffer, pkts_compl, bytes_compl); - - ++tx_queue->read_count; - read_ptr = tx_queue->read_count & tx_queue->ptr_mask; - } -} - /* Initiate a packet transmission. We use one channel per CPU * (sharing when we have more CPUs than channels). On Falcon, the TX * completion events will be directed back to the CPU that transmitted @@ -834,173 +612,3 @@ int efx_setup_tc(struct net_device *net_dev, enum tc_setup_type type, net_dev->num_tc = num_tc; return 0; } - -void efx_xmit_done(struct efx_tx_queue *tx_queue, unsigned int index) -{ - unsigned fill_level; - struct efx_nic *efx = tx_queue->efx; - struct efx_tx_queue *txq2; - unsigned int pkts_compl = 0, bytes_compl = 0; - - EFX_WARN_ON_ONCE_PARANOID(index > tx_queue->ptr_mask); - - efx_dequeue_buffers(tx_queue, index, &pkts_compl, &bytes_compl); - tx_queue->pkts_compl += pkts_compl; - tx_queue->bytes_compl += bytes_compl; - - if (pkts_compl > 1) - ++tx_queue->merge_events; - - /* See if we need to restart the netif queue. This memory - * barrier ensures that we write read_count (inside - * efx_dequeue_buffers()) before reading the queue status. - */ - smp_mb(); - if (unlikely(netif_tx_queue_stopped(tx_queue->core_txq)) && - likely(efx->port_enabled) && - likely(netif_device_present(efx->net_dev))) { - txq2 = efx_tx_queue_partner(tx_queue); - fill_level = max(tx_queue->insert_count - tx_queue->read_count, - txq2->insert_count - txq2->read_count); - if (fill_level <= efx->txq_wake_thresh) - netif_tx_wake_queue(tx_queue->core_txq); - } - - /* Check whether the hardware queue is now empty */ - if ((int)(tx_queue->read_count - tx_queue->old_write_count) >= 0) { - tx_queue->old_write_count = READ_ONCE(tx_queue->write_count); - if (tx_queue->read_count == tx_queue->old_write_count) { - smp_mb(); - tx_queue->empty_read_count = - tx_queue->read_count | EFX_EMPTY_COUNT_VALID; - } - } -} - -static unsigned int efx_tx_cb_page_count(struct efx_tx_queue *tx_queue) -{ - return DIV_ROUND_UP(tx_queue->ptr_mask + 1, PAGE_SIZE >> EFX_TX_CB_ORDER); -} - -int efx_probe_tx_queue(struct efx_tx_queue *tx_queue) -{ - struct efx_nic *efx = tx_queue->efx; - unsigned int entries; - int rc; - - /* Create the smallest power-of-two aligned ring */ - entries = max(roundup_pow_of_two(efx->txq_entries), EFX_MIN_DMAQ_SIZE); - EFX_WARN_ON_PARANOID(entries > EFX_MAX_DMAQ_SIZE); - tx_queue->ptr_mask = entries - 1; - - netif_dbg(efx, probe, efx->net_dev, - "creating TX queue %d size %#x mask %#x\n", - tx_queue->queue, efx->txq_entries, tx_queue->ptr_mask); - - /* Allocate software ring */ - tx_queue->buffer = kcalloc(entries, sizeof(*tx_queue->buffer), - GFP_KERNEL); - if (!tx_queue->buffer) - return -ENOMEM; - - tx_queue->cb_page = kcalloc(efx_tx_cb_page_count(tx_queue), - sizeof(tx_queue->cb_page[0]), GFP_KERNEL); - if (!tx_queue->cb_page) { - rc = -ENOMEM; - goto fail1; - } - - /* Allocate hardware ring */ - rc = efx_nic_probe_tx(tx_queue); - if (rc) - goto fail2; - - return 0; - -fail2: - kfree(tx_queue->cb_page); - tx_queue->cb_page = NULL; -fail1: - kfree(tx_queue->buffer); - tx_queue->buffer = NULL; - return rc; -} - -void efx_init_tx_queue(struct efx_tx_queue *tx_queue) -{ - struct efx_nic *efx = tx_queue->efx; - - netif_dbg(efx, drv, efx->net_dev, - "initialising TX queue %d\n", tx_queue->queue); - - tx_queue->insert_count = 0; - tx_queue->write_count = 0; - tx_queue->packet_write_count = 0; - tx_queue->old_write_count = 0; - tx_queue->read_count = 0; - tx_queue->old_read_count = 0; - tx_queue->empty_read_count = 0 | EFX_EMPTY_COUNT_VALID; - tx_queue->xmit_more_available = false; - tx_queue->timestamping = (efx_ptp_use_mac_tx_timestamps(efx) && - tx_queue->channel == efx_ptp_channel(efx)); - tx_queue->completed_desc_ptr = tx_queue->ptr_mask; - tx_queue->completed_timestamp_major = 0; - tx_queue->completed_timestamp_minor = 0; - - tx_queue->xdp_tx = efx_channel_is_xdp_tx(tx_queue->channel); - - /* Set up default function pointers. These may get replaced by - * efx_nic_init_tx() based off NIC/queue capabilities. - */ - tx_queue->handle_tso = efx_enqueue_skb_tso; - - /* Set up TX descriptor ring */ - efx_nic_init_tx(tx_queue); - - tx_queue->initialised = true; -} - -void efx_fini_tx_queue(struct efx_tx_queue *tx_queue) -{ - struct efx_tx_buffer *buffer; - - netif_dbg(tx_queue->efx, drv, tx_queue->efx->net_dev, - "shutting down TX queue %d\n", tx_queue->queue); - - if (!tx_queue->buffer) - return; - - /* Free any buffers left in the ring */ - while (tx_queue->read_count != tx_queue->write_count) { - unsigned int pkts_compl = 0, bytes_compl = 0; - buffer = &tx_queue->buffer[tx_queue->read_count & tx_queue->ptr_mask]; - efx_dequeue_buffer(tx_queue, buffer, &pkts_compl, &bytes_compl); - - ++tx_queue->read_count; - } - tx_queue->xmit_more_available = false; - netdev_tx_reset_queue(tx_queue->core_txq); -} - -void efx_remove_tx_queue(struct efx_tx_queue *tx_queue) -{ - int i; - - if (!tx_queue->buffer) - return; - - netif_dbg(tx_queue->efx, drv, tx_queue->efx->net_dev, - "destroying TX queue %d\n", tx_queue->queue); - efx_nic_remove_tx(tx_queue); - - if (tx_queue->cb_page) { - for (i = 0; i < efx_tx_cb_page_count(tx_queue); i++) - efx_nic_free_buffer(tx_queue->efx, - &tx_queue->cb_page[i]); - kfree(tx_queue->cb_page); - tx_queue->cb_page = NULL; - } - - kfree(tx_queue->buffer); - tx_queue->buffer = NULL; -} diff --git a/drivers/net/ethernet/sfc/tx_common.c b/drivers/net/ethernet/sfc/tx_common.c new file mode 100644 index 000000000000..b1571e9789d0 --- /dev/null +++ b/drivers/net/ethernet/sfc/tx_common.c @@ -0,0 +1,404 @@ +// SPDX-License-Identifier: GPL-2.0-only +/**************************************************************************** + * Driver for Solarflare network controllers and boards + * Copyright 2018 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 + * by the Free Software Foundation, incorporated herein by reference. + */ + +#include "net_driver.h" +#include "efx.h" +#include "nic.h" +#include "tx_common.h" + +static unsigned int efx_tx_cb_page_count(struct efx_tx_queue *tx_queue) +{ + return DIV_ROUND_UP(tx_queue->ptr_mask + 1, + PAGE_SIZE >> EFX_TX_CB_ORDER); +} + +int efx_probe_tx_queue(struct efx_tx_queue *tx_queue) +{ + struct efx_nic *efx = tx_queue->efx; + unsigned int entries; + int rc; + + /* Create the smallest power-of-two aligned ring */ + entries = max(roundup_pow_of_two(efx->txq_entries), EFX_MIN_DMAQ_SIZE); + EFX_WARN_ON_PARANOID(entries > EFX_MAX_DMAQ_SIZE); + tx_queue->ptr_mask = entries - 1; + + netif_dbg(efx, probe, efx->net_dev, + "creating TX queue %d size %#x mask %#x\n", + tx_queue->queue, efx->txq_entries, tx_queue->ptr_mask); + + /* Allocate software ring */ + tx_queue->buffer = kcalloc(entries, sizeof(*tx_queue->buffer), + GFP_KERNEL); + if (!tx_queue->buffer) + return -ENOMEM; + + tx_queue->cb_page = kcalloc(efx_tx_cb_page_count(tx_queue), + sizeof(tx_queue->cb_page[0]), GFP_KERNEL); + if (!tx_queue->cb_page) { + rc = -ENOMEM; + goto fail1; + } + + /* Allocate hardware ring */ + rc = efx_nic_probe_tx(tx_queue); + if (rc) + goto fail2; + + return 0; + +fail2: + kfree(tx_queue->cb_page); + tx_queue->cb_page = NULL; +fail1: + kfree(tx_queue->buffer); + tx_queue->buffer = NULL; + return rc; +} + +void efx_init_tx_queue(struct efx_tx_queue *tx_queue) +{ + struct efx_nic *efx = tx_queue->efx; + + netif_dbg(efx, drv, efx->net_dev, + "initialising TX queue %d\n", tx_queue->queue); + + tx_queue->insert_count = 0; + tx_queue->write_count = 0; + tx_queue->packet_write_count = 0; + tx_queue->old_write_count = 0; + tx_queue->read_count = 0; + tx_queue->old_read_count = 0; + tx_queue->empty_read_count = 0 | EFX_EMPTY_COUNT_VALID; + tx_queue->xmit_more_available = false; + tx_queue->timestamping = (efx_ptp_use_mac_tx_timestamps(efx) && + tx_queue->channel == efx_ptp_channel(efx)); + tx_queue->completed_desc_ptr = tx_queue->ptr_mask; + tx_queue->completed_timestamp_major = 0; + tx_queue->completed_timestamp_minor = 0; + + tx_queue->xdp_tx = efx_channel_is_xdp_tx(tx_queue->channel); + + /* Set up default function pointers. These may get replaced by + * efx_nic_init_tx() based off NIC/queue capabilities. + */ + tx_queue->handle_tso = efx_enqueue_skb_tso; + + /* Set up TX descriptor ring */ + efx_nic_init_tx(tx_queue); + + tx_queue->initialised = true; +} + +void efx_fini_tx_queue(struct efx_tx_queue *tx_queue) +{ + struct efx_tx_buffer *buffer; + + netif_dbg(tx_queue->efx, drv, tx_queue->efx->net_dev, + "shutting down TX queue %d\n", tx_queue->queue); + + if (!tx_queue->buffer) + return; + + /* Free any buffers left in the ring */ + while (tx_queue->read_count != tx_queue->write_count) { + unsigned int pkts_compl = 0, bytes_compl = 0; + + buffer = &tx_queue->buffer[tx_queue->read_count & tx_queue->ptr_mask]; + efx_dequeue_buffer(tx_queue, buffer, &pkts_compl, &bytes_compl); + + ++tx_queue->read_count; + } + tx_queue->xmit_more_available = false; + netdev_tx_reset_queue(tx_queue->core_txq); +} + +void efx_remove_tx_queue(struct efx_tx_queue *tx_queue) +{ + int i; + + if (!tx_queue->buffer) + return; + + netif_dbg(tx_queue->efx, drv, tx_queue->efx->net_dev, + "destroying TX queue %d\n", tx_queue->queue); + efx_nic_remove_tx(tx_queue); + + if (tx_queue->cb_page) { + for (i = 0; i < efx_tx_cb_page_count(tx_queue); i++) + efx_nic_free_buffer(tx_queue->efx, + &tx_queue->cb_page[i]); + kfree(tx_queue->cb_page); + tx_queue->cb_page = NULL; + } + + kfree(tx_queue->buffer); + tx_queue->buffer = NULL; +} + +void efx_dequeue_buffer(struct efx_tx_queue *tx_queue, + struct efx_tx_buffer *buffer, + unsigned int *pkts_compl, + unsigned int *bytes_compl) +{ + if (buffer->unmap_len) { + struct device *dma_dev = &tx_queue->efx->pci_dev->dev; + dma_addr_t unmap_addr = buffer->dma_addr - buffer->dma_offset; + + if (buffer->flags & EFX_TX_BUF_MAP_SINGLE) + dma_unmap_single(dma_dev, unmap_addr, buffer->unmap_len, + DMA_TO_DEVICE); + else + dma_unmap_page(dma_dev, unmap_addr, buffer->unmap_len, + DMA_TO_DEVICE); + buffer->unmap_len = 0; + } + + if (buffer->flags & EFX_TX_BUF_SKB) { + struct sk_buff *skb = (struct sk_buff *)buffer->skb; + + EFX_WARN_ON_PARANOID(!pkts_compl || !bytes_compl); + (*pkts_compl)++; + (*bytes_compl) += skb->len; + if (tx_queue->timestamping && + (tx_queue->completed_timestamp_major || + tx_queue->completed_timestamp_minor)) { + struct skb_shared_hwtstamps hwtstamp; + + hwtstamp.hwtstamp = + efx_ptp_nic_to_kernel_time(tx_queue); + skb_tstamp_tx(skb, &hwtstamp); + + tx_queue->completed_timestamp_major = 0; + tx_queue->completed_timestamp_minor = 0; + } + dev_consume_skb_any((struct sk_buff *)buffer->skb); + netif_vdbg(tx_queue->efx, tx_done, tx_queue->efx->net_dev, + "TX queue %d transmission id %x complete\n", + tx_queue->queue, tx_queue->read_count); + } else if (buffer->flags & EFX_TX_BUF_XDP) { + xdp_return_frame_rx_napi(buffer->xdpf); + } + + buffer->len = 0; + buffer->flags = 0; +} + +/* Remove packets from the TX queue + * + * This removes packets from the TX queue, up to and including the + * specified index. + */ +static void efx_dequeue_buffers(struct efx_tx_queue *tx_queue, + unsigned int index, + unsigned int *pkts_compl, + unsigned int *bytes_compl) +{ + struct efx_nic *efx = tx_queue->efx; + unsigned int stop_index, read_ptr; + + stop_index = (index + 1) & tx_queue->ptr_mask; + read_ptr = tx_queue->read_count & tx_queue->ptr_mask; + + while (read_ptr != stop_index) { + struct efx_tx_buffer *buffer = &tx_queue->buffer[read_ptr]; + + if (!(buffer->flags & EFX_TX_BUF_OPTION) && + unlikely(buffer->len == 0)) { + netif_err(efx, tx_err, efx->net_dev, + "TX queue %d spurious TX completion id %x\n", + tx_queue->queue, read_ptr); + efx_schedule_reset(efx, RESET_TYPE_TX_SKIP); + return; + } + + efx_dequeue_buffer(tx_queue, buffer, pkts_compl, bytes_compl); + + ++tx_queue->read_count; + read_ptr = tx_queue->read_count & tx_queue->ptr_mask; + } +} + +void efx_xmit_done(struct efx_tx_queue *tx_queue, unsigned int index) +{ + unsigned int fill_level, pkts_compl = 0, bytes_compl = 0; + struct efx_nic *efx = tx_queue->efx; + struct efx_tx_queue *txq2; + + EFX_WARN_ON_ONCE_PARANOID(index > tx_queue->ptr_mask); + + efx_dequeue_buffers(tx_queue, index, &pkts_compl, &bytes_compl); + tx_queue->pkts_compl += pkts_compl; + tx_queue->bytes_compl += bytes_compl; + + if (pkts_compl > 1) + ++tx_queue->merge_events; + + /* See if we need to restart the netif queue. This memory + * barrier ensures that we write read_count (inside + * efx_dequeue_buffers()) before reading the queue status. + */ + smp_mb(); + if (unlikely(netif_tx_queue_stopped(tx_queue->core_txq)) && + likely(efx->port_enabled) && + likely(netif_device_present(efx->net_dev))) { + txq2 = efx_tx_queue_partner(tx_queue); + fill_level = max(tx_queue->insert_count - tx_queue->read_count, + txq2->insert_count - txq2->read_count); + if (fill_level <= efx->txq_wake_thresh) + netif_tx_wake_queue(tx_queue->core_txq); + } + + /* Check whether the hardware queue is now empty */ + if ((int)(tx_queue->read_count - tx_queue->old_write_count) >= 0) { + tx_queue->old_write_count = READ_ONCE(tx_queue->write_count); + if (tx_queue->read_count == tx_queue->old_write_count) { + smp_mb(); + tx_queue->empty_read_count = + tx_queue->read_count | EFX_EMPTY_COUNT_VALID; + } + } +} + +/* Remove buffers put into a tx_queue for the current packet. + * None of the buffers must have an skb attached. + */ +void efx_enqueue_unwind(struct efx_tx_queue *tx_queue, + unsigned int insert_count) +{ + struct efx_tx_buffer *buffer; + unsigned int bytes_compl = 0; + unsigned int pkts_compl = 0; + + /* Work backwards until we hit the original insert pointer value */ + while (tx_queue->insert_count != insert_count) { + --tx_queue->insert_count; + buffer = __efx_tx_queue_get_insert_buffer(tx_queue); + efx_dequeue_buffer(tx_queue, buffer, &pkts_compl, &bytes_compl); + } +} + +struct efx_tx_buffer *efx_tx_map_chunk(struct efx_tx_queue *tx_queue, + dma_addr_t dma_addr, size_t len) +{ + const struct efx_nic_type *nic_type = tx_queue->efx->type; + struct efx_tx_buffer *buffer; + unsigned int dma_len; + + /* Map the fragment taking account of NIC-dependent DMA limits. */ + do { + buffer = efx_tx_queue_get_insert_buffer(tx_queue); + dma_len = nic_type->tx_limit_len(tx_queue, dma_addr, len); + + buffer->len = dma_len; + buffer->dma_addr = dma_addr; + buffer->flags = EFX_TX_BUF_CONT; + len -= dma_len; + dma_addr += dma_len; + ++tx_queue->insert_count; + } while (len); + + return buffer; +} + +/* Map all data from an SKB for DMA and create descriptors on the queue. */ +int efx_tx_map_data(struct efx_tx_queue *tx_queue, struct sk_buff *skb, + unsigned int segment_count) +{ + struct efx_nic *efx = tx_queue->efx; + struct device *dma_dev = &efx->pci_dev->dev; + unsigned int frag_index, nr_frags; + dma_addr_t dma_addr, unmap_addr; + unsigned short dma_flags; + size_t len, unmap_len; + + nr_frags = skb_shinfo(skb)->nr_frags; + frag_index = 0; + + /* Map header data. */ + len = skb_headlen(skb); + dma_addr = dma_map_single(dma_dev, skb->data, len, DMA_TO_DEVICE); + dma_flags = EFX_TX_BUF_MAP_SINGLE; + unmap_len = len; + unmap_addr = dma_addr; + + if (unlikely(dma_mapping_error(dma_dev, dma_addr))) + return -EIO; + + if (segment_count) { + /* For TSO we need to put the header in to a separate + * descriptor. Map this separately if necessary. + */ + size_t header_len = skb_transport_header(skb) - skb->data + + (tcp_hdr(skb)->doff << 2u); + + if (header_len != len) { + tx_queue->tso_long_headers++; + efx_tx_map_chunk(tx_queue, dma_addr, header_len); + len -= header_len; + dma_addr += header_len; + } + } + + /* Add descriptors for each fragment. */ + do { + struct efx_tx_buffer *buffer; + skb_frag_t *fragment; + + buffer = efx_tx_map_chunk(tx_queue, dma_addr, len); + + /* The final descriptor for a fragment is responsible for + * unmapping the whole fragment. + */ + buffer->flags = EFX_TX_BUF_CONT | dma_flags; + buffer->unmap_len = unmap_len; + buffer->dma_offset = buffer->dma_addr - unmap_addr; + + if (frag_index >= nr_frags) { + /* Store SKB details with the final buffer for + * the completion. + */ + buffer->skb = skb; + buffer->flags = EFX_TX_BUF_SKB | dma_flags; + return 0; + } + + /* Move on to the next fragment. */ + fragment = &skb_shinfo(skb)->frags[frag_index++]; + len = skb_frag_size(fragment); + dma_addr = skb_frag_dma_map(dma_dev, fragment, 0, len, + DMA_TO_DEVICE); + dma_flags = 0; + unmap_len = len; + unmap_addr = dma_addr; + + if (unlikely(dma_mapping_error(dma_dev, dma_addr))) + return -EIO; + } while (1); +} + +unsigned int efx_tx_max_skb_descs(struct efx_nic *efx) +{ + /* Header and payload descriptor for each output segment, plus + * one for every input fragment boundary within a segment + */ + unsigned int max_descs = EFX_TSO_MAX_SEGS * 2 + MAX_SKB_FRAGS; + + /* Possibly one more per segment for option descriptors */ + if (efx_nic_rev(efx) >= EFX_REV_HUNT_A0) + max_descs += EFX_TSO_MAX_SEGS; + + /* Possibly more for PCIe page boundaries within input fragments */ + if (PAGE_SIZE > EFX_PAGE_SIZE) + max_descs += max_t(unsigned int, MAX_SKB_FRAGS, + DIV_ROUND_UP(GSO_MAX_SIZE, EFX_PAGE_SIZE)); + + return max_descs; +} diff --git a/drivers/net/ethernet/sfc/tx_common.h b/drivers/net/ethernet/sfc/tx_common.h new file mode 100644 index 000000000000..f92f1fe3a87f --- /dev/null +++ b/drivers/net/ethernet/sfc/tx_common.h @@ -0,0 +1,36 @@ +/* SPDX-License-Identifier: GPL-2.0-only */ +/**************************************************************************** + * Driver for Solarflare network controllers and boards + * Copyright 2018 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 + * by the Free Software Foundation, incorporated herein by reference. + */ + +#ifndef EFX_TX_COMMON_H +#define EFX_TX_COMMON_H + +int efx_probe_tx_queue(struct efx_tx_queue *tx_queue); +void efx_init_tx_queue(struct efx_tx_queue *tx_queue); +void efx_fini_tx_queue(struct efx_tx_queue *tx_queue); +void efx_remove_tx_queue(struct efx_tx_queue *tx_queue); + +void efx_dequeue_buffer(struct efx_tx_queue *tx_queue, + struct efx_tx_buffer *buffer, + unsigned int *pkts_compl, + unsigned int *bytes_compl); + +void efx_xmit_done(struct efx_tx_queue *tx_queue, unsigned int index); + +void efx_enqueue_unwind(struct efx_tx_queue *tx_queue, + unsigned int insert_count); + +struct efx_tx_buffer *efx_tx_map_chunk(struct efx_tx_queue *tx_queue, + dma_addr_t dma_addr, size_t len); +int efx_tx_map_data(struct efx_tx_queue *tx_queue, struct sk_buff *skb, + unsigned int segment_count); + +unsigned int efx_tx_max_skb_descs(struct efx_nic *efx); + +#endif diff --git a/drivers/net/ethernet/socionext/netsec.c b/drivers/net/ethernet/socionext/netsec.c index b5a9e947a4a8..495f6cfdbd91 100644 --- a/drivers/net/ethernet/socionext/netsec.c +++ b/drivers/net/ethernet/socionext/netsec.c @@ -243,6 +243,7 @@ NET_IP_ALIGN) #define NETSEC_RX_BUF_NON_DATA (NETSEC_RXBUF_HEADROOM + \ SKB_DATA_ALIGN(sizeof(struct skb_shared_info))) +#define NETSEC_RX_BUF_SIZE (PAGE_SIZE - NETSEC_RX_BUF_NON_DATA) #define DESC_SZ sizeof(struct netsec_de) @@ -719,7 +720,6 @@ static void *netsec_alloc_rx_data(struct netsec_priv *priv, { struct netsec_desc_ring *dring = &priv->desc_ring[NETSEC_RING_RX]; - enum dma_data_direction dma_dir; struct page *page; page = page_pool_dev_alloc_pages(dring->page_pool); @@ -734,9 +734,7 @@ static void *netsec_alloc_rx_data(struct netsec_priv *priv, /* Make sure the incoming payload fits in the page for XDP and non-XDP * cases and reserve enough space for headroom + skb_shared_info */ - *desc_len = PAGE_SIZE - NETSEC_RX_BUF_NON_DATA; - dma_dir = page_pool_get_dma_dir(dring->page_pool); - dma_sync_single_for_device(priv->dev, *dma_handle, *desc_len, dma_dir); + *desc_len = NETSEC_RX_BUF_SIZE; return page_address(page); } @@ -883,6 +881,8 @@ static u32 netsec_xdp_xmit_back(struct netsec_priv *priv, struct xdp_buff *xdp) static u32 netsec_run_xdp(struct netsec_priv *priv, struct bpf_prog *prog, struct xdp_buff *xdp) { + struct netsec_desc_ring *dring = &priv->desc_ring[NETSEC_RING_RX]; + unsigned int len = xdp->data_end - xdp->data; u32 ret = NETSEC_XDP_PASS; int err; u32 act; @@ -896,7 +896,9 @@ static u32 netsec_run_xdp(struct netsec_priv *priv, struct bpf_prog *prog, case XDP_TX: ret = netsec_xdp_xmit_back(priv, xdp); if (ret != NETSEC_XDP_TX) - xdp_return_buff(xdp); + __page_pool_put_page(dring->page_pool, + virt_to_head_page(xdp->data), + len, true); break; case XDP_REDIRECT: err = xdp_do_redirect(priv->ndev, xdp, prog); @@ -904,7 +906,9 @@ static u32 netsec_run_xdp(struct netsec_priv *priv, struct bpf_prog *prog, ret = NETSEC_XDP_REDIR; } else { ret = NETSEC_XDP_CONSUMED; - xdp_return_buff(xdp); + __page_pool_put_page(dring->page_pool, + virt_to_head_page(xdp->data), + len, true); } break; default: @@ -915,7 +919,9 @@ static u32 netsec_run_xdp(struct netsec_priv *priv, struct bpf_prog *prog, /* fall through -- handle aborts by dropping packet */ case XDP_DROP: ret = NETSEC_XDP_CONSUMED; - xdp_return_buff(xdp); + __page_pool_put_page(dring->page_pool, + virt_to_head_page(xdp->data), + len, true); break; } @@ -1014,7 +1020,8 @@ static int netsec_process_rx(struct netsec_priv *priv, int budget) * cache state. Since we paid the allocation cost if * building an skb fails try to put the page into cache */ - page_pool_recycle_direct(dring->page_pool, page); + __page_pool_put_page(dring->page_pool, page, + pkt_len, true); netif_err(priv, drv, priv->ndev, "rx failed to build skb\n"); break; @@ -1272,17 +1279,19 @@ static int netsec_setup_rx_dring(struct netsec_priv *priv) { struct netsec_desc_ring *dring = &priv->desc_ring[NETSEC_RING_RX]; struct bpf_prog *xdp_prog = READ_ONCE(priv->xdp_prog); - struct page_pool_params pp_params = { 0 }; + struct page_pool_params pp_params = { + .order = 0, + /* internal DMA mapping in page_pool */ + .flags = PP_FLAG_DMA_MAP | PP_FLAG_DMA_SYNC_DEV, + .pool_size = DESC_NUM, + .nid = NUMA_NO_NODE, + .dev = priv->dev, + .dma_dir = xdp_prog ? DMA_BIDIRECTIONAL : DMA_FROM_DEVICE, + .offset = NETSEC_RXBUF_HEADROOM, + .max_len = NETSEC_RX_BUF_SIZE, + }; int i, err; - pp_params.order = 0; - /* internal DMA mapping in page_pool */ - pp_params.flags = PP_FLAG_DMA_MAP; - pp_params.pool_size = DESC_NUM; - pp_params.nid = NUMA_NO_NODE; - pp_params.dev = priv->dev; - pp_params.dma_dir = xdp_prog ? DMA_BIDIRECTIONAL : DMA_FROM_DEVICE; - dring->page_pool = page_pool_create(&pp_params); if (IS_ERR(dring->page_pool)) { err = PTR_ERR(dring->page_pool); @@ -1731,12 +1740,6 @@ static int netsec_netdev_set_features(struct net_device *ndev, return 0; } -static int netsec_netdev_ioctl(struct net_device *ndev, struct ifreq *ifr, - int cmd) -{ - return phy_mii_ioctl(ndev->phydev, ifr, cmd); -} - static int netsec_xdp_xmit(struct net_device *ndev, int n, struct xdp_frame **frames, u32 flags) { @@ -1821,7 +1824,7 @@ static const struct net_device_ops netsec_netdev_ops = { .ndo_set_features = netsec_netdev_set_features, .ndo_set_mac_address = eth_mac_addr, .ndo_validate_addr = eth_validate_addr, - .ndo_do_ioctl = netsec_netdev_ioctl, + .ndo_do_ioctl = phy_do_ioctl, .ndo_xdp_xmit = netsec_xdp_xmit, .ndo_bpf = netsec_xdp, }; diff --git a/drivers/net/ethernet/socionext/sni_ave.c b/drivers/net/ethernet/socionext/sni_ave.c index f7e927ad67fa..b7032422393f 100644 --- a/drivers/net/ethernet/socionext/sni_ave.c +++ b/drivers/net/ethernet/socionext/sni_ave.c @@ -424,16 +424,22 @@ static void ave_ethtool_get_wol(struct net_device *ndev, phy_ethtool_get_wol(ndev->phydev, wol); } -static int ave_ethtool_set_wol(struct net_device *ndev, - struct ethtool_wolinfo *wol) +static int __ave_ethtool_set_wol(struct net_device *ndev, + struct ethtool_wolinfo *wol) { - int ret; - if (!ndev->phydev || (wol->wolopts & (WAKE_ARP | WAKE_MAGICSECURE))) return -EOPNOTSUPP; - ret = phy_ethtool_set_wol(ndev->phydev, wol); + return phy_ethtool_set_wol(ndev->phydev, wol); +} + +static int ave_ethtool_set_wol(struct net_device *ndev, + struct ethtool_wolinfo *wol) +{ + int ret; + + ret = __ave_ethtool_set_wol(ndev, wol); if (!ret) device_set_wakeup_enable(&ndev->dev, !!wol->wolopts); @@ -1216,7 +1222,7 @@ static int ave_init(struct net_device *ndev) /* set wol initial state disabled */ wol.wolopts = 0; - ave_ethtool_set_wol(ndev, &wol); + __ave_ethtool_set_wol(ndev, &wol); if (!phy_interface_is_rgmii(phydev)) phy_set_max_speed(phydev, SPEED_100); @@ -1768,7 +1774,7 @@ static int ave_resume(struct device *dev) ave_ethtool_get_wol(ndev, &wol); wol.wolopts = priv->wolopts; - ave_ethtool_set_wol(ndev, &wol); + __ave_ethtool_set_wol(ndev, &wol); if (ndev->phydev) { ret = phy_resume(ndev->phydev); diff --git a/drivers/net/ethernet/stmicro/stmmac/common.h b/drivers/net/ethernet/stmicro/stmmac/common.h index 31003b67d24f..487099092693 100644 --- a/drivers/net/ethernet/stmicro/stmmac/common.h +++ b/drivers/net/ethernet/stmicro/stmmac/common.h @@ -368,6 +368,7 @@ struct dma_features { unsigned int estdep; unsigned int estsel; unsigned int fpesel; + unsigned int tbssel; }; /* RX Buffer size must be multiple of 4/8/16 bytes */ diff --git a/drivers/net/ethernet/stmicro/stmmac/descs.h b/drivers/net/ethernet/stmicro/stmmac/descs.h index 9f0b9a9e63b3..49d6a866244f 100644 --- a/drivers/net/ethernet/stmicro/stmmac/descs.h +++ b/drivers/net/ethernet/stmicro/stmmac/descs.h @@ -171,6 +171,15 @@ struct dma_extended_desc { __le32 des7; /* Tx/Rx Timestamp High */ }; +/* Enhanced descriptor for TBS */ +struct dma_edesc { + __le32 des4; + __le32 des5; + __le32 des6; + __le32 des7; + struct dma_desc basic; +}; + /* Transmit checksum insertion control */ #define TX_CIC_FULL 3 /* Include IP header and pseudoheader */ diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac-sun8i.c b/drivers/net/ethernet/stmicro/stmmac/dwmac-sun8i.c index 6f834302fda3..58e0511badba 100644 --- a/drivers/net/ethernet/stmicro/stmmac/dwmac-sun8i.c +++ b/drivers/net/ethernet/stmicro/stmmac/dwmac-sun8i.c @@ -973,6 +973,9 @@ static int sun8i_dwmac_set_syscon(struct stmmac_priv *priv) /* default */ break; case PHY_INTERFACE_MODE_RGMII: + case PHY_INTERFACE_MODE_RGMII_ID: + case PHY_INTERFACE_MODE_RGMII_RXID: + case PHY_INTERFACE_MODE_RGMII_TXID: reg |= SYSCON_EPIT | SYSCON_ETCS_INT_GMII; break; case PHY_INTERFACE_MODE_RMII: diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac-sunxi.c b/drivers/net/ethernet/stmicro/stmmac/dwmac-sunxi.c index 26353ef616b8..7d40760e9ba8 100644 --- a/drivers/net/ethernet/stmicro/stmmac/dwmac-sunxi.c +++ b/drivers/net/ethernet/stmicro/stmmac/dwmac-sunxi.c @@ -44,7 +44,7 @@ static int sun7i_gmac_init(struct platform_device *pdev, void *priv) * rate, which then uses the auto-reparenting feature of the * clock driver, and enabling/disabling the clock. */ - if (gmac->interface == PHY_INTERFACE_MODE_RGMII) { + if (phy_interface_mode_is_rgmii(gmac->interface)) { clk_set_rate(gmac->tx_clk, SUN7I_GMAC_GMII_RGMII_RATE); clk_prepare_enable(gmac->tx_clk); gmac->clk_enabled = 1; diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac4.h b/drivers/net/ethernet/stmicro/stmmac/dwmac4.h index 2e6b60a476c6..af50af27550b 100644 --- a/drivers/net/ethernet/stmicro/stmmac/dwmac4.h +++ b/drivers/net/ethernet/stmicro/stmmac/dwmac4.h @@ -239,6 +239,7 @@ enum power_event { /* MAC HW features3 bitmap */ #define GMAC_HW_FEAT_ASP GENMASK(29, 28) +#define GMAC_HW_FEAT_TBSSEL BIT(27) #define GMAC_HW_FEAT_FPESEL BIT(26) #define GMAC_HW_FEAT_ESTWID GENMASK(21, 20) #define GMAC_HW_FEAT_ESTDEP GENMASK(19, 17) diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac4_descs.c b/drivers/net/ethernet/stmicro/stmmac/dwmac4_descs.c index 3e14da69f378..eff82065a501 100644 --- a/drivers/net/ethernet/stmicro/stmmac/dwmac4_descs.c +++ b/drivers/net/ethernet/stmicro/stmmac/dwmac4_descs.c @@ -10,6 +10,7 @@ #include <linux/stmmac.h> #include "common.h" +#include "dwmac4.h" #include "dwmac4_descs.h" static int dwmac4_wrback_get_tx_status(void *data, struct stmmac_extra_stats *x, @@ -505,6 +506,14 @@ static void dwmac4_set_sec_addr(struct dma_desc *p, dma_addr_t addr) p->des3 = cpu_to_le32(upper_32_bits(addr) | RDES3_BUFFER2_VALID_ADDR); } +static void dwmac4_set_tbs(struct dma_edesc *p, u32 sec, u32 nsec) +{ + p->des4 = cpu_to_le32((sec & TDES4_LT) | TDES4_LTV); + p->des5 = cpu_to_le32(nsec & TDES5_LT); + p->des6 = 0; + p->des7 = 0; +} + const struct stmmac_desc_ops dwmac4_desc_ops = { .tx_status = dwmac4_wrback_get_tx_status, .rx_status = dwmac4_wrback_get_rx_status, @@ -534,6 +543,7 @@ const struct stmmac_desc_ops dwmac4_desc_ops = { .set_vlan = dwmac4_set_vlan, .get_rx_header_len = dwmac4_get_rx_header_len, .set_sec_addr = dwmac4_set_sec_addr, + .set_tbs = dwmac4_set_tbs, }; const struct stmmac_mode_ops dwmac4_ring_mode_ops = { diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac4_descs.h b/drivers/net/ethernet/stmicro/stmmac/dwmac4_descs.h index 6d92109dc9aa..6da070ccd737 100644 --- a/drivers/net/ethernet/stmicro/stmmac/dwmac4_descs.h +++ b/drivers/net/ethernet/stmicro/stmmac/dwmac4_descs.h @@ -73,6 +73,13 @@ #define TDES3_CONTEXT_TYPE BIT(30) #define TDES3_CONTEXT_TYPE_SHIFT 30 +/* TDES4 */ +#define TDES4_LTV BIT(31) +#define TDES4_LT GENMASK(7, 0) + +/* TDES5 */ +#define TDES5_LT GENMASK(31, 8) + /* TDS3 use for both format (read and write back) */ #define TDES3_OWN BIT(31) #define TDES3_OWN_SHIFT 31 diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac4_dma.c b/drivers/net/ethernet/stmicro/stmmac/dwmac4_dma.c index 213d44482ffa..bb29bfcd62c3 100644 --- a/drivers/net/ethernet/stmicro/stmmac/dwmac4_dma.c +++ b/drivers/net/ethernet/stmicro/stmmac/dwmac4_dma.c @@ -404,6 +404,7 @@ static void dwmac4_get_hw_feature(void __iomem *ioaddr, /* 5.10 Features */ dma_cap->asp = (hw_cap & GMAC_HW_FEAT_ASP) >> 28; + dma_cap->tbssel = (hw_cap & GMAC_HW_FEAT_TBSSEL) >> 27; dma_cap->fpesel = (hw_cap & GMAC_HW_FEAT_FPESEL) >> 26; dma_cap->estwid = (hw_cap & GMAC_HW_FEAT_ESTWID) >> 20; dma_cap->estdep = (hw_cap & GMAC_HW_FEAT_ESTDEP) >> 17; @@ -471,6 +472,25 @@ static void dwmac4_enable_sph(void __iomem *ioaddr, bool en, u32 chan) writel(value, ioaddr + DMA_CHAN_CONTROL(chan)); } +static int dwmac4_enable_tbs(void __iomem *ioaddr, bool en, u32 chan) +{ + u32 value = readl(ioaddr + DMA_CHAN_TX_CONTROL(chan)); + + if (en) + value |= DMA_CONTROL_EDSE; + else + value &= ~DMA_CONTROL_EDSE; + + writel(value, ioaddr + DMA_CHAN_TX_CONTROL(chan)); + + value = readl(ioaddr + DMA_CHAN_TX_CONTROL(chan)) & DMA_CONTROL_EDSE; + if (en && !value) + return -EIO; + + writel(DMA_TBS_DEF_FTOS, ioaddr + DMA_TBS_CTRL); + return 0; +} + const struct stmmac_dma_ops dwmac4_dma_ops = { .reset = dwmac4_dma_reset, .init = dwmac4_dma_init, @@ -527,4 +547,5 @@ const struct stmmac_dma_ops dwmac410_dma_ops = { .qmode = dwmac4_qmode, .set_bfsize = dwmac4_set_bfsize, .enable_sph = dwmac4_enable_sph, + .enable_tbs = dwmac4_enable_tbs, }; diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac4_dma.h b/drivers/net/ethernet/stmicro/stmmac/dwmac4_dma.h index bcb6d5190f3d..8391ca63d943 100644 --- a/drivers/net/ethernet/stmicro/stmmac/dwmac4_dma.h +++ b/drivers/net/ethernet/stmicro/stmmac/dwmac4_dma.h @@ -22,6 +22,7 @@ #define DMA_DEBUG_STATUS_1 0x00001010 #define DMA_DEBUG_STATUS_2 0x00001014 #define DMA_AXI_BUS_MODE 0x00001028 +#define DMA_TBS_CTRL 0x00001050 /* DMA Bus Mode bitmap */ #define DMA_BUS_MODE_SFT_RESET BIT(0) @@ -82,6 +83,11 @@ #define DMA_AXI_BURST_LEN_MASK 0x000000FE +/* DMA TBS Control */ +#define DMA_TBS_FTOS GENMASK(31, 8) +#define DMA_TBS_FTOV BIT(0) +#define DMA_TBS_DEF_FTOS (DMA_TBS_FTOS | DMA_TBS_FTOV) + /* Following DMA defines are chanels oriented */ #define DMA_CHAN_BASE_ADDR 0x00001100 #define DMA_CHAN_BASE_OFFSET 0x80 @@ -114,6 +120,7 @@ #define DMA_CONTROL_MSS_MASK GENMASK(13, 0) /* DMA Tx Channel X Control register defines */ +#define DMA_CONTROL_EDSE BIT(28) #define DMA_CONTROL_TSE BIT(12) #define DMA_CONTROL_OSP BIT(4) #define DMA_CONTROL_ST BIT(0) diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac5.c b/drivers/net/ethernet/stmicro/stmmac/dwmac5.c index 5d4a3c2458ea..494c859b4ade 100644 --- a/drivers/net/ethernet/stmicro/stmmac/dwmac5.c +++ b/drivers/net/ethernet/stmicro/stmmac/dwmac5.c @@ -657,6 +657,7 @@ void dwmac5_fpe_configure(void __iomem *ioaddr, u32 num_txq, u32 num_rxq, value &= ~EFPE; writel(value, ioaddr + MAC_FPE_CTRL_STS); + return; } value = readl(ioaddr + GMAC_RXQ_CTRL1); diff --git a/drivers/net/ethernet/stmicro/stmmac/dwxgmac2.h b/drivers/net/ethernet/stmicro/stmmac/dwxgmac2.h index 64d13e50e403..6c3b8a950f58 100644 --- a/drivers/net/ethernet/stmicro/stmmac/dwxgmac2.h +++ b/drivers/net/ethernet/stmicro/stmmac/dwxgmac2.h @@ -139,6 +139,7 @@ #define XGMAC_HWFEAT_TXQCNT GENMASK(9, 6) #define XGMAC_HWFEAT_RXQCNT GENMASK(3, 0) #define XGMAC_HW_FEATURE3 0x00000128 +#define XGMAC_HWFEAT_TBSSEL BIT(27) #define XGMAC_HWFEAT_FPESEL BIT(26) #define XGMAC_HWFEAT_ESTWID GENMASK(24, 23) #define XGMAC_HWFEAT_ESTDEP GENMASK(22, 20) @@ -346,6 +347,13 @@ #define XGMAC_TDPS GENMASK(29, 0) #define XGMAC_RX_EDMA_CTRL 0x00003044 #define XGMAC_RDPS GENMASK(29, 0) +#define XGMAC_DMA_TBS_CTRL0 0x00003054 +#define XGMAC_DMA_TBS_CTRL1 0x00003058 +#define XGMAC_DMA_TBS_CTRL2 0x0000305c +#define XGMAC_DMA_TBS_CTRL3 0x00003060 +#define XGMAC_FTOS GENMASK(31, 8) +#define XGMAC_FTOV BIT(0) +#define XGMAC_DEF_FTOS (XGMAC_FTOS | XGMAC_FTOV) #define XGMAC_DMA_SAFETY_INT_STATUS 0x00003064 #define XGMAC_MCSIS BIT(31) #define XGMAC_MSUIS BIT(29) @@ -360,6 +368,7 @@ #define XGMAC_SPH BIT(24) #define XGMAC_PBLx8 BIT(16) #define XGMAC_DMA_CH_TX_CONTROL(x) (0x00003104 + (0x80 * (x))) +#define XGMAC_EDSE BIT(28) #define XGMAC_TxPBL GENMASK(21, 16) #define XGMAC_TxPBL_SHIFT 16 #define XGMAC_TSE BIT(12) @@ -404,6 +413,9 @@ #define XGMAC_REGSIZE ((0x0000317c + (0x80 * 15)) / 4) /* Descriptors */ +#define XGMAC_TDES0_LTV BIT(31) +#define XGMAC_TDES0_LT GENMASK(7, 0) +#define XGMAC_TDES1_LT GENMASK(31, 8) #define XGMAC_TDES2_IVT GENMASK(31, 16) #define XGMAC_TDES2_IVT_SHIFT 16 #define XGMAC_TDES2_IOC BIT(31) @@ -422,6 +434,7 @@ #define XGMAC_TDES3_TCMSSV BIT(26) #define XGMAC_TDES3_SAIC GENMASK(25, 23) #define XGMAC_TDES3_SAIC_SHIFT 23 +#define XGMAC_TDES3_TBSV BIT(24) #define XGMAC_TDES3_THL GENMASK(22, 19) #define XGMAC_TDES3_THL_SHIFT 19 #define XGMAC_TDES3_IVTIR GENMASK(19, 18) diff --git a/drivers/net/ethernet/stmicro/stmmac/dwxgmac2_core.c b/drivers/net/ethernet/stmicro/stmmac/dwxgmac2_core.c index 307105e8dea0..2af3ac5409b7 100644 --- a/drivers/net/ethernet/stmicro/stmmac/dwxgmac2_core.c +++ b/drivers/net/ethernet/stmicro/stmmac/dwxgmac2_core.c @@ -1421,6 +1421,7 @@ static void dwxgmac3_fpe_configure(void __iomem *ioaddr, u32 num_txq, value &= ~XGMAC_EFPE; writel(value, ioaddr + XGMAC_FPE_CTRL_STS); + return; } value = readl(ioaddr + XGMAC_RXQ_CTRL1); diff --git a/drivers/net/ethernet/stmicro/stmmac/dwxgmac2_descs.c b/drivers/net/ethernet/stmicro/stmmac/dwxgmac2_descs.c index bd5838ce1e8a..c3d654cfa9ef 100644 --- a/drivers/net/ethernet/stmicro/stmmac/dwxgmac2_descs.c +++ b/drivers/net/ethernet/stmicro/stmmac/dwxgmac2_descs.c @@ -339,6 +339,14 @@ static void dwxgmac2_set_vlan(struct dma_desc *p, u32 type) p->des2 |= cpu_to_le32(type & XGMAC_TDES2_VTIR); } +static void dwxgmac2_set_tbs(struct dma_edesc *p, u32 sec, u32 nsec) +{ + p->des4 = cpu_to_le32((sec & XGMAC_TDES0_LT) | XGMAC_TDES0_LTV); + p->des5 = cpu_to_le32(nsec & XGMAC_TDES1_LT); + p->des6 = 0; + p->des7 = 0; +} + const struct stmmac_desc_ops dwxgmac210_desc_ops = { .tx_status = dwxgmac2_get_tx_status, .rx_status = dwxgmac2_get_rx_status, @@ -368,4 +376,5 @@ const struct stmmac_desc_ops dwxgmac210_desc_ops = { .set_sarc = dwxgmac2_set_sarc, .set_vlan_tag = dwxgmac2_set_vlan_tag, .set_vlan = dwxgmac2_set_vlan, + .set_tbs = dwxgmac2_set_tbs, }; diff --git a/drivers/net/ethernet/stmicro/stmmac/dwxgmac2_dma.c b/drivers/net/ethernet/stmicro/stmmac/dwxgmac2_dma.c index bbbfa793a367..77308c5c5d29 100644 --- a/drivers/net/ethernet/stmicro/stmmac/dwxgmac2_dma.c +++ b/drivers/net/ethernet/stmicro/stmmac/dwxgmac2_dma.c @@ -429,6 +429,7 @@ static void dwxgmac2_get_hw_feature(void __iomem *ioaddr, /* MAC HW feature 3 */ hw_cap = readl(ioaddr + XGMAC_HW_FEATURE3); + dma_cap->tbssel = (hw_cap & XGMAC_HWFEAT_TBSSEL) >> 27; dma_cap->fpesel = (hw_cap & XGMAC_HWFEAT_FPESEL) >> 26; dma_cap->estwid = (hw_cap & XGMAC_HWFEAT_ESTWID) >> 23; dma_cap->estdep = (hw_cap & XGMAC_HWFEAT_ESTDEP) >> 20; @@ -523,6 +524,28 @@ static void dwxgmac2_enable_sph(void __iomem *ioaddr, bool en, u32 chan) writel(value, ioaddr + XGMAC_DMA_CH_CONTROL(chan)); } +static int dwxgmac2_enable_tbs(void __iomem *ioaddr, bool en, u32 chan) +{ + u32 value = readl(ioaddr + XGMAC_DMA_CH_TX_CONTROL(chan)); + + if (en) + value |= XGMAC_EDSE; + else + value &= ~XGMAC_EDSE; + + writel(value, ioaddr + XGMAC_DMA_CH_TX_CONTROL(chan)); + + value = readl(ioaddr + XGMAC_DMA_CH_TX_CONTROL(chan)) & XGMAC_EDSE; + if (en && !value) + return -EIO; + + writel(XGMAC_DEF_FTOS, ioaddr + XGMAC_DMA_TBS_CTRL0); + writel(XGMAC_DEF_FTOS, ioaddr + XGMAC_DMA_TBS_CTRL1); + writel(XGMAC_DEF_FTOS, ioaddr + XGMAC_DMA_TBS_CTRL2); + writel(XGMAC_DEF_FTOS, ioaddr + XGMAC_DMA_TBS_CTRL3); + return 0; +} + const struct stmmac_dma_ops dwxgmac210_dma_ops = { .reset = dwxgmac2_dma_reset, .init = dwxgmac2_dma_init, @@ -550,4 +573,5 @@ const struct stmmac_dma_ops dwxgmac210_dma_ops = { .qmode = dwxgmac2_qmode, .set_bfsize = dwxgmac2_set_bfsize, .enable_sph = dwxgmac2_enable_sph, + .enable_tbs = dwxgmac2_enable_tbs, }; diff --git a/drivers/net/ethernet/stmicro/stmmac/hwif.h b/drivers/net/ethernet/stmicro/stmmac/hwif.h index 905a6f0edaca..df63b0367aff 100644 --- a/drivers/net/ethernet/stmicro/stmmac/hwif.h +++ b/drivers/net/ethernet/stmicro/stmmac/hwif.h @@ -29,6 +29,7 @@ struct stmmac_extra_stats; struct stmmac_safety_stats; struct dma_desc; struct dma_extended_desc; +struct dma_edesc; /* Descriptors helpers */ struct stmmac_desc_ops { @@ -95,6 +96,7 @@ struct stmmac_desc_ops { void (*set_vlan_tag)(struct dma_desc *p, u16 tag, u16 inner_tag, u32 inner_type); void (*set_vlan)(struct dma_desc *p, u32 type); + void (*set_tbs)(struct dma_edesc *p, u32 sec, u32 nsec); }; #define stmmac_init_rx_desc(__priv, __args...) \ @@ -157,6 +159,8 @@ struct stmmac_desc_ops { stmmac_do_void_callback(__priv, desc, set_vlan_tag, __args) #define stmmac_set_desc_vlan(__priv, __args...) \ stmmac_do_void_callback(__priv, desc, set_vlan, __args) +#define stmmac_set_desc_tbs(__priv, __args...) \ + stmmac_do_void_callback(__priv, desc, set_tbs, __args) struct stmmac_dma_cfg; struct dma_features; @@ -210,6 +214,7 @@ struct stmmac_dma_ops { void (*qmode)(void __iomem *ioaddr, u32 channel, u8 qmode); void (*set_bfsize)(void __iomem *ioaddr, int bfsize, u32 chan); void (*enable_sph)(void __iomem *ioaddr, bool en, u32 chan); + int (*enable_tbs)(void __iomem *ioaddr, bool en, u32 chan); }; #define stmmac_reset(__priv, __args...) \ @@ -268,6 +273,8 @@ struct stmmac_dma_ops { stmmac_do_void_callback(__priv, dma, set_bfsize, __args) #define stmmac_enable_sph(__priv, __args...) \ stmmac_do_void_callback(__priv, dma, enable_sph, __args) +#define stmmac_enable_tbs(__priv, __args...) \ + stmmac_do_callback(__priv, dma, enable_tbs, __args) struct mac_device_info; struct net_device; @@ -526,6 +533,7 @@ struct tc_cls_u32_offload; struct tc_cbs_qopt_offload; struct flow_cls_offload; struct tc_taprio_qopt_offload; +struct tc_etf_qopt_offload; struct stmmac_tc_ops { int (*init)(struct stmmac_priv *priv); @@ -537,6 +545,8 @@ struct stmmac_tc_ops { struct flow_cls_offload *cls); int (*setup_taprio)(struct stmmac_priv *priv, struct tc_taprio_qopt_offload *qopt); + int (*setup_etf)(struct stmmac_priv *priv, + struct tc_etf_qopt_offload *qopt); }; #define stmmac_tc_init(__priv, __args...) \ @@ -549,6 +559,8 @@ struct stmmac_tc_ops { stmmac_do_callback(__priv, tc, setup_cls, __args) #define stmmac_tc_setup_taprio(__priv, __args...) \ stmmac_do_callback(__priv, tc, setup_taprio, __args) +#define stmmac_tc_setup_etf(__priv, __args...) \ + stmmac_do_callback(__priv, tc, setup_etf, __args) struct stmmac_counters; diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac.h b/drivers/net/ethernet/stmicro/stmmac/stmmac.h index f98c5eefb382..9c02fc754bf1 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac.h +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac.h @@ -39,13 +39,18 @@ struct stmmac_tx_info { bool is_jumbo; }; +#define STMMAC_TBS_AVAIL BIT(0) +#define STMMAC_TBS_EN BIT(1) + /* Frequently used values are kept adjacent for cache effect */ struct stmmac_tx_queue { u32 tx_count_frames; + int tbs; struct timer_list txtimer; u32 queue_index; struct stmmac_priv *priv_data; struct dma_extended_desc *dma_etx ____cacheline_aligned_in_smp; + struct dma_edesc *dma_entx; struct dma_desc *dma_tx; struct sk_buff **tx_skbuff; struct stmmac_tx_info *tx_skbuff_dma; diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c index f19b8b15ed80..ff1cbfc834b0 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c @@ -106,6 +106,7 @@ MODULE_PARM_DESC(chain_mode, "To use chain instead of ring mode"); static irqreturn_t stmmac_interrupt(int irq, void *dev_id); #ifdef CONFIG_DEBUG_FS +static const struct net_device_ops stmmac_netdev_ops; static void stmmac_init_fs(struct net_device *dev); static void stmmac_exit_fs(struct net_device *dev); #endif @@ -387,9 +388,8 @@ bool stmmac_eee_init(struct stmmac_priv *priv) /* Using PCS we cannot dial with the phy registers at this stage * so we do not support extra feature like EEE. */ - if ((priv->hw->pcs == STMMAC_PCS_RGMII) || - (priv->hw->pcs == STMMAC_PCS_TBI) || - (priv->hw->pcs == STMMAC_PCS_RTBI)) + if (priv->hw->pcs == STMMAC_PCS_TBI || + priv->hw->pcs == STMMAC_PCS_RTBI) return false; /* Check if MAC core supports the EEE feature. */ @@ -1089,6 +1089,8 @@ static void stmmac_display_tx_rings(struct stmmac_priv *priv) if (priv->extend_desc) head_tx = (void *)tx_q->dma_etx; + else if (tx_q->tbs & STMMAC_TBS_AVAIL) + head_tx = (void *)tx_q->dma_entx; else head_tx = (void *)tx_q->dma_tx; @@ -1162,13 +1164,19 @@ static void stmmac_clear_tx_descriptors(struct stmmac_priv *priv, u32 queue) int i; /* Clear the TX descriptors */ - for (i = 0; i < DMA_TX_SIZE; i++) + for (i = 0; i < DMA_TX_SIZE; i++) { + int last = (i == (DMA_TX_SIZE - 1)); + struct dma_desc *p; + if (priv->extend_desc) - stmmac_init_tx_desc(priv, &tx_q->dma_etx[i].basic, - priv->mode, (i == DMA_TX_SIZE - 1)); + p = &tx_q->dma_etx[i].basic; + else if (tx_q->tbs & STMMAC_TBS_AVAIL) + p = &tx_q->dma_entx[i].basic; else - stmmac_init_tx_desc(priv, &tx_q->dma_tx[i], - priv->mode, (i == DMA_TX_SIZE - 1)); + p = &tx_q->dma_tx[i]; + + stmmac_init_tx_desc(priv, p, priv->mode, last); + } } /** @@ -1382,7 +1390,7 @@ static int init_dma_tx_desc_rings(struct net_device *dev) if (priv->extend_desc) stmmac_mode_init(priv, tx_q->dma_etx, tx_q->dma_tx_phy, DMA_TX_SIZE, 1); - else + else if (!(tx_q->tbs & STMMAC_TBS_AVAIL)) stmmac_mode_init(priv, tx_q->dma_tx, tx_q->dma_tx_phy, DMA_TX_SIZE, 0); } @@ -1391,6 +1399,8 @@ static int init_dma_tx_desc_rings(struct net_device *dev) struct dma_desc *p; if (priv->extend_desc) p = &((tx_q->dma_etx + i)->basic); + else if (tx_q->tbs & STMMAC_TBS_AVAIL) + p = &((tx_q->dma_entx + i)->basic); else p = tx_q->dma_tx + i; @@ -1510,19 +1520,26 @@ static void free_dma_tx_desc_resources(struct stmmac_priv *priv) /* Free TX queue resources */ for (queue = 0; queue < tx_count; queue++) { struct stmmac_tx_queue *tx_q = &priv->tx_queue[queue]; + size_t size; + void *addr; /* Release the DMA TX socket buffers */ dma_free_tx_skbufs(priv, queue); - /* Free DMA regions of consistent memory previously allocated */ - if (!priv->extend_desc) - dma_free_coherent(priv->device, - DMA_TX_SIZE * sizeof(struct dma_desc), - tx_q->dma_tx, tx_q->dma_tx_phy); - else - dma_free_coherent(priv->device, DMA_TX_SIZE * - sizeof(struct dma_extended_desc), - tx_q->dma_etx, tx_q->dma_tx_phy); + if (priv->extend_desc) { + size = sizeof(struct dma_extended_desc); + addr = tx_q->dma_etx; + } else if (tx_q->tbs & STMMAC_TBS_AVAIL) { + size = sizeof(struct dma_edesc); + addr = tx_q->dma_entx; + } else { + size = sizeof(struct dma_desc); + addr = tx_q->dma_tx; + } + + size *= DMA_TX_SIZE; + + dma_free_coherent(priv->device, size, addr, tx_q->dma_tx_phy); kfree(tx_q->tx_skbuff_dma); kfree(tx_q->tx_skbuff); @@ -1615,6 +1632,8 @@ static int alloc_dma_tx_desc_resources(struct stmmac_priv *priv) /* TX queues buffers and DMA */ for (queue = 0; queue < tx_count; queue++) { struct stmmac_tx_queue *tx_q = &priv->tx_queue[queue]; + size_t size; + void *addr; tx_q->queue_index = queue; tx_q->priv_data = priv; @@ -1631,28 +1650,32 @@ static int alloc_dma_tx_desc_resources(struct stmmac_priv *priv) if (!tx_q->tx_skbuff) goto err_dma; - if (priv->extend_desc) { - tx_q->dma_etx = dma_alloc_coherent(priv->device, - DMA_TX_SIZE * sizeof(struct dma_extended_desc), - &tx_q->dma_tx_phy, - GFP_KERNEL); - if (!tx_q->dma_etx) - goto err_dma; - } else { - tx_q->dma_tx = dma_alloc_coherent(priv->device, - DMA_TX_SIZE * sizeof(struct dma_desc), - &tx_q->dma_tx_phy, - GFP_KERNEL); - if (!tx_q->dma_tx) - goto err_dma; - } + if (priv->extend_desc) + size = sizeof(struct dma_extended_desc); + else if (tx_q->tbs & STMMAC_TBS_AVAIL) + size = sizeof(struct dma_edesc); + else + size = sizeof(struct dma_desc); + + size *= DMA_TX_SIZE; + + addr = dma_alloc_coherent(priv->device, size, + &tx_q->dma_tx_phy, GFP_KERNEL); + if (!addr) + goto err_dma; + + if (priv->extend_desc) + tx_q->dma_etx = addr; + else if (tx_q->tbs & STMMAC_TBS_AVAIL) + tx_q->dma_entx = addr; + else + tx_q->dma_tx = addr; } return 0; err_dma: free_dma_tx_desc_resources(priv); - return ret; } @@ -1884,6 +1907,8 @@ static int stmmac_tx_clean(struct stmmac_priv *priv, int budget, u32 queue) if (priv->extend_desc) p = (struct dma_desc *)(tx_q->dma_etx + entry); + else if (tx_q->tbs & STMMAC_TBS_AVAIL) + p = &tx_q->dma_entx[entry].basic; else p = tx_q->dma_tx + entry; @@ -1982,19 +2007,12 @@ static int stmmac_tx_clean(struct stmmac_priv *priv, int budget, u32 queue) static void stmmac_tx_err(struct stmmac_priv *priv, u32 chan) { struct stmmac_tx_queue *tx_q = &priv->tx_queue[chan]; - int i; netif_tx_stop_queue(netdev_get_tx_queue(priv->dev, chan)); stmmac_stop_tx_dma(priv, chan); dma_free_tx_skbufs(priv, chan); - for (i = 0; i < DMA_TX_SIZE; i++) - if (priv->extend_desc) - stmmac_init_tx_desc(priv, &tx_q->dma_etx[i].basic, - priv->mode, (i == DMA_TX_SIZE - 1)); - else - stmmac_init_tx_desc(priv, &tx_q->dma_tx[i], - priv->mode, (i == DMA_TX_SIZE - 1)); + stmmac_clear_tx_descriptors(priv, chan); tx_q->dirty_tx = 0; tx_q->cur_tx = 0; tx_q->mss = 0; @@ -2631,6 +2649,14 @@ static int stmmac_hw_setup(struct net_device *dev, bool init_ptp) if (priv->dma_cap.vlins) stmmac_enable_vlan(priv, priv->hw, STMMAC_VLAN_INSERT); + /* TBS */ + for (chan = 0; chan < tx_cnt; chan++) { + struct stmmac_tx_queue *tx_q = &priv->tx_queue[chan]; + int enable = tx_q->tbs & STMMAC_TBS_AVAIL; + + stmmac_enable_tbs(priv, priv->ioaddr, enable, chan); + } + /* Start the ball rolling... */ stmmac_start_all_dma(priv); @@ -2660,8 +2686,7 @@ static int stmmac_open(struct net_device *dev) u32 chan; int ret; - if (priv->hw->pcs != STMMAC_PCS_RGMII && - priv->hw->pcs != STMMAC_PCS_TBI && + if (priv->hw->pcs != STMMAC_PCS_TBI && priv->hw->pcs != STMMAC_PCS_RTBI) { ret = stmmac_init_phy(dev); if (ret) { @@ -2688,6 +2713,16 @@ static int stmmac_open(struct net_device *dev) priv->rx_copybreak = STMMAC_RX_COPYBREAK; + /* Earlier check for TBS */ + for (chan = 0; chan < priv->plat->tx_queues_to_use; chan++) { + struct stmmac_tx_queue *tx_q = &priv->tx_queue[chan]; + int tbs_en = priv->plat->tx_queues_cfg[chan].tbs_en; + + tx_q->tbs |= tbs_en ? STMMAC_TBS_AVAIL : 0; + if (stmmac_enable_tbs(priv, priv->ioaddr, tbs_en, chan)) + tx_q->tbs &= ~STMMAC_TBS_AVAIL; + } + ret = alloc_dma_desc_resources(priv); if (ret < 0) { netdev_err(priv->dev, "%s: DMA descriptors allocation failed\n", @@ -2836,7 +2871,11 @@ static bool stmmac_vlan_insert(struct stmmac_priv *priv, struct sk_buff *skb, tag = skb_vlan_tag_get(skb); - p = tx_q->dma_tx + tx_q->cur_tx; + if (tx_q->tbs & STMMAC_TBS_AVAIL) + p = &tx_q->dma_entx[tx_q->cur_tx].basic; + else + p = &tx_q->dma_tx[tx_q->cur_tx]; + if (stmmac_set_desc_vlan_tag(priv, p, tag, inner_tag, inner_type)) return false; @@ -2871,7 +2910,11 @@ static void stmmac_tso_allocator(struct stmmac_priv *priv, dma_addr_t des, tx_q->cur_tx = STMMAC_GET_ENTRY(tx_q->cur_tx, DMA_TX_SIZE); WARN_ON(tx_q->tx_skbuff[tx_q->cur_tx]); - desc = tx_q->dma_tx + tx_q->cur_tx; + + if (tx_q->tbs & STMMAC_TBS_AVAIL) + desc = &tx_q->dma_entx[tx_q->cur_tx].basic; + else + desc = &tx_q->dma_tx[tx_q->cur_tx]; curr_addr = des + (total_len - tmp_len); if (priv->dma_cap.addr64 <= 32) @@ -2922,13 +2965,13 @@ static netdev_tx_t stmmac_tso_xmit(struct sk_buff *skb, struct net_device *dev) { struct dma_desc *desc, *first, *mss_desc = NULL; struct stmmac_priv *priv = netdev_priv(dev); + int desc_size, tmp_pay_len = 0, first_tx; int nfrags = skb_shinfo(skb)->nr_frags; u32 queue = skb_get_queue_mapping(skb); unsigned int first_entry, tx_packets; - int tmp_pay_len = 0, first_tx; struct stmmac_tx_queue *tx_q; - u8 proto_hdr_len, hdr; bool has_vlan, set_ic; + u8 proto_hdr_len, hdr; u32 pay_len, mss; dma_addr_t des; int i; @@ -2965,7 +3008,11 @@ static netdev_tx_t stmmac_tso_xmit(struct sk_buff *skb, struct net_device *dev) /* set new MSS value if needed */ if (mss != tx_q->mss) { - mss_desc = tx_q->dma_tx + tx_q->cur_tx; + if (tx_q->tbs & STMMAC_TBS_AVAIL) + mss_desc = &tx_q->dma_entx[tx_q->cur_tx].basic; + else + mss_desc = &tx_q->dma_tx[tx_q->cur_tx]; + stmmac_set_mss(priv, mss_desc, mss); tx_q->mss = mss; tx_q->cur_tx = STMMAC_GET_ENTRY(tx_q->cur_tx, DMA_TX_SIZE); @@ -2985,7 +3032,10 @@ static netdev_tx_t stmmac_tso_xmit(struct sk_buff *skb, struct net_device *dev) first_entry = tx_q->cur_tx; WARN_ON(tx_q->tx_skbuff[first_entry]); - desc = tx_q->dma_tx + first_entry; + if (tx_q->tbs & STMMAC_TBS_AVAIL) + desc = &tx_q->dma_entx[first_entry].basic; + else + desc = &tx_q->dma_tx[first_entry]; first = desc; if (has_vlan) @@ -3057,7 +3107,11 @@ static netdev_tx_t stmmac_tso_xmit(struct sk_buff *skb, struct net_device *dev) set_ic = false; if (set_ic) { - desc = &tx_q->dma_tx[tx_q->cur_tx]; + if (tx_q->tbs & STMMAC_TBS_AVAIL) + desc = &tx_q->dma_entx[tx_q->cur_tx].basic; + else + desc = &tx_q->dma_tx[tx_q->cur_tx]; + tx_q->tx_count_frames = 0; stmmac_set_tx_ic(priv, desc); priv->xstats.tx_set_ic_bit++; @@ -3120,16 +3174,18 @@ static netdev_tx_t stmmac_tso_xmit(struct sk_buff *skb, struct net_device *dev) pr_info("%s: curr=%d dirty=%d f=%d, e=%d, f_p=%p, nfrags %d\n", __func__, tx_q->cur_tx, tx_q->dirty_tx, first_entry, tx_q->cur_tx, first, nfrags); - - stmmac_display_ring(priv, (void *)tx_q->dma_tx, DMA_TX_SIZE, 0); - pr_info(">>> frame to be transmitted: "); print_pkt(skb->data, skb_headlen(skb)); } netdev_tx_sent_queue(netdev_get_tx_queue(dev, queue), skb->len); - tx_q->tx_tail_addr = tx_q->dma_tx_phy + (tx_q->cur_tx * sizeof(*desc)); + if (tx_q->tbs & STMMAC_TBS_AVAIL) + desc_size = sizeof(struct dma_edesc); + else + desc_size = sizeof(struct dma_desc); + + tx_q->tx_tail_addr = tx_q->dma_tx_phy + (tx_q->cur_tx * desc_size); stmmac_set_tx_tail_ptr(priv, priv->ioaddr, tx_q->tx_tail_addr, queue); stmmac_tx_timer_arm(priv, queue); @@ -3159,10 +3215,11 @@ static netdev_tx_t stmmac_xmit(struct sk_buff *skb, struct net_device *dev) u32 queue = skb_get_queue_mapping(skb); int nfrags = skb_shinfo(skb)->nr_frags; int gso = skb_shinfo(skb)->gso_type; + struct dma_edesc *tbs_desc = NULL; + int entry, desc_size, first_tx; struct dma_desc *desc, *first; struct stmmac_tx_queue *tx_q; bool has_vlan, set_ic; - int entry, first_tx; dma_addr_t des; tx_q = &priv->tx_queue[queue]; @@ -3202,6 +3259,8 @@ static netdev_tx_t stmmac_xmit(struct sk_buff *skb, struct net_device *dev) if (likely(priv->extend_desc)) desc = (struct dma_desc *)(tx_q->dma_etx + entry); + else if (tx_q->tbs & STMMAC_TBS_AVAIL) + desc = &tx_q->dma_entx[entry].basic; else desc = tx_q->dma_tx + entry; @@ -3231,6 +3290,8 @@ static netdev_tx_t stmmac_xmit(struct sk_buff *skb, struct net_device *dev) if (likely(priv->extend_desc)) desc = (struct dma_desc *)(tx_q->dma_etx + entry); + else if (tx_q->tbs & STMMAC_TBS_AVAIL) + desc = &tx_q->dma_entx[entry].basic; else desc = tx_q->dma_tx + entry; @@ -3277,6 +3338,8 @@ static netdev_tx_t stmmac_xmit(struct sk_buff *skb, struct net_device *dev) if (set_ic) { if (likely(priv->extend_desc)) desc = &tx_q->dma_etx[entry].basic; + else if (tx_q->tbs & STMMAC_TBS_AVAIL) + desc = &tx_q->dma_entx[entry].basic; else desc = &tx_q->dma_tx[entry]; @@ -3294,20 +3357,11 @@ static netdev_tx_t stmmac_xmit(struct sk_buff *skb, struct net_device *dev) tx_q->cur_tx = entry; if (netif_msg_pktdata(priv)) { - void *tx_head; - netdev_dbg(priv->dev, "%s: curr=%d dirty=%d f=%d, e=%d, first=%p, nfrags=%d", __func__, tx_q->cur_tx, tx_q->dirty_tx, first_entry, entry, first, nfrags); - if (priv->extend_desc) - tx_head = (void *)tx_q->dma_etx; - else - tx_head = (void *)tx_q->dma_tx; - - stmmac_display_ring(priv, tx_head, DMA_TX_SIZE, false); - netdev_dbg(priv->dev, ">>> frame to be transmitted: "); print_pkt(skb->data, skb->len); } @@ -3353,12 +3407,19 @@ static netdev_tx_t stmmac_xmit(struct sk_buff *skb, struct net_device *dev) /* Prepare the first descriptor setting the OWN bit too */ stmmac_prepare_tx_desc(priv, first, 1, nopaged_len, - csum_insertion, priv->mode, 1, last_segment, + csum_insertion, priv->mode, 0, last_segment, skb->len); - } else { - stmmac_set_tx_owner(priv, first); } + if (tx_q->tbs & STMMAC_TBS_EN) { + struct timespec64 ts = ns_to_timespec64(skb->tstamp); + + tbs_desc = &tx_q->dma_entx[first_entry]; + stmmac_set_desc_tbs(priv, tbs_desc, ts.tv_sec, ts.tv_nsec); + } + + stmmac_set_tx_owner(priv, first); + /* The own bit must be the latest setting done when prepare the * descriptor and then barrier is needed to make sure that * all is coherent before granting the DMA engine. @@ -3369,7 +3430,14 @@ static netdev_tx_t stmmac_xmit(struct sk_buff *skb, struct net_device *dev) stmmac_enable_dma_transmission(priv, priv->ioaddr); - tx_q->tx_tail_addr = tx_q->dma_tx_phy + (tx_q->cur_tx * sizeof(*desc)); + if (likely(priv->extend_desc)) + desc_size = sizeof(struct dma_extended_desc); + else if (tx_q->tbs & STMMAC_TBS_AVAIL) + desc_size = sizeof(struct dma_edesc); + else + desc_size = sizeof(struct dma_desc); + + tx_q->tx_tail_addr = tx_q->dma_tx_phy + (tx_q->cur_tx * desc_size); stmmac_set_tx_tail_ptr(priv, priv->ioaddr, tx_q->tx_tail_addr, queue); stmmac_tx_timer_arm(priv, queue); @@ -4089,6 +4157,8 @@ static int stmmac_setup_tc(struct net_device *ndev, enum tc_setup_type type, return stmmac_tc_setup_cbs(priv, priv, type_data); case TC_SETUP_QDISC_TAPRIO: return stmmac_tc_setup_taprio(priv, priv, type_data); + case TC_SETUP_QDISC_ETF: + return stmmac_tc_setup_etf(priv, priv, type_data); default: return -EOPNOTSUPP; } @@ -4192,7 +4262,7 @@ static int stmmac_rings_status_show(struct seq_file *seq, void *v) seq_printf(seq, "Extended descriptor ring:\n"); sysfs_display_ring((void *)tx_q->dma_etx, DMA_TX_SIZE, 1, seq); - } else { + } else if (!(tx_q->tbs & STMMAC_TBS_AVAIL)) { seq_printf(seq, "Descriptor ring:\n"); sysfs_display_ring((void *)tx_q->dma_tx, DMA_TX_SIZE, 0, seq); @@ -4293,10 +4363,44 @@ static int stmmac_dma_cap_show(struct seq_file *seq, void *v) priv->dma_cap.l3l4fnum); seq_printf(seq, "\tARP Offloading: %s\n", priv->dma_cap.arpoffsel ? "Y" : "N"); + seq_printf(seq, "\tEnhancements to Scheduled Traffic (EST): %s\n", + priv->dma_cap.estsel ? "Y" : "N"); + seq_printf(seq, "\tFrame Preemption (FPE): %s\n", + priv->dma_cap.fpesel ? "Y" : "N"); + seq_printf(seq, "\tTime-Based Scheduling (TBS): %s\n", + priv->dma_cap.tbssel ? "Y" : "N"); return 0; } DEFINE_SHOW_ATTRIBUTE(stmmac_dma_cap); +/* Use network device events to rename debugfs file entries. + */ +static int stmmac_device_event(struct notifier_block *unused, + unsigned long event, void *ptr) +{ + struct net_device *dev = netdev_notifier_info_to_dev(ptr); + struct stmmac_priv *priv = netdev_priv(dev); + + if (dev->netdev_ops != &stmmac_netdev_ops) + goto done; + + switch (event) { + case NETDEV_CHANGENAME: + if (priv->dbgfs_dir) + priv->dbgfs_dir = debugfs_rename(stmmac_fs_dir, + priv->dbgfs_dir, + stmmac_fs_dir, + dev->name); + break; + } +done: + return NOTIFY_DONE; +} + +static struct notifier_block stmmac_notifier = { + .notifier_call = stmmac_device_event, +}; + static void stmmac_init_fs(struct net_device *dev) { struct stmmac_priv *priv = netdev_priv(dev); @@ -4311,12 +4415,15 @@ static void stmmac_init_fs(struct net_device *dev) /* Entry to report the DMA HW features */ debugfs_create_file("dma_cap", 0444, priv->dbgfs_dir, dev, &stmmac_dma_cap_fops); + + register_netdevice_notifier(&stmmac_notifier); } static void stmmac_exit_fs(struct net_device *dev) { struct stmmac_priv *priv = netdev_priv(dev); + unregister_netdevice_notifier(&stmmac_notifier); debugfs_remove_recursive(priv->dbgfs_dir); } #endif /* CONFIG_DEBUG_FS */ @@ -4767,8 +4874,7 @@ int stmmac_dvr_probe(struct device *device, stmmac_check_pcs_mode(priv); - if (priv->hw->pcs != STMMAC_PCS_RGMII && - priv->hw->pcs != STMMAC_PCS_TBI && + if (priv->hw->pcs != STMMAC_PCS_TBI && priv->hw->pcs != STMMAC_PCS_RTBI) { /* MDIO bus Registration */ ret = stmmac_mdio_register(ndev); @@ -4802,8 +4908,7 @@ int stmmac_dvr_probe(struct device *device, error_netdev_register: phylink_destroy(priv->phylink); error_phy_setup: - if (priv->hw->pcs != STMMAC_PCS_RGMII && - priv->hw->pcs != STMMAC_PCS_TBI && + if (priv->hw->pcs != STMMAC_PCS_TBI && priv->hw->pcs != STMMAC_PCS_RTBI) stmmac_mdio_unregister(ndev); error_mdio_register: @@ -4848,8 +4953,7 @@ int stmmac_dvr_remove(struct device *dev) reset_control_assert(priv->plat->stmmac_rst); clk_disable_unprepare(priv->plat->pclk); clk_disable_unprepare(priv->plat->stmmac_clk); - if (priv->hw->pcs != STMMAC_PCS_RGMII && - priv->hw->pcs != STMMAC_PCS_TBI && + if (priv->hw->pcs != STMMAC_PCS_TBI && priv->hw->pcs != STMMAC_PCS_RTBI) stmmac_mdio_unregister(ndev); destroy_workqueue(priv->wq); diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_pci.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_pci.c index 8237dbc3e991..623521052152 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_pci.c +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_pci.c @@ -65,7 +65,6 @@ static void common_default_data(struct plat_stmmacenet_data *plat) plat->force_sf_dma_mode = 1; plat->mdio_bus_data->needs_reset = true; - plat->mdio_bus_data->phy_mask = 0; /* Set default value for multicast hash bins */ plat->multicast_filter_bins = HASH_TABLE_SIZE; @@ -154,8 +153,6 @@ static int intel_mgbe_common_data(struct pci_dev *pdev, plat->tx_queues_cfg[6].weight = 0x0F; plat->tx_queues_cfg[7].weight = 0x10; - plat->mdio_bus_data->phy_mask = 0; - plat->dma_cfg->pbl = 32; plat->dma_cfg->pblx8 = true; plat->dma_cfg->fixed_burst = 0; @@ -386,8 +383,6 @@ static int snps_gmac5_default_data(struct pci_dev *pdev, plat->tso_en = 1; plat->pmt = 1; - plat->mdio_bus_data->phy_mask = 0; - /* Set default value for multicast hash bins */ plat->multicast_filter_bins = HASH_TABLE_SIZE; @@ -406,6 +401,8 @@ static int snps_gmac5_default_data(struct pci_dev *pdev, plat->tx_queues_cfg[i].use_prio = false; plat->tx_queues_cfg[i].mode_to_use = MTL_QUEUE_DCB; plat->tx_queues_cfg[i].weight = 25; + if (i > 0) + plat->tx_queues_cfg[i].tbs_en = 1; } plat->rx_sched_algorithm = MTL_RX_ALGORITHM_SP; diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_platform.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_platform.c index cc8d7e7bf9ac..4775f49d7f3b 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_platform.c +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_platform.c @@ -320,7 +320,7 @@ out: static int stmmac_dt_phy(struct plat_stmmacenet_data *plat, struct device_node *np, struct device *dev) { - bool mdio = false; + bool mdio = !of_phy_is_fixed_link(np); static const struct of_device_id need_mdio_ids[] = { { .compatible = "snps,dwc-qos-ethernet-4.10" }, {}, diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c index 13227909287c..2aba2673d6c3 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c @@ -14,6 +14,7 @@ #include <linux/phy.h> #include <linux/udp.h> #include <net/pkt_cls.h> +#include <net/pkt_sched.h> #include <net/tcp.h> #include <net/udp.h> #include <net/tc_act/tc_gact.h> @@ -50,6 +51,7 @@ struct stmmac_packet_attrs { u8 id; int sarc; u16 queue_mapping; + u64 timestamp; }; static u8 stmmac_test_next_id; @@ -80,7 +82,7 @@ static struct sk_buff *stmmac_test_get_udp_skb(struct stmmac_priv *priv, if (attr->max_size && (attr->max_size > size)) size = attr->max_size; - skb = netdev_alloc_skb_ip_align(priv->dev, size); + skb = netdev_alloc_skb(priv->dev, size); if (!skb) return NULL; @@ -208,6 +210,9 @@ static struct sk_buff *stmmac_test_get_udp_skb(struct stmmac_priv *priv, skb->pkt_type = PACKET_HOST; skb->dev = priv->dev; + if (attr->timestamp) + skb->tstamp = ns_to_ktime(attr->timestamp); + return skb; } @@ -244,6 +249,8 @@ static int stmmac_test_loopback_validate(struct sk_buff *skb, struct net_device *orig_ndev) { struct stmmac_test_priv *tpriv = pt->af_packet_priv; + unsigned char *src = tpriv->packet->src; + unsigned char *dst = tpriv->packet->dst; struct stmmachdr *shdr; struct ethhdr *ehdr; struct udphdr *uhdr; @@ -260,15 +267,15 @@ static int stmmac_test_loopback_validate(struct sk_buff *skb, goto out; ehdr = (struct ethhdr *)skb_mac_header(skb); - if (tpriv->packet->dst) { - if (!ether_addr_equal(ehdr->h_dest, tpriv->packet->dst)) + if (dst) { + if (!ether_addr_equal_unaligned(ehdr->h_dest, dst)) goto out; } if (tpriv->packet->sarc) { - if (!ether_addr_equal(ehdr->h_source, ehdr->h_dest)) + if (!ether_addr_equal_unaligned(ehdr->h_source, ehdr->h_dest)) goto out; - } else if (tpriv->packet->src) { - if (!ether_addr_equal(ehdr->h_source, tpriv->packet->src)) + } else if (src) { + if (!ether_addr_equal_unaligned(ehdr->h_source, src)) goto out; } @@ -339,8 +346,7 @@ static int __stmmac_test_loopback(struct stmmac_priv *priv, goto cleanup; } - skb_set_queue_mapping(skb, attr->queue_mapping); - ret = dev_queue_xmit(skb); + ret = dev_direct_xmit(skb, attr->queue_mapping); if (ret) goto cleanup; @@ -714,7 +720,7 @@ static int stmmac_test_flowctrl_validate(struct sk_buff *skb, struct ethhdr *ehdr; ehdr = (struct ethhdr *)skb_mac_header(skb); - if (!ether_addr_equal(ehdr->h_source, orig_ndev->dev_addr)) + if (!ether_addr_equal_unaligned(ehdr->h_source, orig_ndev->dev_addr)) goto out; if (ehdr->h_proto != htons(ETH_P_PAUSE)) goto out; @@ -851,12 +857,16 @@ static int stmmac_test_vlan_validate(struct sk_buff *skb, if (tpriv->vlan_id) { if (skb->vlan_proto != htons(proto)) goto out; - if (skb->vlan_tci != tpriv->vlan_id) + if (skb->vlan_tci != tpriv->vlan_id) { + /* Means filter did not work. */ + tpriv->ok = false; + complete(&tpriv->comp); goto out; + } } ehdr = (struct ethhdr *)skb_mac_header(skb); - if (!ether_addr_equal(ehdr->h_dest, tpriv->packet->dst)) + if (!ether_addr_equal_unaligned(ehdr->h_dest, tpriv->packet->dst)) goto out; ihdr = ip_hdr(skb); @@ -926,8 +936,7 @@ static int __stmmac_test_vlanfilt(struct stmmac_priv *priv) goto vlan_del; } - skb_set_queue_mapping(skb, 0); - ret = dev_queue_xmit(skb); + ret = dev_direct_xmit(skb, 0); if (ret) goto vlan_del; @@ -965,6 +974,9 @@ static int stmmac_test_vlanfilt_perfect(struct stmmac_priv *priv) { int ret, prev_cap = priv->dma_cap.vlhash; + if (!(priv->dev->features & NETIF_F_HW_VLAN_CTAG_FILTER)) + return -EOPNOTSUPP; + priv->dma_cap.vlhash = 0; ret = __stmmac_test_vlanfilt(priv); priv->dma_cap.vlhash = prev_cap; @@ -1018,8 +1030,7 @@ static int __stmmac_test_dvlanfilt(struct stmmac_priv *priv) goto vlan_del; } - skb_set_queue_mapping(skb, 0); - ret = dev_queue_xmit(skb); + ret = dev_direct_xmit(skb, 0); if (ret) goto vlan_del; @@ -1057,6 +1068,9 @@ static int stmmac_test_dvlanfilt_perfect(struct stmmac_priv *priv) { int ret, prev_cap = priv->dma_cap.vlhash; + if (!(priv->dev->features & NETIF_F_HW_VLAN_STAG_FILTER)) + return -EOPNOTSUPP; + priv->dma_cap.vlhash = 0; ret = __stmmac_test_dvlanfilt(priv); priv->dma_cap.vlhash = prev_cap; @@ -1286,8 +1300,7 @@ static int stmmac_test_vlanoff_common(struct stmmac_priv *priv, bool svlan) __vlan_hwaccel_put_tag(skb, htons(proto), tpriv->vlan_id); skb->protocol = htons(proto); - skb_set_queue_mapping(skb, 0); - ret = dev_queue_xmit(skb); + ret = dev_direct_xmit(skb, 0); if (ret) goto vlan_del; @@ -1323,16 +1336,19 @@ static int __stmmac_test_l3filt(struct stmmac_priv *priv, u32 dst, u32 src, struct stmmac_packet_attrs attr = { }; struct flow_dissector *dissector; struct flow_cls_offload *cls; + int ret, old_enable = 0; struct flow_rule *rule; - int ret; if (!tc_can_offload(priv->dev)) return -EOPNOTSUPP; if (!priv->dma_cap.l3l4fnum) return -EOPNOTSUPP; - if (priv->rss.enable) + if (priv->rss.enable) { + old_enable = priv->rss.enable; + priv->rss.enable = false; stmmac_rss_configure(priv, priv->hw, NULL, priv->plat->rx_queues_to_use); + } dissector = kzalloc(sizeof(*dissector), GFP_KERNEL); if (!dissector) { @@ -1399,7 +1415,8 @@ cleanup_cls: cleanup_dissector: kfree(dissector); cleanup_rss: - if (priv->rss.enable) { + if (old_enable) { + priv->rss.enable = old_enable; stmmac_rss_configure(priv, priv->hw, &priv->rss, priv->plat->rx_queues_to_use); } @@ -1444,16 +1461,19 @@ static int __stmmac_test_l4filt(struct stmmac_priv *priv, u32 dst, u32 src, struct stmmac_packet_attrs attr = { }; struct flow_dissector *dissector; struct flow_cls_offload *cls; + int ret, old_enable = 0; struct flow_rule *rule; - int ret; if (!tc_can_offload(priv->dev)) return -EOPNOTSUPP; if (!priv->dma_cap.l3l4fnum) return -EOPNOTSUPP; - if (priv->rss.enable) + if (priv->rss.enable) { + old_enable = priv->rss.enable; + priv->rss.enable = false; stmmac_rss_configure(priv, priv->hw, NULL, priv->plat->rx_queues_to_use); + } dissector = kzalloc(sizeof(*dissector), GFP_KERNEL); if (!dissector) { @@ -1525,7 +1545,8 @@ cleanup_cls: cleanup_dissector: kfree(dissector); cleanup_rss: - if (priv->rss.enable) { + if (old_enable) { + priv->rss.enable = old_enable; stmmac_rss_configure(priv, priv->hw, &priv->rss, priv->plat->rx_queues_to_use); } @@ -1578,7 +1599,7 @@ static int stmmac_test_arp_validate(struct sk_buff *skb, struct arphdr *ahdr; ehdr = (struct ethhdr *)skb_mac_header(skb); - if (!ether_addr_equal(ehdr->h_dest, tpriv->packet->src)) + if (!ether_addr_equal_unaligned(ehdr->h_dest, tpriv->packet->src)) goto out; ahdr = arp_hdr(skb); @@ -1639,8 +1660,7 @@ static int stmmac_test_arpoffload(struct stmmac_priv *priv) if (ret) goto cleanup; - skb_set_queue_mapping(skb, 0); - ret = dev_queue_xmit(skb); + ret = dev_direct_xmit(skb, 0); if (ret) goto cleanup_promisc; @@ -1728,6 +1748,68 @@ static int stmmac_test_sph(struct stmmac_priv *priv) return 0; } +static int stmmac_test_tbs(struct stmmac_priv *priv) +{ +#define STMMAC_TBS_LT_OFFSET (500 * 1000 * 1000) /* 500 ms*/ + struct stmmac_packet_attrs attr = { }; + struct tc_etf_qopt_offload qopt; + u64 start_time, curr_time = 0; + unsigned long flags; + int ret, i; + + if (!priv->hwts_tx_en) + return -EOPNOTSUPP; + + /* Find first TBS enabled Queue, if any */ + for (i = 0; i < priv->plat->tx_queues_to_use; i++) + if (priv->tx_queue[i].tbs & STMMAC_TBS_AVAIL) + break; + + if (i >= priv->plat->tx_queues_to_use) + return -EOPNOTSUPP; + + qopt.enable = true; + qopt.queue = i; + + ret = stmmac_tc_setup_etf(priv, priv, &qopt); + if (ret) + return ret; + + spin_lock_irqsave(&priv->ptp_lock, flags); + stmmac_get_systime(priv, priv->ptpaddr, &curr_time); + spin_unlock_irqrestore(&priv->ptp_lock, flags); + + if (!curr_time) { + ret = -EOPNOTSUPP; + goto fail_disable; + } + + start_time = curr_time; + curr_time += STMMAC_TBS_LT_OFFSET; + + attr.dst = priv->dev->dev_addr; + attr.timestamp = curr_time; + attr.timeout = nsecs_to_jiffies(2 * STMMAC_TBS_LT_OFFSET); + attr.queue_mapping = i; + + ret = __stmmac_test_loopback(priv, &attr); + if (ret) + goto fail_disable; + + /* Check if expected time has elapsed */ + spin_lock_irqsave(&priv->ptp_lock, flags); + stmmac_get_systime(priv, priv->ptpaddr, &curr_time); + spin_unlock_irqrestore(&priv->ptp_lock, flags); + + if ((curr_time - start_time) < STMMAC_TBS_LT_OFFSET) + ret = -EINVAL; + +fail_disable: + qopt.enable = false; + stmmac_tc_setup_etf(priv, priv, &qopt); + return ret; +} + #define STMMAC_LOOPBACK_NONE 0 #define STMMAC_LOOPBACK_MAC 1 #define STMMAC_LOOPBACK_PHY 2 @@ -1861,6 +1943,10 @@ static const struct stmmac_test { .name = "Split Header ", .lb = STMMAC_LOOPBACK_PHY, .fn = stmmac_test_sph, + }, { + .name = "TBS (ETF Scheduler) ", + .lb = STMMAC_LOOPBACK_PHY, + .fn = stmmac_test_tbs, }, }; @@ -1869,7 +1955,6 @@ void stmmac_selftest_run(struct net_device *dev, { struct stmmac_priv *priv = netdev_priv(dev); int count = stmmac_selftest_get_count(priv); - int carrier = netif_carrier_ok(dev); int i, ret; memset(buf, 0, sizeof(*buf) * count); @@ -1879,15 +1964,12 @@ void stmmac_selftest_run(struct net_device *dev, netdev_err(priv->dev, "Only offline tests are supported\n"); etest->flags |= ETH_TEST_FL_FAILED; return; - } else if (!carrier) { + } else if (!netif_carrier_ok(dev)) { netdev_err(priv->dev, "You need valid Link to execute tests\n"); etest->flags |= ETH_TEST_FL_FAILED; return; } - /* We don't want extra traffic */ - netif_carrier_off(dev); - /* Wait for queues drain */ msleep(200); @@ -1942,10 +2024,6 @@ void stmmac_selftest_run(struct net_device *dev, break; } } - - /* Restart everything */ - if (carrier) - netif_carrier_on(dev); } void stmmac_selftest_get_strings(struct stmmac_priv *priv, u8 *data) diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_tc.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_tc.c index 6c4686b77516..7a01dee2f9a8 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_tc.c +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_tc.c @@ -577,6 +577,10 @@ static int tc_setup_cls(struct stmmac_priv *priv, { int ret = 0; + /* When RSS is enabled, the filtering will be bypassed */ + if (priv->rss.enable) + return -EBUSY; + switch (cls->command) { case FLOW_CLS_REPLACE: ret = tc_add_flow(priv, cls); @@ -727,10 +731,31 @@ disable: return ret; } +static int tc_setup_etf(struct stmmac_priv *priv, + struct tc_etf_qopt_offload *qopt) +{ + if (!priv->dma_cap.tbssel) + return -EOPNOTSUPP; + if (qopt->queue >= priv->plat->tx_queues_to_use) + return -EINVAL; + if (!(priv->tx_queue[qopt->queue].tbs & STMMAC_TBS_AVAIL)) + return -EINVAL; + + if (qopt->enable) + priv->tx_queue[qopt->queue].tbs |= STMMAC_TBS_EN; + else + priv->tx_queue[qopt->queue].tbs &= ~STMMAC_TBS_EN; + + netdev_info(priv->dev, "%s ETF for Queue %d\n", + qopt->enable ? "enabled" : "disabled", qopt->queue); + return 0; +} + const struct stmmac_tc_ops dwmac510_tc_ops = { .init = tc_init, .setup_cls_u32 = tc_setup_cls_u32, .setup_cbs = tc_setup_cbs, .setup_cls = tc_setup_cls, .setup_taprio = tc_setup_taprio, + .setup_etf = tc_setup_etf, }; diff --git a/drivers/net/ethernet/sun/sunvnet_common.c b/drivers/net/ethernet/sun/sunvnet_common.c index a601a306f9a5..c23ce838ff63 100644 --- a/drivers/net/ethernet/sun/sunvnet_common.c +++ b/drivers/net/ethernet/sun/sunvnet_common.c @@ -1223,7 +1223,7 @@ vnet_handle_offloads(struct vnet_port *port, struct sk_buff *skb, { struct net_device *dev = VNET_PORT_TO_NET_DEVICE(port); struct vio_dring_state *dr = &port->vio.drings[VIO_DRIVER_TX_RING]; - struct sk_buff *segs; + struct sk_buff *segs, *curr, *next; int maclen, datalen; int status; int gso_size, gso_type, gso_segs; @@ -1282,11 +1282,8 @@ vnet_handle_offloads(struct vnet_port *port, struct sk_buff *skb, skb_reset_mac_header(skb); status = 0; - while (segs) { - struct sk_buff *curr = segs; - - segs = segs->next; - curr->next = NULL; + skb_list_walk_safe(segs, curr, next) { + skb_mark_not_on_list(curr); if (port->tso && curr->len > dev->mtu) { skb_shinfo(curr)->gso_size = gso_size; skb_shinfo(curr)->gso_type = gso_type; diff --git a/drivers/net/ethernet/xscale/Kconfig b/drivers/net/ethernet/xscale/Kconfig index cd0a8f46e7c6..98aa7b8ddb06 100644 --- a/drivers/net/ethernet/xscale/Kconfig +++ b/drivers/net/ethernet/xscale/Kconfig @@ -27,4 +27,18 @@ config IXP4XX_ETH Say Y here if you want to use built-in Ethernet ports on IXP4xx processor. +config PTP_1588_CLOCK_IXP46X + tristate "Intel IXP46x as PTP clock" + depends on IXP4XX_ETH + depends on PTP_1588_CLOCK + default y + help + This driver adds support for using the IXP46X as a PTP + clock. This clock is only useful if your PTP programs are + getting hardware time stamps on the PTP Ethernet packets + using the SO_TIMESTAMPING API. + + To compile this driver as a module, choose M here: the module + will be called ptp_ixp46x. + endif # NET_VENDOR_XSCALE diff --git a/drivers/net/ethernet/xscale/Makefile b/drivers/net/ethernet/xscale/Makefile index 794a519d07b3..607f91b1e878 100644 --- a/drivers/net/ethernet/xscale/Makefile +++ b/drivers/net/ethernet/xscale/Makefile @@ -3,4 +3,5 @@ # Makefile for the Intel XScale IXP device drivers. # -obj-$(CONFIG_IXP4XX_ETH) += ixp4xx_eth.o +obj-$(CONFIG_IXP4XX_ETH) += ixp4xx_eth.o +obj-$(CONFIG_PTP_1588_CLOCK_IXP46X) += ptp_ixp46x.o diff --git a/drivers/net/ethernet/xscale/ixp46x_ts.h b/drivers/net/ethernet/xscale/ixp46x_ts.h new file mode 100644 index 000000000000..d792130e27b0 --- /dev/null +++ b/drivers/net/ethernet/xscale/ixp46x_ts.h @@ -0,0 +1,68 @@ +/* SPDX-License-Identifier: GPL-2.0-or-later */ +/* + * PTP 1588 clock using the IXP46X + * + * Copyright (C) 2010 OMICRON electronics GmbH + */ + +#ifndef _IXP46X_TS_H_ +#define _IXP46X_TS_H_ + +#define DEFAULT_ADDEND 0xF0000029 +#define TICKS_NS_SHIFT 4 + +struct ixp46x_channel_ctl { + u32 ch_control; /* 0x40 Time Synchronization Channel Control */ + u32 ch_event; /* 0x44 Time Synchronization Channel Event */ + u32 tx_snap_lo; /* 0x48 Transmit Snapshot Low Register */ + u32 tx_snap_hi; /* 0x4C Transmit Snapshot High Register */ + u32 rx_snap_lo; /* 0x50 Receive Snapshot Low Register */ + u32 rx_snap_hi; /* 0x54 Receive Snapshot High Register */ + u32 src_uuid_lo; /* 0x58 Source UUID0 Low Register */ + u32 src_uuid_hi; /* 0x5C Sequence Identifier/Source UUID0 High */ +}; + +struct ixp46x_ts_regs { + u32 control; /* 0x00 Time Sync Control Register */ + u32 event; /* 0x04 Time Sync Event Register */ + u32 addend; /* 0x08 Time Sync Addend Register */ + u32 accum; /* 0x0C Time Sync Accumulator Register */ + u32 test; /* 0x10 Time Sync Test Register */ + u32 unused; /* 0x14 */ + u32 rsystime_lo; /* 0x18 RawSystemTime_Low Register */ + u32 rsystime_hi; /* 0x1C RawSystemTime_High Register */ + u32 systime_lo; /* 0x20 SystemTime_Low Register */ + u32 systime_hi; /* 0x24 SystemTime_High Register */ + u32 trgt_lo; /* 0x28 TargetTime_Low Register */ + u32 trgt_hi; /* 0x2C TargetTime_High Register */ + u32 asms_lo; /* 0x30 Auxiliary Slave Mode Snapshot Low */ + u32 asms_hi; /* 0x34 Auxiliary Slave Mode Snapshot High */ + u32 amms_lo; /* 0x38 Auxiliary Master Mode Snapshot Low */ + u32 amms_hi; /* 0x3C Auxiliary Master Mode Snapshot High */ + + struct ixp46x_channel_ctl channel[3]; +}; + +/* 0x00 Time Sync Control Register Bits */ +#define TSCR_AMM (1<<3) +#define TSCR_ASM (1<<2) +#define TSCR_TTM (1<<1) +#define TSCR_RST (1<<0) + +/* 0x04 Time Sync Event Register Bits */ +#define TSER_SNM (1<<3) +#define TSER_SNS (1<<2) +#define TTIPEND (1<<1) + +/* 0x40 Time Synchronization Channel Control Register Bits */ +#define MASTER_MODE (1<<0) +#define TIMESTAMP_ALL (1<<1) + +/* 0x44 Time Synchronization Channel Event Register Bits */ +#define TX_SNAPSHOT_LOCKED (1<<0) +#define RX_SNAPSHOT_LOCKED (1<<1) + +/* The ptp_ixp46x module will set this variable */ +extern int ixp46x_phc_index; + +#endif diff --git a/drivers/net/ethernet/xscale/ixp4xx_eth.c b/drivers/net/ethernet/xscale/ixp4xx_eth.c index 6fc04ffb22c2..269596c15133 100644 --- a/drivers/net/ethernet/xscale/ixp4xx_eth.c +++ b/drivers/net/ethernet/xscale/ixp4xx_eth.c @@ -29,14 +29,16 @@ #include <linux/net_tstamp.h> #include <linux/of.h> #include <linux/phy.h> +#include <linux/platform_data/eth_ixp4xx.h> #include <linux/platform_device.h> #include <linux/ptp_classify.h> #include <linux/slab.h> #include <linux/module.h> -#include <mach/ixp46x_ts.h> #include <linux/soc/ixp4xx/npe.h> #include <linux/soc/ixp4xx/qmgr.h> +#include "ixp46x_ts.h" + #define DEBUG_DESC 0 #define DEBUG_RX 0 #define DEBUG_TX 0 @@ -517,25 +519,14 @@ static int ixp4xx_mdio_write(struct mii_bus *bus, int phy_id, int location, return ret; } -static int ixp4xx_mdio_register(void) +static int ixp4xx_mdio_register(struct eth_regs __iomem *regs) { int err; if (!(mdio_bus = mdiobus_alloc())) return -ENOMEM; - if (cpu_is_ixp43x()) { - /* IXP43x lacks NPE-B and uses NPE-C for MII PHY access */ - if (!(ixp4xx_read_feature_bits() & IXP4XX_FEATURE_NPEC_ETH)) - return -ENODEV; - mdio_regs = (struct eth_regs __iomem *)IXP4XX_EthC_BASE_VIRT; - } else { - /* All MII PHY accesses use NPE-B Ethernet registers */ - if (!(ixp4xx_read_feature_bits() & IXP4XX_FEATURE_NPEB_ETH0)) - return -ENODEV; - mdio_regs = (struct eth_regs __iomem *)IXP4XX_EthB_BASE_VIRT; - } - + mdio_regs = regs; __raw_writel(DEFAULT_CORE_CNTRL, &mdio_regs->core_control); spin_lock_init(&mdio_lock); mdio_bus->name = "IXP4xx MII Bus"; @@ -581,8 +572,8 @@ static void ixp4xx_adjust_link(struct net_device *dev) __raw_writel(DEFAULT_TX_CNTRL0 | TX_CNTRL0_HALFDUPLEX, &port->regs->tx_control[0]); - printk(KERN_INFO "%s: link up, speed %u Mb/s, %s duplex\n", - dev->name, port->speed, port->duplex ? "full" : "half"); + netdev_info(dev, "%s: link up, speed %u Mb/s, %s duplex\n", + dev->name, port->speed, port->duplex ? "full" : "half"); } @@ -592,7 +583,7 @@ static inline void debug_pkt(struct net_device *dev, const char *func, #if DEBUG_PKT_BYTES int i; - printk(KERN_DEBUG "%s: %s(%i) ", dev->name, func, len); + netdev_debug(dev, "%s(%i) ", func, len); for (i = 0; i < len; i++) { if (i >= DEBUG_PKT_BYTES) break; @@ -683,7 +674,7 @@ static int eth_poll(struct napi_struct *napi, int budget) int received = 0; #if DEBUG_RX - printk(KERN_DEBUG "%s: eth_poll\n", dev->name); + netdev_debug(dev, "eth_poll\n"); #endif while (received < budget) { @@ -697,23 +688,20 @@ static int eth_poll(struct napi_struct *napi, int budget) if ((n = queue_get_desc(rxq, port, 0)) < 0) { #if DEBUG_RX - printk(KERN_DEBUG "%s: eth_poll napi_complete\n", - dev->name); + netdev_debug(dev, "eth_poll napi_complete\n"); #endif napi_complete(napi); qmgr_enable_irq(rxq); if (!qmgr_stat_below_low_watermark(rxq) && napi_reschedule(napi)) { /* not empty again */ #if DEBUG_RX - printk(KERN_DEBUG "%s: eth_poll napi_reschedule succeeded\n", - dev->name); + netdev_debug(dev, "eth_poll napi_reschedule succeeded\n"); #endif qmgr_disable_irq(rxq); continue; } #if DEBUG_RX - printk(KERN_DEBUG "%s: eth_poll all done\n", - dev->name); + netdev_debug(dev, "eth_poll all done\n"); #endif return received; /* all work done */ } @@ -778,7 +766,7 @@ static int eth_poll(struct napi_struct *napi, int budget) } #if DEBUG_RX - printk(KERN_DEBUG "eth_poll(): end, not all work done\n"); + netdev_debug(dev, "eth_poll(): end, not all work done\n"); #endif return received; /* not all work done */ } @@ -842,7 +830,7 @@ static int eth_xmit(struct sk_buff *skb, struct net_device *dev) struct desc *desc; #if DEBUG_TX - printk(KERN_DEBUG "%s: eth_xmit\n", dev->name); + netdev_debug(dev, "eth_xmit\n"); #endif if (unlikely(skb->len > MAX_MRU)) { @@ -897,22 +885,21 @@ static int eth_xmit(struct sk_buff *skb, struct net_device *dev) if (qmgr_stat_below_low_watermark(txreadyq)) { /* empty */ #if DEBUG_TX - printk(KERN_DEBUG "%s: eth_xmit queue full\n", dev->name); + netdev_debug(dev, "eth_xmit queue full\n"); #endif netif_stop_queue(dev); /* we could miss TX ready interrupt */ /* really empty in fact */ if (!qmgr_stat_below_low_watermark(txreadyq)) { #if DEBUG_TX - printk(KERN_DEBUG "%s: eth_xmit ready again\n", - dev->name); + netdev_debug(dev, "eth_xmit ready again\n"); #endif netif_wake_queue(dev); } } #if DEBUG_TX - printk(KERN_DEBUG "%s: eth_xmit end\n", dev->name); + netdev_debug(dev, "eth_xmit end\n"); #endif ixp_tx_timestamp(port, skb); @@ -1099,7 +1086,7 @@ static int init_queues(struct port *port) int i; if (!ports_open) { - dma_pool = dma_pool_create(DRV_NAME, &port->netdev->dev, + dma_pool = dma_pool_create(DRV_NAME, port->netdev->dev.parent, POOL_ALLOC_SIZE, 32, 0); if (!dma_pool) return -ENOMEM; @@ -1186,8 +1173,7 @@ static int eth_open(struct net_device *dev) return err; if (npe_recv_message(npe, &msg, "ETH_GET_STATUS")) { - printk(KERN_ERR "%s: %s not responding\n", dev->name, - npe_name(npe)); + netdev_err(dev, "%s not responding\n", npe_name(npe)); return -EIO; } port->firmware[0] = msg.byte4; @@ -1299,7 +1285,7 @@ static int eth_close(struct net_device *dev) msg.eth_id = port->id; msg.byte3 = 1; if (npe_send_recv_message(port->npe, &msg, "ETH_ENABLE_LOOPBACK")) - printk(KERN_CRIT "%s: unable to enable loopback\n", dev->name); + netdev_crit(dev, "unable to enable loopback\n"); i = 0; do { /* drain RX buffers */ @@ -1323,11 +1309,11 @@ static int eth_close(struct net_device *dev) } while (++i < MAX_CLOSE_WAIT); if (buffs) - printk(KERN_CRIT "%s: unable to drain RX queue, %i buffer(s)" - " left in NPE\n", dev->name, buffs); + netdev_crit(dev, "unable to drain RX queue, %i buffer(s)" + " left in NPE\n", buffs); #if DEBUG_CLOSE if (!buffs) - printk(KERN_DEBUG "Draining RX queue took %i cycles\n", i); + netdev_debug(dev, "draining RX queue took %i cycles\n", i); #endif buffs = TX_DESCS; @@ -1343,17 +1329,16 @@ static int eth_close(struct net_device *dev) } while (++i < MAX_CLOSE_WAIT); if (buffs) - printk(KERN_CRIT "%s: unable to drain TX queue, %i buffer(s) " - "left in NPE\n", dev->name, buffs); + netdev_crit(dev, "unable to drain TX queue, %i buffer(s) " + "left in NPE\n", buffs); #if DEBUG_CLOSE if (!buffs) - printk(KERN_DEBUG "Draining TX queues took %i cycles\n", i); + netdev_debug(dev, "draining TX queues took %i cycles\n", i); #endif msg.byte3 = 0; if (npe_send_recv_message(port->npe, &msg, "ETH_DISABLE_LOOPBACK")) - printk(KERN_CRIT "%s: unable to disable loopback\n", - dev->name); + netdev_crit(dev, "unable to disable loopback\n"); phy_stop(dev->phydev); @@ -1374,54 +1359,88 @@ static const struct net_device_ops ixp4xx_netdev_ops = { .ndo_validate_addr = eth_validate_addr, }; -static int eth_init_one(struct platform_device *pdev) +static int ixp4xx_eth_probe(struct platform_device *pdev) { - struct port *port; - struct net_device *dev; - struct eth_plat_info *plat = dev_get_platdata(&pdev->dev); - struct phy_device *phydev = NULL; - u32 regs_phys; char phy_id[MII_BUS_ID_SIZE + 3]; + struct phy_device *phydev = NULL; + struct device *dev = &pdev->dev; + struct eth_plat_info *plat; + resource_size_t regs_phys; + struct net_device *ndev; + struct resource *res; + struct port *port; int err; - if (!(dev = alloc_etherdev(sizeof(struct port)))) + plat = dev_get_platdata(dev); + + if (!(ndev = devm_alloc_etherdev(dev, sizeof(struct port)))) return -ENOMEM; - SET_NETDEV_DEV(dev, &pdev->dev); - port = netdev_priv(dev); - port->netdev = dev; + SET_NETDEV_DEV(ndev, dev); + port = netdev_priv(ndev); + port->netdev = ndev; port->id = pdev->id; + /* Get the port resource and remap */ + res = platform_get_resource(pdev, IORESOURCE_MEM, 0); + if (!res) + return -ENODEV; + regs_phys = res->start; + port->regs = devm_ioremap_resource(dev, res); + switch (port->id) { case IXP4XX_ETH_NPEA: - port->regs = (struct eth_regs __iomem *)IXP4XX_EthA_BASE_VIRT; - regs_phys = IXP4XX_EthA_BASE_PHYS; + /* If the MDIO bus is not up yet, defer probe */ + if (!mdio_bus) + return -EPROBE_DEFER; break; case IXP4XX_ETH_NPEB: - port->regs = (struct eth_regs __iomem *)IXP4XX_EthB_BASE_VIRT; - regs_phys = IXP4XX_EthB_BASE_PHYS; + /* + * On all except IXP43x, NPE-B is used for the MDIO bus. + * If there is no NPE-B in the feature set, bail out, else + * register the MDIO bus. + */ + if (!cpu_is_ixp43x()) { + if (!(ixp4xx_read_feature_bits() & + IXP4XX_FEATURE_NPEB_ETH0)) + return -ENODEV; + /* Else register the MDIO bus on NPE-B */ + if ((err = ixp4xx_mdio_register(port->regs))) + return err; + } + if (!mdio_bus) + return -EPROBE_DEFER; break; case IXP4XX_ETH_NPEC: - port->regs = (struct eth_regs __iomem *)IXP4XX_EthC_BASE_VIRT; - regs_phys = IXP4XX_EthC_BASE_PHYS; + /* + * IXP43x lacks NPE-B and uses NPE-C for the MDIO bus access, + * of there is no NPE-C, no bus, nothing works, so bail out. + */ + if (cpu_is_ixp43x()) { + if (!(ixp4xx_read_feature_bits() & + IXP4XX_FEATURE_NPEC_ETH)) + return -ENODEV; + /* Else register the MDIO bus on NPE-C */ + if ((err = ixp4xx_mdio_register(port->regs))) + return err; + } + if (!mdio_bus) + return -EPROBE_DEFER; break; default: - err = -ENODEV; - goto err_free; + return -ENODEV; } - dev->netdev_ops = &ixp4xx_netdev_ops; - dev->ethtool_ops = &ixp4xx_ethtool_ops; - dev->tx_queue_len = 100; + ndev->netdev_ops = &ixp4xx_netdev_ops; + ndev->ethtool_ops = &ixp4xx_ethtool_ops; + ndev->tx_queue_len = 100; - netif_napi_add(dev, &port->napi, eth_poll, NAPI_WEIGHT); + netif_napi_add(ndev, &port->napi, eth_poll, NAPI_WEIGHT); - if (!(port->npe = npe_request(NPE_ID(port->id)))) { - err = -EIO; - goto err_free; - } + if (!(port->npe = npe_request(NPE_ID(port->id)))) + return -EIO; - port->mem_res = request_mem_region(regs_phys, REGS_SIZE, dev->name); + port->mem_res = request_mem_region(regs_phys, REGS_SIZE, ndev->name); if (!port->mem_res) { err = -EBUSY; goto err_npe_rel; @@ -1429,9 +1448,9 @@ static int eth_init_one(struct platform_device *pdev) port->plat = plat; npe_port_tab[NPE_ID(port->id)] = port; - memcpy(dev->dev_addr, plat->hwaddr, ETH_ALEN); + memcpy(ndev->dev_addr, plat->hwaddr, ETH_ALEN); - platform_set_drvdata(pdev, dev); + platform_set_drvdata(pdev, ndev); __raw_writel(DEFAULT_CORE_CNTRL | CORE_RESET, &port->regs->core_control); @@ -1441,7 +1460,7 @@ static int eth_init_one(struct platform_device *pdev) snprintf(phy_id, MII_BUS_ID_SIZE + 3, PHY_ID_FMT, mdio_bus->id, plat->phy); - phydev = phy_connect(dev, phy_id, &ixp4xx_adjust_link, + phydev = phy_connect(ndev, phy_id, &ixp4xx_adjust_link, PHY_INTERFACE_MODE_MII); if (IS_ERR(phydev)) { err = PTR_ERR(phydev); @@ -1450,11 +1469,11 @@ static int eth_init_one(struct platform_device *pdev) phydev->irq = PHY_POLL; - if ((err = register_netdev(dev))) + if ((err = register_netdev(ndev))) goto err_phy_dis; - printk(KERN_INFO "%s: MII PHY %i on %s\n", dev->name, plat->phy, - npe_name(port->npe)); + netdev_info(ndev, "%s: MII PHY %i on %s\n", ndev->name, plat->phy, + npe_name(port->npe)); return 0; @@ -1465,58 +1484,32 @@ err_free_mem: release_resource(port->mem_res); err_npe_rel: npe_release(port->npe); -err_free: - free_netdev(dev); return err; } -static int eth_remove_one(struct platform_device *pdev) +static int ixp4xx_eth_remove(struct platform_device *pdev) { - struct net_device *dev = platform_get_drvdata(pdev); - struct phy_device *phydev = dev->phydev; - struct port *port = netdev_priv(dev); + struct net_device *ndev = platform_get_drvdata(pdev); + struct phy_device *phydev = ndev->phydev; + struct port *port = netdev_priv(ndev); - unregister_netdev(dev); + unregister_netdev(ndev); phy_disconnect(phydev); + ixp4xx_mdio_remove(); npe_port_tab[NPE_ID(port->id)] = NULL; npe_release(port->npe); release_resource(port->mem_res); - free_netdev(dev); return 0; } static struct platform_driver ixp4xx_eth_driver = { .driver.name = DRV_NAME, - .probe = eth_init_one, - .remove = eth_remove_one, + .probe = ixp4xx_eth_probe, + .remove = ixp4xx_eth_remove, }; - -static int __init eth_init_module(void) -{ - int err; - - /* - * FIXME: we bail out on device tree boot but this really needs - * to be fixed in a nicer way: this registers the MDIO bus before - * even matching the driver infrastructure, we should only probe - * detected hardware. - */ - if (of_have_populated_dt()) - return -ENODEV; - if ((err = ixp4xx_mdio_register())) - return err; - return platform_driver_register(&ixp4xx_eth_driver); -} - -static void __exit eth_cleanup_module(void) -{ - platform_driver_unregister(&ixp4xx_eth_driver); - ixp4xx_mdio_remove(); -} +module_platform_driver(ixp4xx_eth_driver); MODULE_AUTHOR("Krzysztof Halasa"); MODULE_DESCRIPTION("Intel IXP4xx Ethernet driver"); MODULE_LICENSE("GPL v2"); MODULE_ALIAS("platform:ixp4xx_eth"); -module_init(eth_init_module); -module_exit(eth_cleanup_module); diff --git a/drivers/net/ethernet/xscale/ptp_ixp46x.c b/drivers/net/ethernet/xscale/ptp_ixp46x.c new file mode 100644 index 000000000000..9ecc395239e9 --- /dev/null +++ b/drivers/net/ethernet/xscale/ptp_ixp46x.c @@ -0,0 +1,329 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* + * PTP 1588 clock using the IXP46X + * + * Copyright (C) 2010 OMICRON electronics GmbH + */ +#include <linux/device.h> +#include <linux/err.h> +#include <linux/gpio.h> +#include <linux/init.h> +#include <linux/interrupt.h> +#include <linux/io.h> +#include <linux/irq.h> +#include <linux/kernel.h> +#include <linux/module.h> + +#include <linux/ptp_clock_kernel.h> + +#include "ixp46x_ts.h" + +#define DRIVER "ptp_ixp46x" +#define N_EXT_TS 2 +#define MASTER_GPIO 8 +#define MASTER_IRQ 25 +#define SLAVE_GPIO 7 +#define SLAVE_IRQ 24 + +struct ixp_clock { + struct ixp46x_ts_regs *regs; + struct ptp_clock *ptp_clock; + struct ptp_clock_info caps; + int exts0_enabled; + int exts1_enabled; +}; + +DEFINE_SPINLOCK(register_lock); + +/* + * Register access functions + */ + +static u64 ixp_systime_read(struct ixp46x_ts_regs *regs) +{ + u64 ns; + u32 lo, hi; + + lo = __raw_readl(®s->systime_lo); + hi = __raw_readl(®s->systime_hi); + + ns = ((u64) hi) << 32; + ns |= lo; + ns <<= TICKS_NS_SHIFT; + + return ns; +} + +static void ixp_systime_write(struct ixp46x_ts_regs *regs, u64 ns) +{ + u32 hi, lo; + + ns >>= TICKS_NS_SHIFT; + hi = ns >> 32; + lo = ns & 0xffffffff; + + __raw_writel(lo, ®s->systime_lo); + __raw_writel(hi, ®s->systime_hi); +} + +/* + * Interrupt service routine + */ + +static irqreturn_t isr(int irq, void *priv) +{ + struct ixp_clock *ixp_clock = priv; + struct ixp46x_ts_regs *regs = ixp_clock->regs; + struct ptp_clock_event event; + u32 ack = 0, lo, hi, val; + + val = __raw_readl(®s->event); + + if (val & TSER_SNS) { + ack |= TSER_SNS; + if (ixp_clock->exts0_enabled) { + hi = __raw_readl(®s->asms_hi); + lo = __raw_readl(®s->asms_lo); + event.type = PTP_CLOCK_EXTTS; + event.index = 0; + event.timestamp = ((u64) hi) << 32; + event.timestamp |= lo; + event.timestamp <<= TICKS_NS_SHIFT; + ptp_clock_event(ixp_clock->ptp_clock, &event); + } + } + + if (val & TSER_SNM) { + ack |= TSER_SNM; + if (ixp_clock->exts1_enabled) { + hi = __raw_readl(®s->amms_hi); + lo = __raw_readl(®s->amms_lo); + event.type = PTP_CLOCK_EXTTS; + event.index = 1; + event.timestamp = ((u64) hi) << 32; + event.timestamp |= lo; + event.timestamp <<= TICKS_NS_SHIFT; + ptp_clock_event(ixp_clock->ptp_clock, &event); + } + } + + if (val & TTIPEND) + ack |= TTIPEND; /* this bit seems to be always set */ + + if (ack) { + __raw_writel(ack, ®s->event); + return IRQ_HANDLED; + } else + return IRQ_NONE; +} + +/* + * PTP clock operations + */ + +static int ptp_ixp_adjfreq(struct ptp_clock_info *ptp, s32 ppb) +{ + u64 adj; + u32 diff, addend; + int neg_adj = 0; + struct ixp_clock *ixp_clock = container_of(ptp, struct ixp_clock, caps); + struct ixp46x_ts_regs *regs = ixp_clock->regs; + + if (ppb < 0) { + neg_adj = 1; + ppb = -ppb; + } + addend = DEFAULT_ADDEND; + adj = addend; + adj *= ppb; + diff = div_u64(adj, 1000000000ULL); + + addend = neg_adj ? addend - diff : addend + diff; + + __raw_writel(addend, ®s->addend); + + return 0; +} + +static int ptp_ixp_adjtime(struct ptp_clock_info *ptp, s64 delta) +{ + s64 now; + unsigned long flags; + struct ixp_clock *ixp_clock = container_of(ptp, struct ixp_clock, caps); + struct ixp46x_ts_regs *regs = ixp_clock->regs; + + spin_lock_irqsave(®ister_lock, flags); + + now = ixp_systime_read(regs); + now += delta; + ixp_systime_write(regs, now); + + spin_unlock_irqrestore(®ister_lock, flags); + + return 0; +} + +static int ptp_ixp_gettime(struct ptp_clock_info *ptp, struct timespec64 *ts) +{ + u64 ns; + unsigned long flags; + struct ixp_clock *ixp_clock = container_of(ptp, struct ixp_clock, caps); + struct ixp46x_ts_regs *regs = ixp_clock->regs; + + spin_lock_irqsave(®ister_lock, flags); + + ns = ixp_systime_read(regs); + + spin_unlock_irqrestore(®ister_lock, flags); + + *ts = ns_to_timespec64(ns); + return 0; +} + +static int ptp_ixp_settime(struct ptp_clock_info *ptp, + const struct timespec64 *ts) +{ + u64 ns; + unsigned long flags; + struct ixp_clock *ixp_clock = container_of(ptp, struct ixp_clock, caps); + struct ixp46x_ts_regs *regs = ixp_clock->regs; + + ns = timespec64_to_ns(ts); + + spin_lock_irqsave(®ister_lock, flags); + + ixp_systime_write(regs, ns); + + spin_unlock_irqrestore(®ister_lock, flags); + + return 0; +} + +static int ptp_ixp_enable(struct ptp_clock_info *ptp, + struct ptp_clock_request *rq, int on) +{ + struct ixp_clock *ixp_clock = container_of(ptp, struct ixp_clock, caps); + + switch (rq->type) { + case PTP_CLK_REQ_EXTTS: + switch (rq->extts.index) { + case 0: + ixp_clock->exts0_enabled = on ? 1 : 0; + break; + case 1: + ixp_clock->exts1_enabled = on ? 1 : 0; + break; + default: + return -EINVAL; + } + return 0; + default: + break; + } + + return -EOPNOTSUPP; +} + +static const struct ptp_clock_info ptp_ixp_caps = { + .owner = THIS_MODULE, + .name = "IXP46X timer", + .max_adj = 66666655, + .n_ext_ts = N_EXT_TS, + .n_pins = 0, + .pps = 0, + .adjfreq = ptp_ixp_adjfreq, + .adjtime = ptp_ixp_adjtime, + .gettime64 = ptp_ixp_gettime, + .settime64 = ptp_ixp_settime, + .enable = ptp_ixp_enable, +}; + +/* module operations */ + +static struct ixp_clock ixp_clock; + +static int setup_interrupt(int gpio) +{ + int irq; + int err; + + err = gpio_request(gpio, "ixp4-ptp"); + if (err) + return err; + + err = gpio_direction_input(gpio); + if (err) + return err; + + irq = gpio_to_irq(gpio); + if (irq < 0) + return irq; + + err = irq_set_irq_type(irq, IRQF_TRIGGER_FALLING); + if (err) { + pr_err("cannot set trigger type for irq %d\n", irq); + return err; + } + + err = request_irq(irq, isr, 0, DRIVER, &ixp_clock); + if (err) { + pr_err("request_irq failed for irq %d\n", irq); + return err; + } + + return irq; +} + +static void __exit ptp_ixp_exit(void) +{ + free_irq(MASTER_IRQ, &ixp_clock); + free_irq(SLAVE_IRQ, &ixp_clock); + ixp46x_phc_index = -1; + ptp_clock_unregister(ixp_clock.ptp_clock); +} + +static int __init ptp_ixp_init(void) +{ + if (!cpu_is_ixp46x()) + return -ENODEV; + + ixp_clock.regs = + (struct ixp46x_ts_regs __iomem *) IXP4XX_TIMESYNC_BASE_VIRT; + + ixp_clock.caps = ptp_ixp_caps; + + ixp_clock.ptp_clock = ptp_clock_register(&ixp_clock.caps, NULL); + + if (IS_ERR(ixp_clock.ptp_clock)) + return PTR_ERR(ixp_clock.ptp_clock); + + ixp46x_phc_index = ptp_clock_index(ixp_clock.ptp_clock); + + __raw_writel(DEFAULT_ADDEND, &ixp_clock.regs->addend); + __raw_writel(1, &ixp_clock.regs->trgt_lo); + __raw_writel(0, &ixp_clock.regs->trgt_hi); + __raw_writel(TTIPEND, &ixp_clock.regs->event); + + if (MASTER_IRQ != setup_interrupt(MASTER_GPIO)) { + pr_err("failed to setup gpio %d as irq\n", MASTER_GPIO); + goto no_master; + } + if (SLAVE_IRQ != setup_interrupt(SLAVE_GPIO)) { + pr_err("failed to setup gpio %d as irq\n", SLAVE_GPIO); + goto no_slave; + } + + return 0; +no_slave: + free_irq(MASTER_IRQ, &ixp_clock); +no_master: + ptp_clock_unregister(ixp_clock.ptp_clock); + return -ENODEV; +} + +module_init(ptp_ixp_init); +module_exit(ptp_ixp_exit); + +MODULE_AUTHOR("Richard Cochran <richardcochran@gmail.com>"); +MODULE_DESCRIPTION("PTP clock using the IXP46X timer"); +MODULE_LICENSE("GPL"); diff --git a/drivers/net/gtp.c b/drivers/net/gtp.c index dea3978b223c..bc85db8e6cdf 100644 --- a/drivers/net/gtp.c +++ b/drivers/net/gtp.c @@ -813,7 +813,7 @@ static struct sock *gtp_encap_enable_socket(int fd, int type, lock_sock(sock->sk); if (sock->sk->sk_user_data) { sk = ERR_PTR(-EBUSY); - goto out_sock; + goto out_rel_sock; } sk = sock->sk; @@ -826,8 +826,9 @@ static struct sock *gtp_encap_enable_socket(int fd, int type, setup_udp_tunnel_sock(sock_net(sock->sk), sock, &tuncfg); -out_sock: +out_rel_sock: release_sock(sock->sk); +out_sock: sockfd_put(sock); return sk; } diff --git a/drivers/net/hyperv/rndis_filter.c b/drivers/net/hyperv/rndis_filter.c index 857c4bea451c..e66d77dc28c8 100644 --- a/drivers/net/hyperv/rndis_filter.c +++ b/drivers/net/hyperv/rndis_filter.c @@ -1443,8 +1443,6 @@ void rndis_filter_device_remove(struct hv_device *dev, /* Halt and release the rndis device */ rndis_filter_halt_device(net_dev, rndis_dev); - net_dev->extension = NULL; - netvsc_device_remove(dev); } diff --git a/drivers/net/macsec.c b/drivers/net/macsec.c index afd8b2a08245..45bfd99f17fa 100644 --- a/drivers/net/macsec.c +++ b/drivers/net/macsec.c @@ -11,16 +11,17 @@ #include <linux/module.h> #include <crypto/aead.h> #include <linux/etherdevice.h> +#include <linux/netdevice.h> #include <linux/rtnetlink.h> #include <linux/refcount.h> #include <net/genetlink.h> #include <net/sock.h> #include <net/gro_cells.h> +#include <net/macsec.h> +#include <linux/phy.h> #include <uapi/linux/if_macsec.h> -typedef u64 __bitwise sci_t; - #define MACSEC_SCI_LEN 8 /* SecTAG length = macsec_eth_header without the optional SCI */ @@ -58,8 +59,6 @@ struct macsec_eth_header { #define GCM_AES_IV_LEN 12 #define DEFAULT_ICV_LEN 16 -#define MACSEC_NUM_AN 4 /* 2 bits for the association number */ - #define for_each_rxsc(secy, sc) \ for (sc = rcu_dereference_bh(secy->rx_sc); \ sc; \ @@ -77,49 +76,6 @@ struct gcm_iv { __be32 pn; }; -/** - * struct macsec_key - SA key - * @id: user-provided key identifier - * @tfm: crypto struct, key storage - */ -struct macsec_key { - u8 id[MACSEC_KEYID_LEN]; - struct crypto_aead *tfm; -}; - -struct macsec_rx_sc_stats { - __u64 InOctetsValidated; - __u64 InOctetsDecrypted; - __u64 InPktsUnchecked; - __u64 InPktsDelayed; - __u64 InPktsOK; - __u64 InPktsInvalid; - __u64 InPktsLate; - __u64 InPktsNotValid; - __u64 InPktsNotUsingSA; - __u64 InPktsUnusedSA; -}; - -struct macsec_rx_sa_stats { - __u32 InPktsOK; - __u32 InPktsInvalid; - __u32 InPktsNotValid; - __u32 InPktsNotUsingSA; - __u32 InPktsUnusedSA; -}; - -struct macsec_tx_sa_stats { - __u32 OutPktsProtected; - __u32 OutPktsEncrypted; -}; - -struct macsec_tx_sc_stats { - __u64 OutPktsProtected; - __u64 OutPktsEncrypted; - __u64 OutOctetsProtected; - __u64 OutOctetsEncrypted; -}; - struct macsec_dev_stats { __u64 OutPktsUntagged; __u64 InPktsUntagged; @@ -131,124 +87,8 @@ struct macsec_dev_stats { __u64 InPktsOverrun; }; -/** - * struct macsec_rx_sa - receive secure association - * @active: - * @next_pn: packet number expected for the next packet - * @lock: protects next_pn manipulations - * @key: key structure - * @stats: per-SA stats - */ -struct macsec_rx_sa { - struct macsec_key key; - spinlock_t lock; - u32 next_pn; - refcount_t refcnt; - bool active; - struct macsec_rx_sa_stats __percpu *stats; - struct macsec_rx_sc *sc; - struct rcu_head rcu; -}; - -struct pcpu_rx_sc_stats { - struct macsec_rx_sc_stats stats; - struct u64_stats_sync syncp; -}; - -/** - * struct macsec_rx_sc - receive secure channel - * @sci: secure channel identifier for this SC - * @active: channel is active - * @sa: array of secure associations - * @stats: per-SC stats - */ -struct macsec_rx_sc { - struct macsec_rx_sc __rcu *next; - sci_t sci; - bool active; - struct macsec_rx_sa __rcu *sa[MACSEC_NUM_AN]; - struct pcpu_rx_sc_stats __percpu *stats; - refcount_t refcnt; - struct rcu_head rcu_head; -}; - -/** - * struct macsec_tx_sa - transmit secure association - * @active: - * @next_pn: packet number to use for the next packet - * @lock: protects next_pn manipulations - * @key: key structure - * @stats: per-SA stats - */ -struct macsec_tx_sa { - struct macsec_key key; - spinlock_t lock; - u32 next_pn; - refcount_t refcnt; - bool active; - struct macsec_tx_sa_stats __percpu *stats; - struct rcu_head rcu; -}; - -struct pcpu_tx_sc_stats { - struct macsec_tx_sc_stats stats; - struct u64_stats_sync syncp; -}; - -/** - * struct macsec_tx_sc - transmit secure channel - * @active: - * @encoding_sa: association number of the SA currently in use - * @encrypt: encrypt packets on transmit, or authenticate only - * @send_sci: always include the SCI in the SecTAG - * @end_station: - * @scb: single copy broadcast flag - * @sa: array of secure associations - * @stats: stats for this TXSC - */ -struct macsec_tx_sc { - bool active; - u8 encoding_sa; - bool encrypt; - bool send_sci; - bool end_station; - bool scb; - struct macsec_tx_sa __rcu *sa[MACSEC_NUM_AN]; - struct pcpu_tx_sc_stats __percpu *stats; -}; - #define MACSEC_VALIDATE_DEFAULT MACSEC_VALIDATE_STRICT -/** - * struct macsec_secy - MACsec Security Entity - * @netdev: netdevice for this SecY - * @n_rx_sc: number of receive secure channels configured on this SecY - * @sci: secure channel identifier used for tx - * @key_len: length of keys used by the cipher suite - * @icv_len: length of ICV used by the cipher suite - * @validate_frames: validation mode - * @operational: MAC_Operational flag - * @protect_frames: enable protection for this SecY - * @replay_protect: enable packet number checks on receive - * @replay_window: size of the replay window - * @tx_sc: transmit secure channel - * @rx_sc: linked list of receive secure channels - */ -struct macsec_secy { - struct net_device *netdev; - unsigned int n_rx_sc; - sci_t sci; - u16 key_len; - u16 icv_len; - enum macsec_validation_type validate_frames; - bool operational; - bool protect_frames; - bool replay_protect; - u32 replay_window; - struct macsec_tx_sc tx_sc; - struct macsec_rx_sc __rcu *rx_sc; -}; - struct pcpu_secy_stats { struct macsec_dev_stats stats; struct u64_stats_sync syncp; @@ -260,6 +100,7 @@ struct pcpu_secy_stats { * @real_dev: pointer to underlying netdevice * @stats: MACsec device stats * @secys: linked list of SecY's on the underlying device + * @offload: status of offloading on the MACsec device */ struct macsec_dev { struct macsec_secy secy; @@ -267,6 +108,7 @@ struct macsec_dev { struct pcpu_secy_stats __percpu *stats; struct list_head secys; struct gro_cells gro_cells; + enum macsec_offload offload; }; /** @@ -480,6 +322,56 @@ static void macsec_set_shortlen(struct macsec_eth_header *h, size_t data_len) h->short_length = data_len; } +/* Checks if a MACsec interface is being offloaded to an hardware engine */ +static bool macsec_is_offloaded(struct macsec_dev *macsec) +{ + if (macsec->offload == MACSEC_OFFLOAD_PHY) + return true; + + return false; +} + +/* Checks if underlying layers implement MACsec offloading functions. */ +static bool macsec_check_offload(enum macsec_offload offload, + struct macsec_dev *macsec) +{ + if (!macsec || !macsec->real_dev) + return false; + + if (offload == MACSEC_OFFLOAD_PHY) + return macsec->real_dev->phydev && + macsec->real_dev->phydev->macsec_ops; + + return false; +} + +static const struct macsec_ops *__macsec_get_ops(enum macsec_offload offload, + struct macsec_dev *macsec, + struct macsec_context *ctx) +{ + if (ctx) { + memset(ctx, 0, sizeof(*ctx)); + ctx->offload = offload; + + if (offload == MACSEC_OFFLOAD_PHY) + ctx->phydev = macsec->real_dev->phydev; + } + + return macsec->real_dev->phydev->macsec_ops; +} + +/* Returns a pointer to the MACsec ops struct if any and updates the MACsec + * context device reference if provided. + */ +static const struct macsec_ops *macsec_get_ops(struct macsec_dev *macsec, + struct macsec_context *ctx) +{ + if (!macsec_check_offload(macsec->offload, macsec)) + return NULL; + + return __macsec_get_ops(macsec->offload, macsec, ctx); +} + /* validate MACsec packet according to IEEE 802.1AE-2006 9.12 */ static bool macsec_validate_skb(struct sk_buff *skb, u16 icv_len) { @@ -532,6 +424,23 @@ static struct macsec_eth_header *macsec_ethhdr(struct sk_buff *skb) return (struct macsec_eth_header *)skb_mac_header(skb); } +static void __macsec_pn_wrapped(struct macsec_secy *secy, + struct macsec_tx_sa *tx_sa) +{ + pr_debug("PN wrapped, transitioning to !oper\n"); + tx_sa->active = false; + if (secy->protect_frames) + secy->operational = false; +} + +void macsec_pn_wrapped(struct macsec_secy *secy, struct macsec_tx_sa *tx_sa) +{ + spin_lock_bh(&tx_sa->lock); + __macsec_pn_wrapped(secy, tx_sa); + spin_unlock_bh(&tx_sa->lock); +} +EXPORT_SYMBOL_GPL(macsec_pn_wrapped); + static u32 tx_sa_update_pn(struct macsec_tx_sa *tx_sa, struct macsec_secy *secy) { u32 pn; @@ -540,12 +449,8 @@ static u32 tx_sa_update_pn(struct macsec_tx_sa *tx_sa, struct macsec_secy *secy) pn = tx_sa->next_pn; tx_sa->next_pn++; - if (tx_sa->next_pn == 0) { - pr_debug("PN wrapped, transitioning to !oper\n"); - tx_sa->active = false; - if (secy->protect_frames) - secy->operational = false; - } + if (tx_sa->next_pn == 0) + __macsec_pn_wrapped(secy, tx_sa); spin_unlock_bh(&tx_sa->lock); return pn; @@ -1029,8 +934,10 @@ static struct macsec_rx_sc *find_rx_sc_rtnl(struct macsec_secy *secy, sci_t sci) return NULL; } -static void handle_not_macsec(struct sk_buff *skb) +static enum rx_handler_result handle_not_macsec(struct sk_buff *skb) { + /* Deliver to the uncontrolled port by default */ + enum rx_handler_result ret = RX_HANDLER_PASS; struct macsec_rxh_data *rxd; struct macsec_dev *macsec; @@ -1045,7 +952,8 @@ static void handle_not_macsec(struct sk_buff *skb) struct sk_buff *nskb; struct pcpu_secy_stats *secy_stats = this_cpu_ptr(macsec->stats); - if (macsec->secy.validate_frames == MACSEC_VALIDATE_STRICT) { + if (!macsec_is_offloaded(macsec) && + macsec->secy.validate_frames == MACSEC_VALIDATE_STRICT) { u64_stats_update_begin(&secy_stats->syncp); secy_stats->stats.InPktsNoTag++; u64_stats_update_end(&secy_stats->syncp); @@ -1064,9 +972,17 @@ static void handle_not_macsec(struct sk_buff *skb) secy_stats->stats.InPktsUntagged++; u64_stats_update_end(&secy_stats->syncp); } + + if (netif_running(macsec->secy.netdev) && + macsec_is_offloaded(macsec)) { + ret = RX_HANDLER_EXACT; + goto out; + } } +out: rcu_read_unlock(); + return ret; } static rx_handler_result_t macsec_handle_frame(struct sk_buff **pskb) @@ -1091,12 +1007,8 @@ static rx_handler_result_t macsec_handle_frame(struct sk_buff **pskb) goto drop_direct; hdr = macsec_ethhdr(skb); - if (hdr->eth.h_proto != htons(ETH_P_MACSEC)) { - handle_not_macsec(skb); - - /* and deliver to the uncontrolled port */ - return RX_HANDLER_PASS; - } + if (hdr->eth.h_proto != htons(ETH_P_MACSEC)) + return handle_not_macsec(skb); skb = skb_unshare(skb, GFP_ATOMIC); *pskb = skb; @@ -1585,6 +1497,7 @@ static const struct nla_policy macsec_genl_policy[NUM_MACSEC_ATTR] = { [MACSEC_ATTR_IFINDEX] = { .type = NLA_U32 }, [MACSEC_ATTR_RXSC_CONFIG] = { .type = NLA_NESTED }, [MACSEC_ATTR_SA_CONFIG] = { .type = NLA_NESTED }, + [MACSEC_ATTR_OFFLOAD] = { .type = NLA_NESTED }, }; static const struct nla_policy macsec_genl_rxsc_policy[NUM_MACSEC_RXSC_ATTR] = { @@ -1602,6 +1515,44 @@ static const struct nla_policy macsec_genl_sa_policy[NUM_MACSEC_SA_ATTR] = { .len = MACSEC_MAX_KEY_LEN, }, }; +static const struct nla_policy macsec_genl_offload_policy[NUM_MACSEC_OFFLOAD_ATTR] = { + [MACSEC_OFFLOAD_ATTR_TYPE] = { .type = NLA_U8 }, +}; + +/* Offloads an operation to a device driver */ +static int macsec_offload(int (* const func)(struct macsec_context *), + struct macsec_context *ctx) +{ + int ret; + + if (unlikely(!func)) + return 0; + + if (ctx->offload == MACSEC_OFFLOAD_PHY) + mutex_lock(&ctx->phydev->lock); + + /* Phase I: prepare. The drive should fail here if there are going to be + * issues in the commit phase. + */ + ctx->prepare = true; + ret = (*func)(ctx); + if (ret) + goto phy_unlock; + + /* Phase II: commit. This step cannot fail. */ + ctx->prepare = false; + ret = (*func)(ctx); + /* This should never happen: commit is not allowed to fail */ + if (unlikely(ret)) + WARN(1, "MACsec offloading commit failed (%d)\n", ret); + +phy_unlock: + if (ctx->offload == MACSEC_OFFLOAD_PHY) + mutex_unlock(&ctx->phydev->lock); + + return ret; +} + static int parse_sa_config(struct nlattr **attrs, struct nlattr **tb_sa) { if (!attrs[MACSEC_ATTR_SA_CONFIG]) @@ -1717,13 +1668,40 @@ static int macsec_add_rxsa(struct sk_buff *skb, struct genl_info *info) if (tb_sa[MACSEC_SA_ATTR_ACTIVE]) rx_sa->active = !!nla_get_u8(tb_sa[MACSEC_SA_ATTR_ACTIVE]); - nla_memcpy(rx_sa->key.id, tb_sa[MACSEC_SA_ATTR_KEYID], MACSEC_KEYID_LEN); rx_sa->sc = rx_sc; + + /* If h/w offloading is available, propagate to the device */ + if (macsec_is_offloaded(netdev_priv(dev))) { + const struct macsec_ops *ops; + struct macsec_context ctx; + + ops = macsec_get_ops(netdev_priv(dev), &ctx); + if (!ops) { + err = -EOPNOTSUPP; + goto cleanup; + } + + ctx.sa.assoc_num = assoc_num; + ctx.sa.rx_sa = rx_sa; + memcpy(ctx.sa.key, nla_data(tb_sa[MACSEC_SA_ATTR_KEY]), + MACSEC_KEYID_LEN); + + err = macsec_offload(ops->mdo_add_rxsa, &ctx); + if (err) + goto cleanup; + } + + nla_memcpy(rx_sa->key.id, tb_sa[MACSEC_SA_ATTR_KEYID], MACSEC_KEYID_LEN); rcu_assign_pointer(rx_sc->sa[assoc_num], rx_sa); rtnl_unlock(); return 0; + +cleanup: + kfree(rx_sa); + rtnl_unlock(); + return err; } static bool validate_add_rxsc(struct nlattr **attrs) @@ -1746,6 +1724,8 @@ static int macsec_add_rxsc(struct sk_buff *skb, struct genl_info *info) struct nlattr **attrs = info->attrs; struct macsec_rx_sc *rx_sc; struct nlattr *tb_rxsc[MACSEC_RXSC_ATTR_MAX + 1]; + bool was_active; + int ret; if (!attrs[MACSEC_ATTR_IFINDEX]) return -EINVAL; @@ -1771,12 +1751,35 @@ static int macsec_add_rxsc(struct sk_buff *skb, struct genl_info *info) return PTR_ERR(rx_sc); } + was_active = rx_sc->active; if (tb_rxsc[MACSEC_RXSC_ATTR_ACTIVE]) rx_sc->active = !!nla_get_u8(tb_rxsc[MACSEC_RXSC_ATTR_ACTIVE]); + if (macsec_is_offloaded(netdev_priv(dev))) { + const struct macsec_ops *ops; + struct macsec_context ctx; + + ops = macsec_get_ops(netdev_priv(dev), &ctx); + if (!ops) { + ret = -EOPNOTSUPP; + goto cleanup; + } + + ctx.rx_sc = rx_sc; + + ret = macsec_offload(ops->mdo_add_rxsc, &ctx); + if (ret) + goto cleanup; + } + rtnl_unlock(); return 0; + +cleanup: + rx_sc->active = was_active; + rtnl_unlock(); + return ret; } static bool validate_add_txsa(struct nlattr **attrs) @@ -1813,6 +1816,7 @@ static int macsec_add_txsa(struct sk_buff *skb, struct genl_info *info) struct macsec_tx_sa *tx_sa; unsigned char assoc_num; struct nlattr *tb_sa[MACSEC_SA_ATTR_MAX + 1]; + bool was_operational; int err; if (!attrs[MACSEC_ATTR_IFINDEX]) @@ -1863,8 +1867,6 @@ static int macsec_add_txsa(struct sk_buff *skb, struct genl_info *info) return err; } - nla_memcpy(tx_sa->key.id, tb_sa[MACSEC_SA_ATTR_KEYID], MACSEC_KEYID_LEN); - spin_lock_bh(&tx_sa->lock); tx_sa->next_pn = nla_get_u32(tb_sa[MACSEC_SA_ATTR_PN]); spin_unlock_bh(&tx_sa->lock); @@ -1872,14 +1874,43 @@ static int macsec_add_txsa(struct sk_buff *skb, struct genl_info *info) if (tb_sa[MACSEC_SA_ATTR_ACTIVE]) tx_sa->active = !!nla_get_u8(tb_sa[MACSEC_SA_ATTR_ACTIVE]); + was_operational = secy->operational; if (assoc_num == tx_sc->encoding_sa && tx_sa->active) secy->operational = true; + /* If h/w offloading is available, propagate to the device */ + if (macsec_is_offloaded(netdev_priv(dev))) { + const struct macsec_ops *ops; + struct macsec_context ctx; + + ops = macsec_get_ops(netdev_priv(dev), &ctx); + if (!ops) { + err = -EOPNOTSUPP; + goto cleanup; + } + + ctx.sa.assoc_num = assoc_num; + ctx.sa.tx_sa = tx_sa; + memcpy(ctx.sa.key, nla_data(tb_sa[MACSEC_SA_ATTR_KEY]), + MACSEC_KEYID_LEN); + + err = macsec_offload(ops->mdo_add_txsa, &ctx); + if (err) + goto cleanup; + } + + nla_memcpy(tx_sa->key.id, tb_sa[MACSEC_SA_ATTR_KEYID], MACSEC_KEYID_LEN); rcu_assign_pointer(tx_sc->sa[assoc_num], tx_sa); rtnl_unlock(); return 0; + +cleanup: + secy->operational = was_operational; + kfree(tx_sa); + rtnl_unlock(); + return err; } static int macsec_del_rxsa(struct sk_buff *skb, struct genl_info *info) @@ -1892,6 +1923,7 @@ static int macsec_del_rxsa(struct sk_buff *skb, struct genl_info *info) u8 assoc_num; struct nlattr *tb_rxsc[MACSEC_RXSC_ATTR_MAX + 1]; struct nlattr *tb_sa[MACSEC_SA_ATTR_MAX + 1]; + int ret; if (!attrs[MACSEC_ATTR_IFINDEX]) return -EINVAL; @@ -1915,12 +1947,35 @@ static int macsec_del_rxsa(struct sk_buff *skb, struct genl_info *info) return -EBUSY; } + /* If h/w offloading is available, propagate to the device */ + if (macsec_is_offloaded(netdev_priv(dev))) { + const struct macsec_ops *ops; + struct macsec_context ctx; + + ops = macsec_get_ops(netdev_priv(dev), &ctx); + if (!ops) { + ret = -EOPNOTSUPP; + goto cleanup; + } + + ctx.sa.assoc_num = assoc_num; + ctx.sa.rx_sa = rx_sa; + + ret = macsec_offload(ops->mdo_del_rxsa, &ctx); + if (ret) + goto cleanup; + } + RCU_INIT_POINTER(rx_sc->sa[assoc_num], NULL); clear_rx_sa(rx_sa); rtnl_unlock(); return 0; + +cleanup: + rtnl_unlock(); + return ret; } static int macsec_del_rxsc(struct sk_buff *skb, struct genl_info *info) @@ -1931,6 +1986,7 @@ static int macsec_del_rxsc(struct sk_buff *skb, struct genl_info *info) struct macsec_rx_sc *rx_sc; sci_t sci; struct nlattr *tb_rxsc[MACSEC_RXSC_ATTR_MAX + 1]; + int ret; if (!attrs[MACSEC_ATTR_IFINDEX]) return -EINVAL; @@ -1957,10 +2013,31 @@ static int macsec_del_rxsc(struct sk_buff *skb, struct genl_info *info) return -ENODEV; } + /* If h/w offloading is available, propagate to the device */ + if (macsec_is_offloaded(netdev_priv(dev))) { + const struct macsec_ops *ops; + struct macsec_context ctx; + + ops = macsec_get_ops(netdev_priv(dev), &ctx); + if (!ops) { + ret = -EOPNOTSUPP; + goto cleanup; + } + + ctx.rx_sc = rx_sc; + ret = macsec_offload(ops->mdo_del_rxsc, &ctx); + if (ret) + goto cleanup; + } + free_rx_sc(rx_sc); rtnl_unlock(); return 0; + +cleanup: + rtnl_unlock(); + return ret; } static int macsec_del_txsa(struct sk_buff *skb, struct genl_info *info) @@ -1972,6 +2049,7 @@ static int macsec_del_txsa(struct sk_buff *skb, struct genl_info *info) struct macsec_tx_sa *tx_sa; u8 assoc_num; struct nlattr *tb_sa[MACSEC_SA_ATTR_MAX + 1]; + int ret; if (!attrs[MACSEC_ATTR_IFINDEX]) return -EINVAL; @@ -1992,12 +2070,35 @@ static int macsec_del_txsa(struct sk_buff *skb, struct genl_info *info) return -EBUSY; } + /* If h/w offloading is available, propagate to the device */ + if (macsec_is_offloaded(netdev_priv(dev))) { + const struct macsec_ops *ops; + struct macsec_context ctx; + + ops = macsec_get_ops(netdev_priv(dev), &ctx); + if (!ops) { + ret = -EOPNOTSUPP; + goto cleanup; + } + + ctx.sa.assoc_num = assoc_num; + ctx.sa.tx_sa = tx_sa; + + ret = macsec_offload(ops->mdo_del_txsa, &ctx); + if (ret) + goto cleanup; + } + RCU_INIT_POINTER(tx_sc->sa[assoc_num], NULL); clear_tx_sa(tx_sa); rtnl_unlock(); return 0; + +cleanup: + rtnl_unlock(); + return ret; } static bool validate_upd_sa(struct nlattr **attrs) @@ -2030,6 +2131,9 @@ static int macsec_upd_txsa(struct sk_buff *skb, struct genl_info *info) struct macsec_tx_sa *tx_sa; u8 assoc_num; struct nlattr *tb_sa[MACSEC_SA_ATTR_MAX + 1]; + bool was_operational, was_active; + u32 prev_pn = 0; + int ret = 0; if (!attrs[MACSEC_ATTR_IFINDEX]) return -EINVAL; @@ -2050,19 +2154,52 @@ static int macsec_upd_txsa(struct sk_buff *skb, struct genl_info *info) if (tb_sa[MACSEC_SA_ATTR_PN]) { spin_lock_bh(&tx_sa->lock); + prev_pn = tx_sa->next_pn; tx_sa->next_pn = nla_get_u32(tb_sa[MACSEC_SA_ATTR_PN]); spin_unlock_bh(&tx_sa->lock); } + was_active = tx_sa->active; if (tb_sa[MACSEC_SA_ATTR_ACTIVE]) tx_sa->active = nla_get_u8(tb_sa[MACSEC_SA_ATTR_ACTIVE]); + was_operational = secy->operational; if (assoc_num == tx_sc->encoding_sa) secy->operational = tx_sa->active; + /* If h/w offloading is available, propagate to the device */ + if (macsec_is_offloaded(netdev_priv(dev))) { + const struct macsec_ops *ops; + struct macsec_context ctx; + + ops = macsec_get_ops(netdev_priv(dev), &ctx); + if (!ops) { + ret = -EOPNOTSUPP; + goto cleanup; + } + + ctx.sa.assoc_num = assoc_num; + ctx.sa.tx_sa = tx_sa; + + ret = macsec_offload(ops->mdo_upd_txsa, &ctx); + if (ret) + goto cleanup; + } + rtnl_unlock(); return 0; + +cleanup: + if (tb_sa[MACSEC_SA_ATTR_PN]) { + spin_lock_bh(&tx_sa->lock); + tx_sa->next_pn = prev_pn; + spin_unlock_bh(&tx_sa->lock); + } + tx_sa->active = was_active; + secy->operational = was_operational; + rtnl_unlock(); + return ret; } static int macsec_upd_rxsa(struct sk_buff *skb, struct genl_info *info) @@ -2075,6 +2212,9 @@ static int macsec_upd_rxsa(struct sk_buff *skb, struct genl_info *info) u8 assoc_num; struct nlattr *tb_rxsc[MACSEC_RXSC_ATTR_MAX + 1]; struct nlattr *tb_sa[MACSEC_SA_ATTR_MAX + 1]; + bool was_active; + u32 prev_pn = 0; + int ret = 0; if (!attrs[MACSEC_ATTR_IFINDEX]) return -EINVAL; @@ -2098,15 +2238,46 @@ static int macsec_upd_rxsa(struct sk_buff *skb, struct genl_info *info) if (tb_sa[MACSEC_SA_ATTR_PN]) { spin_lock_bh(&rx_sa->lock); + prev_pn = rx_sa->next_pn; rx_sa->next_pn = nla_get_u32(tb_sa[MACSEC_SA_ATTR_PN]); spin_unlock_bh(&rx_sa->lock); } + was_active = rx_sa->active; if (tb_sa[MACSEC_SA_ATTR_ACTIVE]) rx_sa->active = nla_get_u8(tb_sa[MACSEC_SA_ATTR_ACTIVE]); + /* If h/w offloading is available, propagate to the device */ + if (macsec_is_offloaded(netdev_priv(dev))) { + const struct macsec_ops *ops; + struct macsec_context ctx; + + ops = macsec_get_ops(netdev_priv(dev), &ctx); + if (!ops) { + ret = -EOPNOTSUPP; + goto cleanup; + } + + ctx.sa.assoc_num = assoc_num; + ctx.sa.rx_sa = rx_sa; + + ret = macsec_offload(ops->mdo_upd_rxsa, &ctx); + if (ret) + goto cleanup; + } + rtnl_unlock(); return 0; + +cleanup: + if (tb_sa[MACSEC_SA_ATTR_PN]) { + spin_lock_bh(&rx_sa->lock); + rx_sa->next_pn = prev_pn; + spin_unlock_bh(&rx_sa->lock); + } + rx_sa->active = was_active; + rtnl_unlock(); + return ret; } static int macsec_upd_rxsc(struct sk_buff *skb, struct genl_info *info) @@ -2116,6 +2287,9 @@ static int macsec_upd_rxsc(struct sk_buff *skb, struct genl_info *info) struct macsec_secy *secy; struct macsec_rx_sc *rx_sc; struct nlattr *tb_rxsc[MACSEC_RXSC_ATTR_MAX + 1]; + unsigned int prev_n_rx_sc; + bool was_active; + int ret; if (!attrs[MACSEC_ATTR_IFINDEX]) return -EINVAL; @@ -2133,6 +2307,8 @@ static int macsec_upd_rxsc(struct sk_buff *skb, struct genl_info *info) return PTR_ERR(rx_sc); } + was_active = rx_sc->active; + prev_n_rx_sc = secy->n_rx_sc; if (tb_rxsc[MACSEC_RXSC_ATTR_ACTIVE]) { bool new = !!nla_get_u8(tb_rxsc[MACSEC_RXSC_ATTR_ACTIVE]); @@ -2142,9 +2318,153 @@ static int macsec_upd_rxsc(struct sk_buff *skb, struct genl_info *info) rx_sc->active = new; } + /* If h/w offloading is available, propagate to the device */ + if (macsec_is_offloaded(netdev_priv(dev))) { + const struct macsec_ops *ops; + struct macsec_context ctx; + + ops = macsec_get_ops(netdev_priv(dev), &ctx); + if (!ops) { + ret = -EOPNOTSUPP; + goto cleanup; + } + + ctx.rx_sc = rx_sc; + + ret = macsec_offload(ops->mdo_upd_rxsc, &ctx); + if (ret) + goto cleanup; + } + + rtnl_unlock(); + + return 0; + +cleanup: + secy->n_rx_sc = prev_n_rx_sc; + rx_sc->active = was_active; rtnl_unlock(); + return ret; +} + +static bool macsec_is_configured(struct macsec_dev *macsec) +{ + struct macsec_secy *secy = &macsec->secy; + struct macsec_tx_sc *tx_sc = &secy->tx_sc; + int i; + + if (secy->n_rx_sc > 0) + return true; + + for (i = 0; i < MACSEC_NUM_AN; i++) + if (tx_sc->sa[i]) + return true; + + return false; +} + +static int macsec_upd_offload(struct sk_buff *skb, struct genl_info *info) +{ + struct nlattr *tb_offload[MACSEC_OFFLOAD_ATTR_MAX + 1]; + enum macsec_offload offload, prev_offload; + int (*func)(struct macsec_context *ctx); + struct nlattr **attrs = info->attrs; + struct net_device *dev, *loop_dev; + const struct macsec_ops *ops; + struct macsec_context ctx; + struct macsec_dev *macsec; + struct net *loop_net; + int ret; + + if (!attrs[MACSEC_ATTR_IFINDEX]) + return -EINVAL; + + if (!attrs[MACSEC_ATTR_OFFLOAD]) + return -EINVAL; + + if (nla_parse_nested_deprecated(tb_offload, MACSEC_OFFLOAD_ATTR_MAX, + attrs[MACSEC_ATTR_OFFLOAD], + macsec_genl_offload_policy, NULL)) + return -EINVAL; + + dev = get_dev_from_nl(genl_info_net(info), attrs); + if (IS_ERR(dev)) + return PTR_ERR(dev); + macsec = macsec_priv(dev); + + offload = nla_get_u8(tb_offload[MACSEC_OFFLOAD_ATTR_TYPE]); + if (macsec->offload == offload) + return 0; + + /* Check if the offloading mode is supported by the underlying layers */ + if (offload != MACSEC_OFFLOAD_OFF && + !macsec_check_offload(offload, macsec)) + return -EOPNOTSUPP; + + if (offload == MACSEC_OFFLOAD_OFF) + goto skip_limitation; + /* Check the physical interface isn't offloading another interface + * first. + */ + for_each_net(loop_net) { + for_each_netdev(loop_net, loop_dev) { + struct macsec_dev *priv; + + if (!netif_is_macsec(loop_dev)) + continue; + + priv = macsec_priv(loop_dev); + + if (priv->real_dev == macsec->real_dev && + priv->offload != MACSEC_OFFLOAD_OFF) + return -EBUSY; + } + } + +skip_limitation: + /* Check if the net device is busy. */ + if (netif_running(dev)) + return -EBUSY; + + rtnl_lock(); + + prev_offload = macsec->offload; + macsec->offload = offload; + + /* Check if the device already has rules configured: we do not support + * rules migration. + */ + if (macsec_is_configured(macsec)) { + ret = -EBUSY; + goto rollback; + } + + ops = __macsec_get_ops(offload == MACSEC_OFFLOAD_OFF ? prev_offload : offload, + macsec, &ctx); + if (!ops) { + ret = -EOPNOTSUPP; + goto rollback; + } + + if (prev_offload == MACSEC_OFFLOAD_OFF) + func = ops->mdo_add_secy; + else + func = ops->mdo_del_secy; + + ctx.secy = &macsec->secy; + ret = macsec_offload(func, &ctx); + if (ret) + goto rollback; + + rtnl_unlock(); return 0; + +rollback: + macsec->offload = prev_offload; + + rtnl_unlock(); + return ret; } static int copy_tx_sa_stats(struct sk_buff *skb, @@ -2408,12 +2728,13 @@ static noinline_for_stack int dump_secy(struct macsec_secy *secy, struct net_device *dev, struct sk_buff *skb, struct netlink_callback *cb) { - struct macsec_rx_sc *rx_sc; + struct macsec_dev *macsec = netdev_priv(dev); struct macsec_tx_sc *tx_sc = &secy->tx_sc; struct nlattr *txsa_list, *rxsc_list; - int i, j; - void *hdr; + struct macsec_rx_sc *rx_sc; struct nlattr *attr; + void *hdr; + int i, j; hdr = genlmsg_put(skb, NETLINK_CB(cb->skb).portid, cb->nlh->nlmsg_seq, &macsec_fam, NLM_F_MULTI, MACSEC_CMD_GET_TXSC); @@ -2425,6 +2746,13 @@ dump_secy(struct macsec_secy *secy, struct net_device *dev, if (nla_put_u32(skb, MACSEC_ATTR_IFINDEX, dev->ifindex)) goto nla_put_failure; + attr = nla_nest_start_noflag(skb, MACSEC_ATTR_OFFLOAD); + if (!attr) + goto nla_put_failure; + if (nla_put_u8(skb, MACSEC_OFFLOAD_ATTR_TYPE, macsec->offload)) + goto nla_put_failure; + nla_nest_end(skb, attr); + if (nla_put_secy(secy, skb)) goto nla_put_failure; @@ -2690,6 +3018,12 @@ static const struct genl_ops macsec_genl_ops[] = { .doit = macsec_upd_rxsa, .flags = GENL_ADMIN_PERM, }, + { + .cmd = MACSEC_CMD_UPD_OFFLOAD, + .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP, + .doit = macsec_upd_offload, + .flags = GENL_ADMIN_PERM, + }, }; static struct genl_family macsec_fam __ro_after_init = { @@ -2712,6 +3046,11 @@ static netdev_tx_t macsec_start_xmit(struct sk_buff *skb, struct pcpu_secy_stats *secy_stats; int ret, len; + if (macsec_is_offloaded(netdev_priv(dev))) { + skb->dev = macsec->real_dev; + return dev_queue_xmit(skb); + } + /* 10.5 */ if (!secy->protect_frames) { secy_stats = this_cpu_ptr(macsec->stats); @@ -2825,6 +3164,22 @@ static int macsec_dev_open(struct net_device *dev) goto clear_allmulti; } + /* If h/w offloading is available, propagate to the device */ + if (macsec_is_offloaded(macsec)) { + const struct macsec_ops *ops; + struct macsec_context ctx; + + ops = macsec_get_ops(netdev_priv(dev), &ctx); + if (!ops) { + err = -EOPNOTSUPP; + goto clear_allmulti; + } + + err = macsec_offload(ops->mdo_dev_open, &ctx); + if (err) + goto clear_allmulti; + } + if (netif_carrier_ok(real_dev)) netif_carrier_on(dev); @@ -2845,6 +3200,16 @@ static int macsec_dev_stop(struct net_device *dev) netif_carrier_off(dev); + /* If h/w offloading is available, propagate to the device */ + if (macsec_is_offloaded(macsec)) { + const struct macsec_ops *ops; + struct macsec_context ctx; + + ops = macsec_get_ops(macsec, &ctx); + if (ops) + macsec_offload(ops->mdo_dev_stop, &ctx); + } + dev_mc_unsync(real_dev, dev); dev_uc_unsync(real_dev, dev); @@ -3076,6 +3441,11 @@ static int macsec_changelink(struct net_device *dev, struct nlattr *tb[], struct nlattr *data[], struct netlink_ext_ack *extack) { + struct macsec_dev *macsec = macsec_priv(dev); + struct macsec_tx_sa tx_sc; + struct macsec_secy secy; + int ret; + if (!data) return 0; @@ -3085,7 +3455,41 @@ static int macsec_changelink(struct net_device *dev, struct nlattr *tb[], data[IFLA_MACSEC_PORT]) return -EINVAL; - return macsec_changelink_common(dev, data); + /* Keep a copy of unmodified secy and tx_sc, in case the offload + * propagation fails, to revert macsec_changelink_common. + */ + memcpy(&secy, &macsec->secy, sizeof(secy)); + memcpy(&tx_sc, &macsec->secy.tx_sc, sizeof(tx_sc)); + + ret = macsec_changelink_common(dev, data); + if (ret) + return ret; + + /* If h/w offloading is available, propagate to the device */ + if (macsec_is_offloaded(macsec)) { + const struct macsec_ops *ops; + struct macsec_context ctx; + int ret; + + ops = macsec_get_ops(netdev_priv(dev), &ctx); + if (!ops) { + ret = -EOPNOTSUPP; + goto cleanup; + } + + ctx.secy = &macsec->secy; + ret = macsec_offload(ops->mdo_upd_secy, &ctx); + if (ret) + goto cleanup; + } + + return 0; + +cleanup: + memcpy(&macsec->secy.tx_sc, &tx_sc, sizeof(tx_sc)); + memcpy(&macsec->secy, &secy, sizeof(secy)); + + return ret; } static void macsec_del_dev(struct macsec_dev *macsec) @@ -3128,6 +3532,18 @@ static void macsec_dellink(struct net_device *dev, struct list_head *head) struct net_device *real_dev = macsec->real_dev; struct macsec_rxh_data *rxd = macsec_data_rtnl(real_dev); + /* If h/w offloading is available, propagate to the device */ + if (macsec_is_offloaded(macsec)) { + const struct macsec_ops *ops; + struct macsec_context ctx; + + ops = macsec_get_ops(netdev_priv(dev), &ctx); + if (ops) { + ctx.secy = &macsec->secy; + macsec_offload(ops->mdo_del_secy, &ctx); + } + } + macsec_common_dellink(dev, head); if (list_empty(&rxd->secys)) { @@ -3239,6 +3655,9 @@ static int macsec_newlink(struct net *net, struct net_device *dev, macsec->real_dev = real_dev; + /* MACsec offloading is off by default */ + macsec->offload = MACSEC_OFFLOAD_OFF; + if (data && data[IFLA_MACSEC_ICV_LEN]) icv_len = nla_get_u8(data[IFLA_MACSEC_ICV_LEN]); dev->mtu = real_dev->mtu - icv_len - macsec_extra_len(true); diff --git a/drivers/net/macvlan.c b/drivers/net/macvlan.c index d066cf58c926..81aa7adf4801 100644 --- a/drivers/net/macvlan.c +++ b/drivers/net/macvlan.c @@ -513,10 +513,11 @@ static int macvlan_queue_xmit(struct sk_buff *skb, struct net_device *dev) const struct macvlan_dev *dest; if (vlan->mode == MACVLAN_MODE_BRIDGE) { - const struct ethhdr *eth = (void *)skb->data; + const struct ethhdr *eth = skb_eth_hdr(skb); /* send to other bridge ports directly */ if (is_multicast_ether_addr(eth->h_dest)) { + skb_reset_mac_header(skb); macvlan_broadcast(skb, port, dev, MACVLAN_MODE_BRIDGE); goto xmit_world; } diff --git a/drivers/net/netdevsim/dev.c b/drivers/net/netdevsim/dev.c index 059711edfc61..b53fbc06e104 100644 --- a/drivers/net/netdevsim/dev.c +++ b/drivers/net/netdevsim/dev.c @@ -53,7 +53,7 @@ static ssize_t nsim_dev_take_snapshot_write(struct file *file, get_random_bytes(dummy_data, NSIM_DEV_DUMMY_REGION_SIZE); - id = devlink_region_shapshot_id_get(priv_to_devlink(nsim_dev)); + id = devlink_region_snapshot_id_get(priv_to_devlink(nsim_dev)); err = devlink_region_snapshot_create(nsim_dev->dummy_region, dummy_data, id, kfree); if (err) { @@ -270,7 +270,7 @@ struct nsim_trap_data { }; /* All driver-specific traps must be documented in - * Documentation/networking/devlink-trap-netdevsim.rst + * Documentation/networking/devlink/netdevsim.rst */ enum { NSIM_TRAP_ID_BASE = DEVLINK_TRAP_GENERIC_ID_MAX, diff --git a/drivers/net/netdevsim/fib.c b/drivers/net/netdevsim/fib.c index b5df308b4e33..f32d56ac3e80 100644 --- a/drivers/net/netdevsim/fib.c +++ b/drivers/net/netdevsim/fib.c @@ -14,6 +14,12 @@ * THE COST OF ALL NECESSARY SERVICING, REPAIR OR CORRECTION. */ +#include <linux/in6.h> +#include <linux/kernel.h> +#include <linux/list.h> +#include <linux/rhashtable.h> +#include <linux/spinlock_types.h> +#include <linux/types.h> #include <net/fib_notifier.h> #include <net/ip_fib.h> #include <net/ip6_fib.h> @@ -36,6 +42,48 @@ struct nsim_fib_data { struct notifier_block fib_nb; struct nsim_per_fib_data ipv4; struct nsim_per_fib_data ipv6; + struct rhashtable fib_rt_ht; + struct list_head fib_rt_list; + spinlock_t fib_lock; /* Protects hashtable, list and accounting */ + struct devlink *devlink; +}; + +struct nsim_fib_rt_key { + unsigned char addr[sizeof(struct in6_addr)]; + unsigned char prefix_len; + int family; + u32 tb_id; +}; + +struct nsim_fib_rt { + struct nsim_fib_rt_key key; + struct rhash_head ht_node; + struct list_head list; /* Member of fib_rt_list */ +}; + +struct nsim_fib4_rt { + struct nsim_fib_rt common; + struct fib_info *fi; + u8 tos; + u8 type; +}; + +struct nsim_fib6_rt { + struct nsim_fib_rt common; + struct list_head nh_list; + unsigned int nhs; +}; + +struct nsim_fib6_rt_nh { + struct list_head list; /* Member of nh_list */ + struct fib6_info *rt; +}; + +static const struct rhashtable_params nsim_fib_rt_ht_params = { + .key_offset = offsetof(struct nsim_fib_rt, key), + .head_offset = offsetof(struct nsim_fib_rt, ht_node), + .key_len = sizeof(struct nsim_fib_rt_key), + .automatic_shrinking = true, }; u64 nsim_fib_get_val(struct nsim_fib_data *fib_data, @@ -144,18 +192,556 @@ static int nsim_fib_account(struct nsim_fib_entry *entry, bool add, return err; } +static void nsim_fib_rt_init(struct nsim_fib_data *data, + struct nsim_fib_rt *fib_rt, const void *addr, + size_t addr_len, unsigned int prefix_len, + int family, u32 tb_id) +{ + memcpy(fib_rt->key.addr, addr, addr_len); + fib_rt->key.prefix_len = prefix_len; + fib_rt->key.family = family; + fib_rt->key.tb_id = tb_id; + list_add(&fib_rt->list, &data->fib_rt_list); +} + +static void nsim_fib_rt_fini(struct nsim_fib_rt *fib_rt) +{ + list_del(&fib_rt->list); +} + +static struct nsim_fib_rt *nsim_fib_rt_lookup(struct rhashtable *fib_rt_ht, + const void *addr, size_t addr_len, + unsigned int prefix_len, + int family, u32 tb_id) +{ + struct nsim_fib_rt_key key; + + memset(&key, 0, sizeof(key)); + memcpy(key.addr, addr, addr_len); + key.prefix_len = prefix_len; + key.family = family; + key.tb_id = tb_id; + + return rhashtable_lookup_fast(fib_rt_ht, &key, nsim_fib_rt_ht_params); +} + +static struct nsim_fib4_rt * +nsim_fib4_rt_create(struct nsim_fib_data *data, + struct fib_entry_notifier_info *fen_info) +{ + struct nsim_fib4_rt *fib4_rt; + + fib4_rt = kzalloc(sizeof(*fib4_rt), GFP_ATOMIC); + if (!fib4_rt) + return NULL; + + nsim_fib_rt_init(data, &fib4_rt->common, &fen_info->dst, sizeof(u32), + fen_info->dst_len, AF_INET, fen_info->tb_id); + + fib4_rt->fi = fen_info->fi; + fib_info_hold(fib4_rt->fi); + fib4_rt->tos = fen_info->tos; + fib4_rt->type = fen_info->type; + + return fib4_rt; +} + +static void nsim_fib4_rt_destroy(struct nsim_fib4_rt *fib4_rt) +{ + fib_info_put(fib4_rt->fi); + nsim_fib_rt_fini(&fib4_rt->common); + kfree(fib4_rt); +} + +static struct nsim_fib4_rt * +nsim_fib4_rt_lookup(struct rhashtable *fib_rt_ht, + const struct fib_entry_notifier_info *fen_info) +{ + struct nsim_fib_rt *fib_rt; + + fib_rt = nsim_fib_rt_lookup(fib_rt_ht, &fen_info->dst, sizeof(u32), + fen_info->dst_len, AF_INET, + fen_info->tb_id); + if (!fib_rt) + return NULL; + + return container_of(fib_rt, struct nsim_fib4_rt, common); +} + +static void nsim_fib4_rt_hw_flags_set(struct net *net, + const struct nsim_fib4_rt *fib4_rt, + bool trap) +{ + u32 *p_dst = (u32 *) fib4_rt->common.key.addr; + int dst_len = fib4_rt->common.key.prefix_len; + struct fib_rt_info fri; + + fri.fi = fib4_rt->fi; + fri.tb_id = fib4_rt->common.key.tb_id; + fri.dst = cpu_to_be32(*p_dst); + fri.dst_len = dst_len; + fri.tos = fib4_rt->tos; + fri.type = fib4_rt->type; + fri.offload = false; + fri.trap = trap; + fib_alias_hw_flags_set(net, &fri); +} + +static int nsim_fib4_rt_add(struct nsim_fib_data *data, + struct nsim_fib4_rt *fib4_rt, + struct netlink_ext_ack *extack) +{ + struct net *net = devlink_net(data->devlink); + int err; + + err = nsim_fib_account(&data->ipv4.fib, true, extack); + if (err) + return err; + + err = rhashtable_insert_fast(&data->fib_rt_ht, + &fib4_rt->common.ht_node, + nsim_fib_rt_ht_params); + if (err) { + NL_SET_ERR_MSG_MOD(extack, "Failed to insert IPv4 route"); + goto err_fib_dismiss; + } + + nsim_fib4_rt_hw_flags_set(net, fib4_rt, true); + + return 0; + +err_fib_dismiss: + nsim_fib_account(&data->ipv4.fib, false, extack); + return err; +} + +static int nsim_fib4_rt_replace(struct nsim_fib_data *data, + struct nsim_fib4_rt *fib4_rt, + struct nsim_fib4_rt *fib4_rt_old, + struct netlink_ext_ack *extack) +{ + struct net *net = devlink_net(data->devlink); + int err; + + /* We are replacing a route, so no need to change the accounting. */ + err = rhashtable_replace_fast(&data->fib_rt_ht, + &fib4_rt_old->common.ht_node, + &fib4_rt->common.ht_node, + nsim_fib_rt_ht_params); + if (err) { + NL_SET_ERR_MSG_MOD(extack, "Failed to replace IPv4 route"); + return err; + } + + nsim_fib4_rt_hw_flags_set(net, fib4_rt, true); + + nsim_fib4_rt_hw_flags_set(net, fib4_rt_old, false); + nsim_fib4_rt_destroy(fib4_rt_old); + + return 0; +} + +static int nsim_fib4_rt_insert(struct nsim_fib_data *data, + struct fib_entry_notifier_info *fen_info) +{ + struct netlink_ext_ack *extack = fen_info->info.extack; + struct nsim_fib4_rt *fib4_rt, *fib4_rt_old; + int err; + + fib4_rt = nsim_fib4_rt_create(data, fen_info); + if (!fib4_rt) + return -ENOMEM; + + fib4_rt_old = nsim_fib4_rt_lookup(&data->fib_rt_ht, fen_info); + if (!fib4_rt_old) + err = nsim_fib4_rt_add(data, fib4_rt, extack); + else + err = nsim_fib4_rt_replace(data, fib4_rt, fib4_rt_old, extack); + + if (err) + nsim_fib4_rt_destroy(fib4_rt); + + return err; +} + +static void nsim_fib4_rt_remove(struct nsim_fib_data *data, + const struct fib_entry_notifier_info *fen_info) +{ + struct netlink_ext_ack *extack = fen_info->info.extack; + struct nsim_fib4_rt *fib4_rt; + + fib4_rt = nsim_fib4_rt_lookup(&data->fib_rt_ht, fen_info); + if (WARN_ON_ONCE(!fib4_rt)) + return; + + rhashtable_remove_fast(&data->fib_rt_ht, &fib4_rt->common.ht_node, + nsim_fib_rt_ht_params); + nsim_fib_account(&data->ipv4.fib, false, extack); + nsim_fib4_rt_destroy(fib4_rt); +} + +static int nsim_fib4_event(struct nsim_fib_data *data, + struct fib_notifier_info *info, + unsigned long event) +{ + struct fib_entry_notifier_info *fen_info; + int err = 0; + + fen_info = container_of(info, struct fib_entry_notifier_info, info); + + if (fen_info->fi->nh) { + NL_SET_ERR_MSG_MOD(info->extack, "IPv4 route with nexthop objects is not supported"); + return 0; + } + + switch (event) { + case FIB_EVENT_ENTRY_REPLACE: + err = nsim_fib4_rt_insert(data, fen_info); + break; + case FIB_EVENT_ENTRY_DEL: + nsim_fib4_rt_remove(data, fen_info); + break; + default: + break; + } + + return err; +} + +static struct nsim_fib6_rt_nh * +nsim_fib6_rt_nh_find(const struct nsim_fib6_rt *fib6_rt, + const struct fib6_info *rt) +{ + struct nsim_fib6_rt_nh *fib6_rt_nh; + + list_for_each_entry(fib6_rt_nh, &fib6_rt->nh_list, list) { + if (fib6_rt_nh->rt == rt) + return fib6_rt_nh; + } + + return NULL; +} + +static int nsim_fib6_rt_nh_add(struct nsim_fib6_rt *fib6_rt, + struct fib6_info *rt) +{ + struct nsim_fib6_rt_nh *fib6_rt_nh; + + fib6_rt_nh = kzalloc(sizeof(*fib6_rt_nh), GFP_ATOMIC); + if (!fib6_rt_nh) + return -ENOMEM; + + fib6_info_hold(rt); + fib6_rt_nh->rt = rt; + list_add_tail(&fib6_rt_nh->list, &fib6_rt->nh_list); + fib6_rt->nhs++; + + return 0; +} + +static void nsim_fib6_rt_nh_del(struct nsim_fib6_rt *fib6_rt, + const struct fib6_info *rt) +{ + struct nsim_fib6_rt_nh *fib6_rt_nh; + + fib6_rt_nh = nsim_fib6_rt_nh_find(fib6_rt, rt); + if (WARN_ON_ONCE(!fib6_rt_nh)) + return; + + fib6_rt->nhs--; + list_del(&fib6_rt_nh->list); +#if IS_ENABLED(CONFIG_IPV6) + fib6_info_release(fib6_rt_nh->rt); +#endif + kfree(fib6_rt_nh); +} + +static struct nsim_fib6_rt * +nsim_fib6_rt_create(struct nsim_fib_data *data, + struct fib6_entry_notifier_info *fen6_info) +{ + struct fib6_info *iter, *rt = fen6_info->rt; + struct nsim_fib6_rt *fib6_rt; + int i = 0; + int err; + + fib6_rt = kzalloc(sizeof(*fib6_rt), GFP_ATOMIC); + if (!fib6_rt) + return ERR_PTR(-ENOMEM); + + nsim_fib_rt_init(data, &fib6_rt->common, &rt->fib6_dst.addr, + sizeof(rt->fib6_dst.addr), rt->fib6_dst.plen, AF_INET6, + rt->fib6_table->tb6_id); + + /* We consider a multipath IPv6 route as one entry, but it can be made + * up from several fib6_info structs (one for each nexthop), so we + * add them all to the same list under the entry. + */ + INIT_LIST_HEAD(&fib6_rt->nh_list); + + err = nsim_fib6_rt_nh_add(fib6_rt, rt); + if (err) + goto err_fib_rt_fini; + + if (!fen6_info->nsiblings) + return fib6_rt; + + list_for_each_entry(iter, &rt->fib6_siblings, fib6_siblings) { + if (i == fen6_info->nsiblings) + break; + + err = nsim_fib6_rt_nh_add(fib6_rt, iter); + if (err) + goto err_fib6_rt_nh_del; + i++; + } + WARN_ON_ONCE(i != fen6_info->nsiblings); + + return fib6_rt; + +err_fib6_rt_nh_del: + list_for_each_entry_continue_reverse(iter, &rt->fib6_siblings, + fib6_siblings) + nsim_fib6_rt_nh_del(fib6_rt, iter); + nsim_fib6_rt_nh_del(fib6_rt, rt); +err_fib_rt_fini: + nsim_fib_rt_fini(&fib6_rt->common); + kfree(fib6_rt); + return ERR_PTR(err); +} + +static void nsim_fib6_rt_destroy(struct nsim_fib6_rt *fib6_rt) +{ + struct nsim_fib6_rt_nh *iter, *tmp; + + list_for_each_entry_safe(iter, tmp, &fib6_rt->nh_list, list) + nsim_fib6_rt_nh_del(fib6_rt, iter->rt); + WARN_ON_ONCE(!list_empty(&fib6_rt->nh_list)); + nsim_fib_rt_fini(&fib6_rt->common); + kfree(fib6_rt); +} + +static struct nsim_fib6_rt * +nsim_fib6_rt_lookup(struct rhashtable *fib_rt_ht, const struct fib6_info *rt) +{ + struct nsim_fib_rt *fib_rt; + + fib_rt = nsim_fib_rt_lookup(fib_rt_ht, &rt->fib6_dst.addr, + sizeof(rt->fib6_dst.addr), + rt->fib6_dst.plen, AF_INET6, + rt->fib6_table->tb6_id); + if (!fib_rt) + return NULL; + + return container_of(fib_rt, struct nsim_fib6_rt, common); +} + +static int nsim_fib6_rt_append(struct nsim_fib_data *data, + struct fib6_entry_notifier_info *fen6_info) +{ + struct fib6_info *iter, *rt = fen6_info->rt; + struct nsim_fib6_rt *fib6_rt; + int i = 0; + int err; + + fib6_rt = nsim_fib6_rt_lookup(&data->fib_rt_ht, rt); + if (WARN_ON_ONCE(!fib6_rt)) + return -EINVAL; + + err = nsim_fib6_rt_nh_add(fib6_rt, rt); + if (err) + return err; + rt->trap = true; + + if (!fen6_info->nsiblings) + return 0; + + list_for_each_entry(iter, &rt->fib6_siblings, fib6_siblings) { + if (i == fen6_info->nsiblings) + break; + + err = nsim_fib6_rt_nh_add(fib6_rt, iter); + if (err) + goto err_fib6_rt_nh_del; + iter->trap = true; + i++; + } + WARN_ON_ONCE(i != fen6_info->nsiblings); + + return 0; + +err_fib6_rt_nh_del: + list_for_each_entry_continue_reverse(iter, &rt->fib6_siblings, + fib6_siblings) { + iter->trap = false; + nsim_fib6_rt_nh_del(fib6_rt, iter); + } + rt->trap = false; + nsim_fib6_rt_nh_del(fib6_rt, rt); + return err; +} + +static void nsim_fib6_rt_hw_flags_set(const struct nsim_fib6_rt *fib6_rt, + bool trap) +{ + struct nsim_fib6_rt_nh *fib6_rt_nh; + + list_for_each_entry(fib6_rt_nh, &fib6_rt->nh_list, list) + fib6_info_hw_flags_set(fib6_rt_nh->rt, false, trap); +} + +static int nsim_fib6_rt_add(struct nsim_fib_data *data, + struct nsim_fib6_rt *fib6_rt, + struct netlink_ext_ack *extack) +{ + int err; + + err = nsim_fib_account(&data->ipv6.fib, true, extack); + if (err) + return err; + + err = rhashtable_insert_fast(&data->fib_rt_ht, + &fib6_rt->common.ht_node, + nsim_fib_rt_ht_params); + if (err) { + NL_SET_ERR_MSG_MOD(extack, "Failed to insert IPv6 route"); + goto err_fib_dismiss; + } + + nsim_fib6_rt_hw_flags_set(fib6_rt, true); + + return 0; + +err_fib_dismiss: + nsim_fib_account(&data->ipv6.fib, false, extack); + return err; +} + +static int nsim_fib6_rt_replace(struct nsim_fib_data *data, + struct nsim_fib6_rt *fib6_rt, + struct nsim_fib6_rt *fib6_rt_old, + struct netlink_ext_ack *extack) +{ + int err; + + /* We are replacing a route, so no need to change the accounting. */ + err = rhashtable_replace_fast(&data->fib_rt_ht, + &fib6_rt_old->common.ht_node, + &fib6_rt->common.ht_node, + nsim_fib_rt_ht_params); + if (err) { + NL_SET_ERR_MSG_MOD(extack, "Failed to replace IPv6 route"); + return err; + } + + nsim_fib6_rt_hw_flags_set(fib6_rt, true); + + nsim_fib6_rt_hw_flags_set(fib6_rt_old, false); + nsim_fib6_rt_destroy(fib6_rt_old); + + return 0; +} + +static int nsim_fib6_rt_insert(struct nsim_fib_data *data, + struct fib6_entry_notifier_info *fen6_info) +{ + struct netlink_ext_ack *extack = fen6_info->info.extack; + struct nsim_fib6_rt *fib6_rt, *fib6_rt_old; + int err; + + fib6_rt = nsim_fib6_rt_create(data, fen6_info); + if (IS_ERR(fib6_rt)) + return PTR_ERR(fib6_rt); + + fib6_rt_old = nsim_fib6_rt_lookup(&data->fib_rt_ht, fen6_info->rt); + if (!fib6_rt_old) + err = nsim_fib6_rt_add(data, fib6_rt, extack); + else + err = nsim_fib6_rt_replace(data, fib6_rt, fib6_rt_old, extack); + + if (err) + nsim_fib6_rt_destroy(fib6_rt); + + return err; +} + +static void +nsim_fib6_rt_remove(struct nsim_fib_data *data, + const struct fib6_entry_notifier_info *fen6_info) +{ + struct netlink_ext_ack *extack = fen6_info->info.extack; + struct nsim_fib6_rt *fib6_rt; + + /* Multipath routes are first added to the FIB trie and only then + * notified. If we vetoed the addition, we will get a delete + * notification for a route we do not have. Therefore, do not warn if + * route was not found. + */ + fib6_rt = nsim_fib6_rt_lookup(&data->fib_rt_ht, fen6_info->rt); + if (!fib6_rt) + return; + + /* If not all the nexthops are deleted, then only reduce the nexthop + * group. + */ + if (fen6_info->nsiblings + 1 != fib6_rt->nhs) { + nsim_fib6_rt_nh_del(fib6_rt, fen6_info->rt); + return; + } + + rhashtable_remove_fast(&data->fib_rt_ht, &fib6_rt->common.ht_node, + nsim_fib_rt_ht_params); + nsim_fib_account(&data->ipv6.fib, false, extack); + nsim_fib6_rt_destroy(fib6_rt); +} + +static int nsim_fib6_event(struct nsim_fib_data *data, + struct fib_notifier_info *info, + unsigned long event) +{ + struct fib6_entry_notifier_info *fen6_info; + int err = 0; + + fen6_info = container_of(info, struct fib6_entry_notifier_info, info); + + if (fen6_info->rt->nh) { + NL_SET_ERR_MSG_MOD(info->extack, "IPv6 route with nexthop objects is not supported"); + return 0; + } + + if (fen6_info->rt->fib6_src.plen) { + NL_SET_ERR_MSG_MOD(info->extack, "IPv6 source-specific route is not supported"); + return 0; + } + + switch (event) { + case FIB_EVENT_ENTRY_REPLACE: + err = nsim_fib6_rt_insert(data, fen6_info); + break; + case FIB_EVENT_ENTRY_APPEND: + err = nsim_fib6_rt_append(data, fen6_info); + break; + case FIB_EVENT_ENTRY_DEL: + nsim_fib6_rt_remove(data, fen6_info); + break; + default: + break; + } + + return err; +} + static int nsim_fib_event(struct nsim_fib_data *data, - struct fib_notifier_info *info, bool add) + struct fib_notifier_info *info, unsigned long event) { - struct netlink_ext_ack *extack = info->extack; int err = 0; switch (info->family) { case AF_INET: - err = nsim_fib_account(&data->ipv4.fib, add, extack); + err = nsim_fib4_event(data, info, event); break; case AF_INET6: - err = nsim_fib_account(&data->ipv6.fib, add, extack); + err = nsim_fib6_event(data, info, event); break; } @@ -170,6 +756,9 @@ static int nsim_fib_event_nb(struct notifier_block *nb, unsigned long event, struct fib_notifier_info *info = ptr; int err = 0; + /* IPv6 routes can be added via RAs from softIRQ. */ + spin_lock_bh(&data->fib_lock); + switch (event) { case FIB_EVENT_RULE_ADD: /* fall through */ case FIB_EVENT_RULE_DEL: @@ -178,23 +767,74 @@ static int nsim_fib_event_nb(struct notifier_block *nb, unsigned long event, break; case FIB_EVENT_ENTRY_REPLACE: /* fall through */ + case FIB_EVENT_ENTRY_APPEND: /* fall through */ case FIB_EVENT_ENTRY_DEL: - err = nsim_fib_event(data, info, event != FIB_EVENT_ENTRY_DEL); + err = nsim_fib_event(data, info, event); break; } + spin_unlock_bh(&data->fib_lock); + return notifier_from_errno(err); } +static void nsim_fib4_rt_free(struct nsim_fib_rt *fib_rt, + struct nsim_fib_data *data) +{ + struct devlink *devlink = data->devlink; + struct nsim_fib4_rt *fib4_rt; + + fib4_rt = container_of(fib_rt, struct nsim_fib4_rt, common); + nsim_fib4_rt_hw_flags_set(devlink_net(devlink), fib4_rt, false); + nsim_fib_account(&data->ipv4.fib, false, NULL); + nsim_fib4_rt_destroy(fib4_rt); +} + +static void nsim_fib6_rt_free(struct nsim_fib_rt *fib_rt, + struct nsim_fib_data *data) +{ + struct nsim_fib6_rt *fib6_rt; + + fib6_rt = container_of(fib_rt, struct nsim_fib6_rt, common); + nsim_fib6_rt_hw_flags_set(fib6_rt, false); + nsim_fib_account(&data->ipv6.fib, false, NULL); + nsim_fib6_rt_destroy(fib6_rt); +} + +static void nsim_fib_rt_free(void *ptr, void *arg) +{ + struct nsim_fib_rt *fib_rt = ptr; + struct nsim_fib_data *data = arg; + + switch (fib_rt->key.family) { + case AF_INET: + nsim_fib4_rt_free(fib_rt, data); + break; + case AF_INET6: + nsim_fib6_rt_free(fib_rt, data); + break; + default: + WARN_ON_ONCE(1); + } +} + /* inconsistent dump, trying again */ static void nsim_fib_dump_inconsistent(struct notifier_block *nb) { struct nsim_fib_data *data = container_of(nb, struct nsim_fib_data, fib_nb); + struct nsim_fib_rt *fib_rt, *fib_rt_tmp; + + /* The notifier block is still not registered, so we do not need to + * take any locks here. + */ + list_for_each_entry_safe(fib_rt, fib_rt_tmp, &data->fib_rt_list, list) { + rhashtable_remove_fast(&data->fib_rt_ht, &fib_rt->ht_node, + nsim_fib_rt_ht_params); + nsim_fib_rt_free(fib_rt, data); + } - data->ipv4.fib.num = 0ULL; data->ipv4.rules.num = 0ULL; - data->ipv6.fib.num = 0ULL; data->ipv6.rules.num = 0ULL; } @@ -255,6 +895,13 @@ struct nsim_fib_data *nsim_fib_create(struct devlink *devlink, data = kzalloc(sizeof(*data), GFP_KERNEL); if (!data) return ERR_PTR(-ENOMEM); + data->devlink = devlink; + + spin_lock_init(&data->fib_lock); + INIT_LIST_HEAD(&data->fib_rt_list); + err = rhashtable_init(&data->fib_rt_ht, &nsim_fib_rt_ht_params); + if (err) + goto err_data_free; nsim_fib_set_max_all(data, devlink); @@ -263,7 +910,7 @@ struct nsim_fib_data *nsim_fib_create(struct devlink *devlink, nsim_fib_dump_inconsistent, extack); if (err) { pr_err("Failed to register fib notifier\n"); - goto err_out; + goto err_rhashtable_destroy; } devlink_resource_occ_get_register(devlink, @@ -284,7 +931,10 @@ struct nsim_fib_data *nsim_fib_create(struct devlink *devlink, data); return data; -err_out: +err_rhashtable_destroy: + rhashtable_free_and_destroy(&data->fib_rt_ht, nsim_fib_rt_free, + data); +err_data_free: kfree(data); return ERR_PTR(err); } @@ -300,5 +950,8 @@ void nsim_fib_destroy(struct devlink *devlink, struct nsim_fib_data *data) devlink_resource_occ_get_unregister(devlink, NSIM_RESOURCE_IPV4_FIB); unregister_fib_notifier(devlink_net(devlink), &data->fib_nb); + rhashtable_free_and_destroy(&data->fib_rt_ht, nsim_fib_rt_free, + data); + WARN_ON_ONCE(!list_empty(&data->fib_rt_list)); kfree(data); } diff --git a/drivers/net/phy/Kconfig b/drivers/net/phy/Kconfig index 2e016271e126..6b5ee26795a2 100644 --- a/drivers/net/phy/Kconfig +++ b/drivers/net/phy/Kconfig @@ -346,14 +346,14 @@ config DAVICOM_PHY Currently supports dm9161e and dm9131 config DP83822_PHY - tristate "Texas Instruments DP83822 PHY" + tristate "Texas Instruments DP83822/825 PHYs" ---help--- - Supports the DP83822 PHY. + Supports the DP83822 and DP83825I PHYs. config DP83TC811_PHY - tristate "Texas Instruments DP83TC822 PHY" + tristate "Texas Instruments DP83TC811 PHY" ---help--- - Supports the DP83TC822 PHY. + Supports the DP83TC811 PHY. config DP83848_PHY tristate "Texas Instruments DP83848 PHY" @@ -437,6 +437,9 @@ config MICROCHIP_T1_PHY config MICROSEMI_PHY tristate "Microsemi PHYs" + depends on MACSEC || MACSEC=n + select CRYPTO_AES + select CRYPTO_ECB ---help--- Currently supports VSC8514, VSC8530, VSC8531, VSC8540 and VSC8541 PHYs diff --git a/drivers/net/phy/adin.c b/drivers/net/phy/adin.c index cf5a391c93e6..c7eabe4382fb 100644 --- a/drivers/net/phy/adin.c +++ b/drivers/net/phy/adin.c @@ -145,7 +145,7 @@ struct adin_clause45_mmd_map { u16 adin_regnum; }; -static struct adin_clause45_mmd_map adin_clause45_mmd_map[] = { +static const struct adin_clause45_mmd_map adin_clause45_mmd_map[] = { { MDIO_MMD_PCS, MDIO_PCS_EEE_ABLE, ADIN1300_EEE_CAP_REG }, { MDIO_MMD_AN, MDIO_AN_EEE_LPABLE, ADIN1300_EEE_LPABLE_REG }, { MDIO_MMD_AN, MDIO_AN_EEE_ADV, ADIN1300_EEE_ADV_REG }, @@ -159,7 +159,7 @@ struct adin_hw_stat { u16 reg2; }; -static struct adin_hw_stat adin_hw_stats[] = { +static const struct adin_hw_stat adin_hw_stats[] = { { "total_frames_checked_count", 0x940A, 0x940B }, /* hi + lo */ { "length_error_frames_count", 0x940C }, { "alignment_error_frames_count", 0x940D }, @@ -456,7 +456,7 @@ static int adin_phy_config_intr(struct phy_device *phydev) static int adin_cl45_to_adin_reg(struct phy_device *phydev, int devad, u16 cl45_regnum) { - struct adin_clause45_mmd_map *m; + const struct adin_clause45_mmd_map *m; int i; if (devad == MDIO_MMD_VEND1) @@ -625,7 +625,7 @@ static int adin_soft_reset(struct phy_device *phydev) if (rc < 0) return rc; - msleep(10); + msleep(20); /* If we get a read error something may be wrong */ rc = phy_read_mmd(phydev, MDIO_MMD_VEND1, @@ -650,7 +650,7 @@ static void adin_get_strings(struct phy_device *phydev, u8 *data) } static int adin_read_mmd_stat_regs(struct phy_device *phydev, - struct adin_hw_stat *stat, + const struct adin_hw_stat *stat, u32 *val) { int ret; @@ -676,7 +676,7 @@ static int adin_read_mmd_stat_regs(struct phy_device *phydev, static u64 adin_get_stat(struct phy_device *phydev, int i) { - struct adin_hw_stat *stat = &adin_hw_stats[i]; + const struct adin_hw_stat *stat = &adin_hw_stats[i]; struct adin_priv *priv = phydev->priv; u32 val; int ret; diff --git a/drivers/net/phy/dp83867.c b/drivers/net/phy/dp83867.c index adda0d0eab80..967f57ed0b65 100644 --- a/drivers/net/phy/dp83867.c +++ b/drivers/net/phy/dp83867.c @@ -99,6 +99,7 @@ #define DP83867_PHYCR_TX_FIFO_DEPTH_MASK GENMASK(15, 14) #define DP83867_PHYCR_RX_FIFO_DEPTH_MASK GENMASK(13, 12) #define DP83867_PHYCR_RESERVED_MASK BIT(11) +#define DP83867_PHYCR_FORCE_LINK_GOOD BIT(10) /* RGMIIDCTL bits */ #define DP83867_RGMII_TX_CLK_DELAY_MAX 0xf @@ -635,7 +636,12 @@ static int dp83867_phy_reset(struct phy_device *phydev) usleep_range(10, 20); - return 0; + /* After reset FORCE_LINK_GOOD bit is set. Although the + * default value should be unset. Disable FORCE_LINK_GOOD + * for the phy to work properly. + */ + return phy_modify(phydev, MII_DP83867_PHYCTRL, + DP83867_PHYCR_FORCE_LINK_GOOD, 0); } static struct phy_driver dp83867_driver[] = { diff --git a/drivers/net/phy/mdio_bus.c b/drivers/net/phy/mdio_bus.c index 229e480179ff..9bb9f37f21dc 100644 --- a/drivers/net/phy/mdio_bus.c +++ b/drivers/net/phy/mdio_bus.c @@ -59,17 +59,11 @@ static int mdiobus_register_gpiod(struct mdio_device *mdiodev) static int mdiobus_register_reset(struct mdio_device *mdiodev) { - struct reset_control *reset = NULL; - - if (mdiodev->dev.of_node) - reset = of_reset_control_get_exclusive(mdiodev->dev.of_node, - "phy"); - if (IS_ERR(reset)) { - if (PTR_ERR(reset) == -ENOENT || PTR_ERR(reset) == -ENOTSUPP) - reset = NULL; - else - return PTR_ERR(reset); - } + struct reset_control *reset; + + reset = reset_control_get_optional_exclusive(&mdiodev->dev, "phy"); + if (IS_ERR(reset)) + return PTR_ERR(reset); mdiodev->reset_ctrl = reset; @@ -164,9 +158,11 @@ struct mii_bus *mdiobus_alloc_size(size_t size) if (size) bus->priv = (void *)bus + aligned_size; - /* Initialise the interrupts to polling */ - for (i = 0; i < PHY_MAX_ADDR; i++) + /* Initialise the interrupts to polling and 64-bit seqcounts */ + for (i = 0; i < PHY_MAX_ADDR; i++) { bus->irq[i] = PHY_POLL; + u64_stats_init(&bus->stats[i].syncp); + } return bus; } @@ -255,9 +251,215 @@ static void mdiobus_release(struct device *d) kfree(bus); } +struct mdio_bus_stat_attr { + int addr; + unsigned int field_offset; +}; + +static u64 mdio_bus_get_stat(struct mdio_bus_stats *s, unsigned int offset) +{ + const char *p = (const char *)s + offset; + unsigned int start; + u64 val = 0; + + do { + start = u64_stats_fetch_begin(&s->syncp); + val = u64_stats_read((const u64_stats_t *)p); + } while (u64_stats_fetch_retry(&s->syncp, start)); + + return val; +} + +static u64 mdio_bus_get_global_stat(struct mii_bus *bus, unsigned int offset) +{ + unsigned int i; + u64 val = 0; + + for (i = 0; i < PHY_MAX_ADDR; i++) + val += mdio_bus_get_stat(&bus->stats[i], offset); + + return val; +} + +static ssize_t mdio_bus_stat_field_show(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + struct mii_bus *bus = to_mii_bus(dev); + struct mdio_bus_stat_attr *sattr; + struct dev_ext_attribute *eattr; + u64 val; + + eattr = container_of(attr, struct dev_ext_attribute, attr); + sattr = eattr->var; + + if (sattr->addr < 0) + val = mdio_bus_get_global_stat(bus, sattr->field_offset); + else + val = mdio_bus_get_stat(&bus->stats[sattr->addr], + sattr->field_offset); + + return sprintf(buf, "%llu\n", val); +} + +static ssize_t mdio_bus_device_stat_field_show(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + struct mdio_device *mdiodev = to_mdio_device(dev); + struct mii_bus *bus = mdiodev->bus; + struct mdio_bus_stat_attr *sattr; + struct dev_ext_attribute *eattr; + int addr = mdiodev->addr; + u64 val; + + eattr = container_of(attr, struct dev_ext_attribute, attr); + sattr = eattr->var; + + val = mdio_bus_get_stat(&bus->stats[addr], sattr->field_offset); + + return sprintf(buf, "%llu\n", val); +} + +#define MDIO_BUS_STATS_ATTR_DECL(field, file) \ +static struct dev_ext_attribute dev_attr_mdio_bus_##field = { \ + .attr = { .attr = { .name = file, .mode = 0444 }, \ + .show = mdio_bus_stat_field_show, \ + }, \ + .var = &((struct mdio_bus_stat_attr) { \ + -1, offsetof(struct mdio_bus_stats, field) \ + }), \ +}; \ +static struct dev_ext_attribute dev_attr_mdio_bus_device_##field = { \ + .attr = { .attr = { .name = file, .mode = 0444 }, \ + .show = mdio_bus_device_stat_field_show, \ + }, \ + .var = &((struct mdio_bus_stat_attr) { \ + -1, offsetof(struct mdio_bus_stats, field) \ + }), \ +}; + +#define MDIO_BUS_STATS_ATTR(field) \ + MDIO_BUS_STATS_ATTR_DECL(field, __stringify(field)) + +MDIO_BUS_STATS_ATTR(transfers); +MDIO_BUS_STATS_ATTR(errors); +MDIO_BUS_STATS_ATTR(writes); +MDIO_BUS_STATS_ATTR(reads); + +#define MDIO_BUS_STATS_ADDR_ATTR_DECL(field, addr, file) \ +static struct dev_ext_attribute dev_attr_mdio_bus_addr_##field##_##addr = { \ + .attr = { .attr = { .name = file, .mode = 0444 }, \ + .show = mdio_bus_stat_field_show, \ + }, \ + .var = &((struct mdio_bus_stat_attr) { \ + addr, offsetof(struct mdio_bus_stats, field) \ + }), \ +} + +#define MDIO_BUS_STATS_ADDR_ATTR(field, addr) \ + MDIO_BUS_STATS_ADDR_ATTR_DECL(field, addr, \ + __stringify(field) "_" __stringify(addr)) + +#define MDIO_BUS_STATS_ADDR_ATTR_GROUP_DECL(addr) \ + MDIO_BUS_STATS_ADDR_ATTR(transfers, addr); \ + MDIO_BUS_STATS_ADDR_ATTR(errors, addr); \ + MDIO_BUS_STATS_ADDR_ATTR(writes, addr); \ + MDIO_BUS_STATS_ADDR_ATTR(reads, addr) \ + +MDIO_BUS_STATS_ADDR_ATTR_GROUP_DECL(0); +MDIO_BUS_STATS_ADDR_ATTR_GROUP_DECL(1); +MDIO_BUS_STATS_ADDR_ATTR_GROUP_DECL(2); +MDIO_BUS_STATS_ADDR_ATTR_GROUP_DECL(3); +MDIO_BUS_STATS_ADDR_ATTR_GROUP_DECL(4); +MDIO_BUS_STATS_ADDR_ATTR_GROUP_DECL(5); +MDIO_BUS_STATS_ADDR_ATTR_GROUP_DECL(6); +MDIO_BUS_STATS_ADDR_ATTR_GROUP_DECL(7); +MDIO_BUS_STATS_ADDR_ATTR_GROUP_DECL(8); +MDIO_BUS_STATS_ADDR_ATTR_GROUP_DECL(9); +MDIO_BUS_STATS_ADDR_ATTR_GROUP_DECL(10); +MDIO_BUS_STATS_ADDR_ATTR_GROUP_DECL(11); +MDIO_BUS_STATS_ADDR_ATTR_GROUP_DECL(12); +MDIO_BUS_STATS_ADDR_ATTR_GROUP_DECL(13); +MDIO_BUS_STATS_ADDR_ATTR_GROUP_DECL(14); +MDIO_BUS_STATS_ADDR_ATTR_GROUP_DECL(15); +MDIO_BUS_STATS_ADDR_ATTR_GROUP_DECL(16); +MDIO_BUS_STATS_ADDR_ATTR_GROUP_DECL(17); +MDIO_BUS_STATS_ADDR_ATTR_GROUP_DECL(18); +MDIO_BUS_STATS_ADDR_ATTR_GROUP_DECL(19); +MDIO_BUS_STATS_ADDR_ATTR_GROUP_DECL(20); +MDIO_BUS_STATS_ADDR_ATTR_GROUP_DECL(21); +MDIO_BUS_STATS_ADDR_ATTR_GROUP_DECL(22); +MDIO_BUS_STATS_ADDR_ATTR_GROUP_DECL(23); +MDIO_BUS_STATS_ADDR_ATTR_GROUP_DECL(24); +MDIO_BUS_STATS_ADDR_ATTR_GROUP_DECL(25); +MDIO_BUS_STATS_ADDR_ATTR_GROUP_DECL(26); +MDIO_BUS_STATS_ADDR_ATTR_GROUP_DECL(27); +MDIO_BUS_STATS_ADDR_ATTR_GROUP_DECL(28); +MDIO_BUS_STATS_ADDR_ATTR_GROUP_DECL(29); +MDIO_BUS_STATS_ADDR_ATTR_GROUP_DECL(30); +MDIO_BUS_STATS_ADDR_ATTR_GROUP_DECL(31); + +#define MDIO_BUS_STATS_ADDR_ATTR_GROUP(addr) \ + &dev_attr_mdio_bus_addr_transfers_##addr.attr.attr, \ + &dev_attr_mdio_bus_addr_errors_##addr.attr.attr, \ + &dev_attr_mdio_bus_addr_writes_##addr.attr.attr, \ + &dev_attr_mdio_bus_addr_reads_##addr.attr.attr \ + +static struct attribute *mdio_bus_statistics_attrs[] = { + &dev_attr_mdio_bus_transfers.attr.attr, + &dev_attr_mdio_bus_errors.attr.attr, + &dev_attr_mdio_bus_writes.attr.attr, + &dev_attr_mdio_bus_reads.attr.attr, + MDIO_BUS_STATS_ADDR_ATTR_GROUP(0), + MDIO_BUS_STATS_ADDR_ATTR_GROUP(1), + MDIO_BUS_STATS_ADDR_ATTR_GROUP(2), + MDIO_BUS_STATS_ADDR_ATTR_GROUP(3), + MDIO_BUS_STATS_ADDR_ATTR_GROUP(4), + MDIO_BUS_STATS_ADDR_ATTR_GROUP(5), + MDIO_BUS_STATS_ADDR_ATTR_GROUP(6), + MDIO_BUS_STATS_ADDR_ATTR_GROUP(7), + MDIO_BUS_STATS_ADDR_ATTR_GROUP(8), + MDIO_BUS_STATS_ADDR_ATTR_GROUP(9), + MDIO_BUS_STATS_ADDR_ATTR_GROUP(10), + MDIO_BUS_STATS_ADDR_ATTR_GROUP(11), + MDIO_BUS_STATS_ADDR_ATTR_GROUP(12), + MDIO_BUS_STATS_ADDR_ATTR_GROUP(13), + MDIO_BUS_STATS_ADDR_ATTR_GROUP(14), + MDIO_BUS_STATS_ADDR_ATTR_GROUP(15), + MDIO_BUS_STATS_ADDR_ATTR_GROUP(16), + MDIO_BUS_STATS_ADDR_ATTR_GROUP(17), + MDIO_BUS_STATS_ADDR_ATTR_GROUP(18), + MDIO_BUS_STATS_ADDR_ATTR_GROUP(19), + MDIO_BUS_STATS_ADDR_ATTR_GROUP(20), + MDIO_BUS_STATS_ADDR_ATTR_GROUP(21), + MDIO_BUS_STATS_ADDR_ATTR_GROUP(22), + MDIO_BUS_STATS_ADDR_ATTR_GROUP(23), + MDIO_BUS_STATS_ADDR_ATTR_GROUP(24), + MDIO_BUS_STATS_ADDR_ATTR_GROUP(25), + MDIO_BUS_STATS_ADDR_ATTR_GROUP(26), + MDIO_BUS_STATS_ADDR_ATTR_GROUP(27), + MDIO_BUS_STATS_ADDR_ATTR_GROUP(28), + MDIO_BUS_STATS_ADDR_ATTR_GROUP(29), + MDIO_BUS_STATS_ADDR_ATTR_GROUP(30), + MDIO_BUS_STATS_ADDR_ATTR_GROUP(31), + NULL, +}; + +static const struct attribute_group mdio_bus_statistics_group = { + .name = "statistics", + .attrs = mdio_bus_statistics_attrs, +}; + +static const struct attribute_group *mdio_bus_groups[] = { + &mdio_bus_statistics_group, + NULL, +}; + static struct class mdio_bus_class = { .name = "mdio_bus", .dev_release = mdiobus_release, + .dev_groups = mdio_bus_groups, }; #if IS_ENABLED(CONFIG_OF_MDIO) @@ -536,6 +738,24 @@ struct phy_device *mdiobus_scan(struct mii_bus *bus, int addr) } EXPORT_SYMBOL(mdiobus_scan); +static void mdiobus_stats_acct(struct mdio_bus_stats *stats, bool op, int ret) +{ + u64_stats_update_begin(&stats->syncp); + + u64_stats_inc(&stats->transfers); + if (ret < 0) { + u64_stats_inc(&stats->errors); + goto out; + } + + if (op) + u64_stats_inc(&stats->reads); + else + u64_stats_inc(&stats->writes); +out: + u64_stats_update_end(&stats->syncp); +} + /** * __mdiobus_read - Unlocked version of the mdiobus_read function * @bus: the mii_bus struct @@ -555,6 +775,7 @@ int __mdiobus_read(struct mii_bus *bus, int addr, u32 regnum) retval = bus->read(bus, addr, regnum); trace_mdio_access(bus, 1, addr, regnum, retval, retval); + mdiobus_stats_acct(&bus->stats[addr], true, retval); return retval; } @@ -580,6 +801,7 @@ int __mdiobus_write(struct mii_bus *bus, int addr, u32 regnum, u16 val) err = bus->write(bus, addr, regnum, val); trace_mdio_access(bus, 0, addr, regnum, val, err); + mdiobus_stats_acct(&bus->stats[addr], false, err); return err; } @@ -725,8 +947,27 @@ static int mdio_uevent(struct device *dev, struct kobj_uevent_env *env) return 0; } +static struct attribute *mdio_bus_device_statistics_attrs[] = { + &dev_attr_mdio_bus_device_transfers.attr.attr, + &dev_attr_mdio_bus_device_errors.attr.attr, + &dev_attr_mdio_bus_device_writes.attr.attr, + &dev_attr_mdio_bus_device_reads.attr.attr, + NULL, +}; + +static const struct attribute_group mdio_bus_device_statistics_group = { + .name = "statistics", + .attrs = mdio_bus_device_statistics_attrs, +}; + +static const struct attribute_group *mdio_bus_dev_groups[] = { + &mdio_bus_device_statistics_group, + NULL, +}; + struct bus_type mdio_bus_type = { .name = "mdio_bus", + .dev_groups = mdio_bus_dev_groups, .match = mdio_bus_match, .uevent = mdio_uevent, }; diff --git a/drivers/net/phy/mscc.c b/drivers/net/phy/mscc.c index 50214c081164..937ac7da2789 100644 --- a/drivers/net/phy/mscc.c +++ b/drivers/net/phy/mscc.c @@ -18,6 +18,17 @@ #include <linux/netdevice.h> #include <dt-bindings/net/mscc-phy-vsc8531.h> +#include <linux/scatterlist.h> +#include <crypto/skcipher.h> + +#if IS_ENABLED(CONFIG_MACSEC) +#include <net/macsec.h> +#endif + +#include "mscc_macsec.h" +#include "mscc_mac.h" +#include "mscc_fc_buffer.h" + enum rgmii_rx_clock_delay { RGMII_RX_CLK_DELAY_0_2_NS = 0, RGMII_RX_CLK_DELAY_0_8_NS = 1, @@ -69,7 +80,7 @@ enum rgmii_rx_clock_delay { #define MSCC_PHY_EXT_PHY_CNTL_2 24 #define MII_VSC85XX_INT_MASK 25 -#define MII_VSC85XX_INT_MASK_MASK 0xa000 +#define MII_VSC85XX_INT_MASK_MASK 0xa020 #define MII_VSC85XX_INT_MASK_WOL 0x0040 #define MII_VSC85XX_INT_STATUS 26 @@ -121,6 +132,26 @@ enum rgmii_rx_clock_delay { #define PHY_S6G_PLL_FSM_CTRL_DATA_POS 8 #define PHY_S6G_PLL_FSM_ENA_POS 7 +#define MSCC_EXT_PAGE_MACSEC_17 17 +#define MSCC_EXT_PAGE_MACSEC_18 18 + +#define MSCC_EXT_PAGE_MACSEC_19 19 +#define MSCC_PHY_MACSEC_19_REG_ADDR(x) (x) +#define MSCC_PHY_MACSEC_19_TARGET(x) ((x) << 12) +#define MSCC_PHY_MACSEC_19_READ BIT(14) +#define MSCC_PHY_MACSEC_19_CMD BIT(15) + +#define MSCC_EXT_PAGE_MACSEC_20 20 +#define MSCC_PHY_MACSEC_20_TARGET(x) (x) +enum macsec_bank { + FC_BUFFER = 0x04, + HOST_MAC = 0x05, + LINE_MAC = 0x06, + IP_1588 = 0x0e, + MACSEC_INGR = 0x38, + MACSEC_EGR = 0x3c, +}; + #define MSCC_EXT_PAGE_ACCESS 31 #define MSCC_PHY_PAGE_STANDARD 0x0000 /* Standard registers */ #define MSCC_PHY_PAGE_EXTENDED 0x0001 /* Extended registers */ @@ -128,6 +159,7 @@ enum rgmii_rx_clock_delay { #define MSCC_PHY_PAGE_EXTENDED_3 0x0003 /* Extended reg - page 3 */ #define MSCC_PHY_PAGE_EXTENDED_4 0x0004 /* Extended reg - page 4 */ #define MSCC_PHY_PAGE_CSR_CNTL MSCC_PHY_PAGE_EXTENDED_4 +#define MSCC_PHY_PAGE_MACSEC MSCC_PHY_PAGE_EXTENDED_4 /* Extended reg - GPIO; this is a bank of registers that are shared for all PHYs * in the same package. */ @@ -175,6 +207,9 @@ enum rgmii_rx_clock_delay { #define SECURE_ON_ENABLE 0x8000 #define SECURE_ON_PASSWD_LEN_4 0x4000 +#define MSCC_PHY_EXTENDED_INT 28 +#define MSCC_PHY_EXTENDED_INT_MS_EGR BIT(9) + /* Extended Page 3 Registers */ #define MSCC_PHY_SERDES_TX_VALID_CNT 21 #define MSCC_PHY_SERDES_TX_CRC_ERR_CNT 22 @@ -411,6 +446,44 @@ static const struct vsc85xx_hw_stat vsc8584_hw_stats[] = { }, }; +#if IS_ENABLED(CONFIG_MACSEC) +struct macsec_flow { + struct list_head list; + enum mscc_macsec_destination_ports port; + enum macsec_bank bank; + u32 index; + int assoc_num; + bool has_transformation; + + /* Highest takes precedence [0..15] */ + u8 priority; + + u8 key[MACSEC_KEYID_LEN]; + + union { + struct macsec_rx_sa *rx_sa; + struct macsec_tx_sa *tx_sa; + }; + + /* Matching */ + struct { + u8 sci:1; + u8 tagged:1; + u8 untagged:1; + u8 etype:1; + } match; + + u16 etype; + + /* Action */ + struct { + u8 bypass:1; + u8 drop:1; + } action; + +}; +#endif + struct vsc8531_private { int rate_magic; u16 supp_led_modes; @@ -424,6 +497,19 @@ struct vsc8531_private { * package. */ unsigned int base_addr; + +#if IS_ENABLED(CONFIG_MACSEC) + /* MACsec fields: + * - One SecY per device (enforced at the s/w implementation level) + * - macsec_flows: list of h/w flows + * - ingr_flows: bitmap of ingress flows + * - egr_flows: bitmap of egress flows + */ + struct macsec_secy *secy; + struct list_head macsec_flows; + unsigned long ingr_flows; + unsigned long egr_flows; +#endif }; #ifdef CONFIG_OF_MDIO @@ -1584,6 +1670,978 @@ out: return ret; } +#if IS_ENABLED(CONFIG_MACSEC) +static u32 vsc8584_macsec_phy_read(struct phy_device *phydev, + enum macsec_bank bank, u32 reg) +{ + u32 val, val_l = 0, val_h = 0; + unsigned long deadline; + int rc; + + rc = phy_select_page(phydev, MSCC_PHY_PAGE_MACSEC); + if (rc < 0) + goto failed; + + __phy_write(phydev, MSCC_EXT_PAGE_MACSEC_20, + MSCC_PHY_MACSEC_20_TARGET(bank >> 2)); + + if (bank >> 2 == 0x1) + /* non-MACsec access */ + bank &= 0x3; + else + bank = 0; + + __phy_write(phydev, MSCC_EXT_PAGE_MACSEC_19, + MSCC_PHY_MACSEC_19_CMD | MSCC_PHY_MACSEC_19_READ | + MSCC_PHY_MACSEC_19_REG_ADDR(reg) | + MSCC_PHY_MACSEC_19_TARGET(bank)); + + deadline = jiffies + msecs_to_jiffies(PROC_CMD_NCOMPLETED_TIMEOUT_MS); + do { + val = __phy_read(phydev, MSCC_EXT_PAGE_MACSEC_19); + } while (time_before(jiffies, deadline) && !(val & MSCC_PHY_MACSEC_19_CMD)); + + val_l = __phy_read(phydev, MSCC_EXT_PAGE_MACSEC_17); + val_h = __phy_read(phydev, MSCC_EXT_PAGE_MACSEC_18); + +failed: + phy_restore_page(phydev, rc, rc); + + return (val_h << 16) | val_l; +} + +static void vsc8584_macsec_phy_write(struct phy_device *phydev, + enum macsec_bank bank, u32 reg, u32 val) +{ + unsigned long deadline; + int rc; + + rc = phy_select_page(phydev, MSCC_PHY_PAGE_MACSEC); + if (rc < 0) + goto failed; + + __phy_write(phydev, MSCC_EXT_PAGE_MACSEC_20, + MSCC_PHY_MACSEC_20_TARGET(bank >> 2)); + + if ((bank >> 2 == 0x1) || (bank >> 2 == 0x3)) + bank &= 0x3; + else + /* MACsec access */ + bank = 0; + + __phy_write(phydev, MSCC_EXT_PAGE_MACSEC_17, (u16)val); + __phy_write(phydev, MSCC_EXT_PAGE_MACSEC_18, (u16)(val >> 16)); + + __phy_write(phydev, MSCC_EXT_PAGE_MACSEC_19, + MSCC_PHY_MACSEC_19_CMD | MSCC_PHY_MACSEC_19_REG_ADDR(reg) | + MSCC_PHY_MACSEC_19_TARGET(bank)); + + deadline = jiffies + msecs_to_jiffies(PROC_CMD_NCOMPLETED_TIMEOUT_MS); + do { + val = __phy_read(phydev, MSCC_EXT_PAGE_MACSEC_19); + } while (time_before(jiffies, deadline) && !(val & MSCC_PHY_MACSEC_19_CMD)); + +failed: + phy_restore_page(phydev, rc, rc); +} + +static void vsc8584_macsec_classification(struct phy_device *phydev, + enum macsec_bank bank) +{ + /* enable VLAN tag parsing */ + vsc8584_macsec_phy_write(phydev, bank, MSCC_MS_SAM_CP_TAG, + MSCC_MS_SAM_CP_TAG_PARSE_STAG | + MSCC_MS_SAM_CP_TAG_PARSE_QTAG | + MSCC_MS_SAM_CP_TAG_PARSE_QINQ); +} + +static void vsc8584_macsec_flow_default_action(struct phy_device *phydev, + enum macsec_bank bank, + bool block) +{ + u32 port = (bank == MACSEC_INGR) ? + MSCC_MS_PORT_UNCONTROLLED : MSCC_MS_PORT_COMMON; + u32 action = MSCC_MS_FLOW_BYPASS; + + if (block) + action = MSCC_MS_FLOW_DROP; + + vsc8584_macsec_phy_write(phydev, bank, MSCC_MS_SAM_NM_FLOW_NCP, + /* MACsec untagged */ + MSCC_MS_SAM_NM_FLOW_NCP_UNTAGGED_FLOW_TYPE(action) | + MSCC_MS_SAM_NM_FLOW_NCP_UNTAGGED_DROP_ACTION(MSCC_MS_ACTION_DROP) | + MSCC_MS_SAM_NM_FLOW_NCP_UNTAGGED_DEST_PORT(port) | + /* MACsec tagged */ + MSCC_MS_SAM_NM_FLOW_NCP_TAGGED_FLOW_TYPE(action) | + MSCC_MS_SAM_NM_FLOW_NCP_TAGGED_DROP_ACTION(MSCC_MS_ACTION_DROP) | + MSCC_MS_SAM_NM_FLOW_NCP_TAGGED_DEST_PORT(port) | + /* Bad tag */ + MSCC_MS_SAM_NM_FLOW_NCP_BADTAG_FLOW_TYPE(action) | + MSCC_MS_SAM_NM_FLOW_NCP_BADTAG_DROP_ACTION(MSCC_MS_ACTION_DROP) | + MSCC_MS_SAM_NM_FLOW_NCP_BADTAG_DEST_PORT(port) | + /* Kay tag */ + MSCC_MS_SAM_NM_FLOW_NCP_KAY_FLOW_TYPE(action) | + MSCC_MS_SAM_NM_FLOW_NCP_KAY_DROP_ACTION(MSCC_MS_ACTION_DROP) | + MSCC_MS_SAM_NM_FLOW_NCP_KAY_DEST_PORT(port)); + vsc8584_macsec_phy_write(phydev, bank, MSCC_MS_SAM_NM_FLOW_CP, + /* MACsec untagged */ + MSCC_MS_SAM_NM_FLOW_NCP_UNTAGGED_FLOW_TYPE(action) | + MSCC_MS_SAM_NM_FLOW_CP_UNTAGGED_DROP_ACTION(MSCC_MS_ACTION_DROP) | + MSCC_MS_SAM_NM_FLOW_CP_UNTAGGED_DEST_PORT(port) | + /* MACsec tagged */ + MSCC_MS_SAM_NM_FLOW_NCP_TAGGED_FLOW_TYPE(action) | + MSCC_MS_SAM_NM_FLOW_CP_TAGGED_DROP_ACTION(MSCC_MS_ACTION_DROP) | + MSCC_MS_SAM_NM_FLOW_CP_TAGGED_DEST_PORT(port) | + /* Bad tag */ + MSCC_MS_SAM_NM_FLOW_NCP_BADTAG_FLOW_TYPE(action) | + MSCC_MS_SAM_NM_FLOW_CP_BADTAG_DROP_ACTION(MSCC_MS_ACTION_DROP) | + MSCC_MS_SAM_NM_FLOW_CP_BADTAG_DEST_PORT(port) | + /* Kay tag */ + MSCC_MS_SAM_NM_FLOW_NCP_KAY_FLOW_TYPE(action) | + MSCC_MS_SAM_NM_FLOW_CP_KAY_DROP_ACTION(MSCC_MS_ACTION_DROP) | + MSCC_MS_SAM_NM_FLOW_CP_KAY_DEST_PORT(port)); +} + +static void vsc8584_macsec_integrity_checks(struct phy_device *phydev, + enum macsec_bank bank) +{ + u32 val; + + if (bank != MACSEC_INGR) + return; + + /* Set default rules to pass unmatched frames */ + val = vsc8584_macsec_phy_read(phydev, bank, + MSCC_MS_PARAMS2_IG_CC_CONTROL); + val |= MSCC_MS_PARAMS2_IG_CC_CONTROL_NON_MATCH_CTRL_ACT | + MSCC_MS_PARAMS2_IG_CC_CONTROL_NON_MATCH_ACT; + vsc8584_macsec_phy_write(phydev, bank, MSCC_MS_PARAMS2_IG_CC_CONTROL, + val); + + vsc8584_macsec_phy_write(phydev, bank, MSCC_MS_PARAMS2_IG_CP_TAG, + MSCC_MS_PARAMS2_IG_CP_TAG_PARSE_STAG | + MSCC_MS_PARAMS2_IG_CP_TAG_PARSE_QTAG | + MSCC_MS_PARAMS2_IG_CP_TAG_PARSE_QINQ); +} + +static void vsc8584_macsec_block_init(struct phy_device *phydev, + enum macsec_bank bank) +{ + u32 val; + int i; + + vsc8584_macsec_phy_write(phydev, bank, MSCC_MS_ENA_CFG, + MSCC_MS_ENA_CFG_SW_RST | + MSCC_MS_ENA_CFG_MACSEC_BYPASS_ENA); + + /* Set the MACsec block out of s/w reset and enable clocks */ + vsc8584_macsec_phy_write(phydev, bank, MSCC_MS_ENA_CFG, + MSCC_MS_ENA_CFG_CLK_ENA); + + vsc8584_macsec_phy_write(phydev, bank, MSCC_MS_STATUS_CONTEXT_CTRL, + bank == MACSEC_INGR ? 0xe5880214 : 0xe5880218); + vsc8584_macsec_phy_write(phydev, bank, MSCC_MS_MISC_CONTROL, + MSCC_MS_MISC_CONTROL_MC_LATENCY_FIX(bank == MACSEC_INGR ? 57 : 40) | + MSCC_MS_MISC_CONTROL_XFORM_REC_SIZE(bank == MACSEC_INGR ? 1 : 2)); + + /* Clear the counters */ + val = vsc8584_macsec_phy_read(phydev, bank, MSCC_MS_COUNT_CONTROL); + val |= MSCC_MS_COUNT_CONTROL_AUTO_CNTR_RESET; + vsc8584_macsec_phy_write(phydev, bank, MSCC_MS_COUNT_CONTROL, val); + + /* Enable octet increment mode */ + vsc8584_macsec_phy_write(phydev, bank, MSCC_MS_PP_CTRL, + MSCC_MS_PP_CTRL_MACSEC_OCTET_INCR_MODE); + + vsc8584_macsec_phy_write(phydev, bank, MSCC_MS_BLOCK_CTX_UPDATE, 0x3); + + val = vsc8584_macsec_phy_read(phydev, bank, MSCC_MS_COUNT_CONTROL); + val |= MSCC_MS_COUNT_CONTROL_RESET_ALL; + vsc8584_macsec_phy_write(phydev, bank, MSCC_MS_COUNT_CONTROL, val); + + /* Set the MTU */ + vsc8584_macsec_phy_write(phydev, bank, MSCC_MS_NON_VLAN_MTU_CHECK, + MSCC_MS_NON_VLAN_MTU_CHECK_NV_MTU_COMPARE(32761) | + MSCC_MS_NON_VLAN_MTU_CHECK_NV_MTU_COMP_DROP); + + for (i = 0; i < 8; i++) + vsc8584_macsec_phy_write(phydev, bank, MSCC_MS_VLAN_MTU_CHECK(i), + MSCC_MS_VLAN_MTU_CHECK_MTU_COMPARE(32761) | + MSCC_MS_VLAN_MTU_CHECK_MTU_COMP_DROP); + + if (bank == MACSEC_EGR) { + val = vsc8584_macsec_phy_read(phydev, bank, MSCC_MS_INTR_CTRL_STATUS); + val &= ~MSCC_MS_INTR_CTRL_STATUS_INTR_ENABLE_M; + vsc8584_macsec_phy_write(phydev, bank, MSCC_MS_INTR_CTRL_STATUS, val); + + vsc8584_macsec_phy_write(phydev, bank, MSCC_MS_FC_CFG, + MSCC_MS_FC_CFG_FCBUF_ENA | + MSCC_MS_FC_CFG_LOW_THRESH(0x1) | + MSCC_MS_FC_CFG_HIGH_THRESH(0x4) | + MSCC_MS_FC_CFG_LOW_BYTES_VAL(0x4) | + MSCC_MS_FC_CFG_HIGH_BYTES_VAL(0x6)); + } + + vsc8584_macsec_classification(phydev, bank); + vsc8584_macsec_flow_default_action(phydev, bank, false); + vsc8584_macsec_integrity_checks(phydev, bank); + + /* Enable the MACsec block */ + vsc8584_macsec_phy_write(phydev, bank, MSCC_MS_ENA_CFG, + MSCC_MS_ENA_CFG_CLK_ENA | + MSCC_MS_ENA_CFG_MACSEC_ENA | + MSCC_MS_ENA_CFG_MACSEC_SPEED_MODE(0x5)); +} + +static void vsc8584_macsec_mac_init(struct phy_device *phydev, + enum macsec_bank bank) +{ + u32 val; + int i; + + /* Clear host & line stats */ + for (i = 0; i < 36; i++) + vsc8584_macsec_phy_write(phydev, bank, 0x1c + i, 0); + + val = vsc8584_macsec_phy_read(phydev, bank, + MSCC_MAC_PAUSE_CFG_TX_FRAME_CTRL); + val &= ~MSCC_MAC_PAUSE_CFG_TX_FRAME_CTRL_PAUSE_MODE_M; + val |= MSCC_MAC_PAUSE_CFG_TX_FRAME_CTRL_PAUSE_MODE(2) | + MSCC_MAC_PAUSE_CFG_TX_FRAME_CTRL_PAUSE_VALUE(0xffff); + vsc8584_macsec_phy_write(phydev, bank, + MSCC_MAC_PAUSE_CFG_TX_FRAME_CTRL, val); + + val = vsc8584_macsec_phy_read(phydev, bank, + MSCC_MAC_PAUSE_CFG_TX_FRAME_CTRL_2); + val |= 0xffff; + vsc8584_macsec_phy_write(phydev, bank, + MSCC_MAC_PAUSE_CFG_TX_FRAME_CTRL_2, val); + + val = vsc8584_macsec_phy_read(phydev, bank, + MSCC_MAC_PAUSE_CFG_RX_FRAME_CTRL); + if (bank == HOST_MAC) + val |= MSCC_MAC_PAUSE_CFG_RX_FRAME_CTRL_PAUSE_TIMER_ENA | + MSCC_MAC_PAUSE_CFG_RX_FRAME_CTRL_PAUSE_FRAME_DROP_ENA; + else + val |= MSCC_MAC_PAUSE_CFG_RX_FRAME_CTRL_PAUSE_REACT_ENA | + MSCC_MAC_PAUSE_CFG_RX_FRAME_CTRL_PAUSE_FRAME_DROP_ENA | + MSCC_MAC_PAUSE_CFG_RX_FRAME_CTRL_PAUSE_MODE | + MSCC_MAC_PAUSE_CFG_RX_FRAME_CTRL_EARLY_PAUSE_DETECT_ENA; + vsc8584_macsec_phy_write(phydev, bank, + MSCC_MAC_PAUSE_CFG_RX_FRAME_CTRL, val); + + vsc8584_macsec_phy_write(phydev, bank, MSCC_MAC_CFG_PKTINF_CFG, + MSCC_MAC_CFG_PKTINF_CFG_STRIP_FCS_ENA | + MSCC_MAC_CFG_PKTINF_CFG_INSERT_FCS_ENA | + MSCC_MAC_CFG_PKTINF_CFG_LPI_RELAY_ENA | + MSCC_MAC_CFG_PKTINF_CFG_STRIP_PREAMBLE_ENA | + MSCC_MAC_CFG_PKTINF_CFG_INSERT_PREAMBLE_ENA | + (bank == HOST_MAC ? + MSCC_MAC_CFG_PKTINF_CFG_ENABLE_TX_PADDING : 0)); + + val = vsc8584_macsec_phy_read(phydev, bank, MSCC_MAC_CFG_MODE_CFG); + val &= ~MSCC_MAC_CFG_MODE_CFG_DISABLE_DIC; + vsc8584_macsec_phy_write(phydev, bank, MSCC_MAC_CFG_MODE_CFG, val); + + val = vsc8584_macsec_phy_read(phydev, bank, MSCC_MAC_CFG_MAXLEN_CFG); + val &= ~MSCC_MAC_CFG_MAXLEN_CFG_MAX_LEN_M; + val |= MSCC_MAC_CFG_MAXLEN_CFG_MAX_LEN(10240); + vsc8584_macsec_phy_write(phydev, bank, MSCC_MAC_CFG_MAXLEN_CFG, val); + + vsc8584_macsec_phy_write(phydev, bank, MSCC_MAC_CFG_ADV_CHK_CFG, + MSCC_MAC_CFG_ADV_CHK_CFG_SFD_CHK_ENA | + MSCC_MAC_CFG_ADV_CHK_CFG_PRM_CHK_ENA | + MSCC_MAC_CFG_ADV_CHK_CFG_OOR_ERR_ENA | + MSCC_MAC_CFG_ADV_CHK_CFG_INR_ERR_ENA); + + val = vsc8584_macsec_phy_read(phydev, bank, MSCC_MAC_CFG_LFS_CFG); + val &= ~MSCC_MAC_CFG_LFS_CFG_LFS_MODE_ENA; + vsc8584_macsec_phy_write(phydev, bank, MSCC_MAC_CFG_LFS_CFG, val); + + vsc8584_macsec_phy_write(phydev, bank, MSCC_MAC_CFG_ENA_CFG, + MSCC_MAC_CFG_ENA_CFG_RX_CLK_ENA | + MSCC_MAC_CFG_ENA_CFG_TX_CLK_ENA | + MSCC_MAC_CFG_ENA_CFG_RX_ENA | + MSCC_MAC_CFG_ENA_CFG_TX_ENA); +} + +/* Must be called with mdio_lock taken */ +static int vsc8584_macsec_init(struct phy_device *phydev) +{ + u32 val; + + vsc8584_macsec_block_init(phydev, MACSEC_INGR); + vsc8584_macsec_block_init(phydev, MACSEC_EGR); + vsc8584_macsec_mac_init(phydev, HOST_MAC); + vsc8584_macsec_mac_init(phydev, LINE_MAC); + + vsc8584_macsec_phy_write(phydev, FC_BUFFER, + MSCC_FCBUF_FC_READ_THRESH_CFG, + MSCC_FCBUF_FC_READ_THRESH_CFG_TX_THRESH(4) | + MSCC_FCBUF_FC_READ_THRESH_CFG_RX_THRESH(5)); + + val = vsc8584_macsec_phy_read(phydev, FC_BUFFER, MSCC_FCBUF_MODE_CFG); + val |= MSCC_FCBUF_MODE_CFG_PAUSE_GEN_ENA | + MSCC_FCBUF_MODE_CFG_RX_PPM_RATE_ADAPT_ENA | + MSCC_FCBUF_MODE_CFG_TX_PPM_RATE_ADAPT_ENA; + vsc8584_macsec_phy_write(phydev, FC_BUFFER, MSCC_FCBUF_MODE_CFG, val); + + vsc8584_macsec_phy_write(phydev, FC_BUFFER, MSCC_FCBUF_PPM_RATE_ADAPT_THRESH_CFG, + MSCC_FCBUF_PPM_RATE_ADAPT_THRESH_CFG_TX_THRESH(8) | + MSCC_FCBUF_PPM_RATE_ADAPT_THRESH_CFG_TX_OFFSET(9)); + + val = vsc8584_macsec_phy_read(phydev, FC_BUFFER, + MSCC_FCBUF_TX_DATA_QUEUE_CFG); + val &= ~(MSCC_FCBUF_TX_DATA_QUEUE_CFG_START_M | + MSCC_FCBUF_TX_DATA_QUEUE_CFG_END_M); + val |= MSCC_FCBUF_TX_DATA_QUEUE_CFG_START(0) | + MSCC_FCBUF_TX_DATA_QUEUE_CFG_END(5119); + vsc8584_macsec_phy_write(phydev, FC_BUFFER, + MSCC_FCBUF_TX_DATA_QUEUE_CFG, val); + + val = vsc8584_macsec_phy_read(phydev, FC_BUFFER, MSCC_FCBUF_ENA_CFG); + val |= MSCC_FCBUF_ENA_CFG_TX_ENA | MSCC_FCBUF_ENA_CFG_RX_ENA; + vsc8584_macsec_phy_write(phydev, FC_BUFFER, MSCC_FCBUF_ENA_CFG, val); + + val = vsc8584_macsec_phy_read(phydev, IP_1588, + MSCC_PROC_0_IP_1588_TOP_CFG_STAT_MODE_CTL); + val &= ~MSCC_PROC_0_IP_1588_TOP_CFG_STAT_MODE_CTL_PROTOCOL_MODE_M; + val |= MSCC_PROC_0_IP_1588_TOP_CFG_STAT_MODE_CTL_PROTOCOL_MODE(4); + vsc8584_macsec_phy_write(phydev, IP_1588, + MSCC_PROC_0_IP_1588_TOP_CFG_STAT_MODE_CTL, val); + + return 0; +} + +static void vsc8584_macsec_flow(struct phy_device *phydev, + struct macsec_flow *flow) +{ + struct vsc8531_private *priv = phydev->priv; + enum macsec_bank bank = flow->bank; + u32 val, match = 0, mask = 0, action = 0, idx = flow->index; + + if (flow->match.tagged) + match |= MSCC_MS_SAM_MISC_MATCH_TAGGED; + if (flow->match.untagged) + match |= MSCC_MS_SAM_MISC_MATCH_UNTAGGED; + + if (bank == MACSEC_INGR && flow->assoc_num >= 0) { + match |= MSCC_MS_SAM_MISC_MATCH_AN(flow->assoc_num); + mask |= MSCC_MS_SAM_MASK_AN_MASK(0x3); + } + + if (bank == MACSEC_INGR && flow->match.sci && flow->rx_sa->sc->sci) { + match |= MSCC_MS_SAM_MISC_MATCH_TCI(BIT(3)); + mask |= MSCC_MS_SAM_MASK_TCI_MASK(BIT(3)) | + MSCC_MS_SAM_MASK_SCI_MASK; + + vsc8584_macsec_phy_write(phydev, bank, MSCC_MS_SAM_MATCH_SCI_LO(idx), + lower_32_bits(flow->rx_sa->sc->sci)); + vsc8584_macsec_phy_write(phydev, bank, MSCC_MS_SAM_MATCH_SCI_HI(idx), + upper_32_bits(flow->rx_sa->sc->sci)); + } + + if (flow->match.etype) { + mask |= MSCC_MS_SAM_MASK_MAC_ETYPE_MASK; + + vsc8584_macsec_phy_write(phydev, bank, MSCC_MS_SAM_MAC_SA_MATCH_HI(idx), + MSCC_MS_SAM_MAC_SA_MATCH_HI_ETYPE(htons(flow->etype))); + } + + match |= MSCC_MS_SAM_MISC_MATCH_PRIORITY(flow->priority); + + vsc8584_macsec_phy_write(phydev, bank, MSCC_MS_SAM_MISC_MATCH(idx), match); + vsc8584_macsec_phy_write(phydev, bank, MSCC_MS_SAM_MASK(idx), mask); + + /* Action for matching packets */ + if (flow->action.drop) + action = MSCC_MS_FLOW_DROP; + else if (flow->action.bypass || flow->port == MSCC_MS_PORT_UNCONTROLLED) + action = MSCC_MS_FLOW_BYPASS; + else + action = (bank == MACSEC_INGR) ? + MSCC_MS_FLOW_INGRESS : MSCC_MS_FLOW_EGRESS; + + val = MSCC_MS_SAM_FLOW_CTRL_FLOW_TYPE(action) | + MSCC_MS_SAM_FLOW_CTRL_DROP_ACTION(MSCC_MS_ACTION_DROP) | + MSCC_MS_SAM_FLOW_CTRL_DEST_PORT(flow->port); + + if (action == MSCC_MS_FLOW_BYPASS) + goto write_ctrl; + + if (bank == MACSEC_INGR) { + if (priv->secy->replay_protect) + val |= MSCC_MS_SAM_FLOW_CTRL_REPLAY_PROTECT; + if (priv->secy->validate_frames == MACSEC_VALIDATE_STRICT) + val |= MSCC_MS_SAM_FLOW_CTRL_VALIDATE_FRAMES(MSCC_MS_VALIDATE_STRICT); + else if (priv->secy->validate_frames == MACSEC_VALIDATE_CHECK) + val |= MSCC_MS_SAM_FLOW_CTRL_VALIDATE_FRAMES(MSCC_MS_VALIDATE_CHECK); + } else if (bank == MACSEC_EGR) { + if (priv->secy->protect_frames) + val |= MSCC_MS_SAM_FLOW_CTRL_PROTECT_FRAME; + if (priv->secy->tx_sc.encrypt) + val |= MSCC_MS_SAM_FLOW_CTRL_CONF_PROTECT; + if (priv->secy->tx_sc.send_sci) + val |= MSCC_MS_SAM_FLOW_CTRL_INCLUDE_SCI; + } + +write_ctrl: + vsc8584_macsec_phy_write(phydev, bank, MSCC_MS_SAM_FLOW_CTRL(idx), val); +} + +static struct macsec_flow *vsc8584_macsec_find_flow(struct macsec_context *ctx, + enum macsec_bank bank) +{ + struct vsc8531_private *priv = ctx->phydev->priv; + struct macsec_flow *pos, *tmp; + + list_for_each_entry_safe(pos, tmp, &priv->macsec_flows, list) + if (pos->assoc_num == ctx->sa.assoc_num && pos->bank == bank) + return pos; + + return ERR_PTR(-ENOENT); +} + +static void vsc8584_macsec_flow_enable(struct phy_device *phydev, + struct macsec_flow *flow) +{ + enum macsec_bank bank = flow->bank; + u32 val, idx = flow->index; + + if ((flow->bank == MACSEC_INGR && flow->rx_sa && !flow->rx_sa->active) || + (flow->bank == MACSEC_EGR && flow->tx_sa && !flow->tx_sa->active)) + return; + + /* Enable */ + vsc8584_macsec_phy_write(phydev, bank, MSCC_MS_SAM_ENTRY_SET1, BIT(idx)); + + /* Set in-use */ + val = vsc8584_macsec_phy_read(phydev, bank, MSCC_MS_SAM_FLOW_CTRL(idx)); + val |= MSCC_MS_SAM_FLOW_CTRL_SA_IN_USE; + vsc8584_macsec_phy_write(phydev, bank, MSCC_MS_SAM_FLOW_CTRL(idx), val); +} + +static void vsc8584_macsec_flow_disable(struct phy_device *phydev, + struct macsec_flow *flow) +{ + enum macsec_bank bank = flow->bank; + u32 val, idx = flow->index; + + /* Disable */ + vsc8584_macsec_phy_write(phydev, bank, MSCC_MS_SAM_ENTRY_CLEAR1, BIT(idx)); + + /* Clear in-use */ + val = vsc8584_macsec_phy_read(phydev, bank, MSCC_MS_SAM_FLOW_CTRL(idx)); + val &= ~MSCC_MS_SAM_FLOW_CTRL_SA_IN_USE; + vsc8584_macsec_phy_write(phydev, bank, MSCC_MS_SAM_FLOW_CTRL(idx), val); +} + +static u32 vsc8584_macsec_flow_context_id(struct macsec_flow *flow) +{ + if (flow->bank == MACSEC_INGR) + return flow->index + MSCC_MS_MAX_FLOWS; + + return flow->index; +} + +/* Derive the AES key to get a key for the hash autentication */ +static int vsc8584_macsec_derive_key(const u8 key[MACSEC_KEYID_LEN], + u16 key_len, u8 hkey[16]) +{ + struct crypto_skcipher *tfm = crypto_alloc_skcipher("ecb(aes)", 0, 0); + struct skcipher_request *req = NULL; + struct scatterlist src, dst; + DECLARE_CRYPTO_WAIT(wait); + u32 input[4] = {0}; + int ret; + + if (IS_ERR(tfm)) + return PTR_ERR(tfm); + + req = skcipher_request_alloc(tfm, GFP_KERNEL); + if (!req) { + ret = -ENOMEM; + goto out; + } + + skcipher_request_set_callback(req, CRYPTO_TFM_REQ_MAY_BACKLOG | + CRYPTO_TFM_REQ_MAY_SLEEP, crypto_req_done, + &wait); + ret = crypto_skcipher_setkey(tfm, key, key_len); + if (ret < 0) + goto out; + + sg_init_one(&src, input, 16); + sg_init_one(&dst, hkey, 16); + skcipher_request_set_crypt(req, &src, &dst, 16, NULL); + + ret = crypto_wait_req(crypto_skcipher_encrypt(req), &wait); + +out: + skcipher_request_free(req); + crypto_free_skcipher(tfm); + return ret; +} + +static int vsc8584_macsec_transformation(struct phy_device *phydev, + struct macsec_flow *flow) +{ + struct vsc8531_private *priv = phydev->priv; + enum macsec_bank bank = flow->bank; + int i, ret, index = flow->index; + u32 rec = 0, control = 0; + u8 hkey[16]; + sci_t sci; + + ret = vsc8584_macsec_derive_key(flow->key, priv->secy->key_len, hkey); + if (ret) + return ret; + + switch (priv->secy->key_len) { + case 16: + control |= CONTROL_CRYPTO_ALG(CTRYPTO_ALG_AES_CTR_128); + break; + case 32: + control |= CONTROL_CRYPTO_ALG(CTRYPTO_ALG_AES_CTR_256); + break; + default: + return -EINVAL; + } + + control |= (bank == MACSEC_EGR) ? + (CONTROL_TYPE_EGRESS | CONTROL_AN(priv->secy->tx_sc.encoding_sa)) : + (CONTROL_TYPE_INGRESS | CONTROL_SEQ_MASK); + + control |= CONTROL_UPDATE_SEQ | CONTROL_ENCRYPT_AUTH | CONTROL_KEY_IN_CTX | + CONTROL_IV0 | CONTROL_IV1 | CONTROL_IV_IN_SEQ | + CONTROL_DIGEST_TYPE(0x2) | CONTROL_SEQ_TYPE(0x1) | + CONTROL_AUTH_ALG(AUTH_ALG_AES_GHAS) | CONTROL_CONTEXT_ID; + + /* Set the control word */ + vsc8584_macsec_phy_write(phydev, bank, MSCC_MS_XFORM_REC(index, rec++), + control); + + /* Set the context ID. Must be unique. */ + vsc8584_macsec_phy_write(phydev, bank, MSCC_MS_XFORM_REC(index, rec++), + vsc8584_macsec_flow_context_id(flow)); + + /* Set the encryption/decryption key */ + for (i = 0; i < priv->secy->key_len / sizeof(u32); i++) + vsc8584_macsec_phy_write(phydev, bank, + MSCC_MS_XFORM_REC(index, rec++), + ((u32 *)flow->key)[i]); + + /* Set the authentication key */ + for (i = 0; i < 4; i++) + vsc8584_macsec_phy_write(phydev, bank, + MSCC_MS_XFORM_REC(index, rec++), + ((u32 *)hkey)[i]); + + /* Initial sequence number */ + vsc8584_macsec_phy_write(phydev, bank, MSCC_MS_XFORM_REC(index, rec++), + bank == MACSEC_INGR ? + flow->rx_sa->next_pn : flow->tx_sa->next_pn); + + if (bank == MACSEC_INGR) + /* Set the mask (replay window size) */ + vsc8584_macsec_phy_write(phydev, bank, + MSCC_MS_XFORM_REC(index, rec++), + priv->secy->replay_window); + + /* Set the input vectors */ + sci = bank == MACSEC_INGR ? flow->rx_sa->sc->sci : priv->secy->sci; + vsc8584_macsec_phy_write(phydev, bank, MSCC_MS_XFORM_REC(index, rec++), + lower_32_bits(sci)); + vsc8584_macsec_phy_write(phydev, bank, MSCC_MS_XFORM_REC(index, rec++), + upper_32_bits(sci)); + + while (rec < 20) + vsc8584_macsec_phy_write(phydev, bank, MSCC_MS_XFORM_REC(index, rec++), + 0); + + flow->has_transformation = true; + return 0; +} + +static struct macsec_flow *vsc8584_macsec_alloc_flow(struct vsc8531_private *priv, + enum macsec_bank bank) +{ + unsigned long *bitmap = bank == MACSEC_INGR ? + &priv->ingr_flows : &priv->egr_flows; + struct macsec_flow *flow; + int index; + + index = find_first_zero_bit(bitmap, MSCC_MS_MAX_FLOWS); + + if (index == MSCC_MS_MAX_FLOWS) + return ERR_PTR(-ENOMEM); + + flow = kzalloc(sizeof(*flow), GFP_KERNEL); + if (!flow) + return ERR_PTR(-ENOMEM); + + set_bit(index, bitmap); + flow->index = index; + flow->bank = bank; + flow->priority = 8; + flow->assoc_num = -1; + + list_add_tail(&flow->list, &priv->macsec_flows); + return flow; +} + +static void vsc8584_macsec_free_flow(struct vsc8531_private *priv, + struct macsec_flow *flow) +{ + unsigned long *bitmap = flow->bank == MACSEC_INGR ? + &priv->ingr_flows : &priv->egr_flows; + + list_del(&flow->list); + clear_bit(flow->index, bitmap); + kfree(flow); +} + +static int vsc8584_macsec_add_flow(struct phy_device *phydev, + struct macsec_flow *flow, bool update) +{ + int ret; + + flow->port = MSCC_MS_PORT_CONTROLLED; + vsc8584_macsec_flow(phydev, flow); + + if (update) + return 0; + + ret = vsc8584_macsec_transformation(phydev, flow); + if (ret) { + vsc8584_macsec_free_flow(phydev->priv, flow); + return ret; + } + + return 0; +} + +static int vsc8584_macsec_default_flows(struct phy_device *phydev) +{ + struct macsec_flow *flow; + + /* Add a rule to let the MKA traffic go through, ingress */ + flow = vsc8584_macsec_alloc_flow(phydev->priv, MACSEC_INGR); + if (IS_ERR(flow)) + return PTR_ERR(flow); + + flow->priority = 15; + flow->port = MSCC_MS_PORT_UNCONTROLLED; + flow->match.tagged = 1; + flow->match.untagged = 1; + flow->match.etype = 1; + flow->etype = ETH_P_PAE; + flow->action.bypass = 1; + + vsc8584_macsec_flow(phydev, flow); + vsc8584_macsec_flow_enable(phydev, flow); + + /* Add a rule to let the MKA traffic go through, egress */ + flow = vsc8584_macsec_alloc_flow(phydev->priv, MACSEC_EGR); + if (IS_ERR(flow)) + return PTR_ERR(flow); + + flow->priority = 15; + flow->port = MSCC_MS_PORT_COMMON; + flow->match.untagged = 1; + flow->match.etype = 1; + flow->etype = ETH_P_PAE; + flow->action.bypass = 1; + + vsc8584_macsec_flow(phydev, flow); + vsc8584_macsec_flow_enable(phydev, flow); + + return 0; +} + +static void vsc8584_macsec_del_flow(struct phy_device *phydev, + struct macsec_flow *flow) +{ + vsc8584_macsec_flow_disable(phydev, flow); + vsc8584_macsec_free_flow(phydev->priv, flow); +} + +static int __vsc8584_macsec_add_rxsa(struct macsec_context *ctx, + struct macsec_flow *flow, bool update) +{ + struct phy_device *phydev = ctx->phydev; + struct vsc8531_private *priv = phydev->priv; + + if (!flow) { + flow = vsc8584_macsec_alloc_flow(priv, MACSEC_INGR); + if (IS_ERR(flow)) + return PTR_ERR(flow); + + memcpy(flow->key, ctx->sa.key, priv->secy->key_len); + } + + flow->assoc_num = ctx->sa.assoc_num; + flow->rx_sa = ctx->sa.rx_sa; + + /* Always match tagged packets on ingress */ + flow->match.tagged = 1; + flow->match.sci = 1; + + if (priv->secy->validate_frames != MACSEC_VALIDATE_DISABLED) + flow->match.untagged = 1; + + return vsc8584_macsec_add_flow(phydev, flow, update); +} + +static int __vsc8584_macsec_add_txsa(struct macsec_context *ctx, + struct macsec_flow *flow, bool update) +{ + struct phy_device *phydev = ctx->phydev; + struct vsc8531_private *priv = phydev->priv; + + if (!flow) { + flow = vsc8584_macsec_alloc_flow(priv, MACSEC_EGR); + if (IS_ERR(flow)) + return PTR_ERR(flow); + + memcpy(flow->key, ctx->sa.key, priv->secy->key_len); + } + + flow->assoc_num = ctx->sa.assoc_num; + flow->tx_sa = ctx->sa.tx_sa; + + /* Always match untagged packets on egress */ + flow->match.untagged = 1; + + return vsc8584_macsec_add_flow(phydev, flow, update); +} + +static int vsc8584_macsec_dev_open(struct macsec_context *ctx) +{ + struct vsc8531_private *priv = ctx->phydev->priv; + struct macsec_flow *flow, *tmp; + + /* No operation to perform before the commit step */ + if (ctx->prepare) + return 0; + + list_for_each_entry_safe(flow, tmp, &priv->macsec_flows, list) + vsc8584_macsec_flow_enable(ctx->phydev, flow); + + return 0; +} + +static int vsc8584_macsec_dev_stop(struct macsec_context *ctx) +{ + struct vsc8531_private *priv = ctx->phydev->priv; + struct macsec_flow *flow, *tmp; + + /* No operation to perform before the commit step */ + if (ctx->prepare) + return 0; + + list_for_each_entry_safe(flow, tmp, &priv->macsec_flows, list) + vsc8584_macsec_flow_disable(ctx->phydev, flow); + + return 0; +} + +static int vsc8584_macsec_add_secy(struct macsec_context *ctx) +{ + struct vsc8531_private *priv = ctx->phydev->priv; + struct macsec_secy *secy = ctx->secy; + + if (ctx->prepare) { + if (priv->secy) + return -EEXIST; + + return 0; + } + + priv->secy = secy; + + vsc8584_macsec_flow_default_action(ctx->phydev, MACSEC_EGR, + secy->validate_frames != MACSEC_VALIDATE_DISABLED); + vsc8584_macsec_flow_default_action(ctx->phydev, MACSEC_INGR, + secy->validate_frames != MACSEC_VALIDATE_DISABLED); + + return vsc8584_macsec_default_flows(ctx->phydev); +} + +static int vsc8584_macsec_del_secy(struct macsec_context *ctx) +{ + struct vsc8531_private *priv = ctx->phydev->priv; + struct macsec_flow *flow, *tmp; + + /* No operation to perform before the commit step */ + if (ctx->prepare) + return 0; + + list_for_each_entry_safe(flow, tmp, &priv->macsec_flows, list) + vsc8584_macsec_del_flow(ctx->phydev, flow); + + vsc8584_macsec_flow_default_action(ctx->phydev, MACSEC_EGR, false); + vsc8584_macsec_flow_default_action(ctx->phydev, MACSEC_INGR, false); + + priv->secy = NULL; + return 0; +} + +static int vsc8584_macsec_upd_secy(struct macsec_context *ctx) +{ + /* No operation to perform before the commit step */ + if (ctx->prepare) + return 0; + + vsc8584_macsec_del_secy(ctx); + return vsc8584_macsec_add_secy(ctx); +} + +static int vsc8584_macsec_add_rxsc(struct macsec_context *ctx) +{ + /* Nothing to do */ + return 0; +} + +static int vsc8584_macsec_upd_rxsc(struct macsec_context *ctx) +{ + return -EOPNOTSUPP; +} + +static int vsc8584_macsec_del_rxsc(struct macsec_context *ctx) +{ + struct vsc8531_private *priv = ctx->phydev->priv; + struct macsec_flow *flow, *tmp; + + /* No operation to perform before the commit step */ + if (ctx->prepare) + return 0; + + list_for_each_entry_safe(flow, tmp, &priv->macsec_flows, list) { + if (flow->bank == MACSEC_INGR && flow->rx_sa && + flow->rx_sa->sc->sci == ctx->rx_sc->sci) + vsc8584_macsec_del_flow(ctx->phydev, flow); + } + + return 0; +} + +static int vsc8584_macsec_add_rxsa(struct macsec_context *ctx) +{ + struct macsec_flow *flow = NULL; + + if (ctx->prepare) + return __vsc8584_macsec_add_rxsa(ctx, flow, false); + + flow = vsc8584_macsec_find_flow(ctx, MACSEC_INGR); + if (IS_ERR(flow)) + return PTR_ERR(flow); + + vsc8584_macsec_flow_enable(ctx->phydev, flow); + return 0; +} + +static int vsc8584_macsec_upd_rxsa(struct macsec_context *ctx) +{ + struct macsec_flow *flow; + + flow = vsc8584_macsec_find_flow(ctx, MACSEC_INGR); + if (IS_ERR(flow)) + return PTR_ERR(flow); + + if (ctx->prepare) { + /* Make sure the flow is disabled before updating it */ + vsc8584_macsec_flow_disable(ctx->phydev, flow); + + return __vsc8584_macsec_add_rxsa(ctx, flow, true); + } + + vsc8584_macsec_flow_enable(ctx->phydev, flow); + return 0; +} + +static int vsc8584_macsec_del_rxsa(struct macsec_context *ctx) +{ + struct macsec_flow *flow; + + flow = vsc8584_macsec_find_flow(ctx, MACSEC_INGR); + + if (IS_ERR(flow)) + return PTR_ERR(flow); + if (ctx->prepare) + return 0; + + vsc8584_macsec_del_flow(ctx->phydev, flow); + return 0; +} + +static int vsc8584_macsec_add_txsa(struct macsec_context *ctx) +{ + struct macsec_flow *flow = NULL; + + if (ctx->prepare) + return __vsc8584_macsec_add_txsa(ctx, flow, false); + + flow = vsc8584_macsec_find_flow(ctx, MACSEC_EGR); + if (IS_ERR(flow)) + return PTR_ERR(flow); + + vsc8584_macsec_flow_enable(ctx->phydev, flow); + return 0; +} + +static int vsc8584_macsec_upd_txsa(struct macsec_context *ctx) +{ + struct macsec_flow *flow; + + flow = vsc8584_macsec_find_flow(ctx, MACSEC_EGR); + if (IS_ERR(flow)) + return PTR_ERR(flow); + + if (ctx->prepare) { + /* Make sure the flow is disabled before updating it */ + vsc8584_macsec_flow_disable(ctx->phydev, flow); + + return __vsc8584_macsec_add_txsa(ctx, flow, true); + } + + vsc8584_macsec_flow_enable(ctx->phydev, flow); + return 0; +} + +static int vsc8584_macsec_del_txsa(struct macsec_context *ctx) +{ + struct macsec_flow *flow; + + flow = vsc8584_macsec_find_flow(ctx, MACSEC_EGR); + + if (IS_ERR(flow)) + return PTR_ERR(flow); + if (ctx->prepare) + return 0; + + vsc8584_macsec_del_flow(ctx->phydev, flow); + return 0; +} + +static struct macsec_ops vsc8584_macsec_ops = { + .mdo_dev_open = vsc8584_macsec_dev_open, + .mdo_dev_stop = vsc8584_macsec_dev_stop, + .mdo_add_secy = vsc8584_macsec_add_secy, + .mdo_upd_secy = vsc8584_macsec_upd_secy, + .mdo_del_secy = vsc8584_macsec_del_secy, + .mdo_add_rxsc = vsc8584_macsec_add_rxsc, + .mdo_upd_rxsc = vsc8584_macsec_upd_rxsc, + .mdo_del_rxsc = vsc8584_macsec_del_rxsc, + .mdo_add_rxsa = vsc8584_macsec_add_rxsa, + .mdo_upd_rxsa = vsc8584_macsec_upd_rxsa, + .mdo_del_rxsa = vsc8584_macsec_del_rxsa, + .mdo_add_txsa = vsc8584_macsec_add_txsa, + .mdo_upd_txsa = vsc8584_macsec_upd_txsa, + .mdo_del_txsa = vsc8584_macsec_del_txsa, +}; +#endif /* CONFIG_MACSEC */ + /* Check if one PHY has already done the init of the parts common to all PHYs * in the Quad PHY package. */ @@ -1733,6 +2791,24 @@ static int vsc8584_config_init(struct phy_device *phydev) mutex_unlock(&phydev->mdio.bus->mdio_lock); +#if IS_ENABLED(CONFIG_MACSEC) + /* MACsec */ + switch (phydev->phy_id & phydev->drv->phy_id_mask) { + case PHY_ID_VSC856X: + case PHY_ID_VSC8575: + case PHY_ID_VSC8582: + case PHY_ID_VSC8584: + INIT_LIST_HEAD(&vsc8531->macsec_flows); + vsc8531->secy = NULL; + + phydev->macsec_ops = &vsc8584_macsec_ops; + + ret = vsc8584_macsec_init(phydev); + if (ret) + goto err; + } +#endif + phy_write(phydev, MSCC_EXT_PAGE_ACCESS, MSCC_PHY_PAGE_STANDARD); val = phy_read(phydev, MSCC_PHY_EXT_PHY_CNTL_1); @@ -1758,6 +2834,43 @@ err: return ret; } +static int vsc8584_handle_interrupt(struct phy_device *phydev) +{ +#if IS_ENABLED(CONFIG_MACSEC) + struct vsc8531_private *priv = phydev->priv; + struct macsec_flow *flow, *tmp; + u32 cause, rec; + + /* Check MACsec PN rollover */ + cause = vsc8584_macsec_phy_read(phydev, MACSEC_EGR, + MSCC_MS_INTR_CTRL_STATUS); + cause &= MSCC_MS_INTR_CTRL_STATUS_INTR_CLR_STATUS_M; + if (!(cause & MACSEC_INTR_CTRL_STATUS_ROLLOVER)) + goto skip_rollover; + + rec = 6 + priv->secy->key_len / sizeof(u32); + list_for_each_entry_safe(flow, tmp, &priv->macsec_flows, list) { + u32 val; + + if (flow->bank != MACSEC_EGR || !flow->has_transformation) + continue; + + val = vsc8584_macsec_phy_read(phydev, MACSEC_EGR, + MSCC_MS_XFORM_REC(flow->index, rec)); + if (val == 0xffffffff) { + vsc8584_macsec_flow_disable(phydev, flow); + macsec_pn_wrapped(priv->secy, flow->tx_sa); + break; + } + } + +skip_rollover: +#endif + + phy_mac_interrupt(phydev); + return 0; +} + static int vsc85xx_config_init(struct phy_device *phydev) { int rc, i, phy_id; @@ -2201,6 +3314,20 @@ static int vsc85xx_config_intr(struct phy_device *phydev) int rc; if (phydev->interrupts == PHY_INTERRUPT_ENABLED) { +#if IS_ENABLED(CONFIG_MACSEC) + phy_write(phydev, MSCC_EXT_PAGE_ACCESS, + MSCC_PHY_PAGE_EXTENDED_2); + phy_write(phydev, MSCC_PHY_EXTENDED_INT, + MSCC_PHY_EXTENDED_INT_MS_EGR); + phy_write(phydev, MSCC_EXT_PAGE_ACCESS, + MSCC_PHY_PAGE_STANDARD); + + vsc8584_macsec_phy_write(phydev, MACSEC_EGR, + MSCC_MS_AIC_CTRL, 0xf); + vsc8584_macsec_phy_write(phydev, MACSEC_EGR, + MSCC_MS_INTR_CTRL_STATUS, + MSCC_MS_INTR_CTRL_STATUS_INTR_ENABLE(MACSEC_INTR_CTRL_STATUS_ROLLOVER)); +#endif rc = phy_write(phydev, MII_VSC85XX_INT_MASK, MII_VSC85XX_INT_MASK_MASK); } else { @@ -2550,6 +3677,7 @@ static struct phy_driver vsc85xx_driver[] = { .config_aneg = &vsc85xx_config_aneg, .aneg_done = &genphy_aneg_done, .read_status = &vsc85xx_read_status, + .handle_interrupt = &vsc8584_handle_interrupt, .ack_interrupt = &vsc85xx_ack_interrupt, .config_intr = &vsc85xx_config_intr, .did_interrupt = &vsc8584_did_interrupt, @@ -2602,6 +3730,7 @@ static struct phy_driver vsc85xx_driver[] = { .config_aneg = &vsc85xx_config_aneg, .aneg_done = &genphy_aneg_done, .read_status = &vsc85xx_read_status, + .handle_interrupt = &vsc8584_handle_interrupt, .ack_interrupt = &vsc85xx_ack_interrupt, .config_intr = &vsc85xx_config_intr, .did_interrupt = &vsc8584_did_interrupt, @@ -2626,6 +3755,7 @@ static struct phy_driver vsc85xx_driver[] = { .config_aneg = &vsc85xx_config_aneg, .aneg_done = &genphy_aneg_done, .read_status = &vsc85xx_read_status, + .handle_interrupt = &vsc8584_handle_interrupt, .ack_interrupt = &vsc85xx_ack_interrupt, .config_intr = &vsc85xx_config_intr, .did_interrupt = &vsc8584_did_interrupt, @@ -2650,6 +3780,7 @@ static struct phy_driver vsc85xx_driver[] = { .config_aneg = &vsc85xx_config_aneg, .aneg_done = &genphy_aneg_done, .read_status = &vsc85xx_read_status, + .handle_interrupt = &vsc8584_handle_interrupt, .ack_interrupt = &vsc85xx_ack_interrupt, .config_intr = &vsc85xx_config_intr, .did_interrupt = &vsc8584_did_interrupt, diff --git a/drivers/net/phy/mscc_fc_buffer.h b/drivers/net/phy/mscc_fc_buffer.h new file mode 100644 index 000000000000..7e9c0e877895 --- /dev/null +++ b/drivers/net/phy/mscc_fc_buffer.h @@ -0,0 +1,64 @@ +/* SPDX-License-Identifier: (GPL-2.0 OR MIT) */ +/* + * Microsemi Ocelot Switch driver + * + * Copyright (C) 2019 Microsemi Corporation + */ + +#ifndef _MSCC_OCELOT_FC_BUFFER_H_ +#define _MSCC_OCELOT_FC_BUFFER_H_ + +#define MSCC_FCBUF_ENA_CFG 0x00 +#define MSCC_FCBUF_MODE_CFG 0x01 +#define MSCC_FCBUF_PPM_RATE_ADAPT_THRESH_CFG 0x02 +#define MSCC_FCBUF_TX_CTRL_QUEUE_CFG 0x03 +#define MSCC_FCBUF_TX_DATA_QUEUE_CFG 0x04 +#define MSCC_FCBUF_RX_DATA_QUEUE_CFG 0x05 +#define MSCC_FCBUF_TX_BUFF_XON_XOFF_THRESH_CFG 0x06 +#define MSCC_FCBUF_FC_READ_THRESH_CFG 0x07 +#define MSCC_FCBUF_TX_FRM_GAP_COMP 0x08 + +#define MSCC_FCBUF_ENA_CFG_TX_ENA BIT(0) +#define MSCC_FCBUF_ENA_CFG_RX_ENA BIT(4) + +#define MSCC_FCBUF_MODE_CFG_DROP_BEHAVIOUR BIT(4) +#define MSCC_FCBUF_MODE_CFG_PAUSE_REACT_ENA BIT(8) +#define MSCC_FCBUF_MODE_CFG_RX_PPM_RATE_ADAPT_ENA BIT(12) +#define MSCC_FCBUF_MODE_CFG_TX_PPM_RATE_ADAPT_ENA BIT(16) +#define MSCC_FCBUF_MODE_CFG_TX_CTRL_QUEUE_ENA BIT(20) +#define MSCC_FCBUF_MODE_CFG_PAUSE_GEN_ENA BIT(24) +#define MSCC_FCBUF_MODE_CFG_INCLUDE_PAUSE_RCVD_IN_PAUSE_GEN BIT(28) + +#define MSCC_FCBUF_PPM_RATE_ADAPT_THRESH_CFG_TX_THRESH(x) (x) +#define MSCC_FCBUF_PPM_RATE_ADAPT_THRESH_CFG_TX_THRESH_M GENMASK(15, 0) +#define MSCC_FCBUF_PPM_RATE_ADAPT_THRESH_CFG_TX_OFFSET(x) ((x) << 16) +#define MSCC_FCBUF_PPM_RATE_ADAPT_THRESH_CFG_TX_OFFSET_M GENMASK(19, 16) +#define MSCC_FCBUF_PPM_RATE_ADAPT_THRESH_CFG_RX_THRESH(x) ((x) << 20) +#define MSCC_FCBUF_PPM_RATE_ADAPT_THRESH_CFG_RX_THRESH_M GENMASK(31, 20) + +#define MSCC_FCBUF_TX_CTRL_QUEUE_CFG_START(x) (x) +#define MSCC_FCBUF_TX_CTRL_QUEUE_CFG_START_M GENMASK(15, 0) +#define MSCC_FCBUF_TX_CTRL_QUEUE_CFG_END(x) ((x) << 16) +#define MSCC_FCBUF_TX_CTRL_QUEUE_CFG_END_M GENMASK(31, 16) + +#define MSCC_FCBUF_TX_DATA_QUEUE_CFG_START(x) (x) +#define MSCC_FCBUF_TX_DATA_QUEUE_CFG_START_M GENMASK(15, 0) +#define MSCC_FCBUF_TX_DATA_QUEUE_CFG_END(x) ((x) << 16) +#define MSCC_FCBUF_TX_DATA_QUEUE_CFG_END_M GENMASK(31, 16) + +#define MSCC_FCBUF_RX_DATA_QUEUE_CFG_START(x) (x) +#define MSCC_FCBUF_RX_DATA_QUEUE_CFG_START_M GENMASK(15, 0) +#define MSCC_FCBUF_RX_DATA_QUEUE_CFG_END(x) ((x) << 16) +#define MSCC_FCBUF_RX_DATA_QUEUE_CFG_END_M GENMASK(31, 16) + +#define MSCC_FCBUF_TX_BUFF_XON_XOFF_THRESH_CFG_XOFF_THRESH(x) (x) +#define MSCC_FCBUF_TX_BUFF_XON_XOFF_THRESH_CFG_XOFF_THRESH_M GENMASK(15, 0) +#define MSCC_FCBUF_TX_BUFF_XON_XOFF_THRESH_CFG_XON_THRESH(x) ((x) << 16) +#define MSCC_FCBUF_TX_BUFF_XON_XOFF_THRESH_CFG_XON_THRESH_M GENMASK(31, 16) + +#define MSCC_FCBUF_FC_READ_THRESH_CFG_TX_THRESH(x) (x) +#define MSCC_FCBUF_FC_READ_THRESH_CFG_TX_THRESH_M GENMASK(15, 0) +#define MSCC_FCBUF_FC_READ_THRESH_CFG_RX_THRESH(x) ((x) << 16) +#define MSCC_FCBUF_FC_READ_THRESH_CFG_RX_THRESH_M GENMASK(31, 16) + +#endif diff --git a/drivers/net/phy/mscc_mac.h b/drivers/net/phy/mscc_mac.h new file mode 100644 index 000000000000..9420ee5175a6 --- /dev/null +++ b/drivers/net/phy/mscc_mac.h @@ -0,0 +1,159 @@ +/* SPDX-License-Identifier: (GPL-2.0 OR MIT) */ +/* + * Microsemi Ocelot Switch driver + * + * Copyright (c) 2017 Microsemi Corporation + */ + +#ifndef _MSCC_OCELOT_LINE_MAC_H_ +#define _MSCC_OCELOT_LINE_MAC_H_ + +#define MSCC_MAC_CFG_ENA_CFG 0x00 +#define MSCC_MAC_CFG_MODE_CFG 0x01 +#define MSCC_MAC_CFG_MAXLEN_CFG 0x02 +#define MSCC_MAC_CFG_NUM_TAGS_CFG 0x03 +#define MSCC_MAC_CFG_TAGS_CFG 0x04 +#define MSCC_MAC_CFG_ADV_CHK_CFG 0x07 +#define MSCC_MAC_CFG_LFS_CFG 0x08 +#define MSCC_MAC_CFG_LB_CFG 0x09 +#define MSCC_MAC_CFG_PKTINF_CFG 0x0a +#define MSCC_MAC_PAUSE_CFG_TX_FRAME_CTRL 0x0b +#define MSCC_MAC_PAUSE_CFG_TX_FRAME_CTRL_2 0x0c +#define MSCC_MAC_PAUSE_CFG_RX_FRAME_CTRL 0x0d +#define MSCC_MAC_PAUSE_CFG_STATE 0x0e +#define MSCC_MAC_PAUSE_CFG_MAC_ADDRESS_LSB 0x0f +#define MSCC_MAC_PAUSE_CFG_MAC_ADDRESS_MSB 0x10 +#define MSCC_MAC_STATUS_RX_LANE_STICKY_0 0x11 +#define MSCC_MAC_STATUS_RX_LANE_STICKY_1 0x12 +#define MSCC_MAC_STATUS_TX_MONITOR_STICKY 0x13 +#define MSCC_MAC_STATUS_TX_MONITOR_STICKY_MASK 0x14 +#define MSCC_MAC_STATUS_STICKY 0x15 +#define MSCC_MAC_STATUS_STICKY_MASK 0x16 +#define MSCC_MAC_STATS_32BIT_RX_HIH_CKSM_ERR_CNT 0x17 +#define MSCC_MAC_STATS_32BIT_RX_XGMII_PROT_ERR_CNT 0x18 +#define MSCC_MAC_STATS_32BIT_RX_SYMBOL_ERR_CNT 0x19 +#define MSCC_MAC_STATS_32BIT_RX_PAUSE_CNT 0x1a +#define MSCC_MAC_STATS_32BIT_RX_UNSUP_OPCODE_CNT 0x1b +#define MSCC_MAC_STATS_32BIT_RX_UC_CNT 0x1c +#define MSCC_MAC_STATS_32BIT_RX_MC_CNT 0x1d +#define MSCC_MAC_STATS_32BIT_RX_BC_CNT 0x1e +#define MSCC_MAC_STATS_32BIT_RX_CRC_ERR_CNT 0x1f +#define MSCC_MAC_STATS_32BIT_RX_UNDERSIZE_CNT 0x20 +#define MSCC_MAC_STATS_32BIT_RX_FRAGMENTS_CNT 0x21 +#define MSCC_MAC_STATS_32BIT_RX_IN_RANGE_LEN_ERR_CNT 0x22 +#define MSCC_MAC_STATS_32BIT_RX_OUT_OF_RANGE_LEN_ERR_CNT 0x23 +#define MSCC_MAC_STATS_32BIT_RX_OVERSIZE_CNT 0x24 +#define MSCC_MAC_STATS_32BIT_RX_JABBERS_CNT 0x25 +#define MSCC_MAC_STATS_32BIT_RX_SIZE64_CNT 0x26 +#define MSCC_MAC_STATS_32BIT_RX_SIZE65TO127_CNT 0x27 +#define MSCC_MAC_STATS_32BIT_RX_SIZE128TO255_CNT 0x28 +#define MSCC_MAC_STATS_32BIT_RX_SIZE256TO511_CNT 0x29 +#define MSCC_MAC_STATS_32BIT_RX_SIZE512TO1023_CNT 0x2a +#define MSCC_MAC_STATS_32BIT_RX_SIZE1024TO1518_CNT 0x2b +#define MSCC_MAC_STATS_32BIT_RX_SIZE1519TOMAX_CNT 0x2c +#define MSCC_MAC_STATS_32BIT_RX_IPG_SHRINK_CNT 0x2d +#define MSCC_MAC_STATS_32BIT_TX_PAUSE_CNT 0x2e +#define MSCC_MAC_STATS_32BIT_TX_UC_CNT 0x2f +#define MSCC_MAC_STATS_32BIT_TX_MC_CNT 0x30 +#define MSCC_MAC_STATS_32BIT_TX_BC_CNT 0x31 +#define MSCC_MAC_STATS_32BIT_TX_SIZE64_CNT 0x32 +#define MSCC_MAC_STATS_32BIT_TX_SIZE65TO127_CNT 0x33 +#define MSCC_MAC_STATS_32BIT_TX_SIZE128TO255_CNT 0x34 +#define MSCC_MAC_STATS_32BIT_TX_SIZE256TO511_CNT 0x35 +#define MSCC_MAC_STATS_32BIT_TX_SIZE512TO1023_CNT 0x36 +#define MSCC_MAC_STATS_32BIT_TX_SIZE1024TO1518_CNT 0x37 +#define MSCC_MAC_STATS_32BIT_TX_SIZE1519TOMAX_CNT 0x38 +#define MSCC_MAC_STATS_40BIT_RX_BAD_BYTES_CNT 0x39 +#define MSCC_MAC_STATS_40BIT_RX_BAD_BYTES_MSB_CNT 0x3a +#define MSCC_MAC_STATS_40BIT_RX_OK_BYTES_CNT 0x3b +#define MSCC_MAC_STATS_40BIT_RX_OK_BYTES_MSB_CNT 0x3c +#define MSCC_MAC_STATS_40BIT_RX_IN_BYTES_CNT 0x3d +#define MSCC_MAC_STATS_40BIT_RX_IN_BYTES_MSB_CNT 0x3e +#define MSCC_MAC_STATS_40BIT_TX_OK_BYTES_CNT 0x3f +#define MSCC_MAC_STATS_40BIT_TX_OK_BYTES_MSB_CNT 0x40 +#define MSCC_MAC_STATS_40BIT_TX_OUT_BYTES_CNT 0x41 +#define MSCC_MAC_STATS_40BIT_TX_OUT_BYTES_MSB_CNT 0x42 + +#define MSCC_MAC_CFG_ENA_CFG_RX_CLK_ENA BIT(0) +#define MSCC_MAC_CFG_ENA_CFG_TX_CLK_ENA BIT(4) +#define MSCC_MAC_CFG_ENA_CFG_RX_SW_RST BIT(8) +#define MSCC_MAC_CFG_ENA_CFG_TX_SW_RST BIT(12) +#define MSCC_MAC_CFG_ENA_CFG_RX_ENA BIT(16) +#define MSCC_MAC_CFG_ENA_CFG_TX_ENA BIT(20) + +#define MSCC_MAC_CFG_MODE_CFG_FORCE_CW_UPDATE_INTERVAL(x) ((x) << 20) +#define MSCC_MAC_CFG_MODE_CFG_FORCE_CW_UPDATE_INTERVAL_M GENMASK(29, 20) +#define MSCC_MAC_CFG_MODE_CFG_FORCE_CW_UPDATE BIT(16) +#define MSCC_MAC_CFG_MODE_CFG_TUNNEL_PAUSE_FRAMES BIT(14) +#define MSCC_MAC_CFG_MODE_CFG_MAC_PREAMBLE_CFG(x) ((x) << 10) +#define MSCC_MAC_CFG_MODE_CFG_MAC_PREAMBLE_CFG_M GENMASK(12, 10) +#define MSCC_MAC_CFG_MODE_CFG_MAC_IPG_CFG BIT(6) +#define MSCC_MAC_CFG_MODE_CFG_XGMII_GEN_MODE_ENA BIT(4) +#define MSCC_MAC_CFG_MODE_CFG_HIH_CRC_CHECK BIT(2) +#define MSCC_MAC_CFG_MODE_CFG_UNDERSIZED_FRAME_DROP_DIS BIT(1) +#define MSCC_MAC_CFG_MODE_CFG_DISABLE_DIC BIT(0) + +#define MSCC_MAC_CFG_MAXLEN_CFG_MAX_LEN_TAG_CHK BIT(16) +#define MSCC_MAC_CFG_MAXLEN_CFG_MAX_LEN(x) (x) +#define MSCC_MAC_CFG_MAXLEN_CFG_MAX_LEN_M GENMASK(15, 0) + +#define MSCC_MAC_CFG_TAGS_CFG_RSZ 0x4 +#define MSCC_MAC_CFG_TAGS_CFG_TAG_ID(x) ((x) << 16) +#define MSCC_MAC_CFG_TAGS_CFG_TAG_ID_M GENMASK(31, 16) +#define MSCC_MAC_CFG_TAGS_CFG_TAG_ENA BIT(4) + +#define MSCC_MAC_CFG_ADV_CHK_CFG_EXT_EOP_CHK_ENA BIT(24) +#define MSCC_MAC_CFG_ADV_CHK_CFG_EXT_SOP_CHK_ENA BIT(20) +#define MSCC_MAC_CFG_ADV_CHK_CFG_SFD_CHK_ENA BIT(16) +#define MSCC_MAC_CFG_ADV_CHK_CFG_PRM_SHK_CHK_DIS BIT(12) +#define MSCC_MAC_CFG_ADV_CHK_CFG_PRM_CHK_ENA BIT(8) +#define MSCC_MAC_CFG_ADV_CHK_CFG_OOR_ERR_ENA BIT(4) +#define MSCC_MAC_CFG_ADV_CHK_CFG_INR_ERR_ENA BIT(0) + +#define MSCC_MAC_CFG_LFS_CFG_LFS_INH_TX BIT(8) +#define MSCC_MAC_CFG_LFS_CFG_LFS_DIS_TX BIT(4) +#define MSCC_MAC_CFG_LFS_CFG_LFS_UNIDIR_ENA BIT(3) +#define MSCC_MAC_CFG_LFS_CFG_USE_LEADING_EDGE_DETECT BIT(2) +#define MSCC_MAC_CFG_LFS_CFG_SPURIOUS_Q_DIS BIT(1) +#define MSCC_MAC_CFG_LFS_CFG_LFS_MODE_ENA BIT(0) + +#define MSCC_MAC_CFG_LB_CFG_XGMII_HOST_LB_ENA BIT(4) +#define MSCC_MAC_CFG_LB_CFG_XGMII_PHY_LB_ENA BIT(0) + +#define MSCC_MAC_CFG_PKTINF_CFG_STRIP_FCS_ENA BIT(0) +#define MSCC_MAC_CFG_PKTINF_CFG_INSERT_FCS_ENA BIT(4) +#define MSCC_MAC_CFG_PKTINF_CFG_STRIP_PREAMBLE_ENA BIT(8) +#define MSCC_MAC_CFG_PKTINF_CFG_INSERT_PREAMBLE_ENA BIT(12) +#define MSCC_MAC_CFG_PKTINF_CFG_LPI_RELAY_ENA BIT(16) +#define MSCC_MAC_CFG_PKTINF_CFG_LF_RELAY_ENA BIT(20) +#define MSCC_MAC_CFG_PKTINF_CFG_RF_RELAY_ENA BIT(24) +#define MSCC_MAC_CFG_PKTINF_CFG_ENABLE_TX_PADDING BIT(25) +#define MSCC_MAC_CFG_PKTINF_CFG_ENABLE_RX_PADDING BIT(26) +#define MSCC_MAC_CFG_PKTINF_CFG_ENABLE_4BYTE_PREAMBLE BIT(27) +#define MSCC_MAC_CFG_PKTINF_CFG_MACSEC_BYPASS_NUM_PTP_STALL_CLKS(x) ((x) << 28) +#define MSCC_MAC_CFG_PKTINF_CFG_MACSEC_BYPASS_NUM_PTP_STALL_CLKS_M GENMASK(30, 28) + +#define MSCC_MAC_PAUSE_CFG_TX_FRAME_CTRL_PAUSE_VALUE(x) ((x) << 16) +#define MSCC_MAC_PAUSE_CFG_TX_FRAME_CTRL_PAUSE_VALUE_M GENMASK(31, 16) +#define MSCC_MAC_PAUSE_CFG_TX_FRAME_CTRL_WAIT_FOR_LPI_LOW BIT(12) +#define MSCC_MAC_PAUSE_CFG_TX_FRAME_CTRL_USE_PAUSE_STALL_ENA BIT(8) +#define MSCC_MAC_PAUSE_CFG_TX_FRAME_CTRL_PAUSE_REPL_MODE BIT(4) +#define MSCC_MAC_PAUSE_CFG_TX_FRAME_CTRL_PAUSE_FRC_FRAME BIT(2) +#define MSCC_MAC_PAUSE_CFG_TX_FRAME_CTRL_PAUSE_MODE(x) (x) +#define MSCC_MAC_PAUSE_CFG_TX_FRAME_CTRL_PAUSE_MODE_M GENMASK(1, 0) + +#define MSCC_MAC_PAUSE_CFG_RX_FRAME_CTRL_EARLY_PAUSE_DETECT_ENA BIT(16) +#define MSCC_MAC_PAUSE_CFG_RX_FRAME_CTRL_PRE_CRC_MODE BIT(20) +#define MSCC_MAC_PAUSE_CFG_RX_FRAME_CTRL_PAUSE_TIMER_ENA BIT(12) +#define MSCC_MAC_PAUSE_CFG_RX_FRAME_CTRL_PAUSE_REACT_ENA BIT(8) +#define MSCC_MAC_PAUSE_CFG_RX_FRAME_CTRL_PAUSE_FRAME_DROP_ENA BIT(4) +#define MSCC_MAC_PAUSE_CFG_RX_FRAME_CTRL_PAUSE_MODE BIT(0) + +#define MSCC_MAC_PAUSE_CFG_STATE_PAUSE_STATE BIT(0) +#define MSCC_MAC_PAUSE_CFG_STATE_MAC_TX_PAUSE_GEN BIT(4) + +#define MSCC_PROC_0_IP_1588_TOP_CFG_STAT_MODE_CTL 0x2 +#define MSCC_PROC_0_IP_1588_TOP_CFG_STAT_MODE_CTL_PROTOCOL_MODE(x) (x) +#define MSCC_PROC_0_IP_1588_TOP_CFG_STAT_MODE_CTL_PROTOCOL_MODE_M GENMASK(2, 0) + +#endif /* _MSCC_OCELOT_LINE_MAC_H_ */ diff --git a/drivers/net/phy/mscc_macsec.h b/drivers/net/phy/mscc_macsec.h new file mode 100644 index 000000000000..d9ab6aba7482 --- /dev/null +++ b/drivers/net/phy/mscc_macsec.h @@ -0,0 +1,266 @@ +/* SPDX-License-Identifier: (GPL-2.0 OR MIT) */ +/* + * Microsemi Ocelot Switch driver + * + * Copyright (c) 2018 Microsemi Corporation + */ + +#ifndef _MSCC_OCELOT_MACSEC_H_ +#define _MSCC_OCELOT_MACSEC_H_ + +#define MSCC_MS_MAX_FLOWS 16 + +#define CONTROL_TYPE_EGRESS 0x6 +#define CONTROL_TYPE_INGRESS 0xf +#define CONTROL_IV0 BIT(5) +#define CONTROL_IV1 BIT(6) +#define CONTROL_IV2 BIT(7) +#define CONTROL_UPDATE_SEQ BIT(13) +#define CONTROL_IV_IN_SEQ BIT(14) +#define CONTROL_ENCRYPT_AUTH BIT(15) +#define CONTROL_KEY_IN_CTX BIT(16) +#define CONTROL_CRYPTO_ALG(x) ((x) << 17) +#define CTRYPTO_ALG_AES_CTR_128 0x5 +#define CTRYPTO_ALG_AES_CTR_192 0x6 +#define CTRYPTO_ALG_AES_CTR_256 0x7 +#define CONTROL_DIGEST_TYPE(x) ((x) << 21) +#define CONTROL_AUTH_ALG(x) ((x) << 23) +#define AUTH_ALG_AES_GHAS 0x4 +#define CONTROL_AN(x) ((x) << 26) +#define CONTROL_SEQ_TYPE(x) ((x) << 28) +#define CONTROL_SEQ_MASK BIT(30) +#define CONTROL_CONTEXT_ID BIT(31) + +enum mscc_macsec_destination_ports { + MSCC_MS_PORT_COMMON = 0, + MSCC_MS_PORT_RSVD = 1, + MSCC_MS_PORT_CONTROLLED = 2, + MSCC_MS_PORT_UNCONTROLLED = 3, +}; + +enum mscc_macsec_drop_actions { + MSCC_MS_ACTION_BYPASS_CRC = 0, + MSCC_MS_ACTION_BYPASS_BAD = 1, + MSCC_MS_ACTION_DROP = 2, + MSCC_MS_ACTION_BYPASS = 3, +}; + +enum mscc_macsec_flow_types { + MSCC_MS_FLOW_BYPASS = 0, + MSCC_MS_FLOW_DROP = 1, + MSCC_MS_FLOW_INGRESS = 2, + MSCC_MS_FLOW_EGRESS = 3, +}; + +enum mscc_macsec_validate_levels { + MSCC_MS_VALIDATE_DISABLED = 0, + MSCC_MS_VALIDATE_CHECK = 1, + MSCC_MS_VALIDATE_STRICT = 2, +}; + +#define MSCC_MS_XFORM_REC(x, y) (((x) << 5) + (y)) +#define MSCC_MS_ENA_CFG 0x800 +#define MSCC_MS_FC_CFG 0x804 +#define MSCC_MS_SAM_MAC_SA_MATCH_LO(x) (0x1000 + ((x) << 4)) +#define MSCC_MS_SAM_MAC_SA_MATCH_HI(x) (0x1001 + ((x) << 4)) +#define MSCC_MS_SAM_MISC_MATCH(x) (0x1004 + ((x) << 4)) +#define MSCC_MS_SAM_MATCH_SCI_LO(x) (0x1005 + ((x) << 4)) +#define MSCC_MS_SAM_MATCH_SCI_HI(x) (0x1006 + ((x) << 4)) +#define MSCC_MS_SAM_MASK(x) (0x1007 + ((x) << 4)) +#define MSCC_MS_SAM_ENTRY_SET1 0x1808 +#define MSCC_MS_SAM_ENTRY_CLEAR1 0x180c +#define MSCC_MS_SAM_FLOW_CTRL(x) (0x1c00 + (x)) +#define MSCC_MS_SAM_CP_TAG 0x1e40 +#define MSCC_MS_SAM_NM_FLOW_NCP 0x1e51 +#define MSCC_MS_SAM_NM_FLOW_CP 0x1e52 +#define MSCC_MS_MISC_CONTROL 0x1e5f +#define MSCC_MS_COUNT_CONTROL 0x3204 +#define MSCC_MS_PARAMS2_IG_CC_CONTROL 0x3a10 +#define MSCC_MS_PARAMS2_IG_CP_TAG 0x3a14 +#define MSCC_MS_VLAN_MTU_CHECK(x) (0x3c40 + (x)) +#define MSCC_MS_NON_VLAN_MTU_CHECK 0x3c48 +#define MSCC_MS_PP_CTRL 0x3c4b +#define MSCC_MS_STATUS_CONTEXT_CTRL 0x3d02 +#define MSCC_MS_INTR_CTRL_STATUS 0x3d04 +#define MSCC_MS_BLOCK_CTX_UPDATE 0x3d0c +#define MSCC_MS_AIC_CTRL 0x3e02 + +/* MACSEC_ENA_CFG */ +#define MSCC_MS_ENA_CFG_CLK_ENA BIT(0) +#define MSCC_MS_ENA_CFG_SW_RST BIT(1) +#define MSCC_MS_ENA_CFG_MACSEC_BYPASS_ENA BIT(8) +#define MSCC_MS_ENA_CFG_MACSEC_ENA BIT(9) +#define MSCC_MS_ENA_CFG_MACSEC_SPEED_MODE(x) ((x) << 10) +#define MSCC_MS_ENA_CFG_MACSEC_SPEED_MODE_M GENMASK(12, 10) + +/* MACSEC_FC_CFG */ +#define MSCC_MS_FC_CFG_FCBUF_ENA BIT(0) +#define MSCC_MS_FC_CFG_USE_PKT_EXPANSION_INDICATION BIT(1) +#define MSCC_MS_FC_CFG_LOW_THRESH(x) ((x) << 4) +#define MSCC_MS_FC_CFG_LOW_THRESH_M GENMASK(7, 4) +#define MSCC_MS_FC_CFG_HIGH_THRESH(x) ((x) << 8) +#define MSCC_MS_FC_CFG_HIGH_THRESH_M GENMASK(11, 8) +#define MSCC_MS_FC_CFG_LOW_BYTES_VAL(x) ((x) << 12) +#define MSCC_MS_FC_CFG_LOW_BYTES_VAL_M GENMASK(14, 12) +#define MSCC_MS_FC_CFG_HIGH_BYTES_VAL(x) ((x) << 16) +#define MSCC_MS_FC_CFG_HIGH_BYTES_VAL_M GENMASK(18, 16) + +/* MSCC_MS_SAM_MAC_SA_MATCH_HI */ +#define MSCC_MS_SAM_MAC_SA_MATCH_HI_ETYPE(x) ((x) << 16) +#define MSCC_MS_SAM_MAC_SA_MATCH_HI_ETYPE_M GENMASK(31, 16) + +/* MACSEC_SAM_MISC_MATCH */ +#define MSCC_MS_SAM_MISC_MATCH_VLAN_VALID BIT(0) +#define MSCC_MS_SAM_MISC_MATCH_QINQ_FOUND BIT(1) +#define MSCC_MS_SAM_MISC_MATCH_STAG_VALID BIT(2) +#define MSCC_MS_SAM_MISC_MATCH_QTAG_VALID BIT(3) +#define MSCC_MS_SAM_MISC_MATCH_VLAN_UP(x) ((x) << 4) +#define MSCC_MS_SAM_MISC_MATCH_VLAN_UP_M GENMASK(6, 4) +#define MSCC_MS_SAM_MISC_MATCH_CONTROL_PACKET BIT(7) +#define MSCC_MS_SAM_MISC_MATCH_UNTAGGED BIT(8) +#define MSCC_MS_SAM_MISC_MATCH_TAGGED BIT(9) +#define MSCC_MS_SAM_MISC_MATCH_BAD_TAG BIT(10) +#define MSCC_MS_SAM_MISC_MATCH_KAY_TAG BIT(11) +#define MSCC_MS_SAM_MISC_MATCH_SOURCE_PORT(x) ((x) << 12) +#define MSCC_MS_SAM_MISC_MATCH_SOURCE_PORT_M GENMASK(13, 12) +#define MSCC_MS_SAM_MISC_MATCH_PRIORITY(x) ((x) << 16) +#define MSCC_MS_SAM_MISC_MATCH_PRIORITY_M GENMASK(19, 16) +#define MSCC_MS_SAM_MISC_MATCH_AN(x) ((x) << 24) +#define MSCC_MS_SAM_MISC_MATCH_TCI(x) ((x) << 26) + +/* MACSEC_SAM_MASK */ +#define MSCC_MS_SAM_MASK_MAC_SA_MASK(x) (x) +#define MSCC_MS_SAM_MASK_MAC_SA_MASK_M GENMASK(5, 0) +#define MSCC_MS_SAM_MASK_MAC_DA_MASK(x) ((x) << 6) +#define MSCC_MS_SAM_MASK_MAC_DA_MASK_M GENMASK(11, 6) +#define MSCC_MS_SAM_MASK_MAC_ETYPE_MASK BIT(12) +#define MSCC_MS_SAM_MASK_VLAN_VLD_MASK BIT(13) +#define MSCC_MS_SAM_MASK_QINQ_FOUND_MASK BIT(14) +#define MSCC_MS_SAM_MASK_STAG_VLD_MASK BIT(15) +#define MSCC_MS_SAM_MASK_QTAG_VLD_MASK BIT(16) +#define MSCC_MS_SAM_MASK_VLAN_UP_MASK BIT(17) +#define MSCC_MS_SAM_MASK_VLAN_ID_MASK BIT(18) +#define MSCC_MS_SAM_MASK_SOURCE_PORT_MASK BIT(19) +#define MSCC_MS_SAM_MASK_CTL_PACKET_MASK BIT(20) +#define MSCC_MS_SAM_MASK_VLAN_UP_INNER_MASK BIT(21) +#define MSCC_MS_SAM_MASK_VLAN_ID_INNER_MASK BIT(22) +#define MSCC_MS_SAM_MASK_SCI_MASK BIT(23) +#define MSCC_MS_SAM_MASK_AN_MASK(x) ((x) << 24) +#define MSCC_MS_SAM_MASK_TCI_MASK(x) ((x) << 26) + +/* MACSEC_SAM_FLOW_CTRL_EGR */ +#define MSCC_MS_SAM_FLOW_CTRL_FLOW_TYPE(x) (x) +#define MSCC_MS_SAM_FLOW_CTRL_FLOW_TYPE_M GENMASK(1, 0) +#define MSCC_MS_SAM_FLOW_CTRL_DEST_PORT(x) ((x) << 2) +#define MSCC_MS_SAM_FLOW_CTRL_DEST_PORT_M GENMASK(3, 2) +#define MSCC_MS_SAM_FLOW_CTRL_RESV_4 BIT(4) +#define MSCC_MS_SAM_FLOW_CTRL_FLOW_CRYPT_AUTH BIT(5) +#define MSCC_MS_SAM_FLOW_CTRL_DROP_ACTION(x) ((x) << 6) +#define MSCC_MS_SAM_FLOW_CTRL_DROP_ACTION_M GENMASK(7, 6) +#define MSCC_MS_SAM_FLOW_CTRL_RESV_15_TO_8(x) ((x) << 8) +#define MSCC_MS_SAM_FLOW_CTRL_RESV_15_TO_8_M GENMASK(15, 8) +#define MSCC_MS_SAM_FLOW_CTRL_PROTECT_FRAME BIT(16) +#define MSCC_MS_SAM_FLOW_CTRL_REPLAY_PROTECT BIT(16) +#define MSCC_MS_SAM_FLOW_CTRL_SA_IN_USE BIT(17) +#define MSCC_MS_SAM_FLOW_CTRL_INCLUDE_SCI BIT(18) +#define MSCC_MS_SAM_FLOW_CTRL_USE_ES BIT(19) +#define MSCC_MS_SAM_FLOW_CTRL_USE_SCB BIT(20) +#define MSCC_MS_SAM_FLOW_CTRL_VALIDATE_FRAMES(x) ((x) << 19) +#define MSCC_MS_SAM_FLOW_CTRL_TAG_BYPASS_SIZE(x) ((x) << 21) +#define MSCC_MS_SAM_FLOW_CTRL_TAG_BYPASS_SIZE_M GENMASK(22, 21) +#define MSCC_MS_SAM_FLOW_CTRL_RESV_23 BIT(23) +#define MSCC_MS_SAM_FLOW_CTRL_CONFIDENTIALITY_OFFSET(x) ((x) << 24) +#define MSCC_MS_SAM_FLOW_CTRL_CONFIDENTIALITY_OFFSET_M GENMASK(30, 24) +#define MSCC_MS_SAM_FLOW_CTRL_CONF_PROTECT BIT(31) + +/* MACSEC_SAM_CP_TAG */ +#define MSCC_MS_SAM_CP_TAG_MAP_TBL(x) (x) +#define MSCC_MS_SAM_CP_TAG_MAP_TBL_M GENMASK(23, 0) +#define MSCC_MS_SAM_CP_TAG_DEF_UP(x) ((x) << 24) +#define MSCC_MS_SAM_CP_TAG_DEF_UP_M GENMASK(26, 24) +#define MSCC_MS_SAM_CP_TAG_STAG_UP_EN BIT(27) +#define MSCC_MS_SAM_CP_TAG_QTAG_UP_EN BIT(28) +#define MSCC_MS_SAM_CP_TAG_PARSE_QINQ BIT(29) +#define MSCC_MS_SAM_CP_TAG_PARSE_STAG BIT(30) +#define MSCC_MS_SAM_CP_TAG_PARSE_QTAG BIT(31) + +/* MACSEC_SAM_NM_FLOW_NCP */ +#define MSCC_MS_SAM_NM_FLOW_NCP_UNTAGGED_FLOW_TYPE(x) (x) +#define MSCC_MS_SAM_NM_FLOW_NCP_UNTAGGED_DEST_PORT(x) ((x) << 2) +#define MSCC_MS_SAM_NM_FLOW_NCP_UNTAGGED_DROP_ACTION(x) ((x) << 6) +#define MSCC_MS_SAM_NM_FLOW_NCP_TAGGED_FLOW_TYPE(x) ((x) << 8) +#define MSCC_MS_SAM_NM_FLOW_NCP_TAGGED_DEST_PORT(x) ((x) << 10) +#define MSCC_MS_SAM_NM_FLOW_NCP_TAGGED_DROP_ACTION(x) ((x) << 14) +#define MSCC_MS_SAM_NM_FLOW_NCP_BADTAG_FLOW_TYPE(x) ((x) << 16) +#define MSCC_MS_SAM_NM_FLOW_NCP_BADTAG_DEST_PORT(x) ((x) << 18) +#define MSCC_MS_SAM_NM_FLOW_NCP_BADTAG_DROP_ACTION(x) ((x) << 22) +#define MSCC_MS_SAM_NM_FLOW_NCP_KAY_FLOW_TYPE(x) ((x) << 24) +#define MSCC_MS_SAM_NM_FLOW_NCP_KAY_DEST_PORT(x) ((x) << 26) +#define MSCC_MS_SAM_NM_FLOW_NCP_KAY_DROP_ACTION(x) ((x) << 30) + +/* MACSEC_SAM_NM_FLOW_CP */ +#define MSCC_MS_SAM_NM_FLOW_CP_UNTAGGED_FLOW_TYPE(x) (x) +#define MSCC_MS_SAM_NM_FLOW_CP_UNTAGGED_DEST_PORT(x) ((x) << 2) +#define MSCC_MS_SAM_NM_FLOW_CP_UNTAGGED_DROP_ACTION(x) ((x) << 6) +#define MSCC_MS_SAM_NM_FLOW_CP_TAGGED_FLOW_TYPE(x) ((x) << 8) +#define MSCC_MS_SAM_NM_FLOW_CP_TAGGED_DEST_PORT(x) ((x) << 10) +#define MSCC_MS_SAM_NM_FLOW_CP_TAGGED_DROP_ACTION(x) ((x) << 14) +#define MSCC_MS_SAM_NM_FLOW_CP_BADTAG_FLOW_TYPE(x) ((x) << 16) +#define MSCC_MS_SAM_NM_FLOW_CP_BADTAG_DEST_PORT(x) ((x) << 18) +#define MSCC_MS_SAM_NM_FLOW_CP_BADTAG_DROP_ACTION(x) ((x) << 22) +#define MSCC_MS_SAM_NM_FLOW_CP_KAY_FLOW_TYPE(x) ((x) << 24) +#define MSCC_MS_SAM_NM_FLOW_CP_KAY_DEST_PORT(x) ((x) << 26) +#define MSCC_MS_SAM_NM_FLOW_CP_KAY_DROP_ACTION(x) ((x) << 30) + +/* MACSEC_MISC_CONTROL */ +#define MSCC_MS_MISC_CONTROL_MC_LATENCY_FIX(x) (x) +#define MSCC_MS_MISC_CONTROL_MC_LATENCY_FIX_M GENMASK(5, 0) +#define MSCC_MS_MISC_CONTROL_STATIC_BYPASS BIT(8) +#define MSCC_MS_MISC_CONTROL_NM_MACSEC_EN BIT(9) +#define MSCC_MS_MISC_CONTROL_VALIDATE_FRAMES(x) ((x) << 10) +#define MSCC_MS_MISC_CONTROL_VALIDATE_FRAMES_M GENMASK(11, 10) +#define MSCC_MS_MISC_CONTROL_XFORM_REC_SIZE(x) ((x) << 24) +#define MSCC_MS_MISC_CONTROL_XFORM_REC_SIZE_M GENMASK(25, 24) + +/* MACSEC_COUNT_CONTROL */ +#define MSCC_MS_COUNT_CONTROL_RESET_ALL BIT(0) +#define MSCC_MS_COUNT_CONTROL_DEBUG_ACCESS BIT(1) +#define MSCC_MS_COUNT_CONTROL_SATURATE_CNTRS BIT(2) +#define MSCC_MS_COUNT_CONTROL_AUTO_CNTR_RESET BIT(3) + +/* MACSEC_PARAMS2_IG_CC_CONTROL */ +#define MSCC_MS_PARAMS2_IG_CC_CONTROL_NON_MATCH_CTRL_ACT BIT(14) +#define MSCC_MS_PARAMS2_IG_CC_CONTROL_NON_MATCH_ACT BIT(15) + +/* MACSEC_PARAMS2_IG_CP_TAG */ +#define MSCC_MS_PARAMS2_IG_CP_TAG_MAP_TBL(x) (x) +#define MSCC_MS_PARAMS2_IG_CP_TAG_MAP_TBL_M GENMASK(23, 0) +#define MSCC_MS_PARAMS2_IG_CP_TAG_DEF_UP(x) ((x) << 24) +#define MSCC_MS_PARAMS2_IG_CP_TAG_DEF_UP_M GENMASK(26, 24) +#define MSCC_MS_PARAMS2_IG_CP_TAG_STAG_UP_EN BIT(27) +#define MSCC_MS_PARAMS2_IG_CP_TAG_QTAG_UP_EN BIT(28) +#define MSCC_MS_PARAMS2_IG_CP_TAG_PARSE_QINQ BIT(29) +#define MSCC_MS_PARAMS2_IG_CP_TAG_PARSE_STAG BIT(30) +#define MSCC_MS_PARAMS2_IG_CP_TAG_PARSE_QTAG BIT(31) + +/* MACSEC_VLAN_MTU_CHECK */ +#define MSCC_MS_VLAN_MTU_CHECK_MTU_COMPARE(x) (x) +#define MSCC_MS_VLAN_MTU_CHECK_MTU_COMPARE_M GENMASK(14, 0) +#define MSCC_MS_VLAN_MTU_CHECK_MTU_COMP_DROP BIT(15) + +/* MACSEC_NON_VLAN_MTU_CHECK */ +#define MSCC_MS_NON_VLAN_MTU_CHECK_NV_MTU_COMPARE(x) (x) +#define MSCC_MS_NON_VLAN_MTU_CHECK_NV_MTU_COMPARE_M GENMASK(14, 0) +#define MSCC_MS_NON_VLAN_MTU_CHECK_NV_MTU_COMP_DROP BIT(15) + +/* MACSEC_PP_CTRL */ +#define MSCC_MS_PP_CTRL_MACSEC_OCTET_INCR_MODE BIT(0) + +/* MACSEC_INTR_CTRL_STATUS */ +#define MSCC_MS_INTR_CTRL_STATUS_INTR_CLR_STATUS(x) (x) +#define MSCC_MS_INTR_CTRL_STATUS_INTR_CLR_STATUS_M GENMASK(15, 0) +#define MSCC_MS_INTR_CTRL_STATUS_INTR_ENABLE(x) ((x) << 16) +#define MSCC_MS_INTR_CTRL_STATUS_INTR_ENABLE_M GENMASK(31, 16) +#define MACSEC_INTR_CTRL_STATUS_ROLLOVER BIT(5) + +#endif diff --git a/drivers/net/phy/phy-core.c b/drivers/net/phy/phy-core.c index 769a076514b0..a4d2d59fceca 100644 --- a/drivers/net/phy/phy-core.c +++ b/drivers/net/phy/phy-core.c @@ -387,7 +387,7 @@ int __phy_read_mmd(struct phy_device *phydev, int devad, u32 regnum) if (regnum > (u16)~0 || devad > 32) return -EINVAL; - if (phydev->drv->read_mmd) { + if (phydev->drv && phydev->drv->read_mmd) { val = phydev->drv->read_mmd(phydev, devad, regnum); } else if (phydev->is_c45) { u32 addr = MII_ADDR_C45 | (devad << 16) | (regnum & 0xffff); @@ -444,7 +444,7 @@ int __phy_write_mmd(struct phy_device *phydev, int devad, u32 regnum, u16 val) if (regnum > (u16)~0 || devad > 32) return -EINVAL; - if (phydev->drv->write_mmd) { + if (phydev->drv && phydev->drv->write_mmd) { ret = phydev->drv->write_mmd(phydev, devad, regnum, val); } else if (phydev->is_c45) { u32 addr = MII_ADDR_C45 | (devad << 16) | (regnum & 0xffff); diff --git a/drivers/net/phy/phy.c b/drivers/net/phy/phy.c index 541ed01496bf..d76e038cf2cb 100644 --- a/drivers/net/phy/phy.c +++ b/drivers/net/phy/phy.c @@ -432,6 +432,31 @@ int phy_mii_ioctl(struct phy_device *phydev, struct ifreq *ifr, int cmd) } EXPORT_SYMBOL(phy_mii_ioctl); +/** + * phy_do_ioctl - generic ndo_do_ioctl implementation + * @dev: the net_device struct + * @ifr: &struct ifreq for socket ioctl's + * @cmd: ioctl cmd to execute + */ +int phy_do_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd) +{ + if (!dev->phydev) + return -ENODEV; + + return phy_mii_ioctl(dev->phydev, ifr, cmd); +} +EXPORT_SYMBOL(phy_do_ioctl); + +/* same as phy_do_ioctl, but ensures that net_device is running */ +int phy_do_ioctl_running(struct net_device *dev, struct ifreq *ifr, int cmd) +{ + if (!netif_running(dev)) + return -ENODEV; + + return phy_do_ioctl(dev, ifr, cmd); +} +EXPORT_SYMBOL(phy_do_ioctl_running); + void phy_queue_state_machine(struct phy_device *phydev, unsigned long jiffies) { mod_delayed_work(system_power_efficient_wq, &phydev->state_queue, diff --git a/drivers/net/phy/phy_device.c b/drivers/net/phy/phy_device.c index e5dc9f87f495..6a5056e0ae77 100644 --- a/drivers/net/phy/phy_device.c +++ b/drivers/net/phy/phy_device.c @@ -1107,9 +1107,8 @@ void phy_attached_info(struct phy_device *phydev) EXPORT_SYMBOL(phy_attached_info); #define ATTACHED_FMT "attached PHY driver [%s] (mii_bus:phy_addr=%s, irq=%s)" -void phy_attached_print(struct phy_device *phydev, const char *fmt, ...) +char *phy_attached_info_irq(struct phy_device *phydev) { - const char *drv_name = phydev->drv ? phydev->drv->name : "unbound"; char *irq_str; char irq_num[8]; @@ -1126,6 +1125,14 @@ void phy_attached_print(struct phy_device *phydev, const char *fmt, ...) break; } + return kasprintf(GFP_KERNEL, "%s", irq_str); +} +EXPORT_SYMBOL(phy_attached_info_irq); + +void phy_attached_print(struct phy_device *phydev, const char *fmt, ...) +{ + const char *drv_name = phydev->drv ? phydev->drv->name : "unbound"; + char *irq_str = phy_attached_info_irq(phydev); if (!fmt) { phydev_info(phydev, ATTACHED_FMT "\n", @@ -1142,6 +1149,7 @@ void phy_attached_print(struct phy_device *phydev, const char *fmt, ...) vprintk(fmt, ap); va_end(ap); } + kfree(irq_str); } EXPORT_SYMBOL(phy_attached_print); diff --git a/drivers/net/phy/phylink.c b/drivers/net/phy/phylink.c index af914a8842bd..70b9a143db84 100644 --- a/drivers/net/phy/phylink.c +++ b/drivers/net/phy/phylink.c @@ -298,6 +298,7 @@ static int phylink_parse_mode(struct phylink *pl, struct fwnode_handle *fwnode) phylink_set(pl->supported, 2500baseX_Full); break; + case PHY_INTERFACE_MODE_USXGMII: case PHY_INTERFACE_MODE_10GKR: case PHY_INTERFACE_MODE_10GBASER: phylink_set(pl->supported, 10baseT_Half); @@ -307,6 +308,10 @@ static int phylink_parse_mode(struct phylink *pl, struct fwnode_handle *fwnode) phylink_set(pl->supported, 1000baseT_Half); phylink_set(pl->supported, 1000baseT_Full); phylink_set(pl->supported, 1000baseX_Full); + phylink_set(pl->supported, 2500baseT_Full); + phylink_set(pl->supported, 2500baseX_Full); + phylink_set(pl->supported, 5000baseT_Full); + phylink_set(pl->supported, 10000baseT_Full); phylink_set(pl->supported, 10000baseKR_Full); phylink_set(pl->supported, 10000baseCR_Full); phylink_set(pl->supported, 10000baseSR_Full); @@ -572,6 +577,9 @@ static int phylink_register_sfp(struct phylink *pl, struct sfp_bus *bus; int ret; + if (!fwnode) + return 0; + bus = sfp_bus_find_fwnode(fwnode); if (IS_ERR(bus)) { ret = PTR_ERR(bus); @@ -723,6 +731,7 @@ static int phylink_bringup_phy(struct phylink *pl, struct phy_device *phy, { struct phylink_link_state config; __ETHTOOL_DECLARE_LINK_MODE_MASK(supported); + char *irq_str; int ret; /* @@ -758,9 +767,11 @@ static int phylink_bringup_phy(struct phylink *pl, struct phy_device *phy, phy->phylink = pl; phy->phy_link_change = phylink_phy_change; + irq_str = phy_attached_info_irq(phy); phylink_info(pl, - "PHY [%s] driver [%s]\n", dev_name(&phy->mdio.dev), - phy->drv->name); + "PHY [%s] driver [%s] (irq=%s)\n", + dev_name(&phy->mdio.dev), phy->drv->name, irq_str); + kfree(irq_str); mutex_lock(&phy->lock); mutex_lock(&pl->state_mutex); diff --git a/drivers/net/ppp/pptp.c b/drivers/net/ppp/pptp.c index e1fabb3e3246..acccb747aeda 100644 --- a/drivers/net/ppp/pptp.c +++ b/drivers/net/ppp/pptp.c @@ -155,7 +155,7 @@ static int pptp_xmit(struct ppp_channel *chan, struct sk_buff *skb) opt->dst_addr.sin_addr.s_addr, opt->src_addr.sin_addr.s_addr, 0, 0, IPPROTO_GRE, - RT_TOS(0), 0); + RT_TOS(0), sk->sk_bound_dev_if); if (IS_ERR(rt)) goto tx_error; @@ -444,7 +444,8 @@ static int pptp_connect(struct socket *sock, struct sockaddr *uservaddr, opt->dst_addr.sin_addr.s_addr, opt->src_addr.sin_addr.s_addr, 0, 0, - IPPROTO_GRE, RT_CONN_FLAGS(sk), 0); + IPPROTO_GRE, RT_CONN_FLAGS(sk), + sk->sk_bound_dev_if); if (IS_ERR(rt)) { error = -EHOSTUNREACH; goto end; diff --git a/drivers/net/tap.c b/drivers/net/tap.c index a6d63665ad03..1f4bdd94407a 100644 --- a/drivers/net/tap.c +++ b/drivers/net/tap.c @@ -341,6 +341,7 @@ rx_handler_result_t tap_handle_frame(struct sk_buff **pskb) features |= tap->tap_features; if (netif_needs_gso(skb, features)) { struct sk_buff *segs = __skb_gso_segment(skb, features, false); + struct sk_buff *next; if (IS_ERR(segs)) goto drop; @@ -352,16 +353,13 @@ rx_handler_result_t tap_handle_frame(struct sk_buff **pskb) } consume_skb(skb); - while (segs) { - struct sk_buff *nskb = segs->next; - - segs->next = NULL; - if (ptr_ring_produce(&q->ring, segs)) { - kfree_skb(segs); - kfree_skb_list(nskb); + skb_list_walk_safe(segs, skb, next) { + skb_mark_not_on_list(skb); + if (ptr_ring_produce(&q->ring, skb)) { + kfree_skb(skb); + kfree_skb_list(next); break; } - segs = nskb; } } else { /* If we receive a partial checksum and the tap side diff --git a/drivers/net/usb/lan78xx.c b/drivers/net/usb/lan78xx.c index f1a63497235e..c391f2521ba1 100644 --- a/drivers/net/usb/lan78xx.c +++ b/drivers/net/usb/lan78xx.c @@ -2724,11 +2724,6 @@ static int lan78xx_stop(struct net_device *net) return 0; } -static int lan78xx_linearize(struct sk_buff *skb) -{ - return skb_linearize(skb); -} - static struct sk_buff *lan78xx_tx_prep(struct lan78xx_net *dev, struct sk_buff *skb, gfp_t flags) { @@ -2740,8 +2735,10 @@ static struct sk_buff *lan78xx_tx_prep(struct lan78xx_net *dev, return NULL; } - if (lan78xx_linearize(skb) < 0) + if (skb_linearize(skb)) { + dev_kfree_skb_any(skb); return NULL; + } tx_cmd_a = (u32)(skb->len & TX_CMD_A_LEN_MASK_) | TX_CMD_A_FCS_; @@ -3753,6 +3750,7 @@ static int lan78xx_probe(struct usb_interface *intf, /* MTU range: 68 - 9000 */ netdev->max_mtu = MAX_SINGLE_PACKET_SIZE; + netif_set_gso_max_size(netdev, MAX_SINGLE_PACKET_SIZE - MAX_HEADER); dev->ep_blkin = (intf->cur_altsetting)->endpoint + 0; dev->ep_blkout = (intf->cur_altsetting)->endpoint + 1; diff --git a/drivers/net/usb/qmi_wwan.c b/drivers/net/usb/qmi_wwan.c index 4196c0e32740..9485c8d1de8a 100644 --- a/drivers/net/usb/qmi_wwan.c +++ b/drivers/net/usb/qmi_wwan.c @@ -1062,6 +1062,7 @@ static const struct usb_device_id products[] = { {QMI_QUIRK_QUECTEL_DYNCFG(0x2c7c, 0x0125)}, /* Quectel EC25, EC20 R2.0 Mini PCIe */ {QMI_QUIRK_QUECTEL_DYNCFG(0x2c7c, 0x0306)}, /* Quectel EP06/EG06/EM06 */ {QMI_QUIRK_QUECTEL_DYNCFG(0x2c7c, 0x0512)}, /* Quectel EG12/EM12 */ + {QMI_QUIRK_QUECTEL_DYNCFG(0x2c7c, 0x0800)}, /* Quectel RM500Q-GL */ /* 3. Combined interface devices matching on interface number */ {QMI_FIXED_INTF(0x0408, 0xea42, 4)}, /* Yota / Megafon M100-1 */ diff --git a/drivers/net/usb/r8152.c b/drivers/net/usb/r8152.c index 9ec1da429514..36051288034a 100644 --- a/drivers/net/usb/r8152.c +++ b/drivers/net/usb/r8152.c @@ -1897,8 +1897,8 @@ static void r8152_csum_workaround(struct r8152 *tp, struct sk_buff *skb, { if (skb_shinfo(skb)->gso_size) { netdev_features_t features = tp->netdev->features; + struct sk_buff *segs, *seg, *next; struct sk_buff_head seg_list; - struct sk_buff *segs, *nskb; features &= ~(NETIF_F_SG | NETIF_F_IPV6_CSUM | NETIF_F_TSO6); segs = skb_gso_segment(skb, features); @@ -1907,12 +1907,10 @@ static void r8152_csum_workaround(struct r8152 *tp, struct sk_buff *skb, __skb_queue_head_init(&seg_list); - do { - nskb = segs; - segs = segs->next; - nskb->next = NULL; - __skb_queue_tail(&seg_list, nskb); - } while (segs); + skb_list_walk_safe(segs, seg, next) { + skb_mark_not_on_list(seg); + __skb_queue_tail(&seg_list, seg); + } skb_queue_splice(&seg_list, list); dev_kfree_skb(skb); @@ -6597,6 +6595,9 @@ static int rtl8152_probe(struct usb_interface *intf, return -ENODEV; } + if (intf->cur_altsetting->desc.bNumEndpoints < 3) + return -ENODEV; + usb_reset_device(udev); netdev = alloc_etherdev(sizeof(struct r8152)); if (!netdev) { diff --git a/drivers/net/vmxnet3/vmxnet3_ethtool.c b/drivers/net/vmxnet3/vmxnet3_ethtool.c index 0a38c76688ab..1e4b9ba70983 100644 --- a/drivers/net/vmxnet3/vmxnet3_ethtool.c +++ b/drivers/net/vmxnet3/vmxnet3_ethtool.c @@ -555,10 +555,8 @@ vmxnet3_set_ringparam(struct net_device *netdev, } if (VMXNET3_VERSION_GE_3(adapter)) { - if (param->rx_mini_pending < 0 || - param->rx_mini_pending > VMXNET3_RXDATA_DESC_MAX_SIZE) { + if (param->rx_mini_pending > VMXNET3_RXDATA_DESC_MAX_SIZE) return -EINVAL; - } } else if (param->rx_mini_pending != 0) { return -EINVAL; } diff --git a/drivers/net/vxlan.c b/drivers/net/vxlan.c index e95e6585ab82..d3b08b76b1ec 100644 --- a/drivers/net/vxlan.c +++ b/drivers/net/vxlan.c @@ -2542,7 +2542,7 @@ static void vxlan_xmit_one(struct sk_buff *skb, struct net_device *dev, ndst = &rt->dst; skb_tunnel_check_pmtu(skb, ndst, VXLAN_HEADROOM); - tos = ip_tunnel_ecn_encap(tos, old_iph, skb); + tos = ip_tunnel_ecn_encap(RT_TOS(tos), old_iph, skb); ttl = ttl ? : ip4_dst_hoplimit(&rt->dst); err = vxlan_build_skb(skb, ndst, sizeof(struct iphdr), vni, md, flags, udp_sum); @@ -2582,7 +2582,7 @@ static void vxlan_xmit_one(struct sk_buff *skb, struct net_device *dev, skb_tunnel_check_pmtu(skb, ndst, VXLAN6_HEADROOM); - tos = ip_tunnel_ecn_encap(tos, old_iph, skb); + tos = ip_tunnel_ecn_encap(RT_TOS(tos), old_iph, skb); ttl = ttl ? : ip6_dst_hoplimit(ndst); skb_scrub_packet(skb, xnet); err = vxlan_build_skb(skb, ndst, sizeof(struct ipv6hdr), diff --git a/drivers/net/wan/Kconfig b/drivers/net/wan/Kconfig index dd1a147f2971..4530840e15ef 100644 --- a/drivers/net/wan/Kconfig +++ b/drivers/net/wan/Kconfig @@ -315,7 +315,8 @@ config DSCC4_PCI_RST config IXP4XX_HSS tristate "Intel IXP4xx HSS (synchronous serial port) support" - depends on HDLC && ARM && ARCH_IXP4XX && IXP4XX_NPE && IXP4XX_QMGR + depends on HDLC && IXP4XX_NPE && IXP4XX_QMGR + depends on ARCH_IXP4XX help Say Y here if you want to use built-in HSS ports on IXP4xx processor. diff --git a/drivers/net/wan/fsl_ucc_hdlc.c b/drivers/net/wan/fsl_ucc_hdlc.c index c28f8409067e..3998cac49d7f 100644 --- a/drivers/net/wan/fsl_ucc_hdlc.c +++ b/drivers/net/wan/fsl_ucc_hdlc.c @@ -73,7 +73,7 @@ static struct ucc_tdm_info utdm_primary_info = { }, }; -static struct ucc_tdm_info utdm_info[MAX_HDLC_NUM]; +static struct ucc_tdm_info utdm_info[UCC_MAX_NUM]; static int uhdlc_init(struct ucc_hdlc_private *priv) { diff --git a/drivers/net/wan/hdlc_x25.c b/drivers/net/wan/hdlc_x25.c index 5643675ff724..c84536b03aa8 100644 --- a/drivers/net/wan/hdlc_x25.c +++ b/drivers/net/wan/hdlc_x25.c @@ -21,8 +21,17 @@ #include <linux/skbuff.h> #include <net/x25device.h> +struct x25_state { + x25_hdlc_proto settings; +}; + static int x25_ioctl(struct net_device *dev, struct ifreq *ifr); +static struct x25_state *state(hdlc_device *hdlc) +{ + return hdlc->state; +} + /* These functions are callbacks called by LAPB layer */ static void x25_connect_disconnect(struct net_device *dev, int reason, int code) @@ -62,11 +71,12 @@ static int x25_data_indication(struct net_device *dev, struct sk_buff *skb) { unsigned char *ptr; - skb_push(skb, 1); - if (skb_cow(skb, 1)) return NET_RX_DROP; + skb_push(skb, 1); + skb_reset_network_header(skb); + ptr = skb->data; *ptr = X25_IFACE_DATA; @@ -79,6 +89,13 @@ static int x25_data_indication(struct net_device *dev, struct sk_buff *skb) static void x25_data_transmit(struct net_device *dev, struct sk_buff *skb) { hdlc_device *hdlc = dev_to_hdlc(dev); + + skb_reset_network_header(skb); + skb->protocol = hdlc_type_trans(skb, dev); + + if (dev_nit_active(dev)) + dev_queue_xmit_nit(skb, dev); + hdlc->xmit(skb, dev); /* Ignore return value :-( */ } @@ -93,6 +110,7 @@ static netdev_tx_t x25_xmit(struct sk_buff *skb, struct net_device *dev) switch (skb->data[0]) { case X25_IFACE_DATA: /* Data to be transmitted */ skb_pull(skb, 1); + skb_reset_network_header(skb); if ((result = lapb_data_request(dev, skb)) != LAPB_OK) dev_kfree_skb(skb); return NETDEV_TX_OK; @@ -131,7 +149,6 @@ static netdev_tx_t x25_xmit(struct sk_buff *skb, struct net_device *dev) static int x25_open(struct net_device *dev) { - int result; static const struct lapb_register_struct cb = { .connect_confirmation = x25_connected, .connect_indication = x25_connected, @@ -140,10 +157,33 @@ static int x25_open(struct net_device *dev) .data_indication = x25_data_indication, .data_transmit = x25_data_transmit, }; + hdlc_device *hdlc = dev_to_hdlc(dev); + struct lapb_parms_struct params; + int result; result = lapb_register(dev, &cb); if (result != LAPB_OK) return result; + + result = lapb_getparms(dev, ¶ms); + if (result != LAPB_OK) + return result; + + if (state(hdlc)->settings.dce) + params.mode = params.mode | LAPB_DCE; + + if (state(hdlc)->settings.modulo == 128) + params.mode = params.mode | LAPB_EXTENDED; + + params.window = state(hdlc)->settings.window; + params.t1 = state(hdlc)->settings.t1; + params.t2 = state(hdlc)->settings.t2; + params.n2 = state(hdlc)->settings.n2; + + result = lapb_setparms(dev, ¶ms); + if (result != LAPB_OK) + return result; + return 0; } @@ -186,7 +226,10 @@ static struct hdlc_proto proto = { static int x25_ioctl(struct net_device *dev, struct ifreq *ifr) { + x25_hdlc_proto __user *x25_s = ifr->ifr_settings.ifs_ifsu.x25; + const size_t size = sizeof(x25_hdlc_proto); hdlc_device *hdlc = dev_to_hdlc(dev); + x25_hdlc_proto new_settings; int result; switch (ifr->ifr_settings.type) { @@ -194,7 +237,13 @@ static int x25_ioctl(struct net_device *dev, struct ifreq *ifr) if (dev_to_hdlc(dev)->proto != &proto) return -EINVAL; ifr->ifr_settings.type = IF_PROTO_X25; - return 0; /* return protocol only, no settable parameters */ + if (ifr->ifr_settings.size < size) { + ifr->ifr_settings.size = size; /* data size wanted */ + return -ENOBUFS; + } + if (copy_to_user(x25_s, &state(hdlc)->settings, size)) + return -EFAULT; + return 0; case IF_PROTO_X25: if (!capable(CAP_NET_ADMIN)) @@ -203,12 +252,46 @@ static int x25_ioctl(struct net_device *dev, struct ifreq *ifr) if (dev->flags & IFF_UP) return -EBUSY; + /* backward compatibility */ + if (ifr->ifr_settings.size == 0) { + new_settings.dce = 0; + new_settings.modulo = 8; + new_settings.window = 7; + new_settings.t1 = 3; + new_settings.t2 = 1; + new_settings.n2 = 10; + } + else { + if (copy_from_user(&new_settings, x25_s, size)) + return -EFAULT; + + if ((new_settings.dce != 0 && + new_settings.dce != 1) || + (new_settings.modulo != 8 && + new_settings.modulo != 128) || + new_settings.window < 1 || + (new_settings.modulo == 8 && + new_settings.window > 7) || + (new_settings.modulo == 128 && + new_settings.window > 127) || + new_settings.t1 < 1 || + new_settings.t1 > 255 || + new_settings.t2 < 1 || + new_settings.t2 > 255 || + new_settings.n2 < 1 || + new_settings.n2 > 255) + return -EINVAL; + } + result=hdlc->attach(dev, ENCODING_NRZ,PARITY_CRC16_PR1_CCITT); if (result) return result; - if ((result = attach_hdlc_protocol(dev, &proto, 0))) + if ((result = attach_hdlc_protocol(dev, &proto, + sizeof(struct x25_state)))) return result; + + memcpy(&state(hdlc)->settings, &new_settings, size); dev->type = ARPHRD_X25; call_netdevice_notifiers(NETDEV_POST_TYPE_CHANGE, dev); netif_dormant_off(dev); diff --git a/drivers/net/wan/ixp4xx_hss.c b/drivers/net/wan/ixp4xx_hss.c index ea6ee6a608ce..7c5cf77e9ef1 100644 --- a/drivers/net/wan/ixp4xx_hss.c +++ b/drivers/net/wan/ixp4xx_hss.c @@ -17,6 +17,7 @@ #include <linux/io.h> #include <linux/kernel.h> #include <linux/platform_device.h> +#include <linux/platform_data/wan_ixp4xx_hss.h> #include <linux/poll.h> #include <linux/slab.h> #include <linux/soc/ixp4xx/npe.h> @@ -258,7 +259,7 @@ struct port { struct hss_plat_info *plat; buffer_t *rx_buff_tab[RX_DESCS], *tx_buff_tab[TX_DESCS]; struct desc *desc_tab; /* coherent */ - u32 desc_tab_phys; + dma_addr_t desc_tab_phys; unsigned int id; unsigned int clock_type, clock_rate, loopback; unsigned int initialized, carrier; @@ -858,7 +859,7 @@ static int hss_hdlc_xmit(struct sk_buff *skb, struct net_device *dev) dev->stats.tx_dropped++; return NETDEV_TX_OK; } - memcpy_swab32(mem, (u32 *)((int)skb->data & ~3), bytes / 4); + memcpy_swab32(mem, (u32 *)((uintptr_t)skb->data & ~3), bytes / 4); dev_kfree_skb(skb); #endif @@ -1182,14 +1183,14 @@ static int hss_hdlc_attach(struct net_device *dev, unsigned short encoding, } } -static u32 check_clock(u32 rate, u32 a, u32 b, u32 c, +static u32 check_clock(u32 timer_freq, u32 rate, u32 a, u32 b, u32 c, u32 *best, u32 *best_diff, u32 *reg) { /* a is 10-bit, b is 10-bit, c is 12-bit */ u64 new_rate; u32 new_diff; - new_rate = ixp4xx_timer_freq * (u64)(c + 1); + new_rate = timer_freq * (u64)(c + 1); do_div(new_rate, a * (c + 1) + b + 1); new_diff = abs((u32)new_rate - rate); @@ -1201,40 +1202,43 @@ static u32 check_clock(u32 rate, u32 a, u32 b, u32 c, return new_diff; } -static void find_best_clock(u32 rate, u32 *best, u32 *reg) +static void find_best_clock(u32 timer_freq, u32 rate, u32 *best, u32 *reg) { u32 a, b, diff = 0xFFFFFFFF; - a = ixp4xx_timer_freq / rate; + a = timer_freq / rate; if (a > 0x3FF) { /* 10-bit value - we can go as slow as ca. 65 kb/s */ - check_clock(rate, 0x3FF, 1, 1, best, &diff, reg); + check_clock(timer_freq, rate, 0x3FF, 1, 1, best, &diff, reg); return; } if (a == 0) { /* > 66.666 MHz */ a = 1; /* minimum divider is 1 (a = 0, b = 1, c = 1) */ - rate = ixp4xx_timer_freq; + rate = timer_freq; } - if (rate * a == ixp4xx_timer_freq) { /* don't divide by 0 later */ - check_clock(rate, a - 1, 1, 1, best, &diff, reg); + if (rate * a == timer_freq) { /* don't divide by 0 later */ + check_clock(timer_freq, rate, a - 1, 1, 1, best, &diff, reg); return; } for (b = 0; b < 0x400; b++) { u64 c = (b + 1) * (u64)rate; - do_div(c, ixp4xx_timer_freq - rate * a); + do_div(c, timer_freq - rate * a); c--; if (c >= 0xFFF) { /* 12-bit - no need to check more 'b's */ if (b == 0 && /* also try a bit higher rate */ - !check_clock(rate, a - 1, 1, 1, best, &diff, reg)) + !check_clock(timer_freq, rate, a - 1, 1, 1, best, + &diff, reg)) return; - check_clock(rate, a, b, 0xFFF, best, &diff, reg); + check_clock(timer_freq, rate, a, b, 0xFFF, best, + &diff, reg); return; } - if (!check_clock(rate, a, b, c, best, &diff, reg)) + if (!check_clock(timer_freq, rate, a, b, c, best, &diff, reg)) return; - if (!check_clock(rate, a, b, c + 1, best, &diff, reg)) + if (!check_clock(timer_freq, rate, a, b, c + 1, best, &diff, + reg)) return; } } @@ -1285,8 +1289,9 @@ static int hss_hdlc_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd) port->clock_type = clk; /* Update settings */ if (clk == CLOCK_INT) - find_best_clock(new_line.clock_rate, &port->clock_rate, - &port->clock_reg); + find_best_clock(port->plat->timer_freq, + new_line.clock_rate, + &port->clock_rate, &port->clock_reg); else { port->clock_rate = 0; port->clock_reg = CLK42X_SPEED_2048KHZ; diff --git a/drivers/net/wan/lapbether.c b/drivers/net/wan/lapbether.c index 0f1217b506ad..e30d91a38cfb 100644 --- a/drivers/net/wan/lapbether.c +++ b/drivers/net/wan/lapbether.c @@ -64,7 +64,7 @@ static struct lapbethdev *lapbeth_get_x25_dev(struct net_device *dev) { struct lapbethdev *lapbeth; - list_for_each_entry_rcu(lapbeth, &lapbeth_devices, node) { + list_for_each_entry_rcu(lapbeth, &lapbeth_devices, node, lockdep_rtnl_is_held()) { if (lapbeth->ethdev == dev) return lapbeth; } diff --git a/drivers/net/wan/sdla.c b/drivers/net/wan/sdla.c index e2e679a01b65..77ccf3672ede 100644 --- a/drivers/net/wan/sdla.c +++ b/drivers/net/wan/sdla.c @@ -708,7 +708,7 @@ static netdev_tx_t sdla_transmit(struct sk_buff *skb, spin_lock_irqsave(&sdla_lock, flags); SDLA_WINDOW(dev, addr); - pbuf = (void *)(((int) dev->mem_start) + (addr & SDLA_ADDR_MASK)); + pbuf = (void *)(dev->mem_start + (addr & SDLA_ADDR_MASK)); __sdla_write(dev, pbuf->buf_addr, skb->data, skb->len); SDLA_WINDOW(dev, addr); pbuf->opp_flag = 1; diff --git a/drivers/net/wireguard/device.h b/drivers/net/wireguard/device.h index c91f3051c5c7..b15a8be9d816 100644 --- a/drivers/net/wireguard/device.h +++ b/drivers/net/wireguard/device.h @@ -62,12 +62,4 @@ struct wg_device { int wg_device_init(void); void wg_device_uninit(void); -/* Later after the dust settles, this can be moved into include/linux/skbuff.h, - * where virtually all code that deals with GSO segs can benefit, around ~30 - * drivers as of writing. - */ -#define skb_list_walk_safe(first, skb, next) \ - for (skb = first, next = skb->next; skb; \ - skb = next, next = skb ? skb->next : NULL) - #endif /* _WG_DEVICE_H */ diff --git a/drivers/net/wireless/intel/iwlwifi/mvm/tx.c b/drivers/net/wireless/intel/iwlwifi/mvm/tx.c index dc5c02fbc65a..6a241d37a057 100644 --- a/drivers/net/wireless/intel/iwlwifi/mvm/tx.c +++ b/drivers/net/wireless/intel/iwlwifi/mvm/tx.c @@ -847,10 +847,7 @@ iwl_mvm_tx_tso_segment(struct sk_buff *skb, unsigned int num_subframes, else if (next) consume_skb(skb); - while (next) { - tmp = next; - next = tmp->next; - + skb_list_walk_safe(next, tmp, next) { memcpy(tmp->cb, cb, sizeof(tmp->cb)); /* * Compute the length of all the data added for the A-MSDU. @@ -880,9 +877,7 @@ iwl_mvm_tx_tso_segment(struct sk_buff *skb, unsigned int num_subframes, skb_shinfo(tmp)->gso_size = 0; } - tmp->prev = NULL; - tmp->next = NULL; - + skb_mark_not_on_list(tmp); __skb_queue_tail(mpdus_skb, tmp); i++; } diff --git a/drivers/net/xen-netback/hash.c b/drivers/net/xen-netback/hash.c index 10d580c3dea3..6b7532f7c936 100644 --- a/drivers/net/xen-netback/hash.c +++ b/drivers/net/xen-netback/hash.c @@ -51,7 +51,8 @@ static void xenvif_add_hash(struct xenvif *vif, const u8 *tag, found = false; oldest = NULL; - list_for_each_entry_rcu(entry, &vif->hash.cache.list, link) { + list_for_each_entry_rcu(entry, &vif->hash.cache.list, link, + lockdep_is_held(&vif->hash.cache.lock)) { /* Make sure we don't add duplicate entries */ if (entry->len == len && memcmp(entry->tag, tag, len) == 0) @@ -102,7 +103,8 @@ static void xenvif_flush_hash(struct xenvif *vif) spin_lock_irqsave(&vif->hash.cache.lock, flags); - list_for_each_entry_rcu(entry, &vif->hash.cache.list, link) { + list_for_each_entry_rcu(entry, &vif->hash.cache.list, link, + lockdep_is_held(&vif->hash.cache.lock)) { list_del_rcu(&entry->link); vif->hash.cache.count--; kfree_rcu(entry, rcu); |