diff options
Diffstat (limited to 'drivers/net/ethernet/intel/ice/ice_main.c')
| -rw-r--r-- | drivers/net/ethernet/intel/ice/ice_main.c | 3664 |
1 files changed, 2901 insertions, 763 deletions
diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c index 4ee85a217c6f..0f6718719453 100644 --- a/drivers/net/ethernet/intel/ice/ice_main.c +++ b/drivers/net/ethernet/intel/ice/ice_main.c @@ -13,6 +13,15 @@ #include "ice_dcb_lib.h" #include "ice_dcb_nl.h" #include "ice_devlink.h" +/* Including ice_trace.h with CREATE_TRACE_POINTS defined will generate the + * ice tracepoint functions. This must be done exactly once across the + * ice driver. + */ +#define CREATE_TRACE_POINTS +#include "ice_trace.h" +#include "ice_eswitch.h" +#include "ice_tc_lib.h" +#include "ice_vsi_vlan_ops.h" #define DRV_SUMMARY "Intel(R) Ethernet Connection E800 Series Linux Driver" static const char ice_driver_string[] = DRV_SUMMARY; @@ -35,15 +44,42 @@ MODULE_PARM_DESC(debug, "netif level (0=none,...,16=all), hw debug_mask (0x8XXXX MODULE_PARM_DESC(debug, "netif level (0=none,...,16=all)"); #endif /* !CONFIG_DYNAMIC_DEBUG */ +static DEFINE_IDA(ice_aux_ida); +DEFINE_STATIC_KEY_FALSE(ice_xdp_locking_key); +EXPORT_SYMBOL(ice_xdp_locking_key); + +/** + * ice_hw_to_dev - Get device pointer from the hardware structure + * @hw: pointer to the device HW structure + * + * Used to access the device pointer from compilation units which can't easily + * include the definition of struct ice_pf without leading to circular header + * dependencies. + */ +struct device *ice_hw_to_dev(struct ice_hw *hw) +{ + struct ice_pf *pf = container_of(hw, struct ice_pf, hw); + + return &pf->pdev->dev; +} + static struct workqueue_struct *ice_wq; static const struct net_device_ops ice_netdev_safe_mode_ops; static const struct net_device_ops ice_netdev_ops; -static int ice_vsi_open(struct ice_vsi *vsi); static void ice_rebuild(struct ice_pf *pf, enum ice_reset_req reset_type); static void ice_vsi_release_all(struct ice_pf *pf); +static int ice_rebuild_channels(struct ice_pf *pf); +static void ice_remove_q_channels(struct ice_vsi *vsi, bool rem_adv_fltr); + +static int +ice_indr_setup_tc_cb(struct net_device *netdev, struct Qdisc *sch, + void *cb_priv, enum tc_setup_type type, void *type_data, + void *data, + void (*cleanup)(struct flow_block_cb *block_cb)); + bool netif_is_ice(struct net_device *dev) { return dev && (dev->netdev_ops == &ice_netdev_ops); @@ -53,7 +89,7 @@ bool netif_is_ice(struct net_device *dev) * ice_get_tx_pending - returns number of Tx descriptors not processed * @ring: the ring of descriptors */ -static u16 ice_get_tx_pending(struct ice_ring *ring) +static u16 ice_get_tx_pending(struct ice_tx_ring *ring) { u16 head, tail; @@ -92,10 +128,15 @@ static void ice_check_for_hang_subtask(struct ice_pf *pf) hw = &vsi->back->hw; - for (i = 0; i < vsi->num_txq; i++) { - struct ice_ring *tx_ring = vsi->tx_rings[i]; + ice_for_each_txq(vsi, i) { + struct ice_tx_ring *tx_ring = vsi->tx_rings[i]; + + if (!tx_ring) + continue; + if (ice_ring_ch_enabled(tx_ring)) + continue; - if (tx_ring && tx_ring->desc) { + if (tx_ring->desc) { /* If packet counter has not changed the queue is * likely stalled, so force an interrupt for this * queue. @@ -130,7 +171,6 @@ static void ice_check_for_hang_subtask(struct ice_pf *pf) */ static int ice_init_mac_fltr(struct ice_pf *pf) { - enum ice_status status; struct ice_vsi *vsi; u8 *perm_addr; @@ -139,11 +179,7 @@ static int ice_init_mac_fltr(struct ice_pf *pf) return -EINVAL; perm_addr = vsi->port_info->mac.perm_addr; - status = ice_fltr_add_mac_and_broadcast(vsi, perm_addr, ICE_FWD_TO_VSI); - if (status) - return -EIO; - - return 0; + return ice_fltr_add_mac_and_broadcast(vsi, perm_addr, ICE_FWD_TO_VSI); } /** @@ -183,6 +219,14 @@ static int ice_add_mac_to_unsync_list(struct net_device *netdev, const u8 *addr) struct ice_netdev_priv *np = netdev_priv(netdev); struct ice_vsi *vsi = np->vsi; + /* Under some circumstances, we might receive a request to delete our + * own device address from our uc list. Because we store the device + * address in the VSI's MAC filter list, we need to ignore such + * requests and not delete our device address from this list. + */ + if (ether_addr_equal(addr, netdev->dev_addr)) + return 0; + if (ice_fltr_add_mac_to_list(vsi, &vsi->tmp_unsync_list, addr, ICE_FWD_TO_VSI)) return -EINVAL; @@ -199,44 +243,76 @@ static int ice_add_mac_to_unsync_list(struct net_device *netdev, const u8 *addr) static bool ice_vsi_fltr_changed(struct ice_vsi *vsi) { return test_bit(ICE_VSI_UMAC_FLTR_CHANGED, vsi->state) || - test_bit(ICE_VSI_MMAC_FLTR_CHANGED, vsi->state) || - test_bit(ICE_VSI_VLAN_FLTR_CHANGED, vsi->state); + test_bit(ICE_VSI_MMAC_FLTR_CHANGED, vsi->state); } /** - * ice_cfg_promisc - Enable or disable promiscuous mode for a given PF + * ice_set_promisc - Enable promiscuous mode for a given PF * @vsi: the VSI being configured * @promisc_m: mask of promiscuous config bits - * @set_promisc: enable or disable promisc flag request * */ -static int ice_cfg_promisc(struct ice_vsi *vsi, u8 promisc_m, bool set_promisc) +static int ice_set_promisc(struct ice_vsi *vsi, u8 promisc_m) { - struct ice_hw *hw = &vsi->back->hw; - enum ice_status status = 0; + int status; if (vsi->type != ICE_VSI_PF) return 0; - if (vsi->num_vlan > 1) { - status = ice_set_vlan_vsi_promisc(hw, vsi->idx, promisc_m, - set_promisc); + if (ice_vsi_has_non_zero_vlans(vsi)) { + promisc_m |= (ICE_PROMISC_VLAN_RX | ICE_PROMISC_VLAN_TX); + status = ice_fltr_set_vlan_vsi_promisc(&vsi->back->hw, vsi, + promisc_m); } else { - if (set_promisc) - status = ice_set_vsi_promisc(hw, vsi->idx, promisc_m, - 0); - else - status = ice_clear_vsi_promisc(hw, vsi->idx, promisc_m, - 0); + status = ice_fltr_set_vsi_promisc(&vsi->back->hw, vsi->idx, + promisc_m, 0); } - - if (status) - return -EIO; + if (status && status != -EEXIST) + return status; return 0; } /** + * ice_clear_promisc - Disable promiscuous mode for a given PF + * @vsi: the VSI being configured + * @promisc_m: mask of promiscuous config bits + * + */ +static int ice_clear_promisc(struct ice_vsi *vsi, u8 promisc_m) +{ + int status; + + if (vsi->type != ICE_VSI_PF) + return 0; + + if (ice_vsi_has_non_zero_vlans(vsi)) { + promisc_m |= (ICE_PROMISC_VLAN_RX | ICE_PROMISC_VLAN_TX); + status = ice_fltr_clear_vlan_vsi_promisc(&vsi->back->hw, vsi, + promisc_m); + } else { + status = ice_fltr_clear_vsi_promisc(&vsi->back->hw, vsi->idx, + promisc_m, 0); + } + + return status; +} + +/** + * ice_get_devlink_port - Get devlink port from netdev + * @netdev: the netdevice structure + */ +static struct devlink_port *ice_get_devlink_port(struct net_device *netdev) +{ + struct ice_pf *pf = ice_netdev_to_pf(netdev); + + if (!ice_is_switchdev_running(pf)) + return NULL; + + return &pf->devlink_port; +} + +/** * ice_vsi_sync_fltr - Update the VSI filter list to the HW * @vsi: ptr to the VSI * @@ -244,15 +320,14 @@ static int ice_cfg_promisc(struct ice_vsi *vsi, u8 promisc_m, bool set_promisc) */ static int ice_vsi_sync_fltr(struct ice_vsi *vsi) { + struct ice_vsi_vlan_ops *vlan_ops = ice_get_compat_vsi_vlan_ops(vsi); struct device *dev = ice_pf_to_dev(vsi->back); struct net_device *netdev = vsi->netdev; bool promisc_forced_on = false; struct ice_pf *pf = vsi->back; struct ice_hw *hw = &pf->hw; - enum ice_status status = 0; u32 changed_flags = 0; - u8 promisc_m; - int err = 0; + int err; if (!vsi->netdev) return -EINVAL; @@ -269,7 +344,6 @@ static int ice_vsi_sync_fltr(struct ice_vsi *vsi) if (ice_vsi_fltr_changed(vsi)) { clear_bit(ICE_VSI_UMAC_FLTR_CHANGED, vsi->state); clear_bit(ICE_VSI_MMAC_FLTR_CHANGED, vsi->state); - clear_bit(ICE_VSI_VLAN_FLTR_CHANGED, vsi->state); /* grab the netdev's addr_list_lock */ netif_addr_lock_bh(netdev); @@ -282,25 +356,23 @@ static int ice_vsi_sync_fltr(struct ice_vsi *vsi) } /* Remove MAC addresses in the unsync list */ - status = ice_fltr_remove_mac_list(vsi, &vsi->tmp_unsync_list); + err = ice_fltr_remove_mac_list(vsi, &vsi->tmp_unsync_list); ice_fltr_free_list(dev, &vsi->tmp_unsync_list); - if (status) { + if (err) { netdev_err(netdev, "Failed to delete MAC filters\n"); /* if we failed because of alloc failures, just bail */ - if (status == ICE_ERR_NO_MEMORY) { - err = -ENOMEM; + if (err == -ENOMEM) goto out; - } } /* Add MAC addresses in the sync list */ - status = ice_fltr_add_mac_list(vsi, &vsi->tmp_sync_list); + err = ice_fltr_add_mac_list(vsi, &vsi->tmp_sync_list); ice_fltr_free_list(dev, &vsi->tmp_sync_list); /* If filter is added successfully or already exists, do not go into * 'if' condition and report it as error. Instead continue processing * rest of the function. */ - if (status && status != ICE_ERR_ALREADY_EXISTS) { + if (err && err != -EEXIST) { netdev_err(netdev, "Failed to add MAC filters\n"); /* If there is no more space for new umac filters, VSI * should go into promiscuous mode. There should be some @@ -313,36 +385,22 @@ static int ice_vsi_sync_fltr(struct ice_vsi *vsi) netdev_warn(netdev, "Reached MAC filter limit, forcing promisc mode on VSI %d\n", vsi->vsi_num); } else { - err = -EIO; goto out; } } + err = 0; /* check for changes in promiscuous modes */ if (changed_flags & IFF_ALLMULTI) { if (vsi->current_netdev_flags & IFF_ALLMULTI) { - if (vsi->num_vlan > 1) - promisc_m = ICE_MCAST_VLAN_PROMISC_BITS; - else - promisc_m = ICE_MCAST_PROMISC_BITS; - - err = ice_cfg_promisc(vsi, promisc_m, true); + err = ice_set_promisc(vsi, ICE_MCAST_PROMISC_BITS); if (err) { - netdev_err(netdev, "Error setting Multicast promiscuous mode on VSI %i\n", - vsi->vsi_num); vsi->current_netdev_flags &= ~IFF_ALLMULTI; goto out_promisc; } } else { /* !(vsi->current_netdev_flags & IFF_ALLMULTI) */ - if (vsi->num_vlan > 1) - promisc_m = ICE_MCAST_VLAN_PROMISC_BITS; - else - promisc_m = ICE_MCAST_PROMISC_BITS; - - err = ice_cfg_promisc(vsi, promisc_m, false); + err = ice_clear_promisc(vsi, ICE_MCAST_PROMISC_BITS); if (err) { - netdev_err(netdev, "Error clearing Multicast promiscuous mode on VSI %i\n", - vsi->vsi_num); vsi->current_netdev_flags |= IFF_ALLMULTI; goto out_promisc; } @@ -354,8 +412,8 @@ static int ice_vsi_sync_fltr(struct ice_vsi *vsi) clear_bit(ICE_VSI_PROMISC_CHANGED, vsi->state); if (vsi->current_netdev_flags & IFF_PROMISC) { /* Apply Rx filter rule to get traffic from wire */ - if (!ice_is_dflt_vsi_in_use(pf->first_sw)) { - err = ice_set_dflt_vsi(pf->first_sw, vsi); + if (!ice_is_dflt_vsi_in_use(vsi->port_info)) { + err = ice_set_dflt_vsi(vsi); if (err && err != -EEXIST) { netdev_err(netdev, "Error %d setting default VSI %i Rx rule\n", err, vsi->vsi_num); @@ -363,12 +421,13 @@ static int ice_vsi_sync_fltr(struct ice_vsi *vsi) ~IFF_PROMISC; goto out_promisc; } - ice_cfg_vlan_pruning(vsi, false, false); + err = 0; + vlan_ops->dis_rx_filtering(vsi); } } else { /* Clear Rx filter to remove traffic from wire */ - if (ice_is_vsi_dflt_vsi(pf->first_sw, vsi)) { - err = ice_clear_dflt_vsi(pf->first_sw); + if (ice_is_vsi_dflt_vsi(vsi)) { + err = ice_clear_dflt_vsi(vsi); if (err) { netdev_err(netdev, "Error %d clearing default VSI %i Rx rule\n", err, vsi->vsi_num); @@ -376,8 +435,9 @@ static int ice_vsi_sync_fltr(struct ice_vsi *vsi) IFF_PROMISC; goto out_promisc; } - if (vsi->num_vlan > 1) - ice_cfg_vlan_pruning(vsi, true, false); + if (vsi->netdev->features & + NETIF_F_HW_VLAN_CTAG_FILTER) + vlan_ops->ena_rx_filtering(vsi); } } } @@ -439,34 +499,105 @@ static void ice_pf_dis_all_vsi(struct ice_pf *pf, bool locked) } /** - * ice_prepare_for_reset - prep for the core to reset + * ice_clear_sw_switch_recipes - clear switch recipes * @pf: board private structure * + * Mark switch recipes as not created in sw structures. There are cases where + * rules (especially advanced rules) need to be restored, either re-read from + * hardware or added again. For example after the reset. 'recp_created' flag + * prevents from doing that and need to be cleared upfront. + */ +static void ice_clear_sw_switch_recipes(struct ice_pf *pf) +{ + struct ice_sw_recipe *recp; + u8 i; + + recp = pf->hw.switch_info->recp_list; + for (i = 0; i < ICE_MAX_NUM_RECIPES; i++) + recp[i].recp_created = false; +} + +/** + * ice_prepare_for_reset - prep for reset + * @pf: board private structure + * @reset_type: reset type requested + * * Inform or close all dependent features in prep for reset. */ static void -ice_prepare_for_reset(struct ice_pf *pf) +ice_prepare_for_reset(struct ice_pf *pf, enum ice_reset_req reset_type) { struct ice_hw *hw = &pf->hw; - unsigned int i; + struct ice_vsi *vsi; + struct ice_vf *vf; + unsigned int bkt; + + dev_dbg(ice_pf_to_dev(pf), "reset_type=%d\n", reset_type); /* already prepared for reset */ if (test_bit(ICE_PREPARED_FOR_RESET, pf->state)) return; + ice_unplug_aux_dev(pf); + /* Notify VFs of impending reset */ if (ice_check_sq_alive(hw, &hw->mailboxq)) ice_vc_notify_reset(pf); /* Disable VFs until reset is completed */ - ice_for_each_vf(pf, i) - ice_set_vf_state_qs_dis(&pf->vf[i]); + mutex_lock(&pf->vfs.table_lock); + ice_for_each_vf(pf, bkt, vf) + ice_set_vf_state_qs_dis(vf); + mutex_unlock(&pf->vfs.table_lock); + + if (ice_is_eswitch_mode_switchdev(pf)) { + if (reset_type != ICE_RESET_PFR) + ice_clear_sw_switch_recipes(pf); + } + + /* release ADQ specific HW and SW resources */ + vsi = ice_get_main_vsi(pf); + if (!vsi) + goto skip; + + /* to be on safe side, reset orig_rss_size so that normal flow + * of deciding rss_size can take precedence + */ + vsi->orig_rss_size = 0; + + if (test_bit(ICE_FLAG_TC_MQPRIO, pf->flags)) { + if (reset_type == ICE_RESET_PFR) { + vsi->old_ena_tc = vsi->all_enatc; + vsi->old_numtc = vsi->all_numtc; + } else { + ice_remove_q_channels(vsi, true); + + /* for other reset type, do not support channel rebuild + * hence reset needed info + */ + vsi->old_ena_tc = 0; + vsi->all_enatc = 0; + vsi->old_numtc = 0; + vsi->all_numtc = 0; + vsi->req_txq = 0; + vsi->req_rxq = 0; + clear_bit(ICE_FLAG_TC_MQPRIO, pf->flags); + memset(&vsi->mqprio_qopt, 0, sizeof(vsi->mqprio_qopt)); + } + } +skip: /* clear SW filtering DB */ ice_clear_hw_tbls(hw); /* disable the VSIs and their queues that are not already DOWN */ ice_pf_dis_all_vsi(pf, false); + if (test_bit(ICE_FLAG_PTP_SUPPORTED, pf->flags)) + ice_ptp_prepare_for_reset(pf); + + if (ice_is_feature_supported(pf, ICE_F_GNSS)) + ice_gnss_exit(pf); + if (hw->port_info) ice_sched_clear_port(hw->port_info); @@ -478,8 +609,7 @@ ice_prepare_for_reset(struct ice_pf *pf) /** * ice_do_reset - Initiate one of many types of resets * @pf: board private structure - * @reset_type: reset type requested - * before this function was called. + * @reset_type: reset type requested before this function was called. */ static void ice_do_reset(struct ice_pf *pf, enum ice_reset_req reset_type) { @@ -488,7 +618,7 @@ static void ice_do_reset(struct ice_pf *pf, enum ice_reset_req reset_type) dev_dbg(dev, "reset_type 0x%x requested\n", reset_type); - ice_prepare_for_reset(pf); + ice_prepare_for_reset(pf, reset_type); /* trigger the reset */ if (ice_reset(hw, reset_type)) { @@ -499,6 +629,7 @@ static void ice_do_reset(struct ice_pf *pf, enum ice_reset_req reset_type) clear_bit(ICE_PFR_REQ, pf->state); clear_bit(ICE_CORER_REQ, pf->state); clear_bit(ICE_GLOBR_REQ, pf->state); + wake_up(&pf->reset_wait_queue); return; } @@ -511,7 +642,8 @@ static void ice_do_reset(struct ice_pf *pf, enum ice_reset_req reset_type) ice_rebuild(pf, reset_type); clear_bit(ICE_PREPARED_FOR_RESET, pf->state); clear_bit(ICE_PFR_REQ, pf->state); - ice_reset_all_vfs(pf, true); + wake_up(&pf->reset_wait_queue); + ice_reset_all_vfs(pf); } } @@ -544,7 +676,7 @@ static void ice_reset_subtask(struct ice_pf *pf) /* return if no valid reset type requested */ if (reset_type == ICE_RESET_INVAL) return; - ice_prepare_for_reset(pf); + ice_prepare_for_reset(pf, reset_type); /* make sure we are ready to rebuild */ if (ice_check_reset(&pf->hw)) { @@ -561,7 +693,8 @@ static void ice_reset_subtask(struct ice_pf *pf) clear_bit(ICE_PFR_REQ, pf->state); clear_bit(ICE_CORER_REQ, pf->state); clear_bit(ICE_GLOBR_REQ, pf->state); - ice_reset_all_vfs(pf, true); + wake_up(&pf->reset_wait_queue); + ice_reset_all_vfs(pf); } return; @@ -600,7 +733,10 @@ static void ice_print_topo_conflict(struct ice_vsi *vsi) netdev_info(vsi->netdev, "Potential misconfiguration of the Ethernet port detected. If it was not intended, please use the Intel (R) Ethernet Port Configuration Tool to address the issue.\n"); break; case ICE_AQ_LINK_TOPO_UNSUPP_MEDIA: - netdev_info(vsi->netdev, "Rx/Tx is disabled on this device because an unsupported module type was detected. Refer to the Intel(R) Ethernet Adapters and Devices User Guide for a list of supported modules.\n"); + if (test_bit(ICE_FLAG_LINK_LENIENT_MODE_ENA, vsi->back->flags)) + netdev_warn(vsi->netdev, "An unsupported module type was detected. Refer to the Intel(R) Ethernet Adapters and Devices User Guide for a list of supported modules\n"); + else + netdev_err(vsi->netdev, "Rx/Tx is disabled on this device because an unsupported module type was detected. Refer to the Intel(R) Ethernet Adapters and Devices User Guide for a list of supported modules.\n"); break; default: break; @@ -616,12 +752,12 @@ void ice_print_link_msg(struct ice_vsi *vsi, bool isup) { struct ice_aqc_get_phy_caps_data *caps; const char *an_advertised; - enum ice_status status; const char *fec_req; const char *speed; const char *fec; const char *fc; const char *an; + int status; if (!vsi) return; @@ -858,6 +994,75 @@ static void ice_set_dflt_mib(struct ice_pf *pf) } /** + * ice_check_phy_fw_load - check if PHY FW load failed + * @pf: pointer to PF struct + * @link_cfg_err: bitmap from the link info structure + * + * check if external PHY FW load failed and print an error message if it did + */ +static void ice_check_phy_fw_load(struct ice_pf *pf, u8 link_cfg_err) +{ + if (!(link_cfg_err & ICE_AQ_LINK_EXTERNAL_PHY_LOAD_FAILURE)) { + clear_bit(ICE_FLAG_PHY_FW_LOAD_FAILED, pf->flags); + return; + } + + if (test_bit(ICE_FLAG_PHY_FW_LOAD_FAILED, pf->flags)) + return; + + if (link_cfg_err & ICE_AQ_LINK_EXTERNAL_PHY_LOAD_FAILURE) { + dev_err(ice_pf_to_dev(pf), "Device failed to load the FW for the external PHY. Please download and install the latest NVM for your device and try again\n"); + set_bit(ICE_FLAG_PHY_FW_LOAD_FAILED, pf->flags); + } +} + +/** + * ice_check_module_power + * @pf: pointer to PF struct + * @link_cfg_err: bitmap from the link info structure + * + * check module power level returned by a previous call to aq_get_link_info + * and print error messages if module power level is not supported + */ +static void ice_check_module_power(struct ice_pf *pf, u8 link_cfg_err) +{ + /* if module power level is supported, clear the flag */ + if (!(link_cfg_err & (ICE_AQ_LINK_INVAL_MAX_POWER_LIMIT | + ICE_AQ_LINK_MODULE_POWER_UNSUPPORTED))) { + clear_bit(ICE_FLAG_MOD_POWER_UNSUPPORTED, pf->flags); + return; + } + + /* if ICE_FLAG_MOD_POWER_UNSUPPORTED was previously set and the + * above block didn't clear this bit, there's nothing to do + */ + if (test_bit(ICE_FLAG_MOD_POWER_UNSUPPORTED, pf->flags)) + return; + + if (link_cfg_err & ICE_AQ_LINK_INVAL_MAX_POWER_LIMIT) { + dev_err(ice_pf_to_dev(pf), "The installed module is incompatible with the device's NVM image. Cannot start link\n"); + set_bit(ICE_FLAG_MOD_POWER_UNSUPPORTED, pf->flags); + } else if (link_cfg_err & ICE_AQ_LINK_MODULE_POWER_UNSUPPORTED) { + dev_err(ice_pf_to_dev(pf), "The module's power requirements exceed the device's power supply. Cannot start link\n"); + set_bit(ICE_FLAG_MOD_POWER_UNSUPPORTED, pf->flags); + } +} + +/** + * ice_check_link_cfg_err - check if link configuration failed + * @pf: pointer to the PF struct + * @link_cfg_err: bitmap from the link info structure + * + * print if any link configuration failure happens due to the value in the + * link_cfg_err parameter in the link info structure + */ +static void ice_check_link_cfg_err(struct ice_pf *pf, u8 link_cfg_err) +{ + ice_check_module_power(pf, link_cfg_err); + ice_check_phy_fw_load(pf, link_cfg_err); +} + +/** * ice_link_event - process the link event * @pf: PF that the link event is associated with * @pi: port_info for the port that the link event is associated with @@ -872,10 +1077,10 @@ ice_link_event(struct ice_pf *pf, struct ice_port_info *pi, bool link_up, { struct device *dev = ice_pf_to_dev(pf); struct ice_phy_info *phy_info; - enum ice_status status; struct ice_vsi *vsi; u16 old_link_speed; bool old_link; + int status; phy_info = &pi->phy; phy_info->link_info_old = phy_info->link_info; @@ -888,10 +1093,12 @@ ice_link_event(struct ice_pf *pf, struct ice_port_info *pi, bool link_up, */ status = ice_update_link_info(pi); if (status) - dev_dbg(dev, "Failed to update link status on port %d, err %s aq_err %s\n", - pi->lport, ice_stat_str(status), + dev_dbg(dev, "Failed to update link status on port %d, err %d aq_err %s\n", + pi->lport, status, ice_aq_str(pi->hw->adminq.sq_last_status)); + ice_check_link_cfg_err(pf, pi->phy.link_info.link_cfg_err); + /* Check if the link state is up after updating link info, and treat * this event as an UP event since the link is actually UP now. */ @@ -913,6 +1120,9 @@ ice_link_event(struct ice_pf *pf, struct ice_port_info *pi, bool link_up, if (link_up == old_link && link_speed == old_link_speed) return 0; + if (!ice_is_e810(&pf->hw)) + ice_ptp_link_change(pf, pf->hw.pf_id, link_up); + if (ice_is_dcb_active(pf)) { if (test_bit(ICE_FLAG_DCB_ENA, pf->flags)) ice_dcb_rebuild(pf); @@ -968,7 +1178,8 @@ static int ice_init_link_events(struct ice_port_info *pi) u16 mask; mask = ~((u16)(ICE_AQ_LINK_EVENT_UPDOWN | ICE_AQ_LINK_EVENT_MEDIA_NA | - ICE_AQ_LINK_EVENT_MODULE_QUAL_FAIL)); + ICE_AQ_LINK_EVENT_MODULE_QUAL_FAIL | + ICE_AQ_LINK_EVENT_PHY_FW_LOAD_FAIL)); if (ice_aq_set_event_mask(pi->hw, pi->lport, mask, NULL)) { dev_dbg(ice_hw_to_dev(pi->hw), "Failed to set link event mask for port %d\n", @@ -1190,6 +1401,10 @@ static int __ice_clean_ctrlq(struct ice_pf *pf, enum ice_ctl_q q_type) cq = &hw->adminq; qtype = "Admin"; break; + case ICE_CTL_Q_SB: + cq = &hw->sbq; + qtype = "Sideband"; + break; case ICE_CTL_Q_MAILBOX: cq = &hw->mailboxq; qtype = "Mailbox"; @@ -1252,15 +1467,15 @@ static int __ice_clean_ctrlq(struct ice_pf *pf, enum ice_ctl_q q_type) return 0; do { - enum ice_status ret; u16 opcode; + int ret; ret = ice_clean_rq_elem(hw, cq, &event, &pending); - if (ret == ICE_ERR_AQ_NO_WORK) + if (ret == -EALREADY) break; if (ret) { - dev_err(dev, "%s Receive Queue event error %s\n", qtype, - ice_stat_str(ret)); + dev_err(dev, "%s Receive Queue event error %d\n", qtype, + ret); break; } @@ -1364,6 +1579,34 @@ static void ice_clean_mailboxq_subtask(struct ice_pf *pf) } /** + * ice_clean_sbq_subtask - clean the Sideband Queue rings + * @pf: board private structure + */ +static void ice_clean_sbq_subtask(struct ice_pf *pf) +{ + struct ice_hw *hw = &pf->hw; + + /* Nothing to do here if sideband queue is not supported */ + if (!ice_is_sbq_supported(hw)) { + clear_bit(ICE_SIDEBANDQ_EVENT_PENDING, pf->state); + return; + } + + if (!test_bit(ICE_SIDEBANDQ_EVENT_PENDING, pf->state)) + return; + + if (__ice_clean_ctrlq(pf, ICE_CTL_Q_SB)) + return; + + clear_bit(ICE_SIDEBANDQ_EVENT_PENDING, pf->state); + + if (ice_ctrlq_pending(hw, &hw->sbq)) + __ice_clean_ctrlq(pf, ICE_CTL_Q_SB); + + ice_flush(hw); +} + +/** * ice_service_task_schedule - schedule the service task to wake up * @pf: board private structure * @@ -1450,7 +1693,8 @@ static void ice_handle_mdd_event(struct ice_pf *pf) { struct device *dev = ice_pf_to_dev(pf); struct ice_hw *hw = &pf->hw; - unsigned int i; + struct ice_vf *vf; + unsigned int bkt; u32 reg; if (!test_and_clear_bit(ICE_MDD_EVENT_PENDING, pf->state)) { @@ -1538,47 +1782,46 @@ static void ice_handle_mdd_event(struct ice_pf *pf) /* Check to see if one of the VFs caused an MDD event, and then * increment counters and set print pending */ - ice_for_each_vf(pf, i) { - struct ice_vf *vf = &pf->vf[i]; - - reg = rd32(hw, VP_MDET_TX_PQM(i)); + mutex_lock(&pf->vfs.table_lock); + ice_for_each_vf(pf, bkt, vf) { + reg = rd32(hw, VP_MDET_TX_PQM(vf->vf_id)); if (reg & VP_MDET_TX_PQM_VALID_M) { - wr32(hw, VP_MDET_TX_PQM(i), 0xFFFF); + wr32(hw, VP_MDET_TX_PQM(vf->vf_id), 0xFFFF); vf->mdd_tx_events.count++; set_bit(ICE_MDD_VF_PRINT_PENDING, pf->state); if (netif_msg_tx_err(pf)) dev_info(dev, "Malicious Driver Detection event TX_PQM detected on VF %d\n", - i); + vf->vf_id); } - reg = rd32(hw, VP_MDET_TX_TCLAN(i)); + reg = rd32(hw, VP_MDET_TX_TCLAN(vf->vf_id)); if (reg & VP_MDET_TX_TCLAN_VALID_M) { - wr32(hw, VP_MDET_TX_TCLAN(i), 0xFFFF); + wr32(hw, VP_MDET_TX_TCLAN(vf->vf_id), 0xFFFF); vf->mdd_tx_events.count++; set_bit(ICE_MDD_VF_PRINT_PENDING, pf->state); if (netif_msg_tx_err(pf)) dev_info(dev, "Malicious Driver Detection event TX_TCLAN detected on VF %d\n", - i); + vf->vf_id); } - reg = rd32(hw, VP_MDET_TX_TDPU(i)); + reg = rd32(hw, VP_MDET_TX_TDPU(vf->vf_id)); if (reg & VP_MDET_TX_TDPU_VALID_M) { - wr32(hw, VP_MDET_TX_TDPU(i), 0xFFFF); + wr32(hw, VP_MDET_TX_TDPU(vf->vf_id), 0xFFFF); vf->mdd_tx_events.count++; set_bit(ICE_MDD_VF_PRINT_PENDING, pf->state); if (netif_msg_tx_err(pf)) dev_info(dev, "Malicious Driver Detection event TX_TDPU detected on VF %d\n", - i); + vf->vf_id); } - reg = rd32(hw, VP_MDET_RX(i)); + reg = rd32(hw, VP_MDET_RX(vf->vf_id)); if (reg & VP_MDET_RX_VALID_M) { - wr32(hw, VP_MDET_RX(i), 0xFFFF); + wr32(hw, VP_MDET_RX(vf->vf_id), 0xFFFF); vf->mdd_rx_events.count++; set_bit(ICE_MDD_VF_PRINT_PENDING, pf->state); if (netif_msg_rx_err(pf)) dev_info(dev, "Malicious Driver Detection event RX detected on VF %d\n", - i); + vf->vf_id); /* Since the queue is disabled on VF Rx MDD events, the * PF can be configured to reset the VF through ethtool @@ -1589,10 +1832,11 @@ static void ice_handle_mdd_event(struct ice_pf *pf) * reset, so print the event prior to reset. */ ice_print_vf_rx_mdd_event(vf); - ice_reset_vf(&pf->vf[i], false); + ice_reset_vf(vf, ICE_VF_RESET_LOCK); } } } + mutex_unlock(&pf->vfs.table_lock); ice_print_vfs_mdd_events(pf); } @@ -1683,19 +1927,17 @@ static int ice_init_nvm_phy_type(struct ice_port_info *pi) { struct ice_aqc_get_phy_caps_data *pcaps; struct ice_pf *pf = pi->hw->back; - enum ice_status status; - int err = 0; + int err; pcaps = kzalloc(sizeof(*pcaps), GFP_KERNEL); if (!pcaps) return -ENOMEM; - status = ice_aq_get_phy_caps(pi, false, ICE_AQC_REPORT_TOPO_CAP_NO_MEDIA, pcaps, - NULL); + err = ice_aq_get_phy_caps(pi, false, ICE_AQC_REPORT_TOPO_CAP_NO_MEDIA, + pcaps, NULL); - if (status) { + if (err) { dev_err(ice_pf_to_dev(pf), "Get PHY capability failed.\n"); - err = -EIO; goto out; } @@ -1794,8 +2036,7 @@ static int ice_init_phy_user_cfg(struct ice_port_info *pi) struct ice_aqc_get_phy_caps_data *pcaps; struct ice_phy_info *phy = &pi->phy; struct ice_pf *pf = pi->hw->back; - enum ice_status status; - int err = 0; + int err; if (!(phy->link_info.link_info & ICE_AQ_MEDIA_AVAILABLE)) return -EIO; @@ -1805,14 +2046,13 @@ static int ice_init_phy_user_cfg(struct ice_port_info *pi) return -ENOMEM; if (ice_fw_supports_report_dflt_cfg(pi->hw)) - status = ice_aq_get_phy_caps(pi, false, ICE_AQC_REPORT_DFLT_CFG, - pcaps, NULL); + err = ice_aq_get_phy_caps(pi, false, ICE_AQC_REPORT_DFLT_CFG, + pcaps, NULL); else - status = ice_aq_get_phy_caps(pi, false, ICE_AQC_REPORT_TOPO_CAP_MEDIA, - pcaps, NULL); - if (status) { + err = ice_aq_get_phy_caps(pi, false, ICE_AQC_REPORT_TOPO_CAP_MEDIA, + pcaps, NULL); + if (err) { dev_err(ice_pf_to_dev(pf), "Get PHY capability failed.\n"); - err = -EIO; goto err_out; } @@ -1866,8 +2106,7 @@ static int ice_configure_phy(struct ice_vsi *vsi) struct ice_aqc_set_phy_cfg_data *cfg; struct ice_phy_info *phy = &pi->phy; struct ice_pf *pf = vsi->back; - enum ice_status status; - int err = 0; + int err; /* Ensure we have media as we cannot configure a medialess port */ if (!(phy->link_info.link_info & ICE_AQ_MEDIA_AVAILABLE)) @@ -1875,7 +2114,8 @@ static int ice_configure_phy(struct ice_vsi *vsi) ice_print_topo_conflict(vsi); - if (phy->link_info.topo_media_conflict == ICE_AQ_LINK_TOPO_UNSUPP_MEDIA) + if (!test_bit(ICE_FLAG_LINK_LENIENT_MODE_ENA, pf->flags) && + phy->link_info.topo_media_conflict == ICE_AQ_LINK_TOPO_UNSUPP_MEDIA) return -EPERM; if (test_bit(ICE_FLAG_LINK_DOWN_ON_CLOSE_ENA, pf->flags)) @@ -1886,12 +2126,11 @@ static int ice_configure_phy(struct ice_vsi *vsi) return -ENOMEM; /* Get current PHY config */ - status = ice_aq_get_phy_caps(pi, false, ICE_AQC_REPORT_ACTIVE_CFG, pcaps, - NULL); - if (status) { - dev_err(dev, "Failed to get PHY configuration, VSI %d error %s\n", - vsi->vsi_num, ice_stat_str(status)); - err = -EIO; + err = ice_aq_get_phy_caps(pi, false, ICE_AQC_REPORT_ACTIVE_CFG, pcaps, + NULL); + if (err) { + dev_err(dev, "Failed to get PHY configuration, VSI %d error %d\n", + vsi->vsi_num, err); goto done; } @@ -1905,15 +2144,14 @@ static int ice_configure_phy(struct ice_vsi *vsi) /* Use PHY topology as baseline for configuration */ memset(pcaps, 0, sizeof(*pcaps)); if (ice_fw_supports_report_dflt_cfg(pi->hw)) - status = ice_aq_get_phy_caps(pi, false, ICE_AQC_REPORT_DFLT_CFG, - pcaps, NULL); + err = ice_aq_get_phy_caps(pi, false, ICE_AQC_REPORT_DFLT_CFG, + pcaps, NULL); else - status = ice_aq_get_phy_caps(pi, false, ICE_AQC_REPORT_TOPO_CAP_MEDIA, - pcaps, NULL); - if (status) { - dev_err(dev, "Failed to get PHY caps, VSI %d error %s\n", - vsi->vsi_num, ice_stat_str(status)); - err = -EIO; + err = ice_aq_get_phy_caps(pi, false, ICE_AQC_REPORT_TOPO_CAP_MEDIA, + pcaps, NULL); + if (err) { + dev_err(dev, "Failed to get PHY caps, VSI %d error %d\n", + vsi->vsi_num, err); goto done; } @@ -1966,12 +2204,10 @@ static int ice_configure_phy(struct ice_vsi *vsi) /* Enable link and link update */ cfg->caps |= ICE_AQ_PHY_ENA_AUTO_LINK_UPDT | ICE_AQ_PHY_ENA_LINK; - status = ice_aq_set_phy_cfg(&pf->hw, pi, cfg, NULL); - if (status) { - dev_err(dev, "Failed to set phy config, VSI %d error %s\n", - vsi->vsi_num, ice_stat_str(status)); - err = -EIO; - } + err = ice_aq_set_phy_cfg(&pf->hw, pi, cfg, NULL); + if (err) + dev_err(dev, "Failed to set phy config, VSI %d error %d\n", + vsi->vsi_num, err); kfree(cfg); done: @@ -2006,6 +2242,8 @@ static void ice_check_media_subtask(struct ice_pf *pf) if (err) return; + ice_check_link_cfg_err(pf, pi->phy.link_info.link_cfg_err); + if (pi->phy.link_info.link_info & ICE_AQ_MEDIA_AVAILABLE) { if (!test_bit(ICE_PHY_INIT_COMPLETE, pf->state)) ice_init_phy_user_cfg(pi); @@ -2049,6 +2287,43 @@ static void ice_service_task(struct work_struct *work) return; } + if (test_and_clear_bit(ICE_AUX_ERR_PENDING, pf->state)) { + struct iidc_event *event; + + event = kzalloc(sizeof(*event), GFP_KERNEL); + if (event) { + set_bit(IIDC_EVENT_CRIT_ERR, event->type); + /* report the entire OICR value to AUX driver */ + swap(event->reg, pf->oicr_err_reg); + ice_send_event_to_aux(pf, event); + kfree(event); + } + } + + if (test_bit(ICE_FLAG_PLUG_AUX_DEV, pf->flags)) { + /* Plug aux device per request */ + ice_plug_aux_dev(pf); + + /* Mark plugging as done but check whether unplug was + * requested during ice_plug_aux_dev() call + * (e.g. from ice_clear_rdma_cap()) and if so then + * plug aux device. + */ + if (!test_and_clear_bit(ICE_FLAG_PLUG_AUX_DEV, pf->flags)) + ice_unplug_aux_dev(pf); + } + + if (test_and_clear_bit(ICE_FLAG_MTU_CHANGED, pf->flags)) { + struct iidc_event *event; + + event = kzalloc(sizeof(*event), GFP_KERNEL); + if (event) { + set_bit(IIDC_EVENT_AFTER_MTU_CHANGE, event->type); + ice_send_event_to_aux(pf, event); + kfree(event); + } + } + ice_clean_adminq_subtask(pf); ice_check_media_subtask(pf); ice_check_for_hang_subtask(pf); @@ -2063,6 +2338,7 @@ static void ice_service_task(struct work_struct *work) ice_process_vflr_event(pf); ice_clean_mailboxq_subtask(pf); + ice_clean_sbq_subtask(pf); ice_sync_arfs_fltrs(pf); ice_flush_fdir_ctx(pf); @@ -2078,6 +2354,7 @@ static void ice_service_task(struct work_struct *work) test_bit(ICE_VFLR_EVENT_PENDING, pf->state) || test_bit(ICE_MAILBOXQ_EVENT_PENDING, pf->state) || test_bit(ICE_FD_VF_FLUSH_CTX, pf->state) || + test_bit(ICE_SIDEBANDQ_EVENT_PENDING, pf->state) || test_bit(ICE_ADMINQ_EVENT_PENDING, pf->state)) mod_timer(&pf->serv_tmr, jiffies); } @@ -2096,6 +2373,10 @@ static void ice_set_ctrlq_len(struct ice_hw *hw) hw->mailboxq.num_sq_entries = ICE_MBXSQ_LEN; hw->mailboxq.rq_buf_size = ICE_MBXQ_MAX_BUF_LEN; hw->mailboxq.sq_buf_size = ICE_MBXQ_MAX_BUF_LEN; + hw->sbq.num_rq_entries = ICE_SBQ_LEN; + hw->sbq.num_sq_entries = ICE_SBQ_LEN; + hw->sbq.rq_buf_size = ICE_SBQ_MAX_BUF_LEN; + hw->sbq.sq_buf_size = ICE_SBQ_MAX_BUF_LEN; } /** @@ -2202,21 +2483,21 @@ static int ice_vsi_req_irq_msix(struct ice_vsi *vsi, char *basename) irq_num = pf->msix_entries[base + vector].vector; - if (q_vector->tx.ring && q_vector->rx.ring) { + if (q_vector->tx.tx_ring && q_vector->rx.rx_ring) { snprintf(q_vector->name, sizeof(q_vector->name) - 1, "%s-%s-%d", basename, "TxRx", rx_int_idx++); tx_int_idx++; - } else if (q_vector->rx.ring) { + } else if (q_vector->rx.rx_ring) { snprintf(q_vector->name, sizeof(q_vector->name) - 1, "%s-%s-%d", basename, "rx", rx_int_idx++); - } else if (q_vector->tx.ring) { + } else if (q_vector->tx.tx_ring) { snprintf(q_vector->name, sizeof(q_vector->name) - 1, "%s-%s-%d", basename, "tx", tx_int_idx++); } else { /* skip this unused q_vector */ continue; } - if (vsi->type == ICE_VSI_CTRL && vsi->vf_id != ICE_INVAL_VFID) + if (vsi->type == ICE_VSI_CTRL && vsi->vf) err = devm_request_irq(dev, irq_num, vsi->irq_handler, IRQF_SHARED, q_vector->name, q_vector); @@ -2243,6 +2524,13 @@ static int ice_vsi_req_irq_msix(struct ice_vsi *vsi, char *basename) irq_set_affinity_hint(irq_num, &q_vector->affinity_mask); } + err = ice_set_cpu_rx_rmap(vsi); + if (err) { + netdev_err(vsi->netdev, "Failed to setup CPU RMAP on VSI %u: %pe\n", + vsi->vsi_num, ERR_PTR(err)); + goto free_q_irqs; + } + vsi->irqs_ready = true; return 0; @@ -2267,11 +2555,12 @@ free_q_irqs: static int ice_xdp_alloc_setup_rings(struct ice_vsi *vsi) { struct device *dev = ice_pf_to_dev(vsi->back); - int i; + struct ice_tx_desc *tx_desc; + int i, j; - for (i = 0; i < vsi->num_xdp_txq; i++) { + ice_for_each_xdp_txq(vsi, i) { u16 xdp_q_idx = vsi->alloc_txq + i; - struct ice_ring *xdp_ring; + struct ice_tx_ring *xdp_ring; xdp_ring = kzalloc(sizeof(*xdp_ring), GFP_KERNEL); @@ -2280,16 +2569,21 @@ static int ice_xdp_alloc_setup_rings(struct ice_vsi *vsi) xdp_ring->q_index = xdp_q_idx; xdp_ring->reg_idx = vsi->txq_map[xdp_q_idx]; - xdp_ring->ring_active = false; xdp_ring->vsi = vsi; xdp_ring->netdev = NULL; xdp_ring->dev = dev; xdp_ring->count = vsi->num_tx_desc; + xdp_ring->next_dd = ICE_RING_QUARTER(xdp_ring) - 1; + xdp_ring->next_rs = ICE_RING_QUARTER(xdp_ring) - 1; WRITE_ONCE(vsi->xdp_rings[i], xdp_ring); if (ice_setup_tx_ring(xdp_ring)) goto free_xdp_rings; ice_set_ring_xdp(xdp_ring); - xdp_ring->xsk_pool = ice_xsk_pool(xdp_ring); + spin_lock_init(&xdp_ring->tx_lock); + for (j = 0; j < xdp_ring->count; j++) { + tx_desc = ICE_TX_DESC(xdp_ring, j); + tx_desc->cmd_type_offset_bsz = 0; + } } return 0; @@ -2341,9 +2635,9 @@ int ice_prepare_xdp_rings(struct ice_vsi *vsi, struct bpf_prog *prog) .vsi_map_offset = vsi->alloc_txq, .mapping_mode = ICE_VSI_MAP_CONTIG }; - enum ice_status status; struct device *dev; int i, v_idx; + int status; dev = ice_pf_to_dev(pf); vsi->xdp_rings = devm_kcalloc(dev, vsi->num_xdp_txq, @@ -2355,6 +2649,10 @@ int ice_prepare_xdp_rings(struct ice_vsi *vsi, struct bpf_prog *prog) if (__ice_vsi_get_qs(&xdp_qs_cfg)) goto err_map_xdp; + if (static_key_enabled(&ice_xdp_locking_key)) + netdev_warn(vsi->netdev, + "Could not allocate one XDP Tx ring per CPU, XDP_TX/XDP_REDIRECT actions will be slower\n"); + if (ice_xdp_alloc_setup_rings(vsi)) goto clear_xdp_rings; @@ -2368,15 +2666,32 @@ int ice_prepare_xdp_rings(struct ice_vsi *vsi, struct bpf_prog *prog) q_base = vsi->num_xdp_txq - xdp_rings_rem; for (q_id = q_base; q_id < (q_base + xdp_rings_per_v); q_id++) { - struct ice_ring *xdp_ring = vsi->xdp_rings[q_id]; + struct ice_tx_ring *xdp_ring = vsi->xdp_rings[q_id]; xdp_ring->q_vector = q_vector; - xdp_ring->next = q_vector->tx.ring; - q_vector->tx.ring = xdp_ring; + xdp_ring->next = q_vector->tx.tx_ring; + q_vector->tx.tx_ring = xdp_ring; } xdp_rings_rem -= xdp_rings_per_v; } + ice_for_each_rxq(vsi, i) { + if (static_key_enabled(&ice_xdp_locking_key)) { + vsi->rx_rings[i]->xdp_ring = vsi->xdp_rings[i % vsi->num_xdp_txq]; + } else { + struct ice_q_vector *q_vector = vsi->rx_rings[i]->q_vector; + struct ice_tx_ring *ring; + + ice_for_each_tx_ring(ring, q_vector->tx) { + if (ice_ring_is_xdp(ring)) { + vsi->rx_rings[i]->xdp_ring = ring; + break; + } + } + } + ice_tx_xsk_pool(vsi, i); + } + /* omit the scheduler update if in reset path; XDP queues will be * taken into account at the end of ice_vsi_rebuild, where * ice_cfg_vsi_lan is being called @@ -2393,15 +2708,26 @@ int ice_prepare_xdp_rings(struct ice_vsi *vsi, struct bpf_prog *prog) status = ice_cfg_vsi_lan(vsi->port_info, vsi->idx, vsi->tc_cfg.ena_tc, max_txqs); if (status) { - dev_err(dev, "Failed VSI LAN queue config for XDP, error: %s\n", - ice_stat_str(status)); + dev_err(dev, "Failed VSI LAN queue config for XDP, error: %d\n", + status); goto clear_xdp_rings; } - ice_vsi_assign_bpf_prog(vsi, prog); + + /* assign the prog only when it's not already present on VSI; + * this flow is a subject of both ethtool -L and ndo_bpf flows; + * VSI rebuild that happens under ethtool -L can expose us to + * the bpf_prog refcount issues as we would be swapping same + * bpf_prog pointers from vsi->xdp_prog and calling bpf_prog_put + * on it as it would be treated as an 'old_prog'; for ndo_bpf + * this is not harmful as dev_xdp_install bumps the refcount + * before calling the op exposed by the driver; + */ + if (!ice_is_xdp_ena_vsi(vsi)) + ice_vsi_assign_bpf_prog(vsi, prog); return 0; clear_xdp_rings: - for (i = 0; i < vsi->num_xdp_txq; i++) + ice_for_each_xdp_txq(vsi, i) if (vsi->xdp_rings[i]) { kfree_rcu(vsi->xdp_rings[i], rcu); vsi->xdp_rings[i] = NULL; @@ -2409,7 +2735,7 @@ clear_xdp_rings: err_map_xdp: mutex_lock(&pf->avail_q_mutex); - for (i = 0; i < vsi->num_xdp_txq; i++) { + ice_for_each_xdp_txq(vsi, i) { clear_bit(vsi->txq_map[i + vsi->alloc_txq], pf->avail_txqs); vsi->txq_map[i + vsi->alloc_txq] = ICE_INVAL_Q_INDEX; } @@ -2442,28 +2768,30 @@ int ice_destroy_xdp_rings(struct ice_vsi *vsi) ice_for_each_q_vector(vsi, v_idx) { struct ice_q_vector *q_vector = vsi->q_vectors[v_idx]; - struct ice_ring *ring; + struct ice_tx_ring *ring; - ice_for_each_ring(ring, q_vector->tx) + ice_for_each_tx_ring(ring, q_vector->tx) if (!ring->tx_buf || !ice_ring_is_xdp(ring)) break; /* restore the value of last node prior to XDP setup */ - q_vector->tx.ring = ring; + q_vector->tx.tx_ring = ring; } free_qmap: mutex_lock(&pf->avail_q_mutex); - for (i = 0; i < vsi->num_xdp_txq; i++) { + ice_for_each_xdp_txq(vsi, i) { clear_bit(vsi->txq_map[i + vsi->alloc_txq], pf->avail_txqs); vsi->txq_map[i + vsi->alloc_txq] = ICE_INVAL_Q_INDEX; } mutex_unlock(&pf->avail_q_mutex); - for (i = 0; i < vsi->num_xdp_txq; i++) + ice_for_each_xdp_txq(vsi, i) if (vsi->xdp_rings[i]) { - if (vsi->xdp_rings[i]->desc) + if (vsi->xdp_rings[i]->desc) { + synchronize_rcu(); ice_free_tx_ring(vsi->xdp_rings[i]); + } kfree_rcu(vsi->xdp_rings[i], rcu); vsi->xdp_rings[i] = NULL; } @@ -2471,6 +2799,9 @@ free_qmap: devm_kfree(ice_pf_to_dev(pf), vsi->xdp_rings); vsi->xdp_rings = NULL; + if (static_key_enabled(&ice_xdp_locking_key)) + static_branch_dec(&ice_xdp_locking_key); + if (ice_is_reset_in_progress(pf->state) || !vsi->q_vectors[0]) return 0; @@ -2498,7 +2829,7 @@ static void ice_vsi_rx_napi_schedule(struct ice_vsi *vsi) int i; ice_for_each_rxq(vsi, i) { - struct ice_ring *rx_ring = vsi->rx_rings[i]; + struct ice_rx_ring *rx_ring = vsi->rx_rings[i]; if (rx_ring->xsk_pool) napi_schedule(&rx_ring->q_vector->napi); @@ -2506,6 +2837,29 @@ static void ice_vsi_rx_napi_schedule(struct ice_vsi *vsi) } /** + * ice_vsi_determine_xdp_res - figure out how many Tx qs can XDP have + * @vsi: VSI to determine the count of XDP Tx qs + * + * returns 0 if Tx qs count is higher than at least half of CPU count, + * -ENOMEM otherwise + */ +int ice_vsi_determine_xdp_res(struct ice_vsi *vsi) +{ + u16 avail = ice_get_avail_txq_count(vsi->back); + u16 cpus = num_possible_cpus(); + + if (avail < cpus / 2) + return -ENOMEM; + + vsi->num_xdp_txq = min_t(u16, avail, cpus); + + if (vsi->num_xdp_txq < cpus) + static_branch_inc(&ice_xdp_locking_key); + + return 0; +} + +/** * ice_xdp_setup_prog - Add or remove XDP eBPF program * @vsi: VSI to setup XDP for * @prog: XDP program @@ -2534,15 +2888,32 @@ ice_xdp_setup_prog(struct ice_vsi *vsi, struct bpf_prog *prog, } if (!ice_is_xdp_ena_vsi(vsi) && prog) { - vsi->num_xdp_txq = vsi->alloc_rxq; - xdp_ring_err = ice_prepare_xdp_rings(vsi, prog); + xdp_ring_err = ice_vsi_determine_xdp_res(vsi); + if (xdp_ring_err) { + NL_SET_ERR_MSG_MOD(extack, "Not enough Tx resources for XDP"); + } else { + xdp_ring_err = ice_prepare_xdp_rings(vsi, prog); + if (xdp_ring_err) + NL_SET_ERR_MSG_MOD(extack, "Setting up XDP Tx resources failed"); + } + /* reallocate Rx queues that are used for zero-copy */ + xdp_ring_err = ice_realloc_zc_buf(vsi, true); if (xdp_ring_err) - NL_SET_ERR_MSG_MOD(extack, "Setting up XDP Tx resources failed"); + NL_SET_ERR_MSG_MOD(extack, "Setting up XDP Rx resources failed"); } else if (ice_is_xdp_ena_vsi(vsi) && !prog) { xdp_ring_err = ice_destroy_xdp_rings(vsi); if (xdp_ring_err) NL_SET_ERR_MSG_MOD(extack, "Freeing XDP Tx resources failed"); + /* reallocate Rx queues that were used for zero-copy */ + xdp_ring_err = ice_realloc_zc_buf(vsi, false); + if (xdp_ring_err) + NL_SET_ERR_MSG_MOD(extack, "Freeing XDP Rx resources failed"); } else { + /* safe to call even when prog == vsi->xdp_prog as + * dev_xdp_install in net/core/dev.c incremented prog's + * refcount so corresponding bpf_prog_put won't cause + * underflow + */ ice_vsi_assign_bpf_prog(vsi, prog); } @@ -2556,6 +2927,20 @@ ice_xdp_setup_prog(struct ice_vsi *vsi, struct bpf_prog *prog, } /** + * ice_xdp_safe_mode - XDP handler for safe mode + * @dev: netdevice + * @xdp: XDP command + */ +static int ice_xdp_safe_mode(struct net_device __always_unused *dev, + struct netdev_bpf *xdp) +{ + NL_SET_ERR_MSG_MOD(xdp->extack, + "Please provide working DDP firmware package in order to use XDP\n" + "Refer to Documentation/networking/device_drivers/ethernet/intel/ice.rst"); + return -EOPNOTSUPP; +} + +/** * ice_xdp - implements XDP handler * @dev: netdevice * @xdp: XDP command @@ -2608,6 +2993,7 @@ static void ice_ena_misc_vector(struct ice_pf *pf) PFINT_OICR_PCI_EXCEPTION_M | PFINT_OICR_VFLR_M | PFINT_OICR_HMC_ERR_M | + PFINT_OICR_PE_PUSH_M | PFINT_OICR_PE_CRITERR_M); wr32(hw, PFINT_OICR_ENA, val); @@ -2633,6 +3019,7 @@ static irqreturn_t ice_misc_intr(int __always_unused irq, void *data) dev = ice_pf_to_dev(pf); set_bit(ICE_ADMINQ_EVENT_PENDING, pf->state); set_bit(ICE_MAILBOXQ_EVENT_PENDING, pf->state); + set_bit(ICE_SIDEBANDQ_EVENT_PENDING, pf->state); oicr = rd32(hw, PFINT_OICR); ena_mask = rd32(hw, PFINT_OICR_ENA); @@ -2678,8 +3065,6 @@ static irqreturn_t ice_misc_intr(int __always_unused irq, void *data) /* If a reset cycle isn't already in progress, we set a bit in * pf->state so that the service task can start a reset/rebuild. - * We also make note of which reset happened so that peer - * devices/drivers can be informed. */ if (!test_and_set_bit(ICE_RESET_OICR_RECV, pf->state)) { if (reset == ICE_RESET_CORER) @@ -2706,11 +3091,29 @@ static irqreturn_t ice_misc_intr(int __always_unused irq, void *data) } } - if (oicr & PFINT_OICR_HMC_ERR_M) { - ena_mask &= ~PFINT_OICR_HMC_ERR_M; - dev_dbg(dev, "HMC Error interrupt - info 0x%x, data 0x%x\n", - rd32(hw, PFHMC_ERRORINFO), - rd32(hw, PFHMC_ERRORDATA)); + if (oicr & PFINT_OICR_TSYN_TX_M) { + ena_mask &= ~PFINT_OICR_TSYN_TX_M; + if (!hw->reset_ongoing) + ret = IRQ_WAKE_THREAD; + } + + if (oicr & PFINT_OICR_TSYN_EVNT_M) { + u8 tmr_idx = hw->func_caps.ts_func_info.tmr_index_owned; + u32 gltsyn_stat = rd32(hw, GLTSYN_STAT(tmr_idx)); + + /* Save EVENTs from GTSYN register */ + pf->ptp.ext_ts_irq |= gltsyn_stat & (GLTSYN_STAT_EVENT0_M | + GLTSYN_STAT_EVENT1_M | + GLTSYN_STAT_EVENT2_M); + ena_mask &= ~PFINT_OICR_TSYN_EVNT_M; + kthread_queue_work(pf->ptp.kworker, &pf->ptp.extts_work); + } + +#define ICE_AUX_CRIT_ERR (PFINT_OICR_PE_CRITERR_M | PFINT_OICR_HMC_ERR_M | PFINT_OICR_PE_PUSH_M) + if (oicr & ICE_AUX_CRIT_ERR) { + pf->oicr_err_reg |= oicr; + set_bit(ICE_AUX_ERR_PENDING, pf->state); + ena_mask &= ~ICE_AUX_CRIT_ERR; } /* Report any remaining unexpected interrupts */ @@ -2720,14 +3123,14 @@ static irqreturn_t ice_misc_intr(int __always_unused irq, void *data) /* If a critical error is pending there is no choice but to * reset the device. */ - if (oicr & (PFINT_OICR_PE_CRITERR_M | - PFINT_OICR_PCI_EXCEPTION_M | + if (oicr & (PFINT_OICR_PCI_EXCEPTION_M | PFINT_OICR_ECC_ERR_M)) { set_bit(ICE_PFR_REQ, pf->state); ice_service_task_schedule(pf); } } - ret = IRQ_HANDLED; + if (!ret) + ret = IRQ_HANDLED; ice_service_task_schedule(pf); ice_irq_dynamic_ena(hw, NULL, NULL); @@ -2736,6 +3139,24 @@ static irqreturn_t ice_misc_intr(int __always_unused irq, void *data) } /** + * ice_misc_intr_thread_fn - misc interrupt thread function + * @irq: interrupt number + * @data: pointer to a q_vector + */ +static irqreturn_t ice_misc_intr_thread_fn(int __always_unused irq, void *data) +{ + irqreturn_t ret = IRQ_HANDLED; + struct ice_pf *pf = data; + bool irq_handled; + + irq_handled = ice_ptp_process_ts(pf); + if (!irq_handled) + ret = IRQ_WAKE_THREAD; + + return ret; +} + +/** * ice_dis_ctrlq_interrupts - disable control queue interrupts * @hw: pointer to HW structure */ @@ -2749,6 +3170,9 @@ static void ice_dis_ctrlq_interrupts(struct ice_hw *hw) wr32(hw, PFINT_MBX_CTL, rd32(hw, PFINT_MBX_CTL) & ~PFINT_MBX_CTL_CAUSE_ENA_M); + wr32(hw, PFINT_SB_CTL, + rd32(hw, PFINT_SB_CTL) & ~PFINT_SB_CTL_CAUSE_ENA_M); + /* disable Control queue Interrupt causes */ wr32(hw, PFINT_OICR_CTL, rd32(hw, PFINT_OICR_CTL) & ~PFINT_OICR_CTL_CAUSE_ENA_M); @@ -2803,6 +3227,11 @@ static void ice_ena_ctrlq_interrupts(struct ice_hw *hw, u16 reg_idx) PFINT_MBX_CTL_CAUSE_ENA_M); wr32(hw, PFINT_MBX_CTL, val); + /* This enables Sideband queue Interrupt causes */ + val = ((reg_idx & PFINT_SB_CTL_MSIX_INDX_M) | + PFINT_SB_CTL_CAUSE_ENA_M); + wr32(hw, PFINT_SB_CTL, val); + ice_flush(hw); } @@ -2839,10 +3268,12 @@ static int ice_req_irq_msix_misc(struct ice_pf *pf) pf->num_avail_sw_msix -= 1; pf->oicr_idx = (u16)oicr_idx; - err = devm_request_irq(dev, pf->msix_entries[pf->oicr_idx].vector, - ice_misc_intr, 0, pf->int_name, pf); + err = devm_request_threaded_irq(dev, + pf->msix_entries[pf->oicr_idx].vector, + ice_misc_intr, ice_misc_intr_thread_fn, + 0, pf->int_name, pf); if (err) { - dev_err(dev, "devm_request_irq for %s failed: %d\n", + dev_err(dev, "devm_request_threaded_irq for %s failed: %d\n", pf->int_name, err); ice_free_res(pf->irq_tracker, 1, ICE_RES_MISC_VEC_ID); pf->num_avail_sw_msix += 1; @@ -2879,7 +3310,7 @@ static void ice_napi_add(struct ice_vsi *vsi) ice_for_each_q_vector(vsi, v_idx) netif_napi_add(vsi->netdev, &vsi->q_vectors[v_idx]->napi, - ice_napi_poll, NAPI_POLL_WEIGHT); + ice_napi_poll); } /** @@ -2908,6 +3339,7 @@ static void ice_set_ops(struct net_device *netdev) static void ice_set_netdev_features(struct net_device *netdev) { struct ice_pf *pf = ice_netdev_to_pf(netdev); + bool is_dvm_ena = ice_is_dvm_ena(&pf->hw); netdev_features_t csumo_features; netdev_features_t vlano_features; netdev_features_t dflt_features; @@ -2934,6 +3366,10 @@ static void ice_set_netdev_features(struct net_device *netdev) NETIF_F_HW_VLAN_CTAG_TX | NETIF_F_HW_VLAN_CTAG_RX; + /* Enable CTAG/STAG filtering by default in Double VLAN Mode (DVM) */ + if (is_dvm_ena) + vlano_features |= NETIF_F_HW_VLAN_STAG_FILTER; + tso_features = NETIF_F_TSO | NETIF_F_TSO_ECN | NETIF_F_TSO6 | @@ -2953,15 +3389,35 @@ static void ice_set_netdev_features(struct net_device *netdev) vlano_features | tso_features; /* add support for HW_CSUM on packets with MPLS header */ - netdev->mpls_features = NETIF_F_HW_CSUM; + netdev->mpls_features = NETIF_F_HW_CSUM | + NETIF_F_TSO | + NETIF_F_TSO6; /* enable features */ netdev->features |= netdev->hw_features; + + netdev->hw_features |= NETIF_F_HW_TC; + netdev->hw_features |= NETIF_F_LOOPBACK; + /* encap and VLAN devices inherit default, csumo and tso features */ netdev->hw_enc_features |= dflt_features | csumo_features | tso_features; netdev->vlan_features |= dflt_features | csumo_features | tso_features; + + /* advertise support but don't enable by default since only one type of + * VLAN offload can be enabled at a time (i.e. CTAG or STAG). When one + * type turns on the other has to be turned off. This is enforced by the + * ice_fix_features() ndo callback. + */ + if (is_dvm_ena) + netdev->hw_features |= NETIF_F_HW_VLAN_STAG_RX | + NETIF_F_HW_VLAN_STAG_TX; + + /* Leave CRC / FCS stripping enabled by default, but allow the value to + * be changed at runtime + */ + netdev->hw_features |= NETIF_F_RXFCS; } /** @@ -2972,7 +3428,6 @@ static void ice_set_netdev_features(struct net_device *netdev) */ static int ice_cfg_netdev(struct ice_vsi *vsi) { - struct ice_pf *pf = vsi->back; struct ice_netdev_priv *np; struct net_device *netdev; u8 mac_addr[ETH_ALEN]; @@ -2992,9 +3447,9 @@ static int ice_cfg_netdev(struct ice_vsi *vsi) ice_set_ops(netdev); if (vsi->type == ICE_VSI_PF) { - SET_NETDEV_DEV(netdev, ice_pf_to_dev(pf)); + SET_NETDEV_DEV(netdev, ice_pf_to_dev(vsi->back)); ether_addr_copy(mac_addr, vsi->port_info->mac.perm_addr); - ether_addr_copy(netdev->dev_addr, mac_addr); + eth_hw_addr_set(netdev, mac_addr); ether_addr_copy(netdev->perm_addr, mac_addr); } @@ -3037,7 +3492,14 @@ void ice_fill_rss_lut(u8 *lut, u16 rss_table_size, u16 rss_size) static struct ice_vsi * ice_pf_vsi_setup(struct ice_pf *pf, struct ice_port_info *pi) { - return ice_vsi_setup(pf, pi, ICE_VSI_PF, ICE_INVAL_VFID); + return ice_vsi_setup(pf, pi, ICE_VSI_PF, NULL, NULL); +} + +static struct ice_vsi * +ice_chnl_vsi_setup(struct ice_pf *pf, struct ice_port_info *pi, + struct ice_channel *ch) +{ + return ice_vsi_setup(pf, pi, ICE_VSI_CHNL, NULL, ch); } /** @@ -3051,7 +3513,7 @@ ice_pf_vsi_setup(struct ice_pf *pf, struct ice_port_info *pi) static struct ice_vsi * ice_ctrl_vsi_setup(struct ice_pf *pf, struct ice_port_info *pi) { - return ice_vsi_setup(pf, pi, ICE_VSI_CTRL, ICE_INVAL_VFID); + return ice_vsi_setup(pf, pi, ICE_VSI_CTRL, NULL, NULL); } /** @@ -3065,42 +3527,68 @@ ice_ctrl_vsi_setup(struct ice_pf *pf, struct ice_port_info *pi) struct ice_vsi * ice_lb_vsi_setup(struct ice_pf *pf, struct ice_port_info *pi) { - return ice_vsi_setup(pf, pi, ICE_VSI_LB, ICE_INVAL_VFID); + return ice_vsi_setup(pf, pi, ICE_VSI_LB, NULL, NULL); } /** * ice_vlan_rx_add_vid - Add a VLAN ID filter to HW offload * @netdev: network interface to be adjusted - * @proto: unused protocol + * @proto: VLAN TPID * @vid: VLAN ID to be added * * net_device_ops implementation for adding VLAN IDs */ static int -ice_vlan_rx_add_vid(struct net_device *netdev, __always_unused __be16 proto, - u16 vid) +ice_vlan_rx_add_vid(struct net_device *netdev, __be16 proto, u16 vid) { struct ice_netdev_priv *np = netdev_priv(netdev); + struct ice_vsi_vlan_ops *vlan_ops; struct ice_vsi *vsi = np->vsi; + struct ice_vlan vlan; int ret; /* VLAN 0 is added by default during load/reset */ if (!vid) return 0; - /* Enable VLAN pruning when a VLAN other than 0 is added */ - if (!ice_vsi_is_vlan_pruning_ena(vsi)) { - ret = ice_cfg_vlan_pruning(vsi, true, false); + while (test_and_set_bit(ICE_CFG_BUSY, vsi->state)) + usleep_range(1000, 2000); + + /* Add multicast promisc rule for the VLAN ID to be added if + * all-multicast is currently enabled. + */ + if (vsi->current_netdev_flags & IFF_ALLMULTI) { + ret = ice_fltr_set_vsi_promisc(&vsi->back->hw, vsi->idx, + ICE_MCAST_VLAN_PROMISC_BITS, + vid); if (ret) - return ret; + goto finish; } + vlan_ops = ice_get_compat_vsi_vlan_ops(vsi); + /* Add a switch rule for this VLAN ID so its corresponding VLAN tagged * packets aren't pruned by the device's internal switch on Rx */ - ret = ice_vsi_add_vlan(vsi, vid, ICE_FWD_TO_VSI); - if (!ret) - set_bit(ICE_VSI_VLAN_FLTR_CHANGED, vsi->state); + vlan = ICE_VLAN(be16_to_cpu(proto), vid, 0); + ret = vlan_ops->add_vlan(vsi, &vlan); + if (ret) + goto finish; + + /* If all-multicast is currently enabled and this VLAN ID is only one + * besides VLAN-0 we have to update look-up type of multicast promisc + * rule for VLAN-0 from ICE_SW_LKUP_PROMISC to ICE_SW_LKUP_PROMISC_VLAN. + */ + if ((vsi->current_netdev_flags & IFF_ALLMULTI) && + ice_vsi_num_non_zero_vlans(vsi) == 1) { + ice_fltr_clear_vsi_promisc(&vsi->back->hw, vsi->idx, + ICE_MCAST_PROMISC_BITS, 0); + ice_fltr_set_vsi_promisc(&vsi->back->hw, vsi->idx, + ICE_MCAST_VLAN_PROMISC_BITS, 0); + } + +finish: + clear_bit(ICE_CFG_BUSY, vsi->state); return ret; } @@ -3108,39 +3596,130 @@ ice_vlan_rx_add_vid(struct net_device *netdev, __always_unused __be16 proto, /** * ice_vlan_rx_kill_vid - Remove a VLAN ID filter from HW offload * @netdev: network interface to be adjusted - * @proto: unused protocol + * @proto: VLAN TPID * @vid: VLAN ID to be removed * * net_device_ops implementation for removing VLAN IDs */ static int -ice_vlan_rx_kill_vid(struct net_device *netdev, __always_unused __be16 proto, - u16 vid) +ice_vlan_rx_kill_vid(struct net_device *netdev, __be16 proto, u16 vid) { struct ice_netdev_priv *np = netdev_priv(netdev); + struct ice_vsi_vlan_ops *vlan_ops; struct ice_vsi *vsi = np->vsi; + struct ice_vlan vlan; int ret; /* don't allow removal of VLAN 0 */ if (!vid) return 0; - /* Make sure ice_vsi_kill_vlan is successful before updating VLAN + while (test_and_set_bit(ICE_CFG_BUSY, vsi->state)) + usleep_range(1000, 2000); + + ret = ice_clear_vsi_promisc(&vsi->back->hw, vsi->idx, + ICE_MCAST_VLAN_PROMISC_BITS, vid); + if (ret) { + netdev_err(netdev, "Error clearing multicast promiscuous mode on VSI %i\n", + vsi->vsi_num); + vsi->current_netdev_flags |= IFF_ALLMULTI; + } + + vlan_ops = ice_get_compat_vsi_vlan_ops(vsi); + + /* Make sure VLAN delete is successful before updating VLAN * information */ - ret = ice_vsi_kill_vlan(vsi, vid); + vlan = ICE_VLAN(be16_to_cpu(proto), vid, 0); + ret = vlan_ops->del_vlan(vsi, &vlan); if (ret) - return ret; + goto finish; + + /* Remove multicast promisc rule for the removed VLAN ID if + * all-multicast is enabled. + */ + if (vsi->current_netdev_flags & IFF_ALLMULTI) + ice_fltr_clear_vsi_promisc(&vsi->back->hw, vsi->idx, + ICE_MCAST_VLAN_PROMISC_BITS, vid); + + if (!ice_vsi_has_non_zero_vlans(vsi)) { + /* Update look-up type of multicast promisc rule for VLAN 0 + * from ICE_SW_LKUP_PROMISC_VLAN to ICE_SW_LKUP_PROMISC when + * all-multicast is enabled and VLAN 0 is the only VLAN rule. + */ + if (vsi->current_netdev_flags & IFF_ALLMULTI) { + ice_fltr_clear_vsi_promisc(&vsi->back->hw, vsi->idx, + ICE_MCAST_VLAN_PROMISC_BITS, + 0); + ice_fltr_set_vsi_promisc(&vsi->back->hw, vsi->idx, + ICE_MCAST_PROMISC_BITS, 0); + } + } - /* Disable pruning when VLAN 0 is the only VLAN rule */ - if (vsi->num_vlan == 1 && ice_vsi_is_vlan_pruning_ena(vsi)) - ret = ice_cfg_vlan_pruning(vsi, false, false); +finish: + clear_bit(ICE_CFG_BUSY, vsi->state); - set_bit(ICE_VSI_VLAN_FLTR_CHANGED, vsi->state); return ret; } /** + * ice_rep_indr_tc_block_unbind + * @cb_priv: indirection block private data + */ +static void ice_rep_indr_tc_block_unbind(void *cb_priv) +{ + struct ice_indr_block_priv *indr_priv = cb_priv; + + list_del(&indr_priv->list); + kfree(indr_priv); +} + +/** + * ice_tc_indir_block_unregister - Unregister TC indirect block notifications + * @vsi: VSI struct which has the netdev + */ +static void ice_tc_indir_block_unregister(struct ice_vsi *vsi) +{ + struct ice_netdev_priv *np = netdev_priv(vsi->netdev); + + flow_indr_dev_unregister(ice_indr_setup_tc_cb, np, + ice_rep_indr_tc_block_unbind); +} + +/** + * ice_tc_indir_block_remove - clean indirect TC block notifications + * @pf: PF structure + */ +static void ice_tc_indir_block_remove(struct ice_pf *pf) +{ + struct ice_vsi *pf_vsi = ice_get_main_vsi(pf); + + if (!pf_vsi) + return; + + ice_tc_indir_block_unregister(pf_vsi); +} + +/** + * ice_tc_indir_block_register - Register TC indirect block notifications + * @vsi: VSI struct which has the netdev + * + * Returns 0 on success, negative value on failure + */ +static int ice_tc_indir_block_register(struct ice_vsi *vsi) +{ + struct ice_netdev_priv *np; + + if (!vsi || !vsi->netdev) + return -EINVAL; + + np = netdev_priv(vsi->netdev); + + INIT_LIST_HEAD(&np->tc_indr_block_priv_list); + return flow_indr_dev_register(ice_indr_setup_tc_cb, np); +} + +/** * ice_setup_pf_sw - Setup the HW switch on startup or after reset * @pf: board private structure * @@ -3148,24 +3727,38 @@ ice_vlan_rx_kill_vid(struct net_device *netdev, __always_unused __be16 proto, */ static int ice_setup_pf_sw(struct ice_pf *pf) { + struct device *dev = ice_pf_to_dev(pf); + bool dvm = ice_is_dvm_ena(&pf->hw); struct ice_vsi *vsi; - int status = 0; + int status; if (ice_is_reset_in_progress(pf->state)) return -EBUSY; + status = ice_aq_set_port_params(pf->hw.port_info, dvm, NULL); + if (status) + return -EIO; + vsi = ice_pf_vsi_setup(pf, pf->hw.port_info); if (!vsi) return -ENOMEM; + /* init channel list */ + INIT_LIST_HEAD(&vsi->ch_list); + status = ice_cfg_netdev(vsi); - if (status) { - status = -ENODEV; + if (status) goto unroll_vsi_setup; - } /* netdev has to be configured before setting frame size */ ice_vsi_cfg_frame_size(vsi); + /* init indirect block notifications */ + status = ice_tc_indir_block_register(vsi); + if (status) { + dev_err(dev, "Failed to register netdev notifier\n"); + goto unroll_cfg_netdev; + } + /* Setup DCB netlink interface */ ice_dcbnl_setup(vsi); @@ -3175,23 +3768,15 @@ static int ice_setup_pf_sw(struct ice_pf *pf) */ ice_napi_add(vsi); - status = ice_set_cpu_rx_rmap(vsi); - if (status) { - dev_err(ice_pf_to_dev(pf), "Failed to set CPU Rx map VSI %d error %d\n", - vsi->vsi_num, status); - status = -EINVAL; - goto unroll_napi_add; - } status = ice_init_mac_fltr(pf); if (status) - goto free_cpu_rx_map; - - return status; + goto unroll_napi_add; -free_cpu_rx_map: - ice_free_cpu_rx_rmap(vsi); + return 0; unroll_napi_add: + ice_tc_indir_block_unregister(vsi); +unroll_cfg_netdev: if (vsi) { ice_napi_del(vsi); if (vsi->netdev) { @@ -3253,9 +3838,11 @@ u16 ice_get_avail_rxq_count(struct ice_pf *pf) static void ice_deinit_pf(struct ice_pf *pf) { ice_service_task_stop(pf); + mutex_destroy(&pf->adev_mutex); mutex_destroy(&pf->sw_mutex); mutex_destroy(&pf->tc_mutex); mutex_destroy(&pf->avail_q_mutex); + mutex_destroy(&pf->vfs.table_lock); if (pf->avail_txqs) { bitmap_free(pf->avail_txqs); @@ -3266,6 +3853,9 @@ static void ice_deinit_pf(struct ice_pf *pf) bitmap_free(pf->avail_rxqs); pf->avail_rxqs = NULL; } + + if (pf->ptp.clock) + ptp_clock_unregister(pf->ptp.clock); } /** @@ -3276,14 +3866,17 @@ static void ice_set_pf_caps(struct ice_pf *pf) { struct ice_hw_func_caps *func_caps = &pf->hw.func_caps; + clear_bit(ICE_FLAG_RDMA_ENA, pf->flags); + if (func_caps->common_cap.rdma) + set_bit(ICE_FLAG_RDMA_ENA, pf->flags); clear_bit(ICE_FLAG_DCB_CAPABLE, pf->flags); if (func_caps->common_cap.dcb) set_bit(ICE_FLAG_DCB_CAPABLE, pf->flags); clear_bit(ICE_FLAG_SRIOV_CAPABLE, pf->flags); if (func_caps->common_cap.sr_iov_1_1) { set_bit(ICE_FLAG_SRIOV_CAPABLE, pf->flags); - pf->num_vfs_supported = min_t(int, func_caps->num_allocd_vfs, - ICE_MAX_VF_COUNT); + pf->vfs.num_supported = min_t(int, func_caps->num_allocd_vfs, + ICE_MAX_SRIOV_VFS); } clear_bit(ICE_FLAG_RSS_ENA, pf->flags); if (func_caps->common_cap.rss_table_size) @@ -3306,6 +3899,10 @@ static void ice_set_pf_caps(struct ice_pf *pf) func_caps->fd_fltr_best_effort); } + clear_bit(ICE_FLAG_PTP_SUPPORTED, pf->flags); + if (func_caps->common_cap.ieee_1588) + set_bit(ICE_FLAG_PTP_SUPPORTED, pf->flags); + pf->max_pf_txqs = func_caps->common_cap.num_txq; pf->max_pf_rxqs = func_caps->common_cap.num_rxq; } @@ -3320,11 +3917,14 @@ static int ice_init_pf(struct ice_pf *pf) mutex_init(&pf->sw_mutex); mutex_init(&pf->tc_mutex); + mutex_init(&pf->adev_mutex); INIT_HLIST_HEAD(&pf->aq_wait_list); spin_lock_init(&pf->aq_wait_lock); init_waitqueue_head(&pf->aq_wait_queue); + init_waitqueue_head(&pf->reset_wait_queue); + /* setup service timer and periodic service task */ timer_setup(&pf->serv_tmr, ice_service_timer, 0); pf->serv_tmr_period = HZ; @@ -3338,78 +3938,146 @@ static int ice_init_pf(struct ice_pf *pf) pf->avail_rxqs = bitmap_zalloc(pf->max_pf_rxqs, GFP_KERNEL); if (!pf->avail_rxqs) { - devm_kfree(ice_pf_to_dev(pf), pf->avail_txqs); + bitmap_free(pf->avail_txqs); pf->avail_txqs = NULL; return -ENOMEM; } + mutex_init(&pf->vfs.table_lock); + hash_init(pf->vfs.table); + return 0; } /** + * ice_reduce_msix_usage - Reduce usage of MSI-X vectors + * @pf: board private structure + * @v_remain: number of remaining MSI-X vectors to be distributed + * + * Reduce the usage of MSI-X vectors when entire request cannot be fulfilled. + * pf->num_lan_msix and pf->num_rdma_msix values are set based on number of + * remaining vectors. + */ +static void ice_reduce_msix_usage(struct ice_pf *pf, int v_remain) +{ + int v_rdma; + + if (!ice_is_rdma_ena(pf)) { + pf->num_lan_msix = v_remain; + return; + } + + /* RDMA needs at least 1 interrupt in addition to AEQ MSIX */ + v_rdma = ICE_RDMA_NUM_AEQ_MSIX + 1; + + if (v_remain < ICE_MIN_LAN_TXRX_MSIX + ICE_MIN_RDMA_MSIX) { + dev_warn(ice_pf_to_dev(pf), "Not enough MSI-X vectors to support RDMA.\n"); + clear_bit(ICE_FLAG_RDMA_ENA, pf->flags); + + pf->num_rdma_msix = 0; + pf->num_lan_msix = ICE_MIN_LAN_TXRX_MSIX; + } else if ((v_remain < ICE_MIN_LAN_TXRX_MSIX + v_rdma) || + (v_remain - v_rdma < v_rdma)) { + /* Support minimum RDMA and give remaining vectors to LAN MSIX */ + pf->num_rdma_msix = ICE_MIN_RDMA_MSIX; + pf->num_lan_msix = v_remain - ICE_MIN_RDMA_MSIX; + } else { + /* Split remaining MSIX with RDMA after accounting for AEQ MSIX + */ + pf->num_rdma_msix = (v_remain - ICE_RDMA_NUM_AEQ_MSIX) / 2 + + ICE_RDMA_NUM_AEQ_MSIX; + pf->num_lan_msix = v_remain - pf->num_rdma_msix; + } +} + +/** * ice_ena_msix_range - Request a range of MSIX vectors from the OS * @pf: board private structure * - * compute the number of MSIX vectors required (v_budget) and request from - * the OS. Return the number of vectors reserved or negative on failure + * Compute the number of MSIX vectors wanted and request from the OS. Adjust + * device usage if there are not enough vectors. Return the number of vectors + * reserved or negative on failure. */ static int ice_ena_msix_range(struct ice_pf *pf) { - int v_left, v_actual, v_other, v_budget = 0; + int num_cpus, hw_num_msix, v_other, v_wanted, v_actual; struct device *dev = ice_pf_to_dev(pf); - int needed, err, i; + int err, i; - v_left = pf->hw.func_caps.common_cap.num_msix_vectors; + hw_num_msix = pf->hw.func_caps.common_cap.num_msix_vectors; + num_cpus = num_online_cpus(); - /* reserve for LAN miscellaneous handler */ - needed = ICE_MIN_LAN_OICR_MSIX; - if (v_left < needed) - goto no_hw_vecs_left_err; - v_budget += needed; - v_left -= needed; + /* LAN miscellaneous handler */ + v_other = ICE_MIN_LAN_OICR_MSIX; - /* reserve for flow director */ - if (test_bit(ICE_FLAG_FD_ENA, pf->flags)) { - needed = ICE_FDIR_MSIX; - if (v_left < needed) - goto no_hw_vecs_left_err; - v_budget += needed; - v_left -= needed; + /* Flow Director */ + if (test_bit(ICE_FLAG_FD_ENA, pf->flags)) + v_other += ICE_FDIR_MSIX; + + /* switchdev */ + v_other += ICE_ESWITCH_MSIX; + + v_wanted = v_other; + + /* LAN traffic */ + pf->num_lan_msix = num_cpus; + v_wanted += pf->num_lan_msix; + + /* RDMA auxiliary driver */ + if (ice_is_rdma_ena(pf)) { + pf->num_rdma_msix = num_cpus + ICE_RDMA_NUM_AEQ_MSIX; + v_wanted += pf->num_rdma_msix; } - /* total used for non-traffic vectors */ - v_other = v_budget; + if (v_wanted > hw_num_msix) { + int v_remain; - /* reserve vectors for LAN traffic */ - needed = min_t(int, num_online_cpus(), v_left); - if (v_left < needed) - goto no_hw_vecs_left_err; - pf->num_lan_msix = needed; - v_budget += needed; - v_left -= needed; + dev_warn(dev, "not enough device MSI-X vectors. wanted = %d, available = %d\n", + v_wanted, hw_num_msix); - pf->msix_entries = devm_kcalloc(dev, v_budget, + if (hw_num_msix < ICE_MIN_MSIX) { + err = -ERANGE; + goto exit_err; + } + + v_remain = hw_num_msix - v_other; + if (v_remain < ICE_MIN_LAN_TXRX_MSIX) { + v_other = ICE_MIN_MSIX - ICE_MIN_LAN_TXRX_MSIX; + v_remain = ICE_MIN_LAN_TXRX_MSIX; + } + + ice_reduce_msix_usage(pf, v_remain); + v_wanted = pf->num_lan_msix + pf->num_rdma_msix + v_other; + + dev_notice(dev, "Reducing request to %d MSI-X vectors for LAN traffic.\n", + pf->num_lan_msix); + if (ice_is_rdma_ena(pf)) + dev_notice(dev, "Reducing request to %d MSI-X vectors for RDMA.\n", + pf->num_rdma_msix); + } + + pf->msix_entries = devm_kcalloc(dev, v_wanted, sizeof(*pf->msix_entries), GFP_KERNEL); if (!pf->msix_entries) { err = -ENOMEM; goto exit_err; } - for (i = 0; i < v_budget; i++) + for (i = 0; i < v_wanted; i++) pf->msix_entries[i].entry = i; /* actually reserve the vectors */ v_actual = pci_enable_msix_range(pf->pdev, pf->msix_entries, - ICE_MIN_MSIX, v_budget); + ICE_MIN_MSIX, v_wanted); if (v_actual < 0) { dev_err(dev, "unable to reserve MSI-X vectors\n"); err = v_actual; goto msix_err; } - if (v_actual < v_budget) { + if (v_actual < v_wanted) { dev_warn(dev, "not enough OS MSI-X vectors. requested = %d, obtained = %d\n", - v_budget, v_actual); + v_wanted, v_actual); if (v_actual < ICE_MIN_MSIX) { /* error if we can't get minimum vectors */ @@ -3417,16 +4085,19 @@ static int ice_ena_msix_range(struct ice_pf *pf) err = -ERANGE; goto msix_err; } else { - int v_traffic = v_actual - v_other; + int v_remain = v_actual - v_other; - if (v_actual == ICE_MIN_MSIX || - v_traffic < ICE_MIN_LAN_TXRX_MSIX) - pf->num_lan_msix = ICE_MIN_LAN_TXRX_MSIX; - else - pf->num_lan_msix = v_traffic; + if (v_remain < ICE_MIN_LAN_TXRX_MSIX) + v_remain = ICE_MIN_LAN_TXRX_MSIX; + + ice_reduce_msix_usage(pf, v_remain); dev_notice(dev, "Enabled %d MSI-X vectors for LAN traffic.\n", pf->num_lan_msix); + + if (ice_is_rdma_ena(pf)) + dev_notice(dev, "Enabled %d MSI-X vectors for RDMA.\n", + pf->num_rdma_msix); } } @@ -3434,13 +4105,9 @@ static int ice_ena_msix_range(struct ice_pf *pf) msix_err: devm_kfree(dev, pf->msix_entries); - goto exit_err; -no_hw_vecs_left_err: - dev_err(dev, "not enough device MSI-X vectors. requested = %d, available = %d\n", - needed, v_left); - err = -ERANGE; exit_err: + pf->num_rdma_msix = 0; pf->num_lan_msix = 0; return err; } @@ -3577,8 +4244,8 @@ static void ice_set_safe_mode_vlan_cfg(struct ice_pf *pf) { struct ice_vsi *vsi = ice_get_main_vsi(pf); struct ice_vsi_ctx *ctxt; - enum ice_status status; struct ice_hw *hw; + int status; if (!vsi) return; @@ -3603,18 +4270,17 @@ static void ice_set_safe_mode_vlan_cfg(struct ice_pf *pf) ctxt->info.sw_flags2 &= ~ICE_AQ_VSI_SW_FLAG_RX_VLAN_PRUNE_ENA; /* allow all VLANs on Tx and don't strip on Rx */ - ctxt->info.vlan_flags = ICE_AQ_VSI_VLAN_MODE_ALL | - ICE_AQ_VSI_VLAN_EMOD_NOTHING; + ctxt->info.inner_vlan_flags = ICE_AQ_VSI_INNER_VLAN_TX_MODE_ALL | + ICE_AQ_VSI_INNER_VLAN_EMODE_NOTHING; status = ice_update_vsi(hw, vsi->idx, ctxt, NULL); if (status) { - dev_err(ice_pf_to_dev(vsi->back), "Failed to update VSI for safe mode VLANs, err %s aq_err %s\n", - ice_stat_str(status), - ice_aq_str(hw->adminq.sq_last_status)); + dev_err(ice_pf_to_dev(vsi->back), "Failed to update VSI for safe mode VLANs, err %d aq_err %s\n", + status, ice_aq_str(hw->adminq.sq_last_status)); } else { vsi->info.sec_flags = ctxt->info.sec_flags; vsi->info.sw_flags2 = ctxt->info.sw_flags2; - vsi->info.vlan_flags = ctxt->info.vlan_flags; + vsi->info.inner_vlan_flags = ctxt->info.inner_vlan_flags; } kfree(ctxt); @@ -3623,108 +4289,80 @@ static void ice_set_safe_mode_vlan_cfg(struct ice_pf *pf) /** * ice_log_pkg_init - log result of DDP package load * @hw: pointer to hardware info - * @status: status of package load + * @state: state of package load */ -static void -ice_log_pkg_init(struct ice_hw *hw, enum ice_status *status) +static void ice_log_pkg_init(struct ice_hw *hw, enum ice_ddp_state state) { - struct ice_pf *pf = (struct ice_pf *)hw->back; - struct device *dev = ice_pf_to_dev(pf); + struct ice_pf *pf = hw->back; + struct device *dev; - switch (*status) { - case ICE_SUCCESS: - /* The package download AdminQ command returned success because - * this download succeeded or ICE_ERR_AQ_NO_WORK since there is - * already a package loaded on the device. - */ - if (hw->pkg_ver.major == hw->active_pkg_ver.major && - hw->pkg_ver.minor == hw->active_pkg_ver.minor && - hw->pkg_ver.update == hw->active_pkg_ver.update && - hw->pkg_ver.draft == hw->active_pkg_ver.draft && - !memcmp(hw->pkg_name, hw->active_pkg_name, - sizeof(hw->pkg_name))) { - if (hw->pkg_dwnld_status == ICE_AQ_RC_EEXIST) - dev_info(dev, "DDP package already present on device: %s version %d.%d.%d.%d\n", - hw->active_pkg_name, - hw->active_pkg_ver.major, - hw->active_pkg_ver.minor, - hw->active_pkg_ver.update, - hw->active_pkg_ver.draft); - else - dev_info(dev, "The DDP package was successfully loaded: %s version %d.%d.%d.%d\n", - hw->active_pkg_name, - hw->active_pkg_ver.major, - hw->active_pkg_ver.minor, - hw->active_pkg_ver.update, - hw->active_pkg_ver.draft); - } else if (hw->active_pkg_ver.major != ICE_PKG_SUPP_VER_MAJ || - hw->active_pkg_ver.minor != ICE_PKG_SUPP_VER_MNR) { - dev_err(dev, "The device has a DDP package that is not supported by the driver. The device has package '%s' version %d.%d.x.x. The driver requires version %d.%d.x.x. Entering Safe Mode.\n", - hw->active_pkg_name, - hw->active_pkg_ver.major, - hw->active_pkg_ver.minor, - ICE_PKG_SUPP_VER_MAJ, ICE_PKG_SUPP_VER_MNR); - *status = ICE_ERR_NOT_SUPPORTED; - } else if (hw->active_pkg_ver.major == ICE_PKG_SUPP_VER_MAJ && - hw->active_pkg_ver.minor == ICE_PKG_SUPP_VER_MNR) { - dev_info(dev, "The driver could not load the DDP package file because a compatible DDP package is already present on the device. The device has package '%s' version %d.%d.%d.%d. The package file found by the driver: '%s' version %d.%d.%d.%d.\n", - hw->active_pkg_name, - hw->active_pkg_ver.major, - hw->active_pkg_ver.minor, - hw->active_pkg_ver.update, - hw->active_pkg_ver.draft, - hw->pkg_name, - hw->pkg_ver.major, - hw->pkg_ver.minor, - hw->pkg_ver.update, - hw->pkg_ver.draft); - } else { - dev_err(dev, "An unknown error occurred when loading the DDP package, please reboot the system. If the problem persists, update the NVM. Entering Safe Mode.\n"); - *status = ICE_ERR_NOT_SUPPORTED; - } + dev = ice_pf_to_dev(pf); + + switch (state) { + case ICE_DDP_PKG_SUCCESS: + dev_info(dev, "The DDP package was successfully loaded: %s version %d.%d.%d.%d\n", + hw->active_pkg_name, + hw->active_pkg_ver.major, + hw->active_pkg_ver.minor, + hw->active_pkg_ver.update, + hw->active_pkg_ver.draft); + break; + case ICE_DDP_PKG_SAME_VERSION_ALREADY_LOADED: + dev_info(dev, "DDP package already present on device: %s version %d.%d.%d.%d\n", + hw->active_pkg_name, + hw->active_pkg_ver.major, + hw->active_pkg_ver.minor, + hw->active_pkg_ver.update, + hw->active_pkg_ver.draft); + break; + case ICE_DDP_PKG_ALREADY_LOADED_NOT_SUPPORTED: + dev_err(dev, "The device has a DDP package that is not supported by the driver. The device has package '%s' version %d.%d.x.x. The driver requires version %d.%d.x.x. Entering Safe Mode.\n", + hw->active_pkg_name, + hw->active_pkg_ver.major, + hw->active_pkg_ver.minor, + ICE_PKG_SUPP_VER_MAJ, ICE_PKG_SUPP_VER_MNR); break; - case ICE_ERR_FW_DDP_MISMATCH: + case ICE_DDP_PKG_COMPATIBLE_ALREADY_LOADED: + dev_info(dev, "The driver could not load the DDP package file because a compatible DDP package is already present on the device. The device has package '%s' version %d.%d.%d.%d. The package file found by the driver: '%s' version %d.%d.%d.%d.\n", + hw->active_pkg_name, + hw->active_pkg_ver.major, + hw->active_pkg_ver.minor, + hw->active_pkg_ver.update, + hw->active_pkg_ver.draft, + hw->pkg_name, + hw->pkg_ver.major, + hw->pkg_ver.minor, + hw->pkg_ver.update, + hw->pkg_ver.draft); + break; + case ICE_DDP_PKG_FW_MISMATCH: dev_err(dev, "The firmware loaded on the device is not compatible with the DDP package. Please update the device's NVM. Entering safe mode.\n"); break; - case ICE_ERR_BUF_TOO_SHORT: - case ICE_ERR_CFG: + case ICE_DDP_PKG_INVALID_FILE: dev_err(dev, "The DDP package file is invalid. Entering Safe Mode.\n"); break; - case ICE_ERR_NOT_SUPPORTED: - /* Package File version not supported */ - if (hw->pkg_ver.major > ICE_PKG_SUPP_VER_MAJ || - (hw->pkg_ver.major == ICE_PKG_SUPP_VER_MAJ && - hw->pkg_ver.minor > ICE_PKG_SUPP_VER_MNR)) - dev_err(dev, "The DDP package file version is higher than the driver supports. Please use an updated driver. Entering Safe Mode.\n"); - else if (hw->pkg_ver.major < ICE_PKG_SUPP_VER_MAJ || - (hw->pkg_ver.major == ICE_PKG_SUPP_VER_MAJ && - hw->pkg_ver.minor < ICE_PKG_SUPP_VER_MNR)) - dev_err(dev, "The DDP package file version is lower than the driver supports. The driver requires version %d.%d.x.x. Please use an updated DDP Package file. Entering Safe Mode.\n", - ICE_PKG_SUPP_VER_MAJ, ICE_PKG_SUPP_VER_MNR); + case ICE_DDP_PKG_FILE_VERSION_TOO_HIGH: + dev_err(dev, "The DDP package file version is higher than the driver supports. Please use an updated driver. Entering Safe Mode.\n"); break; - case ICE_ERR_AQ_ERROR: - switch (hw->pkg_dwnld_status) { - case ICE_AQ_RC_ENOSEC: - case ICE_AQ_RC_EBADSIG: - dev_err(dev, "The DDP package could not be loaded because its signature is not valid. Please use a valid DDP Package. Entering Safe Mode.\n"); - return; - case ICE_AQ_RC_ESVN: - dev_err(dev, "The DDP Package could not be loaded because its security revision is too low. Please use an updated DDP Package. Entering Safe Mode.\n"); - return; - case ICE_AQ_RC_EBADMAN: - case ICE_AQ_RC_EBADBUF: - dev_err(dev, "An error occurred on the device while loading the DDP package. The device will be reset.\n"); - /* poll for reset to complete */ - if (ice_check_reset(hw)) - dev_err(dev, "Error resetting device. Please reload the driver\n"); - return; - default: - break; - } - fallthrough; + case ICE_DDP_PKG_FILE_VERSION_TOO_LOW: + dev_err(dev, "The DDP package file version is lower than the driver supports. The driver requires version %d.%d.x.x. Please use an updated DDP Package file. Entering Safe Mode.\n", + ICE_PKG_SUPP_VER_MAJ, ICE_PKG_SUPP_VER_MNR); + break; + case ICE_DDP_PKG_FILE_SIGNATURE_INVALID: + dev_err(dev, "The DDP package could not be loaded because its signature is not valid. Please use a valid DDP Package. Entering Safe Mode.\n"); + break; + case ICE_DDP_PKG_FILE_REVISION_TOO_LOW: + dev_err(dev, "The DDP Package could not be loaded because its security revision is too low. Please use an updated DDP Package. Entering Safe Mode.\n"); + break; + case ICE_DDP_PKG_LOAD_ERROR: + dev_err(dev, "An error occurred on the device while loading the DDP package. The device will be reset.\n"); + /* poll for reset to complete */ + if (ice_check_reset(hw)) + dev_err(dev, "Error resetting device. Please reload the driver\n"); + break; + case ICE_DDP_PKG_ERR: default: - dev_err(dev, "An unknown error (%d) occurred when loading the DDP package. Entering Safe Mode.\n", - *status); + dev_err(dev, "An unknown error occurred when loading the DDP package. Entering Safe Mode.\n"); break; } } @@ -3740,24 +4378,24 @@ ice_log_pkg_init(struct ice_hw *hw, enum ice_status *status) static void ice_load_pkg(const struct firmware *firmware, struct ice_pf *pf) { - enum ice_status status = ICE_ERR_PARAM; + enum ice_ddp_state state = ICE_DDP_PKG_ERR; struct device *dev = ice_pf_to_dev(pf); struct ice_hw *hw = &pf->hw; /* Load DDP Package */ if (firmware && !hw->pkg_copy) { - status = ice_copy_and_init_pkg(hw, firmware->data, - firmware->size); - ice_log_pkg_init(hw, &status); + state = ice_copy_and_init_pkg(hw, firmware->data, + firmware->size); + ice_log_pkg_init(hw, state); } else if (!firmware && hw->pkg_copy) { /* Reload package during rebuild after CORER/GLOBR reset */ - status = ice_init_pkg(hw, hw->pkg_copy, hw->pkg_size); - ice_log_pkg_init(hw, &status); + state = ice_init_pkg(hw, hw->pkg_copy, hw->pkg_size); + ice_log_pkg_init(hw, state); } else { dev_err(dev, "The DDP package file failed to load. Entering Safe Mode.\n"); } - if (status) { + if (!ice_is_init_pkg_successful(state)) { /* Safe Mode */ clear_bit(ICE_FLAG_ADV_FEATURES, pf->flags); return; @@ -3788,9 +4426,9 @@ static void ice_verify_cacheline_size(struct ice_pf *pf) * ice_send_version - update firmware with driver version * @pf: PF struct * - * Returns ICE_SUCCESS on success, else error code + * Returns 0 on success, else error code */ -static enum ice_status ice_send_version(struct ice_pf *pf) +static int ice_send_version(struct ice_pf *pf) { struct ice_driver_ver dv; @@ -3961,6 +4599,10 @@ static int ice_register_netdev(struct ice_pf *pf) if (!vsi || !vsi->netdev) return -EIO; + err = ice_devlink_create_pf_port(pf); + if (err) + goto err_devlink_create; + err = register_netdev(vsi->netdev); if (err) goto err_register_netdev; @@ -3968,17 +4610,13 @@ static int ice_register_netdev(struct ice_pf *pf) set_bit(ICE_VSI_NETDEV_REGISTERED, vsi->state); netif_carrier_off(vsi->netdev); netif_tx_stop_all_queues(vsi->netdev); - err = ice_devlink_create_port(vsi); - if (err) - goto err_devlink_create; - devlink_port_type_eth_set(&vsi->devlink_port, vsi->netdev); + devlink_port_type_eth_set(&pf->devlink_port, vsi->netdev); return 0; -err_devlink_create: - unregister_netdev(vsi->netdev); - clear_bit(ICE_VSI_NETDEV_REGISTERED, vsi->state); err_register_netdev: + ice_devlink_destroy_pf_port(pf); +err_devlink_create: free_netdev(vsi->netdev); vsi->netdev = NULL; clear_bit(ICE_VSI_NETDEV_ALLOCD, vsi->state); @@ -4000,6 +4638,11 @@ ice_probe(struct pci_dev *pdev, const struct pci_device_id __always_unused *ent) struct ice_hw *hw; int i, err; + if (pdev->is_virtfn) { + dev_err(dev, "can't probe a virtual function\n"); + return -EINVAL; + } + /* this driver uses devres, see * Documentation/driver-api/driver-model/devres.rst */ @@ -4017,10 +4660,11 @@ ice_probe(struct pci_dev *pdev, const struct pci_device_id __always_unused *ent) if (!pf) return -ENOMEM; + /* initialize Auxiliary index to invalid value */ + pf->aux_idx = -1; + /* set up for high or low DMA */ err = dma_set_mask_and_coherent(dev, DMA_BIT_MASK(64)); - if (err) - err = dma_set_mask_and_coherent(dev, DMA_BIT_MASK(32)); if (err) { dev_err(dev, "DMA configuration failed: 0x%x\n", err); return err; @@ -4051,12 +4695,6 @@ ice_probe(struct pci_dev *pdev, const struct pci_device_id __always_unused *ent) pf->msg_enable = netif_msg_init(debug, ICE_DFLT_NETIF_M); - err = ice_devlink_register(pf); - if (err) { - dev_err(dev, "ice_devlink_register failed: %d\n", err); - goto err_exit_unroll; - } - #ifndef CONFIG_DYNAMIC_DEBUG if (debug < -1) hw->debug_mask = debug; @@ -4069,6 +4707,8 @@ ice_probe(struct pci_dev *pdev, const struct pci_device_id __always_unused *ent) goto err_exit_unroll; } + ice_init_feature_support(pf); + ice_request_fw(pf); /* if ice_request_fw fails, ICE_FLAG_ADV_FEATURES bit won't be @@ -4076,7 +4716,6 @@ ice_probe(struct pci_dev *pdev, const struct pci_device_id __always_unused *ent) * true */ if (ice_is_safe_mode(pf)) { - dev_err(dev, "Package download failed. Advanced features disabled - Device now in Safe Mode\n"); /* we already got function/device capabilities but these don't * reflect what the driver needs to do in safe mode. Instead of * adding conditional logic everywhere to ignore these @@ -4204,6 +4843,9 @@ ice_probe(struct pci_dev *pdev, const struct pci_device_id __always_unused *ent) ice_init_link_dflt_override(pf->hw.port_info); + ice_check_link_cfg_err(pf, + pf->hw.port_info->phy.link_info.link_cfg_err); + /* if media available, initialize PHY settings */ if (pf->hw.port_info->phy.link_info.link_info & ICE_AQ_MEDIA_AVAILABLE) { @@ -4242,6 +4884,11 @@ ice_probe(struct pci_dev *pdev, const struct pci_device_id __always_unused *ent) } /* initialize DDP driven features */ + if (test_bit(ICE_FLAG_PTP_SUPPORTED, pf->flags)) + ice_ptp_init(pf); + + if (ice_is_feature_supported(pf, ICE_F_GNSS)) + ice_gnss_init(pf); /* Note: Flow director init failure is non-fatal to load */ if (ice_init_fdir(pf)) @@ -4266,10 +4913,38 @@ probe_done: if (err) goto err_netdev_reg; + err = ice_devlink_register_params(pf); + if (err) + goto err_netdev_reg; + /* ready to go, so clear down state bit */ clear_bit(ICE_DOWN, pf->state); + if (ice_is_rdma_ena(pf)) { + pf->aux_idx = ida_alloc(&ice_aux_ida, GFP_KERNEL); + if (pf->aux_idx < 0) { + dev_err(dev, "Failed to allocate device ID for AUX driver\n"); + err = -ENOMEM; + goto err_devlink_reg_param; + } + + err = ice_init_rdma(pf); + if (err) { + dev_err(dev, "Failed to initialize RDMA: %d\n", err); + err = -EIO; + goto err_init_aux_unroll; + } + } else { + dev_warn(dev, "RDMA is not supported on this device\n"); + } + + ice_devlink_register(pf); return 0; +err_init_aux_unroll: + pf->adev = NULL; + ida_free(&ice_aux_ida, pf->aux_idx); +err_devlink_reg_param: + ice_devlink_unregister_params(pf); err_netdev_reg: err_send_version_unroll: ice_vsi_release_all(pf); @@ -4288,7 +4963,6 @@ err_init_pf_unroll: ice_devlink_destroy_regions(pf); ice_deinit_hw(hw); err_exit_unroll: - ice_devlink_unregister(pf); pci_disable_pcie_error_reporting(pdev); pci_disable_device(pdev); return err; @@ -4327,9 +5001,9 @@ static void ice_setup_mc_magic_wake(struct ice_pf *pf) { struct device *dev = ice_pf_to_dev(pf); struct ice_hw *hw = &pf->hw; - enum ice_status status; u8 mac_addr[ETH_ALEN]; struct ice_vsi *vsi; + int status; u8 flags; if (!pf->wol_ena) @@ -4351,9 +5025,8 @@ static void ice_setup_mc_magic_wake(struct ice_pf *pf) status = ice_aq_manage_mac_write(hw, mac_addr, flags, NULL); if (status) - dev_err(dev, "Failed to enable Multicast Magic Packet wake, err %s aq_err %s\n", - ice_stat_str(status), - ice_aq_str(hw->adminq.sq_last_status)); + dev_err(dev, "Failed to enable Multicast Magic Packet wake, err %d aq_err %s\n", + status, ice_aq_str(hw->adminq.sq_last_status)); } /** @@ -4365,31 +5038,39 @@ static void ice_remove(struct pci_dev *pdev) struct ice_pf *pf = pci_get_drvdata(pdev); int i; - if (!pf) - return; - + ice_devlink_unregister(pf); for (i = 0; i < ICE_MAX_RESET_WAIT; i++) { if (!ice_is_reset_in_progress(pf->state)) break; msleep(100); } + ice_tc_indir_block_remove(pf); + if (test_bit(ICE_FLAG_SRIOV_ENA, pf->flags)) { set_bit(ICE_VF_RESETS_DISABLED, pf->state); ice_free_vfs(pf); } - set_bit(ICE_DOWN, pf->state); ice_service_task_stop(pf); ice_aq_cancel_waiting_tasks(pf); + ice_unplug_aux_dev(pf); + if (pf->aux_idx >= 0) + ida_free(&ice_aux_ida, pf->aux_idx); + ice_devlink_unregister_params(pf); + set_bit(ICE_DOWN, pf->state); - mutex_destroy(&(&pf->hw)->fdir_fltr_lock); ice_deinit_lag(pf); + if (test_bit(ICE_FLAG_PTP_SUPPORTED, pf->flags)) + ice_ptp_release(pf); + if (ice_is_feature_supported(pf, ICE_F_GNSS)) + ice_gnss_exit(pf); if (!ice_is_safe_mode(pf)) ice_remove_arfs(pf); ice_setup_mc_magic_wake(pf); ice_vsi_release_all(pf); + mutex_destroy(&(&pf->hw)->fdir_fltr_lock); ice_set_wake(pf); ice_free_irq_msix_misc(pf); ice_for_each_vsi(pf, i) { @@ -4400,7 +5081,6 @@ static void ice_remove(struct pci_dev *pdev) ice_deinit_pf(pf); ice_devlink_destroy_regions(pf); ice_deinit_hw(&pf->hw); - ice_devlink_unregister(pf); /* Issue a PFR as part of the prescribed driver unload flow. Do not * do it via ice_schedule_reset() since there is no need to rebuild @@ -4538,6 +5218,8 @@ static int __maybe_unused ice_suspend(struct device *dev) */ disabled = ice_service_task_stop(pf); + ice_unplug_aux_dev(pf); + /* Already suspended?, then there is nothing to do */ if (test_and_set_bit(ICE_SUSPENDED, pf->state)) { if (!disabled) @@ -4570,7 +5252,6 @@ static int __maybe_unused ice_suspend(struct device *dev) continue; ice_vsi_free_q_vectors(pf->vsi[v]); } - ice_free_cpu_rx_rmap(ice_get_main_vsi(pf)); ice_clear_interrupt_scheme(pf); pci_save_state(pdev); @@ -4660,7 +5341,7 @@ ice_pci_err_detected(struct pci_dev *pdev, pci_channel_state_t err) if (!test_bit(ICE_PREPARED_FOR_RESET, pf->state)) { set_bit(ICE_PFR_REQ, pf->state); - ice_prepare_for_reset(pf); + ice_prepare_for_reset(pf, ICE_RESET_PFR); } } @@ -4700,12 +5381,6 @@ static pci_ers_result_t ice_pci_err_slot_reset(struct pci_dev *pdev) result = PCI_ERS_RESULT_DISCONNECT; } - err = pci_aer_clear_nonfatal_status(pdev); - if (err) - dev_dbg(&pdev->dev, "pci_aer_clear_nonfatal_status() failed, error %d\n", - err); - /* non-fatal, continue */ - return result; } @@ -4752,7 +5427,7 @@ static void ice_pci_err_reset_prepare(struct pci_dev *pdev) if (!test_bit(ICE_PREPARED_FOR_RESET, pf->state)) { set_bit(ICE_PFR_REQ, pf->state); - ice_prepare_for_reset(pf); + ice_prepare_for_reset(pf, ICE_RESET_PFR); } } } @@ -4778,6 +5453,8 @@ static const struct pci_device_id ice_pci_tbl[] = { { PCI_VDEVICE(INTEL, ICE_DEV_ID_E810C_BACKPLANE), 0 }, { PCI_VDEVICE(INTEL, ICE_DEV_ID_E810C_QSFP), 0 }, { PCI_VDEVICE(INTEL, ICE_DEV_ID_E810C_SFP), 0 }, + { PCI_VDEVICE(INTEL, ICE_DEV_ID_E810_XXV_BACKPLANE), 0 }, + { PCI_VDEVICE(INTEL, ICE_DEV_ID_E810_XXV_QSFP), 0 }, { PCI_VDEVICE(INTEL, ICE_DEV_ID_E810_XXV_SFP), 0 }, { PCI_VDEVICE(INTEL, ICE_DEV_ID_E823C_BACKPLANE), 0 }, { PCI_VDEVICE(INTEL, ICE_DEV_ID_E823C_QSFP), 0 }, @@ -4798,6 +5475,7 @@ static const struct pci_device_id ice_pci_tbl[] = { { PCI_VDEVICE(INTEL, ICE_DEV_ID_E823L_10G_BASE_T), 0 }, { PCI_VDEVICE(INTEL, ICE_DEV_ID_E823L_1GBE), 0 }, { PCI_VDEVICE(INTEL, ICE_DEV_ID_E823L_QSFP), 0 }, + { PCI_VDEVICE(INTEL, ICE_DEV_ID_E822_SI_DFLT), 0 }, /* required last entry */ { 0, } }; @@ -4883,10 +5561,10 @@ static int ice_set_mac_address(struct net_device *netdev, void *pi) struct ice_pf *pf = vsi->back; struct ice_hw *hw = &pf->hw; struct sockaddr *addr = pi; - enum ice_status status; + u8 old_mac[ETH_ALEN]; u8 flags = 0; - int err = 0; u8 *mac; + int err; mac = (u8 *)addr->sa_data; @@ -4894,7 +5572,7 @@ static int ice_set_mac_address(struct net_device *netdev, void *pi) return -EADDRNOTAVAIL; if (ether_addr_equal(netdev->dev_addr, mac)) { - netdev_warn(netdev, "already using mac %pM\n", mac); + netdev_dbg(netdev, "already using mac %pM\n", mac); return 0; } @@ -4905,48 +5583,60 @@ static int ice_set_mac_address(struct net_device *netdev, void *pi) return -EBUSY; } + if (ice_chnl_dmac_fltr_cnt(pf)) { + netdev_err(netdev, "can't set mac %pM. Device has tc-flower filters, delete all of them and try again\n", + mac); + return -EAGAIN; + } + + netif_addr_lock_bh(netdev); + ether_addr_copy(old_mac, netdev->dev_addr); + /* change the netdev's MAC address */ + eth_hw_addr_set(netdev, mac); + netif_addr_unlock_bh(netdev); + /* Clean up old MAC filter. Not an error if old filter doesn't exist */ - status = ice_fltr_remove_mac(vsi, netdev->dev_addr, ICE_FWD_TO_VSI); - if (status && status != ICE_ERR_DOES_NOT_EXIST) { + err = ice_fltr_remove_mac(vsi, old_mac, ICE_FWD_TO_VSI); + if (err && err != -ENOENT) { err = -EADDRNOTAVAIL; goto err_update_filters; } /* Add filter for new MAC. If filter exists, return success */ - status = ice_fltr_add_mac(vsi, mac, ICE_FWD_TO_VSI); - if (status == ICE_ERR_ALREADY_EXISTS) { + err = ice_fltr_add_mac(vsi, mac, ICE_FWD_TO_VSI); + if (err == -EEXIST) { /* Although this MAC filter is already present in hardware it's * possible in some cases (e.g. bonding) that dev_addr was * modified outside of the driver and needs to be restored back * to this value. */ - memcpy(netdev->dev_addr, mac, netdev->addr_len); netdev_dbg(netdev, "filter for MAC %pM already exists\n", mac); - return 0; - } - /* error if the new filter addition failed */ - if (status) + return 0; + } else if (err) { + /* error if the new filter addition failed */ err = -EADDRNOTAVAIL; + } err_update_filters: if (err) { netdev_err(netdev, "can't set MAC %pM. filter update failed\n", mac); + netif_addr_lock_bh(netdev); + eth_hw_addr_set(netdev, old_mac); + netif_addr_unlock_bh(netdev); return err; } - /* change the netdev's MAC address */ - memcpy(netdev->dev_addr, mac, netdev->addr_len); netdev_dbg(vsi->netdev, "updated MAC address to %pM\n", netdev->dev_addr); /* write new MAC address to the firmware */ flags = ICE_AQC_MAN_MAC_UPDATE_LAA_WOL; - status = ice_aq_manage_mac_write(hw, mac, flags, NULL); - if (status) { - netdev_err(netdev, "can't set MAC %pM. write to firmware failed error %s\n", - mac, ice_stat_str(status)); + err = ice_aq_manage_mac_write(hw, mac, flags, NULL); + if (err) { + netdev_err(netdev, "can't set MAC %pM. write to firmware failed error %d\n", + mac, err); } return 0; } @@ -4988,8 +5678,8 @@ ice_set_tx_maxrate(struct net_device *netdev, int queue_index, u32 maxrate) { struct ice_netdev_priv *np = netdev_priv(netdev); struct ice_vsi *vsi = np->vsi; - enum ice_status status; u16 q_handle; + int status; u8 tc; /* Validate maxrate requested is within permitted range */ @@ -5009,13 +5699,11 @@ ice_set_tx_maxrate(struct net_device *netdev, int queue_index, u32 maxrate) else status = ice_cfg_q_bw_lmt(vsi->port_info, vsi->idx, tc, q_handle, ICE_MAX_BW, maxrate * 1000); - if (status) { - netdev_err(netdev, "Unable to set Tx max rate, error %s\n", - ice_stat_str(status)); - return -EIO; - } + if (status) + netdev_err(netdev, "Unable to set Tx max rate, error %d\n", + status); - return 0; + return status; } /** @@ -5065,11 +5753,12 @@ ice_fdb_add(struct ndmsg *ndm, struct nlattr __always_unused *tb[], * @dev: the net device pointer * @addr: the MAC address entry being added * @vid: VLAN ID + * @extack: netlink extended ack */ static int ice_fdb_del(struct ndmsg *ndm, __always_unused struct nlattr *tb[], struct net_device *dev, const unsigned char *addr, - __always_unused u16 vid) + __always_unused u16 vid, struct netlink_ext_ack *extack) { int err; @@ -5088,6 +5777,251 @@ ice_fdb_del(struct ndmsg *ndm, __always_unused struct nlattr *tb[], return err; } +#define NETIF_VLAN_OFFLOAD_FEATURES (NETIF_F_HW_VLAN_CTAG_RX | \ + NETIF_F_HW_VLAN_CTAG_TX | \ + NETIF_F_HW_VLAN_STAG_RX | \ + NETIF_F_HW_VLAN_STAG_TX) + +#define NETIF_VLAN_STRIPPING_FEATURES (NETIF_F_HW_VLAN_CTAG_RX | \ + NETIF_F_HW_VLAN_STAG_RX) + +#define NETIF_VLAN_FILTERING_FEATURES (NETIF_F_HW_VLAN_CTAG_FILTER | \ + NETIF_F_HW_VLAN_STAG_FILTER) + +/** + * ice_fix_features - fix the netdev features flags based on device limitations + * @netdev: ptr to the netdev that flags are being fixed on + * @features: features that need to be checked and possibly fixed + * + * Make sure any fixups are made to features in this callback. This enables the + * driver to not have to check unsupported configurations throughout the driver + * because that's the responsiblity of this callback. + * + * Single VLAN Mode (SVM) Supported Features: + * NETIF_F_HW_VLAN_CTAG_FILTER + * NETIF_F_HW_VLAN_CTAG_RX + * NETIF_F_HW_VLAN_CTAG_TX + * + * Double VLAN Mode (DVM) Supported Features: + * NETIF_F_HW_VLAN_CTAG_FILTER + * NETIF_F_HW_VLAN_CTAG_RX + * NETIF_F_HW_VLAN_CTAG_TX + * + * NETIF_F_HW_VLAN_STAG_FILTER + * NETIF_HW_VLAN_STAG_RX + * NETIF_HW_VLAN_STAG_TX + * + * Features that need fixing: + * Cannot simultaneously enable CTAG and STAG stripping and/or insertion. + * These are mutually exlusive as the VSI context cannot support multiple + * VLAN ethertypes simultaneously for stripping and/or insertion. If this + * is not done, then default to clearing the requested STAG offload + * settings. + * + * All supported filtering has to be enabled or disabled together. For + * example, in DVM, CTAG and STAG filtering have to be enabled and disabled + * together. If this is not done, then default to VLAN filtering disabled. + * These are mutually exclusive as there is currently no way to + * enable/disable VLAN filtering based on VLAN ethertype when using VLAN + * prune rules. + */ +static netdev_features_t +ice_fix_features(struct net_device *netdev, netdev_features_t features) +{ + struct ice_netdev_priv *np = netdev_priv(netdev); + netdev_features_t req_vlan_fltr, cur_vlan_fltr; + bool cur_ctag, cur_stag, req_ctag, req_stag; + + cur_vlan_fltr = netdev->features & NETIF_VLAN_FILTERING_FEATURES; + cur_ctag = cur_vlan_fltr & NETIF_F_HW_VLAN_CTAG_FILTER; + cur_stag = cur_vlan_fltr & NETIF_F_HW_VLAN_STAG_FILTER; + + req_vlan_fltr = features & NETIF_VLAN_FILTERING_FEATURES; + req_ctag = req_vlan_fltr & NETIF_F_HW_VLAN_CTAG_FILTER; + req_stag = req_vlan_fltr & NETIF_F_HW_VLAN_STAG_FILTER; + + if (req_vlan_fltr != cur_vlan_fltr) { + if (ice_is_dvm_ena(&np->vsi->back->hw)) { + if (req_ctag && req_stag) { + features |= NETIF_VLAN_FILTERING_FEATURES; + } else if (!req_ctag && !req_stag) { + features &= ~NETIF_VLAN_FILTERING_FEATURES; + } else if ((!cur_ctag && req_ctag && !cur_stag) || + (!cur_stag && req_stag && !cur_ctag)) { + features |= NETIF_VLAN_FILTERING_FEATURES; + netdev_warn(netdev, "802.1Q and 802.1ad VLAN filtering must be either both on or both off. VLAN filtering has been enabled for both types.\n"); + } else if ((cur_ctag && !req_ctag && cur_stag) || + (cur_stag && !req_stag && cur_ctag)) { + features &= ~NETIF_VLAN_FILTERING_FEATURES; + netdev_warn(netdev, "802.1Q and 802.1ad VLAN filtering must be either both on or both off. VLAN filtering has been disabled for both types.\n"); + } + } else { + if (req_vlan_fltr & NETIF_F_HW_VLAN_STAG_FILTER) + netdev_warn(netdev, "cannot support requested 802.1ad filtering setting in SVM mode\n"); + + if (req_vlan_fltr & NETIF_F_HW_VLAN_CTAG_FILTER) + features |= NETIF_F_HW_VLAN_CTAG_FILTER; + } + } + + if ((features & (NETIF_F_HW_VLAN_CTAG_RX | NETIF_F_HW_VLAN_CTAG_TX)) && + (features & (NETIF_F_HW_VLAN_STAG_RX | NETIF_F_HW_VLAN_STAG_TX))) { + netdev_warn(netdev, "cannot support CTAG and STAG VLAN stripping and/or insertion simultaneously since CTAG and STAG offloads are mutually exclusive, clearing STAG offload settings\n"); + features &= ~(NETIF_F_HW_VLAN_STAG_RX | + NETIF_F_HW_VLAN_STAG_TX); + } + + if (!(netdev->features & NETIF_F_RXFCS) && + (features & NETIF_F_RXFCS) && + (features & NETIF_VLAN_STRIPPING_FEATURES) && + !ice_vsi_has_non_zero_vlans(np->vsi)) { + netdev_warn(netdev, "Disabling VLAN stripping as FCS/CRC stripping is also disabled and there is no VLAN configured\n"); + features &= ~NETIF_VLAN_STRIPPING_FEATURES; + } + + return features; +} + +/** + * ice_set_vlan_offload_features - set VLAN offload features for the PF VSI + * @vsi: PF's VSI + * @features: features used to determine VLAN offload settings + * + * First, determine the vlan_ethertype based on the VLAN offload bits in + * features. Then determine if stripping and insertion should be enabled or + * disabled. Finally enable or disable VLAN stripping and insertion. + */ +static int +ice_set_vlan_offload_features(struct ice_vsi *vsi, netdev_features_t features) +{ + bool enable_stripping = true, enable_insertion = true; + struct ice_vsi_vlan_ops *vlan_ops; + int strip_err = 0, insert_err = 0; + u16 vlan_ethertype = 0; + + vlan_ops = ice_get_compat_vsi_vlan_ops(vsi); + + if (features & (NETIF_F_HW_VLAN_STAG_RX | NETIF_F_HW_VLAN_STAG_TX)) + vlan_ethertype = ETH_P_8021AD; + else if (features & (NETIF_F_HW_VLAN_CTAG_RX | NETIF_F_HW_VLAN_CTAG_TX)) + vlan_ethertype = ETH_P_8021Q; + + if (!(features & (NETIF_F_HW_VLAN_STAG_RX | NETIF_F_HW_VLAN_CTAG_RX))) + enable_stripping = false; + if (!(features & (NETIF_F_HW_VLAN_STAG_TX | NETIF_F_HW_VLAN_CTAG_TX))) + enable_insertion = false; + + if (enable_stripping) + strip_err = vlan_ops->ena_stripping(vsi, vlan_ethertype); + else + strip_err = vlan_ops->dis_stripping(vsi); + + if (enable_insertion) + insert_err = vlan_ops->ena_insertion(vsi, vlan_ethertype); + else + insert_err = vlan_ops->dis_insertion(vsi); + + if (strip_err || insert_err) + return -EIO; + + return 0; +} + +/** + * ice_set_vlan_filtering_features - set VLAN filtering features for the PF VSI + * @vsi: PF's VSI + * @features: features used to determine VLAN filtering settings + * + * Enable or disable Rx VLAN filtering based on the VLAN filtering bits in the + * features. + */ +static int +ice_set_vlan_filtering_features(struct ice_vsi *vsi, netdev_features_t features) +{ + struct ice_vsi_vlan_ops *vlan_ops = ice_get_compat_vsi_vlan_ops(vsi); + int err = 0; + + /* support Single VLAN Mode (SVM) and Double VLAN Mode (DVM) by checking + * if either bit is set + */ + if (features & + (NETIF_F_HW_VLAN_CTAG_FILTER | NETIF_F_HW_VLAN_STAG_FILTER)) + err = vlan_ops->ena_rx_filtering(vsi); + else + err = vlan_ops->dis_rx_filtering(vsi); + + return err; +} + +/** + * ice_set_vlan_features - set VLAN settings based on suggested feature set + * @netdev: ptr to the netdev being adjusted + * @features: the feature set that the stack is suggesting + * + * Only update VLAN settings if the requested_vlan_features are different than + * the current_vlan_features. + */ +static int +ice_set_vlan_features(struct net_device *netdev, netdev_features_t features) +{ + netdev_features_t current_vlan_features, requested_vlan_features; + struct ice_netdev_priv *np = netdev_priv(netdev); + struct ice_vsi *vsi = np->vsi; + int err; + + current_vlan_features = netdev->features & NETIF_VLAN_OFFLOAD_FEATURES; + requested_vlan_features = features & NETIF_VLAN_OFFLOAD_FEATURES; + if (current_vlan_features ^ requested_vlan_features) { + if ((features & NETIF_F_RXFCS) && + (features & NETIF_VLAN_STRIPPING_FEATURES)) { + dev_err(ice_pf_to_dev(vsi->back), + "To enable VLAN stripping, you must first enable FCS/CRC stripping\n"); + return -EIO; + } + + err = ice_set_vlan_offload_features(vsi, features); + if (err) + return err; + } + + current_vlan_features = netdev->features & + NETIF_VLAN_FILTERING_FEATURES; + requested_vlan_features = features & NETIF_VLAN_FILTERING_FEATURES; + if (current_vlan_features ^ requested_vlan_features) { + err = ice_set_vlan_filtering_features(vsi, features); + if (err) + return err; + } + + return 0; +} + +/** + * ice_set_loopback - turn on/off loopback mode on underlying PF + * @vsi: ptr to VSI + * @ena: flag to indicate the on/off setting + */ +static int ice_set_loopback(struct ice_vsi *vsi, bool ena) +{ + bool if_running = netif_running(vsi->netdev); + int ret; + + if (if_running && !test_and_set_bit(ICE_VSI_DOWN, vsi->state)) { + ret = ice_down(vsi); + if (ret) { + netdev_err(vsi->netdev, "Preparing device to toggle loopback failed\n"); + return ret; + } + } + ret = ice_aq_set_mac_loopback(&vsi->back->hw, ena, NULL); + if (ret) + netdev_err(vsi->netdev, "Failed to toggle loopback state\n"); + if (if_running) + ret = ice_up(vsi); + + return ret; +} + /** * ice_set_features - set the netdev feature flags * @netdev: ptr to the netdev being adjusted @@ -5096,80 +6030,96 @@ ice_fdb_del(struct ndmsg *ndm, __always_unused struct nlattr *tb[], static int ice_set_features(struct net_device *netdev, netdev_features_t features) { + netdev_features_t changed = netdev->features ^ features; struct ice_netdev_priv *np = netdev_priv(netdev); struct ice_vsi *vsi = np->vsi; struct ice_pf *pf = vsi->back; int ret = 0; /* Don't set any netdev advanced features with device in Safe Mode */ - if (ice_is_safe_mode(vsi->back)) { - dev_err(ice_pf_to_dev(vsi->back), "Device is in Safe Mode - not enabling advanced netdev features\n"); + if (ice_is_safe_mode(pf)) { + dev_err(ice_pf_to_dev(pf), + "Device is in Safe Mode - not enabling advanced netdev features\n"); return ret; } /* Do not change setting during reset */ if (ice_is_reset_in_progress(pf->state)) { - dev_err(ice_pf_to_dev(vsi->back), "Device is resetting, changing advanced netdev features temporarily unavailable.\n"); + dev_err(ice_pf_to_dev(pf), + "Device is resetting, changing advanced netdev features temporarily unavailable.\n"); return -EBUSY; } /* Multiple features can be changed in one call so keep features in * separate if/else statements to guarantee each feature is checked */ - if (features & NETIF_F_RXHASH && !(netdev->features & NETIF_F_RXHASH)) - ice_vsi_manage_rss_lut(vsi, true); - else if (!(features & NETIF_F_RXHASH) && - netdev->features & NETIF_F_RXHASH) - ice_vsi_manage_rss_lut(vsi, false); - - if ((features & NETIF_F_HW_VLAN_CTAG_RX) && - !(netdev->features & NETIF_F_HW_VLAN_CTAG_RX)) - ret = ice_vsi_manage_vlan_stripping(vsi, true); - else if (!(features & NETIF_F_HW_VLAN_CTAG_RX) && - (netdev->features & NETIF_F_HW_VLAN_CTAG_RX)) - ret = ice_vsi_manage_vlan_stripping(vsi, false); - - if ((features & NETIF_F_HW_VLAN_CTAG_TX) && - !(netdev->features & NETIF_F_HW_VLAN_CTAG_TX)) - ret = ice_vsi_manage_vlan_insertion(vsi); - else if (!(features & NETIF_F_HW_VLAN_CTAG_TX) && - (netdev->features & NETIF_F_HW_VLAN_CTAG_TX)) - ret = ice_vsi_manage_vlan_insertion(vsi); - - if ((features & NETIF_F_HW_VLAN_CTAG_FILTER) && - !(netdev->features & NETIF_F_HW_VLAN_CTAG_FILTER)) - ret = ice_cfg_vlan_pruning(vsi, true, false); - else if (!(features & NETIF_F_HW_VLAN_CTAG_FILTER) && - (netdev->features & NETIF_F_HW_VLAN_CTAG_FILTER)) - ret = ice_cfg_vlan_pruning(vsi, false, false); - - if ((features & NETIF_F_NTUPLE) && - !(netdev->features & NETIF_F_NTUPLE)) { - ice_vsi_manage_fdir(vsi, true); - ice_init_arfs(vsi); - } else if (!(features & NETIF_F_NTUPLE) && - (netdev->features & NETIF_F_NTUPLE)) { - ice_vsi_manage_fdir(vsi, false); - ice_clear_arfs(vsi); + if (changed & NETIF_F_RXHASH) + ice_vsi_manage_rss_lut(vsi, !!(features & NETIF_F_RXHASH)); + + ret = ice_set_vlan_features(netdev, features); + if (ret) + return ret; + + /* Turn on receive of FCS aka CRC, and after setting this + * flag the packet data will have the 4 byte CRC appended + */ + if (changed & NETIF_F_RXFCS) { + if ((features & NETIF_F_RXFCS) && + (features & NETIF_VLAN_STRIPPING_FEATURES)) { + dev_err(ice_pf_to_dev(vsi->back), + "To disable FCS/CRC stripping, you must first disable VLAN stripping\n"); + return -EIO; + } + + ice_vsi_cfg_crc_strip(vsi, !!(features & NETIF_F_RXFCS)); + ret = ice_down_up(vsi); + if (ret) + return ret; + } + + if (changed & NETIF_F_NTUPLE) { + bool ena = !!(features & NETIF_F_NTUPLE); + + ice_vsi_manage_fdir(vsi, ena); + ena ? ice_init_arfs(vsi) : ice_clear_arfs(vsi); + } + + /* don't turn off hw_tc_offload when ADQ is already enabled */ + if (!(features & NETIF_F_HW_TC) && ice_is_adq_active(pf)) { + dev_err(ice_pf_to_dev(pf), "ADQ is active, can't turn hw_tc_offload off\n"); + return -EACCES; } + if (changed & NETIF_F_HW_TC) { + bool ena = !!(features & NETIF_F_HW_TC); + + ena ? set_bit(ICE_FLAG_CLS_FLOWER, pf->flags) : + clear_bit(ICE_FLAG_CLS_FLOWER, pf->flags); + } + + if (changed & NETIF_F_LOOPBACK) + ret = ice_set_loopback(vsi, !!(features & NETIF_F_LOOPBACK)); + return ret; } /** - * ice_vsi_vlan_setup - Setup VLAN offload properties on a VSI + * ice_vsi_vlan_setup - Setup VLAN offload properties on a PF VSI * @vsi: VSI to setup VLAN properties for */ static int ice_vsi_vlan_setup(struct ice_vsi *vsi) { - int ret = 0; + int err; - if (vsi->netdev->features & NETIF_F_HW_VLAN_CTAG_RX) - ret = ice_vsi_manage_vlan_stripping(vsi, true); - if (vsi->netdev->features & NETIF_F_HW_VLAN_CTAG_TX) - ret = ice_vsi_manage_vlan_insertion(vsi); + err = ice_set_vlan_offload_features(vsi, vsi->netdev->features); + if (err) + return err; - return ret; + err = ice_set_vlan_filtering_features(vsi, vsi->netdev->features); + if (err) + return err; + + return ice_vsi_add_vlan_zero(vsi); } /** @@ -5185,10 +6135,12 @@ int ice_vsi_cfg(struct ice_vsi *vsi) if (vsi->netdev) { ice_set_rx_mode(vsi->netdev); - err = ice_vsi_vlan_setup(vsi); + if (vsi->type != ICE_VSI_LB) { + err = ice_vsi_vlan_setup(vsi); - if (err) - return err; + if (err) + return err; + } } ice_vsi_cfg_dcb_rings(vsi); @@ -5202,76 +6154,59 @@ int ice_vsi_cfg(struct ice_vsi *vsi) } /* THEORY OF MODERATION: - * The below code creates custom DIM profiles for use by this driver, because - * the ice driver hardware works differently than the hardware that DIMLIB was + * The ice driver hardware works differently than the hardware that DIMLIB was * originally made for. ice hardware doesn't have packet count limits that * can trigger an interrupt, but it *does* have interrupt rate limit support, - * and this code adds that capability to be used by the driver when it's using - * DIMLIB. The DIMLIB code was always designed to be a suggestion to the driver - * for how to "respond" to traffic and interrupts, so this driver uses a - * slightly different set of moderation parameters to get best performance. + * which is hard-coded to a limit of 250,000 ints/second. + * If not using dynamic moderation, the INTRL value can be modified + * by ethtool rx-usecs-high. */ struct ice_dim { /* the throttle rate for interrupts, basically worst case delay before * an initial interrupt fires, value is stored in microseconds. */ u16 itr; - /* the rate limit for interrupts, which can cap a delay from a small - * ITR at a certain amount of interrupts per second. f.e. a 2us ITR - * could yield as much as 500,000 interrupts per second, but with a - * 10us rate limit, it limits to 100,000 interrupts per second. Value - * is stored in microseconds. - */ - u16 intrl; }; /* Make a different profile for Rx that doesn't allow quite so aggressive - * moderation at the high end (it maxes out at 128us or about 8k interrupts a - * second. The INTRL/rate parameters here are only useful to cap small ITR - * values, which is why for larger ITR's - like 128, which can only generate - * 8k interrupts per second, there is no point to rate limit and the values - * are set to zero. The rate limit values do affect latency, and so must - * be reasonably small so to not impact latency sensitive tests. + * moderation at the high end (it maxes out at 126us or about 8k interrupts a + * second. */ static const struct ice_dim rx_profile[] = { - {2, 10}, - {8, 16}, - {32, 0}, - {96, 0}, - {128, 0} + {2}, /* 500,000 ints/s, capped at 250K by INTRL */ + {8}, /* 125,000 ints/s */ + {16}, /* 62,500 ints/s */ + {62}, /* 16,129 ints/s */ + {126} /* 7,936 ints/s */ }; /* The transmit profile, which has the same sorts of values * as the previous struct */ static const struct ice_dim tx_profile[] = { - {2, 10}, - {8, 16}, - {64, 0}, - {128, 0}, - {256, 0} + {2}, /* 500,000 ints/s, capped at 250K by INTRL */ + {8}, /* 125,000 ints/s */ + {40}, /* 16,125 ints/s */ + {128}, /* 7,812 ints/s */ + {256} /* 3,906 ints/s */ }; static void ice_tx_dim_work(struct work_struct *work) { struct ice_ring_container *rc; - struct ice_q_vector *q_vector; struct dim *dim; - u16 itr, intrl; + u16 itr; dim = container_of(work, struct dim, work); - rc = container_of(dim, struct ice_ring_container, dim); - q_vector = container_of(rc, struct ice_q_vector, tx); + rc = (struct ice_ring_container *)dim->priv; - if (dim->profile_ix >= ARRAY_SIZE(tx_profile)) - dim->profile_ix = ARRAY_SIZE(tx_profile) - 1; + WARN_ON(dim->profile_ix >= ARRAY_SIZE(tx_profile)); /* look up the values in our local table */ itr = tx_profile[dim->profile_ix].itr; - intrl = tx_profile[dim->profile_ix].intrl; + ice_trace(tx_dim_work, container_of(rc, struct ice_q_vector, tx), dim); ice_write_itr(rc, itr); - ice_write_intrl(q_vector, intrl); dim->state = DIM_START_MEASURE; } @@ -5279,27 +6214,65 @@ static void ice_tx_dim_work(struct work_struct *work) static void ice_rx_dim_work(struct work_struct *work) { struct ice_ring_container *rc; - struct ice_q_vector *q_vector; struct dim *dim; - u16 itr, intrl; + u16 itr; dim = container_of(work, struct dim, work); - rc = container_of(dim, struct ice_ring_container, dim); - q_vector = container_of(rc, struct ice_q_vector, rx); + rc = (struct ice_ring_container *)dim->priv; - if (dim->profile_ix >= ARRAY_SIZE(rx_profile)) - dim->profile_ix = ARRAY_SIZE(rx_profile) - 1; + WARN_ON(dim->profile_ix >= ARRAY_SIZE(rx_profile)); /* look up the values in our local table */ itr = rx_profile[dim->profile_ix].itr; - intrl = rx_profile[dim->profile_ix].intrl; + ice_trace(rx_dim_work, container_of(rc, struct ice_q_vector, rx), dim); ice_write_itr(rc, itr); - ice_write_intrl(q_vector, intrl); dim->state = DIM_START_MEASURE; } +#define ICE_DIM_DEFAULT_PROFILE_IX 1 + +/** + * ice_init_moderation - set up interrupt moderation + * @q_vector: the vector containing rings to be configured + * + * Set up interrupt moderation registers, with the intent to do the right thing + * when called from reset or from probe, and whether or not dynamic moderation + * is enabled or not. Take special care to write all the registers in both + * dynamic moderation mode or not in order to make sure hardware is in a known + * state. + */ +static void ice_init_moderation(struct ice_q_vector *q_vector) +{ + struct ice_ring_container *rc; + bool tx_dynamic, rx_dynamic; + + rc = &q_vector->tx; + INIT_WORK(&rc->dim.work, ice_tx_dim_work); + rc->dim.mode = DIM_CQ_PERIOD_MODE_START_FROM_EQE; + rc->dim.profile_ix = ICE_DIM_DEFAULT_PROFILE_IX; + rc->dim.priv = rc; + tx_dynamic = ITR_IS_DYNAMIC(rc); + + /* set the initial TX ITR to match the above */ + ice_write_itr(rc, tx_dynamic ? + tx_profile[rc->dim.profile_ix].itr : rc->itr_setting); + + rc = &q_vector->rx; + INIT_WORK(&rc->dim.work, ice_rx_dim_work); + rc->dim.mode = DIM_CQ_PERIOD_MODE_START_FROM_EQE; + rc->dim.profile_ix = ICE_DIM_DEFAULT_PROFILE_IX; + rc->dim.priv = rc; + rx_dynamic = ITR_IS_DYNAMIC(rc); + + /* set the initial RX ITR to match the above */ + ice_write_itr(rc, rx_dynamic ? rx_profile[rc->dim.profile_ix].itr : + rc->itr_setting); + + ice_set_q_vector_intrl(q_vector); +} + /** * ice_napi_enable_all - Enable NAPI for all q_vectors in the VSI * @vsi: the VSI being configured @@ -5314,13 +6287,9 @@ static void ice_napi_enable_all(struct ice_vsi *vsi) ice_for_each_q_vector(vsi, q_idx) { struct ice_q_vector *q_vector = vsi->q_vectors[q_idx]; - INIT_WORK(&q_vector->tx.dim.work, ice_tx_dim_work); - q_vector->tx.dim.mode = DIM_CQ_PERIOD_MODE_START_FROM_EQE; - - INIT_WORK(&q_vector->rx.dim.work, ice_rx_dim_work); - q_vector->rx.dim.mode = DIM_CQ_PERIOD_MODE_START_FROM_EQE; + ice_init_moderation(q_vector); - if (q_vector->rx.ring || q_vector->tx.ring) + if (q_vector->rx.rx_ring || q_vector->tx.tx_ring) napi_enable(&q_vector->napi); } } @@ -5356,8 +6325,14 @@ static int ice_up_complete(struct ice_vsi *vsi) ice_print_link_msg(vsi, true); netif_tx_start_all_queues(vsi->netdev); netif_carrier_on(vsi->netdev); + if (!ice_is_e810(&pf->hw)) + ice_ptp_link_change(pf, pf->hw.pf_id, true); } + /* Perform an initial read of the statistics registers now to + * set the baseline so counters are ready when interface is up + */ + ice_update_eth_stats(vsi); ice_service_task_schedule(pf); return 0; @@ -5380,48 +6355,49 @@ int ice_up(struct ice_vsi *vsi) /** * ice_fetch_u64_stats_per_ring - get packets and bytes stats per ring - * @ring: Tx or Rx ring to read stats from + * @syncp: pointer to u64_stats_sync + * @stats: stats that pkts and bytes count will be taken from * @pkts: packets stats counter * @bytes: bytes stats counter * * This function fetches stats from the ring considering the atomic operations * that needs to be performed to read u64 values in 32 bit machine. */ -static void -ice_fetch_u64_stats_per_ring(struct ice_ring *ring, u64 *pkts, u64 *bytes) +void +ice_fetch_u64_stats_per_ring(struct u64_stats_sync *syncp, + struct ice_q_stats stats, u64 *pkts, u64 *bytes) { unsigned int start; - *pkts = 0; - *bytes = 0; - if (!ring) - return; do { - start = u64_stats_fetch_begin_irq(&ring->syncp); - *pkts = ring->stats.pkts; - *bytes = ring->stats.bytes; - } while (u64_stats_fetch_retry_irq(&ring->syncp, start)); + start = u64_stats_fetch_begin_irq(syncp); + *pkts = stats.pkts; + *bytes = stats.bytes; + } while (u64_stats_fetch_retry_irq(syncp, start)); } /** * ice_update_vsi_tx_ring_stats - Update VSI Tx ring stats counters * @vsi: the VSI to be updated + * @vsi_stats: the stats struct to be updated * @rings: rings to work on * @count: number of rings */ static void -ice_update_vsi_tx_ring_stats(struct ice_vsi *vsi, struct ice_ring **rings, - u16 count) +ice_update_vsi_tx_ring_stats(struct ice_vsi *vsi, + struct rtnl_link_stats64 *vsi_stats, + struct ice_tx_ring **rings, u16 count) { - struct rtnl_link_stats64 *vsi_stats = &vsi->net_stats; u16 i; for (i = 0; i < count; i++) { - struct ice_ring *ring; - u64 pkts, bytes; + struct ice_tx_ring *ring; + u64 pkts = 0, bytes = 0; ring = READ_ONCE(rings[i]); - ice_fetch_u64_stats_per_ring(ring, &pkts, &bytes); + if (!ring) + continue; + ice_fetch_u64_stats_per_ring(&ring->syncp, ring->stats, &pkts, &bytes); vsi_stats->tx_packets += pkts; vsi_stats->tx_bytes += bytes; vsi->tx_restart += ring->tx_stats.restart_q; @@ -5436,16 +6412,13 @@ ice_update_vsi_tx_ring_stats(struct ice_vsi *vsi, struct ice_ring **rings, */ static void ice_update_vsi_ring_stats(struct ice_vsi *vsi) { - struct rtnl_link_stats64 *vsi_stats = &vsi->net_stats; - struct ice_ring *ring; + struct rtnl_link_stats64 *vsi_stats; u64 pkts, bytes; int i; - /* reset netdev stats */ - vsi_stats->tx_packets = 0; - vsi_stats->tx_bytes = 0; - vsi_stats->rx_packets = 0; - vsi_stats->rx_bytes = 0; + vsi_stats = kzalloc(sizeof(*vsi_stats), GFP_ATOMIC); + if (!vsi_stats) + return; /* reset non-netdev (extended) stats */ vsi->tx_restart = 0; @@ -5457,12 +6430,14 @@ static void ice_update_vsi_ring_stats(struct ice_vsi *vsi) rcu_read_lock(); /* update Tx rings counters */ - ice_update_vsi_tx_ring_stats(vsi, vsi->tx_rings, vsi->num_txq); + ice_update_vsi_tx_ring_stats(vsi, vsi_stats, vsi->tx_rings, + vsi->num_txq); /* update Rx rings counters */ ice_for_each_rxq(vsi, i) { - ring = READ_ONCE(vsi->rx_rings[i]); - ice_fetch_u64_stats_per_ring(ring, &pkts, &bytes); + struct ice_rx_ring *ring = READ_ONCE(vsi->rx_rings[i]); + + ice_fetch_u64_stats_per_ring(&ring->syncp, ring->stats, &pkts, &bytes); vsi_stats->rx_packets += pkts; vsi_stats->rx_bytes += bytes; vsi->rx_buf_failed += ring->rx_stats.alloc_buf_failed; @@ -5471,10 +6446,17 @@ static void ice_update_vsi_ring_stats(struct ice_vsi *vsi) /* update XDP Tx rings counters */ if (ice_is_xdp_ena_vsi(vsi)) - ice_update_vsi_tx_ring_stats(vsi, vsi->xdp_rings, + ice_update_vsi_tx_ring_stats(vsi, vsi_stats, vsi->xdp_rings, vsi->num_xdp_txq); rcu_read_unlock(); + + vsi->net_stats.tx_packets = vsi_stats->tx_packets; + vsi->net_stats.tx_bytes = vsi_stats->tx_bytes; + vsi->net_stats.rx_packets = vsi_stats->rx_packets; + vsi->net_stats.rx_bytes = vsi_stats->rx_bytes; + + kfree(vsi_stats); } /** @@ -5726,7 +6708,7 @@ static void ice_napi_disable_all(struct ice_vsi *vsi) ice_for_each_q_vector(vsi, q_idx) { struct ice_q_vector *q_vector = vsi->q_vectors[q_idx]; - if (q_vector->rx.ring || q_vector->tx.ring) + if (q_vector->rx.rx_ring || q_vector->tx.tx_ring) napi_disable(&q_vector->napi); cancel_work_sync(&q_vector->tx.dim.work); @@ -5737,17 +6719,23 @@ static void ice_napi_disable_all(struct ice_vsi *vsi) /** * ice_down - Shutdown the connection * @vsi: The VSI being stopped + * + * Caller of this function is expected to set the vsi->state ICE_DOWN bit */ int ice_down(struct ice_vsi *vsi) { - int i, tx_err, rx_err, link_err = 0; + int i, tx_err, rx_err, vlan_err = 0; - /* Caller of this function is expected to set the - * vsi->state ICE_DOWN bit - */ - if (vsi->netdev) { + WARN_ON(!test_bit(ICE_VSI_DOWN, vsi->state)); + + if (vsi->netdev && vsi->type == ICE_VSI_PF) { + vlan_err = ice_vsi_del_vlan_zero(vsi); + if (!ice_is_e810(&vsi->back->hw)) + ice_ptp_link_change(vsi->back, vsi->back->hw.pf_id, false); netif_carrier_off(vsi->netdev); netif_tx_disable(vsi->netdev); + } else if (vsi->type == ICE_VSI_SWITCHDEV_CTRL) { + ice_eswitch_stop_all_tx_queues(vsi->back); } ice_vsi_dis_irq(vsi); @@ -5770,20 +6758,13 @@ int ice_down(struct ice_vsi *vsi) ice_napi_disable_all(vsi); - if (test_bit(ICE_FLAG_LINK_DOWN_ON_CLOSE_ENA, vsi->back->flags)) { - link_err = ice_force_phys_link_state(vsi, false); - if (link_err) - netdev_err(vsi->netdev, "Failed to set physical link down, VSI %d error %d\n", - vsi->vsi_num, link_err); - } - ice_for_each_txq(vsi, i) ice_clean_tx_ring(vsi->tx_rings[i]); ice_for_each_rxq(vsi, i) ice_clean_rx_ring(vsi->rx_rings[i]); - if (tx_err || rx_err || link_err) { + if (tx_err || rx_err || vlan_err) { netdev_err(vsi->netdev, "Failed to close VSI 0x%04X on switch 0x%04X\n", vsi->vsi_num, vsi->vsw->sw_id); return -EIO; @@ -5793,6 +6774,31 @@ int ice_down(struct ice_vsi *vsi) } /** + * ice_down_up - shutdown the VSI connection and bring it up + * @vsi: the VSI to be reconnected + */ +int ice_down_up(struct ice_vsi *vsi) +{ + int ret; + + /* if DOWN already set, nothing to do */ + if (test_and_set_bit(ICE_VSI_DOWN, vsi->state)) + return 0; + + ret = ice_down(vsi); + if (ret) + return ret; + + ret = ice_up(vsi); + if (ret) { + netdev_err(vsi->netdev, "reallocating resources failed during netdev features change, may need to reload driver\n"); + return ret; + } + + return 0; +} + +/** * ice_vsi_setup_tx_rings - Allocate VSI Tx queue resources * @vsi: VSI having resources allocated * @@ -5809,12 +6815,13 @@ int ice_vsi_setup_tx_rings(struct ice_vsi *vsi) } ice_for_each_txq(vsi, i) { - struct ice_ring *ring = vsi->tx_rings[i]; + struct ice_tx_ring *ring = vsi->tx_rings[i]; if (!ring) return -EINVAL; - ring->netdev = vsi->netdev; + if (vsi->netdev) + ring->netdev = vsi->netdev; err = ice_setup_tx_ring(ring); if (err) break; @@ -5840,12 +6847,13 @@ int ice_vsi_setup_rx_rings(struct ice_vsi *vsi) } ice_for_each_rxq(vsi, i) { - struct ice_ring *ring = vsi->rx_rings[i]; + struct ice_rx_ring *ring = vsi->rx_rings[i]; if (!ring) return -EINVAL; - ring->netdev = vsi->netdev; + if (vsi->netdev) + ring->netdev = vsi->netdev; err = ice_setup_rx_ring(ring); if (err) break; @@ -5918,7 +6926,7 @@ err_setup_tx: * * Returns 0 on success, negative value on error */ -static int ice_vsi_open(struct ice_vsi *vsi) +int ice_vsi_open(struct ice_vsi *vsi) { char int_name[ICE_INT_NAME_STR_LEN]; struct ice_pf *pf = vsi->back; @@ -5943,14 +6951,18 @@ static int ice_vsi_open(struct ice_vsi *vsi) if (err) goto err_setup_rx; - /* Notify the stack of the actual queue counts. */ - err = netif_set_real_num_tx_queues(vsi->netdev, vsi->num_txq); - if (err) - goto err_set_qs; + ice_vsi_cfg_netdev_tc(vsi, vsi->tc_cfg.ena_tc); - err = netif_set_real_num_rx_queues(vsi->netdev, vsi->num_rxq); - if (err) - goto err_set_qs; + if (vsi->type == ICE_VSI_PF) { + /* Notify the stack of the actual queue counts. */ + err = netif_set_real_num_tx_queues(vsi->netdev, vsi->num_txq); + if (err) + goto err_set_qs; + + err = netif_set_real_num_rx_queues(vsi->netdev, vsi->num_rxq); + if (err) + goto err_set_qs; + } err = ice_up_complete(vsi); if (err) @@ -5985,6 +6997,9 @@ static void ice_vsi_release_all(struct ice_pf *pf) if (!pf->vsi[i]) continue; + if (pf->vsi[i]->type == ICE_VSI_CHNL) + continue; + err = ice_vsi_release(pf->vsi[i]); if (err) dev_dbg(ice_pf_to_dev(pf), "Failed to release pf->vsi[%d], err %d, vsi_num = %d\n", @@ -6002,7 +7017,6 @@ static void ice_vsi_release_all(struct ice_pf *pf) static int ice_vsi_rebuild_by_type(struct ice_pf *pf, enum ice_vsi_type type) { struct device *dev = ice_pf_to_dev(pf); - enum ice_status status; int i, err; ice_for_each_vsi(pf, i) { @@ -6020,12 +7034,11 @@ static int ice_vsi_rebuild_by_type(struct ice_pf *pf, enum ice_vsi_type type) } /* replay filters for the VSI */ - status = ice_replay_vsi(&pf->hw, vsi->idx); - if (status) { - dev_err(dev, "replay VSI failed, status %s, VSI index %d, type %s\n", - ice_stat_str(status), vsi->idx, - ice_vsi_type_str(type)); - return -EIO; + err = ice_replay_vsi(&pf->hw, vsi->idx); + if (err) { + dev_err(dev, "replay VSI failed, error %d, VSI index %d, type %s\n", + err, vsi->idx, ice_vsi_type_str(type)); + return err; } /* Re-map HW VSI number, using VSI handle that has been @@ -6088,7 +7101,7 @@ static void ice_rebuild(struct ice_pf *pf, enum ice_reset_req reset_type) { struct device *dev = ice_pf_to_dev(pf); struct ice_hw *hw = &pf->hw; - enum ice_status ret; + bool dvm; int err; if (test_bit(ICE_DOWN, pf->state)) @@ -6096,10 +7109,20 @@ static void ice_rebuild(struct ice_pf *pf, enum ice_reset_req reset_type) dev_dbg(dev, "rebuilding PF after reset_type=%d\n", reset_type); - ret = ice_init_all_ctrlq(hw); - if (ret) { - dev_err(dev, "control queues init failed %s\n", - ice_stat_str(ret)); +#define ICE_EMP_RESET_SLEEP_MS 5000 + if (reset_type == ICE_RESET_EMPR) { + /* If an EMP reset has occurred, any previously pending flash + * update will have completed. We no longer know whether or + * not the NVM update EMP reset is restricted. + */ + pf->fw_emp_reset_disabled = false; + + msleep(ICE_EMP_RESET_SLEEP_MS); + } + + err = ice_init_all_ctrlq(hw); + if (err) { + dev_err(dev, "control queues init failed %d\n", err); goto err_init_ctrlq; } @@ -6113,33 +7136,38 @@ static void ice_rebuild(struct ice_pf *pf, enum ice_reset_req reset_type) ice_load_pkg(NULL, pf); } - ret = ice_clear_pf_cfg(hw); - if (ret) { - dev_err(dev, "clear PF configuration failed %s\n", - ice_stat_str(ret)); + err = ice_clear_pf_cfg(hw); + if (err) { + dev_err(dev, "clear PF configuration failed %d\n", err); goto err_init_ctrlq; } - if (pf->first_sw->dflt_vsi_ena) - dev_info(dev, "Clearing default VSI, re-enable after reset completes\n"); - /* clear the default VSI configuration if it exists */ - pf->first_sw->dflt_vsi = NULL; - pf->first_sw->dflt_vsi_ena = false; - ice_clear_pxe_mode(hw); - ret = ice_get_caps(hw); - if (ret) { - dev_err(dev, "ice_get_caps failed %s\n", ice_stat_str(ret)); + err = ice_init_nvm(hw); + if (err) { + dev_err(dev, "ice_init_nvm failed %d\n", err); goto err_init_ctrlq; } - ret = ice_aq_set_mac_cfg(hw, ICE_AQ_SET_MAC_FRAME_SIZE_MAX, NULL); - if (ret) { - dev_err(dev, "set_mac_cfg failed %s\n", ice_stat_str(ret)); + err = ice_get_caps(hw); + if (err) { + dev_err(dev, "ice_get_caps failed %d\n", err); goto err_init_ctrlq; } + err = ice_aq_set_mac_cfg(hw, ICE_AQ_SET_MAC_FRAME_SIZE_MAX, NULL); + if (err) { + dev_err(dev, "set_mac_cfg failed %d\n", err); + goto err_init_ctrlq; + } + + dvm = ice_is_dvm_ena(hw); + + err = ice_aq_set_port_params(pf->hw.port_info, dvm, NULL); + if (err) + goto err_init_ctrlq; + err = ice_sched_init_port(hw->port_info); if (err) goto err_sched_init_port; @@ -6169,6 +7197,16 @@ static void ice_rebuild(struct ice_pf *pf, enum ice_reset_req reset_type) if (test_bit(ICE_FLAG_DCB_ENA, pf->flags)) ice_dcb_rebuild(pf); + /* If the PF previously had enabled PTP, PTP init needs to happen before + * the VSI rebuild. If not, this causes the PTP link status events to + * fail. + */ + if (test_bit(ICE_FLAG_PTP_SUPPORTED, pf->flags)) + ice_ptp_reset(pf); + + if (ice_is_feature_supported(pf, ICE_F_GNSS)) + ice_gnss_init(pf); + /* rebuild PF VSI */ err = ice_vsi_rebuild_by_type(pf, ICE_VSI_PF); if (err) { @@ -6176,6 +7214,25 @@ static void ice_rebuild(struct ice_pf *pf, enum ice_reset_req reset_type) goto err_vsi_rebuild; } + /* configure PTP timestamping after VSI rebuild */ + if (test_bit(ICE_FLAG_PTP_SUPPORTED, pf->flags)) + ice_ptp_cfg_timestamp(pf, false); + + err = ice_vsi_rebuild_by_type(pf, ICE_VSI_SWITCHDEV_CTRL); + if (err) { + dev_err(dev, "Switchdev CTRL VSI rebuild failed: %d\n", err); + goto err_vsi_rebuild; + } + + if (reset_type == ICE_RESET_PFR) { + err = ice_rebuild_channels(pf); + if (err) { + dev_err(dev, "failed to rebuild and replay ADQ VSIs, err %d\n", + err); + goto err_vsi_rebuild; + } + } + /* If Flow Director is active */ if (test_bit(ICE_FLAG_FD_ENA, pf->flags)) { err = ice_vsi_rebuild_by_type(pf, ICE_VSI_CTRL); @@ -6197,10 +7254,10 @@ static void ice_rebuild(struct ice_pf *pf, enum ice_reset_req reset_type) ice_update_pf_netdev_link(pf); /* tell the firmware we are up */ - ret = ice_send_version(pf); - if (ret) { - dev_err(dev, "Rebuild failed due to error sending driver version: %s\n", - ice_stat_str(ret)); + err = ice_send_version(pf); + if (err) { + dev_err(dev, "Rebuild failed due to error sending driver version: %d\n", + err); goto err_vsi_rebuild; } @@ -6208,6 +7265,8 @@ static void ice_rebuild(struct ice_pf *pf, enum ice_reset_req reset_type) /* if we get here, reset flow is successful */ clear_bit(ICE_RESET_FAILED, pf->state); + + ice_plug_aux_dev(pf); return; err_vsi_rebuild: @@ -6247,6 +7306,7 @@ static int ice_change_mtu(struct net_device *netdev, int new_mtu) struct ice_vsi *vsi = np->vsi; struct ice_pf *pf = vsi->back; u8 count = 0; + int err = 0; if (new_mtu == (int)netdev->mtu) { netdev_warn(netdev, "MTU is already %u\n", netdev->mtu); @@ -6283,8 +7343,6 @@ static int ice_change_mtu(struct net_device *netdev, int new_mtu) /* if VSI is up, bring it down and then back up */ if (!test_and_set_bit(ICE_VSI_DOWN, vsi->state)) { - int err; - err = ice_down(vsi); if (err) { netdev_err(netdev, "change MTU if_down err %d\n", err); @@ -6299,7 +7357,30 @@ static int ice_change_mtu(struct net_device *netdev, int new_mtu) } netdev_dbg(netdev, "changed MTU to %d\n", new_mtu); - return 0; + set_bit(ICE_FLAG_MTU_CHANGED, pf->flags); + + return err; +} + +/** + * ice_eth_ioctl - Access the hwtstamp interface + * @netdev: network interface device structure + * @ifr: interface request data + * @cmd: ioctl command + */ +static int ice_eth_ioctl(struct net_device *netdev, struct ifreq *ifr, int cmd) +{ + struct ice_netdev_priv *np = netdev_priv(netdev); + struct ice_pf *pf = np->vsi->back; + + switch (cmd) { + case SIOCGHWTSTAMP: + return ice_ptp_get_ts_config(pf, ifr); + case SIOCSHWTSTAMP: + return ice_ptp_set_ts_config(pf, ifr); + default: + return -EOPNOTSUPP; + } } /** @@ -6345,78 +7426,6 @@ const char *ice_aq_str(enum ice_aq_err aq_err) } /** - * ice_stat_str - convert status err code to a string - * @stat_err: the status error code to convert - */ -const char *ice_stat_str(enum ice_status stat_err) -{ - switch (stat_err) { - case ICE_SUCCESS: - return "OK"; - case ICE_ERR_PARAM: - return "ICE_ERR_PARAM"; - case ICE_ERR_NOT_IMPL: - return "ICE_ERR_NOT_IMPL"; - case ICE_ERR_NOT_READY: - return "ICE_ERR_NOT_READY"; - case ICE_ERR_NOT_SUPPORTED: - return "ICE_ERR_NOT_SUPPORTED"; - case ICE_ERR_BAD_PTR: - return "ICE_ERR_BAD_PTR"; - case ICE_ERR_INVAL_SIZE: - return "ICE_ERR_INVAL_SIZE"; - case ICE_ERR_DEVICE_NOT_SUPPORTED: - return "ICE_ERR_DEVICE_NOT_SUPPORTED"; - case ICE_ERR_RESET_FAILED: - return "ICE_ERR_RESET_FAILED"; - case ICE_ERR_FW_API_VER: - return "ICE_ERR_FW_API_VER"; - case ICE_ERR_NO_MEMORY: - return "ICE_ERR_NO_MEMORY"; - case ICE_ERR_CFG: - return "ICE_ERR_CFG"; - case ICE_ERR_OUT_OF_RANGE: - return "ICE_ERR_OUT_OF_RANGE"; - case ICE_ERR_ALREADY_EXISTS: - return "ICE_ERR_ALREADY_EXISTS"; - case ICE_ERR_NVM: - return "ICE_ERR_NVM"; - case ICE_ERR_NVM_CHECKSUM: - return "ICE_ERR_NVM_CHECKSUM"; - case ICE_ERR_BUF_TOO_SHORT: - return "ICE_ERR_BUF_TOO_SHORT"; - case ICE_ERR_NVM_BLANK_MODE: - return "ICE_ERR_NVM_BLANK_MODE"; - case ICE_ERR_IN_USE: - return "ICE_ERR_IN_USE"; - case ICE_ERR_MAX_LIMIT: - return "ICE_ERR_MAX_LIMIT"; - case ICE_ERR_RESET_ONGOING: - return "ICE_ERR_RESET_ONGOING"; - case ICE_ERR_HW_TABLE: - return "ICE_ERR_HW_TABLE"; - case ICE_ERR_DOES_NOT_EXIST: - return "ICE_ERR_DOES_NOT_EXIST"; - case ICE_ERR_FW_DDP_MISMATCH: - return "ICE_ERR_FW_DDP_MISMATCH"; - case ICE_ERR_AQ_ERROR: - return "ICE_ERR_AQ_ERROR"; - case ICE_ERR_AQ_TIMEOUT: - return "ICE_ERR_AQ_TIMEOUT"; - case ICE_ERR_AQ_FULL: - return "ICE_ERR_AQ_FULL"; - case ICE_ERR_AQ_NO_WORK: - return "ICE_ERR_AQ_NO_WORK"; - case ICE_ERR_AQ_EMPTY: - return "ICE_ERR_AQ_EMPTY"; - case ICE_ERR_AQ_FW_CRITICAL: - return "ICE_ERR_AQ_FW_CRITICAL"; - } - - return "ICE_ERR_UNKNOWN"; -} - -/** * ice_set_rss_lut - Set RSS LUT * @vsi: Pointer to VSI structure * @lut: Lookup table @@ -6428,7 +7437,7 @@ int ice_set_rss_lut(struct ice_vsi *vsi, u8 *lut, u16 lut_size) { struct ice_aq_get_set_rss_lut_params params = {}; struct ice_hw *hw = &vsi->back->hw; - enum ice_status status; + int status; if (!lut) return -EINVAL; @@ -6439,14 +7448,11 @@ int ice_set_rss_lut(struct ice_vsi *vsi, u8 *lut, u16 lut_size) params.lut = lut; status = ice_aq_set_rss_lut(hw, ¶ms); - if (status) { - dev_err(ice_pf_to_dev(vsi->back), "Cannot set RSS lut, err %s aq_err %s\n", - ice_stat_str(status), - ice_aq_str(hw->adminq.sq_last_status)); - return -EIO; - } + if (status) + dev_err(ice_pf_to_dev(vsi->back), "Cannot set RSS lut, err %d aq_err %s\n", + status, ice_aq_str(hw->adminq.sq_last_status)); - return 0; + return status; } /** @@ -6459,20 +7465,17 @@ int ice_set_rss_lut(struct ice_vsi *vsi, u8 *lut, u16 lut_size) int ice_set_rss_key(struct ice_vsi *vsi, u8 *seed) { struct ice_hw *hw = &vsi->back->hw; - enum ice_status status; + int status; if (!seed) return -EINVAL; status = ice_aq_set_rss_key(hw, vsi->idx, (struct ice_aqc_get_set_rss_keys *)seed); - if (status) { - dev_err(ice_pf_to_dev(vsi->back), "Cannot set RSS key, err %s aq_err %s\n", - ice_stat_str(status), - ice_aq_str(hw->adminq.sq_last_status)); - return -EIO; - } + if (status) + dev_err(ice_pf_to_dev(vsi->back), "Cannot set RSS key, err %d aq_err %s\n", + status, ice_aq_str(hw->adminq.sq_last_status)); - return 0; + return status; } /** @@ -6487,7 +7490,7 @@ int ice_get_rss_lut(struct ice_vsi *vsi, u8 *lut, u16 lut_size) { struct ice_aq_get_set_rss_lut_params params = {}; struct ice_hw *hw = &vsi->back->hw; - enum ice_status status; + int status; if (!lut) return -EINVAL; @@ -6498,14 +7501,11 @@ int ice_get_rss_lut(struct ice_vsi *vsi, u8 *lut, u16 lut_size) params.lut = lut; status = ice_aq_get_rss_lut(hw, ¶ms); - if (status) { - dev_err(ice_pf_to_dev(vsi->back), "Cannot get RSS lut, err %s aq_err %s\n", - ice_stat_str(status), - ice_aq_str(hw->adminq.sq_last_status)); - return -EIO; - } + if (status) + dev_err(ice_pf_to_dev(vsi->back), "Cannot get RSS lut, err %d aq_err %s\n", + status, ice_aq_str(hw->adminq.sq_last_status)); - return 0; + return status; } /** @@ -6518,20 +7518,17 @@ int ice_get_rss_lut(struct ice_vsi *vsi, u8 *lut, u16 lut_size) int ice_get_rss_key(struct ice_vsi *vsi, u8 *seed) { struct ice_hw *hw = &vsi->back->hw; - enum ice_status status; + int status; if (!seed) return -EINVAL; status = ice_aq_get_rss_key(hw, vsi->idx, (struct ice_aqc_get_set_rss_keys *)seed); - if (status) { - dev_err(ice_pf_to_dev(vsi->back), "Cannot get RSS key, err %s aq_err %s\n", - ice_stat_str(status), - ice_aq_str(hw->adminq.sq_last_status)); - return -EIO; - } + if (status) + dev_err(ice_pf_to_dev(vsi->back), "Cannot get RSS key, err %d aq_err %s\n", + status, ice_aq_str(hw->adminq.sq_last_status)); - return 0; + return status; } /** @@ -6572,8 +7569,7 @@ static int ice_vsi_update_bridge_mode(struct ice_vsi *vsi, u16 bmode) struct ice_aqc_vsi_props *vsi_props; struct ice_hw *hw = &vsi->back->hw; struct ice_vsi_ctx *ctxt; - enum ice_status status; - int ret = 0; + int ret; vsi_props = &vsi->info; @@ -6591,12 +7587,10 @@ static int ice_vsi_update_bridge_mode(struct ice_vsi *vsi, u16 bmode) ctxt->info.sw_flags &= ~ICE_AQ_VSI_SW_FLAG_ALLOW_LB; ctxt->info.valid_sections = cpu_to_le16(ICE_AQ_VSI_PROP_SW_VALID); - status = ice_update_vsi(hw, vsi->idx, ctxt, NULL); - if (status) { - dev_err(ice_pf_to_dev(vsi->back), "update VSI for bridge mode failed, bmode = %d err %s aq_err %s\n", - bmode, ice_stat_str(status), - ice_aq_str(hw->adminq.sq_last_status)); - ret = -EIO; + ret = ice_update_vsi(hw, vsi->idx, ctxt, NULL); + if (ret) { + dev_err(ice_pf_to_dev(vsi->back), "update VSI for bridge mode failed, bmode = %d err %d aq_err %s\n", + bmode, ret, ice_aq_str(hw->adminq.sq_last_status)); goto out; } /* Update sw flags for book keeping */ @@ -6628,7 +7622,6 @@ ice_bridge_setlink(struct net_device *dev, struct nlmsghdr *nlh, struct ice_pf *pf = np->vsi->back; struct nlattr *attr, *br_spec; struct ice_hw *hw = &pf->hw; - enum ice_status status; struct ice_sw *pf_sw; int rem, v, err = 0; @@ -6662,14 +7655,14 @@ ice_bridge_setlink(struct net_device *dev, struct nlmsghdr *nlh, /* Update the unicast switch filter rules for the corresponding * switch of the netdev */ - status = ice_update_sw_rule_bridge_mode(hw); - if (status) { - netdev_err(dev, "switch rule update failed, mode = %d err %s aq_err %s\n", - mode, ice_stat_str(status), + err = ice_update_sw_rule_bridge_mode(hw); + if (err) { + netdev_err(dev, "switch rule update failed, mode = %d err %d aq_err %s\n", + mode, err, ice_aq_str(hw->adminq.sq_last_status)); /* revert hw->evb_veb */ hw->evb_veb = (pf_sw->bridge_mode == BRIDGE_MODE_VEB); - return -EIO; + return err; } pf_sw->bridge_mode = mode; @@ -6686,7 +7679,7 @@ ice_bridge_setlink(struct net_device *dev, struct nlmsghdr *nlh, static void ice_tx_timeout(struct net_device *netdev, unsigned int txqueue) { struct ice_netdev_priv *np = netdev_priv(netdev); - struct ice_ring *tx_ring = NULL; + struct ice_tx_ring *tx_ring = NULL; struct ice_vsi *vsi = np->vsi; struct ice_pf *pf = vsi->back; u32 i; @@ -6704,7 +7697,7 @@ static void ice_tx_timeout(struct net_device *netdev, unsigned int txqueue) } /* now that we have an index, find the tx_ring struct */ - for (i = 0; i < vsi->num_txq; i++) + ice_for_each_txq(vsi, 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]; @@ -6761,6 +7754,1125 @@ static void ice_tx_timeout(struct net_device *netdev, unsigned int txqueue) } /** + * ice_setup_tc_cls_flower - flower classifier offloads + * @np: net device to configure + * @filter_dev: device on which filter is added + * @cls_flower: offload data + */ +static int +ice_setup_tc_cls_flower(struct ice_netdev_priv *np, + struct net_device *filter_dev, + struct flow_cls_offload *cls_flower) +{ + struct ice_vsi *vsi = np->vsi; + + if (cls_flower->common.chain_index) + return -EOPNOTSUPP; + + switch (cls_flower->command) { + case FLOW_CLS_REPLACE: + return ice_add_cls_flower(filter_dev, vsi, cls_flower); + case FLOW_CLS_DESTROY: + return ice_del_cls_flower(vsi, cls_flower); + default: + return -EINVAL; + } +} + +/** + * ice_setup_tc_block_cb - callback handler registered for TC block + * @type: TC SETUP type + * @type_data: TC flower offload data that contains user input + * @cb_priv: netdev private data + */ +static int +ice_setup_tc_block_cb(enum tc_setup_type type, void *type_data, void *cb_priv) +{ + struct ice_netdev_priv *np = cb_priv; + + switch (type) { + case TC_SETUP_CLSFLOWER: + return ice_setup_tc_cls_flower(np, np->vsi->netdev, + type_data); + default: + return -EOPNOTSUPP; + } +} + +/** + * ice_validate_mqprio_qopt - Validate TCF input parameters + * @vsi: Pointer to VSI + * @mqprio_qopt: input parameters for mqprio queue configuration + * + * This function validates MQPRIO params, such as qcount (power of 2 wherever + * needed), and make sure user doesn't specify qcount and BW rate limit + * for TCs, which are more than "num_tc" + */ +static int +ice_validate_mqprio_qopt(struct ice_vsi *vsi, + struct tc_mqprio_qopt_offload *mqprio_qopt) +{ + u64 sum_max_rate = 0, sum_min_rate = 0; + int non_power_of_2_qcount = 0; + struct ice_pf *pf = vsi->back; + int max_rss_q_cnt = 0; + struct device *dev; + int i, speed; + u8 num_tc; + + if (vsi->type != ICE_VSI_PF) + return -EINVAL; + + if (mqprio_qopt->qopt.offset[0] != 0 || + mqprio_qopt->qopt.num_tc < 1 || + mqprio_qopt->qopt.num_tc > ICE_CHNL_MAX_TC) + return -EINVAL; + + dev = ice_pf_to_dev(pf); + vsi->ch_rss_size = 0; + num_tc = mqprio_qopt->qopt.num_tc; + + for (i = 0; num_tc; i++) { + int qcount = mqprio_qopt->qopt.count[i]; + u64 max_rate, min_rate, rem; + + if (!qcount) + return -EINVAL; + + if (is_power_of_2(qcount)) { + if (non_power_of_2_qcount && + qcount > non_power_of_2_qcount) { + dev_err(dev, "qcount[%d] cannot be greater than non power of 2 qcount[%d]\n", + qcount, non_power_of_2_qcount); + return -EINVAL; + } + if (qcount > max_rss_q_cnt) + max_rss_q_cnt = qcount; + } else { + if (non_power_of_2_qcount && + qcount != non_power_of_2_qcount) { + dev_err(dev, "Only one non power of 2 qcount allowed[%d,%d]\n", + qcount, non_power_of_2_qcount); + return -EINVAL; + } + if (qcount < max_rss_q_cnt) { + dev_err(dev, "non power of 2 qcount[%d] cannot be less than other qcount[%d]\n", + qcount, max_rss_q_cnt); + return -EINVAL; + } + max_rss_q_cnt = qcount; + non_power_of_2_qcount = qcount; + } + + /* TC command takes input in K/N/Gbps or K/M/Gbit etc but + * converts the bandwidth rate limit into Bytes/s when + * passing it down to the driver. So convert input bandwidth + * from Bytes/s to Kbps + */ + max_rate = mqprio_qopt->max_rate[i]; + max_rate = div_u64(max_rate, ICE_BW_KBPS_DIVISOR); + sum_max_rate += max_rate; + + /* min_rate is minimum guaranteed rate and it can't be zero */ + min_rate = mqprio_qopt->min_rate[i]; + min_rate = div_u64(min_rate, ICE_BW_KBPS_DIVISOR); + sum_min_rate += min_rate; + + if (min_rate && min_rate < ICE_MIN_BW_LIMIT) { + dev_err(dev, "TC%d: min_rate(%llu Kbps) < %u Kbps\n", i, + min_rate, ICE_MIN_BW_LIMIT); + return -EINVAL; + } + + iter_div_u64_rem(min_rate, ICE_MIN_BW_LIMIT, &rem); + if (rem) { + dev_err(dev, "TC%d: Min Rate not multiple of %u Kbps", + i, ICE_MIN_BW_LIMIT); + return -EINVAL; + } + + iter_div_u64_rem(max_rate, ICE_MIN_BW_LIMIT, &rem); + if (rem) { + dev_err(dev, "TC%d: Max Rate not multiple of %u Kbps", + i, ICE_MIN_BW_LIMIT); + return -EINVAL; + } + + /* min_rate can't be more than max_rate, except when max_rate + * is zero (implies max_rate sought is max line rate). In such + * a case min_rate can be more than max. + */ + if (max_rate && min_rate > max_rate) { + dev_err(dev, "min_rate %llu Kbps can't be more than max_rate %llu Kbps\n", + min_rate, max_rate); + return -EINVAL; + } + + if (i >= mqprio_qopt->qopt.num_tc - 1) + break; + if (mqprio_qopt->qopt.offset[i + 1] != + (mqprio_qopt->qopt.offset[i] + qcount)) + return -EINVAL; + } + if (vsi->num_rxq < + (mqprio_qopt->qopt.offset[i] + mqprio_qopt->qopt.count[i])) + return -EINVAL; + if (vsi->num_txq < + (mqprio_qopt->qopt.offset[i] + mqprio_qopt->qopt.count[i])) + return -EINVAL; + + speed = ice_get_link_speed_kbps(vsi); + if (sum_max_rate && sum_max_rate > (u64)speed) { + dev_err(dev, "Invalid max Tx rate(%llu) Kbps > speed(%u) Kbps specified\n", + sum_max_rate, speed); + return -EINVAL; + } + if (sum_min_rate && sum_min_rate > (u64)speed) { + dev_err(dev, "Invalid min Tx rate(%llu) Kbps > speed (%u) Kbps specified\n", + sum_min_rate, speed); + return -EINVAL; + } + + /* make sure vsi->ch_rss_size is set correctly based on TC's qcount */ + vsi->ch_rss_size = max_rss_q_cnt; + + return 0; +} + +/** + * ice_add_vsi_to_fdir - add a VSI to the flow director group for PF + * @pf: ptr to PF device + * @vsi: ptr to VSI + */ +static int ice_add_vsi_to_fdir(struct ice_pf *pf, struct ice_vsi *vsi) +{ + struct device *dev = ice_pf_to_dev(pf); + bool added = false; + struct ice_hw *hw; + int flow; + + if (!(vsi->num_gfltr || vsi->num_bfltr)) + return -EINVAL; + + hw = &pf->hw; + for (flow = 0; flow < ICE_FLTR_PTYPE_MAX; flow++) { + struct ice_fd_hw_prof *prof; + int tun, status; + u64 entry_h; + + if (!(hw->fdir_prof && hw->fdir_prof[flow] && + hw->fdir_prof[flow]->cnt)) + continue; + + for (tun = 0; tun < ICE_FD_HW_SEG_MAX; tun++) { + enum ice_flow_priority prio; + u64 prof_id; + + /* add this VSI to FDir profile for this flow */ + prio = ICE_FLOW_PRIO_NORMAL; + prof = hw->fdir_prof[flow]; + prof_id = flow + tun * ICE_FLTR_PTYPE_MAX; + status = ice_flow_add_entry(hw, ICE_BLK_FD, prof_id, + prof->vsi_h[0], vsi->idx, + prio, prof->fdir_seg[tun], + &entry_h); + if (status) { + dev_err(dev, "channel VSI idx %d, not able to add to group %d\n", + vsi->idx, flow); + continue; + } + + prof->entry_h[prof->cnt][tun] = entry_h; + } + + /* store VSI for filter replay and delete */ + prof->vsi_h[prof->cnt] = vsi->idx; + prof->cnt++; + + added = true; + dev_dbg(dev, "VSI idx %d added to fdir group %d\n", vsi->idx, + flow); + } + + if (!added) + dev_dbg(dev, "VSI idx %d not added to fdir groups\n", vsi->idx); + + return 0; +} + +/** + * ice_add_channel - add a channel by adding VSI + * @pf: ptr to PF device + * @sw_id: underlying HW switching element ID + * @ch: ptr to channel structure + * + * Add a channel (VSI) using add_vsi and queue_map + */ +static int ice_add_channel(struct ice_pf *pf, u16 sw_id, struct ice_channel *ch) +{ + struct device *dev = ice_pf_to_dev(pf); + struct ice_vsi *vsi; + + if (ch->type != ICE_VSI_CHNL) { + dev_err(dev, "add new VSI failed, ch->type %d\n", ch->type); + return -EINVAL; + } + + vsi = ice_chnl_vsi_setup(pf, pf->hw.port_info, ch); + if (!vsi || vsi->type != ICE_VSI_CHNL) { + dev_err(dev, "create chnl VSI failure\n"); + return -EINVAL; + } + + ice_add_vsi_to_fdir(pf, vsi); + + ch->sw_id = sw_id; + ch->vsi_num = vsi->vsi_num; + ch->info.mapping_flags = vsi->info.mapping_flags; + ch->ch_vsi = vsi; + /* set the back pointer of channel for newly created VSI */ + vsi->ch = ch; + + memcpy(&ch->info.q_mapping, &vsi->info.q_mapping, + sizeof(vsi->info.q_mapping)); + memcpy(&ch->info.tc_mapping, vsi->info.tc_mapping, + sizeof(vsi->info.tc_mapping)); + + return 0; +} + +/** + * ice_chnl_cfg_res + * @vsi: the VSI being setup + * @ch: ptr to channel structure + * + * Configure channel specific resources such as rings, vector. + */ +static void ice_chnl_cfg_res(struct ice_vsi *vsi, struct ice_channel *ch) +{ + int i; + + for (i = 0; i < ch->num_txq; i++) { + struct ice_q_vector *tx_q_vector, *rx_q_vector; + struct ice_ring_container *rc; + struct ice_tx_ring *tx_ring; + struct ice_rx_ring *rx_ring; + + tx_ring = vsi->tx_rings[ch->base_q + i]; + rx_ring = vsi->rx_rings[ch->base_q + i]; + if (!tx_ring || !rx_ring) + continue; + + /* setup ring being channel enabled */ + tx_ring->ch = ch; + rx_ring->ch = ch; + + /* following code block sets up vector specific attributes */ + tx_q_vector = tx_ring->q_vector; + rx_q_vector = rx_ring->q_vector; + if (!tx_q_vector && !rx_q_vector) + continue; + + if (tx_q_vector) { + tx_q_vector->ch = ch; + /* setup Tx and Rx ITR setting if DIM is off */ + rc = &tx_q_vector->tx; + if (!ITR_IS_DYNAMIC(rc)) + ice_write_itr(rc, rc->itr_setting); + } + if (rx_q_vector) { + rx_q_vector->ch = ch; + /* setup Tx and Rx ITR setting if DIM is off */ + rc = &rx_q_vector->rx; + if (!ITR_IS_DYNAMIC(rc)) + ice_write_itr(rc, rc->itr_setting); + } + } + + /* it is safe to assume that, if channel has non-zero num_t[r]xq, then + * GLINT_ITR register would have written to perform in-context + * update, hence perform flush + */ + if (ch->num_txq || ch->num_rxq) + ice_flush(&vsi->back->hw); +} + +/** + * ice_cfg_chnl_all_res - configure channel resources + * @vsi: pte to main_vsi + * @ch: ptr to channel structure + * + * This function configures channel specific resources such as flow-director + * counter index, and other resources such as queues, vectors, ITR settings + */ +static void +ice_cfg_chnl_all_res(struct ice_vsi *vsi, struct ice_channel *ch) +{ + /* configure channel (aka ADQ) resources such as queues, vectors, + * ITR settings for channel specific vectors and anything else + */ + ice_chnl_cfg_res(vsi, ch); +} + +/** + * ice_setup_hw_channel - setup new channel + * @pf: ptr to PF device + * @vsi: the VSI being setup + * @ch: ptr to channel structure + * @sw_id: underlying HW switching element ID + * @type: type of channel to be created (VMDq2/VF) + * + * Setup new channel (VSI) based on specified type (VMDq2/VF) + * and configures Tx rings accordingly + */ +static int +ice_setup_hw_channel(struct ice_pf *pf, struct ice_vsi *vsi, + struct ice_channel *ch, u16 sw_id, u8 type) +{ + struct device *dev = ice_pf_to_dev(pf); + int ret; + + ch->base_q = vsi->next_base_q; + ch->type = type; + + ret = ice_add_channel(pf, sw_id, ch); + if (ret) { + dev_err(dev, "failed to add_channel using sw_id %u\n", sw_id); + return ret; + } + + /* configure/setup ADQ specific resources */ + ice_cfg_chnl_all_res(vsi, ch); + + /* make sure to update the next_base_q so that subsequent channel's + * (aka ADQ) VSI queue map is correct + */ + vsi->next_base_q = vsi->next_base_q + ch->num_rxq; + dev_dbg(dev, "added channel: vsi_num %u, num_rxq %u\n", ch->vsi_num, + ch->num_rxq); + + return 0; +} + +/** + * ice_setup_channel - setup new channel using uplink element + * @pf: ptr to PF device + * @vsi: the VSI being setup + * @ch: ptr to channel structure + * + * Setup new channel (VSI) based on specified type (VMDq2/VF) + * and uplink switching element + */ +static bool +ice_setup_channel(struct ice_pf *pf, struct ice_vsi *vsi, + struct ice_channel *ch) +{ + struct device *dev = ice_pf_to_dev(pf); + u16 sw_id; + int ret; + + if (vsi->type != ICE_VSI_PF) { + dev_err(dev, "unsupported parent VSI type(%d)\n", vsi->type); + return false; + } + + sw_id = pf->first_sw->sw_id; + + /* create channel (VSI) */ + ret = ice_setup_hw_channel(pf, vsi, ch, sw_id, ICE_VSI_CHNL); + if (ret) { + dev_err(dev, "failed to setup hw_channel\n"); + return false; + } + dev_dbg(dev, "successfully created channel()\n"); + + return ch->ch_vsi ? true : false; +} + +/** + * ice_set_bw_limit - setup BW limit for Tx traffic based on max_tx_rate + * @vsi: VSI to be configured + * @max_tx_rate: max Tx rate in Kbps to be configured as maximum BW limit + * @min_tx_rate: min Tx rate in Kbps to be configured as minimum BW limit + */ +static int +ice_set_bw_limit(struct ice_vsi *vsi, u64 max_tx_rate, u64 min_tx_rate) +{ + int err; + + err = ice_set_min_bw_limit(vsi, min_tx_rate); + if (err) + return err; + + return ice_set_max_bw_limit(vsi, max_tx_rate); +} + +/** + * ice_create_q_channel - function to create channel + * @vsi: VSI to be configured + * @ch: ptr to channel (it contains channel specific params) + * + * This function creates channel (VSI) using num_queues specified by user, + * reconfigs RSS if needed. + */ +static int ice_create_q_channel(struct ice_vsi *vsi, struct ice_channel *ch) +{ + struct ice_pf *pf = vsi->back; + struct device *dev; + + if (!ch) + return -EINVAL; + + dev = ice_pf_to_dev(pf); + if (!ch->num_txq || !ch->num_rxq) { + dev_err(dev, "Invalid num_queues requested: %d\n", ch->num_rxq); + return -EINVAL; + } + + if (!vsi->cnt_q_avail || vsi->cnt_q_avail < ch->num_txq) { + dev_err(dev, "cnt_q_avail (%u) less than num_queues %d\n", + vsi->cnt_q_avail, ch->num_txq); + return -EINVAL; + } + + if (!ice_setup_channel(pf, vsi, ch)) { + dev_info(dev, "Failed to setup channel\n"); + return -EINVAL; + } + /* configure BW rate limit */ + if (ch->ch_vsi && (ch->max_tx_rate || ch->min_tx_rate)) { + int ret; + + ret = ice_set_bw_limit(ch->ch_vsi, ch->max_tx_rate, + ch->min_tx_rate); + if (ret) + dev_err(dev, "failed to set Tx rate of %llu Kbps for VSI(%u)\n", + ch->max_tx_rate, ch->ch_vsi->vsi_num); + else + dev_dbg(dev, "set Tx rate of %llu Kbps for VSI(%u)\n", + ch->max_tx_rate, ch->ch_vsi->vsi_num); + } + + vsi->cnt_q_avail -= ch->num_txq; + + return 0; +} + +/** + * ice_rem_all_chnl_fltrs - removes all channel filters + * @pf: ptr to PF, TC-flower based filter are tracked at PF level + * + * Remove all advanced switch filters only if they are channel specific + * tc-flower based filter + */ +static void ice_rem_all_chnl_fltrs(struct ice_pf *pf) +{ + struct ice_tc_flower_fltr *fltr; + struct hlist_node *node; + + /* to remove all channel filters, iterate an ordered list of filters */ + hlist_for_each_entry_safe(fltr, node, + &pf->tc_flower_fltr_list, + tc_flower_node) { + struct ice_rule_query_data rule; + int status; + + /* for now process only channel specific filters */ + if (!ice_is_chnl_fltr(fltr)) + continue; + + rule.rid = fltr->rid; + rule.rule_id = fltr->rule_id; + rule.vsi_handle = fltr->dest_id; + status = ice_rem_adv_rule_by_id(&pf->hw, &rule); + if (status) { + if (status == -ENOENT) + dev_dbg(ice_pf_to_dev(pf), "TC flower filter (rule_id %u) does not exist\n", + rule.rule_id); + else + dev_err(ice_pf_to_dev(pf), "failed to delete TC flower filter, status %d\n", + status); + } else if (fltr->dest_vsi) { + /* update advanced switch filter count */ + if (fltr->dest_vsi->type == ICE_VSI_CHNL) { + u32 flags = fltr->flags; + + fltr->dest_vsi->num_chnl_fltr--; + if (flags & (ICE_TC_FLWR_FIELD_DST_MAC | + ICE_TC_FLWR_FIELD_ENC_DST_MAC)) + pf->num_dmac_chnl_fltrs--; + } + } + + hlist_del(&fltr->tc_flower_node); + kfree(fltr); + } +} + +/** + * ice_remove_q_channels - Remove queue channels for the TCs + * @vsi: VSI to be configured + * @rem_fltr: delete advanced switch filter or not + * + * Remove queue channels for the TCs + */ +static void ice_remove_q_channels(struct ice_vsi *vsi, bool rem_fltr) +{ + struct ice_channel *ch, *ch_tmp; + struct ice_pf *pf = vsi->back; + int i; + + /* remove all tc-flower based filter if they are channel filters only */ + if (rem_fltr) + ice_rem_all_chnl_fltrs(pf); + + /* remove ntuple filters since queue configuration is being changed */ + if (vsi->netdev->features & NETIF_F_NTUPLE) { + struct ice_hw *hw = &pf->hw; + + mutex_lock(&hw->fdir_fltr_lock); + ice_fdir_del_all_fltrs(vsi); + mutex_unlock(&hw->fdir_fltr_lock); + } + + /* perform cleanup for channels if they exist */ + list_for_each_entry_safe(ch, ch_tmp, &vsi->ch_list, list) { + struct ice_vsi *ch_vsi; + + list_del(&ch->list); + ch_vsi = ch->ch_vsi; + if (!ch_vsi) { + kfree(ch); + continue; + } + + /* Reset queue contexts */ + for (i = 0; i < ch->num_rxq; i++) { + struct ice_tx_ring *tx_ring; + struct ice_rx_ring *rx_ring; + + tx_ring = vsi->tx_rings[ch->base_q + i]; + rx_ring = vsi->rx_rings[ch->base_q + i]; + if (tx_ring) { + tx_ring->ch = NULL; + if (tx_ring->q_vector) + tx_ring->q_vector->ch = NULL; + } + if (rx_ring) { + rx_ring->ch = NULL; + if (rx_ring->q_vector) + rx_ring->q_vector->ch = NULL; + } + } + + /* Release FD resources for the channel VSI */ + ice_fdir_rem_adq_chnl(&pf->hw, ch->ch_vsi->idx); + + /* clear the VSI from scheduler tree */ + ice_rm_vsi_lan_cfg(ch->ch_vsi->port_info, ch->ch_vsi->idx); + + /* Delete VSI from FW */ + ice_vsi_delete(ch->ch_vsi); + + /* Delete VSI from PF and HW VSI arrays */ + ice_vsi_clear(ch->ch_vsi); + + /* free the channel */ + kfree(ch); + } + + /* clear the channel VSI map which is stored in main VSI */ + ice_for_each_chnl_tc(i) + vsi->tc_map_vsi[i] = NULL; + + /* reset main VSI's all TC information */ + vsi->all_enatc = 0; + vsi->all_numtc = 0; +} + +/** + * ice_rebuild_channels - rebuild channel + * @pf: ptr to PF + * + * Recreate channel VSIs and replay filters + */ +static int ice_rebuild_channels(struct ice_pf *pf) +{ + struct device *dev = ice_pf_to_dev(pf); + struct ice_vsi *main_vsi; + bool rem_adv_fltr = true; + struct ice_channel *ch; + struct ice_vsi *vsi; + int tc_idx = 1; + int i, err; + + main_vsi = ice_get_main_vsi(pf); + if (!main_vsi) + return 0; + + if (!test_bit(ICE_FLAG_TC_MQPRIO, pf->flags) || + main_vsi->old_numtc == 1) + return 0; /* nothing to be done */ + + /* reconfigure main VSI based on old value of TC and cached values + * for MQPRIO opts + */ + err = ice_vsi_cfg_tc(main_vsi, main_vsi->old_ena_tc); + if (err) { + dev_err(dev, "failed configuring TC(ena_tc:0x%02x) for HW VSI=%u\n", + main_vsi->old_ena_tc, main_vsi->vsi_num); + return err; + } + + /* rebuild ADQ VSIs */ + ice_for_each_vsi(pf, i) { + enum ice_vsi_type type; + + vsi = pf->vsi[i]; + if (!vsi || vsi->type != ICE_VSI_CHNL) + continue; + + type = vsi->type; + + /* rebuild ADQ VSI */ + err = ice_vsi_rebuild(vsi, true); + if (err) { + dev_err(dev, "VSI (type:%s) at index %d rebuild failed, err %d\n", + ice_vsi_type_str(type), vsi->idx, err); + goto cleanup; + } + + /* Re-map HW VSI number, using VSI handle that has been + * previously validated in ice_replay_vsi() call above + */ + vsi->vsi_num = ice_get_hw_vsi_num(&pf->hw, vsi->idx); + + /* replay filters for the VSI */ + err = ice_replay_vsi(&pf->hw, vsi->idx); + if (err) { + dev_err(dev, "VSI (type:%s) replay failed, err %d, VSI index %d\n", + ice_vsi_type_str(type), err, vsi->idx); + rem_adv_fltr = false; + goto cleanup; + } + dev_info(dev, "VSI (type:%s) at index %d rebuilt successfully\n", + ice_vsi_type_str(type), vsi->idx); + + /* store ADQ VSI at correct TC index in main VSI's + * map of TC to VSI + */ + main_vsi->tc_map_vsi[tc_idx++] = vsi; + } + + /* ADQ VSI(s) has been rebuilt successfully, so setup + * channel for main VSI's Tx and Rx rings + */ + list_for_each_entry(ch, &main_vsi->ch_list, list) { + struct ice_vsi *ch_vsi; + + ch_vsi = ch->ch_vsi; + if (!ch_vsi) + continue; + + /* reconfig channel resources */ + ice_cfg_chnl_all_res(main_vsi, ch); + + /* replay BW rate limit if it is non-zero */ + if (!ch->max_tx_rate && !ch->min_tx_rate) + continue; + + err = ice_set_bw_limit(ch_vsi, ch->max_tx_rate, + ch->min_tx_rate); + if (err) + dev_err(dev, "failed (err:%d) to rebuild BW rate limit, max_tx_rate: %llu Kbps, min_tx_rate: %llu Kbps for VSI(%u)\n", + err, ch->max_tx_rate, ch->min_tx_rate, + ch_vsi->vsi_num); + else + dev_dbg(dev, "successfully rebuild BW rate limit, max_tx_rate: %llu Kbps, min_tx_rate: %llu Kbps for VSI(%u)\n", + ch->max_tx_rate, ch->min_tx_rate, + ch_vsi->vsi_num); + } + + /* reconfig RSS for main VSI */ + if (main_vsi->ch_rss_size) + ice_vsi_cfg_rss_lut_key(main_vsi); + + return 0; + +cleanup: + ice_remove_q_channels(main_vsi, rem_adv_fltr); + return err; +} + +/** + * ice_create_q_channels - Add queue channel for the given TCs + * @vsi: VSI to be configured + * + * Configures queue channel mapping to the given TCs + */ +static int ice_create_q_channels(struct ice_vsi *vsi) +{ + struct ice_pf *pf = vsi->back; + struct ice_channel *ch; + int ret = 0, i; + + ice_for_each_chnl_tc(i) { + if (!(vsi->all_enatc & BIT(i))) + continue; + + ch = kzalloc(sizeof(*ch), GFP_KERNEL); + if (!ch) { + ret = -ENOMEM; + goto err_free; + } + INIT_LIST_HEAD(&ch->list); + ch->num_rxq = vsi->mqprio_qopt.qopt.count[i]; + ch->num_txq = vsi->mqprio_qopt.qopt.count[i]; + ch->base_q = vsi->mqprio_qopt.qopt.offset[i]; + ch->max_tx_rate = vsi->mqprio_qopt.max_rate[i]; + ch->min_tx_rate = vsi->mqprio_qopt.min_rate[i]; + + /* convert to Kbits/s */ + if (ch->max_tx_rate) + ch->max_tx_rate = div_u64(ch->max_tx_rate, + ICE_BW_KBPS_DIVISOR); + if (ch->min_tx_rate) + ch->min_tx_rate = div_u64(ch->min_tx_rate, + ICE_BW_KBPS_DIVISOR); + + ret = ice_create_q_channel(vsi, ch); + if (ret) { + dev_err(ice_pf_to_dev(pf), + "failed creating channel TC:%d\n", i); + kfree(ch); + goto err_free; + } + list_add_tail(&ch->list, &vsi->ch_list); + vsi->tc_map_vsi[i] = ch->ch_vsi; + dev_dbg(ice_pf_to_dev(pf), + "successfully created channel: VSI %pK\n", ch->ch_vsi); + } + return 0; + +err_free: + ice_remove_q_channels(vsi, false); + + return ret; +} + +/** + * ice_setup_tc_mqprio_qdisc - configure multiple traffic classes + * @netdev: net device to configure + * @type_data: TC offload data + */ +static int ice_setup_tc_mqprio_qdisc(struct net_device *netdev, void *type_data) +{ + struct tc_mqprio_qopt_offload *mqprio_qopt = type_data; + struct ice_netdev_priv *np = netdev_priv(netdev); + struct ice_vsi *vsi = np->vsi; + struct ice_pf *pf = vsi->back; + u16 mode, ena_tc_qdisc = 0; + int cur_txq, cur_rxq; + u8 hw = 0, num_tcf; + struct device *dev; + int ret, i; + + dev = ice_pf_to_dev(pf); + num_tcf = mqprio_qopt->qopt.num_tc; + hw = mqprio_qopt->qopt.hw; + mode = mqprio_qopt->mode; + if (!hw) { + clear_bit(ICE_FLAG_TC_MQPRIO, pf->flags); + vsi->ch_rss_size = 0; + memcpy(&vsi->mqprio_qopt, mqprio_qopt, sizeof(*mqprio_qopt)); + goto config_tcf; + } + + /* Generate queue region map for number of TCF requested */ + for (i = 0; i < num_tcf; i++) + ena_tc_qdisc |= BIT(i); + + switch (mode) { + case TC_MQPRIO_MODE_CHANNEL: + + ret = ice_validate_mqprio_qopt(vsi, mqprio_qopt); + if (ret) { + netdev_err(netdev, "failed to validate_mqprio_qopt(), ret %d\n", + ret); + return ret; + } + memcpy(&vsi->mqprio_qopt, mqprio_qopt, sizeof(*mqprio_qopt)); + set_bit(ICE_FLAG_TC_MQPRIO, pf->flags); + /* don't assume state of hw_tc_offload during driver load + * and set the flag for TC flower filter if hw_tc_offload + * already ON + */ + if (vsi->netdev->features & NETIF_F_HW_TC) + set_bit(ICE_FLAG_CLS_FLOWER, pf->flags); + break; + default: + return -EINVAL; + } + +config_tcf: + + /* Requesting same TCF configuration as already enabled */ + if (ena_tc_qdisc == vsi->tc_cfg.ena_tc && + mode != TC_MQPRIO_MODE_CHANNEL) + return 0; + + /* Pause VSI queues */ + ice_dis_vsi(vsi, true); + + if (!hw && !test_bit(ICE_FLAG_TC_MQPRIO, pf->flags)) + ice_remove_q_channels(vsi, true); + + if (!hw && !test_bit(ICE_FLAG_TC_MQPRIO, pf->flags)) { + vsi->req_txq = min_t(int, ice_get_avail_txq_count(pf), + num_online_cpus()); + vsi->req_rxq = min_t(int, ice_get_avail_rxq_count(pf), + num_online_cpus()); + } else { + /* logic to rebuild VSI, same like ethtool -L */ + u16 offset = 0, qcount_tx = 0, qcount_rx = 0; + + for (i = 0; i < num_tcf; i++) { + if (!(ena_tc_qdisc & BIT(i))) + continue; + + offset = vsi->mqprio_qopt.qopt.offset[i]; + qcount_rx = vsi->mqprio_qopt.qopt.count[i]; + qcount_tx = vsi->mqprio_qopt.qopt.count[i]; + } + vsi->req_txq = offset + qcount_tx; + vsi->req_rxq = offset + qcount_rx; + + /* store away original rss_size info, so that it gets reused + * form ice_vsi_rebuild during tc-qdisc delete stage - to + * determine, what should be the rss_sizefor main VSI + */ + vsi->orig_rss_size = vsi->rss_size; + } + + /* save current values of Tx and Rx queues before calling VSI rebuild + * for fallback option + */ + cur_txq = vsi->num_txq; + cur_rxq = vsi->num_rxq; + + /* proceed with rebuild main VSI using correct number of queues */ + ret = ice_vsi_rebuild(vsi, false); + if (ret) { + /* fallback to current number of queues */ + dev_info(dev, "Rebuild failed with new queues, try with current number of queues\n"); + vsi->req_txq = cur_txq; + vsi->req_rxq = cur_rxq; + clear_bit(ICE_RESET_FAILED, pf->state); + if (ice_vsi_rebuild(vsi, false)) { + dev_err(dev, "Rebuild of main VSI failed again\n"); + return ret; + } + } + + vsi->all_numtc = num_tcf; + vsi->all_enatc = ena_tc_qdisc; + ret = ice_vsi_cfg_tc(vsi, ena_tc_qdisc); + if (ret) { + netdev_err(netdev, "failed configuring TC for VSI id=%d\n", + vsi->vsi_num); + goto exit; + } + + if (test_bit(ICE_FLAG_TC_MQPRIO, pf->flags)) { + u64 max_tx_rate = vsi->mqprio_qopt.max_rate[0]; + u64 min_tx_rate = vsi->mqprio_qopt.min_rate[0]; + + /* set TC0 rate limit if specified */ + if (max_tx_rate || min_tx_rate) { + /* convert to Kbits/s */ + if (max_tx_rate) + max_tx_rate = div_u64(max_tx_rate, ICE_BW_KBPS_DIVISOR); + if (min_tx_rate) + min_tx_rate = div_u64(min_tx_rate, ICE_BW_KBPS_DIVISOR); + + ret = ice_set_bw_limit(vsi, max_tx_rate, min_tx_rate); + if (!ret) { + dev_dbg(dev, "set Tx rate max %llu min %llu for VSI(%u)\n", + max_tx_rate, min_tx_rate, vsi->vsi_num); + } else { + dev_err(dev, "failed to set Tx rate max %llu min %llu for VSI(%u)\n", + max_tx_rate, min_tx_rate, vsi->vsi_num); + goto exit; + } + } + ret = ice_create_q_channels(vsi); + if (ret) { + netdev_err(netdev, "failed configuring queue channels\n"); + goto exit; + } else { + netdev_dbg(netdev, "successfully configured channels\n"); + } + } + + if (vsi->ch_rss_size) + ice_vsi_cfg_rss_lut_key(vsi); + +exit: + /* if error, reset the all_numtc and all_enatc */ + if (ret) { + vsi->all_numtc = 0; + vsi->all_enatc = 0; + } + /* resume VSI */ + ice_ena_vsi(vsi, true); + + return ret; +} + +static LIST_HEAD(ice_block_cb_list); + +static int +ice_setup_tc(struct net_device *netdev, enum tc_setup_type type, + void *type_data) +{ + struct ice_netdev_priv *np = netdev_priv(netdev); + struct ice_pf *pf = np->vsi->back; + int err; + + switch (type) { + case TC_SETUP_BLOCK: + return flow_block_cb_setup_simple(type_data, + &ice_block_cb_list, + ice_setup_tc_block_cb, + np, np, true); + case TC_SETUP_QDISC_MQPRIO: + /* setup traffic classifier for receive side */ + mutex_lock(&pf->tc_mutex); + err = ice_setup_tc_mqprio_qdisc(netdev, type_data); + mutex_unlock(&pf->tc_mutex); + return err; + default: + return -EOPNOTSUPP; + } + return -EOPNOTSUPP; +} + +static struct ice_indr_block_priv * +ice_indr_block_priv_lookup(struct ice_netdev_priv *np, + struct net_device *netdev) +{ + struct ice_indr_block_priv *cb_priv; + + list_for_each_entry(cb_priv, &np->tc_indr_block_priv_list, list) { + if (!cb_priv->netdev) + return NULL; + if (cb_priv->netdev == netdev) + return cb_priv; + } + return NULL; +} + +static int +ice_indr_setup_block_cb(enum tc_setup_type type, void *type_data, + void *indr_priv) +{ + struct ice_indr_block_priv *priv = indr_priv; + struct ice_netdev_priv *np = priv->np; + + switch (type) { + case TC_SETUP_CLSFLOWER: + return ice_setup_tc_cls_flower(np, priv->netdev, + (struct flow_cls_offload *) + type_data); + default: + return -EOPNOTSUPP; + } +} + +static int +ice_indr_setup_tc_block(struct net_device *netdev, struct Qdisc *sch, + struct ice_netdev_priv *np, + struct flow_block_offload *f, void *data, + void (*cleanup)(struct flow_block_cb *block_cb)) +{ + struct ice_indr_block_priv *indr_priv; + struct flow_block_cb *block_cb; + + if (!ice_is_tunnel_supported(netdev) && + !(is_vlan_dev(netdev) && + vlan_dev_real_dev(netdev) == np->vsi->netdev)) + return -EOPNOTSUPP; + + if (f->binder_type != FLOW_BLOCK_BINDER_TYPE_CLSACT_INGRESS) + return -EOPNOTSUPP; + + switch (f->command) { + case FLOW_BLOCK_BIND: + indr_priv = ice_indr_block_priv_lookup(np, netdev); + if (indr_priv) + return -EEXIST; + + indr_priv = kzalloc(sizeof(*indr_priv), GFP_KERNEL); + if (!indr_priv) + return -ENOMEM; + + indr_priv->netdev = netdev; + indr_priv->np = np; + list_add(&indr_priv->list, &np->tc_indr_block_priv_list); + + block_cb = + flow_indr_block_cb_alloc(ice_indr_setup_block_cb, + indr_priv, indr_priv, + ice_rep_indr_tc_block_unbind, + f, netdev, sch, data, np, + cleanup); + + if (IS_ERR(block_cb)) { + list_del(&indr_priv->list); + kfree(indr_priv); + return PTR_ERR(block_cb); + } + flow_block_cb_add(block_cb, f); + list_add_tail(&block_cb->driver_list, &ice_block_cb_list); + break; + case FLOW_BLOCK_UNBIND: + indr_priv = ice_indr_block_priv_lookup(np, netdev); + if (!indr_priv) + return -ENOENT; + + block_cb = flow_block_cb_lookup(f->block, + ice_indr_setup_block_cb, + indr_priv); + if (!block_cb) + return -ENOENT; + + flow_indr_block_cb_remove(block_cb, f); + + list_del(&block_cb->driver_list); + break; + default: + return -EOPNOTSUPP; + } + return 0; +} + +static int +ice_indr_setup_tc_cb(struct net_device *netdev, struct Qdisc *sch, + void *cb_priv, enum tc_setup_type type, void *type_data, + void *data, + void (*cleanup)(struct flow_block_cb *block_cb)) +{ + switch (type) { + case TC_SETUP_BLOCK: + return ice_indr_setup_tc_block(netdev, sch, cb_priv, type_data, + data, cleanup); + + default: + return -EOPNOTSUPP; + } +} + +/** * ice_open - Called when a network interface becomes active * @netdev: network interface device structure * @@ -6800,7 +8912,6 @@ int ice_open_internal(struct net_device *netdev) struct ice_vsi *vsi = np->vsi; struct ice_pf *pf = vsi->back; struct ice_port_info *pi; - enum ice_status status; int err; if (test_bit(ICE_NEEDS_RESTART, pf->state)) { @@ -6811,13 +8922,14 @@ int ice_open_internal(struct net_device *netdev) netif_carrier_off(netdev); pi = vsi->port_info; - status = ice_update_link_info(pi); - if (status) { - netdev_err(netdev, "Failed to get link info, error %s\n", - ice_stat_str(status)); - return -EIO; + err = ice_update_link_info(pi); + if (err) { + netdev_err(netdev, "Failed to get link info, error %d\n", err); + return err; } + ice_check_link_cfg_err(pf, pi->phy.link_info.link_cfg_err); + /* Set PHY if there is media, otherwise, turn off PHY */ if (pi->phy.link_info.link_info & ICE_AQ_MEDIA_AVAILABLE) { clear_bit(ICE_FLAG_NO_MEDIA, pf->flags); @@ -6873,6 +8985,16 @@ int ice_stop(struct net_device *netdev) return -EBUSY; } + if (test_bit(ICE_FLAG_LINK_DOWN_ON_CLOSE_ENA, vsi->back->flags)) { + int link_err = ice_force_phys_link_state(vsi, false); + + if (link_err) { + netdev_err(vsi->netdev, "Failed to set physical link down, VSI %d error %d\n", + vsi->vsi_num, link_err); + return -EIO; + } + } + ice_vsi_close(vsi); return 0; @@ -6889,6 +9011,7 @@ ice_features_check(struct sk_buff *skb, struct net_device __always_unused *netdev, netdev_features_t features) { + bool gso = skb_is_gso(skb); size_t len; /* No point in doing any of this if neither checksum nor GSO are @@ -6901,24 +9024,32 @@ ice_features_check(struct sk_buff *skb, /* We cannot support GSO if the MSS is going to be less than * 64 bytes. If it is then we need to drop support for GSO. */ - if (skb_is_gso(skb) && (skb_shinfo(skb)->gso_size < 64)) + if (gso && (skb_shinfo(skb)->gso_size < ICE_TXD_CTX_MIN_MSS)) features &= ~NETIF_F_GSO_MASK; - len = skb_network_header(skb) - skb->data; + len = skb_network_offset(skb); if (len > ICE_TXD_MACLEN_MAX || len & 0x1) goto out_rm_features; - len = skb_transport_header(skb) - skb_network_header(skb); + len = skb_network_header_len(skb); if (len > ICE_TXD_IPLEN_MAX || len & 0x1) goto out_rm_features; if (skb->encapsulation) { - len = skb_inner_network_header(skb) - skb_transport_header(skb); - if (len > ICE_TXD_L4LEN_MAX || len & 0x1) - goto out_rm_features; + /* this must work for VXLAN frames AND IPIP/SIT frames, and in + * the case of IPIP frames, the transport header pointer is + * after the inner header! So check to make sure that this + * is a GRE or UDP_TUNNEL frame before doing that math. + */ + if (gso && (skb_shinfo(skb)->gso_type & + (SKB_GSO_GRE | SKB_GSO_UDP_TUNNEL))) { + len = skb_inner_network_header(skb) - + skb_transport_header(skb); + if (len > ICE_TXD_L4LEN_MAX || len & 0x1) + goto out_rm_features; + } - len = skb_inner_transport_header(skb) - - skb_inner_network_header(skb); + len = skb_inner_network_header_len(skb); if (len > ICE_TXD_IPLEN_MAX || len & 0x1) goto out_rm_features; } @@ -6937,19 +9068,23 @@ static const struct net_device_ops ice_netdev_safe_mode_ops = { .ndo_change_mtu = ice_change_mtu, .ndo_get_stats64 = ice_get_stats64, .ndo_tx_timeout = ice_tx_timeout, + .ndo_bpf = ice_xdp_safe_mode, }; static const struct net_device_ops ice_netdev_ops = { .ndo_open = ice_open, .ndo_stop = ice_stop, .ndo_start_xmit = ice_start_xmit, + .ndo_select_queue = ice_select_queue, .ndo_features_check = ice_features_check, + .ndo_fix_features = ice_fix_features, .ndo_set_rx_mode = ice_set_rx_mode, .ndo_set_mac_address = ice_set_mac_address, .ndo_validate_addr = eth_validate_addr, .ndo_change_mtu = ice_change_mtu, .ndo_get_stats64 = ice_get_stats64, .ndo_set_tx_maxrate = ice_set_tx_maxrate, + .ndo_eth_ioctl = ice_eth_ioctl, .ndo_set_vf_spoofchk = ice_set_vf_spoofchk, .ndo_set_vf_mac = ice_set_vf_mac, .ndo_get_vf_config = ice_get_vf_cfg, @@ -6957,8 +9092,10 @@ static const struct net_device_ops ice_netdev_ops = { .ndo_set_vf_vlan = ice_set_vf_port_vlan, .ndo_set_vf_link_state = ice_set_vf_link_state, .ndo_get_vf_stats = ice_get_vf_stats, + .ndo_set_vf_rate = ice_set_vf_bw, .ndo_vlan_rx_add_vid = ice_vlan_rx_add_vid, .ndo_vlan_rx_kill_vid = ice_vlan_rx_kill_vid, + .ndo_setup_tc = ice_setup_tc, .ndo_set_features = ice_set_features, .ndo_bridge_getlink = ice_bridge_getlink, .ndo_bridge_setlink = ice_bridge_setlink, @@ -6971,4 +9108,5 @@ static const struct net_device_ops ice_netdev_ops = { .ndo_bpf = ice_xdp, .ndo_xdp_xmit = ice_xdp_xmit, .ndo_xsk_wakeup = ice_xsk_wakeup, + .ndo_get_devlink_port = ice_get_devlink_port, }; |