diff options
Diffstat (limited to 'drivers/net/ethernet/intel')
23 files changed, 512 insertions, 240 deletions
diff --git a/drivers/net/ethernet/intel/i40e/i40e_client.c b/drivers/net/ethernet/intel/i40e/i40e_client.c index ea2bb0140a6e..10d7a982a5b9 100644 --- a/drivers/net/ethernet/intel/i40e/i40e_client.c +++ b/drivers/net/ethernet/intel/i40e/i40e_client.c @@ -177,6 +177,10 @@ void i40e_notify_client_of_netdev_close(struct i40e_vsi *vsi, bool reset) "Cannot locate client instance close routine\n"); return; } + if (!test_bit(__I40E_CLIENT_INSTANCE_OPENED, &cdev->state)) { + dev_dbg(&pf->pdev->dev, "Client is not open, abort close\n"); + return; + } cdev->client->ops->close(&cdev->lan_info, cdev->client, reset); clear_bit(__I40E_CLIENT_INSTANCE_OPENED, &cdev->state); i40e_client_release_qvlist(&cdev->lan_info); @@ -429,7 +433,6 @@ void i40e_client_subtask(struct i40e_pf *pf) /* Remove failed client instance */ clear_bit(__I40E_CLIENT_INSTANCE_OPENED, &cdev->state); - i40e_client_del_instance(pf); return; } } diff --git a/drivers/net/ethernet/intel/i40e/i40e_ethtool.c b/drivers/net/ethernet/intel/i40e/i40e_ethtool.c index 156e92c43780..e9cd0fa6a0d2 100644 --- a/drivers/net/ethernet/intel/i40e/i40e_ethtool.c +++ b/drivers/net/ethernet/intel/i40e/i40e_ethtool.c @@ -4485,7 +4485,7 @@ static int i40e_check_fdir_input_set(struct i40e_vsi *vsi, (struct in6_addr *)&ipv6_full_mask)) new_mask |= I40E_L3_V6_DST_MASK; else if (ipv6_addr_any((struct in6_addr *) - &usr_ip6_spec->ip6src)) + &usr_ip6_spec->ip6dst)) new_mask &= ~I40E_L3_V6_DST_MASK; else return -EOPNOTSUPP; diff --git a/drivers/net/ethernet/intel/i40e/i40e_main.c b/drivers/net/ethernet/intel/i40e/i40e_main.c index b36bf9c3e1e4..e3d9804aeb25 100644 --- a/drivers/net/ethernet/intel/i40e/i40e_main.c +++ b/drivers/net/ethernet/intel/i40e/i40e_main.c @@ -384,7 +384,9 @@ static void i40e_tx_timeout(struct net_device *netdev, unsigned int txqueue) set_bit(__I40E_GLOBAL_RESET_REQUESTED, pf->state); break; default: - netdev_err(netdev, "tx_timeout recovery unsuccessful\n"); + netdev_err(netdev, "tx_timeout recovery unsuccessful, device is in non-recoverable state.\n"); + set_bit(__I40E_DOWN_REQUESTED, pf->state); + set_bit(__I40E_VSI_DOWN_REQUESTED, vsi->state); break; } @@ -5907,6 +5909,26 @@ static int i40e_get_link_speed(struct i40e_vsi *vsi) } /** + * i40e_bw_bytes_to_mbits - Convert max_tx_rate from bytes to mbits + * @vsi: Pointer to vsi structure + * @max_tx_rate: max TX rate in bytes to be converted into Mbits + * + * Helper function to convert units before send to set BW limit + **/ +static u64 i40e_bw_bytes_to_mbits(struct i40e_vsi *vsi, u64 max_tx_rate) +{ + if (max_tx_rate < I40E_BW_MBPS_DIVISOR) { + dev_warn(&vsi->back->pdev->dev, + "Setting max tx rate to minimum usable value of 50Mbps.\n"); + max_tx_rate = I40E_BW_CREDIT_DIVISOR; + } else { + do_div(max_tx_rate, I40E_BW_MBPS_DIVISOR); + } + + return max_tx_rate; +} + +/** * i40e_set_bw_limit - setup BW limit for Tx traffic based on max_tx_rate * @vsi: VSI to be configured * @seid: seid of the channel/VSI @@ -5928,10 +5950,10 @@ int i40e_set_bw_limit(struct i40e_vsi *vsi, u16 seid, u64 max_tx_rate) max_tx_rate, seid); return -EINVAL; } - if (max_tx_rate && max_tx_rate < 50) { + if (max_tx_rate && max_tx_rate < I40E_BW_CREDIT_DIVISOR) { dev_warn(&pf->pdev->dev, "Setting max tx rate to minimum usable value of 50Mbps.\n"); - max_tx_rate = 50; + max_tx_rate = I40E_BW_CREDIT_DIVISOR; } /* Tx rate credits are in values of 50Mbps, 0 is disabled */ @@ -6657,6 +6679,9 @@ static int i40e_configure_queue_channels(struct i40e_vsi *vsi) vsi->tc_seid_map[i] = ch->seid; } } + + /* reset to reconfigure TX queue contexts */ + i40e_do_reset(vsi->back, I40E_PF_RESET_FLAG, true); return ret; err_free: @@ -8219,9 +8244,9 @@ config_tc: if (i40e_is_tc_mqprio_enabled(pf)) { if (vsi->mqprio_qopt.max_rate[0]) { - u64 max_tx_rate = vsi->mqprio_qopt.max_rate[0]; + u64 max_tx_rate = i40e_bw_bytes_to_mbits(vsi, + vsi->mqprio_qopt.max_rate[0]); - do_div(max_tx_rate, I40E_BW_MBPS_DIVISOR); ret = i40e_set_bw_limit(vsi, vsi->seid, max_tx_rate); if (!ret) { u64 credits = max_tx_rate; @@ -10966,10 +10991,10 @@ static void i40e_rebuild(struct i40e_pf *pf, bool reinit, bool lock_acquired) } if (vsi->mqprio_qopt.max_rate[0]) { - u64 max_tx_rate = vsi->mqprio_qopt.max_rate[0]; + u64 max_tx_rate = i40e_bw_bytes_to_mbits(vsi, + vsi->mqprio_qopt.max_rate[0]); u64 credits = 0; - do_div(max_tx_rate, I40E_BW_MBPS_DIVISOR); ret = i40e_set_bw_limit(vsi, vsi->seid, max_tx_rate); if (ret) goto end_unlock; diff --git a/drivers/net/ethernet/intel/i40e/i40e_txrx.c b/drivers/net/ethernet/intel/i40e/i40e_txrx.c index f6ba97a0166e..69e67eb6aea7 100644 --- a/drivers/net/ethernet/intel/i40e/i40e_txrx.c +++ b/drivers/net/ethernet/intel/i40e/i40e_txrx.c @@ -3203,11 +3203,13 @@ static int i40e_tx_enable_csum(struct sk_buff *skb, u32 *tx_flags, protocol = vlan_get_protocol(skb); - if (eth_p_mpls(protocol)) + if (eth_p_mpls(protocol)) { ip.hdr = skb_inner_network_header(skb); - else + l4.hdr = skb_checksum_start(skb); + } else { ip.hdr = skb_network_header(skb); - l4.hdr = skb_checksum_start(skb); + l4.hdr = skb_transport_header(skb); + } /* set the tx_flags to indicate the IP protocol type. this is * required so that checksum header computation below is accurate. @@ -3686,7 +3688,8 @@ u16 i40e_lan_select_queue(struct net_device *netdev, u8 prio; /* is DCB enabled at all? */ - if (vsi->tc_config.numtc == 1) + if (vsi->tc_config.numtc == 1 || + i40e_is_tc_mqprio_enabled(vsi->back)) return netdev_pick_tx(netdev, skb, sb_dev); prio = skb->priority; diff --git a/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c b/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c index 4f184c50f6e8..7e9f6a69eb10 100644 --- a/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c +++ b/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c @@ -2039,6 +2039,25 @@ static void i40e_del_qch(struct i40e_vf *vf) } /** + * i40e_vc_get_max_frame_size + * @vf: pointer to the VF + * + * Max frame size is determined based on the current port's max frame size and + * whether a port VLAN is configured on this VF. The VF is not aware whether + * it's in a port VLAN so the PF needs to account for this in max frame size + * checks and sending the max frame size to the VF. + **/ +static u16 i40e_vc_get_max_frame_size(struct i40e_vf *vf) +{ + u16 max_frame_size = vf->pf->hw.phy.link_info.max_frame_size; + + if (vf->port_vlan_id) + max_frame_size -= VLAN_HLEN; + + return max_frame_size; +} + +/** * i40e_vc_get_vf_resources_msg * @vf: pointer to the VF info * @msg: pointer to the msg buffer @@ -2139,6 +2158,7 @@ static int i40e_vc_get_vf_resources_msg(struct i40e_vf *vf, u8 *msg) vfres->max_vectors = pf->hw.func_caps.num_msix_vectors_vf; vfres->rss_key_size = I40E_HKEY_ARRAY_SIZE; vfres->rss_lut_size = I40E_VF_HLUT_ARRAY_SIZE; + vfres->max_mtu = i40e_vc_get_max_frame_size(vf); if (vf->lan_vsi_idx) { vfres->vsi_res[0].vsi_id = vf->lan_vsi_id; diff --git a/drivers/net/ethernet/intel/iavf/iavf_adminq.c b/drivers/net/ethernet/intel/iavf/iavf_adminq.c index cd4e6a22d0f9..9ffbd24d83cb 100644 --- a/drivers/net/ethernet/intel/iavf/iavf_adminq.c +++ b/drivers/net/ethernet/intel/iavf/iavf_adminq.c @@ -324,6 +324,7 @@ static enum iavf_status iavf_config_arq_regs(struct iavf_hw *hw) static enum iavf_status iavf_init_asq(struct iavf_hw *hw) { enum iavf_status ret_code = 0; + int i; if (hw->aq.asq.count > 0) { /* queue already initialized */ @@ -354,12 +355,17 @@ static enum iavf_status iavf_init_asq(struct iavf_hw *hw) /* initialize base registers */ ret_code = iavf_config_asq_regs(hw); if (ret_code) - goto init_adminq_free_rings; + goto init_free_asq_bufs; /* success! */ hw->aq.asq.count = hw->aq.num_asq_entries; goto init_adminq_exit; +init_free_asq_bufs: + for (i = 0; i < hw->aq.num_asq_entries; i++) + iavf_free_dma_mem(hw, &hw->aq.asq.r.asq_bi[i]); + iavf_free_virt_mem(hw, &hw->aq.asq.dma_head); + init_adminq_free_rings: iavf_free_adminq_asq(hw); @@ -383,6 +389,7 @@ init_adminq_exit: static enum iavf_status iavf_init_arq(struct iavf_hw *hw) { enum iavf_status ret_code = 0; + int i; if (hw->aq.arq.count > 0) { /* queue already initialized */ @@ -413,12 +420,16 @@ static enum iavf_status iavf_init_arq(struct iavf_hw *hw) /* initialize base registers */ ret_code = iavf_config_arq_regs(hw); if (ret_code) - goto init_adminq_free_rings; + goto init_free_arq_bufs; /* success! */ hw->aq.arq.count = hw->aq.num_arq_entries; goto init_adminq_exit; +init_free_arq_bufs: + for (i = 0; i < hw->aq.num_arq_entries; i++) + iavf_free_dma_mem(hw, &hw->aq.arq.r.arq_bi[i]); + iavf_free_virt_mem(hw, &hw->aq.arq.dma_head); init_adminq_free_rings: iavf_free_adminq_arq(hw); diff --git a/drivers/net/ethernet/intel/iavf/iavf_main.c b/drivers/net/ethernet/intel/iavf/iavf_main.c index 45d097a164ad..0c89f16bf1e2 100644 --- a/drivers/net/ethernet/intel/iavf/iavf_main.c +++ b/drivers/net/ethernet/intel/iavf/iavf_main.c @@ -1077,7 +1077,6 @@ static int iavf_set_mac(struct net_device *netdev, void *p) { struct iavf_adapter *adapter = netdev_priv(netdev); struct sockaddr *addr = p; - bool handle_mac = iavf_is_mac_set_handled(netdev, addr->sa_data); int ret; if (!is_valid_ether_addr(addr->sa_data)) @@ -1094,10 +1093,9 @@ static int iavf_set_mac(struct net_device *netdev, void *p) return 0; } - if (handle_mac) - goto done; - - ret = wait_event_interruptible_timeout(adapter->vc_waitqueue, false, msecs_to_jiffies(2500)); + ret = wait_event_interruptible_timeout(adapter->vc_waitqueue, + iavf_is_mac_set_handled(netdev, addr->sa_data), + msecs_to_jiffies(2500)); /* If ret < 0 then it means wait was interrupted. * If ret == 0 then it means we got a timeout. @@ -1111,7 +1109,6 @@ static int iavf_set_mac(struct net_device *netdev, void *p) if (!ret) return -EAGAIN; -done: if (!ether_addr_equal(netdev->dev_addr, addr->sa_data)) return -EACCES; @@ -2367,7 +2364,7 @@ static void iavf_init_get_resources(struct iavf_adapter *adapter) err = iavf_get_vf_config(adapter); if (err == -EALREADY) { err = iavf_send_vf_config_msg(adapter); - goto err_alloc; + goto err; } else if (err == -EINVAL) { /* We only get -EINVAL if the device is in a very bad * state or if we've been disabled for previous bad @@ -2877,6 +2874,11 @@ static void iavf_reset_task(struct work_struct *work) int i = 0, err; bool running; + /* Detach interface to avoid subsequent NDO callbacks */ + rtnl_lock(); + netif_device_detach(netdev); + rtnl_unlock(); + /* When device is being removed it doesn't make sense to run the reset * task, just return in such a case. */ @@ -2884,7 +2886,7 @@ static void iavf_reset_task(struct work_struct *work) if (adapter->state != __IAVF_REMOVE) queue_work(iavf_wq, &adapter->reset_task); - return; + goto reset_finish; } while (!mutex_trylock(&adapter->client_lock)) @@ -2954,7 +2956,6 @@ continue_reset: if (running) { netif_carrier_off(netdev); - netif_tx_stop_all_queues(netdev); adapter->link_up = false; iavf_napi_disable_all(adapter); } @@ -3084,14 +3085,21 @@ continue_reset: mutex_unlock(&adapter->client_lock); mutex_unlock(&adapter->crit_lock); - return; + goto reset_finish; reset_err: + if (running) { + set_bit(__IAVF_VSI_DOWN, adapter->vsi.state); + iavf_free_traffic_irqs(adapter); + } + iavf_disable_vf(adapter); + mutex_unlock(&adapter->client_lock); mutex_unlock(&adapter->crit_lock); - if (running) - iavf_change_state(adapter, __IAVF_RUNNING); dev_err(&adapter->pdev->dev, "failed to allocate resources during reinit\n"); - iavf_close(netdev); +reset_finish: + rtnl_lock(); + netif_device_attach(netdev); + rtnl_unlock(); } /** @@ -4085,8 +4093,17 @@ static int iavf_open(struct net_device *netdev) return -EIO; } - while (!mutex_trylock(&adapter->crit_lock)) + while (!mutex_trylock(&adapter->crit_lock)) { + /* If we are in __IAVF_INIT_CONFIG_ADAPTER state the crit_lock + * is already taken and iavf_open is called from an upper + * device's notifier reacting on NETDEV_REGISTER event. + * We have to leave here to avoid dead lock. + */ + if (adapter->state == __IAVF_INIT_CONFIG_ADAPTER) + return -EBUSY; + usleep_range(500, 1000); + } if (adapter->state != __IAVF_DOWN) { err = -EBUSY; diff --git a/drivers/net/ethernet/intel/iavf/iavf_txrx.c b/drivers/net/ethernet/intel/iavf/iavf_txrx.c index 06d18797d25a..18b6a702a1d6 100644 --- a/drivers/net/ethernet/intel/iavf/iavf_txrx.c +++ b/drivers/net/ethernet/intel/iavf/iavf_txrx.c @@ -114,8 +114,11 @@ u32 iavf_get_tx_pending(struct iavf_ring *ring, bool in_sw) { u32 head, tail; + /* underlying hardware might not allow access and/or always return + * 0 for the head/tail registers so just use the cached values + */ head = ring->next_to_clean; - tail = readl(ring->tail); + tail = ring->next_to_use; if (head != tail) return (head < tail) ? @@ -1390,7 +1393,7 @@ static struct sk_buff *iavf_build_skb(struct iavf_ring *rx_ring, #endif struct sk_buff *skb; - if (!rx_buffer) + if (!rx_buffer || !size) return NULL; /* prefetch first cache line of first page */ va = page_address(rx_buffer->page) + rx_buffer->page_offset; @@ -1548,7 +1551,7 @@ static int iavf_clean_rx_irq(struct iavf_ring *rx_ring, int budget) /* exit if we failed to retrieve a buffer */ if (!skb) { rx_ring->rx_stats.alloc_buff_failed++; - if (rx_buffer) + if (rx_buffer && size) rx_buffer->pagecnt_bias++; break; } diff --git a/drivers/net/ethernet/intel/iavf/iavf_virtchnl.c b/drivers/net/ethernet/intel/iavf/iavf_virtchnl.c index 15ee85dc33bd..5a9e6563923e 100644 --- a/drivers/net/ethernet/intel/iavf/iavf_virtchnl.c +++ b/drivers/net/ethernet/intel/iavf/iavf_virtchnl.c @@ -269,11 +269,14 @@ int iavf_get_vf_vlan_v2_caps(struct iavf_adapter *adapter) void iavf_configure_queues(struct iavf_adapter *adapter) { struct virtchnl_vsi_queue_config_info *vqci; - struct virtchnl_queue_pair_info *vqpi; + int i, max_frame = adapter->vf_res->max_mtu; int pairs = adapter->num_active_queues; - int i, max_frame = IAVF_MAX_RXBUFFER; + struct virtchnl_queue_pair_info *vqpi; size_t len; + if (max_frame > IAVF_MAX_RXBUFFER || !max_frame) + max_frame = IAVF_MAX_RXBUFFER; + if (adapter->current_op != VIRTCHNL_OP_UNKNOWN) { /* bail because we already have a command pending */ dev_err(&adapter->pdev->dev, "Cannot configure queues, command %d pending\n", diff --git a/drivers/net/ethernet/intel/ice/ice.h b/drivers/net/ethernet/intel/ice/ice.h index cc5b85afd437..841fa149c407 100644 --- a/drivers/net/ethernet/intel/ice/ice.h +++ b/drivers/net/ethernet/intel/ice/ice.h @@ -684,8 +684,8 @@ static inline void ice_set_ring_xdp(struct ice_tx_ring *ring) * ice_xsk_pool - get XSK buffer pool bound to a ring * @ring: Rx ring to use * - * Returns a pointer to xdp_umem structure if there is a buffer pool present, - * NULL otherwise. + * Returns a pointer to xsk_buff_pool structure if there is a buffer pool + * present, NULL otherwise. */ static inline struct xsk_buff_pool *ice_xsk_pool(struct ice_rx_ring *ring) { @@ -699,23 +699,33 @@ static inline struct xsk_buff_pool *ice_xsk_pool(struct ice_rx_ring *ring) } /** - * ice_tx_xsk_pool - get XSK buffer pool bound to a ring - * @ring: Tx ring to use + * ice_tx_xsk_pool - assign XSK buff pool to XDP ring + * @vsi: pointer to VSI + * @qid: index of a queue to look at XSK buff pool presence * - * Returns a pointer to xdp_umem structure if there is a buffer pool present, - * NULL otherwise. Tx equivalent of ice_xsk_pool. + * Sets XSK buff pool pointer on XDP ring. + * + * XDP ring is picked from Rx ring, whereas Rx ring is picked based on provided + * queue id. Reason for doing so is that queue vectors might have assigned more + * than one XDP ring, e.g. when user reduced the queue count on netdev; Rx ring + * carries a pointer to one of these XDP rings for its own purposes, such as + * handling XDP_TX action, therefore we can piggyback here on the + * rx_ring->xdp_ring assignment that was done during XDP rings initialization. */ -static inline struct xsk_buff_pool *ice_tx_xsk_pool(struct ice_tx_ring *ring) +static inline void ice_tx_xsk_pool(struct ice_vsi *vsi, u16 qid) { - struct ice_vsi *vsi = ring->vsi; - u16 qid; + struct ice_tx_ring *ring; - qid = ring->q_index - vsi->alloc_txq; + ring = vsi->rx_rings[qid]->xdp_ring; + if (!ring) + return; - if (!ice_is_xdp_ena_vsi(vsi) || !test_bit(qid, vsi->af_xdp_zc_qps)) - return NULL; + if (!ice_is_xdp_ena_vsi(vsi) || !test_bit(qid, vsi->af_xdp_zc_qps)) { + ring->xsk_pool = NULL; + return; + } - return xsk_get_pool_from_qid(vsi->netdev, qid); + ring->xsk_pool = xsk_get_pool_from_qid(vsi->netdev, qid); } /** diff --git a/drivers/net/ethernet/intel/ice/ice_base.c b/drivers/net/ethernet/intel/ice/ice_base.c index 136d7911adb4..1e3243808178 100644 --- a/drivers/net/ethernet/intel/ice/ice_base.c +++ b/drivers/net/ethernet/intel/ice/ice_base.c @@ -7,18 +7,6 @@ #include "ice_dcb_lib.h" #include "ice_sriov.h" -static bool ice_alloc_rx_buf_zc(struct ice_rx_ring *rx_ring) -{ - rx_ring->xdp_buf = kcalloc(rx_ring->count, sizeof(*rx_ring->xdp_buf), GFP_KERNEL); - return !!rx_ring->xdp_buf; -} - -static bool ice_alloc_rx_buf(struct ice_rx_ring *rx_ring) -{ - rx_ring->rx_buf = kcalloc(rx_ring->count, sizeof(*rx_ring->rx_buf), GFP_KERNEL); - return !!rx_ring->rx_buf; -} - /** * __ice_vsi_get_qs_contig - Assign a contiguous chunk of queues to VSI * @qs_cfg: gathered variables needed for PF->VSI queues assignment @@ -519,11 +507,8 @@ int ice_vsi_cfg_rxq(struct ice_rx_ring *ring) xdp_rxq_info_reg(&ring->xdp_rxq, ring->netdev, ring->q_index, ring->q_vector->napi.napi_id); - kfree(ring->rx_buf); ring->xsk_pool = ice_xsk_pool(ring); if (ring->xsk_pool) { - if (!ice_alloc_rx_buf_zc(ring)) - return -ENOMEM; xdp_rxq_info_unreg_mem_model(&ring->xdp_rxq); ring->rx_buf_len = @@ -538,8 +523,6 @@ int ice_vsi_cfg_rxq(struct ice_rx_ring *ring) dev_info(dev, "Registered XDP mem model MEM_TYPE_XSK_BUFF_POOL on Rx ring %d\n", ring->q_index); } else { - if (!ice_alloc_rx_buf(ring)) - return -ENOMEM; if (!xdp_rxq_info_is_reg(&ring->xdp_rxq)) /* coverity[check_return] */ xdp_rxq_info_reg(&ring->xdp_rxq, diff --git a/drivers/net/ethernet/intel/ice/ice_fltr.c b/drivers/net/ethernet/intel/ice/ice_fltr.c index 85a94483c2ed..40e678cfb507 100644 --- a/drivers/net/ethernet/intel/ice/ice_fltr.c +++ b/drivers/net/ethernet/intel/ice/ice_fltr.c @@ -62,7 +62,7 @@ ice_fltr_set_vlan_vsi_promisc(struct ice_hw *hw, struct ice_vsi *vsi, int result; result = ice_set_vlan_vsi_promisc(hw, vsi->idx, promisc_mask, false); - if (result) + if (result && result != -EEXIST) dev_err(ice_pf_to_dev(pf), "Error setting promisc mode on VSI %i (rc=%d)\n", vsi->vsi_num, result); @@ -86,7 +86,7 @@ ice_fltr_clear_vlan_vsi_promisc(struct ice_hw *hw, struct ice_vsi *vsi, int result; result = ice_set_vlan_vsi_promisc(hw, vsi->idx, promisc_mask, true); - if (result) + if (result && result != -EEXIST) dev_err(ice_pf_to_dev(pf), "Error clearing promisc mode on VSI %i (rc=%d)\n", vsi->vsi_num, result); @@ -109,7 +109,7 @@ ice_fltr_clear_vsi_promisc(struct ice_hw *hw, u16 vsi_handle, u8 promisc_mask, int result; result = ice_clear_vsi_promisc(hw, vsi_handle, promisc_mask, vid); - if (result) + if (result && result != -EEXIST) dev_err(ice_pf_to_dev(pf), "Error clearing promisc mode on VSI %i for VID %u (rc=%d)\n", ice_get_hw_vsi_num(hw, vsi_handle), vid, result); @@ -132,7 +132,7 @@ ice_fltr_set_vsi_promisc(struct ice_hw *hw, u16 vsi_handle, u8 promisc_mask, int result; result = ice_set_vsi_promisc(hw, vsi_handle, promisc_mask, vid); - if (result) + if (result && result != -EEXIST) dev_err(ice_pf_to_dev(pf), "Error setting promisc mode on VSI %i for VID %u (rc=%d)\n", ice_get_hw_vsi_num(hw, vsi_handle), vid, result); diff --git a/drivers/net/ethernet/intel/ice/ice_lib.c b/drivers/net/ethernet/intel/ice/ice_lib.c index a830f7f9aed0..58d483e2f539 100644 --- a/drivers/net/ethernet/intel/ice/ice_lib.c +++ b/drivers/net/ethernet/intel/ice/ice_lib.c @@ -914,7 +914,7 @@ static void ice_set_dflt_vsi_ctx(struct ice_hw *hw, struct ice_vsi_ctx *ctxt) */ static int ice_vsi_setup_q_map(struct ice_vsi *vsi, struct ice_vsi_ctx *ctxt) { - u16 offset = 0, qmap = 0, tx_count = 0, pow = 0; + u16 offset = 0, qmap = 0, tx_count = 0, rx_count = 0, pow = 0; u16 num_txq_per_tc, num_rxq_per_tc; u16 qcount_tx = vsi->alloc_txq; u16 qcount_rx = vsi->alloc_rxq; @@ -981,23 +981,25 @@ static int ice_vsi_setup_q_map(struct ice_vsi *vsi, struct ice_vsi_ctx *ctxt) * at least 1) */ if (offset) - vsi->num_rxq = offset; + rx_count = offset; else - vsi->num_rxq = num_rxq_per_tc; + rx_count = num_rxq_per_tc; - if (vsi->num_rxq > vsi->alloc_rxq) { + if (rx_count > vsi->alloc_rxq) { dev_err(ice_pf_to_dev(vsi->back), "Trying to use more Rx queues (%u), than were allocated (%u)!\n", - vsi->num_rxq, vsi->alloc_rxq); + rx_count, vsi->alloc_rxq); return -EINVAL; } - vsi->num_txq = tx_count; - if (vsi->num_txq > vsi->alloc_txq) { + if (tx_count > vsi->alloc_txq) { dev_err(ice_pf_to_dev(vsi->back), "Trying to use more Tx queues (%u), than were allocated (%u)!\n", - vsi->num_txq, vsi->alloc_txq); + tx_count, vsi->alloc_txq); return -EINVAL; } + vsi->num_txq = tx_count; + vsi->num_rxq = rx_count; + if (vsi->type == ICE_VSI_VF && vsi->num_txq != vsi->num_rxq) { dev_dbg(ice_pf_to_dev(vsi->back), "VF VSI should have same number of Tx and Rx queues. Hence making them equal\n"); /* since there is a chance that num_rxq could have been changed @@ -1986,8 +1988,8 @@ int ice_vsi_cfg_xdp_txqs(struct ice_vsi *vsi) if (ret) return ret; - ice_for_each_xdp_txq(vsi, i) - vsi->xdp_rings[i]->xsk_pool = ice_tx_xsk_pool(vsi->xdp_rings[i]); + ice_for_each_rxq(vsi, i) + ice_tx_xsk_pool(vsi, i); return ret; } @@ -3181,7 +3183,7 @@ int ice_vsi_rebuild(struct ice_vsi *vsi, bool init_vsi) pf = vsi->back; vtype = vsi->type; - if (WARN_ON(vtype == ICE_VSI_VF) && !vsi->vf) + if (WARN_ON(vtype == ICE_VSI_VF && !vsi->vf)) return -EINVAL; ice_vsi_init_vlan_ops(vsi); @@ -3490,6 +3492,7 @@ ice_vsi_setup_q_map_mqprio(struct ice_vsi *vsi, struct ice_vsi_ctx *ctxt, u16 pow, offset = 0, qcount_tx = 0, qcount_rx = 0, qmap; u16 tc0_offset = vsi->mqprio_qopt.qopt.offset[0]; int tc0_qcount = vsi->mqprio_qopt.qopt.count[0]; + u16 new_txq, new_rxq; u8 netdev_tc = 0; int i; @@ -3530,21 +3533,24 @@ ice_vsi_setup_q_map_mqprio(struct ice_vsi *vsi, struct ice_vsi_ctx *ctxt, } } - /* Set actual Tx/Rx queue pairs */ - vsi->num_txq = offset + qcount_tx; - if (vsi->num_txq > vsi->alloc_txq) { + new_txq = offset + qcount_tx; + if (new_txq > vsi->alloc_txq) { dev_err(ice_pf_to_dev(vsi->back), "Trying to use more Tx queues (%u), than were allocated (%u)!\n", - vsi->num_txq, vsi->alloc_txq); + new_txq, vsi->alloc_txq); return -EINVAL; } - vsi->num_rxq = offset + qcount_rx; - if (vsi->num_rxq > vsi->alloc_rxq) { + new_rxq = offset + qcount_rx; + if (new_rxq > vsi->alloc_rxq) { dev_err(ice_pf_to_dev(vsi->back), "Trying to use more Rx queues (%u), than were allocated (%u)!\n", - vsi->num_rxq, vsi->alloc_rxq); + new_rxq, vsi->alloc_rxq); return -EINVAL; } + /* Set actual Tx/Rx queue pairs */ + vsi->num_txq = new_txq; + vsi->num_rxq = new_rxq; + /* Setup queue TC[0].qmap for given VSI context */ ctxt->info.tc_mapping[0] = cpu_to_le16(qmap); ctxt->info.q_mapping[0] = cpu_to_le16(vsi->rxq_map[0]); @@ -3576,6 +3582,7 @@ int ice_vsi_cfg_tc(struct ice_vsi *vsi, u8 ena_tc) { u16 max_txqs[ICE_MAX_TRAFFIC_CLASS] = { 0 }; struct ice_pf *pf = vsi->back; + struct ice_tc_cfg old_tc_cfg; struct ice_vsi_ctx *ctx; struct device *dev; int i, ret = 0; @@ -3600,6 +3607,7 @@ int ice_vsi_cfg_tc(struct ice_vsi *vsi, u8 ena_tc) max_txqs[i] = vsi->num_txq; } + memcpy(&old_tc_cfg, &vsi->tc_cfg, sizeof(old_tc_cfg)); vsi->tc_cfg.ena_tc = ena_tc; vsi->tc_cfg.numtc = num_tc; @@ -3616,8 +3624,10 @@ int ice_vsi_cfg_tc(struct ice_vsi *vsi, u8 ena_tc) else ret = ice_vsi_setup_q_map(vsi, ctx); - if (ret) + if (ret) { + memcpy(&vsi->tc_cfg, &old_tc_cfg, sizeof(vsi->tc_cfg)); goto out; + } /* must to indicate which section of VSI context are being modified */ ctx->info.valid_sections = cpu_to_le16(ICE_AQ_VSI_PROP_RXQ_MAP_VALID); @@ -4062,7 +4072,11 @@ int ice_vsi_del_vlan_zero(struct ice_vsi *vsi) if (err && err != -EEXIST) return err; - return 0; + /* when deleting the last VLAN filter, make sure to disable the VLAN + * promisc mode so the filter isn't left by accident + */ + return ice_clear_vsi_promisc(&vsi->back->hw, vsi->idx, + ICE_MCAST_VLAN_PROMISC_BITS, 0); } /** diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c index eb40526ee179..e109cb93886b 100644 --- a/drivers/net/ethernet/intel/ice/ice_main.c +++ b/drivers/net/ethernet/intel/ice/ice_main.c @@ -267,8 +267,10 @@ static int ice_set_promisc(struct ice_vsi *vsi, u8 promisc_m) status = ice_fltr_set_vsi_promisc(&vsi->back->hw, vsi->idx, promisc_m, 0); } + if (status && status != -EEXIST) + return status; - return status; + return 0; } /** @@ -2397,8 +2399,6 @@ int ice_schedule_reset(struct ice_pf *pf, enum ice_reset_req reset) return -EBUSY; } - ice_unplug_aux_dev(pf); - switch (reset) { case ICE_RESET_PFR: set_bit(ICE_PFR_REQ, pf->state); @@ -2579,7 +2579,6 @@ static int ice_xdp_alloc_setup_rings(struct ice_vsi *vsi) if (ice_setup_tx_ring(xdp_ring)) goto free_xdp_rings; ice_set_ring_xdp(xdp_ring); - xdp_ring->xsk_pool = ice_tx_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); @@ -2587,13 +2586,6 @@ static int ice_xdp_alloc_setup_rings(struct ice_vsi *vsi) } } - 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 - vsi->rx_rings[i]->xdp_ring = vsi->xdp_rings[i]; - } - return 0; free_xdp_rings: @@ -2683,6 +2675,23 @@ int ice_prepare_xdp_rings(struct ice_vsi *vsi, struct bpf_prog *prog) 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 @@ -2887,10 +2896,18 @@ ice_xdp_setup_prog(struct ice_vsi *vsi, struct bpf_prog *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 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 @@ -3573,6 +3590,14 @@ ice_vlan_rx_kill_vid(struct net_device *netdev, __be16 proto, u16 vid) 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 @@ -3886,7 +3911,7 @@ 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; } @@ -6624,7 +6649,7 @@ static void ice_napi_disable_all(struct ice_vsi *vsi) */ int ice_down(struct ice_vsi *vsi) { - int i, tx_err, rx_err, link_err = 0, vlan_err = 0; + int i, tx_err, rx_err, vlan_err = 0; WARN_ON(!test_bit(ICE_VSI_DOWN, vsi->state)); @@ -6658,20 +6683,13 @@ 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 || vlan_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; @@ -6833,6 +6851,8 @@ int ice_vsi_open(struct ice_vsi *vsi) if (err) goto err_setup_rx; + ice_vsi_cfg_netdev_tc(vsi, vsi->tc_cfg.ena_tc); + 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); @@ -8865,6 +8885,16 @@ 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; diff --git a/drivers/net/ethernet/intel/ice/ice_switch.c b/drivers/net/ethernet/intel/ice/ice_switch.c index 262e553e3b58..3808034f7e7e 100644 --- a/drivers/net/ethernet/intel/ice/ice_switch.c +++ b/drivers/net/ethernet/intel/ice/ice_switch.c @@ -4445,6 +4445,13 @@ ice_set_vlan_vsi_promisc(struct ice_hw *hw, u16 vsi_handle, u8 promisc_mask, goto free_fltr_list; list_for_each_entry(list_itr, &vsi_list_head, list_entry) { + /* Avoid enabling or disabling VLAN zero twice when in double + * VLAN mode + */ + if (ice_is_dvm_ena(hw) && + list_itr->fltr_info.l_data.vlan.tpid == 0) + continue; + vlan_id = list_itr->fltr_info.l_data.vlan.vlan_id; if (rm_vlan_promisc) status = ice_clear_vsi_promisc(hw, vsi_handle, @@ -4452,7 +4459,7 @@ ice_set_vlan_vsi_promisc(struct ice_hw *hw, u16 vsi_handle, u8 promisc_mask, else status = ice_set_vsi_promisc(hw, vsi_handle, promisc_mask, vlan_id); - if (status) + if (status && status != -EEXIST) break; } diff --git a/drivers/net/ethernet/intel/ice/ice_txrx.c b/drivers/net/ethernet/intel/ice/ice_txrx.c index 836dce840712..dd2285d4bef4 100644 --- a/drivers/net/ethernet/intel/ice/ice_txrx.c +++ b/drivers/net/ethernet/intel/ice/ice_txrx.c @@ -610,7 +610,7 @@ ice_xdp_xmit(struct net_device *dev, int n, struct xdp_frame **frames, if (test_bit(ICE_VSI_DOWN, vsi->state)) return -ENETDOWN; - if (!ice_is_xdp_ena_vsi(vsi) || queue_index >= vsi->num_xdp_txq) + if (!ice_is_xdp_ena_vsi(vsi)) return -ENXIO; if (unlikely(flags & ~XDP_XMIT_FLAGS_MASK)) @@ -621,6 +621,9 @@ ice_xdp_xmit(struct net_device *dev, int n, struct xdp_frame **frames, xdp_ring = vsi->xdp_rings[queue_index]; spin_lock(&xdp_ring->tx_lock); } else { + /* Generally, should not happen */ + if (unlikely(queue_index >= vsi->num_xdp_txq)) + return -ENXIO; xdp_ring = vsi->xdp_rings[queue_index]; } @@ -1464,7 +1467,7 @@ int ice_napi_poll(struct napi_struct *napi, int budget) bool wd; if (tx_ring->xsk_pool) - wd = ice_xmit_zc(tx_ring, ICE_DESC_UNUSED(tx_ring), budget); + wd = ice_xmit_zc(tx_ring); else if (ice_ring_is_xdp(tx_ring)) wd = true; else diff --git a/drivers/net/ethernet/intel/ice/ice_vf_lib.c b/drivers/net/ethernet/intel/ice/ice_vf_lib.c index 8fd7c3e37f5e..0abeed092de1 100644 --- a/drivers/net/ethernet/intel/ice/ice_vf_lib.c +++ b/drivers/net/ethernet/intel/ice/ice_vf_lib.c @@ -571,8 +571,10 @@ int ice_reset_vf(struct ice_vf *vf, u32 flags) if (ice_is_vf_disabled(vf)) { vsi = ice_get_vf_vsi(vf); - if (WARN_ON(!vsi)) + if (!vsi) { + dev_dbg(dev, "VF is already removed\n"); return -EINVAL; + } ice_vsi_stop_lan_tx_rings(vsi, ICE_NO_RESET, vf->vf_id); ice_vsi_stop_all_rx_rings(vsi); dev_dbg(dev, "VF is already disabled, there is no need for resetting it, telling VM, all is fine %d\n", @@ -762,13 +764,16 @@ static int ice_cfg_mac_antispoof(struct ice_vsi *vsi, bool enable) static int ice_vsi_ena_spoofchk(struct ice_vsi *vsi) { struct ice_vsi_vlan_ops *vlan_ops; - int err; + int err = 0; vlan_ops = ice_get_compat_vsi_vlan_ops(vsi); - err = vlan_ops->ena_tx_filtering(vsi); - if (err) - return err; + /* Allow VF with VLAN 0 only to send all tagged traffic */ + if (vsi->type != ICE_VSI_VF || ice_vsi_has_non_zero_vlans(vsi)) { + err = vlan_ops->ena_tx_filtering(vsi); + if (err) + return err; + } return ice_cfg_mac_antispoof(vsi, true); } diff --git a/drivers/net/ethernet/intel/ice/ice_virtchnl.c b/drivers/net/ethernet/intel/ice/ice_virtchnl.c index 094e3c97a1ea..2b4c791b6cba 100644 --- a/drivers/net/ethernet/intel/ice/ice_virtchnl.c +++ b/drivers/net/ethernet/intel/ice/ice_virtchnl.c @@ -2288,6 +2288,15 @@ static int ice_vc_process_vlan_msg(struct ice_vf *vf, u8 *msg, bool add_v) /* Enable VLAN filtering on first non-zero VLAN */ if (!vlan_promisc && vid && !ice_is_dvm_ena(&pf->hw)) { + if (vf->spoofchk) { + status = vsi->inner_vlan_ops.ena_tx_filtering(vsi); + if (status) { + v_ret = VIRTCHNL_STATUS_ERR_PARAM; + dev_err(dev, "Enable VLAN anti-spoofing on VLAN ID: %d failed error-%d\n", + vid, status); + goto error_param; + } + } if (vsi->inner_vlan_ops.ena_rx_filtering(vsi)) { v_ret = VIRTCHNL_STATUS_ERR_PARAM; dev_err(dev, "Enable VLAN pruning on VLAN ID: %d failed error-%d\n", @@ -2333,8 +2342,10 @@ static int ice_vc_process_vlan_msg(struct ice_vf *vf, u8 *msg, bool add_v) } /* Disable VLAN filtering when only VLAN 0 is left */ - if (!ice_vsi_has_non_zero_vlans(vsi)) + if (!ice_vsi_has_non_zero_vlans(vsi)) { + vsi->inner_vlan_ops.dis_tx_filtering(vsi); vsi->inner_vlan_ops.dis_rx_filtering(vsi); + } if (vlan_promisc) ice_vf_dis_vlan_promisc(vsi, &vlan); @@ -2838,6 +2849,13 @@ ice_vc_del_vlans(struct ice_vf *vf, struct ice_vsi *vsi, if (vlan_promisc) ice_vf_dis_vlan_promisc(vsi, &vlan); + + /* Disable VLAN filtering when only VLAN 0 is left */ + if (!ice_vsi_has_non_zero_vlans(vsi) && ice_is_dvm_ena(&vsi->back->hw)) { + err = vsi->outer_vlan_ops.dis_tx_filtering(vsi); + if (err) + return err; + } } vc_vlan = &vlan_fltr->inner; @@ -2853,8 +2871,17 @@ ice_vc_del_vlans(struct ice_vf *vf, struct ice_vsi *vsi, /* no support for VLAN promiscuous on inner VLAN unless * we are in Single VLAN Mode (SVM) */ - if (!ice_is_dvm_ena(&vsi->back->hw) && vlan_promisc) - ice_vf_dis_vlan_promisc(vsi, &vlan); + if (!ice_is_dvm_ena(&vsi->back->hw)) { + if (vlan_promisc) + ice_vf_dis_vlan_promisc(vsi, &vlan); + + /* Disable VLAN filtering when only VLAN 0 is left */ + if (!ice_vsi_has_non_zero_vlans(vsi)) { + err = vsi->inner_vlan_ops.dis_tx_filtering(vsi); + if (err) + return err; + } + } } } @@ -2931,6 +2958,13 @@ ice_vc_add_vlans(struct ice_vf *vf, struct ice_vsi *vsi, if (err) return err; } + + /* Enable VLAN filtering on first non-zero VLAN */ + if (vf->spoofchk && vlan.vid && ice_is_dvm_ena(&vsi->back->hw)) { + err = vsi->outer_vlan_ops.ena_tx_filtering(vsi); + if (err) + return err; + } } vc_vlan = &vlan_fltr->inner; @@ -2946,10 +2980,19 @@ ice_vc_add_vlans(struct ice_vf *vf, struct ice_vsi *vsi, /* no support for VLAN promiscuous on inner VLAN unless * we are in Single VLAN Mode (SVM) */ - if (!ice_is_dvm_ena(&vsi->back->hw) && vlan_promisc) { - err = ice_vf_ena_vlan_promisc(vsi, &vlan); - if (err) - return err; + if (!ice_is_dvm_ena(&vsi->back->hw)) { + if (vlan_promisc) { + err = ice_vf_ena_vlan_promisc(vsi, &vlan); + if (err) + return err; + } + + /* Enable VLAN filtering on first non-zero VLAN */ + if (vf->spoofchk && vlan.vid) { + err = vsi->inner_vlan_ops.ena_tx_filtering(vsi); + if (err) + return err; + } } } } diff --git a/drivers/net/ethernet/intel/ice/ice_xsk.c b/drivers/net/ethernet/intel/ice/ice_xsk.c index 49ba8bfdbf04..056c904b83cc 100644 --- a/drivers/net/ethernet/intel/ice/ice_xsk.c +++ b/drivers/net/ethernet/intel/ice/ice_xsk.c @@ -192,6 +192,7 @@ static int ice_qp_dis(struct ice_vsi *vsi, u16 q_idx) err = ice_vsi_ctrl_one_rx_ring(vsi, false, q_idx, true); if (err) return err; + ice_clean_rx_ring(rx_ring); ice_qvec_toggle_napi(vsi, q_vector, false); ice_qp_clean_rings(vsi, q_idx); @@ -243,7 +244,7 @@ static int ice_qp_ena(struct ice_vsi *vsi, u16 q_idx) if (err) goto free_buf; ice_set_ring_xdp(xdp_ring); - xdp_ring->xsk_pool = ice_tx_xsk_pool(xdp_ring); + ice_tx_xsk_pool(vsi, q_idx); } err = ice_vsi_cfg_rxq(rx_ring); @@ -317,6 +318,62 @@ ice_xsk_pool_enable(struct ice_vsi *vsi, struct xsk_buff_pool *pool, u16 qid) } /** + * ice_realloc_rx_xdp_bufs - reallocate for either XSK or normal buffer + * @rx_ring: Rx ring + * @pool_present: is pool for XSK present + * + * Try allocating memory and return ENOMEM, if failed to allocate. + * If allocation was successful, substitute buffer with allocated one. + * Returns 0 on success, negative on failure + */ +static int +ice_realloc_rx_xdp_bufs(struct ice_rx_ring *rx_ring, bool pool_present) +{ + size_t elem_size = pool_present ? sizeof(*rx_ring->xdp_buf) : + sizeof(*rx_ring->rx_buf); + void *sw_ring = kcalloc(rx_ring->count, elem_size, GFP_KERNEL); + + if (!sw_ring) + return -ENOMEM; + + if (pool_present) { + kfree(rx_ring->rx_buf); + rx_ring->rx_buf = NULL; + rx_ring->xdp_buf = sw_ring; + } else { + kfree(rx_ring->xdp_buf); + rx_ring->xdp_buf = NULL; + rx_ring->rx_buf = sw_ring; + } + + return 0; +} + +/** + * ice_realloc_zc_buf - reallocate XDP ZC queue pairs + * @vsi: Current VSI + * @zc: is zero copy set + * + * Reallocate buffer for rx_rings that might be used by XSK. + * XDP requires more memory, than rx_buf provides. + * Returns 0 on success, negative on failure + */ +int ice_realloc_zc_buf(struct ice_vsi *vsi, bool zc) +{ + struct ice_rx_ring *rx_ring; + unsigned long q; + + for_each_set_bit(q, vsi->af_xdp_zc_qps, + max_t(int, vsi->alloc_txq, vsi->alloc_rxq)) { + rx_ring = vsi->rx_rings[q]; + if (ice_realloc_rx_xdp_bufs(rx_ring, zc)) + return -ENOMEM; + } + + return 0; +} + +/** * ice_xsk_pool_setup - enable/disable a buffer pool region depending on its state * @vsi: Current VSI * @pool: buffer pool to enable/associate to a ring, NULL to disable @@ -329,9 +386,8 @@ int ice_xsk_pool_setup(struct ice_vsi *vsi, struct xsk_buff_pool *pool, u16 qid) bool if_running, pool_present = !!pool; int ret = 0, pool_failure = 0; - if (!is_power_of_2(vsi->rx_rings[qid]->count) || - !is_power_of_2(vsi->tx_rings[qid]->count)) { - netdev_err(vsi->netdev, "Please align ring sizes to power of 2\n"); + if (qid >= vsi->num_rxq || qid >= vsi->num_txq) { + netdev_err(vsi->netdev, "Please use queue id in scope of combined queues count\n"); pool_failure = -EINVAL; goto failure; } @@ -339,11 +395,17 @@ int ice_xsk_pool_setup(struct ice_vsi *vsi, struct xsk_buff_pool *pool, u16 qid) if_running = netif_running(vsi->netdev) && ice_is_xdp_ena_vsi(vsi); if (if_running) { + struct ice_rx_ring *rx_ring = vsi->rx_rings[qid]; + ret = ice_qp_dis(vsi, qid); if (ret) { netdev_err(vsi->netdev, "ice_qp_dis error = %d\n", ret); goto xsk_pool_if_up; } + + ret = ice_realloc_rx_xdp_bufs(rx_ring, pool_present); + if (ret) + goto xsk_pool_if_up; } pool_failure = pool_present ? ice_xsk_pool_enable(vsi, pool, qid) : @@ -353,7 +415,7 @@ xsk_pool_if_up: if (if_running) { ret = ice_qp_ena(vsi, qid); if (!ret && pool_present) - napi_schedule(&vsi->xdp_rings[qid]->q_vector->napi); + napi_schedule(&vsi->rx_rings[qid]->xdp_ring->q_vector->napi); else if (ret) netdev_err(vsi->netdev, "ice_qp_ena error = %d\n", ret); } @@ -465,11 +527,10 @@ exit: bool ice_alloc_rx_bufs_zc(struct ice_rx_ring *rx_ring, u16 count) { u16 rx_thresh = ICE_RING_QUARTER(rx_ring); - u16 batched, leftover, i, tail_bumps; + u16 leftover, i, tail_bumps; - batched = ALIGN_DOWN(count, rx_thresh); - tail_bumps = batched / rx_thresh; - leftover = count & (rx_thresh - 1); + tail_bumps = count / rx_thresh; + leftover = count - (tail_bumps * rx_thresh); for (i = 0; i < tail_bumps; i++) if (!__ice_alloc_rx_bufs_zc(rx_ring, rx_thresh)) @@ -719,69 +780,57 @@ ice_clean_xdp_tx_buf(struct ice_tx_ring *xdp_ring, struct ice_tx_buf *tx_buf) } /** - * ice_clean_xdp_irq_zc - Reclaim resources after transmit completes on XDP ring - * @xdp_ring: XDP ring to clean - * @napi_budget: amount of descriptors that NAPI allows us to clean - * - * Returns count of cleaned descriptors + * ice_clean_xdp_irq_zc - produce AF_XDP descriptors to CQ + * @xdp_ring: XDP Tx ring */ -static u16 ice_clean_xdp_irq_zc(struct ice_tx_ring *xdp_ring, int napi_budget) +static void ice_clean_xdp_irq_zc(struct ice_tx_ring *xdp_ring) { - u16 tx_thresh = ICE_RING_QUARTER(xdp_ring); - int budget = napi_budget / tx_thresh; - u16 next_dd = xdp_ring->next_dd; - u16 ntc, cleared_dds = 0; - - do { - struct ice_tx_desc *next_dd_desc; - u16 desc_cnt = xdp_ring->count; - struct ice_tx_buf *tx_buf; - u32 xsk_frames; - u16 i; - - next_dd_desc = ICE_TX_DESC(xdp_ring, next_dd); - if (!(next_dd_desc->cmd_type_offset_bsz & - cpu_to_le64(ICE_TX_DESC_DTYPE_DESC_DONE))) - break; + u16 ntc = xdp_ring->next_to_clean; + struct ice_tx_desc *tx_desc; + u16 cnt = xdp_ring->count; + struct ice_tx_buf *tx_buf; + u16 xsk_frames = 0; + u16 last_rs; + int i; - cleared_dds++; - xsk_frames = 0; - if (likely(!xdp_ring->xdp_tx_active)) { - xsk_frames = tx_thresh; - goto skip; - } + last_rs = xdp_ring->next_to_use ? xdp_ring->next_to_use - 1 : cnt - 1; + tx_desc = ICE_TX_DESC(xdp_ring, last_rs); + if ((tx_desc->cmd_type_offset_bsz & + cpu_to_le64(ICE_TX_DESC_DTYPE_DESC_DONE))) { + if (last_rs >= ntc) + xsk_frames = last_rs - ntc + 1; + else + xsk_frames = last_rs + cnt - ntc + 1; + } - ntc = xdp_ring->next_to_clean; + if (!xsk_frames) + return; - for (i = 0; i < tx_thresh; i++) { - tx_buf = &xdp_ring->tx_buf[ntc]; + if (likely(!xdp_ring->xdp_tx_active)) + goto skip; - if (tx_buf->raw_buf) { - ice_clean_xdp_tx_buf(xdp_ring, tx_buf); - tx_buf->raw_buf = NULL; - } else { - xsk_frames++; - } + ntc = xdp_ring->next_to_clean; + for (i = 0; i < xsk_frames; i++) { + tx_buf = &xdp_ring->tx_buf[ntc]; - ntc++; - if (ntc >= xdp_ring->count) - ntc = 0; + if (tx_buf->raw_buf) { + ice_clean_xdp_tx_buf(xdp_ring, tx_buf); + tx_buf->raw_buf = NULL; + } else { + xsk_frames++; } + + ntc++; + if (ntc >= xdp_ring->count) + ntc = 0; + } skip: - xdp_ring->next_to_clean += tx_thresh; - if (xdp_ring->next_to_clean >= desc_cnt) - xdp_ring->next_to_clean -= desc_cnt; - if (xsk_frames) - xsk_tx_completed(xdp_ring->xsk_pool, xsk_frames); - next_dd_desc->cmd_type_offset_bsz = 0; - next_dd = next_dd + tx_thresh; - if (next_dd >= desc_cnt) - next_dd = tx_thresh - 1; - } while (--budget); - - xdp_ring->next_dd = next_dd; - - return cleared_dds * tx_thresh; + tx_desc->cmd_type_offset_bsz = 0; + xdp_ring->next_to_clean += xsk_frames; + if (xdp_ring->next_to_clean >= cnt) + xdp_ring->next_to_clean -= cnt; + if (xsk_frames) + xsk_tx_completed(xdp_ring->xsk_pool, xsk_frames); } /** @@ -816,7 +865,6 @@ static void ice_xmit_pkt(struct ice_tx_ring *xdp_ring, struct xdp_desc *desc, static void ice_xmit_pkt_batch(struct ice_tx_ring *xdp_ring, struct xdp_desc *descs, unsigned int *total_bytes) { - u16 tx_thresh = ICE_RING_QUARTER(xdp_ring); u16 ntu = xdp_ring->next_to_use; struct ice_tx_desc *tx_desc; u32 i; @@ -836,13 +884,6 @@ static void ice_xmit_pkt_batch(struct ice_tx_ring *xdp_ring, struct xdp_desc *de } xdp_ring->next_to_use = ntu; - - if (xdp_ring->next_to_use > xdp_ring->next_rs) { - tx_desc = ICE_TX_DESC(xdp_ring, xdp_ring->next_rs); - tx_desc->cmd_type_offset_bsz |= - cpu_to_le64(ICE_TX_DESC_CMD_RS << ICE_TXD_QW1_CMD_S); - xdp_ring->next_rs += tx_thresh; - } } /** @@ -855,7 +896,6 @@ static void ice_xmit_pkt_batch(struct ice_tx_ring *xdp_ring, struct xdp_desc *de static void ice_fill_tx_hw_ring(struct ice_tx_ring *xdp_ring, struct xdp_desc *descs, u32 nb_pkts, unsigned int *total_bytes) { - u16 tx_thresh = ICE_RING_QUARTER(xdp_ring); u32 batched, leftover, i; batched = ALIGN_DOWN(nb_pkts, PKTS_PER_BATCH); @@ -864,54 +904,54 @@ static void ice_fill_tx_hw_ring(struct ice_tx_ring *xdp_ring, struct xdp_desc *d ice_xmit_pkt_batch(xdp_ring, &descs[i], total_bytes); for (; i < batched + leftover; i++) ice_xmit_pkt(xdp_ring, &descs[i], total_bytes); +} - if (xdp_ring->next_to_use > xdp_ring->next_rs) { - struct ice_tx_desc *tx_desc; +/** + * ice_set_rs_bit - set RS bit on last produced descriptor (one behind current NTU) + * @xdp_ring: XDP ring to produce the HW Tx descriptors on + */ +static void ice_set_rs_bit(struct ice_tx_ring *xdp_ring) +{ + u16 ntu = xdp_ring->next_to_use ? xdp_ring->next_to_use - 1 : xdp_ring->count - 1; + struct ice_tx_desc *tx_desc; - tx_desc = ICE_TX_DESC(xdp_ring, xdp_ring->next_rs); - tx_desc->cmd_type_offset_bsz |= - cpu_to_le64(ICE_TX_DESC_CMD_RS << ICE_TXD_QW1_CMD_S); - xdp_ring->next_rs += tx_thresh; - } + tx_desc = ICE_TX_DESC(xdp_ring, ntu); + tx_desc->cmd_type_offset_bsz |= + cpu_to_le64(ICE_TX_DESC_CMD_RS << ICE_TXD_QW1_CMD_S); } /** * ice_xmit_zc - take entries from XSK Tx ring and place them onto HW Tx ring * @xdp_ring: XDP ring to produce the HW Tx descriptors on - * @budget: number of free descriptors on HW Tx ring that can be used - * @napi_budget: amount of descriptors that NAPI allows us to clean * * Returns true if there is no more work that needs to be done, false otherwise */ -bool ice_xmit_zc(struct ice_tx_ring *xdp_ring, u32 budget, int napi_budget) +bool ice_xmit_zc(struct ice_tx_ring *xdp_ring) { struct xdp_desc *descs = xdp_ring->xsk_pool->tx_descs; - u16 tx_thresh = ICE_RING_QUARTER(xdp_ring); u32 nb_pkts, nb_processed = 0; unsigned int total_bytes = 0; + int budget; + + ice_clean_xdp_irq_zc(xdp_ring); - if (budget < tx_thresh) - budget += ice_clean_xdp_irq_zc(xdp_ring, napi_budget); + budget = ICE_DESC_UNUSED(xdp_ring); + budget = min_t(u16, budget, ICE_RING_QUARTER(xdp_ring)); nb_pkts = xsk_tx_peek_release_desc_batch(xdp_ring->xsk_pool, budget); if (!nb_pkts) return true; if (xdp_ring->next_to_use + nb_pkts >= xdp_ring->count) { - struct ice_tx_desc *tx_desc; - nb_processed = xdp_ring->count - xdp_ring->next_to_use; ice_fill_tx_hw_ring(xdp_ring, descs, nb_processed, &total_bytes); - tx_desc = ICE_TX_DESC(xdp_ring, xdp_ring->next_rs); - tx_desc->cmd_type_offset_bsz |= - cpu_to_le64(ICE_TX_DESC_CMD_RS << ICE_TXD_QW1_CMD_S); - xdp_ring->next_rs = tx_thresh - 1; xdp_ring->next_to_use = 0; } ice_fill_tx_hw_ring(xdp_ring, &descs[nb_processed], nb_pkts - nb_processed, &total_bytes); + ice_set_rs_bit(xdp_ring); ice_xdp_ring_update_tail(xdp_ring); ice_update_tx_ring_stats(xdp_ring, nb_pkts, total_bytes); @@ -944,13 +984,13 @@ ice_xsk_wakeup(struct net_device *netdev, u32 queue_id, if (!ice_is_xdp_ena_vsi(vsi)) return -EINVAL; - if (queue_id >= vsi->num_txq) + if (queue_id >= vsi->num_txq || queue_id >= vsi->num_rxq) return -EINVAL; - if (!vsi->xdp_rings[queue_id]->xsk_pool) - return -EINVAL; + ring = vsi->rx_rings[queue_id]->xdp_ring; - ring = vsi->xdp_rings[queue_id]; + if (!ring->xsk_pool) + return -EINVAL; /* The idea here is that if NAPI is running, mark a miss, so * it will run again. If not, trigger an interrupt and @@ -989,14 +1029,16 @@ bool ice_xsk_any_rx_ring_ena(struct ice_vsi *vsi) */ void ice_xsk_clean_rx_ring(struct ice_rx_ring *rx_ring) { - u16 count_mask = rx_ring->count - 1; u16 ntc = rx_ring->next_to_clean; u16 ntu = rx_ring->next_to_use; - for ( ; ntc != ntu; ntc = (ntc + 1) & count_mask) { + while (ntc != ntu) { struct xdp_buff *xdp = *ice_xdp_buf(rx_ring, ntc); xsk_buff_free(xdp); + ntc++; + if (ntc >= rx_ring->count) + ntc = 0; } } diff --git a/drivers/net/ethernet/intel/ice/ice_xsk.h b/drivers/net/ethernet/intel/ice/ice_xsk.h index 21faec8e97db..6fa181f080ef 100644 --- a/drivers/net/ethernet/intel/ice/ice_xsk.h +++ b/drivers/net/ethernet/intel/ice/ice_xsk.h @@ -26,12 +26,10 @@ bool ice_alloc_rx_bufs_zc(struct ice_rx_ring *rx_ring, u16 count); bool ice_xsk_any_rx_ring_ena(struct ice_vsi *vsi); void ice_xsk_clean_rx_ring(struct ice_rx_ring *rx_ring); void ice_xsk_clean_xdp_ring(struct ice_tx_ring *xdp_ring); -bool ice_xmit_zc(struct ice_tx_ring *xdp_ring, u32 budget, int napi_budget); +bool ice_xmit_zc(struct ice_tx_ring *xdp_ring); +int ice_realloc_zc_buf(struct ice_vsi *vsi, bool zc); #else -static inline bool -ice_xmit_zc(struct ice_tx_ring __always_unused *xdp_ring, - u32 __always_unused budget, - int __always_unused napi_budget) +static inline bool ice_xmit_zc(struct ice_tx_ring __always_unused *xdp_ring) { return false; } @@ -72,5 +70,12 @@ ice_xsk_wakeup(struct net_device __always_unused *netdev, static inline void ice_xsk_clean_rx_ring(struct ice_rx_ring *rx_ring) { } static inline void ice_xsk_clean_xdp_ring(struct ice_tx_ring *xdp_ring) { } + +static inline int +ice_realloc_zc_buf(struct ice_vsi __always_unused *vsi, + bool __always_unused zc) +{ + return 0; +} #endif /* CONFIG_XDP_SOCKETS */ #endif /* !_ICE_XSK_H_ */ diff --git a/drivers/net/ethernet/intel/igb/igb.h b/drivers/net/ethernet/intel/igb/igb.h index 2d3daf022651..015b78144114 100644 --- a/drivers/net/ethernet/intel/igb/igb.h +++ b/drivers/net/ethernet/intel/igb/igb.h @@ -664,6 +664,8 @@ struct igb_adapter { struct igb_mac_addr *mac_table; struct vf_mac_filter vf_macs; struct vf_mac_filter *vf_mac_list; + /* lock for VF resources */ + spinlock_t vfs_lock; }; /* flags controlling PTP/1588 function */ diff --git a/drivers/net/ethernet/intel/igb/igb_main.c b/drivers/net/ethernet/intel/igb/igb_main.c index d8b836a85cc3..2796e81d2726 100644 --- a/drivers/net/ethernet/intel/igb/igb_main.c +++ b/drivers/net/ethernet/intel/igb/igb_main.c @@ -3637,6 +3637,7 @@ static int igb_disable_sriov(struct pci_dev *pdev) struct net_device *netdev = pci_get_drvdata(pdev); struct igb_adapter *adapter = netdev_priv(netdev); struct e1000_hw *hw = &adapter->hw; + unsigned long flags; /* reclaim resources allocated to VFs */ if (adapter->vf_data) { @@ -3649,12 +3650,13 @@ static int igb_disable_sriov(struct pci_dev *pdev) pci_disable_sriov(pdev); msleep(500); } - + spin_lock_irqsave(&adapter->vfs_lock, flags); kfree(adapter->vf_mac_list); adapter->vf_mac_list = NULL; kfree(adapter->vf_data); adapter->vf_data = NULL; adapter->vfs_allocated_count = 0; + spin_unlock_irqrestore(&adapter->vfs_lock, flags); wr32(E1000_IOVCTL, E1000_IOVCTL_REUSE_VFQ); wrfl(); msleep(100); @@ -3814,7 +3816,9 @@ static void igb_remove(struct pci_dev *pdev) igb_release_hw_control(adapter); #ifdef CONFIG_PCI_IOV + rtnl_lock(); igb_disable_sriov(pdev); + rtnl_unlock(); #endif unregister_netdev(netdev); @@ -3974,6 +3978,9 @@ static int igb_sw_init(struct igb_adapter *adapter) spin_lock_init(&adapter->nfc_lock); spin_lock_init(&adapter->stats64_lock); + + /* init spinlock to avoid concurrency of VF resources */ + spin_lock_init(&adapter->vfs_lock); #ifdef CONFIG_PCI_IOV switch (hw->mac.type) { case e1000_82576: @@ -7958,8 +7965,10 @@ unlock: static void igb_msg_task(struct igb_adapter *adapter) { struct e1000_hw *hw = &adapter->hw; + unsigned long flags; u32 vf; + spin_lock_irqsave(&adapter->vfs_lock, flags); for (vf = 0; vf < adapter->vfs_allocated_count; vf++) { /* process any reset requests */ if (!igb_check_for_rst(hw, vf)) @@ -7973,6 +7982,7 @@ static void igb_msg_task(struct igb_adapter *adapter) if (!igb_check_for_ack(hw, vf)) igb_rcv_ack_from_vf(adapter, vf); } + spin_unlock_irqrestore(&adapter->vfs_lock, flags); } /** diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe_ptp.c b/drivers/net/ethernet/intel/ixgbe/ixgbe_ptp.c index 9f06896a049b..f8605f57bd06 100644 --- a/drivers/net/ethernet/intel/ixgbe/ixgbe_ptp.c +++ b/drivers/net/ethernet/intel/ixgbe/ixgbe_ptp.c @@ -1214,7 +1214,6 @@ void ixgbe_ptp_start_cyclecounter(struct ixgbe_adapter *adapter) struct cyclecounter cc; unsigned long flags; u32 incval = 0; - u32 tsauxc = 0; u32 fuse0 = 0; /* For some of the boards below this mask is technically incorrect. @@ -1249,18 +1248,6 @@ void ixgbe_ptp_start_cyclecounter(struct ixgbe_adapter *adapter) case ixgbe_mac_x550em_a: case ixgbe_mac_X550: cc.read = ixgbe_ptp_read_X550; - - /* enable SYSTIME counter */ - IXGBE_WRITE_REG(hw, IXGBE_SYSTIMR, 0); - IXGBE_WRITE_REG(hw, IXGBE_SYSTIML, 0); - IXGBE_WRITE_REG(hw, IXGBE_SYSTIMH, 0); - tsauxc = IXGBE_READ_REG(hw, IXGBE_TSAUXC); - IXGBE_WRITE_REG(hw, IXGBE_TSAUXC, - tsauxc & ~IXGBE_TSAUXC_DISABLE_SYSTIME); - IXGBE_WRITE_REG(hw, IXGBE_TSIM, IXGBE_TSIM_TXTS); - IXGBE_WRITE_REG(hw, IXGBE_EIMS, IXGBE_EIMS_TIMESYNC); - - IXGBE_WRITE_FLUSH(hw); break; case ixgbe_mac_X540: cc.read = ixgbe_ptp_read_82599; @@ -1293,6 +1280,50 @@ void ixgbe_ptp_start_cyclecounter(struct ixgbe_adapter *adapter) } /** + * ixgbe_ptp_init_systime - Initialize SYSTIME registers + * @adapter: the ixgbe private board structure + * + * Initialize and start the SYSTIME registers. + */ +static void ixgbe_ptp_init_systime(struct ixgbe_adapter *adapter) +{ + struct ixgbe_hw *hw = &adapter->hw; + u32 tsauxc; + + switch (hw->mac.type) { + case ixgbe_mac_X550EM_x: + case ixgbe_mac_x550em_a: + case ixgbe_mac_X550: + tsauxc = IXGBE_READ_REG(hw, IXGBE_TSAUXC); + + /* Reset SYSTIME registers to 0 */ + IXGBE_WRITE_REG(hw, IXGBE_SYSTIMR, 0); + IXGBE_WRITE_REG(hw, IXGBE_SYSTIML, 0); + IXGBE_WRITE_REG(hw, IXGBE_SYSTIMH, 0); + + /* Reset interrupt settings */ + IXGBE_WRITE_REG(hw, IXGBE_TSIM, IXGBE_TSIM_TXTS); + IXGBE_WRITE_REG(hw, IXGBE_EIMS, IXGBE_EIMS_TIMESYNC); + + /* Activate the SYSTIME counter */ + IXGBE_WRITE_REG(hw, IXGBE_TSAUXC, + tsauxc & ~IXGBE_TSAUXC_DISABLE_SYSTIME); + break; + case ixgbe_mac_X540: + case ixgbe_mac_82599EB: + /* Reset SYSTIME registers to 0 */ + IXGBE_WRITE_REG(hw, IXGBE_SYSTIML, 0); + IXGBE_WRITE_REG(hw, IXGBE_SYSTIMH, 0); + break; + default: + /* Other devices aren't supported */ + return; + }; + + IXGBE_WRITE_FLUSH(hw); +} + +/** * ixgbe_ptp_reset * @adapter: the ixgbe private board structure * @@ -1318,6 +1349,8 @@ void ixgbe_ptp_reset(struct ixgbe_adapter *adapter) ixgbe_ptp_start_cyclecounter(adapter); + ixgbe_ptp_init_systime(adapter); + spin_lock_irqsave(&adapter->tmreg_lock, flags); timecounter_init(&adapter->hw_tc, &adapter->hw_cc, ktime_to_ns(ktime_get_real())); |