diff options
Diffstat (limited to 'drivers/net/ethernet/mellanox')
52 files changed, 4058 insertions, 1561 deletions
diff --git a/drivers/net/ethernet/mellanox/mlx4/en_dcb_nl.c b/drivers/net/ethernet/mellanox/mlx4/en_dcb_nl.c index 5f41dc9..1a0c3bf8 100644 --- a/drivers/net/ethernet/mellanox/mlx4/en_dcb_nl.c +++ b/drivers/net/ethernet/mellanox/mlx4/en_dcb_nl.c @@ -310,6 +310,7 @@ static int mlx4_en_ets_validate(struct mlx4_en_priv *priv, struct ieee_ets *ets) } switch (ets->tc_tsa[i]) { + case IEEE_8021QAZ_TSA_VENDOR: case IEEE_8021QAZ_TSA_STRICT: break; case IEEE_8021QAZ_TSA_ETS: @@ -347,6 +348,10 @@ static int mlx4_en_config_port_scheduler(struct mlx4_en_priv *priv, /* higher TC means higher priority => lower pg */ for (i = IEEE_8021QAZ_MAX_TCS - 1; i >= 0; i--) { switch (ets->tc_tsa[i]) { + case IEEE_8021QAZ_TSA_VENDOR: + pg[i] = MLX4_EN_TC_VENDOR; + tc_tx_bw[i] = MLX4_EN_BW_MAX; + break; case IEEE_8021QAZ_TSA_STRICT: pg[i] = num_strict++; tc_tx_bw[i] = MLX4_EN_BW_MAX; diff --git a/drivers/net/ethernet/mellanox/mlx4/en_ethtool.c b/drivers/net/ethernet/mellanox/mlx4/en_ethtool.c index bf1f041..ebc1f56 100644 --- a/drivers/net/ethernet/mellanox/mlx4/en_ethtool.c +++ b/drivers/net/ethernet/mellanox/mlx4/en_ethtool.c @@ -1094,12 +1094,21 @@ static int mlx4_en_set_ringparam(struct net_device *dev, if (param->rx_jumbo_pending || param->rx_mini_pending) return -EINVAL; + if (param->rx_pending < MLX4_EN_MIN_RX_SIZE) { + en_warn(priv, "%s: rx_pending (%d) < min (%d)\n", + __func__, param->rx_pending, + MLX4_EN_MIN_RX_SIZE); + return -EINVAL; + } + if (param->tx_pending < MLX4_EN_MIN_TX_SIZE) { + en_warn(priv, "%s: tx_pending (%d) < min (%lu)\n", + __func__, param->tx_pending, + MLX4_EN_MIN_TX_SIZE); + return -EINVAL; + } + rx_size = roundup_pow_of_two(param->rx_pending); - rx_size = max_t(u32, rx_size, MLX4_EN_MIN_RX_SIZE); - rx_size = min_t(u32, rx_size, MLX4_EN_MAX_RX_SIZE); tx_size = roundup_pow_of_two(param->tx_pending); - tx_size = max_t(u32, tx_size, MLX4_EN_MIN_TX_SIZE); - tx_size = min_t(u32, tx_size, MLX4_EN_MAX_TX_SIZE); if (rx_size == (priv->port_up ? priv->rx_ring[0]->actual_size : priv->rx_ring[0]->size) && diff --git a/drivers/net/ethernet/mellanox/mlx4/en_netdev.c b/drivers/net/ethernet/mellanox/mlx4/en_netdev.c index 99051a2..8fc51bc 100644 --- a/drivers/net/ethernet/mellanox/mlx4/en_netdev.c +++ b/drivers/net/ethernet/mellanox/mlx4/en_netdev.c @@ -2172,8 +2172,9 @@ static int mlx4_en_alloc_resources(struct mlx4_en_priv *priv) if (mlx4_en_create_rx_ring(priv, &priv->rx_ring[i], prof->rx_ring_size, priv->stride, - node)) + node, i)) goto err; + } #ifdef CONFIG_RFS_ACCEL @@ -3336,6 +3337,13 @@ int mlx4_en_init_netdev(struct mlx4_en_dev *mdev, int port, priv->msg_enable = MLX4_EN_MSG_LEVEL; #ifdef CONFIG_MLX4_EN_DCB if (!mlx4_is_slave(priv->mdev->dev)) { + u8 prio; + + for (prio = 0; prio < IEEE_8021QAZ_MAX_TCS; ++prio) { + priv->ets.prio_tc[prio] = prio; + priv->ets.tc_tsa[prio] = IEEE_8021QAZ_TSA_VENDOR; + } + priv->dcbx_cap = DCB_CAP_DCBX_VER_CEE | DCB_CAP_DCBX_HOST | DCB_CAP_DCBX_VER_IEEE; priv->flags |= MLX4_EN_DCB_ENABLED; diff --git a/drivers/net/ethernet/mellanox/mlx4/en_rx.c b/drivers/net/ethernet/mellanox/mlx4/en_rx.c index 85e28ef..b4d144e 100644 --- a/drivers/net/ethernet/mellanox/mlx4/en_rx.c +++ b/drivers/net/ethernet/mellanox/mlx4/en_rx.c @@ -262,7 +262,7 @@ void mlx4_en_set_num_rx_rings(struct mlx4_en_dev *mdev) int mlx4_en_create_rx_ring(struct mlx4_en_priv *priv, struct mlx4_en_rx_ring **pring, - u32 size, u16 stride, int node) + u32 size, u16 stride, int node, int queue_index) { struct mlx4_en_dev *mdev = priv->mdev; struct mlx4_en_rx_ring *ring; @@ -286,6 +286,9 @@ int mlx4_en_create_rx_ring(struct mlx4_en_priv *priv, ring->log_stride = ffs(ring->stride) - 1; ring->buf_size = ring->size * ring->stride + TXBB_SIZE; + if (xdp_rxq_info_reg(&ring->xdp_rxq, priv->dev, queue_index) < 0) + goto err_ring; + tmp = size * roundup_pow_of_two(MLX4_EN_MAX_RX_FRAGS * sizeof(struct mlx4_en_rx_alloc)); ring->rx_info = vzalloc_node(tmp, node); @@ -293,7 +296,7 @@ int mlx4_en_create_rx_ring(struct mlx4_en_priv *priv, ring->rx_info = vzalloc(tmp); if (!ring->rx_info) { err = -ENOMEM; - goto err_ring; + goto err_xdp_info; } } @@ -317,6 +320,8 @@ int mlx4_en_create_rx_ring(struct mlx4_en_priv *priv, err_info: vfree(ring->rx_info); ring->rx_info = NULL; +err_xdp_info: + xdp_rxq_info_unreg(&ring->xdp_rxq); err_ring: kfree(ring); *pring = NULL; @@ -440,6 +445,7 @@ void mlx4_en_destroy_rx_ring(struct mlx4_en_priv *priv, lockdep_is_held(&mdev->state_lock)); if (old_prog) bpf_prog_put(old_prog); + xdp_rxq_info_unreg(&ring->xdp_rxq); mlx4_free_hwq_res(mdev->dev, &ring->wqres, size * stride + TXBB_SIZE); vfree(ring->rx_info); ring->rx_info = NULL; @@ -617,6 +623,10 @@ static int get_fixed_ipv6_csum(__wsum hw_checksum, struct sk_buff *skb, return 0; } #endif + +/* We reach this function only after checking that any of + * the (IPv4 | IPv6) bits are set in cqe->status. + */ static int check_csum(struct mlx4_cqe *cqe, struct sk_buff *skb, void *va, netdev_features_t dev_features) { @@ -632,13 +642,11 @@ static int check_csum(struct mlx4_cqe *cqe, struct sk_buff *skb, void *va, hdr += sizeof(struct vlan_hdr); } - if (cqe->status & cpu_to_be16(MLX4_CQE_STATUS_IPV4)) - return get_fixed_ipv4_csum(hw_checksum, skb, hdr); #if IS_ENABLED(CONFIG_IPV6) if (cqe->status & cpu_to_be16(MLX4_CQE_STATUS_IPV6)) return get_fixed_ipv6_csum(hw_checksum, skb, hdr); #endif - return 0; + return get_fixed_ipv4_csum(hw_checksum, skb, hdr); } int mlx4_en_process_rx_cq(struct net_device *dev, struct mlx4_en_cq *cq, int budget) @@ -650,6 +658,7 @@ int mlx4_en_process_rx_cq(struct net_device *dev, struct mlx4_en_cq *cq, int bud int cq_ring = cq->ring; bool doorbell_pending; struct mlx4_cqe *cqe; + struct xdp_buff xdp; int polled = 0; int index; @@ -664,6 +673,7 @@ int mlx4_en_process_rx_cq(struct net_device *dev, struct mlx4_en_cq *cq, int bud /* Protect accesses to: ring->xdp_prog, priv->mac_hash list */ rcu_read_lock(); xdp_prog = rcu_dereference(ring->xdp_prog); + xdp.rxq = &ring->xdp_rxq; doorbell_pending = 0; /* We assume a 1:1 mapping between CQEs and Rx descriptors, so Rx @@ -748,7 +758,6 @@ int mlx4_en_process_rx_cq(struct net_device *dev, struct mlx4_en_cq *cq, int bud * read bytes but not past the end of the frag. */ if (xdp_prog) { - struct xdp_buff xdp; dma_addr_t dma; void *orig_data; u32 act; @@ -814,33 +823,33 @@ xdp_drop_no_cnt: if (likely(dev->features & NETIF_F_RXCSUM)) { if (cqe->status & cpu_to_be16(MLX4_CQE_STATUS_TCP | MLX4_CQE_STATUS_UDP)) { - if ((cqe->status & cpu_to_be16(MLX4_CQE_STATUS_IPOK)) && - cqe->checksum == cpu_to_be16(0xffff)) { - bool l2_tunnel = (dev->hw_enc_features & NETIF_F_RXCSUM) && - (cqe->vlan_my_qpn & cpu_to_be32(MLX4_CQE_L2_TUNNEL)); - - ip_summed = CHECKSUM_UNNECESSARY; - hash_type = PKT_HASH_TYPE_L4; - if (l2_tunnel) - skb->csum_level = 1; - ring->csum_ok++; - } else { + bool l2_tunnel; + + if (!((cqe->status & cpu_to_be16(MLX4_CQE_STATUS_IPOK)) && + cqe->checksum == cpu_to_be16(0xffff))) goto csum_none; - } + + l2_tunnel = (dev->hw_enc_features & NETIF_F_RXCSUM) && + (cqe->vlan_my_qpn & cpu_to_be32(MLX4_CQE_L2_TUNNEL)); + ip_summed = CHECKSUM_UNNECESSARY; + hash_type = PKT_HASH_TYPE_L4; + if (l2_tunnel) + skb->csum_level = 1; + ring->csum_ok++; } else { - if (priv->flags & MLX4_EN_FLAG_RX_CSUM_NON_TCP_UDP && - (cqe->status & cpu_to_be16(MLX4_CQE_STATUS_IPV4 | - MLX4_CQE_STATUS_IPV6))) { - if (check_csum(cqe, skb, va, dev->features)) { - goto csum_none; - } else { - ip_summed = CHECKSUM_COMPLETE; - hash_type = PKT_HASH_TYPE_L3; - ring->csum_complete++; - } - } else { + if (!(priv->flags & MLX4_EN_FLAG_RX_CSUM_NON_TCP_UDP && + (cqe->status & cpu_to_be16(MLX4_CQE_STATUS_IPV4 | +#if IS_ENABLED(CONFIG_IPV6) + MLX4_CQE_STATUS_IPV6)))) +#else + 0)))) +#endif goto csum_none; - } + if (check_csum(cqe, skb, va, dev->features)) + goto csum_none; + ip_summed = CHECKSUM_COMPLETE; + hash_type = PKT_HASH_TYPE_L3; + ring->csum_complete++; } } else { csum_none: diff --git a/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h b/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h index 2b72677..f470ae3 100644 --- a/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h +++ b/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h @@ -46,6 +46,7 @@ #endif #include <linux/cpu_rmap.h> #include <linux/ptp_clock_kernel.h> +#include <net/xdp.h> #include <linux/mlx4/device.h> #include <linux/mlx4/qp.h> @@ -356,6 +357,7 @@ struct mlx4_en_rx_ring { unsigned long dropped; int hwtstamp_rx_filter; cpumask_var_t affinity_mask; + struct xdp_rxq_info xdp_rxq; }; struct mlx4_en_cq { @@ -479,6 +481,7 @@ struct mlx4_en_frag_info { #define MLX4_EN_BW_MIN 1 #define MLX4_EN_BW_MAX 100 /* Utilize 100% of the line */ +#define MLX4_EN_TC_VENDOR 0 #define MLX4_EN_TC_ETS 7 enum dcb_pfc_type { @@ -719,7 +722,7 @@ void mlx4_en_set_num_rx_rings(struct mlx4_en_dev *mdev); void mlx4_en_recover_from_oom(struct mlx4_en_priv *priv); int mlx4_en_create_rx_ring(struct mlx4_en_priv *priv, struct mlx4_en_rx_ring **pring, - u32 size, u16 stride, int node); + u32 size, u16 stride, int node, int queue_index); void mlx4_en_destroy_rx_ring(struct mlx4_en_priv *priv, struct mlx4_en_rx_ring **pring, u32 size, u16 stride); diff --git a/drivers/net/ethernet/mellanox/mlx4/mr.c b/drivers/net/ethernet/mellanox/mlx4/mr.c index c7c0764..2e84f10 100644 --- a/drivers/net/ethernet/mellanox/mlx4/mr.c +++ b/drivers/net/ethernet/mellanox/mlx4/mr.c @@ -1103,30 +1103,16 @@ EXPORT_SYMBOL_GPL(mlx4_fmr_enable); void mlx4_fmr_unmap(struct mlx4_dev *dev, struct mlx4_fmr *fmr, u32 *lkey, u32 *rkey) { - struct mlx4_cmd_mailbox *mailbox; - int err; - if (!fmr->maps) return; - fmr->maps = 0; + /* To unmap: it is sufficient to take back ownership from HW */ + *(u8 *)fmr->mpt = MLX4_MPT_STATUS_SW; - mailbox = mlx4_alloc_cmd_mailbox(dev); - if (IS_ERR(mailbox)) { - err = PTR_ERR(mailbox); - pr_warn("mlx4_ib: mlx4_alloc_cmd_mailbox failed (%d)\n", err); - return; - } + /* Make sure MPT status is visible */ + wmb(); - err = mlx4_HW2SW_MPT(dev, NULL, - key_to_hw_index(fmr->mr.key) & - (dev->caps.num_mpts - 1)); - mlx4_free_cmd_mailbox(dev, mailbox); - if (err) { - pr_warn("mlx4_ib: mlx4_HW2SW_MPT failed (%d)\n", err); - return; - } - fmr->mr.enabled = MLX4_MPT_EN_SW; + fmr->maps = 0; } EXPORT_SYMBOL_GPL(mlx4_fmr_unmap); @@ -1136,6 +1122,22 @@ int mlx4_fmr_free(struct mlx4_dev *dev, struct mlx4_fmr *fmr) if (fmr->maps) return -EBUSY; + if (fmr->mr.enabled == MLX4_MPT_EN_HW) { + /* In case of FMR was enabled and unmapped + * make sure to give ownership of MPT back to HW + * so HW2SW_MPT command will success. + */ + *(u8 *)fmr->mpt = MLX4_MPT_STATUS_SW; + /* Make sure MPT status is visible before changing MPT fields */ + wmb(); + fmr->mpt->length = 0; + fmr->mpt->start = 0; + /* Make sure MPT data is visible after changing MPT status */ + wmb(); + *(u8 *)fmr->mpt = MLX4_MPT_STATUS_HW; + /* make sure MPT status is visible */ + wmb(); + } ret = mlx4_mr_free(dev, &fmr->mr); if (ret) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/Makefile b/drivers/net/ethernet/mellanox/mlx5/core/Makefile index 19b21b4..c805769 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/Makefile +++ b/drivers/net/ethernet/mellanox/mlx5/core/Makefile @@ -14,7 +14,7 @@ mlx5_core-$(CONFIG_MLX5_FPGA) += fpga/cmd.o fpga/core.o fpga/conn.o fpga/sdk.o \ fpga/ipsec.o mlx5_core-$(CONFIG_MLX5_CORE_EN) += en_main.o en_common.o en_fs.o en_ethtool.o \ - en_tx.o en_rx.o en_rx_am.o en_txrx.o en_stats.o vxlan.o \ + en_tx.o en_rx.o en_dim.o en_txrx.o en_stats.o vxlan.o \ en_arfs.o en_fs_ethtool.o en_selftest.o mlx5_core-$(CONFIG_MLX5_MPFS) += lib/mpfs.o diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en.h b/drivers/net/ethernet/mellanox/mlx5/core/en.h index c2d89bf..4c9360b 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en.h @@ -44,8 +44,11 @@ #include <linux/mlx5/port.h> #include <linux/mlx5/vport.h> #include <linux/mlx5/transobj.h> +#include <linux/mlx5/fs.h> #include <linux/rhashtable.h> #include <net/switchdev.h> +#include <net/xdp.h> +#include <linux/net_dim.h> #include "wq.h" #include "mlx5_core.h" #include "en_stats.h" @@ -226,12 +229,6 @@ enum mlx5e_priv_flag { #define MLX5E_MAX_BW_ALLOC 100 /* Max percentage of BW allocation */ #endif -struct mlx5e_cq_moder { - u16 usec; - u16 pkts; - u8 cq_period_mode; -}; - struct mlx5e_params { u8 log_sq_size; u8 rq_wq_type; @@ -242,8 +239,8 @@ struct mlx5e_params { u16 num_channels; u8 num_tc; bool rx_cqe_compress_def; - struct mlx5e_cq_moder rx_cq_moderation; - struct mlx5e_cq_moder tx_cq_moderation; + struct net_dim_cq_moder rx_cq_moderation; + struct net_dim_cq_moder tx_cq_moderation; bool lro_en; u32 lro_wqe_sz; u16 tx_max_inline; @@ -253,7 +250,7 @@ struct mlx5e_params { u32 indirection_rqt[MLX5E_INDIR_RQT_SIZE]; bool vlan_strip_disable; bool scatter_fcs_en; - bool rx_am_enabled; + bool rx_dim_enabled; u32 lro_timeout; u32 pflags; struct bpf_prog *xdp_prog; @@ -472,32 +469,6 @@ struct mlx5e_mpw_info { u16 skbs_frags[MLX5_MPWRQ_PAGES_PER_WQE]; }; -struct mlx5e_rx_am_stats { - int ppms; /* packets per msec */ - int bpms; /* bytes per msec */ - int epms; /* events per msec */ -}; - -struct mlx5e_rx_am_sample { - ktime_t time; - u32 pkt_ctr; - u32 byte_ctr; - u16 event_ctr; -}; - -struct mlx5e_rx_am { /* Adaptive Moderation */ - u8 state; - struct mlx5e_rx_am_stats prev_stats; - struct mlx5e_rx_am_sample start_sample; - struct work_struct work; - u8 profile_ix; - u8 mode; - u8 tune_state; - u8 steps_right; - u8 steps_left; - u8 tired; -}; - /* a single cache unit is capable to serve one napi call (for non-striding rq) * or a MPWQE (for striding rq). */ @@ -558,7 +529,7 @@ struct mlx5e_rq { unsigned long state; int ix; - struct mlx5e_rx_am am; /* Adaptive Moderation */ + struct net_dim dim; /* Dynamic Interrupt Moderation */ /* XDP */ struct bpf_prog *xdp_prog; @@ -571,6 +542,9 @@ struct mlx5e_rq { u32 rqn; struct mlx5_core_dev *mdev; struct mlx5_core_mkey umr_mkey; + + /* XDP read-mostly */ + struct xdp_rxq_info xdp_rxq; } ____cacheline_aligned_in_smp; struct mlx5e_channel { @@ -587,6 +561,7 @@ struct mlx5e_channel { /* data path - accessed per napi poll */ struct irq_desc *irq_desc; + struct mlx5e_ch_stats stats; /* control */ struct mlx5e_priv *priv; @@ -655,6 +630,7 @@ struct mlx5e_tc_table { struct rhashtable ht; DECLARE_HASHTABLE(mod_hdr_tbl, 8); + DECLARE_HASHTABLE(hairpin_tbl, 8); }; struct mlx5e_vlan_table { @@ -722,6 +698,11 @@ enum { MLX5E_ARFS_FT_LEVEL }; +enum { + MLX5E_TC_FT_LEVEL = 0, + MLX5E_TC_TTC_FT_LEVEL, +}; + struct mlx5e_ethtool_table { struct mlx5_flow_table *ft; int num_rules; @@ -860,11 +841,7 @@ void mlx5e_dealloc_rx_wqe(struct mlx5e_rq *rq, u16 ix); void mlx5e_dealloc_rx_mpwqe(struct mlx5e_rq *rq, u16 ix); void mlx5e_free_rx_mpwqe(struct mlx5e_rq *rq, struct mlx5e_mpw_info *wi); -void mlx5e_rx_am(struct mlx5e_rq *rq); -void mlx5e_rx_am_work(struct work_struct *work); -struct mlx5e_cq_moder mlx5e_am_get_def_profile(u8 rx_cq_period_mode); - -void mlx5e_update_stats(struct mlx5e_priv *priv, bool full); +void mlx5e_update_stats(struct mlx5e_priv *priv); int mlx5e_create_flow_steering(struct mlx5e_priv *priv); void mlx5e_destroy_flow_steering(struct mlx5e_priv *priv); @@ -1054,11 +1031,26 @@ int mlx5e_create_direct_tirs(struct mlx5e_priv *priv); void mlx5e_destroy_direct_tirs(struct mlx5e_priv *priv); void mlx5e_destroy_rqt(struct mlx5e_priv *priv, struct mlx5e_rqt *rqt); -int mlx5e_create_ttc_table(struct mlx5e_priv *priv); -void mlx5e_destroy_ttc_table(struct mlx5e_priv *priv); +struct ttc_params { + struct mlx5_flow_table_attr ft_attr; + u32 any_tt_tirn; + u32 indir_tirn[MLX5E_NUM_INDIR_TIRS]; + struct mlx5e_ttc_table *inner_ttc; +}; + +void mlx5e_set_ttc_basic_params(struct mlx5e_priv *priv, struct ttc_params *ttc_params); +void mlx5e_set_ttc_ft_params(struct ttc_params *ttc_params); +void mlx5e_set_inner_ttc_ft_params(struct ttc_params *ttc_params); -int mlx5e_create_inner_ttc_table(struct mlx5e_priv *priv); -void mlx5e_destroy_inner_ttc_table(struct mlx5e_priv *priv); +int mlx5e_create_ttc_table(struct mlx5e_priv *priv, struct ttc_params *params, + struct mlx5e_ttc_table *ttc); +void mlx5e_destroy_ttc_table(struct mlx5e_priv *priv, + struct mlx5e_ttc_table *ttc); + +int mlx5e_create_inner_ttc_table(struct mlx5e_priv *priv, struct ttc_params *params, + struct mlx5e_ttc_table *ttc); +void mlx5e_destroy_inner_ttc_table(struct mlx5e_priv *priv, + struct mlx5e_ttc_table *ttc); int mlx5e_create_tis(struct mlx5_core_dev *mdev, int tc, u32 underlay_qpn, u32 *tisn); @@ -1071,6 +1063,8 @@ int mlx5e_open(struct net_device *netdev); void mlx5e_update_stats_work(struct work_struct *work); u32 mlx5e_choose_lro_timeout(struct mlx5_core_dev *mdev, u32 wanted_timeout); +int mlx5e_bits_invert(unsigned long a, int size); + /* ethtool helpers */ void mlx5e_ethtool_get_drvinfo(struct mlx5e_priv *priv, struct ethtool_drvinfo *drvinfo); @@ -1110,4 +1104,5 @@ void mlx5e_build_nic_params(struct mlx5_core_dev *mdev, struct mlx5e_params *params, u16 max_channels); u8 mlx5e_params_calculate_tx_min_inline(struct mlx5_core_dev *mdev); +void mlx5e_rx_dim_work(struct work_struct *work); #endif /* __MLX5_EN_H__ */ diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_dim.c b/drivers/net/ethernet/mellanox/mlx5/core/en_dim.c new file mode 100644 index 0000000..602851a --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_dim.c @@ -0,0 +1,48 @@ +/* + * Copyright (c) 2016, Mellanox Technologies. All rights reserved. + * + * This software is available to you under a choice of one of two + * licenses. You may choose to be licensed under the terms of the GNU + * General Public License (GPL) Version 2, available from the file + * COPYING in the main directory of this source tree, or the + * OpenIB.org BSD license below: + * + * Redistribution and use in source and binary forms, with or + * without modification, are permitted provided that the following + * conditions are met: + * + * - Redistributions of source code must retain the above + * copyright notice, this list of conditions and the following + * disclaimer. + * + * - Redistributions in binary form must reproduce the above + * copyright notice, this list of conditions and the following + * disclaimer in the documentation and/or other materials + * provided with the distribution. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + */ + +#include <linux/net_dim.h> +#include "en.h" + +void mlx5e_rx_dim_work(struct work_struct *work) +{ + struct net_dim *dim = container_of(work, struct net_dim, + work); + struct mlx5e_rq *rq = container_of(dim, struct mlx5e_rq, dim); + struct net_dim_cq_moder cur_profile = net_dim_get_profile(dim->mode, + dim->profile_ix); + + mlx5_core_modify_cq_moderation(rq->mdev, &rq->cq.mcq, + cur_profile.usec, cur_profile.pkts); + + dim->state = NET_DIM_START_MEASURE; +} diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c b/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c index ea5fff2..cc8048f 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c @@ -207,7 +207,7 @@ void mlx5e_ethtool_get_ethtool_stats(struct mlx5e_priv *priv, return; mutex_lock(&priv->state_lock); - mlx5e_update_stats(priv, true); + mlx5e_update_stats(priv); mutex_unlock(&priv->state_lock); for (i = 0; i < mlx5e_num_stats_grps; i++) @@ -295,7 +295,6 @@ int mlx5e_ethtool_set_ringparam(struct mlx5e_priv *priv, struct mlx5e_channels new_channels = {}; u32 rx_pending_wqes; u32 min_rq_size; - u32 max_rq_size; u8 log_rq_size; u8 log_sq_size; u32 num_mtts; @@ -314,8 +313,6 @@ int mlx5e_ethtool_set_ringparam(struct mlx5e_priv *priv, min_rq_size = mlx5e_rx_wqes_to_packets(priv, rq_wq_type, 1 << mlx5_min_log_rq_size(rq_wq_type)); - max_rq_size = mlx5e_rx_wqes_to_packets(priv, rq_wq_type, - 1 << mlx5_max_log_rq_size(rq_wq_type)); rx_pending_wqes = mlx5e_packets_to_rx_wqes(priv, rq_wq_type, param->rx_pending); @@ -325,12 +322,6 @@ int mlx5e_ethtool_set_ringparam(struct mlx5e_priv *priv, min_rq_size); return -EINVAL; } - if (param->rx_pending > max_rq_size) { - netdev_info(priv->netdev, "%s: rx_pending (%d) > max (%d)\n", - __func__, param->rx_pending, - max_rq_size); - return -EINVAL; - } num_mtts = MLX5E_REQUIRED_MTTS(rx_pending_wqes); if (priv->channels.params.rq_wq_type == MLX5_WQ_TYPE_LINKED_LIST_STRIDING_RQ && @@ -346,12 +337,6 @@ int mlx5e_ethtool_set_ringparam(struct mlx5e_priv *priv, 1 << MLX5E_PARAMS_MINIMUM_LOG_SQ_SIZE); return -EINVAL; } - if (param->tx_pending > (1 << MLX5E_PARAMS_MAXIMUM_LOG_SQ_SIZE)) { - netdev_info(priv->netdev, "%s: tx_pending (%d) > max (%d)\n", - __func__, param->tx_pending, - 1 << MLX5E_PARAMS_MAXIMUM_LOG_SQ_SIZE); - return -EINVAL; - } log_rq_size = order_base_2(rx_pending_wqes); log_sq_size = order_base_2(param->tx_pending); @@ -479,7 +464,7 @@ int mlx5e_ethtool_get_coalesce(struct mlx5e_priv *priv, coal->rx_max_coalesced_frames = priv->channels.params.rx_cq_moderation.pkts; coal->tx_coalesce_usecs = priv->channels.params.tx_cq_moderation.usec; coal->tx_max_coalesced_frames = priv->channels.params.tx_cq_moderation.pkts; - coal->use_adaptive_rx_coalesce = priv->channels.params.rx_am_enabled; + coal->use_adaptive_rx_coalesce = priv->channels.params.rx_dim_enabled; return 0; } @@ -533,7 +518,7 @@ int mlx5e_ethtool_set_coalesce(struct mlx5e_priv *priv, new_channels.params.tx_cq_moderation.pkts = coal->tx_max_coalesced_frames; new_channels.params.rx_cq_moderation.usec = coal->rx_coalesce_usecs; new_channels.params.rx_cq_moderation.pkts = coal->rx_max_coalesced_frames; - new_channels.params.rx_am_enabled = !!coal->use_adaptive_rx_coalesce; + new_channels.params.rx_dim_enabled = !!coal->use_adaptive_rx_coalesce; if (!test_bit(MLX5E_STATE_OPENED, &priv->state)) { priv->channels.params = new_channels.params; @@ -541,7 +526,7 @@ int mlx5e_ethtool_set_coalesce(struct mlx5e_priv *priv, } /* we are opened */ - reset = !!coal->use_adaptive_rx_coalesce != priv->channels.params.rx_am_enabled; + reset = !!coal->use_adaptive_rx_coalesce != priv->channels.params.rx_dim_enabled; if (!reset) { mlx5e_set_priv_channels_coalesce(priv, coal); priv->channels.params = new_channels.params; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c b/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c index def5134..f64dda2 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c @@ -806,25 +806,25 @@ mlx5e_generate_ttc_rule(struct mlx5e_priv *priv, return err ? ERR_PTR(err) : rule; } -static int mlx5e_generate_ttc_table_rules(struct mlx5e_priv *priv) +static int mlx5e_generate_ttc_table_rules(struct mlx5e_priv *priv, + struct ttc_params *params, + struct mlx5e_ttc_table *ttc) { struct mlx5_flow_destination dest = {}; - struct mlx5e_ttc_table *ttc; struct mlx5_flow_handle **rules; struct mlx5_flow_table *ft; int tt; int err; - ttc = &priv->fs.ttc; ft = ttc->ft.t; rules = ttc->rules; dest.type = MLX5_FLOW_DESTINATION_TYPE_TIR; for (tt = 0; tt < MLX5E_NUM_TT; tt++) { if (tt == MLX5E_TT_ANY) - dest.tir_num = priv->direct_tir[0].tirn; + dest.tir_num = params->any_tt_tirn; else - dest.tir_num = priv->indir_tir[tt].tirn; + dest.tir_num = params->indir_tirn[tt]; rules[tt] = mlx5e_generate_ttc_rule(priv, ft, &dest, ttc_rules[tt].etype, ttc_rules[tt].proto); @@ -832,12 +832,12 @@ static int mlx5e_generate_ttc_table_rules(struct mlx5e_priv *priv) goto del_rules; } - if (!mlx5e_tunnel_inner_ft_supported(priv->mdev)) + if (!params->inner_ttc || !mlx5e_tunnel_inner_ft_supported(priv->mdev)) return 0; rules = ttc->tunnel_rules; dest.type = MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE; - dest.ft = priv->fs.inner_ttc.ft.t; + dest.ft = params->inner_ttc->ft.t; for (tt = 0; tt < MLX5E_NUM_TUNNEL_TT; tt++) { rules[tt] = mlx5e_generate_ttc_rule(priv, ft, &dest, ttc_tunnel_rules[tt].etype, @@ -977,25 +977,25 @@ mlx5e_generate_inner_ttc_rule(struct mlx5e_priv *priv, return err ? ERR_PTR(err) : rule; } -static int mlx5e_generate_inner_ttc_table_rules(struct mlx5e_priv *priv) +static int mlx5e_generate_inner_ttc_table_rules(struct mlx5e_priv *priv, + struct ttc_params *params, + struct mlx5e_ttc_table *ttc) { struct mlx5_flow_destination dest = {}; struct mlx5_flow_handle **rules; - struct mlx5e_ttc_table *ttc; struct mlx5_flow_table *ft; int err; int tt; - ttc = &priv->fs.inner_ttc; ft = ttc->ft.t; rules = ttc->rules; dest.type = MLX5_FLOW_DESTINATION_TYPE_TIR; for (tt = 0; tt < MLX5E_NUM_TT; tt++) { if (tt == MLX5E_TT_ANY) - dest.tir_num = priv->direct_tir[0].tirn; + dest.tir_num = params->any_tt_tirn; else - dest.tir_num = priv->inner_indir_tir[tt].tirn; + dest.tir_num = params->indir_tirn[tt]; rules[tt] = mlx5e_generate_inner_ttc_rule(priv, ft, &dest, ttc_rules[tt].etype, @@ -1075,21 +1075,42 @@ err: return err; } -int mlx5e_create_inner_ttc_table(struct mlx5e_priv *priv) +void mlx5e_set_ttc_basic_params(struct mlx5e_priv *priv, + struct ttc_params *ttc_params) +{ + ttc_params->any_tt_tirn = priv->direct_tir[0].tirn; + ttc_params->inner_ttc = &priv->fs.inner_ttc; +} + +void mlx5e_set_inner_ttc_ft_params(struct ttc_params *ttc_params) +{ + struct mlx5_flow_table_attr *ft_attr = &ttc_params->ft_attr; + + ft_attr->max_fte = MLX5E_INNER_TTC_TABLE_SIZE; + ft_attr->level = MLX5E_INNER_TTC_FT_LEVEL; + ft_attr->prio = MLX5E_NIC_PRIO; +} + +void mlx5e_set_ttc_ft_params(struct ttc_params *ttc_params) + +{ + struct mlx5_flow_table_attr *ft_attr = &ttc_params->ft_attr; + + ft_attr->max_fte = MLX5E_TTC_TABLE_SIZE; + ft_attr->level = MLX5E_TTC_FT_LEVEL; + ft_attr->prio = MLX5E_NIC_PRIO; +} + +int mlx5e_create_inner_ttc_table(struct mlx5e_priv *priv, struct ttc_params *params, + struct mlx5e_ttc_table *ttc) { - struct mlx5e_ttc_table *ttc = &priv->fs.inner_ttc; - struct mlx5_flow_table_attr ft_attr = {}; struct mlx5e_flow_table *ft = &ttc->ft; int err; if (!mlx5e_tunnel_inner_ft_supported(priv->mdev)) return 0; - ft_attr.max_fte = MLX5E_INNER_TTC_TABLE_SIZE; - ft_attr.level = MLX5E_INNER_TTC_FT_LEVEL; - ft_attr.prio = MLX5E_NIC_PRIO; - - ft->t = mlx5_create_flow_table(priv->fs.ns, &ft_attr); + ft->t = mlx5_create_flow_table(priv->fs.ns, ¶ms->ft_attr); if (IS_ERR(ft->t)) { err = PTR_ERR(ft->t); ft->t = NULL; @@ -1100,7 +1121,7 @@ int mlx5e_create_inner_ttc_table(struct mlx5e_priv *priv) if (err) goto err; - err = mlx5e_generate_inner_ttc_table_rules(priv); + err = mlx5e_generate_inner_ttc_table_rules(priv, params, ttc); if (err) goto err; @@ -1111,10 +1132,9 @@ err: return err; } -void mlx5e_destroy_inner_ttc_table(struct mlx5e_priv *priv) +void mlx5e_destroy_inner_ttc_table(struct mlx5e_priv *priv, + struct mlx5e_ttc_table *ttc) { - struct mlx5e_ttc_table *ttc = &priv->fs.inner_ttc; - if (!mlx5e_tunnel_inner_ft_supported(priv->mdev)) return; @@ -1122,27 +1142,21 @@ void mlx5e_destroy_inner_ttc_table(struct mlx5e_priv *priv) mlx5e_destroy_flow_table(&ttc->ft); } -void mlx5e_destroy_ttc_table(struct mlx5e_priv *priv) +void mlx5e_destroy_ttc_table(struct mlx5e_priv *priv, + struct mlx5e_ttc_table *ttc) { - struct mlx5e_ttc_table *ttc = &priv->fs.ttc; - mlx5e_cleanup_ttc_rules(ttc); mlx5e_destroy_flow_table(&ttc->ft); } -int mlx5e_create_ttc_table(struct mlx5e_priv *priv) +int mlx5e_create_ttc_table(struct mlx5e_priv *priv, struct ttc_params *params, + struct mlx5e_ttc_table *ttc) { bool match_ipv_outer = MLX5_CAP_FLOWTABLE_NIC_RX(priv->mdev, ft_field_support.outer_ip_version); - struct mlx5e_ttc_table *ttc = &priv->fs.ttc; - struct mlx5_flow_table_attr ft_attr = {}; struct mlx5e_flow_table *ft = &ttc->ft; int err; - ft_attr.max_fte = MLX5E_TTC_TABLE_SIZE; - ft_attr.level = MLX5E_TTC_FT_LEVEL; - ft_attr.prio = MLX5E_NIC_PRIO; - - ft->t = mlx5_create_flow_table(priv->fs.ns, &ft_attr); + ft->t = mlx5_create_flow_table(priv->fs.ns, ¶ms->ft_attr); if (IS_ERR(ft->t)) { err = PTR_ERR(ft->t); ft->t = NULL; @@ -1153,7 +1167,7 @@ int mlx5e_create_ttc_table(struct mlx5e_priv *priv) if (err) goto err; - err = mlx5e_generate_ttc_table_rules(priv); + err = mlx5e_generate_ttc_table_rules(priv, params, ttc); if (err) goto err; @@ -1474,7 +1488,8 @@ static void mlx5e_destroy_vlan_table(struct mlx5e_priv *priv) int mlx5e_create_flow_steering(struct mlx5e_priv *priv) { - int err; + struct ttc_params ttc_params = {}; + int tt, err; priv->fs.ns = mlx5_get_flow_namespace(priv->mdev, MLX5_FLOW_NAMESPACE_KERNEL); @@ -1489,14 +1504,23 @@ int mlx5e_create_flow_steering(struct mlx5e_priv *priv) priv->netdev->hw_features &= ~NETIF_F_NTUPLE; } - err = mlx5e_create_inner_ttc_table(priv); + mlx5e_set_ttc_basic_params(priv, &ttc_params); + mlx5e_set_inner_ttc_ft_params(&ttc_params); + for (tt = 0; tt < MLX5E_NUM_INDIR_TIRS; tt++) + ttc_params.indir_tirn[tt] = priv->inner_indir_tir[tt].tirn; + + err = mlx5e_create_inner_ttc_table(priv, &ttc_params, &priv->fs.inner_ttc); if (err) { netdev_err(priv->netdev, "Failed to create inner ttc table, err=%d\n", err); goto err_destroy_arfs_tables; } - err = mlx5e_create_ttc_table(priv); + mlx5e_set_ttc_ft_params(&ttc_params); + for (tt = 0; tt < MLX5E_NUM_INDIR_TIRS; tt++) + ttc_params.indir_tirn[tt] = priv->indir_tir[tt].tirn; + + err = mlx5e_create_ttc_table(priv, &ttc_params, &priv->fs.ttc); if (err) { netdev_err(priv->netdev, "Failed to create ttc table, err=%d\n", err); @@ -1524,9 +1548,9 @@ int mlx5e_create_flow_steering(struct mlx5e_priv *priv) err_destroy_l2_table: mlx5e_destroy_l2_table(priv); err_destroy_ttc_table: - mlx5e_destroy_ttc_table(priv); + mlx5e_destroy_ttc_table(priv, &priv->fs.ttc); err_destroy_inner_ttc_table: - mlx5e_destroy_inner_ttc_table(priv); + mlx5e_destroy_inner_ttc_table(priv, &priv->fs.inner_ttc); err_destroy_arfs_tables: mlx5e_arfs_destroy_tables(priv); @@ -1537,8 +1561,8 @@ void mlx5e_destroy_flow_steering(struct mlx5e_priv *priv) { mlx5e_destroy_vlan_table(priv); mlx5e_destroy_l2_table(priv); - mlx5e_destroy_ttc_table(priv); - mlx5e_destroy_inner_ttc_table(priv); + mlx5e_destroy_ttc_table(priv, &priv->fs.ttc); + mlx5e_destroy_inner_ttc_table(priv, &priv->fs.inner_ttc); mlx5e_arfs_destroy_tables(priv); mlx5e_ethtool_cleanup_steering(priv); } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c index d8aefee..47bab84 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c @@ -173,182 +173,23 @@ unlock: rtnl_unlock(); } -static void mlx5e_update_sw_counters(struct mlx5e_priv *priv) +void mlx5e_update_stats(struct mlx5e_priv *priv) { - struct mlx5e_sw_stats temp, *s = &temp; - struct mlx5e_rq_stats *rq_stats; - struct mlx5e_sq_stats *sq_stats; - int i, j; - - memset(s, 0, sizeof(*s)); - for (i = 0; i < priv->channels.num; i++) { - struct mlx5e_channel *c = priv->channels.c[i]; - - rq_stats = &c->rq.stats; - - s->rx_packets += rq_stats->packets; - s->rx_bytes += rq_stats->bytes; - s->rx_lro_packets += rq_stats->lro_packets; - s->rx_lro_bytes += rq_stats->lro_bytes; - s->rx_removed_vlan_packets += rq_stats->removed_vlan_packets; - s->rx_csum_none += rq_stats->csum_none; - s->rx_csum_complete += rq_stats->csum_complete; - s->rx_csum_unnecessary += rq_stats->csum_unnecessary; - s->rx_csum_unnecessary_inner += rq_stats->csum_unnecessary_inner; - s->rx_xdp_drop += rq_stats->xdp_drop; - s->rx_xdp_tx += rq_stats->xdp_tx; - s->rx_xdp_tx_full += rq_stats->xdp_tx_full; - s->rx_wqe_err += rq_stats->wqe_err; - s->rx_mpwqe_filler += rq_stats->mpwqe_filler; - s->rx_buff_alloc_err += rq_stats->buff_alloc_err; - s->rx_cqe_compress_blks += rq_stats->cqe_compress_blks; - s->rx_cqe_compress_pkts += rq_stats->cqe_compress_pkts; - s->rx_page_reuse += rq_stats->page_reuse; - s->rx_cache_reuse += rq_stats->cache_reuse; - s->rx_cache_full += rq_stats->cache_full; - s->rx_cache_empty += rq_stats->cache_empty; - s->rx_cache_busy += rq_stats->cache_busy; - s->rx_cache_waive += rq_stats->cache_waive; - - for (j = 0; j < priv->channels.params.num_tc; j++) { - sq_stats = &c->sq[j].stats; - - s->tx_packets += sq_stats->packets; - s->tx_bytes += sq_stats->bytes; - s->tx_tso_packets += sq_stats->tso_packets; - s->tx_tso_bytes += sq_stats->tso_bytes; - s->tx_tso_inner_packets += sq_stats->tso_inner_packets; - s->tx_tso_inner_bytes += sq_stats->tso_inner_bytes; - s->tx_added_vlan_packets += sq_stats->added_vlan_packets; - s->tx_queue_stopped += sq_stats->stopped; - s->tx_queue_wake += sq_stats->wake; - s->tx_queue_dropped += sq_stats->dropped; - s->tx_xmit_more += sq_stats->xmit_more; - s->tx_csum_partial_inner += sq_stats->csum_partial_inner; - s->tx_csum_none += sq_stats->csum_none; - s->tx_csum_partial += sq_stats->csum_partial; - } - } - - s->link_down_events_phy = MLX5_GET(ppcnt_reg, - priv->stats.pport.phy_counters, - counter_set.phys_layer_cntrs.link_down_events); - memcpy(&priv->stats.sw, s, sizeof(*s)); -} - -static void mlx5e_update_vport_counters(struct mlx5e_priv *priv) -{ - int outlen = MLX5_ST_SZ_BYTES(query_vport_counter_out); - u32 *out = (u32 *)priv->stats.vport.query_vport_out; - u32 in[MLX5_ST_SZ_DW(query_vport_counter_in)] = {0}; - struct mlx5_core_dev *mdev = priv->mdev; - - MLX5_SET(query_vport_counter_in, in, opcode, - MLX5_CMD_OP_QUERY_VPORT_COUNTER); - MLX5_SET(query_vport_counter_in, in, op_mod, 0); - MLX5_SET(query_vport_counter_in, in, other_vport, 0); - - mlx5_cmd_exec(mdev, in, sizeof(in), out, outlen); -} - -static void mlx5e_update_pport_counters(struct mlx5e_priv *priv, bool full) -{ - struct mlx5e_pport_stats *pstats = &priv->stats.pport; - struct mlx5_core_dev *mdev = priv->mdev; - u32 in[MLX5_ST_SZ_DW(ppcnt_reg)] = {0}; - int sz = MLX5_ST_SZ_BYTES(ppcnt_reg); - int prio; - void *out; - - MLX5_SET(ppcnt_reg, in, local_port, 1); - - out = pstats->IEEE_802_3_counters; - MLX5_SET(ppcnt_reg, in, grp, MLX5_IEEE_802_3_COUNTERS_GROUP); - mlx5_core_access_reg(mdev, in, sz, out, sz, MLX5_REG_PPCNT, 0, 0); - - if (!full) - return; - - out = pstats->RFC_2863_counters; - MLX5_SET(ppcnt_reg, in, grp, MLX5_RFC_2863_COUNTERS_GROUP); - mlx5_core_access_reg(mdev, in, sz, out, sz, MLX5_REG_PPCNT, 0, 0); - - out = pstats->RFC_2819_counters; - MLX5_SET(ppcnt_reg, in, grp, MLX5_RFC_2819_COUNTERS_GROUP); - mlx5_core_access_reg(mdev, in, sz, out, sz, MLX5_REG_PPCNT, 0, 0); - - out = pstats->phy_counters; - MLX5_SET(ppcnt_reg, in, grp, MLX5_PHYSICAL_LAYER_COUNTERS_GROUP); - mlx5_core_access_reg(mdev, in, sz, out, sz, MLX5_REG_PPCNT, 0, 0); - - if (MLX5_CAP_PCAM_FEATURE(mdev, ppcnt_statistical_group)) { - out = pstats->phy_statistical_counters; - MLX5_SET(ppcnt_reg, in, grp, MLX5_PHYSICAL_LAYER_STATISTICAL_GROUP); - mlx5_core_access_reg(mdev, in, sz, out, sz, MLX5_REG_PPCNT, 0, 0); - } - - if (MLX5_CAP_PCAM_FEATURE(mdev, rx_buffer_fullness_counters)) { - out = pstats->eth_ext_counters; - MLX5_SET(ppcnt_reg, in, grp, MLX5_ETHERNET_EXTENDED_COUNTERS_GROUP); - mlx5_core_access_reg(mdev, in, sz, out, sz, MLX5_REG_PPCNT, 0, 0); - } - - MLX5_SET(ppcnt_reg, in, grp, MLX5_PER_PRIORITY_COUNTERS_GROUP); - for (prio = 0; prio < NUM_PPORT_PRIO; prio++) { - out = pstats->per_prio_counters[prio]; - MLX5_SET(ppcnt_reg, in, prio_tc, prio); - mlx5_core_access_reg(mdev, in, sz, out, sz, - MLX5_REG_PPCNT, 0, 0); - } -} - -static void mlx5e_update_q_counter(struct mlx5e_priv *priv) -{ - struct mlx5e_qcounter_stats *qcnt = &priv->stats.qcnt; - u32 out[MLX5_ST_SZ_DW(query_q_counter_out)]; - int err; - - if (!priv->q_counter) - return; - - err = mlx5_core_query_q_counter(priv->mdev, priv->q_counter, 0, out, sizeof(out)); - if (err) - return; - - qcnt->rx_out_of_buffer = MLX5_GET(query_q_counter_out, out, out_of_buffer); -} - -static void mlx5e_update_pcie_counters(struct mlx5e_priv *priv) -{ - struct mlx5e_pcie_stats *pcie_stats = &priv->stats.pcie; - struct mlx5_core_dev *mdev = priv->mdev; - u32 in[MLX5_ST_SZ_DW(mpcnt_reg)] = {0}; - int sz = MLX5_ST_SZ_BYTES(mpcnt_reg); - void *out; - - if (!MLX5_CAP_MCAM_FEATURE(mdev, pcie_performance_group)) - return; - - out = pcie_stats->pcie_perf_counters; - MLX5_SET(mpcnt_reg, in, grp, MLX5_PCIE_PERFORMANCE_COUNTERS_GROUP); - mlx5_core_access_reg(mdev, in, sz, out, sz, MLX5_REG_MPCNT, 0, 0); -} + int i; -void mlx5e_update_stats(struct mlx5e_priv *priv, bool full) -{ - if (full) { - mlx5e_update_pcie_counters(priv); - mlx5e_ipsec_update_stats(priv); - } - mlx5e_update_pport_counters(priv, full); - mlx5e_update_vport_counters(priv); - mlx5e_update_q_counter(priv); - mlx5e_update_sw_counters(priv); + for (i = mlx5e_num_stats_grps - 1; i >= 0; i--) + if (mlx5e_stats_grps[i].update_stats) + mlx5e_stats_grps[i].update_stats(priv); } static void mlx5e_update_ndo_stats(struct mlx5e_priv *priv) { - mlx5e_update_stats(priv, false); + int i; + + for (i = mlx5e_num_stats_grps - 1; i >= 0; i--) + if (mlx5e_stats_grps[i].update_stats_mask & + MLX5E_NDO_UPDATE_STATS) + mlx5e_stats_grps[i].update_stats(priv); } void mlx5e_update_stats_work(struct work_struct *work) @@ -582,6 +423,10 @@ static int mlx5e_alloc_rq(struct mlx5e_channel *c, goto err_rq_wq_destroy; } + err = xdp_rxq_info_reg(&rq->xdp_rxq, rq->netdev, rq->ix); + if (err < 0) + goto err_rq_wq_destroy; + rq->buff.map_dir = rq->xdp_prog ? DMA_BIDIRECTIONAL : DMA_FROM_DEVICE; rq->buff.headroom = params->rq_headroom; @@ -674,8 +519,17 @@ static int mlx5e_alloc_rq(struct mlx5e_channel *c, wqe->data.lkey = rq->mkey_be; } - INIT_WORK(&rq->am.work, mlx5e_rx_am_work); - rq->am.mode = params->rx_cq_moderation.cq_period_mode; + INIT_WORK(&rq->dim.work, mlx5e_rx_dim_work); + + switch (params->rx_cq_moderation.cq_period_mode) { + case MLX5_CQ_PERIOD_MODE_START_FROM_CQE: + rq->dim.mode = NET_DIM_CQ_PERIOD_MODE_START_FROM_CQE; + break; + case MLX5_CQ_PERIOD_MODE_START_FROM_EQE: + default: + rq->dim.mode = NET_DIM_CQ_PERIOD_MODE_START_FROM_EQE; + } + rq->page_cache.head = 0; rq->page_cache.tail = 0; @@ -687,6 +541,7 @@ err_destroy_umr_mkey: err_rq_wq_destroy: if (rq->xdp_prog) bpf_prog_put(rq->xdp_prog); + xdp_rxq_info_unreg(&rq->xdp_rxq); mlx5_wq_destroy(&rq->wq_ctrl); return err; @@ -699,6 +554,8 @@ static void mlx5e_free_rq(struct mlx5e_rq *rq) if (rq->xdp_prog) bpf_prog_put(rq->xdp_prog); + xdp_rxq_info_unreg(&rq->xdp_rxq); + switch (rq->wq_type) { case MLX5_WQ_TYPE_LINKED_LIST_STRIDING_RQ: mlx5e_rq_free_mpwqe_info(rq); @@ -919,7 +776,7 @@ static int mlx5e_open_rq(struct mlx5e_channel *c, if (err) goto err_destroy_rq; - if (params->rx_am_enabled) + if (params->rx_dim_enabled) c->rq.state |= BIT(MLX5E_RQ_STATE_AM); return 0; @@ -952,7 +809,7 @@ static void mlx5e_deactivate_rq(struct mlx5e_rq *rq) static void mlx5e_close_rq(struct mlx5e_rq *rq) { - cancel_work_sync(&rq->am.work); + cancel_work_sync(&rq->dim.work); mlx5e_destroy_rq(rq); mlx5e_free_rx_descs(rq); mlx5e_free_rq(rq); @@ -1565,7 +1422,7 @@ static void mlx5e_destroy_cq(struct mlx5e_cq *cq) } static int mlx5e_open_cq(struct mlx5e_channel *c, - struct mlx5e_cq_moder moder, + struct net_dim_cq_moder moder, struct mlx5e_cq_param *param, struct mlx5e_cq *cq) { @@ -1747,7 +1604,7 @@ static int mlx5e_open_channel(struct mlx5e_priv *priv, int ix, struct mlx5e_channel_param *cparam, struct mlx5e_channel **cp) { - struct mlx5e_cq_moder icocq_moder = {0, 0}; + struct net_dim_cq_moder icocq_moder = {0, 0}; struct net_device *netdev = priv->netdev; int cpu = mlx5e_get_cpu(priv, ix); struct mlx5e_channel *c; @@ -1999,7 +1856,7 @@ static void mlx5e_build_ico_cq_param(struct mlx5e_priv *priv, mlx5e_build_common_cq_param(priv, param); - param->cq_period_mode = MLX5_CQ_PERIOD_MODE_START_FROM_EQE; + param->cq_period_mode = NET_DIM_CQ_PERIOD_MODE_START_FROM_EQE; } static void mlx5e_build_icosq_param(struct mlx5e_priv *priv, @@ -2203,7 +2060,7 @@ static int mlx5e_rx_hash_fn(int hfunc) MLX5_RX_HASH_FN_INVERTED_XOR8; } -static int mlx5e_bits_invert(unsigned long a, int size) +int mlx5e_bits_invert(unsigned long a, int size) { int inv = 0; int i; @@ -2765,6 +2622,9 @@ static int mlx5e_alloc_drop_rq(struct mlx5_core_dev *mdev, if (err) return err; + /* Mark as unused given "Drop-RQ" packets never reach XDP */ + xdp_rxq_info_unused(&rq->xdp_rxq); + rq->mdev = mdev; return 0; @@ -3084,9 +2944,6 @@ out: static int mlx5e_setup_tc_cls_flower(struct mlx5e_priv *priv, struct tc_cls_flower_offload *cls_flower) { - if (cls_flower->common.chain_index) - return -EOPNOTSUPP; - switch (cls_flower->command) { case TC_CLSFLOWER_REPLACE: return mlx5e_configure_flower(priv, cls_flower); @@ -3104,7 +2961,7 @@ int mlx5e_setup_tc_block_cb(enum tc_setup_type type, void *type_data, { struct mlx5e_priv *priv = cb_priv; - if (!tc_can_offload(priv->netdev)) + if (!tc_cls_can_offload_and_chain0(priv->netdev, type_data)) return -EOPNOTSUPP; switch (type) { @@ -3738,26 +3595,62 @@ static netdev_features_t mlx5e_features_check(struct sk_buff *skb, return features; } +static bool mlx5e_tx_timeout_eq_recover(struct net_device *dev, + struct mlx5e_txqsq *sq) +{ + struct mlx5e_priv *priv = netdev_priv(dev); + struct mlx5_core_dev *mdev = priv->mdev; + int irqn_not_used, eqn; + struct mlx5_eq *eq; + u32 eqe_count; + + if (mlx5_vector2eqn(mdev, sq->cq.mcq.vector, &eqn, &irqn_not_used)) + return false; + + eq = mlx5_eqn2eq(mdev, eqn); + if (IS_ERR(eq)) + return false; + + netdev_err(dev, "EQ 0x%x: Cons = 0x%x, irqn = 0x%x\n", + eqn, eq->cons_index, eq->irqn); + + eqe_count = mlx5_eq_poll_irq_disabled(eq); + if (!eqe_count) + return false; + + netdev_err(dev, "Recover %d eqes on EQ 0x%x\n", eqe_count, eq->eqn); + sq->channel->stats.eq_rearm++; + return true; +} + static void mlx5e_tx_timeout(struct net_device *dev) { struct mlx5e_priv *priv = netdev_priv(dev); - bool sched_work = false; + bool reopen_channels = false; int i; netdev_err(dev, "TX timeout detected\n"); for (i = 0; i < priv->channels.num * priv->channels.params.num_tc; i++) { + struct netdev_queue *dev_queue = netdev_get_tx_queue(dev, i); struct mlx5e_txqsq *sq = priv->txq2sq[i]; - if (!netif_xmit_stopped(netdev_get_tx_queue(dev, i))) + if (!netif_xmit_stopped(dev_queue)) continue; - sched_work = true; - clear_bit(MLX5E_SQ_STATE_ENABLED, &sq->state); - netdev_err(dev, "TX timeout on queue: %d, SQ: 0x%x, CQ: 0x%x, SQ Cons: 0x%x SQ Prod: 0x%x\n", - i, sq->sqn, sq->cq.mcq.cqn, sq->cc, sq->pc); + netdev_err(dev, "TX timeout on queue: %d, SQ: 0x%x, CQ: 0x%x, SQ Cons: 0x%x SQ Prod: 0x%x, usecs since last trans: %u\n", + i, sq->sqn, sq->cq.mcq.cqn, sq->cc, sq->pc, + jiffies_to_usecs(jiffies - dev_queue->trans_start)); + + /* If we recover a lost interrupt, most likely TX timeout will + * be resolved, skip reopening channels + */ + if (!mlx5e_tx_timeout_eq_recover(dev, sq)) { + clear_bit(MLX5E_SQ_STATE_ENABLED, &sq->state); + reopen_channels = true; + } } - if (sched_work && test_bit(MLX5E_STATE_OPENED, &priv->state)) + if (reopen_channels && test_bit(MLX5E_STATE_OPENED, &priv->state)) schedule_work(&priv->tx_timeout_work); } @@ -4044,9 +3937,18 @@ void mlx5e_set_rx_cq_mode_params(struct mlx5e_params *params, u8 cq_period_mode) params->rx_cq_moderation.usec = MLX5E_PARAMS_DEFAULT_RX_CQ_MODERATION_USEC_FROM_CQE; - if (params->rx_am_enabled) - params->rx_cq_moderation = - mlx5e_am_get_def_profile(cq_period_mode); + if (params->rx_dim_enabled) { + switch (cq_period_mode) { + case MLX5_CQ_PERIOD_MODE_START_FROM_CQE: + params->rx_cq_moderation = + net_dim_get_def_profile(NET_DIM_CQ_PERIOD_MODE_START_FROM_CQE); + break; + case MLX5_CQ_PERIOD_MODE_START_FROM_EQE: + default: + params->rx_cq_moderation = + net_dim_get_def_profile(NET_DIM_CQ_PERIOD_MODE_START_FROM_EQE); + } + } MLX5E_SET_PFLAG(params, MLX5E_PFLAG_RX_CQE_BASED_MODER, params->rx_cq_moderation.cq_period_mode == @@ -4108,7 +4010,7 @@ void mlx5e_build_nic_params(struct mlx5_core_dev *mdev, cq_period_mode = MLX5_CAP_GEN(mdev, cq_period_start_from_cqe) ? MLX5_CQ_PERIOD_MODE_START_FROM_CQE : MLX5_CQ_PERIOD_MODE_START_FROM_EQE; - params->rx_am_enabled = MLX5_CAP_GEN(mdev, cq_moderation); + params->rx_dim_enabled = MLX5_CAP_GEN(mdev, cq_moderation); mlx5e_set_rx_cq_mode_params(params, cq_period_mode); mlx5e_set_tx_cq_mode_params(params, cq_period_mode); @@ -4315,9 +4217,6 @@ static void mlx5e_nic_cleanup(struct mlx5e_priv *priv) { mlx5e_ipsec_cleanup(priv); mlx5e_vxlan_cleanup(priv); - - if (priv->channels.params.xdp_prog) - bpf_prog_put(priv->channels.params.xdp_prog); } static int mlx5e_init_nic_rx(struct mlx5e_priv *priv) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c index 3409d86..363d8dc 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c @@ -190,6 +190,63 @@ int mlx5e_attr_get(struct net_device *dev, struct switchdev_attr *attr) return 0; } +static void mlx5e_sqs2vport_stop(struct mlx5_eswitch *esw, + struct mlx5_eswitch_rep *rep) +{ + struct mlx5e_rep_sq *rep_sq, *tmp; + struct mlx5e_rep_priv *rpriv; + + if (esw->mode != SRIOV_OFFLOADS) + return; + + rpriv = mlx5e_rep_to_rep_priv(rep); + list_for_each_entry_safe(rep_sq, tmp, &rpriv->vport_sqs_list, list) { + mlx5_eswitch_del_send_to_vport_rule(rep_sq->send_to_vport_rule); + list_del(&rep_sq->list); + kfree(rep_sq); + } +} + +static int mlx5e_sqs2vport_start(struct mlx5_eswitch *esw, + struct mlx5_eswitch_rep *rep, + u16 *sqns_array, int sqns_num) +{ + struct mlx5_flow_handle *flow_rule; + struct mlx5e_rep_priv *rpriv; + struct mlx5e_rep_sq *rep_sq; + int err; + int i; + + if (esw->mode != SRIOV_OFFLOADS) + return 0; + + rpriv = mlx5e_rep_to_rep_priv(rep); + for (i = 0; i < sqns_num; i++) { + rep_sq = kzalloc(sizeof(*rep_sq), GFP_KERNEL); + if (!rep_sq) { + err = -ENOMEM; + goto out_err; + } + + /* Add re-inject rule to the PF/representor sqs */ + flow_rule = mlx5_eswitch_add_send_to_vport_rule(esw, + rep->vport, + sqns_array[i]); + if (IS_ERR(flow_rule)) { + err = PTR_ERR(flow_rule); + kfree(rep_sq); + goto out_err; + } + rep_sq->send_to_vport_rule = flow_rule; + list_add(&rep_sq->list, &rpriv->vport_sqs_list); + } + return 0; + +out_err: + mlx5e_sqs2vport_stop(esw, rep); + return err; +} + int mlx5e_add_sqs_fwd_rules(struct mlx5e_priv *priv) { struct mlx5_eswitch *esw = priv->mdev->priv.eswitch; @@ -210,7 +267,7 @@ int mlx5e_add_sqs_fwd_rules(struct mlx5e_priv *priv) sqs[num_sqs++] = c->sq[tc].sqn; } - err = mlx5_eswitch_sqs2vport_start(esw, rep, sqs, num_sqs); + err = mlx5e_sqs2vport_start(esw, rep, sqs, num_sqs); kfree(sqs); out: @@ -225,7 +282,7 @@ void mlx5e_remove_sqs_fwd_rules(struct mlx5e_priv *priv) struct mlx5e_rep_priv *rpriv = priv->ppriv; struct mlx5_eswitch_rep *rep = rpriv->rep; - mlx5_eswitch_sqs2vport_stop(esw, rep); + mlx5e_sqs2vport_stop(esw, rep); } static void mlx5e_rep_neigh_update_init_interval(struct mlx5e_rep_priv *rpriv) @@ -238,7 +295,7 @@ static void mlx5e_rep_neigh_update_init_interval(struct mlx5e_rep_priv *rpriv) #endif unsigned long ipv4_interval = NEIGH_VAR(&arp_tbl.parms, DELAY_PROBE_TIME); - struct net_device *netdev = rpriv->rep->netdev; + struct net_device *netdev = rpriv->netdev; struct mlx5e_priv *priv = netdev_priv(netdev); rpriv->neigh_update.min_interval = min_t(unsigned long, ipv6_interval, ipv4_interval); @@ -259,7 +316,7 @@ static void mlx5e_rep_neigh_stats_work(struct work_struct *work) { struct mlx5e_rep_priv *rpriv = container_of(work, struct mlx5e_rep_priv, neigh_update.neigh_stats_work.work); - struct net_device *netdev = rpriv->rep->netdev; + struct net_device *netdev = rpriv->netdev; struct mlx5e_priv *priv = netdev_priv(netdev); struct mlx5e_neigh_hash_entry *nhe; @@ -355,7 +412,7 @@ static int mlx5e_rep_netevent_event(struct notifier_block *nb, struct mlx5e_rep_priv *rpriv = container_of(nb, struct mlx5e_rep_priv, neigh_update.netevent_nb); struct mlx5e_neigh_update_table *neigh_update = &rpriv->neigh_update; - struct net_device *netdev = rpriv->rep->netdev; + struct net_device *netdev = rpriv->netdev; struct mlx5e_priv *priv = netdev_priv(netdev); struct mlx5e_neigh_hash_entry *nhe = NULL; struct mlx5e_neigh m_neigh = {}; @@ -483,7 +540,7 @@ out_err: static void mlx5e_rep_neigh_cleanup(struct mlx5e_rep_priv *rpriv) { struct mlx5e_neigh_update_table *neigh_update = &rpriv->neigh_update; - struct mlx5e_priv *priv = netdev_priv(rpriv->rep->netdev); + struct mlx5e_priv *priv = netdev_priv(rpriv->netdev); unregister_netevent_notifier(&neigh_update->netevent_nb); @@ -662,9 +719,6 @@ static int mlx5e_rep_setup_tc_cls_flower(struct mlx5e_priv *priv, struct tc_cls_flower_offload *cls_flower) { - if (cls_flower->common.chain_index) - return -EOPNOTSUPP; - switch (cls_flower->command) { case TC_CLSFLOWER_REPLACE: return mlx5e_configure_flower(priv, cls_flower); @@ -682,7 +736,7 @@ static int mlx5e_rep_setup_tc_cb(enum tc_setup_type type, void *type_data, { struct mlx5e_priv *priv = cb_priv; - if (!tc_can_offload(priv->netdev)) + if (!tc_cls_can_offload_and_chain0(priv->netdev, type_data)) return -EOPNOTSUPP; switch (type) { @@ -827,7 +881,7 @@ static void mlx5e_build_rep_params(struct mlx5_core_dev *mdev, params->rq_wq_type = MLX5_WQ_TYPE_LINKED_LIST; params->log_rq_size = MLX5E_PARAMS_MINIMUM_LOG_RQ_SIZE; - params->rx_am_enabled = MLX5_CAP_GEN(mdev, cq_moderation); + params->rx_dim_enabled = MLX5_CAP_GEN(mdev, cq_moderation); mlx5e_set_rx_cq_mode_params(params, cq_period_mode); params->tx_max_inline = mlx5e_get_max_inline_cap(mdev); @@ -906,7 +960,7 @@ static int mlx5e_init_rep_rx(struct mlx5e_priv *priv) err = PTR_ERR(flow_rule); goto err_destroy_direct_tirs; } - rep->vport_rx_rule = flow_rule; + rpriv->vport_rx_rule = flow_rule; err = mlx5e_tc_init(priv); if (err) @@ -915,7 +969,7 @@ static int mlx5e_init_rep_rx(struct mlx5e_priv *priv) return 0; err_del_flow_rule: - mlx5_del_flow_rules(rep->vport_rx_rule); + mlx5_del_flow_rules(rpriv->vport_rx_rule); err_destroy_direct_tirs: mlx5e_destroy_direct_tirs(priv); err_destroy_direct_rqts: @@ -926,10 +980,9 @@ err_destroy_direct_rqts: static void mlx5e_cleanup_rep_rx(struct mlx5e_priv *priv) { struct mlx5e_rep_priv *rpriv = priv->ppriv; - struct mlx5_eswitch_rep *rep = rpriv->rep; mlx5e_tc_cleanup(priv); - mlx5_del_flow_rules(rep->vport_rx_rule); + mlx5_del_flow_rules(rpriv->vport_rx_rule); mlx5e_destroy_direct_tirs(priv); mlx5e_destroy_direct_rqts(priv); } @@ -969,10 +1022,10 @@ static const struct mlx5e_profile mlx5e_rep_profile = { /* e-Switch vport representors */ static int -mlx5e_nic_rep_load(struct mlx5_eswitch *esw, struct mlx5_eswitch_rep *rep) +mlx5e_nic_rep_load(struct mlx5_core_dev *dev, struct mlx5_eswitch_rep *rep) { - struct mlx5e_priv *priv = netdev_priv(rep->netdev); - struct mlx5e_rep_priv *rpriv = priv->ppriv; + struct mlx5e_rep_priv *rpriv = mlx5e_rep_to_rep_priv(rep); + struct mlx5e_priv *priv = netdev_priv(rpriv->netdev); int err; @@ -994,10 +1047,10 @@ err_remove_sqs: } static void -mlx5e_nic_rep_unload(struct mlx5_eswitch *esw, struct mlx5_eswitch_rep *rep) +mlx5e_nic_rep_unload(struct mlx5_eswitch_rep *rep) { - struct mlx5e_priv *priv = netdev_priv(rep->netdev); - struct mlx5e_rep_priv *rpriv = priv->ppriv; + struct mlx5e_rep_priv *rpriv = mlx5e_rep_to_rep_priv(rep); + struct mlx5e_priv *priv = netdev_priv(rpriv->netdev); if (test_bit(MLX5E_STATE_OPENED, &priv->state)) mlx5e_remove_sqs_fwd_rules(priv); @@ -1010,8 +1063,9 @@ mlx5e_nic_rep_unload(struct mlx5_eswitch *esw, struct mlx5_eswitch_rep *rep) } static int -mlx5e_vport_rep_load(struct mlx5_eswitch *esw, struct mlx5_eswitch_rep *rep) +mlx5e_vport_rep_load(struct mlx5_core_dev *dev, struct mlx5_eswitch_rep *rep) { + struct mlx5e_rep_priv *uplink_rpriv; struct mlx5e_rep_priv *rpriv; struct net_device *netdev; struct mlx5e_priv *upriv; @@ -1021,7 +1075,7 @@ mlx5e_vport_rep_load(struct mlx5_eswitch *esw, struct mlx5_eswitch_rep *rep) if (!rpriv) return -ENOMEM; - netdev = mlx5e_create_netdev(esw->dev, &mlx5e_rep_profile, rpriv); + netdev = mlx5e_create_netdev(dev, &mlx5e_rep_profile, rpriv); if (!netdev) { pr_warn("Failed to create representor netdev for vport %d\n", rep->vport); @@ -1029,8 +1083,10 @@ mlx5e_vport_rep_load(struct mlx5_eswitch *esw, struct mlx5_eswitch_rep *rep) return -EINVAL; } - rep->netdev = netdev; + rpriv->netdev = netdev; rpriv->rep = rep; + rep->rep_if[REP_ETH].priv = rpriv; + INIT_LIST_HEAD(&rpriv->vport_sqs_list); err = mlx5e_attach_netdev(netdev_priv(netdev)); if (err) { @@ -1046,7 +1102,8 @@ mlx5e_vport_rep_load(struct mlx5_eswitch *esw, struct mlx5_eswitch_rep *rep) goto err_detach_netdev; } - upriv = netdev_priv(mlx5_eswitch_get_uplink_netdev(esw)); + uplink_rpriv = mlx5_eswitch_get_uplink_priv(dev->priv.eswitch, REP_ETH); + upriv = netdev_priv(uplink_rpriv->netdev); err = tc_setup_cb_egdev_register(netdev, mlx5e_setup_tc_block_cb, upriv); if (err) @@ -1078,16 +1135,19 @@ err_destroy_netdev: } static void -mlx5e_vport_rep_unload(struct mlx5_eswitch *esw, struct mlx5_eswitch_rep *rep) +mlx5e_vport_rep_unload(struct mlx5_eswitch_rep *rep) { - struct net_device *netdev = rep->netdev; + struct mlx5e_rep_priv *rpriv = mlx5e_rep_to_rep_priv(rep); + struct net_device *netdev = rpriv->netdev; struct mlx5e_priv *priv = netdev_priv(netdev); - struct mlx5e_rep_priv *rpriv = priv->ppriv; + struct mlx5e_rep_priv *uplink_rpriv; void *ppriv = priv->ppriv; struct mlx5e_priv *upriv; - unregister_netdev(rep->netdev); - upriv = netdev_priv(mlx5_eswitch_get_uplink_netdev(esw)); + unregister_netdev(netdev); + uplink_rpriv = mlx5_eswitch_get_uplink_priv(priv->mdev->priv.eswitch, + REP_ETH); + upriv = netdev_priv(uplink_rpriv->netdev); tc_setup_cb_egdev_unregister(netdev, mlx5e_setup_tc_block_cb, upriv); mlx5e_rep_neigh_cleanup(rpriv); @@ -1102,18 +1162,13 @@ static void mlx5e_rep_register_vf_vports(struct mlx5e_priv *priv) struct mlx5_eswitch *esw = mdev->priv.eswitch; int total_vfs = MLX5_TOTAL_VPORTS(mdev); int vport; - u8 mac[ETH_ALEN]; - - mlx5_query_nic_vport_mac_address(mdev, 0, mac); for (vport = 1; vport < total_vfs; vport++) { - struct mlx5_eswitch_rep rep; + struct mlx5_eswitch_rep_if rep_if = {}; - rep.load = mlx5e_vport_rep_load; - rep.unload = mlx5e_vport_rep_unload; - rep.vport = vport; - ether_addr_copy(rep.hw_id, mac); - mlx5_eswitch_register_vport_rep(esw, vport, &rep); + rep_if.load = mlx5e_vport_rep_load; + rep_if.unload = mlx5e_vport_rep_unload; + mlx5_eswitch_register_vport_rep(esw, vport, &rep_if, REP_ETH); } } @@ -1125,21 +1180,24 @@ static void mlx5e_rep_unregister_vf_vports(struct mlx5e_priv *priv) int vport; for (vport = 1; vport < total_vfs; vport++) - mlx5_eswitch_unregister_vport_rep(esw, vport); + mlx5_eswitch_unregister_vport_rep(esw, vport, REP_ETH); } void mlx5e_register_vport_reps(struct mlx5e_priv *priv) { struct mlx5_core_dev *mdev = priv->mdev; struct mlx5_eswitch *esw = mdev->priv.eswitch; - struct mlx5_eswitch_rep rep; + struct mlx5_eswitch_rep_if rep_if; + struct mlx5e_rep_priv *rpriv; + + rpriv = priv->ppriv; + rpriv->netdev = priv->netdev; - mlx5_query_nic_vport_mac_address(mdev, 0, rep.hw_id); - rep.load = mlx5e_nic_rep_load; - rep.unload = mlx5e_nic_rep_unload; - rep.vport = FDB_UPLINK_VPORT; - rep.netdev = priv->netdev; - mlx5_eswitch_register_vport_rep(esw, 0, &rep); /* UPLINK PF vport*/ + rep_if.load = mlx5e_nic_rep_load; + rep_if.unload = mlx5e_nic_rep_unload; + rep_if.priv = rpriv; + INIT_LIST_HEAD(&rpriv->vport_sqs_list); + mlx5_eswitch_register_vport_rep(esw, 0, &rep_if, REP_ETH); /* UPLINK PF vport*/ mlx5e_rep_register_vf_vports(priv); /* VFs vports */ } @@ -1150,7 +1208,7 @@ void mlx5e_unregister_vport_reps(struct mlx5e_priv *priv) struct mlx5_eswitch *esw = mdev->priv.eswitch; mlx5e_rep_unregister_vf_vports(priv); /* VFs vports */ - mlx5_eswitch_unregister_vport_rep(esw, 0); /* UPLINK PF*/ + mlx5_eswitch_unregister_vport_rep(esw, 0, REP_ETH); /* UPLINK PF*/ } void *mlx5e_alloc_nic_rep_priv(struct mlx5_core_dev *mdev) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h index 5659ed9..b9b481f 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h @@ -56,8 +56,17 @@ struct mlx5e_neigh_update_table { struct mlx5e_rep_priv { struct mlx5_eswitch_rep *rep; struct mlx5e_neigh_update_table neigh_update; + struct net_device *netdev; + struct mlx5_flow_handle *vport_rx_rule; + struct list_head vport_sqs_list; }; +static inline +struct mlx5e_rep_priv *mlx5e_rep_to_rep_priv(struct mlx5_eswitch_rep *rep) +{ + return (struct mlx5e_rep_priv *)rep->rep_if[REP_ETH].priv; +} + struct mlx5e_neigh { struct net_device *dev; union { @@ -124,6 +133,11 @@ struct mlx5e_encap_entry { int encap_size; }; +struct mlx5e_rep_sq { + struct mlx5_flow_handle *send_to_vport_rule; + struct list_head list; +}; + void *mlx5e_alloc_nic_rep_priv(struct mlx5_core_dev *mdev); void mlx5e_register_vport_reps(struct mlx5e_priv *priv); void mlx5e_unregister_vport_reps(struct mlx5e_priv *priv); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c index 5b499c7..0d4bb06 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c @@ -495,8 +495,8 @@ static inline void mlx5e_poll_ico_single_cqe(struct mlx5e_cq *cq, mlx5_cqwq_pop(&cq->wq); if (unlikely((cqe->op_own >> 4) != MLX5_CQE_REQ)) { - WARN_ONCE(true, "mlx5e: Bad OP in ICOSQ CQE: 0x%x\n", - cqe->op_own); + netdev_WARN_ONCE(cq->channel->netdev, + "Bad OP in ICOSQ CQE: 0x%x\n", cqe->op_own); return; } @@ -506,9 +506,8 @@ static inline void mlx5e_poll_ico_single_cqe(struct mlx5e_cq *cq, } if (unlikely(icowi->opcode != MLX5_OPCODE_NOP)) - WARN_ONCE(true, - "mlx5e: Bad OPCODE in ICOSQ WQE info: 0x%x\n", - icowi->opcode); + netdev_WARN_ONCE(cq->channel->netdev, + "Bad OPCODE in ICOSQ WQE info: 0x%x\n", icowi->opcode); } static void mlx5e_poll_ico_cq(struct mlx5e_cq *cq, struct mlx5e_rq *rq) @@ -632,7 +631,7 @@ static inline void mlx5e_handle_csum(struct net_device *netdev, return; } - if (is_last_ethertype_ip(skb, &network_depth)) { + if (likely(is_last_ethertype_ip(skb, &network_depth))) { skb->ip_summed = CHECKSUM_COMPLETE; skb->csum = csum_unfold((__force __sum16)cqe->check_sum); if (network_depth > ETH_HLEN) @@ -812,6 +811,7 @@ static inline int mlx5e_xdp_handle(struct mlx5e_rq *rq, xdp_set_data_meta_invalid(&xdp); xdp.data_end = xdp.data + *len; xdp.data_hard_start = va; + xdp.rxq = &rq->xdp_rxq; act = bpf_prog_run_xdp(prog, &xdp); switch (act) { @@ -1175,7 +1175,9 @@ static inline void mlx5i_complete_rx_cqe(struct mlx5e_rq *rq, u32 cqe_bcnt, struct sk_buff *skb) { + struct hwtstamp_config *tstamp; struct net_device *netdev; + struct mlx5e_priv *priv; char *pseudo_header; u32 qpn; u8 *dgid; @@ -1194,6 +1196,9 @@ static inline void mlx5i_complete_rx_cqe(struct mlx5e_rq *rq, return; } + priv = mlx5i_epriv(netdev); + tstamp = &priv->tstamp; + g = (be32_to_cpu(cqe->flags_rqpn) >> 28) & 3; dgid = skb->data + MLX5_IB_GRH_DGID_OFFSET; if ((!g) || dgid[0] != 0xff) @@ -1214,7 +1219,7 @@ static inline void mlx5i_complete_rx_cqe(struct mlx5e_rq *rq, skb->ip_summed = CHECKSUM_COMPLETE; skb->csum = csum_unfold((__force __sum16)cqe->check_sum); - if (unlikely(mlx5e_rx_hw_stamp(rq->tstamp))) + if (unlikely(mlx5e_rx_hw_stamp(tstamp))) skb_hwtstamps(skb)->hwtstamp = mlx5_timecounter_cyc2time(rq->clock, get_cqe_ts(cqe)); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rx_am.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rx_am.c deleted file mode 100644 index b69a705..0000000 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_rx_am.c +++ /dev/null @@ -1,347 +0,0 @@ -/* - * Copyright (c) 2016, Mellanox Technologies. All rights reserved. - * - * This software is available to you under a choice of one of two - * licenses. You may choose to be licensed under the terms of the GNU - * General Public License (GPL) Version 2, available from the file - * COPYING in the main directory of this source tree, or the - * OpenIB.org BSD license below: - * - * Redistribution and use in source and binary forms, with or - * without modification, are permitted provided that the following - * conditions are met: - * - * - Redistributions of source code must retain the above - * copyright notice, this list of conditions and the following - * disclaimer. - * - * - Redistributions in binary form must reproduce the above - * copyright notice, this list of conditions and the following - * disclaimer in the documentation and/or other materials - * provided with the distribution. - * - * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, - * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF - * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND - * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS - * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN - * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN - * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE - * SOFTWARE. - */ - -#include "en.h" - -/* Adaptive moderation profiles */ -#define MLX5E_AM_DEFAULT_RX_CQ_MODERATION_PKTS_FROM_EQE 256 -#define MLX5E_RX_AM_DEF_PROFILE_CQE 1 -#define MLX5E_RX_AM_DEF_PROFILE_EQE 1 -#define MLX5E_PARAMS_AM_NUM_PROFILES 5 - -/* All profiles sizes must be MLX5E_PARAMS_AM_NUM_PROFILES */ -#define MLX5_AM_EQE_PROFILES { \ - {1, MLX5E_AM_DEFAULT_RX_CQ_MODERATION_PKTS_FROM_EQE}, \ - {8, MLX5E_AM_DEFAULT_RX_CQ_MODERATION_PKTS_FROM_EQE}, \ - {64, MLX5E_AM_DEFAULT_RX_CQ_MODERATION_PKTS_FROM_EQE}, \ - {128, MLX5E_AM_DEFAULT_RX_CQ_MODERATION_PKTS_FROM_EQE}, \ - {256, MLX5E_AM_DEFAULT_RX_CQ_MODERATION_PKTS_FROM_EQE}, \ -} - -#define MLX5_AM_CQE_PROFILES { \ - {2, 256}, \ - {8, 128}, \ - {16, 64}, \ - {32, 64}, \ - {64, 64} \ -} - -static const struct mlx5e_cq_moder -profile[MLX5_CQ_PERIOD_NUM_MODES][MLX5E_PARAMS_AM_NUM_PROFILES] = { - MLX5_AM_EQE_PROFILES, - MLX5_AM_CQE_PROFILES, -}; - -static inline struct mlx5e_cq_moder mlx5e_am_get_profile(u8 cq_period_mode, int ix) -{ - struct mlx5e_cq_moder cq_moder; - - cq_moder = profile[cq_period_mode][ix]; - cq_moder.cq_period_mode = cq_period_mode; - return cq_moder; -} - -struct mlx5e_cq_moder mlx5e_am_get_def_profile(u8 rx_cq_period_mode) -{ - int default_profile_ix; - - if (rx_cq_period_mode == MLX5_CQ_PERIOD_MODE_START_FROM_CQE) - default_profile_ix = MLX5E_RX_AM_DEF_PROFILE_CQE; - else /* MLX5_CQ_PERIOD_MODE_START_FROM_EQE */ - default_profile_ix = MLX5E_RX_AM_DEF_PROFILE_EQE; - - return mlx5e_am_get_profile(rx_cq_period_mode, default_profile_ix); -} - -/* Adaptive moderation logic */ -enum { - MLX5E_AM_START_MEASURE, - MLX5E_AM_MEASURE_IN_PROGRESS, - MLX5E_AM_APPLY_NEW_PROFILE, -}; - -enum { - MLX5E_AM_PARKING_ON_TOP, - MLX5E_AM_PARKING_TIRED, - MLX5E_AM_GOING_RIGHT, - MLX5E_AM_GOING_LEFT, -}; - -enum { - MLX5E_AM_STATS_WORSE, - MLX5E_AM_STATS_SAME, - MLX5E_AM_STATS_BETTER, -}; - -enum { - MLX5E_AM_STEPPED, - MLX5E_AM_TOO_TIRED, - MLX5E_AM_ON_EDGE, -}; - -static bool mlx5e_am_on_top(struct mlx5e_rx_am *am) -{ - switch (am->tune_state) { - case MLX5E_AM_PARKING_ON_TOP: - case MLX5E_AM_PARKING_TIRED: - return true; - case MLX5E_AM_GOING_RIGHT: - return (am->steps_left > 1) && (am->steps_right == 1); - default: /* MLX5E_AM_GOING_LEFT */ - return (am->steps_right > 1) && (am->steps_left == 1); - } -} - -static void mlx5e_am_turn(struct mlx5e_rx_am *am) -{ - switch (am->tune_state) { - case MLX5E_AM_PARKING_ON_TOP: - case MLX5E_AM_PARKING_TIRED: - break; - case MLX5E_AM_GOING_RIGHT: - am->tune_state = MLX5E_AM_GOING_LEFT; - am->steps_left = 0; - break; - case MLX5E_AM_GOING_LEFT: - am->tune_state = MLX5E_AM_GOING_RIGHT; - am->steps_right = 0; - break; - } -} - -static int mlx5e_am_step(struct mlx5e_rx_am *am) -{ - if (am->tired == (MLX5E_PARAMS_AM_NUM_PROFILES * 2)) - return MLX5E_AM_TOO_TIRED; - - switch (am->tune_state) { - case MLX5E_AM_PARKING_ON_TOP: - case MLX5E_AM_PARKING_TIRED: - break; - case MLX5E_AM_GOING_RIGHT: - if (am->profile_ix == (MLX5E_PARAMS_AM_NUM_PROFILES - 1)) - return MLX5E_AM_ON_EDGE; - am->profile_ix++; - am->steps_right++; - break; - case MLX5E_AM_GOING_LEFT: - if (am->profile_ix == 0) - return MLX5E_AM_ON_EDGE; - am->profile_ix--; - am->steps_left++; - break; - } - - am->tired++; - return MLX5E_AM_STEPPED; -} - -static void mlx5e_am_park_on_top(struct mlx5e_rx_am *am) -{ - am->steps_right = 0; - am->steps_left = 0; - am->tired = 0; - am->tune_state = MLX5E_AM_PARKING_ON_TOP; -} - -static void mlx5e_am_park_tired(struct mlx5e_rx_am *am) -{ - am->steps_right = 0; - am->steps_left = 0; - am->tune_state = MLX5E_AM_PARKING_TIRED; -} - -static void mlx5e_am_exit_parking(struct mlx5e_rx_am *am) -{ - am->tune_state = am->profile_ix ? MLX5E_AM_GOING_LEFT : - MLX5E_AM_GOING_RIGHT; - mlx5e_am_step(am); -} - -#define IS_SIGNIFICANT_DIFF(val, ref) \ - (((100 * abs((val) - (ref))) / (ref)) > 10) /* more than 10% difference */ - -static int mlx5e_am_stats_compare(struct mlx5e_rx_am_stats *curr, - struct mlx5e_rx_am_stats *prev) -{ - if (!prev->bpms) - return curr->bpms ? MLX5E_AM_STATS_BETTER : - MLX5E_AM_STATS_SAME; - - if (IS_SIGNIFICANT_DIFF(curr->bpms, prev->bpms)) - return (curr->bpms > prev->bpms) ? MLX5E_AM_STATS_BETTER : - MLX5E_AM_STATS_WORSE; - - if (!prev->ppms) - return curr->ppms ? MLX5E_AM_STATS_BETTER : - MLX5E_AM_STATS_SAME; - - if (IS_SIGNIFICANT_DIFF(curr->ppms, prev->ppms)) - return (curr->ppms > prev->ppms) ? MLX5E_AM_STATS_BETTER : - MLX5E_AM_STATS_WORSE; - if (!prev->epms) - return MLX5E_AM_STATS_SAME; - - if (IS_SIGNIFICANT_DIFF(curr->epms, prev->epms)) - return (curr->epms < prev->epms) ? MLX5E_AM_STATS_BETTER : - MLX5E_AM_STATS_WORSE; - - return MLX5E_AM_STATS_SAME; -} - -static bool mlx5e_am_decision(struct mlx5e_rx_am_stats *curr_stats, - struct mlx5e_rx_am *am) -{ - int prev_state = am->tune_state; - int prev_ix = am->profile_ix; - int stats_res; - int step_res; - - switch (am->tune_state) { - case MLX5E_AM_PARKING_ON_TOP: - stats_res = mlx5e_am_stats_compare(curr_stats, &am->prev_stats); - if (stats_res != MLX5E_AM_STATS_SAME) - mlx5e_am_exit_parking(am); - break; - - case MLX5E_AM_PARKING_TIRED: - am->tired--; - if (!am->tired) - mlx5e_am_exit_parking(am); - break; - - case MLX5E_AM_GOING_RIGHT: - case MLX5E_AM_GOING_LEFT: - stats_res = mlx5e_am_stats_compare(curr_stats, &am->prev_stats); - if (stats_res != MLX5E_AM_STATS_BETTER) - mlx5e_am_turn(am); - - if (mlx5e_am_on_top(am)) { - mlx5e_am_park_on_top(am); - break; - } - - step_res = mlx5e_am_step(am); - switch (step_res) { - case MLX5E_AM_ON_EDGE: - mlx5e_am_park_on_top(am); - break; - case MLX5E_AM_TOO_TIRED: - mlx5e_am_park_tired(am); - break; - } - - break; - } - - if ((prev_state != MLX5E_AM_PARKING_ON_TOP) || - (am->tune_state != MLX5E_AM_PARKING_ON_TOP)) - am->prev_stats = *curr_stats; - - return am->profile_ix != prev_ix; -} - -static void mlx5e_am_sample(struct mlx5e_rq *rq, - struct mlx5e_rx_am_sample *s) -{ - s->time = ktime_get(); - s->pkt_ctr = rq->stats.packets; - s->byte_ctr = rq->stats.bytes; - s->event_ctr = rq->cq.event_ctr; -} - -#define MLX5E_AM_NEVENTS 64 -#define BITS_PER_TYPE(type) (sizeof(type) * BITS_PER_BYTE) -#define BIT_GAP(bits, end, start) ((((end) - (start)) + BIT_ULL(bits)) & (BIT_ULL(bits) - 1)) - -static void mlx5e_am_calc_stats(struct mlx5e_rx_am_sample *start, - struct mlx5e_rx_am_sample *end, - struct mlx5e_rx_am_stats *curr_stats) -{ - /* u32 holds up to 71 minutes, should be enough */ - u32 delta_us = ktime_us_delta(end->time, start->time); - u32 npkts = BIT_GAP(BITS_PER_TYPE(u32), end->pkt_ctr, start->pkt_ctr); - u32 nbytes = BIT_GAP(BITS_PER_TYPE(u32), end->byte_ctr, - start->byte_ctr); - - if (!delta_us) - return; - - curr_stats->ppms = DIV_ROUND_UP(npkts * USEC_PER_MSEC, delta_us); - curr_stats->bpms = DIV_ROUND_UP(nbytes * USEC_PER_MSEC, delta_us); - curr_stats->epms = DIV_ROUND_UP(MLX5E_AM_NEVENTS * USEC_PER_MSEC, - delta_us); -} - -void mlx5e_rx_am_work(struct work_struct *work) -{ - struct mlx5e_rx_am *am = container_of(work, struct mlx5e_rx_am, - work); - struct mlx5e_rq *rq = container_of(am, struct mlx5e_rq, am); - struct mlx5e_cq_moder cur_profile = profile[am->mode][am->profile_ix]; - - mlx5_core_modify_cq_moderation(rq->mdev, &rq->cq.mcq, - cur_profile.usec, cur_profile.pkts); - - am->state = MLX5E_AM_START_MEASURE; -} - -void mlx5e_rx_am(struct mlx5e_rq *rq) -{ - struct mlx5e_rx_am *am = &rq->am; - struct mlx5e_rx_am_sample end_sample; - struct mlx5e_rx_am_stats curr_stats; - u16 nevents; - - switch (am->state) { - case MLX5E_AM_MEASURE_IN_PROGRESS: - nevents = BIT_GAP(BITS_PER_TYPE(u16), rq->cq.event_ctr, - am->start_sample.event_ctr); - if (nevents < MLX5E_AM_NEVENTS) - break; - mlx5e_am_sample(rq, &end_sample); - mlx5e_am_calc_stats(&am->start_sample, &end_sample, - &curr_stats); - if (mlx5e_am_decision(&curr_stats, am)) { - am->state = MLX5E_AM_APPLY_NEW_PROFILE; - schedule_work(&am->work); - break; - } - /* fall through */ - case MLX5E_AM_START_MEASURE: - mlx5e_am_sample(rq, &am->start_sample); - am->state = MLX5E_AM_MEASURE_IN_PROGRESS; - break; - case MLX5E_AM_APPLY_NEW_PROFILE: - break; - } -} diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c index b74ddc7..5f0f349 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c @@ -71,6 +71,7 @@ static const struct counter_desc sw_stats_desc[] = { { MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, rx_cache_empty) }, { MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, rx_cache_busy) }, { MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, rx_cache_waive) }, + { MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, ch_eq_rearm) }, { MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, link_down_events_phy) }, }; @@ -99,6 +100,72 @@ static int mlx5e_grp_sw_fill_stats(struct mlx5e_priv *priv, u64 *data, int idx) return idx; } +static void mlx5e_grp_sw_update_stats(struct mlx5e_priv *priv) +{ + struct mlx5e_sw_stats temp, *s = &temp; + struct mlx5e_rq_stats *rq_stats; + struct mlx5e_sq_stats *sq_stats; + struct mlx5e_ch_stats *ch_stats; + int i, j; + + memset(s, 0, sizeof(*s)); + for (i = 0; i < priv->channels.num; i++) { + struct mlx5e_channel *c = priv->channels.c[i]; + + rq_stats = &c->rq.stats; + ch_stats = &c->stats; + + s->rx_packets += rq_stats->packets; + s->rx_bytes += rq_stats->bytes; + s->rx_lro_packets += rq_stats->lro_packets; + s->rx_lro_bytes += rq_stats->lro_bytes; + s->rx_removed_vlan_packets += rq_stats->removed_vlan_packets; + s->rx_csum_none += rq_stats->csum_none; + s->rx_csum_complete += rq_stats->csum_complete; + s->rx_csum_unnecessary += rq_stats->csum_unnecessary; + s->rx_csum_unnecessary_inner += rq_stats->csum_unnecessary_inner; + s->rx_xdp_drop += rq_stats->xdp_drop; + s->rx_xdp_tx += rq_stats->xdp_tx; + s->rx_xdp_tx_full += rq_stats->xdp_tx_full; + s->rx_wqe_err += rq_stats->wqe_err; + s->rx_mpwqe_filler += rq_stats->mpwqe_filler; + s->rx_buff_alloc_err += rq_stats->buff_alloc_err; + s->rx_cqe_compress_blks += rq_stats->cqe_compress_blks; + s->rx_cqe_compress_pkts += rq_stats->cqe_compress_pkts; + s->rx_page_reuse += rq_stats->page_reuse; + s->rx_cache_reuse += rq_stats->cache_reuse; + s->rx_cache_full += rq_stats->cache_full; + s->rx_cache_empty += rq_stats->cache_empty; + s->rx_cache_busy += rq_stats->cache_busy; + s->rx_cache_waive += rq_stats->cache_waive; + s->ch_eq_rearm += ch_stats->eq_rearm; + + for (j = 0; j < priv->channels.params.num_tc; j++) { + sq_stats = &c->sq[j].stats; + + s->tx_packets += sq_stats->packets; + s->tx_bytes += sq_stats->bytes; + s->tx_tso_packets += sq_stats->tso_packets; + s->tx_tso_bytes += sq_stats->tso_bytes; + s->tx_tso_inner_packets += sq_stats->tso_inner_packets; + s->tx_tso_inner_bytes += sq_stats->tso_inner_bytes; + s->tx_added_vlan_packets += sq_stats->added_vlan_packets; + s->tx_queue_stopped += sq_stats->stopped; + s->tx_queue_wake += sq_stats->wake; + s->tx_queue_dropped += sq_stats->dropped; + s->tx_xmit_more += sq_stats->xmit_more; + s->tx_csum_partial_inner += sq_stats->csum_partial_inner; + s->tx_csum_none += sq_stats->csum_none; + s->tx_csum_partial += sq_stats->csum_partial; + } + } + + s->link_down_events_phy = MLX5_GET(ppcnt_reg, + priv->stats.pport.phy_counters, + counter_set.phys_layer_cntrs.link_down_events); + memcpy(&priv->stats.sw, s, sizeof(*s)); +} + static const struct counter_desc q_stats_desc[] = { { MLX5E_DECLARE_STAT(struct mlx5e_qcounter_stats, rx_out_of_buffer) }, }; @@ -128,6 +195,22 @@ static int mlx5e_grp_q_fill_stats(struct mlx5e_priv *priv, u64 *data, int idx) return idx; } +static void mlx5e_grp_q_update_stats(struct mlx5e_priv *priv) +{ + struct mlx5e_qcounter_stats *qcnt = &priv->stats.qcnt; + u32 out[MLX5_ST_SZ_DW(query_q_counter_out)]; + int err; + + if (!priv->q_counter) + return; + + err = mlx5_core_query_q_counter(priv->mdev, priv->q_counter, 0, out, sizeof(out)); + if (err) + return; + + qcnt->rx_out_of_buffer = MLX5_GET(query_q_counter_out, out, out_of_buffer); +} + #define VPORT_COUNTER_OFF(c) MLX5_BYTE_OFF(query_vport_counter_out, c) static const struct counter_desc vport_stats_desc[] = { { "rx_vport_unicast_packets", @@ -200,6 +283,19 @@ static int mlx5e_grp_vport_fill_stats(struct mlx5e_priv *priv, u64 *data, return idx; } +static void mlx5e_grp_vport_update_stats(struct mlx5e_priv *priv) +{ + int outlen = MLX5_ST_SZ_BYTES(query_vport_counter_out); + u32 *out = (u32 *)priv->stats.vport.query_vport_out; + u32 in[MLX5_ST_SZ_DW(query_vport_counter_in)] = {0}; + struct mlx5_core_dev *mdev = priv->mdev; + + MLX5_SET(query_vport_counter_in, in, opcode, MLX5_CMD_OP_QUERY_VPORT_COUNTER); + MLX5_SET(query_vport_counter_in, in, op_mod, 0); + MLX5_SET(query_vport_counter_in, in, other_vport, 0); + mlx5_cmd_exec(mdev, in, sizeof(in), out, outlen); +} + #define PPORT_802_3_OFF(c) \ MLX5_BYTE_OFF(ppcnt_reg, \ counter_set.eth_802_3_cntrs_grp_data_layout.c##_high) @@ -252,6 +348,20 @@ static int mlx5e_grp_802_3_fill_stats(struct mlx5e_priv *priv, u64 *data, return idx; } +static void mlx5e_grp_802_3_update_stats(struct mlx5e_priv *priv) +{ + struct mlx5e_pport_stats *pstats = &priv->stats.pport; + struct mlx5_core_dev *mdev = priv->mdev; + u32 in[MLX5_ST_SZ_DW(ppcnt_reg)] = {0}; + int sz = MLX5_ST_SZ_BYTES(ppcnt_reg); + void *out; + + MLX5_SET(ppcnt_reg, in, local_port, 1); + out = pstats->IEEE_802_3_counters; + MLX5_SET(ppcnt_reg, in, grp, MLX5_IEEE_802_3_COUNTERS_GROUP); + mlx5_core_access_reg(mdev, in, sz, out, sz, MLX5_REG_PPCNT, 0, 0); +} + #define PPORT_2863_OFF(c) \ MLX5_BYTE_OFF(ppcnt_reg, \ counter_set.eth_2863_cntrs_grp_data_layout.c##_high) @@ -289,6 +399,20 @@ static int mlx5e_grp_2863_fill_stats(struct mlx5e_priv *priv, u64 *data, return idx; } +static void mlx5e_grp_2863_update_stats(struct mlx5e_priv *priv) +{ + struct mlx5e_pport_stats *pstats = &priv->stats.pport; + struct mlx5_core_dev *mdev = priv->mdev; + u32 in[MLX5_ST_SZ_DW(ppcnt_reg)] = {0}; + int sz = MLX5_ST_SZ_BYTES(ppcnt_reg); + void *out; + + MLX5_SET(ppcnt_reg, in, local_port, 1); + out = pstats->RFC_2863_counters; + MLX5_SET(ppcnt_reg, in, grp, MLX5_RFC_2863_COUNTERS_GROUP); + mlx5_core_access_reg(mdev, in, sz, out, sz, MLX5_REG_PPCNT, 0, 0); +} + #define PPORT_2819_OFF(c) \ MLX5_BYTE_OFF(ppcnt_reg, \ counter_set.eth_2819_cntrs_grp_data_layout.c##_high) @@ -336,6 +460,20 @@ static int mlx5e_grp_2819_fill_stats(struct mlx5e_priv *priv, u64 *data, return idx; } +static void mlx5e_grp_2819_update_stats(struct mlx5e_priv *priv) +{ + struct mlx5e_pport_stats *pstats = &priv->stats.pport; + struct mlx5_core_dev *mdev = priv->mdev; + u32 in[MLX5_ST_SZ_DW(ppcnt_reg)] = {0}; + int sz = MLX5_ST_SZ_BYTES(ppcnt_reg); + void *out; + + MLX5_SET(ppcnt_reg, in, local_port, 1); + out = pstats->RFC_2819_counters; + MLX5_SET(ppcnt_reg, in, grp, MLX5_RFC_2819_COUNTERS_GROUP); + mlx5_core_access_reg(mdev, in, sz, out, sz, MLX5_REG_PPCNT, 0, 0); +} + #define PPORT_PHY_STATISTICAL_OFF(c) \ MLX5_BYTE_OFF(ppcnt_reg, \ counter_set.phys_layer_statistical_cntrs.c##_high) @@ -376,6 +514,27 @@ static int mlx5e_grp_phy_fill_stats(struct mlx5e_priv *priv, u64 *data, int idx) return idx; } +static void mlx5e_grp_phy_update_stats(struct mlx5e_priv *priv) +{ + struct mlx5e_pport_stats *pstats = &priv->stats.pport; + struct mlx5_core_dev *mdev = priv->mdev; + u32 in[MLX5_ST_SZ_DW(ppcnt_reg)] = {0}; + int sz = MLX5_ST_SZ_BYTES(ppcnt_reg); + void *out; + + MLX5_SET(ppcnt_reg, in, local_port, 1); + out = pstats->phy_counters; + MLX5_SET(ppcnt_reg, in, grp, MLX5_PHYSICAL_LAYER_COUNTERS_GROUP); + mlx5_core_access_reg(mdev, in, sz, out, sz, MLX5_REG_PPCNT, 0, 0); + + if (!MLX5_CAP_PCAM_FEATURE(mdev, ppcnt_statistical_group)) + return; + + out = pstats->phy_statistical_counters; + MLX5_SET(ppcnt_reg, in, grp, MLX5_PHYSICAL_LAYER_STATISTICAL_GROUP); + mlx5_core_access_reg(mdev, in, sz, out, sz, MLX5_REG_PPCNT, 0, 0); +} + #define PPORT_ETH_EXT_OFF(c) \ MLX5_BYTE_OFF(ppcnt_reg, \ counter_set.eth_extended_cntrs_grp_data_layout.c##_high) @@ -418,6 +577,23 @@ static int mlx5e_grp_eth_ext_fill_stats(struct mlx5e_priv *priv, u64 *data, return idx; } +static void mlx5e_grp_eth_ext_update_stats(struct mlx5e_priv *priv) +{ + struct mlx5e_pport_stats *pstats = &priv->stats.pport; + struct mlx5_core_dev *mdev = priv->mdev; + u32 in[MLX5_ST_SZ_DW(ppcnt_reg)] = {0}; + int sz = MLX5_ST_SZ_BYTES(ppcnt_reg); + void *out; + + if (!MLX5_CAP_PCAM_FEATURE(mdev, rx_buffer_fullness_counters)) + return; + + MLX5_SET(ppcnt_reg, in, local_port, 1); + out = pstats->eth_ext_counters; + MLX5_SET(ppcnt_reg, in, grp, MLX5_ETHERNET_EXTENDED_COUNTERS_GROUP); + mlx5_core_access_reg(mdev, in, sz, out, sz, MLX5_REG_PPCNT, 0, 0); +} + #define PCIE_PERF_OFF(c) \ MLX5_BYTE_OFF(mpcnt_reg, counter_set.pcie_perf_cntrs_grp_data_layout.c) static const struct counter_desc pcie_perf_stats_desc[] = { @@ -505,6 +681,22 @@ static int mlx5e_grp_pcie_fill_stats(struct mlx5e_priv *priv, u64 *data, return idx; } +static void mlx5e_grp_pcie_update_stats(struct mlx5e_priv *priv) +{ + struct mlx5e_pcie_stats *pcie_stats = &priv->stats.pcie; + struct mlx5_core_dev *mdev = priv->mdev; + u32 in[MLX5_ST_SZ_DW(mpcnt_reg)] = {0}; + int sz = MLX5_ST_SZ_BYTES(mpcnt_reg); + void *out; + + if (!MLX5_CAP_MCAM_FEATURE(mdev, pcie_performance_group)) + return; + + out = pcie_stats->pcie_perf_counters; + MLX5_SET(mpcnt_reg, in, grp, MLX5_PCIE_PERFORMANCE_COUNTERS_GROUP); + mlx5_core_access_reg(mdev, in, sz, out, sz, MLX5_REG_MPCNT, 0, 0); +} + #define PPORT_PER_PRIO_OFF(c) \ MLX5_BYTE_OFF(ppcnt_reg, \ counter_set.eth_per_prio_grp_data_layout.c##_high) @@ -656,6 +848,47 @@ static int mlx5e_grp_per_prio_pfc_fill_stats(struct mlx5e_priv *priv, return idx; } +static int mlx5e_grp_per_prio_get_num_stats(struct mlx5e_priv *priv) +{ + return mlx5e_grp_per_prio_traffic_get_num_stats(priv) + + mlx5e_grp_per_prio_pfc_get_num_stats(priv); +} + +static int mlx5e_grp_per_prio_fill_strings(struct mlx5e_priv *priv, u8 *data, + int idx) +{ + idx = mlx5e_grp_per_prio_traffic_fill_strings(priv, data, idx); + idx = mlx5e_grp_per_prio_pfc_fill_strings(priv, data, idx); + return idx; +} + +static int mlx5e_grp_per_prio_fill_stats(struct mlx5e_priv *priv, u64 *data, + int idx) +{ + idx = mlx5e_grp_per_prio_traffic_fill_stats(priv, data, idx); + idx = mlx5e_grp_per_prio_pfc_fill_stats(priv, data, idx); + return idx; +} + +static void mlx5e_grp_per_prio_update_stats(struct mlx5e_priv *priv) +{ + struct mlx5e_pport_stats *pstats = &priv->stats.pport; + struct mlx5_core_dev *mdev = priv->mdev; + u32 in[MLX5_ST_SZ_DW(ppcnt_reg)] = {0}; + int sz = MLX5_ST_SZ_BYTES(ppcnt_reg); + int prio; + void *out; + + MLX5_SET(ppcnt_reg, in, local_port, 1); + MLX5_SET(ppcnt_reg, in, grp, MLX5_PER_PRIORITY_COUNTERS_GROUP); + for (prio = 0; prio < NUM_PPORT_PRIO; prio++) { + out = pstats->per_prio_counters[prio]; + MLX5_SET(ppcnt_reg, in, prio_tc, prio); + mlx5_core_access_reg(mdev, in, sz, out, sz, + MLX5_REG_PPCNT, 0, 0); + } +} + static const struct counter_desc mlx5e_pme_status_desc[] = { { "module_unplug", 8 }, }; @@ -723,6 +956,11 @@ static int mlx5e_grp_ipsec_fill_stats(struct mlx5e_priv *priv, u64 *data, return idx + mlx5e_ipsec_get_stats(priv, data + idx); } +static void mlx5e_grp_ipsec_update_stats(struct mlx5e_priv *priv) +{ + mlx5e_ipsec_update_stats(priv); +} + static const struct counter_desc rq_stats_desc[] = { { MLX5E_DECLARE_RX_STAT(struct mlx5e_rq_stats, packets) }, { MLX5E_DECLARE_RX_STAT(struct mlx5e_rq_stats, bytes) }, @@ -767,12 +1005,18 @@ static const struct counter_desc sq_stats_desc[] = { { MLX5E_DECLARE_TX_STAT(struct mlx5e_sq_stats, xmit_more) }, }; +static const struct counter_desc ch_stats_desc[] = { + { MLX5E_DECLARE_CH_STAT(struct mlx5e_ch_stats, eq_rearm) }, +}; + #define NUM_RQ_STATS ARRAY_SIZE(rq_stats_desc) #define NUM_SQ_STATS ARRAY_SIZE(sq_stats_desc) +#define NUM_CH_STATS ARRAY_SIZE(ch_stats_desc) static int mlx5e_grp_channels_get_num_stats(struct mlx5e_priv *priv) { return (NUM_RQ_STATS * priv->channels.num) + + (NUM_CH_STATS * priv->channels.num) + (NUM_SQ_STATS * priv->channels.num * priv->channels.params.num_tc); } @@ -785,6 +1029,11 @@ static int mlx5e_grp_channels_fill_strings(struct mlx5e_priv *priv, u8 *data, return idx; for (i = 0; i < priv->channels.num; i++) + for (j = 0; j < NUM_CH_STATS; j++) + sprintf(data + (idx++) * ETH_GSTRING_LEN, + ch_stats_desc[j].format, i); + + for (i = 0; i < priv->channels.num; i++) for (j = 0; j < NUM_RQ_STATS; j++) sprintf(data + (idx++) * ETH_GSTRING_LEN, rq_stats_desc[j].format, i); @@ -808,6 +1057,12 @@ static int mlx5e_grp_channels_fill_stats(struct mlx5e_priv *priv, u64 *data, return idx; for (i = 0; i < channels->num; i++) + for (j = 0; j < NUM_CH_STATS; j++) + data[idx++] = + MLX5E_READ_CTR64_CPU(&channels->c[i]->stats, + ch_stats_desc, j); + + for (i = 0; i < channels->num; i++) for (j = 0; j < NUM_RQ_STATS; j++) data[idx++] = MLX5E_READ_CTR64_CPU(&channels->c[i]->rq.stats, @@ -823,61 +1078,71 @@ static int mlx5e_grp_channels_fill_stats(struct mlx5e_priv *priv, u64 *data, return idx; } +/* The stats groups order is opposite to the update_stats() order calls */ const struct mlx5e_stats_grp mlx5e_stats_grps[] = { { .get_num_stats = mlx5e_grp_sw_get_num_stats, .fill_strings = mlx5e_grp_sw_fill_strings, .fill_stats = mlx5e_grp_sw_fill_stats, + .update_stats_mask = MLX5E_NDO_UPDATE_STATS, + .update_stats = mlx5e_grp_sw_update_stats, }, { .get_num_stats = mlx5e_grp_q_get_num_stats, .fill_strings = mlx5e_grp_q_fill_strings, .fill_stats = mlx5e_grp_q_fill_stats, + .update_stats_mask = MLX5E_NDO_UPDATE_STATS, + .update_stats = mlx5e_grp_q_update_stats, }, { .get_num_stats = mlx5e_grp_vport_get_num_stats, .fill_strings = mlx5e_grp_vport_fill_strings, .fill_stats = mlx5e_grp_vport_fill_stats, + .update_stats_mask = MLX5E_NDO_UPDATE_STATS, + .update_stats = mlx5e_grp_vport_update_stats, }, { .get_num_stats = mlx5e_grp_802_3_get_num_stats, .fill_strings = mlx5e_grp_802_3_fill_strings, .fill_stats = mlx5e_grp_802_3_fill_stats, + .update_stats_mask = MLX5E_NDO_UPDATE_STATS, + .update_stats = mlx5e_grp_802_3_update_stats, }, { .get_num_stats = mlx5e_grp_2863_get_num_stats, .fill_strings = mlx5e_grp_2863_fill_strings, .fill_stats = mlx5e_grp_2863_fill_stats, + .update_stats = mlx5e_grp_2863_update_stats, }, { .get_num_stats = mlx5e_grp_2819_get_num_stats, .fill_strings = mlx5e_grp_2819_fill_strings, .fill_stats = mlx5e_grp_2819_fill_stats, + .update_stats = mlx5e_grp_2819_update_stats, }, { .get_num_stats = mlx5e_grp_phy_get_num_stats, .fill_strings = mlx5e_grp_phy_fill_strings, .fill_stats = mlx5e_grp_phy_fill_stats, + .update_stats = mlx5e_grp_phy_update_stats, }, { .get_num_stats = mlx5e_grp_eth_ext_get_num_stats, .fill_strings = mlx5e_grp_eth_ext_fill_strings, .fill_stats = mlx5e_grp_eth_ext_fill_stats, + .update_stats = mlx5e_grp_eth_ext_update_stats, }, { .get_num_stats = mlx5e_grp_pcie_get_num_stats, .fill_strings = mlx5e_grp_pcie_fill_strings, .fill_stats = mlx5e_grp_pcie_fill_stats, + .update_stats = mlx5e_grp_pcie_update_stats, }, { - .get_num_stats = mlx5e_grp_per_prio_traffic_get_num_stats, - .fill_strings = mlx5e_grp_per_prio_traffic_fill_strings, - .fill_stats = mlx5e_grp_per_prio_traffic_fill_stats, - }, - { - .get_num_stats = mlx5e_grp_per_prio_pfc_get_num_stats, - .fill_strings = mlx5e_grp_per_prio_pfc_fill_strings, - .fill_stats = mlx5e_grp_per_prio_pfc_fill_stats, + .get_num_stats = mlx5e_grp_per_prio_get_num_stats, + .fill_strings = mlx5e_grp_per_prio_fill_strings, + .fill_stats = mlx5e_grp_per_prio_fill_stats, + .update_stats = mlx5e_grp_per_prio_update_stats, }, { .get_num_stats = mlx5e_grp_pme_get_num_stats, @@ -888,6 +1153,7 @@ const struct mlx5e_stats_grp mlx5e_stats_grps[] = { .get_num_stats = mlx5e_grp_ipsec_get_num_stats, .fill_strings = mlx5e_grp_ipsec_fill_strings, .fill_stats = mlx5e_grp_ipsec_fill_stats, + .update_stats = mlx5e_grp_ipsec_update_stats, }, { .get_num_stats = mlx5e_grp_channels_get_num_stats, diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h index d679e21..0b3320a 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h @@ -44,6 +44,7 @@ #define MLX5E_DECLARE_STAT(type, fld) #fld, offsetof(type, fld) #define MLX5E_DECLARE_RX_STAT(type, fld) "rx%d_"#fld, offsetof(type, fld) #define MLX5E_DECLARE_TX_STAT(type, fld) "tx%d_"#fld, offsetof(type, fld) +#define MLX5E_DECLARE_CH_STAT(type, fld) "ch%d_"#fld, offsetof(type, fld) struct counter_desc { char format[ETH_GSTRING_LEN]; @@ -88,6 +89,7 @@ struct mlx5e_sw_stats { u64 rx_cache_empty; u64 rx_cache_busy; u64 rx_cache_waive; + u64 ch_eq_rearm; /* Special handling counters */ u64 link_down_events_phy; @@ -192,6 +194,10 @@ struct mlx5e_sq_stats { u64 dropped; }; +struct mlx5e_ch_stats { + u64 eq_rearm; +}; + struct mlx5e_stats { struct mlx5e_sw_stats sw; struct mlx5e_qcounter_stats qcnt; @@ -201,11 +207,17 @@ struct mlx5e_stats { struct mlx5e_pcie_stats pcie; }; +enum { + MLX5E_NDO_UPDATE_STATS = BIT(0x1), +}; + struct mlx5e_priv; struct mlx5e_stats_grp { + u16 update_stats_mask; int (*get_num_stats)(struct mlx5e_priv *priv); int (*fill_strings)(struct mlx5e_priv *priv, u8 *data, int idx); int (*fill_stats)(struct mlx5e_priv *priv, u64 *data, int idx); + void (*update_stats)(struct mlx5e_priv *priv); }; extern const struct mlx5e_stats_grp mlx5e_stats_grps[]; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c index 55979ec..fd98b0d 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c @@ -51,17 +51,22 @@ #include "en_tc.h" #include "eswitch.h" #include "vxlan.h" +#include "fs_core.h" struct mlx5_nic_flow_attr { u32 action; u32 flow_tag; u32 mod_hdr_id; + u32 hairpin_tirn; + struct mlx5_flow_table *hairpin_ft; }; enum { MLX5E_TC_FLOW_ESWITCH = BIT(0), MLX5E_TC_FLOW_NIC = BIT(1), MLX5E_TC_FLOW_OFFLOADED = BIT(2), + MLX5E_TC_FLOW_HAIRPIN = BIT(3), + MLX5E_TC_FLOW_HAIRPIN_RSS = BIT(4), }; struct mlx5e_tc_flow { @@ -71,6 +76,7 @@ struct mlx5e_tc_flow { struct mlx5_flow_handle *rule; struct list_head encap; /* flows sharing the same encap ID */ struct list_head mod_hdr; /* flows sharing the same mod hdr ID */ + struct list_head hairpin; /* flows sharing the same hairpin */ union { struct mlx5_esw_flow_attr esw_attr[0]; struct mlx5_nic_flow_attr nic_attr[0]; @@ -93,6 +99,32 @@ enum { #define MLX5E_TC_TABLE_NUM_GROUPS 4 #define MLX5E_TC_TABLE_MAX_GROUP_SIZE (1 << 16) +struct mlx5e_hairpin { + struct mlx5_hairpin *pair; + + struct mlx5_core_dev *func_mdev; + struct mlx5e_priv *func_priv; + u32 tdn; + u32 tirn; + + int num_channels; + struct mlx5e_rqt indir_rqt; + u32 indir_tirn[MLX5E_NUM_INDIR_TIRS]; + struct mlx5e_ttc_table ttc; +}; + +struct mlx5e_hairpin_entry { + /* a node of a hash table which keeps all the hairpin entries */ + struct hlist_node hairpin_hlist; + + /* flows sharing the same hairpin */ + struct list_head flows; + + u16 peer_vhca_id; + u8 prio; + struct mlx5e_hairpin *hp; +}; + struct mod_hdr_key { int num_actions; void *actions; @@ -222,6 +254,417 @@ static void mlx5e_detach_mod_hdr(struct mlx5e_priv *priv, } } +static +struct mlx5_core_dev *mlx5e_hairpin_get_mdev(struct net *net, int ifindex) +{ + struct net_device *netdev; + struct mlx5e_priv *priv; + + netdev = __dev_get_by_index(net, ifindex); + priv = netdev_priv(netdev); + return priv->mdev; +} + +static int mlx5e_hairpin_create_transport(struct mlx5e_hairpin *hp) +{ + u32 in[MLX5_ST_SZ_DW(create_tir_in)] = {0}; + void *tirc; + int err; + + err = mlx5_core_alloc_transport_domain(hp->func_mdev, &hp->tdn); + if (err) + goto alloc_tdn_err; + + tirc = MLX5_ADDR_OF(create_tir_in, in, ctx); + + MLX5_SET(tirc, tirc, disp_type, MLX5_TIRC_DISP_TYPE_DIRECT); + MLX5_SET(tirc, tirc, inline_rqn, hp->pair->rqn[0]); + MLX5_SET(tirc, tirc, transport_domain, hp->tdn); + + err = mlx5_core_create_tir(hp->func_mdev, in, MLX5_ST_SZ_BYTES(create_tir_in), &hp->tirn); + if (err) + goto create_tir_err; + + return 0; + +create_tir_err: + mlx5_core_dealloc_transport_domain(hp->func_mdev, hp->tdn); +alloc_tdn_err: + return err; +} + +static void mlx5e_hairpin_destroy_transport(struct mlx5e_hairpin *hp) +{ + mlx5_core_destroy_tir(hp->func_mdev, hp->tirn); + mlx5_core_dealloc_transport_domain(hp->func_mdev, hp->tdn); +} + +static void mlx5e_hairpin_fill_rqt_rqns(struct mlx5e_hairpin *hp, void *rqtc) +{ + u32 indirection_rqt[MLX5E_INDIR_RQT_SIZE], rqn; + struct mlx5e_priv *priv = hp->func_priv; + int i, ix, sz = MLX5E_INDIR_RQT_SIZE; + + mlx5e_build_default_indir_rqt(indirection_rqt, sz, + hp->num_channels); + + for (i = 0; i < sz; i++) { + ix = i; + if (priv->channels.params.rss_hfunc == ETH_RSS_HASH_XOR) + ix = mlx5e_bits_invert(i, ilog2(sz)); + ix = indirection_rqt[ix]; + rqn = hp->pair->rqn[ix]; + MLX5_SET(rqtc, rqtc, rq_num[i], rqn); + } +} + +static int mlx5e_hairpin_create_indirect_rqt(struct mlx5e_hairpin *hp) +{ + int inlen, err, sz = MLX5E_INDIR_RQT_SIZE; + struct mlx5e_priv *priv = hp->func_priv; + struct mlx5_core_dev *mdev = priv->mdev; + void *rqtc; + u32 *in; + + inlen = MLX5_ST_SZ_BYTES(create_rqt_in) + sizeof(u32) * sz; + in = kvzalloc(inlen, GFP_KERNEL); + if (!in) + return -ENOMEM; + + rqtc = MLX5_ADDR_OF(create_rqt_in, in, rqt_context); + + MLX5_SET(rqtc, rqtc, rqt_actual_size, sz); + MLX5_SET(rqtc, rqtc, rqt_max_size, sz); + + mlx5e_hairpin_fill_rqt_rqns(hp, rqtc); + + err = mlx5_core_create_rqt(mdev, in, inlen, &hp->indir_rqt.rqtn); + if (!err) + hp->indir_rqt.enabled = true; + + kvfree(in); + return err; +} + +static int mlx5e_hairpin_create_indirect_tirs(struct mlx5e_hairpin *hp) +{ + struct mlx5e_priv *priv = hp->func_priv; + u32 in[MLX5_ST_SZ_DW(create_tir_in)]; + int tt, i, err; + void *tirc; + + for (tt = 0; tt < MLX5E_NUM_INDIR_TIRS; tt++) { + memset(in, 0, MLX5_ST_SZ_BYTES(create_tir_in)); + tirc = MLX5_ADDR_OF(create_tir_in, in, ctx); + + MLX5_SET(tirc, tirc, transport_domain, hp->tdn); + MLX5_SET(tirc, tirc, disp_type, MLX5_TIRC_DISP_TYPE_INDIRECT); + MLX5_SET(tirc, tirc, indirect_table, hp->indir_rqt.rqtn); + mlx5e_build_indir_tir_ctx_hash(&priv->channels.params, tt, tirc, false); + + err = mlx5_core_create_tir(hp->func_mdev, in, + MLX5_ST_SZ_BYTES(create_tir_in), &hp->indir_tirn[tt]); + if (err) { + mlx5_core_warn(hp->func_mdev, "create indirect tirs failed, %d\n", err); + goto err_destroy_tirs; + } + } + return 0; + +err_destroy_tirs: + for (i = 0; i < tt; i++) + mlx5_core_destroy_tir(hp->func_mdev, hp->indir_tirn[i]); + return err; +} + +static void mlx5e_hairpin_destroy_indirect_tirs(struct mlx5e_hairpin *hp) +{ + int tt; + + for (tt = 0; tt < MLX5E_NUM_INDIR_TIRS; tt++) + mlx5_core_destroy_tir(hp->func_mdev, hp->indir_tirn[tt]); +} + +static void mlx5e_hairpin_set_ttc_params(struct mlx5e_hairpin *hp, + struct ttc_params *ttc_params) +{ + struct mlx5_flow_table_attr *ft_attr = &ttc_params->ft_attr; + int tt; + + memset(ttc_params, 0, sizeof(*ttc_params)); + + ttc_params->any_tt_tirn = hp->tirn; + + for (tt = 0; tt < MLX5E_NUM_INDIR_TIRS; tt++) + ttc_params->indir_tirn[tt] = hp->indir_tirn[tt]; + + ft_attr->max_fte = MLX5E_NUM_TT; + ft_attr->level = MLX5E_TC_TTC_FT_LEVEL; + ft_attr->prio = MLX5E_TC_PRIO; +} + +static int mlx5e_hairpin_rss_init(struct mlx5e_hairpin *hp) +{ + struct mlx5e_priv *priv = hp->func_priv; + struct ttc_params ttc_params; + int err; + + err = mlx5e_hairpin_create_indirect_rqt(hp); + if (err) + return err; + + err = mlx5e_hairpin_create_indirect_tirs(hp); + if (err) + goto err_create_indirect_tirs; + + mlx5e_hairpin_set_ttc_params(hp, &ttc_params); + err = mlx5e_create_ttc_table(priv, &ttc_params, &hp->ttc); + if (err) + goto err_create_ttc_table; + + netdev_dbg(priv->netdev, "add hairpin: using %d channels rss ttc table id %x\n", + hp->num_channels, hp->ttc.ft.t->id); + + return 0; + +err_create_ttc_table: + mlx5e_hairpin_destroy_indirect_tirs(hp); +err_create_indirect_tirs: + mlx5e_destroy_rqt(priv, &hp->indir_rqt); + + return err; +} + +static void mlx5e_hairpin_rss_cleanup(struct mlx5e_hairpin *hp) +{ + struct mlx5e_priv *priv = hp->func_priv; + + mlx5e_destroy_ttc_table(priv, &hp->ttc); + mlx5e_hairpin_destroy_indirect_tirs(hp); + mlx5e_destroy_rqt(priv, &hp->indir_rqt); +} + +static struct mlx5e_hairpin * +mlx5e_hairpin_create(struct mlx5e_priv *priv, struct mlx5_hairpin_params *params, + int peer_ifindex) +{ + struct mlx5_core_dev *func_mdev, *peer_mdev; + struct mlx5e_hairpin *hp; + struct mlx5_hairpin *pair; + int err; + + hp = kzalloc(sizeof(*hp), GFP_KERNEL); + if (!hp) + return ERR_PTR(-ENOMEM); + + func_mdev = priv->mdev; + peer_mdev = mlx5e_hairpin_get_mdev(dev_net(priv->netdev), peer_ifindex); + + pair = mlx5_core_hairpin_create(func_mdev, peer_mdev, params); + if (IS_ERR(pair)) { + err = PTR_ERR(pair); + goto create_pair_err; + } + hp->pair = pair; + hp->func_mdev = func_mdev; + hp->func_priv = priv; + hp->num_channels = params->num_channels; + + err = mlx5e_hairpin_create_transport(hp); + if (err) + goto create_transport_err; + + if (hp->num_channels > 1) { + err = mlx5e_hairpin_rss_init(hp); + if (err) + goto rss_init_err; + } + + return hp; + +rss_init_err: + mlx5e_hairpin_destroy_transport(hp); +create_transport_err: + mlx5_core_hairpin_destroy(hp->pair); +create_pair_err: + kfree(hp); + return ERR_PTR(err); +} + +static void mlx5e_hairpin_destroy(struct mlx5e_hairpin *hp) +{ + if (hp->num_channels > 1) + mlx5e_hairpin_rss_cleanup(hp); + mlx5e_hairpin_destroy_transport(hp); + mlx5_core_hairpin_destroy(hp->pair); + kvfree(hp); +} + +static inline u32 hash_hairpin_info(u16 peer_vhca_id, u8 prio) +{ + return (peer_vhca_id << 16 | prio); +} + +static struct mlx5e_hairpin_entry *mlx5e_hairpin_get(struct mlx5e_priv *priv, + u16 peer_vhca_id, u8 prio) +{ + struct mlx5e_hairpin_entry *hpe; + u32 hash_key = hash_hairpin_info(peer_vhca_id, prio); + + hash_for_each_possible(priv->fs.tc.hairpin_tbl, hpe, + hairpin_hlist, hash_key) { + if (hpe->peer_vhca_id == peer_vhca_id && hpe->prio == prio) + return hpe; + } + + return NULL; +} + +#define UNKNOWN_MATCH_PRIO 8 + +static int mlx5e_hairpin_get_prio(struct mlx5e_priv *priv, + struct mlx5_flow_spec *spec, u8 *match_prio) +{ + void *headers_c, *headers_v; + u8 prio_val, prio_mask = 0; + bool vlan_present; + +#ifdef CONFIG_MLX5_CORE_EN_DCB + if (priv->dcbx_dp.trust_state != MLX5_QPTS_TRUST_PCP) { + netdev_warn(priv->netdev, + "only PCP trust state supported for hairpin\n"); + return -EOPNOTSUPP; + } +#endif + headers_c = MLX5_ADDR_OF(fte_match_param, spec->match_criteria, outer_headers); + headers_v = MLX5_ADDR_OF(fte_match_param, spec->match_value, outer_headers); + + vlan_present = MLX5_GET(fte_match_set_lyr_2_4, headers_v, cvlan_tag); + if (vlan_present) { + prio_mask = MLX5_GET(fte_match_set_lyr_2_4, headers_c, first_prio); + prio_val = MLX5_GET(fte_match_set_lyr_2_4, headers_v, first_prio); + } + + if (!vlan_present || !prio_mask) { + prio_val = UNKNOWN_MATCH_PRIO; + } else if (prio_mask != 0x7) { + netdev_warn(priv->netdev, + "masked priority match not supported for hairpin\n"); + return -EOPNOTSUPP; + } + + *match_prio = prio_val; + return 0; +} + +static int mlx5e_hairpin_flow_add(struct mlx5e_priv *priv, + struct mlx5e_tc_flow *flow, + struct mlx5e_tc_flow_parse_attr *parse_attr) +{ + int peer_ifindex = parse_attr->mirred_ifindex; + struct mlx5_hairpin_params params; + struct mlx5_core_dev *peer_mdev; + struct mlx5e_hairpin_entry *hpe; + struct mlx5e_hairpin *hp; + u64 link_speed64; + u32 link_speed; + u8 match_prio; + u16 peer_id; + int err; + + peer_mdev = mlx5e_hairpin_get_mdev(dev_net(priv->netdev), peer_ifindex); + if (!MLX5_CAP_GEN(priv->mdev, hairpin) || !MLX5_CAP_GEN(peer_mdev, hairpin)) { + netdev_warn(priv->netdev, "hairpin is not supported\n"); + return -EOPNOTSUPP; + } + + peer_id = MLX5_CAP_GEN(peer_mdev, vhca_id); + err = mlx5e_hairpin_get_prio(priv, &parse_attr->spec, &match_prio); + if (err) + return err; + hpe = mlx5e_hairpin_get(priv, peer_id, match_prio); + if (hpe) + goto attach_flow; + + hpe = kzalloc(sizeof(*hpe), GFP_KERNEL); + if (!hpe) + return -ENOMEM; + + INIT_LIST_HEAD(&hpe->flows); + hpe->peer_vhca_id = peer_id; + hpe->prio = match_prio; + + params.log_data_size = 15; + params.log_data_size = min_t(u8, params.log_data_size, + MLX5_CAP_GEN(priv->mdev, log_max_hairpin_wq_data_sz)); + params.log_data_size = max_t(u8, params.log_data_size, + MLX5_CAP_GEN(priv->mdev, log_min_hairpin_wq_data_sz)); + + params.log_num_packets = params.log_data_size - + MLX5_MPWRQ_MIN_LOG_STRIDE_SZ(priv->mdev); + params.log_num_packets = min_t(u8, params.log_num_packets, + MLX5_CAP_GEN(priv->mdev, log_max_hairpin_num_packets)); + + params.q_counter = priv->q_counter; + /* set hairpin pair per each 50Gbs share of the link */ + mlx5e_get_max_linkspeed(priv->mdev, &link_speed); + link_speed = max_t(u32, link_speed, 50000); + link_speed64 = link_speed; + do_div(link_speed64, 50000); + params.num_channels = link_speed64; + + hp = mlx5e_hairpin_create(priv, ¶ms, peer_ifindex); + if (IS_ERR(hp)) { + err = PTR_ERR(hp); + goto create_hairpin_err; + } + + netdev_dbg(priv->netdev, "add hairpin: tirn %x rqn %x peer %s sqn %x prio %d (log) data %d packets %d\n", + hp->tirn, hp->pair->rqn[0], hp->pair->peer_mdev->priv.name, + hp->pair->sqn[0], match_prio, params.log_data_size, params.log_num_packets); + + hpe->hp = hp; + hash_add(priv->fs.tc.hairpin_tbl, &hpe->hairpin_hlist, + hash_hairpin_info(peer_id, match_prio)); + +attach_flow: + if (hpe->hp->num_channels > 1) { + flow->flags |= MLX5E_TC_FLOW_HAIRPIN_RSS; + flow->nic_attr->hairpin_ft = hpe->hp->ttc.ft.t; + } else { + flow->nic_attr->hairpin_tirn = hpe->hp->tirn; + } + list_add(&flow->hairpin, &hpe->flows); + + return 0; + +create_hairpin_err: + kfree(hpe); + return err; +} + +static void mlx5e_hairpin_flow_del(struct mlx5e_priv *priv, + struct mlx5e_tc_flow *flow) +{ + struct list_head *next = flow->hairpin.next; + + list_del(&flow->hairpin); + + /* no more hairpin flows for us, release the hairpin pair */ + if (list_empty(next)) { + struct mlx5e_hairpin_entry *hpe; + + hpe = list_entry(next, struct mlx5e_hairpin_entry, flows); + + netdev_dbg(priv->netdev, "del hairpin: peer %s\n", + hpe->hp->pair->peer_mdev->priv.name); + + mlx5e_hairpin_destroy(hpe->hp); + hash_del(&hpe->hairpin_hlist); + kfree(hpe); + } +} + static struct mlx5_flow_handle * mlx5e_tc_add_nic_flow(struct mlx5e_priv *priv, struct mlx5e_tc_flow_parse_attr *parse_attr, @@ -229,7 +672,7 @@ mlx5e_tc_add_nic_flow(struct mlx5e_priv *priv, { struct mlx5_nic_flow_attr *attr = flow->nic_attr; struct mlx5_core_dev *dev = priv->mdev; - struct mlx5_flow_destination dest = {}; + struct mlx5_flow_destination dest[2] = {}; struct mlx5_flow_act flow_act = { .action = attr->action, .flow_tag = attr->flow_tag, @@ -238,18 +681,37 @@ mlx5e_tc_add_nic_flow(struct mlx5e_priv *priv, struct mlx5_fc *counter = NULL; struct mlx5_flow_handle *rule; bool table_created = false; - int err; + int err, dest_ix = 0; - if (attr->action & MLX5_FLOW_CONTEXT_ACTION_FWD_DEST) { - dest.type = MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE; - dest.ft = priv->fs.vlan.ft.t; - } else if (attr->action & MLX5_FLOW_CONTEXT_ACTION_COUNT) { - counter = mlx5_fc_create(dev, true); - if (IS_ERR(counter)) - return ERR_CAST(counter); + if (flow->flags & MLX5E_TC_FLOW_HAIRPIN) { + err = mlx5e_hairpin_flow_add(priv, flow, parse_attr); + if (err) { + rule = ERR_PTR(err); + goto err_add_hairpin_flow; + } + if (flow->flags & MLX5E_TC_FLOW_HAIRPIN_RSS) { + dest[dest_ix].type = MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE; + dest[dest_ix].ft = attr->hairpin_ft; + } else { + dest[dest_ix].type = MLX5_FLOW_DESTINATION_TYPE_TIR; + dest[dest_ix].tir_num = attr->hairpin_tirn; + } + dest_ix++; + } else if (attr->action & MLX5_FLOW_CONTEXT_ACTION_FWD_DEST) { + dest[dest_ix].type = MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE; + dest[dest_ix].ft = priv->fs.vlan.ft.t; + dest_ix++; + } - dest.type = MLX5_FLOW_DESTINATION_TYPE_COUNTER; - dest.counter = counter; + if (attr->action & MLX5_FLOW_CONTEXT_ACTION_COUNT) { + counter = mlx5_fc_create(dev, true); + if (IS_ERR(counter)) { + rule = ERR_CAST(counter); + goto err_fc_create; + } + dest[dest_ix].type = MLX5_FLOW_DESTINATION_TYPE_COUNTER; + dest[dest_ix].counter = counter; + dest_ix++; } if (attr->action & MLX5_FLOW_CONTEXT_ACTION_MOD_HDR) { @@ -279,7 +741,7 @@ mlx5e_tc_add_nic_flow(struct mlx5e_priv *priv, MLX5E_TC_PRIO, tc_tbl_size, MLX5E_TC_TABLE_NUM_GROUPS, - 0, 0); + MLX5E_TC_FT_LEVEL, 0); if (IS_ERR(priv->fs.tc.t)) { netdev_err(priv->netdev, "Failed to create tc offload table\n"); @@ -292,7 +754,7 @@ mlx5e_tc_add_nic_flow(struct mlx5e_priv *priv, parse_attr->spec.match_criteria_enable = MLX5_MATCH_OUTER_HEADERS; rule = mlx5_add_flow_rules(priv->fs.tc.t, &parse_attr->spec, - &flow_act, &dest, 1); + &flow_act, dest, dest_ix); if (IS_ERR(rule)) goto err_add_rule; @@ -309,7 +771,10 @@ err_create_ft: mlx5e_detach_mod_hdr(priv, flow); err_create_mod_hdr_id: mlx5_fc_destroy(dev, counter); - +err_fc_create: + if (flow->flags & MLX5E_TC_FLOW_HAIRPIN) + mlx5e_hairpin_flow_del(priv, flow); +err_add_hairpin_flow: return rule; } @@ -330,6 +795,9 @@ static void mlx5e_tc_del_nic_flow(struct mlx5e_priv *priv, if (attr->action & MLX5_FLOW_CONTEXT_ACTION_MOD_HDR) mlx5e_detach_mod_hdr(priv, flow); + + if (flow->flags & MLX5E_TC_FLOW_HAIRPIN) + mlx5e_hairpin_flow_del(priv, flow); } static void mlx5e_detach_encap(struct mlx5e_priv *priv, @@ -617,7 +1085,8 @@ static int parse_tunnel_attr(struct mlx5e_priv *priv, FLOW_DISSECTOR_KEY_ENC_PORTS, f->mask); struct mlx5_eswitch *esw = priv->mdev->priv.eswitch; - struct net_device *up_dev = mlx5_eswitch_get_uplink_netdev(esw); + struct mlx5e_rep_priv *uplink_rpriv = mlx5_eswitch_get_uplink_priv(esw, REP_ETH); + struct net_device *up_dev = uplink_rpriv->netdev; struct mlx5e_priv *up_priv = netdev_priv(up_dev); /* Full udp dst port must be given */ @@ -1421,6 +1890,20 @@ static bool actions_match_supported(struct mlx5e_priv *priv, return true; } +static bool same_hw_devs(struct mlx5e_priv *priv, struct mlx5e_priv *peer_priv) +{ + struct mlx5_core_dev *fmdev, *pmdev; + u16 func_id, peer_id; + + fmdev = priv->mdev; + pmdev = peer_priv->mdev; + + func_id = (u16)((fmdev->pdev->bus->number << 8) | PCI_SLOT(fmdev->pdev->devfn)); + peer_id = (u16)((pmdev->pdev->bus->number << 8) | PCI_SLOT(pmdev->pdev->devfn)); + + return (func_id == peer_id); +} + static int parse_tc_nic_actions(struct mlx5e_priv *priv, struct tcf_exts *exts, struct mlx5e_tc_flow_parse_attr *parse_attr, struct mlx5e_tc_flow *flow) @@ -1465,6 +1948,23 @@ static int parse_tc_nic_actions(struct mlx5e_priv *priv, struct tcf_exts *exts, return -EOPNOTSUPP; } + if (is_tcf_mirred_egress_redirect(a)) { + struct net_device *peer_dev = tcf_mirred_dev(a); + + if (priv->netdev->netdev_ops == peer_dev->netdev_ops && + same_hw_devs(priv, netdev_priv(peer_dev))) { + parse_attr->mirred_ifindex = peer_dev->ifindex; + flow->flags |= MLX5E_TC_FLOW_HAIRPIN; + attr->action |= MLX5_FLOW_CONTEXT_ACTION_FWD_DEST | + MLX5_FLOW_CONTEXT_ACTION_COUNT; + } else { + netdev_warn(priv->netdev, "device %s not on same HW, can't offload\n", + peer_dev->name); + return -EINVAL; + } + continue; + } + if (is_tcf_skbedit_mark(a)) { u32 mark = tcf_skbedit_mark(a); @@ -1507,6 +2007,7 @@ static int mlx5e_route_lookup_ipv4(struct mlx5e_priv *priv, int *out_ttl) { struct mlx5_eswitch *esw = priv->mdev->priv.eswitch; + struct mlx5e_rep_priv *uplink_rpriv; struct rtable *rt; struct neighbour *n = NULL; @@ -1520,9 +2021,10 @@ static int mlx5e_route_lookup_ipv4(struct mlx5e_priv *priv, #else return -EOPNOTSUPP; #endif + uplink_rpriv = mlx5_eswitch_get_uplink_priv(esw, REP_ETH); /* if the egress device isn't on the same HW e-switch, we use the uplink */ if (!switchdev_port_same_parent_id(priv->netdev, rt->dst.dev)) - *out_dev = mlx5_eswitch_get_uplink_netdev(esw); + *out_dev = uplink_rpriv->netdev; else *out_dev = rt->dst.dev; @@ -1547,6 +2049,7 @@ static int mlx5e_route_lookup_ipv6(struct mlx5e_priv *priv, struct dst_entry *dst; #if IS_ENABLED(CONFIG_INET) && IS_ENABLED(CONFIG_IPV6) + struct mlx5e_rep_priv *uplink_rpriv; struct mlx5_eswitch *esw = priv->mdev->priv.eswitch; int ret; @@ -1557,9 +2060,10 @@ static int mlx5e_route_lookup_ipv6(struct mlx5e_priv *priv, *out_ttl = ip6_dst_hoplimit(dst); + uplink_rpriv = mlx5_eswitch_get_uplink_priv(esw, REP_ETH); /* if the egress device isn't on the same HW e-switch, we use the uplink */ if (!switchdev_port_same_parent_id(priv->netdev, dst->dev)) - *out_dev = mlx5_eswitch_get_uplink_netdev(esw); + *out_dev = uplink_rpriv->netdev; else *out_dev = dst->dev; #else @@ -1859,7 +2363,9 @@ static int mlx5e_attach_encap(struct mlx5e_priv *priv, struct mlx5e_tc_flow *flow) { struct mlx5_eswitch *esw = priv->mdev->priv.eswitch; - struct net_device *up_dev = mlx5_eswitch_get_uplink_netdev(esw); + struct mlx5e_rep_priv *uplink_rpriv = mlx5_eswitch_get_uplink_priv(esw, + REP_ETH); + struct net_device *up_dev = uplink_rpriv->netdev; unsigned short family = ip_tunnel_info_af(tun_info); struct mlx5e_priv *up_priv = netdev_priv(up_dev); struct mlx5_esw_flow_attr *attr = flow->esw_attr; @@ -1982,11 +2488,10 @@ static int parse_tc_fdb_actions(struct mlx5e_priv *priv, struct tcf_exts *exts, } if (is_tcf_mirred_egress_redirect(a)) { - int ifindex = tcf_mirred_ifindex(a); struct net_device *out_dev; struct mlx5e_priv *out_priv; - out_dev = __dev_get_by_index(dev_net(priv->netdev), ifindex); + out_dev = tcf_mirred_dev(a); if (switchdev_port_same_parent_id(priv->netdev, out_dev)) { @@ -1996,7 +2501,7 @@ static int parse_tc_fdb_actions(struct mlx5e_priv *priv, struct tcf_exts *exts, rpriv = out_priv->ppriv; attr->out_rep = rpriv->rep; } else if (encap) { - parse_attr->mirred_ifindex = ifindex; + parse_attr->mirred_ifindex = out_dev->ifindex; parse_attr->tun_info = *info; attr->parse_attr = parse_attr; attr->action |= MLX5_FLOW_CONTEXT_ACTION_ENCAP | @@ -2182,6 +2687,7 @@ int mlx5e_tc_init(struct mlx5e_priv *priv) struct mlx5e_tc_table *tc = &priv->fs.tc; hash_init(tc->mod_hdr_tbl); + hash_init(tc->hairpin_tbl); tc->ht_params = mlx5e_tc_flow_ht_params; return rhashtable_init(&tc->ht, &tc->ht_params); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_txrx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_txrx.c index ab92298..f292bb3 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_txrx.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_txrx.c @@ -78,8 +78,14 @@ int mlx5e_napi_poll(struct napi_struct *napi, int budget) for (i = 0; i < c->num_tc; i++) mlx5e_cq_arm(&c->sq[i].cq); - if (MLX5E_TEST_BIT(c->rq.state, MLX5E_RQ_STATE_AM)) - mlx5e_rx_am(&c->rq); + if (MLX5E_TEST_BIT(c->rq.state, MLX5E_RQ_STATE_AM)) { + struct net_dim_sample dim_sample; + net_dim_sample(c->rq.cq.event_ctr, + c->rq.stats.packets, + c->rq.stats.bytes, + &dim_sample); + net_dim(&c->rq.dim, dim_sample); + } mlx5e_cq_arm(&c->rq.cq); mlx5e_cq_arm(&c->icosq.cq); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eq.c b/drivers/net/ethernet/mellanox/mlx5/core/eq.c index 14d5782..05c1157 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eq.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/eq.c @@ -534,6 +534,24 @@ static irqreturn_t mlx5_eq_int(int irq, void *eq_ptr) return IRQ_HANDLED; } +/* Some architectures don't latch interrupts when they are disabled, so using + * mlx5_eq_poll_irq_disabled could end up losing interrupts while trying to + * avoid losing them. It is not recommended to use it, unless this is the last + * resort. + */ +u32 mlx5_eq_poll_irq_disabled(struct mlx5_eq *eq) +{ + u32 count_eqe; + + disable_irq(eq->irqn); + count_eqe = eq->cons_index; + mlx5_eq_int(eq->irqn, eq); + count_eqe = eq->cons_index - count_eqe; + enable_irq(eq->irqn); + + return count_eqe; +} + static void init_eq_buf(struct mlx5_eq *eq) { struct mlx5_eqe *eqe; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c index bbb140f..5ecf2cd 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c @@ -37,6 +37,7 @@ #include <linux/mlx5/fs.h> #include "mlx5_core.h" #include "eswitch.h" +#include "fs_core.h" #define UPLINK_VPORT 0xFFFF @@ -867,9 +868,10 @@ static int esw_vport_enable_egress_acl(struct mlx5_eswitch *esw, esw_debug(dev, "Create vport[%d] egress ACL log_max_size(%d)\n", vport->vport, MLX5_CAP_ESW_EGRESS_ACL(dev, log_max_ft_size)); - root_ns = mlx5_get_flow_namespace(dev, MLX5_FLOW_NAMESPACE_ESW_EGRESS); + root_ns = mlx5_get_flow_vport_acl_namespace(dev, MLX5_FLOW_NAMESPACE_ESW_EGRESS, + vport->vport); if (!root_ns) { - esw_warn(dev, "Failed to get E-Switch egress flow namespace\n"); + esw_warn(dev, "Failed to get E-Switch egress flow namespace for vport (%d)\n", vport->vport); return -EOPNOTSUPP; } @@ -984,9 +986,10 @@ static int esw_vport_enable_ingress_acl(struct mlx5_eswitch *esw, esw_debug(dev, "Create vport[%d] ingress ACL log_max_size(%d)\n", vport->vport, MLX5_CAP_ESW_INGRESS_ACL(dev, log_max_ft_size)); - root_ns = mlx5_get_flow_namespace(dev, MLX5_FLOW_NAMESPACE_ESW_INGRESS); + root_ns = mlx5_get_flow_vport_acl_namespace(dev, MLX5_FLOW_NAMESPACE_ESW_INGRESS, + vport->vport); if (!root_ns) { - esw_warn(dev, "Failed to get E-Switch ingress flow namespace\n"); + esw_warn(dev, "Failed to get E-Switch ingress flow namespace for vport (%d)\n", vport->vport); return -EOPNOTSUPP; } @@ -1121,8 +1124,12 @@ static void esw_vport_disable_ingress_acl(struct mlx5_eswitch *esw, static int esw_vport_ingress_config(struct mlx5_eswitch *esw, struct mlx5_vport *vport) { + struct mlx5_fc *counter = vport->ingress.drop_counter; + struct mlx5_flow_destination drop_ctr_dst = {0}; + struct mlx5_flow_destination *dst = NULL; struct mlx5_flow_act flow_act = {0}; struct mlx5_flow_spec *spec; + int dest_num = 0; int err = 0; u8 *smac_v; @@ -1186,9 +1193,18 @@ static int esw_vport_ingress_config(struct mlx5_eswitch *esw, memset(spec, 0, sizeof(*spec)); flow_act.action = MLX5_FLOW_CONTEXT_ACTION_DROP; + + /* Attach drop flow counter */ + if (counter) { + flow_act.action |= MLX5_FLOW_CONTEXT_ACTION_COUNT; + drop_ctr_dst.type = MLX5_FLOW_DESTINATION_TYPE_COUNTER; + drop_ctr_dst.counter = counter; + dst = &drop_ctr_dst; + dest_num++; + } vport->ingress.drop_rule = mlx5_add_flow_rules(vport->ingress.acl, spec, - &flow_act, NULL, 0); + &flow_act, dst, dest_num); if (IS_ERR(vport->ingress.drop_rule)) { err = PTR_ERR(vport->ingress.drop_rule); esw_warn(esw->dev, @@ -1208,8 +1224,12 @@ out: static int esw_vport_egress_config(struct mlx5_eswitch *esw, struct mlx5_vport *vport) { + struct mlx5_fc *counter = vport->egress.drop_counter; + struct mlx5_flow_destination drop_ctr_dst = {0}; + struct mlx5_flow_destination *dst = NULL; struct mlx5_flow_act flow_act = {0}; struct mlx5_flow_spec *spec; + int dest_num = 0; int err = 0; esw_vport_cleanup_egress_rules(esw, vport); @@ -1260,9 +1280,18 @@ static int esw_vport_egress_config(struct mlx5_eswitch *esw, /* Drop others rule (star rule) */ memset(spec, 0, sizeof(*spec)); flow_act.action = MLX5_FLOW_CONTEXT_ACTION_DROP; + + /* Attach egress drop flow counter */ + if (counter) { + flow_act.action |= MLX5_FLOW_CONTEXT_ACTION_COUNT; + drop_ctr_dst.type = MLX5_FLOW_DESTINATION_TYPE_COUNTER; + drop_ctr_dst.counter = counter; + dst = &drop_ctr_dst; + dest_num++; + } vport->egress.drop_rule = mlx5_add_flow_rules(vport->egress.acl, spec, - &flow_act, NULL, 0); + &flow_act, dst, dest_num); if (IS_ERR(vport->egress.drop_rule)) { err = PTR_ERR(vport->egress.drop_rule); esw_warn(esw->dev, @@ -1290,7 +1319,7 @@ static int esw_create_tsar(struct mlx5_eswitch *esw) err = mlx5_create_scheduling_element_cmd(dev, SCHEDULING_HIERARCHY_E_SWITCH, - &tsar_ctx, + tsar_ctx, &esw->qos.root_tsar_id); if (err) { esw_warn(esw->dev, "E-Switch create TSAR failed (%d)\n", err); @@ -1333,20 +1362,20 @@ static int esw_vport_enable_qos(struct mlx5_eswitch *esw, int vport_num, if (vport->qos.enabled) return -EEXIST; - MLX5_SET(scheduling_context, &sched_ctx, element_type, + MLX5_SET(scheduling_context, sched_ctx, element_type, SCHEDULING_CONTEXT_ELEMENT_TYPE_VPORT); - vport_elem = MLX5_ADDR_OF(scheduling_context, &sched_ctx, + vport_elem = MLX5_ADDR_OF(scheduling_context, sched_ctx, element_attributes); MLX5_SET(vport_element, vport_elem, vport_number, vport_num); - MLX5_SET(scheduling_context, &sched_ctx, parent_element_id, + MLX5_SET(scheduling_context, sched_ctx, parent_element_id, esw->qos.root_tsar_id); - MLX5_SET(scheduling_context, &sched_ctx, max_average_bw, + MLX5_SET(scheduling_context, sched_ctx, max_average_bw, initial_max_rate); - MLX5_SET(scheduling_context, &sched_ctx, bw_share, initial_bw_share); + MLX5_SET(scheduling_context, sched_ctx, bw_share, initial_bw_share); err = mlx5_create_scheduling_element_cmd(dev, SCHEDULING_HIERARCHY_E_SWITCH, - &sched_ctx, + sched_ctx, &vport->qos.esw_tsar_ix); if (err) { esw_warn(esw->dev, "E-Switch create TSAR vport element failed (vport=%d,err=%d)\n", @@ -1392,22 +1421,22 @@ static int esw_vport_qos_config(struct mlx5_eswitch *esw, int vport_num, if (!vport->qos.enabled) return -EIO; - MLX5_SET(scheduling_context, &sched_ctx, element_type, + MLX5_SET(scheduling_context, sched_ctx, element_type, SCHEDULING_CONTEXT_ELEMENT_TYPE_VPORT); - vport_elem = MLX5_ADDR_OF(scheduling_context, &sched_ctx, + vport_elem = MLX5_ADDR_OF(scheduling_context, sched_ctx, element_attributes); MLX5_SET(vport_element, vport_elem, vport_number, vport_num); - MLX5_SET(scheduling_context, &sched_ctx, parent_element_id, + MLX5_SET(scheduling_context, sched_ctx, parent_element_id, esw->qos.root_tsar_id); - MLX5_SET(scheduling_context, &sched_ctx, max_average_bw, + MLX5_SET(scheduling_context, sched_ctx, max_average_bw, max_rate); - MLX5_SET(scheduling_context, &sched_ctx, bw_share, bw_share); + MLX5_SET(scheduling_context, sched_ctx, bw_share, bw_share); bitmask |= MODIFY_SCHEDULING_ELEMENT_IN_MODIFY_BITMASK_MAX_AVERAGE_BW; bitmask |= MODIFY_SCHEDULING_ELEMENT_IN_MODIFY_BITMASK_BW_SHARE; err = mlx5_modify_scheduling_element_cmd(dev, SCHEDULING_HIERARCHY_E_SWITCH, - &sched_ctx, + sched_ctx, vport->qos.esw_tsar_ix, bitmask); if (err) { @@ -1455,6 +1484,41 @@ static void esw_apply_vport_conf(struct mlx5_eswitch *esw, } } +static void esw_vport_create_drop_counters(struct mlx5_vport *vport) +{ + struct mlx5_core_dev *dev = vport->dev; + + if (MLX5_CAP_ESW_INGRESS_ACL(dev, flow_counter)) { + vport->ingress.drop_counter = mlx5_fc_create(dev, false); + if (IS_ERR(vport->ingress.drop_counter)) { + esw_warn(dev, + "vport[%d] configure ingress drop rule counter failed\n", + vport->vport); + vport->ingress.drop_counter = NULL; + } + } + + if (MLX5_CAP_ESW_EGRESS_ACL(dev, flow_counter)) { + vport->egress.drop_counter = mlx5_fc_create(dev, false); + if (IS_ERR(vport->egress.drop_counter)) { + esw_warn(dev, + "vport[%d] configure egress drop rule counter failed\n", + vport->vport); + vport->egress.drop_counter = NULL; + } + } +} + +static void esw_vport_destroy_drop_counters(struct mlx5_vport *vport) +{ + struct mlx5_core_dev *dev = vport->dev; + + if (vport->ingress.drop_counter) + mlx5_fc_destroy(dev, vport->ingress.drop_counter); + if (vport->egress.drop_counter) + mlx5_fc_destroy(dev, vport->egress.drop_counter); +} + static void esw_enable_vport(struct mlx5_eswitch *esw, int vport_num, int enable_events) { @@ -1481,6 +1545,10 @@ static void esw_enable_vport(struct mlx5_eswitch *esw, int vport_num, if (!vport_num) vport->info.trusted = true; + /* create steering drop counters for ingress and egress ACLs */ + if (vport_num && esw->mode == SRIOV_LEGACY) + esw_vport_create_drop_counters(vport); + esw_vport_change_handle_locked(vport); esw->enabled_vports++; @@ -1519,6 +1587,7 @@ static void esw_disable_vport(struct mlx5_eswitch *esw, int vport_num) MLX5_ESW_VPORT_ADMIN_STATE_DOWN); esw_vport_disable_egress_acl(esw, vport); esw_vport_disable_ingress_acl(esw, vport); + esw_vport_destroy_drop_counters(vport); } esw->enabled_vports--; mutex_unlock(&esw->state_lock); @@ -1644,13 +1713,9 @@ int mlx5_eswitch_init(struct mlx5_core_dev *dev) goto abort; } - esw->offloads.vport_reps = - kzalloc(total_vports * sizeof(struct mlx5_eswitch_rep), - GFP_KERNEL); - if (!esw->offloads.vport_reps) { - err = -ENOMEM; + err = esw_offloads_init_reps(esw); + if (err) goto abort; - } hash_init(esw->offloads.encap_tbl); hash_init(esw->offloads.mod_hdr_tbl); @@ -1681,8 +1746,8 @@ int mlx5_eswitch_init(struct mlx5_core_dev *dev) abort: if (esw->work_queue) destroy_workqueue(esw->work_queue); + esw_offloads_cleanup_reps(esw); kfree(esw->vports); - kfree(esw->offloads.vport_reps); kfree(esw); return err; } @@ -1696,7 +1761,7 @@ void mlx5_eswitch_cleanup(struct mlx5_eswitch *esw) esw->dev->priv.eswitch = NULL; destroy_workqueue(esw->work_queue); - kfree(esw->offloads.vport_reps); + esw_offloads_cleanup_reps(esw); kfree(esw->vports); kfree(esw); } @@ -2018,12 +2083,36 @@ unlock: return err; } +static void mlx5_eswitch_query_vport_drop_stats(struct mlx5_core_dev *dev, + int vport_idx, + struct mlx5_vport_drop_stats *stats) +{ + struct mlx5_eswitch *esw = dev->priv.eswitch; + struct mlx5_vport *vport = &esw->vports[vport_idx]; + u64 bytes = 0; + u16 idx = 0; + + if (!vport->enabled || esw->mode != SRIOV_LEGACY) + return; + + if (vport->egress.drop_counter) { + idx = vport->egress.drop_counter->id; + mlx5_fc_query(dev, idx, &stats->rx_dropped, &bytes); + } + + if (vport->ingress.drop_counter) { + idx = vport->ingress.drop_counter->id; + mlx5_fc_query(dev, idx, &stats->tx_dropped, &bytes); + } +} + int mlx5_eswitch_get_vport_stats(struct mlx5_eswitch *esw, int vport, struct ifla_vf_stats *vf_stats) { int outlen = MLX5_ST_SZ_BYTES(query_vport_counter_out); u32 in[MLX5_ST_SZ_DW(query_vport_counter_in)] = {0}; + struct mlx5_vport_drop_stats stats = {0}; int err = 0; u32 *out; @@ -2078,6 +2167,10 @@ int mlx5_eswitch_get_vport_stats(struct mlx5_eswitch *esw, vf_stats->broadcast = MLX5_GET_CTR(out, received_eth_broadcast.packets); + mlx5_eswitch_query_vport_drop_stats(esw->dev, vport, &stats); + vf_stats->rx_dropped = stats.rx_dropped; + vf_stats->tx_dropped = stats.tx_dropped; + free_out: kvfree(out); return err; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h index 565c8b7..2fa0370 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h @@ -45,6 +45,11 @@ enum { SRIOV_OFFLOADS }; +enum { + REP_ETH, + NUM_REP_TYPES, +}; + #ifdef CONFIG_MLX5_ESWITCH #define MLX5_MAX_UC_PER_VPORT(dev) \ @@ -68,6 +73,7 @@ struct vport_ingress { struct mlx5_flow_group *drop_grp; struct mlx5_flow_handle *allow_rule; struct mlx5_flow_handle *drop_rule; + struct mlx5_fc *drop_counter; }; struct vport_egress { @@ -76,6 +82,12 @@ struct vport_egress { struct mlx5_flow_group *drop_grp; struct mlx5_flow_handle *allowed_vlan; struct mlx5_flow_handle *drop_rule; + struct mlx5_fc *drop_counter; +}; + +struct mlx5_vport_drop_stats { + u64 rx_dropped; + u64 tx_dropped; }; struct mlx5_vport_info { @@ -133,25 +145,21 @@ struct mlx5_eswitch_fdb { }; }; -struct mlx5_esw_sq { - struct mlx5_flow_handle *send_to_vport_rule; - struct list_head list; +struct mlx5_eswitch_rep; +struct mlx5_eswitch_rep_if { + int (*load)(struct mlx5_core_dev *dev, + struct mlx5_eswitch_rep *rep); + void (*unload)(struct mlx5_eswitch_rep *rep); + void *priv; + bool valid; }; struct mlx5_eswitch_rep { - int (*load)(struct mlx5_eswitch *esw, - struct mlx5_eswitch_rep *rep); - void (*unload)(struct mlx5_eswitch *esw, - struct mlx5_eswitch_rep *rep); + struct mlx5_eswitch_rep_if rep_if[NUM_REP_TYPES]; u16 vport; u8 hw_id[ETH_ALEN]; - struct net_device *netdev; - - struct mlx5_flow_handle *vport_rx_rule; - struct list_head vport_sqs_list; u16 vlan; u32 vlan_refcount; - bool valid; }; struct mlx5_esw_offload { @@ -197,6 +205,8 @@ struct mlx5_eswitch { void esw_offloads_cleanup(struct mlx5_eswitch *esw, int nvports); int esw_offloads_init(struct mlx5_eswitch *esw, int nvports); +void esw_offloads_cleanup_reps(struct mlx5_eswitch *esw); +int esw_offloads_init_reps(struct mlx5_eswitch *esw); /* E-Switch API */ int mlx5_eswitch_init(struct mlx5_core_dev *dev); @@ -221,6 +231,10 @@ int mlx5_eswitch_get_vport_config(struct mlx5_eswitch *esw, int mlx5_eswitch_get_vport_stats(struct mlx5_eswitch *esw, int vport, struct ifla_vf_stats *vf_stats); +struct mlx5_flow_handle * +mlx5_eswitch_add_send_to_vport_rule(struct mlx5_eswitch *esw, int vport, + u32 sqn); +void mlx5_eswitch_del_send_to_vport_rule(struct mlx5_flow_handle *rule); struct mlx5_flow_spec; struct mlx5_esw_flow_attr; @@ -257,12 +271,6 @@ struct mlx5_esw_flow_attr { struct mlx5e_tc_flow_parse_attr *parse_attr; }; -int mlx5_eswitch_sqs2vport_start(struct mlx5_eswitch *esw, - struct mlx5_eswitch_rep *rep, - u16 *sqns_array, int sqns_num); -void mlx5_eswitch_sqs2vport_stop(struct mlx5_eswitch *esw, - struct mlx5_eswitch_rep *rep); - int mlx5_devlink_eswitch_mode_set(struct devlink *devlink, u16 mode); int mlx5_devlink_eswitch_mode_get(struct devlink *devlink, u16 *mode); int mlx5_devlink_eswitch_inline_mode_set(struct devlink *devlink, u8 mode); @@ -272,10 +280,12 @@ int mlx5_devlink_eswitch_encap_mode_set(struct devlink *devlink, u8 encap); int mlx5_devlink_eswitch_encap_mode_get(struct devlink *devlink, u8 *encap); void mlx5_eswitch_register_vport_rep(struct mlx5_eswitch *esw, int vport_index, - struct mlx5_eswitch_rep *rep); + struct mlx5_eswitch_rep_if *rep_if, + u8 rep_type); void mlx5_eswitch_unregister_vport_rep(struct mlx5_eswitch *esw, - int vport_index); -struct net_device *mlx5_eswitch_get_uplink_netdev(struct mlx5_eswitch *esw); + int vport_index, + u8 rep_type); +void *mlx5_eswitch_get_uplink_priv(struct mlx5_eswitch *esw, u8 rep_type); int mlx5_eswitch_add_vlan_action(struct mlx5_eswitch *esw, struct mlx5_esw_flow_attr *attr); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c index 1143d80..99f583a 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c @@ -130,7 +130,7 @@ static int esw_set_global_vlan_pop(struct mlx5_eswitch *esw, u8 val) esw_debug(esw->dev, "%s applying global %s policy\n", __func__, val ? "pop" : "none"); for (vf_vport = 1; vf_vport < esw->enabled_vports; vf_vport++) { rep = &esw->offloads.vport_reps[vf_vport]; - if (!rep->valid) + if (!rep->rep_if[REP_ETH].valid) continue; err = __mlx5_eswitch_set_vport_vlan(esw, rep->vport, 0, 0, val); @@ -302,7 +302,7 @@ out: return err; } -static struct mlx5_flow_handle * +struct mlx5_flow_handle * mlx5_eswitch_add_send_to_vport_rule(struct mlx5_eswitch *esw, int vport, u32 sqn) { struct mlx5_flow_act flow_act = {0}; @@ -339,57 +339,9 @@ out: return flow_rule; } -void mlx5_eswitch_sqs2vport_stop(struct mlx5_eswitch *esw, - struct mlx5_eswitch_rep *rep) -{ - struct mlx5_esw_sq *esw_sq, *tmp; - - if (esw->mode != SRIOV_OFFLOADS) - return; - - list_for_each_entry_safe(esw_sq, tmp, &rep->vport_sqs_list, list) { - mlx5_del_flow_rules(esw_sq->send_to_vport_rule); - list_del(&esw_sq->list); - kfree(esw_sq); - } -} - -int mlx5_eswitch_sqs2vport_start(struct mlx5_eswitch *esw, - struct mlx5_eswitch_rep *rep, - u16 *sqns_array, int sqns_num) +void mlx5_eswitch_del_send_to_vport_rule(struct mlx5_flow_handle *rule) { - struct mlx5_flow_handle *flow_rule; - struct mlx5_esw_sq *esw_sq; - int err; - int i; - - if (esw->mode != SRIOV_OFFLOADS) - return 0; - - for (i = 0; i < sqns_num; i++) { - esw_sq = kzalloc(sizeof(*esw_sq), GFP_KERNEL); - if (!esw_sq) { - err = -ENOMEM; - goto out_err; - } - - /* Add re-inject rule to the PF/representor sqs */ - flow_rule = mlx5_eswitch_add_send_to_vport_rule(esw, - rep->vport, - sqns_array[i]); - if (IS_ERR(flow_rule)) { - err = PTR_ERR(flow_rule); - kfree(esw_sq); - goto out_err; - } - esw_sq->send_to_vport_rule = flow_rule; - list_add(&esw_sq->list, &rep->vport_sqs_list); - } - return 0; - -out_err: - mlx5_eswitch_sqs2vport_stop(esw, rep); - return err; + mlx5_del_flow_rules(rule); } static int esw_add_fdb_miss_rule(struct mlx5_eswitch *esw) @@ -732,12 +684,111 @@ static int esw_offloads_start(struct mlx5_eswitch *esw) return err; } -int esw_offloads_init(struct mlx5_eswitch *esw, int nvports) +void esw_offloads_cleanup_reps(struct mlx5_eswitch *esw) +{ + kfree(esw->offloads.vport_reps); +} + +int esw_offloads_init_reps(struct mlx5_eswitch *esw) +{ + int total_vfs = MLX5_TOTAL_VPORTS(esw->dev); + struct mlx5_core_dev *dev = esw->dev; + struct mlx5_esw_offload *offloads; + struct mlx5_eswitch_rep *rep; + u8 hw_id[ETH_ALEN]; + int vport; + + esw->offloads.vport_reps = kcalloc(total_vfs, + sizeof(struct mlx5_eswitch_rep), + GFP_KERNEL); + if (!esw->offloads.vport_reps) + return -ENOMEM; + + offloads = &esw->offloads; + mlx5_query_nic_vport_mac_address(dev, 0, hw_id); + + for (vport = 0; vport < total_vfs; vport++) { + rep = &offloads->vport_reps[vport]; + + rep->vport = vport; + ether_addr_copy(rep->hw_id, hw_id); + } + + offloads->vport_reps[0].vport = FDB_UPLINK_VPORT; + + return 0; +} + +static void esw_offloads_unload_reps_type(struct mlx5_eswitch *esw, int nvports, + u8 rep_type) +{ + struct mlx5_eswitch_rep *rep; + int vport; + + for (vport = nvports - 1; vport >= 0; vport--) { + rep = &esw->offloads.vport_reps[vport]; + if (!rep->rep_if[rep_type].valid) + continue; + + rep->rep_if[rep_type].unload(rep); + } +} + +static void esw_offloads_unload_reps(struct mlx5_eswitch *esw, int nvports) +{ + u8 rep_type = NUM_REP_TYPES; + + while (rep_type-- > 0) + esw_offloads_unload_reps_type(esw, nvports, rep_type); +} + +static int esw_offloads_load_reps_type(struct mlx5_eswitch *esw, int nvports, + u8 rep_type) { struct mlx5_eswitch_rep *rep; int vport; int err; + for (vport = 0; vport < nvports; vport++) { + rep = &esw->offloads.vport_reps[vport]; + if (!rep->rep_if[rep_type].valid) + continue; + + err = rep->rep_if[rep_type].load(esw->dev, rep); + if (err) + goto err_reps; + } + + return 0; + +err_reps: + esw_offloads_unload_reps_type(esw, vport, rep_type); + return err; +} + +static int esw_offloads_load_reps(struct mlx5_eswitch *esw, int nvports) +{ + u8 rep_type = 0; + int err; + + for (rep_type = 0; rep_type < NUM_REP_TYPES; rep_type++) { + err = esw_offloads_load_reps_type(esw, nvports, rep_type); + if (err) + goto err_reps; + } + + return err; + +err_reps: + while (rep_type-- > 0) + esw_offloads_unload_reps_type(esw, nvports, rep_type); + return err; +} + +int esw_offloads_init(struct mlx5_eswitch *esw, int nvports) +{ + int err; + /* disable PF RoCE so missed packets don't go through RoCE steering */ mlx5_dev_list_lock(); mlx5_remove_dev_by_protocol(esw->dev, MLX5_INTERFACE_PROTOCOL_IB); @@ -755,25 +806,13 @@ int esw_offloads_init(struct mlx5_eswitch *esw, int nvports) if (err) goto create_fg_err; - for (vport = 0; vport < nvports; vport++) { - rep = &esw->offloads.vport_reps[vport]; - if (!rep->valid) - continue; - - err = rep->load(esw, rep); - if (err) - goto err_reps; - } + err = esw_offloads_load_reps(esw, nvports); + if (err) + goto err_reps; return 0; err_reps: - for (vport--; vport >= 0; vport--) { - rep = &esw->offloads.vport_reps[vport]; - if (!rep->valid) - continue; - rep->unload(esw, rep); - } esw_destroy_vport_rx_group(esw); create_fg_err: @@ -814,16 +853,7 @@ static int esw_offloads_stop(struct mlx5_eswitch *esw) void esw_offloads_cleanup(struct mlx5_eswitch *esw, int nvports) { - struct mlx5_eswitch_rep *rep; - int vport; - - for (vport = nvports - 1; vport >= 0; vport--) { - rep = &esw->offloads.vport_reps[vport]; - if (!rep->valid) - continue; - rep->unload(esw, rep); - } - + esw_offloads_unload_reps(esw, nvports); esw_destroy_vport_rx_group(esw); esw_destroy_offloads_table(esw); esw_destroy_offloads_fdb_tables(esw); @@ -1120,27 +1150,23 @@ int mlx5_devlink_eswitch_encap_mode_get(struct devlink *devlink, u8 *encap) void mlx5_eswitch_register_vport_rep(struct mlx5_eswitch *esw, int vport_index, - struct mlx5_eswitch_rep *__rep) + struct mlx5_eswitch_rep_if *__rep_if, + u8 rep_type) { struct mlx5_esw_offload *offloads = &esw->offloads; - struct mlx5_eswitch_rep *rep; - - rep = &offloads->vport_reps[vport_index]; + struct mlx5_eswitch_rep_if *rep_if; - memset(rep, 0, sizeof(*rep)); + rep_if = &offloads->vport_reps[vport_index].rep_if[rep_type]; - rep->load = __rep->load; - rep->unload = __rep->unload; - rep->vport = __rep->vport; - rep->netdev = __rep->netdev; - ether_addr_copy(rep->hw_id, __rep->hw_id); + rep_if->load = __rep_if->load; + rep_if->unload = __rep_if->unload; + rep_if->priv = __rep_if->priv; - INIT_LIST_HEAD(&rep->vport_sqs_list); - rep->valid = true; + rep_if->valid = true; } void mlx5_eswitch_unregister_vport_rep(struct mlx5_eswitch *esw, - int vport_index) + int vport_index, u8 rep_type) { struct mlx5_esw_offload *offloads = &esw->offloads; struct mlx5_eswitch_rep *rep; @@ -1148,17 +1174,17 @@ void mlx5_eswitch_unregister_vport_rep(struct mlx5_eswitch *esw, rep = &offloads->vport_reps[vport_index]; if (esw->mode == SRIOV_OFFLOADS && esw->vports[vport_index].enabled) - rep->unload(esw, rep); + rep->rep_if[rep_type].unload(rep); - rep->valid = false; + rep->rep_if[rep_type].valid = false; } -struct net_device *mlx5_eswitch_get_uplink_netdev(struct mlx5_eswitch *esw) +void *mlx5_eswitch_get_uplink_priv(struct mlx5_eswitch *esw, u8 rep_type) { #define UPLINK_REP_INDEX 0 struct mlx5_esw_offload *offloads = &esw->offloads; struct mlx5_eswitch_rep *rep; rep = &offloads->vport_reps[UPLINK_REP_INDEX]; - return rep->netdev; + return rep->rep_if[rep_type].priv; } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c index dfaad9e..c025c98 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c @@ -89,6 +89,9 @@ /* One more level for tc */ #define KERNEL_MIN_LEVEL (KERNEL_NIC_PRIO_NUM_LEVELS + 1) +#define KERNEL_NIC_TC_NUM_PRIOS 1 +#define KERNEL_NIC_TC_NUM_LEVELS 2 + #define ANCHOR_NUM_LEVELS 1 #define ANCHOR_NUM_PRIOS 1 #define ANCHOR_MIN_LEVEL (BY_PASS_MIN_LEVEL + 1) @@ -134,7 +137,7 @@ static struct init_tree_node { ADD_NS(ADD_MULTIPLE_PRIO(ETHTOOL_NUM_PRIOS, ETHTOOL_PRIO_NUM_LEVELS))), ADD_PRIO(0, KERNEL_MIN_LEVEL, 0, {}, - ADD_NS(ADD_MULTIPLE_PRIO(1, 1), + ADD_NS(ADD_MULTIPLE_PRIO(KERNEL_NIC_TC_NUM_PRIOS, KERNEL_NIC_TC_NUM_LEVELS), ADD_MULTIPLE_PRIO(KERNEL_NIC_NUM_PRIOS, KERNEL_NIC_PRIO_NUM_LEVELS))), ADD_PRIO(0, BY_PASS_MIN_LEVEL, 0, @@ -2026,16 +2029,6 @@ struct mlx5_flow_namespace *mlx5_get_flow_namespace(struct mlx5_core_dev *dev, return &steering->fdb_root_ns->ns; else return NULL; - case MLX5_FLOW_NAMESPACE_ESW_EGRESS: - if (steering->esw_egress_root_ns) - return &steering->esw_egress_root_ns->ns; - else - return NULL; - case MLX5_FLOW_NAMESPACE_ESW_INGRESS: - if (steering->esw_ingress_root_ns) - return &steering->esw_ingress_root_ns->ns; - else - return NULL; case MLX5_FLOW_NAMESPACE_SNIFFER_RX: if (steering->sniffer_rx_root_ns) return &steering->sniffer_rx_root_ns->ns; @@ -2066,6 +2059,33 @@ struct mlx5_flow_namespace *mlx5_get_flow_namespace(struct mlx5_core_dev *dev, } EXPORT_SYMBOL(mlx5_get_flow_namespace); +struct mlx5_flow_namespace *mlx5_get_flow_vport_acl_namespace(struct mlx5_core_dev *dev, + enum mlx5_flow_namespace_type type, + int vport) +{ + struct mlx5_flow_steering *steering = dev->priv.steering; + + if (!steering || vport >= MLX5_TOTAL_VPORTS(dev)) + return NULL; + + switch (type) { + case MLX5_FLOW_NAMESPACE_ESW_EGRESS: + if (steering->esw_egress_root_ns && + steering->esw_egress_root_ns[vport]) + return &steering->esw_egress_root_ns[vport]->ns; + else + return NULL; + case MLX5_FLOW_NAMESPACE_ESW_INGRESS: + if (steering->esw_ingress_root_ns && + steering->esw_ingress_root_ns[vport]) + return &steering->esw_ingress_root_ns[vport]->ns; + else + return NULL; + default: + return NULL; + } +} + static struct fs_prio *fs_create_prio(struct mlx5_flow_namespace *ns, unsigned int prio, int num_levels) { @@ -2343,13 +2363,41 @@ static void cleanup_root_ns(struct mlx5_flow_root_namespace *root_ns) clean_tree(&root_ns->ns.node); } +static void cleanup_egress_acls_root_ns(struct mlx5_core_dev *dev) +{ + struct mlx5_flow_steering *steering = dev->priv.steering; + int i; + + if (!steering->esw_egress_root_ns) + return; + + for (i = 0; i < MLX5_TOTAL_VPORTS(dev); i++) + cleanup_root_ns(steering->esw_egress_root_ns[i]); + + kfree(steering->esw_egress_root_ns); +} + +static void cleanup_ingress_acls_root_ns(struct mlx5_core_dev *dev) +{ + struct mlx5_flow_steering *steering = dev->priv.steering; + int i; + + if (!steering->esw_ingress_root_ns) + return; + + for (i = 0; i < MLX5_TOTAL_VPORTS(dev); i++) + cleanup_root_ns(steering->esw_ingress_root_ns[i]); + + kfree(steering->esw_ingress_root_ns); +} + void mlx5_cleanup_fs(struct mlx5_core_dev *dev) { struct mlx5_flow_steering *steering = dev->priv.steering; cleanup_root_ns(steering->root_ns); - cleanup_root_ns(steering->esw_egress_root_ns); - cleanup_root_ns(steering->esw_ingress_root_ns); + cleanup_egress_acls_root_ns(dev); + cleanup_ingress_acls_root_ns(dev); cleanup_root_ns(steering->fdb_root_ns); cleanup_root_ns(steering->sniffer_rx_root_ns); cleanup_root_ns(steering->sniffer_tx_root_ns); @@ -2418,34 +2466,86 @@ out_err: return PTR_ERR(prio); } -static int init_ingress_acl_root_ns(struct mlx5_flow_steering *steering) +static int init_egress_acl_root_ns(struct mlx5_flow_steering *steering, int vport) { struct fs_prio *prio; - steering->esw_egress_root_ns = create_root_ns(steering, FS_FT_ESW_EGRESS_ACL); - if (!steering->esw_egress_root_ns) + steering->esw_egress_root_ns[vport] = create_root_ns(steering, FS_FT_ESW_EGRESS_ACL); + if (!steering->esw_egress_root_ns[vport]) return -ENOMEM; /* create 1 prio*/ - prio = fs_create_prio(&steering->esw_egress_root_ns->ns, 0, - MLX5_TOTAL_VPORTS(steering->dev)); + prio = fs_create_prio(&steering->esw_egress_root_ns[vport]->ns, 0, 1); return PTR_ERR_OR_ZERO(prio); } -static int init_egress_acl_root_ns(struct mlx5_flow_steering *steering) +static int init_ingress_acl_root_ns(struct mlx5_flow_steering *steering, int vport) { struct fs_prio *prio; - steering->esw_ingress_root_ns = create_root_ns(steering, FS_FT_ESW_INGRESS_ACL); - if (!steering->esw_ingress_root_ns) + steering->esw_ingress_root_ns[vport] = create_root_ns(steering, FS_FT_ESW_INGRESS_ACL); + if (!steering->esw_ingress_root_ns[vport]) return -ENOMEM; /* create 1 prio*/ - prio = fs_create_prio(&steering->esw_ingress_root_ns->ns, 0, - MLX5_TOTAL_VPORTS(steering->dev)); + prio = fs_create_prio(&steering->esw_ingress_root_ns[vport]->ns, 0, 1); return PTR_ERR_OR_ZERO(prio); } +static int init_egress_acls_root_ns(struct mlx5_core_dev *dev) +{ + struct mlx5_flow_steering *steering = dev->priv.steering; + int err; + int i; + + steering->esw_egress_root_ns = kcalloc(MLX5_TOTAL_VPORTS(dev), + sizeof(*steering->esw_egress_root_ns), + GFP_KERNEL); + if (!steering->esw_egress_root_ns) + return -ENOMEM; + + for (i = 0; i < MLX5_TOTAL_VPORTS(dev); i++) { + err = init_egress_acl_root_ns(steering, i); + if (err) + goto cleanup_root_ns; + } + + return 0; + +cleanup_root_ns: + for (i--; i >= 0; i--) + cleanup_root_ns(steering->esw_egress_root_ns[i]); + kfree(steering->esw_egress_root_ns); + return err; +} + +static int init_ingress_acls_root_ns(struct mlx5_core_dev *dev) +{ + struct mlx5_flow_steering *steering = dev->priv.steering; + int err; + int i; + + steering->esw_ingress_root_ns = kcalloc(MLX5_TOTAL_VPORTS(dev), + sizeof(*steering->esw_ingress_root_ns), + GFP_KERNEL); + if (!steering->esw_ingress_root_ns) + return -ENOMEM; + + for (i = 0; i < MLX5_TOTAL_VPORTS(dev); i++) { + err = init_ingress_acl_root_ns(steering, i); + if (err) + goto cleanup_root_ns; + } + + return 0; + +cleanup_root_ns: + for (i--; i >= 0; i--) + cleanup_root_ns(steering->esw_ingress_root_ns[i]); + kfree(steering->esw_ingress_root_ns); + return err; +} + int mlx5_init_fs(struct mlx5_core_dev *dev) { struct mlx5_flow_steering *steering; @@ -2488,12 +2588,12 @@ int mlx5_init_fs(struct mlx5_core_dev *dev) goto err; } if (MLX5_CAP_ESW_EGRESS_ACL(dev, ft_support)) { - err = init_egress_acl_root_ns(steering); + err = init_egress_acls_root_ns(dev); if (err) goto err; } if (MLX5_CAP_ESW_INGRESS_ACL(dev, ft_support)) { - err = init_ingress_acl_root_ns(steering); + err = init_ingress_acls_root_ns(dev); if (err) goto err; } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.h b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.h index 397d24a..0526270 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.h @@ -71,8 +71,8 @@ struct mlx5_flow_steering { struct kmem_cache *ftes_cache; struct mlx5_flow_root_namespace *root_ns; struct mlx5_flow_root_namespace *fdb_root_ns; - struct mlx5_flow_root_namespace *esw_egress_root_ns; - struct mlx5_flow_root_namespace *esw_ingress_root_ns; + struct mlx5_flow_root_namespace **esw_egress_root_ns; + struct mlx5_flow_root_namespace **esw_ingress_root_ns; struct mlx5_flow_root_namespace *sniffer_tx_root_ns; struct mlx5_flow_root_namespace *sniffer_rx_root_ns; }; @@ -233,6 +233,8 @@ void mlx5_fc_queue_stats_work(struct mlx5_core_dev *dev, unsigned long delay); void mlx5_fc_update_sampling_interval(struct mlx5_core_dev *dev, unsigned long interval); +int mlx5_fc_query(struct mlx5_core_dev *dev, u16 id, + u64 *packets, u64 *bytes); int mlx5_init_fs(struct mlx5_core_dev *dev); void mlx5_cleanup_fs(struct mlx5_core_dev *dev); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_counters.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_counters.c index 89d1f865..b7ab929d 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/fs_counters.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_counters.c @@ -312,6 +312,12 @@ void mlx5_cleanup_fc_stats(struct mlx5_core_dev *dev) } } +int mlx5_fc_query(struct mlx5_core_dev *dev, u16 id, + u64 *packets, u64 *bytes) +{ + return mlx5_cmd_fc_query(dev, id, packets, bytes); +} + void mlx5_fc_query_cached(struct mlx5_fc *counter, u64 *bytes, u64 *packets, u64 *lastuse) { diff --git a/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ethtool.c b/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ethtool.c index 6f338a9..90cb50f 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ethtool.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ethtool.c @@ -254,4 +254,5 @@ const struct ethtool_ops mlx5i_ethtool_ops = { const struct ethtool_ops mlx5i_pkey_ethtool_ops = { .get_drvinfo = mlx5i_get_drvinfo, .get_link = ethtool_op_get_link, + .get_ts_info = mlx5i_get_ts_info, }; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib.c b/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib.c index a281d95..f953378 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib.c @@ -41,7 +41,6 @@ static int mlx5i_open(struct net_device *netdev); static int mlx5i_close(struct net_device *netdev); static int mlx5i_change_mtu(struct net_device *netdev, int new_mtu); -static int mlx5i_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd); static const struct net_device_ops mlx5i_netdev_ops = { .ndo_open = mlx5i_open, @@ -242,7 +241,8 @@ static void mlx5i_cleanup_tx(struct mlx5e_priv *priv) static int mlx5i_create_flow_steering(struct mlx5e_priv *priv) { - int err; + struct ttc_params ttc_params = {}; + int tt, err; priv->fs.ns = mlx5_get_flow_namespace(priv->mdev, MLX5_FLOW_NAMESPACE_KERNEL); @@ -257,14 +257,23 @@ static int mlx5i_create_flow_steering(struct mlx5e_priv *priv) priv->netdev->hw_features &= ~NETIF_F_NTUPLE; } - err = mlx5e_create_inner_ttc_table(priv); + mlx5e_set_ttc_basic_params(priv, &ttc_params); + mlx5e_set_inner_ttc_ft_params(&ttc_params); + for (tt = 0; tt < MLX5E_NUM_INDIR_TIRS; tt++) + ttc_params.indir_tirn[tt] = priv->inner_indir_tir[tt].tirn; + + err = mlx5e_create_inner_ttc_table(priv, &ttc_params, &priv->fs.inner_ttc); if (err) { netdev_err(priv->netdev, "Failed to create inner ttc table, err=%d\n", err); goto err_destroy_arfs_tables; } - err = mlx5e_create_ttc_table(priv); + mlx5e_set_ttc_ft_params(&ttc_params); + for (tt = 0; tt < MLX5E_NUM_INDIR_TIRS; tt++) + ttc_params.indir_tirn[tt] = priv->indir_tir[tt].tirn; + + err = mlx5e_create_ttc_table(priv, &ttc_params, &priv->fs.ttc); if (err) { netdev_err(priv->netdev, "Failed to create ttc table, err=%d\n", err); @@ -274,7 +283,7 @@ static int mlx5i_create_flow_steering(struct mlx5e_priv *priv) return 0; err_destroy_inner_ttc_table: - mlx5e_destroy_inner_ttc_table(priv); + mlx5e_destroy_inner_ttc_table(priv, &priv->fs.inner_ttc); err_destroy_arfs_tables: mlx5e_arfs_destroy_tables(priv); @@ -283,8 +292,8 @@ err_destroy_arfs_tables: static void mlx5i_destroy_flow_steering(struct mlx5e_priv *priv) { - mlx5e_destroy_ttc_table(priv); - mlx5e_destroy_inner_ttc_table(priv); + mlx5e_destroy_ttc_table(priv, &priv->fs.ttc); + mlx5e_destroy_inner_ttc_table(priv, &priv->fs.inner_ttc); mlx5e_arfs_destroy_tables(priv); } @@ -398,7 +407,7 @@ int mlx5i_dev_init(struct net_device *dev) return 0; } -static int mlx5i_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd) +int mlx5i_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd) { struct mlx5e_priv *priv = mlx5i_epriv(dev); @@ -486,7 +495,7 @@ static int mlx5i_close(struct net_device *netdev) mlx5_fs_remove_rx_underlay_qpn(mdev, ipriv->qp.qpn); mlx5i_uninit_underlay_qp(epriv); mlx5e_deactivate_priv_channels(epriv); - mlx5e_close_channels(&epriv->channels);; + mlx5e_close_channels(&epriv->channels); unlock: mutex_unlock(&epriv->state_lock); return 0; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib.h b/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib.h index 4900802..6d9053b 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib.h @@ -76,9 +76,10 @@ int mlx5i_pkey_del_qpn(struct net_device *netdev, u32 qpn); /* Get the net-device corresponding to the given underlay QPN */ struct net_device *mlx5i_pkey_get_netdev(struct net_device *netdev, u32 qpn); -/* Shared ndo functionts */ +/* Shared ndo functions */ int mlx5i_dev_init(struct net_device *dev); void mlx5i_dev_cleanup(struct net_device *dev); +int mlx5i_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd); /* Parent profile functions */ void mlx5i_init(struct mlx5_core_dev *mdev, diff --git a/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib_vlan.c b/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib_vlan.c index 531b02c..b69e9d8 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib_vlan.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib_vlan.c @@ -140,6 +140,7 @@ static int mlx5i_pkey_close(struct net_device *netdev); static int mlx5i_pkey_dev_init(struct net_device *dev); static void mlx5i_pkey_dev_cleanup(struct net_device *netdev); static int mlx5i_pkey_change_mtu(struct net_device *netdev, int new_mtu); +static int mlx5i_pkey_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd); static const struct net_device_ops mlx5i_pkey_netdev_ops = { .ndo_open = mlx5i_pkey_open, @@ -147,6 +148,7 @@ static const struct net_device_ops mlx5i_pkey_netdev_ops = { .ndo_init = mlx5i_pkey_dev_init, .ndo_uninit = mlx5i_pkey_dev_cleanup, .ndo_change_mtu = mlx5i_pkey_change_mtu, + .ndo_do_ioctl = mlx5i_pkey_ioctl, }; /* Child NDOs */ @@ -174,6 +176,11 @@ static int mlx5i_pkey_dev_init(struct net_device *dev) return mlx5i_dev_init(dev); } +static int mlx5i_pkey_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd) +{ + return mlx5i_ioctl(dev, ifr, cmd); +} + static void mlx5i_pkey_dev_cleanup(struct net_device *netdev) { return mlx5i_dev_cleanup(netdev); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h b/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h index b058687..394552f 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h @@ -116,6 +116,7 @@ int mlx5_destroy_scheduling_element_cmd(struct mlx5_core_dev *dev, u8 hierarchy, int mlx5_wait_for_vf_pages(struct mlx5_core_dev *dev); u64 mlx5_read_internal_timer(struct mlx5_core_dev *dev); struct mlx5_eq *mlx5_eqn2eq(struct mlx5_core_dev *dev, int eqn); +u32 mlx5_eq_poll_irq_disabled(struct mlx5_eq *eq); void mlx5_cq_tasklet_cb(unsigned long data); int mlx5_query_pcam_reg(struct mlx5_core_dev *dev, u32 *pcam, u8 feature_group, diff --git a/drivers/net/ethernet/mellanox/mlx5/core/transobj.c b/drivers/net/ethernet/mellanox/mlx5/core/transobj.c index 5e128d7..9e38343 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/transobj.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/transobj.c @@ -398,3 +398,217 @@ void mlx5_core_destroy_rqt(struct mlx5_core_dev *dev, u32 rqtn) mlx5_cmd_exec(dev, in, sizeof(in), out, sizeof(out)); } EXPORT_SYMBOL(mlx5_core_destroy_rqt); + +static int mlx5_hairpin_create_rq(struct mlx5_core_dev *mdev, + struct mlx5_hairpin_params *params, u32 *rqn) +{ + u32 in[MLX5_ST_SZ_DW(create_rq_in)] = {0}; + void *rqc, *wq; + + rqc = MLX5_ADDR_OF(create_rq_in, in, ctx); + wq = MLX5_ADDR_OF(rqc, rqc, wq); + + MLX5_SET(rqc, rqc, hairpin, 1); + MLX5_SET(rqc, rqc, state, MLX5_RQC_STATE_RST); + MLX5_SET(rqc, rqc, counter_set_id, params->q_counter); + + MLX5_SET(wq, wq, log_hairpin_data_sz, params->log_data_size); + MLX5_SET(wq, wq, log_hairpin_num_packets, params->log_num_packets); + + return mlx5_core_create_rq(mdev, in, MLX5_ST_SZ_BYTES(create_rq_in), rqn); +} + +static int mlx5_hairpin_create_sq(struct mlx5_core_dev *mdev, + struct mlx5_hairpin_params *params, u32 *sqn) +{ + u32 in[MLX5_ST_SZ_DW(create_sq_in)] = {0}; + void *sqc, *wq; + + sqc = MLX5_ADDR_OF(create_sq_in, in, ctx); + wq = MLX5_ADDR_OF(sqc, sqc, wq); + + MLX5_SET(sqc, sqc, hairpin, 1); + MLX5_SET(sqc, sqc, state, MLX5_SQC_STATE_RST); + + MLX5_SET(wq, wq, log_hairpin_data_sz, params->log_data_size); + MLX5_SET(wq, wq, log_hairpin_num_packets, params->log_num_packets); + + return mlx5_core_create_sq(mdev, in, MLX5_ST_SZ_BYTES(create_sq_in), sqn); +} + +static int mlx5_hairpin_create_queues(struct mlx5_hairpin *hp, + struct mlx5_hairpin_params *params) +{ + int i, j, err; + + for (i = 0; i < hp->num_channels; i++) { + err = mlx5_hairpin_create_rq(hp->func_mdev, params, &hp->rqn[i]); + if (err) + goto out_err_rq; + } + + for (i = 0; i < hp->num_channels; i++) { + err = mlx5_hairpin_create_sq(hp->peer_mdev, params, &hp->sqn[i]); + if (err) + goto out_err_sq; + } + + return 0; + +out_err_sq: + for (j = 0; j < i; j++) + mlx5_core_destroy_sq(hp->peer_mdev, hp->sqn[j]); + i = hp->num_channels; +out_err_rq: + for (j = 0; j < i; j++) + mlx5_core_destroy_rq(hp->func_mdev, hp->rqn[j]); + return err; +} + +static void mlx5_hairpin_destroy_queues(struct mlx5_hairpin *hp) +{ + int i; + + for (i = 0; i < hp->num_channels; i++) { + mlx5_core_destroy_rq(hp->func_mdev, hp->rqn[i]); + mlx5_core_destroy_sq(hp->peer_mdev, hp->sqn[i]); + } +} + +static int mlx5_hairpin_modify_rq(struct mlx5_core_dev *func_mdev, u32 rqn, + int curr_state, int next_state, + u16 peer_vhca, u32 peer_sq) +{ + u32 in[MLX5_ST_SZ_DW(modify_rq_in)] = {0}; + void *rqc; + + rqc = MLX5_ADDR_OF(modify_rq_in, in, ctx); + + if (next_state == MLX5_RQC_STATE_RDY) { + MLX5_SET(rqc, rqc, hairpin_peer_sq, peer_sq); + MLX5_SET(rqc, rqc, hairpin_peer_vhca, peer_vhca); + } + + MLX5_SET(modify_rq_in, in, rq_state, curr_state); + MLX5_SET(rqc, rqc, state, next_state); + + return mlx5_core_modify_rq(func_mdev, rqn, + in, MLX5_ST_SZ_BYTES(modify_rq_in)); +} + +static int mlx5_hairpin_modify_sq(struct mlx5_core_dev *peer_mdev, u32 sqn, + int curr_state, int next_state, + u16 peer_vhca, u32 peer_rq) +{ + u32 in[MLX5_ST_SZ_DW(modify_sq_in)] = {0}; + void *sqc; + + sqc = MLX5_ADDR_OF(modify_sq_in, in, ctx); + + if (next_state == MLX5_RQC_STATE_RDY) { + MLX5_SET(sqc, sqc, hairpin_peer_rq, peer_rq); + MLX5_SET(sqc, sqc, hairpin_peer_vhca, peer_vhca); + } + + MLX5_SET(modify_sq_in, in, sq_state, curr_state); + MLX5_SET(sqc, sqc, state, next_state); + + return mlx5_core_modify_sq(peer_mdev, sqn, + in, MLX5_ST_SZ_BYTES(modify_sq_in)); +} + +static int mlx5_hairpin_pair_queues(struct mlx5_hairpin *hp) +{ + int i, j, err; + + /* set peer SQs */ + for (i = 0; i < hp->num_channels; i++) { + err = mlx5_hairpin_modify_sq(hp->peer_mdev, hp->sqn[i], + MLX5_SQC_STATE_RST, MLX5_SQC_STATE_RDY, + MLX5_CAP_GEN(hp->func_mdev, vhca_id), hp->rqn[i]); + if (err) + goto err_modify_sq; + } + + /* set func RQs */ + for (i = 0; i < hp->num_channels; i++) { + err = mlx5_hairpin_modify_rq(hp->func_mdev, hp->rqn[i], + MLX5_RQC_STATE_RST, MLX5_RQC_STATE_RDY, + MLX5_CAP_GEN(hp->peer_mdev, vhca_id), hp->sqn[i]); + if (err) + goto err_modify_rq; + } + + return 0; + +err_modify_rq: + for (j = 0; j < i; j++) + mlx5_hairpin_modify_rq(hp->func_mdev, hp->rqn[j], MLX5_RQC_STATE_RDY, + MLX5_RQC_STATE_RST, 0, 0); + i = hp->num_channels; +err_modify_sq: + for (j = 0; j < i; j++) + mlx5_hairpin_modify_sq(hp->peer_mdev, hp->sqn[j], MLX5_SQC_STATE_RDY, + MLX5_SQC_STATE_RST, 0, 0); + return err; +} + +static void mlx5_hairpin_unpair_queues(struct mlx5_hairpin *hp) +{ + int i; + + /* unset func RQs */ + for (i = 0; i < hp->num_channels; i++) + mlx5_hairpin_modify_rq(hp->func_mdev, hp->rqn[i], MLX5_RQC_STATE_RDY, + MLX5_RQC_STATE_RST, 0, 0); + + /* unset peer SQs */ + for (i = 0; i < hp->num_channels; i++) + mlx5_hairpin_modify_sq(hp->peer_mdev, hp->sqn[i], MLX5_SQC_STATE_RDY, + MLX5_SQC_STATE_RST, 0, 0); +} + +struct mlx5_hairpin * +mlx5_core_hairpin_create(struct mlx5_core_dev *func_mdev, + struct mlx5_core_dev *peer_mdev, + struct mlx5_hairpin_params *params) +{ + struct mlx5_hairpin *hp; + int size, err; + + size = sizeof(*hp) + params->num_channels * 2 * sizeof(u32); + hp = kzalloc(size, GFP_KERNEL); + if (!hp) + return ERR_PTR(-ENOMEM); + + hp->func_mdev = func_mdev; + hp->peer_mdev = peer_mdev; + hp->num_channels = params->num_channels; + + hp->rqn = (void *)hp + sizeof(*hp); + hp->sqn = hp->rqn + params->num_channels; + + /* alloc and pair func --> peer hairpin */ + err = mlx5_hairpin_create_queues(hp, params); + if (err) + goto err_create_queues; + + err = mlx5_hairpin_pair_queues(hp); + if (err) + goto err_pair_queues; + + return hp; + +err_pair_queues: + mlx5_hairpin_destroy_queues(hp); +err_create_queues: + kfree(hp); + return ERR_PTR(err); +} + +void mlx5_core_hairpin_destroy(struct mlx5_hairpin *hp) +{ + mlx5_hairpin_unpair_queues(hp); + mlx5_hairpin_destroy_queues(hp); + kfree(hp); +} diff --git a/drivers/net/ethernet/mellanox/mlxsw/core.c b/drivers/net/ethernet/mellanox/mlxsw/core.c index f3315bc..3529b545 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/core.c +++ b/drivers/net/ethernet/mellanox/mlxsw/core.c @@ -113,6 +113,7 @@ struct mlxsw_core { struct mlxsw_thermal *thermal; struct mlxsw_core_port *ports; unsigned int max_ports; + bool reload_fail; unsigned long driver_priv[0]; /* driver_priv has to be always the last item */ }; @@ -962,7 +963,28 @@ mlxsw_devlink_sb_occ_tc_port_bind_get(struct devlink_port *devlink_port, pool_type, p_cur, p_max); } +static int mlxsw_devlink_core_bus_device_reload(struct devlink *devlink) +{ + struct mlxsw_core *mlxsw_core = devlink_priv(devlink); + const struct mlxsw_bus *mlxsw_bus = mlxsw_core->bus; + int err; + + if (!mlxsw_bus->reset) + return -EOPNOTSUPP; + + mlxsw_core_bus_device_unregister(mlxsw_core, true); + mlxsw_bus->reset(mlxsw_core->bus_priv); + err = mlxsw_core_bus_device_register(mlxsw_core->bus_info, + mlxsw_core->bus, + mlxsw_core->bus_priv, true, + devlink); + if (err) + mlxsw_core->reload_fail = true; + return err; +} + static const struct devlink_ops mlxsw_devlink_ops = { + .reload = mlxsw_devlink_core_bus_device_reload, .port_type_set = mlxsw_devlink_port_type_set, .port_split = mlxsw_devlink_port_split, .port_unsplit = mlxsw_devlink_port_unsplit, @@ -980,23 +1002,26 @@ static const struct devlink_ops mlxsw_devlink_ops = { int mlxsw_core_bus_device_register(const struct mlxsw_bus_info *mlxsw_bus_info, const struct mlxsw_bus *mlxsw_bus, - void *bus_priv) + void *bus_priv, bool reload, + struct devlink *devlink) { const char *device_kind = mlxsw_bus_info->device_kind; struct mlxsw_core *mlxsw_core; struct mlxsw_driver *mlxsw_driver; - struct devlink *devlink; size_t alloc_size; int err; mlxsw_driver = mlxsw_core_driver_get(device_kind); if (!mlxsw_driver) return -EINVAL; - alloc_size = sizeof(*mlxsw_core) + mlxsw_driver->priv_size; - devlink = devlink_alloc(&mlxsw_devlink_ops, alloc_size); - if (!devlink) { - err = -ENOMEM; - goto err_devlink_alloc; + + if (!reload) { + alloc_size = sizeof(*mlxsw_core) + mlxsw_driver->priv_size; + devlink = devlink_alloc(&mlxsw_devlink_ops, alloc_size); + if (!devlink) { + err = -ENOMEM; + goto err_devlink_alloc; + } } mlxsw_core = devlink_priv(devlink); @@ -1012,6 +1037,12 @@ int mlxsw_core_bus_device_register(const struct mlxsw_bus_info *mlxsw_bus_info, if (err) goto err_bus_init; + if (mlxsw_driver->resources_register && !reload) { + err = mlxsw_driver->resources_register(mlxsw_core); + if (err) + goto err_register_resources; + } + err = mlxsw_ports_init(mlxsw_core); if (err) goto err_ports_init; @@ -1032,9 +1063,11 @@ int mlxsw_core_bus_device_register(const struct mlxsw_bus_info *mlxsw_bus_info, if (err) goto err_emad_init; - err = devlink_register(devlink, mlxsw_bus_info->dev); - if (err) - goto err_devlink_register; + if (!reload) { + err = devlink_register(devlink, mlxsw_bus_info->dev); + if (err) + goto err_devlink_register; + } err = mlxsw_hwmon_init(mlxsw_core, mlxsw_bus_info, &mlxsw_core->hwmon); if (err) @@ -1057,7 +1090,8 @@ err_driver_init: mlxsw_thermal_fini(mlxsw_core->thermal); err_thermal_init: err_hwmon_init: - devlink_unregister(devlink); + if (!reload) + devlink_unregister(devlink); err_devlink_register: mlxsw_emad_fini(mlxsw_core); err_emad_init: @@ -1067,26 +1101,40 @@ err_alloc_lag_mapping: err_ports_init: mlxsw_bus->fini(bus_priv); err_bus_init: - devlink_free(devlink); + if (!reload) + devlink_resources_unregister(devlink, NULL); +err_register_resources: + if (!reload) + devlink_free(devlink); err_devlink_alloc: mlxsw_core_driver_put(device_kind); return err; } EXPORT_SYMBOL(mlxsw_core_bus_device_register); -void mlxsw_core_bus_device_unregister(struct mlxsw_core *mlxsw_core) +void mlxsw_core_bus_device_unregister(struct mlxsw_core *mlxsw_core, + bool reload) { const char *device_kind = mlxsw_core->bus_info->device_kind; struct devlink *devlink = priv_to_devlink(mlxsw_core); + if (mlxsw_core->reload_fail) + goto reload_fail; + if (mlxsw_core->driver->fini) mlxsw_core->driver->fini(mlxsw_core); mlxsw_thermal_fini(mlxsw_core->thermal); - devlink_unregister(devlink); + if (!reload) + devlink_unregister(devlink); mlxsw_emad_fini(mlxsw_core); kfree(mlxsw_core->lag.mapping); mlxsw_ports_fini(mlxsw_core); + if (!reload) + devlink_resources_unregister(devlink, NULL); mlxsw_core->bus->fini(mlxsw_core->bus_priv); + if (reload) + return; +reload_fail: devlink_free(devlink); mlxsw_core_driver_put(device_kind); } @@ -1791,6 +1839,22 @@ void mlxsw_core_flush_owq(void) } EXPORT_SYMBOL(mlxsw_core_flush_owq); +int mlxsw_core_kvd_sizes_get(struct mlxsw_core *mlxsw_core, + const struct mlxsw_config_profile *profile, + u64 *p_single_size, u64 *p_double_size, + u64 *p_linear_size) +{ + struct mlxsw_driver *driver = mlxsw_core->driver; + + if (!driver->kvd_sizes_get) + return -EINVAL; + + return driver->kvd_sizes_get(mlxsw_core, profile, + p_single_size, p_double_size, + p_linear_size); +} +EXPORT_SYMBOL(mlxsw_core_kvd_sizes_get); + static int __init mlxsw_core_module_init(void) { int err; diff --git a/drivers/net/ethernet/mellanox/mlxsw/core.h b/drivers/net/ethernet/mellanox/mlxsw/core.h index 6e966af..5ddafd7 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/core.h +++ b/drivers/net/ethernet/mellanox/mlxsw/core.h @@ -66,8 +66,9 @@ void mlxsw_core_driver_unregister(struct mlxsw_driver *mlxsw_driver); int mlxsw_core_bus_device_register(const struct mlxsw_bus_info *mlxsw_bus_info, const struct mlxsw_bus *mlxsw_bus, - void *bus_priv); -void mlxsw_core_bus_device_unregister(struct mlxsw_core *mlxsw_core); + void *bus_priv, bool reload, + struct devlink *devlink); +void mlxsw_core_bus_device_unregister(struct mlxsw_core *mlxsw_core, bool reload); struct mlxsw_tx_info { u8 local_port; @@ -308,10 +309,20 @@ struct mlxsw_driver { u32 *p_cur, u32 *p_max); void (*txhdr_construct)(struct sk_buff *skb, const struct mlxsw_tx_info *tx_info); + int (*resources_register)(struct mlxsw_core *mlxsw_core); + int (*kvd_sizes_get)(struct mlxsw_core *mlxsw_core, + const struct mlxsw_config_profile *profile, + u64 *p_single_size, u64 *p_double_size, + u64 *p_linear_size); u8 txhdr_len; const struct mlxsw_config_profile *profile; }; +int mlxsw_core_kvd_sizes_get(struct mlxsw_core *mlxsw_core, + const struct mlxsw_config_profile *profile, + u64 *p_single_size, u64 *p_double_size, + u64 *p_linear_size); + bool mlxsw_core_res_valid(struct mlxsw_core *mlxsw_core, enum mlxsw_res_id res_id); @@ -332,6 +343,7 @@ struct mlxsw_bus { const struct mlxsw_config_profile *profile, struct mlxsw_res *res); void (*fini)(void *bus_priv); + void (*reset)(void *bus_priv); bool (*skb_transmit_busy)(void *bus_priv, const struct mlxsw_tx_info *tx_info); int (*skb_transmit)(void *bus_priv, struct sk_buff *skb, diff --git a/drivers/net/ethernet/mellanox/mlxsw/core_acl_flex_actions.c b/drivers/net/ethernet/mellanox/mlxsw/core_acl_flex_actions.c index 6a979a0..b698fb4 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/core_acl_flex_actions.c +++ b/drivers/net/ethernet/mellanox/mlxsw/core_acl_flex_actions.c @@ -310,9 +310,33 @@ struct mlxsw_afa_block { struct mlxsw_afa_set *first_set; struct mlxsw_afa_set *cur_set; unsigned int cur_act_index; /* In current set. */ - struct list_head fwd_entry_ref_list; + struct list_head resource_list; /* List of resources held by actions + * in this block. + */ }; +struct mlxsw_afa_resource { + struct list_head list; + void (*destructor)(struct mlxsw_afa_block *block, + struct mlxsw_afa_resource *resource); +}; + +static void mlxsw_afa_resource_add(struct mlxsw_afa_block *block, + struct mlxsw_afa_resource *resource) +{ + list_add(&resource->list, &block->resource_list); +} + +static void mlxsw_afa_resources_destroy(struct mlxsw_afa_block *block) +{ + struct mlxsw_afa_resource *resource, *tmp; + + list_for_each_entry_safe(resource, tmp, &block->resource_list, list) { + list_del(&resource->list); + resource->destructor(block, resource); + } +} + struct mlxsw_afa_block *mlxsw_afa_block_create(struct mlxsw_afa *mlxsw_afa) { struct mlxsw_afa_block *block; @@ -320,7 +344,7 @@ struct mlxsw_afa_block *mlxsw_afa_block_create(struct mlxsw_afa *mlxsw_afa) block = kzalloc(sizeof(*block), GFP_KERNEL); if (!block) return NULL; - INIT_LIST_HEAD(&block->fwd_entry_ref_list); + INIT_LIST_HEAD(&block->resource_list); block->afa = mlxsw_afa; /* At least one action set is always present, so just create it here */ @@ -336,8 +360,6 @@ err_first_set_create: } EXPORT_SYMBOL(mlxsw_afa_block_create); -static void mlxsw_afa_fwd_entry_refs_destroy(struct mlxsw_afa_block *block); - void mlxsw_afa_block_destroy(struct mlxsw_afa_block *block) { struct mlxsw_afa_set *set = block->first_set; @@ -348,7 +370,7 @@ void mlxsw_afa_block_destroy(struct mlxsw_afa_block *block) mlxsw_afa_set_put(block->afa, set); set = next_set; } while (set); - mlxsw_afa_fwd_entry_refs_destroy(block); + mlxsw_afa_resources_destroy(block); kfree(block); } EXPORT_SYMBOL(mlxsw_afa_block_destroy); @@ -489,10 +511,29 @@ static void mlxsw_afa_fwd_entry_put(struct mlxsw_afa *mlxsw_afa, } struct mlxsw_afa_fwd_entry_ref { - struct list_head list; + struct mlxsw_afa_resource resource; struct mlxsw_afa_fwd_entry *fwd_entry; }; +static void +mlxsw_afa_fwd_entry_ref_destroy(struct mlxsw_afa_block *block, + struct mlxsw_afa_fwd_entry_ref *fwd_entry_ref) +{ + mlxsw_afa_fwd_entry_put(block->afa, fwd_entry_ref->fwd_entry); + kfree(fwd_entry_ref); +} + +static void +mlxsw_afa_fwd_entry_ref_destructor(struct mlxsw_afa_block *block, + struct mlxsw_afa_resource *resource) +{ + struct mlxsw_afa_fwd_entry_ref *fwd_entry_ref; + + fwd_entry_ref = container_of(resource, struct mlxsw_afa_fwd_entry_ref, + resource); + mlxsw_afa_fwd_entry_ref_destroy(block, fwd_entry_ref); +} + static struct mlxsw_afa_fwd_entry_ref * mlxsw_afa_fwd_entry_ref_create(struct mlxsw_afa_block *block, u8 local_port) { @@ -509,7 +550,8 @@ mlxsw_afa_fwd_entry_ref_create(struct mlxsw_afa_block *block, u8 local_port) goto err_fwd_entry_get; } fwd_entry_ref->fwd_entry = fwd_entry; - list_add(&fwd_entry_ref->list, &block->fwd_entry_ref_list); + fwd_entry_ref->resource.destructor = mlxsw_afa_fwd_entry_ref_destructor; + mlxsw_afa_resource_add(block, &fwd_entry_ref->resource); return fwd_entry_ref; err_fwd_entry_get: @@ -517,23 +559,51 @@ err_fwd_entry_get: return ERR_PTR(err); } +struct mlxsw_afa_counter { + struct mlxsw_afa_resource resource; + u32 counter_index; +}; + static void -mlxsw_afa_fwd_entry_ref_destroy(struct mlxsw_afa_block *block, - struct mlxsw_afa_fwd_entry_ref *fwd_entry_ref) +mlxsw_afa_counter_destroy(struct mlxsw_afa_block *block, + struct mlxsw_afa_counter *counter) { - list_del(&fwd_entry_ref->list); - mlxsw_afa_fwd_entry_put(block->afa, fwd_entry_ref->fwd_entry); - kfree(fwd_entry_ref); + block->afa->ops->counter_index_put(block->afa->ops_priv, + counter->counter_index); + kfree(counter); +} + +static void +mlxsw_afa_counter_destructor(struct mlxsw_afa_block *block, + struct mlxsw_afa_resource *resource) +{ + struct mlxsw_afa_counter *counter; + + counter = container_of(resource, struct mlxsw_afa_counter, resource); + mlxsw_afa_counter_destroy(block, counter); } -static void mlxsw_afa_fwd_entry_refs_destroy(struct mlxsw_afa_block *block) +static struct mlxsw_afa_counter * +mlxsw_afa_counter_create(struct mlxsw_afa_block *block) { - struct mlxsw_afa_fwd_entry_ref *fwd_entry_ref; - struct mlxsw_afa_fwd_entry_ref *tmp; + struct mlxsw_afa_counter *counter; + int err; + + counter = kzalloc(sizeof(*counter), GFP_KERNEL); + if (!counter) + return ERR_PTR(-ENOMEM); + + err = block->afa->ops->counter_index_get(block->afa->ops_priv, + &counter->counter_index); + if (err) + goto err_counter_index_get; + counter->resource.destructor = mlxsw_afa_counter_destructor; + mlxsw_afa_resource_add(block, &counter->resource); + return counter; - list_for_each_entry_safe(fwd_entry_ref, tmp, - &block->fwd_entry_ref_list, list) - mlxsw_afa_fwd_entry_ref_destroy(block, fwd_entry_ref); +err_counter_index_get: + kfree(counter); + return ERR_PTR(err); } #define MLXSW_AFA_ONE_ACTION_LEN 32 @@ -690,6 +760,16 @@ MLXSW_ITEM32(afa, trapdisc, forward_action, 0x00, 0, 4); */ MLXSW_ITEM32(afa, trapdisc, trap_id, 0x04, 0, 9); +/* afa_trapdisc_mirror_agent + * Mirror agent. + */ +MLXSW_ITEM32(afa, trapdisc, mirror_agent, 0x08, 29, 3); + +/* afa_trapdisc_mirror_enable + * Mirror enable. + */ +MLXSW_ITEM32(afa, trapdisc, mirror_enable, 0x08, 24, 1); + static inline void mlxsw_afa_trapdisc_pack(char *payload, enum mlxsw_afa_trapdisc_trap_action trap_action, @@ -701,6 +781,14 @@ mlxsw_afa_trapdisc_pack(char *payload, mlxsw_afa_trapdisc_trap_id_set(payload, trap_id); } +static inline void +mlxsw_afa_trapdisc_mirror_pack(char *payload, bool mirror_enable, + u8 mirror_agent) +{ + mlxsw_afa_trapdisc_mirror_enable_set(payload, mirror_enable); + mlxsw_afa_trapdisc_mirror_agent_set(payload, mirror_agent); +} + int mlxsw_afa_block_append_drop(struct mlxsw_afa_block *block) { char *act = mlxsw_afa_block_append_action(block, @@ -746,6 +834,104 @@ int mlxsw_afa_block_append_trap_and_forward(struct mlxsw_afa_block *block, } EXPORT_SYMBOL(mlxsw_afa_block_append_trap_and_forward); +struct mlxsw_afa_mirror { + struct mlxsw_afa_resource resource; + int span_id; + u8 local_in_port; + u8 local_out_port; + bool ingress; +}; + +static void +mlxsw_afa_mirror_destroy(struct mlxsw_afa_block *block, + struct mlxsw_afa_mirror *mirror) +{ + block->afa->ops->mirror_del(block->afa->ops_priv, + mirror->local_in_port, + mirror->local_out_port, + mirror->ingress); + kfree(mirror); +} + +static void +mlxsw_afa_mirror_destructor(struct mlxsw_afa_block *block, + struct mlxsw_afa_resource *resource) +{ + struct mlxsw_afa_mirror *mirror; + + mirror = container_of(resource, struct mlxsw_afa_mirror, resource); + mlxsw_afa_mirror_destroy(block, mirror); +} + +static struct mlxsw_afa_mirror * +mlxsw_afa_mirror_create(struct mlxsw_afa_block *block, + u8 local_in_port, u8 local_out_port, + bool ingress) +{ + struct mlxsw_afa_mirror *mirror; + int err; + + mirror = kzalloc(sizeof(*mirror), GFP_KERNEL); + if (!mirror) + return ERR_PTR(-ENOMEM); + + err = block->afa->ops->mirror_add(block->afa->ops_priv, + local_in_port, local_out_port, + ingress, &mirror->span_id); + if (err) + goto err_mirror_add; + + mirror->ingress = ingress; + mirror->local_out_port = local_out_port; + mirror->local_in_port = local_in_port; + mirror->resource.destructor = mlxsw_afa_mirror_destructor; + mlxsw_afa_resource_add(block, &mirror->resource); + return mirror; + +err_mirror_add: + kfree(mirror); + return ERR_PTR(err); +} + +static int +mlxsw_afa_block_append_allocated_mirror(struct mlxsw_afa_block *block, + u8 mirror_agent) +{ + char *act = mlxsw_afa_block_append_action(block, + MLXSW_AFA_TRAPDISC_CODE, + MLXSW_AFA_TRAPDISC_SIZE); + if (!act) + return -ENOBUFS; + mlxsw_afa_trapdisc_pack(act, MLXSW_AFA_TRAPDISC_TRAP_ACTION_NOP, + MLXSW_AFA_TRAPDISC_FORWARD_ACTION_FORWARD, 0); + mlxsw_afa_trapdisc_mirror_pack(act, true, mirror_agent); + return 0; +} + +int +mlxsw_afa_block_append_mirror(struct mlxsw_afa_block *block, + u8 local_in_port, u8 local_out_port, bool ingress) +{ + struct mlxsw_afa_mirror *mirror; + int err; + + mirror = mlxsw_afa_mirror_create(block, local_in_port, local_out_port, + ingress); + if (IS_ERR(mirror)) + return PTR_ERR(mirror); + + err = mlxsw_afa_block_append_allocated_mirror(block, mirror->span_id); + if (err) + goto err_append_allocated_mirror; + + return 0; + +err_append_allocated_mirror: + mlxsw_afa_mirror_destroy(block, mirror); + return err; +} +EXPORT_SYMBOL(mlxsw_afa_block_append_mirror); + /* Forwarding Action * ----------------- * Forwarding Action can be used to implement Policy Based Switching (PBS) @@ -853,11 +1039,10 @@ mlxsw_afa_polcnt_pack(char *payload, mlxsw_afa_polcnt_counter_index_set(payload, counter_index); } -int mlxsw_afa_block_append_counter(struct mlxsw_afa_block *block, - u32 counter_index) +int mlxsw_afa_block_append_allocated_counter(struct mlxsw_afa_block *block, + u32 counter_index) { - char *act = mlxsw_afa_block_append_action(block, - MLXSW_AFA_POLCNT_CODE, + char *act = mlxsw_afa_block_append_action(block, MLXSW_AFA_POLCNT_CODE, MLXSW_AFA_POLCNT_SIZE); if (!act) return -ENOBUFS; @@ -865,6 +1050,32 @@ int mlxsw_afa_block_append_counter(struct mlxsw_afa_block *block, counter_index); return 0; } +EXPORT_SYMBOL(mlxsw_afa_block_append_allocated_counter); + +int mlxsw_afa_block_append_counter(struct mlxsw_afa_block *block, + u32 *p_counter_index) +{ + struct mlxsw_afa_counter *counter; + u32 counter_index; + int err; + + counter = mlxsw_afa_counter_create(block); + if (IS_ERR(counter)) + return PTR_ERR(counter); + counter_index = counter->counter_index; + + err = mlxsw_afa_block_append_allocated_counter(block, counter_index); + if (err) + goto err_append_allocated_counter; + + if (p_counter_index) + *p_counter_index = counter_index; + return 0; + +err_append_allocated_counter: + mlxsw_afa_counter_destroy(block, counter); + return err; +} EXPORT_SYMBOL(mlxsw_afa_block_append_counter); /* Virtual Router and Forwarding Domain Action diff --git a/drivers/net/ethernet/mellanox/mlxsw/core_acl_flex_actions.h b/drivers/net/ethernet/mellanox/mlxsw/core_acl_flex_actions.h index a8d3314..4313229 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/core_acl_flex_actions.h +++ b/drivers/net/ethernet/mellanox/mlxsw/core_acl_flex_actions.h @@ -46,6 +46,12 @@ struct mlxsw_afa_ops { void (*kvdl_set_del)(void *priv, u32 kvdl_index, bool is_first); int (*kvdl_fwd_entry_add)(void *priv, u32 *p_kvdl_index, u8 local_port); void (*kvdl_fwd_entry_del)(void *priv, u32 kvdl_index); + int (*counter_index_get)(void *priv, unsigned int *p_counter_index); + void (*counter_index_put)(void *priv, unsigned int counter_index); + int (*mirror_add)(void *priv, u8 locol_in_port, u8 local_out_port, + bool ingress, int *p_span_id); + void (*mirror_del)(void *priv, u8 locol_in_port, u8 local_out_port, + bool ingress); }; struct mlxsw_afa *mlxsw_afa_create(unsigned int max_acts_per_set, @@ -63,12 +69,17 @@ int mlxsw_afa_block_append_drop(struct mlxsw_afa_block *block); int mlxsw_afa_block_append_trap(struct mlxsw_afa_block *block, u16 trap_id); int mlxsw_afa_block_append_trap_and_forward(struct mlxsw_afa_block *block, u16 trap_id); +int mlxsw_afa_block_append_mirror(struct mlxsw_afa_block *block, + u8 local_in_port, u8 local_out_port, + bool ingress); int mlxsw_afa_block_append_fwd(struct mlxsw_afa_block *block, u8 local_port, bool in_port); int mlxsw_afa_block_append_vlan_modify(struct mlxsw_afa_block *block, u16 vid, u8 pcp, u8 et); +int mlxsw_afa_block_append_allocated_counter(struct mlxsw_afa_block *block, + u32 counter_index); int mlxsw_afa_block_append_counter(struct mlxsw_afa_block *block, - u32 counter_index); + u32 *p_counter_index); int mlxsw_afa_block_append_fid_set(struct mlxsw_afa_block *block, u16 fid); int mlxsw_afa_block_append_mcrouter(struct mlxsw_afa_block *block, u16 expected_irif, u16 min_mtu, diff --git a/drivers/net/ethernet/mellanox/mlxsw/i2c.c b/drivers/net/ethernet/mellanox/mlxsw/i2c.c index c0dcfa0..25f9915 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/i2c.c +++ b/drivers/net/ethernet/mellanox/mlxsw/i2c.c @@ -539,7 +539,8 @@ static int mlxsw_i2c_probe(struct i2c_client *client, mlxsw_i2c->dev = &client->dev; err = mlxsw_core_bus_device_register(&mlxsw_i2c->bus_info, - &mlxsw_i2c_bus, mlxsw_i2c); + &mlxsw_i2c_bus, mlxsw_i2c, false, + NULL); if (err) { dev_err(&client->dev, "Fail to register core bus\n"); return err; @@ -557,7 +558,7 @@ static int mlxsw_i2c_remove(struct i2c_client *client) { struct mlxsw_i2c *mlxsw_i2c = i2c_get_clientdata(client); - mlxsw_core_bus_device_unregister(mlxsw_i2c->core); + mlxsw_core_bus_device_unregister(mlxsw_i2c->core, false); mutex_destroy(&mlxsw_i2c->cmd.lock); return 0; diff --git a/drivers/net/ethernet/mellanox/mlxsw/item.h b/drivers/net/ethernet/mellanox/mlxsw/item.h index 28427f0..31c886e 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/item.h +++ b/drivers/net/ethernet/mellanox/mlxsw/item.h @@ -42,7 +42,7 @@ struct mlxsw_item { unsigned short offset; /* bytes in container */ - unsigned short step; /* step in bytes for indexed items */ + short step; /* step in bytes for indexed items */ unsigned short in_step_offset; /* offset within one step */ unsigned char shift; /* shift in bits */ unsigned char element_size; /* size of element in bit array */ diff --git a/drivers/net/ethernet/mellanox/mlxsw/pci.c b/drivers/net/ethernet/mellanox/mlxsw/pci.c index 6ef20e5..85faa87 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/pci.c +++ b/drivers/net/ethernet/mellanox/mlxsw/pci.c @@ -154,6 +154,7 @@ struct mlxsw_pci { } comp; } cmd; struct mlxsw_bus_info bus_info; + const struct pci_device_id *id; }; static void mlxsw_pci_queue_tasklet_schedule(struct mlxsw_pci_queue *q) @@ -1052,38 +1053,18 @@ static int mlxsw_pci_resources_query(struct mlxsw_pci *mlxsw_pci, char *mbox, } static int -mlxsw_pci_profile_get_kvd_sizes(const struct mlxsw_config_profile *profile, +mlxsw_pci_profile_get_kvd_sizes(const struct mlxsw_pci *mlxsw_pci, + const struct mlxsw_config_profile *profile, struct mlxsw_res *res) { - u32 single_size, double_size, linear_size; - - if (!MLXSW_RES_VALID(res, KVD_SINGLE_MIN_SIZE) || - !MLXSW_RES_VALID(res, KVD_DOUBLE_MIN_SIZE) || - !profile->used_kvd_split_data) - return -EIO; - - linear_size = profile->kvd_linear_size; + u64 single_size, double_size, linear_size; + int err; - /* The hash part is what left of the kvd without the - * linear part. It is split to the single size and - * double size by the parts ratio from the profile. - * Both sizes must be a multiplications of the - * granularity from the profile. - */ - double_size = MLXSW_RES_GET(res, KVD_SIZE) - linear_size; - double_size *= profile->kvd_hash_double_parts; - double_size /= profile->kvd_hash_double_parts + - profile->kvd_hash_single_parts; - double_size /= profile->kvd_hash_granularity; - double_size *= profile->kvd_hash_granularity; - single_size = MLXSW_RES_GET(res, KVD_SIZE) - double_size - - linear_size; - - /* Check results are legal. */ - if (single_size < MLXSW_RES_GET(res, KVD_SINGLE_MIN_SIZE) || - double_size < MLXSW_RES_GET(res, KVD_DOUBLE_MIN_SIZE) || - MLXSW_RES_GET(res, KVD_SIZE) < linear_size) - return -EIO; + err = mlxsw_core_kvd_sizes_get(mlxsw_pci->core, profile, + &single_size, &double_size, + &linear_size); + if (err) + return err; MLXSW_RES_SET(res, KVD_SINGLE_SIZE, single_size); MLXSW_RES_SET(res, KVD_DOUBLE_SIZE, double_size); @@ -1184,7 +1165,7 @@ static int mlxsw_pci_config_profile(struct mlxsw_pci *mlxsw_pci, char *mbox, mbox, profile->adaptive_routing_group_cap); } if (MLXSW_RES_VALID(res, KVD_SIZE)) { - err = mlxsw_pci_profile_get_kvd_sizes(profile, res); + err = mlxsw_pci_profile_get_kvd_sizes(mlxsw_pci, profile, res); if (err) return err; @@ -1622,16 +1603,6 @@ static int mlxsw_pci_cmd_exec(void *bus_priv, u16 opcode, u8 opcode_mod, return err; } -static const struct mlxsw_bus mlxsw_pci_bus = { - .kind = "pci", - .init = mlxsw_pci_init, - .fini = mlxsw_pci_fini, - .skb_transmit_busy = mlxsw_pci_skb_transmit_busy, - .skb_transmit = mlxsw_pci_skb_transmit, - .cmd_exec = mlxsw_pci_cmd_exec, - .features = MLXSW_BUS_F_TXRX, -}; - static int mlxsw_pci_sw_reset(struct mlxsw_pci *mlxsw_pci, const struct pci_device_id *id) { @@ -1660,6 +1631,41 @@ static int mlxsw_pci_sw_reset(struct mlxsw_pci *mlxsw_pci, return 0; } +static void mlxsw_pci_free_irq_vectors(struct mlxsw_pci *mlxsw_pci) +{ + pci_free_irq_vectors(mlxsw_pci->pdev); +} + +static int mlxsw_pci_alloc_irq_vectors(struct mlxsw_pci *mlxsw_pci) +{ + int err; + + err = pci_alloc_irq_vectors(mlxsw_pci->pdev, 1, 1, PCI_IRQ_MSIX); + if (err < 0) + dev_err(&mlxsw_pci->pdev->dev, "MSI-X init failed\n"); + return err; +} + +static void mlxsw_pci_reset(void *bus_priv) +{ + struct mlxsw_pci *mlxsw_pci = bus_priv; + + mlxsw_pci_free_irq_vectors(mlxsw_pci); + mlxsw_pci_sw_reset(mlxsw_pci, mlxsw_pci->id); + mlxsw_pci_alloc_irq_vectors(mlxsw_pci); +} + +static const struct mlxsw_bus mlxsw_pci_bus = { + .kind = "pci", + .init = mlxsw_pci_init, + .fini = mlxsw_pci_fini, + .skb_transmit_busy = mlxsw_pci_skb_transmit_busy, + .skb_transmit = mlxsw_pci_skb_transmit, + .cmd_exec = mlxsw_pci_cmd_exec, + .features = MLXSW_BUS_F_TXRX, + .reset = mlxsw_pci_reset, +}; + static int mlxsw_pci_probe(struct pci_dev *pdev, const struct pci_device_id *id) { const char *driver_name = pdev->driver->name; @@ -1721,7 +1727,7 @@ static int mlxsw_pci_probe(struct pci_dev *pdev, const struct pci_device_id *id) goto err_sw_reset; } - err = pci_alloc_irq_vectors(pdev, 1, 1, PCI_IRQ_MSIX); + err = mlxsw_pci_alloc_irq_vectors(mlxsw_pci); if (err < 0) { dev_err(&pdev->dev, "MSI-X init failed\n"); goto err_msix_init; @@ -1730,9 +1736,11 @@ static int mlxsw_pci_probe(struct pci_dev *pdev, const struct pci_device_id *id) mlxsw_pci->bus_info.device_kind = driver_name; mlxsw_pci->bus_info.device_name = pci_name(mlxsw_pci->pdev); mlxsw_pci->bus_info.dev = &pdev->dev; + mlxsw_pci->id = id; err = mlxsw_core_bus_device_register(&mlxsw_pci->bus_info, - &mlxsw_pci_bus, mlxsw_pci); + &mlxsw_pci_bus, mlxsw_pci, false, + NULL); if (err) { dev_err(&pdev->dev, "cannot register bus device\n"); goto err_bus_device_register; @@ -1741,7 +1749,7 @@ static int mlxsw_pci_probe(struct pci_dev *pdev, const struct pci_device_id *id) return 0; err_bus_device_register: - pci_free_irq_vectors(mlxsw_pci->pdev); + mlxsw_pci_free_irq_vectors(mlxsw_pci); err_msix_init: err_sw_reset: iounmap(mlxsw_pci->hw_addr); @@ -1760,8 +1768,8 @@ static void mlxsw_pci_remove(struct pci_dev *pdev) { struct mlxsw_pci *mlxsw_pci = pci_get_drvdata(pdev); - mlxsw_core_bus_device_unregister(mlxsw_pci->core); - pci_free_irq_vectors(mlxsw_pci->pdev); + mlxsw_core_bus_device_unregister(mlxsw_pci->core, false); + mlxsw_pci_free_irq_vectors(mlxsw_pci); iounmap(mlxsw_pci->hw_addr); pci_release_regions(mlxsw_pci->pdev); pci_disable_device(mlxsw_pci->pdev); diff --git a/drivers/net/ethernet/mellanox/mlxsw/reg.h b/drivers/net/ethernet/mellanox/mlxsw/reg.h index 6c4e08b..0e08be4 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/reg.h +++ b/drivers/net/ethernet/mellanox/mlxsw/reg.h @@ -4827,6 +4827,42 @@ static inline void mlxsw_reg_ratr_counter_pack(char *payload, u64 counter_index, mlxsw_reg_ratr_counter_set_type_set(payload, set_type); } +/* RDPM - Router DSCP to Priority Mapping + * -------------------------------------- + * Controls the mapping from DSCP field to switch priority on routed packets + */ +#define MLXSW_REG_RDPM_ID 0x8009 +#define MLXSW_REG_RDPM_BASE_LEN 0x00 +#define MLXSW_REG_RDPM_DSCP_ENTRY_REC_LEN 0x01 +#define MLXSW_REG_RDPM_DSCP_ENTRY_REC_MAX_COUNT 64 +#define MLXSW_REG_RDPM_LEN 0x40 +#define MLXSW_REG_RDPM_LAST_ENTRY (MLXSW_REG_RDPM_BASE_LEN + \ + MLXSW_REG_RDPM_LEN - \ + MLXSW_REG_RDPM_DSCP_ENTRY_REC_LEN) + +MLXSW_REG_DEFINE(rdpm, MLXSW_REG_RDPM_ID, MLXSW_REG_RDPM_LEN); + +/* reg_dscp_entry_e + * Enable update of the specific entry + * Access: Index + */ +MLXSW_ITEM8_INDEXED(reg, rdpm, dscp_entry_e, MLXSW_REG_RDPM_LAST_ENTRY, 7, 1, + -MLXSW_REG_RDPM_DSCP_ENTRY_REC_LEN, 0x00, false); + +/* reg_dscp_entry_prio + * Switch Priority + * Access: RW + */ +MLXSW_ITEM8_INDEXED(reg, rdpm, dscp_entry_prio, MLXSW_REG_RDPM_LAST_ENTRY, 0, 4, + -MLXSW_REG_RDPM_DSCP_ENTRY_REC_LEN, 0x00, false); + +static inline void mlxsw_reg_rdpm_pack(char *payload, unsigned short index, + u8 prio) +{ + mlxsw_reg_rdpm_dscp_entry_e_set(payload, index, 1); + mlxsw_reg_rdpm_dscp_entry_prio_set(payload, index, prio); +} + /* RICNT - Router Interface Counter Register * ----------------------------------------- * The RICNT register retrieves per port performance counters @@ -7640,6 +7676,7 @@ static const struct mlxsw_reg_info *mlxsw_reg_infos[] = { MLXSW_REG(rtar), MLXSW_REG(ratr), MLXSW_REG(rtdp), + MLXSW_REG(rdpm), MLXSW_REG(ricnt), MLXSW_REG(rrcr), MLXSW_REG(ralta), diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum.c index c3837ca..3dcc58d 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum.c @@ -76,12 +76,7 @@ #define MLXSW_FWREV_MAJOR 13 #define MLXSW_FWREV_MINOR 1530 #define MLXSW_FWREV_SUBMINOR 152 - -static const struct mlxsw_fw_rev mlxsw_sp_supported_fw_rev = { - .major = MLXSW_FWREV_MAJOR, - .minor = MLXSW_FWREV_MINOR, - .subminor = MLXSW_FWREV_SUBMINOR -}; +#define MLXSW_FWREV_MINOR_TO_BRANCH(minor) ((minor) / 100) #define MLXSW_SP_FW_FILENAME \ "mellanox/mlxsw_spectrum-" __stringify(MLXSW_FWREV_MAJOR) \ @@ -339,28 +334,25 @@ static int mlxsw_sp_firmware_flash(struct mlxsw_sp *mlxsw_sp, return mlxfw_firmware_flash(&mlxsw_sp_mlxfw_dev.mlxfw_dev, firmware); } -static bool mlxsw_sp_fw_rev_ge(const struct mlxsw_fw_rev *a, - const struct mlxsw_fw_rev *b) -{ - if (a->major != b->major) - return a->major > b->major; - if (a->minor != b->minor) - return a->minor > b->minor; - return a->subminor >= b->subminor; -} - static int mlxsw_sp_fw_rev_validate(struct mlxsw_sp *mlxsw_sp) { const struct mlxsw_fw_rev *rev = &mlxsw_sp->bus_info->fw_rev; const struct firmware *firmware; int err; - if (mlxsw_sp_fw_rev_ge(rev, &mlxsw_sp_supported_fw_rev)) + /* Validate driver & FW are compatible */ + if (rev->major != MLXSW_FWREV_MAJOR) { + WARN(1, "Mismatch in major FW version [%d:%d] is never expected; Please contact support\n", + rev->major, MLXSW_FWREV_MAJOR); + return -EINVAL; + } + if (MLXSW_FWREV_MINOR_TO_BRANCH(rev->minor) == + MLXSW_FWREV_MINOR_TO_BRANCH(MLXSW_FWREV_MINOR)) return 0; - dev_info(mlxsw_sp->bus_info->dev, "The firmware version %d.%d.%d out of data\n", + dev_info(mlxsw_sp->bus_info->dev, "The firmware version %d.%d.%d is incompatible with the driver\n", rev->major, rev->minor, rev->subminor); - dev_info(mlxsw_sp->bus_info->dev, "Upgrading firmware using file %s\n", + dev_info(mlxsw_sp->bus_info->dev, "Flashing firmware using file %s\n", MLXSW_SP_FW_FILENAME); err = request_firmware_direct(&firmware, MLXSW_SP_FW_FILENAME, @@ -576,7 +568,7 @@ static void mlxsw_sp_span_entry_destroy(struct mlxsw_sp *mlxsw_sp, span_entry->used = false; } -static struct mlxsw_sp_span_entry * +struct mlxsw_sp_span_entry * mlxsw_sp_span_entry_find(struct mlxsw_sp *mlxsw_sp, u8 local_port) { int i; @@ -677,13 +669,28 @@ mlxsw_sp_span_entry_bound_port_find(struct mlxsw_sp_port *port, static int mlxsw_sp_span_inspected_port_bind(struct mlxsw_sp_port *port, struct mlxsw_sp_span_entry *span_entry, - enum mlxsw_sp_span_type type) + enum mlxsw_sp_span_type type, + bool bind) { - struct mlxsw_sp_span_inspected_port *inspected_port; struct mlxsw_sp *mlxsw_sp = port->mlxsw_sp; char mpar_pl[MLXSW_REG_MPAR_LEN]; - char sbib_pl[MLXSW_REG_SBIB_LEN]; int pa_id = span_entry->id; + + /* bind the port to the SPAN entry */ + mlxsw_reg_mpar_pack(mpar_pl, port->local_port, + (enum mlxsw_reg_mpar_i_e) type, bind, pa_id); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(mpar), mpar_pl); +} + +static int +mlxsw_sp_span_inspected_port_add(struct mlxsw_sp_port *port, + struct mlxsw_sp_span_entry *span_entry, + enum mlxsw_sp_span_type type, + bool bind) +{ + struct mlxsw_sp_span_inspected_port *inspected_port; + struct mlxsw_sp *mlxsw_sp = port->mlxsw_sp; + char sbib_pl[MLXSW_REG_SBIB_LEN]; int err; /* if it is an egress SPAN, bind a shared buffer to it */ @@ -699,12 +706,12 @@ mlxsw_sp_span_inspected_port_bind(struct mlxsw_sp_port *port, } } - /* bind the port to the SPAN entry */ - mlxsw_reg_mpar_pack(mpar_pl, port->local_port, - (enum mlxsw_reg_mpar_i_e) type, true, pa_id); - err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(mpar), mpar_pl); - if (err) - goto err_mpar_reg_write; + if (bind) { + err = mlxsw_sp_span_inspected_port_bind(port, span_entry, type, + true); + if (err) + goto err_port_bind; + } inspected_port = kzalloc(sizeof(*inspected_port), GFP_KERNEL); if (!inspected_port) { @@ -717,8 +724,11 @@ mlxsw_sp_span_inspected_port_bind(struct mlxsw_sp_port *port, return 0; -err_mpar_reg_write: err_inspected_port_alloc: + if (bind) + mlxsw_sp_span_inspected_port_bind(port, span_entry, type, + false); +err_port_bind: if (type == MLXSW_SP_SPAN_EGRESS) { mlxsw_reg_sbib_pack(sbib_pl, port->local_port, 0); mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(sbib), sbib_pl); @@ -727,25 +737,22 @@ err_inspected_port_alloc: } static void -mlxsw_sp_span_inspected_port_unbind(struct mlxsw_sp_port *port, - struct mlxsw_sp_span_entry *span_entry, - enum mlxsw_sp_span_type type) +mlxsw_sp_span_inspected_port_del(struct mlxsw_sp_port *port, + struct mlxsw_sp_span_entry *span_entry, + enum mlxsw_sp_span_type type, + bool bind) { struct mlxsw_sp_span_inspected_port *inspected_port; struct mlxsw_sp *mlxsw_sp = port->mlxsw_sp; - char mpar_pl[MLXSW_REG_MPAR_LEN]; char sbib_pl[MLXSW_REG_SBIB_LEN]; - int pa_id = span_entry->id; inspected_port = mlxsw_sp_span_entry_bound_port_find(port, span_entry); if (!inspected_port) return; - /* remove the inspected port */ - mlxsw_reg_mpar_pack(mpar_pl, port->local_port, - (enum mlxsw_reg_mpar_i_e) type, false, pa_id); - mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(mpar), mpar_pl); - + if (bind) + mlxsw_sp_span_inspected_port_bind(port, span_entry, type, + false); /* remove the SBIB buffer if it was egress SPAN */ if (type == MLXSW_SP_SPAN_EGRESS) { mlxsw_reg_sbib_pack(sbib_pl, port->local_port, 0); @@ -758,9 +765,9 @@ mlxsw_sp_span_inspected_port_unbind(struct mlxsw_sp_port *port, kfree(inspected_port); } -static int mlxsw_sp_span_mirror_add(struct mlxsw_sp_port *from, - struct mlxsw_sp_port *to, - enum mlxsw_sp_span_type type) +int mlxsw_sp_span_mirror_add(struct mlxsw_sp_port *from, + struct mlxsw_sp_port *to, + enum mlxsw_sp_span_type type, bool bind) { struct mlxsw_sp *mlxsw_sp = from->mlxsw_sp; struct mlxsw_sp_span_entry *span_entry; @@ -773,7 +780,7 @@ static int mlxsw_sp_span_mirror_add(struct mlxsw_sp_port *from, netdev_dbg(from->dev, "Adding inspected port to SPAN entry %d\n", span_entry->id); - err = mlxsw_sp_span_inspected_port_bind(from, span_entry, type); + err = mlxsw_sp_span_inspected_port_add(from, span_entry, type, bind); if (err) goto err_port_bind; @@ -784,9 +791,8 @@ err_port_bind: return err; } -static void mlxsw_sp_span_mirror_remove(struct mlxsw_sp_port *from, - u8 destination_port, - enum mlxsw_sp_span_type type) +void mlxsw_sp_span_mirror_del(struct mlxsw_sp_port *from, u8 destination_port, + enum mlxsw_sp_span_type type, bool bind) { struct mlxsw_sp_span_entry *span_entry; @@ -799,7 +805,7 @@ static void mlxsw_sp_span_mirror_remove(struct mlxsw_sp_port *from, netdev_dbg(from->dev, "removing inspected port from SPAN entry %d\n", span_entry->id); - mlxsw_sp_span_inspected_port_unbind(from, span_entry, type); + mlxsw_sp_span_inspected_port_del(from, span_entry, type, bind); } static int mlxsw_sp_port_sample_set(struct mlxsw_sp_port *mlxsw_sp_port, @@ -1571,14 +1577,11 @@ mlxsw_sp_port_add_cls_matchall_mirror(struct mlxsw_sp_port *mlxsw_sp_port, const struct tc_action *a, bool ingress) { - struct net *net = dev_net(mlxsw_sp_port->dev); enum mlxsw_sp_span_type span_type; struct mlxsw_sp_port *to_port; struct net_device *to_dev; - int ifindex; - ifindex = tcf_mirred_ifindex(a); - to_dev = __dev_get_by_index(net, ifindex); + to_dev = tcf_mirred_dev(a); if (!to_dev) { netdev_err(mlxsw_sp_port->dev, "Could not find requested device\n"); return -EINVAL; @@ -1593,7 +1596,8 @@ mlxsw_sp_port_add_cls_matchall_mirror(struct mlxsw_sp_port *mlxsw_sp_port, mirror->to_local_port = to_port->local_port; mirror->ingress = ingress; span_type = ingress ? MLXSW_SP_SPAN_INGRESS : MLXSW_SP_SPAN_EGRESS; - return mlxsw_sp_span_mirror_add(mlxsw_sp_port, to_port, span_type); + return mlxsw_sp_span_mirror_add(mlxsw_sp_port, to_port, span_type, + true); } static void @@ -1604,8 +1608,8 @@ mlxsw_sp_port_del_cls_matchall_mirror(struct mlxsw_sp_port *mlxsw_sp_port, span_type = mirror->ingress ? MLXSW_SP_SPAN_INGRESS : MLXSW_SP_SPAN_EGRESS; - mlxsw_sp_span_mirror_remove(mlxsw_sp_port, mirror->to_local_port, - span_type); + mlxsw_sp_span_mirror_del(mlxsw_sp_port, mirror->to_local_port, + span_type, true); } static int @@ -1734,9 +1738,6 @@ static int mlxsw_sp_setup_tc_cls_matchall(struct mlxsw_sp_port *mlxsw_sp_port, struct tc_cls_matchall_offload *f, bool ingress) { - if (f->common.chain_index) - return -EOPNOTSUPP; - switch (f->command) { case TC_CLSMATCHALL_REPLACE: return mlxsw_sp_port_add_cls_matchall(mlxsw_sp_port, f, @@ -1750,72 +1751,187 @@ static int mlxsw_sp_setup_tc_cls_matchall(struct mlxsw_sp_port *mlxsw_sp_port, } static int -mlxsw_sp_setup_tc_cls_flower(struct mlxsw_sp_port *mlxsw_sp_port, - struct tc_cls_flower_offload *f, - bool ingress) +mlxsw_sp_setup_tc_cls_flower(struct mlxsw_sp_acl_block *acl_block, + struct tc_cls_flower_offload *f) { + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_acl_block_mlxsw_sp(acl_block); + switch (f->command) { case TC_CLSFLOWER_REPLACE: - return mlxsw_sp_flower_replace(mlxsw_sp_port, ingress, f); + return mlxsw_sp_flower_replace(mlxsw_sp, acl_block, f); case TC_CLSFLOWER_DESTROY: - mlxsw_sp_flower_destroy(mlxsw_sp_port, ingress, f); + mlxsw_sp_flower_destroy(mlxsw_sp, acl_block, f); return 0; case TC_CLSFLOWER_STATS: - return mlxsw_sp_flower_stats(mlxsw_sp_port, ingress, f); + return mlxsw_sp_flower_stats(mlxsw_sp, acl_block, f); default: return -EOPNOTSUPP; } } -static int mlxsw_sp_setup_tc_block_cb(enum tc_setup_type type, void *type_data, - void *cb_priv, bool ingress) +static int mlxsw_sp_setup_tc_block_cb_matchall(enum tc_setup_type type, + void *type_data, + void *cb_priv, bool ingress) { struct mlxsw_sp_port *mlxsw_sp_port = cb_priv; - if (!tc_can_offload(mlxsw_sp_port->dev)) - return -EOPNOTSUPP; - switch (type) { case TC_SETUP_CLSMATCHALL: + if (!tc_cls_can_offload_and_chain0(mlxsw_sp_port->dev, + type_data)) + return -EOPNOTSUPP; + return mlxsw_sp_setup_tc_cls_matchall(mlxsw_sp_port, type_data, ingress); case TC_SETUP_CLSFLOWER: - return mlxsw_sp_setup_tc_cls_flower(mlxsw_sp_port, type_data, - ingress); + return 0; + default: + return -EOPNOTSUPP; + } +} + +static int mlxsw_sp_setup_tc_block_cb_matchall_ig(enum tc_setup_type type, + void *type_data, + void *cb_priv) +{ + return mlxsw_sp_setup_tc_block_cb_matchall(type, type_data, + cb_priv, true); +} + +static int mlxsw_sp_setup_tc_block_cb_matchall_eg(enum tc_setup_type type, + void *type_data, + void *cb_priv) +{ + return mlxsw_sp_setup_tc_block_cb_matchall(type, type_data, + cb_priv, false); +} + +static int mlxsw_sp_setup_tc_block_cb_flower(enum tc_setup_type type, + void *type_data, void *cb_priv) +{ + struct mlxsw_sp_acl_block *acl_block = cb_priv; + + switch (type) { + case TC_SETUP_CLSMATCHALL: + return 0; + case TC_SETUP_CLSFLOWER: + if (mlxsw_sp_acl_block_disabled(acl_block)) + return -EOPNOTSUPP; + + return mlxsw_sp_setup_tc_cls_flower(acl_block, type_data); default: return -EOPNOTSUPP; } } -static int mlxsw_sp_setup_tc_block_cb_ig(enum tc_setup_type type, - void *type_data, void *cb_priv) +static int +mlxsw_sp_setup_tc_block_flower_bind(struct mlxsw_sp_port *mlxsw_sp_port, + struct tcf_block *block, bool ingress) { - return mlxsw_sp_setup_tc_block_cb(type, type_data, cb_priv, true); + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + struct mlxsw_sp_acl_block *acl_block; + struct tcf_block_cb *block_cb; + int err; + + block_cb = tcf_block_cb_lookup(block, mlxsw_sp_setup_tc_block_cb_flower, + mlxsw_sp); + if (!block_cb) { + acl_block = mlxsw_sp_acl_block_create(mlxsw_sp, block->net); + if (!acl_block) + return -ENOMEM; + block_cb = __tcf_block_cb_register(block, + mlxsw_sp_setup_tc_block_cb_flower, + mlxsw_sp, acl_block); + if (IS_ERR(block_cb)) { + err = PTR_ERR(block_cb); + goto err_cb_register; + } + } else { + acl_block = tcf_block_cb_priv(block_cb); + } + tcf_block_cb_incref(block_cb); + err = mlxsw_sp_acl_block_bind(mlxsw_sp, acl_block, + mlxsw_sp_port, ingress); + if (err) + goto err_block_bind; + + if (ingress) + mlxsw_sp_port->ing_acl_block = acl_block; + else + mlxsw_sp_port->eg_acl_block = acl_block; + + return 0; + +err_block_bind: + if (!tcf_block_cb_decref(block_cb)) { + __tcf_block_cb_unregister(block_cb); +err_cb_register: + mlxsw_sp_acl_block_destroy(acl_block); + } + return err; } -static int mlxsw_sp_setup_tc_block_cb_eg(enum tc_setup_type type, - void *type_data, void *cb_priv) +static void +mlxsw_sp_setup_tc_block_flower_unbind(struct mlxsw_sp_port *mlxsw_sp_port, + struct tcf_block *block, bool ingress) { - return mlxsw_sp_setup_tc_block_cb(type, type_data, cb_priv, false); + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + struct mlxsw_sp_acl_block *acl_block; + struct tcf_block_cb *block_cb; + int err; + + block_cb = tcf_block_cb_lookup(block, mlxsw_sp_setup_tc_block_cb_flower, + mlxsw_sp); + if (!block_cb) + return; + + if (ingress) + mlxsw_sp_port->ing_acl_block = NULL; + else + mlxsw_sp_port->eg_acl_block = NULL; + + acl_block = tcf_block_cb_priv(block_cb); + err = mlxsw_sp_acl_block_unbind(mlxsw_sp, acl_block, + mlxsw_sp_port, ingress); + if (!err && !tcf_block_cb_decref(block_cb)) { + __tcf_block_cb_unregister(block_cb); + mlxsw_sp_acl_block_destroy(acl_block); + } } static int mlxsw_sp_setup_tc_block(struct mlxsw_sp_port *mlxsw_sp_port, struct tc_block_offload *f) { tc_setup_cb_t *cb; + bool ingress; + int err; - if (f->binder_type == TCF_BLOCK_BINDER_TYPE_CLSACT_INGRESS) - cb = mlxsw_sp_setup_tc_block_cb_ig; - else if (f->binder_type == TCF_BLOCK_BINDER_TYPE_CLSACT_EGRESS) - cb = mlxsw_sp_setup_tc_block_cb_eg; - else + if (f->binder_type == TCF_BLOCK_BINDER_TYPE_CLSACT_INGRESS) { + cb = mlxsw_sp_setup_tc_block_cb_matchall_ig; + ingress = true; + } else if (f->binder_type == TCF_BLOCK_BINDER_TYPE_CLSACT_EGRESS) { + cb = mlxsw_sp_setup_tc_block_cb_matchall_eg; + ingress = false; + } else { return -EOPNOTSUPP; + } switch (f->command) { case TC_BLOCK_BIND: - return tcf_block_cb_register(f->block, cb, mlxsw_sp_port, - mlxsw_sp_port); + err = tcf_block_cb_register(f->block, cb, mlxsw_sp_port, + mlxsw_sp_port); + if (err) + return err; + err = mlxsw_sp_setup_tc_block_flower_bind(mlxsw_sp_port, + f->block, ingress); + if (err) { + tcf_block_cb_unregister(f->block, cb, mlxsw_sp_port); + return err; + } + return 0; case TC_BLOCK_UNBIND: + mlxsw_sp_setup_tc_block_flower_unbind(mlxsw_sp_port, + f->block, ingress); tcf_block_cb_unregister(f->block, cb, mlxsw_sp_port); return 0; default: @@ -1833,11 +1949,69 @@ static int mlxsw_sp_setup_tc(struct net_device *dev, enum tc_setup_type type, return mlxsw_sp_setup_tc_block(mlxsw_sp_port, type_data); case TC_SETUP_QDISC_RED: return mlxsw_sp_setup_tc_red(mlxsw_sp_port, type_data); + case TC_SETUP_QDISC_PRIO: + return mlxsw_sp_setup_tc_prio(mlxsw_sp_port, type_data); default: return -EOPNOTSUPP; } } + +static int mlxsw_sp_feature_hw_tc(struct net_device *dev, bool enable) +{ + struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); + + if (!enable) { + if (mlxsw_sp_acl_block_rule_count(mlxsw_sp_port->ing_acl_block) || + mlxsw_sp_acl_block_rule_count(mlxsw_sp_port->eg_acl_block) || + !list_empty(&mlxsw_sp_port->mall_tc_list)) { + netdev_err(dev, "Active offloaded tc filters, can't turn hw_tc_offload off\n"); + return -EINVAL; + } + mlxsw_sp_acl_block_disable_inc(mlxsw_sp_port->ing_acl_block); + mlxsw_sp_acl_block_disable_inc(mlxsw_sp_port->eg_acl_block); + } else { + mlxsw_sp_acl_block_disable_dec(mlxsw_sp_port->ing_acl_block); + mlxsw_sp_acl_block_disable_dec(mlxsw_sp_port->eg_acl_block); + } + return 0; +} + +typedef int (*mlxsw_sp_feature_handler)(struct net_device *dev, bool enable); + +static int mlxsw_sp_handle_feature(struct net_device *dev, + netdev_features_t wanted_features, + netdev_features_t feature, + mlxsw_sp_feature_handler feature_handler) +{ + netdev_features_t changes = wanted_features ^ dev->features; + bool enable = !!(wanted_features & feature); + int err; + + if (!(changes & feature)) + return 0; + + err = feature_handler(dev, enable); + if (err) { + netdev_err(dev, "%s feature %pNF failed, err %d\n", + enable ? "Enable" : "Disable", &feature, err); + return err; + } + + if (enable) + dev->features |= feature; + else + dev->features &= ~feature; + + return 0; +} +static int mlxsw_sp_set_features(struct net_device *dev, + netdev_features_t features) +{ + return mlxsw_sp_handle_feature(dev, features, NETIF_F_HW_TC, + mlxsw_sp_feature_hw_tc); +} + static const struct net_device_ops mlxsw_sp_port_netdev_ops = { .ndo_open = mlxsw_sp_port_open, .ndo_stop = mlxsw_sp_port_stop, @@ -1852,6 +2026,7 @@ static const struct net_device_ops mlxsw_sp_port_netdev_ops = { .ndo_vlan_rx_add_vid = mlxsw_sp_port_add_vid, .ndo_vlan_rx_kill_vid = mlxsw_sp_port_kill_vid, .ndo_get_phys_port_name = mlxsw_sp_port_get_phys_port_name, + .ndo_set_features = mlxsw_sp_set_features, }; static void mlxsw_sp_port_get_drvinfo(struct net_device *dev, @@ -3039,6 +3214,13 @@ static int mlxsw_sp_port_create(struct mlxsw_sp *mlxsw_sp, u8 local_port, goto err_port_fids_init; } + err = mlxsw_sp_tc_qdisc_init(mlxsw_sp_port); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Port %d: Failed to initialize TC qdiscs\n", + mlxsw_sp_port->local_port); + goto err_port_qdiscs_init; + } + mlxsw_sp_port_vlan = mlxsw_sp_port_vlan_get(mlxsw_sp_port, 1); if (IS_ERR(mlxsw_sp_port_vlan)) { dev_err(mlxsw_sp->bus_info->dev, "Port %d: Failed to create VID 1\n", @@ -3067,6 +3249,8 @@ err_register_netdev: mlxsw_sp_port_switchdev_fini(mlxsw_sp_port); mlxsw_sp_port_vlan_put(mlxsw_sp_port_vlan); err_port_vlan_get: + mlxsw_sp_tc_qdisc_fini(mlxsw_sp_port); +err_port_qdiscs_init: mlxsw_sp_port_fids_fini(mlxsw_sp_port); err_port_fids_init: mlxsw_sp_port_dcb_fini(mlxsw_sp_port); @@ -3102,6 +3286,7 @@ static void mlxsw_sp_port_remove(struct mlxsw_sp *mlxsw_sp, u8 local_port) mlxsw_sp->ports[local_port] = NULL; mlxsw_sp_port_switchdev_fini(mlxsw_sp_port); mlxsw_sp_port_vlan_flush(mlxsw_sp_port); + mlxsw_sp_tc_qdisc_fini(mlxsw_sp_port); mlxsw_sp_port_fids_fini(mlxsw_sp_port); mlxsw_sp_port_dcb_fini(mlxsw_sp_port); mlxsw_sp_port_swid_set(mlxsw_sp_port, MLXSW_PORT_SWID_DISABLED_PORT); @@ -3933,6 +4118,253 @@ static const struct mlxsw_config_profile mlxsw_sp_config_profile = { .resource_query_enable = 1, }; +static bool +mlxsw_sp_resource_kvd_granularity_validate(struct netlink_ext_ack *extack, + u64 size) +{ + const struct mlxsw_config_profile *profile; + + profile = &mlxsw_sp_config_profile; + if (size % profile->kvd_hash_granularity) { + NL_SET_ERR_MSG_MOD(extack, "resource set with wrong granularity"); + return false; + } + return true; +} + +static int +mlxsw_sp_resource_kvd_size_validate(struct devlink *devlink, u64 size, + struct netlink_ext_ack *extack) +{ + NL_SET_ERR_MSG_MOD(extack, "kvd size cannot be changed"); + return -EINVAL; +} + +static int +mlxsw_sp_resource_kvd_linear_size_validate(struct devlink *devlink, u64 size, + struct netlink_ext_ack *extack) +{ + if (!mlxsw_sp_resource_kvd_granularity_validate(extack, size)) + return -EINVAL; + + return 0; +} + +static int +mlxsw_sp_resource_kvd_hash_single_size_validate(struct devlink *devlink, u64 size, + struct netlink_ext_ack *extack) +{ + struct mlxsw_core *mlxsw_core = devlink_priv(devlink); + + if (!mlxsw_sp_resource_kvd_granularity_validate(extack, size)) + return -EINVAL; + + if (size < MLXSW_CORE_RES_GET(mlxsw_core, KVD_SINGLE_MIN_SIZE)) { + NL_SET_ERR_MSG_MOD(extack, "hash single size is smaller than minimum"); + return -EINVAL; + } + return 0; +} + +static int +mlxsw_sp_resource_kvd_hash_double_size_validate(struct devlink *devlink, u64 size, + struct netlink_ext_ack *extack) +{ + struct mlxsw_core *mlxsw_core = devlink_priv(devlink); + + if (!mlxsw_sp_resource_kvd_granularity_validate(extack, size)) + return -EINVAL; + + if (size < MLXSW_CORE_RES_GET(mlxsw_core, KVD_DOUBLE_MIN_SIZE)) { + NL_SET_ERR_MSG_MOD(extack, "hash double size is smaller than minimum"); + return -EINVAL; + } + return 0; +} + +static u64 mlxsw_sp_resource_kvd_linear_occ_get(struct devlink *devlink) +{ + struct mlxsw_core *mlxsw_core = devlink_priv(devlink); + struct mlxsw_sp *mlxsw_sp = mlxsw_core_driver_priv(mlxsw_core); + + return mlxsw_sp_kvdl_occ_get(mlxsw_sp); +} + +static struct devlink_resource_ops mlxsw_sp_resource_kvd_ops = { + .size_validate = mlxsw_sp_resource_kvd_size_validate, +}; + +static struct devlink_resource_ops mlxsw_sp_resource_kvd_linear_ops = { + .size_validate = mlxsw_sp_resource_kvd_linear_size_validate, + .occ_get = mlxsw_sp_resource_kvd_linear_occ_get, +}; + +static struct devlink_resource_ops mlxsw_sp_resource_kvd_hash_single_ops = { + .size_validate = mlxsw_sp_resource_kvd_hash_single_size_validate, +}; + +static struct devlink_resource_ops mlxsw_sp_resource_kvd_hash_double_ops = { + .size_validate = mlxsw_sp_resource_kvd_hash_double_size_validate, +}; + +static struct devlink_resource_size_params mlxsw_sp_kvd_size_params; +static struct devlink_resource_size_params mlxsw_sp_linear_size_params; +static struct devlink_resource_size_params mlxsw_sp_hash_single_size_params; +static struct devlink_resource_size_params mlxsw_sp_hash_double_size_params; + +static void +mlxsw_sp_resource_size_params_prepare(struct mlxsw_core *mlxsw_core) +{ + u32 single_size_min = MLXSW_CORE_RES_GET(mlxsw_core, + KVD_SINGLE_MIN_SIZE); + u32 double_size_min = MLXSW_CORE_RES_GET(mlxsw_core, + KVD_DOUBLE_MIN_SIZE); + u32 kvd_size = MLXSW_CORE_RES_GET(mlxsw_core, KVD_SIZE); + u32 linear_size_min = 0; + + /* KVD top resource */ + mlxsw_sp_kvd_size_params.size_min = kvd_size; + mlxsw_sp_kvd_size_params.size_max = kvd_size; + mlxsw_sp_kvd_size_params.size_granularity = MLXSW_SP_KVD_GRANULARITY; + mlxsw_sp_kvd_size_params.unit = DEVLINK_RESOURCE_UNIT_ENTRY; + + /* Linear part init */ + mlxsw_sp_linear_size_params.size_min = linear_size_min; + mlxsw_sp_linear_size_params.size_max = kvd_size - single_size_min - + double_size_min; + mlxsw_sp_linear_size_params.size_granularity = MLXSW_SP_KVD_GRANULARITY; + mlxsw_sp_linear_size_params.unit = DEVLINK_RESOURCE_UNIT_ENTRY; + + /* Hash double part init */ + mlxsw_sp_hash_double_size_params.size_min = double_size_min; + mlxsw_sp_hash_double_size_params.size_max = kvd_size - single_size_min - + linear_size_min; + mlxsw_sp_hash_double_size_params.size_granularity = MLXSW_SP_KVD_GRANULARITY; + mlxsw_sp_hash_double_size_params.unit = DEVLINK_RESOURCE_UNIT_ENTRY; + + /* Hash single part init */ + mlxsw_sp_hash_single_size_params.size_min = single_size_min; + mlxsw_sp_hash_single_size_params.size_max = kvd_size - double_size_min - + linear_size_min; + mlxsw_sp_hash_single_size_params.size_granularity = MLXSW_SP_KVD_GRANULARITY; + mlxsw_sp_hash_single_size_params.unit = DEVLINK_RESOURCE_UNIT_ENTRY; +} + +static int mlxsw_sp_resources_register(struct mlxsw_core *mlxsw_core) +{ + struct devlink *devlink = priv_to_devlink(mlxsw_core); + u32 kvd_size, single_size, double_size, linear_size; + const struct mlxsw_config_profile *profile; + int err; + + profile = &mlxsw_sp_config_profile; + if (!MLXSW_CORE_RES_VALID(mlxsw_core, KVD_SIZE)) + return -EIO; + + mlxsw_sp_resource_size_params_prepare(mlxsw_core); + kvd_size = MLXSW_CORE_RES_GET(mlxsw_core, KVD_SIZE); + err = devlink_resource_register(devlink, MLXSW_SP_RESOURCE_NAME_KVD, + true, kvd_size, + MLXSW_SP_RESOURCE_KVD, + DEVLINK_RESOURCE_ID_PARENT_TOP, + &mlxsw_sp_kvd_size_params, + &mlxsw_sp_resource_kvd_ops); + if (err) + return err; + + linear_size = profile->kvd_linear_size; + err = devlink_resource_register(devlink, MLXSW_SP_RESOURCE_NAME_KVD_LINEAR, + false, linear_size, + MLXSW_SP_RESOURCE_KVD_LINEAR, + MLXSW_SP_RESOURCE_KVD, + &mlxsw_sp_linear_size_params, + &mlxsw_sp_resource_kvd_linear_ops); + if (err) + return err; + + double_size = kvd_size - linear_size; + double_size *= profile->kvd_hash_double_parts; + double_size /= profile->kvd_hash_double_parts + + profile->kvd_hash_single_parts; + double_size = rounddown(double_size, profile->kvd_hash_granularity); + err = devlink_resource_register(devlink, MLXSW_SP_RESOURCE_NAME_KVD_HASH_DOUBLE, + false, double_size, + MLXSW_SP_RESOURCE_KVD_HASH_DOUBLE, + MLXSW_SP_RESOURCE_KVD, + &mlxsw_sp_hash_double_size_params, + &mlxsw_sp_resource_kvd_hash_double_ops); + if (err) + return err; + + single_size = kvd_size - double_size - linear_size; + err = devlink_resource_register(devlink, MLXSW_SP_RESOURCE_NAME_KVD_HASH_SINGLE, + false, single_size, + MLXSW_SP_RESOURCE_KVD_HASH_SINGLE, + MLXSW_SP_RESOURCE_KVD, + &mlxsw_sp_hash_single_size_params, + &mlxsw_sp_resource_kvd_hash_single_ops); + if (err) + return err; + + return 0; +} + +static int mlxsw_sp_kvd_sizes_get(struct mlxsw_core *mlxsw_core, + const struct mlxsw_config_profile *profile, + u64 *p_single_size, u64 *p_double_size, + u64 *p_linear_size) +{ + struct devlink *devlink = priv_to_devlink(mlxsw_core); + u32 double_size; + int err; + + if (!MLXSW_CORE_RES_VALID(mlxsw_core, KVD_SINGLE_MIN_SIZE) || + !MLXSW_CORE_RES_VALID(mlxsw_core, KVD_DOUBLE_MIN_SIZE) || + !profile->used_kvd_split_data) + return -EIO; + + /* The hash part is what left of the kvd without the + * linear part. It is split to the single size and + * double size by the parts ratio from the profile. + * Both sizes must be a multiplications of the + * granularity from the profile. In case the user + * provided the sizes they are obtained via devlink. + */ + err = devlink_resource_size_get(devlink, + MLXSW_SP_RESOURCE_KVD_LINEAR, + p_linear_size); + if (err) + *p_linear_size = profile->kvd_linear_size; + + err = devlink_resource_size_get(devlink, + MLXSW_SP_RESOURCE_KVD_HASH_DOUBLE, + p_double_size); + if (err) { + double_size = MLXSW_CORE_RES_GET(mlxsw_core, KVD_SIZE) - + *p_linear_size; + double_size *= profile->kvd_hash_double_parts; + double_size /= profile->kvd_hash_double_parts + + profile->kvd_hash_single_parts; + *p_double_size = rounddown(double_size, + profile->kvd_hash_granularity); + } + + err = devlink_resource_size_get(devlink, + MLXSW_SP_RESOURCE_KVD_HASH_SINGLE, + p_single_size); + if (err) + *p_single_size = MLXSW_CORE_RES_GET(mlxsw_core, KVD_SIZE) - + *p_double_size - *p_linear_size; + + /* Check results are legal. */ + if (*p_single_size < MLXSW_CORE_RES_GET(mlxsw_core, KVD_SINGLE_MIN_SIZE) || + *p_double_size < MLXSW_CORE_RES_GET(mlxsw_core, KVD_DOUBLE_MIN_SIZE) || + MLXSW_CORE_RES_GET(mlxsw_core, KVD_SIZE) < *p_linear_size) + return -EIO; + + return 0; +} + static struct mlxsw_driver mlxsw_sp_driver = { .kind = mlxsw_sp_driver_name, .priv_size = sizeof(struct mlxsw_sp), @@ -3952,6 +4384,8 @@ static struct mlxsw_driver mlxsw_sp_driver = { .sb_occ_port_pool_get = mlxsw_sp_sb_occ_port_pool_get, .sb_occ_tc_port_bind_get = mlxsw_sp_sb_occ_tc_port_bind_get, .txhdr_construct = mlxsw_sp_txhdr_construct, + .resources_register = mlxsw_sp_resources_register, + .kvd_sizes_get = mlxsw_sp_kvd_sizes_get, .txhdr_len = MLXSW_TXHDR_LEN, .profile = &mlxsw_sp_config_profile, }; diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum.h b/drivers/net/ethernet/mellanox/mlxsw/spectrum.h index 05ce1be..bdd8f94a 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum.h +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum.h @@ -66,6 +66,18 @@ #define MLXSW_SP_KVD_LINEAR_SIZE 98304 /* entries */ #define MLXSW_SP_KVD_GRANULARITY 128 +#define MLXSW_SP_RESOURCE_NAME_KVD "kvd" +#define MLXSW_SP_RESOURCE_NAME_KVD_LINEAR "linear" +#define MLXSW_SP_RESOURCE_NAME_KVD_HASH_SINGLE "hash_single" +#define MLXSW_SP_RESOURCE_NAME_KVD_HASH_DOUBLE "hash_double" + +enum mlxsw_sp_resource_id { + MLXSW_SP_RESOURCE_KVD, + MLXSW_SP_RESOURCE_KVD_LINEAR, + MLXSW_SP_RESOURCE_KVD_HASH_SINGLE, + MLXSW_SP_RESOURCE_KVD_HASH_DOUBLE, +}; + struct mlxsw_sp_port; struct mlxsw_sp_rif; @@ -204,29 +216,6 @@ struct mlxsw_sp_port_vlan { struct list_head bridge_vlan_node; }; -enum mlxsw_sp_qdisc_type { - MLXSW_SP_QDISC_NO_QDISC, - MLXSW_SP_QDISC_RED, -}; - -struct mlxsw_sp_qdisc { - u32 handle; - enum mlxsw_sp_qdisc_type type; - struct red_stats xstats_base; - union { - struct { - u64 tail_drop_base; - u64 ecn_base; - u64 wred_drop_base; - } red; - } xstats; - - u64 tx_bytes; - u64 tx_packets; - u64 drops; - u64 overlimits; -}; - /* No need an internal lock; At worse - miss a single periodic iteration */ struct mlxsw_sp_port_xstats { u64 ecn; @@ -269,7 +258,10 @@ struct mlxsw_sp_port { } periodic_hw_stats; struct mlxsw_sp_port_sample *sample; struct list_head vlans_list; - struct mlxsw_sp_qdisc root_qdisc; + struct mlxsw_sp_qdisc *root_qdisc; + unsigned acl_rule_count; + struct mlxsw_sp_acl_block *ing_acl_block; + struct mlxsw_sp_acl_block *eg_acl_block; }; static inline bool @@ -404,6 +396,16 @@ struct mlxsw_sp_port *mlxsw_sp_port_dev_lower_find(struct net_device *dev); struct mlxsw_sp_port *mlxsw_sp_port_lower_dev_hold(struct net_device *dev); void mlxsw_sp_port_dev_put(struct mlxsw_sp_port *mlxsw_sp_port); struct mlxsw_sp_port *mlxsw_sp_port_dev_lower_find_rcu(struct net_device *dev); +int mlxsw_sp_span_mirror_add(struct mlxsw_sp_port *from, + struct mlxsw_sp_port *to, + enum mlxsw_sp_span_type type, + bool bind); +void mlxsw_sp_span_mirror_del(struct mlxsw_sp_port *from, + u8 destination_port, + enum mlxsw_sp_span_type type, + bool bind); +struct mlxsw_sp_span_entry * +mlxsw_sp_span_entry_find(struct mlxsw_sp *mlxsw_sp, u8 local_port); /* spectrum_dcb.c */ #ifdef CONFIG_MLXSW_SPECTRUM_DCB @@ -458,13 +460,13 @@ void mlxsw_sp_kvdl_free(struct mlxsw_sp *mlxsw_sp, int entry_index); int mlxsw_sp_kvdl_alloc_size_query(struct mlxsw_sp *mlxsw_sp, unsigned int entry_count, unsigned int *p_alloc_size); +u64 mlxsw_sp_kvdl_occ_get(const struct mlxsw_sp *mlxsw_sp); struct mlxsw_sp_acl_rule_info { unsigned int priority; struct mlxsw_afk_element_values values; struct mlxsw_afa_block *act_block; unsigned int counter_index; - bool counter_valid; }; enum mlxsw_sp_acl_profile { @@ -477,8 +479,11 @@ struct mlxsw_sp_acl_profile_ops { void *priv, void *ruleset_priv); void (*ruleset_del)(struct mlxsw_sp *mlxsw_sp, void *ruleset_priv); int (*ruleset_bind)(struct mlxsw_sp *mlxsw_sp, void *ruleset_priv, - struct net_device *dev, bool ingress); - void (*ruleset_unbind)(struct mlxsw_sp *mlxsw_sp, void *ruleset_priv); + struct mlxsw_sp_port *mlxsw_sp_port, + bool ingress); + void (*ruleset_unbind)(struct mlxsw_sp *mlxsw_sp, void *ruleset_priv, + struct mlxsw_sp_port *mlxsw_sp_port, + bool ingress); u16 (*ruleset_group_id)(void *ruleset_priv); size_t rule_priv_size; int (*rule_add)(struct mlxsw_sp *mlxsw_sp, @@ -498,17 +503,34 @@ struct mlxsw_sp_acl_ops { enum mlxsw_sp_acl_profile profile); }; +struct mlxsw_sp_acl_block; struct mlxsw_sp_acl_ruleset; /* spectrum_acl.c */ struct mlxsw_afk *mlxsw_sp_acl_afk(struct mlxsw_sp_acl *acl); +struct mlxsw_sp *mlxsw_sp_acl_block_mlxsw_sp(struct mlxsw_sp_acl_block *block); +unsigned int mlxsw_sp_acl_block_rule_count(struct mlxsw_sp_acl_block *block); +void mlxsw_sp_acl_block_disable_inc(struct mlxsw_sp_acl_block *block); +void mlxsw_sp_acl_block_disable_dec(struct mlxsw_sp_acl_block *block); +bool mlxsw_sp_acl_block_disabled(struct mlxsw_sp_acl_block *block); +struct mlxsw_sp_acl_block *mlxsw_sp_acl_block_create(struct mlxsw_sp *mlxsw_sp, + struct net *net); +void mlxsw_sp_acl_block_destroy(struct mlxsw_sp_acl_block *block); +int mlxsw_sp_acl_block_bind(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_block *block, + struct mlxsw_sp_port *mlxsw_sp_port, + bool ingress); +int mlxsw_sp_acl_block_unbind(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_block *block, + struct mlxsw_sp_port *mlxsw_sp_port, + bool ingress); struct mlxsw_sp_acl_ruleset * -mlxsw_sp_acl_ruleset_lookup(struct mlxsw_sp *mlxsw_sp, struct net_device *dev, - bool ingress, u32 chain_index, +mlxsw_sp_acl_ruleset_lookup(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_block *block, u32 chain_index, enum mlxsw_sp_acl_profile profile); struct mlxsw_sp_acl_ruleset * -mlxsw_sp_acl_ruleset_get(struct mlxsw_sp *mlxsw_sp, struct net_device *dev, - bool ingress, u32 chain_index, +mlxsw_sp_acl_ruleset_get(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_block *block, u32 chain_index, enum mlxsw_sp_acl_profile profile); void mlxsw_sp_acl_ruleset_put(struct mlxsw_sp *mlxsw_sp, struct mlxsw_sp_acl_ruleset *ruleset); @@ -532,6 +554,10 @@ int mlxsw_sp_acl_rulei_act_jump(struct mlxsw_sp_acl_rule_info *rulei, u16 group_id); int mlxsw_sp_acl_rulei_act_drop(struct mlxsw_sp_acl_rule_info *rulei); int mlxsw_sp_acl_rulei_act_trap(struct mlxsw_sp_acl_rule_info *rulei); +int mlxsw_sp_acl_rulei_act_mirror(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_rule_info *rulei, + struct mlxsw_sp_acl_block *block, + struct net_device *out_dev); int mlxsw_sp_acl_rulei_act_fwd(struct mlxsw_sp *mlxsw_sp, struct mlxsw_sp_acl_rule_info *rulei, struct net_device *out_dev); @@ -575,16 +601,23 @@ void mlxsw_sp_acl_fini(struct mlxsw_sp *mlxsw_sp); extern const struct mlxsw_sp_acl_ops mlxsw_sp_acl_tcam_ops; /* spectrum_flower.c */ -int mlxsw_sp_flower_replace(struct mlxsw_sp_port *mlxsw_sp_port, bool ingress, +int mlxsw_sp_flower_replace(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_block *block, struct tc_cls_flower_offload *f); -void mlxsw_sp_flower_destroy(struct mlxsw_sp_port *mlxsw_sp_port, bool ingress, +void mlxsw_sp_flower_destroy(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_block *block, struct tc_cls_flower_offload *f); -int mlxsw_sp_flower_stats(struct mlxsw_sp_port *mlxsw_sp_port, bool ingress, +int mlxsw_sp_flower_stats(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_block *block, struct tc_cls_flower_offload *f); /* spectrum_qdisc.c */ +int mlxsw_sp_tc_qdisc_init(struct mlxsw_sp_port *mlxsw_sp_port); +void mlxsw_sp_tc_qdisc_fini(struct mlxsw_sp_port *mlxsw_sp_port); int mlxsw_sp_setup_tc_red(struct mlxsw_sp_port *mlxsw_sp_port, struct tc_red_qopt_offload *p); +int mlxsw_sp_setup_tc_prio(struct mlxsw_sp_port *mlxsw_sp_port, + struct tc_prio_qopt_offload *p); /* spectrum_fid.c */ int mlxsw_sp_fid_flood_set(struct mlxsw_sp_fid *fid, diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl.c index 93dcd31..0897a54 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl.c @@ -39,6 +39,7 @@ #include <linux/string.h> #include <linux/rhashtable.h> #include <linux/netdevice.h> +#include <net/net_namespace.h> #include <net/tc_act/tc_vlan.h> #include "reg.h" @@ -70,9 +71,23 @@ struct mlxsw_afk *mlxsw_sp_acl_afk(struct mlxsw_sp_acl *acl) return acl->afk; } -struct mlxsw_sp_acl_ruleset_ht_key { - struct net_device *dev; /* dev this ruleset is bound to */ +struct mlxsw_sp_acl_block_binding { + struct list_head list; + struct net_device *dev; + struct mlxsw_sp_port *mlxsw_sp_port; bool ingress; +}; + +struct mlxsw_sp_acl_block { + struct list_head binding_list; + struct mlxsw_sp_acl_ruleset *ruleset_zero; + struct mlxsw_sp *mlxsw_sp; + unsigned int rule_count; + unsigned int disable_count; +}; + +struct mlxsw_sp_acl_ruleset_ht_key { + struct mlxsw_sp_acl_block *block; u32 chain_index; const struct mlxsw_sp_acl_profile_ops *ops; }; @@ -118,8 +133,185 @@ struct mlxsw_sp_fid *mlxsw_sp_acl_dummy_fid(struct mlxsw_sp *mlxsw_sp) return mlxsw_sp->acl->dummy_fid; } +struct mlxsw_sp *mlxsw_sp_acl_block_mlxsw_sp(struct mlxsw_sp_acl_block *block) +{ + return block->mlxsw_sp; +} + +unsigned int mlxsw_sp_acl_block_rule_count(struct mlxsw_sp_acl_block *block) +{ + return block ? block->rule_count : 0; +} + +void mlxsw_sp_acl_block_disable_inc(struct mlxsw_sp_acl_block *block) +{ + if (block) + block->disable_count++; +} + +void mlxsw_sp_acl_block_disable_dec(struct mlxsw_sp_acl_block *block) +{ + if (block) + block->disable_count--; +} + +bool mlxsw_sp_acl_block_disabled(struct mlxsw_sp_acl_block *block) +{ + return block->disable_count; +} + +static int +mlxsw_sp_acl_ruleset_bind(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_block *block, + struct mlxsw_sp_acl_block_binding *binding) +{ + struct mlxsw_sp_acl_ruleset *ruleset = block->ruleset_zero; + const struct mlxsw_sp_acl_profile_ops *ops = ruleset->ht_key.ops; + + return ops->ruleset_bind(mlxsw_sp, ruleset->priv, + binding->mlxsw_sp_port, binding->ingress); +} + +static void +mlxsw_sp_acl_ruleset_unbind(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_block *block, + struct mlxsw_sp_acl_block_binding *binding) +{ + struct mlxsw_sp_acl_ruleset *ruleset = block->ruleset_zero; + const struct mlxsw_sp_acl_profile_ops *ops = ruleset->ht_key.ops; + + ops->ruleset_unbind(mlxsw_sp, ruleset->priv, + binding->mlxsw_sp_port, binding->ingress); +} + +static bool mlxsw_sp_acl_ruleset_block_bound(struct mlxsw_sp_acl_block *block) +{ + return block->ruleset_zero; +} + +static int +mlxsw_sp_acl_ruleset_block_bind(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_ruleset *ruleset, + struct mlxsw_sp_acl_block *block) +{ + struct mlxsw_sp_acl_block_binding *binding; + int err; + + block->ruleset_zero = ruleset; + list_for_each_entry(binding, &block->binding_list, list) { + err = mlxsw_sp_acl_ruleset_bind(mlxsw_sp, block, binding); + if (err) + goto rollback; + } + return 0; + +rollback: + list_for_each_entry_continue_reverse(binding, &block->binding_list, + list) + mlxsw_sp_acl_ruleset_unbind(mlxsw_sp, block, binding); + block->ruleset_zero = NULL; + + return err; +} + +static void +mlxsw_sp_acl_ruleset_block_unbind(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_ruleset *ruleset, + struct mlxsw_sp_acl_block *block) +{ + struct mlxsw_sp_acl_block_binding *binding; + + list_for_each_entry(binding, &block->binding_list, list) + mlxsw_sp_acl_ruleset_unbind(mlxsw_sp, block, binding); + block->ruleset_zero = NULL; +} + +struct mlxsw_sp_acl_block *mlxsw_sp_acl_block_create(struct mlxsw_sp *mlxsw_sp, + struct net *net) +{ + struct mlxsw_sp_acl_block *block; + + block = kzalloc(sizeof(*block), GFP_KERNEL); + if (!block) + return NULL; + INIT_LIST_HEAD(&block->binding_list); + block->mlxsw_sp = mlxsw_sp; + return block; +} + +void mlxsw_sp_acl_block_destroy(struct mlxsw_sp_acl_block *block) +{ + WARN_ON(!list_empty(&block->binding_list)); + kfree(block); +} + +static struct mlxsw_sp_acl_block_binding * +mlxsw_sp_acl_block_lookup(struct mlxsw_sp_acl_block *block, + struct mlxsw_sp_port *mlxsw_sp_port, bool ingress) +{ + struct mlxsw_sp_acl_block_binding *binding; + + list_for_each_entry(binding, &block->binding_list, list) + if (binding->mlxsw_sp_port == mlxsw_sp_port && + binding->ingress == ingress) + return binding; + return NULL; +} + +int mlxsw_sp_acl_block_bind(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_block *block, + struct mlxsw_sp_port *mlxsw_sp_port, + bool ingress) +{ + struct mlxsw_sp_acl_block_binding *binding; + int err; + + if (WARN_ON(mlxsw_sp_acl_block_lookup(block, mlxsw_sp_port, ingress))) + return -EEXIST; + + binding = kzalloc(sizeof(*binding), GFP_KERNEL); + if (!binding) + return -ENOMEM; + binding->mlxsw_sp_port = mlxsw_sp_port; + binding->ingress = ingress; + + if (mlxsw_sp_acl_ruleset_block_bound(block)) { + err = mlxsw_sp_acl_ruleset_bind(mlxsw_sp, block, binding); + if (err) + goto err_ruleset_bind; + } + + list_add(&binding->list, &block->binding_list); + return 0; + +err_ruleset_bind: + kfree(binding); + return err; +} + +int mlxsw_sp_acl_block_unbind(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_block *block, + struct mlxsw_sp_port *mlxsw_sp_port, + bool ingress) +{ + struct mlxsw_sp_acl_block_binding *binding; + + binding = mlxsw_sp_acl_block_lookup(block, mlxsw_sp_port, ingress); + if (!binding) + return -ENOENT; + + list_del(&binding->list); + + if (mlxsw_sp_acl_ruleset_block_bound(block)) + mlxsw_sp_acl_ruleset_unbind(mlxsw_sp, block, binding); + + kfree(binding); + return 0; +} + static struct mlxsw_sp_acl_ruleset * mlxsw_sp_acl_ruleset_create(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_block *block, u32 chain_index, const struct mlxsw_sp_acl_profile_ops *ops) { struct mlxsw_sp_acl *acl = mlxsw_sp->acl; @@ -132,6 +324,8 @@ mlxsw_sp_acl_ruleset_create(struct mlxsw_sp *mlxsw_sp, if (!ruleset) return ERR_PTR(-ENOMEM); ruleset->ref_count = 1; + ruleset->ht_key.block = block; + ruleset->ht_key.chain_index = chain_index; ruleset->ht_key.ops = ops; err = rhashtable_init(&ruleset->rule_ht, &mlxsw_sp_acl_rule_ht_params); @@ -142,68 +336,50 @@ mlxsw_sp_acl_ruleset_create(struct mlxsw_sp *mlxsw_sp, if (err) goto err_ops_ruleset_add; - return ruleset; - -err_ops_ruleset_add: - rhashtable_destroy(&ruleset->rule_ht); -err_rhashtable_init: - kfree(ruleset); - return ERR_PTR(err); -} - -static void mlxsw_sp_acl_ruleset_destroy(struct mlxsw_sp *mlxsw_sp, - struct mlxsw_sp_acl_ruleset *ruleset) -{ - const struct mlxsw_sp_acl_profile_ops *ops = ruleset->ht_key.ops; - - ops->ruleset_del(mlxsw_sp, ruleset->priv); - rhashtable_destroy(&ruleset->rule_ht); - kfree(ruleset); -} - -static int mlxsw_sp_acl_ruleset_bind(struct mlxsw_sp *mlxsw_sp, - struct mlxsw_sp_acl_ruleset *ruleset, - struct net_device *dev, bool ingress, - u32 chain_index) -{ - const struct mlxsw_sp_acl_profile_ops *ops = ruleset->ht_key.ops; - struct mlxsw_sp_acl *acl = mlxsw_sp->acl; - int err; - - ruleset->ht_key.dev = dev; - ruleset->ht_key.ingress = ingress; - ruleset->ht_key.chain_index = chain_index; err = rhashtable_insert_fast(&acl->ruleset_ht, &ruleset->ht_node, mlxsw_sp_acl_ruleset_ht_params); if (err) - return err; - if (!ruleset->ht_key.chain_index) { + goto err_ht_insert; + + if (!chain_index) { /* We only need ruleset with chain index 0, the implicit one, * to be directly bound to device. The rest of the rulesets * are bound by "Goto action set". */ - err = ops->ruleset_bind(mlxsw_sp, ruleset->priv, dev, ingress); + err = mlxsw_sp_acl_ruleset_block_bind(mlxsw_sp, ruleset, block); if (err) - goto err_ops_ruleset_bind; + goto err_ruleset_bind; } - return 0; -err_ops_ruleset_bind: + return ruleset; + +err_ruleset_bind: rhashtable_remove_fast(&acl->ruleset_ht, &ruleset->ht_node, mlxsw_sp_acl_ruleset_ht_params); - return err; +err_ht_insert: + ops->ruleset_del(mlxsw_sp, ruleset->priv); +err_ops_ruleset_add: + rhashtable_destroy(&ruleset->rule_ht); +err_rhashtable_init: + kfree(ruleset); + return ERR_PTR(err); } -static void mlxsw_sp_acl_ruleset_unbind(struct mlxsw_sp *mlxsw_sp, - struct mlxsw_sp_acl_ruleset *ruleset) +static void mlxsw_sp_acl_ruleset_destroy(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_ruleset *ruleset) { const struct mlxsw_sp_acl_profile_ops *ops = ruleset->ht_key.ops; + struct mlxsw_sp_acl_block *block = ruleset->ht_key.block; + u32 chain_index = ruleset->ht_key.chain_index; struct mlxsw_sp_acl *acl = mlxsw_sp->acl; - if (!ruleset->ht_key.chain_index) - ops->ruleset_unbind(mlxsw_sp, ruleset->priv); + if (!chain_index) + mlxsw_sp_acl_ruleset_block_unbind(mlxsw_sp, ruleset, block); rhashtable_remove_fast(&acl->ruleset_ht, &ruleset->ht_node, mlxsw_sp_acl_ruleset_ht_params); + ops->ruleset_del(mlxsw_sp, ruleset->priv); + rhashtable_destroy(&ruleset->rule_ht); + kfree(ruleset); } static void mlxsw_sp_acl_ruleset_ref_inc(struct mlxsw_sp_acl_ruleset *ruleset) @@ -216,20 +392,18 @@ static void mlxsw_sp_acl_ruleset_ref_dec(struct mlxsw_sp *mlxsw_sp, { if (--ruleset->ref_count) return; - mlxsw_sp_acl_ruleset_unbind(mlxsw_sp, ruleset); mlxsw_sp_acl_ruleset_destroy(mlxsw_sp, ruleset); } static struct mlxsw_sp_acl_ruleset * -__mlxsw_sp_acl_ruleset_lookup(struct mlxsw_sp_acl *acl, struct net_device *dev, - bool ingress, u32 chain_index, +__mlxsw_sp_acl_ruleset_lookup(struct mlxsw_sp_acl *acl, + struct mlxsw_sp_acl_block *block, u32 chain_index, const struct mlxsw_sp_acl_profile_ops *ops) { struct mlxsw_sp_acl_ruleset_ht_key ht_key; memset(&ht_key, 0, sizeof(ht_key)); - ht_key.dev = dev; - ht_key.ingress = ingress; + ht_key.block = block; ht_key.chain_index = chain_index; ht_key.ops = ops; return rhashtable_lookup_fast(&acl->ruleset_ht, &ht_key, @@ -237,8 +411,8 @@ __mlxsw_sp_acl_ruleset_lookup(struct mlxsw_sp_acl *acl, struct net_device *dev, } struct mlxsw_sp_acl_ruleset * -mlxsw_sp_acl_ruleset_lookup(struct mlxsw_sp *mlxsw_sp, struct net_device *dev, - bool ingress, u32 chain_index, +mlxsw_sp_acl_ruleset_lookup(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_block *block, u32 chain_index, enum mlxsw_sp_acl_profile profile) { const struct mlxsw_sp_acl_profile_ops *ops; @@ -248,45 +422,31 @@ mlxsw_sp_acl_ruleset_lookup(struct mlxsw_sp *mlxsw_sp, struct net_device *dev, ops = acl->ops->profile_ops(mlxsw_sp, profile); if (!ops) return ERR_PTR(-EINVAL); - ruleset = __mlxsw_sp_acl_ruleset_lookup(acl, dev, ingress, - chain_index, ops); + ruleset = __mlxsw_sp_acl_ruleset_lookup(acl, block, chain_index, ops); if (!ruleset) return ERR_PTR(-ENOENT); return ruleset; } struct mlxsw_sp_acl_ruleset * -mlxsw_sp_acl_ruleset_get(struct mlxsw_sp *mlxsw_sp, struct net_device *dev, - bool ingress, u32 chain_index, +mlxsw_sp_acl_ruleset_get(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_block *block, u32 chain_index, enum mlxsw_sp_acl_profile profile) { const struct mlxsw_sp_acl_profile_ops *ops; struct mlxsw_sp_acl *acl = mlxsw_sp->acl; struct mlxsw_sp_acl_ruleset *ruleset; - int err; ops = acl->ops->profile_ops(mlxsw_sp, profile); if (!ops) return ERR_PTR(-EINVAL); - ruleset = __mlxsw_sp_acl_ruleset_lookup(acl, dev, ingress, - chain_index, ops); + ruleset = __mlxsw_sp_acl_ruleset_lookup(acl, block, chain_index, ops); if (ruleset) { mlxsw_sp_acl_ruleset_ref_inc(ruleset); return ruleset; } - ruleset = mlxsw_sp_acl_ruleset_create(mlxsw_sp, ops); - if (IS_ERR(ruleset)) - return ruleset; - err = mlxsw_sp_acl_ruleset_bind(mlxsw_sp, ruleset, dev, - ingress, chain_index); - if (err) - goto err_ruleset_bind; - return ruleset; - -err_ruleset_bind: - mlxsw_sp_acl_ruleset_destroy(mlxsw_sp, ruleset); - return ERR_PTR(err); + return mlxsw_sp_acl_ruleset_create(mlxsw_sp, block, chain_index, ops); } void mlxsw_sp_acl_ruleset_put(struct mlxsw_sp *mlxsw_sp, @@ -302,27 +462,6 @@ u16 mlxsw_sp_acl_ruleset_group_id(struct mlxsw_sp_acl_ruleset *ruleset) return ops->ruleset_group_id(ruleset->priv); } -static int -mlxsw_sp_acl_rulei_counter_alloc(struct mlxsw_sp *mlxsw_sp, - struct mlxsw_sp_acl_rule_info *rulei) -{ - int err; - - err = mlxsw_sp_flow_counter_alloc(mlxsw_sp, &rulei->counter_index); - if (err) - return err; - rulei->counter_valid = true; - return 0; -} - -static void -mlxsw_sp_acl_rulei_counter_free(struct mlxsw_sp *mlxsw_sp, - struct mlxsw_sp_acl_rule_info *rulei) -{ - rulei->counter_valid = false; - mlxsw_sp_flow_counter_free(mlxsw_sp, rulei->counter_index); -} - struct mlxsw_sp_acl_rule_info * mlxsw_sp_acl_rulei_create(struct mlxsw_sp_acl *acl) { @@ -427,6 +566,34 @@ int mlxsw_sp_acl_rulei_act_fwd(struct mlxsw_sp *mlxsw_sp, local_port, in_port); } +int mlxsw_sp_acl_rulei_act_mirror(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_rule_info *rulei, + struct mlxsw_sp_acl_block *block, + struct net_device *out_dev) +{ + struct mlxsw_sp_acl_block_binding *binding; + struct mlxsw_sp_port *out_port; + struct mlxsw_sp_port *in_port; + + if (!list_is_singular(&block->binding_list)) + return -EOPNOTSUPP; + + binding = list_first_entry(&block->binding_list, + struct mlxsw_sp_acl_block_binding, list); + in_port = binding->mlxsw_sp_port; + if (!mlxsw_sp_port_dev_check(out_dev)) + return -EINVAL; + + out_port = netdev_priv(out_dev); + if (out_port->mlxsw_sp != mlxsw_sp) + return -EINVAL; + + return mlxsw_afa_block_append_mirror(rulei->act_block, + in_port->local_port, + out_port->local_port, + binding->ingress); +} + int mlxsw_sp_acl_rulei_act_vlan(struct mlxsw_sp *mlxsw_sp, struct mlxsw_sp_acl_rule_info *rulei, u32 action, u16 vid, u16 proto, u8 prio) @@ -459,7 +626,7 @@ int mlxsw_sp_acl_rulei_act_count(struct mlxsw_sp *mlxsw_sp, struct mlxsw_sp_acl_rule_info *rulei) { return mlxsw_afa_block_append_counter(rulei->act_block, - rulei->counter_index); + &rulei->counter_index); } int mlxsw_sp_acl_rulei_act_fid_set(struct mlxsw_sp *mlxsw_sp, @@ -493,13 +660,8 @@ mlxsw_sp_acl_rule_create(struct mlxsw_sp *mlxsw_sp, goto err_rulei_create; } - err = mlxsw_sp_acl_rulei_counter_alloc(mlxsw_sp, rule->rulei); - if (err) - goto err_counter_alloc; return rule; -err_counter_alloc: - mlxsw_sp_acl_rulei_destroy(rule->rulei); err_rulei_create: kfree(rule); err_alloc: @@ -512,7 +674,6 @@ void mlxsw_sp_acl_rule_destroy(struct mlxsw_sp *mlxsw_sp, { struct mlxsw_sp_acl_ruleset *ruleset = rule->ruleset; - mlxsw_sp_acl_rulei_counter_free(mlxsw_sp, rule->rulei); mlxsw_sp_acl_rulei_destroy(rule->rulei); kfree(rule); mlxsw_sp_acl_ruleset_ref_dec(mlxsw_sp, ruleset); @@ -535,6 +696,7 @@ int mlxsw_sp_acl_rule_add(struct mlxsw_sp *mlxsw_sp, goto err_rhashtable_insert; list_add_tail(&rule->list, &mlxsw_sp->acl->rules); + ruleset->ht_key.block->rule_count++; return 0; err_rhashtable_insert: @@ -548,6 +710,7 @@ void mlxsw_sp_acl_rule_del(struct mlxsw_sp *mlxsw_sp, struct mlxsw_sp_acl_ruleset *ruleset = rule->ruleset; const struct mlxsw_sp_acl_profile_ops *ops = ruleset->ht_key.ops; + ruleset->ht_key.block->rule_count--; list_del(&rule->list); rhashtable_remove_fast(&ruleset->rule_ht, &rule->ht_node, mlxsw_sp_acl_rule_ht_params); diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_flex_actions.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_flex_actions.c index 4d3340e..6ca6894 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_flex_actions.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_flex_actions.c @@ -108,11 +108,77 @@ static void mlxsw_sp_act_kvdl_fwd_entry_del(void *priv, u32 kvdl_index) mlxsw_sp_kvdl_free(mlxsw_sp, kvdl_index); } +static int +mlxsw_sp_act_counter_index_get(void *priv, unsigned int *p_counter_index) +{ + struct mlxsw_sp *mlxsw_sp = priv; + + return mlxsw_sp_flow_counter_alloc(mlxsw_sp, p_counter_index); +} + +static void +mlxsw_sp_act_counter_index_put(void *priv, unsigned int counter_index) +{ + struct mlxsw_sp *mlxsw_sp = priv; + + mlxsw_sp_flow_counter_free(mlxsw_sp, counter_index); +} + +static int +mlxsw_sp_act_mirror_add(void *priv, u8 local_in_port, u8 local_out_port, + bool ingress, int *p_span_id) +{ + struct mlxsw_sp_port *in_port, *out_port; + struct mlxsw_sp_span_entry *span_entry; + struct mlxsw_sp *mlxsw_sp = priv; + enum mlxsw_sp_span_type type; + int err; + + type = ingress ? MLXSW_SP_SPAN_INGRESS : MLXSW_SP_SPAN_EGRESS; + out_port = mlxsw_sp->ports[local_out_port]; + in_port = mlxsw_sp->ports[local_in_port]; + + err = mlxsw_sp_span_mirror_add(in_port, out_port, type, false); + if (err) + return err; + + span_entry = mlxsw_sp_span_entry_find(mlxsw_sp, local_out_port); + if (!span_entry) { + err = -ENOENT; + goto err_span_entry_find; + } + + *p_span_id = span_entry->id; + return 0; + +err_span_entry_find: + mlxsw_sp_span_mirror_del(in_port, local_out_port, type, false); + return err; +} + +static void +mlxsw_sp_act_mirror_del(void *priv, u8 local_in_port, u8 local_out_port, + bool ingress) +{ + struct mlxsw_sp *mlxsw_sp = priv; + struct mlxsw_sp_port *in_port; + enum mlxsw_sp_span_type type; + + type = ingress ? MLXSW_SP_SPAN_INGRESS : MLXSW_SP_SPAN_EGRESS; + in_port = mlxsw_sp->ports[local_in_port]; + + mlxsw_sp_span_mirror_del(in_port, local_out_port, type, false); +} + static const struct mlxsw_afa_ops mlxsw_sp_act_afa_ops = { .kvdl_set_add = mlxsw_sp_act_kvdl_set_add, .kvdl_set_del = mlxsw_sp_act_kvdl_set_del, .kvdl_fwd_entry_add = mlxsw_sp_act_kvdl_fwd_entry_add, .kvdl_fwd_entry_del = mlxsw_sp_act_kvdl_fwd_entry_del, + .counter_index_get = mlxsw_sp_act_counter_index_get, + .counter_index_put = mlxsw_sp_act_counter_index_put, + .mirror_add = mlxsw_sp_act_mirror_add, + .mirror_del = mlxsw_sp_act_mirror_del, }; int mlxsw_sp_afa_init(struct mlxsw_sp *mlxsw_sp) diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c index 7e8284b..c6e180c 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c @@ -154,10 +154,6 @@ struct mlxsw_sp_acl_tcam_group { struct list_head region_list; unsigned int region_count; struct rhashtable chunk_ht; - struct { - u16 local_port; - bool ingress; - } bound; struct mlxsw_sp_acl_tcam_group_ops *ops; const struct mlxsw_sp_acl_tcam_pattern *patterns; unsigned int patterns_count; @@ -262,35 +258,29 @@ static void mlxsw_sp_acl_tcam_group_del(struct mlxsw_sp *mlxsw_sp, static int mlxsw_sp_acl_tcam_group_bind(struct mlxsw_sp *mlxsw_sp, struct mlxsw_sp_acl_tcam_group *group, - struct net_device *dev, bool ingress) + struct mlxsw_sp_port *mlxsw_sp_port, + bool ingress) { - struct mlxsw_sp_port *mlxsw_sp_port; char ppbt_pl[MLXSW_REG_PPBT_LEN]; - if (!mlxsw_sp_port_dev_check(dev)) - return -EINVAL; - - mlxsw_sp_port = netdev_priv(dev); - group->bound.local_port = mlxsw_sp_port->local_port; - group->bound.ingress = ingress; - mlxsw_reg_ppbt_pack(ppbt_pl, - group->bound.ingress ? MLXSW_REG_PXBT_E_IACL : - MLXSW_REG_PXBT_E_EACL, - MLXSW_REG_PXBT_OP_BIND, group->bound.local_port, + mlxsw_reg_ppbt_pack(ppbt_pl, ingress ? MLXSW_REG_PXBT_E_IACL : + MLXSW_REG_PXBT_E_EACL, + MLXSW_REG_PXBT_OP_BIND, mlxsw_sp_port->local_port, group->id); return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ppbt), ppbt_pl); } static void mlxsw_sp_acl_tcam_group_unbind(struct mlxsw_sp *mlxsw_sp, - struct mlxsw_sp_acl_tcam_group *group) + struct mlxsw_sp_acl_tcam_group *group, + struct mlxsw_sp_port *mlxsw_sp_port, + bool ingress) { char ppbt_pl[MLXSW_REG_PPBT_LEN]; - mlxsw_reg_ppbt_pack(ppbt_pl, - group->bound.ingress ? MLXSW_REG_PXBT_E_IACL : - MLXSW_REG_PXBT_E_EACL, - MLXSW_REG_PXBT_OP_UNBIND, group->bound.local_port, + mlxsw_reg_ppbt_pack(ppbt_pl, ingress ? MLXSW_REG_PXBT_E_IACL : + MLXSW_REG_PXBT_E_EACL, + MLXSW_REG_PXBT_OP_UNBIND, mlxsw_sp_port->local_port, group->id); mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ppbt), ppbt_pl); } @@ -1056,21 +1046,25 @@ mlxsw_sp_acl_tcam_flower_ruleset_del(struct mlxsw_sp *mlxsw_sp, static int mlxsw_sp_acl_tcam_flower_ruleset_bind(struct mlxsw_sp *mlxsw_sp, void *ruleset_priv, - struct net_device *dev, bool ingress) + struct mlxsw_sp_port *mlxsw_sp_port, + bool ingress) { struct mlxsw_sp_acl_tcam_flower_ruleset *ruleset = ruleset_priv; return mlxsw_sp_acl_tcam_group_bind(mlxsw_sp, &ruleset->group, - dev, ingress); + mlxsw_sp_port, ingress); } static void mlxsw_sp_acl_tcam_flower_ruleset_unbind(struct mlxsw_sp *mlxsw_sp, - void *ruleset_priv) + void *ruleset_priv, + struct mlxsw_sp_port *mlxsw_sp_port, + bool ingress) { struct mlxsw_sp_acl_tcam_flower_ruleset *ruleset = ruleset_priv; - mlxsw_sp_acl_tcam_group_unbind(mlxsw_sp, &ruleset->group); + mlxsw_sp_acl_tcam_group_unbind(mlxsw_sp, &ruleset->group, + mlxsw_sp_port, ingress); } static u16 diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_dpipe.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_dpipe.c index 96fdba7..f56fa18 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_dpipe.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_dpipe.c @@ -771,14 +771,33 @@ static struct devlink_dpipe_table_ops mlxsw_sp_host4_ops = { .size_get = mlxsw_sp_dpipe_table_host4_size_get, }; +#define MLXSW_SP_DPIPE_TABLE_RESOURCE_UNIT_HOST4 1 + static int mlxsw_sp_dpipe_host4_table_init(struct mlxsw_sp *mlxsw_sp) { struct devlink *devlink = priv_to_devlink(mlxsw_sp->core); + int err; - return devlink_dpipe_table_register(devlink, - MLXSW_SP_DPIPE_TABLE_NAME_HOST4, - &mlxsw_sp_host4_ops, - mlxsw_sp, false); + err = devlink_dpipe_table_register(devlink, + MLXSW_SP_DPIPE_TABLE_NAME_HOST4, + &mlxsw_sp_host4_ops, + mlxsw_sp, false); + if (err) + return err; + + err = devlink_dpipe_table_resource_set(devlink, + MLXSW_SP_DPIPE_TABLE_NAME_HOST4, + MLXSW_SP_RESOURCE_KVD_HASH_SINGLE, + MLXSW_SP_DPIPE_TABLE_RESOURCE_UNIT_HOST4); + if (err) + goto err_resource_set; + + return 0; + +err_resource_set: + devlink_dpipe_table_unregister(devlink, + MLXSW_SP_DPIPE_TABLE_NAME_HOST4); + return err; } static void mlxsw_sp_dpipe_host4_table_fini(struct mlxsw_sp *mlxsw_sp) @@ -829,14 +848,33 @@ static struct devlink_dpipe_table_ops mlxsw_sp_host6_ops = { .size_get = mlxsw_sp_dpipe_table_host6_size_get, }; +#define MLXSW_SP_DPIPE_TABLE_RESOURCE_UNIT_HOST6 2 + static int mlxsw_sp_dpipe_host6_table_init(struct mlxsw_sp *mlxsw_sp) { struct devlink *devlink = priv_to_devlink(mlxsw_sp->core); + int err; - return devlink_dpipe_table_register(devlink, - MLXSW_SP_DPIPE_TABLE_NAME_HOST6, - &mlxsw_sp_host6_ops, - mlxsw_sp, false); + err = devlink_dpipe_table_register(devlink, + MLXSW_SP_DPIPE_TABLE_NAME_HOST6, + &mlxsw_sp_host6_ops, + mlxsw_sp, false); + if (err) + return err; + + err = devlink_dpipe_table_resource_set(devlink, + MLXSW_SP_DPIPE_TABLE_NAME_HOST6, + MLXSW_SP_RESOURCE_KVD_HASH_DOUBLE, + MLXSW_SP_DPIPE_TABLE_RESOURCE_UNIT_HOST6); + if (err) + goto err_resource_set; + + return 0; + +err_resource_set: + devlink_dpipe_table_unregister(devlink, + MLXSW_SP_DPIPE_TABLE_NAME_HOST6); + return err; } static void mlxsw_sp_dpipe_host6_table_fini(struct mlxsw_sp *mlxsw_sp) @@ -1213,14 +1251,33 @@ static struct devlink_dpipe_table_ops mlxsw_sp_dpipe_table_adj_ops = { .size_get = mlxsw_sp_dpipe_table_adj_size_get, }; +#define MLXSW_SP_DPIPE_TABLE_RESOURCE_UNIT_ADJ 1 + static int mlxsw_sp_dpipe_adj_table_init(struct mlxsw_sp *mlxsw_sp) { struct devlink *devlink = priv_to_devlink(mlxsw_sp->core); + int err; - return devlink_dpipe_table_register(devlink, - MLXSW_SP_DPIPE_TABLE_NAME_ADJ, - &mlxsw_sp_dpipe_table_adj_ops, - mlxsw_sp, false); + err = devlink_dpipe_table_register(devlink, + MLXSW_SP_DPIPE_TABLE_NAME_ADJ, + &mlxsw_sp_dpipe_table_adj_ops, + mlxsw_sp, false); + if (err) + return err; + + err = devlink_dpipe_table_resource_set(devlink, + MLXSW_SP_DPIPE_TABLE_NAME_ADJ, + MLXSW_SP_RESOURCE_KVD_LINEAR, + MLXSW_SP_DPIPE_TABLE_RESOURCE_UNIT_ADJ); + if (err) + goto err_resource_set; + + return 0; + +err_resource_set: + devlink_dpipe_table_unregister(devlink, + MLXSW_SP_DPIPE_TABLE_NAME_ADJ); + return err; } static void mlxsw_sp_dpipe_adj_table_fini(struct mlxsw_sp *mlxsw_sp) diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_flower.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_flower.c index 2f0e578..6ce00e2 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_flower.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_flower.c @@ -35,6 +35,7 @@ #include <linux/kernel.h> #include <linux/errno.h> #include <linux/netdevice.h> +#include <net/net_namespace.h> #include <net/flow_dissector.h> #include <net/pkt_cls.h> #include <net/tc_act/tc_gact.h> @@ -45,7 +46,7 @@ #include "core_acl_flex_keys.h" static int mlxsw_sp_flower_parse_actions(struct mlxsw_sp *mlxsw_sp, - struct net_device *dev, bool ingress, + struct mlxsw_sp_acl_block *block, struct mlxsw_sp_acl_rule_info *rulei, struct tcf_exts *exts) { @@ -80,8 +81,7 @@ static int mlxsw_sp_flower_parse_actions(struct mlxsw_sp *mlxsw_sp, struct mlxsw_sp_acl_ruleset *ruleset; u16 group_id; - ruleset = mlxsw_sp_acl_ruleset_lookup(mlxsw_sp, dev, - ingress, + ruleset = mlxsw_sp_acl_ruleset_lookup(mlxsw_sp, block, chain_index, MLXSW_SP_ACL_PROFILE_FLOWER); if (IS_ERR(ruleset)) @@ -92,7 +92,6 @@ static int mlxsw_sp_flower_parse_actions(struct mlxsw_sp *mlxsw_sp, if (err) return err; } else if (is_tcf_mirred_egress_redirect(a)) { - int ifindex = tcf_mirred_ifindex(a); struct net_device *out_dev; struct mlxsw_sp_fid *fid; u16 fid_index; @@ -104,14 +103,18 @@ static int mlxsw_sp_flower_parse_actions(struct mlxsw_sp *mlxsw_sp, if (err) return err; - out_dev = __dev_get_by_index(dev_net(dev), ifindex); - if (out_dev == dev) - out_dev = NULL; - + out_dev = tcf_mirred_dev(a); err = mlxsw_sp_acl_rulei_act_fwd(mlxsw_sp, rulei, out_dev); if (err) return err; + } else if (is_tcf_mirred_egress_mirror(a)) { + struct net_device *out_dev = tcf_mirred_dev(a); + + err = mlxsw_sp_acl_rulei_act_mirror(mlxsw_sp, rulei, + block, out_dev); + if (err) + return err; } else if (is_tcf_vlan(a)) { u16 proto = be16_to_cpu(tcf_vlan_push_proto(a)); u32 action = tcf_vlan_action(a); @@ -266,7 +269,7 @@ static int mlxsw_sp_flower_parse_ip(struct mlxsw_sp *mlxsw_sp, } static int mlxsw_sp_flower_parse(struct mlxsw_sp *mlxsw_sp, - struct net_device *dev, bool ingress, + struct mlxsw_sp_acl_block *block, struct mlxsw_sp_acl_rule_info *rulei, struct tc_cls_flower_offload *f) { @@ -384,21 +387,19 @@ static int mlxsw_sp_flower_parse(struct mlxsw_sp *mlxsw_sp, if (err) return err; - return mlxsw_sp_flower_parse_actions(mlxsw_sp, dev, ingress, - rulei, f->exts); + return mlxsw_sp_flower_parse_actions(mlxsw_sp, block, rulei, f->exts); } -int mlxsw_sp_flower_replace(struct mlxsw_sp_port *mlxsw_sp_port, bool ingress, +int mlxsw_sp_flower_replace(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_block *block, struct tc_cls_flower_offload *f) { - struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; - struct net_device *dev = mlxsw_sp_port->dev; struct mlxsw_sp_acl_rule_info *rulei; struct mlxsw_sp_acl_ruleset *ruleset; struct mlxsw_sp_acl_rule *rule; int err; - ruleset = mlxsw_sp_acl_ruleset_get(mlxsw_sp, dev, ingress, + ruleset = mlxsw_sp_acl_ruleset_get(mlxsw_sp, block, f->common.chain_index, MLXSW_SP_ACL_PROFILE_FLOWER); if (IS_ERR(ruleset)) @@ -411,7 +412,7 @@ int mlxsw_sp_flower_replace(struct mlxsw_sp_port *mlxsw_sp_port, bool ingress, } rulei = mlxsw_sp_acl_rule_rulei(rule); - err = mlxsw_sp_flower_parse(mlxsw_sp, dev, ingress, rulei, f); + err = mlxsw_sp_flower_parse(mlxsw_sp, block, rulei, f); if (err) goto err_flower_parse; @@ -435,15 +436,15 @@ err_rule_create: return err; } -void mlxsw_sp_flower_destroy(struct mlxsw_sp_port *mlxsw_sp_port, bool ingress, +void mlxsw_sp_flower_destroy(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_block *block, struct tc_cls_flower_offload *f) { - struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; struct mlxsw_sp_acl_ruleset *ruleset; struct mlxsw_sp_acl_rule *rule; - ruleset = mlxsw_sp_acl_ruleset_get(mlxsw_sp, mlxsw_sp_port->dev, - ingress, f->common.chain_index, + ruleset = mlxsw_sp_acl_ruleset_get(mlxsw_sp, block, + f->common.chain_index, MLXSW_SP_ACL_PROFILE_FLOWER); if (IS_ERR(ruleset)) return; @@ -457,10 +458,10 @@ void mlxsw_sp_flower_destroy(struct mlxsw_sp_port *mlxsw_sp_port, bool ingress, mlxsw_sp_acl_ruleset_put(mlxsw_sp, ruleset); } -int mlxsw_sp_flower_stats(struct mlxsw_sp_port *mlxsw_sp_port, bool ingress, +int mlxsw_sp_flower_stats(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_block *block, struct tc_cls_flower_offload *f) { - struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; struct mlxsw_sp_acl_ruleset *ruleset; struct mlxsw_sp_acl_rule *rule; u64 packets; @@ -468,8 +469,8 @@ int mlxsw_sp_flower_stats(struct mlxsw_sp_port *mlxsw_sp_port, bool ingress, u64 bytes; int err; - ruleset = mlxsw_sp_acl_ruleset_get(mlxsw_sp, mlxsw_sp_port->dev, - ingress, f->common.chain_index, + ruleset = mlxsw_sp_acl_ruleset_get(mlxsw_sp, block, + f->common.chain_index, MLXSW_SP_ACL_PROFILE_FLOWER); if (WARN_ON(IS_ERR(ruleset))) return -EINVAL; diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_kvdl.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_kvdl.c index 310c382..55f9d2d 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_kvdl.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_kvdl.c @@ -286,6 +286,32 @@ static void mlxsw_sp_kvdl_parts_fini(struct mlxsw_sp *mlxsw_sp) mlxsw_sp_kvdl_part_fini(mlxsw_sp, i); } +static u64 mlxsw_sp_kvdl_part_occ(struct mlxsw_sp_kvdl_part *part) +{ + unsigned int nr_entries; + int bit = -1; + u64 occ = 0; + + nr_entries = (part->info->end_index - + part->info->start_index + 1) / + part->info->alloc_size; + while ((bit = find_next_bit(part->usage, nr_entries, bit + 1)) + < nr_entries) + occ += part->info->alloc_size; + return occ; +} + +u64 mlxsw_sp_kvdl_occ_get(const struct mlxsw_sp *mlxsw_sp) +{ + struct mlxsw_sp_kvdl_part *part; + u64 occ = 0; + + list_for_each_entry(part, &mlxsw_sp->kvdl->parts_list, list) + occ += mlxsw_sp_kvdl_part_occ(part); + + return occ; +} + int mlxsw_sp_kvdl_init(struct mlxsw_sp *mlxsw_sp) { struct mlxsw_sp_kvdl *kvdl; diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_mr_tcam.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_mr_tcam.c index 34a0b63..4c7f32d 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_mr_tcam.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_mr_tcam.c @@ -243,7 +243,8 @@ mlxsw_sp_mr_tcam_afa_block_create(struct mlxsw_sp *mlxsw_sp, if (!afa_block) return ERR_PTR(-ENOMEM); - err = mlxsw_afa_block_append_counter(afa_block, counter_index); + err = mlxsw_afa_block_append_allocated_counter(afa_block, + counter_index); if (err) goto err; diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_qdisc.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_qdisc.c index b5397da..0b76704 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_qdisc.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_qdisc.c @@ -41,6 +41,150 @@ #include "spectrum.h" #include "reg.h" +#define MLXSW_SP_PRIO_BAND_TO_TCLASS(band) (IEEE_8021QAZ_MAX_TCS - band - 1) + +enum mlxsw_sp_qdisc_type { + MLXSW_SP_QDISC_NO_QDISC, + MLXSW_SP_QDISC_RED, + MLXSW_SP_QDISC_PRIO, +}; + +struct mlxsw_sp_qdisc_ops { + enum mlxsw_sp_qdisc_type type; + int (*check_params)(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + void *params); + int (*replace)(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, void *params); + int (*destroy)(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc); + int (*get_stats)(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + struct tc_qopt_offload_stats *stats_ptr); + int (*get_xstats)(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + void *xstats_ptr); + void (*clean_stats)(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc); + /* unoffload - to be used for a qdisc that stops being offloaded without + * being destroyed. + */ + void (*unoffload)(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, void *params); +}; + +struct mlxsw_sp_qdisc { + u32 handle; + u8 tclass_num; + union { + struct red_stats red; + } xstats_base; + struct mlxsw_sp_qdisc_stats { + u64 tx_bytes; + u64 tx_packets; + u64 drops; + u64 overlimits; + u64 backlog; + } stats_base; + + struct mlxsw_sp_qdisc_ops *ops; +}; + +static bool +mlxsw_sp_qdisc_compare(struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, u32 handle, + enum mlxsw_sp_qdisc_type type) +{ + return mlxsw_sp_qdisc && mlxsw_sp_qdisc->ops && + mlxsw_sp_qdisc->ops->type == type && + mlxsw_sp_qdisc->handle == handle; +} + +static int +mlxsw_sp_qdisc_destroy(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc) +{ + int err = 0; + + if (!mlxsw_sp_qdisc) + return 0; + + if (mlxsw_sp_qdisc->ops && mlxsw_sp_qdisc->ops->destroy) + err = mlxsw_sp_qdisc->ops->destroy(mlxsw_sp_port, + mlxsw_sp_qdisc); + + mlxsw_sp_qdisc->handle = TC_H_UNSPEC; + mlxsw_sp_qdisc->ops = NULL; + return err; +} + +static int +mlxsw_sp_qdisc_replace(struct mlxsw_sp_port *mlxsw_sp_port, u32 handle, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + struct mlxsw_sp_qdisc_ops *ops, void *params) +{ + int err; + + if (mlxsw_sp_qdisc->ops && mlxsw_sp_qdisc->ops->type != ops->type) + /* In case this location contained a different qdisc of the + * same type we can override the old qdisc configuration. + * Otherwise, we need to remove the old qdisc before setting the + * new one. + */ + mlxsw_sp_qdisc_destroy(mlxsw_sp_port, mlxsw_sp_qdisc); + err = ops->check_params(mlxsw_sp_port, mlxsw_sp_qdisc, params); + if (err) + goto err_bad_param; + + err = ops->replace(mlxsw_sp_port, mlxsw_sp_qdisc, params); + if (err) + goto err_config; + + if (mlxsw_sp_qdisc->handle != handle) { + mlxsw_sp_qdisc->ops = ops; + if (ops->clean_stats) + ops->clean_stats(mlxsw_sp_port, mlxsw_sp_qdisc); + } + + mlxsw_sp_qdisc->handle = handle; + return 0; + +err_bad_param: +err_config: + if (mlxsw_sp_qdisc->handle == handle && ops->unoffload) + ops->unoffload(mlxsw_sp_port, mlxsw_sp_qdisc, params); + + mlxsw_sp_qdisc_destroy(mlxsw_sp_port, mlxsw_sp_qdisc); + return err; +} + +static int +mlxsw_sp_qdisc_get_stats(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + struct tc_qopt_offload_stats *stats_ptr) +{ + if (mlxsw_sp_qdisc && mlxsw_sp_qdisc->ops && + mlxsw_sp_qdisc->ops->get_stats) + return mlxsw_sp_qdisc->ops->get_stats(mlxsw_sp_port, + mlxsw_sp_qdisc, + stats_ptr); + + return -EOPNOTSUPP; +} + +static int +mlxsw_sp_qdisc_get_xstats(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + void *xstats_ptr) +{ + if (mlxsw_sp_qdisc && mlxsw_sp_qdisc->ops && + mlxsw_sp_qdisc->ops->get_xstats) + return mlxsw_sp_qdisc->ops->get_xstats(mlxsw_sp_port, + mlxsw_sp_qdisc, + xstats_ptr); + + return -EOPNOTSUPP; +} + static int mlxsw_sp_tclass_congestion_enable(struct mlxsw_sp_port *mlxsw_sp_port, int tclass_num, u32 min, u32 max, @@ -80,80 +224,78 @@ mlxsw_sp_tclass_congestion_disable(struct mlxsw_sp_port *mlxsw_sp_port, } static void -mlxsw_sp_setup_tc_qdisc_clean_stats(struct mlxsw_sp_port *mlxsw_sp_port, - struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, - int tclass_num) +mlxsw_sp_setup_tc_qdisc_red_clean_stats(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc) { - struct red_stats *xstats_base = &mlxsw_sp_qdisc->xstats_base; + u8 tclass_num = mlxsw_sp_qdisc->tclass_num; + struct mlxsw_sp_qdisc_stats *stats_base; struct mlxsw_sp_port_xstats *xstats; struct rtnl_link_stats64 *stats; + struct red_stats *red_base; xstats = &mlxsw_sp_port->periodic_hw_stats.xstats; stats = &mlxsw_sp_port->periodic_hw_stats.stats; + stats_base = &mlxsw_sp_qdisc->stats_base; + red_base = &mlxsw_sp_qdisc->xstats_base.red; - mlxsw_sp_qdisc->tx_packets = stats->tx_packets; - mlxsw_sp_qdisc->tx_bytes = stats->tx_bytes; + stats_base->tx_packets = stats->tx_packets; + stats_base->tx_bytes = stats->tx_bytes; - switch (mlxsw_sp_qdisc->type) { - case MLXSW_SP_QDISC_RED: - xstats_base->prob_mark = xstats->ecn; - xstats_base->prob_drop = xstats->wred_drop[tclass_num]; - xstats_base->pdrop = xstats->tail_drop[tclass_num]; + red_base->prob_mark = xstats->ecn; + red_base->prob_drop = xstats->wred_drop[tclass_num]; + red_base->pdrop = xstats->tail_drop[tclass_num]; - mlxsw_sp_qdisc->overlimits = xstats_base->prob_drop + - xstats_base->prob_mark; - mlxsw_sp_qdisc->drops = xstats_base->prob_drop + - xstats_base->pdrop; - break; - default: - break; - } + stats_base->overlimits = red_base->prob_drop + red_base->prob_mark; + stats_base->drops = red_base->prob_drop + red_base->pdrop; + + stats_base->backlog = 0; } static int -mlxsw_sp_qdisc_red_destroy(struct mlxsw_sp_port *mlxsw_sp_port, u32 handle, - struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, - int tclass_num) +mlxsw_sp_qdisc_red_destroy(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc) { - int err; - - if (mlxsw_sp_qdisc->handle != handle) - return 0; - - err = mlxsw_sp_tclass_congestion_disable(mlxsw_sp_port, tclass_num); - mlxsw_sp_qdisc->handle = TC_H_UNSPEC; - mlxsw_sp_qdisc->type = MLXSW_SP_QDISC_NO_QDISC; - - return err; + return mlxsw_sp_tclass_congestion_disable(mlxsw_sp_port, + mlxsw_sp_qdisc->tclass_num); } static int -mlxsw_sp_qdisc_red_replace(struct mlxsw_sp_port *mlxsw_sp_port, u32 handle, - struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, - int tclass_num, - struct tc_red_qopt_offload_params *p) +mlxsw_sp_qdisc_red_check_params(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + void *params) { struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; - u32 min, max; - u64 prob; - int err = 0; + struct tc_red_qopt_offload_params *p = params; if (p->min > p->max) { dev_err(mlxsw_sp->bus_info->dev, "spectrum: RED: min %u is bigger then max %u\n", p->min, p->max); - goto err_bad_param; + return -EINVAL; } if (p->max > MLXSW_CORE_RES_GET(mlxsw_sp->core, MAX_BUFFER_SIZE)) { dev_err(mlxsw_sp->bus_info->dev, "spectrum: RED: max value %u is too big\n", p->max); - goto err_bad_param; + return -EINVAL; } if (p->min == 0 || p->max == 0) { dev_err(mlxsw_sp->bus_info->dev, "spectrum: RED: 0 value is illegal for min and max\n"); - goto err_bad_param; + return -EINVAL; } + return 0; +} + +static int +mlxsw_sp_qdisc_red_replace(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + void *params) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + struct tc_red_qopt_offload_params *p = params; + u8 tclass_num = mlxsw_sp_qdisc->tclass_num; + u32 min, max; + u64 prob; /* calculate probability in percentage */ prob = p->probability; @@ -162,116 +304,309 @@ mlxsw_sp_qdisc_red_replace(struct mlxsw_sp_port *mlxsw_sp_port, u32 handle, prob = DIV_ROUND_UP(prob, 1 << 16); min = mlxsw_sp_bytes_cells(mlxsw_sp, p->min); max = mlxsw_sp_bytes_cells(mlxsw_sp, p->max); - err = mlxsw_sp_tclass_congestion_enable(mlxsw_sp_port, tclass_num, min, - max, prob, p->is_ecn); - if (err) - goto err_config; - - mlxsw_sp_qdisc->type = MLXSW_SP_QDISC_RED; - if (mlxsw_sp_qdisc->handle != handle) - mlxsw_sp_setup_tc_qdisc_clean_stats(mlxsw_sp_port, - mlxsw_sp_qdisc, - tclass_num); + return mlxsw_sp_tclass_congestion_enable(mlxsw_sp_port, tclass_num, min, + max, prob, p->is_ecn); +} - mlxsw_sp_qdisc->handle = handle; - return 0; +static void +mlxsw_sp_qdisc_red_unoffload(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + void *params) +{ + struct tc_red_qopt_offload_params *p = params; + u64 backlog; -err_bad_param: - err = -EINVAL; -err_config: - mlxsw_sp_qdisc_red_destroy(mlxsw_sp_port, mlxsw_sp_qdisc->handle, - mlxsw_sp_qdisc, tclass_num); - return err; + backlog = mlxsw_sp_cells_bytes(mlxsw_sp_port->mlxsw_sp, + mlxsw_sp_qdisc->stats_base.backlog); + p->qstats->backlog -= backlog; } static int -mlxsw_sp_qdisc_get_red_xstats(struct mlxsw_sp_port *mlxsw_sp_port, u32 handle, +mlxsw_sp_qdisc_get_red_xstats(struct mlxsw_sp_port *mlxsw_sp_port, struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, - int tclass_num, struct red_stats *res) + void *xstats_ptr) { - struct red_stats *xstats_base = &mlxsw_sp_qdisc->xstats_base; + struct red_stats *xstats_base = &mlxsw_sp_qdisc->xstats_base.red; + u8 tclass_num = mlxsw_sp_qdisc->tclass_num; struct mlxsw_sp_port_xstats *xstats; - - if (mlxsw_sp_qdisc->handle != handle || - mlxsw_sp_qdisc->type != MLXSW_SP_QDISC_RED) - return -EOPNOTSUPP; + struct red_stats *res = xstats_ptr; + int early_drops, marks, pdrops; xstats = &mlxsw_sp_port->periodic_hw_stats.xstats; - res->prob_drop = xstats->wred_drop[tclass_num] - xstats_base->prob_drop; - res->prob_mark = xstats->ecn - xstats_base->prob_mark; - res->pdrop = xstats->tail_drop[tclass_num] - xstats_base->pdrop; + early_drops = xstats->wred_drop[tclass_num] - xstats_base->prob_drop; + marks = xstats->ecn - xstats_base->prob_mark; + pdrops = xstats->tail_drop[tclass_num] - xstats_base->pdrop; + + res->pdrop += pdrops; + res->prob_drop += early_drops; + res->prob_mark += marks; + + xstats_base->pdrop += pdrops; + xstats_base->prob_drop += early_drops; + xstats_base->prob_mark += marks; return 0; } static int -mlxsw_sp_qdisc_get_red_stats(struct mlxsw_sp_port *mlxsw_sp_port, u32 handle, +mlxsw_sp_qdisc_get_red_stats(struct mlxsw_sp_port *mlxsw_sp_port, struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, - int tclass_num, - struct tc_red_qopt_offload_stats *res) + struct tc_qopt_offload_stats *stats_ptr) { - u64 tx_bytes, tx_packets, overlimits, drops; + u64 tx_bytes, tx_packets, overlimits, drops, backlog; + u8 tclass_num = mlxsw_sp_qdisc->tclass_num; + struct mlxsw_sp_qdisc_stats *stats_base; struct mlxsw_sp_port_xstats *xstats; struct rtnl_link_stats64 *stats; - if (mlxsw_sp_qdisc->handle != handle || - mlxsw_sp_qdisc->type != MLXSW_SP_QDISC_RED) - return -EOPNOTSUPP; - xstats = &mlxsw_sp_port->periodic_hw_stats.xstats; stats = &mlxsw_sp_port->periodic_hw_stats.stats; + stats_base = &mlxsw_sp_qdisc->stats_base; - tx_bytes = stats->tx_bytes - mlxsw_sp_qdisc->tx_bytes; - tx_packets = stats->tx_packets - mlxsw_sp_qdisc->tx_packets; + tx_bytes = stats->tx_bytes - stats_base->tx_bytes; + tx_packets = stats->tx_packets - stats_base->tx_packets; overlimits = xstats->wred_drop[tclass_num] + xstats->ecn - - mlxsw_sp_qdisc->overlimits; + stats_base->overlimits; drops = xstats->wred_drop[tclass_num] + xstats->tail_drop[tclass_num] - - mlxsw_sp_qdisc->drops; - - _bstats_update(res->bstats, tx_bytes, tx_packets); - res->qstats->overlimits += overlimits; - res->qstats->drops += drops; - res->qstats->backlog += mlxsw_sp_cells_bytes(mlxsw_sp_port->mlxsw_sp, - xstats->backlog[tclass_num]); - - mlxsw_sp_qdisc->drops += drops; - mlxsw_sp_qdisc->overlimits += overlimits; - mlxsw_sp_qdisc->tx_bytes += tx_bytes; - mlxsw_sp_qdisc->tx_packets += tx_packets; + stats_base->drops; + backlog = xstats->backlog[tclass_num]; + + _bstats_update(stats_ptr->bstats, tx_bytes, tx_packets); + stats_ptr->qstats->overlimits += overlimits; + stats_ptr->qstats->drops += drops; + stats_ptr->qstats->backlog += + mlxsw_sp_cells_bytes(mlxsw_sp_port->mlxsw_sp, + backlog) - + mlxsw_sp_cells_bytes(mlxsw_sp_port->mlxsw_sp, + stats_base->backlog); + + stats_base->backlog = backlog; + stats_base->drops += drops; + stats_base->overlimits += overlimits; + stats_base->tx_bytes += tx_bytes; + stats_base->tx_packets += tx_packets; return 0; } #define MLXSW_SP_PORT_DEFAULT_TCLASS 0 +static struct mlxsw_sp_qdisc_ops mlxsw_sp_qdisc_ops_red = { + .type = MLXSW_SP_QDISC_RED, + .check_params = mlxsw_sp_qdisc_red_check_params, + .replace = mlxsw_sp_qdisc_red_replace, + .unoffload = mlxsw_sp_qdisc_red_unoffload, + .destroy = mlxsw_sp_qdisc_red_destroy, + .get_stats = mlxsw_sp_qdisc_get_red_stats, + .get_xstats = mlxsw_sp_qdisc_get_red_xstats, + .clean_stats = mlxsw_sp_setup_tc_qdisc_red_clean_stats, +}; + int mlxsw_sp_setup_tc_red(struct mlxsw_sp_port *mlxsw_sp_port, struct tc_red_qopt_offload *p) { struct mlxsw_sp_qdisc *mlxsw_sp_qdisc; - int tclass_num; if (p->parent != TC_H_ROOT) return -EOPNOTSUPP; - mlxsw_sp_qdisc = &mlxsw_sp_port->root_qdisc; - tclass_num = MLXSW_SP_PORT_DEFAULT_TCLASS; + mlxsw_sp_qdisc = mlxsw_sp_port->root_qdisc; + + if (p->command == TC_RED_REPLACE) + return mlxsw_sp_qdisc_replace(mlxsw_sp_port, p->handle, + mlxsw_sp_qdisc, + &mlxsw_sp_qdisc_ops_red, + &p->set); + + if (!mlxsw_sp_qdisc_compare(mlxsw_sp_qdisc, p->handle, + MLXSW_SP_QDISC_RED)) + return -EOPNOTSUPP; switch (p->command) { - case TC_RED_REPLACE: - return mlxsw_sp_qdisc_red_replace(mlxsw_sp_port, p->handle, - mlxsw_sp_qdisc, tclass_num, - &p->set); case TC_RED_DESTROY: - return mlxsw_sp_qdisc_red_destroy(mlxsw_sp_port, p->handle, - mlxsw_sp_qdisc, tclass_num); + return mlxsw_sp_qdisc_destroy(mlxsw_sp_port, mlxsw_sp_qdisc); case TC_RED_XSTATS: - return mlxsw_sp_qdisc_get_red_xstats(mlxsw_sp_port, p->handle, - mlxsw_sp_qdisc, tclass_num, - p->xstats); + return mlxsw_sp_qdisc_get_xstats(mlxsw_sp_port, mlxsw_sp_qdisc, + p->xstats); case TC_RED_STATS: - return mlxsw_sp_qdisc_get_red_stats(mlxsw_sp_port, p->handle, - mlxsw_sp_qdisc, tclass_num, - &p->stats); + return mlxsw_sp_qdisc_get_stats(mlxsw_sp_port, mlxsw_sp_qdisc, + &p->stats); default: return -EOPNOTSUPP; } } + +static int +mlxsw_sp_qdisc_prio_destroy(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc) +{ + int i; + + for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++) + mlxsw_sp_port_prio_tc_set(mlxsw_sp_port, i, + MLXSW_SP_PORT_DEFAULT_TCLASS); + + return 0; +} + +static int +mlxsw_sp_qdisc_prio_check_params(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + void *params) +{ + struct tc_prio_qopt_offload_params *p = params; + + if (p->bands > IEEE_8021QAZ_MAX_TCS) + return -EOPNOTSUPP; + + return 0; +} + +static int +mlxsw_sp_qdisc_prio_replace(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + void *params) +{ + struct tc_prio_qopt_offload_params *p = params; + int tclass, i; + int err; + + for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++) { + tclass = MLXSW_SP_PRIO_BAND_TO_TCLASS(p->priomap[i]); + err = mlxsw_sp_port_prio_tc_set(mlxsw_sp_port, i, tclass); + if (err) + return err; + } + + return 0; +} + +static void +mlxsw_sp_qdisc_prio_unoffload(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + void *params) +{ + struct tc_prio_qopt_offload_params *p = params; + u64 backlog; + + backlog = mlxsw_sp_cells_bytes(mlxsw_sp_port->mlxsw_sp, + mlxsw_sp_qdisc->stats_base.backlog); + p->qstats->backlog -= backlog; +} + +static int +mlxsw_sp_qdisc_get_prio_stats(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + struct tc_qopt_offload_stats *stats_ptr) +{ + u64 tx_bytes, tx_packets, drops = 0, backlog = 0; + struct mlxsw_sp_qdisc_stats *stats_base; + struct mlxsw_sp_port_xstats *xstats; + struct rtnl_link_stats64 *stats; + int i; + + xstats = &mlxsw_sp_port->periodic_hw_stats.xstats; + stats = &mlxsw_sp_port->periodic_hw_stats.stats; + stats_base = &mlxsw_sp_qdisc->stats_base; + + tx_bytes = stats->tx_bytes - stats_base->tx_bytes; + tx_packets = stats->tx_packets - stats_base->tx_packets; + + for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++) { + drops += xstats->tail_drop[i]; + backlog += xstats->backlog[i]; + } + drops = drops - stats_base->drops; + + _bstats_update(stats_ptr->bstats, tx_bytes, tx_packets); + stats_ptr->qstats->drops += drops; + stats_ptr->qstats->backlog += + mlxsw_sp_cells_bytes(mlxsw_sp_port->mlxsw_sp, + backlog) - + mlxsw_sp_cells_bytes(mlxsw_sp_port->mlxsw_sp, + stats_base->backlog); + stats_base->backlog = backlog; + stats_base->drops += drops; + stats_base->tx_bytes += tx_bytes; + stats_base->tx_packets += tx_packets; + return 0; +} + +static void +mlxsw_sp_setup_tc_qdisc_prio_clean_stats(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc) +{ + struct mlxsw_sp_qdisc_stats *stats_base; + struct mlxsw_sp_port_xstats *xstats; + struct rtnl_link_stats64 *stats; + int i; + + xstats = &mlxsw_sp_port->periodic_hw_stats.xstats; + stats = &mlxsw_sp_port->periodic_hw_stats.stats; + stats_base = &mlxsw_sp_qdisc->stats_base; + + stats_base->tx_packets = stats->tx_packets; + stats_base->tx_bytes = stats->tx_bytes; + + stats_base->drops = 0; + for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++) + stats_base->drops += xstats->tail_drop[i]; + + mlxsw_sp_qdisc->stats_base.backlog = 0; +} + +static struct mlxsw_sp_qdisc_ops mlxsw_sp_qdisc_ops_prio = { + .type = MLXSW_SP_QDISC_PRIO, + .check_params = mlxsw_sp_qdisc_prio_check_params, + .replace = mlxsw_sp_qdisc_prio_replace, + .unoffload = mlxsw_sp_qdisc_prio_unoffload, + .destroy = mlxsw_sp_qdisc_prio_destroy, + .get_stats = mlxsw_sp_qdisc_get_prio_stats, + .clean_stats = mlxsw_sp_setup_tc_qdisc_prio_clean_stats, +}; + +int mlxsw_sp_setup_tc_prio(struct mlxsw_sp_port *mlxsw_sp_port, + struct tc_prio_qopt_offload *p) +{ + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc; + + if (p->parent != TC_H_ROOT) + return -EOPNOTSUPP; + + mlxsw_sp_qdisc = mlxsw_sp_port->root_qdisc; + if (p->command == TC_PRIO_REPLACE) + return mlxsw_sp_qdisc_replace(mlxsw_sp_port, p->handle, + mlxsw_sp_qdisc, + &mlxsw_sp_qdisc_ops_prio, + &p->replace_params); + + if (!mlxsw_sp_qdisc_compare(mlxsw_sp_qdisc, p->handle, + MLXSW_SP_QDISC_PRIO)) + return -EOPNOTSUPP; + + switch (p->command) { + case TC_PRIO_DESTROY: + return mlxsw_sp_qdisc_destroy(mlxsw_sp_port, mlxsw_sp_qdisc); + case TC_PRIO_STATS: + return mlxsw_sp_qdisc_get_stats(mlxsw_sp_port, mlxsw_sp_qdisc, + &p->stats); + default: + return -EOPNOTSUPP; + } +} + +int mlxsw_sp_tc_qdisc_init(struct mlxsw_sp_port *mlxsw_sp_port) +{ + mlxsw_sp_port->root_qdisc = kzalloc(sizeof(*mlxsw_sp_port->root_qdisc), + GFP_KERNEL); + if (!mlxsw_sp_port->root_qdisc) + return -ENOMEM; + + mlxsw_sp_port->root_qdisc->tclass_num = MLXSW_SP_PORT_DEFAULT_TCLASS; + + return 0; +} + +void mlxsw_sp_tc_qdisc_fini(struct mlxsw_sp_port *mlxsw_sp_port) +{ + kfree(mlxsw_sp_port->root_qdisc); +} diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c index 7042c85..f0b25ba 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c @@ -71,6 +71,7 @@ #include "spectrum_mr_tcam.h" #include "spectrum_router.h" +struct mlxsw_sp_fib; struct mlxsw_sp_vr; struct mlxsw_sp_lpm_tree; struct mlxsw_sp_rif_ops; @@ -84,6 +85,8 @@ struct mlxsw_sp_router { struct rhashtable nexthop_ht; struct list_head nexthop_list; struct { + /* One tree for each protocol: IPv4 and IPv6 */ + struct mlxsw_sp_lpm_tree *proto_trees[2]; struct mlxsw_sp_lpm_tree *trees; unsigned int tree_count; } lpm; @@ -162,6 +165,15 @@ struct mlxsw_sp_rif_ops { struct mlxsw_sp_fid * (*fid_get)(struct mlxsw_sp_rif *rif); }; +static void mlxsw_sp_lpm_tree_hold(struct mlxsw_sp_lpm_tree *lpm_tree); +static void mlxsw_sp_lpm_tree_put(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_lpm_tree *lpm_tree); +static int mlxsw_sp_vr_lpm_tree_bind(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_fib *fib, + u8 tree_id); +static int mlxsw_sp_vr_lpm_tree_unbind(struct mlxsw_sp *mlxsw_sp, + const struct mlxsw_sp_fib *fib); + static unsigned int * mlxsw_sp_rif_p_counter_get(struct mlxsw_sp_rif *rif, enum mlxsw_sp_rif_counter_dir dir) @@ -349,14 +361,6 @@ mlxsw_sp_prefix_usage_eq(struct mlxsw_sp_prefix_usage *prefix_usage1, return !memcmp(prefix_usage1, prefix_usage2, sizeof(*prefix_usage1)); } -static bool -mlxsw_sp_prefix_usage_none(struct mlxsw_sp_prefix_usage *prefix_usage) -{ - struct mlxsw_sp_prefix_usage prefix_usage_none = {{ 0 } }; - - return mlxsw_sp_prefix_usage_eq(prefix_usage, &prefix_usage_none); -} - static void mlxsw_sp_prefix_usage_cpy(struct mlxsw_sp_prefix_usage *prefix_usage1, struct mlxsw_sp_prefix_usage *prefix_usage2) @@ -398,7 +402,6 @@ enum mlxsw_sp_fib_entry_type { }; struct mlxsw_sp_nexthop_group; -struct mlxsw_sp_fib; struct mlxsw_sp_fib_node { struct list_head entry_list; @@ -445,6 +448,7 @@ struct mlxsw_sp_lpm_tree { u8 id; /* tree ID */ unsigned int ref_count; enum mlxsw_sp_l3proto proto; + unsigned long prefix_ref_count[MLXSW_SP_PREFIX_COUNT]; struct mlxsw_sp_prefix_usage prefix_usage; }; @@ -453,8 +457,6 @@ struct mlxsw_sp_fib { struct list_head node_list; struct mlxsw_sp_vr *vr; struct mlxsw_sp_lpm_tree *lpm_tree; - unsigned long prefix_ref_count[MLXSW_SP_PREFIX_COUNT]; - struct mlxsw_sp_prefix_usage prefix_usage; enum mlxsw_sp_l3proto proto; }; @@ -469,12 +471,15 @@ struct mlxsw_sp_vr { static const struct rhashtable_params mlxsw_sp_fib_ht_params; -static struct mlxsw_sp_fib *mlxsw_sp_fib_create(struct mlxsw_sp_vr *vr, +static struct mlxsw_sp_fib *mlxsw_sp_fib_create(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_vr *vr, enum mlxsw_sp_l3proto proto) { + struct mlxsw_sp_lpm_tree *lpm_tree; struct mlxsw_sp_fib *fib; int err; + lpm_tree = mlxsw_sp->router->lpm.proto_trees[proto]; fib = kzalloc(sizeof(*fib), GFP_KERNEL); if (!fib) return ERR_PTR(-ENOMEM); @@ -484,17 +489,26 @@ static struct mlxsw_sp_fib *mlxsw_sp_fib_create(struct mlxsw_sp_vr *vr, INIT_LIST_HEAD(&fib->node_list); fib->proto = proto; fib->vr = vr; + fib->lpm_tree = lpm_tree; + mlxsw_sp_lpm_tree_hold(lpm_tree); + err = mlxsw_sp_vr_lpm_tree_bind(mlxsw_sp, fib, lpm_tree->id); + if (err) + goto err_lpm_tree_bind; return fib; +err_lpm_tree_bind: + mlxsw_sp_lpm_tree_put(mlxsw_sp, lpm_tree); err_rhashtable_init: kfree(fib); return ERR_PTR(err); } -static void mlxsw_sp_fib_destroy(struct mlxsw_sp_fib *fib) +static void mlxsw_sp_fib_destroy(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib *fib) { + mlxsw_sp_vr_lpm_tree_unbind(mlxsw_sp, fib); + mlxsw_sp_lpm_tree_put(mlxsw_sp, fib->lpm_tree); WARN_ON(!list_empty(&fib->node_list)); - WARN_ON(fib->lpm_tree); rhashtable_destroy(&fib->ht); kfree(fib); } @@ -581,6 +595,9 @@ mlxsw_sp_lpm_tree_create(struct mlxsw_sp *mlxsw_sp, goto err_left_struct_set; memcpy(&lpm_tree->prefix_usage, prefix_usage, sizeof(lpm_tree->prefix_usage)); + memset(&lpm_tree->prefix_ref_count, 0, + sizeof(lpm_tree->prefix_ref_count)); + lpm_tree->ref_count = 1; return lpm_tree; err_left_struct_set: @@ -607,8 +624,10 @@ mlxsw_sp_lpm_tree_get(struct mlxsw_sp *mlxsw_sp, if (lpm_tree->ref_count != 0 && lpm_tree->proto == proto && mlxsw_sp_prefix_usage_eq(&lpm_tree->prefix_usage, - prefix_usage)) + prefix_usage)) { + mlxsw_sp_lpm_tree_hold(lpm_tree); return lpm_tree; + } } return mlxsw_sp_lpm_tree_create(mlxsw_sp, prefix_usage, proto); } @@ -629,9 +648,10 @@ static void mlxsw_sp_lpm_tree_put(struct mlxsw_sp *mlxsw_sp, static int mlxsw_sp_lpm_init(struct mlxsw_sp *mlxsw_sp) { + struct mlxsw_sp_prefix_usage req_prefix_usage = {{ 0 } }; struct mlxsw_sp_lpm_tree *lpm_tree; u64 max_trees; - int i; + int err, i; if (!MLXSW_CORE_RES_VALID(mlxsw_sp->core, MAX_LPM_TREES)) return -EIO; @@ -649,11 +669,42 @@ static int mlxsw_sp_lpm_init(struct mlxsw_sp *mlxsw_sp) lpm_tree->id = i + MLXSW_SP_LPM_TREE_MIN; } + lpm_tree = mlxsw_sp_lpm_tree_get(mlxsw_sp, &req_prefix_usage, + MLXSW_SP_L3_PROTO_IPV4); + if (IS_ERR(lpm_tree)) { + err = PTR_ERR(lpm_tree); + goto err_ipv4_tree_get; + } + mlxsw_sp->router->lpm.proto_trees[MLXSW_SP_L3_PROTO_IPV4] = lpm_tree; + + lpm_tree = mlxsw_sp_lpm_tree_get(mlxsw_sp, &req_prefix_usage, + MLXSW_SP_L3_PROTO_IPV6); + if (IS_ERR(lpm_tree)) { + err = PTR_ERR(lpm_tree); + goto err_ipv6_tree_get; + } + mlxsw_sp->router->lpm.proto_trees[MLXSW_SP_L3_PROTO_IPV6] = lpm_tree; + return 0; + +err_ipv6_tree_get: + lpm_tree = mlxsw_sp->router->lpm.proto_trees[MLXSW_SP_L3_PROTO_IPV4]; + mlxsw_sp_lpm_tree_put(mlxsw_sp, lpm_tree); +err_ipv4_tree_get: + kfree(mlxsw_sp->router->lpm.trees); + return err; } static void mlxsw_sp_lpm_fini(struct mlxsw_sp *mlxsw_sp) { + struct mlxsw_sp_lpm_tree *lpm_tree; + + lpm_tree = mlxsw_sp->router->lpm.proto_trees[MLXSW_SP_L3_PROTO_IPV6]; + mlxsw_sp_lpm_tree_put(mlxsw_sp, lpm_tree); + + lpm_tree = mlxsw_sp->router->lpm.proto_trees[MLXSW_SP_L3_PROTO_IPV4]; + mlxsw_sp_lpm_tree_put(mlxsw_sp, lpm_tree); + kfree(mlxsw_sp->router->lpm.trees); } @@ -745,10 +796,10 @@ static struct mlxsw_sp_vr *mlxsw_sp_vr_create(struct mlxsw_sp *mlxsw_sp, NL_SET_ERR_MSG(extack, "spectrum: Exceeded number of supported virtual routers"); return ERR_PTR(-EBUSY); } - vr->fib4 = mlxsw_sp_fib_create(vr, MLXSW_SP_L3_PROTO_IPV4); + vr->fib4 = mlxsw_sp_fib_create(mlxsw_sp, vr, MLXSW_SP_L3_PROTO_IPV4); if (IS_ERR(vr->fib4)) return ERR_CAST(vr->fib4); - vr->fib6 = mlxsw_sp_fib_create(vr, MLXSW_SP_L3_PROTO_IPV6); + vr->fib6 = mlxsw_sp_fib_create(mlxsw_sp, vr, MLXSW_SP_L3_PROTO_IPV6); if (IS_ERR(vr->fib6)) { err = PTR_ERR(vr->fib6); goto err_fib6_create; @@ -763,21 +814,22 @@ static struct mlxsw_sp_vr *mlxsw_sp_vr_create(struct mlxsw_sp *mlxsw_sp, return vr; err_mr_table_create: - mlxsw_sp_fib_destroy(vr->fib6); + mlxsw_sp_fib_destroy(mlxsw_sp, vr->fib6); vr->fib6 = NULL; err_fib6_create: - mlxsw_sp_fib_destroy(vr->fib4); + mlxsw_sp_fib_destroy(mlxsw_sp, vr->fib4); vr->fib4 = NULL; return ERR_PTR(err); } -static void mlxsw_sp_vr_destroy(struct mlxsw_sp_vr *vr) +static void mlxsw_sp_vr_destroy(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_vr *vr) { mlxsw_sp_mr_table_destroy(vr->mr4_table); vr->mr4_table = NULL; - mlxsw_sp_fib_destroy(vr->fib6); + mlxsw_sp_fib_destroy(mlxsw_sp, vr->fib6); vr->fib6 = NULL; - mlxsw_sp_fib_destroy(vr->fib4); + mlxsw_sp_fib_destroy(mlxsw_sp, vr->fib4); vr->fib4 = NULL; } @@ -793,12 +845,12 @@ static struct mlxsw_sp_vr *mlxsw_sp_vr_get(struct mlxsw_sp *mlxsw_sp, u32 tb_id, return vr; } -static void mlxsw_sp_vr_put(struct mlxsw_sp_vr *vr) +static void mlxsw_sp_vr_put(struct mlxsw_sp *mlxsw_sp, struct mlxsw_sp_vr *vr) { if (!vr->rif_count && list_empty(&vr->fib4->node_list) && list_empty(&vr->fib6->node_list) && mlxsw_sp_mr_table_empty(vr->mr4_table)) - mlxsw_sp_vr_destroy(vr); + mlxsw_sp_vr_destroy(mlxsw_sp, vr); } static bool @@ -809,7 +861,7 @@ mlxsw_sp_vr_lpm_tree_should_replace(struct mlxsw_sp_vr *vr, if (!mlxsw_sp_vr_is_used(vr)) return false; - if (fib->lpm_tree && fib->lpm_tree->id == tree_id) + if (fib->lpm_tree->id == tree_id) return true; return false; } @@ -839,14 +891,13 @@ static int mlxsw_sp_vrs_lpm_tree_replace(struct mlxsw_sp *mlxsw_sp, struct mlxsw_sp_fib *fib, struct mlxsw_sp_lpm_tree *new_tree) { - struct mlxsw_sp_lpm_tree *old_tree = fib->lpm_tree; enum mlxsw_sp_l3proto proto = fib->proto; + struct mlxsw_sp_lpm_tree *old_tree; u8 old_id, new_id = new_tree->id; struct mlxsw_sp_vr *vr; int i, err; - if (!old_tree) - goto no_replace; + old_tree = mlxsw_sp->router->lpm.proto_trees[proto]; old_id = old_tree->id; for (i = 0; i < MLXSW_CORE_RES_GET(mlxsw_sp->core, MAX_VRS); i++) { @@ -860,6 +911,11 @@ static int mlxsw_sp_vrs_lpm_tree_replace(struct mlxsw_sp *mlxsw_sp, goto err_tree_replace; } + memcpy(new_tree->prefix_ref_count, old_tree->prefix_ref_count, + sizeof(new_tree->prefix_ref_count)); + mlxsw_sp->router->lpm.proto_trees[proto] = new_tree; + mlxsw_sp_lpm_tree_put(mlxsw_sp, old_tree); + return 0; err_tree_replace: @@ -871,36 +927,6 @@ err_tree_replace: old_tree); } return err; - -no_replace: - fib->lpm_tree = new_tree; - mlxsw_sp_lpm_tree_hold(new_tree); - err = mlxsw_sp_vr_lpm_tree_bind(mlxsw_sp, fib, new_tree->id); - if (err) { - mlxsw_sp_lpm_tree_put(mlxsw_sp, new_tree); - fib->lpm_tree = NULL; - return err; - } - return 0; -} - -static void -mlxsw_sp_vrs_prefixes(struct mlxsw_sp *mlxsw_sp, - enum mlxsw_sp_l3proto proto, - struct mlxsw_sp_prefix_usage *req_prefix_usage) -{ - int i; - - for (i = 0; i < MLXSW_CORE_RES_GET(mlxsw_sp->core, MAX_VRS); i++) { - struct mlxsw_sp_vr *vr = &mlxsw_sp->router->vrs[i]; - struct mlxsw_sp_fib *fib = mlxsw_sp_vr_fib(vr, proto); - unsigned char prefix; - - if (!mlxsw_sp_vr_is_used(vr)) - continue; - mlxsw_sp_prefix_usage_for_each(prefix, &fib->prefix_usage) - mlxsw_sp_prefix_usage_set(req_prefix_usage, prefix); - } } static int mlxsw_sp_vrs_init(struct mlxsw_sp *mlxsw_sp) @@ -2630,7 +2656,8 @@ struct mlxsw_sp_nexthop_group_cmp_arg { static bool mlxsw_sp_nexthop6_group_has_nexthop(const struct mlxsw_sp_nexthop_group *nh_grp, - const struct in6_addr *gw, int ifindex) + const struct in6_addr *gw, int ifindex, + int weight) { int i; @@ -2638,7 +2665,7 @@ mlxsw_sp_nexthop6_group_has_nexthop(const struct mlxsw_sp_nexthop_group *nh_grp, const struct mlxsw_sp_nexthop *nh; nh = &nh_grp->nexthops[i]; - if (nh->ifindex == ifindex && + if (nh->ifindex == ifindex && nh->nh_weight == weight && ipv6_addr_equal(gw, (struct in6_addr *) nh->gw_addr)) return true; } @@ -2657,11 +2684,13 @@ mlxsw_sp_nexthop6_group_cmp(const struct mlxsw_sp_nexthop_group *nh_grp, list_for_each_entry(mlxsw_sp_rt6, &fib6_entry->rt6_list, list) { struct in6_addr *gw; - int ifindex; + int ifindex, weight; ifindex = mlxsw_sp_rt6->rt->dst.dev->ifindex; + weight = mlxsw_sp_rt6->rt->rt6i_nh_weight; gw = &mlxsw_sp_rt6->rt->rt6i_gateway; - if (!mlxsw_sp_nexthop6_group_has_nexthop(nh_grp, gw, ifindex)) + if (!mlxsw_sp_nexthop6_group_has_nexthop(nh_grp, gw, ifindex, + weight)) return false; } @@ -4192,68 +4221,66 @@ mlxsw_sp_fib_node_entry_is_first(const struct mlxsw_sp_fib_node *fib_node, } static int mlxsw_sp_fib_lpm_tree_link(struct mlxsw_sp *mlxsw_sp, - struct mlxsw_sp_fib *fib, struct mlxsw_sp_fib_node *fib_node) { - struct mlxsw_sp_prefix_usage req_prefix_usage = {{ 0 } }; + struct mlxsw_sp_prefix_usage req_prefix_usage; + struct mlxsw_sp_fib *fib = fib_node->fib; struct mlxsw_sp_lpm_tree *lpm_tree; int err; - /* Since the tree is shared between all virtual routers we must - * make sure it contains all the required prefix lengths. This - * can be computed by either adding the new prefix length to the - * existing prefix usage of a bound tree, or by aggregating the - * prefix lengths across all virtual routers and adding the new - * one as well. - */ - if (fib->lpm_tree) - mlxsw_sp_prefix_usage_cpy(&req_prefix_usage, - &fib->lpm_tree->prefix_usage); - else - mlxsw_sp_vrs_prefixes(mlxsw_sp, fib->proto, &req_prefix_usage); - mlxsw_sp_prefix_usage_set(&req_prefix_usage, fib_node->key.prefix_len); + lpm_tree = mlxsw_sp->router->lpm.proto_trees[fib->proto]; + if (lpm_tree->prefix_ref_count[fib_node->key.prefix_len] != 0) + goto out; + mlxsw_sp_prefix_usage_cpy(&req_prefix_usage, &lpm_tree->prefix_usage); + mlxsw_sp_prefix_usage_set(&req_prefix_usage, fib_node->key.prefix_len); lpm_tree = mlxsw_sp_lpm_tree_get(mlxsw_sp, &req_prefix_usage, fib->proto); if (IS_ERR(lpm_tree)) return PTR_ERR(lpm_tree); - if (fib->lpm_tree && fib->lpm_tree->id == lpm_tree->id) - return 0; - err = mlxsw_sp_vrs_lpm_tree_replace(mlxsw_sp, fib, lpm_tree); if (err) - return err; + goto err_lpm_tree_replace; +out: + lpm_tree->prefix_ref_count[fib_node->key.prefix_len]++; return 0; -} -static void mlxsw_sp_fib_lpm_tree_unlink(struct mlxsw_sp *mlxsw_sp, - struct mlxsw_sp_fib *fib) -{ - if (!mlxsw_sp_prefix_usage_none(&fib->prefix_usage)) - return; - mlxsw_sp_vr_lpm_tree_unbind(mlxsw_sp, fib); - mlxsw_sp_lpm_tree_put(mlxsw_sp, fib->lpm_tree); - fib->lpm_tree = NULL; +err_lpm_tree_replace: + mlxsw_sp_lpm_tree_put(mlxsw_sp, lpm_tree); + return err; } -static void mlxsw_sp_fib_node_prefix_inc(struct mlxsw_sp_fib_node *fib_node) +static void mlxsw_sp_fib_lpm_tree_unlink(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_node *fib_node) { - unsigned char prefix_len = fib_node->key.prefix_len; + struct mlxsw_sp_lpm_tree *lpm_tree = fib_node->fib->lpm_tree; + struct mlxsw_sp_prefix_usage req_prefix_usage; struct mlxsw_sp_fib *fib = fib_node->fib; + int err; - if (fib->prefix_ref_count[prefix_len]++ == 0) - mlxsw_sp_prefix_usage_set(&fib->prefix_usage, prefix_len); -} + if (--lpm_tree->prefix_ref_count[fib_node->key.prefix_len] != 0) + return; + /* Try to construct a new LPM tree from the current prefix usage + * minus the unused one. If we fail, continue using the old one. + */ + mlxsw_sp_prefix_usage_cpy(&req_prefix_usage, &lpm_tree->prefix_usage); + mlxsw_sp_prefix_usage_clear(&req_prefix_usage, + fib_node->key.prefix_len); + lpm_tree = mlxsw_sp_lpm_tree_get(mlxsw_sp, &req_prefix_usage, + fib->proto); + if (IS_ERR(lpm_tree)) + return; -static void mlxsw_sp_fib_node_prefix_dec(struct mlxsw_sp_fib_node *fib_node) -{ - unsigned char prefix_len = fib_node->key.prefix_len; - struct mlxsw_sp_fib *fib = fib_node->fib; + err = mlxsw_sp_vrs_lpm_tree_replace(mlxsw_sp, fib, lpm_tree); + if (err) + goto err_lpm_tree_replace; - if (--fib->prefix_ref_count[prefix_len] == 0) - mlxsw_sp_prefix_usage_clear(&fib->prefix_usage, prefix_len); + return; + +err_lpm_tree_replace: + mlxsw_sp_lpm_tree_put(mlxsw_sp, lpm_tree); } static int mlxsw_sp_fib_node_init(struct mlxsw_sp *mlxsw_sp, @@ -4267,12 +4294,10 @@ static int mlxsw_sp_fib_node_init(struct mlxsw_sp *mlxsw_sp, return err; fib_node->fib = fib; - err = mlxsw_sp_fib_lpm_tree_link(mlxsw_sp, fib, fib_node); + err = mlxsw_sp_fib_lpm_tree_link(mlxsw_sp, fib_node); if (err) goto err_fib_lpm_tree_link; - mlxsw_sp_fib_node_prefix_inc(fib_node); - return 0; err_fib_lpm_tree_link: @@ -4286,8 +4311,7 @@ static void mlxsw_sp_fib_node_fini(struct mlxsw_sp *mlxsw_sp, { struct mlxsw_sp_fib *fib = fib_node->fib; - mlxsw_sp_fib_node_prefix_dec(fib_node); - mlxsw_sp_fib_lpm_tree_unlink(mlxsw_sp, fib); + mlxsw_sp_fib_lpm_tree_unlink(mlxsw_sp, fib_node); fib_node->fib = NULL; mlxsw_sp_fib_node_remove(fib, fib_node); } @@ -4326,7 +4350,7 @@ mlxsw_sp_fib_node_get(struct mlxsw_sp *mlxsw_sp, u32 tb_id, const void *addr, err_fib_node_init: mlxsw_sp_fib_node_destroy(fib_node); err_fib_node_create: - mlxsw_sp_vr_put(vr); + mlxsw_sp_vr_put(mlxsw_sp, vr); return ERR_PTR(err); } @@ -4339,7 +4363,7 @@ static void mlxsw_sp_fib_node_put(struct mlxsw_sp *mlxsw_sp, return; mlxsw_sp_fib_node_fini(mlxsw_sp, fib_node); mlxsw_sp_fib_node_destroy(fib_node); - mlxsw_sp_vr_put(vr); + mlxsw_sp_vr_put(mlxsw_sp, vr); } static struct mlxsw_sp_fib4_entry * @@ -4764,7 +4788,7 @@ static int mlxsw_sp_nexthop6_init(struct mlxsw_sp *mlxsw_sp, struct net_device *dev = rt->dst.dev; nh->nh_grp = nh_grp; - nh->nh_weight = 1; + nh->nh_weight = rt->rt6i_nh_weight; memcpy(&nh->gw_addr, &rt->rt6i_gateway, sizeof(nh->gw_addr)); mlxsw_sp_nexthop_counter_alloc(mlxsw_sp, nh); @@ -5362,7 +5386,7 @@ static void mlxsw_sp_router_fibmr_del(struct mlxsw_sp *mlxsw_sp, return; mlxsw_sp_mr_route4_del(vr->mr4_table, men_info->mfc); - mlxsw_sp_vr_put(vr); + mlxsw_sp_vr_put(mlxsw_sp, vr); } static int @@ -5399,7 +5423,7 @@ mlxsw_sp_router_fibmr_vif_del(struct mlxsw_sp *mlxsw_sp, return; mlxsw_sp_mr_vif_del(vr->mr4_table, ven_info->vif_index); - mlxsw_sp_vr_put(vr); + mlxsw_sp_vr_put(mlxsw_sp, vr); } static int mlxsw_sp_router_set_abort_trap(struct mlxsw_sp *mlxsw_sp) @@ -6048,7 +6072,7 @@ err_fid_get: err_rif_alloc: err_rif_index_alloc: vr->rif_count--; - mlxsw_sp_vr_put(vr); + mlxsw_sp_vr_put(mlxsw_sp, vr); return ERR_PTR(err); } @@ -6071,7 +6095,7 @@ void mlxsw_sp_rif_destroy(struct mlxsw_sp_rif *rif) mlxsw_sp_fid_put(fid); kfree(rif); vr->rif_count--; - mlxsw_sp_vr_put(vr); + mlxsw_sp_vr_put(mlxsw_sp, vr); } static void @@ -6861,7 +6885,7 @@ mlxsw_sp_rif_ipip_lb_configure(struct mlxsw_sp_rif *rif) return 0; err_loopback_op: - mlxsw_sp_vr_put(ul_vr); + mlxsw_sp_vr_put(mlxsw_sp, ul_vr); return err; } @@ -6875,7 +6899,7 @@ static void mlxsw_sp_rif_ipip_lb_deconfigure(struct mlxsw_sp_rif *rif) mlxsw_sp_rif_ipip_lb_op(lb_rif, ul_vr, false); --ul_vr->rif_count; - mlxsw_sp_vr_put(ul_vr); + mlxsw_sp_vr_put(mlxsw_sp, ul_vr); } static const struct mlxsw_sp_rif_ops mlxsw_sp_rif_ipip_lb_ops = { @@ -7010,6 +7034,24 @@ static int mlxsw_sp_mp_hash_init(struct mlxsw_sp *mlxsw_sp) } #endif +static int mlxsw_sp_dscp_init(struct mlxsw_sp *mlxsw_sp) +{ + char rdpm_pl[MLXSW_REG_RDPM_LEN]; + unsigned int i; + + MLXSW_REG_ZERO(rdpm, rdpm_pl); + + /* HW is determining switch priority based on DSCP-bits, but the + * kernel is still doing that based on the ToS. Since there's a + * mismatch in bits we need to make sure to translate the right + * value ToS would observe, skipping the 2 least-significant ECN bits. + */ + for (i = 0; i < MLXSW_REG_RDPM_DSCP_ENTRY_REC_MAX_COUNT; i++) + mlxsw_reg_rdpm_pack(rdpm_pl, i, rt_tos2priority(i << 2)); + + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(rdpm), rdpm_pl); +} + static int __mlxsw_sp_router_init(struct mlxsw_sp *mlxsw_sp) { char rgcr_pl[MLXSW_REG_RGCR_LEN]; @@ -7022,6 +7064,7 @@ static int __mlxsw_sp_router_init(struct mlxsw_sp *mlxsw_sp) mlxsw_reg_rgcr_pack(rgcr_pl, true, true); mlxsw_reg_rgcr_max_router_interfaces_set(rgcr_pl, max_rifs); + mlxsw_reg_rgcr_usp_set(rgcr_pl, true); err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(rgcr), rgcr_pl); if (err) return err; @@ -7097,6 +7140,10 @@ int mlxsw_sp_router_init(struct mlxsw_sp *mlxsw_sp) if (err) goto err_mp_hash_init; + err = mlxsw_sp_dscp_init(mlxsw_sp); + if (err) + goto err_dscp_init; + mlxsw_sp->router->fib_nb.notifier_call = mlxsw_sp_router_fib_event; err = register_fib_notifier(&mlxsw_sp->router->fib_nb, mlxsw_sp_router_fib_dump_flush); @@ -7106,6 +7153,7 @@ int mlxsw_sp_router_init(struct mlxsw_sp *mlxsw_sp) return 0; err_register_fib_notifier: +err_dscp_init: err_mp_hash_init: unregister_netevent_notifier(&mlxsw_sp->router->netevent_nb); err_register_netevent_notifier: |