diff options
Diffstat (limited to 'drivers/net/ethernet/mellanox/mlx5')
82 files changed, 6659 insertions, 1738 deletions
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/Kconfig b/drivers/net/ethernet/mellanox/mlx5/core/Kconfig index bfc0cd5ec423..26685fd0fdaa 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/Kconfig +++ b/drivers/net/ethernet/mellanox/mlx5/core/Kconfig @@ -139,6 +139,14 @@ config MLX5_CORE_IPOIB help MLX5 IPoIB offloads & acceleration support. +config MLX5_EN_MACSEC + bool "Connect-X support for MACSec offload" + depends on MLX5_CORE_EN + depends on MACSEC + default n + help + Build support for MACsec cryptography-offload acceleration in the NIC. + config MLX5_EN_IPSEC bool "Mellanox Technologies IPsec Connect-X support" depends on MLX5_CORE_EN diff --git a/drivers/net/ethernet/mellanox/mlx5/core/Makefile b/drivers/net/ethernet/mellanox/mlx5/core/Makefile index a3773a8177ed..a22c32aabf11 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/Makefile +++ b/drivers/net/ethernet/mellanox/mlx5/core/Makefile @@ -92,6 +92,9 @@ mlx5_core-$(CONFIG_MLX5_CORE_IPOIB) += ipoib/ipoib.o ipoib/ethtool.o ipoib/ipoib # mlx5_core-$(CONFIG_MLX5_FPGA) += fpga/cmd.o fpga/core.o fpga/conn.o fpga/sdk.o +mlx5_core-$(CONFIG_MLX5_EN_MACSEC) += en_accel/macsec.o en_accel/macsec_fs.o \ + en_accel/macsec_stats.o + mlx5_core-$(CONFIG_MLX5_EN_IPSEC) += en_accel/ipsec.o en_accel/ipsec_rxtx.o \ en_accel/ipsec_stats.o en_accel/ipsec_fs.o \ en_accel/ipsec_offload.o diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en.h b/drivers/net/ethernet/mellanox/mlx5/core/en.h index a560df446bac..26a23047f1f3 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en.h @@ -93,29 +93,26 @@ struct page_pool; #define MLX5_MPWRQ_DEF_LOG_STRIDE_SZ(mdev) \ MLX5_MPWRQ_LOG_STRIDE_SZ(mdev, order_base_2(MLX5E_RX_MAX_HEAD)) -#define MLX5_MPWRQ_LOG_WQE_SZ 18 -#define MLX5_MPWRQ_WQE_PAGE_ORDER (MLX5_MPWRQ_LOG_WQE_SZ - PAGE_SHIFT > 0 ? \ - MLX5_MPWRQ_LOG_WQE_SZ - PAGE_SHIFT : 0) -#define MLX5_MPWRQ_PAGES_PER_WQE BIT(MLX5_MPWRQ_WQE_PAGE_ORDER) - -#define MLX5_ALIGN_MTTS(mtts) (ALIGN(mtts, 8)) -#define MLX5_ALIGNED_MTTS_OCTW(mtts) ((mtts) / 2) -#define MLX5_MTT_OCTW(mtts) (MLX5_ALIGNED_MTTS_OCTW(MLX5_ALIGN_MTTS(mtts))) -/* Add another page to MLX5E_REQUIRED_WQE_MTTS as a buffer between - * WQEs, This page will absorb write overflow by the hardware, when - * receiving packets larger than MTU. These oversize packets are - * dropped by the driver at a later stage. +#define MLX5_MPWRQ_MAX_LOG_WQE_SZ 18 + +/* Keep in sync with mlx5e_mpwrq_log_wqe_sz. + * These are theoretical maximums, which can be further restricted by + * capabilities. These values are used for static resource allocations and + * sanity checks. + * MLX5_SEND_WQE_MAX_SIZE is a bit bigger than the maximum cacheline-aligned WQE + * size actually used at runtime, but it's not a problem when calculating static + * array sizes. */ -#define MLX5E_REQUIRED_WQE_MTTS (MLX5_ALIGN_MTTS(MLX5_MPWRQ_PAGES_PER_WQE + 1)) -#define MLX5E_REQUIRED_MTTS(wqes) (wqes * MLX5E_REQUIRED_WQE_MTTS) +#define MLX5_UMR_MAX_MTT_SPACE \ + (ALIGN_DOWN(MLX5_SEND_WQE_MAX_SIZE - sizeof(struct mlx5e_umr_wqe), \ + MLX5_UMR_MTT_ALIGNMENT)) +#define MLX5_MPWRQ_MAX_PAGES_PER_WQE \ + rounddown_pow_of_two(MLX5_UMR_MAX_MTT_SPACE / sizeof(struct mlx5_mtt)) + #define MLX5E_MAX_RQ_NUM_MTTS \ - (ALIGN_DOWN(U16_MAX, 4) * 2) /* So that MLX5_MTT_OCTW(num_mtts) fits into u16 */ + (ALIGN_DOWN(U16_MAX, 4) * 2) /* Fits into u16 and aligned by WQEBB. */ +#define MLX5E_MAX_RQ_NUM_KSMS (U16_MAX - 1) /* So that num_ksms fits into u16. */ #define MLX5E_ORDER2_MAX_PACKET_MTU (order_base_2(10 * 1024)) -#define MLX5E_PARAMS_MAXIMUM_LOG_RQ_SIZE_MPW \ - (ilog2(MLX5E_MAX_RQ_NUM_MTTS / MLX5E_REQUIRED_WQE_MTTS)) -#define MLX5E_LOG_MAX_RQ_NUM_PACKETS_MPW \ - (MLX5E_PARAMS_MAXIMUM_LOG_RQ_SIZE_MPW + \ - (MLX5_MPWRQ_LOG_WQE_SZ - MLX5E_ORDER2_MAX_PACKET_MTU)) #define MLX5E_MIN_SKB_FRAG_SZ (MLX5_SKB_FRAG_SZ(MLX5_RX_HEADROOM)) #define MLX5E_LOG_MAX_RX_WQE_BULK \ @@ -127,8 +124,7 @@ struct page_pool; #define MLX5E_PARAMS_MINIMUM_LOG_RQ_SIZE (1 + MLX5E_LOG_MAX_RX_WQE_BULK) #define MLX5E_PARAMS_DEFAULT_LOG_RQ_SIZE 0xa -#define MLX5E_PARAMS_MAXIMUM_LOG_RQ_SIZE min_t(u8, 0xd, \ - MLX5E_LOG_MAX_RQ_NUM_PACKETS_MPW) +#define MLX5E_PARAMS_MAXIMUM_LOG_RQ_SIZE 0xd #define MLX5E_PARAMS_MINIMUM_LOG_RQ_SIZE_MPW 0x2 @@ -150,13 +146,6 @@ struct page_pool; #define MLX5E_TX_XSK_POLL_BUDGET 64 #define MLX5E_SQ_RECOVER_MIN_INTERVAL 500 /* msecs */ -#define MLX5E_UMR_WQE_INLINE_SZ \ - (sizeof(struct mlx5e_umr_wqe) + \ - ALIGN(MLX5_MPWRQ_PAGES_PER_WQE * sizeof(struct mlx5_mtt), \ - MLX5_UMR_MTT_ALIGNMENT)) -#define MLX5E_UMR_WQEBBS \ - (DIV_ROUND_UP(MLX5E_UMR_WQE_INLINE_SZ, MLX5_SEND_WQE_BB)) - #define MLX5E_KLM_UMR_WQE_SZ(sgl_len)\ (sizeof(struct mlx5e_umr_wqe) +\ (sizeof(struct mlx5_klm) * (sgl_len))) @@ -174,8 +163,7 @@ struct page_pool; ALIGN_DOWN(MLX5E_KLM_MAX_ENTRIES_PER_WQE(wqe_size), MLX5_UMR_KLM_ALIGNMENT) #define MLX5E_MAX_KLM_PER_WQE(mdev) \ - MLX5E_KLM_ENTRIES_PER_WQE(MLX5_SEND_WQE_BB * \ - mlx5e_get_sw_max_sq_mpw_wqebbs(mlx5e_get_max_sq_wqebbs(mdev))) + MLX5E_KLM_ENTRIES_PER_WQE(MLX5_SEND_WQE_BB * mlx5e_get_max_sq_aligned_wqebbs(mdev)) #define MLX5E_MSG_LEVEL NETIF_MSG_LINK @@ -189,12 +177,6 @@ do { \ #define mlx5e_state_dereference(priv, p) \ rcu_dereference_protected((p), lockdep_is_held(&(priv)->state_lock)) -enum mlx5e_rq_group { - MLX5E_RQ_GROUP_REGULAR, - MLX5E_RQ_GROUP_XSK, -#define MLX5E_NUM_RQ_GROUPS(g) (1 + MLX5E_RQ_GROUP_##g) -}; - static inline u8 mlx5e_get_num_lag_ports(struct mlx5_core_dev *mdev) { if (mlx5_lag_is_lacp_owner(mdev)) @@ -227,13 +209,15 @@ static inline int mlx5e_get_max_num_channels(struct mlx5_core_dev *mdev) * bytes units. Driver hardens the limitation to 1KB (16 * WQEBBs), unless firmware capability is stricter. */ -static inline u16 mlx5e_get_max_sq_wqebbs(struct mlx5_core_dev *mdev) +static inline u8 mlx5e_get_max_sq_wqebbs(struct mlx5_core_dev *mdev) { - return min_t(u16, MLX5_SEND_WQE_MAX_WQEBBS, - MLX5_CAP_GEN(mdev, max_wqe_sz_sq) / MLX5_SEND_WQE_BB); + BUILD_BUG_ON(MLX5_SEND_WQE_MAX_WQEBBS > U8_MAX); + + return (u8)min_t(u16, MLX5_SEND_WQE_MAX_WQEBBS, + MLX5_CAP_GEN(mdev, max_wqe_sz_sq) / MLX5_SEND_WQE_BB); } -static inline u8 mlx5e_get_sw_max_sq_mpw_wqebbs(u8 max_sq_wqebbs) +static inline u8 mlx5e_get_max_sq_aligned_wqebbs(struct mlx5_core_dev *mdev) { /* The return value will be multiplied by MLX5_SEND_WQEBB_NUM_DS. * Since max_sq_wqebbs may be up to MLX5_SEND_WQE_MAX_WQEBBS == 16, @@ -242,8 +226,9 @@ static inline u8 mlx5e_get_sw_max_sq_mpw_wqebbs(u8 max_sq_wqebbs) * than MLX5_SEND_WQE_MAX_WQEBBS to let a full-session WQE be * cache-aligned. */ - u8 wqebbs = min_t(u8, max_sq_wqebbs, MLX5_SEND_WQE_MAX_WQEBBS - 1); + u8 wqebbs = mlx5e_get_max_sq_wqebbs(mdev); + wqebbs = min_t(u8, wqebbs, MLX5_SEND_WQE_MAX_WQEBBS - 1); #if L1_CACHE_BYTES >= 128 wqebbs = ALIGN_DOWN(wqebbs, 2); #endif @@ -272,6 +257,7 @@ struct mlx5e_umr_wqe { union { DECLARE_FLEX_ARRAY(struct mlx5_mtt, inline_mtts); DECLARE_FLEX_ARRAY(struct mlx5_klm, inline_klms); + DECLARE_FLEX_ARRAY(struct mlx5_ksm, inline_ksms); }; }; @@ -476,15 +462,11 @@ struct mlx5e_txqsq { struct work_struct recover_work; struct mlx5e_ptpsq *ptpsq; cqe_ts_to_ns ptp_cyc2time; - u16 max_sq_wqebbs; } ____cacheline_aligned_in_smp; -struct mlx5e_dma_info { - dma_addr_t addr; - union { - struct page *page; - struct xdp_buff *xsk; - }; +union mlx5e_alloc_unit { + struct page *page; + struct xdp_buff *xsk; }; /* XDP packets can be transmitted in different ways. On completion, we need to @@ -580,7 +562,6 @@ struct mlx5e_xdpsq { /* control path */ struct mlx5_wq_ctrl wq_ctrl; struct mlx5e_channel *channel; - u16 max_sq_wqebbs; } ____cacheline_aligned_in_smp; struct mlx5e_ktls_resync_resp; @@ -609,25 +590,20 @@ struct mlx5e_icosq { /* control path */ struct mlx5_wq_ctrl wq_ctrl; struct mlx5e_channel *channel; - u16 max_sq_wqebbs; struct work_struct recover_work; } ____cacheline_aligned_in_smp; struct mlx5e_wqe_frag_info { - struct mlx5e_dma_info *di; + union mlx5e_alloc_unit *au; u32 offset; bool last_in_page; }; -struct mlx5e_umr_dma_info { - struct mlx5e_dma_info dma_info[MLX5_MPWRQ_PAGES_PER_WQE]; -}; - struct mlx5e_mpw_info { - struct mlx5e_umr_dma_info umr; u16 consumed_strides; - DECLARE_BITMAP(xdp_xmit_bitmap, MLX5_MPWRQ_PAGES_PER_WQE); + DECLARE_BITMAP(xdp_xmit_bitmap, MLX5_MPWRQ_MAX_PAGES_PER_WQE); + union mlx5e_alloc_unit alloc_units[]; }; #define MLX5E_MAX_RX_FRAGS 4 @@ -635,13 +611,13 @@ struct mlx5e_mpw_info { /* a single cache unit is capable to serve one napi call (for non-striding rq) * or a MPWQE (for striding rq). */ -#define MLX5E_CACHE_UNIT (MLX5_MPWRQ_PAGES_PER_WQE > NAPI_POLL_WEIGHT ? \ - MLX5_MPWRQ_PAGES_PER_WQE : NAPI_POLL_WEIGHT) +#define MLX5E_CACHE_UNIT (MLX5_MPWRQ_MAX_PAGES_PER_WQE > NAPI_POLL_WEIGHT ? \ + MLX5_MPWRQ_MAX_PAGES_PER_WQE : NAPI_POLL_WEIGHT) #define MLX5E_CACHE_SIZE (4 * roundup_pow_of_two(MLX5E_CACHE_UNIT)) struct mlx5e_page_cache { u32 head; u32 tail; - struct mlx5e_dma_info page_cache[MLX5E_CACHE_SIZE]; + struct page *page_cache[MLX5E_CACHE_SIZE]; }; struct mlx5e_rq; @@ -674,6 +650,12 @@ struct mlx5e_rq_frags_info { u8 num_frags; u8 log_num_frags; u8 wqe_bulk; + u8 wqe_index_mask; +}; + +struct mlx5e_dma_info { + dma_addr_t addr; + struct page *page; }; struct mlx5e_shampo_hd { @@ -695,13 +677,20 @@ struct mlx5e_hw_gro_data { int second_ip_id; }; +enum mlx5e_mpwrq_umr_mode { + MLX5E_MPWRQ_UMR_MODE_ALIGNED, + MLX5E_MPWRQ_UMR_MODE_UNALIGNED, + MLX5E_MPWRQ_UMR_MODE_OVERSIZED, + MLX5E_MPWRQ_UMR_MODE_TRIPLE, +}; + struct mlx5e_rq { /* data path */ union { struct { struct mlx5_wq_cyc wq; struct mlx5e_wqe_frag_info *frags; - struct mlx5e_dma_info *di; + union mlx5e_alloc_unit *alloc_units; struct mlx5e_rq_frags_info info; mlx5e_fp_skb_from_cqe skb_from_cqe; } wqe; @@ -710,6 +699,7 @@ struct mlx5e_rq { struct mlx5e_umr_wqe umr_wqe; struct mlx5e_mpw_info *info; mlx5e_fp_skb_from_cqe_mpwrq skb_from_cqe_mpwrq; + __be32 umr_mkey_be; u16 num_strides; u16 actual_wq_head; u8 log_stride_sz; @@ -717,6 +707,11 @@ struct mlx5e_rq { u8 umr_last_bulk; u8 umr_completed; u8 min_wqe_bulk; + u8 page_shift; + u8 pages_per_wqe; + u8 umr_wqebbs; + u8 mtts_per_wqe; + u8 umr_mode; struct mlx5e_shampo_hd *shampo; } mpwqe; }; @@ -767,7 +762,6 @@ struct mlx5e_rq { u32 rqn; struct mlx5_core_dev *mdev; struct mlx5e_channel *channel; - u32 umr_mkey; struct mlx5e_dma_info wqe_overflow; /* XDP read-mostly */ @@ -856,11 +850,6 @@ enum { MLX5E_STATE_XDP_ACTIVE, }; -enum { - MLX5E_TC_PRIO = 0, - MLX5E_NIC_PRIO -}; - struct mlx5e_modify_sq_param { int curr_state; int next_state; @@ -959,6 +948,9 @@ struct mlx5e_priv { const struct mlx5e_profile *profile; void *ppriv; +#ifdef CONFIG_MLX5_EN_MACSEC + struct mlx5e_macsec *macsec; +#endif #ifdef CONFIG_MLX5_EN_IPSEC struct mlx5e_ipsec *ipsec; #endif @@ -1010,7 +1002,6 @@ struct mlx5e_profile { mlx5e_stats_grp_t *stats_grps; const struct mlx5e_rx_handlers *rx_handlers; int max_tc; - u8 rq_groups; u32 features; }; @@ -1019,7 +1010,8 @@ struct mlx5e_profile { void mlx5e_build_ptys2ethtool_map(void); -bool mlx5e_check_fragmented_striding_rq_cap(struct mlx5_core_dev *mdev); +bool mlx5e_check_fragmented_striding_rq_cap(struct mlx5_core_dev *mdev, u8 page_shift, + enum mlx5e_mpwrq_umr_mode umr_mode); void mlx5e_shampo_dealloc_hd(struct mlx5e_rq *rq, u16 len, u16 start, bool close); void mlx5e_get_stats(struct net_device *dev, struct rtnl_link_stats64 *stats); @@ -1047,6 +1039,7 @@ struct mlx5e_rq_param; int mlx5e_open_rq(struct mlx5e_params *params, struct mlx5e_rq_param *param, struct mlx5e_xsk_param *xsk, int node, struct mlx5e_rq *rq); +#define MLX5E_RQ_WQES_TIMEOUT 20000 /* msecs */ int mlx5e_wait_for_min_rx_wqes(struct mlx5e_rq *rq, int wait_time); void mlx5e_close_rq(struct mlx5e_rq *rq); int mlx5e_create_rq(struct mlx5e_rq *rq, struct mlx5e_rq_param *param); @@ -1101,7 +1094,7 @@ void mlx5e_activate_priv_channels(struct mlx5e_priv *priv); void mlx5e_deactivate_priv_channels(struct mlx5e_priv *priv); int mlx5e_ptp_rx_manage_fs_ctx(struct mlx5e_priv *priv, void *ctx); -int mlx5e_modify_rq_state(struct mlx5e_rq *rq, int curr_state, int next_state); +int mlx5e_flush_rq(struct mlx5e_rq *rq, int curr_state); void mlx5e_activate_rq(struct mlx5e_rq *rq); void mlx5e_deactivate_rq(struct mlx5e_rq *rq); void mlx5e_activate_icosq(struct mlx5e_icosq *icosq); @@ -1136,6 +1129,7 @@ static inline bool mlx5_tx_swp_supported(struct mlx5_core_dev *mdev) extern const struct ethtool_ops mlx5e_ethtool_ops; +int mlx5e_create_mkey(struct mlx5_core_dev *mdev, u32 pdn, u32 *mkey); int mlx5e_create_mdev_resources(struct mlx5_core_dev *mdev); void mlx5e_destroy_mdev_resources(struct mlx5_core_dev *mdev); int mlx5e_refresh_tirs(struct mlx5e_priv *priv, bool enable_uc_lb, @@ -1148,8 +1142,6 @@ void mlx5e_destroy_q_counters(struct mlx5e_priv *priv); int mlx5e_open_drop_rq(struct mlx5e_priv *priv, struct mlx5e_rq *drop_rq); void mlx5e_close_drop_rq(struct mlx5e_rq *drop_rq); -int mlx5e_init_di_list(struct mlx5e_rq *rq, int wq_sz, int node); -void mlx5e_free_di_list(struct mlx5e_rq *rq); int mlx5e_create_tis(struct mlx5_core_dev *mdev, void *in, u32 *tisn); void mlx5e_destroy_tis(struct mlx5_core_dev *mdev, u32 tisn); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/channels.c b/drivers/net/ethernet/mellanox/mlx5/core/en/channels.c index e7c14c0de0a7..48581ea3adcb 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/channels.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/channels.c @@ -10,28 +10,33 @@ unsigned int mlx5e_channels_get_num(struct mlx5e_channels *chs) return chs->num; } -void mlx5e_channels_get_regular_rqn(struct mlx5e_channels *chs, unsigned int ix, u32 *rqn) +static struct mlx5e_channel *mlx5e_channels_get(struct mlx5e_channels *chs, unsigned int ix) { - struct mlx5e_channel *c; + WARN_ON_ONCE(ix >= mlx5e_channels_get_num(chs)); + return chs->c[ix]; +} - WARN_ON(ix >= mlx5e_channels_get_num(chs)); - c = chs->c[ix]; +bool mlx5e_channels_is_xsk(struct mlx5e_channels *chs, unsigned int ix) +{ + struct mlx5e_channel *c = mlx5e_channels_get(chs, ix); - *rqn = c->rq.rqn; + return test_bit(MLX5E_CHANNEL_STATE_XSK, c->state); } -bool mlx5e_channels_get_xsk_rqn(struct mlx5e_channels *chs, unsigned int ix, u32 *rqn) +void mlx5e_channels_get_regular_rqn(struct mlx5e_channels *chs, unsigned int ix, u32 *rqn) { - struct mlx5e_channel *c; + struct mlx5e_channel *c = mlx5e_channels_get(chs, ix); - WARN_ON(ix >= mlx5e_channels_get_num(chs)); - c = chs->c[ix]; + *rqn = c->rq.rqn; +} - if (!test_bit(MLX5E_CHANNEL_STATE_XSK, c->state)) - return false; +void mlx5e_channels_get_xsk_rqn(struct mlx5e_channels *chs, unsigned int ix, u32 *rqn) +{ + struct mlx5e_channel *c = mlx5e_channels_get(chs, ix); + + WARN_ON_ONCE(!test_bit(MLX5E_CHANNEL_STATE_XSK, c->state)); *rqn = c->xskrq.rqn; - return true; } bool mlx5e_channels_get_ptp_rqn(struct mlx5e_channels *chs, u32 *rqn) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/channels.h b/drivers/net/ethernet/mellanox/mlx5/core/en/channels.h index ca00cbc827cb..637ca90daaa8 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/channels.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/channels.h @@ -9,8 +9,9 @@ struct mlx5e_channels; unsigned int mlx5e_channels_get_num(struct mlx5e_channels *chs); +bool mlx5e_channels_is_xsk(struct mlx5e_channels *chs, unsigned int ix); void mlx5e_channels_get_regular_rqn(struct mlx5e_channels *chs, unsigned int ix, u32 *rqn); -bool mlx5e_channels_get_xsk_rqn(struct mlx5e_channels *chs, unsigned int ix, u32 *rqn); +void mlx5e_channels_get_xsk_rqn(struct mlx5e_channels *chs, unsigned int ix, u32 *rqn); bool mlx5e_channels_get_ptp_rqn(struct mlx5e_channels *chs, u32 *rqn); #endif /* __MLX5_EN_CHANNELS_H__ */ diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/fs.h b/drivers/net/ethernet/mellanox/mlx5/core/en/fs.h index 9b8cdf2e68ad..bf2741eb7f9b 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/fs.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/fs.h @@ -8,6 +8,7 @@ #include "lib/fs_ttc.h" struct mlx5e_post_act; +struct mlx5e_tc_table; enum { MLX5E_TC_FT_LEVEL = 0, @@ -15,6 +16,11 @@ enum { MLX5E_TC_MISS_LEVEL, }; +enum { + MLX5E_TC_PRIO = 0, + MLX5E_NIC_PRIO +}; + struct mlx5e_flow_table { int num_groups; struct mlx5_flow_table *t; @@ -83,54 +89,28 @@ enum { #endif }; -struct mlx5e_priv; - -#ifdef CONFIG_MLX5_EN_RXNFC - -struct mlx5e_ethtool_table { - struct mlx5_flow_table *ft; - int num_rules; -}; - -#define ETHTOOL_NUM_L3_L4_FTS 7 -#define ETHTOOL_NUM_L2_FTS 4 - -struct mlx5e_ethtool_steering { - struct mlx5e_ethtool_table l3_l4_ft[ETHTOOL_NUM_L3_L4_FTS]; - struct mlx5e_ethtool_table l2_ft[ETHTOOL_NUM_L2_FTS]; - struct list_head rules; - int tot_num_rules; -}; - -void mlx5e_ethtool_init_steering(struct mlx5e_priv *priv); -void mlx5e_ethtool_cleanup_steering(struct mlx5e_priv *priv); -int mlx5e_ethtool_set_rxnfc(struct mlx5e_priv *priv, struct ethtool_rxnfc *cmd); -int mlx5e_ethtool_get_rxnfc(struct mlx5e_priv *priv, - struct ethtool_rxnfc *info, u32 *rule_locs); -#else -static inline void mlx5e_ethtool_init_steering(struct mlx5e_priv *priv) { } -static inline void mlx5e_ethtool_cleanup_steering(struct mlx5e_priv *priv) { } -static inline int mlx5e_ethtool_set_rxnfc(struct mlx5e_priv *priv, struct ethtool_rxnfc *cmd) -{ return -EOPNOTSUPP; } -static inline int mlx5e_ethtool_get_rxnfc(struct mlx5e_priv *priv, - struct ethtool_rxnfc *info, u32 *rule_locs) -{ return -EOPNOTSUPP; } -#endif /* CONFIG_MLX5_EN_RXNFC */ +struct mlx5e_flow_steering; +struct mlx5e_rx_res; #ifdef CONFIG_MLX5_EN_ARFS struct mlx5e_arfs_tables; -int mlx5e_arfs_create_tables(struct mlx5e_priv *priv); -void mlx5e_arfs_destroy_tables(struct mlx5e_priv *priv); -int mlx5e_arfs_enable(struct mlx5e_priv *priv); -int mlx5e_arfs_disable(struct mlx5e_priv *priv); +int mlx5e_arfs_create_tables(struct mlx5e_flow_steering *fs, + struct mlx5e_rx_res *rx_res, bool ntuple); +void mlx5e_arfs_destroy_tables(struct mlx5e_flow_steering *fs, bool ntuple); +int mlx5e_arfs_enable(struct mlx5e_flow_steering *fs); +int mlx5e_arfs_disable(struct mlx5e_flow_steering *fs); int mlx5e_rx_flow_steer(struct net_device *dev, const struct sk_buff *skb, u16 rxq_index, u32 flow_id); #else -static inline int mlx5e_arfs_create_tables(struct mlx5e_priv *priv) { return 0; } -static inline void mlx5e_arfs_destroy_tables(struct mlx5e_priv *priv) {} -static inline int mlx5e_arfs_enable(struct mlx5e_priv *priv) { return -EOPNOTSUPP; } -static inline int mlx5e_arfs_disable(struct mlx5e_priv *priv) { return -EOPNOTSUPP; } +static inline int mlx5e_arfs_create_tables(struct mlx5e_flow_steering *fs, + struct mlx5e_rx_res *rx_res, bool ntuple) +{ return 0; } +static inline void mlx5e_arfs_destroy_tables(struct mlx5e_flow_steering *fs, bool ntuple) {} +static inline int mlx5e_arfs_enable(struct mlx5e_flow_steering *fs) +{ return -EOPNOTSUPP; } +static inline int mlx5e_arfs_disable(struct mlx5e_flow_steering *fs) +{ return -EOPNOTSUPP; } #endif #ifdef CONFIG_MLX5_EN_TLS @@ -142,54 +122,63 @@ struct mlx5e_fs_udp; struct mlx5e_fs_any; struct mlx5e_ptp_fs; -struct mlx5e_flow_steering { - bool state_destroy; - bool vlan_strip_disable; - struct mlx5_core_dev *mdev; - struct mlx5_flow_namespace *ns; -#ifdef CONFIG_MLX5_EN_RXNFC - struct mlx5e_ethtool_steering ethtool; -#endif - struct mlx5e_tc_table *tc; - struct mlx5e_promisc_table promisc; - struct mlx5e_vlan_table *vlan; - struct mlx5e_l2_table l2; - struct mlx5_ttc_table *ttc; - struct mlx5_ttc_table *inner_ttc; -#ifdef CONFIG_MLX5_EN_ARFS - struct mlx5e_arfs_tables *arfs; -#endif -#ifdef CONFIG_MLX5_EN_TLS - struct mlx5e_accel_fs_tcp *accel_tcp; -#endif - struct mlx5e_fs_udp *udp; - struct mlx5e_fs_any *any; - struct mlx5e_ptp_fs *ptp_fs; -}; - -void mlx5e_set_ttc_params(struct mlx5e_priv *priv, +void mlx5e_set_ttc_params(struct mlx5e_flow_steering *fs, + struct mlx5e_rx_res *rx_res, struct ttc_params *ttc_params, bool tunnel); -void mlx5e_destroy_ttc_table(struct mlx5e_priv *priv); -int mlx5e_create_ttc_table(struct mlx5e_priv *priv); +void mlx5e_destroy_ttc_table(struct mlx5e_flow_steering *fs); +int mlx5e_create_ttc_table(struct mlx5e_flow_steering *fs, + struct mlx5e_rx_res *rx_res); void mlx5e_destroy_flow_table(struct mlx5e_flow_table *ft); -void mlx5e_enable_cvlan_filter(struct mlx5e_priv *priv); -void mlx5e_disable_cvlan_filter(struct mlx5e_priv *priv); +void mlx5e_enable_cvlan_filter(struct mlx5e_flow_steering *fs, bool promisc); +void mlx5e_disable_cvlan_filter(struct mlx5e_flow_steering *fs, bool promisc); -int mlx5e_create_flow_steering(struct mlx5e_priv *priv); -void mlx5e_destroy_flow_steering(struct mlx5e_priv *priv); +int mlx5e_create_flow_steering(struct mlx5e_flow_steering *fs, + struct mlx5e_rx_res *rx_res, + const struct mlx5e_profile *profile, + struct net_device *netdev); +void mlx5e_destroy_flow_steering(struct mlx5e_flow_steering *fs, bool ntuple, + const struct mlx5e_profile *profile); struct mlx5e_flow_steering *mlx5e_fs_init(const struct mlx5e_profile *profile, struct mlx5_core_dev *mdev, bool state_destroy); void mlx5e_fs_cleanup(struct mlx5e_flow_steering *fs); - -int mlx5e_add_vlan_trap(struct mlx5e_priv *priv, int trap_id, int tir_num); -void mlx5e_remove_vlan_trap(struct mlx5e_priv *priv); -int mlx5e_add_mac_trap(struct mlx5e_priv *priv, int trap_id, int tir_num); -void mlx5e_remove_mac_trap(struct mlx5e_priv *priv); +struct mlx5e_vlan_table *mlx5e_fs_get_vlan(struct mlx5e_flow_steering *fs); +void mlx5e_fs_set_tc(struct mlx5e_flow_steering *fs, struct mlx5e_tc_table *tc); +struct mlx5e_tc_table *mlx5e_fs_get_tc(struct mlx5e_flow_steering *fs); +struct mlx5e_l2_table *mlx5e_fs_get_l2(struct mlx5e_flow_steering *fs); +struct mlx5_flow_namespace *mlx5e_fs_get_ns(struct mlx5e_flow_steering *fs, bool egress); +void mlx5e_fs_set_ns(struct mlx5e_flow_steering *fs, struct mlx5_flow_namespace *ns, bool egress); +#ifdef CONFIG_MLX5_EN_RXNFC +struct mlx5e_ethtool_steering *mlx5e_fs_get_ethtool(struct mlx5e_flow_steering *fs); +#endif +struct mlx5_ttc_table *mlx5e_fs_get_ttc(struct mlx5e_flow_steering *fs, bool inner); +void mlx5e_fs_set_ttc(struct mlx5e_flow_steering *fs, struct mlx5_ttc_table *ttc, bool inner); +#ifdef CONFIG_MLX5_EN_ARFS +struct mlx5e_arfs_tables *mlx5e_fs_get_arfs(struct mlx5e_flow_steering *fs); +void mlx5e_fs_set_arfs(struct mlx5e_flow_steering *fs, struct mlx5e_arfs_tables *arfs); +#endif +struct mlx5e_ptp_fs *mlx5e_fs_get_ptp(struct mlx5e_flow_steering *fs); +void mlx5e_fs_set_ptp(struct mlx5e_flow_steering *fs, struct mlx5e_ptp_fs *ptp_fs); +struct mlx5e_fs_any *mlx5e_fs_get_any(struct mlx5e_flow_steering *fs); +void mlx5e_fs_set_any(struct mlx5e_flow_steering *fs, struct mlx5e_fs_any *any); +struct mlx5e_fs_udp *mlx5e_fs_get_udp(struct mlx5e_flow_steering *fs); +void mlx5e_fs_set_udp(struct mlx5e_flow_steering *fs, struct mlx5e_fs_udp *udp); +#ifdef CONFIG_MLX5_EN_TLS +struct mlx5e_accel_fs_tcp *mlx5e_fs_get_accel_tcp(struct mlx5e_flow_steering *fs); +void mlx5e_fs_set_accel_tcp(struct mlx5e_flow_steering *fs, struct mlx5e_accel_fs_tcp *accel_tcp); +#endif +void mlx5e_fs_set_state_destroy(struct mlx5e_flow_steering *fs, bool state_destroy); +void mlx5e_fs_set_vlan_strip_disable(struct mlx5e_flow_steering *fs, bool vlan_strip_disable); + +struct mlx5_core_dev *mlx5e_fs_get_mdev(struct mlx5e_flow_steering *fs); +int mlx5e_add_vlan_trap(struct mlx5e_flow_steering *fs, int trap_id, int tir_num); +void mlx5e_remove_vlan_trap(struct mlx5e_flow_steering *fs); +int mlx5e_add_mac_trap(struct mlx5e_flow_steering *fs, int trap_id, int tir_num); +void mlx5e_remove_mac_trap(struct mlx5e_flow_steering *fs); void mlx5e_fs_set_rx_mode_work(struct mlx5e_flow_steering *fs, struct net_device *netdev); int mlx5e_fs_vlan_rx_add_vid(struct mlx5e_flow_steering *fs, struct net_device *netdev, @@ -198,5 +187,18 @@ int mlx5e_fs_vlan_rx_kill_vid(struct mlx5e_flow_steering *fs, struct net_device *netdev, __be16 proto, u16 vid); void mlx5e_fs_init_l2_addr(struct mlx5e_flow_steering *fs, struct net_device *netdev); + +#define fs_err(fs, fmt, ...) \ + mlx5_core_err(mlx5e_fs_get_mdev(fs), fmt, ##__VA_ARGS__) + +#define fs_dbg(fs, fmt, ...) \ + mlx5_core_dbg(mlx5e_fs_get_mdev(fs), fmt, ##__VA_ARGS__) + +#define fs_warn(fs, fmt, ...) \ + mlx5_core_warn(mlx5e_fs_get_mdev(fs), fmt, ##__VA_ARGS__) + +#define fs_warn_once(fs, fmt, ...) \ + mlx5_core_warn_once(mlx5e_fs_get_mdev(fs), fmt, ##__VA_ARGS__) + #endif /* __MLX5E_FLOW_STEER_H__ */ diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/fs_ethtool.h b/drivers/net/ethernet/mellanox/mlx5/core/en/fs_ethtool.h new file mode 100644 index 000000000000..9e276fd3c0cf --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/fs_ethtool.h @@ -0,0 +1,29 @@ +/* SPDX-License-Identifier: GPL-2.0 OR Linux-OpenIB */ +/* Copyright (c) 2022, NVIDIA CORPORATION & AFFILIATES. */ + +#ifndef __MLX5E_FS_ETHTOOL_H__ +#define __MLX5E_FS_ETHTOOL_H__ + +struct mlx5e_priv; +struct mlx5e_ethtool_steering; +#ifdef CONFIG_MLX5_EN_RXNFC +int mlx5e_ethtool_alloc(struct mlx5e_ethtool_steering **ethtool); +void mlx5e_ethtool_free(struct mlx5e_ethtool_steering *ethtool); +void mlx5e_ethtool_init_steering(struct mlx5e_flow_steering *fs); +void mlx5e_ethtool_cleanup_steering(struct mlx5e_flow_steering *fs); +int mlx5e_ethtool_set_rxnfc(struct mlx5e_priv *priv, struct ethtool_rxnfc *cmd); +int mlx5e_ethtool_get_rxnfc(struct mlx5e_priv *priv, + struct ethtool_rxnfc *info, u32 *rule_locs); +#else +static inline int mlx5e_ethtool_alloc(struct mlx5e_ethtool_steering **ethtool) +{ return 0; } +static inline void mlx5e_ethtool_free(struct mlx5e_ethtool_steering *ethtool) { } +static inline void mlx5e_ethtool_init_steering(struct mlx5e_flow_steering *fs) { } +static inline void mlx5e_ethtool_cleanup_steering(struct mlx5e_flow_steering *fs) { } +static inline int mlx5e_ethtool_set_rxnfc(struct mlx5e_priv *priv, struct ethtool_rxnfc *cmd) +{ return -EOPNOTSUPP; } +static inline int mlx5e_ethtool_get_rxnfc(struct mlx5e_priv *priv, + struct ethtool_rxnfc *info, u32 *rule_locs) +{ return -EOPNOTSUPP; } +#endif +#endif diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/fs_tt_redirect.c b/drivers/net/ethernet/mellanox/mlx5/core/en/fs_tt_redirect.c index e153d6119e02..03cb79adf912 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/fs_tt_redirect.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/fs_tt_redirect.c @@ -1,9 +1,9 @@ // SPDX-License-Identifier: GPL-2.0 OR Linux-OpenIB /* Copyright (c) 2021, Mellanox Technologies inc. All rights reserved. */ -#include <linux/netdevice.h> #include "en/fs_tt_redirect.h" #include "fs_core.h" +#include "mlx5_core.h" enum fs_udp_type { FS_IPV4_UDP, @@ -74,17 +74,17 @@ static void fs_udp_set_dport_flow(struct mlx5_flow_spec *spec, enum fs_udp_type } struct mlx5_flow_handle * -mlx5e_fs_tt_redirect_udp_add_rule(struct mlx5e_priv *priv, +mlx5e_fs_tt_redirect_udp_add_rule(struct mlx5e_flow_steering *fs, enum mlx5_traffic_types ttc_type, u32 tir_num, u16 d_port) { + struct mlx5e_fs_udp *fs_udp = mlx5e_fs_get_udp(fs); enum fs_udp_type type = tt2fs_udp(ttc_type); struct mlx5_flow_destination dest = {}; struct mlx5_flow_table *ft = NULL; MLX5_DECLARE_FLOW_ACT(flow_act); struct mlx5_flow_handle *rule; struct mlx5_flow_spec *spec; - struct mlx5e_fs_udp *fs_udp; int err; if (type == FS_UDP_NUM_TYPES) @@ -94,7 +94,6 @@ mlx5e_fs_tt_redirect_udp_add_rule(struct mlx5e_priv *priv, if (!spec) return ERR_PTR(-ENOMEM); - fs_udp = priv->fs->udp; ft = fs_udp->tables[type].t; fs_udp_set_dport_flow(spec, type, d_port); @@ -106,31 +105,30 @@ mlx5e_fs_tt_redirect_udp_add_rule(struct mlx5e_priv *priv, if (IS_ERR(rule)) { err = PTR_ERR(rule); - netdev_err(priv->netdev, "%s: add %s rule failed, err %d\n", - __func__, fs_udp_type2str(type), err); + fs_err(fs, "%s: add %s rule failed, err %d\n", + __func__, fs_udp_type2str(type), err); } return rule; } -static int fs_udp_add_default_rule(struct mlx5e_priv *priv, enum fs_udp_type type) +static int fs_udp_add_default_rule(struct mlx5e_flow_steering *fs, enum fs_udp_type type) { + struct mlx5_ttc_table *ttc = mlx5e_fs_get_ttc(fs, false); + struct mlx5e_fs_udp *fs_udp = mlx5e_fs_get_udp(fs); struct mlx5e_flow_table *fs_udp_t; struct mlx5_flow_destination dest; MLX5_DECLARE_FLOW_ACT(flow_act); struct mlx5_flow_handle *rule; - struct mlx5e_fs_udp *fs_udp; int err; - fs_udp = priv->fs->udp; fs_udp_t = &fs_udp->tables[type]; - dest = mlx5_ttc_get_default_dest(priv->fs->ttc, fs_udp2tt(type)); + dest = mlx5_ttc_get_default_dest(ttc, fs_udp2tt(type)); rule = mlx5_add_flow_rules(fs_udp_t->t, NULL, &flow_act, &dest, 1); if (IS_ERR(rule)) { err = PTR_ERR(rule); - netdev_err(priv->netdev, - "%s: add default rule failed, fs type=%d, err %d\n", - __func__, type, err); + fs_err(fs, "%s: add default rule failed, fs type=%d, err %d\n", + __func__, type, err); return err; } @@ -206,33 +204,36 @@ out: return err; } -static int fs_udp_create_table(struct mlx5e_priv *priv, enum fs_udp_type type) +static int fs_udp_create_table(struct mlx5e_flow_steering *fs, enum fs_udp_type type) { - struct mlx5e_flow_table *ft = &priv->fs->udp->tables[type]; + struct mlx5_flow_namespace *ns = mlx5e_fs_get_ns(fs, false); + struct mlx5e_fs_udp *fs_udp = mlx5e_fs_get_udp(fs); struct mlx5_flow_table_attr ft_attr = {}; + struct mlx5e_flow_table *ft; int err; + ft = &fs_udp->tables[type]; ft->num_groups = 0; ft_attr.max_fte = MLX5E_FS_UDP_TABLE_SIZE; ft_attr.level = MLX5E_FS_TT_UDP_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(ns, &ft_attr); if (IS_ERR(ft->t)) { err = PTR_ERR(ft->t); ft->t = NULL; return err; } - netdev_dbg(priv->netdev, "Created fs %s table id %u level %u\n", - fs_udp_type2str(type), ft->t->id, ft->t->level); + mlx5_core_dbg(mlx5e_fs_get_mdev(fs), "Created fs %s table id %u level %u\n", + fs_udp_type2str(type), ft->t->id, ft->t->level); err = fs_udp_create_groups(ft, type); if (err) goto err; - err = fs_udp_add_default_rule(priv, type); + err = fs_udp_add_default_rule(fs, type); if (err) goto err; @@ -253,17 +254,17 @@ static void fs_udp_destroy_table(struct mlx5e_fs_udp *fs_udp, int i) fs_udp->tables[i].t = NULL; } -static int fs_udp_disable(struct mlx5e_priv *priv) +static int fs_udp_disable(struct mlx5e_flow_steering *fs) { + struct mlx5_ttc_table *ttc = mlx5e_fs_get_ttc(fs, false); int err, i; for (i = 0; i < FS_UDP_NUM_TYPES; i++) { /* Modify ttc rules destination to point back to the indir TIRs */ - err = mlx5_ttc_fwd_default_dest(priv->fs->ttc, fs_udp2tt(i)); + err = mlx5_ttc_fwd_default_dest(ttc, fs_udp2tt(i)); if (err) { - netdev_err(priv->netdev, - "%s: modify ttc[%d] default destination failed, err(%d)\n", - __func__, fs_udp2tt(i), err); + fs_err(fs, "%s: modify ttc[%d] default destination failed, err(%d)\n", + __func__, fs_udp2tt(i), err); return err; } } @@ -271,30 +272,31 @@ static int fs_udp_disable(struct mlx5e_priv *priv) return 0; } -static int fs_udp_enable(struct mlx5e_priv *priv) +static int fs_udp_enable(struct mlx5e_flow_steering *fs) { + struct mlx5_ttc_table *ttc = mlx5e_fs_get_ttc(fs, false); + struct mlx5e_fs_udp *udp = mlx5e_fs_get_udp(fs); struct mlx5_flow_destination dest = {}; int err, i; dest.type = MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE; for (i = 0; i < FS_UDP_NUM_TYPES; i++) { - dest.ft = priv->fs->udp->tables[i].t; + dest.ft = udp->tables[i].t; /* Modify ttc rules destination to point on the accel_fs FTs */ - err = mlx5_ttc_fwd_dest(priv->fs->ttc, fs_udp2tt(i), &dest); + err = mlx5_ttc_fwd_dest(ttc, fs_udp2tt(i), &dest); if (err) { - netdev_err(priv->netdev, - "%s: modify ttc[%d] destination to accel failed, err(%d)\n", - __func__, fs_udp2tt(i), err); + fs_err(fs, "%s: modify ttc[%d] destination to accel failed, err(%d)\n", + __func__, fs_udp2tt(i), err); return err; } } return 0; } -void mlx5e_fs_tt_redirect_udp_destroy(struct mlx5e_priv *priv) +void mlx5e_fs_tt_redirect_udp_destroy(struct mlx5e_flow_steering *fs) { - struct mlx5e_fs_udp *fs_udp = priv->fs->udp; + struct mlx5e_fs_udp *fs_udp = mlx5e_fs_get_udp(fs); int i; if (!fs_udp) @@ -303,48 +305,50 @@ void mlx5e_fs_tt_redirect_udp_destroy(struct mlx5e_priv *priv) if (--fs_udp->ref_cnt) return; - fs_udp_disable(priv); + fs_udp_disable(fs); for (i = 0; i < FS_UDP_NUM_TYPES; i++) fs_udp_destroy_table(fs_udp, i); kfree(fs_udp); - priv->fs->udp = NULL; + mlx5e_fs_set_udp(fs, NULL); } -int mlx5e_fs_tt_redirect_udp_create(struct mlx5e_priv *priv) +int mlx5e_fs_tt_redirect_udp_create(struct mlx5e_flow_steering *fs) { + struct mlx5e_fs_udp *udp = mlx5e_fs_get_udp(fs); int i, err; - if (priv->fs->udp) { - priv->fs->udp->ref_cnt++; + if (udp) { + udp->ref_cnt++; return 0; } - priv->fs->udp = kzalloc(sizeof(*priv->fs->udp), GFP_KERNEL); - if (!priv->fs->udp) + udp = kzalloc(sizeof(*udp), GFP_KERNEL); + if (!udp) return -ENOMEM; + mlx5e_fs_set_udp(fs, udp); for (i = 0; i < FS_UDP_NUM_TYPES; i++) { - err = fs_udp_create_table(priv, i); + err = fs_udp_create_table(fs, i); if (err) goto err_destroy_tables; } - err = fs_udp_enable(priv); + err = fs_udp_enable(fs); if (err) goto err_destroy_tables; - priv->fs->udp->ref_cnt = 1; + udp->ref_cnt = 1; return 0; err_destroy_tables: while (--i >= 0) - fs_udp_destroy_table(priv->fs->udp, i); + fs_udp_destroy_table(udp, i); - kfree(priv->fs->udp); - priv->fs->udp = NULL; + kfree(udp); + mlx5e_fs_set_udp(fs, NULL); return err; } @@ -356,22 +360,21 @@ static void fs_any_set_ethertype_flow(struct mlx5_flow_spec *spec, u16 ether_typ } struct mlx5_flow_handle * -mlx5e_fs_tt_redirect_any_add_rule(struct mlx5e_priv *priv, +mlx5e_fs_tt_redirect_any_add_rule(struct mlx5e_flow_steering *fs, u32 tir_num, u16 ether_type) { + struct mlx5e_fs_any *fs_any = mlx5e_fs_get_any(fs); struct mlx5_flow_destination dest = {}; struct mlx5_flow_table *ft = NULL; MLX5_DECLARE_FLOW_ACT(flow_act); struct mlx5_flow_handle *rule; struct mlx5_flow_spec *spec; - struct mlx5e_fs_any *fs_any; int err; spec = kvzalloc(sizeof(*spec), GFP_KERNEL); if (!spec) return ERR_PTR(-ENOMEM); - fs_any = priv->fs->any; ft = fs_any->table.t; fs_any_set_ethertype_flow(spec, ether_type); @@ -383,31 +386,29 @@ mlx5e_fs_tt_redirect_any_add_rule(struct mlx5e_priv *priv, if (IS_ERR(rule)) { err = PTR_ERR(rule); - netdev_err(priv->netdev, "%s: add ANY rule failed, err %d\n", - __func__, err); + fs_err(fs, "%s: add ANY rule failed, err %d\n", + __func__, err); } return rule; } -static int fs_any_add_default_rule(struct mlx5e_priv *priv) +static int fs_any_add_default_rule(struct mlx5e_flow_steering *fs) { + struct mlx5_ttc_table *ttc = mlx5e_fs_get_ttc(fs, false); + struct mlx5e_fs_any *fs_any = mlx5e_fs_get_any(fs); struct mlx5e_flow_table *fs_any_t; struct mlx5_flow_destination dest; MLX5_DECLARE_FLOW_ACT(flow_act); struct mlx5_flow_handle *rule; - struct mlx5e_fs_any *fs_any; int err; - fs_any = priv->fs->any; fs_any_t = &fs_any->table; - - dest = mlx5_ttc_get_default_dest(priv->fs->ttc, MLX5_TT_ANY); + dest = mlx5_ttc_get_default_dest(ttc, MLX5_TT_ANY); rule = mlx5_add_flow_rules(fs_any_t->t, NULL, &flow_act, &dest, 1); if (IS_ERR(rule)) { err = PTR_ERR(rule); - netdev_err(priv->netdev, - "%s: add default rule failed, fs type=ANY, err %d\n", - __func__, err); + fs_err(fs, "%s: add default rule failed, fs type=ANY, err %d\n", + __func__, err); return err; } @@ -472,9 +473,11 @@ err: return err; } -static int fs_any_create_table(struct mlx5e_priv *priv) +static int fs_any_create_table(struct mlx5e_flow_steering *fs) { - struct mlx5e_flow_table *ft = &priv->fs->any->table; + struct mlx5_flow_namespace *ns = mlx5e_fs_get_ns(fs, false); + struct mlx5e_fs_any *fs_any = mlx5e_fs_get_any(fs); + struct mlx5e_flow_table *ft = &fs_any->table; struct mlx5_flow_table_attr ft_attr = {}; int err; @@ -484,21 +487,21 @@ static int fs_any_create_table(struct mlx5e_priv *priv) ft_attr.level = MLX5E_FS_TT_ANY_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(ns, &ft_attr); if (IS_ERR(ft->t)) { err = PTR_ERR(ft->t); ft->t = NULL; return err; } - netdev_dbg(priv->netdev, "Created fs ANY table id %u level %u\n", - ft->t->id, ft->t->level); + mlx5_core_dbg(mlx5e_fs_get_mdev(fs), "Created fs ANY table id %u level %u\n", + ft->t->id, ft->t->level); err = fs_any_create_groups(ft); if (err) goto err; - err = fs_any_add_default_rule(priv); + err = fs_any_add_default_rule(fs); if (err) goto err; @@ -509,35 +512,38 @@ err: return err; } -static int fs_any_disable(struct mlx5e_priv *priv) +static int fs_any_disable(struct mlx5e_flow_steering *fs) { + struct mlx5_ttc_table *ttc = mlx5e_fs_get_ttc(fs, false); int err; /* Modify ttc rules destination to point back to the indir TIRs */ - err = mlx5_ttc_fwd_default_dest(priv->fs->ttc, MLX5_TT_ANY); + err = mlx5_ttc_fwd_default_dest(ttc, MLX5_TT_ANY); if (err) { - netdev_err(priv->netdev, - "%s: modify ttc[%d] default destination failed, err(%d)\n", - __func__, MLX5_TT_ANY, err); + fs_err(fs, + "%s: modify ttc[%d] default destination failed, err(%d)\n", + __func__, MLX5_TT_ANY, err); return err; } return 0; } -static int fs_any_enable(struct mlx5e_priv *priv) +static int fs_any_enable(struct mlx5e_flow_steering *fs) { + struct mlx5_ttc_table *ttc = mlx5e_fs_get_ttc(fs, false); + struct mlx5e_fs_any *any = mlx5e_fs_get_any(fs); struct mlx5_flow_destination dest = {}; int err; dest.type = MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE; - dest.ft = priv->fs->any->table.t; + dest.ft = any->table.t; /* Modify ttc rules destination to point on the accel_fs FTs */ - err = mlx5_ttc_fwd_dest(priv->fs->ttc, MLX5_TT_ANY, &dest); + err = mlx5_ttc_fwd_dest(ttc, MLX5_TT_ANY, &dest); if (err) { - netdev_err(priv->netdev, - "%s: modify ttc[%d] destination to accel failed, err(%d)\n", - __func__, MLX5_TT_ANY, err); + fs_err(fs, + "%s: modify ttc[%d] destination to accel failed, err(%d)\n", + __func__, MLX5_TT_ANY, err); return err; } return 0; @@ -553,9 +559,9 @@ static void fs_any_destroy_table(struct mlx5e_fs_any *fs_any) fs_any->table.t = NULL; } -void mlx5e_fs_tt_redirect_any_destroy(struct mlx5e_priv *priv) +void mlx5e_fs_tt_redirect_any_destroy(struct mlx5e_flow_steering *fs) { - struct mlx5e_fs_any *fs_any = priv->fs->any; + struct mlx5e_fs_any *fs_any = mlx5e_fs_get_any(fs); if (!fs_any) return; @@ -563,43 +569,45 @@ void mlx5e_fs_tt_redirect_any_destroy(struct mlx5e_priv *priv) if (--fs_any->ref_cnt) return; - fs_any_disable(priv); + fs_any_disable(fs); fs_any_destroy_table(fs_any); kfree(fs_any); - priv->fs->any = NULL; + mlx5e_fs_set_any(fs, NULL); } -int mlx5e_fs_tt_redirect_any_create(struct mlx5e_priv *priv) +int mlx5e_fs_tt_redirect_any_create(struct mlx5e_flow_steering *fs) { + struct mlx5e_fs_any *fs_any = mlx5e_fs_get_any(fs); int err; - if (priv->fs->any) { - priv->fs->any->ref_cnt++; + if (fs_any) { + fs_any->ref_cnt++; return 0; } - priv->fs->any = kzalloc(sizeof(*priv->fs->any), GFP_KERNEL); - if (!priv->fs->any) + fs_any = kzalloc(sizeof(*fs_any), GFP_KERNEL); + if (!fs_any) return -ENOMEM; + mlx5e_fs_set_any(fs, fs_any); - err = fs_any_create_table(priv); + err = fs_any_create_table(fs); if (err) return err; - err = fs_any_enable(priv); + err = fs_any_enable(fs); if (err) goto err_destroy_table; - priv->fs->any->ref_cnt = 1; + fs_any->ref_cnt = 1; return 0; err_destroy_table: - fs_any_destroy_table(priv->fs->any); + fs_any_destroy_table(fs_any); - kfree(priv->fs->any); - priv->fs->any = NULL; + kfree(fs_any); + mlx5e_fs_set_any(fs, NULL); return err; } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/fs_tt_redirect.h b/drivers/net/ethernet/mellanox/mlx5/core/en/fs_tt_redirect.h index 7a70c4f38fda..5780fd7ad507 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/fs_tt_redirect.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/fs_tt_redirect.h @@ -4,23 +4,22 @@ #ifndef __MLX5E_FS_TT_REDIRECT_H__ #define __MLX5E_FS_TT_REDIRECT_H__ -#include "en.h" #include "en/fs.h" void mlx5e_fs_tt_redirect_del_rule(struct mlx5_flow_handle *rule); /* UDP traffic type redirect */ struct mlx5_flow_handle * -mlx5e_fs_tt_redirect_udp_add_rule(struct mlx5e_priv *priv, +mlx5e_fs_tt_redirect_udp_add_rule(struct mlx5e_flow_steering *fs, enum mlx5_traffic_types ttc_type, u32 tir_num, u16 d_port); -void mlx5e_fs_tt_redirect_udp_destroy(struct mlx5e_priv *priv); -int mlx5e_fs_tt_redirect_udp_create(struct mlx5e_priv *priv); +void mlx5e_fs_tt_redirect_udp_destroy(struct mlx5e_flow_steering *fs); +int mlx5e_fs_tt_redirect_udp_create(struct mlx5e_flow_steering *fs); /* ANY traffic type redirect*/ struct mlx5_flow_handle * -mlx5e_fs_tt_redirect_any_add_rule(struct mlx5e_priv *priv, +mlx5e_fs_tt_redirect_any_add_rule(struct mlx5e_flow_steering *fs, u32 tir_num, u16 ether_type); -void mlx5e_fs_tt_redirect_any_destroy(struct mlx5e_priv *priv); -int mlx5e_fs_tt_redirect_any_create(struct mlx5e_priv *priv); +void mlx5e_fs_tt_redirect_any_destroy(struct mlx5e_flow_steering *fs); +int mlx5e_fs_tt_redirect_any_create(struct mlx5e_flow_steering *fs); #endif diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/params.c b/drivers/net/ethernet/mellanox/mlx5/core/en/params.c index e025040350ba..29dd3a04c154 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/params.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/params.c @@ -6,11 +6,212 @@ #include "en/port.h" #include "en_accel/en_accel.h" #include "en_accel/ipsec.h" +#include <net/xdp_sock_drv.h> -static bool mlx5e_rx_is_xdp(struct mlx5e_params *params, - struct mlx5e_xsk_param *xsk) +static u8 mlx5e_mpwrq_min_page_shift(struct mlx5_core_dev *mdev) +{ + u8 min_page_shift = MLX5_CAP_GEN_2(mdev, log_min_mkey_entity_size); + + return min_page_shift ? : 12; +} + +u8 mlx5e_mpwrq_page_shift(struct mlx5_core_dev *mdev, struct mlx5e_xsk_param *xsk) +{ + u8 req_page_shift = xsk ? order_base_2(xsk->chunk_size) : PAGE_SHIFT; + u8 min_page_shift = mlx5e_mpwrq_min_page_shift(mdev); + + /* Regular RQ uses order-0 pages, the NIC must be able to map them. */ + if (WARN_ON_ONCE(!xsk && req_page_shift < min_page_shift)) + min_page_shift = req_page_shift; + + return max(req_page_shift, min_page_shift); +} + +enum mlx5e_mpwrq_umr_mode +mlx5e_mpwrq_umr_mode(struct mlx5_core_dev *mdev, struct mlx5e_xsk_param *xsk) +{ + /* Different memory management schemes use different mechanisms to map + * user-mode memory. The stricter guarantees we have, the faster + * mechanisms we use: + * 1. MTT - direct mapping in page granularity. + * 2. KSM - indirect mapping to another MKey to arbitrary addresses, but + * all mappings have the same size. + * 3. KLM - indirect mapping to another MKey to arbitrary addresses, and + * mappings can have different sizes. + */ + u8 page_shift = mlx5e_mpwrq_page_shift(mdev, xsk); + bool unaligned = xsk ? xsk->unaligned : false; + bool oversized = false; + + if (xsk) { + oversized = xsk->chunk_size < (1 << page_shift); + WARN_ON_ONCE(xsk->chunk_size > (1 << page_shift)); + } + + /* XSK frame size doesn't match the UMR page size, either because the + * frame size is not a power of two, or it's smaller than the minimal + * page size supported by the firmware. + * It's possible to receive packets bigger than MTU in certain setups. + * To avoid writing over the XSK frame boundary, the top region of each + * stride is mapped to a garbage page, resulting in two mappings of + * different sizes per frame. + */ + if (oversized) { + /* An optimization for frame sizes equal to 3 * power_of_two. + * 3 KSMs point to the frame, and one KSM points to the garbage + * page, which works faster than KLM. + */ + if (xsk->chunk_size % 3 == 0 && is_power_of_2(xsk->chunk_size / 3)) + return MLX5E_MPWRQ_UMR_MODE_TRIPLE; + + return MLX5E_MPWRQ_UMR_MODE_OVERSIZED; + } + + /* XSK frames can start at arbitrary unaligned locations, but they all + * have the same size which is a power of two. It allows to optimize to + * one KSM per frame. + */ + if (unaligned) + return MLX5E_MPWRQ_UMR_MODE_UNALIGNED; + + /* XSK: frames are naturally aligned, MTT can be used. + * Non-XSK: Allocations happen in units of CPU pages, therefore, the + * mappings are naturally aligned. + */ + return MLX5E_MPWRQ_UMR_MODE_ALIGNED; +} + +u8 mlx5e_mpwrq_umr_entry_size(enum mlx5e_mpwrq_umr_mode mode) +{ + switch (mode) { + case MLX5E_MPWRQ_UMR_MODE_ALIGNED: + return sizeof(struct mlx5_mtt); + case MLX5E_MPWRQ_UMR_MODE_UNALIGNED: + return sizeof(struct mlx5_ksm); + case MLX5E_MPWRQ_UMR_MODE_OVERSIZED: + return sizeof(struct mlx5_klm) * 2; + case MLX5E_MPWRQ_UMR_MODE_TRIPLE: + return sizeof(struct mlx5_ksm) * 4; + } + WARN_ONCE(1, "MPWRQ UMR mode %d is not known\n", mode); + return 0; +} + +u8 mlx5e_mpwrq_log_wqe_sz(struct mlx5_core_dev *mdev, u8 page_shift, + enum mlx5e_mpwrq_umr_mode umr_mode) +{ + u8 umr_entry_size = mlx5e_mpwrq_umr_entry_size(umr_mode); + u8 max_pages_per_wqe, max_log_mpwqe_size; + u16 max_wqe_size; + + /* Keep in sync with MLX5_MPWRQ_MAX_PAGES_PER_WQE. */ + max_wqe_size = mlx5e_get_max_sq_aligned_wqebbs(mdev) * MLX5_SEND_WQE_BB; + max_pages_per_wqe = ALIGN_DOWN(max_wqe_size - sizeof(struct mlx5e_umr_wqe), + MLX5_UMR_MTT_ALIGNMENT) / umr_entry_size; + max_log_mpwqe_size = ilog2(max_pages_per_wqe) + page_shift; + + WARN_ON_ONCE(max_log_mpwqe_size < MLX5E_ORDER2_MAX_PACKET_MTU); + + return min_t(u8, max_log_mpwqe_size, MLX5_MPWRQ_MAX_LOG_WQE_SZ); +} + +u8 mlx5e_mpwrq_pages_per_wqe(struct mlx5_core_dev *mdev, u8 page_shift, + enum mlx5e_mpwrq_umr_mode umr_mode) { - return params->xdp_prog || xsk; + u8 log_wqe_sz = mlx5e_mpwrq_log_wqe_sz(mdev, page_shift, umr_mode); + u8 pages_per_wqe; + + pages_per_wqe = log_wqe_sz > page_shift ? (1 << (log_wqe_sz - page_shift)) : 1; + + /* Two MTTs are needed to form an octword. The number of MTTs is encoded + * in octwords in a UMR WQE, so we need at least two to avoid mapping + * garbage addresses. + */ + if (WARN_ON_ONCE(pages_per_wqe < 2 && umr_mode == MLX5E_MPWRQ_UMR_MODE_ALIGNED)) + pages_per_wqe = 2; + + /* Sanity check for further calculations to succeed. */ + BUILD_BUG_ON(MLX5_MPWRQ_MAX_PAGES_PER_WQE > 64); + if (WARN_ON_ONCE(pages_per_wqe > MLX5_MPWRQ_MAX_PAGES_PER_WQE)) + return MLX5_MPWRQ_MAX_PAGES_PER_WQE; + + return pages_per_wqe; +} + +u16 mlx5e_mpwrq_umr_wqe_sz(struct mlx5_core_dev *mdev, u8 page_shift, + enum mlx5e_mpwrq_umr_mode umr_mode) +{ + u8 pages_per_wqe = mlx5e_mpwrq_pages_per_wqe(mdev, page_shift, umr_mode); + u8 umr_entry_size = mlx5e_mpwrq_umr_entry_size(umr_mode); + u16 umr_wqe_sz; + + umr_wqe_sz = sizeof(struct mlx5e_umr_wqe) + + ALIGN(pages_per_wqe * umr_entry_size, MLX5_UMR_MTT_ALIGNMENT); + + WARN_ON_ONCE(DIV_ROUND_UP(umr_wqe_sz, MLX5_SEND_WQE_DS) > MLX5_WQE_CTRL_DS_MASK); + + return umr_wqe_sz; +} + +u8 mlx5e_mpwrq_umr_wqebbs(struct mlx5_core_dev *mdev, u8 page_shift, + enum mlx5e_mpwrq_umr_mode umr_mode) +{ + return DIV_ROUND_UP(mlx5e_mpwrq_umr_wqe_sz(mdev, page_shift, umr_mode), + MLX5_SEND_WQE_BB); +} + +u8 mlx5e_mpwrq_mtts_per_wqe(struct mlx5_core_dev *mdev, u8 page_shift, + enum mlx5e_mpwrq_umr_mode umr_mode) +{ + u8 pages_per_wqe = mlx5e_mpwrq_pages_per_wqe(mdev, page_shift, umr_mode); + + /* Add another page as a buffer between WQEs. This page will absorb + * write overflow by the hardware, when receiving packets larger than + * MTU. These oversize packets are dropped by the driver at a later + * stage. + */ + return ALIGN(pages_per_wqe + 1, + MLX5_SEND_WQE_BB / mlx5e_mpwrq_umr_entry_size(umr_mode)); +} + +u32 mlx5e_mpwrq_max_num_entries(struct mlx5_core_dev *mdev, + enum mlx5e_mpwrq_umr_mode umr_mode) +{ + /* Same limits apply to KSMs and KLMs. */ + u32 klm_limit = min(MLX5E_MAX_RQ_NUM_KSMS, + 1 << MLX5_CAP_GEN(mdev, log_max_klm_list_size)); + + switch (umr_mode) { + case MLX5E_MPWRQ_UMR_MODE_ALIGNED: + return MLX5E_MAX_RQ_NUM_MTTS; + case MLX5E_MPWRQ_UMR_MODE_UNALIGNED: + return klm_limit; + case MLX5E_MPWRQ_UMR_MODE_OVERSIZED: + /* Each entry is two KLMs. */ + return klm_limit / 2; + case MLX5E_MPWRQ_UMR_MODE_TRIPLE: + /* Each entry is four KSMs. */ + return klm_limit / 4; + } + WARN_ONCE(1, "MPWRQ UMR mode %d is not known\n", umr_mode); + return 0; +} + +static u8 mlx5e_mpwrq_max_log_rq_size(struct mlx5_core_dev *mdev, u8 page_shift, + enum mlx5e_mpwrq_umr_mode umr_mode) +{ + u8 mtts_per_wqe = mlx5e_mpwrq_mtts_per_wqe(mdev, page_shift, umr_mode); + u32 max_entries = mlx5e_mpwrq_max_num_entries(mdev, umr_mode); + + return ilog2(max_entries / mtts_per_wqe); +} + +u8 mlx5e_mpwrq_max_log_rq_pkts(struct mlx5_core_dev *mdev, u8 page_shift, + enum mlx5e_mpwrq_umr_mode umr_mode) +{ + return mlx5e_mpwrq_max_log_rq_size(mdev, page_shift, umr_mode) + + mlx5e_mpwrq_log_wqe_sz(mdev, page_shift, umr_mode) - + MLX5E_ORDER2_MAX_PACKET_MTU; } u16 mlx5e_get_linear_rq_headroom(struct mlx5e_params *params, @@ -22,7 +223,7 @@ u16 mlx5e_get_linear_rq_headroom(struct mlx5e_params *params, return xsk->headroom; headroom = NET_IP_ALIGN; - if (mlx5e_rx_is_xdp(params, xsk)) + if (params->xdp_prog) headroom += XDP_PACKET_HEADROOM; else headroom += MLX5_RX_HEADROOM; @@ -30,70 +231,80 @@ u16 mlx5e_get_linear_rq_headroom(struct mlx5e_params *params, return headroom; } -u32 mlx5e_rx_get_min_frag_sz(struct mlx5e_params *params, - struct mlx5e_xsk_param *xsk) +static u32 mlx5e_rx_get_linear_sz_xsk(struct mlx5e_params *params, + struct mlx5e_xsk_param *xsk) { u32 hw_mtu = MLX5E_SW2HW_MTU(params, params->sw_mtu); - u16 linear_rq_headroom = mlx5e_get_linear_rq_headroom(params, xsk); - return linear_rq_headroom + hw_mtu; + return xsk->headroom + hw_mtu; } -static u32 mlx5e_rx_get_linear_frag_sz(struct mlx5e_params *params, - struct mlx5e_xsk_param *xsk) +static u32 mlx5e_rx_get_linear_sz_skb(struct mlx5e_params *params, bool xsk) { - u32 frag_sz = mlx5e_rx_get_min_frag_sz(params, xsk); - - /* AF_XDP doesn't build SKBs in place. */ - if (!xsk) - frag_sz = MLX5_SKB_FRAG_SZ(frag_sz); + /* SKBs built on XDP_PASS on XSK RQs don't have headroom. */ + u16 headroom = xsk ? 0 : mlx5e_get_linear_rq_headroom(params, NULL); + u32 hw_mtu = MLX5E_SW2HW_MTU(params, params->sw_mtu); - /* XDP in mlx5e doesn't support multiple packets per page. AF_XDP is a - * special case. It can run with frames smaller than a page, as it - * doesn't allocate pages dynamically. However, here we pretend that - * fragments are page-sized: it allows to treat XSK frames like pages - * by redirecting alloc and free operations to XSK rings and by using - * the fact there are no multiple packets per "page" (which is a frame). - * The latter is important, because frames may come in a random order, - * and we will have trouble assemblying a real page of multiple frames. - */ - if (mlx5e_rx_is_xdp(params, xsk)) - frag_sz = max_t(u32, frag_sz, PAGE_SIZE); + return MLX5_SKB_FRAG_SZ(headroom + hw_mtu); +} - /* Even if we can go with a smaller fragment size, we must not put - * multiple packets into a single frame. +static u32 mlx5e_rx_get_linear_stride_sz(struct mlx5_core_dev *mdev, + struct mlx5e_params *params, + struct mlx5e_xsk_param *xsk, + bool mpwqe) +{ + /* XSK frames are mapped as individual pages, because frames may come in + * an arbitrary order from random locations in the UMEM. */ if (xsk) - frag_sz = max_t(u32, frag_sz, xsk->chunk_size); + return mpwqe ? 1 << mlx5e_mpwrq_page_shift(mdev, xsk) : PAGE_SIZE; + + /* XDP in mlx5e doesn't support multiple packets per page. */ + if (params->xdp_prog) + return PAGE_SIZE; - return frag_sz; + return roundup_pow_of_two(mlx5e_rx_get_linear_sz_skb(params, false)); } -u8 mlx5e_mpwqe_log_pkts_per_wqe(struct mlx5e_params *params, - struct mlx5e_xsk_param *xsk) +static u8 mlx5e_mpwqe_log_pkts_per_wqe(struct mlx5_core_dev *mdev, + struct mlx5e_params *params, + struct mlx5e_xsk_param *xsk) { - u32 linear_frag_sz = mlx5e_rx_get_linear_frag_sz(params, xsk); + u32 linear_stride_sz = mlx5e_rx_get_linear_stride_sz(mdev, params, xsk, true); + enum mlx5e_mpwrq_umr_mode umr_mode = mlx5e_mpwrq_umr_mode(mdev, xsk); + u8 page_shift = mlx5e_mpwrq_page_shift(mdev, xsk); - return MLX5_MPWRQ_LOG_WQE_SZ - order_base_2(linear_frag_sz); + return mlx5e_mpwrq_log_wqe_sz(mdev, page_shift, umr_mode) - + order_base_2(linear_stride_sz); } -bool mlx5e_rx_is_linear_skb(struct mlx5e_params *params, +bool mlx5e_rx_is_linear_skb(struct mlx5_core_dev *mdev, + struct mlx5e_params *params, struct mlx5e_xsk_param *xsk) { - /* AF_XDP allocates SKBs on XDP_PASS - ensure they don't occupy more - * than one page. For this, check both with and without xsk. + if (params->packet_merge.type != MLX5E_PACKET_MERGE_NONE) + return false; + + /* Both XSK and non-XSK cases allocate an SKB on XDP_PASS. Packet data + * must fit into a CPU page. */ - u32 linear_frag_sz = max(mlx5e_rx_get_linear_frag_sz(params, xsk), - mlx5e_rx_get_linear_frag_sz(params, NULL)); + if (mlx5e_rx_get_linear_sz_skb(params, xsk) > PAGE_SIZE) + return false; + + /* XSK frames must be big enough to hold the packet data. */ + if (xsk && mlx5e_rx_get_linear_sz_xsk(params, xsk) > xsk->chunk_size) + return false; - return params->packet_merge.type == MLX5E_PACKET_MERGE_NONE && - linear_frag_sz <= PAGE_SIZE; + return true; } -bool mlx5e_verify_rx_mpwqe_strides(struct mlx5_core_dev *mdev, - u8 log_stride_sz, u8 log_num_strides) +static bool mlx5e_verify_rx_mpwqe_strides(struct mlx5_core_dev *mdev, + u8 log_stride_sz, u8 log_num_strides, + u8 page_shift, + enum mlx5e_mpwrq_umr_mode umr_mode) { - if (log_stride_sz + log_num_strides != MLX5_MPWRQ_LOG_WQE_SZ) + if (log_stride_sz + log_num_strides != + mlx5e_mpwrq_log_wqe_sz(mdev, page_shift, umr_mode)) return false; if (log_stride_sz < MLX5_MPWQE_LOG_STRIDE_SZ_BASE || @@ -113,28 +324,53 @@ bool mlx5e_rx_mpwqe_is_linear_skb(struct mlx5_core_dev *mdev, struct mlx5e_params *params, struct mlx5e_xsk_param *xsk) { - s8 log_num_strides; + enum mlx5e_mpwrq_umr_mode umr_mode = mlx5e_mpwrq_umr_mode(mdev, xsk); + u8 page_shift = mlx5e_mpwrq_page_shift(mdev, xsk); + u8 log_num_strides; u8 log_stride_sz; + u8 log_wqe_sz; - if (!mlx5e_rx_is_linear_skb(params, xsk)) + if (!mlx5e_rx_is_linear_skb(mdev, params, xsk)) return false; - log_stride_sz = order_base_2(mlx5e_rx_get_linear_frag_sz(params, xsk)); - log_num_strides = MLX5_MPWRQ_LOG_WQE_SZ - log_stride_sz; + log_stride_sz = order_base_2(mlx5e_rx_get_linear_stride_sz(mdev, params, xsk, true)); + log_wqe_sz = mlx5e_mpwrq_log_wqe_sz(mdev, page_shift, umr_mode); - return mlx5e_verify_rx_mpwqe_strides(mdev, log_stride_sz, log_num_strides); + if (log_wqe_sz < log_stride_sz) + return false; + + log_num_strides = log_wqe_sz - log_stride_sz; + + return mlx5e_verify_rx_mpwqe_strides(mdev, log_stride_sz, + log_num_strides, page_shift, + umr_mode); } -u8 mlx5e_mpwqe_get_log_rq_size(struct mlx5e_params *params, +u8 mlx5e_mpwqe_get_log_rq_size(struct mlx5_core_dev *mdev, + struct mlx5e_params *params, struct mlx5e_xsk_param *xsk) { - u8 log_pkts_per_wqe = mlx5e_mpwqe_log_pkts_per_wqe(params, xsk); + enum mlx5e_mpwrq_umr_mode umr_mode = mlx5e_mpwrq_umr_mode(mdev, xsk); + u8 log_pkts_per_wqe, page_shift, max_log_rq_size; + + log_pkts_per_wqe = mlx5e_mpwqe_log_pkts_per_wqe(mdev, params, xsk); + page_shift = mlx5e_mpwrq_page_shift(mdev, xsk); + max_log_rq_size = mlx5e_mpwrq_max_log_rq_size(mdev, page_shift, umr_mode); /* Numbers are unsigned, don't subtract to avoid underflow. */ if (params->log_rq_mtu_frames < log_pkts_per_wqe + MLX5E_PARAMS_MINIMUM_LOG_RQ_SIZE_MPW) return MLX5E_PARAMS_MINIMUM_LOG_RQ_SIZE_MPW; + /* Ethtool's rx_max_pending is calculated for regular RQ, that uses + * pages of PAGE_SIZE. Max length of an XSK RQ might differ if it uses a + * frame size not equal to PAGE_SIZE. + * A stricter condition is checked in mlx5e_mpwrq_validate_xsk, WARN on + * unexpected failure. + */ + if (WARN_ON_ONCE(params->log_rq_mtu_frames > log_pkts_per_wqe + max_log_rq_size)) + return max_log_rq_size; + return params->log_rq_mtu_frames - log_pkts_per_wqe; } @@ -164,7 +400,7 @@ u8 mlx5e_mpwqe_get_log_stride_size(struct mlx5_core_dev *mdev, struct mlx5e_xsk_param *xsk) { if (mlx5e_rx_mpwqe_is_linear_skb(mdev, params, xsk)) - return order_base_2(mlx5e_rx_get_linear_frag_sz(params, xsk)); + return order_base_2(mlx5e_rx_get_linear_stride_sz(mdev, params, xsk, true)); return MLX5_MPWRQ_DEF_LOG_STRIDE_SZ(mdev); } @@ -173,7 +409,10 @@ u8 mlx5e_mpwqe_get_log_num_strides(struct mlx5_core_dev *mdev, struct mlx5e_params *params, struct mlx5e_xsk_param *xsk) { - return MLX5_MPWRQ_LOG_WQE_SZ - + enum mlx5e_mpwrq_umr_mode umr_mode = mlx5e_mpwrq_umr_mode(mdev, xsk); + u8 page_shift = mlx5e_mpwrq_page_shift(mdev, xsk); + + return mlx5e_mpwrq_log_wqe_sz(mdev, page_shift, umr_mode) - mlx5e_mpwqe_get_log_stride_size(mdev, params, xsk); } @@ -209,11 +448,11 @@ u16 mlx5e_calc_sq_stop_room(struct mlx5_core_dev *mdev, struct mlx5e_params *par stop_room = mlx5e_ktls_get_stop_room(mdev, params); stop_room += mlx5e_stop_room_for_max_wqe(mdev); if (is_mpwqe) - /* A MPWQE can take up to the maximum-sized WQE + all the normal - * stop room can be taken if a new packet breaks the active - * MPWQE session and allocates its WQEs right away. + /* A MPWQE can take up to the maximum cacheline-aligned WQE + + * all the normal stop room can be taken if a new packet breaks + * the active MPWQE session and allocates its WQEs right away. */ - stop_room += mlx5e_stop_room_for_max_wqe(mdev); + stop_room += mlx5e_stop_room_for_mpwqe(mdev); return stop_room; } @@ -320,22 +559,46 @@ bool slow_pci_heuristic(struct mlx5_core_dev *mdev) link_speed > MLX5E_SLOW_PCI_RATIO * pci_bw; } -bool mlx5e_striding_rq_possible(struct mlx5_core_dev *mdev, - struct mlx5e_params *params) +int mlx5e_mpwrq_validate_regular(struct mlx5_core_dev *mdev, struct mlx5e_params *params) { - if (!mlx5e_check_fragmented_striding_rq_cap(mdev)) - return false; + enum mlx5e_mpwrq_umr_mode umr_mode = mlx5e_mpwrq_umr_mode(mdev, NULL); + u8 page_shift = mlx5e_mpwrq_page_shift(mdev, NULL); - if (params->xdp_prog) { - /* XSK params are not considered here. If striding RQ is in use, - * and an XSK is being opened, mlx5e_rx_mpwqe_is_linear_skb will - * be called with the known XSK params. - */ - if (!mlx5e_rx_mpwqe_is_linear_skb(mdev, params, NULL)) - return false; + if (!mlx5e_check_fragmented_striding_rq_cap(mdev, page_shift, umr_mode)) + return -EOPNOTSUPP; + + if (params->xdp_prog && !mlx5e_rx_mpwqe_is_linear_skb(mdev, params, NULL)) + return -EINVAL; + + return 0; +} + +int mlx5e_mpwrq_validate_xsk(struct mlx5_core_dev *mdev, struct mlx5e_params *params, + struct mlx5e_xsk_param *xsk) +{ + enum mlx5e_mpwrq_umr_mode umr_mode = mlx5e_mpwrq_umr_mode(mdev, xsk); + u8 page_shift = mlx5e_mpwrq_page_shift(mdev, xsk); + bool unaligned = xsk ? xsk->unaligned : false; + u16 max_mtu_pkts; + + if (!mlx5e_check_fragmented_striding_rq_cap(mdev, page_shift, umr_mode)) + return -EOPNOTSUPP; + + if (!mlx5e_rx_mpwqe_is_linear_skb(mdev, params, xsk)) + return -EINVAL; + + /* Current RQ length is too big for the given frame size, the + * needed number of WQEs exceeds the maximum. + */ + max_mtu_pkts = min_t(u8, MLX5E_PARAMS_MAXIMUM_LOG_RQ_SIZE, + mlx5e_mpwrq_max_log_rq_pkts(mdev, page_shift, unaligned)); + if (params->log_rq_mtu_frames > max_mtu_pkts) { + mlx5_core_err(mdev, "Current RQ length %d is too big for XSK with given frame size %u\n", + 1 << params->log_rq_mtu_frames, xsk->chunk_size); + return -EINVAL; } - return true; + return 0; } void mlx5e_init_rq_type_params(struct mlx5_core_dev *mdev, @@ -348,7 +611,7 @@ void mlx5e_init_rq_type_params(struct mlx5_core_dev *mdev, mlx5_core_info(mdev, "MLX5E: StrdRq(%d) RqSz(%ld) StrdSz(%ld) RxCqeCmprss(%d)\n", params->rq_wq_type == MLX5_WQ_TYPE_LINKED_LIST_STRIDING_RQ, params->rq_wq_type == MLX5_WQ_TYPE_LINKED_LIST_STRIDING_RQ ? - BIT(mlx5e_mpwqe_get_log_rq_size(params, NULL)) : + BIT(mlx5e_mpwqe_get_log_rq_size(mdev, params, NULL)) : BIT(params->log_rq_mtu_frames), BIT(mlx5e_mpwqe_get_log_stride_size(mdev, params, NULL)), MLX5E_GET_PFLAG(params, MLX5E_PFLAG_RX_CQE_COMPRESS)); @@ -356,8 +619,7 @@ void mlx5e_init_rq_type_params(struct mlx5_core_dev *mdev, void mlx5e_set_rq_type(struct mlx5_core_dev *mdev, struct mlx5e_params *params) { - params->rq_wq_type = mlx5e_striding_rq_possible(mdev, params) && - MLX5E_GET_PFLAG(params, MLX5E_PFLAG_RX_STRIDING_RQ) ? + params->rq_wq_type = MLX5E_GET_PFLAG(params, MLX5E_PFLAG_RX_STRIDING_RQ) ? MLX5_WQ_TYPE_LINKED_LIST_STRIDING_RQ : MLX5_WQ_TYPE_CYCLIC; } @@ -374,9 +636,9 @@ void mlx5e_build_rq_params(struct mlx5_core_dev *mdev, */ if ((!MLX5E_GET_PFLAG(params, MLX5E_PFLAG_RX_CQE_COMPRESS) || MLX5_CAP_GEN(mdev, mini_cqe_resp_stride_index)) && - mlx5e_striding_rq_possible(mdev, params) && + !mlx5e_mpwrq_validate_regular(mdev, params) && (mlx5e_rx_mpwqe_is_linear_skb(mdev, params, NULL) || - !mlx5e_rx_is_linear_skb(params, NULL))) + !mlx5e_rx_is_linear_skb(mdev, params, NULL))) MLX5E_SET_PFLAG(params, MLX5E_PFLAG_RX_STRIDING_RQ, true); mlx5e_set_rq_type(mdev, params); mlx5e_init_rq_type_params(mdev, params); @@ -419,16 +681,22 @@ static int mlx5e_build_rq_frags_info(struct mlx5_core_dev *mdev, int max_mtu; int i; - if (mlx5e_rx_is_linear_skb(params, xsk)) { + if (mlx5e_rx_is_linear_skb(mdev, params, xsk)) { int frag_stride; - frag_stride = mlx5e_rx_get_linear_frag_sz(params, xsk); - frag_stride = roundup_pow_of_two(frag_stride); + frag_stride = mlx5e_rx_get_linear_stride_sz(mdev, params, xsk, false); info->arr[0].frag_size = byte_count; info->arr[0].frag_stride = frag_stride; info->num_frags = 1; - info->wqe_bulk = PAGE_SIZE / frag_stride; + + /* N WQEs share the same page, N = PAGE_SIZE / frag_stride. The + * first WQE in the page is responsible for allocation of this + * page, this WQE's index is k*N. If WQEs [k*N+1; k*N+N-1] are + * still not completed, the allocation must stop before k*N. + */ + info->wqe_index_mask = (PAGE_SIZE / frag_stride) - 1; + goto out; } @@ -477,11 +745,40 @@ static int mlx5e_build_rq_frags_info(struct mlx5_core_dev *mdev, i++; } info->num_frags = i; - /* number of different wqes sharing a page */ - info->wqe_bulk = 1 + (info->num_frags % 2); + + /* The last fragment of WQE with index 2*N may share the page with the + * first fragment of WQE with index 2*N+1 in certain cases. If WQE 2*N+1 + * is not completed yet, WQE 2*N must not be allocated, as it's + * responsible for allocating a new page. + */ + if (frag_size_max == PAGE_SIZE) { + /* No WQE can start in the middle of a page. */ + info->wqe_index_mask = 0; + } else { + /* PAGE_SIZEs starting from 8192 don't use 2K-sized fragments, + * because there would be more than MLX5E_MAX_RX_FRAGS of them. + */ + WARN_ON(PAGE_SIZE != 2 * DEFAULT_FRAG_SIZE); + + /* Odd number of fragments allows to pack the last fragment of + * the previous WQE and the first fragment of the next WQE into + * the same page. + * As long as DEFAULT_FRAG_SIZE is 2048, and MLX5E_MAX_RX_FRAGS + * is 4, the last fragment can be bigger than the rest only if + * it's the fourth one, so WQEs consisting of 3 fragments will + * always share a page. + * When a page is shared, WQE bulk size is 2, otherwise just 1. + */ + info->wqe_index_mask = info->num_frags % 2; + } out: - info->wqe_bulk = max_t(u8, info->wqe_bulk, 8); + /* Bulking optimization to skip allocation until at least 8 WQEs can be + * allocated in a row. At the same time, never start allocation when + * the page is still used by older WQEs. + */ + info->wqe_bulk = max_t(u8, info->wqe_index_mask + 1, 8); + info->log_num_frags = order_base_2(info->num_frags); return 0; @@ -520,7 +817,7 @@ static u32 mlx5e_shampo_get_log_cq_size(struct mlx5_core_dev *mdev, u16 num_strides = BIT(mlx5e_mpwqe_get_log_num_strides(mdev, params, xsk)); int pkt_per_rsrv = BIT(mlx5e_shampo_get_log_pkt_per_rsrv(mdev, params)); u8 log_stride_sz = mlx5e_mpwqe_get_log_stride_size(mdev, params, xsk); - int wq_size = BIT(mlx5e_mpwqe_get_log_rq_size(params, xsk)); + int wq_size = BIT(mlx5e_mpwqe_get_log_rq_size(mdev, params, xsk)); int wqe_size = BIT(log_stride_sz) * num_strides; /* +1 is for the case that the pkt_per_rsrv dont consume the reservation @@ -544,7 +841,7 @@ static void mlx5e_build_rx_cq_param(struct mlx5_core_dev *mdev, if (params->packet_merge.type == MLX5E_PACKET_MERGE_SHAMPO) log_cq_size = mlx5e_shampo_get_log_cq_size(mdev, params, xsk); else - log_cq_size = mlx5e_mpwqe_get_log_rq_size(params, xsk) + + log_cq_size = mlx5e_mpwqe_get_log_rq_size(mdev, params, xsk) + mlx5e_mpwqe_get_log_num_strides(mdev, params, xsk); break; default: /* MLX5_WQ_TYPE_CYCLIC */ @@ -587,12 +884,16 @@ int mlx5e_build_rq_param(struct mlx5_core_dev *mdev, case MLX5_WQ_TYPE_LINKED_LIST_STRIDING_RQ: { u8 log_wqe_num_of_strides = mlx5e_mpwqe_get_log_num_strides(mdev, params, xsk); u8 log_wqe_stride_size = mlx5e_mpwqe_get_log_stride_size(mdev, params, xsk); + enum mlx5e_mpwrq_umr_mode umr_mode = mlx5e_mpwrq_umr_mode(mdev, xsk); + u8 page_shift = mlx5e_mpwrq_page_shift(mdev, xsk); if (!mlx5e_verify_rx_mpwqe_strides(mdev, log_wqe_stride_size, - log_wqe_num_of_strides)) { + log_wqe_num_of_strides, + page_shift, umr_mode)) { mlx5_core_err(mdev, - "Bad RX MPWQE params: log_stride_size %u, log_num_strides %u\n", - log_wqe_stride_size, log_wqe_num_of_strides); + "Bad RX MPWQE params: log_stride_size %u, log_num_strides %u, umr_mode %d\n", + log_wqe_stride_size, log_wqe_num_of_strides, + umr_mode); return -EINVAL; } @@ -600,7 +901,7 @@ int mlx5e_build_rq_param(struct mlx5_core_dev *mdev, log_wqe_num_of_strides - MLX5_MPWQE_LOG_NUM_STRIDES_BASE); MLX5_SET(wq, wq, log_wqe_stride_size, log_wqe_stride_size - MLX5_MPWQE_LOG_STRIDE_SZ_BASE); - MLX5_SET(wq, wq, log_wq_sz, mlx5e_mpwqe_get_log_rq_size(params, xsk)); + MLX5_SET(wq, wq, log_wq_sz, mlx5e_mpwqe_get_log_rq_size(mdev, params, xsk)); if (params->packet_merge.type == MLX5E_PACKET_MERGE_SHAMPO) { MLX5_SET(wq, wq, shampo_enable, true); MLX5_SET(wq, wq, log_reservation_size, @@ -712,13 +1013,6 @@ static void mlx5e_build_ico_cq_param(struct mlx5_core_dev *mdev, param->cq_period_mode = DIM_CQ_PERIOD_MODE_START_FROM_EQE; } -static u8 mlx5e_get_rq_log_wq_sz(void *rqc) -{ - void *wq = MLX5_ADDR_OF(rqc, rqc, wq); - - return MLX5_GET(wq, wq, log_wq_sz); -} - /* This function calculates the maximum number of headers entries that are needed * per WQE, the formula is based on the size of the reservations and the * restriction we have about max packets for reservation that is equal to max @@ -779,31 +1073,92 @@ static u32 mlx5e_shampo_icosq_sz(struct mlx5_core_dev *mdev, return wqebbs; } +static u32 mlx5e_mpwrq_total_umr_wqebbs(struct mlx5_core_dev *mdev, + struct mlx5e_params *params, + struct mlx5e_xsk_param *xsk) +{ + enum mlx5e_mpwrq_umr_mode umr_mode = mlx5e_mpwrq_umr_mode(mdev, xsk); + u8 page_shift = mlx5e_mpwrq_page_shift(mdev, xsk); + u8 umr_wqebbs; + + umr_wqebbs = mlx5e_mpwrq_umr_wqebbs(mdev, page_shift, umr_mode); + + return umr_wqebbs * (1 << mlx5e_mpwqe_get_log_rq_size(mdev, params, xsk)); +} + static u8 mlx5e_build_icosq_log_wq_sz(struct mlx5_core_dev *mdev, struct mlx5e_params *params, struct mlx5e_rq_param *rqp) { - u32 wqebbs; + u32 wqebbs, total_pages, useful_space; /* MLX5_WQ_TYPE_CYCLIC */ if (params->rq_wq_type != MLX5_WQ_TYPE_LINKED_LIST_STRIDING_RQ) return MLX5E_PARAMS_MINIMUM_LOG_SQ_SIZE; - wqebbs = MLX5E_UMR_WQEBBS * BIT(mlx5e_get_rq_log_wq_sz(rqp->rqc)); + /* UMR WQEs for the regular RQ. */ + wqebbs = mlx5e_mpwrq_total_umr_wqebbs(mdev, params, NULL); /* If XDP program is attached, XSK may be turned on at any time without * restarting the channel. ICOSQ must be big enough to fit UMR WQEs of * both regular RQ and XSK RQ. - * Although mlx5e_mpwqe_get_log_rq_size accepts mlx5e_xsk_param, it - * doesn't affect its return value, as long as params->xdp_prog != NULL, - * so we can just multiply by 2. + * + * XSK uses different values of page_shift, and the total number of UMR + * WQEBBs depends on it. This dependency is complex and not monotonic, + * especially taking into consideration that some of the parameters come + * from capabilities. Hence, we have to try all valid values of XSK + * frame size (and page_shift) to find the maximum. */ - if (params->xdp_prog) - wqebbs *= 2; + if (params->xdp_prog) { + u32 max_xsk_wqebbs = 0; + u8 frame_shift; + + for (frame_shift = XDP_UMEM_MIN_CHUNK_SHIFT; + frame_shift <= PAGE_SHIFT; frame_shift++) { + /* The headroom doesn't affect the calculation. */ + struct mlx5e_xsk_param xsk = { + .chunk_size = 1 << frame_shift, + .unaligned = false, + }; + + /* XSK aligned mode. */ + max_xsk_wqebbs = max(max_xsk_wqebbs, + mlx5e_mpwrq_total_umr_wqebbs(mdev, params, &xsk)); + + /* XSK unaligned mode, frame size is a power of two. */ + xsk.unaligned = true; + max_xsk_wqebbs = max(max_xsk_wqebbs, + mlx5e_mpwrq_total_umr_wqebbs(mdev, params, &xsk)); + + /* XSK unaligned mode, frame size is not equal to stride size. */ + xsk.chunk_size -= 1; + max_xsk_wqebbs = max(max_xsk_wqebbs, + mlx5e_mpwrq_total_umr_wqebbs(mdev, params, &xsk)); + + /* XSK unaligned mode, frame size is a triple power of two. */ + xsk.chunk_size = (1 << frame_shift) / 4 * 3; + max_xsk_wqebbs = max(max_xsk_wqebbs, + mlx5e_mpwrq_total_umr_wqebbs(mdev, params, &xsk)); + } + + wqebbs += max_xsk_wqebbs; + } if (params->packet_merge.type == MLX5E_PACKET_MERGE_SHAMPO) wqebbs += mlx5e_shampo_icosq_sz(mdev, params, rqp); + /* UMR WQEs don't cross the page boundary, they are padded with NOPs. + * This padding is always smaller than the max WQE size. That gives us + * at least (PAGE_SIZE - (max WQE size - MLX5_SEND_WQE_BB)) useful bytes + * per page. The number of pages is estimated as the total size of WQEs + * divided by the useful space in page, rounding up. If some WQEs don't + * fully fit into the useful space, they can occupy part of the padding, + * which proves this estimation to be correct (reserve enough space). + */ + useful_space = PAGE_SIZE - mlx5e_get_max_sq_wqebbs(mdev) + MLX5_SEND_WQE_BB; + total_pages = DIV_ROUND_UP(wqebbs * MLX5_SEND_WQE_BB, useful_space); + wqebbs = total_pages * (PAGE_SIZE / MLX5_SEND_WQE_BB); + return max_t(u8, MLX5E_PARAMS_MINIMUM_LOG_SQ_SIZE, order_base_2(wqebbs)); } @@ -857,7 +1212,7 @@ void mlx5e_build_xdpsq_param(struct mlx5_core_dev *mdev, mlx5e_build_sq_param_common(mdev, param); MLX5_SET(wq, wq, log_wq_sz, params->log_sq_size); param->is_mpw = MLX5E_GET_PFLAG(params, MLX5E_PFLAG_XDP_TX_MPWQE); - param->is_xdp_mb = !mlx5e_rx_is_linear_skb(params, xsk); + param->is_xdp_mb = !mlx5e_rx_is_linear_skb(mdev, params, xsk); mlx5e_build_tx_cq_param(mdev, params, ¶m->cqp); } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/params.h b/drivers/net/ethernet/mellanox/mlx5/core/en/params.h index f5c46e78eebc..034debd140bc 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/params.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/params.h @@ -9,6 +9,7 @@ struct mlx5e_xsk_param { u16 headroom; u16 chunk_size; + bool unaligned; }; struct mlx5e_cq_param { @@ -52,37 +53,26 @@ struct mlx5e_create_sq_param { u8 min_inline_mode; }; -static inline bool mlx5e_qid_get_ch_if_in_group(struct mlx5e_params *params, - u16 qid, - enum mlx5e_rq_group group, - u16 *ix) -{ - int nch = params->num_channels; - int ch = qid - nch * group; - - if (ch < 0 || ch >= nch) - return false; - - *ix = ch; - return true; -} - -static inline void mlx5e_qid_get_ch_and_group(struct mlx5e_params *params, - u16 qid, - u16 *ix, - enum mlx5e_rq_group *group) -{ - u16 nch = params->num_channels; - - *ix = qid % nch; - *group = qid / nch; -} - -static inline bool mlx5e_qid_validate(const struct mlx5e_profile *profile, - struct mlx5e_params *params, u64 qid) -{ - return qid < params->num_channels * profile->rq_groups; -} +/* Striding RQ dynamic parameters */ + +u8 mlx5e_mpwrq_page_shift(struct mlx5_core_dev *mdev, struct mlx5e_xsk_param *xsk); +enum mlx5e_mpwrq_umr_mode +mlx5e_mpwrq_umr_mode(struct mlx5_core_dev *mdev, struct mlx5e_xsk_param *xsk); +u8 mlx5e_mpwrq_umr_entry_size(enum mlx5e_mpwrq_umr_mode mode); +u8 mlx5e_mpwrq_log_wqe_sz(struct mlx5_core_dev *mdev, u8 page_shift, + enum mlx5e_mpwrq_umr_mode umr_mode); +u8 mlx5e_mpwrq_pages_per_wqe(struct mlx5_core_dev *mdev, u8 page_shift, + enum mlx5e_mpwrq_umr_mode umr_mode); +u16 mlx5e_mpwrq_umr_wqe_sz(struct mlx5_core_dev *mdev, u8 page_shift, + enum mlx5e_mpwrq_umr_mode umr_mode); +u8 mlx5e_mpwrq_umr_wqebbs(struct mlx5_core_dev *mdev, u8 page_shift, + enum mlx5e_mpwrq_umr_mode umr_mode); +u8 mlx5e_mpwrq_mtts_per_wqe(struct mlx5_core_dev *mdev, u8 page_shift, + enum mlx5e_mpwrq_umr_mode umr_mode); +u32 mlx5e_mpwrq_max_num_entries(struct mlx5_core_dev *mdev, + enum mlx5e_mpwrq_umr_mode umr_mode); +u8 mlx5e_mpwrq_max_log_rq_pkts(struct mlx5_core_dev *mdev, u8 page_shift, + enum mlx5e_mpwrq_umr_mode umr_mode); /* Parameter calculations */ @@ -92,25 +82,23 @@ void mlx5e_set_tx_cq_mode_params(struct mlx5e_params *params, u8 cq_period_mode) void mlx5e_set_rx_cq_mode_params(struct mlx5e_params *params, u8 cq_period_mode); bool slow_pci_heuristic(struct mlx5_core_dev *mdev); -bool mlx5e_striding_rq_possible(struct mlx5_core_dev *mdev, struct mlx5e_params *params); +int mlx5e_mpwrq_validate_regular(struct mlx5_core_dev *mdev, struct mlx5e_params *params); +int mlx5e_mpwrq_validate_xsk(struct mlx5_core_dev *mdev, struct mlx5e_params *params, + struct mlx5e_xsk_param *xsk); void mlx5e_build_rq_params(struct mlx5_core_dev *mdev, struct mlx5e_params *params); void mlx5e_set_rq_type(struct mlx5_core_dev *mdev, struct mlx5e_params *params); void mlx5e_init_rq_type_params(struct mlx5_core_dev *mdev, struct mlx5e_params *params); -bool mlx5e_verify_rx_mpwqe_strides(struct mlx5_core_dev *mdev, - u8 log_stride_sz, u8 log_num_strides); u16 mlx5e_get_linear_rq_headroom(struct mlx5e_params *params, struct mlx5e_xsk_param *xsk); -u32 mlx5e_rx_get_min_frag_sz(struct mlx5e_params *params, - struct mlx5e_xsk_param *xsk); -u8 mlx5e_mpwqe_log_pkts_per_wqe(struct mlx5e_params *params, - struct mlx5e_xsk_param *xsk); -bool mlx5e_rx_is_linear_skb(struct mlx5e_params *params, +bool mlx5e_rx_is_linear_skb(struct mlx5_core_dev *mdev, + struct mlx5e_params *params, struct mlx5e_xsk_param *xsk); bool mlx5e_rx_mpwqe_is_linear_skb(struct mlx5_core_dev *mdev, struct mlx5e_params *params, struct mlx5e_xsk_param *xsk); -u8 mlx5e_mpwqe_get_log_rq_size(struct mlx5e_params *params, +u8 mlx5e_mpwqe_get_log_rq_size(struct mlx5_core_dev *mdev, + struct mlx5e_params *params, struct mlx5e_xsk_param *xsk); u8 mlx5e_shampo_get_log_hd_entry_size(struct mlx5_core_dev *mdev, struct mlx5e_params *params); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/ptp.c b/drivers/net/ethernet/mellanox/mlx5/core/en/ptp.c index 903de88bab53..8469e9c38670 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/ptp.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/ptp.c @@ -622,37 +622,39 @@ static int mlx5e_ptp_set_state(struct mlx5e_ptp *c, struct mlx5e_params *params) return bitmap_empty(c->state, MLX5E_PTP_STATE_NUM_STATES) ? -EINVAL : 0; } -static void mlx5e_ptp_rx_unset_fs(struct mlx5e_priv *priv) +static void mlx5e_ptp_rx_unset_fs(struct mlx5e_flow_steering *fs) { - struct mlx5e_ptp_fs *ptp_fs = priv->fs->ptp_fs; + struct mlx5e_ptp_fs *ptp_fs = mlx5e_fs_get_ptp(fs); if (!ptp_fs->valid) return; mlx5e_fs_tt_redirect_del_rule(ptp_fs->l2_rule); - mlx5e_fs_tt_redirect_any_destroy(priv); + mlx5e_fs_tt_redirect_any_destroy(fs); mlx5e_fs_tt_redirect_del_rule(ptp_fs->udp_v6_rule); mlx5e_fs_tt_redirect_del_rule(ptp_fs->udp_v4_rule); - mlx5e_fs_tt_redirect_udp_destroy(priv); + mlx5e_fs_tt_redirect_udp_destroy(fs); ptp_fs->valid = false; } static int mlx5e_ptp_rx_set_fs(struct mlx5e_priv *priv) { u32 tirn = mlx5e_rx_res_get_tirn_ptp(priv->rx_res); - struct mlx5e_ptp_fs *ptp_fs = priv->fs->ptp_fs; + struct mlx5e_flow_steering *fs = priv->fs; struct mlx5_flow_handle *rule; + struct mlx5e_ptp_fs *ptp_fs; int err; + ptp_fs = mlx5e_fs_get_ptp(fs); if (ptp_fs->valid) return 0; - err = mlx5e_fs_tt_redirect_udp_create(priv); + err = mlx5e_fs_tt_redirect_udp_create(fs); if (err) goto out_free; - rule = mlx5e_fs_tt_redirect_udp_add_rule(priv, MLX5_TT_IPV4_UDP, + rule = mlx5e_fs_tt_redirect_udp_add_rule(fs, MLX5_TT_IPV4_UDP, tirn, PTP_EV_PORT); if (IS_ERR(rule)) { err = PTR_ERR(rule); @@ -660,7 +662,7 @@ static int mlx5e_ptp_rx_set_fs(struct mlx5e_priv *priv) } ptp_fs->udp_v4_rule = rule; - rule = mlx5e_fs_tt_redirect_udp_add_rule(priv, MLX5_TT_IPV6_UDP, + rule = mlx5e_fs_tt_redirect_udp_add_rule(fs, MLX5_TT_IPV6_UDP, tirn, PTP_EV_PORT); if (IS_ERR(rule)) { err = PTR_ERR(rule); @@ -668,11 +670,11 @@ static int mlx5e_ptp_rx_set_fs(struct mlx5e_priv *priv) } ptp_fs->udp_v6_rule = rule; - err = mlx5e_fs_tt_redirect_any_create(priv); + err = mlx5e_fs_tt_redirect_any_create(fs); if (err) goto out_destroy_udp_v6_rule; - rule = mlx5e_fs_tt_redirect_any_add_rule(priv, tirn, ETH_P_1588); + rule = mlx5e_fs_tt_redirect_any_add_rule(fs, tirn, ETH_P_1588); if (IS_ERR(rule)) { err = PTR_ERR(rule); goto out_destroy_fs_any; @@ -683,13 +685,13 @@ static int mlx5e_ptp_rx_set_fs(struct mlx5e_priv *priv) return 0; out_destroy_fs_any: - mlx5e_fs_tt_redirect_any_destroy(priv); + mlx5e_fs_tt_redirect_any_destroy(fs); out_destroy_udp_v6_rule: mlx5e_fs_tt_redirect_del_rule(ptp_fs->udp_v6_rule); out_destroy_udp_v4_rule: mlx5e_fs_tt_redirect_del_rule(ptp_fs->udp_v4_rule); out_destroy_fs_udp: - mlx5e_fs_tt_redirect_udp_destroy(priv); + mlx5e_fs_tt_redirect_udp_destroy(fs); out_free: return err; } @@ -723,7 +725,7 @@ int mlx5e_ptp_open(struct mlx5e_priv *priv, struct mlx5e_params *params, if (err) goto err_free; - netif_napi_add(netdev, &c->napi, mlx5e_ptp_napi_poll, 64); + netif_napi_add(netdev, &c->napi, mlx5e_ptp_napi_poll); mlx5e_ptp_build_params(c, cparams, params); @@ -797,29 +799,31 @@ int mlx5e_ptp_get_rqn(struct mlx5e_ptp *c, u32 *rqn) return 0; } -int mlx5e_ptp_alloc_rx_fs(struct mlx5e_priv *priv) +int mlx5e_ptp_alloc_rx_fs(struct mlx5e_flow_steering *fs, + const struct mlx5e_profile *profile) { struct mlx5e_ptp_fs *ptp_fs; - if (!mlx5e_profile_feature_cap(priv->profile, PTP_RX)) + if (!mlx5e_profile_feature_cap(profile, PTP_RX)) return 0; ptp_fs = kzalloc(sizeof(*ptp_fs), GFP_KERNEL); if (!ptp_fs) return -ENOMEM; + mlx5e_fs_set_ptp(fs, ptp_fs); - priv->fs->ptp_fs = ptp_fs; return 0; } -void mlx5e_ptp_free_rx_fs(struct mlx5e_priv *priv) +void mlx5e_ptp_free_rx_fs(struct mlx5e_flow_steering *fs, + const struct mlx5e_profile *profile) { - struct mlx5e_ptp_fs *ptp_fs = priv->fs->ptp_fs; + struct mlx5e_ptp_fs *ptp_fs = mlx5e_fs_get_ptp(fs); - if (!mlx5e_profile_feature_cap(priv->profile, PTP_RX)) + if (!mlx5e_profile_feature_cap(profile, PTP_RX)) return; - mlx5e_ptp_rx_unset_fs(priv); + mlx5e_ptp_rx_unset_fs(fs); kfree(ptp_fs); } @@ -845,6 +849,6 @@ int mlx5e_ptp_rx_manage_fs(struct mlx5e_priv *priv, bool set) netdev_WARN_ONCE(priv->netdev, "Don't try to remove PTP RX-FS rules"); return -EINVAL; } - mlx5e_ptp_rx_unset_fs(priv); + mlx5e_ptp_rx_unset_fs(priv->fs); return 0; } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/ptp.h b/drivers/net/ethernet/mellanox/mlx5/core/en/ptp.h index 92dbbec472ec..5bce554e131a 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/ptp.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/ptp.h @@ -74,8 +74,10 @@ void mlx5e_ptp_close(struct mlx5e_ptp *c); void mlx5e_ptp_activate_channel(struct mlx5e_ptp *c); void mlx5e_ptp_deactivate_channel(struct mlx5e_ptp *c); int mlx5e_ptp_get_rqn(struct mlx5e_ptp *c, u32 *rqn); -int mlx5e_ptp_alloc_rx_fs(struct mlx5e_priv *priv); -void mlx5e_ptp_free_rx_fs(struct mlx5e_priv *priv); +int mlx5e_ptp_alloc_rx_fs(struct mlx5e_flow_steering *fs, + const struct mlx5e_profile *profile); +void mlx5e_ptp_free_rx_fs(struct mlx5e_flow_steering *fs, + const struct mlx5e_profile *profile); int mlx5e_ptp_rx_manage_fs(struct mlx5e_priv *priv, bool set); enum { diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c b/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c index fc366e66d0b0..5f6f95ad6888 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c @@ -134,38 +134,17 @@ out: return err; } -static int mlx5e_rq_to_ready(struct mlx5e_rq *rq, int curr_state) -{ - struct net_device *dev = rq->netdev; - int err; - - err = mlx5e_modify_rq_state(rq, curr_state, MLX5_RQC_STATE_RST); - if (err) { - netdev_err(dev, "Failed to move rq 0x%x to reset\n", rq->rqn); - return err; - } - err = mlx5e_modify_rq_state(rq, MLX5_RQC_STATE_RST, MLX5_RQC_STATE_RDY); - if (err) { - netdev_err(dev, "Failed to move rq 0x%x to ready\n", rq->rqn); - return err; - } - - return 0; -} - static int mlx5e_rx_reporter_err_rq_cqe_recover(void *ctx) { struct mlx5e_rq *rq = ctx; int err; mlx5e_deactivate_rq(rq); - mlx5e_free_rx_descs(rq); - - err = mlx5e_rq_to_ready(rq, MLX5_RQC_STATE_ERR); + err = mlx5e_flush_rq(rq, MLX5_RQC_STATE_ERR); + clear_bit(MLX5E_RQ_STATE_RECOVERING, &rq->state); if (err) - goto out; + return err; - clear_bit(MLX5E_RQ_STATE_RECOVERING, &rq->state); mlx5e_activate_rq(rq); rq->stats->recover++; if (rq->channel) @@ -173,9 +152,6 @@ static int mlx5e_rx_reporter_err_rq_cqe_recover(void *ctx) else mlx5e_trigger_napi_sched(rq->cq.napi); return 0; -out: - clear_bit(MLX5E_RQ_STATE_RECOVERING, &rq->state); - return err; } static int mlx5e_rx_reporter_timeout_recover(void *ctx) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/rx_res.c b/drivers/net/ethernet/mellanox/mlx5/core/en/rx_res.c index 24c32f73040a..e1095bc36543 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/rx_res.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/rx_res.c @@ -24,8 +24,6 @@ struct mlx5e_rx_res { struct { struct mlx5e_rqt direct_rqt; struct mlx5e_tir direct_tir; - struct mlx5e_rqt xsk_rqt; - struct mlx5e_tir xsk_tir; } *channels; struct { @@ -320,48 +318,8 @@ static int mlx5e_rx_res_channels_init(struct mlx5e_rx_res *res) mlx5e_tir_builder_clear(builder); } - if (!(res->features & MLX5E_RX_RES_FEATURE_XSK)) - goto out; - - for (ix = 0; ix < res->max_nch; ix++) { - err = mlx5e_rqt_init_direct(&res->channels[ix].xsk_rqt, - res->mdev, false, res->drop_rqn); - if (err) { - mlx5_core_warn(res->mdev, "Failed to create an XSK RQT: err = %d, ix = %u\n", - err, ix); - goto err_destroy_xsk_rqts; - } - } - - for (ix = 0; ix < res->max_nch; ix++) { - mlx5e_tir_builder_build_rqt(builder, res->mdev->mlx5e_res.hw_objs.td.tdn, - mlx5e_rqt_get_rqtn(&res->channels[ix].xsk_rqt), - inner_ft_support); - mlx5e_tir_builder_build_packet_merge(builder, &res->pkt_merge_param); - mlx5e_tir_builder_build_direct(builder); - - err = mlx5e_tir_init(&res->channels[ix].xsk_tir, builder, res->mdev, true); - if (err) { - mlx5_core_warn(res->mdev, "Failed to create an XSK TIR: err = %d, ix = %u\n", - err, ix); - goto err_destroy_xsk_tirs; - } - - mlx5e_tir_builder_clear(builder); - } - goto out; -err_destroy_xsk_tirs: - while (--ix >= 0) - mlx5e_tir_destroy(&res->channels[ix].xsk_tir); - - ix = res->max_nch; -err_destroy_xsk_rqts: - while (--ix >= 0) - mlx5e_rqt_destroy(&res->channels[ix].xsk_rqt); - - ix = res->max_nch; err_destroy_direct_tirs: while (--ix >= 0) mlx5e_tir_destroy(&res->channels[ix].direct_tir); @@ -420,12 +378,6 @@ static void mlx5e_rx_res_channels_destroy(struct mlx5e_rx_res *res) for (ix = 0; ix < res->max_nch; ix++) { mlx5e_tir_destroy(&res->channels[ix].direct_tir); mlx5e_rqt_destroy(&res->channels[ix].direct_rqt); - - if (!(res->features & MLX5E_RX_RES_FEATURE_XSK)) - continue; - - mlx5e_tir_destroy(&res->channels[ix].xsk_tir); - mlx5e_rqt_destroy(&res->channels[ix].xsk_rqt); } kvfree(res->channels); @@ -491,13 +443,6 @@ u32 mlx5e_rx_res_get_tirn_direct(struct mlx5e_rx_res *res, unsigned int ix) return mlx5e_tir_get_tirn(&res->channels[ix].direct_tir); } -u32 mlx5e_rx_res_get_tirn_xsk(struct mlx5e_rx_res *res, unsigned int ix) -{ - WARN_ON(!(res->features & MLX5E_RX_RES_FEATURE_XSK)); - - return mlx5e_tir_get_tirn(&res->channels[ix].xsk_tir); -} - u32 mlx5e_rx_res_get_tirn_rss(struct mlx5e_rx_res *res, enum mlx5_traffic_types tt) { struct mlx5e_rss *rss = res->rss[0]; @@ -523,56 +468,53 @@ static u32 mlx5e_rx_res_get_rqtn_direct(struct mlx5e_rx_res *res, unsigned int i return mlx5e_rqt_get_rqtn(&res->channels[ix].direct_rqt); } -void mlx5e_rx_res_channels_activate(struct mlx5e_rx_res *res, struct mlx5e_channels *chs) +static void mlx5e_rx_res_channel_activate_direct(struct mlx5e_rx_res *res, + struct mlx5e_channels *chs, + unsigned int ix) { - unsigned int nch, ix; + u32 rqn = res->rss_rqns[ix]; int err; - nch = mlx5e_channels_get_num(chs); - - for (ix = 0; ix < chs->num; ix++) - mlx5e_channels_get_regular_rqn(chs, ix, &res->rss_rqns[ix]); - res->rss_nch = chs->num; + err = mlx5e_rqt_redirect_direct(&res->channels[ix].direct_rqt, rqn); + if (err) + mlx5_core_warn(res->mdev, "Failed to redirect direct RQT %#x to RQ %#x (channel %u): err = %d\n", + mlx5e_rqt_get_rqtn(&res->channels[ix].direct_rqt), + rqn, ix, err); +} - mlx5e_rx_res_rss_enable(res); +static void mlx5e_rx_res_channel_deactivate_direct(struct mlx5e_rx_res *res, + unsigned int ix) +{ + int err; - for (ix = 0; ix < nch; ix++) { - u32 rqn; + err = mlx5e_rqt_redirect_direct(&res->channels[ix].direct_rqt, res->drop_rqn); + if (err) + mlx5_core_warn(res->mdev, "Failed to redirect direct RQT %#x to drop RQ %#x (channel %u): err = %d\n", + mlx5e_rqt_get_rqtn(&res->channels[ix].direct_rqt), + res->drop_rqn, ix, err); +} - mlx5e_channels_get_regular_rqn(chs, ix, &rqn); - err = mlx5e_rqt_redirect_direct(&res->channels[ix].direct_rqt, rqn); - if (err) - mlx5_core_warn(res->mdev, "Failed to redirect direct RQT %#x to RQ %#x (channel %u): err = %d\n", - mlx5e_rqt_get_rqtn(&res->channels[ix].direct_rqt), - rqn, ix, err); +void mlx5e_rx_res_channels_activate(struct mlx5e_rx_res *res, struct mlx5e_channels *chs) +{ + unsigned int nch, ix; + int err; - if (!(res->features & MLX5E_RX_RES_FEATURE_XSK)) - continue; + nch = mlx5e_channels_get_num(chs); - if (!mlx5e_channels_get_xsk_rqn(chs, ix, &rqn)) - rqn = res->drop_rqn; - err = mlx5e_rqt_redirect_direct(&res->channels[ix].xsk_rqt, rqn); - if (err) - mlx5_core_warn(res->mdev, "Failed to redirect XSK RQT %#x to RQ %#x (channel %u): err = %d\n", - mlx5e_rqt_get_rqtn(&res->channels[ix].xsk_rqt), - rqn, ix, err); + for (ix = 0; ix < chs->num; ix++) { + if (mlx5e_channels_is_xsk(chs, ix)) + mlx5e_channels_get_xsk_rqn(chs, ix, &res->rss_rqns[ix]); + else + mlx5e_channels_get_regular_rqn(chs, ix, &res->rss_rqns[ix]); } - for (ix = nch; ix < res->max_nch; ix++) { - err = mlx5e_rqt_redirect_direct(&res->channels[ix].direct_rqt, res->drop_rqn); - if (err) - mlx5_core_warn(res->mdev, "Failed to redirect direct RQT %#x to drop RQ %#x (channel %u): err = %d\n", - mlx5e_rqt_get_rqtn(&res->channels[ix].direct_rqt), - res->drop_rqn, ix, err); + res->rss_nch = chs->num; - if (!(res->features & MLX5E_RX_RES_FEATURE_XSK)) - continue; + mlx5e_rx_res_rss_enable(res); - err = mlx5e_rqt_redirect_direct(&res->channels[ix].xsk_rqt, res->drop_rqn); - if (err) - mlx5_core_warn(res->mdev, "Failed to redirect XSK RQT %#x to drop RQ %#x (channel %u): err = %d\n", - mlx5e_rqt_get_rqtn(&res->channels[ix].xsk_rqt), - res->drop_rqn, ix, err); - } + for (ix = 0; ix < nch; ix++) + mlx5e_rx_res_channel_activate_direct(res, chs, ix); + for (ix = nch; ix < res->max_nch; ix++) + mlx5e_rx_res_channel_deactivate_direct(res, ix); if (res->features & MLX5E_RX_RES_FEATURE_PTP) { u32 rqn; @@ -595,22 +537,8 @@ void mlx5e_rx_res_channels_deactivate(struct mlx5e_rx_res *res) mlx5e_rx_res_rss_disable(res); - for (ix = 0; ix < res->max_nch; ix++) { - err = mlx5e_rqt_redirect_direct(&res->channels[ix].direct_rqt, res->drop_rqn); - if (err) - mlx5_core_warn(res->mdev, "Failed to redirect direct RQT %#x to drop RQ %#x (channel %u): err = %d\n", - mlx5e_rqt_get_rqtn(&res->channels[ix].direct_rqt), - res->drop_rqn, ix, err); - - if (!(res->features & MLX5E_RX_RES_FEATURE_XSK)) - continue; - - err = mlx5e_rqt_redirect_direct(&res->channels[ix].xsk_rqt, res->drop_rqn); - if (err) - mlx5_core_warn(res->mdev, "Failed to redirect XSK RQT %#x to drop RQ %#x (channel %u): err = %d\n", - mlx5e_rqt_get_rqtn(&res->channels[ix].xsk_rqt), - res->drop_rqn, ix, err); - } + for (ix = 0; ix < res->max_nch; ix++) + mlx5e_rx_res_channel_deactivate_direct(res, ix); if (res->features & MLX5E_RX_RES_FEATURE_PTP) { err = mlx5e_rqt_redirect_direct(&res->ptp.rqt, res->drop_rqn); @@ -621,33 +549,17 @@ void mlx5e_rx_res_channels_deactivate(struct mlx5e_rx_res *res) } } -int mlx5e_rx_res_xsk_activate(struct mlx5e_rx_res *res, struct mlx5e_channels *chs, - unsigned int ix) +void mlx5e_rx_res_xsk_update(struct mlx5e_rx_res *res, struct mlx5e_channels *chs, + unsigned int ix, bool xsk) { - u32 rqn; - int err; - - if (!mlx5e_channels_get_xsk_rqn(chs, ix, &rqn)) - return -EINVAL; - - err = mlx5e_rqt_redirect_direct(&res->channels[ix].xsk_rqt, rqn); - if (err) - mlx5_core_warn(res->mdev, "Failed to redirect XSK RQT %#x to XSK RQ %#x (channel %u): err = %d\n", - mlx5e_rqt_get_rqtn(&res->channels[ix].xsk_rqt), - rqn, ix, err); - return err; -} + if (xsk) + mlx5e_channels_get_xsk_rqn(chs, ix, &res->rss_rqns[ix]); + else + mlx5e_channels_get_regular_rqn(chs, ix, &res->rss_rqns[ix]); -int mlx5e_rx_res_xsk_deactivate(struct mlx5e_rx_res *res, unsigned int ix) -{ - int err; + mlx5e_rx_res_rss_enable(res); - err = mlx5e_rqt_redirect_direct(&res->channels[ix].xsk_rqt, res->drop_rqn); - if (err) - mlx5_core_warn(res->mdev, "Failed to redirect XSK RQT %#x to drop RQ %#x (channel %u): err = %d\n", - mlx5e_rqt_get_rqtn(&res->channels[ix].xsk_rqt), - res->drop_rqn, ix, err); - return err; + mlx5e_rx_res_channel_activate_direct(res, chs, ix); } int mlx5e_rx_res_packet_merge_set_param(struct mlx5e_rx_res *res, diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/rx_res.h b/drivers/net/ethernet/mellanox/mlx5/core/en/rx_res.h index b39b20a720e0..5d5f64fab60f 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/rx_res.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/rx_res.h @@ -17,8 +17,7 @@ struct mlx5e_rss_params_hash; enum mlx5e_rx_res_features { MLX5E_RX_RES_FEATURE_INNER_FT = BIT(0), - MLX5E_RX_RES_FEATURE_XSK = BIT(1), - MLX5E_RX_RES_FEATURE_PTP = BIT(2), + MLX5E_RX_RES_FEATURE_PTP = BIT(1), }; /* Setup */ @@ -32,7 +31,6 @@ void mlx5e_rx_res_free(struct mlx5e_rx_res *res); /* TIRN getters for flow steering */ u32 mlx5e_rx_res_get_tirn_direct(struct mlx5e_rx_res *res, unsigned int ix); -u32 mlx5e_rx_res_get_tirn_xsk(struct mlx5e_rx_res *res, unsigned int ix); u32 mlx5e_rx_res_get_tirn_rss(struct mlx5e_rx_res *res, enum mlx5_traffic_types tt); u32 mlx5e_rx_res_get_tirn_rss_inner(struct mlx5e_rx_res *res, enum mlx5_traffic_types tt); u32 mlx5e_rx_res_get_tirn_ptp(struct mlx5e_rx_res *res); @@ -40,9 +38,8 @@ u32 mlx5e_rx_res_get_tirn_ptp(struct mlx5e_rx_res *res); /* Activate/deactivate API */ void mlx5e_rx_res_channels_activate(struct mlx5e_rx_res *res, struct mlx5e_channels *chs); void mlx5e_rx_res_channels_deactivate(struct mlx5e_rx_res *res); -int mlx5e_rx_res_xsk_activate(struct mlx5e_rx_res *res, struct mlx5e_channels *chs, - unsigned int ix); -int mlx5e_rx_res_xsk_deactivate(struct mlx5e_rx_res *res, unsigned int ix); +void mlx5e_rx_res_xsk_update(struct mlx5e_rx_res *res, struct mlx5e_channels *chs, + unsigned int ix, bool xsk); /* Configuration API */ void mlx5e_rx_res_rss_set_indir_uniform(struct mlx5e_rx_res *res, unsigned int nch); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/goto.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/goto.c index 69949ab830b6..25174f68613e 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/goto.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc/act/goto.c @@ -12,6 +12,7 @@ validate_goto_chain(struct mlx5e_priv *priv, const struct flow_action_entry *act, struct netlink_ext_ack *extack) { + struct mlx5e_tc_table *tc = mlx5e_fs_get_tc(priv->fs); bool is_esw = mlx5e_is_eswitch_flow(flow); bool ft_flow = mlx5e_is_ft_flow(flow); u32 dest_chain = act->chain_index; @@ -21,7 +22,7 @@ validate_goto_chain(struct mlx5e_priv *priv, u32 max_chain; esw = priv->mdev->priv.eswitch; - chains = is_esw ? esw_chains(esw) : mlx5e_nic_chains(priv->fs->tc); + chains = is_esw ? esw_chains(esw) : mlx5e_nic_chains(tc); max_chain = mlx5_chains_get_chain_range(chains); reformat_and_fwd = is_esw ? MLX5_CAP_ESW_FLOWTABLE_FDB(priv->mdev, reformat_and_fwd_to_table) : diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/trap.c b/drivers/net/ethernet/mellanox/mlx5/core/en/trap.c index 11f2a7fb72a9..201ac7dd338f 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/trap.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/trap.c @@ -147,7 +147,7 @@ static struct mlx5e_trap *mlx5e_open_trap(struct mlx5e_priv *priv) t->mkey_be = cpu_to_be32(priv->mdev->mlx5e_res.hw_objs.mkey); t->stats = &priv->trap_stats.ch; - netif_napi_add(netdev, &t->napi, mlx5e_trap_napi_poll, 64); + netif_napi_add(netdev, &t->napi, mlx5e_trap_napi_poll); err = mlx5e_open_trap_rq(priv, t); if (unlikely(err)) @@ -230,12 +230,12 @@ static int mlx5e_handle_action_trap(struct mlx5e_priv *priv, int trap_id) switch (trap_id) { case DEVLINK_TRAP_GENERIC_ID_INGRESS_VLAN_FILTER: - err = mlx5e_add_vlan_trap(priv, trap_id, mlx5e_trap_get_tirn(priv->en_trap)); + err = mlx5e_add_vlan_trap(priv->fs, trap_id, mlx5e_trap_get_tirn(priv->en_trap)); if (err) goto err_out; break; case DEVLINK_TRAP_GENERIC_ID_DMAC_FILTER: - err = mlx5e_add_mac_trap(priv, trap_id, mlx5e_trap_get_tirn(priv->en_trap)); + err = mlx5e_add_mac_trap(priv->fs, trap_id, mlx5e_trap_get_tirn(priv->en_trap)); if (err) goto err_out; break; @@ -256,10 +256,10 @@ static int mlx5e_handle_action_drop(struct mlx5e_priv *priv, int trap_id) { switch (trap_id) { case DEVLINK_TRAP_GENERIC_ID_INGRESS_VLAN_FILTER: - mlx5e_remove_vlan_trap(priv); + mlx5e_remove_vlan_trap(priv->fs); break; case DEVLINK_TRAP_GENERIC_ID_DMAC_FILTER: - mlx5e_remove_mac_trap(priv); + mlx5e_remove_mac_trap(priv->fs); break; default: netdev_warn(priv->netdev, "%s: Unknown trap id %d\n", __func__, trap_id); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/txrx.h b/drivers/net/ethernet/mellanox/mlx5/core/en/txrx.h index c208ea307bff..4456ad5cedf1 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/txrx.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/txrx.h @@ -439,16 +439,24 @@ static inline u16 mlx5e_stop_room_for_max_wqe(struct mlx5_core_dev *mdev) return MLX5E_STOP_ROOM(mlx5e_get_max_sq_wqebbs(mdev)); } -static inline bool mlx5e_icosq_can_post_wqe(struct mlx5e_icosq *sq, u16 wqe_size) +static inline u16 mlx5e_stop_room_for_mpwqe(struct mlx5_core_dev *mdev) { - u16 room = sq->reserved_room; + u8 mpwqe_wqebbs = mlx5e_get_max_sq_aligned_wqebbs(mdev); - WARN_ONCE(wqe_size > sq->max_sq_wqebbs, - "wqe_size %u is greater than max SQ WQEBBs %u", - wqe_size, sq->max_sq_wqebbs); + return mlx5e_stop_room_for_wqe(mdev, mpwqe_wqebbs); +} - room += MLX5E_STOP_ROOM(wqe_size); +static inline bool mlx5e_icosq_can_post_wqe(struct mlx5e_icosq *sq, u16 wqe_size) +{ + u16 room = sq->reserved_room + MLX5E_STOP_ROOM(wqe_size); return mlx5e_wqc_has_room_for(&sq->wq, sq->cc, sq->pc, room); } + +static inline struct mlx5e_mpw_info *mlx5e_get_mpw_info(struct mlx5e_rq *rq, int i) +{ + size_t isz = struct_size(rq->mpwqe.info, alloc_units, rq->mpwqe.pages_per_wqe); + + return (struct mlx5e_mpw_info *)((char *)rq->mpwqe.info + array_size(i, isz)); +} #endif diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c b/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c index 8f321a6c0809..4685c652c97e 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c @@ -333,7 +333,7 @@ mlx5e_xmit_xdp_frame_mpwqe(struct mlx5e_xdpsq *sq, struct mlx5e_xmit_data *xdptx mlx5e_xdp_mpwqe_add_dseg(sq, xdptxd, stats); - if (unlikely(mlx5e_xdp_mpqwe_is_full(session, sq->max_sq_mpw_wqebbs))) + if (unlikely(mlx5e_xdp_mpwqe_is_full(session, sq->max_sq_mpw_wqebbs))) mlx5e_xdp_mpwqe_complete(sq); stats->xmit++; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.h b/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.h index 287e17911251..bc2d9034af5b 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/xdp.h @@ -122,7 +122,7 @@ static inline bool mlx5e_xdp_get_inline_state(struct mlx5e_xdpsq *sq, bool cur) return cur; } -static inline bool mlx5e_xdp_mpqwe_is_full(struct mlx5e_tx_mpwqe *session, u8 max_sq_mpw_wqebbs) +static inline bool mlx5e_xdp_mpwqe_is_full(struct mlx5e_tx_mpwqe *session, u8 max_sq_mpw_wqebbs) { if (session->inline_on) return session->ds_count + MLX5E_XDP_INLINE_WQE_MAX_DS_CNT > diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/pool.c b/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/pool.c index 2c520394aa1d..ebada0c5af3c 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/pool.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/pool.c @@ -72,6 +72,7 @@ void mlx5e_build_xsk_param(struct xsk_buff_pool *pool, struct mlx5e_xsk_param *x { xsk->headroom = xsk_pool_get_headroom(pool); xsk->chunk_size = xsk_pool_get_chunk_size(pool); + xsk->unaligned = pool->unaligned; } static int mlx5e_xsk_enable_locked(struct mlx5e_priv *priv, @@ -98,6 +99,15 @@ static int mlx5e_xsk_enable_locked(struct mlx5e_priv *priv, mlx5e_build_xsk_param(pool, &xsk); + if (priv->channels.params.rq_wq_type == MLX5_WQ_TYPE_LINKED_LIST_STRIDING_RQ && + mlx5e_mpwrq_umr_mode(priv->mdev, &xsk) == MLX5E_MPWRQ_UMR_MODE_OVERSIZED) { + const char *recommendation = is_power_of_2(xsk.chunk_size) ? + "Upgrade firmware" : "Disable striding RQ"; + + mlx5_core_warn(priv->mdev, "Expected slowdown with XSK frame size %u. %s for better performance.\n", + xsk.chunk_size, recommendation); + } + if (!test_bit(MLX5E_STATE_OPENED, &priv->state)) { /* XSK objects will be created on open. */ goto validate_closed; @@ -123,15 +133,12 @@ static int mlx5e_xsk_enable_locked(struct mlx5e_priv *priv, * any Fill Ring entries at the setup stage. */ - err = mlx5e_rx_res_xsk_activate(priv->rx_res, &priv->channels, ix); - if (unlikely(err)) - goto err_deactivate; + mlx5e_rx_res_xsk_update(priv->rx_res, &priv->channels, ix, true); - return 0; + mlx5e_deactivate_rq(&c->rq); + mlx5e_flush_rq(&c->rq, MLX5_RQC_STATE_RDY); -err_deactivate: - mlx5e_deactivate_xsk(c); - mlx5e_close_xsk(c); + return 0; err_remove_pool: mlx5e_xsk_remove_pool(&priv->xsk, ix); @@ -170,7 +177,13 @@ static int mlx5e_xsk_disable_locked(struct mlx5e_priv *priv, u16 ix) goto remove_pool; c = priv->channels.c[ix]; - mlx5e_rx_res_xsk_deactivate(priv->rx_res, ix); + + mlx5e_activate_rq(&c->rq); + mlx5e_trigger_napi_icosq(c); + mlx5e_wait_for_min_rx_wqes(&c->rq, MLX5E_RQ_WQES_TIMEOUT); + + mlx5e_rx_res_xsk_update(priv->rx_res, &priv->channels, ix, false); + mlx5e_deactivate_xsk(c); mlx5e_close_xsk(c); @@ -208,11 +221,10 @@ int mlx5e_xsk_setup_pool(struct net_device *dev, struct xsk_buff_pool *pool, u16 { struct mlx5e_priv *priv = netdev_priv(dev); struct mlx5e_params *params = &priv->channels.params; - u16 ix; - if (unlikely(!mlx5e_qid_get_ch_if_in_group(params, qid, MLX5E_RQ_GROUP_XSK, &ix))) + if (unlikely(qid >= params->num_channels)) return -EINVAL; - return pool ? mlx5e_xsk_enable_pool(priv, pool, ix) : - mlx5e_xsk_disable_pool(priv, ix); + return pool ? mlx5e_xsk_enable_pool(priv, pool, qid) : + mlx5e_xsk_disable_pool(priv, qid); } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/rx.c b/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/rx.c index 9a1553598a7c..c91b54d9ff27 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/rx.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/rx.c @@ -8,18 +8,221 @@ /* RX data path */ -static struct sk_buff *mlx5e_xsk_construct_skb(struct mlx5e_rq *rq, void *data, - u32 cqe_bcnt) +int mlx5e_xsk_alloc_rx_mpwqe(struct mlx5e_rq *rq, u16 ix) { + struct mlx5e_mpw_info *wi = mlx5e_get_mpw_info(rq, ix); + struct mlx5e_icosq *icosq = rq->icosq; + struct mlx5_wq_cyc *wq = &icosq->wq; + struct mlx5e_umr_wqe *umr_wqe; + int batch, i; + u32 offset; /* 17-bit value with MTT. */ + u16 pi; + + if (unlikely(!xsk_buff_can_alloc(rq->xsk_pool, rq->mpwqe.pages_per_wqe))) + goto err; + + BUILD_BUG_ON(sizeof(wi->alloc_units[0]) != sizeof(wi->alloc_units[0].xsk)); + batch = xsk_buff_alloc_batch(rq->xsk_pool, (struct xdp_buff **)wi->alloc_units, + rq->mpwqe.pages_per_wqe); + + /* If batch < pages_per_wqe, either: + * 1. Some (or all) descriptors were invalid. + * 2. dma_need_sync is true, and it fell back to allocating one frame. + * In either case, try to continue allocating frames one by one, until + * the first error, which will mean there are no more valid descriptors. + */ + for (; batch < rq->mpwqe.pages_per_wqe; batch++) { + wi->alloc_units[batch].xsk = xsk_buff_alloc(rq->xsk_pool); + if (unlikely(!wi->alloc_units[batch].xsk)) + goto err_reuse_batch; + } + + pi = mlx5e_icosq_get_next_pi(icosq, rq->mpwqe.umr_wqebbs); + umr_wqe = mlx5_wq_cyc_get_wqe(wq, pi); + memcpy(umr_wqe, &rq->mpwqe.umr_wqe, sizeof(struct mlx5e_umr_wqe)); + + if (likely(rq->mpwqe.umr_mode == MLX5E_MPWRQ_UMR_MODE_ALIGNED)) { + for (i = 0; i < batch; i++) { + dma_addr_t addr = xsk_buff_xdp_get_frame_dma(wi->alloc_units[i].xsk); + + umr_wqe->inline_mtts[i] = (struct mlx5_mtt) { + .ptag = cpu_to_be64(addr | MLX5_EN_WR), + }; + } + } else if (unlikely(rq->mpwqe.umr_mode == MLX5E_MPWRQ_UMR_MODE_UNALIGNED)) { + for (i = 0; i < batch; i++) { + dma_addr_t addr = xsk_buff_xdp_get_frame_dma(wi->alloc_units[i].xsk); + + umr_wqe->inline_ksms[i] = (struct mlx5_ksm) { + .key = rq->mkey_be, + .va = cpu_to_be64(addr), + }; + } + } else if (likely(rq->mpwqe.umr_mode == MLX5E_MPWRQ_UMR_MODE_TRIPLE)) { + u32 mapping_size = 1 << (rq->mpwqe.page_shift - 2); + + for (i = 0; i < batch; i++) { + dma_addr_t addr = xsk_buff_xdp_get_frame_dma(wi->alloc_units[i].xsk); + + umr_wqe->inline_ksms[i << 2] = (struct mlx5_ksm) { + .key = rq->mkey_be, + .va = cpu_to_be64(addr), + }; + umr_wqe->inline_ksms[(i << 2) + 1] = (struct mlx5_ksm) { + .key = rq->mkey_be, + .va = cpu_to_be64(addr + mapping_size), + }; + umr_wqe->inline_ksms[(i << 2) + 2] = (struct mlx5_ksm) { + .key = rq->mkey_be, + .va = cpu_to_be64(addr + mapping_size * 2), + }; + umr_wqe->inline_ksms[(i << 2) + 3] = (struct mlx5_ksm) { + .key = rq->mkey_be, + .va = cpu_to_be64(rq->wqe_overflow.addr), + }; + } + } else { + __be32 pad_size = cpu_to_be32((1 << rq->mpwqe.page_shift) - + rq->xsk_pool->chunk_size); + __be32 frame_size = cpu_to_be32(rq->xsk_pool->chunk_size); + + for (i = 0; i < batch; i++) { + dma_addr_t addr = xsk_buff_xdp_get_frame_dma(wi->alloc_units[i].xsk); + + umr_wqe->inline_klms[i << 1] = (struct mlx5_klm) { + .key = rq->mkey_be, + .va = cpu_to_be64(addr), + .bcount = frame_size, + }; + umr_wqe->inline_klms[(i << 1) + 1] = (struct mlx5_klm) { + .key = rq->mkey_be, + .va = cpu_to_be64(rq->wqe_overflow.addr), + .bcount = pad_size, + }; + } + } + + bitmap_zero(wi->xdp_xmit_bitmap, rq->mpwqe.pages_per_wqe); + wi->consumed_strides = 0; + + umr_wqe->ctrl.opmod_idx_opcode = + cpu_to_be32((icosq->pc << MLX5_WQE_CTRL_WQE_INDEX_SHIFT) | MLX5_OPCODE_UMR); + + /* Optimized for speed: keep in sync with mlx5e_mpwrq_umr_entry_size. */ + offset = ix * rq->mpwqe.mtts_per_wqe; + if (likely(rq->mpwqe.umr_mode == MLX5E_MPWRQ_UMR_MODE_ALIGNED)) + offset = offset * sizeof(struct mlx5_mtt) / MLX5_OCTWORD; + else if (unlikely(rq->mpwqe.umr_mode == MLX5E_MPWRQ_UMR_MODE_OVERSIZED)) + offset = offset * sizeof(struct mlx5_klm) * 2 / MLX5_OCTWORD; + else if (unlikely(rq->mpwqe.umr_mode == MLX5E_MPWRQ_UMR_MODE_TRIPLE)) + offset = offset * sizeof(struct mlx5_ksm) * 4 / MLX5_OCTWORD; + umr_wqe->uctrl.xlt_offset = cpu_to_be16(offset); + + icosq->db.wqe_info[pi] = (struct mlx5e_icosq_wqe_info) { + .wqe_type = MLX5E_ICOSQ_WQE_UMR_RX, + .num_wqebbs = rq->mpwqe.umr_wqebbs, + .umr.rq = rq, + }; + + icosq->pc += rq->mpwqe.umr_wqebbs; + + icosq->doorbell_cseg = &umr_wqe->ctrl; + + return 0; + +err_reuse_batch: + while (--batch >= 0) + xsk_buff_free(wi->alloc_units[batch].xsk); + +err: + rq->stats->buff_alloc_err++; + return -ENOMEM; +} + +int mlx5e_xsk_alloc_rx_wqes_batched(struct mlx5e_rq *rq, u16 ix, int wqe_bulk) +{ + struct mlx5_wq_cyc *wq = &rq->wqe.wq; + struct xdp_buff **buffs; + u32 contig, alloc; + int i; + + /* mlx5e_init_frags_partition creates a 1:1 mapping between + * rq->wqe.frags and rq->wqe.alloc_units, which allows us to + * allocate XDP buffers straight into alloc_units. + */ + BUILD_BUG_ON(sizeof(rq->wqe.alloc_units[0]) != + sizeof(rq->wqe.alloc_units[0].xsk)); + buffs = (struct xdp_buff **)rq->wqe.alloc_units; + contig = mlx5_wq_cyc_get_size(wq) - ix; + if (wqe_bulk <= contig) { + alloc = xsk_buff_alloc_batch(rq->xsk_pool, buffs + ix, wqe_bulk); + } else { + alloc = xsk_buff_alloc_batch(rq->xsk_pool, buffs + ix, contig); + if (likely(alloc == contig)) + alloc += xsk_buff_alloc_batch(rq->xsk_pool, buffs, wqe_bulk - contig); + } + + for (i = 0; i < alloc; i++) { + int j = mlx5_wq_cyc_ctr2ix(wq, ix + i); + struct mlx5e_wqe_frag_info *frag; + struct mlx5e_rx_wqe_cyc *wqe; + dma_addr_t addr; + + wqe = mlx5_wq_cyc_get_wqe(wq, j); + /* Assumes log_num_frags == 0. */ + frag = &rq->wqe.frags[j]; + + addr = xsk_buff_xdp_get_frame_dma(frag->au->xsk); + wqe->data[0].addr = cpu_to_be64(addr + rq->buff.headroom); + } + + return alloc; +} + +int mlx5e_xsk_alloc_rx_wqes(struct mlx5e_rq *rq, u16 ix, int wqe_bulk) +{ + struct mlx5_wq_cyc *wq = &rq->wqe.wq; + int i; + + for (i = 0; i < wqe_bulk; i++) { + int j = mlx5_wq_cyc_ctr2ix(wq, ix + i); + struct mlx5e_wqe_frag_info *frag; + struct mlx5e_rx_wqe_cyc *wqe; + dma_addr_t addr; + + wqe = mlx5_wq_cyc_get_wqe(wq, j); + /* Assumes log_num_frags == 0. */ + frag = &rq->wqe.frags[j]; + + frag->au->xsk = xsk_buff_alloc(rq->xsk_pool); + if (unlikely(!frag->au->xsk)) + return i; + + addr = xsk_buff_xdp_get_frame_dma(frag->au->xsk); + wqe->data[0].addr = cpu_to_be64(addr + rq->buff.headroom); + } + + return wqe_bulk; +} + +static struct sk_buff *mlx5e_xsk_construct_skb(struct mlx5e_rq *rq, struct xdp_buff *xdp) +{ + u32 totallen = xdp->data_end - xdp->data_meta; + u32 metalen = xdp->data - xdp->data_meta; struct sk_buff *skb; - skb = napi_alloc_skb(rq->cq.napi, cqe_bcnt); + skb = napi_alloc_skb(rq->cq.napi, totallen); if (unlikely(!skb)) { rq->stats->buff_alloc_err++; return NULL; } - skb_put_data(skb, data, cqe_bcnt); + skb_put_data(skb, xdp->data_meta, totallen); + + if (metalen) { + skb_metadata_set(skb, metalen); + __skb_pull(skb, metalen); + } return skb; } @@ -30,7 +233,7 @@ struct sk_buff *mlx5e_xsk_skb_from_cqe_mpwrq_linear(struct mlx5e_rq *rq, u32 head_offset, u32 page_idx) { - struct xdp_buff *xdp = wi->umr.dma_info[page_idx].xsk; + struct xdp_buff *xdp = wi->alloc_units[page_idx].xsk; struct bpf_prog *prog; /* Check packet size. Note LRO doesn't use linear SKB */ @@ -46,8 +249,7 @@ struct sk_buff *mlx5e_xsk_skb_from_cqe_mpwrq_linear(struct mlx5e_rq *rq, */ WARN_ON_ONCE(head_offset); - xdp->data_end = xdp->data + cqe_bcnt; - xdp_set_data_meta_invalid(xdp); + xsk_buff_set_size(xdp, cqe_bcnt); xsk_buff_dma_sync_for_cpu(xdp, rq->xsk_pool); net_prefetch(xdp->data); @@ -76,14 +278,14 @@ struct sk_buff *mlx5e_xsk_skb_from_cqe_mpwrq_linear(struct mlx5e_rq *rq, /* XDP_PASS: copy the data from the UMEM to a new SKB and reuse the * frame. On SKB allocation failure, NULL is returned. */ - return mlx5e_xsk_construct_skb(rq, xdp->data, xdp->data_end - xdp->data); + return mlx5e_xsk_construct_skb(rq, xdp); } struct sk_buff *mlx5e_xsk_skb_from_cqe_linear(struct mlx5e_rq *rq, struct mlx5e_wqe_frag_info *wi, u32 cqe_bcnt) { - struct xdp_buff *xdp = wi->di->xsk; + struct xdp_buff *xdp = wi->au->xsk; struct bpf_prog *prog; /* wi->offset is not used in this function, because xdp->data and the @@ -93,8 +295,7 @@ struct sk_buff *mlx5e_xsk_skb_from_cqe_linear(struct mlx5e_rq *rq, */ WARN_ON_ONCE(wi->offset); - xdp->data_end = xdp->data + cqe_bcnt; - xdp_set_data_meta_invalid(xdp); + xsk_buff_set_size(xdp, cqe_bcnt); xsk_buff_dma_sync_for_cpu(xdp, rq->xsk_pool); net_prefetch(xdp->data); @@ -103,8 +304,8 @@ struct sk_buff *mlx5e_xsk_skb_from_cqe_linear(struct mlx5e_rq *rq, return NULL; /* page/packet was consumed by XDP */ /* XDP_PASS: copy the data from the UMEM to a new SKB. The frame reuse - * will be handled by mlx5e_put_rx_frag. + * will be handled by mlx5e_free_rx_wqe. * On SKB allocation failure, NULL is returned. */ - return mlx5e_xsk_construct_skb(rq, xdp->data, xdp->data_end - xdp->data); + return mlx5e_xsk_construct_skb(rq, xdp); } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/rx.h b/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/rx.h index cc18d97d8ee0..087c943bd8e9 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/rx.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/rx.h @@ -5,12 +5,12 @@ #define __MLX5_EN_XSK_RX_H__ #include "en.h" -#include <net/xdp_sock_drv.h> - -#define MLX5E_MTT_PTAG_MASK 0xfffffffffffffff8ULL /* RX data path */ +int mlx5e_xsk_alloc_rx_mpwqe(struct mlx5e_rq *rq, u16 ix); +int mlx5e_xsk_alloc_rx_wqes_batched(struct mlx5e_rq *rq, u16 ix, int wqe_bulk); +int mlx5e_xsk_alloc_rx_wqes(struct mlx5e_rq *rq, u16 ix, int wqe_bulk); struct sk_buff *mlx5e_xsk_skb_from_cqe_mpwrq_linear(struct mlx5e_rq *rq, struct mlx5e_mpw_info *wi, u16 cqe_bcnt, @@ -20,46 +20,4 @@ struct sk_buff *mlx5e_xsk_skb_from_cqe_linear(struct mlx5e_rq *rq, struct mlx5e_wqe_frag_info *wi, u32 cqe_bcnt); -static inline int mlx5e_xsk_page_alloc_pool(struct mlx5e_rq *rq, - struct mlx5e_dma_info *dma_info) -{ -retry: - dma_info->xsk = xsk_buff_alloc(rq->xsk_pool); - if (!dma_info->xsk) - return -ENOMEM; - - /* Store the DMA address without headroom. In striding RQ case, we just - * provide pages for UMR, and headroom is counted at the setup stage - * when creating a WQE. In non-striding RQ case, headroom is accounted - * in mlx5e_alloc_rx_wqe. - */ - dma_info->addr = xsk_buff_xdp_get_frame_dma(dma_info->xsk); - - /* MTT page mapping has alignment requirements. If they are not - * satisfied, leak the descriptor so that it won't come again, and try - * to allocate a new one. - */ - if (rq->wq_type == MLX5_WQ_TYPE_LINKED_LIST_STRIDING_RQ) { - if (unlikely(dma_info->addr & ~MLX5E_MTT_PTAG_MASK)) { - xsk_buff_discard(dma_info->xsk); - goto retry; - } - } - - return 0; -} - -static inline bool mlx5e_xsk_update_rx_wakeup(struct mlx5e_rq *rq, bool alloc_err) -{ - if (!xsk_uses_need_wakeup(rq->xsk_pool)) - return alloc_err; - - if (unlikely(alloc_err)) - xsk_set_rx_need_wakeup(rq->xsk_pool); - else - xsk_clear_rx_need_wakeup(rq->xsk_pool); - - return false; -} - #endif /* __MLX5_EN_XSK_RX_H__ */ diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/setup.c b/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/setup.c index 98ed9ef3a6bd..ff03c43833bb 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/setup.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/setup.c @@ -5,24 +5,19 @@ #include "en/params.h" #include "en/txrx.h" #include "en/health.h" +#include <net/xdp_sock_drv.h> -/* It matches XDP_UMEM_MIN_CHUNK_SIZE, but as this constant is private and may - * change unexpectedly, and mlx5e has a minimum valid stride size for striding - * RQ, keep this check in the driver. +/* The limitation of 2048 can be altered, but shouldn't go beyond the minimal + * stride size of striding RQ. */ -#define MLX5E_MIN_XSK_CHUNK_SIZE 2048 +#define MLX5E_MIN_XSK_CHUNK_SIZE max(2048, XDP_UMEM_MIN_CHUNK_SIZE) bool mlx5e_validate_xsk_param(struct mlx5e_params *params, struct mlx5e_xsk_param *xsk, struct mlx5_core_dev *mdev) { /* AF_XDP doesn't support frames larger than PAGE_SIZE. */ - if (xsk->chunk_size > PAGE_SIZE || - xsk->chunk_size < MLX5E_MIN_XSK_CHUNK_SIZE) - return false; - - /* Current MTU and XSK headroom don't allow packets to fit the frames. */ - if (mlx5e_rx_get_min_frag_sz(params, xsk) > xsk->chunk_size) + if (xsk->chunk_size > PAGE_SIZE || xsk->chunk_size < MLX5E_MIN_XSK_CHUNK_SIZE) return false; /* frag_sz is different for regular and XSK RQs, so ensure that linear @@ -30,9 +25,9 @@ bool mlx5e_validate_xsk_param(struct mlx5e_params *params, */ switch (params->rq_wq_type) { case MLX5_WQ_TYPE_LINKED_LIST_STRIDING_RQ: - return mlx5e_rx_mpwqe_is_linear_skb(mdev, params, xsk); + return !mlx5e_mpwrq_validate_xsk(mdev, params, xsk); default: /* MLX5_WQ_TYPE_CYCLIC */ - return mlx5e_rx_is_linear_skb(params, xsk); + return mlx5e_rx_is_linear_skb(mdev, params, xsk); } } @@ -71,7 +66,7 @@ static int mlx5e_init_xsk_rq(struct mlx5e_channel *c, rq->xsk_pool = pool; rq->stats = &c->priv->channel_stats[c->ix]->xskrq; rq->ptp_cyc2time = mlx5_rq_ts_translator(mdev); - rq_xdp_ix = c->ix + params->num_channels * MLX5E_RQ_GROUP_XSK; + rq_xdp_ix = c->ix; err = mlx5e_rq_set_handlers(rq, params, xsk); if (err) return err; @@ -159,7 +154,7 @@ err_free_cparam: void mlx5e_close_xsk(struct mlx5e_channel *c) { clear_bit(MLX5E_CHANNEL_STATE_XSK, c->state); - synchronize_net(); /* Sync with the XSK wakeup and with NAPI. */ + synchronize_net(); /* Sync with NAPI. */ mlx5e_close_rq(&c->xskrq); mlx5e_close_cq(&c->xskrq.cq); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/tx.c b/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/tx.c index 4902ef74fedf..367a9505ca4f 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/tx.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/tx.c @@ -12,18 +12,14 @@ int mlx5e_xsk_wakeup(struct net_device *dev, u32 qid, u32 flags) struct mlx5e_priv *priv = netdev_priv(dev); struct mlx5e_params *params = &priv->channels.params; struct mlx5e_channel *c; - u16 ix; if (unlikely(!mlx5e_xdp_is_active(priv))) return -ENETDOWN; - if (unlikely(!mlx5e_qid_get_ch_if_in_group(params, qid, MLX5E_RQ_GROUP_XSK, &ix))) + if (unlikely(qid >= params->num_channels)) return -EINVAL; - c = priv->channels.c[ix]; - - if (unlikely(!test_bit(MLX5E_CHANNEL_STATE_XSK, c->state))) - return -EINVAL; + c = priv->channels.c[qid]; if (!napi_if_scheduled_mark_missed(&c->napi)) { /* To avoid WQE overrun, don't post a NOP if async_icosq is not @@ -36,9 +32,7 @@ int mlx5e_xsk_wakeup(struct net_device *dev, u32 qid, u32 flags) if (test_and_set_bit(MLX5E_SQ_STATE_PENDING_XSK_TX, &c->async_icosq.state)) return 0; - spin_lock_bh(&c->async_icosq_lock); - mlx5e_trigger_irq(&c->async_icosq); - spin_unlock_bh(&c->async_icosq_lock); + mlx5e_trigger_napi_icosq(c); } return 0; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/tx.h b/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/tx.h index a05085035f23..9c505158b975 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/tx.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/tx.h @@ -5,7 +5,6 @@ #define __MLX5_EN_XSK_TX_H__ #include "en.h" -#include <net/xdp_sock_drv.h> /* TX data path */ @@ -13,15 +12,4 @@ int mlx5e_xsk_wakeup(struct net_device *dev, u32 qid, u32 flags); bool mlx5e_xsk_tx(struct mlx5e_xdpsq *sq, unsigned int budget); -static inline void mlx5e_xsk_update_tx_wakeup(struct mlx5e_xdpsq *sq) -{ - if (!xsk_uses_need_wakeup(sq->xsk_pool)) - return; - - if (sq->pc != sq->cc) - xsk_clear_tx_need_wakeup(sq->xsk_pool); - else - xsk_set_tx_need_wakeup(sq->xsk_pool); -} - #endif /* __MLX5_EN_XSK_TX_H__ */ diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/en_accel.h b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/en_accel.h index 1839f1ab1ddd..07187028f0d3 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/en_accel.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/en_accel.h @@ -39,6 +39,7 @@ #include "en_accel/ipsec_rxtx.h" #include "en_accel/ktls.h" #include "en_accel/ktls_txrx.h" +#include <en_accel/macsec.h> #include "en.h" #include "en/txrx.h" @@ -137,6 +138,15 @@ static inline bool mlx5e_accel_tx_begin(struct net_device *dev, } #endif +#ifdef CONFIG_MLX5_EN_MACSEC + if (unlikely(mlx5e_macsec_skb_is_offload(skb))) { + struct mlx5e_priv *priv = netdev_priv(dev); + + if (unlikely(!mlx5e_macsec_handle_tx_skb(priv->macsec, skb))) + return false; + } +#endif + return true; } @@ -163,6 +173,11 @@ static inline void mlx5e_accel_tx_eseg(struct mlx5e_priv *priv, mlx5e_ipsec_tx_build_eseg(priv, skb, eseg); #endif +#ifdef CONFIG_MLX5_EN_MACSEC + if (unlikely(mlx5e_macsec_skb_is_offload(skb))) + mlx5e_macsec_tx_build_eseg(priv->macsec, skb, eseg); +#endif + #if IS_ENABLED(CONFIG_GENEVE) if (skb->encapsulation && skb->ip_summed == CHECKSUM_PARTIAL) mlx5e_tx_tunnel_accel(skb, eseg, ihs); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/fs_tcp.c b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/fs_tcp.c index 20a4f1e585af..285d32d2fd08 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/fs_tcp.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/fs_tcp.c @@ -1,7 +1,7 @@ // SPDX-License-Identifier: GPL-2.0 OR Linux-OpenIB /* Copyright (c) 2020, Mellanox Technologies inc. All rights reserved. */ -#include <linux/netdevice.h> +#include <mlx5_core.h> #include "en_accel/fs_tcp.h" #include "fs_core.h" @@ -71,13 +71,13 @@ void mlx5e_accel_fs_del_sk(struct mlx5_flow_handle *rule) mlx5_del_flow_rules(rule); } -struct mlx5_flow_handle *mlx5e_accel_fs_add_sk(struct mlx5e_priv *priv, +struct mlx5_flow_handle *mlx5e_accel_fs_add_sk(struct mlx5e_flow_steering *fs, struct sock *sk, u32 tirn, uint32_t flow_tag) { + struct mlx5e_accel_fs_tcp *fs_tcp = mlx5e_fs_get_accel_tcp(fs); struct mlx5_flow_destination dest = {}; struct mlx5e_flow_table *ft = NULL; - struct mlx5e_accel_fs_tcp *fs_tcp; MLX5_DECLARE_FLOW_ACT(flow_act); struct mlx5_flow_handle *flow; struct mlx5_flow_spec *spec; @@ -86,19 +86,17 @@ struct mlx5_flow_handle *mlx5e_accel_fs_add_sk(struct mlx5e_priv *priv, if (!spec) return ERR_PTR(-ENOMEM); - fs_tcp = priv->fs->accel_tcp; - spec->match_criteria_enable = MLX5_MATCH_OUTER_HEADERS; switch (sk->sk_family) { case AF_INET: accel_fs_tcp_set_ipv4_flow(spec, sk); ft = &fs_tcp->tables[ACCEL_FS_IPV4_TCP]; - mlx5e_dbg(HW, priv, "%s flow is %pI4:%d -> %pI4:%d\n", __func__, - &inet_sk(sk)->inet_rcv_saddr, - inet_sk(sk)->inet_sport, - &inet_sk(sk)->inet_daddr, - inet_sk(sk)->inet_dport); + fs_dbg(fs, "%s flow is %pI4:%d -> %pI4:%d\n", __func__, + &inet_sk(sk)->inet_rcv_saddr, + inet_sk(sk)->inet_sport, + &inet_sk(sk)->inet_daddr, + inet_sk(sk)->inet_dport); break; #if IS_ENABLED(CONFIG_IPV6) case AF_INET6: @@ -140,34 +138,32 @@ struct mlx5_flow_handle *mlx5e_accel_fs_add_sk(struct mlx5e_priv *priv, flow = mlx5_add_flow_rules(ft->t, spec, &flow_act, &dest, 1); if (IS_ERR(flow)) - netdev_err(priv->netdev, "mlx5_add_flow_rules() failed, flow is %ld\n", - PTR_ERR(flow)); + fs_err(fs, "mlx5_add_flow_rules() failed, flow is %ld\n", PTR_ERR(flow)); out: kvfree(spec); return flow; } -static int accel_fs_tcp_add_default_rule(struct mlx5e_priv *priv, +static int accel_fs_tcp_add_default_rule(struct mlx5e_flow_steering *fs, enum accel_fs_tcp_type type) { + struct mlx5e_accel_fs_tcp *fs_tcp = mlx5e_fs_get_accel_tcp(fs); + struct mlx5_ttc_table *ttc = mlx5e_fs_get_ttc(fs, false); struct mlx5e_flow_table *accel_fs_t; struct mlx5_flow_destination dest; - struct mlx5e_accel_fs_tcp *fs_tcp; MLX5_DECLARE_FLOW_ACT(flow_act); struct mlx5_flow_handle *rule; int err = 0; - fs_tcp = priv->fs->accel_tcp; accel_fs_t = &fs_tcp->tables[type]; - dest = mlx5_ttc_get_default_dest(priv->fs->ttc, fs_accel2tt(type)); + dest = mlx5_ttc_get_default_dest(ttc, fs_accel2tt(type)); rule = mlx5_add_flow_rules(accel_fs_t->t, NULL, &flow_act, &dest, 1); if (IS_ERR(rule)) { err = PTR_ERR(rule); - netdev_err(priv->netdev, - "%s: add default rule failed, accel_fs type=%d, err %d\n", - __func__, type, err); + fs_err(fs, "%s: add default rule failed, accel_fs type=%d, err %d\n", + __func__, type, err); return err; } @@ -265,9 +261,11 @@ out: return err; } -static int accel_fs_tcp_create_table(struct mlx5e_priv *priv, enum accel_fs_tcp_type type) +static int accel_fs_tcp_create_table(struct mlx5e_flow_steering *fs, enum accel_fs_tcp_type type) { - struct mlx5e_flow_table *ft = &priv->fs->accel_tcp->tables[type]; + struct mlx5e_accel_fs_tcp *accel_tcp = mlx5e_fs_get_accel_tcp(fs); + struct mlx5_flow_namespace *ns = mlx5e_fs_get_ns(fs, false); + struct mlx5e_flow_table *ft = &accel_tcp->tables[type]; struct mlx5_flow_table_attr ft_attr = {}; int err; @@ -277,21 +275,21 @@ static int accel_fs_tcp_create_table(struct mlx5e_priv *priv, enum accel_fs_tcp_ ft_attr.level = MLX5E_ACCEL_FS_TCP_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(ns, &ft_attr); if (IS_ERR(ft->t)) { err = PTR_ERR(ft->t); ft->t = NULL; return err; } - netdev_dbg(priv->netdev, "Created fs accel table id %u level %u\n", - ft->t->id, ft->t->level); + fs_dbg(fs, "Created fs accel table id %u level %u\n", + ft->t->id, ft->t->level); err = accel_fs_tcp_create_groups(ft, type); if (err) goto err; - err = accel_fs_tcp_add_default_rule(priv, type); + err = accel_fs_tcp_add_default_rule(fs, type); if (err) goto err; @@ -301,17 +299,18 @@ err: return err; } -static int accel_fs_tcp_disable(struct mlx5e_priv *priv) +static int accel_fs_tcp_disable(struct mlx5e_flow_steering *fs) { + struct mlx5_ttc_table *ttc = mlx5e_fs_get_ttc(fs, false); int err, i; for (i = 0; i < ACCEL_FS_TCP_NUM_TYPES; i++) { /* Modify ttc rules destination to point back to the indir TIRs */ - err = mlx5_ttc_fwd_default_dest(priv->fs->ttc, fs_accel2tt(i)); + err = mlx5_ttc_fwd_default_dest(ttc, fs_accel2tt(i)); if (err) { - netdev_err(priv->netdev, - "%s: modify ttc[%d] default destination failed, err(%d)\n", - __func__, fs_accel2tt(i), err); + fs_err(fs, + "%s: modify ttc[%d] default destination failed, err(%d)\n", + __func__, fs_accel2tt(i), err); return err; } } @@ -319,32 +318,32 @@ static int accel_fs_tcp_disable(struct mlx5e_priv *priv) return 0; } -static int accel_fs_tcp_enable(struct mlx5e_priv *priv) +static int accel_fs_tcp_enable(struct mlx5e_flow_steering *fs) { + struct mlx5e_accel_fs_tcp *accel_tcp = mlx5e_fs_get_accel_tcp(fs); + struct mlx5_ttc_table *ttc = mlx5e_fs_get_ttc(fs, false); struct mlx5_flow_destination dest = {}; int err, i; dest.type = MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE; for (i = 0; i < ACCEL_FS_TCP_NUM_TYPES; i++) { - dest.ft = priv->fs->accel_tcp->tables[i].t; + dest.ft = accel_tcp->tables[i].t; /* Modify ttc rules destination to point on the accel_fs FTs */ - err = mlx5_ttc_fwd_dest(priv->fs->ttc, fs_accel2tt(i), &dest); + err = mlx5_ttc_fwd_dest(ttc, fs_accel2tt(i), &dest); if (err) { - netdev_err(priv->netdev, - "%s: modify ttc[%d] destination to accel failed, err(%d)\n", - __func__, fs_accel2tt(i), err); + fs_err(fs, "%s: modify ttc[%d] destination to accel failed, err(%d)\n", + __func__, fs_accel2tt(i), err); return err; } } return 0; } -static void accel_fs_tcp_destroy_table(struct mlx5e_priv *priv, int i) +static void accel_fs_tcp_destroy_table(struct mlx5e_flow_steering *fs, int i) { - struct mlx5e_accel_fs_tcp *fs_tcp; + struct mlx5e_accel_fs_tcp *fs_tcp = mlx5e_fs_get_accel_tcp(fs); - fs_tcp = priv->fs->accel_tcp; if (IS_ERR_OR_NULL(fs_tcp->tables[i].t)) return; @@ -353,40 +352,43 @@ static void accel_fs_tcp_destroy_table(struct mlx5e_priv *priv, int i) fs_tcp->tables[i].t = NULL; } -void mlx5e_accel_fs_tcp_destroy(struct mlx5e_priv *priv) +void mlx5e_accel_fs_tcp_destroy(struct mlx5e_flow_steering *fs) { + struct mlx5e_accel_fs_tcp *accel_tcp = mlx5e_fs_get_accel_tcp(fs); int i; - if (!priv->fs->accel_tcp) + if (!accel_tcp) return; - accel_fs_tcp_disable(priv); + accel_fs_tcp_disable(fs); for (i = 0; i < ACCEL_FS_TCP_NUM_TYPES; i++) - accel_fs_tcp_destroy_table(priv, i); + accel_fs_tcp_destroy_table(fs, i); - kfree(priv->fs->accel_tcp); - priv->fs->accel_tcp = NULL; + kfree(accel_tcp); + mlx5e_fs_set_accel_tcp(fs, NULL); } -int mlx5e_accel_fs_tcp_create(struct mlx5e_priv *priv) +int mlx5e_accel_fs_tcp_create(struct mlx5e_flow_steering *fs) { + struct mlx5e_accel_fs_tcp *accel_tcp; int i, err; - if (!MLX5_CAP_FLOWTABLE_NIC_RX(priv->mdev, ft_field_support.outer_ip_version)) + if (!MLX5_CAP_FLOWTABLE_NIC_RX(mlx5e_fs_get_mdev(fs), ft_field_support.outer_ip_version)) return -EOPNOTSUPP; - priv->fs->accel_tcp = kzalloc(sizeof(*priv->fs->accel_tcp), GFP_KERNEL); - if (!priv->fs->accel_tcp) + accel_tcp = kvzalloc(sizeof(*accel_tcp), GFP_KERNEL); + if (!accel_tcp) return -ENOMEM; + mlx5e_fs_set_accel_tcp(fs, accel_tcp); for (i = 0; i < ACCEL_FS_TCP_NUM_TYPES; i++) { - err = accel_fs_tcp_create_table(priv, i); + err = accel_fs_tcp_create_table(fs, i); if (err) goto err_destroy_tables; } - err = accel_fs_tcp_enable(priv); + err = accel_fs_tcp_enable(fs); if (err) goto err_destroy_tables; @@ -394,9 +396,8 @@ int mlx5e_accel_fs_tcp_create(struct mlx5e_priv *priv) err_destroy_tables: while (--i >= 0) - accel_fs_tcp_destroy_table(priv, i); - - kfree(priv->fs->accel_tcp); - priv->fs->accel_tcp = NULL; + accel_fs_tcp_destroy_table(fs, i); + kfree(accel_tcp); + mlx5e_fs_set_accel_tcp(fs, NULL); return err; } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/fs_tcp.h b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/fs_tcp.h index 589235824543..a032bff482a6 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/fs_tcp.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/fs_tcp.h @@ -4,19 +4,19 @@ #ifndef __MLX5E_ACCEL_FS_TCP_H__ #define __MLX5E_ACCEL_FS_TCP_H__ -#include "en.h" +#include "en/fs.h" #ifdef CONFIG_MLX5_EN_TLS -int mlx5e_accel_fs_tcp_create(struct mlx5e_priv *priv); -void mlx5e_accel_fs_tcp_destroy(struct mlx5e_priv *priv); -struct mlx5_flow_handle *mlx5e_accel_fs_add_sk(struct mlx5e_priv *priv, +int mlx5e_accel_fs_tcp_create(struct mlx5e_flow_steering *fs); +void mlx5e_accel_fs_tcp_destroy(struct mlx5e_flow_steering *fs); +struct mlx5_flow_handle *mlx5e_accel_fs_add_sk(struct mlx5e_flow_steering *fs, struct sock *sk, u32 tirn, uint32_t flow_tag); void mlx5e_accel_fs_del_sk(struct mlx5_flow_handle *rule); #else -static inline int mlx5e_accel_fs_tcp_create(struct mlx5e_priv *priv) { return 0; } -static inline void mlx5e_accel_fs_tcp_destroy(struct mlx5e_priv *priv) {} -static inline struct mlx5_flow_handle *mlx5e_accel_fs_add_sk(struct mlx5e_priv *priv, +static inline int mlx5e_accel_fs_tcp_create(struct mlx5e_flow_steering *fs) { return 0; } +static inline void mlx5e_accel_fs_tcp_destroy(struct mlx5e_flow_steering *fs) {} +static inline struct mlx5_flow_handle *mlx5e_accel_fs_add_sk(struct mlx5e_flow_steering *fs, struct sock *sk, u32 tirn, uint32_t flow_tag) { return ERR_PTR(-EOPNOTSUPP); } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec_fs.c b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec_fs.c index f8113fd23265..b859e4a4c744 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec_fs.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec_fs.c @@ -174,6 +174,8 @@ static void rx_destroy(struct mlx5e_priv *priv, enum accel_fs_esp_type type) static int rx_create(struct mlx5e_priv *priv, enum accel_fs_esp_type type) { + struct mlx5_flow_namespace *ns = mlx5e_fs_get_ns(priv->fs, false); + struct mlx5_ttc_table *ttc = mlx5e_fs_get_ttc(priv->fs, false); struct mlx5_flow_table_attr ft_attr = {}; struct mlx5e_accel_fs_esp_prot *fs_prot; struct mlx5e_accel_fs_esp *accel_esp; @@ -182,15 +184,14 @@ static int rx_create(struct mlx5e_priv *priv, enum accel_fs_esp_type type) accel_esp = priv->ipsec->rx_fs; fs_prot = &accel_esp->fs_prot[type]; - fs_prot->default_dest = - mlx5_ttc_get_default_dest(priv->fs->ttc, fs_esp2tt(type)); + mlx5_ttc_get_default_dest(ttc, fs_esp2tt(type)); ft_attr.max_fte = 1; ft_attr.autogroup.max_num_groups = 1; ft_attr.level = MLX5E_ACCEL_FS_ESP_FT_ERR_LEVEL; ft_attr.prio = MLX5E_NIC_PRIO; - ft = mlx5_create_auto_grouped_flow_table(priv->fs->ns, &ft_attr); + ft = mlx5_create_auto_grouped_flow_table(ns, &ft_attr); if (IS_ERR(ft)) return PTR_ERR(ft); @@ -205,7 +206,7 @@ static int rx_create(struct mlx5e_priv *priv, enum accel_fs_esp_type type) ft_attr.prio = MLX5E_NIC_PRIO; ft_attr.autogroup.num_reserved_entries = 1; ft_attr.autogroup.max_num_groups = 1; - ft = mlx5_create_auto_grouped_flow_table(priv->fs->ns, &ft_attr); + ft = mlx5_create_auto_grouped_flow_table(ns, &ft_attr); if (IS_ERR(ft)) { err = PTR_ERR(ft); goto err_fs_ft; @@ -230,6 +231,7 @@ err_add: static int rx_ft_get(struct mlx5e_priv *priv, enum accel_fs_esp_type type) { + struct mlx5_ttc_table *ttc = mlx5e_fs_get_ttc(priv->fs, false); struct mlx5e_accel_fs_esp_prot *fs_prot; struct mlx5_flow_destination dest = {}; struct mlx5e_accel_fs_esp *accel_esp; @@ -249,7 +251,7 @@ static int rx_ft_get(struct mlx5e_priv *priv, enum accel_fs_esp_type type) /* connect */ dest.type = MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE; dest.ft = fs_prot->ft; - mlx5_ttc_fwd_dest(priv->fs->ttc, fs_esp2tt(type), &dest); + mlx5_ttc_fwd_dest(ttc, fs_esp2tt(type), &dest); skip: fs_prot->refcnt++; @@ -260,6 +262,7 @@ out: static void rx_ft_put(struct mlx5e_priv *priv, enum accel_fs_esp_type type) { + struct mlx5_ttc_table *ttc = mlx5e_fs_get_ttc(priv->fs, false); struct mlx5e_accel_fs_esp_prot *fs_prot; struct mlx5e_accel_fs_esp *accel_esp; @@ -271,7 +274,7 @@ static void rx_ft_put(struct mlx5e_priv *priv, enum accel_fs_esp_type type) goto out; /* disconnect */ - mlx5_ttc_fwd_default_dest(priv->fs->ttc, fs_esp2tt(type)); + mlx5_ttc_fwd_default_dest(ttc, fs_esp2tt(type)); /* remove FT */ rx_destroy(priv, type); @@ -385,7 +388,8 @@ static void setup_fte_common(struct mlx5_accel_esp_xfrm_attrs *attrs, 0xff, 16); } - flow_act->ipsec_obj_id = ipsec_obj_id; + flow_act->crypto.type = MLX5_FLOW_CONTEXT_ENCRYPT_DECRYPT_TYPE_IPSEC; + flow_act->crypto.obj_id = ipsec_obj_id; flow_act->flags |= FLOW_ACT_NO_APPEND; } @@ -441,7 +445,7 @@ static int rx_add_rule(struct mlx5e_priv *priv, } flow_act.action = MLX5_FLOW_CONTEXT_ACTION_FWD_DEST | - MLX5_FLOW_CONTEXT_ACTION_IPSEC_DECRYPT | + MLX5_FLOW_CONTEXT_ACTION_CRYPTO_DECRYPT | MLX5_FLOW_CONTEXT_ACTION_MOD_HDR; dest.type = MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE; flow_act.modify_hdr = modify_hdr; @@ -497,7 +501,7 @@ static int tx_add_rule(struct mlx5e_priv *priv, MLX5_ETH_WQE_FT_META_IPSEC); flow_act.action = MLX5_FLOW_CONTEXT_ACTION_ALLOW | - MLX5_FLOW_CONTEXT_ACTION_IPSEC_ENCRYPT; + MLX5_FLOW_CONTEXT_ACTION_CRYPTO_ENCRYPT; rule = mlx5_add_flow_rules(priv->ipsec->tx_fs->ft, spec, &flow_act, NULL, 0); if (IS_ERR(rule)) { err = PTR_ERR(rule); @@ -573,7 +577,7 @@ int mlx5e_accel_ipsec_fs_init(struct mlx5e_ipsec *ipsec) int err = -ENOMEM; ns = mlx5_get_flow_namespace(ipsec->mdev, - MLX5_FLOW_NAMESPACE_EGRESS_KERNEL); + MLX5_FLOW_NAMESPACE_EGRESS_IPSEC); if (!ns) return -EOPNOTSUPP; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec_rxtx.h b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec_rxtx.h index 0ae4e12ce528..1878a70b9031 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec_rxtx.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec_rxtx.h @@ -39,9 +39,9 @@ #include "en.h" #include "en/txrx.h" -/* Bit31: IPsec marker, Bit30-24: IPsec syndrome, Bit23-0: IPsec obj id */ +/* Bit31: IPsec marker, Bit30: reserved, Bit29-24: IPsec syndrome, Bit23-0: IPsec obj id */ #define MLX5_IPSEC_METADATA_MARKER(metadata) (((metadata) >> 31) & 0x1) -#define MLX5_IPSEC_METADATA_SYNDROM(metadata) (((metadata) >> 24) & GENMASK(6, 0)) +#define MLX5_IPSEC_METADATA_SYNDROM(metadata) (((metadata) >> 24) & GENMASK(5, 0)) #define MLX5_IPSEC_METADATA_HANDLE(metadata) ((metadata) & GENMASK(23, 0)) struct mlx5e_accel_tx_ipsec_state { @@ -77,11 +77,6 @@ static inline bool mlx5_ipsec_is_rx_flow(struct mlx5_cqe64 *cqe) return MLX5_IPSEC_METADATA_MARKER(be32_to_cpu(cqe->ft_metadata)); } -static inline bool mlx5e_ipsec_is_tx_flow(struct mlx5e_accel_tx_ipsec_state *ipsec_st) -{ - return ipsec_st->x; -} - static inline bool mlx5e_ipsec_eseg_meta(struct mlx5_wqe_eth_seg *eseg) { return eseg->flow_table_metadata & cpu_to_be32(MLX5_ETH_WQE_FT_META_IPSEC); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ktls.c b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ktls.c index 30a70d139046..da2184c94203 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ktls.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ktls.c @@ -92,6 +92,24 @@ static const struct tlsdev_ops mlx5e_ktls_ops = { .tls_dev_resync = mlx5e_ktls_resync, }; +bool mlx5e_is_ktls_rx(struct mlx5_core_dev *mdev) +{ + u8 max_sq_wqebbs = mlx5e_get_max_sq_wqebbs(mdev); + + if (is_kdump_kernel() || !MLX5_CAP_GEN(mdev, tls_rx)) + return false; + + /* Check the possibility to post the required ICOSQ WQEs. */ + if (WARN_ON_ONCE(max_sq_wqebbs < MLX5E_TLS_SET_STATIC_PARAMS_WQEBBS)) + return false; + if (WARN_ON_ONCE(max_sq_wqebbs < MLX5E_TLS_SET_PROGRESS_PARAMS_WQEBBS)) + return false; + if (WARN_ON_ONCE(max_sq_wqebbs < MLX5E_KTLS_GET_PROGRESS_WQEBBS)) + return false; + + return true; +} + void mlx5e_ktls_build_netdev(struct mlx5e_priv *priv) { struct net_device *netdev = priv->netdev; @@ -118,9 +136,9 @@ int mlx5e_ktls_set_feature_rx(struct net_device *netdev, bool enable) mutex_lock(&priv->state_lock); if (enable) - err = mlx5e_accel_fs_tcp_create(priv); + err = mlx5e_accel_fs_tcp_create(priv->fs); else - mlx5e_accel_fs_tcp_destroy(priv); + mlx5e_accel_fs_tcp_destroy(priv->fs); mutex_unlock(&priv->state_lock); return err; @@ -138,7 +156,7 @@ int mlx5e_ktls_init_rx(struct mlx5e_priv *priv) return -ENOMEM; if (priv->netdev->features & NETIF_F_HW_TLS_RX) { - err = mlx5e_accel_fs_tcp_create(priv); + err = mlx5e_accel_fs_tcp_create(priv->fs); if (err) { destroy_workqueue(priv->tls->rx_wq); return err; @@ -154,7 +172,7 @@ void mlx5e_ktls_cleanup_rx(struct mlx5e_priv *priv) return; if (priv->netdev->features & NETIF_F_HW_TLS_RX) - mlx5e_accel_fs_tcp_destroy(priv); + mlx5e_accel_fs_tcp_destroy(priv->fs); destroy_workqueue(priv->tls->rx_wq); } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ktls.h b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ktls.h index 948400dee525..1c35045e41fb 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ktls.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ktls.h @@ -25,7 +25,8 @@ static inline bool mlx5e_is_ktls_device(struct mlx5_core_dev *mdev) if (!MLX5_CAP_GEN(mdev, log_max_dek)) return false; - return MLX5_CAP_TLS(mdev, tls_1_2_aes_gcm_128); + return (MLX5_CAP_TLS(mdev, tls_1_2_aes_gcm_128) || + MLX5_CAP_TLS(mdev, tls_1_2_aes_gcm_256)); } static inline bool mlx5e_ktls_type_check(struct mlx5_core_dev *mdev, @@ -36,6 +37,10 @@ static inline bool mlx5e_ktls_type_check(struct mlx5_core_dev *mdev, if (crypto_info->version == TLS_1_2_VERSION) return MLX5_CAP_TLS(mdev, tls_1_2_aes_gcm_128); break; + case TLS_CIPHER_AES_GCM_256: + if (crypto_info->version == TLS_1_2_VERSION) + return MLX5_CAP_TLS(mdev, tls_1_2_aes_gcm_256); + break; } return false; @@ -56,10 +61,7 @@ static inline bool mlx5e_is_ktls_tx(struct mlx5_core_dev *mdev) return !is_kdump_kernel() && MLX5_CAP_GEN(mdev, tls_tx); } -static inline bool mlx5e_is_ktls_rx(struct mlx5_core_dev *mdev) -{ - return !is_kdump_kernel() && MLX5_CAP_GEN(mdev, tls_rx); -} +bool mlx5e_is_ktls_rx(struct mlx5_core_dev *mdev); struct mlx5e_tls_sw_stats { atomic64_t tx_tls_ctx; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ktls_rx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ktls_rx.c index 27483aa7be8a..3e54834747ce 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ktls_rx.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ktls_rx.c @@ -43,7 +43,7 @@ struct mlx5e_ktls_rx_resync_ctx { }; struct mlx5e_ktls_offload_context_rx { - struct tls12_crypto_info_aes_gcm_128 crypto_info; + union mlx5e_crypto_info crypto_info; struct accel_rule rule; struct sock *sk; struct mlx5e_rq_stats *rq_stats; @@ -111,7 +111,7 @@ static void accel_rule_handle_work(struct work_struct *work) if (unlikely(test_bit(MLX5E_PRIV_RX_FLAG_DELETING, priv_rx->flags))) goto out; - rule = mlx5e_accel_fs_add_sk(accel_rule->priv, priv_rx->sk, + rule = mlx5e_accel_fs_add_sk(accel_rule->priv->fs, priv_rx->sk, mlx5e_tir_get_tirn(&priv_rx->tir), MLX5_FS_DEFAULT_FLOW_TAG); if (!IS_ERR_OR_NULL(rule)) @@ -362,7 +362,6 @@ static void resync_init(struct mlx5e_ktls_rx_resync_ctx *resync, static void resync_handle_seq_match(struct mlx5e_ktls_offload_context_rx *priv_rx, struct mlx5e_channel *c) { - struct tls12_crypto_info_aes_gcm_128 *info = &priv_rx->crypto_info; struct mlx5e_ktls_resync_resp *ktls_resync; struct mlx5e_icosq *sq; bool trigger_poll; @@ -373,7 +372,31 @@ static void resync_handle_seq_match(struct mlx5e_ktls_offload_context_rx *priv_r spin_lock_bh(&ktls_resync->lock); spin_lock_bh(&priv_rx->lock); - memcpy(info->rec_seq, &priv_rx->resync.sw_rcd_sn_be, sizeof(info->rec_seq)); + switch (priv_rx->crypto_info.crypto_info.cipher_type) { + case TLS_CIPHER_AES_GCM_128: { + struct tls12_crypto_info_aes_gcm_128 *info = + &priv_rx->crypto_info.crypto_info_128; + + memcpy(info->rec_seq, &priv_rx->resync.sw_rcd_sn_be, + sizeof(info->rec_seq)); + break; + } + case TLS_CIPHER_AES_GCM_256: { + struct tls12_crypto_info_aes_gcm_256 *info = + &priv_rx->crypto_info.crypto_info_256; + + memcpy(info->rec_seq, &priv_rx->resync.sw_rcd_sn_be, + sizeof(info->rec_seq)); + break; + } + default: + WARN_ONCE(1, "Unsupported cipher type %u\n", + priv_rx->crypto_info.crypto_info.cipher_type); + spin_unlock_bh(&priv_rx->lock); + spin_unlock_bh(&ktls_resync->lock); + return; + } + if (list_empty(&priv_rx->list)) { list_add_tail(&priv_rx->list, &ktls_resync->list); trigger_poll = !test_and_set_bit(MLX5E_SQ_STATE_PENDING_TLS_RX_RESYNC, &sq->state); @@ -461,6 +484,7 @@ static void resync_update_sn(struct mlx5e_rq *rq, struct sk_buff *skb) { struct ethhdr *eth = (struct ethhdr *)(skb->data); struct net_device *netdev = rq->netdev; + struct net *net = dev_net(netdev); struct sock *sk = NULL; unsigned int datalen; struct iphdr *iph; @@ -475,7 +499,7 @@ static void resync_update_sn(struct mlx5e_rq *rq, struct sk_buff *skb) depth += sizeof(struct iphdr); th = (void *)iph + sizeof(struct iphdr); - sk = inet_lookup_established(dev_net(netdev), &tcp_hashinfo, + sk = inet_lookup_established(net, net->ipv4.tcp_death_row.hashinfo, iph->saddr, th->source, iph->daddr, th->dest, netdev->ifindex); #if IS_ENABLED(CONFIG_IPV6) @@ -485,7 +509,7 @@ static void resync_update_sn(struct mlx5e_rq *rq, struct sk_buff *skb) depth += sizeof(struct ipv6hdr); th = (void *)ipv6h + sizeof(struct ipv6hdr); - sk = __inet6_lookup_established(dev_net(netdev), &tcp_hashinfo, + sk = __inet6_lookup_established(net, net->ipv4.tcp_death_row.hashinfo, &ipv6h->saddr, th->source, &ipv6h->daddr, ntohs(th->dest), netdev->ifindex, 0); @@ -603,8 +627,20 @@ int mlx5e_ktls_add_rx(struct net_device *netdev, struct sock *sk, INIT_LIST_HEAD(&priv_rx->list); spin_lock_init(&priv_rx->lock); - priv_rx->crypto_info = - *(struct tls12_crypto_info_aes_gcm_128 *)crypto_info; + switch (crypto_info->cipher_type) { + case TLS_CIPHER_AES_GCM_128: + priv_rx->crypto_info.crypto_info_128 = + *(struct tls12_crypto_info_aes_gcm_128 *)crypto_info; + break; + case TLS_CIPHER_AES_GCM_256: + priv_rx->crypto_info.crypto_info_256 = + *(struct tls12_crypto_info_aes_gcm_256 *)crypto_info; + break; + default: + WARN_ONCE(1, "Unsupported cipher type %u\n", + crypto_info->cipher_type); + return -EOPNOTSUPP; + } rxq = mlx5e_ktls_sk_get_rxq(sk); priv_rx->rxq = rxq; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ktls_tx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ktls_tx.c index 3a1f76eac542..2e0335246967 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ktls_tx.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ktls_tx.c @@ -93,7 +93,7 @@ struct mlx5e_ktls_offload_context_tx { bool ctx_post_pending; /* control / resync */ struct list_head list_node; /* member of the pool */ - struct tls12_crypto_info_aes_gcm_128 crypto_info; + union mlx5e_crypto_info crypto_info; struct tls_offload_context_tx *tx_ctx; struct mlx5_core_dev *mdev; struct mlx5e_tls_sw_stats *sw_stats; @@ -485,8 +485,20 @@ int mlx5e_ktls_add_tx(struct net_device *netdev, struct sock *sk, goto err_create_key; priv_tx->expected_seq = start_offload_tcp_sn; - priv_tx->crypto_info = - *(struct tls12_crypto_info_aes_gcm_128 *)crypto_info; + switch (crypto_info->cipher_type) { + case TLS_CIPHER_AES_GCM_128: + priv_tx->crypto_info.crypto_info_128 = + *(struct tls12_crypto_info_aes_gcm_128 *)crypto_info; + break; + case TLS_CIPHER_AES_GCM_256: + priv_tx->crypto_info.crypto_info_256 = + *(struct tls12_crypto_info_aes_gcm_256 *)crypto_info; + break; + default: + WARN_ONCE(1, "Unsupported cipher type %u\n", + crypto_info->cipher_type); + return -EOPNOTSUPP; + } priv_tx->tx_ctx = tls_offload_ctx_tx(tls_ctx); mlx5e_set_ktls_tx_priv_ctx(tls_ctx, priv_tx); @@ -671,14 +683,31 @@ tx_post_resync_params(struct mlx5e_txqsq *sq, struct mlx5e_ktls_offload_context_tx *priv_tx, u64 rcd_sn) { - struct tls12_crypto_info_aes_gcm_128 *info = &priv_tx->crypto_info; __be64 rn_be = cpu_to_be64(rcd_sn); bool skip_static_post; u16 rec_seq_sz; char *rec_seq; - rec_seq = info->rec_seq; - rec_seq_sz = sizeof(info->rec_seq); + switch (priv_tx->crypto_info.crypto_info.cipher_type) { + case TLS_CIPHER_AES_GCM_128: { + struct tls12_crypto_info_aes_gcm_128 *info = &priv_tx->crypto_info.crypto_info_128; + + rec_seq = info->rec_seq; + rec_seq_sz = sizeof(info->rec_seq); + break; + } + case TLS_CIPHER_AES_GCM_256: { + struct tls12_crypto_info_aes_gcm_256 *info = &priv_tx->crypto_info.crypto_info_256; + + rec_seq = info->rec_seq; + rec_seq_sz = sizeof(info->rec_seq); + break; + } + default: + WARN_ONCE(1, "Unsupported cipher type %u\n", + priv_tx->crypto_info.crypto_info.cipher_type); + return; + } skip_static_post = !memcmp(rec_seq, &rn_be, rec_seq_sz); if (!skip_static_post) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ktls_txrx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ktls_txrx.c index ac29aeb8af49..570a912dd6fa 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ktls_txrx.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ktls_txrx.c @@ -21,7 +21,7 @@ enum { static void fill_static_params(struct mlx5_wqe_tls_static_params_seg *params, - struct tls12_crypto_info_aes_gcm_128 *info, + union mlx5e_crypto_info *crypto_info, u32 key_id, u32 resync_tcp_sn) { char *initial_rn, *gcm_iv; @@ -32,7 +32,26 @@ fill_static_params(struct mlx5_wqe_tls_static_params_seg *params, ctx = params->ctx; - EXTRACT_INFO_FIELDS; + switch (crypto_info->crypto_info.cipher_type) { + case TLS_CIPHER_AES_GCM_128: { + struct tls12_crypto_info_aes_gcm_128 *info = + &crypto_info->crypto_info_128; + + EXTRACT_INFO_FIELDS; + break; + } + case TLS_CIPHER_AES_GCM_256: { + struct tls12_crypto_info_aes_gcm_256 *info = + &crypto_info->crypto_info_256; + + EXTRACT_INFO_FIELDS; + break; + } + default: + WARN_ONCE(1, "Unsupported cipher type %u\n", + crypto_info->crypto_info.cipher_type); + return; + } gcm_iv = MLX5_ADDR_OF(tls_static_params, ctx, gcm_iv); initial_rn = MLX5_ADDR_OF(tls_static_params, ctx, initial_record_number); @@ -54,7 +73,7 @@ fill_static_params(struct mlx5_wqe_tls_static_params_seg *params, void mlx5e_ktls_build_static_params(struct mlx5e_set_tls_static_params_wqe *wqe, u16 pc, u32 sqn, - struct tls12_crypto_info_aes_gcm_128 *info, + union mlx5e_crypto_info *crypto_info, u32 tis_tir_num, u32 key_id, u32 resync_tcp_sn, bool fence, enum tls_offload_ctx_dir direction) { @@ -75,7 +94,7 @@ mlx5e_ktls_build_static_params(struct mlx5e_set_tls_static_params_wqe *wqe, ucseg->flags = MLX5_UMR_INLINE; ucseg->bsf_octowords = cpu_to_be16(MLX5_ST_SZ_BYTES(tls_static_params) / 16); - fill_static_params(&wqe->params, info, key_id, resync_tcp_sn); + fill_static_params(&wqe->params, crypto_info, key_id, resync_tcp_sn); } static void diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ktls_utils.h b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ktls_utils.h index 0dc715c4c10d..3d79cd379890 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ktls_utils.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ktls_utils.h @@ -27,6 +27,12 @@ int mlx5e_ktls_add_rx(struct net_device *netdev, struct sock *sk, void mlx5e_ktls_del_rx(struct net_device *netdev, struct tls_context *tls_ctx); void mlx5e_ktls_rx_resync(struct net_device *netdev, struct sock *sk, u32 seq, u8 *rcd_sn); +union mlx5e_crypto_info { + struct tls_crypto_info crypto_info; + struct tls12_crypto_info_aes_gcm_128 crypto_info_128; + struct tls12_crypto_info_aes_gcm_256 crypto_info_256; +}; + struct mlx5e_set_tls_static_params_wqe { struct mlx5_wqe_ctrl_seg ctrl; struct mlx5_wqe_umr_ctrl_seg uctrl; @@ -72,7 +78,7 @@ struct mlx5e_get_tls_progress_params_wqe { void mlx5e_ktls_build_static_params(struct mlx5e_set_tls_static_params_wqe *wqe, u16 pc, u32 sqn, - struct tls12_crypto_info_aes_gcm_128 *info, + union mlx5e_crypto_info *crypto_info, u32 tis_tir_num, u32 key_id, u32 resync_tcp_sn, bool fence, enum tls_offload_ctx_dir direction); void diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/macsec.c b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/macsec.c new file mode 100644 index 000000000000..5da746da898d --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/macsec.c @@ -0,0 +1,1870 @@ +// SPDX-License-Identifier: GPL-2.0 OR Linux-OpenIB +/* Copyright (c) 2022, NVIDIA CORPORATION & AFFILIATES. All rights reserved. */ + +#include <linux/mlx5/device.h> +#include <linux/mlx5/mlx5_ifc.h> +#include <linux/xarray.h> + +#include "en.h" +#include "lib/aso.h" +#include "lib/mlx5.h" +#include "en_accel/macsec.h" +#include "en_accel/macsec_fs.h" + +#define MLX5_MACSEC_EPN_SCOPE_MID 0x80000000L +#define MLX5E_MACSEC_ASO_CTX_SZ MLX5_ST_SZ_BYTES(macsec_aso) + +enum mlx5_macsec_aso_event_arm { + MLX5E_ASO_EPN_ARM = BIT(0), +}; + +enum { + MLX5_MACSEC_ASO_REMOVE_FLOW_PKT_CNT_OFFSET, +}; + +struct mlx5e_macsec_handle { + struct mlx5e_macsec *macsec; + u32 obj_id; + u8 idx; +}; + +enum { + MLX5_MACSEC_EPN, +}; + +struct mlx5e_macsec_aso_out { + u8 event_arm; + u32 mode_param; +}; + +struct mlx5e_macsec_aso_in { + u8 mode; + u32 obj_id; +}; + +struct mlx5e_macsec_epn_state { + u32 epn_msb; + u8 epn_enabled; + u8 overlap; +}; + +struct mlx5e_macsec_async_work { + struct mlx5e_macsec *macsec; + struct mlx5_core_dev *mdev; + struct work_struct work; + u32 obj_id; +}; + +struct mlx5e_macsec_sa { + bool active; + u8 assoc_num; + u32 macsec_obj_id; + u32 enc_key_id; + u32 next_pn; + sci_t sci; + salt_t salt; + + struct rhash_head hash; + u32 fs_id; + union mlx5e_macsec_rule *macsec_rule; + struct rcu_head rcu_head; + struct mlx5e_macsec_epn_state epn_state; +}; + +struct mlx5e_macsec_rx_sc; +struct mlx5e_macsec_rx_sc_xarray_element { + u32 fs_id; + struct mlx5e_macsec_rx_sc *rx_sc; +}; + +struct mlx5e_macsec_rx_sc { + bool active; + sci_t sci; + struct mlx5e_macsec_sa *rx_sa[MACSEC_NUM_AN]; + struct list_head rx_sc_list_element; + struct mlx5e_macsec_rx_sc_xarray_element *sc_xarray_element; + struct metadata_dst *md_dst; + struct rcu_head rcu_head; +}; + +struct mlx5e_macsec_umr { + dma_addr_t dma_addr; + u8 ctx[MLX5_ST_SZ_BYTES(macsec_aso)]; + u32 mkey; +}; + +struct mlx5e_macsec_aso { + /* ASO */ + struct mlx5_aso *maso; + /* Protects macsec ASO */ + struct mutex aso_lock; + /* UMR */ + struct mlx5e_macsec_umr *umr; + + u32 pdn; +}; + +static const struct rhashtable_params rhash_sci = { + .key_len = sizeof_field(struct mlx5e_macsec_sa, sci), + .key_offset = offsetof(struct mlx5e_macsec_sa, sci), + .head_offset = offsetof(struct mlx5e_macsec_sa, hash), + .automatic_shrinking = true, + .min_size = 1, +}; + +struct mlx5e_macsec_device { + const struct net_device *netdev; + struct mlx5e_macsec_sa *tx_sa[MACSEC_NUM_AN]; + struct list_head macsec_rx_sc_list_head; + unsigned char *dev_addr; + struct list_head macsec_device_list_element; +}; + +struct mlx5e_macsec { + struct list_head macsec_device_list_head; + int num_of_devices; + struct mlx5e_macsec_fs *macsec_fs; + struct mutex lock; /* Protects mlx5e_macsec internal contexts */ + + /* Tx sci -> fs id mapping handling */ + struct rhashtable sci_hash; /* sci -> mlx5e_macsec_sa */ + + /* Rx fs_id -> rx_sc mapping */ + struct xarray sc_xarray; + + struct mlx5_core_dev *mdev; + + /* Stats manage */ + struct mlx5e_macsec_stats stats; + + /* ASO */ + struct mlx5e_macsec_aso aso; + + struct notifier_block nb; + struct workqueue_struct *wq; +}; + +struct mlx5_macsec_obj_attrs { + u32 aso_pdn; + u32 next_pn; + __be64 sci; + u32 enc_key_id; + bool encrypt; + struct mlx5e_macsec_epn_state epn_state; + salt_t salt; + __be32 ssci; + bool replay_protect; + u32 replay_window; +}; + +struct mlx5_aso_ctrl_param { + u8 data_mask_mode; + u8 condition_0_operand; + u8 condition_1_operand; + u8 condition_0_offset; + u8 condition_1_offset; + u8 data_offset; + u8 condition_operand; + u32 condition_0_data; + u32 condition_0_mask; + u32 condition_1_data; + u32 condition_1_mask; + u64 bitwise_data; + u64 data_mask; +}; + +static int mlx5e_macsec_aso_reg_mr(struct mlx5_core_dev *mdev, struct mlx5e_macsec_aso *aso) +{ + struct mlx5e_macsec_umr *umr; + struct device *dma_device; + dma_addr_t dma_addr; + int err; + + umr = kzalloc(sizeof(*umr), GFP_KERNEL); + if (!umr) { + err = -ENOMEM; + return err; + } + + dma_device = &mdev->pdev->dev; + dma_addr = dma_map_single(dma_device, umr->ctx, sizeof(umr->ctx), DMA_BIDIRECTIONAL); + err = dma_mapping_error(dma_device, dma_addr); + if (err) { + mlx5_core_err(mdev, "Can't map dma device, err=%d\n", err); + goto out_dma; + } + + err = mlx5e_create_mkey(mdev, aso->pdn, &umr->mkey); + if (err) { + mlx5_core_err(mdev, "Can't create mkey, err=%d\n", err); + goto out_mkey; + } + + umr->dma_addr = dma_addr; + + aso->umr = umr; + + return 0; + +out_mkey: + dma_unmap_single(dma_device, dma_addr, sizeof(umr->ctx), DMA_BIDIRECTIONAL); +out_dma: + kfree(umr); + return err; +} + +static void mlx5e_macsec_aso_dereg_mr(struct mlx5_core_dev *mdev, struct mlx5e_macsec_aso *aso) +{ + struct mlx5e_macsec_umr *umr = aso->umr; + + mlx5_core_destroy_mkey(mdev, umr->mkey); + dma_unmap_single(&mdev->pdev->dev, umr->dma_addr, sizeof(umr->ctx), DMA_BIDIRECTIONAL); + kfree(umr); +} + +static int macsec_set_replay_protection(struct mlx5_macsec_obj_attrs *attrs, void *aso_ctx) +{ + u8 window_sz; + + if (!attrs->replay_protect) + return 0; + + switch (attrs->replay_window) { + case 256: + window_sz = MLX5_MACSEC_ASO_REPLAY_WIN_256BIT; + break; + case 128: + window_sz = MLX5_MACSEC_ASO_REPLAY_WIN_128BIT; + break; + case 64: + window_sz = MLX5_MACSEC_ASO_REPLAY_WIN_64BIT; + break; + case 32: + window_sz = MLX5_MACSEC_ASO_REPLAY_WIN_32BIT; + break; + default: + return -EINVAL; + } + MLX5_SET(macsec_aso, aso_ctx, window_size, window_sz); + MLX5_SET(macsec_aso, aso_ctx, mode, MLX5_MACSEC_ASO_REPLAY_PROTECTION); + + return 0; +} + +static int mlx5e_macsec_create_object(struct mlx5_core_dev *mdev, + struct mlx5_macsec_obj_attrs *attrs, + bool is_tx, + u32 *macsec_obj_id) +{ + u32 in[MLX5_ST_SZ_DW(create_macsec_obj_in)] = {}; + u32 out[MLX5_ST_SZ_DW(general_obj_out_cmd_hdr)]; + void *aso_ctx; + void *obj; + int err; + + obj = MLX5_ADDR_OF(create_macsec_obj_in, in, macsec_object); + aso_ctx = MLX5_ADDR_OF(macsec_offload_obj, obj, macsec_aso); + + MLX5_SET(macsec_offload_obj, obj, confidentiality_en, attrs->encrypt); + MLX5_SET(macsec_offload_obj, obj, dekn, attrs->enc_key_id); + MLX5_SET(macsec_offload_obj, obj, aso_return_reg, MLX5_MACSEC_ASO_REG_C_4_5); + MLX5_SET(macsec_offload_obj, obj, macsec_aso_access_pd, attrs->aso_pdn); + MLX5_SET(macsec_aso, aso_ctx, mode_parameter, attrs->next_pn); + + /* Epn */ + if (attrs->epn_state.epn_enabled) { + void *salt_p; + int i; + + MLX5_SET(macsec_aso, aso_ctx, epn_event_arm, 1); + MLX5_SET(macsec_offload_obj, obj, epn_en, 1); + MLX5_SET(macsec_offload_obj, obj, epn_msb, attrs->epn_state.epn_msb); + MLX5_SET(macsec_offload_obj, obj, epn_overlap, attrs->epn_state.overlap); + MLX5_SET64(macsec_offload_obj, obj, sci, (__force u64)attrs->ssci); + salt_p = MLX5_ADDR_OF(macsec_offload_obj, obj, salt); + for (i = 0; i < 3 ; i++) + memcpy((u32 *)salt_p + i, &attrs->salt.bytes[4 * (2 - i)], 4); + } else { + MLX5_SET64(macsec_offload_obj, obj, sci, (__force u64)(attrs->sci)); + } + + MLX5_SET(macsec_aso, aso_ctx, valid, 0x1); + if (is_tx) { + MLX5_SET(macsec_aso, aso_ctx, mode, MLX5_MACSEC_ASO_INC_SN); + } else { + err = macsec_set_replay_protection(attrs, aso_ctx); + if (err) + return err; + } + + /* general object fields set */ + MLX5_SET(general_obj_in_cmd_hdr, in, opcode, MLX5_CMD_OP_CREATE_GENERAL_OBJECT); + MLX5_SET(general_obj_in_cmd_hdr, in, obj_type, MLX5_GENERAL_OBJECT_TYPES_MACSEC); + + err = mlx5_cmd_exec(mdev, in, sizeof(in), out, sizeof(out)); + if (err) { + mlx5_core_err(mdev, + "MACsec offload: Failed to create MACsec object (err = %d)\n", + err); + return err; + } + + *macsec_obj_id = MLX5_GET(general_obj_out_cmd_hdr, out, obj_id); + + return err; +} + +static void mlx5e_macsec_destroy_object(struct mlx5_core_dev *mdev, u32 macsec_obj_id) +{ + u32 in[MLX5_ST_SZ_DW(general_obj_in_cmd_hdr)] = {}; + u32 out[MLX5_ST_SZ_DW(general_obj_out_cmd_hdr)]; + + MLX5_SET(general_obj_in_cmd_hdr, in, opcode, MLX5_CMD_OP_DESTROY_GENERAL_OBJECT); + MLX5_SET(general_obj_in_cmd_hdr, in, obj_type, MLX5_GENERAL_OBJECT_TYPES_MACSEC); + MLX5_SET(general_obj_in_cmd_hdr, in, obj_id, macsec_obj_id); + + mlx5_cmd_exec(mdev, in, sizeof(in), out, sizeof(out)); +} + +static void mlx5e_macsec_cleanup_sa(struct mlx5e_macsec *macsec, + struct mlx5e_macsec_sa *sa, + bool is_tx) +{ + int action = (is_tx) ? MLX5_ACCEL_MACSEC_ACTION_ENCRYPT : + MLX5_ACCEL_MACSEC_ACTION_DECRYPT; + + if ((is_tx) && sa->fs_id) { + /* Make sure ongoing datapath readers sees a valid SA */ + rhashtable_remove_fast(&macsec->sci_hash, &sa->hash, rhash_sci); + sa->fs_id = 0; + } + + if (!sa->macsec_rule) + return; + + mlx5e_macsec_fs_del_rule(macsec->macsec_fs, sa->macsec_rule, action); + mlx5e_macsec_destroy_object(macsec->mdev, sa->macsec_obj_id); + sa->macsec_rule = NULL; +} + +static int mlx5e_macsec_init_sa(struct macsec_context *ctx, + struct mlx5e_macsec_sa *sa, + bool encrypt, + bool is_tx) +{ + struct mlx5e_priv *priv = netdev_priv(ctx->netdev); + struct mlx5e_macsec *macsec = priv->macsec; + struct mlx5_macsec_rule_attrs rule_attrs; + struct mlx5_core_dev *mdev = priv->mdev; + struct mlx5_macsec_obj_attrs obj_attrs; + union mlx5e_macsec_rule *macsec_rule; + struct macsec_key *key; + int err; + + obj_attrs.next_pn = sa->next_pn; + obj_attrs.sci = cpu_to_be64((__force u64)sa->sci); + obj_attrs.enc_key_id = sa->enc_key_id; + obj_attrs.encrypt = encrypt; + obj_attrs.aso_pdn = macsec->aso.pdn; + obj_attrs.epn_state = sa->epn_state; + + if (is_tx) { + obj_attrs.ssci = cpu_to_be32((__force u32)ctx->sa.tx_sa->ssci); + key = &ctx->sa.tx_sa->key; + } else { + obj_attrs.ssci = cpu_to_be32((__force u32)ctx->sa.rx_sa->ssci); + key = &ctx->sa.rx_sa->key; + } + + memcpy(&obj_attrs.salt, &key->salt, sizeof(key->salt)); + obj_attrs.replay_window = ctx->secy->replay_window; + obj_attrs.replay_protect = ctx->secy->replay_protect; + + err = mlx5e_macsec_create_object(mdev, &obj_attrs, is_tx, &sa->macsec_obj_id); + if (err) + return err; + + rule_attrs.macsec_obj_id = sa->macsec_obj_id; + rule_attrs.sci = sa->sci; + rule_attrs.assoc_num = sa->assoc_num; + rule_attrs.action = (is_tx) ? MLX5_ACCEL_MACSEC_ACTION_ENCRYPT : + MLX5_ACCEL_MACSEC_ACTION_DECRYPT; + + macsec_rule = mlx5e_macsec_fs_add_rule(macsec->macsec_fs, ctx, &rule_attrs, &sa->fs_id); + if (!macsec_rule) { + err = -ENOMEM; + goto destroy_macsec_object; + } + + sa->macsec_rule = macsec_rule; + + if (is_tx) { + err = rhashtable_insert_fast(&macsec->sci_hash, &sa->hash, rhash_sci); + if (err) + goto destroy_macsec_object_and_rule; + } + + return 0; + +destroy_macsec_object_and_rule: + mlx5e_macsec_cleanup_sa(macsec, sa, is_tx); +destroy_macsec_object: + mlx5e_macsec_destroy_object(mdev, sa->macsec_obj_id); + + return err; +} + +static struct mlx5e_macsec_rx_sc * +mlx5e_macsec_get_rx_sc_from_sc_list(const struct list_head *list, sci_t sci) +{ + struct mlx5e_macsec_rx_sc *iter; + + list_for_each_entry_rcu(iter, list, rx_sc_list_element) { + if (iter->sci == sci) + return iter; + } + + return NULL; +} + +static int mlx5e_macsec_update_rx_sa(struct mlx5e_macsec *macsec, + struct mlx5e_macsec_sa *rx_sa, + bool active) +{ + struct mlx5_core_dev *mdev = macsec->mdev; + struct mlx5_macsec_obj_attrs attrs; + int err = 0; + + if (rx_sa->active != active) + return 0; + + rx_sa->active = active; + if (!active) { + mlx5e_macsec_cleanup_sa(macsec, rx_sa, false); + return 0; + } + + attrs.sci = rx_sa->sci; + attrs.enc_key_id = rx_sa->enc_key_id; + err = mlx5e_macsec_create_object(mdev, &attrs, false, &rx_sa->macsec_obj_id); + if (err) + return err; + + return 0; +} + +static bool mlx5e_macsec_secy_features_validate(struct macsec_context *ctx) +{ + const struct net_device *netdev = ctx->netdev; + const struct macsec_secy *secy = ctx->secy; + + if (secy->validate_frames != MACSEC_VALIDATE_STRICT) { + netdev_err(netdev, + "MACsec offload is supported only when validate_frame is in strict mode\n"); + return false; + } + + if (secy->icv_len != MACSEC_DEFAULT_ICV_LEN) { + netdev_err(netdev, "MACsec offload is supported only when icv_len is %d\n", + MACSEC_DEFAULT_ICV_LEN); + return false; + } + + if (!secy->protect_frames) { + netdev_err(netdev, + "MACsec offload is supported only when protect_frames is set\n"); + return false; + } + + return true; +} + +static struct mlx5e_macsec_device * +mlx5e_macsec_get_macsec_device_context(const struct mlx5e_macsec *macsec, + const struct macsec_context *ctx) +{ + struct mlx5e_macsec_device *iter; + const struct list_head *list; + + list = &macsec->macsec_device_list_head; + list_for_each_entry_rcu(iter, list, macsec_device_list_element) { + if (iter->netdev == ctx->secy->netdev) + return iter; + } + + return NULL; +} + +static void update_macsec_epn(struct mlx5e_macsec_sa *sa, const struct macsec_key *key, + const pn_t *next_pn_halves) +{ + struct mlx5e_macsec_epn_state *epn_state = &sa->epn_state; + + sa->salt = key->salt; + epn_state->epn_enabled = 1; + epn_state->epn_msb = next_pn_halves->upper; + epn_state->overlap = next_pn_halves->lower < MLX5_MACSEC_EPN_SCOPE_MID ? 0 : 1; +} + +static int mlx5e_macsec_add_txsa(struct macsec_context *ctx) +{ + const struct macsec_tx_sc *tx_sc = &ctx->secy->tx_sc; + const struct macsec_tx_sa *ctx_tx_sa = ctx->sa.tx_sa; + struct mlx5e_priv *priv = netdev_priv(ctx->netdev); + const struct macsec_secy *secy = ctx->secy; + struct mlx5e_macsec_device *macsec_device; + struct mlx5_core_dev *mdev = priv->mdev; + u8 assoc_num = ctx->sa.assoc_num; + struct mlx5e_macsec_sa *tx_sa; + struct mlx5e_macsec *macsec; + int err = 0; + + mutex_lock(&priv->macsec->lock); + + macsec = priv->macsec; + macsec_device = mlx5e_macsec_get_macsec_device_context(macsec, ctx); + if (!macsec_device) { + netdev_err(ctx->netdev, "MACsec offload: Failed to find device context\n"); + err = -EEXIST; + goto out; + } + + if (macsec_device->tx_sa[assoc_num]) { + netdev_err(ctx->netdev, "MACsec offload tx_sa: %d already exist\n", assoc_num); + err = -EEXIST; + goto out; + } + + tx_sa = kzalloc(sizeof(*tx_sa), GFP_KERNEL); + if (!tx_sa) { + err = -ENOMEM; + goto out; + } + + tx_sa->active = ctx_tx_sa->active; + tx_sa->next_pn = ctx_tx_sa->next_pn_halves.lower; + tx_sa->sci = secy->sci; + tx_sa->assoc_num = assoc_num; + + if (secy->xpn) + update_macsec_epn(tx_sa, &ctx_tx_sa->key, &ctx_tx_sa->next_pn_halves); + + err = mlx5_create_encryption_key(mdev, ctx->sa.key, secy->key_len, + MLX5_ACCEL_OBJ_MACSEC_KEY, + &tx_sa->enc_key_id); + if (err) + goto destroy_sa; + + macsec_device->tx_sa[assoc_num] = tx_sa; + if (!secy->operational || + assoc_num != tx_sc->encoding_sa || + !tx_sa->active) + goto out; + + err = mlx5e_macsec_init_sa(ctx, tx_sa, tx_sc->encrypt, true); + if (err) + goto destroy_encryption_key; + + mutex_unlock(&macsec->lock); + + return 0; + +destroy_encryption_key: + macsec_device->tx_sa[assoc_num] = NULL; + mlx5_destroy_encryption_key(mdev, tx_sa->enc_key_id); +destroy_sa: + kfree(tx_sa); +out: + mutex_unlock(&macsec->lock); + + return err; +} + +static int mlx5e_macsec_upd_txsa(struct macsec_context *ctx) +{ + const struct macsec_tx_sc *tx_sc = &ctx->secy->tx_sc; + const struct macsec_tx_sa *ctx_tx_sa = ctx->sa.tx_sa; + struct mlx5e_priv *priv = netdev_priv(ctx->netdev); + struct mlx5e_macsec_device *macsec_device; + u8 assoc_num = ctx->sa.assoc_num; + struct mlx5e_macsec_sa *tx_sa; + struct mlx5e_macsec *macsec; + struct net_device *netdev; + int err = 0; + + mutex_lock(&priv->macsec->lock); + + macsec = priv->macsec; + netdev = ctx->netdev; + macsec_device = mlx5e_macsec_get_macsec_device_context(macsec, ctx); + if (!macsec_device) { + netdev_err(netdev, "MACsec offload: Failed to find device context\n"); + err = -EINVAL; + goto out; + } + + tx_sa = macsec_device->tx_sa[assoc_num]; + if (!tx_sa) { + netdev_err(netdev, "MACsec offload: TX sa 0x%x doesn't exist\n", assoc_num); + err = -EEXIST; + goto out; + } + + if (tx_sa->next_pn != ctx_tx_sa->next_pn_halves.lower) { + netdev_err(netdev, "MACsec offload: update TX sa %d PN isn't supported\n", + assoc_num); + err = -EINVAL; + goto out; + } + + if (tx_sa->active == ctx_tx_sa->active) + goto out; + + if (tx_sa->assoc_num != tx_sc->encoding_sa) + goto out; + + if (ctx_tx_sa->active) { + err = mlx5e_macsec_init_sa(ctx, tx_sa, tx_sc->encrypt, true); + if (err) + goto out; + } else { + if (!tx_sa->macsec_rule) { + err = -EINVAL; + goto out; + } + + mlx5e_macsec_cleanup_sa(macsec, tx_sa, true); + } + + tx_sa->active = ctx_tx_sa->active; +out: + mutex_unlock(&macsec->lock); + + return err; +} + +static int mlx5e_macsec_del_txsa(struct macsec_context *ctx) +{ + struct mlx5e_priv *priv = netdev_priv(ctx->netdev); + struct mlx5e_macsec_device *macsec_device; + u8 assoc_num = ctx->sa.assoc_num; + struct mlx5e_macsec_sa *tx_sa; + struct mlx5e_macsec *macsec; + int err = 0; + + mutex_lock(&priv->macsec->lock); + macsec = priv->macsec; + macsec_device = mlx5e_macsec_get_macsec_device_context(macsec, ctx); + if (!macsec_device) { + netdev_err(ctx->netdev, "MACsec offload: Failed to find device context\n"); + err = -EINVAL; + goto out; + } + + tx_sa = macsec_device->tx_sa[assoc_num]; + if (!tx_sa) { + netdev_err(ctx->netdev, "MACsec offload: TX sa 0x%x doesn't exist\n", assoc_num); + err = -EEXIST; + goto out; + } + + mlx5e_macsec_cleanup_sa(macsec, tx_sa, true); + mlx5_destroy_encryption_key(macsec->mdev, tx_sa->enc_key_id); + kfree_rcu(tx_sa); + macsec_device->tx_sa[assoc_num] = NULL; + +out: + mutex_unlock(&macsec->lock); + + return err; +} + +static u32 mlx5e_macsec_get_sa_from_hashtable(struct rhashtable *sci_hash, sci_t *sci) +{ + struct mlx5e_macsec_sa *macsec_sa; + u32 fs_id = 0; + + rcu_read_lock(); + macsec_sa = rhashtable_lookup(sci_hash, sci, rhash_sci); + if (macsec_sa) + fs_id = macsec_sa->fs_id; + rcu_read_unlock(); + + return fs_id; +} + +static int mlx5e_macsec_add_rxsc(struct macsec_context *ctx) +{ + struct mlx5e_macsec_rx_sc_xarray_element *sc_xarray_element; + struct mlx5e_priv *priv = netdev_priv(ctx->netdev); + const struct macsec_rx_sc *ctx_rx_sc = ctx->rx_sc; + struct mlx5e_macsec_device *macsec_device; + struct mlx5e_macsec_rx_sc *rx_sc; + struct list_head *rx_sc_list; + struct mlx5e_macsec *macsec; + int err = 0; + + mutex_lock(&priv->macsec->lock); + macsec = priv->macsec; + macsec_device = mlx5e_macsec_get_macsec_device_context(macsec, ctx); + if (!macsec_device) { + netdev_err(ctx->netdev, "MACsec offload: Failed to find device context\n"); + err = -EINVAL; + goto out; + } + + rx_sc_list = &macsec_device->macsec_rx_sc_list_head; + rx_sc = mlx5e_macsec_get_rx_sc_from_sc_list(rx_sc_list, ctx_rx_sc->sci); + if (rx_sc) { + netdev_err(ctx->netdev, "MACsec offload: rx_sc (sci %lld) already exists\n", + ctx_rx_sc->sci); + err = -EEXIST; + goto out; + } + + rx_sc = kzalloc(sizeof(*rx_sc), GFP_KERNEL); + if (!rx_sc) { + err = -ENOMEM; + goto out; + } + + sc_xarray_element = kzalloc(sizeof(*sc_xarray_element), GFP_KERNEL); + if (!sc_xarray_element) { + err = -ENOMEM; + goto destroy_rx_sc; + } + + sc_xarray_element->rx_sc = rx_sc; + err = xa_alloc(&macsec->sc_xarray, &sc_xarray_element->fs_id, sc_xarray_element, + XA_LIMIT(1, USHRT_MAX), GFP_KERNEL); + if (err) + goto destroy_sc_xarray_elemenet; + + rx_sc->md_dst = metadata_dst_alloc(0, METADATA_MACSEC, GFP_KERNEL); + if (!rx_sc->md_dst) { + err = -ENOMEM; + goto erase_xa_alloc; + } + + rx_sc->sci = ctx_rx_sc->sci; + rx_sc->active = ctx_rx_sc->active; + list_add_rcu(&rx_sc->rx_sc_list_element, rx_sc_list); + + rx_sc->sc_xarray_element = sc_xarray_element; + rx_sc->md_dst->u.macsec_info.sci = rx_sc->sci; + mutex_unlock(&macsec->lock); + + return 0; + +erase_xa_alloc: + xa_erase(&macsec->sc_xarray, sc_xarray_element->fs_id); +destroy_sc_xarray_elemenet: + kfree(sc_xarray_element); +destroy_rx_sc: + kfree(rx_sc); + +out: + mutex_unlock(&macsec->lock); + + return err; +} + +static int mlx5e_macsec_upd_rxsc(struct macsec_context *ctx) +{ + struct mlx5e_priv *priv = netdev_priv(ctx->netdev); + const struct macsec_rx_sc *ctx_rx_sc = ctx->rx_sc; + struct mlx5e_macsec_device *macsec_device; + struct mlx5e_macsec_rx_sc *rx_sc; + struct mlx5e_macsec_sa *rx_sa; + struct mlx5e_macsec *macsec; + struct list_head *list; + int i; + int err = 0; + + mutex_lock(&priv->macsec->lock); + + macsec = priv->macsec; + macsec_device = mlx5e_macsec_get_macsec_device_context(macsec, ctx); + if (!macsec_device) { + netdev_err(ctx->netdev, "MACsec offload: Failed to find device context\n"); + err = -EINVAL; + goto out; + } + + list = &macsec_device->macsec_rx_sc_list_head; + rx_sc = mlx5e_macsec_get_rx_sc_from_sc_list(list, ctx_rx_sc->sci); + if (!rx_sc) { + err = -EINVAL; + goto out; + } + + rx_sc->active = ctx_rx_sc->active; + if (rx_sc->active == ctx_rx_sc->active) + goto out; + + for (i = 0; i < MACSEC_NUM_AN; ++i) { + rx_sa = rx_sc->rx_sa[i]; + if (!rx_sa) + continue; + + err = mlx5e_macsec_update_rx_sa(macsec, rx_sa, rx_sa->active && ctx_rx_sc->active); + if (err) + goto out; + } + +out: + mutex_unlock(&macsec->lock); + + return err; +} + +static int mlx5e_macsec_del_rxsc(struct macsec_context *ctx) +{ + struct mlx5e_priv *priv = netdev_priv(ctx->netdev); + struct mlx5e_macsec_device *macsec_device; + struct mlx5e_macsec_rx_sc *rx_sc; + struct mlx5e_macsec_sa *rx_sa; + struct mlx5e_macsec *macsec; + struct list_head *list; + int err = 0; + int i; + + mutex_lock(&priv->macsec->lock); + + macsec = priv->macsec; + macsec_device = mlx5e_macsec_get_macsec_device_context(macsec, ctx); + if (!macsec_device) { + netdev_err(ctx->netdev, "MACsec offload: Failed to find device context\n"); + err = -EINVAL; + goto out; + } + + list = &macsec_device->macsec_rx_sc_list_head; + rx_sc = mlx5e_macsec_get_rx_sc_from_sc_list(list, ctx->rx_sc->sci); + if (!rx_sc) { + netdev_err(ctx->netdev, + "MACsec offload rx_sc sci %lld doesn't exist\n", + ctx->sa.rx_sa->sc->sci); + err = -EINVAL; + goto out; + } + + for (i = 0; i < MACSEC_NUM_AN; ++i) { + rx_sa = rx_sc->rx_sa[i]; + if (!rx_sa) + continue; + + mlx5e_macsec_cleanup_sa(macsec, rx_sa, false); + mlx5_destroy_encryption_key(macsec->mdev, rx_sa->enc_key_id); + + kfree(rx_sa); + rx_sc->rx_sa[i] = NULL; + } + +/* + * At this point the relevant MACsec offload Rx rule already removed at + * mlx5e_macsec_cleanup_sa need to wait for datapath to finish current + * Rx related data propagating using xa_erase which uses rcu to sync, + * once fs_id is erased then this rx_sc is hidden from datapath. + */ + list_del_rcu(&rx_sc->rx_sc_list_element); + xa_erase(&macsec->sc_xarray, rx_sc->sc_xarray_element->fs_id); + metadata_dst_free(rx_sc->md_dst); + kfree(rx_sc->sc_xarray_element); + + kfree_rcu(rx_sc); + +out: + mutex_unlock(&macsec->lock); + + return err; +} + +static int mlx5e_macsec_add_rxsa(struct macsec_context *ctx) +{ + const struct macsec_rx_sa *ctx_rx_sa = ctx->sa.rx_sa; + struct mlx5e_priv *priv = netdev_priv(ctx->netdev); + struct mlx5e_macsec_device *macsec_device; + struct mlx5_core_dev *mdev = priv->mdev; + u8 assoc_num = ctx->sa.assoc_num; + struct mlx5e_macsec_rx_sc *rx_sc; + sci_t sci = ctx_rx_sa->sc->sci; + struct mlx5e_macsec_sa *rx_sa; + struct mlx5e_macsec *macsec; + struct list_head *list; + int err = 0; + + mutex_lock(&priv->macsec->lock); + + macsec = priv->macsec; + macsec_device = mlx5e_macsec_get_macsec_device_context(macsec, ctx); + if (!macsec_device) { + netdev_err(ctx->netdev, "MACsec offload: Failed to find device context\n"); + err = -EINVAL; + goto out; + } + + list = &macsec_device->macsec_rx_sc_list_head; + rx_sc = mlx5e_macsec_get_rx_sc_from_sc_list(list, sci); + if (!rx_sc) { + netdev_err(ctx->netdev, + "MACsec offload rx_sc sci %lld doesn't exist\n", + ctx->sa.rx_sa->sc->sci); + err = -EINVAL; + goto out; + } + + if (rx_sc->rx_sa[assoc_num]) { + netdev_err(ctx->netdev, + "MACsec offload rx_sc sci %lld rx_sa %d already exist\n", + sci, assoc_num); + err = -EEXIST; + goto out; + } + + rx_sa = kzalloc(sizeof(*rx_sa), GFP_KERNEL); + if (!rx_sa) { + err = -ENOMEM; + goto out; + } + + rx_sa->active = ctx_rx_sa->active; + rx_sa->next_pn = ctx_rx_sa->next_pn; + rx_sa->sci = sci; + rx_sa->assoc_num = assoc_num; + rx_sa->fs_id = rx_sc->sc_xarray_element->fs_id; + + if (ctx->secy->xpn) + update_macsec_epn(rx_sa, &ctx_rx_sa->key, &ctx_rx_sa->next_pn_halves); + + err = mlx5_create_encryption_key(mdev, ctx->sa.key, ctx->secy->key_len, + MLX5_ACCEL_OBJ_MACSEC_KEY, + &rx_sa->enc_key_id); + if (err) + goto destroy_sa; + + rx_sc->rx_sa[assoc_num] = rx_sa; + if (!rx_sa->active) + goto out; + + //TODO - add support for both authentication and encryption flows + err = mlx5e_macsec_init_sa(ctx, rx_sa, true, false); + if (err) + goto destroy_encryption_key; + + goto out; + +destroy_encryption_key: + rx_sc->rx_sa[assoc_num] = NULL; + mlx5_destroy_encryption_key(mdev, rx_sa->enc_key_id); +destroy_sa: + kfree(rx_sa); +out: + mutex_unlock(&macsec->lock); + + return err; +} + +static int mlx5e_macsec_upd_rxsa(struct macsec_context *ctx) +{ + const struct macsec_rx_sa *ctx_rx_sa = ctx->sa.rx_sa; + struct mlx5e_priv *priv = netdev_priv(ctx->netdev); + struct mlx5e_macsec_device *macsec_device; + u8 assoc_num = ctx->sa.assoc_num; + struct mlx5e_macsec_rx_sc *rx_sc; + sci_t sci = ctx_rx_sa->sc->sci; + struct mlx5e_macsec_sa *rx_sa; + struct mlx5e_macsec *macsec; + struct list_head *list; + int err = 0; + + mutex_lock(&priv->macsec->lock); + + macsec = priv->macsec; + macsec_device = mlx5e_macsec_get_macsec_device_context(macsec, ctx); + if (!macsec_device) { + netdev_err(ctx->netdev, "MACsec offload: Failed to find device context\n"); + err = -EINVAL; + goto out; + } + + list = &macsec_device->macsec_rx_sc_list_head; + rx_sc = mlx5e_macsec_get_rx_sc_from_sc_list(list, sci); + if (!rx_sc) { + netdev_err(ctx->netdev, + "MACsec offload rx_sc sci %lld doesn't exist\n", + ctx->sa.rx_sa->sc->sci); + err = -EINVAL; + goto out; + } + + rx_sa = rx_sc->rx_sa[assoc_num]; + if (rx_sa) { + netdev_err(ctx->netdev, + "MACsec offload rx_sc sci %lld rx_sa %d already exist\n", + sci, assoc_num); + err = -EEXIST; + goto out; + } + + if (rx_sa->next_pn != ctx_rx_sa->next_pn_halves.lower) { + netdev_err(ctx->netdev, + "MACsec offload update RX sa %d PN isn't supported\n", + assoc_num); + err = -EINVAL; + goto out; + } + + err = mlx5e_macsec_update_rx_sa(macsec, rx_sa, ctx_rx_sa->active); +out: + mutex_unlock(&macsec->lock); + + return err; +} + +static int mlx5e_macsec_del_rxsa(struct macsec_context *ctx) +{ + struct mlx5e_priv *priv = netdev_priv(ctx->netdev); + struct mlx5e_macsec_device *macsec_device; + sci_t sci = ctx->sa.rx_sa->sc->sci; + struct mlx5e_macsec_rx_sc *rx_sc; + u8 assoc_num = ctx->sa.assoc_num; + struct mlx5e_macsec_sa *rx_sa; + struct mlx5e_macsec *macsec; + struct list_head *list; + int err = 0; + + mutex_lock(&priv->macsec->lock); + + macsec = priv->macsec; + macsec_device = mlx5e_macsec_get_macsec_device_context(macsec, ctx); + if (!macsec_device) { + netdev_err(ctx->netdev, "MACsec offload: Failed to find device context\n"); + err = -EINVAL; + goto out; + } + + list = &macsec_device->macsec_rx_sc_list_head; + rx_sc = mlx5e_macsec_get_rx_sc_from_sc_list(list, sci); + if (!rx_sc) { + netdev_err(ctx->netdev, + "MACsec offload rx_sc sci %lld doesn't exist\n", + ctx->sa.rx_sa->sc->sci); + err = -EINVAL; + goto out; + } + + rx_sa = rx_sc->rx_sa[assoc_num]; + if (rx_sa) { + netdev_err(ctx->netdev, + "MACsec offload rx_sc sci %lld rx_sa %d already exist\n", + sci, assoc_num); + err = -EEXIST; + goto out; + } + + mlx5e_macsec_cleanup_sa(macsec, rx_sa, false); + mlx5_destroy_encryption_key(macsec->mdev, rx_sa->enc_key_id); + kfree(rx_sa); + rx_sc->rx_sa[assoc_num] = NULL; + +out: + mutex_unlock(&macsec->lock); + + return err; +} + +static int mlx5e_macsec_add_secy(struct macsec_context *ctx) +{ + struct mlx5e_priv *priv = netdev_priv(ctx->netdev); + const struct net_device *dev = ctx->secy->netdev; + const struct net_device *netdev = ctx->netdev; + struct mlx5e_macsec_device *macsec_device; + struct mlx5e_macsec *macsec; + int err = 0; + + if (!mlx5e_macsec_secy_features_validate(ctx)) + return -EINVAL; + + mutex_lock(&priv->macsec->lock); + macsec = priv->macsec; + if (mlx5e_macsec_get_macsec_device_context(macsec, ctx)) { + netdev_err(netdev, "MACsec offload: MACsec net_device already exist\n"); + goto out; + } + + if (macsec->num_of_devices >= MLX5_MACSEC_NUM_OF_SUPPORTED_INTERFACES) { + netdev_err(netdev, "Currently, only %d MACsec offload devices can be set\n", + MLX5_MACSEC_NUM_OF_SUPPORTED_INTERFACES); + err = -EBUSY; + goto out; + } + + macsec_device = kzalloc(sizeof(*macsec_device), GFP_KERNEL); + if (!macsec_device) { + err = -ENOMEM; + goto out; + } + + macsec_device->dev_addr = kmemdup(dev->dev_addr, dev->addr_len, GFP_KERNEL); + if (!macsec_device->dev_addr) { + kfree(macsec_device); + err = -ENOMEM; + goto out; + } + + macsec_device->netdev = dev; + + INIT_LIST_HEAD_RCU(&macsec_device->macsec_rx_sc_list_head); + list_add_rcu(&macsec_device->macsec_device_list_element, &macsec->macsec_device_list_head); + + ++macsec->num_of_devices; +out: + mutex_unlock(&macsec->lock); + + return err; +} + +static int macsec_upd_secy_hw_address(struct macsec_context *ctx, + struct mlx5e_macsec_device *macsec_device) +{ + struct mlx5e_priv *priv = netdev_priv(ctx->netdev); + const struct net_device *dev = ctx->secy->netdev; + struct mlx5e_macsec *macsec = priv->macsec; + struct mlx5e_macsec_rx_sc *rx_sc, *tmp; + struct mlx5e_macsec_sa *rx_sa; + struct list_head *list; + int i, err = 0; + + + list = &macsec_device->macsec_rx_sc_list_head; + list_for_each_entry_safe(rx_sc, tmp, list, rx_sc_list_element) { + for (i = 0; i < MACSEC_NUM_AN; ++i) { + rx_sa = rx_sc->rx_sa[i]; + if (!rx_sa || !rx_sa->macsec_rule) + continue; + + mlx5e_macsec_cleanup_sa(macsec, rx_sa, false); + } + } + + list_for_each_entry_safe(rx_sc, tmp, list, rx_sc_list_element) { + for (i = 0; i < MACSEC_NUM_AN; ++i) { + rx_sa = rx_sc->rx_sa[i]; + if (!rx_sa) + continue; + + if (rx_sa->active) { + err = mlx5e_macsec_init_sa(ctx, rx_sa, false, false); + if (err) + goto out; + } + } + } + + memcpy(macsec_device->dev_addr, dev->dev_addr, dev->addr_len); +out: + return err; +} + +/* this function is called from 2 macsec ops functions: + * macsec_set_mac_address – MAC address was changed, therefore we need to destroy + * and create new Tx contexts(macsec object + steering). + * macsec_changelink – in this case the tx SC or SecY may be changed, therefore need to + * destroy Tx and Rx contexts(macsec object + steering) + */ +static int mlx5e_macsec_upd_secy(struct macsec_context *ctx) +{ + const struct macsec_tx_sc *tx_sc = &ctx->secy->tx_sc; + struct mlx5e_priv *priv = netdev_priv(ctx->netdev); + const struct net_device *dev = ctx->secy->netdev; + struct mlx5e_macsec_device *macsec_device; + struct mlx5e_macsec_sa *tx_sa; + struct mlx5e_macsec *macsec; + int i, err = 0; + + if (!mlx5e_macsec_secy_features_validate(ctx)) + return -EINVAL; + + mutex_lock(&priv->macsec->lock); + + macsec = priv->macsec; + macsec_device = mlx5e_macsec_get_macsec_device_context(macsec, ctx); + if (!macsec_device) { + netdev_err(ctx->netdev, "MACsec offload: Failed to find device context\n"); + err = -EINVAL; + goto out; + } + + /* if the dev_addr hasn't change, it mean the callback is from macsec_changelink */ + if (!memcmp(macsec_device->dev_addr, dev->dev_addr, dev->addr_len)) { + err = macsec_upd_secy_hw_address(ctx, macsec_device); + if (err) + goto out; + } + + for (i = 0; i < MACSEC_NUM_AN; ++i) { + tx_sa = macsec_device->tx_sa[i]; + if (!tx_sa) + continue; + + mlx5e_macsec_cleanup_sa(macsec, tx_sa, true); + } + + for (i = 0; i < MACSEC_NUM_AN; ++i) { + tx_sa = macsec_device->tx_sa[i]; + if (!tx_sa) + continue; + + if (tx_sa->assoc_num == tx_sc->encoding_sa && tx_sa->active) { + err = mlx5e_macsec_init_sa(ctx, tx_sa, tx_sc->encrypt, true); + if (err) + goto out; + } + } + +out: + mutex_unlock(&macsec->lock); + + return err; +} + +static int mlx5e_macsec_del_secy(struct macsec_context *ctx) +{ + struct mlx5e_priv *priv = netdev_priv(ctx->netdev); + struct mlx5e_macsec_device *macsec_device; + struct mlx5e_macsec_rx_sc *rx_sc, *tmp; + struct mlx5e_macsec_sa *rx_sa; + struct mlx5e_macsec_sa *tx_sa; + struct mlx5e_macsec *macsec; + struct list_head *list; + int err = 0; + int i; + + mutex_lock(&priv->macsec->lock); + macsec = priv->macsec; + macsec_device = mlx5e_macsec_get_macsec_device_context(macsec, ctx); + if (!macsec_device) { + netdev_err(ctx->netdev, "MACsec offload: Failed to find device context\n"); + err = -EINVAL; + + goto out; + } + + for (i = 0; i < MACSEC_NUM_AN; ++i) { + tx_sa = macsec_device->tx_sa[i]; + if (!tx_sa) + continue; + + mlx5e_macsec_cleanup_sa(macsec, tx_sa, true); + mlx5_destroy_encryption_key(macsec->mdev, tx_sa->enc_key_id); + kfree(tx_sa); + macsec_device->tx_sa[i] = NULL; + } + + list = &macsec_device->macsec_rx_sc_list_head; + list_for_each_entry_safe(rx_sc, tmp, list, rx_sc_list_element) { + for (i = 0; i < MACSEC_NUM_AN; ++i) { + rx_sa = rx_sc->rx_sa[i]; + if (!rx_sa) + continue; + + mlx5e_macsec_cleanup_sa(macsec, rx_sa, false); + mlx5_destroy_encryption_key(macsec->mdev, rx_sa->enc_key_id); + kfree(rx_sa); + rx_sc->rx_sa[i] = NULL; + } + + list_del_rcu(&rx_sc->rx_sc_list_element); + + kfree_rcu(rx_sc); + } + + kfree(macsec_device->dev_addr); + macsec_device->dev_addr = NULL; + + list_del_rcu(&macsec_device->macsec_device_list_element); + --macsec->num_of_devices; + +out: + mutex_unlock(&macsec->lock); + + return err; +} + +static void macsec_build_accel_attrs(struct mlx5e_macsec_sa *sa, + struct mlx5_macsec_obj_attrs *attrs) +{ + attrs->epn_state.epn_msb = sa->epn_state.epn_msb; + attrs->epn_state.overlap = sa->epn_state.overlap; +} + +static void macsec_aso_build_wqe_ctrl_seg(struct mlx5e_macsec_aso *macsec_aso, + struct mlx5_wqe_aso_ctrl_seg *aso_ctrl, + struct mlx5_aso_ctrl_param *param) +{ + memset(aso_ctrl, 0, sizeof(*aso_ctrl)); + if (macsec_aso->umr->dma_addr) { + aso_ctrl->va_l = cpu_to_be32(macsec_aso->umr->dma_addr | ASO_CTRL_READ_EN); + aso_ctrl->va_h = cpu_to_be32((u64)macsec_aso->umr->dma_addr >> 32); + aso_ctrl->l_key = cpu_to_be32(macsec_aso->umr->mkey); + } + + if (!param) + return; + + aso_ctrl->data_mask_mode = param->data_mask_mode << 6; + aso_ctrl->condition_1_0_operand = param->condition_1_operand | + param->condition_0_operand << 4; + aso_ctrl->condition_1_0_offset = param->condition_1_offset | + param->condition_0_offset << 4; + aso_ctrl->data_offset_condition_operand = param->data_offset | + param->condition_operand << 6; + aso_ctrl->condition_0_data = cpu_to_be32(param->condition_0_data); + aso_ctrl->condition_0_mask = cpu_to_be32(param->condition_0_mask); + aso_ctrl->condition_1_data = cpu_to_be32(param->condition_1_data); + aso_ctrl->condition_1_mask = cpu_to_be32(param->condition_1_mask); + aso_ctrl->bitwise_data = cpu_to_be64(param->bitwise_data); + aso_ctrl->data_mask = cpu_to_be64(param->data_mask); +} + +static int mlx5e_macsec_modify_obj(struct mlx5_core_dev *mdev, struct mlx5_macsec_obj_attrs *attrs, + u32 macsec_id) +{ + u32 in[MLX5_ST_SZ_DW(modify_macsec_obj_in)] = {}; + u32 out[MLX5_ST_SZ_DW(query_macsec_obj_out)]; + u64 modify_field_select = 0; + void *obj; + int err; + + /* General object fields set */ + MLX5_SET(general_obj_in_cmd_hdr, in, opcode, MLX5_CMD_OP_QUERY_GENERAL_OBJECT); + MLX5_SET(general_obj_in_cmd_hdr, in, obj_type, MLX5_GENERAL_OBJECT_TYPES_MACSEC); + MLX5_SET(general_obj_in_cmd_hdr, in, obj_id, macsec_id); + err = mlx5_cmd_exec(mdev, in, sizeof(in), out, sizeof(out)); + if (err) { + mlx5_core_err(mdev, "Query MACsec object failed (Object id %d), err = %d\n", + macsec_id, err); + return err; + } + + obj = MLX5_ADDR_OF(query_macsec_obj_out, out, macsec_object); + modify_field_select = MLX5_GET64(macsec_offload_obj, obj, modify_field_select); + + /* EPN */ + if (!(modify_field_select & MLX5_MODIFY_MACSEC_BITMASK_EPN_OVERLAP) || + !(modify_field_select & MLX5_MODIFY_MACSEC_BITMASK_EPN_MSB)) { + mlx5_core_dbg(mdev, "MACsec object field is not modifiable (Object id %d)\n", + macsec_id); + return -EOPNOTSUPP; + } + + obj = MLX5_ADDR_OF(modify_macsec_obj_in, in, macsec_object); + MLX5_SET64(macsec_offload_obj, obj, modify_field_select, + MLX5_MODIFY_MACSEC_BITMASK_EPN_OVERLAP | MLX5_MODIFY_MACSEC_BITMASK_EPN_MSB); + MLX5_SET(macsec_offload_obj, obj, epn_msb, attrs->epn_state.epn_msb); + MLX5_SET(macsec_offload_obj, obj, epn_overlap, attrs->epn_state.overlap); + + /* General object fields set */ + MLX5_SET(general_obj_in_cmd_hdr, in, opcode, MLX5_CMD_OP_MODIFY_GENERAL_OBJECT); + + return mlx5_cmd_exec(mdev, in, sizeof(in), out, sizeof(out)); +} + +static void macsec_aso_build_ctrl(struct mlx5e_macsec_aso *aso, + struct mlx5_wqe_aso_ctrl_seg *aso_ctrl, + struct mlx5e_macsec_aso_in *in) +{ + struct mlx5_aso_ctrl_param param = {}; + + param.data_mask_mode = MLX5_ASO_DATA_MASK_MODE_BITWISE_64BIT; + param.condition_0_operand = MLX5_ASO_ALWAYS_TRUE; + param.condition_1_operand = MLX5_ASO_ALWAYS_TRUE; + if (in->mode == MLX5_MACSEC_EPN) { + param.data_offset = MLX5_MACSEC_ASO_REMOVE_FLOW_PKT_CNT_OFFSET; + param.bitwise_data = BIT_ULL(54); + param.data_mask = param.bitwise_data; + } + macsec_aso_build_wqe_ctrl_seg(aso, aso_ctrl, ¶m); +} + +static int macsec_aso_set_arm_event(struct mlx5_core_dev *mdev, struct mlx5e_macsec *macsec, + struct mlx5e_macsec_aso_in *in) +{ + struct mlx5e_macsec_aso *aso; + struct mlx5_aso_wqe *aso_wqe; + struct mlx5_aso *maso; + int err; + + aso = &macsec->aso; + maso = aso->maso; + + mutex_lock(&aso->aso_lock); + aso_wqe = mlx5_aso_get_wqe(maso); + mlx5_aso_build_wqe(maso, MLX5_MACSEC_ASO_DS_CNT, aso_wqe, in->obj_id, + MLX5_ACCESS_ASO_OPC_MOD_MACSEC); + macsec_aso_build_ctrl(aso, &aso_wqe->aso_ctrl, in); + mlx5_aso_post_wqe(maso, false, &aso_wqe->ctrl); + err = mlx5_aso_poll_cq(maso, false, 10); + mutex_unlock(&aso->aso_lock); + + return err; +} + +static int macsec_aso_query(struct mlx5_core_dev *mdev, struct mlx5e_macsec *macsec, + struct mlx5e_macsec_aso_in *in, struct mlx5e_macsec_aso_out *out) +{ + struct mlx5e_macsec_aso *aso; + struct mlx5_aso_wqe *aso_wqe; + struct mlx5_aso *maso; + int err; + + aso = &macsec->aso; + maso = aso->maso; + + mutex_lock(&aso->aso_lock); + + aso_wqe = mlx5_aso_get_wqe(maso); + mlx5_aso_build_wqe(maso, MLX5_MACSEC_ASO_DS_CNT, aso_wqe, in->obj_id, + MLX5_ACCESS_ASO_OPC_MOD_MACSEC); + macsec_aso_build_wqe_ctrl_seg(aso, &aso_wqe->aso_ctrl, NULL); + + mlx5_aso_post_wqe(maso, false, &aso_wqe->ctrl); + err = mlx5_aso_poll_cq(maso, false, 10); + if (err) + goto err_out; + + if (MLX5_GET(macsec_aso, aso->umr->ctx, epn_event_arm)) + out->event_arm |= MLX5E_ASO_EPN_ARM; + + out->mode_param = MLX5_GET(macsec_aso, aso->umr->ctx, mode_parameter); + +err_out: + mutex_unlock(&aso->aso_lock); + return err; +} + +static struct mlx5e_macsec_sa *get_macsec_tx_sa_from_obj_id(const struct mlx5e_macsec *macsec, + const u32 obj_id) +{ + const struct list_head *device_list; + struct mlx5e_macsec_sa *macsec_sa; + struct mlx5e_macsec_device *iter; + int i; + + device_list = &macsec->macsec_device_list_head; + + list_for_each_entry(iter, device_list, macsec_device_list_element) { + for (i = 0; i < MACSEC_NUM_AN; ++i) { + macsec_sa = iter->tx_sa[i]; + if (!macsec_sa || !macsec_sa->active) + continue; + if (macsec_sa->macsec_obj_id == obj_id) + return macsec_sa; + } + } + + return NULL; +} + +static struct mlx5e_macsec_sa *get_macsec_rx_sa_from_obj_id(const struct mlx5e_macsec *macsec, + const u32 obj_id) +{ + const struct list_head *device_list, *sc_list; + struct mlx5e_macsec_rx_sc *mlx5e_rx_sc; + struct mlx5e_macsec_sa *macsec_sa; + struct mlx5e_macsec_device *iter; + int i; + + device_list = &macsec->macsec_device_list_head; + + list_for_each_entry(iter, device_list, macsec_device_list_element) { + sc_list = &iter->macsec_rx_sc_list_head; + list_for_each_entry(mlx5e_rx_sc, sc_list, rx_sc_list_element) { + for (i = 0; i < MACSEC_NUM_AN; ++i) { + macsec_sa = mlx5e_rx_sc->rx_sa[i]; + if (!macsec_sa || !macsec_sa->active) + continue; + if (macsec_sa->macsec_obj_id == obj_id) + return macsec_sa; + } + } + } + + return NULL; +} + +static void macsec_epn_update(struct mlx5e_macsec *macsec, struct mlx5_core_dev *mdev, + struct mlx5e_macsec_sa *sa, u32 obj_id, u32 mode_param) +{ + struct mlx5_macsec_obj_attrs attrs = {}; + struct mlx5e_macsec_aso_in in = {}; + + /* When the bottom of the replay protection window (mode_param) crosses 2^31 (half sequence + * number wraparound) hence mode_param > MLX5_MACSEC_EPN_SCOPE_MID the SW should update the + * esn_overlap to OLD (1). + * When the bottom of the replay protection window (mode_param) crosses 2^32 (full sequence + * number wraparound) hence mode_param < MLX5_MACSEC_EPN_SCOPE_MID since it did a + * wraparound, the SW should update the esn_overlap to NEW (0), and increment the esn_msb. + */ + + if (mode_param < MLX5_MACSEC_EPN_SCOPE_MID) { + sa->epn_state.epn_msb++; + sa->epn_state.overlap = 0; + } else { + sa->epn_state.overlap = 1; + } + + macsec_build_accel_attrs(sa, &attrs); + mlx5e_macsec_modify_obj(mdev, &attrs, obj_id); + + /* Re-set EPN arm event */ + in.obj_id = obj_id; + in.mode = MLX5_MACSEC_EPN; + macsec_aso_set_arm_event(mdev, macsec, &in); +} + +static void macsec_async_event(struct work_struct *work) +{ + struct mlx5e_macsec_async_work *async_work; + struct mlx5e_macsec_aso_out out = {}; + struct mlx5e_macsec_aso_in in = {}; + struct mlx5e_macsec_sa *macsec_sa; + struct mlx5e_macsec *macsec; + struct mlx5_core_dev *mdev; + u32 obj_id; + + async_work = container_of(work, struct mlx5e_macsec_async_work, work); + macsec = async_work->macsec; + mdev = async_work->mdev; + obj_id = async_work->obj_id; + macsec_sa = get_macsec_tx_sa_from_obj_id(macsec, obj_id); + if (!macsec_sa) { + macsec_sa = get_macsec_rx_sa_from_obj_id(macsec, obj_id); + if (!macsec_sa) { + mlx5_core_dbg(mdev, "MACsec SA is not found (SA object id %d)\n", obj_id); + goto out_async_work; + } + } + + /* Query MACsec ASO context */ + in.obj_id = obj_id; + macsec_aso_query(mdev, macsec, &in, &out); + + /* EPN case */ + if (macsec_sa->epn_state.epn_enabled && !(out.event_arm & MLX5E_ASO_EPN_ARM)) + macsec_epn_update(macsec, mdev, macsec_sa, obj_id, out.mode_param); + +out_async_work: + kfree(async_work); +} + +static int macsec_obj_change_event(struct notifier_block *nb, unsigned long event, void *data) +{ + struct mlx5e_macsec *macsec = container_of(nb, struct mlx5e_macsec, nb); + struct mlx5e_macsec_async_work *async_work; + struct mlx5_eqe_obj_change *obj_change; + struct mlx5_eqe *eqe = data; + u16 obj_type; + u32 obj_id; + + if (event != MLX5_EVENT_TYPE_OBJECT_CHANGE) + return NOTIFY_DONE; + + obj_change = &eqe->data.obj_change; + obj_type = be16_to_cpu(obj_change->obj_type); + obj_id = be32_to_cpu(obj_change->obj_id); + + if (obj_type != MLX5_GENERAL_OBJECT_TYPES_MACSEC) + return NOTIFY_DONE; + + async_work = kzalloc(sizeof(*async_work), GFP_ATOMIC); + if (!async_work) + return NOTIFY_DONE; + + async_work->macsec = macsec; + async_work->mdev = macsec->mdev; + async_work->obj_id = obj_id; + + INIT_WORK(&async_work->work, macsec_async_event); + + WARN_ON(!queue_work(macsec->wq, &async_work->work)); + + return NOTIFY_OK; +} + +static int mlx5e_macsec_aso_init(struct mlx5e_macsec_aso *aso, struct mlx5_core_dev *mdev) +{ + struct mlx5_aso *maso; + int err; + + err = mlx5_core_alloc_pd(mdev, &aso->pdn); + if (err) { + mlx5_core_err(mdev, + "MACsec offload: Failed to alloc pd for MACsec ASO, err=%d\n", + err); + return err; + } + + maso = mlx5_aso_create(mdev, aso->pdn); + if (IS_ERR(maso)) { + err = PTR_ERR(maso); + goto err_aso; + } + + err = mlx5e_macsec_aso_reg_mr(mdev, aso); + if (err) + goto err_aso_reg; + + mutex_init(&aso->aso_lock); + + aso->maso = maso; + + return 0; + +err_aso_reg: + mlx5_aso_destroy(maso); +err_aso: + mlx5_core_dealloc_pd(mdev, aso->pdn); + return err; +} + +static void mlx5e_macsec_aso_cleanup(struct mlx5e_macsec_aso *aso, struct mlx5_core_dev *mdev) +{ + if (!aso) + return; + + mlx5e_macsec_aso_dereg_mr(mdev, aso); + + mlx5_aso_destroy(aso->maso); + + mlx5_core_dealloc_pd(mdev, aso->pdn); +} + +bool mlx5e_is_macsec_device(const struct mlx5_core_dev *mdev) +{ + if (!(MLX5_CAP_GEN_64(mdev, general_obj_types) & + MLX5_GENERAL_OBJ_TYPES_CAP_MACSEC_OFFLOAD)) + return false; + + if (!MLX5_CAP_GEN(mdev, log_max_dek)) + return false; + + if (!MLX5_CAP_MACSEC(mdev, log_max_macsec_offload)) + return false; + + if (!MLX5_CAP_FLOWTABLE_NIC_RX(mdev, macsec_decrypt) || + !MLX5_CAP_FLOWTABLE_NIC_RX(mdev, reformat_remove_macsec)) + return false; + + if (!MLX5_CAP_FLOWTABLE_NIC_TX(mdev, macsec_encrypt) || + !MLX5_CAP_FLOWTABLE_NIC_TX(mdev, reformat_add_macsec)) + return false; + + if (!MLX5_CAP_MACSEC(mdev, macsec_crypto_esp_aes_gcm_128_encrypt) && + !MLX5_CAP_MACSEC(mdev, macsec_crypto_esp_aes_gcm_256_encrypt)) + return false; + + if (!MLX5_CAP_MACSEC(mdev, macsec_crypto_esp_aes_gcm_128_decrypt) && + !MLX5_CAP_MACSEC(mdev, macsec_crypto_esp_aes_gcm_256_decrypt)) + return false; + + return true; +} + +void mlx5e_macsec_get_stats_fill(struct mlx5e_macsec *macsec, void *macsec_stats) +{ + mlx5e_macsec_fs_get_stats_fill(macsec->macsec_fs, macsec_stats); +} + +struct mlx5e_macsec_stats *mlx5e_macsec_get_stats(struct mlx5e_macsec *macsec) +{ + if (!macsec) + return NULL; + + return &macsec->stats; +} + +static const struct macsec_ops macsec_offload_ops = { + .mdo_add_txsa = mlx5e_macsec_add_txsa, + .mdo_upd_txsa = mlx5e_macsec_upd_txsa, + .mdo_del_txsa = mlx5e_macsec_del_txsa, + .mdo_add_rxsc = mlx5e_macsec_add_rxsc, + .mdo_upd_rxsc = mlx5e_macsec_upd_rxsc, + .mdo_del_rxsc = mlx5e_macsec_del_rxsc, + .mdo_add_rxsa = mlx5e_macsec_add_rxsa, + .mdo_upd_rxsa = mlx5e_macsec_upd_rxsa, + .mdo_del_rxsa = mlx5e_macsec_del_rxsa, + .mdo_add_secy = mlx5e_macsec_add_secy, + .mdo_upd_secy = mlx5e_macsec_upd_secy, + .mdo_del_secy = mlx5e_macsec_del_secy, +}; + +bool mlx5e_macsec_handle_tx_skb(struct mlx5e_macsec *macsec, struct sk_buff *skb) +{ + struct metadata_dst *md_dst = skb_metadata_dst(skb); + u32 fs_id; + + fs_id = mlx5e_macsec_get_sa_from_hashtable(&macsec->sci_hash, &md_dst->u.macsec_info.sci); + if (!fs_id) + goto err_out; + + return true; + +err_out: + dev_kfree_skb_any(skb); + return false; +} + +void mlx5e_macsec_tx_build_eseg(struct mlx5e_macsec *macsec, + struct sk_buff *skb, + struct mlx5_wqe_eth_seg *eseg) +{ + struct metadata_dst *md_dst = skb_metadata_dst(skb); + u32 fs_id; + + fs_id = mlx5e_macsec_get_sa_from_hashtable(&macsec->sci_hash, &md_dst->u.macsec_info.sci); + if (!fs_id) + return; + + eseg->flow_table_metadata = cpu_to_be32(MLX5_ETH_WQE_FT_META_MACSEC | fs_id << 2); +} + +void mlx5e_macsec_offload_handle_rx_skb(struct net_device *netdev, + struct sk_buff *skb, + struct mlx5_cqe64 *cqe) +{ + struct mlx5e_macsec_rx_sc_xarray_element *sc_xarray_element; + u32 macsec_meta_data = be32_to_cpu(cqe->ft_metadata); + struct mlx5e_priv *priv = netdev_priv(netdev); + struct mlx5e_macsec_rx_sc *rx_sc; + struct mlx5e_macsec *macsec; + u32 fs_id; + + macsec = priv->macsec; + if (!macsec) + return; + + fs_id = MLX5_MACSEC_METADATA_HANDLE(macsec_meta_data); + + rcu_read_lock(); + sc_xarray_element = xa_load(&macsec->sc_xarray, fs_id); + rx_sc = sc_xarray_element->rx_sc; + if (rx_sc) { + dst_hold(&rx_sc->md_dst->dst); + skb_dst_set(skb, &rx_sc->md_dst->dst); + } + + rcu_read_unlock(); +} + +void mlx5e_macsec_build_netdev(struct mlx5e_priv *priv) +{ + struct net_device *netdev = priv->netdev; + + if (!mlx5e_is_macsec_device(priv->mdev)) + return; + + /* Enable MACsec */ + mlx5_core_dbg(priv->mdev, "mlx5e: MACsec acceleration enabled\n"); + netdev->macsec_ops = &macsec_offload_ops; + netdev->features |= NETIF_F_HW_MACSEC; + netif_keep_dst(netdev); +} + +int mlx5e_macsec_init(struct mlx5e_priv *priv) +{ + struct mlx5_core_dev *mdev = priv->mdev; + struct mlx5e_macsec *macsec = NULL; + struct mlx5e_macsec_fs *macsec_fs; + int err; + + if (!mlx5e_is_macsec_device(priv->mdev)) { + mlx5_core_dbg(mdev, "Not a MACsec offload device\n"); + return 0; + } + + macsec = kzalloc(sizeof(*macsec), GFP_KERNEL); + if (!macsec) + return -ENOMEM; + + INIT_LIST_HEAD(&macsec->macsec_device_list_head); + mutex_init(&macsec->lock); + + err = rhashtable_init(&macsec->sci_hash, &rhash_sci); + if (err) { + mlx5_core_err(mdev, "MACsec offload: Failed to init SCI hash table, err=%d\n", + err); + goto err_hash; + } + + err = mlx5e_macsec_aso_init(&macsec->aso, priv->mdev); + if (err) { + mlx5_core_err(mdev, "MACsec offload: Failed to init aso, err=%d\n", err); + goto err_aso; + } + + macsec->wq = alloc_ordered_workqueue("mlx5e_macsec_%s", 0, priv->netdev->name); + if (!macsec->wq) { + err = -ENOMEM; + goto err_wq; + } + + xa_init_flags(&macsec->sc_xarray, XA_FLAGS_ALLOC1); + + priv->macsec = macsec; + + macsec->mdev = mdev; + + macsec_fs = mlx5e_macsec_fs_init(mdev, priv->netdev); + if (!macsec_fs) { + err = -ENOMEM; + goto err_out; + } + + macsec->macsec_fs = macsec_fs; + + macsec->nb.notifier_call = macsec_obj_change_event; + mlx5_notifier_register(mdev, &macsec->nb); + + mlx5_core_dbg(mdev, "MACsec attached to netdevice\n"); + + return 0; + +err_out: + destroy_workqueue(macsec->wq); +err_wq: + mlx5e_macsec_aso_cleanup(&macsec->aso, priv->mdev); +err_aso: + rhashtable_destroy(&macsec->sci_hash); +err_hash: + kfree(macsec); + priv->macsec = NULL; + return err; +} + +void mlx5e_macsec_cleanup(struct mlx5e_priv *priv) +{ + struct mlx5e_macsec *macsec = priv->macsec; + struct mlx5_core_dev *mdev = macsec->mdev; + + if (!macsec) + return; + + mlx5_notifier_unregister(mdev, &macsec->nb); + + mlx5e_macsec_fs_cleanup(macsec->macsec_fs); + + /* Cleanup workqueue */ + destroy_workqueue(macsec->wq); + + mlx5e_macsec_aso_cleanup(&macsec->aso, mdev); + + priv->macsec = NULL; + + rhashtable_destroy(&macsec->sci_hash); + + mutex_destroy(&macsec->lock); + + kfree(macsec); +} diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/macsec.h b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/macsec.h new file mode 100644 index 000000000000..d580b4a91253 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/macsec.h @@ -0,0 +1,71 @@ +/* SPDX-License-Identifier: GPL-2.0 OR Linux-OpenIB */ +/* Copyright (c) 2022, NVIDIA CORPORATION & AFFILIATES. All rights reserved. */ + +#ifndef __MLX5_EN_ACCEL_MACSEC_H__ +#define __MLX5_EN_ACCEL_MACSEC_H__ + +#ifdef CONFIG_MLX5_EN_MACSEC + +#include <linux/mlx5/driver.h> +#include <net/macsec.h> +#include <net/dst_metadata.h> + +/* Bit31 - 30: MACsec marker, Bit3-0: MACsec id */ +#define MLX5_MACSEC_METADATA_MARKER(metadata) ((((metadata) >> 30) & 0x3) == 0x1) +#define MLX5_MACSEC_METADATA_HANDLE(metadata) ((metadata) & GENMASK(3, 0)) + +struct mlx5e_priv; +struct mlx5e_macsec; + +struct mlx5e_macsec_stats { + u64 macsec_rx_pkts; + u64 macsec_rx_bytes; + u64 macsec_rx_pkts_drop; + u64 macsec_rx_bytes_drop; + u64 macsec_tx_pkts; + u64 macsec_tx_bytes; + u64 macsec_tx_pkts_drop; + u64 macsec_tx_bytes_drop; +}; + +void mlx5e_macsec_build_netdev(struct mlx5e_priv *priv); +int mlx5e_macsec_init(struct mlx5e_priv *priv); +void mlx5e_macsec_cleanup(struct mlx5e_priv *priv); +bool mlx5e_macsec_handle_tx_skb(struct mlx5e_macsec *macsec, struct sk_buff *skb); +void mlx5e_macsec_tx_build_eseg(struct mlx5e_macsec *macsec, + struct sk_buff *skb, + struct mlx5_wqe_eth_seg *eseg); + +static inline bool mlx5e_macsec_skb_is_offload(struct sk_buff *skb) +{ + struct metadata_dst *md_dst = skb_metadata_dst(skb); + + return md_dst && (md_dst->type == METADATA_MACSEC); +} + +static inline bool mlx5e_macsec_is_rx_flow(struct mlx5_cqe64 *cqe) +{ + return MLX5_MACSEC_METADATA_MARKER(be32_to_cpu(cqe->ft_metadata)); +} + +void mlx5e_macsec_offload_handle_rx_skb(struct net_device *netdev, struct sk_buff *skb, + struct mlx5_cqe64 *cqe); +bool mlx5e_is_macsec_device(const struct mlx5_core_dev *mdev); +void mlx5e_macsec_get_stats_fill(struct mlx5e_macsec *macsec, void *macsec_stats); +struct mlx5e_macsec_stats *mlx5e_macsec_get_stats(struct mlx5e_macsec *macsec); + +#else + +static inline void mlx5e_macsec_build_netdev(struct mlx5e_priv *priv) {} +static inline int mlx5e_macsec_init(struct mlx5e_priv *priv) { return 0; } +static inline void mlx5e_macsec_cleanup(struct mlx5e_priv *priv) {} +static inline bool mlx5e_macsec_skb_is_offload(struct sk_buff *skb) { return false; } +static inline bool mlx5e_macsec_is_rx_flow(struct mlx5_cqe64 *cqe) { return false; } +static inline void mlx5e_macsec_offload_handle_rx_skb(struct net_device *netdev, + struct sk_buff *skb, + struct mlx5_cqe64 *cqe) +{} +static inline bool mlx5e_is_macsec_device(const struct mlx5_core_dev *mdev) { return false; } +#endif /* CONFIG_MLX5_EN_MACSEC */ + +#endif /* __MLX5_ACCEL_EN_MACSEC_H__ */ diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/macsec_fs.c b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/macsec_fs.c new file mode 100644 index 000000000000..13dc628b988a --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/macsec_fs.c @@ -0,0 +1,1384 @@ +// SPDX-License-Identifier: GPL-2.0 OR Linux-OpenIB +/* Copyright (c) 2022, NVIDIA CORPORATION & AFFILIATES. All rights reserved. */ + +#include <net/macsec.h> +#include <linux/netdevice.h> +#include <linux/mlx5/qp.h> +#include "fs_core.h" +#include "en/fs.h" +#include "en_accel/macsec_fs.h" +#include "mlx5_core.h" + +/* MACsec TX flow steering */ +#define CRYPTO_NUM_MAXSEC_FTE BIT(15) +#define CRYPTO_TABLE_DEFAULT_RULE_GROUP_SIZE 1 + +#define TX_CRYPTO_TABLE_LEVEL 0 +#define TX_CRYPTO_TABLE_NUM_GROUPS 3 +#define TX_CRYPTO_TABLE_MKE_GROUP_SIZE 1 +#define TX_CRYPTO_TABLE_SA_GROUP_SIZE \ + (CRYPTO_NUM_MAXSEC_FTE - (TX_CRYPTO_TABLE_MKE_GROUP_SIZE + \ + CRYPTO_TABLE_DEFAULT_RULE_GROUP_SIZE)) +#define TX_CHECK_TABLE_LEVEL 1 +#define TX_CHECK_TABLE_NUM_FTE 2 +#define RX_CRYPTO_TABLE_LEVEL 0 +#define RX_CHECK_TABLE_LEVEL 1 +#define RX_CHECK_TABLE_NUM_FTE 3 +#define RX_CRYPTO_TABLE_NUM_GROUPS 3 +#define RX_CRYPTO_TABLE_SA_RULE_WITH_SCI_GROUP_SIZE \ + ((CRYPTO_NUM_MAXSEC_FTE - CRYPTO_TABLE_DEFAULT_RULE_GROUP_SIZE) / 2) +#define RX_CRYPTO_TABLE_SA_RULE_WITHOUT_SCI_GROUP_SIZE \ + (CRYPTO_NUM_MAXSEC_FTE - RX_CRYPTO_TABLE_SA_RULE_WITH_SCI_GROUP_SIZE) +#define RX_NUM_OF_RULES_PER_SA 2 + +#define MLX5_MACSEC_TAG_LEN 8 /* SecTAG length with ethertype and without the optional SCI */ +#define MLX5_MACSEC_SECTAG_TCI_AN_FIELD_BITMASK 0x23 +#define MLX5_MACSEC_SECTAG_TCI_AN_FIELD_OFFSET 0x8 +#define MLX5_MACSEC_SECTAG_TCI_SC_FIELD_OFFSET 0x5 +#define MLX5_MACSEC_SECTAG_TCI_SC_FIELD_BIT (0x1 << MLX5_MACSEC_SECTAG_TCI_SC_FIELD_OFFSET) +#define MLX5_SECTAG_HEADER_SIZE_WITHOUT_SCI 0x8 +#define MLX5_SECTAG_HEADER_SIZE_WITH_SCI (MLX5_SECTAG_HEADER_SIZE_WITHOUT_SCI + MACSEC_SCI_LEN) + +/* MACsec RX flow steering */ +#define MLX5_ETH_WQE_FT_META_MACSEC_MASK 0x3E + +struct mlx5_sectag_header { + __be16 ethertype; + u8 tci_an; + u8 sl; + u32 pn; + u8 sci[MACSEC_SCI_LEN]; /* optional */ +} __packed; + +struct mlx5e_macsec_tx_rule { + struct mlx5_flow_handle *rule; + struct mlx5_pkt_reformat *pkt_reformat; + u32 fs_id; +}; + +struct mlx5e_macsec_tables { + struct mlx5e_flow_table ft_crypto; + struct mlx5_flow_handle *crypto_miss_rule; + + struct mlx5_flow_table *ft_check; + struct mlx5_flow_group *ft_check_group; + struct mlx5_fc *check_miss_rule_counter; + struct mlx5_flow_handle *check_miss_rule; + struct mlx5_fc *check_rule_counter; + + u32 refcnt; +}; + +struct mlx5e_macsec_tx { + struct mlx5_flow_handle *crypto_mke_rule; + struct mlx5_flow_handle *check_rule; + + struct ida tx_halloc; + + struct mlx5e_macsec_tables tables; +}; + +struct mlx5e_macsec_rx_rule { + struct mlx5_flow_handle *rule[RX_NUM_OF_RULES_PER_SA]; + struct mlx5_modify_hdr *meta_modhdr; +}; + +struct mlx5e_macsec_rx { + struct mlx5_flow_handle *check_rule[2]; + struct mlx5_pkt_reformat *check_rule_pkt_reformat[2]; + + struct mlx5e_macsec_tables tables; +}; + +union mlx5e_macsec_rule { + struct mlx5e_macsec_tx_rule tx_rule; + struct mlx5e_macsec_rx_rule rx_rule; +}; + +struct mlx5e_macsec_fs { + struct mlx5_core_dev *mdev; + struct net_device *netdev; + struct mlx5e_macsec_tx *tx_fs; + struct mlx5e_macsec_rx *rx_fs; +}; + +static void macsec_fs_tx_destroy(struct mlx5e_macsec_fs *macsec_fs) +{ + struct mlx5e_macsec_tx *tx_fs = macsec_fs->tx_fs; + struct mlx5e_macsec_tables *tx_tables; + + tx_tables = &tx_fs->tables; + + /* Tx check table */ + if (tx_fs->check_rule) { + mlx5_del_flow_rules(tx_fs->check_rule); + tx_fs->check_rule = NULL; + } + + if (tx_tables->check_miss_rule) { + mlx5_del_flow_rules(tx_tables->check_miss_rule); + tx_tables->check_miss_rule = NULL; + } + + if (tx_tables->ft_check_group) { + mlx5_destroy_flow_group(tx_tables->ft_check_group); + tx_tables->ft_check_group = NULL; + } + + if (tx_tables->ft_check) { + mlx5_destroy_flow_table(tx_tables->ft_check); + tx_tables->ft_check = NULL; + } + + /* Tx crypto table */ + if (tx_fs->crypto_mke_rule) { + mlx5_del_flow_rules(tx_fs->crypto_mke_rule); + tx_fs->crypto_mke_rule = NULL; + } + + if (tx_tables->crypto_miss_rule) { + mlx5_del_flow_rules(tx_tables->crypto_miss_rule); + tx_tables->crypto_miss_rule = NULL; + } + + mlx5e_destroy_flow_table(&tx_tables->ft_crypto); +} + +static int macsec_fs_tx_create_crypto_table_groups(struct mlx5e_flow_table *ft) +{ + int inlen = MLX5_ST_SZ_BYTES(create_flow_group_in); + int mclen = MLX5_ST_SZ_BYTES(fte_match_param); + int ix = 0; + u32 *in; + int err; + u8 *mc; + + ft->g = kcalloc(TX_CRYPTO_TABLE_NUM_GROUPS, sizeof(*ft->g), GFP_KERNEL); + if (!ft->g) + return -ENOMEM; + in = kvzalloc(inlen, GFP_KERNEL); + + if (!in) { + kfree(ft->g); + return -ENOMEM; + } + + mc = MLX5_ADDR_OF(create_flow_group_in, in, match_criteria); + + /* Flow Group for MKE match */ + MLX5_SET_CFG(in, match_criteria_enable, MLX5_MATCH_OUTER_HEADERS); + MLX5_SET_TO_ONES(fte_match_param, mc, outer_headers.ethertype); + + MLX5_SET_CFG(in, start_flow_index, ix); + ix += TX_CRYPTO_TABLE_MKE_GROUP_SIZE; + MLX5_SET_CFG(in, end_flow_index, ix - 1); + ft->g[ft->num_groups] = mlx5_create_flow_group(ft->t, in); + if (IS_ERR(ft->g[ft->num_groups])) + goto err; + ft->num_groups++; + + /* Flow Group for SA rules */ + memset(in, 0, inlen); + memset(mc, 0, mclen); + MLX5_SET_CFG(in, match_criteria_enable, MLX5_MATCH_MISC_PARAMETERS_2); + MLX5_SET(fte_match_param, mc, misc_parameters_2.metadata_reg_a, + MLX5_ETH_WQE_FT_META_MACSEC_MASK); + + MLX5_SET_CFG(in, start_flow_index, ix); + ix += TX_CRYPTO_TABLE_SA_GROUP_SIZE; + MLX5_SET_CFG(in, end_flow_index, ix - 1); + ft->g[ft->num_groups] = mlx5_create_flow_group(ft->t, in); + if (IS_ERR(ft->g[ft->num_groups])) + goto err; + ft->num_groups++; + + /* Flow Group for l2 traps */ + memset(in, 0, inlen); + memset(mc, 0, mclen); + MLX5_SET_CFG(in, start_flow_index, ix); + ix += CRYPTO_TABLE_DEFAULT_RULE_GROUP_SIZE; + MLX5_SET_CFG(in, end_flow_index, ix - 1); + ft->g[ft->num_groups] = mlx5_create_flow_group(ft->t, in); + if (IS_ERR(ft->g[ft->num_groups])) + goto err; + ft->num_groups++; + + kvfree(in); + return 0; + +err: + err = PTR_ERR(ft->g[ft->num_groups]); + ft->g[ft->num_groups] = NULL; + kvfree(in); + + return err; +} + +static struct mlx5_flow_table + *macsec_fs_auto_group_table_create(struct mlx5_flow_namespace *ns, int flags, + int level, int max_fte) +{ + struct mlx5_flow_table_attr ft_attr = {}; + struct mlx5_flow_table *fdb = NULL; + + /* reserve entry for the match all miss group and rule */ + ft_attr.autogroup.num_reserved_entries = 1; + ft_attr.autogroup.max_num_groups = 1; + ft_attr.prio = 0; + ft_attr.flags = flags; + ft_attr.level = level; + ft_attr.max_fte = max_fte; + + fdb = mlx5_create_auto_grouped_flow_table(ns, &ft_attr); + + return fdb; +} + +static int macsec_fs_tx_create(struct mlx5e_macsec_fs *macsec_fs) +{ + int inlen = MLX5_ST_SZ_BYTES(create_flow_group_in); + struct mlx5e_macsec_tx *tx_fs = macsec_fs->tx_fs; + struct net_device *netdev = macsec_fs->netdev; + struct mlx5_flow_table_attr ft_attr = {}; + struct mlx5_flow_destination dest = {}; + struct mlx5e_macsec_tables *tx_tables; + struct mlx5_flow_act flow_act = {}; + struct mlx5e_flow_table *ft_crypto; + struct mlx5_flow_table *flow_table; + struct mlx5_flow_group *flow_group; + struct mlx5_flow_namespace *ns; + struct mlx5_flow_handle *rule; + struct mlx5_flow_spec *spec; + u32 *flow_group_in; + int err = 0; + + ns = mlx5_get_flow_namespace(macsec_fs->mdev, MLX5_FLOW_NAMESPACE_EGRESS_MACSEC); + if (!ns) + return -ENOMEM; + + spec = kvzalloc(sizeof(*spec), GFP_KERNEL); + if (!spec) + return -ENOMEM; + + flow_group_in = kvzalloc(inlen, GFP_KERNEL); + if (!flow_group_in) + goto out_spec; + + tx_tables = &tx_fs->tables; + ft_crypto = &tx_tables->ft_crypto; + + /* Tx crypto table */ + ft_attr.flags = MLX5_FLOW_TABLE_TUNNEL_EN_REFORMAT; + ft_attr.level = TX_CRYPTO_TABLE_LEVEL; + ft_attr.max_fte = CRYPTO_NUM_MAXSEC_FTE; + + flow_table = mlx5_create_flow_table(ns, &ft_attr); + if (IS_ERR(flow_table)) { + err = PTR_ERR(flow_table); + netdev_err(netdev, "Failed to create MACsec Tx crypto table err(%d)\n", err); + goto out_flow_group; + } + ft_crypto->t = flow_table; + + /* Tx crypto table groups */ + err = macsec_fs_tx_create_crypto_table_groups(ft_crypto); + if (err) { + netdev_err(netdev, + "Failed to create default flow group for MACsec Tx crypto table err(%d)\n", + err); + goto err; + } + + /* Tx crypto table MKE rule - MKE packets shouldn't be offloaded */ + memset(&flow_act, 0, sizeof(flow_act)); + memset(spec, 0, sizeof(*spec)); + spec->match_criteria_enable = MLX5_MATCH_OUTER_HEADERS; + + MLX5_SET_TO_ONES(fte_match_param, spec->match_criteria, outer_headers.ethertype); + MLX5_SET(fte_match_param, spec->match_value, outer_headers.ethertype, ETH_P_PAE); + flow_act.action = MLX5_FLOW_CONTEXT_ACTION_ALLOW; + + rule = mlx5_add_flow_rules(ft_crypto->t, spec, &flow_act, NULL, 0); + if (IS_ERR(rule)) { + err = PTR_ERR(rule); + netdev_err(netdev, "Failed to add MACsec TX MKE rule, err=%d\n", err); + goto err; + } + tx_fs->crypto_mke_rule = rule; + + /* Tx crypto table Default miss rule */ + memset(&flow_act, 0, sizeof(flow_act)); + flow_act.action = MLX5_FLOW_CONTEXT_ACTION_ALLOW; + rule = mlx5_add_flow_rules(ft_crypto->t, NULL, &flow_act, NULL, 0); + if (IS_ERR(rule)) { + err = PTR_ERR(rule); + netdev_err(netdev, "Failed to add MACsec Tx table default miss rule %d\n", err); + goto err; + } + tx_tables->crypto_miss_rule = rule; + + /* Tx check table */ + flow_table = macsec_fs_auto_group_table_create(ns, 0, TX_CHECK_TABLE_LEVEL, + TX_CHECK_TABLE_NUM_FTE); + if (IS_ERR(flow_table)) { + err = PTR_ERR(flow_table); + netdev_err(netdev, "fail to create MACsec TX check table, err(%d)\n", err); + goto err; + } + tx_tables->ft_check = flow_table; + + /* Tx check table Default miss group/rule */ + memset(flow_group_in, 0, inlen); + MLX5_SET(create_flow_group_in, flow_group_in, start_flow_index, flow_table->max_fte - 1); + MLX5_SET(create_flow_group_in, flow_group_in, end_flow_index, flow_table->max_fte - 1); + flow_group = mlx5_create_flow_group(tx_tables->ft_check, flow_group_in); + if (IS_ERR(flow_group)) { + err = PTR_ERR(flow_group); + netdev_err(netdev, + "Failed to create default flow group for MACsec Tx crypto table err(%d)\n", + err); + goto err; + } + tx_tables->ft_check_group = flow_group; + + /* Tx check table default drop rule */ + memset(&dest, 0, sizeof(struct mlx5_flow_destination)); + memset(&flow_act, 0, sizeof(flow_act)); + dest.type = MLX5_FLOW_DESTINATION_TYPE_COUNTER; + dest.counter_id = mlx5_fc_id(tx_tables->check_miss_rule_counter); + flow_act.action = MLX5_FLOW_CONTEXT_ACTION_DROP | MLX5_FLOW_CONTEXT_ACTION_COUNT; + rule = mlx5_add_flow_rules(tx_tables->ft_check, NULL, &flow_act, &dest, 1); + if (IS_ERR(rule)) { + err = PTR_ERR(rule); + netdev_err(netdev, "Failed to added MACsec tx check drop rule, err(%d)\n", err); + goto err; + } + tx_tables->check_miss_rule = rule; + + /* Tx check table rule */ + memset(spec, 0, sizeof(struct mlx5_flow_spec)); + memset(&dest, 0, sizeof(struct mlx5_flow_destination)); + memset(&flow_act, 0, sizeof(flow_act)); + + MLX5_SET_TO_ONES(fte_match_param, spec->match_criteria, misc_parameters_2.metadata_reg_c_4); + MLX5_SET(fte_match_param, spec->match_value, misc_parameters_2.metadata_reg_c_4, 0); + spec->match_criteria_enable = MLX5_MATCH_MISC_PARAMETERS_2; + + flow_act.flags = FLOW_ACT_NO_APPEND; + flow_act.action = MLX5_FLOW_CONTEXT_ACTION_ALLOW | MLX5_FLOW_CONTEXT_ACTION_COUNT; + dest.type = MLX5_FLOW_DESTINATION_TYPE_COUNTER; + dest.counter_id = mlx5_fc_id(tx_tables->check_rule_counter); + rule = mlx5_add_flow_rules(tx_tables->ft_check, spec, &flow_act, &dest, 1); + if (IS_ERR(rule)) { + err = PTR_ERR(rule); + netdev_err(netdev, "Failed to add MACsec check rule, err=%d\n", err); + goto err; + } + tx_fs->check_rule = rule; + + goto out_flow_group; + +err: + macsec_fs_tx_destroy(macsec_fs); +out_flow_group: + kvfree(flow_group_in); +out_spec: + kvfree(spec); + return err; +} + +static int macsec_fs_tx_ft_get(struct mlx5e_macsec_fs *macsec_fs) +{ + struct mlx5e_macsec_tx *tx_fs = macsec_fs->tx_fs; + struct mlx5e_macsec_tables *tx_tables; + int err = 0; + + tx_tables = &tx_fs->tables; + if (tx_tables->refcnt) + goto out; + + err = macsec_fs_tx_create(macsec_fs); + if (err) + return err; + +out: + tx_tables->refcnt++; + return err; +} + +static void macsec_fs_tx_ft_put(struct mlx5e_macsec_fs *macsec_fs) +{ + struct mlx5e_macsec_tables *tx_tables = &macsec_fs->tx_fs->tables; + + if (--tx_tables->refcnt) + return; + + macsec_fs_tx_destroy(macsec_fs); +} + +static int macsec_fs_tx_setup_fte(struct mlx5e_macsec_fs *macsec_fs, + struct mlx5_flow_spec *spec, + struct mlx5_flow_act *flow_act, + u32 macsec_obj_id, + u32 *fs_id) +{ + struct mlx5e_macsec_tx *tx_fs = macsec_fs->tx_fs; + int err = 0; + u32 id; + + err = ida_alloc_range(&tx_fs->tx_halloc, 1, + MLX5_MACSEC_NUM_OF_SUPPORTED_INTERFACES, + GFP_KERNEL); + if (err < 0) + return err; + + id = err; + spec->match_criteria_enable |= MLX5_MATCH_MISC_PARAMETERS_2; + + /* Metadata match */ + MLX5_SET(fte_match_param, spec->match_criteria, misc_parameters_2.metadata_reg_a, + MLX5_ETH_WQE_FT_META_MACSEC_MASK); + MLX5_SET(fte_match_param, spec->match_value, misc_parameters_2.metadata_reg_a, + MLX5_ETH_WQE_FT_META_MACSEC | id << 2); + + *fs_id = id; + flow_act->crypto.type = MLX5_FLOW_CONTEXT_ENCRYPT_DECRYPT_TYPE_MACSEC; + flow_act->crypto.obj_id = macsec_obj_id; + + mlx5_core_dbg(macsec_fs->mdev, "Tx fte: macsec obj_id %u, fs_id %u\n", macsec_obj_id, id); + return 0; +} + +static void macsec_fs_tx_create_sectag_header(const struct macsec_context *ctx, + char *reformatbf, + size_t *reformat_size) +{ + const struct macsec_secy *secy = ctx->secy; + bool sci_present = macsec_send_sci(secy); + struct mlx5_sectag_header sectag = {}; + const struct macsec_tx_sc *tx_sc; + + tx_sc = &secy->tx_sc; + sectag.ethertype = htons(ETH_P_MACSEC); + + if (sci_present) { + sectag.tci_an |= MACSEC_TCI_SC; + memcpy(§ag.sci, &secy->sci, + sizeof(sectag.sci)); + } else { + if (tx_sc->end_station) + sectag.tci_an |= MACSEC_TCI_ES; + if (tx_sc->scb) + sectag.tci_an |= MACSEC_TCI_SCB; + } + + /* With GCM, C/E clear for !encrypt, both set for encrypt */ + if (tx_sc->encrypt) + sectag.tci_an |= MACSEC_TCI_CONFID; + else if (secy->icv_len != MACSEC_DEFAULT_ICV_LEN) + sectag.tci_an |= MACSEC_TCI_C; + + sectag.tci_an |= tx_sc->encoding_sa; + + *reformat_size = MLX5_MACSEC_TAG_LEN + (sci_present ? MACSEC_SCI_LEN : 0); + + memcpy(reformatbf, §ag, *reformat_size); +} + +static void macsec_fs_tx_del_rule(struct mlx5e_macsec_fs *macsec_fs, + struct mlx5e_macsec_tx_rule *tx_rule) +{ + if (tx_rule->rule) { + mlx5_del_flow_rules(tx_rule->rule); + tx_rule->rule = NULL; + } + + if (tx_rule->pkt_reformat) { + mlx5_packet_reformat_dealloc(macsec_fs->mdev, tx_rule->pkt_reformat); + tx_rule->pkt_reformat = NULL; + } + + if (tx_rule->fs_id) { + ida_free(&macsec_fs->tx_fs->tx_halloc, tx_rule->fs_id); + tx_rule->fs_id = 0; + } + + kfree(tx_rule); + + macsec_fs_tx_ft_put(macsec_fs); +} + +static union mlx5e_macsec_rule * +macsec_fs_tx_add_rule(struct mlx5e_macsec_fs *macsec_fs, + const struct macsec_context *macsec_ctx, + struct mlx5_macsec_rule_attrs *attrs, + u32 *sa_fs_id) +{ + char reformatbf[MLX5_MACSEC_TAG_LEN + MACSEC_SCI_LEN]; + struct mlx5_pkt_reformat_params reformat_params = {}; + struct mlx5e_macsec_tx *tx_fs = macsec_fs->tx_fs; + struct net_device *netdev = macsec_fs->netdev; + union mlx5e_macsec_rule *macsec_rule = NULL; + struct mlx5_flow_destination dest = {}; + struct mlx5e_macsec_tables *tx_tables; + struct mlx5e_macsec_tx_rule *tx_rule; + struct mlx5_flow_act flow_act = {}; + struct mlx5_flow_handle *rule; + struct mlx5_flow_spec *spec; + size_t reformat_size; + int err = 0; + u32 fs_id; + + tx_tables = &tx_fs->tables; + + spec = kvzalloc(sizeof(*spec), GFP_KERNEL); + if (!spec) + return NULL; + + err = macsec_fs_tx_ft_get(macsec_fs); + if (err) + goto out_spec; + + macsec_rule = kzalloc(sizeof(*macsec_rule), GFP_KERNEL); + if (!macsec_rule) { + macsec_fs_tx_ft_put(macsec_fs); + goto out_spec; + } + + tx_rule = &macsec_rule->tx_rule; + + /* Tx crypto table crypto rule */ + macsec_fs_tx_create_sectag_header(macsec_ctx, reformatbf, &reformat_size); + + reformat_params.type = MLX5_REFORMAT_TYPE_ADD_MACSEC; + reformat_params.size = reformat_size; + reformat_params.data = reformatbf; + flow_act.pkt_reformat = mlx5_packet_reformat_alloc(macsec_fs->mdev, + &reformat_params, + MLX5_FLOW_NAMESPACE_EGRESS_MACSEC); + if (IS_ERR(flow_act.pkt_reformat)) { + err = PTR_ERR(flow_act.pkt_reformat); + netdev_err(netdev, "Failed to allocate MACsec Tx reformat context err=%d\n", err); + goto err; + } + tx_rule->pkt_reformat = flow_act.pkt_reformat; + + err = macsec_fs_tx_setup_fte(macsec_fs, spec, &flow_act, attrs->macsec_obj_id, &fs_id); + if (err) { + netdev_err(netdev, + "Failed to add packet reformat for MACsec TX crypto rule, err=%d\n", + err); + goto err; + } + + tx_rule->fs_id = fs_id; + *sa_fs_id = fs_id; + + flow_act.action = MLX5_FLOW_CONTEXT_ACTION_FWD_DEST | + MLX5_FLOW_CONTEXT_ACTION_CRYPTO_ENCRYPT | + MLX5_FLOW_CONTEXT_ACTION_PACKET_REFORMAT; + dest.type = MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE; + dest.ft = tx_tables->ft_check; + rule = mlx5_add_flow_rules(tx_tables->ft_crypto.t, spec, &flow_act, &dest, 1); + if (IS_ERR(rule)) { + err = PTR_ERR(rule); + netdev_err(netdev, "Failed to add MACsec TX crypto rule, err=%d\n", err); + goto err; + } + tx_rule->rule = rule; + + goto out_spec; + +err: + macsec_fs_tx_del_rule(macsec_fs, tx_rule); + macsec_rule = NULL; +out_spec: + kvfree(spec); + + return macsec_rule; +} + +static void macsec_fs_tx_cleanup(struct mlx5e_macsec_fs *macsec_fs) +{ + struct mlx5e_macsec_tx *tx_fs = macsec_fs->tx_fs; + struct mlx5_core_dev *mdev = macsec_fs->mdev; + struct mlx5e_macsec_tables *tx_tables; + + if (!tx_fs) + return; + + tx_tables = &tx_fs->tables; + if (tx_tables->refcnt) { + netdev_err(macsec_fs->netdev, + "Can't destroy MACsec offload tx_fs, refcnt(%u) isn't 0\n", + tx_tables->refcnt); + return; + } + + ida_destroy(&tx_fs->tx_halloc); + + if (tx_tables->check_miss_rule_counter) { + mlx5_fc_destroy(mdev, tx_tables->check_miss_rule_counter); + tx_tables->check_miss_rule_counter = NULL; + } + + if (tx_tables->check_rule_counter) { + mlx5_fc_destroy(mdev, tx_tables->check_rule_counter); + tx_tables->check_rule_counter = NULL; + } + + kfree(tx_fs); + macsec_fs->tx_fs = NULL; +} + +static int macsec_fs_tx_init(struct mlx5e_macsec_fs *macsec_fs) +{ + struct net_device *netdev = macsec_fs->netdev; + struct mlx5_core_dev *mdev = macsec_fs->mdev; + struct mlx5e_macsec_tables *tx_tables; + struct mlx5e_macsec_tx *tx_fs; + struct mlx5_fc *flow_counter; + int err; + + tx_fs = kzalloc(sizeof(*tx_fs), GFP_KERNEL); + if (!tx_fs) + return -ENOMEM; + + tx_tables = &tx_fs->tables; + + flow_counter = mlx5_fc_create(mdev, false); + if (IS_ERR(flow_counter)) { + err = PTR_ERR(flow_counter); + netdev_err(netdev, + "Failed to create MACsec Tx encrypt flow counter, err(%d)\n", + err); + goto err_encrypt_counter; + } + tx_tables->check_rule_counter = flow_counter; + + flow_counter = mlx5_fc_create(mdev, false); + if (IS_ERR(flow_counter)) { + err = PTR_ERR(flow_counter); + netdev_err(netdev, + "Failed to create MACsec Tx drop flow counter, err(%d)\n", + err); + goto err_drop_counter; + } + tx_tables->check_miss_rule_counter = flow_counter; + + ida_init(&tx_fs->tx_halloc); + + macsec_fs->tx_fs = tx_fs; + + return 0; + +err_drop_counter: + mlx5_fc_destroy(mdev, tx_tables->check_rule_counter); + tx_tables->check_rule_counter = NULL; + +err_encrypt_counter: + kfree(tx_fs); + macsec_fs->tx_fs = NULL; + + return err; +} + +static void macsec_fs_rx_destroy(struct mlx5e_macsec_fs *macsec_fs) +{ + struct mlx5e_macsec_rx *rx_fs = macsec_fs->rx_fs; + struct mlx5e_macsec_tables *rx_tables; + int i; + + /* Rx check table */ + for (i = 1; i >= 0; --i) { + if (rx_fs->check_rule[i]) { + mlx5_del_flow_rules(rx_fs->check_rule[i]); + rx_fs->check_rule[i] = NULL; + } + + if (rx_fs->check_rule_pkt_reformat[i]) { + mlx5_packet_reformat_dealloc(macsec_fs->mdev, + rx_fs->check_rule_pkt_reformat[i]); + rx_fs->check_rule_pkt_reformat[i] = NULL; + } + } + + rx_tables = &rx_fs->tables; + + if (rx_tables->check_miss_rule) { + mlx5_del_flow_rules(rx_tables->check_miss_rule); + rx_tables->check_miss_rule = NULL; + } + + if (rx_tables->ft_check_group) { + mlx5_destroy_flow_group(rx_tables->ft_check_group); + rx_tables->ft_check_group = NULL; + } + + if (rx_tables->ft_check) { + mlx5_destroy_flow_table(rx_tables->ft_check); + rx_tables->ft_check = NULL; + } + + /* Rx crypto table */ + if (rx_tables->crypto_miss_rule) { + mlx5_del_flow_rules(rx_tables->crypto_miss_rule); + rx_tables->crypto_miss_rule = NULL; + } + + mlx5e_destroy_flow_table(&rx_tables->ft_crypto); +} + +static int macsec_fs_rx_create_crypto_table_groups(struct mlx5e_flow_table *ft) +{ + int inlen = MLX5_ST_SZ_BYTES(create_flow_group_in); + int mclen = MLX5_ST_SZ_BYTES(fte_match_param); + int ix = 0; + u32 *in; + int err; + u8 *mc; + + ft->g = kcalloc(RX_CRYPTO_TABLE_NUM_GROUPS, sizeof(*ft->g), GFP_KERNEL); + if (!ft->g) + return -ENOMEM; + + in = kvzalloc(inlen, GFP_KERNEL); + if (!in) { + kfree(ft->g); + return -ENOMEM; + } + + mc = MLX5_ADDR_OF(create_flow_group_in, in, match_criteria); + + /* Flow group for SA rule with SCI */ + MLX5_SET_CFG(in, match_criteria_enable, MLX5_MATCH_OUTER_HEADERS | + MLX5_MATCH_MISC_PARAMETERS_5); + MLX5_SET_TO_ONES(fte_match_param, mc, outer_headers.ethertype); + + MLX5_SET(fte_match_param, mc, misc_parameters_5.macsec_tag_0, + MLX5_MACSEC_SECTAG_TCI_AN_FIELD_BITMASK << + MLX5_MACSEC_SECTAG_TCI_AN_FIELD_OFFSET); + MLX5_SET_TO_ONES(fte_match_param, mc, misc_parameters_5.macsec_tag_2); + MLX5_SET_TO_ONES(fte_match_param, mc, misc_parameters_5.macsec_tag_3); + + MLX5_SET_CFG(in, start_flow_index, ix); + ix += RX_CRYPTO_TABLE_SA_RULE_WITH_SCI_GROUP_SIZE; + MLX5_SET_CFG(in, end_flow_index, ix - 1); + ft->g[ft->num_groups] = mlx5_create_flow_group(ft->t, in); + if (IS_ERR(ft->g[ft->num_groups])) + goto err; + ft->num_groups++; + + /* Flow group for SA rule without SCI */ + memset(in, 0, inlen); + memset(mc, 0, mclen); + MLX5_SET_CFG(in, match_criteria_enable, MLX5_MATCH_OUTER_HEADERS | + MLX5_MATCH_MISC_PARAMETERS_5); + MLX5_SET_TO_ONES(fte_match_param, mc, outer_headers.smac_47_16); + MLX5_SET_TO_ONES(fte_match_param, mc, outer_headers.smac_15_0); + MLX5_SET_TO_ONES(fte_match_param, mc, outer_headers.ethertype); + + MLX5_SET(fte_match_param, mc, misc_parameters_5.macsec_tag_0, + MLX5_MACSEC_SECTAG_TCI_AN_FIELD_BITMASK << MLX5_MACSEC_SECTAG_TCI_AN_FIELD_OFFSET); + + MLX5_SET_CFG(in, start_flow_index, ix); + ix += RX_CRYPTO_TABLE_SA_RULE_WITHOUT_SCI_GROUP_SIZE; + MLX5_SET_CFG(in, end_flow_index, ix - 1); + ft->g[ft->num_groups] = mlx5_create_flow_group(ft->t, in); + if (IS_ERR(ft->g[ft->num_groups])) + goto err; + ft->num_groups++; + + /* Flow Group for l2 traps */ + memset(in, 0, inlen); + memset(mc, 0, mclen); + MLX5_SET_CFG(in, start_flow_index, ix); + ix += CRYPTO_TABLE_DEFAULT_RULE_GROUP_SIZE; + MLX5_SET_CFG(in, end_flow_index, ix - 1); + ft->g[ft->num_groups] = mlx5_create_flow_group(ft->t, in); + if (IS_ERR(ft->g[ft->num_groups])) + goto err; + ft->num_groups++; + + kvfree(in); + return 0; + +err: + err = PTR_ERR(ft->g[ft->num_groups]); + ft->g[ft->num_groups] = NULL; + kvfree(in); + + return err; +} + +static int macsec_fs_rx_create_check_decap_rule(struct mlx5e_macsec_fs *macsec_fs, + struct mlx5_flow_destination *dest, + struct mlx5_flow_act *flow_act, + struct mlx5_flow_spec *spec, + int reformat_param_size) +{ + int rule_index = (reformat_param_size == MLX5_SECTAG_HEADER_SIZE_WITH_SCI) ? 0 : 1; + u8 mlx5_reformat_buf[MLX5_SECTAG_HEADER_SIZE_WITH_SCI]; + struct mlx5_pkt_reformat_params reformat_params = {}; + struct mlx5e_macsec_rx *rx_fs = macsec_fs->rx_fs; + struct net_device *netdev = macsec_fs->netdev; + struct mlx5e_macsec_tables *rx_tables; + struct mlx5_flow_handle *rule; + int err = 0; + + rx_tables = &rx_fs->tables; + + /* Rx check table decap 16B rule */ + memset(dest, 0, sizeof(*dest)); + memset(flow_act, 0, sizeof(*flow_act)); + memset(spec, 0, sizeof(*spec)); + + reformat_params.type = MLX5_REFORMAT_TYPE_DEL_MACSEC; + reformat_params.size = reformat_param_size; + reformat_params.data = mlx5_reformat_buf; + flow_act->pkt_reformat = mlx5_packet_reformat_alloc(macsec_fs->mdev, + &reformat_params, + MLX5_FLOW_NAMESPACE_KERNEL_RX_MACSEC); + if (IS_ERR(flow_act->pkt_reformat)) { + err = PTR_ERR(flow_act->pkt_reformat); + netdev_err(netdev, "Failed to allocate MACsec Rx reformat context err=%d\n", err); + return err; + } + rx_fs->check_rule_pkt_reformat[rule_index] = flow_act->pkt_reformat; + + spec->match_criteria_enable = MLX5_MATCH_MISC_PARAMETERS_2; + /* MACsec syndrome match */ + MLX5_SET_TO_ONES(fte_match_param, spec->match_criteria, misc_parameters_2.macsec_syndrome); + MLX5_SET(fte_match_param, spec->match_value, misc_parameters_2.macsec_syndrome, 0); + /* ASO return reg syndrome match */ + MLX5_SET_TO_ONES(fte_match_param, spec->match_criteria, misc_parameters_2.metadata_reg_c_4); + MLX5_SET(fte_match_param, spec->match_value, misc_parameters_2.metadata_reg_c_4, 0); + + spec->match_criteria_enable |= MLX5_MATCH_MISC_PARAMETERS_5; + /* Sectag TCI SC present bit*/ + MLX5_SET(fte_match_param, spec->match_criteria, misc_parameters_5.macsec_tag_0, + MLX5_MACSEC_SECTAG_TCI_SC_FIELD_BIT << MLX5_MACSEC_SECTAG_TCI_AN_FIELD_OFFSET); + + if (reformat_param_size == MLX5_SECTAG_HEADER_SIZE_WITH_SCI) + MLX5_SET(fte_match_param, spec->match_value, misc_parameters_5.macsec_tag_0, + MLX5_MACSEC_SECTAG_TCI_SC_FIELD_BIT << + MLX5_MACSEC_SECTAG_TCI_AN_FIELD_OFFSET); + + flow_act->flags = FLOW_ACT_NO_APPEND; + flow_act->action = MLX5_FLOW_CONTEXT_ACTION_FWD_NEXT_PRIO | + MLX5_FLOW_CONTEXT_ACTION_PACKET_REFORMAT | + MLX5_FLOW_CONTEXT_ACTION_COUNT; + dest->type = MLX5_FLOW_DESTINATION_TYPE_COUNTER; + dest->counter_id = mlx5_fc_id(rx_tables->check_rule_counter); + rule = mlx5_add_flow_rules(rx_tables->ft_check, spec, flow_act, dest, 1); + if (IS_ERR(rule)) { + err = PTR_ERR(rule); + netdev_err(netdev, "Failed to add MACsec Rx check rule, err=%d\n", err); + return err; + } + + rx_fs->check_rule[rule_index] = rule; + + return 0; +} + +static int macsec_fs_rx_create(struct mlx5e_macsec_fs *macsec_fs) +{ + int inlen = MLX5_ST_SZ_BYTES(create_flow_group_in); + struct mlx5e_macsec_rx *rx_fs = macsec_fs->rx_fs; + struct net_device *netdev = macsec_fs->netdev; + struct mlx5_flow_table_attr ft_attr = {}; + struct mlx5_flow_destination dest = {}; + struct mlx5e_macsec_tables *rx_tables; + struct mlx5e_flow_table *ft_crypto; + struct mlx5_flow_table *flow_table; + struct mlx5_flow_group *flow_group; + struct mlx5_flow_act flow_act = {}; + struct mlx5_flow_namespace *ns; + struct mlx5_flow_handle *rule; + struct mlx5_flow_spec *spec; + u32 *flow_group_in; + int err = 0; + + ns = mlx5_get_flow_namespace(macsec_fs->mdev, MLX5_FLOW_NAMESPACE_KERNEL_RX_MACSEC); + if (!ns) + return -ENOMEM; + + spec = kvzalloc(sizeof(*spec), GFP_KERNEL); + if (!spec) + return -ENOMEM; + + flow_group_in = kvzalloc(inlen, GFP_KERNEL); + if (!flow_group_in) + goto free_spec; + + rx_tables = &rx_fs->tables; + ft_crypto = &rx_tables->ft_crypto; + + /* Rx crypto table */ + ft_attr.level = RX_CRYPTO_TABLE_LEVEL; + ft_attr.max_fte = CRYPTO_NUM_MAXSEC_FTE; + + flow_table = mlx5_create_flow_table(ns, &ft_attr); + if (IS_ERR(flow_table)) { + err = PTR_ERR(flow_table); + netdev_err(netdev, "Failed to create MACsec Rx crypto table err(%d)\n", err); + goto out_flow_group; + } + ft_crypto->t = flow_table; + + /* Rx crypto table groups */ + err = macsec_fs_rx_create_crypto_table_groups(ft_crypto); + if (err) { + netdev_err(netdev, + "Failed to create default flow group for MACsec Tx crypto table err(%d)\n", + err); + goto err; + } + + flow_act.action = MLX5_FLOW_CONTEXT_ACTION_FWD_NEXT_PRIO; + rule = mlx5_add_flow_rules(ft_crypto->t, NULL, &flow_act, NULL, 0); + if (IS_ERR(rule)) { + err = PTR_ERR(rule); + netdev_err(netdev, + "Failed to add MACsec Rx crypto table default miss rule %d\n", + err); + goto err; + } + rx_tables->crypto_miss_rule = rule; + + /* Rx check table */ + flow_table = macsec_fs_auto_group_table_create(ns, + MLX5_FLOW_TABLE_TUNNEL_EN_REFORMAT, + RX_CHECK_TABLE_LEVEL, + RX_CHECK_TABLE_NUM_FTE); + if (IS_ERR(flow_table)) { + err = PTR_ERR(flow_table); + netdev_err(netdev, "fail to create MACsec RX check table, err(%d)\n", err); + goto err; + } + rx_tables->ft_check = flow_table; + + /* Rx check table Default miss group/rule */ + MLX5_SET(create_flow_group_in, flow_group_in, start_flow_index, flow_table->max_fte - 1); + MLX5_SET(create_flow_group_in, flow_group_in, end_flow_index, flow_table->max_fte - 1); + flow_group = mlx5_create_flow_group(rx_tables->ft_check, flow_group_in); + if (IS_ERR(flow_group)) { + err = PTR_ERR(flow_group); + netdev_err(netdev, + "Failed to create default flow group for MACsec Rx check table err(%d)\n", + err); + goto err; + } + rx_tables->ft_check_group = flow_group; + + /* Rx check table default drop rule */ + memset(&flow_act, 0, sizeof(flow_act)); + + dest.type = MLX5_FLOW_DESTINATION_TYPE_COUNTER; + dest.counter_id = mlx5_fc_id(rx_tables->check_miss_rule_counter); + flow_act.action = MLX5_FLOW_CONTEXT_ACTION_DROP | MLX5_FLOW_CONTEXT_ACTION_COUNT; + rule = mlx5_add_flow_rules(rx_tables->ft_check, NULL, &flow_act, &dest, 1); + if (IS_ERR(rule)) { + err = PTR_ERR(rule); + netdev_err(netdev, "Failed to added MACsec Rx check drop rule, err(%d)\n", err); + goto err; + } + rx_tables->check_miss_rule = rule; + + /* Rx check table decap rules */ + err = macsec_fs_rx_create_check_decap_rule(macsec_fs, &dest, &flow_act, spec, + MLX5_SECTAG_HEADER_SIZE_WITH_SCI); + if (err) + goto err; + + err = macsec_fs_rx_create_check_decap_rule(macsec_fs, &dest, &flow_act, spec, + MLX5_SECTAG_HEADER_SIZE_WITHOUT_SCI); + if (err) + goto err; + + goto out_flow_group; + +err: + macsec_fs_rx_destroy(macsec_fs); +out_flow_group: + kvfree(flow_group_in); +free_spec: + kvfree(spec); + return err; +} + +static int macsec_fs_rx_ft_get(struct mlx5e_macsec_fs *macsec_fs) +{ + struct mlx5e_macsec_tables *rx_tables = &macsec_fs->rx_fs->tables; + int err = 0; + + if (rx_tables->refcnt) + goto out; + + err = macsec_fs_rx_create(macsec_fs); + if (err) + return err; + +out: + rx_tables->refcnt++; + return err; +} + +static void macsec_fs_rx_ft_put(struct mlx5e_macsec_fs *macsec_fs) +{ + struct mlx5e_macsec_tables *rx_tables = &macsec_fs->rx_fs->tables; + + if (--rx_tables->refcnt) + return; + + macsec_fs_rx_destroy(macsec_fs); +} + +static void macsec_fs_rx_del_rule(struct mlx5e_macsec_fs *macsec_fs, + struct mlx5e_macsec_rx_rule *rx_rule) +{ + int i; + + for (i = 0; i < RX_NUM_OF_RULES_PER_SA; ++i) { + if (rx_rule->rule[i]) { + mlx5_del_flow_rules(rx_rule->rule[i]); + rx_rule->rule[i] = NULL; + } + } + + if (rx_rule->meta_modhdr) { + mlx5_modify_header_dealloc(macsec_fs->mdev, rx_rule->meta_modhdr); + rx_rule->meta_modhdr = NULL; + } + + kfree(rx_rule); + + macsec_fs_rx_ft_put(macsec_fs); +} + +static void macsec_fs_rx_setup_fte(struct mlx5_flow_spec *spec, + struct mlx5_flow_act *flow_act, + struct mlx5_macsec_rule_attrs *attrs, + bool sci_present) +{ + u8 tci_an = (sci_present << MLX5_MACSEC_SECTAG_TCI_SC_FIELD_OFFSET) | attrs->assoc_num; + struct mlx5_flow_act_crypto_params *crypto_params = &flow_act->crypto; + __be32 *sci_p = (__be32 *)(&attrs->sci); + + spec->match_criteria_enable = MLX5_MATCH_OUTER_HEADERS; + + /* MACsec ethertype */ + MLX5_SET_TO_ONES(fte_match_param, spec->match_criteria, outer_headers.ethertype); + MLX5_SET(fte_match_param, spec->match_value, outer_headers.ethertype, ETH_P_MACSEC); + + spec->match_criteria_enable |= MLX5_MATCH_MISC_PARAMETERS_5; + + /* Sectag AN + TCI SC present bit*/ + MLX5_SET(fte_match_param, spec->match_criteria, misc_parameters_5.macsec_tag_0, + MLX5_MACSEC_SECTAG_TCI_AN_FIELD_BITMASK << MLX5_MACSEC_SECTAG_TCI_AN_FIELD_OFFSET); + MLX5_SET(fte_match_param, spec->match_value, misc_parameters_5.macsec_tag_0, + tci_an << MLX5_MACSEC_SECTAG_TCI_AN_FIELD_OFFSET); + + if (sci_present) { + MLX5_SET_TO_ONES(fte_match_param, spec->match_criteria, + misc_parameters_5.macsec_tag_2); + MLX5_SET(fte_match_param, spec->match_value, misc_parameters_5.macsec_tag_2, + be32_to_cpu(sci_p[0])); + + MLX5_SET_TO_ONES(fte_match_param, spec->match_criteria, + misc_parameters_5.macsec_tag_3); + MLX5_SET(fte_match_param, spec->match_value, misc_parameters_5.macsec_tag_3, + be32_to_cpu(sci_p[1])); + } else { + /* When SCI isn't present in the Sectag, need to match the source */ + /* MAC address only if the SCI contains the default MACsec PORT */ + MLX5_SET_TO_ONES(fte_match_param, spec->match_criteria, outer_headers.smac_47_16); + MLX5_SET_TO_ONES(fte_match_param, spec->match_criteria, outer_headers.smac_15_0); + memcpy(MLX5_ADDR_OF(fte_match_param, spec->match_value, outer_headers.smac_47_16), + sci_p, ETH_ALEN); + } + + crypto_params->type = MLX5_FLOW_CONTEXT_ENCRYPT_DECRYPT_TYPE_MACSEC; + crypto_params->obj_id = attrs->macsec_obj_id; +} + +static union mlx5e_macsec_rule * +macsec_fs_rx_add_rule(struct mlx5e_macsec_fs *macsec_fs, + const struct macsec_context *macsec_ctx, + struct mlx5_macsec_rule_attrs *attrs, + u32 fs_id) +{ + u8 action[MLX5_UN_SZ_BYTES(set_add_copy_action_in_auto)] = {}; + struct mlx5e_macsec_rx *rx_fs = macsec_fs->rx_fs; + struct net_device *netdev = macsec_fs->netdev; + union mlx5e_macsec_rule *macsec_rule = NULL; + struct mlx5_modify_hdr *modify_hdr = NULL; + struct mlx5_flow_destination dest = {}; + struct mlx5e_macsec_tables *rx_tables; + struct mlx5e_macsec_rx_rule *rx_rule; + struct mlx5_flow_act flow_act = {}; + struct mlx5e_flow_table *ft_crypto; + struct mlx5_flow_handle *rule; + struct mlx5_flow_spec *spec; + int err = 0; + + spec = kvzalloc(sizeof(*spec), GFP_KERNEL); + if (!spec) + return NULL; + + err = macsec_fs_rx_ft_get(macsec_fs); + if (err) + goto out_spec; + + macsec_rule = kzalloc(sizeof(*macsec_rule), GFP_KERNEL); + if (!macsec_rule) { + macsec_fs_rx_ft_put(macsec_fs); + goto out_spec; + } + + rx_rule = &macsec_rule->rx_rule; + rx_tables = &rx_fs->tables; + ft_crypto = &rx_tables->ft_crypto; + + /* Set bit[31 - 30] macsec marker - 0x01 */ + /* Set bit[3-0] fs id */ + MLX5_SET(set_action_in, action, action_type, MLX5_ACTION_TYPE_SET); + MLX5_SET(set_action_in, action, field, MLX5_ACTION_IN_FIELD_METADATA_REG_B); + MLX5_SET(set_action_in, action, data, fs_id | BIT(30)); + MLX5_SET(set_action_in, action, offset, 0); + MLX5_SET(set_action_in, action, length, 32); + + modify_hdr = mlx5_modify_header_alloc(macsec_fs->mdev, MLX5_FLOW_NAMESPACE_KERNEL_RX_MACSEC, + 1, action); + if (IS_ERR(modify_hdr)) { + err = PTR_ERR(modify_hdr); + netdev_err(netdev, "fail to alloc MACsec set modify_header_id err=%d\n", err); + modify_hdr = NULL; + goto err; + } + rx_rule->meta_modhdr = modify_hdr; + + /* Rx crypto table with SCI rule */ + macsec_fs_rx_setup_fte(spec, &flow_act, attrs, true); + + flow_act.modify_hdr = modify_hdr; + flow_act.action = MLX5_FLOW_CONTEXT_ACTION_FWD_DEST | + MLX5_FLOW_CONTEXT_ACTION_CRYPTO_DECRYPT | + MLX5_FLOW_CONTEXT_ACTION_MOD_HDR; + + dest.type = MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE; + dest.ft = rx_tables->ft_check; + rule = mlx5_add_flow_rules(ft_crypto->t, spec, &flow_act, &dest, 1); + if (IS_ERR(rule)) { + err = PTR_ERR(rule); + netdev_err(netdev, + "Failed to add SA with SCI rule to Rx crypto rule, err=%d\n", + err); + goto err; + } + rx_rule->rule[0] = rule; + + /* Rx crypto table without SCI rule */ + if (cpu_to_be64((__force u64)attrs->sci) & ntohs(MACSEC_PORT_ES)) { + memset(spec, 0, sizeof(struct mlx5_flow_spec)); + memset(&dest, 0, sizeof(struct mlx5_flow_destination)); + memset(&flow_act, 0, sizeof(flow_act)); + + macsec_fs_rx_setup_fte(spec, &flow_act, attrs, false); + + flow_act.modify_hdr = modify_hdr; + flow_act.action = MLX5_FLOW_CONTEXT_ACTION_FWD_DEST | + MLX5_FLOW_CONTEXT_ACTION_CRYPTO_DECRYPT | + MLX5_FLOW_CONTEXT_ACTION_MOD_HDR; + + dest.type = MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE; + dest.ft = rx_tables->ft_check; + rule = mlx5_add_flow_rules(ft_crypto->t, spec, &flow_act, &dest, 1); + if (IS_ERR(rule)) { + err = PTR_ERR(rule); + netdev_err(netdev, + "Failed to add SA without SCI rule to Rx crypto rule, err=%d\n", + err); + goto err; + } + rx_rule->rule[1] = rule; + } + + return macsec_rule; + +err: + macsec_fs_rx_del_rule(macsec_fs, rx_rule); + macsec_rule = NULL; +out_spec: + kvfree(spec); + return macsec_rule; +} + +static int macsec_fs_rx_init(struct mlx5e_macsec_fs *macsec_fs) +{ + struct net_device *netdev = macsec_fs->netdev; + struct mlx5_core_dev *mdev = macsec_fs->mdev; + struct mlx5e_macsec_tables *rx_tables; + struct mlx5e_macsec_rx *rx_fs; + struct mlx5_fc *flow_counter; + int err; + + rx_fs = kzalloc(sizeof(*rx_fs), GFP_KERNEL); + if (!rx_fs) + return -ENOMEM; + + flow_counter = mlx5_fc_create(mdev, false); + if (IS_ERR(flow_counter)) { + err = PTR_ERR(flow_counter); + netdev_err(netdev, + "Failed to create MACsec Rx encrypt flow counter, err(%d)\n", + err); + goto err_encrypt_counter; + } + + rx_tables = &rx_fs->tables; + rx_tables->check_rule_counter = flow_counter; + + flow_counter = mlx5_fc_create(mdev, false); + if (IS_ERR(flow_counter)) { + err = PTR_ERR(flow_counter); + netdev_err(netdev, + "Failed to create MACsec Rx drop flow counter, err(%d)\n", + err); + goto err_drop_counter; + } + rx_tables->check_miss_rule_counter = flow_counter; + + macsec_fs->rx_fs = rx_fs; + + return 0; + +err_drop_counter: + mlx5_fc_destroy(mdev, rx_tables->check_rule_counter); + rx_tables->check_rule_counter = NULL; + +err_encrypt_counter: + kfree(rx_fs); + macsec_fs->rx_fs = NULL; + + return err; +} + +static void macsec_fs_rx_cleanup(struct mlx5e_macsec_fs *macsec_fs) +{ + struct mlx5e_macsec_rx *rx_fs = macsec_fs->rx_fs; + struct mlx5_core_dev *mdev = macsec_fs->mdev; + struct mlx5e_macsec_tables *rx_tables; + + if (!rx_fs) + return; + + rx_tables = &rx_fs->tables; + + if (rx_tables->refcnt) { + netdev_err(macsec_fs->netdev, + "Can't destroy MACsec offload rx_fs, refcnt(%u) isn't 0\n", + rx_tables->refcnt); + return; + } + + if (rx_tables->check_miss_rule_counter) { + mlx5_fc_destroy(mdev, rx_tables->check_miss_rule_counter); + rx_tables->check_miss_rule_counter = NULL; + } + + if (rx_tables->check_rule_counter) { + mlx5_fc_destroy(mdev, rx_tables->check_rule_counter); + rx_tables->check_rule_counter = NULL; + } + + kfree(rx_fs); + macsec_fs->rx_fs = NULL; +} + +void mlx5e_macsec_fs_get_stats_fill(struct mlx5e_macsec_fs *macsec_fs, void *macsec_stats) +{ + struct mlx5e_macsec_stats *stats = (struct mlx5e_macsec_stats *)macsec_stats; + struct mlx5e_macsec_tables *tx_tables = &macsec_fs->tx_fs->tables; + struct mlx5e_macsec_tables *rx_tables = &macsec_fs->rx_fs->tables; + struct mlx5_core_dev *mdev = macsec_fs->mdev; + + if (tx_tables->check_rule_counter) + mlx5_fc_query(mdev, tx_tables->check_rule_counter, + &stats->macsec_tx_pkts, &stats->macsec_tx_bytes); + + if (tx_tables->check_miss_rule_counter) + mlx5_fc_query(mdev, tx_tables->check_miss_rule_counter, + &stats->macsec_tx_pkts_drop, &stats->macsec_tx_bytes_drop); + + if (rx_tables->check_rule_counter) + mlx5_fc_query(mdev, rx_tables->check_rule_counter, + &stats->macsec_rx_pkts, &stats->macsec_rx_bytes); + + if (rx_tables->check_miss_rule_counter) + mlx5_fc_query(mdev, rx_tables->check_miss_rule_counter, + &stats->macsec_rx_pkts_drop, &stats->macsec_rx_bytes_drop); +} + +union mlx5e_macsec_rule * +mlx5e_macsec_fs_add_rule(struct mlx5e_macsec_fs *macsec_fs, + const struct macsec_context *macsec_ctx, + struct mlx5_macsec_rule_attrs *attrs, + u32 *sa_fs_id) +{ + return (attrs->action == MLX5_ACCEL_MACSEC_ACTION_ENCRYPT) ? + macsec_fs_tx_add_rule(macsec_fs, macsec_ctx, attrs, sa_fs_id) : + macsec_fs_rx_add_rule(macsec_fs, macsec_ctx, attrs, *sa_fs_id); +} + +void mlx5e_macsec_fs_del_rule(struct mlx5e_macsec_fs *macsec_fs, + union mlx5e_macsec_rule *macsec_rule, + int action) +{ + (action == MLX5_ACCEL_MACSEC_ACTION_ENCRYPT) ? + macsec_fs_tx_del_rule(macsec_fs, &macsec_rule->tx_rule) : + macsec_fs_rx_del_rule(macsec_fs, &macsec_rule->rx_rule); +} + +void mlx5e_macsec_fs_cleanup(struct mlx5e_macsec_fs *macsec_fs) +{ + macsec_fs_rx_cleanup(macsec_fs); + macsec_fs_tx_cleanup(macsec_fs); + kfree(macsec_fs); +} + +struct mlx5e_macsec_fs * +mlx5e_macsec_fs_init(struct mlx5_core_dev *mdev, + struct net_device *netdev) +{ + struct mlx5e_macsec_fs *macsec_fs; + int err; + + macsec_fs = kzalloc(sizeof(*macsec_fs), GFP_KERNEL); + if (!macsec_fs) + return NULL; + + macsec_fs->mdev = mdev; + macsec_fs->netdev = netdev; + + err = macsec_fs_tx_init(macsec_fs); + if (err) { + netdev_err(netdev, "MACsec offload: Failed to init tx_fs, err=%d\n", err); + goto err; + } + + err = macsec_fs_rx_init(macsec_fs); + if (err) { + netdev_err(netdev, "MACsec offload: Failed to init tx_fs, err=%d\n", err); + goto tx_cleanup; + } + + return macsec_fs; + +tx_cleanup: + macsec_fs_tx_cleanup(macsec_fs); +err: + kfree(macsec_fs); + return NULL; +} diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/macsec_fs.h b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/macsec_fs.h new file mode 100644 index 000000000000..b429648d4ee7 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/macsec_fs.h @@ -0,0 +1,47 @@ +/* SPDX-License-Identifier: GPL-2.0 OR Linux-OpenIB */ +/* Copyright (c) 2022, NVIDIA CORPORATION & AFFILIATES. All rights reserved. */ + +#ifndef __MLX5_MACSEC_STEERING_H__ +#define __MLX5_MACSEC_STEERING_H__ + +#ifdef CONFIG_MLX5_EN_MACSEC + +#include "en_accel/macsec.h" + +#define MLX5_MACSEC_NUM_OF_SUPPORTED_INTERFACES 16 + +struct mlx5e_macsec_fs; +union mlx5e_macsec_rule; + +struct mlx5_macsec_rule_attrs { + sci_t sci; + u32 macsec_obj_id; + u8 assoc_num; + int action; +}; + +enum mlx5_macsec_action { + MLX5_ACCEL_MACSEC_ACTION_ENCRYPT, + MLX5_ACCEL_MACSEC_ACTION_DECRYPT, +}; + +void mlx5e_macsec_fs_cleanup(struct mlx5e_macsec_fs *macsec_fs); + +struct mlx5e_macsec_fs * +mlx5e_macsec_fs_init(struct mlx5_core_dev *mdev, struct net_device *netdev); + +union mlx5e_macsec_rule * +mlx5e_macsec_fs_add_rule(struct mlx5e_macsec_fs *macsec_fs, + const struct macsec_context *ctx, + struct mlx5_macsec_rule_attrs *attrs, + u32 *sa_fs_id); + +void mlx5e_macsec_fs_del_rule(struct mlx5e_macsec_fs *macsec_fs, + union mlx5e_macsec_rule *macsec_rule, + int action); + +void mlx5e_macsec_fs_get_stats_fill(struct mlx5e_macsec_fs *macsec_fs, void *macsec_stats); + +#endif + +#endif /* __MLX5_MACSEC_STEERING_H__ */ diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/macsec_stats.c b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/macsec_stats.c new file mode 100644 index 000000000000..e50a2e3f3d18 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/macsec_stats.c @@ -0,0 +1,72 @@ +// SPDX-License-Identifier: GPL-2.0 OR Linux-OpenIB +/* Copyright (c) 2022, NVIDIA CORPORATION & AFFILIATES. All rights reserved. */ + +#include <linux/ethtool.h> +#include <net/sock.h> + +#include "en.h" +#include "en_accel/macsec.h" + +static const struct counter_desc mlx5e_macsec_hw_stats_desc[] = { + { MLX5E_DECLARE_STAT(struct mlx5e_macsec_stats, macsec_rx_pkts) }, + { MLX5E_DECLARE_STAT(struct mlx5e_macsec_stats, macsec_rx_bytes) }, + { MLX5E_DECLARE_STAT(struct mlx5e_macsec_stats, macsec_rx_pkts_drop) }, + { MLX5E_DECLARE_STAT(struct mlx5e_macsec_stats, macsec_rx_bytes_drop) }, + { MLX5E_DECLARE_STAT(struct mlx5e_macsec_stats, macsec_tx_pkts) }, + { MLX5E_DECLARE_STAT(struct mlx5e_macsec_stats, macsec_tx_bytes) }, + { MLX5E_DECLARE_STAT(struct mlx5e_macsec_stats, macsec_tx_pkts_drop) }, + { MLX5E_DECLARE_STAT(struct mlx5e_macsec_stats, macsec_tx_bytes_drop) }, +}; + +#define NUM_MACSEC_HW_COUNTERS ARRAY_SIZE(mlx5e_macsec_hw_stats_desc) + +static MLX5E_DECLARE_STATS_GRP_OP_NUM_STATS(macsec_hw) +{ + if (!priv->macsec) + return 0; + + if (mlx5e_is_macsec_device(priv->mdev)) + return NUM_MACSEC_HW_COUNTERS; + + return 0; +} + +static MLX5E_DECLARE_STATS_GRP_OP_UPDATE_STATS(macsec_hw) {} + +static MLX5E_DECLARE_STATS_GRP_OP_FILL_STRS(macsec_hw) +{ + unsigned int i; + + if (!priv->macsec) + return idx; + + if (!mlx5e_is_macsec_device(priv->mdev)) + return idx; + + for (i = 0; i < NUM_MACSEC_HW_COUNTERS; i++) + strcpy(data + (idx++) * ETH_GSTRING_LEN, + mlx5e_macsec_hw_stats_desc[i].format); + + return idx; +} + +static MLX5E_DECLARE_STATS_GRP_OP_FILL_STATS(macsec_hw) +{ + int i; + + if (!priv->macsec) + return idx; + + if (!mlx5e_is_macsec_device(priv->mdev)) + return idx; + + mlx5e_macsec_get_stats_fill(priv->macsec, mlx5e_macsec_get_stats(priv->macsec)); + for (i = 0; i < NUM_MACSEC_HW_COUNTERS; i++) + data[idx++] = MLX5E_READ_CTR64_CPU(mlx5e_macsec_get_stats(priv->macsec), + mlx5e_macsec_hw_stats_desc, + i); + + return idx; +} + +MLX5E_DEFINE_STATS_GRP(macsec_hw, 0); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_arfs.c b/drivers/net/ethernet/mellanox/mlx5/core/en_arfs.c index cd7f245dcf14..0ae1865086ff 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_arfs.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_arfs.c @@ -114,47 +114,49 @@ static enum mlx5_traffic_types arfs_get_tt(enum arfs_type type) } } -static int arfs_disable(struct mlx5e_priv *priv) +static int arfs_disable(struct mlx5e_flow_steering *fs) { + struct mlx5_ttc_table *ttc = mlx5e_fs_get_ttc(fs, false); int err, i; for (i = 0; i < ARFS_NUM_TYPES; i++) { /* Modify ttc rules destination back to their default */ - err = mlx5_ttc_fwd_default_dest(priv->fs->ttc, arfs_get_tt(i)); + err = mlx5_ttc_fwd_default_dest(ttc, arfs_get_tt(i)); if (err) { - netdev_err(priv->netdev, - "%s: modify ttc[%d] default destination failed, err(%d)\n", - __func__, arfs_get_tt(i), err); + fs_err(fs, + "%s: modify ttc[%d] default destination failed, err(%d)\n", + __func__, arfs_get_tt(i), err); return err; } } return 0; } -static void arfs_del_rules(struct mlx5e_priv *priv); +static void arfs_del_rules(struct mlx5e_flow_steering *fs); -int mlx5e_arfs_disable(struct mlx5e_priv *priv) +int mlx5e_arfs_disable(struct mlx5e_flow_steering *fs) { - arfs_del_rules(priv); + arfs_del_rules(fs); - return arfs_disable(priv); + return arfs_disable(fs); } -int mlx5e_arfs_enable(struct mlx5e_priv *priv) +int mlx5e_arfs_enable(struct mlx5e_flow_steering *fs) { + struct mlx5_ttc_table *ttc = mlx5e_fs_get_ttc(fs, false); + struct mlx5e_arfs_tables *arfs = mlx5e_fs_get_arfs(fs); struct mlx5_flow_destination dest = {}; int err, i; dest.type = MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE; for (i = 0; i < ARFS_NUM_TYPES; i++) { - dest.ft = priv->fs->arfs->arfs_tables[i].ft.t; + dest.ft = arfs->arfs_tables[i].ft.t; /* Modify ttc rules destination to point on the aRFS FTs */ - err = mlx5_ttc_fwd_dest(priv->fs->ttc, arfs_get_tt(i), &dest); + err = mlx5_ttc_fwd_dest(ttc, arfs_get_tt(i), &dest); if (err) { - netdev_err(priv->netdev, - "%s: modify ttc[%d] dest to arfs, failed err(%d)\n", - __func__, arfs_get_tt(i), err); - arfs_disable(priv); + fs_err(fs, "%s: modify ttc[%d] dest to arfs, failed err(%d)\n", + __func__, arfs_get_tt(i), err); + arfs_disable(fs); return err; } } @@ -167,31 +169,37 @@ static void arfs_destroy_table(struct arfs_table *arfs_t) mlx5e_destroy_flow_table(&arfs_t->ft); } -static void _mlx5e_cleanup_tables(struct mlx5e_priv *priv) +static void _mlx5e_cleanup_tables(struct mlx5e_flow_steering *fs) { + struct mlx5e_arfs_tables *arfs = mlx5e_fs_get_arfs(fs); int i; - arfs_del_rules(priv); - destroy_workqueue(priv->fs->arfs->wq); + arfs_del_rules(fs); + destroy_workqueue(arfs->wq); for (i = 0; i < ARFS_NUM_TYPES; i++) { - if (!IS_ERR_OR_NULL(priv->fs->arfs->arfs_tables[i].ft.t)) - arfs_destroy_table(&priv->fs->arfs->arfs_tables[i]); + if (!IS_ERR_OR_NULL(arfs->arfs_tables[i].ft.t)) + arfs_destroy_table(&arfs->arfs_tables[i]); } } -void mlx5e_arfs_destroy_tables(struct mlx5e_priv *priv) +void mlx5e_arfs_destroy_tables(struct mlx5e_flow_steering *fs, bool ntuple) { - if (!(priv->netdev->hw_features & NETIF_F_NTUPLE)) + struct mlx5e_arfs_tables *arfs = mlx5e_fs_get_arfs(fs); + + if (!ntuple) return; - _mlx5e_cleanup_tables(priv); - kvfree(priv->fs->arfs); + _mlx5e_cleanup_tables(fs); + mlx5e_fs_set_arfs(fs, NULL); + kvfree(arfs); } -static int arfs_add_default_rule(struct mlx5e_priv *priv, +static int arfs_add_default_rule(struct mlx5e_flow_steering *fs, + struct mlx5e_rx_res *rx_res, enum arfs_type type) { - struct arfs_table *arfs_t = &priv->fs->arfs->arfs_tables[type]; + struct mlx5e_arfs_tables *arfs = mlx5e_fs_get_arfs(fs); + struct arfs_table *arfs_t = &arfs->arfs_tables[type]; struct mlx5_flow_destination dest = {}; MLX5_DECLARE_FLOW_ACT(flow_act); enum mlx5_traffic_types tt; @@ -200,23 +208,21 @@ static int arfs_add_default_rule(struct mlx5e_priv *priv, dest.type = MLX5_FLOW_DESTINATION_TYPE_TIR; tt = arfs_get_tt(type); if (tt == -EINVAL) { - netdev_err(priv->netdev, "%s: bad arfs_type: %d\n", - __func__, type); + fs_err(fs, "%s: bad arfs_type: %d\n", __func__, type); return -EINVAL; } /* FIXME: Must use mlx5_ttc_get_default_dest(), * but can't since TTC default is not setup yet ! */ - dest.tir_num = mlx5e_rx_res_get_tirn_rss(priv->rx_res, tt); + dest.tir_num = mlx5e_rx_res_get_tirn_rss(rx_res, tt); arfs_t->default_rule = mlx5_add_flow_rules(arfs_t->ft.t, NULL, &flow_act, &dest, 1); if (IS_ERR(arfs_t->default_rule)) { err = PTR_ERR(arfs_t->default_rule); arfs_t->default_rule = NULL; - netdev_err(priv->netdev, "%s: add rule failed, arfs type=%d\n", - __func__, type); + fs_err(fs, "%s: add rule failed, arfs type=%d\n", __func__, type); } return err; @@ -318,10 +324,12 @@ out: return err; } -static int arfs_create_table(struct mlx5e_priv *priv, +static int arfs_create_table(struct mlx5e_flow_steering *fs, + struct mlx5e_rx_res *rx_res, enum arfs_type type) { - struct mlx5e_arfs_tables *arfs = priv->fs->arfs; + struct mlx5_flow_namespace *ns = mlx5e_fs_get_ns(fs, false); + struct mlx5e_arfs_tables *arfs = mlx5e_fs_get_arfs(fs); struct mlx5e_flow_table *ft = &arfs->arfs_tables[type].ft; struct mlx5_flow_table_attr ft_attr = {}; int err; @@ -332,7 +340,7 @@ static int arfs_create_table(struct mlx5e_priv *priv, ft_attr.level = MLX5E_ARFS_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(ns, &ft_attr); if (IS_ERR(ft->t)) { err = PTR_ERR(ft->t); ft->t = NULL; @@ -343,7 +351,7 @@ static int arfs_create_table(struct mlx5e_priv *priv, if (err) goto err; - err = arfs_add_default_rule(priv, type); + err = arfs_add_default_rule(fs, rx_res, type); if (err) goto err; @@ -353,35 +361,40 @@ err: return err; } -int mlx5e_arfs_create_tables(struct mlx5e_priv *priv) +int mlx5e_arfs_create_tables(struct mlx5e_flow_steering *fs, + struct mlx5e_rx_res *rx_res, bool ntuple) { + struct mlx5e_arfs_tables *arfs; int err = -ENOMEM; int i; - if (!(priv->netdev->hw_features & NETIF_F_NTUPLE)) + if (!ntuple) return 0; - priv->fs->arfs = kvzalloc(sizeof(*priv->fs->arfs), GFP_KERNEL); - if (!priv->fs->arfs) + arfs = kvzalloc(sizeof(*arfs), GFP_KERNEL); + if (!arfs) return -ENOMEM; - spin_lock_init(&priv->fs->arfs->arfs_lock); - INIT_LIST_HEAD(&priv->fs->arfs->rules); - priv->fs->arfs->wq = create_singlethread_workqueue("mlx5e_arfs"); - if (!priv->fs->arfs->wq) + spin_lock_init(&arfs->arfs_lock); + INIT_LIST_HEAD(&arfs->rules); + arfs->wq = create_singlethread_workqueue("mlx5e_arfs"); + if (!arfs->wq) goto err; + mlx5e_fs_set_arfs(fs, arfs); + for (i = 0; i < ARFS_NUM_TYPES; i++) { - err = arfs_create_table(priv, i); + err = arfs_create_table(fs, rx_res, i); if (err) goto err_des; } return 0; err_des: - _mlx5e_cleanup_tables(priv); + _mlx5e_cleanup_tables(fs); err: - kvfree(priv->fs->arfs); + mlx5e_fs_set_arfs(fs, NULL); + kvfree(arfs); return err; } @@ -389,6 +402,7 @@ err: static void arfs_may_expire_flow(struct mlx5e_priv *priv) { + struct mlx5e_arfs_tables *arfs = mlx5e_fs_get_arfs(priv->fs); struct arfs_rule *arfs_rule; struct hlist_node *htmp; HLIST_HEAD(del_list); @@ -396,8 +410,8 @@ static void arfs_may_expire_flow(struct mlx5e_priv *priv) int i; int j; - spin_lock_bh(&priv->fs->arfs->arfs_lock); - mlx5e_for_each_arfs_rule(arfs_rule, htmp, priv->fs->arfs->arfs_tables, i, j) { + spin_lock_bh(&arfs->arfs_lock); + mlx5e_for_each_arfs_rule(arfs_rule, htmp, arfs->arfs_tables, i, j) { if (!work_pending(&arfs_rule->arfs_work) && rps_may_expire_flow(priv->netdev, arfs_rule->rxq, arfs_rule->flow_id, @@ -408,7 +422,7 @@ static void arfs_may_expire_flow(struct mlx5e_priv *priv) break; } } - spin_unlock_bh(&priv->fs->arfs->arfs_lock); + spin_unlock_bh(&arfs->arfs_lock); hlist_for_each_entry_safe(arfs_rule, htmp, &del_list, hlist) { if (arfs_rule->rule) mlx5_del_flow_rules(arfs_rule->rule); @@ -417,20 +431,21 @@ static void arfs_may_expire_flow(struct mlx5e_priv *priv) } } -static void arfs_del_rules(struct mlx5e_priv *priv) +static void arfs_del_rules(struct mlx5e_flow_steering *fs) { + struct mlx5e_arfs_tables *arfs = mlx5e_fs_get_arfs(fs); struct hlist_node *htmp; struct arfs_rule *rule; HLIST_HEAD(del_list); int i; int j; - spin_lock_bh(&priv->fs->arfs->arfs_lock); - mlx5e_for_each_arfs_rule(rule, htmp, priv->fs->arfs->arfs_tables, i, j) { + spin_lock_bh(&arfs->arfs_lock); + mlx5e_for_each_arfs_rule(rule, htmp, arfs->arfs_tables, i, j) { hlist_del_init(&rule->hlist); hlist_add_head(&rule->hlist, &del_list); } - spin_unlock_bh(&priv->fs->arfs->arfs_lock); + spin_unlock_bh(&arfs->arfs_lock); hlist_for_each_entry_safe(rule, htmp, &del_list, hlist) { cancel_work_sync(&rule->arfs_work); @@ -474,7 +489,7 @@ static struct arfs_table *arfs_get_table(struct mlx5e_arfs_tables *arfs, static struct mlx5_flow_handle *arfs_add_rule(struct mlx5e_priv *priv, struct arfs_rule *arfs_rule) { - struct mlx5e_arfs_tables *arfs = priv->fs->arfs; + struct mlx5e_arfs_tables *arfs = mlx5e_fs_get_arfs(priv->fs); struct arfs_tuple *tuple = &arfs_rule->tuple; struct mlx5_flow_handle *rule = NULL; struct mlx5_flow_destination dest = {}; @@ -588,13 +603,15 @@ static void arfs_handle_work(struct work_struct *work) struct arfs_rule, arfs_work); struct mlx5e_priv *priv = arfs_rule->priv; + struct mlx5e_arfs_tables *arfs; struct mlx5_flow_handle *rule; + arfs = mlx5e_fs_get_arfs(priv->fs); mutex_lock(&priv->state_lock); if (!test_bit(MLX5E_STATE_OPENED, &priv->state)) { - spin_lock_bh(&priv->fs->arfs->arfs_lock); + spin_lock_bh(&arfs->arfs_lock); hlist_del(&arfs_rule->hlist); - spin_unlock_bh(&priv->fs->arfs->arfs_lock); + spin_unlock_bh(&arfs->arfs_lock); mutex_unlock(&priv->state_lock); kfree(arfs_rule); @@ -620,6 +637,7 @@ static struct arfs_rule *arfs_alloc_rule(struct mlx5e_priv *priv, const struct flow_keys *fk, u16 rxq, u32 flow_id) { + struct mlx5e_arfs_tables *arfs = mlx5e_fs_get_arfs(priv->fs); struct arfs_rule *rule; struct arfs_tuple *tuple; @@ -647,7 +665,7 @@ static struct arfs_rule *arfs_alloc_rule(struct mlx5e_priv *priv, tuple->dst_port = fk->ports.dst; rule->flow_id = flow_id; - rule->filter_id = priv->fs->arfs->last_filter_id++ % RPS_NO_FILTER; + rule->filter_id = arfs->last_filter_id++ % RPS_NO_FILTER; hlist_add_head(&rule->hlist, arfs_hash_bucket(arfs_t, tuple->src_port, @@ -691,11 +709,12 @@ int mlx5e_rx_flow_steer(struct net_device *dev, const struct sk_buff *skb, u16 rxq_index, u32 flow_id) { struct mlx5e_priv *priv = netdev_priv(dev); - struct mlx5e_arfs_tables *arfs = priv->fs->arfs; - struct arfs_table *arfs_t; + struct mlx5e_arfs_tables *arfs; struct arfs_rule *arfs_rule; + struct arfs_table *arfs_t; struct flow_keys fk; + arfs = mlx5e_fs_get_arfs(priv->fs); if (!skb_flow_dissect_flow_keys(skb, &fk, 0)) return -EPROTONOSUPPORT; @@ -725,7 +744,7 @@ int mlx5e_rx_flow_steer(struct net_device *dev, const struct sk_buff *skb, return -ENOMEM; } } - queue_work(priv->fs->arfs->wq, &arfs_rule->arfs_work); + queue_work(arfs->wq, &arfs_rule->arfs_work); spin_unlock_bh(&arfs->arfs_lock); return arfs_rule->filter_id; } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_common.c b/drivers/net/ethernet/mellanox/mlx5/core/en_common.c index c0f409c195bf..68f19324db93 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_common.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_common.c @@ -46,8 +46,7 @@ void mlx5e_mkey_set_relaxed_ordering(struct mlx5_core_dev *mdev, void *mkc) MLX5_SET(mkc, mkc, relaxed_ordering_write, ro_pci_enable && ro_write); } -static int mlx5e_create_mkey(struct mlx5_core_dev *mdev, u32 pdn, - u32 *mkey) +int mlx5e_create_mkey(struct mlx5_core_dev *mdev, u32 pdn, u32 *mkey) { int inlen = MLX5_ST_SZ_BYTES(create_mkey_in); void *mkc; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c b/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c index b811207fe5ed..24aa25da482b 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c @@ -38,18 +38,19 @@ #include "en/xsk/pool.h" #include "en/ptp.h" #include "lib/clock.h" +#include "en/fs_ethtool.h" void mlx5e_ethtool_get_drvinfo(struct mlx5e_priv *priv, struct ethtool_drvinfo *drvinfo) { struct mlx5_core_dev *mdev = priv->mdev; - strlcpy(drvinfo->driver, KBUILD_MODNAME, sizeof(drvinfo->driver)); + strscpy(drvinfo->driver, KBUILD_MODNAME, sizeof(drvinfo->driver)); snprintf(drvinfo->fw_version, sizeof(drvinfo->fw_version), "%d.%d.%04d (%.16s)", fw_rev_maj(mdev), fw_rev_min(mdev), fw_rev_sub(mdev), mdev->board_id); - strlcpy(drvinfo->bus_info, dev_name(mdev->device), + strscpy(drvinfo->bus_info, dev_name(mdev->device), sizeof(drvinfo->bus_info)); } @@ -310,7 +311,15 @@ void mlx5e_ethtool_get_ringparam(struct mlx5e_priv *priv, struct ethtool_ringparam *param, struct kernel_ethtool_ringparam *kernel_param) { - param->rx_max_pending = 1 << MLX5E_PARAMS_MAXIMUM_LOG_RQ_SIZE; + /* Limitation for regular RQ. XSK RQ may clamp the queue length in + * mlx5e_mpwqe_get_log_rq_size. + */ + u8 max_log_mpwrq_pkts = mlx5e_mpwrq_max_log_rq_pkts(priv->mdev, + PAGE_SHIFT, + MLX5E_MPWRQ_UMR_MODE_ALIGNED); + + param->rx_max_pending = 1 << min_t(u8, MLX5E_PARAMS_MAXIMUM_LOG_RQ_SIZE, + max_log_mpwrq_pkts); param->tx_max_pending = 1 << MLX5E_PARAMS_MAXIMUM_LOG_SQ_SIZE; param->rx_pending = 1 << priv->channels.params.log_rq_mtu_frames; param->tx_pending = 1 << priv->channels.params.log_sq_size; @@ -494,14 +503,14 @@ int mlx5e_ethtool_set_channels(struct mlx5e_priv *priv, arfs_enabled = opened && (priv->netdev->features & NETIF_F_NTUPLE); if (arfs_enabled) - mlx5e_arfs_disable(priv); + mlx5e_arfs_disable(priv->fs); /* Switch to new channels, set new parameters and close old ones */ err = mlx5e_safe_switch_params(priv, &new_params, mlx5e_num_channels_changed_ctx, NULL, true); if (arfs_enabled) { - int err2 = mlx5e_arfs_enable(priv); + int err2 = mlx5e_arfs_enable(priv->fs); if (err2) netdev_err(priv->netdev, "%s: mlx5e_arfs_enable failed: %d\n", @@ -1996,10 +2005,14 @@ static int set_pflag_rx_striding_rq(struct net_device *netdev, bool enable) struct mlx5e_params new_params; if (enable) { - if (!mlx5e_check_fragmented_striding_rq_cap(mdev)) - return -EOPNOTSUPP; - if (!mlx5e_striding_rq_possible(mdev, &priv->channels.params)) - return -EINVAL; + /* Checking the regular RQ here; mlx5e_validate_xsk_param called + * from mlx5e_open_xsk will check for each XSK queue, and + * mlx5e_safe_switch_params will be reverted if any check fails. + */ + int err = mlx5e_mpwrq_validate_regular(mdev, &priv->channels.params); + + if (err) + return err; } else if (priv->channels.params.packet_merge.type != MLX5E_PACKET_MERGE_NONE) { netdev_warn(netdev, "Can't set legacy RQ with HW-GRO/LRO, disable them first\n"); return -EINVAL; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c b/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c index e0ce5a233d0b..1892ccb889b3 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_fs.c @@ -36,10 +36,38 @@ #include <linux/tcp.h> #include <linux/mlx5/fs.h> #include <linux/mlx5/mpfs.h> -#include "en.h" #include "en_tc.h" #include "lib/mpfs.h" #include "en/ptp.h" +#include "en/fs_ethtool.h" + +struct mlx5e_flow_steering { + struct work_struct set_rx_mode_work; + bool state_destroy; + bool vlan_strip_disable; + struct mlx5_core_dev *mdev; + struct net_device *netdev; + struct mlx5_flow_namespace *ns; + struct mlx5_flow_namespace *egress_ns; +#ifdef CONFIG_MLX5_EN_RXNFC + struct mlx5e_ethtool_steering *ethtool; +#endif + struct mlx5e_tc_table *tc; + struct mlx5e_promisc_table promisc; + struct mlx5e_vlan_table *vlan; + struct mlx5e_l2_table l2; + struct mlx5_ttc_table *ttc; + struct mlx5_ttc_table *inner_ttc; +#ifdef CONFIG_MLX5_EN_ARFS + struct mlx5e_arfs_tables *arfs; +#endif +#ifdef CONFIG_MLX5_EN_TLS + struct mlx5e_accel_fs_tcp *accel_tcp; +#endif + struct mlx5e_fs_udp *udp; + struct mlx5e_fs_any *any; + struct mlx5e_ptp_fs *ptp_fs; +}; static int mlx5e_add_l2_flow_rule(struct mlx5e_flow_steering *fs, struct mlx5e_l2_rule *ai, int type); @@ -148,9 +176,8 @@ static int mlx5e_vport_context_update_vlans(struct mlx5e_flow_steering *fs) max_list_size = 1 << MLX5_CAP_GEN(fs->mdev, log_max_vlan_list); if (list_size > max_list_size) { - mlx5_core_warn(fs->mdev, - "netdev vlans list size (%d) > (%d) max vport list size, some vlans will be dropped\n", - list_size, max_list_size); + fs_warn(fs, "netdev vlans list size (%d) > (%d) max vport list size, some vlans will be dropped\n", + list_size, max_list_size); list_size = max_list_size; } @@ -167,8 +194,8 @@ static int mlx5e_vport_context_update_vlans(struct mlx5e_flow_steering *fs) err = mlx5_modify_nic_vport_vlans(fs->mdev, vlans, list_size); if (err) - mlx5_core_err(fs->mdev, "Failed to modify vport vlans list err(%d)\n", - err); + fs_err(fs, "Failed to modify vport vlans list err(%d)\n", + err); kvfree(vlans); return err; @@ -249,7 +276,7 @@ static int __mlx5e_add_vlan_rule(struct mlx5e_flow_steering *fs, if (IS_ERR(*rule_p)) { err = PTR_ERR(*rule_p); *rule_p = NULL; - mlx5_core_err(fs->mdev, "%s: add rule failed\n", __func__); + fs_err(fs, "%s: add rule failed\n", __func__); } return err; @@ -351,78 +378,78 @@ mlx5e_add_trap_rule(struct mlx5_flow_table *ft, int trap_id, int tir_num) return rule; } -int mlx5e_add_vlan_trap(struct mlx5e_priv *priv, int trap_id, int tir_num) +int mlx5e_add_vlan_trap(struct mlx5e_flow_steering *fs, int trap_id, int tir_num) { - struct mlx5_flow_table *ft = priv->fs->vlan->ft.t; + struct mlx5_flow_table *ft = fs->vlan->ft.t; struct mlx5_flow_handle *rule; int err; rule = mlx5e_add_trap_rule(ft, trap_id, tir_num); if (IS_ERR(rule)) { err = PTR_ERR(rule); - priv->fs->vlan->trap_rule = NULL; - mlx5_core_err(priv->fs->mdev, "%s: add VLAN trap rule failed, err %d\n", - __func__, err); + fs->vlan->trap_rule = NULL; + fs_err(fs, "%s: add VLAN trap rule failed, err %d\n", + __func__, err); return err; } - priv->fs->vlan->trap_rule = rule; + fs->vlan->trap_rule = rule; return 0; } -void mlx5e_remove_vlan_trap(struct mlx5e_priv *priv) +void mlx5e_remove_vlan_trap(struct mlx5e_flow_steering *fs) { - if (priv->fs->vlan->trap_rule) { - mlx5_del_flow_rules(priv->fs->vlan->trap_rule); - priv->fs->vlan->trap_rule = NULL; + if (fs->vlan->trap_rule) { + mlx5_del_flow_rules(fs->vlan->trap_rule); + fs->vlan->trap_rule = NULL; } } -int mlx5e_add_mac_trap(struct mlx5e_priv *priv, int trap_id, int tir_num) +int mlx5e_add_mac_trap(struct mlx5e_flow_steering *fs, int trap_id, int tir_num) { - struct mlx5_flow_table *ft = priv->fs->l2.ft.t; + struct mlx5_flow_table *ft = fs->l2.ft.t; struct mlx5_flow_handle *rule; int err; rule = mlx5e_add_trap_rule(ft, trap_id, tir_num); if (IS_ERR(rule)) { err = PTR_ERR(rule); - priv->fs->l2.trap_rule = NULL; - mlx5_core_err(priv->fs->mdev, "%s: add MAC trap rule failed, err %d\n", - __func__, err); + fs->l2.trap_rule = NULL; + fs_err(fs, "%s: add MAC trap rule failed, err %d\n", + __func__, err); return err; } - priv->fs->l2.trap_rule = rule; + fs->l2.trap_rule = rule; return 0; } -void mlx5e_remove_mac_trap(struct mlx5e_priv *priv) +void mlx5e_remove_mac_trap(struct mlx5e_flow_steering *fs) { - if (priv->fs->l2.trap_rule) { - mlx5_del_flow_rules(priv->fs->l2.trap_rule); - priv->fs->l2.trap_rule = NULL; + if (fs->l2.trap_rule) { + mlx5_del_flow_rules(fs->l2.trap_rule); + fs->l2.trap_rule = NULL; } } -void mlx5e_enable_cvlan_filter(struct mlx5e_priv *priv) +void mlx5e_enable_cvlan_filter(struct mlx5e_flow_steering *fs, bool promisc) { - if (!priv->fs->vlan->cvlan_filter_disabled) + if (!fs->vlan->cvlan_filter_disabled) return; - priv->fs->vlan->cvlan_filter_disabled = false; - if (priv->netdev->flags & IFF_PROMISC) + fs->vlan->cvlan_filter_disabled = false; + if (promisc) return; - mlx5e_fs_del_vlan_rule(priv->fs, MLX5E_VLAN_RULE_TYPE_ANY_CTAG_VID, 0); + mlx5e_fs_del_vlan_rule(fs, MLX5E_VLAN_RULE_TYPE_ANY_CTAG_VID, 0); } -void mlx5e_disable_cvlan_filter(struct mlx5e_priv *priv) +void mlx5e_disable_cvlan_filter(struct mlx5e_flow_steering *fs, bool promisc) { - if (priv->fs->vlan->cvlan_filter_disabled) + if (fs->vlan->cvlan_filter_disabled) return; - priv->fs->vlan->cvlan_filter_disabled = true; - if (priv->netdev->flags & IFF_PROMISC) + fs->vlan->cvlan_filter_disabled = true; + if (promisc) return; - mlx5e_add_vlan_rule(priv->fs, MLX5E_VLAN_RULE_TYPE_ANY_CTAG_VID, 0); + mlx5e_add_vlan_rule(fs, MLX5E_VLAN_RULE_TYPE_ANY_CTAG_VID, 0); } static int mlx5e_vlan_rx_add_cvid(struct mlx5e_flow_steering *fs, u16 vid) @@ -462,7 +489,7 @@ int mlx5e_fs_vlan_rx_add_vid(struct mlx5e_flow_steering *fs, { if (!fs->vlan) { - mlx5_core_err(fs->mdev, "Vlan doesn't exist\n"); + fs_err(fs, "Vlan doesn't exist\n"); return -EINVAL; } @@ -479,7 +506,7 @@ int mlx5e_fs_vlan_rx_kill_vid(struct mlx5e_flow_steering *fs, __be16 proto, u16 vid) { if (!fs->vlan) { - mlx5_core_err(fs->mdev, "Vlan doesn't exist\n"); + fs_err(fs, "Vlan doesn't exist\n"); return -EINVAL; } @@ -512,28 +539,28 @@ static void mlx5e_fs_add_vlan_rules(struct mlx5e_flow_steering *fs) mlx5e_fs_add_any_vid_rules(fs); } -static void mlx5e_del_vlan_rules(struct mlx5e_priv *priv) +static void mlx5e_del_vlan_rules(struct mlx5e_flow_steering *fs) { int i; - mlx5e_fs_del_vlan_rule(priv->fs, MLX5E_VLAN_RULE_TYPE_UNTAGGED, 0); + mlx5e_fs_del_vlan_rule(fs, MLX5E_VLAN_RULE_TYPE_UNTAGGED, 0); - for_each_set_bit(i, priv->fs->vlan->active_cvlans, VLAN_N_VID) { - mlx5e_fs_del_vlan_rule(priv->fs, MLX5E_VLAN_RULE_TYPE_MATCH_CTAG_VID, i); + for_each_set_bit(i, fs->vlan->active_cvlans, VLAN_N_VID) { + mlx5e_fs_del_vlan_rule(fs, MLX5E_VLAN_RULE_TYPE_MATCH_CTAG_VID, i); } - for_each_set_bit(i, priv->fs->vlan->active_svlans, VLAN_N_VID) - mlx5e_fs_del_vlan_rule(priv->fs, MLX5E_VLAN_RULE_TYPE_MATCH_STAG_VID, i); + for_each_set_bit(i, fs->vlan->active_svlans, VLAN_N_VID) + mlx5e_fs_del_vlan_rule(fs, MLX5E_VLAN_RULE_TYPE_MATCH_STAG_VID, i); - WARN_ON_ONCE(priv->fs->state_destroy); + WARN_ON_ONCE(fs->state_destroy); - mlx5e_remove_vlan_trap(priv); + mlx5e_remove_vlan_trap(fs); /* must be called after DESTROY bit is set and * set_rx_mode is called and flushed */ - if (priv->fs->vlan->cvlan_filter_disabled) - mlx5e_fs_del_any_vid_rules(priv->fs); + if (fs->vlan->cvlan_filter_disabled) + mlx5e_fs_del_any_vid_rules(fs); } #define mlx5e_for_each_hash_node(hn, tmp, hash, i) \ @@ -568,8 +595,9 @@ static void mlx5e_execute_l2_action(struct mlx5e_flow_steering *fs, } if (l2_err) - mlx5_core_warn(fs->mdev, "MPFS, failed to %s mac %pM, err(%d)\n", - action == MLX5E_ACTION_ADD ? "add" : "del", mac_addr, l2_err); + fs_warn(fs, "MPFS, failed to %s mac %pM, err(%d)\n", + action == MLX5E_ACTION_ADD ? "add" : "del", + mac_addr, l2_err); } static void mlx5e_sync_netdev_addr(struct mlx5e_flow_steering *fs, @@ -640,9 +668,8 @@ static void mlx5e_vport_context_update_addr_list(struct mlx5e_flow_steering *fs, size++; if (size > max_size) { - mlx5_core_warn(fs->mdev, - "mdev %s list size (%d) > (%d) max vport list size, some addresses will be dropped\n", - is_uc ? "UC" : "MC", size, max_size); + fs_warn(fs, "mdev %s list size (%d) > (%d) max vport list size, some addresses will be dropped\n", + is_uc ? "UC" : "MC", size, max_size); size = max_size; } @@ -658,9 +685,8 @@ static void mlx5e_vport_context_update_addr_list(struct mlx5e_flow_steering *fs, err = mlx5_modify_nic_vport_mac_list(fs->mdev, list_type, addr_array, size); out: if (err) - mlx5_core_err(fs->mdev, - "Failed to modify vport %s list err(%d)\n", - is_uc ? "UC" : "MC", err); + fs_err(fs, "Failed to modify vport %s list err(%d)\n", + is_uc ? "UC" : "MC", err); kfree(addr_array); } @@ -730,7 +756,7 @@ static int mlx5e_add_promisc_rule(struct mlx5e_flow_steering *fs) if (IS_ERR(*rule_p)) { err = PTR_ERR(*rule_p); *rule_p = NULL; - mlx5_core_err(fs->mdev, "%s: add promiscuous rule failed\n", __func__); + fs_err(fs, "%s: add promiscuous rule failed\n", __func__); } kvfree(spec); return err; @@ -750,7 +776,7 @@ static int mlx5e_create_promisc_table(struct mlx5e_flow_steering *fs) ft->t = mlx5_create_auto_grouped_flow_table(fs->ns, &ft_attr); if (IS_ERR(ft->t)) { err = PTR_ERR(ft->t); - mlx5_core_err(fs->mdev, "fail to create promisc table err=%d\n", err); + fs_err(fs, "fail to create promisc table err=%d\n", err); return err; } @@ -807,8 +833,8 @@ void mlx5e_fs_set_rx_mode_work(struct mlx5e_flow_steering *fs, if (err) enable_promisc = false; if (!fs->vlan_strip_disable && !err) - mlx5_core_warn_once(fs->mdev, - "S-tagged traffic will be dropped while C-tag vlan stripping is enabled\n"); + fs_warn_once(fs, + "S-tagged traffic will be dropped while C-tag vlan stripping is enabled\n"); } if (enable_allmulti) mlx5e_add_l2_flow_rule(fs, &ea->allmulti, MLX5E_ALLMULTI); @@ -856,14 +882,15 @@ void mlx5e_destroy_flow_table(struct mlx5e_flow_table *ft) ft->t = NULL; } -static void mlx5e_set_inner_ttc_params(struct mlx5e_priv *priv, +static void mlx5e_set_inner_ttc_params(struct mlx5e_flow_steering *fs, + struct mlx5e_rx_res *rx_res, 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->ns = mlx5_get_flow_namespace(priv->fs->mdev, + ttc_params->ns = mlx5_get_flow_namespace(fs->mdev, MLX5_FLOW_NAMESPACE_KERNEL); ft_attr->level = MLX5E_INNER_TTC_FT_LEVEL; ft_attr->prio = MLX5E_NIC_PRIO; @@ -872,13 +899,14 @@ static void mlx5e_set_inner_ttc_params(struct mlx5e_priv *priv, ttc_params->dests[tt].type = MLX5_FLOW_DESTINATION_TYPE_TIR; ttc_params->dests[tt].tir_num = tt == MLX5_TT_ANY ? - mlx5e_rx_res_get_tirn_direct(priv->rx_res, 0) : - mlx5e_rx_res_get_tirn_rss_inner(priv->rx_res, + mlx5e_rx_res_get_tirn_direct(rx_res, 0) : + mlx5e_rx_res_get_tirn_rss_inner(rx_res, tt); } } -void mlx5e_set_ttc_params(struct mlx5e_priv *priv, +void mlx5e_set_ttc_params(struct mlx5e_flow_steering *fs, + struct mlx5e_rx_res *rx_res, struct ttc_params *ttc_params, bool tunnel) { @@ -886,7 +914,7 @@ void mlx5e_set_ttc_params(struct mlx5e_priv *priv, int tt; memset(ttc_params, 0, sizeof(*ttc_params)); - ttc_params->ns = mlx5_get_flow_namespace(priv->fs->mdev, + ttc_params->ns = mlx5_get_flow_namespace(fs->mdev, MLX5_FLOW_NAMESPACE_KERNEL); ft_attr->level = MLX5E_TTC_FT_LEVEL; ft_attr->prio = MLX5E_NIC_PRIO; @@ -895,19 +923,19 @@ void mlx5e_set_ttc_params(struct mlx5e_priv *priv, ttc_params->dests[tt].type = MLX5_FLOW_DESTINATION_TYPE_TIR; ttc_params->dests[tt].tir_num = tt == MLX5_TT_ANY ? - mlx5e_rx_res_get_tirn_direct(priv->rx_res, 0) : - mlx5e_rx_res_get_tirn_rss(priv->rx_res, tt); + mlx5e_rx_res_get_tirn_direct(rx_res, 0) : + mlx5e_rx_res_get_tirn_rss(rx_res, tt); } ttc_params->inner_ttc = tunnel; - if (!tunnel || !mlx5_tunnel_inner_ft_supported(priv->fs->mdev)) + if (!tunnel || !mlx5_tunnel_inner_ft_supported(fs->mdev)) return; for (tt = 0; tt < MLX5_NUM_TUNNEL_TT; tt++) { ttc_params->tunnel_dests[tt].type = MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE; ttc_params->tunnel_dests[tt].ft = - mlx5_get_ttc_flow_table(priv->fs->inner_ttc); + mlx5_get_ttc_flow_table(fs->inner_ttc); } } @@ -959,8 +987,7 @@ static int mlx5e_add_l2_flow_rule(struct mlx5e_flow_steering *fs, ai->rule = mlx5_add_flow_rules(ft, spec, &flow_act, &dest, 1); if (IS_ERR(ai->rule)) { - mlx5_core_err(fs->mdev, "%s: add l2 rule(mac:%pM) failed\n", - __func__, mv_dmac); + fs_err(fs, "%s: add l2 rule(mac:%pM) failed\n", __func__, mv_dmac); err = PTR_ERR(ai->rule); ai->rule = NULL; } @@ -1044,14 +1071,14 @@ err_destroy_groups: return err; } -static void mlx5e_destroy_l2_table(struct mlx5e_priv *priv) +static void mlx5e_destroy_l2_table(struct mlx5e_flow_steering *fs) { - mlx5e_destroy_flow_table(&priv->fs->l2.ft); + mlx5e_destroy_flow_table(&fs->l2.ft); } -static int mlx5e_create_l2_table(struct mlx5e_priv *priv) +static int mlx5e_create_l2_table(struct mlx5e_flow_steering *fs) { - struct mlx5e_l2_table *l2_table = &priv->fs->l2; + struct mlx5e_l2_table *l2_table = &fs->l2; struct mlx5e_flow_table *ft = &l2_table->ft; struct mlx5_flow_table_attr ft_attr = {}; int err; @@ -1062,7 +1089,7 @@ static int mlx5e_create_l2_table(struct mlx5e_priv *priv) ft_attr.level = MLX5E_L2_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(fs->ns, &ft_attr); if (IS_ERR(ft->t)) { err = PTR_ERR(ft->t); ft->t = NULL; @@ -1221,126 +1248,128 @@ err_destroy_vlan_table: return err; } -static void mlx5e_destroy_vlan_table(struct mlx5e_priv *priv) +static void mlx5e_destroy_vlan_table(struct mlx5e_flow_steering *fs) { - mlx5e_del_vlan_rules(priv); - mlx5e_destroy_flow_table(&priv->fs->vlan->ft); + mlx5e_del_vlan_rules(fs); + mlx5e_destroy_flow_table(&fs->vlan->ft); } -static void mlx5e_destroy_inner_ttc_table(struct mlx5e_priv *priv) +static void mlx5e_destroy_inner_ttc_table(struct mlx5e_flow_steering *fs) { - if (!mlx5_tunnel_inner_ft_supported(priv->fs->mdev)) + if (!mlx5_tunnel_inner_ft_supported(fs->mdev)) return; - mlx5_destroy_ttc_table(priv->fs->inner_ttc); + mlx5_destroy_ttc_table(fs->inner_ttc); } -void mlx5e_destroy_ttc_table(struct mlx5e_priv *priv) +void mlx5e_destroy_ttc_table(struct mlx5e_flow_steering *fs) { - mlx5_destroy_ttc_table(priv->fs->ttc); + mlx5_destroy_ttc_table(fs->ttc); } -static int mlx5e_create_inner_ttc_table(struct mlx5e_priv *priv) +static int mlx5e_create_inner_ttc_table(struct mlx5e_flow_steering *fs, + struct mlx5e_rx_res *rx_res) { struct ttc_params ttc_params = {}; - if (!mlx5_tunnel_inner_ft_supported(priv->fs->mdev)) + if (!mlx5_tunnel_inner_ft_supported(fs->mdev)) return 0; - mlx5e_set_inner_ttc_params(priv, &ttc_params); - priv->fs->inner_ttc = mlx5_create_inner_ttc_table(priv->fs->mdev, - &ttc_params); - if (IS_ERR(priv->fs->inner_ttc)) - return PTR_ERR(priv->fs->inner_ttc); + mlx5e_set_inner_ttc_params(fs, rx_res, &ttc_params); + fs->inner_ttc = mlx5_create_inner_ttc_table(fs->mdev, + &ttc_params); + if (IS_ERR(fs->inner_ttc)) + return PTR_ERR(fs->inner_ttc); return 0; } -int mlx5e_create_ttc_table(struct mlx5e_priv *priv) +int mlx5e_create_ttc_table(struct mlx5e_flow_steering *fs, + struct mlx5e_rx_res *rx_res) { struct ttc_params ttc_params = {}; - mlx5e_set_ttc_params(priv, &ttc_params, true); - priv->fs->ttc = mlx5_create_ttc_table(priv->fs->mdev, &ttc_params); - if (IS_ERR(priv->fs->ttc)) - return PTR_ERR(priv->fs->ttc); + mlx5e_set_ttc_params(fs, rx_res, &ttc_params, true); + fs->ttc = mlx5_create_ttc_table(fs->mdev, &ttc_params); + if (IS_ERR(fs->ttc)) + return PTR_ERR(fs->ttc); return 0; } -int mlx5e_create_flow_steering(struct mlx5e_priv *priv) +int mlx5e_create_flow_steering(struct mlx5e_flow_steering *fs, + struct mlx5e_rx_res *rx_res, + const struct mlx5e_profile *profile, + struct net_device *netdev) { + struct mlx5_flow_namespace *ns = mlx5_get_flow_namespace(fs->mdev, + MLX5_FLOW_NAMESPACE_KERNEL); int err; - priv->fs->ns = mlx5_get_flow_namespace(priv->fs->mdev, - MLX5_FLOW_NAMESPACE_KERNEL); - - if (!priv->fs->ns) + if (!ns) return -EOPNOTSUPP; - err = mlx5e_arfs_create_tables(priv); + mlx5e_fs_set_ns(fs, ns, false); + err = mlx5e_arfs_create_tables(fs, rx_res, + !!(netdev->hw_features & NETIF_F_NTUPLE)); if (err) { - mlx5_core_err(priv->fs->mdev, "Failed to create arfs tables, err=%d\n", - err); - priv->netdev->hw_features &= ~NETIF_F_NTUPLE; + fs_err(fs, "Failed to create arfs tables, err=%d\n", err); + netdev->hw_features &= ~NETIF_F_NTUPLE; } - err = mlx5e_create_inner_ttc_table(priv); + err = mlx5e_create_inner_ttc_table(fs, rx_res); if (err) { - mlx5_core_err(priv->fs->mdev, - "Failed to create inner ttc table, err=%d\n", err); + fs_err(fs, "Failed to create inner ttc table, err=%d\n", err); goto err_destroy_arfs_tables; } - err = mlx5e_create_ttc_table(priv); + err = mlx5e_create_ttc_table(fs, rx_res); if (err) { - mlx5_core_err(priv->fs->mdev, "Failed to create ttc table, err=%d\n", - err); + fs_err(fs, "Failed to create ttc table, err=%d\n", err); goto err_destroy_inner_ttc_table; } - err = mlx5e_create_l2_table(priv); + err = mlx5e_create_l2_table(fs); if (err) { - mlx5_core_err(priv->fs->mdev, "Failed to create l2 table, err=%d\n", - err); + fs_err(fs, "Failed to create l2 table, err=%d\n", err); goto err_destroy_ttc_table; } - err = mlx5e_fs_create_vlan_table(priv->fs); + err = mlx5e_fs_create_vlan_table(fs); if (err) { - mlx5_core_err(priv->fs->mdev, "Failed to create vlan table, err=%d\n", - err); + fs_err(fs, "Failed to create vlan table, err=%d\n", err); goto err_destroy_l2_table; } - err = mlx5e_ptp_alloc_rx_fs(priv); + err = mlx5e_ptp_alloc_rx_fs(fs, profile); if (err) goto err_destory_vlan_table; - mlx5e_ethtool_init_steering(priv); + mlx5e_ethtool_init_steering(fs); return 0; err_destory_vlan_table: - mlx5e_destroy_vlan_table(priv); + mlx5e_destroy_vlan_table(fs); err_destroy_l2_table: - mlx5e_destroy_l2_table(priv); + mlx5e_destroy_l2_table(fs); err_destroy_ttc_table: - mlx5e_destroy_ttc_table(priv); + mlx5e_destroy_ttc_table(fs); err_destroy_inner_ttc_table: - mlx5e_destroy_inner_ttc_table(priv); + mlx5e_destroy_inner_ttc_table(fs); err_destroy_arfs_tables: - mlx5e_arfs_destroy_tables(priv); + mlx5e_arfs_destroy_tables(fs, !!(netdev->hw_features & NETIF_F_NTUPLE)); return err; } -void mlx5e_destroy_flow_steering(struct mlx5e_priv *priv) +void mlx5e_destroy_flow_steering(struct mlx5e_flow_steering *fs, bool ntuple, + const struct mlx5e_profile *profile) { - mlx5e_ptp_free_rx_fs(priv); - mlx5e_destroy_vlan_table(priv); - mlx5e_destroy_l2_table(priv); - mlx5e_destroy_ttc_table(priv); - mlx5e_destroy_inner_ttc_table(priv); - mlx5e_arfs_destroy_tables(priv); - mlx5e_ethtool_cleanup_steering(priv); + mlx5e_ptp_free_rx_fs(fs, profile); + mlx5e_destroy_vlan_table(fs); + mlx5e_destroy_l2_table(fs); + mlx5e_destroy_ttc_table(fs); + mlx5e_destroy_inner_ttc_table(fs); + mlx5e_arfs_destroy_tables(fs, ntuple); + mlx5e_ethtool_cleanup_steering(fs); } static int mlx5e_fs_vlan_alloc(struct mlx5e_flow_steering *fs) @@ -1356,6 +1385,11 @@ static void mlx5e_fs_vlan_free(struct mlx5e_flow_steering *fs) kvfree(fs->vlan); } +struct mlx5e_vlan_table *mlx5e_fs_get_vlan(struct mlx5e_flow_steering *fs) +{ + return fs->vlan; +} + static int mlx5e_fs_tc_alloc(struct mlx5e_flow_steering *fs) { fs->tc = mlx5e_tc_table_alloc(); @@ -1369,6 +1403,32 @@ static void mlx5e_fs_tc_free(struct mlx5e_flow_steering *fs) mlx5e_tc_table_free(fs->tc); } +struct mlx5e_tc_table *mlx5e_fs_get_tc(struct mlx5e_flow_steering *fs) +{ + return fs->tc; +} + +#ifdef CONFIG_MLX5_EN_RXNFC +static int mlx5e_fs_ethtool_alloc(struct mlx5e_flow_steering *fs) +{ + return mlx5e_ethtool_alloc(&fs->ethtool); +} + +static void mlx5e_fs_ethtool_free(struct mlx5e_flow_steering *fs) +{ + mlx5e_ethtool_free(fs->ethtool); +} + +struct mlx5e_ethtool_steering *mlx5e_fs_get_ethtool(struct mlx5e_flow_steering *fs) +{ + return fs->ethtool; +} +#else +static int mlx5e_fs_ethtool_alloc(struct mlx5e_flow_steering *fs) +{ return 0; } +static void mlx5e_fs_ethtool_free(struct mlx5e_flow_steering *fs) { } +#endif + struct mlx5e_flow_steering *mlx5e_fs_init(const struct mlx5e_profile *profile, struct mlx5_core_dev *mdev, bool state_destroy) @@ -1394,8 +1454,13 @@ struct mlx5e_flow_steering *mlx5e_fs_init(const struct mlx5e_profile *profile, goto err_free_vlan; } - return fs; + err = mlx5e_fs_ethtool_alloc(fs); + if (err) + goto err_free_tc; + return fs; +err_free_tc: + mlx5e_fs_tc_free(fs); err_free_vlan: mlx5e_fs_vlan_free(fs); err_free_fs: @@ -1406,7 +1471,109 @@ err: void mlx5e_fs_cleanup(struct mlx5e_flow_steering *fs) { + mlx5e_fs_ethtool_free(fs); mlx5e_fs_tc_free(fs); mlx5e_fs_vlan_free(fs); kvfree(fs); } + +struct mlx5e_l2_table *mlx5e_fs_get_l2(struct mlx5e_flow_steering *fs) +{ + return &fs->l2; +} + +struct mlx5_flow_namespace *mlx5e_fs_get_ns(struct mlx5e_flow_steering *fs, bool egress) +{ + return egress ? fs->egress_ns : fs->ns; +} + +void mlx5e_fs_set_ns(struct mlx5e_flow_steering *fs, struct mlx5_flow_namespace *ns, bool egress) +{ + if (!egress) + fs->ns = ns; + else + fs->egress_ns = ns; +} + +struct mlx5_ttc_table *mlx5e_fs_get_ttc(struct mlx5e_flow_steering *fs, bool inner) +{ + return inner ? fs->inner_ttc : fs->ttc; +} + +void mlx5e_fs_set_ttc(struct mlx5e_flow_steering *fs, struct mlx5_ttc_table *ttc, bool inner) +{ + if (!inner) + fs->ttc = ttc; + else + fs->inner_ttc = ttc; +} + +#ifdef CONFIG_MLX5_EN_ARFS +struct mlx5e_arfs_tables *mlx5e_fs_get_arfs(struct mlx5e_flow_steering *fs) +{ + return fs->arfs; +} + +void mlx5e_fs_set_arfs(struct mlx5e_flow_steering *fs, struct mlx5e_arfs_tables *arfs) +{ + fs->arfs = arfs; +} +#endif + +struct mlx5e_ptp_fs *mlx5e_fs_get_ptp(struct mlx5e_flow_steering *fs) +{ + return fs->ptp_fs; +} + +void mlx5e_fs_set_ptp(struct mlx5e_flow_steering *fs, struct mlx5e_ptp_fs *ptp_fs) +{ + fs->ptp_fs = ptp_fs; +} + +struct mlx5e_fs_any *mlx5e_fs_get_any(struct mlx5e_flow_steering *fs) +{ + return fs->any; +} + +void mlx5e_fs_set_any(struct mlx5e_flow_steering *fs, struct mlx5e_fs_any *any) +{ + fs->any = any; +} + +#ifdef CONFIG_MLX5_EN_TLS +struct mlx5e_accel_fs_tcp *mlx5e_fs_get_accel_tcp(struct mlx5e_flow_steering *fs) +{ + return fs->accel_tcp; +} + +void mlx5e_fs_set_accel_tcp(struct mlx5e_flow_steering *fs, struct mlx5e_accel_fs_tcp *accel_tcp) +{ + fs->accel_tcp = accel_tcp; +} +#endif + +void mlx5e_fs_set_state_destroy(struct mlx5e_flow_steering *fs, bool state_destroy) +{ + fs->state_destroy = state_destroy; +} + +void mlx5e_fs_set_vlan_strip_disable(struct mlx5e_flow_steering *fs, + bool vlan_strip_disable) +{ + fs->vlan_strip_disable = vlan_strip_disable; +} + +struct mlx5e_fs_udp *mlx5e_fs_get_udp(struct mlx5e_flow_steering *fs) +{ + return fs->udp; +} + +void mlx5e_fs_set_udp(struct mlx5e_flow_steering *fs, struct mlx5e_fs_udp *udp) +{ + fs->udp = udp; +} + +struct mlx5_core_dev *mlx5e_fs_get_mdev(struct mlx5e_flow_steering *fs) +{ + return fs->mdev; +} diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c b/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c index 3e4bc7836ef4..aac32e505c14 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c @@ -34,6 +34,22 @@ #include "en.h" #include "en/params.h" #include "en/xsk/pool.h" +#include "en/fs_ethtool.h" + +struct mlx5e_ethtool_table { + struct mlx5_flow_table *ft; + int num_rules; +}; + +#define ETHTOOL_NUM_L3_L4_FTS 7 +#define ETHTOOL_NUM_L2_FTS 4 + +struct mlx5e_ethtool_steering { + struct mlx5e_ethtool_table l3_l4_ft[ETHTOOL_NUM_L3_L4_FTS]; + struct mlx5e_ethtool_table l2_ft[ETHTOOL_NUM_L2_FTS]; + struct list_head rules; + int tot_num_rules; +}; static int flow_type_to_traffic_type(u32 flow_type); @@ -66,6 +82,7 @@ static struct mlx5e_ethtool_table *get_flow_table(struct mlx5e_priv *priv, struct ethtool_rx_flow_spec *fs, int num_tuples) { + struct mlx5e_ethtool_steering *ethtool = mlx5e_fs_get_ethtool(priv->fs); struct mlx5_flow_table_attr ft_attr = {}; struct mlx5e_ethtool_table *eth_ft; struct mlx5_flow_namespace *ns; @@ -81,18 +98,18 @@ static struct mlx5e_ethtool_table *get_flow_table(struct mlx5e_priv *priv, case UDP_V6_FLOW: max_tuples = ETHTOOL_NUM_L3_L4_FTS; prio = MLX5E_ETHTOOL_L3_L4_PRIO + (max_tuples - num_tuples); - eth_ft = &priv->fs->ethtool.l3_l4_ft[prio]; + eth_ft = ðtool->l3_l4_ft[prio]; break; case IP_USER_FLOW: case IPV6_USER_FLOW: max_tuples = ETHTOOL_NUM_L3_L4_FTS; prio = MLX5E_ETHTOOL_L3_L4_PRIO + (max_tuples - num_tuples); - eth_ft = &priv->fs->ethtool.l3_l4_ft[prio]; + eth_ft = ðtool->l3_l4_ft[prio]; break; case ETHER_FLOW: max_tuples = ETHTOOL_NUM_L2_FTS; prio = max_tuples - num_tuples; - eth_ft = &priv->fs->ethtool.l2_ft[prio]; + eth_ft = ðtool->l2_ft[prio]; prio += MLX5E_ETHTOOL_L2_PRIO; break; default: @@ -382,15 +399,16 @@ static int set_flow_attrs(u32 *match_c, u32 *match_v, static void add_rule_to_list(struct mlx5e_priv *priv, struct mlx5e_ethtool_rule *rule) { + struct mlx5e_ethtool_steering *ethtool = mlx5e_fs_get_ethtool(priv->fs); + struct list_head *head = ðtool->rules; struct mlx5e_ethtool_rule *iter; - struct list_head *head = &priv->fs->ethtool.rules; - list_for_each_entry(iter, &priv->fs->ethtool.rules, list) { + list_for_each_entry(iter, ðtool->rules, list) { if (iter->flow_spec.location > rule->flow_spec.location) break; head = &iter->list; } - priv->fs->ethtool.tot_num_rules++; + ethtool->tot_num_rules++; list_add(&rule->list, head); } @@ -433,15 +451,7 @@ static int flow_get_tirn(struct mlx5e_priv *priv, eth_rule->rss = rss; mlx5e_rss_refcnt_inc(eth_rule->rss); } else { - struct mlx5e_params *params = &priv->channels.params; - enum mlx5e_rq_group group; - u16 ix; - - mlx5e_qid_get_ch_and_group(params, fs->ring_cookie, &ix, &group); - - *tirn = group == MLX5E_RQ_GROUP_XSK ? - mlx5e_rx_res_get_tirn_xsk(priv->rx_res, ix) : - mlx5e_rx_res_get_tirn_direct(priv->rx_res, ix); + *tirn = mlx5e_rx_res_get_tirn_direct(priv->rx_res, fs->ring_cookie); } return 0; @@ -499,15 +509,16 @@ free: return err ? ERR_PTR(err) : rule; } -static void del_ethtool_rule(struct mlx5e_priv *priv, +static void del_ethtool_rule(struct mlx5e_flow_steering *fs, struct mlx5e_ethtool_rule *eth_rule) { + struct mlx5e_ethtool_steering *ethtool = mlx5e_fs_get_ethtool(fs); if (eth_rule->rule) mlx5_del_flow_rules(eth_rule->rule); if (eth_rule->rss) mlx5e_rss_refcnt_dec(eth_rule->rss); list_del(ð_rule->list); - priv->fs->ethtool.tot_num_rules--; + ethtool->tot_num_rules--; put_flow_table(eth_rule->eth_ft); kfree(eth_rule); } @@ -515,9 +526,10 @@ static void del_ethtool_rule(struct mlx5e_priv *priv, static struct mlx5e_ethtool_rule *find_ethtool_rule(struct mlx5e_priv *priv, int location) { + struct mlx5e_ethtool_steering *ethtool = mlx5e_fs_get_ethtool(priv->fs); struct mlx5e_ethtool_rule *iter; - list_for_each_entry(iter, &priv->fs->ethtool.rules, list) { + list_for_each_entry(iter, ðtool->rules, list) { if (iter->flow_spec.location == location) return iter; } @@ -531,7 +543,7 @@ static struct mlx5e_ethtool_rule *get_ethtool_rule(struct mlx5e_priv *priv, eth_rule = find_ethtool_rule(priv, location); if (eth_rule) - del_ethtool_rule(priv, eth_rule); + del_ethtool_rule(priv->fs, eth_rule); eth_rule = kzalloc(sizeof(*eth_rule), GFP_KERNEL); if (!eth_rule) @@ -662,8 +674,7 @@ static int validate_flow(struct mlx5e_priv *priv, return -ENOSPC; if (fs->ring_cookie != RX_CLS_FLOW_DISC) - if (!mlx5e_qid_validate(priv->profile, &priv->channels.params, - fs->ring_cookie)) + if (fs->ring_cookie >= priv->channels.params.num_channels) return -EINVAL; switch (flow_type_mask(fs->flow_type)) { @@ -754,7 +765,7 @@ mlx5e_ethtool_flow_replace(struct mlx5e_priv *priv, return 0; del_ethtool_rule: - del_ethtool_rule(priv, eth_rule); + del_ethtool_rule(priv->fs, eth_rule); return err; } @@ -774,7 +785,7 @@ mlx5e_ethtool_flow_remove(struct mlx5e_priv *priv, int location) goto out; } - del_ethtool_rule(priv, eth_rule); + del_ethtool_rule(priv->fs, eth_rule); out: return err; } @@ -783,12 +794,13 @@ static int mlx5e_ethtool_get_flow(struct mlx5e_priv *priv, struct ethtool_rxnfc *info, int location) { + struct mlx5e_ethtool_steering *ethtool = mlx5e_fs_get_ethtool(priv->fs); struct mlx5e_ethtool_rule *eth_rule; if (location < 0 || location >= MAX_NUM_OF_ETHTOOL_RULES) return -EINVAL; - list_for_each_entry(eth_rule, &priv->fs->ethtool.rules, list) { + list_for_each_entry(eth_rule, ðtool->rules, list) { int index; if (eth_rule->flow_spec.location != location) @@ -826,18 +838,34 @@ mlx5e_ethtool_get_all_flows(struct mlx5e_priv *priv, return err; } -void mlx5e_ethtool_cleanup_steering(struct mlx5e_priv *priv) +int mlx5e_ethtool_alloc(struct mlx5e_ethtool_steering **ethtool) { + *ethtool = kvzalloc(sizeof(**ethtool), GFP_KERNEL); + if (!*ethtool) + return -ENOMEM; + return 0; +} + +void mlx5e_ethtool_free(struct mlx5e_ethtool_steering *ethtool) +{ + kvfree(ethtool); +} + +void mlx5e_ethtool_cleanup_steering(struct mlx5e_flow_steering *fs) +{ + struct mlx5e_ethtool_steering *ethtool = mlx5e_fs_get_ethtool(fs); struct mlx5e_ethtool_rule *iter; struct mlx5e_ethtool_rule *temp; - list_for_each_entry_safe(iter, temp, &priv->fs->ethtool.rules, list) - del_ethtool_rule(priv, iter); + list_for_each_entry_safe(iter, temp, ðtool->rules, list) + del_ethtool_rule(fs, iter); } -void mlx5e_ethtool_init_steering(struct mlx5e_priv *priv) +void mlx5e_ethtool_init_steering(struct mlx5e_flow_steering *fs) { - INIT_LIST_HEAD(&priv->fs->ethtool.rules); + struct mlx5e_ethtool_steering *ethtool = mlx5e_fs_get_ethtool(fs); + + INIT_LIST_HEAD(ðtool->rules); } static int flow_type_to_traffic_type(u32 flow_type) @@ -959,11 +987,12 @@ int mlx5e_ethtool_set_rxnfc(struct mlx5e_priv *priv, struct ethtool_rxnfc *cmd) int mlx5e_ethtool_get_rxnfc(struct mlx5e_priv *priv, struct ethtool_rxnfc *info, u32 *rule_locs) { + struct mlx5e_ethtool_steering *ethtool = mlx5e_fs_get_ethtool(priv->fs); int err = 0; switch (info->cmd) { case ETHTOOL_GRXCLSRLCNT: - info->rule_cnt = priv->fs->ethtool.tot_num_rules; + info->rule_cnt = ethtool->tot_num_rules; break; case ETHTOOL_GRXCLSRULE: err = mlx5e_ethtool_get_flow(priv, info, info->fs.location); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c index 02eb2f0fa2ae..364f04309149 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c @@ -45,6 +45,7 @@ #include "en_tc.h" #include "en_rep.h" #include "en_accel/ipsec.h" +#include "en_accel/macsec.h" #include "en_accel/en_accel.h" #include "en_accel/ktls.h" #include "lib/vxlan.h" @@ -67,22 +68,25 @@ #include "qos.h" #include "en/trap.h" -bool mlx5e_check_fragmented_striding_rq_cap(struct mlx5_core_dev *mdev) +bool mlx5e_check_fragmented_striding_rq_cap(struct mlx5_core_dev *mdev, u8 page_shift, + enum mlx5e_mpwrq_umr_mode umr_mode) { - bool striding_rq_umr, inline_umr; - u16 max_wqe_sz_cap; + u16 umr_wqebbs, max_wqebbs; + bool striding_rq_umr; striding_rq_umr = MLX5_CAP_GEN(mdev, striding_rq) && MLX5_CAP_GEN(mdev, umr_ptr_rlky) && MLX5_CAP_ETH(mdev, reg_umr_sq); - max_wqe_sz_cap = mlx5e_get_max_sq_wqebbs(mdev) * MLX5_SEND_WQE_BB; - inline_umr = max_wqe_sz_cap >= MLX5E_UMR_WQE_INLINE_SZ; if (!striding_rq_umr) return false; - if (!inline_umr) { - mlx5_core_warn(mdev, "Cannot support Striding RQ: UMR WQE size (%d) exceeds maximum supported (%d).\n", - (int)MLX5E_UMR_WQE_INLINE_SZ, max_wqe_sz_cap); + + umr_wqebbs = mlx5e_mpwrq_umr_wqebbs(mdev, page_shift, umr_mode); + max_wqebbs = mlx5e_get_max_sq_aligned_wqebbs(mdev); + /* Sanity check; should never happen, because mlx5e_mpwrq_umr_wqebbs is + * calculated from mlx5e_get_max_sq_aligned_wqebbs. + */ + if (WARN_ON(umr_wqebbs > max_wqebbs)) return false; - } + return true; } @@ -199,21 +203,35 @@ static void mlx5e_disable_blocking_events(struct mlx5e_priv *priv) mlx5_blocking_notifier_unregister(priv->mdev, &priv->blocking_events_nb); } +static u16 mlx5e_mpwrq_umr_octowords(u32 entries, enum mlx5e_mpwrq_umr_mode umr_mode) +{ + u8 umr_entry_size = mlx5e_mpwrq_umr_entry_size(umr_mode); + + WARN_ON_ONCE(entries * umr_entry_size % MLX5_OCTWORD); + + return entries * umr_entry_size / MLX5_OCTWORD; +} + static inline void mlx5e_build_umr_wqe(struct mlx5e_rq *rq, struct mlx5e_icosq *sq, struct mlx5e_umr_wqe *wqe) { struct mlx5_wqe_ctrl_seg *cseg = &wqe->ctrl; struct mlx5_wqe_umr_ctrl_seg *ucseg = &wqe->uctrl; - u8 ds_cnt = DIV_ROUND_UP(MLX5E_UMR_WQE_INLINE_SZ, MLX5_SEND_WQE_DS); + u16 octowords; + u8 ds_cnt; + + ds_cnt = DIV_ROUND_UP(mlx5e_mpwrq_umr_wqe_sz(rq->mdev, rq->mpwqe.page_shift, + rq->mpwqe.umr_mode), + MLX5_SEND_WQE_DS); cseg->qpn_ds = cpu_to_be32((sq->sqn << MLX5_WQE_CTRL_QPN_SHIFT) | ds_cnt); - cseg->umr_mkey = rq->mkey_be; + cseg->umr_mkey = rq->mpwqe.umr_mkey_be; ucseg->flags = MLX5_UMR_TRANSLATION_OFFSET_EN | MLX5_UMR_INLINE; - ucseg->xlt_octowords = - cpu_to_be16(MLX5_MTT_OCTW(MLX5_MPWRQ_PAGES_PER_WQE)); + octowords = mlx5e_mpwrq_umr_octowords(rq->mpwqe.pages_per_wqe, rq->mpwqe.umr_mode); + ucseg->xlt_octowords = cpu_to_be16(octowords); ucseg->mkey_mask = cpu_to_be64(MLX5_MKEY_MASK_FREE); } @@ -259,10 +277,12 @@ static void mlx5e_rq_shampo_hd_info_free(struct mlx5e_rq *rq) static int mlx5e_rq_alloc_mpwqe_info(struct mlx5e_rq *rq, int node) { int wq_sz = mlx5_wq_ll_get_size(&rq->mpwqe.wq); + size_t alloc_size; + + alloc_size = array_size(wq_sz, struct_size(rq->mpwqe.info, alloc_units, + rq->mpwqe.pages_per_wqe)); - rq->mpwqe.info = kvzalloc_node(array_size(wq_sz, - sizeof(*rq->mpwqe.info)), - GFP_KERNEL, node); + rq->mpwqe.info = kvzalloc_node(alloc_size, GFP_KERNEL, node); if (!rq->mpwqe.info) return -ENOMEM; @@ -271,18 +291,52 @@ static int mlx5e_rq_alloc_mpwqe_info(struct mlx5e_rq *rq, int node) return 0; } -static int mlx5e_create_umr_mtt_mkey(struct mlx5_core_dev *mdev, - u64 npages, u8 page_shift, u32 *umr_mkey, - dma_addr_t filler_addr) + +static u8 mlx5e_mpwrq_access_mode(enum mlx5e_mpwrq_umr_mode umr_mode) +{ + switch (umr_mode) { + case MLX5E_MPWRQ_UMR_MODE_ALIGNED: + return MLX5_MKC_ACCESS_MODE_MTT; + case MLX5E_MPWRQ_UMR_MODE_UNALIGNED: + return MLX5_MKC_ACCESS_MODE_KSM; + case MLX5E_MPWRQ_UMR_MODE_OVERSIZED: + return MLX5_MKC_ACCESS_MODE_KLMS; + case MLX5E_MPWRQ_UMR_MODE_TRIPLE: + return MLX5_MKC_ACCESS_MODE_KSM; + } + WARN_ONCE(1, "MPWRQ UMR mode %d is not known\n", umr_mode); + return 0; +} + +static int mlx5e_create_umr_mkey(struct mlx5_core_dev *mdev, + u32 npages, u8 page_shift, u32 *umr_mkey, + dma_addr_t filler_addr, + enum mlx5e_mpwrq_umr_mode umr_mode, + u32 xsk_chunk_size) { struct mlx5_mtt *mtt; + struct mlx5_ksm *ksm; + struct mlx5_klm *klm; + u32 octwords; int inlen; void *mkc; u32 *in; int err; int i; - inlen = MLX5_ST_SZ_BYTES(create_mkey_in) + sizeof(*mtt) * npages; + if ((umr_mode == MLX5E_MPWRQ_UMR_MODE_UNALIGNED || + umr_mode == MLX5E_MPWRQ_UMR_MODE_TRIPLE) && + !MLX5_CAP_GEN(mdev, fixed_buffer_size)) { + mlx5_core_warn(mdev, "Unaligned AF_XDP requires fixed_buffer_size capability\n"); + return -EINVAL; + } + + octwords = mlx5e_mpwrq_umr_octowords(npages, umr_mode); + + inlen = MLX5_FLEXIBLE_INLEN(mdev, MLX5_ST_SZ_BYTES(create_mkey_in), + MLX5_OCTWORD, octwords); + if (inlen < 0) + return inlen; in = kvzalloc(inlen, GFP_KERNEL); if (!in) @@ -294,16 +348,17 @@ static int mlx5e_create_umr_mtt_mkey(struct mlx5_core_dev *mdev, MLX5_SET(mkc, mkc, umr_en, 1); MLX5_SET(mkc, mkc, lw, 1); MLX5_SET(mkc, mkc, lr, 1); - MLX5_SET(mkc, mkc, access_mode_1_0, MLX5_MKC_ACCESS_MODE_MTT); + MLX5_SET(mkc, mkc, access_mode_1_0, mlx5e_mpwrq_access_mode(umr_mode)); mlx5e_mkey_set_relaxed_ordering(mdev, mkc); MLX5_SET(mkc, mkc, qpn, 0xffffff); MLX5_SET(mkc, mkc, pd, mdev->mlx5e_res.hw_objs.pdn); MLX5_SET64(mkc, mkc, len, npages << page_shift); - MLX5_SET(mkc, mkc, translations_octword_size, - MLX5_MTT_OCTW(npages)); - MLX5_SET(mkc, mkc, log_page_size, page_shift); - MLX5_SET(create_mkey_in, in, translations_octword_actual_size, - MLX5_MTT_OCTW(npages)); + MLX5_SET(mkc, mkc, translations_octword_size, octwords); + if (umr_mode == MLX5E_MPWRQ_UMR_MODE_TRIPLE) + MLX5_SET(mkc, mkc, log_page_size, page_shift - 2); + else if (umr_mode != MLX5E_MPWRQ_UMR_MODE_OVERSIZED) + MLX5_SET(mkc, mkc, log_page_size, page_shift); + MLX5_SET(create_mkey_in, in, translations_octword_actual_size, octwords); /* Initialize the mkey with all MTTs pointing to a default * page (filler_addr). When the channels are activated, UMR @@ -311,9 +366,47 @@ static int mlx5e_create_umr_mtt_mkey(struct mlx5_core_dev *mdev, * the RQ's pool, while the gaps (wqe_overflow) remain mapped * to the default page. */ - mtt = MLX5_ADDR_OF(create_mkey_in, in, klm_pas_mtt); - for (i = 0 ; i < npages ; i++) - mtt[i].ptag = cpu_to_be64(filler_addr); + switch (umr_mode) { + case MLX5E_MPWRQ_UMR_MODE_OVERSIZED: + klm = MLX5_ADDR_OF(create_mkey_in, in, klm_pas_mtt); + for (i = 0; i < npages; i++) { + klm[i << 1] = (struct mlx5_klm) { + .va = cpu_to_be64(filler_addr), + .bcount = cpu_to_be32(xsk_chunk_size), + .key = cpu_to_be32(mdev->mlx5e_res.hw_objs.mkey), + }; + klm[(i << 1) + 1] = (struct mlx5_klm) { + .va = cpu_to_be64(filler_addr), + .bcount = cpu_to_be32((1 << page_shift) - xsk_chunk_size), + .key = cpu_to_be32(mdev->mlx5e_res.hw_objs.mkey), + }; + } + break; + case MLX5E_MPWRQ_UMR_MODE_UNALIGNED: + ksm = MLX5_ADDR_OF(create_mkey_in, in, klm_pas_mtt); + for (i = 0; i < npages; i++) + ksm[i] = (struct mlx5_ksm) { + .key = cpu_to_be32(mdev->mlx5e_res.hw_objs.mkey), + .va = cpu_to_be64(filler_addr), + }; + break; + case MLX5E_MPWRQ_UMR_MODE_ALIGNED: + mtt = MLX5_ADDR_OF(create_mkey_in, in, klm_pas_mtt); + for (i = 0; i < npages; i++) + mtt[i] = (struct mlx5_mtt) { + .ptag = cpu_to_be64(filler_addr), + }; + break; + case MLX5E_MPWRQ_UMR_MODE_TRIPLE: + ksm = MLX5_ADDR_OF(create_mkey_in, in, klm_pas_mtt); + for (i = 0; i < npages * 4; i++) { + ksm[i] = (struct mlx5_ksm) { + .key = cpu_to_be32(mdev->mlx5e_res.hw_objs.mkey), + .va = cpu_to_be64(filler_addr), + }; + } + break; + } err = mlx5_core_create_mkey(mdev, umr_mkey, in, inlen); @@ -356,10 +449,27 @@ static int mlx5e_create_umr_klm_mkey(struct mlx5_core_dev *mdev, static int mlx5e_create_rq_umr_mkey(struct mlx5_core_dev *mdev, struct mlx5e_rq *rq) { - u64 num_mtts = MLX5E_REQUIRED_MTTS(mlx5_wq_ll_get_size(&rq->mpwqe.wq)); + u32 xsk_chunk_size = rq->xsk_pool ? rq->xsk_pool->chunk_size : 0; + u32 wq_size = mlx5_wq_ll_get_size(&rq->mpwqe.wq); + u32 num_entries, max_num_entries; + u32 umr_mkey; + int err; + + max_num_entries = mlx5e_mpwrq_max_num_entries(mdev, rq->mpwqe.umr_mode); + + /* Shouldn't overflow, the result is at most MLX5E_MAX_RQ_NUM_MTTS. */ + if (WARN_ON_ONCE(check_mul_overflow(wq_size, (u32)rq->mpwqe.mtts_per_wqe, + &num_entries) || + num_entries > max_num_entries)) + mlx5_core_err(mdev, "%s: multiplication overflow: %u * %u > %u\n", + __func__, wq_size, rq->mpwqe.mtts_per_wqe, + max_num_entries); - return mlx5e_create_umr_mtt_mkey(mdev, num_mtts, PAGE_SHIFT, - &rq->umr_mkey, rq->wqe_overflow.addr); + err = mlx5e_create_umr_mkey(mdev, num_entries, rq->mpwqe.page_shift, + &umr_mkey, rq->wqe_overflow.addr, + rq->mpwqe.umr_mode, xsk_chunk_size); + rq->mpwqe.umr_mkey_be = cpu_to_be32(umr_mkey); + return err; } static int mlx5e_create_rq_hd_umr_mkey(struct mlx5_core_dev *mdev, @@ -376,18 +486,20 @@ static int mlx5e_create_rq_hd_umr_mkey(struct mlx5_core_dev *mdev, &rq->mpwqe.shampo->mkey); } -static u64 mlx5e_get_mpwqe_offset(u16 wqe_ix) -{ - return MLX5E_REQUIRED_MTTS(wqe_ix) << PAGE_SHIFT; -} - static void mlx5e_init_frags_partition(struct mlx5e_rq *rq) { struct mlx5e_wqe_frag_info next_frag = {}; struct mlx5e_wqe_frag_info *prev = NULL; int i; - next_frag.di = &rq->wqe.di[0]; + if (rq->xsk_pool) { + /* Assumptions used by XSK batched allocator. */ + WARN_ON(rq->wqe.info.num_frags != 1); + WARN_ON(rq->wqe.info.log_num_frags != 0); + WARN_ON(rq->wqe.info.arr[0].frag_stride != PAGE_SIZE); + } + + next_frag.au = &rq->wqe.alloc_units[0]; for (i = 0; i < mlx5_wq_cyc_get_size(&rq->wqe.wq); i++) { struct mlx5e_rq_frag_info *frag_info = &rq->wqe.info.arr[0]; @@ -397,7 +509,7 @@ static void mlx5e_init_frags_partition(struct mlx5e_rq *rq) for (f = 0; f < rq->wqe.info.num_frags; f++, frag++) { if (next_frag.offset + frag_info[f].frag_stride > PAGE_SIZE) { - next_frag.di++; + next_frag.au++; next_frag.offset = 0; if (prev) prev->last_in_page = true; @@ -414,12 +526,13 @@ static void mlx5e_init_frags_partition(struct mlx5e_rq *rq) prev->last_in_page = true; } -int mlx5e_init_di_list(struct mlx5e_rq *rq, int wq_sz, int node) +static int mlx5e_init_au_list(struct mlx5e_rq *rq, int wq_sz, int node) { int len = wq_sz << rq->wqe.info.log_num_frags; - rq->wqe.di = kvzalloc_node(array_size(len, sizeof(*rq->wqe.di)), GFP_KERNEL, node); - if (!rq->wqe.di) + rq->wqe.alloc_units = kvzalloc_node(array_size(len, sizeof(*rq->wqe.alloc_units)), + GFP_KERNEL, node); + if (!rq->wqe.alloc_units) return -ENOMEM; mlx5e_init_frags_partition(rq); @@ -427,9 +540,9 @@ int mlx5e_init_di_list(struct mlx5e_rq *rq, int wq_sz, int node) return 0; } -void mlx5e_free_di_list(struct mlx5e_rq *rq) +static void mlx5e_free_au_list(struct mlx5e_rq *rq) { - kvfree(rq->wqe.di); + kvfree(rq->wqe.alloc_units); } static void mlx5e_rq_err_cqe_work(struct work_struct *recover_work) @@ -485,7 +598,7 @@ static int mlx5e_init_rxq_rq(struct mlx5e_channel *c, struct mlx5e_params *param if (err) return err; - return xdp_rxq_info_reg(&rq->xdp_rxq, rq->netdev, rq->ix, 0); + return xdp_rxq_info_reg(&rq->xdp_rxq, rq->netdev, rq->ix, c->napi.napi_id); } static int mlx5_rq_shampo_alloc(struct mlx5_core_dev *mdev, @@ -572,6 +685,8 @@ static int mlx5e_alloc_rq(struct mlx5e_params *params, rq->buff.headroom = mlx5e_get_rq_headroom(mdev, params, xsk); pool_size = 1 << params->log_rq_mtu_frames; + rq->mkey_be = cpu_to_be32(mdev->mlx5e_res.hw_objs.mkey); + switch (rq->wq_type) { case MLX5_WQ_TYPE_LINKED_LIST_STRIDING_RQ: err = mlx5_wq_ll_create(mdev, &rqp->wq, rqc_wq, &rq->mpwqe.wq, @@ -587,8 +702,20 @@ static int mlx5e_alloc_rq(struct mlx5e_params *params, wq_sz = mlx5_wq_ll_get_size(&rq->mpwqe.wq); - pool_size = MLX5_MPWRQ_PAGES_PER_WQE << - mlx5e_mpwqe_get_log_rq_size(params, xsk); + rq->mpwqe.page_shift = mlx5e_mpwrq_page_shift(mdev, xsk); + rq->mpwqe.umr_mode = mlx5e_mpwrq_umr_mode(mdev, xsk); + rq->mpwqe.pages_per_wqe = + mlx5e_mpwrq_pages_per_wqe(mdev, rq->mpwqe.page_shift, + rq->mpwqe.umr_mode); + rq->mpwqe.umr_wqebbs = + mlx5e_mpwrq_umr_wqebbs(mdev, rq->mpwqe.page_shift, + rq->mpwqe.umr_mode); + rq->mpwqe.mtts_per_wqe = + mlx5e_mpwrq_mtts_per_wqe(mdev, rq->mpwqe.page_shift, + rq->mpwqe.umr_mode); + + pool_size = rq->mpwqe.pages_per_wqe << + mlx5e_mpwqe_get_log_rq_size(mdev, params, xsk); rq->mpwqe.log_stride_sz = mlx5e_mpwqe_get_log_stride_size(mdev, params, xsk); rq->mpwqe.num_strides = @@ -600,7 +727,6 @@ static int mlx5e_alloc_rq(struct mlx5e_params *params, err = mlx5e_create_rq_umr_mkey(mdev, rq); if (err) goto err_rq_drop_page; - rq->mkey_be = cpu_to_be32(rq->umr_mkey); err = mlx5e_rq_alloc_mpwqe_info(rq, node); if (err) @@ -608,7 +734,7 @@ static int mlx5e_alloc_rq(struct mlx5e_params *params, err = mlx5_rq_shampo_alloc(mdev, params, rqp, rq, &pool_size, node); if (err) - goto err_free_by_rq_type; + goto err_free_mpwqe_info; break; default: /* MLX5_WQ_TYPE_CYCLIC */ @@ -633,11 +759,9 @@ static int mlx5e_alloc_rq(struct mlx5e_params *params, goto err_rq_wq_destroy; } - err = mlx5e_init_di_list(rq, wq_sz, node); + err = mlx5e_init_au_list(rq, wq_sz, node); if (err) goto err_rq_frags; - - rq->mkey_be = cpu_to_be32(mdev->mlx5e_res.hw_objs.mkey); } if (xsk) { @@ -662,14 +786,14 @@ static int mlx5e_alloc_rq(struct mlx5e_params *params, if (IS_ERR(rq->page_pool)) { err = PTR_ERR(rq->page_pool); rq->page_pool = NULL; - goto err_free_shampo; + goto err_free_by_rq_type; } if (xdp_rxq_info_is_reg(&rq->xdp_rxq)) err = xdp_rxq_info_reg_mem_model(&rq->xdp_rxq, MEM_TYPE_PAGE_POOL, rq->page_pool); } if (err) - goto err_free_shampo; + goto err_destroy_page_pool; for (i = 0; i < wq_sz; i++) { if (rq->wq_type == MLX5_WQ_TYPE_LINKED_LIST_STRIDING_RQ) { @@ -677,13 +801,14 @@ static int mlx5e_alloc_rq(struct mlx5e_params *params, mlx5_wq_ll_get_wqe(&rq->mpwqe.wq, i); u32 byte_count = rq->mpwqe.num_strides << rq->mpwqe.log_stride_sz; - u64 dma_offset = mlx5e_get_mpwqe_offset(i); + u64 dma_offset = mul_u32_u32(i, rq->mpwqe.mtts_per_wqe) << + rq->mpwqe.page_shift; u16 headroom = test_bit(MLX5E_RQ_STATE_SHAMPO, &rq->state) ? 0 : rq->buff.headroom; wqe->data[0].addr = cpu_to_be64(dma_offset + headroom); wqe->data[0].byte_count = cpu_to_be32(byte_count); - wqe->data[0].lkey = rq->mkey_be; + wqe->data[0].lkey = rq->mpwqe.umr_mkey_be; } else { struct mlx5e_rx_wqe_cyc *wqe = mlx5_wq_cyc_get_wqe(&rq->wqe.wq, i); @@ -721,19 +846,21 @@ static int mlx5e_alloc_rq(struct mlx5e_params *params, return 0; -err_free_shampo: - mlx5e_rq_free_shampo(rq); +err_destroy_page_pool: + page_pool_destroy(rq->page_pool); err_free_by_rq_type: switch (rq->wq_type) { case MLX5_WQ_TYPE_LINKED_LIST_STRIDING_RQ: + mlx5e_rq_free_shampo(rq); +err_free_mpwqe_info: kvfree(rq->mpwqe.info); err_rq_mkey: - mlx5_core_destroy_mkey(mdev, rq->umr_mkey); + mlx5_core_destroy_mkey(mdev, be32_to_cpu(rq->mpwqe.umr_mkey_be)); err_rq_drop_page: mlx5e_free_mpwqe_rq_drop_page(rq); break; default: /* MLX5_WQ_TYPE_CYCLIC */ - mlx5e_free_di_list(rq); + mlx5e_free_au_list(rq); err_rq_frags: kvfree(rq->wqe.frags); } @@ -761,24 +888,22 @@ static void mlx5e_free_rq(struct mlx5e_rq *rq) switch (rq->wq_type) { case MLX5_WQ_TYPE_LINKED_LIST_STRIDING_RQ: kvfree(rq->mpwqe.info); - mlx5_core_destroy_mkey(rq->mdev, rq->umr_mkey); + mlx5_core_destroy_mkey(rq->mdev, be32_to_cpu(rq->mpwqe.umr_mkey_be)); mlx5e_free_mpwqe_rq_drop_page(rq); mlx5e_rq_free_shampo(rq); break; default: /* MLX5_WQ_TYPE_CYCLIC */ kvfree(rq->wqe.frags); - mlx5e_free_di_list(rq); + mlx5e_free_au_list(rq); } for (i = rq->page_cache.head; i != rq->page_cache.tail; i = (i + 1) & (MLX5E_CACHE_SIZE - 1)) { - struct mlx5e_dma_info *dma_info = &rq->page_cache.page_cache[i]; - /* With AF_XDP, page_cache is not used, so this loop is not * entered, and it's safe to call mlx5e_page_release_dynamic * directly. */ - mlx5e_page_release_dynamic(rq, dma_info->page, false); + mlx5e_page_release_dynamic(rq, rq->page_cache.page_cache[i], false); } xdp_rxq_info_unreg(&rq->xdp_rxq); @@ -833,7 +958,7 @@ int mlx5e_create_rq(struct mlx5e_rq *rq, struct mlx5e_rq_param *param) return err; } -int mlx5e_modify_rq_state(struct mlx5e_rq *rq, int curr_state, int next_state) +static int mlx5e_modify_rq_state(struct mlx5e_rq *rq, int curr_state, int next_state) { struct mlx5_core_dev *mdev = rq->mdev; @@ -862,6 +987,32 @@ int mlx5e_modify_rq_state(struct mlx5e_rq *rq, int curr_state, int next_state) return err; } +static int mlx5e_rq_to_ready(struct mlx5e_rq *rq, int curr_state) +{ + struct net_device *dev = rq->netdev; + int err; + + err = mlx5e_modify_rq_state(rq, curr_state, MLX5_RQC_STATE_RST); + if (err) { + netdev_err(dev, "Failed to move rq 0x%x to reset\n", rq->rqn); + return err; + } + err = mlx5e_modify_rq_state(rq, MLX5_RQC_STATE_RST, MLX5_RQC_STATE_RDY); + if (err) { + netdev_err(dev, "Failed to move rq 0x%x to ready\n", rq->rqn); + return err; + } + + return 0; +} + +int mlx5e_flush_rq(struct mlx5e_rq *rq, int curr_state) +{ + mlx5e_free_rx_descs(rq); + + return mlx5e_rq_to_ready(rq, curr_state); +} + static int mlx5e_modify_rq_scatter_fcs(struct mlx5e_rq *rq, bool enable) { struct mlx5_core_dev *mdev = rq->mdev; @@ -1154,9 +1305,9 @@ static int mlx5e_alloc_xdpsq(struct mlx5e_channel *c, is_redirect ? &c->priv->channel_stats[c->ix]->xdpsq : &c->priv->channel_stats[c->ix]->rq_xdpsq; - sq->max_sq_wqebbs = mlx5e_get_max_sq_wqebbs(mdev); - sq->stop_room = MLX5E_STOP_ROOM(sq->max_sq_wqebbs); - sq->max_sq_mpw_wqebbs = mlx5e_get_sw_max_sq_mpw_wqebbs(sq->max_sq_wqebbs); + sq->stop_room = param->is_mpw ? mlx5e_stop_room_for_mpwqe(mdev) : + mlx5e_stop_room_for_max_wqe(mdev); + sq->max_sq_mpw_wqebbs = mlx5e_get_max_sq_aligned_wqebbs(mdev); param->wq.db_numa_node = cpu_to_node(c->cpu); err = mlx5_wq_cyc_create(mdev, ¶m->wq, sqc_wq, wq, &sq->wq_ctrl); @@ -1231,7 +1382,6 @@ static int mlx5e_alloc_icosq(struct mlx5e_channel *c, sq->channel = c; sq->uar_map = mdev->mlx5e_res.hw_objs.bfreg.map; sq->reserved_room = param->stop_room; - sq->max_sq_wqebbs = mlx5e_get_max_sq_wqebbs(mdev); param->wq.db_numa_node = cpu_to_node(c->cpu); err = mlx5_wq_cyc_create(mdev, ¶m->wq, sqc_wq, wq, &sq->wq_ctrl); @@ -1317,8 +1467,7 @@ static int mlx5e_alloc_txqsq(struct mlx5e_channel *c, sq->uar_map = mdev->mlx5e_res.hw_objs.bfreg.map; sq->min_inline_mode = params->tx_min_inline_mode; sq->hw_mtu = MLX5E_SW2HW_MTU(params, params->sw_mtu); - sq->max_sq_wqebbs = mlx5e_get_max_sq_wqebbs(mdev); - sq->max_sq_mpw_wqebbs = mlx5e_get_sw_max_sq_mpw_wqebbs(sq->max_sq_wqebbs); + sq->max_sq_mpw_wqebbs = mlx5e_get_max_sq_aligned_wqebbs(mdev); INIT_WORK(&sq->recover_work, mlx5e_tx_err_cqe_work); if (!MLX5_CAP_ETH(mdev, wqe_vlan_insert)) set_bit(MLX5E_SQ_STATE_VLAN_NEED_L2_INLINE, &sq->state); @@ -2280,7 +2429,7 @@ static int mlx5e_open_channel(struct mlx5e_priv *priv, int ix, c->aff_mask = irq_get_effective_affinity_mask(irq); c->lag_port = mlx5e_enumerate_lag_port(priv->mdev, ix); - netif_napi_add(netdev, &c->napi, mlx5e_napi_poll, 64); + netif_napi_add(netdev, &c->napi, mlx5e_napi_poll); err = mlx5e_open_queues(c, params, cparam); if (unlikely(err)) @@ -2318,10 +2467,11 @@ static void mlx5e_activate_channel(struct mlx5e_channel *c) mlx5e_activate_txqsq(&c->sq[tc]); mlx5e_activate_icosq(&c->icosq); mlx5e_activate_icosq(&c->async_icosq); - mlx5e_activate_rq(&c->rq); if (test_bit(MLX5E_CHANNEL_STATE_XSK, c->state)) mlx5e_activate_xsk(c); + else + mlx5e_activate_rq(&c->rq); mlx5e_trigger_napi_icosq(c); } @@ -2332,8 +2482,9 @@ static void mlx5e_deactivate_channel(struct mlx5e_channel *c) if (test_bit(MLX5E_CHANNEL_STATE_XSK, c->state)) mlx5e_deactivate_xsk(c); + else + mlx5e_deactivate_rq(&c->rq); - mlx5e_deactivate_rq(&c->rq); mlx5e_deactivate_icosq(&c->async_icosq); mlx5e_deactivate_icosq(&c->icosq); for (tc = 0; tc < c->num_tc; tc++) @@ -2425,8 +2576,6 @@ static void mlx5e_activate_channels(struct mlx5e_channels *chs) mlx5e_ptp_activate_channel(chs->ptp); } -#define MLX5E_RQ_WQES_TIMEOUT 20000 /* msecs */ - static int mlx5e_wait_channels_min_rx_wqes(struct mlx5e_channels *chs) { int err = 0; @@ -2434,8 +2583,12 @@ static int mlx5e_wait_channels_min_rx_wqes(struct mlx5e_channels *chs) for (i = 0; i < chs->num; i++) { int timeout = err ? 0 : MLX5E_RQ_WQES_TIMEOUT; + struct mlx5e_channel *c = chs->c[i]; - err |= mlx5e_wait_for_min_rx_wqes(&chs->c[i]->rq, timeout); + if (test_bit(MLX5E_CHANNEL_STATE_XSK, c->state)) + continue; + + err |= mlx5e_wait_for_min_rx_wqes(&c->rq, timeout); /* Don't wait on the XSK RQ, because the newer xdpsock sample * doesn't provide any Fill Ring entries at the setup stage. @@ -2600,7 +2753,7 @@ static int mlx5e_update_netdev_queues(struct mlx5e_priv *priv) struct netdev_tc_txq old_tc_to_txq[TC_MAX_QUEUE], *tc_to_txq; struct net_device *netdev = priv->netdev; int old_num_txqs, old_ntc; - int num_rxqs, nch, ntc; + int nch, ntc; int err; int i; @@ -2611,7 +2764,6 @@ static int mlx5e_update_netdev_queues(struct mlx5e_priv *priv) nch = priv->channels.params.num_channels; ntc = priv->channels.params.mqprio.num_tc; - num_rxqs = nch * priv->profile->rq_groups; tc_to_txq = priv->channels.params.mqprio.tc_to_txq; err = mlx5e_netdev_set_tcs(netdev, nch, ntc, tc_to_txq); @@ -2620,7 +2772,7 @@ static int mlx5e_update_netdev_queues(struct mlx5e_priv *priv) err = mlx5e_update_tx_netdev_queues(priv); if (err) goto err_tcs; - err = netif_set_real_num_rx_queues(netdev, num_rxqs); + err = netif_set_real_num_rx_queues(netdev, nch); if (err) { netdev_warn(netdev, "netif_set_real_num_rx_queues failed, %d\n", err); goto err_txqs; @@ -2738,7 +2890,7 @@ void mlx5e_activate_priv_channels(struct mlx5e_priv *priv) netif_tx_start_all_queues(priv->netdev); if (mlx5e_is_vport_rep(priv)) - mlx5e_add_sqs_fwd_rules(priv); + mlx5e_rep_activate_channels(priv); mlx5e_wait_channels_min_rx_wqes(&priv->channels); @@ -2752,7 +2904,7 @@ void mlx5e_deactivate_priv_channels(struct mlx5e_priv *priv) mlx5e_rx_res_channels_deactivate(priv->rx_res); if (mlx5e_is_vport_rep(priv)) - mlx5e_remove_sqs_fwd_rules(priv); + mlx5e_rep_deactivate_channels(priv); /* The results of ndo_select_queue are unreliable, while netdev config * is being changed (real_num_tx_queues, num_tc). Stop all queues to @@ -3547,7 +3699,8 @@ mlx5e_get_stats(struct net_device *dev, struct rtnl_link_stats64 *stats) stats->rx_length_errors = PPORT_802_3_GET(pstats, a_in_range_length_errors) + PPORT_802_3_GET(pstats, a_out_of_range_length_field) + - PPORT_802_3_GET(pstats, a_frame_too_long_errors); + PPORT_802_3_GET(pstats, a_frame_too_long_errors) + + VNIC_ENV_GET(&priv->stats.vnic, eth_wqe_too_small); stats->rx_crc_errors = PPORT_802_3_GET(pstats, a_frame_check_sequence_errors); stats->rx_frame_errors = PPORT_802_3_GET(pstats, a_alignment_errors); @@ -3669,9 +3822,11 @@ static int set_feature_cvlan_filter(struct net_device *netdev, bool enable) struct mlx5e_priv *priv = netdev_priv(netdev); if (enable) - mlx5e_enable_cvlan_filter(priv); + mlx5e_enable_cvlan_filter(priv->fs, + !!(priv->netdev->flags & IFF_PROMISC)); else - mlx5e_disable_cvlan_filter(priv); + mlx5e_disable_cvlan_filter(priv->fs, + !!(priv->netdev->flags & IFF_PROMISC)); return 0; } @@ -3780,7 +3935,7 @@ static int set_feature_rx_vlan(struct net_device *netdev, bool enable) mutex_lock(&priv->state_lock); - priv->fs->vlan_strip_disable = !enable; + mlx5e_fs_set_vlan_strip_disable(priv->fs, !enable); priv->channels.params.vlan_strip_disable = !enable; if (!test_bit(MLX5E_STATE_OPENED, &priv->state)) @@ -3788,7 +3943,7 @@ static int set_feature_rx_vlan(struct net_device *netdev, bool enable) err = mlx5e_modify_channels_vsd(&priv->channels, !enable); if (err) { - priv->fs->vlan_strip_disable = enable; + mlx5e_fs_set_vlan_strip_disable(priv->fs, enable); priv->channels.params.vlan_strip_disable = enable; } unlock: @@ -3826,9 +3981,9 @@ static int set_feature_arfs(struct net_device *netdev, bool enable) int err; if (enable) - err = mlx5e_arfs_enable(priv); + err = mlx5e_arfs_enable(priv->fs); else - err = mlx5e_arfs_disable(priv); + err = mlx5e_arfs_disable(priv->fs); return err; } @@ -3912,12 +4067,14 @@ static netdev_features_t mlx5e_fix_features(struct net_device *netdev, netdev_features_t features) { struct mlx5e_priv *priv = netdev_priv(netdev); + struct mlx5e_vlan_table *vlan; struct mlx5e_params *params; + vlan = mlx5e_fs_get_vlan(priv->fs); mutex_lock(&priv->state_lock); params = &priv->channels.params; - if (!priv->fs->vlan || - !bitmap_empty(mlx5e_vlan_get_active_svlans(priv->fs->vlan), VLAN_N_VID)) { + if (!vlan || + !bitmap_empty(mlx5e_vlan_get_active_svlans(vlan), VLAN_N_VID)) { /* HW strips the outer C-tag header, this is a problem * for S-tag traffic. */ @@ -4006,7 +4163,7 @@ static bool mlx5e_xsk_validate_mtu(struct net_device *netdev, * 2. Size of SKBs allocated on XDP_PASS <= PAGE_SIZE. */ max_mtu_frame = MLX5E_HW2SW_MTU(new_params, xsk.chunk_size - hr); - max_mtu_page = mlx5e_xdp_max_mtu(new_params, &xsk); + max_mtu_page = MLX5E_HW2SW_MTU(new_params, SKB_MAX_HEAD(0)); max_mtu = min(max_mtu_frame, max_mtu_page); netdev_err(netdev, "MTU %d is too big for an XSK running on channel %u. Try MTU <= %d\n", @@ -4018,14 +4175,16 @@ static bool mlx5e_xsk_validate_mtu(struct net_device *netdev, return true; } -static bool mlx5e_params_validate_xdp(struct net_device *netdev, struct mlx5e_params *params) +static bool mlx5e_params_validate_xdp(struct net_device *netdev, + struct mlx5_core_dev *mdev, + struct mlx5e_params *params) { bool is_linear; /* No XSK params: AF_XDP can't be enabled yet at the point of setting * the XDP program. */ - is_linear = mlx5e_rx_is_linear_skb(params, NULL); + is_linear = mlx5e_rx_is_linear_skb(mdev, params, NULL); if (!is_linear && params->rq_wq_type != MLX5_WQ_TYPE_CYCLIC) { netdev_warn(netdev, "XDP is not allowed with striding RQ and MTU(%d) > %d\n", @@ -4062,7 +4221,8 @@ int mlx5e_change_mtu(struct net_device *netdev, int new_mtu, if (err) goto out; - if (new_params.xdp_prog && !mlx5e_params_validate_xdp(netdev, &new_params)) { + if (new_params.xdp_prog && !mlx5e_params_validate_xdp(netdev, priv->mdev, + &new_params)) { err = -EINVAL; goto out; } @@ -4077,19 +4237,21 @@ int mlx5e_change_mtu(struct net_device *netdev, int new_mtu, if (params->packet_merge.type == MLX5E_PACKET_MERGE_LRO) reset = false; - if (params->rq_wq_type == MLX5_WQ_TYPE_LINKED_LIST_STRIDING_RQ) { + if (params->rq_wq_type == MLX5_WQ_TYPE_LINKED_LIST_STRIDING_RQ && + params->packet_merge.type != MLX5E_PACKET_MERGE_SHAMPO) { bool is_linear_old = mlx5e_rx_mpwqe_is_linear_skb(priv->mdev, params, NULL); bool is_linear_new = mlx5e_rx_mpwqe_is_linear_skb(priv->mdev, &new_params, NULL); - u8 ppw_old = mlx5e_mpwqe_log_pkts_per_wqe(params, NULL); - u8 ppw_new = mlx5e_mpwqe_log_pkts_per_wqe(&new_params, NULL); + u8 sz_old = mlx5e_mpwqe_get_log_rq_size(priv->mdev, params, NULL); + u8 sz_new = mlx5e_mpwqe_get_log_rq_size(priv->mdev, &new_params, NULL); /* Always reset in linear mode - hw_mtu is used in data path. * Check that the mode was non-linear and didn't change. * If XSK is active, XSK RQs are linear. + * Reset if the RQ size changed, even if it's non-linear. */ if (!is_linear_old && !is_linear_new && !priv->xsk.refcnt && - ppw_old == ppw_new) + sz_old == sz_new) reset = false; } @@ -4539,7 +4701,7 @@ static int mlx5e_xdp_allowed(struct mlx5e_priv *priv, struct bpf_prog *prog) new_params = priv->channels.params; new_params.xdp_prog = prog; - if (!mlx5e_params_validate_xdp(netdev, &new_params)) + if (!mlx5e_params_validate_xdp(netdev, priv->mdev, &new_params)) return -EINVAL; return 0; @@ -4577,8 +4739,20 @@ static int mlx5e_xdp_set(struct net_device *netdev, struct bpf_prog *prog) new_params = priv->channels.params; new_params.xdp_prog = prog; - if (reset) - mlx5e_set_rq_type(priv->mdev, &new_params); + + /* XDP affects striding RQ parameters. Block XDP if striding RQ won't be + * supported with the new parameters: if PAGE_SIZE is bigger than + * MLX5_MPWQE_LOG_STRIDE_SZ_MAX, striding RQ can't be used, even though + * the MTU is small enough for the linear mode, because XDP uses strides + * of PAGE_SIZE on regular RQs. + */ + if (reset && MLX5E_GET_PFLAG(&new_params, MLX5E_PFLAG_RX_STRIDING_RQ)) { + /* Checking for regular RQs here; XSK RQs were checked on XSK bind. */ + err = mlx5e_mpwrq_validate_regular(priv->mdev, &new_params); + if (err) + goto unlock; + } + old_prog = priv->channels.params.xdp_prog; err = mlx5e_safe_switch_params(priv, &new_params, NULL, NULL, reset); @@ -4898,7 +5072,8 @@ static void mlx5e_build_nic_netdev(struct net_device *netdev) if (!!MLX5_CAP_ETH(mdev, lro_cap) && !MLX5_CAP_ETH(mdev, tunnel_lro_vxlan) && !MLX5_CAP_ETH(mdev, tunnel_lro_gre) && - mlx5e_check_fragmented_striding_rq_cap(mdev)) + mlx5e_check_fragmented_striding_rq_cap(mdev, PAGE_SHIFT, + MLX5E_MPWRQ_UMR_MODE_ALIGNED)) netdev->vlan_features |= NETIF_F_LRO; netdev->hw_features = netdev->vlan_features; @@ -4986,6 +5161,7 @@ static void mlx5e_build_nic_netdev(struct net_device *netdev) netif_set_tso_max_size(netdev, GSO_MAX_SIZE); mlx5e_set_netdev_dev_addr(netdev); + mlx5e_macsec_build_netdev(priv); mlx5e_ipsec_build_netdev(priv); mlx5e_ktls_build_netdev(priv); } @@ -5087,7 +5263,7 @@ static int mlx5e_init_nic_rx(struct mlx5e_priv *priv) goto err_destroy_q_counters; } - features = MLX5E_RX_RES_FEATURE_XSK | MLX5E_RX_RES_FEATURE_PTP; + features = MLX5E_RX_RES_FEATURE_PTP; if (priv->channels.params.tunneled_offload_en) features |= MLX5E_RX_RES_FEATURE_INNER_FT; err = mlx5e_rx_res_init(priv->rx_res, priv->mdev, features, @@ -5097,7 +5273,8 @@ static int mlx5e_init_nic_rx(struct mlx5e_priv *priv) if (err) goto err_close_drop_rq; - err = mlx5e_create_flow_steering(priv); + err = mlx5e_create_flow_steering(priv->fs, priv->rx_res, priv->profile, + priv->netdev); if (err) { mlx5_core_warn(mdev, "create flow steering failed, %d\n", err); goto err_destroy_rx_res; @@ -5120,7 +5297,8 @@ static int mlx5e_init_nic_rx(struct mlx5e_priv *priv) err_tc_nic_cleanup: mlx5e_tc_nic_cleanup(priv); err_destroy_flow_steering: - mlx5e_destroy_flow_steering(priv); + mlx5e_destroy_flow_steering(priv->fs, !!(priv->netdev->hw_features & NETIF_F_NTUPLE), + priv->profile); err_destroy_rx_res: mlx5e_rx_res_destroy(priv->rx_res); err_close_drop_rq: @@ -5136,7 +5314,8 @@ static void mlx5e_cleanup_nic_rx(struct mlx5e_priv *priv) { mlx5e_accel_cleanup_rx(priv); mlx5e_tc_nic_cleanup(priv); - mlx5e_destroy_flow_steering(priv); + mlx5e_destroy_flow_steering(priv->fs, !!(priv->netdev->hw_features & NETIF_F_NTUPLE), + priv->profile); mlx5e_rx_res_destroy(priv->rx_res); mlx5e_close_drop_rq(&priv->drop_rq); mlx5e_destroy_q_counters(priv); @@ -5188,9 +5367,14 @@ static void mlx5e_nic_enable(struct mlx5e_priv *priv) { struct net_device *netdev = priv->netdev; struct mlx5_core_dev *mdev = priv->mdev; + int err; mlx5e_fs_init_l2_addr(priv->fs, netdev); + err = mlx5e_macsec_init(priv); + if (err) + mlx5_core_err(mdev, "MACsec initialization failed, %d\n", err); + /* Marking the link as currently not needed by the Driver */ if (!netif_running(netdev)) mlx5e_modify_admin_state(mdev, MLX5_PORT_DOWN); @@ -5248,6 +5432,7 @@ static void mlx5e_nic_disable(struct mlx5e_priv *priv) mlx5e_disable_async_events(priv); mlx5_lag_remove_netdev(mdev, priv->netdev); mlx5_vxlan_reset_to_default(mdev->vxlan); + mlx5e_macsec_cleanup(priv); } int mlx5e_update_nic_rx(struct mlx5e_priv *priv) @@ -5269,7 +5454,6 @@ static const struct mlx5e_profile mlx5e_nic_profile = { .update_carrier = mlx5e_update_carrier, .rx_handlers = &mlx5e_rx_handlers_nic, .max_tc = MLX5E_MAX_NUM_TC, - .rq_groups = MLX5E_NUM_RQ_GROUPS(XSK), .stats_grps = mlx5e_nic_stats_grps, .stats_grps_num = mlx5e_nic_stats_grps_num, .features = BIT(MLX5E_PROFILE_FEATURE_PTP_RX) | @@ -5302,8 +5486,7 @@ mlx5e_calc_max_nch(struct mlx5_core_dev *mdev, struct net_device *netdev, max_nch = mlx5e_profile_max_num_channels(mdev, profile); /* netdev rx queues */ - tmp = netdev->num_rx_queues / max_t(u8, profile->rq_groups, 1); - max_nch = min_t(unsigned int, max_nch, tmp); + max_nch = min_t(unsigned int, max_nch, netdev->num_rx_queues); /* netdev tx queues */ tmp = netdev->num_tx_queues; @@ -5447,11 +5630,7 @@ static unsigned int mlx5e_get_max_num_txqs(struct mlx5_core_dev *mdev, static unsigned int mlx5e_get_max_num_rxqs(struct mlx5_core_dev *mdev, const struct mlx5e_profile *profile) { - unsigned int nch; - - nch = mlx5e_profile_max_num_channels(mdev, profile); - - return nch * profile->rq_groups; + return mlx5e_profile_max_num_channels(mdev, profile); } struct net_device * @@ -5512,7 +5691,8 @@ int mlx5e_attach_netdev(struct mlx5e_priv *priv) clear_bit(MLX5E_STATE_DESTROYING, &priv->state); if (priv->fs) - priv->fs->state_destroy = !test_bit(MLX5E_STATE_DESTROYING, &priv->state); + mlx5e_fs_set_state_destroy(priv->fs, + !test_bit(MLX5E_STATE_DESTROYING, &priv->state)); /* max number of channels may have changed */ max_nch = mlx5e_calc_max_nch(priv->mdev, priv->netdev, profile); @@ -5573,7 +5753,8 @@ out: mlx5e_reset_channels(priv->netdev); set_bit(MLX5E_STATE_DESTROYING, &priv->state); if (priv->fs) - priv->fs->state_destroy = !test_bit(MLX5E_STATE_DESTROYING, &priv->state); + mlx5e_fs_set_state_destroy(priv->fs, + !test_bit(MLX5E_STATE_DESTROYING, &priv->state)); cancel_work_sync(&priv->update_stats_work); return err; } @@ -5584,7 +5765,8 @@ void mlx5e_detach_netdev(struct mlx5e_priv *priv) set_bit(MLX5E_STATE_DESTROYING, &priv->state); if (priv->fs) - priv->fs->state_destroy = !test_bit(MLX5E_STATE_DESTROYING, &priv->state); + mlx5e_fs_set_state_destroy(priv->fs, + !test_bit(MLX5E_STATE_DESTROYING, &priv->state)); if (profile->disable) profile->disable(priv); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c index 759f7d3c2cfd..794cd8dfe9c9 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c @@ -56,6 +56,7 @@ #include "en_accel/ipsec.h" #include "en/tc/int_port.h" #include "en/ptp.h" +#include "en/fs_ethtool.h" #define MLX5E_REP_PARAMS_DEF_LOG_SQ_SIZE \ max(0x7, MLX5E_PARAMS_MINIMUM_LOG_SQ_SIZE) @@ -69,7 +70,7 @@ static void mlx5e_rep_get_drvinfo(struct net_device *dev, struct mlx5e_priv *priv = netdev_priv(dev); struct mlx5_core_dev *mdev = priv->mdev; - strlcpy(drvinfo->driver, mlx5e_rep_driver_name, + strscpy(drvinfo->driver, mlx5e_rep_driver_name, sizeof(drvinfo->driver)); snprintf(drvinfo->fw_version, sizeof(drvinfo->fw_version), "%d.%d.%04d (%.16s)", @@ -397,7 +398,8 @@ out_err: return err; } -int mlx5e_add_sqs_fwd_rules(struct mlx5e_priv *priv) +static int +mlx5e_add_sqs_fwd_rules(struct mlx5e_priv *priv) { int sqs_per_channel = mlx5e_get_dcb_num_tc(&priv->channels.params); struct mlx5_eswitch *esw = priv->mdev->priv.eswitch; @@ -451,7 +453,8 @@ out: return err; } -void mlx5e_remove_sqs_fwd_rules(struct mlx5e_priv *priv) +static void +mlx5e_remove_sqs_fwd_rules(struct mlx5e_priv *priv) { struct mlx5_eswitch *esw = priv->mdev->priv.eswitch; struct mlx5e_rep_priv *rpriv = priv->ppriv; @@ -460,6 +463,49 @@ void mlx5e_remove_sqs_fwd_rules(struct mlx5e_priv *priv) mlx5e_sqs2vport_stop(esw, rep); } +static int +mlx5e_rep_add_meta_tunnel_rule(struct mlx5e_priv *priv) +{ + struct mlx5_eswitch *esw = priv->mdev->priv.eswitch; + struct mlx5e_rep_priv *rpriv = priv->ppriv; + struct mlx5_eswitch_rep *rep = rpriv->rep; + struct mlx5_flow_handle *flow_rule; + struct mlx5_flow_group *g; + + g = esw->fdb_table.offloads.send_to_vport_meta_grp; + if (!g) + return 0; + + flow_rule = mlx5_eswitch_add_send_to_vport_meta_rule(esw, rep->vport); + if (IS_ERR(flow_rule)) + return PTR_ERR(flow_rule); + + rpriv->send_to_vport_meta_rule = flow_rule; + + return 0; +} + +static void +mlx5e_rep_del_meta_tunnel_rule(struct mlx5e_priv *priv) +{ + struct mlx5e_rep_priv *rpriv = priv->ppriv; + + if (rpriv->send_to_vport_meta_rule) + mlx5_eswitch_del_send_to_vport_meta_rule(rpriv->send_to_vport_meta_rule); +} + +void mlx5e_rep_activate_channels(struct mlx5e_priv *priv) +{ + mlx5e_add_sqs_fwd_rules(priv); + mlx5e_rep_add_meta_tunnel_rule(priv); +} + +void mlx5e_rep_deactivate_channels(struct mlx5e_priv *priv) +{ + mlx5e_rep_del_meta_tunnel_rule(priv); + mlx5e_remove_sqs_fwd_rules(priv); +} + static int mlx5e_rep_open(struct net_device *dev) { struct mlx5e_priv *priv = netdev_priv(dev); @@ -747,19 +793,20 @@ static int mlx5e_create_rep_ttc_table(struct mlx5e_priv *priv) struct ttc_params ttc_params = {}; int err; - priv->fs->ns = mlx5_get_flow_namespace(priv->mdev, - MLX5_FLOW_NAMESPACE_KERNEL); + mlx5e_fs_set_ns(priv->fs, + mlx5_get_flow_namespace(priv->mdev, + MLX5_FLOW_NAMESPACE_KERNEL), false); /* The inner_ttc in the ttc params is intentionally not set */ - mlx5e_set_ttc_params(priv, &ttc_params, false); + mlx5e_set_ttc_params(priv->fs, priv->rx_res, &ttc_params, false); if (rep->vport != MLX5_VPORT_UPLINK) /* To give uplik rep TTC a lower level for chaining from root ft */ ttc_params.ft_attr.level = MLX5E_TTC_FT_LEVEL + 1; - priv->fs->ttc = mlx5_create_ttc_table(priv->mdev, &ttc_params); - if (IS_ERR(priv->fs->ttc)) { - err = PTR_ERR(priv->fs->ttc); + mlx5e_fs_set_ttc(priv->fs, mlx5_create_ttc_table(priv->mdev, &ttc_params), false); + if (IS_ERR(mlx5e_fs_get_ttc(priv->fs, false))) { + err = PTR_ERR(mlx5e_fs_get_ttc(priv->fs, false)); netdev_err(priv->netdev, "Failed to create rep ttc table, err=%d\n", err); return err; @@ -779,7 +826,7 @@ static int mlx5e_create_rep_root_ft(struct mlx5e_priv *priv) /* non uplik reps will skip any bypass tables and go directly to * their own ttc */ - rpriv->root_ft = mlx5_get_ttc_flow_table(priv->fs->ttc); + rpriv->root_ft = mlx5_get_ttc_flow_table(mlx5e_fs_get_ttc(priv->fs, false)); return 0; } @@ -887,14 +934,14 @@ static int mlx5e_init_rep_rx(struct mlx5e_priv *priv) if (err) goto err_destroy_root_ft; - mlx5e_ethtool_init_steering(priv); + mlx5e_ethtool_init_steering(priv->fs); return 0; err_destroy_root_ft: mlx5e_destroy_rep_root_ft(priv); err_destroy_ttc_table: - mlx5_destroy_ttc_table(priv->fs->ttc); + mlx5_destroy_ttc_table(mlx5e_fs_get_ttc(priv->fs, false)); err_destroy_rx_res: mlx5e_rx_res_destroy(priv->rx_res); err_close_drop_rq: @@ -908,10 +955,10 @@ err_free_fs: static void mlx5e_cleanup_rep_rx(struct mlx5e_priv *priv) { - mlx5e_ethtool_cleanup_steering(priv); + mlx5e_ethtool_cleanup_steering(priv->fs); rep_vport_rx_rule_destroy(priv); mlx5e_destroy_rep_root_ft(priv); - mlx5_destroy_ttc_table(priv->fs->ttc); + mlx5_destroy_ttc_table(mlx5e_fs_get_ttc(priv->fs, false)); mlx5e_rx_res_destroy(priv->rx_res); mlx5e_close_drop_rq(&priv->drop_rq); mlx5e_rx_res_free(priv->rx_res); @@ -1177,7 +1224,6 @@ static const struct mlx5e_profile mlx5e_rep_profile = { .update_stats = mlx5e_stats_update_ndo_stats, .rx_handlers = &mlx5e_rx_handlers_rep, .max_tc = 1, - .rq_groups = MLX5E_NUM_RQ_GROUPS(REGULAR), .stats_grps = mlx5e_rep_stats_grps, .stats_grps_num = mlx5e_rep_stats_grps_num, .max_nch_limit = mlx5e_rep_max_nch_limit, @@ -1197,8 +1243,6 @@ static const struct mlx5e_profile mlx5e_uplink_rep_profile = { .update_carrier = mlx5e_update_carrier, .rx_handlers = &mlx5e_rx_handlers_rep, .max_tc = MLX5E_MAX_NUM_TC, - /* XSK is needed so we can replace profile with NIC netdev */ - .rq_groups = MLX5E_NUM_RQ_GROUPS(XSK), .stats_grps = mlx5e_ul_rep_stats_grps, .stats_grps_num = mlx5e_ul_rep_stats_grps_num, }; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h index dec183ccd4ac..b4e691760da9 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h @@ -111,6 +111,7 @@ struct mlx5e_rep_priv { struct list_head vport_sqs_list; struct mlx5_rep_uplink_priv uplink_priv; /* valid for uplink rep */ struct rtnl_link_stats64 prev_vf_vport_stats; + struct mlx5_flow_handle *send_to_vport_meta_rule; struct rhashtable tc_ht; }; @@ -241,8 +242,8 @@ int mlx5e_rep_get_offload_stats(int attr_id, const struct net_device *dev, void *sp); bool mlx5e_is_uplink_rep(struct mlx5e_priv *priv); -int mlx5e_add_sqs_fwd_rules(struct mlx5e_priv *priv); -void mlx5e_remove_sqs_fwd_rules(struct mlx5e_priv *priv); +void mlx5e_rep_activate_channels(struct mlx5e_priv *priv); +void mlx5e_rep_deactivate_channels(struct mlx5e_priv *priv); void mlx5e_rep_queue_neigh_stats_work(struct mlx5e_priv *priv); @@ -256,8 +257,8 @@ static inline bool mlx5e_eswitch_rep(const struct net_device *netdev) #else /* CONFIG_MLX5_ESWITCH */ static inline bool mlx5e_is_uplink_rep(struct mlx5e_priv *priv) { return false; } -static inline int mlx5e_add_sqs_fwd_rules(struct mlx5e_priv *priv) { return 0; } -static inline void mlx5e_remove_sqs_fwd_rules(struct mlx5e_priv *priv) {} +static inline void mlx5e_rep_activate_channels(struct mlx5e_priv *priv) {} +static inline void mlx5e_rep_deactivate_channels(struct mlx5e_priv *priv) {} static inline int mlx5e_rep_init(void) { return 0; }; static inline void mlx5e_rep_cleanup(void) {}; static inline bool mlx5e_rep_has_offload_stats(const struct net_device *dev, diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c index 24de37b79f5a..58084650151f 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c @@ -41,6 +41,7 @@ #include <net/gro.h> #include <net/udp.h> #include <net/tcp.h> +#include <net/xdp_sock_drv.h> #include "en.h" #include "en/txrx.h" #include "en_tc.h" @@ -49,6 +50,7 @@ #include "en/rep/tc.h" #include "ipoib/ipoib.h" #include "en_accel/ipsec.h" +#include "en_accel/macsec.h" #include "en_accel/ipsec_rxtx.h" #include "en_accel/ktls_txrx.h" #include "en/xdp.h" @@ -237,69 +239,61 @@ static inline bool mlx5e_rx_cache_put(struct mlx5e_rq *rq, struct page *page) return false; } - cache->page_cache[cache->tail].page = page; - cache->page_cache[cache->tail].addr = page_pool_get_dma_addr(page); + cache->page_cache[cache->tail] = page; cache->tail = tail_next; return true; } -static inline bool mlx5e_rx_cache_get(struct mlx5e_rq *rq, - struct mlx5e_dma_info *dma_info) +static inline bool mlx5e_rx_cache_get(struct mlx5e_rq *rq, union mlx5e_alloc_unit *au) { struct mlx5e_page_cache *cache = &rq->page_cache; struct mlx5e_rq_stats *stats = rq->stats; + dma_addr_t addr; if (unlikely(cache->head == cache->tail)) { stats->cache_empty++; return false; } - if (page_ref_count(cache->page_cache[cache->head].page) != 1) { + if (page_ref_count(cache->page_cache[cache->head]) != 1) { stats->cache_busy++; return false; } - *dma_info = cache->page_cache[cache->head]; + au->page = cache->page_cache[cache->head]; cache->head = (cache->head + 1) & (MLX5E_CACHE_SIZE - 1); stats->cache_reuse++; - dma_sync_single_for_device(rq->pdev, dma_info->addr, - PAGE_SIZE, - DMA_FROM_DEVICE); + addr = page_pool_get_dma_addr(au->page); + /* Non-XSK always uses PAGE_SIZE. */ + dma_sync_single_for_device(rq->pdev, addr, PAGE_SIZE, DMA_FROM_DEVICE); return true; } -static inline int mlx5e_page_alloc_pool(struct mlx5e_rq *rq, - struct mlx5e_dma_info *dma_info) +static inline int mlx5e_page_alloc_pool(struct mlx5e_rq *rq, union mlx5e_alloc_unit *au) { - if (mlx5e_rx_cache_get(rq, dma_info)) + dma_addr_t addr; + + if (mlx5e_rx_cache_get(rq, au)) return 0; - dma_info->page = page_pool_dev_alloc_pages(rq->page_pool); - if (unlikely(!dma_info->page)) + au->page = page_pool_dev_alloc_pages(rq->page_pool); + if (unlikely(!au->page)) return -ENOMEM; - dma_info->addr = dma_map_page_attrs(rq->pdev, dma_info->page, 0, PAGE_SIZE, - rq->buff.map_dir, DMA_ATTR_SKIP_CPU_SYNC); - if (unlikely(dma_mapping_error(rq->pdev, dma_info->addr))) { - page_pool_recycle_direct(rq->page_pool, dma_info->page); - dma_info->page = NULL; + /* Non-XSK always uses PAGE_SIZE. */ + addr = dma_map_page_attrs(rq->pdev, au->page, 0, PAGE_SIZE, + rq->buff.map_dir, DMA_ATTR_SKIP_CPU_SYNC); + if (unlikely(dma_mapping_error(rq->pdev, addr))) { + page_pool_recycle_direct(rq->page_pool, au->page); + au->page = NULL; return -ENOMEM; } - page_pool_set_dma_addr(dma_info->page, dma_info->addr); + page_pool_set_dma_addr(au->page, addr); return 0; } -static inline int mlx5e_page_alloc(struct mlx5e_rq *rq, - struct mlx5e_dma_info *dma_info) -{ - if (rq->xsk_pool) - return mlx5e_xsk_page_alloc_pool(rq, dma_info); - else - return mlx5e_page_alloc_pool(rq, dma_info); -} - void mlx5e_page_dma_unmap(struct mlx5e_rq *rq, struct page *page) { dma_addr_t dma_addr = page_pool_get_dma_addr(page); @@ -324,32 +318,18 @@ void mlx5e_page_release_dynamic(struct mlx5e_rq *rq, struct page *page, bool rec } } -static inline void mlx5e_page_release(struct mlx5e_rq *rq, - struct mlx5e_dma_info *dma_info, - bool recycle) -{ - if (rq->xsk_pool) - /* The `recycle` parameter is ignored, and the page is always - * put into the Reuse Ring, because there is no way to return - * the page to the userspace when the interface goes down. - */ - xsk_buff_free(dma_info->xsk); - else - mlx5e_page_release_dynamic(rq, dma_info->page, recycle); -} - static inline int mlx5e_get_rx_frag(struct mlx5e_rq *rq, struct mlx5e_wqe_frag_info *frag) { int err = 0; if (!frag->offset) - /* On first frag (offset == 0), replenish page (dma_info actually). - * Other frags that point to the same dma_info (with a different + /* On first frag (offset == 0), replenish page (alloc_unit actually). + * Other frags that point to the same alloc_unit (with a different * offset) should just use the new one without replenishing again * by themselves. */ - err = mlx5e_page_alloc(rq, frag->di); + err = mlx5e_page_alloc_pool(rq, frag->au); return err; } @@ -359,7 +339,7 @@ static inline void mlx5e_put_rx_frag(struct mlx5e_rq *rq, bool recycle) { if (frag->last_in_page) - mlx5e_page_release(rq, frag->di, recycle); + mlx5e_page_release_dynamic(rq, frag->au->page, recycle); } static inline struct mlx5e_wqe_frag_info *get_frag(struct mlx5e_rq *rq, u16 ix) @@ -375,6 +355,7 @@ static int mlx5e_alloc_rx_wqe(struct mlx5e_rq *rq, struct mlx5e_rx_wqe_cyc *wqe, int i; for (i = 0; i < rq->wqe.info.num_frags; i++, frag++) { + dma_addr_t addr; u16 headroom; err = mlx5e_get_rx_frag(rq, frag); @@ -382,8 +363,8 @@ static int mlx5e_alloc_rx_wqe(struct mlx5e_rq *rq, struct mlx5e_rx_wqe_cyc *wqe, goto free_frags; headroom = i == 0 ? rq->buff.headroom : 0; - wqe->data[i].addr = cpu_to_be64(frag->di->addr + - frag->offset + headroom); + addr = page_pool_get_dma_addr(frag->au->page); + wqe->data[i].addr = cpu_to_be64(addr + frag->offset + headroom); } return 0; @@ -401,6 +382,15 @@ static inline void mlx5e_free_rx_wqe(struct mlx5e_rq *rq, { int i; + if (rq->xsk_pool) { + /* The `recycle` parameter is ignored, and the page is always + * put into the Reuse Ring, because there is no way to return + * the page to the userspace when the interface goes down. + */ + xsk_buff_free(wi->au->xsk); + return; + } + for (i = 0; i < rq->wqe.info.num_frags; i++, wi++) mlx5e_put_rx_frag(rq, wi, recycle); } @@ -412,84 +402,76 @@ static void mlx5e_dealloc_rx_wqe(struct mlx5e_rq *rq, u16 ix) mlx5e_free_rx_wqe(rq, wi, false); } -static int mlx5e_alloc_rx_wqes(struct mlx5e_rq *rq, u16 ix, u8 wqe_bulk) +static int mlx5e_alloc_rx_wqes(struct mlx5e_rq *rq, u16 ix, int wqe_bulk) { struct mlx5_wq_cyc *wq = &rq->wqe.wq; - int err; int i; - if (rq->xsk_pool) { - int pages_desired = wqe_bulk << rq->wqe.info.log_num_frags; - - /* Check in advance that we have enough frames, instead of - * allocating one-by-one, failing and moving frames to the - * Reuse Ring. - */ - if (unlikely(!xsk_buff_can_alloc(rq->xsk_pool, pages_desired))) - return -ENOMEM; - } - for (i = 0; i < wqe_bulk; i++) { - struct mlx5e_rx_wqe_cyc *wqe = mlx5_wq_cyc_get_wqe(wq, ix + i); - - err = mlx5e_alloc_rx_wqe(rq, wqe, ix + i); - if (unlikely(err)) - goto free_wqes; - } + int j = mlx5_wq_cyc_ctr2ix(wq, ix + i); + struct mlx5e_rx_wqe_cyc *wqe; - return 0; + wqe = mlx5_wq_cyc_get_wqe(wq, j); -free_wqes: - while (--i >= 0) - mlx5e_dealloc_rx_wqe(rq, ix + i); + if (unlikely(mlx5e_alloc_rx_wqe(rq, wqe, j))) + break; + } - return err; + return i; } static inline void mlx5e_add_skb_frag(struct mlx5e_rq *rq, struct sk_buff *skb, - struct mlx5e_dma_info *di, u32 frag_offset, u32 len, + union mlx5e_alloc_unit *au, u32 frag_offset, u32 len, unsigned int truesize) { - dma_sync_single_for_cpu(rq->pdev, - di->addr + frag_offset, - len, DMA_FROM_DEVICE); - page_ref_inc(di->page); + dma_addr_t addr = page_pool_get_dma_addr(au->page); + + dma_sync_single_for_cpu(rq->pdev, addr + frag_offset, len, DMA_FROM_DEVICE); + page_ref_inc(au->page); skb_add_rx_frag(skb, skb_shinfo(skb)->nr_frags, - di->page, frag_offset, len, truesize); + au->page, frag_offset, len, truesize); } static inline void mlx5e_copy_skb_header(struct device *pdev, struct sk_buff *skb, - struct mlx5e_dma_info *dma_info, + struct page *page, dma_addr_t addr, int offset_from, int dma_offset, u32 headlen) { - const void *from = page_address(dma_info->page) + offset_from; + const void *from = page_address(page) + offset_from; /* Aligning len to sizeof(long) optimizes memcpy performance */ unsigned int len = ALIGN(headlen, sizeof(long)); - dma_sync_single_for_cpu(pdev, dma_info->addr + dma_offset, len, - DMA_FROM_DEVICE); + dma_sync_single_for_cpu(pdev, addr + dma_offset, len, DMA_FROM_DEVICE); skb_copy_to_linear_data(skb, from, len); } static void mlx5e_free_rx_mpwqe(struct mlx5e_rq *rq, struct mlx5e_mpw_info *wi, bool recycle) { + union mlx5e_alloc_unit *alloc_units = wi->alloc_units; bool no_xdp_xmit; - struct mlx5e_dma_info *dma_info = wi->umr.dma_info; int i; /* A common case for AF_XDP. */ - if (bitmap_full(wi->xdp_xmit_bitmap, MLX5_MPWRQ_PAGES_PER_WQE)) + if (bitmap_full(wi->xdp_xmit_bitmap, rq->mpwqe.pages_per_wqe)) return; - no_xdp_xmit = bitmap_empty(wi->xdp_xmit_bitmap, - MLX5_MPWRQ_PAGES_PER_WQE); + no_xdp_xmit = bitmap_empty(wi->xdp_xmit_bitmap, rq->mpwqe.pages_per_wqe); - for (i = 0; i < MLX5_MPWRQ_PAGES_PER_WQE; i++) - if (no_xdp_xmit || !test_bit(i, wi->xdp_xmit_bitmap)) - mlx5e_page_release(rq, &dma_info[i], recycle); + if (rq->xsk_pool) { + /* The `recycle` parameter is ignored, and the page is always + * put into the Reuse Ring, because there is no way to return + * the page to the userspace when the interface goes down. + */ + for (i = 0; i < rq->mpwqe.pages_per_wqe; i++) + if (no_xdp_xmit || !test_bit(i, wi->xdp_xmit_bitmap)) + xsk_buff_free(alloc_units[i].xsk); + } else { + for (i = 0; i < rq->mpwqe.pages_per_wqe; i++) + if (no_xdp_xmit || !test_bit(i, wi->xdp_xmit_bitmap)) + mlx5e_page_release_dynamic(rq, alloc_units[i].page, recycle); + } } static void mlx5e_post_rx_mpwqe(struct mlx5e_rq *rq, u8 n) @@ -574,11 +556,13 @@ static int mlx5e_build_shampo_hd_umr(struct mlx5e_rq *rq, header_offset = (index & (MLX5E_SHAMPO_WQ_HEADER_PER_PAGE - 1)) << MLX5E_SHAMPO_LOG_MAX_HEADER_ENTRY_SIZE; if (!(header_offset & (PAGE_SIZE - 1))) { - err = mlx5e_page_alloc(rq, dma_info); + union mlx5e_alloc_unit au; + + err = mlx5e_page_alloc_pool(rq, &au); if (unlikely(err)) goto err_unmap; - addr = dma_info->addr; - page = dma_info->page; + page = dma_info->page = au.page; + addr = dma_info->addr = page_pool_get_dma_addr(au.page); } else { dma_info->addr = addr + header_offset; dma_info->page = page; @@ -611,7 +595,7 @@ err_unmap: dma_info = &shampo->info[--index]; if (!(i & (MLX5E_SHAMPO_WQ_HEADER_PER_PAGE - 1))) { dma_info->addr = ALIGN_DOWN(dma_info->addr, PAGE_SIZE); - mlx5e_page_release(rq, dma_info, true); + mlx5e_page_release_dynamic(rq, dma_info->page, true); } } rq->stats->buff_alloc_err++; @@ -659,57 +643,55 @@ static int mlx5e_alloc_rx_hd_mpwqe(struct mlx5e_rq *rq) static int mlx5e_alloc_rx_mpwqe(struct mlx5e_rq *rq, u16 ix) { - struct mlx5e_mpw_info *wi = &rq->mpwqe.info[ix]; - struct mlx5e_dma_info *dma_info = &wi->umr.dma_info[0]; + struct mlx5e_mpw_info *wi = mlx5e_get_mpw_info(rq, ix); + union mlx5e_alloc_unit *au = &wi->alloc_units[0]; struct mlx5e_icosq *sq = rq->icosq; struct mlx5_wq_cyc *wq = &sq->wq; struct mlx5e_umr_wqe *umr_wqe; + u32 offset; /* 17-bit value with MTT. */ u16 pi; int err; int i; - /* Check in advance that we have enough frames, instead of allocating - * one-by-one, failing and moving frames to the Reuse Ring. - */ - if (rq->xsk_pool && - unlikely(!xsk_buff_can_alloc(rq->xsk_pool, MLX5_MPWRQ_PAGES_PER_WQE))) { - err = -ENOMEM; - goto err; - } - if (test_bit(MLX5E_RQ_STATE_SHAMPO, &rq->state)) { err = mlx5e_alloc_rx_hd_mpwqe(rq); if (unlikely(err)) goto err; } - pi = mlx5e_icosq_get_next_pi(sq, MLX5E_UMR_WQEBBS); + pi = mlx5e_icosq_get_next_pi(sq, rq->mpwqe.umr_wqebbs); umr_wqe = mlx5_wq_cyc_get_wqe(wq, pi); - memcpy(umr_wqe, &rq->mpwqe.umr_wqe, offsetof(struct mlx5e_umr_wqe, inline_mtts)); + memcpy(umr_wqe, &rq->mpwqe.umr_wqe, sizeof(struct mlx5e_umr_wqe)); + + for (i = 0; i < rq->mpwqe.pages_per_wqe; i++, au++) { + dma_addr_t addr; - for (i = 0; i < MLX5_MPWRQ_PAGES_PER_WQE; i++, dma_info++) { - err = mlx5e_page_alloc(rq, dma_info); + err = mlx5e_page_alloc_pool(rq, au); if (unlikely(err)) goto err_unmap; - umr_wqe->inline_mtts[i].ptag = cpu_to_be64(dma_info->addr | MLX5_EN_WR); + addr = page_pool_get_dma_addr(au->page); + umr_wqe->inline_mtts[i] = (struct mlx5_mtt) { + .ptag = cpu_to_be64(addr | MLX5_EN_WR), + }; } - bitmap_zero(wi->xdp_xmit_bitmap, MLX5_MPWRQ_PAGES_PER_WQE); + bitmap_zero(wi->xdp_xmit_bitmap, rq->mpwqe.pages_per_wqe); wi->consumed_strides = 0; umr_wqe->ctrl.opmod_idx_opcode = cpu_to_be32((sq->pc << MLX5_WQE_CTRL_WQE_INDEX_SHIFT) | MLX5_OPCODE_UMR); - umr_wqe->uctrl.xlt_offset = - cpu_to_be16(MLX5_ALIGNED_MTTS_OCTW(MLX5E_REQUIRED_MTTS(ix))); + + offset = (ix * rq->mpwqe.mtts_per_wqe) * sizeof(struct mlx5_mtt) / MLX5_OCTWORD; + umr_wqe->uctrl.xlt_offset = cpu_to_be16(offset); sq->db.wqe_info[pi] = (struct mlx5e_icosq_wqe_info) { .wqe_type = MLX5E_ICOSQ_WQE_UMR_RX, - .num_wqebbs = MLX5E_UMR_WQEBBS, + .num_wqebbs = rq->mpwqe.umr_wqebbs, .umr.rq = rq, }; - sq->pc += MLX5E_UMR_WQEBBS; + sq->pc += rq->mpwqe.umr_wqebbs; sq->doorbell_cseg = &umr_wqe->ctrl; @@ -717,8 +699,8 @@ static int mlx5e_alloc_rx_mpwqe(struct mlx5e_rq *rq, u16 ix) err_unmap: while (--i >= 0) { - dma_info--; - mlx5e_page_release(rq, dma_info, true); + au--; + mlx5e_page_release_dynamic(rq, au->page, true); } err: @@ -752,7 +734,7 @@ void mlx5e_shampo_dealloc_hd(struct mlx5e_rq *rq, u16 len, u16 start, bool close hd_info->addr = ALIGN_DOWN(hd_info->addr, PAGE_SIZE); if (hd_info->page != deleted_page) { deleted_page = hd_info->page; - mlx5e_page_release(rq, hd_info, false); + mlx5e_page_release_dynamic(rq, hd_info->page, false); } } @@ -767,7 +749,7 @@ void mlx5e_shampo_dealloc_hd(struct mlx5e_rq *rq, u16 len, u16 start, bool close static void mlx5e_dealloc_rx_mpwqe(struct mlx5e_rq *rq, u16 ix) { - struct mlx5e_mpw_info *wi = &rq->mpwqe.info[ix]; + struct mlx5e_mpw_info *wi = mlx5e_get_mpw_info(rq, ix); /* Don't recycle, this function is called on rq/netdev close */ mlx5e_free_rx_mpwqe(rq, wi, false); } @@ -775,38 +757,51 @@ static void mlx5e_dealloc_rx_mpwqe(struct mlx5e_rq *rq, u16 ix) INDIRECT_CALLABLE_SCOPE bool mlx5e_post_rx_wqes(struct mlx5e_rq *rq) { struct mlx5_wq_cyc *wq = &rq->wqe.wq; - u8 wqe_bulk; - int err; + int wqe_bulk, count; + bool busy = false; + u16 head; if (unlikely(!test_bit(MLX5E_RQ_STATE_ENABLED, &rq->state))) return false; - wqe_bulk = rq->wqe.info.wqe_bulk; - - if (mlx5_wq_cyc_missing(wq) < wqe_bulk) + if (mlx5_wq_cyc_missing(wq) < rq->wqe.info.wqe_bulk) return false; if (rq->page_pool) page_pool_nid_changed(rq->page_pool, numa_mem_id()); - do { - u16 head = mlx5_wq_cyc_get_head(wq); + wqe_bulk = mlx5_wq_cyc_missing(wq); + head = mlx5_wq_cyc_get_head(wq); - err = mlx5e_alloc_rx_wqes(rq, head, wqe_bulk); - if (unlikely(err)) { - rq->stats->buff_alloc_err++; - break; - } + /* Don't allow any newly allocated WQEs to share the same page with old + * WQEs that aren't completed yet. Stop earlier. + */ + wqe_bulk -= (head + wqe_bulk) & rq->wqe.info.wqe_index_mask; - mlx5_wq_cyc_push_n(wq, wqe_bulk); - } while (mlx5_wq_cyc_missing(wq) >= wqe_bulk); + if (!rq->xsk_pool) + count = mlx5e_alloc_rx_wqes(rq, head, wqe_bulk); + else if (likely(!rq->xsk_pool->dma_need_sync)) + count = mlx5e_xsk_alloc_rx_wqes_batched(rq, head, wqe_bulk); + else + /* If dma_need_sync is true, it's more efficient to call + * xsk_buff_alloc in a loop, rather than xsk_buff_alloc_batch, + * because the latter does the same check and returns only one + * frame. + */ + count = mlx5e_xsk_alloc_rx_wqes(rq, head, wqe_bulk); + + mlx5_wq_cyc_push_n(wq, count); + if (unlikely(count != wqe_bulk)) { + rq->stats->buff_alloc_err++; + busy = true; + } /* ensure wqes are visible to device before updating doorbell record */ dma_wmb(); mlx5_wq_cyc_update_db_record(wq); - return !!err; + return busy; } void mlx5e_free_icosq_descs(struct mlx5e_icosq *sq) @@ -974,7 +969,8 @@ INDIRECT_CALLABLE_SCOPE bool mlx5e_post_rx_mpwqes(struct mlx5e_rq *rq) head = rq->mpwqe.actual_wq_head; i = missing; do { - alloc_err = mlx5e_alloc_rx_mpwqe(rq, head); + alloc_err = rq->xsk_pool ? mlx5e_xsk_alloc_rx_mpwqe(rq, head) : + mlx5e_alloc_rx_mpwqe(rq, head); if (unlikely(alloc_err)) break; @@ -1421,6 +1417,9 @@ static inline void mlx5e_build_rx_skb(struct mlx5_cqe64 *cqe, if (unlikely(mlx5_ipsec_is_rx_flow(cqe))) mlx5e_ipsec_offload_handle_rx_skb(netdev, skb, cqe); + if (unlikely(mlx5e_macsec_is_rx_flow(cqe))) + mlx5e_macsec_offload_handle_rx_skb(netdev, skb, cqe); + if (lro_num_seg > 1) { mlx5e_lro_update_hdr(skb, cqe, cqe_bcnt); skb_shinfo(skb)->gso_size = DIV_ROUND_UP(cqe_bcnt, lro_num_seg); @@ -1524,19 +1523,21 @@ static struct sk_buff * mlx5e_skb_from_cqe_linear(struct mlx5e_rq *rq, struct mlx5e_wqe_frag_info *wi, u32 cqe_bcnt) { - struct mlx5e_dma_info *di = wi->di; + union mlx5e_alloc_unit *au = wi->au; u16 rx_headroom = rq->buff.headroom; struct bpf_prog *prog; struct sk_buff *skb; u32 metasize = 0; void *va, *data; + dma_addr_t addr; u32 frag_size; - va = page_address(di->page) + wi->offset; + va = page_address(au->page) + wi->offset; data = va + rx_headroom; frag_size = MLX5_SKB_FRAG_SZ(rx_headroom + cqe_bcnt); - dma_sync_single_range_for_cpu(rq->pdev, di->addr, wi->offset, + addr = page_pool_get_dma_addr(au->page); + dma_sync_single_range_for_cpu(rq->pdev, addr, wi->offset, frag_size, DMA_FROM_DEVICE); net_prefetch(data); @@ -1546,7 +1547,7 @@ mlx5e_skb_from_cqe_linear(struct mlx5e_rq *rq, struct mlx5e_wqe_frag_info *wi, net_prefetchw(va); /* xdp_frame data area */ mlx5e_fill_xdp_buff(rq, va, rx_headroom, cqe_bcnt, &xdp); - if (mlx5e_xdp_handle(rq, di->page, prog, &xdp)) + if (mlx5e_xdp_handle(rq, au->page, prog, &xdp)) return NULL; /* page/packet was consumed by XDP */ rx_headroom = xdp.data - xdp.data_hard_start; @@ -1559,7 +1560,7 @@ mlx5e_skb_from_cqe_linear(struct mlx5e_rq *rq, struct mlx5e_wqe_frag_info *wi, return NULL; /* queue up for recycling/reuse */ - page_ref_inc(di->page); + page_ref_inc(au->page); return skb; } @@ -1570,20 +1571,22 @@ mlx5e_skb_from_cqe_nonlinear(struct mlx5e_rq *rq, struct mlx5e_wqe_frag_info *wi { struct mlx5e_rq_frag_info *frag_info = &rq->wqe.info.arr[0]; struct mlx5e_wqe_frag_info *head_wi = wi; + union mlx5e_alloc_unit *au = wi->au; u16 rx_headroom = rq->buff.headroom; - struct mlx5e_dma_info *di = wi->di; struct skb_shared_info *sinfo; u32 frag_consumed_bytes; struct bpf_prog *prog; struct xdp_buff xdp; struct sk_buff *skb; + dma_addr_t addr; u32 truesize; void *va; - va = page_address(di->page) + wi->offset; + va = page_address(au->page) + wi->offset; frag_consumed_bytes = min_t(u32, frag_info->frag_size, cqe_bcnt); - dma_sync_single_range_for_cpu(rq->pdev, di->addr, wi->offset, + addr = page_pool_get_dma_addr(au->page); + dma_sync_single_range_for_cpu(rq->pdev, addr, wi->offset, rq->buff.frame0_sz, DMA_FROM_DEVICE); net_prefetchw(va); /* xdp_frame data area */ net_prefetch(va + rx_headroom); @@ -1599,11 +1602,12 @@ mlx5e_skb_from_cqe_nonlinear(struct mlx5e_rq *rq, struct mlx5e_wqe_frag_info *wi while (cqe_bcnt) { skb_frag_t *frag; - di = wi->di; + au = wi->au; frag_consumed_bytes = min_t(u32, frag_info->frag_size, cqe_bcnt); - dma_sync_single_for_cpu(rq->pdev, di->addr + wi->offset, + addr = page_pool_get_dma_addr(au->page); + dma_sync_single_for_cpu(rq->pdev, addr + wi->offset, frag_consumed_bytes, DMA_FROM_DEVICE); if (!xdp_buff_has_frags(&xdp)) { @@ -1616,11 +1620,11 @@ mlx5e_skb_from_cqe_nonlinear(struct mlx5e_rq *rq, struct mlx5e_wqe_frag_info *wi } frag = &sinfo->frags[sinfo->nr_frags++]; - __skb_frag_set_page(frag, di->page); + __skb_frag_set_page(frag, au->page); skb_frag_off_set(frag, wi->offset); skb_frag_size_set(frag, frag_consumed_bytes); - if (page_is_pfmemalloc(di->page)) + if (page_is_pfmemalloc(au->page)) xdp_buff_set_frag_pfmemalloc(&xdp); sinfo->xdp_frags_size += frag_consumed_bytes; @@ -1631,10 +1635,10 @@ mlx5e_skb_from_cqe_nonlinear(struct mlx5e_rq *rq, struct mlx5e_wqe_frag_info *wi wi++; } - di = head_wi->di; + au = head_wi->au; prog = rcu_dereference(rq->xdp_prog); - if (prog && mlx5e_xdp_handle(rq, di->page, prog, &xdp)) { + if (prog && mlx5e_xdp_handle(rq, au->page, prog, &xdp)) { if (test_bit(MLX5E_RQ_FLAG_XDP_XMIT, rq->flags)) { int i; @@ -1651,7 +1655,7 @@ mlx5e_skb_from_cqe_nonlinear(struct mlx5e_rq *rq, struct mlx5e_wqe_frag_info *wi if (unlikely(!skb)) return NULL; - page_ref_inc(di->page); + page_ref_inc(au->page); if (unlikely(xdp_buff_has_frags(&xdp))) { int i; @@ -1706,9 +1710,10 @@ static void mlx5e_handle_rx_cqe(struct mlx5e_rq *rq, struct mlx5_cqe64 *cqe) goto free_wqe; } - skb = INDIRECT_CALL_2(rq->wqe.skb_from_cqe, + skb = INDIRECT_CALL_3(rq->wqe.skb_from_cqe, mlx5e_skb_from_cqe_linear, mlx5e_skb_from_cqe_nonlinear, + mlx5e_xsk_skb_from_cqe_linear, rq, wi, cqe_bcnt); if (!skb) { /* probably for XDP */ @@ -1791,11 +1796,11 @@ static void mlx5e_handle_rx_cqe_mpwrq_rep(struct mlx5e_rq *rq, struct mlx5_cqe64 { u16 cstrides = mpwrq_get_cqe_consumed_strides(cqe); u16 wqe_id = be16_to_cpu(cqe->wqe_id); - struct mlx5e_mpw_info *wi = &rq->mpwqe.info[wqe_id]; + struct mlx5e_mpw_info *wi = mlx5e_get_mpw_info(rq, wqe_id); u16 stride_ix = mpwrq_get_cqe_stride_index(cqe); u32 wqe_offset = stride_ix << rq->mpwqe.log_stride_sz; - u32 head_offset = wqe_offset & (PAGE_SIZE - 1); - u32 page_idx = wqe_offset >> PAGE_SHIFT; + u32 head_offset = wqe_offset & ((1 << rq->mpwqe.page_shift) - 1); + u32 page_idx = wqe_offset >> rq->mpwqe.page_shift; struct mlx5e_rx_wqe_ll *wqe; struct mlx5_wq_ll *wq; struct sk_buff *skb; @@ -1846,12 +1851,13 @@ const struct mlx5e_rx_handlers mlx5e_rx_handlers_rep = { #endif static void -mlx5e_fill_skb_data(struct sk_buff *skb, struct mlx5e_rq *rq, struct mlx5e_dma_info *di, - u32 data_bcnt, u32 data_offset) +mlx5e_fill_skb_data(struct sk_buff *skb, struct mlx5e_rq *rq, + union mlx5e_alloc_unit *au, u32 data_bcnt, u32 data_offset) { net_prefetchw(skb->data); while (data_bcnt) { + /* Non-linear mode, hence non-XSK, which always uses PAGE_SIZE. */ u32 pg_consumed_bytes = min_t(u32, PAGE_SIZE - data_offset, data_bcnt); unsigned int truesize; @@ -1860,12 +1866,12 @@ mlx5e_fill_skb_data(struct sk_buff *skb, struct mlx5e_rq *rq, struct mlx5e_dma_i else truesize = ALIGN(pg_consumed_bytes, BIT(rq->mpwqe.log_stride_sz)); - mlx5e_add_skb_frag(rq, skb, di, data_offset, + mlx5e_add_skb_frag(rq, skb, au, data_offset, pg_consumed_bytes, truesize); data_bcnt -= pg_consumed_bytes; data_offset = 0; - di++; + au++; } } @@ -1873,12 +1879,13 @@ static struct sk_buff * mlx5e_skb_from_cqe_mpwrq_nonlinear(struct mlx5e_rq *rq, struct mlx5e_mpw_info *wi, u16 cqe_bcnt, u32 head_offset, u32 page_idx) { + union mlx5e_alloc_unit *au = &wi->alloc_units[page_idx]; u16 headlen = min_t(u16, MLX5E_RX_MAX_HEAD, cqe_bcnt); - struct mlx5e_dma_info *di = &wi->umr.dma_info[page_idx]; u32 frag_offset = head_offset + headlen; u32 byte_cnt = cqe_bcnt - headlen; - struct mlx5e_dma_info *head_di = di; + union mlx5e_alloc_unit *head_au = au; struct sk_buff *skb; + dma_addr_t addr; skb = napi_alloc_skb(rq->cq.napi, ALIGN(MLX5E_RX_MAX_HEAD, sizeof(long))); @@ -1889,14 +1896,17 @@ mlx5e_skb_from_cqe_mpwrq_nonlinear(struct mlx5e_rq *rq, struct mlx5e_mpw_info *w net_prefetchw(skb->data); + /* Non-linear mode, hence non-XSK, which always uses PAGE_SIZE. */ if (unlikely(frag_offset >= PAGE_SIZE)) { - di++; + au++; frag_offset -= PAGE_SIZE; } - mlx5e_fill_skb_data(skb, rq, di, byte_cnt, frag_offset); + mlx5e_fill_skb_data(skb, rq, au, byte_cnt, frag_offset); /* copy header */ - mlx5e_copy_skb_header(rq->pdev, skb, head_di, head_offset, head_offset, headlen); + addr = page_pool_get_dma_addr(head_au->page); + mlx5e_copy_skb_header(rq->pdev, skb, head_au->page, addr, + head_offset, head_offset, headlen); /* skb linear part was allocated with headlen and aligned to long */ skb->tail += headlen; skb->len += headlen; @@ -1908,12 +1918,13 @@ static struct sk_buff * mlx5e_skb_from_cqe_mpwrq_linear(struct mlx5e_rq *rq, struct mlx5e_mpw_info *wi, u16 cqe_bcnt, u32 head_offset, u32 page_idx) { - struct mlx5e_dma_info *di = &wi->umr.dma_info[page_idx]; + union mlx5e_alloc_unit *au = &wi->alloc_units[page_idx]; u16 rx_headroom = rq->buff.headroom; struct bpf_prog *prog; struct sk_buff *skb; u32 metasize = 0; void *va, *data; + dma_addr_t addr; u32 frag_size; /* Check packet size. Note LRO doesn't use linear SKB */ @@ -1922,11 +1933,12 @@ mlx5e_skb_from_cqe_mpwrq_linear(struct mlx5e_rq *rq, struct mlx5e_mpw_info *wi, return NULL; } - va = page_address(di->page) + head_offset; + va = page_address(au->page) + head_offset; data = va + rx_headroom; frag_size = MLX5_SKB_FRAG_SZ(rx_headroom + cqe_bcnt); - dma_sync_single_range_for_cpu(rq->pdev, di->addr, head_offset, + addr = page_pool_get_dma_addr(au->page); + dma_sync_single_range_for_cpu(rq->pdev, addr, head_offset, frag_size, DMA_FROM_DEVICE); net_prefetch(data); @@ -1936,7 +1948,7 @@ mlx5e_skb_from_cqe_mpwrq_linear(struct mlx5e_rq *rq, struct mlx5e_mpw_info *wi, net_prefetchw(va); /* xdp_frame data area */ mlx5e_fill_xdp_buff(rq, va, rx_headroom, cqe_bcnt, &xdp); - if (mlx5e_xdp_handle(rq, di->page, prog, &xdp)) { + if (mlx5e_xdp_handle(rq, au->page, prog, &xdp)) { if (__test_and_clear_bit(MLX5E_RQ_FLAG_XDP_XMIT, rq->flags)) __set_bit(page_idx, wi->xdp_xmit_bitmap); /* non-atomic */ return NULL; /* page/packet was consumed by XDP */ @@ -1952,7 +1964,7 @@ mlx5e_skb_from_cqe_mpwrq_linear(struct mlx5e_rq *rq, struct mlx5e_mpw_info *wi, return NULL; /* queue up for recycling/reuse */ - page_ref_inc(di->page); + page_ref_inc(au->page); return skb; } @@ -1997,7 +2009,7 @@ mlx5e_skb_from_cqe_shampo(struct mlx5e_rq *rq, struct mlx5e_mpw_info *wi, } prefetchw(skb->data); - mlx5e_copy_skb_header(rq->pdev, skb, head, + mlx5e_copy_skb_header(rq->pdev, skb, head->page, head->addr, head_offset + rx_headroom, rx_headroom, head_size); /* skb linear part was allocated with headlen and aligned to long */ @@ -2049,7 +2061,7 @@ mlx5e_free_rx_shampo_hd_entry(struct mlx5e_rq *rq, u16 header_index) if (((header_index + 1) & (MLX5E_SHAMPO_WQ_HEADER_PER_PAGE - 1)) == 0) { shampo->info[header_index].addr = ALIGN_DOWN(addr, PAGE_SIZE); - mlx5e_page_release(rq, &shampo->info[header_index], true); + mlx5e_page_release_dynamic(rq, shampo->info[header_index].page, true); } bitmap_clear(shampo->bitmap, header_index, 1); } @@ -2070,11 +2082,11 @@ static void mlx5e_handle_rx_cqe_mpwrq_shampo(struct mlx5e_rq *rq, struct mlx5_cq bool match = cqe->shampo.match; struct mlx5e_rq_stats *stats = rq->stats; struct mlx5e_rx_wqe_ll *wqe; - struct mlx5e_dma_info *di; + union mlx5e_alloc_unit *au; struct mlx5e_mpw_info *wi; struct mlx5_wq_ll *wq; - wi = &rq->mpwqe.info[wqe_id]; + wi = mlx5e_get_mpw_info(rq, wqe_id); wi->consumed_strides += cstrides; if (unlikely(MLX5E_RX_ERR_CQE(cqe))) { @@ -2120,8 +2132,8 @@ static void mlx5e_handle_rx_cqe_mpwrq_shampo(struct mlx5e_rq *rq, struct mlx5_cq } if (likely(head_size)) { - di = &wi->umr.dma_info[page_idx]; - mlx5e_fill_skb_data(*skb, rq, di, data_bcnt, data_offset); + au = &wi->alloc_units[page_idx]; + mlx5e_fill_skb_data(*skb, rq, au, data_bcnt, data_offset); } mlx5e_shampo_complete_rx_cqe(rq, cqe, cqe_bcnt, *skb); @@ -2143,11 +2155,11 @@ static void mlx5e_handle_rx_cqe_mpwrq(struct mlx5e_rq *rq, struct mlx5_cqe64 *cq { u16 cstrides = mpwrq_get_cqe_consumed_strides(cqe); u16 wqe_id = be16_to_cpu(cqe->wqe_id); - struct mlx5e_mpw_info *wi = &rq->mpwqe.info[wqe_id]; + struct mlx5e_mpw_info *wi = mlx5e_get_mpw_info(rq, wqe_id); u16 stride_ix = mpwrq_get_cqe_stride_index(cqe); u32 wqe_offset = stride_ix << rq->mpwqe.log_stride_sz; - u32 head_offset = wqe_offset & (PAGE_SIZE - 1); - u32 page_idx = wqe_offset >> PAGE_SHIFT; + u32 head_offset = wqe_offset & ((1 << rq->mpwqe.page_shift) - 1); + u32 page_idx = wqe_offset >> rq->mpwqe.page_shift; struct mlx5e_rx_wqe_ll *wqe; struct mlx5_wq_ll *wq; struct sk_buff *skb; @@ -2170,9 +2182,10 @@ static void mlx5e_handle_rx_cqe_mpwrq(struct mlx5e_rq *rq, struct mlx5_cqe64 *cq cqe_bcnt = mpwrq_get_cqe_byte_cnt(cqe); - skb = INDIRECT_CALL_2(rq->mpwqe.skb_from_cqe_mpwrq, + skb = INDIRECT_CALL_3(rq->mpwqe.skb_from_cqe_mpwrq, mlx5e_skb_from_cqe_mpwrq_linear, mlx5e_skb_from_cqe_mpwrq_nonlinear, + mlx5e_xsk_skb_from_cqe_mpwrq_linear, rq, wi, cqe_bcnt, head_offset, page_idx); if (!skb) goto mpwrq_cqe_out; @@ -2417,7 +2430,7 @@ int mlx5e_rq_set_handlers(struct mlx5e_rq *rq, struct mlx5e_params *params, bool default: /* MLX5_WQ_TYPE_CYCLIC */ rq->wqe.skb_from_cqe = xsk ? mlx5e_xsk_skb_from_cqe_linear : - mlx5e_rx_is_linear_skb(params, NULL) ? + mlx5e_rx_is_linear_skb(mdev, params, NULL) ? mlx5e_skb_from_cqe_linear : mlx5e_skb_from_cqe_nonlinear; rq->post_wqes = mlx5e_post_rx_wqes; @@ -2471,7 +2484,7 @@ free_wqe: void mlx5e_rq_set_trap_handlers(struct mlx5e_rq *rq, struct mlx5e_params *params) { - rq->wqe.skb_from_cqe = mlx5e_rx_is_linear_skb(params, NULL) ? + rq->wqe.skb_from_cqe = mlx5e_rx_is_linear_skb(rq->mdev, params, NULL) ? mlx5e_skb_from_cqe_linear : mlx5e_skb_from_cqe_nonlinear; rq->post_wqes = mlx5e_post_rx_wqes; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c index 7409829d1201..03c1841970f1 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c @@ -641,17 +641,26 @@ static const struct counter_desc vnic_env_stats_dev_oob_desc[] = { VNIC_ENV_OFF(vport_env.internal_rq_out_of_buffer) }, }; +static const struct counter_desc vnic_env_stats_drop_desc[] = { + { "rx_oversize_pkts_buffer", + VNIC_ENV_OFF(vport_env.eth_wqe_too_small) }, +}; + #define NUM_VNIC_ENV_STEER_COUNTERS(dev) \ (MLX5_CAP_GEN(dev, nic_receive_steering_discard) ? \ ARRAY_SIZE(vnic_env_stats_steer_desc) : 0) #define NUM_VNIC_ENV_DEV_OOB_COUNTERS(dev) \ (MLX5_CAP_GEN(dev, vnic_env_int_rq_oob) ? \ ARRAY_SIZE(vnic_env_stats_dev_oob_desc) : 0) +#define NUM_VNIC_ENV_DROP_COUNTERS(dev) \ + (MLX5_CAP_GEN(dev, eth_wqe_too_small) ? \ + ARRAY_SIZE(vnic_env_stats_drop_desc) : 0) static MLX5E_DECLARE_STATS_GRP_OP_NUM_STATS(vnic_env) { return NUM_VNIC_ENV_STEER_COUNTERS(priv->mdev) + - NUM_VNIC_ENV_DEV_OOB_COUNTERS(priv->mdev); + NUM_VNIC_ENV_DEV_OOB_COUNTERS(priv->mdev) + + NUM_VNIC_ENV_DROP_COUNTERS(priv->mdev); } static MLX5E_DECLARE_STATS_GRP_OP_FILL_STRS(vnic_env) @@ -665,6 +674,11 @@ static MLX5E_DECLARE_STATS_GRP_OP_FILL_STRS(vnic_env) for (i = 0; i < NUM_VNIC_ENV_DEV_OOB_COUNTERS(priv->mdev); i++) strcpy(data + (idx++) * ETH_GSTRING_LEN, vnic_env_stats_dev_oob_desc[i].format); + + for (i = 0; i < NUM_VNIC_ENV_DROP_COUNTERS(priv->mdev); i++) + strcpy(data + (idx++) * ETH_GSTRING_LEN, + vnic_env_stats_drop_desc[i].format); + return idx; } @@ -679,6 +693,11 @@ static MLX5E_DECLARE_STATS_GRP_OP_FILL_STATS(vnic_env) for (i = 0; i < NUM_VNIC_ENV_DEV_OOB_COUNTERS(priv->mdev); i++) data[idx++] = MLX5E_READ_CTR32_BE(priv->stats.vnic.query_vnic_env_out, vnic_env_stats_dev_oob_desc, i); + + for (i = 0; i < NUM_VNIC_ENV_DROP_COUNTERS(priv->mdev); i++) + data[idx++] = MLX5E_READ_CTR32_BE(priv->stats.vnic.query_vnic_env_out, + vnic_env_stats_drop_desc, i); + return idx; } @@ -2451,6 +2470,9 @@ mlx5e_stats_grp_t mlx5e_nic_stats_grps[] = { &MLX5E_STATS_GRP(per_port_buff_congest), &MLX5E_STATS_GRP(ptp), &MLX5E_STATS_GRP(qos), +#ifdef CONFIG_MLX5_EN_MACSEC + &MLX5E_STATS_GRP(macsec_hw), +#endif }; unsigned int mlx5e_nic_stats_grps_num(struct mlx5e_priv *priv) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h index ed4fc940e4ef..9f781085be47 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h @@ -273,6 +273,10 @@ struct mlx5e_qcounter_stats { u32 rx_if_down_packets; }; +#define VNIC_ENV_GET(vnic_env_stats, c) \ + MLX5_GET(query_vnic_env_out, (vnic_env_stats)->query_vnic_env_out, \ + vport_env.c) + struct mlx5e_vnic_env_stats { __be64 query_vnic_env_out[MLX5_ST_SZ_QW(query_vnic_env_out)]; }; @@ -486,5 +490,6 @@ extern MLX5E_DECLARE_STATS_GRP(channels); extern MLX5E_DECLARE_STATS_GRP(per_port_buff_congest); extern MLX5E_DECLARE_STATS_GRP(ipsec_sw); extern MLX5E_DECLARE_STATS_GRP(ptp); +extern MLX5E_DECLARE_STATS_GRP(macsec_hw); #endif /* __MLX5_EN_STATS_H__ */ diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c index f154bda668ad..70a7a61f9708 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c @@ -311,6 +311,7 @@ mlx5e_get_flow_meters(struct mlx5_core_dev *dev) static struct mlx5_tc_ct_priv * get_ct_priv(struct mlx5e_priv *priv) { + struct mlx5e_tc_table *tc = mlx5e_fs_get_tc(priv->fs); struct mlx5_eswitch *esw = priv->mdev->priv.eswitch; struct mlx5_rep_uplink_priv *uplink_priv; struct mlx5e_rep_priv *uplink_rpriv; @@ -322,7 +323,7 @@ get_ct_priv(struct mlx5e_priv *priv) return uplink_priv->ct_priv; } - return priv->fs->tc->ct; + return tc->ct; } static struct mlx5e_tc_psample * @@ -345,6 +346,7 @@ get_sample_priv(struct mlx5e_priv *priv) static struct mlx5e_post_act * get_post_action(struct mlx5e_priv *priv) { + struct mlx5e_tc_table *tc = mlx5e_fs_get_tc(priv->fs); struct mlx5_eswitch *esw = priv->mdev->priv.eswitch; struct mlx5_rep_uplink_priv *uplink_priv; struct mlx5e_rep_priv *uplink_rpriv; @@ -356,7 +358,7 @@ get_post_action(struct mlx5e_priv *priv) return uplink_priv->post_act; } - return priv->fs->tc->post_act; + return tc->post_act; } struct mlx5_flow_handle * @@ -607,11 +609,12 @@ int mlx5e_get_flow_namespace(struct mlx5e_tc_flow *flow) static struct mod_hdr_tbl * get_mod_hdr_table(struct mlx5e_priv *priv, struct mlx5e_tc_flow *flow) { + struct mlx5e_tc_table *tc = mlx5e_fs_get_tc(priv->fs); struct mlx5_eswitch *esw = priv->mdev->priv.eswitch; return mlx5e_get_flow_namespace(flow) == MLX5_FLOW_NAMESPACE_FDB ? &esw->offloads.mod_hdr : - &priv->fs->tc->mod_hdr; + &tc->mod_hdr; } static int mlx5e_attach_mod_hdr(struct mlx5e_priv *priv, @@ -810,6 +813,7 @@ static int mlx5e_hairpin_rss_init(struct mlx5e_hairpin *hp) { struct mlx5e_priv *priv = hp->func_priv; struct ttc_params ttc_params; + struct mlx5_ttc_table *ttc; int err; err = mlx5e_hairpin_create_indirect_rqt(hp); @@ -827,9 +831,10 @@ static int mlx5e_hairpin_rss_init(struct mlx5e_hairpin *hp) goto err_create_ttc_table; } + ttc = mlx5e_fs_get_ttc(priv->fs, false); netdev_dbg(priv->netdev, "add hairpin: using %d channels rss ttc table id %x\n", hp->num_channels, - mlx5_get_ttc_flow_table(priv->fs->ttc)->id); + mlx5_get_ttc_flow_table(ttc)->id); return 0; @@ -916,10 +921,11 @@ static inline u32 hash_hairpin_info(u16 peer_vhca_id, u8 prio) static struct mlx5e_hairpin_entry *mlx5e_hairpin_get(struct mlx5e_priv *priv, u16 peer_vhca_id, u8 prio) { + struct mlx5e_tc_table *tc = mlx5e_fs_get_tc(priv->fs); 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, + hash_for_each_possible(tc->hairpin_tbl, hpe, hairpin_hlist, hash_key) { if (hpe->peer_vhca_id == peer_vhca_id && hpe->prio == prio) { refcount_inc(&hpe->refcnt); @@ -933,11 +939,12 @@ static struct mlx5e_hairpin_entry *mlx5e_hairpin_get(struct mlx5e_priv *priv, static void mlx5e_hairpin_put(struct mlx5e_priv *priv, struct mlx5e_hairpin_entry *hpe) { + struct mlx5e_tc_table *tc = mlx5e_fs_get_tc(priv->fs); /* no more hairpin flows for us, release the hairpin pair */ - if (!refcount_dec_and_mutex_lock(&hpe->refcnt, &priv->fs->tc->hairpin_tbl_lock)) + if (!refcount_dec_and_mutex_lock(&hpe->refcnt, &tc->hairpin_tbl_lock)) return; hash_del(&hpe->hairpin_hlist); - mutex_unlock(&priv->fs->tc->hairpin_tbl_lock); + mutex_unlock(&tc->hairpin_tbl_lock); if (!IS_ERR_OR_NULL(hpe->hp)) { netdev_dbg(priv->netdev, "del hairpin: peer %s\n", @@ -993,6 +1000,7 @@ static int mlx5e_hairpin_flow_add(struct mlx5e_priv *priv, struct mlx5e_tc_flow_parse_attr *parse_attr, struct netlink_ext_ack *extack) { + struct mlx5e_tc_table *tc = mlx5e_fs_get_tc(priv->fs); int peer_ifindex = parse_attr->mirred_ifindex[0]; struct mlx5_hairpin_params params; struct mlx5_core_dev *peer_mdev; @@ -1021,10 +1029,10 @@ static int mlx5e_hairpin_flow_add(struct mlx5e_priv *priv, if (err) return err; - mutex_lock(&priv->fs->tc->hairpin_tbl_lock); + mutex_lock(&tc->hairpin_tbl_lock); hpe = mlx5e_hairpin_get(priv, peer_id, match_prio); if (hpe) { - mutex_unlock(&priv->fs->tc->hairpin_tbl_lock); + mutex_unlock(&tc->hairpin_tbl_lock); wait_for_completion(&hpe->res_ready); if (IS_ERR(hpe->hp)) { @@ -1036,7 +1044,7 @@ static int mlx5e_hairpin_flow_add(struct mlx5e_priv *priv, hpe = kzalloc(sizeof(*hpe), GFP_KERNEL); if (!hpe) { - mutex_unlock(&priv->fs->tc->hairpin_tbl_lock); + mutex_unlock(&tc->hairpin_tbl_lock); return -ENOMEM; } @@ -1048,9 +1056,9 @@ static int mlx5e_hairpin_flow_add(struct mlx5e_priv *priv, refcount_set(&hpe->refcnt, 1); init_completion(&hpe->res_ready); - hash_add(priv->fs->tc->hairpin_tbl, &hpe->hairpin_hlist, + hash_add(tc->hairpin_tbl, &hpe->hairpin_hlist, hash_hairpin_info(peer_id, match_prio)); - mutex_unlock(&priv->fs->tc->hairpin_tbl_lock); + mutex_unlock(&tc->hairpin_tbl_lock); params.log_data_size = 16; params.log_data_size = min_t(u8, params.log_data_size, @@ -1126,8 +1134,9 @@ mlx5e_add_offloaded_nic_rule(struct mlx5e_priv *priv, struct mlx5_flow_attr *attr) { struct mlx5_flow_context *flow_context = &spec->flow_context; + struct mlx5e_vlan_table *vlan = mlx5e_fs_get_vlan(priv->fs); + struct mlx5e_tc_table *tc = mlx5e_fs_get_tc(priv->fs); struct mlx5_nic_flow_attr *nic_attr = attr->nic_attr; - struct mlx5e_tc_table *tc = priv->fs->tc; struct mlx5_flow_destination dest[2] = {}; struct mlx5_fs_chains *nic_chains; struct mlx5_flow_act flow_act = { @@ -1163,7 +1172,7 @@ mlx5e_add_offloaded_nic_rule(struct mlx5e_priv *priv, if (IS_ERR(dest[dest_ix].ft)) return ERR_CAST(dest[dest_ix].ft); } else { - dest[dest_ix].ft = mlx5e_vlan_get_flowtable(priv->fs->vlan); + dest[dest_ix].ft = mlx5e_vlan_get_flowtable(vlan); } dest_ix++; } @@ -1191,7 +1200,7 @@ mlx5e_add_offloaded_nic_rule(struct mlx5e_priv *priv, mutex_unlock(&tc->t_lock); netdev_err(priv->netdev, "Failed to create tc offload table\n"); - rule = ERR_CAST(priv->fs->tc->t); + rule = ERR_CAST(tc->t); goto err_ft_get; } } @@ -1293,8 +1302,10 @@ void mlx5e_del_offloaded_nic_rule(struct mlx5e_priv *priv, struct mlx5_flow_handle *rule, struct mlx5_flow_attr *attr) { - struct mlx5_fs_chains *nic_chains = mlx5e_nic_chains(priv->fs->tc); + struct mlx5e_tc_table *tc = mlx5e_fs_get_tc(priv->fs); + struct mlx5_fs_chains *nic_chains; + nic_chains = mlx5e_nic_chains(tc); mlx5_del_flow_rules(rule); if (attr->chain || attr->prio) @@ -1309,8 +1320,8 @@ void mlx5e_del_offloaded_nic_rule(struct mlx5e_priv *priv, static void mlx5e_tc_del_nic_flow(struct mlx5e_priv *priv, struct mlx5e_tc_flow *flow) { + struct mlx5e_tc_table *tc = mlx5e_fs_get_tc(priv->fs); struct mlx5_flow_attr *attr = flow->attr; - struct mlx5e_tc_table *tc = priv->fs->tc; flow_flag_clear(flow, OFFLOADED); @@ -1322,13 +1333,13 @@ static void mlx5e_tc_del_nic_flow(struct mlx5e_priv *priv, /* Remove root table if no rules are left to avoid * extra steering hops. */ - mutex_lock(&priv->fs->tc->t_lock); + mutex_lock(&tc->t_lock); if (!mlx5e_tc_num_filters(priv, MLX5_TC_FLAG(NIC_OFFLOAD)) && !IS_ERR_OR_NULL(tc->t)) { mlx5_chains_put_table(mlx5e_nic_chains(tc), 0, 1, MLX5E_TC_FT_LEVEL); - priv->fs->tc->t = NULL; + tc->t = NULL; } - mutex_unlock(&priv->fs->tc->t_lock); + mutex_unlock(&tc->t_lock); if (attr->action & MLX5_FLOW_CONTEXT_ACTION_MOD_HDR) mlx5e_detach_mod_hdr(priv, flow); @@ -1494,8 +1505,11 @@ bool mlx5e_tc_is_vf_tunnel(struct net_device *out_dev, struct net_device *route_ route_priv = netdev_priv(route_dev); route_mdev = route_priv->mdev; - if (out_mdev->coredev_type != MLX5_COREDEV_PF || - route_mdev->coredev_type != MLX5_COREDEV_VF) + if (out_mdev->coredev_type != MLX5_COREDEV_PF) + return false; + + if (route_mdev->coredev_type != MLX5_COREDEV_VF && + route_mdev->coredev_type != MLX5_COREDEV_SF) return false; return mlx5e_same_hw_devs(out_priv, route_priv); @@ -4058,13 +4072,14 @@ static const struct rhashtable_params tc_ht_params = { static struct rhashtable *get_tc_ht(struct mlx5e_priv *priv, unsigned long flags) { + struct mlx5e_tc_table *tc = mlx5e_fs_get_tc(priv->fs); struct mlx5e_rep_priv *rpriv; if (flags & MLX5_TC_FLAG(ESW_OFFLOAD)) { rpriv = priv->ppriv; return &rpriv->tc_ht; } else /* NIC offload */ - return &priv->fs->tc->ht; + return &tc->ht; } static bool is_peer_flow_needed(struct mlx5e_tc_flow *flow) @@ -4448,7 +4463,7 @@ int mlx5e_configure_flower(struct net_device *dev, struct mlx5e_priv *priv, int err = 0; if (!mlx5_esw_hold(priv->mdev)) - return -EAGAIN; + return -EBUSY; mlx5_esw_get(priv->mdev); @@ -4772,6 +4787,7 @@ void mlx5e_tc_stats_matchall(struct mlx5e_priv *priv, static void mlx5e_tc_hairpin_update_dead_peer(struct mlx5e_priv *priv, struct mlx5e_priv *peer_priv) { + struct mlx5e_tc_table *tc = mlx5e_fs_get_tc(priv->fs); struct mlx5_core_dev *peer_mdev = peer_priv->mdev; struct mlx5e_hairpin_entry *hpe, *tmp; LIST_HEAD(init_wait_list); @@ -4783,11 +4799,11 @@ static void mlx5e_tc_hairpin_update_dead_peer(struct mlx5e_priv *priv, peer_vhca_id = MLX5_CAP_GEN(peer_mdev, vhca_id); - mutex_lock(&priv->fs->tc->hairpin_tbl_lock); - hash_for_each(priv->fs->tc->hairpin_tbl, bkt, hpe, hairpin_hlist) + mutex_lock(&tc->hairpin_tbl_lock); + hash_for_each(tc->hairpin_tbl, bkt, hpe, hairpin_hlist) if (refcount_inc_not_zero(&hpe->refcnt)) list_add(&hpe->dead_peer_wait_list, &init_wait_list); - mutex_unlock(&priv->fs->tc->hairpin_tbl_lock); + mutex_unlock(&tc->hairpin_tbl_lock); list_for_each_entry_safe(hpe, tmp, &init_wait_list, dead_peer_wait_list) { wait_for_completion(&hpe->res_ready); @@ -4841,7 +4857,8 @@ static int mlx5e_tc_nic_get_ft_size(struct mlx5_core_dev *dev) static int mlx5e_tc_nic_create_miss_table(struct mlx5e_priv *priv) { - struct mlx5_flow_table **ft = &priv->fs->tc->miss_t; + struct mlx5e_tc_table *tc = mlx5e_fs_get_tc(priv->fs); + struct mlx5_flow_table **ft = &tc->miss_t; struct mlx5_flow_table_attr ft_attr = {}; struct mlx5_flow_namespace *ns; int err = 0; @@ -4863,12 +4880,14 @@ static int mlx5e_tc_nic_create_miss_table(struct mlx5e_priv *priv) static void mlx5e_tc_nic_destroy_miss_table(struct mlx5e_priv *priv) { - mlx5_destroy_flow_table(priv->fs->tc->miss_t); + struct mlx5e_tc_table *tc = mlx5e_fs_get_tc(priv->fs); + + mlx5_destroy_flow_table(tc->miss_t); } int mlx5e_tc_nic_init(struct mlx5e_priv *priv) { - struct mlx5e_tc_table *tc = priv->fs->tc; + struct mlx5e_tc_table *tc = mlx5e_fs_get_tc(priv->fs); struct mlx5_core_dev *dev = priv->mdev; struct mapping_ctx *chains_mapping; struct mlx5_chains_attr attr = {}; @@ -4909,7 +4928,7 @@ int mlx5e_tc_nic_init(struct mlx5e_priv *priv) attr.ns = MLX5_FLOW_NAMESPACE_KERNEL; attr.max_ft_sz = mlx5e_tc_nic_get_ft_size(dev); attr.max_grp_num = MLX5E_TC_TABLE_NUM_GROUPS; - attr.default_ft = priv->fs->tc->miss_t; + attr.default_ft = tc->miss_t; attr.mapping = chains_mapping; tc->chains = mlx5_chains_create(dev, &attr); @@ -4958,7 +4977,7 @@ static void _mlx5e_tc_del_flow(void *ptr, void *arg) void mlx5e_tc_nic_cleanup(struct mlx5e_priv *priv) { - struct mlx5e_tc_table *tc = priv->fs->tc; + struct mlx5e_tc_table *tc = mlx5e_fs_get_tc(priv->fs); if (tc->netdevice_nb.notifier_call) unregister_netdevice_notifier_dev_net(priv->netdev, @@ -5163,13 +5182,13 @@ bool mlx5e_tc_update_skb(struct mlx5_cqe64 *cqe, #if IS_ENABLED(CONFIG_NET_TC_SKB_EXT) u32 chain = 0, chain_tag, reg_b, zone_restore_id; struct mlx5e_priv *priv = netdev_priv(skb->dev); - struct mlx5e_tc_table *tc = priv->fs->tc; struct mlx5_mapped_obj mapped_obj; struct tc_skb_ext *tc_skb_ext; + struct mlx5e_tc_table *tc; int err; reg_b = be32_to_cpu(cqe->ft_metadata); - + tc = mlx5e_fs_get_tc(priv->fs); chain_tag = reg_b & MLX5E_TC_TABLE_CHAIN_TAG_MASK; err = mapping_find(tc->mapping, chain_tag, &mapped_obj); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.h b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.h index 6ce1ab6b86b7..48241317a535 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.h @@ -54,6 +54,7 @@ ESW_FLOW_ATTR_SZ :\ NIC_FLOW_ATTR_SZ) +struct mlx5_fs_chains *mlx5e_nic_chains(struct mlx5e_tc_table *tc); int mlx5e_tc_num_filters(struct mlx5e_priv *priv, unsigned long flags); struct mlx5e_tc_update_priv { diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c index 27f791feb517..bf2232a2a836 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c @@ -39,6 +39,7 @@ #include "ipoib/ipoib.h" #include "en_accel/en_accel.h" #include "en_accel/ipsec_rxtx.h" +#include "en_accel/macsec.h" #include "en/ptp.h" #include <net/ipv6.h> @@ -485,7 +486,7 @@ err_drop: static bool mlx5e_tx_skb_supports_mpwqe(struct sk_buff *skb, struct mlx5e_tx_attr *attr) { return !skb_is_nonlinear(skb) && !skb_vlan_tag_present(skb) && !attr->ihs && - !attr->insz; + !attr->insz && !mlx5e_macsec_skb_is_offload(skb); } static bool mlx5e_tx_mpwqe_same_eseg(struct mlx5e_txqsq *sq, struct mlx5_wqe_eth_seg *eseg) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_txrx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_txrx.c index 833be29170a1..9a458a5d9853 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_txrx.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_txrx.c @@ -31,6 +31,7 @@ */ #include <linux/irq.h> +#include <net/xdp_sock_drv.h> #include "en.h" #include "en/txrx.h" #include "en/xdp.h" @@ -86,26 +87,36 @@ void mlx5e_trigger_irq(struct mlx5e_icosq *sq) static bool mlx5e_napi_xsk_post(struct mlx5e_xdpsq *xsksq, struct mlx5e_rq *xskrq) { + bool need_wakeup = xsk_uses_need_wakeup(xskrq->xsk_pool); bool busy_xsk = false, xsk_rx_alloc_err; - /* Handle the race between the application querying need_wakeup and the - * driver setting it: - * 1. Update need_wakeup both before and after the TX. If it goes to - * "yes", it can only happen with the first update. - * 2. If the application queried need_wakeup before we set it, the - * packets will be transmitted anyway, even w/o a wakeup. - * 3. Give a chance to clear need_wakeup after new packets were queued - * for TX. + /* If SQ is empty, there are no TX completions to trigger NAPI, so set + * need_wakeup. Do it before queuing packets for TX to avoid race + * condition with userspace. */ - mlx5e_xsk_update_tx_wakeup(xsksq); + if (need_wakeup && xsksq->pc == xsksq->cc) + xsk_set_tx_need_wakeup(xsksq->xsk_pool); busy_xsk |= mlx5e_xsk_tx(xsksq, MLX5E_TX_XSK_POLL_BUDGET); - mlx5e_xsk_update_tx_wakeup(xsksq); + /* If we queued some packets for TX, no need for wakeup anymore. */ + if (need_wakeup && xsksq->pc != xsksq->cc) + xsk_clear_tx_need_wakeup(xsksq->xsk_pool); + /* If WQ is empty, RX won't trigger NAPI, so set need_wakeup. Do it + * before refilling to avoid race condition with userspace. + */ + if (need_wakeup && !mlx5e_rqwq_get_cur_sz(xskrq)) + xsk_set_rx_need_wakeup(xskrq->xsk_pool); xsk_rx_alloc_err = INDIRECT_CALL_2(xskrq->post_wqes, mlx5e_post_rx_mpwqes, mlx5e_post_rx_wqes, xskrq); - busy_xsk |= mlx5e_xsk_update_rx_wakeup(xskrq, xsk_rx_alloc_err); + /* Ask for wakeup if WQ is not full after refill. */ + if (!need_wakeup) + busy_xsk |= xsk_rx_alloc_err; + else if (xsk_rx_alloc_err) + xsk_set_rx_need_wakeup(xskrq->xsk_pool); + else + xsk_clear_rx_need_wakeup(xskrq->xsk_pool); return busy_xsk; } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eq.c b/drivers/net/ethernet/mellanox/mlx5/core/eq.c index 229728c80233..a0242dc15741 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eq.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/eq.c @@ -575,6 +575,9 @@ static void gather_async_events_mask(struct mlx5_core_dev *dev, u64 mask[4]) if (MLX5_CAP_GEN_MAX(dev, vhca_state)) async_event_mask |= (1ull << MLX5_EVENT_TYPE_VHCA_STATE_CHANGE); + if (MLX5_CAP_MACSEC(dev, log_max_macsec_offload)) + async_event_mask |= (1ull << MLX5_EVENT_TYPE_OBJECT_CHANGE); + mask[0] = async_event_mask; if (MLX5_CAP_GEN(dev, event_cap)) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/esw/indir_table.c b/drivers/net/ethernet/mellanox/mlx5/core/esw/indir_table.c index 0abef71cb839..c9a91158e99c 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/esw/indir_table.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/esw/indir_table.c @@ -78,12 +78,16 @@ mlx5_esw_indir_table_needed(struct mlx5_eswitch *esw, struct mlx5_core_dev *dest_mdev) { struct mlx5_esw_flow_attr *esw_attr = attr->esw_attr; + bool vf_sf_vport; + + vf_sf_vport = mlx5_eswitch_is_vf_vport(esw, vport_num) || + mlx5_esw_is_sf_vport(esw, vport_num); /* Use indirect table for all IP traffic from UL to VF with vport * destination when source rewrite flag is set. */ return esw_attr->in_rep->vport == MLX5_VPORT_UPLINK && - mlx5_eswitch_is_vf_vport(esw, vport_num) && + vf_sf_vport && esw->dev == dest_mdev && attr->ip_version && attr->flags & MLX5_ATTR_FLAG_SRC_REWRITE; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.c b/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.c index 694c54066955..4f8a24d84a86 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.c @@ -924,12 +924,16 @@ int mlx5_esw_qos_vport_update_group(struct mlx5_eswitch *esw, struct mlx5_esw_rate_group *group, struct netlink_ext_ack *extack) { - int err; + int err = 0; mutex_lock(&esw->state_lock); + if (!vport->qos.enabled && !group) + goto unlock; + err = esw_qos_vport_enable(esw, vport, 0, 0, extack); if (!err) err = esw_qos_vport_update_group(esw, vport, group, extack); +unlock: mutex_unlock(&esw->state_lock); return err; } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c index 6aa58044b949..c59107fa9e6d 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c @@ -1360,7 +1360,6 @@ void mlx5_eswitch_disable_sriov(struct mlx5_eswitch *esw, bool clear_vf) if (esw->mode == MLX5_ESWITCH_OFFLOADS) { struct devlink *devlink = priv_to_devlink(esw->dev); - esw_offloads_del_send_to_vport_meta_rules(esw); devl_rate_nodes_destroy(devlink); } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h index 87ce5a208cb5..f68dc2d0dbe6 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h @@ -244,6 +244,8 @@ struct mlx5_esw_offload { struct mlx5_flow_table *ft_offloads; struct mlx5_flow_group *vport_rx_group; + struct mlx5_flow_group *vport_rx_drop_group; + struct mlx5_flow_handle *vport_rx_drop_rule; struct xarray vport_reps; struct list_head peer_flows; struct mutex peer_mutex; @@ -344,7 +346,10 @@ void esw_offloads_disable(struct mlx5_eswitch *esw); int esw_offloads_enable(struct mlx5_eswitch *esw); void esw_offloads_cleanup_reps(struct mlx5_eswitch *esw); int esw_offloads_init_reps(struct mlx5_eswitch *esw); -void esw_offloads_del_send_to_vport_meta_rules(struct mlx5_eswitch *esw); + +struct mlx5_flow_handle * +mlx5_eswitch_add_send_to_vport_meta_rule(struct mlx5_eswitch *esw, u16 vport_num); +void mlx5_eswitch_del_send_to_vport_meta_rule(struct mlx5_flow_handle *rule); bool mlx5_esw_vport_match_metadata_supported(const struct mlx5_eswitch *esw); int mlx5_esw_offloads_vport_metadata_set(struct mlx5_eswitch *esw, bool enable); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c index a9f4c652f859..4e50df3139c6 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c @@ -70,6 +70,8 @@ #define MLX5_ESW_VPORT_TBL_SIZE 128 #define MLX5_ESW_VPORT_TBL_NUM_GROUPS 4 +#define MLX5_ESW_FT_OFFLOADS_DROP_RULE (1) + static const struct esw_vport_tbl_namespace mlx5_esw_vport_tbl_mirror_ns = { .max_fte = MLX5_ESW_VPORT_TBL_SIZE, .max_num_groups = MLX5_ESW_VPORT_TBL_NUM_GROUPS, @@ -481,25 +483,27 @@ esw_setup_dests(struct mlx5_flow_destination *dest, !(attr->flags & MLX5_ATTR_FLAG_SLOW_PATH)) { esw_setup_sampler_dest(dest, flow_act, attr->sample_attr.sampler_id, *i); (*i)++; - } else if (attr->dest_ft) { - esw_setup_ft_dest(dest, flow_act, esw, attr, spec, *i); - (*i)++; } else if (attr->flags & MLX5_ATTR_FLAG_SLOW_PATH) { esw_setup_slow_path_dest(dest, flow_act, esw, *i); (*i)++; } else if (attr->flags & MLX5_ATTR_FLAG_ACCEPT) { esw_setup_accept_dest(dest, flow_act, chains, *i); (*i)++; - } else if (attr->dest_chain) { - err = esw_setup_chain_dest(dest, flow_act, chains, attr->dest_chain, - 1, 0, *i); - (*i)++; } else if (esw_is_indir_table(esw, attr)) { err = esw_setup_indir_table(dest, flow_act, esw, attr, spec, true, i); } else if (esw_is_chain_src_port_rewrite(esw, esw_attr)) { err = esw_setup_chain_src_port_rewrite(dest, flow_act, esw, chains, attr, i); } else { *i = esw_setup_vport_dests(dest, flow_act, esw, esw_attr, *i); + + if (attr->dest_ft) { + err = esw_setup_ft_dest(dest, flow_act, esw, attr, spec, *i); + (*i)++; + } else if (attr->dest_chain) { + err = esw_setup_chain_dest(dest, flow_act, chains, attr->dest_chain, + 1, 0, *i); + (*i)++; + } } return err; @@ -1058,52 +1062,23 @@ void mlx5_eswitch_del_send_to_vport_rule(struct mlx5_flow_handle *rule) mlx5_del_flow_rules(rule); } -static void mlx5_eswitch_del_send_to_vport_meta_rules(struct mlx5_eswitch *esw) +void mlx5_eswitch_del_send_to_vport_meta_rule(struct mlx5_flow_handle *rule) { - struct mlx5_flow_handle **flows = esw->fdb_table.offloads.send_to_vport_meta_rules; - int i = 0, num_vfs = esw->esw_funcs.num_vfs; - - if (!num_vfs || !flows) - return; - - for (i = 0; i < num_vfs; i++) - mlx5_del_flow_rules(flows[i]); - - kvfree(flows); - /* If changing eswitch mode from switchdev to legacy, but num_vfs is not 0, - * meta rules could be freed again. So set it to NULL. - */ - esw->fdb_table.offloads.send_to_vport_meta_rules = NULL; + if (rule) + mlx5_del_flow_rules(rule); } -void esw_offloads_del_send_to_vport_meta_rules(struct mlx5_eswitch *esw) -{ - mlx5_eswitch_del_send_to_vport_meta_rules(esw); -} - -static int -mlx5_eswitch_add_send_to_vport_meta_rules(struct mlx5_eswitch *esw) +struct mlx5_flow_handle * +mlx5_eswitch_add_send_to_vport_meta_rule(struct mlx5_eswitch *esw, u16 vport_num) { struct mlx5_flow_destination dest = {}; struct mlx5_flow_act flow_act = {0}; - int num_vfs, rule_idx = 0, err = 0; struct mlx5_flow_handle *flow_rule; - struct mlx5_flow_handle **flows; struct mlx5_flow_spec *spec; - struct mlx5_vport *vport; - unsigned long i; - u16 vport_num; - - num_vfs = esw->esw_funcs.num_vfs; - flows = kvcalloc(num_vfs, sizeof(*flows), GFP_KERNEL); - if (!flows) - return -ENOMEM; spec = kvzalloc(sizeof(*spec), GFP_KERNEL); - if (!spec) { - err = -ENOMEM; - goto alloc_err; - } + if (!spec) + return ERR_PTR(-ENOMEM); MLX5_SET(fte_match_param, spec->match_criteria, misc_parameters_2.metadata_reg_c_0, mlx5_eswitch_get_vport_metadata_mask()); @@ -1116,34 +1091,18 @@ mlx5_eswitch_add_send_to_vport_meta_rules(struct mlx5_eswitch *esw) dest.type = MLX5_FLOW_DESTINATION_TYPE_VPORT; flow_act.action = MLX5_FLOW_CONTEXT_ACTION_FWD_DEST; - mlx5_esw_for_each_vf_vport(esw, i, vport, num_vfs) { - vport_num = vport->vport; - MLX5_SET(fte_match_param, spec->match_value, misc_parameters_2.metadata_reg_c_0, - mlx5_eswitch_get_vport_metadata_for_match(esw, vport_num)); - dest.vport.num = vport_num; - - flow_rule = mlx5_add_flow_rules(esw->fdb_table.offloads.slow_fdb, - spec, &flow_act, &dest, 1); - if (IS_ERR(flow_rule)) { - err = PTR_ERR(flow_rule); - esw_warn(esw->dev, "FDB: Failed to add send to vport meta rule idx %d, err %ld\n", - rule_idx, PTR_ERR(flow_rule)); - goto rule_err; - } - flows[rule_idx++] = flow_rule; - } + MLX5_SET(fte_match_param, spec->match_value, misc_parameters_2.metadata_reg_c_0, + mlx5_eswitch_get_vport_metadata_for_match(esw, vport_num)); + dest.vport.num = vport_num; - esw->fdb_table.offloads.send_to_vport_meta_rules = flows; - kvfree(spec); - return 0; + flow_rule = mlx5_add_flow_rules(esw->fdb_table.offloads.slow_fdb, + spec, &flow_act, &dest, 1); + if (IS_ERR(flow_rule)) + esw_warn(esw->dev, "FDB: Failed to add send to vport meta rule vport %d, err %ld\n", + vport_num, PTR_ERR(flow_rule)); -rule_err: - while (--rule_idx >= 0) - mlx5_del_flow_rules(flows[rule_idx]); kvfree(spec); -alloc_err: - kvfree(flows); - return err; + return flow_rule; } static bool mlx5_eswitch_reg_c1_loopback_supported(struct mlx5_eswitch *esw) @@ -1668,18 +1627,200 @@ esw_chains_destroy(struct mlx5_eswitch *esw, struct mlx5_fs_chains *chains) #endif +static int +esw_create_send_to_vport_group(struct mlx5_eswitch *esw, + struct mlx5_flow_table *fdb, + u32 *flow_group_in, + int *ix) +{ + int inlen = MLX5_ST_SZ_BYTES(create_flow_group_in); + struct mlx5_flow_group *g; + void *match_criteria; + int count, err = 0; + + memset(flow_group_in, 0, inlen); + + MLX5_SET(create_flow_group_in, flow_group_in, match_criteria_enable, + MLX5_MATCH_MISC_PARAMETERS); + + match_criteria = MLX5_ADDR_OF(create_flow_group_in, flow_group_in, match_criteria); + + MLX5_SET_TO_ONES(fte_match_param, match_criteria, misc_parameters.source_sqn); + MLX5_SET_TO_ONES(fte_match_param, match_criteria, misc_parameters.source_port); + if (MLX5_CAP_ESW(esw->dev, merged_eswitch)) { + MLX5_SET_TO_ONES(fte_match_param, match_criteria, + misc_parameters.source_eswitch_owner_vhca_id); + MLX5_SET(create_flow_group_in, flow_group_in, + source_eswitch_owner_vhca_id_valid, 1); + } + + /* See comment at table_size calculation */ + count = MLX5_MAX_PORTS * (esw->total_vports * MAX_SQ_NVPORTS + MAX_PF_SQ); + MLX5_SET(create_flow_group_in, flow_group_in, start_flow_index, 0); + MLX5_SET(create_flow_group_in, flow_group_in, end_flow_index, *ix + count - 1); + *ix += count; + + g = mlx5_create_flow_group(fdb, flow_group_in); + if (IS_ERR(g)) { + err = PTR_ERR(g); + esw_warn(esw->dev, "Failed to create send-to-vport flow group err(%d)\n", err); + goto out; + } + esw->fdb_table.offloads.send_to_vport_grp = g; + +out: + return err; +} + +static int +esw_create_meta_send_to_vport_group(struct mlx5_eswitch *esw, + struct mlx5_flow_table *fdb, + u32 *flow_group_in, + int *ix) +{ + int inlen = MLX5_ST_SZ_BYTES(create_flow_group_in); + struct mlx5_flow_group *g; + void *match_criteria; + int err = 0; + + if (!esw_src_port_rewrite_supported(esw)) + return 0; + + memset(flow_group_in, 0, inlen); + + MLX5_SET(create_flow_group_in, flow_group_in, match_criteria_enable, + MLX5_MATCH_MISC_PARAMETERS_2); + + match_criteria = MLX5_ADDR_OF(create_flow_group_in, flow_group_in, match_criteria); + + MLX5_SET(fte_match_param, match_criteria, + misc_parameters_2.metadata_reg_c_0, + mlx5_eswitch_get_vport_metadata_mask()); + MLX5_SET(fte_match_param, match_criteria, + misc_parameters_2.metadata_reg_c_1, ESW_TUN_MASK); + + MLX5_SET(create_flow_group_in, flow_group_in, start_flow_index, *ix); + MLX5_SET(create_flow_group_in, flow_group_in, + end_flow_index, *ix + esw->total_vports - 1); + *ix += esw->total_vports; + + g = mlx5_create_flow_group(fdb, flow_group_in); + if (IS_ERR(g)) { + err = PTR_ERR(g); + esw_warn(esw->dev, + "Failed to create send-to-vport meta flow group err(%d)\n", err); + goto send_vport_meta_err; + } + esw->fdb_table.offloads.send_to_vport_meta_grp = g; + + return 0; + +send_vport_meta_err: + return err; +} + +static int +esw_create_peer_esw_miss_group(struct mlx5_eswitch *esw, + struct mlx5_flow_table *fdb, + u32 *flow_group_in, + int *ix) +{ + int inlen = MLX5_ST_SZ_BYTES(create_flow_group_in); + struct mlx5_flow_group *g; + void *match_criteria; + int err = 0; + + if (!MLX5_CAP_ESW(esw->dev, merged_eswitch)) + return 0; + + memset(flow_group_in, 0, inlen); + + esw_set_flow_group_source_port(esw, flow_group_in); + + if (!mlx5_eswitch_vport_match_metadata_enabled(esw)) { + match_criteria = MLX5_ADDR_OF(create_flow_group_in, + flow_group_in, + match_criteria); + + MLX5_SET_TO_ONES(fte_match_param, match_criteria, + misc_parameters.source_eswitch_owner_vhca_id); + + MLX5_SET(create_flow_group_in, flow_group_in, + source_eswitch_owner_vhca_id_valid, 1); + } + + MLX5_SET(create_flow_group_in, flow_group_in, start_flow_index, *ix); + MLX5_SET(create_flow_group_in, flow_group_in, end_flow_index, + *ix + esw->total_vports - 1); + *ix += esw->total_vports; + + g = mlx5_create_flow_group(fdb, flow_group_in); + if (IS_ERR(g)) { + err = PTR_ERR(g); + esw_warn(esw->dev, "Failed to create peer miss flow group err(%d)\n", err); + goto out; + } + esw->fdb_table.offloads.peer_miss_grp = g; + +out: + return err; +} + +static int +esw_create_miss_group(struct mlx5_eswitch *esw, + struct mlx5_flow_table *fdb, + u32 *flow_group_in, + int *ix) +{ + int inlen = MLX5_ST_SZ_BYTES(create_flow_group_in); + struct mlx5_flow_group *g; + void *match_criteria; + int err = 0; + u8 *dmac; + + memset(flow_group_in, 0, inlen); + + MLX5_SET(create_flow_group_in, flow_group_in, match_criteria_enable, + MLX5_MATCH_OUTER_HEADERS); + match_criteria = MLX5_ADDR_OF(create_flow_group_in, flow_group_in, + match_criteria); + dmac = MLX5_ADDR_OF(fte_match_param, match_criteria, + outer_headers.dmac_47_16); + dmac[0] = 0x01; + + MLX5_SET(create_flow_group_in, flow_group_in, start_flow_index, *ix); + MLX5_SET(create_flow_group_in, flow_group_in, end_flow_index, + *ix + MLX5_ESW_MISS_FLOWS); + + g = mlx5_create_flow_group(fdb, flow_group_in); + if (IS_ERR(g)) { + err = PTR_ERR(g); + esw_warn(esw->dev, "Failed to create miss flow group err(%d)\n", err); + goto miss_err; + } + esw->fdb_table.offloads.miss_grp = g; + + err = esw_add_fdb_miss_rule(esw); + if (err) + goto miss_rule_err; + + return 0; + +miss_rule_err: + mlx5_destroy_flow_group(esw->fdb_table.offloads.miss_grp); +miss_err: + return err; +} + static int esw_create_offloads_fdb_tables(struct mlx5_eswitch *esw) { int inlen = MLX5_ST_SZ_BYTES(create_flow_group_in); struct mlx5_flow_table_attr ft_attr = {}; - int num_vfs, table_size, ix, err = 0; struct mlx5_core_dev *dev = esw->dev; struct mlx5_flow_namespace *root_ns; struct mlx5_flow_table *fdb = NULL; + int table_size, ix = 0, err = 0; u32 flags = 0, *flow_group_in; - struct mlx5_flow_group *g; - void *match_criteria; - u8 *dmac; esw_debug(esw->dev, "Create offloads FDB Tables\n"); @@ -1713,7 +1854,7 @@ static int esw_create_offloads_fdb_tables(struct mlx5_eswitch *esw) * total vports of the peer (currently is also uses esw->total_vports). */ table_size = MLX5_MAX_PORTS * (esw->total_vports * MAX_SQ_NVPORTS + MAX_PF_SQ) + - MLX5_ESW_MISS_FLOWS + esw->total_vports + esw->esw_funcs.num_vfs; + esw->total_vports * 2 + MLX5_ESW_MISS_FLOWS; /* create the slow path fdb with encap set, so further table instances * can be created at run time while VFs are probed if the FW allows that. @@ -1754,139 +1895,29 @@ static int esw_create_offloads_fdb_tables(struct mlx5_eswitch *esw) goto fdb_chains_err; } - /* create send-to-vport group */ - MLX5_SET(create_flow_group_in, flow_group_in, match_criteria_enable, - MLX5_MATCH_MISC_PARAMETERS); - - match_criteria = MLX5_ADDR_OF(create_flow_group_in, flow_group_in, match_criteria); - - MLX5_SET_TO_ONES(fte_match_param, match_criteria, misc_parameters.source_sqn); - MLX5_SET_TO_ONES(fte_match_param, match_criteria, misc_parameters.source_port); - if (MLX5_CAP_ESW(esw->dev, merged_eswitch)) { - MLX5_SET_TO_ONES(fte_match_param, match_criteria, - misc_parameters.source_eswitch_owner_vhca_id); - MLX5_SET(create_flow_group_in, flow_group_in, - source_eswitch_owner_vhca_id_valid, 1); - } - - /* See comment above table_size calculation */ - ix = MLX5_MAX_PORTS * (esw->total_vports * MAX_SQ_NVPORTS + MAX_PF_SQ); - MLX5_SET(create_flow_group_in, flow_group_in, start_flow_index, 0); - MLX5_SET(create_flow_group_in, flow_group_in, end_flow_index, ix - 1); - - g = mlx5_create_flow_group(fdb, flow_group_in); - if (IS_ERR(g)) { - err = PTR_ERR(g); - esw_warn(dev, "Failed to create send-to-vport flow group err(%d)\n", err); + err = esw_create_send_to_vport_group(esw, fdb, flow_group_in, &ix); + if (err) goto send_vport_err; - } - esw->fdb_table.offloads.send_to_vport_grp = g; - - if (esw_src_port_rewrite_supported(esw)) { - /* meta send to vport */ - memset(flow_group_in, 0, inlen); - MLX5_SET(create_flow_group_in, flow_group_in, match_criteria_enable, - MLX5_MATCH_MISC_PARAMETERS_2); - - match_criteria = MLX5_ADDR_OF(create_flow_group_in, flow_group_in, match_criteria); - - MLX5_SET(fte_match_param, match_criteria, - misc_parameters_2.metadata_reg_c_0, - mlx5_eswitch_get_vport_metadata_mask()); - MLX5_SET(fte_match_param, match_criteria, - misc_parameters_2.metadata_reg_c_1, ESW_TUN_MASK); - - num_vfs = esw->esw_funcs.num_vfs; - if (num_vfs) { - MLX5_SET(create_flow_group_in, flow_group_in, start_flow_index, ix); - MLX5_SET(create_flow_group_in, flow_group_in, - end_flow_index, ix + num_vfs - 1); - ix += num_vfs; - - g = mlx5_create_flow_group(fdb, flow_group_in); - if (IS_ERR(g)) { - err = PTR_ERR(g); - esw_warn(dev, "Failed to create send-to-vport meta flow group err(%d)\n", - err); - goto send_vport_meta_err; - } - esw->fdb_table.offloads.send_to_vport_meta_grp = g; - - err = mlx5_eswitch_add_send_to_vport_meta_rules(esw); - if (err) - goto meta_rule_err; - } - } - - if (MLX5_CAP_ESW(esw->dev, merged_eswitch)) { - /* create peer esw miss group */ - memset(flow_group_in, 0, inlen); - - esw_set_flow_group_source_port(esw, flow_group_in); - - if (!mlx5_eswitch_vport_match_metadata_enabled(esw)) { - match_criteria = MLX5_ADDR_OF(create_flow_group_in, - flow_group_in, - match_criteria); - - MLX5_SET_TO_ONES(fte_match_param, match_criteria, - misc_parameters.source_eswitch_owner_vhca_id); - - MLX5_SET(create_flow_group_in, flow_group_in, - source_eswitch_owner_vhca_id_valid, 1); - } - - MLX5_SET(create_flow_group_in, flow_group_in, start_flow_index, ix); - MLX5_SET(create_flow_group_in, flow_group_in, end_flow_index, - ix + esw->total_vports - 1); - ix += esw->total_vports; - - g = mlx5_create_flow_group(fdb, flow_group_in); - if (IS_ERR(g)) { - err = PTR_ERR(g); - esw_warn(dev, "Failed to create peer miss flow group err(%d)\n", err); - goto peer_miss_err; - } - esw->fdb_table.offloads.peer_miss_grp = g; - } - /* create miss group */ - memset(flow_group_in, 0, inlen); - MLX5_SET(create_flow_group_in, flow_group_in, match_criteria_enable, - MLX5_MATCH_OUTER_HEADERS); - match_criteria = MLX5_ADDR_OF(create_flow_group_in, flow_group_in, - match_criteria); - dmac = MLX5_ADDR_OF(fte_match_param, match_criteria, - outer_headers.dmac_47_16); - dmac[0] = 0x01; - - MLX5_SET(create_flow_group_in, flow_group_in, start_flow_index, ix); - MLX5_SET(create_flow_group_in, flow_group_in, end_flow_index, - ix + MLX5_ESW_MISS_FLOWS); + err = esw_create_meta_send_to_vport_group(esw, fdb, flow_group_in, &ix); + if (err) + goto send_vport_meta_err; - g = mlx5_create_flow_group(fdb, flow_group_in); - if (IS_ERR(g)) { - err = PTR_ERR(g); - esw_warn(dev, "Failed to create miss flow group err(%d)\n", err); - goto miss_err; - } - esw->fdb_table.offloads.miss_grp = g; + err = esw_create_peer_esw_miss_group(esw, fdb, flow_group_in, &ix); + if (err) + goto peer_miss_err; - err = esw_add_fdb_miss_rule(esw); + err = esw_create_miss_group(esw, fdb, flow_group_in, &ix); if (err) - goto miss_rule_err; + goto miss_err; kvfree(flow_group_in); return 0; -miss_rule_err: - mlx5_destroy_flow_group(esw->fdb_table.offloads.miss_grp); miss_err: if (MLX5_CAP_ESW(esw->dev, merged_eswitch)) mlx5_destroy_flow_group(esw->fdb_table.offloads.peer_miss_grp); peer_miss_err: - mlx5_eswitch_del_send_to_vport_meta_rules(esw); -meta_rule_err: if (esw->fdb_table.offloads.send_to_vport_meta_grp) mlx5_destroy_flow_group(esw->fdb_table.offloads.send_to_vport_meta_grp); send_vport_meta_err: @@ -1913,7 +1944,6 @@ static void esw_destroy_offloads_fdb_tables(struct mlx5_eswitch *esw) esw_debug(esw->dev, "Destroy offloads FDB Tables\n"); mlx5_del_flow_rules(esw->fdb_table.offloads.miss_rule_multi); mlx5_del_flow_rules(esw->fdb_table.offloads.miss_rule_uni); - mlx5_eswitch_del_send_to_vport_meta_rules(esw); mlx5_destroy_flow_group(esw->fdb_table.offloads.send_to_vport_grp); if (esw->fdb_table.offloads.send_to_vport_meta_grp) mlx5_destroy_flow_group(esw->fdb_table.offloads.send_to_vport_meta_grp); @@ -1931,7 +1961,7 @@ static void esw_destroy_offloads_fdb_tables(struct mlx5_eswitch *esw) atomic64_set(&esw->user_count, 0); } -static int esw_get_offloads_ft_size(struct mlx5_eswitch *esw) +static int esw_get_nr_ft_offloads_steering_src_ports(struct mlx5_eswitch *esw) { int nvports; @@ -1956,7 +1986,8 @@ static int esw_create_offloads_table(struct mlx5_eswitch *esw) return -EOPNOTSUPP; } - ft_attr.max_fte = esw_get_offloads_ft_size(esw); + ft_attr.max_fte = esw_get_nr_ft_offloads_steering_src_ports(esw) + + MLX5_ESW_FT_OFFLOADS_DROP_RULE; ft_attr.prio = 1; ft_offloads = mlx5_create_flow_table(ns, &ft_attr); @@ -1985,7 +2016,7 @@ static int esw_create_vport_rx_group(struct mlx5_eswitch *esw) int nvports; int err = 0; - nvports = esw_get_offloads_ft_size(esw); + nvports = esw_get_nr_ft_offloads_steering_src_ports(esw); flow_group_in = kvzalloc(inlen, GFP_KERNEL); if (!flow_group_in) return -ENOMEM; @@ -2015,6 +2046,52 @@ static void esw_destroy_vport_rx_group(struct mlx5_eswitch *esw) mlx5_destroy_flow_group(esw->offloads.vport_rx_group); } +static int esw_create_vport_rx_drop_rule_index(struct mlx5_eswitch *esw) +{ + /* ft_offloads table is enlarged by MLX5_ESW_FT_OFFLOADS_DROP_RULE (1) + * for the drop rule, which is placed at the end of the table. + * So return the total of vport and int_port as rule index. + */ + return esw_get_nr_ft_offloads_steering_src_ports(esw); +} + +static int esw_create_vport_rx_drop_group(struct mlx5_eswitch *esw) +{ + int inlen = MLX5_ST_SZ_BYTES(create_flow_group_in); + struct mlx5_flow_group *g; + u32 *flow_group_in; + int flow_index; + int err = 0; + + flow_index = esw_create_vport_rx_drop_rule_index(esw); + + flow_group_in = kvzalloc(inlen, GFP_KERNEL); + if (!flow_group_in) + return -ENOMEM; + + MLX5_SET(create_flow_group_in, flow_group_in, start_flow_index, flow_index); + MLX5_SET(create_flow_group_in, flow_group_in, end_flow_index, flow_index); + + g = mlx5_create_flow_group(esw->offloads.ft_offloads, flow_group_in); + + if (IS_ERR(g)) { + err = PTR_ERR(g); + mlx5_core_warn(esw->dev, "Failed to create vport rx drop group err %d\n", err); + goto out; + } + + esw->offloads.vport_rx_drop_group = g; +out: + kvfree(flow_group_in); + return err; +} + +static void esw_destroy_vport_rx_drop_group(struct mlx5_eswitch *esw) +{ + if (esw->offloads.vport_rx_drop_group) + mlx5_destroy_flow_group(esw->offloads.vport_rx_drop_group); +} + struct mlx5_flow_handle * mlx5_eswitch_create_vport_rx_rule(struct mlx5_eswitch *esw, u16 vport, struct mlx5_flow_destination *dest) @@ -2063,6 +2140,32 @@ out: return flow_rule; } +static int esw_create_vport_rx_drop_rule(struct mlx5_eswitch *esw) +{ + struct mlx5_flow_act flow_act = {}; + struct mlx5_flow_handle *flow_rule; + + flow_act.action = MLX5_FLOW_CONTEXT_ACTION_DROP; + flow_rule = mlx5_add_flow_rules(esw->offloads.ft_offloads, NULL, + &flow_act, NULL, 0); + if (IS_ERR(flow_rule)) { + esw_warn(esw->dev, + "fs offloads: Failed to add vport rx drop rule err %ld\n", + PTR_ERR(flow_rule)); + return PTR_ERR(flow_rule); + } + + esw->offloads.vport_rx_drop_rule = flow_rule; + + return 0; +} + +static void esw_destroy_vport_rx_drop_rule(struct mlx5_eswitch *esw) +{ + if (esw->offloads.vport_rx_drop_rule) + mlx5_del_flow_rules(esw->offloads.vport_rx_drop_rule); +} + static int mlx5_eswitch_inline_mode_get(struct mlx5_eswitch *esw, u8 *mode) { u8 prev_mlx5_mode, mlx5_mode = MLX5_INLINE_MODE_L2; @@ -3063,8 +3166,20 @@ static int esw_offloads_steering_init(struct mlx5_eswitch *esw) if (err) goto create_fg_err; + err = esw_create_vport_rx_drop_group(esw); + if (err) + goto create_rx_drop_fg_err; + + err = esw_create_vport_rx_drop_rule(esw); + if (err) + goto create_rx_drop_rule_err; + return 0; +create_rx_drop_rule_err: + esw_destroy_vport_rx_drop_group(esw); +create_rx_drop_fg_err: + esw_destroy_vport_rx_group(esw); create_fg_err: esw_destroy_offloads_fdb_tables(esw); create_fdb_err: @@ -3082,6 +3197,8 @@ create_indir_err: static void esw_offloads_steering_cleanup(struct mlx5_eswitch *esw) { + esw_destroy_vport_rx_drop_rule(esw); + esw_destroy_vport_rx_drop_group(esw); esw_destroy_vport_rx_group(esw); esw_destroy_offloads_fdb_tables(esw); esw_destroy_restore_table(esw); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/events.c b/drivers/net/ethernet/mellanox/mlx5/core/events.c index a1ac3a654962..9459e56ee90a 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/events.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/events.c @@ -36,6 +36,7 @@ static struct mlx5_nb events_nbs_ref[] = { /* Events to be forwarded (as is) to mlx5 core interfaces (mlx5e/mlx5_ib) */ {.nb.notifier_call = forward_event, .event_type = MLX5_EVENT_TYPE_PORT_CHANGE }, {.nb.notifier_call = forward_event, .event_type = MLX5_EVENT_TYPE_GENERAL_EVENT }, + {.nb.notifier_call = forward_event, .event_type = MLX5_EVENT_TYPE_OBJECT_CHANGE }, /* QP/WQ resource events to forward */ {.nb.notifier_call = forward_event, .event_type = MLX5_EVENT_TYPE_DCT_DRAINED }, {.nb.notifier_call = forward_event, .event_type = MLX5_EVENT_TYPE_PATH_MIG }, @@ -132,6 +133,8 @@ static const char *eqe_type_str(u8 type) return "MLX5_EVENT_TYPE_MONITOR_COUNTER"; case MLX5_EVENT_TYPE_DEVICE_TRACER: return "MLX5_EVENT_TYPE_DEVICE_TRACER"; + case MLX5_EVENT_TYPE_OBJECT_CHANGE: + return "MLX5_EVENT_TYPE_OBJECT_CHANGE"; default: return "Unrecognized event"; } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c index e735e19461ba..32d4c967469c 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c @@ -577,7 +577,10 @@ static int mlx5_cmd_set_fte(struct mlx5_core_dev *dev, MLX5_SET(flow_context, in_flow_context, modify_header_id, fte->action.modify_hdr->id); - MLX5_SET(flow_context, in_flow_context, ipsec_obj_id, fte->action.ipsec_obj_id); + MLX5_SET(flow_context, in_flow_context, encrypt_decrypt_type, + fte->action.crypto.type); + MLX5_SET(flow_context, in_flow_context, encrypt_decrypt_obj_id, + fte->action.crypto.obj_id); vlan = MLX5_ADDR_OF(flow_context, in_flow_context, push_vlan); @@ -919,13 +922,15 @@ static int mlx5_cmd_modify_header_alloc(struct mlx5_flow_root_namespace *ns, max_actions = MLX5_CAP_ESW_FLOWTABLE_FDB(dev, max_modify_header_actions); table_type = FS_FT_FDB; break; + case MLX5_FLOW_NAMESPACE_KERNEL_RX_MACSEC: case MLX5_FLOW_NAMESPACE_KERNEL: case MLX5_FLOW_NAMESPACE_BYPASS: max_actions = MLX5_CAP_FLOWTABLE_NIC_RX(dev, max_modify_header_actions); table_type = FS_FT_NIC_RX; break; case MLX5_FLOW_NAMESPACE_EGRESS: - case MLX5_FLOW_NAMESPACE_EGRESS_KERNEL: + case MLX5_FLOW_NAMESPACE_EGRESS_IPSEC: + case MLX5_FLOW_NAMESPACE_EGRESS_MACSEC: max_actions = MLX5_CAP_FLOWTABLE_NIC_TX(dev, max_modify_header_actions); table_type = FS_FT_NIC_TX; break; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c index e3960cdf5131..d53749248fa0 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c @@ -104,6 +104,10 @@ #define BY_PASS_MIN_LEVEL (ETHTOOL_MIN_LEVEL + MLX5_BY_PASS_NUM_PRIOS +\ LEFTOVERS_NUM_PRIOS) +#define KERNEL_RX_MACSEC_NUM_PRIOS 1 +#define KERNEL_RX_MACSEC_NUM_LEVELS 2 +#define KERNEL_RX_MACSEC_MIN_LEVEL (BY_PASS_MIN_LEVEL + KERNEL_RX_MACSEC_NUM_PRIOS) + #define ETHTOOL_PRIO_NUM_LEVELS 1 #define ETHTOOL_NUM_PRIOS 11 #define ETHTOOL_MIN_LEVEL (KERNEL_MIN_LEVEL + ETHTOOL_NUM_PRIOS) @@ -126,11 +130,15 @@ #define LAG_PRIO_NUM_LEVELS 1 #define LAG_NUM_PRIOS 1 -#define LAG_MIN_LEVEL (OFFLOADS_MIN_LEVEL + 1) +#define LAG_MIN_LEVEL (OFFLOADS_MIN_LEVEL + KERNEL_RX_MACSEC_MIN_LEVEL + 1) #define KERNEL_TX_IPSEC_NUM_PRIOS 1 #define KERNEL_TX_IPSEC_NUM_LEVELS 1 -#define KERNEL_TX_MIN_LEVEL (KERNEL_TX_IPSEC_NUM_LEVELS) +#define KERNEL_TX_IPSEC_MIN_LEVEL (KERNEL_TX_IPSEC_NUM_LEVELS) + +#define KERNEL_TX_MACSEC_NUM_PRIOS 1 +#define KERNEL_TX_MACSEC_NUM_LEVELS 2 +#define KERNEL_TX_MACSEC_MIN_LEVEL (KERNEL_TX_IPSEC_MIN_LEVEL + KERNEL_TX_MACSEC_NUM_PRIOS) struct node_caps { size_t arr_sz; @@ -149,12 +157,16 @@ static struct init_tree_node { enum mlx5_flow_table_miss_action def_miss_action; } root_fs = { .type = FS_TYPE_NAMESPACE, - .ar_size = 7, + .ar_size = 8, .children = (struct init_tree_node[]){ ADD_PRIO(0, BY_PASS_MIN_LEVEL, 0, FS_CHAINING_CAPS, ADD_NS(MLX5_FLOW_TABLE_MISS_ACTION_DEF, ADD_MULTIPLE_PRIO(MLX5_BY_PASS_NUM_PRIOS, BY_PASS_PRIO_NUM_LEVELS))), + ADD_PRIO(0, KERNEL_RX_MACSEC_MIN_LEVEL, 0, FS_CHAINING_CAPS, + ADD_NS(MLX5_FLOW_TABLE_MISS_ACTION_DEF, + ADD_MULTIPLE_PRIO(KERNEL_RX_MACSEC_NUM_PRIOS, + KERNEL_RX_MACSEC_NUM_LEVELS))), ADD_PRIO(0, LAG_MIN_LEVEL, 0, FS_CHAINING_CAPS, ADD_NS(MLX5_FLOW_TABLE_MISS_ACTION_DEF, ADD_MULTIPLE_PRIO(LAG_NUM_PRIOS, @@ -186,18 +198,23 @@ static struct init_tree_node { static struct init_tree_node egress_root_fs = { .type = FS_TYPE_NAMESPACE, - .ar_size = 2, + .ar_size = 3, .children = (struct init_tree_node[]) { ADD_PRIO(0, MLX5_BY_PASS_NUM_PRIOS, 0, FS_CHAINING_CAPS_EGRESS, ADD_NS(MLX5_FLOW_TABLE_MISS_ACTION_DEF, ADD_MULTIPLE_PRIO(MLX5_BY_PASS_NUM_PRIOS, BY_PASS_PRIO_NUM_LEVELS))), - ADD_PRIO(0, KERNEL_TX_MIN_LEVEL, 0, + ADD_PRIO(0, KERNEL_TX_IPSEC_MIN_LEVEL, 0, FS_CHAINING_CAPS_EGRESS, ADD_NS(MLX5_FLOW_TABLE_MISS_ACTION_DEF, ADD_MULTIPLE_PRIO(KERNEL_TX_IPSEC_NUM_PRIOS, KERNEL_TX_IPSEC_NUM_LEVELS))), + ADD_PRIO(0, KERNEL_TX_MACSEC_MIN_LEVEL, 0, + FS_CHAINING_CAPS_EGRESS, + ADD_NS(MLX5_FLOW_TABLE_MISS_ACTION_DEF, + ADD_MULTIPLE_PRIO(KERNEL_TX_MACSEC_NUM_PRIOS, + KERNEL_TX_MACSEC_NUM_LEVELS))), } }; @@ -2269,6 +2286,7 @@ static bool is_nic_rx_ns(enum mlx5_flow_namespace_type type) { switch (type) { case MLX5_FLOW_NAMESPACE_BYPASS: + case MLX5_FLOW_NAMESPACE_KERNEL_RX_MACSEC: case MLX5_FLOW_NAMESPACE_LAG: case MLX5_FLOW_NAMESPACE_OFFLOADS: case MLX5_FLOW_NAMESPACE_ETHTOOL: @@ -2315,7 +2333,8 @@ struct mlx5_flow_namespace *mlx5_get_flow_namespace(struct mlx5_core_dev *dev, prio = FDB_BYPASS_PATH; break; case MLX5_FLOW_NAMESPACE_EGRESS: - case MLX5_FLOW_NAMESPACE_EGRESS_KERNEL: + case MLX5_FLOW_NAMESPACE_EGRESS_IPSEC: + case MLX5_FLOW_NAMESPACE_EGRESS_MACSEC: root_ns = steering->egress_root_ns; prio = type - MLX5_FLOW_NAMESPACE_EGRESS; break; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fw.c b/drivers/net/ethernet/mellanox/mlx5/core/fw.c index 079fa44ada71..f34e758a2f1f 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/fw.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/fw.c @@ -273,6 +273,19 @@ int mlx5_query_hca_caps(struct mlx5_core_dev *dev) return err; } + if (MLX5_CAP_GEN_64(dev, general_obj_types) & + MLX5_GENERAL_OBJ_TYPES_CAP_MACSEC_OFFLOAD) { + err = mlx5_core_get_caps(dev, MLX5_CAP_MACSEC); + if (err) + return err; + } + + if (MLX5_CAP_GEN(dev, adv_virtualization)) { + err = mlx5_core_get_caps(dev, MLX5_CAP_ADV_VIRTUALIZATION); + if (err) + return err; + } + return 0; } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/health.c b/drivers/net/ethernet/mellanox/mlx5/core/health.c index 2cf2c9948446..86ed87d704f7 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/health.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/health.c @@ -601,7 +601,7 @@ static void mlx5_fw_reporter_err_work(struct work_struct *work) fw_reporter_ctx.miss_counter = health->miss_counter; if (fw_reporter_ctx.err_synd) { devlink_health_report(health->fw_reporter, - "FW syndrom reported", &fw_reporter_ctx); + "FW syndrome reported", &fw_reporter_ctx); return; } if (fw_reporter_ctx.miss_counter) @@ -702,11 +702,25 @@ static const struct devlink_health_reporter_ops mlx5_fw_fatal_reporter_ops = { .dump = mlx5_fw_fatal_reporter_dump, }; -#define MLX5_REPORTER_FW_GRACEFUL_PERIOD 1200000 +#define MLX5_FW_REPORTER_ECPF_GRACEFUL_PERIOD 180000 +#define MLX5_FW_REPORTER_PF_GRACEFUL_PERIOD 60000 +#define MLX5_FW_REPORTER_VF_GRACEFUL_PERIOD 30000 +#define MLX5_FW_REPORTER_DEFAULT_GRACEFUL_PERIOD MLX5_FW_REPORTER_VF_GRACEFUL_PERIOD + static void mlx5_fw_reporters_create(struct mlx5_core_dev *dev) { struct mlx5_core_health *health = &dev->priv.health; struct devlink *devlink = priv_to_devlink(dev); + u64 grace_period; + + if (mlx5_core_is_ecpf(dev)) { + grace_period = MLX5_FW_REPORTER_ECPF_GRACEFUL_PERIOD; + } else if (mlx5_core_is_pf(dev)) { + grace_period = MLX5_FW_REPORTER_PF_GRACEFUL_PERIOD; + } else { + /* VF or SF */ + grace_period = MLX5_FW_REPORTER_DEFAULT_GRACEFUL_PERIOD; + } health->fw_reporter = devlink_health_reporter_create(devlink, &mlx5_fw_reporter_ops, @@ -718,7 +732,7 @@ static void mlx5_fw_reporters_create(struct mlx5_core_dev *dev) health->fw_fatal_reporter = devlink_health_reporter_create(devlink, &mlx5_fw_fatal_reporter_ops, - MLX5_REPORTER_FW_GRACEFUL_PERIOD, + grace_period, dev); if (IS_ERR(health->fw_fatal_reporter)) mlx5_core_warn(dev, "Failed to create fw fatal reporter, err = %ld\n", @@ -843,9 +857,6 @@ void mlx5_start_health_poll(struct mlx5_core_dev *dev) health->timer.expires = jiffies + msecs_to_jiffies(poll_interval_ms); add_timer(&health->timer); - - if (mlx5_core_is_pf(dev) && MLX5_CAP_MCAM_REG(dev, mrtc)) - queue_delayed_work(health->wq, &health->update_fw_log_ts_work, 0); } void mlx5_stop_health_poll(struct mlx5_core_dev *dev, bool disable_health) @@ -862,6 +873,14 @@ void mlx5_stop_health_poll(struct mlx5_core_dev *dev, bool disable_health) del_timer_sync(&health->timer); } +void mlx5_start_health_fw_log_up(struct mlx5_core_dev *dev) +{ + struct mlx5_core_health *health = &dev->priv.health; + + if (mlx5_core_is_pf(dev) && MLX5_CAP_MCAM_REG(dev, mrtc)) + queue_delayed_work(health->wq, &health->update_fw_log_ts_work, 0); +} + void mlx5_drain_health_wq(struct mlx5_core_dev *dev) { struct mlx5_core_health *health = &dev->priv.health; @@ -875,13 +894,6 @@ void mlx5_drain_health_wq(struct mlx5_core_dev *dev) cancel_work_sync(&health->fatal_report_work); } -void mlx5_health_flush(struct mlx5_core_dev *dev) -{ - struct mlx5_core_health *health = &dev->priv.health; - - flush_workqueue(health->wq); -} - void mlx5_health_cleanup(struct mlx5_core_dev *dev) { struct mlx5_core_health *health = &dev->priv.health; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ethtool.c b/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ethtool.c index ac3757beaea2..c247cca154e9 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ethtool.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ethtool.c @@ -32,6 +32,7 @@ #include "en.h" #include "ipoib.h" +#include "en/fs_ethtool.h" static void mlx5i_get_drvinfo(struct net_device *dev, struct ethtool_drvinfo *drvinfo) @@ -39,7 +40,7 @@ static void mlx5i_get_drvinfo(struct net_device *dev, struct mlx5e_priv *priv = mlx5i_epriv(dev); mlx5e_ethtool_get_drvinfo(priv, drvinfo); - strlcpy(drvinfo->driver, KBUILD_MODNAME "[ib_ipoib]", + strscpy(drvinfo->driver, KBUILD_MODNAME "[ib_ipoib]", sizeof(drvinfo->driver)); } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib.c b/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib.c index c02b7b08fb4c..4e3a75496dd9 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib.c @@ -35,6 +35,7 @@ #include "en.h" #include "en/params.h" #include "ipoib.h" +#include "en/fs_ethtool.h" #define IB_DEFAULT_Q_KEY 0xb1b #define MLX5I_PARAMS_DEFAULT_LOG_RQ_SIZE 9 @@ -320,43 +321,47 @@ static void mlx5i_cleanup_tx(struct mlx5e_priv *priv) static int mlx5i_create_flow_steering(struct mlx5e_priv *priv) { + struct mlx5_flow_namespace *ns = + mlx5_get_flow_namespace(priv->mdev, MLX5_FLOW_NAMESPACE_KERNEL); int err; - priv->fs->ns = mlx5_get_flow_namespace(priv->mdev, - MLX5_FLOW_NAMESPACE_KERNEL); - if (!priv->fs->ns) + if (!ns) return -EINVAL; - err = mlx5e_arfs_create_tables(priv); + mlx5e_fs_set_ns(priv->fs, ns, false); + err = mlx5e_arfs_create_tables(priv->fs, priv->rx_res, + !!(priv->netdev->hw_features & NETIF_F_NTUPLE)); if (err) { netdev_err(priv->netdev, "Failed to create arfs tables, err=%d\n", err); priv->netdev->hw_features &= ~NETIF_F_NTUPLE; } - err = mlx5e_create_ttc_table(priv); + err = mlx5e_create_ttc_table(priv->fs, priv->rx_res); if (err) { netdev_err(priv->netdev, "Failed to create ttc table, err=%d\n", err); goto err_destroy_arfs_tables; } - mlx5e_ethtool_init_steering(priv); + mlx5e_ethtool_init_steering(priv->fs); return 0; err_destroy_arfs_tables: - mlx5e_arfs_destroy_tables(priv); + mlx5e_arfs_destroy_tables(priv->fs, + !!(priv->netdev->hw_features & NETIF_F_NTUPLE)); return err; } static void mlx5i_destroy_flow_steering(struct mlx5e_priv *priv) { - mlx5e_destroy_ttc_table(priv); - mlx5e_arfs_destroy_tables(priv); - mlx5e_ethtool_cleanup_steering(priv); + mlx5e_destroy_ttc_table(priv->fs); + mlx5e_arfs_destroy_tables(priv->fs, + !!(priv->netdev->hw_features & NETIF_F_NTUPLE)); + mlx5e_ethtool_cleanup_steering(priv->fs); } static int mlx5i_init_rx(struct mlx5e_priv *priv) @@ -458,7 +463,6 @@ static const struct mlx5e_profile mlx5i_nic_profile = { .update_carrier = NULL, /* no HW update in IB link */ .rx_handlers = &mlx5i_rx_handlers, .max_tc = MLX5I_MAX_NUM_TC, - .rq_groups = MLX5E_NUM_RQ_GROUPS(REGULAR), .stats_grps = mlx5i_stats_grps, .stats_grps_num = mlx5i_stats_grps_num, }; 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 0b86e78dbc0e..0227a521d301 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib_vlan.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib_vlan.c @@ -349,7 +349,6 @@ static const struct mlx5e_profile mlx5i_pkey_nic_profile = { .update_stats = NULL, .rx_handlers = &mlx5i_rx_handlers, .max_tc = MLX5I_MAX_NUM_TC, - .rq_groups = MLX5E_NUM_RQ_GROUPS(REGULAR), }; const struct mlx5e_profile *mlx5i_pkey_get_profile(void) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/lag/lag.c b/drivers/net/ethernet/mellanox/mlx5/core/lag/lag.c index 065102278cb8..a9f4ede4a9bf 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/lag/lag.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/lag/lag.c @@ -65,6 +65,21 @@ static int get_port_sel_mode(enum mlx5_lag_mode mode, unsigned long flags) return MLX5_LAG_PORT_SELECT_MODE_QUEUE_AFFINITY; } +static u8 lag_active_port_bits(struct mlx5_lag *ldev) +{ + u8 enabled_ports[MLX5_MAX_PORTS] = {}; + u8 active_port = 0; + int num_enabled; + int idx; + + mlx5_infer_tx_enabled(&ldev->tracker, ldev->ports, enabled_ports, + &num_enabled); + for (idx = 0; idx < num_enabled; idx++) + active_port |= BIT_MASK(enabled_ports[idx]); + + return active_port; +} + static int mlx5_cmd_create_lag(struct mlx5_core_dev *dev, u8 *ports, int mode, unsigned long flags) { @@ -77,9 +92,21 @@ static int mlx5_cmd_create_lag(struct mlx5_core_dev *dev, u8 *ports, int mode, lag_ctx = MLX5_ADDR_OF(create_lag_in, in, ctx); MLX5_SET(create_lag_in, in, opcode, MLX5_CMD_OP_CREATE_LAG); MLX5_SET(lagc, lag_ctx, fdb_selection_mode, fdb_sel_mode); - if (port_sel_mode == MLX5_LAG_PORT_SELECT_MODE_QUEUE_AFFINITY) { + + switch (port_sel_mode) { + case MLX5_LAG_PORT_SELECT_MODE_QUEUE_AFFINITY: MLX5_SET(lagc, lag_ctx, tx_remap_affinity_1, ports[0]); MLX5_SET(lagc, lag_ctx, tx_remap_affinity_2, ports[1]); + break; + case MLX5_LAG_PORT_SELECT_MODE_PORT_SELECT_FT: + if (!MLX5_CAP_PORT_SELECTION(dev, port_select_flow_table_bypass)) + break; + + MLX5_SET(lagc, lag_ctx, active_port, + lag_active_port_bits(mlx5_lag_dev(dev))); + break; + default: + break; } MLX5_SET(lagc, lag_ctx, port_select_mode, port_sel_mode); @@ -386,12 +413,37 @@ static void mlx5_lag_drop_rule_setup(struct mlx5_lag *ldev, } } +static int mlx5_cmd_modify_active_port(struct mlx5_core_dev *dev, u8 ports) +{ + u32 in[MLX5_ST_SZ_DW(modify_lag_in)] = {}; + void *lag_ctx; + + lag_ctx = MLX5_ADDR_OF(modify_lag_in, in, ctx); + + MLX5_SET(modify_lag_in, in, opcode, MLX5_CMD_OP_MODIFY_LAG); + MLX5_SET(modify_lag_in, in, field_select, 0x2); + + MLX5_SET(lagc, lag_ctx, active_port, ports); + + return mlx5_cmd_exec_in(dev, modify_lag, in); +} + static int _mlx5_modify_lag(struct mlx5_lag *ldev, u8 *ports) { struct mlx5_core_dev *dev0 = ldev->pf[MLX5_LAG_P1].dev; + u8 active_ports; + int ret; + + if (test_bit(MLX5_LAG_MODE_FLAG_HASH_BASED, &ldev->mode_flags)) { + ret = mlx5_lag_port_sel_modify(ldev, ports); + if (ret || + !MLX5_CAP_PORT_SELECTION(dev0, port_select_flow_table_bypass)) + return ret; - if (test_bit(MLX5_LAG_MODE_FLAG_HASH_BASED, &ldev->mode_flags)) - return mlx5_lag_port_sel_modify(ldev, ports); + active_ports = lag_active_port_bits(ldev); + + return mlx5_cmd_modify_active_port(dev0, active_ports); + } return mlx5_cmd_modify_lag(dev0, ldev->ports, ports); } @@ -432,21 +484,22 @@ void mlx5_modify_lag(struct mlx5_lag *ldev, mlx5_lag_drop_rule_setup(ldev, tracker); } -#define MLX5_LAG_ROCE_HASH_PORTS_SUPPORTED 4 static int mlx5_lag_set_port_sel_mode_roce(struct mlx5_lag *ldev, unsigned long *flags) { - struct lag_func *dev0 = &ldev->pf[MLX5_LAG_P1]; + struct mlx5_core_dev *dev0 = ldev->pf[MLX5_LAG_P1].dev; - if (ldev->ports == MLX5_LAG_ROCE_HASH_PORTS_SUPPORTED) { - /* Four ports are support only in hash mode */ - if (!MLX5_CAP_PORT_SELECTION(dev0->dev, port_select_flow_table)) - return -EINVAL; - set_bit(MLX5_LAG_MODE_FLAG_HASH_BASED, flags); + if (!MLX5_CAP_PORT_SELECTION(dev0, port_select_flow_table)) { if (ldev->ports > 2) - ldev->buckets = MLX5_LAG_MAX_HASH_BUCKETS; + return -EINVAL; + return 0; } + if (ldev->ports > 2) + ldev->buckets = MLX5_LAG_MAX_HASH_BUCKETS; + + set_bit(MLX5_LAG_MODE_FLAG_HASH_BASED, flags); + return 0; } @@ -1275,6 +1328,22 @@ bool mlx5_lag_is_active(struct mlx5_core_dev *dev) } EXPORT_SYMBOL(mlx5_lag_is_active); +bool mlx5_lag_mode_is_hash(struct mlx5_core_dev *dev) +{ + struct mlx5_lag *ldev; + unsigned long flags; + bool res = 0; + + spin_lock_irqsave(&lag_lock, flags); + ldev = mlx5_lag_dev(dev); + if (ldev) + res = test_bit(MLX5_LAG_MODE_FLAG_HASH_BASED, &ldev->mode_flags); + spin_unlock_irqrestore(&lag_lock, flags); + + return res; +} +EXPORT_SYMBOL(mlx5_lag_mode_is_hash); + bool mlx5_lag_is_master(struct mlx5_core_dev *dev) { struct mlx5_lag *ldev; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/lib/aso.h b/drivers/net/ethernet/mellanox/mlx5/core/lib/aso.h index b3bbf284fe71..d854e01d7fc5 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/lib/aso.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/lib/aso.h @@ -11,7 +11,9 @@ (DIV_ROUND_UP(sizeof(struct mlx5_aso_wqe), MLX5_SEND_WQE_BB)) #define MLX5_ASO_WQEBBS_DATA \ (DIV_ROUND_UP(sizeof(struct mlx5_aso_wqe_data), MLX5_SEND_WQE_BB)) +#define ASO_CTRL_READ_EN BIT(0) #define MLX5_WQE_CTRL_WQE_OPC_MOD_SHIFT 24 +#define MLX5_MACSEC_ASO_DS_CNT (DIV_ROUND_UP(sizeof(struct mlx5_aso_wqe), MLX5_SEND_WQE_DS)) struct mlx5_wqe_aso_ctrl_seg { __be32 va_h; @@ -70,6 +72,7 @@ enum { enum { MLX5_ACCESS_ASO_OPC_MOD_FLOW_METER = 0x2, + MLX5_ACCESS_ASO_OPC_MOD_MACSEC = 0x5, }; struct mlx5_aso; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/lib/clock.c b/drivers/net/ethernet/mellanox/mlx5/core/lib/clock.c index 91e806c1aa21..d3a9ae80fd30 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/lib/clock.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/lib/clock.c @@ -65,6 +65,8 @@ enum { MLX5_MTPPS_FS_TIME_STAMP = BIT(0x4), MLX5_MTPPS_FS_OUT_PULSE_DURATION = BIT(0x5), MLX5_MTPPS_FS_ENH_OUT_PER_ADJ = BIT(0x7), + MLX5_MTPPS_FS_NPPS_PERIOD = BIT(0x9), + MLX5_MTPPS_FS_OUT_PULSE_DURATION_NS = BIT(0xa), }; static bool mlx5_real_time_mode(struct mlx5_core_dev *mdev) @@ -72,6 +74,13 @@ static bool mlx5_real_time_mode(struct mlx5_core_dev *mdev) return (mlx5_is_real_time_rq(mdev) || mlx5_is_real_time_sq(mdev)); } +static bool mlx5_npps_real_time_supported(struct mlx5_core_dev *mdev) +{ + return (mlx5_real_time_mode(mdev) && + MLX5_CAP_MCAM_FEATURE(mdev, npps_period) && + MLX5_CAP_MCAM_FEATURE(mdev, out_pulse_duration_ns)); +} + static bool mlx5_modify_mtutc_allowed(struct mlx5_core_dev *mdev) { return MLX5_CAP_MCAM_FEATURE(mdev, ptpcyc2realtime_modify); @@ -459,9 +468,95 @@ static u64 perout_conf_internal_timer(struct mlx5_core_dev *mdev, s64 sec) return find_target_cycles(mdev, target_ns); } -static u64 perout_conf_real_time(s64 sec) +static u64 perout_conf_real_time(s64 sec, u32 nsec) +{ + return (u64)nsec | (u64)sec << 32; +} + +static int perout_conf_1pps(struct mlx5_core_dev *mdev, struct ptp_clock_request *rq, + u64 *time_stamp, bool real_time) +{ + struct timespec64 ts; + s64 ns; + + ts.tv_nsec = rq->perout.period.nsec; + ts.tv_sec = rq->perout.period.sec; + ns = timespec64_to_ns(&ts); + + if ((ns >> 1) != 500000000LL) + return -EINVAL; + + *time_stamp = real_time ? perout_conf_real_time(rq->perout.start.sec, 0) : + perout_conf_internal_timer(mdev, rq->perout.start.sec); + + return 0; +} + +#define MLX5_MAX_PULSE_DURATION (BIT(__mlx5_bit_sz(mtpps_reg, out_pulse_duration_ns)) - 1) +static int mlx5_perout_conf_out_pulse_duration(struct mlx5_core_dev *mdev, + struct ptp_clock_request *rq, + u32 *out_pulse_duration_ns) { - return (u64)sec << 32; + struct mlx5_pps *pps_info = &mdev->clock.pps_info; + u32 out_pulse_duration; + struct timespec64 ts; + + if (rq->perout.flags & PTP_PEROUT_DUTY_CYCLE) { + ts.tv_sec = rq->perout.on.sec; + ts.tv_nsec = rq->perout.on.nsec; + out_pulse_duration = (u32)timespec64_to_ns(&ts); + } else { + /* out_pulse_duration_ns should be up to 50% of the + * pulse period as default + */ + ts.tv_sec = rq->perout.period.sec; + ts.tv_nsec = rq->perout.period.nsec; + out_pulse_duration = (u32)timespec64_to_ns(&ts) >> 1; + } + + if (out_pulse_duration < pps_info->min_out_pulse_duration_ns || + out_pulse_duration > MLX5_MAX_PULSE_DURATION) { + mlx5_core_err(mdev, "NPPS pulse duration %u is not in [%llu, %lu]\n", + out_pulse_duration, pps_info->min_out_pulse_duration_ns, + MLX5_MAX_PULSE_DURATION); + return -EINVAL; + } + *out_pulse_duration_ns = out_pulse_duration; + + return 0; +} + +static int perout_conf_npps_real_time(struct mlx5_core_dev *mdev, struct ptp_clock_request *rq, + u32 *field_select, u32 *out_pulse_duration_ns, + u64 *period, u64 *time_stamp) +{ + struct mlx5_pps *pps_info = &mdev->clock.pps_info; + struct ptp_clock_time *time = &rq->perout.start; + struct timespec64 ts; + + ts.tv_sec = rq->perout.period.sec; + ts.tv_nsec = rq->perout.period.nsec; + if (timespec64_to_ns(&ts) < pps_info->min_npps_period) { + mlx5_core_err(mdev, "NPPS period is lower than minimal npps period %llu\n", + pps_info->min_npps_period); + return -EINVAL; + } + *period = perout_conf_real_time(rq->perout.period.sec, rq->perout.period.nsec); + + if (mlx5_perout_conf_out_pulse_duration(mdev, rq, out_pulse_duration_ns)) + return -EINVAL; + + *time_stamp = perout_conf_real_time(time->sec, time->nsec); + *field_select |= MLX5_MTPPS_FS_NPPS_PERIOD | + MLX5_MTPPS_FS_OUT_PULSE_DURATION_NS; + + return 0; +} + +static bool mlx5_perout_verify_flags(struct mlx5_core_dev *mdev, unsigned int flags) +{ + return ((!mlx5_npps_real_time_supported(mdev) && flags) || + (mlx5_npps_real_time_supported(mdev) && flags & ~PTP_PEROUT_DUTY_CYCLE)); } static int mlx5_perout_configure(struct ptp_clock_info *ptp, @@ -474,20 +569,20 @@ static int mlx5_perout_configure(struct ptp_clock_info *ptp, container_of(clock, struct mlx5_core_dev, clock); bool rt_mode = mlx5_real_time_mode(mdev); u32 in[MLX5_ST_SZ_DW(mtpps_reg)] = {0}; - struct timespec64 ts; + u32 out_pulse_duration_ns = 0; u32 field_select = 0; + u64 npps_period = 0; u64 time_stamp = 0; u8 pin_mode = 0; u8 pattern = 0; int pin = -1; int err = 0; - s64 ns; if (!MLX5_PPS_CAP(mdev)) return -EOPNOTSUPP; /* Reject requests with unsupported flags */ - if (rq->perout.flags) + if (mlx5_perout_verify_flags(mdev, rq->perout.flags)) return -EOPNOTSUPP; if (rq->perout.index >= clock->ptp_info.n_pins) @@ -500,29 +595,25 @@ static int mlx5_perout_configure(struct ptp_clock_info *ptp, if (on) { bool rt_mode = mlx5_real_time_mode(mdev); - s64 sec = rq->perout.start.sec; - - if (rq->perout.start.nsec) - return -EINVAL; pin_mode = MLX5_PIN_MODE_OUT; pattern = MLX5_OUT_PATTERN_PERIODIC; - ts.tv_sec = rq->perout.period.sec; - ts.tv_nsec = rq->perout.period.nsec; - ns = timespec64_to_ns(&ts); - if ((ns >> 1) != 500000000LL) + if (rt_mode && rq->perout.start.sec > U32_MAX) return -EINVAL; - if (rt_mode && sec > U32_MAX) - return -EINVAL; - - time_stamp = rt_mode ? perout_conf_real_time(sec) : - perout_conf_internal_timer(mdev, sec); - field_select |= MLX5_MTPPS_FS_PIN_MODE | MLX5_MTPPS_FS_PATTERN | MLX5_MTPPS_FS_TIME_STAMP; + + if (mlx5_npps_real_time_supported(mdev)) + err = perout_conf_npps_real_time(mdev, rq, &field_select, + &out_pulse_duration_ns, &npps_period, + &time_stamp); + else + err = perout_conf_1pps(mdev, rq, &time_stamp, rt_mode); + if (err) + return err; } MLX5_SET(mtpps_reg, in, pin, pin); @@ -531,7 +622,8 @@ static int mlx5_perout_configure(struct ptp_clock_info *ptp, MLX5_SET(mtpps_reg, in, enable, on); MLX5_SET64(mtpps_reg, in, time_stamp, time_stamp); MLX5_SET(mtpps_reg, in, field_select, field_select); - + MLX5_SET64(mtpps_reg, in, npps_period, npps_period); + MLX5_SET(mtpps_reg, in, out_pulse_duration_ns, out_pulse_duration_ns); err = mlx5_set_mtpps(mdev, in, sizeof(in)); if (err) return err; @@ -687,6 +779,13 @@ static void mlx5_get_pps_caps(struct mlx5_core_dev *mdev) clock->ptp_info.n_per_out = MLX5_GET(mtpps_reg, out, cap_max_num_of_pps_out_pins); + if (MLX5_CAP_MCAM_FEATURE(mdev, npps_period)) + clock->pps_info.min_npps_period = 1 << MLX5_GET(mtpps_reg, out, + cap_log_min_npps_period); + if (MLX5_CAP_MCAM_FEATURE(mdev, out_pulse_duration_ns)) + clock->pps_info.min_out_pulse_duration_ns = 1 << MLX5_GET(mtpps_reg, out, + cap_log_min_out_pulse_duration_ns); + clock->pps_info.pin_caps[0] = MLX5_GET(mtpps_reg, out, cap_pin_0_mode); clock->pps_info.pin_caps[1] = MLX5_GET(mtpps_reg, out, cap_pin_1_mode); clock->pps_info.pin_caps[2] = MLX5_GET(mtpps_reg, out, cap_pin_2_mode); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/lib/mlx5.h b/drivers/net/ethernet/mellanox/mlx5/core/lib/mlx5.h index 2f536c5d30b1..032adb21ad4b 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/lib/mlx5.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/lib/mlx5.h @@ -83,6 +83,7 @@ int mlx5_notifier_call_chain(struct mlx5_events *events, unsigned int event, voi enum { MLX5_ACCEL_OBJ_TLS_KEY = MLX5_GENERAL_OBJECT_TYPE_ENCRYPTION_KEY_TYPE_TLS, MLX5_ACCEL_OBJ_IPSEC_KEY = MLX5_GENERAL_OBJECT_TYPE_ENCRYPTION_KEY_TYPE_IPSEC, + MLX5_ACCEL_OBJ_MACSEC_KEY = MLX5_GENERAL_OBJECT_TYPE_ENCRYPTION_KEY_TYPE_MACSEC, }; int mlx5_create_encryption_key(struct mlx5_core_dev *mdev, diff --git a/drivers/net/ethernet/mellanox/mlx5/core/main.c b/drivers/net/ethernet/mellanox/mlx5/core/main.c index 89b2d9cea33f..0b459d841c3a 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/main.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/main.c @@ -671,6 +671,33 @@ static int handle_hca_cap_roce(struct mlx5_core_dev *dev, void *set_ctx) return err; } +static int handle_hca_cap_port_selection(struct mlx5_core_dev *dev, + void *set_ctx) +{ + void *set_hca_cap; + int err; + + if (!MLX5_CAP_GEN(dev, port_selection_cap)) + return 0; + + err = mlx5_core_get_caps(dev, MLX5_CAP_PORT_SELECTION); + if (err) + return err; + + if (MLX5_CAP_PORT_SELECTION(dev, port_select_flow_table_bypass) || + !MLX5_CAP_PORT_SELECTION_MAX(dev, port_select_flow_table_bypass)) + return 0; + + set_hca_cap = MLX5_ADDR_OF(set_hca_cap_in, set_ctx, capability); + memcpy(set_hca_cap, dev->caps.hca[MLX5_CAP_PORT_SELECTION]->cur, + MLX5_ST_SZ_BYTES(port_selection_cap)); + MLX5_SET(port_selection_cap, set_hca_cap, port_select_flow_table_bypass, 1); + + err = set_caps(dev, set_ctx, MLX5_SET_HCA_CAP_OP_MODE_PORT_SELECTION); + + return err; +} + static int set_hca_cap(struct mlx5_core_dev *dev) { int set_sz = MLX5_ST_SZ_BYTES(set_hca_cap_in); @@ -715,6 +742,13 @@ static int set_hca_cap(struct mlx5_core_dev *dev) goto out; } + memset(set_ctx, 0, set_sz); + err = handle_hca_cap_port_selection(dev, set_ctx); + if (err) { + mlx5_core_err(dev, "handle_hca_cap_port_selection failed\n"); + goto out; + } + out: kfree(set_ctx); return err; @@ -1058,7 +1092,7 @@ static void mlx5_cleanup_once(struct mlx5_core_dev *dev) mlx5_devcom_unregister_device(dev->priv.devcom); } -static int mlx5_function_setup(struct mlx5_core_dev *dev, u64 timeout) +static int mlx5_function_setup(struct mlx5_core_dev *dev, bool boot, u64 timeout) { int err; @@ -1096,10 +1130,12 @@ static int mlx5_function_setup(struct mlx5_core_dev *dev, u64 timeout) mlx5_cmd_set_state(dev, MLX5_CMDIF_STATE_UP); + mlx5_start_health_poll(dev); + err = mlx5_core_enable_hca(dev, 0); if (err) { mlx5_core_err(dev, "enable hca failed\n"); - goto err_cmd_cleanup; + goto stop_health_poll; } err = mlx5_core_set_issi(dev); @@ -1151,8 +1187,7 @@ static int mlx5_function_setup(struct mlx5_core_dev *dev, u64 timeout) mlx5_core_err(dev, "query hca failed\n"); goto reclaim_boot_pages; } - - mlx5_start_health_poll(dev); + mlx5_start_health_fw_log_up(dev); return 0; @@ -1160,6 +1195,8 @@ reclaim_boot_pages: mlx5_reclaim_startup_pages(dev); err_disable_hca: mlx5_core_disable_hca(dev, 0); +stop_health_poll: + mlx5_stop_health_poll(dev, boot); err_cmd_cleanup: mlx5_cmd_set_state(dev, MLX5_CMDIF_STATE_DOWN); mlx5_cmd_cleanup(dev); @@ -1171,7 +1208,6 @@ static int mlx5_function_teardown(struct mlx5_core_dev *dev, bool boot) { int err; - mlx5_stop_health_poll(dev, boot); err = mlx5_cmd_teardown_hca(dev); if (err) { mlx5_core_err(dev, "tear_down_hca failed, skip cleanup\n"); @@ -1179,6 +1215,7 @@ static int mlx5_function_teardown(struct mlx5_core_dev *dev, bool boot) } mlx5_reclaim_startup_pages(dev); mlx5_core_disable_hca(dev, 0); + mlx5_stop_health_poll(dev, boot); mlx5_cmd_set_state(dev, MLX5_CMDIF_STATE_DOWN); mlx5_cmd_cleanup(dev); @@ -1328,7 +1365,7 @@ int mlx5_init_one(struct mlx5_core_dev *dev) mutex_lock(&dev->intf_state_mutex); dev->state = MLX5_DEVICE_STATE_UP; - err = mlx5_function_setup(dev, mlx5_tout_ms(dev, FW_PRE_INIT_TIMEOUT)); + err = mlx5_function_setup(dev, true, mlx5_tout_ms(dev, FW_PRE_INIT_TIMEOUT)); if (err) goto err_function; @@ -1416,7 +1453,7 @@ int mlx5_load_one_devl_locked(struct mlx5_core_dev *dev, bool recovery) timeout = mlx5_tout_ms(dev, FW_PRE_INIT_ON_RECOVERY_TIMEOUT); else timeout = mlx5_tout_ms(dev, FW_PRE_INIT_TIMEOUT); - err = mlx5_function_setup(dev, timeout); + err = mlx5_function_setup(dev, false, timeout); if (err) goto err_function; @@ -1507,6 +1544,8 @@ static const int types[] = { MLX5_CAP_IPSEC, MLX5_CAP_PORT_SELECTION, MLX5_CAP_DEV_SHAMPO, + MLX5_CAP_MACSEC, + MLX5_CAP_ADV_VIRTUALIZATION, }; static void mlx5_hca_caps_free(struct mlx5_core_dev *dev) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h b/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h index ad61b86d5769..a806e3de7b7c 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h @@ -143,6 +143,36 @@ enum mlx5_semaphore_space_address { #define MLX5_DEFAULT_PROF 2 +static inline int mlx5_flexible_inlen(struct mlx5_core_dev *dev, size_t fixed, + size_t item_size, size_t num_items, + const char *func, int line) +{ + int inlen; + + if (fixed > INT_MAX || item_size > INT_MAX || num_items > INT_MAX) { + mlx5_core_err(dev, "%s: %s:%d: input values too big: %zu + %zu * %zu\n", + __func__, func, line, fixed, item_size, num_items); + return -ENOMEM; + } + + if (check_mul_overflow((int)item_size, (int)num_items, &inlen)) { + mlx5_core_err(dev, "%s: %s:%d: multiplication overflow: %zu + %zu * %zu\n", + __func__, func, line, fixed, item_size, num_items); + return -ENOMEM; + } + + if (check_add_overflow((int)fixed, inlen, &inlen)) { + mlx5_core_err(dev, "%s: %s:%d: addition overflow: %zu + %zu * %zu\n", + __func__, func, line, fixed, item_size, num_items); + return -ENOMEM; + } + + return inlen; +} + +#define MLX5_FLEXIBLE_INLEN(dev, fixed, item_size, num_items) \ + mlx5_flexible_inlen(dev, fixed, item_size, num_items, __func__, __LINE__) + int mlx5_query_hca_caps(struct mlx5_core_dev *dev); int mlx5_query_board_id(struct mlx5_core_dev *dev); int mlx5_cmd_init(struct mlx5_core_dev *dev); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/port.c b/drivers/net/ethernet/mellanox/mlx5/core/port.c index e1bd54574ea5..a1548e6bfb35 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/port.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/port.c @@ -493,29 +493,6 @@ int mlx5_query_port_vl_hw_cap(struct mlx5_core_dev *dev, } EXPORT_SYMBOL_GPL(mlx5_query_port_vl_hw_cap); -int mlx5_core_query_ib_ppcnt(struct mlx5_core_dev *dev, - u8 port_num, void *out, size_t sz) -{ - u32 *in; - int err; - - in = kvzalloc(sz, GFP_KERNEL); - if (!in) { - err = -ENOMEM; - return err; - } - - MLX5_SET(ppcnt_reg, in, local_port, port_num); - - MLX5_SET(ppcnt_reg, in, grp, MLX5_INFINIBAND_PORT_COUNTERS_GROUP); - err = mlx5_core_access_reg(dev, in, sz, out, - sz, MLX5_REG_PPCNT, 0, 0); - - kvfree(in); - return err; -} -EXPORT_SYMBOL_GPL(mlx5_core_query_ib_ppcnt); - static int mlx5_query_pfcc_reg(struct mlx5_core_dev *dev, u32 *out, u32 out_size) { diff --git a/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_types.h b/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_types.h index 062c7c74a1f3..1777a1e508e7 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_types.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_types.h @@ -1294,20 +1294,6 @@ struct mlx5dr_cmd_gid_attr { u32 roce_ver; }; -struct mlx5dr_cmd_qp_create_attr { - u32 page_id; - u32 pdn; - u32 cqn; - u32 pm_state; - u32 service_type; - u32 buff_umem_id; - u32 db_umem_id; - u32 sq_wqe_cnt; - u32 rq_wqe_cnt; - u32 rq_wqe_shift; - u8 isolate_vl_tc:1; -}; - int mlx5dr_cmd_query_gid(struct mlx5_core_dev *mdev, u8 vhca_port_num, u16 index, struct mlx5dr_cmd_gid_attr *attr); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/steering/fs_dr.h b/drivers/net/ethernet/mellanox/mlx5/core/steering/fs_dr.h index 1fb185d6ac7f..d168622063d5 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/steering/fs_dr.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/steering/fs_dr.h @@ -14,10 +14,6 @@ struct mlx5_fs_dr_action { struct mlx5dr_action *dr_action; }; -struct mlx5_fs_dr_ns { - struct mlx5_dr_ns *dr_ns; -}; - struct mlx5_fs_dr_rule { struct mlx5dr_rule *dr_rule; /* Only actions created by fs_dr */ diff --git a/drivers/net/ethernet/mellanox/mlx5/core/wq.h b/drivers/net/ethernet/mellanox/mlx5/core/wq.h index e5c4dcd1425e..4d629e5ddbc7 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/wq.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/wq.h @@ -123,7 +123,7 @@ static inline void mlx5_wq_cyc_push(struct mlx5_wq_cyc *wq) wq->cur_sz++; } -static inline void mlx5_wq_cyc_push_n(struct mlx5_wq_cyc *wq, u8 n) +static inline void mlx5_wq_cyc_push_n(struct mlx5_wq_cyc *wq, u16 n) { wq->wqe_ctr += n; wq->cur_sz += n; |