diff options
Diffstat (limited to 'drivers/net/ethernet/mellanox/mlx5/core')
| -rw-r--r-- | drivers/net/ethernet/mellanox/mlx5/core/alloc.c | 31 | ||||
| -rw-r--r-- | drivers/net/ethernet/mellanox/mlx5/core/en.h | 5 | ||||
| -rw-r--r-- | drivers/net/ethernet/mellanox/mlx5/core/en_main.c | 52 | ||||
| -rw-r--r-- | drivers/net/ethernet/mellanox/mlx5/core/en_rep.c | 3 | ||||
| -rw-r--r-- | drivers/net/ethernet/mellanox/mlx5/core/en_tc.c | 5 | ||||
| -rw-r--r-- | drivers/net/ethernet/mellanox/mlx5/core/eswitch.c | 50 | ||||
| -rw-r--r-- | drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c | 3 | ||||
| -rw-r--r-- | drivers/net/ethernet/mellanox/mlx5/core/fs_core.c | 11 | ||||
| -rw-r--r-- | drivers/net/ethernet/mellanox/mlx5/core/fs_counters.c | 1 | ||||
| -rw-r--r-- | drivers/net/ethernet/mellanox/mlx5/core/health.c | 76 | ||||
| -rw-r--r-- | drivers/net/ethernet/mellanox/mlx5/core/main.c | 44 | ||||
| -rw-r--r-- | drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h | 1 | ||||
| -rw-r--r-- | drivers/net/ethernet/mellanox/mlx5/core/pagealloc.c | 26 |
13 files changed, 231 insertions, 77 deletions
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/alloc.c b/drivers/net/ethernet/mellanox/mlx5/core/alloc.c index 6cb38304669f..2c6e3c7b7417 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/alloc.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/alloc.c @@ -41,6 +41,13 @@ #include "mlx5_core.h" +struct mlx5_db_pgdir { + struct list_head list; + unsigned long *bitmap; + __be32 *db_page; + dma_addr_t db_dma; +}; + /* Handling for queue buffers -- we allocate a bunch of memory and * register it in a memory region at HCA virtual address 0. */ @@ -102,17 +109,28 @@ EXPORT_SYMBOL_GPL(mlx5_buf_free); static struct mlx5_db_pgdir *mlx5_alloc_db_pgdir(struct mlx5_core_dev *dev, int node) { + u32 db_per_page = PAGE_SIZE / cache_line_size(); struct mlx5_db_pgdir *pgdir; pgdir = kzalloc(sizeof(*pgdir), GFP_KERNEL); if (!pgdir) return NULL; - bitmap_fill(pgdir->bitmap, MLX5_DB_PER_PAGE); + pgdir->bitmap = kcalloc(BITS_TO_LONGS(db_per_page), + sizeof(unsigned long), + GFP_KERNEL); + + if (!pgdir->bitmap) { + kfree(pgdir); + return NULL; + } + + bitmap_fill(pgdir->bitmap, db_per_page); pgdir->db_page = mlx5_dma_zalloc_coherent_node(dev, PAGE_SIZE, &pgdir->db_dma, node); if (!pgdir->db_page) { + kfree(pgdir->bitmap); kfree(pgdir); return NULL; } @@ -123,18 +141,19 @@ static struct mlx5_db_pgdir *mlx5_alloc_db_pgdir(struct mlx5_core_dev *dev, static int mlx5_alloc_db_from_pgdir(struct mlx5_db_pgdir *pgdir, struct mlx5_db *db) { + u32 db_per_page = PAGE_SIZE / cache_line_size(); int offset; int i; - i = find_first_bit(pgdir->bitmap, MLX5_DB_PER_PAGE); - if (i >= MLX5_DB_PER_PAGE) + i = find_first_bit(pgdir->bitmap, db_per_page); + if (i >= db_per_page) return -ENOMEM; __clear_bit(i, pgdir->bitmap); db->u.pgdir = pgdir; db->index = i; - offset = db->index * L1_CACHE_BYTES; + offset = db->index * cache_line_size(); db->db = pgdir->db_page + offset / sizeof(*pgdir->db_page); db->dma = pgdir->db_dma + offset; @@ -181,14 +200,16 @@ EXPORT_SYMBOL_GPL(mlx5_db_alloc); void mlx5_db_free(struct mlx5_core_dev *dev, struct mlx5_db *db) { + u32 db_per_page = PAGE_SIZE / cache_line_size(); mutex_lock(&dev->priv.pgdir_mutex); __set_bit(db->index, db->u.pgdir->bitmap); - if (bitmap_full(db->u.pgdir->bitmap, MLX5_DB_PER_PAGE)) { + if (bitmap_full(db->u.pgdir->bitmap, db_per_page)) { dma_free_coherent(&(dev->pdev->dev), PAGE_SIZE, db->u.pgdir->db_page, db->u.pgdir->db_dma); list_del(&db->u.pgdir->list); + kfree(db->u.pgdir->bitmap); kfree(db->u.pgdir); } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en.h b/drivers/net/ethernet/mellanox/mlx5/core/en.h index 460363b66cb1..7a43502a89cc 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en.h @@ -85,6 +85,9 @@ #define MLX5_MPWRQ_SMALL_PACKET_THRESHOLD (128) #define MLX5E_PARAMS_DEFAULT_LRO_WQE_SZ (64 * 1024) +#define MLX5E_DEFAULT_LRO_TIMEOUT 32 +#define MLX5E_LRO_TIMEOUT_ARR_SIZE 4 + #define MLX5E_PARAMS_DEFAULT_RX_CQ_MODERATION_USEC 0x10 #define MLX5E_PARAMS_DEFAULT_RX_CQ_MODERATION_USEC_FROM_CQE 0x3 #define MLX5E_PARAMS_DEFAULT_RX_CQ_MODERATION_PKTS 0x20 @@ -221,6 +224,7 @@ struct mlx5e_params { struct ieee_ets ets; #endif bool rx_am_enabled; + u32 lro_timeout; }; struct mlx5e_tstamp { @@ -888,5 +892,6 @@ int mlx5e_attach_netdev(struct mlx5_core_dev *mdev, struct net_device *netdev); void mlx5e_detach_netdev(struct mlx5_core_dev *mdev, struct net_device *netdev); struct rtnl_link_stats64 * mlx5e_get_stats(struct net_device *dev, struct rtnl_link_stats64 *stats); +u32 mlx5e_choose_lro_timeout(struct mlx5_core_dev *mdev, u32 wanted_timeout); #endif /* __MLX5_EN_H__ */ diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c index 7eaf38020a8f..84e8b250e2af 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c @@ -1445,6 +1445,7 @@ static int mlx5e_open_channel(struct mlx5e_priv *priv, int ix, c->netdev = priv->netdev; c->mkey_be = cpu_to_be32(priv->mdev->mlx5e_res.mkey.key); c->num_tc = priv->params.num_tc; + c->xdp = !!priv->xdp_prog; if (priv->params.rx_am_enabled) rx_cq_profile = mlx5e_am_get_def_profile(priv->params.rx_cq_period_mode); @@ -1468,6 +1469,12 @@ static int mlx5e_open_channel(struct mlx5e_priv *priv, int ix, if (err) goto err_close_tx_cqs; + /* XDP SQ CQ params are same as normal TXQ sq CQ params */ + err = c->xdp ? mlx5e_open_cq(c, &cparam->tx_cq, &c->xdp_sq.cq, + priv->params.tx_cq_moderation) : 0; + if (err) + goto err_close_rx_cq; + napi_enable(&c->napi); err = mlx5e_open_sq(c, 0, &cparam->icosq, &c->icosq); @@ -1488,21 +1495,10 @@ static int mlx5e_open_channel(struct mlx5e_priv *priv, int ix, } } - if (priv->xdp_prog) { - /* XDP SQ CQ params are same as normal TXQ sq CQ params */ - err = mlx5e_open_cq(c, &cparam->tx_cq, &c->xdp_sq.cq, - priv->params.tx_cq_moderation); - if (err) - goto err_close_sqs; - - err = mlx5e_open_sq(c, 0, &cparam->xdp_sq, &c->xdp_sq); - if (err) { - mlx5e_close_cq(&c->xdp_sq.cq); - goto err_close_sqs; - } - } + err = c->xdp ? mlx5e_open_sq(c, 0, &cparam->xdp_sq, &c->xdp_sq) : 0; + if (err) + goto err_close_sqs; - c->xdp = !!priv->xdp_prog; err = mlx5e_open_rq(c, &cparam->rq, &c->rq); if (err) goto err_close_xdp_sq; @@ -1512,7 +1508,8 @@ static int mlx5e_open_channel(struct mlx5e_priv *priv, int ix, return 0; err_close_xdp_sq: - mlx5e_close_sq(&c->xdp_sq); + if (c->xdp) + mlx5e_close_sq(&c->xdp_sq); err_close_sqs: mlx5e_close_sqs(c); @@ -1522,6 +1519,10 @@ err_close_icosq: err_disable_napi: napi_disable(&c->napi); + if (c->xdp) + mlx5e_close_cq(&c->xdp_sq.cq); + +err_close_rx_cq: mlx5e_close_cq(&c->rq.cq); err_close_tx_cqs: @@ -1971,9 +1972,7 @@ static void mlx5e_build_tir_ctx_lro(void *tirc, struct mlx5e_priv *priv) MLX5_SET(tirc, tirc, lro_max_ip_payload_size, (priv->params.lro_wqe_sz - ROUGH_MAX_L2_L3_HDR_SZ) >> 8); - MLX5_SET(tirc, tirc, lro_timeout_period_usecs, - MLX5_CAP_ETH(priv->mdev, - lro_timer_supported_periods[2])); + MLX5_SET(tirc, tirc, lro_timeout_period_usecs, priv->params.lro_timeout); } void mlx5e_build_tir_ctx_hash(void *tirc, struct mlx5e_priv *priv) @@ -3401,6 +3400,18 @@ static void mlx5e_query_min_inline(struct mlx5_core_dev *mdev, } } +u32 mlx5e_choose_lro_timeout(struct mlx5_core_dev *mdev, u32 wanted_timeout) +{ + int i; + + /* The supported periods are organized in ascending order */ + for (i = 0; i < MLX5E_LRO_TIMEOUT_ARR_SIZE - 1; i++) + if (MLX5_CAP_ETH(mdev, lro_timer_supported_periods[i]) >= wanted_timeout) + break; + + return MLX5_CAP_ETH(mdev, lro_timer_supported_periods[i]); +} + static void mlx5e_build_nic_netdev_priv(struct mlx5_core_dev *mdev, struct net_device *netdev, const struct mlx5e_profile *profile, @@ -3419,6 +3430,9 @@ static void mlx5e_build_nic_netdev_priv(struct mlx5_core_dev *mdev, priv->profile = profile; priv->ppriv = ppriv; + priv->params.lro_timeout = + mlx5e_choose_lro_timeout(mdev, MLX5E_DEFAULT_LRO_TIMEOUT); + priv->params.log_sq_size = MLX5E_PARAMS_DEFAULT_LOG_SQ_SIZE; /* set CQE compression */ @@ -4035,7 +4049,6 @@ void mlx5e_destroy_netdev(struct mlx5_core_dev *mdev, struct mlx5e_priv *priv) const struct mlx5e_profile *profile = priv->profile; struct net_device *netdev = priv->netdev; - unregister_netdev(netdev); destroy_workqueue(priv->wq); if (profile->cleanup) profile->cleanup(priv); @@ -4052,6 +4065,7 @@ static void mlx5e_remove(struct mlx5_core_dev *mdev, void *vpriv) for (vport = 1; vport < total_vfs; vport++) mlx5_eswitch_unregister_vport_rep(esw, vport); + unregister_netdev(priv->netdev); mlx5e_detach(mdev, vpriv); mlx5e_destroy_netdev(mdev, priv); } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c index 3c97da103d30..bf1c09ca73c0 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c @@ -308,7 +308,7 @@ static void mlx5e_build_rep_netdev(struct net_device *netdev) netdev->switchdev_ops = &mlx5e_rep_switchdev_ops; #endif - netdev->features |= NETIF_F_VLAN_CHALLENGED | NETIF_F_HW_TC; + netdev->features |= NETIF_F_VLAN_CHALLENGED | NETIF_F_HW_TC | NETIF_F_NETNS_LOCAL; netdev->hw_features |= NETIF_F_HW_TC; eth_hw_addr_random(netdev); @@ -457,6 +457,7 @@ void mlx5e_vport_rep_unload(struct mlx5_eswitch *esw, struct mlx5e_priv *priv = rep->priv_data; struct net_device *netdev = priv->netdev; + unregister_netdev(netdev); mlx5e_detach_netdev(esw->dev, netdev); mlx5e_destroy_netdev(esw->dev, priv); } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c index ce8c54d18906..6bb21b31cfeb 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c @@ -237,12 +237,15 @@ static int parse_cls_flower(struct mlx5e_priv *priv, struct mlx5_flow_spec *spec skb_flow_dissector_target(f->dissector, FLOW_DISSECTOR_KEY_VLAN, f->mask); - if (mask->vlan_id) { + if (mask->vlan_id || mask->vlan_priority) { MLX5_SET(fte_match_set_lyr_2_4, headers_c, vlan_tag, 1); MLX5_SET(fte_match_set_lyr_2_4, headers_v, vlan_tag, 1); MLX5_SET(fte_match_set_lyr_2_4, headers_c, first_vid, mask->vlan_id); MLX5_SET(fte_match_set_lyr_2_4, headers_v, first_vid, key->vlan_id); + + MLX5_SET(fte_match_set_lyr_2_4, headers_c, first_prio, mask->vlan_priority); + MLX5_SET(fte_match_set_lyr_2_4, headers_v, first_prio, key->vlan_priority); } } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c index abbf2c369923..be1f7333ab7f 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c @@ -931,8 +931,8 @@ static void esw_vport_change_handler(struct work_struct *work) mutex_unlock(&esw->state_lock); } -static void esw_vport_enable_egress_acl(struct mlx5_eswitch *esw, - struct mlx5_vport *vport) +static int esw_vport_enable_egress_acl(struct mlx5_eswitch *esw, + struct mlx5_vport *vport) { int inlen = MLX5_ST_SZ_BYTES(create_flow_group_in); struct mlx5_flow_group *vlan_grp = NULL; @@ -949,9 +949,11 @@ static void esw_vport_enable_egress_acl(struct mlx5_eswitch *esw, int table_size = 2; int err = 0; - if (!MLX5_CAP_ESW_EGRESS_ACL(dev, ft_support) || - !IS_ERR_OR_NULL(vport->egress.acl)) - return; + if (!MLX5_CAP_ESW_EGRESS_ACL(dev, ft_support)) + return -EOPNOTSUPP; + + if (!IS_ERR_OR_NULL(vport->egress.acl)) + return 0; esw_debug(dev, "Create vport[%d] egress ACL log_max_size(%d)\n", vport->vport, MLX5_CAP_ESW_EGRESS_ACL(dev, log_max_ft_size)); @@ -959,12 +961,12 @@ static void esw_vport_enable_egress_acl(struct mlx5_eswitch *esw, root_ns = mlx5_get_flow_namespace(dev, MLX5_FLOW_NAMESPACE_ESW_EGRESS); if (!root_ns) { esw_warn(dev, "Failed to get E-Switch egress flow namespace\n"); - return; + return -EIO; } flow_group_in = mlx5_vzalloc(inlen); if (!flow_group_in) - return; + return -ENOMEM; acl = mlx5_create_vport_flow_table(root_ns, 0, table_size, 0, vport->vport); if (IS_ERR(acl)) { @@ -1009,6 +1011,7 @@ out: mlx5_destroy_flow_group(vlan_grp); if (err && !IS_ERR_OR_NULL(acl)) mlx5_destroy_flow_table(acl); + return err; } static void esw_vport_cleanup_egress_rules(struct mlx5_eswitch *esw, @@ -1041,8 +1044,8 @@ static void esw_vport_disable_egress_acl(struct mlx5_eswitch *esw, vport->egress.acl = NULL; } -static void esw_vport_enable_ingress_acl(struct mlx5_eswitch *esw, - struct mlx5_vport *vport) +static int esw_vport_enable_ingress_acl(struct mlx5_eswitch *esw, + struct mlx5_vport *vport) { int inlen = MLX5_ST_SZ_BYTES(create_flow_group_in); struct mlx5_core_dev *dev = esw->dev; @@ -1063,9 +1066,11 @@ static void esw_vport_enable_ingress_acl(struct mlx5_eswitch *esw, int table_size = 4; int err = 0; - if (!MLX5_CAP_ESW_INGRESS_ACL(dev, ft_support) || - !IS_ERR_OR_NULL(vport->ingress.acl)) - return; + if (!MLX5_CAP_ESW_INGRESS_ACL(dev, ft_support)) + return -EOPNOTSUPP; + + if (!IS_ERR_OR_NULL(vport->ingress.acl)) + return 0; esw_debug(dev, "Create vport[%d] ingress ACL log_max_size(%d)\n", vport->vport, MLX5_CAP_ESW_INGRESS_ACL(dev, log_max_ft_size)); @@ -1073,12 +1078,12 @@ static void esw_vport_enable_ingress_acl(struct mlx5_eswitch *esw, root_ns = mlx5_get_flow_namespace(dev, MLX5_FLOW_NAMESPACE_ESW_INGRESS); if (!root_ns) { esw_warn(dev, "Failed to get E-Switch ingress flow namespace\n"); - return; + return -EIO; } flow_group_in = mlx5_vzalloc(inlen); if (!flow_group_in) - return; + return -ENOMEM; acl = mlx5_create_vport_flow_table(root_ns, 0, table_size, 0, vport->vport); if (IS_ERR(acl)) { @@ -1167,6 +1172,7 @@ out: } kvfree(flow_group_in); + return err; } static void esw_vport_cleanup_ingress_rules(struct mlx5_eswitch *esw, @@ -1225,7 +1231,13 @@ static int esw_vport_ingress_config(struct mlx5_eswitch *esw, return 0; } - esw_vport_enable_ingress_acl(esw, vport); + err = esw_vport_enable_ingress_acl(esw, vport); + if (err) { + mlx5_core_warn(esw->dev, + "failed to enable ingress acl (%d) on vport[%d]\n", + err, vport->vport); + return err; + } esw_debug(esw->dev, "vport[%d] configure ingress rules, vlan(%d) qos(%d)\n", @@ -1299,7 +1311,13 @@ static int esw_vport_egress_config(struct mlx5_eswitch *esw, return 0; } - esw_vport_enable_egress_acl(esw, vport); + err = esw_vport_enable_egress_acl(esw, vport); + if (err) { + mlx5_core_warn(esw->dev, + "failed to enable egress acl (%d) on vport[%d]\n", + err, vport->vport); + return err; + } esw_debug(esw->dev, "vport[%d] configure egress rules, vlan(%d) qos(%d)\n", diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c index c55ad8d00c05..d239f5d0ea36 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c @@ -57,7 +57,8 @@ mlx5_eswitch_add_offloaded_rule(struct mlx5_eswitch *esw, if (esw->mode != SRIOV_OFFLOADS) return ERR_PTR(-EOPNOTSUPP); - action = attr->action; + /* per flow vlan pop/push is emulated, don't set that into the firmware */ + action = attr->action & ~(MLX5_FLOW_CONTEXT_ACTION_VLAN_PUSH | MLX5_FLOW_CONTEXT_ACTION_VLAN_POP); if (action & MLX5_FLOW_CONTEXT_ACTION_FWD_DEST) { dest.type = MLX5_FLOW_DESTINATION_TYPE_VPORT; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c index 5da2cc878582..914e5466f729 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c @@ -436,6 +436,9 @@ static void del_flow_group(struct fs_node *node) fs_get_obj(ft, fg->node.parent); dev = get_dev(&ft->node); + if (ft->autogroup.active) + ft->autogroup.num_groups--; + if (mlx5_cmd_destroy_flow_group(dev, ft, fg->id)) mlx5_core_warn(dev, "flow steering can't destroy fg %d of ft %d\n", fg->id, ft->id); @@ -879,7 +882,7 @@ static struct mlx5_flow_group *create_flow_group_common(struct mlx5_flow_table * tree_init_node(&fg->node, !is_auto_fg, del_flow_group); tree_add_node(&fg->node, &ft->node); /* Add node to group list */ - list_add(&fg->node.list, ft->node.children.prev); + list_add(&fg->node.list, prev_fg); return fg; } @@ -893,7 +896,7 @@ struct mlx5_flow_group *mlx5_create_flow_group(struct mlx5_flow_table *ft, return ERR_PTR(-EPERM); lock_ref_node(&ft->node); - fg = create_flow_group_common(ft, fg_in, &ft->node.children, false); + fg = create_flow_group_common(ft, fg_in, ft->node.children.prev, false); unlock_ref_node(&ft->node); return fg; @@ -1012,7 +1015,7 @@ static struct mlx5_flow_group *create_autogroup(struct mlx5_flow_table *ft, u32 *match_criteria) { int inlen = MLX5_ST_SZ_BYTES(create_flow_group_in); - struct list_head *prev = &ft->node.children; + struct list_head *prev = ft->node.children.prev; unsigned int candidate_index = 0; struct mlx5_flow_group *fg; void *match_criteria_addr; @@ -1687,7 +1690,7 @@ static int init_root_ns(struct mlx5_flow_steering *steering) { steering->root_ns = create_root_ns(steering, FS_FT_NIC_RX); - if (IS_ERR_OR_NULL(steering->root_ns)) + if (!steering->root_ns) goto cleanup; if (init_root_tree(steering, &root_fs, &steering->root_ns->ns.node)) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_counters.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_counters.c index 3a9195b4169d..3b026c151cf2 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/fs_counters.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_counters.c @@ -218,6 +218,7 @@ struct mlx5_fc *mlx5_fc_create(struct mlx5_core_dev *dev, bool aging) goto err_out; if (aging) { + counter->cache.lastuse = jiffies; counter->aging = true; spin_lock(&fc_stats->addlist_lock); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/health.c b/drivers/net/ethernet/mellanox/mlx5/core/health.c index 1a05fb965c8d..5bcf93422ee0 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/health.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/health.c @@ -61,10 +61,15 @@ enum { enum { MLX5_NIC_IFC_FULL = 0, MLX5_NIC_IFC_DISABLED = 1, - MLX5_NIC_IFC_NO_DRAM_NIC = 2 + MLX5_NIC_IFC_NO_DRAM_NIC = 2, + MLX5_NIC_IFC_INVALID = 3 }; -static u8 get_nic_interface(struct mlx5_core_dev *dev) +enum { + MLX5_DROP_NEW_HEALTH_WORK, +}; + +static u8 get_nic_state(struct mlx5_core_dev *dev) { return (ioread32be(&dev->iseg->cmdq_addr_l_sz) >> 8) & 3; } @@ -97,7 +102,7 @@ static int in_fatal(struct mlx5_core_dev *dev) struct mlx5_core_health *health = &dev->priv.health; struct health_buffer __iomem *h = health->health; - if (get_nic_interface(dev) == MLX5_NIC_IFC_DISABLED) + if (get_nic_state(dev) == MLX5_NIC_IFC_DISABLED) return 1; if (ioread32be(&h->fw_ver) == 0xffffffff) @@ -127,7 +132,7 @@ unlock: static void mlx5_handle_bad_state(struct mlx5_core_dev *dev) { - u8 nic_interface = get_nic_interface(dev); + u8 nic_interface = get_nic_state(dev); switch (nic_interface) { case MLX5_NIC_IFC_FULL: @@ -149,8 +154,34 @@ static void mlx5_handle_bad_state(struct mlx5_core_dev *dev) mlx5_disable_device(dev); } +static void health_recover(struct work_struct *work) +{ + struct mlx5_core_health *health; + struct delayed_work *dwork; + struct mlx5_core_dev *dev; + struct mlx5_priv *priv; + u8 nic_state; + + dwork = container_of(work, struct delayed_work, work); + health = container_of(dwork, struct mlx5_core_health, recover_work); + priv = container_of(health, struct mlx5_priv, health); + dev = container_of(priv, struct mlx5_core_dev, priv); + + nic_state = get_nic_state(dev); + if (nic_state == MLX5_NIC_IFC_INVALID) { + dev_err(&dev->pdev->dev, "health recovery flow aborted since the nic state is invalid\n"); + return; + } + + dev_err(&dev->pdev->dev, "starting health recovery flow\n"); + mlx5_recover_device(dev); +} + +/* How much time to wait until health resetting the driver (in msecs) */ +#define MLX5_RECOVERY_DELAY_MSECS 60000 static void health_care(struct work_struct *work) { + unsigned long recover_delay = msecs_to_jiffies(MLX5_RECOVERY_DELAY_MSECS); struct mlx5_core_health *health; struct mlx5_core_dev *dev; struct mlx5_priv *priv; @@ -160,6 +191,14 @@ static void health_care(struct work_struct *work) dev = container_of(priv, struct mlx5_core_dev, priv); mlx5_core_warn(dev, "handling bad device here\n"); mlx5_handle_bad_state(dev); + + spin_lock(&health->wq_lock); + if (!test_bit(MLX5_DROP_NEW_HEALTH_WORK, &health->flags)) + schedule_delayed_work(&health->recover_work, recover_delay); + else + dev_err(&dev->pdev->dev, + "new health works are not permitted at this stage\n"); + spin_unlock(&health->wq_lock); } static const char *hsynd_str(u8 synd) @@ -272,7 +311,13 @@ static void poll_health(unsigned long data) if (in_fatal(dev) && !health->sick) { health->sick = true; print_health_info(dev); - schedule_work(&health->work); + spin_lock(&health->wq_lock); + if (!test_bit(MLX5_DROP_NEW_HEALTH_WORK, &health->flags)) + queue_work(health->wq, &health->work); + else + dev_err(&dev->pdev->dev, + "new health works are not permitted at this stage\n"); + spin_unlock(&health->wq_lock); } } @@ -281,6 +326,8 @@ void mlx5_start_health_poll(struct mlx5_core_dev *dev) struct mlx5_core_health *health = &dev->priv.health; init_timer(&health->timer); + health->sick = 0; + clear_bit(MLX5_DROP_NEW_HEALTH_WORK, &health->flags); health->health = &dev->iseg->health; health->health_counter = &dev->iseg->health_counter; @@ -297,11 +344,22 @@ void mlx5_stop_health_poll(struct mlx5_core_dev *dev) del_timer_sync(&health->timer); } +void mlx5_drain_health_wq(struct mlx5_core_dev *dev) +{ + struct mlx5_core_health *health = &dev->priv.health; + + spin_lock(&health->wq_lock); + set_bit(MLX5_DROP_NEW_HEALTH_WORK, &health->flags); + spin_unlock(&health->wq_lock); + cancel_delayed_work_sync(&health->recover_work); + cancel_work_sync(&health->work); +} + void mlx5_health_cleanup(struct mlx5_core_dev *dev) { struct mlx5_core_health *health = &dev->priv.health; - flush_work(&health->work); + destroy_workqueue(health->wq); } int mlx5_health_init(struct mlx5_core_dev *dev) @@ -316,9 +374,13 @@ int mlx5_health_init(struct mlx5_core_dev *dev) strcpy(name, "mlx5_health"); strcat(name, dev_name(&dev->pdev->dev)); + health->wq = create_singlethread_workqueue(name); kfree(name); - + if (!health->wq) + return -ENOMEM; + spin_lock_init(&health->wq_lock); INIT_WORK(&health->work, health_care); + INIT_DELAYED_WORK(&health->recover_work, health_recover); return 0; } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/main.c b/drivers/net/ethernet/mellanox/mlx5/core/main.c index d9c3c70b29e4..3eb931585b3e 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/main.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/main.c @@ -844,12 +844,6 @@ static int mlx5_init_once(struct mlx5_core_dev *dev, struct mlx5_priv *priv) struct pci_dev *pdev = dev->pdev; int err; - err = mlx5_query_hca_caps(dev); - if (err) { - dev_err(&pdev->dev, "query hca failed\n"); - goto out; - } - err = mlx5_query_board_id(dev); if (err) { dev_err(&pdev->dev, "query board id failed\n"); @@ -1023,6 +1017,12 @@ static int mlx5_load_one(struct mlx5_core_dev *dev, struct mlx5_priv *priv, mlx5_start_health_poll(dev); + err = mlx5_query_hca_caps(dev); + if (err) { + dev_err(&pdev->dev, "query hca failed\n"); + goto err_stop_poll; + } + if (boot && mlx5_init_once(dev, priv)) { dev_err(&pdev->dev, "sw objs init failed\n"); goto err_stop_poll; @@ -1226,6 +1226,9 @@ static int init_one(struct pci_dev *pdev, pci_set_drvdata(pdev, dev); + dev->pdev = pdev; + dev->event = mlx5_core_event; + if (prof_sel < 0 || prof_sel >= ARRAY_SIZE(profile)) { mlx5_core_warn(dev, "selected profile out of range, selecting default (%d)\n", @@ -1233,8 +1236,6 @@ static int init_one(struct pci_dev *pdev, prof_sel = MLX5_DEFAULT_PROF; } dev->profile = &profile[prof_sel]; - dev->pdev = pdev; - dev->event = mlx5_core_event; INIT_LIST_HEAD(&priv->ctx_list); spin_lock_init(&priv->ctx_lock); @@ -1313,10 +1314,16 @@ static pci_ers_result_t mlx5_pci_err_detected(struct pci_dev *pdev, struct mlx5_priv *priv = &dev->priv; dev_info(&pdev->dev, "%s was called\n", __func__); + mlx5_enter_error_state(dev); mlx5_unload_one(dev, priv, false); - pci_save_state(pdev); - mlx5_pci_disable_device(dev); + /* In case of kernel call save the pci state and drain health wq */ + if (state) { + pci_save_state(pdev); + mlx5_drain_health_wq(dev); + mlx5_pci_disable_device(dev); + } + return state == pci_channel_io_perm_failure ? PCI_ERS_RESULT_DISCONNECT : PCI_ERS_RESULT_NEED_RESET; } @@ -1373,11 +1380,6 @@ static pci_ers_result_t mlx5_pci_slot_reset(struct pci_dev *pdev) return PCI_ERS_RESULT_RECOVERED; } -void mlx5_disable_device(struct mlx5_core_dev *dev) -{ - mlx5_pci_err_detected(dev->pdev, 0); -} - static void mlx5_pci_resume(struct pci_dev *pdev) { struct mlx5_core_dev *dev = pci_get_drvdata(pdev); @@ -1427,6 +1429,18 @@ static const struct pci_device_id mlx5_core_pci_table[] = { MODULE_DEVICE_TABLE(pci, mlx5_core_pci_table); +void mlx5_disable_device(struct mlx5_core_dev *dev) +{ + mlx5_pci_err_detected(dev->pdev, 0); +} + +void mlx5_recover_device(struct mlx5_core_dev *dev) +{ + mlx5_pci_disable_device(dev); + if (mlx5_pci_slot_reset(dev->pdev) == PCI_ERS_RESULT_RECOVERED) + mlx5_pci_resume(dev->pdev); +} + static struct pci_driver mlx5_core_driver = { .name = DRIVER_NAME, .id_table = mlx5_core_pci_table, diff --git a/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h b/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h index 3d0cfb9f18f9..187662c8ea96 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h @@ -83,6 +83,7 @@ void mlx5_core_event(struct mlx5_core_dev *dev, enum mlx5_dev_event event, unsigned long param); void mlx5_enter_error_state(struct mlx5_core_dev *dev); void mlx5_disable_device(struct mlx5_core_dev *dev); +void mlx5_recover_device(struct mlx5_core_dev *dev); int mlx5_sriov_init(struct mlx5_core_dev *dev); void mlx5_sriov_cleanup(struct mlx5_core_dev *dev); int mlx5_sriov_attach(struct mlx5_core_dev *dev); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/pagealloc.c b/drivers/net/ethernet/mellanox/mlx5/core/pagealloc.c index cc4fd61914d3..a57d5a81eb05 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/pagealloc.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/pagealloc.c @@ -209,6 +209,7 @@ static void free_4k(struct mlx5_core_dev *dev, u64 addr) static int alloc_system_page(struct mlx5_core_dev *dev, u16 func_id) { struct page *page; + u64 zero_addr = 1; u64 addr; int err; int nid = dev_to_node(&dev->pdev->dev); @@ -218,26 +219,35 @@ static int alloc_system_page(struct mlx5_core_dev *dev, u16 func_id) mlx5_core_warn(dev, "failed to allocate page\n"); return -ENOMEM; } +map: addr = dma_map_page(&dev->pdev->dev, page, 0, PAGE_SIZE, DMA_BIDIRECTIONAL); if (dma_mapping_error(&dev->pdev->dev, addr)) { mlx5_core_warn(dev, "failed dma mapping page\n"); err = -ENOMEM; - goto out_alloc; + goto err_mapping; } + + /* Firmware doesn't support page with physical address 0 */ + if (addr == 0) { + zero_addr = addr; + goto map; + } + err = insert_page(dev, addr, page, func_id); if (err) { mlx5_core_err(dev, "failed to track allocated page\n"); - goto out_mapping; + dma_unmap_page(&dev->pdev->dev, addr, PAGE_SIZE, + DMA_BIDIRECTIONAL); } - return 0; - -out_mapping: - dma_unmap_page(&dev->pdev->dev, addr, PAGE_SIZE, DMA_BIDIRECTIONAL); +err_mapping: + if (err) + __free_page(page); -out_alloc: - __free_page(page); + if (zero_addr == 0) + dma_unmap_page(&dev->pdev->dev, zero_addr, PAGE_SIZE, + DMA_BIDIRECTIONAL); return err; } |