diff options
-rw-r--r-- | drivers/net/ethernet/mellanox/mlxsw/spectrum.c | 182 | ||||
-rw-r--r-- | drivers/net/ethernet/mellanox/mlxsw/spectrum.h | 43 | ||||
-rw-r--r-- | drivers/net/ethernet/mellanox/mlxsw/spectrum_acl.c | 302 | ||||
-rw-r--r-- | drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c | 44 | ||||
-rw-r--r-- | drivers/net/ethernet/mellanox/mlxsw/spectrum_flower.c | 41 | ||||
-rw-r--r-- | include/net/pkt_cls.h | 8 | ||||
-rw-r--r-- | include/net/sch_generic.h | 34 | ||||
-rw-r--r-- | include/uapi/linux/rtnetlink.h | 12 | ||||
-rw-r--r-- | net/sched/cls_api.c | 595 | ||||
-rw-r--r-- | net/sched/cls_bpf.c | 9 | ||||
-rw-r--r-- | net/sched/cls_flow.c | 2 | ||||
-rw-r--r-- | net/sched/cls_flower.c | 3 | ||||
-rw-r--r-- | net/sched/cls_matchall.c | 3 | ||||
-rw-r--r-- | net/sched/cls_route.c | 2 | ||||
-rw-r--r-- | net/sched/cls_u32.c | 13 | ||||
-rw-r--r-- | net/sched/sch_api.c | 60 | ||||
-rw-r--r-- | net/sched/sch_ingress.c | 80 |
17 files changed, 1111 insertions, 322 deletions
diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum.c index ed92e04309ba..bbe48917dcad 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum.c @@ -1747,72 +1747,186 @@ static int mlxsw_sp_setup_tc_cls_matchall(struct mlxsw_sp_port *mlxsw_sp_port, } static int -mlxsw_sp_setup_tc_cls_flower(struct mlxsw_sp_port *mlxsw_sp_port, - struct tc_cls_flower_offload *f, - bool ingress) +mlxsw_sp_setup_tc_cls_flower(struct mlxsw_sp_acl_block *acl_block, + struct tc_cls_flower_offload *f) { + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_acl_block_mlxsw_sp(acl_block); + switch (f->command) { case TC_CLSFLOWER_REPLACE: - return mlxsw_sp_flower_replace(mlxsw_sp_port, ingress, f); + return mlxsw_sp_flower_replace(mlxsw_sp, acl_block, f); case TC_CLSFLOWER_DESTROY: - mlxsw_sp_flower_destroy(mlxsw_sp_port, ingress, f); + mlxsw_sp_flower_destroy(mlxsw_sp, acl_block, f); return 0; case TC_CLSFLOWER_STATS: - return mlxsw_sp_flower_stats(mlxsw_sp_port, ingress, f); + return mlxsw_sp_flower_stats(mlxsw_sp, acl_block, f); default: return -EOPNOTSUPP; } } -static int mlxsw_sp_setup_tc_block_cb(enum tc_setup_type type, void *type_data, - void *cb_priv, bool ingress) +static int mlxsw_sp_setup_tc_block_cb_matchall(enum tc_setup_type type, + void *type_data, + void *cb_priv, bool ingress) { struct mlxsw_sp_port *mlxsw_sp_port = cb_priv; - if (!tc_can_offload(mlxsw_sp_port->dev)) - return -EOPNOTSUPP; - switch (type) { case TC_SETUP_CLSMATCHALL: + if (!tc_can_offload(mlxsw_sp_port->dev)) + return -EOPNOTSUPP; + return mlxsw_sp_setup_tc_cls_matchall(mlxsw_sp_port, type_data, ingress); case TC_SETUP_CLSFLOWER: - return mlxsw_sp_setup_tc_cls_flower(mlxsw_sp_port, type_data, - ingress); + return 0; default: return -EOPNOTSUPP; } } -static int mlxsw_sp_setup_tc_block_cb_ig(enum tc_setup_type type, - void *type_data, void *cb_priv) +static int mlxsw_sp_setup_tc_block_cb_matchall_ig(enum tc_setup_type type, + void *type_data, + void *cb_priv) { - return mlxsw_sp_setup_tc_block_cb(type, type_data, cb_priv, true); + return mlxsw_sp_setup_tc_block_cb_matchall(type, type_data, + cb_priv, true); } -static int mlxsw_sp_setup_tc_block_cb_eg(enum tc_setup_type type, - void *type_data, void *cb_priv) +static int mlxsw_sp_setup_tc_block_cb_matchall_eg(enum tc_setup_type type, + void *type_data, + void *cb_priv) { - return mlxsw_sp_setup_tc_block_cb(type, type_data, cb_priv, false); + return mlxsw_sp_setup_tc_block_cb_matchall(type, type_data, + cb_priv, false); +} + +static int mlxsw_sp_setup_tc_block_cb_flower(enum tc_setup_type type, + void *type_data, void *cb_priv) +{ + struct mlxsw_sp_acl_block *acl_block = cb_priv; + + switch (type) { + case TC_SETUP_CLSMATCHALL: + return 0; + case TC_SETUP_CLSFLOWER: + if (mlxsw_sp_acl_block_disabled(acl_block)) + return -EOPNOTSUPP; + + return mlxsw_sp_setup_tc_cls_flower(acl_block, type_data); + default: + return -EOPNOTSUPP; + } +} + +static int +mlxsw_sp_setup_tc_block_flower_bind(struct mlxsw_sp_port *mlxsw_sp_port, + struct tcf_block *block, bool ingress) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + struct mlxsw_sp_acl_block *acl_block; + struct tcf_block_cb *block_cb; + int err; + + block_cb = tcf_block_cb_lookup(block, mlxsw_sp_setup_tc_block_cb_flower, + mlxsw_sp); + if (!block_cb) { + acl_block = mlxsw_sp_acl_block_create(mlxsw_sp, block->net); + if (!acl_block) + return -ENOMEM; + block_cb = __tcf_block_cb_register(block, + mlxsw_sp_setup_tc_block_cb_flower, + mlxsw_sp, acl_block); + if (IS_ERR(block_cb)) { + err = PTR_ERR(block_cb); + goto err_cb_register; + } + } else { + acl_block = tcf_block_cb_priv(block_cb); + } + tcf_block_cb_incref(block_cb); + err = mlxsw_sp_acl_block_bind(mlxsw_sp, acl_block, + mlxsw_sp_port, ingress); + if (err) + goto err_block_bind; + + if (ingress) + mlxsw_sp_port->ing_acl_block = acl_block; + else + mlxsw_sp_port->eg_acl_block = acl_block; + + return 0; + +err_block_bind: + if (!tcf_block_cb_decref(block_cb)) { + __tcf_block_cb_unregister(block_cb); +err_cb_register: + mlxsw_sp_acl_block_destroy(acl_block); + } + return err; +} + +static void +mlxsw_sp_setup_tc_block_flower_unbind(struct mlxsw_sp_port *mlxsw_sp_port, + struct tcf_block *block, bool ingress) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + struct mlxsw_sp_acl_block *acl_block; + struct tcf_block_cb *block_cb; + int err; + + block_cb = tcf_block_cb_lookup(block, mlxsw_sp_setup_tc_block_cb_flower, + mlxsw_sp); + if (!block_cb) + return; + + if (ingress) + mlxsw_sp_port->ing_acl_block = NULL; + else + mlxsw_sp_port->eg_acl_block = NULL; + + acl_block = tcf_block_cb_priv(block_cb); + err = mlxsw_sp_acl_block_unbind(mlxsw_sp, acl_block, + mlxsw_sp_port, ingress); + if (!err && !tcf_block_cb_decref(block_cb)) { + __tcf_block_cb_unregister(block_cb); + mlxsw_sp_acl_block_destroy(acl_block); + } } static int mlxsw_sp_setup_tc_block(struct mlxsw_sp_port *mlxsw_sp_port, struct tc_block_offload *f) { tc_setup_cb_t *cb; + bool ingress; + int err; - if (f->binder_type == TCF_BLOCK_BINDER_TYPE_CLSACT_INGRESS) - cb = mlxsw_sp_setup_tc_block_cb_ig; - else if (f->binder_type == TCF_BLOCK_BINDER_TYPE_CLSACT_EGRESS) - cb = mlxsw_sp_setup_tc_block_cb_eg; - else + if (f->binder_type == TCF_BLOCK_BINDER_TYPE_CLSACT_INGRESS) { + cb = mlxsw_sp_setup_tc_block_cb_matchall_ig; + ingress = true; + } else if (f->binder_type == TCF_BLOCK_BINDER_TYPE_CLSACT_EGRESS) { + cb = mlxsw_sp_setup_tc_block_cb_matchall_eg; + ingress = false; + } else { return -EOPNOTSUPP; + } switch (f->command) { case TC_BLOCK_BIND: - return tcf_block_cb_register(f->block, cb, mlxsw_sp_port, - mlxsw_sp_port); + err = tcf_block_cb_register(f->block, cb, mlxsw_sp_port, + mlxsw_sp_port); + if (err) + return err; + err = mlxsw_sp_setup_tc_block_flower_bind(mlxsw_sp_port, + f->block, ingress); + if (err) { + tcf_block_cb_unregister(f->block, cb, mlxsw_sp_port); + return err; + } + return 0; case TC_BLOCK_UNBIND: + mlxsw_sp_setup_tc_block_flower_unbind(mlxsw_sp_port, + f->block, ingress); tcf_block_cb_unregister(f->block, cb, mlxsw_sp_port); return 0; default: @@ -1842,10 +1956,18 @@ static int mlxsw_sp_feature_hw_tc(struct net_device *dev, bool enable) { struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); - if (!enable && (mlxsw_sp_port->acl_rule_count || - !list_empty(&mlxsw_sp_port->mall_tc_list))) { - netdev_err(dev, "Active offloaded tc filters, can't turn hw_tc_offload off\n"); - return -EINVAL; + if (!enable) { + if (mlxsw_sp_acl_block_rule_count(mlxsw_sp_port->ing_acl_block) || + mlxsw_sp_acl_block_rule_count(mlxsw_sp_port->eg_acl_block) || + !list_empty(&mlxsw_sp_port->mall_tc_list)) { + netdev_err(dev, "Active offloaded tc filters, can't turn hw_tc_offload off\n"); + return -EINVAL; + } + mlxsw_sp_acl_block_disable_inc(mlxsw_sp_port->ing_acl_block); + mlxsw_sp_acl_block_disable_inc(mlxsw_sp_port->eg_acl_block); + } else { + mlxsw_sp_acl_block_disable_dec(mlxsw_sp_port->ing_acl_block); + mlxsw_sp_acl_block_disable_dec(mlxsw_sp_port->eg_acl_block); } return 0; } diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum.h b/drivers/net/ethernet/mellanox/mlxsw/spectrum.h index 237cad373dbe..58ff79211c09 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum.h +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum.h @@ -260,6 +260,8 @@ struct mlxsw_sp_port { struct list_head vlans_list; struct mlxsw_sp_qdisc *root_qdisc; unsigned acl_rule_count; + struct mlxsw_sp_acl_block *ing_acl_block; + struct mlxsw_sp_acl_block *eg_acl_block; }; static inline bool @@ -468,8 +470,11 @@ struct mlxsw_sp_acl_profile_ops { void *priv, void *ruleset_priv); void (*ruleset_del)(struct mlxsw_sp *mlxsw_sp, void *ruleset_priv); int (*ruleset_bind)(struct mlxsw_sp *mlxsw_sp, void *ruleset_priv, - struct net_device *dev, bool ingress); - void (*ruleset_unbind)(struct mlxsw_sp *mlxsw_sp, void *ruleset_priv); + struct mlxsw_sp_port *mlxsw_sp_port, + bool ingress); + void (*ruleset_unbind)(struct mlxsw_sp *mlxsw_sp, void *ruleset_priv, + struct mlxsw_sp_port *mlxsw_sp_port, + bool ingress); u16 (*ruleset_group_id)(void *ruleset_priv); size_t rule_priv_size; int (*rule_add)(struct mlxsw_sp *mlxsw_sp, @@ -489,17 +494,34 @@ struct mlxsw_sp_acl_ops { enum mlxsw_sp_acl_profile profile); }; +struct mlxsw_sp_acl_block; struct mlxsw_sp_acl_ruleset; /* spectrum_acl.c */ struct mlxsw_afk *mlxsw_sp_acl_afk(struct mlxsw_sp_acl *acl); +struct mlxsw_sp *mlxsw_sp_acl_block_mlxsw_sp(struct mlxsw_sp_acl_block *block); +unsigned int mlxsw_sp_acl_block_rule_count(struct mlxsw_sp_acl_block *block); +void mlxsw_sp_acl_block_disable_inc(struct mlxsw_sp_acl_block *block); +void mlxsw_sp_acl_block_disable_dec(struct mlxsw_sp_acl_block *block); +bool mlxsw_sp_acl_block_disabled(struct mlxsw_sp_acl_block *block); +struct mlxsw_sp_acl_block *mlxsw_sp_acl_block_create(struct mlxsw_sp *mlxsw_sp, + struct net *net); +void mlxsw_sp_acl_block_destroy(struct mlxsw_sp_acl_block *block); +int mlxsw_sp_acl_block_bind(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_block *block, + struct mlxsw_sp_port *mlxsw_sp_port, + bool ingress); +int mlxsw_sp_acl_block_unbind(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_block *block, + struct mlxsw_sp_port *mlxsw_sp_port, + bool ingress); struct mlxsw_sp_acl_ruleset * -mlxsw_sp_acl_ruleset_lookup(struct mlxsw_sp *mlxsw_sp, struct net_device *dev, - bool ingress, u32 chain_index, +mlxsw_sp_acl_ruleset_lookup(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_block *block, u32 chain_index, enum mlxsw_sp_acl_profile profile); struct mlxsw_sp_acl_ruleset * -mlxsw_sp_acl_ruleset_get(struct mlxsw_sp *mlxsw_sp, struct net_device *dev, - bool ingress, u32 chain_index, +mlxsw_sp_acl_ruleset_get(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_block *block, u32 chain_index, enum mlxsw_sp_acl_profile profile); void mlxsw_sp_acl_ruleset_put(struct mlxsw_sp *mlxsw_sp, struct mlxsw_sp_acl_ruleset *ruleset); @@ -566,11 +588,14 @@ void mlxsw_sp_acl_fini(struct mlxsw_sp *mlxsw_sp); extern const struct mlxsw_sp_acl_ops mlxsw_sp_acl_tcam_ops; /* spectrum_flower.c */ -int mlxsw_sp_flower_replace(struct mlxsw_sp_port *mlxsw_sp_port, bool ingress, +int mlxsw_sp_flower_replace(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_block *block, struct tc_cls_flower_offload *f); -void mlxsw_sp_flower_destroy(struct mlxsw_sp_port *mlxsw_sp_port, bool ingress, +void mlxsw_sp_flower_destroy(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_block *block, struct tc_cls_flower_offload *f); -int mlxsw_sp_flower_stats(struct mlxsw_sp_port *mlxsw_sp_port, bool ingress, +int mlxsw_sp_flower_stats(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_block *block, struct tc_cls_flower_offload *f); /* spectrum_qdisc.c */ diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl.c index 93dcd315f7d6..9439bfa4ecc2 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl.c @@ -39,6 +39,7 @@ #include <linux/string.h> #include <linux/rhashtable.h> #include <linux/netdevice.h> +#include <net/net_namespace.h> #include <net/tc_act/tc_vlan.h> #include "reg.h" @@ -70,9 +71,23 @@ struct mlxsw_afk *mlxsw_sp_acl_afk(struct mlxsw_sp_acl *acl) return acl->afk; } -struct mlxsw_sp_acl_ruleset_ht_key { - struct net_device *dev; /* dev this ruleset is bound to */ +struct mlxsw_sp_acl_block_binding { + struct list_head list; + struct net_device *dev; + struct mlxsw_sp_port *mlxsw_sp_port; bool ingress; +}; + +struct mlxsw_sp_acl_block { + struct list_head binding_list; + struct mlxsw_sp_acl_ruleset *ruleset_zero; + struct mlxsw_sp *mlxsw_sp; + unsigned int rule_count; + unsigned int disable_count; +}; + +struct mlxsw_sp_acl_ruleset_ht_key { + struct mlxsw_sp_acl_block *block; u32 chain_index; const struct mlxsw_sp_acl_profile_ops *ops; }; @@ -118,8 +133,185 @@ struct mlxsw_sp_fid *mlxsw_sp_acl_dummy_fid(struct mlxsw_sp *mlxsw_sp) return mlxsw_sp->acl->dummy_fid; } +struct mlxsw_sp *mlxsw_sp_acl_block_mlxsw_sp(struct mlxsw_sp_acl_block *block) +{ + return block->mlxsw_sp; +} + +unsigned int mlxsw_sp_acl_block_rule_count(struct mlxsw_sp_acl_block *block) +{ + return block ? block->rule_count : 0; +} + +void mlxsw_sp_acl_block_disable_inc(struct mlxsw_sp_acl_block *block) +{ + if (block) + block->disable_count++; +} + +void mlxsw_sp_acl_block_disable_dec(struct mlxsw_sp_acl_block *block) +{ + if (block) + block->disable_count--; +} + +bool mlxsw_sp_acl_block_disabled(struct mlxsw_sp_acl_block *block) +{ + return block->disable_count; +} + +static int +mlxsw_sp_acl_ruleset_bind(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_block *block, + struct mlxsw_sp_acl_block_binding *binding) +{ + struct mlxsw_sp_acl_ruleset *ruleset = block->ruleset_zero; + const struct mlxsw_sp_acl_profile_ops *ops = ruleset->ht_key.ops; + + return ops->ruleset_bind(mlxsw_sp, ruleset->priv, + binding->mlxsw_sp_port, binding->ingress); +} + +static void +mlxsw_sp_acl_ruleset_unbind(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_block *block, + struct mlxsw_sp_acl_block_binding *binding) +{ + struct mlxsw_sp_acl_ruleset *ruleset = block->ruleset_zero; + const struct mlxsw_sp_acl_profile_ops *ops = ruleset->ht_key.ops; + + ops->ruleset_unbind(mlxsw_sp, ruleset->priv, + binding->mlxsw_sp_port, binding->ingress); +} + +static bool mlxsw_sp_acl_ruleset_block_bound(struct mlxsw_sp_acl_block *block) +{ + return block->ruleset_zero; +} + +static int +mlxsw_sp_acl_ruleset_block_bind(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_ruleset *ruleset, + struct mlxsw_sp_acl_block *block) +{ + struct mlxsw_sp_acl_block_binding *binding; + int err; + + block->ruleset_zero = ruleset; + list_for_each_entry(binding, &block->binding_list, list) { + err = mlxsw_sp_acl_ruleset_bind(mlxsw_sp, block, binding); + if (err) + goto rollback; + } + return 0; + +rollback: + list_for_each_entry_continue_reverse(binding, &block->binding_list, + list) + mlxsw_sp_acl_ruleset_unbind(mlxsw_sp, block, binding); + block->ruleset_zero = NULL; + + return err; +} + +static void +mlxsw_sp_acl_ruleset_block_unbind(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_ruleset *ruleset, + struct mlxsw_sp_acl_block *block) +{ + struct mlxsw_sp_acl_block_binding *binding; + + list_for_each_entry(binding, &block->binding_list, list) + mlxsw_sp_acl_ruleset_unbind(mlxsw_sp, block, binding); + block->ruleset_zero = NULL; +} + +struct mlxsw_sp_acl_block *mlxsw_sp_acl_block_create(struct mlxsw_sp *mlxsw_sp, + struct net *net) +{ + struct mlxsw_sp_acl_block *block; + + block = kzalloc(sizeof(*block), GFP_KERNEL); + if (!block) + return NULL; + INIT_LIST_HEAD(&block->binding_list); + block->mlxsw_sp = mlxsw_sp; + return block; +} + +void mlxsw_sp_acl_block_destroy(struct mlxsw_sp_acl_block *block) +{ + WARN_ON(!list_empty(&block->binding_list)); + kfree(block); +} + +static struct mlxsw_sp_acl_block_binding * +mlxsw_sp_acl_block_lookup(struct mlxsw_sp_acl_block *block, + struct mlxsw_sp_port *mlxsw_sp_port, bool ingress) +{ + struct mlxsw_sp_acl_block_binding *binding; + + list_for_each_entry(binding, &block->binding_list, list) + if (binding->mlxsw_sp_port == mlxsw_sp_port && + binding->ingress == ingress) + return binding; + return NULL; +} + +int mlxsw_sp_acl_block_bind(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_block *block, + struct mlxsw_sp_port *mlxsw_sp_port, + bool ingress) +{ + struct mlxsw_sp_acl_block_binding *binding; + int err; + + if (WARN_ON(mlxsw_sp_acl_block_lookup(block, mlxsw_sp_port, ingress))) + return -EEXIST; + + binding = kzalloc(sizeof(*binding), GFP_KERNEL); + if (!binding) + return -ENOMEM; + binding->mlxsw_sp_port = mlxsw_sp_port; + binding->ingress = ingress; + + if (mlxsw_sp_acl_ruleset_block_bound(block)) { + err = mlxsw_sp_acl_ruleset_bind(mlxsw_sp, block, binding); + if (err) + goto err_ruleset_bind; + } + + list_add(&binding->list, &block->binding_list); + return 0; + +err_ruleset_bind: + kfree(binding); + return err; +} + +int mlxsw_sp_acl_block_unbind(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_block *block, + struct mlxsw_sp_port *mlxsw_sp_port, + bool ingress) +{ + struct mlxsw_sp_acl_block_binding *binding; + + binding = mlxsw_sp_acl_block_lookup(block, mlxsw_sp_port, ingress); + if (!binding) + return -ENOENT; + + list_del(&binding->list); + + if (mlxsw_sp_acl_ruleset_block_bound(block)) + mlxsw_sp_acl_ruleset_unbind(mlxsw_sp, block, binding); + + kfree(binding); + return 0; +} + static struct mlxsw_sp_acl_ruleset * mlxsw_sp_acl_ruleset_create(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_block *block, u32 chain_index, const struct mlxsw_sp_acl_profile_ops *ops) { struct mlxsw_sp_acl *acl = mlxsw_sp->acl; @@ -132,6 +324,8 @@ mlxsw_sp_acl_ruleset_create(struct mlxsw_sp *mlxsw_sp, if (!ruleset) return ERR_PTR(-ENOMEM); ruleset->ref_count = 1; + ruleset->ht_key.block = block; + ruleset->ht_key.chain_index = chain_index; ruleset->ht_key.ops = ops; err = rhashtable_init(&ruleset->rule_ht, &mlxsw_sp_acl_rule_ht_params); @@ -142,68 +336,50 @@ mlxsw_sp_acl_ruleset_create(struct mlxsw_sp *mlxsw_sp, if (err) goto err_ops_ruleset_add; - return ruleset; - -err_ops_ruleset_add: - rhashtable_destroy(&ruleset->rule_ht); -err_rhashtable_init: - kfree(ruleset); - return ERR_PTR(err); -} - -static void mlxsw_sp_acl_ruleset_destroy(struct mlxsw_sp *mlxsw_sp, - struct mlxsw_sp_acl_ruleset *ruleset) -{ - const struct mlxsw_sp_acl_profile_ops *ops = ruleset->ht_key.ops; - - ops->ruleset_del(mlxsw_sp, ruleset->priv); - rhashtable_destroy(&ruleset->rule_ht); - kfree(ruleset); -} - -static int mlxsw_sp_acl_ruleset_bind(struct mlxsw_sp *mlxsw_sp, - struct mlxsw_sp_acl_ruleset *ruleset, - struct net_device *dev, bool ingress, - u32 chain_index) -{ - const struct mlxsw_sp_acl_profile_ops *ops = ruleset->ht_key.ops; - struct mlxsw_sp_acl *acl = mlxsw_sp->acl; - int err; - - ruleset->ht_key.dev = dev; - ruleset->ht_key.ingress = ingress; - ruleset->ht_key.chain_index = chain_index; err = rhashtable_insert_fast(&acl->ruleset_ht, &ruleset->ht_node, mlxsw_sp_acl_ruleset_ht_params); if (err) - return err; - if (!ruleset->ht_key.chain_index) { + goto err_ht_insert; + + if (!chain_index) { /* We only need ruleset with chain index 0, the implicit one, * to be directly bound to device. The rest of the rulesets * are bound by "Goto action set". */ - err = ops->ruleset_bind(mlxsw_sp, ruleset->priv, dev, ingress); + err = mlxsw_sp_acl_ruleset_block_bind(mlxsw_sp, ruleset, block); if (err) - goto err_ops_ruleset_bind; + goto err_ruleset_bind; } - return 0; -err_ops_ruleset_bind: + return ruleset; + +err_ruleset_bind: rhashtable_remove_fast(&acl->ruleset_ht, &ruleset->ht_node, mlxsw_sp_acl_ruleset_ht_params); - return err; +err_ht_insert: + ops->ruleset_del(mlxsw_sp, ruleset->priv); +err_ops_ruleset_add: + rhashtable_destroy(&ruleset->rule_ht); +err_rhashtable_init: + kfree(ruleset); + return ERR_PTR(err); } -static void mlxsw_sp_acl_ruleset_unbind(struct mlxsw_sp *mlxsw_sp, - struct mlxsw_sp_acl_ruleset *ruleset) +static void mlxsw_sp_acl_ruleset_destroy(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_ruleset *ruleset) { const struct mlxsw_sp_acl_profile_ops *ops = ruleset->ht_key.ops; + struct mlxsw_sp_acl_block *block = ruleset->ht_key.block; + u32 chain_index = ruleset->ht_key.chain_index; struct mlxsw_sp_acl *acl = mlxsw_sp->acl; - if (!ruleset->ht_key.chain_index) - ops->ruleset_unbind(mlxsw_sp, ruleset->priv); + if (!chain_index) + mlxsw_sp_acl_ruleset_block_unbind(mlxsw_sp, ruleset, block); rhashtable_remove_fast(&acl->ruleset_ht, &ruleset->ht_node, mlxsw_sp_acl_ruleset_ht_params); + ops->ruleset_del(mlxsw_sp, ruleset->priv); + rhashtable_destroy(&ruleset->rule_ht); + kfree(ruleset); } static void mlxsw_sp_acl_ruleset_ref_inc(struct mlxsw_sp_acl_ruleset *ruleset) @@ -216,20 +392,18 @@ static void mlxsw_sp_acl_ruleset_ref_dec(struct mlxsw_sp *mlxsw_sp, { if (--ruleset->ref_count) return; - mlxsw_sp_acl_ruleset_unbind(mlxsw_sp, ruleset); mlxsw_sp_acl_ruleset_destroy(mlxsw_sp, ruleset); } static struct mlxsw_sp_acl_ruleset * -__mlxsw_sp_acl_ruleset_lookup(struct mlxsw_sp_acl *acl, struct net_device *dev, - bool ingress, u32 chain_index, +__mlxsw_sp_acl_ruleset_lookup(struct mlxsw_sp_acl *acl, + struct mlxsw_sp_acl_block *block, u32 chain_index, const struct mlxsw_sp_acl_profile_ops *ops) { struct mlxsw_sp_acl_ruleset_ht_key ht_key; memset(&ht_key, 0, sizeof(ht_key)); - ht_key.dev = dev; - ht_key.ingress = ingress; + ht_key.block = block; ht_key.chain_index = chain_index; ht_key.ops = ops; return rhashtable_lookup_fast(&acl->ruleset_ht, &ht_key, @@ -237,8 +411,8 @@ __mlxsw_sp_acl_ruleset_lookup(struct mlxsw_sp_acl *acl, struct net_device *dev, } struct mlxsw_sp_acl_ruleset * -mlxsw_sp_acl_ruleset_lookup(struct mlxsw_sp *mlxsw_sp, struct net_device *dev, - bool ingress, u32 chain_index, +mlxsw_sp_acl_ruleset_lookup(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_block *block, u32 chain_index, enum mlxsw_sp_acl_profile profile) { const struct mlxsw_sp_acl_profile_ops *ops; @@ -248,45 +422,31 @@ mlxsw_sp_acl_ruleset_lookup(struct mlxsw_sp *mlxsw_sp, struct net_device *dev, ops = acl->ops->profile_ops(mlxsw_sp, profile); if (!ops) return ERR_PTR(-EINVAL); - ruleset = __mlxsw_sp_acl_ruleset_lookup(acl, dev, ingress, - chain_index, ops); + ruleset = __mlxsw_sp_acl_ruleset_lookup(acl, block, chain_index, ops); if (!ruleset) return ERR_PTR(-ENOENT); return ruleset; } struct mlxsw_sp_acl_ruleset * -mlxsw_sp_acl_ruleset_get(struct mlxsw_sp *mlxsw_sp, struct net_device *dev, - bool ingress, u32 chain_index, +mlxsw_sp_acl_ruleset_get(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_block *block, u32 chain_index, enum mlxsw_sp_acl_profile profile) { const struct mlxsw_sp_acl_profile_ops *ops; struct mlxsw_sp_acl *acl = mlxsw_sp->acl; struct mlxsw_sp_acl_ruleset *ruleset; - int err; ops = acl->ops->profile_ops(mlxsw_sp, profile); if (!ops) return ERR_PTR(-EINVAL); - ruleset = __mlxsw_sp_acl_ruleset_lookup(acl, dev, ingress, - chain_index, ops); + ruleset = __mlxsw_sp_acl_ruleset_lookup(acl, block, chain_index, ops); if (ruleset) { mlxsw_sp_acl_ruleset_ref_inc(ruleset); return ruleset; } - ruleset = mlxsw_sp_acl_ruleset_create(mlxsw_sp, ops); - if (IS_ERR(ruleset)) - return ruleset; - err = mlxsw_sp_acl_ruleset_bind(mlxsw_sp, ruleset, dev, - ingress, chain_index); - if (err) - goto err_ruleset_bind; - return ruleset; - -err_ruleset_bind: - mlxsw_sp_acl_ruleset_destroy(mlxsw_sp, ruleset); - return ERR_PTR(err); + return mlxsw_sp_acl_ruleset_create(mlxsw_sp, block, chain_index, ops); } void mlxsw_sp_acl_ruleset_put(struct mlxsw_sp *mlxsw_sp, @@ -535,6 +695,7 @@ int mlxsw_sp_acl_rule_add(struct mlxsw_sp *mlxsw_sp, goto err_rhashtable_insert; list_add_tail(&rule->list, &mlxsw_sp->acl->rules); + ruleset->ht_key.block->rule_count++; return 0; err_rhashtable_insert: @@ -548,6 +709,7 @@ void mlxsw_sp_acl_rule_del(struct mlxsw_sp *mlxsw_sp, struct mlxsw_sp_acl_ruleset *ruleset = rule->ruleset; const struct mlxsw_sp_acl_profile_ops *ops = ruleset->ht_key.ops; + ruleset->ht_key.block->rule_count--; list_del(&rule->list); rhashtable_remove_fast(&ruleset->rule_ht, &rule->ht_node, mlxsw_sp_acl_rule_ht_params); diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c index 7e8284b46968..c6e180c2be1e 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c @@ -154,10 +154,6 @@ struct mlxsw_sp_acl_tcam_group { struct list_head region_list; unsigned int region_count; struct rhashtable chunk_ht; - struct { - u16 local_port; - bool ingress; - } bound; struct mlxsw_sp_acl_tcam_group_ops *ops; const struct mlxsw_sp_acl_tcam_pattern *patterns; unsigned int patterns_count; @@ -262,35 +258,29 @@ static void mlxsw_sp_acl_tcam_group_del(struct mlxsw_sp *mlxsw_sp, static int mlxsw_sp_acl_tcam_group_bind(struct mlxsw_sp *mlxsw_sp, struct mlxsw_sp_acl_tcam_group *group, - struct net_device *dev, bool ingress) + struct mlxsw_sp_port *mlxsw_sp_port, + bool ingress) { - struct mlxsw_sp_port *mlxsw_sp_port; char ppbt_pl[MLXSW_REG_PPBT_LEN]; - if (!mlxsw_sp_port_dev_check(dev)) - return -EINVAL; - - mlxsw_sp_port = netdev_priv(dev); - group->bound.local_port = mlxsw_sp_port->local_port; - group->bound.ingress = ingress; - mlxsw_reg_ppbt_pack(ppbt_pl, - group->bound.ingress ? MLXSW_REG_PXBT_E_IACL : - MLXSW_REG_PXBT_E_EACL, - MLXSW_REG_PXBT_OP_BIND, group->bound.local_port, + mlxsw_reg_ppbt_pack(ppbt_pl, ingress ? MLXSW_REG_PXBT_E_IACL : + MLXSW_REG_PXBT_E_EACL, + MLXSW_REG_PXBT_OP_BIND, mlxsw_sp_port->local_port, group->id); return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ppbt), ppbt_pl); } static void mlxsw_sp_acl_tcam_group_unbind(struct mlxsw_sp *mlxsw_sp, - struct mlxsw_sp_acl_tcam_group *group) + struct mlxsw_sp_acl_tcam_group *group, + struct mlxsw_sp_port *mlxsw_sp_port, + bool ingress) { char ppbt_pl[MLXSW_REG_PPBT_LEN]; - mlxsw_reg_ppbt_pack(ppbt_pl, - group->bound.ingress ? MLXSW_REG_PXBT_E_IACL : - MLXSW_REG_PXBT_E_EACL, - MLXSW_REG_PXBT_OP_UNBIND, group->bound.local_port, + mlxsw_reg_ppbt_pack(ppbt_pl, ingress ? MLXSW_REG_PXBT_E_IACL : + MLXSW_REG_PXBT_E_EACL, + MLXSW_REG_PXBT_OP_UNBIND, mlxsw_sp_port->local_port, group->id); mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ppbt), ppbt_pl); } @@ -1056,21 +1046,25 @@ mlxsw_sp_acl_tcam_flower_ruleset_del(struct mlxsw_sp *mlxsw_sp, static int mlxsw_sp_acl_tcam_flower_ruleset_bind(struct mlxsw_sp *mlxsw_sp, void *ruleset_priv, - struct net_device *dev, bool ingress) + struct mlxsw_sp_port *mlxsw_sp_port, + bool ingress) { struct mlxsw_sp_acl_tcam_flower_ruleset *ruleset = ruleset_priv; return mlxsw_sp_acl_tcam_group_bind(mlxsw_sp, &ruleset->group, - dev, ingress); + mlxsw_sp_port, ingress); } static void mlxsw_sp_acl_tcam_flower_ruleset_unbind(struct mlxsw_sp *mlxsw_sp, - void *ruleset_priv) + void *ruleset_priv, + struct mlxsw_sp_port *mlxsw_sp_port, + bool ingress) { struct mlxsw_sp_acl_tcam_flower_ruleset *ruleset = ruleset_priv; - mlxsw_sp_acl_tcam_group_unbind(mlxsw_sp, &ruleset->group); + mlxsw_sp_acl_tcam_group_unbind(mlxsw_sp, &ruleset->group, + mlxsw_sp_port, ingress); } static u16 diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_flower.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_flower.c index 42e8a36b9b95..cf7b97d40d78 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_flower.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_flower.c @@ -35,6 +35,7 @@ #include <linux/kernel.h> #include <linux/errno.h> #include <linux/netdevice.h> +#include <net/net_namespace.h> #include <net/flow_dissector.h> #include <net/pkt_cls.h> #include <net/tc_act/tc_gact.h> @@ -45,7 +46,7 @@ #include "core_acl_flex_keys.h" static int mlxsw_sp_flower_parse_actions(struct mlxsw_sp *mlxsw_sp, - struct net_device *dev, bool ingress, + struct mlxsw_sp_acl_block *block, struct mlxsw_sp_acl_rule_info *rulei, struct tcf_exts *exts) { @@ -80,8 +81,7 @@ static int mlxsw_sp_flower_parse_actions(struct mlxsw_sp *mlxsw_sp, struct mlxsw_sp_acl_ruleset *ruleset; u16 group_id; - ruleset = mlxsw_sp_acl_ruleset_lookup(mlxsw_sp, dev, - ingress, + ruleset = mlxsw_sp_acl_ruleset_lookup(mlxsw_sp, block, chain_index, MLXSW_SP_ACL_PROFILE_FLOWER); if (IS_ERR(ruleset)) @@ -104,9 +104,6 @@ static int mlxsw_sp_flower_parse_actions(struct mlxsw_sp *mlxsw_sp, return err; out_dev = tcf_mirred_dev(a); - if (out_dev == dev) - out_dev = NULL; - err = mlxsw_sp_acl_rulei_act_fwd(mlxsw_sp, rulei, out_dev); if (err) @@ -265,7 +262,7 @@ static int mlxsw_sp_flower_parse_ip(struct mlxsw_sp *mlxsw_sp, } static int mlxsw_sp_flower_parse(struct mlxsw_sp *mlxsw_sp, - struct net_device *dev, bool ingress, + struct mlxsw_sp_acl_block *block, struct mlxsw_sp_acl_rule_info *rulei, struct tc_cls_flower_offload *f) { @@ -383,21 +380,19 @@ static int mlxsw_sp_flower_parse(struct mlxsw_sp *mlxsw_sp, if (err) return err; - return mlxsw_sp_flower_parse_actions(mlxsw_sp, dev, ingress, - rulei, f->exts); + return mlxsw_sp_flower_parse_actions(mlxsw_sp, block, rulei, f->exts); } -int mlxsw_sp_flower_replace(struct mlxsw_sp_port *mlxsw_sp_port, bool ingress, +int mlxsw_sp_flower_replace(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_block *block, struct tc_cls_flower_offload *f) { - struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; - struct net_device *dev = mlxsw_sp_port->dev; struct mlxsw_sp_acl_rule_info *rulei; struct mlxsw_sp_acl_ruleset *ruleset; struct mlxsw_sp_acl_rule *rule; int err; - ruleset = mlxsw_sp_acl_ruleset_get(mlxsw_sp, dev, ingress, + ruleset = mlxsw_sp_acl_ruleset_get(mlxsw_sp, block, f->common.chain_index, MLXSW_SP_ACL_PROFILE_FLOWER); if (IS_ERR(ruleset)) @@ -410,7 +405,7 @@ int mlxsw_sp_flower_replace(struct mlxsw_sp_port *mlxsw_sp_port, bool ingress, } rulei = mlxsw_sp_acl_rule_rulei(rule); - err = mlxsw_sp_flower_parse(mlxsw_sp, dev, ingress, rulei, f); + err = mlxsw_sp_flower_parse(mlxsw_sp, block, rulei, f); if (err) goto err_flower_parse; @@ -423,7 +418,6 @@ int mlxsw_sp_flower_replace(struct mlxsw_sp_port *mlxsw_sp_port, bool ingress, goto err_rule_add; mlxsw_sp_acl_ruleset_put(mlxsw_sp, ruleset); - mlxsw_sp_port->acl_rule_count++; return 0; err_rule_add: @@ -435,15 +429,15 @@ err_rule_create: return err; } -void mlxsw_sp_flower_destroy(struct mlxsw_sp_port *mlxsw_sp_port, bool ingress, +void mlxsw_sp_flower_destroy(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_block *block, struct tc_cls_flower_offload *f) { - struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; struct mlxsw_sp_acl_ruleset *ruleset; struct mlxsw_sp_acl_rule *rule; - ruleset = mlxsw_sp_acl_ruleset_get(mlxsw_sp, mlxsw_sp_port->dev, - ingress, f->common.chain_index, + ruleset = mlxsw_sp_acl_ruleset_get(mlxsw_sp, block, + f->common.chain_index, MLXSW_SP_ACL_PROFILE_FLOWER); if (IS_ERR(ruleset)) return; @@ -455,13 +449,12 @@ void mlxsw_sp_flower_destroy(struct mlxsw_sp_port *mlxsw_sp_port, bool ingress, } mlxsw_sp_acl_ruleset_put(mlxsw_sp, ruleset); - mlxsw_sp_port->acl_rule_count--; } -int mlxsw_sp_flower_stats(struct mlxsw_sp_port *mlxsw_sp_port, bool ingress, +int mlxsw_sp_flower_stats(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_block *block, struct tc_cls_flower_offload *f) { - struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; struct mlxsw_sp_acl_ruleset *ruleset; struct mlxsw_sp_acl_rule *rule; u64 packets; @@ -469,8 +462,8 @@ int mlxsw_sp_flower_stats(struct mlxsw_sp_port *mlxsw_sp_port, bool ingress, u64 bytes; int err; - ruleset = mlxsw_sp_acl_ruleset_get(mlxsw_sp, mlxsw_sp_port->dev, - ingress, f->common.chain_index, + ruleset = mlxsw_sp_acl_ruleset_get(mlxsw_sp, block, + f->common.chain_index, MLXSW_SP_ACL_PROFILE_FLOWER); if (WARN_ON(IS_ERR(ruleset))) return -EINVAL; diff --git a/include/net/pkt_cls.h b/include/net/pkt_cls.h index cc23c041a6d7..789d818c4a61 100644 --- a/include/net/pkt_cls.h +++ b/include/net/pkt_cls.h @@ -29,6 +29,7 @@ struct tcf_block_ext_info { enum tcf_block_binder_type binder_type; tcf_chain_head_change_t *chain_head_change; void *chain_head_change_priv; + u32 block_index; }; struct tcf_block_cb; @@ -38,6 +39,7 @@ bool tcf_queue_work(struct work_struct *work); struct tcf_chain *tcf_chain_get(struct tcf_block *block, u32 chain_index, bool create); void tcf_chain_put(struct tcf_chain *chain); +void tcf_block_netif_keep_dst(struct tcf_block *block); int tcf_block_get(struct tcf_block **p_block, struct tcf_proto __rcu **p_filter_chain, struct Qdisc *q, struct netlink_ext_ack *extack); @@ -48,8 +50,14 @@ void tcf_block_put(struct tcf_block *block); void tcf_block_put_ext(struct tcf_block *block, struct Qdisc *q, struct tcf_block_ext_info *ei); +static inline bool tcf_block_shared(struct tcf_block *block) +{ + return block->index; +} + static inline struct Qdisc *tcf_block_q(struct tcf_block *block) { + WARN_ON(tcf_block_shared(block)); return block->q; } diff --git a/include/net/sch_generic.h b/include/net/sch_generic.h index bd9125b0481f..cfc19d0ba2ad 100644 --- a/include/net/sch_generic.h +++ b/include/net/sch_generic.h @@ -204,6 +204,13 @@ struct Qdisc_ops { int (*dump)(struct Qdisc *, struct sk_buff *); int (*dump_stats)(struct Qdisc *, struct gnet_dump *); + void (*ingress_block_set)(struct Qdisc *sch, + u32 block_index); + void (*egress_block_set)(struct Qdisc *sch, + u32 block_index); + u32 (*ingress_block_get)(struct Qdisc *sch); + u32 (*egress_block_get)(struct Qdisc *sch); + struct module *owner; }; @@ -255,8 +262,6 @@ struct tcf_proto { /* All the rest */ u32 prio; - u32 classid; - struct Qdisc *q; void *data; const struct tcf_proto_ops *ops; struct tcf_chain *chain; @@ -275,8 +280,7 @@ typedef void tcf_chain_head_change_t(struct tcf_proto *tp_head, void *priv); struct tcf_chain { struct tcf_proto __rcu *filter_chain; - tcf_chain_head_change_t *chain_head_change; - void *chain_head_change_priv; + struct list_head filter_chain_list; struct list_head list; struct tcf_block *block; u32 index; /* chain index */ @@ -285,11 +289,33 @@ struct tcf_chain { struct tcf_block { struct list_head chain_list; + u32 index; /* block index for shared blocks */ + unsigned int refcnt; struct net *net; struct Qdisc *q; struct list_head cb_list; + struct list_head owner_list; + bool keep_dst; + unsigned int offloadcnt; /* Number of oddloaded filters */ + unsigned int nooffloaddevcnt; /* Number of devs unable to do offload */ }; +static inline void tcf_block_offload_inc(struct tcf_block *block, u32 *flags) +{ + if (*flags & TCA_CLS_FLAGS_IN_HW) + return; + *flags |= TCA_CLS_FLAGS_IN_HW; + block->offloadcnt++; +} + +static inline void tcf_block_offload_dec(struct tcf_block *block, u32 *flags) +{ + if (!(*flags & TCA_CLS_FLAGS_IN_HW)) + return; + *flags &= ~TCA_CLS_FLAGS_IN_HW; + block->offloadcnt--; +} + static inline void qdisc_cb_private_validate(const struct sk_buff *skb, int sz) { struct qdisc_skb_cb *qcb; diff --git a/include/uapi/linux/rtnetlink.h b/include/uapi/linux/rtnetlink.h index 843e29aa3cac..9b15005955fa 100644 --- a/include/uapi/linux/rtnetlink.h +++ b/include/uapi/linux/rtnetlink.h @@ -541,9 +541,19 @@ struct tcmsg { int tcm_ifindex; __u32 tcm_handle; __u32 tcm_parent; +/* tcm_block_index is used instead of tcm_parent + * in case tcm_ifindex == TCM_IFINDEX_MAGIC_BLOCK + */ +#define tcm_block_index tcm_parent __u32 tcm_info; }; +/* For manipulation of filters in shared block, tcm_ifindex is set to + * TCM_IFINDEX_MAGIC_BLOCK, and tcm_parent is aliased to tcm_block_index + * which is the block index. + */ +#define TCM_IFINDEX_MAGIC_BLOCK (0xFFFFFFFFU) + enum { TCA_UNSPEC, TCA_KIND, @@ -558,6 +568,8 @@ enum { TCA_DUMP_INVISIBLE, TCA_CHAIN, TCA_HW_OFFLOAD, + TCA_INGRESS_BLOCK, + TCA_EGRESS_BLOCK, __TCA_MAX }; diff --git a/net/sched/cls_api.c b/net/sched/cls_api.c index 6708b6953bfa..e500d11da9cd 100644 --- a/net/sched/cls_api.c +++ b/net/sched/cls_api.c @@ -24,6 +24,7 @@ #include <linux/init.h> #include <linux/kmod.h> #include <linux/slab.h> +#include <linux/idr.h> #include <net/net_namespace.h> #include <net/sock.h> #include <net/netlink.h> @@ -121,8 +122,7 @@ static inline u32 tcf_auto_prio(struct tcf_proto *tp) } static struct tcf_proto *tcf_proto_create(const char *kind, u32 protocol, - u32 prio, u32 parent, struct Qdisc *q, - struct tcf_chain *chain) + u32 prio, struct tcf_chain *chain) { struct tcf_proto *tp; int err; @@ -156,8 +156,6 @@ static struct tcf_proto *tcf_proto_create(const char *kind, u32 protocol, tp->classify = tp->ops->classify; tp->protocol = protocol; tp->prio = prio; - tp->classid = parent; - tp->q = q; tp->chain = chain; err = tp->ops->init(tp); @@ -179,6 +177,12 @@ static void tcf_proto_destroy(struct tcf_proto *tp) kfree_rcu(tp, rcu); } +struct tcf_filter_chain_list_item { + struct list_head list; + tcf_chain_head_change_t *chain_head_change; + void *chain_head_change_priv; +}; + static struct tcf_chain *tcf_chain_create(struct tcf_block *block, u32 chain_index) { @@ -187,6 +191,7 @@ static struct tcf_chain *tcf_chain_create(struct tcf_block *block, chain = kzalloc(sizeof(*chain), GFP_KERNEL); if (!chain) return NULL; + INIT_LIST_HEAD(&chain->filter_chain_list); list_add_tail(&chain->list, &block->chain_list); chain->block = block; chain->index = chain_index; @@ -194,12 +199,19 @@ static struct tcf_chain *tcf_chain_create(struct tcf_block *block, return chain; } +static void tcf_chain_head_change_item(struct tcf_filter_chain_list_item *item, + struct tcf_proto *tp_head) +{ + if (item->chain_head_change) + item->chain_head_change(tp_head, item->chain_head_change_priv); +} static void tcf_chain_head_change(struct tcf_chain *chain, struct tcf_proto *tp_head) { - if (chain->chain_head_change) - chain->chain_head_change(tp_head, - chain->chain_head_change_priv); + struct tcf_filter_chain_list_item *item; + + list_for_each_entry(item, &chain->filter_chain_list, list) + tcf_chain_head_change_item(item, tp_head); } static void tcf_chain_flush(struct tcf_chain *chain) @@ -253,47 +265,149 @@ void tcf_chain_put(struct tcf_chain *chain) } EXPORT_SYMBOL(tcf_chain_put); -static void tcf_block_offload_cmd(struct tcf_block *block, struct Qdisc *q, - struct tcf_block_ext_info *ei, - enum tc_block_command command) +static bool tcf_block_offload_in_use(struct tcf_block *block) +{ + return block->offloadcnt; +} + +static int tcf_block_offload_cmd(struct tcf_block *block, + struct net_device *dev, + struct tcf_block_ext_info *ei, + enum tc_block_command command) { - struct net_device *dev = q->dev_queue->dev; struct tc_block_offload bo = {}; - if (!dev->netdev_ops->ndo_setup_tc) - return; bo.command = command; bo.binder_type = ei->binder_type; bo.block = block; - dev->netdev_ops->ndo_setup_tc(dev, TC_SETUP_BLOCK, &bo); + return dev->netdev_ops->ndo_setup_tc(dev, TC_SETUP_BLOCK, &bo); } -static void tcf_block_offload_bind(struct tcf_block *block, struct Qdisc *q, - struct tcf_block_ext_info *ei) +static int tcf_block_offload_bind(struct tcf_block *block, struct Qdisc *q, + struct tcf_block_ext_info *ei) { - tcf_block_offload_cmd(block, q, ei, TC_BLOCK_BIND); + struct net_device *dev = q->dev_queue->dev; + int err; + + if (!dev->netdev_ops->ndo_setup_tc) + goto no_offload_dev_inc; + + /* If tc offload feature is disabled and the block we try to bind + * to already has some offloaded filters, forbid to bind. + */ + if (!tc_can_offload(dev) && tcf_block_offload_in_use(block)) + return -EOPNOTSUPP; + + err = tcf_block_offload_cmd(block, dev, ei, TC_BLOCK_BIND); + if (err == -EOPNOTSUPP) + goto no_offload_dev_inc; + return err; + +no_offload_dev_inc: + if (tcf_block_offload_in_use(block)) + return -EOPNOTSUPP; + block->nooffloaddevcnt++; + return 0; } static void tcf_block_offload_unbind(struct tcf_block *block, struct Qdisc *q, struct tcf_block_ext_info *ei) { - tcf_block_offload_cmd(block, q, ei, TC_BLOCK_UNBIND); + struct net_device *dev = q->dev_queue->dev; + int err; + + if (!dev->netdev_ops->ndo_setup_tc) + goto no_offload_dev_dec; + err = tcf_block_offload_cmd(block, dev, ei, TC_BLOCK_UNBIND); + if (err == -EOPNOTSUPP) + goto no_offload_dev_dec; + return; + +no_offload_dev_dec: + WARN_ON(block->nooffloaddevcnt-- == 0); +} + +static int +tcf_chain_head_change_cb_add(struct tcf_chain *chain, + struct tcf_block_ext_info *ei, + struct netlink_ext_ack *extack) +{ + struct tcf_filter_chain_list_item *item; + + item = kmalloc(sizeof(*item), GFP_KERNEL); + if (!item) { + NL_SET_ERR_MSG(extack, "Memory allocation for head change callback item failed"); + return -ENOMEM; + } + item->chain_head_change = ei->chain_head_change; + item->chain_head_change_priv = ei->chain_head_change_priv; + if (chain->filter_chain) + tcf_chain_head_change_item(item, chain->filter_chain); + list_add(&item->list, &chain->filter_chain_list); + return 0; } -int tcf_block_get_ext(struct tcf_block **p_block, struct Qdisc *q, - struct tcf_block_ext_info *ei, - struct netlink_ext_ack *extack) +static void +tcf_chain_head_change_cb_del(struct tcf_chain *chain, + struct tcf_block_ext_info *ei) +{ + struct tcf_filter_chain_list_item *item; + + list_for_each_entry(item, &chain->filter_chain_list, list) { + if ((!ei->chain_head_change && !ei->chain_head_change_priv) || + (item->chain_head_change == ei->chain_head_change && + item->chain_head_change_priv == ei->chain_head_change_priv)) { + tcf_chain_head_change_item(item, NULL); + list_del(&item->list); + kfree(item); + return; + } + } + WARN_ON(1); +} + +struct tcf_net { + struct idr idr; +}; + +static unsigned int tcf_net_id; + +static int tcf_block_insert(struct tcf_block *block, struct net *net, + u32 block_index, struct netlink_ext_ack *extack) +{ + struct tcf_net *tn = net_generic(net, tcf_net_id); + int err; + + err = idr_alloc_ext(&tn->idr, block, NULL, block_index, + block_index + 1, GFP_KERNEL); + if (err) + return err; + block->index = block_index; + return 0; +} + +static void tcf_block_remove(struct tcf_block *block, struct net *net) { - struct tcf_block *block = kzalloc(sizeof(*block), GFP_KERNEL); + struct tcf_net *tn = net_generic(net, tcf_net_id); + + idr_remove_ext(&tn->idr, block->index); +} + +static struct tcf_block *tcf_block_create(struct net *net, struct Qdisc *q, + struct netlink_ext_ack *extack) +{ + struct tcf_block *block; struct tcf_chain *chain; int err; + block = kzalloc(sizeof(*block), GFP_KERNEL); if (!block) { NL_SET_ERR_MSG(extack, "Memory allocation for block failed"); - return -ENOMEM; + return ERR_PTR(-ENOMEM); } INIT_LIST_HEAD(&block->chain_list); INIT_LIST_HEAD(&block->cb_list); + INIT_LIST_HEAD(&block->owner_list); /* Create chain 0 by default, it has to be always present. */ chain = tcf_chain_create(block, 0); @@ -302,17 +416,149 @@ int tcf_block_get_ext(struct tcf_block **p_block, struct Qdisc *q, err = -ENOMEM; goto err_chain_create; } - WARN_ON(!ei->chain_head_change); - chain->chain_head_change = ei->chain_head_change; - chain->chain_head_change_priv = ei->chain_head_change_priv; block->net = qdisc_net(q); + block->refcnt = 1; + block->net = net; block->q = q; - tcf_block_offload_bind(block, q, ei); - *p_block = block; - return 0; + return block; err_chain_create: kfree(block); + return ERR_PTR(err); +} + +static struct tcf_block *tcf_block_lookup(struct net *net, u32 block_index) +{ + struct tcf_net *tn = net_generic(net, tcf_net_id); + + return idr_find_ext(&tn->idr, block_index); +} + +static struct tcf_chain *tcf_block_chain_zero(struct tcf_block *block) +{ + return list_first_entry(&block->chain_list, struct tcf_chain, list); +} + +struct tcf_block_owner_item { + struct list_head list; + struct Qdisc *q; + enum tcf_block_binder_type binder_type; +}; + +static void +tcf_block_owner_netif_keep_dst(struct tcf_block *block, + struct Qdisc *q, + enum tcf_block_binder_type binder_type) +{ + if (block->keep_dst && + binder_type != TCF_BLOCK_BINDER_TYPE_CLSACT_INGRESS && + binder_type != TCF_BLOCK_BINDER_TYPE_CLSACT_EGRESS) + netif_keep_dst(qdisc_dev(q)); +} + +void tcf_block_netif_keep_dst(struct tcf_block *block) +{ + struct tcf_block_owner_item *item; + + block->keep_dst = true; + list_for_each_entry(item, &block->owner_list, list) + tcf_block_owner_netif_keep_dst(block, item->q, + item->binder_type); +} +EXPORT_SYMBOL(tcf_block_netif_keep_dst); + +static int tcf_block_owner_add(struct tcf_block *block, + struct Qdisc *q, + enum tcf_block_binder_type binder_type) +{ + struct tcf_block_owner_item *item; + + item = kmalloc(sizeof(*item), GFP_KERNEL); + if (!item) + return -ENOMEM; + item->q = q; + item->binder_type = binder_type; + list_add(&item->list, &block->owner_list); + return 0; +} + +static void tcf_block_owner_del(struct tcf_block *block, + struct Qdisc *q, + enum tcf_block_binder_type binder_type) +{ + struct tcf_block_owner_item *item; + + list_for_each_entry(item, &block->owner_list, list) { + if (item->q == q && item->binder_type == binder_type) { + list_del(&item->list); + kfree(item); + return; + } + } + WARN_ON(1); +} + +int tcf_block_get_ext(struct tcf_block **p_block, struct Qdisc *q, + struct tcf_block_ext_info *ei, + struct netlink_ext_ack *extack) +{ + struct net *net = qdisc_net(q); + struct tcf_block *block = NULL; + bool created = false; + int err; + + if (ei->block_index) { + /* block_index not 0 means the shared block is requested */ + block = tcf_block_lookup(net, ei->block_index); + if (block) + block->refcnt++; + } + + if (!block) { + block = tcf_block_create(net, q, extack); + if (IS_ERR(block)) + return PTR_ERR(block); + created = true; + if (ei->block_index) { + err = tcf_block_insert(block, net, + ei->block_index, extack); + if (err) + goto err_block_insert; + } + } + + err = tcf_block_owner_add(block, q, ei->binder_type); + if (err) + goto err_block_owner_add; + + tcf_block_owner_netif_keep_dst(block, q, ei->binder_type); + + err = tcf_chain_head_change_cb_add(tcf_block_chain_zero(block), + ei, extack); + if (err) + goto err_chain_head_change_cb_add; + + err = tcf_block_offload_bind(block, q, ei); + if (err) + goto err_block_offload_bind; + + *p_block = block; + return 0; + +err_block_offload_bind: + tcf_chain_head_change_cb_del(tcf_block_chain_zero(block), ei); +err_chain_head_change_cb_add: + tcf_block_owner_del(block, q, ei->binder_type); +err_block_owner_add: + if (created) { + if (tcf_block_shared(block)) + tcf_block_remove(block, net); +err_block_insert: + kfree(tcf_block_chain_zero(block)); + kfree(block); + } else { + block->refcnt--; + } return err; } EXPORT_SYMBOL(tcf_block_get_ext); @@ -346,26 +592,35 @@ void tcf_block_put_ext(struct tcf_block *block, struct Qdisc *q, { struct tcf_chain *chain, *tmp; - /* Hold a refcnt for all chains, so that they don't disappear - * while we are iterating. - */ if (!block) return; - list_for_each_entry(chain, &block->chain_list, list) - tcf_chain_hold(chain); + tcf_chain_head_change_cb_del(tcf_block_chain_zero(block), ei); + tcf_block_owner_del(block, q, ei->binder_type); - list_for_each_entry(chain, &block->chain_list, list) - tcf_chain_flush(chain); + if (--block->refcnt == 0) { + if (tcf_block_shared(block)) + tcf_block_remove(block, block->net); + + /* Hold a refcnt for all chains, so that they don't disappear + * while we are iterating. + */ + list_for_each_entry(chain, &block->chain_list, list) + tcf_chain_hold(chain); + + list_for_each_entry(chain, &block->chain_list, list) + tcf_chain_flush(chain); + } tcf_block_offload_unbind(block, q, ei); - /* At this point, all the chains should have refcnt >= 1. */ - list_for_each_entry_safe(chain, tmp, &block->chain_list, list) - tcf_chain_put(chain); + if (block->refcnt == 0) { + /* At this point, all the chains should have refcnt >= 1. */ + list_for_each_entry_safe(chain, tmp, &block->chain_list, list) + tcf_chain_put(chain); - /* Finally, put chain 0 and allow block to be freed. */ - chain = list_first_entry(&block->chain_list, struct tcf_chain, list); - tcf_chain_put(chain); + /* Finally, put chain 0 and allow block to be freed. */ + tcf_chain_put(tcf_block_chain_zero(block)); + } } EXPORT_SYMBOL(tcf_block_put_ext); @@ -423,9 +678,16 @@ struct tcf_block_cb *__tcf_block_cb_register(struct tcf_block *block, { struct tcf_block_cb *block_cb; + /* At this point, playback of previous block cb calls is not supported, + * so forbid to register to block which already has some offloaded + * filters present. + */ + if (tcf_block_offload_in_use(block)) + return ERR_PTR(-EOPNOTSUPP); + block_cb = kzalloc(sizeof(*block_cb), GFP_KERNEL); if (!block_cb) - return NULL; + return ERR_PTR(-ENOMEM); block_cb->cb = cb; block_cb->cb_ident = cb_ident; block_cb->cb_priv = cb_priv; @@ -441,7 +703,7 @@ int tcf_block_cb_register(struct tcf_block *block, struct tcf_block_cb *block_cb; block_cb = __tcf_block_cb_register(block, cb, cb_ident, cb_priv); - return block_cb ? 0 : -ENOMEM; + return IS_ERR(block_cb) ? PTR_ERR(block_cb) : 0; } EXPORT_SYMBOL(tcf_block_cb_register); @@ -471,6 +733,10 @@ static int tcf_block_cb_call(struct tcf_block *block, enum tc_setup_type type, int ok_count = 0; int err; + /* Make sure all netdevs sharing this block are offload-capable. */ + if (block->nooffloaddevcnt && err_stop) + return -EOPNOTSUPP; + list_for_each_entry(block_cb, &block->cb_list, list) { err = block_cb->cb(type, type_data, block_cb->cb_priv); if (err) { @@ -524,8 +790,9 @@ reclassify: #ifdef CONFIG_NET_CLS_ACT reset: if (unlikely(limit++ >= max_reclassify_loop)) { - net_notice_ratelimited("%s: reclassify loop, rule prio %u, protocol %02x\n", - tp->q->ops->id, tp->prio & 0xffff, + net_notice_ratelimited("%u: reclassify loop, rule prio %u, protocol %02x\n", + tp->chain->block->index, + tp->prio & 0xffff, ntohs(tp->protocol)); return TC_ACT_SHOT; } @@ -598,8 +865,9 @@ static struct tcf_proto *tcf_chain_tp_find(struct tcf_chain *chain, } static int tcf_fill_node(struct net *net, struct sk_buff *skb, - struct tcf_proto *tp, struct Qdisc *q, u32 parent, - void *fh, u32 portid, u32 seq, u16 flags, int event) + struct tcf_proto *tp, struct tcf_block *block, + struct Qdisc *q, u32 parent, void *fh, + u32 portid, u32 seq, u16 flags, int event) { struct tcmsg *tcm; struct nlmsghdr *nlh; @@ -612,8 +880,13 @@ static int tcf_fill_node(struct net *net, struct sk_buff *skb, tcm->tcm_family = AF_UNSPEC; tcm->tcm__pad1 = 0; tcm->tcm__pad2 = 0; - tcm->tcm_ifindex = qdisc_dev(q)->ifindex; - tcm->tcm_parent = parent; + if (q) { + tcm->tcm_ifindex = qdisc_dev(q)->ifindex; + tcm->tcm_parent = parent; + } else { + tcm->tcm_ifindex = TCM_IFINDEX_MAGIC_BLOCK; + tcm->tcm_block_index = block->index; + } tcm->tcm_info = TC_H_MAKE(tp->prio, tp->protocol); if (nla_put_string(skb, TCA_KIND, tp->ops->kind)) goto nla_put_failure; @@ -636,8 +909,8 @@ nla_put_failure: static int tfilter_notify(struct net *net, struct sk_buff *oskb, struct nlmsghdr *n, struct tcf_proto *tp, - struct Qdisc *q, u32 parent, - void *fh, int event, bool unicast) + struct tcf_block *block, struct Qdisc *q, + u32 parent, void *fh, int event, bool unicast) { struct sk_buff *skb; u32 portid = oskb ? NETLINK_CB(oskb).portid : 0; @@ -646,8 +919,8 @@ static int tfilter_notify(struct net *net, struct sk_buff *oskb, if (!skb) return -ENOBUFS; - if (tcf_fill_node(net, skb, tp, q, parent, fh, portid, n->nlmsg_seq, - n->nlmsg_flags, event) <= 0) { + if (tcf_fill_node(net, skb, tp, block, q, parent, fh, portid, + n->nlmsg_seq, n->nlmsg_flags, event) <= 0) { kfree_skb(skb); return -EINVAL; } @@ -661,8 +934,8 @@ static int tfilter_notify(struct net *net, struct sk_buff *oskb, static int tfilter_del_notify(struct net *net, struct sk_buff *oskb, struct nlmsghdr *n, struct tcf_proto *tp, - struct Qdisc *q, u32 parent, - void *fh, bool unicast, bool *last) + struct tcf_block *block, struct Qdisc *q, + u32 parent, void *fh, bool unicast, bool *last) { struct sk_buff *skb; u32 portid = oskb ? NETLINK_CB(oskb).portid : 0; @@ -672,8 +945,8 @@ static int tfilter_del_notify(struct net *net, struct sk_buff *oskb, if (!skb) return -ENOBUFS; - if (tcf_fill_node(net, skb, tp, q, parent, fh, portid, n->nlmsg_seq, - n->nlmsg_flags, RTM_DELTFILTER) <= 0) { + if (tcf_fill_node(net, skb, tp, block, q, parent, fh, portid, + n->nlmsg_seq, n->nlmsg_flags, RTM_DELTFILTER) <= 0) { kfree_skb(skb); return -EINVAL; } @@ -692,15 +965,16 @@ static int tfilter_del_notify(struct net *net, struct sk_buff *oskb, } static void tfilter_notify_chain(struct net *net, struct sk_buff *oskb, - struct Qdisc *q, u32 parent, - struct nlmsghdr *n, + struct tcf_block *block, struct Qdisc *q, + u32 parent, struct nlmsghdr *n, struct tcf_chain *chain, int event) { struct tcf_proto *tp; for (tp = rtnl_dereference(chain->filter_chain); tp; tp = rtnl_dereference(tp->next)) - tfilter_notify(net, oskb, n, tp, q, parent, 0, event, false); + tfilter_notify(net, oskb, n, tp, block, + q, parent, 0, event, false); } /* Add/change/delete/get a filter node */ @@ -716,13 +990,11 @@ static int tc_ctl_tfilter(struct sk_buff *skb, struct nlmsghdr *n, bool prio_allocate; u32 parent; u32 chain_index; - struct net_device *dev; - struct Qdisc *q; + struct Qdisc *q = NULL; struct tcf_chain_info chain_info; struct tcf_chain *chain = NULL; struct tcf_block *block; struct tcf_proto *tp; - const struct Qdisc_class_ops *cops; unsigned long cl; void *fh; int err; @@ -769,41 +1041,58 @@ replay: /* Find head of filter chain. */ - /* Find link */ - dev = __dev_get_by_index(net, t->tcm_ifindex); - if (dev == NULL) - return -ENODEV; - - /* Find qdisc */ - if (!parent) { - q = dev->qdisc; - parent = q->handle; + if (t->tcm_ifindex == TCM_IFINDEX_MAGIC_BLOCK) { + block = tcf_block_lookup(net, t->tcm_block_index); + if (!block) { + NL_SET_ERR_MSG(extack, "Block of given index was not found"); + err = -EINVAL; + goto errout; + } } else { - q = qdisc_lookup(dev, TC_H_MAJ(t->tcm_parent)); - if (q == NULL) - return -EINVAL; - } + const struct Qdisc_class_ops *cops; + struct net_device *dev; - /* Is it classful? */ - cops = q->ops->cl_ops; - if (!cops) - return -EINVAL; + /* Find link */ + dev = __dev_get_by_index(net, t->tcm_ifindex); + if (!dev) + return -ENODEV; - if (!cops->tcf_block) - return -EOPNOTSUPP; + /* Find qdisc */ + if (!parent) { + q = dev->qdisc; + parent = q->handle; + } else { + q = qdisc_lookup(dev, TC_H_MAJ(t->tcm_parent)); + if (!q) + return -EINVAL; + } - /* Do we search for filter, attached to class? */ - if (TC_H_MIN(parent)) { - cl = cops->find(q, parent); - if (cl == 0) - return -ENOENT; - } + /* Is it classful? */ + cops = q->ops->cl_ops; + if (!cops) + return -EINVAL; - /* And the last stroke */ - block = cops->tcf_block(q, cl, extack); - if (!block) { - err = -EINVAL; - goto errout; + if (!cops->tcf_block) + return -EOPNOTSUPP; + + /* Do we search for filter, attached to class? */ + if (TC_H_MIN(parent)) { + cl = cops->find(q, parent); + if (cl == 0) + return -ENOENT; + } + + /* And the last stroke */ + block = cops->tcf_block(q, cl, extack); + if (!block) { + err = -EINVAL; + goto errout; + } + if (tcf_block_shared(block)) { + NL_SET_ERR_MSG(extack, "This filter block is shared. Please use the block index to manipulate the filters"); + err = -EOPNOTSUPP; + goto errout; + } } chain_index = tca[TCA_CHAIN] ? nla_get_u32(tca[TCA_CHAIN]) : 0; @@ -819,7 +1108,7 @@ replay: } if (n->nlmsg_type == RTM_DELTFILTER && prio == 0) { - tfilter_notify_chain(net, skb, q, parent, n, + tfilter_notify_chain(net, skb, block, q, parent, n, chain, RTM_DELTFILTER); tcf_chain_flush(chain); err = 0; @@ -851,7 +1140,7 @@ replay: prio = tcf_auto_prio(tcf_chain_tp_prev(&chain_info)); tp = tcf_proto_create(nla_data(tca[TCA_KIND]), - protocol, prio, parent, q, chain); + protocol, prio, chain); if (IS_ERR(tp)) { err = PTR_ERR(tp); goto errout; @@ -867,7 +1156,7 @@ replay: if (!fh) { if (n->nlmsg_type == RTM_DELTFILTER && t->tcm_handle == 0) { tcf_chain_tp_remove(chain, &chain_info, tp); - tfilter_notify(net, skb, n, tp, q, parent, fh, + tfilter_notify(net, skb, n, tp, block, q, parent, fh, RTM_DELTFILTER, false); tcf_proto_destroy(tp); err = 0; @@ -892,8 +1181,8 @@ replay: } break; case RTM_DELTFILTER: - err = tfilter_del_notify(net, skb, n, tp, q, parent, - fh, false, &last); + err = tfilter_del_notify(net, skb, n, tp, block, + q, parent, fh, false, &last); if (err) goto errout; if (last) { @@ -902,8 +1191,8 @@ replay: } goto errout; case RTM_GETTFILTER: - err = tfilter_notify(net, skb, n, tp, q, parent, fh, - RTM_NEWTFILTER, true); + err = tfilter_notify(net, skb, n, tp, block, q, parent, + fh, RTM_NEWTFILTER, true); goto errout; default: err = -EINVAL; @@ -916,7 +1205,7 @@ replay: if (err == 0) { if (tp_created) tcf_chain_tp_insert(chain, &chain_info, tp); - tfilter_notify(net, skb, n, tp, q, parent, fh, + tfilter_notify(net, skb, n, tp, block, q, parent, fh, RTM_NEWTFILTER, false); } else { if (tp_created) @@ -936,6 +1225,7 @@ struct tcf_dump_args { struct tcf_walker w; struct sk_buff *skb; struct netlink_callback *cb; + struct tcf_block *block; struct Qdisc *q; u32 parent; }; @@ -945,7 +1235,7 @@ static int tcf_node_dump(struct tcf_proto *tp, void *n, struct tcf_walker *arg) struct tcf_dump_args *a = (void *)arg; struct net *net = sock_net(a->skb->sk); - return tcf_fill_node(net, a->skb, tp, a->q, a->parent, + return tcf_fill_node(net, a->skb, tp, a->block, a->q, a->parent, n, NETLINK_CB(a->cb->skb).portid, a->cb->nlh->nlmsg_seq, NLM_F_MULTI, RTM_NEWTFILTER); @@ -956,6 +1246,7 @@ static bool tcf_chain_dump(struct tcf_chain *chain, struct Qdisc *q, u32 parent, long index_start, long *p_index) { struct net *net = sock_net(skb->sk); + struct tcf_block *block = chain->block; struct tcmsg *tcm = nlmsg_data(cb->nlh); struct tcf_dump_args arg; struct tcf_proto *tp; @@ -974,7 +1265,7 @@ static bool tcf_chain_dump(struct tcf_chain *chain, struct Qdisc *q, u32 parent, memset(&cb->args[1], 0, sizeof(cb->args) - sizeof(cb->args[0])); if (cb->args[1] == 0) { - if (tcf_fill_node(net, skb, tp, q, parent, 0, + if (tcf_fill_node(net, skb, tp, block, q, parent, 0, NETLINK_CB(cb->skb).portid, cb->nlh->nlmsg_seq, NLM_F_MULTI, RTM_NEWTFILTER) <= 0) @@ -987,6 +1278,7 @@ static bool tcf_chain_dump(struct tcf_chain *chain, struct Qdisc *q, u32 parent, arg.w.fn = tcf_node_dump; arg.skb = skb; arg.cb = cb; + arg.block = block; arg.q = q; arg.parent = parent; arg.w.stop = 0; @@ -1005,13 +1297,10 @@ static int tc_dump_tfilter(struct sk_buff *skb, struct netlink_callback *cb) { struct net *net = sock_net(skb->sk); struct nlattr *tca[TCA_MAX + 1]; - struct net_device *dev; - struct Qdisc *q; + struct Qdisc *q = NULL; struct tcf_block *block; struct tcf_chain *chain; struct tcmsg *tcm = nlmsg_data(cb->nlh); - unsigned long cl = 0; - const struct Qdisc_class_ops *cops; long index_start; long index; u32 parent; @@ -1024,32 +1313,44 @@ static int tc_dump_tfilter(struct sk_buff *skb, struct netlink_callback *cb) if (err) return err; - dev = __dev_get_by_index(net, tcm->tcm_ifindex); - if (!dev) - return skb->len; - - parent = tcm->tcm_parent; - if (!parent) { - q = dev->qdisc; - parent = q->handle; + if (tcm->tcm_ifindex == TCM_IFINDEX_MAGIC_BLOCK) { + block = tcf_block_lookup(net, tcm->tcm_block_index); + if (!block) + goto out; } else { - q = qdisc_lookup(dev, TC_H_MAJ(tcm->tcm_parent)); - } - if (!q) - goto out; - cops = q->ops->cl_ops; - if (!cops) - goto out; - if (!cops->tcf_block) - goto out; - if (TC_H_MIN(tcm->tcm_parent)) { - cl = cops->find(q, tcm->tcm_parent); - if (cl == 0) + const struct Qdisc_class_ops *cops; + struct net_device *dev; + unsigned long cl = 0; + + dev = __dev_get_by_index(net, tcm->tcm_ifindex); + if (!dev) + return skb->len; + + parent = tcm->tcm_parent; + if (!parent) { + q = dev->qdisc; + parent = q->handle; + } else { + q = qdisc_lookup(dev, TC_H_MAJ(tcm->tcm_parent)); + } + if (!q) + goto out; + cops = q->ops->cl_ops; + if (!cops) + goto out; + if (!cops->tcf_block) + goto out; + if (TC_H_MIN(tcm->tcm_parent)) { + cl = cops->find(q, tcm->tcm_parent); + if (cl == 0) + goto out; + } + block = cops->tcf_block(q, cl, NULL); + if (!block) goto out; + if (tcf_block_shared(block)) + q = NULL; } - block = cops->tcf_block(q, cl, NULL); - if (!block) - goto out; index_start = cb->args[0]; index = 0; @@ -1252,18 +1553,50 @@ int tc_setup_cb_call(struct tcf_block *block, struct tcf_exts *exts, } EXPORT_SYMBOL(tc_setup_cb_call); +static __net_init int tcf_net_init(struct net *net) +{ + struct tcf_net *tn = net_generic(net, tcf_net_id); + + idr_init(&tn->idr); + return 0; +} + +static void __net_exit tcf_net_exit(struct net *net) +{ + struct tcf_net *tn = net_generic(net, tcf_net_id); + + idr_destroy(&tn->idr); +} + +static struct pernet_operations tcf_net_ops = { + .init = tcf_net_init, + .exit = tcf_net_exit, + .id = &tcf_net_id, + .size = sizeof(struct tcf_net), +}; + static int __init tc_filter_init(void) { + int err; + tc_filter_wq = alloc_ordered_workqueue("tc_filter_workqueue", 0); if (!tc_filter_wq) return -ENOMEM; + err = register_pernet_subsys(&tcf_net_ops); + if (err) + goto err_register_pernet_subsys; + rtnl_register(PF_UNSPEC, RTM_NEWTFILTER, tc_ctl_tfilter, NULL, 0); rtnl_register(PF_UNSPEC, RTM_DELTFILTER, tc_ctl_tfilter, NULL, 0); rtnl_register(PF_UNSPEC, RTM_GETTFILTER, tc_ctl_tfilter, tc_dump_tfilter, 0); return 0; + +err_register_pernet_subsys: + destroy_workqueue(tc_filter_wq); + return err; } subsys_initcall(tc_filter_init); diff --git a/net/sched/cls_bpf.c b/net/sched/cls_bpf.c index 8d78e7f4ecc3..cf72aefcf98d 100644 --- a/net/sched/cls_bpf.c +++ b/net/sched/cls_bpf.c @@ -167,13 +167,16 @@ static int cls_bpf_offload_cmd(struct tcf_proto *tp, struct cls_bpf_prog *prog, cls_bpf.exts_integrated = obj->exts_integrated; cls_bpf.gen_flags = obj->gen_flags; + if (oldprog) + tcf_block_offload_dec(block, &oldprog->gen_flags); + err = tc_setup_cb_call(block, NULL, TC_SETUP_CLSBPF, &cls_bpf, skip_sw); if (prog) { if (err < 0) { cls_bpf_offload_cmd(tp, oldprog, prog); return err; } else if (err > 0) { - prog->gen_flags |= TCA_CLS_FLAGS_IN_HW; + tcf_block_offload_inc(block, &prog->gen_flags); } } @@ -392,8 +395,8 @@ static int cls_bpf_prog_from_efd(struct nlattr **tb, struct cls_bpf_prog *prog, prog->bpf_name = name; prog->filter = fp; - if (fp->dst_needed && !(tp->q->flags & TCQ_F_INGRESS)) - netif_keep_dst(qdisc_dev(tp->q)); + if (fp->dst_needed) + tcf_block_netif_keep_dst(tp->chain->block); return 0; } diff --git a/net/sched/cls_flow.c b/net/sched/cls_flow.c index 25c2a888e1f0..28cd6fb52c16 100644 --- a/net/sched/cls_flow.c +++ b/net/sched/cls_flow.c @@ -526,7 +526,7 @@ static int flow_change(struct net *net, struct sk_buff *in_skb, timer_setup(&fnew->perturb_timer, flow_perturbation, TIMER_DEFERRABLE); - netif_keep_dst(qdisc_dev(tp->q)); + tcf_block_netif_keep_dst(tp->chain->block); if (tb[TCA_FLOW_KEYS]) { fnew->keymask = keymask; diff --git a/net/sched/cls_flower.c b/net/sched/cls_flower.c index 6132a7317efa..f61df19b1026 100644 --- a/net/sched/cls_flower.c +++ b/net/sched/cls_flower.c @@ -229,6 +229,7 @@ static void fl_hw_destroy_filter(struct tcf_proto *tp, struct cls_fl_filter *f) tc_setup_cb_call(block, &f->exts, TC_SETUP_CLSFLOWER, &cls_flower, false); + tcf_block_offload_dec(block, &f->flags); } static int fl_hw_replace_filter(struct tcf_proto *tp, @@ -256,7 +257,7 @@ static int fl_hw_replace_filter(struct tcf_proto *tp, fl_hw_destroy_filter(tp, f); return err; } else if (err > 0) { - f->flags |= TCA_CLS_FLAGS_IN_HW; + tcf_block_offload_inc(block, &f->flags); } if (skip_sw && !(f->flags & TCA_CLS_FLAGS_IN_HW)) diff --git a/net/sched/cls_matchall.c b/net/sched/cls_matchall.c index 66d4e0099158..d0e57c86636f 100644 --- a/net/sched/cls_matchall.c +++ b/net/sched/cls_matchall.c @@ -81,6 +81,7 @@ static void mall_destroy_hw_filter(struct tcf_proto *tp, cls_mall.cookie = cookie; tc_setup_cb_call(block, NULL, TC_SETUP_CLSMATCHALL, &cls_mall, false); + tcf_block_offload_dec(block, &head->flags); } static int mall_replace_hw_filter(struct tcf_proto *tp, @@ -103,7 +104,7 @@ static int mall_replace_hw_filter(struct tcf_proto *tp, mall_destroy_hw_filter(tp, head, cookie); return err; } else if (err > 0) { - head->flags |= TCA_CLS_FLAGS_IN_HW; + tcf_block_offload_inc(block, &head->flags); } if (skip_sw && !(head->flags & TCA_CLS_FLAGS_IN_HW)) diff --git a/net/sched/cls_route.c b/net/sched/cls_route.c index ac9a5b8825b9..a1f2b1b7c014 100644 --- a/net/sched/cls_route.c +++ b/net/sched/cls_route.c @@ -527,7 +527,7 @@ static int route4_change(struct net *net, struct sk_buff *in_skb, if (f->handle < f1->handle) break; - netif_keep_dst(qdisc_dev(tp->q)); + tcf_block_netif_keep_dst(tp->chain->block); rcu_assign_pointer(f->next, f1); rcu_assign_pointer(*fp, f); diff --git a/net/sched/cls_u32.c b/net/sched/cls_u32.c index 507859cdd1cb..020d328d0afd 100644 --- a/net/sched/cls_u32.c +++ b/net/sched/cls_u32.c @@ -529,16 +529,17 @@ static int u32_replace_hw_hnode(struct tcf_proto *tp, struct tc_u_hnode *h, return 0; } -static void u32_remove_hw_knode(struct tcf_proto *tp, u32 handle) +static void u32_remove_hw_knode(struct tcf_proto *tp, struct tc_u_knode *n) { struct tcf_block *block = tp->chain->block; struct tc_cls_u32_offload cls_u32 = {}; tc_cls_common_offload_init(&cls_u32.common, tp); cls_u32.command = TC_CLSU32_DELETE_KNODE; - cls_u32.knode.handle = handle; + cls_u32.knode.handle = n->handle; tc_setup_cb_call(block, NULL, TC_SETUP_CLSU32, &cls_u32, false); + tcf_block_offload_dec(block, &n->flags); } static int u32_replace_hw_knode(struct tcf_proto *tp, struct tc_u_knode *n, @@ -567,10 +568,10 @@ static int u32_replace_hw_knode(struct tcf_proto *tp, struct tc_u_knode *n, err = tc_setup_cb_call(block, NULL, TC_SETUP_CLSU32, &cls_u32, skip_sw); if (err < 0) { - u32_remove_hw_knode(tp, n->handle); + u32_remove_hw_knode(tp, n); return err; } else if (err > 0) { - n->flags |= TCA_CLS_FLAGS_IN_HW; + tcf_block_offload_inc(block, &n->flags); } if (skip_sw && !(n->flags & TCA_CLS_FLAGS_IN_HW)) @@ -589,7 +590,7 @@ static void u32_clear_hnode(struct tcf_proto *tp, struct tc_u_hnode *ht) RCU_INIT_POINTER(ht->ht[h], rtnl_dereference(n->next)); tcf_unbind_filter(tp, &n->res); - u32_remove_hw_knode(tp, n->handle); + u32_remove_hw_knode(tp, n); idr_remove_ext(&ht->handle_idr, n->handle); if (tcf_exts_get_net(&n->exts)) call_rcu(&n->rcu, u32_delete_key_freepf_rcu); @@ -682,7 +683,7 @@ static int u32_delete(struct tcf_proto *tp, void *arg, bool *last) goto out; if (TC_U32_KEY(ht->handle)) { - u32_remove_hw_knode(tp, ht->handle); + u32_remove_hw_knode(tp, (struct tc_u_knode *)ht); ret = u32_delete_key(tp, (struct tc_u_knode *)ht); goto out; } diff --git a/net/sched/sch_api.c b/net/sched/sch_api.c index 7dffa9dce28b..d512f49ee83c 100644 --- a/net/sched/sch_api.c +++ b/net/sched/sch_api.c @@ -791,6 +791,7 @@ static int tc_fill_qdisc(struct sk_buff *skb, struct Qdisc *q, u32 clid, unsigned char *b = skb_tail_pointer(skb); struct gnet_dump d; struct qdisc_size_table *stab; + u32 block_index; __u32 qlen; cond_resched(); @@ -807,6 +808,18 @@ static int tc_fill_qdisc(struct sk_buff *skb, struct Qdisc *q, u32 clid, tcm->tcm_info = refcount_read(&q->refcnt); if (nla_put_string(skb, TCA_KIND, q->ops->id)) goto nla_put_failure; + if (q->ops->ingress_block_get) { + block_index = q->ops->ingress_block_get(q); + if (block_index && + nla_put_u32(skb, TCA_INGRESS_BLOCK, block_index)) + goto nla_put_failure; + } + if (q->ops->egress_block_get) { + block_index = q->ops->egress_block_get(q); + if (block_index && + nla_put_u32(skb, TCA_EGRESS_BLOCK, block_index)) + goto nla_put_failure; + } if (q->ops->dump && q->ops->dump(q, skb) < 0) goto nla_put_failure; if (nla_put_u8(skb, TCA_HW_OFFLOAD, !!(q->flags & TCQ_F_OFFLOADED))) @@ -994,6 +1007,40 @@ skip: return err; } +static int qdisc_block_indexes_set(struct Qdisc *sch, struct nlattr **tca, + struct netlink_ext_ack *extack) +{ + u32 block_index; + + if (tca[TCA_INGRESS_BLOCK]) { + block_index = nla_get_u32(tca[TCA_INGRESS_BLOCK]); + + if (!block_index) { + NL_SET_ERR_MSG(extack, "Ingress block index cannot be 0"); + return -EINVAL; + } + if (!sch->ops->ingress_block_set) { + NL_SET_ERR_MSG(extack, "Ingress block sharing is not supported"); + return -EOPNOTSUPP; + } + sch->ops->ingress_block_set(sch, block_index); + } + if (tca[TCA_EGRESS_BLOCK]) { + block_index = nla_get_u32(tca[TCA_EGRESS_BLOCK]); + + if (!block_index) { + NL_SET_ERR_MSG(extack, "Egress block index cannot be 0"); + return -EINVAL; + } + if (!sch->ops->egress_block_set) { + NL_SET_ERR_MSG(extack, "Egress block sharing is not supported"); + return -EOPNOTSUPP; + } + sch->ops->egress_block_set(sch, block_index); + } + return 0; +} + /* lockdep annotation is needed for ingress; egress gets it only for name */ static struct lock_class_key qdisc_tx_lock; static struct lock_class_key qdisc_rx_lock; @@ -1088,6 +1135,10 @@ static struct Qdisc *qdisc_create(struct net_device *dev, netdev_info(dev, "Caught tx_queue_len zero misconfig\n"); } + err = qdisc_block_indexes_set(sch, tca, extack); + if (err) + goto err_out3; + if (ops->init) { err = ops->init(sch, tca[TCA_OPTIONS], extack); if (err != 0) @@ -1169,6 +1220,10 @@ static int qdisc_change(struct Qdisc *sch, struct nlattr **tca, NL_SET_ERR_MSG(extack, "Change operation not supported by specified qdisc"); return -EINVAL; } + if (tca[TCA_INGRESS_BLOCK] || tca[TCA_EGRESS_BLOCK]) { + NL_SET_ERR_MSG(extack, "Change of blocks is not supported"); + return -EOPNOTSUPP; + } err = sch->ops->change(sch, tca[TCA_OPTIONS], extack); if (err) return err; @@ -1894,6 +1949,11 @@ static int tc_ctl_tclass(struct sk_buff *skb, struct nlmsghdr *n, } } + if (tca[TCA_INGRESS_BLOCK] || tca[TCA_EGRESS_BLOCK]) { + NL_SET_ERR_MSG(extack, "Shared blocks are not supported for classes"); + return -EOPNOTSUPP; + } + new_cl = cl; err = -EOPNOTSUPP; if (cops->change) diff --git a/net/sched/sch_ingress.c b/net/sched/sch_ingress.c index 3372dd5e984d..ce3f55259d0d 100644 --- a/net/sched/sch_ingress.c +++ b/net/sched/sch_ingress.c @@ -61,6 +61,20 @@ static void clsact_chain_head_change(struct tcf_proto *tp_head, void *priv) struct mini_Qdisc_pair *miniqp = priv; mini_qdisc_pair_swap(miniqp, tp_head); +}; + +static void ingress_ingress_block_set(struct Qdisc *sch, u32 block_index) +{ + struct ingress_sched_data *q = qdisc_priv(sch); + + q->block_info.block_index = block_index; +} + +static u32 ingress_ingress_block_get(struct Qdisc *sch) +{ + struct ingress_sched_data *q = qdisc_priv(sch); + + return q->block_info.block_index; } static int ingress_init(struct Qdisc *sch, struct nlattr *opt, @@ -113,14 +127,16 @@ static const struct Qdisc_class_ops ingress_class_ops = { }; static struct Qdisc_ops ingress_qdisc_ops __read_mostly = { - .cl_ops = &ingress_class_ops, - .id = "ingress", - .priv_size = sizeof(struct ingress_sched_data), - .static_flags = TCQ_F_CPUSTATS, - .init = ingress_init, - .destroy = ingress_destroy, - .dump = ingress_dump, - .owner = THIS_MODULE, + .cl_ops = &ingress_class_ops, + .id = "ingress", + .priv_size = sizeof(struct ingress_sched_data), + .static_flags = TCQ_F_CPUSTATS, + .init = ingress_init, + .destroy = ingress_destroy, + .dump = ingress_dump, + .ingress_block_set = ingress_ingress_block_set, + .ingress_block_get = ingress_ingress_block_get, + .owner = THIS_MODULE, }; struct clsact_sched_data { @@ -164,6 +180,34 @@ static struct tcf_block *clsact_tcf_block(struct Qdisc *sch, unsigned long cl, } } +static void clsact_ingress_block_set(struct Qdisc *sch, u32 block_index) +{ + struct clsact_sched_data *q = qdisc_priv(sch); + + q->ingress_block_info.block_index = block_index; +} + +static void clsact_egress_block_set(struct Qdisc *sch, u32 block_index) +{ + struct clsact_sched_data *q = qdisc_priv(sch); + + q->egress_block_info.block_index = block_index; +} + +static u32 clsact_ingress_block_get(struct Qdisc *sch) +{ + struct clsact_sched_data *q = qdisc_priv(sch); + + return q->ingress_block_info.block_index; +} + +static u32 clsact_egress_block_get(struct Qdisc *sch) +{ + struct clsact_sched_data *q = qdisc_priv(sch); + + return q->egress_block_info.block_index; +} + static int clsact_init(struct Qdisc *sch, struct nlattr *opt, struct netlink_ext_ack *extack) { @@ -215,14 +259,18 @@ static const struct Qdisc_class_ops clsact_class_ops = { }; static struct Qdisc_ops clsact_qdisc_ops __read_mostly = { - .cl_ops = &clsact_class_ops, - .id = "clsact", - .priv_size = sizeof(struct clsact_sched_data), - .static_flags = TCQ_F_CPUSTATS, - .init = clsact_init, - .destroy = clsact_destroy, - .dump = ingress_dump, - .owner = THIS_MODULE, + .cl_ops = &clsact_class_ops, + .id = "clsact", + .priv_size = sizeof(struct clsact_sched_data), + .static_flags = TCQ_F_CPUSTATS, + .init = clsact_init, + .destroy = clsact_destroy, + .dump = ingress_dump, + .ingress_block_set = clsact_ingress_block_set, + .egress_block_set = clsact_egress_block_set, + .ingress_block_get = clsact_ingress_block_get, + .egress_block_get = clsact_egress_block_get, + .owner = THIS_MODULE, }; static int __init ingress_module_init(void) |